diff --git "a/checkpoint-42659/trainer_state.json" "b/checkpoint-42659/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-42659/trainer_state.json" @@ -0,0 +1,298646 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999941396062988, + "eval_steps": 500, + "global_step": 42659, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 2.3441574804995398e-05, + "grad_norm": 0.9414980411529541, + "learning_rate": 2e-05, + "loss": 1.6415, + "step": 1 + }, + { + "epoch": 4.6883149609990796e-05, + "grad_norm": 1.529971957206726, + "learning_rate": 4e-05, + "loss": 0.6305, + "step": 2 + }, + { + "epoch": 7.03247244149862e-05, + "grad_norm": 8.160544395446777, + "learning_rate": 6e-05, + "loss": 2.757, + "step": 3 + }, + { + "epoch": 9.376629921998159e-05, + "grad_norm": 6.365822792053223, + "learning_rate": 8e-05, + "loss": 2.2921, + "step": 4 + }, + { + "epoch": 0.000117207874024977, + "grad_norm": 6.891416549682617, + "learning_rate": 0.0001, + "loss": 2.1023, + "step": 5 + }, + { + "epoch": 0.0001406494488299724, + "grad_norm": 3.6934502124786377, + "learning_rate": 0.00012, + "loss": 1.5827, + "step": 6 + }, + { + "epoch": 0.0001640910236349678, + "grad_norm": 4.106067657470703, + "learning_rate": 0.00014, + "loss": 1.4701, + "step": 7 + }, + { + "epoch": 0.00018753259843996319, + "grad_norm": 4.995994567871094, + "learning_rate": 0.00016, + "loss": 1.6827, + "step": 8 + }, + { + "epoch": 0.00021097417324495858, + "grad_norm": 4.693511962890625, + "learning_rate": 0.00018, + "loss": 1.3214, + "step": 9 + }, + { + "epoch": 0.000234415748049954, + "grad_norm": 0.7005836963653564, + "learning_rate": 0.0002, + "loss": 0.1531, + "step": 10 + }, + { + "epoch": 0.0002578573228549494, + "grad_norm": 3.0225117206573486, + "learning_rate": 0.0001999999997286986, + "loss": 0.8463, + "step": 11 + }, + { + "epoch": 0.0002812988976599448, + "grad_norm": 4.039967060089111, + "learning_rate": 0.0001999999989147944, + "loss": 1.0663, + "step": 12 + }, + { + "epoch": 0.0003047404724649402, + "grad_norm": 2.7937915325164795, + "learning_rate": 0.00019999999755828735, + "loss": 0.4886, + "step": 13 + }, + { + "epoch": 0.0003281820472699356, + "grad_norm": 0.9701707363128662, + "learning_rate": 0.00019999999565917755, + "loss": 0.2213, + "step": 14 + }, + { + "epoch": 0.000351623622074931, + "grad_norm": 5.027869701385498, + "learning_rate": 0.00019999999321746495, + "loss": 1.5261, + "step": 15 + }, + { + "epoch": 0.00037506519687992637, + "grad_norm": 5.671248912811279, + "learning_rate": 0.00019999999023314954, + "loss": 1.3501, + "step": 16 + }, + { + "epoch": 0.00039850677168492177, + "grad_norm": 2.6153297424316406, + "learning_rate": 0.00019999998670623142, + "loss": 0.5921, + "step": 17 + }, + { + "epoch": 0.00042194834648991717, + "grad_norm": 2.784578800201416, + "learning_rate": 0.00019999998263671055, + "loss": 0.8164, + "step": 18 + }, + { + "epoch": 0.00044538992129491257, + "grad_norm": 2.165065050125122, + "learning_rate": 0.00019999997802458693, + "loss": 0.5544, + "step": 19 + }, + { + "epoch": 0.000468831496099908, + "grad_norm": 3.3359906673431396, + "learning_rate": 0.00019999997286986062, + "loss": 1.2165, + "step": 20 + }, + { + "epoch": 0.0004922730709049034, + "grad_norm": 1.6420247554779053, + "learning_rate": 0.00019999996717253167, + "loss": 0.4833, + "step": 21 + }, + { + "epoch": 0.0005157146457098988, + "grad_norm": 2.8957834243774414, + "learning_rate": 0.0001999999609326001, + "loss": 0.6516, + "step": 22 + }, + { + "epoch": 0.0005391562205148942, + "grad_norm": 0.9769586324691772, + "learning_rate": 0.0001999999541500659, + "loss": 0.353, + "step": 23 + }, + { + "epoch": 0.0005625977953198896, + "grad_norm": 3.4573423862457275, + "learning_rate": 0.00019999994682492916, + "loss": 0.9843, + "step": 24 + }, + { + "epoch": 0.000586039370124885, + "grad_norm": 3.3196773529052734, + "learning_rate": 0.00019999993895718988, + "loss": 0.9977, + "step": 25 + }, + { + "epoch": 0.0006094809449298804, + "grad_norm": 2.755599021911621, + "learning_rate": 0.00019999993054684812, + "loss": 0.5983, + "step": 26 + }, + { + "epoch": 0.0006329225197348758, + "grad_norm": 4.189762592315674, + "learning_rate": 0.00019999992159390394, + "loss": 0.8756, + "step": 27 + }, + { + "epoch": 0.0006563640945398711, + "grad_norm": 4.53598165512085, + "learning_rate": 0.00019999991209835734, + "loss": 1.2418, + "step": 28 + }, + { + "epoch": 0.0006798056693448665, + "grad_norm": 4.247065544128418, + "learning_rate": 0.00019999990206020845, + "loss": 0.958, + "step": 29 + }, + { + "epoch": 0.000703247244149862, + "grad_norm": 3.923525094985962, + "learning_rate": 0.00019999989147945722, + "loss": 0.9473, + "step": 30 + }, + { + "epoch": 0.0007266888189548573, + "grad_norm": 1.2849305868148804, + "learning_rate": 0.00019999988035610382, + "loss": 0.3842, + "step": 31 + }, + { + "epoch": 0.0007501303937598527, + "grad_norm": 6.339925765991211, + "learning_rate": 0.00019999986869014824, + "loss": 1.6681, + "step": 32 + }, + { + "epoch": 0.0007735719685648481, + "grad_norm": 5.609142303466797, + "learning_rate": 0.00019999985648159056, + "loss": 1.0795, + "step": 33 + }, + { + "epoch": 0.0007970135433698435, + "grad_norm": 2.2937114238739014, + "learning_rate": 0.00019999984373043087, + "loss": 0.5032, + "step": 34 + }, + { + "epoch": 0.0008204551181748389, + "grad_norm": 4.482852935791016, + "learning_rate": 0.00019999983043666916, + "loss": 0.8596, + "step": 35 + }, + { + "epoch": 0.0008438966929798343, + "grad_norm": 2.2566919326782227, + "learning_rate": 0.0001999998166003056, + "loss": 0.6529, + "step": 36 + }, + { + "epoch": 0.0008673382677848297, + "grad_norm": 0.9397289752960205, + "learning_rate": 0.0001999998022213402, + "loss": 0.2614, + "step": 37 + }, + { + "epoch": 0.0008907798425898251, + "grad_norm": 2.5312840938568115, + "learning_rate": 0.0001999997872997731, + "loss": 1.8483, + "step": 38 + }, + { + "epoch": 0.0009142214173948205, + "grad_norm": 4.078368186950684, + "learning_rate": 0.00019999977183560432, + "loss": 0.9528, + "step": 39 + }, + { + "epoch": 0.000937662992199816, + "grad_norm": 3.1793360710144043, + "learning_rate": 0.00019999975582883398, + "loss": 1.001, + "step": 40 + }, + { + "epoch": 0.0009611045670048114, + "grad_norm": 3.2885262966156006, + "learning_rate": 0.00019999973927946213, + "loss": 1.1364, + "step": 41 + }, + { + "epoch": 0.0009845461418098067, + "grad_norm": 3.7652838230133057, + "learning_rate": 0.0001999997221874889, + "loss": 1.2718, + "step": 42 + }, + { + "epoch": 0.0010079877166148021, + "grad_norm": 3.5545759201049805, + "learning_rate": 0.00019999970455291438, + "loss": 0.8515, + "step": 43 + }, + { + "epoch": 0.0010314292914197975, + "grad_norm": 0.9161327481269836, + "learning_rate": 0.0001999996863757386, + "loss": 0.2663, + "step": 44 + }, + { + "epoch": 0.001054870866224793, + "grad_norm": 1.1715126037597656, + "learning_rate": 0.00019999966765596176, + "loss": 0.1787, + "step": 45 + }, + { + "epoch": 0.0010783124410297883, + "grad_norm": 2.3909995555877686, + "learning_rate": 0.00019999964839358388, + "loss": 1.7456, + "step": 46 + }, + { + "epoch": 0.0011017540158347837, + "grad_norm": 1.2417851686477661, + "learning_rate": 0.0001999996285886051, + "loss": 0.4475, + "step": 47 + }, + { + "epoch": 0.0011251955906397791, + "grad_norm": 0.8984478712081909, + "learning_rate": 0.00019999960824102555, + "loss": 0.1567, + "step": 48 + }, + { + "epoch": 0.0011486371654447745, + "grad_norm": 0.7232295274734497, + "learning_rate": 0.0001999995873508453, + "loss": 0.8102, + "step": 49 + }, + { + "epoch": 0.00117207874024977, + "grad_norm": 2.970090627670288, + "learning_rate": 0.00019999956591806447, + "loss": 0.7218, + "step": 50 + }, + { + "epoch": 0.0011955203150547653, + "grad_norm": 1.6017177104949951, + "learning_rate": 0.00019999954394268315, + "loss": 0.5478, + "step": 51 + }, + { + "epoch": 0.0012189618898597607, + "grad_norm": 3.4306414127349854, + "learning_rate": 0.00019999952142470156, + "loss": 1.0177, + "step": 52 + }, + { + "epoch": 0.001242403464664756, + "grad_norm": 2.3879518508911133, + "learning_rate": 0.00019999949836411971, + "loss": 0.5061, + "step": 53 + }, + { + "epoch": 0.0012658450394697515, + "grad_norm": 3.1621785163879395, + "learning_rate": 0.00019999947476093778, + "loss": 0.9324, + "step": 54 + }, + { + "epoch": 0.001289286614274747, + "grad_norm": 3.3875718116760254, + "learning_rate": 0.00019999945061515591, + "loss": 0.5281, + "step": 55 + }, + { + "epoch": 0.0013127281890797423, + "grad_norm": 1.5043540000915527, + "learning_rate": 0.00019999942592677418, + "loss": 0.3114, + "step": 56 + }, + { + "epoch": 0.0013361697638847377, + "grad_norm": 1.416698932647705, + "learning_rate": 0.00019999940069579278, + "loss": 0.455, + "step": 57 + }, + { + "epoch": 0.001359611338689733, + "grad_norm": 5.548779487609863, + "learning_rate": 0.0001999993749222118, + "loss": 1.2939, + "step": 58 + }, + { + "epoch": 0.0013830529134947285, + "grad_norm": 4.745920658111572, + "learning_rate": 0.0001999993486060314, + "loss": 1.4376, + "step": 59 + }, + { + "epoch": 0.001406494488299724, + "grad_norm": 3.186375856399536, + "learning_rate": 0.00019999932174725175, + "loss": 0.6493, + "step": 60 + }, + { + "epoch": 0.0014299360631047193, + "grad_norm": 3.2452659606933594, + "learning_rate": 0.00019999929434587292, + "loss": 1.3028, + "step": 61 + }, + { + "epoch": 0.0014533776379097147, + "grad_norm": 2.4435737133026123, + "learning_rate": 0.0001999992664018951, + "loss": 0.6101, + "step": 62 + }, + { + "epoch": 0.00147681921271471, + "grad_norm": 3.39219069480896, + "learning_rate": 0.00019999923791531852, + "loss": 0.8325, + "step": 63 + }, + { + "epoch": 0.0015002607875197055, + "grad_norm": 3.422203540802002, + "learning_rate": 0.00019999920888614324, + "loss": 1.0441, + "step": 64 + }, + { + "epoch": 0.0015237023623247009, + "grad_norm": 3.2521800994873047, + "learning_rate": 0.0001999991793143694, + "loss": 0.6576, + "step": 65 + }, + { + "epoch": 0.0015471439371296963, + "grad_norm": 1.1701942682266235, + "learning_rate": 0.00019999914919999724, + "loss": 0.4792, + "step": 66 + }, + { + "epoch": 0.0015705855119346917, + "grad_norm": 3.4976768493652344, + "learning_rate": 0.00019999911854302687, + "loss": 0.753, + "step": 67 + }, + { + "epoch": 0.001594027086739687, + "grad_norm": 4.002079963684082, + "learning_rate": 0.00019999908734345847, + "loss": 1.1852, + "step": 68 + }, + { + "epoch": 0.0016174686615446825, + "grad_norm": 3.4799258708953857, + "learning_rate": 0.00019999905560129223, + "loss": 0.6372, + "step": 69 + }, + { + "epoch": 0.0016409102363496779, + "grad_norm": 1.7246028184890747, + "learning_rate": 0.00019999902331652825, + "loss": 0.5068, + "step": 70 + }, + { + "epoch": 0.0016643518111546733, + "grad_norm": 2.313500165939331, + "learning_rate": 0.00019999899048916682, + "loss": 0.3556, + "step": 71 + }, + { + "epoch": 0.0016877933859596687, + "grad_norm": 3.3656280040740967, + "learning_rate": 0.00019999895711920803, + "loss": 0.9237, + "step": 72 + }, + { + "epoch": 0.001711234960764664, + "grad_norm": 3.471652030944824, + "learning_rate": 0.00019999892320665208, + "loss": 0.8629, + "step": 73 + }, + { + "epoch": 0.0017346765355696595, + "grad_norm": 5.673941612243652, + "learning_rate": 0.00019999888875149916, + "loss": 1.1004, + "step": 74 + }, + { + "epoch": 0.0017581181103746549, + "grad_norm": 3.1473031044006348, + "learning_rate": 0.00019999885375374948, + "loss": 1.3747, + "step": 75 + }, + { + "epoch": 0.0017815596851796503, + "grad_norm": 3.558990240097046, + "learning_rate": 0.00019999881821340318, + "loss": 1.1249, + "step": 76 + }, + { + "epoch": 0.0018050012599846457, + "grad_norm": 2.7892940044403076, + "learning_rate": 0.0001999987821304605, + "loss": 0.8298, + "step": 77 + }, + { + "epoch": 0.001828442834789641, + "grad_norm": 4.292398452758789, + "learning_rate": 0.0001999987455049216, + "loss": 0.9511, + "step": 78 + }, + { + "epoch": 0.0018518844095946367, + "grad_norm": 3.013763427734375, + "learning_rate": 0.00019999870833678672, + "loss": 0.584, + "step": 79 + }, + { + "epoch": 0.001875325984399632, + "grad_norm": 2.5963003635406494, + "learning_rate": 0.00019999867062605602, + "loss": 1.0336, + "step": 80 + }, + { + "epoch": 0.0018987675592046275, + "grad_norm": 2.2973315715789795, + "learning_rate": 0.00019999863237272973, + "loss": 0.5794, + "step": 81 + }, + { + "epoch": 0.0019222091340096229, + "grad_norm": 2.854062080383301, + "learning_rate": 0.00019999859357680805, + "loss": 1.1987, + "step": 82 + }, + { + "epoch": 0.0019456507088146183, + "grad_norm": 3.9188570976257324, + "learning_rate": 0.00019999855423829117, + "loss": 0.7511, + "step": 83 + }, + { + "epoch": 0.0019690922836196134, + "grad_norm": 0.9795140027999878, + "learning_rate": 0.00019999851435717933, + "loss": 0.4941, + "step": 84 + }, + { + "epoch": 0.001992533858424609, + "grad_norm": 2.6419050693511963, + "learning_rate": 0.00019999847393347274, + "loss": 0.8582, + "step": 85 + }, + { + "epoch": 0.0020159754332296042, + "grad_norm": 2.1949875354766846, + "learning_rate": 0.00019999843296717161, + "loss": 0.6067, + "step": 86 + }, + { + "epoch": 0.0020394170080346, + "grad_norm": 3.160332679748535, + "learning_rate": 0.0001999983914582762, + "loss": 1.0286, + "step": 87 + }, + { + "epoch": 0.002062858582839595, + "grad_norm": 1.0186163187026978, + "learning_rate": 0.0001999983494067867, + "loss": 0.2979, + "step": 88 + }, + { + "epoch": 0.0020863001576445907, + "grad_norm": 2.6671669483184814, + "learning_rate": 0.00019999830681270334, + "loss": 1.5101, + "step": 89 + }, + { + "epoch": 0.002109741732449586, + "grad_norm": 3.4849936962127686, + "learning_rate": 0.00019999826367602633, + "loss": 0.8794, + "step": 90 + }, + { + "epoch": 0.0021331833072545815, + "grad_norm": 3.493051767349243, + "learning_rate": 0.00019999821999675595, + "loss": 0.9617, + "step": 91 + }, + { + "epoch": 0.0021566248820595766, + "grad_norm": 3.9024431705474854, + "learning_rate": 0.00019999817577489242, + "loss": 1.0904, + "step": 92 + }, + { + "epoch": 0.0021800664568645723, + "grad_norm": 3.279397487640381, + "learning_rate": 0.00019999813101043598, + "loss": 0.9222, + "step": 93 + }, + { + "epoch": 0.0022035080316695674, + "grad_norm": 3.6635959148406982, + "learning_rate": 0.00019999808570338688, + "loss": 0.863, + "step": 94 + }, + { + "epoch": 0.002226949606474563, + "grad_norm": 3.1693968772888184, + "learning_rate": 0.0001999980398537453, + "loss": 0.4136, + "step": 95 + }, + { + "epoch": 0.0022503911812795582, + "grad_norm": 3.4511115550994873, + "learning_rate": 0.0001999979934615116, + "loss": 0.8527, + "step": 96 + }, + { + "epoch": 0.002273832756084554, + "grad_norm": 3.0370497703552246, + "learning_rate": 0.00019999794652668596, + "loss": 0.8253, + "step": 97 + }, + { + "epoch": 0.002297274330889549, + "grad_norm": 1.1341345310211182, + "learning_rate": 0.00019999789904926867, + "loss": 0.216, + "step": 98 + }, + { + "epoch": 0.0023207159056945446, + "grad_norm": 2.2644119262695312, + "learning_rate": 0.00019999785102925994, + "loss": 1.7495, + "step": 99 + }, + { + "epoch": 0.00234415748049954, + "grad_norm": 2.5959715843200684, + "learning_rate": 0.00019999780246666003, + "loss": 0.8434, + "step": 100 + }, + { + "epoch": 0.0023675990553045354, + "grad_norm": 0.549773633480072, + "learning_rate": 0.00019999775336146928, + "loss": 0.1556, + "step": 101 + }, + { + "epoch": 0.0023910406301095306, + "grad_norm": 0.6018905639648438, + "learning_rate": 0.00019999770371368792, + "loss": 0.1954, + "step": 102 + }, + { + "epoch": 0.0024144822049145262, + "grad_norm": 3.7298941612243652, + "learning_rate": 0.00019999765352331615, + "loss": 0.9033, + "step": 103 + }, + { + "epoch": 0.0024379237797195214, + "grad_norm": 3.3127222061157227, + "learning_rate": 0.00019999760279035435, + "loss": 0.7644, + "step": 104 + }, + { + "epoch": 0.002461365354524517, + "grad_norm": 3.0934574604034424, + "learning_rate": 0.00019999755151480276, + "loss": 1.0507, + "step": 105 + }, + { + "epoch": 0.002484806929329512, + "grad_norm": 0.5371978282928467, + "learning_rate": 0.0001999974996966616, + "loss": 0.132, + "step": 106 + }, + { + "epoch": 0.002508248504134508, + "grad_norm": 0.43871673941612244, + "learning_rate": 0.0001999974473359312, + "loss": 0.111, + "step": 107 + }, + { + "epoch": 0.002531690078939503, + "grad_norm": 0.5490111708641052, + "learning_rate": 0.00019999739443261186, + "loss": 0.1423, + "step": 108 + }, + { + "epoch": 0.0025551316537444986, + "grad_norm": 3.602670431137085, + "learning_rate": 0.00019999734098670381, + "loss": 0.8937, + "step": 109 + }, + { + "epoch": 0.002578573228549494, + "grad_norm": 3.004990816116333, + "learning_rate": 0.00019999728699820742, + "loss": 0.8263, + "step": 110 + }, + { + "epoch": 0.0026020148033544894, + "grad_norm": 2.3443405628204346, + "learning_rate": 0.0001999972324671229, + "loss": 0.6401, + "step": 111 + }, + { + "epoch": 0.0026254563781594846, + "grad_norm": 3.147966146469116, + "learning_rate": 0.0001999971773934506, + "loss": 0.8515, + "step": 112 + }, + { + "epoch": 0.00264889795296448, + "grad_norm": 3.1494622230529785, + "learning_rate": 0.0001999971217771908, + "loss": 0.6162, + "step": 113 + }, + { + "epoch": 0.0026723395277694754, + "grad_norm": 3.338416337966919, + "learning_rate": 0.00019999706561834384, + "loss": 0.6232, + "step": 114 + }, + { + "epoch": 0.002695781102574471, + "grad_norm": 3.4481589794158936, + "learning_rate": 0.00019999700891690995, + "loss": 1.1118, + "step": 115 + }, + { + "epoch": 0.002719222677379466, + "grad_norm": 4.745134353637695, + "learning_rate": 0.0001999969516728895, + "loss": 0.7879, + "step": 116 + }, + { + "epoch": 0.002742664252184462, + "grad_norm": 0.6915575265884399, + "learning_rate": 0.00019999689388628276, + "loss": 0.1843, + "step": 117 + }, + { + "epoch": 0.002766105826989457, + "grad_norm": 3.3723371028900146, + "learning_rate": 0.00019999683555709007, + "loss": 0.8659, + "step": 118 + }, + { + "epoch": 0.0027895474017944526, + "grad_norm": 4.471100330352783, + "learning_rate": 0.00019999677668531174, + "loss": 1.4215, + "step": 119 + }, + { + "epoch": 0.002812988976599448, + "grad_norm": 4.171133041381836, + "learning_rate": 0.0001999967172709481, + "loss": 1.0697, + "step": 120 + }, + { + "epoch": 0.0028364305514044434, + "grad_norm": 3.436962604522705, + "learning_rate": 0.00019999665731399943, + "loss": 1.0117, + "step": 121 + }, + { + "epoch": 0.0028598721262094386, + "grad_norm": 0.9110831022262573, + "learning_rate": 0.00019999659681446612, + "loss": 1.0074, + "step": 122 + }, + { + "epoch": 0.002883313701014434, + "grad_norm": 1.4486567974090576, + "learning_rate": 0.00019999653577234847, + "loss": 0.3569, + "step": 123 + }, + { + "epoch": 0.0029067552758194294, + "grad_norm": 2.279021739959717, + "learning_rate": 0.00019999647418764678, + "loss": 0.5444, + "step": 124 + }, + { + "epoch": 0.002930196850624425, + "grad_norm": 2.7790920734405518, + "learning_rate": 0.00019999641206036142, + "loss": 0.6926, + "step": 125 + }, + { + "epoch": 0.00295363842542942, + "grad_norm": 1.3123496770858765, + "learning_rate": 0.0001999963493904927, + "loss": 0.288, + "step": 126 + }, + { + "epoch": 0.002977080000234416, + "grad_norm": 1.6177434921264648, + "learning_rate": 0.00019999628617804098, + "loss": 1.8384, + "step": 127 + }, + { + "epoch": 0.003000521575039411, + "grad_norm": 1.2375545501708984, + "learning_rate": 0.0001999962224230066, + "loss": 0.1498, + "step": 128 + }, + { + "epoch": 0.0030239631498444066, + "grad_norm": 3.2499520778656006, + "learning_rate": 0.00019999615812538993, + "loss": 0.7805, + "step": 129 + }, + { + "epoch": 0.0030474047246494018, + "grad_norm": 2.0288217067718506, + "learning_rate": 0.00019999609328519126, + "loss": 0.4695, + "step": 130 + }, + { + "epoch": 0.0030708462994543974, + "grad_norm": 1.5481867790222168, + "learning_rate": 0.000199996027902411, + "loss": 0.4491, + "step": 131 + }, + { + "epoch": 0.0030942878742593926, + "grad_norm": 1.6575515270233154, + "learning_rate": 0.00019999596197704946, + "loss": 1.4825, + "step": 132 + }, + { + "epoch": 0.003117729449064388, + "grad_norm": 5.461463451385498, + "learning_rate": 0.00019999589550910704, + "loss": 1.1603, + "step": 133 + }, + { + "epoch": 0.0031411710238693834, + "grad_norm": 5.084515571594238, + "learning_rate": 0.00019999582849858407, + "loss": 0.9814, + "step": 134 + }, + { + "epoch": 0.003164612598674379, + "grad_norm": 3.59169602394104, + "learning_rate": 0.00019999576094548093, + "loss": 1.0247, + "step": 135 + }, + { + "epoch": 0.003188054173479374, + "grad_norm": 3.1497249603271484, + "learning_rate": 0.00019999569284979795, + "loss": 0.7998, + "step": 136 + }, + { + "epoch": 0.0032114957482843698, + "grad_norm": 2.4013352394104004, + "learning_rate": 0.00019999562421153558, + "loss": 0.42, + "step": 137 + }, + { + "epoch": 0.003234937323089365, + "grad_norm": 4.272458553314209, + "learning_rate": 0.0001999955550306941, + "loss": 0.6945, + "step": 138 + }, + { + "epoch": 0.0032583788978943606, + "grad_norm": 10.410347938537598, + "learning_rate": 0.00019999548530727395, + "loss": 0.7239, + "step": 139 + }, + { + "epoch": 0.0032818204726993557, + "grad_norm": 3.93684458732605, + "learning_rate": 0.00019999541504127545, + "loss": 0.9936, + "step": 140 + }, + { + "epoch": 0.0033052620475043514, + "grad_norm": 3.1311206817626953, + "learning_rate": 0.00019999534423269908, + "loss": 0.6423, + "step": 141 + }, + { + "epoch": 0.0033287036223093465, + "grad_norm": 4.408672332763672, + "learning_rate": 0.00019999527288154512, + "loss": 1.1863, + "step": 142 + }, + { + "epoch": 0.003352145197114342, + "grad_norm": 2.047759771347046, + "learning_rate": 0.000199995200987814, + "loss": 0.7054, + "step": 143 + }, + { + "epoch": 0.0033755867719193373, + "grad_norm": 1.5945075750350952, + "learning_rate": 0.0001999951285515061, + "loss": 2.0284, + "step": 144 + }, + { + "epoch": 0.003399028346724333, + "grad_norm": 3.6566858291625977, + "learning_rate": 0.00019999505557262182, + "loss": 0.8915, + "step": 145 + }, + { + "epoch": 0.003422469921529328, + "grad_norm": 2.762582540512085, + "learning_rate": 0.00019999498205116155, + "loss": 0.9052, + "step": 146 + }, + { + "epoch": 0.0034459114963343238, + "grad_norm": 3.0755512714385986, + "learning_rate": 0.00019999490798712572, + "loss": 0.6242, + "step": 147 + }, + { + "epoch": 0.003469353071139319, + "grad_norm": 4.145164489746094, + "learning_rate": 0.0001999948333805147, + "loss": 0.8571, + "step": 148 + }, + { + "epoch": 0.0034927946459443146, + "grad_norm": 0.7031505107879639, + "learning_rate": 0.0001999947582313289, + "loss": 0.2308, + "step": 149 + }, + { + "epoch": 0.0035162362207493097, + "grad_norm": 1.4142907857894897, + "learning_rate": 0.00019999468253956875, + "loss": 0.2692, + "step": 150 + }, + { + "epoch": 0.0035396777955543053, + "grad_norm": 1.6493067741394043, + "learning_rate": 0.0001999946063052346, + "loss": 1.381, + "step": 151 + }, + { + "epoch": 0.0035631193703593005, + "grad_norm": 1.1576677560806274, + "learning_rate": 0.00019999452952832695, + "loss": 1.013, + "step": 152 + }, + { + "epoch": 0.003586560945164296, + "grad_norm": 1.1547516584396362, + "learning_rate": 0.00019999445220884616, + "loss": 0.2943, + "step": 153 + }, + { + "epoch": 0.0036100025199692913, + "grad_norm": 3.512631893157959, + "learning_rate": 0.00019999437434679266, + "loss": 0.5571, + "step": 154 + }, + { + "epoch": 0.003633444094774287, + "grad_norm": 3.626593589782715, + "learning_rate": 0.00019999429594216682, + "loss": 0.7373, + "step": 155 + }, + { + "epoch": 0.003656885669579282, + "grad_norm": 1.0257095098495483, + "learning_rate": 0.0001999942169949692, + "loss": 0.2618, + "step": 156 + }, + { + "epoch": 0.0036803272443842777, + "grad_norm": 3.561086893081665, + "learning_rate": 0.0001999941375052001, + "loss": 1.0892, + "step": 157 + }, + { + "epoch": 0.0037037688191892734, + "grad_norm": 3.4445087909698486, + "learning_rate": 0.00019999405747286, + "loss": 1.1169, + "step": 158 + }, + { + "epoch": 0.0037272103939942685, + "grad_norm": 1.2318432331085205, + "learning_rate": 0.00019999397689794934, + "loss": 0.4446, + "step": 159 + }, + { + "epoch": 0.003750651968799264, + "grad_norm": 3.0985333919525146, + "learning_rate": 0.00019999389578046855, + "loss": 0.401, + "step": 160 + }, + { + "epoch": 0.0037740935436042593, + "grad_norm": 3.303901195526123, + "learning_rate": 0.00019999381412041806, + "loss": 0.7751, + "step": 161 + }, + { + "epoch": 0.003797535118409255, + "grad_norm": 3.3994028568267822, + "learning_rate": 0.00019999373191779832, + "loss": 0.6931, + "step": 162 + }, + { + "epoch": 0.00382097669321425, + "grad_norm": 4.4089555740356445, + "learning_rate": 0.0001999936491726098, + "loss": 0.8465, + "step": 163 + }, + { + "epoch": 0.0038444182680192457, + "grad_norm": 1.2729086875915527, + "learning_rate": 0.00019999356588485293, + "loss": 0.3649, + "step": 164 + }, + { + "epoch": 0.003867859842824241, + "grad_norm": 4.096433639526367, + "learning_rate": 0.00019999348205452814, + "loss": 0.96, + "step": 165 + }, + { + "epoch": 0.0038913014176292365, + "grad_norm": 2.918997049331665, + "learning_rate": 0.00019999339768163592, + "loss": 0.438, + "step": 166 + }, + { + "epoch": 0.003914742992434232, + "grad_norm": 2.831378936767578, + "learning_rate": 0.0001999933127661767, + "loss": 0.5451, + "step": 167 + }, + { + "epoch": 0.003938184567239227, + "grad_norm": 2.142819404602051, + "learning_rate": 0.000199993227308151, + "loss": 0.4142, + "step": 168 + }, + { + "epoch": 0.003961626142044223, + "grad_norm": 3.134617567062378, + "learning_rate": 0.00019999314130755917, + "loss": 0.8618, + "step": 169 + }, + { + "epoch": 0.003985067716849218, + "grad_norm": 0.8980884552001953, + "learning_rate": 0.0001999930547644018, + "loss": 0.9176, + "step": 170 + }, + { + "epoch": 0.004008509291654213, + "grad_norm": 3.433706283569336, + "learning_rate": 0.00019999296767867923, + "loss": 0.7246, + "step": 171 + }, + { + "epoch": 0.0040319508664592085, + "grad_norm": 3.2497293949127197, + "learning_rate": 0.00019999288005039206, + "loss": 0.7549, + "step": 172 + }, + { + "epoch": 0.0040553924412642045, + "grad_norm": 2.5927412509918213, + "learning_rate": 0.0001999927918795407, + "loss": 0.942, + "step": 173 + }, + { + "epoch": 0.0040788340160692, + "grad_norm": 4.375139236450195, + "learning_rate": 0.00019999270316612563, + "loss": 1.1168, + "step": 174 + }, + { + "epoch": 0.004102275590874195, + "grad_norm": 0.35821831226348877, + "learning_rate": 0.00019999261391014737, + "loss": 0.0669, + "step": 175 + }, + { + "epoch": 0.00412571716567919, + "grad_norm": 4.675452709197998, + "learning_rate": 0.00019999252411160635, + "loss": 1.3701, + "step": 176 + }, + { + "epoch": 0.004149158740484186, + "grad_norm": 3.5235989093780518, + "learning_rate": 0.00019999243377050307, + "loss": 1.2818, + "step": 177 + }, + { + "epoch": 0.004172600315289181, + "grad_norm": 1.510743260383606, + "learning_rate": 0.00019999234288683808, + "loss": 1.3675, + "step": 178 + }, + { + "epoch": 0.0041960418900941765, + "grad_norm": 1.0558756589889526, + "learning_rate": 0.0001999922514606118, + "loss": 0.2154, + "step": 179 + }, + { + "epoch": 0.004219483464899172, + "grad_norm": 3.4941329956054688, + "learning_rate": 0.00019999215949182471, + "loss": 1.0471, + "step": 180 + }, + { + "epoch": 0.004242925039704168, + "grad_norm": 1.2503328323364258, + "learning_rate": 0.00019999206698047744, + "loss": 0.3916, + "step": 181 + }, + { + "epoch": 0.004266366614509163, + "grad_norm": 3.2636210918426514, + "learning_rate": 0.0001999919739265703, + "loss": 0.6434, + "step": 182 + }, + { + "epoch": 0.004289808189314158, + "grad_norm": 0.9269946813583374, + "learning_rate": 0.00019999188033010398, + "loss": 0.2627, + "step": 183 + }, + { + "epoch": 0.004313249764119153, + "grad_norm": 2.906816005706787, + "learning_rate": 0.0001999917861910789, + "loss": 1.0169, + "step": 184 + }, + { + "epoch": 0.004336691338924149, + "grad_norm": 1.2440530061721802, + "learning_rate": 0.00019999169150949553, + "loss": 1.3486, + "step": 185 + }, + { + "epoch": 0.0043601329137291445, + "grad_norm": 3.0394861698150635, + "learning_rate": 0.00019999159628535446, + "loss": 0.9434, + "step": 186 + }, + { + "epoch": 0.00438357448853414, + "grad_norm": 3.0136728286743164, + "learning_rate": 0.00019999150051865617, + "loss": 0.9512, + "step": 187 + }, + { + "epoch": 0.004407016063339135, + "grad_norm": 1.159151554107666, + "learning_rate": 0.0001999914042094012, + "loss": 0.4421, + "step": 188 + }, + { + "epoch": 0.004430457638144131, + "grad_norm": 2.746737003326416, + "learning_rate": 0.00019999130735759004, + "loss": 1.1611, + "step": 189 + }, + { + "epoch": 0.004453899212949126, + "grad_norm": 1.422320008277893, + "learning_rate": 0.00019999120996322324, + "loss": 1.5155, + "step": 190 + }, + { + "epoch": 0.004477340787754121, + "grad_norm": 3.09189510345459, + "learning_rate": 0.00019999111202630135, + "loss": 0.7044, + "step": 191 + }, + { + "epoch": 0.0045007823625591165, + "grad_norm": 3.378934383392334, + "learning_rate": 0.00019999101354682483, + "loss": 0.771, + "step": 192 + }, + { + "epoch": 0.0045242239373641125, + "grad_norm": 3.1223819255828857, + "learning_rate": 0.0001999909145247943, + "loss": 0.6903, + "step": 193 + }, + { + "epoch": 0.004547665512169108, + "grad_norm": 2.684309244155884, + "learning_rate": 0.00019999081496021022, + "loss": 0.5681, + "step": 194 + }, + { + "epoch": 0.004571107086974103, + "grad_norm": 5.04423713684082, + "learning_rate": 0.00019999071485307318, + "loss": 0.9943, + "step": 195 + }, + { + "epoch": 0.004594548661779098, + "grad_norm": 1.2168464660644531, + "learning_rate": 0.00019999061420338372, + "loss": 1.1971, + "step": 196 + }, + { + "epoch": 0.004617990236584094, + "grad_norm": 3.9145822525024414, + "learning_rate": 0.00019999051301114234, + "loss": 0.8329, + "step": 197 + }, + { + "epoch": 0.004641431811389089, + "grad_norm": 3.7784316539764404, + "learning_rate": 0.0001999904112763497, + "loss": 0.6413, + "step": 198 + }, + { + "epoch": 0.0046648733861940845, + "grad_norm": 3.233027935028076, + "learning_rate": 0.00019999030899900622, + "loss": 0.8894, + "step": 199 + }, + { + "epoch": 0.00468831496099908, + "grad_norm": 2.9643805027008057, + "learning_rate": 0.00019999020617911253, + "loss": 0.6361, + "step": 200 + }, + { + "epoch": 0.004711756535804076, + "grad_norm": 1.2054120302200317, + "learning_rate": 0.00019999010281666915, + "loss": 0.2399, + "step": 201 + }, + { + "epoch": 0.004735198110609071, + "grad_norm": 0.6971735954284668, + "learning_rate": 0.00019998999891167668, + "loss": 0.1283, + "step": 202 + }, + { + "epoch": 0.004758639685414066, + "grad_norm": 4.170069694519043, + "learning_rate": 0.00019998989446413562, + "loss": 0.8231, + "step": 203 + }, + { + "epoch": 0.004782081260219061, + "grad_norm": 4.747169494628906, + "learning_rate": 0.00019998978947404666, + "loss": 1.1844, + "step": 204 + }, + { + "epoch": 0.004805522835024057, + "grad_norm": 1.5041358470916748, + "learning_rate": 0.00019998968394141023, + "loss": 1.3512, + "step": 205 + }, + { + "epoch": 0.0048289644098290525, + "grad_norm": 3.6519343852996826, + "learning_rate": 0.00019998957786622698, + "loss": 0.6409, + "step": 206 + }, + { + "epoch": 0.004852405984634048, + "grad_norm": 0.8143812417984009, + "learning_rate": 0.00019998947124849747, + "loss": 0.203, + "step": 207 + }, + { + "epoch": 0.004875847559439043, + "grad_norm": 1.4165912866592407, + "learning_rate": 0.0001999893640882223, + "loss": 0.2927, + "step": 208 + }, + { + "epoch": 0.004899289134244039, + "grad_norm": 4.627612590789795, + "learning_rate": 0.000199989256385402, + "loss": 0.9176, + "step": 209 + }, + { + "epoch": 0.004922730709049034, + "grad_norm": 6.754210472106934, + "learning_rate": 0.0001999891481400372, + "loss": 0.9687, + "step": 210 + }, + { + "epoch": 0.004946172283854029, + "grad_norm": 0.4289917051792145, + "learning_rate": 0.00019998903935212846, + "loss": 0.0491, + "step": 211 + }, + { + "epoch": 0.004969613858659024, + "grad_norm": 0.5525429248809814, + "learning_rate": 0.00019998893002167635, + "loss": 0.0595, + "step": 212 + }, + { + "epoch": 0.0049930554334640205, + "grad_norm": 2.861809253692627, + "learning_rate": 0.00019998882014868156, + "loss": 0.5247, + "step": 213 + }, + { + "epoch": 0.005016497008269016, + "grad_norm": 3.691413640975952, + "learning_rate": 0.00019998870973314458, + "loss": 0.7414, + "step": 214 + }, + { + "epoch": 0.005039938583074011, + "grad_norm": 1.4731985330581665, + "learning_rate": 0.00019998859877506605, + "loss": 0.3458, + "step": 215 + }, + { + "epoch": 0.005063380157879006, + "grad_norm": 6.021062850952148, + "learning_rate": 0.00019998848727444658, + "loss": 1.0186, + "step": 216 + }, + { + "epoch": 0.005086821732684002, + "grad_norm": 2.378687858581543, + "learning_rate": 0.00019998837523128678, + "loss": 0.5563, + "step": 217 + }, + { + "epoch": 0.005110263307488997, + "grad_norm": 3.349968433380127, + "learning_rate": 0.0001999882626455872, + "loss": 0.616, + "step": 218 + }, + { + "epoch": 0.005133704882293992, + "grad_norm": 2.9851624965667725, + "learning_rate": 0.00019998814951734855, + "loss": 0.8844, + "step": 219 + }, + { + "epoch": 0.005157146457098988, + "grad_norm": 0.944996178150177, + "learning_rate": 0.00019998803584657138, + "loss": 0.1736, + "step": 220 + }, + { + "epoch": 0.005180588031903984, + "grad_norm": 2.9396016597747803, + "learning_rate": 0.00019998792163325628, + "loss": 0.7601, + "step": 221 + }, + { + "epoch": 0.005204029606708979, + "grad_norm": 1.108014464378357, + "learning_rate": 0.00019998780687740392, + "loss": 0.234, + "step": 222 + }, + { + "epoch": 0.005227471181513974, + "grad_norm": 3.606123685836792, + "learning_rate": 0.00019998769157901492, + "loss": 1.0608, + "step": 223 + }, + { + "epoch": 0.005250912756318969, + "grad_norm": 3.5015721321105957, + "learning_rate": 0.00019998757573808987, + "loss": 1.2222, + "step": 224 + }, + { + "epoch": 0.005274354331123965, + "grad_norm": 3.257699728012085, + "learning_rate": 0.00019998745935462947, + "loss": 0.6892, + "step": 225 + }, + { + "epoch": 0.00529779590592896, + "grad_norm": 2.1960043907165527, + "learning_rate": 0.00019998734242863427, + "loss": 0.2016, + "step": 226 + }, + { + "epoch": 0.005321237480733956, + "grad_norm": 3.1442551612854004, + "learning_rate": 0.00019998722496010496, + "loss": 0.5409, + "step": 227 + }, + { + "epoch": 0.005344679055538951, + "grad_norm": 0.799491822719574, + "learning_rate": 0.00019998710694904212, + "loss": 0.1891, + "step": 228 + }, + { + "epoch": 0.005368120630343947, + "grad_norm": 1.1602922677993774, + "learning_rate": 0.00019998698839544647, + "loss": 0.2324, + "step": 229 + }, + { + "epoch": 0.005391562205148942, + "grad_norm": 3.9545178413391113, + "learning_rate": 0.00019998686929931857, + "loss": 1.1127, + "step": 230 + }, + { + "epoch": 0.005415003779953937, + "grad_norm": 3.911311388015747, + "learning_rate": 0.0001999867496606591, + "loss": 0.898, + "step": 231 + }, + { + "epoch": 0.005438445354758932, + "grad_norm": 2.602107286453247, + "learning_rate": 0.00019998662947946877, + "loss": 0.4738, + "step": 232 + }, + { + "epoch": 0.005461886929563928, + "grad_norm": 1.118180751800537, + "learning_rate": 0.00019998650875574815, + "loss": 0.2761, + "step": 233 + }, + { + "epoch": 0.005485328504368924, + "grad_norm": 3.848644256591797, + "learning_rate": 0.0001999863874894979, + "loss": 1.1279, + "step": 234 + }, + { + "epoch": 0.005508770079173919, + "grad_norm": 0.5552775263786316, + "learning_rate": 0.00019998626568071872, + "loss": 0.106, + "step": 235 + }, + { + "epoch": 0.005532211653978914, + "grad_norm": 0.7378427982330322, + "learning_rate": 0.00019998614332941123, + "loss": 0.0829, + "step": 236 + }, + { + "epoch": 0.00555565322878391, + "grad_norm": 1.1120091676712036, + "learning_rate": 0.00019998602043557614, + "loss": 0.3116, + "step": 237 + }, + { + "epoch": 0.005579094803588905, + "grad_norm": 4.097359657287598, + "learning_rate": 0.0001999858969992141, + "loss": 0.8973, + "step": 238 + }, + { + "epoch": 0.0056025363783939, + "grad_norm": 3.032529354095459, + "learning_rate": 0.00019998577302032575, + "loss": 0.8293, + "step": 239 + }, + { + "epoch": 0.005625977953198896, + "grad_norm": 4.632148742675781, + "learning_rate": 0.0001999856484989118, + "loss": 1.0538, + "step": 240 + }, + { + "epoch": 0.005649419528003892, + "grad_norm": 3.6577985286712646, + "learning_rate": 0.0001999855234349729, + "loss": 0.8604, + "step": 241 + }, + { + "epoch": 0.005672861102808887, + "grad_norm": 4.086501598358154, + "learning_rate": 0.00019998539782850974, + "loss": 1.0126, + "step": 242 + }, + { + "epoch": 0.005696302677613882, + "grad_norm": 2.9697556495666504, + "learning_rate": 0.000199985271679523, + "loss": 0.6216, + "step": 243 + }, + { + "epoch": 0.005719744252418877, + "grad_norm": 1.703215479850769, + "learning_rate": 0.00019998514498801334, + "loss": 1.6783, + "step": 244 + }, + { + "epoch": 0.005743185827223873, + "grad_norm": 3.7474019527435303, + "learning_rate": 0.00019998501775398152, + "loss": 0.7118, + "step": 245 + }, + { + "epoch": 0.005766627402028868, + "grad_norm": 2.5688743591308594, + "learning_rate": 0.00019998488997742815, + "loss": 0.763, + "step": 246 + }, + { + "epoch": 0.005790068976833864, + "grad_norm": 2.611694574356079, + "learning_rate": 0.00019998476165835396, + "loss": 0.4319, + "step": 247 + }, + { + "epoch": 0.005813510551638859, + "grad_norm": 2.884038209915161, + "learning_rate": 0.00019998463279675966, + "loss": 0.763, + "step": 248 + }, + { + "epoch": 0.005836952126443855, + "grad_norm": 1.2166756391525269, + "learning_rate": 0.00019998450339264593, + "loss": 0.3063, + "step": 249 + }, + { + "epoch": 0.00586039370124885, + "grad_norm": 0.9469688534736633, + "learning_rate": 0.00019998437344601343, + "loss": 0.2152, + "step": 250 + }, + { + "epoch": 0.005883835276053845, + "grad_norm": 4.217563152313232, + "learning_rate": 0.00019998424295686294, + "loss": 0.9357, + "step": 251 + }, + { + "epoch": 0.00590727685085884, + "grad_norm": 3.244009017944336, + "learning_rate": 0.00019998411192519516, + "loss": 0.7135, + "step": 252 + }, + { + "epoch": 0.005930718425663836, + "grad_norm": 3.986856698989868, + "learning_rate": 0.00019998398035101072, + "loss": 0.7255, + "step": 253 + }, + { + "epoch": 0.005954160000468832, + "grad_norm": 3.3261687755584717, + "learning_rate": 0.00019998384823431041, + "loss": 0.6257, + "step": 254 + }, + { + "epoch": 0.005977601575273827, + "grad_norm": 2.449284553527832, + "learning_rate": 0.00019998371557509496, + "loss": 0.7563, + "step": 255 + }, + { + "epoch": 0.006001043150078822, + "grad_norm": 1.0738474130630493, + "learning_rate": 0.00019998358237336504, + "loss": 0.3172, + "step": 256 + }, + { + "epoch": 0.006024484724883818, + "grad_norm": 4.216825008392334, + "learning_rate": 0.0001999834486291214, + "loss": 0.9403, + "step": 257 + }, + { + "epoch": 0.006047926299688813, + "grad_norm": 3.1445322036743164, + "learning_rate": 0.00019998331434236476, + "loss": 0.7302, + "step": 258 + }, + { + "epoch": 0.006071367874493808, + "grad_norm": 1.755458950996399, + "learning_rate": 0.0001999831795130958, + "loss": 1.6911, + "step": 259 + }, + { + "epoch": 0.0060948094492988035, + "grad_norm": 4.0836944580078125, + "learning_rate": 0.00019998304414131534, + "loss": 0.971, + "step": 260 + }, + { + "epoch": 0.0061182510241038, + "grad_norm": 4.268324375152588, + "learning_rate": 0.00019998290822702405, + "loss": 0.8027, + "step": 261 + }, + { + "epoch": 0.006141692598908795, + "grad_norm": 2.700157403945923, + "learning_rate": 0.00019998277177022268, + "loss": 0.9411, + "step": 262 + }, + { + "epoch": 0.00616513417371379, + "grad_norm": 0.6474345922470093, + "learning_rate": 0.00019998263477091205, + "loss": 0.149, + "step": 263 + }, + { + "epoch": 0.006188575748518785, + "grad_norm": 2.564171314239502, + "learning_rate": 0.00019998249722909275, + "loss": 0.3961, + "step": 264 + }, + { + "epoch": 0.006212017323323781, + "grad_norm": 3.1684670448303223, + "learning_rate": 0.00019998235914476563, + "loss": 1.0601, + "step": 265 + }, + { + "epoch": 0.006235458898128776, + "grad_norm": 3.464191198348999, + "learning_rate": 0.00019998222051793144, + "loss": 0.7921, + "step": 266 + }, + { + "epoch": 0.0062589004729337715, + "grad_norm": 4.652545928955078, + "learning_rate": 0.0001999820813485909, + "loss": 0.8568, + "step": 267 + }, + { + "epoch": 0.006282342047738767, + "grad_norm": 0.7411962151527405, + "learning_rate": 0.00019998194163674477, + "loss": 0.1092, + "step": 268 + }, + { + "epoch": 0.006305783622543763, + "grad_norm": 3.5599145889282227, + "learning_rate": 0.0001999818013823938, + "loss": 0.7783, + "step": 269 + }, + { + "epoch": 0.006329225197348758, + "grad_norm": 1.0167779922485352, + "learning_rate": 0.00019998166058553878, + "loss": 0.2476, + "step": 270 + }, + { + "epoch": 0.006352666772153753, + "grad_norm": 3.774890184402466, + "learning_rate": 0.00019998151924618046, + "loss": 0.8761, + "step": 271 + }, + { + "epoch": 0.006376108346958748, + "grad_norm": 2.3481650352478027, + "learning_rate": 0.00019998137736431962, + "loss": 1.7017, + "step": 272 + }, + { + "epoch": 0.006399549921763744, + "grad_norm": 5.033433437347412, + "learning_rate": 0.00019998123493995698, + "loss": 1.0395, + "step": 273 + }, + { + "epoch": 0.0064229914965687395, + "grad_norm": 3.9426069259643555, + "learning_rate": 0.00019998109197309338, + "loss": 0.826, + "step": 274 + }, + { + "epoch": 0.006446433071373735, + "grad_norm": 3.388012647628784, + "learning_rate": 0.00019998094846372953, + "loss": 0.8455, + "step": 275 + }, + { + "epoch": 0.00646987464617873, + "grad_norm": 1.1563398838043213, + "learning_rate": 0.0001999808044118663, + "loss": 0.2504, + "step": 276 + }, + { + "epoch": 0.006493316220983726, + "grad_norm": 1.7231266498565674, + "learning_rate": 0.00019998065981750436, + "loss": 1.6053, + "step": 277 + }, + { + "epoch": 0.006516757795788721, + "grad_norm": 1.5560697317123413, + "learning_rate": 0.00019998051468064456, + "loss": 0.6146, + "step": 278 + }, + { + "epoch": 0.006540199370593716, + "grad_norm": 0.7061778903007507, + "learning_rate": 0.0001999803690012877, + "loss": 0.2194, + "step": 279 + }, + { + "epoch": 0.0065636409453987115, + "grad_norm": 3.057804822921753, + "learning_rate": 0.00019998022277943451, + "loss": 1.0212, + "step": 280 + }, + { + "epoch": 0.0065870825202037075, + "grad_norm": 3.3324427604675293, + "learning_rate": 0.00019998007601508585, + "loss": 0.5665, + "step": 281 + }, + { + "epoch": 0.006610524095008703, + "grad_norm": 1.8528079986572266, + "learning_rate": 0.00019997992870824247, + "loss": 0.4408, + "step": 282 + }, + { + "epoch": 0.006633965669813698, + "grad_norm": 1.0375474691390991, + "learning_rate": 0.00019997978085890518, + "loss": 0.2088, + "step": 283 + }, + { + "epoch": 0.006657407244618693, + "grad_norm": 3.2725019454956055, + "learning_rate": 0.0001999796324670748, + "loss": 1.144, + "step": 284 + }, + { + "epoch": 0.006680848819423689, + "grad_norm": 4.565009593963623, + "learning_rate": 0.00019997948353275215, + "loss": 0.9569, + "step": 285 + }, + { + "epoch": 0.006704290394228684, + "grad_norm": 4.577338218688965, + "learning_rate": 0.00019997933405593797, + "loss": 1.176, + "step": 286 + }, + { + "epoch": 0.0067277319690336795, + "grad_norm": 3.8452067375183105, + "learning_rate": 0.0001999791840366331, + "loss": 0.922, + "step": 287 + }, + { + "epoch": 0.006751173543838675, + "grad_norm": 0.977338433265686, + "learning_rate": 0.0001999790334748384, + "loss": 0.1867, + "step": 288 + }, + { + "epoch": 0.006774615118643671, + "grad_norm": 0.8039883971214294, + "learning_rate": 0.00019997888237055468, + "loss": 0.2707, + "step": 289 + }, + { + "epoch": 0.006798056693448666, + "grad_norm": 0.5904078483581543, + "learning_rate": 0.00019997873072378269, + "loss": 0.1062, + "step": 290 + }, + { + "epoch": 0.006821498268253661, + "grad_norm": 1.1991264820098877, + "learning_rate": 0.00019997857853452328, + "loss": 0.1693, + "step": 291 + }, + { + "epoch": 0.006844939843058656, + "grad_norm": 3.9614524841308594, + "learning_rate": 0.0001999784258027773, + "loss": 1.079, + "step": 292 + }, + { + "epoch": 0.006868381417863652, + "grad_norm": 1.9843370914459229, + "learning_rate": 0.0001999782725285456, + "loss": 1.8801, + "step": 293 + }, + { + "epoch": 0.0068918229926686475, + "grad_norm": 1.9271855354309082, + "learning_rate": 0.00019997811871182898, + "loss": 1.5359, + "step": 294 + }, + { + "epoch": 0.006915264567473643, + "grad_norm": 2.1458778381347656, + "learning_rate": 0.00019997796435262824, + "loss": 0.2225, + "step": 295 + }, + { + "epoch": 0.006938706142278638, + "grad_norm": 1.471805453300476, + "learning_rate": 0.00019997780945094431, + "loss": 1.651, + "step": 296 + }, + { + "epoch": 0.006962147717083634, + "grad_norm": 0.989862322807312, + "learning_rate": 0.00019997765400677793, + "loss": 0.1325, + "step": 297 + }, + { + "epoch": 0.006985589291888629, + "grad_norm": 2.977241277694702, + "learning_rate": 0.00019997749802013002, + "loss": 0.9079, + "step": 298 + }, + { + "epoch": 0.007009030866693624, + "grad_norm": 0.31945839524269104, + "learning_rate": 0.00019997734149100136, + "loss": 0.0711, + "step": 299 + }, + { + "epoch": 0.0070324724414986195, + "grad_norm": 2.7630529403686523, + "learning_rate": 0.00019997718441939287, + "loss": 0.5127, + "step": 300 + }, + { + "epoch": 0.0070559140163036155, + "grad_norm": 1.6786962747573853, + "learning_rate": 0.00019997702680530532, + "loss": 1.4113, + "step": 301 + }, + { + "epoch": 0.007079355591108611, + "grad_norm": 0.6178777813911438, + "learning_rate": 0.00019997686864873965, + "loss": 0.1157, + "step": 302 + }, + { + "epoch": 0.007102797165913606, + "grad_norm": 2.226016044616699, + "learning_rate": 0.00019997670994969667, + "loss": 0.3796, + "step": 303 + }, + { + "epoch": 0.007126238740718601, + "grad_norm": 0.9808973670005798, + "learning_rate": 0.00019997655070817723, + "loss": 0.245, + "step": 304 + }, + { + "epoch": 0.007149680315523597, + "grad_norm": 1.6630938053131104, + "learning_rate": 0.00019997639092418225, + "loss": 1.7187, + "step": 305 + }, + { + "epoch": 0.007173121890328592, + "grad_norm": 3.4778554439544678, + "learning_rate": 0.00019997623059771254, + "loss": 0.8666, + "step": 306 + }, + { + "epoch": 0.0071965634651335875, + "grad_norm": 3.1584129333496094, + "learning_rate": 0.000199976069728769, + "loss": 0.6195, + "step": 307 + }, + { + "epoch": 0.007220005039938583, + "grad_norm": 6.077291488647461, + "learning_rate": 0.00019997590831735248, + "loss": 1.2281, + "step": 308 + }, + { + "epoch": 0.007243446614743579, + "grad_norm": 2.6721315383911133, + "learning_rate": 0.00019997574636346387, + "loss": 0.8758, + "step": 309 + }, + { + "epoch": 0.007266888189548574, + "grad_norm": 3.848665237426758, + "learning_rate": 0.00019997558386710407, + "loss": 0.9877, + "step": 310 + }, + { + "epoch": 0.007290329764353569, + "grad_norm": 2.672657012939453, + "learning_rate": 0.00019997542082827394, + "loss": 0.8841, + "step": 311 + }, + { + "epoch": 0.007313771339158564, + "grad_norm": 3.752558946609497, + "learning_rate": 0.00019997525724697437, + "loss": 1.2043, + "step": 312 + }, + { + "epoch": 0.00733721291396356, + "grad_norm": 0.9092252254486084, + "learning_rate": 0.00019997509312320622, + "loss": 0.1726, + "step": 313 + }, + { + "epoch": 0.0073606544887685555, + "grad_norm": 2.4648001194000244, + "learning_rate": 0.0001999749284569704, + "loss": 1.9459, + "step": 314 + }, + { + "epoch": 0.007384096063573551, + "grad_norm": 0.9067240357398987, + "learning_rate": 0.00019997476324826782, + "loss": 0.1851, + "step": 315 + }, + { + "epoch": 0.007407537638378547, + "grad_norm": 3.201259136199951, + "learning_rate": 0.00019997459749709936, + "loss": 0.5057, + "step": 316 + }, + { + "epoch": 0.007430979213183542, + "grad_norm": 3.878216505050659, + "learning_rate": 0.00019997443120346593, + "loss": 1.2643, + "step": 317 + }, + { + "epoch": 0.007454420787988537, + "grad_norm": 2.693251848220825, + "learning_rate": 0.00019997426436736844, + "loss": 0.4157, + "step": 318 + }, + { + "epoch": 0.007477862362793532, + "grad_norm": 2.6063756942749023, + "learning_rate": 0.00019997409698880775, + "loss": 0.452, + "step": 319 + }, + { + "epoch": 0.007501303937598528, + "grad_norm": 3.2739670276641846, + "learning_rate": 0.0001999739290677848, + "loss": 1.1013, + "step": 320 + }, + { + "epoch": 0.0075247455124035235, + "grad_norm": 4.948163032531738, + "learning_rate": 0.00019997376060430054, + "loss": 1.4667, + "step": 321 + }, + { + "epoch": 0.007548187087208519, + "grad_norm": 4.292586803436279, + "learning_rate": 0.0001999735915983558, + "loss": 1.1798, + "step": 322 + }, + { + "epoch": 0.007571628662013514, + "grad_norm": 5.725894927978516, + "learning_rate": 0.00019997342204995156, + "loss": 0.9029, + "step": 323 + }, + { + "epoch": 0.00759507023681851, + "grad_norm": 2.5722458362579346, + "learning_rate": 0.0001999732519590887, + "loss": 0.5229, + "step": 324 + }, + { + "epoch": 0.007618511811623505, + "grad_norm": 3.8627240657806396, + "learning_rate": 0.0001999730813257682, + "loss": 0.6787, + "step": 325 + }, + { + "epoch": 0.0076419533864285, + "grad_norm": 4.343163967132568, + "learning_rate": 0.00019997291014999094, + "loss": 0.9529, + "step": 326 + }, + { + "epoch": 0.007665394961233495, + "grad_norm": 5.247485637664795, + "learning_rate": 0.00019997273843175783, + "loss": 0.9678, + "step": 327 + }, + { + "epoch": 0.0076888365360384915, + "grad_norm": 4.387784957885742, + "learning_rate": 0.00019997256617106985, + "loss": 0.8682, + "step": 328 + }, + { + "epoch": 0.007712278110843487, + "grad_norm": 8.03680419921875, + "learning_rate": 0.00019997239336792791, + "loss": 1.4963, + "step": 329 + }, + { + "epoch": 0.007735719685648482, + "grad_norm": 1.1137197017669678, + "learning_rate": 0.00019997222002233298, + "loss": 0.2482, + "step": 330 + }, + { + "epoch": 0.007759161260453477, + "grad_norm": 3.8608381748199463, + "learning_rate": 0.00019997204613428596, + "loss": 0.8911, + "step": 331 + }, + { + "epoch": 0.007782602835258473, + "grad_norm": 3.5291664600372314, + "learning_rate": 0.00019997187170378778, + "loss": 0.5124, + "step": 332 + }, + { + "epoch": 0.007806044410063468, + "grad_norm": 3.4093239307403564, + "learning_rate": 0.00019997169673083943, + "loss": 0.4571, + "step": 333 + }, + { + "epoch": 0.007829485984868463, + "grad_norm": 3.296489953994751, + "learning_rate": 0.00019997152121544186, + "loss": 1.0596, + "step": 334 + }, + { + "epoch": 0.00785292755967346, + "grad_norm": 3.685343027114868, + "learning_rate": 0.00019997134515759596, + "loss": 0.7916, + "step": 335 + }, + { + "epoch": 0.007876369134478454, + "grad_norm": 0.796917200088501, + "learning_rate": 0.00019997116855730279, + "loss": 0.2602, + "step": 336 + }, + { + "epoch": 0.00789981070928345, + "grad_norm": 3.2621710300445557, + "learning_rate": 0.00019997099141456318, + "loss": 0.839, + "step": 337 + }, + { + "epoch": 0.007923252284088446, + "grad_norm": 2.7043988704681396, + "learning_rate": 0.00019997081372937823, + "loss": 0.7732, + "step": 338 + }, + { + "epoch": 0.00794669385889344, + "grad_norm": 1.0761784315109253, + "learning_rate": 0.0001999706355017488, + "loss": 0.372, + "step": 339 + }, + { + "epoch": 0.007970135433698436, + "grad_norm": 3.354586601257324, + "learning_rate": 0.0001999704567316759, + "loss": 0.9048, + "step": 340 + }, + { + "epoch": 0.00799357700850343, + "grad_norm": 3.1785502433776855, + "learning_rate": 0.00019997027741916045, + "loss": 0.9738, + "step": 341 + }, + { + "epoch": 0.008017018583308427, + "grad_norm": 4.01973819732666, + "learning_rate": 0.0001999700975642035, + "loss": 0.9143, + "step": 342 + }, + { + "epoch": 0.008040460158113423, + "grad_norm": 3.7564003467559814, + "learning_rate": 0.000199969917166806, + "loss": 0.7783, + "step": 343 + }, + { + "epoch": 0.008063901732918417, + "grad_norm": 3.9998128414154053, + "learning_rate": 0.00019996973622696888, + "loss": 0.7879, + "step": 344 + }, + { + "epoch": 0.008087343307723413, + "grad_norm": 2.018810272216797, + "learning_rate": 0.0001999695547446932, + "loss": 0.3707, + "step": 345 + }, + { + "epoch": 0.008110784882528409, + "grad_norm": 2.856811761856079, + "learning_rate": 0.0001999693727199799, + "loss": 0.6792, + "step": 346 + }, + { + "epoch": 0.008134226457333403, + "grad_norm": 5.997770309448242, + "learning_rate": 0.00019996919015282996, + "loss": 0.9062, + "step": 347 + }, + { + "epoch": 0.0081576680321384, + "grad_norm": 4.202394008636475, + "learning_rate": 0.00019996900704324436, + "loss": 0.8561, + "step": 348 + }, + { + "epoch": 0.008181109606943394, + "grad_norm": 1.5376570224761963, + "learning_rate": 0.00019996882339122415, + "loss": 1.5696, + "step": 349 + }, + { + "epoch": 0.00820455118174839, + "grad_norm": 1.2513118982315063, + "learning_rate": 0.0001999686391967703, + "loss": 0.169, + "step": 350 + }, + { + "epoch": 0.008227992756553386, + "grad_norm": 2.5907223224639893, + "learning_rate": 0.00019996845445988378, + "loss": 0.7613, + "step": 351 + }, + { + "epoch": 0.00825143433135838, + "grad_norm": 4.816099166870117, + "learning_rate": 0.00019996826918056563, + "loss": 0.9663, + "step": 352 + }, + { + "epoch": 0.008274875906163376, + "grad_norm": 1.270471215248108, + "learning_rate": 0.00019996808335881686, + "loss": 0.2193, + "step": 353 + }, + { + "epoch": 0.008298317480968372, + "grad_norm": 4.694580078125, + "learning_rate": 0.0001999678969946384, + "loss": 0.9492, + "step": 354 + }, + { + "epoch": 0.008321759055773367, + "grad_norm": 2.7515718936920166, + "learning_rate": 0.00019996771008803138, + "loss": 0.4693, + "step": 355 + }, + { + "epoch": 0.008345200630578363, + "grad_norm": 0.6539778709411621, + "learning_rate": 0.00019996752263899673, + "loss": 0.1887, + "step": 356 + }, + { + "epoch": 0.008368642205383357, + "grad_norm": 3.151926279067993, + "learning_rate": 0.00019996733464753546, + "loss": 0.4862, + "step": 357 + }, + { + "epoch": 0.008392083780188353, + "grad_norm": 2.1386640071868896, + "learning_rate": 0.00019996714611364866, + "loss": 1.7373, + "step": 358 + }, + { + "epoch": 0.008415525354993349, + "grad_norm": 4.365428924560547, + "learning_rate": 0.0001999669570373373, + "loss": 0.858, + "step": 359 + }, + { + "epoch": 0.008438966929798343, + "grad_norm": 4.186253070831299, + "learning_rate": 0.00019996676741860242, + "loss": 1.079, + "step": 360 + }, + { + "epoch": 0.00846240850460334, + "grad_norm": 2.7644131183624268, + "learning_rate": 0.00019996657725744504, + "loss": 0.5859, + "step": 361 + }, + { + "epoch": 0.008485850079408335, + "grad_norm": 1.0490742921829224, + "learning_rate": 0.00019996638655386624, + "loss": 0.2646, + "step": 362 + }, + { + "epoch": 0.00850929165421333, + "grad_norm": 3.4574692249298096, + "learning_rate": 0.000199966195307867, + "loss": 0.9092, + "step": 363 + }, + { + "epoch": 0.008532733229018326, + "grad_norm": 3.587357997894287, + "learning_rate": 0.00019996600351944834, + "loss": 0.7785, + "step": 364 + }, + { + "epoch": 0.00855617480382332, + "grad_norm": 8.444808959960938, + "learning_rate": 0.00019996581118861137, + "loss": 0.5165, + "step": 365 + }, + { + "epoch": 0.008579616378628316, + "grad_norm": 0.584662914276123, + "learning_rate": 0.00019996561831535709, + "loss": 0.1185, + "step": 366 + }, + { + "epoch": 0.008603057953433312, + "grad_norm": 0.33627158403396606, + "learning_rate": 0.00019996542489968656, + "loss": 0.0461, + "step": 367 + }, + { + "epoch": 0.008626499528238307, + "grad_norm": 2.8643531799316406, + "learning_rate": 0.0001999652309416008, + "loss": 0.4661, + "step": 368 + }, + { + "epoch": 0.008649941103043303, + "grad_norm": 1.4946919679641724, + "learning_rate": 0.0001999650364411009, + "loss": 0.2522, + "step": 369 + }, + { + "epoch": 0.008673382677848299, + "grad_norm": 3.534587860107422, + "learning_rate": 0.00019996484139818792, + "loss": 0.9068, + "step": 370 + }, + { + "epoch": 0.008696824252653293, + "grad_norm": 3.9063844680786133, + "learning_rate": 0.00019996464581286286, + "loss": 0.8126, + "step": 371 + }, + { + "epoch": 0.008720265827458289, + "grad_norm": 3.3564581871032715, + "learning_rate": 0.00019996444968512688, + "loss": 0.6186, + "step": 372 + }, + { + "epoch": 0.008743707402263283, + "grad_norm": 2.922025680541992, + "learning_rate": 0.0001999642530149809, + "loss": 0.5178, + "step": 373 + }, + { + "epoch": 0.00876714897706828, + "grad_norm": 0.8486838936805725, + "learning_rate": 0.00019996405580242614, + "loss": 0.2351, + "step": 374 + }, + { + "epoch": 0.008790590551873275, + "grad_norm": 1.1923085451126099, + "learning_rate": 0.0001999638580474636, + "loss": 0.3554, + "step": 375 + }, + { + "epoch": 0.00881403212667827, + "grad_norm": 2.6981770992279053, + "learning_rate": 0.00019996365975009431, + "loss": 0.4951, + "step": 376 + }, + { + "epoch": 0.008837473701483266, + "grad_norm": 1.2821025848388672, + "learning_rate": 0.00019996346091031943, + "loss": 0.4151, + "step": 377 + }, + { + "epoch": 0.008860915276288262, + "grad_norm": 3.061244010925293, + "learning_rate": 0.00019996326152814, + "loss": 0.6682, + "step": 378 + }, + { + "epoch": 0.008884356851093256, + "grad_norm": 3.209812879562378, + "learning_rate": 0.00019996306160355708, + "loss": 0.6568, + "step": 379 + }, + { + "epoch": 0.008907798425898252, + "grad_norm": 3.7013070583343506, + "learning_rate": 0.00019996286113657178, + "loss": 0.8942, + "step": 380 + }, + { + "epoch": 0.008931240000703247, + "grad_norm": 2.0086488723754883, + "learning_rate": 0.0001999626601271852, + "loss": 1.1127, + "step": 381 + }, + { + "epoch": 0.008954681575508243, + "grad_norm": 3.353661298751831, + "learning_rate": 0.00019996245857539842, + "loss": 0.7804, + "step": 382 + }, + { + "epoch": 0.008978123150313239, + "grad_norm": 0.5593745112419128, + "learning_rate": 0.00019996225648121252, + "loss": 0.0688, + "step": 383 + }, + { + "epoch": 0.009001564725118233, + "grad_norm": 2.919527053833008, + "learning_rate": 0.0001999620538446286, + "loss": 0.6436, + "step": 384 + }, + { + "epoch": 0.009025006299923229, + "grad_norm": 3.4976766109466553, + "learning_rate": 0.00019996185066564778, + "loss": 0.6107, + "step": 385 + }, + { + "epoch": 0.009048447874728225, + "grad_norm": 2.887633800506592, + "learning_rate": 0.0001999616469442711, + "loss": 0.6718, + "step": 386 + }, + { + "epoch": 0.00907188944953322, + "grad_norm": 0.4646814167499542, + "learning_rate": 0.00019996144268049978, + "loss": 0.0671, + "step": 387 + }, + { + "epoch": 0.009095331024338215, + "grad_norm": 1.1693416833877563, + "learning_rate": 0.00019996123787433483, + "loss": 0.2514, + "step": 388 + }, + { + "epoch": 0.00911877259914321, + "grad_norm": 3.999157667160034, + "learning_rate": 0.0001999610325257774, + "loss": 0.5505, + "step": 389 + }, + { + "epoch": 0.009142214173948206, + "grad_norm": 3.5973901748657227, + "learning_rate": 0.00019996082663482857, + "loss": 0.6921, + "step": 390 + }, + { + "epoch": 0.009165655748753202, + "grad_norm": 1.2783763408660889, + "learning_rate": 0.0001999606202014895, + "loss": 0.2921, + "step": 391 + }, + { + "epoch": 0.009189097323558196, + "grad_norm": 2.841029167175293, + "learning_rate": 0.0001999604132257613, + "loss": 0.5796, + "step": 392 + }, + { + "epoch": 0.009212538898363192, + "grad_norm": 2.3846139907836914, + "learning_rate": 0.00019996020570764512, + "loss": 0.4119, + "step": 393 + }, + { + "epoch": 0.009235980473168188, + "grad_norm": 3.7081551551818848, + "learning_rate": 0.00019995999764714202, + "loss": 0.9413, + "step": 394 + }, + { + "epoch": 0.009259422047973183, + "grad_norm": 4.156251430511475, + "learning_rate": 0.00019995978904425315, + "loss": 0.8525, + "step": 395 + }, + { + "epoch": 0.009282863622778179, + "grad_norm": 3.5711631774902344, + "learning_rate": 0.00019995957989897965, + "loss": 1.1295, + "step": 396 + }, + { + "epoch": 0.009306305197583173, + "grad_norm": 3.753122568130493, + "learning_rate": 0.00019995937021132266, + "loss": 0.9767, + "step": 397 + }, + { + "epoch": 0.009329746772388169, + "grad_norm": 3.3710715770721436, + "learning_rate": 0.00019995915998128334, + "loss": 0.9098, + "step": 398 + }, + { + "epoch": 0.009353188347193165, + "grad_norm": 1.667123794555664, + "learning_rate": 0.0001999589492088628, + "loss": 0.4543, + "step": 399 + }, + { + "epoch": 0.00937662992199816, + "grad_norm": 2.679596185684204, + "learning_rate": 0.0001999587378940622, + "loss": 0.4119, + "step": 400 + }, + { + "epoch": 0.009400071496803155, + "grad_norm": 2.3200786113739014, + "learning_rate": 0.00019995852603688267, + "loss": 0.3309, + "step": 401 + }, + { + "epoch": 0.009423513071608151, + "grad_norm": 3.605694055557251, + "learning_rate": 0.00019995831363732537, + "loss": 1.3016, + "step": 402 + }, + { + "epoch": 0.009446954646413146, + "grad_norm": 3.516500234603882, + "learning_rate": 0.00019995810069539145, + "loss": 1.1791, + "step": 403 + }, + { + "epoch": 0.009470396221218142, + "grad_norm": 1.056617021560669, + "learning_rate": 0.00019995788721108206, + "loss": 0.2613, + "step": 404 + }, + { + "epoch": 0.009493837796023136, + "grad_norm": 1.1377627849578857, + "learning_rate": 0.00019995767318439835, + "loss": 0.2875, + "step": 405 + }, + { + "epoch": 0.009517279370828132, + "grad_norm": 2.213689088821411, + "learning_rate": 0.00019995745861534151, + "loss": 0.8709, + "step": 406 + }, + { + "epoch": 0.009540720945633128, + "grad_norm": 3.0719752311706543, + "learning_rate": 0.00019995724350391268, + "loss": 0.7656, + "step": 407 + }, + { + "epoch": 0.009564162520438122, + "grad_norm": 0.20363864302635193, + "learning_rate": 0.00019995702785011307, + "loss": 0.0461, + "step": 408 + }, + { + "epoch": 0.009587604095243119, + "grad_norm": 0.7189046144485474, + "learning_rate": 0.00019995681165394378, + "loss": 0.1675, + "step": 409 + }, + { + "epoch": 0.009611045670048115, + "grad_norm": 3.091827630996704, + "learning_rate": 0.00019995659491540602, + "loss": 0.7439, + "step": 410 + }, + { + "epoch": 0.009634487244853109, + "grad_norm": 1.0315017700195312, + "learning_rate": 0.00019995637763450098, + "loss": 0.2366, + "step": 411 + }, + { + "epoch": 0.009657928819658105, + "grad_norm": 2.657601833343506, + "learning_rate": 0.00019995615981122978, + "loss": 0.8087, + "step": 412 + }, + { + "epoch": 0.0096813703944631, + "grad_norm": 1.1227748394012451, + "learning_rate": 0.00019995594144559372, + "loss": 0.2206, + "step": 413 + }, + { + "epoch": 0.009704811969268095, + "grad_norm": 4.402198791503906, + "learning_rate": 0.00019995572253759386, + "loss": 0.854, + "step": 414 + }, + { + "epoch": 0.009728253544073091, + "grad_norm": 1.7824945449829102, + "learning_rate": 0.00019995550308723147, + "loss": 0.3981, + "step": 415 + }, + { + "epoch": 0.009751695118878086, + "grad_norm": 3.1997158527374268, + "learning_rate": 0.0001999552830945077, + "loss": 0.8917, + "step": 416 + }, + { + "epoch": 0.009775136693683082, + "grad_norm": 1.192685842514038, + "learning_rate": 0.0001999550625594237, + "loss": 0.2622, + "step": 417 + }, + { + "epoch": 0.009798578268488078, + "grad_norm": 0.5539443492889404, + "learning_rate": 0.0001999548414819808, + "loss": 0.0884, + "step": 418 + }, + { + "epoch": 0.009822019843293072, + "grad_norm": 3.7812399864196777, + "learning_rate": 0.0001999546198621801, + "loss": 0.8587, + "step": 419 + }, + { + "epoch": 0.009845461418098068, + "grad_norm": 3.040311813354492, + "learning_rate": 0.0001999543977000228, + "loss": 0.9749, + "step": 420 + }, + { + "epoch": 0.009868902992903062, + "grad_norm": 2.9775354862213135, + "learning_rate": 0.00019995417499551012, + "loss": 0.8232, + "step": 421 + }, + { + "epoch": 0.009892344567708058, + "grad_norm": 2.5154356956481934, + "learning_rate": 0.00019995395174864328, + "loss": 0.5278, + "step": 422 + }, + { + "epoch": 0.009915786142513055, + "grad_norm": 1.232162356376648, + "learning_rate": 0.0001999537279594235, + "loss": 0.229, + "step": 423 + }, + { + "epoch": 0.009939227717318049, + "grad_norm": 3.3375113010406494, + "learning_rate": 0.00019995350362785202, + "loss": 0.7459, + "step": 424 + }, + { + "epoch": 0.009962669292123045, + "grad_norm": 1.5528470277786255, + "learning_rate": 0.00019995327875392996, + "loss": 0.4141, + "step": 425 + }, + { + "epoch": 0.009986110866928041, + "grad_norm": 3.9857707023620605, + "learning_rate": 0.00019995305333765862, + "loss": 0.8426, + "step": 426 + }, + { + "epoch": 0.010009552441733035, + "grad_norm": 3.7219483852386475, + "learning_rate": 0.0001999528273790392, + "loss": 0.6501, + "step": 427 + }, + { + "epoch": 0.010032994016538031, + "grad_norm": 4.0438408851623535, + "learning_rate": 0.00019995260087807295, + "loss": 1.4348, + "step": 428 + }, + { + "epoch": 0.010056435591343026, + "grad_norm": 1.4484838247299194, + "learning_rate": 0.00019995237383476104, + "loss": 0.2365, + "step": 429 + }, + { + "epoch": 0.010079877166148022, + "grad_norm": 4.771573543548584, + "learning_rate": 0.00019995214624910477, + "loss": 1.0235, + "step": 430 + }, + { + "epoch": 0.010103318740953018, + "grad_norm": 0.6659011840820312, + "learning_rate": 0.00019995191812110532, + "loss": 0.1894, + "step": 431 + }, + { + "epoch": 0.010126760315758012, + "grad_norm": 2.6806211471557617, + "learning_rate": 0.00019995168945076397, + "loss": 0.5007, + "step": 432 + }, + { + "epoch": 0.010150201890563008, + "grad_norm": 1.0320320129394531, + "learning_rate": 0.00019995146023808194, + "loss": 0.2453, + "step": 433 + }, + { + "epoch": 0.010173643465368004, + "grad_norm": 2.666579008102417, + "learning_rate": 0.00019995123048306045, + "loss": 0.7114, + "step": 434 + }, + { + "epoch": 0.010197085040172998, + "grad_norm": 3.2198822498321533, + "learning_rate": 0.00019995100018570082, + "loss": 0.8049, + "step": 435 + }, + { + "epoch": 0.010220526614977994, + "grad_norm": 2.8378233909606934, + "learning_rate": 0.00019995076934600423, + "loss": 0.6267, + "step": 436 + }, + { + "epoch": 0.010243968189782989, + "grad_norm": 4.8360066413879395, + "learning_rate": 0.00019995053796397193, + "loss": 0.6669, + "step": 437 + }, + { + "epoch": 0.010267409764587985, + "grad_norm": 3.5620100498199463, + "learning_rate": 0.00019995030603960523, + "loss": 0.9458, + "step": 438 + }, + { + "epoch": 0.010290851339392981, + "grad_norm": 2.6390509605407715, + "learning_rate": 0.00019995007357290537, + "loss": 0.3804, + "step": 439 + }, + { + "epoch": 0.010314292914197975, + "grad_norm": 1.9786664247512817, + "learning_rate": 0.00019994984056387357, + "loss": 1.5177, + "step": 440 + }, + { + "epoch": 0.010337734489002971, + "grad_norm": 2.7225747108459473, + "learning_rate": 0.00019994960701251113, + "loss": 0.6738, + "step": 441 + }, + { + "epoch": 0.010361176063807967, + "grad_norm": 3.1037466526031494, + "learning_rate": 0.00019994937291881933, + "loss": 0.8562, + "step": 442 + }, + { + "epoch": 0.010384617638612962, + "grad_norm": 1.8316909074783325, + "learning_rate": 0.00019994913828279943, + "loss": 0.2263, + "step": 443 + }, + { + "epoch": 0.010408059213417958, + "grad_norm": 3.264254570007324, + "learning_rate": 0.00019994890310445266, + "loss": 0.2882, + "step": 444 + }, + { + "epoch": 0.010431500788222952, + "grad_norm": 1.9972991943359375, + "learning_rate": 0.00019994866738378035, + "loss": 0.579, + "step": 445 + }, + { + "epoch": 0.010454942363027948, + "grad_norm": 4.650149822235107, + "learning_rate": 0.00019994843112078377, + "loss": 0.8719, + "step": 446 + }, + { + "epoch": 0.010478383937832944, + "grad_norm": 0.7443316578865051, + "learning_rate": 0.00019994819431546414, + "loss": 0.1028, + "step": 447 + }, + { + "epoch": 0.010501825512637938, + "grad_norm": 0.6186041235923767, + "learning_rate": 0.00019994795696782287, + "loss": 0.0914, + "step": 448 + }, + { + "epoch": 0.010525267087442934, + "grad_norm": 1.2894401550292969, + "learning_rate": 0.00019994771907786113, + "loss": 0.4346, + "step": 449 + }, + { + "epoch": 0.01054870866224793, + "grad_norm": 0.6461196541786194, + "learning_rate": 0.0001999474806455803, + "loss": 0.1585, + "step": 450 + }, + { + "epoch": 0.010572150237052925, + "grad_norm": 2.498717784881592, + "learning_rate": 0.00019994724167098158, + "loss": 0.5469, + "step": 451 + }, + { + "epoch": 0.01059559181185792, + "grad_norm": 1.754331111907959, + "learning_rate": 0.00019994700215406633, + "loss": 0.4043, + "step": 452 + }, + { + "epoch": 0.010619033386662915, + "grad_norm": 2.0058767795562744, + "learning_rate": 0.00019994676209483583, + "loss": 1.784, + "step": 453 + }, + { + "epoch": 0.010642474961467911, + "grad_norm": 4.833071231842041, + "learning_rate": 0.0001999465214932914, + "loss": 0.9288, + "step": 454 + }, + { + "epoch": 0.010665916536272907, + "grad_norm": 3.095392942428589, + "learning_rate": 0.00019994628034943434, + "loss": 0.6431, + "step": 455 + }, + { + "epoch": 0.010689358111077902, + "grad_norm": 3.6067731380462646, + "learning_rate": 0.00019994603866326595, + "loss": 0.7499, + "step": 456 + }, + { + "epoch": 0.010712799685882898, + "grad_norm": 2.782862901687622, + "learning_rate": 0.00019994579643478754, + "loss": 0.623, + "step": 457 + }, + { + "epoch": 0.010736241260687894, + "grad_norm": 2.6268844604492188, + "learning_rate": 0.00019994555366400043, + "loss": 0.618, + "step": 458 + }, + { + "epoch": 0.010759682835492888, + "grad_norm": 3.5612568855285645, + "learning_rate": 0.00019994531035090596, + "loss": 0.9408, + "step": 459 + }, + { + "epoch": 0.010783124410297884, + "grad_norm": 3.5127878189086914, + "learning_rate": 0.00019994506649550537, + "loss": 0.7954, + "step": 460 + }, + { + "epoch": 0.010806565985102878, + "grad_norm": 3.4057395458221436, + "learning_rate": 0.00019994482209780007, + "loss": 0.9987, + "step": 461 + }, + { + "epoch": 0.010830007559907874, + "grad_norm": 3.1516847610473633, + "learning_rate": 0.0001999445771577914, + "loss": 0.7188, + "step": 462 + }, + { + "epoch": 0.01085344913471287, + "grad_norm": 3.005802869796753, + "learning_rate": 0.00019994433167548056, + "loss": 0.6873, + "step": 463 + }, + { + "epoch": 0.010876890709517865, + "grad_norm": 2.760329484939575, + "learning_rate": 0.00019994408565086902, + "loss": 0.5319, + "step": 464 + }, + { + "epoch": 0.01090033228432286, + "grad_norm": 1.8291417360305786, + "learning_rate": 0.00019994383908395803, + "loss": 0.3786, + "step": 465 + }, + { + "epoch": 0.010923773859127857, + "grad_norm": 2.725510597229004, + "learning_rate": 0.00019994359197474898, + "loss": 0.5097, + "step": 466 + }, + { + "epoch": 0.010947215433932851, + "grad_norm": 4.302855491638184, + "learning_rate": 0.0001999433443232432, + "loss": 0.9703, + "step": 467 + }, + { + "epoch": 0.010970657008737847, + "grad_norm": 4.010611057281494, + "learning_rate": 0.000199943096129442, + "loss": 1.1175, + "step": 468 + }, + { + "epoch": 0.010994098583542842, + "grad_norm": 2.5639309883117676, + "learning_rate": 0.00019994284739334675, + "loss": 0.6345, + "step": 469 + }, + { + "epoch": 0.011017540158347838, + "grad_norm": 0.9455622434616089, + "learning_rate": 0.0001999425981149588, + "loss": 0.1757, + "step": 470 + }, + { + "epoch": 0.011040981733152834, + "grad_norm": 4.0731635093688965, + "learning_rate": 0.00019994234829427953, + "loss": 0.5971, + "step": 471 + }, + { + "epoch": 0.011064423307957828, + "grad_norm": 1.3023326396942139, + "learning_rate": 0.00019994209793131024, + "loss": 0.2842, + "step": 472 + }, + { + "epoch": 0.011087864882762824, + "grad_norm": 4.436237335205078, + "learning_rate": 0.00019994184702605235, + "loss": 0.9734, + "step": 473 + }, + { + "epoch": 0.01111130645756782, + "grad_norm": 5.846012115478516, + "learning_rate": 0.00019994159557850717, + "loss": 1.0498, + "step": 474 + }, + { + "epoch": 0.011134748032372814, + "grad_norm": 3.259536027908325, + "learning_rate": 0.00019994134358867607, + "loss": 1.0118, + "step": 475 + }, + { + "epoch": 0.01115818960717781, + "grad_norm": 4.599364757537842, + "learning_rate": 0.0001999410910565604, + "loss": 0.9048, + "step": 476 + }, + { + "epoch": 0.011181631181982806, + "grad_norm": 0.9478965401649475, + "learning_rate": 0.00019994083798216162, + "loss": 0.1808, + "step": 477 + }, + { + "epoch": 0.0112050727567878, + "grad_norm": 1.1827657222747803, + "learning_rate": 0.00019994058436548099, + "loss": 0.2958, + "step": 478 + }, + { + "epoch": 0.011228514331592797, + "grad_norm": 3.024188756942749, + "learning_rate": 0.00019994033020652, + "loss": 0.8686, + "step": 479 + }, + { + "epoch": 0.011251955906397791, + "grad_norm": 3.9531002044677734, + "learning_rate": 0.00019994007550527994, + "loss": 0.9917, + "step": 480 + }, + { + "epoch": 0.011275397481202787, + "grad_norm": 2.3174214363098145, + "learning_rate": 0.0001999398202617622, + "loss": 0.5041, + "step": 481 + }, + { + "epoch": 0.011298839056007783, + "grad_norm": 0.6272441148757935, + "learning_rate": 0.0001999395644759682, + "loss": 0.0694, + "step": 482 + }, + { + "epoch": 0.011322280630812778, + "grad_norm": 3.077817440032959, + "learning_rate": 0.0001999393081478993, + "loss": 0.4793, + "step": 483 + }, + { + "epoch": 0.011345722205617774, + "grad_norm": 5.035027027130127, + "learning_rate": 0.00019993905127755693, + "loss": 1.1935, + "step": 484 + }, + { + "epoch": 0.01136916378042277, + "grad_norm": 3.1806089878082275, + "learning_rate": 0.00019993879386494246, + "loss": 0.5793, + "step": 485 + }, + { + "epoch": 0.011392605355227764, + "grad_norm": 1.5843921899795532, + "learning_rate": 0.00019993853591005728, + "loss": 0.6427, + "step": 486 + }, + { + "epoch": 0.01141604693003276, + "grad_norm": 1.617194652557373, + "learning_rate": 0.0001999382774129028, + "loss": 0.4775, + "step": 487 + }, + { + "epoch": 0.011439488504837754, + "grad_norm": 2.4136698246002197, + "learning_rate": 0.00019993801837348037, + "loss": 0.4977, + "step": 488 + }, + { + "epoch": 0.01146293007964275, + "grad_norm": 3.8453469276428223, + "learning_rate": 0.0001999377587917915, + "loss": 0.9284, + "step": 489 + }, + { + "epoch": 0.011486371654447746, + "grad_norm": 5.037114143371582, + "learning_rate": 0.00019993749866783749, + "loss": 1.4546, + "step": 490 + }, + { + "epoch": 0.01150981322925274, + "grad_norm": 2.6513617038726807, + "learning_rate": 0.00019993723800161984, + "loss": 0.6658, + "step": 491 + }, + { + "epoch": 0.011533254804057737, + "grad_norm": 2.8397130966186523, + "learning_rate": 0.0001999369767931399, + "loss": 1.0097, + "step": 492 + }, + { + "epoch": 0.011556696378862733, + "grad_norm": 3.495123863220215, + "learning_rate": 0.00019993671504239913, + "loss": 0.954, + "step": 493 + }, + { + "epoch": 0.011580137953667727, + "grad_norm": 2.204360008239746, + "learning_rate": 0.00019993645274939893, + "loss": 0.6809, + "step": 494 + }, + { + "epoch": 0.011603579528472723, + "grad_norm": 1.825942039489746, + "learning_rate": 0.0001999361899141407, + "loss": 0.4063, + "step": 495 + }, + { + "epoch": 0.011627021103277718, + "grad_norm": 1.2263531684875488, + "learning_rate": 0.00019993592653662595, + "loss": 0.2512, + "step": 496 + }, + { + "epoch": 0.011650462678082714, + "grad_norm": 2.916378974914551, + "learning_rate": 0.00019993566261685604, + "loss": 0.5516, + "step": 497 + }, + { + "epoch": 0.01167390425288771, + "grad_norm": 1.1778008937835693, + "learning_rate": 0.0001999353981548324, + "loss": 0.2192, + "step": 498 + }, + { + "epoch": 0.011697345827692704, + "grad_norm": 3.336054801940918, + "learning_rate": 0.0001999351331505565, + "loss": 0.7223, + "step": 499 + }, + { + "epoch": 0.0117207874024977, + "grad_norm": 0.313918799161911, + "learning_rate": 0.00019993486760402972, + "loss": 0.0577, + "step": 500 + }, + { + "epoch": 0.011744228977302696, + "grad_norm": 4.796902656555176, + "learning_rate": 0.00019993460151525354, + "loss": 0.9894, + "step": 501 + }, + { + "epoch": 0.01176767055210769, + "grad_norm": 2.943862199783325, + "learning_rate": 0.00019993433488422942, + "loss": 0.6837, + "step": 502 + }, + { + "epoch": 0.011791112126912686, + "grad_norm": 4.514648914337158, + "learning_rate": 0.00019993406771095877, + "loss": 1.0868, + "step": 503 + }, + { + "epoch": 0.01181455370171768, + "grad_norm": 3.6698572635650635, + "learning_rate": 0.0001999337999954431, + "loss": 0.4286, + "step": 504 + }, + { + "epoch": 0.011837995276522677, + "grad_norm": 3.773056745529175, + "learning_rate": 0.0001999335317376838, + "loss": 1.0717, + "step": 505 + }, + { + "epoch": 0.011861436851327673, + "grad_norm": 3.714493751525879, + "learning_rate": 0.00019993326293768233, + "loss": 1.1734, + "step": 506 + }, + { + "epoch": 0.011884878426132667, + "grad_norm": 2.969019651412964, + "learning_rate": 0.0001999329935954402, + "loss": 0.7456, + "step": 507 + }, + { + "epoch": 0.011908320000937663, + "grad_norm": 5.258603096008301, + "learning_rate": 0.0001999327237109588, + "loss": 1.1582, + "step": 508 + }, + { + "epoch": 0.01193176157574266, + "grad_norm": 2.90918231010437, + "learning_rate": 0.00019993245328423967, + "loss": 0.5225, + "step": 509 + }, + { + "epoch": 0.011955203150547654, + "grad_norm": 1.6693941354751587, + "learning_rate": 0.00019993218231528418, + "loss": 1.8258, + "step": 510 + }, + { + "epoch": 0.01197864472535265, + "grad_norm": 1.2220417261123657, + "learning_rate": 0.0001999319108040939, + "loss": 0.2347, + "step": 511 + }, + { + "epoch": 0.012002086300157644, + "grad_norm": 1.5614436864852905, + "learning_rate": 0.00019993163875067026, + "loss": 0.255, + "step": 512 + }, + { + "epoch": 0.01202552787496264, + "grad_norm": 3.1521832942962646, + "learning_rate": 0.00019993136615501476, + "loss": 0.562, + "step": 513 + }, + { + "epoch": 0.012048969449767636, + "grad_norm": 2.0295350551605225, + "learning_rate": 0.0001999310930171288, + "loss": 2.1667, + "step": 514 + }, + { + "epoch": 0.01207241102457263, + "grad_norm": 1.2131860256195068, + "learning_rate": 0.00019993081933701397, + "loss": 0.2799, + "step": 515 + }, + { + "epoch": 0.012095852599377626, + "grad_norm": 1.9515111446380615, + "learning_rate": 0.00019993054511467168, + "loss": 0.3485, + "step": 516 + }, + { + "epoch": 0.012119294174182622, + "grad_norm": 5.496387481689453, + "learning_rate": 0.00019993027035010346, + "loss": 1.3009, + "step": 517 + }, + { + "epoch": 0.012142735748987617, + "grad_norm": 5.023473739624023, + "learning_rate": 0.00019992999504331077, + "loss": 0.5764, + "step": 518 + }, + { + "epoch": 0.012166177323792613, + "grad_norm": 1.2322347164154053, + "learning_rate": 0.00019992971919429513, + "loss": 0.1264, + "step": 519 + }, + { + "epoch": 0.012189618898597607, + "grad_norm": 1.2068305015563965, + "learning_rate": 0.00019992944280305798, + "loss": 0.2655, + "step": 520 + }, + { + "epoch": 0.012213060473402603, + "grad_norm": 3.105640172958374, + "learning_rate": 0.00019992916586960093, + "loss": 0.8939, + "step": 521 + }, + { + "epoch": 0.0122365020482076, + "grad_norm": 1.5062041282653809, + "learning_rate": 0.0001999288883939254, + "loss": 1.5765, + "step": 522 + }, + { + "epoch": 0.012259943623012593, + "grad_norm": 3.2798221111297607, + "learning_rate": 0.00019992861037603285, + "loss": 0.8796, + "step": 523 + }, + { + "epoch": 0.01228338519781759, + "grad_norm": 5.979715824127197, + "learning_rate": 0.0001999283318159249, + "loss": 0.8775, + "step": 524 + }, + { + "epoch": 0.012306826772622586, + "grad_norm": 4.125465393066406, + "learning_rate": 0.00019992805271360303, + "loss": 0.8506, + "step": 525 + }, + { + "epoch": 0.01233026834742758, + "grad_norm": 3.546483278274536, + "learning_rate": 0.0001999277730690687, + "loss": 0.7906, + "step": 526 + }, + { + "epoch": 0.012353709922232576, + "grad_norm": 3.063926935195923, + "learning_rate": 0.00019992749288232347, + "loss": 0.7086, + "step": 527 + }, + { + "epoch": 0.01237715149703757, + "grad_norm": 2.945787191390991, + "learning_rate": 0.00019992721215336887, + "loss": 0.3382, + "step": 528 + }, + { + "epoch": 0.012400593071842566, + "grad_norm": 2.546720504760742, + "learning_rate": 0.0001999269308822064, + "loss": 0.3401, + "step": 529 + }, + { + "epoch": 0.012424034646647562, + "grad_norm": 3.1667747497558594, + "learning_rate": 0.0001999266490688376, + "loss": 0.9582, + "step": 530 + }, + { + "epoch": 0.012447476221452557, + "grad_norm": 0.4716927111148834, + "learning_rate": 0.00019992636671326399, + "loss": 0.0925, + "step": 531 + }, + { + "epoch": 0.012470917796257553, + "grad_norm": 3.1305224895477295, + "learning_rate": 0.00019992608381548709, + "loss": 0.5671, + "step": 532 + }, + { + "epoch": 0.012494359371062549, + "grad_norm": 1.8105392456054688, + "learning_rate": 0.00019992580037550846, + "loss": 1.7116, + "step": 533 + }, + { + "epoch": 0.012517800945867543, + "grad_norm": 3.3704044818878174, + "learning_rate": 0.00019992551639332962, + "loss": 0.6098, + "step": 534 + }, + { + "epoch": 0.012541242520672539, + "grad_norm": 1.0912072658538818, + "learning_rate": 0.00019992523186895215, + "loss": 1.0201, + "step": 535 + }, + { + "epoch": 0.012564684095477533, + "grad_norm": 1.4289659261703491, + "learning_rate": 0.00019992494680237753, + "loss": 0.3369, + "step": 536 + }, + { + "epoch": 0.01258812567028253, + "grad_norm": 2.856288194656372, + "learning_rate": 0.00019992466119360733, + "loss": 0.4903, + "step": 537 + }, + { + "epoch": 0.012611567245087526, + "grad_norm": 3.8211898803710938, + "learning_rate": 0.0001999243750426431, + "loss": 0.6433, + "step": 538 + }, + { + "epoch": 0.01263500881989252, + "grad_norm": 5.33319091796875, + "learning_rate": 0.00019992408834948645, + "loss": 1.1099, + "step": 539 + }, + { + "epoch": 0.012658450394697516, + "grad_norm": 3.478689193725586, + "learning_rate": 0.00019992380111413883, + "loss": 0.6785, + "step": 540 + }, + { + "epoch": 0.012681891969502512, + "grad_norm": 0.6068065166473389, + "learning_rate": 0.00019992351333660192, + "loss": 0.1249, + "step": 541 + }, + { + "epoch": 0.012705333544307506, + "grad_norm": 4.520905017852783, + "learning_rate": 0.00019992322501687716, + "loss": 0.8329, + "step": 542 + }, + { + "epoch": 0.012728775119112502, + "grad_norm": 3.999436140060425, + "learning_rate": 0.0001999229361549662, + "loss": 0.8801, + "step": 543 + }, + { + "epoch": 0.012752216693917497, + "grad_norm": 4.440465450286865, + "learning_rate": 0.00019992264675087058, + "loss": 1.0803, + "step": 544 + }, + { + "epoch": 0.012775658268722493, + "grad_norm": 3.4192161560058594, + "learning_rate": 0.00019992235680459187, + "loss": 0.8086, + "step": 545 + }, + { + "epoch": 0.012799099843527489, + "grad_norm": 2.9772372245788574, + "learning_rate": 0.00019992206631613165, + "loss": 0.4029, + "step": 546 + }, + { + "epoch": 0.012822541418332483, + "grad_norm": 0.5649701952934265, + "learning_rate": 0.00019992177528549147, + "loss": 0.0955, + "step": 547 + }, + { + "epoch": 0.012845982993137479, + "grad_norm": 4.782743453979492, + "learning_rate": 0.00019992148371267293, + "loss": 1.0976, + "step": 548 + }, + { + "epoch": 0.012869424567942475, + "grad_norm": 3.3168396949768066, + "learning_rate": 0.00019992119159767762, + "loss": 0.4073, + "step": 549 + }, + { + "epoch": 0.01289286614274747, + "grad_norm": 3.1493074893951416, + "learning_rate": 0.00019992089894050713, + "loss": 0.6862, + "step": 550 + }, + { + "epoch": 0.012916307717552465, + "grad_norm": 0.7274974584579468, + "learning_rate": 0.00019992060574116298, + "loss": 0.1949, + "step": 551 + }, + { + "epoch": 0.01293974929235746, + "grad_norm": 0.5701420307159424, + "learning_rate": 0.00019992031199964685, + "loss": 0.1032, + "step": 552 + }, + { + "epoch": 0.012963190867162456, + "grad_norm": 0.9744860529899597, + "learning_rate": 0.00019992001771596026, + "loss": 0.1677, + "step": 553 + }, + { + "epoch": 0.012986632441967452, + "grad_norm": 1.936788558959961, + "learning_rate": 0.0001999197228901049, + "loss": 1.7498, + "step": 554 + }, + { + "epoch": 0.013010074016772446, + "grad_norm": 3.3323562145233154, + "learning_rate": 0.00019991942752208228, + "loss": 0.7065, + "step": 555 + }, + { + "epoch": 0.013033515591577442, + "grad_norm": 0.7041716575622559, + "learning_rate": 0.00019991913161189405, + "loss": 0.1613, + "step": 556 + }, + { + "epoch": 0.013056957166382438, + "grad_norm": 3.531940460205078, + "learning_rate": 0.00019991883515954183, + "loss": 0.5985, + "step": 557 + }, + { + "epoch": 0.013080398741187433, + "grad_norm": 5.2083821296691895, + "learning_rate": 0.00019991853816502718, + "loss": 0.9339, + "step": 558 + }, + { + "epoch": 0.013103840315992429, + "grad_norm": 3.0778274536132812, + "learning_rate": 0.00019991824062835168, + "loss": 0.712, + "step": 559 + }, + { + "epoch": 0.013127281890797423, + "grad_norm": 3.846587657928467, + "learning_rate": 0.00019991794254951706, + "loss": 0.7783, + "step": 560 + }, + { + "epoch": 0.013150723465602419, + "grad_norm": 2.4819176197052, + "learning_rate": 0.00019991764392852484, + "loss": 1.84, + "step": 561 + }, + { + "epoch": 0.013174165040407415, + "grad_norm": 4.836848258972168, + "learning_rate": 0.00019991734476537668, + "loss": 1.0002, + "step": 562 + }, + { + "epoch": 0.01319760661521241, + "grad_norm": 3.7360944747924805, + "learning_rate": 0.00019991704506007422, + "loss": 0.7287, + "step": 563 + }, + { + "epoch": 0.013221048190017405, + "grad_norm": 3.159939765930176, + "learning_rate": 0.00019991674481261907, + "loss": 0.748, + "step": 564 + }, + { + "epoch": 0.013244489764822402, + "grad_norm": 4.175021648406982, + "learning_rate": 0.00019991644402301284, + "loss": 1.047, + "step": 565 + }, + { + "epoch": 0.013267931339627396, + "grad_norm": 0.6498627066612244, + "learning_rate": 0.00019991614269125715, + "loss": 0.1682, + "step": 566 + }, + { + "epoch": 0.013291372914432392, + "grad_norm": 2.41520619392395, + "learning_rate": 0.00019991584081735368, + "loss": 0.8274, + "step": 567 + }, + { + "epoch": 0.013314814489237386, + "grad_norm": 3.2510204315185547, + "learning_rate": 0.00019991553840130406, + "loss": 0.5543, + "step": 568 + }, + { + "epoch": 0.013338256064042382, + "grad_norm": 3.1883251667022705, + "learning_rate": 0.00019991523544310989, + "loss": 0.8455, + "step": 569 + }, + { + "epoch": 0.013361697638847378, + "grad_norm": 2.528167963027954, + "learning_rate": 0.00019991493194277285, + "loss": 0.508, + "step": 570 + }, + { + "epoch": 0.013385139213652373, + "grad_norm": 3.5883312225341797, + "learning_rate": 0.0001999146279002946, + "loss": 0.9408, + "step": 571 + }, + { + "epoch": 0.013408580788457369, + "grad_norm": 4.215970039367676, + "learning_rate": 0.00019991432331567676, + "loss": 1.0106, + "step": 572 + }, + { + "epoch": 0.013432022363262365, + "grad_norm": 1.3743406534194946, + "learning_rate": 0.000199914018188921, + "loss": 0.2267, + "step": 573 + }, + { + "epoch": 0.013455463938067359, + "grad_norm": 2.621593713760376, + "learning_rate": 0.00019991371252002895, + "loss": 0.4444, + "step": 574 + }, + { + "epoch": 0.013478905512872355, + "grad_norm": 2.5646424293518066, + "learning_rate": 0.0001999134063090023, + "loss": 0.3006, + "step": 575 + }, + { + "epoch": 0.01350234708767735, + "grad_norm": 3.321333885192871, + "learning_rate": 0.0001999130995558427, + "loss": 0.6203, + "step": 576 + }, + { + "epoch": 0.013525788662482345, + "grad_norm": 3.897853374481201, + "learning_rate": 0.0001999127922605518, + "loss": 1.0283, + "step": 577 + }, + { + "epoch": 0.013549230237287341, + "grad_norm": 1.3120944499969482, + "learning_rate": 0.0001999124844231313, + "loss": 0.3486, + "step": 578 + }, + { + "epoch": 0.013572671812092336, + "grad_norm": 2.608064889907837, + "learning_rate": 0.00019991217604358282, + "loss": 0.6452, + "step": 579 + }, + { + "epoch": 0.013596113386897332, + "grad_norm": 3.7784831523895264, + "learning_rate": 0.0001999118671219081, + "loss": 0.5914, + "step": 580 + }, + { + "epoch": 0.013619554961702328, + "grad_norm": 1.2419171333312988, + "learning_rate": 0.00019991155765810875, + "loss": 0.1869, + "step": 581 + }, + { + "epoch": 0.013642996536507322, + "grad_norm": 3.4880006313323975, + "learning_rate": 0.00019991124765218648, + "loss": 0.7356, + "step": 582 + }, + { + "epoch": 0.013666438111312318, + "grad_norm": 4.0604071617126465, + "learning_rate": 0.00019991093710414298, + "loss": 0.9566, + "step": 583 + }, + { + "epoch": 0.013689879686117313, + "grad_norm": 0.3017435073852539, + "learning_rate": 0.00019991062601397994, + "loss": 0.0772, + "step": 584 + }, + { + "epoch": 0.013713321260922309, + "grad_norm": 3.927288055419922, + "learning_rate": 0.00019991031438169903, + "loss": 1.1211, + "step": 585 + }, + { + "epoch": 0.013736762835727305, + "grad_norm": 0.7316640019416809, + "learning_rate": 0.00019991000220730194, + "loss": 0.1233, + "step": 586 + }, + { + "epoch": 0.013760204410532299, + "grad_norm": 2.5563440322875977, + "learning_rate": 0.00019990968949079038, + "loss": 0.4046, + "step": 587 + }, + { + "epoch": 0.013783645985337295, + "grad_norm": 2.7482974529266357, + "learning_rate": 0.00019990937623216599, + "loss": 0.4226, + "step": 588 + }, + { + "epoch": 0.013807087560142291, + "grad_norm": 1.7107075452804565, + "learning_rate": 0.00019990906243143056, + "loss": 0.3498, + "step": 589 + }, + { + "epoch": 0.013830529134947285, + "grad_norm": 5.233372688293457, + "learning_rate": 0.00019990874808858573, + "loss": 1.1762, + "step": 590 + }, + { + "epoch": 0.013853970709752281, + "grad_norm": 1.1180962324142456, + "learning_rate": 0.00019990843320363323, + "loss": 0.2372, + "step": 591 + }, + { + "epoch": 0.013877412284557276, + "grad_norm": 3.4014084339141846, + "learning_rate": 0.00019990811777657473, + "loss": 0.5447, + "step": 592 + }, + { + "epoch": 0.013900853859362272, + "grad_norm": 1.2476873397827148, + "learning_rate": 0.00019990780180741203, + "loss": 1.339, + "step": 593 + }, + { + "epoch": 0.013924295434167268, + "grad_norm": 3.0703413486480713, + "learning_rate": 0.00019990748529614676, + "loss": 0.6157, + "step": 594 + }, + { + "epoch": 0.013947737008972262, + "grad_norm": 0.6618407964706421, + "learning_rate": 0.00019990716824278066, + "loss": 0.2527, + "step": 595 + }, + { + "epoch": 0.013971178583777258, + "grad_norm": 3.4015214443206787, + "learning_rate": 0.00019990685064731547, + "loss": 0.6687, + "step": 596 + }, + { + "epoch": 0.013994620158582254, + "grad_norm": 3.0110697746276855, + "learning_rate": 0.0001999065325097529, + "loss": 0.6829, + "step": 597 + }, + { + "epoch": 0.014018061733387249, + "grad_norm": 1.4814236164093018, + "learning_rate": 0.00019990621383009466, + "loss": 0.3646, + "step": 598 + }, + { + "epoch": 0.014041503308192245, + "grad_norm": 2.844437837600708, + "learning_rate": 0.00019990589460834252, + "loss": 0.5932, + "step": 599 + }, + { + "epoch": 0.014064944882997239, + "grad_norm": 2.929882526397705, + "learning_rate": 0.00019990557484449815, + "loss": 0.6998, + "step": 600 + }, + { + "epoch": 0.014088386457802235, + "grad_norm": 3.019639492034912, + "learning_rate": 0.00019990525453856335, + "loss": 0.4748, + "step": 601 + }, + { + "epoch": 0.014111828032607231, + "grad_norm": 3.689779043197632, + "learning_rate": 0.00019990493369053983, + "loss": 0.5312, + "step": 602 + }, + { + "epoch": 0.014135269607412225, + "grad_norm": 4.7770819664001465, + "learning_rate": 0.0001999046123004293, + "loss": 0.8734, + "step": 603 + }, + { + "epoch": 0.014158711182217221, + "grad_norm": 2.3012535572052, + "learning_rate": 0.00019990429036823354, + "loss": 0.5477, + "step": 604 + }, + { + "epoch": 0.014182152757022217, + "grad_norm": 2.0262372493743896, + "learning_rate": 0.0001999039678939543, + "loss": 0.4582, + "step": 605 + }, + { + "epoch": 0.014205594331827212, + "grad_norm": 4.538377285003662, + "learning_rate": 0.00019990364487759336, + "loss": 0.7588, + "step": 606 + }, + { + "epoch": 0.014229035906632208, + "grad_norm": 0.5579646825790405, + "learning_rate": 0.00019990332131915238, + "loss": 0.1296, + "step": 607 + }, + { + "epoch": 0.014252477481437202, + "grad_norm": 2.841057538986206, + "learning_rate": 0.0001999029972186332, + "loss": 1.824, + "step": 608 + }, + { + "epoch": 0.014275919056242198, + "grad_norm": 3.593567371368408, + "learning_rate": 0.00019990267257603757, + "loss": 0.7655, + "step": 609 + }, + { + "epoch": 0.014299360631047194, + "grad_norm": 3.569187879562378, + "learning_rate": 0.00019990234739136718, + "loss": 0.7247, + "step": 610 + }, + { + "epoch": 0.014322802205852189, + "grad_norm": 3.6679208278656006, + "learning_rate": 0.00019990202166462386, + "loss": 0.8847, + "step": 611 + }, + { + "epoch": 0.014346243780657185, + "grad_norm": 1.0573006868362427, + "learning_rate": 0.00019990169539580937, + "loss": 0.2201, + "step": 612 + }, + { + "epoch": 0.01436968535546218, + "grad_norm": 4.616801738739014, + "learning_rate": 0.00019990136858492546, + "loss": 0.8586, + "step": 613 + }, + { + "epoch": 0.014393126930267175, + "grad_norm": 3.077793836593628, + "learning_rate": 0.00019990104123197394, + "loss": 0.5088, + "step": 614 + }, + { + "epoch": 0.014416568505072171, + "grad_norm": 0.8590427041053772, + "learning_rate": 0.00019990071333695654, + "loss": 0.2035, + "step": 615 + }, + { + "epoch": 0.014440010079877165, + "grad_norm": 1.120387315750122, + "learning_rate": 0.00019990038489987505, + "loss": 0.1863, + "step": 616 + }, + { + "epoch": 0.014463451654682161, + "grad_norm": 1.4108237028121948, + "learning_rate": 0.00019990005592073127, + "loss": 0.3212, + "step": 617 + }, + { + "epoch": 0.014486893229487157, + "grad_norm": 2.804150342941284, + "learning_rate": 0.000199899726399527, + "loss": 0.6746, + "step": 618 + }, + { + "epoch": 0.014510334804292152, + "grad_norm": 3.7612600326538086, + "learning_rate": 0.000199899396336264, + "loss": 0.7748, + "step": 619 + }, + { + "epoch": 0.014533776379097148, + "grad_norm": 2.723576784133911, + "learning_rate": 0.00019989906573094404, + "loss": 0.594, + "step": 620 + }, + { + "epoch": 0.014557217953902144, + "grad_norm": 1.4310462474822998, + "learning_rate": 0.00019989873458356897, + "loss": 0.4281, + "step": 621 + }, + { + "epoch": 0.014580659528707138, + "grad_norm": 3.791529893875122, + "learning_rate": 0.00019989840289414055, + "loss": 0.6812, + "step": 622 + }, + { + "epoch": 0.014604101103512134, + "grad_norm": 1.3598430156707764, + "learning_rate": 0.00019989807066266057, + "loss": 0.3669, + "step": 623 + }, + { + "epoch": 0.014627542678317128, + "grad_norm": 2.9622597694396973, + "learning_rate": 0.00019989773788913086, + "loss": 0.6834, + "step": 624 + }, + { + "epoch": 0.014650984253122125, + "grad_norm": 2.85011625289917, + "learning_rate": 0.00019989740457355322, + "loss": 0.6088, + "step": 625 + }, + { + "epoch": 0.01467442582792712, + "grad_norm": 1.0705770254135132, + "learning_rate": 0.00019989707071592944, + "loss": 0.2614, + "step": 626 + }, + { + "epoch": 0.014697867402732115, + "grad_norm": 2.4120821952819824, + "learning_rate": 0.00019989673631626135, + "loss": 1.5968, + "step": 627 + }, + { + "epoch": 0.014721308977537111, + "grad_norm": 1.8417965173721313, + "learning_rate": 0.00019989640137455075, + "loss": 1.6297, + "step": 628 + }, + { + "epoch": 0.014744750552342107, + "grad_norm": 4.181495666503906, + "learning_rate": 0.0001998960658907995, + "loss": 0.8279, + "step": 629 + }, + { + "epoch": 0.014768192127147101, + "grad_norm": 2.548844337463379, + "learning_rate": 0.00019989572986500937, + "loss": 0.3643, + "step": 630 + }, + { + "epoch": 0.014791633701952097, + "grad_norm": 4.202001571655273, + "learning_rate": 0.0001998953932971822, + "loss": 0.7979, + "step": 631 + }, + { + "epoch": 0.014815075276757093, + "grad_norm": 2.655012369155884, + "learning_rate": 0.0001998950561873198, + "loss": 0.4193, + "step": 632 + }, + { + "epoch": 0.014838516851562088, + "grad_norm": 3.3833248615264893, + "learning_rate": 0.00019989471853542405, + "loss": 0.3016, + "step": 633 + }, + { + "epoch": 0.014861958426367084, + "grad_norm": 4.390758991241455, + "learning_rate": 0.00019989438034149674, + "loss": 0.9745, + "step": 634 + }, + { + "epoch": 0.014885400001172078, + "grad_norm": 3.5260748863220215, + "learning_rate": 0.0001998940416055397, + "loss": 0.4737, + "step": 635 + }, + { + "epoch": 0.014908841575977074, + "grad_norm": 2.233731269836426, + "learning_rate": 0.0001998937023275548, + "loss": 1.4068, + "step": 636 + }, + { + "epoch": 0.01493228315078207, + "grad_norm": 2.12404727935791, + "learning_rate": 0.00019989336250754385, + "loss": 0.4931, + "step": 637 + }, + { + "epoch": 0.014955724725587064, + "grad_norm": 5.251610279083252, + "learning_rate": 0.00019989302214550869, + "loss": 0.7909, + "step": 638 + }, + { + "epoch": 0.01497916630039206, + "grad_norm": 2.0279910564422607, + "learning_rate": 0.0001998926812414512, + "loss": 0.2835, + "step": 639 + }, + { + "epoch": 0.015002607875197057, + "grad_norm": 1.4447897672653198, + "learning_rate": 0.00019989233979537322, + "loss": 0.1731, + "step": 640 + }, + { + "epoch": 0.015026049450002051, + "grad_norm": 3.500123977661133, + "learning_rate": 0.00019989199780727658, + "loss": 0.5636, + "step": 641 + }, + { + "epoch": 0.015049491024807047, + "grad_norm": 3.6843371391296387, + "learning_rate": 0.00019989165527716315, + "loss": 0.899, + "step": 642 + }, + { + "epoch": 0.015072932599612041, + "grad_norm": 4.509810924530029, + "learning_rate": 0.0001998913122050348, + "loss": 1.0037, + "step": 643 + }, + { + "epoch": 0.015096374174417037, + "grad_norm": 3.773719549179077, + "learning_rate": 0.0001998909685908934, + "loss": 1.0076, + "step": 644 + }, + { + "epoch": 0.015119815749222033, + "grad_norm": 0.9396259784698486, + "learning_rate": 0.00019989062443474077, + "loss": 0.246, + "step": 645 + }, + { + "epoch": 0.015143257324027028, + "grad_norm": 2.532317876815796, + "learning_rate": 0.0001998902797365788, + "loss": 0.4925, + "step": 646 + }, + { + "epoch": 0.015166698898832024, + "grad_norm": 3.6001460552215576, + "learning_rate": 0.0001998899344964094, + "loss": 0.743, + "step": 647 + }, + { + "epoch": 0.01519014047363702, + "grad_norm": 1.993477702140808, + "learning_rate": 0.00019988958871423437, + "loss": 1.5772, + "step": 648 + }, + { + "epoch": 0.015213582048442014, + "grad_norm": 3.3430747985839844, + "learning_rate": 0.00019988924239005561, + "loss": 0.6526, + "step": 649 + }, + { + "epoch": 0.01523702362324701, + "grad_norm": 1.809601902961731, + "learning_rate": 0.00019988889552387503, + "loss": 1.3646, + "step": 650 + }, + { + "epoch": 0.015260465198052004, + "grad_norm": 5.554145812988281, + "learning_rate": 0.00019988854811569452, + "loss": 0.9589, + "step": 651 + }, + { + "epoch": 0.015283906772857, + "grad_norm": 1.333693504333496, + "learning_rate": 0.00019988820016551592, + "loss": 0.3671, + "step": 652 + }, + { + "epoch": 0.015307348347661997, + "grad_norm": 1.835728645324707, + "learning_rate": 0.00019988785167334114, + "loss": 0.2836, + "step": 653 + }, + { + "epoch": 0.01533078992246699, + "grad_norm": 0.4738597571849823, + "learning_rate": 0.00019988750263917206, + "loss": 0.0859, + "step": 654 + }, + { + "epoch": 0.015354231497271987, + "grad_norm": 1.7263009548187256, + "learning_rate": 0.0001998871530630106, + "loss": 1.6454, + "step": 655 + }, + { + "epoch": 0.015377673072076983, + "grad_norm": 3.7225189208984375, + "learning_rate": 0.0001998868029448586, + "loss": 0.7189, + "step": 656 + }, + { + "epoch": 0.015401114646881977, + "grad_norm": 2.649412155151367, + "learning_rate": 0.00019988645228471806, + "loss": 0.6926, + "step": 657 + }, + { + "epoch": 0.015424556221686973, + "grad_norm": 3.538891553878784, + "learning_rate": 0.00019988610108259076, + "loss": 0.7428, + "step": 658 + }, + { + "epoch": 0.015447997796491968, + "grad_norm": 2.1686034202575684, + "learning_rate": 0.0001998857493384787, + "loss": 0.4082, + "step": 659 + }, + { + "epoch": 0.015471439371296964, + "grad_norm": 2.3826324939727783, + "learning_rate": 0.00019988539705238374, + "loss": 0.3868, + "step": 660 + }, + { + "epoch": 0.01549488094610196, + "grad_norm": 3.574753999710083, + "learning_rate": 0.00019988504422430782, + "loss": 0.9349, + "step": 661 + }, + { + "epoch": 0.015518322520906954, + "grad_norm": 4.993885040283203, + "learning_rate": 0.00019988469085425282, + "loss": 0.9787, + "step": 662 + }, + { + "epoch": 0.01554176409571195, + "grad_norm": 0.5984430909156799, + "learning_rate": 0.0001998843369422207, + "loss": 0.1627, + "step": 663 + }, + { + "epoch": 0.015565205670516946, + "grad_norm": 3.687044382095337, + "learning_rate": 0.00019988398248821335, + "loss": 0.5173, + "step": 664 + }, + { + "epoch": 0.01558864724532194, + "grad_norm": 7.957596778869629, + "learning_rate": 0.00019988362749223272, + "loss": 0.4075, + "step": 665 + }, + { + "epoch": 0.015612088820126937, + "grad_norm": 1.0090070962905884, + "learning_rate": 0.0001998832719542807, + "loss": 0.1958, + "step": 666 + }, + { + "epoch": 0.015635530394931933, + "grad_norm": 4.64321756362915, + "learning_rate": 0.00019988291587435923, + "loss": 1.3407, + "step": 667 + }, + { + "epoch": 0.015658971969736927, + "grad_norm": 3.5671229362487793, + "learning_rate": 0.00019988255925247028, + "loss": 1.017, + "step": 668 + }, + { + "epoch": 0.01568241354454192, + "grad_norm": 1.4052520990371704, + "learning_rate": 0.00019988220208861575, + "loss": 0.2485, + "step": 669 + }, + { + "epoch": 0.01570585511934692, + "grad_norm": 2.3216302394866943, + "learning_rate": 0.00019988184438279755, + "loss": 0.4004, + "step": 670 + }, + { + "epoch": 0.015729296694151913, + "grad_norm": 3.951204538345337, + "learning_rate": 0.0001998814861350177, + "loss": 0.9353, + "step": 671 + }, + { + "epoch": 0.015752738268956908, + "grad_norm": 2.7239181995391846, + "learning_rate": 0.00019988112734527805, + "loss": 0.3729, + "step": 672 + }, + { + "epoch": 0.015776179843761905, + "grad_norm": 0.6177278161048889, + "learning_rate": 0.00019988076801358062, + "loss": 0.1221, + "step": 673 + }, + { + "epoch": 0.0157996214185669, + "grad_norm": 2.6318891048431396, + "learning_rate": 0.00019988040813992734, + "loss": 0.5142, + "step": 674 + }, + { + "epoch": 0.015823062993371894, + "grad_norm": 4.524494171142578, + "learning_rate": 0.00019988004772432015, + "loss": 1.2805, + "step": 675 + }, + { + "epoch": 0.015846504568176892, + "grad_norm": 3.153536081314087, + "learning_rate": 0.000199879686766761, + "loss": 0.8438, + "step": 676 + }, + { + "epoch": 0.015869946142981886, + "grad_norm": 3.5087735652923584, + "learning_rate": 0.00019987932526725188, + "loss": 0.68, + "step": 677 + }, + { + "epoch": 0.01589338771778688, + "grad_norm": 4.172877311706543, + "learning_rate": 0.00019987896322579474, + "loss": 0.8095, + "step": 678 + }, + { + "epoch": 0.015916829292591875, + "grad_norm": 2.450995922088623, + "learning_rate": 0.00019987860064239152, + "loss": 0.5069, + "step": 679 + }, + { + "epoch": 0.015940270867396873, + "grad_norm": 1.9025880098342896, + "learning_rate": 0.00019987823751704419, + "loss": 0.3396, + "step": 680 + }, + { + "epoch": 0.015963712442201867, + "grad_norm": 1.0581085681915283, + "learning_rate": 0.00019987787384975476, + "loss": 0.7311, + "step": 681 + }, + { + "epoch": 0.01598715401700686, + "grad_norm": 4.0336127281188965, + "learning_rate": 0.00019987750964052517, + "loss": 0.7315, + "step": 682 + }, + { + "epoch": 0.01601059559181186, + "grad_norm": 0.6190354228019714, + "learning_rate": 0.0001998771448893574, + "loss": 0.1232, + "step": 683 + }, + { + "epoch": 0.016034037166616853, + "grad_norm": 2.631061553955078, + "learning_rate": 0.00019987677959625345, + "loss": 0.5854, + "step": 684 + }, + { + "epoch": 0.016057478741421848, + "grad_norm": 4.337430477142334, + "learning_rate": 0.00019987641376121527, + "loss": 0.7032, + "step": 685 + }, + { + "epoch": 0.016080920316226845, + "grad_norm": 3.4115757942199707, + "learning_rate": 0.00019987604738424487, + "loss": 0.828, + "step": 686 + }, + { + "epoch": 0.01610436189103184, + "grad_norm": 0.7829211354255676, + "learning_rate": 0.00019987568046534422, + "loss": 0.1556, + "step": 687 + }, + { + "epoch": 0.016127803465836834, + "grad_norm": 0.9408995509147644, + "learning_rate": 0.0001998753130045153, + "loss": 0.1574, + "step": 688 + }, + { + "epoch": 0.016151245040641832, + "grad_norm": 4.582515239715576, + "learning_rate": 0.00019987494500176017, + "loss": 0.5379, + "step": 689 + }, + { + "epoch": 0.016174686615446826, + "grad_norm": 5.389693260192871, + "learning_rate": 0.00019987457645708078, + "loss": 0.9874, + "step": 690 + }, + { + "epoch": 0.01619812819025182, + "grad_norm": 2.9044694900512695, + "learning_rate": 0.00019987420737047908, + "loss": 0.7557, + "step": 691 + }, + { + "epoch": 0.016221569765056818, + "grad_norm": 4.0924906730651855, + "learning_rate": 0.00019987383774195715, + "loss": 0.6762, + "step": 692 + }, + { + "epoch": 0.016245011339861812, + "grad_norm": 4.755887031555176, + "learning_rate": 0.00019987346757151695, + "loss": 0.8461, + "step": 693 + }, + { + "epoch": 0.016268452914666807, + "grad_norm": 1.0171352624893188, + "learning_rate": 0.00019987309685916055, + "loss": 0.1608, + "step": 694 + }, + { + "epoch": 0.0162918944894718, + "grad_norm": 1.145023226737976, + "learning_rate": 0.00019987272560488987, + "loss": 0.1792, + "step": 695 + }, + { + "epoch": 0.0163153360642768, + "grad_norm": 3.3476510047912598, + "learning_rate": 0.00019987235380870702, + "loss": 0.7185, + "step": 696 + }, + { + "epoch": 0.016338777639081793, + "grad_norm": 5.01161003112793, + "learning_rate": 0.0001998719814706139, + "loss": 0.8417, + "step": 697 + }, + { + "epoch": 0.016362219213886788, + "grad_norm": 3.0364768505096436, + "learning_rate": 0.00019987160859061268, + "loss": 0.5468, + "step": 698 + }, + { + "epoch": 0.016385660788691785, + "grad_norm": 4.566936016082764, + "learning_rate": 0.00019987123516870527, + "loss": 0.811, + "step": 699 + }, + { + "epoch": 0.01640910236349678, + "grad_norm": 1.4283720254898071, + "learning_rate": 0.00019987086120489372, + "loss": 0.3476, + "step": 700 + }, + { + "epoch": 0.016432543938301774, + "grad_norm": 3.379831075668335, + "learning_rate": 0.00019987048669918006, + "loss": 0.9198, + "step": 701 + }, + { + "epoch": 0.01645598551310677, + "grad_norm": 3.4715847969055176, + "learning_rate": 0.00019987011165156638, + "loss": 0.3886, + "step": 702 + }, + { + "epoch": 0.016479427087911766, + "grad_norm": 2.9008970260620117, + "learning_rate": 0.00019986973606205463, + "loss": 0.5071, + "step": 703 + }, + { + "epoch": 0.01650286866271676, + "grad_norm": 3.033881664276123, + "learning_rate": 0.00019986935993064687, + "loss": 0.64, + "step": 704 + }, + { + "epoch": 0.016526310237521758, + "grad_norm": 0.9203780293464661, + "learning_rate": 0.00019986898325734516, + "loss": 0.1461, + "step": 705 + }, + { + "epoch": 0.016549751812326752, + "grad_norm": 1.1221818923950195, + "learning_rate": 0.00019986860604215157, + "loss": 0.2054, + "step": 706 + }, + { + "epoch": 0.016573193387131747, + "grad_norm": 2.2006046772003174, + "learning_rate": 0.0001998682282850681, + "loss": 1.7157, + "step": 707 + }, + { + "epoch": 0.016596634961936745, + "grad_norm": 4.010159015655518, + "learning_rate": 0.0001998678499860968, + "loss": 0.7209, + "step": 708 + }, + { + "epoch": 0.01662007653674174, + "grad_norm": 1.5641727447509766, + "learning_rate": 0.0001998674711452398, + "loss": 0.3582, + "step": 709 + }, + { + "epoch": 0.016643518111546733, + "grad_norm": 2.5831472873687744, + "learning_rate": 0.000199867091762499, + "loss": 0.7199, + "step": 710 + }, + { + "epoch": 0.016666959686351727, + "grad_norm": 1.3065968751907349, + "learning_rate": 0.0001998667118378766, + "loss": 0.2256, + "step": 711 + }, + { + "epoch": 0.016690401261156725, + "grad_norm": 4.9702653884887695, + "learning_rate": 0.00019986633137137464, + "loss": 0.9025, + "step": 712 + }, + { + "epoch": 0.01671384283596172, + "grad_norm": 4.717034816741943, + "learning_rate": 0.0001998659503629951, + "loss": 0.7251, + "step": 713 + }, + { + "epoch": 0.016737284410766714, + "grad_norm": 4.004701614379883, + "learning_rate": 0.00019986556881274015, + "loss": 0.719, + "step": 714 + }, + { + "epoch": 0.01676072598557171, + "grad_norm": 3.1529784202575684, + "learning_rate": 0.0001998651867206118, + "loss": 0.4704, + "step": 715 + }, + { + "epoch": 0.016784167560376706, + "grad_norm": 1.4435337781906128, + "learning_rate": 0.00019986480408661214, + "loss": 0.1874, + "step": 716 + }, + { + "epoch": 0.0168076091351817, + "grad_norm": 4.5064377784729, + "learning_rate": 0.00019986442091074325, + "loss": 0.8317, + "step": 717 + }, + { + "epoch": 0.016831050709986698, + "grad_norm": 4.12451696395874, + "learning_rate": 0.00019986403719300717, + "loss": 0.9524, + "step": 718 + }, + { + "epoch": 0.016854492284791692, + "grad_norm": 3.4294772148132324, + "learning_rate": 0.00019986365293340607, + "loss": 0.4872, + "step": 719 + }, + { + "epoch": 0.016877933859596687, + "grad_norm": 3.2497715950012207, + "learning_rate": 0.00019986326813194194, + "loss": 0.297, + "step": 720 + }, + { + "epoch": 0.016901375434401684, + "grad_norm": 1.0954872369766235, + "learning_rate": 0.00019986288278861692, + "loss": 0.2138, + "step": 721 + }, + { + "epoch": 0.01692481700920668, + "grad_norm": 3.7556207180023193, + "learning_rate": 0.0001998624969034331, + "loss": 0.6684, + "step": 722 + }, + { + "epoch": 0.016948258584011673, + "grad_norm": 0.8587589263916016, + "learning_rate": 0.00019986211047639255, + "loss": 0.1848, + "step": 723 + }, + { + "epoch": 0.01697170015881667, + "grad_norm": 3.209831476211548, + "learning_rate": 0.00019986172350749742, + "loss": 0.9412, + "step": 724 + }, + { + "epoch": 0.016995141733621665, + "grad_norm": 3.8170716762542725, + "learning_rate": 0.00019986133599674973, + "loss": 0.7744, + "step": 725 + }, + { + "epoch": 0.01701858330842666, + "grad_norm": 2.695244550704956, + "learning_rate": 0.00019986094794415162, + "loss": 0.4124, + "step": 726 + }, + { + "epoch": 0.017042024883231654, + "grad_norm": 2.4596610069274902, + "learning_rate": 0.0001998605593497052, + "loss": 0.6367, + "step": 727 + }, + { + "epoch": 0.01706546645803665, + "grad_norm": 3.1405842304229736, + "learning_rate": 0.0001998601702134126, + "loss": 1.2391, + "step": 728 + }, + { + "epoch": 0.017088908032841646, + "grad_norm": 3.4034948348999023, + "learning_rate": 0.0001998597805352759, + "loss": 0.5992, + "step": 729 + }, + { + "epoch": 0.01711234960764664, + "grad_norm": 3.010585308074951, + "learning_rate": 0.0001998593903152972, + "loss": 0.6458, + "step": 730 + }, + { + "epoch": 0.017135791182451638, + "grad_norm": 0.6958599090576172, + "learning_rate": 0.00019985899955347867, + "loss": 0.0498, + "step": 731 + }, + { + "epoch": 0.017159232757256632, + "grad_norm": 6.610971450805664, + "learning_rate": 0.00019985860824982239, + "loss": 1.6412, + "step": 732 + }, + { + "epoch": 0.017182674332061627, + "grad_norm": 3.3235816955566406, + "learning_rate": 0.00019985821640433051, + "loss": 0.4703, + "step": 733 + }, + { + "epoch": 0.017206115906866624, + "grad_norm": 3.471195697784424, + "learning_rate": 0.00019985782401700514, + "loss": 0.6565, + "step": 734 + }, + { + "epoch": 0.01722955748167162, + "grad_norm": 3.678927421569824, + "learning_rate": 0.00019985743108784837, + "loss": 0.6479, + "step": 735 + }, + { + "epoch": 0.017252999056476613, + "grad_norm": 3.606964349746704, + "learning_rate": 0.00019985703761686238, + "loss": 0.7158, + "step": 736 + }, + { + "epoch": 0.01727644063128161, + "grad_norm": 1.1749365329742432, + "learning_rate": 0.00019985664360404933, + "loss": 0.3249, + "step": 737 + }, + { + "epoch": 0.017299882206086605, + "grad_norm": 4.911717891693115, + "learning_rate": 0.00019985624904941132, + "loss": 0.8851, + "step": 738 + }, + { + "epoch": 0.0173233237808916, + "grad_norm": 2.921438455581665, + "learning_rate": 0.00019985585395295048, + "loss": 0.5533, + "step": 739 + }, + { + "epoch": 0.017346765355696597, + "grad_norm": 2.763897657394409, + "learning_rate": 0.00019985545831466896, + "loss": 0.6724, + "step": 740 + }, + { + "epoch": 0.01737020693050159, + "grad_norm": 3.197723865509033, + "learning_rate": 0.00019985506213456892, + "loss": 0.6156, + "step": 741 + }, + { + "epoch": 0.017393648505306586, + "grad_norm": 4.254197120666504, + "learning_rate": 0.0001998546654126525, + "loss": 1.0161, + "step": 742 + }, + { + "epoch": 0.01741709008011158, + "grad_norm": 4.020381927490234, + "learning_rate": 0.00019985426814892186, + "loss": 0.6435, + "step": 743 + }, + { + "epoch": 0.017440531654916578, + "grad_norm": 5.111537933349609, + "learning_rate": 0.00019985387034337918, + "loss": 0.8873, + "step": 744 + }, + { + "epoch": 0.017463973229721572, + "grad_norm": 3.676116943359375, + "learning_rate": 0.00019985347199602657, + "loss": 0.719, + "step": 745 + }, + { + "epoch": 0.017487414804526567, + "grad_norm": 0.9108142256736755, + "learning_rate": 0.00019985307310686624, + "loss": 0.1481, + "step": 746 + }, + { + "epoch": 0.017510856379331564, + "grad_norm": 4.146887302398682, + "learning_rate": 0.00019985267367590029, + "loss": 1.0344, + "step": 747 + }, + { + "epoch": 0.01753429795413656, + "grad_norm": 2.666170597076416, + "learning_rate": 0.00019985227370313094, + "loss": 0.674, + "step": 748 + }, + { + "epoch": 0.017557739528941553, + "grad_norm": 4.2514472007751465, + "learning_rate": 0.00019985187318856032, + "loss": 0.9445, + "step": 749 + }, + { + "epoch": 0.01758118110374655, + "grad_norm": 2.95397686958313, + "learning_rate": 0.00019985147213219068, + "loss": 0.5851, + "step": 750 + }, + { + "epoch": 0.017604622678551545, + "grad_norm": 0.25846773386001587, + "learning_rate": 0.0001998510705340241, + "loss": 0.0456, + "step": 751 + }, + { + "epoch": 0.01762806425335654, + "grad_norm": 1.9414712190628052, + "learning_rate": 0.00019985066839406281, + "loss": 0.2808, + "step": 752 + }, + { + "epoch": 0.017651505828161537, + "grad_norm": 3.838470220565796, + "learning_rate": 0.00019985026571230901, + "loss": 0.8485, + "step": 753 + }, + { + "epoch": 0.01767494740296653, + "grad_norm": 3.662252902984619, + "learning_rate": 0.00019984986248876486, + "loss": 0.8645, + "step": 754 + }, + { + "epoch": 0.017698388977771526, + "grad_norm": 3.2335212230682373, + "learning_rate": 0.00019984945872343252, + "loss": 0.7914, + "step": 755 + }, + { + "epoch": 0.017721830552576524, + "grad_norm": 1.1272482872009277, + "learning_rate": 0.00019984905441631422, + "loss": 0.2041, + "step": 756 + }, + { + "epoch": 0.017745272127381518, + "grad_norm": 1.0321180820465088, + "learning_rate": 0.00019984864956741215, + "loss": 0.2167, + "step": 757 + }, + { + "epoch": 0.017768713702186512, + "grad_norm": 5.500738620758057, + "learning_rate": 0.0001998482441767285, + "loss": 1.1636, + "step": 758 + }, + { + "epoch": 0.017792155276991507, + "grad_norm": 1.751926064491272, + "learning_rate": 0.0001998478382442655, + "loss": 0.3346, + "step": 759 + }, + { + "epoch": 0.017815596851796504, + "grad_norm": 1.795303225517273, + "learning_rate": 0.00019984743177002526, + "loss": 1.5487, + "step": 760 + }, + { + "epoch": 0.0178390384266015, + "grad_norm": 2.918677568435669, + "learning_rate": 0.00019984702475401008, + "loss": 0.6758, + "step": 761 + }, + { + "epoch": 0.017862480001406493, + "grad_norm": 1.357708215713501, + "learning_rate": 0.00019984661719622216, + "loss": 0.4075, + "step": 762 + }, + { + "epoch": 0.01788592157621149, + "grad_norm": 0.6137160658836365, + "learning_rate": 0.00019984620909666365, + "loss": 0.1053, + "step": 763 + }, + { + "epoch": 0.017909363151016485, + "grad_norm": 0.819684624671936, + "learning_rate": 0.00019984580045533681, + "loss": 0.1623, + "step": 764 + }, + { + "epoch": 0.01793280472582148, + "grad_norm": 1.7066227197647095, + "learning_rate": 0.00019984539127224386, + "loss": 1.3672, + "step": 765 + }, + { + "epoch": 0.017956246300626477, + "grad_norm": 3.544057607650757, + "learning_rate": 0.000199844981547387, + "loss": 0.5546, + "step": 766 + }, + { + "epoch": 0.01797968787543147, + "grad_norm": 0.9549791812896729, + "learning_rate": 0.00019984457128076848, + "loss": 0.177, + "step": 767 + }, + { + "epoch": 0.018003129450236466, + "grad_norm": 3.960338592529297, + "learning_rate": 0.00019984416047239047, + "loss": 1.1116, + "step": 768 + }, + { + "epoch": 0.018026571025041464, + "grad_norm": 3.8108787536621094, + "learning_rate": 0.00019984374912225525, + "loss": 0.5797, + "step": 769 + }, + { + "epoch": 0.018050012599846458, + "grad_norm": 2.3419196605682373, + "learning_rate": 0.00019984333723036506, + "loss": 0.3532, + "step": 770 + }, + { + "epoch": 0.018073454174651452, + "grad_norm": 2.9389138221740723, + "learning_rate": 0.00019984292479672213, + "loss": 0.4717, + "step": 771 + }, + { + "epoch": 0.01809689574945645, + "grad_norm": 4.004184722900391, + "learning_rate": 0.00019984251182132866, + "loss": 0.8848, + "step": 772 + }, + { + "epoch": 0.018120337324261444, + "grad_norm": 1.6122528314590454, + "learning_rate": 0.0001998420983041869, + "loss": 0.386, + "step": 773 + }, + { + "epoch": 0.01814377889906644, + "grad_norm": 0.5022455453872681, + "learning_rate": 0.00019984168424529912, + "loss": 0.0824, + "step": 774 + }, + { + "epoch": 0.018167220473871433, + "grad_norm": 2.7154579162597656, + "learning_rate": 0.00019984126964466757, + "loss": 0.5319, + "step": 775 + }, + { + "epoch": 0.01819066204867643, + "grad_norm": 1.213165283203125, + "learning_rate": 0.00019984085450229445, + "loss": 1.55, + "step": 776 + }, + { + "epoch": 0.018214103623481425, + "grad_norm": 3.8310627937316895, + "learning_rate": 0.00019984043881818208, + "loss": 0.5767, + "step": 777 + }, + { + "epoch": 0.01823754519828642, + "grad_norm": 4.501200199127197, + "learning_rate": 0.00019984002259233267, + "loss": 0.926, + "step": 778 + }, + { + "epoch": 0.018260986773091417, + "grad_norm": 2.6498184204101562, + "learning_rate": 0.0001998396058247485, + "loss": 0.4971, + "step": 779 + }, + { + "epoch": 0.01828442834789641, + "grad_norm": 3.794560432434082, + "learning_rate": 0.00019983918851543178, + "loss": 0.4308, + "step": 780 + }, + { + "epoch": 0.018307869922701406, + "grad_norm": 3.4017245769500732, + "learning_rate": 0.00019983877066438485, + "loss": 0.8845, + "step": 781 + }, + { + "epoch": 0.018331311497506404, + "grad_norm": 3.83811354637146, + "learning_rate": 0.00019983835227160995, + "loss": 0.5923, + "step": 782 + }, + { + "epoch": 0.018354753072311398, + "grad_norm": 3.9773709774017334, + "learning_rate": 0.0001998379333371093, + "loss": 1.0165, + "step": 783 + }, + { + "epoch": 0.018378194647116392, + "grad_norm": 1.4334120750427246, + "learning_rate": 0.00019983751386088528, + "loss": 1.5481, + "step": 784 + }, + { + "epoch": 0.01840163622192139, + "grad_norm": 3.780634641647339, + "learning_rate": 0.00019983709384294006, + "loss": 0.7535, + "step": 785 + }, + { + "epoch": 0.018425077796726384, + "grad_norm": 3.840294599533081, + "learning_rate": 0.00019983667328327598, + "loss": 0.602, + "step": 786 + }, + { + "epoch": 0.01844851937153138, + "grad_norm": 2.836458921432495, + "learning_rate": 0.00019983625218189527, + "loss": 0.5197, + "step": 787 + }, + { + "epoch": 0.018471960946336376, + "grad_norm": 2.0898971557617188, + "learning_rate": 0.0001998358305388003, + "loss": 0.3964, + "step": 788 + }, + { + "epoch": 0.01849540252114137, + "grad_norm": 1.3812047243118286, + "learning_rate": 0.00019983540835399327, + "loss": 0.4073, + "step": 789 + }, + { + "epoch": 0.018518844095946365, + "grad_norm": 1.39218270778656, + "learning_rate": 0.00019983498562747652, + "loss": 1.4752, + "step": 790 + }, + { + "epoch": 0.018542285670751363, + "grad_norm": 3.2904069423675537, + "learning_rate": 0.00019983456235925232, + "loss": 0.5573, + "step": 791 + }, + { + "epoch": 0.018565727245556357, + "grad_norm": 3.3855435848236084, + "learning_rate": 0.00019983413854932297, + "loss": 0.5471, + "step": 792 + }, + { + "epoch": 0.01858916882036135, + "grad_norm": 1.8453420400619507, + "learning_rate": 0.0001998337141976908, + "loss": 2.0914, + "step": 793 + }, + { + "epoch": 0.018612610395166346, + "grad_norm": 3.7439820766448975, + "learning_rate": 0.0001998332893043581, + "loss": 0.7758, + "step": 794 + }, + { + "epoch": 0.018636051969971344, + "grad_norm": 3.4131808280944824, + "learning_rate": 0.00019983286386932714, + "loss": 0.5872, + "step": 795 + }, + { + "epoch": 0.018659493544776338, + "grad_norm": 3.5884950160980225, + "learning_rate": 0.00019983243789260024, + "loss": 0.5135, + "step": 796 + }, + { + "epoch": 0.018682935119581332, + "grad_norm": 1.2532389163970947, + "learning_rate": 0.00019983201137417975, + "loss": 0.2343, + "step": 797 + }, + { + "epoch": 0.01870637669438633, + "grad_norm": 1.1436880826950073, + "learning_rate": 0.00019983158431406797, + "loss": 0.1052, + "step": 798 + }, + { + "epoch": 0.018729818269191324, + "grad_norm": 3.566037654876709, + "learning_rate": 0.00019983115671226716, + "loss": 0.7546, + "step": 799 + }, + { + "epoch": 0.01875325984399632, + "grad_norm": 1.554773211479187, + "learning_rate": 0.00019983072856877976, + "loss": 0.3263, + "step": 800 + }, + { + "epoch": 0.018776701418801316, + "grad_norm": 1.6732664108276367, + "learning_rate": 0.00019983029988360795, + "loss": 0.1999, + "step": 801 + }, + { + "epoch": 0.01880014299360631, + "grad_norm": 0.7734825611114502, + "learning_rate": 0.00019982987065675416, + "loss": 0.0959, + "step": 802 + }, + { + "epoch": 0.018823584568411305, + "grad_norm": 4.037161350250244, + "learning_rate": 0.00019982944088822068, + "loss": 0.8408, + "step": 803 + }, + { + "epoch": 0.018847026143216303, + "grad_norm": 3.4674954414367676, + "learning_rate": 0.00019982901057800985, + "loss": 1.0273, + "step": 804 + }, + { + "epoch": 0.018870467718021297, + "grad_norm": 1.6569288969039917, + "learning_rate": 0.000199828579726124, + "loss": 1.101, + "step": 805 + }, + { + "epoch": 0.01889390929282629, + "grad_norm": 6.66236686706543, + "learning_rate": 0.00019982814833256548, + "loss": 0.8623, + "step": 806 + }, + { + "epoch": 0.01891735086763129, + "grad_norm": 5.618145942687988, + "learning_rate": 0.0001998277163973366, + "loss": 1.0861, + "step": 807 + }, + { + "epoch": 0.018940792442436283, + "grad_norm": 2.691143274307251, + "learning_rate": 0.00019982728392043973, + "loss": 1.9341, + "step": 808 + }, + { + "epoch": 0.018964234017241278, + "grad_norm": 4.114433765411377, + "learning_rate": 0.0001998268509018772, + "loss": 0.7923, + "step": 809 + }, + { + "epoch": 0.018987675592046272, + "grad_norm": 3.4934396743774414, + "learning_rate": 0.0001998264173416514, + "loss": 0.6242, + "step": 810 + }, + { + "epoch": 0.01901111716685127, + "grad_norm": 2.7927749156951904, + "learning_rate": 0.0001998259832397646, + "loss": 0.517, + "step": 811 + }, + { + "epoch": 0.019034558741656264, + "grad_norm": 1.511560082435608, + "learning_rate": 0.00019982554859621928, + "loss": 1.278, + "step": 812 + }, + { + "epoch": 0.01905800031646126, + "grad_norm": 2.6234939098358154, + "learning_rate": 0.00019982511341101766, + "loss": 0.5445, + "step": 813 + }, + { + "epoch": 0.019081441891266256, + "grad_norm": 3.899932622909546, + "learning_rate": 0.0001998246776841622, + "loss": 0.9789, + "step": 814 + }, + { + "epoch": 0.01910488346607125, + "grad_norm": 4.28947114944458, + "learning_rate": 0.00019982424141565524, + "loss": 0.9032, + "step": 815 + }, + { + "epoch": 0.019128325040876245, + "grad_norm": 3.311159372329712, + "learning_rate": 0.00019982380460549913, + "loss": 0.4641, + "step": 816 + }, + { + "epoch": 0.019151766615681243, + "grad_norm": 3.5831122398376465, + "learning_rate": 0.00019982336725369625, + "loss": 0.797, + "step": 817 + }, + { + "epoch": 0.019175208190486237, + "grad_norm": 2.4373152256011963, + "learning_rate": 0.00019982292936024897, + "loss": 0.3105, + "step": 818 + }, + { + "epoch": 0.01919864976529123, + "grad_norm": 5.345475673675537, + "learning_rate": 0.00019982249092515965, + "loss": 1.0226, + "step": 819 + }, + { + "epoch": 0.01922209134009623, + "grad_norm": 1.042815923690796, + "learning_rate": 0.00019982205194843072, + "loss": 0.1824, + "step": 820 + }, + { + "epoch": 0.019245532914901223, + "grad_norm": 3.4731287956237793, + "learning_rate": 0.00019982161243006452, + "loss": 0.5241, + "step": 821 + }, + { + "epoch": 0.019268974489706218, + "grad_norm": 3.0706193447113037, + "learning_rate": 0.0001998211723700634, + "loss": 0.6385, + "step": 822 + }, + { + "epoch": 0.019292416064511216, + "grad_norm": 0.8642702698707581, + "learning_rate": 0.00019982073176842986, + "loss": 0.1217, + "step": 823 + }, + { + "epoch": 0.01931585763931621, + "grad_norm": 6.885644435882568, + "learning_rate": 0.00019982029062516618, + "loss": 1.0191, + "step": 824 + }, + { + "epoch": 0.019339299214121204, + "grad_norm": 1.6395896673202515, + "learning_rate": 0.0001998198489402748, + "loss": 0.31, + "step": 825 + }, + { + "epoch": 0.0193627407889262, + "grad_norm": 1.6511170864105225, + "learning_rate": 0.00019981940671375816, + "loss": 0.3031, + "step": 826 + }, + { + "epoch": 0.019386182363731196, + "grad_norm": 2.981431484222412, + "learning_rate": 0.00019981896394561857, + "loss": 0.5522, + "step": 827 + }, + { + "epoch": 0.01940962393853619, + "grad_norm": 3.8293545246124268, + "learning_rate": 0.0001998185206358585, + "loss": 0.4813, + "step": 828 + }, + { + "epoch": 0.019433065513341185, + "grad_norm": 3.301079034805298, + "learning_rate": 0.00019981807678448027, + "loss": 0.5634, + "step": 829 + }, + { + "epoch": 0.019456507088146183, + "grad_norm": 3.772845983505249, + "learning_rate": 0.00019981763239148637, + "loss": 0.6905, + "step": 830 + }, + { + "epoch": 0.019479948662951177, + "grad_norm": 0.3754485249519348, + "learning_rate": 0.00019981718745687923, + "loss": 0.0509, + "step": 831 + }, + { + "epoch": 0.01950339023775617, + "grad_norm": 1.9945324659347534, + "learning_rate": 0.00019981674198066122, + "loss": 1.5027, + "step": 832 + }, + { + "epoch": 0.01952683181256117, + "grad_norm": 3.9455931186676025, + "learning_rate": 0.00019981629596283474, + "loss": 0.6132, + "step": 833 + }, + { + "epoch": 0.019550273387366163, + "grad_norm": 3.182389259338379, + "learning_rate": 0.00019981584940340224, + "loss": 0.6195, + "step": 834 + }, + { + "epoch": 0.019573714962171158, + "grad_norm": 1.4701480865478516, + "learning_rate": 0.00019981540230236614, + "loss": 0.2082, + "step": 835 + }, + { + "epoch": 0.019597156536976156, + "grad_norm": 3.4364893436431885, + "learning_rate": 0.00019981495465972885, + "loss": 0.4241, + "step": 836 + }, + { + "epoch": 0.01962059811178115, + "grad_norm": 2.0508673191070557, + "learning_rate": 0.0001998145064754928, + "loss": 1.4907, + "step": 837 + }, + { + "epoch": 0.019644039686586144, + "grad_norm": 4.514815807342529, + "learning_rate": 0.00019981405774966045, + "loss": 1.1703, + "step": 838 + }, + { + "epoch": 0.019667481261391142, + "grad_norm": 4.443180084228516, + "learning_rate": 0.0001998136084822342, + "loss": 1.0227, + "step": 839 + }, + { + "epoch": 0.019690922836196136, + "grad_norm": 4.085733890533447, + "learning_rate": 0.0001998131586732165, + "loss": 0.769, + "step": 840 + }, + { + "epoch": 0.01971436441100113, + "grad_norm": 3.9873504638671875, + "learning_rate": 0.00019981270832260983, + "loss": 1.0332, + "step": 841 + }, + { + "epoch": 0.019737805985806125, + "grad_norm": 4.420246124267578, + "learning_rate": 0.00019981225743041656, + "loss": 0.7101, + "step": 842 + }, + { + "epoch": 0.019761247560611123, + "grad_norm": 3.54548978805542, + "learning_rate": 0.0001998118059966392, + "loss": 0.4227, + "step": 843 + }, + { + "epoch": 0.019784689135416117, + "grad_norm": 4.667275905609131, + "learning_rate": 0.00019981135402128016, + "loss": 0.6114, + "step": 844 + }, + { + "epoch": 0.01980813071022111, + "grad_norm": 2.7594118118286133, + "learning_rate": 0.0001998109015043419, + "loss": 0.5259, + "step": 845 + }, + { + "epoch": 0.01983157228502611, + "grad_norm": 5.159000873565674, + "learning_rate": 0.0001998104484458269, + "loss": 0.5538, + "step": 846 + }, + { + "epoch": 0.019855013859831103, + "grad_norm": 3.660999059677124, + "learning_rate": 0.00019980999484573762, + "loss": 0.4063, + "step": 847 + }, + { + "epoch": 0.019878455434636098, + "grad_norm": 4.692598342895508, + "learning_rate": 0.00019980954070407644, + "loss": 0.8022, + "step": 848 + }, + { + "epoch": 0.019901897009441095, + "grad_norm": 3.295741319656372, + "learning_rate": 0.00019980908602084595, + "loss": 0.6867, + "step": 849 + }, + { + "epoch": 0.01992533858424609, + "grad_norm": 3.300396680831909, + "learning_rate": 0.0001998086307960485, + "loss": 0.797, + "step": 850 + }, + { + "epoch": 0.019948780159051084, + "grad_norm": 4.005765438079834, + "learning_rate": 0.00019980817502968665, + "loss": 0.7366, + "step": 851 + }, + { + "epoch": 0.019972221733856082, + "grad_norm": 3.476701259613037, + "learning_rate": 0.00019980771872176283, + "loss": 0.7694, + "step": 852 + }, + { + "epoch": 0.019995663308661076, + "grad_norm": 2.956392765045166, + "learning_rate": 0.00019980726187227948, + "loss": 0.2681, + "step": 853 + }, + { + "epoch": 0.02001910488346607, + "grad_norm": 4.433586120605469, + "learning_rate": 0.00019980680448123918, + "loss": 0.6979, + "step": 854 + }, + { + "epoch": 0.02004254645827107, + "grad_norm": 3.487717390060425, + "learning_rate": 0.00019980634654864435, + "loss": 0.5971, + "step": 855 + }, + { + "epoch": 0.020065988033076063, + "grad_norm": 1.9375693798065186, + "learning_rate": 0.00019980588807449743, + "loss": 0.4065, + "step": 856 + }, + { + "epoch": 0.020089429607881057, + "grad_norm": 3.744464635848999, + "learning_rate": 0.000199805429058801, + "loss": 1.0724, + "step": 857 + }, + { + "epoch": 0.02011287118268605, + "grad_norm": 3.679365396499634, + "learning_rate": 0.00019980496950155747, + "loss": 0.4938, + "step": 858 + }, + { + "epoch": 0.02013631275749105, + "grad_norm": 4.87100076675415, + "learning_rate": 0.0001998045094027694, + "loss": 1.1641, + "step": 859 + }, + { + "epoch": 0.020159754332296043, + "grad_norm": 2.8561036586761475, + "learning_rate": 0.00019980404876243922, + "loss": 0.6451, + "step": 860 + }, + { + "epoch": 0.020183195907101038, + "grad_norm": 1.1598515510559082, + "learning_rate": 0.00019980358758056948, + "loss": 0.221, + "step": 861 + }, + { + "epoch": 0.020206637481906035, + "grad_norm": 2.183331251144409, + "learning_rate": 0.0001998031258571627, + "loss": 1.4465, + "step": 862 + }, + { + "epoch": 0.02023007905671103, + "grad_norm": 1.0717720985412598, + "learning_rate": 0.0001998026635922213, + "loss": 0.2062, + "step": 863 + }, + { + "epoch": 0.020253520631516024, + "grad_norm": 4.106282711029053, + "learning_rate": 0.0001998022007857479, + "loss": 0.7406, + "step": 864 + }, + { + "epoch": 0.020276962206321022, + "grad_norm": 1.7471357583999634, + "learning_rate": 0.0001998017374377449, + "loss": 0.2848, + "step": 865 + }, + { + "epoch": 0.020300403781126016, + "grad_norm": 1.3788355588912964, + "learning_rate": 0.00019980127354821486, + "loss": 0.3309, + "step": 866 + }, + { + "epoch": 0.02032384535593101, + "grad_norm": 1.862851858139038, + "learning_rate": 0.00019980080911716038, + "loss": 0.3575, + "step": 867 + }, + { + "epoch": 0.020347286930736008, + "grad_norm": 0.7169710397720337, + "learning_rate": 0.00019980034414458385, + "loss": 0.1235, + "step": 868 + }, + { + "epoch": 0.020370728505541003, + "grad_norm": 5.789308071136475, + "learning_rate": 0.00019979987863048783, + "loss": 1.3723, + "step": 869 + }, + { + "epoch": 0.020394170080345997, + "grad_norm": 3.398895740509033, + "learning_rate": 0.0001997994125748749, + "loss": 0.5547, + "step": 870 + }, + { + "epoch": 0.020417611655150995, + "grad_norm": 3.243278980255127, + "learning_rate": 0.00019979894597774752, + "loss": 0.5203, + "step": 871 + }, + { + "epoch": 0.02044105322995599, + "grad_norm": 5.593380928039551, + "learning_rate": 0.00019979847883910826, + "loss": 0.9926, + "step": 872 + }, + { + "epoch": 0.020464494804760983, + "grad_norm": 1.834416389465332, + "learning_rate": 0.00019979801115895966, + "loss": 0.2883, + "step": 873 + }, + { + "epoch": 0.020487936379565978, + "grad_norm": 1.248551607131958, + "learning_rate": 0.00019979754293730423, + "loss": 0.3257, + "step": 874 + }, + { + "epoch": 0.020511377954370975, + "grad_norm": 3.3392202854156494, + "learning_rate": 0.00019979707417414454, + "loss": 0.7039, + "step": 875 + }, + { + "epoch": 0.02053481952917597, + "grad_norm": 2.694645881652832, + "learning_rate": 0.00019979660486948314, + "loss": 0.5102, + "step": 876 + }, + { + "epoch": 0.020558261103980964, + "grad_norm": 4.9307050704956055, + "learning_rate": 0.00019979613502332252, + "loss": 0.8215, + "step": 877 + }, + { + "epoch": 0.020581702678785962, + "grad_norm": 3.9345462322235107, + "learning_rate": 0.00019979566463566525, + "loss": 0.5536, + "step": 878 + }, + { + "epoch": 0.020605144253590956, + "grad_norm": 3.733874797821045, + "learning_rate": 0.00019979519370651394, + "loss": 0.6814, + "step": 879 + }, + { + "epoch": 0.02062858582839595, + "grad_norm": 1.3488261699676514, + "learning_rate": 0.0001997947222358711, + "loss": 0.288, + "step": 880 + }, + { + "epoch": 0.020652027403200948, + "grad_norm": 2.853346824645996, + "learning_rate": 0.0001997942502237393, + "loss": 0.4799, + "step": 881 + }, + { + "epoch": 0.020675468978005943, + "grad_norm": 1.5680640935897827, + "learning_rate": 0.00019979377767012104, + "loss": 0.345, + "step": 882 + }, + { + "epoch": 0.020698910552810937, + "grad_norm": 4.030822277069092, + "learning_rate": 0.00019979330457501896, + "loss": 0.7308, + "step": 883 + }, + { + "epoch": 0.020722352127615935, + "grad_norm": 4.445903301239014, + "learning_rate": 0.0001997928309384356, + "loss": 1.1108, + "step": 884 + }, + { + "epoch": 0.02074579370242093, + "grad_norm": 3.0255589485168457, + "learning_rate": 0.00019979235676037356, + "loss": 0.5103, + "step": 885 + }, + { + "epoch": 0.020769235277225923, + "grad_norm": 1.2191728353500366, + "learning_rate": 0.00019979188204083536, + "loss": 0.1654, + "step": 886 + }, + { + "epoch": 0.02079267685203092, + "grad_norm": 5.066000461578369, + "learning_rate": 0.0001997914067798236, + "loss": 0.894, + "step": 887 + }, + { + "epoch": 0.020816118426835915, + "grad_norm": 1.5256781578063965, + "learning_rate": 0.00019979093097734088, + "loss": 0.228, + "step": 888 + }, + { + "epoch": 0.02083956000164091, + "grad_norm": 0.7980531454086304, + "learning_rate": 0.00019979045463338976, + "loss": 0.1343, + "step": 889 + }, + { + "epoch": 0.020863001576445904, + "grad_norm": 3.5532009601593018, + "learning_rate": 0.00019978997774797283, + "loss": 0.4064, + "step": 890 + }, + { + "epoch": 0.020886443151250902, + "grad_norm": 1.5353751182556152, + "learning_rate": 0.00019978950032109268, + "loss": 1.1589, + "step": 891 + }, + { + "epoch": 0.020909884726055896, + "grad_norm": 1.9914871454238892, + "learning_rate": 0.00019978902235275189, + "loss": 0.32, + "step": 892 + }, + { + "epoch": 0.02093332630086089, + "grad_norm": 1.7888782024383545, + "learning_rate": 0.00019978854384295302, + "loss": 0.2393, + "step": 893 + }, + { + "epoch": 0.020956767875665888, + "grad_norm": 1.1168993711471558, + "learning_rate": 0.00019978806479169876, + "loss": 0.1328, + "step": 894 + }, + { + "epoch": 0.020980209450470882, + "grad_norm": 2.645204782485962, + "learning_rate": 0.00019978758519899163, + "loss": 0.5013, + "step": 895 + }, + { + "epoch": 0.021003651025275877, + "grad_norm": 3.0094003677368164, + "learning_rate": 0.00019978710506483424, + "loss": 0.3831, + "step": 896 + }, + { + "epoch": 0.021027092600080875, + "grad_norm": 1.3449766635894775, + "learning_rate": 0.00019978662438922923, + "loss": 0.1682, + "step": 897 + }, + { + "epoch": 0.02105053417488587, + "grad_norm": 1.6438261270523071, + "learning_rate": 0.00019978614317217922, + "loss": 0.3577, + "step": 898 + }, + { + "epoch": 0.021073975749690863, + "grad_norm": 4.519595623016357, + "learning_rate": 0.00019978566141368674, + "loss": 0.726, + "step": 899 + }, + { + "epoch": 0.02109741732449586, + "grad_norm": 1.2467432022094727, + "learning_rate": 0.0001997851791137545, + "loss": 0.9068, + "step": 900 + }, + { + "epoch": 0.021120858899300855, + "grad_norm": 3.3383169174194336, + "learning_rate": 0.00019978469627238504, + "loss": 0.6308, + "step": 901 + }, + { + "epoch": 0.02114430047410585, + "grad_norm": 0.24905528128147125, + "learning_rate": 0.00019978421288958101, + "loss": 0.0511, + "step": 902 + }, + { + "epoch": 0.021167742048910847, + "grad_norm": 1.674675464630127, + "learning_rate": 0.00019978372896534507, + "loss": 0.4157, + "step": 903 + }, + { + "epoch": 0.02119118362371584, + "grad_norm": 1.4674936532974243, + "learning_rate": 0.0001997832444996798, + "loss": 0.448, + "step": 904 + }, + { + "epoch": 0.021214625198520836, + "grad_norm": 4.448566436767578, + "learning_rate": 0.0001997827594925878, + "loss": 0.7048, + "step": 905 + }, + { + "epoch": 0.02123806677332583, + "grad_norm": 4.507232189178467, + "learning_rate": 0.0001997822739440718, + "loss": 0.789, + "step": 906 + }, + { + "epoch": 0.021261508348130828, + "grad_norm": 3.295973062515259, + "learning_rate": 0.00019978178785413434, + "loss": 0.4501, + "step": 907 + }, + { + "epoch": 0.021284949922935822, + "grad_norm": 3.9342076778411865, + "learning_rate": 0.0001997813012227781, + "loss": 0.7231, + "step": 908 + }, + { + "epoch": 0.021308391497740817, + "grad_norm": 4.33222770690918, + "learning_rate": 0.0001997808140500057, + "loss": 0.758, + "step": 909 + }, + { + "epoch": 0.021331833072545815, + "grad_norm": 2.9070684909820557, + "learning_rate": 0.00019978032633581982, + "loss": 0.6521, + "step": 910 + }, + { + "epoch": 0.02135527464735081, + "grad_norm": 3.659457206726074, + "learning_rate": 0.00019977983808022306, + "loss": 0.9319, + "step": 911 + }, + { + "epoch": 0.021378716222155803, + "grad_norm": 4.666127681732178, + "learning_rate": 0.0001997793492832181, + "loss": 0.7746, + "step": 912 + }, + { + "epoch": 0.0214021577969608, + "grad_norm": 0.8995996713638306, + "learning_rate": 0.0001997788599448076, + "loss": 0.6255, + "step": 913 + }, + { + "epoch": 0.021425599371765795, + "grad_norm": 4.290579319000244, + "learning_rate": 0.0001997783700649942, + "loss": 1.1139, + "step": 914 + }, + { + "epoch": 0.02144904094657079, + "grad_norm": 1.0114145278930664, + "learning_rate": 0.00019977787964378055, + "loss": 0.1016, + "step": 915 + }, + { + "epoch": 0.021472482521375787, + "grad_norm": 4.284561634063721, + "learning_rate": 0.00019977738868116933, + "loss": 0.4003, + "step": 916 + }, + { + "epoch": 0.02149592409618078, + "grad_norm": 1.8832639455795288, + "learning_rate": 0.0001997768971771632, + "loss": 1.6588, + "step": 917 + }, + { + "epoch": 0.021519365670985776, + "grad_norm": 5.783070087432861, + "learning_rate": 0.00019977640513176478, + "loss": 0.8146, + "step": 918 + }, + { + "epoch": 0.021542807245790774, + "grad_norm": 6.055897235870361, + "learning_rate": 0.00019977591254497683, + "loss": 1.3802, + "step": 919 + }, + { + "epoch": 0.021566248820595768, + "grad_norm": 3.400578498840332, + "learning_rate": 0.00019977541941680193, + "loss": 0.4699, + "step": 920 + }, + { + "epoch": 0.021589690395400762, + "grad_norm": 1.5787194967269897, + "learning_rate": 0.0001997749257472428, + "loss": 1.4138, + "step": 921 + }, + { + "epoch": 0.021613131970205757, + "grad_norm": 2.395667791366577, + "learning_rate": 0.00019977443153630216, + "loss": 1.7624, + "step": 922 + }, + { + "epoch": 0.021636573545010754, + "grad_norm": 6.09457540512085, + "learning_rate": 0.00019977393678398265, + "loss": 0.5858, + "step": 923 + }, + { + "epoch": 0.02166001511981575, + "grad_norm": 1.4015600681304932, + "learning_rate": 0.0001997734414902869, + "loss": 0.2241, + "step": 924 + }, + { + "epoch": 0.021683456694620743, + "grad_norm": 0.716822624206543, + "learning_rate": 0.00019977294565521768, + "loss": 0.1085, + "step": 925 + }, + { + "epoch": 0.02170689826942574, + "grad_norm": 3.851517677307129, + "learning_rate": 0.00019977244927877764, + "loss": 0.6231, + "step": 926 + }, + { + "epoch": 0.021730339844230735, + "grad_norm": 1.9565775394439697, + "learning_rate": 0.0001997719523609695, + "loss": 0.4502, + "step": 927 + }, + { + "epoch": 0.02175378141903573, + "grad_norm": 3.0964198112487793, + "learning_rate": 0.00019977145490179593, + "loss": 0.6877, + "step": 928 + }, + { + "epoch": 0.021777222993840727, + "grad_norm": 2.6855428218841553, + "learning_rate": 0.00019977095690125967, + "loss": 0.4586, + "step": 929 + }, + { + "epoch": 0.02180066456864572, + "grad_norm": 2.1346964836120605, + "learning_rate": 0.00019977045835936337, + "loss": 1.2541, + "step": 930 + }, + { + "epoch": 0.021824106143450716, + "grad_norm": 1.3967342376708984, + "learning_rate": 0.00019976995927610975, + "loss": 0.2576, + "step": 931 + }, + { + "epoch": 0.021847547718255714, + "grad_norm": 1.31672203540802, + "learning_rate": 0.00019976945965150154, + "loss": 0.2791, + "step": 932 + }, + { + "epoch": 0.021870989293060708, + "grad_norm": 3.2810611724853516, + "learning_rate": 0.0001997689594855414, + "loss": 0.7141, + "step": 933 + }, + { + "epoch": 0.021894430867865702, + "grad_norm": 3.036647081375122, + "learning_rate": 0.00019976845877823214, + "loss": 0.5514, + "step": 934 + }, + { + "epoch": 0.0219178724426707, + "grad_norm": 1.0567086935043335, + "learning_rate": 0.00019976795752957638, + "loss": 0.2288, + "step": 935 + }, + { + "epoch": 0.021941314017475694, + "grad_norm": 0.7677361965179443, + "learning_rate": 0.00019976745573957688, + "loss": 0.0974, + "step": 936 + }, + { + "epoch": 0.02196475559228069, + "grad_norm": 1.5133455991744995, + "learning_rate": 0.00019976695340823638, + "loss": 0.3048, + "step": 937 + }, + { + "epoch": 0.021988197167085683, + "grad_norm": 1.273379921913147, + "learning_rate": 0.00019976645053555755, + "loss": 0.1632, + "step": 938 + }, + { + "epoch": 0.02201163874189068, + "grad_norm": 4.447317123413086, + "learning_rate": 0.00019976594712154316, + "loss": 0.8747, + "step": 939 + }, + { + "epoch": 0.022035080316695675, + "grad_norm": 2.8449673652648926, + "learning_rate": 0.00019976544316619598, + "loss": 0.614, + "step": 940 + }, + { + "epoch": 0.02205852189150067, + "grad_norm": 0.5910099744796753, + "learning_rate": 0.00019976493866951867, + "loss": 0.1177, + "step": 941 + }, + { + "epoch": 0.022081963466305667, + "grad_norm": 3.787677049636841, + "learning_rate": 0.000199764433631514, + "loss": 0.8008, + "step": 942 + }, + { + "epoch": 0.02210540504111066, + "grad_norm": 4.951800346374512, + "learning_rate": 0.0001997639280521847, + "loss": 0.9602, + "step": 943 + }, + { + "epoch": 0.022128846615915656, + "grad_norm": 4.62385892868042, + "learning_rate": 0.0001997634219315335, + "loss": 0.8488, + "step": 944 + }, + { + "epoch": 0.022152288190720654, + "grad_norm": 3.2560789585113525, + "learning_rate": 0.0001997629152695632, + "loss": 0.739, + "step": 945 + }, + { + "epoch": 0.022175729765525648, + "grad_norm": 3.5510380268096924, + "learning_rate": 0.00019976240806627652, + "loss": 0.7722, + "step": 946 + }, + { + "epoch": 0.022199171340330642, + "grad_norm": 2.6201016902923584, + "learning_rate": 0.00019976190032167622, + "loss": 0.7987, + "step": 947 + }, + { + "epoch": 0.02222261291513564, + "grad_norm": 1.3707325458526611, + "learning_rate": 0.000199761392035765, + "loss": 0.4238, + "step": 948 + }, + { + "epoch": 0.022246054489940634, + "grad_norm": 1.4781279563903809, + "learning_rate": 0.0001997608832085457, + "loss": 1.5749, + "step": 949 + }, + { + "epoch": 0.02226949606474563, + "grad_norm": 3.2480838298797607, + "learning_rate": 0.00019976037384002103, + "loss": 0.7215, + "step": 950 + }, + { + "epoch": 0.022292937639550627, + "grad_norm": 1.267634630203247, + "learning_rate": 0.00019975986393019378, + "loss": 0.2496, + "step": 951 + }, + { + "epoch": 0.02231637921435562, + "grad_norm": 3.879708766937256, + "learning_rate": 0.00019975935347906666, + "loss": 0.8148, + "step": 952 + }, + { + "epoch": 0.022339820789160615, + "grad_norm": 1.3519272804260254, + "learning_rate": 0.0001997588424866425, + "loss": 1.0464, + "step": 953 + }, + { + "epoch": 0.022363262363965613, + "grad_norm": 2.5415279865264893, + "learning_rate": 0.0001997583309529241, + "loss": 0.4783, + "step": 954 + }, + { + "epoch": 0.022386703938770607, + "grad_norm": 4.451211452484131, + "learning_rate": 0.00019975781887791416, + "loss": 0.9041, + "step": 955 + }, + { + "epoch": 0.0224101455135756, + "grad_norm": 3.234757900238037, + "learning_rate": 0.0001997573062616155, + "loss": 0.6884, + "step": 956 + }, + { + "epoch": 0.022433587088380596, + "grad_norm": 3.8481006622314453, + "learning_rate": 0.00019975679310403087, + "loss": 0.7637, + "step": 957 + }, + { + "epoch": 0.022457028663185594, + "grad_norm": 3.3961610794067383, + "learning_rate": 0.00019975627940516306, + "loss": 0.7818, + "step": 958 + }, + { + "epoch": 0.022480470237990588, + "grad_norm": 1.8531211614608765, + "learning_rate": 0.0001997557651650149, + "loss": 1.3744, + "step": 959 + }, + { + "epoch": 0.022503911812795582, + "grad_norm": 2.9255177974700928, + "learning_rate": 0.00019975525038358916, + "loss": 0.4672, + "step": 960 + }, + { + "epoch": 0.02252735338760058, + "grad_norm": 3.8063724040985107, + "learning_rate": 0.0001997547350608886, + "loss": 0.6893, + "step": 961 + }, + { + "epoch": 0.022550794962405574, + "grad_norm": 4.615594387054443, + "learning_rate": 0.00019975421919691608, + "loss": 1.1268, + "step": 962 + }, + { + "epoch": 0.02257423653721057, + "grad_norm": 3.784677743911743, + "learning_rate": 0.00019975370279167435, + "loss": 0.7826, + "step": 963 + }, + { + "epoch": 0.022597678112015566, + "grad_norm": 2.80476713180542, + "learning_rate": 0.0001997531858451662, + "loss": 0.3988, + "step": 964 + }, + { + "epoch": 0.02262111968682056, + "grad_norm": 2.6550402641296387, + "learning_rate": 0.00019975266835739447, + "loss": 0.467, + "step": 965 + }, + { + "epoch": 0.022644561261625555, + "grad_norm": 2.958087682723999, + "learning_rate": 0.000199752150328362, + "loss": 0.7867, + "step": 966 + }, + { + "epoch": 0.022668002836430553, + "grad_norm": 3.672670364379883, + "learning_rate": 0.0001997516317580715, + "loss": 0.4953, + "step": 967 + }, + { + "epoch": 0.022691444411235547, + "grad_norm": 2.8535187244415283, + "learning_rate": 0.00019975111264652586, + "loss": 0.4779, + "step": 968 + }, + { + "epoch": 0.02271488598604054, + "grad_norm": 3.5438899993896484, + "learning_rate": 0.00019975059299372788, + "loss": 0.5646, + "step": 969 + }, + { + "epoch": 0.02273832756084554, + "grad_norm": 4.769031524658203, + "learning_rate": 0.00019975007279968038, + "loss": 0.5119, + "step": 970 + }, + { + "epoch": 0.022761769135650534, + "grad_norm": 1.7468395233154297, + "learning_rate": 0.00019974955206438616, + "loss": 0.2187, + "step": 971 + }, + { + "epoch": 0.022785210710455528, + "grad_norm": 4.033405303955078, + "learning_rate": 0.0001997490307878481, + "loss": 0.8366, + "step": 972 + }, + { + "epoch": 0.022808652285260522, + "grad_norm": 2.5264463424682617, + "learning_rate": 0.00019974850897006897, + "loss": 0.4813, + "step": 973 + }, + { + "epoch": 0.02283209386006552, + "grad_norm": 2.3931329250335693, + "learning_rate": 0.00019974798661105166, + "loss": 0.508, + "step": 974 + }, + { + "epoch": 0.022855535434870514, + "grad_norm": 3.445068359375, + "learning_rate": 0.0001997474637107989, + "loss": 0.6536, + "step": 975 + }, + { + "epoch": 0.02287897700967551, + "grad_norm": 2.835681915283203, + "learning_rate": 0.00019974694026931363, + "loss": 0.7497, + "step": 976 + }, + { + "epoch": 0.022902418584480506, + "grad_norm": 5.772027492523193, + "learning_rate": 0.0001997464162865987, + "loss": 1.2001, + "step": 977 + }, + { + "epoch": 0.0229258601592855, + "grad_norm": 3.3635878562927246, + "learning_rate": 0.0001997458917626569, + "loss": 0.8315, + "step": 978 + }, + { + "epoch": 0.022949301734090495, + "grad_norm": 2.8641724586486816, + "learning_rate": 0.00019974536669749104, + "loss": 0.65, + "step": 979 + }, + { + "epoch": 0.022972743308895493, + "grad_norm": 3.382577657699585, + "learning_rate": 0.00019974484109110404, + "loss": 0.5509, + "step": 980 + }, + { + "epoch": 0.022996184883700487, + "grad_norm": 4.494714260101318, + "learning_rate": 0.00019974431494349873, + "loss": 0.7788, + "step": 981 + }, + { + "epoch": 0.02301962645850548, + "grad_norm": 1.191514492034912, + "learning_rate": 0.00019974378825467795, + "loss": 0.1707, + "step": 982 + }, + { + "epoch": 0.02304306803331048, + "grad_norm": 3.2888941764831543, + "learning_rate": 0.00019974326102464458, + "loss": 0.8213, + "step": 983 + }, + { + "epoch": 0.023066509608115474, + "grad_norm": 5.91091775894165, + "learning_rate": 0.00019974273325340148, + "loss": 1.3629, + "step": 984 + }, + { + "epoch": 0.023089951182920468, + "grad_norm": 3.382094144821167, + "learning_rate": 0.0001997422049409515, + "loss": 1.0487, + "step": 985 + }, + { + "epoch": 0.023113392757725466, + "grad_norm": 3.3960788249969482, + "learning_rate": 0.0001997416760872975, + "loss": 0.549, + "step": 986 + }, + { + "epoch": 0.02313683433253046, + "grad_norm": 2.6911587715148926, + "learning_rate": 0.00019974114669244237, + "loss": 0.8003, + "step": 987 + }, + { + "epoch": 0.023160275907335454, + "grad_norm": 3.2989046573638916, + "learning_rate": 0.00019974061675638897, + "loss": 0.8248, + "step": 988 + }, + { + "epoch": 0.02318371748214045, + "grad_norm": 2.4129183292388916, + "learning_rate": 0.00019974008627914017, + "loss": 0.3422, + "step": 989 + }, + { + "epoch": 0.023207159056945446, + "grad_norm": 1.2467447519302368, + "learning_rate": 0.00019973955526069888, + "loss": 0.239, + "step": 990 + }, + { + "epoch": 0.02323060063175044, + "grad_norm": 4.208026885986328, + "learning_rate": 0.00019973902370106796, + "loss": 0.6792, + "step": 991 + }, + { + "epoch": 0.023254042206555435, + "grad_norm": 1.339125633239746, + "learning_rate": 0.00019973849160025026, + "loss": 0.3449, + "step": 992 + }, + { + "epoch": 0.023277483781360433, + "grad_norm": 2.211322546005249, + "learning_rate": 0.00019973795895824874, + "loss": 0.3883, + "step": 993 + }, + { + "epoch": 0.023300925356165427, + "grad_norm": 1.3940914869308472, + "learning_rate": 0.00019973742577506623, + "loss": 0.2456, + "step": 994 + }, + { + "epoch": 0.02332436693097042, + "grad_norm": 3.5918378829956055, + "learning_rate": 0.00019973689205070566, + "loss": 0.7864, + "step": 995 + }, + { + "epoch": 0.02334780850577542, + "grad_norm": 3.4019579887390137, + "learning_rate": 0.00019973635778516987, + "loss": 0.7243, + "step": 996 + }, + { + "epoch": 0.023371250080580414, + "grad_norm": 3.1331140995025635, + "learning_rate": 0.00019973582297846184, + "loss": 0.7892, + "step": 997 + }, + { + "epoch": 0.023394691655385408, + "grad_norm": 0.4645175635814667, + "learning_rate": 0.00019973528763058443, + "loss": 0.0698, + "step": 998 + }, + { + "epoch": 0.023418133230190406, + "grad_norm": 1.6821728944778442, + "learning_rate": 0.0001997347517415405, + "loss": 1.0395, + "step": 999 + }, + { + "epoch": 0.0234415748049954, + "grad_norm": 3.348080635070801, + "learning_rate": 0.00019973421531133302, + "loss": 0.5678, + "step": 1000 + }, + { + "epoch": 0.023465016379800394, + "grad_norm": 1.7352770566940308, + "learning_rate": 0.0001997336783399649, + "loss": 0.4231, + "step": 1001 + }, + { + "epoch": 0.023488457954605392, + "grad_norm": 1.1100043058395386, + "learning_rate": 0.00019973314082743905, + "loss": 0.1648, + "step": 1002 + }, + { + "epoch": 0.023511899529410386, + "grad_norm": 0.8951387405395508, + "learning_rate": 0.00019973260277375836, + "loss": 0.152, + "step": 1003 + }, + { + "epoch": 0.02353534110421538, + "grad_norm": 2.6199376583099365, + "learning_rate": 0.00019973206417892573, + "loss": 0.4222, + "step": 1004 + }, + { + "epoch": 0.023558782679020375, + "grad_norm": 2.8424232006073, + "learning_rate": 0.00019973152504294416, + "loss": 0.6161, + "step": 1005 + }, + { + "epoch": 0.023582224253825373, + "grad_norm": 3.932279109954834, + "learning_rate": 0.00019973098536581652, + "loss": 0.6986, + "step": 1006 + }, + { + "epoch": 0.023605665828630367, + "grad_norm": 4.104268550872803, + "learning_rate": 0.00019973044514754574, + "loss": 0.6664, + "step": 1007 + }, + { + "epoch": 0.02362910740343536, + "grad_norm": 4.805617809295654, + "learning_rate": 0.00019972990438813475, + "loss": 0.7715, + "step": 1008 + }, + { + "epoch": 0.02365254897824036, + "grad_norm": 1.324083924293518, + "learning_rate": 0.00019972936308758652, + "loss": 0.178, + "step": 1009 + }, + { + "epoch": 0.023675990553045353, + "grad_norm": 3.0766732692718506, + "learning_rate": 0.00019972882124590395, + "loss": 0.3602, + "step": 1010 + }, + { + "epoch": 0.023699432127850348, + "grad_norm": 3.4151134490966797, + "learning_rate": 0.00019972827886309, + "loss": 0.7153, + "step": 1011 + }, + { + "epoch": 0.023722873702655346, + "grad_norm": 0.659814715385437, + "learning_rate": 0.0001997277359391476, + "loss": 0.1344, + "step": 1012 + }, + { + "epoch": 0.02374631527746034, + "grad_norm": 3.695916175842285, + "learning_rate": 0.0001997271924740797, + "loss": 0.8238, + "step": 1013 + }, + { + "epoch": 0.023769756852265334, + "grad_norm": 3.6312055587768555, + "learning_rate": 0.00019972664846788925, + "loss": 0.324, + "step": 1014 + }, + { + "epoch": 0.023793198427070332, + "grad_norm": 3.0894525051116943, + "learning_rate": 0.00019972610392057918, + "loss": 0.6426, + "step": 1015 + }, + { + "epoch": 0.023816640001875326, + "grad_norm": 4.854449272155762, + "learning_rate": 0.0001997255588321525, + "loss": 0.6603, + "step": 1016 + }, + { + "epoch": 0.02384008157668032, + "grad_norm": 1.3932342529296875, + "learning_rate": 0.0001997250132026121, + "loss": 0.2622, + "step": 1017 + }, + { + "epoch": 0.02386352315148532, + "grad_norm": 3.5631115436553955, + "learning_rate": 0.00019972446703196098, + "loss": 0.6506, + "step": 1018 + }, + { + "epoch": 0.023886964726290313, + "grad_norm": 3.223653554916382, + "learning_rate": 0.00019972392032020214, + "loss": 0.5047, + "step": 1019 + }, + { + "epoch": 0.023910406301095307, + "grad_norm": 4.570842742919922, + "learning_rate": 0.00019972337306733845, + "loss": 0.9974, + "step": 1020 + }, + { + "epoch": 0.0239338478759003, + "grad_norm": 2.6487083435058594, + "learning_rate": 0.000199722825273373, + "loss": 0.4703, + "step": 1021 + }, + { + "epoch": 0.0239572894507053, + "grad_norm": 3.0478272438049316, + "learning_rate": 0.00019972227693830863, + "loss": 0.5187, + "step": 1022 + }, + { + "epoch": 0.023980731025510293, + "grad_norm": 0.9246007800102234, + "learning_rate": 0.0001997217280621484, + "loss": 0.225, + "step": 1023 + }, + { + "epoch": 0.024004172600315288, + "grad_norm": 3.60613751411438, + "learning_rate": 0.00019972117864489527, + "loss": 0.5017, + "step": 1024 + }, + { + "epoch": 0.024027614175120286, + "grad_norm": 3.4942400455474854, + "learning_rate": 0.00019972062868655225, + "loss": 0.7047, + "step": 1025 + }, + { + "epoch": 0.02405105574992528, + "grad_norm": 3.0333173274993896, + "learning_rate": 0.00019972007818712224, + "loss": 0.4213, + "step": 1026 + }, + { + "epoch": 0.024074497324730274, + "grad_norm": 4.13253116607666, + "learning_rate": 0.00019971952714660832, + "loss": 0.6829, + "step": 1027 + }, + { + "epoch": 0.024097938899535272, + "grad_norm": 3.4068217277526855, + "learning_rate": 0.00019971897556501344, + "loss": 0.5925, + "step": 1028 + }, + { + "epoch": 0.024121380474340266, + "grad_norm": 3.465060234069824, + "learning_rate": 0.00019971842344234059, + "loss": 0.5809, + "step": 1029 + }, + { + "epoch": 0.02414482204914526, + "grad_norm": 1.435651421546936, + "learning_rate": 0.00019971787077859276, + "loss": 1.0425, + "step": 1030 + }, + { + "epoch": 0.02416826362395026, + "grad_norm": 3.146636486053467, + "learning_rate": 0.000199717317573773, + "loss": 0.5932, + "step": 1031 + }, + { + "epoch": 0.024191705198755253, + "grad_norm": 2.7868611812591553, + "learning_rate": 0.00019971676382788423, + "loss": 0.7059, + "step": 1032 + }, + { + "epoch": 0.024215146773560247, + "grad_norm": 4.466297149658203, + "learning_rate": 0.0001997162095409295, + "loss": 0.7014, + "step": 1033 + }, + { + "epoch": 0.024238588348365245, + "grad_norm": 1.3868666887283325, + "learning_rate": 0.0001997156547129118, + "loss": 0.2644, + "step": 1034 + }, + { + "epoch": 0.02426202992317024, + "grad_norm": 3.0148022174835205, + "learning_rate": 0.00019971509934383418, + "loss": 0.5793, + "step": 1035 + }, + { + "epoch": 0.024285471497975233, + "grad_norm": 2.544731855392456, + "learning_rate": 0.00019971454343369963, + "loss": 0.4409, + "step": 1036 + }, + { + "epoch": 0.024308913072780228, + "grad_norm": 0.754603385925293, + "learning_rate": 0.00019971398698251117, + "loss": 0.1136, + "step": 1037 + }, + { + "epoch": 0.024332354647585226, + "grad_norm": 4.877702236175537, + "learning_rate": 0.00019971342999027179, + "loss": 0.5223, + "step": 1038 + }, + { + "epoch": 0.02435579622239022, + "grad_norm": 1.3403840065002441, + "learning_rate": 0.00019971287245698453, + "loss": 0.2734, + "step": 1039 + }, + { + "epoch": 0.024379237797195214, + "grad_norm": 1.6571027040481567, + "learning_rate": 0.00019971231438265243, + "loss": 0.3098, + "step": 1040 + }, + { + "epoch": 0.024402679372000212, + "grad_norm": 3.492420196533203, + "learning_rate": 0.00019971175576727853, + "loss": 0.5859, + "step": 1041 + }, + { + "epoch": 0.024426120946805206, + "grad_norm": 3.028973340988159, + "learning_rate": 0.0001997111966108658, + "loss": 0.5916, + "step": 1042 + }, + { + "epoch": 0.0244495625216102, + "grad_norm": 2.8434407711029053, + "learning_rate": 0.00019971063691341735, + "loss": 0.4233, + "step": 1043 + }, + { + "epoch": 0.0244730040964152, + "grad_norm": 3.865539312362671, + "learning_rate": 0.00019971007667493615, + "loss": 0.5101, + "step": 1044 + }, + { + "epoch": 0.024496445671220193, + "grad_norm": 2.8829662799835205, + "learning_rate": 0.0001997095158954253, + "loss": 0.6563, + "step": 1045 + }, + { + "epoch": 0.024519887246025187, + "grad_norm": 4.796751499176025, + "learning_rate": 0.0001997089545748878, + "loss": 0.5441, + "step": 1046 + }, + { + "epoch": 0.024543328820830185, + "grad_norm": 3.4754583835601807, + "learning_rate": 0.00019970839271332672, + "loss": 0.7872, + "step": 1047 + }, + { + "epoch": 0.02456677039563518, + "grad_norm": 1.201236605644226, + "learning_rate": 0.00019970783031074508, + "loss": 0.2475, + "step": 1048 + }, + { + "epoch": 0.024590211970440173, + "grad_norm": 3.425419330596924, + "learning_rate": 0.00019970726736714592, + "loss": 0.5277, + "step": 1049 + }, + { + "epoch": 0.02461365354524517, + "grad_norm": 5.498969554901123, + "learning_rate": 0.00019970670388253238, + "loss": 0.9936, + "step": 1050 + }, + { + "epoch": 0.024637095120050165, + "grad_norm": 1.0402497053146362, + "learning_rate": 0.00019970613985690743, + "loss": 0.2156, + "step": 1051 + }, + { + "epoch": 0.02466053669485516, + "grad_norm": 3.393717050552368, + "learning_rate": 0.00019970557529027419, + "loss": 0.534, + "step": 1052 + }, + { + "epoch": 0.024683978269660154, + "grad_norm": 3.642763137817383, + "learning_rate": 0.00019970501018263566, + "loss": 0.3157, + "step": 1053 + }, + { + "epoch": 0.024707419844465152, + "grad_norm": 0.5276401042938232, + "learning_rate": 0.00019970444453399497, + "loss": 0.1071, + "step": 1054 + }, + { + "epoch": 0.024730861419270146, + "grad_norm": 4.658675193786621, + "learning_rate": 0.00019970387834435514, + "loss": 1.1764, + "step": 1055 + }, + { + "epoch": 0.02475430299407514, + "grad_norm": 5.439882755279541, + "learning_rate": 0.00019970331161371928, + "loss": 1.3605, + "step": 1056 + }, + { + "epoch": 0.02477774456888014, + "grad_norm": 1.418066382408142, + "learning_rate": 0.0001997027443420904, + "loss": 0.2558, + "step": 1057 + }, + { + "epoch": 0.024801186143685133, + "grad_norm": 4.718294620513916, + "learning_rate": 0.00019970217652947168, + "loss": 0.7885, + "step": 1058 + }, + { + "epoch": 0.024824627718490127, + "grad_norm": 4.589437484741211, + "learning_rate": 0.00019970160817586613, + "loss": 0.3697, + "step": 1059 + }, + { + "epoch": 0.024848069293295125, + "grad_norm": 1.5543129444122314, + "learning_rate": 0.00019970103928127687, + "loss": 0.2312, + "step": 1060 + }, + { + "epoch": 0.02487151086810012, + "grad_norm": 1.1536484956741333, + "learning_rate": 0.00019970046984570694, + "loss": 0.2212, + "step": 1061 + }, + { + "epoch": 0.024894952442905113, + "grad_norm": 2.826371192932129, + "learning_rate": 0.00019969989986915948, + "loss": 0.4443, + "step": 1062 + }, + { + "epoch": 0.02491839401771011, + "grad_norm": 5.020872116088867, + "learning_rate": 0.00019969932935163753, + "loss": 1.1377, + "step": 1063 + }, + { + "epoch": 0.024941835592515105, + "grad_norm": 3.023770809173584, + "learning_rate": 0.00019969875829314423, + "loss": 0.7729, + "step": 1064 + }, + { + "epoch": 0.0249652771673201, + "grad_norm": 4.821421146392822, + "learning_rate": 0.00019969818669368267, + "loss": 0.8183, + "step": 1065 + }, + { + "epoch": 0.024988718742125098, + "grad_norm": 1.5834143161773682, + "learning_rate": 0.00019969761455325595, + "loss": 0.294, + "step": 1066 + }, + { + "epoch": 0.025012160316930092, + "grad_norm": 2.2299208641052246, + "learning_rate": 0.00019969704187186715, + "loss": 1.6396, + "step": 1067 + }, + { + "epoch": 0.025035601891735086, + "grad_norm": 3.1288366317749023, + "learning_rate": 0.00019969646864951944, + "loss": 0.6058, + "step": 1068 + }, + { + "epoch": 0.02505904346654008, + "grad_norm": 5.372637748718262, + "learning_rate": 0.00019969589488621583, + "loss": 0.7131, + "step": 1069 + }, + { + "epoch": 0.025082485041345078, + "grad_norm": 1.770721673965454, + "learning_rate": 0.00019969532058195953, + "loss": 0.1952, + "step": 1070 + }, + { + "epoch": 0.025105926616150073, + "grad_norm": 2.1216142177581787, + "learning_rate": 0.0001996947457367536, + "loss": 0.5161, + "step": 1071 + }, + { + "epoch": 0.025129368190955067, + "grad_norm": 0.5803524255752563, + "learning_rate": 0.00019969417035060118, + "loss": 0.1329, + "step": 1072 + }, + { + "epoch": 0.025152809765760065, + "grad_norm": 1.110059142112732, + "learning_rate": 0.0001996935944235054, + "loss": 1.8042, + "step": 1073 + }, + { + "epoch": 0.02517625134056506, + "grad_norm": 0.7866904139518738, + "learning_rate": 0.00019969301795546936, + "loss": 0.159, + "step": 1074 + }, + { + "epoch": 0.025199692915370053, + "grad_norm": 3.7544429302215576, + "learning_rate": 0.0001996924409464962, + "loss": 0.7249, + "step": 1075 + }, + { + "epoch": 0.02522313449017505, + "grad_norm": 3.580228090286255, + "learning_rate": 0.00019969186339658906, + "loss": 0.5528, + "step": 1076 + }, + { + "epoch": 0.025246576064980045, + "grad_norm": 2.7455544471740723, + "learning_rate": 0.00019969128530575105, + "loss": 0.6134, + "step": 1077 + }, + { + "epoch": 0.02527001763978504, + "grad_norm": 3.23110294342041, + "learning_rate": 0.00019969070667398534, + "loss": 0.8488, + "step": 1078 + }, + { + "epoch": 0.025293459214590037, + "grad_norm": 4.832276344299316, + "learning_rate": 0.00019969012750129507, + "loss": 1.1587, + "step": 1079 + }, + { + "epoch": 0.025316900789395032, + "grad_norm": 0.9913850426673889, + "learning_rate": 0.00019968954778768334, + "loss": 0.0899, + "step": 1080 + }, + { + "epoch": 0.025340342364200026, + "grad_norm": 2.2204015254974365, + "learning_rate": 0.00019968896753315332, + "loss": 0.3876, + "step": 1081 + }, + { + "epoch": 0.025363783939005024, + "grad_norm": 3.523829698562622, + "learning_rate": 0.00019968838673770819, + "loss": 0.6238, + "step": 1082 + }, + { + "epoch": 0.025387225513810018, + "grad_norm": 1.4841229915618896, + "learning_rate": 0.00019968780540135104, + "loss": 0.2876, + "step": 1083 + }, + { + "epoch": 0.025410667088615013, + "grad_norm": 2.3214986324310303, + "learning_rate": 0.00019968722352408505, + "loss": 0.443, + "step": 1084 + }, + { + "epoch": 0.025434108663420007, + "grad_norm": 1.0359705686569214, + "learning_rate": 0.00019968664110591336, + "loss": 0.1593, + "step": 1085 + }, + { + "epoch": 0.025457550238225005, + "grad_norm": 0.5294941663742065, + "learning_rate": 0.0001996860581468392, + "loss": 0.0711, + "step": 1086 + }, + { + "epoch": 0.02548099181303, + "grad_norm": 2.4033849239349365, + "learning_rate": 0.00019968547464686566, + "loss": 0.5067, + "step": 1087 + }, + { + "epoch": 0.025504433387834993, + "grad_norm": 4.359287261962891, + "learning_rate": 0.00019968489060599596, + "loss": 0.8563, + "step": 1088 + }, + { + "epoch": 0.02552787496263999, + "grad_norm": 1.6502361297607422, + "learning_rate": 0.00019968430602423324, + "loss": 1.4724, + "step": 1089 + }, + { + "epoch": 0.025551316537444985, + "grad_norm": 2.774583578109741, + "learning_rate": 0.00019968372090158062, + "loss": 0.3334, + "step": 1090 + }, + { + "epoch": 0.02557475811224998, + "grad_norm": 4.092152118682861, + "learning_rate": 0.00019968313523804137, + "loss": 0.6744, + "step": 1091 + }, + { + "epoch": 0.025598199687054977, + "grad_norm": 1.2351210117340088, + "learning_rate": 0.00019968254903361863, + "loss": 1.3261, + "step": 1092 + }, + { + "epoch": 0.025621641261859972, + "grad_norm": 2.0266213417053223, + "learning_rate": 0.00019968196228831557, + "loss": 1.5877, + "step": 1093 + }, + { + "epoch": 0.025645082836664966, + "grad_norm": 3.7109196186065674, + "learning_rate": 0.00019968137500213536, + "loss": 0.8982, + "step": 1094 + }, + { + "epoch": 0.025668524411469964, + "grad_norm": 3.456163167953491, + "learning_rate": 0.00019968078717508123, + "loss": 0.8667, + "step": 1095 + }, + { + "epoch": 0.025691965986274958, + "grad_norm": 2.0728588104248047, + "learning_rate": 0.00019968019880715635, + "loss": 1.4014, + "step": 1096 + }, + { + "epoch": 0.025715407561079952, + "grad_norm": 4.01130485534668, + "learning_rate": 0.00019967960989836388, + "loss": 0.6151, + "step": 1097 + }, + { + "epoch": 0.02573884913588495, + "grad_norm": 3.9537618160247803, + "learning_rate": 0.00019967902044870704, + "loss": 1.1159, + "step": 1098 + }, + { + "epoch": 0.025762290710689945, + "grad_norm": 3.32470440864563, + "learning_rate": 0.0001996784304581891, + "loss": 0.7356, + "step": 1099 + }, + { + "epoch": 0.02578573228549494, + "grad_norm": 4.559971809387207, + "learning_rate": 0.0001996778399268131, + "loss": 0.7794, + "step": 1100 + }, + { + "epoch": 0.025809173860299937, + "grad_norm": 1.0745630264282227, + "learning_rate": 0.0001996772488545824, + "loss": 0.3047, + "step": 1101 + }, + { + "epoch": 0.02583261543510493, + "grad_norm": 5.339470863342285, + "learning_rate": 0.00019967665724150013, + "loss": 0.608, + "step": 1102 + }, + { + "epoch": 0.025856057009909925, + "grad_norm": 3.0139641761779785, + "learning_rate": 0.0001996760650875695, + "loss": 0.5647, + "step": 1103 + }, + { + "epoch": 0.02587949858471492, + "grad_norm": 2.7131259441375732, + "learning_rate": 0.00019967547239279376, + "loss": 0.6332, + "step": 1104 + }, + { + "epoch": 0.025902940159519917, + "grad_norm": 3.5286412239074707, + "learning_rate": 0.0001996748791571761, + "loss": 0.5435, + "step": 1105 + }, + { + "epoch": 0.02592638173432491, + "grad_norm": 1.417141079902649, + "learning_rate": 0.00019967428538071973, + "loss": 0.1686, + "step": 1106 + }, + { + "epoch": 0.025949823309129906, + "grad_norm": 4.490147590637207, + "learning_rate": 0.0001996736910634279, + "loss": 0.6793, + "step": 1107 + }, + { + "epoch": 0.025973264883934904, + "grad_norm": 1.4076964855194092, + "learning_rate": 0.0001996730962053038, + "loss": 0.2725, + "step": 1108 + }, + { + "epoch": 0.025996706458739898, + "grad_norm": 0.39417871832847595, + "learning_rate": 0.0001996725008063507, + "loss": 0.097, + "step": 1109 + }, + { + "epoch": 0.026020148033544892, + "grad_norm": 1.22563898563385, + "learning_rate": 0.0001996719048665718, + "loss": 0.2156, + "step": 1110 + }, + { + "epoch": 0.02604358960834989, + "grad_norm": 1.294275164604187, + "learning_rate": 0.00019967130838597036, + "loss": 0.3146, + "step": 1111 + }, + { + "epoch": 0.026067031183154885, + "grad_norm": 2.5709245204925537, + "learning_rate": 0.00019967071136454956, + "loss": 0.4785, + "step": 1112 + }, + { + "epoch": 0.02609047275795988, + "grad_norm": 4.972354412078857, + "learning_rate": 0.0001996701138023127, + "loss": 1.0372, + "step": 1113 + }, + { + "epoch": 0.026113914332764877, + "grad_norm": 0.9414512515068054, + "learning_rate": 0.000199669515699263, + "loss": 0.2832, + "step": 1114 + }, + { + "epoch": 0.02613735590756987, + "grad_norm": 2.865295648574829, + "learning_rate": 0.0001996689170554037, + "loss": 0.6096, + "step": 1115 + }, + { + "epoch": 0.026160797482374865, + "grad_norm": 1.7538690567016602, + "learning_rate": 0.00019966831787073805, + "loss": 0.4526, + "step": 1116 + }, + { + "epoch": 0.026184239057179863, + "grad_norm": 2.0934932231903076, + "learning_rate": 0.0001996677181452693, + "loss": 0.2984, + "step": 1117 + }, + { + "epoch": 0.026207680631984857, + "grad_norm": 4.485007286071777, + "learning_rate": 0.00019966711787900072, + "loss": 0.9116, + "step": 1118 + }, + { + "epoch": 0.02623112220678985, + "grad_norm": 2.8767647743225098, + "learning_rate": 0.00019966651707193554, + "loss": 0.573, + "step": 1119 + }, + { + "epoch": 0.026254563781594846, + "grad_norm": 4.529285430908203, + "learning_rate": 0.00019966591572407707, + "loss": 0.9767, + "step": 1120 + }, + { + "epoch": 0.026278005356399844, + "grad_norm": 2.285475969314575, + "learning_rate": 0.00019966531383542853, + "loss": 1.7868, + "step": 1121 + }, + { + "epoch": 0.026301446931204838, + "grad_norm": 1.626612663269043, + "learning_rate": 0.00019966471140599317, + "loss": 1.3974, + "step": 1122 + }, + { + "epoch": 0.026324888506009832, + "grad_norm": 2.756535053253174, + "learning_rate": 0.0001996641084357743, + "loss": 1.8856, + "step": 1123 + }, + { + "epoch": 0.02634833008081483, + "grad_norm": 3.423966407775879, + "learning_rate": 0.00019966350492477516, + "loss": 0.8072, + "step": 1124 + }, + { + "epoch": 0.026371771655619825, + "grad_norm": 4.346046447753906, + "learning_rate": 0.00019966290087299906, + "loss": 0.8455, + "step": 1125 + }, + { + "epoch": 0.02639521323042482, + "grad_norm": 3.7210421562194824, + "learning_rate": 0.00019966229628044926, + "loss": 0.5723, + "step": 1126 + }, + { + "epoch": 0.026418654805229817, + "grad_norm": 4.331787586212158, + "learning_rate": 0.00019966169114712901, + "loss": 0.7159, + "step": 1127 + }, + { + "epoch": 0.02644209638003481, + "grad_norm": 2.8781609535217285, + "learning_rate": 0.00019966108547304166, + "loss": 0.5067, + "step": 1128 + }, + { + "epoch": 0.026465537954839805, + "grad_norm": 3.29894757270813, + "learning_rate": 0.00019966047925819042, + "loss": 0.667, + "step": 1129 + }, + { + "epoch": 0.026488979529644803, + "grad_norm": 2.4833688735961914, + "learning_rate": 0.00019965987250257866, + "loss": 1.9369, + "step": 1130 + }, + { + "epoch": 0.026512421104449797, + "grad_norm": 4.628660678863525, + "learning_rate": 0.00019965926520620962, + "loss": 0.5536, + "step": 1131 + }, + { + "epoch": 0.02653586267925479, + "grad_norm": 1.30317223072052, + "learning_rate": 0.00019965865736908657, + "loss": 0.2317, + "step": 1132 + }, + { + "epoch": 0.02655930425405979, + "grad_norm": 2.2422749996185303, + "learning_rate": 0.00019965804899121285, + "loss": 0.4395, + "step": 1133 + }, + { + "epoch": 0.026582745828864784, + "grad_norm": 5.208040714263916, + "learning_rate": 0.00019965744007259176, + "loss": 1.3013, + "step": 1134 + }, + { + "epoch": 0.026606187403669778, + "grad_norm": 0.5567788481712341, + "learning_rate": 0.00019965683061322662, + "loss": 0.0447, + "step": 1135 + }, + { + "epoch": 0.026629628978474772, + "grad_norm": 2.147249698638916, + "learning_rate": 0.0001996562206131207, + "loss": 0.3882, + "step": 1136 + }, + { + "epoch": 0.02665307055327977, + "grad_norm": 5.000513076782227, + "learning_rate": 0.00019965561007227734, + "loss": 0.6793, + "step": 1137 + }, + { + "epoch": 0.026676512128084764, + "grad_norm": 2.4702889919281006, + "learning_rate": 0.00019965499899069982, + "loss": 0.467, + "step": 1138 + }, + { + "epoch": 0.02669995370288976, + "grad_norm": 1.3538904190063477, + "learning_rate": 0.00019965438736839147, + "loss": 0.1942, + "step": 1139 + }, + { + "epoch": 0.026723395277694757, + "grad_norm": 3.8638691902160645, + "learning_rate": 0.00019965377520535563, + "loss": 0.6482, + "step": 1140 + }, + { + "epoch": 0.02674683685249975, + "grad_norm": 2.369626760482788, + "learning_rate": 0.0001996531625015956, + "loss": 1.772, + "step": 1141 + }, + { + "epoch": 0.026770278427304745, + "grad_norm": 1.9237743616104126, + "learning_rate": 0.00019965254925711468, + "loss": 1.9029, + "step": 1142 + }, + { + "epoch": 0.026793720002109743, + "grad_norm": 3.743640899658203, + "learning_rate": 0.00019965193547191626, + "loss": 0.687, + "step": 1143 + }, + { + "epoch": 0.026817161576914737, + "grad_norm": 3.323002815246582, + "learning_rate": 0.0001996513211460036, + "loss": 0.8012, + "step": 1144 + }, + { + "epoch": 0.02684060315171973, + "grad_norm": 1.2197184562683105, + "learning_rate": 0.00019965070627938012, + "loss": 0.1246, + "step": 1145 + }, + { + "epoch": 0.02686404472652473, + "grad_norm": 3.347782850265503, + "learning_rate": 0.00019965009087204906, + "loss": 0.3814, + "step": 1146 + }, + { + "epoch": 0.026887486301329724, + "grad_norm": 3.836097478866577, + "learning_rate": 0.0001996494749240138, + "loss": 0.7046, + "step": 1147 + }, + { + "epoch": 0.026910927876134718, + "grad_norm": 4.5958075523376465, + "learning_rate": 0.00019964885843527772, + "loss": 0.868, + "step": 1148 + }, + { + "epoch": 0.026934369450939716, + "grad_norm": 1.2364380359649658, + "learning_rate": 0.00019964824140584412, + "loss": 0.2489, + "step": 1149 + }, + { + "epoch": 0.02695781102574471, + "grad_norm": 3.338702917098999, + "learning_rate": 0.00019964762383571638, + "loss": 0.4686, + "step": 1150 + }, + { + "epoch": 0.026981252600549704, + "grad_norm": 5.349024772644043, + "learning_rate": 0.0001996470057248978, + "loss": 0.934, + "step": 1151 + }, + { + "epoch": 0.0270046941753547, + "grad_norm": 0.9302285313606262, + "learning_rate": 0.00019964638707339178, + "loss": 0.1353, + "step": 1152 + }, + { + "epoch": 0.027028135750159697, + "grad_norm": 3.7893481254577637, + "learning_rate": 0.0001996457678812017, + "loss": 0.5245, + "step": 1153 + }, + { + "epoch": 0.02705157732496469, + "grad_norm": 0.7556709051132202, + "learning_rate": 0.00019964514814833082, + "loss": 0.1265, + "step": 1154 + }, + { + "epoch": 0.027075018899769685, + "grad_norm": 3.7760026454925537, + "learning_rate": 0.00019964452787478258, + "loss": 0.7626, + "step": 1155 + }, + { + "epoch": 0.027098460474574683, + "grad_norm": 1.0842119455337524, + "learning_rate": 0.00019964390706056034, + "loss": 0.2208, + "step": 1156 + }, + { + "epoch": 0.027121902049379677, + "grad_norm": 0.3649764955043793, + "learning_rate": 0.00019964328570566746, + "loss": 0.0521, + "step": 1157 + }, + { + "epoch": 0.02714534362418467, + "grad_norm": 0.5912774205207825, + "learning_rate": 0.0001996426638101073, + "loss": 0.1357, + "step": 1158 + }, + { + "epoch": 0.02716878519898967, + "grad_norm": 1.614464521408081, + "learning_rate": 0.00019964204137388325, + "loss": 1.3616, + "step": 1159 + }, + { + "epoch": 0.027192226773794664, + "grad_norm": 2.2469265460968018, + "learning_rate": 0.00019964141839699868, + "loss": 0.3569, + "step": 1160 + }, + { + "epoch": 0.027215668348599658, + "grad_norm": 3.1718735694885254, + "learning_rate": 0.00019964079487945697, + "loss": 0.4649, + "step": 1161 + }, + { + "epoch": 0.027239109923404656, + "grad_norm": 4.00747013092041, + "learning_rate": 0.00019964017082126152, + "loss": 1.0869, + "step": 1162 + }, + { + "epoch": 0.02726255149820965, + "grad_norm": 1.5436204671859741, + "learning_rate": 0.0001996395462224157, + "loss": 0.3654, + "step": 1163 + }, + { + "epoch": 0.027285993073014644, + "grad_norm": 4.701499938964844, + "learning_rate": 0.00019963892108292288, + "loss": 0.8225, + "step": 1164 + }, + { + "epoch": 0.027309434647819642, + "grad_norm": 3.393373489379883, + "learning_rate": 0.0001996382954027865, + "loss": 0.6146, + "step": 1165 + }, + { + "epoch": 0.027332876222624636, + "grad_norm": 3.2566258907318115, + "learning_rate": 0.00019963766918200987, + "loss": 0.5802, + "step": 1166 + }, + { + "epoch": 0.02735631779742963, + "grad_norm": 4.024772644042969, + "learning_rate": 0.0001996370424205965, + "loss": 0.9603, + "step": 1167 + }, + { + "epoch": 0.027379759372234625, + "grad_norm": 0.9802177548408508, + "learning_rate": 0.00019963641511854975, + "loss": 0.1655, + "step": 1168 + }, + { + "epoch": 0.027403200947039623, + "grad_norm": 2.1336331367492676, + "learning_rate": 0.00019963578727587295, + "loss": 0.5783, + "step": 1169 + }, + { + "epoch": 0.027426642521844617, + "grad_norm": 3.7546803951263428, + "learning_rate": 0.0001996351588925696, + "loss": 0.7313, + "step": 1170 + }, + { + "epoch": 0.02745008409664961, + "grad_norm": 2.91355562210083, + "learning_rate": 0.0001996345299686431, + "loss": 0.4653, + "step": 1171 + }, + { + "epoch": 0.02747352567145461, + "grad_norm": 3.9304492473602295, + "learning_rate": 0.0001996339005040968, + "loss": 0.8285, + "step": 1172 + }, + { + "epoch": 0.027496967246259604, + "grad_norm": 1.195072889328003, + "learning_rate": 0.00019963327049893418, + "loss": 0.2021, + "step": 1173 + }, + { + "epoch": 0.027520408821064598, + "grad_norm": 3.5971767902374268, + "learning_rate": 0.00019963263995315862, + "loss": 0.9031, + "step": 1174 + }, + { + "epoch": 0.027543850395869596, + "grad_norm": 3.760336399078369, + "learning_rate": 0.00019963200886677357, + "loss": 0.8326, + "step": 1175 + }, + { + "epoch": 0.02756729197067459, + "grad_norm": 3.2634079456329346, + "learning_rate": 0.00019963137723978244, + "loss": 0.5291, + "step": 1176 + }, + { + "epoch": 0.027590733545479584, + "grad_norm": 1.7774070501327515, + "learning_rate": 0.0001996307450721886, + "loss": 1.5771, + "step": 1177 + }, + { + "epoch": 0.027614175120284582, + "grad_norm": 3.011507511138916, + "learning_rate": 0.0001996301123639956, + "loss": 0.6764, + "step": 1178 + }, + { + "epoch": 0.027637616695089576, + "grad_norm": 1.3557384014129639, + "learning_rate": 0.0001996294791152068, + "loss": 0.2094, + "step": 1179 + }, + { + "epoch": 0.02766105826989457, + "grad_norm": 1.5868808031082153, + "learning_rate": 0.0001996288453258256, + "loss": 2.1165, + "step": 1180 + }, + { + "epoch": 0.02768449984469957, + "grad_norm": 0.8716061115264893, + "learning_rate": 0.0001996282109958555, + "loss": 0.1309, + "step": 1181 + }, + { + "epoch": 0.027707941419504563, + "grad_norm": 3.3663597106933594, + "learning_rate": 0.00019962757612529996, + "loss": 0.6077, + "step": 1182 + }, + { + "epoch": 0.027731382994309557, + "grad_norm": 3.092433214187622, + "learning_rate": 0.00019962694071416236, + "loss": 0.6015, + "step": 1183 + }, + { + "epoch": 0.02775482456911455, + "grad_norm": 4.216604232788086, + "learning_rate": 0.00019962630476244618, + "loss": 1.0276, + "step": 1184 + }, + { + "epoch": 0.02777826614391955, + "grad_norm": 4.159610271453857, + "learning_rate": 0.00019962566827015488, + "loss": 0.7574, + "step": 1185 + }, + { + "epoch": 0.027801707718724544, + "grad_norm": 0.904468834400177, + "learning_rate": 0.0001996250312372919, + "loss": 0.1963, + "step": 1186 + }, + { + "epoch": 0.027825149293529538, + "grad_norm": 2.575761079788208, + "learning_rate": 0.00019962439366386073, + "loss": 0.5932, + "step": 1187 + }, + { + "epoch": 0.027848590868334536, + "grad_norm": 1.7530728578567505, + "learning_rate": 0.00019962375554986473, + "loss": 1.0831, + "step": 1188 + }, + { + "epoch": 0.02787203244313953, + "grad_norm": 5.369051933288574, + "learning_rate": 0.00019962311689530752, + "loss": 0.7738, + "step": 1189 + }, + { + "epoch": 0.027895474017944524, + "grad_norm": 4.034156322479248, + "learning_rate": 0.0001996224777001924, + "loss": 0.5908, + "step": 1190 + }, + { + "epoch": 0.027918915592749522, + "grad_norm": 1.0660951137542725, + "learning_rate": 0.00019962183796452294, + "loss": 0.1637, + "step": 1191 + }, + { + "epoch": 0.027942357167554516, + "grad_norm": 1.3246526718139648, + "learning_rate": 0.00019962119768830258, + "loss": 0.193, + "step": 1192 + }, + { + "epoch": 0.02796579874235951, + "grad_norm": 2.0979034900665283, + "learning_rate": 0.00019962055687153482, + "loss": 0.2846, + "step": 1193 + }, + { + "epoch": 0.02798924031716451, + "grad_norm": 4.004546165466309, + "learning_rate": 0.0001996199155142231, + "loss": 1.0261, + "step": 1194 + }, + { + "epoch": 0.028012681891969503, + "grad_norm": 4.082563877105713, + "learning_rate": 0.00019961927361637094, + "loss": 0.7035, + "step": 1195 + }, + { + "epoch": 0.028036123466774497, + "grad_norm": 4.873593330383301, + "learning_rate": 0.00019961863117798177, + "loss": 0.8728, + "step": 1196 + }, + { + "epoch": 0.028059565041579495, + "grad_norm": 3.0935745239257812, + "learning_rate": 0.00019961798819905912, + "loss": 0.626, + "step": 1197 + }, + { + "epoch": 0.02808300661638449, + "grad_norm": 4.573702812194824, + "learning_rate": 0.00019961734467960647, + "loss": 0.9208, + "step": 1198 + }, + { + "epoch": 0.028106448191189484, + "grad_norm": 3.9455137252807617, + "learning_rate": 0.00019961670061962732, + "loss": 1.0088, + "step": 1199 + }, + { + "epoch": 0.028129889765994478, + "grad_norm": 1.530962586402893, + "learning_rate": 0.0001996160560191251, + "loss": 0.2151, + "step": 1200 + }, + { + "epoch": 0.028153331340799476, + "grad_norm": 2.6433253288269043, + "learning_rate": 0.0001996154108781034, + "loss": 0.3405, + "step": 1201 + }, + { + "epoch": 0.02817677291560447, + "grad_norm": 2.938983201980591, + "learning_rate": 0.00019961476519656567, + "loss": 0.555, + "step": 1202 + }, + { + "epoch": 0.028200214490409464, + "grad_norm": 5.3130269050598145, + "learning_rate": 0.00019961411897451544, + "loss": 0.7517, + "step": 1203 + }, + { + "epoch": 0.028223656065214462, + "grad_norm": 3.3103320598602295, + "learning_rate": 0.00019961347221195616, + "loss": 0.552, + "step": 1204 + }, + { + "epoch": 0.028247097640019456, + "grad_norm": 1.2897450923919678, + "learning_rate": 0.0001996128249088914, + "loss": 0.2406, + "step": 1205 + }, + { + "epoch": 0.02827053921482445, + "grad_norm": 3.0365922451019287, + "learning_rate": 0.00019961217706532467, + "loss": 0.6861, + "step": 1206 + }, + { + "epoch": 0.02829398078962945, + "grad_norm": 5.116939067840576, + "learning_rate": 0.00019961152868125945, + "loss": 0.61, + "step": 1207 + }, + { + "epoch": 0.028317422364434443, + "grad_norm": 3.0416107177734375, + "learning_rate": 0.0001996108797566993, + "loss": 0.6339, + "step": 1208 + }, + { + "epoch": 0.028340863939239437, + "grad_norm": 3.6158649921417236, + "learning_rate": 0.00019961023029164767, + "loss": 0.6336, + "step": 1209 + }, + { + "epoch": 0.028364305514044435, + "grad_norm": 2.6988399028778076, + "learning_rate": 0.00019960958028610818, + "loss": 0.6173, + "step": 1210 + }, + { + "epoch": 0.02838774708884943, + "grad_norm": 2.1576969623565674, + "learning_rate": 0.00019960892974008425, + "loss": 1.956, + "step": 1211 + }, + { + "epoch": 0.028411188663654423, + "grad_norm": 2.2499072551727295, + "learning_rate": 0.0001996082786535795, + "loss": 2.072, + "step": 1212 + }, + { + "epoch": 0.02843463023845942, + "grad_norm": 2.8535263538360596, + "learning_rate": 0.00019960762702659742, + "loss": 0.5957, + "step": 1213 + }, + { + "epoch": 0.028458071813264416, + "grad_norm": 1.6727155447006226, + "learning_rate": 0.00019960697485914156, + "loss": 1.4186, + "step": 1214 + }, + { + "epoch": 0.02848151338806941, + "grad_norm": 4.135761260986328, + "learning_rate": 0.00019960632215121544, + "loss": 0.4956, + "step": 1215 + }, + { + "epoch": 0.028504954962874404, + "grad_norm": 2.6198506355285645, + "learning_rate": 0.00019960566890282264, + "loss": 1.9039, + "step": 1216 + }, + { + "epoch": 0.028528396537679402, + "grad_norm": 4.249276638031006, + "learning_rate": 0.00019960501511396666, + "loss": 0.7713, + "step": 1217 + }, + { + "epoch": 0.028551838112484396, + "grad_norm": 1.8280118703842163, + "learning_rate": 0.00019960436078465108, + "loss": 0.2824, + "step": 1218 + }, + { + "epoch": 0.02857527968728939, + "grad_norm": 1.9950226545333862, + "learning_rate": 0.00019960370591487941, + "loss": 1.8165, + "step": 1219 + }, + { + "epoch": 0.02859872126209439, + "grad_norm": 5.787414073944092, + "learning_rate": 0.00019960305050465527, + "loss": 1.1296, + "step": 1220 + }, + { + "epoch": 0.028622162836899383, + "grad_norm": 1.0362012386322021, + "learning_rate": 0.00019960239455398216, + "loss": 0.1531, + "step": 1221 + }, + { + "epoch": 0.028645604411704377, + "grad_norm": 2.512284755706787, + "learning_rate": 0.00019960173806286364, + "loss": 0.5409, + "step": 1222 + }, + { + "epoch": 0.028669045986509375, + "grad_norm": 3.1950013637542725, + "learning_rate": 0.00019960108103130332, + "loss": 0.4272, + "step": 1223 + }, + { + "epoch": 0.02869248756131437, + "grad_norm": 2.801780939102173, + "learning_rate": 0.00019960042345930473, + "loss": 0.3938, + "step": 1224 + }, + { + "epoch": 0.028715929136119363, + "grad_norm": 3.5640933513641357, + "learning_rate": 0.00019959976534687145, + "loss": 0.5802, + "step": 1225 + }, + { + "epoch": 0.02873937071092436, + "grad_norm": 3.931225299835205, + "learning_rate": 0.00019959910669400702, + "loss": 0.4221, + "step": 1226 + }, + { + "epoch": 0.028762812285729356, + "grad_norm": 2.5221078395843506, + "learning_rate": 0.000199598447500715, + "loss": 0.4491, + "step": 1227 + }, + { + "epoch": 0.02878625386053435, + "grad_norm": 3.7656984329223633, + "learning_rate": 0.00019959778776699905, + "loss": 0.58, + "step": 1228 + }, + { + "epoch": 0.028809695435339348, + "grad_norm": 3.6105501651763916, + "learning_rate": 0.0001995971274928627, + "loss": 0.7504, + "step": 1229 + }, + { + "epoch": 0.028833137010144342, + "grad_norm": 3.3828892707824707, + "learning_rate": 0.00019959646667830954, + "loss": 0.3389, + "step": 1230 + }, + { + "epoch": 0.028856578584949336, + "grad_norm": 0.8096657991409302, + "learning_rate": 0.00019959580532334314, + "loss": 0.1124, + "step": 1231 + }, + { + "epoch": 0.02888002015975433, + "grad_norm": 3.53448224067688, + "learning_rate": 0.00019959514342796712, + "loss": 0.4195, + "step": 1232 + }, + { + "epoch": 0.02890346173455933, + "grad_norm": 0.8397369980812073, + "learning_rate": 0.00019959448099218503, + "loss": 0.1564, + "step": 1233 + }, + { + "epoch": 0.028926903309364323, + "grad_norm": 3.323415517807007, + "learning_rate": 0.0001995938180160005, + "loss": 0.328, + "step": 1234 + }, + { + "epoch": 0.028950344884169317, + "grad_norm": 2.545639753341675, + "learning_rate": 0.00019959315449941708, + "loss": 0.2404, + "step": 1235 + }, + { + "epoch": 0.028973786458974315, + "grad_norm": 4.29265832901001, + "learning_rate": 0.00019959249044243843, + "loss": 0.4407, + "step": 1236 + }, + { + "epoch": 0.02899722803377931, + "grad_norm": 2.369955539703369, + "learning_rate": 0.00019959182584506816, + "loss": 1.9425, + "step": 1237 + }, + { + "epoch": 0.029020669608584303, + "grad_norm": 3.7275729179382324, + "learning_rate": 0.00019959116070730976, + "loss": 0.4, + "step": 1238 + }, + { + "epoch": 0.0290441111833893, + "grad_norm": 2.584988832473755, + "learning_rate": 0.00019959049502916697, + "loss": 0.4591, + "step": 1239 + }, + { + "epoch": 0.029067552758194296, + "grad_norm": 8.300516128540039, + "learning_rate": 0.0001995898288106434, + "loss": 1.106, + "step": 1240 + }, + { + "epoch": 0.02909099433299929, + "grad_norm": 3.465944528579712, + "learning_rate": 0.00019958916205174254, + "loss": 0.5377, + "step": 1241 + }, + { + "epoch": 0.029114435907804288, + "grad_norm": 1.1016298532485962, + "learning_rate": 0.00019958849475246813, + "loss": 0.9138, + "step": 1242 + }, + { + "epoch": 0.029137877482609282, + "grad_norm": 2.2083382606506348, + "learning_rate": 0.00019958782691282374, + "loss": 1.5727, + "step": 1243 + }, + { + "epoch": 0.029161319057414276, + "grad_norm": 1.3672467470169067, + "learning_rate": 0.00019958715853281297, + "loss": 1.1294, + "step": 1244 + }, + { + "epoch": 0.029184760632219274, + "grad_norm": 3.889521598815918, + "learning_rate": 0.0001995864896124395, + "loss": 0.6528, + "step": 1245 + }, + { + "epoch": 0.02920820220702427, + "grad_norm": 1.9500340223312378, + "learning_rate": 0.00019958582015170695, + "loss": 0.3618, + "step": 1246 + }, + { + "epoch": 0.029231643781829263, + "grad_norm": 5.6016764640808105, + "learning_rate": 0.00019958515015061894, + "loss": 1.0557, + "step": 1247 + }, + { + "epoch": 0.029255085356634257, + "grad_norm": 4.664235591888428, + "learning_rate": 0.00019958447960917908, + "loss": 0.7171, + "step": 1248 + }, + { + "epoch": 0.029278526931439255, + "grad_norm": 6.363917350769043, + "learning_rate": 0.00019958380852739106, + "loss": 0.7192, + "step": 1249 + }, + { + "epoch": 0.02930196850624425, + "grad_norm": 5.313285827636719, + "learning_rate": 0.00019958313690525848, + "loss": 0.69, + "step": 1250 + }, + { + "epoch": 0.029325410081049243, + "grad_norm": 1.2469184398651123, + "learning_rate": 0.00019958246474278501, + "loss": 0.1795, + "step": 1251 + }, + { + "epoch": 0.02934885165585424, + "grad_norm": 2.1203901767730713, + "learning_rate": 0.0001995817920399743, + "loss": 1.3037, + "step": 1252 + }, + { + "epoch": 0.029372293230659235, + "grad_norm": 3.7982752323150635, + "learning_rate": 0.00019958111879682996, + "loss": 0.423, + "step": 1253 + }, + { + "epoch": 0.02939573480546423, + "grad_norm": 3.844395399093628, + "learning_rate": 0.00019958044501335566, + "loss": 0.5366, + "step": 1254 + }, + { + "epoch": 0.029419176380269228, + "grad_norm": 1.8826206922531128, + "learning_rate": 0.00019957977068955506, + "loss": 0.5303, + "step": 1255 + }, + { + "epoch": 0.029442617955074222, + "grad_norm": 0.7489196062088013, + "learning_rate": 0.00019957909582543187, + "loss": 0.1129, + "step": 1256 + }, + { + "epoch": 0.029466059529879216, + "grad_norm": 5.450354099273682, + "learning_rate": 0.00019957842042098967, + "loss": 0.7031, + "step": 1257 + }, + { + "epoch": 0.029489501104684214, + "grad_norm": 3.2842183113098145, + "learning_rate": 0.00019957774447623217, + "loss": 0.5718, + "step": 1258 + }, + { + "epoch": 0.02951294267948921, + "grad_norm": 4.383791923522949, + "learning_rate": 0.00019957706799116304, + "loss": 0.4581, + "step": 1259 + }, + { + "epoch": 0.029536384254294203, + "grad_norm": 0.5316231846809387, + "learning_rate": 0.00019957639096578592, + "loss": 0.0863, + "step": 1260 + }, + { + "epoch": 0.0295598258290992, + "grad_norm": 3.0478148460388184, + "learning_rate": 0.00019957571340010452, + "loss": 0.435, + "step": 1261 + }, + { + "epoch": 0.029583267403904195, + "grad_norm": 1.6189783811569214, + "learning_rate": 0.00019957503529412248, + "loss": 0.3437, + "step": 1262 + }, + { + "epoch": 0.02960670897870919, + "grad_norm": 4.555883884429932, + "learning_rate": 0.00019957435664784355, + "loss": 0.9113, + "step": 1263 + }, + { + "epoch": 0.029630150553514187, + "grad_norm": 2.3884212970733643, + "learning_rate": 0.0001995736774612713, + "loss": 0.4824, + "step": 1264 + }, + { + "epoch": 0.02965359212831918, + "grad_norm": 4.178380489349365, + "learning_rate": 0.0001995729977344095, + "loss": 1.0932, + "step": 1265 + }, + { + "epoch": 0.029677033703124175, + "grad_norm": 2.407578468322754, + "learning_rate": 0.0001995723174672618, + "loss": 0.2664, + "step": 1266 + }, + { + "epoch": 0.02970047527792917, + "grad_norm": 3.3854522705078125, + "learning_rate": 0.00019957163665983193, + "loss": 0.5159, + "step": 1267 + }, + { + "epoch": 0.029723916852734168, + "grad_norm": 5.53548002243042, + "learning_rate": 0.00019957095531212356, + "loss": 0.9523, + "step": 1268 + }, + { + "epoch": 0.029747358427539162, + "grad_norm": 1.062099814414978, + "learning_rate": 0.00019957027342414036, + "loss": 0.9297, + "step": 1269 + }, + { + "epoch": 0.029770800002344156, + "grad_norm": 1.220470905303955, + "learning_rate": 0.00019956959099588607, + "loss": 0.1913, + "step": 1270 + }, + { + "epoch": 0.029794241577149154, + "grad_norm": 1.098376750946045, + "learning_rate": 0.00019956890802736438, + "loss": 0.199, + "step": 1271 + }, + { + "epoch": 0.029817683151954148, + "grad_norm": 3.736280918121338, + "learning_rate": 0.000199568224518579, + "loss": 0.7926, + "step": 1272 + }, + { + "epoch": 0.029841124726759143, + "grad_norm": 1.7218469381332397, + "learning_rate": 0.00019956754046953362, + "loss": 1.4121, + "step": 1273 + }, + { + "epoch": 0.02986456630156414, + "grad_norm": 1.0273540019989014, + "learning_rate": 0.00019956685588023198, + "loss": 0.16, + "step": 1274 + }, + { + "epoch": 0.029888007876369135, + "grad_norm": 3.718599557876587, + "learning_rate": 0.0001995661707506778, + "loss": 1.2294, + "step": 1275 + }, + { + "epoch": 0.02991144945117413, + "grad_norm": 3.631891965866089, + "learning_rate": 0.00019956548508087476, + "loss": 0.2993, + "step": 1276 + }, + { + "epoch": 0.029934891025979127, + "grad_norm": 4.8663129806518555, + "learning_rate": 0.00019956479887082656, + "loss": 0.7961, + "step": 1277 + }, + { + "epoch": 0.02995833260078412, + "grad_norm": 3.2172253131866455, + "learning_rate": 0.00019956411212053703, + "loss": 0.4387, + "step": 1278 + }, + { + "epoch": 0.029981774175589115, + "grad_norm": 4.718775272369385, + "learning_rate": 0.00019956342483000978, + "loss": 0.9674, + "step": 1279 + }, + { + "epoch": 0.030005215750394113, + "grad_norm": 0.957049548625946, + "learning_rate": 0.0001995627369992486, + "loss": 0.2014, + "step": 1280 + }, + { + "epoch": 0.030028657325199107, + "grad_norm": 2.982758045196533, + "learning_rate": 0.0001995620486282572, + "loss": 0.6504, + "step": 1281 + }, + { + "epoch": 0.030052098900004102, + "grad_norm": 1.921386480331421, + "learning_rate": 0.00019956135971703934, + "loss": 2.1092, + "step": 1282 + }, + { + "epoch": 0.030075540474809096, + "grad_norm": 3.373594045639038, + "learning_rate": 0.0001995606702655987, + "loss": 0.6348, + "step": 1283 + }, + { + "epoch": 0.030098982049614094, + "grad_norm": 0.9124084711074829, + "learning_rate": 0.00019955998027393915, + "loss": 0.1665, + "step": 1284 + }, + { + "epoch": 0.030122423624419088, + "grad_norm": 2.8184304237365723, + "learning_rate": 0.0001995592897420643, + "loss": 0.5795, + "step": 1285 + }, + { + "epoch": 0.030145865199224083, + "grad_norm": 1.7426024675369263, + "learning_rate": 0.00019955859866997795, + "loss": 1.084, + "step": 1286 + }, + { + "epoch": 0.03016930677402908, + "grad_norm": 2.0016398429870605, + "learning_rate": 0.00019955790705768382, + "loss": 0.2757, + "step": 1287 + }, + { + "epoch": 0.030192748348834075, + "grad_norm": 3.8337831497192383, + "learning_rate": 0.00019955721490518572, + "loss": 0.7244, + "step": 1288 + }, + { + "epoch": 0.03021618992363907, + "grad_norm": 3.004634141921997, + "learning_rate": 0.00019955652221248735, + "loss": 0.7492, + "step": 1289 + }, + { + "epoch": 0.030239631498444067, + "grad_norm": 2.9553799629211426, + "learning_rate": 0.00019955582897959251, + "loss": 0.4268, + "step": 1290 + }, + { + "epoch": 0.03026307307324906, + "grad_norm": 1.8548661470413208, + "learning_rate": 0.0001995551352065049, + "loss": 1.4346, + "step": 1291 + }, + { + "epoch": 0.030286514648054055, + "grad_norm": 3.1662259101867676, + "learning_rate": 0.0001995544408932284, + "loss": 0.5063, + "step": 1292 + }, + { + "epoch": 0.030309956222859053, + "grad_norm": 2.5687153339385986, + "learning_rate": 0.00019955374603976663, + "loss": 0.4244, + "step": 1293 + }, + { + "epoch": 0.030333397797664047, + "grad_norm": 3.339841842651367, + "learning_rate": 0.0001995530506461235, + "loss": 0.6827, + "step": 1294 + }, + { + "epoch": 0.030356839372469042, + "grad_norm": 3.2034354209899902, + "learning_rate": 0.00019955235471230266, + "loss": 0.7984, + "step": 1295 + }, + { + "epoch": 0.03038028094727404, + "grad_norm": 3.4267749786376953, + "learning_rate": 0.00019955165823830798, + "loss": 0.7155, + "step": 1296 + }, + { + "epoch": 0.030403722522079034, + "grad_norm": 3.206069231033325, + "learning_rate": 0.0001995509612241432, + "loss": 0.8928, + "step": 1297 + }, + { + "epoch": 0.030427164096884028, + "grad_norm": 1.331895112991333, + "learning_rate": 0.00019955026366981208, + "loss": 0.2797, + "step": 1298 + }, + { + "epoch": 0.030450605671689022, + "grad_norm": 3.0622735023498535, + "learning_rate": 0.00019954956557531844, + "loss": 0.7566, + "step": 1299 + }, + { + "epoch": 0.03047404724649402, + "grad_norm": 2.7347967624664307, + "learning_rate": 0.0001995488669406661, + "loss": 0.475, + "step": 1300 + }, + { + "epoch": 0.030497488821299015, + "grad_norm": 3.2735531330108643, + "learning_rate": 0.00019954816776585876, + "loss": 0.5487, + "step": 1301 + }, + { + "epoch": 0.03052093039610401, + "grad_norm": 2.8776094913482666, + "learning_rate": 0.0001995474680509003, + "loss": 0.6527, + "step": 1302 + }, + { + "epoch": 0.030544371970909007, + "grad_norm": 4.5176849365234375, + "learning_rate": 0.0001995467677957944, + "loss": 0.6094, + "step": 1303 + }, + { + "epoch": 0.030567813545714, + "grad_norm": 2.5509634017944336, + "learning_rate": 0.00019954606700054502, + "loss": 0.2947, + "step": 1304 + }, + { + "epoch": 0.030591255120518995, + "grad_norm": 2.4524452686309814, + "learning_rate": 0.00019954536566515587, + "loss": 0.4325, + "step": 1305 + }, + { + "epoch": 0.030614696695323993, + "grad_norm": 3.259265899658203, + "learning_rate": 0.00019954466378963078, + "loss": 0.3351, + "step": 1306 + }, + { + "epoch": 0.030638138270128987, + "grad_norm": 3.968864679336548, + "learning_rate": 0.00019954396137397348, + "loss": 0.7963, + "step": 1307 + }, + { + "epoch": 0.03066157984493398, + "grad_norm": 4.201606273651123, + "learning_rate": 0.00019954325841818791, + "loss": 0.8223, + "step": 1308 + }, + { + "epoch": 0.03068502141973898, + "grad_norm": 4.376197814941406, + "learning_rate": 0.00019954255492227781, + "loss": 1.04, + "step": 1309 + }, + { + "epoch": 0.030708462994543974, + "grad_norm": 1.7447738647460938, + "learning_rate": 0.000199541850886247, + "loss": 1.1723, + "step": 1310 + }, + { + "epoch": 0.030731904569348968, + "grad_norm": 3.226315975189209, + "learning_rate": 0.00019954114631009928, + "loss": 0.5228, + "step": 1311 + }, + { + "epoch": 0.030755346144153966, + "grad_norm": 1.1673747301101685, + "learning_rate": 0.00019954044119383855, + "loss": 0.178, + "step": 1312 + }, + { + "epoch": 0.03077878771895896, + "grad_norm": 3.785367727279663, + "learning_rate": 0.00019953973553746854, + "loss": 0.8305, + "step": 1313 + }, + { + "epoch": 0.030802229293763955, + "grad_norm": 5.4921674728393555, + "learning_rate": 0.00019953902934099316, + "loss": 0.9744, + "step": 1314 + }, + { + "epoch": 0.03082567086856895, + "grad_norm": 1.3443583250045776, + "learning_rate": 0.0001995383226044162, + "loss": 0.2919, + "step": 1315 + }, + { + "epoch": 0.030849112443373947, + "grad_norm": 2.4332103729248047, + "learning_rate": 0.0001995376153277415, + "loss": 0.3704, + "step": 1316 + }, + { + "epoch": 0.03087255401817894, + "grad_norm": 0.6665744185447693, + "learning_rate": 0.00019953690751097289, + "loss": 0.0792, + "step": 1317 + }, + { + "epoch": 0.030895995592983935, + "grad_norm": 3.089169979095459, + "learning_rate": 0.00019953619915411423, + "loss": 0.5159, + "step": 1318 + }, + { + "epoch": 0.030919437167788933, + "grad_norm": 0.8186952471733093, + "learning_rate": 0.00019953549025716937, + "loss": 0.2248, + "step": 1319 + }, + { + "epoch": 0.030942878742593927, + "grad_norm": 2.9763758182525635, + "learning_rate": 0.00019953478082014212, + "loss": 0.6449, + "step": 1320 + }, + { + "epoch": 0.03096632031739892, + "grad_norm": 2.7804207801818848, + "learning_rate": 0.00019953407084303634, + "loss": 0.5922, + "step": 1321 + }, + { + "epoch": 0.03098976189220392, + "grad_norm": 2.5185656547546387, + "learning_rate": 0.00019953336032585593, + "loss": 0.2777, + "step": 1322 + }, + { + "epoch": 0.031013203467008914, + "grad_norm": 1.231755256652832, + "learning_rate": 0.0001995326492686047, + "loss": 0.2167, + "step": 1323 + }, + { + "epoch": 0.031036645041813908, + "grad_norm": 0.4627842307090759, + "learning_rate": 0.00019953193767128648, + "loss": 0.0743, + "step": 1324 + }, + { + "epoch": 0.031060086616618906, + "grad_norm": 1.5859616994857788, + "learning_rate": 0.0001995312255339052, + "loss": 1.4148, + "step": 1325 + }, + { + "epoch": 0.0310835281914239, + "grad_norm": 3.4303135871887207, + "learning_rate": 0.0001995305128564647, + "loss": 0.7261, + "step": 1326 + }, + { + "epoch": 0.031106969766228895, + "grad_norm": 1.1717790365219116, + "learning_rate": 0.0001995297996389688, + "loss": 0.2052, + "step": 1327 + }, + { + "epoch": 0.031130411341033892, + "grad_norm": 4.00161600112915, + "learning_rate": 0.00019952908588142143, + "loss": 0.843, + "step": 1328 + }, + { + "epoch": 0.031153852915838887, + "grad_norm": 2.204280376434326, + "learning_rate": 0.00019952837158382646, + "loss": 0.2346, + "step": 1329 + }, + { + "epoch": 0.03117729449064388, + "grad_norm": 0.28113454580307007, + "learning_rate": 0.00019952765674618772, + "loss": 0.0274, + "step": 1330 + }, + { + "epoch": 0.031200736065448875, + "grad_norm": 5.101832389831543, + "learning_rate": 0.00019952694136850913, + "loss": 0.8976, + "step": 1331 + }, + { + "epoch": 0.031224177640253873, + "grad_norm": 2.8737847805023193, + "learning_rate": 0.00019952622545079456, + "loss": 0.3858, + "step": 1332 + }, + { + "epoch": 0.031247619215058867, + "grad_norm": 4.446382522583008, + "learning_rate": 0.0001995255089930479, + "loss": 0.6932, + "step": 1333 + }, + { + "epoch": 0.031271060789863865, + "grad_norm": 2.002861261367798, + "learning_rate": 0.00019952479199527302, + "loss": 0.4474, + "step": 1334 + }, + { + "epoch": 0.03129450236466886, + "grad_norm": 3.0528650283813477, + "learning_rate": 0.0001995240744574738, + "loss": 0.4379, + "step": 1335 + }, + { + "epoch": 0.031317943939473854, + "grad_norm": 1.4831438064575195, + "learning_rate": 0.00019952335637965417, + "loss": 0.9064, + "step": 1336 + }, + { + "epoch": 0.03134138551427885, + "grad_norm": 2.9564366340637207, + "learning_rate": 0.000199522637761818, + "loss": 0.5404, + "step": 1337 + }, + { + "epoch": 0.03136482708908384, + "grad_norm": 2.617810010910034, + "learning_rate": 0.00019952191860396923, + "loss": 1.9126, + "step": 1338 + }, + { + "epoch": 0.03138826866388884, + "grad_norm": 1.5142204761505127, + "learning_rate": 0.0001995211989061117, + "loss": 0.2477, + "step": 1339 + }, + { + "epoch": 0.03141171023869384, + "grad_norm": 2.7810025215148926, + "learning_rate": 0.00019952047866824937, + "loss": 0.5571, + "step": 1340 + }, + { + "epoch": 0.03143515181349883, + "grad_norm": 1.7103018760681152, + "learning_rate": 0.0001995197578903861, + "loss": 1.6394, + "step": 1341 + }, + { + "epoch": 0.03145859338830383, + "grad_norm": 3.589268445968628, + "learning_rate": 0.00019951903657252585, + "loss": 0.4865, + "step": 1342 + }, + { + "epoch": 0.03148203496310882, + "grad_norm": 4.127164363861084, + "learning_rate": 0.0001995183147146725, + "loss": 0.8062, + "step": 1343 + }, + { + "epoch": 0.031505476537913815, + "grad_norm": 1.0827964544296265, + "learning_rate": 0.00019951759231682996, + "loss": 0.1807, + "step": 1344 + }, + { + "epoch": 0.03152891811271881, + "grad_norm": 2.303666830062866, + "learning_rate": 0.00019951686937900217, + "loss": 1.7252, + "step": 1345 + }, + { + "epoch": 0.03155235968752381, + "grad_norm": 1.4435161352157593, + "learning_rate": 0.00019951614590119307, + "loss": 0.3655, + "step": 1346 + }, + { + "epoch": 0.031575801262328805, + "grad_norm": 2.4973623752593994, + "learning_rate": 0.00019951542188340656, + "loss": 0.3375, + "step": 1347 + }, + { + "epoch": 0.0315992428371338, + "grad_norm": 0.675186276435852, + "learning_rate": 0.00019951469732564653, + "loss": 0.1454, + "step": 1348 + }, + { + "epoch": 0.031622684411938794, + "grad_norm": 4.146817684173584, + "learning_rate": 0.000199513972227917, + "loss": 0.7104, + "step": 1349 + }, + { + "epoch": 0.03164612598674379, + "grad_norm": 1.1718103885650635, + "learning_rate": 0.00019951324659022186, + "loss": 0.1629, + "step": 1350 + }, + { + "epoch": 0.03166956756154878, + "grad_norm": 5.800905704498291, + "learning_rate": 0.00019951252041256503, + "loss": 0.6664, + "step": 1351 + }, + { + "epoch": 0.031693009136353784, + "grad_norm": 3.2460618019104004, + "learning_rate": 0.00019951179369495048, + "loss": 0.8048, + "step": 1352 + }, + { + "epoch": 0.03171645071115878, + "grad_norm": 1.8649011850357056, + "learning_rate": 0.00019951106643738213, + "loss": 1.473, + "step": 1353 + }, + { + "epoch": 0.03173989228596377, + "grad_norm": 3.9453015327453613, + "learning_rate": 0.00019951033863986392, + "loss": 0.5881, + "step": 1354 + }, + { + "epoch": 0.031763333860768767, + "grad_norm": 3.2838425636291504, + "learning_rate": 0.00019950961030239982, + "loss": 0.8142, + "step": 1355 + }, + { + "epoch": 0.03178677543557376, + "grad_norm": 3.263368844985962, + "learning_rate": 0.0001995088814249938, + "loss": 0.5935, + "step": 1356 + }, + { + "epoch": 0.031810217010378755, + "grad_norm": 5.236913204193115, + "learning_rate": 0.00019950815200764978, + "loss": 0.5322, + "step": 1357 + }, + { + "epoch": 0.03183365858518375, + "grad_norm": 3.9218990802764893, + "learning_rate": 0.00019950742205037173, + "loss": 0.5372, + "step": 1358 + }, + { + "epoch": 0.03185710015998875, + "grad_norm": 3.695204019546509, + "learning_rate": 0.0001995066915531636, + "loss": 0.4135, + "step": 1359 + }, + { + "epoch": 0.031880541734793745, + "grad_norm": 1.6758335828781128, + "learning_rate": 0.00019950596051602937, + "loss": 1.2613, + "step": 1360 + }, + { + "epoch": 0.03190398330959874, + "grad_norm": 1.0205714702606201, + "learning_rate": 0.000199505228938973, + "loss": 0.2182, + "step": 1361 + }, + { + "epoch": 0.031927424884403734, + "grad_norm": 3.1297802925109863, + "learning_rate": 0.00019950449682199845, + "loss": 0.3025, + "step": 1362 + }, + { + "epoch": 0.03195086645920873, + "grad_norm": 3.969381332397461, + "learning_rate": 0.0001995037641651097, + "loss": 0.9843, + "step": 1363 + }, + { + "epoch": 0.03197430803401372, + "grad_norm": 0.7407707571983337, + "learning_rate": 0.00019950303096831075, + "loss": 0.0854, + "step": 1364 + }, + { + "epoch": 0.031997749608818724, + "grad_norm": 2.938920021057129, + "learning_rate": 0.00019950229723160557, + "loss": 0.4371, + "step": 1365 + }, + { + "epoch": 0.03202119118362372, + "grad_norm": 3.199747323989868, + "learning_rate": 0.0001995015629549981, + "loss": 0.5078, + "step": 1366 + }, + { + "epoch": 0.03204463275842871, + "grad_norm": 0.7376722097396851, + "learning_rate": 0.00019950082813849232, + "loss": 0.0954, + "step": 1367 + }, + { + "epoch": 0.032068074333233706, + "grad_norm": 0.3031788766384125, + "learning_rate": 0.0001995000927820923, + "loss": 0.0445, + "step": 1368 + }, + { + "epoch": 0.0320915159080387, + "grad_norm": 4.37955904006958, + "learning_rate": 0.00019949935688580198, + "loss": 1.0269, + "step": 1369 + }, + { + "epoch": 0.032114957482843695, + "grad_norm": 5.295725345611572, + "learning_rate": 0.00019949862044962532, + "loss": 0.8162, + "step": 1370 + }, + { + "epoch": 0.03213839905764869, + "grad_norm": 0.29413262009620667, + "learning_rate": 0.00019949788347356636, + "loss": 0.0362, + "step": 1371 + }, + { + "epoch": 0.03216184063245369, + "grad_norm": 0.4559694230556488, + "learning_rate": 0.0001994971459576291, + "loss": 0.0608, + "step": 1372 + }, + { + "epoch": 0.032185282207258685, + "grad_norm": 4.697997093200684, + "learning_rate": 0.0001994964079018175, + "loss": 0.6386, + "step": 1373 + }, + { + "epoch": 0.03220872378206368, + "grad_norm": 4.959383964538574, + "learning_rate": 0.00019949566930613563, + "loss": 0.8684, + "step": 1374 + }, + { + "epoch": 0.032232165356868674, + "grad_norm": 2.1338870525360107, + "learning_rate": 0.00019949493017058746, + "loss": 0.3046, + "step": 1375 + }, + { + "epoch": 0.03225560693167367, + "grad_norm": 3.5999794006347656, + "learning_rate": 0.00019949419049517696, + "loss": 0.4364, + "step": 1376 + }, + { + "epoch": 0.03227904850647866, + "grad_norm": 3.4646713733673096, + "learning_rate": 0.00019949345027990822, + "loss": 0.4093, + "step": 1377 + }, + { + "epoch": 0.032302490081283664, + "grad_norm": 2.650416374206543, + "learning_rate": 0.0001994927095247852, + "loss": 0.4411, + "step": 1378 + }, + { + "epoch": 0.03232593165608866, + "grad_norm": 1.0870344638824463, + "learning_rate": 0.00019949196822981196, + "loss": 0.2256, + "step": 1379 + }, + { + "epoch": 0.03234937323089365, + "grad_norm": 0.4421772062778473, + "learning_rate": 0.00019949122639499248, + "loss": 0.0608, + "step": 1380 + }, + { + "epoch": 0.032372814805698646, + "grad_norm": 1.5129591226577759, + "learning_rate": 0.00019949048402033084, + "loss": 0.214, + "step": 1381 + }, + { + "epoch": 0.03239625638050364, + "grad_norm": 2.094219923019409, + "learning_rate": 0.00019948974110583101, + "loss": 0.3059, + "step": 1382 + }, + { + "epoch": 0.032419697955308635, + "grad_norm": 2.4241278171539307, + "learning_rate": 0.00019948899765149706, + "loss": 0.3455, + "step": 1383 + }, + { + "epoch": 0.032443139530113636, + "grad_norm": 1.2164565324783325, + "learning_rate": 0.00019948825365733298, + "loss": 0.2442, + "step": 1384 + }, + { + "epoch": 0.03246658110491863, + "grad_norm": 2.9074339866638184, + "learning_rate": 0.00019948750912334289, + "loss": 0.3364, + "step": 1385 + }, + { + "epoch": 0.032490022679723625, + "grad_norm": 5.021354675292969, + "learning_rate": 0.00019948676404953075, + "loss": 0.7283, + "step": 1386 + }, + { + "epoch": 0.03251346425452862, + "grad_norm": 5.029524326324463, + "learning_rate": 0.00019948601843590064, + "loss": 0.9669, + "step": 1387 + }, + { + "epoch": 0.032536905829333614, + "grad_norm": 1.3941634893417358, + "learning_rate": 0.0001994852722824566, + "loss": 0.2332, + "step": 1388 + }, + { + "epoch": 0.03256034740413861, + "grad_norm": 1.0018341541290283, + "learning_rate": 0.00019948452558920266, + "loss": 0.1626, + "step": 1389 + }, + { + "epoch": 0.0325837889789436, + "grad_norm": 3.6076791286468506, + "learning_rate": 0.00019948377835614286, + "loss": 0.8507, + "step": 1390 + }, + { + "epoch": 0.0326072305537486, + "grad_norm": 1.806356430053711, + "learning_rate": 0.00019948303058328132, + "loss": 1.2653, + "step": 1391 + }, + { + "epoch": 0.0326306721285536, + "grad_norm": 1.4873102903366089, + "learning_rate": 0.00019948228227062206, + "loss": 0.2901, + "step": 1392 + }, + { + "epoch": 0.03265411370335859, + "grad_norm": 3.3170149326324463, + "learning_rate": 0.00019948153341816914, + "loss": 0.5599, + "step": 1393 + }, + { + "epoch": 0.032677555278163586, + "grad_norm": 3.2769227027893066, + "learning_rate": 0.0001994807840259266, + "loss": 0.4107, + "step": 1394 + }, + { + "epoch": 0.03270099685296858, + "grad_norm": 3.7182414531707764, + "learning_rate": 0.00019948003409389858, + "loss": 0.646, + "step": 1395 + }, + { + "epoch": 0.032724438427773575, + "grad_norm": 2.690126895904541, + "learning_rate": 0.00019947928362208902, + "loss": 0.6286, + "step": 1396 + }, + { + "epoch": 0.032747880002578576, + "grad_norm": 2.653827667236328, + "learning_rate": 0.00019947853261050215, + "loss": 0.5395, + "step": 1397 + }, + { + "epoch": 0.03277132157738357, + "grad_norm": 1.1256639957427979, + "learning_rate": 0.0001994777810591419, + "loss": 0.1145, + "step": 1398 + }, + { + "epoch": 0.032794763152188565, + "grad_norm": 1.5252596139907837, + "learning_rate": 0.00019947702896801244, + "loss": 0.3702, + "step": 1399 + }, + { + "epoch": 0.03281820472699356, + "grad_norm": 3.2492897510528564, + "learning_rate": 0.00019947627633711782, + "loss": 0.7, + "step": 1400 + }, + { + "epoch": 0.032841646301798554, + "grad_norm": 1.2494350671768188, + "learning_rate": 0.00019947552316646215, + "loss": 0.173, + "step": 1401 + }, + { + "epoch": 0.03286508787660355, + "grad_norm": 3.330101251602173, + "learning_rate": 0.00019947476945604946, + "loss": 0.6003, + "step": 1402 + }, + { + "epoch": 0.03288852945140854, + "grad_norm": 3.6250064373016357, + "learning_rate": 0.00019947401520588388, + "loss": 0.8172, + "step": 1403 + }, + { + "epoch": 0.03291197102621354, + "grad_norm": 2.326632261276245, + "learning_rate": 0.00019947326041596953, + "loss": 0.3964, + "step": 1404 + }, + { + "epoch": 0.03293541260101854, + "grad_norm": 4.632391452789307, + "learning_rate": 0.00019947250508631044, + "loss": 0.5751, + "step": 1405 + }, + { + "epoch": 0.03295885417582353, + "grad_norm": 3.987525701522827, + "learning_rate": 0.00019947174921691074, + "loss": 0.7748, + "step": 1406 + }, + { + "epoch": 0.032982295750628526, + "grad_norm": 4.419182300567627, + "learning_rate": 0.00019947099280777454, + "loss": 0.8605, + "step": 1407 + }, + { + "epoch": 0.03300573732543352, + "grad_norm": 3.0853471755981445, + "learning_rate": 0.0001994702358589059, + "loss": 0.365, + "step": 1408 + }, + { + "epoch": 0.033029178900238515, + "grad_norm": 3.7426223754882812, + "learning_rate": 0.00019946947837030905, + "loss": 0.715, + "step": 1409 + }, + { + "epoch": 0.033052620475043516, + "grad_norm": 3.497001886367798, + "learning_rate": 0.00019946872034198794, + "loss": 0.5473, + "step": 1410 + }, + { + "epoch": 0.03307606204984851, + "grad_norm": 4.536554336547852, + "learning_rate": 0.00019946796177394677, + "loss": 0.4713, + "step": 1411 + }, + { + "epoch": 0.033099503624653505, + "grad_norm": 3.637399196624756, + "learning_rate": 0.00019946720266618968, + "loss": 0.5133, + "step": 1412 + }, + { + "epoch": 0.0331229451994585, + "grad_norm": 1.1318838596343994, + "learning_rate": 0.0001994664430187207, + "loss": 0.7181, + "step": 1413 + }, + { + "epoch": 0.033146386774263494, + "grad_norm": 5.52673864364624, + "learning_rate": 0.00019946568283154404, + "loss": 0.7933, + "step": 1414 + }, + { + "epoch": 0.03316982834906849, + "grad_norm": 3.617044687271118, + "learning_rate": 0.00019946492210466378, + "loss": 0.3952, + "step": 1415 + }, + { + "epoch": 0.03319326992387349, + "grad_norm": 3.193469524383545, + "learning_rate": 0.00019946416083808406, + "loss": 0.682, + "step": 1416 + }, + { + "epoch": 0.03321671149867848, + "grad_norm": 0.523939311504364, + "learning_rate": 0.00019946339903180896, + "loss": 0.0824, + "step": 1417 + }, + { + "epoch": 0.03324015307348348, + "grad_norm": 1.1391621828079224, + "learning_rate": 0.00019946263668584271, + "loss": 0.1218, + "step": 1418 + }, + { + "epoch": 0.03326359464828847, + "grad_norm": 2.2351529598236084, + "learning_rate": 0.00019946187380018938, + "loss": 0.3513, + "step": 1419 + }, + { + "epoch": 0.033287036223093466, + "grad_norm": 2.248607873916626, + "learning_rate": 0.00019946111037485314, + "loss": 1.5277, + "step": 1420 + }, + { + "epoch": 0.03331047779789846, + "grad_norm": 3.5871942043304443, + "learning_rate": 0.00019946034640983812, + "loss": 0.8638, + "step": 1421 + }, + { + "epoch": 0.033333919372703455, + "grad_norm": 1.849125862121582, + "learning_rate": 0.00019945958190514845, + "loss": 1.8809, + "step": 1422 + }, + { + "epoch": 0.033357360947508456, + "grad_norm": 2.8575823307037354, + "learning_rate": 0.00019945881686078828, + "loss": 0.6109, + "step": 1423 + }, + { + "epoch": 0.03338080252231345, + "grad_norm": 3.4726953506469727, + "learning_rate": 0.0001994580512767618, + "loss": 0.634, + "step": 1424 + }, + { + "epoch": 0.033404244097118445, + "grad_norm": 1.3871433734893799, + "learning_rate": 0.0001994572851530731, + "loss": 0.328, + "step": 1425 + }, + { + "epoch": 0.03342768567192344, + "grad_norm": 3.743231773376465, + "learning_rate": 0.00019945651848972642, + "loss": 0.6417, + "step": 1426 + }, + { + "epoch": 0.03345112724672843, + "grad_norm": 3.814107894897461, + "learning_rate": 0.00019945575128672586, + "loss": 0.5416, + "step": 1427 + }, + { + "epoch": 0.03347456882153343, + "grad_norm": 3.4281623363494873, + "learning_rate": 0.0001994549835440756, + "loss": 0.6875, + "step": 1428 + }, + { + "epoch": 0.03349801039633843, + "grad_norm": 3.5878818035125732, + "learning_rate": 0.00019945421526177976, + "loss": 0.748, + "step": 1429 + }, + { + "epoch": 0.03352145197114342, + "grad_norm": 1.5964261293411255, + "learning_rate": 0.0001994534464398426, + "loss": 0.3373, + "step": 1430 + }, + { + "epoch": 0.03354489354594842, + "grad_norm": 4.978437900543213, + "learning_rate": 0.00019945267707826824, + "loss": 0.8509, + "step": 1431 + }, + { + "epoch": 0.03356833512075341, + "grad_norm": 2.2155754566192627, + "learning_rate": 0.00019945190717706083, + "loss": 0.65, + "step": 1432 + }, + { + "epoch": 0.033591776695558406, + "grad_norm": 3.8218624591827393, + "learning_rate": 0.0001994511367362246, + "loss": 0.577, + "step": 1433 + }, + { + "epoch": 0.0336152182703634, + "grad_norm": 2.737778425216675, + "learning_rate": 0.0001994503657557637, + "loss": 0.4251, + "step": 1434 + }, + { + "epoch": 0.0336386598451684, + "grad_norm": 3.1336545944213867, + "learning_rate": 0.00019944959423568232, + "loss": 0.4902, + "step": 1435 + }, + { + "epoch": 0.033662101419973396, + "grad_norm": 0.6230221390724182, + "learning_rate": 0.00019944882217598465, + "loss": 0.1137, + "step": 1436 + }, + { + "epoch": 0.03368554299477839, + "grad_norm": 2.454787015914917, + "learning_rate": 0.00019944804957667484, + "loss": 0.379, + "step": 1437 + }, + { + "epoch": 0.033708984569583385, + "grad_norm": 3.7318952083587646, + "learning_rate": 0.00019944727643775716, + "loss": 0.7019, + "step": 1438 + }, + { + "epoch": 0.03373242614438838, + "grad_norm": 3.1457650661468506, + "learning_rate": 0.00019944650275923573, + "loss": 0.5748, + "step": 1439 + }, + { + "epoch": 0.03375586771919337, + "grad_norm": 2.658074378967285, + "learning_rate": 0.0001994457285411148, + "loss": 0.4085, + "step": 1440 + }, + { + "epoch": 0.03377930929399837, + "grad_norm": 3.192790985107422, + "learning_rate": 0.00019944495378339854, + "loss": 0.4744, + "step": 1441 + }, + { + "epoch": 0.03380275086880337, + "grad_norm": 5.139120578765869, + "learning_rate": 0.00019944417848609118, + "loss": 0.6974, + "step": 1442 + }, + { + "epoch": 0.03382619244360836, + "grad_norm": 3.478919506072998, + "learning_rate": 0.0001994434026491969, + "loss": 0.6943, + "step": 1443 + }, + { + "epoch": 0.03384963401841336, + "grad_norm": 2.4775888919830322, + "learning_rate": 0.00019944262627271991, + "loss": 0.2381, + "step": 1444 + }, + { + "epoch": 0.03387307559321835, + "grad_norm": 2.495349884033203, + "learning_rate": 0.00019944184935666444, + "loss": 0.2029, + "step": 1445 + }, + { + "epoch": 0.033896517168023346, + "grad_norm": 3.9157803058624268, + "learning_rate": 0.0001994410719010347, + "loss": 0.4578, + "step": 1446 + }, + { + "epoch": 0.03391995874282834, + "grad_norm": 1.1809155941009521, + "learning_rate": 0.00019944029390583495, + "loss": 0.2275, + "step": 1447 + }, + { + "epoch": 0.03394340031763334, + "grad_norm": 5.087894439697266, + "learning_rate": 0.00019943951537106932, + "loss": 0.7105, + "step": 1448 + }, + { + "epoch": 0.033966841892438336, + "grad_norm": 4.154771327972412, + "learning_rate": 0.0001994387362967421, + "loss": 0.6855, + "step": 1449 + }, + { + "epoch": 0.03399028346724333, + "grad_norm": 3.278872013092041, + "learning_rate": 0.0001994379566828575, + "loss": 0.4233, + "step": 1450 + }, + { + "epoch": 0.034013725042048325, + "grad_norm": 2.1884024143218994, + "learning_rate": 0.00019943717652941975, + "loss": 1.9051, + "step": 1451 + }, + { + "epoch": 0.03403716661685332, + "grad_norm": 1.4594060182571411, + "learning_rate": 0.0001994363958364331, + "loss": 0.3966, + "step": 1452 + }, + { + "epoch": 0.03406060819165831, + "grad_norm": 3.6491708755493164, + "learning_rate": 0.0001994356146039017, + "loss": 0.6564, + "step": 1453 + }, + { + "epoch": 0.03408404976646331, + "grad_norm": 2.6777493953704834, + "learning_rate": 0.00019943483283182994, + "loss": 0.3899, + "step": 1454 + }, + { + "epoch": 0.03410749134126831, + "grad_norm": 3.26816725730896, + "learning_rate": 0.00019943405052022195, + "loss": 0.5704, + "step": 1455 + }, + { + "epoch": 0.0341309329160733, + "grad_norm": 0.29589322209358215, + "learning_rate": 0.00019943326766908205, + "loss": 0.0459, + "step": 1456 + }, + { + "epoch": 0.0341543744908783, + "grad_norm": 1.3101202249526978, + "learning_rate": 0.00019943248427841436, + "loss": 0.1652, + "step": 1457 + }, + { + "epoch": 0.03417781606568329, + "grad_norm": 3.65605092048645, + "learning_rate": 0.00019943170034822328, + "loss": 0.395, + "step": 1458 + }, + { + "epoch": 0.034201257640488286, + "grad_norm": 1.3297101259231567, + "learning_rate": 0.00019943091587851296, + "loss": 0.1602, + "step": 1459 + }, + { + "epoch": 0.03422469921529328, + "grad_norm": 4.885553359985352, + "learning_rate": 0.00019943013086928772, + "loss": 0.7049, + "step": 1460 + }, + { + "epoch": 0.03424814079009828, + "grad_norm": 5.8412370681762695, + "learning_rate": 0.0001994293453205518, + "loss": 0.7046, + "step": 1461 + }, + { + "epoch": 0.034271582364903276, + "grad_norm": 4.531640529632568, + "learning_rate": 0.00019942855923230944, + "loss": 0.6286, + "step": 1462 + }, + { + "epoch": 0.03429502393970827, + "grad_norm": 5.4843220710754395, + "learning_rate": 0.00019942777260456492, + "loss": 0.9043, + "step": 1463 + }, + { + "epoch": 0.034318465514513265, + "grad_norm": 1.3446887731552124, + "learning_rate": 0.0001994269854373225, + "loss": 0.308, + "step": 1464 + }, + { + "epoch": 0.03434190708931826, + "grad_norm": 1.0076967477798462, + "learning_rate": 0.00019942619773058646, + "loss": 0.1554, + "step": 1465 + }, + { + "epoch": 0.03436534866412325, + "grad_norm": 3.6478092670440674, + "learning_rate": 0.00019942540948436111, + "loss": 0.2875, + "step": 1466 + }, + { + "epoch": 0.034388790238928255, + "grad_norm": 2.2551612854003906, + "learning_rate": 0.00019942462069865067, + "loss": 0.4187, + "step": 1467 + }, + { + "epoch": 0.03441223181373325, + "grad_norm": 3.1446542739868164, + "learning_rate": 0.00019942383137345945, + "loss": 0.5979, + "step": 1468 + }, + { + "epoch": 0.03443567338853824, + "grad_norm": 2.837878942489624, + "learning_rate": 0.00019942304150879172, + "loss": 0.3711, + "step": 1469 + }, + { + "epoch": 0.03445911496334324, + "grad_norm": 4.5703511238098145, + "learning_rate": 0.0001994222511046518, + "loss": 1.1518, + "step": 1470 + }, + { + "epoch": 0.03448255653814823, + "grad_norm": 1.396170735359192, + "learning_rate": 0.00019942146016104387, + "loss": 0.2972, + "step": 1471 + }, + { + "epoch": 0.034505998112953226, + "grad_norm": 2.2088770866394043, + "learning_rate": 0.00019942066867797238, + "loss": 0.333, + "step": 1472 + }, + { + "epoch": 0.03452943968775822, + "grad_norm": 0.8960858583450317, + "learning_rate": 0.00019941987665544151, + "loss": 0.142, + "step": 1473 + }, + { + "epoch": 0.03455288126256322, + "grad_norm": 4.192626476287842, + "learning_rate": 0.0001994190840934556, + "loss": 0.9101, + "step": 1474 + }, + { + "epoch": 0.034576322837368216, + "grad_norm": 3.0271923542022705, + "learning_rate": 0.00019941829099201895, + "loss": 0.3831, + "step": 1475 + }, + { + "epoch": 0.03459976441217321, + "grad_norm": 1.8091219663619995, + "learning_rate": 0.00019941749735113585, + "loss": 0.4324, + "step": 1476 + }, + { + "epoch": 0.034623205986978205, + "grad_norm": 3.2798268795013428, + "learning_rate": 0.00019941670317081064, + "loss": 0.7653, + "step": 1477 + }, + { + "epoch": 0.0346466475617832, + "grad_norm": 3.4041285514831543, + "learning_rate": 0.00019941590845104758, + "loss": 0.9781, + "step": 1478 + }, + { + "epoch": 0.03467008913658819, + "grad_norm": 2.4428155422210693, + "learning_rate": 0.000199415113191851, + "loss": 0.4566, + "step": 1479 + }, + { + "epoch": 0.034693530711393195, + "grad_norm": 2.8613216876983643, + "learning_rate": 0.00019941431739322525, + "loss": 0.377, + "step": 1480 + }, + { + "epoch": 0.03471697228619819, + "grad_norm": 3.4630138874053955, + "learning_rate": 0.0001994135210551746, + "loss": 0.9277, + "step": 1481 + }, + { + "epoch": 0.03474041386100318, + "grad_norm": 4.424195289611816, + "learning_rate": 0.0001994127241777034, + "loss": 1.1584, + "step": 1482 + }, + { + "epoch": 0.03476385543580818, + "grad_norm": 1.3849427700042725, + "learning_rate": 0.00019941192676081594, + "loss": 0.2796, + "step": 1483 + }, + { + "epoch": 0.03478729701061317, + "grad_norm": 3.410818099975586, + "learning_rate": 0.0001994111288045166, + "loss": 0.7256, + "step": 1484 + }, + { + "epoch": 0.034810738585418166, + "grad_norm": 3.4735777378082275, + "learning_rate": 0.00019941033030880962, + "loss": 0.7442, + "step": 1485 + }, + { + "epoch": 0.03483418016022316, + "grad_norm": 3.7659196853637695, + "learning_rate": 0.00019940953127369943, + "loss": 0.7219, + "step": 1486 + }, + { + "epoch": 0.03485762173502816, + "grad_norm": 1.7158312797546387, + "learning_rate": 0.00019940873169919034, + "loss": 0.3612, + "step": 1487 + }, + { + "epoch": 0.034881063309833156, + "grad_norm": 3.610304355621338, + "learning_rate": 0.00019940793158528666, + "loss": 0.8225, + "step": 1488 + }, + { + "epoch": 0.03490450488463815, + "grad_norm": 1.8344440460205078, + "learning_rate": 0.00019940713093199276, + "loss": 0.3974, + "step": 1489 + }, + { + "epoch": 0.034927946459443145, + "grad_norm": 1.8847626447677612, + "learning_rate": 0.00019940632973931296, + "loss": 1.6139, + "step": 1490 + }, + { + "epoch": 0.03495138803424814, + "grad_norm": 7.464226722717285, + "learning_rate": 0.00019940552800725163, + "loss": 2.151, + "step": 1491 + }, + { + "epoch": 0.03497482960905313, + "grad_norm": 0.4376743733882904, + "learning_rate": 0.0001994047257358131, + "loss": 0.0499, + "step": 1492 + }, + { + "epoch": 0.034998271183858135, + "grad_norm": 5.477962970733643, + "learning_rate": 0.00019940392292500176, + "loss": 0.6688, + "step": 1493 + }, + { + "epoch": 0.03502171275866313, + "grad_norm": 4.330263137817383, + "learning_rate": 0.00019940311957482193, + "loss": 0.6904, + "step": 1494 + }, + { + "epoch": 0.03504515433346812, + "grad_norm": 2.993833541870117, + "learning_rate": 0.00019940231568527797, + "loss": 2.3003, + "step": 1495 + }, + { + "epoch": 0.03506859590827312, + "grad_norm": 3.02665114402771, + "learning_rate": 0.00019940151125637424, + "loss": 0.4326, + "step": 1496 + }, + { + "epoch": 0.03509203748307811, + "grad_norm": 4.358163356781006, + "learning_rate": 0.00019940070628811515, + "loss": 1.0584, + "step": 1497 + }, + { + "epoch": 0.035115479057883106, + "grad_norm": 5.288777828216553, + "learning_rate": 0.00019939990078050496, + "loss": 1.2539, + "step": 1498 + }, + { + "epoch": 0.03513892063268811, + "grad_norm": 4.732823848724365, + "learning_rate": 0.0001993990947335482, + "loss": 0.8955, + "step": 1499 + }, + { + "epoch": 0.0351623622074931, + "grad_norm": 3.8743889331817627, + "learning_rate": 0.0001993982881472491, + "loss": 0.5978, + "step": 1500 + }, + { + "epoch": 0.035185803782298096, + "grad_norm": 3.211594343185425, + "learning_rate": 0.0001993974810216121, + "loss": 1.1828, + "step": 1501 + }, + { + "epoch": 0.03520924535710309, + "grad_norm": 1.0296181440353394, + "learning_rate": 0.00019939667335664162, + "loss": 0.7893, + "step": 1502 + }, + { + "epoch": 0.035232686931908085, + "grad_norm": 2.636326551437378, + "learning_rate": 0.00019939586515234194, + "loss": 0.5179, + "step": 1503 + }, + { + "epoch": 0.03525612850671308, + "grad_norm": 3.1852431297302246, + "learning_rate": 0.0001993950564087175, + "loss": 0.7691, + "step": 1504 + }, + { + "epoch": 0.03527957008151807, + "grad_norm": 3.3381528854370117, + "learning_rate": 0.00019939424712577274, + "loss": 0.5981, + "step": 1505 + }, + { + "epoch": 0.035303011656323074, + "grad_norm": 3.0011544227600098, + "learning_rate": 0.00019939343730351197, + "loss": 0.4911, + "step": 1506 + }, + { + "epoch": 0.03532645323112807, + "grad_norm": 4.317809581756592, + "learning_rate": 0.00019939262694193958, + "loss": 0.6603, + "step": 1507 + }, + { + "epoch": 0.03534989480593306, + "grad_norm": 2.9305100440979004, + "learning_rate": 0.00019939181604106005, + "loss": 0.3619, + "step": 1508 + }, + { + "epoch": 0.03537333638073806, + "grad_norm": 0.941260576248169, + "learning_rate": 0.00019939100460087772, + "loss": 0.1635, + "step": 1509 + }, + { + "epoch": 0.03539677795554305, + "grad_norm": 2.048706531524658, + "learning_rate": 0.00019939019262139702, + "loss": 1.8771, + "step": 1510 + }, + { + "epoch": 0.035420219530348046, + "grad_norm": 2.8300840854644775, + "learning_rate": 0.00019938938010262232, + "loss": 0.8027, + "step": 1511 + }, + { + "epoch": 0.03544366110515305, + "grad_norm": 0.5533415675163269, + "learning_rate": 0.00019938856704455803, + "loss": 0.1387, + "step": 1512 + }, + { + "epoch": 0.03546710267995804, + "grad_norm": 1.8027312755584717, + "learning_rate": 0.0001993877534472086, + "loss": 0.2899, + "step": 1513 + }, + { + "epoch": 0.035490544254763036, + "grad_norm": 3.3885087966918945, + "learning_rate": 0.00019938693931057845, + "loss": 0.5111, + "step": 1514 + }, + { + "epoch": 0.03551398582956803, + "grad_norm": 2.025465488433838, + "learning_rate": 0.00019938612463467193, + "loss": 0.3234, + "step": 1515 + }, + { + "epoch": 0.035537427404373025, + "grad_norm": 3.2452452182769775, + "learning_rate": 0.00019938530941949353, + "loss": 0.7774, + "step": 1516 + }, + { + "epoch": 0.03556086897917802, + "grad_norm": 2.921342372894287, + "learning_rate": 0.00019938449366504766, + "loss": 0.5344, + "step": 1517 + }, + { + "epoch": 0.03558431055398301, + "grad_norm": 0.7436120510101318, + "learning_rate": 0.0001993836773713387, + "loss": 0.2059, + "step": 1518 + }, + { + "epoch": 0.035607752128788014, + "grad_norm": 1.772218108177185, + "learning_rate": 0.00019938286053837112, + "loss": 1.7005, + "step": 1519 + }, + { + "epoch": 0.03563119370359301, + "grad_norm": 1.3337998390197754, + "learning_rate": 0.00019938204316614938, + "loss": 0.1914, + "step": 1520 + }, + { + "epoch": 0.035654635278398, + "grad_norm": 0.9489693641662598, + "learning_rate": 0.0001993812252546778, + "loss": 0.1645, + "step": 1521 + }, + { + "epoch": 0.035678076853203, + "grad_norm": 4.49517822265625, + "learning_rate": 0.00019938040680396094, + "loss": 0.589, + "step": 1522 + }, + { + "epoch": 0.03570151842800799, + "grad_norm": 0.9213501214981079, + "learning_rate": 0.00019937958781400325, + "loss": 0.2287, + "step": 1523 + }, + { + "epoch": 0.035724960002812986, + "grad_norm": 2.9024007320404053, + "learning_rate": 0.00019937876828480906, + "loss": 0.5168, + "step": 1524 + }, + { + "epoch": 0.03574840157761799, + "grad_norm": 2.6230552196502686, + "learning_rate": 0.0001993779482163829, + "loss": 0.4752, + "step": 1525 + }, + { + "epoch": 0.03577184315242298, + "grad_norm": 2.912607192993164, + "learning_rate": 0.0001993771276087292, + "loss": 0.5532, + "step": 1526 + }, + { + "epoch": 0.035795284727227976, + "grad_norm": 4.025883197784424, + "learning_rate": 0.0001993763064618524, + "loss": 0.6679, + "step": 1527 + }, + { + "epoch": 0.03581872630203297, + "grad_norm": 4.5012664794921875, + "learning_rate": 0.00019937548477575695, + "loss": 0.8012, + "step": 1528 + }, + { + "epoch": 0.035842167876837965, + "grad_norm": 2.8747146129608154, + "learning_rate": 0.00019937466255044736, + "loss": 0.6945, + "step": 1529 + }, + { + "epoch": 0.03586560945164296, + "grad_norm": 5.158900260925293, + "learning_rate": 0.00019937383978592806, + "loss": 0.6035, + "step": 1530 + }, + { + "epoch": 0.03588905102644796, + "grad_norm": 1.6030853986740112, + "learning_rate": 0.0001993730164822035, + "loss": 0.2718, + "step": 1531 + }, + { + "epoch": 0.035912492601252954, + "grad_norm": 2.1739726066589355, + "learning_rate": 0.00019937219263927816, + "loss": 1.707, + "step": 1532 + }, + { + "epoch": 0.03593593417605795, + "grad_norm": 3.572493314743042, + "learning_rate": 0.00019937136825715648, + "loss": 0.357, + "step": 1533 + }, + { + "epoch": 0.03595937575086294, + "grad_norm": 4.8702006340026855, + "learning_rate": 0.000199370543335843, + "loss": 1.1275, + "step": 1534 + }, + { + "epoch": 0.03598281732566794, + "grad_norm": 2.7089672088623047, + "learning_rate": 0.00019936971787534212, + "loss": 0.3926, + "step": 1535 + }, + { + "epoch": 0.03600625890047293, + "grad_norm": 1.0660550594329834, + "learning_rate": 0.0001993688918756584, + "loss": 0.1228, + "step": 1536 + }, + { + "epoch": 0.036029700475277926, + "grad_norm": 2.7121355533599854, + "learning_rate": 0.00019936806533679623, + "loss": 0.3325, + "step": 1537 + }, + { + "epoch": 0.03605314205008293, + "grad_norm": 4.01973295211792, + "learning_rate": 0.00019936723825876018, + "loss": 0.6209, + "step": 1538 + }, + { + "epoch": 0.03607658362488792, + "grad_norm": 5.454864978790283, + "learning_rate": 0.0001993664106415547, + "loss": 1.0672, + "step": 1539 + }, + { + "epoch": 0.036100025199692916, + "grad_norm": 0.7747009992599487, + "learning_rate": 0.00019936558248518429, + "loss": 0.1425, + "step": 1540 + }, + { + "epoch": 0.03612346677449791, + "grad_norm": 3.8783576488494873, + "learning_rate": 0.00019936475378965341, + "loss": 0.931, + "step": 1541 + }, + { + "epoch": 0.036146908349302904, + "grad_norm": 5.499748706817627, + "learning_rate": 0.00019936392455496658, + "loss": 0.4081, + "step": 1542 + }, + { + "epoch": 0.0361703499241079, + "grad_norm": 3.7498528957366943, + "learning_rate": 0.00019936309478112832, + "loss": 0.6591, + "step": 1543 + }, + { + "epoch": 0.0361937914989129, + "grad_norm": 3.6355509757995605, + "learning_rate": 0.00019936226446814312, + "loss": 0.6812, + "step": 1544 + }, + { + "epoch": 0.036217233073717894, + "grad_norm": 3.905284881591797, + "learning_rate": 0.00019936143361601547, + "loss": 0.4872, + "step": 1545 + }, + { + "epoch": 0.03624067464852289, + "grad_norm": 2.179368495941162, + "learning_rate": 0.00019936060222474987, + "loss": 0.4179, + "step": 1546 + }, + { + "epoch": 0.03626411622332788, + "grad_norm": 2.9744296073913574, + "learning_rate": 0.00019935977029435087, + "loss": 0.5261, + "step": 1547 + }, + { + "epoch": 0.03628755779813288, + "grad_norm": 1.464565396308899, + "learning_rate": 0.00019935893782482297, + "loss": 0.3295, + "step": 1548 + }, + { + "epoch": 0.03631099937293787, + "grad_norm": 3.2996456623077393, + "learning_rate": 0.00019935810481617065, + "loss": 0.6491, + "step": 1549 + }, + { + "epoch": 0.036334440947742866, + "grad_norm": 4.251966953277588, + "learning_rate": 0.00019935727126839847, + "loss": 0.7955, + "step": 1550 + }, + { + "epoch": 0.03635788252254787, + "grad_norm": 4.8504791259765625, + "learning_rate": 0.00019935643718151097, + "loss": 0.8242, + "step": 1551 + }, + { + "epoch": 0.03638132409735286, + "grad_norm": 3.8871428966522217, + "learning_rate": 0.0001993556025555126, + "loss": 0.562, + "step": 1552 + }, + { + "epoch": 0.036404765672157856, + "grad_norm": 4.574100971221924, + "learning_rate": 0.00019935476739040798, + "loss": 1.0046, + "step": 1553 + }, + { + "epoch": 0.03642820724696285, + "grad_norm": 1.215169906616211, + "learning_rate": 0.0001993539316862016, + "loss": 0.1087, + "step": 1554 + }, + { + "epoch": 0.036451648821767844, + "grad_norm": 4.731846809387207, + "learning_rate": 0.00019935309544289796, + "loss": 0.7309, + "step": 1555 + }, + { + "epoch": 0.03647509039657284, + "grad_norm": 3.63747239112854, + "learning_rate": 0.00019935225866050163, + "loss": 0.6521, + "step": 1556 + }, + { + "epoch": 0.03649853197137784, + "grad_norm": 2.4009838104248047, + "learning_rate": 0.00019935142133901715, + "loss": 0.4547, + "step": 1557 + }, + { + "epoch": 0.036521973546182834, + "grad_norm": 1.1979985237121582, + "learning_rate": 0.00019935058347844912, + "loss": 0.2417, + "step": 1558 + }, + { + "epoch": 0.03654541512098783, + "grad_norm": 3.1498281955718994, + "learning_rate": 0.00019934974507880197, + "loss": 0.804, + "step": 1559 + }, + { + "epoch": 0.03656885669579282, + "grad_norm": 3.7028956413269043, + "learning_rate": 0.00019934890614008036, + "loss": 0.532, + "step": 1560 + }, + { + "epoch": 0.03659229827059782, + "grad_norm": 0.9870073199272156, + "learning_rate": 0.00019934806666228874, + "loss": 0.1195, + "step": 1561 + }, + { + "epoch": 0.03661573984540281, + "grad_norm": 1.9484481811523438, + "learning_rate": 0.00019934722664543175, + "loss": 0.3559, + "step": 1562 + }, + { + "epoch": 0.03663918142020781, + "grad_norm": 3.271857976913452, + "learning_rate": 0.00019934638608951391, + "loss": 0.4686, + "step": 1563 + }, + { + "epoch": 0.03666262299501281, + "grad_norm": 3.3542561531066895, + "learning_rate": 0.00019934554499453976, + "loss": 0.4137, + "step": 1564 + }, + { + "epoch": 0.0366860645698178, + "grad_norm": 3.694692850112915, + "learning_rate": 0.00019934470336051388, + "loss": 0.7635, + "step": 1565 + }, + { + "epoch": 0.036709506144622796, + "grad_norm": 1.8059203624725342, + "learning_rate": 0.0001993438611874409, + "loss": 0.3465, + "step": 1566 + }, + { + "epoch": 0.03673294771942779, + "grad_norm": 4.57277250289917, + "learning_rate": 0.00019934301847532532, + "loss": 0.5461, + "step": 1567 + }, + { + "epoch": 0.036756389294232784, + "grad_norm": 3.622504234313965, + "learning_rate": 0.0001993421752241717, + "loss": 0.5287, + "step": 1568 + }, + { + "epoch": 0.03677983086903778, + "grad_norm": 2.7085230350494385, + "learning_rate": 0.00019934133143398468, + "loss": 0.7102, + "step": 1569 + }, + { + "epoch": 0.03680327244384278, + "grad_norm": 0.9523638486862183, + "learning_rate": 0.0001993404871047688, + "loss": 0.1463, + "step": 1570 + }, + { + "epoch": 0.036826714018647774, + "grad_norm": 4.132114887237549, + "learning_rate": 0.00019933964223652862, + "loss": 0.6207, + "step": 1571 + }, + { + "epoch": 0.03685015559345277, + "grad_norm": 1.0533769130706787, + "learning_rate": 0.00019933879682926874, + "loss": 0.2301, + "step": 1572 + }, + { + "epoch": 0.03687359716825776, + "grad_norm": 3.1163740158081055, + "learning_rate": 0.0001993379508829938, + "loss": 0.8933, + "step": 1573 + }, + { + "epoch": 0.03689703874306276, + "grad_norm": 2.5379116535186768, + "learning_rate": 0.0001993371043977083, + "loss": 0.3475, + "step": 1574 + }, + { + "epoch": 0.03692048031786775, + "grad_norm": 4.52837610244751, + "learning_rate": 0.0001993362573734169, + "loss": 1.084, + "step": 1575 + }, + { + "epoch": 0.03694392189267275, + "grad_norm": 4.706670761108398, + "learning_rate": 0.00019933540981012418, + "loss": 0.6593, + "step": 1576 + }, + { + "epoch": 0.03696736346747775, + "grad_norm": 4.200976848602295, + "learning_rate": 0.00019933456170783472, + "loss": 0.5987, + "step": 1577 + }, + { + "epoch": 0.03699080504228274, + "grad_norm": 4.569613456726074, + "learning_rate": 0.00019933371306655312, + "loss": 0.4935, + "step": 1578 + }, + { + "epoch": 0.037014246617087736, + "grad_norm": 2.401195526123047, + "learning_rate": 0.00019933286388628405, + "loss": 1.8375, + "step": 1579 + }, + { + "epoch": 0.03703768819189273, + "grad_norm": 1.4977973699569702, + "learning_rate": 0.00019933201416703203, + "loss": 0.2516, + "step": 1580 + }, + { + "epoch": 0.037061129766697724, + "grad_norm": 3.331888198852539, + "learning_rate": 0.0001993311639088017, + "loss": 0.6131, + "step": 1581 + }, + { + "epoch": 0.037084571341502726, + "grad_norm": 3.343209981918335, + "learning_rate": 0.0001993303131115977, + "loss": 0.6056, + "step": 1582 + }, + { + "epoch": 0.03710801291630772, + "grad_norm": 2.8723182678222656, + "learning_rate": 0.0001993294617754246, + "loss": 0.3792, + "step": 1583 + }, + { + "epoch": 0.037131454491112714, + "grad_norm": 2.720898151397705, + "learning_rate": 0.0001993286099002871, + "loss": 0.5226, + "step": 1584 + }, + { + "epoch": 0.03715489606591771, + "grad_norm": 1.0999869108200073, + "learning_rate": 0.00019932775748618975, + "loss": 0.1907, + "step": 1585 + }, + { + "epoch": 0.0371783376407227, + "grad_norm": 6.146969795227051, + "learning_rate": 0.0001993269045331372, + "loss": 1.2134, + "step": 1586 + }, + { + "epoch": 0.0372017792155277, + "grad_norm": 2.63632869720459, + "learning_rate": 0.00019932605104113404, + "loss": 0.5414, + "step": 1587 + }, + { + "epoch": 0.03722522079033269, + "grad_norm": 3.6193628311157227, + "learning_rate": 0.00019932519701018495, + "loss": 0.6122, + "step": 1588 + }, + { + "epoch": 0.03724866236513769, + "grad_norm": 1.3879625797271729, + "learning_rate": 0.00019932434244029456, + "loss": 0.1315, + "step": 1589 + }, + { + "epoch": 0.03727210393994269, + "grad_norm": 1.0915324687957764, + "learning_rate": 0.00019932348733146747, + "loss": 0.1724, + "step": 1590 + }, + { + "epoch": 0.03729554551474768, + "grad_norm": 4.329133987426758, + "learning_rate": 0.00019932263168370837, + "loss": 0.5604, + "step": 1591 + }, + { + "epoch": 0.037318987089552676, + "grad_norm": 5.818304538726807, + "learning_rate": 0.0001993217754970219, + "loss": 0.4638, + "step": 1592 + }, + { + "epoch": 0.03734242866435767, + "grad_norm": 0.6013354063034058, + "learning_rate": 0.00019932091877141264, + "loss": 0.0793, + "step": 1593 + }, + { + "epoch": 0.037365870239162664, + "grad_norm": 2.613110303878784, + "learning_rate": 0.0001993200615068853, + "loss": 0.3115, + "step": 1594 + }, + { + "epoch": 0.037389311813967666, + "grad_norm": 2.1498517990112305, + "learning_rate": 0.00019931920370344451, + "loss": 0.4066, + "step": 1595 + }, + { + "epoch": 0.03741275338877266, + "grad_norm": 1.8967490196228027, + "learning_rate": 0.00019931834536109492, + "loss": 1.7505, + "step": 1596 + }, + { + "epoch": 0.037436194963577654, + "grad_norm": 4.186997413635254, + "learning_rate": 0.0001993174864798412, + "loss": 0.6329, + "step": 1597 + }, + { + "epoch": 0.03745963653838265, + "grad_norm": 2.2762510776519775, + "learning_rate": 0.000199316627059688, + "loss": 1.7193, + "step": 1598 + }, + { + "epoch": 0.03748307811318764, + "grad_norm": 3.5428428649902344, + "learning_rate": 0.00019931576710064002, + "loss": 0.3736, + "step": 1599 + }, + { + "epoch": 0.03750651968799264, + "grad_norm": 3.9062206745147705, + "learning_rate": 0.00019931490660270187, + "loss": 0.4803, + "step": 1600 + }, + { + "epoch": 0.03752996126279763, + "grad_norm": 7.052928924560547, + "learning_rate": 0.00019931404556587824, + "loss": 0.6742, + "step": 1601 + }, + { + "epoch": 0.03755340283760263, + "grad_norm": 1.8889349699020386, + "learning_rate": 0.00019931318399017383, + "loss": 1.616, + "step": 1602 + }, + { + "epoch": 0.03757684441240763, + "grad_norm": 2.615560293197632, + "learning_rate": 0.00019931232187559328, + "loss": 0.2983, + "step": 1603 + }, + { + "epoch": 0.03760028598721262, + "grad_norm": 0.5675761103630066, + "learning_rate": 0.00019931145922214127, + "loss": 0.0695, + "step": 1604 + }, + { + "epoch": 0.037623727562017616, + "grad_norm": 1.2566701173782349, + "learning_rate": 0.0001993105960298225, + "loss": 0.2114, + "step": 1605 + }, + { + "epoch": 0.03764716913682261, + "grad_norm": 4.064579486846924, + "learning_rate": 0.00019930973229864162, + "loss": 0.5419, + "step": 1606 + }, + { + "epoch": 0.037670610711627604, + "grad_norm": 1.2629179954528809, + "learning_rate": 0.00019930886802860335, + "loss": 0.2808, + "step": 1607 + }, + { + "epoch": 0.037694052286432606, + "grad_norm": 1.0688257217407227, + "learning_rate": 0.0001993080032197124, + "loss": 0.2506, + "step": 1608 + }, + { + "epoch": 0.0377174938612376, + "grad_norm": 3.5284626483917236, + "learning_rate": 0.00019930713787197338, + "loss": 0.7871, + "step": 1609 + }, + { + "epoch": 0.037740935436042594, + "grad_norm": 2.238752603530884, + "learning_rate": 0.00019930627198539108, + "loss": 0.4927, + "step": 1610 + }, + { + "epoch": 0.03776437701084759, + "grad_norm": 5.3116865158081055, + "learning_rate": 0.00019930540555997017, + "loss": 0.8898, + "step": 1611 + }, + { + "epoch": 0.03778781858565258, + "grad_norm": 4.705801010131836, + "learning_rate": 0.00019930453859571528, + "loss": 0.5959, + "step": 1612 + }, + { + "epoch": 0.03781126016045758, + "grad_norm": 3.6300032138824463, + "learning_rate": 0.0001993036710926312, + "loss": 0.855, + "step": 1613 + }, + { + "epoch": 0.03783470173526258, + "grad_norm": 4.482497215270996, + "learning_rate": 0.00019930280305072265, + "loss": 0.9795, + "step": 1614 + }, + { + "epoch": 0.03785814331006757, + "grad_norm": 1.073758602142334, + "learning_rate": 0.00019930193446999423, + "loss": 0.8102, + "step": 1615 + }, + { + "epoch": 0.03788158488487257, + "grad_norm": 4.212600231170654, + "learning_rate": 0.00019930106535045075, + "loss": 0.4438, + "step": 1616 + }, + { + "epoch": 0.03790502645967756, + "grad_norm": 2.5619020462036133, + "learning_rate": 0.00019930019569209693, + "loss": 0.6408, + "step": 1617 + }, + { + "epoch": 0.037928468034482556, + "grad_norm": 1.988410472869873, + "learning_rate": 0.0001992993254949374, + "loss": 1.6648, + "step": 1618 + }, + { + "epoch": 0.03795190960928755, + "grad_norm": 0.3639293611049652, + "learning_rate": 0.00019929845475897696, + "loss": 0.0645, + "step": 1619 + }, + { + "epoch": 0.037975351184092544, + "grad_norm": 4.297715187072754, + "learning_rate": 0.00019929758348422033, + "loss": 0.9027, + "step": 1620 + }, + { + "epoch": 0.037998792758897545, + "grad_norm": 1.4756914377212524, + "learning_rate": 0.0001992967116706722, + "loss": 0.3415, + "step": 1621 + }, + { + "epoch": 0.03802223433370254, + "grad_norm": 0.8808648586273193, + "learning_rate": 0.00019929583931833732, + "loss": 0.178, + "step": 1622 + }, + { + "epoch": 0.038045675908507534, + "grad_norm": 3.902466058731079, + "learning_rate": 0.00019929496642722043, + "loss": 0.7528, + "step": 1623 + }, + { + "epoch": 0.03806911748331253, + "grad_norm": 1.3593579530715942, + "learning_rate": 0.00019929409299732626, + "loss": 0.3291, + "step": 1624 + }, + { + "epoch": 0.03809255905811752, + "grad_norm": 1.2790881395339966, + "learning_rate": 0.00019929321902865957, + "loss": 0.2259, + "step": 1625 + }, + { + "epoch": 0.03811600063292252, + "grad_norm": 0.9277656078338623, + "learning_rate": 0.00019929234452122505, + "loss": 0.1785, + "step": 1626 + }, + { + "epoch": 0.03813944220772752, + "grad_norm": 2.6077215671539307, + "learning_rate": 0.00019929146947502747, + "loss": 0.5207, + "step": 1627 + }, + { + "epoch": 0.03816288378253251, + "grad_norm": 1.5184082984924316, + "learning_rate": 0.0001992905938900716, + "loss": 0.2466, + "step": 1628 + }, + { + "epoch": 0.03818632535733751, + "grad_norm": 3.1207101345062256, + "learning_rate": 0.00019928971776636217, + "loss": 0.3524, + "step": 1629 + }, + { + "epoch": 0.0382097669321425, + "grad_norm": 3.08658504486084, + "learning_rate": 0.00019928884110390397, + "loss": 0.5348, + "step": 1630 + }, + { + "epoch": 0.038233208506947496, + "grad_norm": 1.0457643270492554, + "learning_rate": 0.0001992879639027017, + "loss": 0.1379, + "step": 1631 + }, + { + "epoch": 0.03825665008175249, + "grad_norm": 1.5702301263809204, + "learning_rate": 0.00019928708616276015, + "loss": 0.3277, + "step": 1632 + }, + { + "epoch": 0.038280091656557484, + "grad_norm": 1.74453604221344, + "learning_rate": 0.00019928620788408407, + "loss": 0.309, + "step": 1633 + }, + { + "epoch": 0.038303533231362485, + "grad_norm": 1.8825715780258179, + "learning_rate": 0.00019928532906667826, + "loss": 1.7052, + "step": 1634 + }, + { + "epoch": 0.03832697480616748, + "grad_norm": 2.4693610668182373, + "learning_rate": 0.00019928444971054746, + "loss": 0.2953, + "step": 1635 + }, + { + "epoch": 0.038350416380972474, + "grad_norm": 1.3485949039459229, + "learning_rate": 0.00019928356981569641, + "loss": 0.3053, + "step": 1636 + }, + { + "epoch": 0.03837385795577747, + "grad_norm": 1.1532704830169678, + "learning_rate": 0.00019928268938212992, + "loss": 0.2823, + "step": 1637 + }, + { + "epoch": 0.03839729953058246, + "grad_norm": 4.878436088562012, + "learning_rate": 0.0001992818084098528, + "loss": 0.8949, + "step": 1638 + }, + { + "epoch": 0.03842074110538746, + "grad_norm": 2.6626088619232178, + "learning_rate": 0.00019928092689886978, + "loss": 0.3938, + "step": 1639 + }, + { + "epoch": 0.03844418268019246, + "grad_norm": 3.903548002243042, + "learning_rate": 0.00019928004484918565, + "loss": 0.7981, + "step": 1640 + }, + { + "epoch": 0.03846762425499745, + "grad_norm": 3.206761598587036, + "learning_rate": 0.00019927916226080523, + "loss": 0.5909, + "step": 1641 + }, + { + "epoch": 0.03849106582980245, + "grad_norm": 4.2860565185546875, + "learning_rate": 0.00019927827913373327, + "loss": 0.7287, + "step": 1642 + }, + { + "epoch": 0.03851450740460744, + "grad_norm": 2.86395263671875, + "learning_rate": 0.00019927739546797457, + "loss": 0.5503, + "step": 1643 + }, + { + "epoch": 0.038537948979412436, + "grad_norm": 0.25649964809417725, + "learning_rate": 0.00019927651126353392, + "loss": 0.0295, + "step": 1644 + }, + { + "epoch": 0.03856139055421743, + "grad_norm": 2.2192842960357666, + "learning_rate": 0.00019927562652041615, + "loss": 0.3567, + "step": 1645 + }, + { + "epoch": 0.03858483212902243, + "grad_norm": 2.2847588062286377, + "learning_rate": 0.00019927474123862605, + "loss": 0.3469, + "step": 1646 + }, + { + "epoch": 0.038608273703827425, + "grad_norm": 5.950641632080078, + "learning_rate": 0.00019927385541816838, + "loss": 1.1413, + "step": 1647 + }, + { + "epoch": 0.03863171527863242, + "grad_norm": 1.1900821924209595, + "learning_rate": 0.00019927296905904803, + "loss": 0.1886, + "step": 1648 + }, + { + "epoch": 0.038655156853437414, + "grad_norm": 3.999992847442627, + "learning_rate": 0.0001992720821612697, + "loss": 0.7641, + "step": 1649 + }, + { + "epoch": 0.03867859842824241, + "grad_norm": 3.3108391761779785, + "learning_rate": 0.00019927119472483827, + "loss": 0.7436, + "step": 1650 + }, + { + "epoch": 0.0387020400030474, + "grad_norm": 0.93287593126297, + "learning_rate": 0.00019927030674975858, + "loss": 0.1994, + "step": 1651 + }, + { + "epoch": 0.0387254815778524, + "grad_norm": 2.1334540843963623, + "learning_rate": 0.00019926941823603537, + "loss": 0.4255, + "step": 1652 + }, + { + "epoch": 0.0387489231526574, + "grad_norm": 1.151183843612671, + "learning_rate": 0.00019926852918367356, + "loss": 0.1817, + "step": 1653 + }, + { + "epoch": 0.03877236472746239, + "grad_norm": 1.3346911668777466, + "learning_rate": 0.00019926763959267788, + "loss": 0.1805, + "step": 1654 + }, + { + "epoch": 0.03879580630226739, + "grad_norm": 2.3302559852600098, + "learning_rate": 0.00019926674946305322, + "loss": 1.515, + "step": 1655 + }, + { + "epoch": 0.03881924787707238, + "grad_norm": 6.932552814483643, + "learning_rate": 0.00019926585879480438, + "loss": 0.4897, + "step": 1656 + }, + { + "epoch": 0.038842689451877375, + "grad_norm": 3.890144109725952, + "learning_rate": 0.00019926496758793615, + "loss": 0.8259, + "step": 1657 + }, + { + "epoch": 0.03886613102668237, + "grad_norm": 3.2277719974517822, + "learning_rate": 0.0001992640758424535, + "loss": 0.5924, + "step": 1658 + }, + { + "epoch": 0.03888957260148737, + "grad_norm": 4.36138391494751, + "learning_rate": 0.00019926318355836112, + "loss": 0.8278, + "step": 1659 + }, + { + "epoch": 0.038913014176292365, + "grad_norm": 4.323398113250732, + "learning_rate": 0.00019926229073566394, + "loss": 0.5105, + "step": 1660 + }, + { + "epoch": 0.03893645575109736, + "grad_norm": 2.592508316040039, + "learning_rate": 0.00019926139737436676, + "loss": 0.2214, + "step": 1661 + }, + { + "epoch": 0.038959897325902354, + "grad_norm": 4.363956451416016, + "learning_rate": 0.00019926050347447447, + "loss": 1.2189, + "step": 1662 + }, + { + "epoch": 0.03898333890070735, + "grad_norm": 0.40727412700653076, + "learning_rate": 0.0001992596090359919, + "loss": 0.0349, + "step": 1663 + }, + { + "epoch": 0.03900678047551234, + "grad_norm": 0.8596699237823486, + "learning_rate": 0.0001992587140589239, + "loss": 0.1432, + "step": 1664 + }, + { + "epoch": 0.03903022205031734, + "grad_norm": 4.283016681671143, + "learning_rate": 0.00019925781854327528, + "loss": 0.6295, + "step": 1665 + }, + { + "epoch": 0.03905366362512234, + "grad_norm": 4.101322650909424, + "learning_rate": 0.000199256922489051, + "loss": 0.6724, + "step": 1666 + }, + { + "epoch": 0.03907710519992733, + "grad_norm": 1.451493501663208, + "learning_rate": 0.00019925602589625582, + "loss": 0.2667, + "step": 1667 + }, + { + "epoch": 0.03910054677473233, + "grad_norm": 4.141050338745117, + "learning_rate": 0.0001992551287648947, + "loss": 0.6801, + "step": 1668 + }, + { + "epoch": 0.03912398834953732, + "grad_norm": 1.1656968593597412, + "learning_rate": 0.00019925423109497245, + "loss": 0.1089, + "step": 1669 + }, + { + "epoch": 0.039147429924342315, + "grad_norm": 5.313874244689941, + "learning_rate": 0.00019925333288649393, + "loss": 0.7337, + "step": 1670 + }, + { + "epoch": 0.03917087149914731, + "grad_norm": 5.131332874298096, + "learning_rate": 0.00019925243413946407, + "loss": 0.7801, + "step": 1671 + }, + { + "epoch": 0.03919431307395231, + "grad_norm": 3.614569664001465, + "learning_rate": 0.00019925153485388767, + "loss": 0.5068, + "step": 1672 + }, + { + "epoch": 0.039217754648757305, + "grad_norm": 1.6065857410430908, + "learning_rate": 0.00019925063502976967, + "loss": 0.2015, + "step": 1673 + }, + { + "epoch": 0.0392411962235623, + "grad_norm": 3.6297078132629395, + "learning_rate": 0.00019924973466711493, + "loss": 0.6075, + "step": 1674 + }, + { + "epoch": 0.039264637798367294, + "grad_norm": 5.403473854064941, + "learning_rate": 0.00019924883376592836, + "loss": 0.769, + "step": 1675 + }, + { + "epoch": 0.03928807937317229, + "grad_norm": 3.3108527660369873, + "learning_rate": 0.00019924793232621482, + "loss": 0.5708, + "step": 1676 + }, + { + "epoch": 0.03931152094797728, + "grad_norm": 3.425447463989258, + "learning_rate": 0.0001992470303479792, + "loss": 0.4843, + "step": 1677 + }, + { + "epoch": 0.039334962522782284, + "grad_norm": 5.548299312591553, + "learning_rate": 0.0001992461278312264, + "loss": 0.7944, + "step": 1678 + }, + { + "epoch": 0.03935840409758728, + "grad_norm": 3.3905587196350098, + "learning_rate": 0.0001992452247759613, + "loss": 0.5375, + "step": 1679 + }, + { + "epoch": 0.03938184567239227, + "grad_norm": 4.980292797088623, + "learning_rate": 0.00019924432118218884, + "loss": 1.0696, + "step": 1680 + }, + { + "epoch": 0.03940528724719727, + "grad_norm": 3.5768187046051025, + "learning_rate": 0.0001992434170499139, + "loss": 0.5115, + "step": 1681 + }, + { + "epoch": 0.03942872882200226, + "grad_norm": 1.8731770515441895, + "learning_rate": 0.0001992425123791414, + "loss": 1.5727, + "step": 1682 + }, + { + "epoch": 0.039452170396807255, + "grad_norm": 2.353754758834839, + "learning_rate": 0.00019924160716987627, + "loss": 1.8148, + "step": 1683 + }, + { + "epoch": 0.03947561197161225, + "grad_norm": 2.468655824661255, + "learning_rate": 0.00019924070142212332, + "loss": 0.5991, + "step": 1684 + }, + { + "epoch": 0.03949905354641725, + "grad_norm": 4.240562915802002, + "learning_rate": 0.0001992397951358876, + "loss": 0.4535, + "step": 1685 + }, + { + "epoch": 0.039522495121222245, + "grad_norm": 3.285759925842285, + "learning_rate": 0.0001992388883111739, + "loss": 0.5671, + "step": 1686 + }, + { + "epoch": 0.03954593669602724, + "grad_norm": 3.8743157386779785, + "learning_rate": 0.00019923798094798723, + "loss": 0.5277, + "step": 1687 + }, + { + "epoch": 0.039569378270832234, + "grad_norm": 2.6001462936401367, + "learning_rate": 0.00019923707304633247, + "loss": 0.3325, + "step": 1688 + }, + { + "epoch": 0.03959281984563723, + "grad_norm": 2.1135094165802, + "learning_rate": 0.00019923616460621458, + "loss": 1.8879, + "step": 1689 + }, + { + "epoch": 0.03961626142044222, + "grad_norm": 1.0850679874420166, + "learning_rate": 0.00019923525562763844, + "loss": 0.1179, + "step": 1690 + }, + { + "epoch": 0.039639702995247224, + "grad_norm": 2.4407753944396973, + "learning_rate": 0.00019923434611060907, + "loss": 1.9153, + "step": 1691 + }, + { + "epoch": 0.03966314457005222, + "grad_norm": 1.3592243194580078, + "learning_rate": 0.0001992334360551313, + "loss": 0.2583, + "step": 1692 + }, + { + "epoch": 0.03968658614485721, + "grad_norm": 3.2671658992767334, + "learning_rate": 0.00019923252546121014, + "loss": 1.0054, + "step": 1693 + }, + { + "epoch": 0.03971002771966221, + "grad_norm": 2.858213186264038, + "learning_rate": 0.0001992316143288505, + "loss": 0.5378, + "step": 1694 + }, + { + "epoch": 0.0397334692944672, + "grad_norm": 5.021307468414307, + "learning_rate": 0.0001992307026580573, + "loss": 0.7382, + "step": 1695 + }, + { + "epoch": 0.039756910869272195, + "grad_norm": 1.9380240440368652, + "learning_rate": 0.00019922979044883555, + "loss": 0.3036, + "step": 1696 + }, + { + "epoch": 0.03978035244407719, + "grad_norm": 1.5784724950790405, + "learning_rate": 0.00019922887770119015, + "loss": 0.345, + "step": 1697 + }, + { + "epoch": 0.03980379401888219, + "grad_norm": 3.165639638900757, + "learning_rate": 0.0001992279644151261, + "loss": 0.6837, + "step": 1698 + }, + { + "epoch": 0.039827235593687185, + "grad_norm": 3.0825865268707275, + "learning_rate": 0.0001992270505906483, + "loss": 0.547, + "step": 1699 + }, + { + "epoch": 0.03985067716849218, + "grad_norm": 2.0825893878936768, + "learning_rate": 0.00019922613622776173, + "loss": 0.5101, + "step": 1700 + }, + { + "epoch": 0.039874118743297174, + "grad_norm": 2.88625168800354, + "learning_rate": 0.00019922522132647137, + "loss": 2.0261, + "step": 1701 + }, + { + "epoch": 0.03989756031810217, + "grad_norm": 3.2512142658233643, + "learning_rate": 0.00019922430588678215, + "loss": 0.6949, + "step": 1702 + }, + { + "epoch": 0.03992100189290716, + "grad_norm": 0.5963603258132935, + "learning_rate": 0.00019922338990869908, + "loss": 0.1086, + "step": 1703 + }, + { + "epoch": 0.039944443467712164, + "grad_norm": 2.2257332801818848, + "learning_rate": 0.0001992224733922271, + "loss": 0.135, + "step": 1704 + }, + { + "epoch": 0.03996788504251716, + "grad_norm": 3.237866163253784, + "learning_rate": 0.00019922155633737121, + "loss": 1.5244, + "step": 1705 + }, + { + "epoch": 0.03999132661732215, + "grad_norm": 6.139493465423584, + "learning_rate": 0.0001992206387441363, + "loss": 0.8022, + "step": 1706 + }, + { + "epoch": 0.04001476819212715, + "grad_norm": 0.5752692222595215, + "learning_rate": 0.00019921972061252746, + "loss": 0.1104, + "step": 1707 + }, + { + "epoch": 0.04003820976693214, + "grad_norm": 3.1548068523406982, + "learning_rate": 0.00019921880194254964, + "loss": 0.451, + "step": 1708 + }, + { + "epoch": 0.040061651341737135, + "grad_norm": 2.960513114929199, + "learning_rate": 0.0001992178827342078, + "loss": 0.4273, + "step": 1709 + }, + { + "epoch": 0.04008509291654214, + "grad_norm": 1.7847850322723389, + "learning_rate": 0.00019921696298750693, + "loss": 0.5212, + "step": 1710 + }, + { + "epoch": 0.04010853449134713, + "grad_norm": 3.5697994232177734, + "learning_rate": 0.00019921604270245203, + "loss": 0.668, + "step": 1711 + }, + { + "epoch": 0.040131976066152125, + "grad_norm": 3.9428346157073975, + "learning_rate": 0.00019921512187904807, + "loss": 0.5975, + "step": 1712 + }, + { + "epoch": 0.04015541764095712, + "grad_norm": 5.952017784118652, + "learning_rate": 0.0001992142005173001, + "loss": 0.7974, + "step": 1713 + }, + { + "epoch": 0.040178859215762114, + "grad_norm": 2.015113115310669, + "learning_rate": 0.00019921327861721307, + "loss": 0.3852, + "step": 1714 + }, + { + "epoch": 0.04020230079056711, + "grad_norm": 2.9007232189178467, + "learning_rate": 0.00019921235617879202, + "loss": 0.4797, + "step": 1715 + }, + { + "epoch": 0.0402257423653721, + "grad_norm": 2.642019033432007, + "learning_rate": 0.00019921143320204192, + "loss": 0.4983, + "step": 1716 + }, + { + "epoch": 0.040249183940177104, + "grad_norm": 1.2786061763763428, + "learning_rate": 0.00019921050968696783, + "loss": 0.1961, + "step": 1717 + }, + { + "epoch": 0.0402726255149821, + "grad_norm": 2.790172576904297, + "learning_rate": 0.00019920958563357467, + "loss": 0.4541, + "step": 1718 + }, + { + "epoch": 0.04029606708978709, + "grad_norm": 2.425659656524658, + "learning_rate": 0.00019920866104186752, + "loss": 0.4586, + "step": 1719 + }, + { + "epoch": 0.04031950866459209, + "grad_norm": 3.090151071548462, + "learning_rate": 0.00019920773591185141, + "loss": 0.5311, + "step": 1720 + }, + { + "epoch": 0.04034295023939708, + "grad_norm": 3.0887348651885986, + "learning_rate": 0.00019920681024353134, + "loss": 0.6401, + "step": 1721 + }, + { + "epoch": 0.040366391814202075, + "grad_norm": 3.6422011852264404, + "learning_rate": 0.00019920588403691232, + "loss": 0.6777, + "step": 1722 + }, + { + "epoch": 0.04038983338900708, + "grad_norm": 1.776387333869934, + "learning_rate": 0.00019920495729199939, + "loss": 0.2509, + "step": 1723 + }, + { + "epoch": 0.04041327496381207, + "grad_norm": 4.199507236480713, + "learning_rate": 0.00019920403000879753, + "loss": 0.7741, + "step": 1724 + }, + { + "epoch": 0.040436716538617065, + "grad_norm": 2.9704649448394775, + "learning_rate": 0.00019920310218731185, + "loss": 0.3341, + "step": 1725 + }, + { + "epoch": 0.04046015811342206, + "grad_norm": 3.4085195064544678, + "learning_rate": 0.00019920217382754735, + "loss": 0.6007, + "step": 1726 + }, + { + "epoch": 0.040483599688227054, + "grad_norm": 2.4703447818756104, + "learning_rate": 0.00019920124492950905, + "loss": 1.4504, + "step": 1727 + }, + { + "epoch": 0.04050704126303205, + "grad_norm": 1.7477608919143677, + "learning_rate": 0.00019920031549320203, + "loss": 0.3023, + "step": 1728 + }, + { + "epoch": 0.04053048283783705, + "grad_norm": 4.138105392456055, + "learning_rate": 0.0001991993855186313, + "loss": 0.5767, + "step": 1729 + }, + { + "epoch": 0.040553924412642044, + "grad_norm": 3.8408639430999756, + "learning_rate": 0.00019919845500580187, + "loss": 0.6135, + "step": 1730 + }, + { + "epoch": 0.04057736598744704, + "grad_norm": 5.089404106140137, + "learning_rate": 0.00019919752395471888, + "loss": 0.5712, + "step": 1731 + }, + { + "epoch": 0.04060080756225203, + "grad_norm": 4.5485711097717285, + "learning_rate": 0.00019919659236538733, + "loss": 0.9145, + "step": 1732 + }, + { + "epoch": 0.04062424913705703, + "grad_norm": 3.5695695877075195, + "learning_rate": 0.00019919566023781228, + "loss": 0.5549, + "step": 1733 + }, + { + "epoch": 0.04064769071186202, + "grad_norm": 2.644179582595825, + "learning_rate": 0.00019919472757199877, + "loss": 0.3291, + "step": 1734 + }, + { + "epoch": 0.040671132286667015, + "grad_norm": 3.425915479660034, + "learning_rate": 0.0001991937943679519, + "loss": 0.6184, + "step": 1735 + }, + { + "epoch": 0.040694573861472016, + "grad_norm": 3.027588129043579, + "learning_rate": 0.0001991928606256767, + "loss": 0.6188, + "step": 1736 + }, + { + "epoch": 0.04071801543627701, + "grad_norm": 3.5407633781433105, + "learning_rate": 0.00019919192634517826, + "loss": 0.5136, + "step": 1737 + }, + { + "epoch": 0.040741457011082005, + "grad_norm": 2.7972559928894043, + "learning_rate": 0.00019919099152646163, + "loss": 0.6499, + "step": 1738 + }, + { + "epoch": 0.040764898585887, + "grad_norm": 1.0976147651672363, + "learning_rate": 0.0001991900561695319, + "loss": 0.229, + "step": 1739 + }, + { + "epoch": 0.040788340160691994, + "grad_norm": 0.9518344402313232, + "learning_rate": 0.00019918912027439413, + "loss": 0.1753, + "step": 1740 + }, + { + "epoch": 0.04081178173549699, + "grad_norm": 4.262446880340576, + "learning_rate": 0.00019918818384105343, + "loss": 0.8548, + "step": 1741 + }, + { + "epoch": 0.04083522331030199, + "grad_norm": 0.65766841173172, + "learning_rate": 0.0001991872468695148, + "loss": 0.1337, + "step": 1742 + }, + { + "epoch": 0.040858664885106984, + "grad_norm": 4.667148113250732, + "learning_rate": 0.00019918630935978344, + "loss": 0.7323, + "step": 1743 + }, + { + "epoch": 0.04088210645991198, + "grad_norm": 3.946516275405884, + "learning_rate": 0.0001991853713118643, + "loss": 0.5915, + "step": 1744 + }, + { + "epoch": 0.04090554803471697, + "grad_norm": 1.3574762344360352, + "learning_rate": 0.00019918443272576265, + "loss": 0.1399, + "step": 1745 + }, + { + "epoch": 0.04092898960952197, + "grad_norm": 13.304699897766113, + "learning_rate": 0.0001991834936014834, + "loss": 0.3641, + "step": 1746 + }, + { + "epoch": 0.04095243118432696, + "grad_norm": 3.4704394340515137, + "learning_rate": 0.00019918255393903175, + "loss": 0.7439, + "step": 1747 + }, + { + "epoch": 0.040975872759131955, + "grad_norm": 3.970155954360962, + "learning_rate": 0.00019918161373841278, + "loss": 0.7563, + "step": 1748 + }, + { + "epoch": 0.040999314333936956, + "grad_norm": 3.413823127746582, + "learning_rate": 0.00019918067299963156, + "loss": 0.5292, + "step": 1749 + }, + { + "epoch": 0.04102275590874195, + "grad_norm": 1.7399957180023193, + "learning_rate": 0.00019917973172269325, + "loss": 0.2294, + "step": 1750 + }, + { + "epoch": 0.041046197483546945, + "grad_norm": 2.534682035446167, + "learning_rate": 0.00019917878990760292, + "loss": 0.4745, + "step": 1751 + }, + { + "epoch": 0.04106963905835194, + "grad_norm": 2.826441764831543, + "learning_rate": 0.00019917784755436565, + "loss": 0.2889, + "step": 1752 + }, + { + "epoch": 0.041093080633156934, + "grad_norm": 4.907565593719482, + "learning_rate": 0.00019917690466298663, + "loss": 0.8449, + "step": 1753 + }, + { + "epoch": 0.04111652220796193, + "grad_norm": 4.20374870300293, + "learning_rate": 0.00019917596123347094, + "loss": 0.5669, + "step": 1754 + }, + { + "epoch": 0.04113996378276693, + "grad_norm": 2.704724073410034, + "learning_rate": 0.00019917501726582368, + "loss": 0.3818, + "step": 1755 + }, + { + "epoch": 0.041163405357571924, + "grad_norm": 4.359482288360596, + "learning_rate": 0.00019917407276005, + "loss": 0.9527, + "step": 1756 + }, + { + "epoch": 0.04118684693237692, + "grad_norm": 4.023073196411133, + "learning_rate": 0.000199173127716155, + "loss": 0.5582, + "step": 1757 + }, + { + "epoch": 0.04121028850718191, + "grad_norm": 3.658841609954834, + "learning_rate": 0.00019917218213414382, + "loss": 0.3538, + "step": 1758 + }, + { + "epoch": 0.04123373008198691, + "grad_norm": 3.518059492111206, + "learning_rate": 0.0001991712360140216, + "loss": 0.8883, + "step": 1759 + }, + { + "epoch": 0.0412571716567919, + "grad_norm": 4.0367536544799805, + "learning_rate": 0.00019917028935579345, + "loss": 0.716, + "step": 1760 + }, + { + "epoch": 0.0412806132315969, + "grad_norm": 1.3410109281539917, + "learning_rate": 0.00019916934215946454, + "loss": 0.2453, + "step": 1761 + }, + { + "epoch": 0.041304054806401896, + "grad_norm": 1.2446757555007935, + "learning_rate": 0.00019916839442504, + "loss": 0.2481, + "step": 1762 + }, + { + "epoch": 0.04132749638120689, + "grad_norm": 1.4948440790176392, + "learning_rate": 0.00019916744615252493, + "loss": 0.162, + "step": 1763 + }, + { + "epoch": 0.041350937956011885, + "grad_norm": 3.583611488342285, + "learning_rate": 0.00019916649734192453, + "loss": 0.3628, + "step": 1764 + }, + { + "epoch": 0.04137437953081688, + "grad_norm": 4.055022716522217, + "learning_rate": 0.00019916554799324394, + "loss": 0.627, + "step": 1765 + }, + { + "epoch": 0.041397821105621874, + "grad_norm": 2.4620354175567627, + "learning_rate": 0.00019916459810648827, + "loss": 0.4256, + "step": 1766 + }, + { + "epoch": 0.04142126268042687, + "grad_norm": 1.8027336597442627, + "learning_rate": 0.0001991636476816627, + "loss": 0.2635, + "step": 1767 + }, + { + "epoch": 0.04144470425523187, + "grad_norm": 1.4931143522262573, + "learning_rate": 0.00019916269671877243, + "loss": 0.2607, + "step": 1768 + }, + { + "epoch": 0.041468145830036864, + "grad_norm": 3.6328792572021484, + "learning_rate": 0.00019916174521782254, + "loss": 0.6454, + "step": 1769 + }, + { + "epoch": 0.04149158740484186, + "grad_norm": 5.680444240570068, + "learning_rate": 0.00019916079317881824, + "loss": 0.6681, + "step": 1770 + }, + { + "epoch": 0.04151502897964685, + "grad_norm": 1.018545389175415, + "learning_rate": 0.0001991598406017647, + "loss": 0.1386, + "step": 1771 + }, + { + "epoch": 0.041538470554451846, + "grad_norm": 0.5720654129981995, + "learning_rate": 0.00019915888748666706, + "loss": 0.0703, + "step": 1772 + }, + { + "epoch": 0.04156191212925684, + "grad_norm": 1.1103003025054932, + "learning_rate": 0.0001991579338335305, + "loss": 0.1972, + "step": 1773 + }, + { + "epoch": 0.04158535370406184, + "grad_norm": 3.2873778343200684, + "learning_rate": 0.00019915697964236024, + "loss": 0.7604, + "step": 1774 + }, + { + "epoch": 0.041608795278866836, + "grad_norm": 4.2877068519592285, + "learning_rate": 0.0001991560249131614, + "loss": 0.6177, + "step": 1775 + }, + { + "epoch": 0.04163223685367183, + "grad_norm": 4.0599565505981445, + "learning_rate": 0.00019915506964593918, + "loss": 0.5922, + "step": 1776 + }, + { + "epoch": 0.041655678428476825, + "grad_norm": 7.062565326690674, + "learning_rate": 0.00019915411384069877, + "loss": 0.6456, + "step": 1777 + }, + { + "epoch": 0.04167912000328182, + "grad_norm": 0.35522493720054626, + "learning_rate": 0.00019915315749744535, + "loss": 0.0705, + "step": 1778 + }, + { + "epoch": 0.041702561578086814, + "grad_norm": 0.568428099155426, + "learning_rate": 0.0001991522006161841, + "loss": 0.0951, + "step": 1779 + }, + { + "epoch": 0.04172600315289181, + "grad_norm": 3.487255096435547, + "learning_rate": 0.00019915124319692023, + "loss": 0.6708, + "step": 1780 + }, + { + "epoch": 0.04174944472769681, + "grad_norm": 2.924098491668701, + "learning_rate": 0.00019915028523965892, + "loss": 1.7166, + "step": 1781 + }, + { + "epoch": 0.041772886302501804, + "grad_norm": 3.9876248836517334, + "learning_rate": 0.00019914932674440538, + "loss": 1.0057, + "step": 1782 + }, + { + "epoch": 0.0417963278773068, + "grad_norm": 1.522481918334961, + "learning_rate": 0.00019914836771116483, + "loss": 0.2872, + "step": 1783 + }, + { + "epoch": 0.04181976945211179, + "grad_norm": 0.5626645088195801, + "learning_rate": 0.00019914740813994238, + "loss": 0.1284, + "step": 1784 + }, + { + "epoch": 0.041843211026916786, + "grad_norm": 5.388179302215576, + "learning_rate": 0.00019914644803074337, + "loss": 0.8437, + "step": 1785 + }, + { + "epoch": 0.04186665260172178, + "grad_norm": 5.2400007247924805, + "learning_rate": 0.00019914548738357292, + "loss": 0.7326, + "step": 1786 + }, + { + "epoch": 0.04189009417652678, + "grad_norm": 3.3692705631256104, + "learning_rate": 0.00019914452619843628, + "loss": 0.5389, + "step": 1787 + }, + { + "epoch": 0.041913535751331776, + "grad_norm": 1.4365339279174805, + "learning_rate": 0.00019914356447533864, + "loss": 0.1934, + "step": 1788 + }, + { + "epoch": 0.04193697732613677, + "grad_norm": 2.5785739421844482, + "learning_rate": 0.00019914260221428525, + "loss": 0.4145, + "step": 1789 + }, + { + "epoch": 0.041960418900941765, + "grad_norm": 1.8355522155761719, + "learning_rate": 0.00019914163941528128, + "loss": 1.8673, + "step": 1790 + }, + { + "epoch": 0.04198386047574676, + "grad_norm": 1.2907265424728394, + "learning_rate": 0.000199140676078332, + "loss": 1.4282, + "step": 1791 + }, + { + "epoch": 0.042007302050551754, + "grad_norm": 2.8721578121185303, + "learning_rate": 0.00019913971220344266, + "loss": 0.3393, + "step": 1792 + }, + { + "epoch": 0.042030743625356755, + "grad_norm": 1.6686944961547852, + "learning_rate": 0.0001991387477906184, + "loss": 0.3274, + "step": 1793 + }, + { + "epoch": 0.04205418520016175, + "grad_norm": 3.194915294647217, + "learning_rate": 0.00019913778283986452, + "loss": 0.662, + "step": 1794 + }, + { + "epoch": 0.042077626774966743, + "grad_norm": 3.4183876514434814, + "learning_rate": 0.00019913681735118628, + "loss": 0.8018, + "step": 1795 + }, + { + "epoch": 0.04210106834977174, + "grad_norm": 2.3349392414093018, + "learning_rate": 0.00019913585132458886, + "loss": 1.5052, + "step": 1796 + }, + { + "epoch": 0.04212450992457673, + "grad_norm": 3.946608304977417, + "learning_rate": 0.00019913488476007752, + "loss": 0.6045, + "step": 1797 + }, + { + "epoch": 0.042147951499381726, + "grad_norm": 1.3856470584869385, + "learning_rate": 0.0001991339176576575, + "loss": 0.2709, + "step": 1798 + }, + { + "epoch": 0.04217139307418672, + "grad_norm": 5.321515083312988, + "learning_rate": 0.00019913295001733406, + "loss": 0.9792, + "step": 1799 + }, + { + "epoch": 0.04219483464899172, + "grad_norm": 2.1704587936401367, + "learning_rate": 0.00019913198183911247, + "loss": 0.3778, + "step": 1800 + }, + { + "epoch": 0.042218276223796716, + "grad_norm": 3.666445255279541, + "learning_rate": 0.0001991310131229979, + "loss": 0.5166, + "step": 1801 + }, + { + "epoch": 0.04224171779860171, + "grad_norm": 2.5983707904815674, + "learning_rate": 0.00019913004386899573, + "loss": 1.8062, + "step": 1802 + }, + { + "epoch": 0.042265159373406705, + "grad_norm": 1.3231679201126099, + "learning_rate": 0.0001991290740771111, + "loss": 0.2478, + "step": 1803 + }, + { + "epoch": 0.0422886009482117, + "grad_norm": 3.9131646156311035, + "learning_rate": 0.00019912810374734936, + "loss": 0.748, + "step": 1804 + }, + { + "epoch": 0.042312042523016694, + "grad_norm": 3.244170904159546, + "learning_rate": 0.00019912713287971573, + "loss": 0.6407, + "step": 1805 + }, + { + "epoch": 0.042335484097821695, + "grad_norm": 3.1505069732666016, + "learning_rate": 0.00019912616147421553, + "loss": 1.617, + "step": 1806 + }, + { + "epoch": 0.04235892567262669, + "grad_norm": 1.2590187788009644, + "learning_rate": 0.00019912518953085393, + "loss": 0.2368, + "step": 1807 + }, + { + "epoch": 0.04238236724743168, + "grad_norm": 2.2956836223602295, + "learning_rate": 0.0001991242170496363, + "loss": 0.448, + "step": 1808 + }, + { + "epoch": 0.04240580882223668, + "grad_norm": 2.7760627269744873, + "learning_rate": 0.00019912324403056789, + "loss": 0.3357, + "step": 1809 + }, + { + "epoch": 0.04242925039704167, + "grad_norm": 3.362239122390747, + "learning_rate": 0.00019912227047365393, + "loss": 1.9966, + "step": 1810 + }, + { + "epoch": 0.042452691971846666, + "grad_norm": 4.24079704284668, + "learning_rate": 0.00019912129637889977, + "loss": 1.0317, + "step": 1811 + }, + { + "epoch": 0.04247613354665166, + "grad_norm": 1.1882001161575317, + "learning_rate": 0.00019912032174631067, + "loss": 0.1662, + "step": 1812 + }, + { + "epoch": 0.04249957512145666, + "grad_norm": 4.35233736038208, + "learning_rate": 0.00019911934657589192, + "loss": 0.6793, + "step": 1813 + }, + { + "epoch": 0.042523016696261656, + "grad_norm": 3.1751437187194824, + "learning_rate": 0.00019911837086764877, + "loss": 0.681, + "step": 1814 + }, + { + "epoch": 0.04254645827106665, + "grad_norm": 3.085864305496216, + "learning_rate": 0.0001991173946215866, + "loss": 0.7833, + "step": 1815 + }, + { + "epoch": 0.042569899845871645, + "grad_norm": 3.850733757019043, + "learning_rate": 0.00019911641783771063, + "loss": 0.7146, + "step": 1816 + }, + { + "epoch": 0.04259334142067664, + "grad_norm": 3.151984214782715, + "learning_rate": 0.00019911544051602622, + "loss": 0.3644, + "step": 1817 + }, + { + "epoch": 0.042616782995481634, + "grad_norm": 3.7506887912750244, + "learning_rate": 0.00019911446265653861, + "loss": 0.6432, + "step": 1818 + }, + { + "epoch": 0.042640224570286635, + "grad_norm": 3.846052646636963, + "learning_rate": 0.00019911348425925317, + "loss": 0.69, + "step": 1819 + }, + { + "epoch": 0.04266366614509163, + "grad_norm": 4.645135402679443, + "learning_rate": 0.00019911250532417512, + "loss": 0.6221, + "step": 1820 + }, + { + "epoch": 0.04268710771989662, + "grad_norm": 2.044588327407837, + "learning_rate": 0.0001991115258513099, + "loss": 0.4662, + "step": 1821 + }, + { + "epoch": 0.04271054929470162, + "grad_norm": 2.2909536361694336, + "learning_rate": 0.00019911054584066274, + "loss": 0.2452, + "step": 1822 + }, + { + "epoch": 0.04273399086950661, + "grad_norm": 4.210455894470215, + "learning_rate": 0.00019910956529223893, + "loss": 0.8567, + "step": 1823 + }, + { + "epoch": 0.042757432444311606, + "grad_norm": 3.9906296730041504, + "learning_rate": 0.00019910858420604387, + "loss": 0.4397, + "step": 1824 + }, + { + "epoch": 0.04278087401911661, + "grad_norm": 1.6123461723327637, + "learning_rate": 0.00019910760258208283, + "loss": 0.2528, + "step": 1825 + }, + { + "epoch": 0.0428043155939216, + "grad_norm": 1.9684536457061768, + "learning_rate": 0.00019910662042036116, + "loss": 0.1386, + "step": 1826 + }, + { + "epoch": 0.042827757168726596, + "grad_norm": 3.741969347000122, + "learning_rate": 0.0001991056377208842, + "loss": 0.5075, + "step": 1827 + }, + { + "epoch": 0.04285119874353159, + "grad_norm": 0.9373268485069275, + "learning_rate": 0.00019910465448365723, + "loss": 0.2021, + "step": 1828 + }, + { + "epoch": 0.042874640318336585, + "grad_norm": 3.441056966781616, + "learning_rate": 0.00019910367070868562, + "loss": 0.5196, + "step": 1829 + }, + { + "epoch": 0.04289808189314158, + "grad_norm": 3.030299425125122, + "learning_rate": 0.00019910268639597472, + "loss": 0.5345, + "step": 1830 + }, + { + "epoch": 0.042921523467946573, + "grad_norm": 3.688680648803711, + "learning_rate": 0.00019910170154552984, + "loss": 0.7151, + "step": 1831 + }, + { + "epoch": 0.042944965042751575, + "grad_norm": 3.029186725616455, + "learning_rate": 0.00019910071615735636, + "loss": 0.6728, + "step": 1832 + }, + { + "epoch": 0.04296840661755657, + "grad_norm": 3.9055347442626953, + "learning_rate": 0.0001990997302314596, + "loss": 0.6409, + "step": 1833 + }, + { + "epoch": 0.04299184819236156, + "grad_norm": 4.962074279785156, + "learning_rate": 0.00019909874376784493, + "loss": 0.6247, + "step": 1834 + }, + { + "epoch": 0.04301528976716656, + "grad_norm": 3.7279019355773926, + "learning_rate": 0.00019909775676651766, + "loss": 0.3841, + "step": 1835 + }, + { + "epoch": 0.04303873134197155, + "grad_norm": 4.612732887268066, + "learning_rate": 0.0001990967692274832, + "loss": 1.1705, + "step": 1836 + }, + { + "epoch": 0.043062172916776546, + "grad_norm": 1.9195265769958496, + "learning_rate": 0.0001990957811507469, + "loss": 1.2194, + "step": 1837 + }, + { + "epoch": 0.04308561449158155, + "grad_norm": 2.870584726333618, + "learning_rate": 0.00019909479253631406, + "loss": 0.5576, + "step": 1838 + }, + { + "epoch": 0.04310905606638654, + "grad_norm": 0.5206186175346375, + "learning_rate": 0.0001990938033841901, + "loss": 0.0942, + "step": 1839 + }, + { + "epoch": 0.043132497641191536, + "grad_norm": 4.054131507873535, + "learning_rate": 0.0001990928136943804, + "loss": 0.6311, + "step": 1840 + }, + { + "epoch": 0.04315593921599653, + "grad_norm": 3.871081590652466, + "learning_rate": 0.00019909182346689028, + "loss": 0.7031, + "step": 1841 + }, + { + "epoch": 0.043179380790801525, + "grad_norm": 4.580228805541992, + "learning_rate": 0.00019909083270172517, + "loss": 0.4825, + "step": 1842 + }, + { + "epoch": 0.04320282236560652, + "grad_norm": 2.1192593574523926, + "learning_rate": 0.00019908984139889037, + "loss": 0.3385, + "step": 1843 + }, + { + "epoch": 0.04322626394041151, + "grad_norm": 2.731684684753418, + "learning_rate": 0.00019908884955839134, + "loss": 0.39, + "step": 1844 + }, + { + "epoch": 0.043249705515216515, + "grad_norm": 1.0201590061187744, + "learning_rate": 0.00019908785718023345, + "loss": 0.2188, + "step": 1845 + }, + { + "epoch": 0.04327314709002151, + "grad_norm": 4.786778926849365, + "learning_rate": 0.000199086864264422, + "loss": 0.8917, + "step": 1846 + }, + { + "epoch": 0.0432965886648265, + "grad_norm": 5.116650104522705, + "learning_rate": 0.00019908587081096248, + "loss": 0.9837, + "step": 1847 + }, + { + "epoch": 0.0433200302396315, + "grad_norm": 3.1677308082580566, + "learning_rate": 0.00019908487681986024, + "loss": 0.726, + "step": 1848 + }, + { + "epoch": 0.04334347181443649, + "grad_norm": 2.257288694381714, + "learning_rate": 0.00019908388229112066, + "loss": 0.4288, + "step": 1849 + }, + { + "epoch": 0.043366913389241486, + "grad_norm": 2.32287859916687, + "learning_rate": 0.00019908288722474916, + "loss": 0.546, + "step": 1850 + }, + { + "epoch": 0.04339035496404649, + "grad_norm": 6.469823837280273, + "learning_rate": 0.0001990818916207511, + "loss": 0.5086, + "step": 1851 + }, + { + "epoch": 0.04341379653885148, + "grad_norm": 4.298024654388428, + "learning_rate": 0.00019908089547913192, + "loss": 0.6765, + "step": 1852 + }, + { + "epoch": 0.043437238113656476, + "grad_norm": 2.4842731952667236, + "learning_rate": 0.00019907989879989704, + "loss": 0.495, + "step": 1853 + }, + { + "epoch": 0.04346067968846147, + "grad_norm": 1.1330324411392212, + "learning_rate": 0.00019907890158305183, + "loss": 0.1968, + "step": 1854 + }, + { + "epoch": 0.043484121263266465, + "grad_norm": 1.0136501789093018, + "learning_rate": 0.0001990779038286017, + "loss": 0.0983, + "step": 1855 + }, + { + "epoch": 0.04350756283807146, + "grad_norm": 4.350482940673828, + "learning_rate": 0.0001990769055365521, + "loss": 1.0488, + "step": 1856 + }, + { + "epoch": 0.04353100441287646, + "grad_norm": 1.4702154397964478, + "learning_rate": 0.0001990759067069084, + "loss": 0.2646, + "step": 1857 + }, + { + "epoch": 0.043554445987681455, + "grad_norm": 2.9104223251342773, + "learning_rate": 0.00019907490733967603, + "loss": 0.5021, + "step": 1858 + }, + { + "epoch": 0.04357788756248645, + "grad_norm": 5.735979080200195, + "learning_rate": 0.00019907390743486049, + "loss": 1.8396, + "step": 1859 + }, + { + "epoch": 0.04360132913729144, + "grad_norm": 2.357088327407837, + "learning_rate": 0.00019907290699246711, + "loss": 0.3904, + "step": 1860 + }, + { + "epoch": 0.04362477071209644, + "grad_norm": 4.3068037033081055, + "learning_rate": 0.00019907190601250134, + "loss": 0.5641, + "step": 1861 + }, + { + "epoch": 0.04364821228690143, + "grad_norm": 2.3295750617980957, + "learning_rate": 0.00019907090449496862, + "loss": 1.3654, + "step": 1862 + }, + { + "epoch": 0.043671653861706426, + "grad_norm": 2.461453437805176, + "learning_rate": 0.0001990699024398744, + "loss": 0.3591, + "step": 1863 + }, + { + "epoch": 0.04369509543651143, + "grad_norm": 3.2223079204559326, + "learning_rate": 0.0001990688998472241, + "loss": 0.4311, + "step": 1864 + }, + { + "epoch": 0.04371853701131642, + "grad_norm": 0.8962258100509644, + "learning_rate": 0.00019906789671702316, + "loss": 0.126, + "step": 1865 + }, + { + "epoch": 0.043741978586121416, + "grad_norm": 4.155466556549072, + "learning_rate": 0.00019906689304927705, + "loss": 0.7165, + "step": 1866 + }, + { + "epoch": 0.04376542016092641, + "grad_norm": 1.325993299484253, + "learning_rate": 0.00019906588884399115, + "loss": 0.2492, + "step": 1867 + }, + { + "epoch": 0.043788861735731405, + "grad_norm": 1.731949806213379, + "learning_rate": 0.00019906488410117098, + "loss": 0.2554, + "step": 1868 + }, + { + "epoch": 0.0438123033105364, + "grad_norm": 1.8778094053268433, + "learning_rate": 0.00019906387882082197, + "loss": 1.2947, + "step": 1869 + }, + { + "epoch": 0.0438357448853414, + "grad_norm": 3.2936747074127197, + "learning_rate": 0.00019906287300294957, + "loss": 0.7184, + "step": 1870 + }, + { + "epoch": 0.043859186460146395, + "grad_norm": 1.0989623069763184, + "learning_rate": 0.00019906186664755922, + "loss": 0.2172, + "step": 1871 + }, + { + "epoch": 0.04388262803495139, + "grad_norm": 5.158141613006592, + "learning_rate": 0.0001990608597546564, + "loss": 1.0967, + "step": 1872 + }, + { + "epoch": 0.04390606960975638, + "grad_norm": 3.304462432861328, + "learning_rate": 0.00019905985232424655, + "loss": 0.5697, + "step": 1873 + }, + { + "epoch": 0.04392951118456138, + "grad_norm": 2.3966503143310547, + "learning_rate": 0.00019905884435633518, + "loss": 0.1992, + "step": 1874 + }, + { + "epoch": 0.04395295275936637, + "grad_norm": 3.825767993927002, + "learning_rate": 0.0001990578358509277, + "loss": 0.8885, + "step": 1875 + }, + { + "epoch": 0.043976394334171366, + "grad_norm": 0.5039419531822205, + "learning_rate": 0.00019905682680802967, + "loss": 0.0823, + "step": 1876 + }, + { + "epoch": 0.04399983590897637, + "grad_norm": 2.9795429706573486, + "learning_rate": 0.0001990558172276465, + "loss": 0.5399, + "step": 1877 + }, + { + "epoch": 0.04402327748378136, + "grad_norm": 1.4624241590499878, + "learning_rate": 0.00019905480710978365, + "loss": 0.3037, + "step": 1878 + }, + { + "epoch": 0.044046719058586356, + "grad_norm": 3.155073404312134, + "learning_rate": 0.00019905379645444667, + "loss": 0.5246, + "step": 1879 + }, + { + "epoch": 0.04407016063339135, + "grad_norm": 1.128584384918213, + "learning_rate": 0.00019905278526164097, + "loss": 0.7013, + "step": 1880 + }, + { + "epoch": 0.044093602208196345, + "grad_norm": 6.48174524307251, + "learning_rate": 0.0001990517735313721, + "loss": 0.8894, + "step": 1881 + }, + { + "epoch": 0.04411704378300134, + "grad_norm": 2.5859341621398926, + "learning_rate": 0.0001990507612636455, + "loss": 0.4254, + "step": 1882 + }, + { + "epoch": 0.04414048535780634, + "grad_norm": 3.9995672702789307, + "learning_rate": 0.00019904974845846672, + "loss": 0.5848, + "step": 1883 + }, + { + "epoch": 0.044163926932611335, + "grad_norm": 1.5115952491760254, + "learning_rate": 0.00019904873511584117, + "loss": 0.2175, + "step": 1884 + }, + { + "epoch": 0.04418736850741633, + "grad_norm": 2.02223539352417, + "learning_rate": 0.0001990477212357744, + "loss": 1.5466, + "step": 1885 + }, + { + "epoch": 0.04421081008222132, + "grad_norm": 2.1875271797180176, + "learning_rate": 0.00019904670681827197, + "loss": 0.2332, + "step": 1886 + }, + { + "epoch": 0.04423425165702632, + "grad_norm": 3.409292221069336, + "learning_rate": 0.0001990456918633393, + "loss": 0.5618, + "step": 1887 + }, + { + "epoch": 0.04425769323183131, + "grad_norm": 0.901576578617096, + "learning_rate": 0.0001990446763709819, + "loss": 0.1678, + "step": 1888 + }, + { + "epoch": 0.04428113480663631, + "grad_norm": 4.01007080078125, + "learning_rate": 0.0001990436603412053, + "loss": 0.7074, + "step": 1889 + }, + { + "epoch": 0.04430457638144131, + "grad_norm": 3.6463429927825928, + "learning_rate": 0.00019904264377401502, + "loss": 0.7437, + "step": 1890 + }, + { + "epoch": 0.0443280179562463, + "grad_norm": 3.081740617752075, + "learning_rate": 0.00019904162666941656, + "loss": 0.6019, + "step": 1891 + }, + { + "epoch": 0.044351459531051296, + "grad_norm": 3.2710726261138916, + "learning_rate": 0.00019904060902741549, + "loss": 0.5521, + "step": 1892 + }, + { + "epoch": 0.04437490110585629, + "grad_norm": 3.938819169998169, + "learning_rate": 0.00019903959084801724, + "loss": 0.6141, + "step": 1893 + }, + { + "epoch": 0.044398342680661285, + "grad_norm": 3.338876724243164, + "learning_rate": 0.00019903857213122742, + "loss": 0.4822, + "step": 1894 + }, + { + "epoch": 0.04442178425546628, + "grad_norm": 3.1500697135925293, + "learning_rate": 0.0001990375528770515, + "loss": 0.5472, + "step": 1895 + }, + { + "epoch": 0.04444522583027128, + "grad_norm": 2.3305907249450684, + "learning_rate": 0.00019903653308549505, + "loss": 1.4763, + "step": 1896 + }, + { + "epoch": 0.044468667405076275, + "grad_norm": 2.894728183746338, + "learning_rate": 0.0001990355127565636, + "loss": 0.57, + "step": 1897 + }, + { + "epoch": 0.04449210897988127, + "grad_norm": 2.236510753631592, + "learning_rate": 0.00019903449189026263, + "loss": 0.3137, + "step": 1898 + }, + { + "epoch": 0.04451555055468626, + "grad_norm": 3.387453317642212, + "learning_rate": 0.00019903347048659775, + "loss": 0.6213, + "step": 1899 + }, + { + "epoch": 0.04453899212949126, + "grad_norm": 3.389199733734131, + "learning_rate": 0.0001990324485455745, + "loss": 0.5607, + "step": 1900 + }, + { + "epoch": 0.04456243370429625, + "grad_norm": 2.194002628326416, + "learning_rate": 0.00019903142606719833, + "loss": 0.3309, + "step": 1901 + }, + { + "epoch": 0.04458587527910125, + "grad_norm": 0.8756288290023804, + "learning_rate": 0.00019903040305147492, + "loss": 0.0959, + "step": 1902 + }, + { + "epoch": 0.04460931685390625, + "grad_norm": 3.709526300430298, + "learning_rate": 0.00019902937949840973, + "loss": 0.7965, + "step": 1903 + }, + { + "epoch": 0.04463275842871124, + "grad_norm": 3.6321568489074707, + "learning_rate": 0.00019902835540800835, + "loss": 0.521, + "step": 1904 + }, + { + "epoch": 0.044656200003516236, + "grad_norm": 2.6276330947875977, + "learning_rate": 0.00019902733078027633, + "loss": 0.4528, + "step": 1905 + }, + { + "epoch": 0.04467964157832123, + "grad_norm": 2.015338182449341, + "learning_rate": 0.00019902630561521924, + "loss": 0.3403, + "step": 1906 + }, + { + "epoch": 0.044703083153126225, + "grad_norm": 4.737597942352295, + "learning_rate": 0.0001990252799128426, + "loss": 0.7267, + "step": 1907 + }, + { + "epoch": 0.044726524727931226, + "grad_norm": 4.252630710601807, + "learning_rate": 0.00019902425367315202, + "loss": 0.5586, + "step": 1908 + }, + { + "epoch": 0.04474996630273622, + "grad_norm": 6.165805816650391, + "learning_rate": 0.00019902322689615307, + "loss": 0.8453, + "step": 1909 + }, + { + "epoch": 0.044773407877541214, + "grad_norm": 0.47179868817329407, + "learning_rate": 0.0001990221995818513, + "loss": 0.0601, + "step": 1910 + }, + { + "epoch": 0.04479684945234621, + "grad_norm": 0.7898948192596436, + "learning_rate": 0.00019902117173025227, + "loss": 0.1081, + "step": 1911 + }, + { + "epoch": 0.0448202910271512, + "grad_norm": 3.134594202041626, + "learning_rate": 0.0001990201433413616, + "loss": 0.37, + "step": 1912 + }, + { + "epoch": 0.0448437326019562, + "grad_norm": 1.1738345623016357, + "learning_rate": 0.00019901911441518486, + "loss": 0.1447, + "step": 1913 + }, + { + "epoch": 0.04486717417676119, + "grad_norm": 1.8307650089263916, + "learning_rate": 0.0001990180849517276, + "loss": 0.3887, + "step": 1914 + }, + { + "epoch": 0.04489061575156619, + "grad_norm": 0.9038785099983215, + "learning_rate": 0.00019901705495099545, + "loss": 0.1057, + "step": 1915 + }, + { + "epoch": 0.04491405732637119, + "grad_norm": 3.467693328857422, + "learning_rate": 0.00019901602441299394, + "loss": 0.6219, + "step": 1916 + }, + { + "epoch": 0.04493749890117618, + "grad_norm": 3.865436553955078, + "learning_rate": 0.00019901499333772874, + "loss": 0.6489, + "step": 1917 + }, + { + "epoch": 0.044960940475981176, + "grad_norm": 1.284315586090088, + "learning_rate": 0.00019901396172520536, + "loss": 0.2653, + "step": 1918 + }, + { + "epoch": 0.04498438205078617, + "grad_norm": 1.8708713054656982, + "learning_rate": 0.00019901292957542945, + "loss": 1.7087, + "step": 1919 + }, + { + "epoch": 0.045007823625591165, + "grad_norm": 0.8018724322319031, + "learning_rate": 0.0001990118968884066, + "loss": 0.1672, + "step": 1920 + }, + { + "epoch": 0.045031265200396166, + "grad_norm": 3.6905570030212402, + "learning_rate": 0.00019901086366414243, + "loss": 0.7683, + "step": 1921 + }, + { + "epoch": 0.04505470677520116, + "grad_norm": 6.042735576629639, + "learning_rate": 0.00019900982990264253, + "loss": 1.0644, + "step": 1922 + }, + { + "epoch": 0.045078148350006154, + "grad_norm": 4.881829261779785, + "learning_rate": 0.00019900879560391248, + "loss": 0.953, + "step": 1923 + }, + { + "epoch": 0.04510158992481115, + "grad_norm": 6.7828874588012695, + "learning_rate": 0.00019900776076795795, + "loss": 0.9301, + "step": 1924 + }, + { + "epoch": 0.04512503149961614, + "grad_norm": 0.526999831199646, + "learning_rate": 0.00019900672539478452, + "loss": 0.0496, + "step": 1925 + }, + { + "epoch": 0.04514847307442114, + "grad_norm": 5.234800815582275, + "learning_rate": 0.0001990056894843978, + "loss": 1.0388, + "step": 1926 + }, + { + "epoch": 0.04517191464922613, + "grad_norm": 0.6824487447738647, + "learning_rate": 0.00019900465303680345, + "loss": 0.0958, + "step": 1927 + }, + { + "epoch": 0.04519535622403113, + "grad_norm": 3.6624536514282227, + "learning_rate": 0.00019900361605200705, + "loss": 0.7808, + "step": 1928 + }, + { + "epoch": 0.04521879779883613, + "grad_norm": 3.089244842529297, + "learning_rate": 0.00019900257853001425, + "loss": 0.7039, + "step": 1929 + }, + { + "epoch": 0.04524223937364112, + "grad_norm": 1.4312801361083984, + "learning_rate": 0.00019900154047083066, + "loss": 0.2534, + "step": 1930 + }, + { + "epoch": 0.045265680948446116, + "grad_norm": 2.815669059753418, + "learning_rate": 0.00019900050187446198, + "loss": 0.2936, + "step": 1931 + }, + { + "epoch": 0.04528912252325111, + "grad_norm": 2.9787778854370117, + "learning_rate": 0.00019899946274091375, + "loss": 0.3619, + "step": 1932 + }, + { + "epoch": 0.045312564098056105, + "grad_norm": 1.2067359685897827, + "learning_rate": 0.00019899842307019165, + "loss": 0.2142, + "step": 1933 + }, + { + "epoch": 0.045336005672861106, + "grad_norm": 2.9148573875427246, + "learning_rate": 0.00019899738286230134, + "loss": 0.3956, + "step": 1934 + }, + { + "epoch": 0.0453594472476661, + "grad_norm": 4.663976192474365, + "learning_rate": 0.00019899634211724843, + "loss": 0.4339, + "step": 1935 + }, + { + "epoch": 0.045382888822471094, + "grad_norm": 1.1214655637741089, + "learning_rate": 0.0001989953008350386, + "loss": 0.1133, + "step": 1936 + }, + { + "epoch": 0.04540633039727609, + "grad_norm": 1.045907735824585, + "learning_rate": 0.00019899425901567748, + "loss": 0.2604, + "step": 1937 + }, + { + "epoch": 0.04542977197208108, + "grad_norm": 2.068523645401001, + "learning_rate": 0.0001989932166591707, + "loss": 0.2214, + "step": 1938 + }, + { + "epoch": 0.04545321354688608, + "grad_norm": 3.7577145099639893, + "learning_rate": 0.00019899217376552396, + "loss": 0.9374, + "step": 1939 + }, + { + "epoch": 0.04547665512169108, + "grad_norm": 1.6368626356124878, + "learning_rate": 0.00019899113033474292, + "loss": 0.3906, + "step": 1940 + }, + { + "epoch": 0.04550009669649607, + "grad_norm": 3.6508450508117676, + "learning_rate": 0.00019899008636683322, + "loss": 0.4699, + "step": 1941 + }, + { + "epoch": 0.04552353827130107, + "grad_norm": 1.4158802032470703, + "learning_rate": 0.0001989890418618005, + "loss": 0.2754, + "step": 1942 + }, + { + "epoch": 0.04554697984610606, + "grad_norm": 2.874915599822998, + "learning_rate": 0.00019898799681965047, + "loss": 1.833, + "step": 1943 + }, + { + "epoch": 0.045570421420911056, + "grad_norm": 2.043616771697998, + "learning_rate": 0.0001989869512403888, + "loss": 0.1977, + "step": 1944 + }, + { + "epoch": 0.04559386299571605, + "grad_norm": 3.6019697189331055, + "learning_rate": 0.00019898590512402114, + "loss": 0.5779, + "step": 1945 + }, + { + "epoch": 0.045617304570521044, + "grad_norm": 3.4524552822113037, + "learning_rate": 0.0001989848584705532, + "loss": 0.5857, + "step": 1946 + }, + { + "epoch": 0.045640746145326046, + "grad_norm": 0.9813228845596313, + "learning_rate": 0.0001989838112799906, + "loss": 0.1549, + "step": 1947 + }, + { + "epoch": 0.04566418772013104, + "grad_norm": 1.0567741394042969, + "learning_rate": 0.00019898276355233906, + "loss": 0.1577, + "step": 1948 + }, + { + "epoch": 0.045687629294936034, + "grad_norm": 5.820920467376709, + "learning_rate": 0.00019898171528760425, + "loss": 0.8192, + "step": 1949 + }, + { + "epoch": 0.04571107086974103, + "grad_norm": 2.671813726425171, + "learning_rate": 0.00019898066648579193, + "loss": 1.6745, + "step": 1950 + }, + { + "epoch": 0.04573451244454602, + "grad_norm": 3.159522294998169, + "learning_rate": 0.00019897961714690767, + "loss": 0.3877, + "step": 1951 + }, + { + "epoch": 0.04575795401935102, + "grad_norm": 1.668663740158081, + "learning_rate": 0.00019897856727095725, + "loss": 1.2664, + "step": 1952 + }, + { + "epoch": 0.04578139559415602, + "grad_norm": 1.3632017374038696, + "learning_rate": 0.0001989775168579463, + "loss": 0.135, + "step": 1953 + }, + { + "epoch": 0.04580483716896101, + "grad_norm": 4.332801342010498, + "learning_rate": 0.00019897646590788062, + "loss": 0.8339, + "step": 1954 + }, + { + "epoch": 0.04582827874376601, + "grad_norm": 4.982700347900391, + "learning_rate": 0.00019897541442076585, + "loss": 1.0218, + "step": 1955 + }, + { + "epoch": 0.045851720318571, + "grad_norm": 6.307053089141846, + "learning_rate": 0.00019897436239660766, + "loss": 0.7275, + "step": 1956 + }, + { + "epoch": 0.045875161893375996, + "grad_norm": 4.37985372543335, + "learning_rate": 0.0001989733098354118, + "loss": 0.8177, + "step": 1957 + }, + { + "epoch": 0.04589860346818099, + "grad_norm": 0.5912351012229919, + "learning_rate": 0.000198972256737184, + "loss": 0.0944, + "step": 1958 + }, + { + "epoch": 0.045922045042985984, + "grad_norm": 2.9566574096679688, + "learning_rate": 0.00019897120310192996, + "loss": 0.668, + "step": 1959 + }, + { + "epoch": 0.045945486617790986, + "grad_norm": 1.2491806745529175, + "learning_rate": 0.00019897014892965537, + "loss": 0.2099, + "step": 1960 + }, + { + "epoch": 0.04596892819259598, + "grad_norm": 1.8690234422683716, + "learning_rate": 0.00019896909422036595, + "loss": 0.2858, + "step": 1961 + }, + { + "epoch": 0.045992369767400974, + "grad_norm": 4.2191691398620605, + "learning_rate": 0.0001989680389740675, + "loss": 0.6668, + "step": 1962 + }, + { + "epoch": 0.04601581134220597, + "grad_norm": 4.2443647384643555, + "learning_rate": 0.00019896698319076564, + "loss": 0.9107, + "step": 1963 + }, + { + "epoch": 0.04603925291701096, + "grad_norm": 3.6453044414520264, + "learning_rate": 0.00019896592687046612, + "loss": 0.6671, + "step": 1964 + }, + { + "epoch": 0.04606269449181596, + "grad_norm": 1.2105393409729004, + "learning_rate": 0.00019896487001317477, + "loss": 0.1937, + "step": 1965 + }, + { + "epoch": 0.04608613606662096, + "grad_norm": 5.605123043060303, + "learning_rate": 0.0001989638126188972, + "loss": 0.8991, + "step": 1966 + }, + { + "epoch": 0.04610957764142595, + "grad_norm": 4.523392200469971, + "learning_rate": 0.00019896275468763924, + "loss": 1.156, + "step": 1967 + }, + { + "epoch": 0.04613301921623095, + "grad_norm": 3.5589911937713623, + "learning_rate": 0.00019896169621940656, + "loss": 0.373, + "step": 1968 + }, + { + "epoch": 0.04615646079103594, + "grad_norm": 5.862678527832031, + "learning_rate": 0.00019896063721420494, + "loss": 0.7661, + "step": 1969 + }, + { + "epoch": 0.046179902365840936, + "grad_norm": 2.655576467514038, + "learning_rate": 0.00019895957767204013, + "loss": 0.4335, + "step": 1970 + }, + { + "epoch": 0.04620334394064593, + "grad_norm": 1.2134720087051392, + "learning_rate": 0.00019895851759291783, + "loss": 0.2029, + "step": 1971 + }, + { + "epoch": 0.04622678551545093, + "grad_norm": 0.6220419406890869, + "learning_rate": 0.00019895745697684388, + "loss": 0.1397, + "step": 1972 + }, + { + "epoch": 0.046250227090255926, + "grad_norm": 4.169680118560791, + "learning_rate": 0.00019895639582382396, + "loss": 0.7678, + "step": 1973 + }, + { + "epoch": 0.04627366866506092, + "grad_norm": 0.7255352735519409, + "learning_rate": 0.00019895533413386387, + "loss": 0.1365, + "step": 1974 + }, + { + "epoch": 0.046297110239865914, + "grad_norm": 1.6660118103027344, + "learning_rate": 0.00019895427190696935, + "loss": 0.3278, + "step": 1975 + }, + { + "epoch": 0.04632055181467091, + "grad_norm": 4.138119220733643, + "learning_rate": 0.00019895320914314617, + "loss": 0.5947, + "step": 1976 + }, + { + "epoch": 0.0463439933894759, + "grad_norm": 1.25360107421875, + "learning_rate": 0.0001989521458424001, + "loss": 0.2393, + "step": 1977 + }, + { + "epoch": 0.0463674349642809, + "grad_norm": 2.272562026977539, + "learning_rate": 0.00019895108200473688, + "loss": 0.5302, + "step": 1978 + }, + { + "epoch": 0.0463908765390859, + "grad_norm": 1.6861728429794312, + "learning_rate": 0.00019895001763016232, + "loss": 0.3677, + "step": 1979 + }, + { + "epoch": 0.04641431811389089, + "grad_norm": 2.9357001781463623, + "learning_rate": 0.0001989489527186822, + "loss": 0.4757, + "step": 1980 + }, + { + "epoch": 0.04643775968869589, + "grad_norm": 0.7304463386535645, + "learning_rate": 0.00019894788727030225, + "loss": 0.0986, + "step": 1981 + }, + { + "epoch": 0.04646120126350088, + "grad_norm": 0.5834060311317444, + "learning_rate": 0.0001989468212850283, + "loss": 0.0641, + "step": 1982 + }, + { + "epoch": 0.046484642838305876, + "grad_norm": 4.86930513381958, + "learning_rate": 0.00019894575476286613, + "loss": 0.6746, + "step": 1983 + }, + { + "epoch": 0.04650808441311087, + "grad_norm": 1.292555332183838, + "learning_rate": 0.00019894468770382146, + "loss": 0.2734, + "step": 1984 + }, + { + "epoch": 0.04653152598791587, + "grad_norm": 1.8444054126739502, + "learning_rate": 0.0001989436201079002, + "loss": 0.4412, + "step": 1985 + }, + { + "epoch": 0.046554967562720866, + "grad_norm": 4.279559135437012, + "learning_rate": 0.00019894255197510803, + "loss": 0.6864, + "step": 1986 + }, + { + "epoch": 0.04657840913752586, + "grad_norm": 2.109689474105835, + "learning_rate": 0.00019894148330545081, + "loss": 0.3036, + "step": 1987 + }, + { + "epoch": 0.046601850712330854, + "grad_norm": 2.260115146636963, + "learning_rate": 0.0001989404140989343, + "loss": 0.2194, + "step": 1988 + }, + { + "epoch": 0.04662529228713585, + "grad_norm": 2.740412473678589, + "learning_rate": 0.00019893934435556433, + "loss": 0.4975, + "step": 1989 + }, + { + "epoch": 0.04664873386194084, + "grad_norm": 4.0555291175842285, + "learning_rate": 0.0001989382740753467, + "loss": 0.8917, + "step": 1990 + }, + { + "epoch": 0.04667217543674584, + "grad_norm": 4.353993892669678, + "learning_rate": 0.00019893720325828723, + "loss": 0.7532, + "step": 1991 + }, + { + "epoch": 0.04669561701155084, + "grad_norm": 1.4959443807601929, + "learning_rate": 0.0001989361319043917, + "loss": 1.1197, + "step": 1992 + }, + { + "epoch": 0.04671905858635583, + "grad_norm": 4.513528823852539, + "learning_rate": 0.0001989350600136659, + "loss": 0.7527, + "step": 1993 + }, + { + "epoch": 0.04674250016116083, + "grad_norm": 3.149977684020996, + "learning_rate": 0.0001989339875861157, + "loss": 0.6642, + "step": 1994 + }, + { + "epoch": 0.04676594173596582, + "grad_norm": 2.4206395149230957, + "learning_rate": 0.00019893291462174694, + "loss": 0.4507, + "step": 1995 + }, + { + "epoch": 0.046789383310770816, + "grad_norm": 3.318343162536621, + "learning_rate": 0.0001989318411205654, + "loss": 0.4209, + "step": 1996 + }, + { + "epoch": 0.04681282488557581, + "grad_norm": 2.328277349472046, + "learning_rate": 0.00019893076708257687, + "loss": 1.6248, + "step": 1997 + }, + { + "epoch": 0.04683626646038081, + "grad_norm": 2.45479154586792, + "learning_rate": 0.00019892969250778723, + "loss": 1.9211, + "step": 1998 + }, + { + "epoch": 0.046859708035185806, + "grad_norm": 1.7095303535461426, + "learning_rate": 0.0001989286173962023, + "loss": 0.2366, + "step": 1999 + }, + { + "epoch": 0.0468831496099908, + "grad_norm": 2.5496957302093506, + "learning_rate": 0.00019892754174782791, + "loss": 1.9055, + "step": 2000 + }, + { + "epoch": 0.046906591184795794, + "grad_norm": 5.1464362144470215, + "learning_rate": 0.00019892646556266992, + "loss": 0.6675, + "step": 2001 + }, + { + "epoch": 0.04693003275960079, + "grad_norm": 2.9941537380218506, + "learning_rate": 0.0001989253888407341, + "loss": 0.4989, + "step": 2002 + }, + { + "epoch": 0.04695347433440578, + "grad_norm": 2.131316661834717, + "learning_rate": 0.00019892431158202638, + "loss": 0.2856, + "step": 2003 + }, + { + "epoch": 0.046976915909210784, + "grad_norm": 1.4928812980651855, + "learning_rate": 0.00019892323378655253, + "loss": 0.3373, + "step": 2004 + }, + { + "epoch": 0.04700035748401578, + "grad_norm": 5.604949474334717, + "learning_rate": 0.00019892215545431845, + "loss": 0.9488, + "step": 2005 + }, + { + "epoch": 0.04702379905882077, + "grad_norm": 2.6205313205718994, + "learning_rate": 0.00019892107658533, + "loss": 0.4126, + "step": 2006 + }, + { + "epoch": 0.04704724063362577, + "grad_norm": 3.136439085006714, + "learning_rate": 0.00019891999717959295, + "loss": 0.5075, + "step": 2007 + }, + { + "epoch": 0.04707068220843076, + "grad_norm": 2.948268413543701, + "learning_rate": 0.00019891891723711327, + "loss": 0.6574, + "step": 2008 + }, + { + "epoch": 0.047094123783235756, + "grad_norm": 3.452359676361084, + "learning_rate": 0.00019891783675789673, + "loss": 0.6238, + "step": 2009 + }, + { + "epoch": 0.04711756535804075, + "grad_norm": 0.9858332276344299, + "learning_rate": 0.0001989167557419492, + "loss": 0.2563, + "step": 2010 + }, + { + "epoch": 0.04714100693284575, + "grad_norm": 4.619190692901611, + "learning_rate": 0.00019891567418927665, + "loss": 0.7949, + "step": 2011 + }, + { + "epoch": 0.047164448507650746, + "grad_norm": 2.620950222015381, + "learning_rate": 0.00019891459209988481, + "loss": 0.4137, + "step": 2012 + }, + { + "epoch": 0.04718789008245574, + "grad_norm": 1.0469448566436768, + "learning_rate": 0.0001989135094737796, + "loss": 0.2056, + "step": 2013 + }, + { + "epoch": 0.047211331657260734, + "grad_norm": 2.4034337997436523, + "learning_rate": 0.00019891242631096695, + "loss": 0.5915, + "step": 2014 + }, + { + "epoch": 0.04723477323206573, + "grad_norm": 6.411406993865967, + "learning_rate": 0.00019891134261145268, + "loss": 1.0111, + "step": 2015 + }, + { + "epoch": 0.04725821480687072, + "grad_norm": 4.5980916023254395, + "learning_rate": 0.00019891025837524267, + "loss": 0.8321, + "step": 2016 + }, + { + "epoch": 0.047281656381675724, + "grad_norm": 3.8257994651794434, + "learning_rate": 0.00019890917360234283, + "loss": 0.783, + "step": 2017 + }, + { + "epoch": 0.04730509795648072, + "grad_norm": 2.5414512157440186, + "learning_rate": 0.00019890808829275904, + "loss": 0.4058, + "step": 2018 + }, + { + "epoch": 0.04732853953128571, + "grad_norm": 3.3008882999420166, + "learning_rate": 0.00019890700244649717, + "loss": 0.547, + "step": 2019 + }, + { + "epoch": 0.04735198110609071, + "grad_norm": 3.9661014080047607, + "learning_rate": 0.00019890591606356312, + "loss": 0.4791, + "step": 2020 + }, + { + "epoch": 0.0473754226808957, + "grad_norm": 1.2006540298461914, + "learning_rate": 0.0001989048291439628, + "loss": 0.2172, + "step": 2021 + }, + { + "epoch": 0.047398864255700696, + "grad_norm": 0.9711479544639587, + "learning_rate": 0.0001989037416877021, + "loss": 0.1267, + "step": 2022 + }, + { + "epoch": 0.04742230583050569, + "grad_norm": 2.8256497383117676, + "learning_rate": 0.0001989026536947869, + "loss": 0.4334, + "step": 2023 + }, + { + "epoch": 0.04744574740531069, + "grad_norm": 4.083033084869385, + "learning_rate": 0.00019890156516522313, + "loss": 0.5829, + "step": 2024 + }, + { + "epoch": 0.047469188980115685, + "grad_norm": 2.862104654312134, + "learning_rate": 0.0001989004760990167, + "loss": 0.4383, + "step": 2025 + }, + { + "epoch": 0.04749263055492068, + "grad_norm": 0.641736626625061, + "learning_rate": 0.00019889938649617347, + "loss": 0.1129, + "step": 2026 + }, + { + "epoch": 0.047516072129725674, + "grad_norm": 3.551321029663086, + "learning_rate": 0.00019889829635669944, + "loss": 0.6869, + "step": 2027 + }, + { + "epoch": 0.04753951370453067, + "grad_norm": 5.52982759475708, + "learning_rate": 0.00019889720568060042, + "loss": 1.0079, + "step": 2028 + }, + { + "epoch": 0.04756295527933566, + "grad_norm": 3.9553635120391846, + "learning_rate": 0.00019889611446788239, + "loss": 0.7148, + "step": 2029 + }, + { + "epoch": 0.047586396854140664, + "grad_norm": 4.536308288574219, + "learning_rate": 0.00019889502271855128, + "loss": 0.688, + "step": 2030 + }, + { + "epoch": 0.04760983842894566, + "grad_norm": 5.682798385620117, + "learning_rate": 0.000198893930432613, + "loss": 0.8158, + "step": 2031 + }, + { + "epoch": 0.04763328000375065, + "grad_norm": 4.526801109313965, + "learning_rate": 0.00019889283761007347, + "loss": 1.063, + "step": 2032 + }, + { + "epoch": 0.04765672157855565, + "grad_norm": 3.2609035968780518, + "learning_rate": 0.0001988917442509386, + "loss": 0.3193, + "step": 2033 + }, + { + "epoch": 0.04768016315336064, + "grad_norm": 4.455193996429443, + "learning_rate": 0.00019889065035521436, + "loss": 0.7071, + "step": 2034 + }, + { + "epoch": 0.047703604728165636, + "grad_norm": 2.9728457927703857, + "learning_rate": 0.0001988895559229067, + "loss": 0.6808, + "step": 2035 + }, + { + "epoch": 0.04772704630297064, + "grad_norm": 3.138340950012207, + "learning_rate": 0.0001988884609540215, + "loss": 0.6074, + "step": 2036 + }, + { + "epoch": 0.04775048787777563, + "grad_norm": 0.8765249848365784, + "learning_rate": 0.00019888736544856472, + "loss": 0.1256, + "step": 2037 + }, + { + "epoch": 0.047773929452580625, + "grad_norm": 2.73201847076416, + "learning_rate": 0.0001988862694065423, + "loss": 0.5248, + "step": 2038 + }, + { + "epoch": 0.04779737102738562, + "grad_norm": 0.7424970865249634, + "learning_rate": 0.00019888517282796024, + "loss": 0.1431, + "step": 2039 + }, + { + "epoch": 0.047820812602190614, + "grad_norm": 1.328683614730835, + "learning_rate": 0.00019888407571282447, + "loss": 0.2995, + "step": 2040 + }, + { + "epoch": 0.04784425417699561, + "grad_norm": 2.6997616291046143, + "learning_rate": 0.0001988829780611409, + "loss": 0.3673, + "step": 2041 + }, + { + "epoch": 0.0478676957518006, + "grad_norm": 2.948439359664917, + "learning_rate": 0.00019888187987291548, + "loss": 0.6457, + "step": 2042 + }, + { + "epoch": 0.047891137326605604, + "grad_norm": 4.480093955993652, + "learning_rate": 0.00019888078114815424, + "loss": 0.6875, + "step": 2043 + }, + { + "epoch": 0.0479145789014106, + "grad_norm": 2.4721639156341553, + "learning_rate": 0.0001988796818868631, + "loss": 0.453, + "step": 2044 + }, + { + "epoch": 0.04793802047621559, + "grad_norm": 4.1926727294921875, + "learning_rate": 0.00019887858208904796, + "loss": 0.9383, + "step": 2045 + }, + { + "epoch": 0.04796146205102059, + "grad_norm": 2.18367600440979, + "learning_rate": 0.00019887748175471495, + "loss": 1.9029, + "step": 2046 + }, + { + "epoch": 0.04798490362582558, + "grad_norm": 3.685011148452759, + "learning_rate": 0.0001988763808838699, + "loss": 0.642, + "step": 2047 + }, + { + "epoch": 0.048008345200630576, + "grad_norm": 2.4710280895233154, + "learning_rate": 0.0001988752794765188, + "loss": 0.2927, + "step": 2048 + }, + { + "epoch": 0.04803178677543558, + "grad_norm": 2.626873254776001, + "learning_rate": 0.0001988741775326677, + "loss": 0.4284, + "step": 2049 + }, + { + "epoch": 0.04805522835024057, + "grad_norm": 2.2888717651367188, + "learning_rate": 0.00019887307505232248, + "loss": 0.4167, + "step": 2050 + }, + { + "epoch": 0.048078669925045565, + "grad_norm": 2.767179012298584, + "learning_rate": 0.00019887197203548925, + "loss": 1.4864, + "step": 2051 + }, + { + "epoch": 0.04810211149985056, + "grad_norm": 2.163522481918335, + "learning_rate": 0.00019887086848217384, + "loss": 2.1529, + "step": 2052 + }, + { + "epoch": 0.048125553074655554, + "grad_norm": 3.785661220550537, + "learning_rate": 0.00019886976439238235, + "loss": 0.6191, + "step": 2053 + }, + { + "epoch": 0.04814899464946055, + "grad_norm": 4.410844802856445, + "learning_rate": 0.00019886865976612077, + "loss": 0.6433, + "step": 2054 + }, + { + "epoch": 0.04817243622426555, + "grad_norm": 3.5481438636779785, + "learning_rate": 0.00019886755460339503, + "loss": 0.4041, + "step": 2055 + }, + { + "epoch": 0.048195877799070544, + "grad_norm": 1.6306651830673218, + "learning_rate": 0.00019886644890421116, + "loss": 0.1843, + "step": 2056 + }, + { + "epoch": 0.04821931937387554, + "grad_norm": 3.349884033203125, + "learning_rate": 0.00019886534266857516, + "loss": 0.6063, + "step": 2057 + }, + { + "epoch": 0.04824276094868053, + "grad_norm": 1.8745616674423218, + "learning_rate": 0.00019886423589649303, + "loss": 1.2432, + "step": 2058 + }, + { + "epoch": 0.04826620252348553, + "grad_norm": 5.258128643035889, + "learning_rate": 0.00019886312858797078, + "loss": 0.6992, + "step": 2059 + }, + { + "epoch": 0.04828964409829052, + "grad_norm": 2.400158643722534, + "learning_rate": 0.00019886202074301444, + "loss": 1.6581, + "step": 2060 + }, + { + "epoch": 0.048313085673095515, + "grad_norm": 4.702617645263672, + "learning_rate": 0.00019886091236162993, + "loss": 0.4424, + "step": 2061 + }, + { + "epoch": 0.04833652724790052, + "grad_norm": 0.7249739170074463, + "learning_rate": 0.00019885980344382337, + "loss": 0.1159, + "step": 2062 + }, + { + "epoch": 0.04835996882270551, + "grad_norm": 0.9436466693878174, + "learning_rate": 0.00019885869398960076, + "loss": 0.8454, + "step": 2063 + }, + { + "epoch": 0.048383410397510505, + "grad_norm": 6.309682846069336, + "learning_rate": 0.00019885758399896806, + "loss": 0.9465, + "step": 2064 + }, + { + "epoch": 0.0484068519723155, + "grad_norm": 0.5063731670379639, + "learning_rate": 0.00019885647347193136, + "loss": 0.1151, + "step": 2065 + }, + { + "epoch": 0.048430293547120494, + "grad_norm": 4.96437931060791, + "learning_rate": 0.00019885536240849664, + "loss": 0.7225, + "step": 2066 + }, + { + "epoch": 0.04845373512192549, + "grad_norm": 4.252410888671875, + "learning_rate": 0.00019885425080866994, + "loss": 0.4421, + "step": 2067 + }, + { + "epoch": 0.04847717669673049, + "grad_norm": 2.2633659839630127, + "learning_rate": 0.0001988531386724573, + "loss": 1.5449, + "step": 2068 + }, + { + "epoch": 0.048500618271535484, + "grad_norm": 1.477129578590393, + "learning_rate": 0.00019885202599986474, + "loss": 0.2168, + "step": 2069 + }, + { + "epoch": 0.04852405984634048, + "grad_norm": 2.895034074783325, + "learning_rate": 0.00019885091279089832, + "loss": 1.2908, + "step": 2070 + }, + { + "epoch": 0.04854750142114547, + "grad_norm": 3.663041830062866, + "learning_rate": 0.00019884979904556406, + "loss": 0.927, + "step": 2071 + }, + { + "epoch": 0.04857094299595047, + "grad_norm": 2.435579299926758, + "learning_rate": 0.000198848684763868, + "loss": 0.4022, + "step": 2072 + }, + { + "epoch": 0.04859438457075546, + "grad_norm": 1.2436853647232056, + "learning_rate": 0.00019884756994581623, + "loss": 0.1107, + "step": 2073 + }, + { + "epoch": 0.048617826145560455, + "grad_norm": 3.0909664630889893, + "learning_rate": 0.00019884645459141472, + "loss": 0.5081, + "step": 2074 + }, + { + "epoch": 0.04864126772036546, + "grad_norm": 4.098504066467285, + "learning_rate": 0.0001988453387006696, + "loss": 0.597, + "step": 2075 + }, + { + "epoch": 0.04866470929517045, + "grad_norm": 1.4181891679763794, + "learning_rate": 0.0001988442222735869, + "loss": 0.2959, + "step": 2076 + }, + { + "epoch": 0.048688150869975445, + "grad_norm": 1.6542125940322876, + "learning_rate": 0.00019884310531017264, + "loss": 0.2772, + "step": 2077 + }, + { + "epoch": 0.04871159244478044, + "grad_norm": 3.1861062049865723, + "learning_rate": 0.00019884198781043293, + "loss": 0.3841, + "step": 2078 + }, + { + "epoch": 0.048735034019585434, + "grad_norm": 1.287946343421936, + "learning_rate": 0.00019884086977437382, + "loss": 0.2768, + "step": 2079 + }, + { + "epoch": 0.04875847559439043, + "grad_norm": 3.3445653915405273, + "learning_rate": 0.00019883975120200137, + "loss": 0.5891, + "step": 2080 + }, + { + "epoch": 0.04878191716919543, + "grad_norm": 0.99541175365448, + "learning_rate": 0.00019883863209332165, + "loss": 0.2835, + "step": 2081 + }, + { + "epoch": 0.048805358744000424, + "grad_norm": 2.7013442516326904, + "learning_rate": 0.00019883751244834073, + "loss": 0.447, + "step": 2082 + }, + { + "epoch": 0.04882880031880542, + "grad_norm": 2.514995813369751, + "learning_rate": 0.00019883639226706467, + "loss": 1.6223, + "step": 2083 + }, + { + "epoch": 0.04885224189361041, + "grad_norm": 3.9279255867004395, + "learning_rate": 0.00019883527154949962, + "loss": 0.7088, + "step": 2084 + }, + { + "epoch": 0.04887568346841541, + "grad_norm": 3.455333948135376, + "learning_rate": 0.00019883415029565155, + "loss": 0.6106, + "step": 2085 + }, + { + "epoch": 0.0488991250432204, + "grad_norm": 3.1652982234954834, + "learning_rate": 0.00019883302850552662, + "loss": 0.7307, + "step": 2086 + }, + { + "epoch": 0.0489225666180254, + "grad_norm": 4.807657718658447, + "learning_rate": 0.00019883190617913093, + "loss": 0.5895, + "step": 2087 + }, + { + "epoch": 0.0489460081928304, + "grad_norm": 3.9949536323547363, + "learning_rate": 0.0001988307833164705, + "loss": 0.6613, + "step": 2088 + }, + { + "epoch": 0.04896944976763539, + "grad_norm": 1.6945747137069702, + "learning_rate": 0.00019882965991755146, + "loss": 0.306, + "step": 2089 + }, + { + "epoch": 0.048992891342440385, + "grad_norm": 1.449420690536499, + "learning_rate": 0.0001988285359823799, + "loss": 0.9912, + "step": 2090 + }, + { + "epoch": 0.04901633291724538, + "grad_norm": 2.4208738803863525, + "learning_rate": 0.00019882741151096197, + "loss": 0.5, + "step": 2091 + }, + { + "epoch": 0.049039774492050374, + "grad_norm": 2.774829626083374, + "learning_rate": 0.0001988262865033037, + "loss": 0.4365, + "step": 2092 + }, + { + "epoch": 0.04906321606685537, + "grad_norm": 1.770066261291504, + "learning_rate": 0.0001988251609594112, + "loss": 0.3451, + "step": 2093 + }, + { + "epoch": 0.04908665764166037, + "grad_norm": 2.8972971439361572, + "learning_rate": 0.00019882403487929065, + "loss": 0.5726, + "step": 2094 + }, + { + "epoch": 0.049110099216465364, + "grad_norm": 3.306472063064575, + "learning_rate": 0.00019882290826294806, + "loss": 0.6007, + "step": 2095 + }, + { + "epoch": 0.04913354079127036, + "grad_norm": 3.4186010360717773, + "learning_rate": 0.0001988217811103896, + "loss": 0.7142, + "step": 2096 + }, + { + "epoch": 0.04915698236607535, + "grad_norm": 3.0693511962890625, + "learning_rate": 0.0001988206534216214, + "loss": 0.4741, + "step": 2097 + }, + { + "epoch": 0.04918042394088035, + "grad_norm": 4.438840389251709, + "learning_rate": 0.00019881952519664955, + "loss": 0.8512, + "step": 2098 + }, + { + "epoch": 0.04920386551568534, + "grad_norm": 3.600585699081421, + "learning_rate": 0.00019881839643548014, + "loss": 0.3935, + "step": 2099 + }, + { + "epoch": 0.04922730709049034, + "grad_norm": 3.5027670860290527, + "learning_rate": 0.00019881726713811936, + "loss": 0.5732, + "step": 2100 + }, + { + "epoch": 0.04925074866529534, + "grad_norm": 2.0314290523529053, + "learning_rate": 0.00019881613730457333, + "loss": 1.4948, + "step": 2101 + }, + { + "epoch": 0.04927419024010033, + "grad_norm": 5.484715938568115, + "learning_rate": 0.00019881500693484815, + "loss": 0.6203, + "step": 2102 + }, + { + "epoch": 0.049297631814905325, + "grad_norm": 4.391053676605225, + "learning_rate": 0.00019881387602894995, + "loss": 0.7043, + "step": 2103 + }, + { + "epoch": 0.04932107338971032, + "grad_norm": 1.698404312133789, + "learning_rate": 0.00019881274458688486, + "loss": 0.3024, + "step": 2104 + }, + { + "epoch": 0.049344514964515314, + "grad_norm": 3.129338026046753, + "learning_rate": 0.00019881161260865908, + "loss": 0.4417, + "step": 2105 + }, + { + "epoch": 0.04936795653932031, + "grad_norm": 2.228116035461426, + "learning_rate": 0.0001988104800942787, + "loss": 0.2843, + "step": 2106 + }, + { + "epoch": 0.04939139811412531, + "grad_norm": 4.283266067504883, + "learning_rate": 0.00019880934704374984, + "loss": 0.4091, + "step": 2107 + }, + { + "epoch": 0.049414839688930304, + "grad_norm": 2.4721879959106445, + "learning_rate": 0.00019880821345707875, + "loss": 0.538, + "step": 2108 + }, + { + "epoch": 0.0494382812637353, + "grad_norm": 4.34276819229126, + "learning_rate": 0.00019880707933427147, + "loss": 0.8661, + "step": 2109 + }, + { + "epoch": 0.04946172283854029, + "grad_norm": 5.659925937652588, + "learning_rate": 0.00019880594467533422, + "loss": 0.8085, + "step": 2110 + }, + { + "epoch": 0.04948516441334529, + "grad_norm": 1.7661466598510742, + "learning_rate": 0.0001988048094802731, + "loss": 1.2025, + "step": 2111 + }, + { + "epoch": 0.04950860598815028, + "grad_norm": 4.080132007598877, + "learning_rate": 0.00019880367374909436, + "loss": 0.5225, + "step": 2112 + }, + { + "epoch": 0.04953204756295528, + "grad_norm": 2.9928104877471924, + "learning_rate": 0.00019880253748180408, + "loss": 0.4025, + "step": 2113 + }, + { + "epoch": 0.04955548913776028, + "grad_norm": 5.416073322296143, + "learning_rate": 0.00019880140067840847, + "loss": 0.9347, + "step": 2114 + }, + { + "epoch": 0.04957893071256527, + "grad_norm": 4.66161584854126, + "learning_rate": 0.00019880026333891367, + "loss": 0.8585, + "step": 2115 + }, + { + "epoch": 0.049602372287370265, + "grad_norm": 2.4515140056610107, + "learning_rate": 0.00019879912546332586, + "loss": 0.5238, + "step": 2116 + }, + { + "epoch": 0.04962581386217526, + "grad_norm": 2.061204671859741, + "learning_rate": 0.0001987979870516512, + "loss": 0.4427, + "step": 2117 + }, + { + "epoch": 0.049649255436980254, + "grad_norm": 4.64752721786499, + "learning_rate": 0.0001987968481038959, + "loss": 0.8156, + "step": 2118 + }, + { + "epoch": 0.049672697011785255, + "grad_norm": 6.201916694641113, + "learning_rate": 0.00019879570862006615, + "loss": 0.7982, + "step": 2119 + }, + { + "epoch": 0.04969613858659025, + "grad_norm": 2.576995849609375, + "learning_rate": 0.00019879456860016808, + "loss": 2.4432, + "step": 2120 + }, + { + "epoch": 0.049719580161395244, + "grad_norm": 3.2885396480560303, + "learning_rate": 0.0001987934280442079, + "loss": 0.4837, + "step": 2121 + }, + { + "epoch": 0.04974302173620024, + "grad_norm": 0.7217739820480347, + "learning_rate": 0.00019879228695219182, + "loss": 0.1477, + "step": 2122 + }, + { + "epoch": 0.04976646331100523, + "grad_norm": 1.0047063827514648, + "learning_rate": 0.00019879114532412597, + "loss": 0.1254, + "step": 2123 + }, + { + "epoch": 0.04978990488581023, + "grad_norm": 0.37407562136650085, + "learning_rate": 0.00019879000316001664, + "loss": 0.0655, + "step": 2124 + }, + { + "epoch": 0.04981334646061522, + "grad_norm": 3.037062644958496, + "learning_rate": 0.00019878886045986995, + "loss": 0.4699, + "step": 2125 + }, + { + "epoch": 0.04983678803542022, + "grad_norm": 0.7832814455032349, + "learning_rate": 0.00019878771722369213, + "loss": 0.164, + "step": 2126 + }, + { + "epoch": 0.04986022961022522, + "grad_norm": 4.524951457977295, + "learning_rate": 0.00019878657345148937, + "loss": 1.1879, + "step": 2127 + }, + { + "epoch": 0.04988367118503021, + "grad_norm": 3.853675603866577, + "learning_rate": 0.0001987854291432679, + "loss": 0.744, + "step": 2128 + }, + { + "epoch": 0.049907112759835205, + "grad_norm": 3.0377328395843506, + "learning_rate": 0.0001987842842990339, + "loss": 0.3801, + "step": 2129 + }, + { + "epoch": 0.0499305543346402, + "grad_norm": 3.570908546447754, + "learning_rate": 0.0001987831389187936, + "loss": 0.4294, + "step": 2130 + }, + { + "epoch": 0.049953995909445194, + "grad_norm": 3.6620419025421143, + "learning_rate": 0.0001987819930025532, + "loss": 0.5752, + "step": 2131 + }, + { + "epoch": 0.049977437484250195, + "grad_norm": 0.4357900619506836, + "learning_rate": 0.00019878084655031896, + "loss": 0.0641, + "step": 2132 + }, + { + "epoch": 0.05000087905905519, + "grad_norm": 3.498227119445801, + "learning_rate": 0.00019877969956209705, + "loss": 0.6002, + "step": 2133 + }, + { + "epoch": 0.050024320633860184, + "grad_norm": 3.219287157058716, + "learning_rate": 0.0001987785520378937, + "loss": 0.6058, + "step": 2134 + }, + { + "epoch": 0.05004776220866518, + "grad_norm": 2.814493417739868, + "learning_rate": 0.00019877740397771518, + "loss": 0.4572, + "step": 2135 + }, + { + "epoch": 0.05007120378347017, + "grad_norm": 1.3239073753356934, + "learning_rate": 0.00019877625538156768, + "loss": 0.1372, + "step": 2136 + }, + { + "epoch": 0.05009464535827517, + "grad_norm": 4.02598762512207, + "learning_rate": 0.00019877510624945744, + "loss": 0.4665, + "step": 2137 + }, + { + "epoch": 0.05011808693308016, + "grad_norm": 1.7031323909759521, + "learning_rate": 0.0001987739565813907, + "loss": 0.2897, + "step": 2138 + }, + { + "epoch": 0.05014152850788516, + "grad_norm": 4.0260748863220215, + "learning_rate": 0.00019877280637737365, + "loss": 0.6476, + "step": 2139 + }, + { + "epoch": 0.050164970082690157, + "grad_norm": 2.5870893001556396, + "learning_rate": 0.00019877165563741262, + "loss": 1.418, + "step": 2140 + }, + { + "epoch": 0.05018841165749515, + "grad_norm": 3.2915995121002197, + "learning_rate": 0.00019877050436151379, + "loss": 0.3573, + "step": 2141 + }, + { + "epoch": 0.050211853232300145, + "grad_norm": 4.719298362731934, + "learning_rate": 0.00019876935254968344, + "loss": 0.7431, + "step": 2142 + }, + { + "epoch": 0.05023529480710514, + "grad_norm": 3.8221373558044434, + "learning_rate": 0.0001987682002019278, + "loss": 0.5391, + "step": 2143 + }, + { + "epoch": 0.050258736381910134, + "grad_norm": 2.9844813346862793, + "learning_rate": 0.00019876704731825313, + "loss": 0.541, + "step": 2144 + }, + { + "epoch": 0.050282177956715135, + "grad_norm": 1.658829689025879, + "learning_rate": 0.00019876589389866567, + "loss": 0.1921, + "step": 2145 + }, + { + "epoch": 0.05030561953152013, + "grad_norm": 3.622758388519287, + "learning_rate": 0.0001987647399431717, + "loss": 0.4537, + "step": 2146 + }, + { + "epoch": 0.050329061106325124, + "grad_norm": 2.4875214099884033, + "learning_rate": 0.0001987635854517775, + "loss": 1.5663, + "step": 2147 + }, + { + "epoch": 0.05035250268113012, + "grad_norm": 4.380819797515869, + "learning_rate": 0.00019876243042448927, + "loss": 0.6986, + "step": 2148 + }, + { + "epoch": 0.05037594425593511, + "grad_norm": 6.251289367675781, + "learning_rate": 0.00019876127486131336, + "loss": 1.1775, + "step": 2149 + }, + { + "epoch": 0.05039938583074011, + "grad_norm": 4.24191951751709, + "learning_rate": 0.00019876011876225594, + "loss": 0.8778, + "step": 2150 + }, + { + "epoch": 0.05042282740554511, + "grad_norm": 4.43391227722168, + "learning_rate": 0.0001987589621273234, + "loss": 0.6536, + "step": 2151 + }, + { + "epoch": 0.0504462689803501, + "grad_norm": 3.34070086479187, + "learning_rate": 0.0001987578049565219, + "loss": 0.4612, + "step": 2152 + }, + { + "epoch": 0.050469710555155096, + "grad_norm": 0.3712240159511566, + "learning_rate": 0.0001987566472498578, + "loss": 0.0603, + "step": 2153 + }, + { + "epoch": 0.05049315212996009, + "grad_norm": 1.1494314670562744, + "learning_rate": 0.00019875548900733737, + "loss": 0.2428, + "step": 2154 + }, + { + "epoch": 0.050516593704765085, + "grad_norm": 2.891000509262085, + "learning_rate": 0.00019875433022896685, + "loss": 0.5416, + "step": 2155 + }, + { + "epoch": 0.05054003527957008, + "grad_norm": 4.250270366668701, + "learning_rate": 0.00019875317091475256, + "loss": 0.6075, + "step": 2156 + }, + { + "epoch": 0.050563476854375074, + "grad_norm": 2.9667434692382812, + "learning_rate": 0.00019875201106470082, + "loss": 0.4704, + "step": 2157 + }, + { + "epoch": 0.050586918429180075, + "grad_norm": 4.531314849853516, + "learning_rate": 0.00019875085067881784, + "loss": 0.6866, + "step": 2158 + }, + { + "epoch": 0.05061036000398507, + "grad_norm": 3.718564033508301, + "learning_rate": 0.00019874968975710998, + "loss": 0.7867, + "step": 2159 + }, + { + "epoch": 0.050633801578790064, + "grad_norm": 5.267390727996826, + "learning_rate": 0.00019874852829958358, + "loss": 0.7295, + "step": 2160 + }, + { + "epoch": 0.05065724315359506, + "grad_norm": 4.319880962371826, + "learning_rate": 0.00019874736630624482, + "loss": 0.8098, + "step": 2161 + }, + { + "epoch": 0.05068068472840005, + "grad_norm": 5.0942206382751465, + "learning_rate": 0.00019874620377710012, + "loss": 0.8372, + "step": 2162 + }, + { + "epoch": 0.05070412630320505, + "grad_norm": 3.4782228469848633, + "learning_rate": 0.00019874504071215574, + "loss": 0.5409, + "step": 2163 + }, + { + "epoch": 0.05072756787801005, + "grad_norm": 3.26560640335083, + "learning_rate": 0.00019874387711141798, + "loss": 0.6283, + "step": 2164 + }, + { + "epoch": 0.05075100945281504, + "grad_norm": 1.089706301689148, + "learning_rate": 0.00019874271297489314, + "loss": 0.1638, + "step": 2165 + }, + { + "epoch": 0.050774451027620036, + "grad_norm": 2.870004892349243, + "learning_rate": 0.0001987415483025876, + "loss": 2.6649, + "step": 2166 + }, + { + "epoch": 0.05079789260242503, + "grad_norm": 4.393591403961182, + "learning_rate": 0.00019874038309450765, + "loss": 0.7021, + "step": 2167 + }, + { + "epoch": 0.050821334177230025, + "grad_norm": 4.454611778259277, + "learning_rate": 0.00019873921735065957, + "loss": 0.7797, + "step": 2168 + }, + { + "epoch": 0.05084477575203502, + "grad_norm": 3.5676136016845703, + "learning_rate": 0.00019873805107104974, + "loss": 0.524, + "step": 2169 + }, + { + "epoch": 0.050868217326840014, + "grad_norm": 4.128134727478027, + "learning_rate": 0.00019873688425568447, + "loss": 0.5543, + "step": 2170 + }, + { + "epoch": 0.050891658901645015, + "grad_norm": 3.874295949935913, + "learning_rate": 0.00019873571690457006, + "loss": 0.7921, + "step": 2171 + }, + { + "epoch": 0.05091510047645001, + "grad_norm": 1.6644428968429565, + "learning_rate": 0.0001987345490177129, + "loss": 2.1954, + "step": 2172 + }, + { + "epoch": 0.050938542051255004, + "grad_norm": 5.1570234298706055, + "learning_rate": 0.0001987333805951193, + "loss": 0.7724, + "step": 2173 + }, + { + "epoch": 0.05096198362606, + "grad_norm": 2.9145593643188477, + "learning_rate": 0.0001987322116367956, + "loss": 0.4753, + "step": 2174 + }, + { + "epoch": 0.05098542520086499, + "grad_norm": 1.538043737411499, + "learning_rate": 0.00019873104214274815, + "loss": 1.1302, + "step": 2175 + }, + { + "epoch": 0.051008866775669987, + "grad_norm": 2.6530885696411133, + "learning_rate": 0.00019872987211298328, + "loss": 0.3205, + "step": 2176 + }, + { + "epoch": 0.05103230835047499, + "grad_norm": 3.2888100147247314, + "learning_rate": 0.00019872870154750735, + "loss": 0.393, + "step": 2177 + }, + { + "epoch": 0.05105574992527998, + "grad_norm": 1.1946911811828613, + "learning_rate": 0.0001987275304463267, + "loss": 0.209, + "step": 2178 + }, + { + "epoch": 0.051079191500084976, + "grad_norm": 4.455893039703369, + "learning_rate": 0.0001987263588094477, + "loss": 0.719, + "step": 2179 + }, + { + "epoch": 0.05110263307488997, + "grad_norm": 1.3918591737747192, + "learning_rate": 0.0001987251866368767, + "loss": 0.1859, + "step": 2180 + }, + { + "epoch": 0.051126074649694965, + "grad_norm": 0.6581496596336365, + "learning_rate": 0.0001987240139286201, + "loss": 0.0974, + "step": 2181 + }, + { + "epoch": 0.05114951622449996, + "grad_norm": 4.9593400955200195, + "learning_rate": 0.0001987228406846842, + "loss": 0.7221, + "step": 2182 + }, + { + "epoch": 0.05117295779930496, + "grad_norm": 3.5930933952331543, + "learning_rate": 0.00019872166690507536, + "loss": 0.4957, + "step": 2183 + }, + { + "epoch": 0.051196399374109955, + "grad_norm": 1.8975884914398193, + "learning_rate": 0.0001987204925898, + "loss": 0.3774, + "step": 2184 + }, + { + "epoch": 0.05121984094891495, + "grad_norm": 3.6871697902679443, + "learning_rate": 0.00019871931773886447, + "loss": 0.4442, + "step": 2185 + }, + { + "epoch": 0.051243282523719944, + "grad_norm": 4.891168117523193, + "learning_rate": 0.00019871814235227517, + "loss": 0.6026, + "step": 2186 + }, + { + "epoch": 0.05126672409852494, + "grad_norm": 2.930398941040039, + "learning_rate": 0.00019871696643003844, + "loss": 0.4239, + "step": 2187 + }, + { + "epoch": 0.05129016567332993, + "grad_norm": 5.807085037231445, + "learning_rate": 0.00019871578997216065, + "loss": 1.1448, + "step": 2188 + }, + { + "epoch": 0.051313607248134926, + "grad_norm": 4.1021728515625, + "learning_rate": 0.0001987146129786482, + "loss": 0.5959, + "step": 2189 + }, + { + "epoch": 0.05133704882293993, + "grad_norm": 3.998405694961548, + "learning_rate": 0.00019871343544950753, + "loss": 0.8123, + "step": 2190 + }, + { + "epoch": 0.05136049039774492, + "grad_norm": 2.0190467834472656, + "learning_rate": 0.00019871225738474498, + "loss": 1.5853, + "step": 2191 + }, + { + "epoch": 0.051383931972549916, + "grad_norm": 2.9910573959350586, + "learning_rate": 0.00019871107878436693, + "loss": 0.2963, + "step": 2192 + }, + { + "epoch": 0.05140737354735491, + "grad_norm": 4.538352012634277, + "learning_rate": 0.00019870989964837978, + "loss": 0.8377, + "step": 2193 + }, + { + "epoch": 0.051430815122159905, + "grad_norm": 1.0785764455795288, + "learning_rate": 0.00019870871997678996, + "loss": 0.1645, + "step": 2194 + }, + { + "epoch": 0.0514542566969649, + "grad_norm": 3.3164823055267334, + "learning_rate": 0.00019870753976960383, + "loss": 0.65, + "step": 2195 + }, + { + "epoch": 0.0514776982717699, + "grad_norm": 1.838984727859497, + "learning_rate": 0.00019870635902682784, + "loss": 0.3426, + "step": 2196 + }, + { + "epoch": 0.051501139846574895, + "grad_norm": 5.662619590759277, + "learning_rate": 0.00019870517774846834, + "loss": 0.9709, + "step": 2197 + }, + { + "epoch": 0.05152458142137989, + "grad_norm": 2.0128400325775146, + "learning_rate": 0.00019870399593453178, + "loss": 1.3721, + "step": 2198 + }, + { + "epoch": 0.051548022996184883, + "grad_norm": 4.40773344039917, + "learning_rate": 0.00019870281358502458, + "loss": 0.6957, + "step": 2199 + }, + { + "epoch": 0.05157146457098988, + "grad_norm": 5.179568767547607, + "learning_rate": 0.0001987016306999531, + "loss": 0.853, + "step": 2200 + }, + { + "epoch": 0.05159490614579487, + "grad_norm": 1.0808535814285278, + "learning_rate": 0.00019870044727932383, + "loss": 0.2117, + "step": 2201 + }, + { + "epoch": 0.05161834772059987, + "grad_norm": 2.146134614944458, + "learning_rate": 0.00019869926332314317, + "loss": 0.2094, + "step": 2202 + }, + { + "epoch": 0.05164178929540487, + "grad_norm": 2.9623539447784424, + "learning_rate": 0.0001986980788314175, + "loss": 0.3184, + "step": 2203 + }, + { + "epoch": 0.05166523087020986, + "grad_norm": 0.30559247732162476, + "learning_rate": 0.00019869689380415327, + "loss": 0.0356, + "step": 2204 + }, + { + "epoch": 0.051688672445014856, + "grad_norm": 3.069544792175293, + "learning_rate": 0.00019869570824135698, + "loss": 0.3638, + "step": 2205 + }, + { + "epoch": 0.05171211401981985, + "grad_norm": 7.221243858337402, + "learning_rate": 0.0001986945221430349, + "loss": 0.7627, + "step": 2206 + }, + { + "epoch": 0.051735555594624845, + "grad_norm": 2.5555641651153564, + "learning_rate": 0.00019869333550919367, + "loss": 0.5347, + "step": 2207 + }, + { + "epoch": 0.05175899716942984, + "grad_norm": 3.0221521854400635, + "learning_rate": 0.00019869214833983959, + "loss": 0.4571, + "step": 2208 + }, + { + "epoch": 0.05178243874423484, + "grad_norm": 3.078815460205078, + "learning_rate": 0.00019869096063497913, + "loss": 0.471, + "step": 2209 + }, + { + "epoch": 0.051805880319039835, + "grad_norm": 7.838615894317627, + "learning_rate": 0.00019868977239461874, + "loss": 0.8159, + "step": 2210 + }, + { + "epoch": 0.05182932189384483, + "grad_norm": 4.158656597137451, + "learning_rate": 0.00019868858361876487, + "loss": 0.7167, + "step": 2211 + }, + { + "epoch": 0.05185276346864982, + "grad_norm": 3.91782283782959, + "learning_rate": 0.00019868739430742397, + "loss": 0.4894, + "step": 2212 + }, + { + "epoch": 0.05187620504345482, + "grad_norm": 3.4934868812561035, + "learning_rate": 0.0001986862044606025, + "loss": 0.3642, + "step": 2213 + }, + { + "epoch": 0.05189964661825981, + "grad_norm": 1.0904358625411987, + "learning_rate": 0.00019868501407830692, + "loss": 0.1791, + "step": 2214 + }, + { + "epoch": 0.05192308819306481, + "grad_norm": 1.6000466346740723, + "learning_rate": 0.00019868382316054367, + "loss": 1.2783, + "step": 2215 + }, + { + "epoch": 0.05194652976786981, + "grad_norm": 4.043782711029053, + "learning_rate": 0.00019868263170731923, + "loss": 0.6414, + "step": 2216 + }, + { + "epoch": 0.0519699713426748, + "grad_norm": 1.5396759510040283, + "learning_rate": 0.00019868143971864006, + "loss": 0.3261, + "step": 2217 + }, + { + "epoch": 0.051993412917479796, + "grad_norm": 2.1355888843536377, + "learning_rate": 0.0001986802471945126, + "loss": 1.6068, + "step": 2218 + }, + { + "epoch": 0.05201685449228479, + "grad_norm": 1.8843214511871338, + "learning_rate": 0.00019867905413494335, + "loss": 1.5929, + "step": 2219 + }, + { + "epoch": 0.052040296067089785, + "grad_norm": 4.1277618408203125, + "learning_rate": 0.00019867786053993878, + "loss": 0.4289, + "step": 2220 + }, + { + "epoch": 0.05206373764189478, + "grad_norm": 1.374852180480957, + "learning_rate": 0.00019867666640950536, + "loss": 0.1534, + "step": 2221 + }, + { + "epoch": 0.05208717921669978, + "grad_norm": 3.1257197856903076, + "learning_rate": 0.0001986754717436496, + "loss": 0.3649, + "step": 2222 + }, + { + "epoch": 0.052110620791504775, + "grad_norm": 0.9387006163597107, + "learning_rate": 0.00019867427654237792, + "loss": 0.2008, + "step": 2223 + }, + { + "epoch": 0.05213406236630977, + "grad_norm": 1.4715886116027832, + "learning_rate": 0.00019867308080569686, + "loss": 0.2561, + "step": 2224 + }, + { + "epoch": 0.05215750394111476, + "grad_norm": 3.782827854156494, + "learning_rate": 0.00019867188453361287, + "loss": 0.6341, + "step": 2225 + }, + { + "epoch": 0.05218094551591976, + "grad_norm": 5.3641839027404785, + "learning_rate": 0.0001986706877261325, + "loss": 0.6611, + "step": 2226 + }, + { + "epoch": 0.05220438709072475, + "grad_norm": 2.4702301025390625, + "learning_rate": 0.00019866949038326216, + "loss": 0.2342, + "step": 2227 + }, + { + "epoch": 0.05222782866552975, + "grad_norm": 2.9183013439178467, + "learning_rate": 0.00019866829250500843, + "loss": 0.5954, + "step": 2228 + }, + { + "epoch": 0.05225127024033475, + "grad_norm": 6.276961326599121, + "learning_rate": 0.00019866709409137773, + "loss": 0.9123, + "step": 2229 + }, + { + "epoch": 0.05227471181513974, + "grad_norm": 2.826490640640259, + "learning_rate": 0.00019866589514237664, + "loss": 0.6247, + "step": 2230 + }, + { + "epoch": 0.052298153389944736, + "grad_norm": 4.191281318664551, + "learning_rate": 0.0001986646956580116, + "loss": 0.6078, + "step": 2231 + }, + { + "epoch": 0.05232159496474973, + "grad_norm": 5.567075729370117, + "learning_rate": 0.00019866349563828916, + "loss": 1.0782, + "step": 2232 + }, + { + "epoch": 0.052345036539554725, + "grad_norm": 3.824962615966797, + "learning_rate": 0.0001986622950832158, + "loss": 0.3594, + "step": 2233 + }, + { + "epoch": 0.052368478114359726, + "grad_norm": 1.5000219345092773, + "learning_rate": 0.0001986610939927981, + "loss": 0.1441, + "step": 2234 + }, + { + "epoch": 0.05239191968916472, + "grad_norm": 2.5606303215026855, + "learning_rate": 0.0001986598923670425, + "loss": 0.4567, + "step": 2235 + }, + { + "epoch": 0.052415361263969715, + "grad_norm": 3.0405235290527344, + "learning_rate": 0.00019865869020595553, + "loss": 0.4786, + "step": 2236 + }, + { + "epoch": 0.05243880283877471, + "grad_norm": 1.4126496315002441, + "learning_rate": 0.00019865748750954375, + "loss": 0.2175, + "step": 2237 + }, + { + "epoch": 0.0524622444135797, + "grad_norm": 3.2982065677642822, + "learning_rate": 0.00019865628427781368, + "loss": 0.8448, + "step": 2238 + }, + { + "epoch": 0.0524856859883847, + "grad_norm": 1.0293220281600952, + "learning_rate": 0.00019865508051077182, + "loss": 0.2045, + "step": 2239 + }, + { + "epoch": 0.05250912756318969, + "grad_norm": 3.350640058517456, + "learning_rate": 0.00019865387620842471, + "loss": 0.5474, + "step": 2240 + }, + { + "epoch": 0.05253256913799469, + "grad_norm": 3.362516164779663, + "learning_rate": 0.00019865267137077894, + "loss": 0.4923, + "step": 2241 + }, + { + "epoch": 0.05255601071279969, + "grad_norm": 2.188168525695801, + "learning_rate": 0.00019865146599784097, + "loss": 0.2634, + "step": 2242 + }, + { + "epoch": 0.05257945228760468, + "grad_norm": 3.0093722343444824, + "learning_rate": 0.00019865026008961738, + "loss": 0.6404, + "step": 2243 + }, + { + "epoch": 0.052602893862409676, + "grad_norm": 2.678436040878296, + "learning_rate": 0.0001986490536461147, + "loss": 0.5383, + "step": 2244 + }, + { + "epoch": 0.05262633543721467, + "grad_norm": 4.059513092041016, + "learning_rate": 0.00019864784666733947, + "loss": 0.6134, + "step": 2245 + }, + { + "epoch": 0.052649777012019665, + "grad_norm": 3.3428940773010254, + "learning_rate": 0.00019864663915329826, + "loss": 0.4904, + "step": 2246 + }, + { + "epoch": 0.052673218586824666, + "grad_norm": 3.846158742904663, + "learning_rate": 0.0001986454311039976, + "loss": 0.5794, + "step": 2247 + }, + { + "epoch": 0.05269666016162966, + "grad_norm": 3.9227519035339355, + "learning_rate": 0.0001986442225194441, + "loss": 0.5637, + "step": 2248 + }, + { + "epoch": 0.052720101736434655, + "grad_norm": 2.3668768405914307, + "learning_rate": 0.00019864301339964424, + "loss": 1.6151, + "step": 2249 + }, + { + "epoch": 0.05274354331123965, + "grad_norm": 2.641728401184082, + "learning_rate": 0.00019864180374460464, + "loss": 0.4861, + "step": 2250 + }, + { + "epoch": 0.05276698488604464, + "grad_norm": 1.882351040840149, + "learning_rate": 0.0001986405935543318, + "loss": 0.3841, + "step": 2251 + }, + { + "epoch": 0.05279042646084964, + "grad_norm": 4.746920108795166, + "learning_rate": 0.00019863938282883235, + "loss": 1.0529, + "step": 2252 + }, + { + "epoch": 0.05281386803565463, + "grad_norm": 3.790473699569702, + "learning_rate": 0.00019863817156811283, + "loss": 0.4989, + "step": 2253 + }, + { + "epoch": 0.05283730961045963, + "grad_norm": 1.4219281673431396, + "learning_rate": 0.0001986369597721798, + "loss": 0.2493, + "step": 2254 + }, + { + "epoch": 0.05286075118526463, + "grad_norm": 4.223896503448486, + "learning_rate": 0.0001986357474410399, + "loss": 0.5946, + "step": 2255 + }, + { + "epoch": 0.05288419276006962, + "grad_norm": 1.7537404298782349, + "learning_rate": 0.00019863453457469966, + "loss": 0.1935, + "step": 2256 + }, + { + "epoch": 0.052907634334874616, + "grad_norm": 2.8455512523651123, + "learning_rate": 0.0001986333211731656, + "loss": 0.4919, + "step": 2257 + }, + { + "epoch": 0.05293107590967961, + "grad_norm": 0.8862664103507996, + "learning_rate": 0.0001986321072364444, + "loss": 0.5131, + "step": 2258 + }, + { + "epoch": 0.052954517484484605, + "grad_norm": 0.5613242387771606, + "learning_rate": 0.00019863089276454262, + "loss": 0.0619, + "step": 2259 + }, + { + "epoch": 0.052977959059289606, + "grad_norm": 3.1631057262420654, + "learning_rate": 0.00019862967775746685, + "loss": 0.5689, + "step": 2260 + }, + { + "epoch": 0.0530014006340946, + "grad_norm": 6.595679759979248, + "learning_rate": 0.00019862846221522365, + "loss": 1.0925, + "step": 2261 + }, + { + "epoch": 0.053024842208899595, + "grad_norm": 2.576502561569214, + "learning_rate": 0.00019862724613781967, + "loss": 1.8313, + "step": 2262 + }, + { + "epoch": 0.05304828378370459, + "grad_norm": 2.8356473445892334, + "learning_rate": 0.00019862602952526145, + "loss": 0.2495, + "step": 2263 + }, + { + "epoch": 0.05307172535850958, + "grad_norm": 1.2205791473388672, + "learning_rate": 0.00019862481237755563, + "loss": 0.3223, + "step": 2264 + }, + { + "epoch": 0.05309516693331458, + "grad_norm": 4.078831672668457, + "learning_rate": 0.00019862359469470883, + "loss": 0.5943, + "step": 2265 + }, + { + "epoch": 0.05311860850811958, + "grad_norm": 1.4120328426361084, + "learning_rate": 0.00019862237647672758, + "loss": 0.3564, + "step": 2266 + }, + { + "epoch": 0.05314205008292457, + "grad_norm": 3.306137800216675, + "learning_rate": 0.0001986211577236186, + "loss": 0.5783, + "step": 2267 + }, + { + "epoch": 0.05316549165772957, + "grad_norm": 0.7686960101127625, + "learning_rate": 0.0001986199384353884, + "loss": 0.1068, + "step": 2268 + }, + { + "epoch": 0.05318893323253456, + "grad_norm": 3.2135963439941406, + "learning_rate": 0.00019861871861204367, + "loss": 0.2788, + "step": 2269 + }, + { + "epoch": 0.053212374807339556, + "grad_norm": 4.120006084442139, + "learning_rate": 0.00019861749825359097, + "loss": 0.8046, + "step": 2270 + }, + { + "epoch": 0.05323581638214455, + "grad_norm": 0.7719650268554688, + "learning_rate": 0.000198616277360037, + "loss": 0.2216, + "step": 2271 + }, + { + "epoch": 0.053259257956949545, + "grad_norm": 3.236184597015381, + "learning_rate": 0.00019861505593138826, + "loss": 0.328, + "step": 2272 + }, + { + "epoch": 0.053282699531754546, + "grad_norm": 1.4998570680618286, + "learning_rate": 0.00019861383396765155, + "loss": 0.161, + "step": 2273 + }, + { + "epoch": 0.05330614110655954, + "grad_norm": 3.4209721088409424, + "learning_rate": 0.00019861261146883333, + "loss": 0.5297, + "step": 2274 + }, + { + "epoch": 0.053329582681364535, + "grad_norm": 0.6550478339195251, + "learning_rate": 0.00019861138843494034, + "loss": 0.1158, + "step": 2275 + }, + { + "epoch": 0.05335302425616953, + "grad_norm": 3.5510549545288086, + "learning_rate": 0.00019861016486597918, + "loss": 0.6461, + "step": 2276 + }, + { + "epoch": 0.05337646583097452, + "grad_norm": 4.171230316162109, + "learning_rate": 0.00019860894076195647, + "loss": 0.9369, + "step": 2277 + }, + { + "epoch": 0.05339990740577952, + "grad_norm": 1.5836173295974731, + "learning_rate": 0.00019860771612287888, + "loss": 0.2576, + "step": 2278 + }, + { + "epoch": 0.05342334898058452, + "grad_norm": 0.5121792554855347, + "learning_rate": 0.0001986064909487531, + "loss": 0.1205, + "step": 2279 + }, + { + "epoch": 0.05344679055538951, + "grad_norm": 0.8345749974250793, + "learning_rate": 0.0001986052652395857, + "loss": 0.1567, + "step": 2280 + }, + { + "epoch": 0.05347023213019451, + "grad_norm": 2.97969913482666, + "learning_rate": 0.00019860403899538332, + "loss": 0.6321, + "step": 2281 + }, + { + "epoch": 0.0534936737049995, + "grad_norm": 3.92561411857605, + "learning_rate": 0.00019860281221615269, + "loss": 0.9832, + "step": 2282 + }, + { + "epoch": 0.053517115279804496, + "grad_norm": 4.4167160987854, + "learning_rate": 0.00019860158490190042, + "loss": 0.8445, + "step": 2283 + }, + { + "epoch": 0.05354055685460949, + "grad_norm": 2.561845302581787, + "learning_rate": 0.00019860035705263318, + "loss": 0.3482, + "step": 2284 + }, + { + "epoch": 0.053563998429414485, + "grad_norm": 3.3450894355773926, + "learning_rate": 0.0001985991286683576, + "loss": 0.4037, + "step": 2285 + }, + { + "epoch": 0.053587440004219486, + "grad_norm": 0.596904993057251, + "learning_rate": 0.0001985978997490804, + "loss": 0.0998, + "step": 2286 + }, + { + "epoch": 0.05361088157902448, + "grad_norm": 2.9960992336273193, + "learning_rate": 0.00019859667029480823, + "loss": 0.5627, + "step": 2287 + }, + { + "epoch": 0.053634323153829475, + "grad_norm": 0.40338134765625, + "learning_rate": 0.00019859544030554775, + "loss": 0.0557, + "step": 2288 + }, + { + "epoch": 0.05365776472863447, + "grad_norm": 2.492493152618408, + "learning_rate": 0.00019859420978130565, + "loss": 0.4531, + "step": 2289 + }, + { + "epoch": 0.05368120630343946, + "grad_norm": 3.90397310256958, + "learning_rate": 0.0001985929787220886, + "loss": 0.8321, + "step": 2290 + }, + { + "epoch": 0.05370464787824446, + "grad_norm": 3.627213478088379, + "learning_rate": 0.00019859174712790324, + "loss": 0.4477, + "step": 2291 + }, + { + "epoch": 0.05372808945304946, + "grad_norm": 3.909595489501953, + "learning_rate": 0.00019859051499875632, + "loss": 0.5821, + "step": 2292 + }, + { + "epoch": 0.05375153102785445, + "grad_norm": 5.182501316070557, + "learning_rate": 0.00019858928233465448, + "loss": 0.8419, + "step": 2293 + }, + { + "epoch": 0.05377497260265945, + "grad_norm": 3.8989369869232178, + "learning_rate": 0.00019858804913560442, + "loss": 0.6302, + "step": 2294 + }, + { + "epoch": 0.05379841417746444, + "grad_norm": 2.3603341579437256, + "learning_rate": 0.00019858681540161285, + "loss": 0.5379, + "step": 2295 + }, + { + "epoch": 0.053821855752269436, + "grad_norm": 2.6117942333221436, + "learning_rate": 0.00019858558113268644, + "loss": 1.3834, + "step": 2296 + }, + { + "epoch": 0.05384529732707443, + "grad_norm": 4.89403772354126, + "learning_rate": 0.00019858434632883187, + "loss": 0.8124, + "step": 2297 + }, + { + "epoch": 0.05386873890187943, + "grad_norm": 6.076624393463135, + "learning_rate": 0.00019858311099005588, + "loss": 0.6466, + "step": 2298 + }, + { + "epoch": 0.053892180476684426, + "grad_norm": 4.739457607269287, + "learning_rate": 0.00019858187511636517, + "loss": 0.9018, + "step": 2299 + }, + { + "epoch": 0.05391562205148942, + "grad_norm": 4.5111212730407715, + "learning_rate": 0.00019858063870776643, + "loss": 0.4831, + "step": 2300 + }, + { + "epoch": 0.053939063626294415, + "grad_norm": 3.0193300247192383, + "learning_rate": 0.00019857940176426638, + "loss": 0.3148, + "step": 2301 + }, + { + "epoch": 0.05396250520109941, + "grad_norm": 6.211028575897217, + "learning_rate": 0.0001985781642858717, + "loss": 0.7455, + "step": 2302 + }, + { + "epoch": 0.0539859467759044, + "grad_norm": 0.9694613814353943, + "learning_rate": 0.00019857692627258912, + "loss": 0.1112, + "step": 2303 + }, + { + "epoch": 0.0540093883507094, + "grad_norm": 2.0403623580932617, + "learning_rate": 0.0001985756877244254, + "loss": 1.4483, + "step": 2304 + }, + { + "epoch": 0.0540328299255144, + "grad_norm": 4.489995002746582, + "learning_rate": 0.00019857444864138722, + "loss": 1.0452, + "step": 2305 + }, + { + "epoch": 0.05405627150031939, + "grad_norm": 0.9427937269210815, + "learning_rate": 0.00019857320902348131, + "loss": 0.1174, + "step": 2306 + }, + { + "epoch": 0.05407971307512439, + "grad_norm": 5.650027751922607, + "learning_rate": 0.00019857196887071438, + "loss": 0.9167, + "step": 2307 + }, + { + "epoch": 0.05410315464992938, + "grad_norm": 3.9392812252044678, + "learning_rate": 0.0001985707281830932, + "loss": 0.5163, + "step": 2308 + }, + { + "epoch": 0.054126596224734376, + "grad_norm": 0.5502675771713257, + "learning_rate": 0.00019856948696062444, + "loss": 0.0872, + "step": 2309 + }, + { + "epoch": 0.05415003779953937, + "grad_norm": 2.1099178791046143, + "learning_rate": 0.00019856824520331491, + "loss": 0.4545, + "step": 2310 + }, + { + "epoch": 0.05417347937434437, + "grad_norm": 3.061521053314209, + "learning_rate": 0.0001985670029111713, + "loss": 0.5528, + "step": 2311 + }, + { + "epoch": 0.054196920949149366, + "grad_norm": 3.4675207138061523, + "learning_rate": 0.00019856576008420035, + "loss": 0.4009, + "step": 2312 + }, + { + "epoch": 0.05422036252395436, + "grad_norm": 3.5451483726501465, + "learning_rate": 0.00019856451672240884, + "loss": 0.4414, + "step": 2313 + }, + { + "epoch": 0.054243804098759354, + "grad_norm": 0.5891457796096802, + "learning_rate": 0.00019856327282580347, + "loss": 0.0685, + "step": 2314 + }, + { + "epoch": 0.05426724567356435, + "grad_norm": 3.593372344970703, + "learning_rate": 0.000198562028394391, + "loss": 0.4871, + "step": 2315 + }, + { + "epoch": 0.05429068724836934, + "grad_norm": 2.239288330078125, + "learning_rate": 0.00019856078342817818, + "loss": 1.3716, + "step": 2316 + }, + { + "epoch": 0.05431412882317434, + "grad_norm": 2.862464427947998, + "learning_rate": 0.0001985595379271718, + "loss": 0.6162, + "step": 2317 + }, + { + "epoch": 0.05433757039797934, + "grad_norm": 4.7928876876831055, + "learning_rate": 0.00019855829189137863, + "loss": 0.3496, + "step": 2318 + }, + { + "epoch": 0.05436101197278433, + "grad_norm": 1.7823638916015625, + "learning_rate": 0.00019855704532080535, + "loss": 1.218, + "step": 2319 + }, + { + "epoch": 0.05438445354758933, + "grad_norm": 3.5785415172576904, + "learning_rate": 0.0001985557982154588, + "loss": 0.6408, + "step": 2320 + }, + { + "epoch": 0.05440789512239432, + "grad_norm": 2.1480531692504883, + "learning_rate": 0.00019855455057534571, + "loss": 1.2361, + "step": 2321 + }, + { + "epoch": 0.054431336697199316, + "grad_norm": 3.813173294067383, + "learning_rate": 0.00019855330240047282, + "loss": 0.7838, + "step": 2322 + }, + { + "epoch": 0.05445477827200431, + "grad_norm": 1.0556882619857788, + "learning_rate": 0.000198552053690847, + "loss": 0.1574, + "step": 2323 + }, + { + "epoch": 0.05447821984680931, + "grad_norm": 4.466336727142334, + "learning_rate": 0.0001985508044464749, + "loss": 0.673, + "step": 2324 + }, + { + "epoch": 0.054501661421614306, + "grad_norm": 2.1200032234191895, + "learning_rate": 0.00019854955466736343, + "loss": 0.169, + "step": 2325 + }, + { + "epoch": 0.0545251029964193, + "grad_norm": 3.985812187194824, + "learning_rate": 0.00019854830435351927, + "loss": 0.6425, + "step": 2326 + }, + { + "epoch": 0.054548544571224294, + "grad_norm": 3.8637828826904297, + "learning_rate": 0.00019854705350494926, + "loss": 0.4983, + "step": 2327 + }, + { + "epoch": 0.05457198614602929, + "grad_norm": 2.720740556716919, + "learning_rate": 0.00019854580212166014, + "loss": 0.4089, + "step": 2328 + }, + { + "epoch": 0.05459542772083428, + "grad_norm": 4.084686756134033, + "learning_rate": 0.00019854455020365874, + "loss": 0.6842, + "step": 2329 + }, + { + "epoch": 0.054618869295639284, + "grad_norm": 4.233062744140625, + "learning_rate": 0.00019854329775095185, + "loss": 0.7882, + "step": 2330 + }, + { + "epoch": 0.05464231087044428, + "grad_norm": 2.8787901401519775, + "learning_rate": 0.00019854204476354626, + "loss": 0.3501, + "step": 2331 + }, + { + "epoch": 0.05466575244524927, + "grad_norm": 2.6811165809631348, + "learning_rate": 0.00019854079124144872, + "loss": 0.3323, + "step": 2332 + }, + { + "epoch": 0.05468919402005427, + "grad_norm": 5.885102272033691, + "learning_rate": 0.00019853953718466613, + "loss": 0.9261, + "step": 2333 + }, + { + "epoch": 0.05471263559485926, + "grad_norm": 0.7082968354225159, + "learning_rate": 0.0001985382825932052, + "loss": 0.1317, + "step": 2334 + }, + { + "epoch": 0.054736077169664256, + "grad_norm": 2.9830563068389893, + "learning_rate": 0.00019853702746707282, + "loss": 0.346, + "step": 2335 + }, + { + "epoch": 0.05475951874446925, + "grad_norm": 4.448103427886963, + "learning_rate": 0.0001985357718062757, + "loss": 0.8351, + "step": 2336 + }, + { + "epoch": 0.05478296031927425, + "grad_norm": 1.5139403343200684, + "learning_rate": 0.00019853451561082074, + "loss": 0.2933, + "step": 2337 + }, + { + "epoch": 0.054806401894079246, + "grad_norm": 5.659348964691162, + "learning_rate": 0.00019853325888071475, + "loss": 0.8879, + "step": 2338 + }, + { + "epoch": 0.05482984346888424, + "grad_norm": 5.10483455657959, + "learning_rate": 0.0001985320016159645, + "loss": 0.5702, + "step": 2339 + }, + { + "epoch": 0.054853285043689234, + "grad_norm": 1.793739914894104, + "learning_rate": 0.00019853074381657684, + "loss": 1.56, + "step": 2340 + }, + { + "epoch": 0.05487672661849423, + "grad_norm": 6.9107160568237305, + "learning_rate": 0.0001985294854825586, + "loss": 0.7156, + "step": 2341 + }, + { + "epoch": 0.05490016819329922, + "grad_norm": 3.9475319385528564, + "learning_rate": 0.0001985282266139166, + "loss": 0.9096, + "step": 2342 + }, + { + "epoch": 0.054923609768104224, + "grad_norm": 3.256603717803955, + "learning_rate": 0.00019852696721065768, + "loss": 0.5047, + "step": 2343 + }, + { + "epoch": 0.05494705134290922, + "grad_norm": 0.9218440055847168, + "learning_rate": 0.00019852570727278866, + "loss": 0.1413, + "step": 2344 + }, + { + "epoch": 0.05497049291771421, + "grad_norm": 0.9719180464744568, + "learning_rate": 0.0001985244468003164, + "loss": 0.1825, + "step": 2345 + }, + { + "epoch": 0.05499393449251921, + "grad_norm": 3.455886125564575, + "learning_rate": 0.00019852318579324769, + "loss": 0.4017, + "step": 2346 + }, + { + "epoch": 0.0550173760673242, + "grad_norm": 3.630028009414673, + "learning_rate": 0.0001985219242515894, + "loss": 0.5361, + "step": 2347 + }, + { + "epoch": 0.055040817642129196, + "grad_norm": 3.2795021533966064, + "learning_rate": 0.00019852066217534842, + "loss": 0.738, + "step": 2348 + }, + { + "epoch": 0.05506425921693419, + "grad_norm": 1.6018657684326172, + "learning_rate": 0.00019851939956453153, + "loss": 0.4547, + "step": 2349 + }, + { + "epoch": 0.05508770079173919, + "grad_norm": 0.6184875965118408, + "learning_rate": 0.0001985181364191456, + "loss": 0.0936, + "step": 2350 + }, + { + "epoch": 0.055111142366544186, + "grad_norm": 1.4120328426361084, + "learning_rate": 0.00019851687273919751, + "loss": 0.199, + "step": 2351 + }, + { + "epoch": 0.05513458394134918, + "grad_norm": 2.217102289199829, + "learning_rate": 0.00019851560852469411, + "loss": 0.3086, + "step": 2352 + }, + { + "epoch": 0.055158025516154174, + "grad_norm": 0.3966127932071686, + "learning_rate": 0.0001985143437756422, + "loss": 0.1033, + "step": 2353 + }, + { + "epoch": 0.05518146709095917, + "grad_norm": 3.5829858779907227, + "learning_rate": 0.00019851307849204873, + "loss": 0.6148, + "step": 2354 + }, + { + "epoch": 0.05520490866576416, + "grad_norm": 3.6326661109924316, + "learning_rate": 0.00019851181267392052, + "loss": 0.6783, + "step": 2355 + }, + { + "epoch": 0.055228350240569164, + "grad_norm": 3.966737985610962, + "learning_rate": 0.00019851054632126443, + "loss": 0.5035, + "step": 2356 + }, + { + "epoch": 0.05525179181537416, + "grad_norm": 3.819857358932495, + "learning_rate": 0.00019850927943408734, + "loss": 0.73, + "step": 2357 + }, + { + "epoch": 0.05527523339017915, + "grad_norm": 0.9542332291603088, + "learning_rate": 0.00019850801201239617, + "loss": 0.1297, + "step": 2358 + }, + { + "epoch": 0.05529867496498415, + "grad_norm": 1.2587593793869019, + "learning_rate": 0.00019850674405619774, + "loss": 0.2878, + "step": 2359 + }, + { + "epoch": 0.05532211653978914, + "grad_norm": 1.2780805826187134, + "learning_rate": 0.00019850547556549893, + "loss": 0.1962, + "step": 2360 + }, + { + "epoch": 0.055345558114594136, + "grad_norm": 1.0629817247390747, + "learning_rate": 0.00019850420654030665, + "loss": 0.156, + "step": 2361 + }, + { + "epoch": 0.05536899968939914, + "grad_norm": 2.4995298385620117, + "learning_rate": 0.00019850293698062777, + "loss": 0.4057, + "step": 2362 + }, + { + "epoch": 0.05539244126420413, + "grad_norm": 2.9040534496307373, + "learning_rate": 0.0001985016668864692, + "loss": 0.4487, + "step": 2363 + }, + { + "epoch": 0.055415882839009126, + "grad_norm": 1.7332035303115845, + "learning_rate": 0.0001985003962578378, + "loss": 1.2425, + "step": 2364 + }, + { + "epoch": 0.05543932441381412, + "grad_norm": 3.948296546936035, + "learning_rate": 0.00019849912509474049, + "loss": 0.9007, + "step": 2365 + }, + { + "epoch": 0.055462765988619114, + "grad_norm": 1.474286675453186, + "learning_rate": 0.00019849785339718414, + "loss": 0.287, + "step": 2366 + }, + { + "epoch": 0.05548620756342411, + "grad_norm": 2.952230930328369, + "learning_rate": 0.00019849658116517568, + "loss": 0.442, + "step": 2367 + }, + { + "epoch": 0.0555096491382291, + "grad_norm": 3.5301365852355957, + "learning_rate": 0.000198495308398722, + "loss": 0.3743, + "step": 2368 + }, + { + "epoch": 0.055533090713034104, + "grad_norm": 4.162972450256348, + "learning_rate": 0.00019849403509783, + "loss": 0.7773, + "step": 2369 + }, + { + "epoch": 0.0555565322878391, + "grad_norm": 1.6715160608291626, + "learning_rate": 0.00019849276126250663, + "loss": 0.3863, + "step": 2370 + }, + { + "epoch": 0.05557997386264409, + "grad_norm": 4.345530033111572, + "learning_rate": 0.00019849148689275877, + "loss": 0.4453, + "step": 2371 + }, + { + "epoch": 0.05560341543744909, + "grad_norm": 5.092195510864258, + "learning_rate": 0.00019849021198859327, + "loss": 0.8778, + "step": 2372 + }, + { + "epoch": 0.05562685701225408, + "grad_norm": 2.337937116622925, + "learning_rate": 0.00019848893655001715, + "loss": 1.8449, + "step": 2373 + }, + { + "epoch": 0.055650298587059076, + "grad_norm": 3.259676456451416, + "learning_rate": 0.00019848766057703728, + "loss": 1.9373, + "step": 2374 + }, + { + "epoch": 0.05567374016186408, + "grad_norm": 0.9288241267204285, + "learning_rate": 0.00019848638406966062, + "loss": 0.1146, + "step": 2375 + }, + { + "epoch": 0.05569718173666907, + "grad_norm": 3.405905246734619, + "learning_rate": 0.00019848510702789404, + "loss": 0.3903, + "step": 2376 + }, + { + "epoch": 0.055720623311474066, + "grad_norm": 1.0392271280288696, + "learning_rate": 0.00019848382945174454, + "loss": 0.6928, + "step": 2377 + }, + { + "epoch": 0.05574406488627906, + "grad_norm": 1.4899851083755493, + "learning_rate": 0.00019848255134121895, + "loss": 1.2732, + "step": 2378 + }, + { + "epoch": 0.055767506461084054, + "grad_norm": 3.842026948928833, + "learning_rate": 0.00019848127269632432, + "loss": 0.6078, + "step": 2379 + }, + { + "epoch": 0.05579094803588905, + "grad_norm": 2.779252290725708, + "learning_rate": 0.00019847999351706755, + "loss": 0.6051, + "step": 2380 + }, + { + "epoch": 0.05581438961069405, + "grad_norm": 2.556439161300659, + "learning_rate": 0.0001984787138034555, + "loss": 0.2968, + "step": 2381 + }, + { + "epoch": 0.055837831185499044, + "grad_norm": 3.8209099769592285, + "learning_rate": 0.0001984774335554952, + "loss": 0.329, + "step": 2382 + }, + { + "epoch": 0.05586127276030404, + "grad_norm": 0.8451496362686157, + "learning_rate": 0.00019847615277319362, + "loss": 0.0637, + "step": 2383 + }, + { + "epoch": 0.05588471433510903, + "grad_norm": 5.812916278839111, + "learning_rate": 0.00019847487145655763, + "loss": 0.9961, + "step": 2384 + }, + { + "epoch": 0.05590815590991403, + "grad_norm": 2.899916172027588, + "learning_rate": 0.00019847358960559423, + "loss": 0.6308, + "step": 2385 + }, + { + "epoch": 0.05593159748471902, + "grad_norm": 3.187938928604126, + "learning_rate": 0.00019847230722031036, + "loss": 0.3508, + "step": 2386 + }, + { + "epoch": 0.055955039059524016, + "grad_norm": 3.8217849731445312, + "learning_rate": 0.00019847102430071298, + "loss": 0.7016, + "step": 2387 + }, + { + "epoch": 0.05597848063432902, + "grad_norm": 4.562125205993652, + "learning_rate": 0.00019846974084680908, + "loss": 0.5013, + "step": 2388 + }, + { + "epoch": 0.05600192220913401, + "grad_norm": 1.9017009735107422, + "learning_rate": 0.00019846845685860556, + "loss": 1.4579, + "step": 2389 + }, + { + "epoch": 0.056025363783939006, + "grad_norm": 3.751936674118042, + "learning_rate": 0.00019846717233610944, + "loss": 0.6318, + "step": 2390 + }, + { + "epoch": 0.056048805358744, + "grad_norm": 3.8351926803588867, + "learning_rate": 0.00019846588727932764, + "loss": 0.5828, + "step": 2391 + }, + { + "epoch": 0.056072246933548994, + "grad_norm": 5.050808906555176, + "learning_rate": 0.00019846460168826722, + "loss": 0.3797, + "step": 2392 + }, + { + "epoch": 0.05609568850835399, + "grad_norm": 4.385185718536377, + "learning_rate": 0.0001984633155629351, + "loss": 0.7533, + "step": 2393 + }, + { + "epoch": 0.05611913008315899, + "grad_norm": 5.193116664886475, + "learning_rate": 0.00019846202890333822, + "loss": 0.8554, + "step": 2394 + }, + { + "epoch": 0.056142571657963984, + "grad_norm": 2.2810263633728027, + "learning_rate": 0.00019846074170948362, + "loss": 0.4096, + "step": 2395 + }, + { + "epoch": 0.05616601323276898, + "grad_norm": 2.7166683673858643, + "learning_rate": 0.0001984594539813783, + "loss": 0.499, + "step": 2396 + }, + { + "epoch": 0.05618945480757397, + "grad_norm": 3.279418468475342, + "learning_rate": 0.00019845816571902915, + "loss": 0.6087, + "step": 2397 + }, + { + "epoch": 0.05621289638237897, + "grad_norm": 1.2650843858718872, + "learning_rate": 0.00019845687692244328, + "loss": 0.1496, + "step": 2398 + }, + { + "epoch": 0.05623633795718396, + "grad_norm": 1.094987154006958, + "learning_rate": 0.00019845558759162759, + "loss": 0.1234, + "step": 2399 + }, + { + "epoch": 0.056259779531988956, + "grad_norm": 2.3562710285186768, + "learning_rate": 0.00019845429772658913, + "loss": 0.4507, + "step": 2400 + }, + { + "epoch": 0.05628322110679396, + "grad_norm": 7.393082618713379, + "learning_rate": 0.0001984530073273349, + "loss": 0.9225, + "step": 2401 + }, + { + "epoch": 0.05630666268159895, + "grad_norm": 2.891610622406006, + "learning_rate": 0.00019845171639387187, + "loss": 0.4557, + "step": 2402 + }, + { + "epoch": 0.056330104256403946, + "grad_norm": 2.369347333908081, + "learning_rate": 0.00019845042492620704, + "loss": 0.5745, + "step": 2403 + }, + { + "epoch": 0.05635354583120894, + "grad_norm": 4.553523540496826, + "learning_rate": 0.00019844913292434746, + "loss": 0.8133, + "step": 2404 + }, + { + "epoch": 0.056376987406013934, + "grad_norm": 1.4001604318618774, + "learning_rate": 0.0001984478403883001, + "loss": 0.2834, + "step": 2405 + }, + { + "epoch": 0.05640042898081893, + "grad_norm": 2.812049150466919, + "learning_rate": 0.000198446547318072, + "loss": 0.6144, + "step": 2406 + }, + { + "epoch": 0.05642387055562393, + "grad_norm": 2.053309917449951, + "learning_rate": 0.0001984452537136702, + "loss": 0.2674, + "step": 2407 + }, + { + "epoch": 0.056447312130428924, + "grad_norm": 4.50997257232666, + "learning_rate": 0.0001984439595751016, + "loss": 0.6821, + "step": 2408 + }, + { + "epoch": 0.05647075370523392, + "grad_norm": 3.2830469608306885, + "learning_rate": 0.0001984426649023734, + "loss": 0.5776, + "step": 2409 + }, + { + "epoch": 0.05649419528003891, + "grad_norm": 3.0662214756011963, + "learning_rate": 0.0001984413696954925, + "loss": 0.3546, + "step": 2410 + }, + { + "epoch": 0.05651763685484391, + "grad_norm": 3.8332130908966064, + "learning_rate": 0.00019844007395446593, + "loss": 0.5242, + "step": 2411 + }, + { + "epoch": 0.0565410784296489, + "grad_norm": 1.048142910003662, + "learning_rate": 0.00019843877767930076, + "loss": 0.1445, + "step": 2412 + }, + { + "epoch": 0.0565645200044539, + "grad_norm": 0.8391729593276978, + "learning_rate": 0.00019843748087000404, + "loss": 0.0749, + "step": 2413 + }, + { + "epoch": 0.0565879615792589, + "grad_norm": 2.9812819957733154, + "learning_rate": 0.00019843618352658276, + "loss": 1.6025, + "step": 2414 + }, + { + "epoch": 0.05661140315406389, + "grad_norm": 3.6503119468688965, + "learning_rate": 0.000198434885649044, + "loss": 0.4573, + "step": 2415 + }, + { + "epoch": 0.056634844728868886, + "grad_norm": 3.148831605911255, + "learning_rate": 0.00019843358723739475, + "loss": 0.8657, + "step": 2416 + }, + { + "epoch": 0.05665828630367388, + "grad_norm": 2.6611685752868652, + "learning_rate": 0.00019843228829164212, + "loss": 0.2984, + "step": 2417 + }, + { + "epoch": 0.056681727878478874, + "grad_norm": 3.0721945762634277, + "learning_rate": 0.00019843098881179312, + "loss": 0.3621, + "step": 2418 + }, + { + "epoch": 0.05670516945328387, + "grad_norm": 1.6624643802642822, + "learning_rate": 0.0001984296887978548, + "loss": 0.1101, + "step": 2419 + }, + { + "epoch": 0.05672861102808887, + "grad_norm": 4.106475353240967, + "learning_rate": 0.0001984283882498342, + "loss": 2.0309, + "step": 2420 + }, + { + "epoch": 0.056752052602893864, + "grad_norm": 4.880541801452637, + "learning_rate": 0.00019842708716773843, + "loss": 0.9287, + "step": 2421 + }, + { + "epoch": 0.05677549417769886, + "grad_norm": 4.950517654418945, + "learning_rate": 0.0001984257855515745, + "loss": 1.0003, + "step": 2422 + }, + { + "epoch": 0.05679893575250385, + "grad_norm": 0.805091142654419, + "learning_rate": 0.00019842448340134952, + "loss": 0.1705, + "step": 2423 + }, + { + "epoch": 0.05682237732730885, + "grad_norm": 1.0719783306121826, + "learning_rate": 0.0001984231807170705, + "loss": 0.6355, + "step": 2424 + }, + { + "epoch": 0.05684581890211384, + "grad_norm": 2.9215593338012695, + "learning_rate": 0.00019842187749874451, + "loss": 0.4229, + "step": 2425 + }, + { + "epoch": 0.05686926047691884, + "grad_norm": 3.333462953567505, + "learning_rate": 0.0001984205737463787, + "loss": 0.5159, + "step": 2426 + }, + { + "epoch": 0.05689270205172384, + "grad_norm": 4.05152702331543, + "learning_rate": 0.00019841926945998007, + "loss": 0.6762, + "step": 2427 + }, + { + "epoch": 0.05691614362652883, + "grad_norm": 3.942709445953369, + "learning_rate": 0.00019841796463955567, + "loss": 0.9034, + "step": 2428 + }, + { + "epoch": 0.056939585201333826, + "grad_norm": 7.020834922790527, + "learning_rate": 0.0001984166592851127, + "loss": 0.7102, + "step": 2429 + }, + { + "epoch": 0.05696302677613882, + "grad_norm": 2.5862808227539062, + "learning_rate": 0.00019841535339665815, + "loss": 0.248, + "step": 2430 + }, + { + "epoch": 0.056986468350943814, + "grad_norm": 1.5248891115188599, + "learning_rate": 0.00019841404697419908, + "loss": 0.2102, + "step": 2431 + }, + { + "epoch": 0.05700990992574881, + "grad_norm": 1.508142352104187, + "learning_rate": 0.00019841274001774266, + "loss": 0.2552, + "step": 2432 + }, + { + "epoch": 0.05703335150055381, + "grad_norm": 1.0925204753875732, + "learning_rate": 0.00019841143252729592, + "loss": 0.1302, + "step": 2433 + }, + { + "epoch": 0.057056793075358804, + "grad_norm": 0.5092865228652954, + "learning_rate": 0.00019841012450286602, + "loss": 0.0892, + "step": 2434 + }, + { + "epoch": 0.0570802346501638, + "grad_norm": 1.4060770273208618, + "learning_rate": 0.00019840881594446, + "loss": 0.2432, + "step": 2435 + }, + { + "epoch": 0.05710367622496879, + "grad_norm": 1.3904392719268799, + "learning_rate": 0.000198407506852085, + "loss": 0.1685, + "step": 2436 + }, + { + "epoch": 0.05712711779977379, + "grad_norm": 1.2677183151245117, + "learning_rate": 0.00019840619722574806, + "loss": 0.2995, + "step": 2437 + }, + { + "epoch": 0.05715055937457878, + "grad_norm": 3.188143014907837, + "learning_rate": 0.00019840488706545633, + "loss": 0.6944, + "step": 2438 + }, + { + "epoch": 0.05717400094938378, + "grad_norm": 1.7454783916473389, + "learning_rate": 0.0001984035763712169, + "loss": 1.4113, + "step": 2439 + }, + { + "epoch": 0.05719744252418878, + "grad_norm": 3.939624786376953, + "learning_rate": 0.00019840226514303695, + "loss": 0.9418, + "step": 2440 + }, + { + "epoch": 0.05722088409899377, + "grad_norm": 3.4657223224639893, + "learning_rate": 0.0001984009533809235, + "loss": 0.7343, + "step": 2441 + }, + { + "epoch": 0.057244325673798765, + "grad_norm": 2.649491786956787, + "learning_rate": 0.00019839964108488376, + "loss": 0.2673, + "step": 2442 + }, + { + "epoch": 0.05726776724860376, + "grad_norm": 2.7261157035827637, + "learning_rate": 0.00019839832825492475, + "loss": 0.4978, + "step": 2443 + }, + { + "epoch": 0.057291208823408754, + "grad_norm": 3.196621894836426, + "learning_rate": 0.00019839701489105367, + "loss": 0.6037, + "step": 2444 + }, + { + "epoch": 0.057314650398213755, + "grad_norm": 1.4878119230270386, + "learning_rate": 0.0001983957009932776, + "loss": 0.1497, + "step": 2445 + }, + { + "epoch": 0.05733809197301875, + "grad_norm": 3.3791069984436035, + "learning_rate": 0.00019839438656160373, + "loss": 0.4679, + "step": 2446 + }, + { + "epoch": 0.057361533547823744, + "grad_norm": 0.7424440383911133, + "learning_rate": 0.0001983930715960391, + "loss": 0.0842, + "step": 2447 + }, + { + "epoch": 0.05738497512262874, + "grad_norm": 2.9404959678649902, + "learning_rate": 0.00019839175609659093, + "loss": 0.4361, + "step": 2448 + }, + { + "epoch": 0.05740841669743373, + "grad_norm": 4.403658866882324, + "learning_rate": 0.00019839044006326632, + "loss": 0.763, + "step": 2449 + }, + { + "epoch": 0.05743185827223873, + "grad_norm": 4.17379903793335, + "learning_rate": 0.0001983891234960724, + "loss": 0.5187, + "step": 2450 + }, + { + "epoch": 0.05745529984704372, + "grad_norm": 3.167968988418579, + "learning_rate": 0.00019838780639501636, + "loss": 0.4916, + "step": 2451 + }, + { + "epoch": 0.05747874142184872, + "grad_norm": 5.784088611602783, + "learning_rate": 0.0001983864887601053, + "loss": 0.6261, + "step": 2452 + }, + { + "epoch": 0.05750218299665372, + "grad_norm": 1.7633564472198486, + "learning_rate": 0.0001983851705913464, + "loss": 0.3653, + "step": 2453 + }, + { + "epoch": 0.05752562457145871, + "grad_norm": 2.9395554065704346, + "learning_rate": 0.00019838385188874677, + "loss": 0.5764, + "step": 2454 + }, + { + "epoch": 0.057549066146263705, + "grad_norm": 3.443956136703491, + "learning_rate": 0.00019838253265231363, + "loss": 0.8233, + "step": 2455 + }, + { + "epoch": 0.0575725077210687, + "grad_norm": 3.610663890838623, + "learning_rate": 0.00019838121288205405, + "loss": 0.7409, + "step": 2456 + }, + { + "epoch": 0.057595949295873694, + "grad_norm": 2.8560101985931396, + "learning_rate": 0.00019837989257797526, + "loss": 0.3478, + "step": 2457 + }, + { + "epoch": 0.057619390870678695, + "grad_norm": 2.9126274585723877, + "learning_rate": 0.00019837857174008443, + "loss": 0.4593, + "step": 2458 + }, + { + "epoch": 0.05764283244548369, + "grad_norm": 5.252607822418213, + "learning_rate": 0.0001983772503683887, + "loss": 0.6616, + "step": 2459 + }, + { + "epoch": 0.057666274020288684, + "grad_norm": 1.556119680404663, + "learning_rate": 0.0001983759284628952, + "loss": 0.2917, + "step": 2460 + }, + { + "epoch": 0.05768971559509368, + "grad_norm": 1.1253559589385986, + "learning_rate": 0.0001983746060236112, + "loss": 0.1889, + "step": 2461 + }, + { + "epoch": 0.05771315716989867, + "grad_norm": 0.981217086315155, + "learning_rate": 0.0001983732830505438, + "loss": 0.1268, + "step": 2462 + }, + { + "epoch": 0.05773659874470367, + "grad_norm": 4.680423736572266, + "learning_rate": 0.0001983719595437002, + "loss": 0.7495, + "step": 2463 + }, + { + "epoch": 0.05776004031950866, + "grad_norm": 4.18715763092041, + "learning_rate": 0.00019837063550308759, + "loss": 0.388, + "step": 2464 + }, + { + "epoch": 0.05778348189431366, + "grad_norm": 1.2031223773956299, + "learning_rate": 0.00019836931092871314, + "loss": 0.2298, + "step": 2465 + }, + { + "epoch": 0.05780692346911866, + "grad_norm": 4.415149211883545, + "learning_rate": 0.000198367985820584, + "loss": 0.6874, + "step": 2466 + }, + { + "epoch": 0.05783036504392365, + "grad_norm": 1.083827257156372, + "learning_rate": 0.0001983666601787075, + "loss": 0.149, + "step": 2467 + }, + { + "epoch": 0.057853806618728645, + "grad_norm": 3.533656597137451, + "learning_rate": 0.00019836533400309066, + "loss": 0.7495, + "step": 2468 + }, + { + "epoch": 0.05787724819353364, + "grad_norm": 0.9895036220550537, + "learning_rate": 0.00019836400729374075, + "loss": 0.1675, + "step": 2469 + }, + { + "epoch": 0.057900689768338634, + "grad_norm": 4.212669372558594, + "learning_rate": 0.000198362680050665, + "loss": 0.8387, + "step": 2470 + }, + { + "epoch": 0.057924131343143635, + "grad_norm": 3.1826870441436768, + "learning_rate": 0.0001983613522738706, + "loss": 0.4916, + "step": 2471 + }, + { + "epoch": 0.05794757291794863, + "grad_norm": 2.740722179412842, + "learning_rate": 0.0001983600239633647, + "loss": 0.4543, + "step": 2472 + }, + { + "epoch": 0.057971014492753624, + "grad_norm": 4.991090297698975, + "learning_rate": 0.00019835869511915455, + "loss": 0.6569, + "step": 2473 + }, + { + "epoch": 0.05799445606755862, + "grad_norm": 5.367294788360596, + "learning_rate": 0.00019835736574124737, + "loss": 0.5843, + "step": 2474 + }, + { + "epoch": 0.05801789764236361, + "grad_norm": 5.408801078796387, + "learning_rate": 0.00019835603582965035, + "loss": 0.6918, + "step": 2475 + }, + { + "epoch": 0.05804133921716861, + "grad_norm": 4.8743391036987305, + "learning_rate": 0.0001983547053843707, + "loss": 0.5006, + "step": 2476 + }, + { + "epoch": 0.05806478079197361, + "grad_norm": 2.8016839027404785, + "learning_rate": 0.00019835337440541568, + "loss": 0.5016, + "step": 2477 + }, + { + "epoch": 0.0580882223667786, + "grad_norm": 2.617351770401001, + "learning_rate": 0.00019835204289279248, + "loss": 0.4223, + "step": 2478 + }, + { + "epoch": 0.0581116639415836, + "grad_norm": 3.6910953521728516, + "learning_rate": 0.0001983507108465083, + "loss": 0.5319, + "step": 2479 + }, + { + "epoch": 0.05813510551638859, + "grad_norm": 3.6320931911468506, + "learning_rate": 0.00019834937826657045, + "loss": 0.345, + "step": 2480 + }, + { + "epoch": 0.058158547091193585, + "grad_norm": 0.7668538093566895, + "learning_rate": 0.00019834804515298605, + "loss": 0.0744, + "step": 2481 + }, + { + "epoch": 0.05818198866599858, + "grad_norm": 4.194398403167725, + "learning_rate": 0.00019834671150576242, + "loss": 0.5449, + "step": 2482 + }, + { + "epoch": 0.058205430240803574, + "grad_norm": 3.830653429031372, + "learning_rate": 0.00019834537732490678, + "loss": 0.3379, + "step": 2483 + }, + { + "epoch": 0.058228871815608575, + "grad_norm": 4.429388046264648, + "learning_rate": 0.00019834404261042635, + "loss": 0.8063, + "step": 2484 + }, + { + "epoch": 0.05825231339041357, + "grad_norm": 4.782171249389648, + "learning_rate": 0.00019834270736232839, + "loss": 0.7166, + "step": 2485 + }, + { + "epoch": 0.058275754965218564, + "grad_norm": 3.245816707611084, + "learning_rate": 0.0001983413715806201, + "loss": 0.5423, + "step": 2486 + }, + { + "epoch": 0.05829919654002356, + "grad_norm": 4.134199142456055, + "learning_rate": 0.00019834003526530877, + "loss": 0.5238, + "step": 2487 + }, + { + "epoch": 0.05832263811482855, + "grad_norm": 7.563695907592773, + "learning_rate": 0.00019833869841640167, + "loss": 0.9251, + "step": 2488 + }, + { + "epoch": 0.05834607968963355, + "grad_norm": 1.4248287677764893, + "learning_rate": 0.000198337361033906, + "loss": 0.0857, + "step": 2489 + }, + { + "epoch": 0.05836952126443855, + "grad_norm": 2.165041446685791, + "learning_rate": 0.00019833602311782907, + "loss": 0.2369, + "step": 2490 + }, + { + "epoch": 0.05839296283924354, + "grad_norm": 3.3313536643981934, + "learning_rate": 0.00019833468466817807, + "loss": 0.5316, + "step": 2491 + }, + { + "epoch": 0.05841640441404854, + "grad_norm": 1.2660059928894043, + "learning_rate": 0.00019833334568496037, + "loss": 0.2172, + "step": 2492 + }, + { + "epoch": 0.05843984598885353, + "grad_norm": 5.048691749572754, + "learning_rate": 0.00019833200616818314, + "loss": 0.8665, + "step": 2493 + }, + { + "epoch": 0.058463287563658525, + "grad_norm": 8.348846435546875, + "learning_rate": 0.00019833066611785367, + "loss": 0.7064, + "step": 2494 + }, + { + "epoch": 0.05848672913846352, + "grad_norm": 4.369143962860107, + "learning_rate": 0.00019832932553397924, + "loss": 0.8092, + "step": 2495 + }, + { + "epoch": 0.058510170713268514, + "grad_norm": 4.8156962394714355, + "learning_rate": 0.00019832798441656713, + "loss": 0.6225, + "step": 2496 + }, + { + "epoch": 0.058533612288073515, + "grad_norm": 6.521709442138672, + "learning_rate": 0.00019832664276562463, + "loss": 1.5231, + "step": 2497 + }, + { + "epoch": 0.05855705386287851, + "grad_norm": 4.553813457489014, + "learning_rate": 0.00019832530058115899, + "loss": 0.6358, + "step": 2498 + }, + { + "epoch": 0.058580495437683504, + "grad_norm": 0.6071985960006714, + "learning_rate": 0.00019832395786317748, + "loss": 0.0624, + "step": 2499 + }, + { + "epoch": 0.0586039370124885, + "grad_norm": 3.295705795288086, + "learning_rate": 0.00019832261461168743, + "loss": 0.4522, + "step": 2500 + }, + { + "epoch": 0.05862737858729349, + "grad_norm": 3.698446273803711, + "learning_rate": 0.0001983212708266961, + "loss": 0.805, + "step": 2501 + }, + { + "epoch": 0.05865082016209849, + "grad_norm": 1.3746200799942017, + "learning_rate": 0.00019831992650821083, + "loss": 0.2621, + "step": 2502 + }, + { + "epoch": 0.05867426173690349, + "grad_norm": 3.067391872406006, + "learning_rate": 0.00019831858165623883, + "loss": 0.4303, + "step": 2503 + }, + { + "epoch": 0.05869770331170848, + "grad_norm": 1.7861844301223755, + "learning_rate": 0.00019831723627078746, + "loss": 0.2824, + "step": 2504 + }, + { + "epoch": 0.05872114488651348, + "grad_norm": 3.479088306427002, + "learning_rate": 0.000198315890351864, + "loss": 0.4909, + "step": 2505 + }, + { + "epoch": 0.05874458646131847, + "grad_norm": 3.943603277206421, + "learning_rate": 0.00019831454389947575, + "loss": 0.9128, + "step": 2506 + }, + { + "epoch": 0.058768028036123465, + "grad_norm": 8.000354766845703, + "learning_rate": 0.00019831319691363002, + "loss": 0.541, + "step": 2507 + }, + { + "epoch": 0.05879146961092846, + "grad_norm": 6.008207321166992, + "learning_rate": 0.00019831184939433412, + "loss": 0.9632, + "step": 2508 + }, + { + "epoch": 0.05881491118573346, + "grad_norm": 2.540651559829712, + "learning_rate": 0.00019831050134159534, + "loss": 0.3135, + "step": 2509 + }, + { + "epoch": 0.058838352760538455, + "grad_norm": 5.47852897644043, + "learning_rate": 0.00019830915275542105, + "loss": 0.5468, + "step": 2510 + }, + { + "epoch": 0.05886179433534345, + "grad_norm": 0.9875124096870422, + "learning_rate": 0.00019830780363581853, + "loss": 0.1384, + "step": 2511 + }, + { + "epoch": 0.058885235910148444, + "grad_norm": 0.8476619124412537, + "learning_rate": 0.0001983064539827951, + "loss": 0.1484, + "step": 2512 + }, + { + "epoch": 0.05890867748495344, + "grad_norm": 3.248272657394409, + "learning_rate": 0.00019830510379635808, + "loss": 0.3989, + "step": 2513 + }, + { + "epoch": 0.05893211905975843, + "grad_norm": 3.61053729057312, + "learning_rate": 0.0001983037530765148, + "loss": 0.4756, + "step": 2514 + }, + { + "epoch": 0.05895556063456343, + "grad_norm": 2.2824645042419434, + "learning_rate": 0.0001983024018232726, + "loss": 1.727, + "step": 2515 + }, + { + "epoch": 0.05897900220936843, + "grad_norm": 3.73927640914917, + "learning_rate": 0.00019830105003663877, + "loss": 0.5737, + "step": 2516 + }, + { + "epoch": 0.05900244378417342, + "grad_norm": 4.847911357879639, + "learning_rate": 0.0001982996977166207, + "loss": 0.7317, + "step": 2517 + }, + { + "epoch": 0.05902588535897842, + "grad_norm": 15.87205982208252, + "learning_rate": 0.00019829834486322575, + "loss": 0.8893, + "step": 2518 + }, + { + "epoch": 0.05904932693378341, + "grad_norm": 3.4882802963256836, + "learning_rate": 0.00019829699147646117, + "loss": 0.5196, + "step": 2519 + }, + { + "epoch": 0.059072768508588405, + "grad_norm": 0.8563350439071655, + "learning_rate": 0.00019829563755633436, + "loss": 0.0865, + "step": 2520 + }, + { + "epoch": 0.0590962100833934, + "grad_norm": 2.6608965396881104, + "learning_rate": 0.00019829428310285264, + "loss": 0.6282, + "step": 2521 + }, + { + "epoch": 0.0591196516581984, + "grad_norm": 3.569288730621338, + "learning_rate": 0.0001982929281160234, + "loss": 0.5918, + "step": 2522 + }, + { + "epoch": 0.059143093233003395, + "grad_norm": 3.2925703525543213, + "learning_rate": 0.00019829157259585397, + "loss": 0.6047, + "step": 2523 + }, + { + "epoch": 0.05916653480780839, + "grad_norm": 3.2503249645233154, + "learning_rate": 0.0001982902165423517, + "loss": 0.2249, + "step": 2524 + }, + { + "epoch": 0.059189976382613384, + "grad_norm": 2.184866189956665, + "learning_rate": 0.00019828885995552394, + "loss": 1.4652, + "step": 2525 + }, + { + "epoch": 0.05921341795741838, + "grad_norm": 3.811246871948242, + "learning_rate": 0.00019828750283537808, + "loss": 0.62, + "step": 2526 + }, + { + "epoch": 0.05923685953222337, + "grad_norm": 3.1616928577423096, + "learning_rate": 0.00019828614518192144, + "loss": 0.5463, + "step": 2527 + }, + { + "epoch": 0.059260301107028374, + "grad_norm": 2.4077086448669434, + "learning_rate": 0.00019828478699516144, + "loss": 0.34, + "step": 2528 + }, + { + "epoch": 0.05928374268183337, + "grad_norm": 0.68716961145401, + "learning_rate": 0.0001982834282751054, + "loss": 0.116, + "step": 2529 + }, + { + "epoch": 0.05930718425663836, + "grad_norm": 1.3642547130584717, + "learning_rate": 0.00019828206902176073, + "loss": 0.2853, + "step": 2530 + }, + { + "epoch": 0.05933062583144336, + "grad_norm": 1.5392407178878784, + "learning_rate": 0.00019828070923513477, + "loss": 0.3042, + "step": 2531 + }, + { + "epoch": 0.05935406740624835, + "grad_norm": 3.485649585723877, + "learning_rate": 0.00019827934891523494, + "loss": 0.6051, + "step": 2532 + }, + { + "epoch": 0.059377508981053345, + "grad_norm": 3.2587618827819824, + "learning_rate": 0.0001982779880620686, + "loss": 0.6706, + "step": 2533 + }, + { + "epoch": 0.05940095055585834, + "grad_norm": 2.5210139751434326, + "learning_rate": 0.0001982766266756431, + "loss": 0.4806, + "step": 2534 + }, + { + "epoch": 0.05942439213066334, + "grad_norm": 2.322009801864624, + "learning_rate": 0.00019827526475596588, + "loss": 0.3576, + "step": 2535 + }, + { + "epoch": 0.059447833705468335, + "grad_norm": 3.2899904251098633, + "learning_rate": 0.00019827390230304435, + "loss": 0.7479, + "step": 2536 + }, + { + "epoch": 0.05947127528027333, + "grad_norm": 3.52691912651062, + "learning_rate": 0.00019827253931688582, + "loss": 0.5639, + "step": 2537 + }, + { + "epoch": 0.059494716855078324, + "grad_norm": 3.261059522628784, + "learning_rate": 0.00019827117579749773, + "loss": 0.4003, + "step": 2538 + }, + { + "epoch": 0.05951815842988332, + "grad_norm": 2.7643799781799316, + "learning_rate": 0.00019826981174488748, + "loss": 1.6106, + "step": 2539 + }, + { + "epoch": 0.05954160000468831, + "grad_norm": 2.1625125408172607, + "learning_rate": 0.00019826844715906247, + "loss": 0.3166, + "step": 2540 + }, + { + "epoch": 0.059565041579493314, + "grad_norm": 3.949794292449951, + "learning_rate": 0.00019826708204003011, + "loss": 0.4747, + "step": 2541 + }, + { + "epoch": 0.05958848315429831, + "grad_norm": 1.0836596488952637, + "learning_rate": 0.0001982657163877978, + "loss": 0.0875, + "step": 2542 + }, + { + "epoch": 0.0596119247291033, + "grad_norm": 0.9202995300292969, + "learning_rate": 0.00019826435020237293, + "loss": 0.1112, + "step": 2543 + }, + { + "epoch": 0.059635366303908297, + "grad_norm": 6.822673320770264, + "learning_rate": 0.00019826298348376295, + "loss": 0.896, + "step": 2544 + }, + { + "epoch": 0.05965880787871329, + "grad_norm": 3.6534571647644043, + "learning_rate": 0.00019826161623197526, + "loss": 0.5728, + "step": 2545 + }, + { + "epoch": 0.059682249453518285, + "grad_norm": 5.242239475250244, + "learning_rate": 0.0001982602484470173, + "loss": 0.6139, + "step": 2546 + }, + { + "epoch": 0.05970569102832328, + "grad_norm": 4.6888427734375, + "learning_rate": 0.00019825888012889642, + "loss": 0.535, + "step": 2547 + }, + { + "epoch": 0.05972913260312828, + "grad_norm": 2.861704111099243, + "learning_rate": 0.00019825751127762012, + "loss": 0.6368, + "step": 2548 + }, + { + "epoch": 0.059752574177933275, + "grad_norm": 3.019803047180176, + "learning_rate": 0.0001982561418931958, + "loss": 0.3035, + "step": 2549 + }, + { + "epoch": 0.05977601575273827, + "grad_norm": 0.7816648483276367, + "learning_rate": 0.00019825477197563088, + "loss": 0.0974, + "step": 2550 + }, + { + "epoch": 0.059799457327543264, + "grad_norm": 3.0429232120513916, + "learning_rate": 0.0001982534015249328, + "loss": 0.545, + "step": 2551 + }, + { + "epoch": 0.05982289890234826, + "grad_norm": 3.5658912658691406, + "learning_rate": 0.00019825203054110903, + "loss": 0.5895, + "step": 2552 + }, + { + "epoch": 0.05984634047715325, + "grad_norm": 5.5616607666015625, + "learning_rate": 0.00019825065902416696, + "loss": 0.6148, + "step": 2553 + }, + { + "epoch": 0.059869782051958254, + "grad_norm": 4.257177829742432, + "learning_rate": 0.00019824928697411407, + "loss": 0.8674, + "step": 2554 + }, + { + "epoch": 0.05989322362676325, + "grad_norm": 2.1640024185180664, + "learning_rate": 0.00019824791439095778, + "loss": 0.1498, + "step": 2555 + }, + { + "epoch": 0.05991666520156824, + "grad_norm": 1.8893073797225952, + "learning_rate": 0.00019824654127470552, + "loss": 0.4178, + "step": 2556 + }, + { + "epoch": 0.059940106776373236, + "grad_norm": 0.7229540944099426, + "learning_rate": 0.00019824516762536478, + "loss": 0.1351, + "step": 2557 + }, + { + "epoch": 0.05996354835117823, + "grad_norm": 3.9254698753356934, + "learning_rate": 0.00019824379344294299, + "loss": 0.678, + "step": 2558 + }, + { + "epoch": 0.059986989925983225, + "grad_norm": 4.8578572273254395, + "learning_rate": 0.00019824241872744763, + "loss": 0.9319, + "step": 2559 + }, + { + "epoch": 0.060010431500788226, + "grad_norm": 3.6958093643188477, + "learning_rate": 0.00019824104347888612, + "loss": 0.4989, + "step": 2560 + }, + { + "epoch": 0.06003387307559322, + "grad_norm": 1.2095853090286255, + "learning_rate": 0.00019823966769726598, + "loss": 0.1761, + "step": 2561 + }, + { + "epoch": 0.060057314650398215, + "grad_norm": 4.374950885772705, + "learning_rate": 0.00019823829138259462, + "loss": 0.7393, + "step": 2562 + }, + { + "epoch": 0.06008075622520321, + "grad_norm": 1.0420325994491577, + "learning_rate": 0.00019823691453487954, + "loss": 0.1315, + "step": 2563 + }, + { + "epoch": 0.060104197800008204, + "grad_norm": 1.8852990865707397, + "learning_rate": 0.0001982355371541282, + "loss": 0.225, + "step": 2564 + }, + { + "epoch": 0.0601276393748132, + "grad_norm": 1.0474774837493896, + "learning_rate": 0.00019823415924034805, + "loss": 0.2346, + "step": 2565 + }, + { + "epoch": 0.06015108094961819, + "grad_norm": 4.545724868774414, + "learning_rate": 0.00019823278079354658, + "loss": 0.7012, + "step": 2566 + }, + { + "epoch": 0.060174522524423193, + "grad_norm": 3.266477584838867, + "learning_rate": 0.00019823140181373132, + "loss": 0.3395, + "step": 2567 + }, + { + "epoch": 0.06019796409922819, + "grad_norm": 4.096229076385498, + "learning_rate": 0.0001982300223009097, + "loss": 0.5501, + "step": 2568 + }, + { + "epoch": 0.06022140567403318, + "grad_norm": 3.893868923187256, + "learning_rate": 0.00019822864225508923, + "loss": 0.6249, + "step": 2569 + }, + { + "epoch": 0.060244847248838176, + "grad_norm": 3.099820375442505, + "learning_rate": 0.00019822726167627734, + "loss": 0.5206, + "step": 2570 + }, + { + "epoch": 0.06026828882364317, + "grad_norm": 3.033341646194458, + "learning_rate": 0.0001982258805644816, + "loss": 0.7742, + "step": 2571 + }, + { + "epoch": 0.060291730398448165, + "grad_norm": 7.566232204437256, + "learning_rate": 0.00019822449891970947, + "loss": 0.5835, + "step": 2572 + }, + { + "epoch": 0.060315171973253166, + "grad_norm": 5.042271137237549, + "learning_rate": 0.00019822311674196844, + "loss": 0.741, + "step": 2573 + }, + { + "epoch": 0.06033861354805816, + "grad_norm": 0.9126322865486145, + "learning_rate": 0.00019822173403126603, + "loss": 0.1401, + "step": 2574 + }, + { + "epoch": 0.060362055122863155, + "grad_norm": 4.724917411804199, + "learning_rate": 0.0001982203507876097, + "loss": 0.7384, + "step": 2575 + }, + { + "epoch": 0.06038549669766815, + "grad_norm": 4.530577659606934, + "learning_rate": 0.00019821896701100703, + "loss": 0.6085, + "step": 2576 + }, + { + "epoch": 0.060408938272473144, + "grad_norm": 3.7867791652679443, + "learning_rate": 0.00019821758270146544, + "loss": 0.5176, + "step": 2577 + }, + { + "epoch": 0.06043237984727814, + "grad_norm": 2.959221601486206, + "learning_rate": 0.0001982161978589925, + "loss": 0.5677, + "step": 2578 + }, + { + "epoch": 0.06045582142208313, + "grad_norm": 4.993811130523682, + "learning_rate": 0.0001982148124835957, + "loss": 0.757, + "step": 2579 + }, + { + "epoch": 0.06047926299688813, + "grad_norm": 2.0361173152923584, + "learning_rate": 0.0001982134265752826, + "loss": 0.2666, + "step": 2580 + }, + { + "epoch": 0.06050270457169313, + "grad_norm": 3.9344563484191895, + "learning_rate": 0.00019821204013406062, + "loss": 0.7492, + "step": 2581 + }, + { + "epoch": 0.06052614614649812, + "grad_norm": 3.313668727874756, + "learning_rate": 0.00019821065315993742, + "loss": 0.4751, + "step": 2582 + }, + { + "epoch": 0.060549587721303116, + "grad_norm": 1.6158103942871094, + "learning_rate": 0.00019820926565292042, + "loss": 1.6791, + "step": 2583 + }, + { + "epoch": 0.06057302929610811, + "grad_norm": 1.027468204498291, + "learning_rate": 0.0001982078776130172, + "loss": 0.1472, + "step": 2584 + }, + { + "epoch": 0.060596470870913105, + "grad_norm": 0.9630619287490845, + "learning_rate": 0.00019820648904023523, + "loss": 0.1476, + "step": 2585 + }, + { + "epoch": 0.060619912445718106, + "grad_norm": 5.787491798400879, + "learning_rate": 0.00019820509993458214, + "loss": 0.9503, + "step": 2586 + }, + { + "epoch": 0.0606433540205231, + "grad_norm": 0.7456610798835754, + "learning_rate": 0.00019820371029606539, + "loss": 0.0767, + "step": 2587 + }, + { + "epoch": 0.060666795595328095, + "grad_norm": 4.380799293518066, + "learning_rate": 0.00019820232012469257, + "loss": 0.9904, + "step": 2588 + }, + { + "epoch": 0.06069023717013309, + "grad_norm": 2.7856175899505615, + "learning_rate": 0.00019820092942047116, + "loss": 0.6081, + "step": 2589 + }, + { + "epoch": 0.060713678744938084, + "grad_norm": 4.230140686035156, + "learning_rate": 0.00019819953818340879, + "loss": 0.4955, + "step": 2590 + }, + { + "epoch": 0.06073712031974308, + "grad_norm": 3.038571834564209, + "learning_rate": 0.00019819814641351292, + "loss": 0.6649, + "step": 2591 + }, + { + "epoch": 0.06076056189454808, + "grad_norm": 2.8972859382629395, + "learning_rate": 0.00019819675411079118, + "loss": 0.6295, + "step": 2592 + }, + { + "epoch": 0.06078400346935307, + "grad_norm": 6.40317964553833, + "learning_rate": 0.00019819536127525108, + "loss": 0.9567, + "step": 2593 + }, + { + "epoch": 0.06080744504415807, + "grad_norm": 2.6298160552978516, + "learning_rate": 0.0001981939679069002, + "loss": 0.4507, + "step": 2594 + }, + { + "epoch": 0.06083088661896306, + "grad_norm": 3.581512689590454, + "learning_rate": 0.00019819257400574607, + "loss": 0.5606, + "step": 2595 + }, + { + "epoch": 0.060854328193768056, + "grad_norm": 3.184875965118408, + "learning_rate": 0.0001981911795717963, + "loss": 0.6431, + "step": 2596 + }, + { + "epoch": 0.06087776976857305, + "grad_norm": 2.967846632003784, + "learning_rate": 0.0001981897846050584, + "loss": 0.5075, + "step": 2597 + }, + { + "epoch": 0.060901211343378045, + "grad_norm": 1.9988670349121094, + "learning_rate": 0.00019818838910554, + "loss": 1.8091, + "step": 2598 + }, + { + "epoch": 0.060924652918183046, + "grad_norm": 4.430943965911865, + "learning_rate": 0.0001981869930732486, + "loss": 0.8645, + "step": 2599 + }, + { + "epoch": 0.06094809449298804, + "grad_norm": 2.1300528049468994, + "learning_rate": 0.00019818559650819183, + "loss": 0.3576, + "step": 2600 + }, + { + "epoch": 0.060971536067793035, + "grad_norm": 4.082230567932129, + "learning_rate": 0.00019818419941037727, + "loss": 0.6078, + "step": 2601 + }, + { + "epoch": 0.06099497764259803, + "grad_norm": 5.6396331787109375, + "learning_rate": 0.00019818280177981245, + "loss": 1.1692, + "step": 2602 + }, + { + "epoch": 0.061018419217403023, + "grad_norm": 1.7372208833694458, + "learning_rate": 0.000198181403616505, + "loss": 0.3391, + "step": 2603 + }, + { + "epoch": 0.06104186079220802, + "grad_norm": 3.3802926540374756, + "learning_rate": 0.0001981800049204625, + "loss": 0.7428, + "step": 2604 + }, + { + "epoch": 0.06106530236701302, + "grad_norm": 3.640833854675293, + "learning_rate": 0.00019817860569169255, + "loss": 0.5997, + "step": 2605 + }, + { + "epoch": 0.06108874394181801, + "grad_norm": 2.66929292678833, + "learning_rate": 0.00019817720593020272, + "loss": 0.4629, + "step": 2606 + }, + { + "epoch": 0.06111218551662301, + "grad_norm": 4.496716022491455, + "learning_rate": 0.00019817580563600057, + "loss": 0.7907, + "step": 2607 + }, + { + "epoch": 0.061135627091428, + "grad_norm": 4.854788303375244, + "learning_rate": 0.0001981744048090938, + "loss": 1.0543, + "step": 2608 + }, + { + "epoch": 0.061159068666232996, + "grad_norm": 3.3835277557373047, + "learning_rate": 0.0001981730034494899, + "loss": 0.7877, + "step": 2609 + }, + { + "epoch": 0.06118251024103799, + "grad_norm": 4.105818271636963, + "learning_rate": 0.00019817160155719655, + "loss": 0.5789, + "step": 2610 + }, + { + "epoch": 0.061205951815842985, + "grad_norm": 1.0513358116149902, + "learning_rate": 0.0001981701991322213, + "loss": 0.2646, + "step": 2611 + }, + { + "epoch": 0.061229393390647986, + "grad_norm": 2.603337049484253, + "learning_rate": 0.00019816879617457183, + "loss": 0.5164, + "step": 2612 + }, + { + "epoch": 0.06125283496545298, + "grad_norm": 1.7730233669281006, + "learning_rate": 0.00019816739268425568, + "loss": 0.0798, + "step": 2613 + }, + { + "epoch": 0.061276276540257975, + "grad_norm": 3.6474087238311768, + "learning_rate": 0.0001981659886612805, + "loss": 0.7625, + "step": 2614 + }, + { + "epoch": 0.06129971811506297, + "grad_norm": 2.704896926879883, + "learning_rate": 0.0001981645841056539, + "loss": 0.5783, + "step": 2615 + }, + { + "epoch": 0.06132315968986796, + "grad_norm": 0.7543970346450806, + "learning_rate": 0.00019816317901738352, + "loss": 0.0786, + "step": 2616 + }, + { + "epoch": 0.06134660126467296, + "grad_norm": 3.658951997756958, + "learning_rate": 0.000198161773396477, + "loss": 0.6155, + "step": 2617 + }, + { + "epoch": 0.06137004283947796, + "grad_norm": 1.7189565896987915, + "learning_rate": 0.00019816036724294188, + "loss": 0.1801, + "step": 2618 + }, + { + "epoch": 0.06139348441428295, + "grad_norm": 4.891151428222656, + "learning_rate": 0.00019815896055678587, + "loss": 1.04, + "step": 2619 + }, + { + "epoch": 0.06141692598908795, + "grad_norm": 3.644812822341919, + "learning_rate": 0.00019815755333801657, + "loss": 0.7125, + "step": 2620 + }, + { + "epoch": 0.06144036756389294, + "grad_norm": 4.840481281280518, + "learning_rate": 0.00019815614558664165, + "loss": 0.9647, + "step": 2621 + }, + { + "epoch": 0.061463809138697936, + "grad_norm": 2.70853590965271, + "learning_rate": 0.00019815473730266873, + "loss": 0.4296, + "step": 2622 + }, + { + "epoch": 0.06148725071350293, + "grad_norm": 3.7486207485198975, + "learning_rate": 0.0001981533284861054, + "loss": 0.5719, + "step": 2623 + }, + { + "epoch": 0.06151069228830793, + "grad_norm": 3.817720890045166, + "learning_rate": 0.0001981519191369594, + "loss": 0.5092, + "step": 2624 + }, + { + "epoch": 0.061534133863112926, + "grad_norm": 2.3450284004211426, + "learning_rate": 0.0001981505092552383, + "loss": 0.3995, + "step": 2625 + }, + { + "epoch": 0.06155757543791792, + "grad_norm": 2.82865309715271, + "learning_rate": 0.00019814909884094978, + "loss": 0.2079, + "step": 2626 + }, + { + "epoch": 0.061581017012722915, + "grad_norm": 3.989802122116089, + "learning_rate": 0.0001981476878941015, + "loss": 0.8813, + "step": 2627 + }, + { + "epoch": 0.06160445858752791, + "grad_norm": 4.400240898132324, + "learning_rate": 0.0001981462764147011, + "loss": 0.9619, + "step": 2628 + }, + { + "epoch": 0.0616279001623329, + "grad_norm": 4.772131443023682, + "learning_rate": 0.00019814486440275625, + "loss": 0.9247, + "step": 2629 + }, + { + "epoch": 0.0616513417371379, + "grad_norm": 4.655982971191406, + "learning_rate": 0.00019814345185827458, + "loss": 1.1244, + "step": 2630 + }, + { + "epoch": 0.0616747833119429, + "grad_norm": 4.436686038970947, + "learning_rate": 0.00019814203878126382, + "loss": 0.7246, + "step": 2631 + }, + { + "epoch": 0.06169822488674789, + "grad_norm": 2.1310760974884033, + "learning_rate": 0.00019814062517173157, + "loss": 0.3047, + "step": 2632 + }, + { + "epoch": 0.06172166646155289, + "grad_norm": 4.1157026290893555, + "learning_rate": 0.00019813921102968554, + "loss": 0.6739, + "step": 2633 + }, + { + "epoch": 0.06174510803635788, + "grad_norm": 1.2241109609603882, + "learning_rate": 0.00019813779635513337, + "loss": 0.3667, + "step": 2634 + }, + { + "epoch": 0.061768549611162876, + "grad_norm": 3.920327663421631, + "learning_rate": 0.00019813638114808278, + "loss": 0.8136, + "step": 2635 + }, + { + "epoch": 0.06179199118596787, + "grad_norm": 3.123114585876465, + "learning_rate": 0.00019813496540854142, + "loss": 0.7347, + "step": 2636 + }, + { + "epoch": 0.06181543276077287, + "grad_norm": 3.2236690521240234, + "learning_rate": 0.000198133549136517, + "loss": 0.6173, + "step": 2637 + }, + { + "epoch": 0.061838874335577866, + "grad_norm": 2.4907360076904297, + "learning_rate": 0.00019813213233201715, + "loss": 0.4875, + "step": 2638 + }, + { + "epoch": 0.06186231591038286, + "grad_norm": 2.8484606742858887, + "learning_rate": 0.00019813071499504958, + "loss": 0.8265, + "step": 2639 + }, + { + "epoch": 0.061885757485187855, + "grad_norm": 4.501418590545654, + "learning_rate": 0.00019812929712562203, + "loss": 0.3605, + "step": 2640 + }, + { + "epoch": 0.06190919905999285, + "grad_norm": 4.248075485229492, + "learning_rate": 0.00019812787872374214, + "loss": 0.7282, + "step": 2641 + }, + { + "epoch": 0.06193264063479784, + "grad_norm": 8.008071899414062, + "learning_rate": 0.0001981264597894176, + "loss": 1.1676, + "step": 2642 + }, + { + "epoch": 0.06195608220960284, + "grad_norm": 4.045396327972412, + "learning_rate": 0.00019812504032265616, + "loss": 0.8713, + "step": 2643 + }, + { + "epoch": 0.06197952378440784, + "grad_norm": 4.1122918128967285, + "learning_rate": 0.0001981236203234655, + "loss": 1.0605, + "step": 2644 + }, + { + "epoch": 0.06200296535921283, + "grad_norm": 1.161489725112915, + "learning_rate": 0.0001981221997918533, + "loss": 0.1473, + "step": 2645 + }, + { + "epoch": 0.06202640693401783, + "grad_norm": 4.03037691116333, + "learning_rate": 0.0001981207787278273, + "loss": 0.9026, + "step": 2646 + }, + { + "epoch": 0.06204984850882282, + "grad_norm": 5.8806939125061035, + "learning_rate": 0.00019811935713139517, + "loss": 1.0431, + "step": 2647 + }, + { + "epoch": 0.062073290083627816, + "grad_norm": 5.241270065307617, + "learning_rate": 0.00019811793500256464, + "loss": 0.6925, + "step": 2648 + }, + { + "epoch": 0.06209673165843281, + "grad_norm": 3.687331199645996, + "learning_rate": 0.00019811651234134345, + "loss": 0.6679, + "step": 2649 + }, + { + "epoch": 0.06212017323323781, + "grad_norm": 3.912712574005127, + "learning_rate": 0.00019811508914773933, + "loss": 0.8863, + "step": 2650 + }, + { + "epoch": 0.062143614808042806, + "grad_norm": 3.842391014099121, + "learning_rate": 0.00019811366542175996, + "loss": 0.7605, + "step": 2651 + }, + { + "epoch": 0.0621670563828478, + "grad_norm": 4.4198527336120605, + "learning_rate": 0.00019811224116341307, + "loss": 0.8508, + "step": 2652 + }, + { + "epoch": 0.062190497957652795, + "grad_norm": 3.976943016052246, + "learning_rate": 0.00019811081637270643, + "loss": 0.9165, + "step": 2653 + }, + { + "epoch": 0.06221393953245779, + "grad_norm": 1.6970176696777344, + "learning_rate": 0.00019810939104964768, + "loss": 0.2338, + "step": 2654 + }, + { + "epoch": 0.06223738110726278, + "grad_norm": 4.223217010498047, + "learning_rate": 0.0001981079651942447, + "loss": 0.5716, + "step": 2655 + }, + { + "epoch": 0.062260822682067785, + "grad_norm": 1.5959312915802002, + "learning_rate": 0.00019810653880650509, + "loss": 1.441, + "step": 2656 + }, + { + "epoch": 0.06228426425687278, + "grad_norm": 0.9784368276596069, + "learning_rate": 0.00019810511188643665, + "loss": 0.1702, + "step": 2657 + }, + { + "epoch": 0.06230770583167777, + "grad_norm": 1.8201487064361572, + "learning_rate": 0.00019810368443404713, + "loss": 0.3596, + "step": 2658 + }, + { + "epoch": 0.06233114740648277, + "grad_norm": 1.5766232013702393, + "learning_rate": 0.00019810225644934423, + "loss": 0.2829, + "step": 2659 + }, + { + "epoch": 0.06235458898128776, + "grad_norm": 0.9600187540054321, + "learning_rate": 0.00019810082793233574, + "loss": 0.1812, + "step": 2660 + }, + { + "epoch": 0.062378030556092756, + "grad_norm": 4.093736171722412, + "learning_rate": 0.00019809939888302942, + "loss": 1.0631, + "step": 2661 + }, + { + "epoch": 0.06240147213089775, + "grad_norm": 3.939683198928833, + "learning_rate": 0.00019809796930143298, + "loss": 0.5278, + "step": 2662 + }, + { + "epoch": 0.06242491370570275, + "grad_norm": 0.8696374297142029, + "learning_rate": 0.00019809653918755423, + "loss": 0.1287, + "step": 2663 + }, + { + "epoch": 0.062448355280507746, + "grad_norm": 2.259690284729004, + "learning_rate": 0.00019809510854140085, + "loss": 1.6025, + "step": 2664 + }, + { + "epoch": 0.06247179685531274, + "grad_norm": 2.929776430130005, + "learning_rate": 0.0001980936773629807, + "loss": 0.6897, + "step": 2665 + }, + { + "epoch": 0.062495238430117735, + "grad_norm": 2.7886900901794434, + "learning_rate": 0.00019809224565230148, + "loss": 0.511, + "step": 2666 + }, + { + "epoch": 0.06251868000492274, + "grad_norm": 6.008452415466309, + "learning_rate": 0.000198090813409371, + "loss": 0.5949, + "step": 2667 + }, + { + "epoch": 0.06254212157972773, + "grad_norm": 3.439229965209961, + "learning_rate": 0.00019808938063419697, + "loss": 0.3161, + "step": 2668 + }, + { + "epoch": 0.06256556315453272, + "grad_norm": 0.7887934446334839, + "learning_rate": 0.0001980879473267872, + "loss": 0.1363, + "step": 2669 + }, + { + "epoch": 0.06258900472933772, + "grad_norm": 3.04499888420105, + "learning_rate": 0.00019808651348714952, + "loss": 0.5798, + "step": 2670 + }, + { + "epoch": 0.06261244630414271, + "grad_norm": 0.9909093976020813, + "learning_rate": 0.0001980850791152916, + "loss": 0.5932, + "step": 2671 + }, + { + "epoch": 0.06263588787894771, + "grad_norm": 0.49539080262184143, + "learning_rate": 0.0001980836442112213, + "loss": 0.0705, + "step": 2672 + }, + { + "epoch": 0.0626593294537527, + "grad_norm": 0.9526858925819397, + "learning_rate": 0.00019808220877494642, + "loss": 0.1563, + "step": 2673 + }, + { + "epoch": 0.0626827710285577, + "grad_norm": 1.5631382465362549, + "learning_rate": 0.00019808077280647468, + "loss": 1.4329, + "step": 2674 + }, + { + "epoch": 0.06270621260336269, + "grad_norm": 2.7091968059539795, + "learning_rate": 0.00019807933630581394, + "loss": 1.7836, + "step": 2675 + }, + { + "epoch": 0.06272965417816768, + "grad_norm": 2.605114698410034, + "learning_rate": 0.00019807789927297193, + "loss": 0.4355, + "step": 2676 + }, + { + "epoch": 0.06275309575297268, + "grad_norm": 2.175694227218628, + "learning_rate": 0.0001980764617079565, + "loss": 0.1724, + "step": 2677 + }, + { + "epoch": 0.06277653732777767, + "grad_norm": 3.240293502807617, + "learning_rate": 0.00019807502361077545, + "loss": 0.4101, + "step": 2678 + }, + { + "epoch": 0.06279997890258268, + "grad_norm": 3.9517619609832764, + "learning_rate": 0.00019807358498143653, + "loss": 0.7007, + "step": 2679 + }, + { + "epoch": 0.06282342047738768, + "grad_norm": 4.458103179931641, + "learning_rate": 0.00019807214581994763, + "loss": 0.8937, + "step": 2680 + }, + { + "epoch": 0.06284686205219267, + "grad_norm": 4.989431858062744, + "learning_rate": 0.00019807070612631647, + "loss": 0.9154, + "step": 2681 + }, + { + "epoch": 0.06287030362699766, + "grad_norm": 6.9971418380737305, + "learning_rate": 0.0001980692659005509, + "loss": 1.1149, + "step": 2682 + }, + { + "epoch": 0.06289374520180266, + "grad_norm": 1.3498432636260986, + "learning_rate": 0.00019806782514265875, + "loss": 0.8741, + "step": 2683 + }, + { + "epoch": 0.06291718677660765, + "grad_norm": 1.2284905910491943, + "learning_rate": 0.0001980663838526478, + "loss": 0.1805, + "step": 2684 + }, + { + "epoch": 0.06294062835141265, + "grad_norm": 3.7031819820404053, + "learning_rate": 0.0001980649420305259, + "loss": 0.6706, + "step": 2685 + }, + { + "epoch": 0.06296406992621764, + "grad_norm": 3.6084344387054443, + "learning_rate": 0.00019806349967630088, + "loss": 0.5699, + "step": 2686 + }, + { + "epoch": 0.06298751150102264, + "grad_norm": 3.720982074737549, + "learning_rate": 0.00019806205678998056, + "loss": 0.5426, + "step": 2687 + }, + { + "epoch": 0.06301095307582763, + "grad_norm": 1.6576722860336304, + "learning_rate": 0.00019806061337157273, + "loss": 0.2755, + "step": 2688 + }, + { + "epoch": 0.06303439465063262, + "grad_norm": 1.4461232423782349, + "learning_rate": 0.00019805916942108528, + "loss": 0.2102, + "step": 2689 + }, + { + "epoch": 0.06305783622543762, + "grad_norm": 1.7573422193527222, + "learning_rate": 0.00019805772493852603, + "loss": 0.1648, + "step": 2690 + }, + { + "epoch": 0.06308127780024263, + "grad_norm": 4.4377546310424805, + "learning_rate": 0.0001980562799239028, + "loss": 0.6552, + "step": 2691 + }, + { + "epoch": 0.06310471937504762, + "grad_norm": 4.232267379760742, + "learning_rate": 0.0001980548343772234, + "loss": 0.4662, + "step": 2692 + }, + { + "epoch": 0.06312816094985262, + "grad_norm": 1.1029483079910278, + "learning_rate": 0.00019805338829849577, + "loss": 0.155, + "step": 2693 + }, + { + "epoch": 0.06315160252465761, + "grad_norm": 3.5994138717651367, + "learning_rate": 0.00019805194168772764, + "loss": 0.6632, + "step": 2694 + }, + { + "epoch": 0.0631750440994626, + "grad_norm": 5.083408832550049, + "learning_rate": 0.00019805049454492697, + "loss": 0.9412, + "step": 2695 + }, + { + "epoch": 0.0631984856742676, + "grad_norm": 2.094444513320923, + "learning_rate": 0.00019804904687010155, + "loss": 0.4027, + "step": 2696 + }, + { + "epoch": 0.0632219272490726, + "grad_norm": 3.2709336280822754, + "learning_rate": 0.00019804759866325923, + "loss": 0.522, + "step": 2697 + }, + { + "epoch": 0.06324536882387759, + "grad_norm": 3.847280502319336, + "learning_rate": 0.00019804614992440786, + "loss": 0.6279, + "step": 2698 + }, + { + "epoch": 0.06326881039868258, + "grad_norm": 2.132373332977295, + "learning_rate": 0.0001980447006535554, + "loss": 0.286, + "step": 2699 + }, + { + "epoch": 0.06329225197348758, + "grad_norm": 3.364121675491333, + "learning_rate": 0.00019804325085070955, + "loss": 0.3751, + "step": 2700 + }, + { + "epoch": 0.06331569354829257, + "grad_norm": 1.4283270835876465, + "learning_rate": 0.0001980418005158783, + "loss": 0.2732, + "step": 2701 + }, + { + "epoch": 0.06333913512309756, + "grad_norm": 3.5582807064056396, + "learning_rate": 0.00019804034964906948, + "loss": 0.7262, + "step": 2702 + }, + { + "epoch": 0.06336257669790256, + "grad_norm": 3.446863889694214, + "learning_rate": 0.00019803889825029098, + "loss": 0.4863, + "step": 2703 + }, + { + "epoch": 0.06338601827270757, + "grad_norm": 3.612922430038452, + "learning_rate": 0.00019803744631955066, + "loss": 0.4613, + "step": 2704 + }, + { + "epoch": 0.06340945984751256, + "grad_norm": 3.143585205078125, + "learning_rate": 0.00019803599385685637, + "loss": 0.3717, + "step": 2705 + }, + { + "epoch": 0.06343290142231756, + "grad_norm": 1.8094532489776611, + "learning_rate": 0.00019803454086221608, + "loss": 0.3542, + "step": 2706 + }, + { + "epoch": 0.06345634299712255, + "grad_norm": 0.33289748430252075, + "learning_rate": 0.00019803308733563758, + "loss": 0.052, + "step": 2707 + }, + { + "epoch": 0.06347978457192754, + "grad_norm": 4.4361419677734375, + "learning_rate": 0.00019803163327712878, + "loss": 0.6189, + "step": 2708 + }, + { + "epoch": 0.06350322614673254, + "grad_norm": 1.6760711669921875, + "learning_rate": 0.00019803017868669757, + "loss": 0.1631, + "step": 2709 + }, + { + "epoch": 0.06352666772153753, + "grad_norm": 2.6738245487213135, + "learning_rate": 0.00019802872356435188, + "loss": 0.4659, + "step": 2710 + }, + { + "epoch": 0.06355010929634253, + "grad_norm": 2.29325795173645, + "learning_rate": 0.00019802726791009958, + "loss": 0.8944, + "step": 2711 + }, + { + "epoch": 0.06357355087114752, + "grad_norm": 1.6378189325332642, + "learning_rate": 0.00019802581172394856, + "loss": 0.2383, + "step": 2712 + }, + { + "epoch": 0.06359699244595252, + "grad_norm": 3.2941293716430664, + "learning_rate": 0.00019802435500590674, + "loss": 0.6663, + "step": 2713 + }, + { + "epoch": 0.06362043402075751, + "grad_norm": 3.936326026916504, + "learning_rate": 0.000198022897755982, + "loss": 0.5829, + "step": 2714 + }, + { + "epoch": 0.0636438755955625, + "grad_norm": 4.9401702880859375, + "learning_rate": 0.00019802143997418224, + "loss": 0.7682, + "step": 2715 + }, + { + "epoch": 0.0636673171703675, + "grad_norm": 3.8095390796661377, + "learning_rate": 0.00019801998166051545, + "loss": 0.8011, + "step": 2716 + }, + { + "epoch": 0.06369075874517251, + "grad_norm": 3.3744473457336426, + "learning_rate": 0.00019801852281498946, + "loss": 0.4431, + "step": 2717 + }, + { + "epoch": 0.0637142003199775, + "grad_norm": 2.9254696369171143, + "learning_rate": 0.00019801706343761216, + "loss": 0.3828, + "step": 2718 + }, + { + "epoch": 0.0637376418947825, + "grad_norm": 3.208838701248169, + "learning_rate": 0.0001980156035283916, + "loss": 0.4984, + "step": 2719 + }, + { + "epoch": 0.06376108346958749, + "grad_norm": 2.722796678543091, + "learning_rate": 0.00019801414308733554, + "loss": 2.1051, + "step": 2720 + }, + { + "epoch": 0.06378452504439248, + "grad_norm": 4.251000881195068, + "learning_rate": 0.00019801268211445204, + "loss": 0.6743, + "step": 2721 + }, + { + "epoch": 0.06380796661919748, + "grad_norm": 0.8315420150756836, + "learning_rate": 0.00019801122060974895, + "loss": 0.1346, + "step": 2722 + }, + { + "epoch": 0.06383140819400247, + "grad_norm": 2.240028142929077, + "learning_rate": 0.00019800975857323422, + "loss": 0.4069, + "step": 2723 + }, + { + "epoch": 0.06385484976880747, + "grad_norm": 1.971676230430603, + "learning_rate": 0.0001980082960049158, + "loss": 0.2689, + "step": 2724 + }, + { + "epoch": 0.06387829134361246, + "grad_norm": 2.222534656524658, + "learning_rate": 0.00019800683290480158, + "loss": 0.3914, + "step": 2725 + }, + { + "epoch": 0.06390173291841746, + "grad_norm": 4.430580139160156, + "learning_rate": 0.00019800536927289954, + "loss": 0.8988, + "step": 2726 + }, + { + "epoch": 0.06392517449322245, + "grad_norm": 2.984745740890503, + "learning_rate": 0.00019800390510921762, + "loss": 0.3863, + "step": 2727 + }, + { + "epoch": 0.06394861606802744, + "grad_norm": 3.1757187843322754, + "learning_rate": 0.00019800244041376375, + "loss": 0.7782, + "step": 2728 + }, + { + "epoch": 0.06397205764283244, + "grad_norm": 1.2372218370437622, + "learning_rate": 0.00019800097518654587, + "loss": 0.2331, + "step": 2729 + }, + { + "epoch": 0.06399549921763745, + "grad_norm": 3.544370651245117, + "learning_rate": 0.00019799950942757198, + "loss": 0.5867, + "step": 2730 + }, + { + "epoch": 0.06401894079244244, + "grad_norm": 3.7154555320739746, + "learning_rate": 0.00019799804313684998, + "loss": 0.6001, + "step": 2731 + }, + { + "epoch": 0.06404238236724744, + "grad_norm": 4.08477258682251, + "learning_rate": 0.00019799657631438785, + "loss": 0.6686, + "step": 2732 + }, + { + "epoch": 0.06406582394205243, + "grad_norm": 0.8795322179794312, + "learning_rate": 0.0001979951089601935, + "loss": 0.0913, + "step": 2733 + }, + { + "epoch": 0.06408926551685742, + "grad_norm": 3.373696804046631, + "learning_rate": 0.00019799364107427495, + "loss": 0.4274, + "step": 2734 + }, + { + "epoch": 0.06411270709166242, + "grad_norm": 2.6959068775177, + "learning_rate": 0.00019799217265664018, + "loss": 0.492, + "step": 2735 + }, + { + "epoch": 0.06413614866646741, + "grad_norm": 3.2580273151397705, + "learning_rate": 0.0001979907037072971, + "loss": 0.4764, + "step": 2736 + }, + { + "epoch": 0.06415959024127241, + "grad_norm": 4.963383674621582, + "learning_rate": 0.0001979892342262537, + "loss": 0.797, + "step": 2737 + }, + { + "epoch": 0.0641830318160774, + "grad_norm": 3.2307090759277344, + "learning_rate": 0.000197987764213518, + "loss": 0.5064, + "step": 2738 + }, + { + "epoch": 0.0642064733908824, + "grad_norm": 6.396978855133057, + "learning_rate": 0.0001979862936690979, + "loss": 0.672, + "step": 2739 + }, + { + "epoch": 0.06422991496568739, + "grad_norm": 2.665236473083496, + "learning_rate": 0.00019798482259300143, + "loss": 0.4815, + "step": 2740 + }, + { + "epoch": 0.06425335654049238, + "grad_norm": 2.8122317790985107, + "learning_rate": 0.00019798335098523655, + "loss": 0.3591, + "step": 2741 + }, + { + "epoch": 0.06427679811529738, + "grad_norm": 1.3731534481048584, + "learning_rate": 0.00019798187884581124, + "loss": 0.1965, + "step": 2742 + }, + { + "epoch": 0.06430023969010239, + "grad_norm": 3.9942774772644043, + "learning_rate": 0.0001979804061747335, + "loss": 0.355, + "step": 2743 + }, + { + "epoch": 0.06432368126490738, + "grad_norm": 3.5812554359436035, + "learning_rate": 0.00019797893297201133, + "loss": 0.7804, + "step": 2744 + }, + { + "epoch": 0.06434712283971238, + "grad_norm": 5.846257209777832, + "learning_rate": 0.00019797745923765274, + "loss": 0.8222, + "step": 2745 + }, + { + "epoch": 0.06437056441451737, + "grad_norm": 4.441455841064453, + "learning_rate": 0.00019797598497166567, + "loss": 0.7547, + "step": 2746 + }, + { + "epoch": 0.06439400598932236, + "grad_norm": 2.733030319213867, + "learning_rate": 0.00019797451017405818, + "loss": 0.3056, + "step": 2747 + }, + { + "epoch": 0.06441744756412736, + "grad_norm": 2.8554110527038574, + "learning_rate": 0.00019797303484483822, + "loss": 1.831, + "step": 2748 + }, + { + "epoch": 0.06444088913893235, + "grad_norm": 4.793402671813965, + "learning_rate": 0.00019797155898401383, + "loss": 0.6129, + "step": 2749 + }, + { + "epoch": 0.06446433071373735, + "grad_norm": 3.131700277328491, + "learning_rate": 0.00019797008259159297, + "loss": 0.3516, + "step": 2750 + }, + { + "epoch": 0.06448777228854234, + "grad_norm": 3.5344512462615967, + "learning_rate": 0.00019796860566758374, + "loss": 0.4792, + "step": 2751 + }, + { + "epoch": 0.06451121386334734, + "grad_norm": 4.269152641296387, + "learning_rate": 0.00019796712821199406, + "loss": 0.6487, + "step": 2752 + }, + { + "epoch": 0.06453465543815233, + "grad_norm": 1.5638982057571411, + "learning_rate": 0.00019796565022483202, + "loss": 0.3193, + "step": 2753 + }, + { + "epoch": 0.06455809701295732, + "grad_norm": 2.5967202186584473, + "learning_rate": 0.0001979641717061056, + "loss": 0.2937, + "step": 2754 + }, + { + "epoch": 0.06458153858776233, + "grad_norm": 1.5150889158248901, + "learning_rate": 0.00019796269265582282, + "loss": 0.3573, + "step": 2755 + }, + { + "epoch": 0.06460498016256733, + "grad_norm": 5.317030429840088, + "learning_rate": 0.00019796121307399172, + "loss": 0.9717, + "step": 2756 + }, + { + "epoch": 0.06462842173737232, + "grad_norm": 5.82094144821167, + "learning_rate": 0.0001979597329606203, + "loss": 0.8095, + "step": 2757 + }, + { + "epoch": 0.06465186331217732, + "grad_norm": 4.819675922393799, + "learning_rate": 0.00019795825231571664, + "loss": 1.8287, + "step": 2758 + }, + { + "epoch": 0.06467530488698231, + "grad_norm": 3.1786468029022217, + "learning_rate": 0.00019795677113928874, + "loss": 1.5311, + "step": 2759 + }, + { + "epoch": 0.0646987464617873, + "grad_norm": 5.053433418273926, + "learning_rate": 0.00019795528943134465, + "loss": 0.7957, + "step": 2760 + }, + { + "epoch": 0.0647221880365923, + "grad_norm": 3.958200693130493, + "learning_rate": 0.0001979538071918924, + "loss": 0.6864, + "step": 2761 + }, + { + "epoch": 0.06474562961139729, + "grad_norm": 1.5436042547225952, + "learning_rate": 0.00019795232442094005, + "loss": 0.2683, + "step": 2762 + }, + { + "epoch": 0.06476907118620229, + "grad_norm": 3.6210246086120605, + "learning_rate": 0.00019795084111849561, + "loss": 0.8451, + "step": 2763 + }, + { + "epoch": 0.06479251276100728, + "grad_norm": 3.972602605819702, + "learning_rate": 0.0001979493572845672, + "loss": 0.5614, + "step": 2764 + }, + { + "epoch": 0.06481595433581228, + "grad_norm": 2.6513068675994873, + "learning_rate": 0.00019794787291916278, + "loss": 0.3883, + "step": 2765 + }, + { + "epoch": 0.06483939591061727, + "grad_norm": 5.577168941497803, + "learning_rate": 0.00019794638802229044, + "loss": 0.9798, + "step": 2766 + }, + { + "epoch": 0.06486283748542226, + "grad_norm": 1.724837064743042, + "learning_rate": 0.0001979449025939583, + "loss": 0.2507, + "step": 2767 + }, + { + "epoch": 0.06488627906022727, + "grad_norm": 1.0417041778564453, + "learning_rate": 0.0001979434166341743, + "loss": 0.6538, + "step": 2768 + }, + { + "epoch": 0.06490972063503227, + "grad_norm": 3.196051597595215, + "learning_rate": 0.00019794193014294659, + "loss": 0.4361, + "step": 2769 + }, + { + "epoch": 0.06493316220983726, + "grad_norm": 3.3163185119628906, + "learning_rate": 0.00019794044312028324, + "loss": 0.4623, + "step": 2770 + }, + { + "epoch": 0.06495660378464226, + "grad_norm": 3.6784985065460205, + "learning_rate": 0.00019793895556619227, + "loss": 0.7495, + "step": 2771 + }, + { + "epoch": 0.06498004535944725, + "grad_norm": 0.9937915205955505, + "learning_rate": 0.00019793746748068177, + "loss": 0.1961, + "step": 2772 + }, + { + "epoch": 0.06500348693425224, + "grad_norm": 3.1808857917785645, + "learning_rate": 0.00019793597886375982, + "loss": 0.2342, + "step": 2773 + }, + { + "epoch": 0.06502692850905724, + "grad_norm": 2.258883476257324, + "learning_rate": 0.0001979344897154345, + "loss": 0.1791, + "step": 2774 + }, + { + "epoch": 0.06505037008386223, + "grad_norm": 3.612659215927124, + "learning_rate": 0.00019793300003571385, + "loss": 0.6043, + "step": 2775 + }, + { + "epoch": 0.06507381165866723, + "grad_norm": 3.87858510017395, + "learning_rate": 0.00019793150982460605, + "loss": 0.4083, + "step": 2776 + }, + { + "epoch": 0.06509725323347222, + "grad_norm": 2.4656822681427, + "learning_rate": 0.00019793001908211907, + "loss": 0.4346, + "step": 2777 + }, + { + "epoch": 0.06512069480827722, + "grad_norm": 1.4105762243270874, + "learning_rate": 0.00019792852780826107, + "loss": 0.2421, + "step": 2778 + }, + { + "epoch": 0.06514413638308221, + "grad_norm": 2.300529956817627, + "learning_rate": 0.00019792703600304015, + "loss": 0.4761, + "step": 2779 + }, + { + "epoch": 0.0651675779578872, + "grad_norm": 2.368147611618042, + "learning_rate": 0.00019792554366646436, + "loss": 0.3913, + "step": 2780 + }, + { + "epoch": 0.06519101953269221, + "grad_norm": 2.84698486328125, + "learning_rate": 0.00019792405079854182, + "loss": 0.2446, + "step": 2781 + }, + { + "epoch": 0.0652144611074972, + "grad_norm": 3.6248390674591064, + "learning_rate": 0.00019792255739928063, + "loss": 0.9062, + "step": 2782 + }, + { + "epoch": 0.0652379026823022, + "grad_norm": 3.3227460384368896, + "learning_rate": 0.00019792106346868887, + "loss": 0.5846, + "step": 2783 + }, + { + "epoch": 0.0652613442571072, + "grad_norm": 2.5104455947875977, + "learning_rate": 0.00019791956900677468, + "loss": 0.2717, + "step": 2784 + }, + { + "epoch": 0.06528478583191219, + "grad_norm": 2.3213982582092285, + "learning_rate": 0.00019791807401354616, + "loss": 1.5047, + "step": 2785 + }, + { + "epoch": 0.06530822740671718, + "grad_norm": 3.1669106483459473, + "learning_rate": 0.0001979165784890114, + "loss": 0.5569, + "step": 2786 + }, + { + "epoch": 0.06533166898152218, + "grad_norm": 3.854616165161133, + "learning_rate": 0.00019791508243317856, + "loss": 0.7861, + "step": 2787 + }, + { + "epoch": 0.06535511055632717, + "grad_norm": 5.166064262390137, + "learning_rate": 0.00019791358584605572, + "loss": 0.5718, + "step": 2788 + }, + { + "epoch": 0.06537855213113217, + "grad_norm": 3.568742513656616, + "learning_rate": 0.00019791208872765103, + "loss": 0.5973, + "step": 2789 + }, + { + "epoch": 0.06540199370593716, + "grad_norm": 3.351240873336792, + "learning_rate": 0.00019791059107797254, + "loss": 0.5671, + "step": 2790 + }, + { + "epoch": 0.06542543528074216, + "grad_norm": 3.3318865299224854, + "learning_rate": 0.00019790909289702847, + "loss": 0.4979, + "step": 2791 + }, + { + "epoch": 0.06544887685554715, + "grad_norm": 1.9181993007659912, + "learning_rate": 0.00019790759418482687, + "loss": 0.2571, + "step": 2792 + }, + { + "epoch": 0.06547231843035214, + "grad_norm": 4.416832447052002, + "learning_rate": 0.00019790609494137594, + "loss": 0.7543, + "step": 2793 + }, + { + "epoch": 0.06549576000515715, + "grad_norm": 2.9129045009613037, + "learning_rate": 0.0001979045951666838, + "loss": 0.5085, + "step": 2794 + }, + { + "epoch": 0.06551920157996215, + "grad_norm": 3.0048928260803223, + "learning_rate": 0.00019790309486075855, + "loss": 0.6071, + "step": 2795 + }, + { + "epoch": 0.06554264315476714, + "grad_norm": 3.180528402328491, + "learning_rate": 0.00019790159402360838, + "loss": 0.4962, + "step": 2796 + }, + { + "epoch": 0.06556608472957214, + "grad_norm": 3.5553269386291504, + "learning_rate": 0.00019790009265524138, + "loss": 0.5357, + "step": 2797 + }, + { + "epoch": 0.06558952630437713, + "grad_norm": 1.8865454196929932, + "learning_rate": 0.00019789859075566572, + "loss": 0.3618, + "step": 2798 + }, + { + "epoch": 0.06561296787918212, + "grad_norm": 3.390040397644043, + "learning_rate": 0.00019789708832488958, + "loss": 0.2819, + "step": 2799 + }, + { + "epoch": 0.06563640945398712, + "grad_norm": 1.086047887802124, + "learning_rate": 0.00019789558536292105, + "loss": 0.1433, + "step": 2800 + }, + { + "epoch": 0.06565985102879211, + "grad_norm": 4.49606466293335, + "learning_rate": 0.00019789408186976837, + "loss": 0.8485, + "step": 2801 + }, + { + "epoch": 0.06568329260359711, + "grad_norm": 2.598175287246704, + "learning_rate": 0.00019789257784543962, + "loss": 1.5838, + "step": 2802 + }, + { + "epoch": 0.0657067341784021, + "grad_norm": 1.6364675760269165, + "learning_rate": 0.000197891073289943, + "loss": 1.2147, + "step": 2803 + }, + { + "epoch": 0.0657301757532071, + "grad_norm": 3.26023006439209, + "learning_rate": 0.00019788956820328667, + "loss": 0.2114, + "step": 2804 + }, + { + "epoch": 0.06575361732801209, + "grad_norm": 4.128780841827393, + "learning_rate": 0.00019788806258547875, + "loss": 0.5613, + "step": 2805 + }, + { + "epoch": 0.06577705890281708, + "grad_norm": 4.8270440101623535, + "learning_rate": 0.0001978865564365275, + "loss": 0.8038, + "step": 2806 + }, + { + "epoch": 0.06580050047762209, + "grad_norm": 0.7508904933929443, + "learning_rate": 0.000197885049756441, + "loss": 0.0934, + "step": 2807 + }, + { + "epoch": 0.06582394205242709, + "grad_norm": 4.854770183563232, + "learning_rate": 0.00019788354254522748, + "loss": 0.5389, + "step": 2808 + }, + { + "epoch": 0.06584738362723208, + "grad_norm": 4.67726993560791, + "learning_rate": 0.00019788203480289513, + "loss": 0.7726, + "step": 2809 + }, + { + "epoch": 0.06587082520203708, + "grad_norm": 4.366601943969727, + "learning_rate": 0.00019788052652945211, + "loss": 0.6285, + "step": 2810 + }, + { + "epoch": 0.06589426677684207, + "grad_norm": 1.614260196685791, + "learning_rate": 0.0001978790177249066, + "loss": 0.313, + "step": 2811 + }, + { + "epoch": 0.06591770835164706, + "grad_norm": 5.316240310668945, + "learning_rate": 0.00019787750838926675, + "loss": 0.7493, + "step": 2812 + }, + { + "epoch": 0.06594114992645206, + "grad_norm": 3.8810324668884277, + "learning_rate": 0.00019787599852254083, + "loss": 0.4002, + "step": 2813 + }, + { + "epoch": 0.06596459150125705, + "grad_norm": 3.9180140495300293, + "learning_rate": 0.00019787448812473696, + "loss": 0.4345, + "step": 2814 + }, + { + "epoch": 0.06598803307606205, + "grad_norm": 3.460683584213257, + "learning_rate": 0.00019787297719586335, + "loss": 0.4064, + "step": 2815 + }, + { + "epoch": 0.06601147465086704, + "grad_norm": 2.8506827354431152, + "learning_rate": 0.00019787146573592825, + "loss": 0.2726, + "step": 2816 + }, + { + "epoch": 0.06603491622567204, + "grad_norm": 4.286886692047119, + "learning_rate": 0.0001978699537449398, + "loss": 0.7216, + "step": 2817 + }, + { + "epoch": 0.06605835780047703, + "grad_norm": 2.5698137283325195, + "learning_rate": 0.00019786844122290624, + "loss": 0.3936, + "step": 2818 + }, + { + "epoch": 0.06608179937528204, + "grad_norm": 3.65193510055542, + "learning_rate": 0.0001978669281698358, + "loss": 0.6232, + "step": 2819 + }, + { + "epoch": 0.06610524095008703, + "grad_norm": 1.2153165340423584, + "learning_rate": 0.0001978654145857366, + "loss": 0.6661, + "step": 2820 + }, + { + "epoch": 0.06612868252489203, + "grad_norm": 5.143307209014893, + "learning_rate": 0.00019786390047061693, + "loss": 0.8445, + "step": 2821 + }, + { + "epoch": 0.06615212409969702, + "grad_norm": 2.556619167327881, + "learning_rate": 0.000197862385824485, + "loss": 0.3955, + "step": 2822 + }, + { + "epoch": 0.06617556567450202, + "grad_norm": 2.0981504917144775, + "learning_rate": 0.00019786087064734898, + "loss": 0.3438, + "step": 2823 + }, + { + "epoch": 0.06619900724930701, + "grad_norm": 0.3194495737552643, + "learning_rate": 0.00019785935493921717, + "loss": 0.0489, + "step": 2824 + }, + { + "epoch": 0.066222448824112, + "grad_norm": 1.2242820262908936, + "learning_rate": 0.00019785783870009772, + "loss": 0.2064, + "step": 2825 + }, + { + "epoch": 0.066245890398917, + "grad_norm": 8.996512413024902, + "learning_rate": 0.00019785632192999888, + "loss": 0.7774, + "step": 2826 + }, + { + "epoch": 0.06626933197372199, + "grad_norm": 1.282038927078247, + "learning_rate": 0.00019785480462892888, + "loss": 0.1695, + "step": 2827 + }, + { + "epoch": 0.06629277354852699, + "grad_norm": 3.834756851196289, + "learning_rate": 0.000197853286796896, + "loss": 0.7222, + "step": 2828 + }, + { + "epoch": 0.06631621512333198, + "grad_norm": 4.475552082061768, + "learning_rate": 0.0001978517684339084, + "loss": 0.7253, + "step": 2829 + }, + { + "epoch": 0.06633965669813698, + "grad_norm": 3.3962535858154297, + "learning_rate": 0.00019785024953997436, + "loss": 0.4048, + "step": 2830 + }, + { + "epoch": 0.06636309827294197, + "grad_norm": 2.5720057487487793, + "learning_rate": 0.0001978487301151021, + "loss": 0.64, + "step": 2831 + }, + { + "epoch": 0.06638653984774698, + "grad_norm": 1.3874224424362183, + "learning_rate": 0.0001978472101592999, + "loss": 0.2059, + "step": 2832 + }, + { + "epoch": 0.06640998142255197, + "grad_norm": 2.594933271408081, + "learning_rate": 0.000197845689672576, + "loss": 0.3934, + "step": 2833 + }, + { + "epoch": 0.06643342299735697, + "grad_norm": 2.9350452423095703, + "learning_rate": 0.0001978441686549386, + "loss": 0.4922, + "step": 2834 + }, + { + "epoch": 0.06645686457216196, + "grad_norm": 4.413056373596191, + "learning_rate": 0.00019784264710639602, + "loss": 0.7062, + "step": 2835 + }, + { + "epoch": 0.06648030614696696, + "grad_norm": 3.5566904544830322, + "learning_rate": 0.00019784112502695646, + "loss": 0.4714, + "step": 2836 + }, + { + "epoch": 0.06650374772177195, + "grad_norm": 3.4465503692626953, + "learning_rate": 0.00019783960241662822, + "loss": 0.5002, + "step": 2837 + }, + { + "epoch": 0.06652718929657694, + "grad_norm": 3.657160758972168, + "learning_rate": 0.00019783807927541955, + "loss": 0.775, + "step": 2838 + }, + { + "epoch": 0.06655063087138194, + "grad_norm": 2.3615200519561768, + "learning_rate": 0.0001978365556033387, + "loss": 1.6707, + "step": 2839 + }, + { + "epoch": 0.06657407244618693, + "grad_norm": 1.6344444751739502, + "learning_rate": 0.00019783503140039397, + "loss": 0.3685, + "step": 2840 + }, + { + "epoch": 0.06659751402099193, + "grad_norm": 1.6559139490127563, + "learning_rate": 0.0001978335066665936, + "loss": 0.2963, + "step": 2841 + }, + { + "epoch": 0.06662095559579692, + "grad_norm": 3.322252035140991, + "learning_rate": 0.00019783198140194586, + "loss": 0.4402, + "step": 2842 + }, + { + "epoch": 0.06664439717060192, + "grad_norm": 4.710109233856201, + "learning_rate": 0.00019783045560645902, + "loss": 0.4538, + "step": 2843 + }, + { + "epoch": 0.06666783874540691, + "grad_norm": 3.738393783569336, + "learning_rate": 0.0001978289292801414, + "loss": 0.4668, + "step": 2844 + }, + { + "epoch": 0.06669128032021192, + "grad_norm": 3.114119529724121, + "learning_rate": 0.00019782740242300126, + "loss": 0.4047, + "step": 2845 + }, + { + "epoch": 0.06671472189501691, + "grad_norm": 3.8203039169311523, + "learning_rate": 0.0001978258750350469, + "loss": 0.4607, + "step": 2846 + }, + { + "epoch": 0.0667381634698219, + "grad_norm": 0.9607769846916199, + "learning_rate": 0.00019782434711628657, + "loss": 0.1171, + "step": 2847 + }, + { + "epoch": 0.0667616050446269, + "grad_norm": 4.086828231811523, + "learning_rate": 0.00019782281866672857, + "loss": 0.6899, + "step": 2848 + }, + { + "epoch": 0.0667850466194319, + "grad_norm": 4.988412857055664, + "learning_rate": 0.00019782128968638124, + "loss": 0.5235, + "step": 2849 + }, + { + "epoch": 0.06680848819423689, + "grad_norm": 5.082761764526367, + "learning_rate": 0.0001978197601752528, + "loss": 0.6351, + "step": 2850 + }, + { + "epoch": 0.06683192976904188, + "grad_norm": 1.3528082370758057, + "learning_rate": 0.00019781823013335163, + "loss": 0.175, + "step": 2851 + }, + { + "epoch": 0.06685537134384688, + "grad_norm": 3.099289894104004, + "learning_rate": 0.00019781669956068595, + "loss": 0.3616, + "step": 2852 + }, + { + "epoch": 0.06687881291865187, + "grad_norm": 4.401792526245117, + "learning_rate": 0.00019781516845726415, + "loss": 0.6762, + "step": 2853 + }, + { + "epoch": 0.06690225449345687, + "grad_norm": 3.6533210277557373, + "learning_rate": 0.00019781363682309447, + "loss": 0.5116, + "step": 2854 + }, + { + "epoch": 0.06692569606826186, + "grad_norm": 3.160377025604248, + "learning_rate": 0.00019781210465818526, + "loss": 0.4525, + "step": 2855 + }, + { + "epoch": 0.06694913764306686, + "grad_norm": 3.3477349281311035, + "learning_rate": 0.00019781057196254482, + "loss": 0.5457, + "step": 2856 + }, + { + "epoch": 0.06697257921787185, + "grad_norm": 1.81168532371521, + "learning_rate": 0.00019780903873618146, + "loss": 0.348, + "step": 2857 + }, + { + "epoch": 0.06699602079267686, + "grad_norm": 6.474959850311279, + "learning_rate": 0.0001978075049791035, + "loss": 0.7023, + "step": 2858 + }, + { + "epoch": 0.06701946236748185, + "grad_norm": 4.862529277801514, + "learning_rate": 0.0001978059706913193, + "loss": 0.6122, + "step": 2859 + }, + { + "epoch": 0.06704290394228685, + "grad_norm": 3.861382007598877, + "learning_rate": 0.00019780443587283714, + "loss": 0.5986, + "step": 2860 + }, + { + "epoch": 0.06706634551709184, + "grad_norm": 7.095008373260498, + "learning_rate": 0.00019780290052366534, + "loss": 1.2436, + "step": 2861 + }, + { + "epoch": 0.06708978709189684, + "grad_norm": 3.7197768688201904, + "learning_rate": 0.00019780136464381224, + "loss": 0.8026, + "step": 2862 + }, + { + "epoch": 0.06711322866670183, + "grad_norm": 2.49086856842041, + "learning_rate": 0.0001977998282332862, + "loss": 1.6209, + "step": 2863 + }, + { + "epoch": 0.06713667024150682, + "grad_norm": 1.5139862298965454, + "learning_rate": 0.00019779829129209557, + "loss": 0.9567, + "step": 2864 + }, + { + "epoch": 0.06716011181631182, + "grad_norm": 3.214686632156372, + "learning_rate": 0.00019779675382024863, + "loss": 0.5524, + "step": 2865 + }, + { + "epoch": 0.06718355339111681, + "grad_norm": 4.732641220092773, + "learning_rate": 0.00019779521581775376, + "loss": 0.5071, + "step": 2866 + }, + { + "epoch": 0.0672069949659218, + "grad_norm": 4.225509166717529, + "learning_rate": 0.0001977936772846193, + "loss": 0.8867, + "step": 2867 + }, + { + "epoch": 0.0672304365407268, + "grad_norm": 4.775800704956055, + "learning_rate": 0.00019779213822085362, + "loss": 0.7963, + "step": 2868 + }, + { + "epoch": 0.0672538781155318, + "grad_norm": 2.4707508087158203, + "learning_rate": 0.00019779059862646504, + "loss": 0.2857, + "step": 2869 + }, + { + "epoch": 0.0672773196903368, + "grad_norm": 4.947341442108154, + "learning_rate": 0.0001977890585014619, + "loss": 0.4463, + "step": 2870 + }, + { + "epoch": 0.0673007612651418, + "grad_norm": 3.4863932132720947, + "learning_rate": 0.0001977875178458526, + "loss": 0.535, + "step": 2871 + }, + { + "epoch": 0.06732420283994679, + "grad_norm": 1.106172800064087, + "learning_rate": 0.00019778597665964546, + "loss": 0.1331, + "step": 2872 + }, + { + "epoch": 0.06734764441475179, + "grad_norm": 4.720610618591309, + "learning_rate": 0.00019778443494284886, + "loss": 0.6721, + "step": 2873 + }, + { + "epoch": 0.06737108598955678, + "grad_norm": 4.263083457946777, + "learning_rate": 0.0001977828926954712, + "loss": 0.5074, + "step": 2874 + }, + { + "epoch": 0.06739452756436178, + "grad_norm": 1.5311895608901978, + "learning_rate": 0.00019778134991752078, + "loss": 0.2164, + "step": 2875 + }, + { + "epoch": 0.06741796913916677, + "grad_norm": 0.6677362322807312, + "learning_rate": 0.00019777980660900602, + "loss": 0.0692, + "step": 2876 + }, + { + "epoch": 0.06744141071397176, + "grad_norm": 4.205953121185303, + "learning_rate": 0.00019777826276993525, + "loss": 0.5483, + "step": 2877 + }, + { + "epoch": 0.06746485228877676, + "grad_norm": 4.410140037536621, + "learning_rate": 0.00019777671840031692, + "loss": 0.427, + "step": 2878 + }, + { + "epoch": 0.06748829386358175, + "grad_norm": 4.117190361022949, + "learning_rate": 0.00019777517350015936, + "loss": 0.6864, + "step": 2879 + }, + { + "epoch": 0.06751173543838675, + "grad_norm": 2.5536081790924072, + "learning_rate": 0.00019777362806947096, + "loss": 0.2607, + "step": 2880 + }, + { + "epoch": 0.06753517701319174, + "grad_norm": 2.2017650604248047, + "learning_rate": 0.00019777208210826008, + "loss": 0.3346, + "step": 2881 + }, + { + "epoch": 0.06755861858799674, + "grad_norm": 3.003631591796875, + "learning_rate": 0.00019777053561653515, + "loss": 0.4617, + "step": 2882 + }, + { + "epoch": 0.06758206016280174, + "grad_norm": 5.049400806427002, + "learning_rate": 0.00019776898859430456, + "loss": 0.4603, + "step": 2883 + }, + { + "epoch": 0.06760550173760674, + "grad_norm": 1.898126244544983, + "learning_rate": 0.00019776744104157664, + "loss": 0.1472, + "step": 2884 + }, + { + "epoch": 0.06762894331241173, + "grad_norm": 2.7748258113861084, + "learning_rate": 0.0001977658929583599, + "loss": 0.4734, + "step": 2885 + }, + { + "epoch": 0.06765238488721673, + "grad_norm": 4.256916046142578, + "learning_rate": 0.00019776434434466263, + "loss": 0.5977, + "step": 2886 + }, + { + "epoch": 0.06767582646202172, + "grad_norm": 4.060443878173828, + "learning_rate": 0.00019776279520049327, + "loss": 1.0068, + "step": 2887 + }, + { + "epoch": 0.06769926803682672, + "grad_norm": 1.4075946807861328, + "learning_rate": 0.0001977612455258603, + "loss": 0.1724, + "step": 2888 + }, + { + "epoch": 0.06772270961163171, + "grad_norm": 2.1409778594970703, + "learning_rate": 0.000197759695320772, + "loss": 0.273, + "step": 2889 + }, + { + "epoch": 0.0677461511864367, + "grad_norm": 1.4824535846710205, + "learning_rate": 0.00019775814458523688, + "loss": 0.1727, + "step": 2890 + }, + { + "epoch": 0.0677695927612417, + "grad_norm": 4.815151214599609, + "learning_rate": 0.0001977565933192633, + "loss": 0.6349, + "step": 2891 + }, + { + "epoch": 0.06779303433604669, + "grad_norm": 5.018848419189453, + "learning_rate": 0.00019775504152285967, + "loss": 0.9658, + "step": 2892 + }, + { + "epoch": 0.06781647591085169, + "grad_norm": 2.8058087825775146, + "learning_rate": 0.00019775348919603446, + "loss": 2.3146, + "step": 2893 + }, + { + "epoch": 0.06783991748565668, + "grad_norm": 1.4663589000701904, + "learning_rate": 0.0001977519363387961, + "loss": 0.2479, + "step": 2894 + }, + { + "epoch": 0.06786335906046168, + "grad_norm": 4.9410810470581055, + "learning_rate": 0.00019775038295115295, + "loss": 0.6865, + "step": 2895 + }, + { + "epoch": 0.06788680063526668, + "grad_norm": 5.368740081787109, + "learning_rate": 0.00019774882903311348, + "loss": 0.8816, + "step": 2896 + }, + { + "epoch": 0.06791024221007168, + "grad_norm": 1.411388874053955, + "learning_rate": 0.00019774727458468614, + "loss": 0.7188, + "step": 2897 + }, + { + "epoch": 0.06793368378487667, + "grad_norm": 1.4218391180038452, + "learning_rate": 0.00019774571960587932, + "loss": 0.2131, + "step": 2898 + }, + { + "epoch": 0.06795712535968167, + "grad_norm": 4.371151924133301, + "learning_rate": 0.00019774416409670147, + "loss": 0.8532, + "step": 2899 + }, + { + "epoch": 0.06798056693448666, + "grad_norm": 1.2067437171936035, + "learning_rate": 0.00019774260805716104, + "loss": 0.1051, + "step": 2900 + }, + { + "epoch": 0.06800400850929166, + "grad_norm": 3.2691757678985596, + "learning_rate": 0.00019774105148726647, + "loss": 0.3736, + "step": 2901 + }, + { + "epoch": 0.06802745008409665, + "grad_norm": 2.7037136554718018, + "learning_rate": 0.00019773949438702622, + "loss": 1.8086, + "step": 2902 + }, + { + "epoch": 0.06805089165890164, + "grad_norm": 3.1099467277526855, + "learning_rate": 0.0001977379367564487, + "loss": 0.5076, + "step": 2903 + }, + { + "epoch": 0.06807433323370664, + "grad_norm": 0.45252034068107605, + "learning_rate": 0.0001977363785955424, + "loss": 0.0688, + "step": 2904 + }, + { + "epoch": 0.06809777480851163, + "grad_norm": 0.8951784372329712, + "learning_rate": 0.00019773481990431577, + "loss": 0.104, + "step": 2905 + }, + { + "epoch": 0.06812121638331663, + "grad_norm": 4.5811309814453125, + "learning_rate": 0.00019773326068277727, + "loss": 0.6235, + "step": 2906 + }, + { + "epoch": 0.06814465795812162, + "grad_norm": 2.1153831481933594, + "learning_rate": 0.00019773170093093532, + "loss": 1.1612, + "step": 2907 + }, + { + "epoch": 0.06816809953292662, + "grad_norm": 6.570570468902588, + "learning_rate": 0.00019773014064879843, + "loss": 0.634, + "step": 2908 + }, + { + "epoch": 0.06819154110773162, + "grad_norm": 2.904690742492676, + "learning_rate": 0.00019772857983637503, + "loss": 0.5454, + "step": 2909 + }, + { + "epoch": 0.06821498268253662, + "grad_norm": 1.4681142568588257, + "learning_rate": 0.0001977270184936736, + "loss": 0.2163, + "step": 2910 + }, + { + "epoch": 0.06823842425734161, + "grad_norm": 2.6314280033111572, + "learning_rate": 0.00019772545662070264, + "loss": 1.5319, + "step": 2911 + }, + { + "epoch": 0.0682618658321466, + "grad_norm": 1.6198997497558594, + "learning_rate": 0.00019772389421747062, + "loss": 1.1422, + "step": 2912 + }, + { + "epoch": 0.0682853074069516, + "grad_norm": 2.8985719680786133, + "learning_rate": 0.00019772233128398598, + "loss": 0.421, + "step": 2913 + }, + { + "epoch": 0.0683087489817566, + "grad_norm": 3.2915356159210205, + "learning_rate": 0.00019772076782025724, + "loss": 0.5169, + "step": 2914 + }, + { + "epoch": 0.06833219055656159, + "grad_norm": 0.8594393730163574, + "learning_rate": 0.00019771920382629285, + "loss": 0.0543, + "step": 2915 + }, + { + "epoch": 0.06835563213136658, + "grad_norm": 3.425419569015503, + "learning_rate": 0.00019771763930210133, + "loss": 0.6154, + "step": 2916 + }, + { + "epoch": 0.06837907370617158, + "grad_norm": 5.833780765533447, + "learning_rate": 0.00019771607424769115, + "loss": 0.7603, + "step": 2917 + }, + { + "epoch": 0.06840251528097657, + "grad_norm": 2.2019829750061035, + "learning_rate": 0.0001977145086630708, + "loss": 1.382, + "step": 2918 + }, + { + "epoch": 0.06842595685578157, + "grad_norm": 6.6841583251953125, + "learning_rate": 0.00019771294254824877, + "loss": 0.7655, + "step": 2919 + }, + { + "epoch": 0.06844939843058656, + "grad_norm": 2.0097577571868896, + "learning_rate": 0.00019771137590323355, + "loss": 1.2247, + "step": 2920 + }, + { + "epoch": 0.06847284000539156, + "grad_norm": 4.167199611663818, + "learning_rate": 0.0001977098087280337, + "loss": 0.6722, + "step": 2921 + }, + { + "epoch": 0.06849628158019656, + "grad_norm": 2.538011074066162, + "learning_rate": 0.00019770824102265767, + "loss": 1.7405, + "step": 2922 + }, + { + "epoch": 0.06851972315500156, + "grad_norm": 4.896349906921387, + "learning_rate": 0.00019770667278711394, + "loss": 0.8591, + "step": 2923 + }, + { + "epoch": 0.06854316472980655, + "grad_norm": 5.318864822387695, + "learning_rate": 0.00019770510402141106, + "loss": 0.849, + "step": 2924 + }, + { + "epoch": 0.06856660630461155, + "grad_norm": 1.971889853477478, + "learning_rate": 0.0001977035347255576, + "loss": 0.3935, + "step": 2925 + }, + { + "epoch": 0.06859004787941654, + "grad_norm": 1.3054215908050537, + "learning_rate": 0.00019770196489956194, + "loss": 0.169, + "step": 2926 + }, + { + "epoch": 0.06861348945422154, + "grad_norm": 3.6158885955810547, + "learning_rate": 0.0001977003945434327, + "loss": 0.7216, + "step": 2927 + }, + { + "epoch": 0.06863693102902653, + "grad_norm": 2.0087435245513916, + "learning_rate": 0.0001976988236571784, + "loss": 1.2246, + "step": 2928 + }, + { + "epoch": 0.06866037260383152, + "grad_norm": 3.7178473472595215, + "learning_rate": 0.0001976972522408075, + "loss": 0.6374, + "step": 2929 + }, + { + "epoch": 0.06868381417863652, + "grad_norm": 3.4921329021453857, + "learning_rate": 0.00019769568029432856, + "loss": 0.3301, + "step": 2930 + }, + { + "epoch": 0.06870725575344151, + "grad_norm": 0.27527105808258057, + "learning_rate": 0.00019769410781775014, + "loss": 0.0309, + "step": 2931 + }, + { + "epoch": 0.0687306973282465, + "grad_norm": 4.371495723724365, + "learning_rate": 0.00019769253481108075, + "loss": 0.7049, + "step": 2932 + }, + { + "epoch": 0.0687541389030515, + "grad_norm": 2.634143114089966, + "learning_rate": 0.00019769096127432887, + "loss": 0.5173, + "step": 2933 + }, + { + "epoch": 0.06877758047785651, + "grad_norm": 2.957249164581299, + "learning_rate": 0.00019768938720750312, + "loss": 0.4828, + "step": 2934 + }, + { + "epoch": 0.0688010220526615, + "grad_norm": 2.6922013759613037, + "learning_rate": 0.000197687812610612, + "loss": 0.3175, + "step": 2935 + }, + { + "epoch": 0.0688244636274665, + "grad_norm": 1.716949462890625, + "learning_rate": 0.00019768623748366406, + "loss": 0.8222, + "step": 2936 + }, + { + "epoch": 0.06884790520227149, + "grad_norm": 2.953608751296997, + "learning_rate": 0.00019768466182666785, + "loss": 0.5501, + "step": 2937 + }, + { + "epoch": 0.06887134677707649, + "grad_norm": 5.45571231842041, + "learning_rate": 0.00019768308563963193, + "loss": 0.6719, + "step": 2938 + }, + { + "epoch": 0.06889478835188148, + "grad_norm": 3.1674087047576904, + "learning_rate": 0.00019768150892256485, + "loss": 0.59, + "step": 2939 + }, + { + "epoch": 0.06891822992668648, + "grad_norm": 4.991539001464844, + "learning_rate": 0.0001976799316754751, + "loss": 0.7738, + "step": 2940 + }, + { + "epoch": 0.06894167150149147, + "grad_norm": 3.3713269233703613, + "learning_rate": 0.00019767835389837134, + "loss": 0.5255, + "step": 2941 + }, + { + "epoch": 0.06896511307629646, + "grad_norm": 3.707261562347412, + "learning_rate": 0.00019767677559126212, + "loss": 0.5776, + "step": 2942 + }, + { + "epoch": 0.06898855465110146, + "grad_norm": 0.7325432300567627, + "learning_rate": 0.0001976751967541559, + "loss": 0.1364, + "step": 2943 + }, + { + "epoch": 0.06901199622590645, + "grad_norm": 3.2982099056243896, + "learning_rate": 0.00019767361738706136, + "loss": 0.3753, + "step": 2944 + }, + { + "epoch": 0.06903543780071145, + "grad_norm": 2.1181015968322754, + "learning_rate": 0.00019767203748998703, + "loss": 1.5177, + "step": 2945 + }, + { + "epoch": 0.06905887937551644, + "grad_norm": 1.9304958581924438, + "learning_rate": 0.00019767045706294142, + "loss": 0.4745, + "step": 2946 + }, + { + "epoch": 0.06908232095032145, + "grad_norm": 2.86069393157959, + "learning_rate": 0.00019766887610593324, + "loss": 0.4247, + "step": 2947 + }, + { + "epoch": 0.06910576252512644, + "grad_norm": 3.716965675354004, + "learning_rate": 0.00019766729461897093, + "loss": 0.5854, + "step": 2948 + }, + { + "epoch": 0.06912920409993144, + "grad_norm": 2.769413948059082, + "learning_rate": 0.00019766571260206316, + "loss": 1.6917, + "step": 2949 + }, + { + "epoch": 0.06915264567473643, + "grad_norm": 3.1892249584198, + "learning_rate": 0.00019766413005521848, + "loss": 0.2629, + "step": 2950 + }, + { + "epoch": 0.06917608724954143, + "grad_norm": 1.369735598564148, + "learning_rate": 0.00019766254697844548, + "loss": 0.1969, + "step": 2951 + }, + { + "epoch": 0.06919952882434642, + "grad_norm": 0.5488333106040955, + "learning_rate": 0.00019766096337175276, + "loss": 0.0769, + "step": 2952 + }, + { + "epoch": 0.06922297039915142, + "grad_norm": 5.152007102966309, + "learning_rate": 0.0001976593792351489, + "loss": 0.6215, + "step": 2953 + }, + { + "epoch": 0.06924641197395641, + "grad_norm": 4.067293643951416, + "learning_rate": 0.00019765779456864255, + "loss": 0.5782, + "step": 2954 + }, + { + "epoch": 0.0692698535487614, + "grad_norm": 3.5846822261810303, + "learning_rate": 0.00019765620937224222, + "loss": 0.569, + "step": 2955 + }, + { + "epoch": 0.0692932951235664, + "grad_norm": 3.4558541774749756, + "learning_rate": 0.00019765462364595656, + "loss": 0.431, + "step": 2956 + }, + { + "epoch": 0.06931673669837139, + "grad_norm": 3.4370455741882324, + "learning_rate": 0.00019765303738979414, + "loss": 0.467, + "step": 2957 + }, + { + "epoch": 0.06934017827317639, + "grad_norm": 4.677647113800049, + "learning_rate": 0.00019765145060376363, + "loss": 0.8636, + "step": 2958 + }, + { + "epoch": 0.06936361984798138, + "grad_norm": 1.6566563844680786, + "learning_rate": 0.00019764986328787358, + "loss": 1.3419, + "step": 2959 + }, + { + "epoch": 0.06938706142278639, + "grad_norm": 4.213840484619141, + "learning_rate": 0.00019764827544213262, + "loss": 0.5964, + "step": 2960 + }, + { + "epoch": 0.06941050299759138, + "grad_norm": 0.9653413891792297, + "learning_rate": 0.00019764668706654938, + "loss": 0.1661, + "step": 2961 + }, + { + "epoch": 0.06943394457239638, + "grad_norm": 1.18633234500885, + "learning_rate": 0.0001976450981611325, + "loss": 0.1525, + "step": 2962 + }, + { + "epoch": 0.06945738614720137, + "grad_norm": 5.891364574432373, + "learning_rate": 0.00019764350872589055, + "loss": 0.8485, + "step": 2963 + }, + { + "epoch": 0.06948082772200637, + "grad_norm": 2.5166444778442383, + "learning_rate": 0.00019764191876083217, + "loss": 0.2279, + "step": 2964 + }, + { + "epoch": 0.06950426929681136, + "grad_norm": 3.0878448486328125, + "learning_rate": 0.000197640328265966, + "loss": 0.4182, + "step": 2965 + }, + { + "epoch": 0.06952771087161635, + "grad_norm": 5.225889205932617, + "learning_rate": 0.00019763873724130068, + "loss": 0.7745, + "step": 2966 + }, + { + "epoch": 0.06955115244642135, + "grad_norm": 4.922793865203857, + "learning_rate": 0.0001976371456868448, + "loss": 0.6813, + "step": 2967 + }, + { + "epoch": 0.06957459402122634, + "grad_norm": 3.8181381225585938, + "learning_rate": 0.00019763555360260702, + "loss": 0.4391, + "step": 2968 + }, + { + "epoch": 0.06959803559603134, + "grad_norm": 3.6696174144744873, + "learning_rate": 0.000197633960988596, + "loss": 0.783, + "step": 2969 + }, + { + "epoch": 0.06962147717083633, + "grad_norm": 3.6537373065948486, + "learning_rate": 0.00019763236784482033, + "loss": 0.3932, + "step": 2970 + }, + { + "epoch": 0.06964491874564133, + "grad_norm": 5.095301151275635, + "learning_rate": 0.0001976307741712887, + "loss": 0.7458, + "step": 2971 + }, + { + "epoch": 0.06966836032044632, + "grad_norm": 3.4317898750305176, + "learning_rate": 0.00019762917996800976, + "loss": 0.5771, + "step": 2972 + }, + { + "epoch": 0.06969180189525133, + "grad_norm": 3.734499931335449, + "learning_rate": 0.00019762758523499216, + "loss": 0.2712, + "step": 2973 + }, + { + "epoch": 0.06971524347005632, + "grad_norm": 5.000802040100098, + "learning_rate": 0.0001976259899722445, + "loss": 0.6229, + "step": 2974 + }, + { + "epoch": 0.06973868504486132, + "grad_norm": 0.8262157440185547, + "learning_rate": 0.0001976243941797755, + "loss": 0.1404, + "step": 2975 + }, + { + "epoch": 0.06976212661966631, + "grad_norm": 3.0640738010406494, + "learning_rate": 0.0001976227978575938, + "loss": 0.3806, + "step": 2976 + }, + { + "epoch": 0.0697855681944713, + "grad_norm": 3.733311176300049, + "learning_rate": 0.000197621201005708, + "loss": 0.5836, + "step": 2977 + }, + { + "epoch": 0.0698090097692763, + "grad_norm": 4.138223171234131, + "learning_rate": 0.00019761960362412687, + "loss": 0.9522, + "step": 2978 + }, + { + "epoch": 0.0698324513440813, + "grad_norm": 3.891056537628174, + "learning_rate": 0.000197618005712859, + "loss": 0.8034, + "step": 2979 + }, + { + "epoch": 0.06985589291888629, + "grad_norm": 3.294159412384033, + "learning_rate": 0.0001976164072719131, + "loss": 0.5301, + "step": 2980 + }, + { + "epoch": 0.06987933449369128, + "grad_norm": 3.231158494949341, + "learning_rate": 0.0001976148083012978, + "loss": 0.5191, + "step": 2981 + }, + { + "epoch": 0.06990277606849628, + "grad_norm": 4.1877546310424805, + "learning_rate": 0.00019761320880102185, + "loss": 0.4732, + "step": 2982 + }, + { + "epoch": 0.06992621764330127, + "grad_norm": 4.249898910522461, + "learning_rate": 0.00019761160877109388, + "loss": 0.7966, + "step": 2983 + }, + { + "epoch": 0.06994965921810627, + "grad_norm": 4.012451648712158, + "learning_rate": 0.00019761000821152255, + "loss": 0.7804, + "step": 2984 + }, + { + "epoch": 0.06997310079291127, + "grad_norm": 4.0353569984436035, + "learning_rate": 0.0001976084071223166, + "loss": 0.4599, + "step": 2985 + }, + { + "epoch": 0.06999654236771627, + "grad_norm": 3.0558290481567383, + "learning_rate": 0.00019760680550348468, + "loss": 0.3447, + "step": 2986 + }, + { + "epoch": 0.07001998394252126, + "grad_norm": 1.8091106414794922, + "learning_rate": 0.00019760520335503548, + "loss": 0.1838, + "step": 2987 + }, + { + "epoch": 0.07004342551732626, + "grad_norm": 2.7514233589172363, + "learning_rate": 0.00019760360067697767, + "loss": 1.489, + "step": 2988 + }, + { + "epoch": 0.07006686709213125, + "grad_norm": 2.7113664150238037, + "learning_rate": 0.00019760199746932003, + "loss": 0.5075, + "step": 2989 + }, + { + "epoch": 0.07009030866693625, + "grad_norm": 5.1008219718933105, + "learning_rate": 0.00019760039373207118, + "loss": 0.5204, + "step": 2990 + }, + { + "epoch": 0.07011375024174124, + "grad_norm": 0.8950518369674683, + "learning_rate": 0.00019759878946523985, + "loss": 0.121, + "step": 2991 + }, + { + "epoch": 0.07013719181654623, + "grad_norm": 3.0973052978515625, + "learning_rate": 0.00019759718466883473, + "loss": 1.7892, + "step": 2992 + }, + { + "epoch": 0.07016063339135123, + "grad_norm": 0.39195963740348816, + "learning_rate": 0.0001975955793428646, + "loss": 0.058, + "step": 2993 + }, + { + "epoch": 0.07018407496615622, + "grad_norm": 0.556922972202301, + "learning_rate": 0.00019759397348733808, + "loss": 0.0417, + "step": 2994 + }, + { + "epoch": 0.07020751654096122, + "grad_norm": 0.8723920583724976, + "learning_rate": 0.00019759236710226389, + "loss": 0.0945, + "step": 2995 + }, + { + "epoch": 0.07023095811576621, + "grad_norm": 3.250676155090332, + "learning_rate": 0.00019759076018765082, + "loss": 0.6703, + "step": 2996 + }, + { + "epoch": 0.0702543996905712, + "grad_norm": 1.3584349155426025, + "learning_rate": 0.00019758915274350748, + "loss": 0.1881, + "step": 2997 + }, + { + "epoch": 0.07027784126537621, + "grad_norm": 1.3496168851852417, + "learning_rate": 0.00019758754476984273, + "loss": 0.1844, + "step": 2998 + }, + { + "epoch": 0.07030128284018121, + "grad_norm": 5.962337970733643, + "learning_rate": 0.00019758593626666515, + "loss": 0.9805, + "step": 2999 + }, + { + "epoch": 0.0703247244149862, + "grad_norm": 4.388025283813477, + "learning_rate": 0.00019758432723398353, + "loss": 0.5056, + "step": 3000 + }, + { + "epoch": 0.0703481659897912, + "grad_norm": 4.204789161682129, + "learning_rate": 0.00019758271767180666, + "loss": 0.8193, + "step": 3001 + }, + { + "epoch": 0.07037160756459619, + "grad_norm": 5.02403450012207, + "learning_rate": 0.0001975811075801432, + "loss": 0.7917, + "step": 3002 + }, + { + "epoch": 0.07039504913940119, + "grad_norm": 3.6647329330444336, + "learning_rate": 0.00019757949695900192, + "loss": 0.6399, + "step": 3003 + }, + { + "epoch": 0.07041849071420618, + "grad_norm": 3.622697591781616, + "learning_rate": 0.0001975778858083915, + "loss": 0.7355, + "step": 3004 + }, + { + "epoch": 0.07044193228901117, + "grad_norm": 3.0928235054016113, + "learning_rate": 0.00019757627412832077, + "loss": 0.7224, + "step": 3005 + }, + { + "epoch": 0.07046537386381617, + "grad_norm": 2.9755699634552, + "learning_rate": 0.00019757466191879842, + "loss": 0.4756, + "step": 3006 + }, + { + "epoch": 0.07048881543862116, + "grad_norm": 3.1342933177948, + "learning_rate": 0.00019757304917983323, + "loss": 0.504, + "step": 3007 + }, + { + "epoch": 0.07051225701342616, + "grad_norm": 3.270977020263672, + "learning_rate": 0.0001975714359114339, + "loss": 0.6719, + "step": 3008 + }, + { + "epoch": 0.07053569858823115, + "grad_norm": 3.8993260860443115, + "learning_rate": 0.00019756982211360924, + "loss": 0.5743, + "step": 3009 + }, + { + "epoch": 0.07055914016303615, + "grad_norm": 1.3528984785079956, + "learning_rate": 0.00019756820778636796, + "loss": 0.2751, + "step": 3010 + }, + { + "epoch": 0.07058258173784115, + "grad_norm": 1.3790682554244995, + "learning_rate": 0.00019756659292971887, + "loss": 0.1708, + "step": 3011 + }, + { + "epoch": 0.07060602331264615, + "grad_norm": 0.9605859518051147, + "learning_rate": 0.00019756497754367067, + "loss": 0.1349, + "step": 3012 + }, + { + "epoch": 0.07062946488745114, + "grad_norm": 3.77885103225708, + "learning_rate": 0.00019756336162823218, + "loss": 0.5192, + "step": 3013 + }, + { + "epoch": 0.07065290646225614, + "grad_norm": 1.3997209072113037, + "learning_rate": 0.00019756174518341214, + "loss": 0.2088, + "step": 3014 + }, + { + "epoch": 0.07067634803706113, + "grad_norm": 2.1074397563934326, + "learning_rate": 0.00019756012820921934, + "loss": 1.5353, + "step": 3015 + }, + { + "epoch": 0.07069978961186613, + "grad_norm": 1.3665833473205566, + "learning_rate": 0.00019755851070566251, + "loss": 0.2215, + "step": 3016 + }, + { + "epoch": 0.07072323118667112, + "grad_norm": 3.3003695011138916, + "learning_rate": 0.0001975568926727505, + "loss": 0.4103, + "step": 3017 + }, + { + "epoch": 0.07074667276147611, + "grad_norm": 3.2066407203674316, + "learning_rate": 0.000197555274110492, + "loss": 1.4591, + "step": 3018 + }, + { + "epoch": 0.07077011433628111, + "grad_norm": 1.00753653049469, + "learning_rate": 0.0001975536550188959, + "loss": 0.1631, + "step": 3019 + }, + { + "epoch": 0.0707935559110861, + "grad_norm": 0.6809327602386475, + "learning_rate": 0.00019755203539797086, + "loss": 0.138, + "step": 3020 + }, + { + "epoch": 0.0708169974858911, + "grad_norm": 2.8224382400512695, + "learning_rate": 0.00019755041524772575, + "loss": 0.588, + "step": 3021 + }, + { + "epoch": 0.07084043906069609, + "grad_norm": 4.466375827789307, + "learning_rate": 0.00019754879456816937, + "loss": 0.7053, + "step": 3022 + }, + { + "epoch": 0.07086388063550109, + "grad_norm": 3.368137836456299, + "learning_rate": 0.00019754717335931048, + "loss": 0.5293, + "step": 3023 + }, + { + "epoch": 0.0708873222103061, + "grad_norm": 4.50410795211792, + "learning_rate": 0.00019754555162115786, + "loss": 0.8176, + "step": 3024 + }, + { + "epoch": 0.07091076378511109, + "grad_norm": 3.489647626876831, + "learning_rate": 0.00019754392935372037, + "loss": 0.4513, + "step": 3025 + }, + { + "epoch": 0.07093420535991608, + "grad_norm": 1.956491231918335, + "learning_rate": 0.00019754230655700676, + "loss": 1.9841, + "step": 3026 + }, + { + "epoch": 0.07095764693472108, + "grad_norm": 3.153742551803589, + "learning_rate": 0.00019754068323102586, + "loss": 0.5771, + "step": 3027 + }, + { + "epoch": 0.07098108850952607, + "grad_norm": 3.51643967628479, + "learning_rate": 0.00019753905937578645, + "loss": 0.5025, + "step": 3028 + }, + { + "epoch": 0.07100453008433107, + "grad_norm": 4.7173967361450195, + "learning_rate": 0.00019753743499129737, + "loss": 0.4657, + "step": 3029 + }, + { + "epoch": 0.07102797165913606, + "grad_norm": 4.43527889251709, + "learning_rate": 0.00019753581007756743, + "loss": 0.4061, + "step": 3030 + }, + { + "epoch": 0.07105141323394105, + "grad_norm": 1.9354318380355835, + "learning_rate": 0.00019753418463460545, + "loss": 1.399, + "step": 3031 + }, + { + "epoch": 0.07107485480874605, + "grad_norm": 3.5455942153930664, + "learning_rate": 0.00019753255866242022, + "loss": 0.7603, + "step": 3032 + }, + { + "epoch": 0.07109829638355104, + "grad_norm": 3.8498613834381104, + "learning_rate": 0.0001975309321610206, + "loss": 0.4056, + "step": 3033 + }, + { + "epoch": 0.07112173795835604, + "grad_norm": 2.8992443084716797, + "learning_rate": 0.0001975293051304154, + "loss": 0.607, + "step": 3034 + }, + { + "epoch": 0.07114517953316103, + "grad_norm": 0.9622387290000916, + "learning_rate": 0.00019752767757061346, + "loss": 0.1547, + "step": 3035 + }, + { + "epoch": 0.07116862110796603, + "grad_norm": 3.7733664512634277, + "learning_rate": 0.00019752604948162356, + "loss": 0.545, + "step": 3036 + }, + { + "epoch": 0.07119206268277103, + "grad_norm": 3.328186273574829, + "learning_rate": 0.00019752442086345463, + "loss": 0.324, + "step": 3037 + }, + { + "epoch": 0.07121550425757603, + "grad_norm": 3.382830858230591, + "learning_rate": 0.00019752279171611538, + "loss": 0.6605, + "step": 3038 + }, + { + "epoch": 0.07123894583238102, + "grad_norm": 1.0017646551132202, + "learning_rate": 0.00019752116203961478, + "loss": 0.1607, + "step": 3039 + }, + { + "epoch": 0.07126238740718602, + "grad_norm": 1.4820880889892578, + "learning_rate": 0.00019751953183396159, + "loss": 0.1535, + "step": 3040 + }, + { + "epoch": 0.07128582898199101, + "grad_norm": 1.4348431825637817, + "learning_rate": 0.00019751790109916466, + "loss": 0.2261, + "step": 3041 + }, + { + "epoch": 0.071309270556796, + "grad_norm": 4.23916482925415, + "learning_rate": 0.0001975162698352329, + "loss": 0.4846, + "step": 3042 + }, + { + "epoch": 0.071332712131601, + "grad_norm": 4.604820251464844, + "learning_rate": 0.00019751463804217506, + "loss": 1.0394, + "step": 3043 + }, + { + "epoch": 0.071356153706406, + "grad_norm": 3.5017871856689453, + "learning_rate": 0.0001975130057200001, + "loss": 0.6399, + "step": 3044 + }, + { + "epoch": 0.07137959528121099, + "grad_norm": 2.584359884262085, + "learning_rate": 0.0001975113728687168, + "loss": 1.6983, + "step": 3045 + }, + { + "epoch": 0.07140303685601598, + "grad_norm": 1.3371156454086304, + "learning_rate": 0.00019750973948833407, + "loss": 0.2912, + "step": 3046 + }, + { + "epoch": 0.07142647843082098, + "grad_norm": 4.373763084411621, + "learning_rate": 0.00019750810557886075, + "loss": 0.6574, + "step": 3047 + }, + { + "epoch": 0.07144992000562597, + "grad_norm": 3.109647750854492, + "learning_rate": 0.0001975064711403057, + "loss": 0.6505, + "step": 3048 + }, + { + "epoch": 0.07147336158043098, + "grad_norm": 5.997776985168457, + "learning_rate": 0.00019750483617267782, + "loss": 1.3422, + "step": 3049 + }, + { + "epoch": 0.07149680315523597, + "grad_norm": 3.013113498687744, + "learning_rate": 0.00019750320067598592, + "loss": 1.5421, + "step": 3050 + }, + { + "epoch": 0.07152024473004097, + "grad_norm": 4.637538909912109, + "learning_rate": 0.0001975015646502389, + "loss": 0.9325, + "step": 3051 + }, + { + "epoch": 0.07154368630484596, + "grad_norm": 0.47693949937820435, + "learning_rate": 0.0001974999280954457, + "loss": 0.0593, + "step": 3052 + }, + { + "epoch": 0.07156712787965096, + "grad_norm": 3.3069539070129395, + "learning_rate": 0.0001974982910116151, + "loss": 0.6779, + "step": 3053 + }, + { + "epoch": 0.07159056945445595, + "grad_norm": 4.041444301605225, + "learning_rate": 0.00019749665339875607, + "loss": 0.7612, + "step": 3054 + }, + { + "epoch": 0.07161401102926095, + "grad_norm": 4.038890361785889, + "learning_rate": 0.00019749501525687745, + "loss": 0.5013, + "step": 3055 + }, + { + "epoch": 0.07163745260406594, + "grad_norm": 2.8629343509674072, + "learning_rate": 0.00019749337658598814, + "loss": 0.5982, + "step": 3056 + }, + { + "epoch": 0.07166089417887093, + "grad_norm": 3.128952980041504, + "learning_rate": 0.000197491737386097, + "loss": 0.5001, + "step": 3057 + }, + { + "epoch": 0.07168433575367593, + "grad_norm": 4.318533420562744, + "learning_rate": 0.00019749009765721297, + "loss": 0.4916, + "step": 3058 + }, + { + "epoch": 0.07170777732848092, + "grad_norm": 4.414656639099121, + "learning_rate": 0.0001974884573993449, + "loss": 0.6219, + "step": 3059 + }, + { + "epoch": 0.07173121890328592, + "grad_norm": 3.65110445022583, + "learning_rate": 0.00019748681661250177, + "loss": 0.6665, + "step": 3060 + }, + { + "epoch": 0.07175466047809091, + "grad_norm": 3.9050166606903076, + "learning_rate": 0.00019748517529669243, + "loss": 0.4731, + "step": 3061 + }, + { + "epoch": 0.07177810205289592, + "grad_norm": 1.4175448417663574, + "learning_rate": 0.00019748353345192573, + "loss": 0.2497, + "step": 3062 + }, + { + "epoch": 0.07180154362770091, + "grad_norm": 4.135795593261719, + "learning_rate": 0.00019748189107821068, + "loss": 0.9301, + "step": 3063 + }, + { + "epoch": 0.07182498520250591, + "grad_norm": 1.9587006568908691, + "learning_rate": 0.00019748024817555618, + "loss": 0.1317, + "step": 3064 + }, + { + "epoch": 0.0718484267773109, + "grad_norm": 1.103444218635559, + "learning_rate": 0.00019747860474397103, + "loss": 0.1245, + "step": 3065 + }, + { + "epoch": 0.0718718683521159, + "grad_norm": 0.8472976088523865, + "learning_rate": 0.0001974769607834643, + "loss": 0.1446, + "step": 3066 + }, + { + "epoch": 0.07189530992692089, + "grad_norm": 2.5073962211608887, + "learning_rate": 0.00019747531629404482, + "loss": 0.3007, + "step": 3067 + }, + { + "epoch": 0.07191875150172589, + "grad_norm": 3.474062204360962, + "learning_rate": 0.00019747367127572152, + "loss": 0.7175, + "step": 3068 + }, + { + "epoch": 0.07194219307653088, + "grad_norm": 3.0705368518829346, + "learning_rate": 0.00019747202572850334, + "loss": 0.2632, + "step": 3069 + }, + { + "epoch": 0.07196563465133587, + "grad_norm": 5.5022430419921875, + "learning_rate": 0.00019747037965239922, + "loss": 0.6514, + "step": 3070 + }, + { + "epoch": 0.07198907622614087, + "grad_norm": 3.9656198024749756, + "learning_rate": 0.00019746873304741806, + "loss": 0.6846, + "step": 3071 + }, + { + "epoch": 0.07201251780094586, + "grad_norm": 2.293653964996338, + "learning_rate": 0.00019746708591356882, + "loss": 0.382, + "step": 3072 + }, + { + "epoch": 0.07203595937575086, + "grad_norm": 3.453681230545044, + "learning_rate": 0.00019746543825086045, + "loss": 0.4797, + "step": 3073 + }, + { + "epoch": 0.07205940095055585, + "grad_norm": 0.8164797425270081, + "learning_rate": 0.00019746379005930185, + "loss": 0.1147, + "step": 3074 + }, + { + "epoch": 0.07208284252536086, + "grad_norm": 1.4320755004882812, + "learning_rate": 0.000197462141338902, + "loss": 0.1956, + "step": 3075 + }, + { + "epoch": 0.07210628410016585, + "grad_norm": 1.810314416885376, + "learning_rate": 0.00019746049208966982, + "loss": 1.2195, + "step": 3076 + }, + { + "epoch": 0.07212972567497085, + "grad_norm": 1.1963770389556885, + "learning_rate": 0.00019745884231161428, + "loss": 0.1805, + "step": 3077 + }, + { + "epoch": 0.07215316724977584, + "grad_norm": 1.492842435836792, + "learning_rate": 0.0001974571920047443, + "loss": 0.2607, + "step": 3078 + }, + { + "epoch": 0.07217660882458084, + "grad_norm": 0.9178451895713806, + "learning_rate": 0.0001974555411690689, + "loss": 0.114, + "step": 3079 + }, + { + "epoch": 0.07220005039938583, + "grad_norm": 5.222541809082031, + "learning_rate": 0.00019745388980459694, + "loss": 0.8084, + "step": 3080 + }, + { + "epoch": 0.07222349197419083, + "grad_norm": 3.8335201740264893, + "learning_rate": 0.00019745223791133745, + "loss": 0.3655, + "step": 3081 + }, + { + "epoch": 0.07224693354899582, + "grad_norm": 3.0693817138671875, + "learning_rate": 0.0001974505854892994, + "loss": 0.2714, + "step": 3082 + }, + { + "epoch": 0.07227037512380081, + "grad_norm": 1.1670581102371216, + "learning_rate": 0.0001974489325384917, + "loss": 0.2572, + "step": 3083 + }, + { + "epoch": 0.07229381669860581, + "grad_norm": 1.600968599319458, + "learning_rate": 0.00019744727905892338, + "loss": 0.2483, + "step": 3084 + }, + { + "epoch": 0.0723172582734108, + "grad_norm": 4.0001959800720215, + "learning_rate": 0.00019744562505060338, + "loss": 0.5, + "step": 3085 + }, + { + "epoch": 0.0723406998482158, + "grad_norm": 4.454598903656006, + "learning_rate": 0.00019744397051354065, + "loss": 0.5815, + "step": 3086 + }, + { + "epoch": 0.07236414142302079, + "grad_norm": 3.403995990753174, + "learning_rate": 0.00019744231544774422, + "loss": 1.8054, + "step": 3087 + }, + { + "epoch": 0.0723875829978258, + "grad_norm": 5.189001560211182, + "learning_rate": 0.00019744065985322307, + "loss": 0.6136, + "step": 3088 + }, + { + "epoch": 0.0724110245726308, + "grad_norm": 1.3006565570831299, + "learning_rate": 0.00019743900372998612, + "loss": 0.1732, + "step": 3089 + }, + { + "epoch": 0.07243446614743579, + "grad_norm": 4.76231575012207, + "learning_rate": 0.00019743734707804243, + "loss": 0.5601, + "step": 3090 + }, + { + "epoch": 0.07245790772224078, + "grad_norm": 1.4000625610351562, + "learning_rate": 0.00019743568989740094, + "loss": 0.2208, + "step": 3091 + }, + { + "epoch": 0.07248134929704578, + "grad_norm": 2.7311699390411377, + "learning_rate": 0.00019743403218807064, + "loss": 0.4315, + "step": 3092 + }, + { + "epoch": 0.07250479087185077, + "grad_norm": 3.2185447216033936, + "learning_rate": 0.00019743237395006055, + "loss": 0.3485, + "step": 3093 + }, + { + "epoch": 0.07252823244665577, + "grad_norm": 2.907665967941284, + "learning_rate": 0.00019743071518337968, + "loss": 2.0075, + "step": 3094 + }, + { + "epoch": 0.07255167402146076, + "grad_norm": 1.953390121459961, + "learning_rate": 0.000197429055888037, + "loss": 0.4199, + "step": 3095 + }, + { + "epoch": 0.07257511559626575, + "grad_norm": 3.437028408050537, + "learning_rate": 0.0001974273960640415, + "loss": 0.5512, + "step": 3096 + }, + { + "epoch": 0.07259855717107075, + "grad_norm": 0.6124019026756287, + "learning_rate": 0.00019742573571140222, + "loss": 0.0939, + "step": 3097 + }, + { + "epoch": 0.07262199874587574, + "grad_norm": 3.409069061279297, + "learning_rate": 0.00019742407483012818, + "loss": 0.4749, + "step": 3098 + }, + { + "epoch": 0.07264544032068074, + "grad_norm": 2.9551782608032227, + "learning_rate": 0.00019742241342022836, + "loss": 0.5966, + "step": 3099 + }, + { + "epoch": 0.07266888189548573, + "grad_norm": 3.9307701587677, + "learning_rate": 0.00019742075148171178, + "loss": 0.5485, + "step": 3100 + }, + { + "epoch": 0.07269232347029074, + "grad_norm": 1.0025194883346558, + "learning_rate": 0.00019741908901458745, + "loss": 0.208, + "step": 3101 + }, + { + "epoch": 0.07271576504509573, + "grad_norm": 1.9326555728912354, + "learning_rate": 0.00019741742601886442, + "loss": 0.2698, + "step": 3102 + }, + { + "epoch": 0.07273920661990073, + "grad_norm": 3.324552536010742, + "learning_rate": 0.0001974157624945517, + "loss": 0.4389, + "step": 3103 + }, + { + "epoch": 0.07276264819470572, + "grad_norm": 1.5904457569122314, + "learning_rate": 0.0001974140984416583, + "loss": 0.1827, + "step": 3104 + }, + { + "epoch": 0.07278608976951072, + "grad_norm": 1.125314712524414, + "learning_rate": 0.00019741243386019327, + "loss": 0.1829, + "step": 3105 + }, + { + "epoch": 0.07280953134431571, + "grad_norm": 1.440781831741333, + "learning_rate": 0.00019741076875016565, + "loss": 0.1805, + "step": 3106 + }, + { + "epoch": 0.0728329729191207, + "grad_norm": 5.86574649810791, + "learning_rate": 0.00019740910311158444, + "loss": 0.5502, + "step": 3107 + }, + { + "epoch": 0.0728564144939257, + "grad_norm": 1.4064639806747437, + "learning_rate": 0.0001974074369444587, + "loss": 0.0917, + "step": 3108 + }, + { + "epoch": 0.0728798560687307, + "grad_norm": 4.474391460418701, + "learning_rate": 0.00019740577024879747, + "loss": 0.7545, + "step": 3109 + }, + { + "epoch": 0.07290329764353569, + "grad_norm": 2.9078257083892822, + "learning_rate": 0.00019740410302460982, + "loss": 0.6297, + "step": 3110 + }, + { + "epoch": 0.07292673921834068, + "grad_norm": 4.592953681945801, + "learning_rate": 0.00019740243527190473, + "loss": 0.8527, + "step": 3111 + }, + { + "epoch": 0.07295018079314568, + "grad_norm": 5.329082489013672, + "learning_rate": 0.0001974007669906913, + "loss": 1.2022, + "step": 3112 + }, + { + "epoch": 0.07297362236795069, + "grad_norm": 3.8359715938568115, + "learning_rate": 0.00019739909818097858, + "loss": 0.7139, + "step": 3113 + }, + { + "epoch": 0.07299706394275568, + "grad_norm": 2.658653974533081, + "learning_rate": 0.00019739742884277562, + "loss": 0.5105, + "step": 3114 + }, + { + "epoch": 0.07302050551756067, + "grad_norm": 4.623725414276123, + "learning_rate": 0.00019739575897609145, + "loss": 0.766, + "step": 3115 + }, + { + "epoch": 0.07304394709236567, + "grad_norm": 4.968339920043945, + "learning_rate": 0.00019739408858093516, + "loss": 0.656, + "step": 3116 + }, + { + "epoch": 0.07306738866717066, + "grad_norm": 3.780139684677124, + "learning_rate": 0.0001973924176573158, + "loss": 0.9216, + "step": 3117 + }, + { + "epoch": 0.07309083024197566, + "grad_norm": 1.8188080787658691, + "learning_rate": 0.00019739074620524246, + "loss": 1.1609, + "step": 3118 + }, + { + "epoch": 0.07311427181678065, + "grad_norm": 0.7982780933380127, + "learning_rate": 0.00019738907422472418, + "loss": 0.1458, + "step": 3119 + }, + { + "epoch": 0.07313771339158565, + "grad_norm": 1.2073945999145508, + "learning_rate": 0.00019738740171577004, + "loss": 0.209, + "step": 3120 + }, + { + "epoch": 0.07316115496639064, + "grad_norm": 6.205429553985596, + "learning_rate": 0.00019738572867838912, + "loss": 0.8449, + "step": 3121 + }, + { + "epoch": 0.07318459654119563, + "grad_norm": 6.546579837799072, + "learning_rate": 0.0001973840551125905, + "loss": 1.1427, + "step": 3122 + }, + { + "epoch": 0.07320803811600063, + "grad_norm": 2.6314432621002197, + "learning_rate": 0.00019738238101838328, + "loss": 0.6124, + "step": 3123 + }, + { + "epoch": 0.07323147969080562, + "grad_norm": 4.407940864562988, + "learning_rate": 0.0001973807063957765, + "loss": 0.8049, + "step": 3124 + }, + { + "epoch": 0.07325492126561062, + "grad_norm": 3.0560076236724854, + "learning_rate": 0.00019737903124477927, + "loss": 0.3649, + "step": 3125 + }, + { + "epoch": 0.07327836284041563, + "grad_norm": 3.2333931922912598, + "learning_rate": 0.00019737735556540068, + "loss": 0.4966, + "step": 3126 + }, + { + "epoch": 0.07330180441522062, + "grad_norm": 1.1359894275665283, + "learning_rate": 0.00019737567935764985, + "loss": 0.1423, + "step": 3127 + }, + { + "epoch": 0.07332524599002561, + "grad_norm": 4.3668293952941895, + "learning_rate": 0.00019737400262153583, + "loss": 0.53, + "step": 3128 + }, + { + "epoch": 0.07334868756483061, + "grad_norm": 1.2722774744033813, + "learning_rate": 0.00019737232535706772, + "loss": 1.4303, + "step": 3129 + }, + { + "epoch": 0.0733721291396356, + "grad_norm": 1.3428926467895508, + "learning_rate": 0.00019737064756425462, + "loss": 0.1706, + "step": 3130 + }, + { + "epoch": 0.0733955707144406, + "grad_norm": 3.68030047416687, + "learning_rate": 0.00019736896924310568, + "loss": 0.5187, + "step": 3131 + }, + { + "epoch": 0.07341901228924559, + "grad_norm": 2.057183265686035, + "learning_rate": 0.00019736729039362998, + "loss": 0.3643, + "step": 3132 + }, + { + "epoch": 0.07344245386405059, + "grad_norm": 4.432425022125244, + "learning_rate": 0.0001973656110158366, + "loss": 0.4645, + "step": 3133 + }, + { + "epoch": 0.07346589543885558, + "grad_norm": 1.1668906211853027, + "learning_rate": 0.00019736393110973468, + "loss": 0.1492, + "step": 3134 + }, + { + "epoch": 0.07348933701366057, + "grad_norm": 2.797551155090332, + "learning_rate": 0.00019736225067533333, + "loss": 0.3903, + "step": 3135 + }, + { + "epoch": 0.07351277858846557, + "grad_norm": 4.0605034828186035, + "learning_rate": 0.00019736056971264167, + "loss": 0.4214, + "step": 3136 + }, + { + "epoch": 0.07353622016327056, + "grad_norm": 3.4964263439178467, + "learning_rate": 0.00019735888822166887, + "loss": 0.4564, + "step": 3137 + }, + { + "epoch": 0.07355966173807556, + "grad_norm": 3.735187530517578, + "learning_rate": 0.00019735720620242395, + "loss": 0.3523, + "step": 3138 + }, + { + "epoch": 0.07358310331288057, + "grad_norm": 3.724329948425293, + "learning_rate": 0.00019735552365491608, + "loss": 0.7037, + "step": 3139 + }, + { + "epoch": 0.07360654488768556, + "grad_norm": 2.493283987045288, + "learning_rate": 0.00019735384057915445, + "loss": 0.4919, + "step": 3140 + }, + { + "epoch": 0.07362998646249055, + "grad_norm": 4.743603706359863, + "learning_rate": 0.0001973521569751481, + "loss": 0.8209, + "step": 3141 + }, + { + "epoch": 0.07365342803729555, + "grad_norm": 6.867105484008789, + "learning_rate": 0.00019735047284290624, + "loss": 0.8248, + "step": 3142 + }, + { + "epoch": 0.07367686961210054, + "grad_norm": 2.830397605895996, + "learning_rate": 0.00019734878818243799, + "loss": 0.6574, + "step": 3143 + }, + { + "epoch": 0.07370031118690554, + "grad_norm": 2.156930446624756, + "learning_rate": 0.00019734710299375242, + "loss": 0.2522, + "step": 3144 + }, + { + "epoch": 0.07372375276171053, + "grad_norm": 2.7094829082489014, + "learning_rate": 0.00019734541727685877, + "loss": 1.9492, + "step": 3145 + }, + { + "epoch": 0.07374719433651553, + "grad_norm": 4.757994651794434, + "learning_rate": 0.00019734373103176615, + "loss": 0.3741, + "step": 3146 + }, + { + "epoch": 0.07377063591132052, + "grad_norm": 3.333188772201538, + "learning_rate": 0.00019734204425848371, + "loss": 0.5884, + "step": 3147 + }, + { + "epoch": 0.07379407748612551, + "grad_norm": 5.421635150909424, + "learning_rate": 0.0001973403569570206, + "loss": 0.3792, + "step": 3148 + }, + { + "epoch": 0.07381751906093051, + "grad_norm": 3.3976573944091797, + "learning_rate": 0.00019733866912738596, + "loss": 0.4051, + "step": 3149 + }, + { + "epoch": 0.0738409606357355, + "grad_norm": 1.5270123481750488, + "learning_rate": 0.00019733698076958897, + "loss": 0.3055, + "step": 3150 + }, + { + "epoch": 0.0738644022105405, + "grad_norm": 3.444303035736084, + "learning_rate": 0.0001973352918836388, + "loss": 0.4035, + "step": 3151 + }, + { + "epoch": 0.0738878437853455, + "grad_norm": 3.417583465576172, + "learning_rate": 0.0001973336024695446, + "loss": 0.2413, + "step": 3152 + }, + { + "epoch": 0.0739112853601505, + "grad_norm": 3.751873254776001, + "learning_rate": 0.0001973319125273155, + "loss": 0.6936, + "step": 3153 + }, + { + "epoch": 0.0739347269349555, + "grad_norm": 1.4764724969863892, + "learning_rate": 0.00019733022205696075, + "loss": 0.2597, + "step": 3154 + }, + { + "epoch": 0.07395816850976049, + "grad_norm": 3.4072487354278564, + "learning_rate": 0.00019732853105848945, + "loss": 0.4092, + "step": 3155 + }, + { + "epoch": 0.07398161008456548, + "grad_norm": 1.004825472831726, + "learning_rate": 0.00019732683953191082, + "loss": 0.1728, + "step": 3156 + }, + { + "epoch": 0.07400505165937048, + "grad_norm": 3.0340769290924072, + "learning_rate": 0.000197325147477234, + "loss": 0.5707, + "step": 3157 + }, + { + "epoch": 0.07402849323417547, + "grad_norm": 2.8767755031585693, + "learning_rate": 0.00019732345489446824, + "loss": 0.4764, + "step": 3158 + }, + { + "epoch": 0.07405193480898047, + "grad_norm": 4.134655952453613, + "learning_rate": 0.00019732176178362262, + "loss": 0.5773, + "step": 3159 + }, + { + "epoch": 0.07407537638378546, + "grad_norm": 0.937510073184967, + "learning_rate": 0.00019732006814470642, + "loss": 0.1817, + "step": 3160 + }, + { + "epoch": 0.07409881795859045, + "grad_norm": 4.054699420928955, + "learning_rate": 0.00019731837397772881, + "loss": 1.7219, + "step": 3161 + }, + { + "epoch": 0.07412225953339545, + "grad_norm": 2.5951054096221924, + "learning_rate": 0.00019731667928269892, + "loss": 0.4521, + "step": 3162 + }, + { + "epoch": 0.07414570110820044, + "grad_norm": 3.044511079788208, + "learning_rate": 0.000197314984059626, + "loss": 0.4031, + "step": 3163 + }, + { + "epoch": 0.07416914268300545, + "grad_norm": 5.586264610290527, + "learning_rate": 0.00019731328830851927, + "loss": 1.5004, + "step": 3164 + }, + { + "epoch": 0.07419258425781045, + "grad_norm": 3.0303897857666016, + "learning_rate": 0.0001973115920293879, + "loss": 0.5531, + "step": 3165 + }, + { + "epoch": 0.07421602583261544, + "grad_norm": 3.060863971710205, + "learning_rate": 0.00019730989522224107, + "loss": 0.5138, + "step": 3166 + }, + { + "epoch": 0.07423946740742043, + "grad_norm": 1.8784626722335815, + "learning_rate": 0.000197308197887088, + "loss": 0.296, + "step": 3167 + }, + { + "epoch": 0.07426290898222543, + "grad_norm": 1.0057820081710815, + "learning_rate": 0.00019730650002393796, + "loss": 0.1029, + "step": 3168 + }, + { + "epoch": 0.07428635055703042, + "grad_norm": 3.1999058723449707, + "learning_rate": 0.00019730480163280008, + "loss": 0.3551, + "step": 3169 + }, + { + "epoch": 0.07430979213183542, + "grad_norm": 1.2579803466796875, + "learning_rate": 0.0001973031027136836, + "loss": 0.2293, + "step": 3170 + }, + { + "epoch": 0.07433323370664041, + "grad_norm": 3.5483362674713135, + "learning_rate": 0.0001973014032665978, + "loss": 0.7492, + "step": 3171 + }, + { + "epoch": 0.0743566752814454, + "grad_norm": 3.4176414012908936, + "learning_rate": 0.00019729970329155182, + "loss": 0.298, + "step": 3172 + }, + { + "epoch": 0.0743801168562504, + "grad_norm": 2.4995615482330322, + "learning_rate": 0.00019729800278855492, + "loss": 1.9618, + "step": 3173 + }, + { + "epoch": 0.0744035584310554, + "grad_norm": 5.2911295890808105, + "learning_rate": 0.0001972963017576163, + "loss": 1.0005, + "step": 3174 + }, + { + "epoch": 0.07442700000586039, + "grad_norm": 4.936560153961182, + "learning_rate": 0.00019729460019874524, + "loss": 0.6634, + "step": 3175 + }, + { + "epoch": 0.07445044158066538, + "grad_norm": 0.7095153331756592, + "learning_rate": 0.00019729289811195092, + "loss": 0.0483, + "step": 3176 + }, + { + "epoch": 0.07447388315547039, + "grad_norm": 2.3539748191833496, + "learning_rate": 0.0001972911954972426, + "loss": 0.4037, + "step": 3177 + }, + { + "epoch": 0.07449732473027539, + "grad_norm": 2.614896774291992, + "learning_rate": 0.00019728949235462952, + "loss": 0.3766, + "step": 3178 + }, + { + "epoch": 0.07452076630508038, + "grad_norm": 3.8647446632385254, + "learning_rate": 0.00019728778868412095, + "loss": 0.5565, + "step": 3179 + }, + { + "epoch": 0.07454420787988537, + "grad_norm": 2.29082989692688, + "learning_rate": 0.00019728608448572605, + "loss": 0.1904, + "step": 3180 + }, + { + "epoch": 0.07456764945469037, + "grad_norm": 3.5330138206481934, + "learning_rate": 0.00019728437975945418, + "loss": 0.6756, + "step": 3181 + }, + { + "epoch": 0.07459109102949536, + "grad_norm": 4.941400051116943, + "learning_rate": 0.0001972826745053145, + "loss": 0.9775, + "step": 3182 + }, + { + "epoch": 0.07461453260430036, + "grad_norm": 3.7536303997039795, + "learning_rate": 0.00019728096872331628, + "loss": 0.2083, + "step": 3183 + }, + { + "epoch": 0.07463797417910535, + "grad_norm": 2.5264840126037598, + "learning_rate": 0.0001972792624134688, + "loss": 0.3675, + "step": 3184 + }, + { + "epoch": 0.07466141575391035, + "grad_norm": 1.5416059494018555, + "learning_rate": 0.0001972775555757813, + "loss": 0.2638, + "step": 3185 + }, + { + "epoch": 0.07468485732871534, + "grad_norm": 3.535674571990967, + "learning_rate": 0.00019727584821026305, + "loss": 0.6964, + "step": 3186 + }, + { + "epoch": 0.07470829890352033, + "grad_norm": 2.637239933013916, + "learning_rate": 0.00019727414031692333, + "loss": 0.6167, + "step": 3187 + }, + { + "epoch": 0.07473174047832533, + "grad_norm": 3.3956871032714844, + "learning_rate": 0.00019727243189577136, + "loss": 0.5044, + "step": 3188 + }, + { + "epoch": 0.07475518205313032, + "grad_norm": 3.9339637756347656, + "learning_rate": 0.00019727072294681647, + "loss": 0.7432, + "step": 3189 + }, + { + "epoch": 0.07477862362793533, + "grad_norm": 5.415685653686523, + "learning_rate": 0.0001972690134700679, + "loss": 0.7797, + "step": 3190 + }, + { + "epoch": 0.07480206520274033, + "grad_norm": 1.2243216037750244, + "learning_rate": 0.0001972673034655349, + "loss": 0.211, + "step": 3191 + }, + { + "epoch": 0.07482550677754532, + "grad_norm": 1.387771725654602, + "learning_rate": 0.00019726559293322678, + "loss": 0.1493, + "step": 3192 + }, + { + "epoch": 0.07484894835235031, + "grad_norm": 2.414520502090454, + "learning_rate": 0.00019726388187315283, + "loss": 0.5604, + "step": 3193 + }, + { + "epoch": 0.07487238992715531, + "grad_norm": 4.356638431549072, + "learning_rate": 0.00019726217028532235, + "loss": 0.6295, + "step": 3194 + }, + { + "epoch": 0.0748958315019603, + "grad_norm": 5.3513503074646, + "learning_rate": 0.00019726045816974454, + "loss": 0.8174, + "step": 3195 + }, + { + "epoch": 0.0749192730767653, + "grad_norm": 2.9714536666870117, + "learning_rate": 0.0001972587455264288, + "loss": 1.6799, + "step": 3196 + }, + { + "epoch": 0.07494271465157029, + "grad_norm": 1.1245075464248657, + "learning_rate": 0.00019725703235538435, + "loss": 0.1893, + "step": 3197 + }, + { + "epoch": 0.07496615622637529, + "grad_norm": 4.748763084411621, + "learning_rate": 0.00019725531865662053, + "loss": 0.9751, + "step": 3198 + }, + { + "epoch": 0.07498959780118028, + "grad_norm": 4.748398303985596, + "learning_rate": 0.00019725360443014658, + "loss": 0.6935, + "step": 3199 + }, + { + "epoch": 0.07501303937598527, + "grad_norm": 3.5679008960723877, + "learning_rate": 0.00019725188967597185, + "loss": 0.4551, + "step": 3200 + }, + { + "epoch": 0.07503648095079027, + "grad_norm": 1.4006848335266113, + "learning_rate": 0.00019725017439410565, + "loss": 0.2108, + "step": 3201 + }, + { + "epoch": 0.07505992252559526, + "grad_norm": 2.389511823654175, + "learning_rate": 0.00019724845858455726, + "loss": 0.3929, + "step": 3202 + }, + { + "epoch": 0.07508336410040027, + "grad_norm": 2.795933961868286, + "learning_rate": 0.00019724674224733598, + "loss": 0.4107, + "step": 3203 + }, + { + "epoch": 0.07510680567520527, + "grad_norm": 0.731053352355957, + "learning_rate": 0.00019724502538245116, + "loss": 0.1828, + "step": 3204 + }, + { + "epoch": 0.07513024725001026, + "grad_norm": 3.1318771839141846, + "learning_rate": 0.00019724330798991208, + "loss": 0.5586, + "step": 3205 + }, + { + "epoch": 0.07515368882481525, + "grad_norm": 4.035830020904541, + "learning_rate": 0.00019724159006972812, + "loss": 0.5047, + "step": 3206 + }, + { + "epoch": 0.07517713039962025, + "grad_norm": 2.2788243293762207, + "learning_rate": 0.0001972398716219085, + "loss": 1.3549, + "step": 3207 + }, + { + "epoch": 0.07520057197442524, + "grad_norm": 1.7746444940567017, + "learning_rate": 0.00019723815264646266, + "loss": 1.4811, + "step": 3208 + }, + { + "epoch": 0.07522401354923024, + "grad_norm": 1.4549061059951782, + "learning_rate": 0.00019723643314339983, + "loss": 0.2604, + "step": 3209 + }, + { + "epoch": 0.07524745512403523, + "grad_norm": 3.638723850250244, + "learning_rate": 0.00019723471311272942, + "loss": 0.4362, + "step": 3210 + }, + { + "epoch": 0.07527089669884023, + "grad_norm": 4.122156620025635, + "learning_rate": 0.00019723299255446067, + "loss": 0.4801, + "step": 3211 + }, + { + "epoch": 0.07529433827364522, + "grad_norm": 3.714695692062378, + "learning_rate": 0.00019723127146860302, + "loss": 0.3652, + "step": 3212 + }, + { + "epoch": 0.07531777984845021, + "grad_norm": 0.44207388162612915, + "learning_rate": 0.00019722954985516573, + "loss": 0.059, + "step": 3213 + }, + { + "epoch": 0.07534122142325521, + "grad_norm": 1.2187762260437012, + "learning_rate": 0.00019722782771415816, + "loss": 0.725, + "step": 3214 + }, + { + "epoch": 0.0753646629980602, + "grad_norm": 1.1809535026550293, + "learning_rate": 0.00019722610504558968, + "loss": 0.1985, + "step": 3215 + }, + { + "epoch": 0.07538810457286521, + "grad_norm": 3.676140069961548, + "learning_rate": 0.00019722438184946957, + "loss": 0.9123, + "step": 3216 + }, + { + "epoch": 0.0754115461476702, + "grad_norm": 3.590996742248535, + "learning_rate": 0.00019722265812580729, + "loss": 0.465, + "step": 3217 + }, + { + "epoch": 0.0754349877224752, + "grad_norm": 2.737100839614868, + "learning_rate": 0.00019722093387461215, + "loss": 1.5645, + "step": 3218 + }, + { + "epoch": 0.0754584292972802, + "grad_norm": 3.2042293548583984, + "learning_rate": 0.00019721920909589346, + "loss": 0.7461, + "step": 3219 + }, + { + "epoch": 0.07548187087208519, + "grad_norm": 1.0656698942184448, + "learning_rate": 0.0001972174837896606, + "loss": 0.1695, + "step": 3220 + }, + { + "epoch": 0.07550531244689018, + "grad_norm": 1.6303967237472534, + "learning_rate": 0.00019721575795592292, + "loss": 0.6497, + "step": 3221 + }, + { + "epoch": 0.07552875402169518, + "grad_norm": 1.0650445222854614, + "learning_rate": 0.00019721403159468984, + "loss": 0.1451, + "step": 3222 + }, + { + "epoch": 0.07555219559650017, + "grad_norm": 2.9793434143066406, + "learning_rate": 0.00019721230470597068, + "loss": 0.4797, + "step": 3223 + }, + { + "epoch": 0.07557563717130517, + "grad_norm": 2.3852591514587402, + "learning_rate": 0.0001972105772897748, + "loss": 0.3228, + "step": 3224 + }, + { + "epoch": 0.07559907874611016, + "grad_norm": 2.789057970046997, + "learning_rate": 0.00019720884934611163, + "loss": 0.4078, + "step": 3225 + }, + { + "epoch": 0.07562252032091515, + "grad_norm": 1.214449405670166, + "learning_rate": 0.00019720712087499046, + "loss": 0.2494, + "step": 3226 + }, + { + "epoch": 0.07564596189572015, + "grad_norm": 1.6643047332763672, + "learning_rate": 0.00019720539187642077, + "loss": 0.8096, + "step": 3227 + }, + { + "epoch": 0.07566940347052516, + "grad_norm": 2.9488747119903564, + "learning_rate": 0.00019720366235041187, + "loss": 0.5168, + "step": 3228 + }, + { + "epoch": 0.07569284504533015, + "grad_norm": 3.2532260417938232, + "learning_rate": 0.0001972019322969732, + "loss": 0.7875, + "step": 3229 + }, + { + "epoch": 0.07571628662013515, + "grad_norm": 3.5245656967163086, + "learning_rate": 0.00019720020171611408, + "loss": 0.6609, + "step": 3230 + }, + { + "epoch": 0.07573972819494014, + "grad_norm": 0.6432678699493408, + "learning_rate": 0.00019719847060784392, + "loss": 0.0528, + "step": 3231 + }, + { + "epoch": 0.07576316976974513, + "grad_norm": 4.253719329833984, + "learning_rate": 0.00019719673897217215, + "loss": 0.4445, + "step": 3232 + }, + { + "epoch": 0.07578661134455013, + "grad_norm": 2.6028926372528076, + "learning_rate": 0.0001971950068091081, + "loss": 1.9585, + "step": 3233 + }, + { + "epoch": 0.07581005291935512, + "grad_norm": 0.7995219826698303, + "learning_rate": 0.00019719327411866125, + "loss": 0.1139, + "step": 3234 + }, + { + "epoch": 0.07583349449416012, + "grad_norm": 4.736549377441406, + "learning_rate": 0.00019719154090084095, + "loss": 0.6885, + "step": 3235 + }, + { + "epoch": 0.07585693606896511, + "grad_norm": 2.1204798221588135, + "learning_rate": 0.00019718980715565662, + "loss": 1.1733, + "step": 3236 + }, + { + "epoch": 0.0758803776437701, + "grad_norm": 6.730458736419678, + "learning_rate": 0.00019718807288311767, + "loss": 1.5374, + "step": 3237 + }, + { + "epoch": 0.0759038192185751, + "grad_norm": 3.031628370285034, + "learning_rate": 0.0001971863380832335, + "loss": 1.3394, + "step": 3238 + }, + { + "epoch": 0.0759272607933801, + "grad_norm": 2.6499483585357666, + "learning_rate": 0.00019718460275601352, + "loss": 1.5504, + "step": 3239 + }, + { + "epoch": 0.07595070236818509, + "grad_norm": 2.4614622592926025, + "learning_rate": 0.00019718286690146715, + "loss": 0.3664, + "step": 3240 + }, + { + "epoch": 0.0759741439429901, + "grad_norm": 3.4609110355377197, + "learning_rate": 0.0001971811305196038, + "loss": 0.5931, + "step": 3241 + }, + { + "epoch": 0.07599758551779509, + "grad_norm": 1.1594136953353882, + "learning_rate": 0.00019717939361043295, + "loss": 0.1956, + "step": 3242 + }, + { + "epoch": 0.07602102709260009, + "grad_norm": 0.7804440259933472, + "learning_rate": 0.00019717765617396392, + "loss": 0.0946, + "step": 3243 + }, + { + "epoch": 0.07604446866740508, + "grad_norm": 4.339264869689941, + "learning_rate": 0.00019717591821020626, + "loss": 0.6186, + "step": 3244 + }, + { + "epoch": 0.07606791024221007, + "grad_norm": 1.619035243988037, + "learning_rate": 0.00019717417971916928, + "loss": 0.2087, + "step": 3245 + }, + { + "epoch": 0.07609135181701507, + "grad_norm": 2.3196535110473633, + "learning_rate": 0.00019717244070086248, + "loss": 1.6126, + "step": 3246 + }, + { + "epoch": 0.07611479339182006, + "grad_norm": 4.574919700622559, + "learning_rate": 0.0001971707011552953, + "loss": 0.7248, + "step": 3247 + }, + { + "epoch": 0.07613823496662506, + "grad_norm": 1.3735592365264893, + "learning_rate": 0.00019716896108247715, + "loss": 0.1347, + "step": 3248 + }, + { + "epoch": 0.07616167654143005, + "grad_norm": 4.329329967498779, + "learning_rate": 0.0001971672204824175, + "loss": 0.8153, + "step": 3249 + }, + { + "epoch": 0.07618511811623505, + "grad_norm": 5.26861047744751, + "learning_rate": 0.00019716547935512578, + "loss": 0.8199, + "step": 3250 + }, + { + "epoch": 0.07620855969104004, + "grad_norm": 3.5594747066497803, + "learning_rate": 0.0001971637377006114, + "loss": 0.6035, + "step": 3251 + }, + { + "epoch": 0.07623200126584503, + "grad_norm": 0.5105588436126709, + "learning_rate": 0.00019716199551888389, + "loss": 0.0646, + "step": 3252 + }, + { + "epoch": 0.07625544284065003, + "grad_norm": 2.6567275524139404, + "learning_rate": 0.00019716025280995262, + "loss": 2.0141, + "step": 3253 + }, + { + "epoch": 0.07627888441545504, + "grad_norm": 0.5451585054397583, + "learning_rate": 0.00019715850957382713, + "loss": 0.0636, + "step": 3254 + }, + { + "epoch": 0.07630232599026003, + "grad_norm": 4.89564323425293, + "learning_rate": 0.00019715676581051682, + "loss": 0.9284, + "step": 3255 + }, + { + "epoch": 0.07632576756506503, + "grad_norm": 3.5309250354766846, + "learning_rate": 0.00019715502152003116, + "loss": 0.2932, + "step": 3256 + }, + { + "epoch": 0.07634920913987002, + "grad_norm": 3.423813819885254, + "learning_rate": 0.0001971532767023796, + "loss": 0.4226, + "step": 3257 + }, + { + "epoch": 0.07637265071467501, + "grad_norm": 5.429090976715088, + "learning_rate": 0.00019715153135757167, + "loss": 0.7532, + "step": 3258 + }, + { + "epoch": 0.07639609228948001, + "grad_norm": 3.550445795059204, + "learning_rate": 0.0001971497854856168, + "loss": 0.5211, + "step": 3259 + }, + { + "epoch": 0.076419533864285, + "grad_norm": 3.2141306400299072, + "learning_rate": 0.00019714803908652443, + "loss": 1.6046, + "step": 3260 + }, + { + "epoch": 0.07644297543909, + "grad_norm": 4.508984088897705, + "learning_rate": 0.0001971462921603041, + "loss": 0.7853, + "step": 3261 + }, + { + "epoch": 0.07646641701389499, + "grad_norm": 1.4092453718185425, + "learning_rate": 0.00019714454470696522, + "loss": 0.1921, + "step": 3262 + }, + { + "epoch": 0.07648985858869999, + "grad_norm": 2.948007583618164, + "learning_rate": 0.0001971427967265173, + "loss": 0.2402, + "step": 3263 + }, + { + "epoch": 0.07651330016350498, + "grad_norm": 1.3856247663497925, + "learning_rate": 0.0001971410482189699, + "loss": 0.2583, + "step": 3264 + }, + { + "epoch": 0.07653674173830997, + "grad_norm": 0.9117443561553955, + "learning_rate": 0.0001971392991843324, + "loss": 0.1059, + "step": 3265 + }, + { + "epoch": 0.07656018331311497, + "grad_norm": 2.693506956100464, + "learning_rate": 0.00019713754962261435, + "loss": 0.445, + "step": 3266 + }, + { + "epoch": 0.07658362488791998, + "grad_norm": 1.3983441591262817, + "learning_rate": 0.0001971357995338252, + "loss": 0.2287, + "step": 3267 + }, + { + "epoch": 0.07660706646272497, + "grad_norm": 2.4008290767669678, + "learning_rate": 0.0001971340489179745, + "loss": 0.3635, + "step": 3268 + }, + { + "epoch": 0.07663050803752997, + "grad_norm": 4.233178615570068, + "learning_rate": 0.00019713229777507167, + "loss": 0.3092, + "step": 3269 + }, + { + "epoch": 0.07665394961233496, + "grad_norm": 0.3522509038448334, + "learning_rate": 0.00019713054610512632, + "loss": 0.0771, + "step": 3270 + }, + { + "epoch": 0.07667739118713995, + "grad_norm": 4.08924674987793, + "learning_rate": 0.00019712879390814785, + "loss": 0.8418, + "step": 3271 + }, + { + "epoch": 0.07670083276194495, + "grad_norm": 1.3578402996063232, + "learning_rate": 0.00019712704118414581, + "loss": 0.3069, + "step": 3272 + }, + { + "epoch": 0.07672427433674994, + "grad_norm": 3.0711007118225098, + "learning_rate": 0.00019712528793312974, + "loss": 0.5878, + "step": 3273 + }, + { + "epoch": 0.07674771591155494, + "grad_norm": 2.260490655899048, + "learning_rate": 0.00019712353415510912, + "loss": 0.272, + "step": 3274 + }, + { + "epoch": 0.07677115748635993, + "grad_norm": 1.1639034748077393, + "learning_rate": 0.0001971217798500935, + "loss": 0.1582, + "step": 3275 + }, + { + "epoch": 0.07679459906116493, + "grad_norm": 4.413294792175293, + "learning_rate": 0.00019712002501809232, + "loss": 0.5043, + "step": 3276 + }, + { + "epoch": 0.07681804063596992, + "grad_norm": 3.8162384033203125, + "learning_rate": 0.00019711826965911517, + "loss": 0.5705, + "step": 3277 + }, + { + "epoch": 0.07684148221077491, + "grad_norm": 3.722275972366333, + "learning_rate": 0.0001971165137731716, + "loss": 0.2086, + "step": 3278 + }, + { + "epoch": 0.07686492378557991, + "grad_norm": 4.664172172546387, + "learning_rate": 0.00019711475736027104, + "loss": 0.731, + "step": 3279 + }, + { + "epoch": 0.07688836536038492, + "grad_norm": 0.7648279070854187, + "learning_rate": 0.0001971130004204231, + "loss": 0.0804, + "step": 3280 + }, + { + "epoch": 0.07691180693518991, + "grad_norm": 2.3681111335754395, + "learning_rate": 0.0001971112429536373, + "loss": 0.223, + "step": 3281 + }, + { + "epoch": 0.0769352485099949, + "grad_norm": 0.6376553773880005, + "learning_rate": 0.00019710948495992313, + "loss": 0.0874, + "step": 3282 + }, + { + "epoch": 0.0769586900847999, + "grad_norm": 1.5566692352294922, + "learning_rate": 0.0001971077264392902, + "loss": 0.206, + "step": 3283 + }, + { + "epoch": 0.0769821316596049, + "grad_norm": 5.274338722229004, + "learning_rate": 0.000197105967391748, + "loss": 0.593, + "step": 3284 + }, + { + "epoch": 0.07700557323440989, + "grad_norm": 4.2469801902771, + "learning_rate": 0.00019710420781730606, + "loss": 0.4025, + "step": 3285 + }, + { + "epoch": 0.07702901480921488, + "grad_norm": 3.5381221771240234, + "learning_rate": 0.00019710244771597402, + "loss": 0.1375, + "step": 3286 + }, + { + "epoch": 0.07705245638401988, + "grad_norm": 5.4306182861328125, + "learning_rate": 0.00019710068708776133, + "loss": 0.8568, + "step": 3287 + }, + { + "epoch": 0.07707589795882487, + "grad_norm": 4.59316349029541, + "learning_rate": 0.0001970989259326776, + "loss": 0.3723, + "step": 3288 + }, + { + "epoch": 0.07709933953362987, + "grad_norm": 3.487334966659546, + "learning_rate": 0.00019709716425073235, + "loss": 0.7376, + "step": 3289 + }, + { + "epoch": 0.07712278110843486, + "grad_norm": 3.9871931076049805, + "learning_rate": 0.00019709540204193517, + "loss": 0.8423, + "step": 3290 + }, + { + "epoch": 0.07714622268323985, + "grad_norm": 2.4334232807159424, + "learning_rate": 0.0001970936393062956, + "loss": 0.3425, + "step": 3291 + }, + { + "epoch": 0.07716966425804486, + "grad_norm": 4.446340560913086, + "learning_rate": 0.00019709187604382322, + "loss": 0.7177, + "step": 3292 + }, + { + "epoch": 0.07719310583284986, + "grad_norm": 3.0128729343414307, + "learning_rate": 0.00019709011225452758, + "loss": 0.5591, + "step": 3293 + }, + { + "epoch": 0.07721654740765485, + "grad_norm": 0.9966609477996826, + "learning_rate": 0.00019708834793841828, + "loss": 0.139, + "step": 3294 + }, + { + "epoch": 0.07723998898245985, + "grad_norm": 4.064762592315674, + "learning_rate": 0.00019708658309550487, + "loss": 0.5926, + "step": 3295 + }, + { + "epoch": 0.07726343055726484, + "grad_norm": 1.0971736907958984, + "learning_rate": 0.0001970848177257969, + "loss": 0.1528, + "step": 3296 + }, + { + "epoch": 0.07728687213206983, + "grad_norm": 4.399998664855957, + "learning_rate": 0.00019708305182930403, + "loss": 0.5533, + "step": 3297 + }, + { + "epoch": 0.07731031370687483, + "grad_norm": 6.376077651977539, + "learning_rate": 0.00019708128540603577, + "loss": 0.6098, + "step": 3298 + }, + { + "epoch": 0.07733375528167982, + "grad_norm": 3.978851556777954, + "learning_rate": 0.0001970795184560017, + "loss": 0.5503, + "step": 3299 + }, + { + "epoch": 0.07735719685648482, + "grad_norm": 3.230573892593384, + "learning_rate": 0.00019707775097921144, + "loss": 0.6421, + "step": 3300 + }, + { + "epoch": 0.07738063843128981, + "grad_norm": 3.1296913623809814, + "learning_rate": 0.0001970759829756746, + "loss": 0.5354, + "step": 3301 + }, + { + "epoch": 0.0774040800060948, + "grad_norm": 1.833777904510498, + "learning_rate": 0.00019707421444540072, + "loss": 1.3733, + "step": 3302 + }, + { + "epoch": 0.0774275215808998, + "grad_norm": 0.8781443238258362, + "learning_rate": 0.00019707244538839943, + "loss": 0.1316, + "step": 3303 + }, + { + "epoch": 0.0774509631557048, + "grad_norm": 1.1023519039154053, + "learning_rate": 0.00019707067580468035, + "loss": 0.1202, + "step": 3304 + }, + { + "epoch": 0.0774744047305098, + "grad_norm": 1.8759880065917969, + "learning_rate": 0.00019706890569425303, + "loss": 0.2823, + "step": 3305 + }, + { + "epoch": 0.0774978463053148, + "grad_norm": 3.397406578063965, + "learning_rate": 0.00019706713505712712, + "loss": 0.4624, + "step": 3306 + }, + { + "epoch": 0.07752128788011979, + "grad_norm": 0.6494703888893127, + "learning_rate": 0.00019706536389331217, + "loss": 0.123, + "step": 3307 + }, + { + "epoch": 0.07754472945492479, + "grad_norm": 3.9695346355438232, + "learning_rate": 0.00019706359220281782, + "loss": 0.6867, + "step": 3308 + }, + { + "epoch": 0.07756817102972978, + "grad_norm": 4.39473819732666, + "learning_rate": 0.0001970618199856537, + "loss": 0.4949, + "step": 3309 + }, + { + "epoch": 0.07759161260453477, + "grad_norm": 4.099949359893799, + "learning_rate": 0.00019706004724182944, + "loss": 0.6325, + "step": 3310 + }, + { + "epoch": 0.07761505417933977, + "grad_norm": 4.542725086212158, + "learning_rate": 0.00019705827397135465, + "loss": 0.7549, + "step": 3311 + }, + { + "epoch": 0.07763849575414476, + "grad_norm": 1.9570930004119873, + "learning_rate": 0.00019705650017423886, + "loss": 1.654, + "step": 3312 + }, + { + "epoch": 0.07766193732894976, + "grad_norm": 3.389601230621338, + "learning_rate": 0.00019705472585049183, + "loss": 0.2577, + "step": 3313 + }, + { + "epoch": 0.07768537890375475, + "grad_norm": 3.053868532180786, + "learning_rate": 0.00019705295100012313, + "loss": 0.3909, + "step": 3314 + }, + { + "epoch": 0.07770882047855975, + "grad_norm": 5.2726240158081055, + "learning_rate": 0.0001970511756231424, + "loss": 0.6318, + "step": 3315 + }, + { + "epoch": 0.07773226205336474, + "grad_norm": 3.7779784202575684, + "learning_rate": 0.0001970493997195592, + "loss": 0.4524, + "step": 3316 + }, + { + "epoch": 0.07775570362816973, + "grad_norm": 3.7696547508239746, + "learning_rate": 0.0001970476232893833, + "loss": 0.6124, + "step": 3317 + }, + { + "epoch": 0.07777914520297474, + "grad_norm": 3.795755624771118, + "learning_rate": 0.00019704584633262423, + "loss": 0.4379, + "step": 3318 + }, + { + "epoch": 0.07780258677777974, + "grad_norm": 1.4719563722610474, + "learning_rate": 0.00019704406884929164, + "loss": 0.2036, + "step": 3319 + }, + { + "epoch": 0.07782602835258473, + "grad_norm": 1.3493127822875977, + "learning_rate": 0.00019704229083939527, + "loss": 0.203, + "step": 3320 + }, + { + "epoch": 0.07784946992738973, + "grad_norm": 2.1864633560180664, + "learning_rate": 0.00019704051230294467, + "loss": 0.1973, + "step": 3321 + }, + { + "epoch": 0.07787291150219472, + "grad_norm": 3.8323850631713867, + "learning_rate": 0.0001970387332399495, + "loss": 0.698, + "step": 3322 + }, + { + "epoch": 0.07789635307699971, + "grad_norm": 2.752837896347046, + "learning_rate": 0.00019703695365041945, + "loss": 1.4294, + "step": 3323 + }, + { + "epoch": 0.07791979465180471, + "grad_norm": 1.4256391525268555, + "learning_rate": 0.00019703517353436415, + "loss": 0.2999, + "step": 3324 + }, + { + "epoch": 0.0779432362266097, + "grad_norm": 4.073864459991455, + "learning_rate": 0.00019703339289179331, + "loss": 0.6261, + "step": 3325 + }, + { + "epoch": 0.0779666778014147, + "grad_norm": 3.4068427085876465, + "learning_rate": 0.00019703161172271653, + "loss": 0.4202, + "step": 3326 + }, + { + "epoch": 0.07799011937621969, + "grad_norm": 1.2189611196517944, + "learning_rate": 0.0001970298300271435, + "loss": 0.2557, + "step": 3327 + }, + { + "epoch": 0.07801356095102469, + "grad_norm": 4.242702960968018, + "learning_rate": 0.0001970280478050839, + "loss": 0.7442, + "step": 3328 + }, + { + "epoch": 0.07803700252582968, + "grad_norm": 3.280014753341675, + "learning_rate": 0.00019702626505654734, + "loss": 0.417, + "step": 3329 + }, + { + "epoch": 0.07806044410063467, + "grad_norm": 0.8249542713165283, + "learning_rate": 0.00019702448178154355, + "loss": 0.1469, + "step": 3330 + }, + { + "epoch": 0.07808388567543968, + "grad_norm": 2.1723999977111816, + "learning_rate": 0.00019702269798008222, + "loss": 0.3011, + "step": 3331 + }, + { + "epoch": 0.07810732725024468, + "grad_norm": 3.6797821521759033, + "learning_rate": 0.00019702091365217297, + "loss": 0.3699, + "step": 3332 + }, + { + "epoch": 0.07813076882504967, + "grad_norm": 4.776339530944824, + "learning_rate": 0.00019701912879782553, + "loss": 1.1715, + "step": 3333 + }, + { + "epoch": 0.07815421039985467, + "grad_norm": 4.033507823944092, + "learning_rate": 0.00019701734341704959, + "loss": 0.583, + "step": 3334 + }, + { + "epoch": 0.07817765197465966, + "grad_norm": 0.764188826084137, + "learning_rate": 0.00019701555750985478, + "loss": 0.0846, + "step": 3335 + }, + { + "epoch": 0.07820109354946465, + "grad_norm": 3.107260227203369, + "learning_rate": 0.00019701377107625084, + "loss": 1.7864, + "step": 3336 + }, + { + "epoch": 0.07822453512426965, + "grad_norm": 3.7390575408935547, + "learning_rate": 0.00019701198411624745, + "loss": 0.6445, + "step": 3337 + }, + { + "epoch": 0.07824797669907464, + "grad_norm": 3.4464213848114014, + "learning_rate": 0.0001970101966298543, + "loss": 0.374, + "step": 3338 + }, + { + "epoch": 0.07827141827387964, + "grad_norm": 0.4983593225479126, + "learning_rate": 0.00019700840861708115, + "loss": 0.0799, + "step": 3339 + }, + { + "epoch": 0.07829485984868463, + "grad_norm": 4.5649919509887695, + "learning_rate": 0.0001970066200779376, + "loss": 0.9391, + "step": 3340 + }, + { + "epoch": 0.07831830142348963, + "grad_norm": 1.5223726034164429, + "learning_rate": 0.0001970048310124334, + "loss": 0.1506, + "step": 3341 + }, + { + "epoch": 0.07834174299829462, + "grad_norm": 4.797775745391846, + "learning_rate": 0.00019700304142057827, + "loss": 0.8783, + "step": 3342 + }, + { + "epoch": 0.07836518457309963, + "grad_norm": 1.7379305362701416, + "learning_rate": 0.0001970012513023819, + "loss": 0.3104, + "step": 3343 + }, + { + "epoch": 0.07838862614790462, + "grad_norm": 0.8747650980949402, + "learning_rate": 0.00019699946065785401, + "loss": 0.1347, + "step": 3344 + }, + { + "epoch": 0.07841206772270962, + "grad_norm": 4.3754048347473145, + "learning_rate": 0.00019699766948700433, + "loss": 0.4797, + "step": 3345 + }, + { + "epoch": 0.07843550929751461, + "grad_norm": 3.397822856903076, + "learning_rate": 0.00019699587778984257, + "loss": 0.8159, + "step": 3346 + }, + { + "epoch": 0.0784589508723196, + "grad_norm": 5.64393424987793, + "learning_rate": 0.00019699408556637845, + "loss": 0.6155, + "step": 3347 + }, + { + "epoch": 0.0784823924471246, + "grad_norm": 2.3781955242156982, + "learning_rate": 0.00019699229281662167, + "loss": 1.7043, + "step": 3348 + }, + { + "epoch": 0.0785058340219296, + "grad_norm": 2.835402250289917, + "learning_rate": 0.00019699049954058202, + "loss": 0.4626, + "step": 3349 + }, + { + "epoch": 0.07852927559673459, + "grad_norm": 5.493829250335693, + "learning_rate": 0.00019698870573826914, + "loss": 0.9131, + "step": 3350 + }, + { + "epoch": 0.07855271717153958, + "grad_norm": 3.3701577186584473, + "learning_rate": 0.00019698691140969288, + "loss": 0.391, + "step": 3351 + }, + { + "epoch": 0.07857615874634458, + "grad_norm": 2.2256031036376953, + "learning_rate": 0.00019698511655486287, + "loss": 1.8641, + "step": 3352 + }, + { + "epoch": 0.07859960032114957, + "grad_norm": 2.8661389350891113, + "learning_rate": 0.0001969833211737889, + "loss": 0.3852, + "step": 3353 + }, + { + "epoch": 0.07862304189595457, + "grad_norm": 3.900240421295166, + "learning_rate": 0.0001969815252664807, + "loss": 0.7618, + "step": 3354 + }, + { + "epoch": 0.07864648347075956, + "grad_norm": 1.4801268577575684, + "learning_rate": 0.00019697972883294803, + "loss": 0.1356, + "step": 3355 + }, + { + "epoch": 0.07866992504556457, + "grad_norm": 3.87096905708313, + "learning_rate": 0.00019697793187320063, + "loss": 0.4113, + "step": 3356 + }, + { + "epoch": 0.07869336662036956, + "grad_norm": 5.49067497253418, + "learning_rate": 0.00019697613438724823, + "loss": 0.6324, + "step": 3357 + }, + { + "epoch": 0.07871680819517456, + "grad_norm": 1.9240081310272217, + "learning_rate": 0.00019697433637510059, + "loss": 0.3364, + "step": 3358 + }, + { + "epoch": 0.07874024976997955, + "grad_norm": 4.622624397277832, + "learning_rate": 0.00019697253783676748, + "loss": 0.55, + "step": 3359 + }, + { + "epoch": 0.07876369134478454, + "grad_norm": 1.4536017179489136, + "learning_rate": 0.00019697073877225868, + "loss": 1.0478, + "step": 3360 + }, + { + "epoch": 0.07878713291958954, + "grad_norm": 4.77266788482666, + "learning_rate": 0.00019696893918158388, + "loss": 0.5234, + "step": 3361 + }, + { + "epoch": 0.07881057449439453, + "grad_norm": 5.710120677947998, + "learning_rate": 0.0001969671390647529, + "loss": 1.1228, + "step": 3362 + }, + { + "epoch": 0.07883401606919953, + "grad_norm": 3.1422526836395264, + "learning_rate": 0.0001969653384217755, + "loss": 0.5721, + "step": 3363 + }, + { + "epoch": 0.07885745764400452, + "grad_norm": 3.9139411449432373, + "learning_rate": 0.00019696353725266147, + "loss": 0.4822, + "step": 3364 + }, + { + "epoch": 0.07888089921880952, + "grad_norm": 3.593029499053955, + "learning_rate": 0.00019696173555742053, + "loss": 0.6775, + "step": 3365 + }, + { + "epoch": 0.07890434079361451, + "grad_norm": 0.9246063232421875, + "learning_rate": 0.0001969599333360625, + "loss": 0.6978, + "step": 3366 + }, + { + "epoch": 0.0789277823684195, + "grad_norm": 1.4742820262908936, + "learning_rate": 0.00019695813058859715, + "loss": 0.1546, + "step": 3367 + }, + { + "epoch": 0.0789512239432245, + "grad_norm": 1.1655353307724, + "learning_rate": 0.00019695632731503426, + "loss": 0.152, + "step": 3368 + }, + { + "epoch": 0.07897466551802951, + "grad_norm": 2.7023372650146484, + "learning_rate": 0.0001969545235153836, + "loss": 0.3473, + "step": 3369 + }, + { + "epoch": 0.0789981070928345, + "grad_norm": 3.1649186611175537, + "learning_rate": 0.00019695271918965495, + "loss": 0.3847, + "step": 3370 + }, + { + "epoch": 0.0790215486676395, + "grad_norm": 0.8743128776550293, + "learning_rate": 0.00019695091433785818, + "loss": 0.1147, + "step": 3371 + }, + { + "epoch": 0.07904499024244449, + "grad_norm": 1.1807293891906738, + "learning_rate": 0.000196949108960003, + "loss": 0.1577, + "step": 3372 + }, + { + "epoch": 0.07906843181724948, + "grad_norm": 3.5463685989379883, + "learning_rate": 0.00019694730305609917, + "loss": 0.539, + "step": 3373 + }, + { + "epoch": 0.07909187339205448, + "grad_norm": 2.655083417892456, + "learning_rate": 0.0001969454966261566, + "loss": 1.5926, + "step": 3374 + }, + { + "epoch": 0.07911531496685947, + "grad_norm": 3.1306419372558594, + "learning_rate": 0.00019694368967018507, + "loss": 0.3812, + "step": 3375 + }, + { + "epoch": 0.07913875654166447, + "grad_norm": 4.31089973449707, + "learning_rate": 0.0001969418821881943, + "loss": 0.4593, + "step": 3376 + }, + { + "epoch": 0.07916219811646946, + "grad_norm": 3.844194173812866, + "learning_rate": 0.00019694007418019415, + "loss": 0.5325, + "step": 3377 + }, + { + "epoch": 0.07918563969127446, + "grad_norm": 0.9273166656494141, + "learning_rate": 0.00019693826564619449, + "loss": 0.113, + "step": 3378 + }, + { + "epoch": 0.07920908126607945, + "grad_norm": 6.921784400939941, + "learning_rate": 0.00019693645658620503, + "loss": 0.9409, + "step": 3379 + }, + { + "epoch": 0.07923252284088445, + "grad_norm": 4.678712368011475, + "learning_rate": 0.00019693464700023562, + "loss": 0.7044, + "step": 3380 + }, + { + "epoch": 0.07925596441568944, + "grad_norm": 3.2085301876068115, + "learning_rate": 0.0001969328368882961, + "loss": 0.54, + "step": 3381 + }, + { + "epoch": 0.07927940599049445, + "grad_norm": 1.9302847385406494, + "learning_rate": 0.0001969310262503963, + "loss": 1.6626, + "step": 3382 + }, + { + "epoch": 0.07930284756529944, + "grad_norm": 3.1491189002990723, + "learning_rate": 0.000196929215086546, + "loss": 0.6941, + "step": 3383 + }, + { + "epoch": 0.07932628914010444, + "grad_norm": 1.5309401750564575, + "learning_rate": 0.00019692740339675508, + "loss": 0.3126, + "step": 3384 + }, + { + "epoch": 0.07934973071490943, + "grad_norm": 2.897447347640991, + "learning_rate": 0.00019692559118103333, + "loss": 0.7303, + "step": 3385 + }, + { + "epoch": 0.07937317228971442, + "grad_norm": 4.344655990600586, + "learning_rate": 0.0001969237784393906, + "loss": 0.4974, + "step": 3386 + }, + { + "epoch": 0.07939661386451942, + "grad_norm": 1.1945652961730957, + "learning_rate": 0.0001969219651718367, + "loss": 0.2274, + "step": 3387 + }, + { + "epoch": 0.07942005543932441, + "grad_norm": 2.374058961868286, + "learning_rate": 0.00019692015137838154, + "loss": 1.1909, + "step": 3388 + }, + { + "epoch": 0.07944349701412941, + "grad_norm": 2.471395254135132, + "learning_rate": 0.00019691833705903487, + "loss": 1.9235, + "step": 3389 + }, + { + "epoch": 0.0794669385889344, + "grad_norm": 3.6952803134918213, + "learning_rate": 0.00019691652221380662, + "loss": 0.3531, + "step": 3390 + }, + { + "epoch": 0.0794903801637394, + "grad_norm": 3.0730128288269043, + "learning_rate": 0.00019691470684270654, + "loss": 0.5909, + "step": 3391 + }, + { + "epoch": 0.07951382173854439, + "grad_norm": 0.5386780500411987, + "learning_rate": 0.00019691289094574458, + "loss": 0.0971, + "step": 3392 + }, + { + "epoch": 0.07953726331334939, + "grad_norm": 3.577707290649414, + "learning_rate": 0.00019691107452293055, + "loss": 0.3252, + "step": 3393 + }, + { + "epoch": 0.07956070488815438, + "grad_norm": 3.4441442489624023, + "learning_rate": 0.00019690925757427427, + "loss": 0.6312, + "step": 3394 + }, + { + "epoch": 0.07958414646295939, + "grad_norm": 3.3743393421173096, + "learning_rate": 0.00019690744009978565, + "loss": 0.593, + "step": 3395 + }, + { + "epoch": 0.07960758803776438, + "grad_norm": 1.0702109336853027, + "learning_rate": 0.00019690562209947454, + "loss": 0.0959, + "step": 3396 + }, + { + "epoch": 0.07963102961256938, + "grad_norm": 2.08569598197937, + "learning_rate": 0.0001969038035733508, + "loss": 0.5471, + "step": 3397 + }, + { + "epoch": 0.07965447118737437, + "grad_norm": 1.5671333074569702, + "learning_rate": 0.0001969019845214243, + "loss": 1.1508, + "step": 3398 + }, + { + "epoch": 0.07967791276217936, + "grad_norm": 2.7173521518707275, + "learning_rate": 0.00019690016494370492, + "loss": 0.4763, + "step": 3399 + }, + { + "epoch": 0.07970135433698436, + "grad_norm": 3.166781187057495, + "learning_rate": 0.0001968983448402025, + "loss": 0.4093, + "step": 3400 + }, + { + "epoch": 0.07972479591178935, + "grad_norm": 4.064182758331299, + "learning_rate": 0.00019689652421092692, + "loss": 0.523, + "step": 3401 + }, + { + "epoch": 0.07974823748659435, + "grad_norm": 4.056604385375977, + "learning_rate": 0.00019689470305588809, + "loss": 0.565, + "step": 3402 + }, + { + "epoch": 0.07977167906139934, + "grad_norm": 1.4154276847839355, + "learning_rate": 0.00019689288137509587, + "loss": 0.1715, + "step": 3403 + }, + { + "epoch": 0.07979512063620434, + "grad_norm": 0.6430423855781555, + "learning_rate": 0.00019689105916856017, + "loss": 0.0852, + "step": 3404 + }, + { + "epoch": 0.07981856221100933, + "grad_norm": 4.903023719787598, + "learning_rate": 0.00019688923643629088, + "loss": 0.8318, + "step": 3405 + }, + { + "epoch": 0.07984200378581433, + "grad_norm": 3.7260184288024902, + "learning_rate": 0.0001968874131782978, + "loss": 0.4601, + "step": 3406 + }, + { + "epoch": 0.07986544536061933, + "grad_norm": 7.2044148445129395, + "learning_rate": 0.00019688558939459094, + "loss": 0.8515, + "step": 3407 + }, + { + "epoch": 0.07988888693542433, + "grad_norm": 0.8865537047386169, + "learning_rate": 0.00019688376508518015, + "loss": 0.0509, + "step": 3408 + }, + { + "epoch": 0.07991232851022932, + "grad_norm": 2.438035249710083, + "learning_rate": 0.00019688194025007532, + "loss": 1.6893, + "step": 3409 + }, + { + "epoch": 0.07993577008503432, + "grad_norm": 4.400259017944336, + "learning_rate": 0.00019688011488928634, + "loss": 0.6468, + "step": 3410 + }, + { + "epoch": 0.07995921165983931, + "grad_norm": 1.4650201797485352, + "learning_rate": 0.00019687828900282316, + "loss": 0.2326, + "step": 3411 + }, + { + "epoch": 0.0799826532346443, + "grad_norm": 3.13712215423584, + "learning_rate": 0.00019687646259069564, + "loss": 1.7503, + "step": 3412 + }, + { + "epoch": 0.0800060948094493, + "grad_norm": 4.35381555557251, + "learning_rate": 0.00019687463565291373, + "loss": 0.1811, + "step": 3413 + }, + { + "epoch": 0.0800295363842543, + "grad_norm": 3.1023056507110596, + "learning_rate": 0.0001968728081894873, + "loss": 0.4001, + "step": 3414 + }, + { + "epoch": 0.08005297795905929, + "grad_norm": 4.253684043884277, + "learning_rate": 0.00019687098020042628, + "loss": 0.6663, + "step": 3415 + }, + { + "epoch": 0.08007641953386428, + "grad_norm": 4.378341197967529, + "learning_rate": 0.00019686915168574062, + "loss": 0.6096, + "step": 3416 + }, + { + "epoch": 0.08009986110866928, + "grad_norm": 4.007127285003662, + "learning_rate": 0.00019686732264544024, + "loss": 0.5044, + "step": 3417 + }, + { + "epoch": 0.08012330268347427, + "grad_norm": 0.8288443088531494, + "learning_rate": 0.000196865493079535, + "loss": 0.1097, + "step": 3418 + }, + { + "epoch": 0.08014674425827926, + "grad_norm": 4.353278636932373, + "learning_rate": 0.0001968636629880349, + "loss": 0.5077, + "step": 3419 + }, + { + "epoch": 0.08017018583308427, + "grad_norm": 4.732459545135498, + "learning_rate": 0.0001968618323709498, + "loss": 0.506, + "step": 3420 + }, + { + "epoch": 0.08019362740788927, + "grad_norm": 2.194497585296631, + "learning_rate": 0.00019686000122828975, + "loss": 0.2532, + "step": 3421 + }, + { + "epoch": 0.08021706898269426, + "grad_norm": 3.7867612838745117, + "learning_rate": 0.00019685816956006456, + "loss": 0.5979, + "step": 3422 + }, + { + "epoch": 0.08024051055749926, + "grad_norm": 2.0326688289642334, + "learning_rate": 0.00019685633736628421, + "loss": 0.1688, + "step": 3423 + }, + { + "epoch": 0.08026395213230425, + "grad_norm": 2.394986867904663, + "learning_rate": 0.00019685450464695866, + "loss": 1.6098, + "step": 3424 + }, + { + "epoch": 0.08028739370710924, + "grad_norm": 3.1658778190612793, + "learning_rate": 0.00019685267140209786, + "loss": 0.2761, + "step": 3425 + }, + { + "epoch": 0.08031083528191424, + "grad_norm": 0.9451344013214111, + "learning_rate": 0.0001968508376317117, + "loss": 0.181, + "step": 3426 + }, + { + "epoch": 0.08033427685671923, + "grad_norm": 1.046204924583435, + "learning_rate": 0.00019684900333581022, + "loss": 0.1815, + "step": 3427 + }, + { + "epoch": 0.08035771843152423, + "grad_norm": 3.0673279762268066, + "learning_rate": 0.00019684716851440333, + "loss": 0.5931, + "step": 3428 + }, + { + "epoch": 0.08038116000632922, + "grad_norm": 5.813241004943848, + "learning_rate": 0.00019684533316750093, + "loss": 0.8476, + "step": 3429 + }, + { + "epoch": 0.08040460158113422, + "grad_norm": 3.6848039627075195, + "learning_rate": 0.00019684349729511306, + "loss": 0.753, + "step": 3430 + }, + { + "epoch": 0.08042804315593921, + "grad_norm": 3.407763719558716, + "learning_rate": 0.00019684166089724964, + "loss": 0.4549, + "step": 3431 + }, + { + "epoch": 0.0804514847307442, + "grad_norm": 1.5829241275787354, + "learning_rate": 0.00019683982397392065, + "loss": 0.938, + "step": 3432 + }, + { + "epoch": 0.08047492630554921, + "grad_norm": 4.062140941619873, + "learning_rate": 0.00019683798652513605, + "loss": 0.521, + "step": 3433 + }, + { + "epoch": 0.08049836788035421, + "grad_norm": 4.3278045654296875, + "learning_rate": 0.00019683614855090583, + "loss": 0.9737, + "step": 3434 + }, + { + "epoch": 0.0805218094551592, + "grad_norm": 4.392317295074463, + "learning_rate": 0.00019683431005123992, + "loss": 0.6513, + "step": 3435 + }, + { + "epoch": 0.0805452510299642, + "grad_norm": 4.149937152862549, + "learning_rate": 0.00019683247102614834, + "loss": 0.6108, + "step": 3436 + }, + { + "epoch": 0.08056869260476919, + "grad_norm": 4.732481956481934, + "learning_rate": 0.00019683063147564104, + "loss": 0.8945, + "step": 3437 + }, + { + "epoch": 0.08059213417957418, + "grad_norm": 4.630643844604492, + "learning_rate": 0.00019682879139972802, + "loss": 0.5825, + "step": 3438 + }, + { + "epoch": 0.08061557575437918, + "grad_norm": 5.28948974609375, + "learning_rate": 0.00019682695079841923, + "loss": 0.8703, + "step": 3439 + }, + { + "epoch": 0.08063901732918417, + "grad_norm": 1.113518238067627, + "learning_rate": 0.00019682510967172472, + "loss": 0.1312, + "step": 3440 + }, + { + "epoch": 0.08066245890398917, + "grad_norm": 4.206676006317139, + "learning_rate": 0.0001968232680196544, + "loss": 0.5171, + "step": 3441 + }, + { + "epoch": 0.08068590047879416, + "grad_norm": 5.169172286987305, + "learning_rate": 0.00019682142584221833, + "loss": 0.7627, + "step": 3442 + }, + { + "epoch": 0.08070934205359916, + "grad_norm": 3.1641182899475098, + "learning_rate": 0.0001968195831394265, + "loss": 0.56, + "step": 3443 + }, + { + "epoch": 0.08073278362840415, + "grad_norm": 3.855290174484253, + "learning_rate": 0.00019681773991128888, + "loss": 0.4701, + "step": 3444 + }, + { + "epoch": 0.08075622520320914, + "grad_norm": 3.374305009841919, + "learning_rate": 0.00019681589615781544, + "loss": 0.6009, + "step": 3445 + }, + { + "epoch": 0.08077966677801415, + "grad_norm": 2.1832504272460938, + "learning_rate": 0.00019681405187901626, + "loss": 1.4531, + "step": 3446 + }, + { + "epoch": 0.08080310835281915, + "grad_norm": 4.421206474304199, + "learning_rate": 0.00019681220707490132, + "loss": 0.7946, + "step": 3447 + }, + { + "epoch": 0.08082654992762414, + "grad_norm": 1.1007286310195923, + "learning_rate": 0.0001968103617454806, + "loss": 0.1453, + "step": 3448 + }, + { + "epoch": 0.08084999150242914, + "grad_norm": 1.0225199460983276, + "learning_rate": 0.00019680851589076414, + "loss": 0.1845, + "step": 3449 + }, + { + "epoch": 0.08087343307723413, + "grad_norm": 0.5960331559181213, + "learning_rate": 0.00019680666951076196, + "loss": 0.0903, + "step": 3450 + }, + { + "epoch": 0.08089687465203912, + "grad_norm": 0.9923271536827087, + "learning_rate": 0.00019680482260548408, + "loss": 0.1276, + "step": 3451 + }, + { + "epoch": 0.08092031622684412, + "grad_norm": 4.1420722007751465, + "learning_rate": 0.0001968029751749405, + "loss": 0.4891, + "step": 3452 + }, + { + "epoch": 0.08094375780164911, + "grad_norm": 2.1190292835235596, + "learning_rate": 0.00019680112721914123, + "loss": 1.5757, + "step": 3453 + }, + { + "epoch": 0.08096719937645411, + "grad_norm": 4.279684066772461, + "learning_rate": 0.00019679927873809635, + "loss": 0.4319, + "step": 3454 + }, + { + "epoch": 0.0809906409512591, + "grad_norm": 0.5991588234901428, + "learning_rate": 0.00019679742973181585, + "loss": 0.1394, + "step": 3455 + }, + { + "epoch": 0.0810140825260641, + "grad_norm": 3.3578407764434814, + "learning_rate": 0.00019679558020030977, + "loss": 0.584, + "step": 3456 + }, + { + "epoch": 0.08103752410086909, + "grad_norm": 3.1595146656036377, + "learning_rate": 0.00019679373014358818, + "loss": 0.3365, + "step": 3457 + }, + { + "epoch": 0.0810609656756741, + "grad_norm": 2.8398497104644775, + "learning_rate": 0.00019679187956166106, + "loss": 0.586, + "step": 3458 + }, + { + "epoch": 0.08108440725047909, + "grad_norm": 0.9501920342445374, + "learning_rate": 0.0001967900284545385, + "loss": 0.1863, + "step": 3459 + }, + { + "epoch": 0.08110784882528409, + "grad_norm": 0.8161033391952515, + "learning_rate": 0.0001967881768222305, + "loss": 0.1339, + "step": 3460 + }, + { + "epoch": 0.08113129040008908, + "grad_norm": 1.1923494338989258, + "learning_rate": 0.00019678632466474714, + "loss": 0.5592, + "step": 3461 + }, + { + "epoch": 0.08115473197489408, + "grad_norm": 3.8800137042999268, + "learning_rate": 0.00019678447198209846, + "loss": 0.632, + "step": 3462 + }, + { + "epoch": 0.08117817354969907, + "grad_norm": 4.179308891296387, + "learning_rate": 0.00019678261877429453, + "loss": 0.7345, + "step": 3463 + }, + { + "epoch": 0.08120161512450406, + "grad_norm": 3.6680335998535156, + "learning_rate": 0.00019678076504134535, + "loss": 0.2037, + "step": 3464 + }, + { + "epoch": 0.08122505669930906, + "grad_norm": 4.803579807281494, + "learning_rate": 0.00019677891078326108, + "loss": 0.9385, + "step": 3465 + }, + { + "epoch": 0.08124849827411405, + "grad_norm": 0.5965617299079895, + "learning_rate": 0.00019677705600005166, + "loss": 0.0754, + "step": 3466 + }, + { + "epoch": 0.08127193984891905, + "grad_norm": 1.9919737577438354, + "learning_rate": 0.00019677520069172723, + "loss": 0.296, + "step": 3467 + }, + { + "epoch": 0.08129538142372404, + "grad_norm": 2.268587112426758, + "learning_rate": 0.00019677334485829784, + "loss": 0.256, + "step": 3468 + }, + { + "epoch": 0.08131882299852904, + "grad_norm": 4.290424346923828, + "learning_rate": 0.00019677148849977355, + "loss": 0.5963, + "step": 3469 + }, + { + "epoch": 0.08134226457333403, + "grad_norm": 2.4352986812591553, + "learning_rate": 0.0001967696316161645, + "loss": 1.7272, + "step": 3470 + }, + { + "epoch": 0.08136570614813904, + "grad_norm": 1.5841773748397827, + "learning_rate": 0.00019676777420748066, + "loss": 0.2934, + "step": 3471 + }, + { + "epoch": 0.08138914772294403, + "grad_norm": 3.1363131999969482, + "learning_rate": 0.00019676591627373216, + "loss": 0.4055, + "step": 3472 + }, + { + "epoch": 0.08141258929774903, + "grad_norm": 3.8610708713531494, + "learning_rate": 0.00019676405781492905, + "loss": 0.5217, + "step": 3473 + }, + { + "epoch": 0.08143603087255402, + "grad_norm": 2.6618497371673584, + "learning_rate": 0.0001967621988310815, + "loss": 0.3333, + "step": 3474 + }, + { + "epoch": 0.08145947244735902, + "grad_norm": 3.221708297729492, + "learning_rate": 0.0001967603393221995, + "loss": 0.2701, + "step": 3475 + }, + { + "epoch": 0.08148291402216401, + "grad_norm": 4.543689250946045, + "learning_rate": 0.0001967584792882932, + "loss": 0.5736, + "step": 3476 + }, + { + "epoch": 0.081506355596969, + "grad_norm": 3.1516454219818115, + "learning_rate": 0.00019675661872937267, + "loss": 0.3919, + "step": 3477 + }, + { + "epoch": 0.081529797171774, + "grad_norm": 3.6271495819091797, + "learning_rate": 0.00019675475764544797, + "loss": 0.6759, + "step": 3478 + }, + { + "epoch": 0.081553238746579, + "grad_norm": 0.5375126600265503, + "learning_rate": 0.00019675289603652928, + "loss": 0.0536, + "step": 3479 + }, + { + "epoch": 0.08157668032138399, + "grad_norm": 3.131989002227783, + "learning_rate": 0.0001967510339026266, + "loss": 0.7964, + "step": 3480 + }, + { + "epoch": 0.08160012189618898, + "grad_norm": 4.8787336349487305, + "learning_rate": 0.00019674917124375013, + "loss": 0.4914, + "step": 3481 + }, + { + "epoch": 0.08162356347099398, + "grad_norm": 2.5133767127990723, + "learning_rate": 0.00019674730805990996, + "loss": 0.3662, + "step": 3482 + }, + { + "epoch": 0.08164700504579897, + "grad_norm": 1.3573399782180786, + "learning_rate": 0.00019674544435111614, + "loss": 0.143, + "step": 3483 + }, + { + "epoch": 0.08167044662060398, + "grad_norm": 4.542501926422119, + "learning_rate": 0.00019674358011737881, + "loss": 0.7743, + "step": 3484 + }, + { + "epoch": 0.08169388819540897, + "grad_norm": 1.9427406787872314, + "learning_rate": 0.00019674171535870812, + "loss": 0.1991, + "step": 3485 + }, + { + "epoch": 0.08171732977021397, + "grad_norm": 4.229214668273926, + "learning_rate": 0.00019673985007511414, + "loss": 0.56, + "step": 3486 + }, + { + "epoch": 0.08174077134501896, + "grad_norm": 3.626979351043701, + "learning_rate": 0.00019673798426660704, + "loss": 0.4925, + "step": 3487 + }, + { + "epoch": 0.08176421291982396, + "grad_norm": 1.3352084159851074, + "learning_rate": 0.00019673611793319688, + "loss": 0.1411, + "step": 3488 + }, + { + "epoch": 0.08178765449462895, + "grad_norm": 3.214003801345825, + "learning_rate": 0.00019673425107489387, + "loss": 0.5573, + "step": 3489 + }, + { + "epoch": 0.08181109606943394, + "grad_norm": 0.9281025528907776, + "learning_rate": 0.00019673238369170804, + "loss": 0.1218, + "step": 3490 + }, + { + "epoch": 0.08183453764423894, + "grad_norm": 2.2301902770996094, + "learning_rate": 0.00019673051578364965, + "loss": 0.292, + "step": 3491 + }, + { + "epoch": 0.08185797921904393, + "grad_norm": 2.3824758529663086, + "learning_rate": 0.00019672864735072868, + "loss": 1.6716, + "step": 3492 + }, + { + "epoch": 0.08188142079384893, + "grad_norm": 3.611151695251465, + "learning_rate": 0.0001967267783929554, + "loss": 0.6585, + "step": 3493 + }, + { + "epoch": 0.08190486236865392, + "grad_norm": 0.36214667558670044, + "learning_rate": 0.0001967249089103399, + "loss": 0.0423, + "step": 3494 + }, + { + "epoch": 0.08192830394345892, + "grad_norm": 4.445809364318848, + "learning_rate": 0.00019672303890289232, + "loss": 0.4599, + "step": 3495 + }, + { + "epoch": 0.08195174551826391, + "grad_norm": 1.3084452152252197, + "learning_rate": 0.0001967211683706228, + "loss": 0.2093, + "step": 3496 + }, + { + "epoch": 0.08197518709306892, + "grad_norm": 1.3788717985153198, + "learning_rate": 0.0001967192973135415, + "loss": 0.2676, + "step": 3497 + }, + { + "epoch": 0.08199862866787391, + "grad_norm": 10.570982933044434, + "learning_rate": 0.0001967174257316586, + "loss": 1.0712, + "step": 3498 + }, + { + "epoch": 0.08202207024267891, + "grad_norm": 3.2099764347076416, + "learning_rate": 0.00019671555362498418, + "loss": 0.4722, + "step": 3499 + }, + { + "epoch": 0.0820455118174839, + "grad_norm": 2.6007275581359863, + "learning_rate": 0.0001967136809935285, + "loss": 0.5081, + "step": 3500 + }, + { + "epoch": 0.0820689533922889, + "grad_norm": 3.7262990474700928, + "learning_rate": 0.00019671180783730163, + "loss": 0.5176, + "step": 3501 + }, + { + "epoch": 0.08209239496709389, + "grad_norm": 3.147481918334961, + "learning_rate": 0.0001967099341563138, + "loss": 0.3049, + "step": 3502 + }, + { + "epoch": 0.08211583654189888, + "grad_norm": 5.602180480957031, + "learning_rate": 0.0001967080599505751, + "loss": 0.8506, + "step": 3503 + }, + { + "epoch": 0.08213927811670388, + "grad_norm": 3.8238775730133057, + "learning_rate": 0.0001967061852200958, + "loss": 0.6849, + "step": 3504 + }, + { + "epoch": 0.08216271969150887, + "grad_norm": 3.7398507595062256, + "learning_rate": 0.00019670430996488597, + "loss": 0.553, + "step": 3505 + }, + { + "epoch": 0.08218616126631387, + "grad_norm": 4.21665620803833, + "learning_rate": 0.00019670243418495588, + "loss": 0.7647, + "step": 3506 + }, + { + "epoch": 0.08220960284111886, + "grad_norm": 2.426018476486206, + "learning_rate": 0.00019670055788031566, + "loss": 2.0673, + "step": 3507 + }, + { + "epoch": 0.08223304441592386, + "grad_norm": 3.640014171600342, + "learning_rate": 0.0001966986810509755, + "loss": 0.3372, + "step": 3508 + }, + { + "epoch": 0.08225648599072885, + "grad_norm": 1.2531019449234009, + "learning_rate": 0.00019669680369694553, + "loss": 0.2089, + "step": 3509 + }, + { + "epoch": 0.08227992756553386, + "grad_norm": 0.3082675337791443, + "learning_rate": 0.000196694925818236, + "loss": 0.0431, + "step": 3510 + }, + { + "epoch": 0.08230336914033885, + "grad_norm": 1.8203505277633667, + "learning_rate": 0.00019669304741485712, + "loss": 1.3717, + "step": 3511 + }, + { + "epoch": 0.08232681071514385, + "grad_norm": 2.7782697677612305, + "learning_rate": 0.000196691168486819, + "loss": 0.6265, + "step": 3512 + }, + { + "epoch": 0.08235025228994884, + "grad_norm": 5.230659008026123, + "learning_rate": 0.00019668928903413191, + "loss": 0.8748, + "step": 3513 + }, + { + "epoch": 0.08237369386475384, + "grad_norm": 0.9989299178123474, + "learning_rate": 0.000196687409056806, + "loss": 0.6005, + "step": 3514 + }, + { + "epoch": 0.08239713543955883, + "grad_norm": 3.725569248199463, + "learning_rate": 0.00019668552855485154, + "loss": 0.4534, + "step": 3515 + }, + { + "epoch": 0.08242057701436382, + "grad_norm": 4.369661808013916, + "learning_rate": 0.00019668364752827862, + "loss": 0.8389, + "step": 3516 + }, + { + "epoch": 0.08244401858916882, + "grad_norm": 4.806898593902588, + "learning_rate": 0.00019668176597709754, + "loss": 0.6402, + "step": 3517 + }, + { + "epoch": 0.08246746016397381, + "grad_norm": 3.1354424953460693, + "learning_rate": 0.0001966798839013185, + "loss": 0.7045, + "step": 3518 + }, + { + "epoch": 0.08249090173877881, + "grad_norm": 4.500845909118652, + "learning_rate": 0.00019667800130095163, + "loss": 0.5706, + "step": 3519 + }, + { + "epoch": 0.0825143433135838, + "grad_norm": 1.1621520519256592, + "learning_rate": 0.00019667611817600727, + "loss": 0.1003, + "step": 3520 + }, + { + "epoch": 0.0825377848883888, + "grad_norm": 3.7855112552642822, + "learning_rate": 0.00019667423452649554, + "loss": 0.5375, + "step": 3521 + }, + { + "epoch": 0.0825612264631938, + "grad_norm": 3.4982492923736572, + "learning_rate": 0.0001966723503524267, + "loss": 0.8496, + "step": 3522 + }, + { + "epoch": 0.0825846680379988, + "grad_norm": 5.399875640869141, + "learning_rate": 0.00019667046565381097, + "loss": 0.9081, + "step": 3523 + }, + { + "epoch": 0.08260810961280379, + "grad_norm": 4.697774410247803, + "learning_rate": 0.00019666858043065858, + "loss": 0.9835, + "step": 3524 + }, + { + "epoch": 0.08263155118760879, + "grad_norm": 0.6922926902770996, + "learning_rate": 0.00019666669468297975, + "loss": 0.1441, + "step": 3525 + }, + { + "epoch": 0.08265499276241378, + "grad_norm": 3.402787446975708, + "learning_rate": 0.00019666480841078472, + "loss": 0.7665, + "step": 3526 + }, + { + "epoch": 0.08267843433721878, + "grad_norm": 3.110142230987549, + "learning_rate": 0.0001966629216140837, + "loss": 0.5972, + "step": 3527 + }, + { + "epoch": 0.08270187591202377, + "grad_norm": 3.144869804382324, + "learning_rate": 0.000196661034292887, + "loss": 1.3197, + "step": 3528 + }, + { + "epoch": 0.08272531748682876, + "grad_norm": 4.109544277191162, + "learning_rate": 0.00019665914644720477, + "loss": 1.0992, + "step": 3529 + }, + { + "epoch": 0.08274875906163376, + "grad_norm": 1.9699876308441162, + "learning_rate": 0.0001966572580770473, + "loss": 0.293, + "step": 3530 + }, + { + "epoch": 0.08277220063643875, + "grad_norm": 3.1023313999176025, + "learning_rate": 0.00019665536918242484, + "loss": 0.6377, + "step": 3531 + }, + { + "epoch": 0.08279564221124375, + "grad_norm": 3.33605694770813, + "learning_rate": 0.0001966534797633476, + "loss": 0.6696, + "step": 3532 + }, + { + "epoch": 0.08281908378604874, + "grad_norm": 4.045075416564941, + "learning_rate": 0.0001966515898198259, + "loss": 0.6965, + "step": 3533 + }, + { + "epoch": 0.08284252536085374, + "grad_norm": 3.7490053176879883, + "learning_rate": 0.00019664969935186995, + "loss": 0.7591, + "step": 3534 + }, + { + "epoch": 0.08286596693565874, + "grad_norm": 3.1613683700561523, + "learning_rate": 0.00019664780835948998, + "loss": 0.4797, + "step": 3535 + }, + { + "epoch": 0.08288940851046374, + "grad_norm": 3.356184959411621, + "learning_rate": 0.0001966459168426963, + "loss": 0.4516, + "step": 3536 + }, + { + "epoch": 0.08291285008526873, + "grad_norm": 4.104770183563232, + "learning_rate": 0.00019664402480149918, + "loss": 0.8461, + "step": 3537 + }, + { + "epoch": 0.08293629166007373, + "grad_norm": 4.725372791290283, + "learning_rate": 0.00019664213223590885, + "loss": 0.7787, + "step": 3538 + }, + { + "epoch": 0.08295973323487872, + "grad_norm": 3.8901560306549072, + "learning_rate": 0.0001966402391459356, + "loss": 0.6328, + "step": 3539 + }, + { + "epoch": 0.08298317480968372, + "grad_norm": 2.327664613723755, + "learning_rate": 0.00019663834553158968, + "loss": 1.5318, + "step": 3540 + }, + { + "epoch": 0.08300661638448871, + "grad_norm": 2.8211002349853516, + "learning_rate": 0.00019663645139288139, + "loss": 0.6804, + "step": 3541 + }, + { + "epoch": 0.0830300579592937, + "grad_norm": 3.631463050842285, + "learning_rate": 0.000196634556729821, + "loss": 0.5918, + "step": 3542 + }, + { + "epoch": 0.0830534995340987, + "grad_norm": 3.6729471683502197, + "learning_rate": 0.00019663266154241877, + "loss": 0.541, + "step": 3543 + }, + { + "epoch": 0.08307694110890369, + "grad_norm": 1.2151833772659302, + "learning_rate": 0.00019663076583068502, + "loss": 0.2356, + "step": 3544 + }, + { + "epoch": 0.08310038268370869, + "grad_norm": 4.365105628967285, + "learning_rate": 0.00019662886959463002, + "loss": 0.8982, + "step": 3545 + }, + { + "epoch": 0.08312382425851368, + "grad_norm": 5.178152561187744, + "learning_rate": 0.00019662697283426406, + "loss": 0.6543, + "step": 3546 + }, + { + "epoch": 0.08314726583331868, + "grad_norm": 0.5532276034355164, + "learning_rate": 0.0001966250755495974, + "loss": 0.0501, + "step": 3547 + }, + { + "epoch": 0.08317070740812368, + "grad_norm": 5.040855407714844, + "learning_rate": 0.0001966231777406404, + "loss": 0.881, + "step": 3548 + }, + { + "epoch": 0.08319414898292868, + "grad_norm": 4.309007167816162, + "learning_rate": 0.0001966212794074033, + "loss": 0.6026, + "step": 3549 + }, + { + "epoch": 0.08321759055773367, + "grad_norm": 0.574213445186615, + "learning_rate": 0.00019661938054989643, + "loss": 0.0854, + "step": 3550 + }, + { + "epoch": 0.08324103213253867, + "grad_norm": 3.586395502090454, + "learning_rate": 0.00019661748116813006, + "loss": 0.6861, + "step": 3551 + }, + { + "epoch": 0.08326447370734366, + "grad_norm": 4.5727105140686035, + "learning_rate": 0.00019661558126211453, + "loss": 0.5639, + "step": 3552 + }, + { + "epoch": 0.08328791528214866, + "grad_norm": 5.5915632247924805, + "learning_rate": 0.00019661368083186012, + "loss": 1.3211, + "step": 3553 + }, + { + "epoch": 0.08331135685695365, + "grad_norm": 3.543743848800659, + "learning_rate": 0.00019661177987737721, + "loss": 0.5243, + "step": 3554 + }, + { + "epoch": 0.08333479843175864, + "grad_norm": 4.140334129333496, + "learning_rate": 0.00019660987839867605, + "loss": 0.6764, + "step": 3555 + }, + { + "epoch": 0.08335824000656364, + "grad_norm": 2.855027437210083, + "learning_rate": 0.00019660797639576696, + "loss": 0.5367, + "step": 3556 + }, + { + "epoch": 0.08338168158136863, + "grad_norm": 1.3382161855697632, + "learning_rate": 0.0001966060738686603, + "loss": 0.159, + "step": 3557 + }, + { + "epoch": 0.08340512315617363, + "grad_norm": 2.788195848464966, + "learning_rate": 0.00019660417081736632, + "loss": 0.4769, + "step": 3558 + }, + { + "epoch": 0.08342856473097862, + "grad_norm": 1.3381305932998657, + "learning_rate": 0.00019660226724189543, + "loss": 0.7224, + "step": 3559 + }, + { + "epoch": 0.08345200630578362, + "grad_norm": 3.2295682430267334, + "learning_rate": 0.00019660036314225792, + "loss": 0.4095, + "step": 3560 + }, + { + "epoch": 0.08347544788058862, + "grad_norm": 1.3464120626449585, + "learning_rate": 0.0001965984585184641, + "loss": 0.1607, + "step": 3561 + }, + { + "epoch": 0.08349888945539362, + "grad_norm": 2.2302069664001465, + "learning_rate": 0.00019659655337052434, + "loss": 0.206, + "step": 3562 + }, + { + "epoch": 0.08352233103019861, + "grad_norm": 1.3210262060165405, + "learning_rate": 0.000196594647698449, + "loss": 0.2631, + "step": 3563 + }, + { + "epoch": 0.08354577260500361, + "grad_norm": 1.6268714666366577, + "learning_rate": 0.00019659274150224835, + "loss": 1.2899, + "step": 3564 + }, + { + "epoch": 0.0835692141798086, + "grad_norm": 3.9666826725006104, + "learning_rate": 0.00019659083478193275, + "loss": 0.7287, + "step": 3565 + }, + { + "epoch": 0.0835926557546136, + "grad_norm": 3.834594488143921, + "learning_rate": 0.0001965889275375126, + "loss": 0.5943, + "step": 3566 + }, + { + "epoch": 0.08361609732941859, + "grad_norm": 3.8186185359954834, + "learning_rate": 0.0001965870197689982, + "loss": 0.6599, + "step": 3567 + }, + { + "epoch": 0.08363953890422358, + "grad_norm": 3.52095890045166, + "learning_rate": 0.0001965851114763999, + "loss": 0.5766, + "step": 3568 + }, + { + "epoch": 0.08366298047902858, + "grad_norm": 3.121852159500122, + "learning_rate": 0.0001965832026597281, + "loss": 0.2394, + "step": 3569 + }, + { + "epoch": 0.08368642205383357, + "grad_norm": 1.4733325242996216, + "learning_rate": 0.0001965812933189931, + "loss": 0.1712, + "step": 3570 + }, + { + "epoch": 0.08370986362863857, + "grad_norm": 4.674515724182129, + "learning_rate": 0.0001965793834542053, + "loss": 0.7287, + "step": 3571 + }, + { + "epoch": 0.08373330520344356, + "grad_norm": 3.5747740268707275, + "learning_rate": 0.00019657747306537507, + "loss": 0.3278, + "step": 3572 + }, + { + "epoch": 0.08375674677824856, + "grad_norm": 1.5613502264022827, + "learning_rate": 0.0001965755621525127, + "loss": 0.1252, + "step": 3573 + }, + { + "epoch": 0.08378018835305356, + "grad_norm": 4.5950775146484375, + "learning_rate": 0.00019657365071562867, + "loss": 0.4474, + "step": 3574 + }, + { + "epoch": 0.08380362992785856, + "grad_norm": 1.9803239107131958, + "learning_rate": 0.00019657173875473326, + "loss": 1.2669, + "step": 3575 + }, + { + "epoch": 0.08382707150266355, + "grad_norm": 3.042163372039795, + "learning_rate": 0.0001965698262698369, + "loss": 0.4992, + "step": 3576 + }, + { + "epoch": 0.08385051307746855, + "grad_norm": 4.016462326049805, + "learning_rate": 0.00019656791326094997, + "loss": 0.4406, + "step": 3577 + }, + { + "epoch": 0.08387395465227354, + "grad_norm": 4.954759120941162, + "learning_rate": 0.0001965659997280828, + "loss": 0.5269, + "step": 3578 + }, + { + "epoch": 0.08389739622707854, + "grad_norm": 5.2274489402771, + "learning_rate": 0.00019656408567124577, + "loss": 0.5676, + "step": 3579 + }, + { + "epoch": 0.08392083780188353, + "grad_norm": 4.376159191131592, + "learning_rate": 0.00019656217109044932, + "loss": 0.6291, + "step": 3580 + }, + { + "epoch": 0.08394427937668852, + "grad_norm": 4.689589500427246, + "learning_rate": 0.00019656025598570383, + "loss": 0.7928, + "step": 3581 + }, + { + "epoch": 0.08396772095149352, + "grad_norm": 3.7995567321777344, + "learning_rate": 0.00019655834035701967, + "loss": 0.5405, + "step": 3582 + }, + { + "epoch": 0.08399116252629851, + "grad_norm": 0.5493695139884949, + "learning_rate": 0.00019655642420440724, + "loss": 0.0702, + "step": 3583 + }, + { + "epoch": 0.08401460410110351, + "grad_norm": 4.495513439178467, + "learning_rate": 0.00019655450752787692, + "loss": 0.6136, + "step": 3584 + }, + { + "epoch": 0.0840380456759085, + "grad_norm": 2.3854849338531494, + "learning_rate": 0.00019655259032743915, + "loss": 1.231, + "step": 3585 + }, + { + "epoch": 0.08406148725071351, + "grad_norm": 2.511749505996704, + "learning_rate": 0.0001965506726031043, + "loss": 0.2683, + "step": 3586 + }, + { + "epoch": 0.0840849288255185, + "grad_norm": 3.5913984775543213, + "learning_rate": 0.0001965487543548828, + "loss": 0.3278, + "step": 3587 + }, + { + "epoch": 0.0841083704003235, + "grad_norm": 5.078485012054443, + "learning_rate": 0.00019654683558278504, + "loss": 0.5535, + "step": 3588 + }, + { + "epoch": 0.08413181197512849, + "grad_norm": 3.0445683002471924, + "learning_rate": 0.00019654491628682142, + "loss": 1.3025, + "step": 3589 + }, + { + "epoch": 0.08415525354993349, + "grad_norm": 4.045226573944092, + "learning_rate": 0.00019654299646700238, + "loss": 0.8895, + "step": 3590 + }, + { + "epoch": 0.08417869512473848, + "grad_norm": 2.8822672367095947, + "learning_rate": 0.0001965410761233383, + "loss": 0.2617, + "step": 3591 + }, + { + "epoch": 0.08420213669954348, + "grad_norm": 4.495993614196777, + "learning_rate": 0.00019653915525583965, + "loss": 0.7024, + "step": 3592 + }, + { + "epoch": 0.08422557827434847, + "grad_norm": 3.5510048866271973, + "learning_rate": 0.00019653723386451683, + "loss": 0.5907, + "step": 3593 + }, + { + "epoch": 0.08424901984915346, + "grad_norm": 3.3124942779541016, + "learning_rate": 0.0001965353119493803, + "loss": 0.4882, + "step": 3594 + }, + { + "epoch": 0.08427246142395846, + "grad_norm": 3.3800759315490723, + "learning_rate": 0.00019653338951044038, + "loss": 0.4544, + "step": 3595 + }, + { + "epoch": 0.08429590299876345, + "grad_norm": 2.5833277702331543, + "learning_rate": 0.00019653146654770762, + "loss": 0.361, + "step": 3596 + }, + { + "epoch": 0.08431934457356845, + "grad_norm": 2.1363584995269775, + "learning_rate": 0.0001965295430611924, + "loss": 1.6392, + "step": 3597 + }, + { + "epoch": 0.08434278614837344, + "grad_norm": 2.0787649154663086, + "learning_rate": 0.00019652761905090515, + "loss": 0.2192, + "step": 3598 + }, + { + "epoch": 0.08436622772317845, + "grad_norm": 3.0229854583740234, + "learning_rate": 0.00019652569451685637, + "loss": 0.3723, + "step": 3599 + }, + { + "epoch": 0.08438966929798344, + "grad_norm": 4.53194522857666, + "learning_rate": 0.00019652376945905638, + "loss": 0.8339, + "step": 3600 + }, + { + "epoch": 0.08441311087278844, + "grad_norm": 3.797182321548462, + "learning_rate": 0.00019652184387751578, + "loss": 0.4599, + "step": 3601 + }, + { + "epoch": 0.08443655244759343, + "grad_norm": 3.6464791297912598, + "learning_rate": 0.0001965199177722449, + "loss": 1.756, + "step": 3602 + }, + { + "epoch": 0.08445999402239843, + "grad_norm": 4.100632190704346, + "learning_rate": 0.00019651799114325425, + "loss": 0.7139, + "step": 3603 + }, + { + "epoch": 0.08448343559720342, + "grad_norm": 1.5954639911651611, + "learning_rate": 0.00019651606399055424, + "loss": 0.1471, + "step": 3604 + }, + { + "epoch": 0.08450687717200842, + "grad_norm": 5.264995098114014, + "learning_rate": 0.0001965141363141554, + "loss": 0.6691, + "step": 3605 + }, + { + "epoch": 0.08453031874681341, + "grad_norm": 3.9446773529052734, + "learning_rate": 0.00019651220811406808, + "loss": 0.431, + "step": 3606 + }, + { + "epoch": 0.0845537603216184, + "grad_norm": 2.247377872467041, + "learning_rate": 0.00019651027939030286, + "loss": 1.101, + "step": 3607 + }, + { + "epoch": 0.0845772018964234, + "grad_norm": 4.1538896560668945, + "learning_rate": 0.00019650835014287013, + "loss": 0.5125, + "step": 3608 + }, + { + "epoch": 0.08460064347122839, + "grad_norm": 3.5536863803863525, + "learning_rate": 0.00019650642037178037, + "loss": 0.3979, + "step": 3609 + }, + { + "epoch": 0.08462408504603339, + "grad_norm": 1.8208982944488525, + "learning_rate": 0.0001965044900770441, + "loss": 0.9596, + "step": 3610 + }, + { + "epoch": 0.08464752662083838, + "grad_norm": 3.995274782180786, + "learning_rate": 0.0001965025592586717, + "loss": 0.6208, + "step": 3611 + }, + { + "epoch": 0.08467096819564339, + "grad_norm": 3.4222335815429688, + "learning_rate": 0.0001965006279166737, + "loss": 0.4571, + "step": 3612 + }, + { + "epoch": 0.08469440977044838, + "grad_norm": 3.6341025829315186, + "learning_rate": 0.00019649869605106065, + "loss": 0.3483, + "step": 3613 + }, + { + "epoch": 0.08471785134525338, + "grad_norm": 3.7179453372955322, + "learning_rate": 0.00019649676366184288, + "loss": 0.6165, + "step": 3614 + }, + { + "epoch": 0.08474129292005837, + "grad_norm": 4.656064033508301, + "learning_rate": 0.000196494830749031, + "loss": 0.534, + "step": 3615 + }, + { + "epoch": 0.08476473449486337, + "grad_norm": 3.104191303253174, + "learning_rate": 0.00019649289731263542, + "loss": 0.3746, + "step": 3616 + }, + { + "epoch": 0.08478817606966836, + "grad_norm": 1.5984156131744385, + "learning_rate": 0.00019649096335266672, + "loss": 0.2043, + "step": 3617 + }, + { + "epoch": 0.08481161764447336, + "grad_norm": 5.379594802856445, + "learning_rate": 0.00019648902886913532, + "loss": 0.5444, + "step": 3618 + }, + { + "epoch": 0.08483505921927835, + "grad_norm": 4.221158981323242, + "learning_rate": 0.00019648709386205172, + "loss": 0.7542, + "step": 3619 + }, + { + "epoch": 0.08485850079408334, + "grad_norm": 3.7386178970336914, + "learning_rate": 0.00019648515833142643, + "loss": 0.348, + "step": 3620 + }, + { + "epoch": 0.08488194236888834, + "grad_norm": 0.9064016342163086, + "learning_rate": 0.00019648322227726998, + "loss": 0.0901, + "step": 3621 + }, + { + "epoch": 0.08490538394369333, + "grad_norm": 5.634291172027588, + "learning_rate": 0.00019648128569959284, + "loss": 0.5087, + "step": 3622 + }, + { + "epoch": 0.08492882551849833, + "grad_norm": 1.7243678569793701, + "learning_rate": 0.00019647934859840554, + "loss": 0.3177, + "step": 3623 + }, + { + "epoch": 0.08495226709330332, + "grad_norm": 3.404330253601074, + "learning_rate": 0.00019647741097371857, + "loss": 0.3976, + "step": 3624 + }, + { + "epoch": 0.08497570866810833, + "grad_norm": 5.168160438537598, + "learning_rate": 0.00019647547282554246, + "loss": 0.6277, + "step": 3625 + }, + { + "epoch": 0.08499915024291332, + "grad_norm": 4.901256084442139, + "learning_rate": 0.00019647353415388775, + "loss": 0.6078, + "step": 3626 + }, + { + "epoch": 0.08502259181771832, + "grad_norm": 3.3751087188720703, + "learning_rate": 0.00019647159495876488, + "loss": 0.3284, + "step": 3627 + }, + { + "epoch": 0.08504603339252331, + "grad_norm": 5.318412780761719, + "learning_rate": 0.00019646965524018444, + "loss": 0.8104, + "step": 3628 + }, + { + "epoch": 0.0850694749673283, + "grad_norm": 3.312452554702759, + "learning_rate": 0.00019646771499815695, + "loss": 0.564, + "step": 3629 + }, + { + "epoch": 0.0850929165421333, + "grad_norm": 2.599127769470215, + "learning_rate": 0.00019646577423269293, + "loss": 0.317, + "step": 3630 + }, + { + "epoch": 0.0851163581169383, + "grad_norm": 4.656958103179932, + "learning_rate": 0.00019646383294380287, + "loss": 0.7541, + "step": 3631 + }, + { + "epoch": 0.08513979969174329, + "grad_norm": 1.3316069841384888, + "learning_rate": 0.0001964618911314974, + "loss": 0.2197, + "step": 3632 + }, + { + "epoch": 0.08516324126654828, + "grad_norm": 1.2798737287521362, + "learning_rate": 0.00019645994879578696, + "loss": 0.1329, + "step": 3633 + }, + { + "epoch": 0.08518668284135328, + "grad_norm": 3.4946038722991943, + "learning_rate": 0.00019645800593668213, + "loss": 0.6409, + "step": 3634 + }, + { + "epoch": 0.08521012441615827, + "grad_norm": 3.8511767387390137, + "learning_rate": 0.00019645606255419345, + "loss": 0.5546, + "step": 3635 + }, + { + "epoch": 0.08523356599096327, + "grad_norm": 3.3006491661071777, + "learning_rate": 0.0001964541186483315, + "loss": 0.4015, + "step": 3636 + }, + { + "epoch": 0.08525700756576828, + "grad_norm": 4.037435054779053, + "learning_rate": 0.0001964521742191067, + "loss": 0.737, + "step": 3637 + }, + { + "epoch": 0.08528044914057327, + "grad_norm": 3.607347011566162, + "learning_rate": 0.00019645022926652973, + "loss": 0.6248, + "step": 3638 + }, + { + "epoch": 0.08530389071537826, + "grad_norm": 0.7934058904647827, + "learning_rate": 0.00019644828379061116, + "loss": 0.1014, + "step": 3639 + }, + { + "epoch": 0.08532733229018326, + "grad_norm": 3.7899441719055176, + "learning_rate": 0.00019644633779136142, + "loss": 0.6453, + "step": 3640 + }, + { + "epoch": 0.08535077386498825, + "grad_norm": 3.5224075317382812, + "learning_rate": 0.00019644439126879118, + "loss": 0.722, + "step": 3641 + }, + { + "epoch": 0.08537421543979325, + "grad_norm": 1.0786101818084717, + "learning_rate": 0.00019644244422291094, + "loss": 0.1475, + "step": 3642 + }, + { + "epoch": 0.08539765701459824, + "grad_norm": 2.9227423667907715, + "learning_rate": 0.0001964404966537313, + "loss": 0.6031, + "step": 3643 + }, + { + "epoch": 0.08542109858940324, + "grad_norm": 2.7054343223571777, + "learning_rate": 0.0001964385485612628, + "loss": 0.4471, + "step": 3644 + }, + { + "epoch": 0.08544454016420823, + "grad_norm": 1.452135682106018, + "learning_rate": 0.00019643659994551604, + "loss": 0.1683, + "step": 3645 + }, + { + "epoch": 0.08546798173901322, + "grad_norm": 4.20982027053833, + "learning_rate": 0.00019643465080650154, + "loss": 0.6093, + "step": 3646 + }, + { + "epoch": 0.08549142331381822, + "grad_norm": 2.113847255706787, + "learning_rate": 0.00019643270114422996, + "loss": 1.7188, + "step": 3647 + }, + { + "epoch": 0.08551486488862321, + "grad_norm": 3.9071242809295654, + "learning_rate": 0.00019643075095871182, + "loss": 0.4996, + "step": 3648 + }, + { + "epoch": 0.08553830646342821, + "grad_norm": 1.180292010307312, + "learning_rate": 0.0001964288002499577, + "loss": 0.2217, + "step": 3649 + }, + { + "epoch": 0.08556174803823322, + "grad_norm": 4.1482133865356445, + "learning_rate": 0.0001964268490179782, + "loss": 0.6209, + "step": 3650 + }, + { + "epoch": 0.08558518961303821, + "grad_norm": 3.826991319656372, + "learning_rate": 0.00019642489726278393, + "loss": 0.8132, + "step": 3651 + }, + { + "epoch": 0.0856086311878432, + "grad_norm": 2.0154383182525635, + "learning_rate": 0.00019642294498438542, + "loss": 1.5762, + "step": 3652 + }, + { + "epoch": 0.0856320727626482, + "grad_norm": 3.211667060852051, + "learning_rate": 0.0001964209921827933, + "loss": 0.5212, + "step": 3653 + }, + { + "epoch": 0.08565551433745319, + "grad_norm": 1.0993330478668213, + "learning_rate": 0.0001964190388580182, + "loss": 0.1346, + "step": 3654 + }, + { + "epoch": 0.08567895591225819, + "grad_norm": 2.3120245933532715, + "learning_rate": 0.00019641708501007065, + "loss": 0.3038, + "step": 3655 + }, + { + "epoch": 0.08570239748706318, + "grad_norm": 2.3911526203155518, + "learning_rate": 0.00019641513063896131, + "loss": 1.9474, + "step": 3656 + }, + { + "epoch": 0.08572583906186818, + "grad_norm": 1.5069551467895508, + "learning_rate": 0.00019641317574470074, + "loss": 0.37, + "step": 3657 + }, + { + "epoch": 0.08574928063667317, + "grad_norm": 2.169790506362915, + "learning_rate": 0.0001964112203272996, + "loss": 0.3942, + "step": 3658 + }, + { + "epoch": 0.08577272221147816, + "grad_norm": 3.713595151901245, + "learning_rate": 0.00019640926438676846, + "loss": 0.5069, + "step": 3659 + }, + { + "epoch": 0.08579616378628316, + "grad_norm": 1.906488299369812, + "learning_rate": 0.0001964073079231179, + "loss": 0.2666, + "step": 3660 + }, + { + "epoch": 0.08581960536108815, + "grad_norm": 3.942178726196289, + "learning_rate": 0.00019640535093635864, + "loss": 0.3827, + "step": 3661 + }, + { + "epoch": 0.08584304693589315, + "grad_norm": 3.3405253887176514, + "learning_rate": 0.00019640339342650117, + "loss": 0.4501, + "step": 3662 + }, + { + "epoch": 0.08586648851069816, + "grad_norm": 4.70729398727417, + "learning_rate": 0.0001964014353935562, + "loss": 1.0207, + "step": 3663 + }, + { + "epoch": 0.08588993008550315, + "grad_norm": 3.1626477241516113, + "learning_rate": 0.00019639947683753434, + "loss": 0.4373, + "step": 3664 + }, + { + "epoch": 0.08591337166030814, + "grad_norm": 2.091336727142334, + "learning_rate": 0.0001963975177584462, + "loss": 0.1358, + "step": 3665 + }, + { + "epoch": 0.08593681323511314, + "grad_norm": 4.210812091827393, + "learning_rate": 0.00019639555815630243, + "loss": 0.536, + "step": 3666 + }, + { + "epoch": 0.08596025480991813, + "grad_norm": 2.90686297416687, + "learning_rate": 0.00019639359803111365, + "loss": 0.4825, + "step": 3667 + }, + { + "epoch": 0.08598369638472313, + "grad_norm": 3.9352357387542725, + "learning_rate": 0.0001963916373828905, + "loss": 0.6799, + "step": 3668 + }, + { + "epoch": 0.08600713795952812, + "grad_norm": 3.2444725036621094, + "learning_rate": 0.00019638967621164358, + "loss": 0.5454, + "step": 3669 + }, + { + "epoch": 0.08603057953433312, + "grad_norm": 4.721273899078369, + "learning_rate": 0.00019638771451738362, + "loss": 0.6456, + "step": 3670 + }, + { + "epoch": 0.08605402110913811, + "grad_norm": 3.3188743591308594, + "learning_rate": 0.00019638575230012116, + "loss": 0.3285, + "step": 3671 + }, + { + "epoch": 0.0860774626839431, + "grad_norm": 2.541529655456543, + "learning_rate": 0.00019638378955986693, + "loss": 0.5424, + "step": 3672 + }, + { + "epoch": 0.0861009042587481, + "grad_norm": 4.504644870758057, + "learning_rate": 0.00019638182629663154, + "loss": 0.4286, + "step": 3673 + }, + { + "epoch": 0.08612434583355309, + "grad_norm": 5.693789005279541, + "learning_rate": 0.00019637986251042565, + "loss": 0.5225, + "step": 3674 + }, + { + "epoch": 0.08614778740835809, + "grad_norm": 5.74249792098999, + "learning_rate": 0.00019637789820125994, + "loss": 0.9174, + "step": 3675 + }, + { + "epoch": 0.0861712289831631, + "grad_norm": 4.980351448059082, + "learning_rate": 0.000196375933369145, + "loss": 0.5432, + "step": 3676 + }, + { + "epoch": 0.08619467055796809, + "grad_norm": 5.283028602600098, + "learning_rate": 0.00019637396801409156, + "loss": 0.6049, + "step": 3677 + }, + { + "epoch": 0.08621811213277308, + "grad_norm": 2.866577386856079, + "learning_rate": 0.00019637200213611027, + "loss": 0.2219, + "step": 3678 + }, + { + "epoch": 0.08624155370757808, + "grad_norm": 6.665127277374268, + "learning_rate": 0.00019637003573521177, + "loss": 0.8569, + "step": 3679 + }, + { + "epoch": 0.08626499528238307, + "grad_norm": 4.864529609680176, + "learning_rate": 0.00019636806881140676, + "loss": 0.6388, + "step": 3680 + }, + { + "epoch": 0.08628843685718807, + "grad_norm": 2.077289342880249, + "learning_rate": 0.0001963661013647059, + "loss": 0.3908, + "step": 3681 + }, + { + "epoch": 0.08631187843199306, + "grad_norm": 4.139244556427002, + "learning_rate": 0.00019636413339511983, + "loss": 0.6601, + "step": 3682 + }, + { + "epoch": 0.08633532000679806, + "grad_norm": 12.4925537109375, + "learning_rate": 0.0001963621649026593, + "loss": 0.9128, + "step": 3683 + }, + { + "epoch": 0.08635876158160305, + "grad_norm": 0.6705361008644104, + "learning_rate": 0.00019636019588733491, + "loss": 0.1171, + "step": 3684 + }, + { + "epoch": 0.08638220315640804, + "grad_norm": 5.538782596588135, + "learning_rate": 0.00019635822634915743, + "loss": 0.6326, + "step": 3685 + }, + { + "epoch": 0.08640564473121304, + "grad_norm": 4.387018203735352, + "learning_rate": 0.00019635625628813748, + "loss": 0.6032, + "step": 3686 + }, + { + "epoch": 0.08642908630601803, + "grad_norm": 2.382192611694336, + "learning_rate": 0.0001963542857042858, + "loss": 1.7803, + "step": 3687 + }, + { + "epoch": 0.08645252788082303, + "grad_norm": 4.882470607757568, + "learning_rate": 0.00019635231459761302, + "loss": 0.456, + "step": 3688 + }, + { + "epoch": 0.08647596945562804, + "grad_norm": 1.950325846672058, + "learning_rate": 0.00019635034296812992, + "loss": 0.179, + "step": 3689 + }, + { + "epoch": 0.08649941103043303, + "grad_norm": 3.1761882305145264, + "learning_rate": 0.00019634837081584713, + "loss": 0.3517, + "step": 3690 + }, + { + "epoch": 0.08652285260523802, + "grad_norm": 4.215000152587891, + "learning_rate": 0.00019634639814077533, + "loss": 0.6843, + "step": 3691 + }, + { + "epoch": 0.08654629418004302, + "grad_norm": 0.9534603357315063, + "learning_rate": 0.00019634442494292532, + "loss": 0.1208, + "step": 3692 + }, + { + "epoch": 0.08656973575484801, + "grad_norm": 1.5382027626037598, + "learning_rate": 0.0001963424512223077, + "loss": 0.2935, + "step": 3693 + }, + { + "epoch": 0.086593177329653, + "grad_norm": 2.4389452934265137, + "learning_rate": 0.0001963404769789333, + "loss": 0.2057, + "step": 3694 + }, + { + "epoch": 0.086616618904458, + "grad_norm": 4.495862007141113, + "learning_rate": 0.0001963385022128127, + "loss": 0.6883, + "step": 3695 + }, + { + "epoch": 0.086640060479263, + "grad_norm": 7.789395809173584, + "learning_rate": 0.00019633652692395672, + "loss": 0.9737, + "step": 3696 + }, + { + "epoch": 0.08666350205406799, + "grad_norm": 0.4011204242706299, + "learning_rate": 0.00019633455111237603, + "loss": 0.0376, + "step": 3697 + }, + { + "epoch": 0.08668694362887298, + "grad_norm": 1.8527859449386597, + "learning_rate": 0.00019633257477808132, + "loss": 0.1716, + "step": 3698 + }, + { + "epoch": 0.08671038520367798, + "grad_norm": 3.059368848800659, + "learning_rate": 0.0001963305979210834, + "loss": 0.4959, + "step": 3699 + }, + { + "epoch": 0.08673382677848297, + "grad_norm": 0.951180100440979, + "learning_rate": 0.0001963286205413929, + "loss": 0.1184, + "step": 3700 + }, + { + "epoch": 0.08675726835328798, + "grad_norm": 4.0666823387146, + "learning_rate": 0.00019632664263902064, + "loss": 0.5431, + "step": 3701 + }, + { + "epoch": 0.08678070992809298, + "grad_norm": 1.952014446258545, + "learning_rate": 0.00019632466421397728, + "loss": 1.542, + "step": 3702 + }, + { + "epoch": 0.08680415150289797, + "grad_norm": 1.351762294769287, + "learning_rate": 0.00019632268526627357, + "loss": 0.2286, + "step": 3703 + }, + { + "epoch": 0.08682759307770296, + "grad_norm": 2.8934285640716553, + "learning_rate": 0.0001963207057959203, + "loss": 1.7657, + "step": 3704 + }, + { + "epoch": 0.08685103465250796, + "grad_norm": 4.030712127685547, + "learning_rate": 0.00019631872580292816, + "loss": 0.5991, + "step": 3705 + }, + { + "epoch": 0.08687447622731295, + "grad_norm": 5.7528605461120605, + "learning_rate": 0.0001963167452873079, + "loss": 0.8096, + "step": 3706 + }, + { + "epoch": 0.08689791780211795, + "grad_norm": 2.1362979412078857, + "learning_rate": 0.00019631476424907028, + "loss": 0.3102, + "step": 3707 + }, + { + "epoch": 0.08692135937692294, + "grad_norm": 3.3964955806732178, + "learning_rate": 0.00019631278268822603, + "loss": 0.3176, + "step": 3708 + }, + { + "epoch": 0.08694480095172794, + "grad_norm": 3.6647989749908447, + "learning_rate": 0.0001963108006047859, + "loss": 0.419, + "step": 3709 + }, + { + "epoch": 0.08696824252653293, + "grad_norm": 1.5147664546966553, + "learning_rate": 0.00019630881799876068, + "loss": 0.1917, + "step": 3710 + }, + { + "epoch": 0.08699168410133792, + "grad_norm": 3.9780938625335693, + "learning_rate": 0.0001963068348701611, + "loss": 0.6039, + "step": 3711 + }, + { + "epoch": 0.08701512567614292, + "grad_norm": 3.190290927886963, + "learning_rate": 0.00019630485121899794, + "loss": 0.3349, + "step": 3712 + }, + { + "epoch": 0.08703856725094791, + "grad_norm": 4.050565719604492, + "learning_rate": 0.0001963028670452819, + "loss": 0.8281, + "step": 3713 + }, + { + "epoch": 0.08706200882575292, + "grad_norm": 2.3831565380096436, + "learning_rate": 0.00019630088234902387, + "loss": 0.3018, + "step": 3714 + }, + { + "epoch": 0.08708545040055791, + "grad_norm": 4.4078779220581055, + "learning_rate": 0.00019629889713023452, + "loss": 0.6443, + "step": 3715 + }, + { + "epoch": 0.08710889197536291, + "grad_norm": 4.375972747802734, + "learning_rate": 0.0001962969113889246, + "loss": 0.8599, + "step": 3716 + }, + { + "epoch": 0.0871323335501679, + "grad_norm": 1.4196285009384155, + "learning_rate": 0.00019629492512510496, + "loss": 0.8482, + "step": 3717 + }, + { + "epoch": 0.0871557751249729, + "grad_norm": 3.4396300315856934, + "learning_rate": 0.00019629293833878637, + "loss": 0.5775, + "step": 3718 + }, + { + "epoch": 0.08717921669977789, + "grad_norm": 3.259840488433838, + "learning_rate": 0.00019629095102997957, + "loss": 0.3774, + "step": 3719 + }, + { + "epoch": 0.08720265827458289, + "grad_norm": 0.4538835287094116, + "learning_rate": 0.00019628896319869537, + "loss": 0.078, + "step": 3720 + }, + { + "epoch": 0.08722609984938788, + "grad_norm": 1.1720930337905884, + "learning_rate": 0.0001962869748449445, + "loss": 0.1378, + "step": 3721 + }, + { + "epoch": 0.08724954142419288, + "grad_norm": 2.589529275894165, + "learning_rate": 0.00019628498596873785, + "loss": 0.2797, + "step": 3722 + }, + { + "epoch": 0.08727298299899787, + "grad_norm": 2.6859078407287598, + "learning_rate": 0.00019628299657008618, + "loss": 1.828, + "step": 3723 + }, + { + "epoch": 0.08729642457380286, + "grad_norm": 2.85537052154541, + "learning_rate": 0.00019628100664900023, + "loss": 0.498, + "step": 3724 + }, + { + "epoch": 0.08731986614860786, + "grad_norm": 1.2802730798721313, + "learning_rate": 0.00019627901620549083, + "loss": 0.1757, + "step": 3725 + }, + { + "epoch": 0.08734330772341285, + "grad_norm": 2.166198492050171, + "learning_rate": 0.00019627702523956878, + "loss": 1.7108, + "step": 3726 + }, + { + "epoch": 0.08736674929821786, + "grad_norm": 5.555452823638916, + "learning_rate": 0.00019627503375124487, + "loss": 0.6139, + "step": 3727 + }, + { + "epoch": 0.08739019087302285, + "grad_norm": 5.319886207580566, + "learning_rate": 0.00019627304174052993, + "loss": 1.0542, + "step": 3728 + }, + { + "epoch": 0.08741363244782785, + "grad_norm": 2.982062339782715, + "learning_rate": 0.0001962710492074348, + "loss": 0.5472, + "step": 3729 + }, + { + "epoch": 0.08743707402263284, + "grad_norm": 3.1944174766540527, + "learning_rate": 0.0001962690561519702, + "loss": 0.6571, + "step": 3730 + }, + { + "epoch": 0.08746051559743784, + "grad_norm": 2.4879744052886963, + "learning_rate": 0.000196267062574147, + "loss": 1.47, + "step": 3731 + }, + { + "epoch": 0.08748395717224283, + "grad_norm": 5.887535572052002, + "learning_rate": 0.00019626506847397602, + "loss": 0.4142, + "step": 3732 + }, + { + "epoch": 0.08750739874704783, + "grad_norm": 3.894951820373535, + "learning_rate": 0.0001962630738514681, + "loss": 0.6309, + "step": 3733 + }, + { + "epoch": 0.08753084032185282, + "grad_norm": 3.262197256088257, + "learning_rate": 0.000196261078706634, + "loss": 0.7606, + "step": 3734 + }, + { + "epoch": 0.08755428189665782, + "grad_norm": 3.962759494781494, + "learning_rate": 0.00019625908303948458, + "loss": 0.4701, + "step": 3735 + }, + { + "epoch": 0.08757772347146281, + "grad_norm": 0.9542123079299927, + "learning_rate": 0.00019625708685003067, + "loss": 0.1095, + "step": 3736 + }, + { + "epoch": 0.0876011650462678, + "grad_norm": 4.098684310913086, + "learning_rate": 0.00019625509013828312, + "loss": 0.5676, + "step": 3737 + }, + { + "epoch": 0.0876246066210728, + "grad_norm": 3.5449371337890625, + "learning_rate": 0.00019625309290425272, + "loss": 0.6424, + "step": 3738 + }, + { + "epoch": 0.08764804819587779, + "grad_norm": 3.8979241847991943, + "learning_rate": 0.00019625109514795035, + "loss": 0.5808, + "step": 3739 + }, + { + "epoch": 0.0876714897706828, + "grad_norm": 4.180144786834717, + "learning_rate": 0.00019624909686938682, + "loss": 0.2654, + "step": 3740 + }, + { + "epoch": 0.0876949313454878, + "grad_norm": 5.477909088134766, + "learning_rate": 0.00019624709806857296, + "loss": 1.3072, + "step": 3741 + }, + { + "epoch": 0.08771837292029279, + "grad_norm": 1.9389069080352783, + "learning_rate": 0.00019624509874551968, + "loss": 0.2578, + "step": 3742 + }, + { + "epoch": 0.08774181449509778, + "grad_norm": 0.635269045829773, + "learning_rate": 0.0001962430989002378, + "loss": 0.071, + "step": 3743 + }, + { + "epoch": 0.08776525606990278, + "grad_norm": 4.165474891662598, + "learning_rate": 0.00019624109853273814, + "loss": 0.4799, + "step": 3744 + }, + { + "epoch": 0.08778869764470777, + "grad_norm": 2.710177421569824, + "learning_rate": 0.00019623909764303156, + "loss": 0.4664, + "step": 3745 + }, + { + "epoch": 0.08781213921951277, + "grad_norm": 4.324915409088135, + "learning_rate": 0.00019623709623112893, + "loss": 0.6979, + "step": 3746 + }, + { + "epoch": 0.08783558079431776, + "grad_norm": 4.394331932067871, + "learning_rate": 0.00019623509429704116, + "loss": 0.4805, + "step": 3747 + }, + { + "epoch": 0.08785902236912276, + "grad_norm": 0.5834438800811768, + "learning_rate": 0.00019623309184077902, + "loss": 0.0797, + "step": 3748 + }, + { + "epoch": 0.08788246394392775, + "grad_norm": 0.8107856512069702, + "learning_rate": 0.0001962310888623534, + "loss": 0.0956, + "step": 3749 + }, + { + "epoch": 0.08790590551873274, + "grad_norm": 4.048373699188232, + "learning_rate": 0.0001962290853617752, + "loss": 0.5716, + "step": 3750 + }, + { + "epoch": 0.08792934709353774, + "grad_norm": 3.711843252182007, + "learning_rate": 0.00019622708133905527, + "loss": 0.4383, + "step": 3751 + }, + { + "epoch": 0.08795278866834273, + "grad_norm": 2.013049364089966, + "learning_rate": 0.00019622507679420453, + "loss": 0.228, + "step": 3752 + }, + { + "epoch": 0.08797623024314774, + "grad_norm": 3.2748029232025146, + "learning_rate": 0.00019622307172723378, + "loss": 0.2307, + "step": 3753 + }, + { + "epoch": 0.08799967181795273, + "grad_norm": 3.6803553104400635, + "learning_rate": 0.00019622106613815393, + "loss": 0.4601, + "step": 3754 + }, + { + "epoch": 0.08802311339275773, + "grad_norm": 3.3987021446228027, + "learning_rate": 0.0001962190600269759, + "loss": 0.619, + "step": 3755 + }, + { + "epoch": 0.08804655496756272, + "grad_norm": 4.623305320739746, + "learning_rate": 0.00019621705339371054, + "loss": 0.7178, + "step": 3756 + }, + { + "epoch": 0.08806999654236772, + "grad_norm": 2.490144968032837, + "learning_rate": 0.00019621504623836873, + "loss": 0.3101, + "step": 3757 + }, + { + "epoch": 0.08809343811717271, + "grad_norm": 3.5187721252441406, + "learning_rate": 0.00019621303856096138, + "loss": 0.444, + "step": 3758 + }, + { + "epoch": 0.0881168796919777, + "grad_norm": 0.9938526153564453, + "learning_rate": 0.00019621103036149938, + "loss": 0.1266, + "step": 3759 + }, + { + "epoch": 0.0881403212667827, + "grad_norm": 2.7949414253234863, + "learning_rate": 0.00019620902163999362, + "loss": 1.7501, + "step": 3760 + }, + { + "epoch": 0.0881637628415877, + "grad_norm": 4.236056804656982, + "learning_rate": 0.000196207012396455, + "loss": 0.4308, + "step": 3761 + }, + { + "epoch": 0.08818720441639269, + "grad_norm": 1.7602834701538086, + "learning_rate": 0.00019620500263089444, + "loss": 0.286, + "step": 3762 + }, + { + "epoch": 0.08821064599119768, + "grad_norm": 3.045759677886963, + "learning_rate": 0.0001962029923433228, + "loss": 0.3251, + "step": 3763 + }, + { + "epoch": 0.08823408756600268, + "grad_norm": 2.508775472640991, + "learning_rate": 0.00019620098153375106, + "loss": 0.451, + "step": 3764 + }, + { + "epoch": 0.08825752914080769, + "grad_norm": 4.346813201904297, + "learning_rate": 0.00019619897020219007, + "loss": 0.6597, + "step": 3765 + }, + { + "epoch": 0.08828097071561268, + "grad_norm": 2.4111833572387695, + "learning_rate": 0.00019619695834865075, + "loss": 0.2558, + "step": 3766 + }, + { + "epoch": 0.08830441229041767, + "grad_norm": 5.455415725708008, + "learning_rate": 0.00019619494597314404, + "loss": 0.5017, + "step": 3767 + }, + { + "epoch": 0.08832785386522267, + "grad_norm": 3.577070951461792, + "learning_rate": 0.00019619293307568086, + "loss": 1.4664, + "step": 3768 + }, + { + "epoch": 0.08835129544002766, + "grad_norm": 4.10477876663208, + "learning_rate": 0.0001961909196562721, + "loss": 0.3783, + "step": 3769 + }, + { + "epoch": 0.08837473701483266, + "grad_norm": 0.9390559196472168, + "learning_rate": 0.00019618890571492873, + "loss": 0.1148, + "step": 3770 + }, + { + "epoch": 0.08839817858963765, + "grad_norm": 2.4144577980041504, + "learning_rate": 0.00019618689125166163, + "loss": 0.3356, + "step": 3771 + }, + { + "epoch": 0.08842162016444265, + "grad_norm": 4.509500980377197, + "learning_rate": 0.00019618487626648176, + "loss": 0.5578, + "step": 3772 + }, + { + "epoch": 0.08844506173924764, + "grad_norm": 2.329207181930542, + "learning_rate": 0.00019618286075940005, + "loss": 0.3808, + "step": 3773 + }, + { + "epoch": 0.08846850331405264, + "grad_norm": 5.464447975158691, + "learning_rate": 0.0001961808447304274, + "loss": 0.4833, + "step": 3774 + }, + { + "epoch": 0.08849194488885763, + "grad_norm": 1.152366280555725, + "learning_rate": 0.00019617882817957484, + "loss": 0.137, + "step": 3775 + }, + { + "epoch": 0.08851538646366262, + "grad_norm": 1.0086610317230225, + "learning_rate": 0.00019617681110685323, + "loss": 0.1058, + "step": 3776 + }, + { + "epoch": 0.08853882803846762, + "grad_norm": 3.4383583068847656, + "learning_rate": 0.00019617479351227352, + "loss": 0.4146, + "step": 3777 + }, + { + "epoch": 0.08856226961327263, + "grad_norm": 3.1780121326446533, + "learning_rate": 0.0001961727753958467, + "loss": 0.6251, + "step": 3778 + }, + { + "epoch": 0.08858571118807762, + "grad_norm": 1.7934889793395996, + "learning_rate": 0.0001961707567575837, + "loss": 0.2439, + "step": 3779 + }, + { + "epoch": 0.08860915276288261, + "grad_norm": 2.4896624088287354, + "learning_rate": 0.00019616873759749546, + "loss": 0.1754, + "step": 3780 + }, + { + "epoch": 0.08863259433768761, + "grad_norm": 4.397201061248779, + "learning_rate": 0.00019616671791559295, + "loss": 0.5677, + "step": 3781 + }, + { + "epoch": 0.0886560359124926, + "grad_norm": 0.8087470531463623, + "learning_rate": 0.00019616469771188713, + "loss": 0.0958, + "step": 3782 + }, + { + "epoch": 0.0886794774872976, + "grad_norm": 4.803033351898193, + "learning_rate": 0.00019616267698638893, + "loss": 0.6452, + "step": 3783 + }, + { + "epoch": 0.08870291906210259, + "grad_norm": 0.42540043592453003, + "learning_rate": 0.00019616065573910935, + "loss": 0.0554, + "step": 3784 + }, + { + "epoch": 0.08872636063690759, + "grad_norm": 2.845632791519165, + "learning_rate": 0.00019615863397005935, + "loss": 0.2977, + "step": 3785 + }, + { + "epoch": 0.08874980221171258, + "grad_norm": 0.6929962635040283, + "learning_rate": 0.0001961566116792499, + "loss": 0.1226, + "step": 3786 + }, + { + "epoch": 0.08877324378651757, + "grad_norm": 3.421199321746826, + "learning_rate": 0.000196154588866692, + "loss": 0.6047, + "step": 3787 + }, + { + "epoch": 0.08879668536132257, + "grad_norm": 5.012919902801514, + "learning_rate": 0.00019615256553239655, + "loss": 0.7467, + "step": 3788 + }, + { + "epoch": 0.08882012693612756, + "grad_norm": 1.2533382177352905, + "learning_rate": 0.00019615054167637462, + "loss": 0.2647, + "step": 3789 + }, + { + "epoch": 0.08884356851093256, + "grad_norm": 0.9880827069282532, + "learning_rate": 0.0001961485172986371, + "loss": 0.1619, + "step": 3790 + }, + { + "epoch": 0.08886701008573757, + "grad_norm": 1.8032840490341187, + "learning_rate": 0.00019614649239919506, + "loss": 0.2666, + "step": 3791 + }, + { + "epoch": 0.08889045166054256, + "grad_norm": 3.2008111476898193, + "learning_rate": 0.00019614446697805945, + "loss": 0.701, + "step": 3792 + }, + { + "epoch": 0.08891389323534755, + "grad_norm": 1.0579092502593994, + "learning_rate": 0.0001961424410352412, + "loss": 0.1747, + "step": 3793 + }, + { + "epoch": 0.08893733481015255, + "grad_norm": 2.338984727859497, + "learning_rate": 0.00019614041457075143, + "loss": 2.0147, + "step": 3794 + }, + { + "epoch": 0.08896077638495754, + "grad_norm": 2.3609135150909424, + "learning_rate": 0.00019613838758460102, + "loss": 0.4213, + "step": 3795 + }, + { + "epoch": 0.08898421795976254, + "grad_norm": 0.9913028478622437, + "learning_rate": 0.00019613636007680105, + "loss": 0.1477, + "step": 3796 + }, + { + "epoch": 0.08900765953456753, + "grad_norm": 3.4566762447357178, + "learning_rate": 0.0001961343320473625, + "loss": 0.5013, + "step": 3797 + }, + { + "epoch": 0.08903110110937253, + "grad_norm": 2.756699562072754, + "learning_rate": 0.00019613230349629633, + "loss": 0.5083, + "step": 3798 + }, + { + "epoch": 0.08905454268417752, + "grad_norm": 2.882474660873413, + "learning_rate": 0.0001961302744236136, + "loss": 0.5069, + "step": 3799 + }, + { + "epoch": 0.08907798425898251, + "grad_norm": 3.0164544582366943, + "learning_rate": 0.00019612824482932525, + "loss": 0.3169, + "step": 3800 + }, + { + "epoch": 0.08910142583378751, + "grad_norm": 2.8138742446899414, + "learning_rate": 0.0001961262147134424, + "loss": 0.4143, + "step": 3801 + }, + { + "epoch": 0.0891248674085925, + "grad_norm": 4.751468181610107, + "learning_rate": 0.00019612418407597595, + "loss": 0.8669, + "step": 3802 + }, + { + "epoch": 0.0891483089833975, + "grad_norm": 0.6620890498161316, + "learning_rate": 0.00019612215291693702, + "loss": 0.0937, + "step": 3803 + }, + { + "epoch": 0.0891717505582025, + "grad_norm": 3.0224149227142334, + "learning_rate": 0.00019612012123633658, + "loss": 0.4438, + "step": 3804 + }, + { + "epoch": 0.0891951921330075, + "grad_norm": 1.5049916505813599, + "learning_rate": 0.00019611808903418566, + "loss": 1.019, + "step": 3805 + }, + { + "epoch": 0.0892186337078125, + "grad_norm": 5.031360149383545, + "learning_rate": 0.00019611605631049525, + "loss": 0.8014, + "step": 3806 + }, + { + "epoch": 0.08924207528261749, + "grad_norm": 4.372645378112793, + "learning_rate": 0.00019611402306527644, + "loss": 0.6527, + "step": 3807 + }, + { + "epoch": 0.08926551685742248, + "grad_norm": 8.646684646606445, + "learning_rate": 0.00019611198929854026, + "loss": 0.7985, + "step": 3808 + }, + { + "epoch": 0.08928895843222748, + "grad_norm": 5.3927717208862305, + "learning_rate": 0.0001961099550102977, + "loss": 0.8114, + "step": 3809 + }, + { + "epoch": 0.08931240000703247, + "grad_norm": 3.7090249061584473, + "learning_rate": 0.00019610792020055982, + "loss": 0.6037, + "step": 3810 + }, + { + "epoch": 0.08933584158183747, + "grad_norm": 3.168682336807251, + "learning_rate": 0.00019610588486933765, + "loss": 0.5464, + "step": 3811 + }, + { + "epoch": 0.08935928315664246, + "grad_norm": 4.365353107452393, + "learning_rate": 0.00019610384901664227, + "loss": 0.5478, + "step": 3812 + }, + { + "epoch": 0.08938272473144745, + "grad_norm": 2.8700921535491943, + "learning_rate": 0.0001961018126424847, + "loss": 0.4112, + "step": 3813 + }, + { + "epoch": 0.08940616630625245, + "grad_norm": 2.8572278022766113, + "learning_rate": 0.00019609977574687598, + "loss": 0.2947, + "step": 3814 + }, + { + "epoch": 0.08942960788105744, + "grad_norm": 1.8562170267105103, + "learning_rate": 0.0001960977383298272, + "loss": 0.3117, + "step": 3815 + }, + { + "epoch": 0.08945304945586245, + "grad_norm": 0.8748646974563599, + "learning_rate": 0.00019609570039134938, + "loss": 0.1269, + "step": 3816 + }, + { + "epoch": 0.08947649103066745, + "grad_norm": 3.1602017879486084, + "learning_rate": 0.00019609366193145357, + "loss": 0.4728, + "step": 3817 + }, + { + "epoch": 0.08949993260547244, + "grad_norm": 5.528581619262695, + "learning_rate": 0.00019609162295015087, + "loss": 0.5548, + "step": 3818 + }, + { + "epoch": 0.08952337418027743, + "grad_norm": 3.4717965126037598, + "learning_rate": 0.0001960895834474523, + "loss": 0.4529, + "step": 3819 + }, + { + "epoch": 0.08954681575508243, + "grad_norm": 4.143022060394287, + "learning_rate": 0.00019608754342336897, + "loss": 0.5535, + "step": 3820 + }, + { + "epoch": 0.08957025732988742, + "grad_norm": 1.5248193740844727, + "learning_rate": 0.00019608550287791192, + "loss": 0.8717, + "step": 3821 + }, + { + "epoch": 0.08959369890469242, + "grad_norm": 2.0727131366729736, + "learning_rate": 0.00019608346181109222, + "loss": 0.213, + "step": 3822 + }, + { + "epoch": 0.08961714047949741, + "grad_norm": 5.8353590965271, + "learning_rate": 0.00019608142022292096, + "loss": 0.9255, + "step": 3823 + }, + { + "epoch": 0.0896405820543024, + "grad_norm": 3.476886749267578, + "learning_rate": 0.00019607937811340925, + "loss": 0.4293, + "step": 3824 + }, + { + "epoch": 0.0896640236291074, + "grad_norm": 3.386876344680786, + "learning_rate": 0.00019607733548256806, + "loss": 2.1494, + "step": 3825 + }, + { + "epoch": 0.0896874652039124, + "grad_norm": 5.637668132781982, + "learning_rate": 0.0001960752923304086, + "loss": 0.7143, + "step": 3826 + }, + { + "epoch": 0.08971090677871739, + "grad_norm": 3.0115318298339844, + "learning_rate": 0.00019607324865694188, + "loss": 1.493, + "step": 3827 + }, + { + "epoch": 0.08973434835352238, + "grad_norm": 3.5252692699432373, + "learning_rate": 0.00019607120446217904, + "loss": 0.383, + "step": 3828 + }, + { + "epoch": 0.08975778992832739, + "grad_norm": 4.732570171356201, + "learning_rate": 0.0001960691597461311, + "loss": 0.5933, + "step": 3829 + }, + { + "epoch": 0.08978123150313239, + "grad_norm": 2.3647916316986084, + "learning_rate": 0.00019606711450880925, + "loss": 1.9678, + "step": 3830 + }, + { + "epoch": 0.08980467307793738, + "grad_norm": 1.0566022396087646, + "learning_rate": 0.00019606506875022447, + "loss": 0.1992, + "step": 3831 + }, + { + "epoch": 0.08982811465274237, + "grad_norm": 3.6253790855407715, + "learning_rate": 0.00019606302247038797, + "loss": 0.1759, + "step": 3832 + }, + { + "epoch": 0.08985155622754737, + "grad_norm": 3.081727981567383, + "learning_rate": 0.00019606097566931082, + "loss": 0.4667, + "step": 3833 + }, + { + "epoch": 0.08987499780235236, + "grad_norm": 3.4823110103607178, + "learning_rate": 0.00019605892834700407, + "loss": 0.5407, + "step": 3834 + }, + { + "epoch": 0.08989843937715736, + "grad_norm": 0.6158022284507751, + "learning_rate": 0.0001960568805034789, + "loss": 0.0656, + "step": 3835 + }, + { + "epoch": 0.08992188095196235, + "grad_norm": 4.056995868682861, + "learning_rate": 0.0001960548321387464, + "loss": 0.5164, + "step": 3836 + }, + { + "epoch": 0.08994532252676735, + "grad_norm": 4.896787166595459, + "learning_rate": 0.00019605278325281767, + "loss": 0.909, + "step": 3837 + }, + { + "epoch": 0.08996876410157234, + "grad_norm": 3.619252920150757, + "learning_rate": 0.00019605073384570384, + "loss": 0.5295, + "step": 3838 + }, + { + "epoch": 0.08999220567637733, + "grad_norm": 3.577321767807007, + "learning_rate": 0.00019604868391741604, + "loss": 0.3833, + "step": 3839 + }, + { + "epoch": 0.09001564725118233, + "grad_norm": 4.629184246063232, + "learning_rate": 0.00019604663346796538, + "loss": 0.6893, + "step": 3840 + }, + { + "epoch": 0.09003908882598732, + "grad_norm": 3.81138014793396, + "learning_rate": 0.00019604458249736296, + "loss": 0.6854, + "step": 3841 + }, + { + "epoch": 0.09006253040079233, + "grad_norm": 3.272611618041992, + "learning_rate": 0.00019604253100561993, + "loss": 0.4, + "step": 3842 + }, + { + "epoch": 0.09008597197559733, + "grad_norm": 2.6452724933624268, + "learning_rate": 0.00019604047899274746, + "loss": 0.4278, + "step": 3843 + }, + { + "epoch": 0.09010941355040232, + "grad_norm": 3.1878931522369385, + "learning_rate": 0.00019603842645875665, + "loss": 0.461, + "step": 3844 + }, + { + "epoch": 0.09013285512520731, + "grad_norm": 3.192089557647705, + "learning_rate": 0.00019603637340365862, + "loss": 0.3943, + "step": 3845 + }, + { + "epoch": 0.09015629670001231, + "grad_norm": 2.131906509399414, + "learning_rate": 0.00019603431982746452, + "loss": 1.3588, + "step": 3846 + }, + { + "epoch": 0.0901797382748173, + "grad_norm": 0.5362159609794617, + "learning_rate": 0.00019603226573018553, + "loss": 0.0777, + "step": 3847 + }, + { + "epoch": 0.0902031798496223, + "grad_norm": 2.2967171669006348, + "learning_rate": 0.00019603021111183275, + "loss": 1.0961, + "step": 3848 + }, + { + "epoch": 0.09022662142442729, + "grad_norm": 4.80722713470459, + "learning_rate": 0.00019602815597241736, + "loss": 0.6672, + "step": 3849 + }, + { + "epoch": 0.09025006299923229, + "grad_norm": 3.2452456951141357, + "learning_rate": 0.00019602610031195045, + "loss": 0.4428, + "step": 3850 + }, + { + "epoch": 0.09027350457403728, + "grad_norm": 3.0578958988189697, + "learning_rate": 0.00019602404413044324, + "loss": 0.3702, + "step": 3851 + }, + { + "epoch": 0.09029694614884227, + "grad_norm": 1.9458588361740112, + "learning_rate": 0.00019602198742790688, + "loss": 1.4988, + "step": 3852 + }, + { + "epoch": 0.09032038772364727, + "grad_norm": 4.665829658508301, + "learning_rate": 0.00019601993020435253, + "loss": 0.572, + "step": 3853 + }, + { + "epoch": 0.09034382929845226, + "grad_norm": 2.6013646125793457, + "learning_rate": 0.0001960178724597913, + "loss": 1.8374, + "step": 3854 + }, + { + "epoch": 0.09036727087325727, + "grad_norm": 3.5464940071105957, + "learning_rate": 0.00019601581419423443, + "loss": 0.456, + "step": 3855 + }, + { + "epoch": 0.09039071244806227, + "grad_norm": 2.364607572555542, + "learning_rate": 0.00019601375540769303, + "loss": 1.6431, + "step": 3856 + }, + { + "epoch": 0.09041415402286726, + "grad_norm": 3.364642381668091, + "learning_rate": 0.00019601169610017832, + "loss": 0.4171, + "step": 3857 + }, + { + "epoch": 0.09043759559767225, + "grad_norm": 4.273921966552734, + "learning_rate": 0.00019600963627170142, + "loss": 0.5234, + "step": 3858 + }, + { + "epoch": 0.09046103717247725, + "grad_norm": 1.3532145023345947, + "learning_rate": 0.00019600757592227352, + "loss": 0.25, + "step": 3859 + }, + { + "epoch": 0.09048447874728224, + "grad_norm": 3.5296456813812256, + "learning_rate": 0.00019600551505190585, + "loss": 0.52, + "step": 3860 + }, + { + "epoch": 0.09050792032208724, + "grad_norm": 5.297824382781982, + "learning_rate": 0.00019600345366060955, + "loss": 0.6799, + "step": 3861 + }, + { + "epoch": 0.09053136189689223, + "grad_norm": 4.252755165100098, + "learning_rate": 0.00019600139174839578, + "loss": 0.6232, + "step": 3862 + }, + { + "epoch": 0.09055480347169723, + "grad_norm": 4.930519104003906, + "learning_rate": 0.0001959993293152758, + "loss": 0.7896, + "step": 3863 + }, + { + "epoch": 0.09057824504650222, + "grad_norm": 3.3964829444885254, + "learning_rate": 0.0001959972663612607, + "loss": 0.5905, + "step": 3864 + }, + { + "epoch": 0.09060168662130721, + "grad_norm": 3.4283177852630615, + "learning_rate": 0.00019599520288636181, + "loss": 0.6008, + "step": 3865 + }, + { + "epoch": 0.09062512819611221, + "grad_norm": 3.1634628772735596, + "learning_rate": 0.0001959931388905902, + "loss": 0.5287, + "step": 3866 + }, + { + "epoch": 0.0906485697709172, + "grad_norm": 2.850149393081665, + "learning_rate": 0.00019599107437395715, + "loss": 0.3921, + "step": 3867 + }, + { + "epoch": 0.09067201134572221, + "grad_norm": 2.5310475826263428, + "learning_rate": 0.00019598900933647382, + "loss": 0.2547, + "step": 3868 + }, + { + "epoch": 0.0906954529205272, + "grad_norm": 4.030781269073486, + "learning_rate": 0.0001959869437781514, + "loss": 0.6981, + "step": 3869 + }, + { + "epoch": 0.0907188944953322, + "grad_norm": 3.3061678409576416, + "learning_rate": 0.00019598487769900116, + "loss": 0.5525, + "step": 3870 + }, + { + "epoch": 0.0907423360701372, + "grad_norm": 5.612123489379883, + "learning_rate": 0.00019598281109903427, + "loss": 0.9021, + "step": 3871 + }, + { + "epoch": 0.09076577764494219, + "grad_norm": 0.7047792077064514, + "learning_rate": 0.00019598074397826195, + "loss": 0.0927, + "step": 3872 + }, + { + "epoch": 0.09078921921974718, + "grad_norm": 2.571810722351074, + "learning_rate": 0.0001959786763366954, + "loss": 0.5072, + "step": 3873 + }, + { + "epoch": 0.09081266079455218, + "grad_norm": 3.674124240875244, + "learning_rate": 0.00019597660817434584, + "loss": 0.4622, + "step": 3874 + }, + { + "epoch": 0.09083610236935717, + "grad_norm": 1.9121911525726318, + "learning_rate": 0.00019597453949122453, + "loss": 0.269, + "step": 3875 + }, + { + "epoch": 0.09085954394416217, + "grad_norm": 2.407320499420166, + "learning_rate": 0.00019597247028734267, + "loss": 0.425, + "step": 3876 + }, + { + "epoch": 0.09088298551896716, + "grad_norm": 1.3616907596588135, + "learning_rate": 0.00019597040056271148, + "loss": 0.1882, + "step": 3877 + }, + { + "epoch": 0.09090642709377215, + "grad_norm": 3.4609997272491455, + "learning_rate": 0.0001959683303173422, + "loss": 0.408, + "step": 3878 + }, + { + "epoch": 0.09092986866857715, + "grad_norm": 3.2543373107910156, + "learning_rate": 0.00019596625955124606, + "loss": 0.3729, + "step": 3879 + }, + { + "epoch": 0.09095331024338216, + "grad_norm": 3.8401741981506348, + "learning_rate": 0.0001959641882644343, + "loss": 0.6278, + "step": 3880 + }, + { + "epoch": 0.09097675181818715, + "grad_norm": 2.5666613578796387, + "learning_rate": 0.00019596211645691813, + "loss": 1.835, + "step": 3881 + }, + { + "epoch": 0.09100019339299215, + "grad_norm": 2.5879549980163574, + "learning_rate": 0.00019596004412870884, + "loss": 0.3579, + "step": 3882 + }, + { + "epoch": 0.09102363496779714, + "grad_norm": 4.043653964996338, + "learning_rate": 0.00019595797127981766, + "loss": 0.6771, + "step": 3883 + }, + { + "epoch": 0.09104707654260213, + "grad_norm": 2.6690917015075684, + "learning_rate": 0.00019595589791025582, + "loss": 1.6091, + "step": 3884 + }, + { + "epoch": 0.09107051811740713, + "grad_norm": 2.7753512859344482, + "learning_rate": 0.0001959538240200346, + "loss": 0.4941, + "step": 3885 + }, + { + "epoch": 0.09109395969221212, + "grad_norm": 2.4478302001953125, + "learning_rate": 0.0001959517496091652, + "loss": 0.2451, + "step": 3886 + }, + { + "epoch": 0.09111740126701712, + "grad_norm": 2.349473714828491, + "learning_rate": 0.00019594967467765893, + "loss": 0.1142, + "step": 3887 + }, + { + "epoch": 0.09114084284182211, + "grad_norm": 4.290885925292969, + "learning_rate": 0.00019594759922552703, + "loss": 0.7619, + "step": 3888 + }, + { + "epoch": 0.0911642844166271, + "grad_norm": 1.7319837808609009, + "learning_rate": 0.00019594552325278075, + "loss": 1.3699, + "step": 3889 + }, + { + "epoch": 0.0911877259914321, + "grad_norm": 3.609398365020752, + "learning_rate": 0.00019594344675943136, + "loss": 0.6094, + "step": 3890 + }, + { + "epoch": 0.0912111675662371, + "grad_norm": 0.4374404847621918, + "learning_rate": 0.00019594136974549013, + "loss": 0.1027, + "step": 3891 + }, + { + "epoch": 0.09123460914104209, + "grad_norm": 4.4581685066223145, + "learning_rate": 0.00019593929221096835, + "loss": 0.7333, + "step": 3892 + }, + { + "epoch": 0.0912580507158471, + "grad_norm": 2.1943793296813965, + "learning_rate": 0.00019593721415587725, + "loss": 1.8792, + "step": 3893 + }, + { + "epoch": 0.09128149229065209, + "grad_norm": 1.505106806755066, + "learning_rate": 0.00019593513558022814, + "loss": 0.2189, + "step": 3894 + }, + { + "epoch": 0.09130493386545709, + "grad_norm": 4.990261554718018, + "learning_rate": 0.00019593305648403226, + "loss": 0.9985, + "step": 3895 + }, + { + "epoch": 0.09132837544026208, + "grad_norm": 4.015891075134277, + "learning_rate": 0.00019593097686730095, + "loss": 0.3768, + "step": 3896 + }, + { + "epoch": 0.09135181701506707, + "grad_norm": 1.6192516088485718, + "learning_rate": 0.00019592889673004544, + "loss": 0.227, + "step": 3897 + }, + { + "epoch": 0.09137525858987207, + "grad_norm": 1.2656649351119995, + "learning_rate": 0.00019592681607227705, + "loss": 0.2292, + "step": 3898 + }, + { + "epoch": 0.09139870016467706, + "grad_norm": 3.553513288497925, + "learning_rate": 0.00019592473489400707, + "loss": 0.6723, + "step": 3899 + }, + { + "epoch": 0.09142214173948206, + "grad_norm": 4.026970386505127, + "learning_rate": 0.0001959226531952468, + "loss": 0.5429, + "step": 3900 + }, + { + "epoch": 0.09144558331428705, + "grad_norm": 3.449059247970581, + "learning_rate": 0.00019592057097600748, + "loss": 0.3761, + "step": 3901 + }, + { + "epoch": 0.09146902488909205, + "grad_norm": 1.66456937789917, + "learning_rate": 0.00019591848823630046, + "loss": 0.2498, + "step": 3902 + }, + { + "epoch": 0.09149246646389704, + "grad_norm": 1.2546275854110718, + "learning_rate": 0.00019591640497613702, + "loss": 0.1612, + "step": 3903 + }, + { + "epoch": 0.09151590803870203, + "grad_norm": 5.843565464019775, + "learning_rate": 0.00019591432119552845, + "loss": 0.6958, + "step": 3904 + }, + { + "epoch": 0.09153934961350703, + "grad_norm": 4.068614482879639, + "learning_rate": 0.00019591223689448611, + "loss": 0.7399, + "step": 3905 + }, + { + "epoch": 0.09156279118831204, + "grad_norm": 3.3525614738464355, + "learning_rate": 0.00019591015207302124, + "loss": 0.4794, + "step": 3906 + }, + { + "epoch": 0.09158623276311703, + "grad_norm": 2.921057939529419, + "learning_rate": 0.00019590806673114522, + "loss": 0.2611, + "step": 3907 + }, + { + "epoch": 0.09160967433792203, + "grad_norm": 3.0329325199127197, + "learning_rate": 0.00019590598086886932, + "loss": 0.6473, + "step": 3908 + }, + { + "epoch": 0.09163311591272702, + "grad_norm": 4.175178050994873, + "learning_rate": 0.00019590389448620487, + "loss": 0.6058, + "step": 3909 + }, + { + "epoch": 0.09165655748753201, + "grad_norm": 4.919735431671143, + "learning_rate": 0.00019590180758316322, + "loss": 0.4899, + "step": 3910 + }, + { + "epoch": 0.09167999906233701, + "grad_norm": 2.3598239421844482, + "learning_rate": 0.00019589972015975562, + "loss": 0.3463, + "step": 3911 + }, + { + "epoch": 0.091703440637142, + "grad_norm": 3.7039945125579834, + "learning_rate": 0.00019589763221599348, + "loss": 0.4799, + "step": 3912 + }, + { + "epoch": 0.091726882211947, + "grad_norm": 3.3117661476135254, + "learning_rate": 0.00019589554375188805, + "loss": 2.2636, + "step": 3913 + }, + { + "epoch": 0.09175032378675199, + "grad_norm": 3.1642541885375977, + "learning_rate": 0.00019589345476745074, + "loss": 0.5517, + "step": 3914 + }, + { + "epoch": 0.09177376536155699, + "grad_norm": 6.389397144317627, + "learning_rate": 0.00019589136526269283, + "loss": 0.5633, + "step": 3915 + }, + { + "epoch": 0.09179720693636198, + "grad_norm": 2.475024700164795, + "learning_rate": 0.00019588927523762567, + "loss": 1.7608, + "step": 3916 + }, + { + "epoch": 0.09182064851116697, + "grad_norm": 4.911717414855957, + "learning_rate": 0.00019588718469226062, + "loss": 0.4531, + "step": 3917 + }, + { + "epoch": 0.09184409008597197, + "grad_norm": 3.1031887531280518, + "learning_rate": 0.00019588509362660901, + "loss": 0.3436, + "step": 3918 + }, + { + "epoch": 0.09186753166077698, + "grad_norm": 4.80194616317749, + "learning_rate": 0.0001958830020406822, + "loss": 0.8178, + "step": 3919 + }, + { + "epoch": 0.09189097323558197, + "grad_norm": 1.8360826969146729, + "learning_rate": 0.0001958809099344915, + "loss": 0.1371, + "step": 3920 + }, + { + "epoch": 0.09191441481038697, + "grad_norm": 4.1721930503845215, + "learning_rate": 0.00019587881730804828, + "loss": 0.7327, + "step": 3921 + }, + { + "epoch": 0.09193785638519196, + "grad_norm": 1.4593827724456787, + "learning_rate": 0.00019587672416136392, + "loss": 0.2459, + "step": 3922 + }, + { + "epoch": 0.09196129795999695, + "grad_norm": 4.721362590789795, + "learning_rate": 0.00019587463049444977, + "loss": 0.5927, + "step": 3923 + }, + { + "epoch": 0.09198473953480195, + "grad_norm": 1.8055260181427002, + "learning_rate": 0.00019587253630731717, + "loss": 0.2981, + "step": 3924 + }, + { + "epoch": 0.09200818110960694, + "grad_norm": 4.356103897094727, + "learning_rate": 0.00019587044159997749, + "loss": 0.7727, + "step": 3925 + }, + { + "epoch": 0.09203162268441194, + "grad_norm": 4.472800254821777, + "learning_rate": 0.0001958683463724421, + "loss": 0.4642, + "step": 3926 + }, + { + "epoch": 0.09205506425921693, + "grad_norm": 0.6976740956306458, + "learning_rate": 0.00019586625062472234, + "loss": 0.1265, + "step": 3927 + }, + { + "epoch": 0.09207850583402193, + "grad_norm": 2.6839864253997803, + "learning_rate": 0.00019586415435682964, + "loss": 1.765, + "step": 3928 + }, + { + "epoch": 0.09210194740882692, + "grad_norm": 3.2114787101745605, + "learning_rate": 0.00019586205756877534, + "loss": 0.5094, + "step": 3929 + }, + { + "epoch": 0.09212538898363191, + "grad_norm": 3.5958023071289062, + "learning_rate": 0.00019585996026057082, + "loss": 0.5383, + "step": 3930 + }, + { + "epoch": 0.09214883055843692, + "grad_norm": 3.0062756538391113, + "learning_rate": 0.00019585786243222746, + "loss": 0.419, + "step": 3931 + }, + { + "epoch": 0.09217227213324192, + "grad_norm": 3.3114655017852783, + "learning_rate": 0.00019585576408375662, + "loss": 0.527, + "step": 3932 + }, + { + "epoch": 0.09219571370804691, + "grad_norm": 3.2749428749084473, + "learning_rate": 0.00019585366521516972, + "loss": 0.504, + "step": 3933 + }, + { + "epoch": 0.0922191552828519, + "grad_norm": 2.2021684646606445, + "learning_rate": 0.00019585156582647815, + "loss": 0.3209, + "step": 3934 + }, + { + "epoch": 0.0922425968576569, + "grad_norm": 0.9741812348365784, + "learning_rate": 0.00019584946591769326, + "loss": 0.121, + "step": 3935 + }, + { + "epoch": 0.0922660384324619, + "grad_norm": 0.4910936951637268, + "learning_rate": 0.0001958473654888265, + "loss": 0.0538, + "step": 3936 + }, + { + "epoch": 0.09228948000726689, + "grad_norm": 4.291139602661133, + "learning_rate": 0.00019584526453988925, + "loss": 0.5691, + "step": 3937 + }, + { + "epoch": 0.09231292158207188, + "grad_norm": 4.35619592666626, + "learning_rate": 0.00019584316307089288, + "loss": 0.5355, + "step": 3938 + }, + { + "epoch": 0.09233636315687688, + "grad_norm": 3.4402480125427246, + "learning_rate": 0.00019584106108184882, + "loss": 0.3155, + "step": 3939 + }, + { + "epoch": 0.09235980473168187, + "grad_norm": 4.246981143951416, + "learning_rate": 0.00019583895857276846, + "loss": 0.6727, + "step": 3940 + }, + { + "epoch": 0.09238324630648687, + "grad_norm": 2.53977108001709, + "learning_rate": 0.00019583685554366324, + "loss": 0.2818, + "step": 3941 + }, + { + "epoch": 0.09240668788129186, + "grad_norm": 4.0912861824035645, + "learning_rate": 0.00019583475199454452, + "loss": 0.8154, + "step": 3942 + }, + { + "epoch": 0.09243012945609685, + "grad_norm": 1.505935549736023, + "learning_rate": 0.00019583264792542374, + "loss": 0.198, + "step": 3943 + }, + { + "epoch": 0.09245357103090186, + "grad_norm": 4.1428046226501465, + "learning_rate": 0.00019583054333631236, + "loss": 0.7283, + "step": 3944 + }, + { + "epoch": 0.09247701260570686, + "grad_norm": 2.9178338050842285, + "learning_rate": 0.0001958284382272217, + "loss": 0.4689, + "step": 3945 + }, + { + "epoch": 0.09250045418051185, + "grad_norm": 2.192211389541626, + "learning_rate": 0.00019582633259816325, + "loss": 0.245, + "step": 3946 + }, + { + "epoch": 0.09252389575531685, + "grad_norm": 3.776125431060791, + "learning_rate": 0.00019582422644914846, + "loss": 0.4034, + "step": 3947 + }, + { + "epoch": 0.09254733733012184, + "grad_norm": 0.6913145780563354, + "learning_rate": 0.0001958221197801887, + "loss": 0.1348, + "step": 3948 + }, + { + "epoch": 0.09257077890492683, + "grad_norm": 1.3530653715133667, + "learning_rate": 0.00019582001259129542, + "loss": 0.2484, + "step": 3949 + }, + { + "epoch": 0.09259422047973183, + "grad_norm": 4.442366123199463, + "learning_rate": 0.00019581790488248008, + "loss": 0.4414, + "step": 3950 + }, + { + "epoch": 0.09261766205453682, + "grad_norm": 4.218746662139893, + "learning_rate": 0.00019581579665375407, + "loss": 0.607, + "step": 3951 + }, + { + "epoch": 0.09264110362934182, + "grad_norm": 3.505988121032715, + "learning_rate": 0.00019581368790512883, + "loss": 0.4289, + "step": 3952 + }, + { + "epoch": 0.09266454520414681, + "grad_norm": 2.4506304264068604, + "learning_rate": 0.0001958115786366159, + "loss": 1.6445, + "step": 3953 + }, + { + "epoch": 0.0926879867789518, + "grad_norm": 3.0508952140808105, + "learning_rate": 0.0001958094688482266, + "loss": 0.4108, + "step": 3954 + }, + { + "epoch": 0.0927114283537568, + "grad_norm": 5.0833563804626465, + "learning_rate": 0.00019580735853997241, + "loss": 0.6313, + "step": 3955 + }, + { + "epoch": 0.0927348699285618, + "grad_norm": 4.165139198303223, + "learning_rate": 0.00019580524771186481, + "loss": 0.8871, + "step": 3956 + }, + { + "epoch": 0.0927583115033668, + "grad_norm": 4.494751930236816, + "learning_rate": 0.00019580313636391529, + "loss": 0.3638, + "step": 3957 + }, + { + "epoch": 0.0927817530781718, + "grad_norm": 1.5672322511672974, + "learning_rate": 0.0001958010244961352, + "loss": 1.3505, + "step": 3958 + }, + { + "epoch": 0.09280519465297679, + "grad_norm": 3.8797686100006104, + "learning_rate": 0.0001957989121085361, + "loss": 0.6766, + "step": 3959 + }, + { + "epoch": 0.09282863622778179, + "grad_norm": 4.445271968841553, + "learning_rate": 0.00019579679920112936, + "loss": 0.3815, + "step": 3960 + }, + { + "epoch": 0.09285207780258678, + "grad_norm": 0.7254324555397034, + "learning_rate": 0.00019579468577392655, + "loss": 0.1315, + "step": 3961 + }, + { + "epoch": 0.09287551937739177, + "grad_norm": 3.731029510498047, + "learning_rate": 0.00019579257182693904, + "loss": 0.406, + "step": 3962 + }, + { + "epoch": 0.09289896095219677, + "grad_norm": 3.5368387699127197, + "learning_rate": 0.00019579045736017837, + "loss": 0.5341, + "step": 3963 + }, + { + "epoch": 0.09292240252700176, + "grad_norm": 4.354273319244385, + "learning_rate": 0.00019578834237365595, + "loss": 0.7182, + "step": 3964 + }, + { + "epoch": 0.09294584410180676, + "grad_norm": 3.2481343746185303, + "learning_rate": 0.00019578622686738332, + "loss": 0.4915, + "step": 3965 + }, + { + "epoch": 0.09296928567661175, + "grad_norm": 2.3303277492523193, + "learning_rate": 0.00019578411084137195, + "loss": 1.9076, + "step": 3966 + }, + { + "epoch": 0.09299272725141675, + "grad_norm": 6.067286491394043, + "learning_rate": 0.00019578199429563323, + "loss": 0.853, + "step": 3967 + }, + { + "epoch": 0.09301616882622174, + "grad_norm": 1.0123051404953003, + "learning_rate": 0.00019577987723017878, + "loss": 0.1532, + "step": 3968 + }, + { + "epoch": 0.09303961040102673, + "grad_norm": 3.0591318607330322, + "learning_rate": 0.00019577775964502002, + "loss": 0.4758, + "step": 3969 + }, + { + "epoch": 0.09306305197583174, + "grad_norm": 4.163092613220215, + "learning_rate": 0.0001957756415401684, + "loss": 0.5116, + "step": 3970 + }, + { + "epoch": 0.09308649355063674, + "grad_norm": 4.144264221191406, + "learning_rate": 0.00019577352291563549, + "loss": 0.345, + "step": 3971 + }, + { + "epoch": 0.09310993512544173, + "grad_norm": 3.0918591022491455, + "learning_rate": 0.00019577140377143276, + "loss": 1.1094, + "step": 3972 + }, + { + "epoch": 0.09313337670024673, + "grad_norm": 2.5669240951538086, + "learning_rate": 0.00019576928410757166, + "loss": 0.3137, + "step": 3973 + }, + { + "epoch": 0.09315681827505172, + "grad_norm": 2.514486789703369, + "learning_rate": 0.0001957671639240638, + "loss": 0.4972, + "step": 3974 + }, + { + "epoch": 0.09318025984985671, + "grad_norm": 2.5804922580718994, + "learning_rate": 0.00019576504322092056, + "loss": 0.3483, + "step": 3975 + }, + { + "epoch": 0.09320370142466171, + "grad_norm": 2.547236680984497, + "learning_rate": 0.00019576292199815348, + "loss": 0.3183, + "step": 3976 + }, + { + "epoch": 0.0932271429994667, + "grad_norm": 2.8263962268829346, + "learning_rate": 0.00019576080025577415, + "loss": 0.4275, + "step": 3977 + }, + { + "epoch": 0.0932505845742717, + "grad_norm": 3.023188352584839, + "learning_rate": 0.000195758677993794, + "loss": 0.3203, + "step": 3978 + }, + { + "epoch": 0.09327402614907669, + "grad_norm": 1.797062635421753, + "learning_rate": 0.00019575655521222456, + "loss": 1.3198, + "step": 3979 + }, + { + "epoch": 0.09329746772388169, + "grad_norm": 1.1505162715911865, + "learning_rate": 0.00019575443191107737, + "loss": 0.1838, + "step": 3980 + }, + { + "epoch": 0.09332090929868668, + "grad_norm": 3.691943407058716, + "learning_rate": 0.00019575230809036395, + "loss": 0.4002, + "step": 3981 + }, + { + "epoch": 0.09334435087349167, + "grad_norm": 4.348879814147949, + "learning_rate": 0.00019575018375009583, + "loss": 0.4684, + "step": 3982 + }, + { + "epoch": 0.09336779244829668, + "grad_norm": 3.012716054916382, + "learning_rate": 0.00019574805889028448, + "loss": 0.6376, + "step": 3983 + }, + { + "epoch": 0.09339123402310168, + "grad_norm": 1.933203935623169, + "learning_rate": 0.00019574593351094147, + "loss": 0.8559, + "step": 3984 + }, + { + "epoch": 0.09341467559790667, + "grad_norm": 3.904578447341919, + "learning_rate": 0.00019574380761207838, + "loss": 0.6482, + "step": 3985 + }, + { + "epoch": 0.09343811717271167, + "grad_norm": 5.2977614402771, + "learning_rate": 0.00019574168119370665, + "loss": 0.7293, + "step": 3986 + }, + { + "epoch": 0.09346155874751666, + "grad_norm": 4.88126277923584, + "learning_rate": 0.00019573955425583787, + "loss": 0.6569, + "step": 3987 + }, + { + "epoch": 0.09348500032232165, + "grad_norm": 0.5743271112442017, + "learning_rate": 0.0001957374267984836, + "loss": 0.0565, + "step": 3988 + }, + { + "epoch": 0.09350844189712665, + "grad_norm": 3.7986841201782227, + "learning_rate": 0.00019573529882165535, + "loss": 0.4989, + "step": 3989 + }, + { + "epoch": 0.09353188347193164, + "grad_norm": 4.708718776702881, + "learning_rate": 0.0001957331703253647, + "loss": 0.6035, + "step": 3990 + }, + { + "epoch": 0.09355532504673664, + "grad_norm": 9.227438926696777, + "learning_rate": 0.0001957310413096231, + "loss": 0.613, + "step": 3991 + }, + { + "epoch": 0.09357876662154163, + "grad_norm": 3.1942310333251953, + "learning_rate": 0.00019572891177444226, + "loss": 0.4401, + "step": 3992 + }, + { + "epoch": 0.09360220819634663, + "grad_norm": 5.4750075340271, + "learning_rate": 0.00019572678171983363, + "loss": 0.7452, + "step": 3993 + }, + { + "epoch": 0.09362564977115162, + "grad_norm": 4.223808765411377, + "learning_rate": 0.0001957246511458088, + "loss": 0.4361, + "step": 3994 + }, + { + "epoch": 0.09364909134595663, + "grad_norm": 4.0656938552856445, + "learning_rate": 0.0001957225200523793, + "loss": 0.52, + "step": 3995 + }, + { + "epoch": 0.09367253292076162, + "grad_norm": 1.168000340461731, + "learning_rate": 0.0001957203884395567, + "loss": 0.1676, + "step": 3996 + }, + { + "epoch": 0.09369597449556662, + "grad_norm": 0.836536705493927, + "learning_rate": 0.00019571825630735262, + "loss": 0.1425, + "step": 3997 + }, + { + "epoch": 0.09371941607037161, + "grad_norm": 4.852011203765869, + "learning_rate": 0.00019571612365577856, + "loss": 0.7814, + "step": 3998 + }, + { + "epoch": 0.0937428576451766, + "grad_norm": 2.1889920234680176, + "learning_rate": 0.00019571399048484613, + "loss": 0.3082, + "step": 3999 + }, + { + "epoch": 0.0937662992199816, + "grad_norm": 11.1832275390625, + "learning_rate": 0.0001957118567945669, + "loss": 0.3473, + "step": 4000 + }, + { + "epoch": 0.0937897407947866, + "grad_norm": 2.619415760040283, + "learning_rate": 0.00019570972258495245, + "loss": 1.4227, + "step": 4001 + }, + { + "epoch": 0.09381318236959159, + "grad_norm": 1.5557345151901245, + "learning_rate": 0.00019570758785601434, + "loss": 0.3096, + "step": 4002 + }, + { + "epoch": 0.09383662394439658, + "grad_norm": 5.561388969421387, + "learning_rate": 0.00019570545260776419, + "loss": 0.6877, + "step": 4003 + }, + { + "epoch": 0.09386006551920158, + "grad_norm": 4.805627822875977, + "learning_rate": 0.00019570331684021355, + "loss": 0.7037, + "step": 4004 + }, + { + "epoch": 0.09388350709400657, + "grad_norm": 4.2289581298828125, + "learning_rate": 0.00019570118055337404, + "loss": 0.5322, + "step": 4005 + }, + { + "epoch": 0.09390694866881157, + "grad_norm": 5.31380558013916, + "learning_rate": 0.00019569904374725717, + "loss": 0.8008, + "step": 4006 + }, + { + "epoch": 0.09393039024361656, + "grad_norm": 4.486664295196533, + "learning_rate": 0.00019569690642187468, + "loss": 0.6564, + "step": 4007 + }, + { + "epoch": 0.09395383181842157, + "grad_norm": 3.125868320465088, + "learning_rate": 0.00019569476857723804, + "loss": 0.3068, + "step": 4008 + }, + { + "epoch": 0.09397727339322656, + "grad_norm": 1.4261916875839233, + "learning_rate": 0.0001956926302133589, + "loss": 0.1167, + "step": 4009 + }, + { + "epoch": 0.09400071496803156, + "grad_norm": 3.8925998210906982, + "learning_rate": 0.00019569049133024886, + "loss": 0.7527, + "step": 4010 + }, + { + "epoch": 0.09402415654283655, + "grad_norm": 3.4423024654388428, + "learning_rate": 0.00019568835192791952, + "loss": 0.5123, + "step": 4011 + }, + { + "epoch": 0.09404759811764155, + "grad_norm": 6.815533638000488, + "learning_rate": 0.00019568621200638248, + "loss": 0.7513, + "step": 4012 + }, + { + "epoch": 0.09407103969244654, + "grad_norm": 0.7801187038421631, + "learning_rate": 0.00019568407156564938, + "loss": 0.1104, + "step": 4013 + }, + { + "epoch": 0.09409448126725153, + "grad_norm": 4.911304950714111, + "learning_rate": 0.0001956819306057318, + "loss": 0.7091, + "step": 4014 + }, + { + "epoch": 0.09411792284205653, + "grad_norm": 1.0616514682769775, + "learning_rate": 0.0001956797891266414, + "loss": 0.1158, + "step": 4015 + }, + { + "epoch": 0.09414136441686152, + "grad_norm": 4.044240474700928, + "learning_rate": 0.00019567764712838978, + "loss": 0.6239, + "step": 4016 + }, + { + "epoch": 0.09416480599166652, + "grad_norm": 2.6762583255767822, + "learning_rate": 0.00019567550461098855, + "loss": 2.0009, + "step": 4017 + }, + { + "epoch": 0.09418824756647151, + "grad_norm": 2.515000104904175, + "learning_rate": 0.0001956733615744493, + "loss": 0.498, + "step": 4018 + }, + { + "epoch": 0.0942116891412765, + "grad_norm": 5.8768630027771, + "learning_rate": 0.00019567121801878376, + "loss": 0.8023, + "step": 4019 + }, + { + "epoch": 0.0942351307160815, + "grad_norm": 3.655174732208252, + "learning_rate": 0.00019566907394400347, + "loss": 0.7786, + "step": 4020 + }, + { + "epoch": 0.09425857229088651, + "grad_norm": 1.340891718864441, + "learning_rate": 0.0001956669293501201, + "loss": 0.2377, + "step": 4021 + }, + { + "epoch": 0.0942820138656915, + "grad_norm": 4.808528900146484, + "learning_rate": 0.00019566478423714526, + "loss": 0.3525, + "step": 4022 + }, + { + "epoch": 0.0943054554404965, + "grad_norm": 2.8597424030303955, + "learning_rate": 0.00019566263860509064, + "loss": 0.5026, + "step": 4023 + }, + { + "epoch": 0.09432889701530149, + "grad_norm": 2.6309614181518555, + "learning_rate": 0.00019566049245396786, + "loss": 0.274, + "step": 4024 + }, + { + "epoch": 0.09435233859010649, + "grad_norm": 3.0832648277282715, + "learning_rate": 0.00019565834578378851, + "loss": 0.4414, + "step": 4025 + }, + { + "epoch": 0.09437578016491148, + "grad_norm": 4.429259300231934, + "learning_rate": 0.00019565619859456432, + "loss": 0.4298, + "step": 4026 + }, + { + "epoch": 0.09439922173971647, + "grad_norm": 5.759054660797119, + "learning_rate": 0.0001956540508863069, + "loss": 0.9246, + "step": 4027 + }, + { + "epoch": 0.09442266331452147, + "grad_norm": 3.0284202098846436, + "learning_rate": 0.00019565190265902793, + "loss": 0.4431, + "step": 4028 + }, + { + "epoch": 0.09444610488932646, + "grad_norm": 4.659571647644043, + "learning_rate": 0.00019564975391273904, + "loss": 0.5005, + "step": 4029 + }, + { + "epoch": 0.09446954646413146, + "grad_norm": 0.7105307579040527, + "learning_rate": 0.0001956476046474519, + "loss": 0.1077, + "step": 4030 + }, + { + "epoch": 0.09449298803893645, + "grad_norm": 4.173056602478027, + "learning_rate": 0.00019564545486317814, + "loss": 0.5371, + "step": 4031 + }, + { + "epoch": 0.09451642961374145, + "grad_norm": 4.515767574310303, + "learning_rate": 0.00019564330455992948, + "loss": 0.9316, + "step": 4032 + }, + { + "epoch": 0.09453987118854644, + "grad_norm": 3.702380657196045, + "learning_rate": 0.00019564115373771755, + "loss": 0.6684, + "step": 4033 + }, + { + "epoch": 0.09456331276335145, + "grad_norm": 1.205735683441162, + "learning_rate": 0.000195639002396554, + "loss": 0.1873, + "step": 4034 + }, + { + "epoch": 0.09458675433815644, + "grad_norm": 4.776761054992676, + "learning_rate": 0.00019563685053645058, + "loss": 0.8388, + "step": 4035 + }, + { + "epoch": 0.09461019591296144, + "grad_norm": 3.515702247619629, + "learning_rate": 0.0001956346981574189, + "loss": 0.7297, + "step": 4036 + }, + { + "epoch": 0.09463363748776643, + "grad_norm": 0.9832911491394043, + "learning_rate": 0.00019563254525947066, + "loss": 0.1359, + "step": 4037 + }, + { + "epoch": 0.09465707906257143, + "grad_norm": 0.6970996856689453, + "learning_rate": 0.00019563039184261754, + "loss": 0.0681, + "step": 4038 + }, + { + "epoch": 0.09468052063737642, + "grad_norm": 3.096966028213501, + "learning_rate": 0.00019562823790687122, + "loss": 0.3439, + "step": 4039 + }, + { + "epoch": 0.09470396221218141, + "grad_norm": 3.68292236328125, + "learning_rate": 0.0001956260834522434, + "loss": 0.7163, + "step": 4040 + }, + { + "epoch": 0.09472740378698641, + "grad_norm": 1.1220272779464722, + "learning_rate": 0.00019562392847874576, + "loss": 0.2049, + "step": 4041 + }, + { + "epoch": 0.0947508453617914, + "grad_norm": 3.7892942428588867, + "learning_rate": 0.00019562177298639, + "loss": 0.6722, + "step": 4042 + }, + { + "epoch": 0.0947742869365964, + "grad_norm": 4.053496837615967, + "learning_rate": 0.0001956196169751878, + "loss": 0.7756, + "step": 4043 + }, + { + "epoch": 0.09479772851140139, + "grad_norm": 3.9538559913635254, + "learning_rate": 0.00019561746044515087, + "loss": 0.4412, + "step": 4044 + }, + { + "epoch": 0.09482117008620639, + "grad_norm": 1.2816717624664307, + "learning_rate": 0.00019561530339629092, + "loss": 0.1664, + "step": 4045 + }, + { + "epoch": 0.09484461166101138, + "grad_norm": 3.7407824993133545, + "learning_rate": 0.00019561314582861966, + "loss": 0.5097, + "step": 4046 + }, + { + "epoch": 0.09486805323581639, + "grad_norm": 2.5864131450653076, + "learning_rate": 0.00019561098774214877, + "loss": 0.2696, + "step": 4047 + }, + { + "epoch": 0.09489149481062138, + "grad_norm": 2.7533206939697266, + "learning_rate": 0.00019560882913688995, + "loss": 0.3544, + "step": 4048 + }, + { + "epoch": 0.09491493638542638, + "grad_norm": 3.887498617172241, + "learning_rate": 0.00019560667001285495, + "loss": 0.8972, + "step": 4049 + }, + { + "epoch": 0.09493837796023137, + "grad_norm": 5.136560440063477, + "learning_rate": 0.00019560451037005548, + "loss": 0.5628, + "step": 4050 + }, + { + "epoch": 0.09496181953503637, + "grad_norm": 0.7721432447433472, + "learning_rate": 0.00019560235020850324, + "loss": 0.1302, + "step": 4051 + }, + { + "epoch": 0.09498526110984136, + "grad_norm": 1.2556167840957642, + "learning_rate": 0.00019560018952820998, + "loss": 0.1543, + "step": 4052 + }, + { + "epoch": 0.09500870268464635, + "grad_norm": 4.194615840911865, + "learning_rate": 0.00019559802832918736, + "loss": 0.5407, + "step": 4053 + }, + { + "epoch": 0.09503214425945135, + "grad_norm": 4.853603839874268, + "learning_rate": 0.0001955958666114472, + "loss": 0.747, + "step": 4054 + }, + { + "epoch": 0.09505558583425634, + "grad_norm": 2.6732823848724365, + "learning_rate": 0.00019559370437500112, + "loss": 1.4498, + "step": 4055 + }, + { + "epoch": 0.09507902740906134, + "grad_norm": 3.9107820987701416, + "learning_rate": 0.00019559154161986094, + "loss": 0.4253, + "step": 4056 + }, + { + "epoch": 0.09510246898386633, + "grad_norm": 2.8623101711273193, + "learning_rate": 0.0001955893783460384, + "loss": 0.4911, + "step": 4057 + }, + { + "epoch": 0.09512591055867133, + "grad_norm": 1.4340877532958984, + "learning_rate": 0.00019558721455354517, + "loss": 0.2946, + "step": 4058 + }, + { + "epoch": 0.09514935213347633, + "grad_norm": 4.561777114868164, + "learning_rate": 0.00019558505024239303, + "loss": 0.4513, + "step": 4059 + }, + { + "epoch": 0.09517279370828133, + "grad_norm": 5.618708610534668, + "learning_rate": 0.00019558288541259375, + "loss": 0.9842, + "step": 4060 + }, + { + "epoch": 0.09519623528308632, + "grad_norm": 2.768148422241211, + "learning_rate": 0.000195580720064159, + "loss": 0.3285, + "step": 4061 + }, + { + "epoch": 0.09521967685789132, + "grad_norm": 3.1113338470458984, + "learning_rate": 0.0001955785541971006, + "loss": 0.5819, + "step": 4062 + }, + { + "epoch": 0.09524311843269631, + "grad_norm": 1.484100341796875, + "learning_rate": 0.00019557638781143028, + "loss": 0.2469, + "step": 4063 + }, + { + "epoch": 0.0952665600075013, + "grad_norm": 3.377092123031616, + "learning_rate": 0.0001955742209071598, + "loss": 0.4874, + "step": 4064 + }, + { + "epoch": 0.0952900015823063, + "grad_norm": 5.206972122192383, + "learning_rate": 0.0001955720534843009, + "loss": 0.8355, + "step": 4065 + }, + { + "epoch": 0.0953134431571113, + "grad_norm": 2.5336716175079346, + "learning_rate": 0.00019556988554286534, + "loss": 0.4351, + "step": 4066 + }, + { + "epoch": 0.09533688473191629, + "grad_norm": 3.339712619781494, + "learning_rate": 0.0001955677170828649, + "loss": 0.4617, + "step": 4067 + }, + { + "epoch": 0.09536032630672128, + "grad_norm": 1.0144985914230347, + "learning_rate": 0.00019556554810431137, + "loss": 0.181, + "step": 4068 + }, + { + "epoch": 0.09538376788152628, + "grad_norm": 2.6533706188201904, + "learning_rate": 0.00019556337860721646, + "loss": 0.3921, + "step": 4069 + }, + { + "epoch": 0.09540720945633127, + "grad_norm": 4.404800891876221, + "learning_rate": 0.00019556120859159194, + "loss": 0.54, + "step": 4070 + }, + { + "epoch": 0.09543065103113627, + "grad_norm": 2.509866237640381, + "learning_rate": 0.00019555903805744965, + "loss": 1.5269, + "step": 4071 + }, + { + "epoch": 0.09545409260594127, + "grad_norm": 1.4465410709381104, + "learning_rate": 0.00019555686700480134, + "loss": 0.1252, + "step": 4072 + }, + { + "epoch": 0.09547753418074627, + "grad_norm": 4.5770182609558105, + "learning_rate": 0.00019555469543365876, + "loss": 0.8064, + "step": 4073 + }, + { + "epoch": 0.09550097575555126, + "grad_norm": 5.261317729949951, + "learning_rate": 0.00019555252334403373, + "loss": 0.93, + "step": 4074 + }, + { + "epoch": 0.09552441733035626, + "grad_norm": 3.5818264484405518, + "learning_rate": 0.00019555035073593802, + "loss": 0.219, + "step": 4075 + }, + { + "epoch": 0.09554785890516125, + "grad_norm": 3.8717658519744873, + "learning_rate": 0.00019554817760938342, + "loss": 0.6702, + "step": 4076 + }, + { + "epoch": 0.09557130047996625, + "grad_norm": 4.515472888946533, + "learning_rate": 0.00019554600396438172, + "loss": 0.4958, + "step": 4077 + }, + { + "epoch": 0.09559474205477124, + "grad_norm": 3.9021127223968506, + "learning_rate": 0.00019554382980094473, + "loss": 0.3425, + "step": 4078 + }, + { + "epoch": 0.09561818362957623, + "grad_norm": 2.9163334369659424, + "learning_rate": 0.0001955416551190842, + "loss": 0.298, + "step": 4079 + }, + { + "epoch": 0.09564162520438123, + "grad_norm": 7.453970432281494, + "learning_rate": 0.000195539479918812, + "loss": 0.9497, + "step": 4080 + }, + { + "epoch": 0.09566506677918622, + "grad_norm": 4.353397846221924, + "learning_rate": 0.00019553730420013987, + "loss": 0.674, + "step": 4081 + }, + { + "epoch": 0.09568850835399122, + "grad_norm": 1.2397657632827759, + "learning_rate": 0.00019553512796307966, + "loss": 0.1422, + "step": 4082 + }, + { + "epoch": 0.09571194992879621, + "grad_norm": 0.7741730213165283, + "learning_rate": 0.00019553295120764314, + "loss": 0.0722, + "step": 4083 + }, + { + "epoch": 0.0957353915036012, + "grad_norm": 5.140786647796631, + "learning_rate": 0.00019553077393384216, + "loss": 0.7114, + "step": 4084 + }, + { + "epoch": 0.09575883307840621, + "grad_norm": 4.915174961090088, + "learning_rate": 0.0001955285961416885, + "loss": 0.7739, + "step": 4085 + }, + { + "epoch": 0.09578227465321121, + "grad_norm": 3.8856537342071533, + "learning_rate": 0.000195526417831194, + "loss": 0.4934, + "step": 4086 + }, + { + "epoch": 0.0958057162280162, + "grad_norm": 5.707530498504639, + "learning_rate": 0.00019552423900237047, + "loss": 0.9804, + "step": 4087 + }, + { + "epoch": 0.0958291578028212, + "grad_norm": 4.063791275024414, + "learning_rate": 0.0001955220596552297, + "loss": 0.5022, + "step": 4088 + }, + { + "epoch": 0.09585259937762619, + "grad_norm": 2.6596696376800537, + "learning_rate": 0.00019551987978978362, + "loss": 0.2734, + "step": 4089 + }, + { + "epoch": 0.09587604095243119, + "grad_norm": 3.372249126434326, + "learning_rate": 0.0001955176994060439, + "loss": 0.5007, + "step": 4090 + }, + { + "epoch": 0.09589948252723618, + "grad_norm": 3.1398439407348633, + "learning_rate": 0.00019551551850402254, + "loss": 0.2786, + "step": 4091 + }, + { + "epoch": 0.09592292410204117, + "grad_norm": 2.5942819118499756, + "learning_rate": 0.00019551333708373126, + "loss": 0.3839, + "step": 4092 + }, + { + "epoch": 0.09594636567684617, + "grad_norm": 3.9748427867889404, + "learning_rate": 0.0001955111551451819, + "loss": 0.6946, + "step": 4093 + }, + { + "epoch": 0.09596980725165116, + "grad_norm": 3.8159778118133545, + "learning_rate": 0.00019550897268838635, + "loss": 0.6379, + "step": 4094 + }, + { + "epoch": 0.09599324882645616, + "grad_norm": 3.044550657272339, + "learning_rate": 0.00019550678971335644, + "loss": 0.3858, + "step": 4095 + }, + { + "epoch": 0.09601669040126115, + "grad_norm": 3.584364652633667, + "learning_rate": 0.00019550460622010397, + "loss": 0.6892, + "step": 4096 + }, + { + "epoch": 0.09604013197606615, + "grad_norm": 5.75269079208374, + "learning_rate": 0.00019550242220864086, + "loss": 0.9742, + "step": 4097 + }, + { + "epoch": 0.09606357355087115, + "grad_norm": 1.5014857053756714, + "learning_rate": 0.00019550023767897888, + "loss": 0.2312, + "step": 4098 + }, + { + "epoch": 0.09608701512567615, + "grad_norm": 4.3184428215026855, + "learning_rate": 0.00019549805263112995, + "loss": 1.0275, + "step": 4099 + }, + { + "epoch": 0.09611045670048114, + "grad_norm": 3.0862138271331787, + "learning_rate": 0.00019549586706510592, + "loss": 2.161, + "step": 4100 + }, + { + "epoch": 0.09613389827528614, + "grad_norm": 0.43744027614593506, + "learning_rate": 0.0001954936809809186, + "loss": 0.0546, + "step": 4101 + }, + { + "epoch": 0.09615733985009113, + "grad_norm": 2.9524929523468018, + "learning_rate": 0.0001954914943785799, + "loss": 0.5202, + "step": 4102 + }, + { + "epoch": 0.09618078142489613, + "grad_norm": 4.094585418701172, + "learning_rate": 0.00019548930725810166, + "loss": 0.6398, + "step": 4103 + }, + { + "epoch": 0.09620422299970112, + "grad_norm": 3.6379456520080566, + "learning_rate": 0.00019548711961949575, + "loss": 0.418, + "step": 4104 + }, + { + "epoch": 0.09622766457450611, + "grad_norm": 1.771546721458435, + "learning_rate": 0.00019548493146277402, + "loss": 0.3477, + "step": 4105 + }, + { + "epoch": 0.09625110614931111, + "grad_norm": 2.2970361709594727, + "learning_rate": 0.00019548274278794841, + "loss": 0.3294, + "step": 4106 + }, + { + "epoch": 0.0962745477241161, + "grad_norm": 0.8705961108207703, + "learning_rate": 0.00019548055359503073, + "loss": 0.1362, + "step": 4107 + }, + { + "epoch": 0.0962979892989211, + "grad_norm": 2.990490674972534, + "learning_rate": 0.00019547836388403292, + "loss": 0.5002, + "step": 4108 + }, + { + "epoch": 0.09632143087372609, + "grad_norm": 1.8001465797424316, + "learning_rate": 0.00019547617365496675, + "loss": 1.1949, + "step": 4109 + }, + { + "epoch": 0.0963448724485311, + "grad_norm": 1.3590950965881348, + "learning_rate": 0.00019547398290784424, + "loss": 0.2002, + "step": 4110 + }, + { + "epoch": 0.0963683140233361, + "grad_norm": 3.6283485889434814, + "learning_rate": 0.0001954717916426772, + "loss": 0.3809, + "step": 4111 + }, + { + "epoch": 0.09639175559814109, + "grad_norm": 0.9415833950042725, + "learning_rate": 0.0001954695998594775, + "loss": 0.1795, + "step": 4112 + }, + { + "epoch": 0.09641519717294608, + "grad_norm": 0.9689575433731079, + "learning_rate": 0.0001954674075582571, + "loss": 0.1453, + "step": 4113 + }, + { + "epoch": 0.09643863874775108, + "grad_norm": 2.6908411979675293, + "learning_rate": 0.00019546521473902786, + "loss": 0.416, + "step": 4114 + }, + { + "epoch": 0.09646208032255607, + "grad_norm": 1.2851063013076782, + "learning_rate": 0.00019546302140180166, + "loss": 0.1996, + "step": 4115 + }, + { + "epoch": 0.09648552189736107, + "grad_norm": 3.180837392807007, + "learning_rate": 0.00019546082754659045, + "loss": 0.4226, + "step": 4116 + }, + { + "epoch": 0.09650896347216606, + "grad_norm": 2.743985176086426, + "learning_rate": 0.0001954586331734061, + "loss": 0.409, + "step": 4117 + }, + { + "epoch": 0.09653240504697105, + "grad_norm": 3.103590488433838, + "learning_rate": 0.00019545643828226052, + "loss": 0.5059, + "step": 4118 + }, + { + "epoch": 0.09655584662177605, + "grad_norm": 3.1948187351226807, + "learning_rate": 0.0001954542428731656, + "loss": 0.4253, + "step": 4119 + }, + { + "epoch": 0.09657928819658104, + "grad_norm": 1.521041989326477, + "learning_rate": 0.0001954520469461333, + "loss": 0.1671, + "step": 4120 + }, + { + "epoch": 0.09660272977138604, + "grad_norm": 2.766371011734009, + "learning_rate": 0.0001954498505011755, + "loss": 0.4584, + "step": 4121 + }, + { + "epoch": 0.09662617134619103, + "grad_norm": 0.8908990621566772, + "learning_rate": 0.0001954476535383041, + "loss": 0.0998, + "step": 4122 + }, + { + "epoch": 0.09664961292099604, + "grad_norm": 4.385994911193848, + "learning_rate": 0.0001954454560575311, + "loss": 0.473, + "step": 4123 + }, + { + "epoch": 0.09667305449580103, + "grad_norm": 2.385542154312134, + "learning_rate": 0.00019544325805886836, + "loss": 0.272, + "step": 4124 + }, + { + "epoch": 0.09669649607060603, + "grad_norm": 2.2174184322357178, + "learning_rate": 0.0001954410595423278, + "loss": 1.0202, + "step": 4125 + }, + { + "epoch": 0.09671993764541102, + "grad_norm": 0.9103114604949951, + "learning_rate": 0.00019543886050792138, + "loss": 0.0852, + "step": 4126 + }, + { + "epoch": 0.09674337922021602, + "grad_norm": 1.0740735530853271, + "learning_rate": 0.000195436660955661, + "loss": 0.1151, + "step": 4127 + }, + { + "epoch": 0.09676682079502101, + "grad_norm": 3.9692604541778564, + "learning_rate": 0.00019543446088555864, + "loss": 0.5162, + "step": 4128 + }, + { + "epoch": 0.096790262369826, + "grad_norm": 2.518552780151367, + "learning_rate": 0.00019543226029762617, + "loss": 0.1427, + "step": 4129 + }, + { + "epoch": 0.096813703944631, + "grad_norm": 3.3688364028930664, + "learning_rate": 0.00019543005919187562, + "loss": 0.5292, + "step": 4130 + }, + { + "epoch": 0.096837145519436, + "grad_norm": 2.8720624446868896, + "learning_rate": 0.00019542785756831887, + "loss": 0.3919, + "step": 4131 + }, + { + "epoch": 0.09686058709424099, + "grad_norm": 3.7415921688079834, + "learning_rate": 0.00019542565542696787, + "loss": 0.7579, + "step": 4132 + }, + { + "epoch": 0.09688402866904598, + "grad_norm": 5.7607293128967285, + "learning_rate": 0.0001954234527678346, + "loss": 1.2534, + "step": 4133 + }, + { + "epoch": 0.09690747024385098, + "grad_norm": 7.64984655380249, + "learning_rate": 0.00019542124959093097, + "loss": 0.7049, + "step": 4134 + }, + { + "epoch": 0.09693091181865597, + "grad_norm": 3.441305637359619, + "learning_rate": 0.00019541904589626897, + "loss": 0.5793, + "step": 4135 + }, + { + "epoch": 0.09695435339346098, + "grad_norm": 3.658139228820801, + "learning_rate": 0.00019541684168386056, + "loss": 0.5695, + "step": 4136 + }, + { + "epoch": 0.09697779496826597, + "grad_norm": 3.6864562034606934, + "learning_rate": 0.00019541463695371767, + "loss": 0.4098, + "step": 4137 + }, + { + "epoch": 0.09700123654307097, + "grad_norm": 2.557016611099243, + "learning_rate": 0.00019541243170585225, + "loss": 0.3307, + "step": 4138 + }, + { + "epoch": 0.09702467811787596, + "grad_norm": 3.2118890285491943, + "learning_rate": 0.0001954102259402763, + "loss": 0.4711, + "step": 4139 + }, + { + "epoch": 0.09704811969268096, + "grad_norm": 4.115218162536621, + "learning_rate": 0.00019540801965700185, + "loss": 0.5644, + "step": 4140 + }, + { + "epoch": 0.09707156126748595, + "grad_norm": 3.564277410507202, + "learning_rate": 0.00019540581285604073, + "loss": 0.4023, + "step": 4141 + }, + { + "epoch": 0.09709500284229095, + "grad_norm": 2.704834222793579, + "learning_rate": 0.000195403605537405, + "loss": 2.0671, + "step": 4142 + }, + { + "epoch": 0.09711844441709594, + "grad_norm": 2.9824225902557373, + "learning_rate": 0.0001954013977011066, + "loss": 0.5294, + "step": 4143 + }, + { + "epoch": 0.09714188599190093, + "grad_norm": 0.6937596201896667, + "learning_rate": 0.0001953991893471576, + "loss": 0.1176, + "step": 4144 + }, + { + "epoch": 0.09716532756670593, + "grad_norm": 4.25129508972168, + "learning_rate": 0.0001953969804755699, + "loss": 0.5524, + "step": 4145 + }, + { + "epoch": 0.09718876914151092, + "grad_norm": 3.443451404571533, + "learning_rate": 0.00019539477108635546, + "loss": 0.5787, + "step": 4146 + }, + { + "epoch": 0.09721221071631592, + "grad_norm": 4.425863265991211, + "learning_rate": 0.00019539256117952634, + "loss": 0.4535, + "step": 4147 + }, + { + "epoch": 0.09723565229112091, + "grad_norm": 2.108067750930786, + "learning_rate": 0.0001953903507550945, + "loss": 0.251, + "step": 4148 + }, + { + "epoch": 0.09725909386592592, + "grad_norm": 2.0763142108917236, + "learning_rate": 0.0001953881398130719, + "loss": 0.2109, + "step": 4149 + }, + { + "epoch": 0.09728253544073091, + "grad_norm": 5.022745609283447, + "learning_rate": 0.0001953859283534706, + "loss": 0.7874, + "step": 4150 + }, + { + "epoch": 0.09730597701553591, + "grad_norm": 1.2931334972381592, + "learning_rate": 0.0001953837163763026, + "loss": 0.1725, + "step": 4151 + }, + { + "epoch": 0.0973294185903409, + "grad_norm": 4.0129594802856445, + "learning_rate": 0.00019538150388157983, + "loss": 0.6219, + "step": 4152 + }, + { + "epoch": 0.0973528601651459, + "grad_norm": 4.725438117980957, + "learning_rate": 0.00019537929086931434, + "loss": 0.5736, + "step": 4153 + }, + { + "epoch": 0.09737630173995089, + "grad_norm": 3.6059396266937256, + "learning_rate": 0.00019537707733951814, + "loss": 0.5308, + "step": 4154 + }, + { + "epoch": 0.09739974331475588, + "grad_norm": 4.642857074737549, + "learning_rate": 0.00019537486329220324, + "loss": 0.4302, + "step": 4155 + }, + { + "epoch": 0.09742318488956088, + "grad_norm": 4.435068607330322, + "learning_rate": 0.00019537264872738165, + "loss": 0.6756, + "step": 4156 + }, + { + "epoch": 0.09744662646436587, + "grad_norm": 4.093033313751221, + "learning_rate": 0.0001953704336450654, + "loss": 0.6007, + "step": 4157 + }, + { + "epoch": 0.09747006803917087, + "grad_norm": 1.444422960281372, + "learning_rate": 0.0001953682180452665, + "loss": 0.3076, + "step": 4158 + }, + { + "epoch": 0.09749350961397586, + "grad_norm": 3.497021198272705, + "learning_rate": 0.00019536600192799695, + "loss": 0.4901, + "step": 4159 + }, + { + "epoch": 0.09751695118878086, + "grad_norm": 3.8483352661132812, + "learning_rate": 0.00019536378529326877, + "loss": 0.5175, + "step": 4160 + }, + { + "epoch": 0.09754039276358585, + "grad_norm": 4.3772687911987305, + "learning_rate": 0.00019536156814109405, + "loss": 0.4849, + "step": 4161 + }, + { + "epoch": 0.09756383433839086, + "grad_norm": 4.973785877227783, + "learning_rate": 0.00019535935047148477, + "loss": 0.5257, + "step": 4162 + }, + { + "epoch": 0.09758727591319585, + "grad_norm": 1.21510910987854, + "learning_rate": 0.00019535713228445296, + "loss": 0.1272, + "step": 4163 + }, + { + "epoch": 0.09761071748800085, + "grad_norm": 0.9724964499473572, + "learning_rate": 0.00019535491358001067, + "loss": 0.1401, + "step": 4164 + }, + { + "epoch": 0.09763415906280584, + "grad_norm": 2.7087912559509277, + "learning_rate": 0.00019535269435816996, + "loss": 0.2623, + "step": 4165 + }, + { + "epoch": 0.09765760063761084, + "grad_norm": 1.4117461442947388, + "learning_rate": 0.0001953504746189428, + "loss": 0.1735, + "step": 4166 + }, + { + "epoch": 0.09768104221241583, + "grad_norm": 2.6059603691101074, + "learning_rate": 0.00019534825436234133, + "loss": 0.2853, + "step": 4167 + }, + { + "epoch": 0.09770448378722082, + "grad_norm": 0.4535065293312073, + "learning_rate": 0.00019534603358837755, + "loss": 0.0809, + "step": 4168 + }, + { + "epoch": 0.09772792536202582, + "grad_norm": 0.501304566860199, + "learning_rate": 0.0001953438122970635, + "loss": 0.0639, + "step": 4169 + }, + { + "epoch": 0.09775136693683081, + "grad_norm": 3.173597574234009, + "learning_rate": 0.00019534159048841123, + "loss": 0.357, + "step": 4170 + }, + { + "epoch": 0.09777480851163581, + "grad_norm": 2.3927419185638428, + "learning_rate": 0.00019533936816243283, + "loss": 0.2707, + "step": 4171 + }, + { + "epoch": 0.0977982500864408, + "grad_norm": 1.9132894277572632, + "learning_rate": 0.00019533714531914034, + "loss": 1.6302, + "step": 4172 + }, + { + "epoch": 0.0978216916612458, + "grad_norm": 0.81331866979599, + "learning_rate": 0.0001953349219585458, + "loss": 0.0913, + "step": 4173 + }, + { + "epoch": 0.0978451332360508, + "grad_norm": 3.617784023284912, + "learning_rate": 0.0001953326980806613, + "loss": 0.396, + "step": 4174 + }, + { + "epoch": 0.0978685748108558, + "grad_norm": 3.8079633712768555, + "learning_rate": 0.0001953304736854989, + "loss": 0.6644, + "step": 4175 + }, + { + "epoch": 0.0978920163856608, + "grad_norm": 4.638308048248291, + "learning_rate": 0.00019532824877307068, + "loss": 0.6831, + "step": 4176 + }, + { + "epoch": 0.09791545796046579, + "grad_norm": 0.6942754983901978, + "learning_rate": 0.00019532602334338867, + "loss": 0.066, + "step": 4177 + }, + { + "epoch": 0.09793889953527078, + "grad_norm": 0.7202619314193726, + "learning_rate": 0.000195323797396465, + "loss": 0.1078, + "step": 4178 + }, + { + "epoch": 0.09796234111007578, + "grad_norm": 3.91386079788208, + "learning_rate": 0.00019532157093231174, + "loss": 0.5161, + "step": 4179 + }, + { + "epoch": 0.09798578268488077, + "grad_norm": 4.256832122802734, + "learning_rate": 0.00019531934395094095, + "loss": 0.4925, + "step": 4180 + }, + { + "epoch": 0.09800922425968576, + "grad_norm": 0.7524421215057373, + "learning_rate": 0.0001953171164523647, + "loss": 0.1187, + "step": 4181 + }, + { + "epoch": 0.09803266583449076, + "grad_norm": 3.509061336517334, + "learning_rate": 0.00019531488843659508, + "loss": 1.8808, + "step": 4182 + }, + { + "epoch": 0.09805610740929575, + "grad_norm": 2.5004794597625732, + "learning_rate": 0.00019531265990364424, + "loss": 0.4365, + "step": 4183 + }, + { + "epoch": 0.09807954898410075, + "grad_norm": 3.4734652042388916, + "learning_rate": 0.00019531043085352422, + "loss": 0.4964, + "step": 4184 + }, + { + "epoch": 0.09810299055890574, + "grad_norm": 4.69498872756958, + "learning_rate": 0.0001953082012862471, + "loss": 0.8035, + "step": 4185 + }, + { + "epoch": 0.09812643213371074, + "grad_norm": 2.9844839572906494, + "learning_rate": 0.00019530597120182502, + "loss": 0.3404, + "step": 4186 + }, + { + "epoch": 0.09814987370851574, + "grad_norm": 0.7292380332946777, + "learning_rate": 0.00019530374060027005, + "loss": 0.0848, + "step": 4187 + }, + { + "epoch": 0.09817331528332074, + "grad_norm": 4.535686016082764, + "learning_rate": 0.0001953015094815943, + "loss": 0.5128, + "step": 4188 + }, + { + "epoch": 0.09819675685812573, + "grad_norm": 4.434859752655029, + "learning_rate": 0.00019529927784580988, + "loss": 0.6587, + "step": 4189 + }, + { + "epoch": 0.09822019843293073, + "grad_norm": 3.119279384613037, + "learning_rate": 0.00019529704569292892, + "loss": 0.4004, + "step": 4190 + }, + { + "epoch": 0.09824364000773572, + "grad_norm": 1.7993850708007812, + "learning_rate": 0.00019529481302296348, + "loss": 0.259, + "step": 4191 + }, + { + "epoch": 0.09826708158254072, + "grad_norm": 1.1204311847686768, + "learning_rate": 0.0001952925798359257, + "loss": 0.1522, + "step": 4192 + }, + { + "epoch": 0.09829052315734571, + "grad_norm": 0.7348646521568298, + "learning_rate": 0.00019529034613182774, + "loss": 0.1028, + "step": 4193 + }, + { + "epoch": 0.0983139647321507, + "grad_norm": 1.0501457452774048, + "learning_rate": 0.00019528811191068165, + "loss": 0.2355, + "step": 4194 + }, + { + "epoch": 0.0983374063069557, + "grad_norm": 0.8055974245071411, + "learning_rate": 0.0001952858771724996, + "loss": 0.0795, + "step": 4195 + }, + { + "epoch": 0.0983608478817607, + "grad_norm": 2.9629273414611816, + "learning_rate": 0.00019528364191729371, + "loss": 0.385, + "step": 4196 + }, + { + "epoch": 0.09838428945656569, + "grad_norm": 3.306702136993408, + "learning_rate": 0.0001952814061450761, + "loss": 0.4692, + "step": 4197 + }, + { + "epoch": 0.09840773103137068, + "grad_norm": 0.31733179092407227, + "learning_rate": 0.00019527916985585887, + "loss": 0.04, + "step": 4198 + }, + { + "epoch": 0.09843117260617568, + "grad_norm": 4.185917377471924, + "learning_rate": 0.00019527693304965423, + "loss": 0.566, + "step": 4199 + }, + { + "epoch": 0.09845461418098068, + "grad_norm": 0.4066653549671173, + "learning_rate": 0.00019527469572647426, + "loss": 0.0376, + "step": 4200 + }, + { + "epoch": 0.09847805575578568, + "grad_norm": 1.4281858205795288, + "learning_rate": 0.0001952724578863311, + "loss": 0.1979, + "step": 4201 + }, + { + "epoch": 0.09850149733059067, + "grad_norm": 4.103750228881836, + "learning_rate": 0.0001952702195292369, + "loss": 0.4718, + "step": 4202 + }, + { + "epoch": 0.09852493890539567, + "grad_norm": 2.3822152614593506, + "learning_rate": 0.0001952679806552038, + "loss": 1.5704, + "step": 4203 + }, + { + "epoch": 0.09854838048020066, + "grad_norm": 0.6771017909049988, + "learning_rate": 0.00019526574126424396, + "loss": 0.0703, + "step": 4204 + }, + { + "epoch": 0.09857182205500566, + "grad_norm": 4.8561811447143555, + "learning_rate": 0.00019526350135636955, + "loss": 0.6602, + "step": 4205 + }, + { + "epoch": 0.09859526362981065, + "grad_norm": 3.7408394813537598, + "learning_rate": 0.0001952612609315927, + "loss": 1.6854, + "step": 4206 + }, + { + "epoch": 0.09861870520461564, + "grad_norm": 0.6832720637321472, + "learning_rate": 0.00019525901998992558, + "loss": 0.0676, + "step": 4207 + }, + { + "epoch": 0.09864214677942064, + "grad_norm": 6.913822174072266, + "learning_rate": 0.0001952567785313803, + "loss": 0.6654, + "step": 4208 + }, + { + "epoch": 0.09866558835422563, + "grad_norm": 2.4867665767669678, + "learning_rate": 0.00019525453655596913, + "loss": 0.3466, + "step": 4209 + }, + { + "epoch": 0.09868902992903063, + "grad_norm": 1.2644264698028564, + "learning_rate": 0.00019525229406370409, + "loss": 0.1518, + "step": 4210 + }, + { + "epoch": 0.09871247150383562, + "grad_norm": 3.8336758613586426, + "learning_rate": 0.00019525005105459746, + "loss": 0.617, + "step": 4211 + }, + { + "epoch": 0.09873591307864062, + "grad_norm": 1.996370553970337, + "learning_rate": 0.00019524780752866138, + "loss": 0.2378, + "step": 4212 + }, + { + "epoch": 0.09875935465344562, + "grad_norm": 1.1426653861999512, + "learning_rate": 0.000195245563485908, + "loss": 0.1602, + "step": 4213 + }, + { + "epoch": 0.09878279622825062, + "grad_norm": 4.7720046043396, + "learning_rate": 0.00019524331892634952, + "loss": 1.0765, + "step": 4214 + }, + { + "epoch": 0.09880623780305561, + "grad_norm": 2.7517600059509277, + "learning_rate": 0.00019524107384999813, + "loss": 0.1699, + "step": 4215 + }, + { + "epoch": 0.09882967937786061, + "grad_norm": 3.052995443344116, + "learning_rate": 0.00019523882825686598, + "loss": 0.5111, + "step": 4216 + }, + { + "epoch": 0.0988531209526656, + "grad_norm": 2.94429874420166, + "learning_rate": 0.00019523658214696526, + "loss": 0.2862, + "step": 4217 + }, + { + "epoch": 0.0988765625274706, + "grad_norm": 4.407861709594727, + "learning_rate": 0.0001952343355203082, + "loss": 0.6734, + "step": 4218 + }, + { + "epoch": 0.09890000410227559, + "grad_norm": 4.074376583099365, + "learning_rate": 0.00019523208837690694, + "loss": 0.5882, + "step": 4219 + }, + { + "epoch": 0.09892344567708058, + "grad_norm": 0.7648013830184937, + "learning_rate": 0.00019522984071677367, + "loss": 0.1433, + "step": 4220 + }, + { + "epoch": 0.09894688725188558, + "grad_norm": 3.616570234298706, + "learning_rate": 0.00019522759253992063, + "loss": 0.3115, + "step": 4221 + }, + { + "epoch": 0.09897032882669057, + "grad_norm": 4.507469177246094, + "learning_rate": 0.00019522534384636, + "loss": 0.8123, + "step": 4222 + }, + { + "epoch": 0.09899377040149557, + "grad_norm": 3.3411953449249268, + "learning_rate": 0.00019522309463610398, + "loss": 0.4515, + "step": 4223 + }, + { + "epoch": 0.09901721197630056, + "grad_norm": 4.229162693023682, + "learning_rate": 0.00019522084490916476, + "loss": 0.5019, + "step": 4224 + }, + { + "epoch": 0.09904065355110556, + "grad_norm": 2.556291341781616, + "learning_rate": 0.00019521859466555458, + "loss": 0.2662, + "step": 4225 + }, + { + "epoch": 0.09906409512591056, + "grad_norm": 4.254819393157959, + "learning_rate": 0.00019521634390528558, + "loss": 1.714, + "step": 4226 + }, + { + "epoch": 0.09908753670071556, + "grad_norm": 3.2832841873168945, + "learning_rate": 0.00019521409262837004, + "loss": 0.679, + "step": 4227 + }, + { + "epoch": 0.09911097827552055, + "grad_norm": 1.0563304424285889, + "learning_rate": 0.00019521184083482016, + "loss": 0.1445, + "step": 4228 + }, + { + "epoch": 0.09913441985032555, + "grad_norm": 4.349493503570557, + "learning_rate": 0.00019520958852464816, + "loss": 0.5684, + "step": 4229 + }, + { + "epoch": 0.09915786142513054, + "grad_norm": 4.3644208908081055, + "learning_rate": 0.00019520733569786626, + "loss": 0.6778, + "step": 4230 + }, + { + "epoch": 0.09918130299993554, + "grad_norm": 2.3454430103302, + "learning_rate": 0.00019520508235448667, + "loss": 1.5823, + "step": 4231 + }, + { + "epoch": 0.09920474457474053, + "grad_norm": 4.488832473754883, + "learning_rate": 0.00019520282849452164, + "loss": 0.8843, + "step": 4232 + }, + { + "epoch": 0.09922818614954552, + "grad_norm": 8.515107154846191, + "learning_rate": 0.00019520057411798338, + "loss": 1.0276, + "step": 4233 + }, + { + "epoch": 0.09925162772435052, + "grad_norm": 2.7127039432525635, + "learning_rate": 0.00019519831922488413, + "loss": 0.3298, + "step": 4234 + }, + { + "epoch": 0.09927506929915551, + "grad_norm": 0.6487286686897278, + "learning_rate": 0.0001951960638152361, + "loss": 0.1308, + "step": 4235 + }, + { + "epoch": 0.09929851087396051, + "grad_norm": 3.256302833557129, + "learning_rate": 0.00019519380788905158, + "loss": 0.6907, + "step": 4236 + }, + { + "epoch": 0.0993219524487655, + "grad_norm": 1.462846279144287, + "learning_rate": 0.00019519155144634277, + "loss": 0.1628, + "step": 4237 + }, + { + "epoch": 0.09934539402357051, + "grad_norm": 4.089395523071289, + "learning_rate": 0.0001951892944871219, + "loss": 0.4167, + "step": 4238 + }, + { + "epoch": 0.0993688355983755, + "grad_norm": 0.8951584696769714, + "learning_rate": 0.00019518703701140125, + "loss": 0.1775, + "step": 4239 + }, + { + "epoch": 0.0993922771731805, + "grad_norm": 4.592530727386475, + "learning_rate": 0.0001951847790191931, + "loss": 0.7531, + "step": 4240 + }, + { + "epoch": 0.0994157187479855, + "grad_norm": 1.2453420162200928, + "learning_rate": 0.00019518252051050963, + "loss": 0.1593, + "step": 4241 + }, + { + "epoch": 0.09943916032279049, + "grad_norm": 2.855205535888672, + "learning_rate": 0.00019518026148536316, + "loss": 0.495, + "step": 4242 + }, + { + "epoch": 0.09946260189759548, + "grad_norm": 4.620323181152344, + "learning_rate": 0.00019517800194376585, + "loss": 0.7329, + "step": 4243 + }, + { + "epoch": 0.09948604347240048, + "grad_norm": 0.38938087224960327, + "learning_rate": 0.00019517574188573008, + "loss": 0.0241, + "step": 4244 + }, + { + "epoch": 0.09950948504720547, + "grad_norm": 3.0181095600128174, + "learning_rate": 0.00019517348131126806, + "loss": 0.4138, + "step": 4245 + }, + { + "epoch": 0.09953292662201046, + "grad_norm": 5.257248878479004, + "learning_rate": 0.00019517122022039203, + "loss": 0.7204, + "step": 4246 + }, + { + "epoch": 0.09955636819681546, + "grad_norm": 1.5333845615386963, + "learning_rate": 0.00019516895861311427, + "loss": 0.2958, + "step": 4247 + }, + { + "epoch": 0.09957980977162045, + "grad_norm": 4.548564434051514, + "learning_rate": 0.0001951666964894471, + "loss": 0.4414, + "step": 4248 + }, + { + "epoch": 0.09960325134642545, + "grad_norm": 3.0993409156799316, + "learning_rate": 0.00019516443384940275, + "loss": 0.367, + "step": 4249 + }, + { + "epoch": 0.09962669292123044, + "grad_norm": 2.52282977104187, + "learning_rate": 0.00019516217069299348, + "loss": 0.3168, + "step": 4250 + }, + { + "epoch": 0.09965013449603545, + "grad_norm": 3.845043182373047, + "learning_rate": 0.00019515990702023163, + "loss": 0.7924, + "step": 4251 + }, + { + "epoch": 0.09967357607084044, + "grad_norm": 2.3072168827056885, + "learning_rate": 0.00019515764283112945, + "loss": 0.4253, + "step": 4252 + }, + { + "epoch": 0.09969701764564544, + "grad_norm": 5.028664588928223, + "learning_rate": 0.0001951553781256992, + "loss": 0.4563, + "step": 4253 + }, + { + "epoch": 0.09972045922045043, + "grad_norm": 4.0353803634643555, + "learning_rate": 0.00019515311290395322, + "loss": 0.417, + "step": 4254 + }, + { + "epoch": 0.09974390079525543, + "grad_norm": 3.200734853744507, + "learning_rate": 0.00019515084716590375, + "loss": 0.4141, + "step": 4255 + }, + { + "epoch": 0.09976734237006042, + "grad_norm": 4.1403913497924805, + "learning_rate": 0.00019514858091156312, + "loss": 0.4715, + "step": 4256 + }, + { + "epoch": 0.09979078394486542, + "grad_norm": 7.932816505432129, + "learning_rate": 0.0001951463141409436, + "loss": 0.6869, + "step": 4257 + }, + { + "epoch": 0.09981422551967041, + "grad_norm": 0.7891521453857422, + "learning_rate": 0.0001951440468540575, + "loss": 0.0971, + "step": 4258 + }, + { + "epoch": 0.0998376670944754, + "grad_norm": 2.613798141479492, + "learning_rate": 0.00019514177905091717, + "loss": 0.2901, + "step": 4259 + }, + { + "epoch": 0.0998611086692804, + "grad_norm": 4.436920642852783, + "learning_rate": 0.00019513951073153485, + "loss": 0.4935, + "step": 4260 + }, + { + "epoch": 0.0998845502440854, + "grad_norm": 3.936314105987549, + "learning_rate": 0.00019513724189592285, + "loss": 0.4982, + "step": 4261 + }, + { + "epoch": 0.09990799181889039, + "grad_norm": 1.5559098720550537, + "learning_rate": 0.0001951349725440935, + "loss": 0.3035, + "step": 4262 + }, + { + "epoch": 0.09993143339369538, + "grad_norm": 4.295936584472656, + "learning_rate": 0.00019513270267605915, + "loss": 0.5141, + "step": 4263 + }, + { + "epoch": 0.09995487496850039, + "grad_norm": 3.7845242023468018, + "learning_rate": 0.00019513043229183206, + "loss": 0.8046, + "step": 4264 + }, + { + "epoch": 0.09997831654330538, + "grad_norm": 3.6566162109375, + "learning_rate": 0.00019512816139142456, + "loss": 0.5721, + "step": 4265 + }, + { + "epoch": 0.10000175811811038, + "grad_norm": 2.5686252117156982, + "learning_rate": 0.00019512588997484898, + "loss": 0.3823, + "step": 4266 + }, + { + "epoch": 0.10002519969291537, + "grad_norm": 4.059031963348389, + "learning_rate": 0.00019512361804211766, + "loss": 0.5158, + "step": 4267 + }, + { + "epoch": 0.10004864126772037, + "grad_norm": 2.4415054321289062, + "learning_rate": 0.0001951213455932429, + "loss": 0.4135, + "step": 4268 + }, + { + "epoch": 0.10007208284252536, + "grad_norm": 3.3202409744262695, + "learning_rate": 0.00019511907262823706, + "loss": 0.5174, + "step": 4269 + }, + { + "epoch": 0.10009552441733036, + "grad_norm": 3.7330846786499023, + "learning_rate": 0.00019511679914711245, + "loss": 0.6817, + "step": 4270 + }, + { + "epoch": 0.10011896599213535, + "grad_norm": 4.540284633636475, + "learning_rate": 0.00019511452514988143, + "loss": 0.7454, + "step": 4271 + }, + { + "epoch": 0.10014240756694034, + "grad_norm": 0.6471840739250183, + "learning_rate": 0.0001951122506365563, + "loss": 0.0863, + "step": 4272 + }, + { + "epoch": 0.10016584914174534, + "grad_norm": 1.147626519203186, + "learning_rate": 0.0001951099756071494, + "loss": 0.0968, + "step": 4273 + }, + { + "epoch": 0.10018929071655033, + "grad_norm": 2.4764015674591064, + "learning_rate": 0.00019510770006167313, + "loss": 0.2117, + "step": 4274 + }, + { + "epoch": 0.10021273229135533, + "grad_norm": 3.491426706314087, + "learning_rate": 0.00019510542400013982, + "loss": 0.5078, + "step": 4275 + }, + { + "epoch": 0.10023617386616032, + "grad_norm": 2.361712694168091, + "learning_rate": 0.0001951031474225618, + "loss": 0.3354, + "step": 4276 + }, + { + "epoch": 0.10025961544096533, + "grad_norm": 1.1832464933395386, + "learning_rate": 0.0001951008703289514, + "loss": 0.2057, + "step": 4277 + }, + { + "epoch": 0.10028305701577032, + "grad_norm": 5.203832626342773, + "learning_rate": 0.00019509859271932102, + "loss": 0.7889, + "step": 4278 + }, + { + "epoch": 0.10030649859057532, + "grad_norm": 3.086122512817383, + "learning_rate": 0.00019509631459368302, + "loss": 0.3715, + "step": 4279 + }, + { + "epoch": 0.10032994016538031, + "grad_norm": 4.328097820281982, + "learning_rate": 0.00019509403595204973, + "loss": 0.6127, + "step": 4280 + }, + { + "epoch": 0.10035338174018531, + "grad_norm": 2.393467426300049, + "learning_rate": 0.0001950917567944335, + "loss": 1.3903, + "step": 4281 + }, + { + "epoch": 0.1003768233149903, + "grad_norm": 3.5433390140533447, + "learning_rate": 0.00019508947712084675, + "loss": 0.698, + "step": 4282 + }, + { + "epoch": 0.1004002648897953, + "grad_norm": 4.163487911224365, + "learning_rate": 0.00019508719693130182, + "loss": 0.5669, + "step": 4283 + }, + { + "epoch": 0.10042370646460029, + "grad_norm": 4.798778533935547, + "learning_rate": 0.00019508491622581109, + "loss": 0.4993, + "step": 4284 + }, + { + "epoch": 0.10044714803940528, + "grad_norm": 3.645676851272583, + "learning_rate": 0.00019508263500438692, + "loss": 0.4267, + "step": 4285 + }, + { + "epoch": 0.10047058961421028, + "grad_norm": 2.9290270805358887, + "learning_rate": 0.00019508035326704171, + "loss": 0.4194, + "step": 4286 + }, + { + "epoch": 0.10049403118901527, + "grad_norm": 2.405632495880127, + "learning_rate": 0.00019507807101378785, + "loss": 1.4772, + "step": 4287 + }, + { + "epoch": 0.10051747276382027, + "grad_norm": 5.299931526184082, + "learning_rate": 0.00019507578824463762, + "loss": 0.8038, + "step": 4288 + }, + { + "epoch": 0.10054091433862528, + "grad_norm": 2.975674867630005, + "learning_rate": 0.00019507350495960357, + "loss": 0.281, + "step": 4289 + }, + { + "epoch": 0.10056435591343027, + "grad_norm": 1.5944476127624512, + "learning_rate": 0.000195071221158698, + "loss": 0.2829, + "step": 4290 + }, + { + "epoch": 0.10058779748823526, + "grad_norm": 1.9144258499145508, + "learning_rate": 0.0001950689368419333, + "loss": 0.2364, + "step": 4291 + }, + { + "epoch": 0.10061123906304026, + "grad_norm": 4.322939872741699, + "learning_rate": 0.00019506665200932186, + "loss": 0.6714, + "step": 4292 + }, + { + "epoch": 0.10063468063784525, + "grad_norm": 1.7608898878097534, + "learning_rate": 0.00019506436666087611, + "loss": 0.2789, + "step": 4293 + }, + { + "epoch": 0.10065812221265025, + "grad_norm": 2.3868892192840576, + "learning_rate": 0.00019506208079660841, + "loss": 1.2832, + "step": 4294 + }, + { + "epoch": 0.10068156378745524, + "grad_norm": 6.180210113525391, + "learning_rate": 0.00019505979441653123, + "loss": 0.7704, + "step": 4295 + }, + { + "epoch": 0.10070500536226024, + "grad_norm": 3.0758090019226074, + "learning_rate": 0.0001950575075206569, + "loss": 0.3659, + "step": 4296 + }, + { + "epoch": 0.10072844693706523, + "grad_norm": 2.3341734409332275, + "learning_rate": 0.00019505522010899786, + "loss": 0.3678, + "step": 4297 + }, + { + "epoch": 0.10075188851187022, + "grad_norm": 4.254685401916504, + "learning_rate": 0.00019505293218156656, + "loss": 0.4317, + "step": 4298 + }, + { + "epoch": 0.10077533008667522, + "grad_norm": 3.458038806915283, + "learning_rate": 0.00019505064373837535, + "loss": 0.5807, + "step": 4299 + }, + { + "epoch": 0.10079877166148021, + "grad_norm": 5.022936820983887, + "learning_rate": 0.00019504835477943667, + "loss": 0.7404, + "step": 4300 + }, + { + "epoch": 0.10082221323628521, + "grad_norm": 0.3368383049964905, + "learning_rate": 0.00019504606530476294, + "loss": 0.0479, + "step": 4301 + }, + { + "epoch": 0.10084565481109022, + "grad_norm": 3.156994342803955, + "learning_rate": 0.0001950437753143666, + "loss": 1.9556, + "step": 4302 + }, + { + "epoch": 0.10086909638589521, + "grad_norm": 3.758267879486084, + "learning_rate": 0.00019504148480826004, + "loss": 0.6125, + "step": 4303 + }, + { + "epoch": 0.1008925379607002, + "grad_norm": 3.464338541030884, + "learning_rate": 0.00019503919378645575, + "loss": 0.6055, + "step": 4304 + }, + { + "epoch": 0.1009159795355052, + "grad_norm": 4.0604143142700195, + "learning_rate": 0.00019503690224896613, + "loss": 0.4207, + "step": 4305 + }, + { + "epoch": 0.10093942111031019, + "grad_norm": 4.037724018096924, + "learning_rate": 0.00019503461019580357, + "loss": 0.4608, + "step": 4306 + }, + { + "epoch": 0.10096286268511519, + "grad_norm": 3.9258906841278076, + "learning_rate": 0.00019503231762698053, + "loss": 0.5158, + "step": 4307 + }, + { + "epoch": 0.10098630425992018, + "grad_norm": 2.503117084503174, + "learning_rate": 0.0001950300245425095, + "loss": 0.3946, + "step": 4308 + }, + { + "epoch": 0.10100974583472518, + "grad_norm": 2.8790948390960693, + "learning_rate": 0.00019502773094240287, + "loss": 0.3664, + "step": 4309 + }, + { + "epoch": 0.10103318740953017, + "grad_norm": 3.1012344360351562, + "learning_rate": 0.0001950254368266731, + "loss": 0.4353, + "step": 4310 + }, + { + "epoch": 0.10105662898433516, + "grad_norm": 1.8213893175125122, + "learning_rate": 0.0001950231421953326, + "loss": 0.3473, + "step": 4311 + }, + { + "epoch": 0.10108007055914016, + "grad_norm": 6.881180286407471, + "learning_rate": 0.00019502084704839395, + "loss": 0.9575, + "step": 4312 + }, + { + "epoch": 0.10110351213394515, + "grad_norm": 2.9913690090179443, + "learning_rate": 0.00019501855138586943, + "loss": 0.2655, + "step": 4313 + }, + { + "epoch": 0.10112695370875015, + "grad_norm": 1.0523152351379395, + "learning_rate": 0.00019501625520777161, + "loss": 0.1345, + "step": 4314 + }, + { + "epoch": 0.10115039528355516, + "grad_norm": 4.6111884117126465, + "learning_rate": 0.0001950139585141129, + "loss": 0.522, + "step": 4315 + }, + { + "epoch": 0.10117383685836015, + "grad_norm": 4.033519268035889, + "learning_rate": 0.00019501166130490584, + "loss": 0.5941, + "step": 4316 + }, + { + "epoch": 0.10119727843316514, + "grad_norm": 4.3444013595581055, + "learning_rate": 0.00019500936358016275, + "loss": 0.5179, + "step": 4317 + }, + { + "epoch": 0.10122072000797014, + "grad_norm": 4.27703857421875, + "learning_rate": 0.00019500706533989623, + "loss": 0.5792, + "step": 4318 + }, + { + "epoch": 0.10124416158277513, + "grad_norm": 3.8675248622894287, + "learning_rate": 0.00019500476658411867, + "loss": 0.428, + "step": 4319 + }, + { + "epoch": 0.10126760315758013, + "grad_norm": 3.5586392879486084, + "learning_rate": 0.0001950024673128426, + "loss": 0.6426, + "step": 4320 + }, + { + "epoch": 0.10129104473238512, + "grad_norm": 1.8421447277069092, + "learning_rate": 0.0001950001675260805, + "loss": 0.302, + "step": 4321 + }, + { + "epoch": 0.10131448630719012, + "grad_norm": 5.831228733062744, + "learning_rate": 0.0001949978672238448, + "loss": 0.7722, + "step": 4322 + }, + { + "epoch": 0.10133792788199511, + "grad_norm": 4.433956146240234, + "learning_rate": 0.00019499556640614795, + "loss": 0.5557, + "step": 4323 + }, + { + "epoch": 0.1013613694568001, + "grad_norm": 1.9788328409194946, + "learning_rate": 0.00019499326507300255, + "loss": 0.3454, + "step": 4324 + }, + { + "epoch": 0.1013848110316051, + "grad_norm": 2.794832229614258, + "learning_rate": 0.000194990963224421, + "loss": 0.2903, + "step": 4325 + }, + { + "epoch": 0.1014082526064101, + "grad_norm": 1.2634862661361694, + "learning_rate": 0.0001949886608604158, + "loss": 0.2168, + "step": 4326 + }, + { + "epoch": 0.10143169418121509, + "grad_norm": 2.8767590522766113, + "learning_rate": 0.00019498635798099944, + "loss": 0.5523, + "step": 4327 + }, + { + "epoch": 0.1014551357560201, + "grad_norm": 0.965247392654419, + "learning_rate": 0.00019498405458618447, + "loss": 0.097, + "step": 4328 + }, + { + "epoch": 0.10147857733082509, + "grad_norm": 2.439540386199951, + "learning_rate": 0.00019498175067598334, + "loss": 0.3791, + "step": 4329 + }, + { + "epoch": 0.10150201890563008, + "grad_norm": 2.088237762451172, + "learning_rate": 0.00019497944625040855, + "loss": 0.2868, + "step": 4330 + }, + { + "epoch": 0.10152546048043508, + "grad_norm": 0.7459337115287781, + "learning_rate": 0.00019497714130947263, + "loss": 0.0746, + "step": 4331 + }, + { + "epoch": 0.10154890205524007, + "grad_norm": 2.251004934310913, + "learning_rate": 0.00019497483585318806, + "loss": 0.2384, + "step": 4332 + }, + { + "epoch": 0.10157234363004507, + "grad_norm": 3.9775772094726562, + "learning_rate": 0.00019497252988156735, + "loss": 0.4208, + "step": 4333 + }, + { + "epoch": 0.10159578520485006, + "grad_norm": 3.7295119762420654, + "learning_rate": 0.00019497022339462304, + "loss": 0.3521, + "step": 4334 + }, + { + "epoch": 0.10161922677965506, + "grad_norm": 3.716491460800171, + "learning_rate": 0.00019496791639236763, + "loss": 0.438, + "step": 4335 + }, + { + "epoch": 0.10164266835446005, + "grad_norm": 3.6029016971588135, + "learning_rate": 0.00019496560887481362, + "loss": 0.6258, + "step": 4336 + }, + { + "epoch": 0.10166610992926504, + "grad_norm": 2.7991676330566406, + "learning_rate": 0.00019496330084197355, + "loss": 0.2219, + "step": 4337 + }, + { + "epoch": 0.10168955150407004, + "grad_norm": 3.159322738647461, + "learning_rate": 0.00019496099229385994, + "loss": 0.6197, + "step": 4338 + }, + { + "epoch": 0.10171299307887503, + "grad_norm": 0.8872643709182739, + "learning_rate": 0.00019495868323048533, + "loss": 0.1261, + "step": 4339 + }, + { + "epoch": 0.10173643465368003, + "grad_norm": 5.045068740844727, + "learning_rate": 0.0001949563736518622, + "loss": 0.8062, + "step": 4340 + }, + { + "epoch": 0.10175987622848504, + "grad_norm": 3.9000682830810547, + "learning_rate": 0.00019495406355800316, + "loss": 0.5335, + "step": 4341 + }, + { + "epoch": 0.10178331780329003, + "grad_norm": 1.7955399751663208, + "learning_rate": 0.00019495175294892067, + "loss": 0.2584, + "step": 4342 + }, + { + "epoch": 0.10180675937809502, + "grad_norm": 2.349799394607544, + "learning_rate": 0.00019494944182462732, + "loss": 1.6845, + "step": 4343 + }, + { + "epoch": 0.10183020095290002, + "grad_norm": 2.5374794006347656, + "learning_rate": 0.00019494713018513562, + "loss": 0.5097, + "step": 4344 + }, + { + "epoch": 0.10185364252770501, + "grad_norm": 2.7010648250579834, + "learning_rate": 0.0001949448180304581, + "loss": 0.2987, + "step": 4345 + }, + { + "epoch": 0.10187708410251001, + "grad_norm": 4.043478965759277, + "learning_rate": 0.00019494250536060737, + "loss": 0.4695, + "step": 4346 + }, + { + "epoch": 0.101900525677315, + "grad_norm": 3.972682237625122, + "learning_rate": 0.0001949401921755959, + "loss": 0.5849, + "step": 4347 + }, + { + "epoch": 0.10192396725212, + "grad_norm": 3.8509521484375, + "learning_rate": 0.0001949378784754363, + "loss": 0.5043, + "step": 4348 + }, + { + "epoch": 0.10194740882692499, + "grad_norm": 2.7967495918273926, + "learning_rate": 0.0001949355642601411, + "loss": 0.3367, + "step": 4349 + }, + { + "epoch": 0.10197085040172998, + "grad_norm": 3.228023052215576, + "learning_rate": 0.00019493324952972284, + "loss": 1.6237, + "step": 4350 + }, + { + "epoch": 0.10199429197653498, + "grad_norm": 1.2995198965072632, + "learning_rate": 0.00019493093428419412, + "loss": 0.1855, + "step": 4351 + }, + { + "epoch": 0.10201773355133997, + "grad_norm": 4.434151649475098, + "learning_rate": 0.00019492861852356746, + "loss": 0.5189, + "step": 4352 + }, + { + "epoch": 0.10204117512614498, + "grad_norm": 3.016669988632202, + "learning_rate": 0.00019492630224785546, + "loss": 0.5612, + "step": 4353 + }, + { + "epoch": 0.10206461670094998, + "grad_norm": 3.1073973178863525, + "learning_rate": 0.00019492398545707066, + "loss": 0.392, + "step": 4354 + }, + { + "epoch": 0.10208805827575497, + "grad_norm": 2.741460084915161, + "learning_rate": 0.00019492166815122567, + "loss": 0.2863, + "step": 4355 + }, + { + "epoch": 0.10211149985055996, + "grad_norm": 3.039761543273926, + "learning_rate": 0.00019491935033033304, + "loss": 0.3137, + "step": 4356 + }, + { + "epoch": 0.10213494142536496, + "grad_norm": 3.7172629833221436, + "learning_rate": 0.00019491703199440535, + "loss": 0.6786, + "step": 4357 + }, + { + "epoch": 0.10215838300016995, + "grad_norm": 1.8550509214401245, + "learning_rate": 0.00019491471314345515, + "loss": 0.2781, + "step": 4358 + }, + { + "epoch": 0.10218182457497495, + "grad_norm": 0.8488295078277588, + "learning_rate": 0.00019491239377749506, + "loss": 0.1044, + "step": 4359 + }, + { + "epoch": 0.10220526614977994, + "grad_norm": 3.93302321434021, + "learning_rate": 0.00019491007389653767, + "loss": 0.3998, + "step": 4360 + }, + { + "epoch": 0.10222870772458494, + "grad_norm": 3.126498222351074, + "learning_rate": 0.00019490775350059552, + "loss": 0.5527, + "step": 4361 + }, + { + "epoch": 0.10225214929938993, + "grad_norm": 4.072888374328613, + "learning_rate": 0.00019490543258968124, + "loss": 0.6867, + "step": 4362 + }, + { + "epoch": 0.10227559087419492, + "grad_norm": 3.546309232711792, + "learning_rate": 0.00019490311116380742, + "loss": 0.4464, + "step": 4363 + }, + { + "epoch": 0.10229903244899992, + "grad_norm": 4.3904876708984375, + "learning_rate": 0.00019490078922298665, + "loss": 0.4308, + "step": 4364 + }, + { + "epoch": 0.10232247402380491, + "grad_norm": 2.364027261734009, + "learning_rate": 0.00019489846676723152, + "loss": 0.4017, + "step": 4365 + }, + { + "epoch": 0.10234591559860992, + "grad_norm": 4.162133693695068, + "learning_rate": 0.00019489614379655465, + "loss": 0.4082, + "step": 4366 + }, + { + "epoch": 0.10236935717341492, + "grad_norm": 1.4913276433944702, + "learning_rate": 0.00019489382031096864, + "loss": 1.2024, + "step": 4367 + }, + { + "epoch": 0.10239279874821991, + "grad_norm": 4.376585006713867, + "learning_rate": 0.00019489149631048608, + "loss": 0.5486, + "step": 4368 + }, + { + "epoch": 0.1024162403230249, + "grad_norm": 2.033086061477661, + "learning_rate": 0.00019488917179511962, + "loss": 0.1902, + "step": 4369 + }, + { + "epoch": 0.1024396818978299, + "grad_norm": 3.2590713500976562, + "learning_rate": 0.0001948868467648818, + "loss": 0.5206, + "step": 4370 + }, + { + "epoch": 0.10246312347263489, + "grad_norm": 4.040148735046387, + "learning_rate": 0.0001948845212197853, + "loss": 0.5512, + "step": 4371 + }, + { + "epoch": 0.10248656504743989, + "grad_norm": 0.7038719654083252, + "learning_rate": 0.00019488219515984275, + "loss": 0.0713, + "step": 4372 + }, + { + "epoch": 0.10251000662224488, + "grad_norm": 5.938430309295654, + "learning_rate": 0.0001948798685850667, + "loss": 0.9138, + "step": 4373 + }, + { + "epoch": 0.10253344819704988, + "grad_norm": 5.909849643707275, + "learning_rate": 0.00019487754149546984, + "loss": 0.4273, + "step": 4374 + }, + { + "epoch": 0.10255688977185487, + "grad_norm": 0.9836872816085815, + "learning_rate": 0.00019487521389106478, + "loss": 0.1597, + "step": 4375 + }, + { + "epoch": 0.10258033134665986, + "grad_norm": 3.9765262603759766, + "learning_rate": 0.0001948728857718641, + "loss": 0.7255, + "step": 4376 + }, + { + "epoch": 0.10260377292146486, + "grad_norm": 2.8810033798217773, + "learning_rate": 0.00019487055713788052, + "loss": 0.4101, + "step": 4377 + }, + { + "epoch": 0.10262721449626985, + "grad_norm": 3.312223196029663, + "learning_rate": 0.0001948682279891266, + "loss": 0.5303, + "step": 4378 + }, + { + "epoch": 0.10265065607107486, + "grad_norm": 4.32851505279541, + "learning_rate": 0.00019486589832561503, + "loss": 0.8264, + "step": 4379 + }, + { + "epoch": 0.10267409764587986, + "grad_norm": 2.9708411693573, + "learning_rate": 0.00019486356814735843, + "loss": 0.5712, + "step": 4380 + }, + { + "epoch": 0.10269753922068485, + "grad_norm": 5.544332981109619, + "learning_rate": 0.0001948612374543694, + "loss": 1.2084, + "step": 4381 + }, + { + "epoch": 0.10272098079548984, + "grad_norm": 3.541839122772217, + "learning_rate": 0.00019485890624666068, + "loss": 0.7015, + "step": 4382 + }, + { + "epoch": 0.10274442237029484, + "grad_norm": 0.8330801129341125, + "learning_rate": 0.00019485657452424483, + "loss": 0.1668, + "step": 4383 + }, + { + "epoch": 0.10276786394509983, + "grad_norm": 0.5520254373550415, + "learning_rate": 0.00019485424228713456, + "loss": 0.0592, + "step": 4384 + }, + { + "epoch": 0.10279130551990483, + "grad_norm": 3.9769461154937744, + "learning_rate": 0.0001948519095353425, + "loss": 0.4945, + "step": 4385 + }, + { + "epoch": 0.10281474709470982, + "grad_norm": 3.299643039703369, + "learning_rate": 0.0001948495762688813, + "loss": 0.518, + "step": 4386 + }, + { + "epoch": 0.10283818866951482, + "grad_norm": 4.519689083099365, + "learning_rate": 0.00019484724248776365, + "loss": 0.3071, + "step": 4387 + }, + { + "epoch": 0.10286163024431981, + "grad_norm": 2.425626516342163, + "learning_rate": 0.0001948449081920022, + "loss": 1.314, + "step": 4388 + }, + { + "epoch": 0.1028850718191248, + "grad_norm": 2.283491849899292, + "learning_rate": 0.00019484257338160958, + "loss": 0.2172, + "step": 4389 + }, + { + "epoch": 0.1029085133939298, + "grad_norm": 3.466493606567383, + "learning_rate": 0.00019484023805659852, + "loss": 0.3552, + "step": 4390 + }, + { + "epoch": 0.10293195496873479, + "grad_norm": 3.681380271911621, + "learning_rate": 0.00019483790221698163, + "loss": 0.7007, + "step": 4391 + }, + { + "epoch": 0.1029553965435398, + "grad_norm": 1.0860849618911743, + "learning_rate": 0.00019483556586277165, + "loss": 0.6838, + "step": 4392 + }, + { + "epoch": 0.1029788381183448, + "grad_norm": 3.1034481525421143, + "learning_rate": 0.0001948332289939812, + "loss": 0.3747, + "step": 4393 + }, + { + "epoch": 0.10300227969314979, + "grad_norm": 3.582338809967041, + "learning_rate": 0.000194830891610623, + "loss": 0.3536, + "step": 4394 + }, + { + "epoch": 0.10302572126795478, + "grad_norm": 0.997823178768158, + "learning_rate": 0.00019482855371270972, + "loss": 0.1221, + "step": 4395 + }, + { + "epoch": 0.10304916284275978, + "grad_norm": 3.1229732036590576, + "learning_rate": 0.00019482621530025402, + "loss": 0.4799, + "step": 4396 + }, + { + "epoch": 0.10307260441756477, + "grad_norm": 2.9469738006591797, + "learning_rate": 0.00019482387637326863, + "loss": 0.157, + "step": 4397 + }, + { + "epoch": 0.10309604599236977, + "grad_norm": 1.3215309381484985, + "learning_rate": 0.00019482153693176622, + "loss": 0.1211, + "step": 4398 + }, + { + "epoch": 0.10311948756717476, + "grad_norm": 2.4355785846710205, + "learning_rate": 0.00019481919697575945, + "loss": 0.2728, + "step": 4399 + }, + { + "epoch": 0.10314292914197976, + "grad_norm": 1.7704304456710815, + "learning_rate": 0.00019481685650526108, + "loss": 0.2688, + "step": 4400 + }, + { + "epoch": 0.10316637071678475, + "grad_norm": 3.593350648880005, + "learning_rate": 0.0001948145155202838, + "loss": 0.4331, + "step": 4401 + }, + { + "epoch": 0.10318981229158974, + "grad_norm": 0.6701017618179321, + "learning_rate": 0.00019481217402084024, + "loss": 0.077, + "step": 4402 + }, + { + "epoch": 0.10321325386639474, + "grad_norm": 1.4283194541931152, + "learning_rate": 0.00019480983200694318, + "loss": 0.2867, + "step": 4403 + }, + { + "epoch": 0.10323669544119975, + "grad_norm": 0.9838966131210327, + "learning_rate": 0.00019480748947860535, + "loss": 0.5061, + "step": 4404 + }, + { + "epoch": 0.10326013701600474, + "grad_norm": 2.2337114810943604, + "learning_rate": 0.00019480514643583935, + "loss": 0.3485, + "step": 4405 + }, + { + "epoch": 0.10328357859080974, + "grad_norm": 2.861220359802246, + "learning_rate": 0.000194802802878658, + "loss": 0.3352, + "step": 4406 + }, + { + "epoch": 0.10330702016561473, + "grad_norm": 3.7242305278778076, + "learning_rate": 0.00019480045880707395, + "loss": 0.6432, + "step": 4407 + }, + { + "epoch": 0.10333046174041972, + "grad_norm": 3.410759449005127, + "learning_rate": 0.00019479811422109996, + "loss": 0.4739, + "step": 4408 + }, + { + "epoch": 0.10335390331522472, + "grad_norm": 3.5230891704559326, + "learning_rate": 0.00019479576912074875, + "loss": 0.4461, + "step": 4409 + }, + { + "epoch": 0.10337734489002971, + "grad_norm": 2.5547962188720703, + "learning_rate": 0.000194793423506033, + "loss": 2.0095, + "step": 4410 + }, + { + "epoch": 0.1034007864648347, + "grad_norm": 3.987353801727295, + "learning_rate": 0.0001947910773769655, + "loss": 0.5332, + "step": 4411 + }, + { + "epoch": 0.1034242280396397, + "grad_norm": 1.7863526344299316, + "learning_rate": 0.00019478873073355891, + "loss": 0.2594, + "step": 4412 + }, + { + "epoch": 0.1034476696144447, + "grad_norm": 2.5270063877105713, + "learning_rate": 0.00019478638357582604, + "loss": 0.2451, + "step": 4413 + }, + { + "epoch": 0.10347111118924969, + "grad_norm": 1.6157119274139404, + "learning_rate": 0.00019478403590377958, + "loss": 0.209, + "step": 4414 + }, + { + "epoch": 0.10349455276405468, + "grad_norm": 2.7946724891662598, + "learning_rate": 0.00019478168771743228, + "loss": 0.524, + "step": 4415 + }, + { + "epoch": 0.10351799433885968, + "grad_norm": 2.312520742416382, + "learning_rate": 0.00019477933901679685, + "loss": 0.3259, + "step": 4416 + }, + { + "epoch": 0.10354143591366469, + "grad_norm": 3.519659996032715, + "learning_rate": 0.0001947769898018861, + "loss": 0.6373, + "step": 4417 + }, + { + "epoch": 0.10356487748846968, + "grad_norm": 3.475236415863037, + "learning_rate": 0.0001947746400727127, + "loss": 0.5922, + "step": 4418 + }, + { + "epoch": 0.10358831906327468, + "grad_norm": 3.26375675201416, + "learning_rate": 0.00019477228982928946, + "loss": 0.3037, + "step": 4419 + }, + { + "epoch": 0.10361176063807967, + "grad_norm": 2.7617881298065186, + "learning_rate": 0.0001947699390716291, + "loss": 0.5038, + "step": 4420 + }, + { + "epoch": 0.10363520221288466, + "grad_norm": 3.6746675968170166, + "learning_rate": 0.0001947675877997444, + "loss": 0.4606, + "step": 4421 + }, + { + "epoch": 0.10365864378768966, + "grad_norm": 1.2732237577438354, + "learning_rate": 0.0001947652360136481, + "loss": 0.1553, + "step": 4422 + }, + { + "epoch": 0.10368208536249465, + "grad_norm": 4.957427024841309, + "learning_rate": 0.00019476288371335297, + "loss": 0.7229, + "step": 4423 + }, + { + "epoch": 0.10370552693729965, + "grad_norm": 3.5019583702087402, + "learning_rate": 0.00019476053089887175, + "loss": 0.4713, + "step": 4424 + }, + { + "epoch": 0.10372896851210464, + "grad_norm": 4.2751665115356445, + "learning_rate": 0.00019475817757021725, + "loss": 0.6061, + "step": 4425 + }, + { + "epoch": 0.10375241008690964, + "grad_norm": 5.545104503631592, + "learning_rate": 0.00019475582372740217, + "loss": 0.6855, + "step": 4426 + }, + { + "epoch": 0.10377585166171463, + "grad_norm": 3.629847764968872, + "learning_rate": 0.00019475346937043936, + "loss": 0.364, + "step": 4427 + }, + { + "epoch": 0.10379929323651962, + "grad_norm": 1.8345186710357666, + "learning_rate": 0.00019475111449934155, + "loss": 1.5691, + "step": 4428 + }, + { + "epoch": 0.10382273481132462, + "grad_norm": 0.7564935088157654, + "learning_rate": 0.00019474875911412156, + "loss": 0.0673, + "step": 4429 + }, + { + "epoch": 0.10384617638612963, + "grad_norm": 3.989239454269409, + "learning_rate": 0.0001947464032147921, + "loss": 0.6078, + "step": 4430 + }, + { + "epoch": 0.10386961796093462, + "grad_norm": 2.25012469291687, + "learning_rate": 0.000194744046801366, + "loss": 1.5168, + "step": 4431 + }, + { + "epoch": 0.10389305953573962, + "grad_norm": 2.923097848892212, + "learning_rate": 0.00019474168987385605, + "loss": 0.4591, + "step": 4432 + }, + { + "epoch": 0.10391650111054461, + "grad_norm": 2.1342875957489014, + "learning_rate": 0.00019473933243227505, + "loss": 0.3365, + "step": 4433 + }, + { + "epoch": 0.1039399426853496, + "grad_norm": 3.2790513038635254, + "learning_rate": 0.00019473697447663574, + "loss": 0.6022, + "step": 4434 + }, + { + "epoch": 0.1039633842601546, + "grad_norm": 4.064525604248047, + "learning_rate": 0.00019473461600695094, + "loss": 0.4354, + "step": 4435 + }, + { + "epoch": 0.10398682583495959, + "grad_norm": 2.599241018295288, + "learning_rate": 0.00019473225702323343, + "loss": 0.3133, + "step": 4436 + }, + { + "epoch": 0.10401026740976459, + "grad_norm": 3.1509668827056885, + "learning_rate": 0.00019472989752549605, + "loss": 1.699, + "step": 4437 + }, + { + "epoch": 0.10403370898456958, + "grad_norm": 3.218092203140259, + "learning_rate": 0.0001947275375137516, + "loss": 0.5257, + "step": 4438 + }, + { + "epoch": 0.10405715055937458, + "grad_norm": 3.3324134349823, + "learning_rate": 0.00019472517698801282, + "loss": 0.4896, + "step": 4439 + }, + { + "epoch": 0.10408059213417957, + "grad_norm": 2.8737549781799316, + "learning_rate": 0.0001947228159482926, + "loss": 0.5298, + "step": 4440 + }, + { + "epoch": 0.10410403370898456, + "grad_norm": 2.952126979827881, + "learning_rate": 0.00019472045439460374, + "loss": 0.3446, + "step": 4441 + }, + { + "epoch": 0.10412747528378956, + "grad_norm": 0.757159948348999, + "learning_rate": 0.00019471809232695901, + "loss": 0.0787, + "step": 4442 + }, + { + "epoch": 0.10415091685859457, + "grad_norm": 2.0200185775756836, + "learning_rate": 0.00019471572974537124, + "loss": 1.224, + "step": 4443 + }, + { + "epoch": 0.10417435843339956, + "grad_norm": 3.978070020675659, + "learning_rate": 0.00019471336664985325, + "loss": 0.5266, + "step": 4444 + }, + { + "epoch": 0.10419780000820456, + "grad_norm": 1.2817808389663696, + "learning_rate": 0.00019471100304041788, + "loss": 0.2088, + "step": 4445 + }, + { + "epoch": 0.10422124158300955, + "grad_norm": 7.082512855529785, + "learning_rate": 0.00019470863891707794, + "loss": 0.4511, + "step": 4446 + }, + { + "epoch": 0.10424468315781454, + "grad_norm": 4.033266067504883, + "learning_rate": 0.00019470627427984626, + "loss": 0.5724, + "step": 4447 + }, + { + "epoch": 0.10426812473261954, + "grad_norm": 4.1532979011535645, + "learning_rate": 0.0001947039091287357, + "loss": 0.3549, + "step": 4448 + }, + { + "epoch": 0.10429156630742453, + "grad_norm": 3.467017412185669, + "learning_rate": 0.00019470154346375903, + "loss": 1.861, + "step": 4449 + }, + { + "epoch": 0.10431500788222953, + "grad_norm": 2.5121798515319824, + "learning_rate": 0.00019469917728492914, + "loss": 0.3147, + "step": 4450 + }, + { + "epoch": 0.10433844945703452, + "grad_norm": 5.282217025756836, + "learning_rate": 0.00019469681059225885, + "loss": 0.7199, + "step": 4451 + }, + { + "epoch": 0.10436189103183952, + "grad_norm": 4.245136260986328, + "learning_rate": 0.00019469444338576103, + "loss": 2.5745, + "step": 4452 + }, + { + "epoch": 0.10438533260664451, + "grad_norm": 4.99698543548584, + "learning_rate": 0.00019469207566544846, + "loss": 0.6485, + "step": 4453 + }, + { + "epoch": 0.1044087741814495, + "grad_norm": 3.740981340408325, + "learning_rate": 0.00019468970743133403, + "loss": 0.6417, + "step": 4454 + }, + { + "epoch": 0.1044322157562545, + "grad_norm": 3.033738136291504, + "learning_rate": 0.0001946873386834306, + "loss": 0.168, + "step": 4455 + }, + { + "epoch": 0.1044556573310595, + "grad_norm": 3.8657755851745605, + "learning_rate": 0.000194684969421751, + "loss": 0.5303, + "step": 4456 + }, + { + "epoch": 0.1044790989058645, + "grad_norm": 2.664991855621338, + "learning_rate": 0.00019468259964630812, + "loss": 0.2828, + "step": 4457 + }, + { + "epoch": 0.1045025404806695, + "grad_norm": 1.6933517456054688, + "learning_rate": 0.00019468022935711473, + "loss": 0.1646, + "step": 4458 + }, + { + "epoch": 0.10452598205547449, + "grad_norm": 3.934016704559326, + "learning_rate": 0.00019467785855418378, + "loss": 0.5066, + "step": 4459 + }, + { + "epoch": 0.10454942363027948, + "grad_norm": 5.064457416534424, + "learning_rate": 0.00019467548723752813, + "loss": 0.6872, + "step": 4460 + }, + { + "epoch": 0.10457286520508448, + "grad_norm": 2.887448787689209, + "learning_rate": 0.0001946731154071606, + "loss": 0.3586, + "step": 4461 + }, + { + "epoch": 0.10459630677988947, + "grad_norm": 4.325528621673584, + "learning_rate": 0.0001946707430630941, + "loss": 0.5639, + "step": 4462 + }, + { + "epoch": 0.10461974835469447, + "grad_norm": 5.576358795166016, + "learning_rate": 0.00019466837020534148, + "loss": 0.6042, + "step": 4463 + }, + { + "epoch": 0.10464318992949946, + "grad_norm": 5.646732330322266, + "learning_rate": 0.00019466599683391563, + "loss": 0.4836, + "step": 4464 + }, + { + "epoch": 0.10466663150430446, + "grad_norm": 1.8294464349746704, + "learning_rate": 0.0001946636229488294, + "loss": 0.3147, + "step": 4465 + }, + { + "epoch": 0.10469007307910945, + "grad_norm": 0.711860716342926, + "learning_rate": 0.00019466124855009568, + "loss": 0.0974, + "step": 4466 + }, + { + "epoch": 0.10471351465391444, + "grad_norm": 2.876539945602417, + "learning_rate": 0.0001946588736377274, + "loss": 0.3588, + "step": 4467 + }, + { + "epoch": 0.10473695622871945, + "grad_norm": 3.954519748687744, + "learning_rate": 0.0001946564982117374, + "loss": 0.5629, + "step": 4468 + }, + { + "epoch": 0.10476039780352445, + "grad_norm": 1.3044356107711792, + "learning_rate": 0.00019465412227213856, + "loss": 0.1747, + "step": 4469 + }, + { + "epoch": 0.10478383937832944, + "grad_norm": 3.67315936088562, + "learning_rate": 0.0001946517458189438, + "loss": 0.5494, + "step": 4470 + }, + { + "epoch": 0.10480728095313444, + "grad_norm": 4.266097068786621, + "learning_rate": 0.00019464936885216604, + "loss": 0.669, + "step": 4471 + }, + { + "epoch": 0.10483072252793943, + "grad_norm": 3.2501301765441895, + "learning_rate": 0.0001946469913718181, + "loss": 0.3539, + "step": 4472 + }, + { + "epoch": 0.10485416410274442, + "grad_norm": 1.923202395439148, + "learning_rate": 0.00019464461337791294, + "loss": 0.2218, + "step": 4473 + }, + { + "epoch": 0.10487760567754942, + "grad_norm": 1.3074671030044556, + "learning_rate": 0.00019464223487046343, + "loss": 0.6064, + "step": 4474 + }, + { + "epoch": 0.10490104725235441, + "grad_norm": 4.282883167266846, + "learning_rate": 0.00019463985584948252, + "loss": 0.3515, + "step": 4475 + }, + { + "epoch": 0.1049244888271594, + "grad_norm": 1.2051419019699097, + "learning_rate": 0.0001946374763149831, + "loss": 0.1261, + "step": 4476 + }, + { + "epoch": 0.1049479304019644, + "grad_norm": 4.8745245933532715, + "learning_rate": 0.00019463509626697803, + "loss": 0.6228, + "step": 4477 + }, + { + "epoch": 0.1049713719767694, + "grad_norm": 0.9530338048934937, + "learning_rate": 0.0001946327157054803, + "loss": 0.1944, + "step": 4478 + }, + { + "epoch": 0.10499481355157439, + "grad_norm": 3.134186029434204, + "learning_rate": 0.00019463033463050278, + "loss": 0.4013, + "step": 4479 + }, + { + "epoch": 0.10501825512637938, + "grad_norm": 2.752563953399658, + "learning_rate": 0.00019462795304205842, + "loss": 0.4721, + "step": 4480 + }, + { + "epoch": 0.10504169670118439, + "grad_norm": 1.174522042274475, + "learning_rate": 0.00019462557094016009, + "loss": 0.2326, + "step": 4481 + }, + { + "epoch": 0.10506513827598939, + "grad_norm": 2.3609731197357178, + "learning_rate": 0.00019462318832482077, + "loss": 1.3474, + "step": 4482 + }, + { + "epoch": 0.10508857985079438, + "grad_norm": 5.016956806182861, + "learning_rate": 0.00019462080519605336, + "loss": 0.6997, + "step": 4483 + }, + { + "epoch": 0.10511202142559938, + "grad_norm": 3.9755561351776123, + "learning_rate": 0.00019461842155387083, + "loss": 0.7501, + "step": 4484 + }, + { + "epoch": 0.10513546300040437, + "grad_norm": 4.18734884262085, + "learning_rate": 0.00019461603739828607, + "loss": 0.4508, + "step": 4485 + }, + { + "epoch": 0.10515890457520936, + "grad_norm": 1.9689347743988037, + "learning_rate": 0.000194613652729312, + "loss": 1.5987, + "step": 4486 + }, + { + "epoch": 0.10518234615001436, + "grad_norm": 3.1462318897247314, + "learning_rate": 0.0001946112675469616, + "loss": 0.5007, + "step": 4487 + }, + { + "epoch": 0.10520578772481935, + "grad_norm": 0.5465784072875977, + "learning_rate": 0.00019460888185124785, + "loss": 0.0617, + "step": 4488 + }, + { + "epoch": 0.10522922929962435, + "grad_norm": 3.57413911819458, + "learning_rate": 0.00019460649564218363, + "loss": 0.5508, + "step": 4489 + }, + { + "epoch": 0.10525267087442934, + "grad_norm": 3.8611488342285156, + "learning_rate": 0.00019460410891978188, + "loss": 0.3564, + "step": 4490 + }, + { + "epoch": 0.10527611244923434, + "grad_norm": 2.7961814403533936, + "learning_rate": 0.00019460172168405556, + "loss": 0.3562, + "step": 4491 + }, + { + "epoch": 0.10529955402403933, + "grad_norm": 2.0203466415405273, + "learning_rate": 0.0001945993339350177, + "loss": 0.1856, + "step": 4492 + }, + { + "epoch": 0.10532299559884432, + "grad_norm": 4.88480806350708, + "learning_rate": 0.00019459694567268115, + "loss": 0.6711, + "step": 4493 + }, + { + "epoch": 0.10534643717364933, + "grad_norm": 3.9526073932647705, + "learning_rate": 0.0001945945568970589, + "loss": 0.4503, + "step": 4494 + }, + { + "epoch": 0.10536987874845433, + "grad_norm": 3.2797513008117676, + "learning_rate": 0.00019459216760816394, + "loss": 0.64, + "step": 4495 + }, + { + "epoch": 0.10539332032325932, + "grad_norm": 3.662564992904663, + "learning_rate": 0.00019458977780600925, + "loss": 0.2177, + "step": 4496 + }, + { + "epoch": 0.10541676189806432, + "grad_norm": 4.315327167510986, + "learning_rate": 0.00019458738749060771, + "loss": 0.5684, + "step": 4497 + }, + { + "epoch": 0.10544020347286931, + "grad_norm": 3.1021196842193604, + "learning_rate": 0.00019458499666197239, + "loss": 0.4914, + "step": 4498 + }, + { + "epoch": 0.1054636450476743, + "grad_norm": 2.0620410442352295, + "learning_rate": 0.00019458260532011622, + "loss": 0.3107, + "step": 4499 + }, + { + "epoch": 0.1054870866224793, + "grad_norm": 3.4076271057128906, + "learning_rate": 0.00019458021346505216, + "loss": 0.3728, + "step": 4500 + }, + { + "epoch": 0.10551052819728429, + "grad_norm": 2.911306381225586, + "learning_rate": 0.00019457782109679318, + "loss": 0.4137, + "step": 4501 + }, + { + "epoch": 0.10553396977208929, + "grad_norm": 3.2432079315185547, + "learning_rate": 0.00019457542821535234, + "loss": 0.4515, + "step": 4502 + }, + { + "epoch": 0.10555741134689428, + "grad_norm": 2.5391347408294678, + "learning_rate": 0.00019457303482074253, + "loss": 0.2253, + "step": 4503 + }, + { + "epoch": 0.10558085292169928, + "grad_norm": 0.40954679250717163, + "learning_rate": 0.00019457064091297675, + "loss": 0.0638, + "step": 4504 + }, + { + "epoch": 0.10560429449650427, + "grad_norm": 3.8162059783935547, + "learning_rate": 0.00019456824649206805, + "loss": 0.572, + "step": 4505 + }, + { + "epoch": 0.10562773607130926, + "grad_norm": 2.7755959033966064, + "learning_rate": 0.00019456585155802936, + "loss": 0.2754, + "step": 4506 + }, + { + "epoch": 0.10565117764611427, + "grad_norm": 0.9107025265693665, + "learning_rate": 0.00019456345611087373, + "loss": 0.1804, + "step": 4507 + }, + { + "epoch": 0.10567461922091927, + "grad_norm": 3.3354949951171875, + "learning_rate": 0.0001945610601506141, + "loss": 0.4685, + "step": 4508 + }, + { + "epoch": 0.10569806079572426, + "grad_norm": 1.0984504222869873, + "learning_rate": 0.0001945586636772635, + "loss": 0.0988, + "step": 4509 + }, + { + "epoch": 0.10572150237052926, + "grad_norm": 2.174159526824951, + "learning_rate": 0.00019455626669083497, + "loss": 0.3043, + "step": 4510 + }, + { + "epoch": 0.10574494394533425, + "grad_norm": 2.8404433727264404, + "learning_rate": 0.00019455386919134144, + "loss": 0.369, + "step": 4511 + }, + { + "epoch": 0.10576838552013924, + "grad_norm": 5.731682300567627, + "learning_rate": 0.000194551471178796, + "loss": 0.8023, + "step": 4512 + }, + { + "epoch": 0.10579182709494424, + "grad_norm": 3.941908359527588, + "learning_rate": 0.00019454907265321157, + "loss": 0.7727, + "step": 4513 + }, + { + "epoch": 0.10581526866974923, + "grad_norm": 3.0890426635742188, + "learning_rate": 0.00019454667361460125, + "loss": 0.3973, + "step": 4514 + }, + { + "epoch": 0.10583871024455423, + "grad_norm": 1.5757752656936646, + "learning_rate": 0.00019454427406297796, + "loss": 0.2145, + "step": 4515 + }, + { + "epoch": 0.10586215181935922, + "grad_norm": 3.543224573135376, + "learning_rate": 0.00019454187399835484, + "loss": 0.555, + "step": 4516 + }, + { + "epoch": 0.10588559339416422, + "grad_norm": 1.7055599689483643, + "learning_rate": 0.00019453947342074483, + "loss": 0.9024, + "step": 4517 + }, + { + "epoch": 0.10590903496896921, + "grad_norm": 0.824569582939148, + "learning_rate": 0.00019453707233016098, + "loss": 0.154, + "step": 4518 + }, + { + "epoch": 0.1059324765437742, + "grad_norm": 1.5696368217468262, + "learning_rate": 0.00019453467072661634, + "loss": 0.1885, + "step": 4519 + }, + { + "epoch": 0.10595591811857921, + "grad_norm": 0.5355730652809143, + "learning_rate": 0.0001945322686101239, + "loss": 0.0737, + "step": 4520 + }, + { + "epoch": 0.1059793596933842, + "grad_norm": 3.5047028064727783, + "learning_rate": 0.00019452986598069672, + "loss": 0.816, + "step": 4521 + }, + { + "epoch": 0.1060028012681892, + "grad_norm": 3.286073684692383, + "learning_rate": 0.0001945274628383478, + "loss": 1.1278, + "step": 4522 + }, + { + "epoch": 0.1060262428429942, + "grad_norm": 0.4326600730419159, + "learning_rate": 0.00019452505918309023, + "loss": 0.0393, + "step": 4523 + }, + { + "epoch": 0.10604968441779919, + "grad_norm": 2.987520217895508, + "learning_rate": 0.00019452265501493704, + "loss": 0.4093, + "step": 4524 + }, + { + "epoch": 0.10607312599260418, + "grad_norm": 3.942842960357666, + "learning_rate": 0.00019452025033390126, + "loss": 0.7236, + "step": 4525 + }, + { + "epoch": 0.10609656756740918, + "grad_norm": 0.9953356385231018, + "learning_rate": 0.00019451784513999594, + "loss": 0.1456, + "step": 4526 + }, + { + "epoch": 0.10612000914221417, + "grad_norm": 3.158226490020752, + "learning_rate": 0.00019451543943323416, + "loss": 0.4177, + "step": 4527 + }, + { + "epoch": 0.10614345071701917, + "grad_norm": 1.5807055234909058, + "learning_rate": 0.0001945130332136289, + "loss": 0.2139, + "step": 4528 + }, + { + "epoch": 0.10616689229182416, + "grad_norm": 2.791921854019165, + "learning_rate": 0.0001945106264811933, + "loss": 0.5311, + "step": 4529 + }, + { + "epoch": 0.10619033386662916, + "grad_norm": 2.983647108078003, + "learning_rate": 0.00019450821923594038, + "loss": 0.3624, + "step": 4530 + }, + { + "epoch": 0.10621377544143415, + "grad_norm": 3.144946336746216, + "learning_rate": 0.0001945058114778832, + "loss": 0.5103, + "step": 4531 + }, + { + "epoch": 0.10623721701623916, + "grad_norm": 3.3644158840179443, + "learning_rate": 0.00019450340320703482, + "loss": 0.6338, + "step": 4532 + }, + { + "epoch": 0.10626065859104415, + "grad_norm": 8.470863342285156, + "learning_rate": 0.00019450099442340832, + "loss": 0.6152, + "step": 4533 + }, + { + "epoch": 0.10628410016584915, + "grad_norm": 4.312603950500488, + "learning_rate": 0.00019449858512701678, + "loss": 0.4566, + "step": 4534 + }, + { + "epoch": 0.10630754174065414, + "grad_norm": 4.901691913604736, + "learning_rate": 0.00019449617531787326, + "loss": 0.8385, + "step": 4535 + }, + { + "epoch": 0.10633098331545913, + "grad_norm": 2.8020479679107666, + "learning_rate": 0.00019449376499599083, + "loss": 0.4038, + "step": 4536 + }, + { + "epoch": 0.10635442489026413, + "grad_norm": 5.009662628173828, + "learning_rate": 0.00019449135416138254, + "loss": 0.7303, + "step": 4537 + }, + { + "epoch": 0.10637786646506912, + "grad_norm": 3.288668155670166, + "learning_rate": 0.00019448894281406156, + "loss": 0.27, + "step": 4538 + }, + { + "epoch": 0.10640130803987412, + "grad_norm": 1.181962490081787, + "learning_rate": 0.00019448653095404091, + "loss": 0.1043, + "step": 4539 + }, + { + "epoch": 0.10642474961467911, + "grad_norm": 2.1907970905303955, + "learning_rate": 0.00019448411858133367, + "loss": 2.0723, + "step": 4540 + }, + { + "epoch": 0.1064481911894841, + "grad_norm": 4.311917304992676, + "learning_rate": 0.00019448170569595294, + "loss": 0.4329, + "step": 4541 + }, + { + "epoch": 0.1064716327642891, + "grad_norm": 5.509061336517334, + "learning_rate": 0.00019447929229791182, + "loss": 0.6937, + "step": 4542 + }, + { + "epoch": 0.1064950743390941, + "grad_norm": 2.710225820541382, + "learning_rate": 0.0001944768783872234, + "loss": 0.3569, + "step": 4543 + }, + { + "epoch": 0.10651851591389909, + "grad_norm": 0.734624445438385, + "learning_rate": 0.0001944744639639008, + "loss": 0.0938, + "step": 4544 + }, + { + "epoch": 0.1065419574887041, + "grad_norm": 4.026118278503418, + "learning_rate": 0.0001944720490279571, + "loss": 0.5532, + "step": 4545 + }, + { + "epoch": 0.10656539906350909, + "grad_norm": 1.8776347637176514, + "learning_rate": 0.0001944696335794054, + "loss": 2.303, + "step": 4546 + }, + { + "epoch": 0.10658884063831409, + "grad_norm": 4.13695764541626, + "learning_rate": 0.00019446721761825884, + "loss": 0.5436, + "step": 4547 + }, + { + "epoch": 0.10661228221311908, + "grad_norm": 1.5934412479400635, + "learning_rate": 0.00019446480114453045, + "loss": 0.2483, + "step": 4548 + }, + { + "epoch": 0.10663572378792407, + "grad_norm": 6.401281356811523, + "learning_rate": 0.00019446238415823342, + "loss": 0.4934, + "step": 4549 + }, + { + "epoch": 0.10665916536272907, + "grad_norm": 4.614541053771973, + "learning_rate": 0.00019445996665938084, + "loss": 0.6327, + "step": 4550 + }, + { + "epoch": 0.10668260693753406, + "grad_norm": 5.2608819007873535, + "learning_rate": 0.00019445754864798583, + "loss": 0.8492, + "step": 4551 + }, + { + "epoch": 0.10670604851233906, + "grad_norm": 3.1333467960357666, + "learning_rate": 0.00019445513012406146, + "loss": 0.4301, + "step": 4552 + }, + { + "epoch": 0.10672949008714405, + "grad_norm": 1.3045049905776978, + "learning_rate": 0.00019445271108762096, + "loss": 0.1796, + "step": 4553 + }, + { + "epoch": 0.10675293166194905, + "grad_norm": 0.4404686689376831, + "learning_rate": 0.00019445029153867735, + "loss": 0.0497, + "step": 4554 + }, + { + "epoch": 0.10677637323675404, + "grad_norm": 3.6873691082000732, + "learning_rate": 0.00019444787147724382, + "loss": 0.69, + "step": 4555 + }, + { + "epoch": 0.10679981481155904, + "grad_norm": 4.157344341278076, + "learning_rate": 0.00019444545090333348, + "loss": 0.6025, + "step": 4556 + }, + { + "epoch": 0.10682325638636403, + "grad_norm": 2.3160440921783447, + "learning_rate": 0.00019444302981695948, + "loss": 0.1663, + "step": 4557 + }, + { + "epoch": 0.10684669796116904, + "grad_norm": 2.932129144668579, + "learning_rate": 0.00019444060821813492, + "loss": 0.3785, + "step": 4558 + }, + { + "epoch": 0.10687013953597403, + "grad_norm": 3.8880741596221924, + "learning_rate": 0.00019443818610687298, + "loss": 0.5592, + "step": 4559 + }, + { + "epoch": 0.10689358111077903, + "grad_norm": 3.415672540664673, + "learning_rate": 0.0001944357634831868, + "loss": 0.5473, + "step": 4560 + }, + { + "epoch": 0.10691702268558402, + "grad_norm": 3.0003559589385986, + "learning_rate": 0.0001944333403470895, + "loss": 2.0116, + "step": 4561 + }, + { + "epoch": 0.10694046426038901, + "grad_norm": 3.927236318588257, + "learning_rate": 0.00019443091669859422, + "loss": 0.5722, + "step": 4562 + }, + { + "epoch": 0.10696390583519401, + "grad_norm": 2.9114632606506348, + "learning_rate": 0.00019442849253771416, + "loss": 2.1065, + "step": 4563 + }, + { + "epoch": 0.106987347409999, + "grad_norm": 1.3424534797668457, + "learning_rate": 0.00019442606786446243, + "loss": 0.2576, + "step": 4564 + }, + { + "epoch": 0.107010788984804, + "grad_norm": 1.5679043531417847, + "learning_rate": 0.0001944236426788522, + "loss": 0.2628, + "step": 4565 + }, + { + "epoch": 0.10703423055960899, + "grad_norm": 3.617549419403076, + "learning_rate": 0.00019442121698089664, + "loss": 0.3073, + "step": 4566 + }, + { + "epoch": 0.10705767213441399, + "grad_norm": 1.5112600326538086, + "learning_rate": 0.0001944187907706089, + "loss": 0.3549, + "step": 4567 + }, + { + "epoch": 0.10708111370921898, + "grad_norm": 1.8858658075332642, + "learning_rate": 0.00019441636404800213, + "loss": 0.278, + "step": 4568 + }, + { + "epoch": 0.10710455528402398, + "grad_norm": 1.449567198753357, + "learning_rate": 0.00019441393681308953, + "loss": 0.2658, + "step": 4569 + }, + { + "epoch": 0.10712799685882897, + "grad_norm": 0.9510715007781982, + "learning_rate": 0.00019441150906588425, + "loss": 0.0845, + "step": 4570 + }, + { + "epoch": 0.10715143843363398, + "grad_norm": 4.494024276733398, + "learning_rate": 0.00019440908080639946, + "loss": 0.6154, + "step": 4571 + }, + { + "epoch": 0.10717488000843897, + "grad_norm": 1.978559970855713, + "learning_rate": 0.00019440665203464836, + "loss": 0.1581, + "step": 4572 + }, + { + "epoch": 0.10719832158324397, + "grad_norm": 3.1369998455047607, + "learning_rate": 0.0001944042227506441, + "loss": 0.424, + "step": 4573 + }, + { + "epoch": 0.10722176315804896, + "grad_norm": 5.0218987464904785, + "learning_rate": 0.00019440179295439987, + "loss": 0.7877, + "step": 4574 + }, + { + "epoch": 0.10724520473285395, + "grad_norm": 4.2564496994018555, + "learning_rate": 0.00019439936264592885, + "loss": 0.536, + "step": 4575 + }, + { + "epoch": 0.10726864630765895, + "grad_norm": 3.1536202430725098, + "learning_rate": 0.00019439693182524424, + "loss": 0.4151, + "step": 4576 + }, + { + "epoch": 0.10729208788246394, + "grad_norm": 3.1658666133880615, + "learning_rate": 0.00019439450049235925, + "loss": 0.4981, + "step": 4577 + }, + { + "epoch": 0.10731552945726894, + "grad_norm": 1.4734041690826416, + "learning_rate": 0.000194392068647287, + "loss": 0.2248, + "step": 4578 + }, + { + "epoch": 0.10733897103207393, + "grad_norm": 1.3297489881515503, + "learning_rate": 0.00019438963629004076, + "loss": 0.2434, + "step": 4579 + }, + { + "epoch": 0.10736241260687893, + "grad_norm": 0.49598556756973267, + "learning_rate": 0.0001943872034206337, + "loss": 0.0871, + "step": 4580 + }, + { + "epoch": 0.10738585418168392, + "grad_norm": 2.0527729988098145, + "learning_rate": 0.00019438477003907902, + "loss": 0.4957, + "step": 4581 + }, + { + "epoch": 0.10740929575648892, + "grad_norm": 4.410436630249023, + "learning_rate": 0.00019438233614538995, + "loss": 0.6741, + "step": 4582 + }, + { + "epoch": 0.10743273733129392, + "grad_norm": 2.7142422199249268, + "learning_rate": 0.00019437990173957963, + "loss": 0.3042, + "step": 4583 + }, + { + "epoch": 0.10745617890609892, + "grad_norm": 3.997013568878174, + "learning_rate": 0.00019437746682166132, + "loss": 0.8002, + "step": 4584 + }, + { + "epoch": 0.10747962048090391, + "grad_norm": 3.1534855365753174, + "learning_rate": 0.0001943750313916482, + "loss": 0.5271, + "step": 4585 + }, + { + "epoch": 0.1075030620557089, + "grad_norm": 3.32224702835083, + "learning_rate": 0.00019437259544955352, + "loss": 0.4353, + "step": 4586 + }, + { + "epoch": 0.1075265036305139, + "grad_norm": 4.206940650939941, + "learning_rate": 0.00019437015899539053, + "loss": 0.6844, + "step": 4587 + }, + { + "epoch": 0.1075499452053189, + "grad_norm": 0.5070697069168091, + "learning_rate": 0.00019436772202917235, + "loss": 0.0691, + "step": 4588 + }, + { + "epoch": 0.10757338678012389, + "grad_norm": 2.8335680961608887, + "learning_rate": 0.00019436528455091227, + "loss": 0.2003, + "step": 4589 + }, + { + "epoch": 0.10759682835492888, + "grad_norm": 3.373429775238037, + "learning_rate": 0.0001943628465606235, + "loss": 0.302, + "step": 4590 + }, + { + "epoch": 0.10762026992973388, + "grad_norm": 3.81118106842041, + "learning_rate": 0.0001943604080583193, + "loss": 0.3887, + "step": 4591 + }, + { + "epoch": 0.10764371150453887, + "grad_norm": 1.3534821271896362, + "learning_rate": 0.00019435796904401285, + "loss": 0.1778, + "step": 4592 + }, + { + "epoch": 0.10766715307934387, + "grad_norm": 1.4148623943328857, + "learning_rate": 0.00019435552951771738, + "loss": 0.1665, + "step": 4593 + }, + { + "epoch": 0.10769059465414886, + "grad_norm": 3.5716135501861572, + "learning_rate": 0.00019435308947944618, + "loss": 0.4973, + "step": 4594 + }, + { + "epoch": 0.10771403622895385, + "grad_norm": 4.018433570861816, + "learning_rate": 0.00019435064892921246, + "loss": 0.6319, + "step": 4595 + }, + { + "epoch": 0.10773747780375886, + "grad_norm": 3.1589243412017822, + "learning_rate": 0.00019434820786702947, + "loss": 1.5576, + "step": 4596 + }, + { + "epoch": 0.10776091937856386, + "grad_norm": 1.7206850051879883, + "learning_rate": 0.00019434576629291046, + "loss": 0.1744, + "step": 4597 + }, + { + "epoch": 0.10778436095336885, + "grad_norm": 2.2582666873931885, + "learning_rate": 0.00019434332420686866, + "loss": 0.2479, + "step": 4598 + }, + { + "epoch": 0.10780780252817385, + "grad_norm": 3.8683829307556152, + "learning_rate": 0.0001943408816089173, + "loss": 0.6581, + "step": 4599 + }, + { + "epoch": 0.10783124410297884, + "grad_norm": 4.660354137420654, + "learning_rate": 0.00019433843849906972, + "loss": 0.5659, + "step": 4600 + }, + { + "epoch": 0.10785468567778383, + "grad_norm": 3.998892068862915, + "learning_rate": 0.00019433599487733908, + "loss": 0.5205, + "step": 4601 + }, + { + "epoch": 0.10787812725258883, + "grad_norm": 3.516840696334839, + "learning_rate": 0.0001943335507437387, + "loss": 0.3539, + "step": 4602 + }, + { + "epoch": 0.10790156882739382, + "grad_norm": 3.615819215774536, + "learning_rate": 0.00019433110609828183, + "loss": 0.6437, + "step": 4603 + }, + { + "epoch": 0.10792501040219882, + "grad_norm": 1.990490198135376, + "learning_rate": 0.0001943286609409817, + "loss": 0.3436, + "step": 4604 + }, + { + "epoch": 0.10794845197700381, + "grad_norm": 2.0033047199249268, + "learning_rate": 0.0001943262152718516, + "loss": 0.2771, + "step": 4605 + }, + { + "epoch": 0.1079718935518088, + "grad_norm": 4.1904401779174805, + "learning_rate": 0.00019432376909090483, + "loss": 0.4284, + "step": 4606 + }, + { + "epoch": 0.1079953351266138, + "grad_norm": 2.178443670272827, + "learning_rate": 0.00019432132239815463, + "loss": 0.2789, + "step": 4607 + }, + { + "epoch": 0.1080187767014188, + "grad_norm": 0.6703108549118042, + "learning_rate": 0.00019431887519361427, + "loss": 0.106, + "step": 4608 + }, + { + "epoch": 0.1080422182762238, + "grad_norm": 3.1210596561431885, + "learning_rate": 0.00019431642747729706, + "loss": 1.5141, + "step": 4609 + }, + { + "epoch": 0.1080656598510288, + "grad_norm": 6.5374932289123535, + "learning_rate": 0.00019431397924921624, + "loss": 0.9763, + "step": 4610 + }, + { + "epoch": 0.10808910142583379, + "grad_norm": 4.8021111488342285, + "learning_rate": 0.00019431153050938514, + "loss": 0.7326, + "step": 4611 + }, + { + "epoch": 0.10811254300063879, + "grad_norm": 2.089214563369751, + "learning_rate": 0.000194309081257817, + "loss": 0.2132, + "step": 4612 + }, + { + "epoch": 0.10813598457544378, + "grad_norm": 5.6426897048950195, + "learning_rate": 0.00019430663149452512, + "loss": 0.9567, + "step": 4613 + }, + { + "epoch": 0.10815942615024877, + "grad_norm": 1.37900972366333, + "learning_rate": 0.00019430418121952286, + "loss": 0.1058, + "step": 4614 + }, + { + "epoch": 0.10818286772505377, + "grad_norm": 3.6029744148254395, + "learning_rate": 0.00019430173043282343, + "loss": 0.3904, + "step": 4615 + }, + { + "epoch": 0.10820630929985876, + "grad_norm": 0.8917244672775269, + "learning_rate": 0.00019429927913444014, + "loss": 0.1246, + "step": 4616 + }, + { + "epoch": 0.10822975087466376, + "grad_norm": 3.0992751121520996, + "learning_rate": 0.00019429682732438635, + "loss": 0.3867, + "step": 4617 + }, + { + "epoch": 0.10825319244946875, + "grad_norm": 1.0497921705245972, + "learning_rate": 0.00019429437500267528, + "loss": 0.1097, + "step": 4618 + }, + { + "epoch": 0.10827663402427375, + "grad_norm": 2.1302244663238525, + "learning_rate": 0.00019429192216932032, + "loss": 0.4086, + "step": 4619 + }, + { + "epoch": 0.10830007559907874, + "grad_norm": 3.5771121978759766, + "learning_rate": 0.0001942894688243347, + "loss": 0.4617, + "step": 4620 + }, + { + "epoch": 0.10832351717388373, + "grad_norm": 1.4791687726974487, + "learning_rate": 0.00019428701496773183, + "loss": 0.2371, + "step": 4621 + }, + { + "epoch": 0.10834695874868874, + "grad_norm": 4.991132736206055, + "learning_rate": 0.00019428456059952492, + "loss": 0.5517, + "step": 4622 + }, + { + "epoch": 0.10837040032349374, + "grad_norm": 4.545502185821533, + "learning_rate": 0.00019428210571972735, + "loss": 0.6258, + "step": 4623 + }, + { + "epoch": 0.10839384189829873, + "grad_norm": 3.272508382797241, + "learning_rate": 0.0001942796503283524, + "loss": 0.3312, + "step": 4624 + }, + { + "epoch": 0.10841728347310373, + "grad_norm": 2.6390163898468018, + "learning_rate": 0.00019427719442541343, + "loss": 0.2458, + "step": 4625 + }, + { + "epoch": 0.10844072504790872, + "grad_norm": 1.8613133430480957, + "learning_rate": 0.00019427473801092378, + "loss": 1.6987, + "step": 4626 + }, + { + "epoch": 0.10846416662271371, + "grad_norm": 2.4081006050109863, + "learning_rate": 0.0001942722810848967, + "loss": 0.4174, + "step": 4627 + }, + { + "epoch": 0.10848760819751871, + "grad_norm": 4.298653602600098, + "learning_rate": 0.00019426982364734562, + "loss": 1.0236, + "step": 4628 + }, + { + "epoch": 0.1085110497723237, + "grad_norm": 2.9146108627319336, + "learning_rate": 0.00019426736569828378, + "loss": 0.5404, + "step": 4629 + }, + { + "epoch": 0.1085344913471287, + "grad_norm": 4.04832124710083, + "learning_rate": 0.0001942649072377246, + "loss": 0.529, + "step": 4630 + }, + { + "epoch": 0.10855793292193369, + "grad_norm": 1.6630549430847168, + "learning_rate": 0.00019426244826568136, + "loss": 1.0171, + "step": 4631 + }, + { + "epoch": 0.10858137449673869, + "grad_norm": 3.5287296772003174, + "learning_rate": 0.00019425998878216746, + "loss": 0.5889, + "step": 4632 + }, + { + "epoch": 0.10860481607154368, + "grad_norm": 3.1063451766967773, + "learning_rate": 0.00019425752878719617, + "loss": 0.2278, + "step": 4633 + }, + { + "epoch": 0.10862825764634867, + "grad_norm": 1.9254578351974487, + "learning_rate": 0.00019425506828078088, + "loss": 0.1424, + "step": 4634 + }, + { + "epoch": 0.10865169922115368, + "grad_norm": 2.4408116340637207, + "learning_rate": 0.00019425260726293496, + "loss": 1.1791, + "step": 4635 + }, + { + "epoch": 0.10867514079595868, + "grad_norm": 0.9414684176445007, + "learning_rate": 0.00019425014573367174, + "loss": 0.1357, + "step": 4636 + }, + { + "epoch": 0.10869858237076367, + "grad_norm": 4.05545711517334, + "learning_rate": 0.00019424768369300457, + "loss": 0.5143, + "step": 4637 + }, + { + "epoch": 0.10872202394556867, + "grad_norm": 3.4019200801849365, + "learning_rate": 0.00019424522114094684, + "loss": 0.4074, + "step": 4638 + }, + { + "epoch": 0.10874546552037366, + "grad_norm": 7.605015277862549, + "learning_rate": 0.00019424275807751186, + "loss": 0.7407, + "step": 4639 + }, + { + "epoch": 0.10876890709517865, + "grad_norm": 3.6271560192108154, + "learning_rate": 0.00019424029450271303, + "loss": 0.3743, + "step": 4640 + }, + { + "epoch": 0.10879234866998365, + "grad_norm": 5.103428363800049, + "learning_rate": 0.00019423783041656374, + "loss": 0.8048, + "step": 4641 + }, + { + "epoch": 0.10881579024478864, + "grad_norm": 2.626757860183716, + "learning_rate": 0.00019423536581907727, + "loss": 0.3863, + "step": 4642 + }, + { + "epoch": 0.10883923181959364, + "grad_norm": 7.595884323120117, + "learning_rate": 0.00019423290071026713, + "loss": 0.8052, + "step": 4643 + }, + { + "epoch": 0.10886267339439863, + "grad_norm": 1.1639875173568726, + "learning_rate": 0.00019423043509014657, + "loss": 0.1459, + "step": 4644 + }, + { + "epoch": 0.10888611496920363, + "grad_norm": 0.4647503197193146, + "learning_rate": 0.00019422796895872906, + "loss": 0.0414, + "step": 4645 + }, + { + "epoch": 0.10890955654400862, + "grad_norm": 2.875582456588745, + "learning_rate": 0.00019422550231602793, + "loss": 1.5565, + "step": 4646 + }, + { + "epoch": 0.10893299811881363, + "grad_norm": 2.3699867725372314, + "learning_rate": 0.00019422303516205655, + "loss": 1.3, + "step": 4647 + }, + { + "epoch": 0.10895643969361862, + "grad_norm": 4.81992244720459, + "learning_rate": 0.00019422056749682836, + "loss": 0.6012, + "step": 4648 + }, + { + "epoch": 0.10897988126842362, + "grad_norm": 2.9745430946350098, + "learning_rate": 0.0001942180993203567, + "loss": 0.3142, + "step": 4649 + }, + { + "epoch": 0.10900332284322861, + "grad_norm": 3.7824535369873047, + "learning_rate": 0.000194215630632655, + "loss": 0.5977, + "step": 4650 + }, + { + "epoch": 0.1090267644180336, + "grad_norm": 4.857028484344482, + "learning_rate": 0.00019421316143373662, + "loss": 0.3467, + "step": 4651 + }, + { + "epoch": 0.1090502059928386, + "grad_norm": 0.44682684540748596, + "learning_rate": 0.000194210691723615, + "loss": 0.0452, + "step": 4652 + }, + { + "epoch": 0.1090736475676436, + "grad_norm": 2.7070281505584717, + "learning_rate": 0.0001942082215023035, + "loss": 0.1817, + "step": 4653 + }, + { + "epoch": 0.10909708914244859, + "grad_norm": 4.768310546875, + "learning_rate": 0.00019420575076981554, + "loss": 0.8961, + "step": 4654 + }, + { + "epoch": 0.10912053071725358, + "grad_norm": 4.909120082855225, + "learning_rate": 0.00019420327952616453, + "loss": 0.7056, + "step": 4655 + }, + { + "epoch": 0.10914397229205858, + "grad_norm": 4.345457077026367, + "learning_rate": 0.00019420080777136388, + "loss": 0.3617, + "step": 4656 + }, + { + "epoch": 0.10916741386686357, + "grad_norm": 1.9122755527496338, + "learning_rate": 0.00019419833550542702, + "loss": 0.2281, + "step": 4657 + }, + { + "epoch": 0.10919085544166857, + "grad_norm": 3.8195273876190186, + "learning_rate": 0.00019419586272836732, + "loss": 0.6163, + "step": 4658 + }, + { + "epoch": 0.10921429701647356, + "grad_norm": 3.715640068054199, + "learning_rate": 0.00019419338944019822, + "loss": 0.5282, + "step": 4659 + }, + { + "epoch": 0.10923773859127857, + "grad_norm": 2.5492563247680664, + "learning_rate": 0.00019419091564093315, + "loss": 1.2688, + "step": 4660 + }, + { + "epoch": 0.10926118016608356, + "grad_norm": 0.6103790402412415, + "learning_rate": 0.00019418844133058553, + "loss": 0.0877, + "step": 4661 + }, + { + "epoch": 0.10928462174088856, + "grad_norm": 3.5431129932403564, + "learning_rate": 0.00019418596650916877, + "loss": 0.5419, + "step": 4662 + }, + { + "epoch": 0.10930806331569355, + "grad_norm": 5.066803932189941, + "learning_rate": 0.00019418349117669627, + "loss": 0.8855, + "step": 4663 + }, + { + "epoch": 0.10933150489049855, + "grad_norm": 2.6255860328674316, + "learning_rate": 0.00019418101533318156, + "loss": 0.3082, + "step": 4664 + }, + { + "epoch": 0.10935494646530354, + "grad_norm": 4.252445697784424, + "learning_rate": 0.00019417853897863796, + "loss": 0.5463, + "step": 4665 + }, + { + "epoch": 0.10937838804010853, + "grad_norm": 2.475870132446289, + "learning_rate": 0.00019417606211307898, + "loss": 1.7512, + "step": 4666 + }, + { + "epoch": 0.10940182961491353, + "grad_norm": 2.045170545578003, + "learning_rate": 0.00019417358473651806, + "loss": 0.3449, + "step": 4667 + }, + { + "epoch": 0.10942527118971852, + "grad_norm": 0.8280559778213501, + "learning_rate": 0.0001941711068489686, + "loss": 0.1179, + "step": 4668 + }, + { + "epoch": 0.10944871276452352, + "grad_norm": 4.327867031097412, + "learning_rate": 0.00019416862845044407, + "loss": 0.6957, + "step": 4669 + }, + { + "epoch": 0.10947215433932851, + "grad_norm": 2.892526865005493, + "learning_rate": 0.00019416614954095792, + "loss": 1.4803, + "step": 4670 + }, + { + "epoch": 0.1094955959141335, + "grad_norm": 2.661949872970581, + "learning_rate": 0.0001941636701205236, + "loss": 1.754, + "step": 4671 + }, + { + "epoch": 0.1095190374889385, + "grad_norm": 1.1246662139892578, + "learning_rate": 0.00019416119018915453, + "loss": 0.1622, + "step": 4672 + }, + { + "epoch": 0.10954247906374351, + "grad_norm": 4.208556175231934, + "learning_rate": 0.00019415870974686422, + "loss": 0.5921, + "step": 4673 + }, + { + "epoch": 0.1095659206385485, + "grad_norm": 2.376265525817871, + "learning_rate": 0.0001941562287936661, + "loss": 0.1498, + "step": 4674 + }, + { + "epoch": 0.1095893622133535, + "grad_norm": 3.8704631328582764, + "learning_rate": 0.00019415374732957363, + "loss": 0.4945, + "step": 4675 + }, + { + "epoch": 0.10961280378815849, + "grad_norm": 2.1673786640167236, + "learning_rate": 0.00019415126535460026, + "loss": 1.4285, + "step": 4676 + }, + { + "epoch": 0.10963624536296349, + "grad_norm": 2.470911979675293, + "learning_rate": 0.0001941487828687595, + "loss": 2.0729, + "step": 4677 + }, + { + "epoch": 0.10965968693776848, + "grad_norm": 1.0434554815292358, + "learning_rate": 0.0001941462998720648, + "loss": 0.142, + "step": 4678 + }, + { + "epoch": 0.10968312851257347, + "grad_norm": 1.191402792930603, + "learning_rate": 0.00019414381636452963, + "loss": 0.2365, + "step": 4679 + }, + { + "epoch": 0.10970657008737847, + "grad_norm": 1.3028502464294434, + "learning_rate": 0.00019414133234616748, + "loss": 0.1548, + "step": 4680 + }, + { + "epoch": 0.10973001166218346, + "grad_norm": 2.9256081581115723, + "learning_rate": 0.0001941388478169918, + "loss": 0.3304, + "step": 4681 + }, + { + "epoch": 0.10975345323698846, + "grad_norm": 2.1645162105560303, + "learning_rate": 0.00019413636277701605, + "loss": 0.3277, + "step": 4682 + }, + { + "epoch": 0.10977689481179345, + "grad_norm": 0.5096428394317627, + "learning_rate": 0.00019413387722625377, + "loss": 0.0884, + "step": 4683 + }, + { + "epoch": 0.10980033638659845, + "grad_norm": 2.9843404293060303, + "learning_rate": 0.00019413139116471846, + "loss": 0.3378, + "step": 4684 + }, + { + "epoch": 0.10982377796140344, + "grad_norm": 3.2125015258789062, + "learning_rate": 0.00019412890459242355, + "loss": 0.3286, + "step": 4685 + }, + { + "epoch": 0.10984721953620845, + "grad_norm": 6.550032138824463, + "learning_rate": 0.0001941264175093825, + "loss": 0.9101, + "step": 4686 + }, + { + "epoch": 0.10987066111101344, + "grad_norm": 2.663926839828491, + "learning_rate": 0.00019412392991560894, + "loss": 2.0155, + "step": 4687 + }, + { + "epoch": 0.10989410268581844, + "grad_norm": 1.617882251739502, + "learning_rate": 0.00019412144181111626, + "loss": 0.2442, + "step": 4688 + }, + { + "epoch": 0.10991754426062343, + "grad_norm": 1.2460960149765015, + "learning_rate": 0.000194118953195918, + "loss": 0.2382, + "step": 4689 + }, + { + "epoch": 0.10994098583542843, + "grad_norm": 3.8224964141845703, + "learning_rate": 0.00019411646407002767, + "loss": 0.7443, + "step": 4690 + }, + { + "epoch": 0.10996442741023342, + "grad_norm": 5.81940221786499, + "learning_rate": 0.00019411397443345874, + "loss": 0.6266, + "step": 4691 + }, + { + "epoch": 0.10998786898503841, + "grad_norm": 1.072910189628601, + "learning_rate": 0.00019411148428622474, + "loss": 0.1293, + "step": 4692 + }, + { + "epoch": 0.11001131055984341, + "grad_norm": 4.382613182067871, + "learning_rate": 0.00019410899362833916, + "loss": 0.6953, + "step": 4693 + }, + { + "epoch": 0.1100347521346484, + "grad_norm": 4.4215264320373535, + "learning_rate": 0.00019410650245981558, + "loss": 0.7317, + "step": 4694 + }, + { + "epoch": 0.1100581937094534, + "grad_norm": 3.942124366760254, + "learning_rate": 0.0001941040107806674, + "loss": 0.5638, + "step": 4695 + }, + { + "epoch": 0.11008163528425839, + "grad_norm": 1.665434718132019, + "learning_rate": 0.00019410151859090827, + "loss": 0.2268, + "step": 4696 + }, + { + "epoch": 0.11010507685906339, + "grad_norm": 0.838366687297821, + "learning_rate": 0.00019409902589055165, + "loss": 0.1214, + "step": 4697 + }, + { + "epoch": 0.11012851843386838, + "grad_norm": 3.9268481731414795, + "learning_rate": 0.00019409653267961105, + "loss": 0.5405, + "step": 4698 + }, + { + "epoch": 0.11015196000867339, + "grad_norm": 3.8268826007843018, + "learning_rate": 0.00019409403895810003, + "loss": 0.3311, + "step": 4699 + }, + { + "epoch": 0.11017540158347838, + "grad_norm": 5.821200847625732, + "learning_rate": 0.0001940915447260321, + "loss": 0.8906, + "step": 4700 + }, + { + "epoch": 0.11019884315828338, + "grad_norm": 1.3211886882781982, + "learning_rate": 0.00019408904998342082, + "loss": 0.1896, + "step": 4701 + }, + { + "epoch": 0.11022228473308837, + "grad_norm": 3.5227153301239014, + "learning_rate": 0.0001940865547302797, + "loss": 0.3775, + "step": 4702 + }, + { + "epoch": 0.11024572630789337, + "grad_norm": 2.830242156982422, + "learning_rate": 0.00019408405896662227, + "loss": 0.4781, + "step": 4703 + }, + { + "epoch": 0.11026916788269836, + "grad_norm": 3.729276657104492, + "learning_rate": 0.00019408156269246212, + "loss": 0.4705, + "step": 4704 + }, + { + "epoch": 0.11029260945750335, + "grad_norm": 2.9881930351257324, + "learning_rate": 0.00019407906590781276, + "loss": 0.5297, + "step": 4705 + }, + { + "epoch": 0.11031605103230835, + "grad_norm": 0.8602973222732544, + "learning_rate": 0.00019407656861268773, + "loss": 0.1571, + "step": 4706 + }, + { + "epoch": 0.11033949260711334, + "grad_norm": 2.6357059478759766, + "learning_rate": 0.00019407407080710058, + "loss": 0.4274, + "step": 4707 + }, + { + "epoch": 0.11036293418191834, + "grad_norm": 4.563599586486816, + "learning_rate": 0.00019407157249106492, + "loss": 0.7459, + "step": 4708 + }, + { + "epoch": 0.11038637575672333, + "grad_norm": 2.8869383335113525, + "learning_rate": 0.00019406907366459423, + "loss": 0.3997, + "step": 4709 + }, + { + "epoch": 0.11040981733152833, + "grad_norm": 6.28264045715332, + "learning_rate": 0.0001940665743277021, + "loss": 1.0633, + "step": 4710 + }, + { + "epoch": 0.11043325890633333, + "grad_norm": 4.746960639953613, + "learning_rate": 0.0001940640744804021, + "loss": 0.5054, + "step": 4711 + }, + { + "epoch": 0.11045670048113833, + "grad_norm": 1.3330200910568237, + "learning_rate": 0.0001940615741227078, + "loss": 0.2179, + "step": 4712 + }, + { + "epoch": 0.11048014205594332, + "grad_norm": 4.75632381439209, + "learning_rate": 0.00019405907325463275, + "loss": 0.3184, + "step": 4713 + }, + { + "epoch": 0.11050358363074832, + "grad_norm": 2.246796131134033, + "learning_rate": 0.00019405657187619051, + "loss": 1.6782, + "step": 4714 + }, + { + "epoch": 0.11052702520555331, + "grad_norm": 3.977586269378662, + "learning_rate": 0.0001940540699873947, + "loss": 0.4902, + "step": 4715 + }, + { + "epoch": 0.1105504667803583, + "grad_norm": 2.236563205718994, + "learning_rate": 0.00019405156758825883, + "loss": 1.274, + "step": 4716 + }, + { + "epoch": 0.1105739083551633, + "grad_norm": 1.3044465780258179, + "learning_rate": 0.0001940490646787965, + "loss": 0.2386, + "step": 4717 + }, + { + "epoch": 0.1105973499299683, + "grad_norm": 6.292466163635254, + "learning_rate": 0.00019404656125902132, + "loss": 0.3862, + "step": 4718 + }, + { + "epoch": 0.11062079150477329, + "grad_norm": 2.145906686782837, + "learning_rate": 0.00019404405732894686, + "loss": 0.4508, + "step": 4719 + }, + { + "epoch": 0.11064423307957828, + "grad_norm": 3.2148475646972656, + "learning_rate": 0.0001940415528885867, + "loss": 0.4705, + "step": 4720 + }, + { + "epoch": 0.11066767465438328, + "grad_norm": 0.9406640529632568, + "learning_rate": 0.0001940390479379544, + "loss": 0.1494, + "step": 4721 + }, + { + "epoch": 0.11069111622918827, + "grad_norm": 3.9866394996643066, + "learning_rate": 0.0001940365424770636, + "loss": 0.522, + "step": 4722 + }, + { + "epoch": 0.11071455780399327, + "grad_norm": 4.704040050506592, + "learning_rate": 0.00019403403650592788, + "loss": 0.7344, + "step": 4723 + }, + { + "epoch": 0.11073799937879827, + "grad_norm": 0.9622827768325806, + "learning_rate": 0.00019403153002456086, + "loss": 0.6485, + "step": 4724 + }, + { + "epoch": 0.11076144095360327, + "grad_norm": 3.4872963428497314, + "learning_rate": 0.00019402902303297605, + "loss": 0.4531, + "step": 4725 + }, + { + "epoch": 0.11078488252840826, + "grad_norm": 3.823617935180664, + "learning_rate": 0.00019402651553118716, + "loss": 0.5229, + "step": 4726 + }, + { + "epoch": 0.11080832410321326, + "grad_norm": 3.7733407020568848, + "learning_rate": 0.00019402400751920775, + "loss": 0.5627, + "step": 4727 + }, + { + "epoch": 0.11083176567801825, + "grad_norm": 4.370152473449707, + "learning_rate": 0.00019402149899705142, + "loss": 0.6185, + "step": 4728 + }, + { + "epoch": 0.11085520725282325, + "grad_norm": 5.048326015472412, + "learning_rate": 0.0001940189899647318, + "loss": 0.6148, + "step": 4729 + }, + { + "epoch": 0.11087864882762824, + "grad_norm": 3.201695680618286, + "learning_rate": 0.00019401648042226253, + "loss": 0.4007, + "step": 4730 + }, + { + "epoch": 0.11090209040243323, + "grad_norm": 3.792189836502075, + "learning_rate": 0.00019401397036965715, + "loss": 0.6217, + "step": 4731 + }, + { + "epoch": 0.11092553197723823, + "grad_norm": 3.1534996032714844, + "learning_rate": 0.00019401145980692933, + "loss": 0.3243, + "step": 4732 + }, + { + "epoch": 0.11094897355204322, + "grad_norm": 3.894463300704956, + "learning_rate": 0.0001940089487340927, + "loss": 0.5796, + "step": 4733 + }, + { + "epoch": 0.11097241512684822, + "grad_norm": 1.0123875141143799, + "learning_rate": 0.00019400643715116085, + "loss": 0.1658, + "step": 4734 + }, + { + "epoch": 0.11099585670165321, + "grad_norm": 2.200620174407959, + "learning_rate": 0.00019400392505814745, + "loss": 1.4599, + "step": 4735 + }, + { + "epoch": 0.1110192982764582, + "grad_norm": 3.7677130699157715, + "learning_rate": 0.0001940014124550661, + "loss": 0.6174, + "step": 4736 + }, + { + "epoch": 0.11104273985126321, + "grad_norm": 3.9698333740234375, + "learning_rate": 0.00019399889934193043, + "loss": 0.4002, + "step": 4737 + }, + { + "epoch": 0.11106618142606821, + "grad_norm": 0.6772447228431702, + "learning_rate": 0.00019399638571875413, + "loss": 0.0856, + "step": 4738 + }, + { + "epoch": 0.1110896230008732, + "grad_norm": 3.3111770153045654, + "learning_rate": 0.00019399387158555075, + "loss": 0.3742, + "step": 4739 + }, + { + "epoch": 0.1111130645756782, + "grad_norm": 4.04017448425293, + "learning_rate": 0.000193991356942334, + "loss": 0.6449, + "step": 4740 + }, + { + "epoch": 0.11113650615048319, + "grad_norm": 5.244414329528809, + "learning_rate": 0.0001939888417891175, + "loss": 0.7339, + "step": 4741 + }, + { + "epoch": 0.11115994772528819, + "grad_norm": 1.3967535495758057, + "learning_rate": 0.0001939863261259149, + "loss": 0.1792, + "step": 4742 + }, + { + "epoch": 0.11118338930009318, + "grad_norm": 1.3831822872161865, + "learning_rate": 0.0001939838099527399, + "loss": 0.1647, + "step": 4743 + }, + { + "epoch": 0.11120683087489817, + "grad_norm": 3.218700408935547, + "learning_rate": 0.00019398129326960603, + "loss": 0.5384, + "step": 4744 + }, + { + "epoch": 0.11123027244970317, + "grad_norm": 2.924917459487915, + "learning_rate": 0.00019397877607652702, + "loss": 0.4366, + "step": 4745 + }, + { + "epoch": 0.11125371402450816, + "grad_norm": 3.2811083793640137, + "learning_rate": 0.0001939762583735166, + "loss": 0.5079, + "step": 4746 + }, + { + "epoch": 0.11127715559931316, + "grad_norm": 3.363945245742798, + "learning_rate": 0.00019397374016058832, + "loss": 1.8579, + "step": 4747 + }, + { + "epoch": 0.11130059717411815, + "grad_norm": 3.4677867889404297, + "learning_rate": 0.00019397122143775588, + "loss": 0.4843, + "step": 4748 + }, + { + "epoch": 0.11132403874892315, + "grad_norm": 3.914205312728882, + "learning_rate": 0.00019396870220503296, + "loss": 0.635, + "step": 4749 + }, + { + "epoch": 0.11134748032372815, + "grad_norm": 0.9468575716018677, + "learning_rate": 0.00019396618246243322, + "loss": 0.123, + "step": 4750 + }, + { + "epoch": 0.11137092189853315, + "grad_norm": 3.4242444038391113, + "learning_rate": 0.0001939636622099703, + "loss": 0.3378, + "step": 4751 + }, + { + "epoch": 0.11139436347333814, + "grad_norm": 1.5683869123458862, + "learning_rate": 0.00019396114144765796, + "loss": 0.0658, + "step": 4752 + }, + { + "epoch": 0.11141780504814314, + "grad_norm": 4.0018415451049805, + "learning_rate": 0.00019395862017550979, + "loss": 0.3618, + "step": 4753 + }, + { + "epoch": 0.11144124662294813, + "grad_norm": 3.499089479446411, + "learning_rate": 0.00019395609839353953, + "loss": 0.2955, + "step": 4754 + }, + { + "epoch": 0.11146468819775313, + "grad_norm": 0.6246259808540344, + "learning_rate": 0.0001939535761017608, + "loss": 0.0608, + "step": 4755 + }, + { + "epoch": 0.11148812977255812, + "grad_norm": 3.6428403854370117, + "learning_rate": 0.00019395105330018736, + "loss": 0.7221, + "step": 4756 + }, + { + "epoch": 0.11151157134736311, + "grad_norm": 3.7597663402557373, + "learning_rate": 0.00019394852998883284, + "loss": 1.7604, + "step": 4757 + }, + { + "epoch": 0.11153501292216811, + "grad_norm": 2.971101760864258, + "learning_rate": 0.00019394600616771097, + "loss": 2.1691, + "step": 4758 + }, + { + "epoch": 0.1115584544969731, + "grad_norm": 3.9259161949157715, + "learning_rate": 0.00019394348183683543, + "loss": 0.3371, + "step": 4759 + }, + { + "epoch": 0.1115818960717781, + "grad_norm": 4.6689019203186035, + "learning_rate": 0.00019394095699621988, + "loss": 0.6022, + "step": 4760 + }, + { + "epoch": 0.11160533764658309, + "grad_norm": 4.5018486976623535, + "learning_rate": 0.0001939384316458781, + "loss": 0.6616, + "step": 4761 + }, + { + "epoch": 0.1116287792213881, + "grad_norm": 2.8381662368774414, + "learning_rate": 0.0001939359057858237, + "loss": 1.4339, + "step": 4762 + }, + { + "epoch": 0.1116522207961931, + "grad_norm": 4.016060829162598, + "learning_rate": 0.00019393337941607048, + "loss": 0.7771, + "step": 4763 + }, + { + "epoch": 0.11167566237099809, + "grad_norm": 0.7907111048698425, + "learning_rate": 0.0001939308525366321, + "loss": 0.1027, + "step": 4764 + }, + { + "epoch": 0.11169910394580308, + "grad_norm": 3.4349992275238037, + "learning_rate": 0.00019392832514752227, + "loss": 0.2685, + "step": 4765 + }, + { + "epoch": 0.11172254552060808, + "grad_norm": 2.6771764755249023, + "learning_rate": 0.0001939257972487547, + "loss": 1.3258, + "step": 4766 + }, + { + "epoch": 0.11174598709541307, + "grad_norm": 1.1621674299240112, + "learning_rate": 0.00019392326884034311, + "loss": 0.0703, + "step": 4767 + }, + { + "epoch": 0.11176942867021807, + "grad_norm": 1.0048761367797852, + "learning_rate": 0.00019392073992230123, + "loss": 0.0913, + "step": 4768 + }, + { + "epoch": 0.11179287024502306, + "grad_norm": 0.46406590938568115, + "learning_rate": 0.00019391821049464275, + "loss": 0.0892, + "step": 4769 + }, + { + "epoch": 0.11181631181982805, + "grad_norm": 3.2105908393859863, + "learning_rate": 0.00019391568055738146, + "loss": 0.3165, + "step": 4770 + }, + { + "epoch": 0.11183975339463305, + "grad_norm": 0.927761435508728, + "learning_rate": 0.000193913150110531, + "loss": 0.1354, + "step": 4771 + }, + { + "epoch": 0.11186319496943804, + "grad_norm": 5.064220428466797, + "learning_rate": 0.00019391061915410517, + "loss": 0.8527, + "step": 4772 + }, + { + "epoch": 0.11188663654424304, + "grad_norm": 3.2706921100616455, + "learning_rate": 0.0001939080876881177, + "loss": 0.4355, + "step": 4773 + }, + { + "epoch": 0.11191007811904803, + "grad_norm": 3.0479419231414795, + "learning_rate": 0.00019390555571258228, + "loss": 0.4739, + "step": 4774 + }, + { + "epoch": 0.11193351969385304, + "grad_norm": 3.6238954067230225, + "learning_rate": 0.00019390302322751268, + "loss": 0.6205, + "step": 4775 + }, + { + "epoch": 0.11195696126865803, + "grad_norm": 4.740664005279541, + "learning_rate": 0.00019390049023292264, + "loss": 0.8269, + "step": 4776 + }, + { + "epoch": 0.11198040284346303, + "grad_norm": 1.4834243059158325, + "learning_rate": 0.0001938979567288259, + "loss": 0.1508, + "step": 4777 + }, + { + "epoch": 0.11200384441826802, + "grad_norm": 5.727362632751465, + "learning_rate": 0.00019389542271523621, + "loss": 1.4017, + "step": 4778 + }, + { + "epoch": 0.11202728599307302, + "grad_norm": 4.031205177307129, + "learning_rate": 0.00019389288819216728, + "loss": 0.5777, + "step": 4779 + }, + { + "epoch": 0.11205072756787801, + "grad_norm": 4.232468128204346, + "learning_rate": 0.00019389035315963294, + "loss": 0.5948, + "step": 4780 + }, + { + "epoch": 0.112074169142683, + "grad_norm": 3.7182068824768066, + "learning_rate": 0.00019388781761764688, + "loss": 0.3998, + "step": 4781 + }, + { + "epoch": 0.112097610717488, + "grad_norm": 2.1681370735168457, + "learning_rate": 0.00019388528156622292, + "loss": 0.4025, + "step": 4782 + }, + { + "epoch": 0.112121052292293, + "grad_norm": 3.625161647796631, + "learning_rate": 0.00019388274500537476, + "loss": 0.543, + "step": 4783 + }, + { + "epoch": 0.11214449386709799, + "grad_norm": 4.30872106552124, + "learning_rate": 0.00019388020793511615, + "loss": 0.5293, + "step": 4784 + }, + { + "epoch": 0.11216793544190298, + "grad_norm": 1.2617061138153076, + "learning_rate": 0.00019387767035546096, + "loss": 0.0952, + "step": 4785 + }, + { + "epoch": 0.11219137701670798, + "grad_norm": 2.7843291759490967, + "learning_rate": 0.00019387513226642284, + "loss": 0.3477, + "step": 4786 + }, + { + "epoch": 0.11221481859151297, + "grad_norm": 7.08816385269165, + "learning_rate": 0.00019387259366801564, + "loss": 1.2495, + "step": 4787 + }, + { + "epoch": 0.11223826016631798, + "grad_norm": 1.4248144626617432, + "learning_rate": 0.00019387005456025308, + "loss": 0.1729, + "step": 4788 + }, + { + "epoch": 0.11226170174112297, + "grad_norm": 3.5130019187927246, + "learning_rate": 0.00019386751494314899, + "loss": 0.4616, + "step": 4789 + }, + { + "epoch": 0.11228514331592797, + "grad_norm": 3.044651746749878, + "learning_rate": 0.00019386497481671712, + "loss": 0.2366, + "step": 4790 + }, + { + "epoch": 0.11230858489073296, + "grad_norm": 3.8706111907958984, + "learning_rate": 0.00019386243418097125, + "loss": 0.4233, + "step": 4791 + }, + { + "epoch": 0.11233202646553796, + "grad_norm": 5.083907127380371, + "learning_rate": 0.00019385989303592519, + "loss": 0.367, + "step": 4792 + }, + { + "epoch": 0.11235546804034295, + "grad_norm": 5.321723937988281, + "learning_rate": 0.00019385735138159266, + "loss": 0.6495, + "step": 4793 + }, + { + "epoch": 0.11237890961514795, + "grad_norm": 3.477627754211426, + "learning_rate": 0.00019385480921798754, + "loss": 0.5999, + "step": 4794 + }, + { + "epoch": 0.11240235118995294, + "grad_norm": 1.2016377449035645, + "learning_rate": 0.0001938522665451236, + "loss": 0.2096, + "step": 4795 + }, + { + "epoch": 0.11242579276475793, + "grad_norm": 1.008474349975586, + "learning_rate": 0.0001938497233630146, + "loss": 0.13, + "step": 4796 + }, + { + "epoch": 0.11244923433956293, + "grad_norm": 3.034963369369507, + "learning_rate": 0.00019384717967167435, + "loss": 0.4397, + "step": 4797 + }, + { + "epoch": 0.11247267591436792, + "grad_norm": 1.747316837310791, + "learning_rate": 0.0001938446354711167, + "loss": 0.1809, + "step": 4798 + }, + { + "epoch": 0.11249611748917292, + "grad_norm": 3.8825414180755615, + "learning_rate": 0.00019384209076135539, + "loss": 0.3897, + "step": 4799 + }, + { + "epoch": 0.11251955906397791, + "grad_norm": 4.42098331451416, + "learning_rate": 0.00019383954554240429, + "loss": 0.7056, + "step": 4800 + }, + { + "epoch": 0.11254300063878292, + "grad_norm": 2.4228627681732178, + "learning_rate": 0.00019383699981427715, + "loss": 0.3863, + "step": 4801 + }, + { + "epoch": 0.11256644221358791, + "grad_norm": 2.8875069618225098, + "learning_rate": 0.0001938344535769878, + "loss": 0.2605, + "step": 4802 + }, + { + "epoch": 0.11258988378839291, + "grad_norm": 4.404762268066406, + "learning_rate": 0.00019383190683055013, + "loss": 0.6136, + "step": 4803 + }, + { + "epoch": 0.1126133253631979, + "grad_norm": 3.037381649017334, + "learning_rate": 0.00019382935957497783, + "loss": 0.31, + "step": 4804 + }, + { + "epoch": 0.1126367669380029, + "grad_norm": 0.8962299227714539, + "learning_rate": 0.00019382681181028482, + "loss": 0.1143, + "step": 4805 + }, + { + "epoch": 0.11266020851280789, + "grad_norm": 3.4492807388305664, + "learning_rate": 0.00019382426353648486, + "loss": 0.5081, + "step": 4806 + }, + { + "epoch": 0.11268365008761289, + "grad_norm": 6.9080305099487305, + "learning_rate": 0.00019382171475359184, + "loss": 0.7727, + "step": 4807 + }, + { + "epoch": 0.11270709166241788, + "grad_norm": 2.293320894241333, + "learning_rate": 0.00019381916546161953, + "loss": 0.3929, + "step": 4808 + }, + { + "epoch": 0.11273053323722287, + "grad_norm": 2.709681510925293, + "learning_rate": 0.0001938166156605818, + "loss": 0.2116, + "step": 4809 + }, + { + "epoch": 0.11275397481202787, + "grad_norm": 1.2922232151031494, + "learning_rate": 0.00019381406535049249, + "loss": 0.2133, + "step": 4810 + }, + { + "epoch": 0.11277741638683286, + "grad_norm": 4.693761825561523, + "learning_rate": 0.0001938115145313654, + "loss": 0.4862, + "step": 4811 + }, + { + "epoch": 0.11280085796163786, + "grad_norm": 4.582939147949219, + "learning_rate": 0.0001938089632032144, + "loss": 0.6832, + "step": 4812 + }, + { + "epoch": 0.11282429953644285, + "grad_norm": 1.7204355001449585, + "learning_rate": 0.00019380641136605334, + "loss": 0.2736, + "step": 4813 + }, + { + "epoch": 0.11284774111124786, + "grad_norm": 4.077214241027832, + "learning_rate": 0.00019380385901989606, + "loss": 0.7, + "step": 4814 + }, + { + "epoch": 0.11287118268605285, + "grad_norm": 1.0568678379058838, + "learning_rate": 0.00019380130616475637, + "loss": 0.1556, + "step": 4815 + }, + { + "epoch": 0.11289462426085785, + "grad_norm": 1.2476812601089478, + "learning_rate": 0.0001937987528006482, + "loss": 0.1896, + "step": 4816 + }, + { + "epoch": 0.11291806583566284, + "grad_norm": 3.5820696353912354, + "learning_rate": 0.00019379619892758533, + "loss": 0.5253, + "step": 4817 + }, + { + "epoch": 0.11294150741046784, + "grad_norm": 4.658504009246826, + "learning_rate": 0.00019379364454558167, + "loss": 0.5019, + "step": 4818 + }, + { + "epoch": 0.11296494898527283, + "grad_norm": 3.271056652069092, + "learning_rate": 0.00019379108965465104, + "loss": 0.2936, + "step": 4819 + }, + { + "epoch": 0.11298839056007783, + "grad_norm": 4.580488204956055, + "learning_rate": 0.00019378853425480733, + "loss": 0.4601, + "step": 4820 + }, + { + "epoch": 0.11301183213488282, + "grad_norm": 3.236931800842285, + "learning_rate": 0.00019378597834606437, + "loss": 0.5493, + "step": 4821 + }, + { + "epoch": 0.11303527370968781, + "grad_norm": 6.286581039428711, + "learning_rate": 0.0001937834219284361, + "loss": 0.9995, + "step": 4822 + }, + { + "epoch": 0.11305871528449281, + "grad_norm": 6.3536272048950195, + "learning_rate": 0.0001937808650019363, + "loss": 0.7357, + "step": 4823 + }, + { + "epoch": 0.1130821568592978, + "grad_norm": 2.9671103954315186, + "learning_rate": 0.00019377830756657892, + "loss": 1.8189, + "step": 4824 + }, + { + "epoch": 0.1131055984341028, + "grad_norm": 3.9765052795410156, + "learning_rate": 0.0001937757496223778, + "loss": 0.4221, + "step": 4825 + }, + { + "epoch": 0.1131290400089078, + "grad_norm": 2.952934503555298, + "learning_rate": 0.0001937731911693468, + "loss": 0.5542, + "step": 4826 + }, + { + "epoch": 0.1131524815837128, + "grad_norm": 2.912959098815918, + "learning_rate": 0.00019377063220749988, + "loss": 0.3994, + "step": 4827 + }, + { + "epoch": 0.1131759231585178, + "grad_norm": 1.0153518915176392, + "learning_rate": 0.0001937680727368508, + "loss": 0.1716, + "step": 4828 + }, + { + "epoch": 0.11319936473332279, + "grad_norm": 3.237086057662964, + "learning_rate": 0.00019376551275741358, + "loss": 0.4044, + "step": 4829 + }, + { + "epoch": 0.11322280630812778, + "grad_norm": 5.07950496673584, + "learning_rate": 0.00019376295226920202, + "loss": 0.7663, + "step": 4830 + }, + { + "epoch": 0.11324624788293278, + "grad_norm": 3.0521016120910645, + "learning_rate": 0.00019376039127223007, + "loss": 0.4812, + "step": 4831 + }, + { + "epoch": 0.11326968945773777, + "grad_norm": 1.3830927610397339, + "learning_rate": 0.0001937578297665116, + "loss": 0.1448, + "step": 4832 + }, + { + "epoch": 0.11329313103254277, + "grad_norm": 3.4559497833251953, + "learning_rate": 0.00019375526775206048, + "loss": 0.6621, + "step": 4833 + }, + { + "epoch": 0.11331657260734776, + "grad_norm": 4.654809951782227, + "learning_rate": 0.0001937527052288907, + "loss": 0.5211, + "step": 4834 + }, + { + "epoch": 0.11334001418215275, + "grad_norm": 6.114596366882324, + "learning_rate": 0.00019375014219701604, + "loss": 0.3958, + "step": 4835 + }, + { + "epoch": 0.11336345575695775, + "grad_norm": 1.6020206212997437, + "learning_rate": 0.00019374757865645047, + "loss": 0.2239, + "step": 4836 + }, + { + "epoch": 0.11338689733176274, + "grad_norm": 3.6166210174560547, + "learning_rate": 0.00019374501460720796, + "loss": 0.6773, + "step": 4837 + }, + { + "epoch": 0.11341033890656774, + "grad_norm": 2.2476353645324707, + "learning_rate": 0.00019374245004930233, + "loss": 1.3673, + "step": 4838 + }, + { + "epoch": 0.11343378048137275, + "grad_norm": 1.0229398012161255, + "learning_rate": 0.00019373988498274752, + "loss": 0.133, + "step": 4839 + }, + { + "epoch": 0.11345722205617774, + "grad_norm": 4.175358295440674, + "learning_rate": 0.00019373731940755748, + "loss": 0.549, + "step": 4840 + }, + { + "epoch": 0.11348066363098273, + "grad_norm": 3.7584269046783447, + "learning_rate": 0.00019373475332374611, + "loss": 0.4592, + "step": 4841 + }, + { + "epoch": 0.11350410520578773, + "grad_norm": 3.548752546310425, + "learning_rate": 0.00019373218673132732, + "loss": 0.3991, + "step": 4842 + }, + { + "epoch": 0.11352754678059272, + "grad_norm": 1.149796962738037, + "learning_rate": 0.00019372961963031506, + "loss": 0.1187, + "step": 4843 + }, + { + "epoch": 0.11355098835539772, + "grad_norm": 1.372865915298462, + "learning_rate": 0.00019372705202072324, + "loss": 0.1665, + "step": 4844 + }, + { + "epoch": 0.11357442993020271, + "grad_norm": 2.1148645877838135, + "learning_rate": 0.00019372448390256582, + "loss": 0.1737, + "step": 4845 + }, + { + "epoch": 0.1135978715050077, + "grad_norm": 4.812147617340088, + "learning_rate": 0.00019372191527585668, + "loss": 0.4437, + "step": 4846 + }, + { + "epoch": 0.1136213130798127, + "grad_norm": 2.587566375732422, + "learning_rate": 0.0001937193461406098, + "loss": 0.4663, + "step": 4847 + }, + { + "epoch": 0.1136447546546177, + "grad_norm": 5.24409818649292, + "learning_rate": 0.00019371677649683916, + "loss": 0.595, + "step": 4848 + }, + { + "epoch": 0.11366819622942269, + "grad_norm": 1.233995795249939, + "learning_rate": 0.0001937142063445586, + "loss": 0.1412, + "step": 4849 + }, + { + "epoch": 0.11369163780422768, + "grad_norm": 3.900810718536377, + "learning_rate": 0.00019371163568378212, + "loss": 0.3064, + "step": 4850 + }, + { + "epoch": 0.11371507937903268, + "grad_norm": 4.044016361236572, + "learning_rate": 0.0001937090645145237, + "loss": 0.5292, + "step": 4851 + }, + { + "epoch": 0.11373852095383769, + "grad_norm": 4.296700954437256, + "learning_rate": 0.00019370649283679728, + "loss": 0.5844, + "step": 4852 + }, + { + "epoch": 0.11376196252864268, + "grad_norm": 4.378707408905029, + "learning_rate": 0.00019370392065061674, + "loss": 0.5228, + "step": 4853 + }, + { + "epoch": 0.11378540410344767, + "grad_norm": 1.1584100723266602, + "learning_rate": 0.00019370134795599614, + "loss": 0.1112, + "step": 4854 + }, + { + "epoch": 0.11380884567825267, + "grad_norm": 4.494021892547607, + "learning_rate": 0.00019369877475294936, + "loss": 0.7986, + "step": 4855 + }, + { + "epoch": 0.11383228725305766, + "grad_norm": 3.522052526473999, + "learning_rate": 0.0001936962010414904, + "loss": 0.4239, + "step": 4856 + }, + { + "epoch": 0.11385572882786266, + "grad_norm": 4.909821033477783, + "learning_rate": 0.00019369362682163322, + "loss": 0.5326, + "step": 4857 + }, + { + "epoch": 0.11387917040266765, + "grad_norm": 3.6769165992736816, + "learning_rate": 0.0001936910520933918, + "loss": 0.4855, + "step": 4858 + }, + { + "epoch": 0.11390261197747265, + "grad_norm": 5.859513759613037, + "learning_rate": 0.00019368847685678008, + "loss": 0.8094, + "step": 4859 + }, + { + "epoch": 0.11392605355227764, + "grad_norm": 2.880467414855957, + "learning_rate": 0.00019368590111181204, + "loss": 0.2598, + "step": 4860 + }, + { + "epoch": 0.11394949512708263, + "grad_norm": 3.870816707611084, + "learning_rate": 0.0001936833248585017, + "loss": 0.6575, + "step": 4861 + }, + { + "epoch": 0.11397293670188763, + "grad_norm": 1.0930049419403076, + "learning_rate": 0.000193680748096863, + "loss": 0.1397, + "step": 4862 + }, + { + "epoch": 0.11399637827669262, + "grad_norm": 3.6533043384552, + "learning_rate": 0.00019367817082690988, + "loss": 1.547, + "step": 4863 + }, + { + "epoch": 0.11401981985149762, + "grad_norm": 3.397817373275757, + "learning_rate": 0.0001936755930486564, + "loss": 0.5509, + "step": 4864 + }, + { + "epoch": 0.11404326142630263, + "grad_norm": 4.91722297668457, + "learning_rate": 0.00019367301476211655, + "loss": 0.5469, + "step": 4865 + }, + { + "epoch": 0.11406670300110762, + "grad_norm": 4.667994976043701, + "learning_rate": 0.00019367043596730426, + "loss": 0.756, + "step": 4866 + }, + { + "epoch": 0.11409014457591261, + "grad_norm": 1.2139304876327515, + "learning_rate": 0.00019366785666423354, + "loss": 0.1577, + "step": 4867 + }, + { + "epoch": 0.11411358615071761, + "grad_norm": 4.928985595703125, + "learning_rate": 0.0001936652768529184, + "loss": 1.552, + "step": 4868 + }, + { + "epoch": 0.1141370277255226, + "grad_norm": 3.0352752208709717, + "learning_rate": 0.00019366269653337285, + "loss": 0.1929, + "step": 4869 + }, + { + "epoch": 0.1141604693003276, + "grad_norm": 2.8956480026245117, + "learning_rate": 0.00019366011570561088, + "loss": 0.3991, + "step": 4870 + }, + { + "epoch": 0.11418391087513259, + "grad_norm": 8.28152084350586, + "learning_rate": 0.00019365753436964645, + "loss": 0.3446, + "step": 4871 + }, + { + "epoch": 0.11420735244993759, + "grad_norm": 4.873141765594482, + "learning_rate": 0.00019365495252549365, + "loss": 0.7242, + "step": 4872 + }, + { + "epoch": 0.11423079402474258, + "grad_norm": 3.9419009685516357, + "learning_rate": 0.00019365237017316642, + "loss": 0.5489, + "step": 4873 + }, + { + "epoch": 0.11425423559954757, + "grad_norm": 5.327589988708496, + "learning_rate": 0.00019364978731267879, + "loss": 0.5326, + "step": 4874 + }, + { + "epoch": 0.11427767717435257, + "grad_norm": 5.652124881744385, + "learning_rate": 0.0001936472039440448, + "loss": 0.7548, + "step": 4875 + }, + { + "epoch": 0.11430111874915756, + "grad_norm": 3.727339744567871, + "learning_rate": 0.00019364462006727842, + "loss": 0.3758, + "step": 4876 + }, + { + "epoch": 0.11432456032396257, + "grad_norm": 4.952976226806641, + "learning_rate": 0.00019364203568239375, + "loss": 0.345, + "step": 4877 + }, + { + "epoch": 0.11434800189876757, + "grad_norm": 1.1591377258300781, + "learning_rate": 0.00019363945078940473, + "loss": 0.1298, + "step": 4878 + }, + { + "epoch": 0.11437144347357256, + "grad_norm": 1.0816435813903809, + "learning_rate": 0.0001936368653883254, + "loss": 0.1238, + "step": 4879 + }, + { + "epoch": 0.11439488504837755, + "grad_norm": 1.7061020135879517, + "learning_rate": 0.00019363427947916985, + "loss": 0.188, + "step": 4880 + }, + { + "epoch": 0.11441832662318255, + "grad_norm": 1.72874116897583, + "learning_rate": 0.00019363169306195203, + "loss": 0.2781, + "step": 4881 + }, + { + "epoch": 0.11444176819798754, + "grad_norm": 3.251680850982666, + "learning_rate": 0.00019362910613668604, + "loss": 1.8652, + "step": 4882 + }, + { + "epoch": 0.11446520977279254, + "grad_norm": 4.046894550323486, + "learning_rate": 0.00019362651870338585, + "loss": 0.4581, + "step": 4883 + }, + { + "epoch": 0.11448865134759753, + "grad_norm": 3.2123403549194336, + "learning_rate": 0.00019362393076206556, + "loss": 0.3139, + "step": 4884 + }, + { + "epoch": 0.11451209292240253, + "grad_norm": 4.373807907104492, + "learning_rate": 0.00019362134231273918, + "loss": 0.6521, + "step": 4885 + }, + { + "epoch": 0.11453553449720752, + "grad_norm": 1.973531723022461, + "learning_rate": 0.00019361875335542077, + "loss": 0.3804, + "step": 4886 + }, + { + "epoch": 0.11455897607201251, + "grad_norm": 1.7596625089645386, + "learning_rate": 0.00019361616389012437, + "loss": 0.3052, + "step": 4887 + }, + { + "epoch": 0.11458241764681751, + "grad_norm": 4.9958086013793945, + "learning_rate": 0.00019361357391686404, + "loss": 0.6266, + "step": 4888 + }, + { + "epoch": 0.1146058592216225, + "grad_norm": 4.361586570739746, + "learning_rate": 0.00019361098343565383, + "loss": 0.617, + "step": 4889 + }, + { + "epoch": 0.11462930079642751, + "grad_norm": 1.7542054653167725, + "learning_rate": 0.00019360839244650775, + "loss": 0.3214, + "step": 4890 + }, + { + "epoch": 0.1146527423712325, + "grad_norm": 0.9041233658790588, + "learning_rate": 0.00019360580094943995, + "loss": 0.1455, + "step": 4891 + }, + { + "epoch": 0.1146761839460375, + "grad_norm": 2.584833860397339, + "learning_rate": 0.00019360320894446441, + "loss": 1.1358, + "step": 4892 + }, + { + "epoch": 0.1146996255208425, + "grad_norm": 5.330487251281738, + "learning_rate": 0.00019360061643159524, + "loss": 0.5236, + "step": 4893 + }, + { + "epoch": 0.11472306709564749, + "grad_norm": 3.799936294555664, + "learning_rate": 0.0001935980234108465, + "loss": 0.5756, + "step": 4894 + }, + { + "epoch": 0.11474650867045248, + "grad_norm": 1.6830437183380127, + "learning_rate": 0.00019359542988223226, + "loss": 0.2829, + "step": 4895 + }, + { + "epoch": 0.11476995024525748, + "grad_norm": 0.46544530987739563, + "learning_rate": 0.00019359283584576654, + "loss": 0.0985, + "step": 4896 + }, + { + "epoch": 0.11479339182006247, + "grad_norm": 0.8699294924736023, + "learning_rate": 0.0001935902413014635, + "loss": 0.1525, + "step": 4897 + }, + { + "epoch": 0.11481683339486747, + "grad_norm": 3.9422292709350586, + "learning_rate": 0.0001935876462493372, + "loss": 0.3505, + "step": 4898 + }, + { + "epoch": 0.11484027496967246, + "grad_norm": 2.3785831928253174, + "learning_rate": 0.00019358505068940165, + "loss": 0.2956, + "step": 4899 + }, + { + "epoch": 0.11486371654447745, + "grad_norm": 1.7647427320480347, + "learning_rate": 0.000193582454621671, + "loss": 0.344, + "step": 4900 + }, + { + "epoch": 0.11488715811928245, + "grad_norm": 4.695413589477539, + "learning_rate": 0.00019357985804615934, + "loss": 1.0656, + "step": 4901 + }, + { + "epoch": 0.11491059969408744, + "grad_norm": 4.443937301635742, + "learning_rate": 0.0001935772609628807, + "loss": 0.6332, + "step": 4902 + }, + { + "epoch": 0.11493404126889245, + "grad_norm": 4.651393413543701, + "learning_rate": 0.00019357466337184925, + "loss": 0.5342, + "step": 4903 + }, + { + "epoch": 0.11495748284369744, + "grad_norm": 2.728917360305786, + "learning_rate": 0.00019357206527307902, + "loss": 0.3182, + "step": 4904 + }, + { + "epoch": 0.11498092441850244, + "grad_norm": 2.560635566711426, + "learning_rate": 0.00019356946666658416, + "loss": 2.1155, + "step": 4905 + }, + { + "epoch": 0.11500436599330743, + "grad_norm": 2.7059810161590576, + "learning_rate": 0.00019356686755237873, + "loss": 1.8635, + "step": 4906 + }, + { + "epoch": 0.11502780756811243, + "grad_norm": 4.059035301208496, + "learning_rate": 0.00019356426793047684, + "loss": 0.4104, + "step": 4907 + }, + { + "epoch": 0.11505124914291742, + "grad_norm": 3.517320394515991, + "learning_rate": 0.0001935616678008926, + "loss": 0.4422, + "step": 4908 + }, + { + "epoch": 0.11507469071772242, + "grad_norm": 2.9365928173065186, + "learning_rate": 0.00019355906716364012, + "loss": 0.4604, + "step": 4909 + }, + { + "epoch": 0.11509813229252741, + "grad_norm": 2.985799551010132, + "learning_rate": 0.00019355646601873353, + "loss": 0.6031, + "step": 4910 + }, + { + "epoch": 0.1151215738673324, + "grad_norm": 1.1689813137054443, + "learning_rate": 0.0001935538643661869, + "loss": 0.1429, + "step": 4911 + }, + { + "epoch": 0.1151450154421374, + "grad_norm": 3.2328131198883057, + "learning_rate": 0.0001935512622060144, + "loss": 0.411, + "step": 4912 + }, + { + "epoch": 0.1151684570169424, + "grad_norm": 4.9115681648254395, + "learning_rate": 0.00019354865953823008, + "loss": 0.9021, + "step": 4913 + }, + { + "epoch": 0.11519189859174739, + "grad_norm": 0.6803158521652222, + "learning_rate": 0.00019354605636284812, + "loss": 0.0608, + "step": 4914 + }, + { + "epoch": 0.11521534016655238, + "grad_norm": 1.6034064292907715, + "learning_rate": 0.00019354345267988263, + "loss": 0.2183, + "step": 4915 + }, + { + "epoch": 0.11523878174135739, + "grad_norm": 3.7023260593414307, + "learning_rate": 0.00019354084848934776, + "loss": 0.5621, + "step": 4916 + }, + { + "epoch": 0.11526222331616238, + "grad_norm": 7.051976203918457, + "learning_rate": 0.00019353824379125757, + "loss": 0.7818, + "step": 4917 + }, + { + "epoch": 0.11528566489096738, + "grad_norm": 3.7292368412017822, + "learning_rate": 0.00019353563858562627, + "loss": 0.8042, + "step": 4918 + }, + { + "epoch": 0.11530910646577237, + "grad_norm": 3.8502047061920166, + "learning_rate": 0.00019353303287246795, + "loss": 0.5647, + "step": 4919 + }, + { + "epoch": 0.11533254804057737, + "grad_norm": 7.359594345092773, + "learning_rate": 0.0001935304266517968, + "loss": 0.4777, + "step": 4920 + }, + { + "epoch": 0.11535598961538236, + "grad_norm": 4.655484676361084, + "learning_rate": 0.00019352781992362688, + "loss": 0.54, + "step": 4921 + }, + { + "epoch": 0.11537943119018736, + "grad_norm": 4.524214267730713, + "learning_rate": 0.00019352521268797238, + "loss": 0.7014, + "step": 4922 + }, + { + "epoch": 0.11540287276499235, + "grad_norm": 3.2214105129241943, + "learning_rate": 0.00019352260494484748, + "loss": 0.4275, + "step": 4923 + }, + { + "epoch": 0.11542631433979735, + "grad_norm": 3.2914035320281982, + "learning_rate": 0.00019351999669426628, + "loss": 0.3317, + "step": 4924 + }, + { + "epoch": 0.11544975591460234, + "grad_norm": 2.3084044456481934, + "learning_rate": 0.00019351738793624295, + "loss": 1.3596, + "step": 4925 + }, + { + "epoch": 0.11547319748940733, + "grad_norm": 2.782069206237793, + "learning_rate": 0.00019351477867079163, + "loss": 0.3352, + "step": 4926 + }, + { + "epoch": 0.11549663906421233, + "grad_norm": 3.132642984390259, + "learning_rate": 0.00019351216889792653, + "loss": 0.3448, + "step": 4927 + }, + { + "epoch": 0.11552008063901732, + "grad_norm": 2.9051263332366943, + "learning_rate": 0.00019350955861766172, + "loss": 0.3534, + "step": 4928 + }, + { + "epoch": 0.11554352221382233, + "grad_norm": 4.980300426483154, + "learning_rate": 0.00019350694783001147, + "loss": 0.7069, + "step": 4929 + }, + { + "epoch": 0.11556696378862732, + "grad_norm": 5.206672668457031, + "learning_rate": 0.00019350433653498984, + "loss": 0.5857, + "step": 4930 + }, + { + "epoch": 0.11559040536343232, + "grad_norm": 1.5211607217788696, + "learning_rate": 0.00019350172473261108, + "loss": 0.2391, + "step": 4931 + }, + { + "epoch": 0.11561384693823731, + "grad_norm": 4.602694511413574, + "learning_rate": 0.00019349911242288933, + "loss": 0.5899, + "step": 4932 + }, + { + "epoch": 0.11563728851304231, + "grad_norm": 5.442195415496826, + "learning_rate": 0.00019349649960583875, + "loss": 0.498, + "step": 4933 + }, + { + "epoch": 0.1156607300878473, + "grad_norm": 2.868908166885376, + "learning_rate": 0.00019349388628147357, + "loss": 0.2641, + "step": 4934 + }, + { + "epoch": 0.1156841716626523, + "grad_norm": 1.609654426574707, + "learning_rate": 0.0001934912724498079, + "loss": 0.2546, + "step": 4935 + }, + { + "epoch": 0.11570761323745729, + "grad_norm": 2.1092019081115723, + "learning_rate": 0.00019348865811085597, + "loss": 0.1377, + "step": 4936 + }, + { + "epoch": 0.11573105481226229, + "grad_norm": 2.9149532318115234, + "learning_rate": 0.00019348604326463193, + "loss": 0.3525, + "step": 4937 + }, + { + "epoch": 0.11575449638706728, + "grad_norm": 3.6730453968048096, + "learning_rate": 0.00019348342791115003, + "loss": 0.3312, + "step": 4938 + }, + { + "epoch": 0.11577793796187227, + "grad_norm": 2.972499370574951, + "learning_rate": 0.00019348081205042438, + "loss": 0.6093, + "step": 4939 + }, + { + "epoch": 0.11580137953667727, + "grad_norm": 4.551824569702148, + "learning_rate": 0.00019347819568246924, + "loss": 0.7694, + "step": 4940 + }, + { + "epoch": 0.11582482111148228, + "grad_norm": 4.705928325653076, + "learning_rate": 0.00019347557880729878, + "loss": 0.6375, + "step": 4941 + }, + { + "epoch": 0.11584826268628727, + "grad_norm": 4.254760265350342, + "learning_rate": 0.0001934729614249272, + "loss": 0.707, + "step": 4942 + }, + { + "epoch": 0.11587170426109226, + "grad_norm": 2.714816093444824, + "learning_rate": 0.00019347034353536872, + "loss": 0.2678, + "step": 4943 + }, + { + "epoch": 0.11589514583589726, + "grad_norm": 3.1765480041503906, + "learning_rate": 0.00019346772513863751, + "loss": 0.5698, + "step": 4944 + }, + { + "epoch": 0.11591858741070225, + "grad_norm": 4.949816703796387, + "learning_rate": 0.0001934651062347478, + "loss": 0.725, + "step": 4945 + }, + { + "epoch": 0.11594202898550725, + "grad_norm": 3.6688766479492188, + "learning_rate": 0.00019346248682371379, + "loss": 0.5291, + "step": 4946 + }, + { + "epoch": 0.11596547056031224, + "grad_norm": 0.5024372339248657, + "learning_rate": 0.00019345986690554972, + "loss": 0.0609, + "step": 4947 + }, + { + "epoch": 0.11598891213511724, + "grad_norm": 2.8329625129699707, + "learning_rate": 0.00019345724648026975, + "loss": 0.3845, + "step": 4948 + }, + { + "epoch": 0.11601235370992223, + "grad_norm": 4.960634708404541, + "learning_rate": 0.00019345462554788814, + "loss": 0.7238, + "step": 4949 + }, + { + "epoch": 0.11603579528472723, + "grad_norm": 2.620201826095581, + "learning_rate": 0.00019345200410841913, + "loss": 1.7061, + "step": 4950 + }, + { + "epoch": 0.11605923685953222, + "grad_norm": 6.5832390785217285, + "learning_rate": 0.0001934493821618769, + "loss": 1.574, + "step": 4951 + }, + { + "epoch": 0.11608267843433721, + "grad_norm": 3.6522464752197266, + "learning_rate": 0.00019344675970827568, + "loss": 0.5019, + "step": 4952 + }, + { + "epoch": 0.11610612000914221, + "grad_norm": 3.4096035957336426, + "learning_rate": 0.00019344413674762974, + "loss": 0.4107, + "step": 4953 + }, + { + "epoch": 0.11612956158394722, + "grad_norm": 2.665546178817749, + "learning_rate": 0.00019344151327995327, + "loss": 1.662, + "step": 4954 + }, + { + "epoch": 0.11615300315875221, + "grad_norm": 0.7846449017524719, + "learning_rate": 0.00019343888930526052, + "loss": 0.1384, + "step": 4955 + }, + { + "epoch": 0.1161764447335572, + "grad_norm": 0.3746916651725769, + "learning_rate": 0.00019343626482356573, + "loss": 0.0562, + "step": 4956 + }, + { + "epoch": 0.1161998863083622, + "grad_norm": 4.9468255043029785, + "learning_rate": 0.00019343363983488317, + "loss": 0.5041, + "step": 4957 + }, + { + "epoch": 0.1162233278831672, + "grad_norm": 3.674363374710083, + "learning_rate": 0.000193431014339227, + "loss": 0.5435, + "step": 4958 + }, + { + "epoch": 0.11624676945797219, + "grad_norm": 1.5486905574798584, + "learning_rate": 0.00019342838833661156, + "loss": 0.2076, + "step": 4959 + }, + { + "epoch": 0.11627021103277718, + "grad_norm": 1.8325188159942627, + "learning_rate": 0.00019342576182705107, + "loss": 0.3416, + "step": 4960 + }, + { + "epoch": 0.11629365260758218, + "grad_norm": 2.1773314476013184, + "learning_rate": 0.00019342313481055974, + "loss": 1.3404, + "step": 4961 + }, + { + "epoch": 0.11631709418238717, + "grad_norm": 2.8430473804473877, + "learning_rate": 0.00019342050728715188, + "loss": 0.2611, + "step": 4962 + }, + { + "epoch": 0.11634053575719217, + "grad_norm": 0.524067759513855, + "learning_rate": 0.0001934178792568417, + "loss": 0.0434, + "step": 4963 + }, + { + "epoch": 0.11636397733199716, + "grad_norm": 2.1855506896972656, + "learning_rate": 0.00019341525071964346, + "loss": 0.2985, + "step": 4964 + }, + { + "epoch": 0.11638741890680215, + "grad_norm": 3.9246203899383545, + "learning_rate": 0.00019341262167557147, + "loss": 0.6599, + "step": 4965 + }, + { + "epoch": 0.11641086048160715, + "grad_norm": 2.2771170139312744, + "learning_rate": 0.00019340999212463995, + "loss": 1.5503, + "step": 4966 + }, + { + "epoch": 0.11643430205641216, + "grad_norm": 3.6866133213043213, + "learning_rate": 0.0001934073620668632, + "loss": 0.6407, + "step": 4967 + }, + { + "epoch": 0.11645774363121715, + "grad_norm": 4.644825458526611, + "learning_rate": 0.0001934047315022555, + "loss": 0.7226, + "step": 4968 + }, + { + "epoch": 0.11648118520602214, + "grad_norm": 0.6321972608566284, + "learning_rate": 0.00019340210043083107, + "loss": 0.0558, + "step": 4969 + }, + { + "epoch": 0.11650462678082714, + "grad_norm": 1.3972989320755005, + "learning_rate": 0.00019339946885260422, + "loss": 0.2119, + "step": 4970 + }, + { + "epoch": 0.11652806835563213, + "grad_norm": 3.161525249481201, + "learning_rate": 0.00019339683676758922, + "loss": 0.4509, + "step": 4971 + }, + { + "epoch": 0.11655150993043713, + "grad_norm": 0.518220841884613, + "learning_rate": 0.0001933942041758004, + "loss": 0.0464, + "step": 4972 + }, + { + "epoch": 0.11657495150524212, + "grad_norm": 2.082932710647583, + "learning_rate": 0.00019339157107725195, + "loss": 0.3005, + "step": 4973 + }, + { + "epoch": 0.11659839308004712, + "grad_norm": 2.6950511932373047, + "learning_rate": 0.00019338893747195823, + "loss": 1.8413, + "step": 4974 + }, + { + "epoch": 0.11662183465485211, + "grad_norm": 1.553317904472351, + "learning_rate": 0.0001933863033599335, + "loss": 0.2623, + "step": 4975 + }, + { + "epoch": 0.1166452762296571, + "grad_norm": 4.106377124786377, + "learning_rate": 0.00019338366874119208, + "loss": 0.5968, + "step": 4976 + }, + { + "epoch": 0.1166687178044621, + "grad_norm": 1.9848631620407104, + "learning_rate": 0.00019338103361574822, + "loss": 0.3349, + "step": 4977 + }, + { + "epoch": 0.1166921593792671, + "grad_norm": 1.5337114334106445, + "learning_rate": 0.00019337839798361627, + "loss": 0.2332, + "step": 4978 + }, + { + "epoch": 0.11671560095407209, + "grad_norm": 2.055036783218384, + "learning_rate": 0.0001933757618448105, + "loss": 1.8196, + "step": 4979 + }, + { + "epoch": 0.1167390425288771, + "grad_norm": 2.481196880340576, + "learning_rate": 0.00019337312519934524, + "loss": 0.4384, + "step": 4980 + }, + { + "epoch": 0.11676248410368209, + "grad_norm": 3.7965481281280518, + "learning_rate": 0.00019337048804723473, + "loss": 0.5699, + "step": 4981 + }, + { + "epoch": 0.11678592567848708, + "grad_norm": 2.7882561683654785, + "learning_rate": 0.00019336785038849335, + "loss": 1.8858, + "step": 4982 + }, + { + "epoch": 0.11680936725329208, + "grad_norm": 3.341756582260132, + "learning_rate": 0.0001933652122231354, + "loss": 0.4139, + "step": 4983 + }, + { + "epoch": 0.11683280882809707, + "grad_norm": 4.958105087280273, + "learning_rate": 0.00019336257355117518, + "loss": 0.7196, + "step": 4984 + }, + { + "epoch": 0.11685625040290207, + "grad_norm": 3.8486411571502686, + "learning_rate": 0.00019335993437262703, + "loss": 0.6033, + "step": 4985 + }, + { + "epoch": 0.11687969197770706, + "grad_norm": 5.451600074768066, + "learning_rate": 0.0001933572946875052, + "loss": 0.7773, + "step": 4986 + }, + { + "epoch": 0.11690313355251206, + "grad_norm": 2.863995313644409, + "learning_rate": 0.0001933546544958241, + "loss": 1.7224, + "step": 4987 + }, + { + "epoch": 0.11692657512731705, + "grad_norm": 0.6095690131187439, + "learning_rate": 0.000193352013797598, + "loss": 0.0885, + "step": 4988 + }, + { + "epoch": 0.11695001670212204, + "grad_norm": 6.443140506744385, + "learning_rate": 0.00019334937259284129, + "loss": 0.8564, + "step": 4989 + }, + { + "epoch": 0.11697345827692704, + "grad_norm": 10.759397506713867, + "learning_rate": 0.00019334673088156824, + "loss": 0.7335, + "step": 4990 + }, + { + "epoch": 0.11699689985173203, + "grad_norm": 1.3430949449539185, + "learning_rate": 0.00019334408866379316, + "loss": 0.3135, + "step": 4991 + }, + { + "epoch": 0.11702034142653703, + "grad_norm": 2.2502384185791016, + "learning_rate": 0.00019334144593953048, + "loss": 0.3344, + "step": 4992 + }, + { + "epoch": 0.11704378300134204, + "grad_norm": 3.5627825260162354, + "learning_rate": 0.00019333880270879446, + "loss": 0.6554, + "step": 4993 + }, + { + "epoch": 0.11706722457614703, + "grad_norm": 4.554779052734375, + "learning_rate": 0.00019333615897159948, + "loss": 0.9406, + "step": 4994 + }, + { + "epoch": 0.11709066615095202, + "grad_norm": 1.3066747188568115, + "learning_rate": 0.0001933335147279599, + "loss": 0.1919, + "step": 4995 + }, + { + "epoch": 0.11711410772575702, + "grad_norm": 2.873270034790039, + "learning_rate": 0.00019333086997789001, + "loss": 0.3335, + "step": 4996 + }, + { + "epoch": 0.11713754930056201, + "grad_norm": 0.9805269837379456, + "learning_rate": 0.00019332822472140423, + "loss": 0.1197, + "step": 4997 + }, + { + "epoch": 0.11716099087536701, + "grad_norm": 4.639997959136963, + "learning_rate": 0.00019332557895851687, + "loss": 0.6184, + "step": 4998 + }, + { + "epoch": 0.117184432450172, + "grad_norm": 2.8219194412231445, + "learning_rate": 0.00019332293268924228, + "loss": 0.2173, + "step": 4999 + }, + { + "epoch": 0.117207874024977, + "grad_norm": 4.0290093421936035, + "learning_rate": 0.00019332028591359483, + "loss": 0.6477, + "step": 5000 + }, + { + "epoch": 0.11723131559978199, + "grad_norm": 2.6990230083465576, + "learning_rate": 0.00019331763863158892, + "loss": 1.325, + "step": 5001 + }, + { + "epoch": 0.11725475717458698, + "grad_norm": 1.7448837757110596, + "learning_rate": 0.00019331499084323882, + "loss": 0.3511, + "step": 5002 + }, + { + "epoch": 0.11727819874939198, + "grad_norm": 3.454327344894409, + "learning_rate": 0.000193312342548559, + "loss": 0.4428, + "step": 5003 + }, + { + "epoch": 0.11730164032419697, + "grad_norm": 2.185593843460083, + "learning_rate": 0.00019330969374756378, + "loss": 0.2878, + "step": 5004 + }, + { + "epoch": 0.11732508189900198, + "grad_norm": 3.9613146781921387, + "learning_rate": 0.00019330704444026756, + "loss": 0.5353, + "step": 5005 + }, + { + "epoch": 0.11734852347380698, + "grad_norm": 4.16701602935791, + "learning_rate": 0.00019330439462668465, + "loss": 0.3105, + "step": 5006 + }, + { + "epoch": 0.11737196504861197, + "grad_norm": 3.8844826221466064, + "learning_rate": 0.0001933017443068295, + "loss": 1.392, + "step": 5007 + }, + { + "epoch": 0.11739540662341696, + "grad_norm": 4.6862263679504395, + "learning_rate": 0.00019329909348071643, + "loss": 0.5012, + "step": 5008 + }, + { + "epoch": 0.11741884819822196, + "grad_norm": 5.843983173370361, + "learning_rate": 0.0001932964421483599, + "loss": 0.8127, + "step": 5009 + }, + { + "epoch": 0.11744228977302695, + "grad_norm": 3.97699236869812, + "learning_rate": 0.0001932937903097742, + "loss": 0.263, + "step": 5010 + }, + { + "epoch": 0.11746573134783195, + "grad_norm": 2.1210784912109375, + "learning_rate": 0.0001932911379649738, + "loss": 0.3574, + "step": 5011 + }, + { + "epoch": 0.11748917292263694, + "grad_norm": 4.009335517883301, + "learning_rate": 0.00019328848511397306, + "loss": 0.6823, + "step": 5012 + }, + { + "epoch": 0.11751261449744194, + "grad_norm": 3.7381997108459473, + "learning_rate": 0.00019328583175678637, + "loss": 0.4485, + "step": 5013 + }, + { + "epoch": 0.11753605607224693, + "grad_norm": 5.411229133605957, + "learning_rate": 0.00019328317789342814, + "loss": 0.5957, + "step": 5014 + }, + { + "epoch": 0.11755949764705192, + "grad_norm": 0.9830082058906555, + "learning_rate": 0.00019328052352391277, + "loss": 0.1231, + "step": 5015 + }, + { + "epoch": 0.11758293922185692, + "grad_norm": 4.488698482513428, + "learning_rate": 0.00019327786864825464, + "loss": 0.8921, + "step": 5016 + }, + { + "epoch": 0.11760638079666191, + "grad_norm": 3.61740779876709, + "learning_rate": 0.00019327521326646817, + "loss": 0.4326, + "step": 5017 + }, + { + "epoch": 0.11762982237146692, + "grad_norm": 4.786810398101807, + "learning_rate": 0.0001932725573785678, + "loss": 0.34, + "step": 5018 + }, + { + "epoch": 0.11765326394627192, + "grad_norm": 2.0915160179138184, + "learning_rate": 0.0001932699009845679, + "loss": 1.3, + "step": 5019 + }, + { + "epoch": 0.11767670552107691, + "grad_norm": 1.085757851600647, + "learning_rate": 0.00019326724408448288, + "loss": 0.1615, + "step": 5020 + }, + { + "epoch": 0.1177001470958819, + "grad_norm": 4.1671905517578125, + "learning_rate": 0.0001932645866783272, + "loss": 0.7896, + "step": 5021 + }, + { + "epoch": 0.1177235886706869, + "grad_norm": 5.13135290145874, + "learning_rate": 0.00019326192876611522, + "loss": 0.7181, + "step": 5022 + }, + { + "epoch": 0.1177470302454919, + "grad_norm": 4.433331489562988, + "learning_rate": 0.0001932592703478614, + "loss": 0.465, + "step": 5023 + }, + { + "epoch": 0.11777047182029689, + "grad_norm": 1.6161880493164062, + "learning_rate": 0.00019325661142358016, + "loss": 0.3093, + "step": 5024 + }, + { + "epoch": 0.11779391339510188, + "grad_norm": 6.353994369506836, + "learning_rate": 0.00019325395199328593, + "loss": 0.717, + "step": 5025 + }, + { + "epoch": 0.11781735496990688, + "grad_norm": 3.580312967300415, + "learning_rate": 0.00019325129205699312, + "loss": 0.5758, + "step": 5026 + }, + { + "epoch": 0.11784079654471187, + "grad_norm": 4.117129802703857, + "learning_rate": 0.00019324863161471618, + "loss": 0.5333, + "step": 5027 + }, + { + "epoch": 0.11786423811951686, + "grad_norm": 4.300075054168701, + "learning_rate": 0.00019324597066646955, + "loss": 0.5901, + "step": 5028 + }, + { + "epoch": 0.11788767969432186, + "grad_norm": 3.6428205966949463, + "learning_rate": 0.00019324330921226764, + "loss": 0.5874, + "step": 5029 + }, + { + "epoch": 0.11791112126912685, + "grad_norm": 3.003829002380371, + "learning_rate": 0.00019324064725212496, + "loss": 0.2828, + "step": 5030 + }, + { + "epoch": 0.11793456284393186, + "grad_norm": 2.986520767211914, + "learning_rate": 0.0001932379847860559, + "loss": 0.381, + "step": 5031 + }, + { + "epoch": 0.11795800441873686, + "grad_norm": 0.5935752391815186, + "learning_rate": 0.00019323532181407486, + "loss": 0.0672, + "step": 5032 + }, + { + "epoch": 0.11798144599354185, + "grad_norm": 3.964919328689575, + "learning_rate": 0.00019323265833619637, + "loss": 0.5919, + "step": 5033 + }, + { + "epoch": 0.11800488756834684, + "grad_norm": 4.082042217254639, + "learning_rate": 0.00019322999435243484, + "loss": 0.5649, + "step": 5034 + }, + { + "epoch": 0.11802832914315184, + "grad_norm": 0.9682613015174866, + "learning_rate": 0.00019322732986280477, + "loss": 0.0845, + "step": 5035 + }, + { + "epoch": 0.11805177071795683, + "grad_norm": 0.8874209523200989, + "learning_rate": 0.00019322466486732057, + "loss": 0.0962, + "step": 5036 + }, + { + "epoch": 0.11807521229276183, + "grad_norm": 2.074007272720337, + "learning_rate": 0.0001932219993659967, + "loss": 0.1913, + "step": 5037 + }, + { + "epoch": 0.11809865386756682, + "grad_norm": 1.5964094400405884, + "learning_rate": 0.00019321933335884767, + "loss": 1.3163, + "step": 5038 + }, + { + "epoch": 0.11812209544237182, + "grad_norm": 2.786330461502075, + "learning_rate": 0.00019321666684588793, + "loss": 1.156, + "step": 5039 + }, + { + "epoch": 0.11814553701717681, + "grad_norm": 1.3477058410644531, + "learning_rate": 0.00019321399982713192, + "loss": 0.2006, + "step": 5040 + }, + { + "epoch": 0.1181689785919818, + "grad_norm": 3.0732202529907227, + "learning_rate": 0.0001932113323025941, + "loss": 0.3791, + "step": 5041 + }, + { + "epoch": 0.1181924201667868, + "grad_norm": 3.3938655853271484, + "learning_rate": 0.000193208664272289, + "loss": 0.4317, + "step": 5042 + }, + { + "epoch": 0.1182158617415918, + "grad_norm": 4.262282848358154, + "learning_rate": 0.00019320599573623105, + "loss": 0.4897, + "step": 5043 + }, + { + "epoch": 0.1182393033163968, + "grad_norm": 2.7661497592926025, + "learning_rate": 0.00019320332669443474, + "loss": 0.5219, + "step": 5044 + }, + { + "epoch": 0.1182627448912018, + "grad_norm": 1.1637953519821167, + "learning_rate": 0.00019320065714691458, + "loss": 0.1713, + "step": 5045 + }, + { + "epoch": 0.11828618646600679, + "grad_norm": 3.3835911750793457, + "learning_rate": 0.00019319798709368504, + "loss": 0.6482, + "step": 5046 + }, + { + "epoch": 0.11830962804081178, + "grad_norm": 1.605420708656311, + "learning_rate": 0.00019319531653476057, + "loss": 0.1912, + "step": 5047 + }, + { + "epoch": 0.11833306961561678, + "grad_norm": 1.352358102798462, + "learning_rate": 0.00019319264547015572, + "loss": 0.2148, + "step": 5048 + }, + { + "epoch": 0.11835651119042177, + "grad_norm": 2.814657688140869, + "learning_rate": 0.00019318997389988494, + "loss": 1.6788, + "step": 5049 + }, + { + "epoch": 0.11837995276522677, + "grad_norm": 4.334560394287109, + "learning_rate": 0.0001931873018239627, + "loss": 0.656, + "step": 5050 + }, + { + "epoch": 0.11840339434003176, + "grad_norm": 1.7013001441955566, + "learning_rate": 0.00019318462924240362, + "loss": 0.2239, + "step": 5051 + }, + { + "epoch": 0.11842683591483676, + "grad_norm": 3.9043140411376953, + "learning_rate": 0.0001931819561552221, + "loss": 0.5148, + "step": 5052 + }, + { + "epoch": 0.11845027748964175, + "grad_norm": 3.865997791290283, + "learning_rate": 0.00019317928256243265, + "loss": 0.5567, + "step": 5053 + }, + { + "epoch": 0.11847371906444674, + "grad_norm": 4.590125560760498, + "learning_rate": 0.0001931766084640498, + "loss": 0.4689, + "step": 5054 + }, + { + "epoch": 0.11849716063925174, + "grad_norm": 5.3909101486206055, + "learning_rate": 0.00019317393386008802, + "loss": 0.8494, + "step": 5055 + }, + { + "epoch": 0.11852060221405675, + "grad_norm": 3.463456153869629, + "learning_rate": 0.0001931712587505619, + "loss": 0.6316, + "step": 5056 + }, + { + "epoch": 0.11854404378886174, + "grad_norm": 3.782198905944824, + "learning_rate": 0.00019316858313548588, + "loss": 0.7478, + "step": 5057 + }, + { + "epoch": 0.11856748536366674, + "grad_norm": 2.15399169921875, + "learning_rate": 0.00019316590701487453, + "loss": 0.2991, + "step": 5058 + }, + { + "epoch": 0.11859092693847173, + "grad_norm": 5.681337833404541, + "learning_rate": 0.00019316323038874234, + "loss": 0.9276, + "step": 5059 + }, + { + "epoch": 0.11861436851327672, + "grad_norm": 4.037329196929932, + "learning_rate": 0.0001931605532571038, + "loss": 0.6226, + "step": 5060 + }, + { + "epoch": 0.11863781008808172, + "grad_norm": 3.9202136993408203, + "learning_rate": 0.00019315787561997354, + "loss": 0.5195, + "step": 5061 + }, + { + "epoch": 0.11866125166288671, + "grad_norm": 2.165375232696533, + "learning_rate": 0.00019315519747736603, + "loss": 0.2912, + "step": 5062 + }, + { + "epoch": 0.11868469323769171, + "grad_norm": 4.419837951660156, + "learning_rate": 0.00019315251882929578, + "loss": 0.4233, + "step": 5063 + }, + { + "epoch": 0.1187081348124967, + "grad_norm": 7.266263008117676, + "learning_rate": 0.00019314983967577733, + "loss": 0.5193, + "step": 5064 + }, + { + "epoch": 0.1187315763873017, + "grad_norm": 3.882683515548706, + "learning_rate": 0.00019314716001682522, + "loss": 0.6386, + "step": 5065 + }, + { + "epoch": 0.11875501796210669, + "grad_norm": 1.6187522411346436, + "learning_rate": 0.00019314447985245404, + "loss": 0.234, + "step": 5066 + }, + { + "epoch": 0.11877845953691168, + "grad_norm": 1.1746371984481812, + "learning_rate": 0.00019314179918267829, + "loss": 0.2047, + "step": 5067 + }, + { + "epoch": 0.11880190111171668, + "grad_norm": 0.9505111575126648, + "learning_rate": 0.0001931391180075125, + "loss": 0.0824, + "step": 5068 + }, + { + "epoch": 0.11882534268652169, + "grad_norm": 2.666029930114746, + "learning_rate": 0.00019313643632697125, + "loss": 1.4427, + "step": 5069 + }, + { + "epoch": 0.11884878426132668, + "grad_norm": 3.4717652797698975, + "learning_rate": 0.00019313375414106907, + "loss": 0.321, + "step": 5070 + }, + { + "epoch": 0.11887222583613168, + "grad_norm": 3.593245506286621, + "learning_rate": 0.0001931310714498205, + "loss": 0.5477, + "step": 5071 + }, + { + "epoch": 0.11889566741093667, + "grad_norm": 2.8473494052886963, + "learning_rate": 0.00019312838825324014, + "loss": 0.3946, + "step": 5072 + }, + { + "epoch": 0.11891910898574166, + "grad_norm": 3.325827121734619, + "learning_rate": 0.00019312570455134252, + "loss": 0.5613, + "step": 5073 + }, + { + "epoch": 0.11894255056054666, + "grad_norm": 4.6893181800842285, + "learning_rate": 0.00019312302034414224, + "loss": 0.5385, + "step": 5074 + }, + { + "epoch": 0.11896599213535165, + "grad_norm": 4.670650005340576, + "learning_rate": 0.00019312033563165376, + "loss": 0.6505, + "step": 5075 + }, + { + "epoch": 0.11898943371015665, + "grad_norm": 4.264636039733887, + "learning_rate": 0.0001931176504138918, + "loss": 0.5428, + "step": 5076 + }, + { + "epoch": 0.11901287528496164, + "grad_norm": 0.5027433633804321, + "learning_rate": 0.00019311496469087081, + "loss": 0.0909, + "step": 5077 + }, + { + "epoch": 0.11903631685976664, + "grad_norm": 1.2874222993850708, + "learning_rate": 0.00019311227846260544, + "loss": 0.1487, + "step": 5078 + }, + { + "epoch": 0.11905975843457163, + "grad_norm": 3.5926568508148193, + "learning_rate": 0.0001931095917291102, + "loss": 0.6984, + "step": 5079 + }, + { + "epoch": 0.11908320000937662, + "grad_norm": 2.315789222717285, + "learning_rate": 0.00019310690449039964, + "loss": 0.2552, + "step": 5080 + }, + { + "epoch": 0.11910664158418162, + "grad_norm": 4.4254231452941895, + "learning_rate": 0.00019310421674648847, + "loss": 0.8582, + "step": 5081 + }, + { + "epoch": 0.11913008315898663, + "grad_norm": 3.026198625564575, + "learning_rate": 0.00019310152849739122, + "loss": 0.4937, + "step": 5082 + }, + { + "epoch": 0.11915352473379162, + "grad_norm": 2.2788522243499756, + "learning_rate": 0.00019309883974312242, + "loss": 1.6164, + "step": 5083 + }, + { + "epoch": 0.11917696630859662, + "grad_norm": 3.4166460037231445, + "learning_rate": 0.00019309615048369668, + "loss": 0.4789, + "step": 5084 + }, + { + "epoch": 0.11920040788340161, + "grad_norm": 4.785659313201904, + "learning_rate": 0.00019309346071912865, + "loss": 0.636, + "step": 5085 + }, + { + "epoch": 0.1192238494582066, + "grad_norm": 3.5432190895080566, + "learning_rate": 0.00019309077044943286, + "loss": 0.5854, + "step": 5086 + }, + { + "epoch": 0.1192472910330116, + "grad_norm": 1.261125087738037, + "learning_rate": 0.00019308807967462395, + "loss": 0.2011, + "step": 5087 + }, + { + "epoch": 0.11927073260781659, + "grad_norm": 3.1470487117767334, + "learning_rate": 0.0001930853883947165, + "loss": 0.4463, + "step": 5088 + }, + { + "epoch": 0.11929417418262159, + "grad_norm": 4.815899848937988, + "learning_rate": 0.00019308269660972508, + "loss": 0.4079, + "step": 5089 + }, + { + "epoch": 0.11931761575742658, + "grad_norm": 1.334166407585144, + "learning_rate": 0.00019308000431966438, + "loss": 0.1422, + "step": 5090 + }, + { + "epoch": 0.11934105733223158, + "grad_norm": 0.4689319133758545, + "learning_rate": 0.00019307731152454894, + "loss": 0.0385, + "step": 5091 + }, + { + "epoch": 0.11936449890703657, + "grad_norm": 4.746934413909912, + "learning_rate": 0.0001930746182243934, + "loss": 0.7006, + "step": 5092 + }, + { + "epoch": 0.11938794048184156, + "grad_norm": 3.1028354167938232, + "learning_rate": 0.00019307192441921233, + "loss": 0.4628, + "step": 5093 + }, + { + "epoch": 0.11941138205664656, + "grad_norm": 2.3379111289978027, + "learning_rate": 0.0001930692301090204, + "loss": 1.3129, + "step": 5094 + }, + { + "epoch": 0.11943482363145157, + "grad_norm": 3.021148443222046, + "learning_rate": 0.00019306653529383222, + "loss": 0.383, + "step": 5095 + }, + { + "epoch": 0.11945826520625656, + "grad_norm": 1.8342299461364746, + "learning_rate": 0.0001930638399736624, + "loss": 1.3043, + "step": 5096 + }, + { + "epoch": 0.11948170678106156, + "grad_norm": 4.46923303604126, + "learning_rate": 0.0001930611441485256, + "loss": 0.3899, + "step": 5097 + }, + { + "epoch": 0.11950514835586655, + "grad_norm": 3.41015362739563, + "learning_rate": 0.00019305844781843633, + "loss": 0.849, + "step": 5098 + }, + { + "epoch": 0.11952858993067154, + "grad_norm": 2.3795242309570312, + "learning_rate": 0.00019305575098340938, + "loss": 0.1725, + "step": 5099 + }, + { + "epoch": 0.11955203150547654, + "grad_norm": 4.43122673034668, + "learning_rate": 0.0001930530536434593, + "loss": 0.5517, + "step": 5100 + }, + { + "epoch": 0.11957547308028153, + "grad_norm": 3.5575695037841797, + "learning_rate": 0.0001930503557986007, + "loss": 0.5088, + "step": 5101 + }, + { + "epoch": 0.11959891465508653, + "grad_norm": 3.4715683460235596, + "learning_rate": 0.00019304765744884828, + "loss": 0.4059, + "step": 5102 + }, + { + "epoch": 0.11962235622989152, + "grad_norm": 4.267117023468018, + "learning_rate": 0.00019304495859421664, + "loss": 0.4054, + "step": 5103 + }, + { + "epoch": 0.11964579780469652, + "grad_norm": 2.688842296600342, + "learning_rate": 0.00019304225923472043, + "loss": 1.7373, + "step": 5104 + }, + { + "epoch": 0.11966923937950151, + "grad_norm": 0.8712986707687378, + "learning_rate": 0.00019303955937037434, + "loss": 0.1416, + "step": 5105 + }, + { + "epoch": 0.1196926809543065, + "grad_norm": 5.43213415145874, + "learning_rate": 0.00019303685900119298, + "loss": 0.669, + "step": 5106 + }, + { + "epoch": 0.1197161225291115, + "grad_norm": 1.8179699182510376, + "learning_rate": 0.00019303415812719097, + "loss": 0.2365, + "step": 5107 + }, + { + "epoch": 0.11973956410391651, + "grad_norm": 1.3966070413589478, + "learning_rate": 0.00019303145674838304, + "loss": 0.4027, + "step": 5108 + }, + { + "epoch": 0.1197630056787215, + "grad_norm": 1.9179513454437256, + "learning_rate": 0.00019302875486478378, + "loss": 0.1985, + "step": 5109 + }, + { + "epoch": 0.1197864472535265, + "grad_norm": 0.6679768562316895, + "learning_rate": 0.0001930260524764079, + "loss": 0.0996, + "step": 5110 + }, + { + "epoch": 0.11980988882833149, + "grad_norm": 1.195780634880066, + "learning_rate": 0.00019302334958327003, + "loss": 0.1334, + "step": 5111 + }, + { + "epoch": 0.11983333040313648, + "grad_norm": 3.8701260089874268, + "learning_rate": 0.00019302064618538487, + "loss": 0.257, + "step": 5112 + }, + { + "epoch": 0.11985677197794148, + "grad_norm": 1.9872376918792725, + "learning_rate": 0.00019301794228276706, + "loss": 0.2492, + "step": 5113 + }, + { + "epoch": 0.11988021355274647, + "grad_norm": 1.5443023443222046, + "learning_rate": 0.00019301523787543126, + "loss": 0.2681, + "step": 5114 + }, + { + "epoch": 0.11990365512755147, + "grad_norm": 2.0612528324127197, + "learning_rate": 0.0001930125329633922, + "loss": 0.2719, + "step": 5115 + }, + { + "epoch": 0.11992709670235646, + "grad_norm": 3.9999759197235107, + "learning_rate": 0.00019300982754666448, + "loss": 0.6544, + "step": 5116 + }, + { + "epoch": 0.11995053827716146, + "grad_norm": 2.379199981689453, + "learning_rate": 0.00019300712162526287, + "loss": 0.2613, + "step": 5117 + }, + { + "epoch": 0.11997397985196645, + "grad_norm": 5.262441635131836, + "learning_rate": 0.00019300441519920196, + "loss": 0.9653, + "step": 5118 + }, + { + "epoch": 0.11999742142677144, + "grad_norm": 3.1451375484466553, + "learning_rate": 0.0001930017082684965, + "loss": 0.2881, + "step": 5119 + }, + { + "epoch": 0.12002086300157645, + "grad_norm": 4.297578811645508, + "learning_rate": 0.00019299900083316114, + "loss": 0.6238, + "step": 5120 + }, + { + "epoch": 0.12004430457638145, + "grad_norm": 3.1761655807495117, + "learning_rate": 0.0001929962928932106, + "loss": 1.669, + "step": 5121 + }, + { + "epoch": 0.12006774615118644, + "grad_norm": 1.576858639717102, + "learning_rate": 0.00019299358444865953, + "loss": 0.2315, + "step": 5122 + }, + { + "epoch": 0.12009118772599144, + "grad_norm": 2.9734385013580322, + "learning_rate": 0.0001929908754995227, + "loss": 0.3186, + "step": 5123 + }, + { + "epoch": 0.12011462930079643, + "grad_norm": 4.327017307281494, + "learning_rate": 0.00019298816604581474, + "loss": 0.5671, + "step": 5124 + }, + { + "epoch": 0.12013807087560142, + "grad_norm": 4.790115833282471, + "learning_rate": 0.00019298545608755039, + "loss": 0.5424, + "step": 5125 + }, + { + "epoch": 0.12016151245040642, + "grad_norm": 2.119192361831665, + "learning_rate": 0.00019298274562474434, + "loss": 1.2097, + "step": 5126 + }, + { + "epoch": 0.12018495402521141, + "grad_norm": 3.314409017562866, + "learning_rate": 0.00019298003465741128, + "loss": 0.2823, + "step": 5127 + }, + { + "epoch": 0.12020839560001641, + "grad_norm": 0.3819439113140106, + "learning_rate": 0.00019297732318556595, + "loss": 0.0403, + "step": 5128 + }, + { + "epoch": 0.1202318371748214, + "grad_norm": 0.897767961025238, + "learning_rate": 0.00019297461120922304, + "loss": 0.1349, + "step": 5129 + }, + { + "epoch": 0.1202552787496264, + "grad_norm": 1.6853077411651611, + "learning_rate": 0.0001929718987283973, + "loss": 0.2305, + "step": 5130 + }, + { + "epoch": 0.12027872032443139, + "grad_norm": 4.378454685211182, + "learning_rate": 0.00019296918574310338, + "loss": 0.564, + "step": 5131 + }, + { + "epoch": 0.12030216189923638, + "grad_norm": 3.363821268081665, + "learning_rate": 0.00019296647225335609, + "loss": 0.4897, + "step": 5132 + }, + { + "epoch": 0.12032560347404139, + "grad_norm": 1.2593390941619873, + "learning_rate": 0.00019296375825917007, + "loss": 0.1428, + "step": 5133 + }, + { + "epoch": 0.12034904504884639, + "grad_norm": 2.773955821990967, + "learning_rate": 0.00019296104376056013, + "loss": 0.286, + "step": 5134 + }, + { + "epoch": 0.12037248662365138, + "grad_norm": 4.088048458099365, + "learning_rate": 0.00019295832875754093, + "loss": 0.6538, + "step": 5135 + }, + { + "epoch": 0.12039592819845638, + "grad_norm": 1.4108229875564575, + "learning_rate": 0.00019295561325012722, + "loss": 0.1975, + "step": 5136 + }, + { + "epoch": 0.12041936977326137, + "grad_norm": 1.6852904558181763, + "learning_rate": 0.00019295289723833374, + "loss": 0.3625, + "step": 5137 + }, + { + "epoch": 0.12044281134806636, + "grad_norm": 1.7299243211746216, + "learning_rate": 0.0001929501807221752, + "loss": 0.9395, + "step": 5138 + }, + { + "epoch": 0.12046625292287136, + "grad_norm": 2.7829477787017822, + "learning_rate": 0.00019294746370166637, + "loss": 0.3675, + "step": 5139 + }, + { + "epoch": 0.12048969449767635, + "grad_norm": 4.1792826652526855, + "learning_rate": 0.00019294474617682203, + "loss": 0.4841, + "step": 5140 + }, + { + "epoch": 0.12051313607248135, + "grad_norm": 4.758777141571045, + "learning_rate": 0.00019294202814765688, + "loss": 0.4429, + "step": 5141 + }, + { + "epoch": 0.12053657764728634, + "grad_norm": 4.210111618041992, + "learning_rate": 0.00019293930961418565, + "loss": 0.3811, + "step": 5142 + }, + { + "epoch": 0.12056001922209134, + "grad_norm": 2.5281410217285156, + "learning_rate": 0.0001929365905764231, + "loss": 0.3196, + "step": 5143 + }, + { + "epoch": 0.12058346079689633, + "grad_norm": 2.227980613708496, + "learning_rate": 0.00019293387103438403, + "loss": 0.2374, + "step": 5144 + }, + { + "epoch": 0.12060690237170132, + "grad_norm": 4.450150489807129, + "learning_rate": 0.00019293115098808315, + "loss": 0.7716, + "step": 5145 + }, + { + "epoch": 0.12063034394650633, + "grad_norm": 5.166140556335449, + "learning_rate": 0.0001929284304375352, + "loss": 0.7439, + "step": 5146 + }, + { + "epoch": 0.12065378552131133, + "grad_norm": 4.566688060760498, + "learning_rate": 0.00019292570938275503, + "loss": 0.4822, + "step": 5147 + }, + { + "epoch": 0.12067722709611632, + "grad_norm": 3.59098482131958, + "learning_rate": 0.00019292298782375733, + "loss": 0.6024, + "step": 5148 + }, + { + "epoch": 0.12070066867092132, + "grad_norm": 3.957559823989868, + "learning_rate": 0.00019292026576055686, + "loss": 0.4273, + "step": 5149 + }, + { + "epoch": 0.12072411024572631, + "grad_norm": 1.9218089580535889, + "learning_rate": 0.00019291754319316842, + "loss": 0.1933, + "step": 5150 + }, + { + "epoch": 0.1207475518205313, + "grad_norm": 1.492562174797058, + "learning_rate": 0.0001929148201216068, + "loss": 0.1297, + "step": 5151 + }, + { + "epoch": 0.1207709933953363, + "grad_norm": 4.189767837524414, + "learning_rate": 0.00019291209654588674, + "loss": 0.5454, + "step": 5152 + }, + { + "epoch": 0.12079443497014129, + "grad_norm": 5.319922924041748, + "learning_rate": 0.00019290937246602306, + "loss": 0.6125, + "step": 5153 + }, + { + "epoch": 0.12081787654494629, + "grad_norm": 3.860521078109741, + "learning_rate": 0.00019290664788203045, + "loss": 0.4903, + "step": 5154 + }, + { + "epoch": 0.12084131811975128, + "grad_norm": 4.556212902069092, + "learning_rate": 0.00019290392279392381, + "loss": 0.9447, + "step": 5155 + }, + { + "epoch": 0.12086475969455628, + "grad_norm": 2.789879322052002, + "learning_rate": 0.00019290119720171783, + "loss": 2.246, + "step": 5156 + }, + { + "epoch": 0.12088820126936127, + "grad_norm": 1.4807645082473755, + "learning_rate": 0.0001928984711054274, + "loss": 0.2386, + "step": 5157 + }, + { + "epoch": 0.12091164284416626, + "grad_norm": 2.6204960346221924, + "learning_rate": 0.0001928957445050672, + "loss": 0.4431, + "step": 5158 + }, + { + "epoch": 0.12093508441897127, + "grad_norm": 3.5653843879699707, + "learning_rate": 0.0001928930174006521, + "loss": 0.3532, + "step": 5159 + }, + { + "epoch": 0.12095852599377627, + "grad_norm": 2.0286967754364014, + "learning_rate": 0.0001928902897921969, + "loss": 1.2228, + "step": 5160 + }, + { + "epoch": 0.12098196756858126, + "grad_norm": 3.25285267829895, + "learning_rate": 0.00019288756167971634, + "loss": 0.4596, + "step": 5161 + }, + { + "epoch": 0.12100540914338626, + "grad_norm": 2.287423849105835, + "learning_rate": 0.00019288483306322527, + "loss": 1.5116, + "step": 5162 + }, + { + "epoch": 0.12102885071819125, + "grad_norm": 0.8856050968170166, + "learning_rate": 0.00019288210394273846, + "loss": 0.1181, + "step": 5163 + }, + { + "epoch": 0.12105229229299624, + "grad_norm": 1.305045247077942, + "learning_rate": 0.0001928793743182708, + "loss": 0.1911, + "step": 5164 + }, + { + "epoch": 0.12107573386780124, + "grad_norm": 0.9116883873939514, + "learning_rate": 0.000192876644189837, + "loss": 0.107, + "step": 5165 + }, + { + "epoch": 0.12109917544260623, + "grad_norm": 4.960405349731445, + "learning_rate": 0.0001928739135574519, + "loss": 0.4129, + "step": 5166 + }, + { + "epoch": 0.12112261701741123, + "grad_norm": 0.5697091221809387, + "learning_rate": 0.00019287118242113037, + "loss": 0.0723, + "step": 5167 + }, + { + "epoch": 0.12114605859221622, + "grad_norm": 3.7665319442749023, + "learning_rate": 0.0001928684507808872, + "loss": 0.282, + "step": 5168 + }, + { + "epoch": 0.12116950016702122, + "grad_norm": 2.9131691455841064, + "learning_rate": 0.00019286571863673716, + "loss": 0.3855, + "step": 5169 + }, + { + "epoch": 0.12119294174182621, + "grad_norm": 3.374995708465576, + "learning_rate": 0.00019286298598869517, + "loss": 0.4642, + "step": 5170 + }, + { + "epoch": 0.1212163833166312, + "grad_norm": 5.649351119995117, + "learning_rate": 0.00019286025283677598, + "loss": 0.623, + "step": 5171 + }, + { + "epoch": 0.12123982489143621, + "grad_norm": 3.404935836791992, + "learning_rate": 0.00019285751918099445, + "loss": 0.5212, + "step": 5172 + }, + { + "epoch": 0.1212632664662412, + "grad_norm": 3.3708930015563965, + "learning_rate": 0.00019285478502136544, + "loss": 0.6463, + "step": 5173 + }, + { + "epoch": 0.1212867080410462, + "grad_norm": 3.2551937103271484, + "learning_rate": 0.0001928520503579037, + "loss": 2.0811, + "step": 5174 + }, + { + "epoch": 0.1213101496158512, + "grad_norm": 7.410305976867676, + "learning_rate": 0.00019284931519062415, + "loss": 0.7384, + "step": 5175 + }, + { + "epoch": 0.12133359119065619, + "grad_norm": 0.3876854181289673, + "learning_rate": 0.00019284657951954158, + "loss": 0.0436, + "step": 5176 + }, + { + "epoch": 0.12135703276546118, + "grad_norm": 2.01578688621521, + "learning_rate": 0.00019284384334467092, + "loss": 0.2312, + "step": 5177 + }, + { + "epoch": 0.12138047434026618, + "grad_norm": 1.6954429149627686, + "learning_rate": 0.00019284110666602691, + "loss": 0.1512, + "step": 5178 + }, + { + "epoch": 0.12140391591507117, + "grad_norm": 4.295651912689209, + "learning_rate": 0.00019283836948362447, + "loss": 0.7374, + "step": 5179 + }, + { + "epoch": 0.12142735748987617, + "grad_norm": 2.535520553588867, + "learning_rate": 0.00019283563179747838, + "loss": 1.8605, + "step": 5180 + }, + { + "epoch": 0.12145079906468116, + "grad_norm": 1.6472972631454468, + "learning_rate": 0.0001928328936076036, + "loss": 0.3986, + "step": 5181 + }, + { + "epoch": 0.12147424063948616, + "grad_norm": 4.5939621925354, + "learning_rate": 0.0001928301549140149, + "loss": 0.8078, + "step": 5182 + }, + { + "epoch": 0.12149768221429115, + "grad_norm": 2.7061970233917236, + "learning_rate": 0.00019282741571672717, + "loss": 0.2986, + "step": 5183 + }, + { + "epoch": 0.12152112378909616, + "grad_norm": 3.621101140975952, + "learning_rate": 0.00019282467601575528, + "loss": 0.4387, + "step": 5184 + }, + { + "epoch": 0.12154456536390115, + "grad_norm": 4.839608669281006, + "learning_rate": 0.00019282193581111407, + "loss": 0.7152, + "step": 5185 + }, + { + "epoch": 0.12156800693870615, + "grad_norm": 3.84578275680542, + "learning_rate": 0.00019281919510281842, + "loss": 0.7588, + "step": 5186 + }, + { + "epoch": 0.12159144851351114, + "grad_norm": 2.578991413116455, + "learning_rate": 0.00019281645389088324, + "loss": 0.3611, + "step": 5187 + }, + { + "epoch": 0.12161489008831614, + "grad_norm": 0.8778987526893616, + "learning_rate": 0.0001928137121753234, + "loss": 0.188, + "step": 5188 + }, + { + "epoch": 0.12163833166312113, + "grad_norm": 3.3273746967315674, + "learning_rate": 0.0001928109699561537, + "loss": 0.463, + "step": 5189 + }, + { + "epoch": 0.12166177323792612, + "grad_norm": 2.9136760234832764, + "learning_rate": 0.00019280822723338905, + "loss": 0.4939, + "step": 5190 + }, + { + "epoch": 0.12168521481273112, + "grad_norm": 3.758068799972534, + "learning_rate": 0.00019280548400704437, + "loss": 0.5672, + "step": 5191 + }, + { + "epoch": 0.12170865638753611, + "grad_norm": 1.1193772554397583, + "learning_rate": 0.00019280274027713452, + "loss": 0.1312, + "step": 5192 + }, + { + "epoch": 0.12173209796234111, + "grad_norm": 3.0786330699920654, + "learning_rate": 0.00019279999604367441, + "loss": 0.5575, + "step": 5193 + }, + { + "epoch": 0.1217555395371461, + "grad_norm": 0.7963334918022156, + "learning_rate": 0.0001927972513066789, + "loss": 0.1193, + "step": 5194 + }, + { + "epoch": 0.1217789811119511, + "grad_norm": 3.1034231185913086, + "learning_rate": 0.0001927945060661629, + "loss": 0.4037, + "step": 5195 + }, + { + "epoch": 0.12180242268675609, + "grad_norm": 2.755701780319214, + "learning_rate": 0.00019279176032214128, + "loss": 0.3544, + "step": 5196 + }, + { + "epoch": 0.1218258642615611, + "grad_norm": 4.993815898895264, + "learning_rate": 0.00019278901407462898, + "loss": 1.0353, + "step": 5197 + }, + { + "epoch": 0.12184930583636609, + "grad_norm": 2.139310121536255, + "learning_rate": 0.00019278626732364086, + "loss": 1.4109, + "step": 5198 + }, + { + "epoch": 0.12187274741117109, + "grad_norm": 4.790261268615723, + "learning_rate": 0.00019278352006919187, + "loss": 0.5347, + "step": 5199 + }, + { + "epoch": 0.12189618898597608, + "grad_norm": 2.485617160797119, + "learning_rate": 0.0001927807723112969, + "loss": 0.3292, + "step": 5200 + }, + { + "epoch": 0.12191963056078108, + "grad_norm": 4.998195648193359, + "learning_rate": 0.00019277802404997084, + "loss": 0.7162, + "step": 5201 + }, + { + "epoch": 0.12194307213558607, + "grad_norm": 1.8175958395004272, + "learning_rate": 0.0001927752752852286, + "loss": 0.2835, + "step": 5202 + }, + { + "epoch": 0.12196651371039106, + "grad_norm": 5.033877849578857, + "learning_rate": 0.0001927725260170851, + "loss": 0.5232, + "step": 5203 + }, + { + "epoch": 0.12198995528519606, + "grad_norm": 3.718238592147827, + "learning_rate": 0.00019276977624555527, + "loss": 0.4689, + "step": 5204 + }, + { + "epoch": 0.12201339686000105, + "grad_norm": 3.9318318367004395, + "learning_rate": 0.00019276702597065403, + "loss": 0.4647, + "step": 5205 + }, + { + "epoch": 0.12203683843480605, + "grad_norm": 2.8560664653778076, + "learning_rate": 0.00019276427519239632, + "loss": 0.3053, + "step": 5206 + }, + { + "epoch": 0.12206028000961104, + "grad_norm": 1.7433125972747803, + "learning_rate": 0.00019276152391079702, + "loss": 0.1619, + "step": 5207 + }, + { + "epoch": 0.12208372158441604, + "grad_norm": 1.9423115253448486, + "learning_rate": 0.0001927587721258711, + "loss": 1.1014, + "step": 5208 + }, + { + "epoch": 0.12210716315922103, + "grad_norm": 2.647308111190796, + "learning_rate": 0.00019275601983763347, + "loss": 0.4099, + "step": 5209 + }, + { + "epoch": 0.12213060473402604, + "grad_norm": 1.2918505668640137, + "learning_rate": 0.00019275326704609906, + "loss": 0.1506, + "step": 5210 + }, + { + "epoch": 0.12215404630883103, + "grad_norm": 4.118402004241943, + "learning_rate": 0.00019275051375128283, + "loss": 0.6498, + "step": 5211 + }, + { + "epoch": 0.12217748788363603, + "grad_norm": 3.870096445083618, + "learning_rate": 0.00019274775995319966, + "loss": 0.5328, + "step": 5212 + }, + { + "epoch": 0.12220092945844102, + "grad_norm": 2.4277024269104004, + "learning_rate": 0.00019274500565186458, + "loss": 0.2805, + "step": 5213 + }, + { + "epoch": 0.12222437103324602, + "grad_norm": 2.3907172679901123, + "learning_rate": 0.0001927422508472925, + "loss": 1.7803, + "step": 5214 + }, + { + "epoch": 0.12224781260805101, + "grad_norm": 3.099332571029663, + "learning_rate": 0.00019273949553949834, + "loss": 0.3895, + "step": 5215 + }, + { + "epoch": 0.122271254182856, + "grad_norm": 2.0672945976257324, + "learning_rate": 0.00019273673972849706, + "loss": 1.604, + "step": 5216 + }, + { + "epoch": 0.122294695757661, + "grad_norm": 2.6970133781433105, + "learning_rate": 0.00019273398341430363, + "loss": 0.4251, + "step": 5217 + }, + { + "epoch": 0.12231813733246599, + "grad_norm": 1.2014873027801514, + "learning_rate": 0.00019273122659693303, + "loss": 0.1434, + "step": 5218 + }, + { + "epoch": 0.12234157890727099, + "grad_norm": 3.40950345993042, + "learning_rate": 0.00019272846927640013, + "loss": 0.2444, + "step": 5219 + }, + { + "epoch": 0.12236502048207598, + "grad_norm": 3.579690933227539, + "learning_rate": 0.00019272571145272, + "loss": 0.4961, + "step": 5220 + }, + { + "epoch": 0.12238846205688098, + "grad_norm": 2.4778764247894287, + "learning_rate": 0.0001927229531259075, + "loss": 0.2107, + "step": 5221 + }, + { + "epoch": 0.12241190363168597, + "grad_norm": 4.047978401184082, + "learning_rate": 0.00019272019429597767, + "loss": 0.6039, + "step": 5222 + }, + { + "epoch": 0.12243534520649098, + "grad_norm": 3.642988681793213, + "learning_rate": 0.00019271743496294548, + "loss": 0.7073, + "step": 5223 + }, + { + "epoch": 0.12245878678129597, + "grad_norm": 4.79019021987915, + "learning_rate": 0.0001927146751268259, + "loss": 0.5087, + "step": 5224 + }, + { + "epoch": 0.12248222835610097, + "grad_norm": 3.69994854927063, + "learning_rate": 0.0001927119147876338, + "loss": 0.5604, + "step": 5225 + }, + { + "epoch": 0.12250566993090596, + "grad_norm": 4.527801036834717, + "learning_rate": 0.00019270915394538433, + "loss": 0.4915, + "step": 5226 + }, + { + "epoch": 0.12252911150571096, + "grad_norm": 4.350104808807373, + "learning_rate": 0.00019270639260009236, + "loss": 0.3399, + "step": 5227 + }, + { + "epoch": 0.12255255308051595, + "grad_norm": 0.5001720190048218, + "learning_rate": 0.0001927036307517729, + "loss": 0.0496, + "step": 5228 + }, + { + "epoch": 0.12257599465532094, + "grad_norm": 4.213543891906738, + "learning_rate": 0.0001927008684004409, + "loss": 0.709, + "step": 5229 + }, + { + "epoch": 0.12259943623012594, + "grad_norm": 4.350273132324219, + "learning_rate": 0.00019269810554611142, + "loss": 0.6009, + "step": 5230 + }, + { + "epoch": 0.12262287780493093, + "grad_norm": 3.9800422191619873, + "learning_rate": 0.00019269534218879937, + "loss": 0.4795, + "step": 5231 + }, + { + "epoch": 0.12264631937973593, + "grad_norm": 4.882312774658203, + "learning_rate": 0.00019269257832851984, + "loss": 0.8452, + "step": 5232 + }, + { + "epoch": 0.12266976095454092, + "grad_norm": 3.245232582092285, + "learning_rate": 0.00019268981396528773, + "loss": 0.3043, + "step": 5233 + }, + { + "epoch": 0.12269320252934592, + "grad_norm": 5.960949420928955, + "learning_rate": 0.0001926870490991181, + "loss": 0.8027, + "step": 5234 + }, + { + "epoch": 0.12271664410415092, + "grad_norm": 2.8760838508605957, + "learning_rate": 0.00019268428373002595, + "loss": 0.3903, + "step": 5235 + }, + { + "epoch": 0.12274008567895592, + "grad_norm": 4.307777404785156, + "learning_rate": 0.00019268151785802623, + "loss": 0.6016, + "step": 5236 + }, + { + "epoch": 0.12276352725376091, + "grad_norm": 3.215859889984131, + "learning_rate": 0.00019267875148313402, + "loss": 0.2711, + "step": 5237 + }, + { + "epoch": 0.1227869688285659, + "grad_norm": 1.8986161947250366, + "learning_rate": 0.0001926759846053643, + "loss": 1.5314, + "step": 5238 + }, + { + "epoch": 0.1228104104033709, + "grad_norm": 2.8011109828948975, + "learning_rate": 0.00019267321722473207, + "loss": 0.274, + "step": 5239 + }, + { + "epoch": 0.1228338519781759, + "grad_norm": 4.05674934387207, + "learning_rate": 0.00019267044934125237, + "loss": 0.9318, + "step": 5240 + }, + { + "epoch": 0.12285729355298089, + "grad_norm": 3.744093894958496, + "learning_rate": 0.00019266768095494018, + "loss": 0.5232, + "step": 5241 + }, + { + "epoch": 0.12288073512778588, + "grad_norm": 0.6755735874176025, + "learning_rate": 0.00019266491206581057, + "loss": 0.0823, + "step": 5242 + }, + { + "epoch": 0.12290417670259088, + "grad_norm": 2.131330966949463, + "learning_rate": 0.00019266214267387854, + "loss": 0.2683, + "step": 5243 + }, + { + "epoch": 0.12292761827739587, + "grad_norm": 2.1826062202453613, + "learning_rate": 0.0001926593727791591, + "loss": 0.3837, + "step": 5244 + }, + { + "epoch": 0.12295105985220087, + "grad_norm": 1.8280352354049683, + "learning_rate": 0.00019265660238166734, + "loss": 0.2625, + "step": 5245 + }, + { + "epoch": 0.12297450142700586, + "grad_norm": 1.1126728057861328, + "learning_rate": 0.00019265383148141824, + "loss": 0.1079, + "step": 5246 + }, + { + "epoch": 0.12299794300181086, + "grad_norm": 2.9702351093292236, + "learning_rate": 0.0001926510600784268, + "loss": 0.3431, + "step": 5247 + }, + { + "epoch": 0.12302138457661586, + "grad_norm": 3.055699586868286, + "learning_rate": 0.00019264828817270812, + "loss": 0.3688, + "step": 5248 + }, + { + "epoch": 0.12304482615142086, + "grad_norm": 3.7910351753234863, + "learning_rate": 0.00019264551576427726, + "loss": 0.4981, + "step": 5249 + }, + { + "epoch": 0.12306826772622585, + "grad_norm": 3.6829631328582764, + "learning_rate": 0.0001926427428531492, + "loss": 0.5188, + "step": 5250 + }, + { + "epoch": 0.12309170930103085, + "grad_norm": 1.6352310180664062, + "learning_rate": 0.000192639969439339, + "loss": 0.2282, + "step": 5251 + }, + { + "epoch": 0.12311515087583584, + "grad_norm": 3.2726333141326904, + "learning_rate": 0.00019263719552286171, + "loss": 0.3147, + "step": 5252 + }, + { + "epoch": 0.12313859245064084, + "grad_norm": 1.3662350177764893, + "learning_rate": 0.00019263442110373244, + "loss": 0.2247, + "step": 5253 + }, + { + "epoch": 0.12316203402544583, + "grad_norm": 3.4750494956970215, + "learning_rate": 0.0001926316461819662, + "loss": 0.5959, + "step": 5254 + }, + { + "epoch": 0.12318547560025082, + "grad_norm": 6.212866306304932, + "learning_rate": 0.000192628870757578, + "loss": 0.4664, + "step": 5255 + }, + { + "epoch": 0.12320891717505582, + "grad_norm": 4.687678337097168, + "learning_rate": 0.00019262609483058297, + "loss": 0.4026, + "step": 5256 + }, + { + "epoch": 0.12323235874986081, + "grad_norm": 2.3445677757263184, + "learning_rate": 0.0001926233184009961, + "loss": 0.3013, + "step": 5257 + }, + { + "epoch": 0.1232558003246658, + "grad_norm": 4.680914402008057, + "learning_rate": 0.00019262054146883256, + "loss": 0.3207, + "step": 5258 + }, + { + "epoch": 0.1232792418994708, + "grad_norm": 5.248358249664307, + "learning_rate": 0.00019261776403410732, + "loss": 0.9452, + "step": 5259 + }, + { + "epoch": 0.1233026834742758, + "grad_norm": 2.188157796859741, + "learning_rate": 0.0001926149860968355, + "loss": 0.3162, + "step": 5260 + }, + { + "epoch": 0.1233261250490808, + "grad_norm": 1.2918038368225098, + "learning_rate": 0.00019261220765703216, + "loss": 0.1477, + "step": 5261 + }, + { + "epoch": 0.1233495666238858, + "grad_norm": 4.630037784576416, + "learning_rate": 0.00019260942871471238, + "loss": 0.4185, + "step": 5262 + }, + { + "epoch": 0.12337300819869079, + "grad_norm": 1.3669320344924927, + "learning_rate": 0.00019260664926989125, + "loss": 0.2461, + "step": 5263 + }, + { + "epoch": 0.12339644977349579, + "grad_norm": 2.9373996257781982, + "learning_rate": 0.00019260386932258382, + "loss": 0.3791, + "step": 5264 + }, + { + "epoch": 0.12341989134830078, + "grad_norm": 2.6919045448303223, + "learning_rate": 0.0001926010888728052, + "loss": 0.3015, + "step": 5265 + }, + { + "epoch": 0.12344333292310578, + "grad_norm": 8.836000442504883, + "learning_rate": 0.00019259830792057046, + "loss": 0.4971, + "step": 5266 + }, + { + "epoch": 0.12346677449791077, + "grad_norm": 3.1254990100860596, + "learning_rate": 0.0001925955264658947, + "loss": 0.706, + "step": 5267 + }, + { + "epoch": 0.12349021607271576, + "grad_norm": 4.013479232788086, + "learning_rate": 0.000192592744508793, + "loss": 0.2054, + "step": 5268 + }, + { + "epoch": 0.12351365764752076, + "grad_norm": 4.851461887359619, + "learning_rate": 0.00019258996204928047, + "loss": 0.4641, + "step": 5269 + }, + { + "epoch": 0.12353709922232575, + "grad_norm": 3.986480474472046, + "learning_rate": 0.00019258717908737221, + "loss": 0.5927, + "step": 5270 + }, + { + "epoch": 0.12356054079713075, + "grad_norm": 1.7406206130981445, + "learning_rate": 0.00019258439562308332, + "loss": 0.1915, + "step": 5271 + }, + { + "epoch": 0.12358398237193574, + "grad_norm": 3.858194351196289, + "learning_rate": 0.00019258161165642887, + "loss": 0.5579, + "step": 5272 + }, + { + "epoch": 0.12360742394674074, + "grad_norm": 2.3911867141723633, + "learning_rate": 0.00019257882718742402, + "loss": 0.3635, + "step": 5273 + }, + { + "epoch": 0.12363086552154574, + "grad_norm": 1.6596949100494385, + "learning_rate": 0.00019257604221608384, + "loss": 0.2988, + "step": 5274 + }, + { + "epoch": 0.12365430709635074, + "grad_norm": 4.292860507965088, + "learning_rate": 0.00019257325674242344, + "loss": 0.2921, + "step": 5275 + }, + { + "epoch": 0.12367774867115573, + "grad_norm": 3.6226701736450195, + "learning_rate": 0.00019257047076645794, + "loss": 0.244, + "step": 5276 + }, + { + "epoch": 0.12370119024596073, + "grad_norm": 3.362299680709839, + "learning_rate": 0.00019256768428820248, + "loss": 0.1815, + "step": 5277 + }, + { + "epoch": 0.12372463182076572, + "grad_norm": 3.6526987552642822, + "learning_rate": 0.00019256489730767217, + "loss": 0.4422, + "step": 5278 + }, + { + "epoch": 0.12374807339557072, + "grad_norm": 3.57814359664917, + "learning_rate": 0.00019256210982488212, + "loss": 0.4591, + "step": 5279 + }, + { + "epoch": 0.12377151497037571, + "grad_norm": 2.326932668685913, + "learning_rate": 0.00019255932183984747, + "loss": 1.308, + "step": 5280 + }, + { + "epoch": 0.1237949565451807, + "grad_norm": 3.3705389499664307, + "learning_rate": 0.0001925565333525833, + "loss": 0.2571, + "step": 5281 + }, + { + "epoch": 0.1238183981199857, + "grad_norm": 8.304683685302734, + "learning_rate": 0.0001925537443631048, + "loss": 0.7654, + "step": 5282 + }, + { + "epoch": 0.12384183969479069, + "grad_norm": 1.3623539209365845, + "learning_rate": 0.0001925509548714271, + "loss": 0.1807, + "step": 5283 + }, + { + "epoch": 0.12386528126959569, + "grad_norm": 4.835574150085449, + "learning_rate": 0.00019254816487756528, + "loss": 0.9697, + "step": 5284 + }, + { + "epoch": 0.12388872284440068, + "grad_norm": 4.481378078460693, + "learning_rate": 0.00019254537438153456, + "loss": 0.3999, + "step": 5285 + }, + { + "epoch": 0.12391216441920568, + "grad_norm": 3.6992568969726562, + "learning_rate": 0.00019254258338334998, + "loss": 0.5586, + "step": 5286 + }, + { + "epoch": 0.12393560599401068, + "grad_norm": 3.520524740219116, + "learning_rate": 0.00019253979188302677, + "loss": 0.5638, + "step": 5287 + }, + { + "epoch": 0.12395904756881568, + "grad_norm": 4.502151012420654, + "learning_rate": 0.00019253699988058004, + "loss": 0.3382, + "step": 5288 + }, + { + "epoch": 0.12398248914362067, + "grad_norm": 1.9231741428375244, + "learning_rate": 0.00019253420737602497, + "loss": 0.1669, + "step": 5289 + }, + { + "epoch": 0.12400593071842567, + "grad_norm": 3.237022876739502, + "learning_rate": 0.00019253141436937666, + "loss": 0.3865, + "step": 5290 + }, + { + "epoch": 0.12402937229323066, + "grad_norm": 2.547663927078247, + "learning_rate": 0.00019252862086065032, + "loss": 1.3809, + "step": 5291 + }, + { + "epoch": 0.12405281386803566, + "grad_norm": 4.1524739265441895, + "learning_rate": 0.00019252582684986104, + "loss": 0.3562, + "step": 5292 + }, + { + "epoch": 0.12407625544284065, + "grad_norm": 3.3114588260650635, + "learning_rate": 0.00019252303233702406, + "loss": 0.4103, + "step": 5293 + }, + { + "epoch": 0.12409969701764564, + "grad_norm": 5.032567501068115, + "learning_rate": 0.0001925202373221545, + "loss": 0.4424, + "step": 5294 + }, + { + "epoch": 0.12412313859245064, + "grad_norm": 4.425757884979248, + "learning_rate": 0.00019251744180526754, + "loss": 0.3452, + "step": 5295 + }, + { + "epoch": 0.12414658016725563, + "grad_norm": 4.779648780822754, + "learning_rate": 0.00019251464578637833, + "loss": 0.5728, + "step": 5296 + }, + { + "epoch": 0.12417002174206063, + "grad_norm": 3.4437198638916016, + "learning_rate": 0.00019251184926550203, + "loss": 0.5232, + "step": 5297 + }, + { + "epoch": 0.12419346331686562, + "grad_norm": 3.7553603649139404, + "learning_rate": 0.00019250905224265386, + "loss": 0.3036, + "step": 5298 + }, + { + "epoch": 0.12421690489167063, + "grad_norm": 2.7803714275360107, + "learning_rate": 0.00019250625471784895, + "loss": 0.2007, + "step": 5299 + }, + { + "epoch": 0.12424034646647562, + "grad_norm": 0.920725405216217, + "learning_rate": 0.0001925034566911025, + "loss": 0.1198, + "step": 5300 + }, + { + "epoch": 0.12426378804128062, + "grad_norm": 3.416349172592163, + "learning_rate": 0.00019250065816242974, + "loss": 0.3984, + "step": 5301 + }, + { + "epoch": 0.12428722961608561, + "grad_norm": 1.537530541419983, + "learning_rate": 0.00019249785913184574, + "loss": 0.2493, + "step": 5302 + }, + { + "epoch": 0.1243106711908906, + "grad_norm": 2.1825904846191406, + "learning_rate": 0.0001924950595993658, + "loss": 1.4382, + "step": 5303 + }, + { + "epoch": 0.1243341127656956, + "grad_norm": 4.035836219787598, + "learning_rate": 0.00019249225956500504, + "loss": 0.4469, + "step": 5304 + }, + { + "epoch": 0.1243575543405006, + "grad_norm": 3.261247158050537, + "learning_rate": 0.0001924894590287787, + "loss": 0.4724, + "step": 5305 + }, + { + "epoch": 0.12438099591530559, + "grad_norm": 3.077909231185913, + "learning_rate": 0.00019248665799070195, + "loss": 0.4071, + "step": 5306 + }, + { + "epoch": 0.12440443749011058, + "grad_norm": 3.770190715789795, + "learning_rate": 0.00019248385645079, + "loss": 0.7282, + "step": 5307 + }, + { + "epoch": 0.12442787906491558, + "grad_norm": 4.710041046142578, + "learning_rate": 0.00019248105440905803, + "loss": 0.7317, + "step": 5308 + }, + { + "epoch": 0.12445132063972057, + "grad_norm": 2.4447712898254395, + "learning_rate": 0.00019247825186552127, + "loss": 0.3891, + "step": 5309 + }, + { + "epoch": 0.12447476221452557, + "grad_norm": 1.4915785789489746, + "learning_rate": 0.0001924754488201949, + "loss": 0.2134, + "step": 5310 + }, + { + "epoch": 0.12449820378933056, + "grad_norm": 2.5004796981811523, + "learning_rate": 0.00019247264527309416, + "loss": 0.3806, + "step": 5311 + }, + { + "epoch": 0.12452164536413557, + "grad_norm": 3.514031171798706, + "learning_rate": 0.00019246984122423425, + "loss": 0.4169, + "step": 5312 + }, + { + "epoch": 0.12454508693894056, + "grad_norm": 3.498211622238159, + "learning_rate": 0.00019246703667363035, + "loss": 0.4925, + "step": 5313 + }, + { + "epoch": 0.12456852851374556, + "grad_norm": 1.9950660467147827, + "learning_rate": 0.00019246423162129776, + "loss": 0.1773, + "step": 5314 + }, + { + "epoch": 0.12459197008855055, + "grad_norm": 1.9366044998168945, + "learning_rate": 0.00019246142606725163, + "loss": 0.2659, + "step": 5315 + }, + { + "epoch": 0.12461541166335555, + "grad_norm": 3.837714672088623, + "learning_rate": 0.00019245862001150718, + "loss": 0.5505, + "step": 5316 + }, + { + "epoch": 0.12463885323816054, + "grad_norm": 3.15505313873291, + "learning_rate": 0.00019245581345407966, + "loss": 0.331, + "step": 5317 + }, + { + "epoch": 0.12466229481296554, + "grad_norm": 3.701467275619507, + "learning_rate": 0.00019245300639498432, + "loss": 0.4755, + "step": 5318 + }, + { + "epoch": 0.12468573638777053, + "grad_norm": 5.560680866241455, + "learning_rate": 0.00019245019883423634, + "loss": 0.5563, + "step": 5319 + }, + { + "epoch": 0.12470917796257552, + "grad_norm": 3.6290957927703857, + "learning_rate": 0.00019244739077185102, + "loss": 0.6461, + "step": 5320 + }, + { + "epoch": 0.12473261953738052, + "grad_norm": 2.424398183822632, + "learning_rate": 0.00019244458220784352, + "loss": 1.0575, + "step": 5321 + }, + { + "epoch": 0.12475606111218551, + "grad_norm": 4.703105449676514, + "learning_rate": 0.00019244177314222912, + "loss": 0.6553, + "step": 5322 + }, + { + "epoch": 0.1247795026869905, + "grad_norm": 1.3302853107452393, + "learning_rate": 0.0001924389635750231, + "loss": 0.2049, + "step": 5323 + }, + { + "epoch": 0.1248029442617955, + "grad_norm": 1.1996370553970337, + "learning_rate": 0.00019243615350624064, + "loss": 0.1762, + "step": 5324 + }, + { + "epoch": 0.12482638583660051, + "grad_norm": 0.5169705748558044, + "learning_rate": 0.00019243334293589698, + "loss": 0.0361, + "step": 5325 + }, + { + "epoch": 0.1248498274114055, + "grad_norm": 3.7529640197753906, + "learning_rate": 0.00019243053186400742, + "loss": 0.4384, + "step": 5326 + }, + { + "epoch": 0.1248732689862105, + "grad_norm": 5.5800323486328125, + "learning_rate": 0.00019242772029058722, + "loss": 0.7309, + "step": 5327 + }, + { + "epoch": 0.12489671056101549, + "grad_norm": 4.4570393562316895, + "learning_rate": 0.0001924249082156516, + "loss": 0.732, + "step": 5328 + }, + { + "epoch": 0.12492015213582049, + "grad_norm": 3.6024703979492188, + "learning_rate": 0.00019242209563921585, + "loss": 0.4852, + "step": 5329 + }, + { + "epoch": 0.12494359371062548, + "grad_norm": 4.4734601974487305, + "learning_rate": 0.00019241928256129515, + "loss": 0.5414, + "step": 5330 + }, + { + "epoch": 0.12496703528543048, + "grad_norm": 3.0069661140441895, + "learning_rate": 0.00019241646898190487, + "loss": 0.3018, + "step": 5331 + }, + { + "epoch": 0.12499047686023547, + "grad_norm": 2.1102778911590576, + "learning_rate": 0.00019241365490106022, + "loss": 0.3179, + "step": 5332 + }, + { + "epoch": 0.12501391843504048, + "grad_norm": 0.44966790080070496, + "learning_rate": 0.00019241084031877649, + "loss": 0.0659, + "step": 5333 + }, + { + "epoch": 0.12503736000984547, + "grad_norm": 1.0394254922866821, + "learning_rate": 0.00019240802523506892, + "loss": 0.109, + "step": 5334 + }, + { + "epoch": 0.12506080158465047, + "grad_norm": 3.3454015254974365, + "learning_rate": 0.00019240520964995285, + "loss": 0.4337, + "step": 5335 + }, + { + "epoch": 0.12508424315945546, + "grad_norm": 4.076927185058594, + "learning_rate": 0.00019240239356344346, + "loss": 0.486, + "step": 5336 + }, + { + "epoch": 0.12510768473426045, + "grad_norm": 2.6283349990844727, + "learning_rate": 0.00019239957697555612, + "loss": 0.3411, + "step": 5337 + }, + { + "epoch": 0.12513112630906545, + "grad_norm": 4.575443744659424, + "learning_rate": 0.00019239675988630607, + "loss": 0.4236, + "step": 5338 + }, + { + "epoch": 0.12515456788387044, + "grad_norm": 2.4547863006591797, + "learning_rate": 0.00019239394229570862, + "loss": 0.2396, + "step": 5339 + }, + { + "epoch": 0.12517800945867544, + "grad_norm": 8.545140266418457, + "learning_rate": 0.000192391124203779, + "loss": 0.8049, + "step": 5340 + }, + { + "epoch": 0.12520145103348043, + "grad_norm": 0.7436292171478271, + "learning_rate": 0.00019238830561053258, + "loss": 0.1014, + "step": 5341 + }, + { + "epoch": 0.12522489260828543, + "grad_norm": 1.6275453567504883, + "learning_rate": 0.0001923854865159846, + "loss": 0.2599, + "step": 5342 + }, + { + "epoch": 0.12524833418309042, + "grad_norm": 3.677664041519165, + "learning_rate": 0.00019238266692015037, + "loss": 0.5631, + "step": 5343 + }, + { + "epoch": 0.12527177575789541, + "grad_norm": 3.406972885131836, + "learning_rate": 0.0001923798468230452, + "loss": 0.5769, + "step": 5344 + }, + { + "epoch": 0.1252952173327004, + "grad_norm": 2.396353006362915, + "learning_rate": 0.00019237702622468438, + "loss": 0.4002, + "step": 5345 + }, + { + "epoch": 0.1253186589075054, + "grad_norm": 3.6800618171691895, + "learning_rate": 0.0001923742051250832, + "loss": 0.7351, + "step": 5346 + }, + { + "epoch": 0.1253421004823104, + "grad_norm": 3.823559522628784, + "learning_rate": 0.00019237138352425702, + "loss": 0.4264, + "step": 5347 + }, + { + "epoch": 0.1253655420571154, + "grad_norm": 1.0808395147323608, + "learning_rate": 0.00019236856142222108, + "loss": 0.2494, + "step": 5348 + }, + { + "epoch": 0.1253889836319204, + "grad_norm": 3.1814355850219727, + "learning_rate": 0.00019236573881899075, + "loss": 1.4185, + "step": 5349 + }, + { + "epoch": 0.12541242520672538, + "grad_norm": 3.7996373176574707, + "learning_rate": 0.0001923629157145813, + "loss": 0.6165, + "step": 5350 + }, + { + "epoch": 0.12543586678153038, + "grad_norm": 3.786621332168579, + "learning_rate": 0.00019236009210900812, + "loss": 0.5585, + "step": 5351 + }, + { + "epoch": 0.12545930835633537, + "grad_norm": 6.094004154205322, + "learning_rate": 0.00019235726800228645, + "loss": 0.8011, + "step": 5352 + }, + { + "epoch": 0.12548274993114036, + "grad_norm": 1.0710312128067017, + "learning_rate": 0.00019235444339443163, + "loss": 0.2086, + "step": 5353 + }, + { + "epoch": 0.12550619150594536, + "grad_norm": 0.6860663294792175, + "learning_rate": 0.00019235161828545905, + "loss": 0.0874, + "step": 5354 + }, + { + "epoch": 0.12552963308075035, + "grad_norm": 1.8904749155044556, + "learning_rate": 0.00019234879267538395, + "loss": 0.1961, + "step": 5355 + }, + { + "epoch": 0.12555307465555535, + "grad_norm": 0.6701537370681763, + "learning_rate": 0.00019234596656422172, + "loss": 0.0839, + "step": 5356 + }, + { + "epoch": 0.12557651623036037, + "grad_norm": 3.903346538543701, + "learning_rate": 0.00019234313995198767, + "loss": 0.4336, + "step": 5357 + }, + { + "epoch": 0.12559995780516536, + "grad_norm": 4.893550872802734, + "learning_rate": 0.00019234031283869719, + "loss": 0.6452, + "step": 5358 + }, + { + "epoch": 0.12562339937997036, + "grad_norm": 5.178609371185303, + "learning_rate": 0.00019233748522436551, + "loss": 0.8171, + "step": 5359 + }, + { + "epoch": 0.12564684095477535, + "grad_norm": 0.7552732229232788, + "learning_rate": 0.00019233465710900808, + "loss": 0.0836, + "step": 5360 + }, + { + "epoch": 0.12567028252958035, + "grad_norm": 3.5889956951141357, + "learning_rate": 0.00019233182849264023, + "loss": 0.3628, + "step": 5361 + }, + { + "epoch": 0.12569372410438534, + "grad_norm": 6.1767377853393555, + "learning_rate": 0.00019232899937527721, + "loss": 0.8156, + "step": 5362 + }, + { + "epoch": 0.12571716567919033, + "grad_norm": 0.9692939519882202, + "learning_rate": 0.0001923261697569345, + "loss": 0.086, + "step": 5363 + }, + { + "epoch": 0.12574060725399533, + "grad_norm": 5.853118419647217, + "learning_rate": 0.0001923233396376274, + "loss": 0.9274, + "step": 5364 + }, + { + "epoch": 0.12576404882880032, + "grad_norm": 2.7545080184936523, + "learning_rate": 0.00019232050901737122, + "loss": 0.2608, + "step": 5365 + }, + { + "epoch": 0.12578749040360532, + "grad_norm": 2.42246413230896, + "learning_rate": 0.0001923176778961814, + "loss": 0.0664, + "step": 5366 + }, + { + "epoch": 0.1258109319784103, + "grad_norm": 1.4056663513183594, + "learning_rate": 0.00019231484627407324, + "loss": 0.1479, + "step": 5367 + }, + { + "epoch": 0.1258343735532153, + "grad_norm": 1.820221185684204, + "learning_rate": 0.00019231201415106212, + "loss": 1.9392, + "step": 5368 + }, + { + "epoch": 0.1258578151280203, + "grad_norm": 4.191267490386963, + "learning_rate": 0.00019230918152716344, + "loss": 0.341, + "step": 5369 + }, + { + "epoch": 0.1258812567028253, + "grad_norm": 4.264024257659912, + "learning_rate": 0.00019230634840239255, + "loss": 0.5663, + "step": 5370 + }, + { + "epoch": 0.1259046982776303, + "grad_norm": 5.076449394226074, + "learning_rate": 0.0001923035147767648, + "loss": 0.4036, + "step": 5371 + }, + { + "epoch": 0.12592813985243528, + "grad_norm": 3.5913422107696533, + "learning_rate": 0.00019230068065029558, + "loss": 0.4468, + "step": 5372 + }, + { + "epoch": 0.12595158142724028, + "grad_norm": 3.2174971103668213, + "learning_rate": 0.0001922978460230003, + "loss": 1.4069, + "step": 5373 + }, + { + "epoch": 0.12597502300204527, + "grad_norm": 6.23480749130249, + "learning_rate": 0.0001922950108948943, + "loss": 0.894, + "step": 5374 + }, + { + "epoch": 0.12599846457685027, + "grad_norm": 3.201329231262207, + "learning_rate": 0.00019229217526599296, + "loss": 0.3981, + "step": 5375 + }, + { + "epoch": 0.12602190615165526, + "grad_norm": 0.6499501466751099, + "learning_rate": 0.00019228933913631167, + "loss": 0.1101, + "step": 5376 + }, + { + "epoch": 0.12604534772646026, + "grad_norm": 3.296085834503174, + "learning_rate": 0.00019228650250586587, + "loss": 0.5652, + "step": 5377 + }, + { + "epoch": 0.12606878930126525, + "grad_norm": 1.4293179512023926, + "learning_rate": 0.0001922836653746709, + "loss": 0.1629, + "step": 5378 + }, + { + "epoch": 0.12609223087607024, + "grad_norm": 5.1960673332214355, + "learning_rate": 0.00019228082774274216, + "loss": 0.4848, + "step": 5379 + }, + { + "epoch": 0.12611567245087524, + "grad_norm": 2.735008955001831, + "learning_rate": 0.00019227798961009507, + "loss": 0.3597, + "step": 5380 + }, + { + "epoch": 0.12613911402568023, + "grad_norm": 13.298147201538086, + "learning_rate": 0.00019227515097674497, + "loss": 0.4192, + "step": 5381 + }, + { + "epoch": 0.12616255560048525, + "grad_norm": 3.7266626358032227, + "learning_rate": 0.00019227231184270734, + "loss": 0.4973, + "step": 5382 + }, + { + "epoch": 0.12618599717529025, + "grad_norm": 1.7521320581436157, + "learning_rate": 0.00019226947220799754, + "loss": 0.1719, + "step": 5383 + }, + { + "epoch": 0.12620943875009524, + "grad_norm": 4.106724739074707, + "learning_rate": 0.00019226663207263102, + "loss": 0.5785, + "step": 5384 + }, + { + "epoch": 0.12623288032490024, + "grad_norm": 6.246294021606445, + "learning_rate": 0.00019226379143662313, + "loss": 0.5414, + "step": 5385 + }, + { + "epoch": 0.12625632189970523, + "grad_norm": 2.1994242668151855, + "learning_rate": 0.0001922609502999893, + "loss": 0.1582, + "step": 5386 + }, + { + "epoch": 0.12627976347451023, + "grad_norm": 4.247507095336914, + "learning_rate": 0.00019225810866274498, + "loss": 0.5986, + "step": 5387 + }, + { + "epoch": 0.12630320504931522, + "grad_norm": 4.24169921875, + "learning_rate": 0.0001922552665249056, + "loss": 0.7005, + "step": 5388 + }, + { + "epoch": 0.12632664662412021, + "grad_norm": 0.9371757507324219, + "learning_rate": 0.0001922524238864865, + "loss": 0.1003, + "step": 5389 + }, + { + "epoch": 0.1263500881989252, + "grad_norm": 3.8460466861724854, + "learning_rate": 0.0001922495807475032, + "loss": 0.4531, + "step": 5390 + }, + { + "epoch": 0.1263735297737302, + "grad_norm": 2.4550275802612305, + "learning_rate": 0.00019224673710797104, + "loss": 0.2566, + "step": 5391 + }, + { + "epoch": 0.1263969713485352, + "grad_norm": 4.325490951538086, + "learning_rate": 0.0001922438929679055, + "loss": 0.3344, + "step": 5392 + }, + { + "epoch": 0.1264204129233402, + "grad_norm": 5.494930744171143, + "learning_rate": 0.000192241048327322, + "loss": 0.4367, + "step": 5393 + }, + { + "epoch": 0.1264438544981452, + "grad_norm": 1.0995805263519287, + "learning_rate": 0.00019223820318623597, + "loss": 0.1715, + "step": 5394 + }, + { + "epoch": 0.12646729607295018, + "grad_norm": 1.3811357021331787, + "learning_rate": 0.0001922353575446629, + "loss": 0.1794, + "step": 5395 + }, + { + "epoch": 0.12649073764775517, + "grad_norm": 3.379024028778076, + "learning_rate": 0.00019223251140261814, + "loss": 0.5266, + "step": 5396 + }, + { + "epoch": 0.12651417922256017, + "grad_norm": 0.8586273193359375, + "learning_rate": 0.0001922296647601172, + "loss": 0.1217, + "step": 5397 + }, + { + "epoch": 0.12653762079736516, + "grad_norm": 4.490147113800049, + "learning_rate": 0.0001922268176171755, + "loss": 0.6464, + "step": 5398 + }, + { + "epoch": 0.12656106237217016, + "grad_norm": 4.93589448928833, + "learning_rate": 0.00019222396997380851, + "loss": 0.8969, + "step": 5399 + }, + { + "epoch": 0.12658450394697515, + "grad_norm": 1.573250651359558, + "learning_rate": 0.00019222112183003165, + "loss": 0.1888, + "step": 5400 + }, + { + "epoch": 0.12660794552178015, + "grad_norm": 1.5946646928787231, + "learning_rate": 0.00019221827318586038, + "loss": 0.1988, + "step": 5401 + }, + { + "epoch": 0.12663138709658514, + "grad_norm": 3.3362648487091064, + "learning_rate": 0.00019221542404131018, + "loss": 0.5068, + "step": 5402 + }, + { + "epoch": 0.12665482867139014, + "grad_norm": 2.3907670974731445, + "learning_rate": 0.0001922125743963965, + "loss": 0.1203, + "step": 5403 + }, + { + "epoch": 0.12667827024619513, + "grad_norm": 2.0780975818634033, + "learning_rate": 0.0001922097242511348, + "loss": 0.1881, + "step": 5404 + }, + { + "epoch": 0.12670171182100012, + "grad_norm": 3.0166561603546143, + "learning_rate": 0.00019220687360554055, + "loss": 0.4375, + "step": 5405 + }, + { + "epoch": 0.12672515339580512, + "grad_norm": 4.712376594543457, + "learning_rate": 0.00019220402245962917, + "loss": 0.703, + "step": 5406 + }, + { + "epoch": 0.1267485949706101, + "grad_norm": 4.006109714508057, + "learning_rate": 0.00019220117081341623, + "loss": 0.64, + "step": 5407 + }, + { + "epoch": 0.12677203654541513, + "grad_norm": 4.844362258911133, + "learning_rate": 0.0001921983186669171, + "loss": 0.6448, + "step": 5408 + }, + { + "epoch": 0.12679547812022013, + "grad_norm": 3.672222852706909, + "learning_rate": 0.00019219546602014732, + "loss": 0.5646, + "step": 5409 + }, + { + "epoch": 0.12681891969502512, + "grad_norm": 1.855161190032959, + "learning_rate": 0.00019219261287312236, + "loss": 0.2054, + "step": 5410 + }, + { + "epoch": 0.12684236126983012, + "grad_norm": 3.878720760345459, + "learning_rate": 0.00019218975922585766, + "loss": 0.6729, + "step": 5411 + }, + { + "epoch": 0.1268658028446351, + "grad_norm": 3.714399576187134, + "learning_rate": 0.00019218690507836876, + "loss": 0.3961, + "step": 5412 + }, + { + "epoch": 0.1268892444194401, + "grad_norm": 4.445343494415283, + "learning_rate": 0.00019218405043067112, + "loss": 0.7446, + "step": 5413 + }, + { + "epoch": 0.1269126859942451, + "grad_norm": 3.4789113998413086, + "learning_rate": 0.0001921811952827802, + "loss": 0.7041, + "step": 5414 + }, + { + "epoch": 0.1269361275690501, + "grad_norm": 0.9722551107406616, + "learning_rate": 0.00019217833963471154, + "loss": 0.0877, + "step": 5415 + }, + { + "epoch": 0.1269595691438551, + "grad_norm": 2.3941781520843506, + "learning_rate": 0.0001921754834864806, + "loss": 0.2513, + "step": 5416 + }, + { + "epoch": 0.12698301071866008, + "grad_norm": 4.063004970550537, + "learning_rate": 0.00019217262683810297, + "loss": 0.5586, + "step": 5417 + }, + { + "epoch": 0.12700645229346508, + "grad_norm": 1.5475122928619385, + "learning_rate": 0.00019216976968959402, + "loss": 0.1479, + "step": 5418 + }, + { + "epoch": 0.12702989386827007, + "grad_norm": 4.524455547332764, + "learning_rate": 0.0001921669120409693, + "loss": 0.5547, + "step": 5419 + }, + { + "epoch": 0.12705333544307507, + "grad_norm": 0.6978267431259155, + "learning_rate": 0.00019216405389224433, + "loss": 0.1156, + "step": 5420 + }, + { + "epoch": 0.12707677701788006, + "grad_norm": 4.479223251342773, + "learning_rate": 0.0001921611952434346, + "loss": 0.7109, + "step": 5421 + }, + { + "epoch": 0.12710021859268505, + "grad_norm": 4.699766635894775, + "learning_rate": 0.00019215833609455567, + "loss": 0.587, + "step": 5422 + }, + { + "epoch": 0.12712366016749005, + "grad_norm": 3.8878333568573, + "learning_rate": 0.000192155476445623, + "loss": 0.462, + "step": 5423 + }, + { + "epoch": 0.12714710174229504, + "grad_norm": 4.1294684410095215, + "learning_rate": 0.0001921526162966521, + "loss": 0.4158, + "step": 5424 + }, + { + "epoch": 0.12717054331710004, + "grad_norm": 4.38166618347168, + "learning_rate": 0.00019214975564765853, + "loss": 0.4363, + "step": 5425 + }, + { + "epoch": 0.12719398489190503, + "grad_norm": 4.556490898132324, + "learning_rate": 0.0001921468944986578, + "loss": 0.7201, + "step": 5426 + }, + { + "epoch": 0.12721742646671003, + "grad_norm": 3.772183418273926, + "learning_rate": 0.00019214403284966542, + "loss": 0.3892, + "step": 5427 + }, + { + "epoch": 0.12724086804151502, + "grad_norm": 3.68218994140625, + "learning_rate": 0.00019214117070069696, + "loss": 0.573, + "step": 5428 + }, + { + "epoch": 0.12726430961632001, + "grad_norm": 4.240017414093018, + "learning_rate": 0.00019213830805176786, + "loss": 0.3736, + "step": 5429 + }, + { + "epoch": 0.127287751191125, + "grad_norm": 3.281435012817383, + "learning_rate": 0.00019213544490289376, + "loss": 0.4958, + "step": 5430 + }, + { + "epoch": 0.12731119276593, + "grad_norm": 3.5968520641326904, + "learning_rate": 0.00019213258125409013, + "loss": 0.6076, + "step": 5431 + }, + { + "epoch": 0.127334634340735, + "grad_norm": 3.3862030506134033, + "learning_rate": 0.0001921297171053725, + "loss": 0.3116, + "step": 5432 + }, + { + "epoch": 0.12735807591554, + "grad_norm": 0.569004237651825, + "learning_rate": 0.00019212685245675645, + "loss": 0.0857, + "step": 5433 + }, + { + "epoch": 0.12738151749034501, + "grad_norm": 1.024612307548523, + "learning_rate": 0.00019212398730825752, + "loss": 0.0911, + "step": 5434 + }, + { + "epoch": 0.12740495906515, + "grad_norm": 3.600463628768921, + "learning_rate": 0.0001921211216598912, + "loss": 0.5055, + "step": 5435 + }, + { + "epoch": 0.127428400639955, + "grad_norm": 4.000565052032471, + "learning_rate": 0.00019211825551167315, + "loss": 0.4539, + "step": 5436 + }, + { + "epoch": 0.12745184221476, + "grad_norm": 3.655919075012207, + "learning_rate": 0.00019211538886361882, + "loss": 2.0098, + "step": 5437 + }, + { + "epoch": 0.127475283789565, + "grad_norm": 0.9119240045547485, + "learning_rate": 0.0001921125217157438, + "loss": 0.1494, + "step": 5438 + }, + { + "epoch": 0.12749872536436999, + "grad_norm": 5.2644524574279785, + "learning_rate": 0.00019210965406806367, + "loss": 0.696, + "step": 5439 + }, + { + "epoch": 0.12752216693917498, + "grad_norm": 4.958714485168457, + "learning_rate": 0.00019210678592059395, + "loss": 0.9401, + "step": 5440 + }, + { + "epoch": 0.12754560851397997, + "grad_norm": 1.237000823020935, + "learning_rate": 0.0001921039172733502, + "loss": 0.1024, + "step": 5441 + }, + { + "epoch": 0.12756905008878497, + "grad_norm": 3.322707176208496, + "learning_rate": 0.00019210104812634804, + "loss": 0.4256, + "step": 5442 + }, + { + "epoch": 0.12759249166358996, + "grad_norm": 1.0368260145187378, + "learning_rate": 0.00019209817847960298, + "loss": 0.1769, + "step": 5443 + }, + { + "epoch": 0.12761593323839496, + "grad_norm": 2.6283226013183594, + "learning_rate": 0.00019209530833313063, + "loss": 0.368, + "step": 5444 + }, + { + "epoch": 0.12763937481319995, + "grad_norm": 4.637679576873779, + "learning_rate": 0.00019209243768694654, + "loss": 0.7071, + "step": 5445 + }, + { + "epoch": 0.12766281638800495, + "grad_norm": 4.594549179077148, + "learning_rate": 0.00019208956654106628, + "loss": 0.5811, + "step": 5446 + }, + { + "epoch": 0.12768625796280994, + "grad_norm": 3.448939323425293, + "learning_rate": 0.00019208669489550548, + "loss": 0.5743, + "step": 5447 + }, + { + "epoch": 0.12770969953761493, + "grad_norm": 4.586919784545898, + "learning_rate": 0.00019208382275027965, + "loss": 0.647, + "step": 5448 + }, + { + "epoch": 0.12773314111241993, + "grad_norm": 3.327876329421997, + "learning_rate": 0.00019208095010540444, + "loss": 0.5431, + "step": 5449 + }, + { + "epoch": 0.12775658268722492, + "grad_norm": 3.999682903289795, + "learning_rate": 0.00019207807696089538, + "loss": 0.4065, + "step": 5450 + }, + { + "epoch": 0.12778002426202992, + "grad_norm": 6.637759208679199, + "learning_rate": 0.00019207520331676806, + "loss": 0.9384, + "step": 5451 + }, + { + "epoch": 0.1278034658368349, + "grad_norm": 4.6070556640625, + "learning_rate": 0.00019207232917303814, + "loss": 0.6109, + "step": 5452 + }, + { + "epoch": 0.1278269074116399, + "grad_norm": 3.226757526397705, + "learning_rate": 0.00019206945452972116, + "loss": 0.3751, + "step": 5453 + }, + { + "epoch": 0.1278503489864449, + "grad_norm": 3.329894542694092, + "learning_rate": 0.00019206657938683271, + "loss": 0.4518, + "step": 5454 + }, + { + "epoch": 0.1278737905612499, + "grad_norm": 3.0750043392181396, + "learning_rate": 0.00019206370374438847, + "loss": 0.3578, + "step": 5455 + }, + { + "epoch": 0.1278972321360549, + "grad_norm": 2.589266777038574, + "learning_rate": 0.00019206082760240393, + "loss": 1.4073, + "step": 5456 + }, + { + "epoch": 0.12792067371085988, + "grad_norm": 2.226141929626465, + "learning_rate": 0.00019205795096089475, + "loss": 0.1968, + "step": 5457 + }, + { + "epoch": 0.12794411528566488, + "grad_norm": 4.323095321655273, + "learning_rate": 0.00019205507381987657, + "loss": 0.7163, + "step": 5458 + }, + { + "epoch": 0.1279675568604699, + "grad_norm": 0.427564799785614, + "learning_rate": 0.00019205219617936497, + "loss": 0.0657, + "step": 5459 + }, + { + "epoch": 0.1279909984352749, + "grad_norm": 1.1487563848495483, + "learning_rate": 0.00019204931803937556, + "loss": 0.1931, + "step": 5460 + }, + { + "epoch": 0.1280144400100799, + "grad_norm": 0.586914598941803, + "learning_rate": 0.00019204643939992394, + "loss": 0.106, + "step": 5461 + }, + { + "epoch": 0.12803788158488488, + "grad_norm": 5.703159332275391, + "learning_rate": 0.00019204356026102576, + "loss": 0.6633, + "step": 5462 + }, + { + "epoch": 0.12806132315968988, + "grad_norm": 2.6330783367156982, + "learning_rate": 0.00019204068062269667, + "loss": 0.3706, + "step": 5463 + }, + { + "epoch": 0.12808476473449487, + "grad_norm": 6.716858386993408, + "learning_rate": 0.0001920378004849522, + "loss": 0.8808, + "step": 5464 + }, + { + "epoch": 0.12810820630929987, + "grad_norm": 3.406663656234741, + "learning_rate": 0.0001920349198478081, + "loss": 0.3587, + "step": 5465 + }, + { + "epoch": 0.12813164788410486, + "grad_norm": 3.4595110416412354, + "learning_rate": 0.00019203203871127992, + "loss": 0.4874, + "step": 5466 + }, + { + "epoch": 0.12815508945890985, + "grad_norm": 1.0873397588729858, + "learning_rate": 0.00019202915707538327, + "loss": 0.1381, + "step": 5467 + }, + { + "epoch": 0.12817853103371485, + "grad_norm": 5.237660884857178, + "learning_rate": 0.00019202627494013385, + "loss": 1.0503, + "step": 5468 + }, + { + "epoch": 0.12820197260851984, + "grad_norm": 2.331836223602295, + "learning_rate": 0.00019202339230554727, + "loss": 0.3457, + "step": 5469 + }, + { + "epoch": 0.12822541418332484, + "grad_norm": 2.8346495628356934, + "learning_rate": 0.0001920205091716392, + "loss": 0.4275, + "step": 5470 + }, + { + "epoch": 0.12824885575812983, + "grad_norm": 3.5788044929504395, + "learning_rate": 0.00019201762553842523, + "loss": 0.3713, + "step": 5471 + }, + { + "epoch": 0.12827229733293483, + "grad_norm": 5.390326976776123, + "learning_rate": 0.00019201474140592107, + "loss": 0.619, + "step": 5472 + }, + { + "epoch": 0.12829573890773982, + "grad_norm": 1.1518428325653076, + "learning_rate": 0.0001920118567741423, + "loss": 0.1575, + "step": 5473 + }, + { + "epoch": 0.12831918048254481, + "grad_norm": 1.1861623525619507, + "learning_rate": 0.00019200897164310462, + "loss": 0.1501, + "step": 5474 + }, + { + "epoch": 0.1283426220573498, + "grad_norm": 1.8583793640136719, + "learning_rate": 0.00019200608601282367, + "loss": 0.2462, + "step": 5475 + }, + { + "epoch": 0.1283660636321548, + "grad_norm": 4.136524200439453, + "learning_rate": 0.0001920031998833151, + "loss": 0.4348, + "step": 5476 + }, + { + "epoch": 0.1283895052069598, + "grad_norm": 1.1678894758224487, + "learning_rate": 0.00019200031325459463, + "loss": 0.1513, + "step": 5477 + }, + { + "epoch": 0.1284129467817648, + "grad_norm": 4.444204330444336, + "learning_rate": 0.00019199742612667785, + "loss": 0.3906, + "step": 5478 + }, + { + "epoch": 0.1284363883565698, + "grad_norm": 4.388692378997803, + "learning_rate": 0.0001919945384995804, + "loss": 0.5147, + "step": 5479 + }, + { + "epoch": 0.12845982993137478, + "grad_norm": 4.92144250869751, + "learning_rate": 0.00019199165037331802, + "loss": 0.4642, + "step": 5480 + }, + { + "epoch": 0.12848327150617977, + "grad_norm": 1.7424904108047485, + "learning_rate": 0.0001919887617479064, + "loss": 0.2154, + "step": 5481 + }, + { + "epoch": 0.12850671308098477, + "grad_norm": 4.374907493591309, + "learning_rate": 0.00019198587262336117, + "loss": 0.9225, + "step": 5482 + }, + { + "epoch": 0.12853015465578976, + "grad_norm": 4.5620880126953125, + "learning_rate": 0.00019198298299969796, + "loss": 1.0714, + "step": 5483 + }, + { + "epoch": 0.12855359623059476, + "grad_norm": 3.4184584617614746, + "learning_rate": 0.00019198009287693253, + "loss": 1.3336, + "step": 5484 + }, + { + "epoch": 0.12857703780539978, + "grad_norm": 4.150393009185791, + "learning_rate": 0.00019197720225508054, + "loss": 0.5684, + "step": 5485 + }, + { + "epoch": 0.12860047938020477, + "grad_norm": 4.973428249359131, + "learning_rate": 0.00019197431113415764, + "loss": 0.7028, + "step": 5486 + }, + { + "epoch": 0.12862392095500977, + "grad_norm": 3.6581084728240967, + "learning_rate": 0.00019197141951417953, + "loss": 0.3798, + "step": 5487 + }, + { + "epoch": 0.12864736252981476, + "grad_norm": 4.891041278839111, + "learning_rate": 0.00019196852739516195, + "loss": 0.6026, + "step": 5488 + }, + { + "epoch": 0.12867080410461976, + "grad_norm": 3.1429882049560547, + "learning_rate": 0.00019196563477712054, + "loss": 1.3597, + "step": 5489 + }, + { + "epoch": 0.12869424567942475, + "grad_norm": 4.066983699798584, + "learning_rate": 0.000191962741660071, + "loss": 0.636, + "step": 5490 + }, + { + "epoch": 0.12871768725422975, + "grad_norm": 0.8664042353630066, + "learning_rate": 0.00019195984804402905, + "loss": 0.11, + "step": 5491 + }, + { + "epoch": 0.12874112882903474, + "grad_norm": 6.244593620300293, + "learning_rate": 0.00019195695392901034, + "loss": 0.7577, + "step": 5492 + }, + { + "epoch": 0.12876457040383973, + "grad_norm": 2.0502333641052246, + "learning_rate": 0.00019195405931503068, + "loss": 0.1445, + "step": 5493 + }, + { + "epoch": 0.12878801197864473, + "grad_norm": 3.355222225189209, + "learning_rate": 0.00019195116420210565, + "loss": 0.3, + "step": 5494 + }, + { + "epoch": 0.12881145355344972, + "grad_norm": 4.019442081451416, + "learning_rate": 0.00019194826859025103, + "loss": 0.4726, + "step": 5495 + }, + { + "epoch": 0.12883489512825472, + "grad_norm": 5.924292087554932, + "learning_rate": 0.00019194537247948254, + "loss": 0.7351, + "step": 5496 + }, + { + "epoch": 0.1288583367030597, + "grad_norm": 0.8884099125862122, + "learning_rate": 0.0001919424758698159, + "loss": 0.1001, + "step": 5497 + }, + { + "epoch": 0.1288817782778647, + "grad_norm": 4.220831394195557, + "learning_rate": 0.00019193957876126674, + "loss": 0.4326, + "step": 5498 + }, + { + "epoch": 0.1289052198526697, + "grad_norm": 2.8175268173217773, + "learning_rate": 0.00019193668115385085, + "loss": 0.2356, + "step": 5499 + }, + { + "epoch": 0.1289286614274747, + "grad_norm": 4.4538798332214355, + "learning_rate": 0.00019193378304758398, + "loss": 0.5329, + "step": 5500 + }, + { + "epoch": 0.1289521030022797, + "grad_norm": 4.678678035736084, + "learning_rate": 0.00019193088444248178, + "loss": 0.8081, + "step": 5501 + }, + { + "epoch": 0.12897554457708468, + "grad_norm": 3.0985004901885986, + "learning_rate": 0.00019192798533856003, + "loss": 0.617, + "step": 5502 + }, + { + "epoch": 0.12899898615188968, + "grad_norm": 0.8740888833999634, + "learning_rate": 0.00019192508573583446, + "loss": 0.2352, + "step": 5503 + }, + { + "epoch": 0.12902242772669467, + "grad_norm": 3.589644193649292, + "learning_rate": 0.0001919221856343208, + "loss": 0.3608, + "step": 5504 + }, + { + "epoch": 0.12904586930149967, + "grad_norm": 0.39207208156585693, + "learning_rate": 0.00019191928503403474, + "loss": 0.0913, + "step": 5505 + }, + { + "epoch": 0.12906931087630466, + "grad_norm": 1.4443398714065552, + "learning_rate": 0.00019191638393499204, + "loss": 0.1337, + "step": 5506 + }, + { + "epoch": 0.12909275245110965, + "grad_norm": 2.699610471725464, + "learning_rate": 0.0001919134823372085, + "loss": 0.32, + "step": 5507 + }, + { + "epoch": 0.12911619402591465, + "grad_norm": 5.211273670196533, + "learning_rate": 0.0001919105802406998, + "loss": 0.556, + "step": 5508 + }, + { + "epoch": 0.12913963560071964, + "grad_norm": 4.040342807769775, + "learning_rate": 0.00019190767764548175, + "loss": 0.5372, + "step": 5509 + }, + { + "epoch": 0.12916307717552467, + "grad_norm": 3.4067165851593018, + "learning_rate": 0.00019190477455157, + "loss": 0.5362, + "step": 5510 + }, + { + "epoch": 0.12918651875032966, + "grad_norm": 1.4297778606414795, + "learning_rate": 0.00019190187095898038, + "loss": 0.1757, + "step": 5511 + }, + { + "epoch": 0.12920996032513465, + "grad_norm": 3.657208204269409, + "learning_rate": 0.00019189896686772865, + "loss": 0.5077, + "step": 5512 + }, + { + "epoch": 0.12923340189993965, + "grad_norm": 2.925689220428467, + "learning_rate": 0.00019189606227783048, + "loss": 1.629, + "step": 5513 + }, + { + "epoch": 0.12925684347474464, + "grad_norm": 1.7877399921417236, + "learning_rate": 0.00019189315718930175, + "loss": 0.2742, + "step": 5514 + }, + { + "epoch": 0.12928028504954964, + "grad_norm": 2.3762214183807373, + "learning_rate": 0.00019189025160215811, + "loss": 0.1289, + "step": 5515 + }, + { + "epoch": 0.12930372662435463, + "grad_norm": 3.5070078372955322, + "learning_rate": 0.00019188734551641546, + "loss": 0.537, + "step": 5516 + }, + { + "epoch": 0.12932716819915963, + "grad_norm": 2.9784092903137207, + "learning_rate": 0.00019188443893208943, + "loss": 0.549, + "step": 5517 + }, + { + "epoch": 0.12935060977396462, + "grad_norm": 6.09552526473999, + "learning_rate": 0.00019188153184919587, + "loss": 0.5007, + "step": 5518 + }, + { + "epoch": 0.12937405134876961, + "grad_norm": 3.510218858718872, + "learning_rate": 0.0001918786242677505, + "loss": 0.476, + "step": 5519 + }, + { + "epoch": 0.1293974929235746, + "grad_norm": 4.726041793823242, + "learning_rate": 0.00019187571618776918, + "loss": 0.5047, + "step": 5520 + }, + { + "epoch": 0.1294209344983796, + "grad_norm": 3.52398943901062, + "learning_rate": 0.00019187280760926762, + "loss": 0.3562, + "step": 5521 + }, + { + "epoch": 0.1294443760731846, + "grad_norm": 5.065188407897949, + "learning_rate": 0.00019186989853226162, + "loss": 0.7721, + "step": 5522 + }, + { + "epoch": 0.1294678176479896, + "grad_norm": 1.6874276399612427, + "learning_rate": 0.000191866988956767, + "loss": 0.9644, + "step": 5523 + }, + { + "epoch": 0.12949125922279459, + "grad_norm": 4.595978260040283, + "learning_rate": 0.00019186407888279947, + "loss": 0.7058, + "step": 5524 + }, + { + "epoch": 0.12951470079759958, + "grad_norm": 3.558330774307251, + "learning_rate": 0.00019186116831037487, + "loss": 0.6699, + "step": 5525 + }, + { + "epoch": 0.12953814237240457, + "grad_norm": 4.620254993438721, + "learning_rate": 0.000191858257239509, + "loss": 0.7759, + "step": 5526 + }, + { + "epoch": 0.12956158394720957, + "grad_norm": 1.1197266578674316, + "learning_rate": 0.00019185534567021763, + "loss": 0.168, + "step": 5527 + }, + { + "epoch": 0.12958502552201456, + "grad_norm": 3.5416014194488525, + "learning_rate": 0.0001918524336025166, + "loss": 0.3108, + "step": 5528 + }, + { + "epoch": 0.12960846709681956, + "grad_norm": 6.969895362854004, + "learning_rate": 0.00019184952103642168, + "loss": 0.8009, + "step": 5529 + }, + { + "epoch": 0.12963190867162455, + "grad_norm": 4.0142316818237305, + "learning_rate": 0.00019184660797194866, + "loss": 0.7207, + "step": 5530 + }, + { + "epoch": 0.12965535024642955, + "grad_norm": 5.189810276031494, + "learning_rate": 0.00019184369440911335, + "loss": 0.4852, + "step": 5531 + }, + { + "epoch": 0.12967879182123454, + "grad_norm": 2.5031614303588867, + "learning_rate": 0.0001918407803479316, + "loss": 1.782, + "step": 5532 + }, + { + "epoch": 0.12970223339603953, + "grad_norm": 1.4851930141448975, + "learning_rate": 0.0001918378657884192, + "loss": 0.2424, + "step": 5533 + }, + { + "epoch": 0.12972567497084453, + "grad_norm": 3.777930736541748, + "learning_rate": 0.00019183495073059193, + "loss": 0.4084, + "step": 5534 + }, + { + "epoch": 0.12974911654564952, + "grad_norm": 2.7937259674072266, + "learning_rate": 0.00019183203517446562, + "loss": 0.4701, + "step": 5535 + }, + { + "epoch": 0.12977255812045455, + "grad_norm": 1.4345884323120117, + "learning_rate": 0.00019182911912005615, + "loss": 0.3258, + "step": 5536 + }, + { + "epoch": 0.12979599969525954, + "grad_norm": 4.827003002166748, + "learning_rate": 0.0001918262025673793, + "loss": 0.6917, + "step": 5537 + }, + { + "epoch": 0.12981944127006453, + "grad_norm": 3.106877326965332, + "learning_rate": 0.00019182328551645082, + "loss": 0.4333, + "step": 5538 + }, + { + "epoch": 0.12984288284486953, + "grad_norm": 2.8981919288635254, + "learning_rate": 0.00019182036796728668, + "loss": 0.3363, + "step": 5539 + }, + { + "epoch": 0.12986632441967452, + "grad_norm": 1.9344927072525024, + "learning_rate": 0.00019181744991990263, + "loss": 0.2045, + "step": 5540 + }, + { + "epoch": 0.12988976599447952, + "grad_norm": 3.814289093017578, + "learning_rate": 0.0001918145313743145, + "loss": 0.9279, + "step": 5541 + }, + { + "epoch": 0.1299132075692845, + "grad_norm": 2.3187289237976074, + "learning_rate": 0.00019181161233053815, + "loss": 0.3777, + "step": 5542 + }, + { + "epoch": 0.1299366491440895, + "grad_norm": 3.052093982696533, + "learning_rate": 0.00019180869278858943, + "loss": 0.4438, + "step": 5543 + }, + { + "epoch": 0.1299600907188945, + "grad_norm": 2.97876238822937, + "learning_rate": 0.00019180577274848412, + "loss": 0.4438, + "step": 5544 + }, + { + "epoch": 0.1299835322936995, + "grad_norm": 3.760000228881836, + "learning_rate": 0.00019180285221023812, + "loss": 0.5996, + "step": 5545 + }, + { + "epoch": 0.1300069738685045, + "grad_norm": 2.786653757095337, + "learning_rate": 0.00019179993117386727, + "loss": 0.2851, + "step": 5546 + }, + { + "epoch": 0.13003041544330948, + "grad_norm": 2.654981851577759, + "learning_rate": 0.0001917970096393874, + "loss": 0.3061, + "step": 5547 + }, + { + "epoch": 0.13005385701811448, + "grad_norm": 4.587459564208984, + "learning_rate": 0.00019179408760681442, + "loss": 0.6746, + "step": 5548 + }, + { + "epoch": 0.13007729859291947, + "grad_norm": 4.097029685974121, + "learning_rate": 0.00019179116507616408, + "loss": 0.6019, + "step": 5549 + }, + { + "epoch": 0.13010074016772447, + "grad_norm": 1.7740784883499146, + "learning_rate": 0.00019178824204745234, + "loss": 0.203, + "step": 5550 + }, + { + "epoch": 0.13012418174252946, + "grad_norm": 3.8810908794403076, + "learning_rate": 0.000191785318520695, + "loss": 0.3934, + "step": 5551 + }, + { + "epoch": 0.13014762331733445, + "grad_norm": 3.701183795928955, + "learning_rate": 0.00019178239449590793, + "loss": 0.6661, + "step": 5552 + }, + { + "epoch": 0.13017106489213945, + "grad_norm": 2.2086246013641357, + "learning_rate": 0.00019177946997310703, + "loss": 0.3287, + "step": 5553 + }, + { + "epoch": 0.13019450646694444, + "grad_norm": 3.882302761077881, + "learning_rate": 0.00019177654495230813, + "loss": 0.4118, + "step": 5554 + }, + { + "epoch": 0.13021794804174944, + "grad_norm": 6.014747619628906, + "learning_rate": 0.00019177361943352713, + "loss": 0.6931, + "step": 5555 + }, + { + "epoch": 0.13024138961655443, + "grad_norm": 4.652583599090576, + "learning_rate": 0.00019177069341677988, + "loss": 0.7962, + "step": 5556 + }, + { + "epoch": 0.13026483119135943, + "grad_norm": 5.7193284034729, + "learning_rate": 0.00019176776690208228, + "loss": 0.6703, + "step": 5557 + }, + { + "epoch": 0.13028827276616442, + "grad_norm": 4.165899276733398, + "learning_rate": 0.00019176483988945017, + "loss": 0.6323, + "step": 5558 + }, + { + "epoch": 0.13031171434096941, + "grad_norm": 3.958939790725708, + "learning_rate": 0.00019176191237889946, + "loss": 0.4141, + "step": 5559 + }, + { + "epoch": 0.1303351559157744, + "grad_norm": 3.5286407470703125, + "learning_rate": 0.00019175898437044607, + "loss": 0.3333, + "step": 5560 + }, + { + "epoch": 0.13035859749057943, + "grad_norm": 4.277728080749512, + "learning_rate": 0.0001917560558641058, + "loss": 0.6183, + "step": 5561 + }, + { + "epoch": 0.13038203906538443, + "grad_norm": 1.3482986688613892, + "learning_rate": 0.00019175312685989462, + "loss": 0.1399, + "step": 5562 + }, + { + "epoch": 0.13040548064018942, + "grad_norm": 2.2094311714172363, + "learning_rate": 0.0001917501973578284, + "loss": 1.2112, + "step": 5563 + }, + { + "epoch": 0.1304289222149944, + "grad_norm": 4.057412624359131, + "learning_rate": 0.00019174726735792302, + "loss": 0.3292, + "step": 5564 + }, + { + "epoch": 0.1304523637897994, + "grad_norm": 2.850083112716675, + "learning_rate": 0.00019174433686019437, + "loss": 0.53, + "step": 5565 + }, + { + "epoch": 0.1304758053646044, + "grad_norm": 2.620043992996216, + "learning_rate": 0.00019174140586465842, + "loss": 0.3237, + "step": 5566 + }, + { + "epoch": 0.1304992469394094, + "grad_norm": 3.3741626739501953, + "learning_rate": 0.00019173847437133098, + "loss": 1.6643, + "step": 5567 + }, + { + "epoch": 0.1305226885142144, + "grad_norm": 1.8358029127120972, + "learning_rate": 0.000191735542380228, + "loss": 0.2534, + "step": 5568 + }, + { + "epoch": 0.13054613008901939, + "grad_norm": 5.055637359619141, + "learning_rate": 0.0001917326098913654, + "loss": 0.8507, + "step": 5569 + }, + { + "epoch": 0.13056957166382438, + "grad_norm": 3.0446863174438477, + "learning_rate": 0.00019172967690475907, + "loss": 0.3996, + "step": 5570 + }, + { + "epoch": 0.13059301323862937, + "grad_norm": 6.241936683654785, + "learning_rate": 0.00019172674342042494, + "loss": 0.5428, + "step": 5571 + }, + { + "epoch": 0.13061645481343437, + "grad_norm": 4.812552452087402, + "learning_rate": 0.0001917238094383789, + "loss": 0.4082, + "step": 5572 + }, + { + "epoch": 0.13063989638823936, + "grad_norm": 3.031972885131836, + "learning_rate": 0.00019172087495863694, + "loss": 0.413, + "step": 5573 + }, + { + "epoch": 0.13066333796304436, + "grad_norm": 5.822864055633545, + "learning_rate": 0.00019171793998121487, + "loss": 0.6329, + "step": 5574 + }, + { + "epoch": 0.13068677953784935, + "grad_norm": 0.7075181007385254, + "learning_rate": 0.00019171500450612875, + "loss": 0.095, + "step": 5575 + }, + { + "epoch": 0.13071022111265435, + "grad_norm": 4.9719109535217285, + "learning_rate": 0.0001917120685333944, + "loss": 0.6781, + "step": 5576 + }, + { + "epoch": 0.13073366268745934, + "grad_norm": 1.3048603534698486, + "learning_rate": 0.00019170913206302775, + "loss": 0.5313, + "step": 5577 + }, + { + "epoch": 0.13075710426226433, + "grad_norm": 2.9399425983428955, + "learning_rate": 0.00019170619509504484, + "loss": 1.942, + "step": 5578 + }, + { + "epoch": 0.13078054583706933, + "grad_norm": 3.9560670852661133, + "learning_rate": 0.0001917032576294615, + "loss": 0.3871, + "step": 5579 + }, + { + "epoch": 0.13080398741187432, + "grad_norm": 4.1618781089782715, + "learning_rate": 0.0001917003196662937, + "loss": 0.2741, + "step": 5580 + }, + { + "epoch": 0.13082742898667932, + "grad_norm": 3.2054738998413086, + "learning_rate": 0.00019169738120555741, + "loss": 0.3058, + "step": 5581 + }, + { + "epoch": 0.1308508705614843, + "grad_norm": 2.4579172134399414, + "learning_rate": 0.00019169444224726855, + "loss": 0.3889, + "step": 5582 + }, + { + "epoch": 0.1308743121362893, + "grad_norm": 1.2322782278060913, + "learning_rate": 0.00019169150279144305, + "loss": 0.1674, + "step": 5583 + }, + { + "epoch": 0.1308977537110943, + "grad_norm": 1.5184987783432007, + "learning_rate": 0.00019168856283809687, + "loss": 0.137, + "step": 5584 + }, + { + "epoch": 0.1309211952858993, + "grad_norm": 0.7840717434883118, + "learning_rate": 0.000191685622387246, + "loss": 0.0606, + "step": 5585 + }, + { + "epoch": 0.1309446368607043, + "grad_norm": 0.8870894312858582, + "learning_rate": 0.00019168268143890632, + "loss": 0.1048, + "step": 5586 + }, + { + "epoch": 0.1309680784355093, + "grad_norm": 0.7572756409645081, + "learning_rate": 0.0001916797399930939, + "loss": 0.043, + "step": 5587 + }, + { + "epoch": 0.1309915200103143, + "grad_norm": 2.6466686725616455, + "learning_rate": 0.0001916767980498246, + "loss": 0.2252, + "step": 5588 + }, + { + "epoch": 0.1310149615851193, + "grad_norm": 6.009920597076416, + "learning_rate": 0.0001916738556091144, + "loss": 0.9155, + "step": 5589 + }, + { + "epoch": 0.1310384031599243, + "grad_norm": 3.4649972915649414, + "learning_rate": 0.0001916709126709793, + "loss": 0.523, + "step": 5590 + }, + { + "epoch": 0.1310618447347293, + "grad_norm": 3.8324801921844482, + "learning_rate": 0.00019166796923543523, + "loss": 0.6562, + "step": 5591 + }, + { + "epoch": 0.13108528630953428, + "grad_norm": 3.3295531272888184, + "learning_rate": 0.0001916650253024982, + "loss": 0.5101, + "step": 5592 + }, + { + "epoch": 0.13110872788433928, + "grad_norm": 3.267829656600952, + "learning_rate": 0.00019166208087218416, + "loss": 0.4832, + "step": 5593 + }, + { + "epoch": 0.13113216945914427, + "grad_norm": 3.824313163757324, + "learning_rate": 0.0001916591359445091, + "loss": 0.605, + "step": 5594 + }, + { + "epoch": 0.13115561103394927, + "grad_norm": 3.5503499507904053, + "learning_rate": 0.00019165619051948897, + "loss": 0.5085, + "step": 5595 + }, + { + "epoch": 0.13117905260875426, + "grad_norm": 3.4979989528656006, + "learning_rate": 0.0001916532445971398, + "loss": 0.2254, + "step": 5596 + }, + { + "epoch": 0.13120249418355925, + "grad_norm": 3.6161110401153564, + "learning_rate": 0.00019165029817747752, + "loss": 0.4156, + "step": 5597 + }, + { + "epoch": 0.13122593575836425, + "grad_norm": 0.3125230073928833, + "learning_rate": 0.00019164735126051815, + "loss": 0.0262, + "step": 5598 + }, + { + "epoch": 0.13124937733316924, + "grad_norm": 5.243563175201416, + "learning_rate": 0.0001916444038462777, + "loss": 0.8783, + "step": 5599 + }, + { + "epoch": 0.13127281890797424, + "grad_norm": 1.1601189374923706, + "learning_rate": 0.00019164145593477213, + "loss": 0.1704, + "step": 5600 + }, + { + "epoch": 0.13129626048277923, + "grad_norm": 3.5892436504364014, + "learning_rate": 0.0001916385075260174, + "loss": 0.4507, + "step": 5601 + }, + { + "epoch": 0.13131970205758423, + "grad_norm": 5.1601691246032715, + "learning_rate": 0.00019163555862002963, + "loss": 0.5578, + "step": 5602 + }, + { + "epoch": 0.13134314363238922, + "grad_norm": 13.136245727539062, + "learning_rate": 0.00019163260921682468, + "loss": 0.4348, + "step": 5603 + }, + { + "epoch": 0.13136658520719421, + "grad_norm": 6.162680625915527, + "learning_rate": 0.00019162965931641862, + "loss": 0.5412, + "step": 5604 + }, + { + "epoch": 0.1313900267819992, + "grad_norm": 3.3306920528411865, + "learning_rate": 0.0001916267089188275, + "loss": 0.2699, + "step": 5605 + }, + { + "epoch": 0.1314134683568042, + "grad_norm": 4.922996997833252, + "learning_rate": 0.0001916237580240672, + "loss": 0.6185, + "step": 5606 + }, + { + "epoch": 0.1314369099316092, + "grad_norm": 2.772913694381714, + "learning_rate": 0.0001916208066321539, + "loss": 0.3881, + "step": 5607 + }, + { + "epoch": 0.1314603515064142, + "grad_norm": 1.0837359428405762, + "learning_rate": 0.00019161785474310346, + "loss": 0.1976, + "step": 5608 + }, + { + "epoch": 0.13148379308121919, + "grad_norm": 5.177427291870117, + "learning_rate": 0.000191614902356932, + "loss": 0.5851, + "step": 5609 + }, + { + "epoch": 0.13150723465602418, + "grad_norm": 3.93945050239563, + "learning_rate": 0.0001916119494736555, + "loss": 0.6964, + "step": 5610 + }, + { + "epoch": 0.13153067623082917, + "grad_norm": 4.097017288208008, + "learning_rate": 0.00019160899609329, + "loss": 0.379, + "step": 5611 + }, + { + "epoch": 0.13155411780563417, + "grad_norm": 3.5362987518310547, + "learning_rate": 0.0001916060422158515, + "loss": 0.5782, + "step": 5612 + }, + { + "epoch": 0.1315775593804392, + "grad_norm": 5.49861478805542, + "learning_rate": 0.00019160308784135603, + "loss": 0.6003, + "step": 5613 + }, + { + "epoch": 0.13160100095524419, + "grad_norm": 0.8242112398147583, + "learning_rate": 0.00019160013296981964, + "loss": 0.0682, + "step": 5614 + }, + { + "epoch": 0.13162444253004918, + "grad_norm": 2.7527148723602295, + "learning_rate": 0.00019159717760125835, + "loss": 0.4691, + "step": 5615 + }, + { + "epoch": 0.13164788410485417, + "grad_norm": 2.744248628616333, + "learning_rate": 0.00019159422173568817, + "loss": 0.2422, + "step": 5616 + }, + { + "epoch": 0.13167132567965917, + "grad_norm": 3.2670247554779053, + "learning_rate": 0.00019159126537312518, + "loss": 0.3648, + "step": 5617 + }, + { + "epoch": 0.13169476725446416, + "grad_norm": 3.9253203868865967, + "learning_rate": 0.00019158830851358544, + "loss": 0.5299, + "step": 5618 + }, + { + "epoch": 0.13171820882926916, + "grad_norm": 2.7981395721435547, + "learning_rate": 0.00019158535115708497, + "loss": 1.7332, + "step": 5619 + }, + { + "epoch": 0.13174165040407415, + "grad_norm": 7.760585308074951, + "learning_rate": 0.00019158239330363977, + "loss": 0.8061, + "step": 5620 + }, + { + "epoch": 0.13176509197887915, + "grad_norm": 2.964780330657959, + "learning_rate": 0.00019157943495326594, + "loss": 0.2738, + "step": 5621 + }, + { + "epoch": 0.13178853355368414, + "grad_norm": 4.2174811363220215, + "learning_rate": 0.0001915764761059795, + "loss": 0.3842, + "step": 5622 + }, + { + "epoch": 0.13181197512848913, + "grad_norm": 3.9401745796203613, + "learning_rate": 0.00019157351676179655, + "loss": 0.3677, + "step": 5623 + }, + { + "epoch": 0.13183541670329413, + "grad_norm": 3.6994051933288574, + "learning_rate": 0.00019157055692073315, + "loss": 0.3645, + "step": 5624 + }, + { + "epoch": 0.13185885827809912, + "grad_norm": 4.7099456787109375, + "learning_rate": 0.0001915675965828053, + "loss": 0.6583, + "step": 5625 + }, + { + "epoch": 0.13188229985290412, + "grad_norm": 3.806821346282959, + "learning_rate": 0.0001915646357480291, + "loss": 0.591, + "step": 5626 + }, + { + "epoch": 0.1319057414277091, + "grad_norm": 2.180884838104248, + "learning_rate": 0.0001915616744164206, + "loss": 0.2149, + "step": 5627 + }, + { + "epoch": 0.1319291830025141, + "grad_norm": 0.5978502631187439, + "learning_rate": 0.0001915587125879959, + "loss": 0.0568, + "step": 5628 + }, + { + "epoch": 0.1319526245773191, + "grad_norm": 1.0479029417037964, + "learning_rate": 0.00019155575026277104, + "loss": 0.1835, + "step": 5629 + }, + { + "epoch": 0.1319760661521241, + "grad_norm": 3.2811012268066406, + "learning_rate": 0.00019155278744076212, + "loss": 0.2174, + "step": 5630 + }, + { + "epoch": 0.1319995077269291, + "grad_norm": 4.563112735748291, + "learning_rate": 0.00019154982412198518, + "loss": 0.5798, + "step": 5631 + }, + { + "epoch": 0.13202294930173408, + "grad_norm": 5.623298645019531, + "learning_rate": 0.00019154686030645635, + "loss": 0.6685, + "step": 5632 + }, + { + "epoch": 0.13204639087653908, + "grad_norm": 1.2044005393981934, + "learning_rate": 0.00019154389599419167, + "loss": 0.2649, + "step": 5633 + }, + { + "epoch": 0.13206983245134407, + "grad_norm": 3.9350168704986572, + "learning_rate": 0.00019154093118520723, + "loss": 0.6019, + "step": 5634 + }, + { + "epoch": 0.13209327402614907, + "grad_norm": 4.179288864135742, + "learning_rate": 0.0001915379658795191, + "loss": 0.4597, + "step": 5635 + }, + { + "epoch": 0.13211671560095406, + "grad_norm": 17.511531829833984, + "learning_rate": 0.00019153500007714343, + "loss": 0.5212, + "step": 5636 + }, + { + "epoch": 0.13214015717575905, + "grad_norm": 3.001122236251831, + "learning_rate": 0.00019153203377809627, + "loss": 0.1439, + "step": 5637 + }, + { + "epoch": 0.13216359875056408, + "grad_norm": 0.654132068157196, + "learning_rate": 0.00019152906698239371, + "loss": 0.0665, + "step": 5638 + }, + { + "epoch": 0.13218704032536907, + "grad_norm": 3.7404372692108154, + "learning_rate": 0.00019152609969005186, + "loss": 0.6708, + "step": 5639 + }, + { + "epoch": 0.13221048190017406, + "grad_norm": 2.783845901489258, + "learning_rate": 0.00019152313190108682, + "loss": 1.9118, + "step": 5640 + }, + { + "epoch": 0.13223392347497906, + "grad_norm": 5.633434772491455, + "learning_rate": 0.0001915201636155147, + "loss": 0.4431, + "step": 5641 + }, + { + "epoch": 0.13225736504978405, + "grad_norm": 3.8737757205963135, + "learning_rate": 0.0001915171948333516, + "loss": 0.2913, + "step": 5642 + }, + { + "epoch": 0.13228080662458905, + "grad_norm": 3.792231798171997, + "learning_rate": 0.0001915142255546136, + "loss": 0.5251, + "step": 5643 + }, + { + "epoch": 0.13230424819939404, + "grad_norm": 4.963063716888428, + "learning_rate": 0.00019151125577931687, + "loss": 0.7114, + "step": 5644 + }, + { + "epoch": 0.13232768977419904, + "grad_norm": 3.903263568878174, + "learning_rate": 0.00019150828550747748, + "loss": 0.5652, + "step": 5645 + }, + { + "epoch": 0.13235113134900403, + "grad_norm": 3.4254209995269775, + "learning_rate": 0.00019150531473911158, + "loss": 0.6108, + "step": 5646 + }, + { + "epoch": 0.13237457292380903, + "grad_norm": 4.556787490844727, + "learning_rate": 0.00019150234347423522, + "loss": 0.5631, + "step": 5647 + }, + { + "epoch": 0.13239801449861402, + "grad_norm": 3.4745066165924072, + "learning_rate": 0.00019149937171286462, + "loss": 0.3602, + "step": 5648 + }, + { + "epoch": 0.132421456073419, + "grad_norm": 1.2165379524230957, + "learning_rate": 0.00019149639945501584, + "loss": 0.1386, + "step": 5649 + }, + { + "epoch": 0.132444897648224, + "grad_norm": 2.203702688217163, + "learning_rate": 0.000191493426700705, + "loss": 0.19, + "step": 5650 + }, + { + "epoch": 0.132468339223029, + "grad_norm": 4.800155162811279, + "learning_rate": 0.0001914904534499483, + "loss": 0.8179, + "step": 5651 + }, + { + "epoch": 0.132491780797834, + "grad_norm": 1.071025013923645, + "learning_rate": 0.00019148747970276178, + "loss": 0.1485, + "step": 5652 + }, + { + "epoch": 0.132515222372639, + "grad_norm": 1.643738865852356, + "learning_rate": 0.00019148450545916168, + "loss": 0.3472, + "step": 5653 + }, + { + "epoch": 0.13253866394744399, + "grad_norm": 3.9100427627563477, + "learning_rate": 0.00019148153071916403, + "loss": 0.4775, + "step": 5654 + }, + { + "epoch": 0.13256210552224898, + "grad_norm": 4.721468925476074, + "learning_rate": 0.00019147855548278507, + "loss": 0.5681, + "step": 5655 + }, + { + "epoch": 0.13258554709705397, + "grad_norm": 1.7577561140060425, + "learning_rate": 0.00019147557975004084, + "loss": 0.2855, + "step": 5656 + }, + { + "epoch": 0.13260898867185897, + "grad_norm": 3.618921995162964, + "learning_rate": 0.00019147260352094759, + "loss": 0.4596, + "step": 5657 + }, + { + "epoch": 0.13263243024666396, + "grad_norm": 3.2898781299591064, + "learning_rate": 0.00019146962679552138, + "loss": 0.4774, + "step": 5658 + }, + { + "epoch": 0.13265587182146896, + "grad_norm": 3.908334255218506, + "learning_rate": 0.00019146664957377842, + "loss": 0.3254, + "step": 5659 + }, + { + "epoch": 0.13267931339627395, + "grad_norm": 2.040494918823242, + "learning_rate": 0.00019146367185573487, + "loss": 0.2518, + "step": 5660 + }, + { + "epoch": 0.13270275497107895, + "grad_norm": 1.888426423072815, + "learning_rate": 0.00019146069364140683, + "loss": 1.6422, + "step": 5661 + }, + { + "epoch": 0.13272619654588394, + "grad_norm": 1.6433755159378052, + "learning_rate": 0.0001914577149308105, + "loss": 0.2251, + "step": 5662 + }, + { + "epoch": 0.13274963812068893, + "grad_norm": 4.154087543487549, + "learning_rate": 0.00019145473572396206, + "loss": 0.547, + "step": 5663 + }, + { + "epoch": 0.13277307969549396, + "grad_norm": 2.1795642375946045, + "learning_rate": 0.00019145175602087764, + "loss": 0.3419, + "step": 5664 + }, + { + "epoch": 0.13279652127029895, + "grad_norm": 2.492608070373535, + "learning_rate": 0.00019144877582157342, + "loss": 0.4076, + "step": 5665 + }, + { + "epoch": 0.13281996284510394, + "grad_norm": 3.037895441055298, + "learning_rate": 0.00019144579512606557, + "loss": 0.4492, + "step": 5666 + }, + { + "epoch": 0.13284340441990894, + "grad_norm": 2.7070984840393066, + "learning_rate": 0.0001914428139343703, + "loss": 1.3571, + "step": 5667 + }, + { + "epoch": 0.13286684599471393, + "grad_norm": 3.9129130840301514, + "learning_rate": 0.00019143983224650368, + "loss": 1.4847, + "step": 5668 + }, + { + "epoch": 0.13289028756951893, + "grad_norm": 4.580489158630371, + "learning_rate": 0.000191436850062482, + "loss": 0.6526, + "step": 5669 + }, + { + "epoch": 0.13291372914432392, + "grad_norm": 3.6839005947113037, + "learning_rate": 0.0001914338673823214, + "loss": 0.3718, + "step": 5670 + }, + { + "epoch": 0.13293717071912892, + "grad_norm": 1.40336012840271, + "learning_rate": 0.0001914308842060381, + "loss": 0.2601, + "step": 5671 + }, + { + "epoch": 0.1329606122939339, + "grad_norm": 3.6137211322784424, + "learning_rate": 0.00019142790053364818, + "loss": 0.3054, + "step": 5672 + }, + { + "epoch": 0.1329840538687389, + "grad_norm": 4.670192718505859, + "learning_rate": 0.00019142491636516793, + "loss": 0.9238, + "step": 5673 + }, + { + "epoch": 0.1330074954435439, + "grad_norm": 3.657031536102295, + "learning_rate": 0.00019142193170061352, + "loss": 0.3558, + "step": 5674 + }, + { + "epoch": 0.1330309370183489, + "grad_norm": 3.3766591548919678, + "learning_rate": 0.00019141894654000114, + "loss": 0.3959, + "step": 5675 + }, + { + "epoch": 0.1330543785931539, + "grad_norm": 3.4588265419006348, + "learning_rate": 0.00019141596088334697, + "loss": 0.4765, + "step": 5676 + }, + { + "epoch": 0.13307782016795888, + "grad_norm": 3.6882686614990234, + "learning_rate": 0.00019141297473066724, + "loss": 0.4937, + "step": 5677 + }, + { + "epoch": 0.13310126174276388, + "grad_norm": 0.7585864067077637, + "learning_rate": 0.00019140998808197813, + "loss": 0.09, + "step": 5678 + }, + { + "epoch": 0.13312470331756887, + "grad_norm": 3.4800069332122803, + "learning_rate": 0.00019140700093729586, + "loss": 0.3854, + "step": 5679 + }, + { + "epoch": 0.13314814489237387, + "grad_norm": 3.402125597000122, + "learning_rate": 0.00019140401329663663, + "loss": 0.4771, + "step": 5680 + }, + { + "epoch": 0.13317158646717886, + "grad_norm": 3.099869966506958, + "learning_rate": 0.00019140102516001668, + "loss": 0.4154, + "step": 5681 + }, + { + "epoch": 0.13319502804198385, + "grad_norm": 4.389246940612793, + "learning_rate": 0.00019139803652745216, + "loss": 0.8162, + "step": 5682 + }, + { + "epoch": 0.13321846961678885, + "grad_norm": 6.0280303955078125, + "learning_rate": 0.00019139504739895932, + "loss": 0.9016, + "step": 5683 + }, + { + "epoch": 0.13324191119159384, + "grad_norm": 4.688984394073486, + "learning_rate": 0.0001913920577745544, + "loss": 0.5322, + "step": 5684 + }, + { + "epoch": 0.13326535276639884, + "grad_norm": 4.524352073669434, + "learning_rate": 0.00019138906765425363, + "loss": 0.5069, + "step": 5685 + }, + { + "epoch": 0.13328879434120383, + "grad_norm": 2.649183750152588, + "learning_rate": 0.00019138607703807317, + "loss": 0.2683, + "step": 5686 + }, + { + "epoch": 0.13331223591600883, + "grad_norm": 3.3728044033050537, + "learning_rate": 0.0001913830859260293, + "loss": 0.4694, + "step": 5687 + }, + { + "epoch": 0.13333567749081382, + "grad_norm": 5.419027805328369, + "learning_rate": 0.00019138009431813826, + "loss": 0.3157, + "step": 5688 + }, + { + "epoch": 0.13335911906561884, + "grad_norm": 1.6700735092163086, + "learning_rate": 0.00019137710221441622, + "loss": 0.2056, + "step": 5689 + }, + { + "epoch": 0.13338256064042384, + "grad_norm": 3.415947675704956, + "learning_rate": 0.00019137410961487947, + "loss": 0.3693, + "step": 5690 + }, + { + "epoch": 0.13340600221522883, + "grad_norm": 4.138092041015625, + "learning_rate": 0.00019137111651954425, + "loss": 0.6885, + "step": 5691 + }, + { + "epoch": 0.13342944379003382, + "grad_norm": 4.806591033935547, + "learning_rate": 0.00019136812292842677, + "loss": 0.5055, + "step": 5692 + }, + { + "epoch": 0.13345288536483882, + "grad_norm": 1.1628056764602661, + "learning_rate": 0.00019136512884154326, + "loss": 0.1735, + "step": 5693 + }, + { + "epoch": 0.1334763269396438, + "grad_norm": 3.4334335327148438, + "learning_rate": 0.00019136213425891002, + "loss": 0.6782, + "step": 5694 + }, + { + "epoch": 0.1334997685144488, + "grad_norm": 3.3387866020202637, + "learning_rate": 0.00019135913918054325, + "loss": 0.2413, + "step": 5695 + }, + { + "epoch": 0.1335232100892538, + "grad_norm": 4.0287556648254395, + "learning_rate": 0.00019135614360645925, + "loss": 0.4852, + "step": 5696 + }, + { + "epoch": 0.1335466516640588, + "grad_norm": 4.1928606033325195, + "learning_rate": 0.00019135314753667424, + "loss": 0.6336, + "step": 5697 + }, + { + "epoch": 0.1335700932388638, + "grad_norm": 5.0874528884887695, + "learning_rate": 0.00019135015097120447, + "loss": 0.7082, + "step": 5698 + }, + { + "epoch": 0.13359353481366879, + "grad_norm": 1.6935151815414429, + "learning_rate": 0.0001913471539100662, + "loss": 0.1775, + "step": 5699 + }, + { + "epoch": 0.13361697638847378, + "grad_norm": 4.892632484436035, + "learning_rate": 0.0001913441563532757, + "loss": 0.5681, + "step": 5700 + }, + { + "epoch": 0.13364041796327877, + "grad_norm": 1.205151081085205, + "learning_rate": 0.0001913411583008493, + "loss": 0.132, + "step": 5701 + }, + { + "epoch": 0.13366385953808377, + "grad_norm": 5.923111915588379, + "learning_rate": 0.00019133815975280313, + "loss": 0.919, + "step": 5702 + }, + { + "epoch": 0.13368730111288876, + "grad_norm": 2.9997365474700928, + "learning_rate": 0.00019133516070915358, + "loss": 0.2707, + "step": 5703 + }, + { + "epoch": 0.13371074268769376, + "grad_norm": 4.083303451538086, + "learning_rate": 0.0001913321611699169, + "loss": 0.5389, + "step": 5704 + }, + { + "epoch": 0.13373418426249875, + "grad_norm": 4.372618675231934, + "learning_rate": 0.0001913291611351093, + "loss": 0.6478, + "step": 5705 + }, + { + "epoch": 0.13375762583730375, + "grad_norm": 3.271857738494873, + "learning_rate": 0.00019132616060474713, + "loss": 0.4527, + "step": 5706 + }, + { + "epoch": 0.13378106741210874, + "grad_norm": 3.6351115703582764, + "learning_rate": 0.00019132315957884664, + "loss": 1.458, + "step": 5707 + }, + { + "epoch": 0.13380450898691373, + "grad_norm": 3.2615182399749756, + "learning_rate": 0.00019132015805742411, + "loss": 1.9263, + "step": 5708 + }, + { + "epoch": 0.13382795056171873, + "grad_norm": 3.7227413654327393, + "learning_rate": 0.00019131715604049586, + "loss": 0.5544, + "step": 5709 + }, + { + "epoch": 0.13385139213652372, + "grad_norm": 4.703038692474365, + "learning_rate": 0.0001913141535280781, + "loss": 0.535, + "step": 5710 + }, + { + "epoch": 0.13387483371132872, + "grad_norm": 4.184515476226807, + "learning_rate": 0.00019131115052018721, + "loss": 0.4099, + "step": 5711 + }, + { + "epoch": 0.1338982752861337, + "grad_norm": 5.569277763366699, + "learning_rate": 0.00019130814701683946, + "loss": 0.8405, + "step": 5712 + }, + { + "epoch": 0.1339217168609387, + "grad_norm": 3.5067615509033203, + "learning_rate": 0.00019130514301805114, + "loss": 0.3507, + "step": 5713 + }, + { + "epoch": 0.1339451584357437, + "grad_norm": 0.9241841435432434, + "learning_rate": 0.00019130213852383852, + "loss": 0.1279, + "step": 5714 + }, + { + "epoch": 0.13396860001054872, + "grad_norm": 2.7690961360931396, + "learning_rate": 0.00019129913353421794, + "loss": 1.7359, + "step": 5715 + }, + { + "epoch": 0.13399204158535372, + "grad_norm": 1.038130283355713, + "learning_rate": 0.0001912961280492057, + "loss": 0.1692, + "step": 5716 + }, + { + "epoch": 0.1340154831601587, + "grad_norm": 2.8661162853240967, + "learning_rate": 0.00019129312206881812, + "loss": 0.3416, + "step": 5717 + }, + { + "epoch": 0.1340389247349637, + "grad_norm": 1.6386866569519043, + "learning_rate": 0.00019129011559307146, + "loss": 0.2234, + "step": 5718 + }, + { + "epoch": 0.1340623663097687, + "grad_norm": 5.0237226486206055, + "learning_rate": 0.00019128710862198207, + "loss": 0.778, + "step": 5719 + }, + { + "epoch": 0.1340858078845737, + "grad_norm": 3.898543119430542, + "learning_rate": 0.00019128410115556627, + "loss": 0.4554, + "step": 5720 + }, + { + "epoch": 0.1341092494593787, + "grad_norm": 4.468048572540283, + "learning_rate": 0.00019128109319384038, + "loss": 0.5286, + "step": 5721 + }, + { + "epoch": 0.13413269103418368, + "grad_norm": 4.9489827156066895, + "learning_rate": 0.0001912780847368207, + "loss": 0.7113, + "step": 5722 + }, + { + "epoch": 0.13415613260898868, + "grad_norm": 2.438969135284424, + "learning_rate": 0.00019127507578452357, + "loss": 0.2704, + "step": 5723 + }, + { + "epoch": 0.13417957418379367, + "grad_norm": 1.6556596755981445, + "learning_rate": 0.0001912720663369653, + "loss": 0.1461, + "step": 5724 + }, + { + "epoch": 0.13420301575859866, + "grad_norm": 3.739076614379883, + "learning_rate": 0.00019126905639416225, + "loss": 0.6495, + "step": 5725 + }, + { + "epoch": 0.13422645733340366, + "grad_norm": 4.4100494384765625, + "learning_rate": 0.00019126604595613071, + "loss": 0.7764, + "step": 5726 + }, + { + "epoch": 0.13424989890820865, + "grad_norm": 0.9139430522918701, + "learning_rate": 0.00019126303502288707, + "loss": 0.155, + "step": 5727 + }, + { + "epoch": 0.13427334048301365, + "grad_norm": 1.1212704181671143, + "learning_rate": 0.00019126002359444764, + "loss": 0.2373, + "step": 5728 + }, + { + "epoch": 0.13429678205781864, + "grad_norm": 3.7563161849975586, + "learning_rate": 0.00019125701167082873, + "loss": 0.6142, + "step": 5729 + }, + { + "epoch": 0.13432022363262364, + "grad_norm": 3.7600109577178955, + "learning_rate": 0.0001912539992520467, + "loss": 0.4689, + "step": 5730 + }, + { + "epoch": 0.13434366520742863, + "grad_norm": 4.4140753746032715, + "learning_rate": 0.0001912509863381179, + "loss": 0.4393, + "step": 5731 + }, + { + "epoch": 0.13436710678223363, + "grad_norm": 5.131381511688232, + "learning_rate": 0.0001912479729290587, + "loss": 0.923, + "step": 5732 + }, + { + "epoch": 0.13439054835703862, + "grad_norm": 2.178806781768799, + "learning_rate": 0.00019124495902488547, + "loss": 1.3974, + "step": 5733 + }, + { + "epoch": 0.1344139899318436, + "grad_norm": 3.882258892059326, + "learning_rate": 0.0001912419446256145, + "loss": 0.2933, + "step": 5734 + }, + { + "epoch": 0.1344374315066486, + "grad_norm": 4.692087173461914, + "learning_rate": 0.00019123892973126215, + "loss": 0.6008, + "step": 5735 + }, + { + "epoch": 0.1344608730814536, + "grad_norm": 4.944162368774414, + "learning_rate": 0.0001912359143418448, + "loss": 0.474, + "step": 5736 + }, + { + "epoch": 0.1344843146562586, + "grad_norm": 0.5140254497528076, + "learning_rate": 0.00019123289845737884, + "loss": 0.0711, + "step": 5737 + }, + { + "epoch": 0.1345077562310636, + "grad_norm": 2.6386353969573975, + "learning_rate": 0.0001912298820778806, + "loss": 1.53, + "step": 5738 + }, + { + "epoch": 0.13453119780586859, + "grad_norm": 2.1081511974334717, + "learning_rate": 0.00019122686520336646, + "loss": 0.4559, + "step": 5739 + }, + { + "epoch": 0.1345546393806736, + "grad_norm": 2.8989102840423584, + "learning_rate": 0.00019122384783385276, + "loss": 0.3637, + "step": 5740 + }, + { + "epoch": 0.1345780809554786, + "grad_norm": 4.407965660095215, + "learning_rate": 0.0001912208299693559, + "loss": 0.7795, + "step": 5741 + }, + { + "epoch": 0.1346015225302836, + "grad_norm": 2.017198085784912, + "learning_rate": 0.00019121781160989228, + "loss": 0.2758, + "step": 5742 + }, + { + "epoch": 0.1346249641050886, + "grad_norm": 3.0667335987091064, + "learning_rate": 0.00019121479275547825, + "loss": 1.9554, + "step": 5743 + }, + { + "epoch": 0.13464840567989358, + "grad_norm": 3.8699352741241455, + "learning_rate": 0.00019121177340613015, + "loss": 0.4463, + "step": 5744 + }, + { + "epoch": 0.13467184725469858, + "grad_norm": 3.5196776390075684, + "learning_rate": 0.00019120875356186444, + "loss": 0.2735, + "step": 5745 + }, + { + "epoch": 0.13469528882950357, + "grad_norm": 3.0254762172698975, + "learning_rate": 0.00019120573322269744, + "loss": 0.391, + "step": 5746 + }, + { + "epoch": 0.13471873040430857, + "grad_norm": 0.4305045008659363, + "learning_rate": 0.0001912027123886456, + "loss": 0.0387, + "step": 5747 + }, + { + "epoch": 0.13474217197911356, + "grad_norm": 1.1564196348190308, + "learning_rate": 0.00019119969105972526, + "loss": 0.1059, + "step": 5748 + }, + { + "epoch": 0.13476561355391856, + "grad_norm": 4.643134117126465, + "learning_rate": 0.00019119666923595282, + "loss": 0.57, + "step": 5749 + }, + { + "epoch": 0.13478905512872355, + "grad_norm": 2.805372714996338, + "learning_rate": 0.0001911936469173447, + "loss": 0.4489, + "step": 5750 + }, + { + "epoch": 0.13481249670352854, + "grad_norm": 2.4048879146575928, + "learning_rate": 0.00019119062410391727, + "loss": 0.3531, + "step": 5751 + }, + { + "epoch": 0.13483593827833354, + "grad_norm": 2.751483678817749, + "learning_rate": 0.000191187600795687, + "loss": 0.4155, + "step": 5752 + }, + { + "epoch": 0.13485937985313853, + "grad_norm": 5.093166828155518, + "learning_rate": 0.0001911845769926702, + "loss": 0.561, + "step": 5753 + }, + { + "epoch": 0.13488282142794353, + "grad_norm": 6.397964000701904, + "learning_rate": 0.00019118155269488333, + "loss": 0.7517, + "step": 5754 + }, + { + "epoch": 0.13490626300274852, + "grad_norm": 2.9176909923553467, + "learning_rate": 0.00019117852790234282, + "loss": 0.4667, + "step": 5755 + }, + { + "epoch": 0.13492970457755352, + "grad_norm": 2.4192049503326416, + "learning_rate": 0.00019117550261506502, + "loss": 0.2573, + "step": 5756 + }, + { + "epoch": 0.1349531461523585, + "grad_norm": 4.065261363983154, + "learning_rate": 0.0001911724768330664, + "loss": 0.524, + "step": 5757 + }, + { + "epoch": 0.1349765877271635, + "grad_norm": 4.2815351486206055, + "learning_rate": 0.00019116945055636334, + "loss": 0.4783, + "step": 5758 + }, + { + "epoch": 0.1350000293019685, + "grad_norm": 4.8050127029418945, + "learning_rate": 0.00019116642378497228, + "loss": 0.7183, + "step": 5759 + }, + { + "epoch": 0.1350234708767735, + "grad_norm": 3.8151488304138184, + "learning_rate": 0.00019116339651890964, + "loss": 0.5256, + "step": 5760 + }, + { + "epoch": 0.1350469124515785, + "grad_norm": 3.0808303356170654, + "learning_rate": 0.00019116036875819185, + "loss": 0.4587, + "step": 5761 + }, + { + "epoch": 0.13507035402638348, + "grad_norm": 0.9249255061149597, + "learning_rate": 0.00019115734050283533, + "loss": 0.0687, + "step": 5762 + }, + { + "epoch": 0.13509379560118848, + "grad_norm": 2.612412214279175, + "learning_rate": 0.00019115431175285651, + "loss": 0.2971, + "step": 5763 + }, + { + "epoch": 0.13511723717599347, + "grad_norm": 3.98099946975708, + "learning_rate": 0.00019115128250827185, + "loss": 0.7126, + "step": 5764 + }, + { + "epoch": 0.13514067875079847, + "grad_norm": 5.558365821838379, + "learning_rate": 0.00019114825276909776, + "loss": 0.8299, + "step": 5765 + }, + { + "epoch": 0.1351641203256035, + "grad_norm": 4.207444667816162, + "learning_rate": 0.0001911452225353507, + "loss": 0.673, + "step": 5766 + }, + { + "epoch": 0.13518756190040848, + "grad_norm": 3.0284183025360107, + "learning_rate": 0.0001911421918070471, + "loss": 1.3032, + "step": 5767 + }, + { + "epoch": 0.13521100347521348, + "grad_norm": 1.3893787860870361, + "learning_rate": 0.0001911391605842034, + "loss": 0.153, + "step": 5768 + }, + { + "epoch": 0.13523444505001847, + "grad_norm": 0.6592244505882263, + "learning_rate": 0.00019113612886683602, + "loss": 0.0715, + "step": 5769 + }, + { + "epoch": 0.13525788662482346, + "grad_norm": 6.493566513061523, + "learning_rate": 0.00019113309665496147, + "loss": 0.8699, + "step": 5770 + }, + { + "epoch": 0.13528132819962846, + "grad_norm": 1.0921682119369507, + "learning_rate": 0.0001911300639485962, + "loss": 0.1458, + "step": 5771 + }, + { + "epoch": 0.13530476977443345, + "grad_norm": 2.7688357830047607, + "learning_rate": 0.0001911270307477566, + "loss": 0.3126, + "step": 5772 + }, + { + "epoch": 0.13532821134923845, + "grad_norm": 4.165993690490723, + "learning_rate": 0.0001911239970524592, + "loss": 0.5386, + "step": 5773 + }, + { + "epoch": 0.13535165292404344, + "grad_norm": 1.0577524900436401, + "learning_rate": 0.0001911209628627204, + "loss": 0.1247, + "step": 5774 + }, + { + "epoch": 0.13537509449884844, + "grad_norm": 4.4297943115234375, + "learning_rate": 0.00019111792817855673, + "loss": 0.6153, + "step": 5775 + }, + { + "epoch": 0.13539853607365343, + "grad_norm": 3.394402027130127, + "learning_rate": 0.0001911148929999846, + "loss": 0.3875, + "step": 5776 + }, + { + "epoch": 0.13542197764845842, + "grad_norm": 3.653637170791626, + "learning_rate": 0.00019111185732702049, + "loss": 0.6738, + "step": 5777 + }, + { + "epoch": 0.13544541922326342, + "grad_norm": 4.353789806365967, + "learning_rate": 0.00019110882115968086, + "loss": 0.7522, + "step": 5778 + }, + { + "epoch": 0.1354688607980684, + "grad_norm": 1.442158818244934, + "learning_rate": 0.00019110578449798226, + "loss": 0.2059, + "step": 5779 + }, + { + "epoch": 0.1354923023728734, + "grad_norm": 3.1653711795806885, + "learning_rate": 0.00019110274734194106, + "loss": 0.4322, + "step": 5780 + }, + { + "epoch": 0.1355157439476784, + "grad_norm": 0.6812604665756226, + "learning_rate": 0.0001910997096915738, + "loss": 0.104, + "step": 5781 + }, + { + "epoch": 0.1355391855224834, + "grad_norm": 2.31019926071167, + "learning_rate": 0.000191096671546897, + "loss": 0.4812, + "step": 5782 + }, + { + "epoch": 0.1355626270972884, + "grad_norm": 2.8659849166870117, + "learning_rate": 0.00019109363290792706, + "loss": 0.2072, + "step": 5783 + }, + { + "epoch": 0.13558606867209338, + "grad_norm": 2.887300729751587, + "learning_rate": 0.00019109059377468054, + "loss": 0.3381, + "step": 5784 + }, + { + "epoch": 0.13560951024689838, + "grad_norm": 1.9069099426269531, + "learning_rate": 0.00019108755414717386, + "loss": 1.4705, + "step": 5785 + }, + { + "epoch": 0.13563295182170337, + "grad_norm": 4.22961950302124, + "learning_rate": 0.00019108451402542355, + "loss": 0.8157, + "step": 5786 + }, + { + "epoch": 0.13565639339650837, + "grad_norm": 2.9792706966400146, + "learning_rate": 0.00019108147340944612, + "loss": 0.4325, + "step": 5787 + }, + { + "epoch": 0.13567983497131336, + "grad_norm": 3.254302740097046, + "learning_rate": 0.00019107843229925804, + "loss": 0.5179, + "step": 5788 + }, + { + "epoch": 0.13570327654611836, + "grad_norm": 3.7032668590545654, + "learning_rate": 0.00019107539069487586, + "loss": 0.5569, + "step": 5789 + }, + { + "epoch": 0.13572671812092335, + "grad_norm": 3.066588878631592, + "learning_rate": 0.000191072348596316, + "loss": 0.4036, + "step": 5790 + }, + { + "epoch": 0.13575015969572835, + "grad_norm": 3.1417696475982666, + "learning_rate": 0.00019106930600359506, + "loss": 0.4311, + "step": 5791 + }, + { + "epoch": 0.13577360127053337, + "grad_norm": 0.5595006942749023, + "learning_rate": 0.00019106626291672947, + "loss": 0.0741, + "step": 5792 + }, + { + "epoch": 0.13579704284533836, + "grad_norm": 4.306867599487305, + "learning_rate": 0.0001910632193357358, + "loss": 0.4975, + "step": 5793 + }, + { + "epoch": 0.13582048442014336, + "grad_norm": 1.4385496377944946, + "learning_rate": 0.0001910601752606305, + "loss": 0.1758, + "step": 5794 + }, + { + "epoch": 0.13584392599494835, + "grad_norm": 3.620126485824585, + "learning_rate": 0.0001910571306914302, + "loss": 0.3514, + "step": 5795 + }, + { + "epoch": 0.13586736756975334, + "grad_norm": 0.8350220322608948, + "learning_rate": 0.00019105408562815132, + "loss": 0.0868, + "step": 5796 + }, + { + "epoch": 0.13589080914455834, + "grad_norm": 3.9375667572021484, + "learning_rate": 0.00019105104007081038, + "loss": 0.4994, + "step": 5797 + }, + { + "epoch": 0.13591425071936333, + "grad_norm": 1.8405214548110962, + "learning_rate": 0.00019104799401942397, + "loss": 1.7368, + "step": 5798 + }, + { + "epoch": 0.13593769229416833, + "grad_norm": 6.725493431091309, + "learning_rate": 0.0001910449474740086, + "loss": 0.401, + "step": 5799 + }, + { + "epoch": 0.13596113386897332, + "grad_norm": 4.034055709838867, + "learning_rate": 0.00019104190043458074, + "loss": 0.3782, + "step": 5800 + }, + { + "epoch": 0.13598457544377832, + "grad_norm": 6.974907398223877, + "learning_rate": 0.00019103885290115696, + "loss": 0.2416, + "step": 5801 + }, + { + "epoch": 0.1360080170185833, + "grad_norm": 3.3598265647888184, + "learning_rate": 0.00019103580487375384, + "loss": 0.3022, + "step": 5802 + }, + { + "epoch": 0.1360314585933883, + "grad_norm": 4.575523853302002, + "learning_rate": 0.0001910327563523879, + "loss": 0.6807, + "step": 5803 + }, + { + "epoch": 0.1360549001681933, + "grad_norm": 5.021275043487549, + "learning_rate": 0.0001910297073370756, + "loss": 0.4585, + "step": 5804 + }, + { + "epoch": 0.1360783417429983, + "grad_norm": 1.9688955545425415, + "learning_rate": 0.0001910266578278336, + "loss": 0.3311, + "step": 5805 + }, + { + "epoch": 0.1361017833178033, + "grad_norm": 4.1261138916015625, + "learning_rate": 0.00019102360782467838, + "loss": 0.7129, + "step": 5806 + }, + { + "epoch": 0.13612522489260828, + "grad_norm": 5.280327320098877, + "learning_rate": 0.0001910205573276265, + "loss": 0.5882, + "step": 5807 + }, + { + "epoch": 0.13614866646741328, + "grad_norm": 5.467422962188721, + "learning_rate": 0.00019101750633669453, + "loss": 0.6286, + "step": 5808 + }, + { + "epoch": 0.13617210804221827, + "grad_norm": 3.113391399383545, + "learning_rate": 0.00019101445485189902, + "loss": 0.3674, + "step": 5809 + }, + { + "epoch": 0.13619554961702326, + "grad_norm": 4.084746837615967, + "learning_rate": 0.0001910114028732565, + "loss": 0.5574, + "step": 5810 + }, + { + "epoch": 0.13621899119182826, + "grad_norm": 1.2738901376724243, + "learning_rate": 0.00019100835040078355, + "loss": 0.1566, + "step": 5811 + }, + { + "epoch": 0.13624243276663325, + "grad_norm": 3.0938382148742676, + "learning_rate": 0.00019100529743449674, + "loss": 1.8751, + "step": 5812 + }, + { + "epoch": 0.13626587434143825, + "grad_norm": 5.861605167388916, + "learning_rate": 0.00019100224397441262, + "loss": 1.075, + "step": 5813 + }, + { + "epoch": 0.13628931591624324, + "grad_norm": 3.8388540744781494, + "learning_rate": 0.00019099919002054776, + "loss": 0.5786, + "step": 5814 + }, + { + "epoch": 0.13631275749104824, + "grad_norm": 3.5372235774993896, + "learning_rate": 0.00019099613557291875, + "loss": 0.3285, + "step": 5815 + }, + { + "epoch": 0.13633619906585323, + "grad_norm": 3.5815210342407227, + "learning_rate": 0.00019099308063154214, + "loss": 0.7127, + "step": 5816 + }, + { + "epoch": 0.13635964064065825, + "grad_norm": 2.224491596221924, + "learning_rate": 0.00019099002519643452, + "loss": 1.776, + "step": 5817 + }, + { + "epoch": 0.13638308221546325, + "grad_norm": 2.1878061294555664, + "learning_rate": 0.00019098696926761246, + "loss": 0.2453, + "step": 5818 + }, + { + "epoch": 0.13640652379026824, + "grad_norm": 2.037760019302368, + "learning_rate": 0.00019098391284509255, + "loss": 0.2546, + "step": 5819 + }, + { + "epoch": 0.13642996536507324, + "grad_norm": 2.5148720741271973, + "learning_rate": 0.0001909808559288914, + "loss": 0.3137, + "step": 5820 + }, + { + "epoch": 0.13645340693987823, + "grad_norm": 5.30387544631958, + "learning_rate": 0.00019097779851902553, + "loss": 0.3126, + "step": 5821 + }, + { + "epoch": 0.13647684851468322, + "grad_norm": 2.048903703689575, + "learning_rate": 0.00019097474061551158, + "loss": 0.2585, + "step": 5822 + }, + { + "epoch": 0.13650029008948822, + "grad_norm": 4.164310455322266, + "learning_rate": 0.00019097168221836612, + "loss": 0.5359, + "step": 5823 + }, + { + "epoch": 0.1365237316642932, + "grad_norm": 0.9102200269699097, + "learning_rate": 0.00019096862332760575, + "loss": 0.1126, + "step": 5824 + }, + { + "epoch": 0.1365471732390982, + "grad_norm": 3.25555682182312, + "learning_rate": 0.0001909655639432471, + "loss": 0.5488, + "step": 5825 + }, + { + "epoch": 0.1365706148139032, + "grad_norm": 2.02301287651062, + "learning_rate": 0.0001909625040653067, + "loss": 2.011, + "step": 5826 + }, + { + "epoch": 0.1365940563887082, + "grad_norm": 1.9350557327270508, + "learning_rate": 0.00019095944369380122, + "loss": 0.2126, + "step": 5827 + }, + { + "epoch": 0.1366174979635132, + "grad_norm": 2.7084782123565674, + "learning_rate": 0.00019095638282874724, + "loss": 0.4191, + "step": 5828 + }, + { + "epoch": 0.13664093953831818, + "grad_norm": 3.2610208988189697, + "learning_rate": 0.00019095332147016136, + "loss": 0.3664, + "step": 5829 + }, + { + "epoch": 0.13666438111312318, + "grad_norm": 4.119907855987549, + "learning_rate": 0.00019095025961806024, + "loss": 0.478, + "step": 5830 + }, + { + "epoch": 0.13668782268792817, + "grad_norm": 3.87508225440979, + "learning_rate": 0.0001909471972724604, + "loss": 0.498, + "step": 5831 + }, + { + "epoch": 0.13671126426273317, + "grad_norm": 0.6018820405006409, + "learning_rate": 0.00019094413443337854, + "loss": 0.0491, + "step": 5832 + }, + { + "epoch": 0.13673470583753816, + "grad_norm": 0.6964131593704224, + "learning_rate": 0.00019094107110083125, + "loss": 0.0689, + "step": 5833 + }, + { + "epoch": 0.13675814741234316, + "grad_norm": 5.0468268394470215, + "learning_rate": 0.00019093800727483515, + "loss": 0.633, + "step": 5834 + }, + { + "epoch": 0.13678158898714815, + "grad_norm": 0.6307666301727295, + "learning_rate": 0.00019093494295540687, + "loss": 0.0704, + "step": 5835 + }, + { + "epoch": 0.13680503056195314, + "grad_norm": 3.28391695022583, + "learning_rate": 0.000190931878142563, + "loss": 0.5441, + "step": 5836 + }, + { + "epoch": 0.13682847213675814, + "grad_norm": 3.165555238723755, + "learning_rate": 0.00019092881283632024, + "loss": 2.0183, + "step": 5837 + }, + { + "epoch": 0.13685191371156313, + "grad_norm": 5.792149066925049, + "learning_rate": 0.00019092574703669516, + "loss": 1.0656, + "step": 5838 + }, + { + "epoch": 0.13687535528636813, + "grad_norm": 3.2096216678619385, + "learning_rate": 0.00019092268074370443, + "loss": 1.9128, + "step": 5839 + }, + { + "epoch": 0.13689879686117312, + "grad_norm": 3.5925140380859375, + "learning_rate": 0.0001909196139573647, + "loss": 0.3691, + "step": 5840 + }, + { + "epoch": 0.13692223843597812, + "grad_norm": 1.2278680801391602, + "learning_rate": 0.00019091654667769255, + "loss": 0.106, + "step": 5841 + }, + { + "epoch": 0.1369456800107831, + "grad_norm": 3.5572667121887207, + "learning_rate": 0.00019091347890470468, + "loss": 0.7763, + "step": 5842 + }, + { + "epoch": 0.13696912158558813, + "grad_norm": 4.959041595458984, + "learning_rate": 0.00019091041063841772, + "loss": 0.4544, + "step": 5843 + }, + { + "epoch": 0.13699256316039313, + "grad_norm": 3.5820982456207275, + "learning_rate": 0.00019090734187884832, + "loss": 0.4437, + "step": 5844 + }, + { + "epoch": 0.13701600473519812, + "grad_norm": 2.172464370727539, + "learning_rate": 0.00019090427262601311, + "loss": 0.2751, + "step": 5845 + }, + { + "epoch": 0.13703944631000312, + "grad_norm": 1.685261845588684, + "learning_rate": 0.00019090120287992875, + "loss": 0.2239, + "step": 5846 + }, + { + "epoch": 0.1370628878848081, + "grad_norm": 0.6897400617599487, + "learning_rate": 0.00019089813264061193, + "loss": 0.0724, + "step": 5847 + }, + { + "epoch": 0.1370863294596131, + "grad_norm": 3.003521203994751, + "learning_rate": 0.00019089506190807928, + "loss": 0.3529, + "step": 5848 + }, + { + "epoch": 0.1371097710344181, + "grad_norm": 4.602240085601807, + "learning_rate": 0.0001908919906823475, + "loss": 0.7583, + "step": 5849 + }, + { + "epoch": 0.1371332126092231, + "grad_norm": 3.724825382232666, + "learning_rate": 0.00019088891896343316, + "loss": 0.4938, + "step": 5850 + }, + { + "epoch": 0.1371566541840281, + "grad_norm": 4.299793243408203, + "learning_rate": 0.00019088584675135305, + "loss": 0.6893, + "step": 5851 + }, + { + "epoch": 0.13718009575883308, + "grad_norm": 4.681862831115723, + "learning_rate": 0.00019088277404612374, + "loss": 0.2324, + "step": 5852 + }, + { + "epoch": 0.13720353733363808, + "grad_norm": 3.43617844581604, + "learning_rate": 0.00019087970084776196, + "loss": 1.4061, + "step": 5853 + }, + { + "epoch": 0.13722697890844307, + "grad_norm": 4.916350364685059, + "learning_rate": 0.00019087662715628438, + "loss": 0.2445, + "step": 5854 + }, + { + "epoch": 0.13725042048324806, + "grad_norm": 2.9579148292541504, + "learning_rate": 0.00019087355297170764, + "loss": 0.3311, + "step": 5855 + }, + { + "epoch": 0.13727386205805306, + "grad_norm": 5.610457897186279, + "learning_rate": 0.00019087047829404846, + "loss": 0.6571, + "step": 5856 + }, + { + "epoch": 0.13729730363285805, + "grad_norm": 1.3474565744400024, + "learning_rate": 0.0001908674031233235, + "loss": 0.2464, + "step": 5857 + }, + { + "epoch": 0.13732074520766305, + "grad_norm": 3.2118561267852783, + "learning_rate": 0.00019086432745954945, + "loss": 0.5122, + "step": 5858 + }, + { + "epoch": 0.13734418678246804, + "grad_norm": 3.14266300201416, + "learning_rate": 0.000190861251302743, + "loss": 0.3498, + "step": 5859 + }, + { + "epoch": 0.13736762835727304, + "grad_norm": 2.292092800140381, + "learning_rate": 0.0001908581746529209, + "loss": 1.8935, + "step": 5860 + }, + { + "epoch": 0.13739106993207803, + "grad_norm": 0.4954896569252014, + "learning_rate": 0.00019085509751009974, + "loss": 0.1005, + "step": 5861 + }, + { + "epoch": 0.13741451150688302, + "grad_norm": 4.221159934997559, + "learning_rate": 0.00019085201987429628, + "loss": 0.7525, + "step": 5862 + }, + { + "epoch": 0.13743795308168802, + "grad_norm": 0.7621211409568787, + "learning_rate": 0.0001908489417455272, + "loss": 0.0993, + "step": 5863 + }, + { + "epoch": 0.137461394656493, + "grad_norm": 3.6484153270721436, + "learning_rate": 0.00019084586312380922, + "loss": 0.6276, + "step": 5864 + }, + { + "epoch": 0.137484836231298, + "grad_norm": 3.1458022594451904, + "learning_rate": 0.000190842784009159, + "loss": 0.5521, + "step": 5865 + }, + { + "epoch": 0.137508277806103, + "grad_norm": 0.8419052958488464, + "learning_rate": 0.0001908397044015933, + "loss": 0.1584, + "step": 5866 + }, + { + "epoch": 0.137531719380908, + "grad_norm": 3.865562915802002, + "learning_rate": 0.00019083662430112883, + "loss": 0.5398, + "step": 5867 + }, + { + "epoch": 0.13755516095571302, + "grad_norm": 3.452037811279297, + "learning_rate": 0.00019083354370778225, + "loss": 0.4286, + "step": 5868 + }, + { + "epoch": 0.137578602530518, + "grad_norm": 2.2317304611206055, + "learning_rate": 0.00019083046262157033, + "loss": 1.4746, + "step": 5869 + }, + { + "epoch": 0.137602044105323, + "grad_norm": 4.012791633605957, + "learning_rate": 0.00019082738104250977, + "loss": 0.6306, + "step": 5870 + }, + { + "epoch": 0.137625485680128, + "grad_norm": 3.360262393951416, + "learning_rate": 0.00019082429897061725, + "loss": 0.446, + "step": 5871 + }, + { + "epoch": 0.137648927254933, + "grad_norm": 2.840240955352783, + "learning_rate": 0.00019082121640590958, + "loss": 0.2842, + "step": 5872 + }, + { + "epoch": 0.137672368829738, + "grad_norm": 3.1466643810272217, + "learning_rate": 0.0001908181333484034, + "loss": 0.2465, + "step": 5873 + }, + { + "epoch": 0.13769581040454298, + "grad_norm": 4.968297481536865, + "learning_rate": 0.0001908150497981155, + "loss": 0.6122, + "step": 5874 + }, + { + "epoch": 0.13771925197934798, + "grad_norm": 0.7300992608070374, + "learning_rate": 0.00019081196575506258, + "loss": 0.0651, + "step": 5875 + }, + { + "epoch": 0.13774269355415297, + "grad_norm": 3.865704298019409, + "learning_rate": 0.00019080888121926137, + "loss": 0.4825, + "step": 5876 + }, + { + "epoch": 0.13776613512895797, + "grad_norm": 1.9414236545562744, + "learning_rate": 0.00019080579619072864, + "loss": 0.3013, + "step": 5877 + }, + { + "epoch": 0.13778957670376296, + "grad_norm": 4.4607086181640625, + "learning_rate": 0.00019080271066948106, + "loss": 0.5431, + "step": 5878 + }, + { + "epoch": 0.13781301827856796, + "grad_norm": 4.480865001678467, + "learning_rate": 0.00019079962465553545, + "loss": 0.6938, + "step": 5879 + }, + { + "epoch": 0.13783645985337295, + "grad_norm": 1.7428252696990967, + "learning_rate": 0.0001907965381489085, + "loss": 1.1385, + "step": 5880 + }, + { + "epoch": 0.13785990142817794, + "grad_norm": 1.8594192266464233, + "learning_rate": 0.00019079345114961701, + "loss": 0.285, + "step": 5881 + }, + { + "epoch": 0.13788334300298294, + "grad_norm": 1.1797133684158325, + "learning_rate": 0.0001907903636576777, + "loss": 0.1263, + "step": 5882 + }, + { + "epoch": 0.13790678457778793, + "grad_norm": 4.288512706756592, + "learning_rate": 0.00019078727567310732, + "loss": 0.4581, + "step": 5883 + }, + { + "epoch": 0.13793022615259293, + "grad_norm": 4.42967414855957, + "learning_rate": 0.0001907841871959226, + "loss": 0.6627, + "step": 5884 + }, + { + "epoch": 0.13795366772739792, + "grad_norm": 1.3397706747055054, + "learning_rate": 0.00019078109822614039, + "loss": 0.1308, + "step": 5885 + }, + { + "epoch": 0.13797710930220292, + "grad_norm": 0.886443018913269, + "learning_rate": 0.00019077800876377734, + "loss": 0.1258, + "step": 5886 + }, + { + "epoch": 0.1380005508770079, + "grad_norm": 3.419341564178467, + "learning_rate": 0.00019077491880885027, + "loss": 0.4697, + "step": 5887 + }, + { + "epoch": 0.1380239924518129, + "grad_norm": 4.194849014282227, + "learning_rate": 0.00019077182836137595, + "loss": 0.6329, + "step": 5888 + }, + { + "epoch": 0.1380474340266179, + "grad_norm": 2.629101037979126, + "learning_rate": 0.00019076873742137113, + "loss": 1.8683, + "step": 5889 + }, + { + "epoch": 0.1380708756014229, + "grad_norm": 2.9792368412017822, + "learning_rate": 0.00019076564598885257, + "loss": 0.3443, + "step": 5890 + }, + { + "epoch": 0.1380943171762279, + "grad_norm": 2.7985997200012207, + "learning_rate": 0.0001907625540638371, + "loss": 0.3425, + "step": 5891 + }, + { + "epoch": 0.13811775875103288, + "grad_norm": 4.104649066925049, + "learning_rate": 0.00019075946164634141, + "loss": 0.5469, + "step": 5892 + }, + { + "epoch": 0.13814120032583788, + "grad_norm": 1.9632139205932617, + "learning_rate": 0.00019075636873638237, + "loss": 0.1493, + "step": 5893 + }, + { + "epoch": 0.1381646419006429, + "grad_norm": 0.6769276261329651, + "learning_rate": 0.00019075327533397672, + "loss": 0.1068, + "step": 5894 + }, + { + "epoch": 0.1381880834754479, + "grad_norm": 3.833958864212036, + "learning_rate": 0.00019075018143914125, + "loss": 0.2594, + "step": 5895 + }, + { + "epoch": 0.1382115250502529, + "grad_norm": 1.9139395952224731, + "learning_rate": 0.00019074708705189272, + "loss": 0.215, + "step": 5896 + }, + { + "epoch": 0.13823496662505788, + "grad_norm": 3.6012794971466064, + "learning_rate": 0.00019074399217224795, + "loss": 0.5602, + "step": 5897 + }, + { + "epoch": 0.13825840819986288, + "grad_norm": 0.3128526210784912, + "learning_rate": 0.00019074089680022374, + "loss": 0.0326, + "step": 5898 + }, + { + "epoch": 0.13828184977466787, + "grad_norm": 0.9795439839363098, + "learning_rate": 0.00019073780093583685, + "loss": 0.1481, + "step": 5899 + }, + { + "epoch": 0.13830529134947286, + "grad_norm": 3.518580675125122, + "learning_rate": 0.00019073470457910412, + "loss": 0.6462, + "step": 5900 + }, + { + "epoch": 0.13832873292427786, + "grad_norm": 0.53682941198349, + "learning_rate": 0.00019073160773004232, + "loss": 0.0469, + "step": 5901 + }, + { + "epoch": 0.13835217449908285, + "grad_norm": 5.5705952644348145, + "learning_rate": 0.00019072851038866827, + "loss": 0.5452, + "step": 5902 + }, + { + "epoch": 0.13837561607388785, + "grad_norm": 3.3501791954040527, + "learning_rate": 0.00019072541255499878, + "loss": 0.3439, + "step": 5903 + }, + { + "epoch": 0.13839905764869284, + "grad_norm": 2.3465030193328857, + "learning_rate": 0.00019072231422905065, + "loss": 1.251, + "step": 5904 + }, + { + "epoch": 0.13842249922349784, + "grad_norm": 1.2031677961349487, + "learning_rate": 0.00019071921541084067, + "loss": 0.0876, + "step": 5905 + }, + { + "epoch": 0.13844594079830283, + "grad_norm": 5.216805458068848, + "learning_rate": 0.0001907161161003857, + "loss": 0.6561, + "step": 5906 + }, + { + "epoch": 0.13846938237310782, + "grad_norm": 4.349993705749512, + "learning_rate": 0.00019071301629770254, + "loss": 0.6102, + "step": 5907 + }, + { + "epoch": 0.13849282394791282, + "grad_norm": 3.2330923080444336, + "learning_rate": 0.00019070991600280798, + "loss": 0.4465, + "step": 5908 + }, + { + "epoch": 0.1385162655227178, + "grad_norm": 4.270643711090088, + "learning_rate": 0.00019070681521571886, + "loss": 0.5181, + "step": 5909 + }, + { + "epoch": 0.1385397070975228, + "grad_norm": 1.1905450820922852, + "learning_rate": 0.00019070371393645202, + "loss": 0.1099, + "step": 5910 + }, + { + "epoch": 0.1385631486723278, + "grad_norm": 0.8688432574272156, + "learning_rate": 0.0001907006121650243, + "loss": 0.1315, + "step": 5911 + }, + { + "epoch": 0.1385865902471328, + "grad_norm": 1.0660972595214844, + "learning_rate": 0.00019069750990145248, + "loss": 0.1443, + "step": 5912 + }, + { + "epoch": 0.1386100318219378, + "grad_norm": 2.5919370651245117, + "learning_rate": 0.00019069440714575343, + "loss": 0.4211, + "step": 5913 + }, + { + "epoch": 0.13863347339674278, + "grad_norm": 2.4959559440612793, + "learning_rate": 0.000190691303897944, + "loss": 0.2941, + "step": 5914 + }, + { + "epoch": 0.13865691497154778, + "grad_norm": 4.184417247772217, + "learning_rate": 0.00019068820015804097, + "loss": 0.3882, + "step": 5915 + }, + { + "epoch": 0.13868035654635277, + "grad_norm": 2.6019558906555176, + "learning_rate": 0.00019068509592606124, + "loss": 1.5969, + "step": 5916 + }, + { + "epoch": 0.13870379812115777, + "grad_norm": 0.7132258415222168, + "learning_rate": 0.00019068199120202158, + "loss": 0.1035, + "step": 5917 + }, + { + "epoch": 0.13872723969596276, + "grad_norm": 4.654780864715576, + "learning_rate": 0.00019067888598593894, + "loss": 0.5991, + "step": 5918 + }, + { + "epoch": 0.13875068127076778, + "grad_norm": 2.9337897300720215, + "learning_rate": 0.0001906757802778301, + "loss": 0.4689, + "step": 5919 + }, + { + "epoch": 0.13877412284557278, + "grad_norm": 7.175079345703125, + "learning_rate": 0.0001906726740777119, + "loss": 0.8589, + "step": 5920 + }, + { + "epoch": 0.13879756442037777, + "grad_norm": 3.2335705757141113, + "learning_rate": 0.00019066956738560128, + "loss": 0.4271, + "step": 5921 + }, + { + "epoch": 0.13882100599518277, + "grad_norm": 2.721104621887207, + "learning_rate": 0.000190666460201515, + "loss": 0.437, + "step": 5922 + }, + { + "epoch": 0.13884444756998776, + "grad_norm": 1.0899624824523926, + "learning_rate": 0.00019066335252546996, + "loss": 0.194, + "step": 5923 + }, + { + "epoch": 0.13886788914479276, + "grad_norm": 5.431762218475342, + "learning_rate": 0.00019066024435748303, + "loss": 0.9182, + "step": 5924 + }, + { + "epoch": 0.13889133071959775, + "grad_norm": 4.64771032333374, + "learning_rate": 0.00019065713569757105, + "loss": 0.6205, + "step": 5925 + }, + { + "epoch": 0.13891477229440274, + "grad_norm": 2.261612892150879, + "learning_rate": 0.0001906540265457509, + "loss": 1.6317, + "step": 5926 + }, + { + "epoch": 0.13893821386920774, + "grad_norm": 4.171730041503906, + "learning_rate": 0.00019065091690203948, + "loss": 0.4856, + "step": 5927 + }, + { + "epoch": 0.13896165544401273, + "grad_norm": 5.485893726348877, + "learning_rate": 0.00019064780676645362, + "loss": 0.901, + "step": 5928 + }, + { + "epoch": 0.13898509701881773, + "grad_norm": 3.6850011348724365, + "learning_rate": 0.00019064469613901022, + "loss": 0.589, + "step": 5929 + }, + { + "epoch": 0.13900853859362272, + "grad_norm": 1.5236396789550781, + "learning_rate": 0.00019064158501972612, + "loss": 0.252, + "step": 5930 + }, + { + "epoch": 0.13903198016842772, + "grad_norm": 2.7616751194000244, + "learning_rate": 0.00019063847340861827, + "loss": 0.4065, + "step": 5931 + }, + { + "epoch": 0.1390554217432327, + "grad_norm": 3.1800057888031006, + "learning_rate": 0.0001906353613057035, + "loss": 0.3725, + "step": 5932 + }, + { + "epoch": 0.1390788633180377, + "grad_norm": 1.1704940795898438, + "learning_rate": 0.0001906322487109987, + "loss": 0.1046, + "step": 5933 + }, + { + "epoch": 0.1391023048928427, + "grad_norm": 4.166507244110107, + "learning_rate": 0.00019062913562452078, + "loss": 0.518, + "step": 5934 + }, + { + "epoch": 0.1391257464676477, + "grad_norm": 1.505953073501587, + "learning_rate": 0.00019062602204628664, + "loss": 0.121, + "step": 5935 + }, + { + "epoch": 0.1391491880424527, + "grad_norm": 0.8894460797309875, + "learning_rate": 0.0001906229079763131, + "loss": 0.14, + "step": 5936 + }, + { + "epoch": 0.13917262961725768, + "grad_norm": 2.1819136142730713, + "learning_rate": 0.00019061979341461718, + "loss": 0.319, + "step": 5937 + }, + { + "epoch": 0.13919607119206268, + "grad_norm": 6.352422714233398, + "learning_rate": 0.00019061667836121565, + "loss": 0.7691, + "step": 5938 + }, + { + "epoch": 0.13921951276686767, + "grad_norm": 3.082948923110962, + "learning_rate": 0.00019061356281612553, + "loss": 0.5879, + "step": 5939 + }, + { + "epoch": 0.13924295434167266, + "grad_norm": 3.430968761444092, + "learning_rate": 0.00019061044677936363, + "loss": 0.3225, + "step": 5940 + }, + { + "epoch": 0.13926639591647766, + "grad_norm": 5.7378458976745605, + "learning_rate": 0.00019060733025094692, + "loss": 0.8118, + "step": 5941 + }, + { + "epoch": 0.13928983749128265, + "grad_norm": 4.694289207458496, + "learning_rate": 0.0001906042132308923, + "loss": 0.9117, + "step": 5942 + }, + { + "epoch": 0.13931327906608765, + "grad_norm": 2.741875410079956, + "learning_rate": 0.00019060109571921664, + "loss": 1.0831, + "step": 5943 + }, + { + "epoch": 0.13933672064089264, + "grad_norm": 2.986882448196411, + "learning_rate": 0.00019059797771593692, + "loss": 0.4247, + "step": 5944 + }, + { + "epoch": 0.13936016221569766, + "grad_norm": 1.300188422203064, + "learning_rate": 0.00019059485922107, + "loss": 0.2195, + "step": 5945 + }, + { + "epoch": 0.13938360379050266, + "grad_norm": 3.904057025909424, + "learning_rate": 0.00019059174023463283, + "loss": 0.7974, + "step": 5946 + }, + { + "epoch": 0.13940704536530765, + "grad_norm": 1.6627253293991089, + "learning_rate": 0.00019058862075664235, + "loss": 1.062, + "step": 5947 + }, + { + "epoch": 0.13943048694011265, + "grad_norm": 2.673105239868164, + "learning_rate": 0.00019058550078711542, + "loss": 0.384, + "step": 5948 + }, + { + "epoch": 0.13945392851491764, + "grad_norm": 4.453822612762451, + "learning_rate": 0.00019058238032606907, + "loss": 0.8097, + "step": 5949 + }, + { + "epoch": 0.13947737008972264, + "grad_norm": 4.72755765914917, + "learning_rate": 0.00019057925937352016, + "loss": 0.815, + "step": 5950 + }, + { + "epoch": 0.13950081166452763, + "grad_norm": 3.196220874786377, + "learning_rate": 0.00019057613792948562, + "loss": 0.4682, + "step": 5951 + }, + { + "epoch": 0.13952425323933262, + "grad_norm": 3.6864640712738037, + "learning_rate": 0.00019057301599398243, + "loss": 0.5204, + "step": 5952 + }, + { + "epoch": 0.13954769481413762, + "grad_norm": 4.4878034591674805, + "learning_rate": 0.00019056989356702752, + "loss": 0.6381, + "step": 5953 + }, + { + "epoch": 0.1395711363889426, + "grad_norm": 1.0140241384506226, + "learning_rate": 0.0001905667706486378, + "loss": 0.1478, + "step": 5954 + }, + { + "epoch": 0.1395945779637476, + "grad_norm": 2.6290628910064697, + "learning_rate": 0.00019056364723883022, + "loss": 0.3394, + "step": 5955 + }, + { + "epoch": 0.1396180195385526, + "grad_norm": 0.5904884934425354, + "learning_rate": 0.00019056052333762177, + "loss": 0.0724, + "step": 5956 + }, + { + "epoch": 0.1396414611133576, + "grad_norm": 2.742896795272827, + "learning_rate": 0.00019055739894502938, + "loss": 0.5146, + "step": 5957 + }, + { + "epoch": 0.1396649026881626, + "grad_norm": 3.4406259059906006, + "learning_rate": 0.00019055427406106999, + "loss": 0.5843, + "step": 5958 + }, + { + "epoch": 0.13968834426296758, + "grad_norm": 4.278403282165527, + "learning_rate": 0.00019055114868576055, + "loss": 0.6594, + "step": 5959 + }, + { + "epoch": 0.13971178583777258, + "grad_norm": 1.6655641794204712, + "learning_rate": 0.00019054802281911803, + "loss": 0.1594, + "step": 5960 + }, + { + "epoch": 0.13973522741257757, + "grad_norm": 0.5905565023422241, + "learning_rate": 0.0001905448964611594, + "loss": 0.0571, + "step": 5961 + }, + { + "epoch": 0.13975866898738257, + "grad_norm": 3.7156858444213867, + "learning_rate": 0.00019054176961190162, + "loss": 0.4374, + "step": 5962 + }, + { + "epoch": 0.13978211056218756, + "grad_norm": 3.770982265472412, + "learning_rate": 0.00019053864227136166, + "loss": 0.4235, + "step": 5963 + }, + { + "epoch": 0.13980555213699256, + "grad_norm": 2.3241970539093018, + "learning_rate": 0.00019053551443955647, + "loss": 0.3133, + "step": 5964 + }, + { + "epoch": 0.13982899371179755, + "grad_norm": 3.263777017593384, + "learning_rate": 0.00019053238611650304, + "loss": 0.517, + "step": 5965 + }, + { + "epoch": 0.13985243528660254, + "grad_norm": 2.216353178024292, + "learning_rate": 0.00019052925730221834, + "loss": 1.2686, + "step": 5966 + }, + { + "epoch": 0.13987587686140754, + "grad_norm": 1.694501519203186, + "learning_rate": 0.00019052612799671931, + "loss": 0.2048, + "step": 5967 + }, + { + "epoch": 0.13989931843621253, + "grad_norm": 3.081667423248291, + "learning_rate": 0.000190522998200023, + "loss": 0.3875, + "step": 5968 + }, + { + "epoch": 0.13992276001101753, + "grad_norm": 3.4496285915374756, + "learning_rate": 0.00019051986791214636, + "loss": 0.295, + "step": 5969 + }, + { + "epoch": 0.13994620158582255, + "grad_norm": 2.926934242248535, + "learning_rate": 0.00019051673713310635, + "loss": 0.3302, + "step": 5970 + }, + { + "epoch": 0.13996964316062754, + "grad_norm": 3.5278565883636475, + "learning_rate": 0.00019051360586291998, + "loss": 2.1167, + "step": 5971 + }, + { + "epoch": 0.13999308473543254, + "grad_norm": 0.8770354390144348, + "learning_rate": 0.00019051047410160427, + "loss": 0.122, + "step": 5972 + }, + { + "epoch": 0.14001652631023753, + "grad_norm": 3.0445241928100586, + "learning_rate": 0.00019050734184917616, + "loss": 0.3757, + "step": 5973 + }, + { + "epoch": 0.14003996788504253, + "grad_norm": 5.984563827514648, + "learning_rate": 0.00019050420910565267, + "loss": 0.9868, + "step": 5974 + }, + { + "epoch": 0.14006340945984752, + "grad_norm": 2.119518756866455, + "learning_rate": 0.00019050107587105077, + "loss": 0.1886, + "step": 5975 + }, + { + "epoch": 0.14008685103465252, + "grad_norm": 1.297054409980774, + "learning_rate": 0.00019049794214538753, + "loss": 0.2438, + "step": 5976 + }, + { + "epoch": 0.1401102926094575, + "grad_norm": 1.9358365535736084, + "learning_rate": 0.00019049480792867989, + "loss": 0.2028, + "step": 5977 + }, + { + "epoch": 0.1401337341842625, + "grad_norm": 4.485905170440674, + "learning_rate": 0.00019049167322094488, + "loss": 0.4513, + "step": 5978 + }, + { + "epoch": 0.1401571757590675, + "grad_norm": 2.007038116455078, + "learning_rate": 0.00019048853802219952, + "loss": 0.2354, + "step": 5979 + }, + { + "epoch": 0.1401806173338725, + "grad_norm": 4.885457992553711, + "learning_rate": 0.00019048540233246078, + "loss": 0.4234, + "step": 5980 + }, + { + "epoch": 0.1402040589086775, + "grad_norm": 1.8054333925247192, + "learning_rate": 0.00019048226615174573, + "loss": 0.2294, + "step": 5981 + }, + { + "epoch": 0.14022750048348248, + "grad_norm": 4.73178243637085, + "learning_rate": 0.00019047912948007132, + "loss": 0.6061, + "step": 5982 + }, + { + "epoch": 0.14025094205828748, + "grad_norm": 6.89371395111084, + "learning_rate": 0.00019047599231745464, + "loss": 0.9401, + "step": 5983 + }, + { + "epoch": 0.14027438363309247, + "grad_norm": 3.8654379844665527, + "learning_rate": 0.00019047285466391268, + "loss": 0.4745, + "step": 5984 + }, + { + "epoch": 0.14029782520789746, + "grad_norm": 2.2664411067962646, + "learning_rate": 0.0001904697165194625, + "loss": 0.3762, + "step": 5985 + }, + { + "epoch": 0.14032126678270246, + "grad_norm": 1.4009792804718018, + "learning_rate": 0.00019046657788412104, + "loss": 0.1025, + "step": 5986 + }, + { + "epoch": 0.14034470835750745, + "grad_norm": 2.6906983852386475, + "learning_rate": 0.0001904634387579054, + "loss": 1.4269, + "step": 5987 + }, + { + "epoch": 0.14036814993231245, + "grad_norm": 4.662012577056885, + "learning_rate": 0.00019046029914083262, + "loss": 0.6591, + "step": 5988 + }, + { + "epoch": 0.14039159150711744, + "grad_norm": 3.7391738891601562, + "learning_rate": 0.0001904571590329197, + "loss": 0.4664, + "step": 5989 + }, + { + "epoch": 0.14041503308192244, + "grad_norm": 4.43838357925415, + "learning_rate": 0.00019045401843418368, + "loss": 0.6618, + "step": 5990 + }, + { + "epoch": 0.14043847465672743, + "grad_norm": 4.6764068603515625, + "learning_rate": 0.00019045087734464162, + "loss": 0.5134, + "step": 5991 + }, + { + "epoch": 0.14046191623153242, + "grad_norm": 2.875753879547119, + "learning_rate": 0.00019044773576431054, + "loss": 0.4262, + "step": 5992 + }, + { + "epoch": 0.14048535780633742, + "grad_norm": 5.297171115875244, + "learning_rate": 0.00019044459369320753, + "loss": 0.9635, + "step": 5993 + }, + { + "epoch": 0.1405087993811424, + "grad_norm": 3.5833332538604736, + "learning_rate": 0.0001904414511313496, + "loss": 0.5195, + "step": 5994 + }, + { + "epoch": 0.1405322409559474, + "grad_norm": 2.697847366333008, + "learning_rate": 0.00019043830807875383, + "loss": 0.453, + "step": 5995 + }, + { + "epoch": 0.14055568253075243, + "grad_norm": 2.4783389568328857, + "learning_rate": 0.00019043516453543725, + "loss": 0.2869, + "step": 5996 + }, + { + "epoch": 0.14057912410555742, + "grad_norm": 3.446756601333618, + "learning_rate": 0.0001904320205014169, + "loss": 0.5884, + "step": 5997 + }, + { + "epoch": 0.14060256568036242, + "grad_norm": 3.4651951789855957, + "learning_rate": 0.0001904288759767099, + "loss": 0.4, + "step": 5998 + }, + { + "epoch": 0.1406260072551674, + "grad_norm": 0.9076058864593506, + "learning_rate": 0.00019042573096133322, + "loss": 0.0931, + "step": 5999 + }, + { + "epoch": 0.1406494488299724, + "grad_norm": 2.319754123687744, + "learning_rate": 0.00019042258545530402, + "loss": 0.41, + "step": 6000 + }, + { + "epoch": 0.1406728904047774, + "grad_norm": 1.3493123054504395, + "learning_rate": 0.00019041943945863934, + "loss": 0.1989, + "step": 6001 + }, + { + "epoch": 0.1406963319795824, + "grad_norm": 1.4439451694488525, + "learning_rate": 0.00019041629297135621, + "loss": 0.183, + "step": 6002 + }, + { + "epoch": 0.1407197735543874, + "grad_norm": 2.9652585983276367, + "learning_rate": 0.00019041314599347174, + "loss": 0.2769, + "step": 6003 + }, + { + "epoch": 0.14074321512919238, + "grad_norm": 3.5046041011810303, + "learning_rate": 0.00019040999852500302, + "loss": 0.3512, + "step": 6004 + }, + { + "epoch": 0.14076665670399738, + "grad_norm": 2.8345439434051514, + "learning_rate": 0.0001904068505659671, + "loss": 0.3663, + "step": 6005 + }, + { + "epoch": 0.14079009827880237, + "grad_norm": 1.8152110576629639, + "learning_rate": 0.00019040370211638105, + "loss": 0.4006, + "step": 6006 + }, + { + "epoch": 0.14081353985360737, + "grad_norm": 7.405877590179443, + "learning_rate": 0.000190400553176262, + "loss": 0.9978, + "step": 6007 + }, + { + "epoch": 0.14083698142841236, + "grad_norm": 1.2299712896347046, + "learning_rate": 0.00019039740374562696, + "loss": 0.1457, + "step": 6008 + }, + { + "epoch": 0.14086042300321736, + "grad_norm": 1.3928451538085938, + "learning_rate": 0.0001903942538244931, + "loss": 0.2527, + "step": 6009 + }, + { + "epoch": 0.14088386457802235, + "grad_norm": 0.6485745906829834, + "learning_rate": 0.00019039110341287748, + "loss": 0.081, + "step": 6010 + }, + { + "epoch": 0.14090730615282734, + "grad_norm": 1.1208350658416748, + "learning_rate": 0.00019038795251079718, + "loss": 0.111, + "step": 6011 + }, + { + "epoch": 0.14093074772763234, + "grad_norm": 3.1892926692962646, + "learning_rate": 0.0001903848011182693, + "loss": 0.4002, + "step": 6012 + }, + { + "epoch": 0.14095418930243733, + "grad_norm": 2.4915547370910645, + "learning_rate": 0.00019038164923531093, + "loss": 0.2666, + "step": 6013 + }, + { + "epoch": 0.14097763087724233, + "grad_norm": 4.2535600662231445, + "learning_rate": 0.00019037849686193923, + "loss": 0.328, + "step": 6014 + }, + { + "epoch": 0.14100107245204732, + "grad_norm": 3.0017354488372803, + "learning_rate": 0.00019037534399817126, + "loss": 0.4837, + "step": 6015 + }, + { + "epoch": 0.14102451402685232, + "grad_norm": 3.633310556411743, + "learning_rate": 0.00019037219064402413, + "loss": 0.4456, + "step": 6016 + }, + { + "epoch": 0.1410479556016573, + "grad_norm": 4.20515251159668, + "learning_rate": 0.00019036903679951493, + "loss": 0.7055, + "step": 6017 + }, + { + "epoch": 0.1410713971764623, + "grad_norm": 0.7325958013534546, + "learning_rate": 0.00019036588246466082, + "loss": 0.0957, + "step": 6018 + }, + { + "epoch": 0.1410948387512673, + "grad_norm": 3.914619207382202, + "learning_rate": 0.00019036272763947885, + "loss": 0.6034, + "step": 6019 + }, + { + "epoch": 0.1411182803260723, + "grad_norm": 6.87346887588501, + "learning_rate": 0.00019035957232398624, + "loss": 0.9494, + "step": 6020 + }, + { + "epoch": 0.1411417219008773, + "grad_norm": 3.6264634132385254, + "learning_rate": 0.0001903564165182, + "loss": 0.5822, + "step": 6021 + }, + { + "epoch": 0.1411651634756823, + "grad_norm": 2.245224952697754, + "learning_rate": 0.00019035326022213736, + "loss": 0.3802, + "step": 6022 + }, + { + "epoch": 0.1411886050504873, + "grad_norm": 1.5141451358795166, + "learning_rate": 0.00019035010343581532, + "loss": 0.1159, + "step": 6023 + }, + { + "epoch": 0.1412120466252923, + "grad_norm": 2.8847508430480957, + "learning_rate": 0.00019034694615925114, + "loss": 0.4105, + "step": 6024 + }, + { + "epoch": 0.1412354882000973, + "grad_norm": 8.090692520141602, + "learning_rate": 0.00019034378839246182, + "loss": 0.5907, + "step": 6025 + }, + { + "epoch": 0.1412589297749023, + "grad_norm": 1.712733507156372, + "learning_rate": 0.0001903406301354646, + "loss": 0.223, + "step": 6026 + }, + { + "epoch": 0.14128237134970728, + "grad_norm": 2.618776798248291, + "learning_rate": 0.00019033747138827662, + "loss": 0.3874, + "step": 6027 + }, + { + "epoch": 0.14130581292451228, + "grad_norm": 0.8943840265274048, + "learning_rate": 0.00019033431215091496, + "loss": 0.1378, + "step": 6028 + }, + { + "epoch": 0.14132925449931727, + "grad_norm": 3.3130381107330322, + "learning_rate": 0.00019033115242339673, + "loss": 0.1987, + "step": 6029 + }, + { + "epoch": 0.14135269607412226, + "grad_norm": 3.480587959289551, + "learning_rate": 0.00019032799220573918, + "loss": 0.4773, + "step": 6030 + }, + { + "epoch": 0.14137613764892726, + "grad_norm": 3.9038634300231934, + "learning_rate": 0.00019032483149795942, + "loss": 0.6841, + "step": 6031 + }, + { + "epoch": 0.14139957922373225, + "grad_norm": 2.490114212036133, + "learning_rate": 0.00019032167030007453, + "loss": 0.4017, + "step": 6032 + }, + { + "epoch": 0.14142302079853725, + "grad_norm": 3.20070743560791, + "learning_rate": 0.00019031850861210174, + "loss": 0.4613, + "step": 6033 + }, + { + "epoch": 0.14144646237334224, + "grad_norm": 4.313375473022461, + "learning_rate": 0.00019031534643405818, + "loss": 0.6165, + "step": 6034 + }, + { + "epoch": 0.14146990394814724, + "grad_norm": 4.461740970611572, + "learning_rate": 0.00019031218376596102, + "loss": 0.6922, + "step": 6035 + }, + { + "epoch": 0.14149334552295223, + "grad_norm": 3.535060167312622, + "learning_rate": 0.0001903090206078274, + "loss": 0.5708, + "step": 6036 + }, + { + "epoch": 0.14151678709775722, + "grad_norm": 2.843160629272461, + "learning_rate": 0.00019030585695967452, + "loss": 1.5515, + "step": 6037 + }, + { + "epoch": 0.14154022867256222, + "grad_norm": 1.4796226024627686, + "learning_rate": 0.0001903026928215195, + "loss": 0.276, + "step": 6038 + }, + { + "epoch": 0.1415636702473672, + "grad_norm": 4.277408599853516, + "learning_rate": 0.00019029952819337953, + "loss": 0.5529, + "step": 6039 + }, + { + "epoch": 0.1415871118221722, + "grad_norm": 1.7070525884628296, + "learning_rate": 0.0001902963630752718, + "loss": 0.2095, + "step": 6040 + }, + { + "epoch": 0.1416105533969772, + "grad_norm": 4.5393571853637695, + "learning_rate": 0.00019029319746721344, + "loss": 0.8356, + "step": 6041 + }, + { + "epoch": 0.1416339949717822, + "grad_norm": 2.8010730743408203, + "learning_rate": 0.00019029003136922164, + "loss": 0.2672, + "step": 6042 + }, + { + "epoch": 0.1416574365465872, + "grad_norm": 1.1894433498382568, + "learning_rate": 0.00019028686478131363, + "loss": 0.1665, + "step": 6043 + }, + { + "epoch": 0.14168087812139218, + "grad_norm": 3.4912312030792236, + "learning_rate": 0.0001902836977035065, + "loss": 0.6047, + "step": 6044 + }, + { + "epoch": 0.14170431969619718, + "grad_norm": 3.913236141204834, + "learning_rate": 0.00019028053013581752, + "loss": 0.6191, + "step": 6045 + }, + { + "epoch": 0.14172776127100217, + "grad_norm": 2.649179697036743, + "learning_rate": 0.00019027736207826386, + "loss": 0.3325, + "step": 6046 + }, + { + "epoch": 0.1417512028458072, + "grad_norm": 1.424628496170044, + "learning_rate": 0.00019027419353086268, + "loss": 0.8705, + "step": 6047 + }, + { + "epoch": 0.1417746444206122, + "grad_norm": 3.4843506813049316, + "learning_rate": 0.0001902710244936312, + "loss": 0.4233, + "step": 6048 + }, + { + "epoch": 0.14179808599541718, + "grad_norm": 5.590022087097168, + "learning_rate": 0.00019026785496658658, + "loss": 0.6526, + "step": 6049 + }, + { + "epoch": 0.14182152757022218, + "grad_norm": 5.585735321044922, + "learning_rate": 0.00019026468494974608, + "loss": 0.7809, + "step": 6050 + }, + { + "epoch": 0.14184496914502717, + "grad_norm": 2.255760669708252, + "learning_rate": 0.00019026151444312682, + "loss": 1.1649, + "step": 6051 + }, + { + "epoch": 0.14186841071983217, + "grad_norm": 1.0918442010879517, + "learning_rate": 0.00019025834344674607, + "loss": 0.183, + "step": 6052 + }, + { + "epoch": 0.14189185229463716, + "grad_norm": 0.7867177128791809, + "learning_rate": 0.000190255171960621, + "loss": 0.1343, + "step": 6053 + }, + { + "epoch": 0.14191529386944216, + "grad_norm": 4.013588905334473, + "learning_rate": 0.00019025199998476882, + "loss": 0.57, + "step": 6054 + }, + { + "epoch": 0.14193873544424715, + "grad_norm": 2.106926918029785, + "learning_rate": 0.00019024882751920676, + "loss": 0.309, + "step": 6055 + }, + { + "epoch": 0.14196217701905214, + "grad_norm": 3.1443569660186768, + "learning_rate": 0.00019024565456395202, + "loss": 0.5391, + "step": 6056 + }, + { + "epoch": 0.14198561859385714, + "grad_norm": 3.473388433456421, + "learning_rate": 0.00019024248111902182, + "loss": 0.4352, + "step": 6057 + }, + { + "epoch": 0.14200906016866213, + "grad_norm": 4.952681541442871, + "learning_rate": 0.0001902393071844334, + "loss": 0.5163, + "step": 6058 + }, + { + "epoch": 0.14203250174346713, + "grad_norm": 1.7912516593933105, + "learning_rate": 0.00019023613276020394, + "loss": 1.3705, + "step": 6059 + }, + { + "epoch": 0.14205594331827212, + "grad_norm": 2.311964988708496, + "learning_rate": 0.0001902329578463507, + "loss": 0.1818, + "step": 6060 + }, + { + "epoch": 0.14207938489307712, + "grad_norm": 2.841867446899414, + "learning_rate": 0.00019022978244289087, + "loss": 1.6235, + "step": 6061 + }, + { + "epoch": 0.1421028264678821, + "grad_norm": 1.6299337148666382, + "learning_rate": 0.00019022660654984176, + "loss": 0.2322, + "step": 6062 + }, + { + "epoch": 0.1421262680426871, + "grad_norm": 6.025466442108154, + "learning_rate": 0.00019022343016722047, + "loss": 1.0632, + "step": 6063 + }, + { + "epoch": 0.1421497096174921, + "grad_norm": 6.167788028717041, + "learning_rate": 0.00019022025329504437, + "loss": 0.6694, + "step": 6064 + }, + { + "epoch": 0.1421731511922971, + "grad_norm": 2.7054998874664307, + "learning_rate": 0.0001902170759333306, + "loss": 0.2192, + "step": 6065 + }, + { + "epoch": 0.1421965927671021, + "grad_norm": 4.485992431640625, + "learning_rate": 0.00019021389808209646, + "loss": 0.5633, + "step": 6066 + }, + { + "epoch": 0.14222003434190708, + "grad_norm": 0.9668811559677124, + "learning_rate": 0.0001902107197413592, + "loss": 0.1249, + "step": 6067 + }, + { + "epoch": 0.14224347591671208, + "grad_norm": 3.04433274269104, + "learning_rate": 0.00019020754091113597, + "loss": 0.4123, + "step": 6068 + }, + { + "epoch": 0.14226691749151707, + "grad_norm": 5.030875205993652, + "learning_rate": 0.00019020436159144412, + "loss": 0.4786, + "step": 6069 + }, + { + "epoch": 0.14229035906632206, + "grad_norm": 2.8502845764160156, + "learning_rate": 0.00019020118178230089, + "loss": 0.3585, + "step": 6070 + }, + { + "epoch": 0.14231380064112706, + "grad_norm": 4.197671413421631, + "learning_rate": 0.00019019800148372347, + "loss": 0.6736, + "step": 6071 + }, + { + "epoch": 0.14233724221593205, + "grad_norm": 3.232283115386963, + "learning_rate": 0.00019019482069572918, + "loss": 0.6071, + "step": 6072 + }, + { + "epoch": 0.14236068379073707, + "grad_norm": 4.449595928192139, + "learning_rate": 0.00019019163941833526, + "loss": 0.6744, + "step": 6073 + }, + { + "epoch": 0.14238412536554207, + "grad_norm": 1.336146354675293, + "learning_rate": 0.00019018845765155897, + "loss": 0.1576, + "step": 6074 + }, + { + "epoch": 0.14240756694034706, + "grad_norm": 0.7760056257247925, + "learning_rate": 0.00019018527539541754, + "loss": 0.1003, + "step": 6075 + }, + { + "epoch": 0.14243100851515206, + "grad_norm": 5.109124660491943, + "learning_rate": 0.00019018209264992831, + "loss": 0.585, + "step": 6076 + }, + { + "epoch": 0.14245445008995705, + "grad_norm": 3.0228211879730225, + "learning_rate": 0.0001901789094151085, + "loss": 0.4591, + "step": 6077 + }, + { + "epoch": 0.14247789166476205, + "grad_norm": 0.951941192150116, + "learning_rate": 0.00019017572569097538, + "loss": 0.1213, + "step": 6078 + }, + { + "epoch": 0.14250133323956704, + "grad_norm": 3.4700443744659424, + "learning_rate": 0.00019017254147754627, + "loss": 0.4185, + "step": 6079 + }, + { + "epoch": 0.14252477481437203, + "grad_norm": 2.9848835468292236, + "learning_rate": 0.0001901693567748384, + "loss": 0.4798, + "step": 6080 + }, + { + "epoch": 0.14254821638917703, + "grad_norm": 4.5064287185668945, + "learning_rate": 0.00019016617158286906, + "loss": 0.7395, + "step": 6081 + }, + { + "epoch": 0.14257165796398202, + "grad_norm": 4.6179609298706055, + "learning_rate": 0.00019016298590165552, + "loss": 0.4634, + "step": 6082 + }, + { + "epoch": 0.14259509953878702, + "grad_norm": 2.197465419769287, + "learning_rate": 0.00019015979973121508, + "loss": 0.1769, + "step": 6083 + }, + { + "epoch": 0.142618541113592, + "grad_norm": 5.681013107299805, + "learning_rate": 0.00019015661307156506, + "loss": 0.8496, + "step": 6084 + }, + { + "epoch": 0.142641982688397, + "grad_norm": 3.514406442642212, + "learning_rate": 0.00019015342592272272, + "loss": 0.5102, + "step": 6085 + }, + { + "epoch": 0.142665424263202, + "grad_norm": 3.8733842372894287, + "learning_rate": 0.00019015023828470535, + "loss": 0.5526, + "step": 6086 + }, + { + "epoch": 0.142688865838007, + "grad_norm": 1.3471633195877075, + "learning_rate": 0.00019014705015753025, + "loss": 0.1786, + "step": 6087 + }, + { + "epoch": 0.142712307412812, + "grad_norm": 3.9704747200012207, + "learning_rate": 0.0001901438615412147, + "loss": 0.483, + "step": 6088 + }, + { + "epoch": 0.14273574898761698, + "grad_norm": 2.2603635787963867, + "learning_rate": 0.00019014067243577605, + "loss": 0.2648, + "step": 6089 + }, + { + "epoch": 0.14275919056242198, + "grad_norm": 6.1695556640625, + "learning_rate": 0.0001901374828412316, + "loss": 0.5085, + "step": 6090 + }, + { + "epoch": 0.14278263213722697, + "grad_norm": 1.3391385078430176, + "learning_rate": 0.00019013429275759855, + "loss": 0.1498, + "step": 6091 + }, + { + "epoch": 0.14280607371203197, + "grad_norm": 4.541357517242432, + "learning_rate": 0.00019013110218489438, + "loss": 0.6905, + "step": 6092 + }, + { + "epoch": 0.14282951528683696, + "grad_norm": 2.0434937477111816, + "learning_rate": 0.00019012791112313625, + "loss": 1.41, + "step": 6093 + }, + { + "epoch": 0.14285295686164196, + "grad_norm": 9.39345932006836, + "learning_rate": 0.00019012471957234158, + "loss": 0.4837, + "step": 6094 + }, + { + "epoch": 0.14287639843644695, + "grad_norm": 4.869535446166992, + "learning_rate": 0.00019012152753252766, + "loss": 0.5172, + "step": 6095 + }, + { + "epoch": 0.14289984001125194, + "grad_norm": 1.274759292602539, + "learning_rate": 0.00019011833500371175, + "loss": 0.1343, + "step": 6096 + }, + { + "epoch": 0.14292328158605694, + "grad_norm": 3.150392532348633, + "learning_rate": 0.00019011514198591126, + "loss": 1.8335, + "step": 6097 + }, + { + "epoch": 0.14294672316086196, + "grad_norm": 5.548162937164307, + "learning_rate": 0.00019011194847914347, + "loss": 0.5557, + "step": 6098 + }, + { + "epoch": 0.14297016473566695, + "grad_norm": 3.4483187198638916, + "learning_rate": 0.0001901087544834257, + "loss": 2.1863, + "step": 6099 + }, + { + "epoch": 0.14299360631047195, + "grad_norm": 1.522087812423706, + "learning_rate": 0.0001901055599987753, + "loss": 0.251, + "step": 6100 + }, + { + "epoch": 0.14301704788527694, + "grad_norm": 0.8394440412521362, + "learning_rate": 0.00019010236502520962, + "loss": 0.1181, + "step": 6101 + }, + { + "epoch": 0.14304048946008194, + "grad_norm": 3.942963123321533, + "learning_rate": 0.00019009916956274595, + "loss": 0.3853, + "step": 6102 + }, + { + "epoch": 0.14306393103488693, + "grad_norm": 4.8394694328308105, + "learning_rate": 0.00019009597361140164, + "loss": 0.5487, + "step": 6103 + }, + { + "epoch": 0.14308737260969193, + "grad_norm": 1.8886314630508423, + "learning_rate": 0.00019009277717119404, + "loss": 0.1733, + "step": 6104 + }, + { + "epoch": 0.14311081418449692, + "grad_norm": 4.606661319732666, + "learning_rate": 0.00019008958024214055, + "loss": 0.3913, + "step": 6105 + }, + { + "epoch": 0.14313425575930191, + "grad_norm": 1.3871973752975464, + "learning_rate": 0.00019008638282425841, + "loss": 0.1633, + "step": 6106 + }, + { + "epoch": 0.1431576973341069, + "grad_norm": 3.5226492881774902, + "learning_rate": 0.00019008318491756505, + "loss": 0.6364, + "step": 6107 + }, + { + "epoch": 0.1431811389089119, + "grad_norm": 6.237707138061523, + "learning_rate": 0.0001900799865220778, + "loss": 0.7074, + "step": 6108 + }, + { + "epoch": 0.1432045804837169, + "grad_norm": 0.919215738773346, + "learning_rate": 0.000190076787637814, + "loss": 0.1127, + "step": 6109 + }, + { + "epoch": 0.1432280220585219, + "grad_norm": 1.26544988155365, + "learning_rate": 0.000190073588264791, + "loss": 0.2331, + "step": 6110 + }, + { + "epoch": 0.1432514636333269, + "grad_norm": 4.798377990722656, + "learning_rate": 0.00019007038840302621, + "loss": 0.6072, + "step": 6111 + }, + { + "epoch": 0.14327490520813188, + "grad_norm": 4.140806198120117, + "learning_rate": 0.00019006718805253696, + "loss": 0.8074, + "step": 6112 + }, + { + "epoch": 0.14329834678293688, + "grad_norm": 3.340538501739502, + "learning_rate": 0.0001900639872133406, + "loss": 0.5589, + "step": 6113 + }, + { + "epoch": 0.14332178835774187, + "grad_norm": 3.3681015968322754, + "learning_rate": 0.0001900607858854545, + "loss": 0.413, + "step": 6114 + }, + { + "epoch": 0.14334522993254686, + "grad_norm": 4.0625457763671875, + "learning_rate": 0.00019005758406889607, + "loss": 0.4139, + "step": 6115 + }, + { + "epoch": 0.14336867150735186, + "grad_norm": 1.1992398500442505, + "learning_rate": 0.00019005438176368264, + "loss": 0.083, + "step": 6116 + }, + { + "epoch": 0.14339211308215685, + "grad_norm": 4.640687465667725, + "learning_rate": 0.0001900511789698316, + "loss": 0.5553, + "step": 6117 + }, + { + "epoch": 0.14341555465696185, + "grad_norm": 4.023862838745117, + "learning_rate": 0.00019004797568736035, + "loss": 0.6548, + "step": 6118 + }, + { + "epoch": 0.14343899623176684, + "grad_norm": 4.064443588256836, + "learning_rate": 0.00019004477191628623, + "loss": 0.3775, + "step": 6119 + }, + { + "epoch": 0.14346243780657184, + "grad_norm": 4.0276641845703125, + "learning_rate": 0.00019004156765662668, + "loss": 0.5762, + "step": 6120 + }, + { + "epoch": 0.14348587938137683, + "grad_norm": 2.7903175354003906, + "learning_rate": 0.000190038362908399, + "loss": 1.694, + "step": 6121 + }, + { + "epoch": 0.14350932095618182, + "grad_norm": 1.4036197662353516, + "learning_rate": 0.00019003515767162068, + "loss": 0.2159, + "step": 6122 + }, + { + "epoch": 0.14353276253098682, + "grad_norm": 3.575101375579834, + "learning_rate": 0.00019003195194630906, + "loss": 0.3774, + "step": 6123 + }, + { + "epoch": 0.14355620410579184, + "grad_norm": 3.834010601043701, + "learning_rate": 0.0001900287457324815, + "loss": 0.6179, + "step": 6124 + }, + { + "epoch": 0.14357964568059683, + "grad_norm": 3.997783899307251, + "learning_rate": 0.00019002553903015544, + "loss": 0.5687, + "step": 6125 + }, + { + "epoch": 0.14360308725540183, + "grad_norm": 1.4770817756652832, + "learning_rate": 0.0001900223318393483, + "loss": 0.2116, + "step": 6126 + }, + { + "epoch": 0.14362652883020682, + "grad_norm": 5.74007511138916, + "learning_rate": 0.00019001912416007747, + "loss": 0.9723, + "step": 6127 + }, + { + "epoch": 0.14364997040501182, + "grad_norm": 4.180772304534912, + "learning_rate": 0.00019001591599236028, + "loss": 0.2637, + "step": 6128 + }, + { + "epoch": 0.1436734119798168, + "grad_norm": 2.2217750549316406, + "learning_rate": 0.00019001270733621427, + "loss": 0.2239, + "step": 6129 + }, + { + "epoch": 0.1436968535546218, + "grad_norm": 5.083395481109619, + "learning_rate": 0.0001900094981916567, + "loss": 0.4264, + "step": 6130 + }, + { + "epoch": 0.1437202951294268, + "grad_norm": 3.601029396057129, + "learning_rate": 0.00019000628855870514, + "loss": 0.59, + "step": 6131 + }, + { + "epoch": 0.1437437367042318, + "grad_norm": 3.9023263454437256, + "learning_rate": 0.00019000307843737688, + "loss": 0.7055, + "step": 6132 + }, + { + "epoch": 0.1437671782790368, + "grad_norm": 2.806509494781494, + "learning_rate": 0.0001899998678276894, + "loss": 1.6114, + "step": 6133 + }, + { + "epoch": 0.14379061985384178, + "grad_norm": 4.346275806427002, + "learning_rate": 0.0001899966567296601, + "loss": 0.5942, + "step": 6134 + }, + { + "epoch": 0.14381406142864678, + "grad_norm": 4.2325263023376465, + "learning_rate": 0.00018999344514330642, + "loss": 0.5963, + "step": 6135 + }, + { + "epoch": 0.14383750300345177, + "grad_norm": 3.7400121688842773, + "learning_rate": 0.00018999023306864575, + "loss": 0.2898, + "step": 6136 + }, + { + "epoch": 0.14386094457825677, + "grad_norm": 4.219250202178955, + "learning_rate": 0.00018998702050569558, + "loss": 0.9614, + "step": 6137 + }, + { + "epoch": 0.14388438615306176, + "grad_norm": 4.611416816711426, + "learning_rate": 0.0001899838074544733, + "loss": 0.5318, + "step": 6138 + }, + { + "epoch": 0.14390782772786676, + "grad_norm": 3.34065580368042, + "learning_rate": 0.00018998059391499633, + "loss": 0.4834, + "step": 6139 + }, + { + "epoch": 0.14393126930267175, + "grad_norm": 0.6886817812919617, + "learning_rate": 0.00018997737988728214, + "loss": 0.0449, + "step": 6140 + }, + { + "epoch": 0.14395471087747674, + "grad_norm": 0.9263150691986084, + "learning_rate": 0.0001899741653713482, + "loss": 0.0909, + "step": 6141 + }, + { + "epoch": 0.14397815245228174, + "grad_norm": 4.169551372528076, + "learning_rate": 0.00018997095036721185, + "loss": 0.6387, + "step": 6142 + }, + { + "epoch": 0.14400159402708673, + "grad_norm": 5.093668460845947, + "learning_rate": 0.0001899677348748906, + "loss": 0.9691, + "step": 6143 + }, + { + "epoch": 0.14402503560189173, + "grad_norm": 3.0306975841522217, + "learning_rate": 0.0001899645188944019, + "loss": 0.1613, + "step": 6144 + }, + { + "epoch": 0.14404847717669672, + "grad_norm": 3.9057633876800537, + "learning_rate": 0.0001899613024257632, + "loss": 0.5138, + "step": 6145 + }, + { + "epoch": 0.14407191875150172, + "grad_norm": 4.729757308959961, + "learning_rate": 0.0001899580854689919, + "loss": 0.8175, + "step": 6146 + }, + { + "epoch": 0.1440953603263067, + "grad_norm": 4.141150951385498, + "learning_rate": 0.00018995486802410555, + "loss": 0.551, + "step": 6147 + }, + { + "epoch": 0.1441188019011117, + "grad_norm": 2.841459274291992, + "learning_rate": 0.00018995165009112156, + "loss": 1.6079, + "step": 6148 + }, + { + "epoch": 0.14414224347591673, + "grad_norm": 3.1270151138305664, + "learning_rate": 0.00018994843167005737, + "loss": 0.5495, + "step": 6149 + }, + { + "epoch": 0.14416568505072172, + "grad_norm": 4.337297439575195, + "learning_rate": 0.00018994521276093044, + "loss": 0.7486, + "step": 6150 + }, + { + "epoch": 0.14418912662552671, + "grad_norm": 2.803800582885742, + "learning_rate": 0.00018994199336375828, + "loss": 0.4262, + "step": 6151 + }, + { + "epoch": 0.1442125682003317, + "grad_norm": 2.8591854572296143, + "learning_rate": 0.00018993877347855834, + "loss": 0.2944, + "step": 6152 + }, + { + "epoch": 0.1442360097751367, + "grad_norm": 1.033485770225525, + "learning_rate": 0.00018993555310534808, + "loss": 0.1714, + "step": 6153 + }, + { + "epoch": 0.1442594513499417, + "grad_norm": 0.7562428116798401, + "learning_rate": 0.00018993233224414495, + "loss": 0.0932, + "step": 6154 + }, + { + "epoch": 0.1442828929247467, + "grad_norm": 1.9847712516784668, + "learning_rate": 0.0001899291108949665, + "loss": 1.4917, + "step": 6155 + }, + { + "epoch": 0.1443063344995517, + "grad_norm": 1.986077904701233, + "learning_rate": 0.00018992588905783014, + "loss": 0.1787, + "step": 6156 + }, + { + "epoch": 0.14432977607435668, + "grad_norm": 3.4260222911834717, + "learning_rate": 0.00018992266673275337, + "loss": 0.3992, + "step": 6157 + }, + { + "epoch": 0.14435321764916167, + "grad_norm": 3.8250629901885986, + "learning_rate": 0.00018991944391975372, + "loss": 0.5141, + "step": 6158 + }, + { + "epoch": 0.14437665922396667, + "grad_norm": 1.2551013231277466, + "learning_rate": 0.00018991622061884862, + "loss": 0.1169, + "step": 6159 + }, + { + "epoch": 0.14440010079877166, + "grad_norm": 2.9464540481567383, + "learning_rate": 0.00018991299683005557, + "loss": 0.6238, + "step": 6160 + }, + { + "epoch": 0.14442354237357666, + "grad_norm": 2.6107237339019775, + "learning_rate": 0.0001899097725533921, + "loss": 0.1977, + "step": 6161 + }, + { + "epoch": 0.14444698394838165, + "grad_norm": 4.709012985229492, + "learning_rate": 0.00018990654778887564, + "loss": 0.6843, + "step": 6162 + }, + { + "epoch": 0.14447042552318665, + "grad_norm": 2.295891046524048, + "learning_rate": 0.00018990332253652375, + "loss": 0.952, + "step": 6163 + }, + { + "epoch": 0.14449386709799164, + "grad_norm": 1.4330990314483643, + "learning_rate": 0.00018990009679635387, + "loss": 0.1711, + "step": 6164 + }, + { + "epoch": 0.14451730867279663, + "grad_norm": 3.821744918823242, + "learning_rate": 0.00018989687056838358, + "loss": 0.4805, + "step": 6165 + }, + { + "epoch": 0.14454075024760163, + "grad_norm": 7.112173557281494, + "learning_rate": 0.00018989364385263032, + "loss": 0.7071, + "step": 6166 + }, + { + "epoch": 0.14456419182240662, + "grad_norm": 1.131948471069336, + "learning_rate": 0.0001898904166491116, + "loss": 0.117, + "step": 6167 + }, + { + "epoch": 0.14458763339721162, + "grad_norm": 4.339391708374023, + "learning_rate": 0.000189887188957845, + "loss": 0.3173, + "step": 6168 + }, + { + "epoch": 0.1446110749720166, + "grad_norm": 2.9363903999328613, + "learning_rate": 0.00018988396077884793, + "loss": 0.3678, + "step": 6169 + }, + { + "epoch": 0.1446345165468216, + "grad_norm": 5.134754180908203, + "learning_rate": 0.000189880732112138, + "loss": 0.4996, + "step": 6170 + }, + { + "epoch": 0.1446579581216266, + "grad_norm": 4.003210067749023, + "learning_rate": 0.0001898775029577327, + "loss": 0.4589, + "step": 6171 + }, + { + "epoch": 0.1446813996964316, + "grad_norm": 3.4099674224853516, + "learning_rate": 0.0001898742733156495, + "loss": 0.2135, + "step": 6172 + }, + { + "epoch": 0.1447048412712366, + "grad_norm": 5.305039405822754, + "learning_rate": 0.000189871043185906, + "loss": 0.6745, + "step": 6173 + }, + { + "epoch": 0.14472828284604158, + "grad_norm": 1.2709258794784546, + "learning_rate": 0.0001898678125685197, + "loss": 0.1042, + "step": 6174 + }, + { + "epoch": 0.1447517244208466, + "grad_norm": 3.822474718093872, + "learning_rate": 0.0001898645814635081, + "loss": 0.463, + "step": 6175 + }, + { + "epoch": 0.1447751659956516, + "grad_norm": 4.654453277587891, + "learning_rate": 0.00018986134987088873, + "loss": 0.4279, + "step": 6176 + }, + { + "epoch": 0.1447986075704566, + "grad_norm": 5.1960368156433105, + "learning_rate": 0.00018985811779067917, + "loss": 0.7448, + "step": 6177 + }, + { + "epoch": 0.1448220491452616, + "grad_norm": 4.005566120147705, + "learning_rate": 0.00018985488522289695, + "loss": 0.3367, + "step": 6178 + }, + { + "epoch": 0.14484549072006658, + "grad_norm": 1.8047767877578735, + "learning_rate": 0.0001898516521675596, + "loss": 0.2563, + "step": 6179 + }, + { + "epoch": 0.14486893229487158, + "grad_norm": 3.9863028526306152, + "learning_rate": 0.00018984841862468464, + "loss": 0.6883, + "step": 6180 + }, + { + "epoch": 0.14489237386967657, + "grad_norm": 2.4185774326324463, + "learning_rate": 0.00018984518459428962, + "loss": 1.2859, + "step": 6181 + }, + { + "epoch": 0.14491581544448157, + "grad_norm": 0.6245895028114319, + "learning_rate": 0.00018984195007639215, + "loss": 0.0831, + "step": 6182 + }, + { + "epoch": 0.14493925701928656, + "grad_norm": 3.3550760746002197, + "learning_rate": 0.0001898387150710097, + "loss": 0.4086, + "step": 6183 + }, + { + "epoch": 0.14496269859409155, + "grad_norm": 0.5554403066635132, + "learning_rate": 0.0001898354795781599, + "loss": 0.0643, + "step": 6184 + }, + { + "epoch": 0.14498614016889655, + "grad_norm": 2.385735034942627, + "learning_rate": 0.0001898322435978602, + "loss": 0.2581, + "step": 6185 + }, + { + "epoch": 0.14500958174370154, + "grad_norm": 1.1105644702911377, + "learning_rate": 0.00018982900713012825, + "loss": 0.1405, + "step": 6186 + }, + { + "epoch": 0.14503302331850654, + "grad_norm": 3.6685709953308105, + "learning_rate": 0.0001898257701749816, + "loss": 0.5691, + "step": 6187 + }, + { + "epoch": 0.14505646489331153, + "grad_norm": 1.292009949684143, + "learning_rate": 0.00018982253273243775, + "loss": 0.1238, + "step": 6188 + }, + { + "epoch": 0.14507990646811653, + "grad_norm": 2.1307811737060547, + "learning_rate": 0.00018981929480251433, + "loss": 0.2779, + "step": 6189 + }, + { + "epoch": 0.14510334804292152, + "grad_norm": 3.2182202339172363, + "learning_rate": 0.00018981605638522888, + "loss": 0.3378, + "step": 6190 + }, + { + "epoch": 0.14512678961772651, + "grad_norm": 2.65482234954834, + "learning_rate": 0.000189812817480599, + "loss": 0.2012, + "step": 6191 + }, + { + "epoch": 0.1451502311925315, + "grad_norm": 1.4383797645568848, + "learning_rate": 0.00018980957808864223, + "loss": 0.2427, + "step": 6192 + }, + { + "epoch": 0.1451736727673365, + "grad_norm": 4.357830047607422, + "learning_rate": 0.00018980633820937621, + "loss": 0.5723, + "step": 6193 + }, + { + "epoch": 0.1451971143421415, + "grad_norm": 5.959588527679443, + "learning_rate": 0.00018980309784281843, + "loss": 0.9401, + "step": 6194 + }, + { + "epoch": 0.1452205559169465, + "grad_norm": 2.06075382232666, + "learning_rate": 0.00018979985698898655, + "loss": 0.112, + "step": 6195 + }, + { + "epoch": 0.1452439974917515, + "grad_norm": 1.3602553606033325, + "learning_rate": 0.00018979661564789808, + "loss": 0.2005, + "step": 6196 + }, + { + "epoch": 0.14526743906655648, + "grad_norm": 3.454878807067871, + "learning_rate": 0.00018979337381957065, + "loss": 0.5007, + "step": 6197 + }, + { + "epoch": 0.14529088064136148, + "grad_norm": 3.7426156997680664, + "learning_rate": 0.0001897901315040219, + "loss": 1.8941, + "step": 6198 + }, + { + "epoch": 0.14531432221616647, + "grad_norm": 2.771448850631714, + "learning_rate": 0.00018978688870126932, + "loss": 0.3362, + "step": 6199 + }, + { + "epoch": 0.14533776379097146, + "grad_norm": 5.70946741104126, + "learning_rate": 0.0001897836454113306, + "loss": 0.8204, + "step": 6200 + }, + { + "epoch": 0.14536120536577649, + "grad_norm": 1.2896205186843872, + "learning_rate": 0.00018978040163422326, + "loss": 0.5102, + "step": 6201 + }, + { + "epoch": 0.14538464694058148, + "grad_norm": 2.574155330657959, + "learning_rate": 0.00018977715736996493, + "loss": 0.2881, + "step": 6202 + }, + { + "epoch": 0.14540808851538647, + "grad_norm": 1.8764623403549194, + "learning_rate": 0.00018977391261857326, + "loss": 0.2291, + "step": 6203 + }, + { + "epoch": 0.14543153009019147, + "grad_norm": 3.141251802444458, + "learning_rate": 0.00018977066738006578, + "loss": 0.297, + "step": 6204 + }, + { + "epoch": 0.14545497166499646, + "grad_norm": 2.5512502193450928, + "learning_rate": 0.00018976742165446016, + "loss": 0.2424, + "step": 6205 + }, + { + "epoch": 0.14547841323980146, + "grad_norm": 1.5879541635513306, + "learning_rate": 0.00018976417544177396, + "loss": 0.1792, + "step": 6206 + }, + { + "epoch": 0.14550185481460645, + "grad_norm": 2.3356029987335205, + "learning_rate": 0.00018976092874202485, + "loss": 1.2367, + "step": 6207 + }, + { + "epoch": 0.14552529638941145, + "grad_norm": 1.1285219192504883, + "learning_rate": 0.0001897576815552304, + "loss": 0.1474, + "step": 6208 + }, + { + "epoch": 0.14554873796421644, + "grad_norm": 1.748698115348816, + "learning_rate": 0.00018975443388140825, + "loss": 0.1805, + "step": 6209 + }, + { + "epoch": 0.14557217953902143, + "grad_norm": 4.652560234069824, + "learning_rate": 0.000189751185720576, + "loss": 0.5997, + "step": 6210 + }, + { + "epoch": 0.14559562111382643, + "grad_norm": 0.9083337783813477, + "learning_rate": 0.0001897479370727513, + "loss": 0.1445, + "step": 6211 + }, + { + "epoch": 0.14561906268863142, + "grad_norm": 3.4472296237945557, + "learning_rate": 0.0001897446879379518, + "loss": 1.3259, + "step": 6212 + }, + { + "epoch": 0.14564250426343642, + "grad_norm": 1.209312915802002, + "learning_rate": 0.00018974143831619505, + "loss": 0.1208, + "step": 6213 + }, + { + "epoch": 0.1456659458382414, + "grad_norm": 3.7928242683410645, + "learning_rate": 0.00018973818820749877, + "loss": 0.575, + "step": 6214 + }, + { + "epoch": 0.1456893874130464, + "grad_norm": 1.0545825958251953, + "learning_rate": 0.00018973493761188057, + "loss": 0.1309, + "step": 6215 + }, + { + "epoch": 0.1457128289878514, + "grad_norm": 6.309964656829834, + "learning_rate": 0.00018973168652935805, + "loss": 0.7545, + "step": 6216 + }, + { + "epoch": 0.1457362705626564, + "grad_norm": 7.85974645614624, + "learning_rate": 0.00018972843495994887, + "loss": 0.6162, + "step": 6217 + }, + { + "epoch": 0.1457597121374614, + "grad_norm": 3.8539698123931885, + "learning_rate": 0.0001897251829036707, + "loss": 0.2638, + "step": 6218 + }, + { + "epoch": 0.14578315371226638, + "grad_norm": 2.976486921310425, + "learning_rate": 0.00018972193036054114, + "loss": 0.4219, + "step": 6219 + }, + { + "epoch": 0.14580659528707138, + "grad_norm": 1.7973823547363281, + "learning_rate": 0.00018971867733057787, + "loss": 0.7315, + "step": 6220 + }, + { + "epoch": 0.14583003686187637, + "grad_norm": 4.999703407287598, + "learning_rate": 0.00018971542381379853, + "loss": 0.6704, + "step": 6221 + }, + { + "epoch": 0.14585347843668137, + "grad_norm": 2.463521718978882, + "learning_rate": 0.0001897121698102208, + "loss": 0.2408, + "step": 6222 + }, + { + "epoch": 0.14587692001148636, + "grad_norm": 3.2777552604675293, + "learning_rate": 0.0001897089153198623, + "loss": 0.4622, + "step": 6223 + }, + { + "epoch": 0.14590036158629135, + "grad_norm": 4.955059051513672, + "learning_rate": 0.00018970566034274072, + "loss": 0.6674, + "step": 6224 + }, + { + "epoch": 0.14592380316109635, + "grad_norm": 2.745553731918335, + "learning_rate": 0.00018970240487887368, + "loss": 1.5722, + "step": 6225 + }, + { + "epoch": 0.14594724473590137, + "grad_norm": 0.7390286922454834, + "learning_rate": 0.0001896991489282789, + "loss": 0.109, + "step": 6226 + }, + { + "epoch": 0.14597068631070637, + "grad_norm": 3.58891224861145, + "learning_rate": 0.00018969589249097398, + "loss": 0.4719, + "step": 6227 + }, + { + "epoch": 0.14599412788551136, + "grad_norm": 5.872560501098633, + "learning_rate": 0.00018969263556697664, + "loss": 0.7954, + "step": 6228 + }, + { + "epoch": 0.14601756946031635, + "grad_norm": 2.0625693798065186, + "learning_rate": 0.00018968937815630455, + "loss": 1.3589, + "step": 6229 + }, + { + "epoch": 0.14604101103512135, + "grad_norm": 0.7487462162971497, + "learning_rate": 0.00018968612025897533, + "loss": 0.1385, + "step": 6230 + }, + { + "epoch": 0.14606445260992634, + "grad_norm": 3.5573890209198, + "learning_rate": 0.00018968286187500674, + "loss": 0.4837, + "step": 6231 + }, + { + "epoch": 0.14608789418473134, + "grad_norm": 2.272461175918579, + "learning_rate": 0.00018967960300441642, + "loss": 0.3514, + "step": 6232 + }, + { + "epoch": 0.14611133575953633, + "grad_norm": 0.6888598799705505, + "learning_rate": 0.00018967634364722204, + "loss": 0.049, + "step": 6233 + }, + { + "epoch": 0.14613477733434133, + "grad_norm": 2.1730740070343018, + "learning_rate": 0.00018967308380344128, + "loss": 0.4486, + "step": 6234 + }, + { + "epoch": 0.14615821890914632, + "grad_norm": 3.6417174339294434, + "learning_rate": 0.00018966982347309188, + "loss": 0.3601, + "step": 6235 + }, + { + "epoch": 0.14618166048395131, + "grad_norm": 5.228623390197754, + "learning_rate": 0.0001896665626561915, + "loss": 0.8495, + "step": 6236 + }, + { + "epoch": 0.1462051020587563, + "grad_norm": 3.0185017585754395, + "learning_rate": 0.0001896633013527578, + "loss": 0.6111, + "step": 6237 + }, + { + "epoch": 0.1462285436335613, + "grad_norm": 0.43576905131340027, + "learning_rate": 0.0001896600395628085, + "loss": 0.0384, + "step": 6238 + }, + { + "epoch": 0.1462519852083663, + "grad_norm": 1.383472204208374, + "learning_rate": 0.00018965677728636132, + "loss": 0.2028, + "step": 6239 + }, + { + "epoch": 0.1462754267831713, + "grad_norm": 2.080173969268799, + "learning_rate": 0.00018965351452343397, + "loss": 0.199, + "step": 6240 + }, + { + "epoch": 0.14629886835797629, + "grad_norm": 4.841480731964111, + "learning_rate": 0.0001896502512740441, + "loss": 0.5964, + "step": 6241 + }, + { + "epoch": 0.14632230993278128, + "grad_norm": 2.417283535003662, + "learning_rate": 0.00018964698753820946, + "loss": 0.4143, + "step": 6242 + }, + { + "epoch": 0.14634575150758627, + "grad_norm": 1.399476408958435, + "learning_rate": 0.00018964372331594772, + "loss": 0.1513, + "step": 6243 + }, + { + "epoch": 0.14636919308239127, + "grad_norm": 3.2308850288391113, + "learning_rate": 0.00018964045860727665, + "loss": 0.5714, + "step": 6244 + }, + { + "epoch": 0.14639263465719626, + "grad_norm": 3.424546957015991, + "learning_rate": 0.0001896371934122139, + "loss": 0.5949, + "step": 6245 + }, + { + "epoch": 0.14641607623200126, + "grad_norm": 4.078087329864502, + "learning_rate": 0.00018963392773077727, + "loss": 0.5958, + "step": 6246 + }, + { + "epoch": 0.14643951780680625, + "grad_norm": 3.150435209274292, + "learning_rate": 0.00018963066156298437, + "loss": 0.3459, + "step": 6247 + }, + { + "epoch": 0.14646295938161125, + "grad_norm": 3.364677667617798, + "learning_rate": 0.000189627394908853, + "loss": 1.495, + "step": 6248 + }, + { + "epoch": 0.14648640095641624, + "grad_norm": 3.483063220977783, + "learning_rate": 0.00018962412776840087, + "loss": 0.5308, + "step": 6249 + }, + { + "epoch": 0.14650984253122123, + "grad_norm": 5.276522159576416, + "learning_rate": 0.0001896208601416457, + "loss": 1.0263, + "step": 6250 + }, + { + "epoch": 0.14653328410602623, + "grad_norm": 3.0681679248809814, + "learning_rate": 0.00018961759202860523, + "loss": 2.1933, + "step": 6251 + }, + { + "epoch": 0.14655672568083125, + "grad_norm": 0.599101185798645, + "learning_rate": 0.00018961432342929716, + "loss": 0.055, + "step": 6252 + }, + { + "epoch": 0.14658016725563625, + "grad_norm": 1.255797028541565, + "learning_rate": 0.00018961105434373926, + "loss": 0.1592, + "step": 6253 + }, + { + "epoch": 0.14660360883044124, + "grad_norm": 2.900144577026367, + "learning_rate": 0.0001896077847719493, + "loss": 0.4625, + "step": 6254 + }, + { + "epoch": 0.14662705040524623, + "grad_norm": 0.6329066753387451, + "learning_rate": 0.00018960451471394492, + "loss": 0.0869, + "step": 6255 + }, + { + "epoch": 0.14665049198005123, + "grad_norm": 0.6076514720916748, + "learning_rate": 0.00018960124416974398, + "loss": 0.0531, + "step": 6256 + }, + { + "epoch": 0.14667393355485622, + "grad_norm": 3.3014767169952393, + "learning_rate": 0.00018959797313936413, + "loss": 0.4986, + "step": 6257 + }, + { + "epoch": 0.14669737512966122, + "grad_norm": 3.69583797454834, + "learning_rate": 0.00018959470162282317, + "loss": 0.5314, + "step": 6258 + }, + { + "epoch": 0.1467208167044662, + "grad_norm": 3.710310220718384, + "learning_rate": 0.00018959142962013886, + "loss": 0.771, + "step": 6259 + }, + { + "epoch": 0.1467442582792712, + "grad_norm": 2.230738401412964, + "learning_rate": 0.0001895881571313289, + "loss": 1.516, + "step": 6260 + }, + { + "epoch": 0.1467676998540762, + "grad_norm": 2.3869431018829346, + "learning_rate": 0.00018958488415641108, + "loss": 0.5008, + "step": 6261 + }, + { + "epoch": 0.1467911414288812, + "grad_norm": 2.6558430194854736, + "learning_rate": 0.00018958161069540318, + "loss": 0.6348, + "step": 6262 + }, + { + "epoch": 0.1468145830036862, + "grad_norm": 4.912554740905762, + "learning_rate": 0.00018957833674832295, + "loss": 0.6169, + "step": 6263 + }, + { + "epoch": 0.14683802457849118, + "grad_norm": 3.6883139610290527, + "learning_rate": 0.00018957506231518813, + "loss": 0.4847, + "step": 6264 + }, + { + "epoch": 0.14686146615329618, + "grad_norm": 2.6423232555389404, + "learning_rate": 0.0001895717873960165, + "loss": 0.4331, + "step": 6265 + }, + { + "epoch": 0.14688490772810117, + "grad_norm": 3.7152247428894043, + "learning_rate": 0.00018956851199082583, + "loss": 1.2712, + "step": 6266 + }, + { + "epoch": 0.14690834930290617, + "grad_norm": 2.6864404678344727, + "learning_rate": 0.00018956523609963392, + "loss": 0.3425, + "step": 6267 + }, + { + "epoch": 0.14693179087771116, + "grad_norm": 2.2936205863952637, + "learning_rate": 0.0001895619597224585, + "loss": 0.2503, + "step": 6268 + }, + { + "epoch": 0.14695523245251615, + "grad_norm": 3.178586483001709, + "learning_rate": 0.0001895586828593174, + "loss": 0.3926, + "step": 6269 + }, + { + "epoch": 0.14697867402732115, + "grad_norm": 2.7563884258270264, + "learning_rate": 0.0001895554055102283, + "loss": 0.3927, + "step": 6270 + }, + { + "epoch": 0.14700211560212614, + "grad_norm": 2.990784168243408, + "learning_rate": 0.0001895521276752091, + "loss": 0.5326, + "step": 6271 + }, + { + "epoch": 0.14702555717693114, + "grad_norm": 3.801961660385132, + "learning_rate": 0.00018954884935427755, + "loss": 0.3058, + "step": 6272 + }, + { + "epoch": 0.14704899875173613, + "grad_norm": 1.151078701019287, + "learning_rate": 0.0001895455705474514, + "loss": 0.1115, + "step": 6273 + }, + { + "epoch": 0.14707244032654113, + "grad_norm": 2.5659077167510986, + "learning_rate": 0.00018954229125474846, + "loss": 0.2463, + "step": 6274 + }, + { + "epoch": 0.14709588190134612, + "grad_norm": 4.506509304046631, + "learning_rate": 0.00018953901147618655, + "loss": 0.4546, + "step": 6275 + }, + { + "epoch": 0.14711932347615111, + "grad_norm": 3.79587459564209, + "learning_rate": 0.00018953573121178346, + "loss": 0.3916, + "step": 6276 + }, + { + "epoch": 0.14714276505095614, + "grad_norm": 2.561408042907715, + "learning_rate": 0.00018953245046155693, + "loss": 1.5174, + "step": 6277 + }, + { + "epoch": 0.14716620662576113, + "grad_norm": 3.325272560119629, + "learning_rate": 0.00018952916922552484, + "loss": 0.3756, + "step": 6278 + }, + { + "epoch": 0.14718964820056613, + "grad_norm": 1.86470365524292, + "learning_rate": 0.00018952588750370495, + "loss": 1.3279, + "step": 6279 + }, + { + "epoch": 0.14721308977537112, + "grad_norm": 3.475175380706787, + "learning_rate": 0.00018952260529611506, + "loss": 0.4597, + "step": 6280 + }, + { + "epoch": 0.14723653135017611, + "grad_norm": 5.24402379989624, + "learning_rate": 0.00018951932260277303, + "loss": 0.8501, + "step": 6281 + }, + { + "epoch": 0.1472599729249811, + "grad_norm": 2.3837709426879883, + "learning_rate": 0.00018951603942369663, + "loss": 0.1726, + "step": 6282 + }, + { + "epoch": 0.1472834144997861, + "grad_norm": 4.204874515533447, + "learning_rate": 0.00018951275575890367, + "loss": 0.5736, + "step": 6283 + }, + { + "epoch": 0.1473068560745911, + "grad_norm": 3.2695932388305664, + "learning_rate": 0.000189509471608412, + "loss": 1.2637, + "step": 6284 + }, + { + "epoch": 0.1473302976493961, + "grad_norm": 1.691390037536621, + "learning_rate": 0.00018950618697223938, + "loss": 1.1076, + "step": 6285 + }, + { + "epoch": 0.14735373922420109, + "grad_norm": 4.148710250854492, + "learning_rate": 0.0001895029018504037, + "loss": 0.6885, + "step": 6286 + }, + { + "epoch": 0.14737718079900608, + "grad_norm": 4.690939903259277, + "learning_rate": 0.00018949961624292278, + "loss": 0.599, + "step": 6287 + }, + { + "epoch": 0.14740062237381107, + "grad_norm": 0.8353723287582397, + "learning_rate": 0.0001894963301498144, + "loss": 0.0757, + "step": 6288 + }, + { + "epoch": 0.14742406394861607, + "grad_norm": 2.385003089904785, + "learning_rate": 0.00018949304357109643, + "loss": 1.1515, + "step": 6289 + }, + { + "epoch": 0.14744750552342106, + "grad_norm": 4.923844337463379, + "learning_rate": 0.00018948975650678667, + "loss": 0.6112, + "step": 6290 + }, + { + "epoch": 0.14747094709822606, + "grad_norm": 1.5846205949783325, + "learning_rate": 0.00018948646895690295, + "loss": 0.9857, + "step": 6291 + }, + { + "epoch": 0.14749438867303105, + "grad_norm": 4.460725784301758, + "learning_rate": 0.0001894831809214632, + "loss": 0.4767, + "step": 6292 + }, + { + "epoch": 0.14751783024783605, + "grad_norm": 3.5825047492980957, + "learning_rate": 0.00018947989240048515, + "loss": 0.4859, + "step": 6293 + }, + { + "epoch": 0.14754127182264104, + "grad_norm": 3.8814377784729004, + "learning_rate": 0.0001894766033939867, + "loss": 0.4908, + "step": 6294 + }, + { + "epoch": 0.14756471339744603, + "grad_norm": 2.7566263675689697, + "learning_rate": 0.0001894733139019857, + "loss": 1.456, + "step": 6295 + }, + { + "epoch": 0.14758815497225103, + "grad_norm": 1.6598891019821167, + "learning_rate": 0.00018947002392449995, + "loss": 0.1739, + "step": 6296 + }, + { + "epoch": 0.14761159654705602, + "grad_norm": 0.8084732294082642, + "learning_rate": 0.00018946673346154736, + "loss": 0.0576, + "step": 6297 + }, + { + "epoch": 0.14763503812186102, + "grad_norm": 1.0338473320007324, + "learning_rate": 0.00018946344251314577, + "loss": 0.1301, + "step": 6298 + }, + { + "epoch": 0.147658479696666, + "grad_norm": 3.202601909637451, + "learning_rate": 0.00018946015107931298, + "loss": 0.588, + "step": 6299 + }, + { + "epoch": 0.147681921271471, + "grad_norm": 0.5732741951942444, + "learning_rate": 0.0001894568591600669, + "loss": 0.0553, + "step": 6300 + }, + { + "epoch": 0.147705362846276, + "grad_norm": 3.4298715591430664, + "learning_rate": 0.00018945356675542543, + "loss": 0.5739, + "step": 6301 + }, + { + "epoch": 0.147728804421081, + "grad_norm": 2.3694446086883545, + "learning_rate": 0.00018945027386540637, + "loss": 0.2912, + "step": 6302 + }, + { + "epoch": 0.14775224599588602, + "grad_norm": 2.753002882003784, + "learning_rate": 0.0001894469804900276, + "loss": 1.7624, + "step": 6303 + }, + { + "epoch": 0.147775687570691, + "grad_norm": 1.4653515815734863, + "learning_rate": 0.000189443686629307, + "loss": 0.1822, + "step": 6304 + }, + { + "epoch": 0.147799129145496, + "grad_norm": 2.877586603164673, + "learning_rate": 0.00018944039228326246, + "loss": 1.6285, + "step": 6305 + }, + { + "epoch": 0.147822570720301, + "grad_norm": 1.4771103858947754, + "learning_rate": 0.00018943709745191184, + "loss": 0.1913, + "step": 6306 + }, + { + "epoch": 0.147846012295106, + "grad_norm": 3.4064133167266846, + "learning_rate": 0.00018943380213527297, + "loss": 0.5072, + "step": 6307 + }, + { + "epoch": 0.147869453869911, + "grad_norm": 3.1606311798095703, + "learning_rate": 0.0001894305063333638, + "loss": 1.717, + "step": 6308 + }, + { + "epoch": 0.14789289544471598, + "grad_norm": 0.5533521771430969, + "learning_rate": 0.0001894272100462022, + "loss": 0.042, + "step": 6309 + }, + { + "epoch": 0.14791633701952098, + "grad_norm": 3.8673171997070312, + "learning_rate": 0.00018942391327380603, + "loss": 0.5938, + "step": 6310 + }, + { + "epoch": 0.14793977859432597, + "grad_norm": 3.338597059249878, + "learning_rate": 0.00018942061601619317, + "loss": 0.3464, + "step": 6311 + }, + { + "epoch": 0.14796322016913097, + "grad_norm": 0.7684953808784485, + "learning_rate": 0.0001894173182733816, + "loss": 0.1092, + "step": 6312 + }, + { + "epoch": 0.14798666174393596, + "grad_norm": 2.9394543170928955, + "learning_rate": 0.0001894140200453891, + "loss": 0.3838, + "step": 6313 + }, + { + "epoch": 0.14801010331874095, + "grad_norm": 0.7511798143386841, + "learning_rate": 0.00018941072133223363, + "loss": 0.0902, + "step": 6314 + }, + { + "epoch": 0.14803354489354595, + "grad_norm": 3.5860981941223145, + "learning_rate": 0.00018940742213393306, + "loss": 0.3626, + "step": 6315 + }, + { + "epoch": 0.14805698646835094, + "grad_norm": 2.9110231399536133, + "learning_rate": 0.0001894041224505053, + "loss": 0.4443, + "step": 6316 + }, + { + "epoch": 0.14808042804315594, + "grad_norm": 3.4347546100616455, + "learning_rate": 0.00018940082228196826, + "loss": 0.4643, + "step": 6317 + }, + { + "epoch": 0.14810386961796093, + "grad_norm": 2.4171345233917236, + "learning_rate": 0.00018939752162833986, + "loss": 1.7557, + "step": 6318 + }, + { + "epoch": 0.14812731119276593, + "grad_norm": 4.703575611114502, + "learning_rate": 0.00018939422048963798, + "loss": 0.5705, + "step": 6319 + }, + { + "epoch": 0.14815075276757092, + "grad_norm": 1.657586932182312, + "learning_rate": 0.00018939091886588055, + "loss": 0.1108, + "step": 6320 + }, + { + "epoch": 0.14817419434237591, + "grad_norm": 4.285704612731934, + "learning_rate": 0.00018938761675708546, + "loss": 0.4729, + "step": 6321 + }, + { + "epoch": 0.1481976359171809, + "grad_norm": 3.1436567306518555, + "learning_rate": 0.0001893843141632707, + "loss": 0.4115, + "step": 6322 + }, + { + "epoch": 0.1482210774919859, + "grad_norm": 4.610388278961182, + "learning_rate": 0.0001893810110844541, + "loss": 0.488, + "step": 6323 + }, + { + "epoch": 0.1482445190667909, + "grad_norm": 3.567758321762085, + "learning_rate": 0.00018937770752065362, + "loss": 0.5779, + "step": 6324 + }, + { + "epoch": 0.1482679606415959, + "grad_norm": 2.9093809127807617, + "learning_rate": 0.0001893744034718872, + "loss": 0.358, + "step": 6325 + }, + { + "epoch": 0.14829140221640089, + "grad_norm": 3.33756422996521, + "learning_rate": 0.00018937109893817273, + "loss": 0.4565, + "step": 6326 + }, + { + "epoch": 0.14831484379120588, + "grad_norm": 3.658787488937378, + "learning_rate": 0.0001893677939195282, + "loss": 0.4926, + "step": 6327 + }, + { + "epoch": 0.1483382853660109, + "grad_norm": 3.3993723392486572, + "learning_rate": 0.00018936448841597146, + "loss": 0.546, + "step": 6328 + }, + { + "epoch": 0.1483617269408159, + "grad_norm": 1.8822218179702759, + "learning_rate": 0.00018936118242752055, + "loss": 0.3035, + "step": 6329 + }, + { + "epoch": 0.1483851685156209, + "grad_norm": 2.270955801010132, + "learning_rate": 0.0001893578759541933, + "loss": 0.3111, + "step": 6330 + }, + { + "epoch": 0.14840861009042589, + "grad_norm": 0.8852485418319702, + "learning_rate": 0.00018935456899600773, + "loss": 0.0649, + "step": 6331 + }, + { + "epoch": 0.14843205166523088, + "grad_norm": 2.6494340896606445, + "learning_rate": 0.00018935126155298176, + "loss": 0.528, + "step": 6332 + }, + { + "epoch": 0.14845549324003587, + "grad_norm": 2.4970178604125977, + "learning_rate": 0.00018934795362513332, + "loss": 0.2835, + "step": 6333 + }, + { + "epoch": 0.14847893481484087, + "grad_norm": 3.2204573154449463, + "learning_rate": 0.00018934464521248038, + "loss": 0.3914, + "step": 6334 + }, + { + "epoch": 0.14850237638964586, + "grad_norm": 3.33884859085083, + "learning_rate": 0.00018934133631504088, + "loss": 0.467, + "step": 6335 + }, + { + "epoch": 0.14852581796445086, + "grad_norm": 2.172884941101074, + "learning_rate": 0.00018933802693283275, + "loss": 1.4859, + "step": 6336 + }, + { + "epoch": 0.14854925953925585, + "grad_norm": 1.1216787099838257, + "learning_rate": 0.000189334717065874, + "loss": 0.2001, + "step": 6337 + }, + { + "epoch": 0.14857270111406085, + "grad_norm": 5.102049827575684, + "learning_rate": 0.00018933140671418256, + "loss": 0.5056, + "step": 6338 + }, + { + "epoch": 0.14859614268886584, + "grad_norm": 0.4071093201637268, + "learning_rate": 0.00018932809587777642, + "loss": 0.0288, + "step": 6339 + }, + { + "epoch": 0.14861958426367083, + "grad_norm": 5.075652599334717, + "learning_rate": 0.00018932478455667347, + "loss": 0.8305, + "step": 6340 + }, + { + "epoch": 0.14864302583847583, + "grad_norm": 2.43643856048584, + "learning_rate": 0.00018932147275089176, + "loss": 0.1388, + "step": 6341 + }, + { + "epoch": 0.14866646741328082, + "grad_norm": 2.994877815246582, + "learning_rate": 0.00018931816046044924, + "loss": 0.3821, + "step": 6342 + }, + { + "epoch": 0.14868990898808582, + "grad_norm": 4.558137893676758, + "learning_rate": 0.00018931484768536385, + "loss": 0.7397, + "step": 6343 + }, + { + "epoch": 0.1487133505628908, + "grad_norm": 1.9389086961746216, + "learning_rate": 0.00018931153442565356, + "loss": 0.4227, + "step": 6344 + }, + { + "epoch": 0.1487367921376958, + "grad_norm": 4.285367012023926, + "learning_rate": 0.00018930822068133642, + "loss": 0.6048, + "step": 6345 + }, + { + "epoch": 0.1487602337125008, + "grad_norm": 3.686720848083496, + "learning_rate": 0.00018930490645243034, + "loss": 0.4858, + "step": 6346 + }, + { + "epoch": 0.1487836752873058, + "grad_norm": 2.4032623767852783, + "learning_rate": 0.00018930159173895334, + "loss": 0.406, + "step": 6347 + }, + { + "epoch": 0.1488071168621108, + "grad_norm": 2.1513168811798096, + "learning_rate": 0.00018929827654092337, + "loss": 0.9083, + "step": 6348 + }, + { + "epoch": 0.14883055843691578, + "grad_norm": 3.5344269275665283, + "learning_rate": 0.00018929496085835846, + "loss": 0.66, + "step": 6349 + }, + { + "epoch": 0.14885400001172078, + "grad_norm": 1.3238799571990967, + "learning_rate": 0.0001892916446912766, + "loss": 0.1511, + "step": 6350 + }, + { + "epoch": 0.14887744158652577, + "grad_norm": 3.305725336074829, + "learning_rate": 0.00018928832803969572, + "loss": 0.3159, + "step": 6351 + }, + { + "epoch": 0.14890088316133077, + "grad_norm": 2.1096608638763428, + "learning_rate": 0.00018928501090363387, + "loss": 0.1468, + "step": 6352 + }, + { + "epoch": 0.14892432473613576, + "grad_norm": 3.1512153148651123, + "learning_rate": 0.00018928169328310905, + "loss": 0.4622, + "step": 6353 + }, + { + "epoch": 0.14894776631094078, + "grad_norm": 4.0651936531066895, + "learning_rate": 0.00018927837517813926, + "loss": 0.3865, + "step": 6354 + }, + { + "epoch": 0.14897120788574578, + "grad_norm": 3.3631842136383057, + "learning_rate": 0.0001892750565887425, + "loss": 0.3508, + "step": 6355 + }, + { + "epoch": 0.14899464946055077, + "grad_norm": 4.227780342102051, + "learning_rate": 0.00018927173751493676, + "loss": 1.9495, + "step": 6356 + }, + { + "epoch": 0.14901809103535577, + "grad_norm": 5.516700267791748, + "learning_rate": 0.00018926841795674005, + "loss": 0.6147, + "step": 6357 + }, + { + "epoch": 0.14904153261016076, + "grad_norm": 3.7404110431671143, + "learning_rate": 0.0001892650979141704, + "loss": 0.3085, + "step": 6358 + }, + { + "epoch": 0.14906497418496575, + "grad_norm": 3.2962193489074707, + "learning_rate": 0.00018926177738724586, + "loss": 0.4574, + "step": 6359 + }, + { + "epoch": 0.14908841575977075, + "grad_norm": 0.6538308262825012, + "learning_rate": 0.00018925845637598435, + "loss": 0.0569, + "step": 6360 + }, + { + "epoch": 0.14911185733457574, + "grad_norm": 4.036275863647461, + "learning_rate": 0.00018925513488040398, + "loss": 0.6334, + "step": 6361 + }, + { + "epoch": 0.14913529890938074, + "grad_norm": 0.4269184470176697, + "learning_rate": 0.00018925181290052273, + "loss": 0.0294, + "step": 6362 + }, + { + "epoch": 0.14915874048418573, + "grad_norm": 0.7161809206008911, + "learning_rate": 0.00018924849043635864, + "loss": 0.1148, + "step": 6363 + }, + { + "epoch": 0.14918218205899073, + "grad_norm": 5.4686713218688965, + "learning_rate": 0.00018924516748792974, + "loss": 0.9079, + "step": 6364 + }, + { + "epoch": 0.14920562363379572, + "grad_norm": 2.835397481918335, + "learning_rate": 0.000189241844055254, + "loss": 0.3183, + "step": 6365 + }, + { + "epoch": 0.14922906520860071, + "grad_norm": 4.6673994064331055, + "learning_rate": 0.00018923852013834954, + "loss": 0.3652, + "step": 6366 + }, + { + "epoch": 0.1492525067834057, + "grad_norm": 4.706110000610352, + "learning_rate": 0.0001892351957372344, + "loss": 0.7477, + "step": 6367 + }, + { + "epoch": 0.1492759483582107, + "grad_norm": 1.40652334690094, + "learning_rate": 0.0001892318708519265, + "loss": 0.221, + "step": 6368 + }, + { + "epoch": 0.1492993899330157, + "grad_norm": 2.4052045345306396, + "learning_rate": 0.00018922854548244402, + "loss": 0.2373, + "step": 6369 + }, + { + "epoch": 0.1493228315078207, + "grad_norm": 4.181654930114746, + "learning_rate": 0.0001892252196288049, + "loss": 0.4962, + "step": 6370 + }, + { + "epoch": 0.14934627308262569, + "grad_norm": 1.5578770637512207, + "learning_rate": 0.00018922189329102727, + "loss": 0.2202, + "step": 6371 + }, + { + "epoch": 0.14936971465743068, + "grad_norm": 1.468347191810608, + "learning_rate": 0.00018921856646912912, + "loss": 0.2054, + "step": 6372 + }, + { + "epoch": 0.14939315623223567, + "grad_norm": 3.2305779457092285, + "learning_rate": 0.00018921523916312852, + "loss": 1.7062, + "step": 6373 + }, + { + "epoch": 0.14941659780704067, + "grad_norm": 2.9775471687316895, + "learning_rate": 0.00018921191137304352, + "loss": 0.5086, + "step": 6374 + }, + { + "epoch": 0.14944003938184566, + "grad_norm": 3.1282193660736084, + "learning_rate": 0.0001892085830988922, + "loss": 0.2912, + "step": 6375 + }, + { + "epoch": 0.14946348095665066, + "grad_norm": 0.7049998641014099, + "learning_rate": 0.0001892052543406926, + "loss": 0.0695, + "step": 6376 + }, + { + "epoch": 0.14948692253145565, + "grad_norm": 5.141985893249512, + "learning_rate": 0.00018920192509846274, + "loss": 0.6333, + "step": 6377 + }, + { + "epoch": 0.14951036410626065, + "grad_norm": 4.346166133880615, + "learning_rate": 0.00018919859537222075, + "loss": 0.611, + "step": 6378 + }, + { + "epoch": 0.14953380568106564, + "grad_norm": 1.835456371307373, + "learning_rate": 0.0001891952651619847, + "loss": 0.3235, + "step": 6379 + }, + { + "epoch": 0.14955724725587066, + "grad_norm": 3.9049391746520996, + "learning_rate": 0.0001891919344677726, + "loss": 0.6976, + "step": 6380 + }, + { + "epoch": 0.14958068883067566, + "grad_norm": 1.499922752380371, + "learning_rate": 0.00018918860328960254, + "loss": 0.1398, + "step": 6381 + }, + { + "epoch": 0.14960413040548065, + "grad_norm": 4.286501407623291, + "learning_rate": 0.00018918527162749266, + "loss": 0.4569, + "step": 6382 + }, + { + "epoch": 0.14962757198028565, + "grad_norm": 4.320180416107178, + "learning_rate": 0.00018918193948146093, + "loss": 0.585, + "step": 6383 + }, + { + "epoch": 0.14965101355509064, + "grad_norm": 2.9158012866973877, + "learning_rate": 0.00018917860685152554, + "loss": 0.4443, + "step": 6384 + }, + { + "epoch": 0.14967445512989563, + "grad_norm": 4.849606037139893, + "learning_rate": 0.00018917527373770452, + "loss": 0.7136, + "step": 6385 + }, + { + "epoch": 0.14969789670470063, + "grad_norm": 3.4454526901245117, + "learning_rate": 0.00018917194014001594, + "loss": 0.3346, + "step": 6386 + }, + { + "epoch": 0.14972133827950562, + "grad_norm": 3.7650177478790283, + "learning_rate": 0.0001891686060584779, + "loss": 0.4065, + "step": 6387 + }, + { + "epoch": 0.14974477985431062, + "grad_norm": 3.3685643672943115, + "learning_rate": 0.00018916527149310852, + "loss": 0.52, + "step": 6388 + }, + { + "epoch": 0.1497682214291156, + "grad_norm": 2.5150771141052246, + "learning_rate": 0.00018916193644392585, + "loss": 0.3971, + "step": 6389 + }, + { + "epoch": 0.1497916630039206, + "grad_norm": 4.402888774871826, + "learning_rate": 0.000189158600910948, + "loss": 0.6768, + "step": 6390 + }, + { + "epoch": 0.1498151045787256, + "grad_norm": 2.388983964920044, + "learning_rate": 0.00018915526489419307, + "loss": 0.3179, + "step": 6391 + }, + { + "epoch": 0.1498385461535306, + "grad_norm": 2.5521795749664307, + "learning_rate": 0.0001891519283936792, + "loss": 0.3171, + "step": 6392 + }, + { + "epoch": 0.1498619877283356, + "grad_norm": 2.861541748046875, + "learning_rate": 0.00018914859140942443, + "loss": 0.3376, + "step": 6393 + }, + { + "epoch": 0.14988542930314058, + "grad_norm": 4.30867338180542, + "learning_rate": 0.0001891452539414469, + "loss": 0.5162, + "step": 6394 + }, + { + "epoch": 0.14990887087794558, + "grad_norm": 1.8002935647964478, + "learning_rate": 0.0001891419159897647, + "loss": 1.5151, + "step": 6395 + }, + { + "epoch": 0.14993231245275057, + "grad_norm": 2.9855403900146484, + "learning_rate": 0.00018913857755439598, + "loss": 0.3997, + "step": 6396 + }, + { + "epoch": 0.14995575402755557, + "grad_norm": 6.473509788513184, + "learning_rate": 0.00018913523863535887, + "loss": 0.5774, + "step": 6397 + }, + { + "epoch": 0.14997919560236056, + "grad_norm": 4.2177534103393555, + "learning_rate": 0.0001891318992326714, + "loss": 0.742, + "step": 6398 + }, + { + "epoch": 0.15000263717716555, + "grad_norm": 4.464262962341309, + "learning_rate": 0.00018912855934635173, + "loss": 0.5281, + "step": 6399 + }, + { + "epoch": 0.15002607875197055, + "grad_norm": 2.855088710784912, + "learning_rate": 0.00018912521897641803, + "loss": 0.4057, + "step": 6400 + }, + { + "epoch": 0.15004952032677554, + "grad_norm": 4.3128981590271, + "learning_rate": 0.00018912187812288836, + "loss": 0.3768, + "step": 6401 + }, + { + "epoch": 0.15007296190158054, + "grad_norm": 3.686772346496582, + "learning_rate": 0.0001891185367857809, + "loss": 0.6586, + "step": 6402 + }, + { + "epoch": 0.15009640347638553, + "grad_norm": 2.381586790084839, + "learning_rate": 0.00018911519496511372, + "loss": 0.1703, + "step": 6403 + }, + { + "epoch": 0.15011984505119053, + "grad_norm": 4.226410388946533, + "learning_rate": 0.000189111852660905, + "loss": 0.6395, + "step": 6404 + }, + { + "epoch": 0.15014328662599555, + "grad_norm": 4.137767314910889, + "learning_rate": 0.00018910850987317285, + "loss": 0.6039, + "step": 6405 + }, + { + "epoch": 0.15016672820080054, + "grad_norm": 3.347668409347534, + "learning_rate": 0.00018910516660193544, + "loss": 0.3535, + "step": 6406 + }, + { + "epoch": 0.15019016977560554, + "grad_norm": 3.379310369491577, + "learning_rate": 0.00018910182284721086, + "loss": 0.4233, + "step": 6407 + }, + { + "epoch": 0.15021361135041053, + "grad_norm": 0.9364840984344482, + "learning_rate": 0.0001890984786090173, + "loss": 0.1011, + "step": 6408 + }, + { + "epoch": 0.15023705292521553, + "grad_norm": 2.5541555881500244, + "learning_rate": 0.00018909513388737289, + "loss": 0.4776, + "step": 6409 + }, + { + "epoch": 0.15026049450002052, + "grad_norm": 4.563040733337402, + "learning_rate": 0.0001890917886822958, + "loss": 0.7309, + "step": 6410 + }, + { + "epoch": 0.1502839360748255, + "grad_norm": 1.305227279663086, + "learning_rate": 0.00018908844299380415, + "loss": 0.1387, + "step": 6411 + }, + { + "epoch": 0.1503073776496305, + "grad_norm": 3.2806448936462402, + "learning_rate": 0.0001890850968219161, + "loss": 0.4704, + "step": 6412 + }, + { + "epoch": 0.1503308192244355, + "grad_norm": 0.5935465097427368, + "learning_rate": 0.0001890817501666498, + "loss": 0.0774, + "step": 6413 + }, + { + "epoch": 0.1503542607992405, + "grad_norm": 0.847435712814331, + "learning_rate": 0.0001890784030280234, + "loss": 0.0788, + "step": 6414 + }, + { + "epoch": 0.1503777023740455, + "grad_norm": 4.995213031768799, + "learning_rate": 0.00018907505540605513, + "loss": 0.6171, + "step": 6415 + }, + { + "epoch": 0.15040114394885049, + "grad_norm": 3.463813304901123, + "learning_rate": 0.00018907170730076306, + "loss": 1.3956, + "step": 6416 + }, + { + "epoch": 0.15042458552365548, + "grad_norm": 4.286025047302246, + "learning_rate": 0.00018906835871216544, + "loss": 0.4011, + "step": 6417 + }, + { + "epoch": 0.15044802709846047, + "grad_norm": 2.007506847381592, + "learning_rate": 0.0001890650096402804, + "loss": 0.1361, + "step": 6418 + }, + { + "epoch": 0.15047146867326547, + "grad_norm": 2.860818862915039, + "learning_rate": 0.0001890616600851261, + "loss": 0.3832, + "step": 6419 + }, + { + "epoch": 0.15049491024807046, + "grad_norm": 0.8277170658111572, + "learning_rate": 0.0001890583100467207, + "loss": 0.1096, + "step": 6420 + }, + { + "epoch": 0.15051835182287546, + "grad_norm": 6.284062385559082, + "learning_rate": 0.00018905495952508243, + "loss": 0.6181, + "step": 6421 + }, + { + "epoch": 0.15054179339768045, + "grad_norm": 3.6155855655670166, + "learning_rate": 0.00018905160852022946, + "loss": 0.4224, + "step": 6422 + }, + { + "epoch": 0.15056523497248545, + "grad_norm": 3.298292875289917, + "learning_rate": 0.00018904825703217992, + "loss": 0.6586, + "step": 6423 + }, + { + "epoch": 0.15058867654729044, + "grad_norm": 2.7884743213653564, + "learning_rate": 0.00018904490506095207, + "loss": 0.3587, + "step": 6424 + }, + { + "epoch": 0.15061211812209543, + "grad_norm": 5.162617206573486, + "learning_rate": 0.00018904155260656403, + "loss": 0.7438, + "step": 6425 + }, + { + "epoch": 0.15063555969690043, + "grad_norm": 2.3049933910369873, + "learning_rate": 0.00018903819966903406, + "loss": 1.5121, + "step": 6426 + }, + { + "epoch": 0.15065900127170542, + "grad_norm": 0.7361961603164673, + "learning_rate": 0.00018903484624838027, + "loss": 0.0578, + "step": 6427 + }, + { + "epoch": 0.15068244284651042, + "grad_norm": 2.607004404067993, + "learning_rate": 0.00018903149234462095, + "loss": 0.4791, + "step": 6428 + }, + { + "epoch": 0.1507058844213154, + "grad_norm": 6.581178665161133, + "learning_rate": 0.00018902813795777424, + "loss": 0.9373, + "step": 6429 + }, + { + "epoch": 0.1507293259961204, + "grad_norm": 1.6051608324050903, + "learning_rate": 0.00018902478308785835, + "loss": 0.2287, + "step": 6430 + }, + { + "epoch": 0.15075276757092543, + "grad_norm": 3.35074782371521, + "learning_rate": 0.00018902142773489144, + "loss": 0.4637, + "step": 6431 + }, + { + "epoch": 0.15077620914573042, + "grad_norm": 4.29120397567749, + "learning_rate": 0.0001890180718988918, + "loss": 0.7805, + "step": 6432 + }, + { + "epoch": 0.15079965072053542, + "grad_norm": 0.8038742542266846, + "learning_rate": 0.0001890147155798776, + "loss": 0.108, + "step": 6433 + }, + { + "epoch": 0.1508230922953404, + "grad_norm": 3.2010440826416016, + "learning_rate": 0.00018901135877786703, + "loss": 0.3557, + "step": 6434 + }, + { + "epoch": 0.1508465338701454, + "grad_norm": 2.6482818126678467, + "learning_rate": 0.00018900800149287833, + "loss": 0.312, + "step": 6435 + }, + { + "epoch": 0.1508699754449504, + "grad_norm": 4.284143447875977, + "learning_rate": 0.00018900464372492972, + "loss": 0.6529, + "step": 6436 + }, + { + "epoch": 0.1508934170197554, + "grad_norm": 2.720712900161743, + "learning_rate": 0.0001890012854740394, + "loss": 0.4891, + "step": 6437 + }, + { + "epoch": 0.1509168585945604, + "grad_norm": 2.4389255046844482, + "learning_rate": 0.0001889979267402256, + "loss": 0.6359, + "step": 6438 + }, + { + "epoch": 0.15094030016936538, + "grad_norm": 0.9698941707611084, + "learning_rate": 0.00018899456752350657, + "loss": 0.1275, + "step": 6439 + }, + { + "epoch": 0.15096374174417038, + "grad_norm": 0.853050708770752, + "learning_rate": 0.00018899120782390048, + "loss": 0.093, + "step": 6440 + }, + { + "epoch": 0.15098718331897537, + "grad_norm": 2.7237160205841064, + "learning_rate": 0.0001889878476414256, + "loss": 0.4249, + "step": 6441 + }, + { + "epoch": 0.15101062489378037, + "grad_norm": 0.8735179305076599, + "learning_rate": 0.0001889844869761002, + "loss": 0.0867, + "step": 6442 + }, + { + "epoch": 0.15103406646858536, + "grad_norm": 3.769896984100342, + "learning_rate": 0.00018898112582794242, + "loss": 0.5992, + "step": 6443 + }, + { + "epoch": 0.15105750804339035, + "grad_norm": 2.502342939376831, + "learning_rate": 0.0001889777641969706, + "loss": 0.2855, + "step": 6444 + }, + { + "epoch": 0.15108094961819535, + "grad_norm": 4.54252815246582, + "learning_rate": 0.0001889744020832029, + "loss": 0.4963, + "step": 6445 + }, + { + "epoch": 0.15110439119300034, + "grad_norm": 2.651399850845337, + "learning_rate": 0.00018897103948665759, + "loss": 2.0291, + "step": 6446 + }, + { + "epoch": 0.15112783276780534, + "grad_norm": 4.228682041168213, + "learning_rate": 0.0001889676764073529, + "loss": 0.3914, + "step": 6447 + }, + { + "epoch": 0.15115127434261033, + "grad_norm": 1.0216671228408813, + "learning_rate": 0.00018896431284530713, + "loss": 0.1343, + "step": 6448 + }, + { + "epoch": 0.15117471591741533, + "grad_norm": 1.9670131206512451, + "learning_rate": 0.00018896094880053847, + "loss": 0.2596, + "step": 6449 + }, + { + "epoch": 0.15119815749222032, + "grad_norm": 3.1633083820343018, + "learning_rate": 0.00018895758427306522, + "loss": 0.4683, + "step": 6450 + }, + { + "epoch": 0.15122159906702531, + "grad_norm": 3.024763345718384, + "learning_rate": 0.00018895421926290562, + "loss": 0.3367, + "step": 6451 + }, + { + "epoch": 0.1512450406418303, + "grad_norm": 2.718050241470337, + "learning_rate": 0.00018895085377007792, + "loss": 0.3023, + "step": 6452 + }, + { + "epoch": 0.1512684822166353, + "grad_norm": 4.2214460372924805, + "learning_rate": 0.00018894748779460039, + "loss": 0.4621, + "step": 6453 + }, + { + "epoch": 0.1512919237914403, + "grad_norm": 1.903099775314331, + "learning_rate": 0.00018894412133649125, + "loss": 0.2119, + "step": 6454 + }, + { + "epoch": 0.1513153653662453, + "grad_norm": 4.005669593811035, + "learning_rate": 0.00018894075439576885, + "loss": 0.7045, + "step": 6455 + }, + { + "epoch": 0.1513388069410503, + "grad_norm": 1.899996280670166, + "learning_rate": 0.00018893738697245142, + "loss": 0.3112, + "step": 6456 + }, + { + "epoch": 0.1513622485158553, + "grad_norm": 2.4842028617858887, + "learning_rate": 0.0001889340190665572, + "loss": 1.9671, + "step": 6457 + }, + { + "epoch": 0.1513856900906603, + "grad_norm": 2.8524844646453857, + "learning_rate": 0.00018893065067810452, + "loss": 0.4442, + "step": 6458 + }, + { + "epoch": 0.1514091316654653, + "grad_norm": 5.581087589263916, + "learning_rate": 0.0001889272818071116, + "loss": 1.3507, + "step": 6459 + }, + { + "epoch": 0.1514325732402703, + "grad_norm": 4.007874965667725, + "learning_rate": 0.00018892391245359678, + "loss": 0.3911, + "step": 6460 + }, + { + "epoch": 0.15145601481507528, + "grad_norm": 2.933826208114624, + "learning_rate": 0.00018892054261757828, + "loss": 0.1942, + "step": 6461 + }, + { + "epoch": 0.15147945638988028, + "grad_norm": 4.019176959991455, + "learning_rate": 0.00018891717229907443, + "loss": 0.5879, + "step": 6462 + }, + { + "epoch": 0.15150289796468527, + "grad_norm": 3.9696834087371826, + "learning_rate": 0.0001889138014981035, + "loss": 0.5978, + "step": 6463 + }, + { + "epoch": 0.15152633953949027, + "grad_norm": 5.1955976486206055, + "learning_rate": 0.0001889104302146838, + "loss": 0.4847, + "step": 6464 + }, + { + "epoch": 0.15154978111429526, + "grad_norm": 4.169005870819092, + "learning_rate": 0.0001889070584488336, + "loss": 0.6883, + "step": 6465 + }, + { + "epoch": 0.15157322268910026, + "grad_norm": 1.0957926511764526, + "learning_rate": 0.00018890368620057117, + "loss": 0.1384, + "step": 6466 + }, + { + "epoch": 0.15159666426390525, + "grad_norm": 1.4981046915054321, + "learning_rate": 0.00018890031346991486, + "loss": 0.2567, + "step": 6467 + }, + { + "epoch": 0.15162010583871025, + "grad_norm": 3.647078037261963, + "learning_rate": 0.00018889694025688296, + "loss": 0.6172, + "step": 6468 + }, + { + "epoch": 0.15164354741351524, + "grad_norm": 1.741212010383606, + "learning_rate": 0.00018889356656149375, + "loss": 0.3173, + "step": 6469 + }, + { + "epoch": 0.15166698898832023, + "grad_norm": 2.950044870376587, + "learning_rate": 0.00018889019238376556, + "loss": 0.3059, + "step": 6470 + }, + { + "epoch": 0.15169043056312523, + "grad_norm": 3.1461851596832275, + "learning_rate": 0.00018888681772371668, + "loss": 0.5213, + "step": 6471 + }, + { + "epoch": 0.15171387213793022, + "grad_norm": 2.7837393283843994, + "learning_rate": 0.00018888344258136542, + "loss": 0.4796, + "step": 6472 + }, + { + "epoch": 0.15173731371273522, + "grad_norm": 4.098050594329834, + "learning_rate": 0.0001888800669567301, + "loss": 0.2219, + "step": 6473 + }, + { + "epoch": 0.1517607552875402, + "grad_norm": 0.8524942994117737, + "learning_rate": 0.00018887669084982904, + "loss": 0.0987, + "step": 6474 + }, + { + "epoch": 0.1517841968623452, + "grad_norm": 4.320672035217285, + "learning_rate": 0.00018887331426068057, + "loss": 0.6807, + "step": 6475 + }, + { + "epoch": 0.1518076384371502, + "grad_norm": 3.0845069885253906, + "learning_rate": 0.00018886993718930298, + "loss": 0.3757, + "step": 6476 + }, + { + "epoch": 0.1518310800119552, + "grad_norm": 1.5158560276031494, + "learning_rate": 0.00018886655963571466, + "loss": 0.2469, + "step": 6477 + }, + { + "epoch": 0.1518545215867602, + "grad_norm": 3.07717227935791, + "learning_rate": 0.00018886318159993382, + "loss": 0.3661, + "step": 6478 + }, + { + "epoch": 0.15187796316156518, + "grad_norm": 3.704768657684326, + "learning_rate": 0.0001888598030819789, + "loss": 0.4924, + "step": 6479 + }, + { + "epoch": 0.15190140473637018, + "grad_norm": 5.035215854644775, + "learning_rate": 0.00018885642408186817, + "loss": 0.5743, + "step": 6480 + }, + { + "epoch": 0.15192484631117517, + "grad_norm": 5.741992950439453, + "learning_rate": 0.00018885304459961998, + "loss": 1.2666, + "step": 6481 + }, + { + "epoch": 0.1519482878859802, + "grad_norm": 4.741471290588379, + "learning_rate": 0.0001888496646352527, + "loss": 0.7284, + "step": 6482 + }, + { + "epoch": 0.1519717294607852, + "grad_norm": 0.9686075448989868, + "learning_rate": 0.00018884628418878462, + "loss": 0.1802, + "step": 6483 + }, + { + "epoch": 0.15199517103559018, + "grad_norm": 1.7396405935287476, + "learning_rate": 0.0001888429032602341, + "loss": 0.2348, + "step": 6484 + }, + { + "epoch": 0.15201861261039518, + "grad_norm": 1.8885287046432495, + "learning_rate": 0.0001888395218496195, + "loss": 1.3699, + "step": 6485 + }, + { + "epoch": 0.15204205418520017, + "grad_norm": 0.879044771194458, + "learning_rate": 0.00018883613995695914, + "loss": 0.3334, + "step": 6486 + }, + { + "epoch": 0.15206549576000516, + "grad_norm": 3.604846715927124, + "learning_rate": 0.00018883275758227144, + "loss": 0.4534, + "step": 6487 + }, + { + "epoch": 0.15208893733481016, + "grad_norm": 2.7100043296813965, + "learning_rate": 0.00018882937472557466, + "loss": 1.269, + "step": 6488 + }, + { + "epoch": 0.15211237890961515, + "grad_norm": 6.674753189086914, + "learning_rate": 0.00018882599138688718, + "loss": 1.1721, + "step": 6489 + }, + { + "epoch": 0.15213582048442015, + "grad_norm": 3.7770230770111084, + "learning_rate": 0.0001888226075662274, + "loss": 0.4668, + "step": 6490 + }, + { + "epoch": 0.15215926205922514, + "grad_norm": 2.4136126041412354, + "learning_rate": 0.00018881922326361363, + "loss": 1.0818, + "step": 6491 + }, + { + "epoch": 0.15218270363403014, + "grad_norm": 3.809812545776367, + "learning_rate": 0.00018881583847906428, + "loss": 0.4551, + "step": 6492 + }, + { + "epoch": 0.15220614520883513, + "grad_norm": 3.1160504817962646, + "learning_rate": 0.00018881245321259767, + "loss": 0.3315, + "step": 6493 + }, + { + "epoch": 0.15222958678364013, + "grad_norm": 2.0443358421325684, + "learning_rate": 0.0001888090674642322, + "loss": 0.4287, + "step": 6494 + }, + { + "epoch": 0.15225302835844512, + "grad_norm": 2.8271384239196777, + "learning_rate": 0.00018880568123398622, + "loss": 0.2031, + "step": 6495 + }, + { + "epoch": 0.1522764699332501, + "grad_norm": 2.291574239730835, + "learning_rate": 0.00018880229452187817, + "loss": 1.5288, + "step": 6496 + }, + { + "epoch": 0.1522999115080551, + "grad_norm": 3.4924488067626953, + "learning_rate": 0.00018879890732792634, + "loss": 0.3743, + "step": 6497 + }, + { + "epoch": 0.1523233530828601, + "grad_norm": 2.611806869506836, + "learning_rate": 0.00018879551965214912, + "loss": 0.3564, + "step": 6498 + }, + { + "epoch": 0.1523467946576651, + "grad_norm": 2.6336939334869385, + "learning_rate": 0.00018879213149456493, + "loss": 0.342, + "step": 6499 + }, + { + "epoch": 0.1523702362324701, + "grad_norm": 2.6253747940063477, + "learning_rate": 0.00018878874285519216, + "loss": 1.2275, + "step": 6500 + }, + { + "epoch": 0.15239367780727509, + "grad_norm": 4.757302761077881, + "learning_rate": 0.00018878535373404915, + "loss": 0.4563, + "step": 6501 + }, + { + "epoch": 0.15241711938208008, + "grad_norm": 2.882302761077881, + "learning_rate": 0.00018878196413115431, + "loss": 1.687, + "step": 6502 + }, + { + "epoch": 0.15244056095688507, + "grad_norm": 2.6405248641967773, + "learning_rate": 0.00018877857404652607, + "loss": 0.3535, + "step": 6503 + }, + { + "epoch": 0.15246400253169007, + "grad_norm": 4.955999851226807, + "learning_rate": 0.00018877518348018276, + "loss": 1.0517, + "step": 6504 + }, + { + "epoch": 0.15248744410649506, + "grad_norm": 1.132886290550232, + "learning_rate": 0.0001887717924321428, + "loss": 0.1456, + "step": 6505 + }, + { + "epoch": 0.15251088568130006, + "grad_norm": 2.433198928833008, + "learning_rate": 0.0001887684009024246, + "loss": 0.2262, + "step": 6506 + }, + { + "epoch": 0.15253432725610508, + "grad_norm": 1.4772950410842896, + "learning_rate": 0.0001887650088910466, + "loss": 0.1997, + "step": 6507 + }, + { + "epoch": 0.15255776883091007, + "grad_norm": 3.404388427734375, + "learning_rate": 0.00018876161639802715, + "loss": 0.497, + "step": 6508 + }, + { + "epoch": 0.15258121040571507, + "grad_norm": 3.0631103515625, + "learning_rate": 0.00018875822342338466, + "loss": 0.353, + "step": 6509 + }, + { + "epoch": 0.15260465198052006, + "grad_norm": 0.43619614839553833, + "learning_rate": 0.00018875482996713757, + "loss": 0.0261, + "step": 6510 + }, + { + "epoch": 0.15262809355532506, + "grad_norm": 0.36199232935905457, + "learning_rate": 0.00018875143602930424, + "loss": 0.0327, + "step": 6511 + }, + { + "epoch": 0.15265153513013005, + "grad_norm": 5.203220844268799, + "learning_rate": 0.00018874804160990316, + "loss": 0.5734, + "step": 6512 + }, + { + "epoch": 0.15267497670493504, + "grad_norm": 2.837286949157715, + "learning_rate": 0.0001887446467089527, + "loss": 0.3455, + "step": 6513 + }, + { + "epoch": 0.15269841827974004, + "grad_norm": 3.881457805633545, + "learning_rate": 0.0001887412513264713, + "loss": 0.7078, + "step": 6514 + }, + { + "epoch": 0.15272185985454503, + "grad_norm": 4.334044456481934, + "learning_rate": 0.00018873785546247735, + "loss": 0.583, + "step": 6515 + }, + { + "epoch": 0.15274530142935003, + "grad_norm": 3.7763874530792236, + "learning_rate": 0.00018873445911698933, + "loss": 0.4974, + "step": 6516 + }, + { + "epoch": 0.15276874300415502, + "grad_norm": 6.021674156188965, + "learning_rate": 0.00018873106229002563, + "loss": 0.7493, + "step": 6517 + }, + { + "epoch": 0.15279218457896002, + "grad_norm": 4.199775695800781, + "learning_rate": 0.0001887276649816047, + "loss": 0.4729, + "step": 6518 + }, + { + "epoch": 0.152815626153765, + "grad_norm": 4.3652238845825195, + "learning_rate": 0.00018872426719174492, + "loss": 0.6276, + "step": 6519 + }, + { + "epoch": 0.15283906772857, + "grad_norm": 4.8949360847473145, + "learning_rate": 0.00018872086892046482, + "loss": 0.4143, + "step": 6520 + }, + { + "epoch": 0.152862509303375, + "grad_norm": 4.910192966461182, + "learning_rate": 0.00018871747016778277, + "loss": 0.3299, + "step": 6521 + }, + { + "epoch": 0.15288595087818, + "grad_norm": 1.4995743036270142, + "learning_rate": 0.00018871407093371722, + "loss": 0.2181, + "step": 6522 + }, + { + "epoch": 0.152909392452985, + "grad_norm": 0.8095302581787109, + "learning_rate": 0.00018871067121828665, + "loss": 0.0822, + "step": 6523 + }, + { + "epoch": 0.15293283402778998, + "grad_norm": 4.6965484619140625, + "learning_rate": 0.00018870727102150947, + "loss": 0.6854, + "step": 6524 + }, + { + "epoch": 0.15295627560259498, + "grad_norm": 3.993690252304077, + "learning_rate": 0.00018870387034340417, + "loss": 0.302, + "step": 6525 + }, + { + "epoch": 0.15297971717739997, + "grad_norm": 1.3505336046218872, + "learning_rate": 0.00018870046918398914, + "loss": 0.2698, + "step": 6526 + }, + { + "epoch": 0.15300315875220497, + "grad_norm": 2.642805576324463, + "learning_rate": 0.0001886970675432829, + "loss": 0.2407, + "step": 6527 + }, + { + "epoch": 0.15302660032700996, + "grad_norm": 4.676273822784424, + "learning_rate": 0.00018869366542130386, + "loss": 0.3948, + "step": 6528 + }, + { + "epoch": 0.15305004190181495, + "grad_norm": 0.9997730255126953, + "learning_rate": 0.00018869026281807049, + "loss": 0.121, + "step": 6529 + }, + { + "epoch": 0.15307348347661995, + "grad_norm": 0.7620260715484619, + "learning_rate": 0.0001886868597336013, + "loss": 0.0606, + "step": 6530 + }, + { + "epoch": 0.15309692505142494, + "grad_norm": 4.764498233795166, + "learning_rate": 0.00018868345616791468, + "loss": 0.5136, + "step": 6531 + }, + { + "epoch": 0.15312036662622994, + "grad_norm": 2.208838701248169, + "learning_rate": 0.00018868005212102913, + "loss": 0.1349, + "step": 6532 + }, + { + "epoch": 0.15314380820103496, + "grad_norm": 2.6760663986206055, + "learning_rate": 0.00018867664759296315, + "loss": 1.445, + "step": 6533 + }, + { + "epoch": 0.15316724977583995, + "grad_norm": 1.808800220489502, + "learning_rate": 0.00018867324258373515, + "loss": 0.2379, + "step": 6534 + }, + { + "epoch": 0.15319069135064495, + "grad_norm": 2.560117483139038, + "learning_rate": 0.00018866983709336368, + "loss": 0.4457, + "step": 6535 + }, + { + "epoch": 0.15321413292544994, + "grad_norm": 3.2290520668029785, + "learning_rate": 0.00018866643112186714, + "loss": 0.2835, + "step": 6536 + }, + { + "epoch": 0.15323757450025494, + "grad_norm": 0.43796855211257935, + "learning_rate": 0.0001886630246692641, + "loss": 0.0636, + "step": 6537 + }, + { + "epoch": 0.15326101607505993, + "grad_norm": 3.8621959686279297, + "learning_rate": 0.00018865961773557296, + "loss": 0.3186, + "step": 6538 + }, + { + "epoch": 0.15328445764986492, + "grad_norm": 1.372256875038147, + "learning_rate": 0.00018865621032081227, + "loss": 0.1955, + "step": 6539 + }, + { + "epoch": 0.15330789922466992, + "grad_norm": 1.9189550876617432, + "learning_rate": 0.00018865280242500045, + "loss": 0.2091, + "step": 6540 + }, + { + "epoch": 0.1533313407994749, + "grad_norm": 3.451192617416382, + "learning_rate": 0.00018864939404815605, + "loss": 0.4137, + "step": 6541 + }, + { + "epoch": 0.1533547823742799, + "grad_norm": 4.209476470947266, + "learning_rate": 0.00018864598519029757, + "loss": 0.4867, + "step": 6542 + }, + { + "epoch": 0.1533782239490849, + "grad_norm": 4.378663539886475, + "learning_rate": 0.00018864257585144345, + "loss": 0.4385, + "step": 6543 + }, + { + "epoch": 0.1534016655238899, + "grad_norm": 1.7354602813720703, + "learning_rate": 0.00018863916603161223, + "loss": 0.225, + "step": 6544 + }, + { + "epoch": 0.1534251070986949, + "grad_norm": 1.2986046075820923, + "learning_rate": 0.0001886357557308224, + "loss": 0.193, + "step": 6545 + }, + { + "epoch": 0.15344854867349988, + "grad_norm": 3.668513774871826, + "learning_rate": 0.00018863234494909248, + "loss": 0.5102, + "step": 6546 + }, + { + "epoch": 0.15347199024830488, + "grad_norm": 3.215503692626953, + "learning_rate": 0.00018862893368644094, + "loss": 0.317, + "step": 6547 + }, + { + "epoch": 0.15349543182310987, + "grad_norm": 1.0738050937652588, + "learning_rate": 0.00018862552194288634, + "loss": 0.2077, + "step": 6548 + }, + { + "epoch": 0.15351887339791487, + "grad_norm": 2.772411823272705, + "learning_rate": 0.00018862210971844716, + "loss": 0.226, + "step": 6549 + }, + { + "epoch": 0.15354231497271986, + "grad_norm": 3.5092012882232666, + "learning_rate": 0.0001886186970131419, + "loss": 0.4901, + "step": 6550 + }, + { + "epoch": 0.15356575654752486, + "grad_norm": 4.859691619873047, + "learning_rate": 0.00018861528382698914, + "loss": 0.7296, + "step": 6551 + }, + { + "epoch": 0.15358919812232985, + "grad_norm": 4.014619827270508, + "learning_rate": 0.00018861187016000732, + "loss": 0.6792, + "step": 6552 + }, + { + "epoch": 0.15361263969713485, + "grad_norm": 1.598304271697998, + "learning_rate": 0.00018860845601221502, + "loss": 0.234, + "step": 6553 + }, + { + "epoch": 0.15363608127193984, + "grad_norm": 2.911593198776245, + "learning_rate": 0.00018860504138363076, + "loss": 2.1259, + "step": 6554 + }, + { + "epoch": 0.15365952284674483, + "grad_norm": 1.5224138498306274, + "learning_rate": 0.000188601626274273, + "loss": 0.1804, + "step": 6555 + }, + { + "epoch": 0.15368296442154983, + "grad_norm": 5.272702693939209, + "learning_rate": 0.00018859821068416036, + "loss": 0.699, + "step": 6556 + }, + { + "epoch": 0.15370640599635482, + "grad_norm": 4.734034538269043, + "learning_rate": 0.00018859479461331137, + "loss": 0.7474, + "step": 6557 + }, + { + "epoch": 0.15372984757115982, + "grad_norm": 4.559349536895752, + "learning_rate": 0.0001885913780617445, + "loss": 0.561, + "step": 6558 + }, + { + "epoch": 0.15375328914596484, + "grad_norm": 2.8232085704803467, + "learning_rate": 0.0001885879610294783, + "loss": 0.3342, + "step": 6559 + }, + { + "epoch": 0.15377673072076983, + "grad_norm": 0.9810433983802795, + "learning_rate": 0.00018858454351653134, + "loss": 0.154, + "step": 6560 + }, + { + "epoch": 0.15380017229557483, + "grad_norm": 3.581730604171753, + "learning_rate": 0.00018858112552292216, + "loss": 0.4416, + "step": 6561 + }, + { + "epoch": 0.15382361387037982, + "grad_norm": 1.2862236499786377, + "learning_rate": 0.00018857770704866936, + "loss": 0.217, + "step": 6562 + }, + { + "epoch": 0.15384705544518482, + "grad_norm": 1.463435411453247, + "learning_rate": 0.00018857428809379136, + "loss": 0.23, + "step": 6563 + }, + { + "epoch": 0.1538704970199898, + "grad_norm": 2.5354692935943604, + "learning_rate": 0.0001885708686583068, + "loss": 1.2259, + "step": 6564 + }, + { + "epoch": 0.1538939385947948, + "grad_norm": 1.0163196325302124, + "learning_rate": 0.00018856744874223423, + "loss": 0.1411, + "step": 6565 + }, + { + "epoch": 0.1539173801695998, + "grad_norm": 4.772700309753418, + "learning_rate": 0.00018856402834559218, + "loss": 0.7195, + "step": 6566 + }, + { + "epoch": 0.1539408217444048, + "grad_norm": 1.1490161418914795, + "learning_rate": 0.00018856060746839923, + "loss": 0.1661, + "step": 6567 + }, + { + "epoch": 0.1539642633192098, + "grad_norm": 1.458905816078186, + "learning_rate": 0.00018855718611067394, + "loss": 0.2919, + "step": 6568 + }, + { + "epoch": 0.15398770489401478, + "grad_norm": 3.0046603679656982, + "learning_rate": 0.00018855376427243486, + "loss": 0.5279, + "step": 6569 + }, + { + "epoch": 0.15401114646881978, + "grad_norm": 2.560224771499634, + "learning_rate": 0.00018855034195370055, + "loss": 1.9099, + "step": 6570 + }, + { + "epoch": 0.15403458804362477, + "grad_norm": 1.7657179832458496, + "learning_rate": 0.0001885469191544896, + "loss": 1.6983, + "step": 6571 + }, + { + "epoch": 0.15405802961842976, + "grad_norm": 1.0307191610336304, + "learning_rate": 0.0001885434958748206, + "loss": 0.1301, + "step": 6572 + }, + { + "epoch": 0.15408147119323476, + "grad_norm": 3.273507595062256, + "learning_rate": 0.0001885400721147121, + "loss": 0.6608, + "step": 6573 + }, + { + "epoch": 0.15410491276803975, + "grad_norm": 6.015194892883301, + "learning_rate": 0.00018853664787418266, + "loss": 0.6393, + "step": 6574 + }, + { + "epoch": 0.15412835434284475, + "grad_norm": 3.870643377304077, + "learning_rate": 0.00018853322315325087, + "loss": 0.5863, + "step": 6575 + }, + { + "epoch": 0.15415179591764974, + "grad_norm": 0.7475870847702026, + "learning_rate": 0.00018852979795193535, + "loss": 0.0728, + "step": 6576 + }, + { + "epoch": 0.15417523749245474, + "grad_norm": 2.9306094646453857, + "learning_rate": 0.00018852637227025464, + "loss": 0.4087, + "step": 6577 + }, + { + "epoch": 0.15419867906725973, + "grad_norm": 5.105018615722656, + "learning_rate": 0.00018852294610822735, + "loss": 0.5371, + "step": 6578 + }, + { + "epoch": 0.15422212064206473, + "grad_norm": 3.2514944076538086, + "learning_rate": 0.00018851951946587207, + "loss": 0.3624, + "step": 6579 + }, + { + "epoch": 0.15424556221686972, + "grad_norm": 4.947850227355957, + "learning_rate": 0.0001885160923432074, + "loss": 0.9126, + "step": 6580 + }, + { + "epoch": 0.1542690037916747, + "grad_norm": 0.4491809010505676, + "learning_rate": 0.00018851266474025188, + "loss": 0.0437, + "step": 6581 + }, + { + "epoch": 0.1542924453664797, + "grad_norm": 3.133453607559204, + "learning_rate": 0.0001885092366570242, + "loss": 0.3899, + "step": 6582 + }, + { + "epoch": 0.1543158869412847, + "grad_norm": 4.013339519500732, + "learning_rate": 0.00018850580809354287, + "loss": 0.6459, + "step": 6583 + }, + { + "epoch": 0.15433932851608972, + "grad_norm": 2.7115330696105957, + "learning_rate": 0.00018850237904982658, + "loss": 0.3728, + "step": 6584 + }, + { + "epoch": 0.15436277009089472, + "grad_norm": 3.54714035987854, + "learning_rate": 0.00018849894952589386, + "loss": 0.7312, + "step": 6585 + }, + { + "epoch": 0.1543862116656997, + "grad_norm": 1.4172412157058716, + "learning_rate": 0.00018849551952176337, + "loss": 0.215, + "step": 6586 + }, + { + "epoch": 0.1544096532405047, + "grad_norm": 5.2452311515808105, + "learning_rate": 0.0001884920890374537, + "loss": 0.598, + "step": 6587 + }, + { + "epoch": 0.1544330948153097, + "grad_norm": 2.771033763885498, + "learning_rate": 0.00018848865807298345, + "loss": 0.2828, + "step": 6588 + }, + { + "epoch": 0.1544565363901147, + "grad_norm": 1.1727592945098877, + "learning_rate": 0.0001884852266283713, + "loss": 0.141, + "step": 6589 + }, + { + "epoch": 0.1544799779649197, + "grad_norm": 1.703285574913025, + "learning_rate": 0.00018848179470363577, + "loss": 0.151, + "step": 6590 + }, + { + "epoch": 0.15450341953972468, + "grad_norm": 6.121667861938477, + "learning_rate": 0.00018847836229879557, + "loss": 0.512, + "step": 6591 + }, + { + "epoch": 0.15452686111452968, + "grad_norm": 6.7871294021606445, + "learning_rate": 0.00018847492941386924, + "loss": 0.99, + "step": 6592 + }, + { + "epoch": 0.15455030268933467, + "grad_norm": 3.9334373474121094, + "learning_rate": 0.0001884714960488755, + "loss": 0.6455, + "step": 6593 + }, + { + "epoch": 0.15457374426413967, + "grad_norm": 5.721231460571289, + "learning_rate": 0.00018846806220383292, + "loss": 0.7645, + "step": 6594 + }, + { + "epoch": 0.15459718583894466, + "grad_norm": 3.14770770072937, + "learning_rate": 0.00018846462787876016, + "loss": 0.5223, + "step": 6595 + }, + { + "epoch": 0.15462062741374966, + "grad_norm": 2.662346363067627, + "learning_rate": 0.00018846119307367582, + "loss": 0.1807, + "step": 6596 + }, + { + "epoch": 0.15464406898855465, + "grad_norm": 3.315558910369873, + "learning_rate": 0.00018845775778859856, + "loss": 0.3662, + "step": 6597 + }, + { + "epoch": 0.15466751056335964, + "grad_norm": 3.410670280456543, + "learning_rate": 0.00018845432202354702, + "loss": 0.4343, + "step": 6598 + }, + { + "epoch": 0.15469095213816464, + "grad_norm": 3.7625744342803955, + "learning_rate": 0.00018845088577853987, + "loss": 1.6121, + "step": 6599 + }, + { + "epoch": 0.15471439371296963, + "grad_norm": 0.6552702784538269, + "learning_rate": 0.0001884474490535957, + "loss": 0.0673, + "step": 6600 + }, + { + "epoch": 0.15473783528777463, + "grad_norm": 7.281239986419678, + "learning_rate": 0.0001884440118487332, + "loss": 0.6053, + "step": 6601 + }, + { + "epoch": 0.15476127686257962, + "grad_norm": 1.3506041765213013, + "learning_rate": 0.00018844057416397098, + "loss": 0.2325, + "step": 6602 + }, + { + "epoch": 0.15478471843738462, + "grad_norm": 3.6111059188842773, + "learning_rate": 0.00018843713599932775, + "loss": 0.418, + "step": 6603 + }, + { + "epoch": 0.1548081600121896, + "grad_norm": 2.1393449306488037, + "learning_rate": 0.0001884336973548221, + "loss": 1.0208, + "step": 6604 + }, + { + "epoch": 0.1548316015869946, + "grad_norm": 1.3313368558883667, + "learning_rate": 0.00018843025823047277, + "loss": 0.1362, + "step": 6605 + }, + { + "epoch": 0.1548550431617996, + "grad_norm": 3.0442543029785156, + "learning_rate": 0.00018842681862629834, + "loss": 0.2649, + "step": 6606 + }, + { + "epoch": 0.1548784847366046, + "grad_norm": 2.3405208587646484, + "learning_rate": 0.0001884233785423175, + "loss": 1.5873, + "step": 6607 + }, + { + "epoch": 0.1549019263114096, + "grad_norm": 3.7374627590179443, + "learning_rate": 0.00018841993797854892, + "loss": 0.4394, + "step": 6608 + }, + { + "epoch": 0.15492536788621458, + "grad_norm": 2.7228457927703857, + "learning_rate": 0.0001884164969350113, + "loss": 0.4559, + "step": 6609 + }, + { + "epoch": 0.1549488094610196, + "grad_norm": 3.521993398666382, + "learning_rate": 0.00018841305541172324, + "loss": 0.3655, + "step": 6610 + }, + { + "epoch": 0.1549722510358246, + "grad_norm": 4.024219036102295, + "learning_rate": 0.0001884096134087035, + "loss": 0.4922, + "step": 6611 + }, + { + "epoch": 0.1549956926106296, + "grad_norm": 3.642008066177368, + "learning_rate": 0.00018840617092597065, + "loss": 0.4422, + "step": 6612 + }, + { + "epoch": 0.1550191341854346, + "grad_norm": 1.513096809387207, + "learning_rate": 0.0001884027279635435, + "loss": 0.2659, + "step": 6613 + }, + { + "epoch": 0.15504257576023958, + "grad_norm": 4.193077087402344, + "learning_rate": 0.00018839928452144062, + "loss": 0.7264, + "step": 6614 + }, + { + "epoch": 0.15506601733504458, + "grad_norm": 3.1695122718811035, + "learning_rate": 0.00018839584059968077, + "loss": 0.3527, + "step": 6615 + }, + { + "epoch": 0.15508945890984957, + "grad_norm": 5.769690036773682, + "learning_rate": 0.00018839239619828258, + "loss": 0.4449, + "step": 6616 + }, + { + "epoch": 0.15511290048465456, + "grad_norm": 0.6703829169273376, + "learning_rate": 0.00018838895131726476, + "loss": 0.0776, + "step": 6617 + }, + { + "epoch": 0.15513634205945956, + "grad_norm": 4.336231231689453, + "learning_rate": 0.000188385505956646, + "loss": 0.4771, + "step": 6618 + }, + { + "epoch": 0.15515978363426455, + "grad_norm": 1.7213939428329468, + "learning_rate": 0.00018838206011644498, + "loss": 0.286, + "step": 6619 + }, + { + "epoch": 0.15518322520906955, + "grad_norm": 3.520561933517456, + "learning_rate": 0.00018837861379668048, + "loss": 0.546, + "step": 6620 + }, + { + "epoch": 0.15520666678387454, + "grad_norm": 2.932447910308838, + "learning_rate": 0.0001883751669973711, + "loss": 0.3529, + "step": 6621 + }, + { + "epoch": 0.15523010835867954, + "grad_norm": 3.0983169078826904, + "learning_rate": 0.00018837171971853559, + "loss": 0.4556, + "step": 6622 + }, + { + "epoch": 0.15525354993348453, + "grad_norm": 3.319148063659668, + "learning_rate": 0.00018836827196019263, + "loss": 0.4566, + "step": 6623 + }, + { + "epoch": 0.15527699150828952, + "grad_norm": 3.1179747581481934, + "learning_rate": 0.00018836482372236094, + "loss": 0.5771, + "step": 6624 + }, + { + "epoch": 0.15530043308309452, + "grad_norm": 2.132791519165039, + "learning_rate": 0.00018836137500505924, + "loss": 0.3823, + "step": 6625 + }, + { + "epoch": 0.1553238746578995, + "grad_norm": 3.009895086288452, + "learning_rate": 0.0001883579258083062, + "loss": 0.5318, + "step": 6626 + }, + { + "epoch": 0.1553473162327045, + "grad_norm": 4.077467441558838, + "learning_rate": 0.00018835447613212063, + "loss": 0.7331, + "step": 6627 + }, + { + "epoch": 0.1553707578075095, + "grad_norm": 1.393869400024414, + "learning_rate": 0.00018835102597652116, + "loss": 0.1356, + "step": 6628 + }, + { + "epoch": 0.1553941993823145, + "grad_norm": 3.09358549118042, + "learning_rate": 0.00018834757534152654, + "loss": 0.2464, + "step": 6629 + }, + { + "epoch": 0.1554176409571195, + "grad_norm": 0.7961544990539551, + "learning_rate": 0.00018834412422715548, + "loss": 0.0669, + "step": 6630 + }, + { + "epoch": 0.15544108253192448, + "grad_norm": 3.7204744815826416, + "learning_rate": 0.00018834067263342674, + "loss": 1.5215, + "step": 6631 + }, + { + "epoch": 0.15546452410672948, + "grad_norm": 3.7631585597991943, + "learning_rate": 0.00018833722056035902, + "loss": 0.5242, + "step": 6632 + }, + { + "epoch": 0.15548796568153447, + "grad_norm": 4.040266513824463, + "learning_rate": 0.00018833376800797105, + "loss": 0.7928, + "step": 6633 + }, + { + "epoch": 0.15551140725633947, + "grad_norm": 2.0119540691375732, + "learning_rate": 0.00018833031497628158, + "loss": 0.3239, + "step": 6634 + }, + { + "epoch": 0.1555348488311445, + "grad_norm": 4.8258748054504395, + "learning_rate": 0.00018832686146530932, + "loss": 0.5907, + "step": 6635 + }, + { + "epoch": 0.15555829040594948, + "grad_norm": 0.8466546535491943, + "learning_rate": 0.00018832340747507303, + "loss": 0.0608, + "step": 6636 + }, + { + "epoch": 0.15558173198075448, + "grad_norm": 5.42258882522583, + "learning_rate": 0.00018831995300559149, + "loss": 0.7143, + "step": 6637 + }, + { + "epoch": 0.15560517355555947, + "grad_norm": 2.466977596282959, + "learning_rate": 0.00018831649805688336, + "loss": 0.2918, + "step": 6638 + }, + { + "epoch": 0.15562861513036447, + "grad_norm": 2.4421675205230713, + "learning_rate": 0.00018831304262896742, + "loss": 0.2447, + "step": 6639 + }, + { + "epoch": 0.15565205670516946, + "grad_norm": 4.361330032348633, + "learning_rate": 0.00018830958672186245, + "loss": 0.5521, + "step": 6640 + }, + { + "epoch": 0.15567549827997446, + "grad_norm": 4.760359764099121, + "learning_rate": 0.00018830613033558717, + "loss": 0.5191, + "step": 6641 + }, + { + "epoch": 0.15569893985477945, + "grad_norm": 2.7038769721984863, + "learning_rate": 0.00018830267347016036, + "loss": 0.4636, + "step": 6642 + }, + { + "epoch": 0.15572238142958444, + "grad_norm": 6.0091447830200195, + "learning_rate": 0.00018829921612560074, + "loss": 0.3531, + "step": 6643 + }, + { + "epoch": 0.15574582300438944, + "grad_norm": 4.238377571105957, + "learning_rate": 0.00018829575830192708, + "loss": 0.5553, + "step": 6644 + }, + { + "epoch": 0.15576926457919443, + "grad_norm": 1.5447975397109985, + "learning_rate": 0.0001882922999991582, + "loss": 0.2132, + "step": 6645 + }, + { + "epoch": 0.15579270615399943, + "grad_norm": 3.5315048694610596, + "learning_rate": 0.00018828884121731282, + "loss": 0.4654, + "step": 6646 + }, + { + "epoch": 0.15581614772880442, + "grad_norm": 4.49525260925293, + "learning_rate": 0.00018828538195640965, + "loss": 0.7361, + "step": 6647 + }, + { + "epoch": 0.15583958930360942, + "grad_norm": 4.1602091789245605, + "learning_rate": 0.00018828192221646752, + "loss": 0.5338, + "step": 6648 + }, + { + "epoch": 0.1558630308784144, + "grad_norm": 2.433619499206543, + "learning_rate": 0.00018827846199750522, + "loss": 1.5644, + "step": 6649 + }, + { + "epoch": 0.1558864724532194, + "grad_norm": 2.272134304046631, + "learning_rate": 0.0001882750012995415, + "loss": 1.6166, + "step": 6650 + }, + { + "epoch": 0.1559099140280244, + "grad_norm": 3.9475107192993164, + "learning_rate": 0.00018827154012259517, + "loss": 0.5636, + "step": 6651 + }, + { + "epoch": 0.1559333556028294, + "grad_norm": 3.9744415283203125, + "learning_rate": 0.00018826807846668495, + "loss": 0.5778, + "step": 6652 + }, + { + "epoch": 0.1559567971776344, + "grad_norm": 3.870194435119629, + "learning_rate": 0.00018826461633182965, + "loss": 0.4434, + "step": 6653 + }, + { + "epoch": 0.15598023875243938, + "grad_norm": 0.4144372344017029, + "learning_rate": 0.00018826115371804803, + "loss": 0.0488, + "step": 6654 + }, + { + "epoch": 0.15600368032724438, + "grad_norm": 0.9298803210258484, + "learning_rate": 0.00018825769062535898, + "loss": 0.0828, + "step": 6655 + }, + { + "epoch": 0.15602712190204937, + "grad_norm": 1.6070833206176758, + "learning_rate": 0.00018825422705378116, + "loss": 0.2414, + "step": 6656 + }, + { + "epoch": 0.15605056347685436, + "grad_norm": 2.696646213531494, + "learning_rate": 0.00018825076300333347, + "loss": 0.2676, + "step": 6657 + }, + { + "epoch": 0.15607400505165936, + "grad_norm": 4.281284809112549, + "learning_rate": 0.0001882472984740346, + "loss": 0.3454, + "step": 6658 + }, + { + "epoch": 0.15609744662646435, + "grad_norm": 5.847055912017822, + "learning_rate": 0.0001882438334659035, + "loss": 0.7445, + "step": 6659 + }, + { + "epoch": 0.15612088820126935, + "grad_norm": 3.194196939468384, + "learning_rate": 0.0001882403679789588, + "loss": 0.2208, + "step": 6660 + }, + { + "epoch": 0.15614432977607437, + "grad_norm": 3.3137688636779785, + "learning_rate": 0.0001882369020132194, + "loss": 0.2998, + "step": 6661 + }, + { + "epoch": 0.15616777135087936, + "grad_norm": 1.2369190454483032, + "learning_rate": 0.0001882334355687041, + "loss": 0.1855, + "step": 6662 + }, + { + "epoch": 0.15619121292568436, + "grad_norm": 5.8953704833984375, + "learning_rate": 0.00018822996864543168, + "loss": 0.686, + "step": 6663 + }, + { + "epoch": 0.15621465450048935, + "grad_norm": 3.435192346572876, + "learning_rate": 0.00018822650124342098, + "loss": 0.3702, + "step": 6664 + }, + { + "epoch": 0.15623809607529435, + "grad_norm": 3.1304564476013184, + "learning_rate": 0.0001882230333626908, + "loss": 0.4989, + "step": 6665 + }, + { + "epoch": 0.15626153765009934, + "grad_norm": 3.073530435562134, + "learning_rate": 0.00018821956500325997, + "loss": 0.5034, + "step": 6666 + }, + { + "epoch": 0.15628497922490434, + "grad_norm": 1.7630666494369507, + "learning_rate": 0.00018821609616514727, + "loss": 0.2157, + "step": 6667 + }, + { + "epoch": 0.15630842079970933, + "grad_norm": 2.3358678817749023, + "learning_rate": 0.00018821262684837157, + "loss": 1.562, + "step": 6668 + }, + { + "epoch": 0.15633186237451432, + "grad_norm": 5.02272367477417, + "learning_rate": 0.0001882091570529517, + "loss": 0.6258, + "step": 6669 + }, + { + "epoch": 0.15635530394931932, + "grad_norm": 2.3877503871917725, + "learning_rate": 0.00018820568677890644, + "loss": 1.3904, + "step": 6670 + }, + { + "epoch": 0.1563787455241243, + "grad_norm": 2.2056896686553955, + "learning_rate": 0.00018820221602625466, + "loss": 0.2161, + "step": 6671 + }, + { + "epoch": 0.1564021870989293, + "grad_norm": 3.0271239280700684, + "learning_rate": 0.00018819874479501515, + "loss": 0.4965, + "step": 6672 + }, + { + "epoch": 0.1564256286737343, + "grad_norm": 3.2108042240142822, + "learning_rate": 0.0001881952730852068, + "loss": 0.4918, + "step": 6673 + }, + { + "epoch": 0.1564490702485393, + "grad_norm": 3.6582794189453125, + "learning_rate": 0.0001881918008968484, + "loss": 0.6266, + "step": 6674 + }, + { + "epoch": 0.1564725118233443, + "grad_norm": 3.848017692565918, + "learning_rate": 0.0001881883282299588, + "loss": 0.7587, + "step": 6675 + }, + { + "epoch": 0.15649595339814928, + "grad_norm": 2.0787675380706787, + "learning_rate": 0.00018818485508455686, + "loss": 0.2318, + "step": 6676 + }, + { + "epoch": 0.15651939497295428, + "grad_norm": 6.563187599182129, + "learning_rate": 0.00018818138146066142, + "loss": 0.7287, + "step": 6677 + }, + { + "epoch": 0.15654283654775927, + "grad_norm": 4.494446754455566, + "learning_rate": 0.00018817790735829132, + "loss": 0.5276, + "step": 6678 + }, + { + "epoch": 0.15656627812256427, + "grad_norm": 5.260522842407227, + "learning_rate": 0.0001881744327774654, + "loss": 0.6182, + "step": 6679 + }, + { + "epoch": 0.15658971969736926, + "grad_norm": 1.3497004508972168, + "learning_rate": 0.00018817095771820258, + "loss": 0.1934, + "step": 6680 + }, + { + "epoch": 0.15661316127217426, + "grad_norm": 3.5648305416107178, + "learning_rate": 0.00018816748218052163, + "loss": 0.4975, + "step": 6681 + }, + { + "epoch": 0.15663660284697925, + "grad_norm": 2.799830198287964, + "learning_rate": 0.00018816400616444146, + "loss": 0.3613, + "step": 6682 + }, + { + "epoch": 0.15666004442178424, + "grad_norm": 3.508605718612671, + "learning_rate": 0.00018816052966998088, + "loss": 0.4868, + "step": 6683 + }, + { + "epoch": 0.15668348599658924, + "grad_norm": 3.831678867340088, + "learning_rate": 0.00018815705269715883, + "loss": 0.6863, + "step": 6684 + }, + { + "epoch": 0.15670692757139423, + "grad_norm": 4.123430252075195, + "learning_rate": 0.00018815357524599408, + "loss": 0.3754, + "step": 6685 + }, + { + "epoch": 0.15673036914619926, + "grad_norm": 4.263240337371826, + "learning_rate": 0.0001881500973165056, + "loss": 0.5527, + "step": 6686 + }, + { + "epoch": 0.15675381072100425, + "grad_norm": 3.9882447719573975, + "learning_rate": 0.00018814661890871216, + "loss": 0.3752, + "step": 6687 + }, + { + "epoch": 0.15677725229580924, + "grad_norm": 1.9714512825012207, + "learning_rate": 0.00018814314002263275, + "loss": 0.1691, + "step": 6688 + }, + { + "epoch": 0.15680069387061424, + "grad_norm": 5.776281356811523, + "learning_rate": 0.00018813966065828614, + "loss": 0.8131, + "step": 6689 + }, + { + "epoch": 0.15682413544541923, + "grad_norm": 3.188065767288208, + "learning_rate": 0.00018813618081569127, + "loss": 0.4497, + "step": 6690 + }, + { + "epoch": 0.15684757702022423, + "grad_norm": 1.6827105283737183, + "learning_rate": 0.000188132700494867, + "loss": 1.2933, + "step": 6691 + }, + { + "epoch": 0.15687101859502922, + "grad_norm": 3.213111162185669, + "learning_rate": 0.00018812921969583223, + "loss": 0.3954, + "step": 6692 + }, + { + "epoch": 0.15689446016983422, + "grad_norm": 2.6665124893188477, + "learning_rate": 0.00018812573841860583, + "loss": 0.3484, + "step": 6693 + }, + { + "epoch": 0.1569179017446392, + "grad_norm": 3.370572805404663, + "learning_rate": 0.00018812225666320668, + "loss": 0.3538, + "step": 6694 + }, + { + "epoch": 0.1569413433194442, + "grad_norm": 0.8857362866401672, + "learning_rate": 0.0001881187744296537, + "loss": 0.1211, + "step": 6695 + }, + { + "epoch": 0.1569647848942492, + "grad_norm": 4.949525356292725, + "learning_rate": 0.0001881152917179658, + "loss": 0.3752, + "step": 6696 + }, + { + "epoch": 0.1569882264690542, + "grad_norm": 4.303220748901367, + "learning_rate": 0.00018811180852816182, + "loss": 0.5746, + "step": 6697 + }, + { + "epoch": 0.1570116680438592, + "grad_norm": 2.586298942565918, + "learning_rate": 0.0001881083248602607, + "loss": 0.2329, + "step": 6698 + }, + { + "epoch": 0.15703510961866418, + "grad_norm": 4.020472049713135, + "learning_rate": 0.0001881048407142813, + "loss": 0.5028, + "step": 6699 + }, + { + "epoch": 0.15705855119346918, + "grad_norm": 2.8300962448120117, + "learning_rate": 0.00018810135609024259, + "loss": 1.9234, + "step": 6700 + }, + { + "epoch": 0.15708199276827417, + "grad_norm": 2.5252645015716553, + "learning_rate": 0.00018809787098816343, + "loss": 0.2045, + "step": 6701 + }, + { + "epoch": 0.15710543434307916, + "grad_norm": 0.5777347683906555, + "learning_rate": 0.00018809438540806276, + "loss": 0.0821, + "step": 6702 + }, + { + "epoch": 0.15712887591788416, + "grad_norm": 3.1279964447021484, + "learning_rate": 0.00018809089934995946, + "loss": 0.3556, + "step": 6703 + }, + { + "epoch": 0.15715231749268915, + "grad_norm": 3.8963022232055664, + "learning_rate": 0.0001880874128138725, + "loss": 0.5856, + "step": 6704 + }, + { + "epoch": 0.15717575906749415, + "grad_norm": 4.094825267791748, + "learning_rate": 0.00018808392579982073, + "loss": 0.3281, + "step": 6705 + }, + { + "epoch": 0.15719920064229914, + "grad_norm": 2.983279228210449, + "learning_rate": 0.0001880804383078231, + "loss": 0.4176, + "step": 6706 + }, + { + "epoch": 0.15722264221710414, + "grad_norm": 3.7188785076141357, + "learning_rate": 0.00018807695033789854, + "loss": 0.5388, + "step": 6707 + }, + { + "epoch": 0.15724608379190913, + "grad_norm": 2.9006588459014893, + "learning_rate": 0.00018807346189006596, + "loss": 0.3552, + "step": 6708 + }, + { + "epoch": 0.15726952536671412, + "grad_norm": 2.893770456314087, + "learning_rate": 0.0001880699729643443, + "loss": 0.3649, + "step": 6709 + }, + { + "epoch": 0.15729296694151912, + "grad_norm": 4.539196014404297, + "learning_rate": 0.00018806648356075253, + "loss": 0.5288, + "step": 6710 + }, + { + "epoch": 0.1573164085163241, + "grad_norm": 2.5656819343566895, + "learning_rate": 0.0001880629936793095, + "loss": 0.324, + "step": 6711 + }, + { + "epoch": 0.15733985009112914, + "grad_norm": 6.259242057800293, + "learning_rate": 0.00018805950332003422, + "loss": 0.5664, + "step": 6712 + }, + { + "epoch": 0.15736329166593413, + "grad_norm": 3.597630739212036, + "learning_rate": 0.0001880560124829456, + "loss": 0.3413, + "step": 6713 + }, + { + "epoch": 0.15738673324073912, + "grad_norm": 3.7209155559539795, + "learning_rate": 0.00018805252116806253, + "loss": 0.4679, + "step": 6714 + }, + { + "epoch": 0.15741017481554412, + "grad_norm": 5.052282333374023, + "learning_rate": 0.00018804902937540402, + "loss": 0.4751, + "step": 6715 + }, + { + "epoch": 0.1574336163903491, + "grad_norm": 4.815585613250732, + "learning_rate": 0.00018804553710498906, + "loss": 0.5555, + "step": 6716 + }, + { + "epoch": 0.1574570579651541, + "grad_norm": 1.5252978801727295, + "learning_rate": 0.00018804204435683648, + "loss": 0.2188, + "step": 6717 + }, + { + "epoch": 0.1574804995399591, + "grad_norm": 2.641719102859497, + "learning_rate": 0.00018803855113096532, + "loss": 1.6977, + "step": 6718 + }, + { + "epoch": 0.1575039411147641, + "grad_norm": 4.66179084777832, + "learning_rate": 0.0001880350574273945, + "loss": 0.6664, + "step": 6719 + }, + { + "epoch": 0.1575273826895691, + "grad_norm": 0.9912242889404297, + "learning_rate": 0.00018803156324614297, + "loss": 0.1763, + "step": 6720 + }, + { + "epoch": 0.15755082426437408, + "grad_norm": 3.6084911823272705, + "learning_rate": 0.00018802806858722972, + "loss": 0.5656, + "step": 6721 + }, + { + "epoch": 0.15757426583917908, + "grad_norm": 0.5970975756645203, + "learning_rate": 0.00018802457345067368, + "loss": 0.0729, + "step": 6722 + }, + { + "epoch": 0.15759770741398407, + "grad_norm": 0.5581799149513245, + "learning_rate": 0.00018802107783649384, + "loss": 0.0841, + "step": 6723 + }, + { + "epoch": 0.15762114898878907, + "grad_norm": 4.932704448699951, + "learning_rate": 0.00018801758174470917, + "loss": 0.5689, + "step": 6724 + }, + { + "epoch": 0.15764459056359406, + "grad_norm": 2.132763385772705, + "learning_rate": 0.00018801408517533862, + "loss": 0.2354, + "step": 6725 + }, + { + "epoch": 0.15766803213839906, + "grad_norm": 2.5552523136138916, + "learning_rate": 0.00018801058812840116, + "loss": 0.268, + "step": 6726 + }, + { + "epoch": 0.15769147371320405, + "grad_norm": 3.3848538398742676, + "learning_rate": 0.00018800709060391575, + "loss": 0.268, + "step": 6727 + }, + { + "epoch": 0.15771491528800904, + "grad_norm": 4.342230319976807, + "learning_rate": 0.00018800359260190142, + "loss": 0.5614, + "step": 6728 + }, + { + "epoch": 0.15773835686281404, + "grad_norm": 4.680584907531738, + "learning_rate": 0.00018800009412237714, + "loss": 0.8084, + "step": 6729 + }, + { + "epoch": 0.15776179843761903, + "grad_norm": 3.3201568126678467, + "learning_rate": 0.00018799659516536183, + "loss": 0.5763, + "step": 6730 + }, + { + "epoch": 0.15778524001242403, + "grad_norm": 0.8416464924812317, + "learning_rate": 0.00018799309573087455, + "loss": 0.0966, + "step": 6731 + }, + { + "epoch": 0.15780868158722902, + "grad_norm": 1.9936736822128296, + "learning_rate": 0.00018798959581893426, + "loss": 0.1945, + "step": 6732 + }, + { + "epoch": 0.15783212316203402, + "grad_norm": 3.2619826793670654, + "learning_rate": 0.00018798609542955994, + "loss": 0.3547, + "step": 6733 + }, + { + "epoch": 0.157855564736839, + "grad_norm": 3.3410120010375977, + "learning_rate": 0.00018798259456277058, + "loss": 0.3324, + "step": 6734 + }, + { + "epoch": 0.157879006311644, + "grad_norm": 2.371123790740967, + "learning_rate": 0.0001879790932185852, + "loss": 0.1272, + "step": 6735 + }, + { + "epoch": 0.157902447886449, + "grad_norm": 4.857202529907227, + "learning_rate": 0.0001879755913970228, + "loss": 0.5319, + "step": 6736 + }, + { + "epoch": 0.157925889461254, + "grad_norm": 3.2496285438537598, + "learning_rate": 0.00018797208909810236, + "loss": 1.7081, + "step": 6737 + }, + { + "epoch": 0.15794933103605902, + "grad_norm": 3.0543084144592285, + "learning_rate": 0.00018796858632184286, + "loss": 0.1963, + "step": 6738 + }, + { + "epoch": 0.157972772610864, + "grad_norm": 0.9047747254371643, + "learning_rate": 0.00018796508306826338, + "loss": 0.1554, + "step": 6739 + }, + { + "epoch": 0.157996214185669, + "grad_norm": 3.2505416870117188, + "learning_rate": 0.0001879615793373829, + "loss": 0.2503, + "step": 6740 + }, + { + "epoch": 0.158019655760474, + "grad_norm": 3.453690767288208, + "learning_rate": 0.00018795807512922034, + "loss": 0.2336, + "step": 6741 + }, + { + "epoch": 0.158043097335279, + "grad_norm": 5.041258811950684, + "learning_rate": 0.00018795457044379484, + "loss": 0.3667, + "step": 6742 + }, + { + "epoch": 0.158066538910084, + "grad_norm": 0.5015707612037659, + "learning_rate": 0.00018795106528112537, + "loss": 0.0368, + "step": 6743 + }, + { + "epoch": 0.15808998048488898, + "grad_norm": 6.452219486236572, + "learning_rate": 0.00018794755964123098, + "loss": 0.4757, + "step": 6744 + }, + { + "epoch": 0.15811342205969398, + "grad_norm": 2.3386192321777344, + "learning_rate": 0.00018794405352413058, + "loss": 0.2681, + "step": 6745 + }, + { + "epoch": 0.15813686363449897, + "grad_norm": 3.102177381515503, + "learning_rate": 0.00018794054692984334, + "loss": 0.3342, + "step": 6746 + }, + { + "epoch": 0.15816030520930396, + "grad_norm": 0.9867315292358398, + "learning_rate": 0.00018793703985838818, + "loss": 0.1522, + "step": 6747 + }, + { + "epoch": 0.15818374678410896, + "grad_norm": 4.193514823913574, + "learning_rate": 0.00018793353230978416, + "loss": 0.4738, + "step": 6748 + }, + { + "epoch": 0.15820718835891395, + "grad_norm": 5.290258884429932, + "learning_rate": 0.00018793002428405035, + "loss": 0.7423, + "step": 6749 + }, + { + "epoch": 0.15823062993371895, + "grad_norm": 1.541534662246704, + "learning_rate": 0.00018792651578120574, + "loss": 0.2586, + "step": 6750 + }, + { + "epoch": 0.15825407150852394, + "grad_norm": 4.433980464935303, + "learning_rate": 0.00018792300680126942, + "loss": 0.4282, + "step": 6751 + }, + { + "epoch": 0.15827751308332894, + "grad_norm": 3.7031710147857666, + "learning_rate": 0.00018791949734426035, + "loss": 0.7758, + "step": 6752 + }, + { + "epoch": 0.15830095465813393, + "grad_norm": 2.06569504737854, + "learning_rate": 0.0001879159874101976, + "loss": 1.4453, + "step": 6753 + }, + { + "epoch": 0.15832439623293892, + "grad_norm": 3.554231643676758, + "learning_rate": 0.0001879124769991003, + "loss": 0.522, + "step": 6754 + }, + { + "epoch": 0.15834783780774392, + "grad_norm": 7.797680854797363, + "learning_rate": 0.0001879089661109874, + "loss": 0.5118, + "step": 6755 + }, + { + "epoch": 0.1583712793825489, + "grad_norm": 1.8538542985916138, + "learning_rate": 0.00018790545474587798, + "loss": 0.2056, + "step": 6756 + }, + { + "epoch": 0.1583947209573539, + "grad_norm": 1.4355664253234863, + "learning_rate": 0.00018790194290379107, + "loss": 0.2646, + "step": 6757 + }, + { + "epoch": 0.1584181625321589, + "grad_norm": 3.811107873916626, + "learning_rate": 0.00018789843058474574, + "loss": 0.2325, + "step": 6758 + }, + { + "epoch": 0.1584416041069639, + "grad_norm": 2.91279673576355, + "learning_rate": 0.00018789491778876106, + "loss": 0.3173, + "step": 6759 + }, + { + "epoch": 0.1584650456817689, + "grad_norm": 1.4889116287231445, + "learning_rate": 0.0001878914045158561, + "loss": 0.2831, + "step": 6760 + }, + { + "epoch": 0.15848848725657388, + "grad_norm": 0.9444628357887268, + "learning_rate": 0.00018788789076604992, + "loss": 0.1146, + "step": 6761 + }, + { + "epoch": 0.15851192883137888, + "grad_norm": 1.3100026845932007, + "learning_rate": 0.00018788437653936155, + "loss": 0.2302, + "step": 6762 + }, + { + "epoch": 0.1585353704061839, + "grad_norm": 2.815960645675659, + "learning_rate": 0.00018788086183581007, + "loss": 1.6272, + "step": 6763 + }, + { + "epoch": 0.1585588119809889, + "grad_norm": 0.8166928887367249, + "learning_rate": 0.00018787734665541458, + "loss": 0.1345, + "step": 6764 + }, + { + "epoch": 0.1585822535557939, + "grad_norm": 1.516648530960083, + "learning_rate": 0.00018787383099819413, + "loss": 0.232, + "step": 6765 + }, + { + "epoch": 0.15860569513059888, + "grad_norm": 3.4603476524353027, + "learning_rate": 0.0001878703148641678, + "loss": 0.4819, + "step": 6766 + }, + { + "epoch": 0.15862913670540388, + "grad_norm": 2.8944873809814453, + "learning_rate": 0.00018786679825335469, + "loss": 0.6084, + "step": 6767 + }, + { + "epoch": 0.15865257828020887, + "grad_norm": 0.4156653881072998, + "learning_rate": 0.00018786328116577385, + "loss": 0.069, + "step": 6768 + }, + { + "epoch": 0.15867601985501387, + "grad_norm": 2.872591018676758, + "learning_rate": 0.00018785976360144438, + "loss": 0.427, + "step": 6769 + }, + { + "epoch": 0.15869946142981886, + "grad_norm": 3.258772611618042, + "learning_rate": 0.00018785624556038534, + "loss": 0.4061, + "step": 6770 + }, + { + "epoch": 0.15872290300462386, + "grad_norm": 3.396561622619629, + "learning_rate": 0.0001878527270426159, + "loss": 1.098, + "step": 6771 + }, + { + "epoch": 0.15874634457942885, + "grad_norm": 4.067331790924072, + "learning_rate": 0.00018784920804815504, + "loss": 0.3997, + "step": 6772 + }, + { + "epoch": 0.15876978615423384, + "grad_norm": 4.312479019165039, + "learning_rate": 0.00018784568857702191, + "loss": 0.5978, + "step": 6773 + }, + { + "epoch": 0.15879322772903884, + "grad_norm": 4.08172607421875, + "learning_rate": 0.00018784216862923562, + "loss": 0.508, + "step": 6774 + }, + { + "epoch": 0.15881666930384383, + "grad_norm": 3.5605549812316895, + "learning_rate": 0.00018783864820481525, + "loss": 0.3587, + "step": 6775 + }, + { + "epoch": 0.15884011087864883, + "grad_norm": 2.556417226791382, + "learning_rate": 0.0001878351273037799, + "loss": 0.4294, + "step": 6776 + }, + { + "epoch": 0.15886355245345382, + "grad_norm": 3.663546562194824, + "learning_rate": 0.00018783160592614868, + "loss": 0.4539, + "step": 6777 + }, + { + "epoch": 0.15888699402825882, + "grad_norm": 3.7209787368774414, + "learning_rate": 0.00018782808407194067, + "loss": 0.8165, + "step": 6778 + }, + { + "epoch": 0.1589104356030638, + "grad_norm": 3.0020265579223633, + "learning_rate": 0.00018782456174117504, + "loss": 1.9408, + "step": 6779 + }, + { + "epoch": 0.1589338771778688, + "grad_norm": 2.6909303665161133, + "learning_rate": 0.00018782103893387087, + "loss": 0.4301, + "step": 6780 + }, + { + "epoch": 0.1589573187526738, + "grad_norm": 3.54498291015625, + "learning_rate": 0.00018781751565004727, + "loss": 0.4643, + "step": 6781 + }, + { + "epoch": 0.1589807603274788, + "grad_norm": 1.1282696723937988, + "learning_rate": 0.00018781399188972335, + "loss": 0.1737, + "step": 6782 + }, + { + "epoch": 0.1590042019022838, + "grad_norm": 2.5895297527313232, + "learning_rate": 0.00018781046765291824, + "loss": 1.6868, + "step": 6783 + }, + { + "epoch": 0.15902764347708878, + "grad_norm": 0.8095921874046326, + "learning_rate": 0.00018780694293965108, + "loss": 0.1426, + "step": 6784 + }, + { + "epoch": 0.15905108505189378, + "grad_norm": 4.486494541168213, + "learning_rate": 0.00018780341774994098, + "loss": 0.8359, + "step": 6785 + }, + { + "epoch": 0.15907452662669877, + "grad_norm": 3.1224944591522217, + "learning_rate": 0.00018779989208380705, + "loss": 0.5372, + "step": 6786 + }, + { + "epoch": 0.15909796820150376, + "grad_norm": 3.282881498336792, + "learning_rate": 0.00018779636594126845, + "loss": 0.3034, + "step": 6787 + }, + { + "epoch": 0.15912140977630876, + "grad_norm": 2.9344606399536133, + "learning_rate": 0.0001877928393223443, + "loss": 0.2055, + "step": 6788 + }, + { + "epoch": 0.15914485135111378, + "grad_norm": 3.1451821327209473, + "learning_rate": 0.00018778931222705373, + "loss": 1.5676, + "step": 6789 + }, + { + "epoch": 0.15916829292591878, + "grad_norm": 3.5404884815216064, + "learning_rate": 0.00018778578465541588, + "loss": 0.6933, + "step": 6790 + }, + { + "epoch": 0.15919173450072377, + "grad_norm": 3.9534859657287598, + "learning_rate": 0.00018778225660744991, + "loss": 0.5889, + "step": 6791 + }, + { + "epoch": 0.15921517607552876, + "grad_norm": 2.372102975845337, + "learning_rate": 0.0001877787280831749, + "loss": 0.3931, + "step": 6792 + }, + { + "epoch": 0.15923861765033376, + "grad_norm": 1.8682196140289307, + "learning_rate": 0.00018777519908261007, + "loss": 0.191, + "step": 6793 + }, + { + "epoch": 0.15926205922513875, + "grad_norm": 3.3765311241149902, + "learning_rate": 0.00018777166960577457, + "loss": 0.5288, + "step": 6794 + }, + { + "epoch": 0.15928550079994375, + "grad_norm": 1.6600018739700317, + "learning_rate": 0.0001877681396526875, + "loss": 0.2752, + "step": 6795 + }, + { + "epoch": 0.15930894237474874, + "grad_norm": 2.5783071517944336, + "learning_rate": 0.00018776460922336803, + "loss": 1.5641, + "step": 6796 + }, + { + "epoch": 0.15933238394955374, + "grad_norm": 1.9406784772872925, + "learning_rate": 0.00018776107831783534, + "loss": 0.2064, + "step": 6797 + }, + { + "epoch": 0.15935582552435873, + "grad_norm": 5.031788349151611, + "learning_rate": 0.00018775754693610856, + "loss": 0.729, + "step": 6798 + }, + { + "epoch": 0.15937926709916372, + "grad_norm": 3.844219923019409, + "learning_rate": 0.00018775401507820688, + "loss": 0.458, + "step": 6799 + }, + { + "epoch": 0.15940270867396872, + "grad_norm": 3.274338960647583, + "learning_rate": 0.0001877504827441494, + "loss": 0.4388, + "step": 6800 + }, + { + "epoch": 0.1594261502487737, + "grad_norm": 1.6571929454803467, + "learning_rate": 0.0001877469499339554, + "loss": 1.159, + "step": 6801 + }, + { + "epoch": 0.1594495918235787, + "grad_norm": 2.6788313388824463, + "learning_rate": 0.0001877434166476439, + "loss": 1.4315, + "step": 6802 + }, + { + "epoch": 0.1594730333983837, + "grad_norm": 0.7255390286445618, + "learning_rate": 0.00018773988288523424, + "loss": 0.0911, + "step": 6803 + }, + { + "epoch": 0.1594964749731887, + "grad_norm": 2.878915548324585, + "learning_rate": 0.00018773634864674549, + "loss": 0.4771, + "step": 6804 + }, + { + "epoch": 0.1595199165479937, + "grad_norm": 2.2764892578125, + "learning_rate": 0.00018773281393219679, + "loss": 0.2648, + "step": 6805 + }, + { + "epoch": 0.15954335812279868, + "grad_norm": 2.2801296710968018, + "learning_rate": 0.0001877292787416074, + "loss": 0.298, + "step": 6806 + }, + { + "epoch": 0.15956679969760368, + "grad_norm": 1.503864049911499, + "learning_rate": 0.0001877257430749965, + "loss": 0.2532, + "step": 6807 + }, + { + "epoch": 0.15959024127240867, + "grad_norm": 3.374361991882324, + "learning_rate": 0.00018772220693238323, + "loss": 0.4285, + "step": 6808 + }, + { + "epoch": 0.15961368284721367, + "grad_norm": 2.8431074619293213, + "learning_rate": 0.0001877186703137868, + "loss": 0.471, + "step": 6809 + }, + { + "epoch": 0.15963712442201866, + "grad_norm": 3.7619524002075195, + "learning_rate": 0.00018771513321922642, + "loss": 0.5781, + "step": 6810 + }, + { + "epoch": 0.15966056599682366, + "grad_norm": 0.6081570386886597, + "learning_rate": 0.00018771159564872126, + "loss": 0.071, + "step": 6811 + }, + { + "epoch": 0.15968400757162865, + "grad_norm": 2.9692773818969727, + "learning_rate": 0.00018770805760229048, + "loss": 0.3186, + "step": 6812 + }, + { + "epoch": 0.15970744914643364, + "grad_norm": 4.446171283721924, + "learning_rate": 0.00018770451907995335, + "loss": 0.5754, + "step": 6813 + }, + { + "epoch": 0.15973089072123867, + "grad_norm": 3.0877089500427246, + "learning_rate": 0.000187700980081729, + "loss": 0.337, + "step": 6814 + }, + { + "epoch": 0.15975433229604366, + "grad_norm": 4.512960433959961, + "learning_rate": 0.0001876974406076367, + "loss": 0.599, + "step": 6815 + }, + { + "epoch": 0.15977777387084866, + "grad_norm": 2.3081202507019043, + "learning_rate": 0.00018769390065769562, + "loss": 1.9751, + "step": 6816 + }, + { + "epoch": 0.15980121544565365, + "grad_norm": 2.4854660034179688, + "learning_rate": 0.00018769036023192492, + "loss": 0.2334, + "step": 6817 + }, + { + "epoch": 0.15982465702045864, + "grad_norm": 1.2965514659881592, + "learning_rate": 0.00018768681933034392, + "loss": 0.1319, + "step": 6818 + }, + { + "epoch": 0.15984809859526364, + "grad_norm": 4.602838039398193, + "learning_rate": 0.00018768327795297177, + "loss": 0.8998, + "step": 6819 + }, + { + "epoch": 0.15987154017006863, + "grad_norm": 1.1832062005996704, + "learning_rate": 0.00018767973609982766, + "loss": 0.1334, + "step": 6820 + }, + { + "epoch": 0.15989498174487363, + "grad_norm": 3.055259943008423, + "learning_rate": 0.00018767619377093082, + "loss": 0.2653, + "step": 6821 + }, + { + "epoch": 0.15991842331967862, + "grad_norm": 0.6761828660964966, + "learning_rate": 0.0001876726509663005, + "loss": 0.0889, + "step": 6822 + }, + { + "epoch": 0.15994186489448362, + "grad_norm": 1.03993821144104, + "learning_rate": 0.00018766910768595595, + "loss": 0.1398, + "step": 6823 + }, + { + "epoch": 0.1599653064692886, + "grad_norm": 2.040389060974121, + "learning_rate": 0.0001876655639299163, + "loss": 1.6541, + "step": 6824 + }, + { + "epoch": 0.1599887480440936, + "grad_norm": 4.325235843658447, + "learning_rate": 0.00018766201969820087, + "loss": 0.4129, + "step": 6825 + }, + { + "epoch": 0.1600121896188986, + "grad_norm": 5.071159839630127, + "learning_rate": 0.00018765847499082884, + "loss": 0.5733, + "step": 6826 + }, + { + "epoch": 0.1600356311937036, + "grad_norm": 2.89520525932312, + "learning_rate": 0.0001876549298078195, + "loss": 0.4919, + "step": 6827 + }, + { + "epoch": 0.1600590727685086, + "grad_norm": 2.8149635791778564, + "learning_rate": 0.000187651384149192, + "loss": 0.3376, + "step": 6828 + }, + { + "epoch": 0.16008251434331358, + "grad_norm": 3.518031597137451, + "learning_rate": 0.00018764783801496565, + "loss": 0.5641, + "step": 6829 + }, + { + "epoch": 0.16010595591811858, + "grad_norm": 2.419015884399414, + "learning_rate": 0.00018764429140515965, + "loss": 0.3637, + "step": 6830 + }, + { + "epoch": 0.16012939749292357, + "grad_norm": 0.8463472127914429, + "learning_rate": 0.00018764074431979326, + "loss": 0.0945, + "step": 6831 + }, + { + "epoch": 0.16015283906772856, + "grad_norm": 2.45259428024292, + "learning_rate": 0.00018763719675888573, + "loss": 1.372, + "step": 6832 + }, + { + "epoch": 0.16017628064253356, + "grad_norm": 1.0085617303848267, + "learning_rate": 0.00018763364872245633, + "loss": 0.1207, + "step": 6833 + }, + { + "epoch": 0.16019972221733855, + "grad_norm": 3.369480609893799, + "learning_rate": 0.00018763010021052425, + "loss": 0.3093, + "step": 6834 + }, + { + "epoch": 0.16022316379214355, + "grad_norm": 4.1837921142578125, + "learning_rate": 0.00018762655122310878, + "loss": 0.4294, + "step": 6835 + }, + { + "epoch": 0.16024660536694854, + "grad_norm": 0.8684640526771545, + "learning_rate": 0.0001876230017602292, + "loss": 0.0586, + "step": 6836 + }, + { + "epoch": 0.16027004694175354, + "grad_norm": 4.40787935256958, + "learning_rate": 0.00018761945182190477, + "loss": 0.6092, + "step": 6837 + }, + { + "epoch": 0.16029348851655853, + "grad_norm": 2.791942596435547, + "learning_rate": 0.00018761590140815471, + "loss": 0.4858, + "step": 6838 + }, + { + "epoch": 0.16031693009136352, + "grad_norm": 2.0626912117004395, + "learning_rate": 0.0001876123505189983, + "loss": 0.243, + "step": 6839 + }, + { + "epoch": 0.16034037166616855, + "grad_norm": 2.654374599456787, + "learning_rate": 0.00018760879915445485, + "loss": 0.1925, + "step": 6840 + }, + { + "epoch": 0.16036381324097354, + "grad_norm": 1.7189244031906128, + "learning_rate": 0.00018760524731454355, + "loss": 0.2223, + "step": 6841 + }, + { + "epoch": 0.16038725481577853, + "grad_norm": 0.8199900984764099, + "learning_rate": 0.0001876016949992837, + "loss": 0.1172, + "step": 6842 + }, + { + "epoch": 0.16041069639058353, + "grad_norm": 3.919570207595825, + "learning_rate": 0.00018759814220869464, + "loss": 0.5801, + "step": 6843 + }, + { + "epoch": 0.16043413796538852, + "grad_norm": 3.881253719329834, + "learning_rate": 0.00018759458894279557, + "loss": 0.434, + "step": 6844 + }, + { + "epoch": 0.16045757954019352, + "grad_norm": 3.145691156387329, + "learning_rate": 0.0001875910352016058, + "loss": 0.3767, + "step": 6845 + }, + { + "epoch": 0.1604810211149985, + "grad_norm": 4.811984539031982, + "learning_rate": 0.0001875874809851446, + "loss": 0.6223, + "step": 6846 + }, + { + "epoch": 0.1605044626898035, + "grad_norm": 1.8230440616607666, + "learning_rate": 0.00018758392629343126, + "loss": 0.1101, + "step": 6847 + }, + { + "epoch": 0.1605279042646085, + "grad_norm": 3.548105239868164, + "learning_rate": 0.0001875803711264851, + "loss": 0.4228, + "step": 6848 + }, + { + "epoch": 0.1605513458394135, + "grad_norm": 0.5361230969429016, + "learning_rate": 0.00018757681548432537, + "loss": 0.05, + "step": 6849 + }, + { + "epoch": 0.1605747874142185, + "grad_norm": 4.835147857666016, + "learning_rate": 0.00018757325936697138, + "loss": 0.7486, + "step": 6850 + }, + { + "epoch": 0.16059822898902348, + "grad_norm": 4.43109130859375, + "learning_rate": 0.00018756970277444243, + "loss": 0.6127, + "step": 6851 + }, + { + "epoch": 0.16062167056382848, + "grad_norm": 4.252618789672852, + "learning_rate": 0.00018756614570675782, + "loss": 0.8316, + "step": 6852 + }, + { + "epoch": 0.16064511213863347, + "grad_norm": 3.185988187789917, + "learning_rate": 0.0001875625881639368, + "loss": 0.3555, + "step": 6853 + }, + { + "epoch": 0.16066855371343847, + "grad_norm": 5.534113883972168, + "learning_rate": 0.00018755903014599875, + "loss": 0.5949, + "step": 6854 + }, + { + "epoch": 0.16069199528824346, + "grad_norm": 4.048070907592773, + "learning_rate": 0.00018755547165296292, + "loss": 0.663, + "step": 6855 + }, + { + "epoch": 0.16071543686304846, + "grad_norm": 2.9127275943756104, + "learning_rate": 0.00018755191268484864, + "loss": 0.2314, + "step": 6856 + }, + { + "epoch": 0.16073887843785345, + "grad_norm": 4.766898155212402, + "learning_rate": 0.00018754835324167522, + "loss": 0.671, + "step": 6857 + }, + { + "epoch": 0.16076232001265844, + "grad_norm": 0.6714709401130676, + "learning_rate": 0.00018754479332346198, + "loss": 0.0657, + "step": 6858 + }, + { + "epoch": 0.16078576158746344, + "grad_norm": 3.0317728519439697, + "learning_rate": 0.00018754123293022822, + "loss": 0.4778, + "step": 6859 + }, + { + "epoch": 0.16080920316226843, + "grad_norm": 0.9630699753761292, + "learning_rate": 0.00018753767206199329, + "loss": 0.1356, + "step": 6860 + }, + { + "epoch": 0.16083264473707343, + "grad_norm": 1.07575523853302, + "learning_rate": 0.00018753411071877648, + "loss": 0.2167, + "step": 6861 + }, + { + "epoch": 0.16085608631187842, + "grad_norm": 3.0159425735473633, + "learning_rate": 0.00018753054890059714, + "loss": 0.3891, + "step": 6862 + }, + { + "epoch": 0.16087952788668342, + "grad_norm": 2.4413869380950928, + "learning_rate": 0.00018752698660747457, + "loss": 1.7344, + "step": 6863 + }, + { + "epoch": 0.1609029694614884, + "grad_norm": 2.746506690979004, + "learning_rate": 0.00018752342383942809, + "loss": 0.3562, + "step": 6864 + }, + { + "epoch": 0.16092641103629343, + "grad_norm": 2.19954514503479, + "learning_rate": 0.00018751986059647706, + "loss": 0.3676, + "step": 6865 + }, + { + "epoch": 0.16094985261109843, + "grad_norm": 2.181717872619629, + "learning_rate": 0.0001875162968786408, + "loss": 0.2927, + "step": 6866 + }, + { + "epoch": 0.16097329418590342, + "grad_norm": 2.098580837249756, + "learning_rate": 0.00018751273268593867, + "loss": 0.3118, + "step": 6867 + }, + { + "epoch": 0.16099673576070841, + "grad_norm": 3.3717024326324463, + "learning_rate": 0.00018750916801838998, + "loss": 0.4163, + "step": 6868 + }, + { + "epoch": 0.1610201773355134, + "grad_norm": 3.733100652694702, + "learning_rate": 0.00018750560287601409, + "loss": 0.477, + "step": 6869 + }, + { + "epoch": 0.1610436189103184, + "grad_norm": 3.558356285095215, + "learning_rate": 0.00018750203725883036, + "loss": 0.495, + "step": 6870 + }, + { + "epoch": 0.1610670604851234, + "grad_norm": 0.839953601360321, + "learning_rate": 0.00018749847116685808, + "loss": 0.0654, + "step": 6871 + }, + { + "epoch": 0.1610905020599284, + "grad_norm": 1.5791326761245728, + "learning_rate": 0.00018749490460011665, + "loss": 0.3197, + "step": 6872 + }, + { + "epoch": 0.1611139436347334, + "grad_norm": 4.123174667358398, + "learning_rate": 0.0001874913375586254, + "loss": 0.557, + "step": 6873 + }, + { + "epoch": 0.16113738520953838, + "grad_norm": 3.3931047916412354, + "learning_rate": 0.0001874877700424037, + "loss": 2.0629, + "step": 6874 + }, + { + "epoch": 0.16116082678434338, + "grad_norm": 1.3071175813674927, + "learning_rate": 0.00018748420205147087, + "loss": 0.2041, + "step": 6875 + }, + { + "epoch": 0.16118426835914837, + "grad_norm": 5.918080806732178, + "learning_rate": 0.00018748063358584635, + "loss": 0.4723, + "step": 6876 + }, + { + "epoch": 0.16120770993395336, + "grad_norm": 4.49368143081665, + "learning_rate": 0.00018747706464554943, + "loss": 0.4512, + "step": 6877 + }, + { + "epoch": 0.16123115150875836, + "grad_norm": 4.624247074127197, + "learning_rate": 0.00018747349523059952, + "loss": 0.3482, + "step": 6878 + }, + { + "epoch": 0.16125459308356335, + "grad_norm": 4.965755939483643, + "learning_rate": 0.00018746992534101592, + "loss": 0.6019, + "step": 6879 + }, + { + "epoch": 0.16127803465836835, + "grad_norm": 2.7878384590148926, + "learning_rate": 0.00018746635497681808, + "loss": 0.2913, + "step": 6880 + }, + { + "epoch": 0.16130147623317334, + "grad_norm": 3.7184555530548096, + "learning_rate": 0.00018746278413802532, + "loss": 0.5354, + "step": 6881 + }, + { + "epoch": 0.16132491780797834, + "grad_norm": 4.3441267013549805, + "learning_rate": 0.00018745921282465705, + "loss": 0.3029, + "step": 6882 + }, + { + "epoch": 0.16134835938278333, + "grad_norm": 4.189855575561523, + "learning_rate": 0.00018745564103673262, + "loss": 0.5566, + "step": 6883 + }, + { + "epoch": 0.16137180095758832, + "grad_norm": 0.6384453773498535, + "learning_rate": 0.00018745206877427141, + "loss": 0.0851, + "step": 6884 + }, + { + "epoch": 0.16139524253239332, + "grad_norm": 3.209728479385376, + "learning_rate": 0.0001874484960372928, + "loss": 0.4942, + "step": 6885 + }, + { + "epoch": 0.1614186841071983, + "grad_norm": 8.003084182739258, + "learning_rate": 0.00018744492282581624, + "loss": 0.5362, + "step": 6886 + }, + { + "epoch": 0.1614421256820033, + "grad_norm": 2.4146273136138916, + "learning_rate": 0.00018744134913986106, + "loss": 0.503, + "step": 6887 + }, + { + "epoch": 0.1614655672568083, + "grad_norm": 3.5257060527801514, + "learning_rate": 0.00018743777497944665, + "loss": 0.5158, + "step": 6888 + }, + { + "epoch": 0.1614890088316133, + "grad_norm": 3.71964693069458, + "learning_rate": 0.00018743420034459245, + "loss": 0.6539, + "step": 6889 + }, + { + "epoch": 0.1615124504064183, + "grad_norm": 4.377289295196533, + "learning_rate": 0.00018743062523531777, + "loss": 0.8405, + "step": 6890 + }, + { + "epoch": 0.1615358919812233, + "grad_norm": 2.169290781021118, + "learning_rate": 0.00018742704965164208, + "loss": 0.2157, + "step": 6891 + }, + { + "epoch": 0.1615593335560283, + "grad_norm": 3.3647398948669434, + "learning_rate": 0.00018742347359358477, + "loss": 0.6681, + "step": 6892 + }, + { + "epoch": 0.1615827751308333, + "grad_norm": 3.7663025856018066, + "learning_rate": 0.0001874198970611652, + "loss": 0.4549, + "step": 6893 + }, + { + "epoch": 0.1616062167056383, + "grad_norm": 3.9765355587005615, + "learning_rate": 0.00018741632005440283, + "loss": 0.5215, + "step": 6894 + }, + { + "epoch": 0.1616296582804433, + "grad_norm": 2.7390730381011963, + "learning_rate": 0.00018741274257331707, + "loss": 0.2915, + "step": 6895 + }, + { + "epoch": 0.16165309985524828, + "grad_norm": 2.3227932453155518, + "learning_rate": 0.0001874091646179273, + "loss": 1.5386, + "step": 6896 + }, + { + "epoch": 0.16167654143005328, + "grad_norm": 2.7468347549438477, + "learning_rate": 0.0001874055861882529, + "loss": 0.2982, + "step": 6897 + }, + { + "epoch": 0.16169998300485827, + "grad_norm": 1.704259991645813, + "learning_rate": 0.0001874020072843134, + "loss": 0.0928, + "step": 6898 + }, + { + "epoch": 0.16172342457966327, + "grad_norm": 3.6776299476623535, + "learning_rate": 0.00018739842790612812, + "loss": 0.7476, + "step": 6899 + }, + { + "epoch": 0.16174686615446826, + "grad_norm": 4.18056058883667, + "learning_rate": 0.00018739484805371651, + "loss": 0.4708, + "step": 6900 + }, + { + "epoch": 0.16177030772927325, + "grad_norm": 1.1599581241607666, + "learning_rate": 0.000187391267727098, + "loss": 0.1465, + "step": 6901 + }, + { + "epoch": 0.16179374930407825, + "grad_norm": 4.009170055389404, + "learning_rate": 0.000187387686926292, + "loss": 0.4872, + "step": 6902 + }, + { + "epoch": 0.16181719087888324, + "grad_norm": 4.803140163421631, + "learning_rate": 0.00018738410565131798, + "loss": 0.6563, + "step": 6903 + }, + { + "epoch": 0.16184063245368824, + "grad_norm": 3.0724565982818604, + "learning_rate": 0.00018738052390219535, + "loss": 0.2323, + "step": 6904 + }, + { + "epoch": 0.16186407402849323, + "grad_norm": 2.26539945602417, + "learning_rate": 0.00018737694167894354, + "loss": 0.2699, + "step": 6905 + }, + { + "epoch": 0.16188751560329823, + "grad_norm": 3.430572271347046, + "learning_rate": 0.00018737335898158194, + "loss": 0.4696, + "step": 6906 + }, + { + "epoch": 0.16191095717810322, + "grad_norm": 0.8938559293746948, + "learning_rate": 0.0001873697758101301, + "loss": 0.1397, + "step": 6907 + }, + { + "epoch": 0.16193439875290822, + "grad_norm": 4.294753551483154, + "learning_rate": 0.00018736619216460737, + "loss": 0.6782, + "step": 6908 + }, + { + "epoch": 0.1619578403277132, + "grad_norm": 3.333498477935791, + "learning_rate": 0.00018736260804503321, + "loss": 0.3379, + "step": 6909 + }, + { + "epoch": 0.1619812819025182, + "grad_norm": 4.846035957336426, + "learning_rate": 0.00018735902345142709, + "loss": 0.3481, + "step": 6910 + }, + { + "epoch": 0.1620047234773232, + "grad_norm": 2.493544340133667, + "learning_rate": 0.00018735543838380845, + "loss": 1.2361, + "step": 6911 + }, + { + "epoch": 0.1620281650521282, + "grad_norm": 1.497177243232727, + "learning_rate": 0.00018735185284219675, + "loss": 0.2227, + "step": 6912 + }, + { + "epoch": 0.1620516066269332, + "grad_norm": 0.46453857421875, + "learning_rate": 0.00018734826682661147, + "loss": 0.0521, + "step": 6913 + }, + { + "epoch": 0.16207504820173818, + "grad_norm": 8.20438289642334, + "learning_rate": 0.000187344680337072, + "loss": 0.8276, + "step": 6914 + }, + { + "epoch": 0.16209848977654318, + "grad_norm": 5.854777812957764, + "learning_rate": 0.00018734109337359786, + "loss": 0.4966, + "step": 6915 + }, + { + "epoch": 0.1621219313513482, + "grad_norm": 0.8763144612312317, + "learning_rate": 0.0001873375059362085, + "loss": 0.0764, + "step": 6916 + }, + { + "epoch": 0.1621453729261532, + "grad_norm": 3.0119872093200684, + "learning_rate": 0.00018733391802492334, + "loss": 0.2895, + "step": 6917 + }, + { + "epoch": 0.16216881450095819, + "grad_norm": 5.2372870445251465, + "learning_rate": 0.0001873303296397619, + "loss": 0.6782, + "step": 6918 + }, + { + "epoch": 0.16219225607576318, + "grad_norm": 4.671289443969727, + "learning_rate": 0.00018732674078074366, + "loss": 0.4938, + "step": 6919 + }, + { + "epoch": 0.16221569765056817, + "grad_norm": 4.839365005493164, + "learning_rate": 0.00018732315144788804, + "loss": 0.4346, + "step": 6920 + }, + { + "epoch": 0.16223913922537317, + "grad_norm": 5.218401908874512, + "learning_rate": 0.00018731956164121455, + "loss": 0.7668, + "step": 6921 + }, + { + "epoch": 0.16226258080017816, + "grad_norm": 4.269512176513672, + "learning_rate": 0.00018731597136074267, + "loss": 0.7513, + "step": 6922 + }, + { + "epoch": 0.16228602237498316, + "grad_norm": 5.3519721031188965, + "learning_rate": 0.00018731238060649185, + "loss": 0.4898, + "step": 6923 + }, + { + "epoch": 0.16230946394978815, + "grad_norm": 4.103187084197998, + "learning_rate": 0.00018730878937848165, + "loss": 0.5545, + "step": 6924 + }, + { + "epoch": 0.16233290552459315, + "grad_norm": 4.444677829742432, + "learning_rate": 0.00018730519767673148, + "loss": 0.5723, + "step": 6925 + }, + { + "epoch": 0.16235634709939814, + "grad_norm": 3.127633810043335, + "learning_rate": 0.00018730160550126083, + "loss": 0.2688, + "step": 6926 + }, + { + "epoch": 0.16237978867420313, + "grad_norm": 3.034205675125122, + "learning_rate": 0.00018729801285208923, + "loss": 0.3294, + "step": 6927 + }, + { + "epoch": 0.16240323024900813, + "grad_norm": 4.457058429718018, + "learning_rate": 0.00018729441972923616, + "loss": 0.6851, + "step": 6928 + }, + { + "epoch": 0.16242667182381312, + "grad_norm": 2.739057779312134, + "learning_rate": 0.0001872908261327211, + "loss": 1.954, + "step": 6929 + }, + { + "epoch": 0.16245011339861812, + "grad_norm": 0.840255618095398, + "learning_rate": 0.0001872872320625636, + "loss": 0.077, + "step": 6930 + }, + { + "epoch": 0.1624735549734231, + "grad_norm": 4.114185810089111, + "learning_rate": 0.00018728363751878307, + "loss": 0.6475, + "step": 6931 + }, + { + "epoch": 0.1624969965482281, + "grad_norm": 3.3610165119171143, + "learning_rate": 0.0001872800425013991, + "loss": 0.4137, + "step": 6932 + }, + { + "epoch": 0.1625204381230331, + "grad_norm": 2.098424196243286, + "learning_rate": 0.00018727644701043117, + "loss": 0.2043, + "step": 6933 + }, + { + "epoch": 0.1625438796978381, + "grad_norm": 4.284010887145996, + "learning_rate": 0.0001872728510458988, + "loss": 0.7643, + "step": 6934 + }, + { + "epoch": 0.1625673212726431, + "grad_norm": 0.9566344022750854, + "learning_rate": 0.00018726925460782145, + "loss": 0.1232, + "step": 6935 + }, + { + "epoch": 0.16259076284744808, + "grad_norm": 3.242241144180298, + "learning_rate": 0.00018726565769621868, + "loss": 0.4707, + "step": 6936 + }, + { + "epoch": 0.16261420442225308, + "grad_norm": 1.823304533958435, + "learning_rate": 0.00018726206031111003, + "loss": 0.3122, + "step": 6937 + }, + { + "epoch": 0.16263764599705807, + "grad_norm": 2.9208977222442627, + "learning_rate": 0.00018725846245251494, + "loss": 1.8758, + "step": 6938 + }, + { + "epoch": 0.16266108757186307, + "grad_norm": 3.1351263523101807, + "learning_rate": 0.00018725486412045301, + "loss": 0.4872, + "step": 6939 + }, + { + "epoch": 0.16268452914666806, + "grad_norm": 4.463995933532715, + "learning_rate": 0.00018725126531494372, + "loss": 0.4037, + "step": 6940 + }, + { + "epoch": 0.16270797072147306, + "grad_norm": 4.918294906616211, + "learning_rate": 0.0001872476660360066, + "loss": 0.5813, + "step": 6941 + }, + { + "epoch": 0.16273141229627808, + "grad_norm": 4.974679946899414, + "learning_rate": 0.0001872440662836612, + "loss": 0.6876, + "step": 6942 + }, + { + "epoch": 0.16275485387108307, + "grad_norm": 3.324361562728882, + "learning_rate": 0.00018724046605792706, + "loss": 0.5673, + "step": 6943 + }, + { + "epoch": 0.16277829544588807, + "grad_norm": 2.804893732070923, + "learning_rate": 0.0001872368653588237, + "loss": 0.5229, + "step": 6944 + }, + { + "epoch": 0.16280173702069306, + "grad_norm": 4.696619987487793, + "learning_rate": 0.00018723326418637064, + "loss": 0.5152, + "step": 6945 + }, + { + "epoch": 0.16282517859549805, + "grad_norm": 4.495310306549072, + "learning_rate": 0.00018722966254058744, + "loss": 0.4101, + "step": 6946 + }, + { + "epoch": 0.16284862017030305, + "grad_norm": 2.9609272480010986, + "learning_rate": 0.00018722606042149363, + "loss": 0.6334, + "step": 6947 + }, + { + "epoch": 0.16287206174510804, + "grad_norm": 4.309670925140381, + "learning_rate": 0.00018722245782910876, + "loss": 0.5945, + "step": 6948 + }, + { + "epoch": 0.16289550331991304, + "grad_norm": 1.6093270778656006, + "learning_rate": 0.0001872188547634524, + "loss": 0.1609, + "step": 6949 + }, + { + "epoch": 0.16291894489471803, + "grad_norm": 5.455915451049805, + "learning_rate": 0.00018721525122454406, + "loss": 0.4248, + "step": 6950 + }, + { + "epoch": 0.16294238646952303, + "grad_norm": 3.283003568649292, + "learning_rate": 0.00018721164721240333, + "loss": 0.3353, + "step": 6951 + }, + { + "epoch": 0.16296582804432802, + "grad_norm": 4.1042962074279785, + "learning_rate": 0.00018720804272704978, + "loss": 0.4195, + "step": 6952 + }, + { + "epoch": 0.16298926961913301, + "grad_norm": 5.963765621185303, + "learning_rate": 0.0001872044377685029, + "loss": 0.5949, + "step": 6953 + }, + { + "epoch": 0.163012711193938, + "grad_norm": 1.2336176633834839, + "learning_rate": 0.00018720083233678228, + "loss": 0.2107, + "step": 6954 + }, + { + "epoch": 0.163036152768743, + "grad_norm": 3.1301770210266113, + "learning_rate": 0.00018719722643190752, + "loss": 0.4436, + "step": 6955 + }, + { + "epoch": 0.163059594343548, + "grad_norm": 3.5253782272338867, + "learning_rate": 0.00018719362005389814, + "loss": 0.6934, + "step": 6956 + }, + { + "epoch": 0.163083035918353, + "grad_norm": 2.8185133934020996, + "learning_rate": 0.00018719001320277373, + "loss": 0.4963, + "step": 6957 + }, + { + "epoch": 0.163106477493158, + "grad_norm": 5.208103656768799, + "learning_rate": 0.00018718640587855385, + "loss": 0.7221, + "step": 6958 + }, + { + "epoch": 0.16312991906796298, + "grad_norm": 3.306638240814209, + "learning_rate": 0.0001871827980812581, + "loss": 0.4914, + "step": 6959 + }, + { + "epoch": 0.16315336064276798, + "grad_norm": 1.6654868125915527, + "learning_rate": 0.000187179189810906, + "loss": 0.2949, + "step": 6960 + }, + { + "epoch": 0.16317680221757297, + "grad_norm": 5.097724437713623, + "learning_rate": 0.0001871755810675172, + "loss": 0.6775, + "step": 6961 + }, + { + "epoch": 0.16320024379237796, + "grad_norm": 5.385200500488281, + "learning_rate": 0.00018717197185111125, + "loss": 0.6286, + "step": 6962 + }, + { + "epoch": 0.16322368536718296, + "grad_norm": 1.9677852392196655, + "learning_rate": 0.0001871683621617077, + "loss": 1.5149, + "step": 6963 + }, + { + "epoch": 0.16324712694198795, + "grad_norm": 1.8778945207595825, + "learning_rate": 0.00018716475199932617, + "loss": 0.2865, + "step": 6964 + }, + { + "epoch": 0.16327056851679295, + "grad_norm": 4.286211013793945, + "learning_rate": 0.00018716114136398625, + "loss": 0.4451, + "step": 6965 + }, + { + "epoch": 0.16329401009159794, + "grad_norm": 0.694399893283844, + "learning_rate": 0.00018715753025570756, + "loss": 0.0584, + "step": 6966 + }, + { + "epoch": 0.16331745166640294, + "grad_norm": 1.0929183959960938, + "learning_rate": 0.0001871539186745096, + "loss": 0.1705, + "step": 6967 + }, + { + "epoch": 0.16334089324120796, + "grad_norm": 3.213071584701538, + "learning_rate": 0.00018715030662041205, + "loss": 1.4741, + "step": 6968 + }, + { + "epoch": 0.16336433481601295, + "grad_norm": 1.1215240955352783, + "learning_rate": 0.00018714669409343452, + "loss": 0.1743, + "step": 6969 + }, + { + "epoch": 0.16338777639081795, + "grad_norm": 3.27016019821167, + "learning_rate": 0.00018714308109359654, + "loss": 0.5405, + "step": 6970 + }, + { + "epoch": 0.16341121796562294, + "grad_norm": 2.603362798690796, + "learning_rate": 0.00018713946762091778, + "loss": 0.2472, + "step": 6971 + }, + { + "epoch": 0.16343465954042793, + "grad_norm": 2.4548838138580322, + "learning_rate": 0.0001871358536754178, + "loss": 1.5991, + "step": 6972 + }, + { + "epoch": 0.16345810111523293, + "grad_norm": 1.9304916858673096, + "learning_rate": 0.00018713223925711622, + "loss": 0.1358, + "step": 6973 + }, + { + "epoch": 0.16348154269003792, + "grad_norm": 3.183286190032959, + "learning_rate": 0.00018712862436603267, + "loss": 0.7239, + "step": 6974 + }, + { + "epoch": 0.16350498426484292, + "grad_norm": 3.4457051753997803, + "learning_rate": 0.00018712500900218678, + "loss": 0.2288, + "step": 6975 + }, + { + "epoch": 0.1635284258396479, + "grad_norm": 3.910534143447876, + "learning_rate": 0.0001871213931655981, + "loss": 0.4933, + "step": 6976 + }, + { + "epoch": 0.1635518674144529, + "grad_norm": 4.5530009269714355, + "learning_rate": 0.00018711777685628632, + "loss": 0.7751, + "step": 6977 + }, + { + "epoch": 0.1635753089892579, + "grad_norm": 5.91217565536499, + "learning_rate": 0.000187114160074271, + "loss": 0.69, + "step": 6978 + }, + { + "epoch": 0.1635987505640629, + "grad_norm": 6.983726501464844, + "learning_rate": 0.0001871105428195718, + "loss": 0.9242, + "step": 6979 + }, + { + "epoch": 0.1636221921388679, + "grad_norm": 3.7416703701019287, + "learning_rate": 0.00018710692509220838, + "loss": 0.4434, + "step": 6980 + }, + { + "epoch": 0.16364563371367288, + "grad_norm": 4.456112384796143, + "learning_rate": 0.0001871033068922003, + "loss": 0.5297, + "step": 6981 + }, + { + "epoch": 0.16366907528847788, + "grad_norm": 1.0446851253509521, + "learning_rate": 0.00018709968821956725, + "loss": 0.1449, + "step": 6982 + }, + { + "epoch": 0.16369251686328287, + "grad_norm": 3.880063533782959, + "learning_rate": 0.00018709606907432887, + "loss": 0.4546, + "step": 6983 + }, + { + "epoch": 0.16371595843808787, + "grad_norm": 1.4936050176620483, + "learning_rate": 0.00018709244945650473, + "loss": 0.1811, + "step": 6984 + }, + { + "epoch": 0.16373940001289286, + "grad_norm": 3.2328433990478516, + "learning_rate": 0.00018708882936611452, + "loss": 0.3689, + "step": 6985 + }, + { + "epoch": 0.16376284158769785, + "grad_norm": 4.271566867828369, + "learning_rate": 0.00018708520880317785, + "loss": 0.5428, + "step": 6986 + }, + { + "epoch": 0.16378628316250285, + "grad_norm": 3.3595516681671143, + "learning_rate": 0.00018708158776771443, + "loss": 0.4844, + "step": 6987 + }, + { + "epoch": 0.16380972473730784, + "grad_norm": 4.205976963043213, + "learning_rate": 0.00018707796625974383, + "loss": 0.6311, + "step": 6988 + }, + { + "epoch": 0.16383316631211284, + "grad_norm": 0.8122702836990356, + "learning_rate": 0.00018707434427928576, + "loss": 0.3642, + "step": 6989 + }, + { + "epoch": 0.16385660788691783, + "grad_norm": 1.8268462419509888, + "learning_rate": 0.00018707072182635984, + "loss": 0.2529, + "step": 6990 + }, + { + "epoch": 0.16388004946172283, + "grad_norm": 3.1466333866119385, + "learning_rate": 0.00018706709890098573, + "loss": 1.591, + "step": 6991 + }, + { + "epoch": 0.16390349103652782, + "grad_norm": 3.944744110107422, + "learning_rate": 0.00018706347550318311, + "loss": 0.5416, + "step": 6992 + }, + { + "epoch": 0.16392693261133284, + "grad_norm": 4.2694549560546875, + "learning_rate": 0.0001870598516329716, + "loss": 0.3875, + "step": 6993 + }, + { + "epoch": 0.16395037418613784, + "grad_norm": 3.349848508834839, + "learning_rate": 0.0001870562272903709, + "loss": 0.4899, + "step": 6994 + }, + { + "epoch": 0.16397381576094283, + "grad_norm": 1.6887520551681519, + "learning_rate": 0.00018705260247540067, + "loss": 0.3133, + "step": 6995 + }, + { + "epoch": 0.16399725733574783, + "grad_norm": 0.910698413848877, + "learning_rate": 0.00018704897718808058, + "loss": 0.0921, + "step": 6996 + }, + { + "epoch": 0.16402069891055282, + "grad_norm": 1.5491026639938354, + "learning_rate": 0.00018704535142843027, + "loss": 0.156, + "step": 6997 + }, + { + "epoch": 0.16404414048535781, + "grad_norm": 4.106495380401611, + "learning_rate": 0.00018704172519646946, + "loss": 0.5956, + "step": 6998 + }, + { + "epoch": 0.1640675820601628, + "grad_norm": 3.5482420921325684, + "learning_rate": 0.0001870380984922178, + "loss": 1.7721, + "step": 6999 + }, + { + "epoch": 0.1640910236349678, + "grad_norm": 2.202728509902954, + "learning_rate": 0.00018703447131569493, + "loss": 0.2526, + "step": 7000 + }, + { + "epoch": 0.1641144652097728, + "grad_norm": 4.7902069091796875, + "learning_rate": 0.00018703084366692062, + "loss": 0.4175, + "step": 7001 + }, + { + "epoch": 0.1641379067845778, + "grad_norm": 0.8958625197410583, + "learning_rate": 0.00018702721554591446, + "loss": 0.0634, + "step": 7002 + }, + { + "epoch": 0.16416134835938279, + "grad_norm": 3.5156803131103516, + "learning_rate": 0.00018702358695269622, + "loss": 0.4722, + "step": 7003 + }, + { + "epoch": 0.16418478993418778, + "grad_norm": 1.207939863204956, + "learning_rate": 0.00018701995788728552, + "loss": 0.1625, + "step": 7004 + }, + { + "epoch": 0.16420823150899277, + "grad_norm": 2.4723801612854004, + "learning_rate": 0.00018701632834970208, + "loss": 0.1792, + "step": 7005 + }, + { + "epoch": 0.16423167308379777, + "grad_norm": 2.8807578086853027, + "learning_rate": 0.00018701269833996565, + "loss": 0.3372, + "step": 7006 + }, + { + "epoch": 0.16425511465860276, + "grad_norm": 3.5638375282287598, + "learning_rate": 0.00018700906785809582, + "loss": 0.3561, + "step": 7007 + }, + { + "epoch": 0.16427855623340776, + "grad_norm": 4.726449966430664, + "learning_rate": 0.00018700543690411235, + "loss": 0.5216, + "step": 7008 + }, + { + "epoch": 0.16430199780821275, + "grad_norm": 0.7601968050003052, + "learning_rate": 0.00018700180547803495, + "loss": 0.1598, + "step": 7009 + }, + { + "epoch": 0.16432543938301775, + "grad_norm": 4.342578887939453, + "learning_rate": 0.00018699817357988328, + "loss": 0.4515, + "step": 7010 + }, + { + "epoch": 0.16434888095782274, + "grad_norm": 2.869462728500366, + "learning_rate": 0.0001869945412096771, + "loss": 0.26, + "step": 7011 + }, + { + "epoch": 0.16437232253262773, + "grad_norm": 3.874112129211426, + "learning_rate": 0.00018699090836743606, + "loss": 0.3625, + "step": 7012 + }, + { + "epoch": 0.16439576410743273, + "grad_norm": 4.854094505310059, + "learning_rate": 0.00018698727505317994, + "loss": 0.6489, + "step": 7013 + }, + { + "epoch": 0.16441920568223772, + "grad_norm": 3.9871861934661865, + "learning_rate": 0.00018698364126692842, + "loss": 0.3481, + "step": 7014 + }, + { + "epoch": 0.16444264725704272, + "grad_norm": 8.303004264831543, + "learning_rate": 0.00018698000700870118, + "loss": 0.674, + "step": 7015 + }, + { + "epoch": 0.1644660888318477, + "grad_norm": 4.433371543884277, + "learning_rate": 0.00018697637227851804, + "loss": 0.7314, + "step": 7016 + }, + { + "epoch": 0.1644895304066527, + "grad_norm": 3.984304428100586, + "learning_rate": 0.0001869727370763986, + "loss": 0.3211, + "step": 7017 + }, + { + "epoch": 0.1645129719814577, + "grad_norm": 4.351452827453613, + "learning_rate": 0.00018696910140236268, + "loss": 0.1888, + "step": 7018 + }, + { + "epoch": 0.16453641355626272, + "grad_norm": 4.498950481414795, + "learning_rate": 0.00018696546525642994, + "loss": 0.5012, + "step": 7019 + }, + { + "epoch": 0.16455985513106772, + "grad_norm": 3.7338755130767822, + "learning_rate": 0.00018696182863862015, + "loss": 0.4258, + "step": 7020 + }, + { + "epoch": 0.1645832967058727, + "grad_norm": 1.144775629043579, + "learning_rate": 0.00018695819154895303, + "loss": 0.1841, + "step": 7021 + }, + { + "epoch": 0.1646067382806777, + "grad_norm": 2.922595262527466, + "learning_rate": 0.00018695455398744834, + "loss": 0.6076, + "step": 7022 + }, + { + "epoch": 0.1646301798554827, + "grad_norm": 6.27885627746582, + "learning_rate": 0.00018695091595412577, + "loss": 0.9137, + "step": 7023 + }, + { + "epoch": 0.1646536214302877, + "grad_norm": 1.1327990293502808, + "learning_rate": 0.0001869472774490051, + "loss": 0.1333, + "step": 7024 + }, + { + "epoch": 0.1646770630050927, + "grad_norm": 3.4738895893096924, + "learning_rate": 0.00018694363847210607, + "loss": 0.6241, + "step": 7025 + }, + { + "epoch": 0.16470050457989768, + "grad_norm": 3.1582465171813965, + "learning_rate": 0.0001869399990234484, + "loss": 0.2545, + "step": 7026 + }, + { + "epoch": 0.16472394615470268, + "grad_norm": 4.065396785736084, + "learning_rate": 0.00018693635910305183, + "loss": 0.4401, + "step": 7027 + }, + { + "epoch": 0.16474738772950767, + "grad_norm": 2.1092159748077393, + "learning_rate": 0.00018693271871093615, + "loss": 0.1659, + "step": 7028 + }, + { + "epoch": 0.16477082930431267, + "grad_norm": 4.399582386016846, + "learning_rate": 0.00018692907784712107, + "loss": 0.9717, + "step": 7029 + }, + { + "epoch": 0.16479427087911766, + "grad_norm": 0.6296824216842651, + "learning_rate": 0.00018692543651162638, + "loss": 0.101, + "step": 7030 + }, + { + "epoch": 0.16481771245392265, + "grad_norm": 3.843524694442749, + "learning_rate": 0.00018692179470447187, + "loss": 0.5356, + "step": 7031 + }, + { + "epoch": 0.16484115402872765, + "grad_norm": 3.8372132778167725, + "learning_rate": 0.0001869181524256772, + "loss": 0.4353, + "step": 7032 + }, + { + "epoch": 0.16486459560353264, + "grad_norm": 2.809476613998413, + "learning_rate": 0.00018691450967526222, + "loss": 0.333, + "step": 7033 + }, + { + "epoch": 0.16488803717833764, + "grad_norm": 4.974763870239258, + "learning_rate": 0.00018691086645324667, + "loss": 0.3347, + "step": 7034 + }, + { + "epoch": 0.16491147875314263, + "grad_norm": 4.01249885559082, + "learning_rate": 0.00018690722275965032, + "loss": 0.414, + "step": 7035 + }, + { + "epoch": 0.16493492032794763, + "grad_norm": 3.911670446395874, + "learning_rate": 0.0001869035785944929, + "loss": 0.4399, + "step": 7036 + }, + { + "epoch": 0.16495836190275262, + "grad_norm": 1.4770286083221436, + "learning_rate": 0.00018689993395779424, + "loss": 0.1802, + "step": 7037 + }, + { + "epoch": 0.16498180347755761, + "grad_norm": 1.2887388467788696, + "learning_rate": 0.00018689628884957412, + "loss": 0.2125, + "step": 7038 + }, + { + "epoch": 0.1650052450523626, + "grad_norm": 4.4017839431762695, + "learning_rate": 0.00018689264326985226, + "loss": 0.5578, + "step": 7039 + }, + { + "epoch": 0.1650286866271676, + "grad_norm": 0.9110023975372314, + "learning_rate": 0.0001868889972186485, + "loss": 0.0801, + "step": 7040 + }, + { + "epoch": 0.1650521282019726, + "grad_norm": 3.1139626502990723, + "learning_rate": 0.00018688535069598259, + "loss": 0.3039, + "step": 7041 + }, + { + "epoch": 0.1650755697767776, + "grad_norm": 3.014309883117676, + "learning_rate": 0.0001868817037018743, + "loss": 0.2889, + "step": 7042 + }, + { + "epoch": 0.1650990113515826, + "grad_norm": 4.818223476409912, + "learning_rate": 0.00018687805623634346, + "loss": 0.5419, + "step": 7043 + }, + { + "epoch": 0.1651224529263876, + "grad_norm": 3.281452178955078, + "learning_rate": 0.00018687440829940984, + "loss": 0.4203, + "step": 7044 + }, + { + "epoch": 0.1651458945011926, + "grad_norm": 0.7974584102630615, + "learning_rate": 0.00018687075989109325, + "loss": 0.1597, + "step": 7045 + }, + { + "epoch": 0.1651693360759976, + "grad_norm": 6.263547897338867, + "learning_rate": 0.00018686711101141348, + "loss": 0.8479, + "step": 7046 + }, + { + "epoch": 0.1651927776508026, + "grad_norm": 3.076540231704712, + "learning_rate": 0.00018686346166039029, + "loss": 0.3394, + "step": 7047 + }, + { + "epoch": 0.16521621922560759, + "grad_norm": 3.097778081893921, + "learning_rate": 0.00018685981183804354, + "loss": 0.3932, + "step": 7048 + }, + { + "epoch": 0.16523966080041258, + "grad_norm": 3.591061592102051, + "learning_rate": 0.00018685616154439302, + "loss": 0.3351, + "step": 7049 + }, + { + "epoch": 0.16526310237521757, + "grad_norm": 1.7501376867294312, + "learning_rate": 0.0001868525107794585, + "loss": 0.2766, + "step": 7050 + }, + { + "epoch": 0.16528654395002257, + "grad_norm": 2.6465370655059814, + "learning_rate": 0.00018684885954325982, + "loss": 0.333, + "step": 7051 + }, + { + "epoch": 0.16530998552482756, + "grad_norm": 3.8131942749023438, + "learning_rate": 0.00018684520783581677, + "loss": 0.6106, + "step": 7052 + }, + { + "epoch": 0.16533342709963256, + "grad_norm": 4.30000638961792, + "learning_rate": 0.0001868415556571492, + "loss": 0.5869, + "step": 7053 + }, + { + "epoch": 0.16535686867443755, + "grad_norm": 5.31809139251709, + "learning_rate": 0.00018683790300727692, + "loss": 0.7507, + "step": 7054 + }, + { + "epoch": 0.16538031024924255, + "grad_norm": 2.6059768199920654, + "learning_rate": 0.00018683424988621972, + "loss": 0.2705, + "step": 7055 + }, + { + "epoch": 0.16540375182404754, + "grad_norm": 1.401807427406311, + "learning_rate": 0.00018683059629399744, + "loss": 0.1749, + "step": 7056 + }, + { + "epoch": 0.16542719339885253, + "grad_norm": 2.927190065383911, + "learning_rate": 0.0001868269422306299, + "loss": 1.7751, + "step": 7057 + }, + { + "epoch": 0.16545063497365753, + "grad_norm": 2.6553447246551514, + "learning_rate": 0.00018682328769613693, + "loss": 0.3561, + "step": 7058 + }, + { + "epoch": 0.16547407654846252, + "grad_norm": 2.1632659435272217, + "learning_rate": 0.0001868196326905384, + "loss": 0.1975, + "step": 7059 + }, + { + "epoch": 0.16549751812326752, + "grad_norm": 2.8363728523254395, + "learning_rate": 0.00018681597721385403, + "loss": 0.3647, + "step": 7060 + }, + { + "epoch": 0.1655209596980725, + "grad_norm": 2.465811014175415, + "learning_rate": 0.00018681232126610377, + "loss": 0.275, + "step": 7061 + }, + { + "epoch": 0.1655444012728775, + "grad_norm": 0.7076861262321472, + "learning_rate": 0.00018680866484730742, + "loss": 0.0574, + "step": 7062 + }, + { + "epoch": 0.1655678428476825, + "grad_norm": 4.042705059051514, + "learning_rate": 0.00018680500795748478, + "loss": 0.3699, + "step": 7063 + }, + { + "epoch": 0.1655912844224875, + "grad_norm": 1.6024335622787476, + "learning_rate": 0.00018680135059665577, + "loss": 0.1993, + "step": 7064 + }, + { + "epoch": 0.1656147259972925, + "grad_norm": 3.663200616836548, + "learning_rate": 0.00018679769276484017, + "loss": 0.381, + "step": 7065 + }, + { + "epoch": 0.16563816757209748, + "grad_norm": 4.307976245880127, + "learning_rate": 0.00018679403446205783, + "loss": 0.5841, + "step": 7066 + }, + { + "epoch": 0.16566160914690248, + "grad_norm": 0.6960292458534241, + "learning_rate": 0.00018679037568832864, + "loss": 0.0566, + "step": 7067 + }, + { + "epoch": 0.16568505072170747, + "grad_norm": 4.294863700866699, + "learning_rate": 0.00018678671644367244, + "loss": 0.572, + "step": 7068 + }, + { + "epoch": 0.16570849229651247, + "grad_norm": 4.492240905761719, + "learning_rate": 0.00018678305672810902, + "loss": 0.622, + "step": 7069 + }, + { + "epoch": 0.1657319338713175, + "grad_norm": 5.514212131500244, + "learning_rate": 0.00018677939654165835, + "loss": 0.4101, + "step": 7070 + }, + { + "epoch": 0.16575537544612248, + "grad_norm": 4.704708576202393, + "learning_rate": 0.00018677573588434023, + "loss": 0.6557, + "step": 7071 + }, + { + "epoch": 0.16577881702092748, + "grad_norm": 4.79706335067749, + "learning_rate": 0.00018677207475617454, + "loss": 0.7464, + "step": 7072 + }, + { + "epoch": 0.16580225859573247, + "grad_norm": 4.0985331535339355, + "learning_rate": 0.0001867684131571811, + "loss": 0.7331, + "step": 7073 + }, + { + "epoch": 0.16582570017053747, + "grad_norm": 11.817131042480469, + "learning_rate": 0.0001867647510873798, + "loss": 0.7329, + "step": 7074 + }, + { + "epoch": 0.16584914174534246, + "grad_norm": 4.716949462890625, + "learning_rate": 0.00018676108854679056, + "loss": 0.4496, + "step": 7075 + }, + { + "epoch": 0.16587258332014745, + "grad_norm": 4.63118314743042, + "learning_rate": 0.00018675742553543319, + "loss": 0.8133, + "step": 7076 + }, + { + "epoch": 0.16589602489495245, + "grad_norm": 2.872573137283325, + "learning_rate": 0.0001867537620533276, + "loss": 0.4245, + "step": 7077 + }, + { + "epoch": 0.16591946646975744, + "grad_norm": 2.8804893493652344, + "learning_rate": 0.00018675009810049367, + "loss": 0.5344, + "step": 7078 + }, + { + "epoch": 0.16594290804456244, + "grad_norm": 3.004187822341919, + "learning_rate": 0.00018674643367695123, + "loss": 0.2589, + "step": 7079 + }, + { + "epoch": 0.16596634961936743, + "grad_norm": 3.4553353786468506, + "learning_rate": 0.0001867427687827202, + "loss": 0.4851, + "step": 7080 + }, + { + "epoch": 0.16598979119417243, + "grad_norm": 4.483937740325928, + "learning_rate": 0.00018673910341782053, + "loss": 0.5006, + "step": 7081 + }, + { + "epoch": 0.16601323276897742, + "grad_norm": 4.236230373382568, + "learning_rate": 0.00018673543758227198, + "loss": 0.3985, + "step": 7082 + }, + { + "epoch": 0.16603667434378241, + "grad_norm": 1.4578739404678345, + "learning_rate": 0.00018673177127609454, + "loss": 0.1824, + "step": 7083 + }, + { + "epoch": 0.1660601159185874, + "grad_norm": 1.2773596048355103, + "learning_rate": 0.00018672810449930807, + "loss": 0.1677, + "step": 7084 + }, + { + "epoch": 0.1660835574933924, + "grad_norm": 3.4752819538116455, + "learning_rate": 0.00018672443725193248, + "loss": 0.3226, + "step": 7085 + }, + { + "epoch": 0.1661069990681974, + "grad_norm": 2.946415662765503, + "learning_rate": 0.0001867207695339876, + "loss": 0.4971, + "step": 7086 + }, + { + "epoch": 0.1661304406430024, + "grad_norm": 4.343348026275635, + "learning_rate": 0.00018671710134549345, + "loss": 0.5339, + "step": 7087 + }, + { + "epoch": 0.16615388221780739, + "grad_norm": 4.245337963104248, + "learning_rate": 0.00018671343268646982, + "loss": 0.4312, + "step": 7088 + }, + { + "epoch": 0.16617732379261238, + "grad_norm": 3.64882493019104, + "learning_rate": 0.0001867097635569367, + "loss": 0.5221, + "step": 7089 + }, + { + "epoch": 0.16620076536741737, + "grad_norm": 1.6832886934280396, + "learning_rate": 0.00018670609395691394, + "loss": 0.2118, + "step": 7090 + }, + { + "epoch": 0.16622420694222237, + "grad_norm": 5.753878116607666, + "learning_rate": 0.0001867024238864215, + "loss": 0.9356, + "step": 7091 + }, + { + "epoch": 0.16624764851702736, + "grad_norm": 2.825453281402588, + "learning_rate": 0.00018669875334547927, + "loss": 0.4523, + "step": 7092 + }, + { + "epoch": 0.16627109009183236, + "grad_norm": 0.87437504529953, + "learning_rate": 0.00018669508233410716, + "loss": 0.1274, + "step": 7093 + }, + { + "epoch": 0.16629453166663735, + "grad_norm": 0.723307728767395, + "learning_rate": 0.0001866914108523251, + "loss": 0.0994, + "step": 7094 + }, + { + "epoch": 0.16631797324144237, + "grad_norm": 3.3801259994506836, + "learning_rate": 0.00018668773890015297, + "loss": 0.3636, + "step": 7095 + }, + { + "epoch": 0.16634141481624737, + "grad_norm": 3.316112995147705, + "learning_rate": 0.00018668406647761077, + "loss": 0.599, + "step": 7096 + }, + { + "epoch": 0.16636485639105236, + "grad_norm": 2.968672752380371, + "learning_rate": 0.0001866803935847184, + "loss": 0.3289, + "step": 7097 + }, + { + "epoch": 0.16638829796585736, + "grad_norm": 4.4015116691589355, + "learning_rate": 0.0001866767202214957, + "loss": 0.5363, + "step": 7098 + }, + { + "epoch": 0.16641173954066235, + "grad_norm": 1.8601229190826416, + "learning_rate": 0.00018667304638796277, + "loss": 1.3603, + "step": 7099 + }, + { + "epoch": 0.16643518111546735, + "grad_norm": 3.0682008266448975, + "learning_rate": 0.0001866693720841394, + "loss": 1.4126, + "step": 7100 + }, + { + "epoch": 0.16645862269027234, + "grad_norm": 2.701401948928833, + "learning_rate": 0.00018666569731004562, + "loss": 0.4551, + "step": 7101 + }, + { + "epoch": 0.16648206426507733, + "grad_norm": 2.84517502784729, + "learning_rate": 0.0001866620220657013, + "loss": 0.2665, + "step": 7102 + }, + { + "epoch": 0.16650550583988233, + "grad_norm": 2.8226242065429688, + "learning_rate": 0.0001866583463511264, + "loss": 0.424, + "step": 7103 + }, + { + "epoch": 0.16652894741468732, + "grad_norm": 4.662168502807617, + "learning_rate": 0.0001866546701663409, + "loss": 0.6681, + "step": 7104 + }, + { + "epoch": 0.16655238898949232, + "grad_norm": 2.4044029712677, + "learning_rate": 0.0001866509935113647, + "loss": 0.3098, + "step": 7105 + }, + { + "epoch": 0.1665758305642973, + "grad_norm": 4.376627445220947, + "learning_rate": 0.0001866473163862178, + "loss": 0.4481, + "step": 7106 + }, + { + "epoch": 0.1665992721391023, + "grad_norm": 1.2600878477096558, + "learning_rate": 0.00018664363879092013, + "loss": 0.1933, + "step": 7107 + }, + { + "epoch": 0.1666227137139073, + "grad_norm": 2.705554485321045, + "learning_rate": 0.00018663996072549162, + "loss": 1.5567, + "step": 7108 + }, + { + "epoch": 0.1666461552887123, + "grad_norm": 2.132108211517334, + "learning_rate": 0.00018663628218995224, + "loss": 0.2234, + "step": 7109 + }, + { + "epoch": 0.1666695968635173, + "grad_norm": 3.8066163063049316, + "learning_rate": 0.000186632603184322, + "loss": 0.3274, + "step": 7110 + }, + { + "epoch": 0.16669303843832228, + "grad_norm": 1.1996577978134155, + "learning_rate": 0.00018662892370862079, + "loss": 0.2323, + "step": 7111 + }, + { + "epoch": 0.16671648001312728, + "grad_norm": 3.806180715560913, + "learning_rate": 0.00018662524376286858, + "loss": 0.3823, + "step": 7112 + }, + { + "epoch": 0.16673992158793227, + "grad_norm": 5.0218000411987305, + "learning_rate": 0.0001866215633470854, + "loss": 0.6749, + "step": 7113 + }, + { + "epoch": 0.16676336316273727, + "grad_norm": 2.8253414630889893, + "learning_rate": 0.00018661788246129114, + "loss": 0.2895, + "step": 7114 + }, + { + "epoch": 0.16678680473754226, + "grad_norm": 4.232754707336426, + "learning_rate": 0.00018661420110550585, + "loss": 0.4867, + "step": 7115 + }, + { + "epoch": 0.16681024631234725, + "grad_norm": 5.078305244445801, + "learning_rate": 0.00018661051927974947, + "loss": 0.8513, + "step": 7116 + }, + { + "epoch": 0.16683368788715225, + "grad_norm": 3.351325511932373, + "learning_rate": 0.00018660683698404197, + "loss": 0.391, + "step": 7117 + }, + { + "epoch": 0.16685712946195724, + "grad_norm": 3.072930097579956, + "learning_rate": 0.00018660315421840332, + "loss": 0.3115, + "step": 7118 + }, + { + "epoch": 0.16688057103676224, + "grad_norm": 2.2935376167297363, + "learning_rate": 0.00018659947098285354, + "loss": 0.2366, + "step": 7119 + }, + { + "epoch": 0.16690401261156723, + "grad_norm": 4.13768196105957, + "learning_rate": 0.00018659578727741257, + "loss": 0.5804, + "step": 7120 + }, + { + "epoch": 0.16692745418637225, + "grad_norm": 1.547142505645752, + "learning_rate": 0.00018659210310210047, + "loss": 0.1605, + "step": 7121 + }, + { + "epoch": 0.16695089576117725, + "grad_norm": 1.0988801717758179, + "learning_rate": 0.00018658841845693716, + "loss": 0.1065, + "step": 7122 + }, + { + "epoch": 0.16697433733598224, + "grad_norm": 6.2959465980529785, + "learning_rate": 0.00018658473334194265, + "loss": 0.6278, + "step": 7123 + }, + { + "epoch": 0.16699777891078724, + "grad_norm": 3.2885591983795166, + "learning_rate": 0.00018658104775713693, + "loss": 0.4713, + "step": 7124 + }, + { + "epoch": 0.16702122048559223, + "grad_norm": 1.4174696207046509, + "learning_rate": 0.00018657736170254004, + "loss": 0.0709, + "step": 7125 + }, + { + "epoch": 0.16704466206039723, + "grad_norm": 5.515402793884277, + "learning_rate": 0.00018657367517817193, + "loss": 0.4533, + "step": 7126 + }, + { + "epoch": 0.16706810363520222, + "grad_norm": 4.289803504943848, + "learning_rate": 0.00018656998818405263, + "loss": 0.5431, + "step": 7127 + }, + { + "epoch": 0.16709154521000721, + "grad_norm": 3.8424057960510254, + "learning_rate": 0.00018656630072020215, + "loss": 0.6157, + "step": 7128 + }, + { + "epoch": 0.1671149867848122, + "grad_norm": 3.1432342529296875, + "learning_rate": 0.0001865626127866405, + "loss": 1.4664, + "step": 7129 + }, + { + "epoch": 0.1671384283596172, + "grad_norm": 5.748993396759033, + "learning_rate": 0.00018655892438338762, + "loss": 0.7363, + "step": 7130 + }, + { + "epoch": 0.1671618699344222, + "grad_norm": 2.8664121627807617, + "learning_rate": 0.00018655523551046362, + "loss": 1.469, + "step": 7131 + }, + { + "epoch": 0.1671853115092272, + "grad_norm": 1.506700038909912, + "learning_rate": 0.0001865515461678885, + "loss": 0.1578, + "step": 7132 + }, + { + "epoch": 0.16720875308403219, + "grad_norm": 3.0315732955932617, + "learning_rate": 0.0001865478563556822, + "loss": 0.2507, + "step": 7133 + }, + { + "epoch": 0.16723219465883718, + "grad_norm": 4.37895393371582, + "learning_rate": 0.00018654416607386486, + "loss": 0.4325, + "step": 7134 + }, + { + "epoch": 0.16725563623364217, + "grad_norm": 0.5983082056045532, + "learning_rate": 0.00018654047532245643, + "loss": 0.0674, + "step": 7135 + }, + { + "epoch": 0.16727907780844717, + "grad_norm": 1.0708576440811157, + "learning_rate": 0.00018653678410147692, + "loss": 0.1705, + "step": 7136 + }, + { + "epoch": 0.16730251938325216, + "grad_norm": 2.1433093547821045, + "learning_rate": 0.00018653309241094642, + "loss": 0.1779, + "step": 7137 + }, + { + "epoch": 0.16732596095805716, + "grad_norm": 4.927791118621826, + "learning_rate": 0.0001865294002508849, + "loss": 0.7145, + "step": 7138 + }, + { + "epoch": 0.16734940253286215, + "grad_norm": 4.683916091918945, + "learning_rate": 0.0001865257076213124, + "loss": 0.4359, + "step": 7139 + }, + { + "epoch": 0.16737284410766715, + "grad_norm": 1.5950886011123657, + "learning_rate": 0.00018652201452224905, + "loss": 0.1548, + "step": 7140 + }, + { + "epoch": 0.16739628568247214, + "grad_norm": 1.141335129737854, + "learning_rate": 0.00018651832095371474, + "loss": 0.1499, + "step": 7141 + }, + { + "epoch": 0.16741972725727713, + "grad_norm": 3.3990020751953125, + "learning_rate": 0.00018651462691572963, + "loss": 0.5328, + "step": 7142 + }, + { + "epoch": 0.16744316883208213, + "grad_norm": 4.633116722106934, + "learning_rate": 0.00018651093240831373, + "loss": 0.5099, + "step": 7143 + }, + { + "epoch": 0.16746661040688712, + "grad_norm": 4.823581218719482, + "learning_rate": 0.00018650723743148702, + "loss": 0.5405, + "step": 7144 + }, + { + "epoch": 0.16749005198169212, + "grad_norm": 1.0859545469284058, + "learning_rate": 0.00018650354198526968, + "loss": 0.1213, + "step": 7145 + }, + { + "epoch": 0.1675134935564971, + "grad_norm": 3.0252530574798584, + "learning_rate": 0.00018649984606968165, + "loss": 0.2144, + "step": 7146 + }, + { + "epoch": 0.16753693513130213, + "grad_norm": 3.171546697616577, + "learning_rate": 0.00018649614968474303, + "loss": 0.3778, + "step": 7147 + }, + { + "epoch": 0.16756037670610713, + "grad_norm": 4.588571071624756, + "learning_rate": 0.00018649245283047387, + "loss": 0.7628, + "step": 7148 + }, + { + "epoch": 0.16758381828091212, + "grad_norm": 2.7525529861450195, + "learning_rate": 0.00018648875550689425, + "loss": 0.2078, + "step": 7149 + }, + { + "epoch": 0.16760725985571712, + "grad_norm": 3.0999746322631836, + "learning_rate": 0.00018648505771402417, + "loss": 0.3195, + "step": 7150 + }, + { + "epoch": 0.1676307014305221, + "grad_norm": 2.182866096496582, + "learning_rate": 0.00018648135945188376, + "loss": 0.3584, + "step": 7151 + }, + { + "epoch": 0.1676541430053271, + "grad_norm": 1.3921046257019043, + "learning_rate": 0.00018647766072049304, + "loss": 0.2002, + "step": 7152 + }, + { + "epoch": 0.1676775845801321, + "grad_norm": 3.517162322998047, + "learning_rate": 0.00018647396151987212, + "loss": 0.6173, + "step": 7153 + }, + { + "epoch": 0.1677010261549371, + "grad_norm": 4.624171733856201, + "learning_rate": 0.00018647026185004105, + "loss": 0.5829, + "step": 7154 + }, + { + "epoch": 0.1677244677297421, + "grad_norm": 5.341935157775879, + "learning_rate": 0.00018646656171101991, + "loss": 0.8869, + "step": 7155 + }, + { + "epoch": 0.16774790930454708, + "grad_norm": 1.0835070610046387, + "learning_rate": 0.00018646286110282878, + "loss": 0.168, + "step": 7156 + }, + { + "epoch": 0.16777135087935208, + "grad_norm": 1.1033780574798584, + "learning_rate": 0.00018645916002548775, + "loss": 0.1782, + "step": 7157 + }, + { + "epoch": 0.16779479245415707, + "grad_norm": 4.057155609130859, + "learning_rate": 0.00018645545847901685, + "loss": 0.5933, + "step": 7158 + }, + { + "epoch": 0.16781823402896207, + "grad_norm": 1.6548995971679688, + "learning_rate": 0.0001864517564634362, + "loss": 0.2744, + "step": 7159 + }, + { + "epoch": 0.16784167560376706, + "grad_norm": 1.7994784116744995, + "learning_rate": 0.00018644805397876593, + "loss": 0.1196, + "step": 7160 + }, + { + "epoch": 0.16786511717857205, + "grad_norm": 2.792032480239868, + "learning_rate": 0.00018644435102502604, + "loss": 0.1437, + "step": 7161 + }, + { + "epoch": 0.16788855875337705, + "grad_norm": 3.5861141681671143, + "learning_rate": 0.0001864406476022367, + "loss": 0.3275, + "step": 7162 + }, + { + "epoch": 0.16791200032818204, + "grad_norm": 1.0089961290359497, + "learning_rate": 0.00018643694371041797, + "loss": 0.1578, + "step": 7163 + }, + { + "epoch": 0.16793544190298704, + "grad_norm": 3.8008806705474854, + "learning_rate": 0.00018643323934958996, + "loss": 1.7473, + "step": 7164 + }, + { + "epoch": 0.16795888347779203, + "grad_norm": 1.8919882774353027, + "learning_rate": 0.00018642953451977276, + "loss": 0.2547, + "step": 7165 + }, + { + "epoch": 0.16798232505259703, + "grad_norm": 5.373246192932129, + "learning_rate": 0.0001864258292209865, + "loss": 0.9348, + "step": 7166 + }, + { + "epoch": 0.16800576662740202, + "grad_norm": 3.883751153945923, + "learning_rate": 0.00018642212345325117, + "loss": 0.5199, + "step": 7167 + }, + { + "epoch": 0.16802920820220701, + "grad_norm": 2.6789329051971436, + "learning_rate": 0.00018641841721658706, + "loss": 0.3934, + "step": 7168 + }, + { + "epoch": 0.168052649777012, + "grad_norm": 0.827060878276825, + "learning_rate": 0.00018641471051101415, + "loss": 0.144, + "step": 7169 + }, + { + "epoch": 0.168076091351817, + "grad_norm": 0.774883508682251, + "learning_rate": 0.00018641100333655262, + "loss": 0.1149, + "step": 7170 + }, + { + "epoch": 0.168099532926622, + "grad_norm": 3.30903959274292, + "learning_rate": 0.00018640729569322252, + "loss": 0.6285, + "step": 7171 + }, + { + "epoch": 0.16812297450142702, + "grad_norm": 2.715195655822754, + "learning_rate": 0.000186403587581044, + "loss": 0.3583, + "step": 7172 + }, + { + "epoch": 0.168146416076232, + "grad_norm": 3.173231601715088, + "learning_rate": 0.00018639987900003724, + "loss": 0.2102, + "step": 7173 + }, + { + "epoch": 0.168169857651037, + "grad_norm": 4.469428062438965, + "learning_rate": 0.00018639616995022227, + "loss": 0.5887, + "step": 7174 + }, + { + "epoch": 0.168193299225842, + "grad_norm": 1.401708722114563, + "learning_rate": 0.00018639246043161925, + "loss": 0.9537, + "step": 7175 + }, + { + "epoch": 0.168216740800647, + "grad_norm": 2.1270499229431152, + "learning_rate": 0.0001863887504442483, + "loss": 0.122, + "step": 7176 + }, + { + "epoch": 0.168240182375452, + "grad_norm": 2.6454224586486816, + "learning_rate": 0.0001863850399881296, + "loss": 0.4063, + "step": 7177 + }, + { + "epoch": 0.16826362395025699, + "grad_norm": 1.916988730430603, + "learning_rate": 0.00018638132906328322, + "loss": 0.2747, + "step": 7178 + }, + { + "epoch": 0.16828706552506198, + "grad_norm": 4.058072090148926, + "learning_rate": 0.0001863776176697293, + "loss": 0.6283, + "step": 7179 + }, + { + "epoch": 0.16831050709986697, + "grad_norm": 2.885026216506958, + "learning_rate": 0.000186373905807488, + "loss": 0.2553, + "step": 7180 + }, + { + "epoch": 0.16833394867467197, + "grad_norm": 4.159292221069336, + "learning_rate": 0.0001863701934765795, + "loss": 0.4962, + "step": 7181 + }, + { + "epoch": 0.16835739024947696, + "grad_norm": 1.6155487298965454, + "learning_rate": 0.00018636648067702386, + "loss": 0.0909, + "step": 7182 + }, + { + "epoch": 0.16838083182428196, + "grad_norm": 1.1007460355758667, + "learning_rate": 0.00018636276740884128, + "loss": 0.0805, + "step": 7183 + }, + { + "epoch": 0.16840427339908695, + "grad_norm": 3.5663301944732666, + "learning_rate": 0.00018635905367205192, + "loss": 0.3617, + "step": 7184 + }, + { + "epoch": 0.16842771497389195, + "grad_norm": 3.950833559036255, + "learning_rate": 0.00018635533946667588, + "loss": 0.515, + "step": 7185 + }, + { + "epoch": 0.16845115654869694, + "grad_norm": 4.814545631408691, + "learning_rate": 0.00018635162479273336, + "loss": 0.2837, + "step": 7186 + }, + { + "epoch": 0.16847459812350193, + "grad_norm": 2.2003443241119385, + "learning_rate": 0.00018634790965024446, + "loss": 0.2378, + "step": 7187 + }, + { + "epoch": 0.16849803969830693, + "grad_norm": 4.155704498291016, + "learning_rate": 0.0001863441940392294, + "loss": 0.2494, + "step": 7188 + }, + { + "epoch": 0.16852148127311192, + "grad_norm": 6.0453782081604, + "learning_rate": 0.00018634047795970833, + "loss": 0.7125, + "step": 7189 + }, + { + "epoch": 0.16854492284791692, + "grad_norm": 4.738590717315674, + "learning_rate": 0.00018633676141170137, + "loss": 0.5708, + "step": 7190 + }, + { + "epoch": 0.1685683644227219, + "grad_norm": 3.1067605018615723, + "learning_rate": 0.00018633304439522874, + "loss": 0.3651, + "step": 7191 + }, + { + "epoch": 0.1685918059975269, + "grad_norm": 6.412949562072754, + "learning_rate": 0.0001863293269103106, + "loss": 0.7792, + "step": 7192 + }, + { + "epoch": 0.1686152475723319, + "grad_norm": 2.2849645614624023, + "learning_rate": 0.00018632560895696703, + "loss": 0.2504, + "step": 7193 + }, + { + "epoch": 0.1686386891471369, + "grad_norm": 3.286504030227661, + "learning_rate": 0.00018632189053521833, + "loss": 0.3025, + "step": 7194 + }, + { + "epoch": 0.1686621307219419, + "grad_norm": 2.1405248641967773, + "learning_rate": 0.00018631817164508463, + "loss": 0.3309, + "step": 7195 + }, + { + "epoch": 0.16868557229674688, + "grad_norm": 4.089468955993652, + "learning_rate": 0.0001863144522865861, + "loss": 0.3407, + "step": 7196 + }, + { + "epoch": 0.16870901387155188, + "grad_norm": 3.7188327312469482, + "learning_rate": 0.0001863107324597429, + "loss": 0.5439, + "step": 7197 + }, + { + "epoch": 0.1687324554463569, + "grad_norm": 4.888668060302734, + "learning_rate": 0.00018630701216457523, + "loss": 0.3222, + "step": 7198 + }, + { + "epoch": 0.1687558970211619, + "grad_norm": 4.543917179107666, + "learning_rate": 0.00018630329140110333, + "loss": 0.6045, + "step": 7199 + }, + { + "epoch": 0.1687793385959669, + "grad_norm": 2.2287545204162598, + "learning_rate": 0.00018629957016934735, + "loss": 0.224, + "step": 7200 + }, + { + "epoch": 0.16880278017077188, + "grad_norm": 3.517305850982666, + "learning_rate": 0.00018629584846932743, + "loss": 0.4458, + "step": 7201 + }, + { + "epoch": 0.16882622174557688, + "grad_norm": 4.202040195465088, + "learning_rate": 0.00018629212630106383, + "loss": 0.4853, + "step": 7202 + }, + { + "epoch": 0.16884966332038187, + "grad_norm": 3.225316286087036, + "learning_rate": 0.00018628840366457673, + "loss": 0.5422, + "step": 7203 + }, + { + "epoch": 0.16887310489518687, + "grad_norm": 1.428105354309082, + "learning_rate": 0.0001862846805598863, + "loss": 0.8681, + "step": 7204 + }, + { + "epoch": 0.16889654646999186, + "grad_norm": 2.1443705558776855, + "learning_rate": 0.0001862809569870128, + "loss": 0.2893, + "step": 7205 + }, + { + "epoch": 0.16891998804479685, + "grad_norm": 4.651994228363037, + "learning_rate": 0.0001862772329459764, + "loss": 0.5112, + "step": 7206 + }, + { + "epoch": 0.16894342961960185, + "grad_norm": 4.783205032348633, + "learning_rate": 0.00018627350843679729, + "loss": 0.6358, + "step": 7207 + }, + { + "epoch": 0.16896687119440684, + "grad_norm": 4.381723403930664, + "learning_rate": 0.0001862697834594957, + "loss": 0.8694, + "step": 7208 + }, + { + "epoch": 0.16899031276921184, + "grad_norm": 4.673503875732422, + "learning_rate": 0.00018626605801409186, + "loss": 0.4902, + "step": 7209 + }, + { + "epoch": 0.16901375434401683, + "grad_norm": 4.305656433105469, + "learning_rate": 0.00018626233210060594, + "loss": 0.3675, + "step": 7210 + }, + { + "epoch": 0.16903719591882183, + "grad_norm": 3.1346566677093506, + "learning_rate": 0.0001862586057190582, + "loss": 0.5167, + "step": 7211 + }, + { + "epoch": 0.16906063749362682, + "grad_norm": 0.46454131603240967, + "learning_rate": 0.00018625487886946883, + "loss": 0.0477, + "step": 7212 + }, + { + "epoch": 0.16908407906843181, + "grad_norm": 3.540377140045166, + "learning_rate": 0.00018625115155185805, + "loss": 0.406, + "step": 7213 + }, + { + "epoch": 0.1691075206432368, + "grad_norm": 2.7916836738586426, + "learning_rate": 0.00018624742376624611, + "loss": 0.1603, + "step": 7214 + }, + { + "epoch": 0.1691309622180418, + "grad_norm": 3.747492551803589, + "learning_rate": 0.00018624369551265325, + "loss": 0.5761, + "step": 7215 + }, + { + "epoch": 0.1691544037928468, + "grad_norm": 5.3566975593566895, + "learning_rate": 0.00018623996679109964, + "loss": 0.5397, + "step": 7216 + }, + { + "epoch": 0.1691778453676518, + "grad_norm": 2.3854799270629883, + "learning_rate": 0.00018623623760160554, + "loss": 0.1761, + "step": 7217 + }, + { + "epoch": 0.16920128694245679, + "grad_norm": 4.001609802246094, + "learning_rate": 0.0001862325079441912, + "loss": 0.513, + "step": 7218 + }, + { + "epoch": 0.16922472851726178, + "grad_norm": 3.406538248062134, + "learning_rate": 0.00018622877781887683, + "loss": 0.4348, + "step": 7219 + }, + { + "epoch": 0.16924817009206677, + "grad_norm": 2.577540397644043, + "learning_rate": 0.00018622504722568272, + "loss": 0.2947, + "step": 7220 + }, + { + "epoch": 0.16927161166687177, + "grad_norm": 1.5019338130950928, + "learning_rate": 0.00018622131616462907, + "loss": 0.0546, + "step": 7221 + }, + { + "epoch": 0.16929505324167676, + "grad_norm": 3.5709104537963867, + "learning_rate": 0.00018621758463573614, + "loss": 0.4736, + "step": 7222 + }, + { + "epoch": 0.16931849481648178, + "grad_norm": 1.1135072708129883, + "learning_rate": 0.00018621385263902415, + "loss": 0.12, + "step": 7223 + }, + { + "epoch": 0.16934193639128678, + "grad_norm": 3.396007537841797, + "learning_rate": 0.00018621012017451337, + "loss": 0.5231, + "step": 7224 + }, + { + "epoch": 0.16936537796609177, + "grad_norm": 0.9098303914070129, + "learning_rate": 0.00018620638724222409, + "loss": 0.0617, + "step": 7225 + }, + { + "epoch": 0.16938881954089677, + "grad_norm": 5.65400505065918, + "learning_rate": 0.00018620265384217648, + "loss": 0.7268, + "step": 7226 + }, + { + "epoch": 0.16941226111570176, + "grad_norm": 0.4799584746360779, + "learning_rate": 0.0001861989199743909, + "loss": 0.0428, + "step": 7227 + }, + { + "epoch": 0.16943570269050676, + "grad_norm": 3.890293836593628, + "learning_rate": 0.00018619518563888748, + "loss": 0.7169, + "step": 7228 + }, + { + "epoch": 0.16945914426531175, + "grad_norm": 3.811370372772217, + "learning_rate": 0.00018619145083568662, + "loss": 0.3092, + "step": 7229 + }, + { + "epoch": 0.16948258584011675, + "grad_norm": 1.194105863571167, + "learning_rate": 0.00018618771556480854, + "loss": 0.1324, + "step": 7230 + }, + { + "epoch": 0.16950602741492174, + "grad_norm": 0.9112176299095154, + "learning_rate": 0.00018618397982627345, + "loss": 0.1041, + "step": 7231 + }, + { + "epoch": 0.16952946898972673, + "grad_norm": 5.229382038116455, + "learning_rate": 0.00018618024362010167, + "loss": 0.8101, + "step": 7232 + }, + { + "epoch": 0.16955291056453173, + "grad_norm": 2.689318895339966, + "learning_rate": 0.00018617650694631346, + "loss": 1.5817, + "step": 7233 + }, + { + "epoch": 0.16957635213933672, + "grad_norm": 2.972576379776001, + "learning_rate": 0.0001861727698049291, + "loss": 0.3589, + "step": 7234 + }, + { + "epoch": 0.16959979371414172, + "grad_norm": 3.862354040145874, + "learning_rate": 0.0001861690321959689, + "loss": 0.5037, + "step": 7235 + }, + { + "epoch": 0.1696232352889467, + "grad_norm": 3.65657639503479, + "learning_rate": 0.00018616529411945307, + "loss": 0.4249, + "step": 7236 + }, + { + "epoch": 0.1696466768637517, + "grad_norm": 1.3993932008743286, + "learning_rate": 0.00018616155557540196, + "loss": 0.132, + "step": 7237 + }, + { + "epoch": 0.1696701184385567, + "grad_norm": 4.240357875823975, + "learning_rate": 0.0001861578165638358, + "loss": 0.4419, + "step": 7238 + }, + { + "epoch": 0.1696935600133617, + "grad_norm": 4.337752819061279, + "learning_rate": 0.00018615407708477496, + "loss": 0.5932, + "step": 7239 + }, + { + "epoch": 0.1697170015881667, + "grad_norm": 3.650844097137451, + "learning_rate": 0.0001861503371382396, + "loss": 0.3112, + "step": 7240 + }, + { + "epoch": 0.16974044316297168, + "grad_norm": 5.46611213684082, + "learning_rate": 0.00018614659672425013, + "loss": 0.7769, + "step": 7241 + }, + { + "epoch": 0.16976388473777668, + "grad_norm": 5.388636112213135, + "learning_rate": 0.0001861428558428268, + "loss": 0.406, + "step": 7242 + }, + { + "epoch": 0.16978732631258167, + "grad_norm": 0.6974598169326782, + "learning_rate": 0.0001861391144939899, + "loss": 0.0636, + "step": 7243 + }, + { + "epoch": 0.16981076788738667, + "grad_norm": 2.2765605449676514, + "learning_rate": 0.00018613537267775975, + "loss": 0.2465, + "step": 7244 + }, + { + "epoch": 0.16983420946219166, + "grad_norm": 4.424454212188721, + "learning_rate": 0.00018613163039415666, + "loss": 0.6137, + "step": 7245 + }, + { + "epoch": 0.16985765103699665, + "grad_norm": 3.9598991870880127, + "learning_rate": 0.0001861278876432009, + "loss": 0.4482, + "step": 7246 + }, + { + "epoch": 0.16988109261180165, + "grad_norm": 2.710127115249634, + "learning_rate": 0.0001861241444249128, + "loss": 0.2444, + "step": 7247 + }, + { + "epoch": 0.16990453418660664, + "grad_norm": 6.593414306640625, + "learning_rate": 0.0001861204007393127, + "loss": 0.6922, + "step": 7248 + }, + { + "epoch": 0.16992797576141166, + "grad_norm": 2.2063143253326416, + "learning_rate": 0.00018611665658642084, + "loss": 0.2068, + "step": 7249 + }, + { + "epoch": 0.16995141733621666, + "grad_norm": 4.434950351715088, + "learning_rate": 0.0001861129119662576, + "loss": 0.4299, + "step": 7250 + }, + { + "epoch": 0.16997485891102165, + "grad_norm": 4.129875183105469, + "learning_rate": 0.00018610916687884332, + "loss": 0.589, + "step": 7251 + }, + { + "epoch": 0.16999830048582665, + "grad_norm": 3.445819616317749, + "learning_rate": 0.0001861054213241982, + "loss": 0.4804, + "step": 7252 + }, + { + "epoch": 0.17002174206063164, + "grad_norm": 3.7318060398101807, + "learning_rate": 0.0001861016753023427, + "loss": 0.7133, + "step": 7253 + }, + { + "epoch": 0.17004518363543664, + "grad_norm": 4.587769031524658, + "learning_rate": 0.00018609792881329708, + "loss": 0.4795, + "step": 7254 + }, + { + "epoch": 0.17006862521024163, + "grad_norm": 8.836363792419434, + "learning_rate": 0.00018609418185708168, + "loss": 0.4659, + "step": 7255 + }, + { + "epoch": 0.17009206678504663, + "grad_norm": 3.9562411308288574, + "learning_rate": 0.00018609043443371684, + "loss": 0.4469, + "step": 7256 + }, + { + "epoch": 0.17011550835985162, + "grad_norm": 3.6359169483184814, + "learning_rate": 0.00018608668654322285, + "loss": 0.3205, + "step": 7257 + }, + { + "epoch": 0.1701389499346566, + "grad_norm": 1.1652781963348389, + "learning_rate": 0.00018608293818562008, + "loss": 0.1568, + "step": 7258 + }, + { + "epoch": 0.1701623915094616, + "grad_norm": 2.756019115447998, + "learning_rate": 0.00018607918936092887, + "loss": 0.3309, + "step": 7259 + }, + { + "epoch": 0.1701858330842666, + "grad_norm": 0.4620025157928467, + "learning_rate": 0.00018607544006916956, + "loss": 0.0769, + "step": 7260 + }, + { + "epoch": 0.1702092746590716, + "grad_norm": 4.728423595428467, + "learning_rate": 0.0001860716903103625, + "loss": 0.7287, + "step": 7261 + }, + { + "epoch": 0.1702327162338766, + "grad_norm": 2.9607326984405518, + "learning_rate": 0.00018606794008452804, + "loss": 0.2835, + "step": 7262 + }, + { + "epoch": 0.17025615780868159, + "grad_norm": 6.301852703094482, + "learning_rate": 0.0001860641893916865, + "loss": 0.6657, + "step": 7263 + }, + { + "epoch": 0.17027959938348658, + "grad_norm": 3.5639193058013916, + "learning_rate": 0.00018606043823185826, + "loss": 0.5465, + "step": 7264 + }, + { + "epoch": 0.17030304095829157, + "grad_norm": 3.7831990718841553, + "learning_rate": 0.00018605668660506367, + "loss": 0.445, + "step": 7265 + }, + { + "epoch": 0.17032648253309657, + "grad_norm": 3.1563680171966553, + "learning_rate": 0.000186052934511323, + "loss": 0.5453, + "step": 7266 + }, + { + "epoch": 0.17034992410790156, + "grad_norm": 4.454673767089844, + "learning_rate": 0.00018604918195065678, + "loss": 0.6947, + "step": 7267 + }, + { + "epoch": 0.17037336568270656, + "grad_norm": 4.467764854431152, + "learning_rate": 0.00018604542892308525, + "loss": 0.5441, + "step": 7268 + }, + { + "epoch": 0.17039680725751155, + "grad_norm": 4.247813701629639, + "learning_rate": 0.00018604167542862882, + "loss": 0.6502, + "step": 7269 + }, + { + "epoch": 0.17042024883231655, + "grad_norm": 3.149829387664795, + "learning_rate": 0.00018603792146730782, + "loss": 0.3255, + "step": 7270 + }, + { + "epoch": 0.17044369040712154, + "grad_norm": 4.3747124671936035, + "learning_rate": 0.00018603416703914262, + "loss": 0.629, + "step": 7271 + }, + { + "epoch": 0.17046713198192653, + "grad_norm": 3.062584638595581, + "learning_rate": 0.0001860304121441536, + "loss": 0.3318, + "step": 7272 + }, + { + "epoch": 0.17049057355673153, + "grad_norm": 5.072171211242676, + "learning_rate": 0.0001860266567823612, + "loss": 0.3341, + "step": 7273 + }, + { + "epoch": 0.17051401513153655, + "grad_norm": 5.177538871765137, + "learning_rate": 0.0001860229009537857, + "loss": 1.6744, + "step": 7274 + }, + { + "epoch": 0.17053745670634154, + "grad_norm": 4.213603496551514, + "learning_rate": 0.00018601914465844757, + "loss": 0.4134, + "step": 7275 + }, + { + "epoch": 0.17056089828114654, + "grad_norm": 1.3795462846755981, + "learning_rate": 0.0001860153878963671, + "loss": 0.2138, + "step": 7276 + }, + { + "epoch": 0.17058433985595153, + "grad_norm": 2.7727174758911133, + "learning_rate": 0.00018601163066756473, + "loss": 0.2325, + "step": 7277 + }, + { + "epoch": 0.17060778143075653, + "grad_norm": 0.7187258005142212, + "learning_rate": 0.00018600787297206085, + "loss": 0.1016, + "step": 7278 + }, + { + "epoch": 0.17063122300556152, + "grad_norm": 6.400426864624023, + "learning_rate": 0.0001860041148098758, + "loss": 0.9064, + "step": 7279 + }, + { + "epoch": 0.17065466458036652, + "grad_norm": 1.0538597106933594, + "learning_rate": 0.00018600035618103003, + "loss": 0.1725, + "step": 7280 + }, + { + "epoch": 0.1706781061551715, + "grad_norm": 1.1975361108779907, + "learning_rate": 0.0001859965970855439, + "loss": 0.1279, + "step": 7281 + }, + { + "epoch": 0.1707015477299765, + "grad_norm": 4.654057502746582, + "learning_rate": 0.00018599283752343779, + "loss": 0.5084, + "step": 7282 + }, + { + "epoch": 0.1707249893047815, + "grad_norm": 2.729679584503174, + "learning_rate": 0.00018598907749473216, + "loss": 0.2911, + "step": 7283 + }, + { + "epoch": 0.1707484308795865, + "grad_norm": 0.997306227684021, + "learning_rate": 0.0001859853169994474, + "loss": 0.1544, + "step": 7284 + }, + { + "epoch": 0.1707718724543915, + "grad_norm": 3.499056577682495, + "learning_rate": 0.00018598155603760384, + "loss": 0.3017, + "step": 7285 + }, + { + "epoch": 0.17079531402919648, + "grad_norm": 4.658726692199707, + "learning_rate": 0.00018597779460922197, + "loss": 1.0792, + "step": 7286 + }, + { + "epoch": 0.17081875560400148, + "grad_norm": 2.105886459350586, + "learning_rate": 0.00018597403271432216, + "loss": 0.8859, + "step": 7287 + }, + { + "epoch": 0.17084219717880647, + "grad_norm": 1.7345420122146606, + "learning_rate": 0.00018597027035292482, + "loss": 0.9453, + "step": 7288 + }, + { + "epoch": 0.17086563875361147, + "grad_norm": 2.002912998199463, + "learning_rate": 0.0001859665075250504, + "loss": 1.377, + "step": 7289 + }, + { + "epoch": 0.17088908032841646, + "grad_norm": 2.6610655784606934, + "learning_rate": 0.00018596274423071925, + "loss": 0.2003, + "step": 7290 + }, + { + "epoch": 0.17091252190322145, + "grad_norm": 1.6247305870056152, + "learning_rate": 0.00018595898046995185, + "loss": 0.3702, + "step": 7291 + }, + { + "epoch": 0.17093596347802645, + "grad_norm": 1.5453822612762451, + "learning_rate": 0.00018595521624276864, + "loss": 0.1967, + "step": 7292 + }, + { + "epoch": 0.17095940505283144, + "grad_norm": 1.8261698484420776, + "learning_rate": 0.00018595145154918998, + "loss": 0.3097, + "step": 7293 + }, + { + "epoch": 0.17098284662763644, + "grad_norm": 1.2697534561157227, + "learning_rate": 0.00018594768638923633, + "loss": 0.1463, + "step": 7294 + }, + { + "epoch": 0.17100628820244143, + "grad_norm": 4.495408058166504, + "learning_rate": 0.0001859439207629281, + "loss": 0.7801, + "step": 7295 + }, + { + "epoch": 0.17102972977724643, + "grad_norm": 4.362523078918457, + "learning_rate": 0.00018594015467028577, + "loss": 0.7053, + "step": 7296 + }, + { + "epoch": 0.17105317135205142, + "grad_norm": 2.5200815200805664, + "learning_rate": 0.00018593638811132974, + "loss": 1.4965, + "step": 7297 + }, + { + "epoch": 0.17107661292685641, + "grad_norm": 3.4044532775878906, + "learning_rate": 0.00018593262108608044, + "loss": 2.1704, + "step": 7298 + }, + { + "epoch": 0.1711000545016614, + "grad_norm": 3.53843355178833, + "learning_rate": 0.00018592885359455834, + "loss": 0.6572, + "step": 7299 + }, + { + "epoch": 0.17112349607646643, + "grad_norm": 4.105966567993164, + "learning_rate": 0.00018592508563678383, + "loss": 0.5423, + "step": 7300 + }, + { + "epoch": 0.17114693765127142, + "grad_norm": 4.201674938201904, + "learning_rate": 0.00018592131721277745, + "loss": 0.6847, + "step": 7301 + }, + { + "epoch": 0.17117037922607642, + "grad_norm": 3.5300447940826416, + "learning_rate": 0.00018591754832255952, + "loss": 0.5878, + "step": 7302 + }, + { + "epoch": 0.1711938208008814, + "grad_norm": 2.4711246490478516, + "learning_rate": 0.0001859137789661506, + "loss": 0.2069, + "step": 7303 + }, + { + "epoch": 0.1712172623756864, + "grad_norm": 1.828276515007019, + "learning_rate": 0.0001859100091435711, + "loss": 0.9176, + "step": 7304 + }, + { + "epoch": 0.1712407039504914, + "grad_norm": 1.7461262941360474, + "learning_rate": 0.00018590623885484143, + "loss": 0.9481, + "step": 7305 + }, + { + "epoch": 0.1712641455252964, + "grad_norm": 3.4894862174987793, + "learning_rate": 0.00018590246809998214, + "loss": 0.2225, + "step": 7306 + }, + { + "epoch": 0.1712875871001014, + "grad_norm": 1.7019249200820923, + "learning_rate": 0.0001858986968790136, + "loss": 0.1405, + "step": 7307 + }, + { + "epoch": 0.17131102867490638, + "grad_norm": 2.1520180702209473, + "learning_rate": 0.00018589492519195634, + "loss": 1.6512, + "step": 7308 + }, + { + "epoch": 0.17133447024971138, + "grad_norm": 3.1699249744415283, + "learning_rate": 0.0001858911530388308, + "loss": 0.423, + "step": 7309 + }, + { + "epoch": 0.17135791182451637, + "grad_norm": 0.8282530307769775, + "learning_rate": 0.00018588738041965748, + "loss": 0.115, + "step": 7310 + }, + { + "epoch": 0.17138135339932137, + "grad_norm": 3.753955841064453, + "learning_rate": 0.00018588360733445676, + "loss": 0.3704, + "step": 7311 + }, + { + "epoch": 0.17140479497412636, + "grad_norm": 1.1250977516174316, + "learning_rate": 0.0001858798337832492, + "loss": 0.1665, + "step": 7312 + }, + { + "epoch": 0.17142823654893136, + "grad_norm": 2.3368687629699707, + "learning_rate": 0.00018587605976605524, + "loss": 0.2663, + "step": 7313 + }, + { + "epoch": 0.17145167812373635, + "grad_norm": 3.196676254272461, + "learning_rate": 0.0001858722852828954, + "loss": 0.3375, + "step": 7314 + }, + { + "epoch": 0.17147511969854135, + "grad_norm": 3.0855162143707275, + "learning_rate": 0.0001858685103337901, + "loss": 0.3807, + "step": 7315 + }, + { + "epoch": 0.17149856127334634, + "grad_norm": 4.64393424987793, + "learning_rate": 0.00018586473491875985, + "loss": 0.6492, + "step": 7316 + }, + { + "epoch": 0.17152200284815133, + "grad_norm": 5.084710121154785, + "learning_rate": 0.00018586095903782513, + "loss": 0.7355, + "step": 7317 + }, + { + "epoch": 0.17154544442295633, + "grad_norm": 1.0842854976654053, + "learning_rate": 0.0001858571826910064, + "loss": 0.0939, + "step": 7318 + }, + { + "epoch": 0.17156888599776132, + "grad_norm": 3.5019564628601074, + "learning_rate": 0.00018585340587832423, + "loss": 0.645, + "step": 7319 + }, + { + "epoch": 0.17159232757256632, + "grad_norm": 4.931536674499512, + "learning_rate": 0.00018584962859979903, + "loss": 0.5082, + "step": 7320 + }, + { + "epoch": 0.1716157691473713, + "grad_norm": 2.9192075729370117, + "learning_rate": 0.00018584585085545138, + "loss": 0.4018, + "step": 7321 + }, + { + "epoch": 0.1716392107221763, + "grad_norm": 1.3793444633483887, + "learning_rate": 0.0001858420726453017, + "loss": 0.2079, + "step": 7322 + }, + { + "epoch": 0.1716626522969813, + "grad_norm": 2.113192319869995, + "learning_rate": 0.0001858382939693705, + "loss": 1.2806, + "step": 7323 + }, + { + "epoch": 0.1716860938717863, + "grad_norm": 4.433186054229736, + "learning_rate": 0.00018583451482767833, + "loss": 0.6064, + "step": 7324 + }, + { + "epoch": 0.1717095354465913, + "grad_norm": 0.578018844127655, + "learning_rate": 0.0001858307352202457, + "loss": 0.06, + "step": 7325 + }, + { + "epoch": 0.1717329770213963, + "grad_norm": 1.4810616970062256, + "learning_rate": 0.00018582695514709303, + "loss": 0.2204, + "step": 7326 + }, + { + "epoch": 0.1717564185962013, + "grad_norm": 2.819854259490967, + "learning_rate": 0.00018582317460824092, + "loss": 1.7566, + "step": 7327 + }, + { + "epoch": 0.1717798601710063, + "grad_norm": 3.3489770889282227, + "learning_rate": 0.0001858193936037098, + "loss": 0.5881, + "step": 7328 + }, + { + "epoch": 0.1718033017458113, + "grad_norm": 2.6894073486328125, + "learning_rate": 0.00018581561213352031, + "loss": 0.1671, + "step": 7329 + }, + { + "epoch": 0.1718267433206163, + "grad_norm": 0.8077715635299683, + "learning_rate": 0.00018581183019769287, + "loss": 0.1066, + "step": 7330 + }, + { + "epoch": 0.17185018489542128, + "grad_norm": 3.420206308364868, + "learning_rate": 0.000185808047796248, + "loss": 0.4041, + "step": 7331 + }, + { + "epoch": 0.17187362647022628, + "grad_norm": 2.4949207305908203, + "learning_rate": 0.00018580426492920628, + "loss": 0.2866, + "step": 7332 + }, + { + "epoch": 0.17189706804503127, + "grad_norm": 0.6988838911056519, + "learning_rate": 0.00018580048159658817, + "loss": 0.0865, + "step": 7333 + }, + { + "epoch": 0.17192050961983626, + "grad_norm": 4.702048301696777, + "learning_rate": 0.0001857966977984143, + "loss": 0.5789, + "step": 7334 + }, + { + "epoch": 0.17194395119464126, + "grad_norm": 1.2761895656585693, + "learning_rate": 0.0001857929135347051, + "loss": 0.8679, + "step": 7335 + }, + { + "epoch": 0.17196739276944625, + "grad_norm": 3.109799861907959, + "learning_rate": 0.00018578912880548113, + "loss": 0.3016, + "step": 7336 + }, + { + "epoch": 0.17199083434425125, + "grad_norm": 3.8382749557495117, + "learning_rate": 0.00018578534361076291, + "loss": 0.3416, + "step": 7337 + }, + { + "epoch": 0.17201427591905624, + "grad_norm": 2.4270412921905518, + "learning_rate": 0.00018578155795057107, + "loss": 0.349, + "step": 7338 + }, + { + "epoch": 0.17203771749386124, + "grad_norm": 1.126242756843567, + "learning_rate": 0.00018577777182492603, + "loss": 0.1515, + "step": 7339 + }, + { + "epoch": 0.17206115906866623, + "grad_norm": 6.200927257537842, + "learning_rate": 0.0001857739852338484, + "loss": 0.7743, + "step": 7340 + }, + { + "epoch": 0.17208460064347122, + "grad_norm": 4.22707986831665, + "learning_rate": 0.00018577019817735872, + "loss": 0.6686, + "step": 7341 + }, + { + "epoch": 0.17210804221827622, + "grad_norm": 4.163585662841797, + "learning_rate": 0.00018576641065547756, + "loss": 0.4591, + "step": 7342 + }, + { + "epoch": 0.1721314837930812, + "grad_norm": 5.144712448120117, + "learning_rate": 0.0001857626226682254, + "loss": 0.9538, + "step": 7343 + }, + { + "epoch": 0.1721549253678862, + "grad_norm": 0.3436484634876251, + "learning_rate": 0.00018575883421562286, + "loss": 0.0398, + "step": 7344 + }, + { + "epoch": 0.1721783669426912, + "grad_norm": 2.339747190475464, + "learning_rate": 0.00018575504529769047, + "loss": 0.1753, + "step": 7345 + }, + { + "epoch": 0.1722018085174962, + "grad_norm": 4.940325736999512, + "learning_rate": 0.0001857512559144488, + "loss": 0.3617, + "step": 7346 + }, + { + "epoch": 0.1722252500923012, + "grad_norm": 1.0558693408966064, + "learning_rate": 0.00018574746606591838, + "loss": 0.21, + "step": 7347 + }, + { + "epoch": 0.17224869166710619, + "grad_norm": 3.1571381092071533, + "learning_rate": 0.00018574367575211983, + "loss": 0.4909, + "step": 7348 + }, + { + "epoch": 0.17227213324191118, + "grad_norm": 3.782111167907715, + "learning_rate": 0.00018573988497307367, + "loss": 0.7559, + "step": 7349 + }, + { + "epoch": 0.17229557481671617, + "grad_norm": 2.5473954677581787, + "learning_rate": 0.00018573609372880048, + "loss": 0.1817, + "step": 7350 + }, + { + "epoch": 0.1723190163915212, + "grad_norm": 2.518723249435425, + "learning_rate": 0.00018573230201932082, + "loss": 0.3076, + "step": 7351 + }, + { + "epoch": 0.1723424579663262, + "grad_norm": 3.9220945835113525, + "learning_rate": 0.0001857285098446553, + "loss": 0.5609, + "step": 7352 + }, + { + "epoch": 0.17236589954113118, + "grad_norm": 1.1952677965164185, + "learning_rate": 0.00018572471720482446, + "loss": 0.1934, + "step": 7353 + }, + { + "epoch": 0.17238934111593618, + "grad_norm": 3.9028513431549072, + "learning_rate": 0.00018572092409984895, + "loss": 0.6874, + "step": 7354 + }, + { + "epoch": 0.17241278269074117, + "grad_norm": 1.0970548391342163, + "learning_rate": 0.00018571713052974924, + "loss": 0.151, + "step": 7355 + }, + { + "epoch": 0.17243622426554617, + "grad_norm": 3.000720262527466, + "learning_rate": 0.000185713336494546, + "loss": 0.416, + "step": 7356 + }, + { + "epoch": 0.17245966584035116, + "grad_norm": 2.554553747177124, + "learning_rate": 0.00018570954199425974, + "loss": 0.1653, + "step": 7357 + }, + { + "epoch": 0.17248310741515616, + "grad_norm": 1.8580546379089355, + "learning_rate": 0.0001857057470289111, + "loss": 0.1412, + "step": 7358 + }, + { + "epoch": 0.17250654898996115, + "grad_norm": 2.381589412689209, + "learning_rate": 0.00018570195159852072, + "loss": 0.2574, + "step": 7359 + }, + { + "epoch": 0.17252999056476614, + "grad_norm": 3.225353717803955, + "learning_rate": 0.00018569815570310908, + "loss": 0.6176, + "step": 7360 + }, + { + "epoch": 0.17255343213957114, + "grad_norm": 0.4864685833454132, + "learning_rate": 0.00018569435934269688, + "loss": 0.0466, + "step": 7361 + }, + { + "epoch": 0.17257687371437613, + "grad_norm": 1.3133151531219482, + "learning_rate": 0.00018569056251730465, + "loss": 0.1115, + "step": 7362 + }, + { + "epoch": 0.17260031528918113, + "grad_norm": 2.8178529739379883, + "learning_rate": 0.00018568676522695305, + "loss": 0.3423, + "step": 7363 + }, + { + "epoch": 0.17262375686398612, + "grad_norm": 4.304673194885254, + "learning_rate": 0.00018568296747166265, + "loss": 0.2205, + "step": 7364 + }, + { + "epoch": 0.17264719843879112, + "grad_norm": 3.66237211227417, + "learning_rate": 0.00018567916925145403, + "loss": 0.5146, + "step": 7365 + }, + { + "epoch": 0.1726706400135961, + "grad_norm": 3.948988676071167, + "learning_rate": 0.00018567537056634782, + "loss": 0.3679, + "step": 7366 + }, + { + "epoch": 0.1726940815884011, + "grad_norm": 3.438138246536255, + "learning_rate": 0.00018567157141636468, + "loss": 0.3465, + "step": 7367 + }, + { + "epoch": 0.1727175231632061, + "grad_norm": 5.212063789367676, + "learning_rate": 0.00018566777180152518, + "loss": 0.5359, + "step": 7368 + }, + { + "epoch": 0.1727409647380111, + "grad_norm": 2.943286418914795, + "learning_rate": 0.00018566397172184993, + "loss": 1.4969, + "step": 7369 + }, + { + "epoch": 0.1727644063128161, + "grad_norm": 4.153974533081055, + "learning_rate": 0.00018566017117735957, + "loss": 0.7282, + "step": 7370 + }, + { + "epoch": 0.17278784788762108, + "grad_norm": 3.639549493789673, + "learning_rate": 0.00018565637016807469, + "loss": 0.5784, + "step": 7371 + }, + { + "epoch": 0.17281128946242608, + "grad_norm": 3.3826563358306885, + "learning_rate": 0.00018565256869401594, + "loss": 0.3856, + "step": 7372 + }, + { + "epoch": 0.17283473103723107, + "grad_norm": 2.3811416625976562, + "learning_rate": 0.00018564876675520396, + "loss": 0.2715, + "step": 7373 + }, + { + "epoch": 0.17285817261203607, + "grad_norm": 4.7680206298828125, + "learning_rate": 0.00018564496435165935, + "loss": 0.4713, + "step": 7374 + }, + { + "epoch": 0.17288161418684106, + "grad_norm": 2.695671558380127, + "learning_rate": 0.00018564116148340277, + "loss": 0.4189, + "step": 7375 + }, + { + "epoch": 0.17290505576164605, + "grad_norm": 1.4368984699249268, + "learning_rate": 0.00018563735815045485, + "loss": 0.1622, + "step": 7376 + }, + { + "epoch": 0.17292849733645108, + "grad_norm": 5.010223388671875, + "learning_rate": 0.0001856335543528362, + "loss": 0.4642, + "step": 7377 + }, + { + "epoch": 0.17295193891125607, + "grad_norm": 4.082458972930908, + "learning_rate": 0.00018562975009056747, + "loss": 0.4939, + "step": 7378 + }, + { + "epoch": 0.17297538048606106, + "grad_norm": 5.337831497192383, + "learning_rate": 0.00018562594536366932, + "loss": 0.4531, + "step": 7379 + }, + { + "epoch": 0.17299882206086606, + "grad_norm": 3.5182220935821533, + "learning_rate": 0.00018562214017216238, + "loss": 0.4888, + "step": 7380 + }, + { + "epoch": 0.17302226363567105, + "grad_norm": 0.6257554292678833, + "learning_rate": 0.00018561833451606727, + "loss": 0.063, + "step": 7381 + }, + { + "epoch": 0.17304570521047605, + "grad_norm": 1.812251329421997, + "learning_rate": 0.00018561452839540473, + "loss": 0.2442, + "step": 7382 + }, + { + "epoch": 0.17306914678528104, + "grad_norm": 4.64284610748291, + "learning_rate": 0.00018561072181019528, + "loss": 0.4946, + "step": 7383 + }, + { + "epoch": 0.17309258836008604, + "grad_norm": 1.251177430152893, + "learning_rate": 0.00018560691476045968, + "loss": 0.1145, + "step": 7384 + }, + { + "epoch": 0.17311602993489103, + "grad_norm": 2.381119728088379, + "learning_rate": 0.00018560310724621858, + "loss": 1.385, + "step": 7385 + }, + { + "epoch": 0.17313947150969602, + "grad_norm": 4.5922088623046875, + "learning_rate": 0.0001855992992674926, + "loss": 0.6334, + "step": 7386 + }, + { + "epoch": 0.17316291308450102, + "grad_norm": 1.4713913202285767, + "learning_rate": 0.00018559549082430239, + "loss": 0.1518, + "step": 7387 + }, + { + "epoch": 0.173186354659306, + "grad_norm": 3.6472384929656982, + "learning_rate": 0.00018559168191666867, + "loss": 0.3962, + "step": 7388 + }, + { + "epoch": 0.173209796234111, + "grad_norm": 1.9965853691101074, + "learning_rate": 0.00018558787254461205, + "loss": 1.3423, + "step": 7389 + }, + { + "epoch": 0.173233237808916, + "grad_norm": 3.1443560123443604, + "learning_rate": 0.00018558406270815324, + "loss": 0.3702, + "step": 7390 + }, + { + "epoch": 0.173256679383721, + "grad_norm": 0.7310524582862854, + "learning_rate": 0.00018558025240731293, + "loss": 0.0854, + "step": 7391 + }, + { + "epoch": 0.173280120958526, + "grad_norm": 3.652709722518921, + "learning_rate": 0.0001855764416421117, + "loss": 0.4655, + "step": 7392 + }, + { + "epoch": 0.17330356253333098, + "grad_norm": 3.6880669593811035, + "learning_rate": 0.00018557263041257032, + "loss": 0.4421, + "step": 7393 + }, + { + "epoch": 0.17332700410813598, + "grad_norm": 2.332014560699463, + "learning_rate": 0.00018556881871870947, + "loss": 0.2811, + "step": 7394 + }, + { + "epoch": 0.17335044568294097, + "grad_norm": 0.6253783106803894, + "learning_rate": 0.00018556500656054977, + "loss": 0.0844, + "step": 7395 + }, + { + "epoch": 0.17337388725774597, + "grad_norm": 1.4663673639297485, + "learning_rate": 0.00018556119393811193, + "loss": 0.2324, + "step": 7396 + }, + { + "epoch": 0.17339732883255096, + "grad_norm": 2.0093789100646973, + "learning_rate": 0.0001855573808514167, + "loss": 1.1072, + "step": 7397 + }, + { + "epoch": 0.17342077040735596, + "grad_norm": 4.556207180023193, + "learning_rate": 0.00018555356730048464, + "loss": 0.556, + "step": 7398 + }, + { + "epoch": 0.17344421198216095, + "grad_norm": 1.1356697082519531, + "learning_rate": 0.00018554975328533658, + "loss": 0.1141, + "step": 7399 + }, + { + "epoch": 0.17346765355696595, + "grad_norm": 4.106128692626953, + "learning_rate": 0.00018554593880599314, + "loss": 0.66, + "step": 7400 + }, + { + "epoch": 0.17349109513177094, + "grad_norm": 4.996467113494873, + "learning_rate": 0.000185542123862475, + "loss": 0.7531, + "step": 7401 + }, + { + "epoch": 0.17351453670657596, + "grad_norm": 2.727290391921997, + "learning_rate": 0.00018553830845480294, + "loss": 1.9379, + "step": 7402 + }, + { + "epoch": 0.17353797828138096, + "grad_norm": 1.9840331077575684, + "learning_rate": 0.0001855344925829976, + "loss": 0.1987, + "step": 7403 + }, + { + "epoch": 0.17356141985618595, + "grad_norm": 5.159419536590576, + "learning_rate": 0.00018553067624707965, + "loss": 0.6057, + "step": 7404 + }, + { + "epoch": 0.17358486143099094, + "grad_norm": 5.029829025268555, + "learning_rate": 0.0001855268594470699, + "loss": 0.6838, + "step": 7405 + }, + { + "epoch": 0.17360830300579594, + "grad_norm": 3.677671432495117, + "learning_rate": 0.00018552304218298897, + "loss": 0.5561, + "step": 7406 + }, + { + "epoch": 0.17363174458060093, + "grad_norm": 3.337019681930542, + "learning_rate": 0.00018551922445485762, + "loss": 0.3294, + "step": 7407 + }, + { + "epoch": 0.17365518615540593, + "grad_norm": 2.4299466609954834, + "learning_rate": 0.00018551540626269655, + "loss": 0.4753, + "step": 7408 + }, + { + "epoch": 0.17367862773021092, + "grad_norm": 3.9190192222595215, + "learning_rate": 0.0001855115876065265, + "loss": 0.5746, + "step": 7409 + }, + { + "epoch": 0.17370206930501592, + "grad_norm": 1.3173195123672485, + "learning_rate": 0.00018550776848636814, + "loss": 0.1038, + "step": 7410 + }, + { + "epoch": 0.1737255108798209, + "grad_norm": 2.2511284351348877, + "learning_rate": 0.00018550394890224227, + "loss": 1.7409, + "step": 7411 + }, + { + "epoch": 0.1737489524546259, + "grad_norm": 3.6388304233551025, + "learning_rate": 0.00018550012885416953, + "loss": 0.5311, + "step": 7412 + }, + { + "epoch": 0.1737723940294309, + "grad_norm": 2.258371353149414, + "learning_rate": 0.0001854963083421707, + "loss": 0.3271, + "step": 7413 + }, + { + "epoch": 0.1737958356042359, + "grad_norm": 0.8091490864753723, + "learning_rate": 0.0001854924873662665, + "loss": 0.1585, + "step": 7414 + }, + { + "epoch": 0.1738192771790409, + "grad_norm": 2.4709837436676025, + "learning_rate": 0.00018548866592647765, + "loss": 0.2845, + "step": 7415 + }, + { + "epoch": 0.17384271875384588, + "grad_norm": 3.295060157775879, + "learning_rate": 0.0001854848440228249, + "loss": 0.4231, + "step": 7416 + }, + { + "epoch": 0.17386616032865088, + "grad_norm": 4.9303388595581055, + "learning_rate": 0.00018548102165532901, + "loss": 0.8329, + "step": 7417 + }, + { + "epoch": 0.17388960190345587, + "grad_norm": 2.0194759368896484, + "learning_rate": 0.00018547719882401067, + "loss": 0.1911, + "step": 7418 + }, + { + "epoch": 0.17391304347826086, + "grad_norm": 0.7307173609733582, + "learning_rate": 0.00018547337552889064, + "loss": 0.1033, + "step": 7419 + }, + { + "epoch": 0.17393648505306586, + "grad_norm": 2.616753101348877, + "learning_rate": 0.00018546955176998968, + "loss": 0.3465, + "step": 7420 + }, + { + "epoch": 0.17395992662787085, + "grad_norm": 4.5033183097839355, + "learning_rate": 0.0001854657275473285, + "loss": 0.4166, + "step": 7421 + }, + { + "epoch": 0.17398336820267585, + "grad_norm": 4.282111167907715, + "learning_rate": 0.0001854619028609279, + "loss": 0.7422, + "step": 7422 + }, + { + "epoch": 0.17400680977748084, + "grad_norm": 3.639540195465088, + "learning_rate": 0.00018545807771080863, + "loss": 0.578, + "step": 7423 + }, + { + "epoch": 0.17403025135228584, + "grad_norm": 3.1065640449523926, + "learning_rate": 0.00018545425209699144, + "loss": 0.3013, + "step": 7424 + }, + { + "epoch": 0.17405369292709083, + "grad_norm": 2.8131263256073, + "learning_rate": 0.00018545042601949703, + "loss": 1.1296, + "step": 7425 + }, + { + "epoch": 0.17407713450189582, + "grad_norm": 3.4492554664611816, + "learning_rate": 0.00018544659947834622, + "loss": 0.3671, + "step": 7426 + }, + { + "epoch": 0.17410057607670082, + "grad_norm": 0.4789977967739105, + "learning_rate": 0.00018544277247355977, + "loss": 0.0797, + "step": 7427 + }, + { + "epoch": 0.17412401765150584, + "grad_norm": 2.607667922973633, + "learning_rate": 0.00018543894500515842, + "loss": 0.272, + "step": 7428 + }, + { + "epoch": 0.17414745922631084, + "grad_norm": 3.5319042205810547, + "learning_rate": 0.00018543511707316297, + "loss": 0.7549, + "step": 7429 + }, + { + "epoch": 0.17417090080111583, + "grad_norm": 4.021003246307373, + "learning_rate": 0.00018543128867759415, + "loss": 0.5746, + "step": 7430 + }, + { + "epoch": 0.17419434237592082, + "grad_norm": 5.27031135559082, + "learning_rate": 0.00018542745981847277, + "loss": 0.7442, + "step": 7431 + }, + { + "epoch": 0.17421778395072582, + "grad_norm": 2.6622393131256104, + "learning_rate": 0.00018542363049581958, + "loss": 0.1657, + "step": 7432 + }, + { + "epoch": 0.1742412255255308, + "grad_norm": 3.430176019668579, + "learning_rate": 0.00018541980070965538, + "loss": 0.4944, + "step": 7433 + }, + { + "epoch": 0.1742646671003358, + "grad_norm": 1.5112106800079346, + "learning_rate": 0.00018541597046000093, + "loss": 1.128, + "step": 7434 + }, + { + "epoch": 0.1742881086751408, + "grad_norm": 2.37322735786438, + "learning_rate": 0.00018541213974687703, + "loss": 0.2154, + "step": 7435 + }, + { + "epoch": 0.1743115502499458, + "grad_norm": 4.923589706420898, + "learning_rate": 0.00018540830857030445, + "loss": 0.5, + "step": 7436 + }, + { + "epoch": 0.1743349918247508, + "grad_norm": 2.618436574935913, + "learning_rate": 0.000185404476930304, + "loss": 1.8348, + "step": 7437 + }, + { + "epoch": 0.17435843339955578, + "grad_norm": 2.4752626419067383, + "learning_rate": 0.00018540064482689645, + "loss": 0.2493, + "step": 7438 + }, + { + "epoch": 0.17438187497436078, + "grad_norm": 2.702643871307373, + "learning_rate": 0.0001853968122601026, + "loss": 0.4038, + "step": 7439 + }, + { + "epoch": 0.17440531654916577, + "grad_norm": 0.7150228023529053, + "learning_rate": 0.00018539297922994324, + "loss": 0.0776, + "step": 7440 + }, + { + "epoch": 0.17442875812397077, + "grad_norm": 0.6488556861877441, + "learning_rate": 0.00018538914573643916, + "loss": 0.0686, + "step": 7441 + }, + { + "epoch": 0.17445219969877576, + "grad_norm": 1.5595736503601074, + "learning_rate": 0.00018538531177961123, + "loss": 0.2668, + "step": 7442 + }, + { + "epoch": 0.17447564127358076, + "grad_norm": 3.5419223308563232, + "learning_rate": 0.00018538147735948016, + "loss": 0.5738, + "step": 7443 + }, + { + "epoch": 0.17449908284838575, + "grad_norm": 3.6624956130981445, + "learning_rate": 0.00018537764247606678, + "loss": 0.2995, + "step": 7444 + }, + { + "epoch": 0.17452252442319074, + "grad_norm": 7.2621026039123535, + "learning_rate": 0.00018537380712939193, + "loss": 0.6441, + "step": 7445 + }, + { + "epoch": 0.17454596599799574, + "grad_norm": 4.340719699859619, + "learning_rate": 0.0001853699713194764, + "loss": 0.4558, + "step": 7446 + }, + { + "epoch": 0.17456940757280073, + "grad_norm": 4.398372173309326, + "learning_rate": 0.00018536613504634099, + "loss": 0.587, + "step": 7447 + }, + { + "epoch": 0.17459284914760573, + "grad_norm": 3.7512059211730957, + "learning_rate": 0.00018536229831000656, + "loss": 0.9892, + "step": 7448 + }, + { + "epoch": 0.17461629072241072, + "grad_norm": 1.0588274002075195, + "learning_rate": 0.0001853584611104939, + "loss": 0.1449, + "step": 7449 + }, + { + "epoch": 0.17463973229721572, + "grad_norm": 3.9517321586608887, + "learning_rate": 0.0001853546234478238, + "loss": 0.3728, + "step": 7450 + }, + { + "epoch": 0.1746631738720207, + "grad_norm": 1.8486579656600952, + "learning_rate": 0.0001853507853220171, + "loss": 1.1471, + "step": 7451 + }, + { + "epoch": 0.1746866154468257, + "grad_norm": 3.3581602573394775, + "learning_rate": 0.00018534694673309467, + "loss": 0.4951, + "step": 7452 + }, + { + "epoch": 0.17471005702163073, + "grad_norm": 4.506094932556152, + "learning_rate": 0.0001853431076810773, + "loss": 0.9299, + "step": 7453 + }, + { + "epoch": 0.17473349859643572, + "grad_norm": 3.3179423809051514, + "learning_rate": 0.00018533926816598583, + "loss": 0.4044, + "step": 7454 + }, + { + "epoch": 0.17475694017124072, + "grad_norm": 4.152040481567383, + "learning_rate": 0.0001853354281878411, + "loss": 0.7187, + "step": 7455 + }, + { + "epoch": 0.1747803817460457, + "grad_norm": 4.080030918121338, + "learning_rate": 0.00018533158774666392, + "loss": 0.6086, + "step": 7456 + }, + { + "epoch": 0.1748038233208507, + "grad_norm": 2.051302433013916, + "learning_rate": 0.00018532774684247517, + "loss": 0.258, + "step": 7457 + }, + { + "epoch": 0.1748272648956557, + "grad_norm": 3.329591989517212, + "learning_rate": 0.00018532390547529564, + "loss": 0.442, + "step": 7458 + }, + { + "epoch": 0.1748507064704607, + "grad_norm": 1.6798173189163208, + "learning_rate": 0.0001853200636451462, + "loss": 0.1836, + "step": 7459 + }, + { + "epoch": 0.1748741480452657, + "grad_norm": 1.628954291343689, + "learning_rate": 0.0001853162213520477, + "loss": 0.1811, + "step": 7460 + }, + { + "epoch": 0.17489758962007068, + "grad_norm": 1.8658241033554077, + "learning_rate": 0.00018531237859602102, + "loss": 0.2026, + "step": 7461 + }, + { + "epoch": 0.17492103119487568, + "grad_norm": 4.6830830574035645, + "learning_rate": 0.00018530853537708692, + "loss": 0.4859, + "step": 7462 + }, + { + "epoch": 0.17494447276968067, + "grad_norm": 3.541849374771118, + "learning_rate": 0.00018530469169526637, + "loss": 0.3837, + "step": 7463 + }, + { + "epoch": 0.17496791434448566, + "grad_norm": 0.5930203199386597, + "learning_rate": 0.0001853008475505801, + "loss": 0.0688, + "step": 7464 + }, + { + "epoch": 0.17499135591929066, + "grad_norm": 7.014890193939209, + "learning_rate": 0.00018529700294304906, + "loss": 0.4437, + "step": 7465 + }, + { + "epoch": 0.17501479749409565, + "grad_norm": 3.6147146224975586, + "learning_rate": 0.00018529315787269411, + "loss": 0.6346, + "step": 7466 + }, + { + "epoch": 0.17503823906890065, + "grad_norm": 1.4463415145874023, + "learning_rate": 0.00018528931233953606, + "loss": 0.1313, + "step": 7467 + }, + { + "epoch": 0.17506168064370564, + "grad_norm": 6.445090293884277, + "learning_rate": 0.0001852854663435958, + "loss": 0.3993, + "step": 7468 + }, + { + "epoch": 0.17508512221851064, + "grad_norm": 2.6231329441070557, + "learning_rate": 0.0001852816198848942, + "loss": 1.853, + "step": 7469 + }, + { + "epoch": 0.17510856379331563, + "grad_norm": 1.59513521194458, + "learning_rate": 0.00018527777296345214, + "loss": 0.1712, + "step": 7470 + }, + { + "epoch": 0.17513200536812062, + "grad_norm": 3.9317312240600586, + "learning_rate": 0.0001852739255792905, + "loss": 0.5483, + "step": 7471 + }, + { + "epoch": 0.17515544694292562, + "grad_norm": 3.3264822959899902, + "learning_rate": 0.0001852700777324301, + "loss": 0.4493, + "step": 7472 + }, + { + "epoch": 0.1751788885177306, + "grad_norm": 1.3919799327850342, + "learning_rate": 0.0001852662294228919, + "loss": 0.2092, + "step": 7473 + }, + { + "epoch": 0.1752023300925356, + "grad_norm": 0.3675731420516968, + "learning_rate": 0.00018526238065069673, + "loss": 0.0368, + "step": 7474 + }, + { + "epoch": 0.1752257716673406, + "grad_norm": 3.596566677093506, + "learning_rate": 0.0001852585314158655, + "loss": 0.7144, + "step": 7475 + }, + { + "epoch": 0.1752492132421456, + "grad_norm": 4.035824775695801, + "learning_rate": 0.00018525468171841903, + "loss": 0.3719, + "step": 7476 + }, + { + "epoch": 0.1752726548169506, + "grad_norm": 2.8365626335144043, + "learning_rate": 0.0001852508315583783, + "loss": 0.2652, + "step": 7477 + }, + { + "epoch": 0.17529609639175558, + "grad_norm": 3.174607276916504, + "learning_rate": 0.00018524698093576417, + "loss": 0.489, + "step": 7478 + }, + { + "epoch": 0.1753195379665606, + "grad_norm": 1.5347671508789062, + "learning_rate": 0.0001852431298505975, + "loss": 0.2602, + "step": 7479 + }, + { + "epoch": 0.1753429795413656, + "grad_norm": 3.992595911026001, + "learning_rate": 0.00018523927830289925, + "loss": 1.8754, + "step": 7480 + }, + { + "epoch": 0.1753664211161706, + "grad_norm": 5.644362926483154, + "learning_rate": 0.00018523542629269023, + "loss": 0.6037, + "step": 7481 + }, + { + "epoch": 0.1753898626909756, + "grad_norm": 2.397017478942871, + "learning_rate": 0.00018523157381999142, + "loss": 0.2053, + "step": 7482 + }, + { + "epoch": 0.17541330426578058, + "grad_norm": 3.2418324947357178, + "learning_rate": 0.0001852277208848237, + "loss": 0.4113, + "step": 7483 + }, + { + "epoch": 0.17543674584058558, + "grad_norm": 3.1870901584625244, + "learning_rate": 0.00018522386748720795, + "loss": 0.3026, + "step": 7484 + }, + { + "epoch": 0.17546018741539057, + "grad_norm": 3.4119815826416016, + "learning_rate": 0.0001852200136271651, + "loss": 0.5343, + "step": 7485 + }, + { + "epoch": 0.17548362899019557, + "grad_norm": 1.4589177370071411, + "learning_rate": 0.00018521615930471605, + "loss": 0.1812, + "step": 7486 + }, + { + "epoch": 0.17550707056500056, + "grad_norm": 1.7809621095657349, + "learning_rate": 0.00018521230451988175, + "loss": 1.315, + "step": 7487 + }, + { + "epoch": 0.17553051213980556, + "grad_norm": 3.1948859691619873, + "learning_rate": 0.00018520844927268307, + "loss": 0.5078, + "step": 7488 + }, + { + "epoch": 0.17555395371461055, + "grad_norm": 1.5550307035446167, + "learning_rate": 0.00018520459356314095, + "loss": 1.3837, + "step": 7489 + }, + { + "epoch": 0.17557739528941554, + "grad_norm": 2.015099048614502, + "learning_rate": 0.0001852007373912763, + "loss": 0.3617, + "step": 7490 + }, + { + "epoch": 0.17560083686422054, + "grad_norm": 1.204246163368225, + "learning_rate": 0.00018519688075711011, + "loss": 0.1735, + "step": 7491 + }, + { + "epoch": 0.17562427843902553, + "grad_norm": 0.20102837681770325, + "learning_rate": 0.0001851930236606632, + "loss": 0.0112, + "step": 7492 + }, + { + "epoch": 0.17564772001383053, + "grad_norm": 3.165010929107666, + "learning_rate": 0.00018518916610195654, + "loss": 0.5077, + "step": 7493 + }, + { + "epoch": 0.17567116158863552, + "grad_norm": 2.4412624835968018, + "learning_rate": 0.0001851853080810111, + "loss": 0.2506, + "step": 7494 + }, + { + "epoch": 0.17569460316344052, + "grad_norm": 1.2155674695968628, + "learning_rate": 0.00018518144959784775, + "loss": 0.1496, + "step": 7495 + }, + { + "epoch": 0.1757180447382455, + "grad_norm": 0.7813835740089417, + "learning_rate": 0.00018517759065248747, + "loss": 0.1009, + "step": 7496 + }, + { + "epoch": 0.1757414863130505, + "grad_norm": 2.2838664054870605, + "learning_rate": 0.0001851737312449512, + "loss": 0.2407, + "step": 7497 + }, + { + "epoch": 0.1757649278878555, + "grad_norm": 3.3614859580993652, + "learning_rate": 0.00018516987137525987, + "loss": 0.3676, + "step": 7498 + }, + { + "epoch": 0.1757883694626605, + "grad_norm": 4.05367374420166, + "learning_rate": 0.00018516601104343443, + "loss": 0.2726, + "step": 7499 + }, + { + "epoch": 0.1758118110374655, + "grad_norm": 1.5923576354980469, + "learning_rate": 0.0001851621502494958, + "loss": 0.2567, + "step": 7500 + }, + { + "epoch": 0.17583525261227048, + "grad_norm": 1.6686216592788696, + "learning_rate": 0.00018515828899346497, + "loss": 0.3148, + "step": 7501 + }, + { + "epoch": 0.17585869418707548, + "grad_norm": 0.5886324644088745, + "learning_rate": 0.00018515442727536286, + "loss": 0.0829, + "step": 7502 + }, + { + "epoch": 0.17588213576188047, + "grad_norm": 0.3562447726726532, + "learning_rate": 0.00018515056509521043, + "loss": 0.0346, + "step": 7503 + }, + { + "epoch": 0.17590557733668546, + "grad_norm": 2.375408411026001, + "learning_rate": 0.00018514670245302863, + "loss": 0.1572, + "step": 7504 + }, + { + "epoch": 0.1759290189114905, + "grad_norm": 2.6985955238342285, + "learning_rate": 0.00018514283934883844, + "loss": 0.2029, + "step": 7505 + }, + { + "epoch": 0.17595246048629548, + "grad_norm": 3.555413246154785, + "learning_rate": 0.00018513897578266078, + "loss": 0.5114, + "step": 7506 + }, + { + "epoch": 0.17597590206110048, + "grad_norm": 0.7900478839874268, + "learning_rate": 0.00018513511175451667, + "loss": 0.1152, + "step": 7507 + }, + { + "epoch": 0.17599934363590547, + "grad_norm": 2.9512898921966553, + "learning_rate": 0.00018513124726442706, + "loss": 1.9434, + "step": 7508 + }, + { + "epoch": 0.17602278521071046, + "grad_norm": 8.015401840209961, + "learning_rate": 0.00018512738231241292, + "loss": 0.4827, + "step": 7509 + }, + { + "epoch": 0.17604622678551546, + "grad_norm": 4.026829242706299, + "learning_rate": 0.00018512351689849518, + "loss": 0.445, + "step": 7510 + }, + { + "epoch": 0.17606966836032045, + "grad_norm": 6.306526184082031, + "learning_rate": 0.00018511965102269487, + "loss": 0.9189, + "step": 7511 + }, + { + "epoch": 0.17609310993512545, + "grad_norm": 4.318136692047119, + "learning_rate": 0.0001851157846850329, + "loss": 0.726, + "step": 7512 + }, + { + "epoch": 0.17611655150993044, + "grad_norm": 6.959718704223633, + "learning_rate": 0.00018511191788553032, + "loss": 0.9324, + "step": 7513 + }, + { + "epoch": 0.17613999308473544, + "grad_norm": 3.7295725345611572, + "learning_rate": 0.00018510805062420805, + "loss": 0.7687, + "step": 7514 + }, + { + "epoch": 0.17616343465954043, + "grad_norm": 0.45214325189590454, + "learning_rate": 0.00018510418290108716, + "loss": 0.0539, + "step": 7515 + }, + { + "epoch": 0.17618687623434542, + "grad_norm": 1.4230390787124634, + "learning_rate": 0.00018510031471618855, + "loss": 0.2091, + "step": 7516 + }, + { + "epoch": 0.17621031780915042, + "grad_norm": 4.244172096252441, + "learning_rate": 0.0001850964460695332, + "loss": 0.3526, + "step": 7517 + }, + { + "epoch": 0.1762337593839554, + "grad_norm": 1.822052240371704, + "learning_rate": 0.0001850925769611422, + "loss": 0.1877, + "step": 7518 + }, + { + "epoch": 0.1762572009587604, + "grad_norm": 3.7298154830932617, + "learning_rate": 0.00018508870739103645, + "loss": 0.304, + "step": 7519 + }, + { + "epoch": 0.1762806425335654, + "grad_norm": 5.0709757804870605, + "learning_rate": 0.000185084837359237, + "loss": 0.3746, + "step": 7520 + }, + { + "epoch": 0.1763040841083704, + "grad_norm": 4.094529628753662, + "learning_rate": 0.00018508096686576484, + "loss": 0.5903, + "step": 7521 + }, + { + "epoch": 0.1763275256831754, + "grad_norm": 1.328049898147583, + "learning_rate": 0.00018507709591064094, + "loss": 0.9525, + "step": 7522 + }, + { + "epoch": 0.17635096725798038, + "grad_norm": 4.193729877471924, + "learning_rate": 0.00018507322449388633, + "loss": 0.3714, + "step": 7523 + }, + { + "epoch": 0.17637440883278538, + "grad_norm": 3.410963773727417, + "learning_rate": 0.000185069352615522, + "loss": 0.5259, + "step": 7524 + }, + { + "epoch": 0.17639785040759037, + "grad_norm": 1.2584691047668457, + "learning_rate": 0.00018506548027556898, + "loss": 0.1895, + "step": 7525 + }, + { + "epoch": 0.17642129198239537, + "grad_norm": 3.1800010204315186, + "learning_rate": 0.0001850616074740483, + "loss": 0.3578, + "step": 7526 + }, + { + "epoch": 0.17644473355720036, + "grad_norm": 2.845639705657959, + "learning_rate": 0.00018505773421098089, + "loss": 0.4243, + "step": 7527 + }, + { + "epoch": 0.17646817513200536, + "grad_norm": 4.27915096282959, + "learning_rate": 0.0001850538604863879, + "loss": 0.6559, + "step": 7528 + }, + { + "epoch": 0.17649161670681035, + "grad_norm": 1.823394775390625, + "learning_rate": 0.0001850499863002902, + "loss": 2.1907, + "step": 7529 + }, + { + "epoch": 0.17651505828161537, + "grad_norm": 2.882761001586914, + "learning_rate": 0.0001850461116527089, + "loss": 0.3239, + "step": 7530 + }, + { + "epoch": 0.17653849985642037, + "grad_norm": 4.452301979064941, + "learning_rate": 0.00018504223654366503, + "loss": 0.5104, + "step": 7531 + }, + { + "epoch": 0.17656194143122536, + "grad_norm": 3.8173117637634277, + "learning_rate": 0.00018503836097317957, + "loss": 0.6515, + "step": 7532 + }, + { + "epoch": 0.17658538300603036, + "grad_norm": 5.062869071960449, + "learning_rate": 0.00018503448494127357, + "loss": 0.7248, + "step": 7533 + }, + { + "epoch": 0.17660882458083535, + "grad_norm": 1.807593822479248, + "learning_rate": 0.0001850306084479681, + "loss": 1.0691, + "step": 7534 + }, + { + "epoch": 0.17663226615564034, + "grad_norm": 3.975198268890381, + "learning_rate": 0.00018502673149328413, + "loss": 0.5695, + "step": 7535 + }, + { + "epoch": 0.17665570773044534, + "grad_norm": 1.1530272960662842, + "learning_rate": 0.0001850228540772427, + "loss": 0.1311, + "step": 7536 + }, + { + "epoch": 0.17667914930525033, + "grad_norm": 2.747549295425415, + "learning_rate": 0.00018501897619986493, + "loss": 0.2361, + "step": 7537 + }, + { + "epoch": 0.17670259088005533, + "grad_norm": 3.705885171890259, + "learning_rate": 0.00018501509786117177, + "loss": 0.5667, + "step": 7538 + }, + { + "epoch": 0.17672603245486032, + "grad_norm": 3.8229968547821045, + "learning_rate": 0.0001850112190611843, + "loss": 0.3391, + "step": 7539 + }, + { + "epoch": 0.17674947402966532, + "grad_norm": 5.1686482429504395, + "learning_rate": 0.00018500733979992355, + "loss": 0.5937, + "step": 7540 + }, + { + "epoch": 0.1767729156044703, + "grad_norm": 0.7915934920310974, + "learning_rate": 0.0001850034600774106, + "loss": 0.0778, + "step": 7541 + }, + { + "epoch": 0.1767963571792753, + "grad_norm": 2.229332208633423, + "learning_rate": 0.00018499957989366651, + "loss": 0.1261, + "step": 7542 + }, + { + "epoch": 0.1768197987540803, + "grad_norm": 5.05634880065918, + "learning_rate": 0.0001849956992487123, + "loss": 0.6438, + "step": 7543 + }, + { + "epoch": 0.1768432403288853, + "grad_norm": 3.98004412651062, + "learning_rate": 0.000184991818142569, + "loss": 0.4724, + "step": 7544 + }, + { + "epoch": 0.1768666819036903, + "grad_norm": 5.0156965255737305, + "learning_rate": 0.00018498793657525773, + "loss": 0.5569, + "step": 7545 + }, + { + "epoch": 0.17689012347849528, + "grad_norm": 2.2214419841766357, + "learning_rate": 0.00018498405454679952, + "loss": 0.1233, + "step": 7546 + }, + { + "epoch": 0.17691356505330028, + "grad_norm": 3.6738858222961426, + "learning_rate": 0.00018498017205721546, + "loss": 0.4168, + "step": 7547 + }, + { + "epoch": 0.17693700662810527, + "grad_norm": 3.7562994956970215, + "learning_rate": 0.0001849762891065266, + "loss": 0.5488, + "step": 7548 + }, + { + "epoch": 0.17696044820291026, + "grad_norm": 1.125771164894104, + "learning_rate": 0.00018497240569475399, + "loss": 0.1543, + "step": 7549 + }, + { + "epoch": 0.17698388977771526, + "grad_norm": 1.065678596496582, + "learning_rate": 0.0001849685218219187, + "loss": 0.1398, + "step": 7550 + }, + { + "epoch": 0.17700733135252025, + "grad_norm": 3.8345749378204346, + "learning_rate": 0.00018496463748804182, + "loss": 0.5255, + "step": 7551 + }, + { + "epoch": 0.17703077292732525, + "grad_norm": 1.1521961688995361, + "learning_rate": 0.00018496075269314446, + "loss": 0.1792, + "step": 7552 + }, + { + "epoch": 0.17705421450213024, + "grad_norm": 3.3279948234558105, + "learning_rate": 0.00018495686743724764, + "loss": 0.6426, + "step": 7553 + }, + { + "epoch": 0.17707765607693524, + "grad_norm": 4.1458611488342285, + "learning_rate": 0.00018495298172037248, + "loss": 0.5279, + "step": 7554 + }, + { + "epoch": 0.17710109765174023, + "grad_norm": 3.9096405506134033, + "learning_rate": 0.00018494909554254006, + "loss": 0.3958, + "step": 7555 + }, + { + "epoch": 0.17712453922654525, + "grad_norm": 2.784217596054077, + "learning_rate": 0.00018494520890377145, + "loss": 0.4618, + "step": 7556 + }, + { + "epoch": 0.17714798080135025, + "grad_norm": 4.713284492492676, + "learning_rate": 0.00018494132180408776, + "loss": 0.6621, + "step": 7557 + }, + { + "epoch": 0.17717142237615524, + "grad_norm": 3.829817533493042, + "learning_rate": 0.00018493743424351005, + "loss": 0.4166, + "step": 7558 + }, + { + "epoch": 0.17719486395096024, + "grad_norm": 3.7031095027923584, + "learning_rate": 0.00018493354622205944, + "loss": 0.4493, + "step": 7559 + }, + { + "epoch": 0.17721830552576523, + "grad_norm": 4.523824691772461, + "learning_rate": 0.00018492965773975705, + "loss": 0.686, + "step": 7560 + }, + { + "epoch": 0.17724174710057022, + "grad_norm": 3.49696683883667, + "learning_rate": 0.00018492576879662392, + "loss": 0.3615, + "step": 7561 + }, + { + "epoch": 0.17726518867537522, + "grad_norm": 3.749058246612549, + "learning_rate": 0.00018492187939268115, + "loss": 0.5386, + "step": 7562 + }, + { + "epoch": 0.1772886302501802, + "grad_norm": 3.6249306201934814, + "learning_rate": 0.00018491798952794991, + "loss": 0.3981, + "step": 7563 + }, + { + "epoch": 0.1773120718249852, + "grad_norm": 0.7801098823547363, + "learning_rate": 0.0001849140992024513, + "loss": 0.0954, + "step": 7564 + }, + { + "epoch": 0.1773355133997902, + "grad_norm": 4.452328681945801, + "learning_rate": 0.00018491020841620635, + "loss": 0.6151, + "step": 7565 + }, + { + "epoch": 0.1773589549745952, + "grad_norm": 1.0716465711593628, + "learning_rate": 0.00018490631716923622, + "loss": 0.1447, + "step": 7566 + }, + { + "epoch": 0.1773823965494002, + "grad_norm": 3.850545883178711, + "learning_rate": 0.00018490242546156206, + "loss": 0.4565, + "step": 7567 + }, + { + "epoch": 0.17740583812420518, + "grad_norm": 3.734919548034668, + "learning_rate": 0.00018489853329320494, + "loss": 0.4349, + "step": 7568 + }, + { + "epoch": 0.17742927969901018, + "grad_norm": 0.9347181916236877, + "learning_rate": 0.000184894640664186, + "loss": 0.0963, + "step": 7569 + }, + { + "epoch": 0.17745272127381517, + "grad_norm": 3.489861488342285, + "learning_rate": 0.00018489074757452634, + "loss": 1.3051, + "step": 7570 + }, + { + "epoch": 0.17747616284862017, + "grad_norm": 3.263580799102783, + "learning_rate": 0.0001848868540242471, + "loss": 0.5344, + "step": 7571 + }, + { + "epoch": 0.17749960442342516, + "grad_norm": 1.3446416854858398, + "learning_rate": 0.00018488296001336943, + "loss": 0.1732, + "step": 7572 + }, + { + "epoch": 0.17752304599823016, + "grad_norm": 2.6759660243988037, + "learning_rate": 0.0001848790655419144, + "loss": 0.4251, + "step": 7573 + }, + { + "epoch": 0.17754648757303515, + "grad_norm": 0.8745677471160889, + "learning_rate": 0.00018487517060990317, + "loss": 0.1009, + "step": 7574 + }, + { + "epoch": 0.17756992914784014, + "grad_norm": 3.196749210357666, + "learning_rate": 0.00018487127521735688, + "loss": 0.4147, + "step": 7575 + }, + { + "epoch": 0.17759337072264514, + "grad_norm": 4.252834796905518, + "learning_rate": 0.00018486737936429667, + "loss": 0.6619, + "step": 7576 + }, + { + "epoch": 0.17761681229745013, + "grad_norm": 1.0415124893188477, + "learning_rate": 0.00018486348305074367, + "loss": 0.134, + "step": 7577 + }, + { + "epoch": 0.17764025387225513, + "grad_norm": 1.2324280738830566, + "learning_rate": 0.000184859586276719, + "loss": 0.0638, + "step": 7578 + }, + { + "epoch": 0.17766369544706012, + "grad_norm": 5.194002628326416, + "learning_rate": 0.00018485568904224388, + "loss": 0.9051, + "step": 7579 + }, + { + "epoch": 0.17768713702186512, + "grad_norm": 2.9424242973327637, + "learning_rate": 0.0001848517913473394, + "loss": 0.331, + "step": 7580 + }, + { + "epoch": 0.17771057859667014, + "grad_norm": 2.9222214221954346, + "learning_rate": 0.0001848478931920267, + "loss": 0.6274, + "step": 7581 + }, + { + "epoch": 0.17773402017147513, + "grad_norm": 0.6493744850158691, + "learning_rate": 0.00018484399457632694, + "loss": 0.0594, + "step": 7582 + }, + { + "epoch": 0.17775746174628013, + "grad_norm": 1.2816500663757324, + "learning_rate": 0.0001848400955002613, + "loss": 0.252, + "step": 7583 + }, + { + "epoch": 0.17778090332108512, + "grad_norm": 3.422057628631592, + "learning_rate": 0.00018483619596385088, + "loss": 0.5175, + "step": 7584 + }, + { + "epoch": 0.17780434489589012, + "grad_norm": 2.366410255432129, + "learning_rate": 0.0001848322959671169, + "loss": 0.3464, + "step": 7585 + }, + { + "epoch": 0.1778277864706951, + "grad_norm": 2.2450993061065674, + "learning_rate": 0.00018482839551008047, + "loss": 0.3367, + "step": 7586 + }, + { + "epoch": 0.1778512280455001, + "grad_norm": 3.2632603645324707, + "learning_rate": 0.00018482449459276281, + "loss": 1.8684, + "step": 7587 + }, + { + "epoch": 0.1778746696203051, + "grad_norm": 8.201645851135254, + "learning_rate": 0.00018482059321518505, + "loss": 0.9752, + "step": 7588 + }, + { + "epoch": 0.1778981111951101, + "grad_norm": 2.3456428050994873, + "learning_rate": 0.00018481669137736836, + "loss": 0.256, + "step": 7589 + }, + { + "epoch": 0.1779215527699151, + "grad_norm": 2.7904775142669678, + "learning_rate": 0.00018481278907933396, + "loss": 0.4011, + "step": 7590 + }, + { + "epoch": 0.17794499434472008, + "grad_norm": 2.088571548461914, + "learning_rate": 0.00018480888632110292, + "loss": 1.6692, + "step": 7591 + }, + { + "epoch": 0.17796843591952508, + "grad_norm": 4.337948799133301, + "learning_rate": 0.00018480498310269653, + "loss": 0.7658, + "step": 7592 + }, + { + "epoch": 0.17799187749433007, + "grad_norm": 3.3832204341888428, + "learning_rate": 0.00018480107942413587, + "loss": 0.4389, + "step": 7593 + }, + { + "epoch": 0.17801531906913506, + "grad_norm": 4.058834075927734, + "learning_rate": 0.0001847971752854422, + "loss": 0.9637, + "step": 7594 + }, + { + "epoch": 0.17803876064394006, + "grad_norm": 2.8309006690979004, + "learning_rate": 0.00018479327068663664, + "loss": 2.4012, + "step": 7595 + }, + { + "epoch": 0.17806220221874505, + "grad_norm": 3.8187448978424072, + "learning_rate": 0.00018478936562774044, + "loss": 0.69, + "step": 7596 + }, + { + "epoch": 0.17808564379355005, + "grad_norm": 4.139993667602539, + "learning_rate": 0.00018478546010877476, + "loss": 0.5771, + "step": 7597 + }, + { + "epoch": 0.17810908536835504, + "grad_norm": 3.8790881633758545, + "learning_rate": 0.00018478155412976076, + "loss": 0.5413, + "step": 7598 + }, + { + "epoch": 0.17813252694316004, + "grad_norm": 4.673426628112793, + "learning_rate": 0.0001847776476907197, + "loss": 0.5232, + "step": 7599 + }, + { + "epoch": 0.17815596851796503, + "grad_norm": 4.148377895355225, + "learning_rate": 0.00018477374079167272, + "loss": 0.6294, + "step": 7600 + }, + { + "epoch": 0.17817941009277002, + "grad_norm": 3.85493803024292, + "learning_rate": 0.00018476983343264104, + "loss": 0.5173, + "step": 7601 + }, + { + "epoch": 0.17820285166757502, + "grad_norm": 4.2015509605407715, + "learning_rate": 0.00018476592561364585, + "loss": 0.4663, + "step": 7602 + }, + { + "epoch": 0.17822629324238, + "grad_norm": 2.2440855503082275, + "learning_rate": 0.0001847620173347084, + "loss": 0.4063, + "step": 7603 + }, + { + "epoch": 0.178249734817185, + "grad_norm": 3.6690878868103027, + "learning_rate": 0.00018475810859584982, + "loss": 0.4685, + "step": 7604 + }, + { + "epoch": 0.17827317639199, + "grad_norm": 1.1707855463027954, + "learning_rate": 0.00018475419939709137, + "loss": 0.1258, + "step": 7605 + }, + { + "epoch": 0.178296617966795, + "grad_norm": 4.676024913787842, + "learning_rate": 0.00018475028973845425, + "loss": 0.4794, + "step": 7606 + }, + { + "epoch": 0.17832005954160002, + "grad_norm": 2.6180646419525146, + "learning_rate": 0.0001847463796199597, + "loss": 0.2103, + "step": 7607 + }, + { + "epoch": 0.178343501116405, + "grad_norm": 1.6610658168792725, + "learning_rate": 0.0001847424690416289, + "loss": 0.1957, + "step": 7608 + }, + { + "epoch": 0.17836694269121, + "grad_norm": 3.1924757957458496, + "learning_rate": 0.00018473855800348307, + "loss": 0.4741, + "step": 7609 + }, + { + "epoch": 0.178390384266015, + "grad_norm": 1.1958268880844116, + "learning_rate": 0.00018473464650554342, + "loss": 0.1096, + "step": 7610 + }, + { + "epoch": 0.17841382584082, + "grad_norm": 5.831681728363037, + "learning_rate": 0.00018473073454783123, + "loss": 0.7317, + "step": 7611 + }, + { + "epoch": 0.178437267415625, + "grad_norm": 5.471249103546143, + "learning_rate": 0.00018472682213036768, + "loss": 0.9038, + "step": 7612 + }, + { + "epoch": 0.17846070899042998, + "grad_norm": 4.547652244567871, + "learning_rate": 0.000184722909253174, + "loss": 0.4905, + "step": 7613 + }, + { + "epoch": 0.17848415056523498, + "grad_norm": 1.1849945783615112, + "learning_rate": 0.00018471899591627144, + "loss": 0.1726, + "step": 7614 + }, + { + "epoch": 0.17850759214003997, + "grad_norm": 2.3355932235717773, + "learning_rate": 0.00018471508211968124, + "loss": 0.2192, + "step": 7615 + }, + { + "epoch": 0.17853103371484497, + "grad_norm": 3.7327678203582764, + "learning_rate": 0.0001847111678634246, + "loss": 0.174, + "step": 7616 + }, + { + "epoch": 0.17855447528964996, + "grad_norm": 5.58084774017334, + "learning_rate": 0.00018470725314752282, + "loss": 0.5377, + "step": 7617 + }, + { + "epoch": 0.17857791686445496, + "grad_norm": 4.991566181182861, + "learning_rate": 0.00018470333797199708, + "loss": 0.824, + "step": 7618 + }, + { + "epoch": 0.17860135843925995, + "grad_norm": 2.066537380218506, + "learning_rate": 0.00018469942233686863, + "loss": 1.693, + "step": 7619 + }, + { + "epoch": 0.17862480001406494, + "grad_norm": 3.8446784019470215, + "learning_rate": 0.00018469550624215875, + "loss": 0.3288, + "step": 7620 + }, + { + "epoch": 0.17864824158886994, + "grad_norm": 3.595726490020752, + "learning_rate": 0.0001846915896878887, + "loss": 0.2655, + "step": 7621 + }, + { + "epoch": 0.17867168316367493, + "grad_norm": 1.9958113431930542, + "learning_rate": 0.00018468767267407963, + "loss": 1.289, + "step": 7622 + }, + { + "epoch": 0.17869512473847993, + "grad_norm": 3.1739895343780518, + "learning_rate": 0.00018468375520075295, + "loss": 0.4647, + "step": 7623 + }, + { + "epoch": 0.17871856631328492, + "grad_norm": 2.087878704071045, + "learning_rate": 0.00018467983726792978, + "loss": 0.2114, + "step": 7624 + }, + { + "epoch": 0.17874200788808992, + "grad_norm": 2.5972535610198975, + "learning_rate": 0.00018467591887563143, + "loss": 1.8483, + "step": 7625 + }, + { + "epoch": 0.1787654494628949, + "grad_norm": 1.20168936252594, + "learning_rate": 0.00018467200002387916, + "loss": 0.1353, + "step": 7626 + }, + { + "epoch": 0.1787888910376999, + "grad_norm": 3.056978225708008, + "learning_rate": 0.0001846680807126943, + "loss": 0.415, + "step": 7627 + }, + { + "epoch": 0.1788123326125049, + "grad_norm": 3.247296094894409, + "learning_rate": 0.000184664160942098, + "loss": 0.4226, + "step": 7628 + }, + { + "epoch": 0.1788357741873099, + "grad_norm": 4.098209381103516, + "learning_rate": 0.00018466024071211158, + "loss": 0.4229, + "step": 7629 + }, + { + "epoch": 0.1788592157621149, + "grad_norm": 3.7995080947875977, + "learning_rate": 0.00018465632002275632, + "loss": 0.4994, + "step": 7630 + }, + { + "epoch": 0.17888265733691988, + "grad_norm": 4.7607855796813965, + "learning_rate": 0.0001846523988740535, + "loss": 0.9576, + "step": 7631 + }, + { + "epoch": 0.1789060989117249, + "grad_norm": 4.406134605407715, + "learning_rate": 0.00018464847726602437, + "loss": 0.6719, + "step": 7632 + }, + { + "epoch": 0.1789295404865299, + "grad_norm": 0.9775751829147339, + "learning_rate": 0.00018464455519869022, + "loss": 0.1082, + "step": 7633 + }, + { + "epoch": 0.1789529820613349, + "grad_norm": 0.37647727131843567, + "learning_rate": 0.00018464063267207233, + "loss": 0.0382, + "step": 7634 + }, + { + "epoch": 0.1789764236361399, + "grad_norm": 2.406374931335449, + "learning_rate": 0.000184636709686192, + "loss": 1.7089, + "step": 7635 + }, + { + "epoch": 0.17899986521094488, + "grad_norm": 2.6104447841644287, + "learning_rate": 0.0001846327862410705, + "loss": 1.4504, + "step": 7636 + }, + { + "epoch": 0.17902330678574987, + "grad_norm": 4.822855472564697, + "learning_rate": 0.00018462886233672915, + "loss": 0.6137, + "step": 7637 + }, + { + "epoch": 0.17904674836055487, + "grad_norm": 3.808997392654419, + "learning_rate": 0.00018462493797318915, + "loss": 0.4157, + "step": 7638 + }, + { + "epoch": 0.17907018993535986, + "grad_norm": 5.631366729736328, + "learning_rate": 0.00018462101315047192, + "loss": 0.7256, + "step": 7639 + }, + { + "epoch": 0.17909363151016486, + "grad_norm": 2.3259530067443848, + "learning_rate": 0.00018461708786859867, + "loss": 1.6957, + "step": 7640 + }, + { + "epoch": 0.17911707308496985, + "grad_norm": 1.456604242324829, + "learning_rate": 0.0001846131621275907, + "loss": 0.1435, + "step": 7641 + }, + { + "epoch": 0.17914051465977485, + "grad_norm": 3.365809440612793, + "learning_rate": 0.00018460923592746937, + "loss": 1.4114, + "step": 7642 + }, + { + "epoch": 0.17916395623457984, + "grad_norm": 3.9830188751220703, + "learning_rate": 0.0001846053092682559, + "loss": 0.4831, + "step": 7643 + }, + { + "epoch": 0.17918739780938484, + "grad_norm": 1.4492765665054321, + "learning_rate": 0.0001846013821499717, + "loss": 0.2422, + "step": 7644 + }, + { + "epoch": 0.17921083938418983, + "grad_norm": 3.6714746952056885, + "learning_rate": 0.00018459745457263796, + "loss": 0.5845, + "step": 7645 + }, + { + "epoch": 0.17923428095899482, + "grad_norm": 2.826793670654297, + "learning_rate": 0.0001845935265362761, + "loss": 1.6543, + "step": 7646 + }, + { + "epoch": 0.17925772253379982, + "grad_norm": 0.7964304089546204, + "learning_rate": 0.00018458959804090735, + "loss": 0.0885, + "step": 7647 + }, + { + "epoch": 0.1792811641086048, + "grad_norm": 4.0048136711120605, + "learning_rate": 0.0001845856690865531, + "loss": 0.7217, + "step": 7648 + }, + { + "epoch": 0.1793046056834098, + "grad_norm": 3.5845890045166016, + "learning_rate": 0.00018458173967323465, + "loss": 0.4667, + "step": 7649 + }, + { + "epoch": 0.1793280472582148, + "grad_norm": 2.278355360031128, + "learning_rate": 0.00018457780980097324, + "loss": 0.3136, + "step": 7650 + }, + { + "epoch": 0.1793514888330198, + "grad_norm": 2.6399176120758057, + "learning_rate": 0.0001845738794697903, + "loss": 0.3549, + "step": 7651 + }, + { + "epoch": 0.1793749304078248, + "grad_norm": 0.796334981918335, + "learning_rate": 0.0001845699486797071, + "loss": 0.1194, + "step": 7652 + }, + { + "epoch": 0.17939837198262978, + "grad_norm": 4.0382490158081055, + "learning_rate": 0.00018456601743074497, + "loss": 0.471, + "step": 7653 + }, + { + "epoch": 0.17942181355743478, + "grad_norm": 2.090301275253296, + "learning_rate": 0.00018456208572292524, + "loss": 0.2605, + "step": 7654 + }, + { + "epoch": 0.17944525513223977, + "grad_norm": 0.6141961216926575, + "learning_rate": 0.0001845581535562693, + "loss": 0.1067, + "step": 7655 + }, + { + "epoch": 0.17946869670704477, + "grad_norm": 3.499063014984131, + "learning_rate": 0.0001845542209307984, + "loss": 0.5521, + "step": 7656 + }, + { + "epoch": 0.17949213828184976, + "grad_norm": 3.5388495922088623, + "learning_rate": 0.00018455028784653395, + "loss": 0.4746, + "step": 7657 + }, + { + "epoch": 0.17951557985665478, + "grad_norm": 4.052797794342041, + "learning_rate": 0.00018454635430349722, + "loss": 0.7159, + "step": 7658 + }, + { + "epoch": 0.17953902143145978, + "grad_norm": 1.0228524208068848, + "learning_rate": 0.00018454242030170964, + "loss": 0.1677, + "step": 7659 + }, + { + "epoch": 0.17956246300626477, + "grad_norm": 0.914531946182251, + "learning_rate": 0.0001845384858411925, + "loss": 0.1525, + "step": 7660 + }, + { + "epoch": 0.17958590458106977, + "grad_norm": 4.595582485198975, + "learning_rate": 0.00018453455092196715, + "loss": 0.7107, + "step": 7661 + }, + { + "epoch": 0.17960934615587476, + "grad_norm": 4.072437763214111, + "learning_rate": 0.00018453061554405493, + "loss": 0.4922, + "step": 7662 + }, + { + "epoch": 0.17963278773067975, + "grad_norm": 2.824446439743042, + "learning_rate": 0.00018452667970747725, + "loss": 1.2375, + "step": 7663 + }, + { + "epoch": 0.17965622930548475, + "grad_norm": 4.176913261413574, + "learning_rate": 0.0001845227434122554, + "loss": 0.6975, + "step": 7664 + }, + { + "epoch": 0.17967967088028974, + "grad_norm": 3.8739025592803955, + "learning_rate": 0.00018451880665841077, + "loss": 0.5115, + "step": 7665 + }, + { + "epoch": 0.17970311245509474, + "grad_norm": 1.782457709312439, + "learning_rate": 0.00018451486944596473, + "loss": 0.2841, + "step": 7666 + }, + { + "epoch": 0.17972655402989973, + "grad_norm": 4.5528178215026855, + "learning_rate": 0.00018451093177493863, + "loss": 0.6537, + "step": 7667 + }, + { + "epoch": 0.17974999560470473, + "grad_norm": 3.5891997814178467, + "learning_rate": 0.00018450699364535385, + "loss": 0.4566, + "step": 7668 + }, + { + "epoch": 0.17977343717950972, + "grad_norm": 2.334895372390747, + "learning_rate": 0.0001845030550572317, + "loss": 0.2374, + "step": 7669 + }, + { + "epoch": 0.17979687875431472, + "grad_norm": 4.6420440673828125, + "learning_rate": 0.00018449911601059365, + "loss": 0.5618, + "step": 7670 + }, + { + "epoch": 0.1798203203291197, + "grad_norm": 3.437105178833008, + "learning_rate": 0.000184495176505461, + "loss": 0.3128, + "step": 7671 + }, + { + "epoch": 0.1798437619039247, + "grad_norm": 4.106106281280518, + "learning_rate": 0.00018449123654185515, + "loss": 0.5974, + "step": 7672 + }, + { + "epoch": 0.1798672034787297, + "grad_norm": 3.7310268878936768, + "learning_rate": 0.00018448729611979746, + "loss": 0.4942, + "step": 7673 + }, + { + "epoch": 0.1798906450535347, + "grad_norm": 2.799182653427124, + "learning_rate": 0.00018448335523930937, + "loss": 1.2149, + "step": 7674 + }, + { + "epoch": 0.1799140866283397, + "grad_norm": 2.2546944618225098, + "learning_rate": 0.0001844794139004122, + "loss": 1.5764, + "step": 7675 + }, + { + "epoch": 0.17993752820314468, + "grad_norm": 1.4461839199066162, + "learning_rate": 0.00018447547210312731, + "loss": 0.1473, + "step": 7676 + }, + { + "epoch": 0.17996096977794968, + "grad_norm": 2.8295435905456543, + "learning_rate": 0.00018447152984747619, + "loss": 0.479, + "step": 7677 + }, + { + "epoch": 0.17998441135275467, + "grad_norm": 3.286742687225342, + "learning_rate": 0.00018446758713348013, + "loss": 0.5628, + "step": 7678 + }, + { + "epoch": 0.18000785292755966, + "grad_norm": 4.061708450317383, + "learning_rate": 0.00018446364396116057, + "loss": 0.7042, + "step": 7679 + }, + { + "epoch": 0.18003129450236466, + "grad_norm": 5.969931602478027, + "learning_rate": 0.00018445970033053894, + "loss": 0.4451, + "step": 7680 + }, + { + "epoch": 0.18005473607716965, + "grad_norm": 2.298541307449341, + "learning_rate": 0.00018445575624163657, + "loss": 0.2269, + "step": 7681 + }, + { + "epoch": 0.18007817765197465, + "grad_norm": 2.1070921421051025, + "learning_rate": 0.00018445181169447493, + "loss": 0.2922, + "step": 7682 + }, + { + "epoch": 0.18010161922677964, + "grad_norm": 4.752819061279297, + "learning_rate": 0.00018444786668907534, + "loss": 0.328, + "step": 7683 + }, + { + "epoch": 0.18012506080158466, + "grad_norm": 1.0503785610198975, + "learning_rate": 0.00018444392122545926, + "loss": 0.168, + "step": 7684 + }, + { + "epoch": 0.18014850237638966, + "grad_norm": 3.2924046516418457, + "learning_rate": 0.00018443997530364812, + "loss": 0.4543, + "step": 7685 + }, + { + "epoch": 0.18017194395119465, + "grad_norm": 3.9587202072143555, + "learning_rate": 0.00018443602892366328, + "loss": 0.5685, + "step": 7686 + }, + { + "epoch": 0.18019538552599965, + "grad_norm": 4.133760929107666, + "learning_rate": 0.00018443208208552615, + "loss": 0.5027, + "step": 7687 + }, + { + "epoch": 0.18021882710080464, + "grad_norm": 4.421917915344238, + "learning_rate": 0.0001844281347892582, + "loss": 0.2942, + "step": 7688 + }, + { + "epoch": 0.18024226867560963, + "grad_norm": 3.648876905441284, + "learning_rate": 0.0001844241870348808, + "loss": 0.5821, + "step": 7689 + }, + { + "epoch": 0.18026571025041463, + "grad_norm": 2.3154208660125732, + "learning_rate": 0.0001844202388224154, + "loss": 0.319, + "step": 7690 + }, + { + "epoch": 0.18028915182521962, + "grad_norm": 5.071732044219971, + "learning_rate": 0.0001844162901518834, + "loss": 0.6541, + "step": 7691 + }, + { + "epoch": 0.18031259340002462, + "grad_norm": 1.312842607498169, + "learning_rate": 0.00018441234102330626, + "loss": 0.1549, + "step": 7692 + }, + { + "epoch": 0.1803360349748296, + "grad_norm": 3.223773717880249, + "learning_rate": 0.00018440839143670537, + "loss": 0.2532, + "step": 7693 + }, + { + "epoch": 0.1803594765496346, + "grad_norm": 2.8890933990478516, + "learning_rate": 0.00018440444139210214, + "loss": 0.3911, + "step": 7694 + }, + { + "epoch": 0.1803829181244396, + "grad_norm": 1.015298843383789, + "learning_rate": 0.00018440049088951806, + "loss": 0.0773, + "step": 7695 + }, + { + "epoch": 0.1804063596992446, + "grad_norm": 1.9541279077529907, + "learning_rate": 0.00018439653992897458, + "loss": 0.3708, + "step": 7696 + }, + { + "epoch": 0.1804298012740496, + "grad_norm": 1.2652636766433716, + "learning_rate": 0.00018439258851049307, + "loss": 0.1316, + "step": 7697 + }, + { + "epoch": 0.18045324284885458, + "grad_norm": 2.079652786254883, + "learning_rate": 0.000184388636634095, + "loss": 0.2451, + "step": 7698 + }, + { + "epoch": 0.18047668442365958, + "grad_norm": 2.595780372619629, + "learning_rate": 0.0001843846842998018, + "loss": 0.3005, + "step": 7699 + }, + { + "epoch": 0.18050012599846457, + "grad_norm": 2.9711315631866455, + "learning_rate": 0.00018438073150763495, + "loss": 0.3336, + "step": 7700 + }, + { + "epoch": 0.18052356757326957, + "grad_norm": 4.227585315704346, + "learning_rate": 0.0001843767782576159, + "loss": 1.6528, + "step": 7701 + }, + { + "epoch": 0.18054700914807456, + "grad_norm": 3.6506619453430176, + "learning_rate": 0.00018437282454976603, + "loss": 0.4423, + "step": 7702 + }, + { + "epoch": 0.18057045072287956, + "grad_norm": 3.758969783782959, + "learning_rate": 0.00018436887038410688, + "loss": 0.4478, + "step": 7703 + }, + { + "epoch": 0.18059389229768455, + "grad_norm": 3.2298362255096436, + "learning_rate": 0.00018436491576065986, + "loss": 0.3329, + "step": 7704 + }, + { + "epoch": 0.18061733387248954, + "grad_norm": 4.139408111572266, + "learning_rate": 0.00018436096067944642, + "loss": 0.3838, + "step": 7705 + }, + { + "epoch": 0.18064077544729454, + "grad_norm": 5.615427017211914, + "learning_rate": 0.00018435700514048803, + "loss": 0.4357, + "step": 7706 + }, + { + "epoch": 0.18066421702209953, + "grad_norm": 3.3710811138153076, + "learning_rate": 0.0001843530491438062, + "loss": 0.3332, + "step": 7707 + }, + { + "epoch": 0.18068765859690453, + "grad_norm": 4.173172950744629, + "learning_rate": 0.00018434909268942232, + "loss": 0.505, + "step": 7708 + }, + { + "epoch": 0.18071110017170955, + "grad_norm": 2.281137704849243, + "learning_rate": 0.00018434513577735787, + "loss": 1.4348, + "step": 7709 + }, + { + "epoch": 0.18073454174651454, + "grad_norm": 4.639439582824707, + "learning_rate": 0.0001843411784076344, + "loss": 0.6258, + "step": 7710 + }, + { + "epoch": 0.18075798332131954, + "grad_norm": 2.577627182006836, + "learning_rate": 0.00018433722058027328, + "loss": 0.2827, + "step": 7711 + }, + { + "epoch": 0.18078142489612453, + "grad_norm": 5.871138095855713, + "learning_rate": 0.00018433326229529606, + "loss": 0.6516, + "step": 7712 + }, + { + "epoch": 0.18080486647092953, + "grad_norm": 3.6624257564544678, + "learning_rate": 0.00018432930355272417, + "loss": 0.5499, + "step": 7713 + }, + { + "epoch": 0.18082830804573452, + "grad_norm": 0.9711064100265503, + "learning_rate": 0.00018432534435257908, + "loss": 0.1229, + "step": 7714 + }, + { + "epoch": 0.18085174962053951, + "grad_norm": 3.0973711013793945, + "learning_rate": 0.00018432138469488233, + "loss": 1.9064, + "step": 7715 + }, + { + "epoch": 0.1808751911953445, + "grad_norm": 2.7116572856903076, + "learning_rate": 0.00018431742457965537, + "loss": 0.3216, + "step": 7716 + }, + { + "epoch": 0.1808986327701495, + "grad_norm": 2.662811040878296, + "learning_rate": 0.00018431346400691972, + "loss": 1.8443, + "step": 7717 + }, + { + "epoch": 0.1809220743449545, + "grad_norm": 2.542177677154541, + "learning_rate": 0.0001843095029766968, + "loss": 0.2239, + "step": 7718 + }, + { + "epoch": 0.1809455159197595, + "grad_norm": 3.9957501888275146, + "learning_rate": 0.00018430554148900818, + "loss": 0.2475, + "step": 7719 + }, + { + "epoch": 0.1809689574945645, + "grad_norm": 1.3491483926773071, + "learning_rate": 0.00018430157954387532, + "loss": 0.1586, + "step": 7720 + }, + { + "epoch": 0.18099239906936948, + "grad_norm": 2.44317364692688, + "learning_rate": 0.0001842976171413197, + "loss": 1.4617, + "step": 7721 + }, + { + "epoch": 0.18101584064417447, + "grad_norm": 3.864539623260498, + "learning_rate": 0.00018429365428136285, + "loss": 0.5984, + "step": 7722 + }, + { + "epoch": 0.18103928221897947, + "grad_norm": 4.993529319763184, + "learning_rate": 0.00018428969096402625, + "loss": 0.4937, + "step": 7723 + }, + { + "epoch": 0.18106272379378446, + "grad_norm": 1.006774663925171, + "learning_rate": 0.00018428572718933144, + "loss": 0.1272, + "step": 7724 + }, + { + "epoch": 0.18108616536858946, + "grad_norm": 4.487819671630859, + "learning_rate": 0.00018428176295729988, + "loss": 0.6263, + "step": 7725 + }, + { + "epoch": 0.18110960694339445, + "grad_norm": 5.814844608306885, + "learning_rate": 0.0001842777982679531, + "loss": 0.5034, + "step": 7726 + }, + { + "epoch": 0.18113304851819945, + "grad_norm": 5.261858940124512, + "learning_rate": 0.00018427383312131265, + "loss": 0.7781, + "step": 7727 + }, + { + "epoch": 0.18115649009300444, + "grad_norm": 3.804511785507202, + "learning_rate": 0.00018426986751740002, + "loss": 0.443, + "step": 7728 + }, + { + "epoch": 0.18117993166780944, + "grad_norm": 5.863221168518066, + "learning_rate": 0.00018426590145623667, + "loss": 0.6691, + "step": 7729 + }, + { + "epoch": 0.18120337324261443, + "grad_norm": 3.2371771335601807, + "learning_rate": 0.00018426193493784417, + "loss": 0.2967, + "step": 7730 + }, + { + "epoch": 0.18122681481741942, + "grad_norm": 3.6033599376678467, + "learning_rate": 0.00018425796796224408, + "loss": 0.4404, + "step": 7731 + }, + { + "epoch": 0.18125025639222442, + "grad_norm": 3.9316816329956055, + "learning_rate": 0.00018425400052945787, + "loss": 0.3439, + "step": 7732 + }, + { + "epoch": 0.1812736979670294, + "grad_norm": 2.5522825717926025, + "learning_rate": 0.0001842500326395071, + "loss": 0.309, + "step": 7733 + }, + { + "epoch": 0.1812971395418344, + "grad_norm": 3.5895748138427734, + "learning_rate": 0.00018424606429241327, + "loss": 0.4594, + "step": 7734 + }, + { + "epoch": 0.18132058111663943, + "grad_norm": 1.0601873397827148, + "learning_rate": 0.0001842420954881979, + "loss": 0.0945, + "step": 7735 + }, + { + "epoch": 0.18134402269144442, + "grad_norm": 5.942542552947998, + "learning_rate": 0.00018423812622688258, + "loss": 0.8154, + "step": 7736 + }, + { + "epoch": 0.18136746426624942, + "grad_norm": 1.1108468770980835, + "learning_rate": 0.00018423415650848887, + "loss": 0.0961, + "step": 7737 + }, + { + "epoch": 0.1813909058410544, + "grad_norm": 0.7837164402008057, + "learning_rate": 0.00018423018633303818, + "loss": 0.1103, + "step": 7738 + }, + { + "epoch": 0.1814143474158594, + "grad_norm": 3.1006317138671875, + "learning_rate": 0.00018422621570055215, + "loss": 0.5055, + "step": 7739 + }, + { + "epoch": 0.1814377889906644, + "grad_norm": 1.5453295707702637, + "learning_rate": 0.0001842222446110523, + "loss": 0.7533, + "step": 7740 + }, + { + "epoch": 0.1814612305654694, + "grad_norm": 1.2058064937591553, + "learning_rate": 0.0001842182730645602, + "loss": 0.1464, + "step": 7741 + }, + { + "epoch": 0.1814846721402744, + "grad_norm": 2.2911529541015625, + "learning_rate": 0.00018421430106109737, + "loss": 0.2034, + "step": 7742 + }, + { + "epoch": 0.18150811371507938, + "grad_norm": 3.332000970840454, + "learning_rate": 0.00018421032860068537, + "loss": 0.2731, + "step": 7743 + }, + { + "epoch": 0.18153155528988438, + "grad_norm": 5.045856952667236, + "learning_rate": 0.00018420635568334578, + "loss": 0.6001, + "step": 7744 + }, + { + "epoch": 0.18155499686468937, + "grad_norm": 4.2185258865356445, + "learning_rate": 0.00018420238230910013, + "loss": 0.4522, + "step": 7745 + }, + { + "epoch": 0.18157843843949437, + "grad_norm": 2.6526026725769043, + "learning_rate": 0.00018419840847796998, + "loss": 0.1948, + "step": 7746 + }, + { + "epoch": 0.18160188001429936, + "grad_norm": 4.136519432067871, + "learning_rate": 0.00018419443418997688, + "loss": 0.5813, + "step": 7747 + }, + { + "epoch": 0.18162532158910435, + "grad_norm": 4.695374965667725, + "learning_rate": 0.00018419045944514245, + "loss": 0.5052, + "step": 7748 + }, + { + "epoch": 0.18164876316390935, + "grad_norm": 1.3308054208755493, + "learning_rate": 0.00018418648424348822, + "loss": 0.1663, + "step": 7749 + }, + { + "epoch": 0.18167220473871434, + "grad_norm": 1.5165867805480957, + "learning_rate": 0.0001841825085850357, + "loss": 0.1813, + "step": 7750 + }, + { + "epoch": 0.18169564631351934, + "grad_norm": 4.17962646484375, + "learning_rate": 0.00018417853246980657, + "loss": 0.4812, + "step": 7751 + }, + { + "epoch": 0.18171908788832433, + "grad_norm": 1.4802237749099731, + "learning_rate": 0.00018417455589782235, + "loss": 0.1655, + "step": 7752 + }, + { + "epoch": 0.18174252946312933, + "grad_norm": 5.771732807159424, + "learning_rate": 0.00018417057886910463, + "loss": 0.4599, + "step": 7753 + }, + { + "epoch": 0.18176597103793432, + "grad_norm": 4.583521842956543, + "learning_rate": 0.00018416660138367497, + "loss": 0.6662, + "step": 7754 + }, + { + "epoch": 0.18178941261273932, + "grad_norm": 2.0982210636138916, + "learning_rate": 0.00018416262344155498, + "loss": 0.2497, + "step": 7755 + }, + { + "epoch": 0.1818128541875443, + "grad_norm": 3.8138606548309326, + "learning_rate": 0.0001841586450427662, + "loss": 0.5071, + "step": 7756 + }, + { + "epoch": 0.1818362957623493, + "grad_norm": 2.6862576007843018, + "learning_rate": 0.00018415466618733025, + "loss": 0.3919, + "step": 7757 + }, + { + "epoch": 0.1818597373371543, + "grad_norm": 3.231865406036377, + "learning_rate": 0.00018415068687526874, + "loss": 0.4345, + "step": 7758 + }, + { + "epoch": 0.1818831789119593, + "grad_norm": 1.3215006589889526, + "learning_rate": 0.00018414670710660325, + "loss": 0.1843, + "step": 7759 + }, + { + "epoch": 0.18190662048676431, + "grad_norm": 3.319042444229126, + "learning_rate": 0.00018414272688135533, + "loss": 0.3725, + "step": 7760 + }, + { + "epoch": 0.1819300620615693, + "grad_norm": 5.168032646179199, + "learning_rate": 0.0001841387461995466, + "loss": 0.5301, + "step": 7761 + }, + { + "epoch": 0.1819535036363743, + "grad_norm": 0.8526667952537537, + "learning_rate": 0.0001841347650611987, + "loss": 0.0966, + "step": 7762 + }, + { + "epoch": 0.1819769452111793, + "grad_norm": 3.522207021713257, + "learning_rate": 0.00018413078346633316, + "loss": 0.4795, + "step": 7763 + }, + { + "epoch": 0.1820003867859843, + "grad_norm": 1.8172789812088013, + "learning_rate": 0.00018412680141497166, + "loss": 0.7925, + "step": 7764 + }, + { + "epoch": 0.18202382836078929, + "grad_norm": 2.5432748794555664, + "learning_rate": 0.00018412281890713573, + "loss": 0.4279, + "step": 7765 + }, + { + "epoch": 0.18204726993559428, + "grad_norm": 5.081185340881348, + "learning_rate": 0.00018411883594284705, + "loss": 0.7431, + "step": 7766 + }, + { + "epoch": 0.18207071151039927, + "grad_norm": 4.47662353515625, + "learning_rate": 0.0001841148525221272, + "loss": 0.6928, + "step": 7767 + }, + { + "epoch": 0.18209415308520427, + "grad_norm": 3.0275685787200928, + "learning_rate": 0.0001841108686449978, + "loss": 0.1795, + "step": 7768 + }, + { + "epoch": 0.18211759466000926, + "grad_norm": 0.9418619275093079, + "learning_rate": 0.00018410688431148042, + "loss": 0.1277, + "step": 7769 + }, + { + "epoch": 0.18214103623481426, + "grad_norm": 3.6840896606445312, + "learning_rate": 0.00018410289952159676, + "loss": 0.341, + "step": 7770 + }, + { + "epoch": 0.18216447780961925, + "grad_norm": 3.0398268699645996, + "learning_rate": 0.0001840989142753684, + "loss": 1.7672, + "step": 7771 + }, + { + "epoch": 0.18218791938442425, + "grad_norm": 1.6627984046936035, + "learning_rate": 0.00018409492857281696, + "loss": 0.1418, + "step": 7772 + }, + { + "epoch": 0.18221136095922924, + "grad_norm": 3.9322097301483154, + "learning_rate": 0.00018409094241396407, + "loss": 0.572, + "step": 7773 + }, + { + "epoch": 0.18223480253403423, + "grad_norm": 3.4216575622558594, + "learning_rate": 0.00018408695579883138, + "loss": 0.3424, + "step": 7774 + }, + { + "epoch": 0.18225824410883923, + "grad_norm": 2.820587396621704, + "learning_rate": 0.00018408296872744049, + "loss": 0.2875, + "step": 7775 + }, + { + "epoch": 0.18228168568364422, + "grad_norm": 4.5933003425598145, + "learning_rate": 0.00018407898119981303, + "loss": 0.8493, + "step": 7776 + }, + { + "epoch": 0.18230512725844922, + "grad_norm": 3.1085314750671387, + "learning_rate": 0.00018407499321597068, + "loss": 0.129, + "step": 7777 + }, + { + "epoch": 0.1823285688332542, + "grad_norm": 3.8639423847198486, + "learning_rate": 0.00018407100477593505, + "loss": 1.8979, + "step": 7778 + }, + { + "epoch": 0.1823520104080592, + "grad_norm": 2.215442180633545, + "learning_rate": 0.00018406701587972776, + "loss": 1.5768, + "step": 7779 + }, + { + "epoch": 0.1823754519828642, + "grad_norm": 0.7799462676048279, + "learning_rate": 0.0001840630265273705, + "loss": 0.0969, + "step": 7780 + }, + { + "epoch": 0.1823988935576692, + "grad_norm": 5.046205520629883, + "learning_rate": 0.00018405903671888488, + "loss": 0.7074, + "step": 7781 + }, + { + "epoch": 0.1824223351324742, + "grad_norm": 3.4244794845581055, + "learning_rate": 0.0001840550464542926, + "loss": 1.5488, + "step": 7782 + }, + { + "epoch": 0.18244577670727918, + "grad_norm": 2.1171913146972656, + "learning_rate": 0.00018405105573361525, + "loss": 0.26, + "step": 7783 + }, + { + "epoch": 0.18246921828208418, + "grad_norm": 4.851024627685547, + "learning_rate": 0.0001840470645568745, + "loss": 0.4335, + "step": 7784 + }, + { + "epoch": 0.18249265985688917, + "grad_norm": 1.1848095655441284, + "learning_rate": 0.00018404307292409203, + "loss": 0.1026, + "step": 7785 + }, + { + "epoch": 0.1825161014316942, + "grad_norm": 2.609877109527588, + "learning_rate": 0.00018403908083528946, + "loss": 1.3445, + "step": 7786 + }, + { + "epoch": 0.1825395430064992, + "grad_norm": 3.8700613975524902, + "learning_rate": 0.00018403508829048848, + "loss": 0.4739, + "step": 7787 + }, + { + "epoch": 0.18256298458130418, + "grad_norm": 5.7742838859558105, + "learning_rate": 0.00018403109528971077, + "loss": 0.7133, + "step": 7788 + }, + { + "epoch": 0.18258642615610918, + "grad_norm": 4.072441101074219, + "learning_rate": 0.00018402710183297797, + "loss": 0.6112, + "step": 7789 + }, + { + "epoch": 0.18260986773091417, + "grad_norm": 0.501250147819519, + "learning_rate": 0.00018402310792031172, + "loss": 0.0825, + "step": 7790 + }, + { + "epoch": 0.18263330930571917, + "grad_norm": 4.007122993469238, + "learning_rate": 0.00018401911355173375, + "loss": 0.4598, + "step": 7791 + }, + { + "epoch": 0.18265675088052416, + "grad_norm": 5.74495267868042, + "learning_rate": 0.00018401511872726572, + "loss": 0.7578, + "step": 7792 + }, + { + "epoch": 0.18268019245532915, + "grad_norm": 5.036840915679932, + "learning_rate": 0.00018401112344692928, + "loss": 0.3934, + "step": 7793 + }, + { + "epoch": 0.18270363403013415, + "grad_norm": 4.1147847175598145, + "learning_rate": 0.00018400712771074611, + "loss": 0.3164, + "step": 7794 + }, + { + "epoch": 0.18272707560493914, + "grad_norm": 2.230642795562744, + "learning_rate": 0.00018400313151873792, + "loss": 0.2847, + "step": 7795 + }, + { + "epoch": 0.18275051717974414, + "grad_norm": 2.982261896133423, + "learning_rate": 0.00018399913487092637, + "loss": 0.3234, + "step": 7796 + }, + { + "epoch": 0.18277395875454913, + "grad_norm": 3.879378318786621, + "learning_rate": 0.00018399513776733313, + "loss": 0.6938, + "step": 7797 + }, + { + "epoch": 0.18279740032935413, + "grad_norm": 2.810499429702759, + "learning_rate": 0.00018399114020797993, + "loss": 0.4939, + "step": 7798 + }, + { + "epoch": 0.18282084190415912, + "grad_norm": 5.2584733963012695, + "learning_rate": 0.00018398714219288844, + "loss": 0.4681, + "step": 7799 + }, + { + "epoch": 0.18284428347896411, + "grad_norm": 1.6187763214111328, + "learning_rate": 0.00018398314372208038, + "loss": 0.2371, + "step": 7800 + }, + { + "epoch": 0.1828677250537691, + "grad_norm": 3.5531115531921387, + "learning_rate": 0.00018397914479557738, + "loss": 0.3668, + "step": 7801 + }, + { + "epoch": 0.1828911666285741, + "grad_norm": 1.8615859746932983, + "learning_rate": 0.00018397514541340122, + "loss": 0.1601, + "step": 7802 + }, + { + "epoch": 0.1829146082033791, + "grad_norm": 1.5881773233413696, + "learning_rate": 0.00018397114557557355, + "loss": 0.1964, + "step": 7803 + }, + { + "epoch": 0.1829380497781841, + "grad_norm": 3.8407540321350098, + "learning_rate": 0.00018396714528211606, + "loss": 0.3958, + "step": 7804 + }, + { + "epoch": 0.1829614913529891, + "grad_norm": 3.018475294113159, + "learning_rate": 0.00018396314453305047, + "loss": 1.854, + "step": 7805 + }, + { + "epoch": 0.18298493292779408, + "grad_norm": 2.195589065551758, + "learning_rate": 0.00018395914332839855, + "loss": 0.9062, + "step": 7806 + }, + { + "epoch": 0.18300837450259907, + "grad_norm": 3.943194627761841, + "learning_rate": 0.0001839551416681819, + "loss": 0.5574, + "step": 7807 + }, + { + "epoch": 0.18303181607740407, + "grad_norm": 0.9496101140975952, + "learning_rate": 0.0001839511395524223, + "loss": 0.1471, + "step": 7808 + }, + { + "epoch": 0.18305525765220906, + "grad_norm": 1.6916009187698364, + "learning_rate": 0.0001839471369811415, + "loss": 0.4683, + "step": 7809 + }, + { + "epoch": 0.18307869922701406, + "grad_norm": 4.05734395980835, + "learning_rate": 0.00018394313395436114, + "loss": 0.5771, + "step": 7810 + }, + { + "epoch": 0.18310214080181908, + "grad_norm": 5.372860908508301, + "learning_rate": 0.00018393913047210297, + "loss": 0.7957, + "step": 7811 + }, + { + "epoch": 0.18312558237662407, + "grad_norm": 4.423242568969727, + "learning_rate": 0.00018393512653438877, + "loss": 0.5241, + "step": 7812 + }, + { + "epoch": 0.18314902395142907, + "grad_norm": 4.717846870422363, + "learning_rate": 0.00018393112214124015, + "loss": 0.7954, + "step": 7813 + }, + { + "epoch": 0.18317246552623406, + "grad_norm": 3.999915838241577, + "learning_rate": 0.0001839271172926789, + "loss": 0.4873, + "step": 7814 + }, + { + "epoch": 0.18319590710103906, + "grad_norm": 0.803394079208374, + "learning_rate": 0.0001839231119887268, + "loss": 0.1314, + "step": 7815 + }, + { + "epoch": 0.18321934867584405, + "grad_norm": 0.792658805847168, + "learning_rate": 0.00018391910622940548, + "loss": 0.0491, + "step": 7816 + }, + { + "epoch": 0.18324279025064905, + "grad_norm": 1.8992003202438354, + "learning_rate": 0.00018391510001473677, + "loss": 0.3569, + "step": 7817 + }, + { + "epoch": 0.18326623182545404, + "grad_norm": 3.530883550643921, + "learning_rate": 0.00018391109334474237, + "loss": 0.3627, + "step": 7818 + }, + { + "epoch": 0.18328967340025903, + "grad_norm": 1.677117109298706, + "learning_rate": 0.00018390708621944402, + "loss": 0.1044, + "step": 7819 + }, + { + "epoch": 0.18331311497506403, + "grad_norm": 3.458371639251709, + "learning_rate": 0.00018390307863886345, + "loss": 0.5352, + "step": 7820 + }, + { + "epoch": 0.18333655654986902, + "grad_norm": 4.712778091430664, + "learning_rate": 0.0001838990706030224, + "loss": 0.7521, + "step": 7821 + }, + { + "epoch": 0.18335999812467402, + "grad_norm": 3.9319262504577637, + "learning_rate": 0.00018389506211194265, + "loss": 0.4452, + "step": 7822 + }, + { + "epoch": 0.183383439699479, + "grad_norm": 5.013703346252441, + "learning_rate": 0.00018389105316564592, + "loss": 0.4104, + "step": 7823 + }, + { + "epoch": 0.183406881274284, + "grad_norm": 4.198429584503174, + "learning_rate": 0.00018388704376415399, + "loss": 0.6017, + "step": 7824 + }, + { + "epoch": 0.183430322849089, + "grad_norm": 3.615180253982544, + "learning_rate": 0.00018388303390748863, + "loss": 0.4443, + "step": 7825 + }, + { + "epoch": 0.183453764423894, + "grad_norm": 4.972517490386963, + "learning_rate": 0.0001838790235956715, + "loss": 0.6375, + "step": 7826 + }, + { + "epoch": 0.183477205998699, + "grad_norm": 3.0455973148345947, + "learning_rate": 0.00018387501282872448, + "loss": 0.4092, + "step": 7827 + }, + { + "epoch": 0.18350064757350398, + "grad_norm": 3.376786708831787, + "learning_rate": 0.00018387100160666928, + "loss": 0.4264, + "step": 7828 + }, + { + "epoch": 0.18352408914830898, + "grad_norm": 1.7105792760849, + "learning_rate": 0.00018386698992952765, + "loss": 0.2539, + "step": 7829 + }, + { + "epoch": 0.18354753072311397, + "grad_norm": 1.347306251525879, + "learning_rate": 0.0001838629777973214, + "loss": 0.173, + "step": 7830 + }, + { + "epoch": 0.18357097229791897, + "grad_norm": 4.702389240264893, + "learning_rate": 0.00018385896521007226, + "loss": 0.7029, + "step": 7831 + }, + { + "epoch": 0.18359441387272396, + "grad_norm": 5.478905200958252, + "learning_rate": 0.000183854952167802, + "loss": 1.0436, + "step": 7832 + }, + { + "epoch": 0.18361785544752895, + "grad_norm": 4.262477397918701, + "learning_rate": 0.0001838509386705324, + "loss": 0.5244, + "step": 7833 + }, + { + "epoch": 0.18364129702233395, + "grad_norm": 1.4861435890197754, + "learning_rate": 0.0001838469247182853, + "loss": 0.2055, + "step": 7834 + }, + { + "epoch": 0.18366473859713894, + "grad_norm": 4.733738899230957, + "learning_rate": 0.0001838429103110824, + "loss": 0.5542, + "step": 7835 + }, + { + "epoch": 0.18368818017194394, + "grad_norm": 2.5434141159057617, + "learning_rate": 0.0001838388954489455, + "loss": 0.3696, + "step": 7836 + }, + { + "epoch": 0.18371162174674896, + "grad_norm": 2.4528958797454834, + "learning_rate": 0.00018383488013189642, + "loss": 1.1999, + "step": 7837 + }, + { + "epoch": 0.18373506332155395, + "grad_norm": 4.10908842086792, + "learning_rate": 0.00018383086435995693, + "loss": 0.5257, + "step": 7838 + }, + { + "epoch": 0.18375850489635895, + "grad_norm": 0.7889413833618164, + "learning_rate": 0.00018382684813314878, + "loss": 0.1136, + "step": 7839 + }, + { + "epoch": 0.18378194647116394, + "grad_norm": 2.8588359355926514, + "learning_rate": 0.00018382283145149383, + "loss": 2.0856, + "step": 7840 + }, + { + "epoch": 0.18380538804596894, + "grad_norm": 3.4870197772979736, + "learning_rate": 0.0001838188143150138, + "loss": 0.3498, + "step": 7841 + }, + { + "epoch": 0.18382882962077393, + "grad_norm": 4.179501533508301, + "learning_rate": 0.00018381479672373055, + "loss": 0.6381, + "step": 7842 + }, + { + "epoch": 0.18385227119557893, + "grad_norm": 4.4070143699646, + "learning_rate": 0.00018381077867766585, + "loss": 0.5328, + "step": 7843 + }, + { + "epoch": 0.18387571277038392, + "grad_norm": 2.409409999847412, + "learning_rate": 0.00018380676017684154, + "loss": 1.1386, + "step": 7844 + }, + { + "epoch": 0.18389915434518891, + "grad_norm": 4.409891605377197, + "learning_rate": 0.00018380274122127937, + "loss": 0.5796, + "step": 7845 + }, + { + "epoch": 0.1839225959199939, + "grad_norm": 0.5921149849891663, + "learning_rate": 0.00018379872181100115, + "loss": 0.0597, + "step": 7846 + }, + { + "epoch": 0.1839460374947989, + "grad_norm": 4.682120323181152, + "learning_rate": 0.00018379470194602874, + "loss": 0.7001, + "step": 7847 + }, + { + "epoch": 0.1839694790696039, + "grad_norm": 4.279823303222656, + "learning_rate": 0.0001837906816263839, + "loss": 0.4486, + "step": 7848 + }, + { + "epoch": 0.1839929206444089, + "grad_norm": 3.2597532272338867, + "learning_rate": 0.00018378666085208843, + "loss": 0.4817, + "step": 7849 + }, + { + "epoch": 0.18401636221921389, + "grad_norm": 3.7163946628570557, + "learning_rate": 0.00018378263962316423, + "loss": 0.5645, + "step": 7850 + }, + { + "epoch": 0.18403980379401888, + "grad_norm": 1.3238991498947144, + "learning_rate": 0.00018377861793963309, + "loss": 0.1051, + "step": 7851 + }, + { + "epoch": 0.18406324536882387, + "grad_norm": 3.1107773780822754, + "learning_rate": 0.00018377459580151679, + "loss": 0.403, + "step": 7852 + }, + { + "epoch": 0.18408668694362887, + "grad_norm": 0.8748726844787598, + "learning_rate": 0.0001837705732088372, + "loss": 0.0877, + "step": 7853 + }, + { + "epoch": 0.18411012851843386, + "grad_norm": 2.5429911613464355, + "learning_rate": 0.00018376655016161605, + "loss": 0.326, + "step": 7854 + }, + { + "epoch": 0.18413357009323886, + "grad_norm": 0.5753430128097534, + "learning_rate": 0.00018376252665987532, + "loss": 0.0789, + "step": 7855 + }, + { + "epoch": 0.18415701166804385, + "grad_norm": 3.6964192390441895, + "learning_rate": 0.00018375850270363673, + "loss": 0.4356, + "step": 7856 + }, + { + "epoch": 0.18418045324284885, + "grad_norm": 3.0187389850616455, + "learning_rate": 0.00018375447829292217, + "loss": 0.3753, + "step": 7857 + }, + { + "epoch": 0.18420389481765384, + "grad_norm": 3.9107022285461426, + "learning_rate": 0.00018375045342775343, + "loss": 0.3394, + "step": 7858 + }, + { + "epoch": 0.18422733639245883, + "grad_norm": 3.102466106414795, + "learning_rate": 0.0001837464281081524, + "loss": 0.3774, + "step": 7859 + }, + { + "epoch": 0.18425077796726383, + "grad_norm": 1.3703128099441528, + "learning_rate": 0.0001837424023341409, + "loss": 0.1418, + "step": 7860 + }, + { + "epoch": 0.18427421954206882, + "grad_norm": 4.965177059173584, + "learning_rate": 0.00018373837610574073, + "loss": 0.4364, + "step": 7861 + }, + { + "epoch": 0.18429766111687385, + "grad_norm": 3.455256223678589, + "learning_rate": 0.00018373434942297383, + "loss": 0.4911, + "step": 7862 + }, + { + "epoch": 0.18432110269167884, + "grad_norm": 1.050402283668518, + "learning_rate": 0.00018373032228586196, + "loss": 0.115, + "step": 7863 + }, + { + "epoch": 0.18434454426648383, + "grad_norm": 4.554635047912598, + "learning_rate": 0.000183726294694427, + "loss": 0.379, + "step": 7864 + }, + { + "epoch": 0.18436798584128883, + "grad_norm": 3.6749536991119385, + "learning_rate": 0.00018372226664869083, + "loss": 0.399, + "step": 7865 + }, + { + "epoch": 0.18439142741609382, + "grad_norm": 5.610636234283447, + "learning_rate": 0.0001837182381486753, + "loss": 0.5505, + "step": 7866 + }, + { + "epoch": 0.18441486899089882, + "grad_norm": 3.741015911102295, + "learning_rate": 0.00018371420919440222, + "loss": 1.5616, + "step": 7867 + }, + { + "epoch": 0.1844383105657038, + "grad_norm": 3.7777910232543945, + "learning_rate": 0.00018371017978589352, + "loss": 0.4432, + "step": 7868 + }, + { + "epoch": 0.1844617521405088, + "grad_norm": 3.597914695739746, + "learning_rate": 0.00018370614992317102, + "loss": 1.6019, + "step": 7869 + }, + { + "epoch": 0.1844851937153138, + "grad_norm": 4.566976547241211, + "learning_rate": 0.0001837021196062566, + "loss": 0.5459, + "step": 7870 + }, + { + "epoch": 0.1845086352901188, + "grad_norm": 3.4919304847717285, + "learning_rate": 0.00018369808883517212, + "loss": 0.372, + "step": 7871 + }, + { + "epoch": 0.1845320768649238, + "grad_norm": 4.424670219421387, + "learning_rate": 0.00018369405760993942, + "loss": 0.6697, + "step": 7872 + }, + { + "epoch": 0.18455551843972878, + "grad_norm": 3.402625322341919, + "learning_rate": 0.00018369002593058046, + "loss": 0.311, + "step": 7873 + }, + { + "epoch": 0.18457896001453378, + "grad_norm": 4.008194446563721, + "learning_rate": 0.00018368599379711706, + "loss": 0.464, + "step": 7874 + }, + { + "epoch": 0.18460240158933877, + "grad_norm": 4.723906993865967, + "learning_rate": 0.00018368196120957108, + "loss": 0.6478, + "step": 7875 + }, + { + "epoch": 0.18462584316414377, + "grad_norm": 0.9050693511962891, + "learning_rate": 0.00018367792816796445, + "loss": 0.102, + "step": 7876 + }, + { + "epoch": 0.18464928473894876, + "grad_norm": 4.9756999015808105, + "learning_rate": 0.00018367389467231903, + "loss": 0.7578, + "step": 7877 + }, + { + "epoch": 0.18467272631375375, + "grad_norm": 4.819327354431152, + "learning_rate": 0.00018366986072265671, + "loss": 0.4947, + "step": 7878 + }, + { + "epoch": 0.18469616788855875, + "grad_norm": 1.5062010288238525, + "learning_rate": 0.00018366582631899935, + "loss": 0.2263, + "step": 7879 + }, + { + "epoch": 0.18471960946336374, + "grad_norm": 2.642751693725586, + "learning_rate": 0.00018366179146136888, + "loss": 0.3281, + "step": 7880 + }, + { + "epoch": 0.18474305103816874, + "grad_norm": 5.260924816131592, + "learning_rate": 0.00018365775614978715, + "loss": 0.3307, + "step": 7881 + }, + { + "epoch": 0.18476649261297373, + "grad_norm": 0.6471335887908936, + "learning_rate": 0.00018365372038427614, + "loss": 0.1274, + "step": 7882 + }, + { + "epoch": 0.18478993418777873, + "grad_norm": 3.742987871170044, + "learning_rate": 0.0001836496841648576, + "loss": 0.6347, + "step": 7883 + }, + { + "epoch": 0.18481337576258372, + "grad_norm": 6.8267903327941895, + "learning_rate": 0.00018364564749155359, + "loss": 0.4104, + "step": 7884 + }, + { + "epoch": 0.18483681733738871, + "grad_norm": 1.4376213550567627, + "learning_rate": 0.00018364161036438594, + "loss": 0.1886, + "step": 7885 + }, + { + "epoch": 0.1848602589121937, + "grad_norm": 1.3108258247375488, + "learning_rate": 0.00018363757278337656, + "loss": 0.1146, + "step": 7886 + }, + { + "epoch": 0.1848837004869987, + "grad_norm": 3.3931264877319336, + "learning_rate": 0.00018363353474854733, + "loss": 0.4873, + "step": 7887 + }, + { + "epoch": 0.18490714206180373, + "grad_norm": 1.5774649381637573, + "learning_rate": 0.00018362949625992024, + "loss": 0.1676, + "step": 7888 + }, + { + "epoch": 0.18493058363660872, + "grad_norm": 2.245400905609131, + "learning_rate": 0.0001836254573175171, + "loss": 1.5076, + "step": 7889 + }, + { + "epoch": 0.18495402521141371, + "grad_norm": 5.043694972991943, + "learning_rate": 0.0001836214179213599, + "loss": 0.4649, + "step": 7890 + }, + { + "epoch": 0.1849774667862187, + "grad_norm": 6.316823959350586, + "learning_rate": 0.00018361737807147055, + "loss": 0.6792, + "step": 7891 + }, + { + "epoch": 0.1850009083610237, + "grad_norm": 0.5556317567825317, + "learning_rate": 0.00018361333776787092, + "loss": 0.0332, + "step": 7892 + }, + { + "epoch": 0.1850243499358287, + "grad_norm": 3.0785577297210693, + "learning_rate": 0.00018360929701058298, + "loss": 0.3426, + "step": 7893 + }, + { + "epoch": 0.1850477915106337, + "grad_norm": 1.738364577293396, + "learning_rate": 0.00018360525579962867, + "loss": 0.1966, + "step": 7894 + }, + { + "epoch": 0.18507123308543869, + "grad_norm": 3.893324613571167, + "learning_rate": 0.00018360121413502984, + "loss": 1.864, + "step": 7895 + }, + { + "epoch": 0.18509467466024368, + "grad_norm": 3.6473426818847656, + "learning_rate": 0.0001835971720168085, + "loss": 0.3047, + "step": 7896 + }, + { + "epoch": 0.18511811623504867, + "grad_norm": 5.91462516784668, + "learning_rate": 0.0001835931294449866, + "loss": 0.324, + "step": 7897 + }, + { + "epoch": 0.18514155780985367, + "grad_norm": 2.5898635387420654, + "learning_rate": 0.00018358908641958596, + "loss": 1.4529, + "step": 7898 + }, + { + "epoch": 0.18516499938465866, + "grad_norm": 4.949781894683838, + "learning_rate": 0.00018358504294062862, + "loss": 0.6462, + "step": 7899 + }, + { + "epoch": 0.18518844095946366, + "grad_norm": 2.188462018966675, + "learning_rate": 0.00018358099900813647, + "loss": 1.6303, + "step": 7900 + }, + { + "epoch": 0.18521188253426865, + "grad_norm": 1.918216586112976, + "learning_rate": 0.0001835769546221315, + "loss": 1.9221, + "step": 7901 + }, + { + "epoch": 0.18523532410907365, + "grad_norm": 3.241572618484497, + "learning_rate": 0.00018357290978263558, + "loss": 0.2873, + "step": 7902 + }, + { + "epoch": 0.18525876568387864, + "grad_norm": 4.185405254364014, + "learning_rate": 0.00018356886448967074, + "loss": 0.5277, + "step": 7903 + }, + { + "epoch": 0.18528220725868363, + "grad_norm": 4.764857769012451, + "learning_rate": 0.00018356481874325887, + "loss": 0.4863, + "step": 7904 + }, + { + "epoch": 0.18530564883348863, + "grad_norm": 1.1356441974639893, + "learning_rate": 0.00018356077254342193, + "loss": 0.0993, + "step": 7905 + }, + { + "epoch": 0.18532909040829362, + "grad_norm": 5.425755023956299, + "learning_rate": 0.00018355672589018192, + "loss": 0.6527, + "step": 7906 + }, + { + "epoch": 0.18535253198309862, + "grad_norm": 5.148467540740967, + "learning_rate": 0.00018355267878356077, + "loss": 0.8307, + "step": 7907 + }, + { + "epoch": 0.1853759735579036, + "grad_norm": 1.8372606039047241, + "learning_rate": 0.0001835486312235804, + "loss": 0.2505, + "step": 7908 + }, + { + "epoch": 0.1853994151327086, + "grad_norm": 0.9356855750083923, + "learning_rate": 0.00018354458321026286, + "loss": 0.1012, + "step": 7909 + }, + { + "epoch": 0.1854228567075136, + "grad_norm": 1.9950687885284424, + "learning_rate": 0.00018354053474362999, + "loss": 2.1298, + "step": 7910 + }, + { + "epoch": 0.1854462982823186, + "grad_norm": 0.6520652770996094, + "learning_rate": 0.0001835364858237039, + "loss": 0.0451, + "step": 7911 + }, + { + "epoch": 0.1854697398571236, + "grad_norm": 3.7565388679504395, + "learning_rate": 0.00018353243645050644, + "loss": 0.4628, + "step": 7912 + }, + { + "epoch": 0.18549318143192858, + "grad_norm": 4.039470195770264, + "learning_rate": 0.00018352838662405967, + "loss": 0.3236, + "step": 7913 + }, + { + "epoch": 0.1855166230067336, + "grad_norm": 4.1596760749816895, + "learning_rate": 0.0001835243363443855, + "loss": 0.5294, + "step": 7914 + }, + { + "epoch": 0.1855400645815386, + "grad_norm": 3.592768430709839, + "learning_rate": 0.00018352028561150594, + "loss": 0.5424, + "step": 7915 + }, + { + "epoch": 0.1855635061563436, + "grad_norm": 3.7367067337036133, + "learning_rate": 0.00018351623442544295, + "loss": 0.6495, + "step": 7916 + }, + { + "epoch": 0.1855869477311486, + "grad_norm": 2.5662946701049805, + "learning_rate": 0.00018351218278621855, + "loss": 0.3841, + "step": 7917 + }, + { + "epoch": 0.18561038930595358, + "grad_norm": 4.017256259918213, + "learning_rate": 0.0001835081306938547, + "loss": 0.6315, + "step": 7918 + }, + { + "epoch": 0.18563383088075858, + "grad_norm": 0.853878915309906, + "learning_rate": 0.00018350407814837335, + "loss": 0.0867, + "step": 7919 + }, + { + "epoch": 0.18565727245556357, + "grad_norm": 0.6375055909156799, + "learning_rate": 0.0001835000251497965, + "loss": 0.0574, + "step": 7920 + }, + { + "epoch": 0.18568071403036857, + "grad_norm": 2.973253011703491, + "learning_rate": 0.00018349597169814623, + "loss": 0.369, + "step": 7921 + }, + { + "epoch": 0.18570415560517356, + "grad_norm": 5.128635406494141, + "learning_rate": 0.00018349191779344447, + "loss": 0.715, + "step": 7922 + }, + { + "epoch": 0.18572759717997855, + "grad_norm": 3.3036952018737793, + "learning_rate": 0.00018348786343571317, + "loss": 0.432, + "step": 7923 + }, + { + "epoch": 0.18575103875478355, + "grad_norm": 6.241300106048584, + "learning_rate": 0.00018348380862497437, + "loss": 0.8731, + "step": 7924 + }, + { + "epoch": 0.18577448032958854, + "grad_norm": 0.9252178072929382, + "learning_rate": 0.00018347975336125013, + "loss": 0.0952, + "step": 7925 + }, + { + "epoch": 0.18579792190439354, + "grad_norm": 1.957898736000061, + "learning_rate": 0.00018347569764456237, + "loss": 0.2634, + "step": 7926 + }, + { + "epoch": 0.18582136347919853, + "grad_norm": 1.5167279243469238, + "learning_rate": 0.00018347164147493314, + "loss": 0.2345, + "step": 7927 + }, + { + "epoch": 0.18584480505400353, + "grad_norm": 1.4080673456192017, + "learning_rate": 0.00018346758485238444, + "loss": 0.2057, + "step": 7928 + }, + { + "epoch": 0.18586824662880852, + "grad_norm": 1.3733590841293335, + "learning_rate": 0.00018346352777693828, + "loss": 0.2426, + "step": 7929 + }, + { + "epoch": 0.18589168820361351, + "grad_norm": 3.6950314044952393, + "learning_rate": 0.00018345947024861665, + "loss": 0.4936, + "step": 7930 + }, + { + "epoch": 0.1859151297784185, + "grad_norm": 1.0575355291366577, + "learning_rate": 0.00018345541226744158, + "loss": 0.1225, + "step": 7931 + }, + { + "epoch": 0.1859385713532235, + "grad_norm": 2.722637891769409, + "learning_rate": 0.00018345135383343513, + "loss": 0.3187, + "step": 7932 + }, + { + "epoch": 0.1859620129280285, + "grad_norm": 4.664611339569092, + "learning_rate": 0.0001834472949466193, + "loss": 0.482, + "step": 7933 + }, + { + "epoch": 0.1859854545028335, + "grad_norm": 1.0421658754348755, + "learning_rate": 0.00018344323560701603, + "loss": 0.5476, + "step": 7934 + }, + { + "epoch": 0.18600889607763849, + "grad_norm": 4.37137508392334, + "learning_rate": 0.00018343917581464746, + "loss": 0.6564, + "step": 7935 + }, + { + "epoch": 0.18603233765244348, + "grad_norm": 5.434080600738525, + "learning_rate": 0.0001834351155695356, + "loss": 0.6815, + "step": 7936 + }, + { + "epoch": 0.18605577922724847, + "grad_norm": 1.6702616214752197, + "learning_rate": 0.00018343105487170244, + "loss": 0.1712, + "step": 7937 + }, + { + "epoch": 0.18607922080205347, + "grad_norm": 3.5159451961517334, + "learning_rate": 0.00018342699372117, + "loss": 0.5062, + "step": 7938 + }, + { + "epoch": 0.1861026623768585, + "grad_norm": 3.1560842990875244, + "learning_rate": 0.00018342293211796035, + "loss": 0.4193, + "step": 7939 + }, + { + "epoch": 0.18612610395166349, + "grad_norm": 1.0455974340438843, + "learning_rate": 0.00018341887006209558, + "loss": 0.0852, + "step": 7940 + }, + { + "epoch": 0.18614954552646848, + "grad_norm": 3.524667978286743, + "learning_rate": 0.0001834148075535976, + "loss": 0.4408, + "step": 7941 + }, + { + "epoch": 0.18617298710127347, + "grad_norm": 3.8458800315856934, + "learning_rate": 0.00018341074459248859, + "loss": 0.5377, + "step": 7942 + }, + { + "epoch": 0.18619642867607847, + "grad_norm": 1.7499290704727173, + "learning_rate": 0.0001834066811787905, + "loss": 1.1192, + "step": 7943 + }, + { + "epoch": 0.18621987025088346, + "grad_norm": 2.9339871406555176, + "learning_rate": 0.00018340261731252544, + "loss": 0.4776, + "step": 7944 + }, + { + "epoch": 0.18624331182568846, + "grad_norm": 6.141181468963623, + "learning_rate": 0.00018339855299371542, + "loss": 0.6985, + "step": 7945 + }, + { + "epoch": 0.18626675340049345, + "grad_norm": 5.0622239112854, + "learning_rate": 0.0001833944882223825, + "loss": 0.5722, + "step": 7946 + }, + { + "epoch": 0.18629019497529845, + "grad_norm": 3.878542900085449, + "learning_rate": 0.00018339042299854875, + "loss": 0.3916, + "step": 7947 + }, + { + "epoch": 0.18631363655010344, + "grad_norm": 3.378941059112549, + "learning_rate": 0.0001833863573222362, + "loss": 0.3184, + "step": 7948 + }, + { + "epoch": 0.18633707812490843, + "grad_norm": 2.3453197479248047, + "learning_rate": 0.00018338229119346696, + "loss": 0.3699, + "step": 7949 + }, + { + "epoch": 0.18636051969971343, + "grad_norm": 1.4591448307037354, + "learning_rate": 0.00018337822461226305, + "loss": 0.1056, + "step": 7950 + }, + { + "epoch": 0.18638396127451842, + "grad_norm": 3.2216339111328125, + "learning_rate": 0.00018337415757864657, + "loss": 0.5521, + "step": 7951 + }, + { + "epoch": 0.18640740284932342, + "grad_norm": 2.0826685428619385, + "learning_rate": 0.00018337009009263954, + "loss": 0.2446, + "step": 7952 + }, + { + "epoch": 0.1864308444241284, + "grad_norm": 6.418205261230469, + "learning_rate": 0.0001833660221542641, + "loss": 0.2274, + "step": 7953 + }, + { + "epoch": 0.1864542859989334, + "grad_norm": 2.4913554191589355, + "learning_rate": 0.0001833619537635422, + "loss": 0.3389, + "step": 7954 + }, + { + "epoch": 0.1864777275737384, + "grad_norm": 2.7072596549987793, + "learning_rate": 0.00018335788492049602, + "loss": 0.469, + "step": 7955 + }, + { + "epoch": 0.1865011691485434, + "grad_norm": 3.5820860862731934, + "learning_rate": 0.00018335381562514767, + "loss": 0.3928, + "step": 7956 + }, + { + "epoch": 0.1865246107233484, + "grad_norm": 2.4412126541137695, + "learning_rate": 0.0001833497458775191, + "loss": 0.3941, + "step": 7957 + }, + { + "epoch": 0.18654805229815338, + "grad_norm": 3.219623565673828, + "learning_rate": 0.0001833456756776325, + "loss": 0.5252, + "step": 7958 + }, + { + "epoch": 0.18657149387295838, + "grad_norm": 3.7274580001831055, + "learning_rate": 0.00018334160502550995, + "loss": 0.7656, + "step": 7959 + }, + { + "epoch": 0.18659493544776337, + "grad_norm": 4.493252277374268, + "learning_rate": 0.00018333753392117346, + "loss": 0.7437, + "step": 7960 + }, + { + "epoch": 0.18661837702256837, + "grad_norm": 2.190955400466919, + "learning_rate": 0.0001833334623646452, + "loss": 0.1899, + "step": 7961 + }, + { + "epoch": 0.18664181859737336, + "grad_norm": 4.20254373550415, + "learning_rate": 0.0001833293903559472, + "loss": 0.5058, + "step": 7962 + }, + { + "epoch": 0.18666526017217835, + "grad_norm": 3.9895172119140625, + "learning_rate": 0.00018332531789510162, + "loss": 0.65, + "step": 7963 + }, + { + "epoch": 0.18668870174698335, + "grad_norm": 3.126006603240967, + "learning_rate": 0.0001833212449821305, + "loss": 1.9988, + "step": 7964 + }, + { + "epoch": 0.18671214332178837, + "grad_norm": 3.912039041519165, + "learning_rate": 0.00018331717161705594, + "loss": 0.438, + "step": 7965 + }, + { + "epoch": 0.18673558489659337, + "grad_norm": 3.7807016372680664, + "learning_rate": 0.00018331309779990012, + "loss": 0.4407, + "step": 7966 + }, + { + "epoch": 0.18675902647139836, + "grad_norm": 5.6320085525512695, + "learning_rate": 0.00018330902353068504, + "loss": 0.8341, + "step": 7967 + }, + { + "epoch": 0.18678246804620335, + "grad_norm": 0.8664217591285706, + "learning_rate": 0.0001833049488094329, + "loss": 0.0644, + "step": 7968 + }, + { + "epoch": 0.18680590962100835, + "grad_norm": 0.6070694923400879, + "learning_rate": 0.0001833008736361657, + "loss": 0.0403, + "step": 7969 + }, + { + "epoch": 0.18682935119581334, + "grad_norm": 1.6093604564666748, + "learning_rate": 0.00018329679801090568, + "loss": 0.123, + "step": 7970 + }, + { + "epoch": 0.18685279277061834, + "grad_norm": 2.8985402584075928, + "learning_rate": 0.00018329272193367486, + "loss": 0.4011, + "step": 7971 + }, + { + "epoch": 0.18687623434542333, + "grad_norm": 2.7391774654388428, + "learning_rate": 0.0001832886454044954, + "loss": 0.2759, + "step": 7972 + }, + { + "epoch": 0.18689967592022833, + "grad_norm": 5.784007549285889, + "learning_rate": 0.0001832845684233894, + "loss": 0.4916, + "step": 7973 + }, + { + "epoch": 0.18692311749503332, + "grad_norm": 2.3510801792144775, + "learning_rate": 0.00018328049099037902, + "loss": 2.0609, + "step": 7974 + }, + { + "epoch": 0.18694655906983831, + "grad_norm": 4.741113185882568, + "learning_rate": 0.0001832764131054863, + "loss": 0.7528, + "step": 7975 + }, + { + "epoch": 0.1869700006446433, + "grad_norm": 4.298549652099609, + "learning_rate": 0.00018327233476873346, + "loss": 0.6955, + "step": 7976 + }, + { + "epoch": 0.1869934422194483, + "grad_norm": 2.6612350940704346, + "learning_rate": 0.0001832682559801426, + "loss": 0.3367, + "step": 7977 + }, + { + "epoch": 0.1870168837942533, + "grad_norm": 2.214129686355591, + "learning_rate": 0.0001832641767397358, + "loss": 0.9233, + "step": 7978 + }, + { + "epoch": 0.1870403253690583, + "grad_norm": 2.304678440093994, + "learning_rate": 0.00018326009704753527, + "loss": 0.25, + "step": 7979 + }, + { + "epoch": 0.18706376694386329, + "grad_norm": 2.483935832977295, + "learning_rate": 0.0001832560169035631, + "loss": 1.6908, + "step": 7980 + }, + { + "epoch": 0.18708720851866828, + "grad_norm": 3.707329511642456, + "learning_rate": 0.00018325193630784142, + "loss": 1.4663, + "step": 7981 + }, + { + "epoch": 0.18711065009347327, + "grad_norm": 4.328886032104492, + "learning_rate": 0.00018324785526039244, + "loss": 0.6152, + "step": 7982 + }, + { + "epoch": 0.18713409166827827, + "grad_norm": 1.5713194608688354, + "learning_rate": 0.00018324377376123822, + "loss": 0.274, + "step": 7983 + }, + { + "epoch": 0.18715753324308326, + "grad_norm": 4.123602390289307, + "learning_rate": 0.00018323969181040096, + "loss": 0.741, + "step": 7984 + }, + { + "epoch": 0.18718097481788826, + "grad_norm": 3.290159225463867, + "learning_rate": 0.0001832356094079028, + "loss": 0.3945, + "step": 7985 + }, + { + "epoch": 0.18720441639269325, + "grad_norm": 3.919074773788452, + "learning_rate": 0.00018323152655376585, + "loss": 0.4399, + "step": 7986 + }, + { + "epoch": 0.18722785796749825, + "grad_norm": 2.8369064331054688, + "learning_rate": 0.00018322744324801232, + "loss": 0.9985, + "step": 7987 + }, + { + "epoch": 0.18725129954230324, + "grad_norm": 1.7633440494537354, + "learning_rate": 0.00018322335949066433, + "loss": 0.3604, + "step": 7988 + }, + { + "epoch": 0.18727474111710823, + "grad_norm": 2.667966842651367, + "learning_rate": 0.00018321927528174405, + "loss": 0.4254, + "step": 7989 + }, + { + "epoch": 0.18729818269191326, + "grad_norm": 3.4763004779815674, + "learning_rate": 0.00018321519062127365, + "loss": 0.2673, + "step": 7990 + }, + { + "epoch": 0.18732162426671825, + "grad_norm": 2.781324625015259, + "learning_rate": 0.0001832111055092753, + "loss": 0.2866, + "step": 7991 + }, + { + "epoch": 0.18734506584152325, + "grad_norm": 3.6761322021484375, + "learning_rate": 0.00018320701994577115, + "loss": 0.4839, + "step": 7992 + }, + { + "epoch": 0.18736850741632824, + "grad_norm": 3.466891288757324, + "learning_rate": 0.00018320293393078336, + "loss": 0.556, + "step": 7993 + }, + { + "epoch": 0.18739194899113323, + "grad_norm": 5.091977119445801, + "learning_rate": 0.00018319884746433412, + "loss": 0.5586, + "step": 7994 + }, + { + "epoch": 0.18741539056593823, + "grad_norm": 4.65836238861084, + "learning_rate": 0.00018319476054644559, + "loss": 0.4203, + "step": 7995 + }, + { + "epoch": 0.18743883214074322, + "grad_norm": 2.4780797958374023, + "learning_rate": 0.0001831906731771399, + "loss": 0.296, + "step": 7996 + }, + { + "epoch": 0.18746227371554822, + "grad_norm": 0.7377234697341919, + "learning_rate": 0.00018318658535643935, + "loss": 0.0736, + "step": 7997 + }, + { + "epoch": 0.1874857152903532, + "grad_norm": 4.094782829284668, + "learning_rate": 0.00018318249708436604, + "loss": 0.5332, + "step": 7998 + }, + { + "epoch": 0.1875091568651582, + "grad_norm": 3.029855728149414, + "learning_rate": 0.00018317840836094213, + "loss": 0.3131, + "step": 7999 + }, + { + "epoch": 0.1875325984399632, + "grad_norm": 3.107243537902832, + "learning_rate": 0.00018317431918618983, + "loss": 0.6843, + "step": 8000 + }, + { + "epoch": 0.1875560400147682, + "grad_norm": 2.6747474670410156, + "learning_rate": 0.00018317022956013137, + "loss": 0.3835, + "step": 8001 + }, + { + "epoch": 0.1875794815895732, + "grad_norm": 2.7831239700317383, + "learning_rate": 0.0001831661394827889, + "loss": 0.3266, + "step": 8002 + }, + { + "epoch": 0.18760292316437818, + "grad_norm": 4.558957099914551, + "learning_rate": 0.0001831620489541846, + "loss": 0.5653, + "step": 8003 + }, + { + "epoch": 0.18762636473918318, + "grad_norm": 4.316844940185547, + "learning_rate": 0.00018315795797434073, + "loss": 0.519, + "step": 8004 + }, + { + "epoch": 0.18764980631398817, + "grad_norm": 3.426910638809204, + "learning_rate": 0.0001831538665432794, + "loss": 0.5218, + "step": 8005 + }, + { + "epoch": 0.18767324788879317, + "grad_norm": 2.9157676696777344, + "learning_rate": 0.00018314977466102285, + "loss": 0.9343, + "step": 8006 + }, + { + "epoch": 0.18769668946359816, + "grad_norm": 0.8566151261329651, + "learning_rate": 0.0001831456823275933, + "loss": 0.0983, + "step": 8007 + }, + { + "epoch": 0.18772013103840315, + "grad_norm": 3.5337793827056885, + "learning_rate": 0.00018314158954301293, + "loss": 0.3524, + "step": 8008 + }, + { + "epoch": 0.18774357261320815, + "grad_norm": 3.536728620529175, + "learning_rate": 0.00018313749630730395, + "loss": 0.4712, + "step": 8009 + }, + { + "epoch": 0.18776701418801314, + "grad_norm": 0.9559693932533264, + "learning_rate": 0.00018313340262048858, + "loss": 0.1441, + "step": 8010 + }, + { + "epoch": 0.18779045576281814, + "grad_norm": 4.272099018096924, + "learning_rate": 0.00018312930848258906, + "loss": 0.4867, + "step": 8011 + }, + { + "epoch": 0.18781389733762313, + "grad_norm": 3.18192195892334, + "learning_rate": 0.00018312521389362755, + "loss": 0.2757, + "step": 8012 + }, + { + "epoch": 0.18783733891242813, + "grad_norm": 3.681485176086426, + "learning_rate": 0.00018312111885362628, + "loss": 0.3725, + "step": 8013 + }, + { + "epoch": 0.18786078048723312, + "grad_norm": 4.2526092529296875, + "learning_rate": 0.00018311702336260753, + "loss": 0.4289, + "step": 8014 + }, + { + "epoch": 0.18788422206203811, + "grad_norm": 4.692405700683594, + "learning_rate": 0.00018311292742059342, + "loss": 0.3839, + "step": 8015 + }, + { + "epoch": 0.18790766363684314, + "grad_norm": 2.6751015186309814, + "learning_rate": 0.00018310883102760626, + "loss": 1.8126, + "step": 8016 + }, + { + "epoch": 0.18793110521164813, + "grad_norm": 1.6835548877716064, + "learning_rate": 0.00018310473418366825, + "loss": 0.1523, + "step": 8017 + }, + { + "epoch": 0.18795454678645312, + "grad_norm": 4.115335464477539, + "learning_rate": 0.00018310063688880158, + "loss": 0.4663, + "step": 8018 + }, + { + "epoch": 0.18797798836125812, + "grad_norm": 3.457360029220581, + "learning_rate": 0.00018309653914302856, + "loss": 0.4173, + "step": 8019 + }, + { + "epoch": 0.1880014299360631, + "grad_norm": 2.163174867630005, + "learning_rate": 0.00018309244094637134, + "loss": 0.2879, + "step": 8020 + }, + { + "epoch": 0.1880248715108681, + "grad_norm": 4.632167339324951, + "learning_rate": 0.00018308834229885222, + "loss": 0.8123, + "step": 8021 + }, + { + "epoch": 0.1880483130856731, + "grad_norm": 5.07067346572876, + "learning_rate": 0.00018308424320049343, + "loss": 0.5379, + "step": 8022 + }, + { + "epoch": 0.1880717546604781, + "grad_norm": 1.5591753721237183, + "learning_rate": 0.0001830801436513172, + "loss": 0.1915, + "step": 8023 + }, + { + "epoch": 0.1880951962352831, + "grad_norm": 5.787128925323486, + "learning_rate": 0.00018307604365134576, + "loss": 0.6104, + "step": 8024 + }, + { + "epoch": 0.18811863781008809, + "grad_norm": 2.1711015701293945, + "learning_rate": 0.00018307194320060137, + "loss": 0.2529, + "step": 8025 + }, + { + "epoch": 0.18814207938489308, + "grad_norm": 6.635408878326416, + "learning_rate": 0.0001830678422991063, + "loss": 1.1655, + "step": 8026 + }, + { + "epoch": 0.18816552095969807, + "grad_norm": 4.4216485023498535, + "learning_rate": 0.00018306374094688278, + "loss": 0.4486, + "step": 8027 + }, + { + "epoch": 0.18818896253450307, + "grad_norm": 1.022283673286438, + "learning_rate": 0.00018305963914395302, + "loss": 0.174, + "step": 8028 + }, + { + "epoch": 0.18821240410930806, + "grad_norm": 1.6664142608642578, + "learning_rate": 0.00018305553689033937, + "loss": 0.1034, + "step": 8029 + }, + { + "epoch": 0.18823584568411306, + "grad_norm": 2.2735865116119385, + "learning_rate": 0.00018305143418606402, + "loss": 0.1093, + "step": 8030 + }, + { + "epoch": 0.18825928725891805, + "grad_norm": 6.076618671417236, + "learning_rate": 0.00018304733103114925, + "loss": 0.5812, + "step": 8031 + }, + { + "epoch": 0.18828272883372305, + "grad_norm": 0.9799401164054871, + "learning_rate": 0.00018304322742561735, + "loss": 0.1602, + "step": 8032 + }, + { + "epoch": 0.18830617040852804, + "grad_norm": 3.6131985187530518, + "learning_rate": 0.00018303912336949053, + "loss": 0.4805, + "step": 8033 + }, + { + "epoch": 0.18832961198333303, + "grad_norm": 1.4959568977355957, + "learning_rate": 0.00018303501886279114, + "loss": 0.2007, + "step": 8034 + }, + { + "epoch": 0.18835305355813803, + "grad_norm": 3.249072790145874, + "learning_rate": 0.00018303091390554137, + "loss": 0.4459, + "step": 8035 + }, + { + "epoch": 0.18837649513294302, + "grad_norm": 5.234620094299316, + "learning_rate": 0.00018302680849776354, + "loss": 0.7234, + "step": 8036 + }, + { + "epoch": 0.18839993670774802, + "grad_norm": 2.8000457286834717, + "learning_rate": 0.0001830227026394799, + "loss": 0.3195, + "step": 8037 + }, + { + "epoch": 0.188423378282553, + "grad_norm": 1.5214385986328125, + "learning_rate": 0.00018301859633071274, + "loss": 0.2507, + "step": 8038 + }, + { + "epoch": 0.188446819857358, + "grad_norm": 3.697662591934204, + "learning_rate": 0.00018301448957148435, + "loss": 0.3147, + "step": 8039 + }, + { + "epoch": 0.188470261432163, + "grad_norm": 0.7852441072463989, + "learning_rate": 0.000183010382361817, + "loss": 0.0894, + "step": 8040 + }, + { + "epoch": 0.18849370300696802, + "grad_norm": 2.1234912872314453, + "learning_rate": 0.000183006274701733, + "loss": 0.2244, + "step": 8041 + }, + { + "epoch": 0.18851714458177302, + "grad_norm": 3.059842348098755, + "learning_rate": 0.0001830021665912546, + "loss": 0.3769, + "step": 8042 + }, + { + "epoch": 0.188540586156578, + "grad_norm": 3.475214719772339, + "learning_rate": 0.0001829980580304041, + "loss": 2.2873, + "step": 8043 + }, + { + "epoch": 0.188564027731383, + "grad_norm": 2.926262617111206, + "learning_rate": 0.00018299394901920383, + "loss": 0.3542, + "step": 8044 + }, + { + "epoch": 0.188587469306188, + "grad_norm": 3.091479539871216, + "learning_rate": 0.00018298983955767603, + "loss": 0.3327, + "step": 8045 + }, + { + "epoch": 0.188610910880993, + "grad_norm": 3.310358762741089, + "learning_rate": 0.00018298572964584305, + "loss": 0.3969, + "step": 8046 + }, + { + "epoch": 0.188634352455798, + "grad_norm": 3.962733507156372, + "learning_rate": 0.00018298161928372717, + "loss": 0.3794, + "step": 8047 + }, + { + "epoch": 0.18865779403060298, + "grad_norm": 4.527224063873291, + "learning_rate": 0.00018297750847135066, + "loss": 0.5529, + "step": 8048 + }, + { + "epoch": 0.18868123560540798, + "grad_norm": 3.451622486114502, + "learning_rate": 0.0001829733972087359, + "loss": 0.6228, + "step": 8049 + }, + { + "epoch": 0.18870467718021297, + "grad_norm": 1.2683042287826538, + "learning_rate": 0.00018296928549590513, + "loss": 0.1677, + "step": 8050 + }, + { + "epoch": 0.18872811875501797, + "grad_norm": 2.596212863922119, + "learning_rate": 0.00018296517333288068, + "loss": 0.2115, + "step": 8051 + }, + { + "epoch": 0.18875156032982296, + "grad_norm": 2.6858327388763428, + "learning_rate": 0.00018296106071968488, + "loss": 0.2311, + "step": 8052 + }, + { + "epoch": 0.18877500190462795, + "grad_norm": 3.507584571838379, + "learning_rate": 0.00018295694765634003, + "loss": 0.283, + "step": 8053 + }, + { + "epoch": 0.18879844347943295, + "grad_norm": 4.22708797454834, + "learning_rate": 0.00018295283414286844, + "loss": 0.44, + "step": 8054 + }, + { + "epoch": 0.18882188505423794, + "grad_norm": 1.6796973943710327, + "learning_rate": 0.00018294872017929245, + "loss": 0.1797, + "step": 8055 + }, + { + "epoch": 0.18884532662904294, + "grad_norm": 4.148077011108398, + "learning_rate": 0.00018294460576563437, + "loss": 0.4167, + "step": 8056 + }, + { + "epoch": 0.18886876820384793, + "grad_norm": 1.3995277881622314, + "learning_rate": 0.0001829404909019165, + "loss": 0.3965, + "step": 8057 + }, + { + "epoch": 0.18889220977865293, + "grad_norm": 0.7731505632400513, + "learning_rate": 0.00018293637558816124, + "loss": 0.0753, + "step": 8058 + }, + { + "epoch": 0.18891565135345792, + "grad_norm": 4.745385646820068, + "learning_rate": 0.00018293225982439083, + "loss": 0.7245, + "step": 8059 + }, + { + "epoch": 0.18893909292826291, + "grad_norm": 1.3229414224624634, + "learning_rate": 0.00018292814361062768, + "loss": 0.1408, + "step": 8060 + }, + { + "epoch": 0.1889625345030679, + "grad_norm": 3.236034393310547, + "learning_rate": 0.00018292402694689407, + "loss": 1.2948, + "step": 8061 + }, + { + "epoch": 0.1889859760778729, + "grad_norm": 4.970794677734375, + "learning_rate": 0.00018291990983321238, + "loss": 0.5781, + "step": 8062 + }, + { + "epoch": 0.1890094176526779, + "grad_norm": 2.581281900405884, + "learning_rate": 0.00018291579226960492, + "loss": 1.4787, + "step": 8063 + }, + { + "epoch": 0.1890328592274829, + "grad_norm": 3.058695077896118, + "learning_rate": 0.00018291167425609403, + "loss": 0.2326, + "step": 8064 + }, + { + "epoch": 0.18905630080228789, + "grad_norm": 8.498015403747559, + "learning_rate": 0.00018290755579270207, + "loss": 0.5776, + "step": 8065 + }, + { + "epoch": 0.18907974237709288, + "grad_norm": 1.5406075716018677, + "learning_rate": 0.00018290343687945136, + "loss": 0.1691, + "step": 8066 + }, + { + "epoch": 0.1891031839518979, + "grad_norm": 1.4644067287445068, + "learning_rate": 0.0001828993175163643, + "loss": 0.1673, + "step": 8067 + }, + { + "epoch": 0.1891266255267029, + "grad_norm": 0.4934784770011902, + "learning_rate": 0.0001828951977034632, + "loss": 0.048, + "step": 8068 + }, + { + "epoch": 0.1891500671015079, + "grad_norm": 3.6484627723693848, + "learning_rate": 0.00018289107744077042, + "loss": 0.3903, + "step": 8069 + }, + { + "epoch": 0.18917350867631288, + "grad_norm": 2.8430988788604736, + "learning_rate": 0.00018288695672830834, + "loss": 0.2977, + "step": 8070 + }, + { + "epoch": 0.18919695025111788, + "grad_norm": 3.7870328426361084, + "learning_rate": 0.0001828828355660993, + "loss": 0.3211, + "step": 8071 + }, + { + "epoch": 0.18922039182592287, + "grad_norm": 2.5635859966278076, + "learning_rate": 0.00018287871395416568, + "loss": 0.1655, + "step": 8072 + }, + { + "epoch": 0.18924383340072787, + "grad_norm": 3.2048301696777344, + "learning_rate": 0.00018287459189252978, + "loss": 0.3747, + "step": 8073 + }, + { + "epoch": 0.18926727497553286, + "grad_norm": 4.189146041870117, + "learning_rate": 0.00018287046938121403, + "loss": 0.6702, + "step": 8074 + }, + { + "epoch": 0.18929071655033786, + "grad_norm": 1.468000054359436, + "learning_rate": 0.0001828663464202408, + "loss": 0.1909, + "step": 8075 + }, + { + "epoch": 0.18931415812514285, + "grad_norm": 2.026289939880371, + "learning_rate": 0.00018286222300963246, + "loss": 1.0139, + "step": 8076 + }, + { + "epoch": 0.18933759969994785, + "grad_norm": 2.191338539123535, + "learning_rate": 0.00018285809914941133, + "loss": 1.3172, + "step": 8077 + }, + { + "epoch": 0.18936104127475284, + "grad_norm": 1.4379663467407227, + "learning_rate": 0.00018285397483959982, + "loss": 0.1932, + "step": 8078 + }, + { + "epoch": 0.18938448284955783, + "grad_norm": 0.9143485426902771, + "learning_rate": 0.00018284985008022037, + "loss": 0.078, + "step": 8079 + }, + { + "epoch": 0.18940792442436283, + "grad_norm": 5.155972957611084, + "learning_rate": 0.00018284572487129523, + "loss": 0.7138, + "step": 8080 + }, + { + "epoch": 0.18943136599916782, + "grad_norm": 1.4426709413528442, + "learning_rate": 0.0001828415992128469, + "loss": 0.1879, + "step": 8081 + }, + { + "epoch": 0.18945480757397282, + "grad_norm": 2.2632999420166016, + "learning_rate": 0.00018283747310489772, + "loss": 0.2193, + "step": 8082 + }, + { + "epoch": 0.1894782491487778, + "grad_norm": 4.673175811767578, + "learning_rate": 0.00018283334654747008, + "loss": 0.424, + "step": 8083 + }, + { + "epoch": 0.1895016907235828, + "grad_norm": 4.4812445640563965, + "learning_rate": 0.0001828292195405864, + "loss": 0.5874, + "step": 8084 + }, + { + "epoch": 0.1895251322983878, + "grad_norm": 7.555484294891357, + "learning_rate": 0.000182825092084269, + "loss": 0.5642, + "step": 8085 + }, + { + "epoch": 0.1895485738731928, + "grad_norm": 6.827380657196045, + "learning_rate": 0.00018282096417854033, + "loss": 0.7559, + "step": 8086 + }, + { + "epoch": 0.1895720154479978, + "grad_norm": 0.8326976299285889, + "learning_rate": 0.0001828168358234228, + "loss": 0.1361, + "step": 8087 + }, + { + "epoch": 0.18959545702280278, + "grad_norm": 2.8957135677337646, + "learning_rate": 0.00018281270701893875, + "loss": 0.3495, + "step": 8088 + }, + { + "epoch": 0.18961889859760778, + "grad_norm": 2.758544921875, + "learning_rate": 0.00018280857776511065, + "loss": 0.2788, + "step": 8089 + }, + { + "epoch": 0.18964234017241277, + "grad_norm": 3.961387872695923, + "learning_rate": 0.00018280444806196089, + "loss": 0.3896, + "step": 8090 + }, + { + "epoch": 0.18966578174721777, + "grad_norm": 3.0590131282806396, + "learning_rate": 0.00018280031790951186, + "loss": 0.3245, + "step": 8091 + }, + { + "epoch": 0.18968922332202276, + "grad_norm": 3.4877562522888184, + "learning_rate": 0.00018279618730778597, + "loss": 1.8712, + "step": 8092 + }, + { + "epoch": 0.18971266489682778, + "grad_norm": 1.8944131135940552, + "learning_rate": 0.00018279205625680565, + "loss": 0.2327, + "step": 8093 + }, + { + "epoch": 0.18973610647163278, + "grad_norm": 4.770325183868408, + "learning_rate": 0.0001827879247565933, + "loss": 0.5315, + "step": 8094 + }, + { + "epoch": 0.18975954804643777, + "grad_norm": 4.019551753997803, + "learning_rate": 0.00018278379280717134, + "loss": 0.3706, + "step": 8095 + }, + { + "epoch": 0.18978298962124276, + "grad_norm": 6.115249156951904, + "learning_rate": 0.0001827796604085622, + "loss": 1.1271, + "step": 8096 + }, + { + "epoch": 0.18980643119604776, + "grad_norm": 4.9107890129089355, + "learning_rate": 0.00018277552756078825, + "loss": 0.7444, + "step": 8097 + }, + { + "epoch": 0.18982987277085275, + "grad_norm": 2.907905340194702, + "learning_rate": 0.000182771394263872, + "loss": 0.4012, + "step": 8098 + }, + { + "epoch": 0.18985331434565775, + "grad_norm": 2.344961404800415, + "learning_rate": 0.00018276726051783587, + "loss": 0.467, + "step": 8099 + }, + { + "epoch": 0.18987675592046274, + "grad_norm": 3.083918809890747, + "learning_rate": 0.00018276312632270223, + "loss": 0.4583, + "step": 8100 + }, + { + "epoch": 0.18990019749526774, + "grad_norm": 4.327312469482422, + "learning_rate": 0.00018275899167849354, + "loss": 0.5369, + "step": 8101 + }, + { + "epoch": 0.18992363907007273, + "grad_norm": 1.5082985162734985, + "learning_rate": 0.00018275485658523222, + "loss": 0.2468, + "step": 8102 + }, + { + "epoch": 0.18994708064487772, + "grad_norm": 0.934784471988678, + "learning_rate": 0.0001827507210429407, + "loss": 0.0913, + "step": 8103 + }, + { + "epoch": 0.18997052221968272, + "grad_norm": 3.306467056274414, + "learning_rate": 0.00018274658505164148, + "loss": 0.2055, + "step": 8104 + }, + { + "epoch": 0.1899939637944877, + "grad_norm": 4.236787796020508, + "learning_rate": 0.00018274244861135697, + "loss": 0.5893, + "step": 8105 + }, + { + "epoch": 0.1900174053692927, + "grad_norm": 4.22376823425293, + "learning_rate": 0.00018273831172210958, + "loss": 0.5013, + "step": 8106 + }, + { + "epoch": 0.1900408469440977, + "grad_norm": 1.6352649927139282, + "learning_rate": 0.0001827341743839218, + "loss": 0.2542, + "step": 8107 + }, + { + "epoch": 0.1900642885189027, + "grad_norm": 2.1716787815093994, + "learning_rate": 0.00018273003659681605, + "loss": 0.4163, + "step": 8108 + }, + { + "epoch": 0.1900877300937077, + "grad_norm": 3.3006763458251953, + "learning_rate": 0.00018272589836081478, + "loss": 1.6408, + "step": 8109 + }, + { + "epoch": 0.19011117166851269, + "grad_norm": 2.761338233947754, + "learning_rate": 0.00018272175967594048, + "loss": 0.3537, + "step": 8110 + }, + { + "epoch": 0.19013461324331768, + "grad_norm": 2.4548206329345703, + "learning_rate": 0.0001827176205422156, + "loss": 0.3216, + "step": 8111 + }, + { + "epoch": 0.19015805481812267, + "grad_norm": 4.073875427246094, + "learning_rate": 0.00018271348095966257, + "loss": 0.5921, + "step": 8112 + }, + { + "epoch": 0.19018149639292767, + "grad_norm": 3.020048141479492, + "learning_rate": 0.00018270934092830387, + "loss": 0.3918, + "step": 8113 + }, + { + "epoch": 0.19020493796773266, + "grad_norm": 3.998600959777832, + "learning_rate": 0.00018270520044816197, + "loss": 0.8193, + "step": 8114 + }, + { + "epoch": 0.19022837954253766, + "grad_norm": 0.8865960240364075, + "learning_rate": 0.0001827010595192593, + "loss": 0.0896, + "step": 8115 + }, + { + "epoch": 0.19025182111734265, + "grad_norm": 2.221240520477295, + "learning_rate": 0.00018269691814161838, + "loss": 1.2697, + "step": 8116 + }, + { + "epoch": 0.19027526269214765, + "grad_norm": 1.7123924493789673, + "learning_rate": 0.00018269277631526165, + "loss": 0.2426, + "step": 8117 + }, + { + "epoch": 0.19029870426695267, + "grad_norm": 4.954214096069336, + "learning_rate": 0.0001826886340402116, + "loss": 0.6789, + "step": 8118 + }, + { + "epoch": 0.19032214584175766, + "grad_norm": 3.566227674484253, + "learning_rate": 0.00018268449131649068, + "loss": 0.3025, + "step": 8119 + }, + { + "epoch": 0.19034558741656266, + "grad_norm": 1.3829888105392456, + "learning_rate": 0.0001826803481441214, + "loss": 0.1267, + "step": 8120 + }, + { + "epoch": 0.19036902899136765, + "grad_norm": 3.451915740966797, + "learning_rate": 0.0001826762045231262, + "loss": 0.3315, + "step": 8121 + }, + { + "epoch": 0.19039247056617264, + "grad_norm": 2.74985671043396, + "learning_rate": 0.00018267206045352762, + "loss": 0.3672, + "step": 8122 + }, + { + "epoch": 0.19041591214097764, + "grad_norm": 1.1405346393585205, + "learning_rate": 0.00018266791593534812, + "loss": 0.1534, + "step": 8123 + }, + { + "epoch": 0.19043935371578263, + "grad_norm": 3.934452772140503, + "learning_rate": 0.00018266377096861014, + "loss": 0.6272, + "step": 8124 + }, + { + "epoch": 0.19046279529058763, + "grad_norm": 2.5839779376983643, + "learning_rate": 0.00018265962555333624, + "loss": 0.3429, + "step": 8125 + }, + { + "epoch": 0.19048623686539262, + "grad_norm": 5.025559902191162, + "learning_rate": 0.0001826554796895489, + "loss": 1.0164, + "step": 8126 + }, + { + "epoch": 0.19050967844019762, + "grad_norm": 5.272496223449707, + "learning_rate": 0.0001826513333772706, + "loss": 0.7435, + "step": 8127 + }, + { + "epoch": 0.1905331200150026, + "grad_norm": 1.3432966470718384, + "learning_rate": 0.0001826471866165238, + "loss": 0.1825, + "step": 8128 + }, + { + "epoch": 0.1905565615898076, + "grad_norm": 5.14666748046875, + "learning_rate": 0.00018264303940733107, + "loss": 0.7395, + "step": 8129 + }, + { + "epoch": 0.1905800031646126, + "grad_norm": 4.29060173034668, + "learning_rate": 0.00018263889174971492, + "loss": 0.605, + "step": 8130 + }, + { + "epoch": 0.1906034447394176, + "grad_norm": 2.8576130867004395, + "learning_rate": 0.00018263474364369777, + "loss": 0.3614, + "step": 8131 + }, + { + "epoch": 0.1906268863142226, + "grad_norm": 1.9667378664016724, + "learning_rate": 0.0001826305950893022, + "loss": 0.3211, + "step": 8132 + }, + { + "epoch": 0.19065032788902758, + "grad_norm": 2.0691652297973633, + "learning_rate": 0.0001826264460865507, + "loss": 1.1604, + "step": 8133 + }, + { + "epoch": 0.19067376946383258, + "grad_norm": 3.101213216781616, + "learning_rate": 0.00018262229663546577, + "loss": 0.4793, + "step": 8134 + }, + { + "epoch": 0.19069721103863757, + "grad_norm": 2.551982879638672, + "learning_rate": 0.00018261814673606994, + "loss": 0.2175, + "step": 8135 + }, + { + "epoch": 0.19072065261344257, + "grad_norm": 0.7638635635375977, + "learning_rate": 0.00018261399638838573, + "loss": 0.082, + "step": 8136 + }, + { + "epoch": 0.19074409418824756, + "grad_norm": 2.561755657196045, + "learning_rate": 0.00018260984559243564, + "loss": 0.3691, + "step": 8137 + }, + { + "epoch": 0.19076753576305255, + "grad_norm": 3.1514601707458496, + "learning_rate": 0.0001826056943482422, + "loss": 0.4227, + "step": 8138 + }, + { + "epoch": 0.19079097733785755, + "grad_norm": 6.040432929992676, + "learning_rate": 0.00018260154265582794, + "loss": 0.5272, + "step": 8139 + }, + { + "epoch": 0.19081441891266254, + "grad_norm": 1.344957947731018, + "learning_rate": 0.0001825973905152154, + "loss": 0.2316, + "step": 8140 + }, + { + "epoch": 0.19083786048746754, + "grad_norm": 2.7151756286621094, + "learning_rate": 0.00018259323792642711, + "loss": 0.2482, + "step": 8141 + }, + { + "epoch": 0.19086130206227253, + "grad_norm": 3.12266206741333, + "learning_rate": 0.00018258908488948556, + "loss": 0.4525, + "step": 8142 + }, + { + "epoch": 0.19088474363707753, + "grad_norm": 1.2526087760925293, + "learning_rate": 0.00018258493140441332, + "loss": 0.2097, + "step": 8143 + }, + { + "epoch": 0.19090818521188255, + "grad_norm": 1.9451724290847778, + "learning_rate": 0.00018258077747123293, + "loss": 0.2951, + "step": 8144 + }, + { + "epoch": 0.19093162678668754, + "grad_norm": 3.789463996887207, + "learning_rate": 0.00018257662308996693, + "loss": 0.5289, + "step": 8145 + }, + { + "epoch": 0.19095506836149254, + "grad_norm": 0.49139484763145447, + "learning_rate": 0.00018257246826063782, + "loss": 0.0526, + "step": 8146 + }, + { + "epoch": 0.19097850993629753, + "grad_norm": 0.5126243829727173, + "learning_rate": 0.0001825683129832682, + "loss": 0.0719, + "step": 8147 + }, + { + "epoch": 0.19100195151110252, + "grad_norm": 4.1847429275512695, + "learning_rate": 0.0001825641572578806, + "loss": 0.4215, + "step": 8148 + }, + { + "epoch": 0.19102539308590752, + "grad_norm": 4.2205729484558105, + "learning_rate": 0.00018256000108449752, + "loss": 0.5872, + "step": 8149 + }, + { + "epoch": 0.1910488346607125, + "grad_norm": 3.815051555633545, + "learning_rate": 0.00018255584446314157, + "loss": 0.534, + "step": 8150 + }, + { + "epoch": 0.1910722762355175, + "grad_norm": 1.3718372583389282, + "learning_rate": 0.00018255168739383533, + "loss": 0.1807, + "step": 8151 + }, + { + "epoch": 0.1910957178103225, + "grad_norm": 3.7556960582733154, + "learning_rate": 0.00018254752987660125, + "loss": 0.4409, + "step": 8152 + }, + { + "epoch": 0.1911191593851275, + "grad_norm": 4.1616950035095215, + "learning_rate": 0.00018254337191146199, + "loss": 0.3052, + "step": 8153 + }, + { + "epoch": 0.1911426009599325, + "grad_norm": 3.4247336387634277, + "learning_rate": 0.00018253921349844005, + "loss": 0.4975, + "step": 8154 + }, + { + "epoch": 0.19116604253473748, + "grad_norm": 3.7968757152557373, + "learning_rate": 0.00018253505463755804, + "loss": 0.3059, + "step": 8155 + }, + { + "epoch": 0.19118948410954248, + "grad_norm": 6.0842061042785645, + "learning_rate": 0.0001825308953288385, + "loss": 0.9979, + "step": 8156 + }, + { + "epoch": 0.19121292568434747, + "grad_norm": 3.1095638275146484, + "learning_rate": 0.000182526735572304, + "loss": 0.3126, + "step": 8157 + }, + { + "epoch": 0.19123636725915247, + "grad_norm": 2.3102500438690186, + "learning_rate": 0.00018252257536797707, + "loss": 0.2489, + "step": 8158 + }, + { + "epoch": 0.19125980883395746, + "grad_norm": 3.27386474609375, + "learning_rate": 0.00018251841471588037, + "loss": 0.4449, + "step": 8159 + }, + { + "epoch": 0.19128325040876246, + "grad_norm": 2.9594473838806152, + "learning_rate": 0.00018251425361603645, + "loss": 0.4576, + "step": 8160 + }, + { + "epoch": 0.19130669198356745, + "grad_norm": 5.700272083282471, + "learning_rate": 0.00018251009206846783, + "loss": 0.6423, + "step": 8161 + }, + { + "epoch": 0.19133013355837244, + "grad_norm": 1.1768531799316406, + "learning_rate": 0.00018250593007319713, + "loss": 0.136, + "step": 8162 + }, + { + "epoch": 0.19135357513317744, + "grad_norm": 2.620954990386963, + "learning_rate": 0.00018250176763024694, + "loss": 0.2779, + "step": 8163 + }, + { + "epoch": 0.19137701670798243, + "grad_norm": 2.998833656311035, + "learning_rate": 0.00018249760473963987, + "loss": 0.4892, + "step": 8164 + }, + { + "epoch": 0.19140045828278743, + "grad_norm": 2.6575706005096436, + "learning_rate": 0.00018249344140139847, + "loss": 1.5837, + "step": 8165 + }, + { + "epoch": 0.19142389985759242, + "grad_norm": 3.490852117538452, + "learning_rate": 0.0001824892776155453, + "loss": 1.4492, + "step": 8166 + }, + { + "epoch": 0.19144734143239742, + "grad_norm": 3.8820888996124268, + "learning_rate": 0.00018248511338210302, + "loss": 0.3628, + "step": 8167 + }, + { + "epoch": 0.1914707830072024, + "grad_norm": 3.7970597743988037, + "learning_rate": 0.00018248094870109417, + "loss": 0.3806, + "step": 8168 + }, + { + "epoch": 0.19149422458200743, + "grad_norm": 3.8369693756103516, + "learning_rate": 0.0001824767835725414, + "loss": 0.4567, + "step": 8169 + }, + { + "epoch": 0.19151766615681243, + "grad_norm": 6.102745532989502, + "learning_rate": 0.00018247261799646727, + "loss": 0.708, + "step": 8170 + }, + { + "epoch": 0.19154110773161742, + "grad_norm": 4.730750560760498, + "learning_rate": 0.0001824684519728944, + "loss": 0.3089, + "step": 8171 + }, + { + "epoch": 0.19156454930642242, + "grad_norm": 3.5175507068634033, + "learning_rate": 0.00018246428550184538, + "loss": 0.406, + "step": 8172 + }, + { + "epoch": 0.1915879908812274, + "grad_norm": 4.674038410186768, + "learning_rate": 0.00018246011858334284, + "loss": 0.4797, + "step": 8173 + }, + { + "epoch": 0.1916114324560324, + "grad_norm": 3.976970672607422, + "learning_rate": 0.00018245595121740938, + "loss": 0.5665, + "step": 8174 + }, + { + "epoch": 0.1916348740308374, + "grad_norm": 1.4020980596542358, + "learning_rate": 0.00018245178340406762, + "loss": 0.2441, + "step": 8175 + }, + { + "epoch": 0.1916583156056424, + "grad_norm": 2.94789981842041, + "learning_rate": 0.00018244761514334013, + "loss": 0.3847, + "step": 8176 + }, + { + "epoch": 0.1916817571804474, + "grad_norm": 2.382178544998169, + "learning_rate": 0.00018244344643524957, + "loss": 0.2669, + "step": 8177 + }, + { + "epoch": 0.19170519875525238, + "grad_norm": 2.071241617202759, + "learning_rate": 0.0001824392772798186, + "loss": 0.2751, + "step": 8178 + }, + { + "epoch": 0.19172864033005738, + "grad_norm": 2.1719844341278076, + "learning_rate": 0.00018243510767706976, + "loss": 0.2936, + "step": 8179 + }, + { + "epoch": 0.19175208190486237, + "grad_norm": 2.768054246902466, + "learning_rate": 0.0001824309376270257, + "loss": 0.4839, + "step": 8180 + }, + { + "epoch": 0.19177552347966736, + "grad_norm": 1.7359782457351685, + "learning_rate": 0.00018242676712970903, + "loss": 1.2178, + "step": 8181 + }, + { + "epoch": 0.19179896505447236, + "grad_norm": 0.46586373448371887, + "learning_rate": 0.00018242259618514247, + "loss": 0.0476, + "step": 8182 + }, + { + "epoch": 0.19182240662927735, + "grad_norm": 3.7222585678100586, + "learning_rate": 0.00018241842479334853, + "loss": 0.4261, + "step": 8183 + }, + { + "epoch": 0.19184584820408235, + "grad_norm": 3.857729911804199, + "learning_rate": 0.00018241425295434992, + "loss": 0.3642, + "step": 8184 + }, + { + "epoch": 0.19186928977888734, + "grad_norm": 5.092158317565918, + "learning_rate": 0.00018241008066816926, + "loss": 0.507, + "step": 8185 + }, + { + "epoch": 0.19189273135369234, + "grad_norm": 4.316263675689697, + "learning_rate": 0.00018240590793482917, + "loss": 0.5673, + "step": 8186 + }, + { + "epoch": 0.19191617292849733, + "grad_norm": 3.4336564540863037, + "learning_rate": 0.00018240173475435234, + "loss": 0.2694, + "step": 8187 + }, + { + "epoch": 0.19193961450330232, + "grad_norm": 2.954364776611328, + "learning_rate": 0.00018239756112676134, + "loss": 0.2719, + "step": 8188 + }, + { + "epoch": 0.19196305607810732, + "grad_norm": 4.657783508300781, + "learning_rate": 0.00018239338705207888, + "loss": 0.5202, + "step": 8189 + }, + { + "epoch": 0.1919864976529123, + "grad_norm": 1.1493316888809204, + "learning_rate": 0.00018238921253032757, + "loss": 0.1274, + "step": 8190 + }, + { + "epoch": 0.1920099392277173, + "grad_norm": 4.0095367431640625, + "learning_rate": 0.00018238503756153008, + "loss": 0.4747, + "step": 8191 + }, + { + "epoch": 0.1920333808025223, + "grad_norm": 2.9697587490081787, + "learning_rate": 0.0001823808621457091, + "loss": 0.366, + "step": 8192 + }, + { + "epoch": 0.1920568223773273, + "grad_norm": 3.165579080581665, + "learning_rate": 0.0001823766862828872, + "loss": 2.0246, + "step": 8193 + }, + { + "epoch": 0.1920802639521323, + "grad_norm": 2.0253381729125977, + "learning_rate": 0.00018237250997308707, + "loss": 0.2331, + "step": 8194 + }, + { + "epoch": 0.1921037055269373, + "grad_norm": 2.263251543045044, + "learning_rate": 0.00018236833321633143, + "loss": 0.2663, + "step": 8195 + }, + { + "epoch": 0.1921271471017423, + "grad_norm": 3.156978130340576, + "learning_rate": 0.00018236415601264287, + "loss": 0.3759, + "step": 8196 + }, + { + "epoch": 0.1921505886765473, + "grad_norm": 4.480553150177002, + "learning_rate": 0.00018235997836204407, + "loss": 1.7567, + "step": 8197 + }, + { + "epoch": 0.1921740302513523, + "grad_norm": 0.7849575281143188, + "learning_rate": 0.00018235580026455773, + "loss": 0.1332, + "step": 8198 + }, + { + "epoch": 0.1921974718261573, + "grad_norm": 5.049940586090088, + "learning_rate": 0.00018235162172020652, + "loss": 0.5923, + "step": 8199 + }, + { + "epoch": 0.19222091340096228, + "grad_norm": 5.527289390563965, + "learning_rate": 0.00018234744272901309, + "loss": 0.6159, + "step": 8200 + }, + { + "epoch": 0.19224435497576728, + "grad_norm": 3.4859516620635986, + "learning_rate": 0.0001823432632910001, + "loss": 0.2708, + "step": 8201 + }, + { + "epoch": 0.19226779655057227, + "grad_norm": 3.477398157119751, + "learning_rate": 0.00018233908340619027, + "loss": 0.4363, + "step": 8202 + }, + { + "epoch": 0.19229123812537727, + "grad_norm": 3.3477284908294678, + "learning_rate": 0.00018233490307460626, + "loss": 0.5733, + "step": 8203 + }, + { + "epoch": 0.19231467970018226, + "grad_norm": 3.6331324577331543, + "learning_rate": 0.00018233072229627072, + "loss": 0.3565, + "step": 8204 + }, + { + "epoch": 0.19233812127498726, + "grad_norm": 2.403096914291382, + "learning_rate": 0.00018232654107120638, + "loss": 0.2472, + "step": 8205 + }, + { + "epoch": 0.19236156284979225, + "grad_norm": 1.962714672088623, + "learning_rate": 0.00018232235939943592, + "loss": 1.586, + "step": 8206 + }, + { + "epoch": 0.19238500442459724, + "grad_norm": 2.3255107402801514, + "learning_rate": 0.00018231817728098202, + "loss": 1.6792, + "step": 8207 + }, + { + "epoch": 0.19240844599940224, + "grad_norm": 6.157334327697754, + "learning_rate": 0.00018231399471586737, + "loss": 0.7696, + "step": 8208 + }, + { + "epoch": 0.19243188757420723, + "grad_norm": 7.991459369659424, + "learning_rate": 0.00018230981170411467, + "loss": 0.3712, + "step": 8209 + }, + { + "epoch": 0.19245532914901223, + "grad_norm": 4.084059238433838, + "learning_rate": 0.00018230562824574661, + "loss": 0.4584, + "step": 8210 + }, + { + "epoch": 0.19247877072381722, + "grad_norm": 4.85173225402832, + "learning_rate": 0.00018230144434078593, + "loss": 0.6664, + "step": 8211 + }, + { + "epoch": 0.19250221229862222, + "grad_norm": 2.103391647338867, + "learning_rate": 0.00018229725998925527, + "loss": 1.1631, + "step": 8212 + }, + { + "epoch": 0.1925256538734272, + "grad_norm": 0.9162964820861816, + "learning_rate": 0.00018229307519117738, + "loss": 0.1638, + "step": 8213 + }, + { + "epoch": 0.1925490954482322, + "grad_norm": 1.305213451385498, + "learning_rate": 0.00018228888994657494, + "loss": 0.6555, + "step": 8214 + }, + { + "epoch": 0.1925725370230372, + "grad_norm": 0.8721217513084412, + "learning_rate": 0.00018228470425547068, + "loss": 0.0962, + "step": 8215 + }, + { + "epoch": 0.1925959785978422, + "grad_norm": 4.016368389129639, + "learning_rate": 0.0001822805181178873, + "loss": 0.5222, + "step": 8216 + }, + { + "epoch": 0.1926194201726472, + "grad_norm": 2.921002149581909, + "learning_rate": 0.0001822763315338475, + "loss": 0.397, + "step": 8217 + }, + { + "epoch": 0.19264286174745218, + "grad_norm": 3.403395652770996, + "learning_rate": 0.000182272144503374, + "loss": 1.4853, + "step": 8218 + }, + { + "epoch": 0.19266630332225718, + "grad_norm": 2.4396588802337646, + "learning_rate": 0.00018226795702648959, + "loss": 0.3351, + "step": 8219 + }, + { + "epoch": 0.1926897448970622, + "grad_norm": 5.7723259925842285, + "learning_rate": 0.00018226376910321688, + "loss": 1.2477, + "step": 8220 + }, + { + "epoch": 0.1927131864718672, + "grad_norm": 1.3859238624572754, + "learning_rate": 0.00018225958073357865, + "loss": 0.1815, + "step": 8221 + }, + { + "epoch": 0.1927366280466722, + "grad_norm": 1.1813862323760986, + "learning_rate": 0.00018225539191759764, + "loss": 0.1715, + "step": 8222 + }, + { + "epoch": 0.19276006962147718, + "grad_norm": 5.57703161239624, + "learning_rate": 0.00018225120265529654, + "loss": 0.528, + "step": 8223 + }, + { + "epoch": 0.19278351119628218, + "grad_norm": 2.8846395015716553, + "learning_rate": 0.00018224701294669812, + "loss": 0.2197, + "step": 8224 + }, + { + "epoch": 0.19280695277108717, + "grad_norm": 1.0051600933074951, + "learning_rate": 0.00018224282279182506, + "loss": 0.1594, + "step": 8225 + }, + { + "epoch": 0.19283039434589216, + "grad_norm": 4.120222568511963, + "learning_rate": 0.0001822386321907002, + "loss": 0.4008, + "step": 8226 + }, + { + "epoch": 0.19285383592069716, + "grad_norm": 2.2807257175445557, + "learning_rate": 0.00018223444114334614, + "loss": 0.1417, + "step": 8227 + }, + { + "epoch": 0.19287727749550215, + "grad_norm": 2.2976672649383545, + "learning_rate": 0.00018223024964978573, + "loss": 0.2839, + "step": 8228 + }, + { + "epoch": 0.19290071907030715, + "grad_norm": 2.6585192680358887, + "learning_rate": 0.00018222605771004165, + "loss": 0.3664, + "step": 8229 + }, + { + "epoch": 0.19292416064511214, + "grad_norm": 0.9379342794418335, + "learning_rate": 0.0001822218653241367, + "loss": 1.089, + "step": 8230 + }, + { + "epoch": 0.19294760221991714, + "grad_norm": 2.0060501098632812, + "learning_rate": 0.00018221767249209357, + "loss": 0.2379, + "step": 8231 + }, + { + "epoch": 0.19297104379472213, + "grad_norm": 0.5524731278419495, + "learning_rate": 0.00018221347921393503, + "loss": 0.0536, + "step": 8232 + }, + { + "epoch": 0.19299448536952712, + "grad_norm": 2.684556007385254, + "learning_rate": 0.00018220928548968388, + "loss": 0.2247, + "step": 8233 + }, + { + "epoch": 0.19301792694433212, + "grad_norm": 4.587380886077881, + "learning_rate": 0.00018220509131936278, + "loss": 0.4099, + "step": 8234 + }, + { + "epoch": 0.1930413685191371, + "grad_norm": 2.445356845855713, + "learning_rate": 0.00018220089670299458, + "loss": 0.2806, + "step": 8235 + }, + { + "epoch": 0.1930648100939421, + "grad_norm": 1.321282148361206, + "learning_rate": 0.000182196701640602, + "loss": 0.1637, + "step": 8236 + }, + { + "epoch": 0.1930882516687471, + "grad_norm": 3.626436710357666, + "learning_rate": 0.0001821925061322078, + "loss": 0.2794, + "step": 8237 + }, + { + "epoch": 0.1931116932435521, + "grad_norm": 3.2493884563446045, + "learning_rate": 0.00018218831017783476, + "loss": 0.6074, + "step": 8238 + }, + { + "epoch": 0.1931351348183571, + "grad_norm": 3.0923445224761963, + "learning_rate": 0.00018218411377750565, + "loss": 0.3369, + "step": 8239 + }, + { + "epoch": 0.19315857639316208, + "grad_norm": 4.312252998352051, + "learning_rate": 0.00018217991693124318, + "loss": 0.546, + "step": 8240 + }, + { + "epoch": 0.19318201796796708, + "grad_norm": 2.41629958152771, + "learning_rate": 0.00018217571963907017, + "loss": 1.1757, + "step": 8241 + }, + { + "epoch": 0.19320545954277207, + "grad_norm": 1.5723322629928589, + "learning_rate": 0.00018217152190100946, + "loss": 0.239, + "step": 8242 + }, + { + "epoch": 0.19322890111757707, + "grad_norm": 0.7201161980628967, + "learning_rate": 0.00018216732371708372, + "loss": 0.1267, + "step": 8243 + }, + { + "epoch": 0.19325234269238206, + "grad_norm": 2.2793385982513428, + "learning_rate": 0.0001821631250873158, + "loss": 0.2037, + "step": 8244 + }, + { + "epoch": 0.19327578426718706, + "grad_norm": 1.6601624488830566, + "learning_rate": 0.00018215892601172842, + "loss": 0.1989, + "step": 8245 + }, + { + "epoch": 0.19329922584199208, + "grad_norm": 2.669005870819092, + "learning_rate": 0.0001821547264903444, + "loss": 1.3386, + "step": 8246 + }, + { + "epoch": 0.19332266741679707, + "grad_norm": 1.9419901371002197, + "learning_rate": 0.00018215052652318652, + "loss": 1.2978, + "step": 8247 + }, + { + "epoch": 0.19334610899160207, + "grad_norm": 4.350695610046387, + "learning_rate": 0.0001821463261102776, + "loss": 0.5982, + "step": 8248 + }, + { + "epoch": 0.19336955056640706, + "grad_norm": 2.215693235397339, + "learning_rate": 0.00018214212525164038, + "loss": 0.2568, + "step": 8249 + }, + { + "epoch": 0.19339299214121206, + "grad_norm": 4.702511310577393, + "learning_rate": 0.0001821379239472977, + "loss": 0.7132, + "step": 8250 + }, + { + "epoch": 0.19341643371601705, + "grad_norm": 0.9792750477790833, + "learning_rate": 0.00018213372219727232, + "loss": 0.0631, + "step": 8251 + }, + { + "epoch": 0.19343987529082204, + "grad_norm": 1.19298255443573, + "learning_rate": 0.00018212952000158704, + "loss": 0.0708, + "step": 8252 + }, + { + "epoch": 0.19346331686562704, + "grad_norm": 2.8371996879577637, + "learning_rate": 0.0001821253173602647, + "loss": 1.8226, + "step": 8253 + }, + { + "epoch": 0.19348675844043203, + "grad_norm": 1.7817823886871338, + "learning_rate": 0.00018212111427332806, + "loss": 0.2752, + "step": 8254 + }, + { + "epoch": 0.19351020001523703, + "grad_norm": 2.8738338947296143, + "learning_rate": 0.00018211691074079996, + "loss": 0.3175, + "step": 8255 + }, + { + "epoch": 0.19353364159004202, + "grad_norm": 0.7314059734344482, + "learning_rate": 0.00018211270676270317, + "loss": 0.0583, + "step": 8256 + }, + { + "epoch": 0.19355708316484702, + "grad_norm": 2.770390510559082, + "learning_rate": 0.00018210850233906057, + "loss": 0.1476, + "step": 8257 + }, + { + "epoch": 0.193580524739652, + "grad_norm": 6.573898792266846, + "learning_rate": 0.0001821042974698949, + "loss": 0.8369, + "step": 8258 + }, + { + "epoch": 0.193603966314457, + "grad_norm": 3.145540237426758, + "learning_rate": 0.00018210009215522899, + "loss": 0.4025, + "step": 8259 + }, + { + "epoch": 0.193627407889262, + "grad_norm": 1.0719010829925537, + "learning_rate": 0.00018209588639508568, + "loss": 0.0981, + "step": 8260 + }, + { + "epoch": 0.193650849464067, + "grad_norm": 2.7858052253723145, + "learning_rate": 0.0001820916801894878, + "loss": 0.0733, + "step": 8261 + }, + { + "epoch": 0.193674291038872, + "grad_norm": 2.6951558589935303, + "learning_rate": 0.00018208747353845814, + "loss": 0.2204, + "step": 8262 + }, + { + "epoch": 0.19369773261367698, + "grad_norm": 3.889528274536133, + "learning_rate": 0.00018208326644201953, + "loss": 0.3146, + "step": 8263 + }, + { + "epoch": 0.19372117418848198, + "grad_norm": 2.2287192344665527, + "learning_rate": 0.00018207905890019484, + "loss": 0.2193, + "step": 8264 + }, + { + "epoch": 0.19374461576328697, + "grad_norm": 3.9825217723846436, + "learning_rate": 0.00018207485091300684, + "loss": 0.3759, + "step": 8265 + }, + { + "epoch": 0.19376805733809196, + "grad_norm": 1.5551276206970215, + "learning_rate": 0.0001820706424804784, + "loss": 0.1711, + "step": 8266 + }, + { + "epoch": 0.19379149891289696, + "grad_norm": 3.4095749855041504, + "learning_rate": 0.00018206643360263236, + "loss": 0.4125, + "step": 8267 + }, + { + "epoch": 0.19381494048770195, + "grad_norm": 3.827840805053711, + "learning_rate": 0.00018206222427949152, + "loss": 0.4659, + "step": 8268 + }, + { + "epoch": 0.19383838206250695, + "grad_norm": 3.330747127532959, + "learning_rate": 0.00018205801451107875, + "loss": 0.4008, + "step": 8269 + }, + { + "epoch": 0.19386182363731194, + "grad_norm": 3.689049243927002, + "learning_rate": 0.0001820538042974169, + "loss": 0.5159, + "step": 8270 + }, + { + "epoch": 0.19388526521211694, + "grad_norm": 3.807082414627075, + "learning_rate": 0.0001820495936385288, + "loss": 0.4245, + "step": 8271 + }, + { + "epoch": 0.19390870678692196, + "grad_norm": 3.6602306365966797, + "learning_rate": 0.00018204538253443727, + "loss": 0.5926, + "step": 8272 + }, + { + "epoch": 0.19393214836172695, + "grad_norm": 6.409303188323975, + "learning_rate": 0.0001820411709851652, + "loss": 0.3039, + "step": 8273 + }, + { + "epoch": 0.19395558993653195, + "grad_norm": 3.2458319664001465, + "learning_rate": 0.00018203695899073546, + "loss": 0.3993, + "step": 8274 + }, + { + "epoch": 0.19397903151133694, + "grad_norm": 10.778141975402832, + "learning_rate": 0.00018203274655117084, + "loss": 0.5739, + "step": 8275 + }, + { + "epoch": 0.19400247308614194, + "grad_norm": 2.9276583194732666, + "learning_rate": 0.00018202853366649423, + "loss": 1.3011, + "step": 8276 + }, + { + "epoch": 0.19402591466094693, + "grad_norm": 2.847435235977173, + "learning_rate": 0.00018202432033672853, + "loss": 0.3593, + "step": 8277 + }, + { + "epoch": 0.19404935623575192, + "grad_norm": 3.796370267868042, + "learning_rate": 0.0001820201065618965, + "loss": 0.5293, + "step": 8278 + }, + { + "epoch": 0.19407279781055692, + "grad_norm": 4.276762008666992, + "learning_rate": 0.00018201589234202113, + "loss": 0.5491, + "step": 8279 + }, + { + "epoch": 0.1940962393853619, + "grad_norm": 1.3877688646316528, + "learning_rate": 0.00018201167767712518, + "loss": 0.2176, + "step": 8280 + }, + { + "epoch": 0.1941196809601669, + "grad_norm": 3.3138539791107178, + "learning_rate": 0.00018200746256723157, + "loss": 0.4581, + "step": 8281 + }, + { + "epoch": 0.1941431225349719, + "grad_norm": 1.1007359027862549, + "learning_rate": 0.00018200324701236314, + "loss": 0.1311, + "step": 8282 + }, + { + "epoch": 0.1941665641097769, + "grad_norm": 4.256620407104492, + "learning_rate": 0.00018199903101254284, + "loss": 0.5319, + "step": 8283 + }, + { + "epoch": 0.1941900056845819, + "grad_norm": 1.5431901216506958, + "learning_rate": 0.00018199481456779347, + "loss": 0.2538, + "step": 8284 + }, + { + "epoch": 0.19421344725938688, + "grad_norm": 3.6220226287841797, + "learning_rate": 0.0001819905976781379, + "loss": 0.4123, + "step": 8285 + }, + { + "epoch": 0.19423688883419188, + "grad_norm": 3.2630038261413574, + "learning_rate": 0.00018198638034359907, + "loss": 0.2906, + "step": 8286 + }, + { + "epoch": 0.19426033040899687, + "grad_norm": 3.5652096271514893, + "learning_rate": 0.00018198216256419983, + "loss": 0.5587, + "step": 8287 + }, + { + "epoch": 0.19428377198380187, + "grad_norm": 2.1140663623809814, + "learning_rate": 0.00018197794433996305, + "loss": 0.21, + "step": 8288 + }, + { + "epoch": 0.19430721355860686, + "grad_norm": 4.404540061950684, + "learning_rate": 0.00018197372567091167, + "loss": 0.751, + "step": 8289 + }, + { + "epoch": 0.19433065513341186, + "grad_norm": 3.525296926498413, + "learning_rate": 0.00018196950655706854, + "loss": 0.4289, + "step": 8290 + }, + { + "epoch": 0.19435409670821685, + "grad_norm": 5.58989953994751, + "learning_rate": 0.00018196528699845655, + "loss": 0.6513, + "step": 8291 + }, + { + "epoch": 0.19437753828302184, + "grad_norm": 1.903665542602539, + "learning_rate": 0.00018196106699509864, + "loss": 1.3302, + "step": 8292 + }, + { + "epoch": 0.19440097985782684, + "grad_norm": 2.337427854537964, + "learning_rate": 0.00018195684654701764, + "loss": 0.188, + "step": 8293 + }, + { + "epoch": 0.19442442143263183, + "grad_norm": 4.618991851806641, + "learning_rate": 0.0001819526256542365, + "loss": 0.3647, + "step": 8294 + }, + { + "epoch": 0.19444786300743683, + "grad_norm": 3.085859775543213, + "learning_rate": 0.0001819484043167781, + "loss": 0.3596, + "step": 8295 + }, + { + "epoch": 0.19447130458224182, + "grad_norm": 4.816216945648193, + "learning_rate": 0.00018194418253466538, + "loss": 0.5594, + "step": 8296 + }, + { + "epoch": 0.19449474615704684, + "grad_norm": 6.143423080444336, + "learning_rate": 0.00018193996030792123, + "loss": 0.8982, + "step": 8297 + }, + { + "epoch": 0.19451818773185184, + "grad_norm": 4.323556423187256, + "learning_rate": 0.0001819357376365685, + "loss": 0.6761, + "step": 8298 + }, + { + "epoch": 0.19454162930665683, + "grad_norm": 2.430506467819214, + "learning_rate": 0.00018193151452063018, + "loss": 0.2305, + "step": 8299 + }, + { + "epoch": 0.19456507088146183, + "grad_norm": 2.8533244132995605, + "learning_rate": 0.00018192729096012914, + "loss": 0.372, + "step": 8300 + }, + { + "epoch": 0.19458851245626682, + "grad_norm": 3.9739036560058594, + "learning_rate": 0.00018192306695508835, + "loss": 0.4232, + "step": 8301 + }, + { + "epoch": 0.19461195403107182, + "grad_norm": 4.775707244873047, + "learning_rate": 0.0001819188425055307, + "loss": 0.7355, + "step": 8302 + }, + { + "epoch": 0.1946353956058768, + "grad_norm": 5.689848899841309, + "learning_rate": 0.00018191461761147908, + "loss": 0.5985, + "step": 8303 + }, + { + "epoch": 0.1946588371806818, + "grad_norm": 4.118288993835449, + "learning_rate": 0.00018191039227295647, + "loss": 0.4391, + "step": 8304 + }, + { + "epoch": 0.1946822787554868, + "grad_norm": 3.509659767150879, + "learning_rate": 0.00018190616648998571, + "loss": 0.394, + "step": 8305 + }, + { + "epoch": 0.1947057203302918, + "grad_norm": 2.9709949493408203, + "learning_rate": 0.00018190194026258984, + "loss": 0.3279, + "step": 8306 + }, + { + "epoch": 0.1947291619050968, + "grad_norm": 2.7870705127716064, + "learning_rate": 0.00018189771359079172, + "loss": 1.8549, + "step": 8307 + }, + { + "epoch": 0.19475260347990178, + "grad_norm": 4.135568618774414, + "learning_rate": 0.0001818934864746143, + "loss": 0.6435, + "step": 8308 + }, + { + "epoch": 0.19477604505470678, + "grad_norm": 1.3541315793991089, + "learning_rate": 0.00018188925891408054, + "loss": 0.1161, + "step": 8309 + }, + { + "epoch": 0.19479948662951177, + "grad_norm": 3.3042690753936768, + "learning_rate": 0.00018188503090921335, + "loss": 0.4379, + "step": 8310 + }, + { + "epoch": 0.19482292820431676, + "grad_norm": 1.6480633020401, + "learning_rate": 0.00018188080246003566, + "loss": 0.122, + "step": 8311 + }, + { + "epoch": 0.19484636977912176, + "grad_norm": 3.7570345401763916, + "learning_rate": 0.00018187657356657044, + "loss": 0.4404, + "step": 8312 + }, + { + "epoch": 0.19486981135392675, + "grad_norm": 2.715052843093872, + "learning_rate": 0.00018187234422884066, + "loss": 0.3489, + "step": 8313 + }, + { + "epoch": 0.19489325292873175, + "grad_norm": 3.316866636276245, + "learning_rate": 0.0001818681144468692, + "loss": 0.3645, + "step": 8314 + }, + { + "epoch": 0.19491669450353674, + "grad_norm": 4.983136177062988, + "learning_rate": 0.0001818638842206791, + "loss": 0.5786, + "step": 8315 + }, + { + "epoch": 0.19494013607834174, + "grad_norm": 3.2185733318328857, + "learning_rate": 0.0001818596535502932, + "loss": 0.4714, + "step": 8316 + }, + { + "epoch": 0.19496357765314673, + "grad_norm": 2.691375494003296, + "learning_rate": 0.00018185542243573456, + "loss": 0.3857, + "step": 8317 + }, + { + "epoch": 0.19498701922795172, + "grad_norm": 4.643527507781982, + "learning_rate": 0.0001818511908770261, + "loss": 0.6161, + "step": 8318 + }, + { + "epoch": 0.19501046080275672, + "grad_norm": 5.47890043258667, + "learning_rate": 0.0001818469588741907, + "loss": 0.3951, + "step": 8319 + }, + { + "epoch": 0.1950339023775617, + "grad_norm": 3.309525728225708, + "learning_rate": 0.00018184272642725148, + "loss": 0.4393, + "step": 8320 + }, + { + "epoch": 0.1950573439523667, + "grad_norm": 4.601022720336914, + "learning_rate": 0.0001818384935362313, + "loss": 0.4045, + "step": 8321 + }, + { + "epoch": 0.1950807855271717, + "grad_norm": 4.024118423461914, + "learning_rate": 0.00018183426020115318, + "loss": 0.4231, + "step": 8322 + }, + { + "epoch": 0.19510422710197672, + "grad_norm": 0.27819541096687317, + "learning_rate": 0.00018183002642204002, + "loss": 0.0138, + "step": 8323 + }, + { + "epoch": 0.19512766867678172, + "grad_norm": 4.277885437011719, + "learning_rate": 0.00018182579219891488, + "loss": 0.5254, + "step": 8324 + }, + { + "epoch": 0.1951511102515867, + "grad_norm": 2.4436259269714355, + "learning_rate": 0.00018182155753180065, + "loss": 0.2511, + "step": 8325 + }, + { + "epoch": 0.1951745518263917, + "grad_norm": 1.759124994277954, + "learning_rate": 0.00018181732242072038, + "loss": 0.136, + "step": 8326 + }, + { + "epoch": 0.1951979934011967, + "grad_norm": 3.4728007316589355, + "learning_rate": 0.000181813086865697, + "loss": 0.4899, + "step": 8327 + }, + { + "epoch": 0.1952214349760017, + "grad_norm": 1.276572346687317, + "learning_rate": 0.0001818088508667535, + "loss": 0.1318, + "step": 8328 + }, + { + "epoch": 0.1952448765508067, + "grad_norm": 4.690038204193115, + "learning_rate": 0.0001818046144239129, + "loss": 0.6308, + "step": 8329 + }, + { + "epoch": 0.19526831812561168, + "grad_norm": 4.936946392059326, + "learning_rate": 0.00018180037753719815, + "loss": 0.6968, + "step": 8330 + }, + { + "epoch": 0.19529175970041668, + "grad_norm": 1.6208164691925049, + "learning_rate": 0.00018179614020663224, + "loss": 0.2546, + "step": 8331 + }, + { + "epoch": 0.19531520127522167, + "grad_norm": 3.145534038543701, + "learning_rate": 0.0001817919024322382, + "loss": 0.3936, + "step": 8332 + }, + { + "epoch": 0.19533864285002667, + "grad_norm": 2.4470465183258057, + "learning_rate": 0.00018178766421403898, + "loss": 1.832, + "step": 8333 + }, + { + "epoch": 0.19536208442483166, + "grad_norm": 2.0433456897735596, + "learning_rate": 0.00018178342555205757, + "loss": 1.7109, + "step": 8334 + }, + { + "epoch": 0.19538552599963666, + "grad_norm": 3.3389370441436768, + "learning_rate": 0.00018177918644631707, + "loss": 0.3536, + "step": 8335 + }, + { + "epoch": 0.19540896757444165, + "grad_norm": 5.7633209228515625, + "learning_rate": 0.00018177494689684034, + "loss": 0.7989, + "step": 8336 + }, + { + "epoch": 0.19543240914924664, + "grad_norm": 5.354474067687988, + "learning_rate": 0.00018177070690365046, + "loss": 0.4822, + "step": 8337 + }, + { + "epoch": 0.19545585072405164, + "grad_norm": 3.2159838676452637, + "learning_rate": 0.00018176646646677044, + "loss": 0.424, + "step": 8338 + }, + { + "epoch": 0.19547929229885663, + "grad_norm": 0.7194211483001709, + "learning_rate": 0.00018176222558622326, + "loss": 0.058, + "step": 8339 + }, + { + "epoch": 0.19550273387366163, + "grad_norm": 1.0311689376831055, + "learning_rate": 0.00018175798426203193, + "loss": 0.1191, + "step": 8340 + }, + { + "epoch": 0.19552617544846662, + "grad_norm": 1.2953382730484009, + "learning_rate": 0.0001817537424942195, + "loss": 0.0493, + "step": 8341 + }, + { + "epoch": 0.19554961702327162, + "grad_norm": 4.443248271942139, + "learning_rate": 0.00018174950028280898, + "loss": 0.6475, + "step": 8342 + }, + { + "epoch": 0.1955730585980766, + "grad_norm": 1.1794159412384033, + "learning_rate": 0.00018174525762782335, + "loss": 0.1728, + "step": 8343 + }, + { + "epoch": 0.1955965001728816, + "grad_norm": 3.9549880027770996, + "learning_rate": 0.00018174101452928567, + "loss": 0.4925, + "step": 8344 + }, + { + "epoch": 0.1956199417476866, + "grad_norm": 4.045703411102295, + "learning_rate": 0.00018173677098721892, + "loss": 0.464, + "step": 8345 + }, + { + "epoch": 0.1956433833224916, + "grad_norm": 2.504551410675049, + "learning_rate": 0.00018173252700164617, + "loss": 0.2358, + "step": 8346 + }, + { + "epoch": 0.1956668248972966, + "grad_norm": 1.8069416284561157, + "learning_rate": 0.0001817282825725904, + "loss": 0.0981, + "step": 8347 + }, + { + "epoch": 0.1956902664721016, + "grad_norm": 4.720643043518066, + "learning_rate": 0.00018172403770007472, + "loss": 0.399, + "step": 8348 + }, + { + "epoch": 0.1957137080469066, + "grad_norm": 0.4649502635002136, + "learning_rate": 0.0001817197923841221, + "loss": 0.0312, + "step": 8349 + }, + { + "epoch": 0.1957371496217116, + "grad_norm": 5.323171138763428, + "learning_rate": 0.0001817155466247556, + "loss": 0.5343, + "step": 8350 + }, + { + "epoch": 0.1957605911965166, + "grad_norm": 2.4208309650421143, + "learning_rate": 0.00018171130042199821, + "loss": 1.3346, + "step": 8351 + }, + { + "epoch": 0.1957840327713216, + "grad_norm": 4.244905471801758, + "learning_rate": 0.00018170705377587302, + "loss": 0.4501, + "step": 8352 + }, + { + "epoch": 0.19580747434612658, + "grad_norm": 3.554569721221924, + "learning_rate": 0.00018170280668640304, + "loss": 0.363, + "step": 8353 + }, + { + "epoch": 0.19583091592093158, + "grad_norm": 3.9916062355041504, + "learning_rate": 0.00018169855915361137, + "loss": 0.6311, + "step": 8354 + }, + { + "epoch": 0.19585435749573657, + "grad_norm": 1.5335355997085571, + "learning_rate": 0.00018169431117752102, + "loss": 0.2184, + "step": 8355 + }, + { + "epoch": 0.19587779907054156, + "grad_norm": 4.678846836090088, + "learning_rate": 0.00018169006275815504, + "loss": 0.5926, + "step": 8356 + }, + { + "epoch": 0.19590124064534656, + "grad_norm": 2.129997730255127, + "learning_rate": 0.00018168581389553645, + "loss": 1.7039, + "step": 8357 + }, + { + "epoch": 0.19592468222015155, + "grad_norm": 1.4155476093292236, + "learning_rate": 0.00018168156458968837, + "loss": 0.2182, + "step": 8358 + }, + { + "epoch": 0.19594812379495655, + "grad_norm": 3.6847498416900635, + "learning_rate": 0.0001816773148406338, + "loss": 0.424, + "step": 8359 + }, + { + "epoch": 0.19597156536976154, + "grad_norm": 0.8578221201896667, + "learning_rate": 0.00018167306464839583, + "loss": 0.0726, + "step": 8360 + }, + { + "epoch": 0.19599500694456654, + "grad_norm": 2.4241528511047363, + "learning_rate": 0.00018166881401299753, + "loss": 0.1683, + "step": 8361 + }, + { + "epoch": 0.19601844851937153, + "grad_norm": 4.872349262237549, + "learning_rate": 0.00018166456293446196, + "loss": 0.5682, + "step": 8362 + }, + { + "epoch": 0.19604189009417652, + "grad_norm": 3.1468613147735596, + "learning_rate": 0.00018166031141281214, + "loss": 0.3642, + "step": 8363 + }, + { + "epoch": 0.19606533166898152, + "grad_norm": 1.6630185842514038, + "learning_rate": 0.00018165605944807118, + "loss": 0.3266, + "step": 8364 + }, + { + "epoch": 0.1960887732437865, + "grad_norm": 4.189899444580078, + "learning_rate": 0.00018165180704026213, + "loss": 0.5267, + "step": 8365 + }, + { + "epoch": 0.1961122148185915, + "grad_norm": 4.106740474700928, + "learning_rate": 0.00018164755418940813, + "loss": 0.6421, + "step": 8366 + }, + { + "epoch": 0.1961356563933965, + "grad_norm": 5.814898490905762, + "learning_rate": 0.00018164330089553217, + "loss": 0.3413, + "step": 8367 + }, + { + "epoch": 0.1961590979682015, + "grad_norm": 1.5499114990234375, + "learning_rate": 0.00018163904715865736, + "loss": 0.1761, + "step": 8368 + }, + { + "epoch": 0.1961825395430065, + "grad_norm": 3.281024217605591, + "learning_rate": 0.0001816347929788068, + "loss": 0.3866, + "step": 8369 + }, + { + "epoch": 0.19620598111781148, + "grad_norm": 7.213140487670898, + "learning_rate": 0.00018163053835600356, + "loss": 0.7499, + "step": 8370 + }, + { + "epoch": 0.19622942269261648, + "grad_norm": 3.1801817417144775, + "learning_rate": 0.00018162628329027072, + "loss": 0.4908, + "step": 8371 + }, + { + "epoch": 0.19625286426742147, + "grad_norm": 3.706164598464966, + "learning_rate": 0.00018162202778163134, + "loss": 0.3416, + "step": 8372 + }, + { + "epoch": 0.19627630584222647, + "grad_norm": 5.277427673339844, + "learning_rate": 0.0001816177718301086, + "loss": 0.6936, + "step": 8373 + }, + { + "epoch": 0.1962997474170315, + "grad_norm": 1.7497210502624512, + "learning_rate": 0.00018161351543572552, + "loss": 1.1822, + "step": 8374 + }, + { + "epoch": 0.19632318899183648, + "grad_norm": 1.3327113389968872, + "learning_rate": 0.00018160925859850516, + "loss": 0.1281, + "step": 8375 + }, + { + "epoch": 0.19634663056664148, + "grad_norm": 0.9544697999954224, + "learning_rate": 0.00018160500131847074, + "loss": 0.1001, + "step": 8376 + }, + { + "epoch": 0.19637007214144647, + "grad_norm": 3.2684600353240967, + "learning_rate": 0.00018160074359564523, + "loss": 0.2373, + "step": 8377 + }, + { + "epoch": 0.19639351371625147, + "grad_norm": 1.0792109966278076, + "learning_rate": 0.0001815964854300518, + "loss": 0.1436, + "step": 8378 + }, + { + "epoch": 0.19641695529105646, + "grad_norm": 0.5469432473182678, + "learning_rate": 0.00018159222682171357, + "loss": 0.0741, + "step": 8379 + }, + { + "epoch": 0.19644039686586146, + "grad_norm": 4.027271270751953, + "learning_rate": 0.0001815879677706536, + "loss": 0.3474, + "step": 8380 + }, + { + "epoch": 0.19646383844066645, + "grad_norm": 3.2310450077056885, + "learning_rate": 0.00018158370827689506, + "loss": 0.3374, + "step": 8381 + }, + { + "epoch": 0.19648728001547144, + "grad_norm": 2.9734671115875244, + "learning_rate": 0.00018157944834046102, + "loss": 0.1954, + "step": 8382 + }, + { + "epoch": 0.19651072159027644, + "grad_norm": 4.175528526306152, + "learning_rate": 0.00018157518796137456, + "loss": 0.6198, + "step": 8383 + }, + { + "epoch": 0.19653416316508143, + "grad_norm": 3.1879489421844482, + "learning_rate": 0.00018157092713965887, + "loss": 0.376, + "step": 8384 + }, + { + "epoch": 0.19655760473988643, + "grad_norm": 1.0771490335464478, + "learning_rate": 0.00018156666587533702, + "loss": 0.1267, + "step": 8385 + }, + { + "epoch": 0.19658104631469142, + "grad_norm": 3.353818655014038, + "learning_rate": 0.00018156240416843214, + "loss": 0.3521, + "step": 8386 + }, + { + "epoch": 0.19660448788949642, + "grad_norm": 5.624483585357666, + "learning_rate": 0.00018155814201896737, + "loss": 0.5106, + "step": 8387 + }, + { + "epoch": 0.1966279294643014, + "grad_norm": 3.493704080581665, + "learning_rate": 0.00018155387942696583, + "loss": 0.4153, + "step": 8388 + }, + { + "epoch": 0.1966513710391064, + "grad_norm": 1.049575924873352, + "learning_rate": 0.00018154961639245065, + "loss": 0.1097, + "step": 8389 + }, + { + "epoch": 0.1966748126139114, + "grad_norm": 4.544390678405762, + "learning_rate": 0.00018154535291544494, + "loss": 0.624, + "step": 8390 + }, + { + "epoch": 0.1966982541887164, + "grad_norm": 4.16227388381958, + "learning_rate": 0.00018154108899597188, + "loss": 0.3097, + "step": 8391 + }, + { + "epoch": 0.1967216957635214, + "grad_norm": 1.4805206060409546, + "learning_rate": 0.00018153682463405458, + "loss": 0.1996, + "step": 8392 + }, + { + "epoch": 0.19674513733832638, + "grad_norm": 2.2478978633880615, + "learning_rate": 0.00018153255982971615, + "loss": 0.1532, + "step": 8393 + }, + { + "epoch": 0.19676857891313138, + "grad_norm": 4.679776668548584, + "learning_rate": 0.00018152829458297978, + "loss": 0.5503, + "step": 8394 + }, + { + "epoch": 0.19679202048793637, + "grad_norm": 3.555732011795044, + "learning_rate": 0.00018152402889386857, + "loss": 0.3138, + "step": 8395 + }, + { + "epoch": 0.19681546206274136, + "grad_norm": 1.9497718811035156, + "learning_rate": 0.0001815197627624057, + "loss": 0.1754, + "step": 8396 + }, + { + "epoch": 0.19683890363754636, + "grad_norm": 3.2072908878326416, + "learning_rate": 0.0001815154961886143, + "loss": 0.2989, + "step": 8397 + }, + { + "epoch": 0.19686234521235135, + "grad_norm": 3.2783122062683105, + "learning_rate": 0.00018151122917251753, + "loss": 0.5352, + "step": 8398 + }, + { + "epoch": 0.19688578678715637, + "grad_norm": 4.0317206382751465, + "learning_rate": 0.00018150696171413854, + "loss": 0.4605, + "step": 8399 + }, + { + "epoch": 0.19690922836196137, + "grad_norm": 4.221811771392822, + "learning_rate": 0.0001815026938135005, + "loss": 0.4834, + "step": 8400 + }, + { + "epoch": 0.19693266993676636, + "grad_norm": 1.551482081413269, + "learning_rate": 0.00018149842547062652, + "loss": 0.2907, + "step": 8401 + }, + { + "epoch": 0.19695611151157136, + "grad_norm": 2.8093695640563965, + "learning_rate": 0.0001814941566855398, + "loss": 0.263, + "step": 8402 + }, + { + "epoch": 0.19697955308637635, + "grad_norm": 3.3465535640716553, + "learning_rate": 0.00018148988745826352, + "loss": 0.3785, + "step": 8403 + }, + { + "epoch": 0.19700299466118135, + "grad_norm": 1.0773165225982666, + "learning_rate": 0.00018148561778882078, + "loss": 0.1459, + "step": 8404 + }, + { + "epoch": 0.19702643623598634, + "grad_norm": 3.649078369140625, + "learning_rate": 0.00018148134767723481, + "loss": 0.5179, + "step": 8405 + }, + { + "epoch": 0.19704987781079134, + "grad_norm": 4.143102169036865, + "learning_rate": 0.00018147707712352874, + "loss": 0.4992, + "step": 8406 + }, + { + "epoch": 0.19707331938559633, + "grad_norm": 4.830834865570068, + "learning_rate": 0.00018147280612772576, + "loss": 0.5101, + "step": 8407 + }, + { + "epoch": 0.19709676096040132, + "grad_norm": 1.2163265943527222, + "learning_rate": 0.00018146853468984907, + "loss": 0.1394, + "step": 8408 + }, + { + "epoch": 0.19712020253520632, + "grad_norm": 3.1642203330993652, + "learning_rate": 0.00018146426280992182, + "loss": 0.4325, + "step": 8409 + }, + { + "epoch": 0.1971436441100113, + "grad_norm": 5.906159400939941, + "learning_rate": 0.00018145999048796717, + "loss": 0.4076, + "step": 8410 + }, + { + "epoch": 0.1971670856848163, + "grad_norm": 3.8811967372894287, + "learning_rate": 0.00018145571772400833, + "loss": 0.3987, + "step": 8411 + }, + { + "epoch": 0.1971905272596213, + "grad_norm": 0.9233996272087097, + "learning_rate": 0.0001814514445180685, + "loss": 0.1162, + "step": 8412 + }, + { + "epoch": 0.1972139688344263, + "grad_norm": 2.0165271759033203, + "learning_rate": 0.00018144717087017077, + "loss": 1.6143, + "step": 8413 + }, + { + "epoch": 0.1972374104092313, + "grad_norm": 0.72254878282547, + "learning_rate": 0.00018144289678033847, + "loss": 0.11, + "step": 8414 + }, + { + "epoch": 0.19726085198403628, + "grad_norm": 0.4534737765789032, + "learning_rate": 0.00018143862224859468, + "loss": 0.0549, + "step": 8415 + }, + { + "epoch": 0.19728429355884128, + "grad_norm": 1.2309130430221558, + "learning_rate": 0.00018143434727496265, + "loss": 0.154, + "step": 8416 + }, + { + "epoch": 0.19730773513364627, + "grad_norm": 2.79465913772583, + "learning_rate": 0.00018143007185946556, + "loss": 0.3878, + "step": 8417 + }, + { + "epoch": 0.19733117670845127, + "grad_norm": 1.1748743057250977, + "learning_rate": 0.0001814257960021266, + "loss": 0.1297, + "step": 8418 + }, + { + "epoch": 0.19735461828325626, + "grad_norm": 2.449286699295044, + "learning_rate": 0.00018142151970296902, + "loss": 1.158, + "step": 8419 + }, + { + "epoch": 0.19737805985806126, + "grad_norm": 1.3724969625473022, + "learning_rate": 0.00018141724296201595, + "loss": 0.1713, + "step": 8420 + }, + { + "epoch": 0.19740150143286625, + "grad_norm": 1.6929051876068115, + "learning_rate": 0.00018141296577929062, + "loss": 0.1391, + "step": 8421 + }, + { + "epoch": 0.19742494300767124, + "grad_norm": 5.045629501342773, + "learning_rate": 0.0001814086881548163, + "loss": 0.6523, + "step": 8422 + }, + { + "epoch": 0.19744838458247624, + "grad_norm": 2.53692889213562, + "learning_rate": 0.00018140441008861611, + "loss": 1.4425, + "step": 8423 + }, + { + "epoch": 0.19747182615728123, + "grad_norm": 2.65120530128479, + "learning_rate": 0.00018140013158071333, + "loss": 0.2919, + "step": 8424 + }, + { + "epoch": 0.19749526773208625, + "grad_norm": 2.6220297813415527, + "learning_rate": 0.0001813958526311311, + "loss": 1.5471, + "step": 8425 + }, + { + "epoch": 0.19751870930689125, + "grad_norm": 2.1391468048095703, + "learning_rate": 0.0001813915732398927, + "loss": 0.2565, + "step": 8426 + }, + { + "epoch": 0.19754215088169624, + "grad_norm": 1.3481000661849976, + "learning_rate": 0.00018138729340702137, + "loss": 0.1338, + "step": 8427 + }, + { + "epoch": 0.19756559245650124, + "grad_norm": 2.774489402770996, + "learning_rate": 0.00018138301313254024, + "loss": 0.3602, + "step": 8428 + }, + { + "epoch": 0.19758903403130623, + "grad_norm": 3.0193092823028564, + "learning_rate": 0.00018137873241647266, + "loss": 0.5204, + "step": 8429 + }, + { + "epoch": 0.19761247560611123, + "grad_norm": 3.0750789642333984, + "learning_rate": 0.00018137445125884172, + "loss": 0.4023, + "step": 8430 + }, + { + "epoch": 0.19763591718091622, + "grad_norm": 3.6247386932373047, + "learning_rate": 0.00018137016965967079, + "loss": 0.5302, + "step": 8431 + }, + { + "epoch": 0.19765935875572122, + "grad_norm": 2.3355178833007812, + "learning_rate": 0.000181365887618983, + "loss": 1.6665, + "step": 8432 + }, + { + "epoch": 0.1976828003305262, + "grad_norm": 4.124922275543213, + "learning_rate": 0.00018136160513680162, + "loss": 0.3186, + "step": 8433 + }, + { + "epoch": 0.1977062419053312, + "grad_norm": 2.455526113510132, + "learning_rate": 0.00018135732221314984, + "loss": 0.1757, + "step": 8434 + }, + { + "epoch": 0.1977296834801362, + "grad_norm": 3.46744704246521, + "learning_rate": 0.000181353038848051, + "loss": 0.498, + "step": 8435 + }, + { + "epoch": 0.1977531250549412, + "grad_norm": 3.0897045135498047, + "learning_rate": 0.00018134875504152825, + "loss": 0.2609, + "step": 8436 + }, + { + "epoch": 0.1977765666297462, + "grad_norm": 1.4402042627334595, + "learning_rate": 0.0001813444707936049, + "loss": 0.7489, + "step": 8437 + }, + { + "epoch": 0.19780000820455118, + "grad_norm": 5.697366714477539, + "learning_rate": 0.00018134018610430414, + "loss": 0.6488, + "step": 8438 + }, + { + "epoch": 0.19782344977935618, + "grad_norm": 4.301728248596191, + "learning_rate": 0.00018133590097364923, + "loss": 0.4442, + "step": 8439 + }, + { + "epoch": 0.19784689135416117, + "grad_norm": 3.658968448638916, + "learning_rate": 0.00018133161540166346, + "loss": 0.4151, + "step": 8440 + }, + { + "epoch": 0.19787033292896616, + "grad_norm": 4.224647045135498, + "learning_rate": 0.00018132732938837008, + "loss": 0.2839, + "step": 8441 + }, + { + "epoch": 0.19789377450377116, + "grad_norm": 3.539133071899414, + "learning_rate": 0.00018132304293379226, + "loss": 0.2873, + "step": 8442 + }, + { + "epoch": 0.19791721607857615, + "grad_norm": 4.124135971069336, + "learning_rate": 0.00018131875603795335, + "loss": 0.514, + "step": 8443 + }, + { + "epoch": 0.19794065765338115, + "grad_norm": 0.8741663694381714, + "learning_rate": 0.0001813144687008766, + "loss": 0.0663, + "step": 8444 + }, + { + "epoch": 0.19796409922818614, + "grad_norm": 2.0937016010284424, + "learning_rate": 0.00018131018092258525, + "loss": 0.2824, + "step": 8445 + }, + { + "epoch": 0.19798754080299114, + "grad_norm": 3.2653121948242188, + "learning_rate": 0.00018130589270310256, + "loss": 1.6247, + "step": 8446 + }, + { + "epoch": 0.19801098237779613, + "grad_norm": 1.4409840106964111, + "learning_rate": 0.00018130160404245183, + "loss": 0.0701, + "step": 8447 + }, + { + "epoch": 0.19803442395260112, + "grad_norm": 2.3982224464416504, + "learning_rate": 0.0001812973149406563, + "loss": 0.2491, + "step": 8448 + }, + { + "epoch": 0.19805786552740612, + "grad_norm": 2.841919422149658, + "learning_rate": 0.00018129302539773924, + "loss": 0.3681, + "step": 8449 + }, + { + "epoch": 0.1980813071022111, + "grad_norm": 1.5836714506149292, + "learning_rate": 0.00018128873541372397, + "loss": 0.2548, + "step": 8450 + }, + { + "epoch": 0.19810474867701613, + "grad_norm": 4.3248701095581055, + "learning_rate": 0.0001812844449886337, + "loss": 0.5974, + "step": 8451 + }, + { + "epoch": 0.19812819025182113, + "grad_norm": 2.923739194869995, + "learning_rate": 0.00018128015412249179, + "loss": 0.3535, + "step": 8452 + }, + { + "epoch": 0.19815163182662612, + "grad_norm": 1.0761473178863525, + "learning_rate": 0.00018127586281532142, + "loss": 0.0644, + "step": 8453 + }, + { + "epoch": 0.19817507340143112, + "grad_norm": 3.316239356994629, + "learning_rate": 0.00018127157106714598, + "loss": 1.2783, + "step": 8454 + }, + { + "epoch": 0.1981985149762361, + "grad_norm": 3.4570839405059814, + "learning_rate": 0.0001812672788779887, + "loss": 0.2576, + "step": 8455 + }, + { + "epoch": 0.1982219565510411, + "grad_norm": 4.186557769775391, + "learning_rate": 0.00018126298624787288, + "loss": 0.3582, + "step": 8456 + }, + { + "epoch": 0.1982453981258461, + "grad_norm": 4.620722770690918, + "learning_rate": 0.00018125869317682177, + "loss": 0.7524, + "step": 8457 + }, + { + "epoch": 0.1982688397006511, + "grad_norm": 2.850470781326294, + "learning_rate": 0.00018125439966485877, + "loss": 0.4534, + "step": 8458 + }, + { + "epoch": 0.1982922812754561, + "grad_norm": 4.237627983093262, + "learning_rate": 0.00018125010571200706, + "loss": 0.4389, + "step": 8459 + }, + { + "epoch": 0.19831572285026108, + "grad_norm": 4.794408321380615, + "learning_rate": 0.00018124581131829005, + "loss": 0.6782, + "step": 8460 + }, + { + "epoch": 0.19833916442506608, + "grad_norm": 2.8954384326934814, + "learning_rate": 0.00018124151648373093, + "loss": 1.5291, + "step": 8461 + }, + { + "epoch": 0.19836260599987107, + "grad_norm": 3.0285439491271973, + "learning_rate": 0.00018123722120835307, + "loss": 0.3467, + "step": 8462 + }, + { + "epoch": 0.19838604757467607, + "grad_norm": 5.564121246337891, + "learning_rate": 0.00018123292549217975, + "loss": 0.9423, + "step": 8463 + }, + { + "epoch": 0.19840948914948106, + "grad_norm": 4.0979180335998535, + "learning_rate": 0.00018122862933523434, + "loss": 0.4078, + "step": 8464 + }, + { + "epoch": 0.19843293072428606, + "grad_norm": 4.518852233886719, + "learning_rate": 0.00018122433273754006, + "loss": 0.5841, + "step": 8465 + }, + { + "epoch": 0.19845637229909105, + "grad_norm": 1.5209566354751587, + "learning_rate": 0.00018122003569912029, + "loss": 0.1691, + "step": 8466 + }, + { + "epoch": 0.19847981387389604, + "grad_norm": 5.376167297363281, + "learning_rate": 0.00018121573821999832, + "loss": 0.7789, + "step": 8467 + }, + { + "epoch": 0.19850325544870104, + "grad_norm": 5.503620147705078, + "learning_rate": 0.00018121144030019748, + "loss": 0.535, + "step": 8468 + }, + { + "epoch": 0.19852669702350603, + "grad_norm": 4.148711681365967, + "learning_rate": 0.00018120714193974106, + "loss": 0.4633, + "step": 8469 + }, + { + "epoch": 0.19855013859831103, + "grad_norm": 2.230454444885254, + "learning_rate": 0.00018120284313865242, + "loss": 0.1171, + "step": 8470 + }, + { + "epoch": 0.19857358017311602, + "grad_norm": 3.966909408569336, + "learning_rate": 0.00018119854389695483, + "loss": 0.4406, + "step": 8471 + }, + { + "epoch": 0.19859702174792102, + "grad_norm": 4.663771152496338, + "learning_rate": 0.0001811942442146717, + "loss": 0.3746, + "step": 8472 + }, + { + "epoch": 0.198620463322726, + "grad_norm": 3.0428664684295654, + "learning_rate": 0.0001811899440918263, + "loss": 2.0215, + "step": 8473 + }, + { + "epoch": 0.198643904897531, + "grad_norm": 2.4268290996551514, + "learning_rate": 0.00018118564352844198, + "loss": 1.4276, + "step": 8474 + }, + { + "epoch": 0.198667346472336, + "grad_norm": 1.8015204668045044, + "learning_rate": 0.0001811813425245421, + "loss": 0.238, + "step": 8475 + }, + { + "epoch": 0.19869078804714102, + "grad_norm": 1.3928383588790894, + "learning_rate": 0.00018117704108014997, + "loss": 0.2262, + "step": 8476 + }, + { + "epoch": 0.19871422962194601, + "grad_norm": 3.0012476444244385, + "learning_rate": 0.00018117273919528892, + "loss": 1.834, + "step": 8477 + }, + { + "epoch": 0.198737671196751, + "grad_norm": 5.3822221755981445, + "learning_rate": 0.00018116843686998227, + "loss": 0.6247, + "step": 8478 + }, + { + "epoch": 0.198761112771556, + "grad_norm": 4.215364456176758, + "learning_rate": 0.00018116413410425346, + "loss": 0.5642, + "step": 8479 + }, + { + "epoch": 0.198784554346361, + "grad_norm": 4.258223533630371, + "learning_rate": 0.00018115983089812574, + "loss": 0.5456, + "step": 8480 + }, + { + "epoch": 0.198807995921166, + "grad_norm": 2.9005625247955322, + "learning_rate": 0.0001811555272516225, + "loss": 0.244, + "step": 8481 + }, + { + "epoch": 0.198831437495971, + "grad_norm": 7.953525066375732, + "learning_rate": 0.0001811512231647671, + "loss": 0.5483, + "step": 8482 + }, + { + "epoch": 0.19885487907077598, + "grad_norm": 2.936493396759033, + "learning_rate": 0.0001811469186375829, + "loss": 0.3341, + "step": 8483 + }, + { + "epoch": 0.19887832064558097, + "grad_norm": 1.8522372245788574, + "learning_rate": 0.0001811426136700932, + "loss": 0.1532, + "step": 8484 + }, + { + "epoch": 0.19890176222038597, + "grad_norm": 3.662040948867798, + "learning_rate": 0.0001811383082623214, + "loss": 0.5104, + "step": 8485 + }, + { + "epoch": 0.19892520379519096, + "grad_norm": 2.313901424407959, + "learning_rate": 0.00018113400241429086, + "loss": 0.2504, + "step": 8486 + }, + { + "epoch": 0.19894864536999596, + "grad_norm": 3.856459617614746, + "learning_rate": 0.00018112969612602494, + "loss": 0.4619, + "step": 8487 + }, + { + "epoch": 0.19897208694480095, + "grad_norm": 4.924591541290283, + "learning_rate": 0.00018112538939754703, + "loss": 0.9701, + "step": 8488 + }, + { + "epoch": 0.19899552851960595, + "grad_norm": 2.7012507915496826, + "learning_rate": 0.00018112108222888046, + "loss": 0.2758, + "step": 8489 + }, + { + "epoch": 0.19901897009441094, + "grad_norm": 2.269728183746338, + "learning_rate": 0.00018111677462004862, + "loss": 0.2443, + "step": 8490 + }, + { + "epoch": 0.19904241166921594, + "grad_norm": 3.551684856414795, + "learning_rate": 0.0001811124665710749, + "loss": 0.4601, + "step": 8491 + }, + { + "epoch": 0.19906585324402093, + "grad_norm": 1.7922632694244385, + "learning_rate": 0.0001811081580819826, + "loss": 0.1263, + "step": 8492 + }, + { + "epoch": 0.19908929481882592, + "grad_norm": 2.716282367706299, + "learning_rate": 0.0001811038491527952, + "loss": 0.292, + "step": 8493 + }, + { + "epoch": 0.19911273639363092, + "grad_norm": 3.6688313484191895, + "learning_rate": 0.00018109953978353603, + "loss": 0.4146, + "step": 8494 + }, + { + "epoch": 0.1991361779684359, + "grad_norm": 0.9224792718887329, + "learning_rate": 0.00018109522997422842, + "loss": 0.1311, + "step": 8495 + }, + { + "epoch": 0.1991596195432409, + "grad_norm": 3.1066558361053467, + "learning_rate": 0.0001810909197248959, + "loss": 0.4709, + "step": 8496 + }, + { + "epoch": 0.1991830611180459, + "grad_norm": 4.8794474601745605, + "learning_rate": 0.0001810866090355617, + "loss": 0.3772, + "step": 8497 + }, + { + "epoch": 0.1992065026928509, + "grad_norm": 5.349462032318115, + "learning_rate": 0.00018108229790624934, + "loss": 0.7907, + "step": 8498 + }, + { + "epoch": 0.1992299442676559, + "grad_norm": 4.226968765258789, + "learning_rate": 0.00018107798633698212, + "loss": 0.4959, + "step": 8499 + }, + { + "epoch": 0.19925338584246088, + "grad_norm": 3.533554792404175, + "learning_rate": 0.00018107367432778346, + "loss": 0.2543, + "step": 8500 + }, + { + "epoch": 0.19927682741726588, + "grad_norm": 2.3366506099700928, + "learning_rate": 0.00018106936187867677, + "loss": 0.4179, + "step": 8501 + }, + { + "epoch": 0.1993002689920709, + "grad_norm": 2.914031982421875, + "learning_rate": 0.00018106504898968547, + "loss": 0.4605, + "step": 8502 + }, + { + "epoch": 0.1993237105668759, + "grad_norm": 2.2662503719329834, + "learning_rate": 0.00018106073566083287, + "loss": 0.2707, + "step": 8503 + }, + { + "epoch": 0.1993471521416809, + "grad_norm": 3.4070563316345215, + "learning_rate": 0.0001810564218921425, + "loss": 0.5513, + "step": 8504 + }, + { + "epoch": 0.19937059371648588, + "grad_norm": 3.034191131591797, + "learning_rate": 0.00018105210768363768, + "loss": 1.9688, + "step": 8505 + }, + { + "epoch": 0.19939403529129088, + "grad_norm": 2.8660054206848145, + "learning_rate": 0.00018104779303534184, + "loss": 0.3498, + "step": 8506 + }, + { + "epoch": 0.19941747686609587, + "grad_norm": 3.461639881134033, + "learning_rate": 0.00018104347794727841, + "loss": 0.3179, + "step": 8507 + }, + { + "epoch": 0.19944091844090087, + "grad_norm": 0.8827561140060425, + "learning_rate": 0.00018103916241947078, + "loss": 0.1461, + "step": 8508 + }, + { + "epoch": 0.19946436001570586, + "grad_norm": 3.4200332164764404, + "learning_rate": 0.00018103484645194237, + "loss": 0.6148, + "step": 8509 + }, + { + "epoch": 0.19948780159051085, + "grad_norm": 3.072753667831421, + "learning_rate": 0.0001810305300447166, + "loss": 0.4487, + "step": 8510 + }, + { + "epoch": 0.19951124316531585, + "grad_norm": 3.874542236328125, + "learning_rate": 0.00018102621319781692, + "loss": 0.4613, + "step": 8511 + }, + { + "epoch": 0.19953468474012084, + "grad_norm": 5.884461402893066, + "learning_rate": 0.00018102189591126674, + "loss": 0.6568, + "step": 8512 + }, + { + "epoch": 0.19955812631492584, + "grad_norm": 5.446394443511963, + "learning_rate": 0.00018101757818508944, + "loss": 0.6782, + "step": 8513 + }, + { + "epoch": 0.19958156788973083, + "grad_norm": 1.8680925369262695, + "learning_rate": 0.0001810132600193085, + "loss": 0.1624, + "step": 8514 + }, + { + "epoch": 0.19960500946453583, + "grad_norm": 2.6510448455810547, + "learning_rate": 0.00018100894141394734, + "loss": 0.279, + "step": 8515 + }, + { + "epoch": 0.19962845103934082, + "grad_norm": 2.6539368629455566, + "learning_rate": 0.00018100462236902938, + "loss": 1.9476, + "step": 8516 + }, + { + "epoch": 0.19965189261414582, + "grad_norm": 4.817783832550049, + "learning_rate": 0.00018100030288457805, + "loss": 0.4348, + "step": 8517 + }, + { + "epoch": 0.1996753341889508, + "grad_norm": 3.9882776737213135, + "learning_rate": 0.0001809959829606168, + "loss": 0.3777, + "step": 8518 + }, + { + "epoch": 0.1996987757637558, + "grad_norm": 2.8009493350982666, + "learning_rate": 0.00018099166259716908, + "loss": 0.2979, + "step": 8519 + }, + { + "epoch": 0.1997222173385608, + "grad_norm": 1.0306977033615112, + "learning_rate": 0.0001809873417942583, + "loss": 0.1893, + "step": 8520 + }, + { + "epoch": 0.1997456589133658, + "grad_norm": 5.019053936004639, + "learning_rate": 0.00018098302055190798, + "loss": 0.5246, + "step": 8521 + }, + { + "epoch": 0.1997691004881708, + "grad_norm": 0.5866532921791077, + "learning_rate": 0.00018097869887014146, + "loss": 0.047, + "step": 8522 + }, + { + "epoch": 0.19979254206297578, + "grad_norm": 0.37968024611473083, + "learning_rate": 0.00018097437674898225, + "loss": 0.021, + "step": 8523 + }, + { + "epoch": 0.19981598363778078, + "grad_norm": 4.8523640632629395, + "learning_rate": 0.0001809700541884538, + "loss": 0.4619, + "step": 8524 + }, + { + "epoch": 0.19983942521258577, + "grad_norm": 3.604440689086914, + "learning_rate": 0.00018096573118857955, + "loss": 0.2827, + "step": 8525 + }, + { + "epoch": 0.19986286678739076, + "grad_norm": 3.0357396602630615, + "learning_rate": 0.000180961407749383, + "loss": 1.5849, + "step": 8526 + }, + { + "epoch": 0.19988630836219579, + "grad_norm": 4.901516914367676, + "learning_rate": 0.00018095708387088757, + "loss": 0.6217, + "step": 8527 + }, + { + "epoch": 0.19990974993700078, + "grad_norm": 6.5516557693481445, + "learning_rate": 0.0001809527595531167, + "loss": 1.1058, + "step": 8528 + }, + { + "epoch": 0.19993319151180577, + "grad_norm": 2.3407957553863525, + "learning_rate": 0.0001809484347960939, + "loss": 0.2513, + "step": 8529 + }, + { + "epoch": 0.19995663308661077, + "grad_norm": 0.5711522102355957, + "learning_rate": 0.0001809441095998426, + "loss": 0.0599, + "step": 8530 + }, + { + "epoch": 0.19998007466141576, + "grad_norm": 3.650317907333374, + "learning_rate": 0.0001809397839643863, + "loss": 0.7192, + "step": 8531 + }, + { + "epoch": 0.20000351623622076, + "grad_norm": 3.6959147453308105, + "learning_rate": 0.00018093545788974846, + "loss": 0.4558, + "step": 8532 + }, + { + "epoch": 0.20002695781102575, + "grad_norm": 3.5132133960723877, + "learning_rate": 0.00018093113137595252, + "loss": 0.5113, + "step": 8533 + }, + { + "epoch": 0.20005039938583075, + "grad_norm": 2.499417781829834, + "learning_rate": 0.00018092680442302202, + "loss": 0.3097, + "step": 8534 + }, + { + "epoch": 0.20007384096063574, + "grad_norm": 0.42315196990966797, + "learning_rate": 0.0001809224770309804, + "loss": 0.0578, + "step": 8535 + }, + { + "epoch": 0.20009728253544073, + "grad_norm": 2.952342987060547, + "learning_rate": 0.00018091814919985111, + "loss": 0.4626, + "step": 8536 + }, + { + "epoch": 0.20012072411024573, + "grad_norm": 3.0822362899780273, + "learning_rate": 0.0001809138209296577, + "loss": 0.3659, + "step": 8537 + }, + { + "epoch": 0.20014416568505072, + "grad_norm": 3.9131507873535156, + "learning_rate": 0.0001809094922204236, + "loss": 0.4524, + "step": 8538 + }, + { + "epoch": 0.20016760725985572, + "grad_norm": 6.787663459777832, + "learning_rate": 0.00018090516307217236, + "loss": 0.6991, + "step": 8539 + }, + { + "epoch": 0.2001910488346607, + "grad_norm": 0.7405194640159607, + "learning_rate": 0.0001809008334849274, + "loss": 0.0656, + "step": 8540 + }, + { + "epoch": 0.2002144904094657, + "grad_norm": 1.1310961246490479, + "learning_rate": 0.00018089650345871222, + "loss": 0.125, + "step": 8541 + }, + { + "epoch": 0.2002379319842707, + "grad_norm": 4.445096492767334, + "learning_rate": 0.00018089217299355038, + "loss": 0.6313, + "step": 8542 + }, + { + "epoch": 0.2002613735590757, + "grad_norm": 1.0967686176300049, + "learning_rate": 0.00018088784208946534, + "loss": 0.1023, + "step": 8543 + }, + { + "epoch": 0.2002848151338807, + "grad_norm": 3.0183181762695312, + "learning_rate": 0.00018088351074648055, + "loss": 0.2139, + "step": 8544 + }, + { + "epoch": 0.20030825670868568, + "grad_norm": 4.042293548583984, + "learning_rate": 0.0001808791789646196, + "loss": 0.306, + "step": 8545 + }, + { + "epoch": 0.20033169828349068, + "grad_norm": 1.0509521961212158, + "learning_rate": 0.00018087484674390592, + "loss": 0.1963, + "step": 8546 + }, + { + "epoch": 0.20035513985829567, + "grad_norm": 3.507862091064453, + "learning_rate": 0.00018087051408436307, + "loss": 0.4196, + "step": 8547 + }, + { + "epoch": 0.20037858143310067, + "grad_norm": 1.6442135572433472, + "learning_rate": 0.00018086618098601452, + "loss": 1.1017, + "step": 8548 + }, + { + "epoch": 0.20040202300790566, + "grad_norm": 5.407462120056152, + "learning_rate": 0.00018086184744888381, + "loss": 0.8361, + "step": 8549 + }, + { + "epoch": 0.20042546458271066, + "grad_norm": 6.331055641174316, + "learning_rate": 0.0001808575134729944, + "loss": 0.5183, + "step": 8550 + }, + { + "epoch": 0.20044890615751565, + "grad_norm": 3.938673973083496, + "learning_rate": 0.0001808531790583699, + "loss": 0.6597, + "step": 8551 + }, + { + "epoch": 0.20047234773232064, + "grad_norm": 1.6371556520462036, + "learning_rate": 0.0001808488442050338, + "loss": 0.2844, + "step": 8552 + }, + { + "epoch": 0.20049578930712567, + "grad_norm": 5.380971908569336, + "learning_rate": 0.00018084450891300952, + "loss": 0.4964, + "step": 8553 + }, + { + "epoch": 0.20051923088193066, + "grad_norm": 2.565505266189575, + "learning_rate": 0.0001808401731823207, + "loss": 0.3775, + "step": 8554 + }, + { + "epoch": 0.20054267245673565, + "grad_norm": 3.6494579315185547, + "learning_rate": 0.00018083583701299083, + "loss": 0.2675, + "step": 8555 + }, + { + "epoch": 0.20056611403154065, + "grad_norm": 4.634739875793457, + "learning_rate": 0.00018083150040504344, + "loss": 0.5832, + "step": 8556 + }, + { + "epoch": 0.20058955560634564, + "grad_norm": 3.52504563331604, + "learning_rate": 0.00018082716335850204, + "loss": 1.1669, + "step": 8557 + }, + { + "epoch": 0.20061299718115064, + "grad_norm": 6.933376312255859, + "learning_rate": 0.0001808228258733902, + "loss": 1.3313, + "step": 8558 + }, + { + "epoch": 0.20063643875595563, + "grad_norm": 2.357759952545166, + "learning_rate": 0.00018081848794973144, + "loss": 0.3012, + "step": 8559 + }, + { + "epoch": 0.20065988033076063, + "grad_norm": 3.356494665145874, + "learning_rate": 0.00018081414958754928, + "loss": 0.3773, + "step": 8560 + }, + { + "epoch": 0.20068332190556562, + "grad_norm": 2.6618969440460205, + "learning_rate": 0.00018080981078686726, + "loss": 0.291, + "step": 8561 + }, + { + "epoch": 0.20070676348037061, + "grad_norm": 3.1751210689544678, + "learning_rate": 0.00018080547154770892, + "loss": 0.4052, + "step": 8562 + }, + { + "epoch": 0.2007302050551756, + "grad_norm": 1.0064936876296997, + "learning_rate": 0.00018080113187009787, + "loss": 0.1411, + "step": 8563 + }, + { + "epoch": 0.2007536466299806, + "grad_norm": 3.556878089904785, + "learning_rate": 0.00018079679175405757, + "loss": 0.329, + "step": 8564 + }, + { + "epoch": 0.2007770882047856, + "grad_norm": 2.8958899974823, + "learning_rate": 0.00018079245119961162, + "loss": 0.26, + "step": 8565 + }, + { + "epoch": 0.2008005297795906, + "grad_norm": 2.075852632522583, + "learning_rate": 0.00018078811020678354, + "loss": 0.2625, + "step": 8566 + }, + { + "epoch": 0.2008239713543956, + "grad_norm": 3.8716654777526855, + "learning_rate": 0.00018078376877559693, + "loss": 0.6357, + "step": 8567 + }, + { + "epoch": 0.20084741292920058, + "grad_norm": 3.9629416465759277, + "learning_rate": 0.00018077942690607527, + "loss": 0.7084, + "step": 8568 + }, + { + "epoch": 0.20087085450400557, + "grad_norm": 4.069786548614502, + "learning_rate": 0.0001807750845982422, + "loss": 0.4391, + "step": 8569 + }, + { + "epoch": 0.20089429607881057, + "grad_norm": 4.32474946975708, + "learning_rate": 0.00018077074185212126, + "loss": 0.6865, + "step": 8570 + }, + { + "epoch": 0.20091773765361556, + "grad_norm": 1.007210373878479, + "learning_rate": 0.000180766398667736, + "loss": 0.1195, + "step": 8571 + }, + { + "epoch": 0.20094117922842056, + "grad_norm": 2.7691545486450195, + "learning_rate": 0.00018076205504511, + "loss": 0.2557, + "step": 8572 + }, + { + "epoch": 0.20096462080322555, + "grad_norm": 3.5143120288848877, + "learning_rate": 0.0001807577109842668, + "loss": 0.5532, + "step": 8573 + }, + { + "epoch": 0.20098806237803055, + "grad_norm": 1.5306414365768433, + "learning_rate": 0.00018075336648523, + "loss": 0.2882, + "step": 8574 + }, + { + "epoch": 0.20101150395283554, + "grad_norm": 2.8647119998931885, + "learning_rate": 0.00018074902154802315, + "loss": 0.4052, + "step": 8575 + }, + { + "epoch": 0.20103494552764054, + "grad_norm": 4.215633392333984, + "learning_rate": 0.00018074467617266985, + "loss": 0.562, + "step": 8576 + }, + { + "epoch": 0.20105838710244553, + "grad_norm": 0.7165574431419373, + "learning_rate": 0.00018074033035919365, + "loss": 0.0546, + "step": 8577 + }, + { + "epoch": 0.20108182867725055, + "grad_norm": 1.7002066373825073, + "learning_rate": 0.00018073598410761816, + "loss": 0.1703, + "step": 8578 + }, + { + "epoch": 0.20110527025205555, + "grad_norm": 3.4542131423950195, + "learning_rate": 0.00018073163741796693, + "loss": 0.3542, + "step": 8579 + }, + { + "epoch": 0.20112871182686054, + "grad_norm": 2.7049596309661865, + "learning_rate": 0.0001807272902902636, + "loss": 0.3263, + "step": 8580 + }, + { + "epoch": 0.20115215340166553, + "grad_norm": 1.2498279809951782, + "learning_rate": 0.0001807229427245317, + "loss": 0.1638, + "step": 8581 + }, + { + "epoch": 0.20117559497647053, + "grad_norm": 4.141992092132568, + "learning_rate": 0.00018071859472079487, + "loss": 0.7952, + "step": 8582 + }, + { + "epoch": 0.20119903655127552, + "grad_norm": 3.4248924255371094, + "learning_rate": 0.00018071424627907665, + "loss": 0.36, + "step": 8583 + }, + { + "epoch": 0.20122247812608052, + "grad_norm": 1.4214057922363281, + "learning_rate": 0.00018070989739940065, + "loss": 0.1469, + "step": 8584 + }, + { + "epoch": 0.2012459197008855, + "grad_norm": 6.773070812225342, + "learning_rate": 0.00018070554808179048, + "loss": 0.6458, + "step": 8585 + }, + { + "epoch": 0.2012693612756905, + "grad_norm": 0.8048047423362732, + "learning_rate": 0.00018070119832626978, + "loss": 0.0759, + "step": 8586 + }, + { + "epoch": 0.2012928028504955, + "grad_norm": 6.621984004974365, + "learning_rate": 0.00018069684813286206, + "loss": 0.7591, + "step": 8587 + }, + { + "epoch": 0.2013162444253005, + "grad_norm": 0.9569252133369446, + "learning_rate": 0.000180692497501591, + "loss": 0.1122, + "step": 8588 + }, + { + "epoch": 0.2013396860001055, + "grad_norm": 3.3918027877807617, + "learning_rate": 0.00018068814643248016, + "loss": 0.285, + "step": 8589 + }, + { + "epoch": 0.20136312757491048, + "grad_norm": 3.2031097412109375, + "learning_rate": 0.00018068379492555318, + "loss": 1.5641, + "step": 8590 + }, + { + "epoch": 0.20138656914971548, + "grad_norm": 3.369513988494873, + "learning_rate": 0.00018067944298083365, + "loss": 0.5555, + "step": 8591 + }, + { + "epoch": 0.20141001072452047, + "grad_norm": 3.57370924949646, + "learning_rate": 0.00018067509059834522, + "loss": 0.4754, + "step": 8592 + }, + { + "epoch": 0.20143345229932547, + "grad_norm": 4.449038028717041, + "learning_rate": 0.0001806707377781115, + "loss": 0.6581, + "step": 8593 + }, + { + "epoch": 0.20145689387413046, + "grad_norm": 5.672210216522217, + "learning_rate": 0.00018066638452015604, + "loss": 0.7642, + "step": 8594 + }, + { + "epoch": 0.20148033544893545, + "grad_norm": 1.8071132898330688, + "learning_rate": 0.00018066203082450248, + "loss": 1.2938, + "step": 8595 + }, + { + "epoch": 0.20150377702374045, + "grad_norm": 4.707385063171387, + "learning_rate": 0.00018065767669117453, + "loss": 0.6492, + "step": 8596 + }, + { + "epoch": 0.20152721859854544, + "grad_norm": 4.276747703552246, + "learning_rate": 0.00018065332212019574, + "loss": 0.4974, + "step": 8597 + }, + { + "epoch": 0.20155066017335044, + "grad_norm": 1.7520917654037476, + "learning_rate": 0.00018064896711158975, + "loss": 0.3479, + "step": 8598 + }, + { + "epoch": 0.20157410174815543, + "grad_norm": 5.408090114593506, + "learning_rate": 0.00018064461166538022, + "loss": 0.472, + "step": 8599 + }, + { + "epoch": 0.20159754332296043, + "grad_norm": 1.6161692142486572, + "learning_rate": 0.00018064025578159076, + "loss": 0.22, + "step": 8600 + }, + { + "epoch": 0.20162098489776542, + "grad_norm": 2.4248361587524414, + "learning_rate": 0.00018063589946024498, + "loss": 0.1635, + "step": 8601 + }, + { + "epoch": 0.20164442647257041, + "grad_norm": 3.658451795578003, + "learning_rate": 0.00018063154270136657, + "loss": 0.1796, + "step": 8602 + }, + { + "epoch": 0.2016678680473754, + "grad_norm": 4.055794715881348, + "learning_rate": 0.00018062718550497913, + "loss": 0.6775, + "step": 8603 + }, + { + "epoch": 0.20169130962218043, + "grad_norm": 3.7566261291503906, + "learning_rate": 0.0001806228278711063, + "loss": 0.4893, + "step": 8604 + }, + { + "epoch": 0.20171475119698543, + "grad_norm": 4.9247918128967285, + "learning_rate": 0.00018061846979977174, + "loss": 0.6686, + "step": 8605 + }, + { + "epoch": 0.20173819277179042, + "grad_norm": 0.6711252927780151, + "learning_rate": 0.0001806141112909991, + "loss": 0.0827, + "step": 8606 + }, + { + "epoch": 0.20176163434659541, + "grad_norm": 4.044301986694336, + "learning_rate": 0.00018060975234481202, + "loss": 0.5512, + "step": 8607 + }, + { + "epoch": 0.2017850759214004, + "grad_norm": 3.5322885513305664, + "learning_rate": 0.0001806053929612342, + "loss": 0.5013, + "step": 8608 + }, + { + "epoch": 0.2018085174962054, + "grad_norm": 6.7514238357543945, + "learning_rate": 0.0001806010331402892, + "loss": 0.8274, + "step": 8609 + }, + { + "epoch": 0.2018319590710104, + "grad_norm": 3.6775429248809814, + "learning_rate": 0.00018059667288200076, + "loss": 2.2019, + "step": 8610 + }, + { + "epoch": 0.2018554006458154, + "grad_norm": 4.393826484680176, + "learning_rate": 0.00018059231218639247, + "loss": 0.254, + "step": 8611 + }, + { + "epoch": 0.20187884222062039, + "grad_norm": 1.0414988994598389, + "learning_rate": 0.00018058795105348806, + "loss": 0.1151, + "step": 8612 + }, + { + "epoch": 0.20190228379542538, + "grad_norm": 1.9811079502105713, + "learning_rate": 0.00018058358948331115, + "loss": 0.2726, + "step": 8613 + }, + { + "epoch": 0.20192572537023037, + "grad_norm": 5.388841152191162, + "learning_rate": 0.00018057922747588543, + "loss": 0.5067, + "step": 8614 + }, + { + "epoch": 0.20194916694503537, + "grad_norm": 4.256533145904541, + "learning_rate": 0.00018057486503123452, + "loss": 0.7616, + "step": 8615 + }, + { + "epoch": 0.20197260851984036, + "grad_norm": 1.988050103187561, + "learning_rate": 0.00018057050214938215, + "loss": 0.2232, + "step": 8616 + }, + { + "epoch": 0.20199605009464536, + "grad_norm": 3.3432469367980957, + "learning_rate": 0.00018056613883035198, + "loss": 0.6348, + "step": 8617 + }, + { + "epoch": 0.20201949166945035, + "grad_norm": 2.28440523147583, + "learning_rate": 0.00018056177507416767, + "loss": 1.5542, + "step": 8618 + }, + { + "epoch": 0.20204293324425535, + "grad_norm": 2.91536808013916, + "learning_rate": 0.00018055741088085291, + "loss": 0.3629, + "step": 8619 + }, + { + "epoch": 0.20206637481906034, + "grad_norm": 4.708189010620117, + "learning_rate": 0.00018055304625043137, + "loss": 0.8365, + "step": 8620 + }, + { + "epoch": 0.20208981639386533, + "grad_norm": 4.419531345367432, + "learning_rate": 0.00018054868118292674, + "loss": 0.6177, + "step": 8621 + }, + { + "epoch": 0.20211325796867033, + "grad_norm": 1.057521939277649, + "learning_rate": 0.0001805443156783627, + "loss": 0.1085, + "step": 8622 + }, + { + "epoch": 0.20213669954347532, + "grad_norm": 2.275635242462158, + "learning_rate": 0.0001805399497367629, + "loss": 0.2311, + "step": 8623 + }, + { + "epoch": 0.20216014111828032, + "grad_norm": 5.171917915344238, + "learning_rate": 0.0001805355833581511, + "loss": 0.6866, + "step": 8624 + }, + { + "epoch": 0.2021835826930853, + "grad_norm": 3.3950328826904297, + "learning_rate": 0.00018053121654255095, + "loss": 0.3595, + "step": 8625 + }, + { + "epoch": 0.2022070242678903, + "grad_norm": 3.840280294418335, + "learning_rate": 0.00018052684928998618, + "loss": 0.42, + "step": 8626 + }, + { + "epoch": 0.2022304658426953, + "grad_norm": 7.6023945808410645, + "learning_rate": 0.00018052248160048046, + "loss": 0.4754, + "step": 8627 + }, + { + "epoch": 0.2022539074175003, + "grad_norm": 2.031001091003418, + "learning_rate": 0.00018051811347405747, + "loss": 0.3496, + "step": 8628 + }, + { + "epoch": 0.2022773489923053, + "grad_norm": 1.3859272003173828, + "learning_rate": 0.00018051374491074093, + "loss": 0.8582, + "step": 8629 + }, + { + "epoch": 0.2023007905671103, + "grad_norm": 3.931617021560669, + "learning_rate": 0.00018050937591055456, + "loss": 0.4563, + "step": 8630 + }, + { + "epoch": 0.2023242321419153, + "grad_norm": 1.4839898347854614, + "learning_rate": 0.00018050500647352204, + "loss": 0.2104, + "step": 8631 + }, + { + "epoch": 0.2023476737167203, + "grad_norm": 1.8803437948226929, + "learning_rate": 0.0001805006365996671, + "loss": 1.265, + "step": 8632 + }, + { + "epoch": 0.2023711152915253, + "grad_norm": 2.7273900508880615, + "learning_rate": 0.00018049626628901345, + "loss": 0.3244, + "step": 8633 + }, + { + "epoch": 0.2023945568663303, + "grad_norm": 2.172921657562256, + "learning_rate": 0.00018049189554158478, + "loss": 0.1916, + "step": 8634 + }, + { + "epoch": 0.20241799844113528, + "grad_norm": 5.128176212310791, + "learning_rate": 0.00018048752435740484, + "loss": 0.7443, + "step": 8635 + }, + { + "epoch": 0.20244144001594028, + "grad_norm": 2.359210252761841, + "learning_rate": 0.00018048315273649733, + "loss": 0.3559, + "step": 8636 + }, + { + "epoch": 0.20246488159074527, + "grad_norm": 4.003720760345459, + "learning_rate": 0.00018047878067888596, + "loss": 0.4174, + "step": 8637 + }, + { + "epoch": 0.20248832316555027, + "grad_norm": 2.85103702545166, + "learning_rate": 0.00018047440818459444, + "loss": 0.2842, + "step": 8638 + }, + { + "epoch": 0.20251176474035526, + "grad_norm": 1.0277701616287231, + "learning_rate": 0.00018047003525364653, + "loss": 0.0971, + "step": 8639 + }, + { + "epoch": 0.20253520631516025, + "grad_norm": 2.5605506896972656, + "learning_rate": 0.00018046566188606596, + "loss": 0.3459, + "step": 8640 + }, + { + "epoch": 0.20255864788996525, + "grad_norm": 1.570191740989685, + "learning_rate": 0.00018046128808187645, + "loss": 0.1746, + "step": 8641 + }, + { + "epoch": 0.20258208946477024, + "grad_norm": 3.291217803955078, + "learning_rate": 0.00018045691384110173, + "loss": 0.4901, + "step": 8642 + }, + { + "epoch": 0.20260553103957524, + "grad_norm": 2.0638880729675293, + "learning_rate": 0.00018045253916376553, + "loss": 1.0958, + "step": 8643 + }, + { + "epoch": 0.20262897261438023, + "grad_norm": 4.025032043457031, + "learning_rate": 0.00018044816404989158, + "loss": 0.3736, + "step": 8644 + }, + { + "epoch": 0.20265241418918523, + "grad_norm": 1.7042176723480225, + "learning_rate": 0.00018044378849950364, + "loss": 0.1156, + "step": 8645 + }, + { + "epoch": 0.20267585576399022, + "grad_norm": 6.882967472076416, + "learning_rate": 0.00018043941251262542, + "loss": 0.7187, + "step": 8646 + }, + { + "epoch": 0.20269929733879521, + "grad_norm": 4.484349727630615, + "learning_rate": 0.0001804350360892807, + "loss": 0.8899, + "step": 8647 + }, + { + "epoch": 0.2027227389136002, + "grad_norm": 2.443936586380005, + "learning_rate": 0.00018043065922949324, + "loss": 0.3035, + "step": 8648 + }, + { + "epoch": 0.2027461804884052, + "grad_norm": 3.3945834636688232, + "learning_rate": 0.00018042628193328672, + "loss": 0.2938, + "step": 8649 + }, + { + "epoch": 0.2027696220632102, + "grad_norm": 4.397811412811279, + "learning_rate": 0.00018042190420068495, + "loss": 0.6982, + "step": 8650 + }, + { + "epoch": 0.2027930636380152, + "grad_norm": 2.684661626815796, + "learning_rate": 0.00018041752603171167, + "loss": 0.2246, + "step": 8651 + }, + { + "epoch": 0.2028165052128202, + "grad_norm": 4.227006912231445, + "learning_rate": 0.0001804131474263906, + "loss": 0.4707, + "step": 8652 + }, + { + "epoch": 0.20283994678762518, + "grad_norm": 2.1924171447753906, + "learning_rate": 0.00018040876838474557, + "loss": 0.2863, + "step": 8653 + }, + { + "epoch": 0.20286338836243017, + "grad_norm": 3.342683792114258, + "learning_rate": 0.0001804043889068003, + "loss": 0.3357, + "step": 8654 + }, + { + "epoch": 0.2028868299372352, + "grad_norm": 6.245827674865723, + "learning_rate": 0.00018040000899257854, + "loss": 0.2329, + "step": 8655 + }, + { + "epoch": 0.2029102715120402, + "grad_norm": 2.09515380859375, + "learning_rate": 0.00018039562864210408, + "loss": 0.3669, + "step": 8656 + }, + { + "epoch": 0.20293371308684519, + "grad_norm": 3.9545280933380127, + "learning_rate": 0.00018039124785540066, + "loss": 0.6412, + "step": 8657 + }, + { + "epoch": 0.20295715466165018, + "grad_norm": 4.712405681610107, + "learning_rate": 0.0001803868666324921, + "loss": 0.5496, + "step": 8658 + }, + { + "epoch": 0.20298059623645517, + "grad_norm": 1.042830228805542, + "learning_rate": 0.0001803824849734021, + "loss": 0.1292, + "step": 8659 + }, + { + "epoch": 0.20300403781126017, + "grad_norm": 4.038870334625244, + "learning_rate": 0.00018037810287815454, + "loss": 0.4029, + "step": 8660 + }, + { + "epoch": 0.20302747938606516, + "grad_norm": 0.7249897718429565, + "learning_rate": 0.00018037372034677306, + "loss": 0.0653, + "step": 8661 + }, + { + "epoch": 0.20305092096087016, + "grad_norm": 0.9976685047149658, + "learning_rate": 0.00018036933737928157, + "loss": 0.0969, + "step": 8662 + }, + { + "epoch": 0.20307436253567515, + "grad_norm": 4.903814315795898, + "learning_rate": 0.0001803649539757038, + "loss": 0.5566, + "step": 8663 + }, + { + "epoch": 0.20309780411048015, + "grad_norm": 3.9949095249176025, + "learning_rate": 0.00018036057013606348, + "loss": 0.3398, + "step": 8664 + }, + { + "epoch": 0.20312124568528514, + "grad_norm": 3.221855878829956, + "learning_rate": 0.00018035618586038446, + "loss": 0.3251, + "step": 8665 + }, + { + "epoch": 0.20314468726009013, + "grad_norm": 2.3791966438293457, + "learning_rate": 0.00018035180114869053, + "loss": 0.1596, + "step": 8666 + }, + { + "epoch": 0.20316812883489513, + "grad_norm": 1.5395102500915527, + "learning_rate": 0.00018034741600100545, + "loss": 0.2501, + "step": 8667 + }, + { + "epoch": 0.20319157040970012, + "grad_norm": 3.8249237537384033, + "learning_rate": 0.00018034303041735306, + "loss": 0.463, + "step": 8668 + }, + { + "epoch": 0.20321501198450512, + "grad_norm": 5.227147579193115, + "learning_rate": 0.00018033864439775713, + "loss": 0.7244, + "step": 8669 + }, + { + "epoch": 0.2032384535593101, + "grad_norm": 2.891500234603882, + "learning_rate": 0.00018033425794224144, + "loss": 0.4641, + "step": 8670 + }, + { + "epoch": 0.2032618951341151, + "grad_norm": 0.6396132707595825, + "learning_rate": 0.0001803298710508298, + "loss": 0.064, + "step": 8671 + }, + { + "epoch": 0.2032853367089201, + "grad_norm": 3.9205381870269775, + "learning_rate": 0.00018032548372354606, + "loss": 0.3658, + "step": 8672 + }, + { + "epoch": 0.2033087782837251, + "grad_norm": 2.0540473461151123, + "learning_rate": 0.00018032109596041393, + "loss": 0.277, + "step": 8673 + }, + { + "epoch": 0.2033322198585301, + "grad_norm": 1.256531834602356, + "learning_rate": 0.00018031670776145727, + "loss": 0.1706, + "step": 8674 + }, + { + "epoch": 0.20335566143333508, + "grad_norm": 2.219677448272705, + "learning_rate": 0.00018031231912669994, + "loss": 1.4417, + "step": 8675 + }, + { + "epoch": 0.20337910300814008, + "grad_norm": 2.3156168460845947, + "learning_rate": 0.00018030793005616573, + "loss": 1.5963, + "step": 8676 + }, + { + "epoch": 0.20340254458294507, + "grad_norm": 3.938015937805176, + "learning_rate": 0.0001803035405498784, + "loss": 0.5156, + "step": 8677 + }, + { + "epoch": 0.20342598615775007, + "grad_norm": 1.6114164590835571, + "learning_rate": 0.00018029915060786177, + "loss": 0.2775, + "step": 8678 + }, + { + "epoch": 0.20344942773255506, + "grad_norm": 3.75022292137146, + "learning_rate": 0.00018029476023013972, + "loss": 0.4299, + "step": 8679 + }, + { + "epoch": 0.20347286930736005, + "grad_norm": 3.600663423538208, + "learning_rate": 0.00018029036941673608, + "loss": 0.3164, + "step": 8680 + }, + { + "epoch": 0.20349631088216508, + "grad_norm": 4.639988899230957, + "learning_rate": 0.0001802859781676746, + "loss": 0.4244, + "step": 8681 + }, + { + "epoch": 0.20351975245697007, + "grad_norm": 3.442795515060425, + "learning_rate": 0.00018028158648297912, + "loss": 0.5959, + "step": 8682 + }, + { + "epoch": 0.20354319403177507, + "grad_norm": 5.632937908172607, + "learning_rate": 0.00018027719436267353, + "loss": 0.8064, + "step": 8683 + }, + { + "epoch": 0.20356663560658006, + "grad_norm": 4.870721340179443, + "learning_rate": 0.0001802728018067816, + "loss": 0.5752, + "step": 8684 + }, + { + "epoch": 0.20359007718138505, + "grad_norm": 2.8035688400268555, + "learning_rate": 0.00018026840881532722, + "loss": 0.3146, + "step": 8685 + }, + { + "epoch": 0.20361351875619005, + "grad_norm": 4.272928714752197, + "learning_rate": 0.00018026401538833418, + "loss": 0.551, + "step": 8686 + }, + { + "epoch": 0.20363696033099504, + "grad_norm": 3.1713709831237793, + "learning_rate": 0.00018025962152582633, + "loss": 0.3181, + "step": 8687 + }, + { + "epoch": 0.20366040190580004, + "grad_norm": 2.378801107406616, + "learning_rate": 0.00018025522722782755, + "loss": 0.0999, + "step": 8688 + }, + { + "epoch": 0.20368384348060503, + "grad_norm": 5.68488883972168, + "learning_rate": 0.0001802508324943616, + "loss": 0.682, + "step": 8689 + }, + { + "epoch": 0.20370728505541003, + "grad_norm": 4.11115837097168, + "learning_rate": 0.00018024643732545244, + "loss": 0.3383, + "step": 8690 + }, + { + "epoch": 0.20373072663021502, + "grad_norm": 2.6942648887634277, + "learning_rate": 0.00018024204172112382, + "loss": 0.1786, + "step": 8691 + }, + { + "epoch": 0.20375416820502001, + "grad_norm": 3.9198577404022217, + "learning_rate": 0.00018023764568139963, + "loss": 0.6355, + "step": 8692 + }, + { + "epoch": 0.203777609779825, + "grad_norm": 2.723729372024536, + "learning_rate": 0.0001802332492063037, + "loss": 0.4383, + "step": 8693 + }, + { + "epoch": 0.20380105135463, + "grad_norm": 4.328869819641113, + "learning_rate": 0.00018022885229585991, + "loss": 0.5411, + "step": 8694 + }, + { + "epoch": 0.203824492929435, + "grad_norm": 5.405379772186279, + "learning_rate": 0.00018022445495009213, + "loss": 0.4683, + "step": 8695 + }, + { + "epoch": 0.20384793450424, + "grad_norm": 3.097668170928955, + "learning_rate": 0.00018022005716902419, + "loss": 0.3832, + "step": 8696 + }, + { + "epoch": 0.20387137607904499, + "grad_norm": 4.1894049644470215, + "learning_rate": 0.00018021565895267995, + "loss": 0.6156, + "step": 8697 + }, + { + "epoch": 0.20389481765384998, + "grad_norm": 1.1406188011169434, + "learning_rate": 0.0001802112603010833, + "loss": 0.1705, + "step": 8698 + }, + { + "epoch": 0.20391825922865497, + "grad_norm": 4.358364105224609, + "learning_rate": 0.00018020686121425809, + "loss": 0.6194, + "step": 8699 + }, + { + "epoch": 0.20394170080345997, + "grad_norm": 1.1589889526367188, + "learning_rate": 0.0001802024616922282, + "loss": 0.1373, + "step": 8700 + }, + { + "epoch": 0.20396514237826496, + "grad_norm": 4.109663963317871, + "learning_rate": 0.00018019806173501753, + "loss": 0.4857, + "step": 8701 + }, + { + "epoch": 0.20398858395306996, + "grad_norm": 4.046588897705078, + "learning_rate": 0.0001801936613426499, + "loss": 0.5616, + "step": 8702 + }, + { + "epoch": 0.20401202552787495, + "grad_norm": 3.908587694168091, + "learning_rate": 0.0001801892605151492, + "loss": 0.3281, + "step": 8703 + }, + { + "epoch": 0.20403546710267995, + "grad_norm": 4.552195072174072, + "learning_rate": 0.0001801848592525393, + "loss": 0.7021, + "step": 8704 + }, + { + "epoch": 0.20405890867748494, + "grad_norm": 2.7087864875793457, + "learning_rate": 0.0001801804575548441, + "loss": 0.3589, + "step": 8705 + }, + { + "epoch": 0.20408235025228996, + "grad_norm": 3.1351816654205322, + "learning_rate": 0.00018017605542208754, + "loss": 1.4835, + "step": 8706 + }, + { + "epoch": 0.20410579182709496, + "grad_norm": 5.323589324951172, + "learning_rate": 0.00018017165285429343, + "loss": 0.4735, + "step": 8707 + }, + { + "epoch": 0.20412923340189995, + "grad_norm": 4.149742603302002, + "learning_rate": 0.00018016724985148562, + "loss": 0.3142, + "step": 8708 + }, + { + "epoch": 0.20415267497670495, + "grad_norm": 0.8142613172531128, + "learning_rate": 0.0001801628464136881, + "loss": 0.1005, + "step": 8709 + }, + { + "epoch": 0.20417611655150994, + "grad_norm": 1.461038589477539, + "learning_rate": 0.00018015844254092476, + "loss": 0.2149, + "step": 8710 + }, + { + "epoch": 0.20419955812631493, + "grad_norm": 4.521058082580566, + "learning_rate": 0.0001801540382332194, + "loss": 0.5888, + "step": 8711 + }, + { + "epoch": 0.20422299970111993, + "grad_norm": 3.424164295196533, + "learning_rate": 0.00018014963349059597, + "loss": 0.3173, + "step": 8712 + }, + { + "epoch": 0.20424644127592492, + "grad_norm": 3.443427801132202, + "learning_rate": 0.0001801452283130784, + "loss": 0.3058, + "step": 8713 + }, + { + "epoch": 0.20426988285072992, + "grad_norm": 3.716745376586914, + "learning_rate": 0.00018014082270069058, + "loss": 0.3748, + "step": 8714 + }, + { + "epoch": 0.2042933244255349, + "grad_norm": 3.0533761978149414, + "learning_rate": 0.0001801364166534564, + "loss": 0.3598, + "step": 8715 + }, + { + "epoch": 0.2043167660003399, + "grad_norm": 0.7665550708770752, + "learning_rate": 0.00018013201017139972, + "loss": 0.0943, + "step": 8716 + }, + { + "epoch": 0.2043402075751449, + "grad_norm": 4.721924304962158, + "learning_rate": 0.00018012760325454455, + "loss": 0.5375, + "step": 8717 + }, + { + "epoch": 0.2043636491499499, + "grad_norm": 3.264051914215088, + "learning_rate": 0.00018012319590291474, + "loss": 0.4494, + "step": 8718 + }, + { + "epoch": 0.2043870907247549, + "grad_norm": 3.427891731262207, + "learning_rate": 0.00018011878811653422, + "loss": 0.2975, + "step": 8719 + }, + { + "epoch": 0.20441053229955988, + "grad_norm": 4.00446081161499, + "learning_rate": 0.00018011437989542687, + "loss": 0.324, + "step": 8720 + }, + { + "epoch": 0.20443397387436488, + "grad_norm": 3.6117281913757324, + "learning_rate": 0.0001801099712396167, + "loss": 0.6599, + "step": 8721 + }, + { + "epoch": 0.20445741544916987, + "grad_norm": 0.4577687680721283, + "learning_rate": 0.00018010556214912752, + "loss": 0.038, + "step": 8722 + }, + { + "epoch": 0.20448085702397487, + "grad_norm": 2.5170738697052, + "learning_rate": 0.00018010115262398336, + "loss": 0.206, + "step": 8723 + }, + { + "epoch": 0.20450429859877986, + "grad_norm": 1.011415719985962, + "learning_rate": 0.00018009674266420805, + "loss": 0.1401, + "step": 8724 + }, + { + "epoch": 0.20452774017358485, + "grad_norm": 3.7372376918792725, + "learning_rate": 0.00018009233226982558, + "loss": 0.5454, + "step": 8725 + }, + { + "epoch": 0.20455118174838985, + "grad_norm": 2.4443891048431396, + "learning_rate": 0.00018008792144085986, + "loss": 1.6749, + "step": 8726 + }, + { + "epoch": 0.20457462332319484, + "grad_norm": 3.816779851913452, + "learning_rate": 0.00018008351017733483, + "loss": 0.298, + "step": 8727 + }, + { + "epoch": 0.20459806489799984, + "grad_norm": 4.303123474121094, + "learning_rate": 0.0001800790984792744, + "loss": 0.4527, + "step": 8728 + }, + { + "epoch": 0.20462150647280483, + "grad_norm": 3.142202138900757, + "learning_rate": 0.00018007468634670255, + "loss": 0.3327, + "step": 8729 + }, + { + "epoch": 0.20464494804760983, + "grad_norm": 0.9356063604354858, + "learning_rate": 0.0001800702737796432, + "loss": 0.0931, + "step": 8730 + }, + { + "epoch": 0.20466838962241482, + "grad_norm": 3.024524450302124, + "learning_rate": 0.00018006586077812028, + "loss": 0.2316, + "step": 8731 + }, + { + "epoch": 0.20469183119721984, + "grad_norm": 4.247762203216553, + "learning_rate": 0.00018006144734215775, + "loss": 0.4368, + "step": 8732 + }, + { + "epoch": 0.20471527277202484, + "grad_norm": 4.563129425048828, + "learning_rate": 0.0001800570334717796, + "loss": 0.5786, + "step": 8733 + }, + { + "epoch": 0.20473871434682983, + "grad_norm": 2.2656638622283936, + "learning_rate": 0.00018005261916700967, + "loss": 0.2315, + "step": 8734 + }, + { + "epoch": 0.20476215592163483, + "grad_norm": 3.1596248149871826, + "learning_rate": 0.000180048204427872, + "loss": 0.4792, + "step": 8735 + }, + { + "epoch": 0.20478559749643982, + "grad_norm": 1.886001706123352, + "learning_rate": 0.00018004378925439056, + "loss": 0.7212, + "step": 8736 + }, + { + "epoch": 0.20480903907124481, + "grad_norm": 3.6778531074523926, + "learning_rate": 0.00018003937364658923, + "loss": 0.4297, + "step": 8737 + }, + { + "epoch": 0.2048324806460498, + "grad_norm": 2.6637864112854004, + "learning_rate": 0.00018003495760449202, + "loss": 0.2919, + "step": 8738 + }, + { + "epoch": 0.2048559222208548, + "grad_norm": 6.025853633880615, + "learning_rate": 0.00018003054112812286, + "loss": 0.6164, + "step": 8739 + }, + { + "epoch": 0.2048793637956598, + "grad_norm": 6.5041656494140625, + "learning_rate": 0.0001800261242175058, + "loss": 0.7029, + "step": 8740 + }, + { + "epoch": 0.2049028053704648, + "grad_norm": 1.7569289207458496, + "learning_rate": 0.00018002170687266464, + "loss": 0.1268, + "step": 8741 + }, + { + "epoch": 0.20492624694526979, + "grad_norm": 6.341728210449219, + "learning_rate": 0.0001800172890936235, + "loss": 0.5679, + "step": 8742 + }, + { + "epoch": 0.20494968852007478, + "grad_norm": 1.1765860319137573, + "learning_rate": 0.0001800128708804063, + "loss": 0.1374, + "step": 8743 + }, + { + "epoch": 0.20497313009487977, + "grad_norm": 1.756359577178955, + "learning_rate": 0.00018000845223303702, + "loss": 0.1283, + "step": 8744 + }, + { + "epoch": 0.20499657166968477, + "grad_norm": 0.43854498863220215, + "learning_rate": 0.0001800040331515396, + "loss": 0.053, + "step": 8745 + }, + { + "epoch": 0.20502001324448976, + "grad_norm": 4.148134231567383, + "learning_rate": 0.0001799996136359381, + "loss": 0.345, + "step": 8746 + }, + { + "epoch": 0.20504345481929476, + "grad_norm": 8.004212379455566, + "learning_rate": 0.0001799951936862564, + "loss": 0.6769, + "step": 8747 + }, + { + "epoch": 0.20506689639409975, + "grad_norm": 4.6142048835754395, + "learning_rate": 0.00017999077330251853, + "loss": 0.4821, + "step": 8748 + }, + { + "epoch": 0.20509033796890475, + "grad_norm": 4.359156131744385, + "learning_rate": 0.00017998635248474848, + "loss": 0.5312, + "step": 8749 + }, + { + "epoch": 0.20511377954370974, + "grad_norm": 3.4753916263580322, + "learning_rate": 0.00017998193123297025, + "loss": 0.5325, + "step": 8750 + }, + { + "epoch": 0.20513722111851473, + "grad_norm": 2.874377727508545, + "learning_rate": 0.0001799775095472078, + "loss": 0.2381, + "step": 8751 + }, + { + "epoch": 0.20516066269331973, + "grad_norm": 3.0294227600097656, + "learning_rate": 0.00017997308742748517, + "loss": 0.3807, + "step": 8752 + }, + { + "epoch": 0.20518410426812472, + "grad_norm": 3.2929537296295166, + "learning_rate": 0.00017996866487382626, + "loss": 0.3466, + "step": 8753 + }, + { + "epoch": 0.20520754584292972, + "grad_norm": 3.5089519023895264, + "learning_rate": 0.00017996424188625513, + "loss": 1.8414, + "step": 8754 + }, + { + "epoch": 0.2052309874177347, + "grad_norm": 2.317336320877075, + "learning_rate": 0.00017995981846479582, + "loss": 0.2974, + "step": 8755 + }, + { + "epoch": 0.2052544289925397, + "grad_norm": 8.574871063232422, + "learning_rate": 0.00017995539460947225, + "loss": 0.3266, + "step": 8756 + }, + { + "epoch": 0.20527787056734473, + "grad_norm": 4.340612888336182, + "learning_rate": 0.0001799509703203085, + "loss": 0.4657, + "step": 8757 + }, + { + "epoch": 0.20530131214214972, + "grad_norm": 4.034926414489746, + "learning_rate": 0.0001799465455973285, + "loss": 0.3893, + "step": 8758 + }, + { + "epoch": 0.20532475371695472, + "grad_norm": 1.06405508518219, + "learning_rate": 0.00017994212044055628, + "loss": 0.0922, + "step": 8759 + }, + { + "epoch": 0.2053481952917597, + "grad_norm": 4.349960803985596, + "learning_rate": 0.0001799376948500159, + "loss": 0.43, + "step": 8760 + }, + { + "epoch": 0.2053716368665647, + "grad_norm": 3.0393505096435547, + "learning_rate": 0.00017993326882573136, + "loss": 0.3761, + "step": 8761 + }, + { + "epoch": 0.2053950784413697, + "grad_norm": 2.148635149002075, + "learning_rate": 0.00017992884236772661, + "loss": 1.5235, + "step": 8762 + }, + { + "epoch": 0.2054185200161747, + "grad_norm": 1.910295844078064, + "learning_rate": 0.00017992441547602574, + "loss": 0.2904, + "step": 8763 + }, + { + "epoch": 0.2054419615909797, + "grad_norm": 2.1322848796844482, + "learning_rate": 0.00017991998815065273, + "loss": 0.1637, + "step": 8764 + }, + { + "epoch": 0.20546540316578468, + "grad_norm": 1.697203516960144, + "learning_rate": 0.00017991556039163165, + "loss": 0.1831, + "step": 8765 + }, + { + "epoch": 0.20548884474058968, + "grad_norm": 1.8069462776184082, + "learning_rate": 0.00017991113219898645, + "loss": 0.269, + "step": 8766 + }, + { + "epoch": 0.20551228631539467, + "grad_norm": 5.168079853057861, + "learning_rate": 0.00017990670357274121, + "loss": 0.6689, + "step": 8767 + }, + { + "epoch": 0.20553572789019967, + "grad_norm": 3.6211373805999756, + "learning_rate": 0.00017990227451291998, + "loss": 0.42, + "step": 8768 + }, + { + "epoch": 0.20555916946500466, + "grad_norm": 4.523114204406738, + "learning_rate": 0.00017989784501954673, + "loss": 0.5483, + "step": 8769 + }, + { + "epoch": 0.20558261103980965, + "grad_norm": 3.9037222862243652, + "learning_rate": 0.0001798934150926455, + "loss": 0.526, + "step": 8770 + }, + { + "epoch": 0.20560605261461465, + "grad_norm": 3.4473495483398438, + "learning_rate": 0.0001798889847322404, + "loss": 0.7745, + "step": 8771 + }, + { + "epoch": 0.20562949418941964, + "grad_norm": 4.3832106590271, + "learning_rate": 0.00017988455393835542, + "loss": 0.5509, + "step": 8772 + }, + { + "epoch": 0.20565293576422464, + "grad_norm": 3.314271926879883, + "learning_rate": 0.0001798801227110146, + "loss": 0.3441, + "step": 8773 + }, + { + "epoch": 0.20567637733902963, + "grad_norm": 3.6638739109039307, + "learning_rate": 0.00017987569105024198, + "loss": 0.2953, + "step": 8774 + }, + { + "epoch": 0.20569981891383463, + "grad_norm": 4.714745998382568, + "learning_rate": 0.0001798712589560616, + "loss": 0.7937, + "step": 8775 + }, + { + "epoch": 0.20572326048863962, + "grad_norm": 5.680431842803955, + "learning_rate": 0.00017986682642849754, + "loss": 0.4997, + "step": 8776 + }, + { + "epoch": 0.20574670206344461, + "grad_norm": 2.5607714653015137, + "learning_rate": 0.00017986239346757382, + "loss": 0.3091, + "step": 8777 + }, + { + "epoch": 0.2057701436382496, + "grad_norm": 5.091336250305176, + "learning_rate": 0.00017985796007331456, + "loss": 0.6518, + "step": 8778 + }, + { + "epoch": 0.2057935852130546, + "grad_norm": 0.9134320616722107, + "learning_rate": 0.0001798535262457437, + "loss": 0.084, + "step": 8779 + }, + { + "epoch": 0.2058170267878596, + "grad_norm": 1.74507474899292, + "learning_rate": 0.0001798490919848854, + "loss": 0.2139, + "step": 8780 + }, + { + "epoch": 0.2058404683626646, + "grad_norm": 5.639239311218262, + "learning_rate": 0.00017984465729076367, + "loss": 0.5719, + "step": 8781 + }, + { + "epoch": 0.20586390993746959, + "grad_norm": 1.977541208267212, + "learning_rate": 0.00017984022216340257, + "loss": 1.1128, + "step": 8782 + }, + { + "epoch": 0.2058873515122746, + "grad_norm": 4.0313029289245605, + "learning_rate": 0.0001798357866028262, + "loss": 0.4507, + "step": 8783 + }, + { + "epoch": 0.2059107930870796, + "grad_norm": 3.247027635574341, + "learning_rate": 0.0001798313506090586, + "loss": 0.2351, + "step": 8784 + }, + { + "epoch": 0.2059342346618846, + "grad_norm": 3.4667389392852783, + "learning_rate": 0.00017982691418212384, + "loss": 0.4761, + "step": 8785 + }, + { + "epoch": 0.2059576762366896, + "grad_norm": 1.6390011310577393, + "learning_rate": 0.000179822477322046, + "loss": 0.1998, + "step": 8786 + }, + { + "epoch": 0.20598111781149459, + "grad_norm": 3.240854501724243, + "learning_rate": 0.00017981804002884918, + "loss": 0.2896, + "step": 8787 + }, + { + "epoch": 0.20600455938629958, + "grad_norm": 3.586644172668457, + "learning_rate": 0.0001798136023025574, + "loss": 0.5934, + "step": 8788 + }, + { + "epoch": 0.20602800096110457, + "grad_norm": 2.921510934829712, + "learning_rate": 0.0001798091641431948, + "loss": 0.2218, + "step": 8789 + }, + { + "epoch": 0.20605144253590957, + "grad_norm": 3.7071759700775146, + "learning_rate": 0.0001798047255507854, + "loss": 0.2821, + "step": 8790 + }, + { + "epoch": 0.20607488411071456, + "grad_norm": 6.285760879516602, + "learning_rate": 0.00017980028652535334, + "loss": 0.6124, + "step": 8791 + }, + { + "epoch": 0.20609832568551956, + "grad_norm": 1.948602318763733, + "learning_rate": 0.00017979584706692266, + "loss": 0.2511, + "step": 8792 + }, + { + "epoch": 0.20612176726032455, + "grad_norm": 3.3054771423339844, + "learning_rate": 0.00017979140717551748, + "loss": 0.3279, + "step": 8793 + }, + { + "epoch": 0.20614520883512955, + "grad_norm": 2.300992965698242, + "learning_rate": 0.00017978696685116185, + "loss": 1.6841, + "step": 8794 + }, + { + "epoch": 0.20616865040993454, + "grad_norm": 4.168566703796387, + "learning_rate": 0.00017978252609387993, + "loss": 0.4671, + "step": 8795 + }, + { + "epoch": 0.20619209198473953, + "grad_norm": 5.5192551612854, + "learning_rate": 0.00017977808490369577, + "loss": 0.6177, + "step": 8796 + }, + { + "epoch": 0.20621553355954453, + "grad_norm": 1.0852068662643433, + "learning_rate": 0.00017977364328063347, + "loss": 0.1238, + "step": 8797 + }, + { + "epoch": 0.20623897513434952, + "grad_norm": 4.386826038360596, + "learning_rate": 0.00017976920122471715, + "loss": 0.3918, + "step": 8798 + }, + { + "epoch": 0.20626241670915452, + "grad_norm": 3.3379199504852295, + "learning_rate": 0.00017976475873597092, + "loss": 0.3953, + "step": 8799 + }, + { + "epoch": 0.2062858582839595, + "grad_norm": 4.009288787841797, + "learning_rate": 0.00017976031581441886, + "loss": 0.3427, + "step": 8800 + }, + { + "epoch": 0.2063092998587645, + "grad_norm": 1.0066859722137451, + "learning_rate": 0.00017975587246008506, + "loss": 0.0926, + "step": 8801 + }, + { + "epoch": 0.2063327414335695, + "grad_norm": 1.4230495691299438, + "learning_rate": 0.00017975142867299366, + "loss": 0.1614, + "step": 8802 + }, + { + "epoch": 0.2063561830083745, + "grad_norm": 4.703097820281982, + "learning_rate": 0.00017974698445316877, + "loss": 0.6913, + "step": 8803 + }, + { + "epoch": 0.2063796245831795, + "grad_norm": 1.0652116537094116, + "learning_rate": 0.0001797425398006345, + "loss": 0.0845, + "step": 8804 + }, + { + "epoch": 0.20640306615798448, + "grad_norm": 1.9948735237121582, + "learning_rate": 0.000179738094715415, + "loss": 0.3101, + "step": 8805 + }, + { + "epoch": 0.20642650773278948, + "grad_norm": 4.631972789764404, + "learning_rate": 0.00017973364919753428, + "loss": 0.6687, + "step": 8806 + }, + { + "epoch": 0.20644994930759447, + "grad_norm": 5.198658466339111, + "learning_rate": 0.00017972920324701657, + "loss": 0.4013, + "step": 8807 + }, + { + "epoch": 0.2064733908823995, + "grad_norm": 0.8769065737724304, + "learning_rate": 0.00017972475686388602, + "loss": 0.096, + "step": 8808 + }, + { + "epoch": 0.2064968324572045, + "grad_norm": 2.615644693374634, + "learning_rate": 0.00017972031004816663, + "loss": 0.3303, + "step": 8809 + }, + { + "epoch": 0.20652027403200948, + "grad_norm": 0.7838465571403503, + "learning_rate": 0.00017971586279988262, + "loss": 0.0832, + "step": 8810 + }, + { + "epoch": 0.20654371560681448, + "grad_norm": 3.2978899478912354, + "learning_rate": 0.00017971141511905807, + "loss": 0.352, + "step": 8811 + }, + { + "epoch": 0.20656715718161947, + "grad_norm": 0.6344325542449951, + "learning_rate": 0.0001797069670057172, + "loss": 0.1292, + "step": 8812 + }, + { + "epoch": 0.20659059875642447, + "grad_norm": 4.00587797164917, + "learning_rate": 0.00017970251845988403, + "loss": 0.575, + "step": 8813 + }, + { + "epoch": 0.20661404033122946, + "grad_norm": 1.6543737649917603, + "learning_rate": 0.00017969806948158276, + "loss": 0.1626, + "step": 8814 + }, + { + "epoch": 0.20663748190603445, + "grad_norm": 4.112926959991455, + "learning_rate": 0.00017969362007083753, + "loss": 0.3978, + "step": 8815 + }, + { + "epoch": 0.20666092348083945, + "grad_norm": 3.897669792175293, + "learning_rate": 0.0001796891702276725, + "loss": 0.6125, + "step": 8816 + }, + { + "epoch": 0.20668436505564444, + "grad_norm": 3.1457252502441406, + "learning_rate": 0.00017968471995211179, + "loss": 0.1234, + "step": 8817 + }, + { + "epoch": 0.20670780663044944, + "grad_norm": 3.3841822147369385, + "learning_rate": 0.0001796802692441795, + "loss": 0.4517, + "step": 8818 + }, + { + "epoch": 0.20673124820525443, + "grad_norm": 4.153479099273682, + "learning_rate": 0.00017967581810389988, + "loss": 0.3232, + "step": 8819 + }, + { + "epoch": 0.20675468978005943, + "grad_norm": 3.2243168354034424, + "learning_rate": 0.00017967136653129702, + "loss": 0.4129, + "step": 8820 + }, + { + "epoch": 0.20677813135486442, + "grad_norm": 4.08613920211792, + "learning_rate": 0.00017966691452639508, + "loss": 0.3991, + "step": 8821 + }, + { + "epoch": 0.2068015729296694, + "grad_norm": 6.8505682945251465, + "learning_rate": 0.0001796624620892182, + "loss": 0.7228, + "step": 8822 + }, + { + "epoch": 0.2068250145044744, + "grad_norm": 3.361558437347412, + "learning_rate": 0.0001796580092197906, + "loss": 1.952, + "step": 8823 + }, + { + "epoch": 0.2068484560792794, + "grad_norm": 5.138154029846191, + "learning_rate": 0.0001796535559181364, + "loss": 0.7307, + "step": 8824 + }, + { + "epoch": 0.2068718976540844, + "grad_norm": 1.1427029371261597, + "learning_rate": 0.00017964910218427974, + "loss": 0.1286, + "step": 8825 + }, + { + "epoch": 0.2068953392288894, + "grad_norm": 0.9518641829490662, + "learning_rate": 0.00017964464801824482, + "loss": 0.0501, + "step": 8826 + }, + { + "epoch": 0.20691878080369439, + "grad_norm": 1.627130389213562, + "learning_rate": 0.0001796401934200558, + "loss": 0.1826, + "step": 8827 + }, + { + "epoch": 0.20694222237849938, + "grad_norm": 8.009007453918457, + "learning_rate": 0.00017963573838973687, + "loss": 0.3307, + "step": 8828 + }, + { + "epoch": 0.20696566395330437, + "grad_norm": 1.1895630359649658, + "learning_rate": 0.00017963128292731214, + "loss": 0.1352, + "step": 8829 + }, + { + "epoch": 0.20698910552810937, + "grad_norm": 1.956465721130371, + "learning_rate": 0.00017962682703280588, + "loss": 1.0162, + "step": 8830 + }, + { + "epoch": 0.20701254710291436, + "grad_norm": 2.8462138175964355, + "learning_rate": 0.0001796223707062422, + "loss": 0.3673, + "step": 8831 + }, + { + "epoch": 0.20703598867771936, + "grad_norm": 1.2200853824615479, + "learning_rate": 0.00017961791394764526, + "loss": 0.1349, + "step": 8832 + }, + { + "epoch": 0.20705943025252435, + "grad_norm": 4.383017539978027, + "learning_rate": 0.00017961345675703933, + "loss": 0.7219, + "step": 8833 + }, + { + "epoch": 0.20708287182732937, + "grad_norm": 5.232057571411133, + "learning_rate": 0.00017960899913444854, + "loss": 0.286, + "step": 8834 + }, + { + "epoch": 0.20710631340213437, + "grad_norm": 2.260690689086914, + "learning_rate": 0.00017960454107989704, + "loss": 0.3002, + "step": 8835 + }, + { + "epoch": 0.20712975497693936, + "grad_norm": 4.764307498931885, + "learning_rate": 0.0001796000825934091, + "loss": 0.4801, + "step": 8836 + }, + { + "epoch": 0.20715319655174436, + "grad_norm": 0.43238529562950134, + "learning_rate": 0.00017959562367500889, + "loss": 0.0544, + "step": 8837 + }, + { + "epoch": 0.20717663812654935, + "grad_norm": 4.594024658203125, + "learning_rate": 0.00017959116432472055, + "loss": 0.5141, + "step": 8838 + }, + { + "epoch": 0.20720007970135434, + "grad_norm": 1.3002245426177979, + "learning_rate": 0.00017958670454256833, + "loss": 0.1749, + "step": 8839 + }, + { + "epoch": 0.20722352127615934, + "grad_norm": 4.073544025421143, + "learning_rate": 0.0001795822443285764, + "loss": 0.4229, + "step": 8840 + }, + { + "epoch": 0.20724696285096433, + "grad_norm": 3.978513717651367, + "learning_rate": 0.000179577783682769, + "loss": 0.3211, + "step": 8841 + }, + { + "epoch": 0.20727040442576933, + "grad_norm": 1.8922039270401, + "learning_rate": 0.0001795733226051703, + "loss": 1.1298, + "step": 8842 + }, + { + "epoch": 0.20729384600057432, + "grad_norm": 3.813842296600342, + "learning_rate": 0.00017956886109580448, + "loss": 0.4784, + "step": 8843 + }, + { + "epoch": 0.20731728757537932, + "grad_norm": 2.313121795654297, + "learning_rate": 0.00017956439915469582, + "loss": 0.2122, + "step": 8844 + }, + { + "epoch": 0.2073407291501843, + "grad_norm": 5.938575267791748, + "learning_rate": 0.0001795599367818685, + "loss": 0.4253, + "step": 8845 + }, + { + "epoch": 0.2073641707249893, + "grad_norm": 1.437057614326477, + "learning_rate": 0.0001795554739773467, + "loss": 0.1826, + "step": 8846 + }, + { + "epoch": 0.2073876122997943, + "grad_norm": 1.4623358249664307, + "learning_rate": 0.00017955101074115465, + "loss": 0.1403, + "step": 8847 + }, + { + "epoch": 0.2074110538745993, + "grad_norm": 4.322736740112305, + "learning_rate": 0.00017954654707331665, + "loss": 1.4204, + "step": 8848 + }, + { + "epoch": 0.2074344954494043, + "grad_norm": 4.223784923553467, + "learning_rate": 0.00017954208297385677, + "loss": 0.6092, + "step": 8849 + }, + { + "epoch": 0.20745793702420928, + "grad_norm": 3.9745869636535645, + "learning_rate": 0.00017953761844279934, + "loss": 0.4062, + "step": 8850 + }, + { + "epoch": 0.20748137859901428, + "grad_norm": 4.408514022827148, + "learning_rate": 0.00017953315348016857, + "loss": 0.3858, + "step": 8851 + }, + { + "epoch": 0.20750482017381927, + "grad_norm": 6.777222633361816, + "learning_rate": 0.00017952868808598864, + "loss": 0.5425, + "step": 8852 + }, + { + "epoch": 0.20752826174862427, + "grad_norm": 3.295032501220703, + "learning_rate": 0.00017952422226028385, + "loss": 0.3341, + "step": 8853 + }, + { + "epoch": 0.20755170332342926, + "grad_norm": 1.4580563306808472, + "learning_rate": 0.00017951975600307836, + "loss": 0.1801, + "step": 8854 + }, + { + "epoch": 0.20757514489823425, + "grad_norm": 1.442577600479126, + "learning_rate": 0.00017951528931439646, + "loss": 0.1489, + "step": 8855 + }, + { + "epoch": 0.20759858647303925, + "grad_norm": 4.198638916015625, + "learning_rate": 0.00017951082219426237, + "loss": 0.699, + "step": 8856 + }, + { + "epoch": 0.20762202804784424, + "grad_norm": 3.3785974979400635, + "learning_rate": 0.00017950635464270028, + "loss": 0.4286, + "step": 8857 + }, + { + "epoch": 0.20764546962264924, + "grad_norm": 2.3941192626953125, + "learning_rate": 0.00017950188665973453, + "loss": 0.1409, + "step": 8858 + }, + { + "epoch": 0.20766891119745423, + "grad_norm": 1.0018246173858643, + "learning_rate": 0.00017949741824538928, + "loss": 0.157, + "step": 8859 + }, + { + "epoch": 0.20769235277225925, + "grad_norm": 3.2837345600128174, + "learning_rate": 0.0001794929493996888, + "loss": 0.3724, + "step": 8860 + }, + { + "epoch": 0.20771579434706425, + "grad_norm": 4.399288177490234, + "learning_rate": 0.00017948848012265734, + "loss": 0.4553, + "step": 8861 + }, + { + "epoch": 0.20773923592186924, + "grad_norm": 3.4843201637268066, + "learning_rate": 0.00017948401041431916, + "loss": 0.323, + "step": 8862 + }, + { + "epoch": 0.20776267749667424, + "grad_norm": 4.673262119293213, + "learning_rate": 0.0001794795402746985, + "loss": 0.373, + "step": 8863 + }, + { + "epoch": 0.20778611907147923, + "grad_norm": 4.276426315307617, + "learning_rate": 0.00017947506970381963, + "loss": 0.7561, + "step": 8864 + }, + { + "epoch": 0.20780956064628422, + "grad_norm": 7.570444107055664, + "learning_rate": 0.00017947059870170682, + "loss": 0.6538, + "step": 8865 + }, + { + "epoch": 0.20783300222108922, + "grad_norm": 1.2113267183303833, + "learning_rate": 0.00017946612726838426, + "loss": 0.0949, + "step": 8866 + }, + { + "epoch": 0.2078564437958942, + "grad_norm": 1.2834351062774658, + "learning_rate": 0.00017946165540387628, + "loss": 0.19, + "step": 8867 + }, + { + "epoch": 0.2078798853706992, + "grad_norm": 1.20646071434021, + "learning_rate": 0.00017945718310820712, + "loss": 0.1823, + "step": 8868 + }, + { + "epoch": 0.2079033269455042, + "grad_norm": 3.1459810733795166, + "learning_rate": 0.00017945271038140103, + "loss": 1.9585, + "step": 8869 + }, + { + "epoch": 0.2079267685203092, + "grad_norm": 2.2986056804656982, + "learning_rate": 0.00017944823722348233, + "loss": 0.1347, + "step": 8870 + }, + { + "epoch": 0.2079502100951142, + "grad_norm": 4.21511697769165, + "learning_rate": 0.00017944376363447526, + "loss": 0.5749, + "step": 8871 + }, + { + "epoch": 0.20797365166991919, + "grad_norm": 0.867743730545044, + "learning_rate": 0.0001794392896144041, + "loss": 0.0736, + "step": 8872 + }, + { + "epoch": 0.20799709324472418, + "grad_norm": 0.5412430167198181, + "learning_rate": 0.0001794348151632931, + "loss": 0.0592, + "step": 8873 + }, + { + "epoch": 0.20802053481952917, + "grad_norm": 4.927217483520508, + "learning_rate": 0.00017943034028116656, + "loss": 0.5648, + "step": 8874 + }, + { + "epoch": 0.20804397639433417, + "grad_norm": 4.574788570404053, + "learning_rate": 0.00017942586496804878, + "loss": 0.3826, + "step": 8875 + }, + { + "epoch": 0.20806741796913916, + "grad_norm": 2.492753505706787, + "learning_rate": 0.000179421389223964, + "loss": 2.079, + "step": 8876 + }, + { + "epoch": 0.20809085954394416, + "grad_norm": 5.546772003173828, + "learning_rate": 0.00017941691304893654, + "loss": 0.6651, + "step": 8877 + }, + { + "epoch": 0.20811430111874915, + "grad_norm": 4.176913261413574, + "learning_rate": 0.00017941243644299066, + "loss": 0.4629, + "step": 8878 + }, + { + "epoch": 0.20813774269355415, + "grad_norm": 3.23330020904541, + "learning_rate": 0.00017940795940615066, + "loss": 0.2252, + "step": 8879 + }, + { + "epoch": 0.20816118426835914, + "grad_norm": 2.6194944381713867, + "learning_rate": 0.00017940348193844089, + "loss": 2.1219, + "step": 8880 + }, + { + "epoch": 0.20818462584316413, + "grad_norm": 3.553870439529419, + "learning_rate": 0.00017939900403988557, + "loss": 0.3297, + "step": 8881 + }, + { + "epoch": 0.20820806741796913, + "grad_norm": 5.720286846160889, + "learning_rate": 0.00017939452571050902, + "loss": 0.7758, + "step": 8882 + }, + { + "epoch": 0.20823150899277412, + "grad_norm": 3.922968864440918, + "learning_rate": 0.00017939004695033555, + "loss": 0.5509, + "step": 8883 + }, + { + "epoch": 0.20825495056757912, + "grad_norm": 4.751302242279053, + "learning_rate": 0.00017938556775938942, + "loss": 0.7679, + "step": 8884 + }, + { + "epoch": 0.20827839214238414, + "grad_norm": 3.440640687942505, + "learning_rate": 0.00017938108813769497, + "loss": 0.5567, + "step": 8885 + }, + { + "epoch": 0.20830183371718913, + "grad_norm": 1.1765187978744507, + "learning_rate": 0.00017937660808527657, + "loss": 0.1489, + "step": 8886 + }, + { + "epoch": 0.20832527529199413, + "grad_norm": 1.1163592338562012, + "learning_rate": 0.0001793721276021584, + "loss": 0.1532, + "step": 8887 + }, + { + "epoch": 0.20834871686679912, + "grad_norm": 2.17842960357666, + "learning_rate": 0.00017936764668836485, + "loss": 0.282, + "step": 8888 + }, + { + "epoch": 0.20837215844160412, + "grad_norm": 0.523939311504364, + "learning_rate": 0.0001793631653439202, + "loss": 0.0398, + "step": 8889 + }, + { + "epoch": 0.2083956000164091, + "grad_norm": 2.664750099182129, + "learning_rate": 0.0001793586835688488, + "loss": 0.3578, + "step": 8890 + }, + { + "epoch": 0.2084190415912141, + "grad_norm": 4.296961784362793, + "learning_rate": 0.00017935420136317494, + "loss": 0.5844, + "step": 8891 + }, + { + "epoch": 0.2084424831660191, + "grad_norm": 4.555168628692627, + "learning_rate": 0.00017934971872692298, + "loss": 0.5879, + "step": 8892 + }, + { + "epoch": 0.2084659247408241, + "grad_norm": 4.302995681762695, + "learning_rate": 0.0001793452356601172, + "loss": 0.4553, + "step": 8893 + }, + { + "epoch": 0.2084893663156291, + "grad_norm": 4.904271125793457, + "learning_rate": 0.00017934075216278196, + "loss": 0.4151, + "step": 8894 + }, + { + "epoch": 0.20851280789043408, + "grad_norm": 3.037782907485962, + "learning_rate": 0.00017933626823494155, + "loss": 0.2576, + "step": 8895 + }, + { + "epoch": 0.20853624946523908, + "grad_norm": 1.3642995357513428, + "learning_rate": 0.00017933178387662032, + "loss": 0.1622, + "step": 8896 + }, + { + "epoch": 0.20855969104004407, + "grad_norm": 2.73824143409729, + "learning_rate": 0.00017932729908784259, + "loss": 0.205, + "step": 8897 + }, + { + "epoch": 0.20858313261484906, + "grad_norm": 2.469611406326294, + "learning_rate": 0.00017932281386863269, + "loss": 0.2316, + "step": 8898 + }, + { + "epoch": 0.20860657418965406, + "grad_norm": 4.022745132446289, + "learning_rate": 0.000179318328219015, + "loss": 0.4961, + "step": 8899 + }, + { + "epoch": 0.20863001576445905, + "grad_norm": 3.7558135986328125, + "learning_rate": 0.00017931384213901383, + "loss": 0.4487, + "step": 8900 + }, + { + "epoch": 0.20865345733926405, + "grad_norm": 0.5400941967964172, + "learning_rate": 0.0001793093556286535, + "loss": 0.0526, + "step": 8901 + }, + { + "epoch": 0.20867689891406904, + "grad_norm": 1.4844951629638672, + "learning_rate": 0.0001793048686879584, + "loss": 0.1367, + "step": 8902 + }, + { + "epoch": 0.20870034048887404, + "grad_norm": 2.34501314163208, + "learning_rate": 0.00017930038131695283, + "loss": 0.1787, + "step": 8903 + }, + { + "epoch": 0.20872378206367903, + "grad_norm": 2.8547229766845703, + "learning_rate": 0.0001792958935156612, + "loss": 0.3073, + "step": 8904 + }, + { + "epoch": 0.20874722363848403, + "grad_norm": 1.02041494846344, + "learning_rate": 0.0001792914052841078, + "loss": 0.1299, + "step": 8905 + }, + { + "epoch": 0.20877066521328902, + "grad_norm": 5.912494659423828, + "learning_rate": 0.000179286916622317, + "loss": 1.3108, + "step": 8906 + }, + { + "epoch": 0.208794106788094, + "grad_norm": 3.5685412883758545, + "learning_rate": 0.00017928242753031314, + "loss": 0.4688, + "step": 8907 + }, + { + "epoch": 0.208817548362899, + "grad_norm": 3.3999361991882324, + "learning_rate": 0.00017927793800812062, + "loss": 0.3888, + "step": 8908 + }, + { + "epoch": 0.208840989937704, + "grad_norm": 1.4006503820419312, + "learning_rate": 0.0001792734480557638, + "loss": 0.1318, + "step": 8909 + }, + { + "epoch": 0.208864431512509, + "grad_norm": 1.6528843641281128, + "learning_rate": 0.000179268957673267, + "loss": 0.2005, + "step": 8910 + }, + { + "epoch": 0.20888787308731402, + "grad_norm": 3.8757059574127197, + "learning_rate": 0.00017926446686065462, + "loss": 0.402, + "step": 8911 + }, + { + "epoch": 0.208911314662119, + "grad_norm": 3.771888256072998, + "learning_rate": 0.000179259975617951, + "loss": 0.4761, + "step": 8912 + }, + { + "epoch": 0.208934756236924, + "grad_norm": 3.878031015396118, + "learning_rate": 0.00017925548394518052, + "loss": 0.5664, + "step": 8913 + }, + { + "epoch": 0.208958197811729, + "grad_norm": 2.386692762374878, + "learning_rate": 0.0001792509918423676, + "loss": 0.177, + "step": 8914 + }, + { + "epoch": 0.208981639386534, + "grad_norm": 3.7576088905334473, + "learning_rate": 0.0001792464993095365, + "loss": 0.3673, + "step": 8915 + }, + { + "epoch": 0.209005080961339, + "grad_norm": 2.2664167881011963, + "learning_rate": 0.0001792420063467117, + "loss": 0.403, + "step": 8916 + }, + { + "epoch": 0.20902852253614398, + "grad_norm": 4.0586771965026855, + "learning_rate": 0.00017923751295391755, + "loss": 0.4579, + "step": 8917 + }, + { + "epoch": 0.20905196411094898, + "grad_norm": 1.8850414752960205, + "learning_rate": 0.00017923301913117844, + "loss": 1.3147, + "step": 8918 + }, + { + "epoch": 0.20907540568575397, + "grad_norm": 1.1764743328094482, + "learning_rate": 0.0001792285248785187, + "loss": 0.1523, + "step": 8919 + }, + { + "epoch": 0.20909884726055897, + "grad_norm": 4.496969699859619, + "learning_rate": 0.0001792240301959628, + "loss": 0.6789, + "step": 8920 + }, + { + "epoch": 0.20912228883536396, + "grad_norm": 4.433244705200195, + "learning_rate": 0.00017921953508353506, + "loss": 0.7434, + "step": 8921 + }, + { + "epoch": 0.20914573041016896, + "grad_norm": 3.1074635982513428, + "learning_rate": 0.0001792150395412599, + "loss": 0.447, + "step": 8922 + }, + { + "epoch": 0.20916917198497395, + "grad_norm": 2.619562864303589, + "learning_rate": 0.0001792105435691617, + "loss": 0.2321, + "step": 8923 + }, + { + "epoch": 0.20919261355977894, + "grad_norm": 1.409883975982666, + "learning_rate": 0.0001792060471672649, + "loss": 0.1493, + "step": 8924 + }, + { + "epoch": 0.20921605513458394, + "grad_norm": 1.0045536756515503, + "learning_rate": 0.0001792015503355938, + "loss": 0.1634, + "step": 8925 + }, + { + "epoch": 0.20923949670938893, + "grad_norm": 2.0390584468841553, + "learning_rate": 0.0001791970530741729, + "loss": 0.3181, + "step": 8926 + }, + { + "epoch": 0.20926293828419393, + "grad_norm": 3.922198534011841, + "learning_rate": 0.00017919255538302654, + "loss": 0.5434, + "step": 8927 + }, + { + "epoch": 0.20928637985899892, + "grad_norm": 0.2812798023223877, + "learning_rate": 0.00017918805726217918, + "loss": 0.0241, + "step": 8928 + }, + { + "epoch": 0.20930982143380392, + "grad_norm": 3.1326630115509033, + "learning_rate": 0.00017918355871165517, + "loss": 0.4085, + "step": 8929 + }, + { + "epoch": 0.2093332630086089, + "grad_norm": 5.945324897766113, + "learning_rate": 0.00017917905973147895, + "loss": 0.5397, + "step": 8930 + }, + { + "epoch": 0.2093567045834139, + "grad_norm": 2.7271833419799805, + "learning_rate": 0.00017917456032167493, + "loss": 0.2671, + "step": 8931 + }, + { + "epoch": 0.2093801461582189, + "grad_norm": 5.411003589630127, + "learning_rate": 0.00017917006048226753, + "loss": 0.4254, + "step": 8932 + }, + { + "epoch": 0.2094035877330239, + "grad_norm": 3.474438428878784, + "learning_rate": 0.00017916556021328113, + "loss": 0.4005, + "step": 8933 + }, + { + "epoch": 0.2094270293078289, + "grad_norm": 3.3175718784332275, + "learning_rate": 0.0001791610595147402, + "loss": 2.3756, + "step": 8934 + }, + { + "epoch": 0.20945047088263388, + "grad_norm": 0.9321720600128174, + "learning_rate": 0.00017915655838666913, + "loss": 0.0936, + "step": 8935 + }, + { + "epoch": 0.2094739124574389, + "grad_norm": 3.9679315090179443, + "learning_rate": 0.00017915205682909232, + "loss": 0.4047, + "step": 8936 + }, + { + "epoch": 0.2094973540322439, + "grad_norm": 1.4495940208435059, + "learning_rate": 0.00017914755484203427, + "loss": 0.2309, + "step": 8937 + }, + { + "epoch": 0.2095207956070489, + "grad_norm": 4.154691696166992, + "learning_rate": 0.00017914305242551932, + "loss": 0.5585, + "step": 8938 + }, + { + "epoch": 0.2095442371818539, + "grad_norm": 2.625340700149536, + "learning_rate": 0.00017913854957957197, + "loss": 1.5585, + "step": 8939 + }, + { + "epoch": 0.20956767875665888, + "grad_norm": 1.481631875038147, + "learning_rate": 0.0001791340463042166, + "loss": 0.1304, + "step": 8940 + }, + { + "epoch": 0.20959112033146388, + "grad_norm": 3.3292932510375977, + "learning_rate": 0.0001791295425994777, + "loss": 0.3737, + "step": 8941 + }, + { + "epoch": 0.20961456190626887, + "grad_norm": 2.4782626628875732, + "learning_rate": 0.00017912503846537968, + "loss": 1.1883, + "step": 8942 + }, + { + "epoch": 0.20963800348107386, + "grad_norm": 1.3281515836715698, + "learning_rate": 0.0001791205339019469, + "loss": 0.101, + "step": 8943 + }, + { + "epoch": 0.20966144505587886, + "grad_norm": 2.2462501525878906, + "learning_rate": 0.000179116028909204, + "loss": 0.2978, + "step": 8944 + }, + { + "epoch": 0.20968488663068385, + "grad_norm": 1.6572705507278442, + "learning_rate": 0.00017911152348717518, + "loss": 0.6507, + "step": 8945 + }, + { + "epoch": 0.20970832820548885, + "grad_norm": 0.5683172941207886, + "learning_rate": 0.0001791070176358851, + "loss": 0.0832, + "step": 8946 + }, + { + "epoch": 0.20973176978029384, + "grad_norm": 3.9532852172851562, + "learning_rate": 0.00017910251135535806, + "loss": 0.4561, + "step": 8947 + }, + { + "epoch": 0.20975521135509884, + "grad_norm": 3.436107635498047, + "learning_rate": 0.00017909800464561862, + "loss": 0.3517, + "step": 8948 + }, + { + "epoch": 0.20977865292990383, + "grad_norm": 5.2249226570129395, + "learning_rate": 0.00017909349750669116, + "loss": 0.6164, + "step": 8949 + }, + { + "epoch": 0.20980209450470882, + "grad_norm": 1.6799598932266235, + "learning_rate": 0.00017908898993860016, + "loss": 0.1549, + "step": 8950 + }, + { + "epoch": 0.20982553607951382, + "grad_norm": 4.303152561187744, + "learning_rate": 0.00017908448194137007, + "loss": 0.2525, + "step": 8951 + }, + { + "epoch": 0.2098489776543188, + "grad_norm": 1.1294702291488647, + "learning_rate": 0.00017907997351502536, + "loss": 0.0928, + "step": 8952 + }, + { + "epoch": 0.2098724192291238, + "grad_norm": 4.381962776184082, + "learning_rate": 0.00017907546465959048, + "loss": 0.4163, + "step": 8953 + }, + { + "epoch": 0.2098958608039288, + "grad_norm": 4.572017192840576, + "learning_rate": 0.00017907095537508992, + "loss": 0.3255, + "step": 8954 + }, + { + "epoch": 0.2099193023787338, + "grad_norm": 3.4385385513305664, + "learning_rate": 0.00017906644566154813, + "loss": 0.3912, + "step": 8955 + }, + { + "epoch": 0.2099427439535388, + "grad_norm": 1.5258647203445435, + "learning_rate": 0.00017906193551898959, + "loss": 0.1263, + "step": 8956 + }, + { + "epoch": 0.20996618552834379, + "grad_norm": 7.830618858337402, + "learning_rate": 0.00017905742494743875, + "loss": 0.517, + "step": 8957 + }, + { + "epoch": 0.20998962710314878, + "grad_norm": 3.431480646133423, + "learning_rate": 0.0001790529139469201, + "loss": 0.4181, + "step": 8958 + }, + { + "epoch": 0.21001306867795377, + "grad_norm": 4.156547546386719, + "learning_rate": 0.00017904840251745812, + "loss": 0.7024, + "step": 8959 + }, + { + "epoch": 0.21003651025275877, + "grad_norm": 3.2362639904022217, + "learning_rate": 0.00017904389065907727, + "loss": 0.4281, + "step": 8960 + }, + { + "epoch": 0.21005995182756376, + "grad_norm": 1.963738203048706, + "learning_rate": 0.0001790393783718021, + "loss": 0.2734, + "step": 8961 + }, + { + "epoch": 0.21008339340236878, + "grad_norm": 1.6459286212921143, + "learning_rate": 0.000179034865655657, + "loss": 0.1066, + "step": 8962 + }, + { + "epoch": 0.21010683497717378, + "grad_norm": 4.495154857635498, + "learning_rate": 0.0001790303525106665, + "loss": 0.557, + "step": 8963 + }, + { + "epoch": 0.21013027655197877, + "grad_norm": 3.784656047821045, + "learning_rate": 0.0001790258389368551, + "loss": 0.3707, + "step": 8964 + }, + { + "epoch": 0.21015371812678377, + "grad_norm": 3.7109100818634033, + "learning_rate": 0.00017902132493424723, + "loss": 0.4414, + "step": 8965 + }, + { + "epoch": 0.21017715970158876, + "grad_norm": 4.0255584716796875, + "learning_rate": 0.00017901681050286748, + "loss": 0.485, + "step": 8966 + }, + { + "epoch": 0.21020060127639376, + "grad_norm": 7.759304046630859, + "learning_rate": 0.00017901229564274026, + "loss": 0.8104, + "step": 8967 + }, + { + "epoch": 0.21022404285119875, + "grad_norm": 1.7843356132507324, + "learning_rate": 0.0001790077803538901, + "loss": 0.241, + "step": 8968 + }, + { + "epoch": 0.21024748442600374, + "grad_norm": 5.969215393066406, + "learning_rate": 0.00017900326463634155, + "loss": 0.891, + "step": 8969 + }, + { + "epoch": 0.21027092600080874, + "grad_norm": 2.7047133445739746, + "learning_rate": 0.00017899874849011904, + "loss": 0.201, + "step": 8970 + }, + { + "epoch": 0.21029436757561373, + "grad_norm": 4.221607208251953, + "learning_rate": 0.00017899423191524707, + "loss": 0.6809, + "step": 8971 + }, + { + "epoch": 0.21031780915041873, + "grad_norm": 3.004457950592041, + "learning_rate": 0.00017898971491175022, + "loss": 0.2609, + "step": 8972 + }, + { + "epoch": 0.21034125072522372, + "grad_norm": 4.764808177947998, + "learning_rate": 0.00017898519747965293, + "loss": 0.7186, + "step": 8973 + }, + { + "epoch": 0.21036469230002872, + "grad_norm": 3.938603162765503, + "learning_rate": 0.00017898067961897974, + "loss": 0.4232, + "step": 8974 + }, + { + "epoch": 0.2103881338748337, + "grad_norm": 2.3023879528045654, + "learning_rate": 0.00017897616132975518, + "loss": 0.1998, + "step": 8975 + }, + { + "epoch": 0.2104115754496387, + "grad_norm": 4.791967391967773, + "learning_rate": 0.00017897164261200372, + "loss": 0.6805, + "step": 8976 + }, + { + "epoch": 0.2104350170244437, + "grad_norm": 3.60532808303833, + "learning_rate": 0.00017896712346574992, + "loss": 0.55, + "step": 8977 + }, + { + "epoch": 0.2104584585992487, + "grad_norm": 4.5364460945129395, + "learning_rate": 0.00017896260389101829, + "loss": 0.5769, + "step": 8978 + }, + { + "epoch": 0.2104819001740537, + "grad_norm": 0.7355962991714478, + "learning_rate": 0.00017895808388783334, + "loss": 0.0894, + "step": 8979 + }, + { + "epoch": 0.21050534174885868, + "grad_norm": 2.0485103130340576, + "learning_rate": 0.00017895356345621965, + "loss": 0.2345, + "step": 8980 + }, + { + "epoch": 0.21052878332366368, + "grad_norm": 2.8547255992889404, + "learning_rate": 0.00017894904259620166, + "loss": 1.7381, + "step": 8981 + }, + { + "epoch": 0.21055222489846867, + "grad_norm": 2.9295876026153564, + "learning_rate": 0.00017894452130780394, + "loss": 0.3736, + "step": 8982 + }, + { + "epoch": 0.21057566647327366, + "grad_norm": 4.262102127075195, + "learning_rate": 0.00017893999959105105, + "loss": 0.5281, + "step": 8983 + }, + { + "epoch": 0.21059910804807866, + "grad_norm": 3.6850690841674805, + "learning_rate": 0.0001789354774459675, + "loss": 0.379, + "step": 8984 + }, + { + "epoch": 0.21062254962288365, + "grad_norm": 3.0037758350372314, + "learning_rate": 0.00017893095487257784, + "loss": 1.7461, + "step": 8985 + }, + { + "epoch": 0.21064599119768865, + "grad_norm": 3.5761477947235107, + "learning_rate": 0.00017892643187090656, + "loss": 0.5916, + "step": 8986 + }, + { + "epoch": 0.21066943277249367, + "grad_norm": 1.5747116804122925, + "learning_rate": 0.0001789219084409783, + "loss": 0.2017, + "step": 8987 + }, + { + "epoch": 0.21069287434729866, + "grad_norm": 2.9666965007781982, + "learning_rate": 0.0001789173845828175, + "loss": 0.4657, + "step": 8988 + }, + { + "epoch": 0.21071631592210366, + "grad_norm": 4.639419078826904, + "learning_rate": 0.00017891286029644875, + "loss": 0.5758, + "step": 8989 + }, + { + "epoch": 0.21073975749690865, + "grad_norm": 4.705573558807373, + "learning_rate": 0.0001789083355818966, + "loss": 0.6678, + "step": 8990 + }, + { + "epoch": 0.21076319907171365, + "grad_norm": 4.511171817779541, + "learning_rate": 0.00017890381043918566, + "loss": 0.7747, + "step": 8991 + }, + { + "epoch": 0.21078664064651864, + "grad_norm": 4.429278373718262, + "learning_rate": 0.00017889928486834037, + "loss": 0.4918, + "step": 8992 + }, + { + "epoch": 0.21081008222132364, + "grad_norm": 3.6780648231506348, + "learning_rate": 0.00017889475886938535, + "loss": 0.5004, + "step": 8993 + }, + { + "epoch": 0.21083352379612863, + "grad_norm": 3.1607813835144043, + "learning_rate": 0.00017889023244234515, + "loss": 0.3131, + "step": 8994 + }, + { + "epoch": 0.21085696537093362, + "grad_norm": 4.70522403717041, + "learning_rate": 0.00017888570558724433, + "loss": 0.554, + "step": 8995 + }, + { + "epoch": 0.21088040694573862, + "grad_norm": 3.1147522926330566, + "learning_rate": 0.0001788811783041075, + "loss": 0.3691, + "step": 8996 + }, + { + "epoch": 0.2109038485205436, + "grad_norm": 2.9702374935150146, + "learning_rate": 0.0001788766505929591, + "loss": 0.3206, + "step": 8997 + }, + { + "epoch": 0.2109272900953486, + "grad_norm": 3.8366503715515137, + "learning_rate": 0.00017887212245382383, + "loss": 0.5807, + "step": 8998 + }, + { + "epoch": 0.2109507316701536, + "grad_norm": 2.5594706535339355, + "learning_rate": 0.0001788675938867262, + "loss": 0.2529, + "step": 8999 + }, + { + "epoch": 0.2109741732449586, + "grad_norm": 3.2037675380706787, + "learning_rate": 0.0001788630648916908, + "loss": 0.3895, + "step": 9000 + }, + { + "epoch": 0.2109976148197636, + "grad_norm": 6.47538948059082, + "learning_rate": 0.00017885853546874216, + "loss": 0.7998, + "step": 9001 + }, + { + "epoch": 0.21102105639456858, + "grad_norm": 3.902883768081665, + "learning_rate": 0.0001788540056179049, + "loss": 0.544, + "step": 9002 + }, + { + "epoch": 0.21104449796937358, + "grad_norm": 2.565062999725342, + "learning_rate": 0.00017884947533920357, + "loss": 0.386, + "step": 9003 + }, + { + "epoch": 0.21106793954417857, + "grad_norm": 4.846789836883545, + "learning_rate": 0.0001788449446326628, + "loss": 0.5498, + "step": 9004 + }, + { + "epoch": 0.21109138111898357, + "grad_norm": 2.309720754623413, + "learning_rate": 0.00017884041349830714, + "loss": 0.2705, + "step": 9005 + }, + { + "epoch": 0.21111482269378856, + "grad_norm": 3.313608169555664, + "learning_rate": 0.00017883588193616116, + "loss": 0.4031, + "step": 9006 + }, + { + "epoch": 0.21113826426859356, + "grad_norm": 3.6421053409576416, + "learning_rate": 0.00017883134994624948, + "loss": 0.3713, + "step": 9007 + }, + { + "epoch": 0.21116170584339855, + "grad_norm": 1.7161030769348145, + "learning_rate": 0.00017882681752859667, + "loss": 0.1777, + "step": 9008 + }, + { + "epoch": 0.21118514741820354, + "grad_norm": 0.8155103921890259, + "learning_rate": 0.00017882228468322735, + "loss": 0.1054, + "step": 9009 + }, + { + "epoch": 0.21120858899300854, + "grad_norm": 3.057894229888916, + "learning_rate": 0.0001788177514101661, + "loss": 0.2671, + "step": 9010 + }, + { + "epoch": 0.21123203056781353, + "grad_norm": 3.173007011413574, + "learning_rate": 0.0001788132177094375, + "loss": 0.2796, + "step": 9011 + }, + { + "epoch": 0.21125547214261853, + "grad_norm": 1.8929650783538818, + "learning_rate": 0.00017880868358106615, + "loss": 1.4751, + "step": 9012 + }, + { + "epoch": 0.21127891371742355, + "grad_norm": 3.8979759216308594, + "learning_rate": 0.0001788041490250767, + "loss": 0.3518, + "step": 9013 + }, + { + "epoch": 0.21130235529222854, + "grad_norm": 3.1945583820343018, + "learning_rate": 0.0001787996140414937, + "loss": 0.2813, + "step": 9014 + }, + { + "epoch": 0.21132579686703354, + "grad_norm": 4.435130596160889, + "learning_rate": 0.0001787950786303418, + "loss": 0.612, + "step": 9015 + }, + { + "epoch": 0.21134923844183853, + "grad_norm": 2.7351737022399902, + "learning_rate": 0.00017879054279164555, + "loss": 0.3713, + "step": 9016 + }, + { + "epoch": 0.21137268001664353, + "grad_norm": 4.610549449920654, + "learning_rate": 0.00017878600652542962, + "loss": 2.6175, + "step": 9017 + }, + { + "epoch": 0.21139612159144852, + "grad_norm": 1.583490014076233, + "learning_rate": 0.0001787814698317186, + "loss": 0.1836, + "step": 9018 + }, + { + "epoch": 0.21141956316625352, + "grad_norm": 3.450472116470337, + "learning_rate": 0.00017877693271053713, + "loss": 0.331, + "step": 9019 + }, + { + "epoch": 0.2114430047410585, + "grad_norm": 6.417930603027344, + "learning_rate": 0.0001787723951619098, + "loss": 0.4416, + "step": 9020 + }, + { + "epoch": 0.2114664463158635, + "grad_norm": 1.5035609006881714, + "learning_rate": 0.00017876785718586122, + "loss": 0.1883, + "step": 9021 + }, + { + "epoch": 0.2114898878906685, + "grad_norm": 3.960949182510376, + "learning_rate": 0.00017876331878241604, + "loss": 0.5115, + "step": 9022 + }, + { + "epoch": 0.2115133294654735, + "grad_norm": 5.738574981689453, + "learning_rate": 0.00017875877995159888, + "loss": 0.6348, + "step": 9023 + }, + { + "epoch": 0.2115367710402785, + "grad_norm": 4.19630765914917, + "learning_rate": 0.00017875424069343436, + "loss": 0.48, + "step": 9024 + }, + { + "epoch": 0.21156021261508348, + "grad_norm": 3.216885566711426, + "learning_rate": 0.00017874970100794713, + "loss": 0.2289, + "step": 9025 + }, + { + "epoch": 0.21158365418988848, + "grad_norm": 3.4761271476745605, + "learning_rate": 0.0001787451608951618, + "loss": 0.2602, + "step": 9026 + }, + { + "epoch": 0.21160709576469347, + "grad_norm": 2.531904458999634, + "learning_rate": 0.00017874062035510302, + "loss": 0.3832, + "step": 9027 + }, + { + "epoch": 0.21163053733949846, + "grad_norm": 2.6228294372558594, + "learning_rate": 0.00017873607938779541, + "loss": 0.2976, + "step": 9028 + }, + { + "epoch": 0.21165397891430346, + "grad_norm": 1.3822706937789917, + "learning_rate": 0.0001787315379932636, + "loss": 0.1169, + "step": 9029 + }, + { + "epoch": 0.21167742048910845, + "grad_norm": 1.2178705930709839, + "learning_rate": 0.0001787269961715323, + "loss": 0.19, + "step": 9030 + }, + { + "epoch": 0.21170086206391345, + "grad_norm": 1.7436885833740234, + "learning_rate": 0.00017872245392262608, + "loss": 0.2106, + "step": 9031 + }, + { + "epoch": 0.21172430363871844, + "grad_norm": 1.681545615196228, + "learning_rate": 0.0001787179112465696, + "loss": 0.1802, + "step": 9032 + }, + { + "epoch": 0.21174774521352344, + "grad_norm": 3.7398860454559326, + "learning_rate": 0.00017871336814338751, + "loss": 0.4622, + "step": 9033 + }, + { + "epoch": 0.21177118678832843, + "grad_norm": 3.1594715118408203, + "learning_rate": 0.00017870882461310452, + "loss": 0.3865, + "step": 9034 + }, + { + "epoch": 0.21179462836313342, + "grad_norm": 1.4852524995803833, + "learning_rate": 0.00017870428065574521, + "loss": 0.186, + "step": 9035 + }, + { + "epoch": 0.21181806993793842, + "grad_norm": 3.1666736602783203, + "learning_rate": 0.00017869973627133426, + "loss": 0.5041, + "step": 9036 + }, + { + "epoch": 0.2118415115127434, + "grad_norm": 4.245513916015625, + "learning_rate": 0.0001786951914598963, + "loss": 0.4452, + "step": 9037 + }, + { + "epoch": 0.2118649530875484, + "grad_norm": 4.495754241943359, + "learning_rate": 0.00017869064622145605, + "loss": 0.4332, + "step": 9038 + }, + { + "epoch": 0.21188839466235343, + "grad_norm": 3.1068179607391357, + "learning_rate": 0.00017868610055603814, + "loss": 0.377, + "step": 9039 + }, + { + "epoch": 0.21191183623715842, + "grad_norm": 4.397881984710693, + "learning_rate": 0.0001786815544636672, + "loss": 0.3154, + "step": 9040 + }, + { + "epoch": 0.21193527781196342, + "grad_norm": 3.902369737625122, + "learning_rate": 0.00017867700794436798, + "loss": 0.3164, + "step": 9041 + }, + { + "epoch": 0.2119587193867684, + "grad_norm": 2.01408314704895, + "learning_rate": 0.0001786724609981651, + "loss": 0.1638, + "step": 9042 + }, + { + "epoch": 0.2119821609615734, + "grad_norm": 2.5986626148223877, + "learning_rate": 0.00017866791362508322, + "loss": 0.3369, + "step": 9043 + }, + { + "epoch": 0.2120056025363784, + "grad_norm": 1.9559270143508911, + "learning_rate": 0.00017866336582514702, + "loss": 0.2708, + "step": 9044 + }, + { + "epoch": 0.2120290441111834, + "grad_norm": 2.213836193084717, + "learning_rate": 0.00017865881759838122, + "loss": 0.2228, + "step": 9045 + }, + { + "epoch": 0.2120524856859884, + "grad_norm": 2.3350632190704346, + "learning_rate": 0.00017865426894481042, + "loss": 0.3286, + "step": 9046 + }, + { + "epoch": 0.21207592726079338, + "grad_norm": 4.223564624786377, + "learning_rate": 0.00017864971986445935, + "loss": 0.6047, + "step": 9047 + }, + { + "epoch": 0.21209936883559838, + "grad_norm": 2.89571475982666, + "learning_rate": 0.00017864517035735272, + "loss": 1.5928, + "step": 9048 + }, + { + "epoch": 0.21212281041040337, + "grad_norm": 1.171129822731018, + "learning_rate": 0.00017864062042351515, + "loss": 0.1387, + "step": 9049 + }, + { + "epoch": 0.21214625198520837, + "grad_norm": 2.846914768218994, + "learning_rate": 0.0001786360700629714, + "loss": 0.3399, + "step": 9050 + }, + { + "epoch": 0.21216969356001336, + "grad_norm": 5.1941423416137695, + "learning_rate": 0.00017863151927574608, + "loss": 0.7891, + "step": 9051 + }, + { + "epoch": 0.21219313513481836, + "grad_norm": 6.65235710144043, + "learning_rate": 0.00017862696806186397, + "loss": 1.0608, + "step": 9052 + }, + { + "epoch": 0.21221657670962335, + "grad_norm": 2.113300085067749, + "learning_rate": 0.00017862241642134969, + "loss": 0.1767, + "step": 9053 + }, + { + "epoch": 0.21224001828442834, + "grad_norm": 1.3074705600738525, + "learning_rate": 0.00017861786435422795, + "loss": 0.0841, + "step": 9054 + }, + { + "epoch": 0.21226345985923334, + "grad_norm": 2.7034571170806885, + "learning_rate": 0.0001786133118605235, + "loss": 0.3073, + "step": 9055 + }, + { + "epoch": 0.21228690143403833, + "grad_norm": 5.193404197692871, + "learning_rate": 0.000178608758940261, + "loss": 0.6354, + "step": 9056 + }, + { + "epoch": 0.21231034300884333, + "grad_norm": 1.830773949623108, + "learning_rate": 0.00017860420559346517, + "loss": 0.1746, + "step": 9057 + }, + { + "epoch": 0.21233378458364832, + "grad_norm": 4.237575054168701, + "learning_rate": 0.00017859965182016071, + "loss": 0.8722, + "step": 9058 + }, + { + "epoch": 0.21235722615845332, + "grad_norm": 1.0532681941986084, + "learning_rate": 0.00017859509762037235, + "loss": 0.1246, + "step": 9059 + }, + { + "epoch": 0.2123806677332583, + "grad_norm": 4.517276287078857, + "learning_rate": 0.00017859054299412472, + "loss": 0.3057, + "step": 9060 + }, + { + "epoch": 0.2124041093080633, + "grad_norm": 4.009767532348633, + "learning_rate": 0.00017858598794144267, + "loss": 0.6594, + "step": 9061 + }, + { + "epoch": 0.2124275508828683, + "grad_norm": 4.219710350036621, + "learning_rate": 0.00017858143246235076, + "loss": 0.4077, + "step": 9062 + }, + { + "epoch": 0.2124509924576733, + "grad_norm": 2.5491533279418945, + "learning_rate": 0.00017857687655687383, + "loss": 1.531, + "step": 9063 + }, + { + "epoch": 0.21247443403247832, + "grad_norm": 4.1400275230407715, + "learning_rate": 0.0001785723202250366, + "loss": 0.5252, + "step": 9064 + }, + { + "epoch": 0.2124978756072833, + "grad_norm": 5.3289384841918945, + "learning_rate": 0.00017856776346686368, + "loss": 0.7673, + "step": 9065 + }, + { + "epoch": 0.2125213171820883, + "grad_norm": 4.3242316246032715, + "learning_rate": 0.00017856320628237987, + "loss": 0.2845, + "step": 9066 + }, + { + "epoch": 0.2125447587568933, + "grad_norm": 4.714540958404541, + "learning_rate": 0.00017855864867160994, + "loss": 0.6693, + "step": 9067 + }, + { + "epoch": 0.2125682003316983, + "grad_norm": 2.5945258140563965, + "learning_rate": 0.00017855409063457852, + "loss": 1.6407, + "step": 9068 + }, + { + "epoch": 0.2125916419065033, + "grad_norm": 3.8835031986236572, + "learning_rate": 0.00017854953217131045, + "loss": 0.7762, + "step": 9069 + }, + { + "epoch": 0.21261508348130828, + "grad_norm": 4.7618184089660645, + "learning_rate": 0.00017854497328183036, + "loss": 0.6257, + "step": 9070 + }, + { + "epoch": 0.21263852505611328, + "grad_norm": 6.451463222503662, + "learning_rate": 0.0001785404139661631, + "loss": 0.6146, + "step": 9071 + }, + { + "epoch": 0.21266196663091827, + "grad_norm": 0.898201048374176, + "learning_rate": 0.0001785358542243333, + "loss": 0.0963, + "step": 9072 + }, + { + "epoch": 0.21268540820572326, + "grad_norm": 2.816620349884033, + "learning_rate": 0.00017853129405636574, + "loss": 1.9253, + "step": 9073 + }, + { + "epoch": 0.21270884978052826, + "grad_norm": 4.353067874908447, + "learning_rate": 0.0001785267334622852, + "loss": 0.4954, + "step": 9074 + }, + { + "epoch": 0.21273229135533325, + "grad_norm": 3.5822410583496094, + "learning_rate": 0.0001785221724421164, + "loss": 0.3572, + "step": 9075 + }, + { + "epoch": 0.21275573293013825, + "grad_norm": 1.3176385164260864, + "learning_rate": 0.00017851761099588404, + "loss": 0.1934, + "step": 9076 + }, + { + "epoch": 0.21277917450494324, + "grad_norm": 0.9711645841598511, + "learning_rate": 0.00017851304912361296, + "loss": 0.1209, + "step": 9077 + }, + { + "epoch": 0.21280261607974824, + "grad_norm": 0.3500994145870209, + "learning_rate": 0.00017850848682532785, + "loss": 0.0251, + "step": 9078 + }, + { + "epoch": 0.21282605765455323, + "grad_norm": 4.153914928436279, + "learning_rate": 0.00017850392410105346, + "loss": 0.7845, + "step": 9079 + }, + { + "epoch": 0.21284949922935822, + "grad_norm": 1.9686939716339111, + "learning_rate": 0.00017849936095081458, + "loss": 0.1442, + "step": 9080 + }, + { + "epoch": 0.21287294080416322, + "grad_norm": 2.2252941131591797, + "learning_rate": 0.00017849479737463598, + "loss": 0.2925, + "step": 9081 + }, + { + "epoch": 0.2128963823789682, + "grad_norm": 5.092584609985352, + "learning_rate": 0.0001784902333725424, + "loss": 0.7865, + "step": 9082 + }, + { + "epoch": 0.2129198239537732, + "grad_norm": 5.456280708312988, + "learning_rate": 0.0001784856689445586, + "loss": 0.7438, + "step": 9083 + }, + { + "epoch": 0.2129432655285782, + "grad_norm": 3.8680622577667236, + "learning_rate": 0.00017848110409070936, + "loss": 0.5947, + "step": 9084 + }, + { + "epoch": 0.2129667071033832, + "grad_norm": 2.30660080909729, + "learning_rate": 0.00017847653881101943, + "loss": 1.5918, + "step": 9085 + }, + { + "epoch": 0.2129901486781882, + "grad_norm": 4.649816989898682, + "learning_rate": 0.0001784719731055136, + "loss": 0.4892, + "step": 9086 + }, + { + "epoch": 0.21301359025299318, + "grad_norm": 0.572515070438385, + "learning_rate": 0.00017846740697421664, + "loss": 0.0415, + "step": 9087 + }, + { + "epoch": 0.21303703182779818, + "grad_norm": 6.115943908691406, + "learning_rate": 0.00017846284041715332, + "loss": 0.6581, + "step": 9088 + }, + { + "epoch": 0.21306047340260317, + "grad_norm": 2.312655210494995, + "learning_rate": 0.00017845827343434842, + "loss": 1.7405, + "step": 9089 + }, + { + "epoch": 0.2130839149774082, + "grad_norm": 0.32670825719833374, + "learning_rate": 0.00017845370602582677, + "loss": 0.0197, + "step": 9090 + }, + { + "epoch": 0.2131073565522132, + "grad_norm": 4.475550651550293, + "learning_rate": 0.00017844913819161303, + "loss": 0.6232, + "step": 9091 + }, + { + "epoch": 0.21313079812701818, + "grad_norm": 1.0810558795928955, + "learning_rate": 0.00017844456993173209, + "loss": 0.1177, + "step": 9092 + }, + { + "epoch": 0.21315423970182318, + "grad_norm": 0.8899589776992798, + "learning_rate": 0.0001784400012462087, + "loss": 0.0817, + "step": 9093 + }, + { + "epoch": 0.21317768127662817, + "grad_norm": 3.2798686027526855, + "learning_rate": 0.00017843543213506767, + "loss": 0.2583, + "step": 9094 + }, + { + "epoch": 0.21320112285143317, + "grad_norm": 3.21588397026062, + "learning_rate": 0.00017843086259833378, + "loss": 0.3677, + "step": 9095 + }, + { + "epoch": 0.21322456442623816, + "grad_norm": 4.57167387008667, + "learning_rate": 0.00017842629263603184, + "loss": 0.5557, + "step": 9096 + }, + { + "epoch": 0.21324800600104316, + "grad_norm": 3.596109390258789, + "learning_rate": 0.00017842172224818658, + "loss": 0.3648, + "step": 9097 + }, + { + "epoch": 0.21327144757584815, + "grad_norm": 1.6251286268234253, + "learning_rate": 0.0001784171514348229, + "loss": 0.2692, + "step": 9098 + }, + { + "epoch": 0.21329488915065314, + "grad_norm": 5.037043571472168, + "learning_rate": 0.00017841258019596554, + "loss": 0.6375, + "step": 9099 + }, + { + "epoch": 0.21331833072545814, + "grad_norm": 4.232018947601318, + "learning_rate": 0.0001784080085316393, + "loss": 0.5603, + "step": 9100 + }, + { + "epoch": 0.21334177230026313, + "grad_norm": 3.103832960128784, + "learning_rate": 0.00017840343644186902, + "loss": 0.2614, + "step": 9101 + }, + { + "epoch": 0.21336521387506813, + "grad_norm": 2.9992752075195312, + "learning_rate": 0.00017839886392667949, + "loss": 0.3459, + "step": 9102 + }, + { + "epoch": 0.21338865544987312, + "grad_norm": 2.5085415840148926, + "learning_rate": 0.00017839429098609548, + "loss": 0.3375, + "step": 9103 + }, + { + "epoch": 0.21341209702467812, + "grad_norm": 3.3194146156311035, + "learning_rate": 0.0001783897176201419, + "loss": 0.326, + "step": 9104 + }, + { + "epoch": 0.2134355385994831, + "grad_norm": 2.93380069732666, + "learning_rate": 0.00017838514382884348, + "loss": 0.1921, + "step": 9105 + }, + { + "epoch": 0.2134589801742881, + "grad_norm": 3.3455207347869873, + "learning_rate": 0.00017838056961222507, + "loss": 0.6752, + "step": 9106 + }, + { + "epoch": 0.2134824217490931, + "grad_norm": 1.35897958278656, + "learning_rate": 0.00017837599497031148, + "loss": 0.1889, + "step": 9107 + }, + { + "epoch": 0.2135058633238981, + "grad_norm": 4.5258989334106445, + "learning_rate": 0.00017837141990312754, + "loss": 0.5624, + "step": 9108 + }, + { + "epoch": 0.2135293048987031, + "grad_norm": 3.788527250289917, + "learning_rate": 0.00017836684441069805, + "loss": 0.3554, + "step": 9109 + }, + { + "epoch": 0.21355274647350808, + "grad_norm": 1.7670706510543823, + "learning_rate": 0.0001783622684930479, + "loss": 0.2336, + "step": 9110 + }, + { + "epoch": 0.21357618804831308, + "grad_norm": 3.849514961242676, + "learning_rate": 0.00017835769215020185, + "loss": 0.5603, + "step": 9111 + }, + { + "epoch": 0.21359962962311807, + "grad_norm": 2.4763598442077637, + "learning_rate": 0.00017835311538218475, + "loss": 0.3682, + "step": 9112 + }, + { + "epoch": 0.21362307119792306, + "grad_norm": 2.9950459003448486, + "learning_rate": 0.00017834853818902147, + "loss": 0.4516, + "step": 9113 + }, + { + "epoch": 0.21364651277272806, + "grad_norm": 3.2573001384735107, + "learning_rate": 0.0001783439605707368, + "loss": 0.4624, + "step": 9114 + }, + { + "epoch": 0.21366995434753308, + "grad_norm": 1.0973484516143799, + "learning_rate": 0.0001783393825273556, + "loss": 0.1378, + "step": 9115 + }, + { + "epoch": 0.21369339592233808, + "grad_norm": 1.2658371925354004, + "learning_rate": 0.00017833480405890273, + "loss": 0.1678, + "step": 9116 + }, + { + "epoch": 0.21371683749714307, + "grad_norm": 4.674359321594238, + "learning_rate": 0.00017833022516540301, + "loss": 0.6454, + "step": 9117 + }, + { + "epoch": 0.21374027907194806, + "grad_norm": 1.212883472442627, + "learning_rate": 0.00017832564584688125, + "loss": 0.1336, + "step": 9118 + }, + { + "epoch": 0.21376372064675306, + "grad_norm": 4.106048583984375, + "learning_rate": 0.00017832106610336234, + "loss": 0.3587, + "step": 9119 + }, + { + "epoch": 0.21378716222155805, + "grad_norm": 3.6229472160339355, + "learning_rate": 0.00017831648593487115, + "loss": 0.6352, + "step": 9120 + }, + { + "epoch": 0.21381060379636305, + "grad_norm": 3.6432082653045654, + "learning_rate": 0.00017831190534143247, + "loss": 0.4821, + "step": 9121 + }, + { + "epoch": 0.21383404537116804, + "grad_norm": 0.9357970356941223, + "learning_rate": 0.00017830732432307122, + "loss": 0.1137, + "step": 9122 + }, + { + "epoch": 0.21385748694597304, + "grad_norm": 4.173257827758789, + "learning_rate": 0.00017830274287981225, + "loss": 0.5663, + "step": 9123 + }, + { + "epoch": 0.21388092852077803, + "grad_norm": 1.7975343465805054, + "learning_rate": 0.00017829816101168033, + "loss": 0.2724, + "step": 9124 + }, + { + "epoch": 0.21390437009558302, + "grad_norm": 0.6228080987930298, + "learning_rate": 0.00017829357871870043, + "loss": 0.1137, + "step": 9125 + }, + { + "epoch": 0.21392781167038802, + "grad_norm": 2.587043285369873, + "learning_rate": 0.00017828899600089737, + "loss": 0.2282, + "step": 9126 + }, + { + "epoch": 0.213951253245193, + "grad_norm": 3.32633900642395, + "learning_rate": 0.000178284412858296, + "loss": 0.2821, + "step": 9127 + }, + { + "epoch": 0.213974694819998, + "grad_norm": 1.2827000617980957, + "learning_rate": 0.00017827982929092122, + "loss": 0.214, + "step": 9128 + }, + { + "epoch": 0.213998136394803, + "grad_norm": 4.7786760330200195, + "learning_rate": 0.00017827524529879786, + "loss": 0.6731, + "step": 9129 + }, + { + "epoch": 0.214021577969608, + "grad_norm": 1.7058035135269165, + "learning_rate": 0.00017827066088195085, + "loss": 0.2076, + "step": 9130 + }, + { + "epoch": 0.214045019544413, + "grad_norm": 3.8251736164093018, + "learning_rate": 0.000178266076040405, + "loss": 0.4496, + "step": 9131 + }, + { + "epoch": 0.21406846111921798, + "grad_norm": 2.1928677558898926, + "learning_rate": 0.00017826149077418527, + "loss": 0.3701, + "step": 9132 + }, + { + "epoch": 0.21409190269402298, + "grad_norm": 0.8303082585334778, + "learning_rate": 0.00017825690508331644, + "loss": 0.1482, + "step": 9133 + }, + { + "epoch": 0.21411534426882797, + "grad_norm": 3.7330586910247803, + "learning_rate": 0.00017825231896782343, + "loss": 0.36, + "step": 9134 + }, + { + "epoch": 0.21413878584363297, + "grad_norm": 4.203818321228027, + "learning_rate": 0.00017824773242773118, + "loss": 0.5005, + "step": 9135 + }, + { + "epoch": 0.21416222741843796, + "grad_norm": 3.0732200145721436, + "learning_rate": 0.00017824314546306452, + "loss": 0.4696, + "step": 9136 + }, + { + "epoch": 0.21418566899324296, + "grad_norm": 4.252399921417236, + "learning_rate": 0.00017823855807384836, + "loss": 0.6825, + "step": 9137 + }, + { + "epoch": 0.21420911056804795, + "grad_norm": 1.9473857879638672, + "learning_rate": 0.0001782339702601076, + "loss": 0.1909, + "step": 9138 + }, + { + "epoch": 0.21423255214285294, + "grad_norm": 4.6035895347595215, + "learning_rate": 0.0001782293820218671, + "loss": 0.4929, + "step": 9139 + }, + { + "epoch": 0.21425599371765794, + "grad_norm": 3.371150255203247, + "learning_rate": 0.00017822479335915175, + "loss": 0.4753, + "step": 9140 + }, + { + "epoch": 0.21427943529246296, + "grad_norm": 2.6262693405151367, + "learning_rate": 0.0001782202042719865, + "loss": 0.2555, + "step": 9141 + }, + { + "epoch": 0.21430287686726796, + "grad_norm": 3.4853081703186035, + "learning_rate": 0.00017821561476039622, + "loss": 0.3059, + "step": 9142 + }, + { + "epoch": 0.21432631844207295, + "grad_norm": 1.4359065294265747, + "learning_rate": 0.0001782110248244058, + "loss": 0.1886, + "step": 9143 + }, + { + "epoch": 0.21434976001687794, + "grad_norm": 1.7953599691390991, + "learning_rate": 0.00017820643446404016, + "loss": 0.3427, + "step": 9144 + }, + { + "epoch": 0.21437320159168294, + "grad_norm": 3.3048512935638428, + "learning_rate": 0.00017820184367932422, + "loss": 0.3737, + "step": 9145 + }, + { + "epoch": 0.21439664316648793, + "grad_norm": 1.1847610473632812, + "learning_rate": 0.00017819725247028288, + "loss": 0.1606, + "step": 9146 + }, + { + "epoch": 0.21442008474129293, + "grad_norm": 4.976329326629639, + "learning_rate": 0.00017819266083694103, + "loss": 0.5561, + "step": 9147 + }, + { + "epoch": 0.21444352631609792, + "grad_norm": 3.4354677200317383, + "learning_rate": 0.00017818806877932365, + "loss": 0.5179, + "step": 9148 + }, + { + "epoch": 0.21446696789090292, + "grad_norm": 4.209543228149414, + "learning_rate": 0.00017818347629745555, + "loss": 0.5597, + "step": 9149 + }, + { + "epoch": 0.2144904094657079, + "grad_norm": 1.6484007835388184, + "learning_rate": 0.00017817888339136174, + "loss": 0.2272, + "step": 9150 + }, + { + "epoch": 0.2145138510405129, + "grad_norm": 0.9880679845809937, + "learning_rate": 0.00017817429006106712, + "loss": 0.153, + "step": 9151 + }, + { + "epoch": 0.2145372926153179, + "grad_norm": 2.796154737472534, + "learning_rate": 0.00017816969630659658, + "loss": 0.3984, + "step": 9152 + }, + { + "epoch": 0.2145607341901229, + "grad_norm": 2.607649803161621, + "learning_rate": 0.00017816510212797512, + "loss": 1.9048, + "step": 9153 + }, + { + "epoch": 0.2145841757649279, + "grad_norm": 2.4985501766204834, + "learning_rate": 0.00017816050752522756, + "loss": 0.2672, + "step": 9154 + }, + { + "epoch": 0.21460761733973288, + "grad_norm": 2.9178121089935303, + "learning_rate": 0.00017815591249837888, + "loss": 0.1558, + "step": 9155 + }, + { + "epoch": 0.21463105891453788, + "grad_norm": 3.902846336364746, + "learning_rate": 0.00017815131704745408, + "loss": 0.3635, + "step": 9156 + }, + { + "epoch": 0.21465450048934287, + "grad_norm": 4.197511672973633, + "learning_rate": 0.000178146721172478, + "loss": 0.4569, + "step": 9157 + }, + { + "epoch": 0.21467794206414786, + "grad_norm": 3.5203959941864014, + "learning_rate": 0.0001781421248734756, + "loss": 0.5118, + "step": 9158 + }, + { + "epoch": 0.21470138363895286, + "grad_norm": 3.1451237201690674, + "learning_rate": 0.00017813752815047188, + "loss": 0.3199, + "step": 9159 + }, + { + "epoch": 0.21472482521375785, + "grad_norm": 3.6452229022979736, + "learning_rate": 0.00017813293100349167, + "loss": 0.4329, + "step": 9160 + }, + { + "epoch": 0.21474826678856285, + "grad_norm": 1.9847536087036133, + "learning_rate": 0.00017812833343256004, + "loss": 0.2874, + "step": 9161 + }, + { + "epoch": 0.21477170836336784, + "grad_norm": 4.114924907684326, + "learning_rate": 0.00017812373543770184, + "loss": 0.4738, + "step": 9162 + }, + { + "epoch": 0.21479514993817284, + "grad_norm": 3.585874319076538, + "learning_rate": 0.00017811913701894208, + "loss": 0.3831, + "step": 9163 + }, + { + "epoch": 0.21481859151297783, + "grad_norm": 2.6608734130859375, + "learning_rate": 0.00017811453817630568, + "loss": 0.2093, + "step": 9164 + }, + { + "epoch": 0.21484203308778282, + "grad_norm": 3.7637505531311035, + "learning_rate": 0.00017810993890981757, + "loss": 0.3329, + "step": 9165 + }, + { + "epoch": 0.21486547466258785, + "grad_norm": 2.8130128383636475, + "learning_rate": 0.0001781053392195028, + "loss": 0.1946, + "step": 9166 + }, + { + "epoch": 0.21488891623739284, + "grad_norm": 2.2565195560455322, + "learning_rate": 0.0001781007391053862, + "loss": 0.3697, + "step": 9167 + }, + { + "epoch": 0.21491235781219784, + "grad_norm": 3.3409323692321777, + "learning_rate": 0.0001780961385674928, + "loss": 0.3854, + "step": 9168 + }, + { + "epoch": 0.21493579938700283, + "grad_norm": 3.0096049308776855, + "learning_rate": 0.00017809153760584757, + "loss": 1.7659, + "step": 9169 + }, + { + "epoch": 0.21495924096180782, + "grad_norm": 5.432003498077393, + "learning_rate": 0.00017808693622047548, + "loss": 0.8516, + "step": 9170 + }, + { + "epoch": 0.21498268253661282, + "grad_norm": 3.2203893661499023, + "learning_rate": 0.00017808233441140146, + "loss": 0.4606, + "step": 9171 + }, + { + "epoch": 0.2150061241114178, + "grad_norm": 5.538031101226807, + "learning_rate": 0.0001780777321786505, + "loss": 0.379, + "step": 9172 + }, + { + "epoch": 0.2150295656862228, + "grad_norm": 1.6890509128570557, + "learning_rate": 0.00017807312952224758, + "loss": 0.186, + "step": 9173 + }, + { + "epoch": 0.2150530072610278, + "grad_norm": 2.9729154109954834, + "learning_rate": 0.00017806852644221765, + "loss": 0.439, + "step": 9174 + }, + { + "epoch": 0.2150764488358328, + "grad_norm": 3.4990668296813965, + "learning_rate": 0.0001780639229385857, + "loss": 0.4288, + "step": 9175 + }, + { + "epoch": 0.2150998904106378, + "grad_norm": 3.585759162902832, + "learning_rate": 0.00017805931901137672, + "loss": 0.5022, + "step": 9176 + }, + { + "epoch": 0.21512333198544278, + "grad_norm": 1.2016675472259521, + "learning_rate": 0.00017805471466061568, + "loss": 0.1228, + "step": 9177 + }, + { + "epoch": 0.21514677356024778, + "grad_norm": 2.2827565670013428, + "learning_rate": 0.00017805010988632757, + "loss": 0.1855, + "step": 9178 + }, + { + "epoch": 0.21517021513505277, + "grad_norm": 1.2853962182998657, + "learning_rate": 0.00017804550468853735, + "loss": 0.2119, + "step": 9179 + }, + { + "epoch": 0.21519365670985777, + "grad_norm": 3.0210883617401123, + "learning_rate": 0.00017804089906727, + "loss": 0.4761, + "step": 9180 + }, + { + "epoch": 0.21521709828466276, + "grad_norm": 1.0182342529296875, + "learning_rate": 0.00017803629302255057, + "loss": 0.08, + "step": 9181 + }, + { + "epoch": 0.21524053985946776, + "grad_norm": 0.8325868844985962, + "learning_rate": 0.00017803168655440403, + "loss": 0.068, + "step": 9182 + }, + { + "epoch": 0.21526398143427275, + "grad_norm": 1.767912745475769, + "learning_rate": 0.00017802707966285537, + "loss": 0.1441, + "step": 9183 + }, + { + "epoch": 0.21528742300907774, + "grad_norm": 3.599644660949707, + "learning_rate": 0.00017802247234792958, + "loss": 0.5197, + "step": 9184 + }, + { + "epoch": 0.21531086458388274, + "grad_norm": 2.531155824661255, + "learning_rate": 0.00017801786460965163, + "loss": 1.3368, + "step": 9185 + }, + { + "epoch": 0.21533430615868773, + "grad_norm": 4.795960903167725, + "learning_rate": 0.00017801325644804656, + "loss": 0.4383, + "step": 9186 + }, + { + "epoch": 0.21535774773349273, + "grad_norm": 3.600813865661621, + "learning_rate": 0.00017800864786313936, + "loss": 0.2502, + "step": 9187 + }, + { + "epoch": 0.21538118930829772, + "grad_norm": 3.5864861011505127, + "learning_rate": 0.00017800403885495509, + "loss": 0.3662, + "step": 9188 + }, + { + "epoch": 0.21540463088310272, + "grad_norm": 6.596198081970215, + "learning_rate": 0.0001779994294235187, + "loss": 0.6549, + "step": 9189 + }, + { + "epoch": 0.2154280724579077, + "grad_norm": 4.457853317260742, + "learning_rate": 0.00017799481956885517, + "loss": 0.433, + "step": 9190 + }, + { + "epoch": 0.2154515140327127, + "grad_norm": 2.1976187229156494, + "learning_rate": 0.00017799020929098956, + "loss": 1.533, + "step": 9191 + }, + { + "epoch": 0.21547495560751773, + "grad_norm": 3.4269793033599854, + "learning_rate": 0.00017798559858994692, + "loss": 0.4151, + "step": 9192 + }, + { + "epoch": 0.21549839718232272, + "grad_norm": 3.1305572986602783, + "learning_rate": 0.00017798098746575223, + "loss": 0.3167, + "step": 9193 + }, + { + "epoch": 0.21552183875712771, + "grad_norm": 4.012729167938232, + "learning_rate": 0.0001779763759184305, + "loss": 0.4718, + "step": 9194 + }, + { + "epoch": 0.2155452803319327, + "grad_norm": 3.583699941635132, + "learning_rate": 0.00017797176394800674, + "loss": 0.5187, + "step": 9195 + }, + { + "epoch": 0.2155687219067377, + "grad_norm": 2.7406842708587646, + "learning_rate": 0.00017796715155450602, + "loss": 0.3627, + "step": 9196 + }, + { + "epoch": 0.2155921634815427, + "grad_norm": 0.916742742061615, + "learning_rate": 0.00017796253873795332, + "loss": 0.0727, + "step": 9197 + }, + { + "epoch": 0.2156156050563477, + "grad_norm": 3.431245803833008, + "learning_rate": 0.0001779579254983737, + "loss": 0.2118, + "step": 9198 + }, + { + "epoch": 0.2156390466311527, + "grad_norm": 1.39991295337677, + "learning_rate": 0.0001779533118357922, + "loss": 0.2158, + "step": 9199 + }, + { + "epoch": 0.21566248820595768, + "grad_norm": 4.002741813659668, + "learning_rate": 0.00017794869775023382, + "loss": 0.3619, + "step": 9200 + }, + { + "epoch": 0.21568592978076268, + "grad_norm": 3.7973296642303467, + "learning_rate": 0.00017794408324172363, + "loss": 0.4277, + "step": 9201 + }, + { + "epoch": 0.21570937135556767, + "grad_norm": 3.786595106124878, + "learning_rate": 0.00017793946831028666, + "loss": 0.275, + "step": 9202 + }, + { + "epoch": 0.21573281293037266, + "grad_norm": 2.739191770553589, + "learning_rate": 0.0001779348529559479, + "loss": 0.3502, + "step": 9203 + }, + { + "epoch": 0.21575625450517766, + "grad_norm": 3.837468385696411, + "learning_rate": 0.0001779302371787325, + "loss": 0.4699, + "step": 9204 + }, + { + "epoch": 0.21577969607998265, + "grad_norm": 6.287978649139404, + "learning_rate": 0.0001779256209786654, + "loss": 0.9013, + "step": 9205 + }, + { + "epoch": 0.21580313765478765, + "grad_norm": 3.6506261825561523, + "learning_rate": 0.0001779210043557717, + "loss": 0.3351, + "step": 9206 + }, + { + "epoch": 0.21582657922959264, + "grad_norm": 4.549388885498047, + "learning_rate": 0.00017791638731007642, + "loss": 0.4953, + "step": 9207 + }, + { + "epoch": 0.21585002080439764, + "grad_norm": 1.9085530042648315, + "learning_rate": 0.00017791176984160466, + "loss": 0.1428, + "step": 9208 + }, + { + "epoch": 0.21587346237920263, + "grad_norm": 5.8407440185546875, + "learning_rate": 0.00017790715195038144, + "loss": 0.5125, + "step": 9209 + }, + { + "epoch": 0.21589690395400762, + "grad_norm": 3.992950201034546, + "learning_rate": 0.0001779025336364318, + "loss": 0.4841, + "step": 9210 + }, + { + "epoch": 0.21592034552881262, + "grad_norm": 3.5131561756134033, + "learning_rate": 0.00017789791489978085, + "loss": 0.4616, + "step": 9211 + }, + { + "epoch": 0.2159437871036176, + "grad_norm": 3.104310989379883, + "learning_rate": 0.00017789329574045362, + "loss": 0.3805, + "step": 9212 + }, + { + "epoch": 0.2159672286784226, + "grad_norm": 3.107032537460327, + "learning_rate": 0.00017788867615847515, + "loss": 0.3082, + "step": 9213 + }, + { + "epoch": 0.2159906702532276, + "grad_norm": 1.7976548671722412, + "learning_rate": 0.00017788405615387057, + "loss": 0.1484, + "step": 9214 + }, + { + "epoch": 0.2160141118280326, + "grad_norm": 2.8517282009124756, + "learning_rate": 0.00017787943572666488, + "loss": 0.4011, + "step": 9215 + }, + { + "epoch": 0.2160375534028376, + "grad_norm": 2.6364758014678955, + "learning_rate": 0.0001778748148768832, + "loss": 0.2394, + "step": 9216 + }, + { + "epoch": 0.21606099497764258, + "grad_norm": 4.281954288482666, + "learning_rate": 0.0001778701936045506, + "loss": 0.4534, + "step": 9217 + }, + { + "epoch": 0.2160844365524476, + "grad_norm": 2.60758638381958, + "learning_rate": 0.0001778655719096921, + "loss": 1.9869, + "step": 9218 + }, + { + "epoch": 0.2161078781272526, + "grad_norm": 5.265790939331055, + "learning_rate": 0.00017786094979233284, + "loss": 0.5427, + "step": 9219 + }, + { + "epoch": 0.2161313197020576, + "grad_norm": 7.228806495666504, + "learning_rate": 0.00017785632725249789, + "loss": 0.6742, + "step": 9220 + }, + { + "epoch": 0.2161547612768626, + "grad_norm": 6.532603740692139, + "learning_rate": 0.0001778517042902123, + "loss": 0.5829, + "step": 9221 + }, + { + "epoch": 0.21617820285166758, + "grad_norm": 4.974170684814453, + "learning_rate": 0.0001778470809055012, + "loss": 0.6914, + "step": 9222 + }, + { + "epoch": 0.21620164442647258, + "grad_norm": 5.185410499572754, + "learning_rate": 0.0001778424570983896, + "loss": 0.4288, + "step": 9223 + }, + { + "epoch": 0.21622508600127757, + "grad_norm": 4.880720615386963, + "learning_rate": 0.00017783783286890267, + "loss": 0.647, + "step": 9224 + }, + { + "epoch": 0.21624852757608257, + "grad_norm": 2.4200544357299805, + "learning_rate": 0.00017783320821706546, + "loss": 0.3432, + "step": 9225 + }, + { + "epoch": 0.21627196915088756, + "grad_norm": 3.811389684677124, + "learning_rate": 0.0001778285831429031, + "loss": 1.5657, + "step": 9226 + }, + { + "epoch": 0.21629541072569256, + "grad_norm": 1.0748931169509888, + "learning_rate": 0.00017782395764644063, + "loss": 0.0944, + "step": 9227 + }, + { + "epoch": 0.21631885230049755, + "grad_norm": 0.7816444635391235, + "learning_rate": 0.0001778193317277032, + "loss": 0.1153, + "step": 9228 + }, + { + "epoch": 0.21634229387530254, + "grad_norm": 4.410451412200928, + "learning_rate": 0.00017781470538671588, + "loss": 0.3554, + "step": 9229 + }, + { + "epoch": 0.21636573545010754, + "grad_norm": 5.774175643920898, + "learning_rate": 0.0001778100786235038, + "loss": 0.8836, + "step": 9230 + }, + { + "epoch": 0.21638917702491253, + "grad_norm": 3.521592617034912, + "learning_rate": 0.000177805451438092, + "loss": 0.3267, + "step": 9231 + }, + { + "epoch": 0.21641261859971753, + "grad_norm": 0.790245771408081, + "learning_rate": 0.00017780082383050564, + "loss": 0.051, + "step": 9232 + }, + { + "epoch": 0.21643606017452252, + "grad_norm": 3.3017067909240723, + "learning_rate": 0.00017779619580076986, + "loss": 0.3766, + "step": 9233 + }, + { + "epoch": 0.21645950174932752, + "grad_norm": 3.699653387069702, + "learning_rate": 0.00017779156734890972, + "loss": 0.4228, + "step": 9234 + }, + { + "epoch": 0.2164829433241325, + "grad_norm": 4.5368733406066895, + "learning_rate": 0.00017778693847495032, + "loss": 0.6717, + "step": 9235 + }, + { + "epoch": 0.2165063848989375, + "grad_norm": 2.6501541137695312, + "learning_rate": 0.00017778230917891682, + "loss": 0.2149, + "step": 9236 + }, + { + "epoch": 0.2165298264737425, + "grad_norm": 5.4246416091918945, + "learning_rate": 0.00017777767946083433, + "loss": 0.657, + "step": 9237 + }, + { + "epoch": 0.2165532680485475, + "grad_norm": 4.976633548736572, + "learning_rate": 0.00017777304932072797, + "loss": 0.527, + "step": 9238 + }, + { + "epoch": 0.2165767096233525, + "grad_norm": 1.5317058563232422, + "learning_rate": 0.00017776841875862288, + "loss": 0.2107, + "step": 9239 + }, + { + "epoch": 0.21660015119815748, + "grad_norm": 0.998485803604126, + "learning_rate": 0.0001777637877745441, + "loss": 0.1601, + "step": 9240 + }, + { + "epoch": 0.21662359277296248, + "grad_norm": 3.5180938243865967, + "learning_rate": 0.00017775915636851687, + "loss": 1.6112, + "step": 9241 + }, + { + "epoch": 0.21664703434776747, + "grad_norm": 7.437420845031738, + "learning_rate": 0.00017775452454056625, + "loss": 1.0047, + "step": 9242 + }, + { + "epoch": 0.2166704759225725, + "grad_norm": 4.440474033355713, + "learning_rate": 0.00017774989229071738, + "loss": 0.7331, + "step": 9243 + }, + { + "epoch": 0.2166939174973775, + "grad_norm": 2.1364529132843018, + "learning_rate": 0.00017774525961899545, + "loss": 1.0595, + "step": 9244 + }, + { + "epoch": 0.21671735907218248, + "grad_norm": 3.3771262168884277, + "learning_rate": 0.00017774062652542552, + "loss": 0.3606, + "step": 9245 + }, + { + "epoch": 0.21674080064698747, + "grad_norm": 1.1200131177902222, + "learning_rate": 0.00017773599301003275, + "loss": 0.169, + "step": 9246 + }, + { + "epoch": 0.21676424222179247, + "grad_norm": 3.6009533405303955, + "learning_rate": 0.00017773135907284235, + "loss": 0.5025, + "step": 9247 + }, + { + "epoch": 0.21678768379659746, + "grad_norm": 6.058791637420654, + "learning_rate": 0.0001777267247138794, + "loss": 0.3134, + "step": 9248 + }, + { + "epoch": 0.21681112537140246, + "grad_norm": 4.1628241539001465, + "learning_rate": 0.00017772208993316904, + "loss": 0.6452, + "step": 9249 + }, + { + "epoch": 0.21683456694620745, + "grad_norm": 6.117635726928711, + "learning_rate": 0.00017771745473073644, + "loss": 0.5261, + "step": 9250 + }, + { + "epoch": 0.21685800852101245, + "grad_norm": 6.036595344543457, + "learning_rate": 0.00017771281910660674, + "loss": 0.7825, + "step": 9251 + }, + { + "epoch": 0.21688145009581744, + "grad_norm": 4.008380889892578, + "learning_rate": 0.00017770818306080507, + "loss": 0.4734, + "step": 9252 + }, + { + "epoch": 0.21690489167062244, + "grad_norm": 1.5561180114746094, + "learning_rate": 0.00017770354659335665, + "loss": 0.1967, + "step": 9253 + }, + { + "epoch": 0.21692833324542743, + "grad_norm": 4.742888450622559, + "learning_rate": 0.0001776989097042866, + "loss": 0.3146, + "step": 9254 + }, + { + "epoch": 0.21695177482023242, + "grad_norm": 3.197455406188965, + "learning_rate": 0.00017769427239362008, + "loss": 0.6284, + "step": 9255 + }, + { + "epoch": 0.21697521639503742, + "grad_norm": 1.4470221996307373, + "learning_rate": 0.00017768963466138223, + "loss": 0.2554, + "step": 9256 + }, + { + "epoch": 0.2169986579698424, + "grad_norm": 3.848971366882324, + "learning_rate": 0.00017768499650759826, + "loss": 0.5683, + "step": 9257 + }, + { + "epoch": 0.2170220995446474, + "grad_norm": 3.761971950531006, + "learning_rate": 0.00017768035793229334, + "loss": 0.4515, + "step": 9258 + }, + { + "epoch": 0.2170455411194524, + "grad_norm": 4.898472309112549, + "learning_rate": 0.00017767571893549254, + "loss": 0.7423, + "step": 9259 + }, + { + "epoch": 0.2170689826942574, + "grad_norm": 3.73530912399292, + "learning_rate": 0.00017767107951722116, + "loss": 0.571, + "step": 9260 + }, + { + "epoch": 0.2170924242690624, + "grad_norm": 1.1039148569107056, + "learning_rate": 0.00017766643967750433, + "loss": 0.0914, + "step": 9261 + }, + { + "epoch": 0.21711586584386738, + "grad_norm": 4.8433380126953125, + "learning_rate": 0.0001776617994163672, + "loss": 0.5805, + "step": 9262 + }, + { + "epoch": 0.21713930741867238, + "grad_norm": 1.7059460878372192, + "learning_rate": 0.00017765715873383495, + "loss": 0.1087, + "step": 9263 + }, + { + "epoch": 0.21716274899347737, + "grad_norm": 5.4700846672058105, + "learning_rate": 0.00017765251762993274, + "loss": 0.4498, + "step": 9264 + }, + { + "epoch": 0.21718619056828237, + "grad_norm": 1.8919740915298462, + "learning_rate": 0.00017764787610468586, + "loss": 0.2236, + "step": 9265 + }, + { + "epoch": 0.21720963214308736, + "grad_norm": 3.0336217880249023, + "learning_rate": 0.0001776432341581194, + "loss": 0.4765, + "step": 9266 + }, + { + "epoch": 0.21723307371789236, + "grad_norm": 0.9010722041130066, + "learning_rate": 0.00017763859179025854, + "loss": 0.0612, + "step": 9267 + }, + { + "epoch": 0.21725651529269735, + "grad_norm": 3.6476078033447266, + "learning_rate": 0.0001776339490011285, + "loss": 0.3094, + "step": 9268 + }, + { + "epoch": 0.21727995686750237, + "grad_norm": 4.181624889373779, + "learning_rate": 0.00017762930579075448, + "loss": 0.5366, + "step": 9269 + }, + { + "epoch": 0.21730339844230737, + "grad_norm": 1.1481126546859741, + "learning_rate": 0.00017762466215916168, + "loss": 0.1129, + "step": 9270 + }, + { + "epoch": 0.21732684001711236, + "grad_norm": 4.427577972412109, + "learning_rate": 0.00017762001810637528, + "loss": 0.6074, + "step": 9271 + }, + { + "epoch": 0.21735028159191735, + "grad_norm": 3.2488510608673096, + "learning_rate": 0.00017761537363242047, + "loss": 0.6137, + "step": 9272 + }, + { + "epoch": 0.21737372316672235, + "grad_norm": 5.264405727386475, + "learning_rate": 0.00017761072873732244, + "loss": 0.4724, + "step": 9273 + }, + { + "epoch": 0.21739716474152734, + "grad_norm": 4.304667949676514, + "learning_rate": 0.00017760608342110644, + "loss": 0.2193, + "step": 9274 + }, + { + "epoch": 0.21742060631633234, + "grad_norm": 1.7277759313583374, + "learning_rate": 0.00017760143768379762, + "loss": 0.1641, + "step": 9275 + }, + { + "epoch": 0.21744404789113733, + "grad_norm": 1.8046865463256836, + "learning_rate": 0.00017759679152542127, + "loss": 0.1683, + "step": 9276 + }, + { + "epoch": 0.21746748946594233, + "grad_norm": 0.8671835660934448, + "learning_rate": 0.00017759214494600248, + "loss": 0.0971, + "step": 9277 + }, + { + "epoch": 0.21749093104074732, + "grad_norm": 6.921824932098389, + "learning_rate": 0.00017758749794556658, + "loss": 0.2732, + "step": 9278 + }, + { + "epoch": 0.21751437261555231, + "grad_norm": 4.641030311584473, + "learning_rate": 0.00017758285052413872, + "loss": 0.6512, + "step": 9279 + }, + { + "epoch": 0.2175378141903573, + "grad_norm": 2.0387814044952393, + "learning_rate": 0.00017757820268174412, + "loss": 1.2912, + "step": 9280 + }, + { + "epoch": 0.2175612557651623, + "grad_norm": 1.5979775190353394, + "learning_rate": 0.000177573554418408, + "loss": 0.2771, + "step": 9281 + }, + { + "epoch": 0.2175846973399673, + "grad_norm": 3.2838616371154785, + "learning_rate": 0.00017756890573415564, + "loss": 1.2964, + "step": 9282 + }, + { + "epoch": 0.2176081389147723, + "grad_norm": 2.7282330989837646, + "learning_rate": 0.00017756425662901219, + "loss": 0.1915, + "step": 9283 + }, + { + "epoch": 0.2176315804895773, + "grad_norm": 1.6118035316467285, + "learning_rate": 0.0001775596071030029, + "loss": 0.1563, + "step": 9284 + }, + { + "epoch": 0.21765502206438228, + "grad_norm": 6.973618984222412, + "learning_rate": 0.000177554957156153, + "loss": 0.6327, + "step": 9285 + }, + { + "epoch": 0.21767846363918728, + "grad_norm": 0.39490818977355957, + "learning_rate": 0.0001775503067884877, + "loss": 0.0377, + "step": 9286 + }, + { + "epoch": 0.21770190521399227, + "grad_norm": 5.6531219482421875, + "learning_rate": 0.0001775456560000323, + "loss": 0.5564, + "step": 9287 + }, + { + "epoch": 0.21772534678879726, + "grad_norm": 1.1600103378295898, + "learning_rate": 0.00017754100479081198, + "loss": 0.1435, + "step": 9288 + }, + { + "epoch": 0.21774878836360226, + "grad_norm": 4.052919387817383, + "learning_rate": 0.00017753635316085196, + "loss": 0.4034, + "step": 9289 + }, + { + "epoch": 0.21777222993840725, + "grad_norm": 3.860772132873535, + "learning_rate": 0.00017753170111017754, + "loss": 0.5589, + "step": 9290 + }, + { + "epoch": 0.21779567151321225, + "grad_norm": 3.0485522747039795, + "learning_rate": 0.00017752704863881392, + "loss": 0.3506, + "step": 9291 + }, + { + "epoch": 0.21781911308801724, + "grad_norm": 2.698960065841675, + "learning_rate": 0.00017752239574678633, + "loss": 0.2361, + "step": 9292 + }, + { + "epoch": 0.21784255466282224, + "grad_norm": 4.828877925872803, + "learning_rate": 0.00017751774243412007, + "loss": 0.7671, + "step": 9293 + }, + { + "epoch": 0.21786599623762726, + "grad_norm": 4.44006872177124, + "learning_rate": 0.00017751308870084037, + "loss": 0.5673, + "step": 9294 + }, + { + "epoch": 0.21788943781243225, + "grad_norm": 3.7330384254455566, + "learning_rate": 0.00017750843454697243, + "loss": 1.7447, + "step": 9295 + }, + { + "epoch": 0.21791287938723725, + "grad_norm": 3.8166584968566895, + "learning_rate": 0.00017750377997254155, + "loss": 0.441, + "step": 9296 + }, + { + "epoch": 0.21793632096204224, + "grad_norm": 3.035008192062378, + "learning_rate": 0.00017749912497757298, + "loss": 0.2667, + "step": 9297 + }, + { + "epoch": 0.21795976253684723, + "grad_norm": 3.8581085205078125, + "learning_rate": 0.00017749446956209203, + "loss": 0.403, + "step": 9298 + }, + { + "epoch": 0.21798320411165223, + "grad_norm": 1.7902817726135254, + "learning_rate": 0.00017748981372612388, + "loss": 0.2619, + "step": 9299 + }, + { + "epoch": 0.21800664568645722, + "grad_norm": 5.608353614807129, + "learning_rate": 0.00017748515746969382, + "loss": 0.7975, + "step": 9300 + }, + { + "epoch": 0.21803008726126222, + "grad_norm": 2.9123432636260986, + "learning_rate": 0.00017748050079282712, + "loss": 1.3816, + "step": 9301 + }, + { + "epoch": 0.2180535288360672, + "grad_norm": 1.2511318922042847, + "learning_rate": 0.00017747584369554902, + "loss": 1.0632, + "step": 9302 + }, + { + "epoch": 0.2180769704108722, + "grad_norm": 1.1890511512756348, + "learning_rate": 0.00017747118617788483, + "loss": 0.1999, + "step": 9303 + }, + { + "epoch": 0.2181004119856772, + "grad_norm": 2.6676905155181885, + "learning_rate": 0.0001774665282398598, + "loss": 1.3575, + "step": 9304 + }, + { + "epoch": 0.2181238535604822, + "grad_norm": 2.801098346710205, + "learning_rate": 0.00017746186988149923, + "loss": 1.531, + "step": 9305 + }, + { + "epoch": 0.2181472951352872, + "grad_norm": 2.2918262481689453, + "learning_rate": 0.00017745721110282837, + "loss": 0.3374, + "step": 9306 + }, + { + "epoch": 0.21817073671009218, + "grad_norm": 0.5570820569992065, + "learning_rate": 0.0001774525519038725, + "loss": 0.0631, + "step": 9307 + }, + { + "epoch": 0.21819417828489718, + "grad_norm": 3.776197671890259, + "learning_rate": 0.00017744789228465687, + "loss": 0.3866, + "step": 9308 + }, + { + "epoch": 0.21821761985970217, + "grad_norm": 1.9766219854354858, + "learning_rate": 0.00017744323224520687, + "loss": 0.2369, + "step": 9309 + }, + { + "epoch": 0.21824106143450717, + "grad_norm": 4.2757487297058105, + "learning_rate": 0.00017743857178554765, + "loss": 0.5621, + "step": 9310 + }, + { + "epoch": 0.21826450300931216, + "grad_norm": 1.2299962043762207, + "learning_rate": 0.0001774339109057046, + "loss": 0.1008, + "step": 9311 + }, + { + "epoch": 0.21828794458411716, + "grad_norm": 4.358275413513184, + "learning_rate": 0.00017742924960570297, + "loss": 0.4377, + "step": 9312 + }, + { + "epoch": 0.21831138615892215, + "grad_norm": 2.9394326210021973, + "learning_rate": 0.00017742458788556803, + "loss": 0.2898, + "step": 9313 + }, + { + "epoch": 0.21833482773372714, + "grad_norm": 3.6946492195129395, + "learning_rate": 0.0001774199257453251, + "loss": 0.4408, + "step": 9314 + }, + { + "epoch": 0.21835826930853214, + "grad_norm": 3.440873384475708, + "learning_rate": 0.0001774152631849995, + "loss": 0.2944, + "step": 9315 + }, + { + "epoch": 0.21838171088333713, + "grad_norm": 4.323828220367432, + "learning_rate": 0.00017741060020461647, + "loss": 0.6587, + "step": 9316 + }, + { + "epoch": 0.21840515245814213, + "grad_norm": 1.35249662399292, + "learning_rate": 0.0001774059368042014, + "loss": 0.188, + "step": 9317 + }, + { + "epoch": 0.21842859403294712, + "grad_norm": 4.096333026885986, + "learning_rate": 0.0001774012729837795, + "loss": 0.4238, + "step": 9318 + }, + { + "epoch": 0.21845203560775212, + "grad_norm": 0.37602800130844116, + "learning_rate": 0.00017739660874337615, + "loss": 0.0459, + "step": 9319 + }, + { + "epoch": 0.21847547718255714, + "grad_norm": 3.891909122467041, + "learning_rate": 0.0001773919440830166, + "loss": 0.219, + "step": 9320 + }, + { + "epoch": 0.21849891875736213, + "grad_norm": 4.8811421394348145, + "learning_rate": 0.00017738727900272618, + "loss": 0.5313, + "step": 9321 + }, + { + "epoch": 0.21852236033216713, + "grad_norm": 0.7717480659484863, + "learning_rate": 0.00017738261350253023, + "loss": 0.0695, + "step": 9322 + }, + { + "epoch": 0.21854580190697212, + "grad_norm": 4.286736488342285, + "learning_rate": 0.000177377947582454, + "loss": 0.5859, + "step": 9323 + }, + { + "epoch": 0.21856924348177711, + "grad_norm": 0.8398900628089905, + "learning_rate": 0.0001773732812425229, + "loss": 0.0894, + "step": 9324 + }, + { + "epoch": 0.2185926850565821, + "grad_norm": 3.12785267829895, + "learning_rate": 0.00017736861448276219, + "loss": 0.2558, + "step": 9325 + }, + { + "epoch": 0.2186161266313871, + "grad_norm": 2.6091909408569336, + "learning_rate": 0.0001773639473031972, + "loss": 0.2621, + "step": 9326 + }, + { + "epoch": 0.2186395682061921, + "grad_norm": 2.377802848815918, + "learning_rate": 0.00017735927970385324, + "loss": 0.2468, + "step": 9327 + }, + { + "epoch": 0.2186630097809971, + "grad_norm": 4.440091133117676, + "learning_rate": 0.00017735461168475567, + "loss": 0.5647, + "step": 9328 + }, + { + "epoch": 0.2186864513558021, + "grad_norm": 2.0840253829956055, + "learning_rate": 0.0001773499432459298, + "loss": 1.1659, + "step": 9329 + }, + { + "epoch": 0.21870989293060708, + "grad_norm": 1.0185494422912598, + "learning_rate": 0.00017734527438740093, + "loss": 0.1164, + "step": 9330 + }, + { + "epoch": 0.21873333450541207, + "grad_norm": 1.2081735134124756, + "learning_rate": 0.00017734060510919445, + "loss": 0.1272, + "step": 9331 + }, + { + "epoch": 0.21875677608021707, + "grad_norm": 2.76602840423584, + "learning_rate": 0.00017733593541133572, + "loss": 0.2601, + "step": 9332 + }, + { + "epoch": 0.21878021765502206, + "grad_norm": 8.444408416748047, + "learning_rate": 0.00017733126529384996, + "loss": 0.9437, + "step": 9333 + }, + { + "epoch": 0.21880365922982706, + "grad_norm": 5.188808917999268, + "learning_rate": 0.00017732659475676261, + "loss": 0.4034, + "step": 9334 + }, + { + "epoch": 0.21882710080463205, + "grad_norm": 3.569105386734009, + "learning_rate": 0.00017732192380009898, + "loss": 0.3261, + "step": 9335 + }, + { + "epoch": 0.21885054237943705, + "grad_norm": 5.013146877288818, + "learning_rate": 0.00017731725242388442, + "loss": 0.3801, + "step": 9336 + }, + { + "epoch": 0.21887398395424204, + "grad_norm": 4.454406261444092, + "learning_rate": 0.00017731258062814428, + "loss": 0.3216, + "step": 9337 + }, + { + "epoch": 0.21889742552904703, + "grad_norm": 3.6166558265686035, + "learning_rate": 0.0001773079084129039, + "loss": 0.5804, + "step": 9338 + }, + { + "epoch": 0.21892086710385203, + "grad_norm": 1.8040882349014282, + "learning_rate": 0.0001773032357781886, + "loss": 0.145, + "step": 9339 + }, + { + "epoch": 0.21894430867865702, + "grad_norm": 1.8425483703613281, + "learning_rate": 0.00017729856272402384, + "loss": 1.04, + "step": 9340 + }, + { + "epoch": 0.21896775025346202, + "grad_norm": 2.5342118740081787, + "learning_rate": 0.00017729388925043487, + "loss": 0.1479, + "step": 9341 + }, + { + "epoch": 0.218991191828267, + "grad_norm": 2.3108441829681396, + "learning_rate": 0.00017728921535744706, + "loss": 0.1986, + "step": 9342 + }, + { + "epoch": 0.219014633403072, + "grad_norm": 3.7497665882110596, + "learning_rate": 0.00017728454104508581, + "loss": 0.4962, + "step": 9343 + }, + { + "epoch": 0.219038074977877, + "grad_norm": 4.218701362609863, + "learning_rate": 0.00017727986631337647, + "loss": 0.5345, + "step": 9344 + }, + { + "epoch": 0.21906151655268202, + "grad_norm": 4.3610310554504395, + "learning_rate": 0.00017727519116234438, + "loss": 0.4072, + "step": 9345 + }, + { + "epoch": 0.21908495812748702, + "grad_norm": 1.0071086883544922, + "learning_rate": 0.00017727051559201498, + "loss": 0.0798, + "step": 9346 + }, + { + "epoch": 0.219108399702292, + "grad_norm": 4.240580081939697, + "learning_rate": 0.00017726583960241357, + "loss": 0.297, + "step": 9347 + }, + { + "epoch": 0.219131841277097, + "grad_norm": 4.681236743927002, + "learning_rate": 0.00017726116319356552, + "loss": 0.6574, + "step": 9348 + }, + { + "epoch": 0.219155282851902, + "grad_norm": 3.8949670791625977, + "learning_rate": 0.00017725648636549623, + "loss": 0.2446, + "step": 9349 + }, + { + "epoch": 0.219178724426707, + "grad_norm": 1.7018481492996216, + "learning_rate": 0.0001772518091182311, + "loss": 0.1761, + "step": 9350 + }, + { + "epoch": 0.219202166001512, + "grad_norm": 1.5164743661880493, + "learning_rate": 0.00017724713145179545, + "loss": 0.2267, + "step": 9351 + }, + { + "epoch": 0.21922560757631698, + "grad_norm": 6.818190097808838, + "learning_rate": 0.0001772424533662147, + "loss": 0.2639, + "step": 9352 + }, + { + "epoch": 0.21924904915112198, + "grad_norm": 3.7216320037841797, + "learning_rate": 0.00017723777486151424, + "loss": 0.4828, + "step": 9353 + }, + { + "epoch": 0.21927249072592697, + "grad_norm": 3.8834102153778076, + "learning_rate": 0.00017723309593771944, + "loss": 0.3928, + "step": 9354 + }, + { + "epoch": 0.21929593230073197, + "grad_norm": 3.3456521034240723, + "learning_rate": 0.00017722841659485568, + "loss": 0.3434, + "step": 9355 + }, + { + "epoch": 0.21931937387553696, + "grad_norm": 0.7265111207962036, + "learning_rate": 0.00017722373683294837, + "loss": 0.1059, + "step": 9356 + }, + { + "epoch": 0.21934281545034195, + "grad_norm": 1.1527167558670044, + "learning_rate": 0.0001772190566520229, + "loss": 0.1311, + "step": 9357 + }, + { + "epoch": 0.21936625702514695, + "grad_norm": 2.2152421474456787, + "learning_rate": 0.00017721437605210466, + "loss": 0.2934, + "step": 9358 + }, + { + "epoch": 0.21938969859995194, + "grad_norm": 4.52987813949585, + "learning_rate": 0.000177209695033219, + "loss": 0.4762, + "step": 9359 + }, + { + "epoch": 0.21941314017475694, + "grad_norm": 1.797623872756958, + "learning_rate": 0.0001772050135953914, + "loss": 0.2232, + "step": 9360 + }, + { + "epoch": 0.21943658174956193, + "grad_norm": 4.968474388122559, + "learning_rate": 0.0001772003317386472, + "loss": 0.3686, + "step": 9361 + }, + { + "epoch": 0.21946002332436693, + "grad_norm": 4.458798885345459, + "learning_rate": 0.00017719564946301185, + "loss": 0.3811, + "step": 9362 + }, + { + "epoch": 0.21948346489917192, + "grad_norm": 1.585570216178894, + "learning_rate": 0.00017719096676851076, + "loss": 0.2188, + "step": 9363 + }, + { + "epoch": 0.21950690647397691, + "grad_norm": 4.922850131988525, + "learning_rate": 0.00017718628365516928, + "loss": 0.7806, + "step": 9364 + }, + { + "epoch": 0.2195303480487819, + "grad_norm": 0.7576619982719421, + "learning_rate": 0.00017718160012301286, + "loss": 0.0808, + "step": 9365 + }, + { + "epoch": 0.2195537896235869, + "grad_norm": 3.413921594619751, + "learning_rate": 0.00017717691617206692, + "loss": 0.3951, + "step": 9366 + }, + { + "epoch": 0.2195772311983919, + "grad_norm": 1.2412641048431396, + "learning_rate": 0.00017717223180235684, + "loss": 0.185, + "step": 9367 + }, + { + "epoch": 0.2196006727731969, + "grad_norm": 4.210721969604492, + "learning_rate": 0.00017716754701390807, + "loss": 0.3027, + "step": 9368 + }, + { + "epoch": 0.2196241143480019, + "grad_norm": 4.188389778137207, + "learning_rate": 0.00017716286180674602, + "loss": 0.4023, + "step": 9369 + }, + { + "epoch": 0.21964755592280688, + "grad_norm": 6.383159637451172, + "learning_rate": 0.0001771581761808961, + "loss": 1.0381, + "step": 9370 + }, + { + "epoch": 0.2196709974976119, + "grad_norm": 3.538133144378662, + "learning_rate": 0.00017715349013638378, + "loss": 1.4638, + "step": 9371 + }, + { + "epoch": 0.2196944390724169, + "grad_norm": 3.74774169921875, + "learning_rate": 0.00017714880367323445, + "loss": 0.2758, + "step": 9372 + }, + { + "epoch": 0.2197178806472219, + "grad_norm": 4.4296112060546875, + "learning_rate": 0.0001771441167914735, + "loss": 0.7204, + "step": 9373 + }, + { + "epoch": 0.21974132222202689, + "grad_norm": 3.3690035343170166, + "learning_rate": 0.00017713942949112646, + "loss": 0.2879, + "step": 9374 + }, + { + "epoch": 0.21976476379683188, + "grad_norm": 1.9672398567199707, + "learning_rate": 0.00017713474177221867, + "loss": 0.1412, + "step": 9375 + }, + { + "epoch": 0.21978820537163687, + "grad_norm": 2.561570882797241, + "learning_rate": 0.0001771300536347756, + "loss": 0.3476, + "step": 9376 + }, + { + "epoch": 0.21981164694644187, + "grad_norm": 4.404412269592285, + "learning_rate": 0.0001771253650788227, + "loss": 0.5412, + "step": 9377 + }, + { + "epoch": 0.21983508852124686, + "grad_norm": 0.6329814791679382, + "learning_rate": 0.0001771206761043854, + "loss": 0.061, + "step": 9378 + }, + { + "epoch": 0.21985853009605186, + "grad_norm": 1.463951587677002, + "learning_rate": 0.00017711598671148914, + "loss": 0.1126, + "step": 9379 + }, + { + "epoch": 0.21988197167085685, + "grad_norm": 1.9914095401763916, + "learning_rate": 0.00017711129690015938, + "loss": 0.328, + "step": 9380 + }, + { + "epoch": 0.21990541324566185, + "grad_norm": 5.697826862335205, + "learning_rate": 0.00017710660667042154, + "loss": 0.3681, + "step": 9381 + }, + { + "epoch": 0.21992885482046684, + "grad_norm": 4.628183364868164, + "learning_rate": 0.00017710191602230107, + "loss": 0.469, + "step": 9382 + }, + { + "epoch": 0.21995229639527183, + "grad_norm": 3.9090654850006104, + "learning_rate": 0.00017709722495582345, + "loss": 0.5294, + "step": 9383 + }, + { + "epoch": 0.21997573797007683, + "grad_norm": 2.1042044162750244, + "learning_rate": 0.00017709253347101417, + "loss": 0.2154, + "step": 9384 + }, + { + "epoch": 0.21999917954488182, + "grad_norm": 1.7876410484313965, + "learning_rate": 0.00017708784156789857, + "loss": 0.1109, + "step": 9385 + }, + { + "epoch": 0.22002262111968682, + "grad_norm": 3.68536376953125, + "learning_rate": 0.0001770831492465022, + "loss": 0.5311, + "step": 9386 + }, + { + "epoch": 0.2200460626944918, + "grad_norm": 1.4132508039474487, + "learning_rate": 0.0001770784565068505, + "loss": 0.1735, + "step": 9387 + }, + { + "epoch": 0.2200695042692968, + "grad_norm": 4.543850421905518, + "learning_rate": 0.00017707376334896892, + "loss": 0.4853, + "step": 9388 + }, + { + "epoch": 0.2200929458441018, + "grad_norm": 4.139974594116211, + "learning_rate": 0.00017706906977288291, + "loss": 0.3407, + "step": 9389 + }, + { + "epoch": 0.2201163874189068, + "grad_norm": 4.819606304168701, + "learning_rate": 0.000177064375778618, + "loss": 0.343, + "step": 9390 + }, + { + "epoch": 0.2201398289937118, + "grad_norm": 6.629724502563477, + "learning_rate": 0.0001770596813661996, + "loss": 0.5555, + "step": 9391 + }, + { + "epoch": 0.22016327056851678, + "grad_norm": 3.5300910472869873, + "learning_rate": 0.00017705498653565317, + "loss": 0.4678, + "step": 9392 + }, + { + "epoch": 0.22018671214332178, + "grad_norm": 4.298120498657227, + "learning_rate": 0.00017705029128700428, + "loss": 0.2973, + "step": 9393 + }, + { + "epoch": 0.22021015371812677, + "grad_norm": 3.5900344848632812, + "learning_rate": 0.0001770455956202783, + "loss": 0.3374, + "step": 9394 + }, + { + "epoch": 0.22023359529293177, + "grad_norm": 3.017334222793579, + "learning_rate": 0.00017704089953550075, + "loss": 1.4444, + "step": 9395 + }, + { + "epoch": 0.22025703686773676, + "grad_norm": 3.7830333709716797, + "learning_rate": 0.00017703620303269714, + "loss": 0.5629, + "step": 9396 + }, + { + "epoch": 0.22028047844254178, + "grad_norm": 1.4798853397369385, + "learning_rate": 0.00017703150611189292, + "loss": 0.1714, + "step": 9397 + }, + { + "epoch": 0.22030392001734678, + "grad_norm": 2.5807793140411377, + "learning_rate": 0.00017702680877311358, + "loss": 0.3576, + "step": 9398 + }, + { + "epoch": 0.22032736159215177, + "grad_norm": 5.5280046463012695, + "learning_rate": 0.0001770221110163846, + "loss": 0.6043, + "step": 9399 + }, + { + "epoch": 0.22035080316695677, + "grad_norm": 3.053173303604126, + "learning_rate": 0.0001770174128417315, + "loss": 0.2589, + "step": 9400 + }, + { + "epoch": 0.22037424474176176, + "grad_norm": 0.929924726486206, + "learning_rate": 0.0001770127142491797, + "loss": 0.0688, + "step": 9401 + }, + { + "epoch": 0.22039768631656675, + "grad_norm": 3.410640001296997, + "learning_rate": 0.00017700801523875482, + "loss": 0.3945, + "step": 9402 + }, + { + "epoch": 0.22042112789137175, + "grad_norm": 4.095454692840576, + "learning_rate": 0.00017700331581048224, + "loss": 0.4888, + "step": 9403 + }, + { + "epoch": 0.22044456946617674, + "grad_norm": 5.028672695159912, + "learning_rate": 0.0001769986159643875, + "loss": 0.6366, + "step": 9404 + }, + { + "epoch": 0.22046801104098174, + "grad_norm": 4.060763359069824, + "learning_rate": 0.00017699391570049613, + "loss": 0.7755, + "step": 9405 + }, + { + "epoch": 0.22049145261578673, + "grad_norm": 1.351911187171936, + "learning_rate": 0.0001769892150188336, + "loss": 0.1389, + "step": 9406 + }, + { + "epoch": 0.22051489419059173, + "grad_norm": 3.427110195159912, + "learning_rate": 0.00017698451391942542, + "loss": 0.5348, + "step": 9407 + }, + { + "epoch": 0.22053833576539672, + "grad_norm": 3.0840604305267334, + "learning_rate": 0.0001769798124022971, + "loss": 0.368, + "step": 9408 + }, + { + "epoch": 0.22056177734020171, + "grad_norm": 4.292536735534668, + "learning_rate": 0.00017697511046747418, + "loss": 0.6221, + "step": 9409 + }, + { + "epoch": 0.2205852189150067, + "grad_norm": 3.000502347946167, + "learning_rate": 0.00017697040811498208, + "loss": 1.3776, + "step": 9410 + }, + { + "epoch": 0.2206086604898117, + "grad_norm": 3.612877368927002, + "learning_rate": 0.00017696570534484645, + "loss": 0.5416, + "step": 9411 + }, + { + "epoch": 0.2206321020646167, + "grad_norm": 5.562022686004639, + "learning_rate": 0.0001769610021570927, + "loss": 0.6829, + "step": 9412 + }, + { + "epoch": 0.2206555436394217, + "grad_norm": 2.764847755432129, + "learning_rate": 0.0001769562985517464, + "loss": 0.334, + "step": 9413 + }, + { + "epoch": 0.22067898521422669, + "grad_norm": 3.380491018295288, + "learning_rate": 0.00017695159452883305, + "loss": 1.4446, + "step": 9414 + }, + { + "epoch": 0.22070242678903168, + "grad_norm": 3.195192813873291, + "learning_rate": 0.0001769468900883782, + "loss": 0.3796, + "step": 9415 + }, + { + "epoch": 0.22072586836383667, + "grad_norm": 3.5884108543395996, + "learning_rate": 0.00017694218523040735, + "loss": 0.3513, + "step": 9416 + }, + { + "epoch": 0.22074930993864167, + "grad_norm": 3.7727549076080322, + "learning_rate": 0.00017693747995494604, + "loss": 0.466, + "step": 9417 + }, + { + "epoch": 0.22077275151344666, + "grad_norm": 1.9071838855743408, + "learning_rate": 0.00017693277426201985, + "loss": 0.2048, + "step": 9418 + }, + { + "epoch": 0.22079619308825166, + "grad_norm": 2.835007429122925, + "learning_rate": 0.0001769280681516542, + "loss": 0.1673, + "step": 9419 + }, + { + "epoch": 0.22081963466305665, + "grad_norm": 4.301575660705566, + "learning_rate": 0.0001769233616238747, + "loss": 0.5265, + "step": 9420 + }, + { + "epoch": 0.22084307623786165, + "grad_norm": 4.572850704193115, + "learning_rate": 0.0001769186546787069, + "loss": 0.4902, + "step": 9421 + }, + { + "epoch": 0.22086651781266667, + "grad_norm": 6.036021709442139, + "learning_rate": 0.0001769139473161763, + "loss": 0.8714, + "step": 9422 + }, + { + "epoch": 0.22088995938747166, + "grad_norm": 3.229555130004883, + "learning_rate": 0.00017690923953630846, + "loss": 1.8526, + "step": 9423 + }, + { + "epoch": 0.22091340096227666, + "grad_norm": 0.7599018812179565, + "learning_rate": 0.00017690453133912892, + "loss": 0.0501, + "step": 9424 + }, + { + "epoch": 0.22093684253708165, + "grad_norm": 4.000246524810791, + "learning_rate": 0.00017689982272466325, + "loss": 0.4591, + "step": 9425 + }, + { + "epoch": 0.22096028411188665, + "grad_norm": 4.201990127563477, + "learning_rate": 0.00017689511369293698, + "loss": 0.4805, + "step": 9426 + }, + { + "epoch": 0.22098372568669164, + "grad_norm": 1.047379493713379, + "learning_rate": 0.00017689040424397564, + "loss": 0.6096, + "step": 9427 + }, + { + "epoch": 0.22100716726149663, + "grad_norm": 5.3000335693359375, + "learning_rate": 0.0001768856943778048, + "loss": 0.5985, + "step": 9428 + }, + { + "epoch": 0.22103060883630163, + "grad_norm": 3.877664804458618, + "learning_rate": 0.00017688098409445005, + "loss": 0.5474, + "step": 9429 + }, + { + "epoch": 0.22105405041110662, + "grad_norm": 0.9188741445541382, + "learning_rate": 0.00017687627339393688, + "loss": 0.1558, + "step": 9430 + }, + { + "epoch": 0.22107749198591162, + "grad_norm": 3.8900675773620605, + "learning_rate": 0.00017687156227629092, + "loss": 0.4342, + "step": 9431 + }, + { + "epoch": 0.2211009335607166, + "grad_norm": 1.9862639904022217, + "learning_rate": 0.00017686685074153768, + "loss": 0.2631, + "step": 9432 + }, + { + "epoch": 0.2211243751355216, + "grad_norm": 5.369093894958496, + "learning_rate": 0.00017686213878970278, + "loss": 0.7878, + "step": 9433 + }, + { + "epoch": 0.2211478167103266, + "grad_norm": 3.239262580871582, + "learning_rate": 0.0001768574264208117, + "loss": 0.4526, + "step": 9434 + }, + { + "epoch": 0.2211712582851316, + "grad_norm": 3.450467348098755, + "learning_rate": 0.0001768527136348901, + "loss": 0.3633, + "step": 9435 + }, + { + "epoch": 0.2211946998599366, + "grad_norm": 3.1007232666015625, + "learning_rate": 0.00017684800043196353, + "loss": 1.7363, + "step": 9436 + }, + { + "epoch": 0.22121814143474158, + "grad_norm": 5.949241638183594, + "learning_rate": 0.00017684328681205752, + "loss": 0.755, + "step": 9437 + }, + { + "epoch": 0.22124158300954658, + "grad_norm": 3.868243455886841, + "learning_rate": 0.00017683857277519768, + "loss": 2.3395, + "step": 9438 + }, + { + "epoch": 0.22126502458435157, + "grad_norm": 2.7576019763946533, + "learning_rate": 0.0001768338583214096, + "loss": 0.2703, + "step": 9439 + }, + { + "epoch": 0.22128846615915657, + "grad_norm": 2.458479642868042, + "learning_rate": 0.00017682914345071882, + "loss": 0.1518, + "step": 9440 + }, + { + "epoch": 0.22131190773396156, + "grad_norm": 3.6413040161132812, + "learning_rate": 0.00017682442816315097, + "loss": 0.5176, + "step": 9441 + }, + { + "epoch": 0.22133534930876655, + "grad_norm": 1.6724822521209717, + "learning_rate": 0.0001768197124587316, + "loss": 0.2892, + "step": 9442 + }, + { + "epoch": 0.22135879088357155, + "grad_norm": 1.4654453992843628, + "learning_rate": 0.0001768149963374863, + "loss": 0.2732, + "step": 9443 + }, + { + "epoch": 0.22138223245837654, + "grad_norm": 3.9618113040924072, + "learning_rate": 0.00017681027979944068, + "loss": 0.4041, + "step": 9444 + }, + { + "epoch": 0.22140567403318154, + "grad_norm": 0.7504937648773193, + "learning_rate": 0.00017680556284462034, + "loss": 0.1073, + "step": 9445 + }, + { + "epoch": 0.22142911560798653, + "grad_norm": 4.273165702819824, + "learning_rate": 0.00017680084547305085, + "loss": 0.2951, + "step": 9446 + }, + { + "epoch": 0.22145255718279153, + "grad_norm": 2.2739667892456055, + "learning_rate": 0.0001767961276847578, + "loss": 0.3095, + "step": 9447 + }, + { + "epoch": 0.22147599875759655, + "grad_norm": 3.7347893714904785, + "learning_rate": 0.00017679140947976678, + "loss": 0.3405, + "step": 9448 + }, + { + "epoch": 0.22149944033240154, + "grad_norm": 3.915639877319336, + "learning_rate": 0.00017678669085810346, + "loss": 0.4109, + "step": 9449 + }, + { + "epoch": 0.22152288190720654, + "grad_norm": 1.159680724143982, + "learning_rate": 0.00017678197181979338, + "loss": 0.123, + "step": 9450 + }, + { + "epoch": 0.22154632348201153, + "grad_norm": 1.7769908905029297, + "learning_rate": 0.00017677725236486214, + "loss": 1.1831, + "step": 9451 + }, + { + "epoch": 0.22156976505681653, + "grad_norm": 3.8079452514648438, + "learning_rate": 0.00017677253249333538, + "loss": 0.5643, + "step": 9452 + }, + { + "epoch": 0.22159320663162152, + "grad_norm": 3.982390880584717, + "learning_rate": 0.0001767678122052387, + "loss": 0.453, + "step": 9453 + }, + { + "epoch": 0.22161664820642651, + "grad_norm": 1.9203358888626099, + "learning_rate": 0.00017676309150059774, + "loss": 0.2537, + "step": 9454 + }, + { + "epoch": 0.2216400897812315, + "grad_norm": 1.666935920715332, + "learning_rate": 0.0001767583703794381, + "loss": 0.2612, + "step": 9455 + }, + { + "epoch": 0.2216635313560365, + "grad_norm": 3.132702350616455, + "learning_rate": 0.00017675364884178532, + "loss": 0.2277, + "step": 9456 + }, + { + "epoch": 0.2216869729308415, + "grad_norm": 4.179004192352295, + "learning_rate": 0.0001767489268876651, + "loss": 0.3392, + "step": 9457 + }, + { + "epoch": 0.2217104145056465, + "grad_norm": 2.09566068649292, + "learning_rate": 0.00017674420451710307, + "loss": 0.1941, + "step": 9458 + }, + { + "epoch": 0.22173385608045149, + "grad_norm": 0.9522867798805237, + "learning_rate": 0.0001767394817301248, + "loss": 0.103, + "step": 9459 + }, + { + "epoch": 0.22175729765525648, + "grad_norm": 1.981986165046692, + "learning_rate": 0.00017673475852675598, + "loss": 1.4938, + "step": 9460 + }, + { + "epoch": 0.22178073923006147, + "grad_norm": 4.734005928039551, + "learning_rate": 0.00017673003490702218, + "loss": 0.701, + "step": 9461 + }, + { + "epoch": 0.22180418080486647, + "grad_norm": 6.899311542510986, + "learning_rate": 0.00017672531087094905, + "loss": 0.7887, + "step": 9462 + }, + { + "epoch": 0.22182762237967146, + "grad_norm": 4.342239856719971, + "learning_rate": 0.00017672058641856223, + "loss": 0.4488, + "step": 9463 + }, + { + "epoch": 0.22185106395447646, + "grad_norm": 4.750999450683594, + "learning_rate": 0.00017671586154988737, + "loss": 0.7393, + "step": 9464 + }, + { + "epoch": 0.22187450552928145, + "grad_norm": 4.527132034301758, + "learning_rate": 0.00017671113626495006, + "loss": 0.4761, + "step": 9465 + }, + { + "epoch": 0.22189794710408645, + "grad_norm": 7.627859115600586, + "learning_rate": 0.00017670641056377598, + "loss": 0.4581, + "step": 9466 + }, + { + "epoch": 0.22192138867889144, + "grad_norm": 0.5907341241836548, + "learning_rate": 0.00017670168444639076, + "loss": 0.07, + "step": 9467 + }, + { + "epoch": 0.22194483025369643, + "grad_norm": 5.047616004943848, + "learning_rate": 0.00017669695791282006, + "loss": 0.8581, + "step": 9468 + }, + { + "epoch": 0.22196827182850143, + "grad_norm": 5.869987487792969, + "learning_rate": 0.0001766922309630895, + "loss": 0.6766, + "step": 9469 + }, + { + "epoch": 0.22199171340330642, + "grad_norm": 2.483415126800537, + "learning_rate": 0.00017668750359722474, + "loss": 0.2425, + "step": 9470 + }, + { + "epoch": 0.22201515497811142, + "grad_norm": 6.47411584854126, + "learning_rate": 0.0001766827758152514, + "loss": 0.785, + "step": 9471 + }, + { + "epoch": 0.2220385965529164, + "grad_norm": 3.4677910804748535, + "learning_rate": 0.00017667804761719517, + "loss": 0.5246, + "step": 9472 + }, + { + "epoch": 0.22206203812772143, + "grad_norm": 1.6443036794662476, + "learning_rate": 0.00017667331900308173, + "loss": 0.8334, + "step": 9473 + }, + { + "epoch": 0.22208547970252643, + "grad_norm": 4.6977386474609375, + "learning_rate": 0.00017666858997293668, + "loss": 0.4846, + "step": 9474 + }, + { + "epoch": 0.22210892127733142, + "grad_norm": 2.894625425338745, + "learning_rate": 0.0001766638605267857, + "loss": 0.3611, + "step": 9475 + }, + { + "epoch": 0.22213236285213642, + "grad_norm": 0.4520629346370697, + "learning_rate": 0.00017665913066465446, + "loss": 0.042, + "step": 9476 + }, + { + "epoch": 0.2221558044269414, + "grad_norm": 1.5746784210205078, + "learning_rate": 0.00017665440038656865, + "loss": 0.2382, + "step": 9477 + }, + { + "epoch": 0.2221792460017464, + "grad_norm": 2.4192464351654053, + "learning_rate": 0.00017664966969255388, + "loss": 0.2706, + "step": 9478 + }, + { + "epoch": 0.2222026875765514, + "grad_norm": 4.578423500061035, + "learning_rate": 0.00017664493858263587, + "loss": 0.6705, + "step": 9479 + }, + { + "epoch": 0.2222261291513564, + "grad_norm": 2.6545073986053467, + "learning_rate": 0.00017664020705684022, + "loss": 0.2675, + "step": 9480 + }, + { + "epoch": 0.2222495707261614, + "grad_norm": 1.0317128896713257, + "learning_rate": 0.0001766354751151927, + "loss": 0.1119, + "step": 9481 + }, + { + "epoch": 0.22227301230096638, + "grad_norm": 2.4130659103393555, + "learning_rate": 0.0001766307427577189, + "loss": 0.2614, + "step": 9482 + }, + { + "epoch": 0.22229645387577138, + "grad_norm": 6.357079982757568, + "learning_rate": 0.00017662600998444457, + "loss": 0.5424, + "step": 9483 + }, + { + "epoch": 0.22231989545057637, + "grad_norm": 2.539450168609619, + "learning_rate": 0.00017662127679539538, + "loss": 0.2883, + "step": 9484 + }, + { + "epoch": 0.22234333702538137, + "grad_norm": 1.4211851358413696, + "learning_rate": 0.00017661654319059693, + "loss": 0.1715, + "step": 9485 + }, + { + "epoch": 0.22236677860018636, + "grad_norm": 4.980060577392578, + "learning_rate": 0.000176611809170075, + "loss": 0.64, + "step": 9486 + }, + { + "epoch": 0.22239022017499135, + "grad_norm": 0.47556087374687195, + "learning_rate": 0.0001766070747338552, + "loss": 0.0511, + "step": 9487 + }, + { + "epoch": 0.22241366174979635, + "grad_norm": 2.022810220718384, + "learning_rate": 0.00017660233988196327, + "loss": 0.2068, + "step": 9488 + }, + { + "epoch": 0.22243710332460134, + "grad_norm": 5.783531665802002, + "learning_rate": 0.00017659760461442492, + "loss": 0.7797, + "step": 9489 + }, + { + "epoch": 0.22246054489940634, + "grad_norm": 0.41539838910102844, + "learning_rate": 0.0001765928689312658, + "loss": 0.0322, + "step": 9490 + }, + { + "epoch": 0.22248398647421133, + "grad_norm": 1.146157145500183, + "learning_rate": 0.00017658813283251158, + "loss": 0.1481, + "step": 9491 + }, + { + "epoch": 0.22250742804901633, + "grad_norm": 3.5572800636291504, + "learning_rate": 0.00017658339631818802, + "loss": 0.5246, + "step": 9492 + }, + { + "epoch": 0.22253086962382132, + "grad_norm": 3.0740721225738525, + "learning_rate": 0.00017657865938832084, + "loss": 0.3211, + "step": 9493 + }, + { + "epoch": 0.22255431119862631, + "grad_norm": 4.119112491607666, + "learning_rate": 0.00017657392204293565, + "loss": 0.257, + "step": 9494 + }, + { + "epoch": 0.2225777527734313, + "grad_norm": 4.872756004333496, + "learning_rate": 0.00017656918428205824, + "loss": 0.4925, + "step": 9495 + }, + { + "epoch": 0.2226011943482363, + "grad_norm": 4.298824787139893, + "learning_rate": 0.00017656444610571428, + "loss": 0.5559, + "step": 9496 + }, + { + "epoch": 0.2226246359230413, + "grad_norm": 3.1542892456054688, + "learning_rate": 0.00017655970751392946, + "loss": 0.2476, + "step": 9497 + }, + { + "epoch": 0.2226480774978463, + "grad_norm": 4.590725898742676, + "learning_rate": 0.00017655496850672953, + "loss": 0.4945, + "step": 9498 + }, + { + "epoch": 0.2226715190726513, + "grad_norm": 1.0941636562347412, + "learning_rate": 0.00017655022908414018, + "loss": 0.1232, + "step": 9499 + }, + { + "epoch": 0.2226949606474563, + "grad_norm": 3.229300022125244, + "learning_rate": 0.00017654548924618714, + "loss": 0.5409, + "step": 9500 + }, + { + "epoch": 0.2227184022222613, + "grad_norm": 4.130615711212158, + "learning_rate": 0.00017654074899289612, + "loss": 0.3759, + "step": 9501 + }, + { + "epoch": 0.2227418437970663, + "grad_norm": 2.6936864852905273, + "learning_rate": 0.00017653600832429284, + "loss": 0.254, + "step": 9502 + }, + { + "epoch": 0.2227652853718713, + "grad_norm": 3.630629301071167, + "learning_rate": 0.00017653126724040305, + "loss": 0.4886, + "step": 9503 + }, + { + "epoch": 0.22278872694667629, + "grad_norm": 2.464172601699829, + "learning_rate": 0.0001765265257412524, + "loss": 0.328, + "step": 9504 + }, + { + "epoch": 0.22281216852148128, + "grad_norm": 0.7886558175086975, + "learning_rate": 0.00017652178382686673, + "loss": 0.0783, + "step": 9505 + }, + { + "epoch": 0.22283561009628627, + "grad_norm": 2.5769472122192383, + "learning_rate": 0.00017651704149727171, + "loss": 0.292, + "step": 9506 + }, + { + "epoch": 0.22285905167109127, + "grad_norm": 4.783803462982178, + "learning_rate": 0.00017651229875249302, + "loss": 0.3988, + "step": 9507 + }, + { + "epoch": 0.22288249324589626, + "grad_norm": 0.7147454619407654, + "learning_rate": 0.00017650755559255648, + "loss": 0.0935, + "step": 9508 + }, + { + "epoch": 0.22290593482070126, + "grad_norm": 1.3028610944747925, + "learning_rate": 0.00017650281201748777, + "loss": 0.1946, + "step": 9509 + }, + { + "epoch": 0.22292937639550625, + "grad_norm": 3.0056793689727783, + "learning_rate": 0.00017649806802731268, + "loss": 1.4735, + "step": 9510 + }, + { + "epoch": 0.22295281797031125, + "grad_norm": 5.396501064300537, + "learning_rate": 0.0001764933236220569, + "loss": 0.7208, + "step": 9511 + }, + { + "epoch": 0.22297625954511624, + "grad_norm": 0.8220706582069397, + "learning_rate": 0.0001764885788017462, + "loss": 0.083, + "step": 9512 + }, + { + "epoch": 0.22299970111992123, + "grad_norm": 1.6966543197631836, + "learning_rate": 0.0001764838335664063, + "loss": 0.1896, + "step": 9513 + }, + { + "epoch": 0.22302314269472623, + "grad_norm": 3.795574426651001, + "learning_rate": 0.00017647908791606302, + "loss": 0.5398, + "step": 9514 + }, + { + "epoch": 0.22304658426953122, + "grad_norm": 2.2146403789520264, + "learning_rate": 0.000176474341850742, + "loss": 0.2141, + "step": 9515 + }, + { + "epoch": 0.22307002584433622, + "grad_norm": 3.330007314682007, + "learning_rate": 0.0001764695953704691, + "loss": 0.3789, + "step": 9516 + }, + { + "epoch": 0.2230934674191412, + "grad_norm": 1.5121766328811646, + "learning_rate": 0.00017646484847527, + "loss": 0.2249, + "step": 9517 + }, + { + "epoch": 0.2231169089939462, + "grad_norm": 1.06991708278656, + "learning_rate": 0.00017646010116517048, + "loss": 0.0921, + "step": 9518 + }, + { + "epoch": 0.2231403505687512, + "grad_norm": 2.930628538131714, + "learning_rate": 0.0001764553534401963, + "loss": 0.2841, + "step": 9519 + }, + { + "epoch": 0.2231637921435562, + "grad_norm": 1.238290786743164, + "learning_rate": 0.00017645060530037325, + "loss": 0.1095, + "step": 9520 + }, + { + "epoch": 0.2231872337183612, + "grad_norm": 2.596048355102539, + "learning_rate": 0.00017644585674572705, + "loss": 0.3241, + "step": 9521 + }, + { + "epoch": 0.22321067529316618, + "grad_norm": 2.46893048286438, + "learning_rate": 0.00017644110777628348, + "loss": 0.2426, + "step": 9522 + }, + { + "epoch": 0.22323411686797118, + "grad_norm": 1.2907739877700806, + "learning_rate": 0.0001764363583920683, + "loss": 0.1328, + "step": 9523 + }, + { + "epoch": 0.2232575584427762, + "grad_norm": 2.690446376800537, + "learning_rate": 0.0001764316085931073, + "loss": 0.3375, + "step": 9524 + }, + { + "epoch": 0.2232810000175812, + "grad_norm": 4.032814025878906, + "learning_rate": 0.00017642685837942622, + "loss": 0.4054, + "step": 9525 + }, + { + "epoch": 0.2233044415923862, + "grad_norm": 2.754598617553711, + "learning_rate": 0.0001764221077510509, + "loss": 0.995, + "step": 9526 + }, + { + "epoch": 0.22332788316719118, + "grad_norm": 3.6108133792877197, + "learning_rate": 0.00017641735670800705, + "loss": 0.5368, + "step": 9527 + }, + { + "epoch": 0.22335132474199618, + "grad_norm": 4.350020408630371, + "learning_rate": 0.0001764126052503205, + "loss": 0.37, + "step": 9528 + }, + { + "epoch": 0.22337476631680117, + "grad_norm": 2.7012131214141846, + "learning_rate": 0.00017640785337801694, + "loss": 0.4289, + "step": 9529 + }, + { + "epoch": 0.22339820789160617, + "grad_norm": 1.7993690967559814, + "learning_rate": 0.0001764031010911223, + "loss": 1.1107, + "step": 9530 + }, + { + "epoch": 0.22342164946641116, + "grad_norm": 3.7535057067871094, + "learning_rate": 0.00017639834838966225, + "loss": 0.6023, + "step": 9531 + }, + { + "epoch": 0.22344509104121615, + "grad_norm": 3.427419900894165, + "learning_rate": 0.00017639359527366261, + "loss": 0.2002, + "step": 9532 + }, + { + "epoch": 0.22346853261602115, + "grad_norm": 3.3283772468566895, + "learning_rate": 0.00017638884174314917, + "loss": 1.3838, + "step": 9533 + }, + { + "epoch": 0.22349197419082614, + "grad_norm": 0.7527942061424255, + "learning_rate": 0.00017638408779814774, + "loss": 0.0606, + "step": 9534 + }, + { + "epoch": 0.22351541576563114, + "grad_norm": 2.8070342540740967, + "learning_rate": 0.00017637933343868408, + "loss": 0.3007, + "step": 9535 + }, + { + "epoch": 0.22353885734043613, + "grad_norm": 2.8917489051818848, + "learning_rate": 0.00017637457866478403, + "loss": 0.3043, + "step": 9536 + }, + { + "epoch": 0.22356229891524113, + "grad_norm": 5.214942455291748, + "learning_rate": 0.00017636982347647338, + "loss": 0.7641, + "step": 9537 + }, + { + "epoch": 0.22358574049004612, + "grad_norm": 2.9985671043395996, + "learning_rate": 0.00017636506787377786, + "loss": 0.2953, + "step": 9538 + }, + { + "epoch": 0.22360918206485111, + "grad_norm": 2.443506956100464, + "learning_rate": 0.0001763603118567234, + "loss": 0.2647, + "step": 9539 + }, + { + "epoch": 0.2236326236396561, + "grad_norm": 3.3735785484313965, + "learning_rate": 0.00017635555542533572, + "loss": 0.5483, + "step": 9540 + }, + { + "epoch": 0.2236560652144611, + "grad_norm": 1.3766158819198608, + "learning_rate": 0.00017635079857964065, + "loss": 0.2383, + "step": 9541 + }, + { + "epoch": 0.2236795067892661, + "grad_norm": 3.6673715114593506, + "learning_rate": 0.000176346041319664, + "loss": 0.4438, + "step": 9542 + }, + { + "epoch": 0.2237029483640711, + "grad_norm": 3.714938163757324, + "learning_rate": 0.00017634128364543158, + "loss": 0.3028, + "step": 9543 + }, + { + "epoch": 0.22372638993887609, + "grad_norm": 3.327895402908325, + "learning_rate": 0.0001763365255569692, + "loss": 0.3341, + "step": 9544 + }, + { + "epoch": 0.22374983151368108, + "grad_norm": 3.3185622692108154, + "learning_rate": 0.00017633176705430267, + "loss": 1.4991, + "step": 9545 + }, + { + "epoch": 0.22377327308848607, + "grad_norm": 4.384639739990234, + "learning_rate": 0.00017632700813745785, + "loss": 0.4879, + "step": 9546 + }, + { + "epoch": 0.22379671466329107, + "grad_norm": 2.8098909854888916, + "learning_rate": 0.00017632224880646057, + "loss": 0.2217, + "step": 9547 + }, + { + "epoch": 0.22382015623809606, + "grad_norm": 2.0918021202087402, + "learning_rate": 0.00017631748906133657, + "loss": 0.3516, + "step": 9548 + }, + { + "epoch": 0.22384359781290106, + "grad_norm": 2.3806042671203613, + "learning_rate": 0.00017631272890211177, + "loss": 0.2757, + "step": 9549 + }, + { + "epoch": 0.22386703938770608, + "grad_norm": 1.633609652519226, + "learning_rate": 0.00017630796832881191, + "loss": 0.2886, + "step": 9550 + }, + { + "epoch": 0.22389048096251107, + "grad_norm": 4.579191207885742, + "learning_rate": 0.0001763032073414629, + "loss": 0.6174, + "step": 9551 + }, + { + "epoch": 0.22391392253731607, + "grad_norm": 3.608792781829834, + "learning_rate": 0.00017629844594009052, + "loss": 0.4818, + "step": 9552 + }, + { + "epoch": 0.22393736411212106, + "grad_norm": 2.5467827320098877, + "learning_rate": 0.00017629368412472066, + "loss": 0.1857, + "step": 9553 + }, + { + "epoch": 0.22396080568692606, + "grad_norm": 4.505461692810059, + "learning_rate": 0.0001762889218953791, + "loss": 0.5296, + "step": 9554 + }, + { + "epoch": 0.22398424726173105, + "grad_norm": 2.747990131378174, + "learning_rate": 0.0001762841592520917, + "loss": 0.2998, + "step": 9555 + }, + { + "epoch": 0.22400768883653605, + "grad_norm": 3.9835009574890137, + "learning_rate": 0.00017627939619488432, + "loss": 0.7695, + "step": 9556 + }, + { + "epoch": 0.22403113041134104, + "grad_norm": 2.2050516605377197, + "learning_rate": 0.0001762746327237828, + "loss": 0.1863, + "step": 9557 + }, + { + "epoch": 0.22405457198614603, + "grad_norm": 3.982494831085205, + "learning_rate": 0.000176269868838813, + "loss": 0.4327, + "step": 9558 + }, + { + "epoch": 0.22407801356095103, + "grad_norm": 2.3587887287139893, + "learning_rate": 0.0001762651045400007, + "loss": 0.1808, + "step": 9559 + }, + { + "epoch": 0.22410145513575602, + "grad_norm": 3.9348654747009277, + "learning_rate": 0.0001762603398273718, + "loss": 0.6444, + "step": 9560 + }, + { + "epoch": 0.22412489671056102, + "grad_norm": 5.530002117156982, + "learning_rate": 0.00017625557470095216, + "loss": 0.8348, + "step": 9561 + }, + { + "epoch": 0.224148338285366, + "grad_norm": 3.753384590148926, + "learning_rate": 0.00017625080916076763, + "loss": 0.3938, + "step": 9562 + }, + { + "epoch": 0.224171779860171, + "grad_norm": 5.616943359375, + "learning_rate": 0.00017624604320684407, + "loss": 0.6233, + "step": 9563 + }, + { + "epoch": 0.224195221434976, + "grad_norm": 5.56081485748291, + "learning_rate": 0.00017624127683920734, + "loss": 0.6177, + "step": 9564 + }, + { + "epoch": 0.224218663009781, + "grad_norm": 3.4683384895324707, + "learning_rate": 0.0001762365100578833, + "loss": 0.4473, + "step": 9565 + }, + { + "epoch": 0.224242104584586, + "grad_norm": 1.6801568269729614, + "learning_rate": 0.0001762317428628978, + "loss": 1.1934, + "step": 9566 + }, + { + "epoch": 0.22426554615939098, + "grad_norm": 5.5203633308410645, + "learning_rate": 0.00017622697525427673, + "loss": 0.3416, + "step": 9567 + }, + { + "epoch": 0.22428898773419598, + "grad_norm": 1.2583835124969482, + "learning_rate": 0.00017622220723204594, + "loss": 0.1852, + "step": 9568 + }, + { + "epoch": 0.22431242930900097, + "grad_norm": 4.020737648010254, + "learning_rate": 0.0001762174387962313, + "loss": 0.5616, + "step": 9569 + }, + { + "epoch": 0.22433587088380597, + "grad_norm": 3.281731367111206, + "learning_rate": 0.00017621266994685873, + "loss": 0.3346, + "step": 9570 + }, + { + "epoch": 0.22435931245861096, + "grad_norm": 4.518829345703125, + "learning_rate": 0.00017620790068395404, + "loss": 0.3892, + "step": 9571 + }, + { + "epoch": 0.22438275403341595, + "grad_norm": 4.08638334274292, + "learning_rate": 0.00017620313100754316, + "loss": 0.5224, + "step": 9572 + }, + { + "epoch": 0.22440619560822095, + "grad_norm": 3.7656354904174805, + "learning_rate": 0.00017619836091765193, + "loss": 0.2861, + "step": 9573 + }, + { + "epoch": 0.22442963718302594, + "grad_norm": 3.8102753162384033, + "learning_rate": 0.00017619359041430626, + "loss": 0.4092, + "step": 9574 + }, + { + "epoch": 0.22445307875783094, + "grad_norm": 1.6831790208816528, + "learning_rate": 0.000176188819497532, + "loss": 0.3013, + "step": 9575 + }, + { + "epoch": 0.22447652033263596, + "grad_norm": 3.270087718963623, + "learning_rate": 0.00017618404816735513, + "loss": 0.2554, + "step": 9576 + }, + { + "epoch": 0.22449996190744095, + "grad_norm": 1.3503849506378174, + "learning_rate": 0.0001761792764238014, + "loss": 0.201, + "step": 9577 + }, + { + "epoch": 0.22452340348224595, + "grad_norm": 2.3284010887145996, + "learning_rate": 0.0001761745042668968, + "loss": 1.2727, + "step": 9578 + }, + { + "epoch": 0.22454684505705094, + "grad_norm": 1.797737956047058, + "learning_rate": 0.0001761697316966672, + "loss": 0.1771, + "step": 9579 + }, + { + "epoch": 0.22457028663185594, + "grad_norm": 3.367799997329712, + "learning_rate": 0.0001761649587131385, + "loss": 0.3465, + "step": 9580 + }, + { + "epoch": 0.22459372820666093, + "grad_norm": 3.6950623989105225, + "learning_rate": 0.0001761601853163366, + "loss": 0.4751, + "step": 9581 + }, + { + "epoch": 0.22461716978146593, + "grad_norm": 1.4437977075576782, + "learning_rate": 0.00017615541150628735, + "loss": 0.1656, + "step": 9582 + }, + { + "epoch": 0.22464061135627092, + "grad_norm": 3.4979474544525146, + "learning_rate": 0.00017615063728301672, + "loss": 0.574, + "step": 9583 + }, + { + "epoch": 0.2246640529310759, + "grad_norm": 0.3752288818359375, + "learning_rate": 0.00017614586264655063, + "loss": 0.0554, + "step": 9584 + }, + { + "epoch": 0.2246874945058809, + "grad_norm": 0.7064067721366882, + "learning_rate": 0.0001761410875969149, + "loss": 0.0698, + "step": 9585 + }, + { + "epoch": 0.2247109360806859, + "grad_norm": 0.7301836609840393, + "learning_rate": 0.0001761363121341355, + "loss": 0.0648, + "step": 9586 + }, + { + "epoch": 0.2247343776554909, + "grad_norm": 5.392033576965332, + "learning_rate": 0.00017613153625823836, + "loss": 0.8708, + "step": 9587 + }, + { + "epoch": 0.2247578192302959, + "grad_norm": 3.3154146671295166, + "learning_rate": 0.00017612675996924932, + "loss": 0.4535, + "step": 9588 + }, + { + "epoch": 0.22478126080510089, + "grad_norm": 2.3887124061584473, + "learning_rate": 0.00017612198326719437, + "loss": 1.6376, + "step": 9589 + }, + { + "epoch": 0.22480470237990588, + "grad_norm": 3.3965272903442383, + "learning_rate": 0.00017611720615209935, + "loss": 0.3091, + "step": 9590 + }, + { + "epoch": 0.22482814395471087, + "grad_norm": 4.400455951690674, + "learning_rate": 0.00017611242862399027, + "loss": 0.6581, + "step": 9591 + }, + { + "epoch": 0.22485158552951587, + "grad_norm": 0.9573008418083191, + "learning_rate": 0.000176107650682893, + "loss": 0.0945, + "step": 9592 + }, + { + "epoch": 0.22487502710432086, + "grad_norm": 1.1052939891815186, + "learning_rate": 0.0001761028723288335, + "loss": 0.1189, + "step": 9593 + }, + { + "epoch": 0.22489846867912586, + "grad_norm": 2.5999739170074463, + "learning_rate": 0.00017609809356183766, + "loss": 0.3234, + "step": 9594 + }, + { + "epoch": 0.22492191025393085, + "grad_norm": 4.705221652984619, + "learning_rate": 0.0001760933143819314, + "loss": 0.6809, + "step": 9595 + }, + { + "epoch": 0.22494535182873585, + "grad_norm": 3.311014175415039, + "learning_rate": 0.0001760885347891407, + "loss": 0.4022, + "step": 9596 + }, + { + "epoch": 0.22496879340354084, + "grad_norm": 3.9151394367218018, + "learning_rate": 0.0001760837547834915, + "loss": 0.4877, + "step": 9597 + }, + { + "epoch": 0.22499223497834583, + "grad_norm": 3.9919400215148926, + "learning_rate": 0.00017607897436500966, + "loss": 0.2952, + "step": 9598 + }, + { + "epoch": 0.22501567655315083, + "grad_norm": 2.3461506366729736, + "learning_rate": 0.00017607419353372116, + "loss": 0.2578, + "step": 9599 + }, + { + "epoch": 0.22503911812795582, + "grad_norm": 3.5994675159454346, + "learning_rate": 0.000176069412289652, + "loss": 0.4139, + "step": 9600 + }, + { + "epoch": 0.22506255970276084, + "grad_norm": 0.6415466070175171, + "learning_rate": 0.00017606463063282802, + "loss": 0.0412, + "step": 9601 + }, + { + "epoch": 0.22508600127756584, + "grad_norm": 3.895587205886841, + "learning_rate": 0.00017605984856327523, + "loss": 1.6045, + "step": 9602 + }, + { + "epoch": 0.22510944285237083, + "grad_norm": 3.962325096130371, + "learning_rate": 0.00017605506608101958, + "loss": 0.4391, + "step": 9603 + }, + { + "epoch": 0.22513288442717583, + "grad_norm": 4.1279191970825195, + "learning_rate": 0.000176050283186087, + "loss": 0.4991, + "step": 9604 + }, + { + "epoch": 0.22515632600198082, + "grad_norm": 2.7206554412841797, + "learning_rate": 0.00017604549987850344, + "loss": 1.7665, + "step": 9605 + }, + { + "epoch": 0.22517976757678582, + "grad_norm": 5.080018997192383, + "learning_rate": 0.00017604071615829486, + "loss": 0.7534, + "step": 9606 + }, + { + "epoch": 0.2252032091515908, + "grad_norm": 2.3885035514831543, + "learning_rate": 0.00017603593202548717, + "loss": 0.1625, + "step": 9607 + }, + { + "epoch": 0.2252266507263958, + "grad_norm": 4.423253059387207, + "learning_rate": 0.00017603114748010644, + "loss": 0.4639, + "step": 9608 + }, + { + "epoch": 0.2252500923012008, + "grad_norm": 3.0465309619903564, + "learning_rate": 0.00017602636252217856, + "loss": 1.7661, + "step": 9609 + }, + { + "epoch": 0.2252735338760058, + "grad_norm": 4.394673824310303, + "learning_rate": 0.00017602157715172946, + "loss": 0.372, + "step": 9610 + }, + { + "epoch": 0.2252969754508108, + "grad_norm": 3.8085122108459473, + "learning_rate": 0.00017601679136878518, + "loss": 0.4839, + "step": 9611 + }, + { + "epoch": 0.22532041702561578, + "grad_norm": 3.0400681495666504, + "learning_rate": 0.00017601200517337165, + "loss": 0.3894, + "step": 9612 + }, + { + "epoch": 0.22534385860042078, + "grad_norm": 3.8292951583862305, + "learning_rate": 0.00017600721856551483, + "loss": 0.3551, + "step": 9613 + }, + { + "epoch": 0.22536730017522577, + "grad_norm": 0.9210419058799744, + "learning_rate": 0.0001760024315452407, + "loss": 0.1258, + "step": 9614 + }, + { + "epoch": 0.22539074175003077, + "grad_norm": 3.703834295272827, + "learning_rate": 0.00017599764411257524, + "loss": 0.4635, + "step": 9615 + }, + { + "epoch": 0.22541418332483576, + "grad_norm": 4.204129219055176, + "learning_rate": 0.00017599285626754445, + "loss": 0.5713, + "step": 9616 + }, + { + "epoch": 0.22543762489964075, + "grad_norm": 2.8650529384613037, + "learning_rate": 0.00017598806801017426, + "loss": 0.3452, + "step": 9617 + }, + { + "epoch": 0.22546106647444575, + "grad_norm": 1.3016471862792969, + "learning_rate": 0.00017598327934049068, + "loss": 0.2741, + "step": 9618 + }, + { + "epoch": 0.22548450804925074, + "grad_norm": 4.396504878997803, + "learning_rate": 0.0001759784902585197, + "loss": 0.4077, + "step": 9619 + }, + { + "epoch": 0.22550794962405574, + "grad_norm": 3.2075531482696533, + "learning_rate": 0.0001759737007642873, + "loss": 0.3828, + "step": 9620 + }, + { + "epoch": 0.22553139119886073, + "grad_norm": 3.858039140701294, + "learning_rate": 0.00017596891085781943, + "loss": 0.5781, + "step": 9621 + }, + { + "epoch": 0.22555483277366573, + "grad_norm": 6.315362930297852, + "learning_rate": 0.00017596412053914214, + "loss": 0.4686, + "step": 9622 + }, + { + "epoch": 0.22557827434847072, + "grad_norm": 5.67409610748291, + "learning_rate": 0.0001759593298082814, + "loss": 0.3773, + "step": 9623 + }, + { + "epoch": 0.22560171592327571, + "grad_norm": 3.744018793106079, + "learning_rate": 0.0001759545386652632, + "loss": 0.3767, + "step": 9624 + }, + { + "epoch": 0.2256251574980807, + "grad_norm": 1.5481631755828857, + "learning_rate": 0.00017594974711011353, + "loss": 0.0991, + "step": 9625 + }, + { + "epoch": 0.2256485990728857, + "grad_norm": 3.5155694484710693, + "learning_rate": 0.00017594495514285837, + "loss": 0.4262, + "step": 9626 + }, + { + "epoch": 0.22567204064769072, + "grad_norm": 4.664625644683838, + "learning_rate": 0.0001759401627635238, + "loss": 0.6702, + "step": 9627 + }, + { + "epoch": 0.22569548222249572, + "grad_norm": 2.924626350402832, + "learning_rate": 0.00017593536997213575, + "loss": 0.3717, + "step": 9628 + }, + { + "epoch": 0.2257189237973007, + "grad_norm": 1.6425894498825073, + "learning_rate": 0.00017593057676872025, + "loss": 1.1231, + "step": 9629 + }, + { + "epoch": 0.2257423653721057, + "grad_norm": 2.268681287765503, + "learning_rate": 0.0001759257831533033, + "loss": 0.4943, + "step": 9630 + }, + { + "epoch": 0.2257658069469107, + "grad_norm": 2.523632764816284, + "learning_rate": 0.0001759209891259109, + "loss": 0.4149, + "step": 9631 + }, + { + "epoch": 0.2257892485217157, + "grad_norm": 1.1642584800720215, + "learning_rate": 0.0001759161946865691, + "loss": 0.1741, + "step": 9632 + }, + { + "epoch": 0.2258126900965207, + "grad_norm": 1.4702134132385254, + "learning_rate": 0.0001759113998353039, + "loss": 0.0766, + "step": 9633 + }, + { + "epoch": 0.22583613167132568, + "grad_norm": 3.168189287185669, + "learning_rate": 0.0001759066045721413, + "loss": 0.1858, + "step": 9634 + }, + { + "epoch": 0.22585957324613068, + "grad_norm": 0.6311425566673279, + "learning_rate": 0.00017590180889710732, + "loss": 0.047, + "step": 9635 + }, + { + "epoch": 0.22588301482093567, + "grad_norm": 3.6063263416290283, + "learning_rate": 0.000175897012810228, + "loss": 0.3687, + "step": 9636 + }, + { + "epoch": 0.22590645639574067, + "grad_norm": 4.403641700744629, + "learning_rate": 0.00017589221631152935, + "loss": 0.5861, + "step": 9637 + }, + { + "epoch": 0.22592989797054566, + "grad_norm": 3.8400797843933105, + "learning_rate": 0.00017588741940103742, + "loss": 0.5037, + "step": 9638 + }, + { + "epoch": 0.22595333954535066, + "grad_norm": 3.491088628768921, + "learning_rate": 0.0001758826220787782, + "loss": 0.2993, + "step": 9639 + }, + { + "epoch": 0.22597678112015565, + "grad_norm": 3.873422145843506, + "learning_rate": 0.0001758778243447777, + "loss": 0.4014, + "step": 9640 + }, + { + "epoch": 0.22600022269496065, + "grad_norm": 0.8917987942695618, + "learning_rate": 0.00017587302619906205, + "loss": 0.0944, + "step": 9641 + }, + { + "epoch": 0.22602366426976564, + "grad_norm": 1.9699227809906006, + "learning_rate": 0.0001758682276416572, + "loss": 0.1886, + "step": 9642 + }, + { + "epoch": 0.22604710584457063, + "grad_norm": 3.8798909187316895, + "learning_rate": 0.0001758634286725892, + "loss": 0.2563, + "step": 9643 + }, + { + "epoch": 0.22607054741937563, + "grad_norm": 6.303243637084961, + "learning_rate": 0.00017585862929188416, + "loss": 0.2923, + "step": 9644 + }, + { + "epoch": 0.22609398899418062, + "grad_norm": 4.672501564025879, + "learning_rate": 0.000175853829499568, + "loss": 0.619, + "step": 9645 + }, + { + "epoch": 0.22611743056898562, + "grad_norm": 3.7041711807250977, + "learning_rate": 0.0001758490292956669, + "loss": 0.3126, + "step": 9646 + }, + { + "epoch": 0.2261408721437906, + "grad_norm": 3.620699882507324, + "learning_rate": 0.00017584422868020678, + "loss": 0.3004, + "step": 9647 + }, + { + "epoch": 0.2261643137185956, + "grad_norm": 0.8314475417137146, + "learning_rate": 0.00017583942765321374, + "loss": 0.0983, + "step": 9648 + }, + { + "epoch": 0.2261877552934006, + "grad_norm": 5.041469097137451, + "learning_rate": 0.00017583462621471388, + "loss": 0.3446, + "step": 9649 + }, + { + "epoch": 0.2262111968682056, + "grad_norm": 2.1683154106140137, + "learning_rate": 0.00017582982436473317, + "loss": 0.294, + "step": 9650 + }, + { + "epoch": 0.2262346384430106, + "grad_norm": 1.521804928779602, + "learning_rate": 0.00017582502210329769, + "loss": 0.2082, + "step": 9651 + }, + { + "epoch": 0.2262580800178156, + "grad_norm": 4.8588480949401855, + "learning_rate": 0.00017582021943043352, + "loss": 0.53, + "step": 9652 + }, + { + "epoch": 0.2262815215926206, + "grad_norm": 2.9443345069885254, + "learning_rate": 0.00017581541634616673, + "loss": 1.352, + "step": 9653 + }, + { + "epoch": 0.2263049631674256, + "grad_norm": 3.948054313659668, + "learning_rate": 0.00017581061285052333, + "loss": 0.5138, + "step": 9654 + }, + { + "epoch": 0.2263284047422306, + "grad_norm": 2.9485232830047607, + "learning_rate": 0.0001758058089435294, + "loss": 0.2171, + "step": 9655 + }, + { + "epoch": 0.2263518463170356, + "grad_norm": 4.170777797698975, + "learning_rate": 0.00017580100462521105, + "loss": 0.2679, + "step": 9656 + }, + { + "epoch": 0.22637528789184058, + "grad_norm": 2.8671069145202637, + "learning_rate": 0.00017579619989559434, + "loss": 0.3739, + "step": 9657 + }, + { + "epoch": 0.22639872946664558, + "grad_norm": 1.143437147140503, + "learning_rate": 0.00017579139475470528, + "loss": 0.1113, + "step": 9658 + }, + { + "epoch": 0.22642217104145057, + "grad_norm": 7.1081647872924805, + "learning_rate": 0.00017578658920256997, + "loss": 0.6517, + "step": 9659 + }, + { + "epoch": 0.22644561261625556, + "grad_norm": 5.642260551452637, + "learning_rate": 0.00017578178323921452, + "loss": 0.9359, + "step": 9660 + }, + { + "epoch": 0.22646905419106056, + "grad_norm": 0.8438354134559631, + "learning_rate": 0.00017577697686466497, + "loss": 0.0956, + "step": 9661 + }, + { + "epoch": 0.22649249576586555, + "grad_norm": 0.8255476355552673, + "learning_rate": 0.0001757721700789474, + "loss": 0.048, + "step": 9662 + }, + { + "epoch": 0.22651593734067055, + "grad_norm": 5.343852519989014, + "learning_rate": 0.00017576736288208794, + "loss": 0.6563, + "step": 9663 + }, + { + "epoch": 0.22653937891547554, + "grad_norm": 3.924098253250122, + "learning_rate": 0.00017576255527411262, + "loss": 0.5887, + "step": 9664 + }, + { + "epoch": 0.22656282049028054, + "grad_norm": 3.035616874694824, + "learning_rate": 0.00017575774725504757, + "loss": 1.719, + "step": 9665 + }, + { + "epoch": 0.22658626206508553, + "grad_norm": 4.397839546203613, + "learning_rate": 0.00017575293882491881, + "loss": 0.3668, + "step": 9666 + }, + { + "epoch": 0.22660970363989053, + "grad_norm": 4.517941474914551, + "learning_rate": 0.0001757481299837525, + "loss": 0.5971, + "step": 9667 + }, + { + "epoch": 0.22663314521469552, + "grad_norm": 2.0730931758880615, + "learning_rate": 0.00017574332073157472, + "loss": 0.3046, + "step": 9668 + }, + { + "epoch": 0.2266565867895005, + "grad_norm": 2.71124267578125, + "learning_rate": 0.0001757385110684115, + "loss": 0.2057, + "step": 9669 + }, + { + "epoch": 0.2266800283643055, + "grad_norm": 4.8838300704956055, + "learning_rate": 0.00017573370099428906, + "loss": 0.4514, + "step": 9670 + }, + { + "epoch": 0.2267034699391105, + "grad_norm": 1.887589454650879, + "learning_rate": 0.00017572889050923335, + "loss": 0.1869, + "step": 9671 + }, + { + "epoch": 0.2267269115139155, + "grad_norm": 1.6771142482757568, + "learning_rate": 0.0001757240796132706, + "loss": 0.1442, + "step": 9672 + }, + { + "epoch": 0.2267503530887205, + "grad_norm": 1.4257725477218628, + "learning_rate": 0.00017571926830642687, + "loss": 0.1987, + "step": 9673 + }, + { + "epoch": 0.22677379466352549, + "grad_norm": 2.5116915702819824, + "learning_rate": 0.00017571445658872824, + "loss": 0.164, + "step": 9674 + }, + { + "epoch": 0.22679723623833048, + "grad_norm": 12.59536075592041, + "learning_rate": 0.00017570964446020088, + "loss": 0.6122, + "step": 9675 + }, + { + "epoch": 0.22682067781313547, + "grad_norm": 2.4549193382263184, + "learning_rate": 0.0001757048319208708, + "loss": 0.247, + "step": 9676 + }, + { + "epoch": 0.22684411938794047, + "grad_norm": 0.568520724773407, + "learning_rate": 0.0001757000189707642, + "loss": 0.0845, + "step": 9677 + }, + { + "epoch": 0.2268675609627455, + "grad_norm": 6.7254109382629395, + "learning_rate": 0.00017569520560990718, + "loss": 0.6849, + "step": 9678 + }, + { + "epoch": 0.22689100253755048, + "grad_norm": 0.45199233293533325, + "learning_rate": 0.00017569039183832586, + "loss": 0.0747, + "step": 9679 + }, + { + "epoch": 0.22691444411235548, + "grad_norm": 2.7325708866119385, + "learning_rate": 0.00017568557765604628, + "loss": 0.7898, + "step": 9680 + }, + { + "epoch": 0.22693788568716047, + "grad_norm": 7.127826690673828, + "learning_rate": 0.0001756807630630947, + "loss": 0.4463, + "step": 9681 + }, + { + "epoch": 0.22696132726196547, + "grad_norm": 4.6052398681640625, + "learning_rate": 0.00017567594805949715, + "loss": 0.3614, + "step": 9682 + }, + { + "epoch": 0.22698476883677046, + "grad_norm": 4.382532596588135, + "learning_rate": 0.00017567113264527976, + "loss": 0.5706, + "step": 9683 + }, + { + "epoch": 0.22700821041157546, + "grad_norm": 4.231978416442871, + "learning_rate": 0.00017566631682046867, + "loss": 0.3593, + "step": 9684 + }, + { + "epoch": 0.22703165198638045, + "grad_norm": 1.419419765472412, + "learning_rate": 0.00017566150058509006, + "loss": 0.2152, + "step": 9685 + }, + { + "epoch": 0.22705509356118544, + "grad_norm": 1.527647852897644, + "learning_rate": 0.00017565668393916997, + "loss": 0.2141, + "step": 9686 + }, + { + "epoch": 0.22707853513599044, + "grad_norm": 1.3708000183105469, + "learning_rate": 0.00017565186688273463, + "loss": 0.1801, + "step": 9687 + }, + { + "epoch": 0.22710197671079543, + "grad_norm": 2.9673643112182617, + "learning_rate": 0.00017564704941581012, + "loss": 0.291, + "step": 9688 + }, + { + "epoch": 0.22712541828560043, + "grad_norm": 1.562644362449646, + "learning_rate": 0.00017564223153842257, + "loss": 0.2481, + "step": 9689 + }, + { + "epoch": 0.22714885986040542, + "grad_norm": 3.533859968185425, + "learning_rate": 0.00017563741325059818, + "loss": 0.3944, + "step": 9690 + }, + { + "epoch": 0.22717230143521042, + "grad_norm": 3.4988396167755127, + "learning_rate": 0.00017563259455236302, + "loss": 0.2511, + "step": 9691 + }, + { + "epoch": 0.2271957430100154, + "grad_norm": 4.691984176635742, + "learning_rate": 0.00017562777544374332, + "loss": 0.6325, + "step": 9692 + }, + { + "epoch": 0.2272191845848204, + "grad_norm": 3.113992929458618, + "learning_rate": 0.00017562295592476513, + "loss": 0.2912, + "step": 9693 + }, + { + "epoch": 0.2272426261596254, + "grad_norm": 2.210906744003296, + "learning_rate": 0.0001756181359954547, + "loss": 0.2744, + "step": 9694 + }, + { + "epoch": 0.2272660677344304, + "grad_norm": 6.428399562835693, + "learning_rate": 0.0001756133156558381, + "loss": 0.5606, + "step": 9695 + }, + { + "epoch": 0.2272895093092354, + "grad_norm": 1.4970844984054565, + "learning_rate": 0.00017560849490594156, + "loss": 0.1692, + "step": 9696 + }, + { + "epoch": 0.22731295088404038, + "grad_norm": 2.4071738719940186, + "learning_rate": 0.0001756036737457912, + "loss": 0.2317, + "step": 9697 + }, + { + "epoch": 0.22733639245884538, + "grad_norm": 0.7440665364265442, + "learning_rate": 0.00017559885217541317, + "loss": 0.0562, + "step": 9698 + }, + { + "epoch": 0.22735983403365037, + "grad_norm": 3.1529998779296875, + "learning_rate": 0.0001755940301948336, + "loss": 0.3546, + "step": 9699 + }, + { + "epoch": 0.22738327560845537, + "grad_norm": 2.6935715675354004, + "learning_rate": 0.00017558920780407877, + "loss": 0.2299, + "step": 9700 + }, + { + "epoch": 0.22740671718326036, + "grad_norm": 6.7733845710754395, + "learning_rate": 0.00017558438500317475, + "loss": 0.8499, + "step": 9701 + }, + { + "epoch": 0.22743015875806535, + "grad_norm": 3.3495631217956543, + "learning_rate": 0.00017557956179214772, + "loss": 0.2311, + "step": 9702 + }, + { + "epoch": 0.22745360033287038, + "grad_norm": 3.56537127494812, + "learning_rate": 0.00017557473817102386, + "loss": 0.4835, + "step": 9703 + }, + { + "epoch": 0.22747704190767537, + "grad_norm": 3.8808937072753906, + "learning_rate": 0.00017556991413982936, + "loss": 0.291, + "step": 9704 + }, + { + "epoch": 0.22750048348248036, + "grad_norm": 3.4644229412078857, + "learning_rate": 0.00017556508969859036, + "loss": 0.3485, + "step": 9705 + }, + { + "epoch": 0.22752392505728536, + "grad_norm": 2.574524164199829, + "learning_rate": 0.00017556026484733306, + "loss": 0.2869, + "step": 9706 + }, + { + "epoch": 0.22754736663209035, + "grad_norm": 2.3036890029907227, + "learning_rate": 0.00017555543958608363, + "loss": 1.1984, + "step": 9707 + }, + { + "epoch": 0.22757080820689535, + "grad_norm": 1.1550978422164917, + "learning_rate": 0.00017555061391486828, + "loss": 0.106, + "step": 9708 + }, + { + "epoch": 0.22759424978170034, + "grad_norm": 0.6678931713104248, + "learning_rate": 0.00017554578783371318, + "loss": 0.0703, + "step": 9709 + }, + { + "epoch": 0.22761769135650534, + "grad_norm": 3.7904911041259766, + "learning_rate": 0.0001755409613426445, + "loss": 0.2436, + "step": 9710 + }, + { + "epoch": 0.22764113293131033, + "grad_norm": 3.086648464202881, + "learning_rate": 0.00017553613444168845, + "loss": 0.2761, + "step": 9711 + }, + { + "epoch": 0.22766457450611532, + "grad_norm": 3.2360634803771973, + "learning_rate": 0.0001755313071308712, + "loss": 0.3605, + "step": 9712 + }, + { + "epoch": 0.22768801608092032, + "grad_norm": 2.7673096656799316, + "learning_rate": 0.00017552647941021897, + "loss": 0.2663, + "step": 9713 + }, + { + "epoch": 0.2277114576557253, + "grad_norm": 3.4510974884033203, + "learning_rate": 0.0001755216512797579, + "loss": 1.2172, + "step": 9714 + }, + { + "epoch": 0.2277348992305303, + "grad_norm": 3.117629051208496, + "learning_rate": 0.00017551682273951426, + "loss": 0.3221, + "step": 9715 + }, + { + "epoch": 0.2277583408053353, + "grad_norm": 2.798318386077881, + "learning_rate": 0.00017551199378951422, + "loss": 1.2285, + "step": 9716 + }, + { + "epoch": 0.2277817823801403, + "grad_norm": 6.155160427093506, + "learning_rate": 0.00017550716442978396, + "loss": 0.5283, + "step": 9717 + }, + { + "epoch": 0.2278052239549453, + "grad_norm": 1.5045950412750244, + "learning_rate": 0.0001755023346603497, + "loss": 0.1833, + "step": 9718 + }, + { + "epoch": 0.22782866552975028, + "grad_norm": 5.799234867095947, + "learning_rate": 0.0001754975044812377, + "loss": 0.6502, + "step": 9719 + }, + { + "epoch": 0.22785210710455528, + "grad_norm": 5.588581562042236, + "learning_rate": 0.00017549267389247404, + "loss": 0.6277, + "step": 9720 + }, + { + "epoch": 0.22787554867936027, + "grad_norm": 0.8273077011108398, + "learning_rate": 0.00017548784289408506, + "loss": 0.1038, + "step": 9721 + }, + { + "epoch": 0.22789899025416527, + "grad_norm": 4.023476600646973, + "learning_rate": 0.00017548301148609692, + "loss": 0.4418, + "step": 9722 + }, + { + "epoch": 0.22792243182897026, + "grad_norm": 3.592327117919922, + "learning_rate": 0.00017547817966853582, + "loss": 0.4191, + "step": 9723 + }, + { + "epoch": 0.22794587340377526, + "grad_norm": 6.851102352142334, + "learning_rate": 0.00017547334744142797, + "loss": 0.8967, + "step": 9724 + }, + { + "epoch": 0.22796931497858025, + "grad_norm": 1.5012609958648682, + "learning_rate": 0.00017546851480479967, + "loss": 0.2004, + "step": 9725 + }, + { + "epoch": 0.22799275655338525, + "grad_norm": 4.064640522003174, + "learning_rate": 0.00017546368175867704, + "loss": 0.6532, + "step": 9726 + }, + { + "epoch": 0.22801619812819024, + "grad_norm": 3.572133779525757, + "learning_rate": 0.00017545884830308639, + "loss": 0.6227, + "step": 9727 + }, + { + "epoch": 0.22803963970299523, + "grad_norm": 2.849034309387207, + "learning_rate": 0.0001754540144380539, + "loss": 0.222, + "step": 9728 + }, + { + "epoch": 0.22806308127780026, + "grad_norm": 3.151911973953247, + "learning_rate": 0.0001754491801636058, + "loss": 0.2659, + "step": 9729 + }, + { + "epoch": 0.22808652285260525, + "grad_norm": 3.1440060138702393, + "learning_rate": 0.00017544434547976832, + "loss": 0.3338, + "step": 9730 + }, + { + "epoch": 0.22810996442741024, + "grad_norm": 1.3826864957809448, + "learning_rate": 0.0001754395103865677, + "loss": 0.1862, + "step": 9731 + }, + { + "epoch": 0.22813340600221524, + "grad_norm": 5.152402400970459, + "learning_rate": 0.00017543467488403017, + "loss": 0.6575, + "step": 9732 + }, + { + "epoch": 0.22815684757702023, + "grad_norm": 4.899844169616699, + "learning_rate": 0.00017542983897218196, + "loss": 0.6423, + "step": 9733 + }, + { + "epoch": 0.22818028915182523, + "grad_norm": 4.883747577667236, + "learning_rate": 0.00017542500265104938, + "loss": 0.7501, + "step": 9734 + }, + { + "epoch": 0.22820373072663022, + "grad_norm": 1.6233134269714355, + "learning_rate": 0.00017542016592065858, + "loss": 0.2593, + "step": 9735 + }, + { + "epoch": 0.22822717230143522, + "grad_norm": 0.8445978760719299, + "learning_rate": 0.00017541532878103584, + "loss": 0.1081, + "step": 9736 + }, + { + "epoch": 0.2282506138762402, + "grad_norm": 3.443966865539551, + "learning_rate": 0.0001754104912322074, + "loss": 0.3505, + "step": 9737 + }, + { + "epoch": 0.2282740554510452, + "grad_norm": 4.01920747756958, + "learning_rate": 0.0001754056532741995, + "loss": 0.331, + "step": 9738 + }, + { + "epoch": 0.2282974970258502, + "grad_norm": 6.090052127838135, + "learning_rate": 0.00017540081490703842, + "loss": 1.0028, + "step": 9739 + }, + { + "epoch": 0.2283209386006552, + "grad_norm": 2.2054622173309326, + "learning_rate": 0.00017539597613075043, + "loss": 0.3676, + "step": 9740 + }, + { + "epoch": 0.2283443801754602, + "grad_norm": 4.396396636962891, + "learning_rate": 0.0001753911369453617, + "loss": 0.5762, + "step": 9741 + }, + { + "epoch": 0.22836782175026518, + "grad_norm": 3.2062623500823975, + "learning_rate": 0.00017538629735089857, + "loss": 0.3751, + "step": 9742 + }, + { + "epoch": 0.22839126332507018, + "grad_norm": 0.9244531393051147, + "learning_rate": 0.00017538145734738726, + "loss": 0.1369, + "step": 9743 + }, + { + "epoch": 0.22841470489987517, + "grad_norm": 4.372313022613525, + "learning_rate": 0.00017537661693485407, + "loss": 0.6888, + "step": 9744 + }, + { + "epoch": 0.22843814647468016, + "grad_norm": 1.7155784368515015, + "learning_rate": 0.0001753717761133252, + "loss": 0.2724, + "step": 9745 + }, + { + "epoch": 0.22846158804948516, + "grad_norm": 5.00444221496582, + "learning_rate": 0.00017536693488282698, + "loss": 0.557, + "step": 9746 + }, + { + "epoch": 0.22848502962429015, + "grad_norm": 2.140669345855713, + "learning_rate": 0.00017536209324338565, + "loss": 1.4133, + "step": 9747 + }, + { + "epoch": 0.22850847119909515, + "grad_norm": 3.298447370529175, + "learning_rate": 0.0001753572511950275, + "loss": 0.4306, + "step": 9748 + }, + { + "epoch": 0.22853191277390014, + "grad_norm": 4.118767738342285, + "learning_rate": 0.00017535240873777874, + "loss": 0.4022, + "step": 9749 + }, + { + "epoch": 0.22855535434870514, + "grad_norm": 2.637734889984131, + "learning_rate": 0.0001753475658716657, + "loss": 0.2205, + "step": 9750 + }, + { + "epoch": 0.22857879592351013, + "grad_norm": 3.3144354820251465, + "learning_rate": 0.00017534272259671466, + "loss": 0.3547, + "step": 9751 + }, + { + "epoch": 0.22860223749831513, + "grad_norm": 6.092996120452881, + "learning_rate": 0.0001753378789129519, + "loss": 0.5312, + "step": 9752 + }, + { + "epoch": 0.22862567907312012, + "grad_norm": 1.1164021492004395, + "learning_rate": 0.00017533303482040368, + "loss": 0.1584, + "step": 9753 + }, + { + "epoch": 0.22864912064792514, + "grad_norm": 2.5904202461242676, + "learning_rate": 0.0001753281903190963, + "loss": 0.2508, + "step": 9754 + }, + { + "epoch": 0.22867256222273014, + "grad_norm": 4.850028991699219, + "learning_rate": 0.00017532334540905603, + "loss": 0.6466, + "step": 9755 + }, + { + "epoch": 0.22869600379753513, + "grad_norm": 2.401052713394165, + "learning_rate": 0.0001753185000903092, + "loss": 0.276, + "step": 9756 + }, + { + "epoch": 0.22871944537234012, + "grad_norm": 3.106074810028076, + "learning_rate": 0.00017531365436288204, + "loss": 1.4354, + "step": 9757 + }, + { + "epoch": 0.22874288694714512, + "grad_norm": 3.4730417728424072, + "learning_rate": 0.00017530880822680088, + "loss": 0.3716, + "step": 9758 + }, + { + "epoch": 0.2287663285219501, + "grad_norm": 8.706677436828613, + "learning_rate": 0.000175303961682092, + "loss": 0.8708, + "step": 9759 + }, + { + "epoch": 0.2287897700967551, + "grad_norm": 3.2396743297576904, + "learning_rate": 0.00017529911472878172, + "loss": 0.2435, + "step": 9760 + }, + { + "epoch": 0.2288132116715601, + "grad_norm": 1.213956356048584, + "learning_rate": 0.00017529426736689633, + "loss": 0.105, + "step": 9761 + }, + { + "epoch": 0.2288366532463651, + "grad_norm": 4.350886821746826, + "learning_rate": 0.00017528941959646212, + "loss": 0.5122, + "step": 9762 + }, + { + "epoch": 0.2288600948211701, + "grad_norm": 3.4597883224487305, + "learning_rate": 0.0001752845714175054, + "loss": 0.3158, + "step": 9763 + }, + { + "epoch": 0.22888353639597508, + "grad_norm": 3.5818095207214355, + "learning_rate": 0.00017527972283005249, + "loss": 0.2843, + "step": 9764 + }, + { + "epoch": 0.22890697797078008, + "grad_norm": 0.8435932993888855, + "learning_rate": 0.00017527487383412968, + "loss": 0.0706, + "step": 9765 + }, + { + "epoch": 0.22893041954558507, + "grad_norm": 4.06694221496582, + "learning_rate": 0.00017527002442976327, + "loss": 0.619, + "step": 9766 + }, + { + "epoch": 0.22895386112039007, + "grad_norm": 6.322482109069824, + "learning_rate": 0.00017526517461697965, + "loss": 0.6042, + "step": 9767 + }, + { + "epoch": 0.22897730269519506, + "grad_norm": 3.027390480041504, + "learning_rate": 0.00017526032439580503, + "loss": 0.4568, + "step": 9768 + }, + { + "epoch": 0.22900074427000006, + "grad_norm": 2.7180533409118652, + "learning_rate": 0.00017525547376626578, + "loss": 1.321, + "step": 9769 + }, + { + "epoch": 0.22902418584480505, + "grad_norm": 2.648104190826416, + "learning_rate": 0.0001752506227283882, + "loss": 0.2609, + "step": 9770 + }, + { + "epoch": 0.22904762741961004, + "grad_norm": 4.96546745300293, + "learning_rate": 0.00017524577128219865, + "loss": 0.2982, + "step": 9771 + }, + { + "epoch": 0.22907106899441504, + "grad_norm": 3.509068012237549, + "learning_rate": 0.0001752409194277234, + "loss": 0.4144, + "step": 9772 + }, + { + "epoch": 0.22909451056922003, + "grad_norm": 5.355566024780273, + "learning_rate": 0.00017523606716498885, + "loss": 0.8856, + "step": 9773 + }, + { + "epoch": 0.22911795214402503, + "grad_norm": 3.2105600833892822, + "learning_rate": 0.00017523121449402123, + "loss": 0.1897, + "step": 9774 + }, + { + "epoch": 0.22914139371883002, + "grad_norm": 2.950623035430908, + "learning_rate": 0.00017522636141484695, + "loss": 1.6572, + "step": 9775 + }, + { + "epoch": 0.22916483529363502, + "grad_norm": 3.626753807067871, + "learning_rate": 0.0001752215079274923, + "loss": 0.3102, + "step": 9776 + }, + { + "epoch": 0.22918827686844, + "grad_norm": 2.253960132598877, + "learning_rate": 0.0001752166540319837, + "loss": 0.2708, + "step": 9777 + }, + { + "epoch": 0.229211718443245, + "grad_norm": 3.442406177520752, + "learning_rate": 0.00017521179972834736, + "loss": 0.436, + "step": 9778 + }, + { + "epoch": 0.22923516001805, + "grad_norm": 1.6721389293670654, + "learning_rate": 0.00017520694501660971, + "loss": 0.2045, + "step": 9779 + }, + { + "epoch": 0.22925860159285502, + "grad_norm": 5.7049970626831055, + "learning_rate": 0.00017520208989679702, + "loss": 0.4038, + "step": 9780 + }, + { + "epoch": 0.22928204316766002, + "grad_norm": 4.53150749206543, + "learning_rate": 0.00017519723436893573, + "loss": 0.3081, + "step": 9781 + }, + { + "epoch": 0.229305484742465, + "grad_norm": 3.297991991043091, + "learning_rate": 0.00017519237843305208, + "loss": 0.345, + "step": 9782 + }, + { + "epoch": 0.22932892631727, + "grad_norm": 3.4203267097473145, + "learning_rate": 0.0001751875220891725, + "loss": 1.8801, + "step": 9783 + }, + { + "epoch": 0.229352367892075, + "grad_norm": 3.749314546585083, + "learning_rate": 0.0001751826653373233, + "loss": 0.3518, + "step": 9784 + }, + { + "epoch": 0.22937580946688, + "grad_norm": 4.200500965118408, + "learning_rate": 0.00017517780817753083, + "loss": 0.3625, + "step": 9785 + }, + { + "epoch": 0.229399251041685, + "grad_norm": 4.1396484375, + "learning_rate": 0.00017517295060982151, + "loss": 0.4444, + "step": 9786 + }, + { + "epoch": 0.22942269261648998, + "grad_norm": 9.673492431640625, + "learning_rate": 0.0001751680926342216, + "loss": 1.2791, + "step": 9787 + }, + { + "epoch": 0.22944613419129498, + "grad_norm": 1.163757562637329, + "learning_rate": 0.0001751632342507575, + "loss": 0.1709, + "step": 9788 + }, + { + "epoch": 0.22946957576609997, + "grad_norm": 3.3965089321136475, + "learning_rate": 0.00017515837545945565, + "loss": 0.4819, + "step": 9789 + }, + { + "epoch": 0.22949301734090496, + "grad_norm": 0.6195621490478516, + "learning_rate": 0.00017515351626034225, + "loss": 0.0799, + "step": 9790 + }, + { + "epoch": 0.22951645891570996, + "grad_norm": 3.274946689605713, + "learning_rate": 0.0001751486566534438, + "loss": 0.3837, + "step": 9791 + }, + { + "epoch": 0.22953990049051495, + "grad_norm": 3.24137806892395, + "learning_rate": 0.00017514379663878665, + "loss": 0.3426, + "step": 9792 + }, + { + "epoch": 0.22956334206531995, + "grad_norm": 4.842894077301025, + "learning_rate": 0.0001751389362163971, + "loss": 0.6492, + "step": 9793 + }, + { + "epoch": 0.22958678364012494, + "grad_norm": 0.6064428091049194, + "learning_rate": 0.0001751340753863016, + "loss": 0.0449, + "step": 9794 + }, + { + "epoch": 0.22961022521492994, + "grad_norm": 4.155772686004639, + "learning_rate": 0.00017512921414852646, + "loss": 0.6105, + "step": 9795 + }, + { + "epoch": 0.22963366678973493, + "grad_norm": 1.0392087697982788, + "learning_rate": 0.0001751243525030981, + "loss": 0.1548, + "step": 9796 + }, + { + "epoch": 0.22965710836453992, + "grad_norm": 2.9546923637390137, + "learning_rate": 0.0001751194904500429, + "loss": 0.3127, + "step": 9797 + }, + { + "epoch": 0.22968054993934492, + "grad_norm": 2.0943117141723633, + "learning_rate": 0.00017511462798938725, + "loss": 1.1881, + "step": 9798 + }, + { + "epoch": 0.2297039915141499, + "grad_norm": 2.319295883178711, + "learning_rate": 0.00017510976512115752, + "loss": 0.365, + "step": 9799 + }, + { + "epoch": 0.2297274330889549, + "grad_norm": 3.6148271560668945, + "learning_rate": 0.00017510490184538007, + "loss": 0.505, + "step": 9800 + }, + { + "epoch": 0.2297508746637599, + "grad_norm": 3.5614826679229736, + "learning_rate": 0.00017510003816208132, + "loss": 0.4388, + "step": 9801 + }, + { + "epoch": 0.2297743162385649, + "grad_norm": 2.486636161804199, + "learning_rate": 0.00017509517407128766, + "loss": 0.3294, + "step": 9802 + }, + { + "epoch": 0.2297977578133699, + "grad_norm": 2.7553346157073975, + "learning_rate": 0.00017509030957302545, + "loss": 0.2496, + "step": 9803 + }, + { + "epoch": 0.22982119938817488, + "grad_norm": 3.970930814743042, + "learning_rate": 0.00017508544466732112, + "loss": 0.3764, + "step": 9804 + }, + { + "epoch": 0.22984464096297988, + "grad_norm": 1.226364254951477, + "learning_rate": 0.00017508057935420104, + "loss": 0.1779, + "step": 9805 + }, + { + "epoch": 0.2298680825377849, + "grad_norm": 2.5968194007873535, + "learning_rate": 0.00017507571363369165, + "loss": 1.5146, + "step": 9806 + }, + { + "epoch": 0.2298915241125899, + "grad_norm": 6.522576808929443, + "learning_rate": 0.00017507084750581934, + "loss": 0.6913, + "step": 9807 + }, + { + "epoch": 0.2299149656873949, + "grad_norm": 2.4729602336883545, + "learning_rate": 0.00017506598097061045, + "loss": 1.4566, + "step": 9808 + }, + { + "epoch": 0.22993840726219988, + "grad_norm": 5.754672527313232, + "learning_rate": 0.00017506111402809148, + "loss": 0.7127, + "step": 9809 + }, + { + "epoch": 0.22996184883700488, + "grad_norm": 1.6198409795761108, + "learning_rate": 0.00017505624667828878, + "loss": 0.0887, + "step": 9810 + }, + { + "epoch": 0.22998529041180987, + "grad_norm": 2.1126976013183594, + "learning_rate": 0.00017505137892122877, + "loss": 0.2988, + "step": 9811 + }, + { + "epoch": 0.23000873198661487, + "grad_norm": 2.9219017028808594, + "learning_rate": 0.0001750465107569379, + "loss": 0.3637, + "step": 9812 + }, + { + "epoch": 0.23003217356141986, + "grad_norm": 3.647840738296509, + "learning_rate": 0.0001750416421854425, + "loss": 0.633, + "step": 9813 + }, + { + "epoch": 0.23005561513622486, + "grad_norm": 3.1692020893096924, + "learning_rate": 0.0001750367732067691, + "loss": 1.9103, + "step": 9814 + }, + { + "epoch": 0.23007905671102985, + "grad_norm": 2.8460707664489746, + "learning_rate": 0.000175031903820944, + "loss": 0.5632, + "step": 9815 + }, + { + "epoch": 0.23010249828583484, + "grad_norm": 0.7255671620368958, + "learning_rate": 0.0001750270340279937, + "loss": 0.0816, + "step": 9816 + }, + { + "epoch": 0.23012593986063984, + "grad_norm": 3.1212339401245117, + "learning_rate": 0.0001750221638279446, + "loss": 0.5102, + "step": 9817 + }, + { + "epoch": 0.23014938143544483, + "grad_norm": 3.7327516078948975, + "learning_rate": 0.00017501729322082316, + "loss": 0.4091, + "step": 9818 + }, + { + "epoch": 0.23017282301024983, + "grad_norm": 5.083209037780762, + "learning_rate": 0.00017501242220665576, + "loss": 0.5985, + "step": 9819 + }, + { + "epoch": 0.23019626458505482, + "grad_norm": 1.1727885007858276, + "learning_rate": 0.00017500755078546886, + "loss": 0.1054, + "step": 9820 + }, + { + "epoch": 0.23021970615985982, + "grad_norm": 3.893388032913208, + "learning_rate": 0.00017500267895728886, + "loss": 0.5816, + "step": 9821 + }, + { + "epoch": 0.2302431477346648, + "grad_norm": 3.0457541942596436, + "learning_rate": 0.00017499780672214222, + "loss": 0.3003, + "step": 9822 + }, + { + "epoch": 0.2302665893094698, + "grad_norm": 2.8596901893615723, + "learning_rate": 0.00017499293408005534, + "loss": 0.3977, + "step": 9823 + }, + { + "epoch": 0.2302900308842748, + "grad_norm": 2.588630437850952, + "learning_rate": 0.00017498806103105477, + "loss": 0.3055, + "step": 9824 + }, + { + "epoch": 0.2303134724590798, + "grad_norm": 4.38480281829834, + "learning_rate": 0.0001749831875751668, + "loss": 0.6207, + "step": 9825 + }, + { + "epoch": 0.2303369140338848, + "grad_norm": 3.687453508377075, + "learning_rate": 0.00017497831371241797, + "loss": 1.7609, + "step": 9826 + }, + { + "epoch": 0.23036035560868978, + "grad_norm": 2.711721420288086, + "learning_rate": 0.00017497343944283472, + "loss": 0.2549, + "step": 9827 + }, + { + "epoch": 0.23038379718349478, + "grad_norm": 2.5142290592193604, + "learning_rate": 0.00017496856476644345, + "loss": 0.2536, + "step": 9828 + }, + { + "epoch": 0.23040723875829977, + "grad_norm": 0.890921950340271, + "learning_rate": 0.00017496368968327066, + "loss": 0.092, + "step": 9829 + }, + { + "epoch": 0.23043068033310476, + "grad_norm": 3.8325679302215576, + "learning_rate": 0.00017495881419334278, + "loss": 0.53, + "step": 9830 + }, + { + "epoch": 0.2304541219079098, + "grad_norm": 3.52247953414917, + "learning_rate": 0.00017495393829668624, + "loss": 0.3089, + "step": 9831 + }, + { + "epoch": 0.23047756348271478, + "grad_norm": 3.3508288860321045, + "learning_rate": 0.00017494906199332755, + "loss": 1.6089, + "step": 9832 + }, + { + "epoch": 0.23050100505751978, + "grad_norm": 2.297902822494507, + "learning_rate": 0.00017494418528329314, + "loss": 0.2479, + "step": 9833 + }, + { + "epoch": 0.23052444663232477, + "grad_norm": 3.3567564487457275, + "learning_rate": 0.00017493930816660946, + "loss": 1.6459, + "step": 9834 + }, + { + "epoch": 0.23054788820712976, + "grad_norm": 2.793034553527832, + "learning_rate": 0.000174934430643303, + "loss": 0.3223, + "step": 9835 + }, + { + "epoch": 0.23057132978193476, + "grad_norm": 3.4039533138275146, + "learning_rate": 0.0001749295527134002, + "loss": 0.4792, + "step": 9836 + }, + { + "epoch": 0.23059477135673975, + "grad_norm": 2.68912410736084, + "learning_rate": 0.00017492467437692756, + "loss": 0.2884, + "step": 9837 + }, + { + "epoch": 0.23061821293154475, + "grad_norm": 2.9851272106170654, + "learning_rate": 0.0001749197956339115, + "loss": 0.2782, + "step": 9838 + }, + { + "epoch": 0.23064165450634974, + "grad_norm": 7.536771297454834, + "learning_rate": 0.00017491491648437854, + "loss": 0.759, + "step": 9839 + }, + { + "epoch": 0.23066509608115474, + "grad_norm": 2.6133925914764404, + "learning_rate": 0.00017491003692835517, + "loss": 0.2825, + "step": 9840 + }, + { + "epoch": 0.23068853765595973, + "grad_norm": 3.2114410400390625, + "learning_rate": 0.0001749051569658678, + "loss": 0.337, + "step": 9841 + }, + { + "epoch": 0.23071197923076472, + "grad_norm": 3.087833881378174, + "learning_rate": 0.00017490027659694294, + "loss": 1.8102, + "step": 9842 + }, + { + "epoch": 0.23073542080556972, + "grad_norm": 4.800046920776367, + "learning_rate": 0.00017489539582160707, + "loss": 0.485, + "step": 9843 + }, + { + "epoch": 0.2307588623803747, + "grad_norm": 4.462162494659424, + "learning_rate": 0.00017489051463988667, + "loss": 0.447, + "step": 9844 + }, + { + "epoch": 0.2307823039551797, + "grad_norm": 1.7760027647018433, + "learning_rate": 0.00017488563305180827, + "loss": 0.1928, + "step": 9845 + }, + { + "epoch": 0.2308057455299847, + "grad_norm": 1.8371264934539795, + "learning_rate": 0.0001748807510573983, + "loss": 0.2243, + "step": 9846 + }, + { + "epoch": 0.2308291871047897, + "grad_norm": 2.5816211700439453, + "learning_rate": 0.00017487586865668324, + "loss": 2.0727, + "step": 9847 + }, + { + "epoch": 0.2308526286795947, + "grad_norm": 4.839433670043945, + "learning_rate": 0.00017487098584968963, + "loss": 0.502, + "step": 9848 + }, + { + "epoch": 0.23087607025439968, + "grad_norm": 1.501596450805664, + "learning_rate": 0.00017486610263644397, + "loss": 0.1676, + "step": 9849 + }, + { + "epoch": 0.23089951182920468, + "grad_norm": 1.6642374992370605, + "learning_rate": 0.00017486121901697272, + "loss": 0.2278, + "step": 9850 + }, + { + "epoch": 0.23092295340400967, + "grad_norm": 3.721971273422241, + "learning_rate": 0.0001748563349913024, + "loss": 0.4544, + "step": 9851 + }, + { + "epoch": 0.23094639497881467, + "grad_norm": 1.516884446144104, + "learning_rate": 0.00017485145055945944, + "loss": 0.1602, + "step": 9852 + }, + { + "epoch": 0.23096983655361966, + "grad_norm": 2.8697004318237305, + "learning_rate": 0.00017484656572147047, + "loss": 0.3375, + "step": 9853 + }, + { + "epoch": 0.23099327812842466, + "grad_norm": 5.654581069946289, + "learning_rate": 0.00017484168047736192, + "loss": 0.7859, + "step": 9854 + }, + { + "epoch": 0.23101671970322965, + "grad_norm": 3.1152279376983643, + "learning_rate": 0.00017483679482716027, + "loss": 0.3003, + "step": 9855 + }, + { + "epoch": 0.23104016127803464, + "grad_norm": 0.6574581265449524, + "learning_rate": 0.0001748319087708921, + "loss": 0.0652, + "step": 9856 + }, + { + "epoch": 0.23106360285283967, + "grad_norm": 3.7165403366088867, + "learning_rate": 0.0001748270223085839, + "loss": 0.4495, + "step": 9857 + }, + { + "epoch": 0.23108704442764466, + "grad_norm": 1.2489670515060425, + "learning_rate": 0.00017482213544026216, + "loss": 0.1068, + "step": 9858 + }, + { + "epoch": 0.23111048600244966, + "grad_norm": 3.023651599884033, + "learning_rate": 0.0001748172481659534, + "loss": 0.2101, + "step": 9859 + }, + { + "epoch": 0.23113392757725465, + "grad_norm": 2.983691692352295, + "learning_rate": 0.00017481236048568413, + "loss": 0.3216, + "step": 9860 + }, + { + "epoch": 0.23115736915205964, + "grad_norm": 4.5025129318237305, + "learning_rate": 0.0001748074723994809, + "loss": 0.6332, + "step": 9861 + }, + { + "epoch": 0.23118081072686464, + "grad_norm": 2.7137792110443115, + "learning_rate": 0.0001748025839073703, + "loss": 0.2926, + "step": 9862 + }, + { + "epoch": 0.23120425230166963, + "grad_norm": 0.2597787380218506, + "learning_rate": 0.00017479769500937866, + "loss": 0.0125, + "step": 9863 + }, + { + "epoch": 0.23122769387647463, + "grad_norm": 1.265475869178772, + "learning_rate": 0.00017479280570553268, + "loss": 0.1828, + "step": 9864 + }, + { + "epoch": 0.23125113545127962, + "grad_norm": 10.837438583374023, + "learning_rate": 0.00017478791599585884, + "loss": 0.7568, + "step": 9865 + }, + { + "epoch": 0.23127457702608462, + "grad_norm": 0.4027213156223297, + "learning_rate": 0.00017478302588038363, + "loss": 0.045, + "step": 9866 + }, + { + "epoch": 0.2312980186008896, + "grad_norm": 4.010488510131836, + "learning_rate": 0.00017477813535913365, + "loss": 0.3029, + "step": 9867 + }, + { + "epoch": 0.2313214601756946, + "grad_norm": 3.6698532104492188, + "learning_rate": 0.0001747732444321354, + "loss": 0.4882, + "step": 9868 + }, + { + "epoch": 0.2313449017504996, + "grad_norm": 3.675091505050659, + "learning_rate": 0.0001747683530994154, + "loss": 0.3193, + "step": 9869 + }, + { + "epoch": 0.2313683433253046, + "grad_norm": 0.6696430444717407, + "learning_rate": 0.00017476346136100025, + "loss": 0.088, + "step": 9870 + }, + { + "epoch": 0.2313917849001096, + "grad_norm": 4.528181552886963, + "learning_rate": 0.00017475856921691642, + "loss": 0.51, + "step": 9871 + }, + { + "epoch": 0.23141522647491458, + "grad_norm": 3.2598440647125244, + "learning_rate": 0.00017475367666719053, + "loss": 0.3662, + "step": 9872 + }, + { + "epoch": 0.23143866804971958, + "grad_norm": 1.624556064605713, + "learning_rate": 0.00017474878371184907, + "loss": 0.2214, + "step": 9873 + }, + { + "epoch": 0.23146210962452457, + "grad_norm": 4.058978080749512, + "learning_rate": 0.0001747438903509186, + "loss": 0.4357, + "step": 9874 + }, + { + "epoch": 0.23148555119932956, + "grad_norm": 4.247859954833984, + "learning_rate": 0.00017473899658442568, + "loss": 0.4292, + "step": 9875 + }, + { + "epoch": 0.23150899277413456, + "grad_norm": 2.7516255378723145, + "learning_rate": 0.00017473410241239688, + "loss": 1.7913, + "step": 9876 + }, + { + "epoch": 0.23153243434893955, + "grad_norm": 4.581638336181641, + "learning_rate": 0.0001747292078348587, + "loss": 0.4303, + "step": 9877 + }, + { + "epoch": 0.23155587592374455, + "grad_norm": 4.162230968475342, + "learning_rate": 0.0001747243128518378, + "loss": 0.4416, + "step": 9878 + }, + { + "epoch": 0.23157931749854954, + "grad_norm": 3.2866523265838623, + "learning_rate": 0.0001747194174633606, + "loss": 0.5632, + "step": 9879 + }, + { + "epoch": 0.23160275907335454, + "grad_norm": 0.46876558661460876, + "learning_rate": 0.0001747145216694538, + "loss": 0.0308, + "step": 9880 + }, + { + "epoch": 0.23162620064815953, + "grad_norm": 4.8024773597717285, + "learning_rate": 0.00017470962547014388, + "loss": 0.7323, + "step": 9881 + }, + { + "epoch": 0.23164964222296455, + "grad_norm": 1.1264134645462036, + "learning_rate": 0.00017470472886545744, + "loss": 0.0877, + "step": 9882 + }, + { + "epoch": 0.23167308379776955, + "grad_norm": 3.18318247795105, + "learning_rate": 0.00017469983185542108, + "loss": 0.4927, + "step": 9883 + }, + { + "epoch": 0.23169652537257454, + "grad_norm": 4.017026901245117, + "learning_rate": 0.00017469493444006125, + "loss": 0.4886, + "step": 9884 + }, + { + "epoch": 0.23171996694737954, + "grad_norm": 1.5337992906570435, + "learning_rate": 0.00017469003661940463, + "loss": 0.1492, + "step": 9885 + }, + { + "epoch": 0.23174340852218453, + "grad_norm": 3.755640745162964, + "learning_rate": 0.00017468513839347783, + "loss": 0.4907, + "step": 9886 + }, + { + "epoch": 0.23176685009698952, + "grad_norm": 4.334150791168213, + "learning_rate": 0.00017468023976230729, + "loss": 0.4287, + "step": 9887 + }, + { + "epoch": 0.23179029167179452, + "grad_norm": 3.4152660369873047, + "learning_rate": 0.0001746753407259197, + "loss": 0.3797, + "step": 9888 + }, + { + "epoch": 0.2318137332465995, + "grad_norm": 1.0149093866348267, + "learning_rate": 0.00017467044128434165, + "loss": 0.1409, + "step": 9889 + }, + { + "epoch": 0.2318371748214045, + "grad_norm": 1.714176893234253, + "learning_rate": 0.00017466554143759963, + "loss": 0.2642, + "step": 9890 + }, + { + "epoch": 0.2318606163962095, + "grad_norm": 5.2776031494140625, + "learning_rate": 0.00017466064118572033, + "loss": 0.7522, + "step": 9891 + }, + { + "epoch": 0.2318840579710145, + "grad_norm": 4.065779685974121, + "learning_rate": 0.00017465574052873026, + "loss": 0.5624, + "step": 9892 + }, + { + "epoch": 0.2319074995458195, + "grad_norm": 2.7304749488830566, + "learning_rate": 0.00017465083946665604, + "loss": 0.2286, + "step": 9893 + }, + { + "epoch": 0.23193094112062448, + "grad_norm": 6.142971992492676, + "learning_rate": 0.0001746459379995243, + "loss": 0.661, + "step": 9894 + }, + { + "epoch": 0.23195438269542948, + "grad_norm": 5.24378776550293, + "learning_rate": 0.00017464103612736156, + "loss": 0.4695, + "step": 9895 + }, + { + "epoch": 0.23197782427023447, + "grad_norm": 0.64528489112854, + "learning_rate": 0.00017463613385019448, + "loss": 0.0854, + "step": 9896 + }, + { + "epoch": 0.23200126584503947, + "grad_norm": 4.659130573272705, + "learning_rate": 0.00017463123116804964, + "loss": 0.6552, + "step": 9897 + }, + { + "epoch": 0.23202470741984446, + "grad_norm": 2.8870351314544678, + "learning_rate": 0.00017462632808095364, + "loss": 0.3472, + "step": 9898 + }, + { + "epoch": 0.23204814899464946, + "grad_norm": 1.8775709867477417, + "learning_rate": 0.0001746214245889331, + "loss": 0.1684, + "step": 9899 + }, + { + "epoch": 0.23207159056945445, + "grad_norm": 4.034999847412109, + "learning_rate": 0.0001746165206920146, + "loss": 0.4654, + "step": 9900 + }, + { + "epoch": 0.23209503214425944, + "grad_norm": 5.249146938323975, + "learning_rate": 0.00017461161639022475, + "loss": 0.6907, + "step": 9901 + }, + { + "epoch": 0.23211847371906444, + "grad_norm": 2.673628807067871, + "learning_rate": 0.00017460671168359015, + "loss": 0.3453, + "step": 9902 + }, + { + "epoch": 0.23214191529386943, + "grad_norm": 1.0986446142196655, + "learning_rate": 0.00017460180657213748, + "loss": 0.0406, + "step": 9903 + }, + { + "epoch": 0.23216535686867443, + "grad_norm": 4.677333354949951, + "learning_rate": 0.0001745969010558933, + "loss": 0.5074, + "step": 9904 + }, + { + "epoch": 0.23218879844347942, + "grad_norm": 4.219696998596191, + "learning_rate": 0.00017459199513488423, + "loss": 0.5251, + "step": 9905 + }, + { + "epoch": 0.23221224001828442, + "grad_norm": 1.6727081537246704, + "learning_rate": 0.00017458708880913688, + "loss": 0.1066, + "step": 9906 + }, + { + "epoch": 0.2322356815930894, + "grad_norm": 0.8603912591934204, + "learning_rate": 0.0001745821820786779, + "loss": 0.0801, + "step": 9907 + }, + { + "epoch": 0.23225912316789443, + "grad_norm": 3.953604221343994, + "learning_rate": 0.00017457727494353393, + "loss": 0.6168, + "step": 9908 + }, + { + "epoch": 0.23228256474269943, + "grad_norm": 3.986922025680542, + "learning_rate": 0.00017457236740373152, + "loss": 0.6719, + "step": 9909 + }, + { + "epoch": 0.23230600631750442, + "grad_norm": 3.15336012840271, + "learning_rate": 0.0001745674594592974, + "loss": 0.3285, + "step": 9910 + }, + { + "epoch": 0.23232944789230942, + "grad_norm": 3.391258955001831, + "learning_rate": 0.0001745625511102581, + "loss": 0.3757, + "step": 9911 + }, + { + "epoch": 0.2323528894671144, + "grad_norm": 4.114316940307617, + "learning_rate": 0.00017455764235664033, + "loss": 0.5068, + "step": 9912 + }, + { + "epoch": 0.2323763310419194, + "grad_norm": 3.5448057651519775, + "learning_rate": 0.0001745527331984707, + "loss": 1.4941, + "step": 9913 + }, + { + "epoch": 0.2323997726167244, + "grad_norm": 3.105358123779297, + "learning_rate": 0.00017454782363577581, + "loss": 0.2502, + "step": 9914 + }, + { + "epoch": 0.2324232141915294, + "grad_norm": 4.309828281402588, + "learning_rate": 0.00017454291366858238, + "loss": 0.4416, + "step": 9915 + }, + { + "epoch": 0.2324466557663344, + "grad_norm": 2.645643949508667, + "learning_rate": 0.00017453800329691698, + "loss": 0.3351, + "step": 9916 + }, + { + "epoch": 0.23247009734113938, + "grad_norm": 3.797116756439209, + "learning_rate": 0.00017453309252080627, + "loss": 0.4981, + "step": 9917 + }, + { + "epoch": 0.23249353891594438, + "grad_norm": 1.3044023513793945, + "learning_rate": 0.00017452818134027694, + "loss": 0.1972, + "step": 9918 + }, + { + "epoch": 0.23251698049074937, + "grad_norm": 2.4297447204589844, + "learning_rate": 0.00017452326975535555, + "loss": 0.1643, + "step": 9919 + }, + { + "epoch": 0.23254042206555436, + "grad_norm": 4.079602241516113, + "learning_rate": 0.00017451835776606883, + "loss": 0.7373, + "step": 9920 + }, + { + "epoch": 0.23256386364035936, + "grad_norm": 4.787021636962891, + "learning_rate": 0.00017451344537244343, + "loss": 0.5198, + "step": 9921 + }, + { + "epoch": 0.23258730521516435, + "grad_norm": 3.072213888168335, + "learning_rate": 0.00017450853257450596, + "loss": 0.3935, + "step": 9922 + }, + { + "epoch": 0.23261074678996935, + "grad_norm": 0.934235155582428, + "learning_rate": 0.0001745036193722831, + "loss": 0.0924, + "step": 9923 + }, + { + "epoch": 0.23263418836477434, + "grad_norm": 3.8833987712860107, + "learning_rate": 0.0001744987057658015, + "loss": 0.3197, + "step": 9924 + }, + { + "epoch": 0.23265762993957934, + "grad_norm": 3.727555513381958, + "learning_rate": 0.00017449379175508783, + "loss": 0.2673, + "step": 9925 + }, + { + "epoch": 0.23268107151438433, + "grad_norm": 2.1721527576446533, + "learning_rate": 0.0001744888773401687, + "loss": 0.2821, + "step": 9926 + }, + { + "epoch": 0.23270451308918932, + "grad_norm": 2.981782913208008, + "learning_rate": 0.00017448396252107092, + "loss": 0.2281, + "step": 9927 + }, + { + "epoch": 0.23272795466399432, + "grad_norm": 3.3351590633392334, + "learning_rate": 0.000174479047297821, + "loss": 0.3648, + "step": 9928 + }, + { + "epoch": 0.2327513962387993, + "grad_norm": 3.518697500228882, + "learning_rate": 0.00017447413167044572, + "loss": 0.4157, + "step": 9929 + }, + { + "epoch": 0.2327748378136043, + "grad_norm": 1.597933292388916, + "learning_rate": 0.00017446921563897164, + "loss": 0.2102, + "step": 9930 + }, + { + "epoch": 0.2327982793884093, + "grad_norm": 2.9897074699401855, + "learning_rate": 0.0001744642992034256, + "loss": 0.2851, + "step": 9931 + }, + { + "epoch": 0.2328217209632143, + "grad_norm": 2.626847743988037, + "learning_rate": 0.0001744593823638341, + "loss": 0.3781, + "step": 9932 + }, + { + "epoch": 0.23284516253801932, + "grad_norm": 0.6824419498443604, + "learning_rate": 0.00017445446512022393, + "loss": 0.0688, + "step": 9933 + }, + { + "epoch": 0.2328686041128243, + "grad_norm": 3.9664061069488525, + "learning_rate": 0.00017444954747262173, + "loss": 0.3901, + "step": 9934 + }, + { + "epoch": 0.2328920456876293, + "grad_norm": 3.2781949043273926, + "learning_rate": 0.00017444462942105418, + "loss": 2.2742, + "step": 9935 + }, + { + "epoch": 0.2329154872624343, + "grad_norm": 4.776200771331787, + "learning_rate": 0.000174439710965548, + "loss": 0.2577, + "step": 9936 + }, + { + "epoch": 0.2329389288372393, + "grad_norm": 6.12936544418335, + "learning_rate": 0.00017443479210612982, + "loss": 0.5577, + "step": 9937 + }, + { + "epoch": 0.2329623704120443, + "grad_norm": 2.4293160438537598, + "learning_rate": 0.00017442987284282642, + "loss": 1.5041, + "step": 9938 + }, + { + "epoch": 0.23298581198684928, + "grad_norm": 1.5471471548080444, + "learning_rate": 0.0001744249531756644, + "loss": 0.181, + "step": 9939 + }, + { + "epoch": 0.23300925356165428, + "grad_norm": 2.90171480178833, + "learning_rate": 0.0001744200331046705, + "loss": 0.2471, + "step": 9940 + }, + { + "epoch": 0.23303269513645927, + "grad_norm": 3.3365981578826904, + "learning_rate": 0.00017441511262987142, + "loss": 1.9066, + "step": 9941 + }, + { + "epoch": 0.23305613671126427, + "grad_norm": 5.074118614196777, + "learning_rate": 0.00017441019175129383, + "loss": 0.6889, + "step": 9942 + }, + { + "epoch": 0.23307957828606926, + "grad_norm": 8.031164169311523, + "learning_rate": 0.00017440527046896445, + "loss": 0.6468, + "step": 9943 + }, + { + "epoch": 0.23310301986087426, + "grad_norm": 2.56691837310791, + "learning_rate": 0.00017440034878290995, + "loss": 0.2713, + "step": 9944 + }, + { + "epoch": 0.23312646143567925, + "grad_norm": 1.5662083625793457, + "learning_rate": 0.0001743954266931571, + "loss": 0.2677, + "step": 9945 + }, + { + "epoch": 0.23314990301048424, + "grad_norm": 4.345058917999268, + "learning_rate": 0.0001743905041997326, + "loss": 0.4839, + "step": 9946 + }, + { + "epoch": 0.23317334458528924, + "grad_norm": 4.233199596405029, + "learning_rate": 0.0001743855813026631, + "loss": 0.4916, + "step": 9947 + }, + { + "epoch": 0.23319678616009423, + "grad_norm": 3.9655344486236572, + "learning_rate": 0.00017438065800197533, + "loss": 0.4502, + "step": 9948 + }, + { + "epoch": 0.23322022773489923, + "grad_norm": 3.550719976425171, + "learning_rate": 0.00017437573429769602, + "loss": 0.3636, + "step": 9949 + }, + { + "epoch": 0.23324366930970422, + "grad_norm": 1.400319218635559, + "learning_rate": 0.00017437081018985192, + "loss": 1.3332, + "step": 9950 + }, + { + "epoch": 0.23326711088450922, + "grad_norm": 4.114656448364258, + "learning_rate": 0.0001743658856784697, + "loss": 0.3915, + "step": 9951 + }, + { + "epoch": 0.2332905524593142, + "grad_norm": 4.736314296722412, + "learning_rate": 0.00017436096076357607, + "loss": 0.5217, + "step": 9952 + }, + { + "epoch": 0.2333139940341192, + "grad_norm": 4.173443794250488, + "learning_rate": 0.0001743560354451978, + "loss": 0.5115, + "step": 9953 + }, + { + "epoch": 0.2333374356089242, + "grad_norm": 3.7457118034362793, + "learning_rate": 0.00017435110972336158, + "loss": 0.4963, + "step": 9954 + }, + { + "epoch": 0.2333608771837292, + "grad_norm": 4.801143169403076, + "learning_rate": 0.00017434618359809414, + "loss": 0.3439, + "step": 9955 + }, + { + "epoch": 0.2333843187585342, + "grad_norm": 2.3637499809265137, + "learning_rate": 0.00017434125706942223, + "loss": 1.1989, + "step": 9956 + }, + { + "epoch": 0.23340776033333918, + "grad_norm": 2.365157127380371, + "learning_rate": 0.00017433633013737255, + "loss": 0.2746, + "step": 9957 + }, + { + "epoch": 0.23343120190814418, + "grad_norm": 7.679777145385742, + "learning_rate": 0.00017433140280197188, + "loss": 0.801, + "step": 9958 + }, + { + "epoch": 0.2334546434829492, + "grad_norm": 5.722248077392578, + "learning_rate": 0.00017432647506324694, + "loss": 0.5674, + "step": 9959 + }, + { + "epoch": 0.2334780850577542, + "grad_norm": 2.5686819553375244, + "learning_rate": 0.00017432154692122441, + "loss": 0.2376, + "step": 9960 + }, + { + "epoch": 0.2335015266325592, + "grad_norm": 2.8242733478546143, + "learning_rate": 0.00017431661837593112, + "loss": 1.6134, + "step": 9961 + }, + { + "epoch": 0.23352496820736418, + "grad_norm": 3.045286178588867, + "learning_rate": 0.00017431168942739375, + "loss": 0.2913, + "step": 9962 + }, + { + "epoch": 0.23354840978216918, + "grad_norm": 3.6092264652252197, + "learning_rate": 0.00017430676007563906, + "loss": 0.3873, + "step": 9963 + }, + { + "epoch": 0.23357185135697417, + "grad_norm": 3.8150341510772705, + "learning_rate": 0.00017430183032069382, + "loss": 0.4406, + "step": 9964 + }, + { + "epoch": 0.23359529293177916, + "grad_norm": 2.5462427139282227, + "learning_rate": 0.00017429690016258474, + "loss": 1.4888, + "step": 9965 + }, + { + "epoch": 0.23361873450658416, + "grad_norm": 3.3802857398986816, + "learning_rate": 0.00017429196960133857, + "loss": 0.2415, + "step": 9966 + }, + { + "epoch": 0.23364217608138915, + "grad_norm": 4.195024013519287, + "learning_rate": 0.00017428703863698212, + "loss": 0.536, + "step": 9967 + }, + { + "epoch": 0.23366561765619415, + "grad_norm": 1.208540916442871, + "learning_rate": 0.0001742821072695421, + "loss": 0.0732, + "step": 9968 + }, + { + "epoch": 0.23368905923099914, + "grad_norm": 1.1818063259124756, + "learning_rate": 0.00017427717549904528, + "loss": 0.1209, + "step": 9969 + }, + { + "epoch": 0.23371250080580414, + "grad_norm": 7.184091091156006, + "learning_rate": 0.00017427224332551843, + "loss": 0.5022, + "step": 9970 + }, + { + "epoch": 0.23373594238060913, + "grad_norm": 2.7205560207366943, + "learning_rate": 0.00017426731074898829, + "loss": 0.3638, + "step": 9971 + }, + { + "epoch": 0.23375938395541412, + "grad_norm": 4.55029296875, + "learning_rate": 0.0001742623777694816, + "loss": 0.7588, + "step": 9972 + }, + { + "epoch": 0.23378282553021912, + "grad_norm": 1.6668490171432495, + "learning_rate": 0.00017425744438702518, + "loss": 0.1257, + "step": 9973 + }, + { + "epoch": 0.2338062671050241, + "grad_norm": 4.3736982345581055, + "learning_rate": 0.0001742525106016458, + "loss": 0.5336, + "step": 9974 + }, + { + "epoch": 0.2338297086798291, + "grad_norm": 4.042030334472656, + "learning_rate": 0.0001742475764133702, + "loss": 0.5146, + "step": 9975 + }, + { + "epoch": 0.2338531502546341, + "grad_norm": 3.618689775466919, + "learning_rate": 0.00017424264182222513, + "loss": 0.4647, + "step": 9976 + }, + { + "epoch": 0.2338765918294391, + "grad_norm": 2.659057378768921, + "learning_rate": 0.00017423770682823743, + "loss": 0.3295, + "step": 9977 + }, + { + "epoch": 0.2339000334042441, + "grad_norm": 3.224095582962036, + "learning_rate": 0.00017423277143143384, + "loss": 0.3051, + "step": 9978 + }, + { + "epoch": 0.23392347497904908, + "grad_norm": 1.495659351348877, + "learning_rate": 0.00017422783563184115, + "loss": 0.1771, + "step": 9979 + }, + { + "epoch": 0.23394691655385408, + "grad_norm": 1.8015254735946655, + "learning_rate": 0.00017422289942948616, + "loss": 0.3085, + "step": 9980 + }, + { + "epoch": 0.23397035812865907, + "grad_norm": 0.7044347524642944, + "learning_rate": 0.00017421796282439557, + "loss": 0.0764, + "step": 9981 + }, + { + "epoch": 0.23399379970346407, + "grad_norm": 3.679143190383911, + "learning_rate": 0.00017421302581659628, + "loss": 0.4993, + "step": 9982 + }, + { + "epoch": 0.23401724127826906, + "grad_norm": 4.0300726890563965, + "learning_rate": 0.00017420808840611497, + "loss": 0.7719, + "step": 9983 + }, + { + "epoch": 0.23404068285307406, + "grad_norm": 4.907808303833008, + "learning_rate": 0.00017420315059297854, + "loss": 0.5465, + "step": 9984 + }, + { + "epoch": 0.23406412442787908, + "grad_norm": 3.7765398025512695, + "learning_rate": 0.00017419821237721372, + "loss": 0.4036, + "step": 9985 + }, + { + "epoch": 0.23408756600268407, + "grad_norm": 4.033378601074219, + "learning_rate": 0.00017419327375884727, + "loss": 0.4463, + "step": 9986 + }, + { + "epoch": 0.23411100757748907, + "grad_norm": 3.2335684299468994, + "learning_rate": 0.00017418833473790607, + "loss": 0.3213, + "step": 9987 + }, + { + "epoch": 0.23413444915229406, + "grad_norm": 3.6453120708465576, + "learning_rate": 0.00017418339531441687, + "loss": 0.4524, + "step": 9988 + }, + { + "epoch": 0.23415789072709906, + "grad_norm": 5.37164831161499, + "learning_rate": 0.0001741784554884065, + "loss": 0.4276, + "step": 9989 + }, + { + "epoch": 0.23418133230190405, + "grad_norm": 2.6369926929473877, + "learning_rate": 0.00017417351525990174, + "loss": 0.3286, + "step": 9990 + }, + { + "epoch": 0.23420477387670904, + "grad_norm": 5.471822738647461, + "learning_rate": 0.0001741685746289294, + "loss": 0.6622, + "step": 9991 + }, + { + "epoch": 0.23422821545151404, + "grad_norm": 2.2428693771362305, + "learning_rate": 0.00017416363359551627, + "loss": 1.1291, + "step": 9992 + }, + { + "epoch": 0.23425165702631903, + "grad_norm": 2.6567230224609375, + "learning_rate": 0.00017415869215968915, + "loss": 2.1709, + "step": 9993 + }, + { + "epoch": 0.23427509860112403, + "grad_norm": 5.2668375968933105, + "learning_rate": 0.00017415375032147495, + "loss": 0.5777, + "step": 9994 + }, + { + "epoch": 0.23429854017592902, + "grad_norm": 5.192957878112793, + "learning_rate": 0.00017414880808090037, + "loss": 0.5224, + "step": 9995 + }, + { + "epoch": 0.23432198175073402, + "grad_norm": 1.8548758029937744, + "learning_rate": 0.00017414386543799227, + "loss": 0.2032, + "step": 9996 + }, + { + "epoch": 0.234345423325539, + "grad_norm": 2.5812621116638184, + "learning_rate": 0.00017413892239277748, + "loss": 0.3909, + "step": 9997 + }, + { + "epoch": 0.234368864900344, + "grad_norm": 3.4611623287200928, + "learning_rate": 0.00017413397894528284, + "loss": 0.3278, + "step": 9998 + }, + { + "epoch": 0.234392306475149, + "grad_norm": 4.287395477294922, + "learning_rate": 0.00017412903509553512, + "loss": 0.3902, + "step": 9999 + }, + { + "epoch": 0.234415748049954, + "grad_norm": 6.3582916259765625, + "learning_rate": 0.00017412409084356122, + "loss": 0.7235, + "step": 10000 + }, + { + "epoch": 0.234439189624759, + "grad_norm": 0.9962946176528931, + "learning_rate": 0.00017411914618938786, + "loss": 0.1531, + "step": 10001 + }, + { + "epoch": 0.23446263119956398, + "grad_norm": 5.068014621734619, + "learning_rate": 0.00017411420113304192, + "loss": 0.6029, + "step": 10002 + }, + { + "epoch": 0.23448607277436898, + "grad_norm": 3.3055191040039062, + "learning_rate": 0.00017410925567455028, + "loss": 0.2363, + "step": 10003 + }, + { + "epoch": 0.23450951434917397, + "grad_norm": 2.2498254776000977, + "learning_rate": 0.0001741043098139397, + "loss": 0.5985, + "step": 10004 + }, + { + "epoch": 0.23453295592397896, + "grad_norm": 4.2765116691589355, + "learning_rate": 0.0001740993635512371, + "loss": 0.6747, + "step": 10005 + }, + { + "epoch": 0.23455639749878396, + "grad_norm": 3.911414384841919, + "learning_rate": 0.00017409441688646924, + "loss": 0.6569, + "step": 10006 + }, + { + "epoch": 0.23457983907358895, + "grad_norm": 2.9614577293395996, + "learning_rate": 0.00017408946981966298, + "loss": 0.3754, + "step": 10007 + }, + { + "epoch": 0.23460328064839395, + "grad_norm": 2.905491590499878, + "learning_rate": 0.0001740845223508452, + "loss": 1.4849, + "step": 10008 + }, + { + "epoch": 0.23462672222319894, + "grad_norm": 0.8036805391311646, + "learning_rate": 0.0001740795744800427, + "loss": 0.0958, + "step": 10009 + }, + { + "epoch": 0.23465016379800396, + "grad_norm": 3.670513153076172, + "learning_rate": 0.00017407462620728235, + "loss": 0.2846, + "step": 10010 + }, + { + "epoch": 0.23467360537280896, + "grad_norm": 3.26336932182312, + "learning_rate": 0.00017406967753259102, + "loss": 0.2754, + "step": 10011 + }, + { + "epoch": 0.23469704694761395, + "grad_norm": 3.2403130531311035, + "learning_rate": 0.00017406472845599548, + "loss": 0.3811, + "step": 10012 + }, + { + "epoch": 0.23472048852241895, + "grad_norm": 4.50806188583374, + "learning_rate": 0.00017405977897752269, + "loss": 0.3188, + "step": 10013 + }, + { + "epoch": 0.23474393009722394, + "grad_norm": 1.05461585521698, + "learning_rate": 0.0001740548290971994, + "loss": 0.1263, + "step": 10014 + }, + { + "epoch": 0.23476737167202893, + "grad_norm": 3.396860122680664, + "learning_rate": 0.0001740498788150526, + "loss": 0.2935, + "step": 10015 + }, + { + "epoch": 0.23479081324683393, + "grad_norm": 1.6841868162155151, + "learning_rate": 0.00017404492813110904, + "loss": 0.1645, + "step": 10016 + }, + { + "epoch": 0.23481425482163892, + "grad_norm": 5.041261672973633, + "learning_rate": 0.00017403997704539562, + "loss": 0.472, + "step": 10017 + }, + { + "epoch": 0.23483769639644392, + "grad_norm": 8.480001449584961, + "learning_rate": 0.00017403502555793918, + "loss": 0.6254, + "step": 10018 + }, + { + "epoch": 0.2348611379712489, + "grad_norm": 2.662806510925293, + "learning_rate": 0.00017403007366876666, + "loss": 0.2617, + "step": 10019 + }, + { + "epoch": 0.2348845795460539, + "grad_norm": 4.304887294769287, + "learning_rate": 0.00017402512137790484, + "loss": 0.5977, + "step": 10020 + }, + { + "epoch": 0.2349080211208589, + "grad_norm": 1.5489981174468994, + "learning_rate": 0.00017402016868538062, + "loss": 0.1659, + "step": 10021 + }, + { + "epoch": 0.2349314626956639, + "grad_norm": 3.7370173931121826, + "learning_rate": 0.00017401521559122092, + "loss": 0.4302, + "step": 10022 + }, + { + "epoch": 0.2349549042704689, + "grad_norm": 2.4113268852233887, + "learning_rate": 0.00017401026209545254, + "loss": 1.5462, + "step": 10023 + }, + { + "epoch": 0.23497834584527388, + "grad_norm": 1.0395652055740356, + "learning_rate": 0.00017400530819810243, + "loss": 0.1223, + "step": 10024 + }, + { + "epoch": 0.23500178742007888, + "grad_norm": 4.306096076965332, + "learning_rate": 0.00017400035389919744, + "loss": 0.489, + "step": 10025 + }, + { + "epoch": 0.23502522899488387, + "grad_norm": 0.8536907434463501, + "learning_rate": 0.00017399539919876443, + "loss": 0.0856, + "step": 10026 + }, + { + "epoch": 0.23504867056968887, + "grad_norm": 2.2087814807891846, + "learning_rate": 0.0001739904440968303, + "loss": 0.2101, + "step": 10027 + }, + { + "epoch": 0.23507211214449386, + "grad_norm": 0.7367491126060486, + "learning_rate": 0.00017398548859342197, + "loss": 0.0555, + "step": 10028 + }, + { + "epoch": 0.23509555371929886, + "grad_norm": 4.483216762542725, + "learning_rate": 0.00017398053268856626, + "loss": 0.4929, + "step": 10029 + }, + { + "epoch": 0.23511899529410385, + "grad_norm": 3.950061321258545, + "learning_rate": 0.0001739755763822901, + "loss": 0.5417, + "step": 10030 + }, + { + "epoch": 0.23514243686890884, + "grad_norm": 3.375553607940674, + "learning_rate": 0.00017397061967462042, + "loss": 0.3794, + "step": 10031 + }, + { + "epoch": 0.23516587844371384, + "grad_norm": 3.1309263706207275, + "learning_rate": 0.00017396566256558407, + "loss": 0.3228, + "step": 10032 + }, + { + "epoch": 0.23518932001851883, + "grad_norm": 2.326352596282959, + "learning_rate": 0.00017396070505520795, + "loss": 0.2278, + "step": 10033 + }, + { + "epoch": 0.23521276159332383, + "grad_norm": 5.439675807952881, + "learning_rate": 0.00017395574714351895, + "loss": 0.7497, + "step": 10034 + }, + { + "epoch": 0.23523620316812882, + "grad_norm": 1.2464900016784668, + "learning_rate": 0.000173950788830544, + "loss": 0.1557, + "step": 10035 + }, + { + "epoch": 0.23525964474293384, + "grad_norm": 4.143996715545654, + "learning_rate": 0.00017394583011630995, + "loss": 0.5496, + "step": 10036 + }, + { + "epoch": 0.23528308631773884, + "grad_norm": 2.7545909881591797, + "learning_rate": 0.0001739408710008438, + "loss": 0.4741, + "step": 10037 + }, + { + "epoch": 0.23530652789254383, + "grad_norm": 0.8976296782493591, + "learning_rate": 0.00017393591148417238, + "loss": 0.1199, + "step": 10038 + }, + { + "epoch": 0.23532996946734883, + "grad_norm": 3.662771701812744, + "learning_rate": 0.00017393095156632263, + "loss": 0.2533, + "step": 10039 + }, + { + "epoch": 0.23535341104215382, + "grad_norm": 5.263194561004639, + "learning_rate": 0.00017392599124732146, + "loss": 0.7515, + "step": 10040 + }, + { + "epoch": 0.23537685261695881, + "grad_norm": 3.2896676063537598, + "learning_rate": 0.0001739210305271958, + "loss": 0.3455, + "step": 10041 + }, + { + "epoch": 0.2354002941917638, + "grad_norm": 3.0754666328430176, + "learning_rate": 0.00017391606940597254, + "loss": 0.332, + "step": 10042 + }, + { + "epoch": 0.2354237357665688, + "grad_norm": 6.269508361816406, + "learning_rate": 0.0001739111078836786, + "loss": 0.8342, + "step": 10043 + }, + { + "epoch": 0.2354471773413738, + "grad_norm": 3.5364911556243896, + "learning_rate": 0.00017390614596034088, + "loss": 0.3474, + "step": 10044 + }, + { + "epoch": 0.2354706189161788, + "grad_norm": 5.442033767700195, + "learning_rate": 0.00017390118363598637, + "loss": 0.6116, + "step": 10045 + }, + { + "epoch": 0.2354940604909838, + "grad_norm": 1.0323044061660767, + "learning_rate": 0.00017389622091064198, + "loss": 0.1772, + "step": 10046 + }, + { + "epoch": 0.23551750206578878, + "grad_norm": 4.911376953125, + "learning_rate": 0.00017389125778433456, + "loss": 0.3224, + "step": 10047 + }, + { + "epoch": 0.23554094364059378, + "grad_norm": 2.9864847660064697, + "learning_rate": 0.00017388629425709113, + "loss": 0.3998, + "step": 10048 + }, + { + "epoch": 0.23556438521539877, + "grad_norm": 0.8661161065101624, + "learning_rate": 0.0001738813303289386, + "loss": 0.1594, + "step": 10049 + }, + { + "epoch": 0.23558782679020376, + "grad_norm": 1.8176053762435913, + "learning_rate": 0.0001738763659999039, + "loss": 0.2057, + "step": 10050 + }, + { + "epoch": 0.23561126836500876, + "grad_norm": 3.3831281661987305, + "learning_rate": 0.00017387140127001394, + "loss": 0.3, + "step": 10051 + }, + { + "epoch": 0.23563470993981375, + "grad_norm": 3.6191840171813965, + "learning_rate": 0.00017386643613929566, + "loss": 0.3763, + "step": 10052 + }, + { + "epoch": 0.23565815151461875, + "grad_norm": 2.6901066303253174, + "learning_rate": 0.00017386147060777606, + "loss": 0.4014, + "step": 10053 + }, + { + "epoch": 0.23568159308942374, + "grad_norm": 7.578917980194092, + "learning_rate": 0.00017385650467548205, + "loss": 1.1586, + "step": 10054 + }, + { + "epoch": 0.23570503466422874, + "grad_norm": 1.3408117294311523, + "learning_rate": 0.00017385153834244052, + "loss": 0.127, + "step": 10055 + }, + { + "epoch": 0.23572847623903373, + "grad_norm": 3.185823917388916, + "learning_rate": 0.0001738465716086785, + "loss": 0.3399, + "step": 10056 + }, + { + "epoch": 0.23575191781383872, + "grad_norm": 3.36484694480896, + "learning_rate": 0.00017384160447422292, + "loss": 0.2032, + "step": 10057 + }, + { + "epoch": 0.23577535938864372, + "grad_norm": 2.499682664871216, + "learning_rate": 0.00017383663693910069, + "loss": 0.2638, + "step": 10058 + }, + { + "epoch": 0.2357988009634487, + "grad_norm": 5.015158176422119, + "learning_rate": 0.0001738316690033388, + "loss": 0.6673, + "step": 10059 + }, + { + "epoch": 0.2358222425382537, + "grad_norm": 4.315722465515137, + "learning_rate": 0.00017382670066696417, + "loss": 0.642, + "step": 10060 + }, + { + "epoch": 0.23584568411305873, + "grad_norm": 4.134505271911621, + "learning_rate": 0.00017382173193000383, + "loss": 0.3717, + "step": 10061 + }, + { + "epoch": 0.23586912568786372, + "grad_norm": 3.7153451442718506, + "learning_rate": 0.0001738167627924847, + "loss": 0.3552, + "step": 10062 + }, + { + "epoch": 0.23589256726266872, + "grad_norm": 3.8447067737579346, + "learning_rate": 0.00017381179325443372, + "loss": 0.6924, + "step": 10063 + }, + { + "epoch": 0.2359160088374737, + "grad_norm": 2.075697660446167, + "learning_rate": 0.00017380682331587789, + "loss": 0.2331, + "step": 10064 + }, + { + "epoch": 0.2359394504122787, + "grad_norm": 3.4934091567993164, + "learning_rate": 0.00017380185297684412, + "loss": 1.7258, + "step": 10065 + }, + { + "epoch": 0.2359628919870837, + "grad_norm": 2.729421854019165, + "learning_rate": 0.00017379688223735946, + "loss": 0.3554, + "step": 10066 + }, + { + "epoch": 0.2359863335618887, + "grad_norm": 3.1929712295532227, + "learning_rate": 0.0001737919110974508, + "loss": 0.3611, + "step": 10067 + }, + { + "epoch": 0.2360097751366937, + "grad_norm": 3.2134859561920166, + "learning_rate": 0.0001737869395571452, + "loss": 1.475, + "step": 10068 + }, + { + "epoch": 0.23603321671149868, + "grad_norm": 3.2424585819244385, + "learning_rate": 0.00017378196761646958, + "loss": 0.1908, + "step": 10069 + }, + { + "epoch": 0.23605665828630368, + "grad_norm": 4.9886884689331055, + "learning_rate": 0.00017377699527545092, + "loss": 0.6771, + "step": 10070 + }, + { + "epoch": 0.23608009986110867, + "grad_norm": 1.4223676919937134, + "learning_rate": 0.00017377202253411622, + "loss": 0.1113, + "step": 10071 + }, + { + "epoch": 0.23610354143591367, + "grad_norm": 2.699191093444824, + "learning_rate": 0.00017376704939249247, + "loss": 1.9588, + "step": 10072 + }, + { + "epoch": 0.23612698301071866, + "grad_norm": 3.538755178451538, + "learning_rate": 0.00017376207585060658, + "loss": 0.5066, + "step": 10073 + }, + { + "epoch": 0.23615042458552366, + "grad_norm": 3.162243366241455, + "learning_rate": 0.00017375710190848563, + "loss": 0.2545, + "step": 10074 + }, + { + "epoch": 0.23617386616032865, + "grad_norm": 3.7012109756469727, + "learning_rate": 0.00017375212756615658, + "loss": 0.479, + "step": 10075 + }, + { + "epoch": 0.23619730773513364, + "grad_norm": 2.860335350036621, + "learning_rate": 0.00017374715282364637, + "loss": 0.3206, + "step": 10076 + }, + { + "epoch": 0.23622074930993864, + "grad_norm": 3.9979777336120605, + "learning_rate": 0.0001737421776809821, + "loss": 0.4124, + "step": 10077 + }, + { + "epoch": 0.23624419088474363, + "grad_norm": 1.6827547550201416, + "learning_rate": 0.00017373720213819066, + "loss": 0.1902, + "step": 10078 + }, + { + "epoch": 0.23626763245954863, + "grad_norm": 4.916856288909912, + "learning_rate": 0.00017373222619529907, + "loss": 0.7253, + "step": 10079 + }, + { + "epoch": 0.23629107403435362, + "grad_norm": 3.1073431968688965, + "learning_rate": 0.0001737272498523344, + "loss": 1.7199, + "step": 10080 + }, + { + "epoch": 0.23631451560915862, + "grad_norm": 2.7708311080932617, + "learning_rate": 0.00017372227310932353, + "loss": 0.2757, + "step": 10081 + }, + { + "epoch": 0.2363379571839636, + "grad_norm": 8.092836380004883, + "learning_rate": 0.0001737172959662936, + "loss": 0.8641, + "step": 10082 + }, + { + "epoch": 0.2363613987587686, + "grad_norm": 4.633243083953857, + "learning_rate": 0.00017371231842327152, + "loss": 0.5583, + "step": 10083 + }, + { + "epoch": 0.2363848403335736, + "grad_norm": 4.283880233764648, + "learning_rate": 0.00017370734048028432, + "loss": 0.6854, + "step": 10084 + }, + { + "epoch": 0.2364082819083786, + "grad_norm": 2.4526515007019043, + "learning_rate": 0.00017370236213735903, + "loss": 0.2886, + "step": 10085 + }, + { + "epoch": 0.2364317234831836, + "grad_norm": 1.665914535522461, + "learning_rate": 0.00017369738339452265, + "loss": 0.2431, + "step": 10086 + }, + { + "epoch": 0.2364551650579886, + "grad_norm": 1.8723973035812378, + "learning_rate": 0.00017369240425180218, + "loss": 0.144, + "step": 10087 + }, + { + "epoch": 0.2364786066327936, + "grad_norm": 0.7998409867286682, + "learning_rate": 0.00017368742470922466, + "loss": 0.0624, + "step": 10088 + }, + { + "epoch": 0.2365020482075986, + "grad_norm": 1.0000075101852417, + "learning_rate": 0.0001736824447668171, + "loss": 0.1556, + "step": 10089 + }, + { + "epoch": 0.2365254897824036, + "grad_norm": 1.6020883321762085, + "learning_rate": 0.00017367746442460656, + "loss": 0.1155, + "step": 10090 + }, + { + "epoch": 0.23654893135720859, + "grad_norm": 0.8989757895469666, + "learning_rate": 0.00017367248368261996, + "loss": 0.1026, + "step": 10091 + }, + { + "epoch": 0.23657237293201358, + "grad_norm": 1.647414207458496, + "learning_rate": 0.00017366750254088444, + "loss": 0.3303, + "step": 10092 + }, + { + "epoch": 0.23659581450681857, + "grad_norm": 4.042585849761963, + "learning_rate": 0.00017366252099942697, + "loss": 0.5023, + "step": 10093 + }, + { + "epoch": 0.23661925608162357, + "grad_norm": 1.3598554134368896, + "learning_rate": 0.0001736575390582746, + "loss": 1.2587, + "step": 10094 + }, + { + "epoch": 0.23664269765642856, + "grad_norm": 3.18576717376709, + "learning_rate": 0.00017365255671745433, + "loss": 1.7602, + "step": 10095 + }, + { + "epoch": 0.23666613923123356, + "grad_norm": 1.0535670518875122, + "learning_rate": 0.00017364757397699324, + "loss": 0.0774, + "step": 10096 + }, + { + "epoch": 0.23668958080603855, + "grad_norm": 3.1194865703582764, + "learning_rate": 0.00017364259083691832, + "loss": 0.2691, + "step": 10097 + }, + { + "epoch": 0.23671302238084355, + "grad_norm": 3.545267105102539, + "learning_rate": 0.00017363760729725662, + "loss": 0.4262, + "step": 10098 + }, + { + "epoch": 0.23673646395564854, + "grad_norm": 1.5456688404083252, + "learning_rate": 0.00017363262335803523, + "loss": 0.5066, + "step": 10099 + }, + { + "epoch": 0.23675990553045353, + "grad_norm": 1.2808881998062134, + "learning_rate": 0.00017362763901928113, + "loss": 0.1783, + "step": 10100 + }, + { + "epoch": 0.23678334710525853, + "grad_norm": 4.273186206817627, + "learning_rate": 0.0001736226542810214, + "loss": 0.3904, + "step": 10101 + }, + { + "epoch": 0.23680678868006352, + "grad_norm": 1.362988829612732, + "learning_rate": 0.0001736176691432831, + "loss": 0.1764, + "step": 10102 + }, + { + "epoch": 0.23683023025486852, + "grad_norm": 2.8696489334106445, + "learning_rate": 0.00017361268360609322, + "loss": 0.3562, + "step": 10103 + }, + { + "epoch": 0.2368536718296735, + "grad_norm": 3.629133462905884, + "learning_rate": 0.00017360769766947888, + "loss": 0.5171, + "step": 10104 + }, + { + "epoch": 0.2368771134044785, + "grad_norm": 3.769366979598999, + "learning_rate": 0.00017360271133346706, + "loss": 0.4751, + "step": 10105 + }, + { + "epoch": 0.2369005549792835, + "grad_norm": 3.4723293781280518, + "learning_rate": 0.00017359772459808487, + "loss": 0.4743, + "step": 10106 + }, + { + "epoch": 0.2369239965540885, + "grad_norm": 4.378584861755371, + "learning_rate": 0.00017359273746335937, + "loss": 0.5055, + "step": 10107 + }, + { + "epoch": 0.2369474381288935, + "grad_norm": 2.1470680236816406, + "learning_rate": 0.00017358774992931762, + "loss": 0.2079, + "step": 10108 + }, + { + "epoch": 0.23697087970369848, + "grad_norm": 1.4712125062942505, + "learning_rate": 0.00017358276199598666, + "loss": 0.1342, + "step": 10109 + }, + { + "epoch": 0.23699432127850348, + "grad_norm": 3.2156918048858643, + "learning_rate": 0.00017357777366339356, + "loss": 0.4522, + "step": 10110 + }, + { + "epoch": 0.23701776285330847, + "grad_norm": 0.6372902393341064, + "learning_rate": 0.00017357278493156542, + "loss": 0.0719, + "step": 10111 + }, + { + "epoch": 0.2370412044281135, + "grad_norm": 1.3034477233886719, + "learning_rate": 0.00017356779580052925, + "loss": 0.2002, + "step": 10112 + }, + { + "epoch": 0.2370646460029185, + "grad_norm": 3.0891146659851074, + "learning_rate": 0.00017356280627031214, + "loss": 0.2658, + "step": 10113 + }, + { + "epoch": 0.23708808757772348, + "grad_norm": 2.651665687561035, + "learning_rate": 0.00017355781634094118, + "loss": 0.2232, + "step": 10114 + }, + { + "epoch": 0.23711152915252848, + "grad_norm": 3.007124662399292, + "learning_rate": 0.00017355282601244344, + "loss": 0.2339, + "step": 10115 + }, + { + "epoch": 0.23713497072733347, + "grad_norm": 0.7425210475921631, + "learning_rate": 0.00017354783528484603, + "loss": 0.0754, + "step": 10116 + }, + { + "epoch": 0.23715841230213847, + "grad_norm": 3.729933023452759, + "learning_rate": 0.00017354284415817598, + "loss": 0.3428, + "step": 10117 + }, + { + "epoch": 0.23718185387694346, + "grad_norm": 3.63676118850708, + "learning_rate": 0.00017353785263246038, + "loss": 0.431, + "step": 10118 + }, + { + "epoch": 0.23720529545174845, + "grad_norm": 4.454409122467041, + "learning_rate": 0.00017353286070772631, + "loss": 0.4694, + "step": 10119 + }, + { + "epoch": 0.23722873702655345, + "grad_norm": 3.684250593185425, + "learning_rate": 0.0001735278683840009, + "loss": 0.5781, + "step": 10120 + }, + { + "epoch": 0.23725217860135844, + "grad_norm": 1.38555109500885, + "learning_rate": 0.0001735228756613112, + "loss": 0.1036, + "step": 10121 + }, + { + "epoch": 0.23727562017616344, + "grad_norm": 2.791501045227051, + "learning_rate": 0.00017351788253968433, + "loss": 0.371, + "step": 10122 + }, + { + "epoch": 0.23729906175096843, + "grad_norm": 4.820021152496338, + "learning_rate": 0.00017351288901914732, + "loss": 0.6087, + "step": 10123 + }, + { + "epoch": 0.23732250332577343, + "grad_norm": 3.765479326248169, + "learning_rate": 0.00017350789509972735, + "loss": 0.4589, + "step": 10124 + }, + { + "epoch": 0.23734594490057842, + "grad_norm": 2.090442657470703, + "learning_rate": 0.00017350290078145143, + "loss": 1.0201, + "step": 10125 + }, + { + "epoch": 0.23736938647538341, + "grad_norm": 2.6733498573303223, + "learning_rate": 0.00017349790606434673, + "loss": 1.3842, + "step": 10126 + }, + { + "epoch": 0.2373928280501884, + "grad_norm": 8.018216133117676, + "learning_rate": 0.00017349291094844032, + "loss": 0.2359, + "step": 10127 + }, + { + "epoch": 0.2374162696249934, + "grad_norm": 3.2644970417022705, + "learning_rate": 0.00017348791543375932, + "loss": 0.3938, + "step": 10128 + }, + { + "epoch": 0.2374397111997984, + "grad_norm": 4.997384071350098, + "learning_rate": 0.0001734829195203308, + "loss": 0.575, + "step": 10129 + }, + { + "epoch": 0.2374631527746034, + "grad_norm": 5.978076457977295, + "learning_rate": 0.0001734779232081819, + "loss": 0.7335, + "step": 10130 + }, + { + "epoch": 0.2374865943494084, + "grad_norm": 5.764984607696533, + "learning_rate": 0.00017347292649733974, + "loss": 0.6224, + "step": 10131 + }, + { + "epoch": 0.23751003592421338, + "grad_norm": 3.464031219482422, + "learning_rate": 0.0001734679293878314, + "loss": 0.4809, + "step": 10132 + }, + { + "epoch": 0.23753347749901838, + "grad_norm": 3.2502505779266357, + "learning_rate": 0.000173462931879684, + "loss": 0.4499, + "step": 10133 + }, + { + "epoch": 0.23755691907382337, + "grad_norm": 4.934238910675049, + "learning_rate": 0.0001734579339729247, + "loss": 0.4379, + "step": 10134 + }, + { + "epoch": 0.23758036064862836, + "grad_norm": 2.9161462783813477, + "learning_rate": 0.00017345293566758057, + "loss": 2.0153, + "step": 10135 + }, + { + "epoch": 0.23760380222343336, + "grad_norm": 1.6613279581069946, + "learning_rate": 0.00017344793696367873, + "loss": 0.1788, + "step": 10136 + }, + { + "epoch": 0.23762724379823835, + "grad_norm": 3.51501202583313, + "learning_rate": 0.00017344293786124632, + "loss": 0.3393, + "step": 10137 + }, + { + "epoch": 0.23765068537304337, + "grad_norm": 1.5219857692718506, + "learning_rate": 0.0001734379383603105, + "loss": 0.2084, + "step": 10138 + }, + { + "epoch": 0.23767412694784837, + "grad_norm": 3.0996010303497314, + "learning_rate": 0.00017343293846089832, + "loss": 0.3042, + "step": 10139 + }, + { + "epoch": 0.23769756852265336, + "grad_norm": 3.7821578979492188, + "learning_rate": 0.00017342793816303695, + "loss": 0.1653, + "step": 10140 + }, + { + "epoch": 0.23772101009745836, + "grad_norm": 4.00208044052124, + "learning_rate": 0.00017342293746675357, + "loss": 0.4373, + "step": 10141 + }, + { + "epoch": 0.23774445167226335, + "grad_norm": 1.5800219774246216, + "learning_rate": 0.0001734179363720752, + "loss": 0.1364, + "step": 10142 + }, + { + "epoch": 0.23776789324706835, + "grad_norm": 4.063129425048828, + "learning_rate": 0.0001734129348790291, + "loss": 0.5224, + "step": 10143 + }, + { + "epoch": 0.23779133482187334, + "grad_norm": 2.4841842651367188, + "learning_rate": 0.00017340793298764232, + "loss": 0.2884, + "step": 10144 + }, + { + "epoch": 0.23781477639667833, + "grad_norm": 2.6238224506378174, + "learning_rate": 0.00017340293069794204, + "loss": 0.2541, + "step": 10145 + }, + { + "epoch": 0.23783821797148333, + "grad_norm": 5.891435146331787, + "learning_rate": 0.00017339792800995541, + "loss": 0.3684, + "step": 10146 + }, + { + "epoch": 0.23786165954628832, + "grad_norm": 0.8371698260307312, + "learning_rate": 0.00017339292492370953, + "loss": 0.1025, + "step": 10147 + }, + { + "epoch": 0.23788510112109332, + "grad_norm": 3.377436399459839, + "learning_rate": 0.00017338792143923158, + "loss": 0.2127, + "step": 10148 + }, + { + "epoch": 0.2379085426958983, + "grad_norm": 5.372186183929443, + "learning_rate": 0.00017338291755654873, + "loss": 0.4715, + "step": 10149 + }, + { + "epoch": 0.2379319842707033, + "grad_norm": 3.006120204925537, + "learning_rate": 0.0001733779132756881, + "loss": 0.2733, + "step": 10150 + }, + { + "epoch": 0.2379554258455083, + "grad_norm": 4.399667263031006, + "learning_rate": 0.00017337290859667683, + "loss": 0.6758, + "step": 10151 + }, + { + "epoch": 0.2379788674203133, + "grad_norm": 3.3022546768188477, + "learning_rate": 0.0001733679035195421, + "loss": 0.3599, + "step": 10152 + }, + { + "epoch": 0.2380023089951183, + "grad_norm": 4.098820686340332, + "learning_rate": 0.00017336289804431107, + "loss": 0.3637, + "step": 10153 + }, + { + "epoch": 0.23802575056992328, + "grad_norm": 3.292390823364258, + "learning_rate": 0.00017335789217101088, + "loss": 0.375, + "step": 10154 + }, + { + "epoch": 0.23804919214472828, + "grad_norm": 4.609374523162842, + "learning_rate": 0.00017335288589966874, + "loss": 0.4671, + "step": 10155 + }, + { + "epoch": 0.23807263371953327, + "grad_norm": 3.8056488037109375, + "learning_rate": 0.00017334787923031176, + "loss": 0.3502, + "step": 10156 + }, + { + "epoch": 0.23809607529433827, + "grad_norm": 3.7470335960388184, + "learning_rate": 0.00017334287216296709, + "loss": 0.4109, + "step": 10157 + }, + { + "epoch": 0.23811951686914326, + "grad_norm": 2.2660763263702393, + "learning_rate": 0.00017333786469766197, + "loss": 0.1955, + "step": 10158 + }, + { + "epoch": 0.23814295844394825, + "grad_norm": 1.2295589447021484, + "learning_rate": 0.00017333285683442354, + "loss": 0.1245, + "step": 10159 + }, + { + "epoch": 0.23816640001875325, + "grad_norm": 5.996531009674072, + "learning_rate": 0.00017332784857327895, + "loss": 0.6666, + "step": 10160 + }, + { + "epoch": 0.23818984159355824, + "grad_norm": 3.276516914367676, + "learning_rate": 0.00017332283991425537, + "loss": 0.3201, + "step": 10161 + }, + { + "epoch": 0.23821328316836324, + "grad_norm": 7.420753002166748, + "learning_rate": 0.00017331783085738006, + "loss": 0.7148, + "step": 10162 + }, + { + "epoch": 0.23823672474316823, + "grad_norm": 2.2490150928497314, + "learning_rate": 0.00017331282140268007, + "loss": 1.5934, + "step": 10163 + }, + { + "epoch": 0.23826016631797325, + "grad_norm": 2.45770263671875, + "learning_rate": 0.00017330781155018267, + "loss": 1.1509, + "step": 10164 + }, + { + "epoch": 0.23828360789277825, + "grad_norm": 3.6628050804138184, + "learning_rate": 0.00017330280129991505, + "loss": 0.3108, + "step": 10165 + }, + { + "epoch": 0.23830704946758324, + "grad_norm": 3.5077807903289795, + "learning_rate": 0.00017329779065190434, + "loss": 0.4023, + "step": 10166 + }, + { + "epoch": 0.23833049104238824, + "grad_norm": 3.4914562702178955, + "learning_rate": 0.00017329277960617778, + "loss": 1.9322, + "step": 10167 + }, + { + "epoch": 0.23835393261719323, + "grad_norm": 1.047777771949768, + "learning_rate": 0.00017328776816276253, + "loss": 0.1127, + "step": 10168 + }, + { + "epoch": 0.23837737419199823, + "grad_norm": 4.25213098526001, + "learning_rate": 0.0001732827563216858, + "loss": 0.4126, + "step": 10169 + }, + { + "epoch": 0.23840081576680322, + "grad_norm": 0.47927719354629517, + "learning_rate": 0.00017327774408297473, + "loss": 0.0656, + "step": 10170 + }, + { + "epoch": 0.23842425734160821, + "grad_norm": 3.000150442123413, + "learning_rate": 0.0001732727314466566, + "loss": 0.2028, + "step": 10171 + }, + { + "epoch": 0.2384476989164132, + "grad_norm": 1.5498861074447632, + "learning_rate": 0.00017326771841275857, + "loss": 0.2089, + "step": 10172 + }, + { + "epoch": 0.2384711404912182, + "grad_norm": 4.904553413391113, + "learning_rate": 0.00017326270498130782, + "loss": 0.524, + "step": 10173 + }, + { + "epoch": 0.2384945820660232, + "grad_norm": 2.334737539291382, + "learning_rate": 0.00017325769115233159, + "loss": 0.4187, + "step": 10174 + }, + { + "epoch": 0.2385180236408282, + "grad_norm": 1.0787075757980347, + "learning_rate": 0.00017325267692585707, + "loss": 0.1104, + "step": 10175 + }, + { + "epoch": 0.23854146521563319, + "grad_norm": 0.6822150349617004, + "learning_rate": 0.00017324766230191144, + "loss": 0.049, + "step": 10176 + }, + { + "epoch": 0.23856490679043818, + "grad_norm": 0.4300934970378876, + "learning_rate": 0.00017324264728052198, + "loss": 0.0564, + "step": 10177 + }, + { + "epoch": 0.23858834836524317, + "grad_norm": 1.3464628458023071, + "learning_rate": 0.0001732376318617158, + "loss": 0.1071, + "step": 10178 + }, + { + "epoch": 0.23861178994004817, + "grad_norm": 5.345390796661377, + "learning_rate": 0.0001732326160455202, + "loss": 0.6673, + "step": 10179 + }, + { + "epoch": 0.23863523151485316, + "grad_norm": 3.5866661071777344, + "learning_rate": 0.00017322759983196234, + "loss": 0.5124, + "step": 10180 + }, + { + "epoch": 0.23865867308965816, + "grad_norm": 2.9387476444244385, + "learning_rate": 0.00017322258322106953, + "loss": 1.9837, + "step": 10181 + }, + { + "epoch": 0.23868211466446315, + "grad_norm": 1.2410794496536255, + "learning_rate": 0.00017321756621286886, + "loss": 0.1563, + "step": 10182 + }, + { + "epoch": 0.23870555623926815, + "grad_norm": 2.498551368713379, + "learning_rate": 0.00017321254880738764, + "loss": 0.2601, + "step": 10183 + }, + { + "epoch": 0.23872899781407314, + "grad_norm": 1.120159387588501, + "learning_rate": 0.00017320753100465307, + "loss": 0.1414, + "step": 10184 + }, + { + "epoch": 0.23875243938887813, + "grad_norm": 0.7163384556770325, + "learning_rate": 0.00017320251280469239, + "loss": 0.0984, + "step": 10185 + }, + { + "epoch": 0.23877588096368313, + "grad_norm": 1.7982769012451172, + "learning_rate": 0.0001731974942075328, + "loss": 0.8669, + "step": 10186 + }, + { + "epoch": 0.23879932253848812, + "grad_norm": 5.3567657470703125, + "learning_rate": 0.00017319247521320157, + "loss": 0.4768, + "step": 10187 + }, + { + "epoch": 0.23882276411329312, + "grad_norm": 2.6404833793640137, + "learning_rate": 0.00017318745582172587, + "loss": 0.2703, + "step": 10188 + }, + { + "epoch": 0.23884620568809814, + "grad_norm": 3.7942862510681152, + "learning_rate": 0.00017318243603313302, + "loss": 2.0487, + "step": 10189 + }, + { + "epoch": 0.23886964726290313, + "grad_norm": 1.0863975286483765, + "learning_rate": 0.00017317741584745017, + "loss": 0.1171, + "step": 10190 + }, + { + "epoch": 0.23889308883770813, + "grad_norm": 2.6159188747406006, + "learning_rate": 0.00017317239526470462, + "loss": 1.0168, + "step": 10191 + }, + { + "epoch": 0.23891653041251312, + "grad_norm": 8.475822448730469, + "learning_rate": 0.00017316737428492362, + "loss": 0.8007, + "step": 10192 + }, + { + "epoch": 0.23893997198731812, + "grad_norm": 1.1373966932296753, + "learning_rate": 0.00017316235290813442, + "loss": 0.1156, + "step": 10193 + }, + { + "epoch": 0.2389634135621231, + "grad_norm": 2.23590087890625, + "learning_rate": 0.00017315733113436416, + "loss": 0.2887, + "step": 10194 + }, + { + "epoch": 0.2389868551369281, + "grad_norm": 3.202348470687866, + "learning_rate": 0.00017315230896364022, + "loss": 0.4672, + "step": 10195 + }, + { + "epoch": 0.2390102967117331, + "grad_norm": 3.85345196723938, + "learning_rate": 0.00017314728639598976, + "loss": 0.4696, + "step": 10196 + }, + { + "epoch": 0.2390337382865381, + "grad_norm": 2.8821449279785156, + "learning_rate": 0.0001731422634314401, + "loss": 0.3476, + "step": 10197 + }, + { + "epoch": 0.2390571798613431, + "grad_norm": 0.5198764801025391, + "learning_rate": 0.0001731372400700184, + "loss": 0.055, + "step": 10198 + }, + { + "epoch": 0.23908062143614808, + "grad_norm": 1.4237476587295532, + "learning_rate": 0.00017313221631175206, + "loss": 0.2286, + "step": 10199 + }, + { + "epoch": 0.23910406301095308, + "grad_norm": 2.014848470687866, + "learning_rate": 0.0001731271921566682, + "loss": 1.6116, + "step": 10200 + }, + { + "epoch": 0.23912750458575807, + "grad_norm": 3.5590078830718994, + "learning_rate": 0.00017312216760479417, + "loss": 2.1709, + "step": 10201 + }, + { + "epoch": 0.23915094616056307, + "grad_norm": 2.6029133796691895, + "learning_rate": 0.00017311714265615718, + "loss": 0.2486, + "step": 10202 + }, + { + "epoch": 0.23917438773536806, + "grad_norm": 3.012662887573242, + "learning_rate": 0.00017311211731078452, + "loss": 0.3268, + "step": 10203 + }, + { + "epoch": 0.23919782931017305, + "grad_norm": 2.7501728534698486, + "learning_rate": 0.00017310709156870347, + "loss": 0.5313, + "step": 10204 + }, + { + "epoch": 0.23922127088497805, + "grad_norm": 3.1120927333831787, + "learning_rate": 0.00017310206542994126, + "loss": 0.2894, + "step": 10205 + }, + { + "epoch": 0.23924471245978304, + "grad_norm": 3.5033509731292725, + "learning_rate": 0.00017309703889452523, + "loss": 0.3413, + "step": 10206 + }, + { + "epoch": 0.23926815403458804, + "grad_norm": 4.1789870262146, + "learning_rate": 0.0001730920119624826, + "loss": 0.3764, + "step": 10207 + }, + { + "epoch": 0.23929159560939303, + "grad_norm": 3.189434766769409, + "learning_rate": 0.00017308698463384064, + "loss": 0.3054, + "step": 10208 + }, + { + "epoch": 0.23931503718419803, + "grad_norm": 1.6771109104156494, + "learning_rate": 0.00017308195690862664, + "loss": 0.1578, + "step": 10209 + }, + { + "epoch": 0.23933847875900302, + "grad_norm": 3.242142915725708, + "learning_rate": 0.0001730769287868679, + "loss": 0.3322, + "step": 10210 + }, + { + "epoch": 0.23936192033380801, + "grad_norm": 2.763909339904785, + "learning_rate": 0.0001730719002685917, + "loss": 0.3745, + "step": 10211 + }, + { + "epoch": 0.239385361908613, + "grad_norm": 3.9988393783569336, + "learning_rate": 0.0001730668713538253, + "loss": 0.3891, + "step": 10212 + }, + { + "epoch": 0.239408803483418, + "grad_norm": 0.7477399110794067, + "learning_rate": 0.000173061842042596, + "loss": 0.0839, + "step": 10213 + }, + { + "epoch": 0.239432245058223, + "grad_norm": 3.449376106262207, + "learning_rate": 0.00017305681233493108, + "loss": 0.2695, + "step": 10214 + }, + { + "epoch": 0.23945568663302802, + "grad_norm": 4.944403171539307, + "learning_rate": 0.00017305178223085788, + "loss": 0.4775, + "step": 10215 + }, + { + "epoch": 0.23947912820783301, + "grad_norm": 2.564826726913452, + "learning_rate": 0.00017304675173040362, + "loss": 0.423, + "step": 10216 + }, + { + "epoch": 0.239502569782638, + "grad_norm": 0.8993893265724182, + "learning_rate": 0.00017304172083359567, + "loss": 0.1289, + "step": 10217 + }, + { + "epoch": 0.239526011357443, + "grad_norm": 3.7810659408569336, + "learning_rate": 0.00017303668954046124, + "loss": 0.6698, + "step": 10218 + }, + { + "epoch": 0.239549452932248, + "grad_norm": 5.372539043426514, + "learning_rate": 0.0001730316578510277, + "loss": 0.8614, + "step": 10219 + }, + { + "epoch": 0.239572894507053, + "grad_norm": 2.358245372772217, + "learning_rate": 0.00017302662576532235, + "loss": 0.163, + "step": 10220 + }, + { + "epoch": 0.23959633608185799, + "grad_norm": 3.7984230518341064, + "learning_rate": 0.00017302159328337244, + "loss": 0.5595, + "step": 10221 + }, + { + "epoch": 0.23961977765666298, + "grad_norm": 0.6240661144256592, + "learning_rate": 0.00017301656040520533, + "loss": 0.054, + "step": 10222 + }, + { + "epoch": 0.23964321923146797, + "grad_norm": 2.801571846008301, + "learning_rate": 0.00017301152713084834, + "loss": 0.2792, + "step": 10223 + }, + { + "epoch": 0.23966666080627297, + "grad_norm": 3.170409917831421, + "learning_rate": 0.0001730064934603287, + "loss": 0.2619, + "step": 10224 + }, + { + "epoch": 0.23969010238107796, + "grad_norm": 1.042422890663147, + "learning_rate": 0.0001730014593936738, + "loss": 0.0868, + "step": 10225 + }, + { + "epoch": 0.23971354395588296, + "grad_norm": 3.9818522930145264, + "learning_rate": 0.00017299642493091093, + "loss": 0.4903, + "step": 10226 + }, + { + "epoch": 0.23973698553068795, + "grad_norm": 3.4805490970611572, + "learning_rate": 0.0001729913900720674, + "loss": 0.4758, + "step": 10227 + }, + { + "epoch": 0.23976042710549295, + "grad_norm": 1.2752028703689575, + "learning_rate": 0.00017298635481717055, + "loss": 0.1753, + "step": 10228 + }, + { + "epoch": 0.23978386868029794, + "grad_norm": 1.0124915838241577, + "learning_rate": 0.00017298131916624766, + "loss": 0.132, + "step": 10229 + }, + { + "epoch": 0.23980731025510293, + "grad_norm": 2.637683629989624, + "learning_rate": 0.0001729762831193261, + "loss": 0.3175, + "step": 10230 + }, + { + "epoch": 0.23983075182990793, + "grad_norm": 1.466436743736267, + "learning_rate": 0.00017297124667643317, + "loss": 0.1408, + "step": 10231 + }, + { + "epoch": 0.23985419340471292, + "grad_norm": 4.692999362945557, + "learning_rate": 0.0001729662098375962, + "loss": 0.4323, + "step": 10232 + }, + { + "epoch": 0.23987763497951792, + "grad_norm": 3.5552990436553955, + "learning_rate": 0.00017296117260284251, + "loss": 0.4196, + "step": 10233 + }, + { + "epoch": 0.2399010765543229, + "grad_norm": 3.302058696746826, + "learning_rate": 0.00017295613497219948, + "loss": 0.4447, + "step": 10234 + }, + { + "epoch": 0.2399245181291279, + "grad_norm": 2.570584535598755, + "learning_rate": 0.00017295109694569438, + "loss": 0.3293, + "step": 10235 + }, + { + "epoch": 0.2399479597039329, + "grad_norm": 2.8950986862182617, + "learning_rate": 0.00017294605852335459, + "loss": 0.3641, + "step": 10236 + }, + { + "epoch": 0.2399714012787379, + "grad_norm": 5.211170673370361, + "learning_rate": 0.00017294101970520743, + "loss": 0.7764, + "step": 10237 + }, + { + "epoch": 0.2399948428535429, + "grad_norm": 2.6641647815704346, + "learning_rate": 0.00017293598049128025, + "loss": 0.2592, + "step": 10238 + }, + { + "epoch": 0.24001828442834788, + "grad_norm": 2.187476873397827, + "learning_rate": 0.0001729309408816004, + "loss": 0.2473, + "step": 10239 + }, + { + "epoch": 0.2400417260031529, + "grad_norm": 4.449288368225098, + "learning_rate": 0.0001729259008761952, + "loss": 0.2878, + "step": 10240 + }, + { + "epoch": 0.2400651675779579, + "grad_norm": 3.9186184406280518, + "learning_rate": 0.000172920860475092, + "loss": 0.5078, + "step": 10241 + }, + { + "epoch": 0.2400886091527629, + "grad_norm": 1.7898319959640503, + "learning_rate": 0.0001729158196783182, + "loss": 0.0743, + "step": 10242 + }, + { + "epoch": 0.2401120507275679, + "grad_norm": 5.148569107055664, + "learning_rate": 0.00017291077848590106, + "loss": 0.3938, + "step": 10243 + }, + { + "epoch": 0.24013549230237288, + "grad_norm": 4.068948268890381, + "learning_rate": 0.00017290573689786803, + "loss": 1.9372, + "step": 10244 + }, + { + "epoch": 0.24015893387717788, + "grad_norm": 1.4349441528320312, + "learning_rate": 0.0001729006949142464, + "loss": 0.2662, + "step": 10245 + }, + { + "epoch": 0.24018237545198287, + "grad_norm": 3.430161952972412, + "learning_rate": 0.00017289565253506356, + "loss": 0.4551, + "step": 10246 + }, + { + "epoch": 0.24020581702678787, + "grad_norm": 2.921745777130127, + "learning_rate": 0.00017289060976034685, + "loss": 0.3977, + "step": 10247 + }, + { + "epoch": 0.24022925860159286, + "grad_norm": 5.1028218269348145, + "learning_rate": 0.00017288556659012368, + "loss": 0.6772, + "step": 10248 + }, + { + "epoch": 0.24025270017639785, + "grad_norm": 1.1764615774154663, + "learning_rate": 0.0001728805230244213, + "loss": 0.1243, + "step": 10249 + }, + { + "epoch": 0.24027614175120285, + "grad_norm": 0.9800693392753601, + "learning_rate": 0.00017287547906326723, + "loss": 0.0779, + "step": 10250 + }, + { + "epoch": 0.24029958332600784, + "grad_norm": 4.277832984924316, + "learning_rate": 0.0001728704347066887, + "loss": 0.5914, + "step": 10251 + }, + { + "epoch": 0.24032302490081284, + "grad_norm": 2.6566014289855957, + "learning_rate": 0.00017286538995471318, + "loss": 1.6953, + "step": 10252 + }, + { + "epoch": 0.24034646647561783, + "grad_norm": 2.499713182449341, + "learning_rate": 0.00017286034480736798, + "loss": 1.7017, + "step": 10253 + }, + { + "epoch": 0.24036990805042283, + "grad_norm": 2.095945358276367, + "learning_rate": 0.00017285529926468053, + "loss": 1.3191, + "step": 10254 + }, + { + "epoch": 0.24039334962522782, + "grad_norm": 3.334779977798462, + "learning_rate": 0.00017285025332667816, + "loss": 0.228, + "step": 10255 + }, + { + "epoch": 0.24041679120003281, + "grad_norm": 3.291370391845703, + "learning_rate": 0.00017284520699338825, + "loss": 0.4805, + "step": 10256 + }, + { + "epoch": 0.2404402327748378, + "grad_norm": 1.6724494695663452, + "learning_rate": 0.00017284016026483822, + "loss": 0.2617, + "step": 10257 + }, + { + "epoch": 0.2404636743496428, + "grad_norm": 2.4084534645080566, + "learning_rate": 0.00017283511314105545, + "loss": 1.8646, + "step": 10258 + }, + { + "epoch": 0.2404871159244478, + "grad_norm": 0.8494754433631897, + "learning_rate": 0.0001728300656220673, + "loss": 0.0467, + "step": 10259 + }, + { + "epoch": 0.2405105574992528, + "grad_norm": 3.754269599914551, + "learning_rate": 0.00017282501770790116, + "loss": 0.4245, + "step": 10260 + }, + { + "epoch": 0.24053399907405779, + "grad_norm": 3.0612661838531494, + "learning_rate": 0.0001728199693985844, + "loss": 0.3805, + "step": 10261 + }, + { + "epoch": 0.24055744064886278, + "grad_norm": 3.346980094909668, + "learning_rate": 0.00017281492069414446, + "loss": 0.4188, + "step": 10262 + }, + { + "epoch": 0.24058088222366777, + "grad_norm": 3.3325157165527344, + "learning_rate": 0.00017280987159460872, + "loss": 0.316, + "step": 10263 + }, + { + "epoch": 0.24060432379847277, + "grad_norm": 4.325028896331787, + "learning_rate": 0.00017280482210000456, + "loss": 0.6279, + "step": 10264 + }, + { + "epoch": 0.24062776537327776, + "grad_norm": 1.3768677711486816, + "learning_rate": 0.0001727997722103594, + "loss": 0.1771, + "step": 10265 + }, + { + "epoch": 0.24065120694808279, + "grad_norm": 3.799011707305908, + "learning_rate": 0.00017279472192570063, + "loss": 0.4457, + "step": 10266 + }, + { + "epoch": 0.24067464852288778, + "grad_norm": 1.9641278982162476, + "learning_rate": 0.00017278967124605563, + "loss": 1.238, + "step": 10267 + }, + { + "epoch": 0.24069809009769277, + "grad_norm": 1.0290260314941406, + "learning_rate": 0.00017278462017145184, + "loss": 0.1975, + "step": 10268 + }, + { + "epoch": 0.24072153167249777, + "grad_norm": 0.5392546057701111, + "learning_rate": 0.00017277956870191666, + "loss": 0.0518, + "step": 10269 + }, + { + "epoch": 0.24074497324730276, + "grad_norm": 3.3548531532287598, + "learning_rate": 0.00017277451683747746, + "loss": 0.4866, + "step": 10270 + }, + { + "epoch": 0.24076841482210776, + "grad_norm": 4.0643792152404785, + "learning_rate": 0.00017276946457816172, + "loss": 0.3532, + "step": 10271 + }, + { + "epoch": 0.24079185639691275, + "grad_norm": 3.6302967071533203, + "learning_rate": 0.00017276441192399684, + "loss": 0.6007, + "step": 10272 + }, + { + "epoch": 0.24081529797171775, + "grad_norm": 3.3809406757354736, + "learning_rate": 0.00017275935887501017, + "loss": 0.3222, + "step": 10273 + }, + { + "epoch": 0.24083873954652274, + "grad_norm": 3.7020697593688965, + "learning_rate": 0.0001727543054312292, + "loss": 0.4185, + "step": 10274 + }, + { + "epoch": 0.24086218112132773, + "grad_norm": 3.92685604095459, + "learning_rate": 0.00017274925159268132, + "loss": 0.5499, + "step": 10275 + }, + { + "epoch": 0.24088562269613273, + "grad_norm": 2.292750120162964, + "learning_rate": 0.00017274419735939392, + "loss": 0.2434, + "step": 10276 + }, + { + "epoch": 0.24090906427093772, + "grad_norm": 3.2807867527008057, + "learning_rate": 0.0001727391427313945, + "loss": 0.2222, + "step": 10277 + }, + { + "epoch": 0.24093250584574272, + "grad_norm": 3.359142541885376, + "learning_rate": 0.00017273408770871043, + "loss": 0.4327, + "step": 10278 + }, + { + "epoch": 0.2409559474205477, + "grad_norm": 3.573967218399048, + "learning_rate": 0.00017272903229136917, + "loss": 0.5162, + "step": 10279 + }, + { + "epoch": 0.2409793889953527, + "grad_norm": 5.628706932067871, + "learning_rate": 0.0001727239764793981, + "loss": 0.5098, + "step": 10280 + }, + { + "epoch": 0.2410028305701577, + "grad_norm": 3.8770480155944824, + "learning_rate": 0.00017271892027282472, + "loss": 0.5811, + "step": 10281 + }, + { + "epoch": 0.2410262721449627, + "grad_norm": 1.6515671014785767, + "learning_rate": 0.00017271386367167642, + "loss": 0.2373, + "step": 10282 + }, + { + "epoch": 0.2410497137197677, + "grad_norm": 3.579789638519287, + "learning_rate": 0.0001727088066759807, + "loss": 0.5043, + "step": 10283 + }, + { + "epoch": 0.24107315529457268, + "grad_norm": 3.8274312019348145, + "learning_rate": 0.00017270374928576488, + "loss": 0.3682, + "step": 10284 + }, + { + "epoch": 0.24109659686937768, + "grad_norm": 0.5269601941108704, + "learning_rate": 0.00017269869150105652, + "loss": 0.0367, + "step": 10285 + }, + { + "epoch": 0.24112003844418267, + "grad_norm": 2.643402338027954, + "learning_rate": 0.00017269363332188297, + "loss": 1.8817, + "step": 10286 + }, + { + "epoch": 0.24114348001898767, + "grad_norm": 0.6676424741744995, + "learning_rate": 0.00017268857474827176, + "loss": 0.0892, + "step": 10287 + }, + { + "epoch": 0.24116692159379266, + "grad_norm": 2.8473756313323975, + "learning_rate": 0.0001726835157802503, + "loss": 0.1576, + "step": 10288 + }, + { + "epoch": 0.24119036316859765, + "grad_norm": 4.562469959259033, + "learning_rate": 0.000172678456417846, + "loss": 0.5504, + "step": 10289 + }, + { + "epoch": 0.24121380474340265, + "grad_norm": 5.834909915924072, + "learning_rate": 0.0001726733966610864, + "loss": 0.4867, + "step": 10290 + }, + { + "epoch": 0.24123724631820767, + "grad_norm": 5.082122802734375, + "learning_rate": 0.00017266833650999888, + "loss": 0.7405, + "step": 10291 + }, + { + "epoch": 0.24126068789301267, + "grad_norm": 4.102801322937012, + "learning_rate": 0.00017266327596461093, + "loss": 0.7918, + "step": 10292 + }, + { + "epoch": 0.24128412946781766, + "grad_norm": 2.9521892070770264, + "learning_rate": 0.00017265821502495, + "loss": 0.341, + "step": 10293 + }, + { + "epoch": 0.24130757104262265, + "grad_norm": 3.0745482444763184, + "learning_rate": 0.00017265315369104355, + "loss": 0.4669, + "step": 10294 + }, + { + "epoch": 0.24133101261742765, + "grad_norm": 0.82533860206604, + "learning_rate": 0.00017264809196291907, + "loss": 0.1074, + "step": 10295 + }, + { + "epoch": 0.24135445419223264, + "grad_norm": 2.8416924476623535, + "learning_rate": 0.00017264302984060397, + "loss": 0.2147, + "step": 10296 + }, + { + "epoch": 0.24137789576703764, + "grad_norm": 2.2591612339019775, + "learning_rate": 0.00017263796732412576, + "loss": 1.6097, + "step": 10297 + }, + { + "epoch": 0.24140133734184263, + "grad_norm": 1.5315184593200684, + "learning_rate": 0.0001726329044135119, + "loss": 0.58, + "step": 10298 + }, + { + "epoch": 0.24142477891664763, + "grad_norm": 0.7699747681617737, + "learning_rate": 0.00017262784110878983, + "loss": 0.1238, + "step": 10299 + }, + { + "epoch": 0.24144822049145262, + "grad_norm": 0.9093198776245117, + "learning_rate": 0.00017262277740998707, + "loss": 0.1052, + "step": 10300 + }, + { + "epoch": 0.24147166206625761, + "grad_norm": 2.892099142074585, + "learning_rate": 0.00017261771331713107, + "loss": 1.8763, + "step": 10301 + }, + { + "epoch": 0.2414951036410626, + "grad_norm": 3.216157913208008, + "learning_rate": 0.00017261264883024934, + "loss": 0.4051, + "step": 10302 + }, + { + "epoch": 0.2415185452158676, + "grad_norm": 2.2501654624938965, + "learning_rate": 0.00017260758394936932, + "loss": 1.1102, + "step": 10303 + }, + { + "epoch": 0.2415419867906726, + "grad_norm": 1.1480932235717773, + "learning_rate": 0.0001726025186745185, + "loss": 0.1398, + "step": 10304 + }, + { + "epoch": 0.2415654283654776, + "grad_norm": 3.3728625774383545, + "learning_rate": 0.0001725974530057244, + "loss": 0.4826, + "step": 10305 + }, + { + "epoch": 0.24158886994028259, + "grad_norm": 1.882275938987732, + "learning_rate": 0.00017259238694301447, + "loss": 0.1337, + "step": 10306 + }, + { + "epoch": 0.24161231151508758, + "grad_norm": 1.2195683717727661, + "learning_rate": 0.0001725873204864162, + "loss": 0.1364, + "step": 10307 + }, + { + "epoch": 0.24163575308989257, + "grad_norm": 0.7169654965400696, + "learning_rate": 0.00017258225363595708, + "loss": 0.055, + "step": 10308 + }, + { + "epoch": 0.24165919466469757, + "grad_norm": 1.5429613590240479, + "learning_rate": 0.00017257718639166462, + "loss": 1.1726, + "step": 10309 + }, + { + "epoch": 0.24168263623950256, + "grad_norm": 3.542684555053711, + "learning_rate": 0.00017257211875356632, + "loss": 0.4155, + "step": 10310 + }, + { + "epoch": 0.24170607781430756, + "grad_norm": 15.670870780944824, + "learning_rate": 0.00017256705072168964, + "loss": 0.7331, + "step": 10311 + }, + { + "epoch": 0.24172951938911255, + "grad_norm": 3.194561243057251, + "learning_rate": 0.00017256198229606212, + "loss": 0.3262, + "step": 10312 + }, + { + "epoch": 0.24175296096391755, + "grad_norm": 2.9404706954956055, + "learning_rate": 0.00017255691347671122, + "loss": 0.3587, + "step": 10313 + }, + { + "epoch": 0.24177640253872254, + "grad_norm": 3.7417099475860596, + "learning_rate": 0.0001725518442636645, + "loss": 0.5285, + "step": 10314 + }, + { + "epoch": 0.24179984411352753, + "grad_norm": 4.459891319274902, + "learning_rate": 0.00017254677465694938, + "loss": 0.5372, + "step": 10315 + }, + { + "epoch": 0.24182328568833253, + "grad_norm": 2.4900693893432617, + "learning_rate": 0.00017254170465659345, + "loss": 0.1744, + "step": 10316 + }, + { + "epoch": 0.24184672726313755, + "grad_norm": 0.6971591114997864, + "learning_rate": 0.00017253663426262424, + "loss": 0.0936, + "step": 10317 + }, + { + "epoch": 0.24187016883794255, + "grad_norm": 2.8082363605499268, + "learning_rate": 0.00017253156347506914, + "loss": 0.3142, + "step": 10318 + }, + { + "epoch": 0.24189361041274754, + "grad_norm": 2.781240701675415, + "learning_rate": 0.00017252649229395576, + "loss": 0.4691, + "step": 10319 + }, + { + "epoch": 0.24191705198755253, + "grad_norm": 2.058694839477539, + "learning_rate": 0.0001725214207193116, + "loss": 0.1229, + "step": 10320 + }, + { + "epoch": 0.24194049356235753, + "grad_norm": 3.367305278778076, + "learning_rate": 0.00017251634875116418, + "loss": 0.4797, + "step": 10321 + }, + { + "epoch": 0.24196393513716252, + "grad_norm": 3.5535526275634766, + "learning_rate": 0.00017251127638954098, + "loss": 0.3913, + "step": 10322 + }, + { + "epoch": 0.24198737671196752, + "grad_norm": 4.307450294494629, + "learning_rate": 0.00017250620363446962, + "loss": 0.6298, + "step": 10323 + }, + { + "epoch": 0.2420108182867725, + "grad_norm": 3.873098611831665, + "learning_rate": 0.00017250113048597753, + "loss": 1.7942, + "step": 10324 + }, + { + "epoch": 0.2420342598615775, + "grad_norm": 4.273306846618652, + "learning_rate": 0.00017249605694409222, + "loss": 0.5242, + "step": 10325 + }, + { + "epoch": 0.2420577014363825, + "grad_norm": 4.366900444030762, + "learning_rate": 0.0001724909830088413, + "loss": 0.7432, + "step": 10326 + }, + { + "epoch": 0.2420811430111875, + "grad_norm": 0.6931254863739014, + "learning_rate": 0.0001724859086802523, + "loss": 0.0484, + "step": 10327 + }, + { + "epoch": 0.2421045845859925, + "grad_norm": 2.3261239528656006, + "learning_rate": 0.00017248083395835265, + "loss": 1.8969, + "step": 10328 + }, + { + "epoch": 0.24212802616079748, + "grad_norm": 5.0721755027771, + "learning_rate": 0.00017247575884317003, + "loss": 0.5237, + "step": 10329 + }, + { + "epoch": 0.24215146773560248, + "grad_norm": 3.8068339824676514, + "learning_rate": 0.00017247068333473185, + "loss": 0.3239, + "step": 10330 + }, + { + "epoch": 0.24217490931040747, + "grad_norm": 0.8974172472953796, + "learning_rate": 0.00017246560743306574, + "loss": 0.1102, + "step": 10331 + }, + { + "epoch": 0.24219835088521247, + "grad_norm": 4.169158935546875, + "learning_rate": 0.00017246053113819922, + "loss": 0.4122, + "step": 10332 + }, + { + "epoch": 0.24222179246001746, + "grad_norm": 5.147360801696777, + "learning_rate": 0.00017245545445015978, + "loss": 0.2713, + "step": 10333 + }, + { + "epoch": 0.24224523403482245, + "grad_norm": 0.27422791719436646, + "learning_rate": 0.00017245037736897503, + "loss": 0.0261, + "step": 10334 + }, + { + "epoch": 0.24226867560962745, + "grad_norm": 3.251164197921753, + "learning_rate": 0.00017244529989467248, + "loss": 0.3385, + "step": 10335 + }, + { + "epoch": 0.24229211718443244, + "grad_norm": 1.0247259140014648, + "learning_rate": 0.0001724402220272797, + "loss": 0.0986, + "step": 10336 + }, + { + "epoch": 0.24231555875923744, + "grad_norm": 0.9981327652931213, + "learning_rate": 0.00017243514376682426, + "loss": 0.0642, + "step": 10337 + }, + { + "epoch": 0.24233900033404243, + "grad_norm": 2.1100502014160156, + "learning_rate": 0.00017243006511333367, + "loss": 0.2225, + "step": 10338 + }, + { + "epoch": 0.24236244190884743, + "grad_norm": 5.158194541931152, + "learning_rate": 0.0001724249860668355, + "loss": 0.5788, + "step": 10339 + }, + { + "epoch": 0.24238588348365242, + "grad_norm": 9.379340171813965, + "learning_rate": 0.00017241990662735735, + "loss": 0.3662, + "step": 10340 + }, + { + "epoch": 0.24240932505845741, + "grad_norm": 1.4040632247924805, + "learning_rate": 0.00017241482679492673, + "loss": 0.2272, + "step": 10341 + }, + { + "epoch": 0.2424327666332624, + "grad_norm": 4.212676048278809, + "learning_rate": 0.00017240974656957124, + "loss": 0.6112, + "step": 10342 + }, + { + "epoch": 0.24245620820806743, + "grad_norm": 1.3886080980300903, + "learning_rate": 0.00017240466595131842, + "loss": 0.1479, + "step": 10343 + }, + { + "epoch": 0.24247964978287243, + "grad_norm": 6.070070743560791, + "learning_rate": 0.00017239958494019586, + "loss": 0.5772, + "step": 10344 + }, + { + "epoch": 0.24250309135767742, + "grad_norm": 1.1971179246902466, + "learning_rate": 0.0001723945035362311, + "loss": 0.0897, + "step": 10345 + }, + { + "epoch": 0.2425265329324824, + "grad_norm": 3.709554672241211, + "learning_rate": 0.00017238942173945172, + "loss": 0.328, + "step": 10346 + }, + { + "epoch": 0.2425499745072874, + "grad_norm": 2.059701442718506, + "learning_rate": 0.00017238433954988531, + "loss": 1.4687, + "step": 10347 + }, + { + "epoch": 0.2425734160820924, + "grad_norm": 2.0242843627929688, + "learning_rate": 0.00017237925696755944, + "loss": 0.2029, + "step": 10348 + }, + { + "epoch": 0.2425968576568974, + "grad_norm": 3.0004279613494873, + "learning_rate": 0.00017237417399250168, + "loss": 0.3671, + "step": 10349 + }, + { + "epoch": 0.2426202992317024, + "grad_norm": 2.353278398513794, + "learning_rate": 0.00017236909062473964, + "loss": 1.8485, + "step": 10350 + }, + { + "epoch": 0.24264374080650739, + "grad_norm": 3.5377726554870605, + "learning_rate": 0.00017236400686430084, + "loss": 0.4143, + "step": 10351 + }, + { + "epoch": 0.24266718238131238, + "grad_norm": 0.48000001907348633, + "learning_rate": 0.00017235892271121292, + "loss": 0.0616, + "step": 10352 + }, + { + "epoch": 0.24269062395611737, + "grad_norm": 4.7433648109436035, + "learning_rate": 0.00017235383816550348, + "loss": 0.6304, + "step": 10353 + }, + { + "epoch": 0.24271406553092237, + "grad_norm": 1.4999781847000122, + "learning_rate": 0.00017234875322720005, + "loss": 0.2182, + "step": 10354 + }, + { + "epoch": 0.24273750710572736, + "grad_norm": 2.311464786529541, + "learning_rate": 0.00017234366789633024, + "loss": 1.1193, + "step": 10355 + }, + { + "epoch": 0.24276094868053236, + "grad_norm": 3.787825584411621, + "learning_rate": 0.0001723385821729217, + "loss": 0.3567, + "step": 10356 + }, + { + "epoch": 0.24278439025533735, + "grad_norm": 2.2460570335388184, + "learning_rate": 0.00017233349605700194, + "loss": 0.4507, + "step": 10357 + }, + { + "epoch": 0.24280783183014235, + "grad_norm": 3.803023099899292, + "learning_rate": 0.00017232840954859861, + "loss": 0.5271, + "step": 10358 + }, + { + "epoch": 0.24283127340494734, + "grad_norm": 5.106382369995117, + "learning_rate": 0.0001723233226477393, + "loss": 0.6896, + "step": 10359 + }, + { + "epoch": 0.24285471497975233, + "grad_norm": 3.9180612564086914, + "learning_rate": 0.00017231823535445158, + "loss": 0.355, + "step": 10360 + }, + { + "epoch": 0.24287815655455733, + "grad_norm": 3.271493434906006, + "learning_rate": 0.00017231314766876312, + "loss": 0.5129, + "step": 10361 + }, + { + "epoch": 0.24290159812936232, + "grad_norm": 5.610190391540527, + "learning_rate": 0.0001723080595907015, + "loss": 0.9717, + "step": 10362 + }, + { + "epoch": 0.24292503970416732, + "grad_norm": 3.9811336994171143, + "learning_rate": 0.00017230297112029428, + "loss": 0.5959, + "step": 10363 + }, + { + "epoch": 0.2429484812789723, + "grad_norm": 2.8528549671173096, + "learning_rate": 0.00017229788225756908, + "loss": 0.2899, + "step": 10364 + }, + { + "epoch": 0.2429719228537773, + "grad_norm": 0.6789891123771667, + "learning_rate": 0.0001722927930025536, + "loss": 0.0708, + "step": 10365 + }, + { + "epoch": 0.2429953644285823, + "grad_norm": 2.935027599334717, + "learning_rate": 0.00017228770335527536, + "loss": 0.4012, + "step": 10366 + }, + { + "epoch": 0.2430188060033873, + "grad_norm": 2.119697332382202, + "learning_rate": 0.000172282613315762, + "loss": 0.1571, + "step": 10367 + }, + { + "epoch": 0.24304224757819232, + "grad_norm": 6.365975379943848, + "learning_rate": 0.00017227752288404116, + "loss": 0.6123, + "step": 10368 + }, + { + "epoch": 0.2430656891529973, + "grad_norm": 4.318436622619629, + "learning_rate": 0.00017227243206014047, + "loss": 0.6132, + "step": 10369 + }, + { + "epoch": 0.2430891307278023, + "grad_norm": 3.3115618228912354, + "learning_rate": 0.0001722673408440875, + "loss": 0.3194, + "step": 10370 + }, + { + "epoch": 0.2431125723026073, + "grad_norm": 2.952946901321411, + "learning_rate": 0.00017226224923590992, + "loss": 0.538, + "step": 10371 + }, + { + "epoch": 0.2431360138774123, + "grad_norm": 2.2324483394622803, + "learning_rate": 0.00017225715723563533, + "loss": 0.2986, + "step": 10372 + }, + { + "epoch": 0.2431594554522173, + "grad_norm": 2.374114751815796, + "learning_rate": 0.00017225206484329137, + "loss": 0.4275, + "step": 10373 + }, + { + "epoch": 0.24318289702702228, + "grad_norm": 3.241374969482422, + "learning_rate": 0.0001722469720589057, + "loss": 0.4557, + "step": 10374 + }, + { + "epoch": 0.24320633860182728, + "grad_norm": 3.9688963890075684, + "learning_rate": 0.00017224187888250592, + "loss": 0.3434, + "step": 10375 + }, + { + "epoch": 0.24322978017663227, + "grad_norm": 2.6563966274261475, + "learning_rate": 0.00017223678531411965, + "loss": 0.6195, + "step": 10376 + }, + { + "epoch": 0.24325322175143727, + "grad_norm": 5.3492326736450195, + "learning_rate": 0.00017223169135377456, + "loss": 0.5005, + "step": 10377 + }, + { + "epoch": 0.24327666332624226, + "grad_norm": 3.638124704360962, + "learning_rate": 0.00017222659700149833, + "loss": 0.5633, + "step": 10378 + }, + { + "epoch": 0.24330010490104725, + "grad_norm": 3.14570689201355, + "learning_rate": 0.0001722215022573185, + "loss": 0.3755, + "step": 10379 + }, + { + "epoch": 0.24332354647585225, + "grad_norm": 2.804182291030884, + "learning_rate": 0.0001722164071212628, + "loss": 0.2412, + "step": 10380 + }, + { + "epoch": 0.24334698805065724, + "grad_norm": 2.8769567012786865, + "learning_rate": 0.00017221131159335883, + "loss": 0.2561, + "step": 10381 + }, + { + "epoch": 0.24337042962546224, + "grad_norm": 1.5181621313095093, + "learning_rate": 0.00017220621567363425, + "loss": 0.7267, + "step": 10382 + }, + { + "epoch": 0.24339387120026723, + "grad_norm": 3.2943646907806396, + "learning_rate": 0.0001722011193621167, + "loss": 0.3232, + "step": 10383 + }, + { + "epoch": 0.24341731277507223, + "grad_norm": 3.2224438190460205, + "learning_rate": 0.00017219602265883388, + "loss": 0.3046, + "step": 10384 + }, + { + "epoch": 0.24344075434987722, + "grad_norm": 4.16360330581665, + "learning_rate": 0.0001721909255638134, + "loss": 0.5394, + "step": 10385 + }, + { + "epoch": 0.24346419592468221, + "grad_norm": 2.3971290588378906, + "learning_rate": 0.0001721858280770829, + "loss": 1.0892, + "step": 10386 + }, + { + "epoch": 0.2434876374994872, + "grad_norm": 1.4591717720031738, + "learning_rate": 0.0001721807301986701, + "loss": 0.2586, + "step": 10387 + }, + { + "epoch": 0.2435110790742922, + "grad_norm": 2.646986246109009, + "learning_rate": 0.00017217563192860262, + "loss": 0.2211, + "step": 10388 + }, + { + "epoch": 0.2435345206490972, + "grad_norm": 6.308166027069092, + "learning_rate": 0.00017217053326690812, + "loss": 0.637, + "step": 10389 + }, + { + "epoch": 0.2435579622239022, + "grad_norm": 3.9037065505981445, + "learning_rate": 0.0001721654342136143, + "loss": 0.4801, + "step": 10390 + }, + { + "epoch": 0.24358140379870719, + "grad_norm": 0.31266459822654724, + "learning_rate": 0.00017216033476874878, + "loss": 0.0217, + "step": 10391 + }, + { + "epoch": 0.24360484537351218, + "grad_norm": 0.818505585193634, + "learning_rate": 0.00017215523493233924, + "loss": 0.1087, + "step": 10392 + }, + { + "epoch": 0.24362828694831717, + "grad_norm": 3.752897262573242, + "learning_rate": 0.00017215013470441343, + "loss": 0.6796, + "step": 10393 + }, + { + "epoch": 0.2436517285231222, + "grad_norm": 2.7391064167022705, + "learning_rate": 0.00017214503408499891, + "loss": 0.3109, + "step": 10394 + }, + { + "epoch": 0.2436751700979272, + "grad_norm": 4.800084114074707, + "learning_rate": 0.00017213993307412345, + "loss": 0.6068, + "step": 10395 + }, + { + "epoch": 0.24369861167273218, + "grad_norm": 5.350166320800781, + "learning_rate": 0.00017213483167181464, + "loss": 0.5116, + "step": 10396 + }, + { + "epoch": 0.24372205324753718, + "grad_norm": 3.575467824935913, + "learning_rate": 0.0001721297298781002, + "loss": 0.6044, + "step": 10397 + }, + { + "epoch": 0.24374549482234217, + "grad_norm": 3.775006055831909, + "learning_rate": 0.00017212462769300782, + "loss": 0.3907, + "step": 10398 + }, + { + "epoch": 0.24376893639714717, + "grad_norm": 2.7508621215820312, + "learning_rate": 0.0001721195251165652, + "loss": 0.3342, + "step": 10399 + }, + { + "epoch": 0.24379237797195216, + "grad_norm": 1.0080580711364746, + "learning_rate": 0.00017211442214880003, + "loss": 0.1109, + "step": 10400 + }, + { + "epoch": 0.24381581954675716, + "grad_norm": 3.7323617935180664, + "learning_rate": 0.00017210931878973998, + "loss": 0.5696, + "step": 10401 + }, + { + "epoch": 0.24383926112156215, + "grad_norm": 4.790065765380859, + "learning_rate": 0.00017210421503941267, + "loss": 0.561, + "step": 10402 + }, + { + "epoch": 0.24386270269636715, + "grad_norm": 1.526132583618164, + "learning_rate": 0.00017209911089784594, + "loss": 0.1212, + "step": 10403 + }, + { + "epoch": 0.24388614427117214, + "grad_norm": 2.2842884063720703, + "learning_rate": 0.00017209400636506735, + "loss": 0.2316, + "step": 10404 + }, + { + "epoch": 0.24390958584597713, + "grad_norm": 4.17081356048584, + "learning_rate": 0.0001720889014411047, + "loss": 0.3989, + "step": 10405 + }, + { + "epoch": 0.24393302742078213, + "grad_norm": 3.940253973007202, + "learning_rate": 0.00017208379612598562, + "loss": 0.4821, + "step": 10406 + }, + { + "epoch": 0.24395646899558712, + "grad_norm": 4.253598213195801, + "learning_rate": 0.00017207869041973783, + "loss": 0.4956, + "step": 10407 + }, + { + "epoch": 0.24397991057039212, + "grad_norm": 2.952440023422241, + "learning_rate": 0.00017207358432238902, + "loss": 0.3546, + "step": 10408 + }, + { + "epoch": 0.2440033521451971, + "grad_norm": 3.987302541732788, + "learning_rate": 0.00017206847783396697, + "loss": 0.5405, + "step": 10409 + }, + { + "epoch": 0.2440267937200021, + "grad_norm": 1.482561469078064, + "learning_rate": 0.0001720633709544993, + "loss": 0.1744, + "step": 10410 + }, + { + "epoch": 0.2440502352948071, + "grad_norm": 3.3099722862243652, + "learning_rate": 0.00017205826368401375, + "loss": 0.3408, + "step": 10411 + }, + { + "epoch": 0.2440736768696121, + "grad_norm": 1.679532766342163, + "learning_rate": 0.00017205315602253804, + "loss": 0.1047, + "step": 10412 + }, + { + "epoch": 0.2440971184444171, + "grad_norm": 1.499985933303833, + "learning_rate": 0.00017204804797009987, + "loss": 0.1702, + "step": 10413 + }, + { + "epoch": 0.24412056001922208, + "grad_norm": 3.928321361541748, + "learning_rate": 0.000172042939526727, + "loss": 0.4393, + "step": 10414 + }, + { + "epoch": 0.24414400159402708, + "grad_norm": 3.8636250495910645, + "learning_rate": 0.0001720378306924471, + "loss": 0.4245, + "step": 10415 + }, + { + "epoch": 0.24416744316883207, + "grad_norm": 3.1953229904174805, + "learning_rate": 0.00017203272146728787, + "loss": 0.3859, + "step": 10416 + }, + { + "epoch": 0.24419088474363707, + "grad_norm": 2.0928502082824707, + "learning_rate": 0.00017202761185127708, + "loss": 0.1333, + "step": 10417 + }, + { + "epoch": 0.24421432631844206, + "grad_norm": 2.516705274581909, + "learning_rate": 0.0001720225018444425, + "loss": 0.1129, + "step": 10418 + }, + { + "epoch": 0.24423776789324708, + "grad_norm": 5.62601375579834, + "learning_rate": 0.00017201739144681174, + "loss": 0.5886, + "step": 10419 + }, + { + "epoch": 0.24426120946805208, + "grad_norm": 3.6763510704040527, + "learning_rate": 0.0001720122806584126, + "loss": 0.362, + "step": 10420 + }, + { + "epoch": 0.24428465104285707, + "grad_norm": 2.5856149196624756, + "learning_rate": 0.00017200716947927283, + "loss": 1.0927, + "step": 10421 + }, + { + "epoch": 0.24430809261766206, + "grad_norm": 3.2309885025024414, + "learning_rate": 0.0001720020579094201, + "loss": 0.3616, + "step": 10422 + }, + { + "epoch": 0.24433153419246706, + "grad_norm": 5.321370601654053, + "learning_rate": 0.00017199694594888218, + "loss": 0.7307, + "step": 10423 + }, + { + "epoch": 0.24435497576727205, + "grad_norm": 5.217658519744873, + "learning_rate": 0.0001719918335976868, + "loss": 0.6393, + "step": 10424 + }, + { + "epoch": 0.24437841734207705, + "grad_norm": 2.2840936183929443, + "learning_rate": 0.00017198672085586174, + "loss": 0.1053, + "step": 10425 + }, + { + "epoch": 0.24440185891688204, + "grad_norm": 5.098169803619385, + "learning_rate": 0.00017198160772343468, + "loss": 0.4465, + "step": 10426 + }, + { + "epoch": 0.24442530049168704, + "grad_norm": 3.6092398166656494, + "learning_rate": 0.00017197649420043343, + "loss": 0.4391, + "step": 10427 + }, + { + "epoch": 0.24444874206649203, + "grad_norm": 1.3985059261322021, + "learning_rate": 0.00017197138028688565, + "loss": 0.1361, + "step": 10428 + }, + { + "epoch": 0.24447218364129703, + "grad_norm": 5.4274373054504395, + "learning_rate": 0.00017196626598281917, + "loss": 0.3258, + "step": 10429 + }, + { + "epoch": 0.24449562521610202, + "grad_norm": 1.3062217235565186, + "learning_rate": 0.0001719611512882617, + "loss": 0.1641, + "step": 10430 + }, + { + "epoch": 0.244519066790907, + "grad_norm": 3.775522232055664, + "learning_rate": 0.000171956036203241, + "loss": 0.4071, + "step": 10431 + }, + { + "epoch": 0.244542508365712, + "grad_norm": 2.806967258453369, + "learning_rate": 0.0001719509207277848, + "loss": 0.2177, + "step": 10432 + }, + { + "epoch": 0.244565949940517, + "grad_norm": 1.2285270690917969, + "learning_rate": 0.0001719458048619209, + "loss": 0.1987, + "step": 10433 + }, + { + "epoch": 0.244589391515322, + "grad_norm": 2.8279635906219482, + "learning_rate": 0.00017194068860567705, + "loss": 0.2753, + "step": 10434 + }, + { + "epoch": 0.244612833090127, + "grad_norm": 5.97285795211792, + "learning_rate": 0.000171935571959081, + "loss": 0.6699, + "step": 10435 + }, + { + "epoch": 0.24463627466493199, + "grad_norm": 4.043333053588867, + "learning_rate": 0.0001719304549221605, + "loss": 0.5221, + "step": 10436 + }, + { + "epoch": 0.24465971623973698, + "grad_norm": 3.5374186038970947, + "learning_rate": 0.00017192533749494336, + "loss": 0.4082, + "step": 10437 + }, + { + "epoch": 0.24468315781454197, + "grad_norm": 3.176089286804199, + "learning_rate": 0.0001719202196774573, + "loss": 1.4776, + "step": 10438 + }, + { + "epoch": 0.24470659938934697, + "grad_norm": 0.9842707514762878, + "learning_rate": 0.00017191510146973008, + "loss": 0.1312, + "step": 10439 + }, + { + "epoch": 0.24473004096415196, + "grad_norm": 1.1685994863510132, + "learning_rate": 0.0001719099828717895, + "loss": 0.155, + "step": 10440 + }, + { + "epoch": 0.24475348253895696, + "grad_norm": 2.0555105209350586, + "learning_rate": 0.00017190486388366335, + "loss": 0.1905, + "step": 10441 + }, + { + "epoch": 0.24477692411376195, + "grad_norm": 2.708756685256958, + "learning_rate": 0.00017189974450537937, + "loss": 0.3008, + "step": 10442 + }, + { + "epoch": 0.24480036568856695, + "grad_norm": 5.199579238891602, + "learning_rate": 0.00017189462473696538, + "loss": 0.5553, + "step": 10443 + }, + { + "epoch": 0.24482380726337194, + "grad_norm": 1.638135313987732, + "learning_rate": 0.0001718895045784491, + "loss": 0.1472, + "step": 10444 + }, + { + "epoch": 0.24484724883817696, + "grad_norm": 3.4845340251922607, + "learning_rate": 0.0001718843840298584, + "loss": 0.3089, + "step": 10445 + }, + { + "epoch": 0.24487069041298196, + "grad_norm": 4.5113701820373535, + "learning_rate": 0.00017187926309122098, + "loss": 0.4896, + "step": 10446 + }, + { + "epoch": 0.24489413198778695, + "grad_norm": 3.287579298019409, + "learning_rate": 0.00017187414176256466, + "loss": 0.3238, + "step": 10447 + }, + { + "epoch": 0.24491757356259194, + "grad_norm": 4.127485275268555, + "learning_rate": 0.00017186902004391721, + "loss": 0.3489, + "step": 10448 + }, + { + "epoch": 0.24494101513739694, + "grad_norm": 4.474411964416504, + "learning_rate": 0.00017186389793530645, + "loss": 0.479, + "step": 10449 + }, + { + "epoch": 0.24496445671220193, + "grad_norm": 0.9448495507240295, + "learning_rate": 0.00017185877543676018, + "loss": 0.0746, + "step": 10450 + }, + { + "epoch": 0.24498789828700693, + "grad_norm": 3.6901769638061523, + "learning_rate": 0.00017185365254830613, + "loss": 0.4864, + "step": 10451 + }, + { + "epoch": 0.24501133986181192, + "grad_norm": 2.9969637393951416, + "learning_rate": 0.00017184852926997217, + "loss": 0.2892, + "step": 10452 + }, + { + "epoch": 0.24503478143661692, + "grad_norm": 4.293153285980225, + "learning_rate": 0.00017184340560178607, + "loss": 0.5495, + "step": 10453 + }, + { + "epoch": 0.2450582230114219, + "grad_norm": 3.2806622982025146, + "learning_rate": 0.0001718382815437756, + "loss": 0.3869, + "step": 10454 + }, + { + "epoch": 0.2450816645862269, + "grad_norm": 4.318261623382568, + "learning_rate": 0.00017183315709596865, + "loss": 0.7848, + "step": 10455 + }, + { + "epoch": 0.2451051061610319, + "grad_norm": 5.149486064910889, + "learning_rate": 0.00017182803225839293, + "loss": 0.5547, + "step": 10456 + }, + { + "epoch": 0.2451285477358369, + "grad_norm": 1.4958014488220215, + "learning_rate": 0.0001718229070310763, + "loss": 0.2286, + "step": 10457 + }, + { + "epoch": 0.2451519893106419, + "grad_norm": 3.3022453784942627, + "learning_rate": 0.00017181778141404656, + "loss": 1.4784, + "step": 10458 + }, + { + "epoch": 0.24517543088544688, + "grad_norm": 4.036571979522705, + "learning_rate": 0.00017181265540733152, + "loss": 0.4412, + "step": 10459 + }, + { + "epoch": 0.24519887246025188, + "grad_norm": 4.310460090637207, + "learning_rate": 0.00017180752901095898, + "loss": 0.4599, + "step": 10460 + }, + { + "epoch": 0.24522231403505687, + "grad_norm": 2.31145977973938, + "learning_rate": 0.00017180240222495677, + "loss": 1.7179, + "step": 10461 + }, + { + "epoch": 0.24524575560986187, + "grad_norm": 4.245484352111816, + "learning_rate": 0.00017179727504935271, + "loss": 0.2841, + "step": 10462 + }, + { + "epoch": 0.24526919718466686, + "grad_norm": 3.5028345584869385, + "learning_rate": 0.0001717921474841746, + "loss": 0.5384, + "step": 10463 + }, + { + "epoch": 0.24529263875947185, + "grad_norm": 2.708036184310913, + "learning_rate": 0.0001717870195294503, + "loss": 0.2513, + "step": 10464 + }, + { + "epoch": 0.24531608033427685, + "grad_norm": 4.162304401397705, + "learning_rate": 0.0001717818911852076, + "loss": 0.2953, + "step": 10465 + }, + { + "epoch": 0.24533952190908184, + "grad_norm": 0.8992190957069397, + "learning_rate": 0.00017177676245147438, + "loss": 0.0569, + "step": 10466 + }, + { + "epoch": 0.24536296348388684, + "grad_norm": 1.9412925243377686, + "learning_rate": 0.0001717716333282784, + "loss": 1.6096, + "step": 10467 + }, + { + "epoch": 0.24538640505869183, + "grad_norm": 0.9003173112869263, + "learning_rate": 0.00017176650381564752, + "loss": 0.1079, + "step": 10468 + }, + { + "epoch": 0.24540984663349683, + "grad_norm": 3.570984125137329, + "learning_rate": 0.00017176137391360956, + "loss": 0.567, + "step": 10469 + }, + { + "epoch": 0.24543328820830185, + "grad_norm": 3.310093879699707, + "learning_rate": 0.0001717562436221924, + "loss": 0.4205, + "step": 10470 + }, + { + "epoch": 0.24545672978310684, + "grad_norm": 1.3894767761230469, + "learning_rate": 0.0001717511129414238, + "loss": 0.1839, + "step": 10471 + }, + { + "epoch": 0.24548017135791184, + "grad_norm": 2.6133270263671875, + "learning_rate": 0.00017174598187133167, + "loss": 1.5932, + "step": 10472 + }, + { + "epoch": 0.24550361293271683, + "grad_norm": 2.994788408279419, + "learning_rate": 0.0001717408504119438, + "loss": 0.3089, + "step": 10473 + }, + { + "epoch": 0.24552705450752182, + "grad_norm": 4.502529144287109, + "learning_rate": 0.0001717357185632881, + "loss": 0.4756, + "step": 10474 + }, + { + "epoch": 0.24555049608232682, + "grad_norm": 2.0500354766845703, + "learning_rate": 0.00017173058632539236, + "loss": 0.2061, + "step": 10475 + }, + { + "epoch": 0.2455739376571318, + "grad_norm": 4.547725677490234, + "learning_rate": 0.0001717254536982844, + "loss": 0.6182, + "step": 10476 + }, + { + "epoch": 0.2455973792319368, + "grad_norm": 4.582825660705566, + "learning_rate": 0.00017172032068199218, + "loss": 0.3401, + "step": 10477 + }, + { + "epoch": 0.2456208208067418, + "grad_norm": 3.665342330932617, + "learning_rate": 0.00017171518727654347, + "loss": 0.3874, + "step": 10478 + }, + { + "epoch": 0.2456442623815468, + "grad_norm": 3.828535556793213, + "learning_rate": 0.00017171005348196608, + "loss": 0.3715, + "step": 10479 + }, + { + "epoch": 0.2456677039563518, + "grad_norm": 3.959672451019287, + "learning_rate": 0.00017170491929828797, + "loss": 0.4211, + "step": 10480 + }, + { + "epoch": 0.24569114553115678, + "grad_norm": 3.20959734916687, + "learning_rate": 0.00017169978472553695, + "loss": 0.2739, + "step": 10481 + }, + { + "epoch": 0.24571458710596178, + "grad_norm": 4.2556023597717285, + "learning_rate": 0.00017169464976374087, + "loss": 0.5676, + "step": 10482 + }, + { + "epoch": 0.24573802868076677, + "grad_norm": 5.524584770202637, + "learning_rate": 0.0001716895144129276, + "loss": 0.5217, + "step": 10483 + }, + { + "epoch": 0.24576147025557177, + "grad_norm": 2.3967013359069824, + "learning_rate": 0.00017168437867312505, + "loss": 0.1768, + "step": 10484 + }, + { + "epoch": 0.24578491183037676, + "grad_norm": 0.7836630940437317, + "learning_rate": 0.000171679242544361, + "loss": 0.03, + "step": 10485 + }, + { + "epoch": 0.24580835340518176, + "grad_norm": 4.969437122344971, + "learning_rate": 0.00017167410602666336, + "loss": 0.4795, + "step": 10486 + }, + { + "epoch": 0.24583179497998675, + "grad_norm": 3.071107864379883, + "learning_rate": 0.00017166896912006006, + "loss": 0.3419, + "step": 10487 + }, + { + "epoch": 0.24585523655479175, + "grad_norm": 0.7835255861282349, + "learning_rate": 0.00017166383182457885, + "loss": 0.1273, + "step": 10488 + }, + { + "epoch": 0.24587867812959674, + "grad_norm": 4.136954307556152, + "learning_rate": 0.0001716586941402477, + "loss": 0.5295, + "step": 10489 + }, + { + "epoch": 0.24590211970440173, + "grad_norm": 5.496549606323242, + "learning_rate": 0.0001716535560670945, + "loss": 0.6913, + "step": 10490 + }, + { + "epoch": 0.24592556127920673, + "grad_norm": 3.5438196659088135, + "learning_rate": 0.00017164841760514702, + "loss": 0.3926, + "step": 10491 + }, + { + "epoch": 0.24594900285401172, + "grad_norm": 2.7116429805755615, + "learning_rate": 0.00017164327875443327, + "loss": 2.4484, + "step": 10492 + }, + { + "epoch": 0.24597244442881672, + "grad_norm": 3.6637611389160156, + "learning_rate": 0.00017163813951498106, + "loss": 0.4945, + "step": 10493 + }, + { + "epoch": 0.2459958860036217, + "grad_norm": 11.600799560546875, + "learning_rate": 0.0001716329998868183, + "loss": 0.3467, + "step": 10494 + }, + { + "epoch": 0.2460193275784267, + "grad_norm": 4.859008312225342, + "learning_rate": 0.00017162785986997282, + "loss": 0.554, + "step": 10495 + }, + { + "epoch": 0.24604276915323173, + "grad_norm": 2.2812132835388184, + "learning_rate": 0.0001716227194644726, + "loss": 1.328, + "step": 10496 + }, + { + "epoch": 0.24606621072803672, + "grad_norm": 2.7324960231781006, + "learning_rate": 0.00017161757867034547, + "loss": 1.2865, + "step": 10497 + }, + { + "epoch": 0.24608965230284172, + "grad_norm": 5.5076189041137695, + "learning_rate": 0.00017161243748761937, + "loss": 0.7003, + "step": 10498 + }, + { + "epoch": 0.2461130938776467, + "grad_norm": 2.5836191177368164, + "learning_rate": 0.00017160729591632215, + "loss": 0.165, + "step": 10499 + }, + { + "epoch": 0.2461365354524517, + "grad_norm": 4.53024959564209, + "learning_rate": 0.00017160215395648173, + "loss": 0.5386, + "step": 10500 + }, + { + "epoch": 0.2461599770272567, + "grad_norm": 3.4440808296203613, + "learning_rate": 0.00017159701160812603, + "loss": 1.7021, + "step": 10501 + }, + { + "epoch": 0.2461834186020617, + "grad_norm": 3.633204221725464, + "learning_rate": 0.0001715918688712829, + "loss": 0.6327, + "step": 10502 + }, + { + "epoch": 0.2462068601768667, + "grad_norm": 2.1238646507263184, + "learning_rate": 0.00017158672574598031, + "loss": 0.1997, + "step": 10503 + }, + { + "epoch": 0.24623030175167168, + "grad_norm": 2.651806592941284, + "learning_rate": 0.0001715815822322461, + "loss": 0.2557, + "step": 10504 + }, + { + "epoch": 0.24625374332647668, + "grad_norm": 1.587121605873108, + "learning_rate": 0.00017157643833010824, + "loss": 0.2368, + "step": 10505 + }, + { + "epoch": 0.24627718490128167, + "grad_norm": 3.3312861919403076, + "learning_rate": 0.0001715712940395946, + "loss": 0.2597, + "step": 10506 + }, + { + "epoch": 0.24630062647608666, + "grad_norm": 0.7532564997673035, + "learning_rate": 0.0001715661493607331, + "loss": 0.0437, + "step": 10507 + }, + { + "epoch": 0.24632406805089166, + "grad_norm": 2.8409953117370605, + "learning_rate": 0.00017156100429355169, + "loss": 0.3236, + "step": 10508 + }, + { + "epoch": 0.24634750962569665, + "grad_norm": 1.2005068063735962, + "learning_rate": 0.00017155585883807823, + "loss": 0.1922, + "step": 10509 + }, + { + "epoch": 0.24637095120050165, + "grad_norm": 3.4725825786590576, + "learning_rate": 0.00017155071299434067, + "loss": 0.3418, + "step": 10510 + }, + { + "epoch": 0.24639439277530664, + "grad_norm": 24.70591926574707, + "learning_rate": 0.00017154556676236694, + "loss": 0.5777, + "step": 10511 + }, + { + "epoch": 0.24641783435011164, + "grad_norm": 2.7458536624908447, + "learning_rate": 0.00017154042014218493, + "loss": 0.3509, + "step": 10512 + }, + { + "epoch": 0.24644127592491663, + "grad_norm": 0.999567449092865, + "learning_rate": 0.00017153527313382264, + "loss": 0.0508, + "step": 10513 + }, + { + "epoch": 0.24646471749972163, + "grad_norm": 4.547652721405029, + "learning_rate": 0.0001715301257373079, + "loss": 0.5007, + "step": 10514 + }, + { + "epoch": 0.24648815907452662, + "grad_norm": 1.3619980812072754, + "learning_rate": 0.00017152497795266868, + "loss": 0.8603, + "step": 10515 + }, + { + "epoch": 0.2465116006493316, + "grad_norm": 1.7210279703140259, + "learning_rate": 0.00017151982977993294, + "loss": 0.2259, + "step": 10516 + }, + { + "epoch": 0.2465350422241366, + "grad_norm": 3.2496156692504883, + "learning_rate": 0.0001715146812191286, + "loss": 0.4332, + "step": 10517 + }, + { + "epoch": 0.2465584837989416, + "grad_norm": 3.7136120796203613, + "learning_rate": 0.0001715095322702836, + "loss": 0.2789, + "step": 10518 + }, + { + "epoch": 0.2465819253737466, + "grad_norm": 0.8449889421463013, + "learning_rate": 0.00017150438293342583, + "loss": 0.5624, + "step": 10519 + }, + { + "epoch": 0.2466053669485516, + "grad_norm": 1.4792163372039795, + "learning_rate": 0.00017149923320858328, + "loss": 0.2401, + "step": 10520 + }, + { + "epoch": 0.2466288085233566, + "grad_norm": 4.402800559997559, + "learning_rate": 0.0001714940830957839, + "loss": 0.3583, + "step": 10521 + }, + { + "epoch": 0.2466522500981616, + "grad_norm": 2.7724640369415283, + "learning_rate": 0.0001714889325950556, + "loss": 0.2507, + "step": 10522 + }, + { + "epoch": 0.2466756916729666, + "grad_norm": 4.330069541931152, + "learning_rate": 0.0001714837817064263, + "loss": 0.4793, + "step": 10523 + }, + { + "epoch": 0.2466991332477716, + "grad_norm": 3.039907693862915, + "learning_rate": 0.00017147863042992406, + "loss": 0.3655, + "step": 10524 + }, + { + "epoch": 0.2467225748225766, + "grad_norm": 6.902236461639404, + "learning_rate": 0.00017147347876557675, + "loss": 0.4256, + "step": 10525 + }, + { + "epoch": 0.24674601639738158, + "grad_norm": 2.9454853534698486, + "learning_rate": 0.0001714683267134123, + "loss": 2.1241, + "step": 10526 + }, + { + "epoch": 0.24676945797218658, + "grad_norm": 4.0161638259887695, + "learning_rate": 0.00017146317427345874, + "loss": 0.4218, + "step": 10527 + }, + { + "epoch": 0.24679289954699157, + "grad_norm": 3.3442635536193848, + "learning_rate": 0.00017145802144574395, + "loss": 0.4321, + "step": 10528 + }, + { + "epoch": 0.24681634112179657, + "grad_norm": 4.071963310241699, + "learning_rate": 0.00017145286823029592, + "loss": 0.549, + "step": 10529 + }, + { + "epoch": 0.24683978269660156, + "grad_norm": 3.8910880088806152, + "learning_rate": 0.00017144771462714267, + "loss": 0.5113, + "step": 10530 + }, + { + "epoch": 0.24686322427140656, + "grad_norm": 4.002236843109131, + "learning_rate": 0.00017144256063631206, + "loss": 0.4858, + "step": 10531 + }, + { + "epoch": 0.24688666584621155, + "grad_norm": 5.033992767333984, + "learning_rate": 0.00017143740625783212, + "loss": 0.7357, + "step": 10532 + }, + { + "epoch": 0.24691010742101654, + "grad_norm": 5.509345054626465, + "learning_rate": 0.00017143225149173082, + "loss": 0.6915, + "step": 10533 + }, + { + "epoch": 0.24693354899582154, + "grad_norm": 3.0738606452941895, + "learning_rate": 0.00017142709633803614, + "loss": 0.335, + "step": 10534 + }, + { + "epoch": 0.24695699057062653, + "grad_norm": 3.3417723178863525, + "learning_rate": 0.00017142194079677596, + "loss": 0.3364, + "step": 10535 + }, + { + "epoch": 0.24698043214543153, + "grad_norm": 2.271965980529785, + "learning_rate": 0.00017141678486797836, + "loss": 0.188, + "step": 10536 + }, + { + "epoch": 0.24700387372023652, + "grad_norm": 1.0633209943771362, + "learning_rate": 0.00017141162855167127, + "loss": 0.1493, + "step": 10537 + }, + { + "epoch": 0.24702731529504152, + "grad_norm": 1.1947948932647705, + "learning_rate": 0.0001714064718478827, + "loss": 0.1216, + "step": 10538 + }, + { + "epoch": 0.2470507568698465, + "grad_norm": 1.7517319917678833, + "learning_rate": 0.00017140131475664057, + "loss": 0.1933, + "step": 10539 + }, + { + "epoch": 0.2470741984446515, + "grad_norm": 5.092103481292725, + "learning_rate": 0.00017139615727797294, + "loss": 0.5422, + "step": 10540 + }, + { + "epoch": 0.2470976400194565, + "grad_norm": 2.067392110824585, + "learning_rate": 0.0001713909994119077, + "loss": 0.244, + "step": 10541 + }, + { + "epoch": 0.2471210815942615, + "grad_norm": 3.447909355163574, + "learning_rate": 0.00017138584115847293, + "loss": 0.474, + "step": 10542 + }, + { + "epoch": 0.2471445231690665, + "grad_norm": 3.224863052368164, + "learning_rate": 0.00017138068251769655, + "loss": 0.4081, + "step": 10543 + }, + { + "epoch": 0.24716796474387148, + "grad_norm": 1.9571921825408936, + "learning_rate": 0.00017137552348960662, + "loss": 0.2053, + "step": 10544 + }, + { + "epoch": 0.24719140631867648, + "grad_norm": 2.4121203422546387, + "learning_rate": 0.00017137036407423106, + "loss": 0.2579, + "step": 10545 + }, + { + "epoch": 0.24721484789348147, + "grad_norm": 3.6198666095733643, + "learning_rate": 0.0001713652042715979, + "loss": 0.4654, + "step": 10546 + }, + { + "epoch": 0.2472382894682865, + "grad_norm": 7.957564830780029, + "learning_rate": 0.00017136004408173513, + "loss": 0.7023, + "step": 10547 + }, + { + "epoch": 0.2472617310430915, + "grad_norm": 5.678893566131592, + "learning_rate": 0.00017135488350467076, + "loss": 0.5636, + "step": 10548 + }, + { + "epoch": 0.24728517261789648, + "grad_norm": 2.525055170059204, + "learning_rate": 0.0001713497225404328, + "loss": 0.233, + "step": 10549 + }, + { + "epoch": 0.24730861419270148, + "grad_norm": 1.32651948928833, + "learning_rate": 0.00017134456118904923, + "loss": 0.0876, + "step": 10550 + }, + { + "epoch": 0.24733205576750647, + "grad_norm": 2.1413753032684326, + "learning_rate": 0.00017133939945054806, + "loss": 0.2625, + "step": 10551 + }, + { + "epoch": 0.24735549734231146, + "grad_norm": 1.1430517435073853, + "learning_rate": 0.0001713342373249573, + "loss": 0.1505, + "step": 10552 + }, + { + "epoch": 0.24737893891711646, + "grad_norm": 3.254343032836914, + "learning_rate": 0.000171329074812305, + "loss": 0.409, + "step": 10553 + }, + { + "epoch": 0.24740238049192145, + "grad_norm": 3.4017505645751953, + "learning_rate": 0.0001713239119126191, + "loss": 0.4522, + "step": 10554 + }, + { + "epoch": 0.24742582206672645, + "grad_norm": 3.8615024089813232, + "learning_rate": 0.00017131874862592766, + "loss": 0.6509, + "step": 10555 + }, + { + "epoch": 0.24744926364153144, + "grad_norm": 3.7779252529144287, + "learning_rate": 0.00017131358495225865, + "loss": 0.3624, + "step": 10556 + }, + { + "epoch": 0.24747270521633644, + "grad_norm": 0.9492151141166687, + "learning_rate": 0.00017130842089164015, + "loss": 0.0987, + "step": 10557 + }, + { + "epoch": 0.24749614679114143, + "grad_norm": 3.652393341064453, + "learning_rate": 0.00017130325644410018, + "loss": 0.5799, + "step": 10558 + }, + { + "epoch": 0.24751958836594642, + "grad_norm": 3.015437126159668, + "learning_rate": 0.00017129809160966667, + "loss": 0.3157, + "step": 10559 + }, + { + "epoch": 0.24754302994075142, + "grad_norm": 2.5544273853302, + "learning_rate": 0.00017129292638836776, + "loss": 0.3639, + "step": 10560 + }, + { + "epoch": 0.2475664715155564, + "grad_norm": 6.495603084564209, + "learning_rate": 0.0001712877607802314, + "loss": 0.6109, + "step": 10561 + }, + { + "epoch": 0.2475899130903614, + "grad_norm": 4.279823303222656, + "learning_rate": 0.00017128259478528567, + "loss": 0.4924, + "step": 10562 + }, + { + "epoch": 0.2476133546651664, + "grad_norm": 2.8157942295074463, + "learning_rate": 0.0001712774284035586, + "loss": 0.3641, + "step": 10563 + }, + { + "epoch": 0.2476367962399714, + "grad_norm": 0.6187936067581177, + "learning_rate": 0.00017127226163507814, + "loss": 0.056, + "step": 10564 + }, + { + "epoch": 0.2476602378147764, + "grad_norm": 3.400256395339966, + "learning_rate": 0.00017126709447987242, + "loss": 0.3983, + "step": 10565 + }, + { + "epoch": 0.24768367938958138, + "grad_norm": 4.746751308441162, + "learning_rate": 0.00017126192693796944, + "loss": 0.5135, + "step": 10566 + }, + { + "epoch": 0.24770712096438638, + "grad_norm": 3.302039861679077, + "learning_rate": 0.00017125675900939725, + "loss": 0.4308, + "step": 10567 + }, + { + "epoch": 0.24773056253919137, + "grad_norm": 3.448542833328247, + "learning_rate": 0.00017125159069418384, + "loss": 0.5457, + "step": 10568 + }, + { + "epoch": 0.24775400411399637, + "grad_norm": 5.644287109375, + "learning_rate": 0.00017124642199235732, + "loss": 0.6094, + "step": 10569 + }, + { + "epoch": 0.24777744568880136, + "grad_norm": 2.9851884841918945, + "learning_rate": 0.00017124125290394573, + "loss": 0.3424, + "step": 10570 + }, + { + "epoch": 0.24780088726360636, + "grad_norm": 0.9398232698440552, + "learning_rate": 0.00017123608342897706, + "loss": 0.0836, + "step": 10571 + }, + { + "epoch": 0.24782432883841135, + "grad_norm": 3.8579673767089844, + "learning_rate": 0.00017123091356747942, + "loss": 0.3366, + "step": 10572 + }, + { + "epoch": 0.24784777041321637, + "grad_norm": 4.669535160064697, + "learning_rate": 0.00017122574331948084, + "loss": 0.5485, + "step": 10573 + }, + { + "epoch": 0.24787121198802137, + "grad_norm": 4.548976898193359, + "learning_rate": 0.00017122057268500936, + "loss": 0.6547, + "step": 10574 + }, + { + "epoch": 0.24789465356282636, + "grad_norm": 2.808269500732422, + "learning_rate": 0.00017121540166409308, + "loss": 0.2941, + "step": 10575 + }, + { + "epoch": 0.24791809513763136, + "grad_norm": 2.731926918029785, + "learning_rate": 0.00017121023025676, + "loss": 0.2243, + "step": 10576 + }, + { + "epoch": 0.24794153671243635, + "grad_norm": 3.5647592544555664, + "learning_rate": 0.00017120505846303824, + "loss": 0.3929, + "step": 10577 + }, + { + "epoch": 0.24796497828724134, + "grad_norm": 3.917583465576172, + "learning_rate": 0.0001711998862829558, + "loss": 2.3465, + "step": 10578 + }, + { + "epoch": 0.24798841986204634, + "grad_norm": 3.1390087604522705, + "learning_rate": 0.00017119471371654078, + "loss": 0.2976, + "step": 10579 + }, + { + "epoch": 0.24801186143685133, + "grad_norm": 3.287426710128784, + "learning_rate": 0.00017118954076382126, + "loss": 0.2919, + "step": 10580 + }, + { + "epoch": 0.24803530301165633, + "grad_norm": 1.0448198318481445, + "learning_rate": 0.0001711843674248253, + "loss": 0.131, + "step": 10581 + }, + { + "epoch": 0.24805874458646132, + "grad_norm": 2.8084418773651123, + "learning_rate": 0.00017117919369958093, + "loss": 0.1854, + "step": 10582 + }, + { + "epoch": 0.24808218616126632, + "grad_norm": 4.7960286140441895, + "learning_rate": 0.00017117401958811624, + "loss": 0.3952, + "step": 10583 + }, + { + "epoch": 0.2481056277360713, + "grad_norm": 3.3388373851776123, + "learning_rate": 0.00017116884509045933, + "loss": 0.3264, + "step": 10584 + }, + { + "epoch": 0.2481290693108763, + "grad_norm": 3.785815477371216, + "learning_rate": 0.0001711636702066383, + "loss": 0.309, + "step": 10585 + }, + { + "epoch": 0.2481525108856813, + "grad_norm": 4.302329063415527, + "learning_rate": 0.00017115849493668116, + "loss": 0.6254, + "step": 10586 + }, + { + "epoch": 0.2481759524604863, + "grad_norm": 4.663466930389404, + "learning_rate": 0.00017115331928061603, + "loss": 0.5554, + "step": 10587 + }, + { + "epoch": 0.2481993940352913, + "grad_norm": 2.703779935836792, + "learning_rate": 0.000171148143238471, + "loss": 0.4323, + "step": 10588 + }, + { + "epoch": 0.24822283561009628, + "grad_norm": 3.0907487869262695, + "learning_rate": 0.00017114296681027414, + "loss": 0.3667, + "step": 10589 + }, + { + "epoch": 0.24824627718490128, + "grad_norm": 4.532028675079346, + "learning_rate": 0.00017113778999605354, + "loss": 0.5894, + "step": 10590 + }, + { + "epoch": 0.24826971875970627, + "grad_norm": 1.6028547286987305, + "learning_rate": 0.0001711326127958373, + "loss": 0.1557, + "step": 10591 + }, + { + "epoch": 0.24829316033451126, + "grad_norm": 2.0401411056518555, + "learning_rate": 0.0001711274352096535, + "loss": 0.8655, + "step": 10592 + }, + { + "epoch": 0.24831660190931626, + "grad_norm": 3.1864888668060303, + "learning_rate": 0.00017112225723753022, + "loss": 0.4891, + "step": 10593 + }, + { + "epoch": 0.24834004348412125, + "grad_norm": 4.194793224334717, + "learning_rate": 0.00017111707887949558, + "loss": 0.4687, + "step": 10594 + }, + { + "epoch": 0.24836348505892625, + "grad_norm": 4.223206520080566, + "learning_rate": 0.0001711119001355777, + "loss": 0.4413, + "step": 10595 + }, + { + "epoch": 0.24838692663373124, + "grad_norm": 3.944131374359131, + "learning_rate": 0.0001711067210058046, + "loss": 0.4793, + "step": 10596 + }, + { + "epoch": 0.24841036820853624, + "grad_norm": 1.1551555395126343, + "learning_rate": 0.00017110154149020446, + "loss": 0.1568, + "step": 10597 + }, + { + "epoch": 0.24843380978334126, + "grad_norm": 3.8288967609405518, + "learning_rate": 0.00017109636158880534, + "loss": 0.4803, + "step": 10598 + }, + { + "epoch": 0.24845725135814625, + "grad_norm": 3.885451555252075, + "learning_rate": 0.00017109118130163539, + "loss": 0.5365, + "step": 10599 + }, + { + "epoch": 0.24848069293295125, + "grad_norm": 3.0675647258758545, + "learning_rate": 0.0001710860006287227, + "loss": 0.406, + "step": 10600 + }, + { + "epoch": 0.24850413450775624, + "grad_norm": 3.4904489517211914, + "learning_rate": 0.00017108081957009535, + "loss": 0.5105, + "step": 10601 + }, + { + "epoch": 0.24852757608256124, + "grad_norm": 5.6445159912109375, + "learning_rate": 0.00017107563812578146, + "loss": 0.8438, + "step": 10602 + }, + { + "epoch": 0.24855101765736623, + "grad_norm": 3.5645790100097656, + "learning_rate": 0.0001710704562958092, + "loss": 0.2351, + "step": 10603 + }, + { + "epoch": 0.24857445923217122, + "grad_norm": 1.1031913757324219, + "learning_rate": 0.0001710652740802066, + "loss": 0.1491, + "step": 10604 + }, + { + "epoch": 0.24859790080697622, + "grad_norm": 6.618751525878906, + "learning_rate": 0.00017106009147900184, + "loss": 0.6, + "step": 10605 + }, + { + "epoch": 0.2486213423817812, + "grad_norm": 6.349356174468994, + "learning_rate": 0.00017105490849222304, + "loss": 0.6652, + "step": 10606 + }, + { + "epoch": 0.2486447839565862, + "grad_norm": 5.496310234069824, + "learning_rate": 0.00017104972511989828, + "loss": 0.6156, + "step": 10607 + }, + { + "epoch": 0.2486682255313912, + "grad_norm": 0.7226913571357727, + "learning_rate": 0.00017104454136205574, + "loss": 0.1291, + "step": 10608 + }, + { + "epoch": 0.2486916671061962, + "grad_norm": 4.394158363342285, + "learning_rate": 0.00017103935721872354, + "loss": 0.5544, + "step": 10609 + }, + { + "epoch": 0.2487151086810012, + "grad_norm": 5.4171648025512695, + "learning_rate": 0.00017103417268992974, + "loss": 0.8426, + "step": 10610 + }, + { + "epoch": 0.24873855025580618, + "grad_norm": 3.494245767593384, + "learning_rate": 0.00017102898777570253, + "loss": 0.3912, + "step": 10611 + }, + { + "epoch": 0.24876199183061118, + "grad_norm": 5.697723388671875, + "learning_rate": 0.00017102380247607007, + "loss": 0.5775, + "step": 10612 + }, + { + "epoch": 0.24878543340541617, + "grad_norm": 6.290359973907471, + "learning_rate": 0.00017101861679106044, + "loss": 0.6154, + "step": 10613 + }, + { + "epoch": 0.24880887498022117, + "grad_norm": 3.381927967071533, + "learning_rate": 0.0001710134307207018, + "loss": 0.3073, + "step": 10614 + }, + { + "epoch": 0.24883231655502616, + "grad_norm": 2.393418312072754, + "learning_rate": 0.0001710082442650223, + "loss": 0.2243, + "step": 10615 + }, + { + "epoch": 0.24885575812983116, + "grad_norm": 5.543331623077393, + "learning_rate": 0.00017100305742405003, + "loss": 0.3621, + "step": 10616 + }, + { + "epoch": 0.24887919970463615, + "grad_norm": 1.4657996892929077, + "learning_rate": 0.0001709978701978132, + "loss": 0.1036, + "step": 10617 + }, + { + "epoch": 0.24890264127944114, + "grad_norm": 3.0921807289123535, + "learning_rate": 0.00017099268258633993, + "loss": 0.3019, + "step": 10618 + }, + { + "epoch": 0.24892608285424614, + "grad_norm": 3.948413372039795, + "learning_rate": 0.0001709874945896584, + "loss": 0.4594, + "step": 10619 + }, + { + "epoch": 0.24894952442905113, + "grad_norm": 1.9081754684448242, + "learning_rate": 0.0001709823062077967, + "loss": 0.2274, + "step": 10620 + }, + { + "epoch": 0.24897296600385613, + "grad_norm": 7.151295185089111, + "learning_rate": 0.000170977117440783, + "loss": 0.6478, + "step": 10621 + }, + { + "epoch": 0.24899640757866112, + "grad_norm": 2.6719281673431396, + "learning_rate": 0.00017097192828864546, + "loss": 1.5854, + "step": 10622 + }, + { + "epoch": 0.24901984915346612, + "grad_norm": 4.75155782699585, + "learning_rate": 0.00017096673875141227, + "loss": 0.4677, + "step": 10623 + }, + { + "epoch": 0.24904329072827114, + "grad_norm": 1.4494426250457764, + "learning_rate": 0.00017096154882911152, + "loss": 0.1243, + "step": 10624 + }, + { + "epoch": 0.24906673230307613, + "grad_norm": 3.3608815670013428, + "learning_rate": 0.00017095635852177146, + "loss": 0.3424, + "step": 10625 + }, + { + "epoch": 0.24909017387788113, + "grad_norm": 8.521550178527832, + "learning_rate": 0.00017095116782942017, + "loss": 0.6851, + "step": 10626 + }, + { + "epoch": 0.24911361545268612, + "grad_norm": 0.645046055316925, + "learning_rate": 0.00017094597675208585, + "loss": 0.0648, + "step": 10627 + }, + { + "epoch": 0.24913705702749112, + "grad_norm": 3.7757887840270996, + "learning_rate": 0.0001709407852897967, + "loss": 0.3216, + "step": 10628 + }, + { + "epoch": 0.2491604986022961, + "grad_norm": 2.66923189163208, + "learning_rate": 0.00017093559344258082, + "loss": 0.2932, + "step": 10629 + }, + { + "epoch": 0.2491839401771011, + "grad_norm": 3.3181238174438477, + "learning_rate": 0.00017093040121046643, + "loss": 0.4484, + "step": 10630 + }, + { + "epoch": 0.2492073817519061, + "grad_norm": 2.7071008682250977, + "learning_rate": 0.00017092520859348167, + "loss": 1.7888, + "step": 10631 + }, + { + "epoch": 0.2492308233267111, + "grad_norm": 2.9168317317962646, + "learning_rate": 0.00017092001559165473, + "loss": 0.2703, + "step": 10632 + }, + { + "epoch": 0.2492542649015161, + "grad_norm": 5.5388569831848145, + "learning_rate": 0.00017091482220501383, + "loss": 0.511, + "step": 10633 + }, + { + "epoch": 0.24927770647632108, + "grad_norm": 3.8715567588806152, + "learning_rate": 0.00017090962843358708, + "loss": 0.4156, + "step": 10634 + }, + { + "epoch": 0.24930114805112608, + "grad_norm": 0.9934820532798767, + "learning_rate": 0.0001709044342774027, + "loss": 0.1053, + "step": 10635 + }, + { + "epoch": 0.24932458962593107, + "grad_norm": 4.026389122009277, + "learning_rate": 0.00017089923973648887, + "loss": 0.4918, + "step": 10636 + }, + { + "epoch": 0.24934803120073606, + "grad_norm": 2.6660842895507812, + "learning_rate": 0.00017089404481087376, + "loss": 0.2603, + "step": 10637 + }, + { + "epoch": 0.24937147277554106, + "grad_norm": 0.8831287026405334, + "learning_rate": 0.00017088884950058558, + "loss": 0.1313, + "step": 10638 + }, + { + "epoch": 0.24939491435034605, + "grad_norm": 2.9272725582122803, + "learning_rate": 0.0001708836538056525, + "loss": 0.3923, + "step": 10639 + }, + { + "epoch": 0.24941835592515105, + "grad_norm": 5.2243876457214355, + "learning_rate": 0.00017087845772610275, + "loss": 0.3343, + "step": 10640 + }, + { + "epoch": 0.24944179749995604, + "grad_norm": 1.5353842973709106, + "learning_rate": 0.00017087326126196447, + "loss": 0.1523, + "step": 10641 + }, + { + "epoch": 0.24946523907476104, + "grad_norm": 2.5010199546813965, + "learning_rate": 0.0001708680644132659, + "loss": 1.1115, + "step": 10642 + }, + { + "epoch": 0.24948868064956603, + "grad_norm": 3.0726523399353027, + "learning_rate": 0.0001708628671800352, + "loss": 0.3806, + "step": 10643 + }, + { + "epoch": 0.24951212222437102, + "grad_norm": 2.6556570529937744, + "learning_rate": 0.00017085766956230058, + "loss": 1.4059, + "step": 10644 + }, + { + "epoch": 0.24953556379917602, + "grad_norm": 3.1843953132629395, + "learning_rate": 0.00017085247156009028, + "loss": 0.2296, + "step": 10645 + }, + { + "epoch": 0.249559005373981, + "grad_norm": 2.862370014190674, + "learning_rate": 0.00017084727317343246, + "loss": 0.3105, + "step": 10646 + }, + { + "epoch": 0.249582446948786, + "grad_norm": 2.5660927295684814, + "learning_rate": 0.00017084207440235533, + "loss": 0.2597, + "step": 10647 + }, + { + "epoch": 0.249605888523591, + "grad_norm": 4.605475425720215, + "learning_rate": 0.00017083687524688715, + "loss": 0.2522, + "step": 10648 + }, + { + "epoch": 0.24962933009839602, + "grad_norm": 3.9763705730438232, + "learning_rate": 0.00017083167570705607, + "loss": 0.5041, + "step": 10649 + }, + { + "epoch": 0.24965277167320102, + "grad_norm": 1.1020382642745972, + "learning_rate": 0.00017082647578289033, + "loss": 0.1042, + "step": 10650 + }, + { + "epoch": 0.249676213248006, + "grad_norm": 2.8219237327575684, + "learning_rate": 0.00017082127547441816, + "loss": 0.2197, + "step": 10651 + }, + { + "epoch": 0.249699654822811, + "grad_norm": 2.327345371246338, + "learning_rate": 0.00017081607478166771, + "loss": 0.1609, + "step": 10652 + }, + { + "epoch": 0.249723096397616, + "grad_norm": 4.448132514953613, + "learning_rate": 0.0001708108737046673, + "loss": 0.4808, + "step": 10653 + }, + { + "epoch": 0.249746537972421, + "grad_norm": 3.607783794403076, + "learning_rate": 0.00017080567224344505, + "loss": 0.5489, + "step": 10654 + }, + { + "epoch": 0.249769979547226, + "grad_norm": 2.9663877487182617, + "learning_rate": 0.00017080047039802926, + "loss": 1.5865, + "step": 10655 + }, + { + "epoch": 0.24979342112203098, + "grad_norm": 8.924843788146973, + "learning_rate": 0.00017079526816844812, + "loss": 0.5378, + "step": 10656 + }, + { + "epoch": 0.24981686269683598, + "grad_norm": 4.445219039916992, + "learning_rate": 0.00017079006555472987, + "loss": 0.569, + "step": 10657 + }, + { + "epoch": 0.24984030427164097, + "grad_norm": 3.9000771045684814, + "learning_rate": 0.00017078486255690274, + "loss": 0.3922, + "step": 10658 + }, + { + "epoch": 0.24986374584644597, + "grad_norm": 2.804028034210205, + "learning_rate": 0.00017077965917499494, + "loss": 0.3184, + "step": 10659 + }, + { + "epoch": 0.24988718742125096, + "grad_norm": 2.3274381160736084, + "learning_rate": 0.00017077445540903473, + "loss": 0.2509, + "step": 10660 + }, + { + "epoch": 0.24991062899605596, + "grad_norm": 3.077070713043213, + "learning_rate": 0.00017076925125905033, + "loss": 0.4971, + "step": 10661 + }, + { + "epoch": 0.24993407057086095, + "grad_norm": 4.696497917175293, + "learning_rate": 0.00017076404672506997, + "loss": 0.5049, + "step": 10662 + }, + { + "epoch": 0.24995751214566594, + "grad_norm": 4.974542617797852, + "learning_rate": 0.00017075884180712194, + "loss": 0.4043, + "step": 10663 + }, + { + "epoch": 0.24998095372047094, + "grad_norm": 2.7851593494415283, + "learning_rate": 0.00017075363650523442, + "loss": 1.436, + "step": 10664 + }, + { + "epoch": 0.25000439529527596, + "grad_norm": 8.774445533752441, + "learning_rate": 0.00017074843081943568, + "loss": 0.4082, + "step": 10665 + }, + { + "epoch": 0.25002783687008096, + "grad_norm": 2.6020193099975586, + "learning_rate": 0.00017074322474975397, + "loss": 0.2493, + "step": 10666 + }, + { + "epoch": 0.25005127844488595, + "grad_norm": 1.0569010972976685, + "learning_rate": 0.00017073801829621752, + "loss": 0.0726, + "step": 10667 + }, + { + "epoch": 0.25007472001969094, + "grad_norm": 1.4790490865707397, + "learning_rate": 0.0001707328114588546, + "loss": 0.1901, + "step": 10668 + }, + { + "epoch": 0.25009816159449594, + "grad_norm": 0.7550191879272461, + "learning_rate": 0.00017072760423769348, + "loss": 0.0948, + "step": 10669 + }, + { + "epoch": 0.25012160316930093, + "grad_norm": 2.5350303649902344, + "learning_rate": 0.00017072239663276236, + "loss": 0.2394, + "step": 10670 + }, + { + "epoch": 0.2501450447441059, + "grad_norm": 5.956517696380615, + "learning_rate": 0.00017071718864408952, + "loss": 0.6766, + "step": 10671 + }, + { + "epoch": 0.2501684863189109, + "grad_norm": 3.4611260890960693, + "learning_rate": 0.00017071198027170326, + "loss": 0.3271, + "step": 10672 + }, + { + "epoch": 0.2501919278937159, + "grad_norm": 4.637082576751709, + "learning_rate": 0.00017070677151563178, + "loss": 0.3579, + "step": 10673 + }, + { + "epoch": 0.2502153694685209, + "grad_norm": 5.158398628234863, + "learning_rate": 0.00017070156237590338, + "loss": 0.5121, + "step": 10674 + }, + { + "epoch": 0.2502388110433259, + "grad_norm": 5.010222434997559, + "learning_rate": 0.0001706963528525463, + "loss": 0.7651, + "step": 10675 + }, + { + "epoch": 0.2502622526181309, + "grad_norm": 4.277024269104004, + "learning_rate": 0.00017069114294558885, + "loss": 0.3266, + "step": 10676 + }, + { + "epoch": 0.2502856941929359, + "grad_norm": 3.2027065753936768, + "learning_rate": 0.00017068593265505926, + "loss": 0.2776, + "step": 10677 + }, + { + "epoch": 0.2503091357677409, + "grad_norm": 3.967597723007202, + "learning_rate": 0.0001706807219809858, + "loss": 0.5572, + "step": 10678 + }, + { + "epoch": 0.2503325773425459, + "grad_norm": 4.441722393035889, + "learning_rate": 0.00017067551092339676, + "loss": 0.5735, + "step": 10679 + }, + { + "epoch": 0.2503560189173509, + "grad_norm": 0.7599077820777893, + "learning_rate": 0.00017067029948232042, + "loss": 0.0812, + "step": 10680 + }, + { + "epoch": 0.25037946049215587, + "grad_norm": 5.73379373550415, + "learning_rate": 0.00017066508765778505, + "loss": 0.4631, + "step": 10681 + }, + { + "epoch": 0.25040290206696086, + "grad_norm": 0.943263590335846, + "learning_rate": 0.0001706598754498189, + "loss": 0.1211, + "step": 10682 + }, + { + "epoch": 0.25042634364176586, + "grad_norm": 3.765394687652588, + "learning_rate": 0.00017065466285845032, + "loss": 0.4489, + "step": 10683 + }, + { + "epoch": 0.25044978521657085, + "grad_norm": 3.640193462371826, + "learning_rate": 0.0001706494498837075, + "loss": 0.4394, + "step": 10684 + }, + { + "epoch": 0.25047322679137585, + "grad_norm": 1.0139360427856445, + "learning_rate": 0.00017064423652561882, + "loss": 0.1321, + "step": 10685 + }, + { + "epoch": 0.25049666836618084, + "grad_norm": 4.691286087036133, + "learning_rate": 0.00017063902278421252, + "loss": 0.7126, + "step": 10686 + }, + { + "epoch": 0.25052010994098584, + "grad_norm": 1.7617881298065186, + "learning_rate": 0.0001706338086595169, + "loss": 0.2749, + "step": 10687 + }, + { + "epoch": 0.25054355151579083, + "grad_norm": 4.088214874267578, + "learning_rate": 0.00017062859415156023, + "loss": 0.5383, + "step": 10688 + }, + { + "epoch": 0.2505669930905958, + "grad_norm": 1.8897905349731445, + "learning_rate": 0.0001706233792603708, + "loss": 0.2128, + "step": 10689 + }, + { + "epoch": 0.2505904346654008, + "grad_norm": 4.394333362579346, + "learning_rate": 0.00017061816398597702, + "loss": 0.2031, + "step": 10690 + }, + { + "epoch": 0.2506138762402058, + "grad_norm": 4.187958240509033, + "learning_rate": 0.00017061294832840698, + "loss": 0.5135, + "step": 10691 + }, + { + "epoch": 0.2506373178150108, + "grad_norm": 2.328199625015259, + "learning_rate": 0.00017060773228768917, + "loss": 0.1363, + "step": 10692 + }, + { + "epoch": 0.2506607593898158, + "grad_norm": 6.735607147216797, + "learning_rate": 0.00017060251586385183, + "loss": 0.8947, + "step": 10693 + }, + { + "epoch": 0.2506842009646208, + "grad_norm": 2.534771203994751, + "learning_rate": 0.00017059729905692322, + "loss": 1.2224, + "step": 10694 + }, + { + "epoch": 0.2507076425394258, + "grad_norm": 0.5250878930091858, + "learning_rate": 0.00017059208186693167, + "loss": 0.0548, + "step": 10695 + }, + { + "epoch": 0.2507310841142308, + "grad_norm": 2.7629051208496094, + "learning_rate": 0.0001705868642939055, + "loss": 0.3303, + "step": 10696 + }, + { + "epoch": 0.2507545256890358, + "grad_norm": 1.5835216045379639, + "learning_rate": 0.00017058164633787306, + "loss": 0.8913, + "step": 10697 + }, + { + "epoch": 0.2507779672638408, + "grad_norm": 3.02032208442688, + "learning_rate": 0.0001705764279988626, + "loss": 0.3221, + "step": 10698 + }, + { + "epoch": 0.25080140883864577, + "grad_norm": 2.065187692642212, + "learning_rate": 0.00017057120927690248, + "loss": 0.2126, + "step": 10699 + }, + { + "epoch": 0.25082485041345076, + "grad_norm": 1.498763918876648, + "learning_rate": 0.00017056599017202095, + "loss": 0.1381, + "step": 10700 + }, + { + "epoch": 0.25084829198825576, + "grad_norm": 3.0397019386291504, + "learning_rate": 0.0001705607706842464, + "loss": 0.4135, + "step": 10701 + }, + { + "epoch": 0.25087173356306075, + "grad_norm": 3.783210277557373, + "learning_rate": 0.0001705555508136071, + "loss": 0.4474, + "step": 10702 + }, + { + "epoch": 0.25089517513786574, + "grad_norm": 0.7285793423652649, + "learning_rate": 0.00017055033056013143, + "loss": 0.125, + "step": 10703 + }, + { + "epoch": 0.25091861671267074, + "grad_norm": 1.0170085430145264, + "learning_rate": 0.00017054510992384767, + "loss": 0.1096, + "step": 10704 + }, + { + "epoch": 0.25094205828747573, + "grad_norm": 1.9792555570602417, + "learning_rate": 0.00017053988890478413, + "loss": 1.1212, + "step": 10705 + }, + { + "epoch": 0.2509654998622807, + "grad_norm": 0.9028052091598511, + "learning_rate": 0.00017053466750296923, + "loss": 0.0874, + "step": 10706 + }, + { + "epoch": 0.2509889414370857, + "grad_norm": 3.3826770782470703, + "learning_rate": 0.00017052944571843118, + "loss": 0.3527, + "step": 10707 + }, + { + "epoch": 0.2510123830118907, + "grad_norm": 5.072949409484863, + "learning_rate": 0.0001705242235511984, + "loss": 0.6245, + "step": 10708 + }, + { + "epoch": 0.2510358245866957, + "grad_norm": 1.7907534837722778, + "learning_rate": 0.00017051900100129922, + "loss": 0.2301, + "step": 10709 + }, + { + "epoch": 0.2510592661615007, + "grad_norm": 4.28651762008667, + "learning_rate": 0.0001705137780687619, + "loss": 0.7485, + "step": 10710 + }, + { + "epoch": 0.2510827077363057, + "grad_norm": 2.916290283203125, + "learning_rate": 0.0001705085547536149, + "loss": 0.1794, + "step": 10711 + }, + { + "epoch": 0.2511061493111107, + "grad_norm": 1.0496412515640259, + "learning_rate": 0.00017050333105588647, + "loss": 0.1124, + "step": 10712 + }, + { + "epoch": 0.25112959088591574, + "grad_norm": 1.79952871799469, + "learning_rate": 0.000170498106975605, + "loss": 0.1539, + "step": 10713 + }, + { + "epoch": 0.25115303246072074, + "grad_norm": 6.095558166503906, + "learning_rate": 0.0001704928825127988, + "loss": 0.8566, + "step": 10714 + }, + { + "epoch": 0.25117647403552573, + "grad_norm": 3.755906105041504, + "learning_rate": 0.00017048765766749626, + "loss": 0.4496, + "step": 10715 + }, + { + "epoch": 0.2511999156103307, + "grad_norm": 1.2824680805206299, + "learning_rate": 0.0001704824324397257, + "loss": 0.1744, + "step": 10716 + }, + { + "epoch": 0.2512233571851357, + "grad_norm": 4.194718360900879, + "learning_rate": 0.00017047720682951547, + "loss": 0.4509, + "step": 10717 + }, + { + "epoch": 0.2512467987599407, + "grad_norm": 3.3930439949035645, + "learning_rate": 0.00017047198083689394, + "loss": 0.4773, + "step": 10718 + }, + { + "epoch": 0.2512702403347457, + "grad_norm": 0.7334219217300415, + "learning_rate": 0.00017046675446188948, + "loss": 0.0811, + "step": 10719 + }, + { + "epoch": 0.2512936819095507, + "grad_norm": 3.001210927963257, + "learning_rate": 0.0001704615277045304, + "loss": 0.5186, + "step": 10720 + }, + { + "epoch": 0.2513171234843557, + "grad_norm": 4.473063945770264, + "learning_rate": 0.0001704563005648451, + "loss": 0.3916, + "step": 10721 + }, + { + "epoch": 0.2513405650591607, + "grad_norm": 3.7479872703552246, + "learning_rate": 0.00017045107304286192, + "loss": 0.3644, + "step": 10722 + }, + { + "epoch": 0.2513640066339657, + "grad_norm": 3.113328218460083, + "learning_rate": 0.00017044584513860924, + "loss": 1.9653, + "step": 10723 + }, + { + "epoch": 0.2513874482087707, + "grad_norm": 3.3819491863250732, + "learning_rate": 0.00017044061685211542, + "loss": 1.6255, + "step": 10724 + }, + { + "epoch": 0.2514108897835757, + "grad_norm": 3.1303300857543945, + "learning_rate": 0.00017043538818340884, + "loss": 0.3753, + "step": 10725 + }, + { + "epoch": 0.25143433135838067, + "grad_norm": 2.035670042037964, + "learning_rate": 0.00017043015913251788, + "loss": 0.2433, + "step": 10726 + }, + { + "epoch": 0.25145777293318566, + "grad_norm": 4.942036151885986, + "learning_rate": 0.00017042492969947088, + "loss": 0.6149, + "step": 10727 + }, + { + "epoch": 0.25148121450799066, + "grad_norm": 2.321070671081543, + "learning_rate": 0.0001704196998842962, + "loss": 0.1969, + "step": 10728 + }, + { + "epoch": 0.25150465608279565, + "grad_norm": 3.332958221435547, + "learning_rate": 0.00017041446968702232, + "loss": 0.3188, + "step": 10729 + }, + { + "epoch": 0.25152809765760065, + "grad_norm": 6.033059597015381, + "learning_rate": 0.00017040923910767748, + "loss": 0.3455, + "step": 10730 + }, + { + "epoch": 0.25155153923240564, + "grad_norm": 5.577470779418945, + "learning_rate": 0.00017040400814629018, + "loss": 1.0605, + "step": 10731 + }, + { + "epoch": 0.25157498080721064, + "grad_norm": 2.13800311088562, + "learning_rate": 0.00017039877680288871, + "loss": 0.3215, + "step": 10732 + }, + { + "epoch": 0.25159842238201563, + "grad_norm": 5.295889377593994, + "learning_rate": 0.00017039354507750155, + "loss": 0.5138, + "step": 10733 + }, + { + "epoch": 0.2516218639568206, + "grad_norm": 1.3443254232406616, + "learning_rate": 0.00017038831297015697, + "loss": 0.2485, + "step": 10734 + }, + { + "epoch": 0.2516453055316256, + "grad_norm": 2.5504062175750732, + "learning_rate": 0.00017038308048088348, + "loss": 1.7717, + "step": 10735 + }, + { + "epoch": 0.2516687471064306, + "grad_norm": 2.488337516784668, + "learning_rate": 0.0001703778476097094, + "loss": 0.334, + "step": 10736 + }, + { + "epoch": 0.2516921886812356, + "grad_norm": 3.0536139011383057, + "learning_rate": 0.0001703726143566631, + "loss": 0.2991, + "step": 10737 + }, + { + "epoch": 0.2517156302560406, + "grad_norm": 2.4417717456817627, + "learning_rate": 0.00017036738072177308, + "loss": 1.4097, + "step": 10738 + }, + { + "epoch": 0.2517390718308456, + "grad_norm": 3.5904667377471924, + "learning_rate": 0.00017036214670506763, + "loss": 0.4489, + "step": 10739 + }, + { + "epoch": 0.2517625134056506, + "grad_norm": 1.7722653150558472, + "learning_rate": 0.0001703569123065752, + "loss": 1.2309, + "step": 10740 + }, + { + "epoch": 0.2517859549804556, + "grad_norm": 5.246716499328613, + "learning_rate": 0.00017035167752632422, + "loss": 0.4858, + "step": 10741 + }, + { + "epoch": 0.2518093965552606, + "grad_norm": 5.619309425354004, + "learning_rate": 0.000170346442364343, + "loss": 0.8942, + "step": 10742 + }, + { + "epoch": 0.2518328381300656, + "grad_norm": 3.10082745552063, + "learning_rate": 0.00017034120682066006, + "loss": 0.29, + "step": 10743 + }, + { + "epoch": 0.25185627970487057, + "grad_norm": 1.2298496961593628, + "learning_rate": 0.00017033597089530372, + "loss": 0.1939, + "step": 10744 + }, + { + "epoch": 0.25187972127967556, + "grad_norm": 5.501379013061523, + "learning_rate": 0.0001703307345883024, + "loss": 0.997, + "step": 10745 + }, + { + "epoch": 0.25190316285448056, + "grad_norm": 2.9713332653045654, + "learning_rate": 0.00017032549789968462, + "loss": 0.3282, + "step": 10746 + }, + { + "epoch": 0.25192660442928555, + "grad_norm": 0.7320088744163513, + "learning_rate": 0.00017032026082947866, + "loss": 0.1228, + "step": 10747 + }, + { + "epoch": 0.25195004600409054, + "grad_norm": 3.717006206512451, + "learning_rate": 0.00017031502337771297, + "loss": 0.2428, + "step": 10748 + }, + { + "epoch": 0.25197348757889554, + "grad_norm": 1.645101547241211, + "learning_rate": 0.000170309785544416, + "loss": 0.2243, + "step": 10749 + }, + { + "epoch": 0.25199692915370053, + "grad_norm": 4.933157920837402, + "learning_rate": 0.00017030454732961617, + "loss": 0.5462, + "step": 10750 + }, + { + "epoch": 0.2520203707285055, + "grad_norm": 1.2771470546722412, + "learning_rate": 0.0001702993087333419, + "loss": 0.1387, + "step": 10751 + }, + { + "epoch": 0.2520438123033105, + "grad_norm": 3.4901037216186523, + "learning_rate": 0.0001702940697556216, + "loss": 0.4653, + "step": 10752 + }, + { + "epoch": 0.2520672538781155, + "grad_norm": 0.8848809003829956, + "learning_rate": 0.00017028883039648366, + "loss": 0.1301, + "step": 10753 + }, + { + "epoch": 0.2520906954529205, + "grad_norm": 2.3483426570892334, + "learning_rate": 0.0001702835906559566, + "loss": 0.1957, + "step": 10754 + }, + { + "epoch": 0.2521141370277255, + "grad_norm": 2.6564440727233887, + "learning_rate": 0.00017027835053406876, + "loss": 0.3293, + "step": 10755 + }, + { + "epoch": 0.2521375786025305, + "grad_norm": 3.5953562259674072, + "learning_rate": 0.00017027311003084864, + "loss": 0.1506, + "step": 10756 + }, + { + "epoch": 0.2521610201773355, + "grad_norm": 4.499357223510742, + "learning_rate": 0.00017026786914632464, + "loss": 0.4833, + "step": 10757 + }, + { + "epoch": 0.2521844617521405, + "grad_norm": 3.366858720779419, + "learning_rate": 0.0001702626278805252, + "loss": 0.3804, + "step": 10758 + }, + { + "epoch": 0.2522079033269455, + "grad_norm": 2.1713674068450928, + "learning_rate": 0.00017025738623347879, + "loss": 1.802, + "step": 10759 + }, + { + "epoch": 0.2522313449017505, + "grad_norm": 3.3024771213531494, + "learning_rate": 0.0001702521442052138, + "loss": 0.5444, + "step": 10760 + }, + { + "epoch": 0.25225478647655547, + "grad_norm": 5.848537921905518, + "learning_rate": 0.0001702469017957587, + "loss": 1.3135, + "step": 10761 + }, + { + "epoch": 0.25227822805136046, + "grad_norm": 4.287049293518066, + "learning_rate": 0.00017024165900514195, + "loss": 0.4775, + "step": 10762 + }, + { + "epoch": 0.25230166962616546, + "grad_norm": 2.3119218349456787, + "learning_rate": 0.00017023641583339199, + "loss": 0.2516, + "step": 10763 + }, + { + "epoch": 0.2523251112009705, + "grad_norm": 2.908111572265625, + "learning_rate": 0.00017023117228053724, + "loss": 0.376, + "step": 10764 + }, + { + "epoch": 0.2523485527757755, + "grad_norm": 3.7010891437530518, + "learning_rate": 0.0001702259283466062, + "loss": 0.3706, + "step": 10765 + }, + { + "epoch": 0.2523719943505805, + "grad_norm": 3.7700841426849365, + "learning_rate": 0.0001702206840316273, + "loss": 0.3707, + "step": 10766 + }, + { + "epoch": 0.2523954359253855, + "grad_norm": 2.4306912422180176, + "learning_rate": 0.00017021543933562895, + "loss": 0.2056, + "step": 10767 + }, + { + "epoch": 0.2524188775001905, + "grad_norm": 2.197645902633667, + "learning_rate": 0.00017021019425863968, + "loss": 0.4283, + "step": 10768 + }, + { + "epoch": 0.2524423190749955, + "grad_norm": 4.428414344787598, + "learning_rate": 0.00017020494880068792, + "loss": 0.4435, + "step": 10769 + }, + { + "epoch": 0.2524657606498005, + "grad_norm": 3.3367700576782227, + "learning_rate": 0.00017019970296180213, + "loss": 0.5155, + "step": 10770 + }, + { + "epoch": 0.25248920222460547, + "grad_norm": 1.4559931755065918, + "learning_rate": 0.0001701944567420108, + "loss": 0.1022, + "step": 10771 + }, + { + "epoch": 0.25251264379941046, + "grad_norm": 3.866422176361084, + "learning_rate": 0.00017018921014134238, + "loss": 1.4803, + "step": 10772 + }, + { + "epoch": 0.25253608537421546, + "grad_norm": 0.9944096207618713, + "learning_rate": 0.0001701839631598253, + "loss": 0.0876, + "step": 10773 + }, + { + "epoch": 0.25255952694902045, + "grad_norm": 3.5275042057037354, + "learning_rate": 0.00017017871579748806, + "loss": 0.3637, + "step": 10774 + }, + { + "epoch": 0.25258296852382545, + "grad_norm": 1.0632047653198242, + "learning_rate": 0.00017017346805435913, + "loss": 0.125, + "step": 10775 + }, + { + "epoch": 0.25260641009863044, + "grad_norm": 2.9472618103027344, + "learning_rate": 0.00017016821993046704, + "loss": 0.3332, + "step": 10776 + }, + { + "epoch": 0.25262985167343543, + "grad_norm": 3.3457834720611572, + "learning_rate": 0.0001701629714258402, + "loss": 0.288, + "step": 10777 + }, + { + "epoch": 0.25265329324824043, + "grad_norm": 4.473034381866455, + "learning_rate": 0.00017015772254050707, + "loss": 0.3257, + "step": 10778 + }, + { + "epoch": 0.2526767348230454, + "grad_norm": 2.6041648387908936, + "learning_rate": 0.00017015247327449617, + "loss": 0.3747, + "step": 10779 + }, + { + "epoch": 0.2527001763978504, + "grad_norm": 3.928475856781006, + "learning_rate": 0.000170147223627836, + "loss": 0.4467, + "step": 10780 + }, + { + "epoch": 0.2527236179726554, + "grad_norm": 1.1515867710113525, + "learning_rate": 0.000170141973600555, + "loss": 0.1516, + "step": 10781 + }, + { + "epoch": 0.2527470595474604, + "grad_norm": 2.4375686645507812, + "learning_rate": 0.00017013672319268167, + "loss": 0.1917, + "step": 10782 + }, + { + "epoch": 0.2527705011222654, + "grad_norm": 3.332443952560425, + "learning_rate": 0.00017013147240424456, + "loss": 0.3648, + "step": 10783 + }, + { + "epoch": 0.2527939426970704, + "grad_norm": 3.8366637229919434, + "learning_rate": 0.00017012622123527206, + "loss": 0.4762, + "step": 10784 + }, + { + "epoch": 0.2528173842718754, + "grad_norm": 1.7736088037490845, + "learning_rate": 0.00017012096968579274, + "loss": 0.216, + "step": 10785 + }, + { + "epoch": 0.2528408258466804, + "grad_norm": 4.961889266967773, + "learning_rate": 0.00017011571775583505, + "loss": 0.5215, + "step": 10786 + }, + { + "epoch": 0.2528642674214854, + "grad_norm": 0.8564236760139465, + "learning_rate": 0.00017011046544542752, + "loss": 0.09, + "step": 10787 + }, + { + "epoch": 0.2528877089962904, + "grad_norm": 6.779266357421875, + "learning_rate": 0.00017010521275459862, + "loss": 0.7558, + "step": 10788 + }, + { + "epoch": 0.25291115057109537, + "grad_norm": 5.914055824279785, + "learning_rate": 0.00017009995968337687, + "loss": 0.4897, + "step": 10789 + }, + { + "epoch": 0.25293459214590036, + "grad_norm": 5.584639072418213, + "learning_rate": 0.00017009470623179077, + "loss": 0.8634, + "step": 10790 + }, + { + "epoch": 0.25295803372070536, + "grad_norm": 5.6814775466918945, + "learning_rate": 0.0001700894523998688, + "loss": 0.8736, + "step": 10791 + }, + { + "epoch": 0.25298147529551035, + "grad_norm": 3.215266704559326, + "learning_rate": 0.0001700841981876395, + "loss": 0.4987, + "step": 10792 + }, + { + "epoch": 0.25300491687031534, + "grad_norm": 5.2673139572143555, + "learning_rate": 0.0001700789435951314, + "loss": 0.4079, + "step": 10793 + }, + { + "epoch": 0.25302835844512034, + "grad_norm": 9.678335189819336, + "learning_rate": 0.00017007368862237296, + "loss": 0.4444, + "step": 10794 + }, + { + "epoch": 0.25305180001992533, + "grad_norm": 1.1318409442901611, + "learning_rate": 0.00017006843326939272, + "loss": 0.0882, + "step": 10795 + }, + { + "epoch": 0.2530752415947303, + "grad_norm": 4.006050109863281, + "learning_rate": 0.00017006317753621918, + "loss": 0.3551, + "step": 10796 + }, + { + "epoch": 0.2530986831695353, + "grad_norm": 3.02883243560791, + "learning_rate": 0.00017005792142288088, + "loss": 1.0738, + "step": 10797 + }, + { + "epoch": 0.2531221247443403, + "grad_norm": 1.5305012464523315, + "learning_rate": 0.0001700526649294063, + "loss": 0.1942, + "step": 10798 + }, + { + "epoch": 0.2531455663191453, + "grad_norm": 2.1118032932281494, + "learning_rate": 0.00017004740805582404, + "loss": 0.2139, + "step": 10799 + }, + { + "epoch": 0.2531690078939503, + "grad_norm": 3.6276512145996094, + "learning_rate": 0.00017004215080216256, + "loss": 0.411, + "step": 10800 + }, + { + "epoch": 0.2531924494687553, + "grad_norm": 4.34694242477417, + "learning_rate": 0.00017003689316845037, + "loss": 2.2186, + "step": 10801 + }, + { + "epoch": 0.2532158910435603, + "grad_norm": 0.6989090442657471, + "learning_rate": 0.00017003163515471605, + "loss": 0.1177, + "step": 10802 + }, + { + "epoch": 0.2532393326183653, + "grad_norm": 3.424560308456421, + "learning_rate": 0.00017002637676098812, + "loss": 0.392, + "step": 10803 + }, + { + "epoch": 0.2532627741931703, + "grad_norm": 1.0462137460708618, + "learning_rate": 0.00017002111798729508, + "loss": 0.1723, + "step": 10804 + }, + { + "epoch": 0.2532862157679753, + "grad_norm": 2.7650771141052246, + "learning_rate": 0.00017001585883366548, + "loss": 0.2799, + "step": 10805 + }, + { + "epoch": 0.25330965734278027, + "grad_norm": 3.0457212924957275, + "learning_rate": 0.00017001059930012787, + "loss": 0.337, + "step": 10806 + }, + { + "epoch": 0.25333309891758526, + "grad_norm": 3.9854378700256348, + "learning_rate": 0.0001700053393867108, + "loss": 0.7295, + "step": 10807 + }, + { + "epoch": 0.25335654049239026, + "grad_norm": 2.500356674194336, + "learning_rate": 0.00017000007909344274, + "loss": 0.3047, + "step": 10808 + }, + { + "epoch": 0.25337998206719525, + "grad_norm": 4.560007095336914, + "learning_rate": 0.00016999481842035238, + "loss": 0.6715, + "step": 10809 + }, + { + "epoch": 0.25340342364200025, + "grad_norm": 2.890786647796631, + "learning_rate": 0.00016998955736746804, + "loss": 0.2789, + "step": 10810 + }, + { + "epoch": 0.25342686521680524, + "grad_norm": 0.5536222457885742, + "learning_rate": 0.00016998429593481848, + "loss": 0.0752, + "step": 10811 + }, + { + "epoch": 0.25345030679161024, + "grad_norm": 1.5000027418136597, + "learning_rate": 0.00016997903412243215, + "loss": 0.1772, + "step": 10812 + }, + { + "epoch": 0.25347374836641523, + "grad_norm": 3.131770372390747, + "learning_rate": 0.0001699737719303376, + "loss": 0.2894, + "step": 10813 + }, + { + "epoch": 0.2534971899412202, + "grad_norm": 2.520824670791626, + "learning_rate": 0.0001699685093585634, + "loss": 0.3232, + "step": 10814 + }, + { + "epoch": 0.2535206315160252, + "grad_norm": 0.40512171387672424, + "learning_rate": 0.00016996324640713807, + "loss": 0.0159, + "step": 10815 + }, + { + "epoch": 0.25354407309083027, + "grad_norm": 5.45858907699585, + "learning_rate": 0.00016995798307609025, + "loss": 0.4704, + "step": 10816 + }, + { + "epoch": 0.25356751466563526, + "grad_norm": 6.084419250488281, + "learning_rate": 0.0001699527193654484, + "loss": 0.708, + "step": 10817 + }, + { + "epoch": 0.25359095624044026, + "grad_norm": 4.132588863372803, + "learning_rate": 0.00016994745527524116, + "loss": 0.5563, + "step": 10818 + }, + { + "epoch": 0.25361439781524525, + "grad_norm": 2.7761006355285645, + "learning_rate": 0.00016994219080549707, + "loss": 0.4629, + "step": 10819 + }, + { + "epoch": 0.25363783939005025, + "grad_norm": 3.775132179260254, + "learning_rate": 0.00016993692595624466, + "loss": 0.7807, + "step": 10820 + }, + { + "epoch": 0.25366128096485524, + "grad_norm": 4.657879829406738, + "learning_rate": 0.00016993166072751253, + "loss": 0.3484, + "step": 10821 + }, + { + "epoch": 0.25368472253966023, + "grad_norm": 4.6779046058654785, + "learning_rate": 0.00016992639511932925, + "loss": 0.4237, + "step": 10822 + }, + { + "epoch": 0.25370816411446523, + "grad_norm": 3.5550220012664795, + "learning_rate": 0.00016992112913172337, + "loss": 0.5175, + "step": 10823 + }, + { + "epoch": 0.2537316056892702, + "grad_norm": 3.1820027828216553, + "learning_rate": 0.0001699158627647235, + "loss": 0.3468, + "step": 10824 + }, + { + "epoch": 0.2537550472640752, + "grad_norm": 2.8294806480407715, + "learning_rate": 0.00016991059601835818, + "loss": 0.3443, + "step": 10825 + }, + { + "epoch": 0.2537784888388802, + "grad_norm": 4.885488986968994, + "learning_rate": 0.00016990532889265598, + "loss": 0.482, + "step": 10826 + }, + { + "epoch": 0.2538019304136852, + "grad_norm": 2.5853183269500732, + "learning_rate": 0.00016990006138764555, + "loss": 0.3405, + "step": 10827 + }, + { + "epoch": 0.2538253719884902, + "grad_norm": 6.344964027404785, + "learning_rate": 0.00016989479350335536, + "loss": 0.4927, + "step": 10828 + }, + { + "epoch": 0.2538488135632952, + "grad_norm": 4.325407028198242, + "learning_rate": 0.0001698895252398141, + "loss": 0.4452, + "step": 10829 + }, + { + "epoch": 0.2538722551381002, + "grad_norm": 1.4620401859283447, + "learning_rate": 0.0001698842565970503, + "loss": 0.1634, + "step": 10830 + }, + { + "epoch": 0.2538956967129052, + "grad_norm": 1.965445876121521, + "learning_rate": 0.00016987898757509252, + "loss": 1.2039, + "step": 10831 + }, + { + "epoch": 0.2539191382877102, + "grad_norm": 4.806118011474609, + "learning_rate": 0.00016987371817396945, + "loss": 0.6514, + "step": 10832 + }, + { + "epoch": 0.25394257986251517, + "grad_norm": 3.8381340503692627, + "learning_rate": 0.00016986844839370958, + "loss": 0.7382, + "step": 10833 + }, + { + "epoch": 0.25396602143732017, + "grad_norm": 4.270989418029785, + "learning_rate": 0.00016986317823434156, + "loss": 0.3883, + "step": 10834 + }, + { + "epoch": 0.25398946301212516, + "grad_norm": 4.4516706466674805, + "learning_rate": 0.000169857907695894, + "loss": 0.5469, + "step": 10835 + }, + { + "epoch": 0.25401290458693015, + "grad_norm": 2.614938735961914, + "learning_rate": 0.00016985263677839543, + "loss": 0.383, + "step": 10836 + }, + { + "epoch": 0.25403634616173515, + "grad_norm": 1.7471727132797241, + "learning_rate": 0.0001698473654818745, + "loss": 0.2638, + "step": 10837 + }, + { + "epoch": 0.25405978773654014, + "grad_norm": 2.2731659412384033, + "learning_rate": 0.00016984209380635978, + "loss": 0.213, + "step": 10838 + }, + { + "epoch": 0.25408322931134514, + "grad_norm": 1.6675890684127808, + "learning_rate": 0.00016983682175187993, + "loss": 0.1844, + "step": 10839 + }, + { + "epoch": 0.25410667088615013, + "grad_norm": 1.7831416130065918, + "learning_rate": 0.00016983154931846347, + "loss": 0.1869, + "step": 10840 + }, + { + "epoch": 0.2541301124609551, + "grad_norm": 4.772267818450928, + "learning_rate": 0.0001698262765061391, + "loss": 0.2243, + "step": 10841 + }, + { + "epoch": 0.2541535540357601, + "grad_norm": 4.568443775177002, + "learning_rate": 0.0001698210033149354, + "loss": 0.3149, + "step": 10842 + }, + { + "epoch": 0.2541769956105651, + "grad_norm": 4.865123271942139, + "learning_rate": 0.00016981572974488092, + "loss": 0.6501, + "step": 10843 + }, + { + "epoch": 0.2542004371853701, + "grad_norm": 1.9580453634262085, + "learning_rate": 0.00016981045579600438, + "loss": 1.7103, + "step": 10844 + }, + { + "epoch": 0.2542238787601751, + "grad_norm": 0.5886850953102112, + "learning_rate": 0.00016980518146833434, + "loss": 0.0618, + "step": 10845 + }, + { + "epoch": 0.2542473203349801, + "grad_norm": 0.9487919807434082, + "learning_rate": 0.00016979990676189942, + "loss": 0.06, + "step": 10846 + }, + { + "epoch": 0.2542707619097851, + "grad_norm": 3.4635932445526123, + "learning_rate": 0.00016979463167672823, + "loss": 0.2816, + "step": 10847 + }, + { + "epoch": 0.2542942034845901, + "grad_norm": 2.496002197265625, + "learning_rate": 0.00016978935621284942, + "loss": 0.3251, + "step": 10848 + }, + { + "epoch": 0.2543176450593951, + "grad_norm": 3.8327953815460205, + "learning_rate": 0.00016978408037029156, + "loss": 0.256, + "step": 10849 + }, + { + "epoch": 0.2543410866342001, + "grad_norm": 3.613725423812866, + "learning_rate": 0.00016977880414908335, + "loss": 0.4439, + "step": 10850 + }, + { + "epoch": 0.25436452820900507, + "grad_norm": 3.5779478549957275, + "learning_rate": 0.0001697735275492534, + "loss": 0.2513, + "step": 10851 + }, + { + "epoch": 0.25438796978381006, + "grad_norm": 3.592717409133911, + "learning_rate": 0.00016976825057083032, + "loss": 0.3074, + "step": 10852 + }, + { + "epoch": 0.25441141135861506, + "grad_norm": 1.3671354055404663, + "learning_rate": 0.00016976297321384274, + "loss": 0.1826, + "step": 10853 + }, + { + "epoch": 0.25443485293342005, + "grad_norm": 0.39820489287376404, + "learning_rate": 0.0001697576954783193, + "loss": 0.0508, + "step": 10854 + }, + { + "epoch": 0.25445829450822505, + "grad_norm": 3.958451986312866, + "learning_rate": 0.00016975241736428868, + "loss": 0.1986, + "step": 10855 + }, + { + "epoch": 0.25448173608303004, + "grad_norm": 4.252499580383301, + "learning_rate": 0.00016974713887177943, + "loss": 0.4067, + "step": 10856 + }, + { + "epoch": 0.25450517765783504, + "grad_norm": 3.3300323486328125, + "learning_rate": 0.00016974186000082029, + "loss": 0.4355, + "step": 10857 + }, + { + "epoch": 0.25452861923264003, + "grad_norm": 0.7095337510108948, + "learning_rate": 0.0001697365807514398, + "loss": 0.0871, + "step": 10858 + }, + { + "epoch": 0.254552060807445, + "grad_norm": 2.220064878463745, + "learning_rate": 0.0001697313011236667, + "loss": 0.1969, + "step": 10859 + }, + { + "epoch": 0.25457550238225, + "grad_norm": 1.401726484298706, + "learning_rate": 0.0001697260211175296, + "loss": 0.2167, + "step": 10860 + }, + { + "epoch": 0.254598943957055, + "grad_norm": 3.1469955444335938, + "learning_rate": 0.00016972074073305716, + "loss": 0.0888, + "step": 10861 + }, + { + "epoch": 0.25462238553186, + "grad_norm": 3.6981961727142334, + "learning_rate": 0.00016971545997027804, + "loss": 0.3636, + "step": 10862 + }, + { + "epoch": 0.254645827106665, + "grad_norm": 3.832794666290283, + "learning_rate": 0.0001697101788292208, + "loss": 0.5174, + "step": 10863 + }, + { + "epoch": 0.25466926868147, + "grad_norm": 6.9761481285095215, + "learning_rate": 0.0001697048973099142, + "loss": 0.8064, + "step": 10864 + }, + { + "epoch": 0.254692710256275, + "grad_norm": 4.5554327964782715, + "learning_rate": 0.0001696996154123869, + "loss": 0.6466, + "step": 10865 + }, + { + "epoch": 0.25471615183108, + "grad_norm": 4.0095624923706055, + "learning_rate": 0.0001696943331366675, + "loss": 0.9761, + "step": 10866 + }, + { + "epoch": 0.25473959340588503, + "grad_norm": 3.463549852371216, + "learning_rate": 0.0001696890504827847, + "loss": 0.3293, + "step": 10867 + }, + { + "epoch": 0.25476303498069003, + "grad_norm": 5.688555717468262, + "learning_rate": 0.00016968376745076712, + "loss": 0.6549, + "step": 10868 + }, + { + "epoch": 0.254786476555495, + "grad_norm": 3.92714262008667, + "learning_rate": 0.0001696784840406435, + "loss": 0.6192, + "step": 10869 + }, + { + "epoch": 0.2548099181303, + "grad_norm": 1.3943853378295898, + "learning_rate": 0.00016967320025244244, + "loss": 0.0661, + "step": 10870 + }, + { + "epoch": 0.254833359705105, + "grad_norm": 3.350572347640991, + "learning_rate": 0.00016966791608619266, + "loss": 0.2341, + "step": 10871 + }, + { + "epoch": 0.25485680127991, + "grad_norm": 2.3360848426818848, + "learning_rate": 0.0001696626315419228, + "loss": 0.2842, + "step": 10872 + }, + { + "epoch": 0.254880242854715, + "grad_norm": 1.453931212425232, + "learning_rate": 0.00016965734661966153, + "loss": 0.1707, + "step": 10873 + }, + { + "epoch": 0.25490368442952, + "grad_norm": 2.589510917663574, + "learning_rate": 0.00016965206131943755, + "loss": 0.2759, + "step": 10874 + }, + { + "epoch": 0.254927126004325, + "grad_norm": 1.4278335571289062, + "learning_rate": 0.0001696467756412795, + "loss": 0.1491, + "step": 10875 + }, + { + "epoch": 0.25495056757913, + "grad_norm": 4.181624889373779, + "learning_rate": 0.00016964148958521612, + "loss": 0.3341, + "step": 10876 + }, + { + "epoch": 0.254974009153935, + "grad_norm": 4.460717678070068, + "learning_rate": 0.00016963620315127605, + "loss": 0.4032, + "step": 10877 + }, + { + "epoch": 0.25499745072873997, + "grad_norm": 2.770134925842285, + "learning_rate": 0.00016963091633948797, + "loss": 0.208, + "step": 10878 + }, + { + "epoch": 0.25502089230354497, + "grad_norm": 2.933640480041504, + "learning_rate": 0.0001696256291498806, + "loss": 0.2661, + "step": 10879 + }, + { + "epoch": 0.25504433387834996, + "grad_norm": 1.572586178779602, + "learning_rate": 0.0001696203415824826, + "loss": 0.1891, + "step": 10880 + }, + { + "epoch": 0.25506777545315495, + "grad_norm": 0.8956462144851685, + "learning_rate": 0.0001696150536373227, + "loss": 0.1245, + "step": 10881 + }, + { + "epoch": 0.25509121702795995, + "grad_norm": 2.1378347873687744, + "learning_rate": 0.00016960976531442951, + "loss": 0.6574, + "step": 10882 + }, + { + "epoch": 0.25511465860276494, + "grad_norm": 3.244009017944336, + "learning_rate": 0.00016960447661383177, + "loss": 0.3967, + "step": 10883 + }, + { + "epoch": 0.25513810017756994, + "grad_norm": 2.7718687057495117, + "learning_rate": 0.00016959918753555822, + "loss": 0.3606, + "step": 10884 + }, + { + "epoch": 0.25516154175237493, + "grad_norm": 1.7531107664108276, + "learning_rate": 0.0001695938980796375, + "loss": 0.1852, + "step": 10885 + }, + { + "epoch": 0.2551849833271799, + "grad_norm": 4.82444429397583, + "learning_rate": 0.00016958860824609838, + "loss": 0.5905, + "step": 10886 + }, + { + "epoch": 0.2552084249019849, + "grad_norm": 3.1709201335906982, + "learning_rate": 0.0001695833180349695, + "loss": 0.4154, + "step": 10887 + }, + { + "epoch": 0.2552318664767899, + "grad_norm": 4.22412109375, + "learning_rate": 0.00016957802744627952, + "loss": 0.4361, + "step": 10888 + }, + { + "epoch": 0.2552553080515949, + "grad_norm": 2.6678555011749268, + "learning_rate": 0.00016957273648005727, + "loss": 0.4215, + "step": 10889 + }, + { + "epoch": 0.2552787496263999, + "grad_norm": 4.357763290405273, + "learning_rate": 0.00016956744513633136, + "loss": 0.359, + "step": 10890 + }, + { + "epoch": 0.2553021912012049, + "grad_norm": 4.374691486358643, + "learning_rate": 0.00016956215341513056, + "loss": 0.8407, + "step": 10891 + }, + { + "epoch": 0.2553256327760099, + "grad_norm": 3.7889792919158936, + "learning_rate": 0.00016955686131648354, + "loss": 0.4084, + "step": 10892 + }, + { + "epoch": 0.2553490743508149, + "grad_norm": 4.173784255981445, + "learning_rate": 0.00016955156884041904, + "loss": 0.7451, + "step": 10893 + }, + { + "epoch": 0.2553725159256199, + "grad_norm": 3.537997007369995, + "learning_rate": 0.00016954627598696578, + "loss": 1.3988, + "step": 10894 + }, + { + "epoch": 0.2553959575004249, + "grad_norm": 5.4539055824279785, + "learning_rate": 0.00016954098275615245, + "loss": 0.3054, + "step": 10895 + }, + { + "epoch": 0.25541939907522987, + "grad_norm": 1.157199740409851, + "learning_rate": 0.00016953568914800782, + "loss": 0.1319, + "step": 10896 + }, + { + "epoch": 0.25544284065003486, + "grad_norm": 3.374922752380371, + "learning_rate": 0.00016953039516256056, + "loss": 0.4333, + "step": 10897 + }, + { + "epoch": 0.25546628222483986, + "grad_norm": 3.8583984375, + "learning_rate": 0.00016952510079983944, + "loss": 0.4588, + "step": 10898 + }, + { + "epoch": 0.25548972379964485, + "grad_norm": 0.4755569398403168, + "learning_rate": 0.00016951980605987317, + "loss": 0.0515, + "step": 10899 + }, + { + "epoch": 0.25551316537444985, + "grad_norm": 1.2283384799957275, + "learning_rate": 0.00016951451094269045, + "loss": 0.1099, + "step": 10900 + }, + { + "epoch": 0.25553660694925484, + "grad_norm": 4.066287040710449, + "learning_rate": 0.00016950921544832008, + "loss": 0.3983, + "step": 10901 + }, + { + "epoch": 0.25556004852405984, + "grad_norm": 0.9121137261390686, + "learning_rate": 0.00016950391957679074, + "loss": 0.0842, + "step": 10902 + }, + { + "epoch": 0.25558349009886483, + "grad_norm": 1.0043092966079712, + "learning_rate": 0.00016949862332813116, + "loss": 0.1259, + "step": 10903 + }, + { + "epoch": 0.2556069316736698, + "grad_norm": 3.2697222232818604, + "learning_rate": 0.0001694933267023701, + "loss": 0.3018, + "step": 10904 + }, + { + "epoch": 0.2556303732484748, + "grad_norm": 3.2736129760742188, + "learning_rate": 0.0001694880296995363, + "loss": 0.4159, + "step": 10905 + }, + { + "epoch": 0.2556538148232798, + "grad_norm": 2.429497241973877, + "learning_rate": 0.00016948273231965852, + "loss": 0.2179, + "step": 10906 + }, + { + "epoch": 0.2556772563980848, + "grad_norm": 2.7617123126983643, + "learning_rate": 0.00016947743456276543, + "loss": 0.2558, + "step": 10907 + }, + { + "epoch": 0.2557006979728898, + "grad_norm": 0.747725784778595, + "learning_rate": 0.00016947213642888587, + "loss": 0.0431, + "step": 10908 + }, + { + "epoch": 0.2557241395476948, + "grad_norm": 2.7170138359069824, + "learning_rate": 0.0001694668379180485, + "loss": 0.2848, + "step": 10909 + }, + { + "epoch": 0.2557475811224998, + "grad_norm": 3.2078771591186523, + "learning_rate": 0.00016946153903028217, + "loss": 0.1667, + "step": 10910 + }, + { + "epoch": 0.2557710226973048, + "grad_norm": 3.756894826889038, + "learning_rate": 0.00016945623976561555, + "loss": 0.465, + "step": 10911 + }, + { + "epoch": 0.2557944642721098, + "grad_norm": 3.29341983795166, + "learning_rate": 0.0001694509401240774, + "loss": 0.4754, + "step": 10912 + }, + { + "epoch": 0.2558179058469148, + "grad_norm": 4.6859331130981445, + "learning_rate": 0.00016944564010569653, + "loss": 0.2311, + "step": 10913 + }, + { + "epoch": 0.25584134742171977, + "grad_norm": 2.539548397064209, + "learning_rate": 0.00016944033971050165, + "loss": 0.3384, + "step": 10914 + }, + { + "epoch": 0.25586478899652476, + "grad_norm": 4.378772735595703, + "learning_rate": 0.00016943503893852152, + "loss": 0.4092, + "step": 10915 + }, + { + "epoch": 0.25588823057132976, + "grad_norm": 0.9437618255615234, + "learning_rate": 0.00016942973778978496, + "loss": 0.1303, + "step": 10916 + }, + { + "epoch": 0.25591167214613475, + "grad_norm": 2.7149107456207275, + "learning_rate": 0.00016942443626432064, + "loss": 1.9142, + "step": 10917 + }, + { + "epoch": 0.2559351137209398, + "grad_norm": 3.8091654777526855, + "learning_rate": 0.00016941913436215739, + "loss": 0.6118, + "step": 10918 + }, + { + "epoch": 0.2559585552957448, + "grad_norm": 3.557682514190674, + "learning_rate": 0.00016941383208332397, + "loss": 0.2971, + "step": 10919 + }, + { + "epoch": 0.2559819968705498, + "grad_norm": 3.3232290744781494, + "learning_rate": 0.00016940852942784918, + "loss": 0.4225, + "step": 10920 + }, + { + "epoch": 0.2560054384453548, + "grad_norm": 3.4879512786865234, + "learning_rate": 0.00016940322639576172, + "loss": 0.5807, + "step": 10921 + }, + { + "epoch": 0.2560288800201598, + "grad_norm": 2.2498104572296143, + "learning_rate": 0.0001693979229870904, + "loss": 0.2417, + "step": 10922 + }, + { + "epoch": 0.25605232159496477, + "grad_norm": 3.8146002292633057, + "learning_rate": 0.000169392619201864, + "loss": 0.5471, + "step": 10923 + }, + { + "epoch": 0.25607576316976977, + "grad_norm": 3.628427505493164, + "learning_rate": 0.0001693873150401113, + "loss": 0.3697, + "step": 10924 + }, + { + "epoch": 0.25609920474457476, + "grad_norm": 2.9885733127593994, + "learning_rate": 0.00016938201050186108, + "loss": 1.9924, + "step": 10925 + }, + { + "epoch": 0.25612264631937975, + "grad_norm": 2.4284920692443848, + "learning_rate": 0.00016937670558714213, + "loss": 0.2251, + "step": 10926 + }, + { + "epoch": 0.25614608789418475, + "grad_norm": 0.7918274998664856, + "learning_rate": 0.0001693714002959832, + "loss": 0.1033, + "step": 10927 + }, + { + "epoch": 0.25616952946898974, + "grad_norm": 2.750791549682617, + "learning_rate": 0.00016936609462841316, + "loss": 1.9729, + "step": 10928 + }, + { + "epoch": 0.25619297104379474, + "grad_norm": 3.6097283363342285, + "learning_rate": 0.00016936078858446066, + "loss": 0.3635, + "step": 10929 + }, + { + "epoch": 0.25621641261859973, + "grad_norm": 2.1221492290496826, + "learning_rate": 0.00016935548216415463, + "loss": 1.908, + "step": 10930 + }, + { + "epoch": 0.2562398541934047, + "grad_norm": 1.129491925239563, + "learning_rate": 0.00016935017536752378, + "loss": 0.1573, + "step": 10931 + }, + { + "epoch": 0.2562632957682097, + "grad_norm": 3.640172004699707, + "learning_rate": 0.00016934486819459693, + "loss": 0.3897, + "step": 10932 + }, + { + "epoch": 0.2562867373430147, + "grad_norm": 3.2783608436584473, + "learning_rate": 0.00016933956064540286, + "loss": 0.3639, + "step": 10933 + }, + { + "epoch": 0.2563101789178197, + "grad_norm": 2.5080294609069824, + "learning_rate": 0.00016933425271997043, + "loss": 0.258, + "step": 10934 + }, + { + "epoch": 0.2563336204926247, + "grad_norm": 2.428196907043457, + "learning_rate": 0.00016932894441832837, + "loss": 0.2608, + "step": 10935 + }, + { + "epoch": 0.2563570620674297, + "grad_norm": 1.744881272315979, + "learning_rate": 0.0001693236357405055, + "loss": 0.1562, + "step": 10936 + }, + { + "epoch": 0.2563805036422347, + "grad_norm": 4.797714710235596, + "learning_rate": 0.00016931832668653062, + "loss": 0.4172, + "step": 10937 + }, + { + "epoch": 0.2564039452170397, + "grad_norm": 0.5502845644950867, + "learning_rate": 0.00016931301725643257, + "loss": 0.0507, + "step": 10938 + }, + { + "epoch": 0.2564273867918447, + "grad_norm": 4.353276252746582, + "learning_rate": 0.00016930770745024015, + "loss": 0.7294, + "step": 10939 + }, + { + "epoch": 0.2564508283666497, + "grad_norm": 1.4797645807266235, + "learning_rate": 0.00016930239726798213, + "loss": 0.1155, + "step": 10940 + }, + { + "epoch": 0.25647426994145467, + "grad_norm": 1.328088641166687, + "learning_rate": 0.0001692970867096874, + "loss": 0.1546, + "step": 10941 + }, + { + "epoch": 0.25649771151625966, + "grad_norm": 1.2151963710784912, + "learning_rate": 0.00016929177577538468, + "loss": 0.0782, + "step": 10942 + }, + { + "epoch": 0.25652115309106466, + "grad_norm": 4.138044834136963, + "learning_rate": 0.00016928646446510284, + "loss": 0.425, + "step": 10943 + }, + { + "epoch": 0.25654459466586965, + "grad_norm": 1.2078585624694824, + "learning_rate": 0.00016928115277887074, + "loss": 0.1292, + "step": 10944 + }, + { + "epoch": 0.25656803624067465, + "grad_norm": 3.5245869159698486, + "learning_rate": 0.00016927584071671712, + "loss": 0.3441, + "step": 10945 + }, + { + "epoch": 0.25659147781547964, + "grad_norm": 0.765212893486023, + "learning_rate": 0.00016927052827867085, + "loss": 0.0519, + "step": 10946 + }, + { + "epoch": 0.25661491939028463, + "grad_norm": 3.645944118499756, + "learning_rate": 0.00016926521546476074, + "loss": 0.3452, + "step": 10947 + }, + { + "epoch": 0.25663836096508963, + "grad_norm": 2.796257495880127, + "learning_rate": 0.00016925990227501563, + "loss": 1.1637, + "step": 10948 + }, + { + "epoch": 0.2566618025398946, + "grad_norm": 1.7298712730407715, + "learning_rate": 0.00016925458870946435, + "loss": 0.124, + "step": 10949 + }, + { + "epoch": 0.2566852441146996, + "grad_norm": 4.601357936859131, + "learning_rate": 0.0001692492747681357, + "loss": 0.49, + "step": 10950 + }, + { + "epoch": 0.2567086856895046, + "grad_norm": 6.640171527862549, + "learning_rate": 0.00016924396045105857, + "loss": 0.4404, + "step": 10951 + }, + { + "epoch": 0.2567321272643096, + "grad_norm": 1.1093485355377197, + "learning_rate": 0.00016923864575826177, + "loss": 0.1277, + "step": 10952 + }, + { + "epoch": 0.2567555688391146, + "grad_norm": 1.9975864887237549, + "learning_rate": 0.00016923333068977406, + "loss": 0.2049, + "step": 10953 + }, + { + "epoch": 0.2567790104139196, + "grad_norm": 1.1737124919891357, + "learning_rate": 0.00016922801524562445, + "loss": 0.1296, + "step": 10954 + }, + { + "epoch": 0.2568024519887246, + "grad_norm": 3.5405771732330322, + "learning_rate": 0.00016922269942584158, + "loss": 0.4556, + "step": 10955 + }, + { + "epoch": 0.2568258935635296, + "grad_norm": 2.8017401695251465, + "learning_rate": 0.0001692173832304545, + "loss": 0.2255, + "step": 10956 + }, + { + "epoch": 0.2568493351383346, + "grad_norm": 3.727368116378784, + "learning_rate": 0.0001692120666594919, + "loss": 0.2232, + "step": 10957 + }, + { + "epoch": 0.2568727767131396, + "grad_norm": 2.847536325454712, + "learning_rate": 0.00016920674971298265, + "loss": 1.7606, + "step": 10958 + }, + { + "epoch": 0.25689621828794457, + "grad_norm": 4.031620025634766, + "learning_rate": 0.00016920143239095568, + "loss": 0.5428, + "step": 10959 + }, + { + "epoch": 0.25691965986274956, + "grad_norm": 5.6143412590026855, + "learning_rate": 0.00016919611469343976, + "loss": 0.4941, + "step": 10960 + }, + { + "epoch": 0.25694310143755456, + "grad_norm": 4.621371746063232, + "learning_rate": 0.0001691907966204638, + "loss": 0.3921, + "step": 10961 + }, + { + "epoch": 0.25696654301235955, + "grad_norm": 3.848818063735962, + "learning_rate": 0.00016918547817205664, + "loss": 0.4761, + "step": 10962 + }, + { + "epoch": 0.25698998458716454, + "grad_norm": 3.674565315246582, + "learning_rate": 0.00016918015934824708, + "loss": 0.3359, + "step": 10963 + }, + { + "epoch": 0.25701342616196954, + "grad_norm": 2.0143280029296875, + "learning_rate": 0.00016917484014906408, + "loss": 1.0547, + "step": 10964 + }, + { + "epoch": 0.25703686773677453, + "grad_norm": 0.7707948088645935, + "learning_rate": 0.00016916952057453646, + "loss": 0.1085, + "step": 10965 + }, + { + "epoch": 0.2570603093115795, + "grad_norm": 3.4174811840057373, + "learning_rate": 0.00016916420062469304, + "loss": 0.2994, + "step": 10966 + }, + { + "epoch": 0.2570837508863845, + "grad_norm": 2.5156338214874268, + "learning_rate": 0.00016915888029956272, + "loss": 1.4854, + "step": 10967 + }, + { + "epoch": 0.2571071924611895, + "grad_norm": 2.9187498092651367, + "learning_rate": 0.00016915355959917438, + "loss": 0.219, + "step": 10968 + }, + { + "epoch": 0.25713063403599457, + "grad_norm": 1.572941541671753, + "learning_rate": 0.0001691482385235569, + "loss": 0.1336, + "step": 10969 + }, + { + "epoch": 0.25715407561079956, + "grad_norm": 2.277573823928833, + "learning_rate": 0.00016914291707273912, + "loss": 1.9831, + "step": 10970 + }, + { + "epoch": 0.25717751718560455, + "grad_norm": 0.5037795305252075, + "learning_rate": 0.00016913759524674992, + "loss": 0.0759, + "step": 10971 + }, + { + "epoch": 0.25720095876040955, + "grad_norm": 3.1240499019622803, + "learning_rate": 0.00016913227304561819, + "loss": 1.9572, + "step": 10972 + }, + { + "epoch": 0.25722440033521454, + "grad_norm": 1.45736563205719, + "learning_rate": 0.00016912695046937282, + "loss": 0.1468, + "step": 10973 + }, + { + "epoch": 0.25724784191001954, + "grad_norm": 1.3525973558425903, + "learning_rate": 0.00016912162751804262, + "loss": 0.1205, + "step": 10974 + }, + { + "epoch": 0.25727128348482453, + "grad_norm": 4.46381139755249, + "learning_rate": 0.0001691163041916566, + "loss": 0.366, + "step": 10975 + }, + { + "epoch": 0.2572947250596295, + "grad_norm": 5.092264652252197, + "learning_rate": 0.00016911098049024353, + "loss": 0.5388, + "step": 10976 + }, + { + "epoch": 0.2573181666344345, + "grad_norm": 1.8191310167312622, + "learning_rate": 0.00016910565641383235, + "loss": 0.3662, + "step": 10977 + }, + { + "epoch": 0.2573416082092395, + "grad_norm": 1.315192699432373, + "learning_rate": 0.00016910033196245193, + "loss": 0.1315, + "step": 10978 + }, + { + "epoch": 0.2573650497840445, + "grad_norm": 1.056933879852295, + "learning_rate": 0.00016909500713613114, + "loss": 0.1336, + "step": 10979 + }, + { + "epoch": 0.2573884913588495, + "grad_norm": 0.5893889665603638, + "learning_rate": 0.0001690896819348989, + "loss": 0.0558, + "step": 10980 + }, + { + "epoch": 0.2574119329336545, + "grad_norm": 1.187488079071045, + "learning_rate": 0.00016908435635878415, + "loss": 0.1498, + "step": 10981 + }, + { + "epoch": 0.2574353745084595, + "grad_norm": 3.1168675422668457, + "learning_rate": 0.0001690790304078157, + "loss": 0.4133, + "step": 10982 + }, + { + "epoch": 0.2574588160832645, + "grad_norm": 2.4612016677856445, + "learning_rate": 0.0001690737040820225, + "loss": 0.2905, + "step": 10983 + }, + { + "epoch": 0.2574822576580695, + "grad_norm": 0.33746466040611267, + "learning_rate": 0.00016906837738143345, + "loss": 0.0378, + "step": 10984 + }, + { + "epoch": 0.2575056992328745, + "grad_norm": 2.8858864307403564, + "learning_rate": 0.00016906305030607745, + "loss": 0.5358, + "step": 10985 + }, + { + "epoch": 0.25752914080767947, + "grad_norm": 1.034847378730774, + "learning_rate": 0.00016905772285598337, + "loss": 0.0695, + "step": 10986 + }, + { + "epoch": 0.25755258238248446, + "grad_norm": 3.458220958709717, + "learning_rate": 0.00016905239503118015, + "loss": 1.6374, + "step": 10987 + }, + { + "epoch": 0.25757602395728946, + "grad_norm": 2.181791305541992, + "learning_rate": 0.0001690470668316967, + "loss": 0.3519, + "step": 10988 + }, + { + "epoch": 0.25759946553209445, + "grad_norm": 2.6994786262512207, + "learning_rate": 0.00016904173825756195, + "loss": 0.224, + "step": 10989 + }, + { + "epoch": 0.25762290710689945, + "grad_norm": 4.321134090423584, + "learning_rate": 0.00016903640930880474, + "loss": 0.5093, + "step": 10990 + }, + { + "epoch": 0.25764634868170444, + "grad_norm": 2.3801450729370117, + "learning_rate": 0.00016903107998545404, + "loss": 0.3025, + "step": 10991 + }, + { + "epoch": 0.25766979025650943, + "grad_norm": 2.1191370487213135, + "learning_rate": 0.00016902575028753875, + "loss": 1.2224, + "step": 10992 + }, + { + "epoch": 0.25769323183131443, + "grad_norm": 0.6014828681945801, + "learning_rate": 0.00016902042021508783, + "loss": 0.0724, + "step": 10993 + }, + { + "epoch": 0.2577166734061194, + "grad_norm": 5.02375602722168, + "learning_rate": 0.00016901508976813017, + "loss": 0.5635, + "step": 10994 + }, + { + "epoch": 0.2577401149809244, + "grad_norm": 3.234407663345337, + "learning_rate": 0.00016900975894669466, + "loss": 0.3861, + "step": 10995 + }, + { + "epoch": 0.2577635565557294, + "grad_norm": 4.293354511260986, + "learning_rate": 0.0001690044277508103, + "loss": 0.4122, + "step": 10996 + }, + { + "epoch": 0.2577869981305344, + "grad_norm": 2.9903602600097656, + "learning_rate": 0.00016899909618050594, + "loss": 2.6033, + "step": 10997 + }, + { + "epoch": 0.2578104397053394, + "grad_norm": 0.4160778820514679, + "learning_rate": 0.00016899376423581055, + "loss": 0.0554, + "step": 10998 + }, + { + "epoch": 0.2578338812801444, + "grad_norm": 3.680614948272705, + "learning_rate": 0.00016898843191675306, + "loss": 0.4229, + "step": 10999 + }, + { + "epoch": 0.2578573228549494, + "grad_norm": 5.174761772155762, + "learning_rate": 0.0001689830992233624, + "loss": 0.8579, + "step": 11000 + }, + { + "epoch": 0.2578807644297544, + "grad_norm": 4.502912998199463, + "learning_rate": 0.0001689777661556675, + "loss": 0.4596, + "step": 11001 + }, + { + "epoch": 0.2579042060045594, + "grad_norm": 1.1889094114303589, + "learning_rate": 0.00016897243271369732, + "loss": 0.1433, + "step": 11002 + }, + { + "epoch": 0.25792764757936437, + "grad_norm": 2.189387559890747, + "learning_rate": 0.00016896709889748074, + "loss": 0.8852, + "step": 11003 + }, + { + "epoch": 0.25795108915416937, + "grad_norm": 0.4458956718444824, + "learning_rate": 0.00016896176470704676, + "loss": 0.0291, + "step": 11004 + }, + { + "epoch": 0.25797453072897436, + "grad_norm": 3.9035935401916504, + "learning_rate": 0.00016895643014242434, + "loss": 0.4436, + "step": 11005 + }, + { + "epoch": 0.25799797230377935, + "grad_norm": 1.667365550994873, + "learning_rate": 0.00016895109520364236, + "loss": 0.2427, + "step": 11006 + }, + { + "epoch": 0.25802141387858435, + "grad_norm": 3.5321450233459473, + "learning_rate": 0.0001689457598907298, + "loss": 0.3015, + "step": 11007 + }, + { + "epoch": 0.25804485545338934, + "grad_norm": 4.914587020874023, + "learning_rate": 0.00016894042420371563, + "loss": 0.4749, + "step": 11008 + }, + { + "epoch": 0.25806829702819434, + "grad_norm": 5.946348190307617, + "learning_rate": 0.00016893508814262874, + "loss": 0.5359, + "step": 11009 + }, + { + "epoch": 0.25809173860299933, + "grad_norm": 4.789889812469482, + "learning_rate": 0.00016892975170749815, + "loss": 0.2656, + "step": 11010 + }, + { + "epoch": 0.2581151801778043, + "grad_norm": 4.928236484527588, + "learning_rate": 0.00016892441489835277, + "loss": 0.6441, + "step": 11011 + }, + { + "epoch": 0.2581386217526093, + "grad_norm": 1.5057893991470337, + "learning_rate": 0.0001689190777152216, + "loss": 0.233, + "step": 11012 + }, + { + "epoch": 0.2581620633274143, + "grad_norm": 2.209590196609497, + "learning_rate": 0.0001689137401581336, + "loss": 0.2453, + "step": 11013 + }, + { + "epoch": 0.2581855049022193, + "grad_norm": 0.9651123881340027, + "learning_rate": 0.00016890840222711766, + "loss": 0.0644, + "step": 11014 + }, + { + "epoch": 0.2582089464770243, + "grad_norm": 2.3625340461730957, + "learning_rate": 0.0001689030639222028, + "loss": 0.1713, + "step": 11015 + }, + { + "epoch": 0.2582323880518293, + "grad_norm": 3.9339752197265625, + "learning_rate": 0.000168897725243418, + "loss": 0.4292, + "step": 11016 + }, + { + "epoch": 0.2582558296266343, + "grad_norm": 0.550619900226593, + "learning_rate": 0.00016889238619079215, + "loss": 0.045, + "step": 11017 + }, + { + "epoch": 0.2582792712014393, + "grad_norm": 2.530076742172241, + "learning_rate": 0.00016888704676435432, + "loss": 1.3224, + "step": 11018 + }, + { + "epoch": 0.2583027127762443, + "grad_norm": 2.6762876510620117, + "learning_rate": 0.00016888170696413346, + "loss": 0.1751, + "step": 11019 + }, + { + "epoch": 0.25832615435104933, + "grad_norm": 3.171308755874634, + "learning_rate": 0.0001688763667901585, + "loss": 0.3848, + "step": 11020 + }, + { + "epoch": 0.2583495959258543, + "grad_norm": 1.3143072128295898, + "learning_rate": 0.00016887102624245842, + "loss": 0.2245, + "step": 11021 + }, + { + "epoch": 0.2583730375006593, + "grad_norm": 4.171568870544434, + "learning_rate": 0.00016886568532106222, + "loss": 0.5265, + "step": 11022 + }, + { + "epoch": 0.2583964790754643, + "grad_norm": 4.720742225646973, + "learning_rate": 0.00016886034402599888, + "loss": 0.6039, + "step": 11023 + }, + { + "epoch": 0.2584199206502693, + "grad_norm": 1.8149305582046509, + "learning_rate": 0.0001688550023572974, + "loss": 0.0933, + "step": 11024 + }, + { + "epoch": 0.2584433622250743, + "grad_norm": 3.5793614387512207, + "learning_rate": 0.0001688496603149867, + "loss": 0.2736, + "step": 11025 + }, + { + "epoch": 0.2584668037998793, + "grad_norm": 3.1150403022766113, + "learning_rate": 0.00016884431789909585, + "loss": 0.2298, + "step": 11026 + }, + { + "epoch": 0.2584902453746843, + "grad_norm": 0.40749409794807434, + "learning_rate": 0.00016883897510965376, + "loss": 0.0519, + "step": 11027 + }, + { + "epoch": 0.2585136869494893, + "grad_norm": 2.227969169616699, + "learning_rate": 0.00016883363194668944, + "loss": 0.2474, + "step": 11028 + }, + { + "epoch": 0.2585371285242943, + "grad_norm": 4.948520660400391, + "learning_rate": 0.00016882828841023194, + "loss": 0.4332, + "step": 11029 + }, + { + "epoch": 0.2585605700990993, + "grad_norm": 6.533908843994141, + "learning_rate": 0.00016882294450031022, + "loss": 0.7117, + "step": 11030 + }, + { + "epoch": 0.25858401167390427, + "grad_norm": 3.8031415939331055, + "learning_rate": 0.00016881760021695324, + "loss": 0.4176, + "step": 11031 + }, + { + "epoch": 0.25860745324870926, + "grad_norm": 3.83647084236145, + "learning_rate": 0.00016881225556019004, + "loss": 0.3612, + "step": 11032 + }, + { + "epoch": 0.25863089482351426, + "grad_norm": 3.3827271461486816, + "learning_rate": 0.0001688069105300496, + "loss": 0.4184, + "step": 11033 + }, + { + "epoch": 0.25865433639831925, + "grad_norm": 3.640986680984497, + "learning_rate": 0.00016880156512656092, + "loss": 0.4899, + "step": 11034 + }, + { + "epoch": 0.25867777797312425, + "grad_norm": 3.409520149230957, + "learning_rate": 0.00016879621934975302, + "loss": 0.4348, + "step": 11035 + }, + { + "epoch": 0.25870121954792924, + "grad_norm": 4.982442378997803, + "learning_rate": 0.0001687908731996549, + "loss": 0.4416, + "step": 11036 + }, + { + "epoch": 0.25872466112273423, + "grad_norm": 3.750357151031494, + "learning_rate": 0.00016878552667629556, + "loss": 0.4851, + "step": 11037 + }, + { + "epoch": 0.25874810269753923, + "grad_norm": 6.644514560699463, + "learning_rate": 0.00016878017977970403, + "loss": 0.7446, + "step": 11038 + }, + { + "epoch": 0.2587715442723442, + "grad_norm": 6.230194091796875, + "learning_rate": 0.0001687748325099093, + "loss": 0.7147, + "step": 11039 + }, + { + "epoch": 0.2587949858471492, + "grad_norm": 1.0543088912963867, + "learning_rate": 0.00016876948486694042, + "loss": 0.0946, + "step": 11040 + }, + { + "epoch": 0.2588184274219542, + "grad_norm": 3.5846729278564453, + "learning_rate": 0.00016876413685082634, + "loss": 0.3301, + "step": 11041 + }, + { + "epoch": 0.2588418689967592, + "grad_norm": 3.2106871604919434, + "learning_rate": 0.00016875878846159617, + "loss": 0.2058, + "step": 11042 + }, + { + "epoch": 0.2588653105715642, + "grad_norm": 4.097505569458008, + "learning_rate": 0.00016875343969927883, + "loss": 0.385, + "step": 11043 + }, + { + "epoch": 0.2588887521463692, + "grad_norm": 1.9228847026824951, + "learning_rate": 0.00016874809056390344, + "loss": 0.2142, + "step": 11044 + }, + { + "epoch": 0.2589121937211742, + "grad_norm": 0.5636836886405945, + "learning_rate": 0.0001687427410554989, + "loss": 0.0617, + "step": 11045 + }, + { + "epoch": 0.2589356352959792, + "grad_norm": 5.615505218505859, + "learning_rate": 0.0001687373911740944, + "loss": 0.8421, + "step": 11046 + }, + { + "epoch": 0.2589590768707842, + "grad_norm": 0.682884693145752, + "learning_rate": 0.00016873204091971884, + "loss": 0.0436, + "step": 11047 + }, + { + "epoch": 0.25898251844558917, + "grad_norm": 3.5855088233947754, + "learning_rate": 0.00016872669029240132, + "loss": 0.3475, + "step": 11048 + }, + { + "epoch": 0.25900596002039417, + "grad_norm": 2.682522773742676, + "learning_rate": 0.00016872133929217084, + "loss": 0.2856, + "step": 11049 + }, + { + "epoch": 0.25902940159519916, + "grad_norm": 3.7518417835235596, + "learning_rate": 0.00016871598791905643, + "loss": 0.2723, + "step": 11050 + }, + { + "epoch": 0.25905284317000415, + "grad_norm": 1.5461657047271729, + "learning_rate": 0.00016871063617308714, + "loss": 0.1292, + "step": 11051 + }, + { + "epoch": 0.25907628474480915, + "grad_norm": 4.132984638214111, + "learning_rate": 0.000168705284054292, + "loss": 0.4794, + "step": 11052 + }, + { + "epoch": 0.25909972631961414, + "grad_norm": 3.7273049354553223, + "learning_rate": 0.00016869993156270005, + "loss": 0.4161, + "step": 11053 + }, + { + "epoch": 0.25912316789441914, + "grad_norm": 3.1064867973327637, + "learning_rate": 0.00016869457869834035, + "loss": 0.4763, + "step": 11054 + }, + { + "epoch": 0.25914660946922413, + "grad_norm": 7.261490345001221, + "learning_rate": 0.00016868922546124196, + "loss": 1.1005, + "step": 11055 + }, + { + "epoch": 0.2591700510440291, + "grad_norm": 1.8216367959976196, + "learning_rate": 0.00016868387185143386, + "loss": 0.1899, + "step": 11056 + }, + { + "epoch": 0.2591934926188341, + "grad_norm": 4.225410461425781, + "learning_rate": 0.00016867851786894517, + "loss": 0.3901, + "step": 11057 + }, + { + "epoch": 0.2592169341936391, + "grad_norm": 1.9944863319396973, + "learning_rate": 0.0001686731635138049, + "loss": 1.0405, + "step": 11058 + }, + { + "epoch": 0.2592403757684441, + "grad_norm": 6.801360607147217, + "learning_rate": 0.0001686678087860421, + "loss": 0.8835, + "step": 11059 + }, + { + "epoch": 0.2592638173432491, + "grad_norm": 1.579161524772644, + "learning_rate": 0.00016866245368568584, + "loss": 0.1702, + "step": 11060 + }, + { + "epoch": 0.2592872589180541, + "grad_norm": 4.357707500457764, + "learning_rate": 0.00016865709821276519, + "loss": 0.5009, + "step": 11061 + }, + { + "epoch": 0.2593107004928591, + "grad_norm": 1.0506179332733154, + "learning_rate": 0.0001686517423673092, + "loss": 0.1057, + "step": 11062 + }, + { + "epoch": 0.2593341420676641, + "grad_norm": 2.615570306777954, + "learning_rate": 0.00016864638614934693, + "loss": 1.5359, + "step": 11063 + }, + { + "epoch": 0.2593575836424691, + "grad_norm": 2.868037223815918, + "learning_rate": 0.0001686410295589074, + "loss": 0.3014, + "step": 11064 + }, + { + "epoch": 0.2593810252172741, + "grad_norm": 1.9505164623260498, + "learning_rate": 0.00016863567259601975, + "loss": 1.5291, + "step": 11065 + }, + { + "epoch": 0.25940446679207907, + "grad_norm": 2.3285045623779297, + "learning_rate": 0.00016863031526071298, + "loss": 1.6051, + "step": 11066 + }, + { + "epoch": 0.25942790836688406, + "grad_norm": 4.639303207397461, + "learning_rate": 0.00016862495755301624, + "loss": 0.481, + "step": 11067 + }, + { + "epoch": 0.25945134994168906, + "grad_norm": 4.1145920753479, + "learning_rate": 0.0001686195994729585, + "loss": 0.4717, + "step": 11068 + }, + { + "epoch": 0.25947479151649405, + "grad_norm": 1.8552757501602173, + "learning_rate": 0.0001686142410205689, + "loss": 1.2514, + "step": 11069 + }, + { + "epoch": 0.25949823309129905, + "grad_norm": 3.0947606563568115, + "learning_rate": 0.0001686088821958765, + "loss": 0.308, + "step": 11070 + }, + { + "epoch": 0.2595216746661041, + "grad_norm": 0.23579131066799164, + "learning_rate": 0.00016860352299891037, + "loss": 0.0141, + "step": 11071 + }, + { + "epoch": 0.2595451162409091, + "grad_norm": 5.096927642822266, + "learning_rate": 0.00016859816342969962, + "loss": 0.732, + "step": 11072 + }, + { + "epoch": 0.2595685578157141, + "grad_norm": 1.1349624395370483, + "learning_rate": 0.0001685928034882733, + "loss": 0.1337, + "step": 11073 + }, + { + "epoch": 0.2595919993905191, + "grad_norm": 3.9617087841033936, + "learning_rate": 0.00016858744317466048, + "loss": 0.3572, + "step": 11074 + }, + { + "epoch": 0.2596154409653241, + "grad_norm": 1.9557923078536987, + "learning_rate": 0.00016858208248889028, + "loss": 0.1548, + "step": 11075 + }, + { + "epoch": 0.25963888254012907, + "grad_norm": 0.5751790404319763, + "learning_rate": 0.00016857672143099178, + "loss": 0.0471, + "step": 11076 + }, + { + "epoch": 0.25966232411493406, + "grad_norm": 3.669613838195801, + "learning_rate": 0.00016857136000099403, + "loss": 0.4667, + "step": 11077 + }, + { + "epoch": 0.25968576568973906, + "grad_norm": 4.354867458343506, + "learning_rate": 0.00016856599819892618, + "loss": 0.5636, + "step": 11078 + }, + { + "epoch": 0.25970920726454405, + "grad_norm": 3.0738487243652344, + "learning_rate": 0.00016856063602481726, + "loss": 1.9712, + "step": 11079 + }, + { + "epoch": 0.25973264883934905, + "grad_norm": 2.0428130626678467, + "learning_rate": 0.00016855527347869645, + "loss": 0.2495, + "step": 11080 + }, + { + "epoch": 0.25975609041415404, + "grad_norm": 2.7766807079315186, + "learning_rate": 0.00016854991056059276, + "loss": 0.1882, + "step": 11081 + }, + { + "epoch": 0.25977953198895903, + "grad_norm": 3.6226389408111572, + "learning_rate": 0.00016854454727053536, + "loss": 0.7546, + "step": 11082 + }, + { + "epoch": 0.25980297356376403, + "grad_norm": 2.94696044921875, + "learning_rate": 0.0001685391836085533, + "loss": 0.3185, + "step": 11083 + }, + { + "epoch": 0.259826415138569, + "grad_norm": 3.1630120277404785, + "learning_rate": 0.0001685338195746757, + "loss": 0.3387, + "step": 11084 + }, + { + "epoch": 0.259849856713374, + "grad_norm": 3.3106529712677, + "learning_rate": 0.00016852845516893166, + "loss": 0.2384, + "step": 11085 + }, + { + "epoch": 0.259873298288179, + "grad_norm": 2.668532133102417, + "learning_rate": 0.00016852309039135034, + "loss": 0.3707, + "step": 11086 + }, + { + "epoch": 0.259896739862984, + "grad_norm": 2.64019513130188, + "learning_rate": 0.00016851772524196078, + "loss": 0.207, + "step": 11087 + }, + { + "epoch": 0.259920181437789, + "grad_norm": 2.6793572902679443, + "learning_rate": 0.00016851235972079208, + "loss": 0.1449, + "step": 11088 + }, + { + "epoch": 0.259943623012594, + "grad_norm": 2.3210363388061523, + "learning_rate": 0.00016850699382787342, + "loss": 0.3105, + "step": 11089 + }, + { + "epoch": 0.259967064587399, + "grad_norm": 4.300725936889648, + "learning_rate": 0.0001685016275632339, + "loss": 0.438, + "step": 11090 + }, + { + "epoch": 0.259990506162204, + "grad_norm": 6.566486358642578, + "learning_rate": 0.0001684962609269026, + "loss": 0.7507, + "step": 11091 + }, + { + "epoch": 0.260013947737009, + "grad_norm": 4.7516608238220215, + "learning_rate": 0.00016849089391890868, + "loss": 0.4653, + "step": 11092 + }, + { + "epoch": 0.26003738931181397, + "grad_norm": 2.5593974590301514, + "learning_rate": 0.00016848552653928123, + "loss": 1.3361, + "step": 11093 + }, + { + "epoch": 0.26006083088661897, + "grad_norm": 4.2208709716796875, + "learning_rate": 0.0001684801587880494, + "loss": 0.5194, + "step": 11094 + }, + { + "epoch": 0.26008427246142396, + "grad_norm": 2.42668080329895, + "learning_rate": 0.0001684747906652423, + "loss": 0.193, + "step": 11095 + }, + { + "epoch": 0.26010771403622895, + "grad_norm": 1.447228193283081, + "learning_rate": 0.00016846942217088904, + "loss": 0.1692, + "step": 11096 + }, + { + "epoch": 0.26013115561103395, + "grad_norm": 4.061676502227783, + "learning_rate": 0.0001684640533050188, + "loss": 0.5649, + "step": 11097 + }, + { + "epoch": 0.26015459718583894, + "grad_norm": 2.4801511764526367, + "learning_rate": 0.00016845868406766066, + "loss": 0.2787, + "step": 11098 + }, + { + "epoch": 0.26017803876064394, + "grad_norm": 2.4849729537963867, + "learning_rate": 0.00016845331445884378, + "loss": 0.4046, + "step": 11099 + }, + { + "epoch": 0.26020148033544893, + "grad_norm": 0.6791114807128906, + "learning_rate": 0.0001684479444785973, + "loss": 0.0807, + "step": 11100 + }, + { + "epoch": 0.2602249219102539, + "grad_norm": 3.3867123126983643, + "learning_rate": 0.00016844257412695034, + "loss": 0.5462, + "step": 11101 + }, + { + "epoch": 0.2602483634850589, + "grad_norm": 4.810451984405518, + "learning_rate": 0.00016843720340393206, + "loss": 0.6306, + "step": 11102 + }, + { + "epoch": 0.2602718050598639, + "grad_norm": 3.198477029800415, + "learning_rate": 0.00016843183230957158, + "loss": 0.4166, + "step": 11103 + }, + { + "epoch": 0.2602952466346689, + "grad_norm": 2.3528285026550293, + "learning_rate": 0.00016842646084389804, + "loss": 0.2924, + "step": 11104 + }, + { + "epoch": 0.2603186882094739, + "grad_norm": 3.8504419326782227, + "learning_rate": 0.00016842108900694063, + "loss": 0.393, + "step": 11105 + }, + { + "epoch": 0.2603421297842789, + "grad_norm": 4.018064498901367, + "learning_rate": 0.00016841571679872844, + "loss": 0.3393, + "step": 11106 + }, + { + "epoch": 0.2603655713590839, + "grad_norm": 0.35928621888160706, + "learning_rate": 0.00016841034421929065, + "loss": 0.0451, + "step": 11107 + }, + { + "epoch": 0.2603890129338889, + "grad_norm": 3.212064743041992, + "learning_rate": 0.0001684049712686564, + "loss": 1.963, + "step": 11108 + }, + { + "epoch": 0.2604124545086939, + "grad_norm": 5.283070087432861, + "learning_rate": 0.00016839959794685488, + "loss": 0.7364, + "step": 11109 + }, + { + "epoch": 0.2604358960834989, + "grad_norm": 1.3187260627746582, + "learning_rate": 0.00016839422425391523, + "loss": 0.136, + "step": 11110 + }, + { + "epoch": 0.26045933765830387, + "grad_norm": 1.1420079469680786, + "learning_rate": 0.00016838885018986656, + "loss": 0.0929, + "step": 11111 + }, + { + "epoch": 0.26048277923310886, + "grad_norm": 3.533684253692627, + "learning_rate": 0.0001683834757547381, + "loss": 0.4637, + "step": 11112 + }, + { + "epoch": 0.26050622080791386, + "grad_norm": 3.8342084884643555, + "learning_rate": 0.00016837810094855895, + "loss": 0.5293, + "step": 11113 + }, + { + "epoch": 0.26052966238271885, + "grad_norm": 5.051239490509033, + "learning_rate": 0.0001683727257713583, + "loss": 0.7016, + "step": 11114 + }, + { + "epoch": 0.26055310395752385, + "grad_norm": 0.9574806690216064, + "learning_rate": 0.00016836735022316532, + "loss": 0.1788, + "step": 11115 + }, + { + "epoch": 0.26057654553232884, + "grad_norm": 1.7650238275527954, + "learning_rate": 0.0001683619743040092, + "loss": 0.1073, + "step": 11116 + }, + { + "epoch": 0.26059998710713383, + "grad_norm": 3.9171669483184814, + "learning_rate": 0.00016835659801391905, + "loss": 0.5162, + "step": 11117 + }, + { + "epoch": 0.26062342868193883, + "grad_norm": 1.3016324043273926, + "learning_rate": 0.0001683512213529241, + "loss": 0.1461, + "step": 11118 + }, + { + "epoch": 0.2606468702567438, + "grad_norm": 0.8918131589889526, + "learning_rate": 0.0001683458443210535, + "loss": 0.0903, + "step": 11119 + }, + { + "epoch": 0.2606703118315488, + "grad_norm": 1.2185924053192139, + "learning_rate": 0.0001683404669183364, + "loss": 0.1473, + "step": 11120 + }, + { + "epoch": 0.2606937534063538, + "grad_norm": 4.0610270500183105, + "learning_rate": 0.00016833508914480203, + "loss": 0.5017, + "step": 11121 + }, + { + "epoch": 0.26071719498115886, + "grad_norm": 3.3118908405303955, + "learning_rate": 0.00016832971100047957, + "loss": 1.6601, + "step": 11122 + }, + { + "epoch": 0.26074063655596386, + "grad_norm": 4.4513163566589355, + "learning_rate": 0.00016832433248539812, + "loss": 0.4934, + "step": 11123 + }, + { + "epoch": 0.26076407813076885, + "grad_norm": 2.601754665374756, + "learning_rate": 0.00016831895359958695, + "loss": 1.8716, + "step": 11124 + }, + { + "epoch": 0.26078751970557384, + "grad_norm": 3.301736831665039, + "learning_rate": 0.00016831357434307526, + "loss": 0.4061, + "step": 11125 + }, + { + "epoch": 0.26081096128037884, + "grad_norm": 2.812753200531006, + "learning_rate": 0.0001683081947158921, + "loss": 0.2906, + "step": 11126 + }, + { + "epoch": 0.26083440285518383, + "grad_norm": 3.0366175174713135, + "learning_rate": 0.00016830281471806683, + "loss": 0.2307, + "step": 11127 + }, + { + "epoch": 0.2608578444299888, + "grad_norm": 1.1542515754699707, + "learning_rate": 0.00016829743434962853, + "loss": 0.1115, + "step": 11128 + }, + { + "epoch": 0.2608812860047938, + "grad_norm": 4.100831508636475, + "learning_rate": 0.00016829205361060647, + "loss": 0.4961, + "step": 11129 + }, + { + "epoch": 0.2609047275795988, + "grad_norm": 7.322399139404297, + "learning_rate": 0.0001682866725010298, + "loss": 0.7656, + "step": 11130 + }, + { + "epoch": 0.2609281691544038, + "grad_norm": 0.9674264192581177, + "learning_rate": 0.00016828129102092768, + "loss": 0.14, + "step": 11131 + }, + { + "epoch": 0.2609516107292088, + "grad_norm": 1.1992696523666382, + "learning_rate": 0.0001682759091703294, + "loss": 0.0728, + "step": 11132 + }, + { + "epoch": 0.2609750523040138, + "grad_norm": 3.638275623321533, + "learning_rate": 0.00016827052694926405, + "loss": 0.4166, + "step": 11133 + }, + { + "epoch": 0.2609984938788188, + "grad_norm": 4.919182777404785, + "learning_rate": 0.00016826514435776098, + "loss": 0.3094, + "step": 11134 + }, + { + "epoch": 0.2610219354536238, + "grad_norm": 2.354595184326172, + "learning_rate": 0.00016825976139584929, + "loss": 0.2121, + "step": 11135 + }, + { + "epoch": 0.2610453770284288, + "grad_norm": 1.3325307369232178, + "learning_rate": 0.0001682543780635582, + "loss": 0.1497, + "step": 11136 + }, + { + "epoch": 0.2610688186032338, + "grad_norm": 3.033045530319214, + "learning_rate": 0.00016824899436091695, + "loss": 0.2935, + "step": 11137 + }, + { + "epoch": 0.26109226017803877, + "grad_norm": 4.53505277633667, + "learning_rate": 0.00016824361028795472, + "loss": 0.6575, + "step": 11138 + }, + { + "epoch": 0.26111570175284377, + "grad_norm": 5.323051452636719, + "learning_rate": 0.00016823822584470074, + "loss": 0.6868, + "step": 11139 + }, + { + "epoch": 0.26113914332764876, + "grad_norm": 0.8363083600997925, + "learning_rate": 0.0001682328410311843, + "loss": 0.0869, + "step": 11140 + }, + { + "epoch": 0.26116258490245375, + "grad_norm": 5.516524791717529, + "learning_rate": 0.00016822745584743447, + "loss": 0.5178, + "step": 11141 + }, + { + "epoch": 0.26118602647725875, + "grad_norm": 2.837552547454834, + "learning_rate": 0.00016822207029348055, + "loss": 0.2093, + "step": 11142 + }, + { + "epoch": 0.26120946805206374, + "grad_norm": 0.45194658637046814, + "learning_rate": 0.00016821668436935177, + "loss": 0.0464, + "step": 11143 + }, + { + "epoch": 0.26123290962686874, + "grad_norm": 2.7307610511779785, + "learning_rate": 0.00016821129807507735, + "loss": 0.3299, + "step": 11144 + }, + { + "epoch": 0.26125635120167373, + "grad_norm": 3.7239036560058594, + "learning_rate": 0.00016820591141068652, + "loss": 0.2404, + "step": 11145 + }, + { + "epoch": 0.2612797927764787, + "grad_norm": 4.013222694396973, + "learning_rate": 0.00016820052437620847, + "loss": 0.4713, + "step": 11146 + }, + { + "epoch": 0.2613032343512837, + "grad_norm": 4.789562702178955, + "learning_rate": 0.00016819513697167247, + "loss": 0.4685, + "step": 11147 + }, + { + "epoch": 0.2613266759260887, + "grad_norm": 4.8545732498168945, + "learning_rate": 0.00016818974919710774, + "loss": 0.5002, + "step": 11148 + }, + { + "epoch": 0.2613501175008937, + "grad_norm": 5.2062153816223145, + "learning_rate": 0.00016818436105254354, + "loss": 0.6237, + "step": 11149 + }, + { + "epoch": 0.2613735590756987, + "grad_norm": 5.181457996368408, + "learning_rate": 0.00016817897253800903, + "loss": 0.6495, + "step": 11150 + }, + { + "epoch": 0.2613970006505037, + "grad_norm": 3.376127243041992, + "learning_rate": 0.0001681735836535335, + "loss": 0.3389, + "step": 11151 + }, + { + "epoch": 0.2614204422253087, + "grad_norm": 3.9693071842193604, + "learning_rate": 0.00016816819439914623, + "loss": 0.4218, + "step": 11152 + }, + { + "epoch": 0.2614438838001137, + "grad_norm": 3.969653844833374, + "learning_rate": 0.0001681628047748764, + "loss": 0.3453, + "step": 11153 + }, + { + "epoch": 0.2614673253749187, + "grad_norm": 3.771509885787964, + "learning_rate": 0.00016815741478075328, + "loss": 0.6081, + "step": 11154 + }, + { + "epoch": 0.2614907669497237, + "grad_norm": 4.531705856323242, + "learning_rate": 0.00016815202441680611, + "loss": 0.3538, + "step": 11155 + }, + { + "epoch": 0.26151420852452867, + "grad_norm": 3.2548024654388428, + "learning_rate": 0.00016814663368306412, + "loss": 0.2774, + "step": 11156 + }, + { + "epoch": 0.26153765009933366, + "grad_norm": 4.398310661315918, + "learning_rate": 0.00016814124257955658, + "loss": 0.3327, + "step": 11157 + }, + { + "epoch": 0.26156109167413866, + "grad_norm": 3.68890643119812, + "learning_rate": 0.0001681358511063128, + "loss": 1.4335, + "step": 11158 + }, + { + "epoch": 0.26158453324894365, + "grad_norm": 5.761255741119385, + "learning_rate": 0.0001681304592633619, + "loss": 0.6435, + "step": 11159 + }, + { + "epoch": 0.26160797482374865, + "grad_norm": 3.3166890144348145, + "learning_rate": 0.00016812506705073325, + "loss": 0.446, + "step": 11160 + }, + { + "epoch": 0.26163141639855364, + "grad_norm": 1.163335919380188, + "learning_rate": 0.00016811967446845607, + "loss": 0.1358, + "step": 11161 + }, + { + "epoch": 0.26165485797335863, + "grad_norm": 1.3082066774368286, + "learning_rate": 0.00016811428151655963, + "loss": 0.1574, + "step": 11162 + }, + { + "epoch": 0.26167829954816363, + "grad_norm": 1.1967964172363281, + "learning_rate": 0.00016810888819507316, + "loss": 0.169, + "step": 11163 + }, + { + "epoch": 0.2617017411229686, + "grad_norm": 3.9756546020507812, + "learning_rate": 0.00016810349450402593, + "loss": 0.3284, + "step": 11164 + }, + { + "epoch": 0.2617251826977736, + "grad_norm": 4.330644130706787, + "learning_rate": 0.00016809810044344728, + "loss": 0.4609, + "step": 11165 + }, + { + "epoch": 0.2617486242725786, + "grad_norm": 5.577029228210449, + "learning_rate": 0.0001680927060133664, + "loss": 0.5863, + "step": 11166 + }, + { + "epoch": 0.2617720658473836, + "grad_norm": 1.3145499229431152, + "learning_rate": 0.00016808731121381257, + "loss": 0.1645, + "step": 11167 + }, + { + "epoch": 0.2617955074221886, + "grad_norm": 2.851594924926758, + "learning_rate": 0.00016808191604481505, + "loss": 0.8324, + "step": 11168 + }, + { + "epoch": 0.2618189489969936, + "grad_norm": 0.6636165976524353, + "learning_rate": 0.0001680765205064032, + "loss": 0.0545, + "step": 11169 + }, + { + "epoch": 0.2618423905717986, + "grad_norm": 2.6534032821655273, + "learning_rate": 0.0001680711245986062, + "loss": 0.4203, + "step": 11170 + }, + { + "epoch": 0.2618658321466036, + "grad_norm": 1.288886308670044, + "learning_rate": 0.00016806572832145337, + "loss": 0.1195, + "step": 11171 + }, + { + "epoch": 0.2618892737214086, + "grad_norm": 3.870230197906494, + "learning_rate": 0.00016806033167497396, + "loss": 0.5232, + "step": 11172 + }, + { + "epoch": 0.2619127152962136, + "grad_norm": 0.9719544053077698, + "learning_rate": 0.00016805493465919732, + "loss": 0.2062, + "step": 11173 + }, + { + "epoch": 0.2619361568710186, + "grad_norm": 11.14246940612793, + "learning_rate": 0.00016804953727415266, + "loss": 0.9147, + "step": 11174 + }, + { + "epoch": 0.2619595984458236, + "grad_norm": 2.833695411682129, + "learning_rate": 0.0001680441395198693, + "loss": 0.3976, + "step": 11175 + }, + { + "epoch": 0.2619830400206286, + "grad_norm": 4.285853862762451, + "learning_rate": 0.00016803874139637658, + "loss": 0.2489, + "step": 11176 + }, + { + "epoch": 0.2620064815954336, + "grad_norm": 3.6961960792541504, + "learning_rate": 0.00016803334290370366, + "loss": 1.5369, + "step": 11177 + }, + { + "epoch": 0.2620299231702386, + "grad_norm": 4.345904350280762, + "learning_rate": 0.00016802794404187994, + "loss": 0.5267, + "step": 11178 + }, + { + "epoch": 0.2620533647450436, + "grad_norm": 4.346803188323975, + "learning_rate": 0.00016802254481093469, + "loss": 0.6472, + "step": 11179 + }, + { + "epoch": 0.2620768063198486, + "grad_norm": 3.1650784015655518, + "learning_rate": 0.00016801714521089717, + "loss": 0.144, + "step": 11180 + }, + { + "epoch": 0.2621002478946536, + "grad_norm": 1.050139307975769, + "learning_rate": 0.00016801174524179671, + "loss": 0.0847, + "step": 11181 + }, + { + "epoch": 0.2621236894694586, + "grad_norm": 3.1514039039611816, + "learning_rate": 0.00016800634490366265, + "loss": 2.1512, + "step": 11182 + }, + { + "epoch": 0.26214713104426357, + "grad_norm": 4.696746349334717, + "learning_rate": 0.00016800094419652423, + "loss": 0.4758, + "step": 11183 + }, + { + "epoch": 0.26217057261906856, + "grad_norm": 4.193345546722412, + "learning_rate": 0.00016799554312041078, + "loss": 0.3224, + "step": 11184 + }, + { + "epoch": 0.26219401419387356, + "grad_norm": 4.78197717666626, + "learning_rate": 0.0001679901416753516, + "loss": 0.4007, + "step": 11185 + }, + { + "epoch": 0.26221745576867855, + "grad_norm": 4.266528129577637, + "learning_rate": 0.00016798473986137598, + "loss": 0.4105, + "step": 11186 + }, + { + "epoch": 0.26224089734348355, + "grad_norm": 2.2339675426483154, + "learning_rate": 0.00016797933767851326, + "loss": 1.0956, + "step": 11187 + }, + { + "epoch": 0.26226433891828854, + "grad_norm": 4.402099132537842, + "learning_rate": 0.00016797393512679273, + "loss": 0.4328, + "step": 11188 + }, + { + "epoch": 0.26228778049309354, + "grad_norm": 5.446609973907471, + "learning_rate": 0.00016796853220624373, + "loss": 0.5083, + "step": 11189 + }, + { + "epoch": 0.26231122206789853, + "grad_norm": 5.424993515014648, + "learning_rate": 0.00016796312891689558, + "loss": 0.7195, + "step": 11190 + }, + { + "epoch": 0.2623346636427035, + "grad_norm": 2.3916614055633545, + "learning_rate": 0.00016795772525877759, + "loss": 1.0362, + "step": 11191 + }, + { + "epoch": 0.2623581052175085, + "grad_norm": 5.472197532653809, + "learning_rate": 0.00016795232123191902, + "loss": 0.4883, + "step": 11192 + }, + { + "epoch": 0.2623815467923135, + "grad_norm": 3.2190101146698, + "learning_rate": 0.00016794691683634925, + "loss": 0.4212, + "step": 11193 + }, + { + "epoch": 0.2624049883671185, + "grad_norm": 3.593409538269043, + "learning_rate": 0.00016794151207209765, + "loss": 0.3718, + "step": 11194 + }, + { + "epoch": 0.2624284299419235, + "grad_norm": 4.547365188598633, + "learning_rate": 0.00016793610693919346, + "loss": 0.6264, + "step": 11195 + }, + { + "epoch": 0.2624518715167285, + "grad_norm": 0.8618916273117065, + "learning_rate": 0.00016793070143766606, + "loss": 0.1251, + "step": 11196 + }, + { + "epoch": 0.2624753130915335, + "grad_norm": 1.0494602918624878, + "learning_rate": 0.00016792529556754478, + "loss": 0.1524, + "step": 11197 + }, + { + "epoch": 0.2624987546663385, + "grad_norm": 2.380138874053955, + "learning_rate": 0.0001679198893288589, + "loss": 0.1288, + "step": 11198 + }, + { + "epoch": 0.2625221962411435, + "grad_norm": 1.714882493019104, + "learning_rate": 0.00016791448272163777, + "loss": 0.1132, + "step": 11199 + }, + { + "epoch": 0.2625456378159485, + "grad_norm": 2.748016834259033, + "learning_rate": 0.00016790907574591079, + "loss": 0.4426, + "step": 11200 + }, + { + "epoch": 0.26256907939075347, + "grad_norm": 2.756174087524414, + "learning_rate": 0.00016790366840170723, + "loss": 1.8722, + "step": 11201 + }, + { + "epoch": 0.26259252096555846, + "grad_norm": 3.452531099319458, + "learning_rate": 0.0001678982606890565, + "loss": 0.5808, + "step": 11202 + }, + { + "epoch": 0.26261596254036346, + "grad_norm": 3.577582836151123, + "learning_rate": 0.00016789285260798787, + "loss": 0.4458, + "step": 11203 + }, + { + "epoch": 0.26263940411516845, + "grad_norm": 1.385758399963379, + "learning_rate": 0.0001678874441585307, + "loss": 0.1171, + "step": 11204 + }, + { + "epoch": 0.26266284568997345, + "grad_norm": 4.225988388061523, + "learning_rate": 0.00016788203534071435, + "loss": 0.5508, + "step": 11205 + }, + { + "epoch": 0.26268628726477844, + "grad_norm": 1.4060245752334595, + "learning_rate": 0.00016787662615456815, + "loss": 1.3438, + "step": 11206 + }, + { + "epoch": 0.26270972883958343, + "grad_norm": 4.565658092498779, + "learning_rate": 0.00016787121660012154, + "loss": 0.4795, + "step": 11207 + }, + { + "epoch": 0.26273317041438843, + "grad_norm": 3.728384017944336, + "learning_rate": 0.00016786580667740376, + "loss": 0.6173, + "step": 11208 + }, + { + "epoch": 0.2627566119891934, + "grad_norm": 4.101186752319336, + "learning_rate": 0.00016786039638644418, + "loss": 0.3955, + "step": 11209 + }, + { + "epoch": 0.2627800535639984, + "grad_norm": 1.1370351314544678, + "learning_rate": 0.0001678549857272722, + "loss": 0.1247, + "step": 11210 + }, + { + "epoch": 0.2628034951388034, + "grad_norm": 3.8792264461517334, + "learning_rate": 0.00016784957469991716, + "loss": 0.4359, + "step": 11211 + }, + { + "epoch": 0.2628269367136084, + "grad_norm": 1.4271351099014282, + "learning_rate": 0.0001678441633044084, + "loss": 0.1693, + "step": 11212 + }, + { + "epoch": 0.2628503782884134, + "grad_norm": 1.9902005195617676, + "learning_rate": 0.00016783875154077535, + "loss": 0.1484, + "step": 11213 + }, + { + "epoch": 0.2628738198632184, + "grad_norm": 2.017333745956421, + "learning_rate": 0.0001678333394090473, + "loss": 0.171, + "step": 11214 + }, + { + "epoch": 0.2628972614380234, + "grad_norm": 3.886845827102661, + "learning_rate": 0.00016782792690925365, + "loss": 0.3169, + "step": 11215 + }, + { + "epoch": 0.2629207030128284, + "grad_norm": 4.316601276397705, + "learning_rate": 0.00016782251404142375, + "loss": 0.4982, + "step": 11216 + }, + { + "epoch": 0.2629441445876334, + "grad_norm": 3.479069709777832, + "learning_rate": 0.00016781710080558698, + "loss": 0.5694, + "step": 11217 + }, + { + "epoch": 0.26296758616243837, + "grad_norm": 1.008195161819458, + "learning_rate": 0.00016781168720177272, + "loss": 0.1238, + "step": 11218 + }, + { + "epoch": 0.26299102773724337, + "grad_norm": 3.398691415786743, + "learning_rate": 0.00016780627323001034, + "loss": 0.2961, + "step": 11219 + }, + { + "epoch": 0.26301446931204836, + "grad_norm": 2.5703024864196777, + "learning_rate": 0.0001678008588903292, + "loss": 0.2228, + "step": 11220 + }, + { + "epoch": 0.26303791088685335, + "grad_norm": 2.5853114128112793, + "learning_rate": 0.00016779544418275873, + "loss": 0.2853, + "step": 11221 + }, + { + "epoch": 0.26306135246165835, + "grad_norm": 1.6850346326828003, + "learning_rate": 0.00016779002910732827, + "loss": 1.1874, + "step": 11222 + }, + { + "epoch": 0.26308479403646334, + "grad_norm": 3.940012216567993, + "learning_rate": 0.00016778461366406716, + "loss": 0.5418, + "step": 11223 + }, + { + "epoch": 0.26310823561126834, + "grad_norm": 6.498040676116943, + "learning_rate": 0.00016777919785300487, + "loss": 0.7447, + "step": 11224 + }, + { + "epoch": 0.2631316771860734, + "grad_norm": 1.562121868133545, + "learning_rate": 0.00016777378167417073, + "loss": 0.068, + "step": 11225 + }, + { + "epoch": 0.2631551187608784, + "grad_norm": 4.770116806030273, + "learning_rate": 0.00016776836512759415, + "loss": 0.4753, + "step": 11226 + }, + { + "epoch": 0.2631785603356834, + "grad_norm": 3.408480167388916, + "learning_rate": 0.00016776294821330452, + "loss": 0.278, + "step": 11227 + }, + { + "epoch": 0.26320200191048837, + "grad_norm": 3.5532186031341553, + "learning_rate": 0.0001677575309313312, + "loss": 0.4119, + "step": 11228 + }, + { + "epoch": 0.26322544348529336, + "grad_norm": 1.4115076065063477, + "learning_rate": 0.00016775211328170364, + "loss": 0.1383, + "step": 11229 + }, + { + "epoch": 0.26324888506009836, + "grad_norm": 5.18150520324707, + "learning_rate": 0.0001677466952644512, + "loss": 0.5656, + "step": 11230 + }, + { + "epoch": 0.26327232663490335, + "grad_norm": 5.74652099609375, + "learning_rate": 0.00016774127687960332, + "loss": 0.7887, + "step": 11231 + }, + { + "epoch": 0.26329576820970835, + "grad_norm": 3.9259979724884033, + "learning_rate": 0.00016773585812718934, + "loss": 0.3861, + "step": 11232 + }, + { + "epoch": 0.26331920978451334, + "grad_norm": 3.551698684692383, + "learning_rate": 0.00016773043900723865, + "loss": 0.4602, + "step": 11233 + }, + { + "epoch": 0.26334265135931834, + "grad_norm": 4.381046295166016, + "learning_rate": 0.00016772501951978073, + "loss": 0.6047, + "step": 11234 + }, + { + "epoch": 0.26336609293412333, + "grad_norm": 1.1494321823120117, + "learning_rate": 0.00016771959966484495, + "loss": 0.2268, + "step": 11235 + }, + { + "epoch": 0.2633895345089283, + "grad_norm": 2.0290162563323975, + "learning_rate": 0.00016771417944246071, + "loss": 1.26, + "step": 11236 + }, + { + "epoch": 0.2634129760837333, + "grad_norm": 1.88089120388031, + "learning_rate": 0.00016770875885265744, + "loss": 2.0494, + "step": 11237 + }, + { + "epoch": 0.2634364176585383, + "grad_norm": 3.3469784259796143, + "learning_rate": 0.00016770333789546454, + "loss": 0.3675, + "step": 11238 + }, + { + "epoch": 0.2634598592333433, + "grad_norm": 4.525656700134277, + "learning_rate": 0.00016769791657091142, + "loss": 0.6792, + "step": 11239 + }, + { + "epoch": 0.2634833008081483, + "grad_norm": 2.956454038619995, + "learning_rate": 0.0001676924948790275, + "loss": 0.3208, + "step": 11240 + }, + { + "epoch": 0.2635067423829533, + "grad_norm": 7.631062984466553, + "learning_rate": 0.0001676870728198422, + "loss": 0.7358, + "step": 11241 + }, + { + "epoch": 0.2635301839577583, + "grad_norm": 3.1180319786071777, + "learning_rate": 0.00016768165039338494, + "loss": 0.3482, + "step": 11242 + }, + { + "epoch": 0.2635536255325633, + "grad_norm": 5.345396518707275, + "learning_rate": 0.00016767622759968512, + "loss": 0.574, + "step": 11243 + }, + { + "epoch": 0.2635770671073683, + "grad_norm": 5.260560989379883, + "learning_rate": 0.00016767080443877224, + "loss": 0.73, + "step": 11244 + }, + { + "epoch": 0.2636005086821733, + "grad_norm": 4.50935697555542, + "learning_rate": 0.00016766538091067562, + "loss": 0.4872, + "step": 11245 + }, + { + "epoch": 0.26362395025697827, + "grad_norm": 0.9253526329994202, + "learning_rate": 0.0001676599570154248, + "loss": 0.1259, + "step": 11246 + }, + { + "epoch": 0.26364739183178326, + "grad_norm": 3.9277286529541016, + "learning_rate": 0.00016765453275304908, + "loss": 0.5172, + "step": 11247 + }, + { + "epoch": 0.26367083340658826, + "grad_norm": 3.7686314582824707, + "learning_rate": 0.00016764910812357798, + "loss": 0.4901, + "step": 11248 + }, + { + "epoch": 0.26369427498139325, + "grad_norm": 0.8251849412918091, + "learning_rate": 0.00016764368312704094, + "loss": 0.1115, + "step": 11249 + }, + { + "epoch": 0.26371771655619825, + "grad_norm": 4.571662902832031, + "learning_rate": 0.00016763825776346737, + "loss": 0.4794, + "step": 11250 + }, + { + "epoch": 0.26374115813100324, + "grad_norm": 1.1411364078521729, + "learning_rate": 0.00016763283203288673, + "loss": 0.1747, + "step": 11251 + }, + { + "epoch": 0.26376459970580823, + "grad_norm": 2.336256980895996, + "learning_rate": 0.00016762740593532843, + "loss": 0.2554, + "step": 11252 + }, + { + "epoch": 0.26378804128061323, + "grad_norm": 2.889045000076294, + "learning_rate": 0.0001676219794708219, + "loss": 0.4108, + "step": 11253 + }, + { + "epoch": 0.2638114828554182, + "grad_norm": 3.4819462299346924, + "learning_rate": 0.0001676165526393966, + "loss": 0.427, + "step": 11254 + }, + { + "epoch": 0.2638349244302232, + "grad_norm": 1.2322946786880493, + "learning_rate": 0.00016761112544108202, + "loss": 0.2611, + "step": 11255 + }, + { + "epoch": 0.2638583660050282, + "grad_norm": 3.858879804611206, + "learning_rate": 0.00016760569787590757, + "loss": 0.2718, + "step": 11256 + }, + { + "epoch": 0.2638818075798332, + "grad_norm": 1.1262555122375488, + "learning_rate": 0.0001676002699439027, + "loss": 0.1202, + "step": 11257 + }, + { + "epoch": 0.2639052491546382, + "grad_norm": 0.720244824886322, + "learning_rate": 0.00016759484164509685, + "loss": 0.0671, + "step": 11258 + }, + { + "epoch": 0.2639286907294432, + "grad_norm": 3.7417666912078857, + "learning_rate": 0.00016758941297951948, + "loss": 0.5849, + "step": 11259 + }, + { + "epoch": 0.2639521323042482, + "grad_norm": 1.6921793222427368, + "learning_rate": 0.00016758398394720007, + "loss": 0.1961, + "step": 11260 + }, + { + "epoch": 0.2639755738790532, + "grad_norm": 2.180525541305542, + "learning_rate": 0.00016757855454816806, + "loss": 0.3749, + "step": 11261 + }, + { + "epoch": 0.2639990154538582, + "grad_norm": 3.640331745147705, + "learning_rate": 0.00016757312478245291, + "loss": 0.7013, + "step": 11262 + }, + { + "epoch": 0.26402245702866317, + "grad_norm": 3.7735114097595215, + "learning_rate": 0.0001675676946500841, + "loss": 0.3463, + "step": 11263 + }, + { + "epoch": 0.26404589860346817, + "grad_norm": 4.211430549621582, + "learning_rate": 0.00016756226415109106, + "loss": 0.4927, + "step": 11264 + }, + { + "epoch": 0.26406934017827316, + "grad_norm": 3.065673589706421, + "learning_rate": 0.00016755683328550325, + "loss": 1.5916, + "step": 11265 + }, + { + "epoch": 0.26409278175307815, + "grad_norm": 2.6154673099517822, + "learning_rate": 0.00016755140205335022, + "loss": 0.1976, + "step": 11266 + }, + { + "epoch": 0.26411622332788315, + "grad_norm": 1.567014217376709, + "learning_rate": 0.00016754597045466136, + "loss": 0.1924, + "step": 11267 + }, + { + "epoch": 0.26413966490268814, + "grad_norm": 4.329446792602539, + "learning_rate": 0.00016754053848946616, + "loss": 0.5136, + "step": 11268 + }, + { + "epoch": 0.26416310647749314, + "grad_norm": 4.154630661010742, + "learning_rate": 0.0001675351061577941, + "loss": 0.4265, + "step": 11269 + }, + { + "epoch": 0.26418654805229813, + "grad_norm": 3.801506280899048, + "learning_rate": 0.00016752967345967467, + "loss": 0.3883, + "step": 11270 + }, + { + "epoch": 0.2642099896271031, + "grad_norm": 1.1503273248672485, + "learning_rate": 0.0001675242403951373, + "loss": 0.1152, + "step": 11271 + }, + { + "epoch": 0.2642334312019081, + "grad_norm": 0.5793562531471252, + "learning_rate": 0.00016751880696421154, + "loss": 0.0873, + "step": 11272 + }, + { + "epoch": 0.2642568727767131, + "grad_norm": 1.5305556058883667, + "learning_rate": 0.00016751337316692682, + "loss": 0.1896, + "step": 11273 + }, + { + "epoch": 0.2642803143515181, + "grad_norm": 4.788054943084717, + "learning_rate": 0.00016750793900331264, + "loss": 0.9188, + "step": 11274 + }, + { + "epoch": 0.2643037559263231, + "grad_norm": 1.5058724880218506, + "learning_rate": 0.00016750250447339848, + "loss": 0.9084, + "step": 11275 + }, + { + "epoch": 0.26432719750112815, + "grad_norm": 5.908663749694824, + "learning_rate": 0.00016749706957721385, + "loss": 0.721, + "step": 11276 + }, + { + "epoch": 0.26435063907593315, + "grad_norm": 4.010836601257324, + "learning_rate": 0.0001674916343147882, + "loss": 0.2946, + "step": 11277 + }, + { + "epoch": 0.26437408065073814, + "grad_norm": 2.201465129852295, + "learning_rate": 0.00016748619868615106, + "loss": 0.263, + "step": 11278 + }, + { + "epoch": 0.26439752222554314, + "grad_norm": 6.153519630432129, + "learning_rate": 0.0001674807626913319, + "loss": 0.4005, + "step": 11279 + }, + { + "epoch": 0.26442096380034813, + "grad_norm": 3.488053798675537, + "learning_rate": 0.00016747532633036021, + "loss": 0.4665, + "step": 11280 + }, + { + "epoch": 0.2644444053751531, + "grad_norm": 1.1103477478027344, + "learning_rate": 0.00016746988960326552, + "loss": 0.1553, + "step": 11281 + }, + { + "epoch": 0.2644678469499581, + "grad_norm": 0.5985179543495178, + "learning_rate": 0.00016746445251007732, + "loss": 0.0349, + "step": 11282 + }, + { + "epoch": 0.2644912885247631, + "grad_norm": 0.9107652902603149, + "learning_rate": 0.0001674590150508251, + "loss": 0.0729, + "step": 11283 + }, + { + "epoch": 0.2645147300995681, + "grad_norm": 4.538906574249268, + "learning_rate": 0.00016745357722553833, + "loss": 0.4029, + "step": 11284 + }, + { + "epoch": 0.2645381716743731, + "grad_norm": 4.897666931152344, + "learning_rate": 0.0001674481390342466, + "loss": 0.6042, + "step": 11285 + }, + { + "epoch": 0.2645616132491781, + "grad_norm": 3.1764228343963623, + "learning_rate": 0.00016744270047697936, + "loss": 1.5499, + "step": 11286 + }, + { + "epoch": 0.2645850548239831, + "grad_norm": 1.6555521488189697, + "learning_rate": 0.00016743726155376613, + "loss": 0.1312, + "step": 11287 + }, + { + "epoch": 0.2646084963987881, + "grad_norm": 2.6151208877563477, + "learning_rate": 0.0001674318222646364, + "loss": 0.4857, + "step": 11288 + }, + { + "epoch": 0.2646319379735931, + "grad_norm": 1.3085964918136597, + "learning_rate": 0.00016742638260961975, + "loss": 0.1785, + "step": 11289 + }, + { + "epoch": 0.2646553795483981, + "grad_norm": 3.507124662399292, + "learning_rate": 0.0001674209425887456, + "loss": 0.5159, + "step": 11290 + }, + { + "epoch": 0.26467882112320307, + "grad_norm": 3.9660542011260986, + "learning_rate": 0.00016741550220204356, + "loss": 1.6025, + "step": 11291 + }, + { + "epoch": 0.26470226269800806, + "grad_norm": 3.0474648475646973, + "learning_rate": 0.0001674100614495431, + "loss": 0.3852, + "step": 11292 + }, + { + "epoch": 0.26472570427281306, + "grad_norm": 3.5665972232818604, + "learning_rate": 0.00016740462033127375, + "loss": 0.4335, + "step": 11293 + }, + { + "epoch": 0.26474914584761805, + "grad_norm": 4.206067085266113, + "learning_rate": 0.00016739917884726503, + "loss": 0.3386, + "step": 11294 + }, + { + "epoch": 0.26477258742242304, + "grad_norm": 4.61043119430542, + "learning_rate": 0.0001673937369975465, + "loss": 1.613, + "step": 11295 + }, + { + "epoch": 0.26479602899722804, + "grad_norm": 2.756504774093628, + "learning_rate": 0.0001673882947821476, + "loss": 0.3446, + "step": 11296 + }, + { + "epoch": 0.26481947057203303, + "grad_norm": 2.6543524265289307, + "learning_rate": 0.00016738285220109796, + "loss": 1.7642, + "step": 11297 + }, + { + "epoch": 0.264842912146838, + "grad_norm": 0.48188483715057373, + "learning_rate": 0.00016737740925442705, + "loss": 0.0596, + "step": 11298 + }, + { + "epoch": 0.264866353721643, + "grad_norm": 3.0744941234588623, + "learning_rate": 0.00016737196594216445, + "loss": 0.2578, + "step": 11299 + }, + { + "epoch": 0.264889795296448, + "grad_norm": 5.182616710662842, + "learning_rate": 0.00016736652226433964, + "loss": 0.5779, + "step": 11300 + }, + { + "epoch": 0.264913236871253, + "grad_norm": 3.3239736557006836, + "learning_rate": 0.0001673610782209822, + "loss": 0.3691, + "step": 11301 + }, + { + "epoch": 0.264936678446058, + "grad_norm": 1.9016562700271606, + "learning_rate": 0.00016735563381212166, + "loss": 0.1895, + "step": 11302 + }, + { + "epoch": 0.264960120020863, + "grad_norm": 3.054351806640625, + "learning_rate": 0.00016735018903778751, + "loss": 0.437, + "step": 11303 + }, + { + "epoch": 0.264983561595668, + "grad_norm": 2.8701677322387695, + "learning_rate": 0.0001673447438980094, + "loss": 0.3636, + "step": 11304 + }, + { + "epoch": 0.265007003170473, + "grad_norm": 1.6295902729034424, + "learning_rate": 0.0001673392983928168, + "loss": 1.5548, + "step": 11305 + }, + { + "epoch": 0.265030444745278, + "grad_norm": 3.7572457790374756, + "learning_rate": 0.00016733385252223925, + "loss": 0.5324, + "step": 11306 + }, + { + "epoch": 0.265053886320083, + "grad_norm": 2.867645025253296, + "learning_rate": 0.00016732840628630634, + "loss": 0.1721, + "step": 11307 + }, + { + "epoch": 0.26507732789488797, + "grad_norm": 1.811562418937683, + "learning_rate": 0.0001673229596850476, + "loss": 0.2636, + "step": 11308 + }, + { + "epoch": 0.26510076946969297, + "grad_norm": 4.533321380615234, + "learning_rate": 0.00016731751271849258, + "loss": 0.8349, + "step": 11309 + }, + { + "epoch": 0.26512421104449796, + "grad_norm": 2.0138652324676514, + "learning_rate": 0.00016731206538667088, + "loss": 1.4102, + "step": 11310 + }, + { + "epoch": 0.26514765261930295, + "grad_norm": 1.3410019874572754, + "learning_rate": 0.000167306617689612, + "loss": 0.0935, + "step": 11311 + }, + { + "epoch": 0.26517109419410795, + "grad_norm": 3.3538010120391846, + "learning_rate": 0.00016730116962734553, + "loss": 0.276, + "step": 11312 + }, + { + "epoch": 0.26519453576891294, + "grad_norm": 3.4931914806365967, + "learning_rate": 0.00016729572119990102, + "loss": 0.2205, + "step": 11313 + }, + { + "epoch": 0.26521797734371794, + "grad_norm": 2.788646697998047, + "learning_rate": 0.00016729027240730798, + "loss": 1.7357, + "step": 11314 + }, + { + "epoch": 0.26524141891852293, + "grad_norm": 7.254907131195068, + "learning_rate": 0.0001672848232495961, + "loss": 0.6913, + "step": 11315 + }, + { + "epoch": 0.2652648604933279, + "grad_norm": 2.70596981048584, + "learning_rate": 0.00016727937372679484, + "loss": 1.6916, + "step": 11316 + }, + { + "epoch": 0.2652883020681329, + "grad_norm": 4.158058166503906, + "learning_rate": 0.0001672739238389338, + "loss": 0.619, + "step": 11317 + }, + { + "epoch": 0.2653117436429379, + "grad_norm": 0.675500214099884, + "learning_rate": 0.0001672684735860426, + "loss": 0.0916, + "step": 11318 + }, + { + "epoch": 0.2653351852177429, + "grad_norm": 3.040356159210205, + "learning_rate": 0.00016726302296815075, + "loss": 0.3813, + "step": 11319 + }, + { + "epoch": 0.2653586267925479, + "grad_norm": 3.8566012382507324, + "learning_rate": 0.00016725757198528783, + "loss": 0.4598, + "step": 11320 + }, + { + "epoch": 0.2653820683673529, + "grad_norm": 0.49508702754974365, + "learning_rate": 0.00016725212063748343, + "loss": 0.0549, + "step": 11321 + }, + { + "epoch": 0.2654055099421579, + "grad_norm": 3.1151697635650635, + "learning_rate": 0.00016724666892476716, + "loss": 0.2102, + "step": 11322 + }, + { + "epoch": 0.2654289515169629, + "grad_norm": 3.664588451385498, + "learning_rate": 0.00016724121684716855, + "loss": 0.402, + "step": 11323 + }, + { + "epoch": 0.2654523930917679, + "grad_norm": 7.7288360595703125, + "learning_rate": 0.00016723576440471724, + "loss": 0.401, + "step": 11324 + }, + { + "epoch": 0.2654758346665729, + "grad_norm": 3.6006693840026855, + "learning_rate": 0.00016723031159744276, + "loss": 0.4129, + "step": 11325 + }, + { + "epoch": 0.26549927624137787, + "grad_norm": 2.338822841644287, + "learning_rate": 0.0001672248584253747, + "loss": 0.2673, + "step": 11326 + }, + { + "epoch": 0.2655227178161829, + "grad_norm": 2.9510674476623535, + "learning_rate": 0.0001672194048885427, + "loss": 0.2729, + "step": 11327 + }, + { + "epoch": 0.2655461593909879, + "grad_norm": 2.775380849838257, + "learning_rate": 0.0001672139509869763, + "loss": 0.3239, + "step": 11328 + }, + { + "epoch": 0.2655696009657929, + "grad_norm": 5.921899318695068, + "learning_rate": 0.00016720849672070514, + "loss": 0.6311, + "step": 11329 + }, + { + "epoch": 0.2655930425405979, + "grad_norm": 1.8378422260284424, + "learning_rate": 0.00016720304208975875, + "loss": 0.1658, + "step": 11330 + }, + { + "epoch": 0.2656164841154029, + "grad_norm": 3.2923123836517334, + "learning_rate": 0.00016719758709416674, + "loss": 0.266, + "step": 11331 + }, + { + "epoch": 0.2656399256902079, + "grad_norm": 2.395059585571289, + "learning_rate": 0.0001671921317339588, + "loss": 0.2183, + "step": 11332 + }, + { + "epoch": 0.2656633672650129, + "grad_norm": 3.360440969467163, + "learning_rate": 0.0001671866760091644, + "loss": 1.4753, + "step": 11333 + }, + { + "epoch": 0.2656868088398179, + "grad_norm": 2.7532951831817627, + "learning_rate": 0.0001671812199198133, + "loss": 0.3941, + "step": 11334 + }, + { + "epoch": 0.2657102504146229, + "grad_norm": 3.471541404724121, + "learning_rate": 0.00016717576346593493, + "loss": 0.4079, + "step": 11335 + }, + { + "epoch": 0.26573369198942787, + "grad_norm": 1.3224129676818848, + "learning_rate": 0.000167170306647559, + "loss": 0.1237, + "step": 11336 + }, + { + "epoch": 0.26575713356423286, + "grad_norm": 1.3133823871612549, + "learning_rate": 0.00016716484946471512, + "loss": 0.1767, + "step": 11337 + }, + { + "epoch": 0.26578057513903786, + "grad_norm": 0.48046043515205383, + "learning_rate": 0.00016715939191743284, + "loss": 0.0668, + "step": 11338 + }, + { + "epoch": 0.26580401671384285, + "grad_norm": 4.626680374145508, + "learning_rate": 0.00016715393400574185, + "loss": 0.8539, + "step": 11339 + }, + { + "epoch": 0.26582745828864784, + "grad_norm": 3.454254627227783, + "learning_rate": 0.00016714847572967169, + "loss": 0.4343, + "step": 11340 + }, + { + "epoch": 0.26585089986345284, + "grad_norm": 5.995578765869141, + "learning_rate": 0.00016714301708925202, + "loss": 0.4473, + "step": 11341 + }, + { + "epoch": 0.26587434143825783, + "grad_norm": 4.517911434173584, + "learning_rate": 0.00016713755808451245, + "loss": 0.5595, + "step": 11342 + }, + { + "epoch": 0.2658977830130628, + "grad_norm": 1.4597218036651611, + "learning_rate": 0.0001671320987154826, + "loss": 0.2342, + "step": 11343 + }, + { + "epoch": 0.2659212245878678, + "grad_norm": 1.3086397647857666, + "learning_rate": 0.0001671266389821921, + "loss": 0.1156, + "step": 11344 + }, + { + "epoch": 0.2659446661626728, + "grad_norm": 4.264547824859619, + "learning_rate": 0.0001671211788846706, + "loss": 0.5428, + "step": 11345 + }, + { + "epoch": 0.2659681077374778, + "grad_norm": 1.0107485055923462, + "learning_rate": 0.00016711571842294767, + "loss": 0.1333, + "step": 11346 + }, + { + "epoch": 0.2659915493122828, + "grad_norm": 3.650139808654785, + "learning_rate": 0.00016711025759705297, + "loss": 0.5718, + "step": 11347 + }, + { + "epoch": 0.2660149908870878, + "grad_norm": 0.6822779774665833, + "learning_rate": 0.00016710479640701613, + "loss": 0.0578, + "step": 11348 + }, + { + "epoch": 0.2660384324618928, + "grad_norm": 2.191542625427246, + "learning_rate": 0.00016709933485286674, + "loss": 1.4805, + "step": 11349 + }, + { + "epoch": 0.2660618740366978, + "grad_norm": 3.72298264503479, + "learning_rate": 0.0001670938729346345, + "loss": 1.3753, + "step": 11350 + }, + { + "epoch": 0.2660853156115028, + "grad_norm": 2.616868495941162, + "learning_rate": 0.00016708841065234904, + "loss": 0.2145, + "step": 11351 + }, + { + "epoch": 0.2661087571863078, + "grad_norm": 3.0473151206970215, + "learning_rate": 0.00016708294800603995, + "loss": 0.3945, + "step": 11352 + }, + { + "epoch": 0.26613219876111277, + "grad_norm": 4.086672782897949, + "learning_rate": 0.00016707748499573692, + "loss": 0.4327, + "step": 11353 + }, + { + "epoch": 0.26615564033591776, + "grad_norm": 4.442220687866211, + "learning_rate": 0.00016707202162146955, + "loss": 0.615, + "step": 11354 + }, + { + "epoch": 0.26617908191072276, + "grad_norm": 5.699915409088135, + "learning_rate": 0.00016706655788326753, + "loss": 0.725, + "step": 11355 + }, + { + "epoch": 0.26620252348552775, + "grad_norm": 3.0132124423980713, + "learning_rate": 0.00016706109378116047, + "loss": 0.216, + "step": 11356 + }, + { + "epoch": 0.26622596506033275, + "grad_norm": 4.394841194152832, + "learning_rate": 0.00016705562931517798, + "loss": 0.4041, + "step": 11357 + }, + { + "epoch": 0.26624940663513774, + "grad_norm": 3.580371618270874, + "learning_rate": 0.00016705016448534981, + "loss": 0.4672, + "step": 11358 + }, + { + "epoch": 0.26627284820994274, + "grad_norm": 4.151409149169922, + "learning_rate": 0.00016704469929170558, + "loss": 0.4761, + "step": 11359 + }, + { + "epoch": 0.26629628978474773, + "grad_norm": 4.818532466888428, + "learning_rate": 0.0001670392337342749, + "loss": 0.3931, + "step": 11360 + }, + { + "epoch": 0.2663197313595527, + "grad_norm": 3.8577356338500977, + "learning_rate": 0.00016703376781308744, + "loss": 0.5178, + "step": 11361 + }, + { + "epoch": 0.2663431729343577, + "grad_norm": 2.322172164916992, + "learning_rate": 0.00016702830152817287, + "loss": 1.4172, + "step": 11362 + }, + { + "epoch": 0.2663666145091627, + "grad_norm": 5.289988994598389, + "learning_rate": 0.00016702283487956087, + "loss": 0.2966, + "step": 11363 + }, + { + "epoch": 0.2663900560839677, + "grad_norm": 1.104849100112915, + "learning_rate": 0.0001670173678672811, + "loss": 0.1211, + "step": 11364 + }, + { + "epoch": 0.2664134976587727, + "grad_norm": 3.72017765045166, + "learning_rate": 0.00016701190049136317, + "loss": 0.3898, + "step": 11365 + }, + { + "epoch": 0.2664369392335777, + "grad_norm": 3.281156539916992, + "learning_rate": 0.0001670064327518368, + "loss": 1.6196, + "step": 11366 + }, + { + "epoch": 0.2664603808083827, + "grad_norm": 2.2405331134796143, + "learning_rate": 0.00016700096464873166, + "loss": 0.1939, + "step": 11367 + }, + { + "epoch": 0.2664838223831877, + "grad_norm": 5.015546798706055, + "learning_rate": 0.00016699549618207738, + "loss": 0.5896, + "step": 11368 + }, + { + "epoch": 0.2665072639579927, + "grad_norm": 6.6228718757629395, + "learning_rate": 0.00016699002735190365, + "loss": 0.5552, + "step": 11369 + }, + { + "epoch": 0.2665307055327977, + "grad_norm": 2.883103847503662, + "learning_rate": 0.00016698455815824015, + "loss": 1.9457, + "step": 11370 + }, + { + "epoch": 0.26655414710760267, + "grad_norm": 2.5682499408721924, + "learning_rate": 0.00016697908860111658, + "loss": 0.2363, + "step": 11371 + }, + { + "epoch": 0.26657758868240766, + "grad_norm": 2.3184847831726074, + "learning_rate": 0.00016697361868056257, + "loss": 0.2391, + "step": 11372 + }, + { + "epoch": 0.26660103025721266, + "grad_norm": 1.755744218826294, + "learning_rate": 0.0001669681483966078, + "loss": 0.2046, + "step": 11373 + }, + { + "epoch": 0.26662447183201765, + "grad_norm": 4.386747360229492, + "learning_rate": 0.000166962677749282, + "loss": 0.6577, + "step": 11374 + }, + { + "epoch": 0.26664791340682265, + "grad_norm": 0.5202924013137817, + "learning_rate": 0.00016695720673861483, + "loss": 0.0351, + "step": 11375 + }, + { + "epoch": 0.26667135498162764, + "grad_norm": 2.8228466510772705, + "learning_rate": 0.00016695173536463597, + "loss": 0.1846, + "step": 11376 + }, + { + "epoch": 0.26669479655643263, + "grad_norm": 3.4379560947418213, + "learning_rate": 0.00016694626362737512, + "loss": 0.3547, + "step": 11377 + }, + { + "epoch": 0.2667182381312377, + "grad_norm": 0.5492350459098816, + "learning_rate": 0.00016694079152686193, + "loss": 0.0512, + "step": 11378 + }, + { + "epoch": 0.2667416797060427, + "grad_norm": 0.8171244263648987, + "learning_rate": 0.00016693531906312615, + "loss": 0.1303, + "step": 11379 + }, + { + "epoch": 0.2667651212808477, + "grad_norm": 2.965054512023926, + "learning_rate": 0.00016692984623619745, + "loss": 0.2437, + "step": 11380 + }, + { + "epoch": 0.26678856285565267, + "grad_norm": 3.622948408126831, + "learning_rate": 0.0001669243730461055, + "loss": 0.2212, + "step": 11381 + }, + { + "epoch": 0.26681200443045766, + "grad_norm": 5.0112175941467285, + "learning_rate": 0.00016691889949288003, + "loss": 0.3893, + "step": 11382 + }, + { + "epoch": 0.26683544600526266, + "grad_norm": 1.5445432662963867, + "learning_rate": 0.00016691342557655076, + "loss": 0.2028, + "step": 11383 + }, + { + "epoch": 0.26685888758006765, + "grad_norm": 6.000972270965576, + "learning_rate": 0.00016690795129714733, + "loss": 0.6888, + "step": 11384 + }, + { + "epoch": 0.26688232915487264, + "grad_norm": 0.8933048248291016, + "learning_rate": 0.00016690247665469946, + "loss": 0.0973, + "step": 11385 + }, + { + "epoch": 0.26690577072967764, + "grad_norm": 1.2695050239562988, + "learning_rate": 0.00016689700164923692, + "loss": 0.8219, + "step": 11386 + }, + { + "epoch": 0.26692921230448263, + "grad_norm": 4.299123287200928, + "learning_rate": 0.0001668915262807893, + "loss": 0.2782, + "step": 11387 + }, + { + "epoch": 0.2669526538792876, + "grad_norm": 5.901642799377441, + "learning_rate": 0.00016688605054938644, + "loss": 0.4047, + "step": 11388 + }, + { + "epoch": 0.2669760954540926, + "grad_norm": 2.093686819076538, + "learning_rate": 0.00016688057445505796, + "loss": 1.5956, + "step": 11389 + }, + { + "epoch": 0.2669995370288976, + "grad_norm": 3.182051420211792, + "learning_rate": 0.00016687509799783362, + "loss": 0.3776, + "step": 11390 + }, + { + "epoch": 0.2670229786037026, + "grad_norm": 1.701793909072876, + "learning_rate": 0.0001668696211777431, + "loss": 0.2118, + "step": 11391 + }, + { + "epoch": 0.2670464201785076, + "grad_norm": 2.0405032634735107, + "learning_rate": 0.00016686414399481615, + "loss": 0.1923, + "step": 11392 + }, + { + "epoch": 0.2670698617533126, + "grad_norm": 4.039942741394043, + "learning_rate": 0.00016685866644908245, + "loss": 0.4296, + "step": 11393 + }, + { + "epoch": 0.2670933033281176, + "grad_norm": 1.2432613372802734, + "learning_rate": 0.00016685318854057176, + "loss": 0.1596, + "step": 11394 + }, + { + "epoch": 0.2671167449029226, + "grad_norm": 2.257645606994629, + "learning_rate": 0.0001668477102693138, + "loss": 1.0592, + "step": 11395 + }, + { + "epoch": 0.2671401864777276, + "grad_norm": 2.840613603591919, + "learning_rate": 0.00016684223163533827, + "loss": 0.2167, + "step": 11396 + }, + { + "epoch": 0.2671636280525326, + "grad_norm": 2.2631492614746094, + "learning_rate": 0.00016683675263867492, + "loss": 0.3213, + "step": 11397 + }, + { + "epoch": 0.26718706962733757, + "grad_norm": 3.762122631072998, + "learning_rate": 0.00016683127327935345, + "loss": 0.4755, + "step": 11398 + }, + { + "epoch": 0.26721051120214256, + "grad_norm": 1.0902355909347534, + "learning_rate": 0.00016682579355740363, + "loss": 0.0453, + "step": 11399 + }, + { + "epoch": 0.26723395277694756, + "grad_norm": 1.8945401906967163, + "learning_rate": 0.00016682031347285518, + "loss": 0.3284, + "step": 11400 + }, + { + "epoch": 0.26725739435175255, + "grad_norm": 2.8972504138946533, + "learning_rate": 0.00016681483302573785, + "loss": 2.1018, + "step": 11401 + }, + { + "epoch": 0.26728083592655755, + "grad_norm": 3.7115440368652344, + "learning_rate": 0.00016680935221608133, + "loss": 0.3156, + "step": 11402 + }, + { + "epoch": 0.26730427750136254, + "grad_norm": 2.171055316925049, + "learning_rate": 0.0001668038710439154, + "loss": 0.1882, + "step": 11403 + }, + { + "epoch": 0.26732771907616754, + "grad_norm": 2.0490028858184814, + "learning_rate": 0.00016679838950926977, + "loss": 0.2516, + "step": 11404 + }, + { + "epoch": 0.26735116065097253, + "grad_norm": 0.9976841807365417, + "learning_rate": 0.0001667929076121742, + "loss": 0.088, + "step": 11405 + }, + { + "epoch": 0.2673746022257775, + "grad_norm": 6.0223708152771, + "learning_rate": 0.00016678742535265842, + "loss": 0.4002, + "step": 11406 + }, + { + "epoch": 0.2673980438005825, + "grad_norm": 3.008680582046509, + "learning_rate": 0.00016678194273075223, + "loss": 1.5955, + "step": 11407 + }, + { + "epoch": 0.2674214853753875, + "grad_norm": 2.6373515129089355, + "learning_rate": 0.00016677645974648534, + "loss": 0.1298, + "step": 11408 + }, + { + "epoch": 0.2674449269501925, + "grad_norm": 5.072791576385498, + "learning_rate": 0.0001667709763998875, + "loss": 0.5064, + "step": 11409 + }, + { + "epoch": 0.2674683685249975, + "grad_norm": 2.2673490047454834, + "learning_rate": 0.0001667654926909884, + "loss": 1.6206, + "step": 11410 + }, + { + "epoch": 0.2674918100998025, + "grad_norm": 4.111397743225098, + "learning_rate": 0.00016676000861981792, + "loss": 0.2453, + "step": 11411 + }, + { + "epoch": 0.2675152516746075, + "grad_norm": 1.2277801036834717, + "learning_rate": 0.00016675452418640576, + "loss": 0.1176, + "step": 11412 + }, + { + "epoch": 0.2675386932494125, + "grad_norm": 3.9165425300598145, + "learning_rate": 0.00016674903939078165, + "loss": 0.3145, + "step": 11413 + }, + { + "epoch": 0.2675621348242175, + "grad_norm": 3.7525346279144287, + "learning_rate": 0.0001667435542329754, + "loss": 0.5372, + "step": 11414 + }, + { + "epoch": 0.2675855763990225, + "grad_norm": 2.6425580978393555, + "learning_rate": 0.00016673806871301674, + "loss": 0.3579, + "step": 11415 + }, + { + "epoch": 0.26760901797382747, + "grad_norm": 2.8901784420013428, + "learning_rate": 0.00016673258283093545, + "loss": 0.1896, + "step": 11416 + }, + { + "epoch": 0.26763245954863246, + "grad_norm": 3.728825569152832, + "learning_rate": 0.00016672709658676124, + "loss": 1.7222, + "step": 11417 + }, + { + "epoch": 0.26765590112343746, + "grad_norm": 0.5859540700912476, + "learning_rate": 0.00016672160998052398, + "loss": 0.0384, + "step": 11418 + }, + { + "epoch": 0.26767934269824245, + "grad_norm": 4.947568416595459, + "learning_rate": 0.00016671612301225337, + "loss": 0.4541, + "step": 11419 + }, + { + "epoch": 0.26770278427304744, + "grad_norm": 7.843740940093994, + "learning_rate": 0.0001667106356819792, + "loss": 0.6897, + "step": 11420 + }, + { + "epoch": 0.26772622584785244, + "grad_norm": 1.3639204502105713, + "learning_rate": 0.00016670514798973124, + "loss": 0.1663, + "step": 11421 + }, + { + "epoch": 0.26774966742265743, + "grad_norm": 2.9537158012390137, + "learning_rate": 0.00016669965993553932, + "loss": 0.2583, + "step": 11422 + }, + { + "epoch": 0.26777310899746243, + "grad_norm": 5.8973917961120605, + "learning_rate": 0.0001666941715194331, + "loss": 0.6109, + "step": 11423 + }, + { + "epoch": 0.2677965505722674, + "grad_norm": 4.28165340423584, + "learning_rate": 0.00016668868274144244, + "loss": 0.4422, + "step": 11424 + }, + { + "epoch": 0.2678199921470724, + "grad_norm": 3.2317471504211426, + "learning_rate": 0.00016668319360159717, + "loss": 0.451, + "step": 11425 + }, + { + "epoch": 0.2678434337218774, + "grad_norm": 2.4471490383148193, + "learning_rate": 0.00016667770409992697, + "loss": 1.4023, + "step": 11426 + }, + { + "epoch": 0.2678668752966824, + "grad_norm": 3.5961174964904785, + "learning_rate": 0.00016667221423646167, + "loss": 0.3409, + "step": 11427 + }, + { + "epoch": 0.2678903168714874, + "grad_norm": 3.5287210941314697, + "learning_rate": 0.0001666667240112311, + "loss": 0.2859, + "step": 11428 + }, + { + "epoch": 0.26791375844629245, + "grad_norm": 3.055205821990967, + "learning_rate": 0.000166661233424265, + "loss": 0.2623, + "step": 11429 + }, + { + "epoch": 0.26793720002109744, + "grad_norm": 2.2377326488494873, + "learning_rate": 0.0001666557424755931, + "loss": 0.2138, + "step": 11430 + }, + { + "epoch": 0.26796064159590244, + "grad_norm": 2.487004518508911, + "learning_rate": 0.00016665025116524533, + "loss": 0.2353, + "step": 11431 + }, + { + "epoch": 0.26798408317070743, + "grad_norm": 0.7238145470619202, + "learning_rate": 0.0001666447594932514, + "loss": 0.0637, + "step": 11432 + }, + { + "epoch": 0.2680075247455124, + "grad_norm": 3.0807297229766846, + "learning_rate": 0.0001666392674596412, + "loss": 0.2871, + "step": 11433 + }, + { + "epoch": 0.2680309663203174, + "grad_norm": 2.5804786682128906, + "learning_rate": 0.0001666337750644444, + "loss": 0.19, + "step": 11434 + }, + { + "epoch": 0.2680544078951224, + "grad_norm": 10.026653289794922, + "learning_rate": 0.0001666282823076909, + "loss": 0.8932, + "step": 11435 + }, + { + "epoch": 0.2680778494699274, + "grad_norm": 4.991759777069092, + "learning_rate": 0.00016662278918941041, + "loss": 0.4245, + "step": 11436 + }, + { + "epoch": 0.2681012910447324, + "grad_norm": 4.191219329833984, + "learning_rate": 0.0001666172957096328, + "loss": 0.327, + "step": 11437 + }, + { + "epoch": 0.2681247326195374, + "grad_norm": 0.9217866659164429, + "learning_rate": 0.0001666118018683879, + "loss": 0.0693, + "step": 11438 + }, + { + "epoch": 0.2681481741943424, + "grad_norm": 3.336639881134033, + "learning_rate": 0.0001666063076657055, + "loss": 0.2938, + "step": 11439 + }, + { + "epoch": 0.2681716157691474, + "grad_norm": 1.4613276720046997, + "learning_rate": 0.00016660081310161537, + "loss": 0.1923, + "step": 11440 + }, + { + "epoch": 0.2681950573439524, + "grad_norm": 6.48388147354126, + "learning_rate": 0.00016659531817614736, + "loss": 1.0427, + "step": 11441 + }, + { + "epoch": 0.2682184989187574, + "grad_norm": 0.7069831490516663, + "learning_rate": 0.0001665898228893313, + "loss": 0.076, + "step": 11442 + }, + { + "epoch": 0.26824194049356237, + "grad_norm": 1.9403573274612427, + "learning_rate": 0.00016658432724119698, + "loss": 0.3219, + "step": 11443 + }, + { + "epoch": 0.26826538206836736, + "grad_norm": 4.1441121101379395, + "learning_rate": 0.0001665788312317742, + "loss": 0.4268, + "step": 11444 + }, + { + "epoch": 0.26828882364317236, + "grad_norm": 1.1938247680664062, + "learning_rate": 0.00016657333486109287, + "loss": 0.1827, + "step": 11445 + }, + { + "epoch": 0.26831226521797735, + "grad_norm": 3.3205533027648926, + "learning_rate": 0.0001665678381291827, + "loss": 0.3979, + "step": 11446 + }, + { + "epoch": 0.26833570679278235, + "grad_norm": 2.9862864017486572, + "learning_rate": 0.0001665623410360736, + "loss": 2.1085, + "step": 11447 + }, + { + "epoch": 0.26835914836758734, + "grad_norm": 4.7348952293396, + "learning_rate": 0.00016655684358179534, + "loss": 0.5285, + "step": 11448 + }, + { + "epoch": 0.26838258994239234, + "grad_norm": 1.5622751712799072, + "learning_rate": 0.0001665513457663778, + "loss": 0.1502, + "step": 11449 + }, + { + "epoch": 0.26840603151719733, + "grad_norm": 2.8424131870269775, + "learning_rate": 0.0001665458475898508, + "loss": 0.361, + "step": 11450 + }, + { + "epoch": 0.2684294730920023, + "grad_norm": 0.6887962818145752, + "learning_rate": 0.00016654034905224413, + "loss": 0.057, + "step": 11451 + }, + { + "epoch": 0.2684529146668073, + "grad_norm": 1.637563943862915, + "learning_rate": 0.00016653485015358764, + "loss": 0.8322, + "step": 11452 + }, + { + "epoch": 0.2684763562416123, + "grad_norm": 3.95335054397583, + "learning_rate": 0.0001665293508939112, + "loss": 0.4869, + "step": 11453 + }, + { + "epoch": 0.2684997978164173, + "grad_norm": 3.5431203842163086, + "learning_rate": 0.00016652385127324462, + "loss": 0.4558, + "step": 11454 + }, + { + "epoch": 0.2685232393912223, + "grad_norm": 1.4439412355422974, + "learning_rate": 0.0001665183512916178, + "loss": 0.1762, + "step": 11455 + }, + { + "epoch": 0.2685466809660273, + "grad_norm": 3.6544406414031982, + "learning_rate": 0.00016651285094906052, + "loss": 0.3614, + "step": 11456 + }, + { + "epoch": 0.2685701225408323, + "grad_norm": 1.8210467100143433, + "learning_rate": 0.0001665073502456026, + "loss": 0.1384, + "step": 11457 + }, + { + "epoch": 0.2685935641156373, + "grad_norm": 1.9753345251083374, + "learning_rate": 0.00016650184918127393, + "loss": 1.6968, + "step": 11458 + }, + { + "epoch": 0.2686170056904423, + "grad_norm": 0.5930209159851074, + "learning_rate": 0.00016649634775610438, + "loss": 0.0313, + "step": 11459 + }, + { + "epoch": 0.2686404472652473, + "grad_norm": 5.26356840133667, + "learning_rate": 0.0001664908459701238, + "loss": 0.5494, + "step": 11460 + }, + { + "epoch": 0.26866388884005227, + "grad_norm": 0.5208472013473511, + "learning_rate": 0.000166485343823362, + "loss": 0.0692, + "step": 11461 + }, + { + "epoch": 0.26868733041485726, + "grad_norm": 3.0304248332977295, + "learning_rate": 0.0001664798413158488, + "loss": 0.5277, + "step": 11462 + }, + { + "epoch": 0.26871077198966226, + "grad_norm": 2.7314212322235107, + "learning_rate": 0.0001664743384476142, + "loss": 1.6823, + "step": 11463 + }, + { + "epoch": 0.26873421356446725, + "grad_norm": 2.4778249263763428, + "learning_rate": 0.00016646883521868793, + "loss": 0.2309, + "step": 11464 + }, + { + "epoch": 0.26875765513927224, + "grad_norm": 1.9471142292022705, + "learning_rate": 0.0001664633316290999, + "loss": 0.2453, + "step": 11465 + }, + { + "epoch": 0.26878109671407724, + "grad_norm": 1.2693976163864136, + "learning_rate": 0.00016645782767887995, + "loss": 0.1245, + "step": 11466 + }, + { + "epoch": 0.26880453828888223, + "grad_norm": 4.648494720458984, + "learning_rate": 0.00016645232336805796, + "loss": 0.6882, + "step": 11467 + }, + { + "epoch": 0.2688279798636872, + "grad_norm": 2.695791244506836, + "learning_rate": 0.00016644681869666382, + "loss": 0.2397, + "step": 11468 + }, + { + "epoch": 0.2688514214384922, + "grad_norm": 4.843663692474365, + "learning_rate": 0.00016644131366472736, + "loss": 0.7238, + "step": 11469 + }, + { + "epoch": 0.2688748630132972, + "grad_norm": 2.245313882827759, + "learning_rate": 0.00016643580827227841, + "loss": 0.2544, + "step": 11470 + }, + { + "epoch": 0.2688983045881022, + "grad_norm": 3.591893434524536, + "learning_rate": 0.000166430302519347, + "loss": 0.326, + "step": 11471 + }, + { + "epoch": 0.2689217461629072, + "grad_norm": 3.2510969638824463, + "learning_rate": 0.00016642479640596283, + "loss": 0.4586, + "step": 11472 + }, + { + "epoch": 0.2689451877377122, + "grad_norm": 0.5668479800224304, + "learning_rate": 0.00016641928993215585, + "loss": 0.0354, + "step": 11473 + }, + { + "epoch": 0.2689686293125172, + "grad_norm": 1.11821711063385, + "learning_rate": 0.00016641378309795594, + "loss": 0.5235, + "step": 11474 + }, + { + "epoch": 0.2689920708873222, + "grad_norm": 4.961427211761475, + "learning_rate": 0.000166408275903393, + "loss": 0.6608, + "step": 11475 + }, + { + "epoch": 0.2690155124621272, + "grad_norm": 1.5004589557647705, + "learning_rate": 0.00016640276834849687, + "loss": 0.9999, + "step": 11476 + }, + { + "epoch": 0.2690389540369322, + "grad_norm": 5.862856864929199, + "learning_rate": 0.00016639726043329744, + "loss": 0.303, + "step": 11477 + }, + { + "epoch": 0.26906239561173717, + "grad_norm": 6.372698783874512, + "learning_rate": 0.0001663917521578246, + "loss": 0.8052, + "step": 11478 + }, + { + "epoch": 0.26908583718654216, + "grad_norm": 3.797569513320923, + "learning_rate": 0.00016638624352210827, + "loss": 0.2598, + "step": 11479 + }, + { + "epoch": 0.2691092787613472, + "grad_norm": 3.1261954307556152, + "learning_rate": 0.00016638073452617834, + "loss": 0.2692, + "step": 11480 + }, + { + "epoch": 0.2691327203361522, + "grad_norm": 2.526991844177246, + "learning_rate": 0.00016637522517006463, + "loss": 0.2085, + "step": 11481 + }, + { + "epoch": 0.2691561619109572, + "grad_norm": 3.1394753456115723, + "learning_rate": 0.00016636971545379712, + "loss": 0.1794, + "step": 11482 + }, + { + "epoch": 0.2691796034857622, + "grad_norm": 4.017633438110352, + "learning_rate": 0.00016636420537740566, + "loss": 0.4256, + "step": 11483 + }, + { + "epoch": 0.2692030450605672, + "grad_norm": 2.6744801998138428, + "learning_rate": 0.0001663586949409201, + "loss": 0.2322, + "step": 11484 + }, + { + "epoch": 0.2692264866353722, + "grad_norm": 3.3944270610809326, + "learning_rate": 0.00016635318414437046, + "loss": 0.4328, + "step": 11485 + }, + { + "epoch": 0.2692499282101772, + "grad_norm": 1.344457983970642, + "learning_rate": 0.00016634767298778658, + "loss": 0.1583, + "step": 11486 + }, + { + "epoch": 0.2692733697849822, + "grad_norm": 3.3734171390533447, + "learning_rate": 0.00016634216147119832, + "loss": 0.4842, + "step": 11487 + }, + { + "epoch": 0.26929681135978717, + "grad_norm": 3.6718249320983887, + "learning_rate": 0.00016633664959463565, + "loss": 0.3772, + "step": 11488 + }, + { + "epoch": 0.26932025293459216, + "grad_norm": 4.446484088897705, + "learning_rate": 0.00016633113735812846, + "loss": 0.4226, + "step": 11489 + }, + { + "epoch": 0.26934369450939716, + "grad_norm": 2.4087014198303223, + "learning_rate": 0.00016632562476170663, + "loss": 0.2895, + "step": 11490 + }, + { + "epoch": 0.26936713608420215, + "grad_norm": 3.6476168632507324, + "learning_rate": 0.00016632011180540014, + "loss": 0.2206, + "step": 11491 + }, + { + "epoch": 0.26939057765900715, + "grad_norm": 3.931349277496338, + "learning_rate": 0.0001663145984892388, + "loss": 0.2446, + "step": 11492 + }, + { + "epoch": 0.26941401923381214, + "grad_norm": 4.606611251831055, + "learning_rate": 0.0001663090848132526, + "loss": 0.408, + "step": 11493 + }, + { + "epoch": 0.26943746080861714, + "grad_norm": 5.1672844886779785, + "learning_rate": 0.00016630357077747148, + "loss": 0.4885, + "step": 11494 + }, + { + "epoch": 0.26946090238342213, + "grad_norm": 4.035650253295898, + "learning_rate": 0.00016629805638192528, + "loss": 0.3782, + "step": 11495 + }, + { + "epoch": 0.2694843439582271, + "grad_norm": 4.696628570556641, + "learning_rate": 0.00016629254162664397, + "loss": 0.724, + "step": 11496 + }, + { + "epoch": 0.2695077855330321, + "grad_norm": 1.0514252185821533, + "learning_rate": 0.0001662870265116575, + "loss": 0.1557, + "step": 11497 + }, + { + "epoch": 0.2695312271078371, + "grad_norm": 2.5733695030212402, + "learning_rate": 0.0001662815110369957, + "loss": 0.2501, + "step": 11498 + }, + { + "epoch": 0.2695546686826421, + "grad_norm": 3.4074344635009766, + "learning_rate": 0.00016627599520268859, + "loss": 0.3747, + "step": 11499 + }, + { + "epoch": 0.2695781102574471, + "grad_norm": 4.076681613922119, + "learning_rate": 0.00016627047900876608, + "loss": 0.6096, + "step": 11500 + }, + { + "epoch": 0.2696015518322521, + "grad_norm": 4.710137844085693, + "learning_rate": 0.00016626496245525808, + "loss": 0.5077, + "step": 11501 + }, + { + "epoch": 0.2696249934070571, + "grad_norm": 0.7789344191551208, + "learning_rate": 0.0001662594455421945, + "loss": 0.0496, + "step": 11502 + }, + { + "epoch": 0.2696484349818621, + "grad_norm": 1.6782773733139038, + "learning_rate": 0.0001662539282696053, + "loss": 1.3372, + "step": 11503 + }, + { + "epoch": 0.2696718765566671, + "grad_norm": 5.516061305999756, + "learning_rate": 0.00016624841063752049, + "loss": 0.4725, + "step": 11504 + }, + { + "epoch": 0.2696953181314721, + "grad_norm": 0.8215967416763306, + "learning_rate": 0.0001662428926459699, + "loss": 0.0992, + "step": 11505 + }, + { + "epoch": 0.26971875970627707, + "grad_norm": 3.867401123046875, + "learning_rate": 0.0001662373742949835, + "loss": 0.3431, + "step": 11506 + }, + { + "epoch": 0.26974220128108206, + "grad_norm": 1.2337393760681152, + "learning_rate": 0.00016623185558459127, + "loss": 0.1184, + "step": 11507 + }, + { + "epoch": 0.26976564285588706, + "grad_norm": 6.799383163452148, + "learning_rate": 0.00016622633651482312, + "loss": 0.5422, + "step": 11508 + }, + { + "epoch": 0.26978908443069205, + "grad_norm": 1.8531440496444702, + "learning_rate": 0.000166220817085709, + "loss": 0.2025, + "step": 11509 + }, + { + "epoch": 0.26981252600549704, + "grad_norm": 3.4774539470672607, + "learning_rate": 0.00016621529729727886, + "loss": 0.3704, + "step": 11510 + }, + { + "epoch": 0.26983596758030204, + "grad_norm": 5.0007805824279785, + "learning_rate": 0.00016620977714956267, + "loss": 0.4827, + "step": 11511 + }, + { + "epoch": 0.26985940915510703, + "grad_norm": 1.2028515338897705, + "learning_rate": 0.00016620425664259035, + "loss": 0.1339, + "step": 11512 + }, + { + "epoch": 0.269882850729912, + "grad_norm": 5.224799156188965, + "learning_rate": 0.00016619873577639188, + "loss": 0.5087, + "step": 11513 + }, + { + "epoch": 0.269906292304717, + "grad_norm": 4.4931535720825195, + "learning_rate": 0.00016619321455099722, + "loss": 0.256, + "step": 11514 + }, + { + "epoch": 0.269929733879522, + "grad_norm": 1.419734001159668, + "learning_rate": 0.0001661876929664363, + "loss": 0.2043, + "step": 11515 + }, + { + "epoch": 0.269953175454327, + "grad_norm": 6.698697566986084, + "learning_rate": 0.0001661821710227391, + "loss": 0.5789, + "step": 11516 + }, + { + "epoch": 0.269976617029132, + "grad_norm": 0.35641103982925415, + "learning_rate": 0.0001661766487199356, + "loss": 0.0418, + "step": 11517 + }, + { + "epoch": 0.270000058603937, + "grad_norm": 2.9618165493011475, + "learning_rate": 0.0001661711260580557, + "loss": 0.3419, + "step": 11518 + }, + { + "epoch": 0.270023500178742, + "grad_norm": 5.131359100341797, + "learning_rate": 0.00016616560303712947, + "loss": 0.761, + "step": 11519 + }, + { + "epoch": 0.270046941753547, + "grad_norm": 1.124335527420044, + "learning_rate": 0.00016616007965718677, + "loss": 0.0902, + "step": 11520 + }, + { + "epoch": 0.270070383328352, + "grad_norm": 0.7595116496086121, + "learning_rate": 0.00016615455591825762, + "loss": 0.0601, + "step": 11521 + }, + { + "epoch": 0.270093824903157, + "grad_norm": 3.4244813919067383, + "learning_rate": 0.000166149031820372, + "loss": 0.3573, + "step": 11522 + }, + { + "epoch": 0.27011726647796197, + "grad_norm": 1.4193319082260132, + "learning_rate": 0.0001661435073635599, + "loss": 0.2013, + "step": 11523 + }, + { + "epoch": 0.27014070805276696, + "grad_norm": 0.801612138748169, + "learning_rate": 0.00016613798254785125, + "loss": 0.086, + "step": 11524 + }, + { + "epoch": 0.27016414962757196, + "grad_norm": 0.8559765219688416, + "learning_rate": 0.000166132457373276, + "loss": 0.0717, + "step": 11525 + }, + { + "epoch": 0.27018759120237695, + "grad_norm": 4.0844550132751465, + "learning_rate": 0.00016612693183986425, + "loss": 0.3967, + "step": 11526 + }, + { + "epoch": 0.27021103277718195, + "grad_norm": 5.3272318840026855, + "learning_rate": 0.00016612140594764586, + "loss": 0.5082, + "step": 11527 + }, + { + "epoch": 0.27023447435198694, + "grad_norm": 3.6264193058013916, + "learning_rate": 0.00016611587969665088, + "loss": 0.3254, + "step": 11528 + }, + { + "epoch": 0.27025791592679194, + "grad_norm": 1.0564789772033691, + "learning_rate": 0.00016611035308690929, + "loss": 0.0963, + "step": 11529 + }, + { + "epoch": 0.27028135750159693, + "grad_norm": 0.49455156922340393, + "learning_rate": 0.00016610482611845103, + "loss": 0.0599, + "step": 11530 + }, + { + "epoch": 0.270304799076402, + "grad_norm": 2.6127729415893555, + "learning_rate": 0.00016609929879130613, + "loss": 1.134, + "step": 11531 + }, + { + "epoch": 0.270328240651207, + "grad_norm": 22.735525131225586, + "learning_rate": 0.00016609377110550461, + "loss": 0.3821, + "step": 11532 + }, + { + "epoch": 0.27035168222601197, + "grad_norm": 4.062930583953857, + "learning_rate": 0.0001660882430610764, + "loss": 0.4088, + "step": 11533 + }, + { + "epoch": 0.27037512380081696, + "grad_norm": 16.678789138793945, + "learning_rate": 0.00016608271465805155, + "loss": 0.7458, + "step": 11534 + }, + { + "epoch": 0.27039856537562196, + "grad_norm": 3.7411000728607178, + "learning_rate": 0.00016607718589646003, + "loss": 0.451, + "step": 11535 + }, + { + "epoch": 0.27042200695042695, + "grad_norm": 1.2958863973617554, + "learning_rate": 0.0001660716567763318, + "loss": 0.1244, + "step": 11536 + }, + { + "epoch": 0.27044544852523195, + "grad_norm": 4.818488121032715, + "learning_rate": 0.00016606612729769697, + "loss": 0.5608, + "step": 11537 + }, + { + "epoch": 0.27046889010003694, + "grad_norm": 3.0708107948303223, + "learning_rate": 0.00016606059746058542, + "loss": 0.4999, + "step": 11538 + }, + { + "epoch": 0.27049233167484193, + "grad_norm": 6.20689582824707, + "learning_rate": 0.00016605506726502723, + "loss": 0.6465, + "step": 11539 + }, + { + "epoch": 0.27051577324964693, + "grad_norm": 1.7549889087677002, + "learning_rate": 0.0001660495367110524, + "loss": 0.2446, + "step": 11540 + }, + { + "epoch": 0.2705392148244519, + "grad_norm": 2.9353830814361572, + "learning_rate": 0.0001660440057986909, + "loss": 0.3928, + "step": 11541 + }, + { + "epoch": 0.2705626563992569, + "grad_norm": 10.763973236083984, + "learning_rate": 0.0001660384745279728, + "loss": 0.9415, + "step": 11542 + }, + { + "epoch": 0.2705860979740619, + "grad_norm": 9.874831199645996, + "learning_rate": 0.00016603294289892806, + "loss": 0.7302, + "step": 11543 + }, + { + "epoch": 0.2706095395488669, + "grad_norm": 1.7232760190963745, + "learning_rate": 0.0001660274109115867, + "loss": 0.2132, + "step": 11544 + }, + { + "epoch": 0.2706329811236719, + "grad_norm": 1.0749322175979614, + "learning_rate": 0.00016602187856597876, + "loss": 0.1157, + "step": 11545 + }, + { + "epoch": 0.2706564226984769, + "grad_norm": 1.8194615840911865, + "learning_rate": 0.00016601634586213427, + "loss": 0.0622, + "step": 11546 + }, + { + "epoch": 0.2706798642732819, + "grad_norm": 0.7798598408699036, + "learning_rate": 0.00016601081280008322, + "loss": 0.0679, + "step": 11547 + }, + { + "epoch": 0.2707033058480869, + "grad_norm": 4.101974964141846, + "learning_rate": 0.00016600527937985564, + "loss": 0.5154, + "step": 11548 + }, + { + "epoch": 0.2707267474228919, + "grad_norm": 1.382574439048767, + "learning_rate": 0.00016599974560148155, + "loss": 0.2144, + "step": 11549 + }, + { + "epoch": 0.2707501889976969, + "grad_norm": 0.9630309343338013, + "learning_rate": 0.000165994211464991, + "loss": 0.1033, + "step": 11550 + }, + { + "epoch": 0.27077363057250187, + "grad_norm": 3.3431589603424072, + "learning_rate": 0.00016598867697041397, + "loss": 0.3109, + "step": 11551 + }, + { + "epoch": 0.27079707214730686, + "grad_norm": 3.9265544414520264, + "learning_rate": 0.0001659831421177806, + "loss": 0.6891, + "step": 11552 + }, + { + "epoch": 0.27082051372211186, + "grad_norm": 3.6541216373443604, + "learning_rate": 0.0001659776069071208, + "loss": 0.3721, + "step": 11553 + }, + { + "epoch": 0.27084395529691685, + "grad_norm": 3.1108806133270264, + "learning_rate": 0.0001659720713384646, + "loss": 0.2099, + "step": 11554 + }, + { + "epoch": 0.27086739687172184, + "grad_norm": 2.9800877571105957, + "learning_rate": 0.00016596653541184217, + "loss": 0.3462, + "step": 11555 + }, + { + "epoch": 0.27089083844652684, + "grad_norm": 2.927981376647949, + "learning_rate": 0.00016596099912728342, + "loss": 0.2558, + "step": 11556 + }, + { + "epoch": 0.27091428002133183, + "grad_norm": 1.0396445989608765, + "learning_rate": 0.00016595546248481845, + "loss": 0.0761, + "step": 11557 + }, + { + "epoch": 0.2709377215961368, + "grad_norm": 7.088354587554932, + "learning_rate": 0.00016594992548447728, + "loss": 0.6736, + "step": 11558 + }, + { + "epoch": 0.2709611631709418, + "grad_norm": 4.346349716186523, + "learning_rate": 0.00016594438812628996, + "loss": 0.5469, + "step": 11559 + }, + { + "epoch": 0.2709846047457468, + "grad_norm": 0.6499861478805542, + "learning_rate": 0.0001659388504102865, + "loss": 0.0736, + "step": 11560 + }, + { + "epoch": 0.2710080463205518, + "grad_norm": 5.945803165435791, + "learning_rate": 0.00016593331233649705, + "loss": 0.8241, + "step": 11561 + }, + { + "epoch": 0.2710314878953568, + "grad_norm": 1.546257734298706, + "learning_rate": 0.00016592777390495156, + "loss": 0.1761, + "step": 11562 + }, + { + "epoch": 0.2710549294701618, + "grad_norm": 4.477145195007324, + "learning_rate": 0.00016592223511568011, + "loss": 0.46, + "step": 11563 + }, + { + "epoch": 0.2710783710449668, + "grad_norm": 1.0312696695327759, + "learning_rate": 0.00016591669596871276, + "loss": 0.0754, + "step": 11564 + }, + { + "epoch": 0.2711018126197718, + "grad_norm": 4.991665363311768, + "learning_rate": 0.00016591115646407957, + "loss": 0.5587, + "step": 11565 + }, + { + "epoch": 0.2711252541945768, + "grad_norm": 3.2862887382507324, + "learning_rate": 0.00016590561660181062, + "loss": 0.4286, + "step": 11566 + }, + { + "epoch": 0.2711486957693818, + "grad_norm": 1.3122782707214355, + "learning_rate": 0.00016590007638193587, + "loss": 0.1552, + "step": 11567 + }, + { + "epoch": 0.27117213734418677, + "grad_norm": 3.6340901851654053, + "learning_rate": 0.00016589453580448553, + "loss": 0.2581, + "step": 11568 + }, + { + "epoch": 0.27119557891899176, + "grad_norm": 2.2553887367248535, + "learning_rate": 0.00016588899486948952, + "loss": 0.2954, + "step": 11569 + }, + { + "epoch": 0.27121902049379676, + "grad_norm": 3.059995651245117, + "learning_rate": 0.000165883453576978, + "loss": 0.3537, + "step": 11570 + }, + { + "epoch": 0.27124246206860175, + "grad_norm": 5.168790817260742, + "learning_rate": 0.00016587791192698102, + "loss": 0.5759, + "step": 11571 + }, + { + "epoch": 0.27126590364340675, + "grad_norm": 6.328413963317871, + "learning_rate": 0.00016587236991952864, + "loss": 1.0294, + "step": 11572 + }, + { + "epoch": 0.27128934521821174, + "grad_norm": 1.814542531967163, + "learning_rate": 0.0001658668275546509, + "loss": 0.3236, + "step": 11573 + }, + { + "epoch": 0.27131278679301674, + "grad_norm": 3.1603856086730957, + "learning_rate": 0.00016586128483237793, + "loss": 0.3775, + "step": 11574 + }, + { + "epoch": 0.27133622836782173, + "grad_norm": 3.753563404083252, + "learning_rate": 0.00016585574175273976, + "loss": 0.3972, + "step": 11575 + }, + { + "epoch": 0.2713596699426267, + "grad_norm": 0.6739986538887024, + "learning_rate": 0.0001658501983157665, + "loss": 0.1106, + "step": 11576 + }, + { + "epoch": 0.2713831115174317, + "grad_norm": 0.5484141111373901, + "learning_rate": 0.00016584465452148818, + "loss": 0.0741, + "step": 11577 + }, + { + "epoch": 0.2714065530922367, + "grad_norm": 0.7264254093170166, + "learning_rate": 0.00016583911036993496, + "loss": 0.1047, + "step": 11578 + }, + { + "epoch": 0.2714299946670417, + "grad_norm": 3.382807970046997, + "learning_rate": 0.00016583356586113683, + "loss": 0.5131, + "step": 11579 + }, + { + "epoch": 0.2714534362418467, + "grad_norm": 2.37996244430542, + "learning_rate": 0.00016582802099512396, + "loss": 1.6661, + "step": 11580 + }, + { + "epoch": 0.2714768778166517, + "grad_norm": 3.6411941051483154, + "learning_rate": 0.00016582247577192634, + "loss": 0.6339, + "step": 11581 + }, + { + "epoch": 0.2715003193914567, + "grad_norm": 2.8023529052734375, + "learning_rate": 0.00016581693019157418, + "loss": 0.2529, + "step": 11582 + }, + { + "epoch": 0.27152376096626174, + "grad_norm": 3.958524227142334, + "learning_rate": 0.00016581138425409746, + "loss": 0.4785, + "step": 11583 + }, + { + "epoch": 0.27154720254106673, + "grad_norm": 4.157865524291992, + "learning_rate": 0.00016580583795952633, + "loss": 0.5918, + "step": 11584 + }, + { + "epoch": 0.27157064411587173, + "grad_norm": 9.111495018005371, + "learning_rate": 0.00016580029130789091, + "loss": 0.6253, + "step": 11585 + }, + { + "epoch": 0.2715940856906767, + "grad_norm": 5.1370368003845215, + "learning_rate": 0.00016579474429922125, + "loss": 0.6264, + "step": 11586 + }, + { + "epoch": 0.2716175272654817, + "grad_norm": 1.0822948217391968, + "learning_rate": 0.0001657891969335474, + "loss": 0.1122, + "step": 11587 + }, + { + "epoch": 0.2716409688402867, + "grad_norm": 1.9715042114257812, + "learning_rate": 0.00016578364921089955, + "loss": 0.1262, + "step": 11588 + }, + { + "epoch": 0.2716644104150917, + "grad_norm": 2.6948084831237793, + "learning_rate": 0.00016577810113130778, + "loss": 0.4819, + "step": 11589 + }, + { + "epoch": 0.2716878519898967, + "grad_norm": 3.7890002727508545, + "learning_rate": 0.00016577255269480218, + "loss": 0.5925, + "step": 11590 + }, + { + "epoch": 0.2717112935647017, + "grad_norm": 1.3588820695877075, + "learning_rate": 0.00016576700390141287, + "loss": 0.9542, + "step": 11591 + }, + { + "epoch": 0.2717347351395067, + "grad_norm": 2.777108907699585, + "learning_rate": 0.00016576145475116994, + "loss": 0.2946, + "step": 11592 + }, + { + "epoch": 0.2717581767143117, + "grad_norm": 6.576705455780029, + "learning_rate": 0.00016575590524410351, + "loss": 0.6151, + "step": 11593 + }, + { + "epoch": 0.2717816182891167, + "grad_norm": 4.47195291519165, + "learning_rate": 0.00016575035538024367, + "loss": 0.3594, + "step": 11594 + }, + { + "epoch": 0.27180505986392167, + "grad_norm": 4.402732849121094, + "learning_rate": 0.00016574480515962057, + "loss": 0.5866, + "step": 11595 + }, + { + "epoch": 0.27182850143872667, + "grad_norm": 3.524780511856079, + "learning_rate": 0.0001657392545822643, + "loss": 0.2688, + "step": 11596 + }, + { + "epoch": 0.27185194301353166, + "grad_norm": 3.6058719158172607, + "learning_rate": 0.000165733703648205, + "loss": 0.3308, + "step": 11597 + }, + { + "epoch": 0.27187538458833665, + "grad_norm": 2.7166833877563477, + "learning_rate": 0.00016572815235747274, + "loss": 0.2001, + "step": 11598 + }, + { + "epoch": 0.27189882616314165, + "grad_norm": 4.10493278503418, + "learning_rate": 0.00016572260071009772, + "loss": 0.3833, + "step": 11599 + }, + { + "epoch": 0.27192226773794664, + "grad_norm": 1.9937989711761475, + "learning_rate": 0.00016571704870610999, + "loss": 0.2423, + "step": 11600 + }, + { + "epoch": 0.27194570931275164, + "grad_norm": 0.5576638579368591, + "learning_rate": 0.00016571149634553973, + "loss": 0.0764, + "step": 11601 + }, + { + "epoch": 0.27196915088755663, + "grad_norm": 4.311731815338135, + "learning_rate": 0.00016570594362841704, + "loss": 0.5702, + "step": 11602 + }, + { + "epoch": 0.2719925924623616, + "grad_norm": 2.868906259536743, + "learning_rate": 0.00016570039055477205, + "loss": 0.2321, + "step": 11603 + }, + { + "epoch": 0.2720160340371666, + "grad_norm": 5.3827643394470215, + "learning_rate": 0.00016569483712463486, + "loss": 0.8557, + "step": 11604 + }, + { + "epoch": 0.2720394756119716, + "grad_norm": 3.982213020324707, + "learning_rate": 0.00016568928333803568, + "loss": 0.702, + "step": 11605 + }, + { + "epoch": 0.2720629171867766, + "grad_norm": 3.039019823074341, + "learning_rate": 0.00016568372919500458, + "loss": 0.4339, + "step": 11606 + }, + { + "epoch": 0.2720863587615816, + "grad_norm": 3.9764459133148193, + "learning_rate": 0.00016567817469557175, + "loss": 0.5644, + "step": 11607 + }, + { + "epoch": 0.2721098003363866, + "grad_norm": 3.6080000400543213, + "learning_rate": 0.00016567261983976725, + "loss": 0.3948, + "step": 11608 + }, + { + "epoch": 0.2721332419111916, + "grad_norm": 2.914947271347046, + "learning_rate": 0.00016566706462762133, + "loss": 0.3146, + "step": 11609 + }, + { + "epoch": 0.2721566834859966, + "grad_norm": 4.5836100578308105, + "learning_rate": 0.000165661509059164, + "loss": 2.3902, + "step": 11610 + }, + { + "epoch": 0.2721801250608016, + "grad_norm": 5.1166534423828125, + "learning_rate": 0.0001656559531344255, + "loss": 0.48, + "step": 11611 + }, + { + "epoch": 0.2722035666356066, + "grad_norm": 0.6078246235847473, + "learning_rate": 0.00016565039685343598, + "loss": 0.0614, + "step": 11612 + }, + { + "epoch": 0.27222700821041157, + "grad_norm": 8.378833770751953, + "learning_rate": 0.00016564484021622554, + "loss": 0.4675, + "step": 11613 + }, + { + "epoch": 0.27225044978521656, + "grad_norm": 3.3219244480133057, + "learning_rate": 0.00016563928322282435, + "loss": 0.3095, + "step": 11614 + }, + { + "epoch": 0.27227389136002156, + "grad_norm": 1.0637719631195068, + "learning_rate": 0.00016563372587326258, + "loss": 0.083, + "step": 11615 + }, + { + "epoch": 0.27229733293482655, + "grad_norm": 3.78702974319458, + "learning_rate": 0.00016562816816757033, + "loss": 0.1855, + "step": 11616 + }, + { + "epoch": 0.27232077450963155, + "grad_norm": 5.516921520233154, + "learning_rate": 0.0001656226101057778, + "loss": 1.8689, + "step": 11617 + }, + { + "epoch": 0.27234421608443654, + "grad_norm": 0.862687349319458, + "learning_rate": 0.00016561705168791516, + "loss": 0.0903, + "step": 11618 + }, + { + "epoch": 0.27236765765924154, + "grad_norm": 1.3618077039718628, + "learning_rate": 0.00016561149291401255, + "loss": 0.177, + "step": 11619 + }, + { + "epoch": 0.27239109923404653, + "grad_norm": 1.3854626417160034, + "learning_rate": 0.00016560593378410016, + "loss": 0.2119, + "step": 11620 + }, + { + "epoch": 0.2724145408088515, + "grad_norm": 5.750204086303711, + "learning_rate": 0.00016560037429820808, + "loss": 0.3659, + "step": 11621 + }, + { + "epoch": 0.2724379823836565, + "grad_norm": 4.4301958084106445, + "learning_rate": 0.00016559481445636658, + "loss": 0.2681, + "step": 11622 + }, + { + "epoch": 0.2724614239584615, + "grad_norm": 2.924126625061035, + "learning_rate": 0.00016558925425860576, + "loss": 0.1664, + "step": 11623 + }, + { + "epoch": 0.2724848655332665, + "grad_norm": 2.9702513217926025, + "learning_rate": 0.00016558369370495575, + "loss": 0.3724, + "step": 11624 + }, + { + "epoch": 0.2725083071080715, + "grad_norm": 2.793651580810547, + "learning_rate": 0.00016557813279544683, + "loss": 0.3381, + "step": 11625 + }, + { + "epoch": 0.2725317486828765, + "grad_norm": 5.024477958679199, + "learning_rate": 0.00016557257153010908, + "loss": 0.4823, + "step": 11626 + }, + { + "epoch": 0.2725551902576815, + "grad_norm": 4.372223854064941, + "learning_rate": 0.00016556700990897274, + "loss": 0.553, + "step": 11627 + }, + { + "epoch": 0.2725786318324865, + "grad_norm": 4.786637783050537, + "learning_rate": 0.00016556144793206795, + "loss": 0.4211, + "step": 11628 + }, + { + "epoch": 0.2726020734072915, + "grad_norm": 2.596653938293457, + "learning_rate": 0.00016555588559942489, + "loss": 0.2107, + "step": 11629 + }, + { + "epoch": 0.2726255149820965, + "grad_norm": 2.316810131072998, + "learning_rate": 0.00016555032291107376, + "loss": 1.5621, + "step": 11630 + }, + { + "epoch": 0.27264895655690147, + "grad_norm": 3.4516286849975586, + "learning_rate": 0.00016554475986704477, + "loss": 1.7858, + "step": 11631 + }, + { + "epoch": 0.27267239813170646, + "grad_norm": 1.6012815237045288, + "learning_rate": 0.000165539196467368, + "loss": 0.2166, + "step": 11632 + }, + { + "epoch": 0.27269583970651146, + "grad_norm": 3.733854293823242, + "learning_rate": 0.00016553363271207375, + "loss": 0.3648, + "step": 11633 + }, + { + "epoch": 0.2727192812813165, + "grad_norm": 3.4672040939331055, + "learning_rate": 0.00016552806860119218, + "loss": 0.4005, + "step": 11634 + }, + { + "epoch": 0.2727427228561215, + "grad_norm": 2.4831221103668213, + "learning_rate": 0.00016552250413475345, + "loss": 0.1633, + "step": 11635 + }, + { + "epoch": 0.2727661644309265, + "grad_norm": 4.806029319763184, + "learning_rate": 0.00016551693931278775, + "loss": 0.6541, + "step": 11636 + }, + { + "epoch": 0.2727896060057315, + "grad_norm": 1.1747642755508423, + "learning_rate": 0.0001655113741353253, + "loss": 0.1298, + "step": 11637 + }, + { + "epoch": 0.2728130475805365, + "grad_norm": 3.2044942378997803, + "learning_rate": 0.0001655058086023963, + "loss": 0.4138, + "step": 11638 + }, + { + "epoch": 0.2728364891553415, + "grad_norm": 0.7032532095909119, + "learning_rate": 0.000165500242714031, + "loss": 0.0673, + "step": 11639 + }, + { + "epoch": 0.27285993073014647, + "grad_norm": 1.0938161611557007, + "learning_rate": 0.00016549467647025945, + "loss": 0.0821, + "step": 11640 + }, + { + "epoch": 0.27288337230495147, + "grad_norm": 4.687009811401367, + "learning_rate": 0.000165489109871112, + "loss": 0.4552, + "step": 11641 + }, + { + "epoch": 0.27290681387975646, + "grad_norm": 2.6350889205932617, + "learning_rate": 0.00016548354291661877, + "loss": 0.3851, + "step": 11642 + }, + { + "epoch": 0.27293025545456145, + "grad_norm": 4.645069599151611, + "learning_rate": 0.00016547797560680997, + "loss": 0.5981, + "step": 11643 + }, + { + "epoch": 0.27295369702936645, + "grad_norm": 4.313838481903076, + "learning_rate": 0.0001654724079417159, + "loss": 0.522, + "step": 11644 + }, + { + "epoch": 0.27297713860417144, + "grad_norm": 1.5775891542434692, + "learning_rate": 0.00016546683992136666, + "loss": 0.1645, + "step": 11645 + }, + { + "epoch": 0.27300058017897644, + "grad_norm": 3.3349592685699463, + "learning_rate": 0.00016546127154579252, + "loss": 0.4478, + "step": 11646 + }, + { + "epoch": 0.27302402175378143, + "grad_norm": 3.9109537601470947, + "learning_rate": 0.0001654557028150237, + "loss": 0.359, + "step": 11647 + }, + { + "epoch": 0.2730474633285864, + "grad_norm": 0.7146241664886475, + "learning_rate": 0.00016545013372909034, + "loss": 0.0385, + "step": 11648 + }, + { + "epoch": 0.2730709049033914, + "grad_norm": 3.1760432720184326, + "learning_rate": 0.00016544456428802277, + "loss": 0.5931, + "step": 11649 + }, + { + "epoch": 0.2730943464781964, + "grad_norm": 4.660726547241211, + "learning_rate": 0.00016543899449185113, + "loss": 0.6159, + "step": 11650 + }, + { + "epoch": 0.2731177880530014, + "grad_norm": 0.62176913022995, + "learning_rate": 0.00016543342434060566, + "loss": 0.0645, + "step": 11651 + }, + { + "epoch": 0.2731412296278064, + "grad_norm": 2.5728039741516113, + "learning_rate": 0.0001654278538343166, + "loss": 0.2505, + "step": 11652 + }, + { + "epoch": 0.2731646712026114, + "grad_norm": 3.7659780979156494, + "learning_rate": 0.00016542228297301416, + "loss": 0.381, + "step": 11653 + }, + { + "epoch": 0.2731881127774164, + "grad_norm": 2.9650278091430664, + "learning_rate": 0.00016541671175672855, + "loss": 0.4467, + "step": 11654 + }, + { + "epoch": 0.2732115543522214, + "grad_norm": 5.920913219451904, + "learning_rate": 0.00016541114018549005, + "loss": 0.5557, + "step": 11655 + }, + { + "epoch": 0.2732349959270264, + "grad_norm": 5.673497676849365, + "learning_rate": 0.00016540556825932887, + "loss": 0.8393, + "step": 11656 + }, + { + "epoch": 0.2732584375018314, + "grad_norm": 2.9733545780181885, + "learning_rate": 0.00016539999597827522, + "loss": 0.1823, + "step": 11657 + }, + { + "epoch": 0.27328187907663637, + "grad_norm": 4.275442123413086, + "learning_rate": 0.0001653944233423594, + "loss": 0.0816, + "step": 11658 + }, + { + "epoch": 0.27330532065144136, + "grad_norm": 4.740694046020508, + "learning_rate": 0.00016538885035161154, + "loss": 0.3886, + "step": 11659 + }, + { + "epoch": 0.27332876222624636, + "grad_norm": 2.911752462387085, + "learning_rate": 0.00016538327700606198, + "loss": 0.334, + "step": 11660 + }, + { + "epoch": 0.27335220380105135, + "grad_norm": 4.495076656341553, + "learning_rate": 0.00016537770330574087, + "loss": 0.3123, + "step": 11661 + }, + { + "epoch": 0.27337564537585635, + "grad_norm": 0.8751235008239746, + "learning_rate": 0.00016537212925067855, + "loss": 0.1186, + "step": 11662 + }, + { + "epoch": 0.27339908695066134, + "grad_norm": 3.9822275638580322, + "learning_rate": 0.00016536655484090522, + "loss": 0.3354, + "step": 11663 + }, + { + "epoch": 0.27342252852546634, + "grad_norm": 4.478166580200195, + "learning_rate": 0.00016536098007645112, + "loss": 0.7214, + "step": 11664 + }, + { + "epoch": 0.27344597010027133, + "grad_norm": 3.721811294555664, + "learning_rate": 0.00016535540495734647, + "loss": 0.5024, + "step": 11665 + }, + { + "epoch": 0.2734694116750763, + "grad_norm": 4.986618995666504, + "learning_rate": 0.0001653498294836216, + "loss": 0.4704, + "step": 11666 + }, + { + "epoch": 0.2734928532498813, + "grad_norm": 2.947293281555176, + "learning_rate": 0.00016534425365530666, + "loss": 0.3379, + "step": 11667 + }, + { + "epoch": 0.2735162948246863, + "grad_norm": 5.742562294006348, + "learning_rate": 0.000165338677472432, + "loss": 0.4743, + "step": 11668 + }, + { + "epoch": 0.2735397363994913, + "grad_norm": 4.110475540161133, + "learning_rate": 0.00016533310093502784, + "loss": 0.504, + "step": 11669 + }, + { + "epoch": 0.2735631779742963, + "grad_norm": 1.3474823236465454, + "learning_rate": 0.00016532752404312446, + "loss": 0.1546, + "step": 11670 + }, + { + "epoch": 0.2735866195491013, + "grad_norm": 3.3385813236236572, + "learning_rate": 0.00016532194679675207, + "loss": 0.3253, + "step": 11671 + }, + { + "epoch": 0.2736100611239063, + "grad_norm": 3.565397262573242, + "learning_rate": 0.00016531636919594095, + "loss": 0.3727, + "step": 11672 + }, + { + "epoch": 0.2736335026987113, + "grad_norm": 3.279970645904541, + "learning_rate": 0.00016531079124072138, + "loss": 0.3868, + "step": 11673 + }, + { + "epoch": 0.2736569442735163, + "grad_norm": 0.7977884411811829, + "learning_rate": 0.00016530521293112364, + "loss": 0.0553, + "step": 11674 + }, + { + "epoch": 0.2736803858483213, + "grad_norm": 2.7418055534362793, + "learning_rate": 0.000165299634267178, + "loss": 0.2802, + "step": 11675 + }, + { + "epoch": 0.27370382742312627, + "grad_norm": 10.932791709899902, + "learning_rate": 0.00016529405524891465, + "loss": 0.3579, + "step": 11676 + }, + { + "epoch": 0.27372726899793126, + "grad_norm": 1.812669038772583, + "learning_rate": 0.00016528847587636393, + "loss": 1.1381, + "step": 11677 + }, + { + "epoch": 0.27375071057273626, + "grad_norm": 0.9957375526428223, + "learning_rate": 0.00016528289614955614, + "loss": 0.0533, + "step": 11678 + }, + { + "epoch": 0.27377415214754125, + "grad_norm": 0.7582118511199951, + "learning_rate": 0.00016527731606852148, + "loss": 0.1077, + "step": 11679 + }, + { + "epoch": 0.27379759372234624, + "grad_norm": 2.71523118019104, + "learning_rate": 0.00016527173563329027, + "loss": 1.8629, + "step": 11680 + }, + { + "epoch": 0.27382103529715124, + "grad_norm": 1.9905266761779785, + "learning_rate": 0.00016526615484389282, + "loss": 0.2097, + "step": 11681 + }, + { + "epoch": 0.27384447687195623, + "grad_norm": 2.3262672424316406, + "learning_rate": 0.00016526057370035935, + "loss": 0.2119, + "step": 11682 + }, + { + "epoch": 0.2738679184467612, + "grad_norm": 5.2617573738098145, + "learning_rate": 0.00016525499220272023, + "loss": 0.8679, + "step": 11683 + }, + { + "epoch": 0.2738913600215662, + "grad_norm": 4.418025493621826, + "learning_rate": 0.00016524941035100563, + "loss": 0.5232, + "step": 11684 + }, + { + "epoch": 0.27391480159637127, + "grad_norm": 3.2692208290100098, + "learning_rate": 0.00016524382814524588, + "loss": 0.2492, + "step": 11685 + }, + { + "epoch": 0.27393824317117627, + "grad_norm": 3.36667537689209, + "learning_rate": 0.00016523824558547132, + "loss": 0.3528, + "step": 11686 + }, + { + "epoch": 0.27396168474598126, + "grad_norm": 3.895042896270752, + "learning_rate": 0.0001652326626717122, + "loss": 0.4859, + "step": 11687 + }, + { + "epoch": 0.27398512632078625, + "grad_norm": 1.3804891109466553, + "learning_rate": 0.0001652270794039988, + "loss": 0.1679, + "step": 11688 + }, + { + "epoch": 0.27400856789559125, + "grad_norm": 4.434449672698975, + "learning_rate": 0.00016522149578236146, + "loss": 0.5462, + "step": 11689 + }, + { + "epoch": 0.27403200947039624, + "grad_norm": 3.6037821769714355, + "learning_rate": 0.00016521591180683045, + "loss": 0.5045, + "step": 11690 + }, + { + "epoch": 0.27405545104520124, + "grad_norm": 1.036257266998291, + "learning_rate": 0.00016521032747743606, + "loss": 0.1261, + "step": 11691 + }, + { + "epoch": 0.27407889262000623, + "grad_norm": 5.084034442901611, + "learning_rate": 0.0001652047427942086, + "loss": 0.4082, + "step": 11692 + }, + { + "epoch": 0.2741023341948112, + "grad_norm": 3.8721344470977783, + "learning_rate": 0.00016519915775717835, + "loss": 1.5192, + "step": 11693 + }, + { + "epoch": 0.2741257757696162, + "grad_norm": 3.714193344116211, + "learning_rate": 0.00016519357236637565, + "loss": 0.5408, + "step": 11694 + }, + { + "epoch": 0.2741492173444212, + "grad_norm": 2.4899473190307617, + "learning_rate": 0.0001651879866218308, + "loss": 0.3471, + "step": 11695 + }, + { + "epoch": 0.2741726589192262, + "grad_norm": 1.0796512365341187, + "learning_rate": 0.00016518240052357412, + "loss": 0.1369, + "step": 11696 + }, + { + "epoch": 0.2741961004940312, + "grad_norm": 5.409360885620117, + "learning_rate": 0.00016517681407163585, + "loss": 0.5822, + "step": 11697 + }, + { + "epoch": 0.2742195420688362, + "grad_norm": 3.63012433052063, + "learning_rate": 0.0001651712272660464, + "loss": 0.4241, + "step": 11698 + }, + { + "epoch": 0.2742429836436412, + "grad_norm": 2.8706939220428467, + "learning_rate": 0.00016516564010683605, + "loss": 0.1727, + "step": 11699 + }, + { + "epoch": 0.2742664252184462, + "grad_norm": 4.695960998535156, + "learning_rate": 0.00016516005259403505, + "loss": 0.4479, + "step": 11700 + }, + { + "epoch": 0.2742898667932512, + "grad_norm": 5.339205741882324, + "learning_rate": 0.0001651544647276738, + "loss": 0.3466, + "step": 11701 + }, + { + "epoch": 0.2743133083680562, + "grad_norm": 3.058565378189087, + "learning_rate": 0.0001651488765077826, + "loss": 0.3555, + "step": 11702 + }, + { + "epoch": 0.27433674994286117, + "grad_norm": 4.726642608642578, + "learning_rate": 0.00016514328793439174, + "loss": 0.5953, + "step": 11703 + }, + { + "epoch": 0.27436019151766616, + "grad_norm": 4.95785665512085, + "learning_rate": 0.0001651376990075316, + "loss": 0.6074, + "step": 11704 + }, + { + "epoch": 0.27438363309247116, + "grad_norm": 5.773004531860352, + "learning_rate": 0.00016513210972723247, + "loss": 0.4804, + "step": 11705 + }, + { + "epoch": 0.27440707466727615, + "grad_norm": 3.5947906970977783, + "learning_rate": 0.00016512652009352468, + "loss": 0.1623, + "step": 11706 + }, + { + "epoch": 0.27443051624208115, + "grad_norm": 2.741117238998413, + "learning_rate": 0.00016512093010643856, + "loss": 2.139, + "step": 11707 + }, + { + "epoch": 0.27445395781688614, + "grad_norm": 4.1552534103393555, + "learning_rate": 0.00016511533976600441, + "loss": 0.5271, + "step": 11708 + }, + { + "epoch": 0.27447739939169113, + "grad_norm": 0.5763924717903137, + "learning_rate": 0.00016510974907225264, + "loss": 0.0983, + "step": 11709 + }, + { + "epoch": 0.27450084096649613, + "grad_norm": 1.6082746982574463, + "learning_rate": 0.0001651041580252135, + "loss": 0.1773, + "step": 11710 + }, + { + "epoch": 0.2745242825413011, + "grad_norm": 1.6813290119171143, + "learning_rate": 0.0001650985666249174, + "loss": 0.1628, + "step": 11711 + }, + { + "epoch": 0.2745477241161061, + "grad_norm": 1.1592154502868652, + "learning_rate": 0.0001650929748713946, + "loss": 0.1336, + "step": 11712 + }, + { + "epoch": 0.2745711656909111, + "grad_norm": 3.1040704250335693, + "learning_rate": 0.00016508738276467555, + "loss": 1.3789, + "step": 11713 + }, + { + "epoch": 0.2745946072657161, + "grad_norm": 4.776487827301025, + "learning_rate": 0.0001650817903047905, + "loss": 0.6128, + "step": 11714 + }, + { + "epoch": 0.2746180488405211, + "grad_norm": 3.782550096511841, + "learning_rate": 0.00016507619749176985, + "loss": 0.4599, + "step": 11715 + }, + { + "epoch": 0.2746414904153261, + "grad_norm": 3.4787349700927734, + "learning_rate": 0.00016507060432564388, + "loss": 0.3068, + "step": 11716 + }, + { + "epoch": 0.2746649319901311, + "grad_norm": 5.345816135406494, + "learning_rate": 0.000165065010806443, + "loss": 0.5872, + "step": 11717 + }, + { + "epoch": 0.2746883735649361, + "grad_norm": 5.829366683959961, + "learning_rate": 0.00016505941693419754, + "loss": 0.7081, + "step": 11718 + }, + { + "epoch": 0.2747118151397411, + "grad_norm": 1.3515658378601074, + "learning_rate": 0.0001650538227089379, + "loss": 0.1573, + "step": 11719 + }, + { + "epoch": 0.27473525671454607, + "grad_norm": 2.2637743949890137, + "learning_rate": 0.00016504822813069434, + "loss": 0.1742, + "step": 11720 + }, + { + "epoch": 0.27475869828935107, + "grad_norm": 0.48638367652893066, + "learning_rate": 0.00016504263319949724, + "loss": 0.0398, + "step": 11721 + }, + { + "epoch": 0.27478213986415606, + "grad_norm": 5.572722911834717, + "learning_rate": 0.00016503703791537704, + "loss": 0.4799, + "step": 11722 + }, + { + "epoch": 0.27480558143896106, + "grad_norm": 2.727847099304199, + "learning_rate": 0.00016503144227836402, + "loss": 0.3098, + "step": 11723 + }, + { + "epoch": 0.27482902301376605, + "grad_norm": 2.11106014251709, + "learning_rate": 0.00016502584628848857, + "loss": 0.1934, + "step": 11724 + }, + { + "epoch": 0.27485246458857104, + "grad_norm": 1.4944088459014893, + "learning_rate": 0.00016502024994578105, + "loss": 0.2486, + "step": 11725 + }, + { + "epoch": 0.27487590616337604, + "grad_norm": 4.544767379760742, + "learning_rate": 0.00016501465325027183, + "loss": 0.4301, + "step": 11726 + }, + { + "epoch": 0.27489934773818103, + "grad_norm": 1.9330872297286987, + "learning_rate": 0.00016500905620199125, + "loss": 1.2801, + "step": 11727 + }, + { + "epoch": 0.274922789312986, + "grad_norm": 6.359213829040527, + "learning_rate": 0.00016500345880096973, + "loss": 0.8494, + "step": 11728 + }, + { + "epoch": 0.274946230887791, + "grad_norm": 0.9567705988883972, + "learning_rate": 0.00016499786104723759, + "loss": 0.1065, + "step": 11729 + }, + { + "epoch": 0.274969672462596, + "grad_norm": 3.5357158184051514, + "learning_rate": 0.00016499226294082524, + "loss": 0.3975, + "step": 11730 + }, + { + "epoch": 0.274993114037401, + "grad_norm": 0.9779132604598999, + "learning_rate": 0.00016498666448176307, + "loss": 0.1044, + "step": 11731 + }, + { + "epoch": 0.275016555612206, + "grad_norm": 4.632685661315918, + "learning_rate": 0.00016498106567008138, + "loss": 0.456, + "step": 11732 + }, + { + "epoch": 0.275039997187011, + "grad_norm": 2.0149552822113037, + "learning_rate": 0.00016497546650581064, + "loss": 0.1797, + "step": 11733 + }, + { + "epoch": 0.275063438761816, + "grad_norm": 1.9488389492034912, + "learning_rate": 0.00016496986698898115, + "loss": 0.1531, + "step": 11734 + }, + { + "epoch": 0.275086880336621, + "grad_norm": 4.160424709320068, + "learning_rate": 0.00016496426711962338, + "loss": 0.7162, + "step": 11735 + }, + { + "epoch": 0.27511032191142604, + "grad_norm": 2.4071667194366455, + "learning_rate": 0.00016495866689776766, + "loss": 1.5863, + "step": 11736 + }, + { + "epoch": 0.27513376348623103, + "grad_norm": 2.4477455615997314, + "learning_rate": 0.0001649530663234444, + "loss": 0.2202, + "step": 11737 + }, + { + "epoch": 0.275157205061036, + "grad_norm": 0.4567331373691559, + "learning_rate": 0.00016494746539668395, + "loss": 0.0497, + "step": 11738 + }, + { + "epoch": 0.275180646635841, + "grad_norm": 4.780508041381836, + "learning_rate": 0.00016494186411751676, + "loss": 0.6, + "step": 11739 + }, + { + "epoch": 0.275204088210646, + "grad_norm": 1.553331971168518, + "learning_rate": 0.0001649362624859731, + "loss": 0.1981, + "step": 11740 + }, + { + "epoch": 0.275227529785451, + "grad_norm": 4.5305705070495605, + "learning_rate": 0.00016493066050208354, + "loss": 0.3951, + "step": 11741 + }, + { + "epoch": 0.275250971360256, + "grad_norm": 1.0703076124191284, + "learning_rate": 0.00016492505816587837, + "loss": 0.1725, + "step": 11742 + }, + { + "epoch": 0.275274412935061, + "grad_norm": 2.870835542678833, + "learning_rate": 0.00016491945547738804, + "loss": 0.4513, + "step": 11743 + }, + { + "epoch": 0.275297854509866, + "grad_norm": 3.447300434112549, + "learning_rate": 0.00016491385243664285, + "loss": 0.6333, + "step": 11744 + }, + { + "epoch": 0.275321296084671, + "grad_norm": 0.6024847030639648, + "learning_rate": 0.0001649082490436733, + "loss": 0.0362, + "step": 11745 + }, + { + "epoch": 0.275344737659476, + "grad_norm": 4.388982772827148, + "learning_rate": 0.0001649026452985098, + "loss": 0.6152, + "step": 11746 + }, + { + "epoch": 0.275368179234281, + "grad_norm": 4.637937068939209, + "learning_rate": 0.0001648970412011827, + "loss": 0.471, + "step": 11747 + }, + { + "epoch": 0.27539162080908597, + "grad_norm": 3.5747880935668945, + "learning_rate": 0.0001648914367517224, + "loss": 0.3961, + "step": 11748 + }, + { + "epoch": 0.27541506238389096, + "grad_norm": 2.7770400047302246, + "learning_rate": 0.00016488583195015938, + "loss": 0.2774, + "step": 11749 + }, + { + "epoch": 0.27543850395869596, + "grad_norm": 0.8044431805610657, + "learning_rate": 0.000164880226796524, + "loss": 0.0644, + "step": 11750 + }, + { + "epoch": 0.27546194553350095, + "grad_norm": 5.196315288543701, + "learning_rate": 0.00016487462129084667, + "loss": 0.7589, + "step": 11751 + }, + { + "epoch": 0.27548538710830595, + "grad_norm": 4.2340874671936035, + "learning_rate": 0.0001648690154331578, + "loss": 0.441, + "step": 11752 + }, + { + "epoch": 0.27550882868311094, + "grad_norm": 4.645424842834473, + "learning_rate": 0.0001648634092234879, + "loss": 0.5602, + "step": 11753 + }, + { + "epoch": 0.27553227025791593, + "grad_norm": 1.483383297920227, + "learning_rate": 0.00016485780266186728, + "loss": 0.3105, + "step": 11754 + }, + { + "epoch": 0.27555571183272093, + "grad_norm": 3.9463675022125244, + "learning_rate": 0.0001648521957483264, + "loss": 0.4628, + "step": 11755 + }, + { + "epoch": 0.2755791534075259, + "grad_norm": 2.5696792602539062, + "learning_rate": 0.00016484658848289567, + "loss": 0.2622, + "step": 11756 + }, + { + "epoch": 0.2756025949823309, + "grad_norm": 0.43454957008361816, + "learning_rate": 0.00016484098086560554, + "loss": 0.0365, + "step": 11757 + }, + { + "epoch": 0.2756260365571359, + "grad_norm": 1.4674533605575562, + "learning_rate": 0.0001648353728964864, + "loss": 0.248, + "step": 11758 + }, + { + "epoch": 0.2756494781319409, + "grad_norm": 1.0484580993652344, + "learning_rate": 0.0001648297645755687, + "loss": 0.0753, + "step": 11759 + }, + { + "epoch": 0.2756729197067459, + "grad_norm": 3.562096118927002, + "learning_rate": 0.00016482415590288294, + "loss": 0.4751, + "step": 11760 + }, + { + "epoch": 0.2756963612815509, + "grad_norm": 2.5924017429351807, + "learning_rate": 0.00016481854687845946, + "loss": 0.2063, + "step": 11761 + }, + { + "epoch": 0.2757198028563559, + "grad_norm": 3.8203091621398926, + "learning_rate": 0.0001648129375023287, + "loss": 0.5056, + "step": 11762 + }, + { + "epoch": 0.2757432444311609, + "grad_norm": 4.442646026611328, + "learning_rate": 0.0001648073277745211, + "loss": 0.3653, + "step": 11763 + }, + { + "epoch": 0.2757666860059659, + "grad_norm": 3.1511993408203125, + "learning_rate": 0.00016480171769506715, + "loss": 0.3934, + "step": 11764 + }, + { + "epoch": 0.27579012758077087, + "grad_norm": 4.49225378036499, + "learning_rate": 0.00016479610726399726, + "loss": 0.4653, + "step": 11765 + }, + { + "epoch": 0.27581356915557587, + "grad_norm": 9.574816703796387, + "learning_rate": 0.00016479049648134185, + "loss": 0.4957, + "step": 11766 + }, + { + "epoch": 0.27583701073038086, + "grad_norm": 3.552462100982666, + "learning_rate": 0.00016478488534713142, + "loss": 0.4636, + "step": 11767 + }, + { + "epoch": 0.27586045230518585, + "grad_norm": 3.6994290351867676, + "learning_rate": 0.00016477927386139633, + "loss": 0.2101, + "step": 11768 + }, + { + "epoch": 0.27588389387999085, + "grad_norm": 3.8684728145599365, + "learning_rate": 0.00016477366202416712, + "loss": 0.4739, + "step": 11769 + }, + { + "epoch": 0.27590733545479584, + "grad_norm": 5.77083683013916, + "learning_rate": 0.00016476804983547419, + "loss": 0.3642, + "step": 11770 + }, + { + "epoch": 0.27593077702960084, + "grad_norm": 3.1712324619293213, + "learning_rate": 0.00016476243729534798, + "loss": 0.474, + "step": 11771 + }, + { + "epoch": 0.27595421860440583, + "grad_norm": 3.5853302478790283, + "learning_rate": 0.00016475682440381898, + "loss": 0.4597, + "step": 11772 + }, + { + "epoch": 0.2759776601792108, + "grad_norm": 5.888030052185059, + "learning_rate": 0.00016475121116091764, + "loss": 0.735, + "step": 11773 + }, + { + "epoch": 0.2760011017540158, + "grad_norm": 1.951325535774231, + "learning_rate": 0.00016474559756667442, + "loss": 0.1806, + "step": 11774 + }, + { + "epoch": 0.2760245433288208, + "grad_norm": 1.309533953666687, + "learning_rate": 0.00016473998362111973, + "loss": 0.1568, + "step": 11775 + }, + { + "epoch": 0.2760479849036258, + "grad_norm": 3.6997158527374268, + "learning_rate": 0.00016473436932428409, + "loss": 0.3702, + "step": 11776 + }, + { + "epoch": 0.2760714264784308, + "grad_norm": 4.5228986740112305, + "learning_rate": 0.00016472875467619792, + "loss": 0.558, + "step": 11777 + }, + { + "epoch": 0.2760948680532358, + "grad_norm": 3.5538339614868164, + "learning_rate": 0.0001647231396768917, + "loss": 0.4639, + "step": 11778 + }, + { + "epoch": 0.2761183096280408, + "grad_norm": 1.7783632278442383, + "learning_rate": 0.00016471752432639596, + "loss": 0.1832, + "step": 11779 + }, + { + "epoch": 0.2761417512028458, + "grad_norm": 4.138528823852539, + "learning_rate": 0.00016471190862474107, + "loss": 0.3562, + "step": 11780 + }, + { + "epoch": 0.2761651927776508, + "grad_norm": 3.909104108810425, + "learning_rate": 0.00016470629257195755, + "loss": 0.3442, + "step": 11781 + }, + { + "epoch": 0.2761886343524558, + "grad_norm": 4.221010208129883, + "learning_rate": 0.00016470067616807587, + "loss": 0.5708, + "step": 11782 + }, + { + "epoch": 0.27621207592726077, + "grad_norm": 4.213803768157959, + "learning_rate": 0.0001646950594131265, + "loss": 0.5755, + "step": 11783 + }, + { + "epoch": 0.27623551750206576, + "grad_norm": 4.24983024597168, + "learning_rate": 0.00016468944230713992, + "loss": 0.4292, + "step": 11784 + }, + { + "epoch": 0.27625895907687076, + "grad_norm": 8.146309852600098, + "learning_rate": 0.0001646838248501466, + "loss": 0.3686, + "step": 11785 + }, + { + "epoch": 0.27628240065167575, + "grad_norm": 2.1377177238464355, + "learning_rate": 0.00016467820704217705, + "loss": 0.2326, + "step": 11786 + }, + { + "epoch": 0.2763058422264808, + "grad_norm": 5.246634483337402, + "learning_rate": 0.00016467258888326172, + "loss": 0.7235, + "step": 11787 + }, + { + "epoch": 0.2763292838012858, + "grad_norm": 3.229743719100952, + "learning_rate": 0.00016466697037343108, + "loss": 0.2875, + "step": 11788 + }, + { + "epoch": 0.2763527253760908, + "grad_norm": 3.525642156600952, + "learning_rate": 0.00016466135151271566, + "loss": 0.4451, + "step": 11789 + }, + { + "epoch": 0.2763761669508958, + "grad_norm": 0.3588204085826874, + "learning_rate": 0.00016465573230114595, + "loss": 0.0375, + "step": 11790 + }, + { + "epoch": 0.2763996085257008, + "grad_norm": 2.1729745864868164, + "learning_rate": 0.0001646501127387524, + "loss": 1.2037, + "step": 11791 + }, + { + "epoch": 0.2764230501005058, + "grad_norm": 3.737996816635132, + "learning_rate": 0.0001646444928255655, + "loss": 0.5958, + "step": 11792 + }, + { + "epoch": 0.27644649167531077, + "grad_norm": 1.7165427207946777, + "learning_rate": 0.00016463887256161578, + "loss": 0.1788, + "step": 11793 + }, + { + "epoch": 0.27646993325011576, + "grad_norm": 4.771229267120361, + "learning_rate": 0.00016463325194693371, + "loss": 0.3511, + "step": 11794 + }, + { + "epoch": 0.27649337482492076, + "grad_norm": 3.0575051307678223, + "learning_rate": 0.00016462763098154984, + "loss": 2.07, + "step": 11795 + }, + { + "epoch": 0.27651681639972575, + "grad_norm": 3.5463905334472656, + "learning_rate": 0.0001646220096654946, + "loss": 0.5238, + "step": 11796 + }, + { + "epoch": 0.27654025797453075, + "grad_norm": 0.7289366722106934, + "learning_rate": 0.00016461638799879848, + "loss": 0.076, + "step": 11797 + }, + { + "epoch": 0.27656369954933574, + "grad_norm": 4.227173328399658, + "learning_rate": 0.00016461076598149209, + "loss": 0.6216, + "step": 11798 + }, + { + "epoch": 0.27658714112414073, + "grad_norm": 0.6121331453323364, + "learning_rate": 0.00016460514361360583, + "loss": 0.0649, + "step": 11799 + }, + { + "epoch": 0.27661058269894573, + "grad_norm": 6.232282638549805, + "learning_rate": 0.00016459952089517023, + "loss": 0.4518, + "step": 11800 + }, + { + "epoch": 0.2766340242737507, + "grad_norm": 3.0394375324249268, + "learning_rate": 0.00016459389782621584, + "loss": 0.2113, + "step": 11801 + }, + { + "epoch": 0.2766574658485557, + "grad_norm": 5.157254695892334, + "learning_rate": 0.00016458827440677315, + "loss": 0.4673, + "step": 11802 + }, + { + "epoch": 0.2766809074233607, + "grad_norm": 4.5081787109375, + "learning_rate": 0.00016458265063687264, + "loss": 0.4234, + "step": 11803 + }, + { + "epoch": 0.2767043489981657, + "grad_norm": 2.3389225006103516, + "learning_rate": 0.00016457702651654488, + "loss": 0.2485, + "step": 11804 + }, + { + "epoch": 0.2767277905729707, + "grad_norm": 2.6932380199432373, + "learning_rate": 0.00016457140204582032, + "loss": 0.299, + "step": 11805 + }, + { + "epoch": 0.2767512321477757, + "grad_norm": 2.562570810317993, + "learning_rate": 0.00016456577722472954, + "loss": 0.1706, + "step": 11806 + }, + { + "epoch": 0.2767746737225807, + "grad_norm": 3.0630433559417725, + "learning_rate": 0.00016456015205330304, + "loss": 1.8042, + "step": 11807 + }, + { + "epoch": 0.2767981152973857, + "grad_norm": 0.72077476978302, + "learning_rate": 0.00016455452653157132, + "loss": 0.0819, + "step": 11808 + }, + { + "epoch": 0.2768215568721907, + "grad_norm": 3.286994457244873, + "learning_rate": 0.00016454890065956495, + "loss": 0.2693, + "step": 11809 + }, + { + "epoch": 0.27684499844699567, + "grad_norm": 4.352688312530518, + "learning_rate": 0.00016454327443731447, + "loss": 0.562, + "step": 11810 + }, + { + "epoch": 0.27686844002180067, + "grad_norm": 3.149963855743408, + "learning_rate": 0.00016453764786485028, + "loss": 0.3098, + "step": 11811 + }, + { + "epoch": 0.27689188159660566, + "grad_norm": 3.089362859725952, + "learning_rate": 0.00016453202094220305, + "loss": 0.3986, + "step": 11812 + }, + { + "epoch": 0.27691532317141065, + "grad_norm": 3.689380645751953, + "learning_rate": 0.00016452639366940325, + "loss": 0.6462, + "step": 11813 + }, + { + "epoch": 0.27693876474621565, + "grad_norm": 1.0199850797653198, + "learning_rate": 0.0001645207660464814, + "loss": 0.1556, + "step": 11814 + }, + { + "epoch": 0.27696220632102064, + "grad_norm": 3.6644904613494873, + "learning_rate": 0.0001645151380734681, + "loss": 0.3049, + "step": 11815 + }, + { + "epoch": 0.27698564789582564, + "grad_norm": 1.535426139831543, + "learning_rate": 0.00016450950975039383, + "loss": 0.2354, + "step": 11816 + }, + { + "epoch": 0.27700908947063063, + "grad_norm": 2.7858810424804688, + "learning_rate": 0.00016450388107728912, + "loss": 0.2359, + "step": 11817 + }, + { + "epoch": 0.2770325310454356, + "grad_norm": 2.4572432041168213, + "learning_rate": 0.00016449825205418458, + "loss": 0.371, + "step": 11818 + }, + { + "epoch": 0.2770559726202406, + "grad_norm": 1.8249731063842773, + "learning_rate": 0.00016449262268111066, + "loss": 0.1697, + "step": 11819 + }, + { + "epoch": 0.2770794141950456, + "grad_norm": 3.147878408432007, + "learning_rate": 0.00016448699295809803, + "loss": 0.4251, + "step": 11820 + }, + { + "epoch": 0.2771028557698506, + "grad_norm": 4.033063888549805, + "learning_rate": 0.0001644813628851771, + "loss": 0.3226, + "step": 11821 + }, + { + "epoch": 0.2771262973446556, + "grad_norm": 3.3053956031799316, + "learning_rate": 0.0001644757324623785, + "loss": 2.1108, + "step": 11822 + }, + { + "epoch": 0.2771497389194606, + "grad_norm": 2.596482753753662, + "learning_rate": 0.00016447010168973275, + "loss": 0.2194, + "step": 11823 + }, + { + "epoch": 0.2771731804942656, + "grad_norm": 2.6678476333618164, + "learning_rate": 0.00016446447056727047, + "loss": 0.3234, + "step": 11824 + }, + { + "epoch": 0.2771966220690706, + "grad_norm": 4.533397197723389, + "learning_rate": 0.0001644588390950221, + "loss": 0.5139, + "step": 11825 + }, + { + "epoch": 0.2772200636438756, + "grad_norm": 1.570068359375, + "learning_rate": 0.00016445320727301828, + "loss": 0.1288, + "step": 11826 + }, + { + "epoch": 0.2772435052186806, + "grad_norm": 3.7822630405426025, + "learning_rate": 0.00016444757510128955, + "loss": 0.3919, + "step": 11827 + }, + { + "epoch": 0.27726694679348557, + "grad_norm": 2.1307480335235596, + "learning_rate": 0.00016444194257986646, + "loss": 1.5297, + "step": 11828 + }, + { + "epoch": 0.27729038836829056, + "grad_norm": 1.7077909708023071, + "learning_rate": 0.00016443630970877957, + "loss": 0.2599, + "step": 11829 + }, + { + "epoch": 0.27731382994309556, + "grad_norm": 1.307924509048462, + "learning_rate": 0.00016443067648805944, + "loss": 0.194, + "step": 11830 + }, + { + "epoch": 0.27733727151790055, + "grad_norm": 4.26580286026001, + "learning_rate": 0.00016442504291773666, + "loss": 0.3131, + "step": 11831 + }, + { + "epoch": 0.27736071309270555, + "grad_norm": 6.2613959312438965, + "learning_rate": 0.00016441940899784182, + "loss": 0.5965, + "step": 11832 + }, + { + "epoch": 0.27738415466751054, + "grad_norm": 3.904940605163574, + "learning_rate": 0.0001644137747284054, + "loss": 0.2566, + "step": 11833 + }, + { + "epoch": 0.27740759624231554, + "grad_norm": 4.06005859375, + "learning_rate": 0.00016440814010945805, + "loss": 0.514, + "step": 11834 + }, + { + "epoch": 0.27743103781712053, + "grad_norm": 3.447145938873291, + "learning_rate": 0.00016440250514103034, + "loss": 0.4769, + "step": 11835 + }, + { + "epoch": 0.2774544793919255, + "grad_norm": 5.072834014892578, + "learning_rate": 0.00016439686982315283, + "loss": 0.5637, + "step": 11836 + }, + { + "epoch": 0.2774779209667305, + "grad_norm": 3.8402280807495117, + "learning_rate": 0.00016439123415585603, + "loss": 0.4438, + "step": 11837 + }, + { + "epoch": 0.27750136254153557, + "grad_norm": 3.638636827468872, + "learning_rate": 0.00016438559813917065, + "loss": 0.6244, + "step": 11838 + }, + { + "epoch": 0.27752480411634056, + "grad_norm": 4.601718902587891, + "learning_rate": 0.00016437996177312718, + "loss": 0.6588, + "step": 11839 + }, + { + "epoch": 0.27754824569114556, + "grad_norm": 4.122956275939941, + "learning_rate": 0.00016437432505775622, + "loss": 0.7025, + "step": 11840 + }, + { + "epoch": 0.27757168726595055, + "grad_norm": 4.813201427459717, + "learning_rate": 0.00016436868799308839, + "loss": 0.59, + "step": 11841 + }, + { + "epoch": 0.27759512884075555, + "grad_norm": 8.818459510803223, + "learning_rate": 0.00016436305057915425, + "loss": 0.8577, + "step": 11842 + }, + { + "epoch": 0.27761857041556054, + "grad_norm": 4.295646667480469, + "learning_rate": 0.00016435741281598434, + "loss": 0.4635, + "step": 11843 + }, + { + "epoch": 0.27764201199036553, + "grad_norm": 1.085044503211975, + "learning_rate": 0.0001643517747036093, + "loss": 0.0922, + "step": 11844 + }, + { + "epoch": 0.27766545356517053, + "grad_norm": 5.204589366912842, + "learning_rate": 0.00016434613624205976, + "loss": 0.4433, + "step": 11845 + }, + { + "epoch": 0.2776888951399755, + "grad_norm": 4.559200286865234, + "learning_rate": 0.00016434049743136627, + "loss": 0.7271, + "step": 11846 + }, + { + "epoch": 0.2777123367147805, + "grad_norm": 0.5783183574676514, + "learning_rate": 0.00016433485827155943, + "loss": 0.0989, + "step": 11847 + }, + { + "epoch": 0.2777357782895855, + "grad_norm": 4.044670104980469, + "learning_rate": 0.00016432921876266985, + "loss": 0.2166, + "step": 11848 + }, + { + "epoch": 0.2777592198643905, + "grad_norm": 1.1681814193725586, + "learning_rate": 0.00016432357890472807, + "loss": 0.1232, + "step": 11849 + }, + { + "epoch": 0.2777826614391955, + "grad_norm": 3.6535260677337646, + "learning_rate": 0.0001643179386977648, + "loss": 0.449, + "step": 11850 + }, + { + "epoch": 0.2778061030140005, + "grad_norm": 4.67120361328125, + "learning_rate": 0.00016431229814181054, + "loss": 0.4334, + "step": 11851 + }, + { + "epoch": 0.2778295445888055, + "grad_norm": 6.640106678009033, + "learning_rate": 0.000164306657236896, + "loss": 0.6514, + "step": 11852 + }, + { + "epoch": 0.2778529861636105, + "grad_norm": 2.0391502380371094, + "learning_rate": 0.00016430101598305169, + "loss": 0.1219, + "step": 11853 + }, + { + "epoch": 0.2778764277384155, + "grad_norm": 6.029616832733154, + "learning_rate": 0.00016429537438030826, + "loss": 1.3715, + "step": 11854 + }, + { + "epoch": 0.27789986931322047, + "grad_norm": 3.576155185699463, + "learning_rate": 0.00016428973242869633, + "loss": 0.516, + "step": 11855 + }, + { + "epoch": 0.27792331088802547, + "grad_norm": 4.042182922363281, + "learning_rate": 0.00016428409012824647, + "loss": 0.3445, + "step": 11856 + }, + { + "epoch": 0.27794675246283046, + "grad_norm": 4.261000156402588, + "learning_rate": 0.00016427844747898937, + "loss": 0.4196, + "step": 11857 + }, + { + "epoch": 0.27797019403763545, + "grad_norm": 3.7726962566375732, + "learning_rate": 0.00016427280448095562, + "loss": 0.3817, + "step": 11858 + }, + { + "epoch": 0.27799363561244045, + "grad_norm": 1.4340461492538452, + "learning_rate": 0.00016426716113417578, + "loss": 0.1684, + "step": 11859 + }, + { + "epoch": 0.27801707718724544, + "grad_norm": 2.405831813812256, + "learning_rate": 0.00016426151743868052, + "loss": 0.2263, + "step": 11860 + }, + { + "epoch": 0.27804051876205044, + "grad_norm": 1.9880402088165283, + "learning_rate": 0.00016425587339450047, + "loss": 0.2013, + "step": 11861 + }, + { + "epoch": 0.27806396033685543, + "grad_norm": 3.5864593982696533, + "learning_rate": 0.00016425022900166627, + "loss": 0.46, + "step": 11862 + }, + { + "epoch": 0.2780874019116604, + "grad_norm": 3.4575603008270264, + "learning_rate": 0.00016424458426020851, + "loss": 0.4704, + "step": 11863 + }, + { + "epoch": 0.2781108434864654, + "grad_norm": 1.2915343046188354, + "learning_rate": 0.0001642389391701578, + "loss": 0.1337, + "step": 11864 + }, + { + "epoch": 0.2781342850612704, + "grad_norm": 3.1447365283966064, + "learning_rate": 0.00016423329373154482, + "loss": 1.7987, + "step": 11865 + }, + { + "epoch": 0.2781577266360754, + "grad_norm": 3.6815834045410156, + "learning_rate": 0.00016422764794440023, + "loss": 0.4867, + "step": 11866 + }, + { + "epoch": 0.2781811682108804, + "grad_norm": 3.253899574279785, + "learning_rate": 0.00016422200180875454, + "loss": 0.3236, + "step": 11867 + }, + { + "epoch": 0.2782046097856854, + "grad_norm": 2.3621649742126465, + "learning_rate": 0.0001642163553246385, + "loss": 0.2543, + "step": 11868 + }, + { + "epoch": 0.2782280513604904, + "grad_norm": 3.4923958778381348, + "learning_rate": 0.00016421070849208268, + "loss": 0.3139, + "step": 11869 + }, + { + "epoch": 0.2782514929352954, + "grad_norm": 3.086707830429077, + "learning_rate": 0.00016420506131111778, + "loss": 0.4483, + "step": 11870 + }, + { + "epoch": 0.2782749345101004, + "grad_norm": 0.6821690201759338, + "learning_rate": 0.0001641994137817744, + "loss": 0.0835, + "step": 11871 + }, + { + "epoch": 0.2782983760849054, + "grad_norm": 4.531782627105713, + "learning_rate": 0.00016419376590408322, + "loss": 0.5991, + "step": 11872 + }, + { + "epoch": 0.27832181765971037, + "grad_norm": 1.4357823133468628, + "learning_rate": 0.00016418811767807484, + "loss": 0.1505, + "step": 11873 + }, + { + "epoch": 0.27834525923451536, + "grad_norm": 1.3863394260406494, + "learning_rate": 0.00016418246910377993, + "loss": 0.1968, + "step": 11874 + }, + { + "epoch": 0.27836870080932036, + "grad_norm": 1.3974403142929077, + "learning_rate": 0.00016417682018122912, + "loss": 0.1816, + "step": 11875 + }, + { + "epoch": 0.27839214238412535, + "grad_norm": 3.8030107021331787, + "learning_rate": 0.00016417117091045312, + "loss": 0.4948, + "step": 11876 + }, + { + "epoch": 0.27841558395893035, + "grad_norm": 1.3799368143081665, + "learning_rate": 0.0001641655212914825, + "loss": 0.1533, + "step": 11877 + }, + { + "epoch": 0.27843902553373534, + "grad_norm": 2.3518776893615723, + "learning_rate": 0.000164159871324348, + "loss": 0.216, + "step": 11878 + }, + { + "epoch": 0.27846246710854033, + "grad_norm": 4.096179962158203, + "learning_rate": 0.0001641542210090802, + "loss": 0.4454, + "step": 11879 + }, + { + "epoch": 0.27848590868334533, + "grad_norm": 1.150369644165039, + "learning_rate": 0.0001641485703457098, + "loss": 0.1884, + "step": 11880 + }, + { + "epoch": 0.2785093502581503, + "grad_norm": 1.4553277492523193, + "learning_rate": 0.00016414291933426745, + "loss": 0.052, + "step": 11881 + }, + { + "epoch": 0.2785327918329553, + "grad_norm": 1.331937313079834, + "learning_rate": 0.00016413726797478385, + "loss": 0.1803, + "step": 11882 + }, + { + "epoch": 0.2785562334077603, + "grad_norm": 5.515158653259277, + "learning_rate": 0.0001641316162672896, + "loss": 0.4949, + "step": 11883 + }, + { + "epoch": 0.2785796749825653, + "grad_norm": 4.783926010131836, + "learning_rate": 0.0001641259642118154, + "loss": 0.592, + "step": 11884 + }, + { + "epoch": 0.2786031165573703, + "grad_norm": 3.572340726852417, + "learning_rate": 0.00016412031180839193, + "loss": 0.3104, + "step": 11885 + }, + { + "epoch": 0.2786265581321753, + "grad_norm": 3.0526540279388428, + "learning_rate": 0.0001641146590570498, + "loss": 0.3384, + "step": 11886 + }, + { + "epoch": 0.2786499997069803, + "grad_norm": 2.7812390327453613, + "learning_rate": 0.00016410900595781978, + "loss": 0.3304, + "step": 11887 + }, + { + "epoch": 0.2786734412817853, + "grad_norm": 3.2613635063171387, + "learning_rate": 0.00016410335251073246, + "loss": 0.3332, + "step": 11888 + }, + { + "epoch": 0.27869688285659033, + "grad_norm": 3.500430107116699, + "learning_rate": 0.00016409769871581856, + "loss": 0.4437, + "step": 11889 + }, + { + "epoch": 0.2787203244313953, + "grad_norm": 1.7642326354980469, + "learning_rate": 0.00016409204457310875, + "loss": 0.1043, + "step": 11890 + }, + { + "epoch": 0.2787437660062003, + "grad_norm": 2.5871663093566895, + "learning_rate": 0.0001640863900826337, + "loss": 0.2249, + "step": 11891 + }, + { + "epoch": 0.2787672075810053, + "grad_norm": 1.723088026046753, + "learning_rate": 0.00016408073524442406, + "loss": 0.2291, + "step": 11892 + }, + { + "epoch": 0.2787906491558103, + "grad_norm": 3.4374241828918457, + "learning_rate": 0.0001640750800585106, + "loss": 0.4133, + "step": 11893 + }, + { + "epoch": 0.2788140907306153, + "grad_norm": 1.0010406970977783, + "learning_rate": 0.0001640694245249239, + "loss": 0.0959, + "step": 11894 + }, + { + "epoch": 0.2788375323054203, + "grad_norm": 0.9447909593582153, + "learning_rate": 0.00016406376864369477, + "loss": 0.0829, + "step": 11895 + }, + { + "epoch": 0.2788609738802253, + "grad_norm": 5.6235737800598145, + "learning_rate": 0.00016405811241485379, + "loss": 0.8716, + "step": 11896 + }, + { + "epoch": 0.2788844154550303, + "grad_norm": 1.093934416770935, + "learning_rate": 0.0001640524558384317, + "loss": 0.1274, + "step": 11897 + }, + { + "epoch": 0.2789078570298353, + "grad_norm": 5.998189926147461, + "learning_rate": 0.00016404679891445919, + "loss": 0.7196, + "step": 11898 + }, + { + "epoch": 0.2789312986046403, + "grad_norm": 3.9994494915008545, + "learning_rate": 0.0001640411416429669, + "loss": 0.4824, + "step": 11899 + }, + { + "epoch": 0.27895474017944527, + "grad_norm": 4.933227062225342, + "learning_rate": 0.00016403548402398564, + "loss": 0.4002, + "step": 11900 + }, + { + "epoch": 0.27897818175425027, + "grad_norm": 7.092327117919922, + "learning_rate": 0.00016402982605754599, + "loss": 0.5844, + "step": 11901 + }, + { + "epoch": 0.27900162332905526, + "grad_norm": 3.1026148796081543, + "learning_rate": 0.00016402416774367876, + "loss": 0.4144, + "step": 11902 + }, + { + "epoch": 0.27902506490386025, + "grad_norm": 1.59959077835083, + "learning_rate": 0.00016401850908241453, + "loss": 0.7911, + "step": 11903 + }, + { + "epoch": 0.27904850647866525, + "grad_norm": 4.346713066101074, + "learning_rate": 0.00016401285007378412, + "loss": 0.2384, + "step": 11904 + }, + { + "epoch": 0.27907194805347024, + "grad_norm": 1.653442144393921, + "learning_rate": 0.00016400719071781817, + "loss": 0.187, + "step": 11905 + }, + { + "epoch": 0.27909538962827524, + "grad_norm": 3.0317835807800293, + "learning_rate": 0.0001640015310145474, + "loss": 0.351, + "step": 11906 + }, + { + "epoch": 0.27911883120308023, + "grad_norm": 4.693864345550537, + "learning_rate": 0.00016399587096400256, + "loss": 0.5299, + "step": 11907 + }, + { + "epoch": 0.2791422727778852, + "grad_norm": 2.4839770793914795, + "learning_rate": 0.00016399021056621427, + "loss": 0.2784, + "step": 11908 + }, + { + "epoch": 0.2791657143526902, + "grad_norm": 2.569459915161133, + "learning_rate": 0.00016398454982121332, + "loss": 0.1876, + "step": 11909 + }, + { + "epoch": 0.2791891559274952, + "grad_norm": 4.494004726409912, + "learning_rate": 0.00016397888872903037, + "loss": 0.4823, + "step": 11910 + }, + { + "epoch": 0.2792125975023002, + "grad_norm": 1.2278214693069458, + "learning_rate": 0.00016397322728969624, + "loss": 0.198, + "step": 11911 + }, + { + "epoch": 0.2792360390771052, + "grad_norm": 1.4206769466400146, + "learning_rate": 0.00016396756550324156, + "loss": 0.0996, + "step": 11912 + }, + { + "epoch": 0.2792594806519102, + "grad_norm": 3.7325806617736816, + "learning_rate": 0.00016396190336969704, + "loss": 0.2011, + "step": 11913 + }, + { + "epoch": 0.2792829222267152, + "grad_norm": 4.019488334655762, + "learning_rate": 0.00016395624088909344, + "loss": 0.401, + "step": 11914 + }, + { + "epoch": 0.2793063638015202, + "grad_norm": 4.896472930908203, + "learning_rate": 0.0001639505780614615, + "loss": 0.5446, + "step": 11915 + }, + { + "epoch": 0.2793298053763252, + "grad_norm": 1.1819851398468018, + "learning_rate": 0.0001639449148868319, + "loss": 0.1019, + "step": 11916 + }, + { + "epoch": 0.2793532469511302, + "grad_norm": 3.2614047527313232, + "learning_rate": 0.0001639392513652354, + "loss": 0.4495, + "step": 11917 + }, + { + "epoch": 0.27937668852593517, + "grad_norm": 4.55186653137207, + "learning_rate": 0.00016393358749670274, + "loss": 0.3754, + "step": 11918 + }, + { + "epoch": 0.27940013010074016, + "grad_norm": 0.8096957802772522, + "learning_rate": 0.00016392792328126464, + "loss": 0.0621, + "step": 11919 + }, + { + "epoch": 0.27942357167554516, + "grad_norm": 4.033642292022705, + "learning_rate": 0.00016392225871895183, + "loss": 0.3267, + "step": 11920 + }, + { + "epoch": 0.27944701325035015, + "grad_norm": 3.2648496627807617, + "learning_rate": 0.00016391659380979502, + "loss": 0.276, + "step": 11921 + }, + { + "epoch": 0.27947045482515515, + "grad_norm": 2.9588418006896973, + "learning_rate": 0.00016391092855382497, + "loss": 0.2146, + "step": 11922 + }, + { + "epoch": 0.27949389639996014, + "grad_norm": 5.931580066680908, + "learning_rate": 0.00016390526295107245, + "loss": 0.478, + "step": 11923 + }, + { + "epoch": 0.27951733797476513, + "grad_norm": 1.3951414823532104, + "learning_rate": 0.00016389959700156817, + "loss": 0.1927, + "step": 11924 + }, + { + "epoch": 0.27954077954957013, + "grad_norm": 1.9271212816238403, + "learning_rate": 0.0001638939307053429, + "loss": 0.9923, + "step": 11925 + }, + { + "epoch": 0.2795642211243751, + "grad_norm": 2.555861711502075, + "learning_rate": 0.00016388826406242733, + "loss": 0.2621, + "step": 11926 + }, + { + "epoch": 0.2795876626991801, + "grad_norm": 2.655851364135742, + "learning_rate": 0.00016388259707285227, + "loss": 1.2112, + "step": 11927 + }, + { + "epoch": 0.2796111042739851, + "grad_norm": 3.996485948562622, + "learning_rate": 0.0001638769297366484, + "loss": 0.5277, + "step": 11928 + }, + { + "epoch": 0.2796345458487901, + "grad_norm": 0.8860064148902893, + "learning_rate": 0.00016387126205384656, + "loss": 0.1104, + "step": 11929 + }, + { + "epoch": 0.2796579874235951, + "grad_norm": 3.79239821434021, + "learning_rate": 0.00016386559402447747, + "loss": 0.479, + "step": 11930 + }, + { + "epoch": 0.2796814289984001, + "grad_norm": 3.5654456615448, + "learning_rate": 0.00016385992564857182, + "loss": 0.3117, + "step": 11931 + }, + { + "epoch": 0.2797048705732051, + "grad_norm": 6.044042587280273, + "learning_rate": 0.00016385425692616042, + "loss": 0.6952, + "step": 11932 + }, + { + "epoch": 0.2797283121480101, + "grad_norm": 3.9924166202545166, + "learning_rate": 0.00016384858785727406, + "loss": 0.5925, + "step": 11933 + }, + { + "epoch": 0.2797517537228151, + "grad_norm": 4.791006088256836, + "learning_rate": 0.00016384291844194344, + "loss": 0.4334, + "step": 11934 + }, + { + "epoch": 0.27977519529762007, + "grad_norm": 2.750216007232666, + "learning_rate": 0.00016383724868019938, + "loss": 0.2986, + "step": 11935 + }, + { + "epoch": 0.27979863687242507, + "grad_norm": 2.374952554702759, + "learning_rate": 0.00016383157857207257, + "loss": 0.2384, + "step": 11936 + }, + { + "epoch": 0.27982207844723006, + "grad_norm": 4.3760504722595215, + "learning_rate": 0.00016382590811759386, + "loss": 0.5124, + "step": 11937 + }, + { + "epoch": 0.27984552002203505, + "grad_norm": 1.3628660440444946, + "learning_rate": 0.00016382023731679397, + "loss": 0.1395, + "step": 11938 + }, + { + "epoch": 0.27986896159684005, + "grad_norm": 4.760045528411865, + "learning_rate": 0.0001638145661697037, + "loss": 0.3494, + "step": 11939 + }, + { + "epoch": 0.2798924031716451, + "grad_norm": 3.021390199661255, + "learning_rate": 0.00016380889467635371, + "loss": 1.0943, + "step": 11940 + }, + { + "epoch": 0.2799158447464501, + "grad_norm": 3.7804477214813232, + "learning_rate": 0.00016380322283677496, + "loss": 0.3329, + "step": 11941 + }, + { + "epoch": 0.2799392863212551, + "grad_norm": 7.535068035125732, + "learning_rate": 0.00016379755065099808, + "loss": 0.5732, + "step": 11942 + }, + { + "epoch": 0.2799627278960601, + "grad_norm": 1.8542633056640625, + "learning_rate": 0.00016379187811905392, + "loss": 0.1185, + "step": 11943 + }, + { + "epoch": 0.2799861694708651, + "grad_norm": 3.4873406887054443, + "learning_rate": 0.00016378620524097324, + "loss": 0.3766, + "step": 11944 + }, + { + "epoch": 0.28000961104567007, + "grad_norm": 3.101522445678711, + "learning_rate": 0.0001637805320167868, + "loss": 0.2786, + "step": 11945 + }, + { + "epoch": 0.28003305262047506, + "grad_norm": 2.5718955993652344, + "learning_rate": 0.0001637748584465254, + "loss": 0.2914, + "step": 11946 + }, + { + "epoch": 0.28005649419528006, + "grad_norm": 4.151373386383057, + "learning_rate": 0.00016376918453021982, + "loss": 0.2463, + "step": 11947 + }, + { + "epoch": 0.28007993577008505, + "grad_norm": 3.153489112854004, + "learning_rate": 0.00016376351026790083, + "loss": 0.1443, + "step": 11948 + }, + { + "epoch": 0.28010337734489005, + "grad_norm": 5.9818878173828125, + "learning_rate": 0.0001637578356595993, + "loss": 0.5259, + "step": 11949 + }, + { + "epoch": 0.28012681891969504, + "grad_norm": 3.1886651515960693, + "learning_rate": 0.00016375216070534595, + "loss": 0.2752, + "step": 11950 + }, + { + "epoch": 0.28015026049450004, + "grad_norm": 3.3241047859191895, + "learning_rate": 0.00016374648540517156, + "loss": 0.5516, + "step": 11951 + }, + { + "epoch": 0.28017370206930503, + "grad_norm": 5.050118923187256, + "learning_rate": 0.00016374080975910694, + "loss": 0.4332, + "step": 11952 + }, + { + "epoch": 0.28019714364411, + "grad_norm": 2.43755841255188, + "learning_rate": 0.0001637351337671829, + "loss": 0.2709, + "step": 11953 + }, + { + "epoch": 0.280220585218915, + "grad_norm": 0.5843149423599243, + "learning_rate": 0.00016372945742943023, + "loss": 0.054, + "step": 11954 + }, + { + "epoch": 0.28024402679372, + "grad_norm": 4.213972568511963, + "learning_rate": 0.00016372378074587976, + "loss": 0.231, + "step": 11955 + }, + { + "epoch": 0.280267468368525, + "grad_norm": 5.361030578613281, + "learning_rate": 0.00016371810371656223, + "loss": 0.631, + "step": 11956 + }, + { + "epoch": 0.28029090994333, + "grad_norm": 2.605189561843872, + "learning_rate": 0.0001637124263415085, + "loss": 0.2859, + "step": 11957 + }, + { + "epoch": 0.280314351518135, + "grad_norm": 2.9014484882354736, + "learning_rate": 0.00016370674862074933, + "loss": 0.1848, + "step": 11958 + }, + { + "epoch": 0.28033779309294, + "grad_norm": 5.28140115737915, + "learning_rate": 0.00016370107055431557, + "loss": 0.4479, + "step": 11959 + }, + { + "epoch": 0.280361234667745, + "grad_norm": 1.204899549484253, + "learning_rate": 0.000163695392142238, + "loss": 0.0786, + "step": 11960 + }, + { + "epoch": 0.28038467624255, + "grad_norm": 4.439672946929932, + "learning_rate": 0.00016368971338454746, + "loss": 1.6973, + "step": 11961 + }, + { + "epoch": 0.280408117817355, + "grad_norm": 4.056375980377197, + "learning_rate": 0.00016368403428127472, + "loss": 0.3309, + "step": 11962 + }, + { + "epoch": 0.28043155939215997, + "grad_norm": 0.8552486300468445, + "learning_rate": 0.00016367835483245064, + "loss": 0.0591, + "step": 11963 + }, + { + "epoch": 0.28045500096696496, + "grad_norm": 0.7000647783279419, + "learning_rate": 0.000163672675038106, + "loss": 0.0507, + "step": 11964 + }, + { + "epoch": 0.28047844254176996, + "grad_norm": 1.7231696844100952, + "learning_rate": 0.00016366699489827165, + "loss": 0.2053, + "step": 11965 + }, + { + "epoch": 0.28050188411657495, + "grad_norm": 1.5359371900558472, + "learning_rate": 0.00016366131441297838, + "loss": 0.1917, + "step": 11966 + }, + { + "epoch": 0.28052532569137995, + "grad_norm": 3.369607925415039, + "learning_rate": 0.00016365563358225704, + "loss": 0.3012, + "step": 11967 + }, + { + "epoch": 0.28054876726618494, + "grad_norm": 5.049404621124268, + "learning_rate": 0.00016364995240613842, + "loss": 0.4838, + "step": 11968 + }, + { + "epoch": 0.28057220884098993, + "grad_norm": 3.643632650375366, + "learning_rate": 0.00016364427088465342, + "loss": 0.4215, + "step": 11969 + }, + { + "epoch": 0.28059565041579493, + "grad_norm": 2.3399271965026855, + "learning_rate": 0.00016363858901783277, + "loss": 0.2859, + "step": 11970 + }, + { + "epoch": 0.2806190919905999, + "grad_norm": 2.941676616668701, + "learning_rate": 0.00016363290680570735, + "loss": 0.2543, + "step": 11971 + }, + { + "epoch": 0.2806425335654049, + "grad_norm": 3.4982144832611084, + "learning_rate": 0.00016362722424830798, + "loss": 0.4634, + "step": 11972 + }, + { + "epoch": 0.2806659751402099, + "grad_norm": 3.0090770721435547, + "learning_rate": 0.00016362154134566553, + "loss": 2.2034, + "step": 11973 + }, + { + "epoch": 0.2806894167150149, + "grad_norm": 5.242271423339844, + "learning_rate": 0.0001636158580978108, + "loss": 0.6251, + "step": 11974 + }, + { + "epoch": 0.2807128582898199, + "grad_norm": 1.853679895401001, + "learning_rate": 0.00016361017450477466, + "loss": 0.1598, + "step": 11975 + }, + { + "epoch": 0.2807362998646249, + "grad_norm": 1.958455204963684, + "learning_rate": 0.0001636044905665879, + "loss": 0.174, + "step": 11976 + }, + { + "epoch": 0.2807597414394299, + "grad_norm": 4.97987699508667, + "learning_rate": 0.00016359880628328134, + "loss": 0.5536, + "step": 11977 + }, + { + "epoch": 0.2807831830142349, + "grad_norm": 5.844173908233643, + "learning_rate": 0.00016359312165488592, + "loss": 0.5895, + "step": 11978 + }, + { + "epoch": 0.2808066245890399, + "grad_norm": 3.128035306930542, + "learning_rate": 0.00016358743668143245, + "loss": 1.7634, + "step": 11979 + }, + { + "epoch": 0.28083006616384487, + "grad_norm": 3.946955442428589, + "learning_rate": 0.00016358175136295172, + "loss": 0.262, + "step": 11980 + }, + { + "epoch": 0.28085350773864987, + "grad_norm": 3.429810047149658, + "learning_rate": 0.00016357606569947466, + "loss": 0.2778, + "step": 11981 + }, + { + "epoch": 0.28087694931345486, + "grad_norm": 4.941678047180176, + "learning_rate": 0.00016357037969103204, + "loss": 0.5266, + "step": 11982 + }, + { + "epoch": 0.28090039088825985, + "grad_norm": 3.6320412158966064, + "learning_rate": 0.00016356469333765475, + "loss": 0.4068, + "step": 11983 + }, + { + "epoch": 0.28092383246306485, + "grad_norm": 1.418473482131958, + "learning_rate": 0.0001635590066393737, + "loss": 0.1754, + "step": 11984 + }, + { + "epoch": 0.28094727403786984, + "grad_norm": 3.6873223781585693, + "learning_rate": 0.00016355331959621964, + "loss": 0.4259, + "step": 11985 + }, + { + "epoch": 0.28097071561267484, + "grad_norm": 3.0210835933685303, + "learning_rate": 0.0001635476322082235, + "loss": 0.2125, + "step": 11986 + }, + { + "epoch": 0.28099415718747983, + "grad_norm": 0.7550795674324036, + "learning_rate": 0.00016354194447541613, + "loss": 0.0644, + "step": 11987 + }, + { + "epoch": 0.2810175987622848, + "grad_norm": 0.7351778149604797, + "learning_rate": 0.00016353625639782836, + "loss": 0.0383, + "step": 11988 + }, + { + "epoch": 0.2810410403370898, + "grad_norm": 3.572455883026123, + "learning_rate": 0.00016353056797549108, + "loss": 0.4723, + "step": 11989 + }, + { + "epoch": 0.2810644819118948, + "grad_norm": 3.4654695987701416, + "learning_rate": 0.00016352487920843518, + "loss": 0.3522, + "step": 11990 + }, + { + "epoch": 0.2810879234866998, + "grad_norm": 4.52689266204834, + "learning_rate": 0.00016351919009669147, + "loss": 0.4658, + "step": 11991 + }, + { + "epoch": 0.28111136506150486, + "grad_norm": 1.4241935014724731, + "learning_rate": 0.00016351350064029084, + "loss": 0.1126, + "step": 11992 + }, + { + "epoch": 0.28113480663630985, + "grad_norm": 4.227110385894775, + "learning_rate": 0.0001635078108392642, + "loss": 0.5859, + "step": 11993 + }, + { + "epoch": 0.28115824821111485, + "grad_norm": 1.311898946762085, + "learning_rate": 0.00016350212069364238, + "loss": 0.1267, + "step": 11994 + }, + { + "epoch": 0.28118168978591984, + "grad_norm": 4.369434356689453, + "learning_rate": 0.00016349643020345628, + "loss": 0.3114, + "step": 11995 + }, + { + "epoch": 0.28120513136072484, + "grad_norm": 4.433341979980469, + "learning_rate": 0.00016349073936873672, + "loss": 0.5841, + "step": 11996 + }, + { + "epoch": 0.28122857293552983, + "grad_norm": 2.914567232131958, + "learning_rate": 0.00016348504818951466, + "loss": 0.3266, + "step": 11997 + }, + { + "epoch": 0.2812520145103348, + "grad_norm": 3.7371153831481934, + "learning_rate": 0.00016347935666582097, + "loss": 0.2634, + "step": 11998 + }, + { + "epoch": 0.2812754560851398, + "grad_norm": 3.1902987957000732, + "learning_rate": 0.00016347366479768646, + "loss": 0.3576, + "step": 11999 + }, + { + "epoch": 0.2812988976599448, + "grad_norm": 3.9321703910827637, + "learning_rate": 0.00016346797258514207, + "loss": 0.7328, + "step": 12000 + }, + { + "epoch": 0.2813223392347498, + "grad_norm": 3.9443817138671875, + "learning_rate": 0.00016346228002821867, + "loss": 0.4718, + "step": 12001 + }, + { + "epoch": 0.2813457808095548, + "grad_norm": 2.765082359313965, + "learning_rate": 0.0001634565871269472, + "loss": 0.1768, + "step": 12002 + }, + { + "epoch": 0.2813692223843598, + "grad_norm": 2.864250421524048, + "learning_rate": 0.00016345089388135846, + "loss": 2.1305, + "step": 12003 + }, + { + "epoch": 0.2813926639591648, + "grad_norm": 1.0128557682037354, + "learning_rate": 0.00016344520029148337, + "loss": 0.109, + "step": 12004 + }, + { + "epoch": 0.2814161055339698, + "grad_norm": 3.515653371810913, + "learning_rate": 0.0001634395063573529, + "loss": 0.347, + "step": 12005 + }, + { + "epoch": 0.2814395471087748, + "grad_norm": 5.089418411254883, + "learning_rate": 0.00016343381207899784, + "loss": 0.3016, + "step": 12006 + }, + { + "epoch": 0.2814629886835798, + "grad_norm": 3.5413997173309326, + "learning_rate": 0.00016342811745644913, + "loss": 0.3026, + "step": 12007 + }, + { + "epoch": 0.28148643025838477, + "grad_norm": 2.5937070846557617, + "learning_rate": 0.0001634224224897377, + "loss": 0.3944, + "step": 12008 + }, + { + "epoch": 0.28150987183318976, + "grad_norm": 0.8249749541282654, + "learning_rate": 0.0001634167271788944, + "loss": 0.0648, + "step": 12009 + }, + { + "epoch": 0.28153331340799476, + "grad_norm": 4.533718585968018, + "learning_rate": 0.00016341103152395014, + "loss": 0.3804, + "step": 12010 + }, + { + "epoch": 0.28155675498279975, + "grad_norm": 1.6821593046188354, + "learning_rate": 0.00016340533552493586, + "loss": 0.1744, + "step": 12011 + }, + { + "epoch": 0.28158019655760474, + "grad_norm": 4.228200435638428, + "learning_rate": 0.00016339963918188247, + "loss": 0.5245, + "step": 12012 + }, + { + "epoch": 0.28160363813240974, + "grad_norm": 6.081993103027344, + "learning_rate": 0.0001633939424948208, + "loss": 0.3938, + "step": 12013 + }, + { + "epoch": 0.28162707970721473, + "grad_norm": 4.062871932983398, + "learning_rate": 0.00016338824546378186, + "loss": 0.7317, + "step": 12014 + }, + { + "epoch": 0.28165052128201973, + "grad_norm": 3.1869006156921387, + "learning_rate": 0.0001633825480887965, + "loss": 1.6483, + "step": 12015 + }, + { + "epoch": 0.2816739628568247, + "grad_norm": 3.935791015625, + "learning_rate": 0.00016337685036989566, + "loss": 0.4816, + "step": 12016 + }, + { + "epoch": 0.2816974044316297, + "grad_norm": 5.214559555053711, + "learning_rate": 0.00016337115230711027, + "loss": 0.6262, + "step": 12017 + }, + { + "epoch": 0.2817208460064347, + "grad_norm": 2.1099555492401123, + "learning_rate": 0.0001633654539004712, + "loss": 0.2169, + "step": 12018 + }, + { + "epoch": 0.2817442875812397, + "grad_norm": 2.4270215034484863, + "learning_rate": 0.00016335975515000936, + "loss": 0.1762, + "step": 12019 + }, + { + "epoch": 0.2817677291560447, + "grad_norm": 1.2101086378097534, + "learning_rate": 0.00016335405605575576, + "loss": 0.1658, + "step": 12020 + }, + { + "epoch": 0.2817911707308497, + "grad_norm": 4.7626447677612305, + "learning_rate": 0.00016334835661774126, + "loss": 0.6457, + "step": 12021 + }, + { + "epoch": 0.2818146123056547, + "grad_norm": 5.642075061798096, + "learning_rate": 0.00016334265683599678, + "loss": 0.6238, + "step": 12022 + }, + { + "epoch": 0.2818380538804597, + "grad_norm": 2.780057430267334, + "learning_rate": 0.0001633369567105533, + "loss": 0.2471, + "step": 12023 + }, + { + "epoch": 0.2818614954552647, + "grad_norm": 2.5307366847991943, + "learning_rate": 0.00016333125624144165, + "loss": 0.305, + "step": 12024 + }, + { + "epoch": 0.28188493703006967, + "grad_norm": 4.468667030334473, + "learning_rate": 0.00016332555542869287, + "loss": 0.5861, + "step": 12025 + }, + { + "epoch": 0.28190837860487467, + "grad_norm": 1.4517751932144165, + "learning_rate": 0.00016331985427233783, + "loss": 0.1595, + "step": 12026 + }, + { + "epoch": 0.28193182017967966, + "grad_norm": 4.416571140289307, + "learning_rate": 0.00016331415277240748, + "loss": 0.9089, + "step": 12027 + }, + { + "epoch": 0.28195526175448465, + "grad_norm": 2.093569278717041, + "learning_rate": 0.00016330845092893276, + "loss": 1.1122, + "step": 12028 + }, + { + "epoch": 0.28197870332928965, + "grad_norm": 3.5417821407318115, + "learning_rate": 0.0001633027487419446, + "loss": 0.3016, + "step": 12029 + }, + { + "epoch": 0.28200214490409464, + "grad_norm": 2.454735040664673, + "learning_rate": 0.00016329704621147396, + "loss": 2.2493, + "step": 12030 + }, + { + "epoch": 0.28202558647889964, + "grad_norm": 2.8262548446655273, + "learning_rate": 0.00016329134333755175, + "loss": 0.2118, + "step": 12031 + }, + { + "epoch": 0.28204902805370463, + "grad_norm": 0.743179440498352, + "learning_rate": 0.00016328564012020895, + "loss": 0.0696, + "step": 12032 + }, + { + "epoch": 0.2820724696285096, + "grad_norm": 4.884112358093262, + "learning_rate": 0.00016327993655947646, + "loss": 0.7876, + "step": 12033 + }, + { + "epoch": 0.2820959112033146, + "grad_norm": 3.5804250240325928, + "learning_rate": 0.00016327423265538528, + "loss": 0.3041, + "step": 12034 + }, + { + "epoch": 0.2821193527781196, + "grad_norm": 1.5424550771713257, + "learning_rate": 0.00016326852840796633, + "loss": 0.2383, + "step": 12035 + }, + { + "epoch": 0.2821427943529246, + "grad_norm": 2.8763599395751953, + "learning_rate": 0.00016326282381725057, + "loss": 0.42, + "step": 12036 + }, + { + "epoch": 0.2821662359277296, + "grad_norm": 2.2851455211639404, + "learning_rate": 0.0001632571188832689, + "loss": 1.3607, + "step": 12037 + }, + { + "epoch": 0.2821896775025346, + "grad_norm": 1.233875036239624, + "learning_rate": 0.00016325141360605237, + "loss": 0.0566, + "step": 12038 + }, + { + "epoch": 0.2822131190773396, + "grad_norm": 6.599882125854492, + "learning_rate": 0.0001632457079856319, + "loss": 0.4557, + "step": 12039 + }, + { + "epoch": 0.2822365606521446, + "grad_norm": 4.3303303718566895, + "learning_rate": 0.00016324000202203845, + "loss": 0.6464, + "step": 12040 + }, + { + "epoch": 0.2822600022269496, + "grad_norm": 2.937556743621826, + "learning_rate": 0.00016323429571530294, + "loss": 0.3016, + "step": 12041 + }, + { + "epoch": 0.2822834438017546, + "grad_norm": 1.948973536491394, + "learning_rate": 0.0001632285890654564, + "loss": 0.2286, + "step": 12042 + }, + { + "epoch": 0.2823068853765596, + "grad_norm": 3.9444899559020996, + "learning_rate": 0.0001632228820725297, + "loss": 0.5205, + "step": 12043 + }, + { + "epoch": 0.2823303269513646, + "grad_norm": 1.9234592914581299, + "learning_rate": 0.0001632171747365539, + "loss": 0.1769, + "step": 12044 + }, + { + "epoch": 0.2823537685261696, + "grad_norm": 2.687767744064331, + "learning_rate": 0.0001632114670575599, + "loss": 0.2714, + "step": 12045 + }, + { + "epoch": 0.2823772101009746, + "grad_norm": 1.1260707378387451, + "learning_rate": 0.00016320575903557875, + "loss": 0.1873, + "step": 12046 + }, + { + "epoch": 0.2824006516757796, + "grad_norm": 3.396545886993408, + "learning_rate": 0.00016320005067064136, + "loss": 0.5422, + "step": 12047 + }, + { + "epoch": 0.2824240932505846, + "grad_norm": 2.576338529586792, + "learning_rate": 0.0001631943419627787, + "loss": 0.3804, + "step": 12048 + }, + { + "epoch": 0.2824475348253896, + "grad_norm": 3.4757797718048096, + "learning_rate": 0.0001631886329120218, + "loss": 0.3812, + "step": 12049 + }, + { + "epoch": 0.2824709764001946, + "grad_norm": 4.712403297424316, + "learning_rate": 0.00016318292351840153, + "loss": 0.3601, + "step": 12050 + }, + { + "epoch": 0.2824944179749996, + "grad_norm": 2.457105875015259, + "learning_rate": 0.000163177213781949, + "loss": 0.2739, + "step": 12051 + }, + { + "epoch": 0.2825178595498046, + "grad_norm": 2.060117721557617, + "learning_rate": 0.00016317150370269513, + "loss": 1.6281, + "step": 12052 + }, + { + "epoch": 0.28254130112460957, + "grad_norm": 1.674613118171692, + "learning_rate": 0.0001631657932806709, + "loss": 0.204, + "step": 12053 + }, + { + "epoch": 0.28256474269941456, + "grad_norm": 2.2982826232910156, + "learning_rate": 0.00016316008251590728, + "loss": 1.4627, + "step": 12054 + }, + { + "epoch": 0.28258818427421956, + "grad_norm": 2.995060920715332, + "learning_rate": 0.0001631543714084353, + "loss": 0.2634, + "step": 12055 + }, + { + "epoch": 0.28261162584902455, + "grad_norm": 5.358144283294678, + "learning_rate": 0.0001631486599582859, + "loss": 0.5134, + "step": 12056 + }, + { + "epoch": 0.28263506742382954, + "grad_norm": 2.7659082412719727, + "learning_rate": 0.00016314294816549012, + "loss": 0.2613, + "step": 12057 + }, + { + "epoch": 0.28265850899863454, + "grad_norm": 5.295100688934326, + "learning_rate": 0.0001631372360300789, + "loss": 0.2264, + "step": 12058 + }, + { + "epoch": 0.28268195057343953, + "grad_norm": 1.2507890462875366, + "learning_rate": 0.00016313152355208329, + "loss": 0.125, + "step": 12059 + }, + { + "epoch": 0.2827053921482445, + "grad_norm": 4.986114025115967, + "learning_rate": 0.00016312581073153423, + "loss": 0.5307, + "step": 12060 + }, + { + "epoch": 0.2827288337230495, + "grad_norm": 0.7451363801956177, + "learning_rate": 0.00016312009756846277, + "loss": 0.0799, + "step": 12061 + }, + { + "epoch": 0.2827522752978545, + "grad_norm": 3.0338847637176514, + "learning_rate": 0.00016311438406289987, + "loss": 0.3991, + "step": 12062 + }, + { + "epoch": 0.2827757168726595, + "grad_norm": 2.6928844451904297, + "learning_rate": 0.00016310867021487657, + "loss": 0.2945, + "step": 12063 + }, + { + "epoch": 0.2827991584474645, + "grad_norm": 1.3724607229232788, + "learning_rate": 0.00016310295602442384, + "loss": 0.1735, + "step": 12064 + }, + { + "epoch": 0.2828226000222695, + "grad_norm": 3.455836057662964, + "learning_rate": 0.0001630972414915727, + "loss": 0.2004, + "step": 12065 + }, + { + "epoch": 0.2828460415970745, + "grad_norm": 2.8855714797973633, + "learning_rate": 0.00016309152661635415, + "loss": 1.5599, + "step": 12066 + }, + { + "epoch": 0.2828694831718795, + "grad_norm": 3.510080337524414, + "learning_rate": 0.00016308581139879918, + "loss": 0.2761, + "step": 12067 + }, + { + "epoch": 0.2828929247466845, + "grad_norm": 3.6474602222442627, + "learning_rate": 0.00016308009583893885, + "loss": 0.642, + "step": 12068 + }, + { + "epoch": 0.2829163663214895, + "grad_norm": 4.202176094055176, + "learning_rate": 0.00016307437993680414, + "loss": 0.3109, + "step": 12069 + }, + { + "epoch": 0.28293980789629447, + "grad_norm": 0.8113855719566345, + "learning_rate": 0.0001630686636924261, + "loss": 0.0433, + "step": 12070 + }, + { + "epoch": 0.28296324947109947, + "grad_norm": 1.6375123262405396, + "learning_rate": 0.00016306294710583568, + "loss": 0.2783, + "step": 12071 + }, + { + "epoch": 0.28298669104590446, + "grad_norm": 3.753845691680908, + "learning_rate": 0.0001630572301770639, + "loss": 0.2939, + "step": 12072 + }, + { + "epoch": 0.28301013262070945, + "grad_norm": 6.219955921173096, + "learning_rate": 0.0001630515129061419, + "loss": 0.4966, + "step": 12073 + }, + { + "epoch": 0.28303357419551445, + "grad_norm": 1.0846561193466187, + "learning_rate": 0.00016304579529310056, + "loss": 0.1513, + "step": 12074 + }, + { + "epoch": 0.28305701577031944, + "grad_norm": 4.280412197113037, + "learning_rate": 0.00016304007733797095, + "loss": 0.5166, + "step": 12075 + }, + { + "epoch": 0.28308045734512444, + "grad_norm": 1.0387296676635742, + "learning_rate": 0.00016303435904078416, + "loss": 0.0788, + "step": 12076 + }, + { + "epoch": 0.28310389891992943, + "grad_norm": 3.525465726852417, + "learning_rate": 0.00016302864040157114, + "loss": 0.2295, + "step": 12077 + }, + { + "epoch": 0.2831273404947344, + "grad_norm": 3.6143832206726074, + "learning_rate": 0.00016302292142036294, + "loss": 0.2876, + "step": 12078 + }, + { + "epoch": 0.2831507820695394, + "grad_norm": 5.0757341384887695, + "learning_rate": 0.00016301720209719063, + "loss": 0.7168, + "step": 12079 + }, + { + "epoch": 0.2831742236443444, + "grad_norm": 1.744248628616333, + "learning_rate": 0.00016301148243208515, + "loss": 0.2911, + "step": 12080 + }, + { + "epoch": 0.2831976652191494, + "grad_norm": 3.279407024383545, + "learning_rate": 0.00016300576242507764, + "loss": 0.3162, + "step": 12081 + }, + { + "epoch": 0.2832211067939544, + "grad_norm": 4.276885509490967, + "learning_rate": 0.0001630000420761991, + "loss": 0.2961, + "step": 12082 + }, + { + "epoch": 0.2832445483687594, + "grad_norm": 3.5292274951934814, + "learning_rate": 0.00016299432138548053, + "loss": 0.316, + "step": 12083 + }, + { + "epoch": 0.2832679899435644, + "grad_norm": 0.7440919280052185, + "learning_rate": 0.000162988600352953, + "loss": 0.0819, + "step": 12084 + }, + { + "epoch": 0.2832914315183694, + "grad_norm": 5.207919120788574, + "learning_rate": 0.00016298287897864756, + "loss": 0.633, + "step": 12085 + }, + { + "epoch": 0.2833148730931744, + "grad_norm": 4.322846412658691, + "learning_rate": 0.00016297715726259525, + "loss": 0.4838, + "step": 12086 + }, + { + "epoch": 0.2833383146679794, + "grad_norm": 3.8616201877593994, + "learning_rate": 0.00016297143520482713, + "loss": 0.4036, + "step": 12087 + }, + { + "epoch": 0.28336175624278437, + "grad_norm": 1.636712908744812, + "learning_rate": 0.00016296571280537424, + "loss": 0.1406, + "step": 12088 + }, + { + "epoch": 0.28338519781758936, + "grad_norm": 5.092071056365967, + "learning_rate": 0.00016295999006426759, + "loss": 0.3591, + "step": 12089 + }, + { + "epoch": 0.28340863939239436, + "grad_norm": 1.1888058185577393, + "learning_rate": 0.00016295426698153826, + "loss": 0.0936, + "step": 12090 + }, + { + "epoch": 0.28343208096719935, + "grad_norm": 0.8357040882110596, + "learning_rate": 0.00016294854355721733, + "loss": 0.0867, + "step": 12091 + }, + { + "epoch": 0.28345552254200435, + "grad_norm": 2.198524236679077, + "learning_rate": 0.00016294281979133582, + "loss": 0.2855, + "step": 12092 + }, + { + "epoch": 0.28347896411680934, + "grad_norm": 5.255663871765137, + "learning_rate": 0.0001629370956839248, + "loss": 0.47, + "step": 12093 + }, + { + "epoch": 0.2835024056916144, + "grad_norm": 3.4949305057525635, + "learning_rate": 0.00016293137123501533, + "loss": 1.9355, + "step": 12094 + }, + { + "epoch": 0.2835258472664194, + "grad_norm": 3.9283316135406494, + "learning_rate": 0.00016292564644463848, + "loss": 0.4146, + "step": 12095 + }, + { + "epoch": 0.2835492888412244, + "grad_norm": 3.078990936279297, + "learning_rate": 0.0001629199213128253, + "loss": 0.2498, + "step": 12096 + }, + { + "epoch": 0.2835727304160294, + "grad_norm": 3.8467836380004883, + "learning_rate": 0.00016291419583960685, + "loss": 0.3766, + "step": 12097 + }, + { + "epoch": 0.28359617199083437, + "grad_norm": 3.7643773555755615, + "learning_rate": 0.00016290847002501418, + "loss": 0.2508, + "step": 12098 + }, + { + "epoch": 0.28361961356563936, + "grad_norm": 6.321682453155518, + "learning_rate": 0.00016290274386907843, + "loss": 0.5405, + "step": 12099 + }, + { + "epoch": 0.28364305514044436, + "grad_norm": 2.7628438472747803, + "learning_rate": 0.0001628970173718306, + "loss": 1.2159, + "step": 12100 + }, + { + "epoch": 0.28366649671524935, + "grad_norm": 1.1447962522506714, + "learning_rate": 0.0001628912905333018, + "loss": 0.094, + "step": 12101 + }, + { + "epoch": 0.28368993829005434, + "grad_norm": 1.718528151512146, + "learning_rate": 0.00016288556335352307, + "loss": 0.2113, + "step": 12102 + }, + { + "epoch": 0.28371337986485934, + "grad_norm": 1.8764944076538086, + "learning_rate": 0.00016287983583252553, + "loss": 0.8329, + "step": 12103 + }, + { + "epoch": 0.28373682143966433, + "grad_norm": 4.039949893951416, + "learning_rate": 0.0001628741079703402, + "loss": 0.4577, + "step": 12104 + }, + { + "epoch": 0.2837602630144693, + "grad_norm": 7.679848670959473, + "learning_rate": 0.00016286837976699822, + "loss": 0.6931, + "step": 12105 + }, + { + "epoch": 0.2837837045892743, + "grad_norm": 1.3429385423660278, + "learning_rate": 0.00016286265122253063, + "loss": 0.1217, + "step": 12106 + }, + { + "epoch": 0.2838071461640793, + "grad_norm": 3.8108608722686768, + "learning_rate": 0.00016285692233696853, + "loss": 0.5234, + "step": 12107 + }, + { + "epoch": 0.2838305877388843, + "grad_norm": 0.6818922162055969, + "learning_rate": 0.000162851193110343, + "loss": 0.0904, + "step": 12108 + }, + { + "epoch": 0.2838540293136893, + "grad_norm": 3.147899866104126, + "learning_rate": 0.00016284546354268512, + "loss": 0.397, + "step": 12109 + }, + { + "epoch": 0.2838774708884943, + "grad_norm": 3.4991648197174072, + "learning_rate": 0.000162839733634026, + "loss": 0.3126, + "step": 12110 + }, + { + "epoch": 0.2839009124632993, + "grad_norm": 1.0203945636749268, + "learning_rate": 0.0001628340033843967, + "loss": 0.0409, + "step": 12111 + }, + { + "epoch": 0.2839243540381043, + "grad_norm": 2.58914852142334, + "learning_rate": 0.00016282827279382836, + "loss": 0.2318, + "step": 12112 + }, + { + "epoch": 0.2839477956129093, + "grad_norm": 2.142861843109131, + "learning_rate": 0.00016282254186235205, + "loss": 0.2242, + "step": 12113 + }, + { + "epoch": 0.2839712371877143, + "grad_norm": 3.007132053375244, + "learning_rate": 0.00016281681058999886, + "loss": 0.3983, + "step": 12114 + }, + { + "epoch": 0.28399467876251927, + "grad_norm": 3.315934181213379, + "learning_rate": 0.00016281107897679986, + "loss": 0.6151, + "step": 12115 + }, + { + "epoch": 0.28401812033732426, + "grad_norm": 7.467740058898926, + "learning_rate": 0.00016280534702278618, + "loss": 0.6552, + "step": 12116 + }, + { + "epoch": 0.28404156191212926, + "grad_norm": 1.8465213775634766, + "learning_rate": 0.00016279961472798893, + "loss": 0.2934, + "step": 12117 + }, + { + "epoch": 0.28406500348693425, + "grad_norm": 3.5204432010650635, + "learning_rate": 0.0001627938820924392, + "loss": 0.2376, + "step": 12118 + }, + { + "epoch": 0.28408844506173925, + "grad_norm": 4.088506698608398, + "learning_rate": 0.00016278814911616814, + "loss": 0.4225, + "step": 12119 + }, + { + "epoch": 0.28411188663654424, + "grad_norm": 3.7734620571136475, + "learning_rate": 0.0001627824157992068, + "loss": 0.5232, + "step": 12120 + }, + { + "epoch": 0.28413532821134924, + "grad_norm": 4.212916374206543, + "learning_rate": 0.00016277668214158627, + "loss": 0.4962, + "step": 12121 + }, + { + "epoch": 0.28415876978615423, + "grad_norm": 2.6228513717651367, + "learning_rate": 0.0001627709481433377, + "loss": 0.4265, + "step": 12122 + }, + { + "epoch": 0.2841822113609592, + "grad_norm": 3.9077515602111816, + "learning_rate": 0.00016276521380449222, + "loss": 0.3354, + "step": 12123 + }, + { + "epoch": 0.2842056529357642, + "grad_norm": 1.7914544343948364, + "learning_rate": 0.00016275947912508095, + "loss": 0.2246, + "step": 12124 + }, + { + "epoch": 0.2842290945105692, + "grad_norm": 4.3035664558410645, + "learning_rate": 0.00016275374410513495, + "loss": 0.4946, + "step": 12125 + }, + { + "epoch": 0.2842525360853742, + "grad_norm": 4.765383243560791, + "learning_rate": 0.00016274800874468536, + "loss": 0.4018, + "step": 12126 + }, + { + "epoch": 0.2842759776601792, + "grad_norm": 0.5809012055397034, + "learning_rate": 0.00016274227304376333, + "loss": 0.0583, + "step": 12127 + }, + { + "epoch": 0.2842994192349842, + "grad_norm": 3.183858871459961, + "learning_rate": 0.00016273653700239998, + "loss": 0.1621, + "step": 12128 + }, + { + "epoch": 0.2843228608097892, + "grad_norm": 0.600658118724823, + "learning_rate": 0.0001627308006206264, + "loss": 0.028, + "step": 12129 + }, + { + "epoch": 0.2843463023845942, + "grad_norm": 1.9862308502197266, + "learning_rate": 0.00016272506389847374, + "loss": 0.2653, + "step": 12130 + }, + { + "epoch": 0.2843697439593992, + "grad_norm": 4.912814617156982, + "learning_rate": 0.00016271932683597308, + "loss": 0.5769, + "step": 12131 + }, + { + "epoch": 0.2843931855342042, + "grad_norm": 2.8782620429992676, + "learning_rate": 0.00016271358943315562, + "loss": 0.2568, + "step": 12132 + }, + { + "epoch": 0.28441662710900917, + "grad_norm": 4.8440937995910645, + "learning_rate": 0.00016270785169005247, + "loss": 0.3338, + "step": 12133 + }, + { + "epoch": 0.28444006868381416, + "grad_norm": 7.174877643585205, + "learning_rate": 0.00016270211360669475, + "loss": 0.472, + "step": 12134 + }, + { + "epoch": 0.28446351025861916, + "grad_norm": 3.8355014324188232, + "learning_rate": 0.00016269637518311357, + "loss": 0.3317, + "step": 12135 + }, + { + "epoch": 0.28448695183342415, + "grad_norm": 5.750827312469482, + "learning_rate": 0.00016269063641934014, + "loss": 0.8388, + "step": 12136 + }, + { + "epoch": 0.28451039340822915, + "grad_norm": 3.581021785736084, + "learning_rate": 0.0001626848973154055, + "loss": 0.4459, + "step": 12137 + }, + { + "epoch": 0.28453383498303414, + "grad_norm": 3.180285930633545, + "learning_rate": 0.00016267915787134088, + "loss": 0.3602, + "step": 12138 + }, + { + "epoch": 0.28455727655783913, + "grad_norm": 3.047466278076172, + "learning_rate": 0.0001626734180871774, + "loss": 0.2642, + "step": 12139 + }, + { + "epoch": 0.28458071813264413, + "grad_norm": 1.283584475517273, + "learning_rate": 0.00016266767796294617, + "loss": 0.2093, + "step": 12140 + }, + { + "epoch": 0.2846041597074491, + "grad_norm": 4.464139938354492, + "learning_rate": 0.00016266193749867833, + "loss": 0.3361, + "step": 12141 + }, + { + "epoch": 0.2846276012822541, + "grad_norm": 2.712961196899414, + "learning_rate": 0.00016265619669440508, + "loss": 0.2113, + "step": 12142 + }, + { + "epoch": 0.2846510428570591, + "grad_norm": 3.241766929626465, + "learning_rate": 0.00016265045555015757, + "loss": 0.4571, + "step": 12143 + }, + { + "epoch": 0.2846744844318641, + "grad_norm": 2.2664144039154053, + "learning_rate": 0.0001626447140659669, + "loss": 0.229, + "step": 12144 + }, + { + "epoch": 0.28469792600666916, + "grad_norm": 10.193751335144043, + "learning_rate": 0.00016263897224186427, + "loss": 0.7393, + "step": 12145 + }, + { + "epoch": 0.28472136758147415, + "grad_norm": 3.384058952331543, + "learning_rate": 0.0001626332300778808, + "loss": 0.3285, + "step": 12146 + }, + { + "epoch": 0.28474480915627914, + "grad_norm": 2.1202585697174072, + "learning_rate": 0.00016262748757404764, + "loss": 1.4148, + "step": 12147 + }, + { + "epoch": 0.28476825073108414, + "grad_norm": 6.08840274810791, + "learning_rate": 0.00016262174473039603, + "loss": 0.2842, + "step": 12148 + }, + { + "epoch": 0.28479169230588913, + "grad_norm": 5.053438186645508, + "learning_rate": 0.000162616001546957, + "loss": 0.628, + "step": 12149 + }, + { + "epoch": 0.2848151338806941, + "grad_norm": 0.49196186661720276, + "learning_rate": 0.00016261025802376185, + "loss": 0.0238, + "step": 12150 + }, + { + "epoch": 0.2848385754554991, + "grad_norm": 5.723717212677002, + "learning_rate": 0.00016260451416084165, + "loss": 0.7721, + "step": 12151 + }, + { + "epoch": 0.2848620170303041, + "grad_norm": 2.824878692626953, + "learning_rate": 0.0001625987699582276, + "loss": 0.2763, + "step": 12152 + }, + { + "epoch": 0.2848854586051091, + "grad_norm": 1.8637243509292603, + "learning_rate": 0.00016259302541595086, + "loss": 0.1897, + "step": 12153 + }, + { + "epoch": 0.2849089001799141, + "grad_norm": 4.705180644989014, + "learning_rate": 0.0001625872805340426, + "loss": 0.4745, + "step": 12154 + }, + { + "epoch": 0.2849323417547191, + "grad_norm": 1.0636014938354492, + "learning_rate": 0.00016258153531253399, + "loss": 0.1582, + "step": 12155 + }, + { + "epoch": 0.2849557833295241, + "grad_norm": 4.413730621337891, + "learning_rate": 0.00016257578975145624, + "loss": 0.3878, + "step": 12156 + }, + { + "epoch": 0.2849792249043291, + "grad_norm": 3.7894816398620605, + "learning_rate": 0.00016257004385084048, + "loss": 0.4476, + "step": 12157 + }, + { + "epoch": 0.2850026664791341, + "grad_norm": 2.347118616104126, + "learning_rate": 0.0001625642976107179, + "loss": 0.1529, + "step": 12158 + }, + { + "epoch": 0.2850261080539391, + "grad_norm": 3.753016233444214, + "learning_rate": 0.00016255855103111964, + "loss": 0.5092, + "step": 12159 + }, + { + "epoch": 0.28504954962874407, + "grad_norm": 4.383737087249756, + "learning_rate": 0.000162552804112077, + "loss": 0.4807, + "step": 12160 + }, + { + "epoch": 0.28507299120354906, + "grad_norm": 2.567685604095459, + "learning_rate": 0.00016254705685362104, + "loss": 0.4327, + "step": 12161 + }, + { + "epoch": 0.28509643277835406, + "grad_norm": 6.068490982055664, + "learning_rate": 0.00016254130925578302, + "loss": 0.5495, + "step": 12162 + }, + { + "epoch": 0.28511987435315905, + "grad_norm": 1.1191232204437256, + "learning_rate": 0.00016253556131859406, + "loss": 0.1117, + "step": 12163 + }, + { + "epoch": 0.28514331592796405, + "grad_norm": 4.572361946105957, + "learning_rate": 0.00016252981304208541, + "loss": 0.3782, + "step": 12164 + }, + { + "epoch": 0.28516675750276904, + "grad_norm": 1.7503043413162231, + "learning_rate": 0.00016252406442628824, + "loss": 0.2528, + "step": 12165 + }, + { + "epoch": 0.28519019907757404, + "grad_norm": 4.316670894622803, + "learning_rate": 0.0001625183154712337, + "loss": 0.4511, + "step": 12166 + }, + { + "epoch": 0.28521364065237903, + "grad_norm": 3.3499844074249268, + "learning_rate": 0.0001625125661769531, + "loss": 0.4302, + "step": 12167 + }, + { + "epoch": 0.285237082227184, + "grad_norm": 0.9922764301300049, + "learning_rate": 0.0001625068165434775, + "loss": 0.0834, + "step": 12168 + }, + { + "epoch": 0.285260523801989, + "grad_norm": 2.8411662578582764, + "learning_rate": 0.00016250106657083815, + "loss": 0.367, + "step": 12169 + }, + { + "epoch": 0.285283965376794, + "grad_norm": 3.3222837448120117, + "learning_rate": 0.0001624953162590663, + "loss": 0.2651, + "step": 12170 + }, + { + "epoch": 0.285307406951599, + "grad_norm": 2.472827911376953, + "learning_rate": 0.00016248956560819305, + "loss": 0.2351, + "step": 12171 + }, + { + "epoch": 0.285330848526404, + "grad_norm": 3.872544527053833, + "learning_rate": 0.0001624838146182497, + "loss": 0.5398, + "step": 12172 + }, + { + "epoch": 0.285354290101209, + "grad_norm": 4.1192626953125, + "learning_rate": 0.0001624780632892674, + "loss": 0.5065, + "step": 12173 + }, + { + "epoch": 0.285377731676014, + "grad_norm": 1.04744291305542, + "learning_rate": 0.00016247231162127735, + "loss": 0.0796, + "step": 12174 + }, + { + "epoch": 0.285401173250819, + "grad_norm": 3.4285755157470703, + "learning_rate": 0.00016246655961431084, + "loss": 0.33, + "step": 12175 + }, + { + "epoch": 0.285424614825624, + "grad_norm": 2.9454538822174072, + "learning_rate": 0.00016246080726839898, + "loss": 1.815, + "step": 12176 + }, + { + "epoch": 0.285448056400429, + "grad_norm": 4.463327884674072, + "learning_rate": 0.00016245505458357303, + "loss": 0.2961, + "step": 12177 + }, + { + "epoch": 0.28547149797523397, + "grad_norm": 4.7885332107543945, + "learning_rate": 0.00016244930155986419, + "loss": 0.4107, + "step": 12178 + }, + { + "epoch": 0.28549493955003896, + "grad_norm": 1.1129512786865234, + "learning_rate": 0.00016244354819730373, + "loss": 0.0981, + "step": 12179 + }, + { + "epoch": 0.28551838112484396, + "grad_norm": 3.6327152252197266, + "learning_rate": 0.00016243779449592275, + "loss": 0.5315, + "step": 12180 + }, + { + "epoch": 0.28554182269964895, + "grad_norm": 4.365536689758301, + "learning_rate": 0.00016243204045575258, + "loss": 0.379, + "step": 12181 + }, + { + "epoch": 0.28556526427445394, + "grad_norm": 4.748720645904541, + "learning_rate": 0.0001624262860768244, + "loss": 0.5434, + "step": 12182 + }, + { + "epoch": 0.28558870584925894, + "grad_norm": 4.259218692779541, + "learning_rate": 0.00016242053135916943, + "loss": 0.4253, + "step": 12183 + }, + { + "epoch": 0.28561214742406393, + "grad_norm": 1.4003690481185913, + "learning_rate": 0.00016241477630281892, + "loss": 0.1654, + "step": 12184 + }, + { + "epoch": 0.28563558899886893, + "grad_norm": 5.11175537109375, + "learning_rate": 0.00016240902090780405, + "loss": 0.7362, + "step": 12185 + }, + { + "epoch": 0.2856590305736739, + "grad_norm": 1.0259629487991333, + "learning_rate": 0.00016240326517415608, + "loss": 0.0505, + "step": 12186 + }, + { + "epoch": 0.2856824721484789, + "grad_norm": 1.2626930475234985, + "learning_rate": 0.00016239750910190627, + "loss": 0.197, + "step": 12187 + }, + { + "epoch": 0.2857059137232839, + "grad_norm": 3.5085508823394775, + "learning_rate": 0.00016239175269108582, + "loss": 0.3198, + "step": 12188 + }, + { + "epoch": 0.2857293552980889, + "grad_norm": 1.8730580806732178, + "learning_rate": 0.00016238599594172595, + "loss": 1.0729, + "step": 12189 + }, + { + "epoch": 0.2857527968728939, + "grad_norm": 5.574343204498291, + "learning_rate": 0.0001623802388538579, + "loss": 0.7564, + "step": 12190 + }, + { + "epoch": 0.2857762384476989, + "grad_norm": 2.5882203578948975, + "learning_rate": 0.00016237448142751291, + "loss": 1.5666, + "step": 12191 + }, + { + "epoch": 0.2857996800225039, + "grad_norm": 5.051502704620361, + "learning_rate": 0.00016236872366272228, + "loss": 0.5834, + "step": 12192 + }, + { + "epoch": 0.2858231215973089, + "grad_norm": 4.392299652099609, + "learning_rate": 0.00016236296555951718, + "loss": 0.5673, + "step": 12193 + }, + { + "epoch": 0.2858465631721139, + "grad_norm": 3.754777193069458, + "learning_rate": 0.00016235720711792886, + "loss": 0.2958, + "step": 12194 + }, + { + "epoch": 0.28587000474691887, + "grad_norm": 5.496826171875, + "learning_rate": 0.00016235144833798857, + "loss": 0.5953, + "step": 12195 + }, + { + "epoch": 0.2858934463217239, + "grad_norm": 2.3358681201934814, + "learning_rate": 0.0001623456892197276, + "loss": 1.2399, + "step": 12196 + }, + { + "epoch": 0.2859168878965289, + "grad_norm": 1.868019938468933, + "learning_rate": 0.00016233992976317715, + "loss": 0.193, + "step": 12197 + }, + { + "epoch": 0.2859403294713339, + "grad_norm": 4.156998157501221, + "learning_rate": 0.0001623341699683685, + "loss": 0.5091, + "step": 12198 + }, + { + "epoch": 0.2859637710461389, + "grad_norm": 5.09243106842041, + "learning_rate": 0.00016232840983533287, + "loss": 0.7773, + "step": 12199 + }, + { + "epoch": 0.2859872126209439, + "grad_norm": 7.41767692565918, + "learning_rate": 0.00016232264936410156, + "loss": 0.8125, + "step": 12200 + }, + { + "epoch": 0.2860106541957489, + "grad_norm": 0.7890149354934692, + "learning_rate": 0.0001623168885547058, + "loss": 0.0945, + "step": 12201 + }, + { + "epoch": 0.2860340957705539, + "grad_norm": 3.44203519821167, + "learning_rate": 0.00016231112740717682, + "loss": 0.3291, + "step": 12202 + }, + { + "epoch": 0.2860575373453589, + "grad_norm": 3.1155612468719482, + "learning_rate": 0.00016230536592154592, + "loss": 0.2515, + "step": 12203 + }, + { + "epoch": 0.2860809789201639, + "grad_norm": 4.212446689605713, + "learning_rate": 0.0001622996040978444, + "loss": 0.3247, + "step": 12204 + }, + { + "epoch": 0.28610442049496887, + "grad_norm": 0.8638741970062256, + "learning_rate": 0.00016229384193610345, + "loss": 0.0776, + "step": 12205 + }, + { + "epoch": 0.28612786206977386, + "grad_norm": 7.493319988250732, + "learning_rate": 0.00016228807943635435, + "loss": 0.8933, + "step": 12206 + }, + { + "epoch": 0.28615130364457886, + "grad_norm": 2.995007276535034, + "learning_rate": 0.0001622823165986284, + "loss": 0.2422, + "step": 12207 + }, + { + "epoch": 0.28617474521938385, + "grad_norm": 6.674158096313477, + "learning_rate": 0.0001622765534229568, + "loss": 0.437, + "step": 12208 + }, + { + "epoch": 0.28619818679418885, + "grad_norm": 4.348272323608398, + "learning_rate": 0.0001622707899093709, + "loss": 0.3347, + "step": 12209 + }, + { + "epoch": 0.28622162836899384, + "grad_norm": 1.5964986085891724, + "learning_rate": 0.00016226502605790194, + "loss": 0.1195, + "step": 12210 + }, + { + "epoch": 0.28624506994379884, + "grad_norm": 1.7361260652542114, + "learning_rate": 0.0001622592618685812, + "loss": 0.1832, + "step": 12211 + }, + { + "epoch": 0.28626851151860383, + "grad_norm": 1.2812163829803467, + "learning_rate": 0.00016225349734143998, + "loss": 0.145, + "step": 12212 + }, + { + "epoch": 0.2862919530934088, + "grad_norm": 7.577844619750977, + "learning_rate": 0.0001622477324765095, + "loss": 0.3142, + "step": 12213 + }, + { + "epoch": 0.2863153946682138, + "grad_norm": 2.7103631496429443, + "learning_rate": 0.00016224196727382108, + "loss": 1.1286, + "step": 12214 + }, + { + "epoch": 0.2863388362430188, + "grad_norm": 3.5754354000091553, + "learning_rate": 0.00016223620173340598, + "loss": 0.5363, + "step": 12215 + }, + { + "epoch": 0.2863622778178238, + "grad_norm": 2.959949254989624, + "learning_rate": 0.00016223043585529556, + "loss": 0.3694, + "step": 12216 + }, + { + "epoch": 0.2863857193926288, + "grad_norm": 4.089199542999268, + "learning_rate": 0.00016222466963952098, + "loss": 0.3849, + "step": 12217 + }, + { + "epoch": 0.2864091609674338, + "grad_norm": 4.62436580657959, + "learning_rate": 0.00016221890308611362, + "loss": 0.3971, + "step": 12218 + }, + { + "epoch": 0.2864326025422388, + "grad_norm": 5.169435024261475, + "learning_rate": 0.00016221313619510475, + "loss": 0.6657, + "step": 12219 + }, + { + "epoch": 0.2864560441170438, + "grad_norm": 2.64106822013855, + "learning_rate": 0.00016220736896652564, + "loss": 0.2052, + "step": 12220 + }, + { + "epoch": 0.2864794856918488, + "grad_norm": 5.234435558319092, + "learning_rate": 0.0001622016014004076, + "loss": 0.5748, + "step": 12221 + }, + { + "epoch": 0.2865029272666538, + "grad_norm": 2.4839353561401367, + "learning_rate": 0.00016219583349678192, + "loss": 0.3017, + "step": 12222 + }, + { + "epoch": 0.28652636884145877, + "grad_norm": 0.6308846473693848, + "learning_rate": 0.0001621900652556799, + "loss": 0.0556, + "step": 12223 + }, + { + "epoch": 0.28654981041626376, + "grad_norm": 4.220087051391602, + "learning_rate": 0.00016218429667713284, + "loss": 0.5097, + "step": 12224 + }, + { + "epoch": 0.28657325199106876, + "grad_norm": 3.9073245525360107, + "learning_rate": 0.00016217852776117205, + "loss": 2.0821, + "step": 12225 + }, + { + "epoch": 0.28659669356587375, + "grad_norm": 0.7033354043960571, + "learning_rate": 0.00016217275850782882, + "loss": 0.0544, + "step": 12226 + }, + { + "epoch": 0.28662013514067874, + "grad_norm": 3.3298776149749756, + "learning_rate": 0.0001621669889171344, + "loss": 0.5375, + "step": 12227 + }, + { + "epoch": 0.28664357671548374, + "grad_norm": 3.9473283290863037, + "learning_rate": 0.0001621612189891202, + "loss": 0.5155, + "step": 12228 + }, + { + "epoch": 0.28666701829028873, + "grad_norm": 3.879424810409546, + "learning_rate": 0.00016215544872381747, + "loss": 0.6668, + "step": 12229 + }, + { + "epoch": 0.2866904598650937, + "grad_norm": 2.3465142250061035, + "learning_rate": 0.00016214967812125752, + "loss": 2.427, + "step": 12230 + }, + { + "epoch": 0.2867139014398987, + "grad_norm": 5.804582118988037, + "learning_rate": 0.0001621439071814717, + "loss": 0.6945, + "step": 12231 + }, + { + "epoch": 0.2867373430147037, + "grad_norm": 4.933794021606445, + "learning_rate": 0.00016213813590449126, + "loss": 0.5854, + "step": 12232 + }, + { + "epoch": 0.2867607845895087, + "grad_norm": 1.3725095987319946, + "learning_rate": 0.00016213236429034753, + "loss": 0.2117, + "step": 12233 + }, + { + "epoch": 0.2867842261643137, + "grad_norm": 3.415828227996826, + "learning_rate": 0.00016212659233907187, + "loss": 0.4354, + "step": 12234 + }, + { + "epoch": 0.2868076677391187, + "grad_norm": 4.481562614440918, + "learning_rate": 0.0001621208200506956, + "loss": 0.4193, + "step": 12235 + }, + { + "epoch": 0.2868311093139237, + "grad_norm": 4.01175594329834, + "learning_rate": 0.00016211504742524995, + "loss": 0.4005, + "step": 12236 + }, + { + "epoch": 0.2868545508887287, + "grad_norm": 1.9242349863052368, + "learning_rate": 0.00016210927446276636, + "loss": 0.3321, + "step": 12237 + }, + { + "epoch": 0.2868779924635337, + "grad_norm": 2.5107131004333496, + "learning_rate": 0.00016210350116327606, + "loss": 0.1707, + "step": 12238 + }, + { + "epoch": 0.2869014340383387, + "grad_norm": 5.455077171325684, + "learning_rate": 0.00016209772752681046, + "loss": 0.7086, + "step": 12239 + }, + { + "epoch": 0.28692487561314367, + "grad_norm": 4.588451385498047, + "learning_rate": 0.00016209195355340082, + "loss": 0.4724, + "step": 12240 + }, + { + "epoch": 0.28694831718794866, + "grad_norm": 1.2565630674362183, + "learning_rate": 0.00016208617924307853, + "loss": 0.1537, + "step": 12241 + }, + { + "epoch": 0.28697175876275366, + "grad_norm": 1.1736940145492554, + "learning_rate": 0.0001620804045958748, + "loss": 0.1112, + "step": 12242 + }, + { + "epoch": 0.28699520033755865, + "grad_norm": 2.6486220359802246, + "learning_rate": 0.00016207462961182113, + "loss": 0.3639, + "step": 12243 + }, + { + "epoch": 0.28701864191236365, + "grad_norm": 3.0965890884399414, + "learning_rate": 0.00016206885429094874, + "loss": 0.373, + "step": 12244 + }, + { + "epoch": 0.28704208348716864, + "grad_norm": 3.013275623321533, + "learning_rate": 0.000162063078633289, + "loss": 0.3854, + "step": 12245 + }, + { + "epoch": 0.28706552506197364, + "grad_norm": 2.6056978702545166, + "learning_rate": 0.00016205730263887326, + "loss": 0.3428, + "step": 12246 + }, + { + "epoch": 0.2870889666367787, + "grad_norm": 2.432659149169922, + "learning_rate": 0.00016205152630773285, + "loss": 0.3947, + "step": 12247 + }, + { + "epoch": 0.2871124082115837, + "grad_norm": 4.094363212585449, + "learning_rate": 0.00016204574963989913, + "loss": 0.3348, + "step": 12248 + }, + { + "epoch": 0.2871358497863887, + "grad_norm": 1.9898573160171509, + "learning_rate": 0.00016203997263540342, + "loss": 0.1525, + "step": 12249 + }, + { + "epoch": 0.28715929136119367, + "grad_norm": 3.4583332538604736, + "learning_rate": 0.00016203419529427703, + "loss": 0.4693, + "step": 12250 + }, + { + "epoch": 0.28718273293599866, + "grad_norm": 4.441073417663574, + "learning_rate": 0.0001620284176165514, + "loss": 0.6193, + "step": 12251 + }, + { + "epoch": 0.28720617451080366, + "grad_norm": 3.497612237930298, + "learning_rate": 0.0001620226396022578, + "loss": 0.5902, + "step": 12252 + }, + { + "epoch": 0.28722961608560865, + "grad_norm": 2.8405447006225586, + "learning_rate": 0.00016201686125142767, + "loss": 0.395, + "step": 12253 + }, + { + "epoch": 0.28725305766041365, + "grad_norm": 3.099391222000122, + "learning_rate": 0.00016201108256409227, + "loss": 0.3252, + "step": 12254 + }, + { + "epoch": 0.28727649923521864, + "grad_norm": 4.842339038848877, + "learning_rate": 0.00016200530354028297, + "loss": 0.6431, + "step": 12255 + }, + { + "epoch": 0.28729994081002364, + "grad_norm": 0.38482600450515747, + "learning_rate": 0.00016199952418003116, + "loss": 0.021, + "step": 12256 + }, + { + "epoch": 0.28732338238482863, + "grad_norm": 2.5909454822540283, + "learning_rate": 0.00016199374448336822, + "loss": 1.1192, + "step": 12257 + }, + { + "epoch": 0.2873468239596336, + "grad_norm": 3.6061391830444336, + "learning_rate": 0.00016198796445032545, + "loss": 0.5977, + "step": 12258 + }, + { + "epoch": 0.2873702655344386, + "grad_norm": 3.4309628009796143, + "learning_rate": 0.00016198218408093427, + "loss": 1.3014, + "step": 12259 + }, + { + "epoch": 0.2873937071092436, + "grad_norm": 4.739203929901123, + "learning_rate": 0.000161976403375226, + "loss": 0.7611, + "step": 12260 + }, + { + "epoch": 0.2874171486840486, + "grad_norm": 6.65562105178833, + "learning_rate": 0.000161970622333232, + "loss": 0.4047, + "step": 12261 + }, + { + "epoch": 0.2874405902588536, + "grad_norm": 1.314356803894043, + "learning_rate": 0.00016196484095498367, + "loss": 0.0647, + "step": 12262 + }, + { + "epoch": 0.2874640318336586, + "grad_norm": 3.8544061183929443, + "learning_rate": 0.00016195905924051238, + "loss": 0.4171, + "step": 12263 + }, + { + "epoch": 0.2874874734084636, + "grad_norm": 1.498811960220337, + "learning_rate": 0.0001619532771898495, + "loss": 0.1532, + "step": 12264 + }, + { + "epoch": 0.2875109149832686, + "grad_norm": 2.6354901790618896, + "learning_rate": 0.0001619474948030264, + "loss": 0.1946, + "step": 12265 + }, + { + "epoch": 0.2875343565580736, + "grad_norm": 1.7941932678222656, + "learning_rate": 0.00016194171208007446, + "loss": 0.1651, + "step": 12266 + }, + { + "epoch": 0.2875577981328786, + "grad_norm": 3.845456123352051, + "learning_rate": 0.000161935929021025, + "loss": 0.7444, + "step": 12267 + }, + { + "epoch": 0.28758123970768357, + "grad_norm": 4.110740661621094, + "learning_rate": 0.00016193014562590946, + "loss": 0.5732, + "step": 12268 + }, + { + "epoch": 0.28760468128248856, + "grad_norm": 2.5978729724884033, + "learning_rate": 0.00016192436189475923, + "loss": 0.2375, + "step": 12269 + }, + { + "epoch": 0.28762812285729356, + "grad_norm": 1.7217308282852173, + "learning_rate": 0.00016191857782760569, + "loss": 0.1472, + "step": 12270 + }, + { + "epoch": 0.28765156443209855, + "grad_norm": 3.522146224975586, + "learning_rate": 0.00016191279342448017, + "loss": 0.2951, + "step": 12271 + }, + { + "epoch": 0.28767500600690354, + "grad_norm": 0.707249641418457, + "learning_rate": 0.0001619070086854141, + "loss": 0.0755, + "step": 12272 + }, + { + "epoch": 0.28769844758170854, + "grad_norm": 3.9796485900878906, + "learning_rate": 0.00016190122361043892, + "loss": 0.2862, + "step": 12273 + }, + { + "epoch": 0.28772188915651353, + "grad_norm": 4.341625690460205, + "learning_rate": 0.0001618954381995859, + "loss": 0.4001, + "step": 12274 + }, + { + "epoch": 0.2877453307313185, + "grad_norm": 9.006056785583496, + "learning_rate": 0.00016188965245288647, + "loss": 0.9625, + "step": 12275 + }, + { + "epoch": 0.2877687723061235, + "grad_norm": 4.122885227203369, + "learning_rate": 0.00016188386637037207, + "loss": 0.4129, + "step": 12276 + }, + { + "epoch": 0.2877922138809285, + "grad_norm": 3.2739369869232178, + "learning_rate": 0.00016187807995207408, + "loss": 0.3239, + "step": 12277 + }, + { + "epoch": 0.2878156554557335, + "grad_norm": 0.9007963538169861, + "learning_rate": 0.00016187229319802388, + "loss": 0.089, + "step": 12278 + }, + { + "epoch": 0.2878390970305385, + "grad_norm": 4.003042697906494, + "learning_rate": 0.0001618665061082529, + "loss": 0.4824, + "step": 12279 + }, + { + "epoch": 0.2878625386053435, + "grad_norm": 3.702439069747925, + "learning_rate": 0.0001618607186827925, + "loss": 0.4281, + "step": 12280 + }, + { + "epoch": 0.2878859801801485, + "grad_norm": 1.4831360578536987, + "learning_rate": 0.00016185493092167407, + "loss": 0.2045, + "step": 12281 + }, + { + "epoch": 0.2879094217549535, + "grad_norm": 5.309216499328613, + "learning_rate": 0.0001618491428249291, + "loss": 0.4363, + "step": 12282 + }, + { + "epoch": 0.2879328633297585, + "grad_norm": 3.1521623134613037, + "learning_rate": 0.0001618433543925889, + "loss": 0.3263, + "step": 12283 + }, + { + "epoch": 0.2879563049045635, + "grad_norm": 3.4571428298950195, + "learning_rate": 0.00016183756562468494, + "loss": 0.2772, + "step": 12284 + }, + { + "epoch": 0.28797974647936847, + "grad_norm": 2.568530321121216, + "learning_rate": 0.00016183177652124862, + "loss": 0.4423, + "step": 12285 + }, + { + "epoch": 0.28800318805417346, + "grad_norm": 4.523317813873291, + "learning_rate": 0.0001618259870823113, + "loss": 0.6779, + "step": 12286 + }, + { + "epoch": 0.28802662962897846, + "grad_norm": 2.000028371810913, + "learning_rate": 0.00016182019730790446, + "loss": 0.3964, + "step": 12287 + }, + { + "epoch": 0.28805007120378345, + "grad_norm": 4.894166946411133, + "learning_rate": 0.00016181440719805948, + "loss": 0.633, + "step": 12288 + }, + { + "epoch": 0.28807351277858845, + "grad_norm": 2.4052786827087402, + "learning_rate": 0.00016180861675280786, + "loss": 0.329, + "step": 12289 + }, + { + "epoch": 0.28809695435339344, + "grad_norm": 0.8569680452346802, + "learning_rate": 0.00016180282597218088, + "loss": 0.1106, + "step": 12290 + }, + { + "epoch": 0.28812039592819844, + "grad_norm": 3.445383310317993, + "learning_rate": 0.00016179703485621002, + "loss": 0.5322, + "step": 12291 + }, + { + "epoch": 0.28814383750300343, + "grad_norm": 1.8275201320648193, + "learning_rate": 0.00016179124340492674, + "loss": 0.1662, + "step": 12292 + }, + { + "epoch": 0.2881672790778084, + "grad_norm": 2.574200391769409, + "learning_rate": 0.00016178545161836243, + "loss": 0.4233, + "step": 12293 + }, + { + "epoch": 0.2881907206526134, + "grad_norm": 1.071976900100708, + "learning_rate": 0.00016177965949654851, + "loss": 0.1617, + "step": 12294 + }, + { + "epoch": 0.2882141622274184, + "grad_norm": 2.6961894035339355, + "learning_rate": 0.00016177386703951645, + "loss": 2.3408, + "step": 12295 + }, + { + "epoch": 0.2882376038022234, + "grad_norm": 3.443228006362915, + "learning_rate": 0.00016176807424729764, + "loss": 0.3429, + "step": 12296 + }, + { + "epoch": 0.2882610453770284, + "grad_norm": 1.1143162250518799, + "learning_rate": 0.0001617622811199235, + "loss": 0.1069, + "step": 12297 + }, + { + "epoch": 0.28828448695183345, + "grad_norm": 4.010329723358154, + "learning_rate": 0.0001617564876574255, + "loss": 0.4212, + "step": 12298 + }, + { + "epoch": 0.28830792852663845, + "grad_norm": 5.454514980316162, + "learning_rate": 0.00016175069385983507, + "loss": 0.608, + "step": 12299 + }, + { + "epoch": 0.28833137010144344, + "grad_norm": 2.4695138931274414, + "learning_rate": 0.00016174489972718364, + "loss": 0.2525, + "step": 12300 + }, + { + "epoch": 0.28835481167624843, + "grad_norm": 0.9298647046089172, + "learning_rate": 0.00016173910525950263, + "loss": 0.0606, + "step": 12301 + }, + { + "epoch": 0.28837825325105343, + "grad_norm": 3.229358196258545, + "learning_rate": 0.0001617333104568235, + "loss": 0.2765, + "step": 12302 + }, + { + "epoch": 0.2884016948258584, + "grad_norm": 1.9842257499694824, + "learning_rate": 0.0001617275153191777, + "loss": 0.1824, + "step": 12303 + }, + { + "epoch": 0.2884251364006634, + "grad_norm": 3.545198917388916, + "learning_rate": 0.00016172171984659664, + "loss": 0.3899, + "step": 12304 + }, + { + "epoch": 0.2884485779754684, + "grad_norm": 0.4824358820915222, + "learning_rate": 0.00016171592403911182, + "loss": 0.0349, + "step": 12305 + }, + { + "epoch": 0.2884720195502734, + "grad_norm": 4.3613810539245605, + "learning_rate": 0.00016171012789675468, + "loss": 0.3886, + "step": 12306 + }, + { + "epoch": 0.2884954611250784, + "grad_norm": 3.2373080253601074, + "learning_rate": 0.00016170433141955663, + "loss": 0.3397, + "step": 12307 + }, + { + "epoch": 0.2885189026998834, + "grad_norm": 2.40633487701416, + "learning_rate": 0.00016169853460754914, + "loss": 0.2516, + "step": 12308 + }, + { + "epoch": 0.2885423442746884, + "grad_norm": 4.974930763244629, + "learning_rate": 0.00016169273746076366, + "loss": 0.4407, + "step": 12309 + }, + { + "epoch": 0.2885657858494934, + "grad_norm": 3.278369426727295, + "learning_rate": 0.00016168693997923165, + "loss": 0.3249, + "step": 12310 + }, + { + "epoch": 0.2885892274242984, + "grad_norm": 2.970327377319336, + "learning_rate": 0.0001616811421629846, + "loss": 1.3278, + "step": 12311 + }, + { + "epoch": 0.2886126689991034, + "grad_norm": 3.00182843208313, + "learning_rate": 0.0001616753440120539, + "loss": 0.3103, + "step": 12312 + }, + { + "epoch": 0.28863611057390837, + "grad_norm": 3.8496170043945312, + "learning_rate": 0.00016166954552647108, + "loss": 0.4377, + "step": 12313 + }, + { + "epoch": 0.28865955214871336, + "grad_norm": 0.8147679567337036, + "learning_rate": 0.00016166374670626755, + "loss": 0.0792, + "step": 12314 + }, + { + "epoch": 0.28868299372351836, + "grad_norm": 4.0551042556762695, + "learning_rate": 0.0001616579475514748, + "loss": 0.3702, + "step": 12315 + }, + { + "epoch": 0.28870643529832335, + "grad_norm": 2.5529251098632812, + "learning_rate": 0.00016165214806212433, + "loss": 0.1807, + "step": 12316 + }, + { + "epoch": 0.28872987687312834, + "grad_norm": 6.381176471710205, + "learning_rate": 0.0001616463482382475, + "loss": 0.6869, + "step": 12317 + }, + { + "epoch": 0.28875331844793334, + "grad_norm": 2.2540340423583984, + "learning_rate": 0.0001616405480798759, + "loss": 0.1907, + "step": 12318 + }, + { + "epoch": 0.28877676002273833, + "grad_norm": 0.8242038488388062, + "learning_rate": 0.00016163474758704096, + "loss": 0.0948, + "step": 12319 + }, + { + "epoch": 0.2888002015975433, + "grad_norm": 5.121989727020264, + "learning_rate": 0.00016162894675977413, + "loss": 0.2389, + "step": 12320 + }, + { + "epoch": 0.2888236431723483, + "grad_norm": 1.4542900323867798, + "learning_rate": 0.0001616231455981069, + "loss": 0.2347, + "step": 12321 + }, + { + "epoch": 0.2888470847471533, + "grad_norm": 4.4299635887146, + "learning_rate": 0.00016161734410207075, + "loss": 0.4878, + "step": 12322 + }, + { + "epoch": 0.2888705263219583, + "grad_norm": 4.188438415527344, + "learning_rate": 0.00016161154227169715, + "loss": 0.5961, + "step": 12323 + }, + { + "epoch": 0.2888939678967633, + "grad_norm": 5.437929630279541, + "learning_rate": 0.00016160574010701758, + "loss": 0.7455, + "step": 12324 + }, + { + "epoch": 0.2889174094715683, + "grad_norm": 4.333773136138916, + "learning_rate": 0.00016159993760806357, + "loss": 0.637, + "step": 12325 + }, + { + "epoch": 0.2889408510463733, + "grad_norm": 0.6766412854194641, + "learning_rate": 0.00016159413477486654, + "loss": 0.0251, + "step": 12326 + }, + { + "epoch": 0.2889642926211783, + "grad_norm": 3.414496898651123, + "learning_rate": 0.00016158833160745804, + "loss": 0.3659, + "step": 12327 + }, + { + "epoch": 0.2889877341959833, + "grad_norm": 2.8129754066467285, + "learning_rate": 0.00016158252810586946, + "loss": 0.2633, + "step": 12328 + }, + { + "epoch": 0.2890111757707883, + "grad_norm": 1.7337424755096436, + "learning_rate": 0.0001615767242701324, + "loss": 0.2752, + "step": 12329 + }, + { + "epoch": 0.28903461734559327, + "grad_norm": 1.6183220148086548, + "learning_rate": 0.0001615709201002783, + "loss": 0.1541, + "step": 12330 + }, + { + "epoch": 0.28905805892039826, + "grad_norm": 5.771915435791016, + "learning_rate": 0.00016156511559633862, + "loss": 0.1855, + "step": 12331 + }, + { + "epoch": 0.28908150049520326, + "grad_norm": 4.6503214836120605, + "learning_rate": 0.00016155931075834492, + "loss": 0.6229, + "step": 12332 + }, + { + "epoch": 0.28910494207000825, + "grad_norm": 3.5812551975250244, + "learning_rate": 0.00016155350558632868, + "loss": 0.3393, + "step": 12333 + }, + { + "epoch": 0.28912838364481325, + "grad_norm": 0.25818932056427, + "learning_rate": 0.00016154770008032133, + "loss": 0.029, + "step": 12334 + }, + { + "epoch": 0.28915182521961824, + "grad_norm": 3.8678321838378906, + "learning_rate": 0.0001615418942403545, + "loss": 0.3689, + "step": 12335 + }, + { + "epoch": 0.28917526679442324, + "grad_norm": 4.435031414031982, + "learning_rate": 0.00016153608806645958, + "loss": 0.655, + "step": 12336 + }, + { + "epoch": 0.28919870836922823, + "grad_norm": 7.21351957321167, + "learning_rate": 0.00016153028155866812, + "loss": 0.9243, + "step": 12337 + }, + { + "epoch": 0.2892221499440332, + "grad_norm": 2.631561040878296, + "learning_rate": 0.00016152447471701163, + "loss": 0.2834, + "step": 12338 + }, + { + "epoch": 0.2892455915188382, + "grad_norm": 5.12104606628418, + "learning_rate": 0.00016151866754152163, + "loss": 0.4952, + "step": 12339 + }, + { + "epoch": 0.2892690330936432, + "grad_norm": 1.6224082708358765, + "learning_rate": 0.0001615128600322296, + "loss": 0.2023, + "step": 12340 + }, + { + "epoch": 0.2892924746684482, + "grad_norm": 5.996196269989014, + "learning_rate": 0.00016150705218916702, + "loss": 0.5398, + "step": 12341 + }, + { + "epoch": 0.2893159162432532, + "grad_norm": 6.0327043533325195, + "learning_rate": 0.00016150124401236547, + "loss": 0.3592, + "step": 12342 + }, + { + "epoch": 0.2893393578180582, + "grad_norm": 3.014017105102539, + "learning_rate": 0.00016149543550185644, + "loss": 1.7031, + "step": 12343 + }, + { + "epoch": 0.2893627993928632, + "grad_norm": 6.5519609451293945, + "learning_rate": 0.00016148962665767146, + "loss": 0.5071, + "step": 12344 + }, + { + "epoch": 0.2893862409676682, + "grad_norm": 4.250945568084717, + "learning_rate": 0.00016148381747984204, + "loss": 0.4505, + "step": 12345 + }, + { + "epoch": 0.2894096825424732, + "grad_norm": 3.6490590572357178, + "learning_rate": 0.0001614780079683997, + "loss": 0.4695, + "step": 12346 + }, + { + "epoch": 0.2894331241172782, + "grad_norm": 3.783397912979126, + "learning_rate": 0.00016147219812337593, + "loss": 0.386, + "step": 12347 + }, + { + "epoch": 0.28945656569208317, + "grad_norm": 5.559905052185059, + "learning_rate": 0.00016146638794480232, + "loss": 0.7126, + "step": 12348 + }, + { + "epoch": 0.28948000726688816, + "grad_norm": 3.2048912048339844, + "learning_rate": 0.00016146057743271036, + "loss": 0.3607, + "step": 12349 + }, + { + "epoch": 0.2895034488416932, + "grad_norm": 1.3059667348861694, + "learning_rate": 0.00016145476658713156, + "loss": 0.2092, + "step": 12350 + }, + { + "epoch": 0.2895268904164982, + "grad_norm": 3.5692906379699707, + "learning_rate": 0.00016144895540809746, + "loss": 0.4067, + "step": 12351 + }, + { + "epoch": 0.2895503319913032, + "grad_norm": 2.6801867485046387, + "learning_rate": 0.00016144314389563962, + "loss": 0.3379, + "step": 12352 + }, + { + "epoch": 0.2895737735661082, + "grad_norm": 3.4703121185302734, + "learning_rate": 0.00016143733204978955, + "loss": 0.5467, + "step": 12353 + }, + { + "epoch": 0.2895972151409132, + "grad_norm": 1.2892144918441772, + "learning_rate": 0.00016143151987057876, + "loss": 0.1352, + "step": 12354 + }, + { + "epoch": 0.2896206567157182, + "grad_norm": 2.3453140258789062, + "learning_rate": 0.00016142570735803885, + "loss": 1.5974, + "step": 12355 + }, + { + "epoch": 0.2896440982905232, + "grad_norm": 5.132394790649414, + "learning_rate": 0.0001614198945122013, + "loss": 1.4698, + "step": 12356 + }, + { + "epoch": 0.28966753986532817, + "grad_norm": 1.4304794073104858, + "learning_rate": 0.0001614140813330977, + "loss": 0.145, + "step": 12357 + }, + { + "epoch": 0.28969098144013317, + "grad_norm": 5.9646100997924805, + "learning_rate": 0.00016140826782075959, + "loss": 0.798, + "step": 12358 + }, + { + "epoch": 0.28971442301493816, + "grad_norm": 5.259458065032959, + "learning_rate": 0.00016140245397521842, + "loss": 0.4415, + "step": 12359 + }, + { + "epoch": 0.28973786458974315, + "grad_norm": 1.5509523153305054, + "learning_rate": 0.00016139663979650585, + "loss": 0.0874, + "step": 12360 + }, + { + "epoch": 0.28976130616454815, + "grad_norm": 4.786919116973877, + "learning_rate": 0.00016139082528465339, + "loss": 0.5474, + "step": 12361 + }, + { + "epoch": 0.28978474773935314, + "grad_norm": 2.206470251083374, + "learning_rate": 0.0001613850104396926, + "loss": 0.2033, + "step": 12362 + }, + { + "epoch": 0.28980818931415814, + "grad_norm": 3.785780429840088, + "learning_rate": 0.00016137919526165498, + "loss": 0.4956, + "step": 12363 + }, + { + "epoch": 0.28983163088896313, + "grad_norm": 1.901374340057373, + "learning_rate": 0.00016137337975057214, + "loss": 0.1482, + "step": 12364 + }, + { + "epoch": 0.2898550724637681, + "grad_norm": 4.244401931762695, + "learning_rate": 0.00016136756390647562, + "loss": 0.3481, + "step": 12365 + }, + { + "epoch": 0.2898785140385731, + "grad_norm": 5.429022312164307, + "learning_rate": 0.00016136174772939693, + "loss": 0.8684, + "step": 12366 + }, + { + "epoch": 0.2899019556133781, + "grad_norm": 1.2230620384216309, + "learning_rate": 0.00016135593121936775, + "loss": 0.1413, + "step": 12367 + }, + { + "epoch": 0.2899253971881831, + "grad_norm": 4.732159614562988, + "learning_rate": 0.00016135011437641951, + "loss": 0.5154, + "step": 12368 + }, + { + "epoch": 0.2899488387629881, + "grad_norm": 3.3929874897003174, + "learning_rate": 0.00016134429720058384, + "loss": 0.4226, + "step": 12369 + }, + { + "epoch": 0.2899722803377931, + "grad_norm": 2.2015557289123535, + "learning_rate": 0.0001613384796918923, + "loss": 1.4555, + "step": 12370 + }, + { + "epoch": 0.2899957219125981, + "grad_norm": 3.45186185836792, + "learning_rate": 0.0001613326618503764, + "loss": 0.4589, + "step": 12371 + }, + { + "epoch": 0.2900191634874031, + "grad_norm": 0.6212366819381714, + "learning_rate": 0.00016132684367606778, + "loss": 0.087, + "step": 12372 + }, + { + "epoch": 0.2900426050622081, + "grad_norm": 4.2289628982543945, + "learning_rate": 0.00016132102516899799, + "loss": 0.4952, + "step": 12373 + }, + { + "epoch": 0.2900660466370131, + "grad_norm": 0.4014628231525421, + "learning_rate": 0.00016131520632919858, + "loss": 0.0315, + "step": 12374 + }, + { + "epoch": 0.29008948821181807, + "grad_norm": 3.3445444107055664, + "learning_rate": 0.00016130938715670117, + "loss": 0.4274, + "step": 12375 + }, + { + "epoch": 0.29011292978662306, + "grad_norm": 4.276950836181641, + "learning_rate": 0.00016130356765153728, + "loss": 0.3791, + "step": 12376 + }, + { + "epoch": 0.29013637136142806, + "grad_norm": 3.270540475845337, + "learning_rate": 0.00016129774781373848, + "loss": 0.4147, + "step": 12377 + }, + { + "epoch": 0.29015981293623305, + "grad_norm": 5.203123092651367, + "learning_rate": 0.00016129192764333637, + "loss": 0.511, + "step": 12378 + }, + { + "epoch": 0.29018325451103805, + "grad_norm": 2.159574270248413, + "learning_rate": 0.0001612861071403626, + "loss": 0.2813, + "step": 12379 + }, + { + "epoch": 0.29020669608584304, + "grad_norm": 1.075139045715332, + "learning_rate": 0.00016128028630484864, + "loss": 0.0903, + "step": 12380 + }, + { + "epoch": 0.29023013766064804, + "grad_norm": 4.38131046295166, + "learning_rate": 0.00016127446513682612, + "loss": 0.4205, + "step": 12381 + }, + { + "epoch": 0.29025357923545303, + "grad_norm": 3.0065643787384033, + "learning_rate": 0.00016126864363632666, + "loss": 0.2012, + "step": 12382 + }, + { + "epoch": 0.290277020810258, + "grad_norm": 3.338089942932129, + "learning_rate": 0.00016126282180338182, + "loss": 0.3618, + "step": 12383 + }, + { + "epoch": 0.290300462385063, + "grad_norm": 4.996243953704834, + "learning_rate": 0.00016125699963802313, + "loss": 0.7359, + "step": 12384 + }, + { + "epoch": 0.290323903959868, + "grad_norm": 2.739165782928467, + "learning_rate": 0.0001612511771402823, + "loss": 0.3132, + "step": 12385 + }, + { + "epoch": 0.290347345534673, + "grad_norm": 3.4720423221588135, + "learning_rate": 0.00016124535431019085, + "loss": 0.3871, + "step": 12386 + }, + { + "epoch": 0.290370787109478, + "grad_norm": 3.026646614074707, + "learning_rate": 0.00016123953114778039, + "loss": 0.3144, + "step": 12387 + }, + { + "epoch": 0.290394228684283, + "grad_norm": 0.6124436855316162, + "learning_rate": 0.0001612337076530825, + "loss": 0.069, + "step": 12388 + }, + { + "epoch": 0.290417670259088, + "grad_norm": 0.7830132246017456, + "learning_rate": 0.0001612278838261288, + "loss": 0.0745, + "step": 12389 + }, + { + "epoch": 0.290441111833893, + "grad_norm": 6.461580276489258, + "learning_rate": 0.00016122205966695083, + "loss": 0.638, + "step": 12390 + }, + { + "epoch": 0.290464553408698, + "grad_norm": 2.7556538581848145, + "learning_rate": 0.00016121623517558029, + "loss": 0.2292, + "step": 12391 + }, + { + "epoch": 0.290487994983503, + "grad_norm": 3.6764934062957764, + "learning_rate": 0.00016121041035204874, + "loss": 0.33, + "step": 12392 + }, + { + "epoch": 0.29051143655830797, + "grad_norm": 1.663802146911621, + "learning_rate": 0.00016120458519638774, + "loss": 0.1901, + "step": 12393 + }, + { + "epoch": 0.29053487813311296, + "grad_norm": 3.4207723140716553, + "learning_rate": 0.000161198759708629, + "loss": 0.3844, + "step": 12394 + }, + { + "epoch": 0.29055831970791796, + "grad_norm": 2.4901435375213623, + "learning_rate": 0.00016119293388880404, + "loss": 0.3961, + "step": 12395 + }, + { + "epoch": 0.29058176128272295, + "grad_norm": 3.4414708614349365, + "learning_rate": 0.0001611871077369445, + "loss": 0.3701, + "step": 12396 + }, + { + "epoch": 0.29060520285752794, + "grad_norm": 3.555058240890503, + "learning_rate": 0.00016118128125308195, + "loss": 0.3509, + "step": 12397 + }, + { + "epoch": 0.29062864443233294, + "grad_norm": 1.590979814529419, + "learning_rate": 0.00016117545443724811, + "loss": 0.2005, + "step": 12398 + }, + { + "epoch": 0.29065208600713793, + "grad_norm": 1.9702069759368896, + "learning_rate": 0.0001611696272894745, + "loss": 1.0164, + "step": 12399 + }, + { + "epoch": 0.2906755275819429, + "grad_norm": 4.305152416229248, + "learning_rate": 0.0001611637998097928, + "loss": 0.6943, + "step": 12400 + }, + { + "epoch": 0.290698969156748, + "grad_norm": 3.2068731784820557, + "learning_rate": 0.0001611579719982346, + "loss": 0.3244, + "step": 12401 + }, + { + "epoch": 0.29072241073155297, + "grad_norm": 1.0546631813049316, + "learning_rate": 0.0001611521438548315, + "loss": 0.0918, + "step": 12402 + }, + { + "epoch": 0.29074585230635797, + "grad_norm": 2.1763410568237305, + "learning_rate": 0.00016114631537961514, + "loss": 1.2815, + "step": 12403 + }, + { + "epoch": 0.29076929388116296, + "grad_norm": 5.519796848297119, + "learning_rate": 0.0001611404865726172, + "loss": 0.3, + "step": 12404 + }, + { + "epoch": 0.29079273545596795, + "grad_norm": 3.4669106006622314, + "learning_rate": 0.00016113465743386923, + "loss": 0.4602, + "step": 12405 + }, + { + "epoch": 0.29081617703077295, + "grad_norm": 4.2863287925720215, + "learning_rate": 0.00016112882796340288, + "loss": 0.3898, + "step": 12406 + }, + { + "epoch": 0.29083961860557794, + "grad_norm": 3.269568920135498, + "learning_rate": 0.00016112299816124981, + "loss": 0.4078, + "step": 12407 + }, + { + "epoch": 0.29086306018038294, + "grad_norm": 2.780024290084839, + "learning_rate": 0.00016111716802744166, + "loss": 0.1754, + "step": 12408 + }, + { + "epoch": 0.29088650175518793, + "grad_norm": 2.834998846054077, + "learning_rate": 0.00016111133756201, + "loss": 0.3273, + "step": 12409 + }, + { + "epoch": 0.2909099433299929, + "grad_norm": 4.373642444610596, + "learning_rate": 0.0001611055067649865, + "loss": 0.4371, + "step": 12410 + }, + { + "epoch": 0.2909333849047979, + "grad_norm": 8.4756498336792, + "learning_rate": 0.0001610996756364028, + "loss": 0.6746, + "step": 12411 + }, + { + "epoch": 0.2909568264796029, + "grad_norm": 6.781175136566162, + "learning_rate": 0.0001610938441762906, + "loss": 0.7668, + "step": 12412 + }, + { + "epoch": 0.2909802680544079, + "grad_norm": 6.421189308166504, + "learning_rate": 0.0001610880123846814, + "loss": 0.6405, + "step": 12413 + }, + { + "epoch": 0.2910037096292129, + "grad_norm": 4.684745788574219, + "learning_rate": 0.00016108218026160697, + "loss": 0.7597, + "step": 12414 + }, + { + "epoch": 0.2910271512040179, + "grad_norm": 4.077468395233154, + "learning_rate": 0.0001610763478070989, + "loss": 0.4583, + "step": 12415 + }, + { + "epoch": 0.2910505927788229, + "grad_norm": 4.124349594116211, + "learning_rate": 0.00016107051502118886, + "loss": 0.3209, + "step": 12416 + }, + { + "epoch": 0.2910740343536279, + "grad_norm": 3.1005046367645264, + "learning_rate": 0.00016106468190390852, + "loss": 0.2906, + "step": 12417 + }, + { + "epoch": 0.2910974759284329, + "grad_norm": 3.409604787826538, + "learning_rate": 0.00016105884845528945, + "loss": 0.2752, + "step": 12418 + }, + { + "epoch": 0.2911209175032379, + "grad_norm": 4.216002941131592, + "learning_rate": 0.00016105301467536336, + "loss": 0.3799, + "step": 12419 + }, + { + "epoch": 0.29114435907804287, + "grad_norm": 4.40647554397583, + "learning_rate": 0.0001610471805641619, + "loss": 0.6001, + "step": 12420 + }, + { + "epoch": 0.29116780065284786, + "grad_norm": 4.323115348815918, + "learning_rate": 0.00016104134612171674, + "loss": 0.6219, + "step": 12421 + }, + { + "epoch": 0.29119124222765286, + "grad_norm": 4.737237930297852, + "learning_rate": 0.0001610355113480595, + "loss": 0.5906, + "step": 12422 + }, + { + "epoch": 0.29121468380245785, + "grad_norm": 3.3167941570281982, + "learning_rate": 0.00016102967624322185, + "loss": 0.282, + "step": 12423 + }, + { + "epoch": 0.29123812537726285, + "grad_norm": 4.467666149139404, + "learning_rate": 0.0001610238408072355, + "loss": 0.5149, + "step": 12424 + }, + { + "epoch": 0.29126156695206784, + "grad_norm": 0.8940538763999939, + "learning_rate": 0.00016101800504013202, + "loss": 0.0927, + "step": 12425 + }, + { + "epoch": 0.29128500852687284, + "grad_norm": 2.3382763862609863, + "learning_rate": 0.00016101216894194315, + "loss": 0.3222, + "step": 12426 + }, + { + "epoch": 0.29130845010167783, + "grad_norm": 3.901000738143921, + "learning_rate": 0.00016100633251270056, + "loss": 0.4245, + "step": 12427 + }, + { + "epoch": 0.2913318916764828, + "grad_norm": 2.2369534969329834, + "learning_rate": 0.00016100049575243586, + "loss": 0.2743, + "step": 12428 + }, + { + "epoch": 0.2913553332512878, + "grad_norm": 4.971261024475098, + "learning_rate": 0.0001609946586611808, + "loss": 0.8547, + "step": 12429 + }, + { + "epoch": 0.2913787748260928, + "grad_norm": 2.1594910621643066, + "learning_rate": 0.00016098882123896698, + "loss": 0.3029, + "step": 12430 + }, + { + "epoch": 0.2914022164008978, + "grad_norm": 2.9697208404541016, + "learning_rate": 0.0001609829834858261, + "loss": 0.2437, + "step": 12431 + }, + { + "epoch": 0.2914256579757028, + "grad_norm": 2.2640979290008545, + "learning_rate": 0.00016097714540178984, + "loss": 0.1898, + "step": 12432 + }, + { + "epoch": 0.2914490995505078, + "grad_norm": 3.6495776176452637, + "learning_rate": 0.00016097130698688983, + "loss": 0.5619, + "step": 12433 + }, + { + "epoch": 0.2914725411253128, + "grad_norm": 1.1302169561386108, + "learning_rate": 0.00016096546824115784, + "loss": 0.1452, + "step": 12434 + }, + { + "epoch": 0.2914959827001178, + "grad_norm": 4.184727191925049, + "learning_rate": 0.0001609596291646255, + "loss": 0.5433, + "step": 12435 + }, + { + "epoch": 0.2915194242749228, + "grad_norm": 1.947924256324768, + "learning_rate": 0.0001609537897573245, + "loss": 1.437, + "step": 12436 + }, + { + "epoch": 0.2915428658497278, + "grad_norm": 3.0776422023773193, + "learning_rate": 0.00016094795001928653, + "loss": 0.2679, + "step": 12437 + }, + { + "epoch": 0.29156630742453277, + "grad_norm": 2.4604175090789795, + "learning_rate": 0.00016094210995054326, + "loss": 0.1934, + "step": 12438 + }, + { + "epoch": 0.29158974899933776, + "grad_norm": 5.107697486877441, + "learning_rate": 0.00016093626955112638, + "loss": 0.4113, + "step": 12439 + }, + { + "epoch": 0.29161319057414276, + "grad_norm": 5.66780948638916, + "learning_rate": 0.00016093042882106762, + "loss": 0.6531, + "step": 12440 + }, + { + "epoch": 0.29163663214894775, + "grad_norm": 3.7272326946258545, + "learning_rate": 0.0001609245877603986, + "loss": 0.6492, + "step": 12441 + }, + { + "epoch": 0.29166007372375274, + "grad_norm": 3.7382330894470215, + "learning_rate": 0.00016091874636915107, + "loss": 0.4798, + "step": 12442 + }, + { + "epoch": 0.29168351529855774, + "grad_norm": 2.5636563301086426, + "learning_rate": 0.0001609129046473567, + "loss": 0.3038, + "step": 12443 + }, + { + "epoch": 0.29170695687336273, + "grad_norm": 4.0180158615112305, + "learning_rate": 0.00016090706259504724, + "loss": 0.4727, + "step": 12444 + }, + { + "epoch": 0.2917303984481677, + "grad_norm": 5.610132694244385, + "learning_rate": 0.0001609012202122543, + "loss": 0.592, + "step": 12445 + }, + { + "epoch": 0.2917538400229727, + "grad_norm": 4.610368251800537, + "learning_rate": 0.00016089537749900963, + "loss": 0.7002, + "step": 12446 + }, + { + "epoch": 0.2917772815977777, + "grad_norm": 4.611566543579102, + "learning_rate": 0.00016088953445534494, + "loss": 0.456, + "step": 12447 + }, + { + "epoch": 0.2918007231725827, + "grad_norm": 2.1371920108795166, + "learning_rate": 0.00016088369108129192, + "loss": 1.0181, + "step": 12448 + }, + { + "epoch": 0.2918241647473877, + "grad_norm": 3.3712828159332275, + "learning_rate": 0.00016087784737688228, + "loss": 0.2932, + "step": 12449 + }, + { + "epoch": 0.2918476063221927, + "grad_norm": 2.4461567401885986, + "learning_rate": 0.00016087200334214777, + "loss": 0.1451, + "step": 12450 + }, + { + "epoch": 0.2918710478969977, + "grad_norm": 0.6631313562393188, + "learning_rate": 0.00016086615897711998, + "loss": 0.0675, + "step": 12451 + }, + { + "epoch": 0.29189448947180274, + "grad_norm": 5.6234283447265625, + "learning_rate": 0.00016086031428183076, + "loss": 0.5028, + "step": 12452 + }, + { + "epoch": 0.29191793104660774, + "grad_norm": 1.0241289138793945, + "learning_rate": 0.00016085446925631175, + "loss": 0.1194, + "step": 12453 + }, + { + "epoch": 0.29194137262141273, + "grad_norm": 3.816603899002075, + "learning_rate": 0.00016084862390059466, + "loss": 0.3938, + "step": 12454 + }, + { + "epoch": 0.2919648141962177, + "grad_norm": 3.822376012802124, + "learning_rate": 0.00016084277821471122, + "loss": 0.2423, + "step": 12455 + }, + { + "epoch": 0.2919882557710227, + "grad_norm": 5.221504211425781, + "learning_rate": 0.0001608369321986932, + "loss": 0.5046, + "step": 12456 + }, + { + "epoch": 0.2920116973458277, + "grad_norm": 1.890256404876709, + "learning_rate": 0.00016083108585257224, + "loss": 0.2873, + "step": 12457 + }, + { + "epoch": 0.2920351389206327, + "grad_norm": 6.686407566070557, + "learning_rate": 0.00016082523917638008, + "loss": 0.6596, + "step": 12458 + }, + { + "epoch": 0.2920585804954377, + "grad_norm": 5.679031848907471, + "learning_rate": 0.0001608193921701485, + "loss": 0.615, + "step": 12459 + }, + { + "epoch": 0.2920820220702427, + "grad_norm": 5.034817218780518, + "learning_rate": 0.00016081354483390918, + "loss": 0.5257, + "step": 12460 + }, + { + "epoch": 0.2921054636450477, + "grad_norm": 5.50150203704834, + "learning_rate": 0.00016080769716769386, + "loss": 0.5407, + "step": 12461 + }, + { + "epoch": 0.2921289052198527, + "grad_norm": 8.317105293273926, + "learning_rate": 0.00016080184917153427, + "loss": 0.4615, + "step": 12462 + }, + { + "epoch": 0.2921523467946577, + "grad_norm": 2.2773706912994385, + "learning_rate": 0.0001607960008454621, + "loss": 0.1881, + "step": 12463 + }, + { + "epoch": 0.2921757883694627, + "grad_norm": 3.0550334453582764, + "learning_rate": 0.00016079015218950913, + "loss": 0.21, + "step": 12464 + }, + { + "epoch": 0.29219922994426767, + "grad_norm": 1.4816687107086182, + "learning_rate": 0.00016078430320370712, + "loss": 0.0876, + "step": 12465 + }, + { + "epoch": 0.29222267151907266, + "grad_norm": 3.4549481868743896, + "learning_rate": 0.00016077845388808774, + "loss": 0.237, + "step": 12466 + }, + { + "epoch": 0.29224611309387766, + "grad_norm": 2.198986053466797, + "learning_rate": 0.0001607726042426828, + "loss": 0.2486, + "step": 12467 + }, + { + "epoch": 0.29226955466868265, + "grad_norm": 6.62671422958374, + "learning_rate": 0.00016076675426752394, + "loss": 0.3752, + "step": 12468 + }, + { + "epoch": 0.29229299624348765, + "grad_norm": 2.626688003540039, + "learning_rate": 0.000160760903962643, + "loss": 0.2167, + "step": 12469 + }, + { + "epoch": 0.29231643781829264, + "grad_norm": 1.929295539855957, + "learning_rate": 0.00016075505332807167, + "loss": 1.1076, + "step": 12470 + }, + { + "epoch": 0.29233987939309763, + "grad_norm": 3.4845046997070312, + "learning_rate": 0.0001607492023638417, + "loss": 0.318, + "step": 12471 + }, + { + "epoch": 0.29236332096790263, + "grad_norm": 4.1524176597595215, + "learning_rate": 0.0001607433510699849, + "loss": 0.4404, + "step": 12472 + }, + { + "epoch": 0.2923867625427076, + "grad_norm": 2.0080373287200928, + "learning_rate": 0.00016073749944653295, + "loss": 0.1977, + "step": 12473 + }, + { + "epoch": 0.2924102041175126, + "grad_norm": 2.325291395187378, + "learning_rate": 0.0001607316474935176, + "loss": 0.1773, + "step": 12474 + }, + { + "epoch": 0.2924336456923176, + "grad_norm": 0.8733806610107422, + "learning_rate": 0.00016072579521097063, + "loss": 0.0781, + "step": 12475 + }, + { + "epoch": 0.2924570872671226, + "grad_norm": 2.325727701187134, + "learning_rate": 0.0001607199425989238, + "loss": 0.1923, + "step": 12476 + }, + { + "epoch": 0.2924805288419276, + "grad_norm": 0.9042177796363831, + "learning_rate": 0.00016071408965740884, + "loss": 0.0989, + "step": 12477 + }, + { + "epoch": 0.2925039704167326, + "grad_norm": 3.6470088958740234, + "learning_rate": 0.00016070823638645755, + "loss": 0.4012, + "step": 12478 + }, + { + "epoch": 0.2925274119915376, + "grad_norm": 5.6527910232543945, + "learning_rate": 0.00016070238278610162, + "loss": 0.4787, + "step": 12479 + }, + { + "epoch": 0.2925508535663426, + "grad_norm": 1.5135135650634766, + "learning_rate": 0.0001606965288563729, + "loss": 0.1899, + "step": 12480 + }, + { + "epoch": 0.2925742951411476, + "grad_norm": 3.7178096771240234, + "learning_rate": 0.00016069067459730305, + "loss": 0.5812, + "step": 12481 + }, + { + "epoch": 0.29259773671595257, + "grad_norm": 2.8150274753570557, + "learning_rate": 0.00016068482000892393, + "loss": 0.1877, + "step": 12482 + }, + { + "epoch": 0.29262117829075757, + "grad_norm": 3.3764090538024902, + "learning_rate": 0.00016067896509126727, + "loss": 0.2322, + "step": 12483 + }, + { + "epoch": 0.29264461986556256, + "grad_norm": 1.8588519096374512, + "learning_rate": 0.00016067310984436484, + "loss": 0.1761, + "step": 12484 + }, + { + "epoch": 0.29266806144036756, + "grad_norm": 2.6887145042419434, + "learning_rate": 0.0001606672542682484, + "loss": 0.241, + "step": 12485 + }, + { + "epoch": 0.29269150301517255, + "grad_norm": 4.127101421356201, + "learning_rate": 0.00016066139836294976, + "loss": 1.6564, + "step": 12486 + }, + { + "epoch": 0.29271494458997754, + "grad_norm": 4.849165916442871, + "learning_rate": 0.00016065554212850062, + "loss": 0.663, + "step": 12487 + }, + { + "epoch": 0.29273838616478254, + "grad_norm": 1.0847139358520508, + "learning_rate": 0.00016064968556493283, + "loss": 0.0978, + "step": 12488 + }, + { + "epoch": 0.29276182773958753, + "grad_norm": 3.873542308807373, + "learning_rate": 0.00016064382867227815, + "loss": 0.5589, + "step": 12489 + }, + { + "epoch": 0.2927852693143925, + "grad_norm": 5.022409915924072, + "learning_rate": 0.00016063797145056836, + "loss": 0.3674, + "step": 12490 + }, + { + "epoch": 0.2928087108891975, + "grad_norm": 1.8507214784622192, + "learning_rate": 0.0001606321138998352, + "loss": 0.2226, + "step": 12491 + }, + { + "epoch": 0.2928321524640025, + "grad_norm": 3.6049914360046387, + "learning_rate": 0.0001606262560201105, + "loss": 0.4141, + "step": 12492 + }, + { + "epoch": 0.2928555940388075, + "grad_norm": 3.5163583755493164, + "learning_rate": 0.00016062039781142602, + "loss": 0.2288, + "step": 12493 + }, + { + "epoch": 0.2928790356136125, + "grad_norm": 2.2924559116363525, + "learning_rate": 0.00016061453927381357, + "loss": 0.2955, + "step": 12494 + }, + { + "epoch": 0.2929024771884175, + "grad_norm": 6.105802059173584, + "learning_rate": 0.0001606086804073049, + "loss": 0.6325, + "step": 12495 + }, + { + "epoch": 0.2929259187632225, + "grad_norm": 5.084641456604004, + "learning_rate": 0.00016060282121193185, + "loss": 0.4567, + "step": 12496 + }, + { + "epoch": 0.2929493603380275, + "grad_norm": 0.920385479927063, + "learning_rate": 0.0001605969616877262, + "loss": 0.0482, + "step": 12497 + }, + { + "epoch": 0.2929728019128325, + "grad_norm": 2.9290480613708496, + "learning_rate": 0.00016059110183471971, + "loss": 0.345, + "step": 12498 + }, + { + "epoch": 0.2929962434876375, + "grad_norm": 4.874450206756592, + "learning_rate": 0.0001605852416529442, + "loss": 0.4035, + "step": 12499 + }, + { + "epoch": 0.29301968506244247, + "grad_norm": 3.4259722232818604, + "learning_rate": 0.00016057938114243147, + "loss": 0.4319, + "step": 12500 + }, + { + "epoch": 0.29304312663724746, + "grad_norm": 4.413903713226318, + "learning_rate": 0.0001605735203032133, + "loss": 0.4287, + "step": 12501 + }, + { + "epoch": 0.29306656821205246, + "grad_norm": 0.9576515555381775, + "learning_rate": 0.00016056765913532151, + "loss": 0.0734, + "step": 12502 + }, + { + "epoch": 0.2930900097868575, + "grad_norm": 3.224271297454834, + "learning_rate": 0.00016056179763878793, + "loss": 0.2952, + "step": 12503 + }, + { + "epoch": 0.2931134513616625, + "grad_norm": 10.289584159851074, + "learning_rate": 0.00016055593581364432, + "loss": 0.629, + "step": 12504 + }, + { + "epoch": 0.2931368929364675, + "grad_norm": 2.5027835369110107, + "learning_rate": 0.0001605500736599225, + "loss": 1.8601, + "step": 12505 + }, + { + "epoch": 0.2931603345112725, + "grad_norm": 3.141723394393921, + "learning_rate": 0.00016054421117765423, + "loss": 0.3538, + "step": 12506 + }, + { + "epoch": 0.2931837760860775, + "grad_norm": 4.250760078430176, + "learning_rate": 0.00016053834836687143, + "loss": 0.3561, + "step": 12507 + }, + { + "epoch": 0.2932072176608825, + "grad_norm": 3.0597896575927734, + "learning_rate": 0.00016053248522760583, + "loss": 0.3402, + "step": 12508 + }, + { + "epoch": 0.2932306592356875, + "grad_norm": 3.187898874282837, + "learning_rate": 0.00016052662175988926, + "loss": 0.3855, + "step": 12509 + }, + { + "epoch": 0.29325410081049247, + "grad_norm": 4.348997116088867, + "learning_rate": 0.00016052075796375352, + "loss": 0.6547, + "step": 12510 + }, + { + "epoch": 0.29327754238529746, + "grad_norm": 10.643501281738281, + "learning_rate": 0.00016051489383923047, + "loss": 0.2843, + "step": 12511 + }, + { + "epoch": 0.29330098396010246, + "grad_norm": 1.4523086547851562, + "learning_rate": 0.00016050902938635192, + "loss": 0.1688, + "step": 12512 + }, + { + "epoch": 0.29332442553490745, + "grad_norm": 1.5799168348312378, + "learning_rate": 0.00016050316460514962, + "loss": 0.1033, + "step": 12513 + }, + { + "epoch": 0.29334786710971245, + "grad_norm": 0.3238163888454437, + "learning_rate": 0.0001604972994956555, + "loss": 0.0216, + "step": 12514 + }, + { + "epoch": 0.29337130868451744, + "grad_norm": 1.8320757150650024, + "learning_rate": 0.0001604914340579013, + "loss": 0.1506, + "step": 12515 + }, + { + "epoch": 0.29339475025932243, + "grad_norm": 4.704261779785156, + "learning_rate": 0.0001604855682919189, + "loss": 0.6083, + "step": 12516 + }, + { + "epoch": 0.29341819183412743, + "grad_norm": 2.8638412952423096, + "learning_rate": 0.00016047970219774008, + "loss": 0.2504, + "step": 12517 + }, + { + "epoch": 0.2934416334089324, + "grad_norm": 2.3495869636535645, + "learning_rate": 0.00016047383577539673, + "loss": 1.6323, + "step": 12518 + }, + { + "epoch": 0.2934650749837374, + "grad_norm": 5.335143566131592, + "learning_rate": 0.0001604679690249206, + "loss": 1.1327, + "step": 12519 + }, + { + "epoch": 0.2934885165585424, + "grad_norm": 5.14103364944458, + "learning_rate": 0.00016046210194634362, + "loss": 0.5787, + "step": 12520 + }, + { + "epoch": 0.2935119581333474, + "grad_norm": 3.0981504917144775, + "learning_rate": 0.00016045623453969757, + "loss": 0.3687, + "step": 12521 + }, + { + "epoch": 0.2935353997081524, + "grad_norm": 3.173283338546753, + "learning_rate": 0.00016045036680501426, + "loss": 0.1988, + "step": 12522 + }, + { + "epoch": 0.2935588412829574, + "grad_norm": 3.9482648372650146, + "learning_rate": 0.00016044449874232555, + "loss": 0.3008, + "step": 12523 + }, + { + "epoch": 0.2935822828577624, + "grad_norm": 4.533076763153076, + "learning_rate": 0.0001604386303516633, + "loss": 0.6954, + "step": 12524 + }, + { + "epoch": 0.2936057244325674, + "grad_norm": 3.623067617416382, + "learning_rate": 0.0001604327616330594, + "loss": 0.428, + "step": 12525 + }, + { + "epoch": 0.2936291660073724, + "grad_norm": 2.5572664737701416, + "learning_rate": 0.00016042689258654558, + "loss": 0.3314, + "step": 12526 + }, + { + "epoch": 0.29365260758217737, + "grad_norm": 1.968448519706726, + "learning_rate": 0.00016042102321215374, + "loss": 0.2315, + "step": 12527 + }, + { + "epoch": 0.29367604915698237, + "grad_norm": 1.6412532329559326, + "learning_rate": 0.00016041515350991573, + "loss": 0.1416, + "step": 12528 + }, + { + "epoch": 0.29369949073178736, + "grad_norm": 2.453782320022583, + "learning_rate": 0.00016040928347986345, + "loss": 1.414, + "step": 12529 + }, + { + "epoch": 0.29372293230659235, + "grad_norm": 6.9986162185668945, + "learning_rate": 0.00016040341312202863, + "loss": 0.6773, + "step": 12530 + }, + { + "epoch": 0.29374637388139735, + "grad_norm": 0.8150745630264282, + "learning_rate": 0.00016039754243644323, + "loss": 0.1386, + "step": 12531 + }, + { + "epoch": 0.29376981545620234, + "grad_norm": 3.119523286819458, + "learning_rate": 0.00016039167142313905, + "loss": 0.3591, + "step": 12532 + }, + { + "epoch": 0.29379325703100734, + "grad_norm": 1.9627985954284668, + "learning_rate": 0.000160385800082148, + "loss": 0.2305, + "step": 12533 + }, + { + "epoch": 0.29381669860581233, + "grad_norm": 3.5259041786193848, + "learning_rate": 0.00016037992841350186, + "loss": 0.4793, + "step": 12534 + }, + { + "epoch": 0.2938401401806173, + "grad_norm": 2.816338062286377, + "learning_rate": 0.00016037405641723256, + "loss": 0.3736, + "step": 12535 + }, + { + "epoch": 0.2938635817554223, + "grad_norm": 3.4649107456207275, + "learning_rate": 0.00016036818409337187, + "loss": 0.3919, + "step": 12536 + }, + { + "epoch": 0.2938870233302273, + "grad_norm": 3.1410422325134277, + "learning_rate": 0.00016036231144195178, + "loss": 0.3325, + "step": 12537 + }, + { + "epoch": 0.2939104649050323, + "grad_norm": 1.6376475095748901, + "learning_rate": 0.00016035643846300408, + "loss": 0.156, + "step": 12538 + }, + { + "epoch": 0.2939339064798373, + "grad_norm": 0.6369478106498718, + "learning_rate": 0.00016035056515656065, + "loss": 0.0274, + "step": 12539 + }, + { + "epoch": 0.2939573480546423, + "grad_norm": 3.37610125541687, + "learning_rate": 0.00016034469152265336, + "loss": 0.3999, + "step": 12540 + }, + { + "epoch": 0.2939807896294473, + "grad_norm": 2.534210681915283, + "learning_rate": 0.00016033881756131405, + "loss": 0.1652, + "step": 12541 + }, + { + "epoch": 0.2940042312042523, + "grad_norm": 2.7313177585601807, + "learning_rate": 0.00016033294327257464, + "loss": 1.4943, + "step": 12542 + }, + { + "epoch": 0.2940276727790573, + "grad_norm": 4.877679824829102, + "learning_rate": 0.00016032706865646697, + "loss": 0.8759, + "step": 12543 + }, + { + "epoch": 0.2940511143538623, + "grad_norm": 4.383555889129639, + "learning_rate": 0.00016032119371302294, + "loss": 0.5728, + "step": 12544 + }, + { + "epoch": 0.29407455592866727, + "grad_norm": 1.0762919187545776, + "learning_rate": 0.0001603153184422744, + "loss": 0.1004, + "step": 12545 + }, + { + "epoch": 0.29409799750347226, + "grad_norm": 3.641768217086792, + "learning_rate": 0.00016030944284425326, + "loss": 0.4079, + "step": 12546 + }, + { + "epoch": 0.29412143907827726, + "grad_norm": 3.6129326820373535, + "learning_rate": 0.0001603035669189914, + "loss": 0.6073, + "step": 12547 + }, + { + "epoch": 0.29414488065308225, + "grad_norm": 1.9612236022949219, + "learning_rate": 0.00016029769066652064, + "loss": 0.164, + "step": 12548 + }, + { + "epoch": 0.29416832222788725, + "grad_norm": 3.4245951175689697, + "learning_rate": 0.00016029181408687296, + "loss": 0.4857, + "step": 12549 + }, + { + "epoch": 0.29419176380269224, + "grad_norm": 4.144081115722656, + "learning_rate": 0.00016028593718008023, + "loss": 0.3877, + "step": 12550 + }, + { + "epoch": 0.29421520537749724, + "grad_norm": 3.002833604812622, + "learning_rate": 0.00016028005994617428, + "loss": 2.1484, + "step": 12551 + }, + { + "epoch": 0.29423864695230223, + "grad_norm": 1.4640405178070068, + "learning_rate": 0.000160274182385187, + "loss": 0.1162, + "step": 12552 + }, + { + "epoch": 0.2942620885271072, + "grad_norm": 5.027734756469727, + "learning_rate": 0.00016026830449715034, + "loss": 0.4793, + "step": 12553 + }, + { + "epoch": 0.2942855301019123, + "grad_norm": 1.6641966104507446, + "learning_rate": 0.00016026242628209615, + "loss": 0.1251, + "step": 12554 + }, + { + "epoch": 0.29430897167671727, + "grad_norm": 3.4473626613616943, + "learning_rate": 0.00016025654774005633, + "loss": 0.215, + "step": 12555 + }, + { + "epoch": 0.29433241325152226, + "grad_norm": 4.016225337982178, + "learning_rate": 0.00016025066887106283, + "loss": 0.4772, + "step": 12556 + }, + { + "epoch": 0.29435585482632726, + "grad_norm": 4.428532123565674, + "learning_rate": 0.0001602447896751475, + "loss": 0.7409, + "step": 12557 + }, + { + "epoch": 0.29437929640113225, + "grad_norm": 3.2243614196777344, + "learning_rate": 0.00016023891015234224, + "loss": 1.856, + "step": 12558 + }, + { + "epoch": 0.29440273797593725, + "grad_norm": 1.5795726776123047, + "learning_rate": 0.00016023303030267893, + "loss": 1.2423, + "step": 12559 + }, + { + "epoch": 0.29442617955074224, + "grad_norm": 2.851325035095215, + "learning_rate": 0.0001602271501261895, + "loss": 0.2257, + "step": 12560 + }, + { + "epoch": 0.29444962112554723, + "grad_norm": 2.2001726627349854, + "learning_rate": 0.00016022126962290592, + "loss": 0.2398, + "step": 12561 + }, + { + "epoch": 0.29447306270035223, + "grad_norm": 2.7885591983795166, + "learning_rate": 0.00016021538879286, + "loss": 0.4547, + "step": 12562 + }, + { + "epoch": 0.2944965042751572, + "grad_norm": 3.615466594696045, + "learning_rate": 0.0001602095076360837, + "loss": 0.5664, + "step": 12563 + }, + { + "epoch": 0.2945199458499622, + "grad_norm": 2.8059887886047363, + "learning_rate": 0.00016020362615260889, + "loss": 0.2141, + "step": 12564 + }, + { + "epoch": 0.2945433874247672, + "grad_norm": 2.1261165142059326, + "learning_rate": 0.00016019774434246754, + "loss": 1.7989, + "step": 12565 + }, + { + "epoch": 0.2945668289995722, + "grad_norm": 5.369345664978027, + "learning_rate": 0.00016019186220569154, + "loss": 0.5281, + "step": 12566 + }, + { + "epoch": 0.2945902705743772, + "grad_norm": 2.2804319858551025, + "learning_rate": 0.00016018597974231278, + "loss": 1.906, + "step": 12567 + }, + { + "epoch": 0.2946137121491822, + "grad_norm": 24.618234634399414, + "learning_rate": 0.0001601800969523632, + "loss": 0.4889, + "step": 12568 + }, + { + "epoch": 0.2946371537239872, + "grad_norm": 2.470151662826538, + "learning_rate": 0.00016017421383587475, + "loss": 0.3426, + "step": 12569 + }, + { + "epoch": 0.2946605952987922, + "grad_norm": 3.2441630363464355, + "learning_rate": 0.00016016833039287932, + "loss": 0.4511, + "step": 12570 + }, + { + "epoch": 0.2946840368735972, + "grad_norm": 0.5964793562889099, + "learning_rate": 0.0001601624466234088, + "loss": 0.0914, + "step": 12571 + }, + { + "epoch": 0.29470747844840217, + "grad_norm": 4.323572635650635, + "learning_rate": 0.00016015656252749518, + "loss": 0.4548, + "step": 12572 + }, + { + "epoch": 0.29473092002320717, + "grad_norm": 3.788886070251465, + "learning_rate": 0.00016015067810517034, + "loss": 0.3437, + "step": 12573 + }, + { + "epoch": 0.29475436159801216, + "grad_norm": 6.801175117492676, + "learning_rate": 0.00016014479335646627, + "loss": 0.6534, + "step": 12574 + }, + { + "epoch": 0.29477780317281715, + "grad_norm": 2.2079787254333496, + "learning_rate": 0.00016013890828141483, + "loss": 0.3404, + "step": 12575 + }, + { + "epoch": 0.29480124474762215, + "grad_norm": 0.9256333112716675, + "learning_rate": 0.000160133022880048, + "loss": 0.1042, + "step": 12576 + }, + { + "epoch": 0.29482468632242714, + "grad_norm": 3.9768025875091553, + "learning_rate": 0.00016012713715239768, + "loss": 0.4939, + "step": 12577 + }, + { + "epoch": 0.29484812789723214, + "grad_norm": 5.6815924644470215, + "learning_rate": 0.00016012125109849583, + "loss": 0.4871, + "step": 12578 + }, + { + "epoch": 0.29487156947203713, + "grad_norm": 5.450558662414551, + "learning_rate": 0.00016011536471837438, + "loss": 0.7517, + "step": 12579 + }, + { + "epoch": 0.2948950110468421, + "grad_norm": 3.009956121444702, + "learning_rate": 0.00016010947801206528, + "loss": 0.3043, + "step": 12580 + }, + { + "epoch": 0.2949184526216471, + "grad_norm": 2.4090871810913086, + "learning_rate": 0.00016010359097960047, + "loss": 0.2755, + "step": 12581 + }, + { + "epoch": 0.2949418941964521, + "grad_norm": 3.1906726360321045, + "learning_rate": 0.00016009770362101187, + "loss": 0.3332, + "step": 12582 + }, + { + "epoch": 0.2949653357712571, + "grad_norm": 4.029193878173828, + "learning_rate": 0.00016009181593633145, + "loss": 0.6419, + "step": 12583 + }, + { + "epoch": 0.2949887773460621, + "grad_norm": 2.681089401245117, + "learning_rate": 0.00016008592792559113, + "loss": 0.3625, + "step": 12584 + }, + { + "epoch": 0.2950122189208671, + "grad_norm": 2.9569928646087646, + "learning_rate": 0.00016008003958882286, + "loss": 0.2694, + "step": 12585 + }, + { + "epoch": 0.2950356604956721, + "grad_norm": 1.1264874935150146, + "learning_rate": 0.00016007415092605864, + "loss": 0.0943, + "step": 12586 + }, + { + "epoch": 0.2950591020704771, + "grad_norm": 3.8021457195281982, + "learning_rate": 0.0001600682619373304, + "loss": 0.5697, + "step": 12587 + }, + { + "epoch": 0.2950825436452821, + "grad_norm": 4.2089409828186035, + "learning_rate": 0.00016006237262267004, + "loss": 0.5483, + "step": 12588 + }, + { + "epoch": 0.2951059852200871, + "grad_norm": 1.955635905265808, + "learning_rate": 0.00016005648298210957, + "loss": 0.1854, + "step": 12589 + }, + { + "epoch": 0.29512942679489207, + "grad_norm": 6.123908996582031, + "learning_rate": 0.00016005059301568095, + "loss": 0.9377, + "step": 12590 + }, + { + "epoch": 0.29515286836969706, + "grad_norm": 1.1277505159378052, + "learning_rate": 0.0001600447027234161, + "loss": 0.1183, + "step": 12591 + }, + { + "epoch": 0.29517630994450206, + "grad_norm": 1.3411849737167358, + "learning_rate": 0.00016003881210534704, + "loss": 0.1875, + "step": 12592 + }, + { + "epoch": 0.29519975151930705, + "grad_norm": 0.8526687622070312, + "learning_rate": 0.00016003292116150566, + "loss": 0.0978, + "step": 12593 + }, + { + "epoch": 0.29522319309411205, + "grad_norm": 8.315047264099121, + "learning_rate": 0.00016002702989192398, + "loss": 0.3047, + "step": 12594 + }, + { + "epoch": 0.29524663466891704, + "grad_norm": 1.3840309381484985, + "learning_rate": 0.00016002113829663396, + "loss": 0.2482, + "step": 12595 + }, + { + "epoch": 0.29527007624372203, + "grad_norm": 1.2027446031570435, + "learning_rate": 0.00016001524637566754, + "loss": 0.1995, + "step": 12596 + }, + { + "epoch": 0.29529351781852703, + "grad_norm": 3.1408050060272217, + "learning_rate": 0.00016000935412905672, + "loss": 0.3534, + "step": 12597 + }, + { + "epoch": 0.295316959393332, + "grad_norm": 3.9760868549346924, + "learning_rate": 0.00016000346155683342, + "loss": 0.367, + "step": 12598 + }, + { + "epoch": 0.295340400968137, + "grad_norm": 3.980142593383789, + "learning_rate": 0.0001599975686590297, + "loss": 0.4732, + "step": 12599 + }, + { + "epoch": 0.295363842542942, + "grad_norm": 2.9999313354492188, + "learning_rate": 0.00015999167543567745, + "loss": 0.4491, + "step": 12600 + }, + { + "epoch": 0.295387284117747, + "grad_norm": 1.0760698318481445, + "learning_rate": 0.0001599857818868087, + "loss": 0.0989, + "step": 12601 + }, + { + "epoch": 0.295410725692552, + "grad_norm": 0.4763219356536865, + "learning_rate": 0.0001599798880124554, + "loss": 0.0673, + "step": 12602 + }, + { + "epoch": 0.295434167267357, + "grad_norm": 0.8100576996803284, + "learning_rate": 0.00015997399381264952, + "loss": 0.0885, + "step": 12603 + }, + { + "epoch": 0.295457608842162, + "grad_norm": 0.4774542450904846, + "learning_rate": 0.00015996809928742312, + "loss": 0.0347, + "step": 12604 + }, + { + "epoch": 0.29548105041696704, + "grad_norm": 2.2430503368377686, + "learning_rate": 0.0001599622044368081, + "loss": 0.1707, + "step": 12605 + }, + { + "epoch": 0.29550449199177203, + "grad_norm": 3.1580629348754883, + "learning_rate": 0.00015995630926083647, + "loss": 0.3997, + "step": 12606 + }, + { + "epoch": 0.29552793356657703, + "grad_norm": 7.443869590759277, + "learning_rate": 0.0001599504137595402, + "loss": 0.682, + "step": 12607 + }, + { + "epoch": 0.295551375141382, + "grad_norm": 2.489003896713257, + "learning_rate": 0.00015994451793295133, + "loss": 0.2889, + "step": 12608 + }, + { + "epoch": 0.295574816716187, + "grad_norm": 4.543376922607422, + "learning_rate": 0.00015993862178110182, + "loss": 0.5225, + "step": 12609 + }, + { + "epoch": 0.295598258290992, + "grad_norm": 1.872839331626892, + "learning_rate": 0.00015993272530402365, + "loss": 0.1916, + "step": 12610 + }, + { + "epoch": 0.295621699865797, + "grad_norm": 3.4306368827819824, + "learning_rate": 0.00015992682850174883, + "loss": 0.373, + "step": 12611 + }, + { + "epoch": 0.295645141440602, + "grad_norm": 2.091989755630493, + "learning_rate": 0.00015992093137430936, + "loss": 0.1897, + "step": 12612 + }, + { + "epoch": 0.295668583015407, + "grad_norm": 2.771418333053589, + "learning_rate": 0.00015991503392173722, + "loss": 0.3069, + "step": 12613 + }, + { + "epoch": 0.295692024590212, + "grad_norm": 4.368398189544678, + "learning_rate": 0.0001599091361440644, + "loss": 0.539, + "step": 12614 + }, + { + "epoch": 0.295715466165017, + "grad_norm": 3.6910600662231445, + "learning_rate": 0.00015990323804132296, + "loss": 0.4634, + "step": 12615 + }, + { + "epoch": 0.295738907739822, + "grad_norm": 2.8223764896392822, + "learning_rate": 0.00015989733961354485, + "loss": 1.7433, + "step": 12616 + }, + { + "epoch": 0.29576234931462697, + "grad_norm": 1.557157039642334, + "learning_rate": 0.0001598914408607621, + "loss": 0.1615, + "step": 12617 + }, + { + "epoch": 0.29578579088943197, + "grad_norm": 0.8278332352638245, + "learning_rate": 0.0001598855417830067, + "loss": 0.1337, + "step": 12618 + }, + { + "epoch": 0.29580923246423696, + "grad_norm": 3.4563589096069336, + "learning_rate": 0.00015987964238031068, + "loss": 0.2147, + "step": 12619 + }, + { + "epoch": 0.29583267403904195, + "grad_norm": 3.723952531814575, + "learning_rate": 0.00015987374265270603, + "loss": 0.4406, + "step": 12620 + }, + { + "epoch": 0.29585611561384695, + "grad_norm": 1.9693092107772827, + "learning_rate": 0.00015986784260022473, + "loss": 0.1212, + "step": 12621 + }, + { + "epoch": 0.29587955718865194, + "grad_norm": 3.645474433898926, + "learning_rate": 0.0001598619422228989, + "loss": 0.3953, + "step": 12622 + }, + { + "epoch": 0.29590299876345694, + "grad_norm": 1.2371652126312256, + "learning_rate": 0.00015985604152076046, + "loss": 0.1327, + "step": 12623 + }, + { + "epoch": 0.29592644033826193, + "grad_norm": 3.8121767044067383, + "learning_rate": 0.00015985014049384145, + "loss": 0.223, + "step": 12624 + }, + { + "epoch": 0.2959498819130669, + "grad_norm": 3.0857510566711426, + "learning_rate": 0.00015984423914217391, + "loss": 0.3854, + "step": 12625 + }, + { + "epoch": 0.2959733234878719, + "grad_norm": 4.076742172241211, + "learning_rate": 0.00015983833746578983, + "loss": 0.5602, + "step": 12626 + }, + { + "epoch": 0.2959967650626769, + "grad_norm": 5.54632568359375, + "learning_rate": 0.00015983243546472122, + "loss": 0.6945, + "step": 12627 + }, + { + "epoch": 0.2960202066374819, + "grad_norm": 3.834141969680786, + "learning_rate": 0.0001598265331390002, + "loss": 0.4599, + "step": 12628 + }, + { + "epoch": 0.2960436482122869, + "grad_norm": 4.046360969543457, + "learning_rate": 0.00015982063048865868, + "loss": 0.5665, + "step": 12629 + }, + { + "epoch": 0.2960670897870919, + "grad_norm": 0.7834426760673523, + "learning_rate": 0.00015981472751372872, + "loss": 0.0796, + "step": 12630 + }, + { + "epoch": 0.2960905313618969, + "grad_norm": 3.826997995376587, + "learning_rate": 0.0001598088242142424, + "loss": 0.4015, + "step": 12631 + }, + { + "epoch": 0.2961139729367019, + "grad_norm": 2.818070650100708, + "learning_rate": 0.00015980292059023174, + "loss": 0.2286, + "step": 12632 + }, + { + "epoch": 0.2961374145115069, + "grad_norm": 0.9400811791419983, + "learning_rate": 0.0001597970166417287, + "loss": 0.0711, + "step": 12633 + }, + { + "epoch": 0.2961608560863119, + "grad_norm": 2.546660900115967, + "learning_rate": 0.00015979111236876542, + "loss": 1.4639, + "step": 12634 + }, + { + "epoch": 0.29618429766111687, + "grad_norm": 4.634129524230957, + "learning_rate": 0.00015978520777137385, + "loss": 0.6292, + "step": 12635 + }, + { + "epoch": 0.29620773923592186, + "grad_norm": 1.637756109237671, + "learning_rate": 0.00015977930284958607, + "loss": 1.0705, + "step": 12636 + }, + { + "epoch": 0.29623118081072686, + "grad_norm": 4.681230545043945, + "learning_rate": 0.0001597733976034341, + "loss": 0.5054, + "step": 12637 + }, + { + "epoch": 0.29625462238553185, + "grad_norm": 4.34780216217041, + "learning_rate": 0.00015976749203295, + "loss": 0.53, + "step": 12638 + }, + { + "epoch": 0.29627806396033685, + "grad_norm": 3.8737409114837646, + "learning_rate": 0.0001597615861381658, + "loss": 0.7194, + "step": 12639 + }, + { + "epoch": 0.29630150553514184, + "grad_norm": 3.5869171619415283, + "learning_rate": 0.00015975567991911358, + "loss": 1.9678, + "step": 12640 + }, + { + "epoch": 0.29632494710994683, + "grad_norm": 2.274397611618042, + "learning_rate": 0.00015974977337582534, + "loss": 0.0993, + "step": 12641 + }, + { + "epoch": 0.29634838868475183, + "grad_norm": 3.083543539047241, + "learning_rate": 0.00015974386650833316, + "loss": 0.2792, + "step": 12642 + }, + { + "epoch": 0.2963718302595568, + "grad_norm": 2.9276790618896484, + "learning_rate": 0.00015973795931666906, + "loss": 0.4016, + "step": 12643 + }, + { + "epoch": 0.2963952718343618, + "grad_norm": 4.25276517868042, + "learning_rate": 0.00015973205180086513, + "loss": 0.6322, + "step": 12644 + }, + { + "epoch": 0.2964187134091668, + "grad_norm": 0.8940116167068481, + "learning_rate": 0.0001597261439609534, + "loss": 0.0982, + "step": 12645 + }, + { + "epoch": 0.2964421549839718, + "grad_norm": 3.7816219329833984, + "learning_rate": 0.00015972023579696596, + "loss": 0.5377, + "step": 12646 + }, + { + "epoch": 0.2964655965587768, + "grad_norm": 3.754666805267334, + "learning_rate": 0.00015971432730893482, + "loss": 0.5386, + "step": 12647 + }, + { + "epoch": 0.2964890381335818, + "grad_norm": 3.4494667053222656, + "learning_rate": 0.00015970841849689203, + "loss": 0.1918, + "step": 12648 + }, + { + "epoch": 0.2965124797083868, + "grad_norm": 3.1876704692840576, + "learning_rate": 0.00015970250936086974, + "loss": 0.4296, + "step": 12649 + }, + { + "epoch": 0.2965359212831918, + "grad_norm": 0.856220006942749, + "learning_rate": 0.00015969659990089992, + "loss": 0.0652, + "step": 12650 + }, + { + "epoch": 0.2965593628579968, + "grad_norm": 3.6782093048095703, + "learning_rate": 0.00015969069011701468, + "loss": 0.3285, + "step": 12651 + }, + { + "epoch": 0.29658280443280177, + "grad_norm": 4.322514057159424, + "learning_rate": 0.00015968478000924607, + "loss": 0.4186, + "step": 12652 + }, + { + "epoch": 0.29660624600760677, + "grad_norm": 0.29998090863227844, + "learning_rate": 0.0001596788695776262, + "loss": 0.0397, + "step": 12653 + }, + { + "epoch": 0.29662968758241176, + "grad_norm": 3.2042107582092285, + "learning_rate": 0.00015967295882218706, + "loss": 0.2857, + "step": 12654 + }, + { + "epoch": 0.29665312915721676, + "grad_norm": 4.9945831298828125, + "learning_rate": 0.0001596670477429608, + "loss": 0.5678, + "step": 12655 + }, + { + "epoch": 0.2966765707320218, + "grad_norm": 0.6862280368804932, + "learning_rate": 0.00015966113633997947, + "loss": 0.0677, + "step": 12656 + }, + { + "epoch": 0.2967000123068268, + "grad_norm": 3.7272472381591797, + "learning_rate": 0.00015965522461327506, + "loss": 0.6701, + "step": 12657 + }, + { + "epoch": 0.2967234538816318, + "grad_norm": 2.2275338172912598, + "learning_rate": 0.00015964931256287978, + "loss": 0.2862, + "step": 12658 + }, + { + "epoch": 0.2967468954564368, + "grad_norm": 2.640104055404663, + "learning_rate": 0.00015964340018882566, + "loss": 0.3512, + "step": 12659 + }, + { + "epoch": 0.2967703370312418, + "grad_norm": 4.144100666046143, + "learning_rate": 0.00015963748749114477, + "loss": 0.3705, + "step": 12660 + }, + { + "epoch": 0.2967937786060468, + "grad_norm": 0.6782616972923279, + "learning_rate": 0.0001596315744698692, + "loss": 0.0452, + "step": 12661 + }, + { + "epoch": 0.29681722018085177, + "grad_norm": 1.294809103012085, + "learning_rate": 0.000159625661125031, + "loss": 0.1262, + "step": 12662 + }, + { + "epoch": 0.29684066175565677, + "grad_norm": 2.6273036003112793, + "learning_rate": 0.0001596197474566623, + "loss": 1.8659, + "step": 12663 + }, + { + "epoch": 0.29686410333046176, + "grad_norm": 0.3988010883331299, + "learning_rate": 0.0001596138334647952, + "loss": 0.0353, + "step": 12664 + }, + { + "epoch": 0.29688754490526675, + "grad_norm": 3.862962007522583, + "learning_rate": 0.0001596079191494617, + "loss": 0.3753, + "step": 12665 + }, + { + "epoch": 0.29691098648007175, + "grad_norm": 3.3660805225372314, + "learning_rate": 0.00015960200451069402, + "loss": 0.3157, + "step": 12666 + }, + { + "epoch": 0.29693442805487674, + "grad_norm": 1.853967308998108, + "learning_rate": 0.0001595960895485242, + "loss": 0.1268, + "step": 12667 + }, + { + "epoch": 0.29695786962968174, + "grad_norm": 3.266422748565674, + "learning_rate": 0.00015959017426298426, + "loss": 0.2805, + "step": 12668 + }, + { + "epoch": 0.29698131120448673, + "grad_norm": 0.8176315426826477, + "learning_rate": 0.0001595842586541064, + "loss": 0.0714, + "step": 12669 + }, + { + "epoch": 0.2970047527792917, + "grad_norm": 2.9986572265625, + "learning_rate": 0.00015957834272192266, + "loss": 0.4221, + "step": 12670 + }, + { + "epoch": 0.2970281943540967, + "grad_norm": 1.273358941078186, + "learning_rate": 0.00015957242646646517, + "loss": 0.1339, + "step": 12671 + }, + { + "epoch": 0.2970516359289017, + "grad_norm": 2.244235038757324, + "learning_rate": 0.000159566509887766, + "loss": 0.2207, + "step": 12672 + }, + { + "epoch": 0.2970750775037067, + "grad_norm": 1.8974815607070923, + "learning_rate": 0.00015956059298585727, + "loss": 1.1761, + "step": 12673 + }, + { + "epoch": 0.2970985190785117, + "grad_norm": 2.088444709777832, + "learning_rate": 0.0001595546757607711, + "loss": 0.1131, + "step": 12674 + }, + { + "epoch": 0.2971219606533167, + "grad_norm": 0.5390763282775879, + "learning_rate": 0.0001595487582125396, + "loss": 0.0502, + "step": 12675 + }, + { + "epoch": 0.2971454022281217, + "grad_norm": 0.9115976095199585, + "learning_rate": 0.00015954284034119483, + "loss": 0.1283, + "step": 12676 + }, + { + "epoch": 0.2971688438029267, + "grad_norm": 3.3607113361358643, + "learning_rate": 0.00015953692214676896, + "loss": 0.4731, + "step": 12677 + }, + { + "epoch": 0.2971922853777317, + "grad_norm": 2.8338167667388916, + "learning_rate": 0.00015953100362929408, + "loss": 0.3016, + "step": 12678 + }, + { + "epoch": 0.2972157269525367, + "grad_norm": 2.3275723457336426, + "learning_rate": 0.00015952508478880227, + "loss": 0.2513, + "step": 12679 + }, + { + "epoch": 0.29723916852734167, + "grad_norm": 1.4207178354263306, + "learning_rate": 0.0001595191656253257, + "loss": 0.7814, + "step": 12680 + }, + { + "epoch": 0.29726261010214666, + "grad_norm": 2.8204925060272217, + "learning_rate": 0.0001595132461388965, + "loss": 0.2003, + "step": 12681 + }, + { + "epoch": 0.29728605167695166, + "grad_norm": 3.6989526748657227, + "learning_rate": 0.00015950732632954667, + "loss": 0.3984, + "step": 12682 + }, + { + "epoch": 0.29730949325175665, + "grad_norm": 0.2590528726577759, + "learning_rate": 0.00015950140619730847, + "loss": 0.0122, + "step": 12683 + }, + { + "epoch": 0.29733293482656165, + "grad_norm": 2.6526546478271484, + "learning_rate": 0.00015949548574221395, + "loss": 0.2848, + "step": 12684 + }, + { + "epoch": 0.29735637640136664, + "grad_norm": 4.558508396148682, + "learning_rate": 0.00015948956496429525, + "loss": 0.2511, + "step": 12685 + }, + { + "epoch": 0.29737981797617163, + "grad_norm": 3.82989501953125, + "learning_rate": 0.00015948364386358452, + "loss": 0.4231, + "step": 12686 + }, + { + "epoch": 0.29740325955097663, + "grad_norm": 3.0525383949279785, + "learning_rate": 0.00015947772244011385, + "loss": 0.341, + "step": 12687 + }, + { + "epoch": 0.2974267011257816, + "grad_norm": 3.93638014793396, + "learning_rate": 0.00015947180069391536, + "loss": 0.5306, + "step": 12688 + }, + { + "epoch": 0.2974501427005866, + "grad_norm": 1.4987781047821045, + "learning_rate": 0.00015946587862502126, + "loss": 0.205, + "step": 12689 + }, + { + "epoch": 0.2974735842753916, + "grad_norm": 3.40502667427063, + "learning_rate": 0.0001594599562334636, + "loss": 0.2228, + "step": 12690 + }, + { + "epoch": 0.2974970258501966, + "grad_norm": 0.6174047589302063, + "learning_rate": 0.00015945403351927455, + "loss": 0.0427, + "step": 12691 + }, + { + "epoch": 0.2975204674250016, + "grad_norm": 4.117316722869873, + "learning_rate": 0.00015944811048248622, + "loss": 0.465, + "step": 12692 + }, + { + "epoch": 0.2975439089998066, + "grad_norm": 3.296656370162964, + "learning_rate": 0.0001594421871231308, + "loss": 0.5247, + "step": 12693 + }, + { + "epoch": 0.2975673505746116, + "grad_norm": 1.6436244249343872, + "learning_rate": 0.0001594362634412404, + "loss": 0.2209, + "step": 12694 + }, + { + "epoch": 0.2975907921494166, + "grad_norm": 3.9000630378723145, + "learning_rate": 0.00015943033943684713, + "loss": 0.5258, + "step": 12695 + }, + { + "epoch": 0.2976142337242216, + "grad_norm": 4.245753765106201, + "learning_rate": 0.0001594244151099832, + "loss": 0.4562, + "step": 12696 + }, + { + "epoch": 0.29763767529902657, + "grad_norm": 1.7077809572219849, + "learning_rate": 0.00015941849046068073, + "loss": 0.2137, + "step": 12697 + }, + { + "epoch": 0.29766111687383157, + "grad_norm": 2.567509889602661, + "learning_rate": 0.00015941256548897186, + "loss": 0.1397, + "step": 12698 + }, + { + "epoch": 0.29768455844863656, + "grad_norm": 2.801459789276123, + "learning_rate": 0.00015940664019488873, + "loss": 0.1882, + "step": 12699 + }, + { + "epoch": 0.29770800002344155, + "grad_norm": 3.1217024326324463, + "learning_rate": 0.00015940071457846346, + "loss": 0.5372, + "step": 12700 + }, + { + "epoch": 0.29773144159824655, + "grad_norm": 2.9595468044281006, + "learning_rate": 0.0001593947886397283, + "loss": 0.3989, + "step": 12701 + }, + { + "epoch": 0.29775488317305154, + "grad_norm": 2.1720316410064697, + "learning_rate": 0.00015938886237871534, + "loss": 1.7488, + "step": 12702 + }, + { + "epoch": 0.29777832474785654, + "grad_norm": 1.4459384679794312, + "learning_rate": 0.0001593829357954567, + "loss": 0.2194, + "step": 12703 + }, + { + "epoch": 0.29780176632266153, + "grad_norm": 2.916857957839966, + "learning_rate": 0.0001593770088899846, + "loss": 1.2404, + "step": 12704 + }, + { + "epoch": 0.2978252078974665, + "grad_norm": 1.1629467010498047, + "learning_rate": 0.00015937108166233122, + "loss": 0.1412, + "step": 12705 + }, + { + "epoch": 0.2978486494722715, + "grad_norm": 1.1435093879699707, + "learning_rate": 0.00015936515411252864, + "loss": 0.1181, + "step": 12706 + }, + { + "epoch": 0.2978720910470765, + "grad_norm": 3.6757543087005615, + "learning_rate": 0.0001593592262406091, + "loss": 0.3637, + "step": 12707 + }, + { + "epoch": 0.29789553262188156, + "grad_norm": 6.252946853637695, + "learning_rate": 0.0001593532980466047, + "loss": 0.5437, + "step": 12708 + }, + { + "epoch": 0.29791897419668656, + "grad_norm": 2.9695088863372803, + "learning_rate": 0.0001593473695305476, + "loss": 0.3112, + "step": 12709 + }, + { + "epoch": 0.29794241577149155, + "grad_norm": 0.885482668876648, + "learning_rate": 0.00015934144069247006, + "loss": 0.0573, + "step": 12710 + }, + { + "epoch": 0.29796585734629655, + "grad_norm": 2.5131912231445312, + "learning_rate": 0.00015933551153240417, + "loss": 0.1388, + "step": 12711 + }, + { + "epoch": 0.29798929892110154, + "grad_norm": 8.905912399291992, + "learning_rate": 0.00015932958205038214, + "loss": 0.7809, + "step": 12712 + }, + { + "epoch": 0.29801274049590654, + "grad_norm": 3.5786831378936768, + "learning_rate": 0.00015932365224643612, + "loss": 0.3141, + "step": 12713 + }, + { + "epoch": 0.29803618207071153, + "grad_norm": 0.6970001459121704, + "learning_rate": 0.0001593177221205983, + "loss": 0.0997, + "step": 12714 + }, + { + "epoch": 0.2980596236455165, + "grad_norm": 1.519110083580017, + "learning_rate": 0.00015931179167290088, + "loss": 0.2287, + "step": 12715 + }, + { + "epoch": 0.2980830652203215, + "grad_norm": 3.572620153427124, + "learning_rate": 0.00015930586090337595, + "loss": 0.4494, + "step": 12716 + }, + { + "epoch": 0.2981065067951265, + "grad_norm": 3.0755393505096436, + "learning_rate": 0.00015929992981205578, + "loss": 0.2142, + "step": 12717 + }, + { + "epoch": 0.2981299483699315, + "grad_norm": 6.133994102478027, + "learning_rate": 0.0001592939983989725, + "loss": 0.3919, + "step": 12718 + }, + { + "epoch": 0.2981533899447365, + "grad_norm": 2.870511054992676, + "learning_rate": 0.00015928806666415838, + "loss": 1.7467, + "step": 12719 + }, + { + "epoch": 0.2981768315195415, + "grad_norm": 4.734531879425049, + "learning_rate": 0.0001592821346076455, + "loss": 0.6936, + "step": 12720 + }, + { + "epoch": 0.2982002730943465, + "grad_norm": 4.26240348815918, + "learning_rate": 0.00015927620222946608, + "loss": 0.492, + "step": 12721 + }, + { + "epoch": 0.2982237146691515, + "grad_norm": 1.097465991973877, + "learning_rate": 0.00015927026952965235, + "loss": 0.1505, + "step": 12722 + }, + { + "epoch": 0.2982471562439565, + "grad_norm": 4.964351654052734, + "learning_rate": 0.00015926433650823648, + "loss": 0.4028, + "step": 12723 + }, + { + "epoch": 0.2982705978187615, + "grad_norm": 3.2444279193878174, + "learning_rate": 0.0001592584031652506, + "loss": 0.2778, + "step": 12724 + }, + { + "epoch": 0.29829403939356647, + "grad_norm": 3.7883505821228027, + "learning_rate": 0.000159252469500727, + "loss": 0.5816, + "step": 12725 + }, + { + "epoch": 0.29831748096837146, + "grad_norm": 3.98710298538208, + "learning_rate": 0.00015924653551469784, + "loss": 0.4382, + "step": 12726 + }, + { + "epoch": 0.29834092254317646, + "grad_norm": 4.651364803314209, + "learning_rate": 0.00015924060120719527, + "loss": 0.2797, + "step": 12727 + }, + { + "epoch": 0.29836436411798145, + "grad_norm": 3.2234113216400146, + "learning_rate": 0.0001592346665782516, + "loss": 0.3495, + "step": 12728 + }, + { + "epoch": 0.29838780569278645, + "grad_norm": 1.1149822473526, + "learning_rate": 0.00015922873162789892, + "loss": 0.1568, + "step": 12729 + }, + { + "epoch": 0.29841124726759144, + "grad_norm": 4.402176380157471, + "learning_rate": 0.00015922279635616945, + "loss": 0.5391, + "step": 12730 + }, + { + "epoch": 0.29843468884239643, + "grad_norm": 3.85465669631958, + "learning_rate": 0.00015921686076309546, + "loss": 0.5253, + "step": 12731 + }, + { + "epoch": 0.29845813041720143, + "grad_norm": 2.2382287979125977, + "learning_rate": 0.0001592109248487091, + "loss": 0.2958, + "step": 12732 + }, + { + "epoch": 0.2984815719920064, + "grad_norm": 1.5134798288345337, + "learning_rate": 0.00015920498861304261, + "loss": 0.2575, + "step": 12733 + }, + { + "epoch": 0.2985050135668114, + "grad_norm": 5.130469799041748, + "learning_rate": 0.0001591990520561282, + "loss": 0.4941, + "step": 12734 + }, + { + "epoch": 0.2985284551416164, + "grad_norm": 4.022972106933594, + "learning_rate": 0.00015919311517799804, + "loss": 0.4599, + "step": 12735 + }, + { + "epoch": 0.2985518967164214, + "grad_norm": 2.9328348636627197, + "learning_rate": 0.0001591871779786844, + "loss": 0.3849, + "step": 12736 + }, + { + "epoch": 0.2985753382912264, + "grad_norm": 4.706015586853027, + "learning_rate": 0.00015918124045821944, + "loss": 0.5437, + "step": 12737 + }, + { + "epoch": 0.2985987798660314, + "grad_norm": 3.0521090030670166, + "learning_rate": 0.00015917530261663542, + "loss": 0.2799, + "step": 12738 + }, + { + "epoch": 0.2986222214408364, + "grad_norm": 2.764064311981201, + "learning_rate": 0.00015916936445396456, + "loss": 0.3302, + "step": 12739 + }, + { + "epoch": 0.2986456630156414, + "grad_norm": 3.4114937782287598, + "learning_rate": 0.00015916342597023904, + "loss": 0.3179, + "step": 12740 + }, + { + "epoch": 0.2986691045904464, + "grad_norm": 0.5022053122520447, + "learning_rate": 0.00015915748716549113, + "loss": 0.0598, + "step": 12741 + }, + { + "epoch": 0.29869254616525137, + "grad_norm": 4.61980676651001, + "learning_rate": 0.00015915154803975302, + "loss": 0.3492, + "step": 12742 + }, + { + "epoch": 0.29871598774005637, + "grad_norm": 2.5399606227874756, + "learning_rate": 0.0001591456085930569, + "loss": 0.2856, + "step": 12743 + }, + { + "epoch": 0.29873942931486136, + "grad_norm": 2.211947441101074, + "learning_rate": 0.0001591396688254351, + "loss": 0.273, + "step": 12744 + }, + { + "epoch": 0.29876287088966635, + "grad_norm": 0.49707838892936707, + "learning_rate": 0.00015913372873691982, + "loss": 0.0504, + "step": 12745 + }, + { + "epoch": 0.29878631246447135, + "grad_norm": 2.6215081214904785, + "learning_rate": 0.0001591277883275432, + "loss": 1.9034, + "step": 12746 + }, + { + "epoch": 0.29880975403927634, + "grad_norm": 5.376967430114746, + "learning_rate": 0.00015912184759733758, + "loss": 0.6158, + "step": 12747 + }, + { + "epoch": 0.29883319561408134, + "grad_norm": 0.6662097573280334, + "learning_rate": 0.00015911590654633513, + "loss": 0.0932, + "step": 12748 + }, + { + "epoch": 0.29885663718888633, + "grad_norm": 4.9397807121276855, + "learning_rate": 0.00015910996517456811, + "loss": 0.5743, + "step": 12749 + }, + { + "epoch": 0.2988800787636913, + "grad_norm": 2.3245832920074463, + "learning_rate": 0.00015910402348206877, + "loss": 0.2193, + "step": 12750 + }, + { + "epoch": 0.2989035203384963, + "grad_norm": 1.6666696071624756, + "learning_rate": 0.00015909808146886934, + "loss": 0.2083, + "step": 12751 + }, + { + "epoch": 0.2989269619133013, + "grad_norm": 2.1977295875549316, + "learning_rate": 0.00015909213913500202, + "loss": 0.1428, + "step": 12752 + }, + { + "epoch": 0.2989504034881063, + "grad_norm": 3.317932605743408, + "learning_rate": 0.00015908619648049915, + "loss": 0.2958, + "step": 12753 + }, + { + "epoch": 0.2989738450629113, + "grad_norm": 3.803799629211426, + "learning_rate": 0.0001590802535053929, + "loss": 0.5189, + "step": 12754 + }, + { + "epoch": 0.2989972866377163, + "grad_norm": 0.918038010597229, + "learning_rate": 0.00015907431020971552, + "loss": 0.0764, + "step": 12755 + }, + { + "epoch": 0.2990207282125213, + "grad_norm": 0.3271523118019104, + "learning_rate": 0.00015906836659349924, + "loss": 0.0383, + "step": 12756 + }, + { + "epoch": 0.2990441697873263, + "grad_norm": 2.9910850524902344, + "learning_rate": 0.00015906242265677637, + "loss": 0.2517, + "step": 12757 + }, + { + "epoch": 0.2990676113621313, + "grad_norm": 1.9720476865768433, + "learning_rate": 0.00015905647839957915, + "loss": 0.2941, + "step": 12758 + }, + { + "epoch": 0.29909105293693633, + "grad_norm": 0.8423129320144653, + "learning_rate": 0.00015905053382193982, + "loss": 0.0676, + "step": 12759 + }, + { + "epoch": 0.2991144945117413, + "grad_norm": 4.600614070892334, + "learning_rate": 0.00015904458892389062, + "loss": 0.7887, + "step": 12760 + }, + { + "epoch": 0.2991379360865463, + "grad_norm": 1.5609660148620605, + "learning_rate": 0.0001590386437054638, + "loss": 0.961, + "step": 12761 + }, + { + "epoch": 0.2991613776613513, + "grad_norm": 4.293834209442139, + "learning_rate": 0.00015903269816669167, + "loss": 0.4756, + "step": 12762 + }, + { + "epoch": 0.2991848192361563, + "grad_norm": 4.5373454093933105, + "learning_rate": 0.00015902675230760645, + "loss": 0.3177, + "step": 12763 + }, + { + "epoch": 0.2992082608109613, + "grad_norm": 4.039884567260742, + "learning_rate": 0.00015902080612824042, + "loss": 0.4336, + "step": 12764 + }, + { + "epoch": 0.2992317023857663, + "grad_norm": 3.1260108947753906, + "learning_rate": 0.00015901485962862584, + "loss": 0.518, + "step": 12765 + }, + { + "epoch": 0.2992551439605713, + "grad_norm": 2.9918196201324463, + "learning_rate": 0.00015900891280879495, + "loss": 0.3995, + "step": 12766 + }, + { + "epoch": 0.2992785855353763, + "grad_norm": 5.398778438568115, + "learning_rate": 0.00015900296566878, + "loss": 0.6081, + "step": 12767 + }, + { + "epoch": 0.2993020271101813, + "grad_norm": 2.300884485244751, + "learning_rate": 0.00015899701820861338, + "loss": 0.1732, + "step": 12768 + }, + { + "epoch": 0.2993254686849863, + "grad_norm": 3.299259901046753, + "learning_rate": 0.00015899107042832724, + "loss": 0.4609, + "step": 12769 + }, + { + "epoch": 0.29934891025979127, + "grad_norm": 0.6912385821342468, + "learning_rate": 0.00015898512232795392, + "loss": 0.0571, + "step": 12770 + }, + { + "epoch": 0.29937235183459626, + "grad_norm": 1.7312391996383667, + "learning_rate": 0.00015897917390752565, + "loss": 0.1758, + "step": 12771 + }, + { + "epoch": 0.29939579340940126, + "grad_norm": 1.2669200897216797, + "learning_rate": 0.0001589732251670747, + "loss": 0.728, + "step": 12772 + }, + { + "epoch": 0.29941923498420625, + "grad_norm": 4.589970588684082, + "learning_rate": 0.00015896727610663342, + "loss": 0.549, + "step": 12773 + }, + { + "epoch": 0.29944267655901124, + "grad_norm": 0.45347902178764343, + "learning_rate": 0.00015896132672623404, + "loss": 0.0423, + "step": 12774 + }, + { + "epoch": 0.29946611813381624, + "grad_norm": 2.7670302391052246, + "learning_rate": 0.0001589553770259088, + "loss": 1.765, + "step": 12775 + }, + { + "epoch": 0.29948955970862123, + "grad_norm": 1.9739664793014526, + "learning_rate": 0.00015894942700569006, + "loss": 0.2225, + "step": 12776 + }, + { + "epoch": 0.29951300128342623, + "grad_norm": 0.8119016289710999, + "learning_rate": 0.00015894347666561004, + "loss": 0.0847, + "step": 12777 + }, + { + "epoch": 0.2995364428582312, + "grad_norm": 2.2560172080993652, + "learning_rate": 0.0001589375260057011, + "loss": 0.2685, + "step": 12778 + }, + { + "epoch": 0.2995598844330362, + "grad_norm": 0.39415618777275085, + "learning_rate": 0.00015893157502599542, + "loss": 0.0611, + "step": 12779 + }, + { + "epoch": 0.2995833260078412, + "grad_norm": 4.282174110412598, + "learning_rate": 0.00015892562372652543, + "loss": 0.479, + "step": 12780 + }, + { + "epoch": 0.2996067675826462, + "grad_norm": 3.1683976650238037, + "learning_rate": 0.00015891967210732332, + "loss": 1.1828, + "step": 12781 + }, + { + "epoch": 0.2996302091574512, + "grad_norm": 3.388253927230835, + "learning_rate": 0.0001589137201684214, + "loss": 1.4848, + "step": 12782 + }, + { + "epoch": 0.2996536507322562, + "grad_norm": 3.1664159297943115, + "learning_rate": 0.00015890776790985201, + "loss": 0.3803, + "step": 12783 + }, + { + "epoch": 0.2996770923070612, + "grad_norm": 0.7864231467247009, + "learning_rate": 0.0001589018153316474, + "loss": 0.0707, + "step": 12784 + }, + { + "epoch": 0.2997005338818662, + "grad_norm": 3.631267547607422, + "learning_rate": 0.00015889586243383988, + "loss": 0.4105, + "step": 12785 + }, + { + "epoch": 0.2997239754566712, + "grad_norm": 2.4784927368164062, + "learning_rate": 0.00015888990921646175, + "loss": 0.242, + "step": 12786 + }, + { + "epoch": 0.29974741703147617, + "grad_norm": 3.327369213104248, + "learning_rate": 0.0001588839556795453, + "loss": 0.4098, + "step": 12787 + }, + { + "epoch": 0.29977085860628117, + "grad_norm": 3.2525718212127686, + "learning_rate": 0.0001588780018231229, + "loss": 0.4606, + "step": 12788 + }, + { + "epoch": 0.29979430018108616, + "grad_norm": 2.8823201656341553, + "learning_rate": 0.00015887204764722677, + "loss": 0.331, + "step": 12789 + }, + { + "epoch": 0.29981774175589115, + "grad_norm": 3.744457960128784, + "learning_rate": 0.00015886609315188924, + "loss": 0.697, + "step": 12790 + }, + { + "epoch": 0.29984118333069615, + "grad_norm": 2.6874396800994873, + "learning_rate": 0.00015886013833714267, + "loss": 0.2144, + "step": 12791 + }, + { + "epoch": 0.29986462490550114, + "grad_norm": 1.1983872652053833, + "learning_rate": 0.0001588541832030193, + "loss": 0.095, + "step": 12792 + }, + { + "epoch": 0.29988806648030614, + "grad_norm": 3.0678505897521973, + "learning_rate": 0.0001588482277495515, + "loss": 0.1892, + "step": 12793 + }, + { + "epoch": 0.29991150805511113, + "grad_norm": 8.362386703491211, + "learning_rate": 0.00015884227197677158, + "loss": 0.6636, + "step": 12794 + }, + { + "epoch": 0.2999349496299161, + "grad_norm": 3.902515172958374, + "learning_rate": 0.0001588363158847118, + "loss": 0.2842, + "step": 12795 + }, + { + "epoch": 0.2999583912047211, + "grad_norm": 0.34036919474601746, + "learning_rate": 0.00015883035947340454, + "loss": 0.0355, + "step": 12796 + }, + { + "epoch": 0.2999818327795261, + "grad_norm": 1.7489573955535889, + "learning_rate": 0.00015882440274288205, + "loss": 0.1726, + "step": 12797 + }, + { + "epoch": 0.3000052743543311, + "grad_norm": 2.8553662300109863, + "learning_rate": 0.00015881844569317678, + "loss": 0.2035, + "step": 12798 + }, + { + "epoch": 0.3000287159291361, + "grad_norm": 4.697045803070068, + "learning_rate": 0.0001588124883243209, + "loss": 0.3039, + "step": 12799 + }, + { + "epoch": 0.3000521575039411, + "grad_norm": 3.9982616901397705, + "learning_rate": 0.00015880653063634682, + "loss": 0.6996, + "step": 12800 + }, + { + "epoch": 0.3000755990787461, + "grad_norm": 3.4229800701141357, + "learning_rate": 0.00015880057262928688, + "loss": 0.5204, + "step": 12801 + }, + { + "epoch": 0.3000990406535511, + "grad_norm": 1.6325725317001343, + "learning_rate": 0.00015879461430317338, + "loss": 0.3162, + "step": 12802 + }, + { + "epoch": 0.3001224822283561, + "grad_norm": 2.4449450969696045, + "learning_rate": 0.0001587886556580386, + "loss": 1.5522, + "step": 12803 + }, + { + "epoch": 0.3001459238031611, + "grad_norm": 3.5964548587799072, + "learning_rate": 0.00015878269669391493, + "loss": 0.4853, + "step": 12804 + }, + { + "epoch": 0.30016936537796607, + "grad_norm": 6.073265552520752, + "learning_rate": 0.00015877673741083475, + "loss": 0.8536, + "step": 12805 + }, + { + "epoch": 0.30019280695277106, + "grad_norm": 1.3161940574645996, + "learning_rate": 0.00015877077780883028, + "loss": 0.1614, + "step": 12806 + }, + { + "epoch": 0.30021624852757606, + "grad_norm": 3.6980395317077637, + "learning_rate": 0.00015876481788793396, + "loss": 0.46, + "step": 12807 + }, + { + "epoch": 0.30023969010238105, + "grad_norm": 2.2370049953460693, + "learning_rate": 0.0001587588576481781, + "loss": 0.2159, + "step": 12808 + }, + { + "epoch": 0.30026313167718605, + "grad_norm": 2.873857259750366, + "learning_rate": 0.00015875289708959498, + "loss": 1.6941, + "step": 12809 + }, + { + "epoch": 0.3002865732519911, + "grad_norm": 1.581242322921753, + "learning_rate": 0.00015874693621221696, + "loss": 0.18, + "step": 12810 + }, + { + "epoch": 0.3003100148267961, + "grad_norm": 4.908534049987793, + "learning_rate": 0.00015874097501607646, + "loss": 0.3587, + "step": 12811 + }, + { + "epoch": 0.3003334564016011, + "grad_norm": 3.3388521671295166, + "learning_rate": 0.00015873501350120579, + "loss": 0.3251, + "step": 12812 + }, + { + "epoch": 0.3003568979764061, + "grad_norm": 4.82692289352417, + "learning_rate": 0.00015872905166763727, + "loss": 0.6077, + "step": 12813 + }, + { + "epoch": 0.3003803395512111, + "grad_norm": 4.08208703994751, + "learning_rate": 0.00015872308951540327, + "loss": 0.5787, + "step": 12814 + }, + { + "epoch": 0.30040378112601607, + "grad_norm": 1.2428630590438843, + "learning_rate": 0.00015871712704453614, + "loss": 0.0947, + "step": 12815 + }, + { + "epoch": 0.30042722270082106, + "grad_norm": 4.933893203735352, + "learning_rate": 0.00015871116425506817, + "loss": 0.5406, + "step": 12816 + }, + { + "epoch": 0.30045066427562606, + "grad_norm": 3.881359815597534, + "learning_rate": 0.00015870520114703184, + "loss": 0.2751, + "step": 12817 + }, + { + "epoch": 0.30047410585043105, + "grad_norm": 2.8451414108276367, + "learning_rate": 0.0001586992377204594, + "loss": 0.443, + "step": 12818 + }, + { + "epoch": 0.30049754742523604, + "grad_norm": 1.4420242309570312, + "learning_rate": 0.0001586932739753833, + "loss": 0.0895, + "step": 12819 + }, + { + "epoch": 0.30052098900004104, + "grad_norm": 1.6243423223495483, + "learning_rate": 0.00015868730991183577, + "loss": 0.209, + "step": 12820 + }, + { + "epoch": 0.30054443057484603, + "grad_norm": 3.117821455001831, + "learning_rate": 0.0001586813455298493, + "loss": 0.2677, + "step": 12821 + }, + { + "epoch": 0.300567872149651, + "grad_norm": 0.5601708889007568, + "learning_rate": 0.00015867538082945618, + "loss": 0.0345, + "step": 12822 + }, + { + "epoch": 0.300591313724456, + "grad_norm": 1.1126530170440674, + "learning_rate": 0.0001586694158106888, + "loss": 0.1046, + "step": 12823 + }, + { + "epoch": 0.300614755299261, + "grad_norm": 1.5528078079223633, + "learning_rate": 0.00015866345047357952, + "loss": 0.1583, + "step": 12824 + }, + { + "epoch": 0.300638196874066, + "grad_norm": 2.917387008666992, + "learning_rate": 0.0001586574848181607, + "loss": 0.348, + "step": 12825 + }, + { + "epoch": 0.300661638448871, + "grad_norm": 5.728091716766357, + "learning_rate": 0.00015865151884446473, + "loss": 0.5139, + "step": 12826 + }, + { + "epoch": 0.300685080023676, + "grad_norm": 2.144482135772705, + "learning_rate": 0.00015864555255252396, + "loss": 1.958, + "step": 12827 + }, + { + "epoch": 0.300708521598481, + "grad_norm": 0.919154167175293, + "learning_rate": 0.0001586395859423708, + "loss": 0.1051, + "step": 12828 + }, + { + "epoch": 0.300731963173286, + "grad_norm": 4.983508586883545, + "learning_rate": 0.00015863361901403752, + "loss": 0.5548, + "step": 12829 + }, + { + "epoch": 0.300755404748091, + "grad_norm": 2.386291742324829, + "learning_rate": 0.00015862765176755664, + "loss": 0.2031, + "step": 12830 + }, + { + "epoch": 0.300778846322896, + "grad_norm": 2.240051507949829, + "learning_rate": 0.00015862168420296047, + "loss": 0.1231, + "step": 12831 + }, + { + "epoch": 0.30080228789770097, + "grad_norm": 3.4994723796844482, + "learning_rate": 0.00015861571632028138, + "loss": 0.3851, + "step": 12832 + }, + { + "epoch": 0.30082572947250596, + "grad_norm": 4.555578708648682, + "learning_rate": 0.00015860974811955178, + "loss": 0.4575, + "step": 12833 + }, + { + "epoch": 0.30084917104731096, + "grad_norm": 3.9771480560302734, + "learning_rate": 0.00015860377960080402, + "loss": 0.3564, + "step": 12834 + }, + { + "epoch": 0.30087261262211595, + "grad_norm": 2.0176849365234375, + "learning_rate": 0.00015859781076407053, + "loss": 1.3814, + "step": 12835 + }, + { + "epoch": 0.30089605419692095, + "grad_norm": 2.8618485927581787, + "learning_rate": 0.00015859184160938364, + "loss": 0.352, + "step": 12836 + }, + { + "epoch": 0.30091949577172594, + "grad_norm": 0.8914855718612671, + "learning_rate": 0.0001585858721367758, + "loss": 0.0974, + "step": 12837 + }, + { + "epoch": 0.30094293734653094, + "grad_norm": 1.245627999305725, + "learning_rate": 0.00015857990234627932, + "loss": 0.1013, + "step": 12838 + }, + { + "epoch": 0.30096637892133593, + "grad_norm": 3.766355514526367, + "learning_rate": 0.00015857393223792668, + "loss": 0.3744, + "step": 12839 + }, + { + "epoch": 0.3009898204961409, + "grad_norm": 4.462657451629639, + "learning_rate": 0.00015856796181175025, + "loss": 0.6091, + "step": 12840 + }, + { + "epoch": 0.3010132620709459, + "grad_norm": 3.0744128227233887, + "learning_rate": 0.00015856199106778238, + "loss": 0.5026, + "step": 12841 + }, + { + "epoch": 0.3010367036457509, + "grad_norm": 4.7720770835876465, + "learning_rate": 0.0001585560200060555, + "loss": 0.4286, + "step": 12842 + }, + { + "epoch": 0.3010601452205559, + "grad_norm": 1.9747908115386963, + "learning_rate": 0.00015855004862660204, + "loss": 1.2653, + "step": 12843 + }, + { + "epoch": 0.3010835867953609, + "grad_norm": 0.7772402167320251, + "learning_rate": 0.00015854407692945434, + "loss": 0.0698, + "step": 12844 + }, + { + "epoch": 0.3011070283701659, + "grad_norm": 1.8331561088562012, + "learning_rate": 0.00015853810491464488, + "loss": 0.2297, + "step": 12845 + }, + { + "epoch": 0.3011304699449709, + "grad_norm": 4.276230812072754, + "learning_rate": 0.00015853213258220595, + "loss": 0.567, + "step": 12846 + }, + { + "epoch": 0.3011539115197759, + "grad_norm": 0.3763076364994049, + "learning_rate": 0.00015852615993217005, + "loss": 0.0289, + "step": 12847 + }, + { + "epoch": 0.3011773530945809, + "grad_norm": 2.182229995727539, + "learning_rate": 0.00015852018696456958, + "loss": 1.7087, + "step": 12848 + }, + { + "epoch": 0.3012007946693859, + "grad_norm": 4.38364315032959, + "learning_rate": 0.00015851421367943692, + "loss": 0.4219, + "step": 12849 + }, + { + "epoch": 0.30122423624419087, + "grad_norm": 3.706148386001587, + "learning_rate": 0.00015850824007680449, + "loss": 0.3184, + "step": 12850 + }, + { + "epoch": 0.30124767781899586, + "grad_norm": 3.6383869647979736, + "learning_rate": 0.0001585022661567047, + "loss": 0.4094, + "step": 12851 + }, + { + "epoch": 0.30127111939380086, + "grad_norm": 3.188795804977417, + "learning_rate": 0.00015849629191916996, + "loss": 0.3277, + "step": 12852 + }, + { + "epoch": 0.30129456096860585, + "grad_norm": 5.1270880699157715, + "learning_rate": 0.0001584903173642327, + "loss": 0.4913, + "step": 12853 + }, + { + "epoch": 0.30131800254341085, + "grad_norm": 0.8686815500259399, + "learning_rate": 0.00015848434249192533, + "loss": 0.1001, + "step": 12854 + }, + { + "epoch": 0.30134144411821584, + "grad_norm": 0.7112259268760681, + "learning_rate": 0.00015847836730228032, + "loss": 0.1414, + "step": 12855 + }, + { + "epoch": 0.30136488569302083, + "grad_norm": 3.3572139739990234, + "learning_rate": 0.00015847239179533, + "loss": 0.3352, + "step": 12856 + }, + { + "epoch": 0.30138832726782583, + "grad_norm": 3.320774555206299, + "learning_rate": 0.00015846641597110685, + "loss": 0.3413, + "step": 12857 + }, + { + "epoch": 0.3014117688426308, + "grad_norm": 1.3581411838531494, + "learning_rate": 0.00015846043982964324, + "loss": 0.1955, + "step": 12858 + }, + { + "epoch": 0.3014352104174358, + "grad_norm": 3.13454008102417, + "learning_rate": 0.0001584544633709717, + "loss": 0.2726, + "step": 12859 + }, + { + "epoch": 0.3014586519922408, + "grad_norm": 3.7731080055236816, + "learning_rate": 0.00015844848659512457, + "loss": 0.4281, + "step": 12860 + }, + { + "epoch": 0.30148209356704586, + "grad_norm": 4.572572231292725, + "learning_rate": 0.00015844250950213432, + "loss": 0.565, + "step": 12861 + }, + { + "epoch": 0.30150553514185086, + "grad_norm": 0.5009912848472595, + "learning_rate": 0.00015843653209203338, + "loss": 0.0653, + "step": 12862 + }, + { + "epoch": 0.30152897671665585, + "grad_norm": 3.3770008087158203, + "learning_rate": 0.00015843055436485414, + "loss": 0.3867, + "step": 12863 + }, + { + "epoch": 0.30155241829146084, + "grad_norm": 2.7251837253570557, + "learning_rate": 0.00015842457632062908, + "loss": 0.413, + "step": 12864 + }, + { + "epoch": 0.30157585986626584, + "grad_norm": 3.4613828659057617, + "learning_rate": 0.00015841859795939065, + "loss": 0.2924, + "step": 12865 + }, + { + "epoch": 0.30159930144107083, + "grad_norm": 6.147182941436768, + "learning_rate": 0.00015841261928117124, + "loss": 0.4734, + "step": 12866 + }, + { + "epoch": 0.3016227430158758, + "grad_norm": 2.12894868850708, + "learning_rate": 0.00015840664028600335, + "loss": 0.8344, + "step": 12867 + }, + { + "epoch": 0.3016461845906808, + "grad_norm": 2.7460668087005615, + "learning_rate": 0.00015840066097391934, + "loss": 0.2869, + "step": 12868 + }, + { + "epoch": 0.3016696261654858, + "grad_norm": 1.1135413646697998, + "learning_rate": 0.00015839468134495172, + "loss": 0.1191, + "step": 12869 + }, + { + "epoch": 0.3016930677402908, + "grad_norm": 2.8568780422210693, + "learning_rate": 0.00015838870139913292, + "loss": 0.3233, + "step": 12870 + }, + { + "epoch": 0.3017165093150958, + "grad_norm": 3.654026508331299, + "learning_rate": 0.0001583827211364954, + "loss": 0.3913, + "step": 12871 + }, + { + "epoch": 0.3017399508899008, + "grad_norm": 4.465559005737305, + "learning_rate": 0.00015837674055707156, + "loss": 0.403, + "step": 12872 + }, + { + "epoch": 0.3017633924647058, + "grad_norm": 1.2765895128250122, + "learning_rate": 0.0001583707596608939, + "loss": 0.15, + "step": 12873 + }, + { + "epoch": 0.3017868340395108, + "grad_norm": 2.6535003185272217, + "learning_rate": 0.00015836477844799487, + "loss": 0.3129, + "step": 12874 + }, + { + "epoch": 0.3018102756143158, + "grad_norm": 0.731243908405304, + "learning_rate": 0.0001583587969184069, + "loss": 0.0905, + "step": 12875 + }, + { + "epoch": 0.3018337171891208, + "grad_norm": 0.47628286480903625, + "learning_rate": 0.00015835281507216244, + "loss": 0.0455, + "step": 12876 + }, + { + "epoch": 0.30185715876392577, + "grad_norm": 3.2463717460632324, + "learning_rate": 0.000158346832909294, + "loss": 0.1666, + "step": 12877 + }, + { + "epoch": 0.30188060033873076, + "grad_norm": 1.4661420583724976, + "learning_rate": 0.000158340850429834, + "loss": 0.0612, + "step": 12878 + }, + { + "epoch": 0.30190404191353576, + "grad_norm": 3.432722330093384, + "learning_rate": 0.00015833486763381486, + "loss": 0.3921, + "step": 12879 + }, + { + "epoch": 0.30192748348834075, + "grad_norm": 0.8718516230583191, + "learning_rate": 0.00015832888452126914, + "loss": 0.0843, + "step": 12880 + }, + { + "epoch": 0.30195092506314575, + "grad_norm": 6.110569953918457, + "learning_rate": 0.00015832290109222923, + "loss": 0.4985, + "step": 12881 + }, + { + "epoch": 0.30197436663795074, + "grad_norm": 4.935126781463623, + "learning_rate": 0.0001583169173467276, + "loss": 0.4677, + "step": 12882 + }, + { + "epoch": 0.30199780821275574, + "grad_norm": 1.3095883131027222, + "learning_rate": 0.00015831093328479677, + "loss": 0.1404, + "step": 12883 + }, + { + "epoch": 0.30202124978756073, + "grad_norm": 4.529463768005371, + "learning_rate": 0.00015830494890646917, + "loss": 0.4564, + "step": 12884 + }, + { + "epoch": 0.3020446913623657, + "grad_norm": 4.124884605407715, + "learning_rate": 0.00015829896421177728, + "loss": 0.4667, + "step": 12885 + }, + { + "epoch": 0.3020681329371707, + "grad_norm": 3.323864221572876, + "learning_rate": 0.00015829297920075356, + "loss": 0.2091, + "step": 12886 + }, + { + "epoch": 0.3020915745119757, + "grad_norm": 2.667241334915161, + "learning_rate": 0.00015828699387343052, + "loss": 0.2329, + "step": 12887 + }, + { + "epoch": 0.3021150160867807, + "grad_norm": 2.612368583679199, + "learning_rate": 0.00015828100822984055, + "loss": 0.2423, + "step": 12888 + }, + { + "epoch": 0.3021384576615857, + "grad_norm": 4.067419528961182, + "learning_rate": 0.00015827502227001622, + "loss": 0.4905, + "step": 12889 + }, + { + "epoch": 0.3021618992363907, + "grad_norm": 2.9439098834991455, + "learning_rate": 0.00015826903599399001, + "loss": 0.4505, + "step": 12890 + }, + { + "epoch": 0.3021853408111957, + "grad_norm": 2.8727381229400635, + "learning_rate": 0.00015826304940179435, + "loss": 0.442, + "step": 12891 + }, + { + "epoch": 0.3022087823860007, + "grad_norm": 1.0824872255325317, + "learning_rate": 0.00015825706249346174, + "loss": 0.0913, + "step": 12892 + }, + { + "epoch": 0.3022322239608057, + "grad_norm": 3.5497798919677734, + "learning_rate": 0.00015825107526902472, + "loss": 0.4776, + "step": 12893 + }, + { + "epoch": 0.3022556655356107, + "grad_norm": 3.3869192600250244, + "learning_rate": 0.00015824508772851567, + "loss": 0.3467, + "step": 12894 + }, + { + "epoch": 0.30227910711041567, + "grad_norm": 3.091761350631714, + "learning_rate": 0.00015823909987196716, + "loss": 0.5216, + "step": 12895 + }, + { + "epoch": 0.30230254868522066, + "grad_norm": 2.1262168884277344, + "learning_rate": 0.00015823311169941163, + "loss": 0.1937, + "step": 12896 + }, + { + "epoch": 0.30232599026002566, + "grad_norm": 3.770604133605957, + "learning_rate": 0.00015822712321088162, + "loss": 0.4889, + "step": 12897 + }, + { + "epoch": 0.30234943183483065, + "grad_norm": 1.280078411102295, + "learning_rate": 0.00015822113440640961, + "loss": 0.3011, + "step": 12898 + }, + { + "epoch": 0.30237287340963565, + "grad_norm": 2.8080315589904785, + "learning_rate": 0.0001582151452860281, + "loss": 0.2448, + "step": 12899 + }, + { + "epoch": 0.30239631498444064, + "grad_norm": 3.573981523513794, + "learning_rate": 0.00015820915584976953, + "loss": 0.3247, + "step": 12900 + }, + { + "epoch": 0.30241975655924563, + "grad_norm": 1.8148442506790161, + "learning_rate": 0.00015820316609766648, + "loss": 0.2305, + "step": 12901 + }, + { + "epoch": 0.30244319813405063, + "grad_norm": 3.4766972064971924, + "learning_rate": 0.0001581971760297514, + "loss": 0.3551, + "step": 12902 + }, + { + "epoch": 0.3024666397088556, + "grad_norm": 4.552981853485107, + "learning_rate": 0.0001581911856460568, + "loss": 0.5036, + "step": 12903 + }, + { + "epoch": 0.3024900812836606, + "grad_norm": 1.600021481513977, + "learning_rate": 0.00015818519494661522, + "loss": 0.1982, + "step": 12904 + }, + { + "epoch": 0.3025135228584656, + "grad_norm": 5.252697467803955, + "learning_rate": 0.0001581792039314591, + "loss": 0.5434, + "step": 12905 + }, + { + "epoch": 0.3025369644332706, + "grad_norm": 2.625737428665161, + "learning_rate": 0.000158173212600621, + "loss": 0.2978, + "step": 12906 + }, + { + "epoch": 0.3025604060080756, + "grad_norm": 4.6256303787231445, + "learning_rate": 0.0001581672209541334, + "loss": 0.4743, + "step": 12907 + }, + { + "epoch": 0.3025838475828806, + "grad_norm": 4.216610908508301, + "learning_rate": 0.00015816122899202885, + "loss": 0.4705, + "step": 12908 + }, + { + "epoch": 0.3026072891576856, + "grad_norm": 4.439990520477295, + "learning_rate": 0.00015815523671433983, + "loss": 0.571, + "step": 12909 + }, + { + "epoch": 0.3026307307324906, + "grad_norm": 0.5545346140861511, + "learning_rate": 0.00015814924412109886, + "loss": 0.0489, + "step": 12910 + }, + { + "epoch": 0.3026541723072956, + "grad_norm": 0.5364713668823242, + "learning_rate": 0.00015814325121233844, + "loss": 0.0447, + "step": 12911 + }, + { + "epoch": 0.3026776138821006, + "grad_norm": 2.527804136276245, + "learning_rate": 0.0001581372579880911, + "loss": 2.0459, + "step": 12912 + }, + { + "epoch": 0.3027010554569056, + "grad_norm": 0.4896607995033264, + "learning_rate": 0.0001581312644483894, + "loss": 0.0532, + "step": 12913 + }, + { + "epoch": 0.3027244970317106, + "grad_norm": 6.045191764831543, + "learning_rate": 0.0001581252705932658, + "loss": 0.7379, + "step": 12914 + }, + { + "epoch": 0.3027479386065156, + "grad_norm": 2.609619379043579, + "learning_rate": 0.00015811927642275285, + "loss": 0.2459, + "step": 12915 + }, + { + "epoch": 0.3027713801813206, + "grad_norm": 2.8290693759918213, + "learning_rate": 0.00015811328193688306, + "loss": 1.6153, + "step": 12916 + }, + { + "epoch": 0.3027948217561256, + "grad_norm": 3.2187178134918213, + "learning_rate": 0.00015810728713568897, + "loss": 0.3324, + "step": 12917 + }, + { + "epoch": 0.3028182633309306, + "grad_norm": 4.3474202156066895, + "learning_rate": 0.00015810129201920315, + "loss": 0.378, + "step": 12918 + }, + { + "epoch": 0.3028417049057356, + "grad_norm": 4.326859951019287, + "learning_rate": 0.000158095296587458, + "loss": 0.5106, + "step": 12919 + }, + { + "epoch": 0.3028651464805406, + "grad_norm": 2.5304343700408936, + "learning_rate": 0.00015808930084048621, + "loss": 0.2601, + "step": 12920 + }, + { + "epoch": 0.3028885880553456, + "grad_norm": 3.6868503093719482, + "learning_rate": 0.0001580833047783202, + "loss": 0.2319, + "step": 12921 + }, + { + "epoch": 0.30291202963015057, + "grad_norm": 2.660188674926758, + "learning_rate": 0.00015807730840099258, + "loss": 0.1833, + "step": 12922 + }, + { + "epoch": 0.30293547120495556, + "grad_norm": 2.315044641494751, + "learning_rate": 0.00015807131170853585, + "loss": 0.326, + "step": 12923 + }, + { + "epoch": 0.30295891277976056, + "grad_norm": 6.1324615478515625, + "learning_rate": 0.00015806531470098252, + "loss": 0.6499, + "step": 12924 + }, + { + "epoch": 0.30298235435456555, + "grad_norm": 5.68229866027832, + "learning_rate": 0.00015805931737836518, + "loss": 0.8495, + "step": 12925 + }, + { + "epoch": 0.30300579592937055, + "grad_norm": 4.678194999694824, + "learning_rate": 0.00015805331974071634, + "loss": 0.6649, + "step": 12926 + }, + { + "epoch": 0.30302923750417554, + "grad_norm": 5.600282669067383, + "learning_rate": 0.00015804732178806855, + "loss": 0.6285, + "step": 12927 + }, + { + "epoch": 0.30305267907898054, + "grad_norm": 1.930583119392395, + "learning_rate": 0.0001580413235204544, + "loss": 0.2731, + "step": 12928 + }, + { + "epoch": 0.30307612065378553, + "grad_norm": 2.043874502182007, + "learning_rate": 0.00015803532493790634, + "loss": 0.2039, + "step": 12929 + }, + { + "epoch": 0.3030995622285905, + "grad_norm": 2.346958875656128, + "learning_rate": 0.00015802932604045702, + "loss": 0.2055, + "step": 12930 + }, + { + "epoch": 0.3031230038033955, + "grad_norm": 0.6925447583198547, + "learning_rate": 0.0001580233268281389, + "loss": 0.0942, + "step": 12931 + }, + { + "epoch": 0.3031464453782005, + "grad_norm": 2.873481035232544, + "learning_rate": 0.00015801732730098462, + "loss": 0.3791, + "step": 12932 + }, + { + "epoch": 0.3031698869530055, + "grad_norm": 3.3124911785125732, + "learning_rate": 0.0001580113274590267, + "loss": 0.5058, + "step": 12933 + }, + { + "epoch": 0.3031933285278105, + "grad_norm": 4.521693706512451, + "learning_rate": 0.0001580053273022976, + "loss": 0.3926, + "step": 12934 + }, + { + "epoch": 0.3032167701026155, + "grad_norm": 2.9186439514160156, + "learning_rate": 0.00015799932683083006, + "loss": 0.4426, + "step": 12935 + }, + { + "epoch": 0.3032402116774205, + "grad_norm": 0.7870088815689087, + "learning_rate": 0.00015799332604465651, + "loss": 0.1279, + "step": 12936 + }, + { + "epoch": 0.3032636532522255, + "grad_norm": 5.079643726348877, + "learning_rate": 0.00015798732494380952, + "loss": 0.3677, + "step": 12937 + }, + { + "epoch": 0.3032870948270305, + "grad_norm": 2.6547012329101562, + "learning_rate": 0.00015798132352832172, + "loss": 0.273, + "step": 12938 + }, + { + "epoch": 0.3033105364018355, + "grad_norm": 2.5537943840026855, + "learning_rate": 0.00015797532179822558, + "loss": 0.3142, + "step": 12939 + }, + { + "epoch": 0.30333397797664047, + "grad_norm": 2.55100154876709, + "learning_rate": 0.00015796931975355376, + "loss": 1.9508, + "step": 12940 + }, + { + "epoch": 0.30335741955144546, + "grad_norm": 1.527693271636963, + "learning_rate": 0.00015796331739433874, + "loss": 0.2011, + "step": 12941 + }, + { + "epoch": 0.30338086112625046, + "grad_norm": 2.4444260597229004, + "learning_rate": 0.00015795731472061317, + "loss": 0.2496, + "step": 12942 + }, + { + "epoch": 0.30340430270105545, + "grad_norm": 3.404930591583252, + "learning_rate": 0.00015795131173240953, + "loss": 0.3562, + "step": 12943 + }, + { + "epoch": 0.30342774427586044, + "grad_norm": 0.8958269357681274, + "learning_rate": 0.00015794530842976048, + "loss": 0.0843, + "step": 12944 + }, + { + "epoch": 0.30345118585066544, + "grad_norm": 4.742453098297119, + "learning_rate": 0.00015793930481269855, + "loss": 0.2552, + "step": 12945 + }, + { + "epoch": 0.30347462742547043, + "grad_norm": 0.7123630046844482, + "learning_rate": 0.00015793330088125632, + "loss": 0.0743, + "step": 12946 + }, + { + "epoch": 0.30349806900027543, + "grad_norm": 0.8188419938087463, + "learning_rate": 0.00015792729663546636, + "loss": 0.0816, + "step": 12947 + }, + { + "epoch": 0.3035215105750804, + "grad_norm": 3.158621072769165, + "learning_rate": 0.00015792129207536126, + "loss": 0.4646, + "step": 12948 + }, + { + "epoch": 0.3035449521498854, + "grad_norm": 6.215979099273682, + "learning_rate": 0.00015791528720097362, + "loss": 0.7503, + "step": 12949 + }, + { + "epoch": 0.3035683937246904, + "grad_norm": 3.3859941959381104, + "learning_rate": 0.00015790928201233598, + "loss": 0.2745, + "step": 12950 + }, + { + "epoch": 0.3035918352994954, + "grad_norm": 1.3995282649993896, + "learning_rate": 0.00015790327650948096, + "loss": 0.2068, + "step": 12951 + }, + { + "epoch": 0.3036152768743004, + "grad_norm": 3.0908803939819336, + "learning_rate": 0.00015789727069244117, + "loss": 0.4648, + "step": 12952 + }, + { + "epoch": 0.3036387184491054, + "grad_norm": 3.448707103729248, + "learning_rate": 0.00015789126456124908, + "loss": 1.4623, + "step": 12953 + }, + { + "epoch": 0.3036621600239104, + "grad_norm": 0.9121044874191284, + "learning_rate": 0.00015788525811593745, + "loss": 0.0867, + "step": 12954 + }, + { + "epoch": 0.3036856015987154, + "grad_norm": 3.1540334224700928, + "learning_rate": 0.0001578792513565387, + "loss": 0.3699, + "step": 12955 + }, + { + "epoch": 0.3037090431735204, + "grad_norm": 3.2022461891174316, + "learning_rate": 0.00015787324428308557, + "loss": 0.421, + "step": 12956 + }, + { + "epoch": 0.30373248474832537, + "grad_norm": 1.3196152448654175, + "learning_rate": 0.00015786723689561053, + "loss": 0.1219, + "step": 12957 + }, + { + "epoch": 0.30375592632313037, + "grad_norm": 4.609962463378906, + "learning_rate": 0.0001578612291941463, + "loss": 0.459, + "step": 12958 + }, + { + "epoch": 0.30377936789793536, + "grad_norm": 2.9600515365600586, + "learning_rate": 0.00015785522117872539, + "loss": 0.2968, + "step": 12959 + }, + { + "epoch": 0.30380280947274035, + "grad_norm": 3.5712523460388184, + "learning_rate": 0.0001578492128493804, + "loss": 0.4116, + "step": 12960 + }, + { + "epoch": 0.30382625104754535, + "grad_norm": 2.9396400451660156, + "learning_rate": 0.00015784320420614395, + "loss": 0.1719, + "step": 12961 + }, + { + "epoch": 0.30384969262235034, + "grad_norm": 3.203843355178833, + "learning_rate": 0.00015783719524904867, + "loss": 0.3565, + "step": 12962 + }, + { + "epoch": 0.3038731341971554, + "grad_norm": 1.2079070806503296, + "learning_rate": 0.00015783118597812717, + "loss": 0.1651, + "step": 12963 + }, + { + "epoch": 0.3038965757719604, + "grad_norm": 3.9448773860931396, + "learning_rate": 0.00015782517639341198, + "loss": 0.2922, + "step": 12964 + }, + { + "epoch": 0.3039200173467654, + "grad_norm": 5.6903510093688965, + "learning_rate": 0.00015781916649493577, + "loss": 0.4558, + "step": 12965 + }, + { + "epoch": 0.3039434589215704, + "grad_norm": 2.970167636871338, + "learning_rate": 0.00015781315628273117, + "loss": 0.2302, + "step": 12966 + }, + { + "epoch": 0.30396690049637537, + "grad_norm": 2.6052963733673096, + "learning_rate": 0.00015780714575683072, + "loss": 0.1904, + "step": 12967 + }, + { + "epoch": 0.30399034207118036, + "grad_norm": 2.8278584480285645, + "learning_rate": 0.00015780113491726709, + "loss": 0.3539, + "step": 12968 + }, + { + "epoch": 0.30401378364598536, + "grad_norm": 4.290644645690918, + "learning_rate": 0.00015779512376407287, + "loss": 0.4212, + "step": 12969 + }, + { + "epoch": 0.30403722522079035, + "grad_norm": 1.8355588912963867, + "learning_rate": 0.0001577891122972807, + "loss": 0.143, + "step": 12970 + }, + { + "epoch": 0.30406066679559535, + "grad_norm": 3.246516704559326, + "learning_rate": 0.00015778310051692314, + "loss": 0.2675, + "step": 12971 + }, + { + "epoch": 0.30408410837040034, + "grad_norm": 1.6182879209518433, + "learning_rate": 0.00015777708842303287, + "loss": 0.1723, + "step": 12972 + }, + { + "epoch": 0.30410754994520534, + "grad_norm": 5.582156181335449, + "learning_rate": 0.00015777107601564253, + "loss": 0.4468, + "step": 12973 + }, + { + "epoch": 0.30413099152001033, + "grad_norm": 0.3119869828224182, + "learning_rate": 0.00015776506329478465, + "loss": 0.0325, + "step": 12974 + }, + { + "epoch": 0.3041544330948153, + "grad_norm": 1.5690902471542358, + "learning_rate": 0.00015775905026049194, + "loss": 1.1216, + "step": 12975 + }, + { + "epoch": 0.3041778746696203, + "grad_norm": 2.4796223640441895, + "learning_rate": 0.000157753036912797, + "loss": 0.1615, + "step": 12976 + }, + { + "epoch": 0.3042013162444253, + "grad_norm": 0.5561808347702026, + "learning_rate": 0.00015774702325173245, + "loss": 0.042, + "step": 12977 + }, + { + "epoch": 0.3042247578192303, + "grad_norm": 2.941089391708374, + "learning_rate": 0.00015774100927733094, + "loss": 0.2796, + "step": 12978 + }, + { + "epoch": 0.3042481993940353, + "grad_norm": 1.9393956661224365, + "learning_rate": 0.0001577349949896251, + "loss": 0.1712, + "step": 12979 + }, + { + "epoch": 0.3042716409688403, + "grad_norm": 4.1806769371032715, + "learning_rate": 0.0001577289803886475, + "loss": 0.5684, + "step": 12980 + }, + { + "epoch": 0.3042950825436453, + "grad_norm": 4.1224894523620605, + "learning_rate": 0.00015772296547443088, + "loss": 0.3369, + "step": 12981 + }, + { + "epoch": 0.3043185241184503, + "grad_norm": 6.33953857421875, + "learning_rate": 0.0001577169502470078, + "loss": 0.7346, + "step": 12982 + }, + { + "epoch": 0.3043419656932553, + "grad_norm": 1.426039457321167, + "learning_rate": 0.00015771093470641096, + "loss": 0.122, + "step": 12983 + }, + { + "epoch": 0.3043654072680603, + "grad_norm": 0.7092618942260742, + "learning_rate": 0.00015770491885267295, + "loss": 0.1004, + "step": 12984 + }, + { + "epoch": 0.30438884884286527, + "grad_norm": 2.3524718284606934, + "learning_rate": 0.00015769890268582643, + "loss": 0.9058, + "step": 12985 + }, + { + "epoch": 0.30441229041767026, + "grad_norm": 2.7974658012390137, + "learning_rate": 0.000157692886205904, + "loss": 0.4644, + "step": 12986 + }, + { + "epoch": 0.30443573199247526, + "grad_norm": 1.0836219787597656, + "learning_rate": 0.0001576868694129384, + "loss": 0.1287, + "step": 12987 + }, + { + "epoch": 0.30445917356728025, + "grad_norm": 1.9382604360580444, + "learning_rate": 0.00015768085230696222, + "loss": 0.246, + "step": 12988 + }, + { + "epoch": 0.30448261514208524, + "grad_norm": 3.89339542388916, + "learning_rate": 0.00015767483488800807, + "loss": 0.391, + "step": 12989 + }, + { + "epoch": 0.30450605671689024, + "grad_norm": 2.313446521759033, + "learning_rate": 0.0001576688171561087, + "loss": 0.2977, + "step": 12990 + }, + { + "epoch": 0.30452949829169523, + "grad_norm": 3.4790170192718506, + "learning_rate": 0.00015766279911129667, + "loss": 0.2987, + "step": 12991 + }, + { + "epoch": 0.3045529398665002, + "grad_norm": 2.6876895427703857, + "learning_rate": 0.0001576567807536047, + "loss": 0.2995, + "step": 12992 + }, + { + "epoch": 0.3045763814413052, + "grad_norm": 7.472467422485352, + "learning_rate": 0.00015765076208306536, + "loss": 0.6141, + "step": 12993 + }, + { + "epoch": 0.3045998230161102, + "grad_norm": 4.607261657714844, + "learning_rate": 0.00015764474309971143, + "loss": 0.6613, + "step": 12994 + }, + { + "epoch": 0.3046232645909152, + "grad_norm": 3.1312637329101562, + "learning_rate": 0.00015763872380357546, + "loss": 0.3285, + "step": 12995 + }, + { + "epoch": 0.3046467061657202, + "grad_norm": 5.001219272613525, + "learning_rate": 0.0001576327041946902, + "loss": 0.2938, + "step": 12996 + }, + { + "epoch": 0.3046701477405252, + "grad_norm": 1.4034700393676758, + "learning_rate": 0.0001576266842730882, + "loss": 0.188, + "step": 12997 + }, + { + "epoch": 0.3046935893153302, + "grad_norm": 3.2750320434570312, + "learning_rate": 0.00015762066403880222, + "loss": 0.3672, + "step": 12998 + }, + { + "epoch": 0.3047170308901352, + "grad_norm": 2.617598533630371, + "learning_rate": 0.00015761464349186492, + "loss": 0.3034, + "step": 12999 + }, + { + "epoch": 0.3047404724649402, + "grad_norm": 2.685668468475342, + "learning_rate": 0.00015760862263230894, + "loss": 0.1772, + "step": 13000 + }, + { + "epoch": 0.3047639140397452, + "grad_norm": 1.1169592142105103, + "learning_rate": 0.00015760260146016695, + "loss": 0.1551, + "step": 13001 + }, + { + "epoch": 0.30478735561455017, + "grad_norm": 3.21205735206604, + "learning_rate": 0.0001575965799754716, + "loss": 0.3291, + "step": 13002 + }, + { + "epoch": 0.30481079718935516, + "grad_norm": 2.334256887435913, + "learning_rate": 0.0001575905581782556, + "loss": 0.4198, + "step": 13003 + }, + { + "epoch": 0.30483423876416016, + "grad_norm": 1.5405398607254028, + "learning_rate": 0.00015758453606855158, + "loss": 0.1886, + "step": 13004 + }, + { + "epoch": 0.30485768033896515, + "grad_norm": 7.212883949279785, + "learning_rate": 0.0001575785136463923, + "loss": 0.6386, + "step": 13005 + }, + { + "epoch": 0.30488112191377015, + "grad_norm": 1.8533668518066406, + "learning_rate": 0.0001575724909118104, + "loss": 0.1129, + "step": 13006 + }, + { + "epoch": 0.30490456348857514, + "grad_norm": 11.866270065307617, + "learning_rate": 0.0001575664678648385, + "loss": 0.4983, + "step": 13007 + }, + { + "epoch": 0.30492800506338014, + "grad_norm": 3.7612831592559814, + "learning_rate": 0.00015756044450550936, + "loss": 2.3043, + "step": 13008 + }, + { + "epoch": 0.30495144663818513, + "grad_norm": 4.188495635986328, + "learning_rate": 0.0001575544208338556, + "loss": 0.4684, + "step": 13009 + }, + { + "epoch": 0.3049748882129901, + "grad_norm": 0.7865586280822754, + "learning_rate": 0.00015754839684990995, + "loss": 0.1461, + "step": 13010 + }, + { + "epoch": 0.3049983297877951, + "grad_norm": 1.6805667877197266, + "learning_rate": 0.00015754237255370508, + "loss": 0.1406, + "step": 13011 + }, + { + "epoch": 0.3050217713626001, + "grad_norm": 2.924321174621582, + "learning_rate": 0.00015753634794527364, + "loss": 0.2415, + "step": 13012 + }, + { + "epoch": 0.3050452129374051, + "grad_norm": 3.8499152660369873, + "learning_rate": 0.00015753032302464842, + "loss": 0.3945, + "step": 13013 + }, + { + "epoch": 0.30506865451221016, + "grad_norm": 3.611582040786743, + "learning_rate": 0.00015752429779186203, + "loss": 0.3376, + "step": 13014 + }, + { + "epoch": 0.30509209608701515, + "grad_norm": 2.5485808849334717, + "learning_rate": 0.00015751827224694717, + "loss": 0.3331, + "step": 13015 + }, + { + "epoch": 0.30511553766182015, + "grad_norm": 5.13325834274292, + "learning_rate": 0.00015751224638993654, + "loss": 0.5459, + "step": 13016 + }, + { + "epoch": 0.30513897923662514, + "grad_norm": 1.3068557977676392, + "learning_rate": 0.00015750622022086286, + "loss": 0.168, + "step": 13017 + }, + { + "epoch": 0.30516242081143014, + "grad_norm": 3.8091952800750732, + "learning_rate": 0.0001575001937397588, + "loss": 0.3879, + "step": 13018 + }, + { + "epoch": 0.30518586238623513, + "grad_norm": 3.280207633972168, + "learning_rate": 0.0001574941669466571, + "loss": 0.2699, + "step": 13019 + }, + { + "epoch": 0.3052093039610401, + "grad_norm": 4.7299675941467285, + "learning_rate": 0.0001574881398415904, + "loss": 0.3854, + "step": 13020 + }, + { + "epoch": 0.3052327455358451, + "grad_norm": 4.6769585609436035, + "learning_rate": 0.00015748211242459147, + "loss": 0.4564, + "step": 13021 + }, + { + "epoch": 0.3052561871106501, + "grad_norm": 2.459540605545044, + "learning_rate": 0.00015747608469569296, + "loss": 0.3489, + "step": 13022 + }, + { + "epoch": 0.3052796286854551, + "grad_norm": 1.7251605987548828, + "learning_rate": 0.0001574700566549276, + "loss": 0.278, + "step": 13023 + }, + { + "epoch": 0.3053030702602601, + "grad_norm": 2.4117939472198486, + "learning_rate": 0.0001574640283023281, + "loss": 0.3956, + "step": 13024 + }, + { + "epoch": 0.3053265118350651, + "grad_norm": 3.3226513862609863, + "learning_rate": 0.00015745799963792716, + "loss": 0.3273, + "step": 13025 + }, + { + "epoch": 0.3053499534098701, + "grad_norm": 2.1819143295288086, + "learning_rate": 0.0001574519706617575, + "loss": 0.2427, + "step": 13026 + }, + { + "epoch": 0.3053733949846751, + "grad_norm": 9.684500694274902, + "learning_rate": 0.00015744594137385182, + "loss": 0.5158, + "step": 13027 + }, + { + "epoch": 0.3053968365594801, + "grad_norm": 3.6101772785186768, + "learning_rate": 0.00015743991177424285, + "loss": 0.438, + "step": 13028 + }, + { + "epoch": 0.3054202781342851, + "grad_norm": 3.520308494567871, + "learning_rate": 0.00015743388186296333, + "loss": 0.2773, + "step": 13029 + }, + { + "epoch": 0.30544371970909007, + "grad_norm": 1.744227647781372, + "learning_rate": 0.00015742785164004594, + "loss": 0.1973, + "step": 13030 + }, + { + "epoch": 0.30546716128389506, + "grad_norm": 3.07138729095459, + "learning_rate": 0.00015742182110552343, + "loss": 0.3435, + "step": 13031 + }, + { + "epoch": 0.30549060285870006, + "grad_norm": 3.5519943237304688, + "learning_rate": 0.00015741579025942848, + "loss": 0.2466, + "step": 13032 + }, + { + "epoch": 0.30551404443350505, + "grad_norm": 3.7083067893981934, + "learning_rate": 0.00015740975910179383, + "loss": 0.3681, + "step": 13033 + }, + { + "epoch": 0.30553748600831004, + "grad_norm": 4.043458461761475, + "learning_rate": 0.00015740372763265223, + "loss": 0.9978, + "step": 13034 + }, + { + "epoch": 0.30556092758311504, + "grad_norm": 1.7925206422805786, + "learning_rate": 0.00015739769585203638, + "loss": 1.6878, + "step": 13035 + }, + { + "epoch": 0.30558436915792003, + "grad_norm": 5.116499900817871, + "learning_rate": 0.00015739166375997903, + "loss": 0.4021, + "step": 13036 + }, + { + "epoch": 0.305607810732725, + "grad_norm": 3.8821043968200684, + "learning_rate": 0.00015738563135651287, + "loss": 0.2827, + "step": 13037 + }, + { + "epoch": 0.30563125230753, + "grad_norm": 3.1444568634033203, + "learning_rate": 0.0001573795986416707, + "loss": 0.2264, + "step": 13038 + }, + { + "epoch": 0.305654693882335, + "grad_norm": 4.694176197052002, + "learning_rate": 0.0001573735656154852, + "loss": 0.5624, + "step": 13039 + }, + { + "epoch": 0.30567813545714, + "grad_norm": 1.2008908987045288, + "learning_rate": 0.0001573675322779891, + "loss": 0.121, + "step": 13040 + }, + { + "epoch": 0.305701577031945, + "grad_norm": 1.4218193292617798, + "learning_rate": 0.00015736149862921516, + "loss": 0.1766, + "step": 13041 + }, + { + "epoch": 0.30572501860675, + "grad_norm": 2.332042932510376, + "learning_rate": 0.00015735546466919616, + "loss": 1.5065, + "step": 13042 + }, + { + "epoch": 0.305748460181555, + "grad_norm": 0.8897563815116882, + "learning_rate": 0.00015734943039796476, + "loss": 0.0754, + "step": 13043 + }, + { + "epoch": 0.30577190175636, + "grad_norm": 5.209662437438965, + "learning_rate": 0.00015734339581555373, + "loss": 0.5369, + "step": 13044 + }, + { + "epoch": 0.305795343331165, + "grad_norm": 4.927392959594727, + "learning_rate": 0.00015733736092199584, + "loss": 0.9044, + "step": 13045 + }, + { + "epoch": 0.30581878490597, + "grad_norm": 1.6712838411331177, + "learning_rate": 0.0001573313257173238, + "loss": 0.1241, + "step": 13046 + }, + { + "epoch": 0.30584222648077497, + "grad_norm": 3.7500414848327637, + "learning_rate": 0.00015732529020157042, + "loss": 0.4852, + "step": 13047 + }, + { + "epoch": 0.30586566805557996, + "grad_norm": 3.3712387084960938, + "learning_rate": 0.00015731925437476835, + "loss": 0.1627, + "step": 13048 + }, + { + "epoch": 0.30588910963038496, + "grad_norm": 1.4478892087936401, + "learning_rate": 0.00015731321823695042, + "loss": 0.0911, + "step": 13049 + }, + { + "epoch": 0.30591255120518995, + "grad_norm": 4.1349263191223145, + "learning_rate": 0.00015730718178814934, + "loss": 0.4397, + "step": 13050 + }, + { + "epoch": 0.30593599277999495, + "grad_norm": 0.9967558979988098, + "learning_rate": 0.0001573011450283979, + "loss": 0.0951, + "step": 13051 + }, + { + "epoch": 0.30595943435479994, + "grad_norm": 2.5446176528930664, + "learning_rate": 0.0001572951079577288, + "loss": 1.337, + "step": 13052 + }, + { + "epoch": 0.30598287592960494, + "grad_norm": 2.357550621032715, + "learning_rate": 0.0001572890705761749, + "loss": 0.2193, + "step": 13053 + }, + { + "epoch": 0.30600631750440993, + "grad_norm": 3.905801773071289, + "learning_rate": 0.00015728303288376885, + "loss": 0.5459, + "step": 13054 + }, + { + "epoch": 0.3060297590792149, + "grad_norm": 6.318211078643799, + "learning_rate": 0.00015727699488054343, + "loss": 0.823, + "step": 13055 + }, + { + "epoch": 0.3060532006540199, + "grad_norm": 3.3812954425811768, + "learning_rate": 0.0001572709565665315, + "loss": 0.4546, + "step": 13056 + }, + { + "epoch": 0.3060766422288249, + "grad_norm": 2.866792917251587, + "learning_rate": 0.00015726491794176565, + "loss": 0.4294, + "step": 13057 + }, + { + "epoch": 0.3061000838036299, + "grad_norm": 3.1981253623962402, + "learning_rate": 0.00015725887900627883, + "loss": 0.3755, + "step": 13058 + }, + { + "epoch": 0.3061235253784349, + "grad_norm": 5.491255283355713, + "learning_rate": 0.00015725283976010366, + "loss": 0.4109, + "step": 13059 + }, + { + "epoch": 0.3061469669532399, + "grad_norm": 5.632997989654541, + "learning_rate": 0.00015724680020327302, + "loss": 0.294, + "step": 13060 + }, + { + "epoch": 0.3061704085280449, + "grad_norm": 5.188924312591553, + "learning_rate": 0.0001572407603358196, + "loss": 0.3977, + "step": 13061 + }, + { + "epoch": 0.3061938501028499, + "grad_norm": 4.686427593231201, + "learning_rate": 0.00015723472015777623, + "loss": 0.3925, + "step": 13062 + }, + { + "epoch": 0.3062172916776549, + "grad_norm": 2.606210947036743, + "learning_rate": 0.00015722867966917566, + "loss": 0.3138, + "step": 13063 + }, + { + "epoch": 0.3062407332524599, + "grad_norm": 1.5591448545455933, + "learning_rate": 0.00015722263887005063, + "loss": 1.0131, + "step": 13064 + }, + { + "epoch": 0.3062641748272649, + "grad_norm": 2.8896522521972656, + "learning_rate": 0.00015721659776043399, + "loss": 2.0801, + "step": 13065 + }, + { + "epoch": 0.3062876164020699, + "grad_norm": 3.36745548248291, + "learning_rate": 0.00015721055634035843, + "loss": 0.464, + "step": 13066 + }, + { + "epoch": 0.3063110579768749, + "grad_norm": 5.134283542633057, + "learning_rate": 0.00015720451460985682, + "loss": 0.6079, + "step": 13067 + }, + { + "epoch": 0.3063344995516799, + "grad_norm": 4.286677837371826, + "learning_rate": 0.00015719847256896193, + "loss": 0.2331, + "step": 13068 + }, + { + "epoch": 0.3063579411264849, + "grad_norm": 1.5407682657241821, + "learning_rate": 0.0001571924302177065, + "loss": 0.093, + "step": 13069 + }, + { + "epoch": 0.3063813827012899, + "grad_norm": 4.348327159881592, + "learning_rate": 0.0001571863875561233, + "loss": 0.5935, + "step": 13070 + }, + { + "epoch": 0.3064048242760949, + "grad_norm": 4.692074775695801, + "learning_rate": 0.0001571803445842452, + "loss": 0.6605, + "step": 13071 + }, + { + "epoch": 0.3064282658508999, + "grad_norm": 2.3648204803466797, + "learning_rate": 0.0001571743013021049, + "loss": 0.3081, + "step": 13072 + }, + { + "epoch": 0.3064517074257049, + "grad_norm": 6.4400248527526855, + "learning_rate": 0.00015716825770973526, + "loss": 0.3873, + "step": 13073 + }, + { + "epoch": 0.30647514900050987, + "grad_norm": 3.419287919998169, + "learning_rate": 0.00015716221380716907, + "loss": 0.4451, + "step": 13074 + }, + { + "epoch": 0.30649859057531487, + "grad_norm": 4.087954521179199, + "learning_rate": 0.00015715616959443906, + "loss": 0.4517, + "step": 13075 + }, + { + "epoch": 0.30652203215011986, + "grad_norm": 3.1969659328460693, + "learning_rate": 0.00015715012507157807, + "loss": 0.2256, + "step": 13076 + }, + { + "epoch": 0.30654547372492486, + "grad_norm": 4.314979553222656, + "learning_rate": 0.0001571440802386189, + "loss": 0.4445, + "step": 13077 + }, + { + "epoch": 0.30656891529972985, + "grad_norm": 3.093845844268799, + "learning_rate": 0.00015713803509559434, + "loss": 0.2769, + "step": 13078 + }, + { + "epoch": 0.30659235687453484, + "grad_norm": 3.2291440963745117, + "learning_rate": 0.00015713198964253724, + "loss": 0.3689, + "step": 13079 + }, + { + "epoch": 0.30661579844933984, + "grad_norm": 1.5750656127929688, + "learning_rate": 0.0001571259438794803, + "loss": 0.1509, + "step": 13080 + }, + { + "epoch": 0.30663924002414483, + "grad_norm": 2.299370050430298, + "learning_rate": 0.00015711989780645643, + "loss": 0.2198, + "step": 13081 + }, + { + "epoch": 0.3066626815989498, + "grad_norm": 3.0665597915649414, + "learning_rate": 0.00015711385142349834, + "loss": 1.9537, + "step": 13082 + }, + { + "epoch": 0.3066861231737548, + "grad_norm": 4.92396879196167, + "learning_rate": 0.0001571078047306389, + "loss": 0.5555, + "step": 13083 + }, + { + "epoch": 0.3067095647485598, + "grad_norm": 2.453888416290283, + "learning_rate": 0.00015710175772791095, + "loss": 0.2266, + "step": 13084 + }, + { + "epoch": 0.3067330063233648, + "grad_norm": 0.8452664017677307, + "learning_rate": 0.0001570957104153472, + "loss": 0.0785, + "step": 13085 + }, + { + "epoch": 0.3067564478981698, + "grad_norm": 2.9844484329223633, + "learning_rate": 0.00015708966279298053, + "loss": 0.2713, + "step": 13086 + }, + { + "epoch": 0.3067798894729748, + "grad_norm": 2.991699457168579, + "learning_rate": 0.00015708361486084377, + "loss": 0.429, + "step": 13087 + }, + { + "epoch": 0.3068033310477798, + "grad_norm": 5.327384948730469, + "learning_rate": 0.0001570775666189697, + "loss": 0.6841, + "step": 13088 + }, + { + "epoch": 0.3068267726225848, + "grad_norm": 3.386223077774048, + "learning_rate": 0.00015707151806739113, + "loss": 0.2814, + "step": 13089 + }, + { + "epoch": 0.3068502141973898, + "grad_norm": 4.223571300506592, + "learning_rate": 0.00015706546920614094, + "loss": 0.6205, + "step": 13090 + }, + { + "epoch": 0.3068736557721948, + "grad_norm": 3.7272872924804688, + "learning_rate": 0.00015705942003525186, + "loss": 0.4859, + "step": 13091 + }, + { + "epoch": 0.30689709734699977, + "grad_norm": 4.508453845977783, + "learning_rate": 0.0001570533705547568, + "loss": 0.5817, + "step": 13092 + }, + { + "epoch": 0.30692053892180476, + "grad_norm": 3.5341715812683105, + "learning_rate": 0.00015704732076468852, + "loss": 0.3839, + "step": 13093 + }, + { + "epoch": 0.30694398049660976, + "grad_norm": 3.617151975631714, + "learning_rate": 0.0001570412706650799, + "loss": 0.5835, + "step": 13094 + }, + { + "epoch": 0.30696742207141475, + "grad_norm": 2.5389044284820557, + "learning_rate": 0.00015703522025596373, + "loss": 1.6544, + "step": 13095 + }, + { + "epoch": 0.30699086364621975, + "grad_norm": 2.9751250743865967, + "learning_rate": 0.00015702916953737285, + "loss": 0.4059, + "step": 13096 + }, + { + "epoch": 0.30701430522102474, + "grad_norm": 1.389869213104248, + "learning_rate": 0.0001570231185093401, + "loss": 0.1276, + "step": 13097 + }, + { + "epoch": 0.30703774679582974, + "grad_norm": 1.268041729927063, + "learning_rate": 0.00015701706717189827, + "loss": 0.142, + "step": 13098 + }, + { + "epoch": 0.30706118837063473, + "grad_norm": 1.0017237663269043, + "learning_rate": 0.00015701101552508028, + "loss": 0.0713, + "step": 13099 + }, + { + "epoch": 0.3070846299454397, + "grad_norm": 3.7008438110351562, + "learning_rate": 0.00015700496356891889, + "loss": 0.3767, + "step": 13100 + }, + { + "epoch": 0.3071080715202447, + "grad_norm": 2.4698057174682617, + "learning_rate": 0.00015699891130344695, + "loss": 0.2141, + "step": 13101 + }, + { + "epoch": 0.3071315130950497, + "grad_norm": 1.0150212049484253, + "learning_rate": 0.00015699285872869734, + "loss": 0.0759, + "step": 13102 + }, + { + "epoch": 0.3071549546698547, + "grad_norm": 1.4783453941345215, + "learning_rate": 0.0001569868058447029, + "loss": 0.2457, + "step": 13103 + }, + { + "epoch": 0.3071783962446597, + "grad_norm": 4.774276256561279, + "learning_rate": 0.00015698075265149642, + "loss": 0.7711, + "step": 13104 + }, + { + "epoch": 0.3072018378194647, + "grad_norm": 2.393702268600464, + "learning_rate": 0.00015697469914911075, + "loss": 0.313, + "step": 13105 + }, + { + "epoch": 0.3072252793942697, + "grad_norm": 0.8550882339477539, + "learning_rate": 0.0001569686453375788, + "loss": 0.1133, + "step": 13106 + }, + { + "epoch": 0.3072487209690747, + "grad_norm": 4.454981327056885, + "learning_rate": 0.00015696259121693334, + "loss": 0.5955, + "step": 13107 + }, + { + "epoch": 0.3072721625438797, + "grad_norm": 2.2832281589508057, + "learning_rate": 0.00015695653678720727, + "loss": 0.2619, + "step": 13108 + }, + { + "epoch": 0.3072956041186847, + "grad_norm": 2.878871440887451, + "learning_rate": 0.00015695048204843344, + "loss": 0.3408, + "step": 13109 + }, + { + "epoch": 0.30731904569348967, + "grad_norm": 4.585505485534668, + "learning_rate": 0.00015694442700064473, + "loss": 0.6275, + "step": 13110 + }, + { + "epoch": 0.30734248726829466, + "grad_norm": 3.704110860824585, + "learning_rate": 0.00015693837164387393, + "loss": 0.4363, + "step": 13111 + }, + { + "epoch": 0.30736592884309966, + "grad_norm": 3.1435611248016357, + "learning_rate": 0.0001569323159781539, + "loss": 0.3659, + "step": 13112 + }, + { + "epoch": 0.30738937041790465, + "grad_norm": 3.3133134841918945, + "learning_rate": 0.00015692626000351752, + "loss": 0.4517, + "step": 13113 + }, + { + "epoch": 0.30741281199270964, + "grad_norm": 1.2139443159103394, + "learning_rate": 0.00015692020371999767, + "loss": 0.8322, + "step": 13114 + }, + { + "epoch": 0.30743625356751464, + "grad_norm": 2.9909543991088867, + "learning_rate": 0.0001569141471276272, + "loss": 1.8838, + "step": 13115 + }, + { + "epoch": 0.30745969514231963, + "grad_norm": 1.3470433950424194, + "learning_rate": 0.000156908090226439, + "loss": 0.1018, + "step": 13116 + }, + { + "epoch": 0.3074831367171247, + "grad_norm": 0.8857746720314026, + "learning_rate": 0.00015690203301646583, + "loss": 0.1494, + "step": 13117 + }, + { + "epoch": 0.3075065782919297, + "grad_norm": 1.1380983591079712, + "learning_rate": 0.00015689597549774069, + "loss": 0.1213, + "step": 13118 + }, + { + "epoch": 0.30753001986673467, + "grad_norm": 4.654052257537842, + "learning_rate": 0.00015688991767029634, + "loss": 0.3537, + "step": 13119 + }, + { + "epoch": 0.30755346144153967, + "grad_norm": 1.9167054891586304, + "learning_rate": 0.00015688385953416573, + "loss": 1.0893, + "step": 13120 + }, + { + "epoch": 0.30757690301634466, + "grad_norm": 5.4762282371521, + "learning_rate": 0.0001568778010893817, + "loss": 0.5133, + "step": 13121 + }, + { + "epoch": 0.30760034459114965, + "grad_norm": 4.715665340423584, + "learning_rate": 0.0001568717423359771, + "loss": 0.4909, + "step": 13122 + }, + { + "epoch": 0.30762378616595465, + "grad_norm": 3.0050222873687744, + "learning_rate": 0.00015686568327398482, + "loss": 1.4136, + "step": 13123 + }, + { + "epoch": 0.30764722774075964, + "grad_norm": 3.890684127807617, + "learning_rate": 0.00015685962390343775, + "loss": 0.5024, + "step": 13124 + }, + { + "epoch": 0.30767066931556464, + "grad_norm": 3.948065757751465, + "learning_rate": 0.0001568535642243688, + "loss": 0.311, + "step": 13125 + }, + { + "epoch": 0.30769411089036963, + "grad_norm": 2.927128553390503, + "learning_rate": 0.00015684750423681075, + "loss": 0.2205, + "step": 13126 + }, + { + "epoch": 0.3077175524651746, + "grad_norm": 3.1218996047973633, + "learning_rate": 0.0001568414439407966, + "loss": 0.3695, + "step": 13127 + }, + { + "epoch": 0.3077409940399796, + "grad_norm": 0.7845509648323059, + "learning_rate": 0.00015683538333635915, + "loss": 0.0581, + "step": 13128 + }, + { + "epoch": 0.3077644356147846, + "grad_norm": 4.232983112335205, + "learning_rate": 0.00015682932242353133, + "loss": 0.64, + "step": 13129 + }, + { + "epoch": 0.3077878771895896, + "grad_norm": 7.637691020965576, + "learning_rate": 0.000156823261202346, + "loss": 0.3158, + "step": 13130 + }, + { + "epoch": 0.3078113187643946, + "grad_norm": 6.89250373840332, + "learning_rate": 0.00015681719967283606, + "loss": 0.5325, + "step": 13131 + }, + { + "epoch": 0.3078347603391996, + "grad_norm": 5.086641311645508, + "learning_rate": 0.00015681113783503435, + "loss": 0.6322, + "step": 13132 + }, + { + "epoch": 0.3078582019140046, + "grad_norm": 2.4903223514556885, + "learning_rate": 0.00015680507568897387, + "loss": 0.2783, + "step": 13133 + }, + { + "epoch": 0.3078816434888096, + "grad_norm": 4.79203987121582, + "learning_rate": 0.00015679901323468744, + "loss": 0.5298, + "step": 13134 + }, + { + "epoch": 0.3079050850636146, + "grad_norm": 4.313962936401367, + "learning_rate": 0.00015679295047220794, + "loss": 0.6698, + "step": 13135 + }, + { + "epoch": 0.3079285266384196, + "grad_norm": 2.6534945964813232, + "learning_rate": 0.00015678688740156834, + "loss": 0.2014, + "step": 13136 + }, + { + "epoch": 0.30795196821322457, + "grad_norm": 2.8199570178985596, + "learning_rate": 0.00015678082402280144, + "loss": 0.3338, + "step": 13137 + }, + { + "epoch": 0.30797540978802956, + "grad_norm": 2.0239338874816895, + "learning_rate": 0.00015677476033594022, + "loss": 0.1435, + "step": 13138 + }, + { + "epoch": 0.30799885136283456, + "grad_norm": 3.92634916305542, + "learning_rate": 0.00015676869634101754, + "loss": 0.6222, + "step": 13139 + }, + { + "epoch": 0.30802229293763955, + "grad_norm": 3.442645311355591, + "learning_rate": 0.00015676263203806632, + "loss": 0.2763, + "step": 13140 + }, + { + "epoch": 0.30804573451244455, + "grad_norm": 1.5919380187988281, + "learning_rate": 0.00015675656742711948, + "loss": 1.1544, + "step": 13141 + }, + { + "epoch": 0.30806917608724954, + "grad_norm": 3.902064561843872, + "learning_rate": 0.00015675050250820992, + "loss": 0.4945, + "step": 13142 + }, + { + "epoch": 0.30809261766205454, + "grad_norm": 5.847886085510254, + "learning_rate": 0.00015674443728137049, + "loss": 0.8235, + "step": 13143 + }, + { + "epoch": 0.30811605923685953, + "grad_norm": 6.988833427429199, + "learning_rate": 0.00015673837174663417, + "loss": 0.5267, + "step": 13144 + }, + { + "epoch": 0.3081395008116645, + "grad_norm": 4.786613464355469, + "learning_rate": 0.00015673230590403385, + "loss": 0.8186, + "step": 13145 + }, + { + "epoch": 0.3081629423864695, + "grad_norm": 4.349836826324463, + "learning_rate": 0.00015672623975360246, + "loss": 0.6636, + "step": 13146 + }, + { + "epoch": 0.3081863839612745, + "grad_norm": 2.057619094848633, + "learning_rate": 0.00015672017329537287, + "loss": 0.2139, + "step": 13147 + }, + { + "epoch": 0.3082098255360795, + "grad_norm": 3.3151509761810303, + "learning_rate": 0.00015671410652937806, + "loss": 0.4111, + "step": 13148 + }, + { + "epoch": 0.3082332671108845, + "grad_norm": 6.011219501495361, + "learning_rate": 0.00015670803945565088, + "loss": 0.6319, + "step": 13149 + }, + { + "epoch": 0.3082567086856895, + "grad_norm": 3.0401268005371094, + "learning_rate": 0.0001567019720742243, + "loss": 0.3519, + "step": 13150 + }, + { + "epoch": 0.3082801502604945, + "grad_norm": 4.008392333984375, + "learning_rate": 0.00015669590438513123, + "loss": 0.3188, + "step": 13151 + }, + { + "epoch": 0.3083035918352995, + "grad_norm": 3.6576108932495117, + "learning_rate": 0.00015668983638840458, + "loss": 0.2327, + "step": 13152 + }, + { + "epoch": 0.3083270334101045, + "grad_norm": 4.467263221740723, + "learning_rate": 0.0001566837680840773, + "loss": 0.4256, + "step": 13153 + }, + { + "epoch": 0.3083504749849095, + "grad_norm": 2.508894205093384, + "learning_rate": 0.00015667769947218227, + "loss": 1.6228, + "step": 13154 + }, + { + "epoch": 0.30837391655971447, + "grad_norm": 2.6265904903411865, + "learning_rate": 0.00015667163055275247, + "loss": 0.2893, + "step": 13155 + }, + { + "epoch": 0.30839735813451946, + "grad_norm": 3.4746134281158447, + "learning_rate": 0.00015666556132582076, + "loss": 0.3675, + "step": 13156 + }, + { + "epoch": 0.30842079970932446, + "grad_norm": 2.557178497314453, + "learning_rate": 0.0001566594917914202, + "loss": 0.2662, + "step": 13157 + }, + { + "epoch": 0.30844424128412945, + "grad_norm": 1.281740427017212, + "learning_rate": 0.00015665342194958362, + "loss": 0.1203, + "step": 13158 + }, + { + "epoch": 0.30846768285893444, + "grad_norm": 2.855689287185669, + "learning_rate": 0.00015664735180034395, + "loss": 0.364, + "step": 13159 + }, + { + "epoch": 0.30849112443373944, + "grad_norm": 5.86847448348999, + "learning_rate": 0.00015664128134373418, + "loss": 0.6962, + "step": 13160 + }, + { + "epoch": 0.30851456600854443, + "grad_norm": 3.9584550857543945, + "learning_rate": 0.00015663521057978722, + "loss": 0.4456, + "step": 13161 + }, + { + "epoch": 0.3085380075833494, + "grad_norm": 0.9089537858963013, + "learning_rate": 0.00015662913950853602, + "loss": 0.094, + "step": 13162 + }, + { + "epoch": 0.3085614491581544, + "grad_norm": 0.9966199398040771, + "learning_rate": 0.0001566230681300135, + "loss": 0.1149, + "step": 13163 + }, + { + "epoch": 0.3085848907329594, + "grad_norm": 3.6382033824920654, + "learning_rate": 0.00015661699644425262, + "loss": 0.4816, + "step": 13164 + }, + { + "epoch": 0.3086083323077644, + "grad_norm": 1.2518894672393799, + "learning_rate": 0.00015661092445128633, + "loss": 0.1112, + "step": 13165 + }, + { + "epoch": 0.3086317738825694, + "grad_norm": 4.46196985244751, + "learning_rate": 0.0001566048521511476, + "loss": 0.1912, + "step": 13166 + }, + { + "epoch": 0.3086552154573744, + "grad_norm": 3.8541054725646973, + "learning_rate": 0.00015659877954386935, + "loss": 0.2711, + "step": 13167 + }, + { + "epoch": 0.30867865703217945, + "grad_norm": 3.6194822788238525, + "learning_rate": 0.00015659270662948447, + "loss": 0.3804, + "step": 13168 + }, + { + "epoch": 0.30870209860698444, + "grad_norm": 2.648888349533081, + "learning_rate": 0.00015658663340802602, + "loss": 1.4604, + "step": 13169 + }, + { + "epoch": 0.30872554018178944, + "grad_norm": 2.8590149879455566, + "learning_rate": 0.0001565805598795269, + "loss": 1.6296, + "step": 13170 + }, + { + "epoch": 0.30874898175659443, + "grad_norm": 1.2009460926055908, + "learning_rate": 0.00015657448604402006, + "loss": 0.169, + "step": 13171 + }, + { + "epoch": 0.3087724233313994, + "grad_norm": 0.398364394903183, + "learning_rate": 0.00015656841190153848, + "loss": 0.0424, + "step": 13172 + }, + { + "epoch": 0.3087958649062044, + "grad_norm": 0.5189386606216431, + "learning_rate": 0.00015656233745211512, + "loss": 0.0369, + "step": 13173 + }, + { + "epoch": 0.3088193064810094, + "grad_norm": 2.709351062774658, + "learning_rate": 0.0001565562626957829, + "loss": 0.3611, + "step": 13174 + }, + { + "epoch": 0.3088427480558144, + "grad_norm": 2.2894976139068604, + "learning_rate": 0.00015655018763257483, + "loss": 0.286, + "step": 13175 + }, + { + "epoch": 0.3088661896306194, + "grad_norm": 4.002047538757324, + "learning_rate": 0.00015654411226252385, + "loss": 0.2562, + "step": 13176 + }, + { + "epoch": 0.3088896312054244, + "grad_norm": 2.3292582035064697, + "learning_rate": 0.0001565380365856629, + "loss": 0.3705, + "step": 13177 + }, + { + "epoch": 0.3089130727802294, + "grad_norm": 3.2292897701263428, + "learning_rate": 0.000156531960602025, + "loss": 0.3121, + "step": 13178 + }, + { + "epoch": 0.3089365143550344, + "grad_norm": 8.318307876586914, + "learning_rate": 0.0001565258843116431, + "loss": 0.4332, + "step": 13179 + }, + { + "epoch": 0.3089599559298394, + "grad_norm": 3.498368501663208, + "learning_rate": 0.00015651980771455012, + "loss": 0.2873, + "step": 13180 + }, + { + "epoch": 0.3089833975046444, + "grad_norm": 3.236994504928589, + "learning_rate": 0.0001565137308107791, + "loss": 1.0654, + "step": 13181 + }, + { + "epoch": 0.30900683907944937, + "grad_norm": 1.3784229755401611, + "learning_rate": 0.000156507653600363, + "loss": 0.2229, + "step": 13182 + }, + { + "epoch": 0.30903028065425436, + "grad_norm": 3.4401943683624268, + "learning_rate": 0.0001565015760833348, + "loss": 0.2876, + "step": 13183 + }, + { + "epoch": 0.30905372222905936, + "grad_norm": 4.049440860748291, + "learning_rate": 0.0001564954982597274, + "loss": 0.6578, + "step": 13184 + }, + { + "epoch": 0.30907716380386435, + "grad_norm": 3.3327693939208984, + "learning_rate": 0.00015648942012957387, + "loss": 0.4553, + "step": 13185 + }, + { + "epoch": 0.30910060537866935, + "grad_norm": 5.493252754211426, + "learning_rate": 0.00015648334169290714, + "loss": 0.4085, + "step": 13186 + }, + { + "epoch": 0.30912404695347434, + "grad_norm": 0.9740351438522339, + "learning_rate": 0.00015647726294976025, + "loss": 0.0781, + "step": 13187 + }, + { + "epoch": 0.30914748852827934, + "grad_norm": 2.28568696975708, + "learning_rate": 0.0001564711839001661, + "loss": 0.218, + "step": 13188 + }, + { + "epoch": 0.30917093010308433, + "grad_norm": 4.101069927215576, + "learning_rate": 0.00015646510454415773, + "loss": 0.2597, + "step": 13189 + }, + { + "epoch": 0.3091943716778893, + "grad_norm": 6.16671085357666, + "learning_rate": 0.00015645902488176816, + "loss": 0.5687, + "step": 13190 + }, + { + "epoch": 0.3092178132526943, + "grad_norm": 5.06424617767334, + "learning_rate": 0.00015645294491303027, + "loss": 0.6129, + "step": 13191 + }, + { + "epoch": 0.3092412548274993, + "grad_norm": 1.179687261581421, + "learning_rate": 0.00015644686463797713, + "loss": 0.1331, + "step": 13192 + }, + { + "epoch": 0.3092646964023043, + "grad_norm": 3.6960196495056152, + "learning_rate": 0.00015644078405664173, + "loss": 1.3233, + "step": 13193 + }, + { + "epoch": 0.3092881379771093, + "grad_norm": 1.0615882873535156, + "learning_rate": 0.00015643470316905705, + "loss": 0.0878, + "step": 13194 + }, + { + "epoch": 0.3093115795519143, + "grad_norm": 2.289703369140625, + "learning_rate": 0.00015642862197525606, + "loss": 0.2239, + "step": 13195 + }, + { + "epoch": 0.3093350211267193, + "grad_norm": 1.2601733207702637, + "learning_rate": 0.0001564225404752718, + "loss": 0.1271, + "step": 13196 + }, + { + "epoch": 0.3093584627015243, + "grad_norm": 3.043696641921997, + "learning_rate": 0.00015641645866913727, + "loss": 0.2529, + "step": 13197 + }, + { + "epoch": 0.3093819042763293, + "grad_norm": 5.338229656219482, + "learning_rate": 0.00015641037655688542, + "loss": 0.5502, + "step": 13198 + }, + { + "epoch": 0.3094053458511343, + "grad_norm": 3.953930377960205, + "learning_rate": 0.00015640429413854928, + "loss": 0.4814, + "step": 13199 + }, + { + "epoch": 0.30942878742593927, + "grad_norm": 4.189572334289551, + "learning_rate": 0.00015639821141416187, + "loss": 0.2624, + "step": 13200 + }, + { + "epoch": 0.30945222900074426, + "grad_norm": 2.9655330181121826, + "learning_rate": 0.00015639212838375617, + "loss": 1.463, + "step": 13201 + }, + { + "epoch": 0.30947567057554926, + "grad_norm": 4.344783306121826, + "learning_rate": 0.00015638604504736517, + "loss": 0.499, + "step": 13202 + }, + { + "epoch": 0.30949911215035425, + "grad_norm": 3.5900330543518066, + "learning_rate": 0.00015637996140502195, + "loss": 0.4365, + "step": 13203 + }, + { + "epoch": 0.30952255372515924, + "grad_norm": 0.7408573627471924, + "learning_rate": 0.00015637387745675948, + "loss": 0.1009, + "step": 13204 + }, + { + "epoch": 0.30954599529996424, + "grad_norm": 1.4822150468826294, + "learning_rate": 0.00015636779320261072, + "loss": 0.1572, + "step": 13205 + }, + { + "epoch": 0.30956943687476923, + "grad_norm": 2.5739693641662598, + "learning_rate": 0.00015636170864260875, + "loss": 0.2925, + "step": 13206 + }, + { + "epoch": 0.3095928784495742, + "grad_norm": 2.883760690689087, + "learning_rate": 0.00015635562377678657, + "loss": 0.4501, + "step": 13207 + }, + { + "epoch": 0.3096163200243792, + "grad_norm": 4.548754692077637, + "learning_rate": 0.00015634953860517716, + "loss": 0.6319, + "step": 13208 + }, + { + "epoch": 0.3096397615991842, + "grad_norm": 3.4490370750427246, + "learning_rate": 0.00015634345312781362, + "loss": 0.4026, + "step": 13209 + }, + { + "epoch": 0.3096632031739892, + "grad_norm": 3.2342729568481445, + "learning_rate": 0.0001563373673447289, + "loss": 0.3489, + "step": 13210 + }, + { + "epoch": 0.3096866447487942, + "grad_norm": 2.6852781772613525, + "learning_rate": 0.000156331281255956, + "loss": 0.2257, + "step": 13211 + }, + { + "epoch": 0.3097100863235992, + "grad_norm": 2.8602020740509033, + "learning_rate": 0.00015632519486152803, + "loss": 0.3079, + "step": 13212 + }, + { + "epoch": 0.3097335278984042, + "grad_norm": 0.9801938533782959, + "learning_rate": 0.00015631910816147796, + "loss": 0.1421, + "step": 13213 + }, + { + "epoch": 0.3097569694732092, + "grad_norm": 1.3829556703567505, + "learning_rate": 0.00015631302115583885, + "loss": 0.1442, + "step": 13214 + }, + { + "epoch": 0.3097804110480142, + "grad_norm": 1.3851423263549805, + "learning_rate": 0.00015630693384464364, + "loss": 0.1781, + "step": 13215 + }, + { + "epoch": 0.3098038526228192, + "grad_norm": 5.228463172912598, + "learning_rate": 0.00015630084622792547, + "loss": 0.6945, + "step": 13216 + }, + { + "epoch": 0.30982729419762417, + "grad_norm": 3.24194073677063, + "learning_rate": 0.0001562947583057173, + "loss": 0.1714, + "step": 13217 + }, + { + "epoch": 0.30985073577242916, + "grad_norm": 2.728036880493164, + "learning_rate": 0.0001562886700780522, + "loss": 0.327, + "step": 13218 + }, + { + "epoch": 0.3098741773472342, + "grad_norm": 3.839594602584839, + "learning_rate": 0.0001562825815449632, + "loss": 0.3494, + "step": 13219 + }, + { + "epoch": 0.3098976189220392, + "grad_norm": 2.6294753551483154, + "learning_rate": 0.00015627649270648334, + "loss": 0.2248, + "step": 13220 + }, + { + "epoch": 0.3099210604968442, + "grad_norm": 2.491339683532715, + "learning_rate": 0.00015627040356264561, + "loss": 0.2369, + "step": 13221 + }, + { + "epoch": 0.3099445020716492, + "grad_norm": 2.5280685424804688, + "learning_rate": 0.00015626431411348314, + "loss": 0.1557, + "step": 13222 + }, + { + "epoch": 0.3099679436464542, + "grad_norm": 1.015594482421875, + "learning_rate": 0.00015625822435902888, + "loss": 0.0669, + "step": 13223 + }, + { + "epoch": 0.3099913852212592, + "grad_norm": 0.8718229532241821, + "learning_rate": 0.0001562521342993159, + "loss": 0.1152, + "step": 13224 + }, + { + "epoch": 0.3100148267960642, + "grad_norm": 2.386749029159546, + "learning_rate": 0.00015624604393437732, + "loss": 0.2499, + "step": 13225 + }, + { + "epoch": 0.3100382683708692, + "grad_norm": 4.081607818603516, + "learning_rate": 0.00015623995326424607, + "loss": 0.5173, + "step": 13226 + }, + { + "epoch": 0.31006170994567417, + "grad_norm": 3.376530408859253, + "learning_rate": 0.00015623386228895528, + "loss": 1.8559, + "step": 13227 + }, + { + "epoch": 0.31008515152047916, + "grad_norm": 3.218621015548706, + "learning_rate": 0.00015622777100853796, + "loss": 0.37, + "step": 13228 + }, + { + "epoch": 0.31010859309528416, + "grad_norm": 3.324063777923584, + "learning_rate": 0.00015622167942302718, + "loss": 0.4749, + "step": 13229 + }, + { + "epoch": 0.31013203467008915, + "grad_norm": 1.3355103731155396, + "learning_rate": 0.00015621558753245598, + "loss": 0.8399, + "step": 13230 + }, + { + "epoch": 0.31015547624489415, + "grad_norm": 4.141170024871826, + "learning_rate": 0.0001562094953368574, + "loss": 0.4454, + "step": 13231 + }, + { + "epoch": 0.31017891781969914, + "grad_norm": 4.584170818328857, + "learning_rate": 0.00015620340283626453, + "loss": 0.4914, + "step": 13232 + }, + { + "epoch": 0.31020235939450413, + "grad_norm": 3.1299846172332764, + "learning_rate": 0.00015619731003071045, + "loss": 0.3096, + "step": 13233 + }, + { + "epoch": 0.31022580096930913, + "grad_norm": 2.420320510864258, + "learning_rate": 0.00015619121692022817, + "loss": 0.2204, + "step": 13234 + }, + { + "epoch": 0.3102492425441141, + "grad_norm": 4.157581329345703, + "learning_rate": 0.00015618512350485075, + "loss": 0.2278, + "step": 13235 + }, + { + "epoch": 0.3102726841189191, + "grad_norm": 4.050528049468994, + "learning_rate": 0.00015617902978461127, + "loss": 0.4755, + "step": 13236 + }, + { + "epoch": 0.3102961256937241, + "grad_norm": 7.556407928466797, + "learning_rate": 0.00015617293575954282, + "loss": 0.6738, + "step": 13237 + }, + { + "epoch": 0.3103195672685291, + "grad_norm": 3.8518946170806885, + "learning_rate": 0.0001561668414296784, + "loss": 0.3636, + "step": 13238 + }, + { + "epoch": 0.3103430088433341, + "grad_norm": 3.4574928283691406, + "learning_rate": 0.00015616074679505117, + "loss": 0.4528, + "step": 13239 + }, + { + "epoch": 0.3103664504181391, + "grad_norm": 1.2386573553085327, + "learning_rate": 0.00015615465185569413, + "loss": 0.1406, + "step": 13240 + }, + { + "epoch": 0.3103898919929441, + "grad_norm": 4.523463726043701, + "learning_rate": 0.00015614855661164035, + "loss": 0.2629, + "step": 13241 + }, + { + "epoch": 0.3104133335677491, + "grad_norm": 1.4362441301345825, + "learning_rate": 0.00015614246106292293, + "loss": 0.1154, + "step": 13242 + }, + { + "epoch": 0.3104367751425541, + "grad_norm": 3.185777425765991, + "learning_rate": 0.00015613636520957495, + "loss": 0.3358, + "step": 13243 + }, + { + "epoch": 0.31046021671735907, + "grad_norm": 5.340826511383057, + "learning_rate": 0.00015613026905162946, + "loss": 0.3623, + "step": 13244 + }, + { + "epoch": 0.31048365829216407, + "grad_norm": 3.4842987060546875, + "learning_rate": 0.00015612417258911956, + "loss": 0.2594, + "step": 13245 + }, + { + "epoch": 0.31050709986696906, + "grad_norm": 2.833892822265625, + "learning_rate": 0.00015611807582207834, + "loss": 0.317, + "step": 13246 + }, + { + "epoch": 0.31053054144177406, + "grad_norm": 3.083203077316284, + "learning_rate": 0.00015611197875053882, + "loss": 0.3818, + "step": 13247 + }, + { + "epoch": 0.31055398301657905, + "grad_norm": 3.9721550941467285, + "learning_rate": 0.00015610588137453414, + "loss": 0.6688, + "step": 13248 + }, + { + "epoch": 0.31057742459138404, + "grad_norm": 1.7222124338150024, + "learning_rate": 0.0001560997836940974, + "loss": 0.2578, + "step": 13249 + }, + { + "epoch": 0.31060086616618904, + "grad_norm": 2.217811346054077, + "learning_rate": 0.00015609368570926164, + "loss": 0.1343, + "step": 13250 + }, + { + "epoch": 0.31062430774099403, + "grad_norm": 1.2781246900558472, + "learning_rate": 0.00015608758742005995, + "loss": 0.1472, + "step": 13251 + }, + { + "epoch": 0.310647749315799, + "grad_norm": 3.548696279525757, + "learning_rate": 0.00015608148882652545, + "loss": 0.3423, + "step": 13252 + }, + { + "epoch": 0.310671190890604, + "grad_norm": 0.42229798436164856, + "learning_rate": 0.0001560753899286912, + "loss": 0.0323, + "step": 13253 + }, + { + "epoch": 0.310694632465409, + "grad_norm": 4.742948532104492, + "learning_rate": 0.00015606929072659034, + "loss": 0.7086, + "step": 13254 + }, + { + "epoch": 0.310718074040214, + "grad_norm": 2.7925667762756348, + "learning_rate": 0.00015606319122025587, + "loss": 0.36, + "step": 13255 + }, + { + "epoch": 0.310741515615019, + "grad_norm": 2.4195427894592285, + "learning_rate": 0.000156057091409721, + "loss": 0.2637, + "step": 13256 + }, + { + "epoch": 0.310764957189824, + "grad_norm": 2.5286405086517334, + "learning_rate": 0.00015605099129501878, + "loss": 0.2408, + "step": 13257 + }, + { + "epoch": 0.310788398764629, + "grad_norm": 7.444258689880371, + "learning_rate": 0.00015604489087618227, + "loss": 0.6076, + "step": 13258 + }, + { + "epoch": 0.310811840339434, + "grad_norm": 3.0045042037963867, + "learning_rate": 0.00015603879015324462, + "loss": 0.1935, + "step": 13259 + }, + { + "epoch": 0.310835281914239, + "grad_norm": 4.278181552886963, + "learning_rate": 0.00015603268912623891, + "loss": 0.6837, + "step": 13260 + }, + { + "epoch": 0.310858723489044, + "grad_norm": 1.2511273622512817, + "learning_rate": 0.00015602658779519828, + "loss": 0.7391, + "step": 13261 + }, + { + "epoch": 0.31088216506384897, + "grad_norm": 3.3318822383880615, + "learning_rate": 0.0001560204861601558, + "loss": 0.2782, + "step": 13262 + }, + { + "epoch": 0.31090560663865396, + "grad_norm": 2.793731689453125, + "learning_rate": 0.0001560143842211446, + "loss": 0.2873, + "step": 13263 + }, + { + "epoch": 0.31092904821345896, + "grad_norm": 2.7389743328094482, + "learning_rate": 0.00015600828197819775, + "loss": 0.3749, + "step": 13264 + }, + { + "epoch": 0.31095248978826395, + "grad_norm": 2.378342866897583, + "learning_rate": 0.0001560021794313484, + "loss": 1.5055, + "step": 13265 + }, + { + "epoch": 0.31097593136306895, + "grad_norm": 4.664928913116455, + "learning_rate": 0.00015599607658062963, + "loss": 0.4463, + "step": 13266 + }, + { + "epoch": 0.31099937293787394, + "grad_norm": 4.909641742706299, + "learning_rate": 0.00015598997342607457, + "loss": 0.6602, + "step": 13267 + }, + { + "epoch": 0.31102281451267894, + "grad_norm": 1.0239156484603882, + "learning_rate": 0.00015598386996771638, + "loss": 0.1024, + "step": 13268 + }, + { + "epoch": 0.31104625608748393, + "grad_norm": 6.461612224578857, + "learning_rate": 0.00015597776620558814, + "loss": 1.2521, + "step": 13269 + }, + { + "epoch": 0.311069697662289, + "grad_norm": 3.7974724769592285, + "learning_rate": 0.0001559716621397229, + "loss": 0.3618, + "step": 13270 + }, + { + "epoch": 0.311093139237094, + "grad_norm": 4.667267322540283, + "learning_rate": 0.0001559655577701539, + "loss": 0.343, + "step": 13271 + }, + { + "epoch": 0.31111658081189897, + "grad_norm": 4.252705097198486, + "learning_rate": 0.0001559594530969142, + "loss": 0.4919, + "step": 13272 + }, + { + "epoch": 0.31114002238670396, + "grad_norm": 0.9296920299530029, + "learning_rate": 0.00015595334812003694, + "loss": 0.0927, + "step": 13273 + }, + { + "epoch": 0.31116346396150896, + "grad_norm": 3.693671703338623, + "learning_rate": 0.00015594724283955524, + "loss": 0.4582, + "step": 13274 + }, + { + "epoch": 0.31118690553631395, + "grad_norm": 1.6655436754226685, + "learning_rate": 0.0001559411372555022, + "loss": 0.2095, + "step": 13275 + }, + { + "epoch": 0.31121034711111895, + "grad_norm": 2.5857553482055664, + "learning_rate": 0.000155935031367911, + "loss": 1.7145, + "step": 13276 + }, + { + "epoch": 0.31123378868592394, + "grad_norm": 6.290258407592773, + "learning_rate": 0.00015592892517681477, + "loss": 0.7807, + "step": 13277 + }, + { + "epoch": 0.31125723026072893, + "grad_norm": 4.296854019165039, + "learning_rate": 0.00015592281868224655, + "loss": 0.565, + "step": 13278 + }, + { + "epoch": 0.31128067183553393, + "grad_norm": 3.0824732780456543, + "learning_rate": 0.00015591671188423963, + "loss": 0.2009, + "step": 13279 + }, + { + "epoch": 0.3113041134103389, + "grad_norm": 1.3508968353271484, + "learning_rate": 0.00015591060478282703, + "loss": 0.2857, + "step": 13280 + }, + { + "epoch": 0.3113275549851439, + "grad_norm": 4.425258636474609, + "learning_rate": 0.00015590449737804186, + "loss": 1.9218, + "step": 13281 + }, + { + "epoch": 0.3113509965599489, + "grad_norm": 4.88780403137207, + "learning_rate": 0.00015589838966991736, + "loss": 0.5486, + "step": 13282 + }, + { + "epoch": 0.3113744381347539, + "grad_norm": 1.9435473680496216, + "learning_rate": 0.00015589228165848663, + "loss": 1.4117, + "step": 13283 + }, + { + "epoch": 0.3113978797095589, + "grad_norm": 2.857465982437134, + "learning_rate": 0.00015588617334378281, + "loss": 0.3727, + "step": 13284 + }, + { + "epoch": 0.3114213212843639, + "grad_norm": 7.021263599395752, + "learning_rate": 0.00015588006472583903, + "loss": 0.4833, + "step": 13285 + }, + { + "epoch": 0.3114447628591689, + "grad_norm": 0.9047451615333557, + "learning_rate": 0.00015587395580468844, + "loss": 0.0776, + "step": 13286 + }, + { + "epoch": 0.3114682044339739, + "grad_norm": 5.34367036819458, + "learning_rate": 0.00015586784658036423, + "loss": 0.5856, + "step": 13287 + }, + { + "epoch": 0.3114916460087789, + "grad_norm": 4.093212604522705, + "learning_rate": 0.0001558617370528995, + "loss": 0.4794, + "step": 13288 + }, + { + "epoch": 0.31151508758358387, + "grad_norm": 1.353659749031067, + "learning_rate": 0.0001558556272223274, + "loss": 0.2035, + "step": 13289 + }, + { + "epoch": 0.31153852915838887, + "grad_norm": 4.522278785705566, + "learning_rate": 0.00015584951708868107, + "loss": 0.4083, + "step": 13290 + }, + { + "epoch": 0.31156197073319386, + "grad_norm": 2.0594534873962402, + "learning_rate": 0.00015584340665199373, + "loss": 0.1373, + "step": 13291 + }, + { + "epoch": 0.31158541230799885, + "grad_norm": 3.828216314315796, + "learning_rate": 0.0001558372959122985, + "loss": 0.481, + "step": 13292 + }, + { + "epoch": 0.31160885388280385, + "grad_norm": 2.1761348247528076, + "learning_rate": 0.00015583118486962852, + "loss": 0.3035, + "step": 13293 + }, + { + "epoch": 0.31163229545760884, + "grad_norm": 1.2470581531524658, + "learning_rate": 0.00015582507352401693, + "loss": 0.1611, + "step": 13294 + }, + { + "epoch": 0.31165573703241384, + "grad_norm": 3.9042446613311768, + "learning_rate": 0.00015581896187549695, + "loss": 0.5109, + "step": 13295 + }, + { + "epoch": 0.31167917860721883, + "grad_norm": 4.2009100914001465, + "learning_rate": 0.00015581284992410174, + "loss": 0.3453, + "step": 13296 + }, + { + "epoch": 0.3117026201820238, + "grad_norm": 3.7656753063201904, + "learning_rate": 0.00015580673766986443, + "loss": 0.6033, + "step": 13297 + }, + { + "epoch": 0.3117260617568288, + "grad_norm": 2.6140756607055664, + "learning_rate": 0.00015580062511281816, + "loss": 0.2265, + "step": 13298 + }, + { + "epoch": 0.3117495033316338, + "grad_norm": 2.440180778503418, + "learning_rate": 0.00015579451225299616, + "loss": 0.2075, + "step": 13299 + }, + { + "epoch": 0.3117729449064388, + "grad_norm": 1.913957953453064, + "learning_rate": 0.0001557883990904315, + "loss": 0.31, + "step": 13300 + }, + { + "epoch": 0.3117963864812438, + "grad_norm": 1.5304763317108154, + "learning_rate": 0.00015578228562515748, + "loss": 0.2977, + "step": 13301 + }, + { + "epoch": 0.3118198280560488, + "grad_norm": 2.2308876514434814, + "learning_rate": 0.0001557761718572072, + "loss": 0.3385, + "step": 13302 + }, + { + "epoch": 0.3118432696308538, + "grad_norm": 3.329040765762329, + "learning_rate": 0.00015577005778661385, + "loss": 0.3722, + "step": 13303 + }, + { + "epoch": 0.3118667112056588, + "grad_norm": 3.3437089920043945, + "learning_rate": 0.00015576394341341058, + "loss": 0.3702, + "step": 13304 + }, + { + "epoch": 0.3118901527804638, + "grad_norm": 0.987066924571991, + "learning_rate": 0.0001557578287376306, + "loss": 0.0974, + "step": 13305 + }, + { + "epoch": 0.3119135943552688, + "grad_norm": 2.904384136199951, + "learning_rate": 0.0001557517137593071, + "loss": 0.17, + "step": 13306 + }, + { + "epoch": 0.31193703593007377, + "grad_norm": 2.5352466106414795, + "learning_rate": 0.00015574559847847318, + "loss": 0.1135, + "step": 13307 + }, + { + "epoch": 0.31196047750487876, + "grad_norm": 3.926934242248535, + "learning_rate": 0.0001557394828951621, + "loss": 0.2929, + "step": 13308 + }, + { + "epoch": 0.31198391907968376, + "grad_norm": 3.6315841674804688, + "learning_rate": 0.000155733367009407, + "loss": 0.4068, + "step": 13309 + }, + { + "epoch": 0.31200736065448875, + "grad_norm": 3.2139475345611572, + "learning_rate": 0.00015572725082124112, + "loss": 1.4025, + "step": 13310 + }, + { + "epoch": 0.31203080222929375, + "grad_norm": 3.8446452617645264, + "learning_rate": 0.0001557211343306976, + "loss": 0.4664, + "step": 13311 + }, + { + "epoch": 0.31205424380409874, + "grad_norm": 2.54649019241333, + "learning_rate": 0.00015571501753780962, + "loss": 0.1481, + "step": 13312 + }, + { + "epoch": 0.31207768537890374, + "grad_norm": 4.136443138122559, + "learning_rate": 0.0001557089004426104, + "loss": 0.524, + "step": 13313 + }, + { + "epoch": 0.31210112695370873, + "grad_norm": 3.3364198207855225, + "learning_rate": 0.00015570278304513307, + "loss": 0.4672, + "step": 13314 + }, + { + "epoch": 0.3121245685285137, + "grad_norm": 2.4503579139709473, + "learning_rate": 0.00015569666534541094, + "loss": 0.2089, + "step": 13315 + }, + { + "epoch": 0.3121480101033187, + "grad_norm": 3.6165812015533447, + "learning_rate": 0.00015569054734347712, + "loss": 0.3715, + "step": 13316 + }, + { + "epoch": 0.3121714516781237, + "grad_norm": 3.3952689170837402, + "learning_rate": 0.0001556844290393648, + "loss": 0.2982, + "step": 13317 + }, + { + "epoch": 0.3121948932529287, + "grad_norm": 4.701176166534424, + "learning_rate": 0.00015567831043310725, + "loss": 0.76, + "step": 13318 + }, + { + "epoch": 0.3122183348277337, + "grad_norm": 4.095539093017578, + "learning_rate": 0.00015567219152473757, + "loss": 0.5178, + "step": 13319 + }, + { + "epoch": 0.3122417764025387, + "grad_norm": 2.449740409851074, + "learning_rate": 0.00015566607231428905, + "loss": 0.2024, + "step": 13320 + }, + { + "epoch": 0.31226521797734375, + "grad_norm": 3.4223711490631104, + "learning_rate": 0.00015565995280179484, + "loss": 0.343, + "step": 13321 + }, + { + "epoch": 0.31228865955214874, + "grad_norm": 3.1235575675964355, + "learning_rate": 0.00015565383298728818, + "loss": 0.348, + "step": 13322 + }, + { + "epoch": 0.31231210112695373, + "grad_norm": 3.2884299755096436, + "learning_rate": 0.00015564771287080223, + "loss": 0.339, + "step": 13323 + }, + { + "epoch": 0.31233554270175873, + "grad_norm": 3.308687448501587, + "learning_rate": 0.00015564159245237024, + "loss": 0.3181, + "step": 13324 + }, + { + "epoch": 0.3123589842765637, + "grad_norm": 3.6956589221954346, + "learning_rate": 0.00015563547173202543, + "loss": 0.3951, + "step": 13325 + }, + { + "epoch": 0.3123824258513687, + "grad_norm": 5.799474239349365, + "learning_rate": 0.00015562935070980092, + "loss": 0.7703, + "step": 13326 + }, + { + "epoch": 0.3124058674261737, + "grad_norm": 5.0654826164245605, + "learning_rate": 0.00015562322938573005, + "loss": 0.3817, + "step": 13327 + }, + { + "epoch": 0.3124293090009787, + "grad_norm": 4.638621807098389, + "learning_rate": 0.00015561710775984595, + "loss": 0.4376, + "step": 13328 + }, + { + "epoch": 0.3124527505757837, + "grad_norm": 4.649801731109619, + "learning_rate": 0.00015561098583218185, + "loss": 0.4563, + "step": 13329 + }, + { + "epoch": 0.3124761921505887, + "grad_norm": 2.8644063472747803, + "learning_rate": 0.000155604863602771, + "loss": 1.2958, + "step": 13330 + }, + { + "epoch": 0.3124996337253937, + "grad_norm": 3.8756301403045654, + "learning_rate": 0.00015559874107164658, + "loss": 0.3758, + "step": 13331 + }, + { + "epoch": 0.3125230753001987, + "grad_norm": 2.219435691833496, + "learning_rate": 0.00015559261823884182, + "loss": 0.1832, + "step": 13332 + }, + { + "epoch": 0.3125465168750037, + "grad_norm": 1.2792448997497559, + "learning_rate": 0.00015558649510438998, + "loss": 0.0925, + "step": 13333 + }, + { + "epoch": 0.31256995844980867, + "grad_norm": 3.3689002990722656, + "learning_rate": 0.0001555803716683242, + "loss": 0.3381, + "step": 13334 + }, + { + "epoch": 0.31259340002461367, + "grad_norm": 1.306784987449646, + "learning_rate": 0.00015557424793067782, + "loss": 0.1512, + "step": 13335 + }, + { + "epoch": 0.31261684159941866, + "grad_norm": 1.577097773551941, + "learning_rate": 0.000155568123891484, + "loss": 0.1552, + "step": 13336 + }, + { + "epoch": 0.31264028317422365, + "grad_norm": 0.7717241048812866, + "learning_rate": 0.00015556199955077594, + "loss": 0.0773, + "step": 13337 + }, + { + "epoch": 0.31266372474902865, + "grad_norm": 4.448779106140137, + "learning_rate": 0.00015555587490858694, + "loss": 0.5426, + "step": 13338 + }, + { + "epoch": 0.31268716632383364, + "grad_norm": 4.256948947906494, + "learning_rate": 0.0001555497499649502, + "loss": 0.4336, + "step": 13339 + }, + { + "epoch": 0.31271060789863864, + "grad_norm": 1.3347671031951904, + "learning_rate": 0.00015554362471989895, + "loss": 0.1741, + "step": 13340 + }, + { + "epoch": 0.31273404947344363, + "grad_norm": 1.4072208404541016, + "learning_rate": 0.00015553749917346642, + "loss": 0.1735, + "step": 13341 + }, + { + "epoch": 0.3127574910482486, + "grad_norm": 2.8973116874694824, + "learning_rate": 0.0001555313733256859, + "loss": 0.3059, + "step": 13342 + }, + { + "epoch": 0.3127809326230536, + "grad_norm": 2.28214693069458, + "learning_rate": 0.00015552524717659053, + "loss": 0.1366, + "step": 13343 + }, + { + "epoch": 0.3128043741978586, + "grad_norm": 0.34427210688591003, + "learning_rate": 0.00015551912072621363, + "loss": 0.0226, + "step": 13344 + }, + { + "epoch": 0.3128278157726636, + "grad_norm": 7.255062103271484, + "learning_rate": 0.0001555129939745884, + "loss": 0.4666, + "step": 13345 + }, + { + "epoch": 0.3128512573474686, + "grad_norm": 0.7491086721420288, + "learning_rate": 0.00015550686692174815, + "loss": 0.0522, + "step": 13346 + }, + { + "epoch": 0.3128746989222736, + "grad_norm": 4.065040111541748, + "learning_rate": 0.00015550073956772603, + "loss": 0.3823, + "step": 13347 + }, + { + "epoch": 0.3128981404970786, + "grad_norm": 2.97971510887146, + "learning_rate": 0.00015549461191255537, + "loss": 0.2926, + "step": 13348 + }, + { + "epoch": 0.3129215820718836, + "grad_norm": 3.173257350921631, + "learning_rate": 0.00015548848395626939, + "loss": 0.2188, + "step": 13349 + }, + { + "epoch": 0.3129450236466886, + "grad_norm": 3.094327688217163, + "learning_rate": 0.0001554823556989013, + "loss": 0.3664, + "step": 13350 + }, + { + "epoch": 0.3129684652214936, + "grad_norm": 2.4149439334869385, + "learning_rate": 0.0001554762271404844, + "loss": 1.7051, + "step": 13351 + }, + { + "epoch": 0.31299190679629857, + "grad_norm": 0.588941216468811, + "learning_rate": 0.00015547009828105197, + "loss": 0.099, + "step": 13352 + }, + { + "epoch": 0.31301534837110356, + "grad_norm": 3.3144755363464355, + "learning_rate": 0.0001554639691206372, + "loss": 0.2873, + "step": 13353 + }, + { + "epoch": 0.31303878994590856, + "grad_norm": 3.776721239089966, + "learning_rate": 0.00015545783965927336, + "loss": 0.4015, + "step": 13354 + }, + { + "epoch": 0.31306223152071355, + "grad_norm": 4.415811061859131, + "learning_rate": 0.00015545170989699373, + "loss": 0.6537, + "step": 13355 + }, + { + "epoch": 0.31308567309551855, + "grad_norm": 3.904247760772705, + "learning_rate": 0.0001554455798338316, + "loss": 0.3665, + "step": 13356 + }, + { + "epoch": 0.31310911467032354, + "grad_norm": 3.491992235183716, + "learning_rate": 0.00015543944946982014, + "loss": 0.4676, + "step": 13357 + }, + { + "epoch": 0.31313255624512853, + "grad_norm": 1.260392427444458, + "learning_rate": 0.00015543331880499267, + "loss": 0.0972, + "step": 13358 + }, + { + "epoch": 0.31315599781993353, + "grad_norm": 1.7764230966567993, + "learning_rate": 0.00015542718783938248, + "loss": 0.1553, + "step": 13359 + }, + { + "epoch": 0.3131794393947385, + "grad_norm": 4.410138130187988, + "learning_rate": 0.0001554210565730228, + "loss": 0.3815, + "step": 13360 + }, + { + "epoch": 0.3132028809695435, + "grad_norm": 2.889234781265259, + "learning_rate": 0.00015541492500594693, + "loss": 0.3639, + "step": 13361 + }, + { + "epoch": 0.3132263225443485, + "grad_norm": 3.088320016860962, + "learning_rate": 0.0001554087931381881, + "loss": 0.3493, + "step": 13362 + }, + { + "epoch": 0.3132497641191535, + "grad_norm": 3.7171413898468018, + "learning_rate": 0.00015540266096977957, + "loss": 0.4058, + "step": 13363 + }, + { + "epoch": 0.3132732056939585, + "grad_norm": 4.070467472076416, + "learning_rate": 0.0001553965285007547, + "loss": 0.4639, + "step": 13364 + }, + { + "epoch": 0.3132966472687635, + "grad_norm": 2.3032264709472656, + "learning_rate": 0.00015539039573114668, + "loss": 0.3785, + "step": 13365 + }, + { + "epoch": 0.3133200888435685, + "grad_norm": 3.948943614959717, + "learning_rate": 0.00015538426266098883, + "loss": 0.4659, + "step": 13366 + }, + { + "epoch": 0.3133435304183735, + "grad_norm": 1.0715699195861816, + "learning_rate": 0.0001553781292903144, + "loss": 0.0265, + "step": 13367 + }, + { + "epoch": 0.3133669719931785, + "grad_norm": 3.320490598678589, + "learning_rate": 0.00015537199561915669, + "loss": 0.3187, + "step": 13368 + }, + { + "epoch": 0.3133904135679835, + "grad_norm": 5.728999614715576, + "learning_rate": 0.00015536586164754897, + "loss": 0.4782, + "step": 13369 + }, + { + "epoch": 0.31341385514278847, + "grad_norm": 3.6733105182647705, + "learning_rate": 0.00015535972737552452, + "loss": 0.6839, + "step": 13370 + }, + { + "epoch": 0.31343729671759346, + "grad_norm": 2.5784912109375, + "learning_rate": 0.00015535359280311665, + "loss": 0.2286, + "step": 13371 + }, + { + "epoch": 0.3134607382923985, + "grad_norm": 3.810098886489868, + "learning_rate": 0.0001553474579303586, + "loss": 0.5754, + "step": 13372 + }, + { + "epoch": 0.3134841798672035, + "grad_norm": 2.724341630935669, + "learning_rate": 0.00015534132275728374, + "loss": 0.1989, + "step": 13373 + }, + { + "epoch": 0.3135076214420085, + "grad_norm": 2.419018030166626, + "learning_rate": 0.00015533518728392526, + "loss": 0.2164, + "step": 13374 + }, + { + "epoch": 0.3135310630168135, + "grad_norm": 2.7969250679016113, + "learning_rate": 0.00015532905151031652, + "loss": 0.1702, + "step": 13375 + }, + { + "epoch": 0.3135545045916185, + "grad_norm": 4.3417253494262695, + "learning_rate": 0.00015532291543649077, + "loss": 0.3276, + "step": 13376 + }, + { + "epoch": 0.3135779461664235, + "grad_norm": 2.075204849243164, + "learning_rate": 0.00015531677906248134, + "loss": 0.2433, + "step": 13377 + }, + { + "epoch": 0.3136013877412285, + "grad_norm": 0.773219108581543, + "learning_rate": 0.00015531064238832152, + "loss": 0.0516, + "step": 13378 + }, + { + "epoch": 0.31362482931603347, + "grad_norm": 3.4976308345794678, + "learning_rate": 0.0001553045054140446, + "loss": 0.3938, + "step": 13379 + }, + { + "epoch": 0.31364827089083847, + "grad_norm": 0.8577802777290344, + "learning_rate": 0.00015529836813968386, + "loss": 0.0785, + "step": 13380 + }, + { + "epoch": 0.31367171246564346, + "grad_norm": 3.176140308380127, + "learning_rate": 0.00015529223056527262, + "loss": 0.186, + "step": 13381 + }, + { + "epoch": 0.31369515404044845, + "grad_norm": 2.4292922019958496, + "learning_rate": 0.0001552860926908442, + "loss": 0.2395, + "step": 13382 + }, + { + "epoch": 0.31371859561525345, + "grad_norm": 4.030838489532471, + "learning_rate": 0.00015527995451643188, + "loss": 0.6548, + "step": 13383 + }, + { + "epoch": 0.31374203719005844, + "grad_norm": 2.2016446590423584, + "learning_rate": 0.00015527381604206898, + "loss": 0.1041, + "step": 13384 + }, + { + "epoch": 0.31376547876486344, + "grad_norm": 3.466881036758423, + "learning_rate": 0.00015526767726778878, + "loss": 0.2401, + "step": 13385 + }, + { + "epoch": 0.31378892033966843, + "grad_norm": 2.662975549697876, + "learning_rate": 0.0001552615381936246, + "loss": 0.1264, + "step": 13386 + }, + { + "epoch": 0.3138123619144734, + "grad_norm": 1.9436272382736206, + "learning_rate": 0.0001552553988196098, + "loss": 1.5459, + "step": 13387 + }, + { + "epoch": 0.3138358034892784, + "grad_norm": 2.79659366607666, + "learning_rate": 0.0001552492591457776, + "loss": 0.1814, + "step": 13388 + }, + { + "epoch": 0.3138592450640834, + "grad_norm": 0.7134408950805664, + "learning_rate": 0.0001552431191721614, + "loss": 0.0489, + "step": 13389 + }, + { + "epoch": 0.3138826866388884, + "grad_norm": 4.816735744476318, + "learning_rate": 0.0001552369788987945, + "loss": 0.4102, + "step": 13390 + }, + { + "epoch": 0.3139061282136934, + "grad_norm": 3.2244529724121094, + "learning_rate": 0.00015523083832571016, + "loss": 0.242, + "step": 13391 + }, + { + "epoch": 0.3139295697884984, + "grad_norm": 1.5618159770965576, + "learning_rate": 0.00015522469745294176, + "loss": 0.1632, + "step": 13392 + }, + { + "epoch": 0.3139530113633034, + "grad_norm": 3.5497758388519287, + "learning_rate": 0.00015521855628052259, + "loss": 0.3272, + "step": 13393 + }, + { + "epoch": 0.3139764529381084, + "grad_norm": 4.606188774108887, + "learning_rate": 0.000155212414808486, + "loss": 0.7531, + "step": 13394 + }, + { + "epoch": 0.3139998945129134, + "grad_norm": 4.519540786743164, + "learning_rate": 0.00015520627303686527, + "loss": 0.653, + "step": 13395 + }, + { + "epoch": 0.3140233360877184, + "grad_norm": 2.9801135063171387, + "learning_rate": 0.00015520013096569378, + "loss": 0.2915, + "step": 13396 + }, + { + "epoch": 0.31404677766252337, + "grad_norm": 3.089778423309326, + "learning_rate": 0.0001551939885950048, + "loss": 0.3943, + "step": 13397 + }, + { + "epoch": 0.31407021923732836, + "grad_norm": 4.288496017456055, + "learning_rate": 0.00015518784592483172, + "loss": 0.4365, + "step": 13398 + }, + { + "epoch": 0.31409366081213336, + "grad_norm": 2.5169620513916016, + "learning_rate": 0.00015518170295520777, + "loss": 0.3446, + "step": 13399 + }, + { + "epoch": 0.31411710238693835, + "grad_norm": 2.3969926834106445, + "learning_rate": 0.0001551755596861664, + "loss": 0.1478, + "step": 13400 + }, + { + "epoch": 0.31414054396174335, + "grad_norm": 1.1785073280334473, + "learning_rate": 0.00015516941611774088, + "loss": 0.1162, + "step": 13401 + }, + { + "epoch": 0.31416398553654834, + "grad_norm": 4.621003150939941, + "learning_rate": 0.00015516327224996456, + "loss": 0.137, + "step": 13402 + }, + { + "epoch": 0.31418742711135333, + "grad_norm": 4.282768726348877, + "learning_rate": 0.00015515712808287079, + "loss": 0.3788, + "step": 13403 + }, + { + "epoch": 0.31421086868615833, + "grad_norm": 2.601290464401245, + "learning_rate": 0.00015515098361649287, + "loss": 0.156, + "step": 13404 + }, + { + "epoch": 0.3142343102609633, + "grad_norm": 3.7186179161071777, + "learning_rate": 0.00015514483885086414, + "loss": 0.5208, + "step": 13405 + }, + { + "epoch": 0.3142577518357683, + "grad_norm": 3.0331056118011475, + "learning_rate": 0.000155138693786018, + "loss": 0.2676, + "step": 13406 + }, + { + "epoch": 0.3142811934105733, + "grad_norm": 4.566150188446045, + "learning_rate": 0.00015513254842198775, + "loss": 0.5407, + "step": 13407 + }, + { + "epoch": 0.3143046349853783, + "grad_norm": 2.6656694412231445, + "learning_rate": 0.00015512640275880675, + "loss": 0.3946, + "step": 13408 + }, + { + "epoch": 0.3143280765601833, + "grad_norm": 1.3218823671340942, + "learning_rate": 0.00015512025679650834, + "loss": 0.2082, + "step": 13409 + }, + { + "epoch": 0.3143515181349883, + "grad_norm": 3.6735994815826416, + "learning_rate": 0.00015511411053512584, + "loss": 0.2812, + "step": 13410 + }, + { + "epoch": 0.3143749597097933, + "grad_norm": 1.7046817541122437, + "learning_rate": 0.00015510796397469264, + "loss": 0.1506, + "step": 13411 + }, + { + "epoch": 0.3143984012845983, + "grad_norm": 1.6060582399368286, + "learning_rate": 0.00015510181711524208, + "loss": 0.2264, + "step": 13412 + }, + { + "epoch": 0.3144218428594033, + "grad_norm": 4.453358173370361, + "learning_rate": 0.00015509566995680748, + "loss": 0.5428, + "step": 13413 + }, + { + "epoch": 0.31444528443420827, + "grad_norm": 2.3698532581329346, + "learning_rate": 0.0001550895224994223, + "loss": 0.9977, + "step": 13414 + }, + { + "epoch": 0.31446872600901327, + "grad_norm": 1.5163949728012085, + "learning_rate": 0.0001550833747431198, + "loss": 0.1709, + "step": 13415 + }, + { + "epoch": 0.31449216758381826, + "grad_norm": 7.262165546417236, + "learning_rate": 0.0001550772266879333, + "loss": 0.4232, + "step": 13416 + }, + { + "epoch": 0.31451560915862325, + "grad_norm": 1.7686796188354492, + "learning_rate": 0.00015507107833389623, + "loss": 0.265, + "step": 13417 + }, + { + "epoch": 0.31453905073342825, + "grad_norm": 2.8163366317749023, + "learning_rate": 0.00015506492968104198, + "loss": 0.2925, + "step": 13418 + }, + { + "epoch": 0.31456249230823324, + "grad_norm": 1.2123401165008545, + "learning_rate": 0.00015505878072940386, + "loss": 0.1403, + "step": 13419 + }, + { + "epoch": 0.31458593388303824, + "grad_norm": 3.0275912284851074, + "learning_rate": 0.00015505263147901527, + "loss": 1.2041, + "step": 13420 + }, + { + "epoch": 0.31460937545784323, + "grad_norm": 0.7542906999588013, + "learning_rate": 0.00015504648192990952, + "loss": 0.0589, + "step": 13421 + }, + { + "epoch": 0.3146328170326482, + "grad_norm": 4.5258612632751465, + "learning_rate": 0.00015504033208212006, + "loss": 0.5675, + "step": 13422 + }, + { + "epoch": 0.3146562586074533, + "grad_norm": 2.2908716201782227, + "learning_rate": 0.00015503418193568014, + "loss": 0.1847, + "step": 13423 + }, + { + "epoch": 0.31467970018225827, + "grad_norm": 4.297276020050049, + "learning_rate": 0.00015502803149062324, + "loss": 0.4075, + "step": 13424 + }, + { + "epoch": 0.31470314175706326, + "grad_norm": 1.056183934211731, + "learning_rate": 0.00015502188074698268, + "loss": 0.0769, + "step": 13425 + }, + { + "epoch": 0.31472658333186826, + "grad_norm": 0.7166838049888611, + "learning_rate": 0.0001550157297047919, + "loss": 0.0941, + "step": 13426 + }, + { + "epoch": 0.31475002490667325, + "grad_norm": 3.7399678230285645, + "learning_rate": 0.00015500957836408414, + "loss": 0.4343, + "step": 13427 + }, + { + "epoch": 0.31477346648147825, + "grad_norm": 3.788792371749878, + "learning_rate": 0.0001550034267248929, + "loss": 0.4159, + "step": 13428 + }, + { + "epoch": 0.31479690805628324, + "grad_norm": 3.731241464614868, + "learning_rate": 0.00015499727478725153, + "loss": 0.5493, + "step": 13429 + }, + { + "epoch": 0.31482034963108824, + "grad_norm": 5.385744094848633, + "learning_rate": 0.0001549911225511934, + "loss": 0.7494, + "step": 13430 + }, + { + "epoch": 0.31484379120589323, + "grad_norm": 1.5155560970306396, + "learning_rate": 0.00015498497001675186, + "loss": 0.1447, + "step": 13431 + }, + { + "epoch": 0.3148672327806982, + "grad_norm": 2.4077792167663574, + "learning_rate": 0.00015497881718396037, + "loss": 0.272, + "step": 13432 + }, + { + "epoch": 0.3148906743555032, + "grad_norm": 4.3817596435546875, + "learning_rate": 0.00015497266405285225, + "loss": 0.3913, + "step": 13433 + }, + { + "epoch": 0.3149141159303082, + "grad_norm": 2.9987525939941406, + "learning_rate": 0.00015496651062346092, + "loss": 0.378, + "step": 13434 + }, + { + "epoch": 0.3149375575051132, + "grad_norm": 11.854973793029785, + "learning_rate": 0.00015496035689581975, + "loss": 0.4562, + "step": 13435 + }, + { + "epoch": 0.3149609990799182, + "grad_norm": 3.0544168949127197, + "learning_rate": 0.00015495420286996213, + "loss": 0.418, + "step": 13436 + }, + { + "epoch": 0.3149844406547232, + "grad_norm": 2.890888214111328, + "learning_rate": 0.00015494804854592145, + "loss": 0.3454, + "step": 13437 + }, + { + "epoch": 0.3150078822295282, + "grad_norm": 1.4288899898529053, + "learning_rate": 0.00015494189392373117, + "loss": 0.2002, + "step": 13438 + }, + { + "epoch": 0.3150313238043332, + "grad_norm": 0.35454145073890686, + "learning_rate": 0.0001549357390034246, + "loss": 0.0299, + "step": 13439 + }, + { + "epoch": 0.3150547653791382, + "grad_norm": 1.6267485618591309, + "learning_rate": 0.00015492958378503512, + "loss": 0.1951, + "step": 13440 + }, + { + "epoch": 0.3150782069539432, + "grad_norm": 1.1008620262145996, + "learning_rate": 0.00015492342826859624, + "loss": 0.1737, + "step": 13441 + }, + { + "epoch": 0.31510164852874817, + "grad_norm": 2.6258418560028076, + "learning_rate": 0.00015491727245414125, + "loss": 0.3827, + "step": 13442 + }, + { + "epoch": 0.31512509010355316, + "grad_norm": 1.5840694904327393, + "learning_rate": 0.0001549111163417036, + "loss": 0.1475, + "step": 13443 + }, + { + "epoch": 0.31514853167835816, + "grad_norm": 9.12407112121582, + "learning_rate": 0.00015490495993131673, + "loss": 0.505, + "step": 13444 + }, + { + "epoch": 0.31517197325316315, + "grad_norm": 1.185088872909546, + "learning_rate": 0.00015489880322301395, + "loss": 0.1294, + "step": 13445 + }, + { + "epoch": 0.31519541482796815, + "grad_norm": 3.519721746444702, + "learning_rate": 0.00015489264621682875, + "loss": 0.4146, + "step": 13446 + }, + { + "epoch": 0.31521885640277314, + "grad_norm": 1.0985825061798096, + "learning_rate": 0.00015488648891279452, + "loss": 0.1739, + "step": 13447 + }, + { + "epoch": 0.31524229797757813, + "grad_norm": 0.4710003137588501, + "learning_rate": 0.00015488033131094462, + "loss": 0.0656, + "step": 13448 + }, + { + "epoch": 0.31526573955238313, + "grad_norm": 2.18000864982605, + "learning_rate": 0.0001548741734113125, + "loss": 0.2586, + "step": 13449 + }, + { + "epoch": 0.3152891811271881, + "grad_norm": 2.52321195602417, + "learning_rate": 0.0001548680152139316, + "loss": 0.2244, + "step": 13450 + }, + { + "epoch": 0.3153126227019931, + "grad_norm": 4.389456272125244, + "learning_rate": 0.0001548618567188353, + "loss": 0.2947, + "step": 13451 + }, + { + "epoch": 0.3153360642767981, + "grad_norm": 2.2608492374420166, + "learning_rate": 0.00015485569792605703, + "loss": 0.2178, + "step": 13452 + }, + { + "epoch": 0.3153595058516031, + "grad_norm": 3.7407379150390625, + "learning_rate": 0.00015484953883563018, + "loss": 0.4912, + "step": 13453 + }, + { + "epoch": 0.3153829474264081, + "grad_norm": 4.176374435424805, + "learning_rate": 0.0001548433794475882, + "loss": 0.5015, + "step": 13454 + }, + { + "epoch": 0.3154063890012131, + "grad_norm": 3.926887035369873, + "learning_rate": 0.0001548372197619645, + "loss": 0.425, + "step": 13455 + }, + { + "epoch": 0.3154298305760181, + "grad_norm": 2.5447447299957275, + "learning_rate": 0.0001548310597787925, + "loss": 0.3488, + "step": 13456 + }, + { + "epoch": 0.3154532721508231, + "grad_norm": 2.0904953479766846, + "learning_rate": 0.00015482489949810563, + "loss": 0.2453, + "step": 13457 + }, + { + "epoch": 0.3154767137256281, + "grad_norm": 0.9159684777259827, + "learning_rate": 0.00015481873891993732, + "loss": 0.0843, + "step": 13458 + }, + { + "epoch": 0.31550015530043307, + "grad_norm": 3.0920183658599854, + "learning_rate": 0.00015481257804432098, + "loss": 0.3533, + "step": 13459 + }, + { + "epoch": 0.31552359687523807, + "grad_norm": 3.6190268993377686, + "learning_rate": 0.00015480641687129007, + "loss": 0.3949, + "step": 13460 + }, + { + "epoch": 0.31554703845004306, + "grad_norm": 3.0879812240600586, + "learning_rate": 0.00015480025540087799, + "loss": 0.2802, + "step": 13461 + }, + { + "epoch": 0.31557048002484805, + "grad_norm": 5.099351406097412, + "learning_rate": 0.00015479409363311822, + "loss": 0.6171, + "step": 13462 + }, + { + "epoch": 0.31559392159965305, + "grad_norm": 1.4696718454360962, + "learning_rate": 0.0001547879315680441, + "loss": 0.149, + "step": 13463 + }, + { + "epoch": 0.31561736317445804, + "grad_norm": 4.045238494873047, + "learning_rate": 0.00015478176920568916, + "loss": 0.4039, + "step": 13464 + }, + { + "epoch": 0.31564080474926304, + "grad_norm": 1.3815112113952637, + "learning_rate": 0.00015477560654608675, + "loss": 0.1595, + "step": 13465 + }, + { + "epoch": 0.31566424632406803, + "grad_norm": 0.632976770401001, + "learning_rate": 0.0001547694435892704, + "loss": 0.0366, + "step": 13466 + }, + { + "epoch": 0.315687687898873, + "grad_norm": 6.339687347412109, + "learning_rate": 0.0001547632803352735, + "loss": 0.7157, + "step": 13467 + }, + { + "epoch": 0.315711129473678, + "grad_norm": 2.440159320831299, + "learning_rate": 0.0001547571167841295, + "loss": 1.6896, + "step": 13468 + }, + { + "epoch": 0.315734571048483, + "grad_norm": 1.4732213020324707, + "learning_rate": 0.00015475095293587186, + "loss": 0.1704, + "step": 13469 + }, + { + "epoch": 0.315758012623288, + "grad_norm": 5.275175094604492, + "learning_rate": 0.000154744788790534, + "loss": 0.4853, + "step": 13470 + }, + { + "epoch": 0.315781454198093, + "grad_norm": 3.292440176010132, + "learning_rate": 0.00015473862434814937, + "loss": 0.282, + "step": 13471 + }, + { + "epoch": 0.315804895772898, + "grad_norm": 3.3033759593963623, + "learning_rate": 0.00015473245960875143, + "loss": 0.2201, + "step": 13472 + }, + { + "epoch": 0.315828337347703, + "grad_norm": 5.068772792816162, + "learning_rate": 0.00015472629457237362, + "loss": 0.396, + "step": 13473 + }, + { + "epoch": 0.315851778922508, + "grad_norm": 2.0226187705993652, + "learning_rate": 0.0001547201292390494, + "loss": 0.336, + "step": 13474 + }, + { + "epoch": 0.31587522049731304, + "grad_norm": 2.8392844200134277, + "learning_rate": 0.00015471396360881222, + "loss": 0.3805, + "step": 13475 + }, + { + "epoch": 0.31589866207211803, + "grad_norm": 4.908032417297363, + "learning_rate": 0.0001547077976816955, + "loss": 0.2815, + "step": 13476 + }, + { + "epoch": 0.315922103646923, + "grad_norm": 3.8345589637756348, + "learning_rate": 0.0001547016314577328, + "loss": 0.4501, + "step": 13477 + }, + { + "epoch": 0.315945545221728, + "grad_norm": 2.880880355834961, + "learning_rate": 0.00015469546493695745, + "loss": 0.1528, + "step": 13478 + }, + { + "epoch": 0.315968986796533, + "grad_norm": 3.505460739135742, + "learning_rate": 0.00015468929811940303, + "loss": 0.3035, + "step": 13479 + }, + { + "epoch": 0.315992428371338, + "grad_norm": 0.743166446685791, + "learning_rate": 0.0001546831310051029, + "loss": 0.1256, + "step": 13480 + }, + { + "epoch": 0.316015869946143, + "grad_norm": 3.557039976119995, + "learning_rate": 0.00015467696359409055, + "loss": 0.5004, + "step": 13481 + }, + { + "epoch": 0.316039311520948, + "grad_norm": 1.2316402196884155, + "learning_rate": 0.0001546707958863995, + "loss": 0.1067, + "step": 13482 + }, + { + "epoch": 0.316062753095753, + "grad_norm": 6.3709821701049805, + "learning_rate": 0.00015466462788206315, + "loss": 0.881, + "step": 13483 + }, + { + "epoch": 0.316086194670558, + "grad_norm": 5.090885639190674, + "learning_rate": 0.000154658459581115, + "loss": 0.6119, + "step": 13484 + }, + { + "epoch": 0.316109636245363, + "grad_norm": 2.4433610439300537, + "learning_rate": 0.00015465229098358848, + "loss": 1.452, + "step": 13485 + }, + { + "epoch": 0.316133077820168, + "grad_norm": 4.419124126434326, + "learning_rate": 0.00015464612208951714, + "loss": 0.4085, + "step": 13486 + }, + { + "epoch": 0.31615651939497297, + "grad_norm": 3.903341770172119, + "learning_rate": 0.00015463995289893438, + "loss": 0.4879, + "step": 13487 + }, + { + "epoch": 0.31617996096977796, + "grad_norm": 6.773998737335205, + "learning_rate": 0.00015463378341187372, + "loss": 0.6009, + "step": 13488 + }, + { + "epoch": 0.31620340254458296, + "grad_norm": 4.017707824707031, + "learning_rate": 0.00015462761362836859, + "loss": 0.3758, + "step": 13489 + }, + { + "epoch": 0.31622684411938795, + "grad_norm": 5.955153465270996, + "learning_rate": 0.0001546214435484525, + "loss": 0.5244, + "step": 13490 + }, + { + "epoch": 0.31625028569419295, + "grad_norm": 4.124872207641602, + "learning_rate": 0.0001546152731721589, + "loss": 0.6017, + "step": 13491 + }, + { + "epoch": 0.31627372726899794, + "grad_norm": 4.19096565246582, + "learning_rate": 0.00015460910249952132, + "loss": 0.6664, + "step": 13492 + }, + { + "epoch": 0.31629716884380293, + "grad_norm": 3.9143025875091553, + "learning_rate": 0.0001546029315305732, + "loss": 0.5293, + "step": 13493 + }, + { + "epoch": 0.31632061041860793, + "grad_norm": 0.9585302472114563, + "learning_rate": 0.00015459676026534805, + "loss": 0.1485, + "step": 13494 + }, + { + "epoch": 0.3163440519934129, + "grad_norm": 1.463449239730835, + "learning_rate": 0.00015459058870387937, + "loss": 0.1332, + "step": 13495 + }, + { + "epoch": 0.3163674935682179, + "grad_norm": 3.874760389328003, + "learning_rate": 0.00015458441684620055, + "loss": 0.2651, + "step": 13496 + }, + { + "epoch": 0.3163909351430229, + "grad_norm": 0.5531072616577148, + "learning_rate": 0.0001545782446923452, + "loss": 0.0405, + "step": 13497 + }, + { + "epoch": 0.3164143767178279, + "grad_norm": 1.6976969242095947, + "learning_rate": 0.00015457207224234674, + "loss": 0.2742, + "step": 13498 + }, + { + "epoch": 0.3164378182926329, + "grad_norm": 2.7495293617248535, + "learning_rate": 0.00015456589949623873, + "loss": 0.3278, + "step": 13499 + }, + { + "epoch": 0.3164612598674379, + "grad_norm": 1.370778203010559, + "learning_rate": 0.0001545597264540546, + "loss": 0.1704, + "step": 13500 + }, + { + "epoch": 0.3164847014422429, + "grad_norm": 4.419716835021973, + "learning_rate": 0.0001545535531158278, + "loss": 0.472, + "step": 13501 + }, + { + "epoch": 0.3165081430170479, + "grad_norm": 2.6487367153167725, + "learning_rate": 0.00015454737948159193, + "loss": 0.2793, + "step": 13502 + }, + { + "epoch": 0.3165315845918529, + "grad_norm": 3.1535961627960205, + "learning_rate": 0.00015454120555138045, + "loss": 0.3017, + "step": 13503 + }, + { + "epoch": 0.31655502616665787, + "grad_norm": 4.321781635284424, + "learning_rate": 0.00015453503132522683, + "loss": 0.5342, + "step": 13504 + }, + { + "epoch": 0.31657846774146287, + "grad_norm": 2.895543336868286, + "learning_rate": 0.00015452885680316463, + "loss": 0.1891, + "step": 13505 + }, + { + "epoch": 0.31660190931626786, + "grad_norm": 1.8613337278366089, + "learning_rate": 0.00015452268198522728, + "loss": 0.2175, + "step": 13506 + }, + { + "epoch": 0.31662535089107285, + "grad_norm": 0.9997182488441467, + "learning_rate": 0.00015451650687144835, + "loss": 0.0741, + "step": 13507 + }, + { + "epoch": 0.31664879246587785, + "grad_norm": 4.148270130157471, + "learning_rate": 0.0001545103314618613, + "loss": 0.4086, + "step": 13508 + }, + { + "epoch": 0.31667223404068284, + "grad_norm": 2.79237961769104, + "learning_rate": 0.00015450415575649968, + "loss": 0.2636, + "step": 13509 + }, + { + "epoch": 0.31669567561548784, + "grad_norm": 4.942378044128418, + "learning_rate": 0.00015449797975539698, + "loss": 0.5437, + "step": 13510 + }, + { + "epoch": 0.31671911719029283, + "grad_norm": 4.351325035095215, + "learning_rate": 0.00015449180345858673, + "loss": 0.5043, + "step": 13511 + }, + { + "epoch": 0.3167425587650978, + "grad_norm": 4.544153213500977, + "learning_rate": 0.00015448562686610237, + "loss": 0.2889, + "step": 13512 + }, + { + "epoch": 0.3167660003399028, + "grad_norm": 2.3363871574401855, + "learning_rate": 0.0001544794499779775, + "loss": 1.6667, + "step": 13513 + }, + { + "epoch": 0.3167894419147078, + "grad_norm": 3.7462010383605957, + "learning_rate": 0.0001544732727942456, + "loss": 0.3708, + "step": 13514 + }, + { + "epoch": 0.3168128834895128, + "grad_norm": 1.4196631908416748, + "learning_rate": 0.00015446709531494017, + "loss": 0.2075, + "step": 13515 + }, + { + "epoch": 0.3168363250643178, + "grad_norm": 2.7305102348327637, + "learning_rate": 0.00015446091754009477, + "loss": 0.2926, + "step": 13516 + }, + { + "epoch": 0.3168597666391228, + "grad_norm": 4.038724899291992, + "learning_rate": 0.0001544547394697429, + "loss": 0.5868, + "step": 13517 + }, + { + "epoch": 0.3168832082139278, + "grad_norm": 2.061258554458618, + "learning_rate": 0.00015444856110391806, + "loss": 0.2798, + "step": 13518 + }, + { + "epoch": 0.3169066497887328, + "grad_norm": 2.1704397201538086, + "learning_rate": 0.00015444238244265382, + "loss": 0.1967, + "step": 13519 + }, + { + "epoch": 0.3169300913635378, + "grad_norm": 2.33829927444458, + "learning_rate": 0.0001544362034859837, + "loss": 1.0948, + "step": 13520 + }, + { + "epoch": 0.3169535329383428, + "grad_norm": 3.2887122631073, + "learning_rate": 0.00015443002423394116, + "loss": 0.2405, + "step": 13521 + }, + { + "epoch": 0.31697697451314777, + "grad_norm": 2.6082382202148438, + "learning_rate": 0.0001544238446865598, + "loss": 1.6085, + "step": 13522 + }, + { + "epoch": 0.31700041608795276, + "grad_norm": 1.0821714401245117, + "learning_rate": 0.00015441766484387313, + "loss": 0.1279, + "step": 13523 + }, + { + "epoch": 0.31702385766275776, + "grad_norm": 1.363377332687378, + "learning_rate": 0.0001544114847059147, + "loss": 0.1224, + "step": 13524 + }, + { + "epoch": 0.31704729923756275, + "grad_norm": 1.9650319814682007, + "learning_rate": 0.000154405304272718, + "loss": 1.4237, + "step": 13525 + }, + { + "epoch": 0.3170707408123678, + "grad_norm": 1.3321845531463623, + "learning_rate": 0.0001543991235443166, + "loss": 0.1432, + "step": 13526 + }, + { + "epoch": 0.3170941823871728, + "grad_norm": 1.3657926321029663, + "learning_rate": 0.00015439294252074398, + "loss": 0.6932, + "step": 13527 + }, + { + "epoch": 0.3171176239619778, + "grad_norm": 3.4503321647644043, + "learning_rate": 0.0001543867612020338, + "loss": 0.3341, + "step": 13528 + }, + { + "epoch": 0.3171410655367828, + "grad_norm": 2.327160358428955, + "learning_rate": 0.00015438057958821948, + "loss": 0.1836, + "step": 13529 + }, + { + "epoch": 0.3171645071115878, + "grad_norm": 4.138939380645752, + "learning_rate": 0.00015437439767933465, + "loss": 0.4672, + "step": 13530 + }, + { + "epoch": 0.3171879486863928, + "grad_norm": 1.0688265562057495, + "learning_rate": 0.00015436821547541274, + "loss": 0.048, + "step": 13531 + }, + { + "epoch": 0.31721139026119777, + "grad_norm": 0.9870208501815796, + "learning_rate": 0.00015436203297648738, + "loss": 0.1074, + "step": 13532 + }, + { + "epoch": 0.31723483183600276, + "grad_norm": 3.5587141513824463, + "learning_rate": 0.00015435585018259214, + "loss": 0.3335, + "step": 13533 + }, + { + "epoch": 0.31725827341080776, + "grad_norm": 3.627990961074829, + "learning_rate": 0.00015434966709376052, + "loss": 0.3369, + "step": 13534 + }, + { + "epoch": 0.31728171498561275, + "grad_norm": 2.735363483428955, + "learning_rate": 0.00015434348371002606, + "loss": 0.2386, + "step": 13535 + }, + { + "epoch": 0.31730515656041774, + "grad_norm": 2.62638258934021, + "learning_rate": 0.00015433730003142234, + "loss": 0.1829, + "step": 13536 + }, + { + "epoch": 0.31732859813522274, + "grad_norm": 2.434312105178833, + "learning_rate": 0.0001543311160579829, + "loss": 0.3182, + "step": 13537 + }, + { + "epoch": 0.31735203971002773, + "grad_norm": 2.4426121711730957, + "learning_rate": 0.0001543249317897413, + "loss": 0.3615, + "step": 13538 + }, + { + "epoch": 0.31737548128483273, + "grad_norm": 2.715813636779785, + "learning_rate": 0.0001543187472267311, + "loss": 0.2647, + "step": 13539 + }, + { + "epoch": 0.3173989228596377, + "grad_norm": 3.0850069522857666, + "learning_rate": 0.0001543125623689858, + "loss": 0.3601, + "step": 13540 + }, + { + "epoch": 0.3174223644344427, + "grad_norm": 1.542203664779663, + "learning_rate": 0.00015430637721653903, + "loss": 0.1312, + "step": 13541 + }, + { + "epoch": 0.3174458060092477, + "grad_norm": 1.4657740592956543, + "learning_rate": 0.00015430019176942435, + "loss": 0.1145, + "step": 13542 + }, + { + "epoch": 0.3174692475840527, + "grad_norm": 1.0350631475448608, + "learning_rate": 0.00015429400602767532, + "loss": 0.0836, + "step": 13543 + }, + { + "epoch": 0.3174926891588577, + "grad_norm": 2.939012289047241, + "learning_rate": 0.00015428781999132549, + "loss": 0.4125, + "step": 13544 + }, + { + "epoch": 0.3175161307336627, + "grad_norm": 1.8503727912902832, + "learning_rate": 0.0001542816336604084, + "loss": 0.2822, + "step": 13545 + }, + { + "epoch": 0.3175395723084677, + "grad_norm": 1.2592206001281738, + "learning_rate": 0.00015427544703495763, + "loss": 0.153, + "step": 13546 + }, + { + "epoch": 0.3175630138832727, + "grad_norm": 1.1589056253433228, + "learning_rate": 0.00015426926011500675, + "loss": 0.079, + "step": 13547 + }, + { + "epoch": 0.3175864554580777, + "grad_norm": 2.416243314743042, + "learning_rate": 0.00015426307290058934, + "loss": 1.2839, + "step": 13548 + }, + { + "epoch": 0.31760989703288267, + "grad_norm": 3.406595468521118, + "learning_rate": 0.000154256885391739, + "loss": 0.465, + "step": 13549 + }, + { + "epoch": 0.31763333860768767, + "grad_norm": 5.172327518463135, + "learning_rate": 0.0001542506975884893, + "loss": 0.5908, + "step": 13550 + }, + { + "epoch": 0.31765678018249266, + "grad_norm": 2.269139289855957, + "learning_rate": 0.00015424450949087372, + "loss": 1.1782, + "step": 13551 + }, + { + "epoch": 0.31768022175729765, + "grad_norm": 0.5485760569572449, + "learning_rate": 0.00015423832109892594, + "loss": 0.1019, + "step": 13552 + }, + { + "epoch": 0.31770366333210265, + "grad_norm": 3.0996601581573486, + "learning_rate": 0.0001542321324126795, + "loss": 1.8829, + "step": 13553 + }, + { + "epoch": 0.31772710490690764, + "grad_norm": 1.2964849472045898, + "learning_rate": 0.000154225943432168, + "loss": 0.1318, + "step": 13554 + }, + { + "epoch": 0.31775054648171264, + "grad_norm": 3.2043981552124023, + "learning_rate": 0.00015421975415742497, + "loss": 0.3467, + "step": 13555 + }, + { + "epoch": 0.31777398805651763, + "grad_norm": 3.726407051086426, + "learning_rate": 0.00015421356458848406, + "loss": 0.4022, + "step": 13556 + }, + { + "epoch": 0.3177974296313226, + "grad_norm": 3.4876277446746826, + "learning_rate": 0.0001542073747253788, + "loss": 0.2943, + "step": 13557 + }, + { + "epoch": 0.3178208712061276, + "grad_norm": 2.9859349727630615, + "learning_rate": 0.0001542011845681428, + "loss": 0.4592, + "step": 13558 + }, + { + "epoch": 0.3178443127809326, + "grad_norm": 2.8646397590637207, + "learning_rate": 0.0001541949941168097, + "loss": 0.3187, + "step": 13559 + }, + { + "epoch": 0.3178677543557376, + "grad_norm": 1.5400443077087402, + "learning_rate": 0.000154188803371413, + "loss": 0.2085, + "step": 13560 + }, + { + "epoch": 0.3178911959305426, + "grad_norm": 1.892572283744812, + "learning_rate": 0.00015418261233198633, + "loss": 0.1647, + "step": 13561 + }, + { + "epoch": 0.3179146375053476, + "grad_norm": 2.9545469284057617, + "learning_rate": 0.00015417642099856328, + "loss": 0.3006, + "step": 13562 + }, + { + "epoch": 0.3179380790801526, + "grad_norm": 2.864410161972046, + "learning_rate": 0.00015417022937117746, + "loss": 2.379, + "step": 13563 + }, + { + "epoch": 0.3179615206549576, + "grad_norm": 3.2326860427856445, + "learning_rate": 0.00015416403744986242, + "loss": 0.3381, + "step": 13564 + }, + { + "epoch": 0.3179849622297626, + "grad_norm": 4.238204479217529, + "learning_rate": 0.0001541578452346518, + "loss": 0.5094, + "step": 13565 + }, + { + "epoch": 0.3180084038045676, + "grad_norm": 3.8479554653167725, + "learning_rate": 0.00015415165272557923, + "loss": 0.3154, + "step": 13566 + }, + { + "epoch": 0.31803184537937257, + "grad_norm": 0.7121980786323547, + "learning_rate": 0.00015414545992267824, + "loss": 0.0771, + "step": 13567 + }, + { + "epoch": 0.31805528695417756, + "grad_norm": 5.200536251068115, + "learning_rate": 0.00015413926682598245, + "loss": 0.5268, + "step": 13568 + }, + { + "epoch": 0.31807872852898256, + "grad_norm": 0.8515006303787231, + "learning_rate": 0.00015413307343552547, + "loss": 0.0571, + "step": 13569 + }, + { + "epoch": 0.31810217010378755, + "grad_norm": 8.485260963439941, + "learning_rate": 0.00015412687975134096, + "loss": 0.6339, + "step": 13570 + }, + { + "epoch": 0.31812561167859255, + "grad_norm": 0.562802255153656, + "learning_rate": 0.00015412068577346245, + "loss": 0.0628, + "step": 13571 + }, + { + "epoch": 0.31814905325339754, + "grad_norm": 2.9718832969665527, + "learning_rate": 0.00015411449150192356, + "loss": 2.0943, + "step": 13572 + }, + { + "epoch": 0.31817249482820253, + "grad_norm": 2.6955087184906006, + "learning_rate": 0.00015410829693675795, + "loss": 0.271, + "step": 13573 + }, + { + "epoch": 0.31819593640300753, + "grad_norm": 1.056794285774231, + "learning_rate": 0.00015410210207799917, + "loss": 0.1052, + "step": 13574 + }, + { + "epoch": 0.3182193779778125, + "grad_norm": 2.617037773132324, + "learning_rate": 0.00015409590692568088, + "loss": 0.1642, + "step": 13575 + }, + { + "epoch": 0.3182428195526175, + "grad_norm": 3.7189393043518066, + "learning_rate": 0.00015408971147983668, + "loss": 0.4427, + "step": 13576 + }, + { + "epoch": 0.31826626112742257, + "grad_norm": 1.2340930700302124, + "learning_rate": 0.00015408351574050017, + "loss": 0.1298, + "step": 13577 + }, + { + "epoch": 0.31828970270222756, + "grad_norm": 1.098104476928711, + "learning_rate": 0.000154077319707705, + "loss": 0.1183, + "step": 13578 + }, + { + "epoch": 0.31831314427703256, + "grad_norm": 3.228487014770508, + "learning_rate": 0.00015407112338148478, + "loss": 0.4119, + "step": 13579 + }, + { + "epoch": 0.31833658585183755, + "grad_norm": 3.122802972793579, + "learning_rate": 0.0001540649267618731, + "loss": 0.2766, + "step": 13580 + }, + { + "epoch": 0.31836002742664254, + "grad_norm": 3.3195383548736572, + "learning_rate": 0.0001540587298489036, + "loss": 0.313, + "step": 13581 + }, + { + "epoch": 0.31838346900144754, + "grad_norm": 5.0591230392456055, + "learning_rate": 0.00015405253264260993, + "loss": 0.4353, + "step": 13582 + }, + { + "epoch": 0.31840691057625253, + "grad_norm": 2.869882345199585, + "learning_rate": 0.0001540463351430257, + "loss": 0.2831, + "step": 13583 + }, + { + "epoch": 0.3184303521510575, + "grad_norm": 3.2961347103118896, + "learning_rate": 0.00015404013735018454, + "loss": 1.782, + "step": 13584 + }, + { + "epoch": 0.3184537937258625, + "grad_norm": 3.8192124366760254, + "learning_rate": 0.00015403393926412007, + "loss": 0.523, + "step": 13585 + }, + { + "epoch": 0.3184772353006675, + "grad_norm": 4.144343376159668, + "learning_rate": 0.00015402774088486593, + "loss": 0.5168, + "step": 13586 + }, + { + "epoch": 0.3185006768754725, + "grad_norm": 2.9149398803710938, + "learning_rate": 0.00015402154221245571, + "loss": 0.1811, + "step": 13587 + }, + { + "epoch": 0.3185241184502775, + "grad_norm": 2.1031763553619385, + "learning_rate": 0.00015401534324692312, + "loss": 0.2257, + "step": 13588 + }, + { + "epoch": 0.3185475600250825, + "grad_norm": 3.0816144943237305, + "learning_rate": 0.0001540091439883018, + "loss": 0.3438, + "step": 13589 + }, + { + "epoch": 0.3185710015998875, + "grad_norm": 4.958035945892334, + "learning_rate": 0.0001540029444366253, + "loss": 0.6397, + "step": 13590 + }, + { + "epoch": 0.3185944431746925, + "grad_norm": 2.5124008655548096, + "learning_rate": 0.00015399674459192728, + "loss": 1.2764, + "step": 13591 + }, + { + "epoch": 0.3186178847494975, + "grad_norm": 3.4726016521453857, + "learning_rate": 0.00015399054445424144, + "loss": 0.2912, + "step": 13592 + }, + { + "epoch": 0.3186413263243025, + "grad_norm": 4.6860671043396, + "learning_rate": 0.00015398434402360135, + "loss": 0.4108, + "step": 13593 + }, + { + "epoch": 0.31866476789910747, + "grad_norm": 5.088877201080322, + "learning_rate": 0.00015397814330004073, + "loss": 0.5389, + "step": 13594 + }, + { + "epoch": 0.31868820947391246, + "grad_norm": 4.889698505401611, + "learning_rate": 0.00015397194228359317, + "loss": 0.3354, + "step": 13595 + }, + { + "epoch": 0.31871165104871746, + "grad_norm": 0.4863000810146332, + "learning_rate": 0.00015396574097429236, + "loss": 0.0442, + "step": 13596 + }, + { + "epoch": 0.31873509262352245, + "grad_norm": 2.900012254714966, + "learning_rate": 0.00015395953937217188, + "loss": 0.3239, + "step": 13597 + }, + { + "epoch": 0.31875853419832745, + "grad_norm": 6.11453104019165, + "learning_rate": 0.00015395333747726545, + "loss": 0.7386, + "step": 13598 + }, + { + "epoch": 0.31878197577313244, + "grad_norm": 2.3041863441467285, + "learning_rate": 0.0001539471352896067, + "loss": 0.3134, + "step": 13599 + }, + { + "epoch": 0.31880541734793744, + "grad_norm": 4.731856822967529, + "learning_rate": 0.00015394093280922925, + "loss": 0.5335, + "step": 13600 + }, + { + "epoch": 0.31882885892274243, + "grad_norm": 4.848690032958984, + "learning_rate": 0.0001539347300361668, + "loss": 0.3551, + "step": 13601 + }, + { + "epoch": 0.3188523004975474, + "grad_norm": 2.720207929611206, + "learning_rate": 0.00015392852697045299, + "loss": 0.3071, + "step": 13602 + }, + { + "epoch": 0.3188757420723524, + "grad_norm": 7.97993278503418, + "learning_rate": 0.00015392232361212147, + "loss": 0.9964, + "step": 13603 + }, + { + "epoch": 0.3188991836471574, + "grad_norm": 3.006547451019287, + "learning_rate": 0.0001539161199612059, + "loss": 0.3494, + "step": 13604 + }, + { + "epoch": 0.3189226252219624, + "grad_norm": 3.6100668907165527, + "learning_rate": 0.00015390991601773997, + "loss": 0.275, + "step": 13605 + }, + { + "epoch": 0.3189460667967674, + "grad_norm": 2.1375088691711426, + "learning_rate": 0.0001539037117817573, + "loss": 0.228, + "step": 13606 + }, + { + "epoch": 0.3189695083715724, + "grad_norm": 5.287006378173828, + "learning_rate": 0.00015389750725329157, + "loss": 0.4509, + "step": 13607 + }, + { + "epoch": 0.3189929499463774, + "grad_norm": 1.7914068698883057, + "learning_rate": 0.00015389130243237647, + "loss": 0.2026, + "step": 13608 + }, + { + "epoch": 0.3190163915211824, + "grad_norm": 4.32578706741333, + "learning_rate": 0.00015388509731904563, + "loss": 0.4399, + "step": 13609 + }, + { + "epoch": 0.3190398330959874, + "grad_norm": 3.3937783241271973, + "learning_rate": 0.00015387889191333275, + "loss": 0.396, + "step": 13610 + }, + { + "epoch": 0.3190632746707924, + "grad_norm": 5.434929370880127, + "learning_rate": 0.0001538726862152715, + "loss": 0.5346, + "step": 13611 + }, + { + "epoch": 0.31908671624559737, + "grad_norm": 3.8000550270080566, + "learning_rate": 0.0001538664802248955, + "loss": 0.4361, + "step": 13612 + }, + { + "epoch": 0.31911015782040236, + "grad_norm": 5.551680564880371, + "learning_rate": 0.00015386027394223846, + "loss": 0.6941, + "step": 13613 + }, + { + "epoch": 0.31913359939520736, + "grad_norm": 5.399113655090332, + "learning_rate": 0.0001538540673673341, + "loss": 0.3322, + "step": 13614 + }, + { + "epoch": 0.31915704097001235, + "grad_norm": 2.600740909576416, + "learning_rate": 0.00015384786050021604, + "loss": 0.2527, + "step": 13615 + }, + { + "epoch": 0.31918048254481735, + "grad_norm": 5.486682891845703, + "learning_rate": 0.00015384165334091798, + "loss": 0.6043, + "step": 13616 + }, + { + "epoch": 0.31920392411962234, + "grad_norm": 1.457072377204895, + "learning_rate": 0.0001538354458894736, + "loss": 0.1515, + "step": 13617 + }, + { + "epoch": 0.31922736569442733, + "grad_norm": 4.334706783294678, + "learning_rate": 0.00015382923814591655, + "loss": 0.3953, + "step": 13618 + }, + { + "epoch": 0.31925080726923233, + "grad_norm": 1.9511449337005615, + "learning_rate": 0.00015382303011028054, + "loss": 0.1314, + "step": 13619 + }, + { + "epoch": 0.3192742488440373, + "grad_norm": 4.118494510650635, + "learning_rate": 0.0001538168217825993, + "loss": 0.4242, + "step": 13620 + }, + { + "epoch": 0.3192976904188423, + "grad_norm": 2.7118377685546875, + "learning_rate": 0.00015381061316290645, + "loss": 0.2727, + "step": 13621 + }, + { + "epoch": 0.3193211319936473, + "grad_norm": 1.2999366521835327, + "learning_rate": 0.00015380440425123568, + "loss": 0.1808, + "step": 13622 + }, + { + "epoch": 0.3193445735684523, + "grad_norm": 1.6703301668167114, + "learning_rate": 0.0001537981950476207, + "loss": 0.8041, + "step": 13623 + }, + { + "epoch": 0.3193680151432573, + "grad_norm": 3.346311569213867, + "learning_rate": 0.00015379198555209525, + "loss": 0.346, + "step": 13624 + }, + { + "epoch": 0.3193914567180623, + "grad_norm": 3.6915335655212402, + "learning_rate": 0.00015378577576469294, + "loss": 0.4174, + "step": 13625 + }, + { + "epoch": 0.3194148982928673, + "grad_norm": 3.9978480339050293, + "learning_rate": 0.00015377956568544748, + "loss": 0.4482, + "step": 13626 + }, + { + "epoch": 0.3194383398676723, + "grad_norm": 3.268421173095703, + "learning_rate": 0.0001537733553143926, + "loss": 0.3283, + "step": 13627 + }, + { + "epoch": 0.31946178144247733, + "grad_norm": 4.490615367889404, + "learning_rate": 0.00015376714465156202, + "loss": 0.5077, + "step": 13628 + }, + { + "epoch": 0.3194852230172823, + "grad_norm": 5.66698694229126, + "learning_rate": 0.00015376093369698932, + "loss": 0.6137, + "step": 13629 + }, + { + "epoch": 0.3195086645920873, + "grad_norm": 1.464684009552002, + "learning_rate": 0.0001537547224507083, + "loss": 0.2702, + "step": 13630 + }, + { + "epoch": 0.3195321061668923, + "grad_norm": 2.6475489139556885, + "learning_rate": 0.00015374851091275272, + "loss": 0.2356, + "step": 13631 + }, + { + "epoch": 0.3195555477416973, + "grad_norm": 1.759431004524231, + "learning_rate": 0.00015374229908315614, + "loss": 0.1439, + "step": 13632 + }, + { + "epoch": 0.3195789893165023, + "grad_norm": 1.250913143157959, + "learning_rate": 0.00015373608696195236, + "loss": 0.116, + "step": 13633 + }, + { + "epoch": 0.3196024308913073, + "grad_norm": 4.33125114440918, + "learning_rate": 0.00015372987454917507, + "loss": 0.4262, + "step": 13634 + }, + { + "epoch": 0.3196258724661123, + "grad_norm": 4.013427257537842, + "learning_rate": 0.00015372366184485795, + "loss": 0.3103, + "step": 13635 + }, + { + "epoch": 0.3196493140409173, + "grad_norm": 2.9135870933532715, + "learning_rate": 0.00015371744884903473, + "loss": 0.3437, + "step": 13636 + }, + { + "epoch": 0.3196727556157223, + "grad_norm": 2.7634623050689697, + "learning_rate": 0.00015371123556173912, + "loss": 0.3006, + "step": 13637 + }, + { + "epoch": 0.3196961971905273, + "grad_norm": 4.047855377197266, + "learning_rate": 0.00015370502198300482, + "loss": 0.4219, + "step": 13638 + }, + { + "epoch": 0.31971963876533227, + "grad_norm": 4.250809669494629, + "learning_rate": 0.00015369880811286562, + "loss": 0.4587, + "step": 13639 + }, + { + "epoch": 0.31974308034013726, + "grad_norm": 1.0111111402511597, + "learning_rate": 0.00015369259395135515, + "loss": 0.1238, + "step": 13640 + }, + { + "epoch": 0.31976652191494226, + "grad_norm": 4.392712116241455, + "learning_rate": 0.00015368637949850713, + "loss": 0.5193, + "step": 13641 + }, + { + "epoch": 0.31978996348974725, + "grad_norm": 3.8965678215026855, + "learning_rate": 0.00015368016475435532, + "loss": 0.4066, + "step": 13642 + }, + { + "epoch": 0.31981340506455225, + "grad_norm": 2.4478514194488525, + "learning_rate": 0.00015367394971893344, + "loss": 0.2031, + "step": 13643 + }, + { + "epoch": 0.31983684663935724, + "grad_norm": 2.0572104454040527, + "learning_rate": 0.00015366773439227517, + "loss": 0.1623, + "step": 13644 + }, + { + "epoch": 0.31986028821416224, + "grad_norm": 4.471080780029297, + "learning_rate": 0.00015366151877441427, + "loss": 0.4308, + "step": 13645 + }, + { + "epoch": 0.31988372978896723, + "grad_norm": 3.450451612472534, + "learning_rate": 0.0001536553028653845, + "loss": 0.194, + "step": 13646 + }, + { + "epoch": 0.3199071713637722, + "grad_norm": 1.9723550081253052, + "learning_rate": 0.00015364908666521952, + "loss": 0.1617, + "step": 13647 + }, + { + "epoch": 0.3199306129385772, + "grad_norm": 6.152344226837158, + "learning_rate": 0.00015364287017395308, + "loss": 0.3793, + "step": 13648 + }, + { + "epoch": 0.3199540545133822, + "grad_norm": 0.7102668285369873, + "learning_rate": 0.00015363665339161893, + "loss": 0.0434, + "step": 13649 + }, + { + "epoch": 0.3199774960881872, + "grad_norm": 1.0323107242584229, + "learning_rate": 0.0001536304363182508, + "loss": 0.078, + "step": 13650 + }, + { + "epoch": 0.3200009376629922, + "grad_norm": 3.1885628700256348, + "learning_rate": 0.0001536242189538824, + "loss": 0.3202, + "step": 13651 + }, + { + "epoch": 0.3200243792377972, + "grad_norm": 3.3645570278167725, + "learning_rate": 0.00015361800129854748, + "loss": 2.3458, + "step": 13652 + }, + { + "epoch": 0.3200478208126022, + "grad_norm": 0.9141591191291809, + "learning_rate": 0.0001536117833522798, + "loss": 0.0959, + "step": 13653 + }, + { + "epoch": 0.3200712623874072, + "grad_norm": 2.3391988277435303, + "learning_rate": 0.000153605565115113, + "loss": 0.2223, + "step": 13654 + }, + { + "epoch": 0.3200947039622122, + "grad_norm": 3.747738838195801, + "learning_rate": 0.00015359934658708098, + "loss": 0.5701, + "step": 13655 + }, + { + "epoch": 0.3201181455370172, + "grad_norm": 4.0175604820251465, + "learning_rate": 0.00015359312776821738, + "loss": 0.4953, + "step": 13656 + }, + { + "epoch": 0.32014158711182217, + "grad_norm": 6.337192058563232, + "learning_rate": 0.0001535869086585559, + "loss": 0.838, + "step": 13657 + }, + { + "epoch": 0.32016502868662716, + "grad_norm": 5.355804443359375, + "learning_rate": 0.00015358068925813045, + "loss": 0.4581, + "step": 13658 + }, + { + "epoch": 0.32018847026143216, + "grad_norm": 3.553316354751587, + "learning_rate": 0.00015357446956697462, + "loss": 0.4294, + "step": 13659 + }, + { + "epoch": 0.32021191183623715, + "grad_norm": 1.061957597732544, + "learning_rate": 0.00015356824958512219, + "loss": 0.15, + "step": 13660 + }, + { + "epoch": 0.32023535341104215, + "grad_norm": 2.502744197845459, + "learning_rate": 0.00015356202931260694, + "loss": 1.0109, + "step": 13661 + }, + { + "epoch": 0.32025879498584714, + "grad_norm": 3.900844097137451, + "learning_rate": 0.00015355580874946268, + "loss": 0.4609, + "step": 13662 + }, + { + "epoch": 0.32028223656065213, + "grad_norm": 3.2203307151794434, + "learning_rate": 0.00015354958789572302, + "loss": 1.3184, + "step": 13663 + }, + { + "epoch": 0.32030567813545713, + "grad_norm": 1.230124831199646, + "learning_rate": 0.0001535433667514218, + "loss": 0.1444, + "step": 13664 + }, + { + "epoch": 0.3203291197102621, + "grad_norm": 1.5048927068710327, + "learning_rate": 0.0001535371453165928, + "loss": 0.4092, + "step": 13665 + }, + { + "epoch": 0.3203525612850671, + "grad_norm": 4.418081283569336, + "learning_rate": 0.00015353092359126974, + "loss": 0.6057, + "step": 13666 + }, + { + "epoch": 0.3203760028598721, + "grad_norm": 2.922806978225708, + "learning_rate": 0.00015352470157548633, + "loss": 0.3095, + "step": 13667 + }, + { + "epoch": 0.3203994444346771, + "grad_norm": 2.76381516456604, + "learning_rate": 0.00015351847926927645, + "loss": 0.3037, + "step": 13668 + }, + { + "epoch": 0.3204228860094821, + "grad_norm": 6.470213890075684, + "learning_rate": 0.00015351225667267377, + "loss": 0.4343, + "step": 13669 + }, + { + "epoch": 0.3204463275842871, + "grad_norm": 3.780303478240967, + "learning_rate": 0.00015350603378571208, + "loss": 0.3252, + "step": 13670 + }, + { + "epoch": 0.3204697691590921, + "grad_norm": 2.5800423622131348, + "learning_rate": 0.00015349981060842517, + "loss": 0.2335, + "step": 13671 + }, + { + "epoch": 0.3204932107338971, + "grad_norm": 0.6130505204200745, + "learning_rate": 0.00015349358714084673, + "loss": 0.0329, + "step": 13672 + }, + { + "epoch": 0.3205166523087021, + "grad_norm": 2.7277965545654297, + "learning_rate": 0.0001534873633830106, + "loss": 0.2854, + "step": 13673 + }, + { + "epoch": 0.32054009388350707, + "grad_norm": 4.732657432556152, + "learning_rate": 0.00015348113933495055, + "loss": 0.3277, + "step": 13674 + }, + { + "epoch": 0.32056353545831207, + "grad_norm": 5.4166951179504395, + "learning_rate": 0.00015347491499670036, + "loss": 0.5081, + "step": 13675 + }, + { + "epoch": 0.32058697703311706, + "grad_norm": 5.013097763061523, + "learning_rate": 0.00015346869036829373, + "loss": 0.4697, + "step": 13676 + }, + { + "epoch": 0.32061041860792205, + "grad_norm": 4.552703857421875, + "learning_rate": 0.0001534624654497645, + "loss": 0.61, + "step": 13677 + }, + { + "epoch": 0.32063386018272705, + "grad_norm": 0.7817431688308716, + "learning_rate": 0.00015345624024114645, + "loss": 0.0687, + "step": 13678 + }, + { + "epoch": 0.3206573017575321, + "grad_norm": 0.3921999931335449, + "learning_rate": 0.0001534500147424733, + "loss": 0.0348, + "step": 13679 + }, + { + "epoch": 0.3206807433323371, + "grad_norm": 3.028202772140503, + "learning_rate": 0.00015344378895377884, + "loss": 0.422, + "step": 13680 + }, + { + "epoch": 0.3207041849071421, + "grad_norm": 4.301206588745117, + "learning_rate": 0.00015343756287509693, + "loss": 0.7533, + "step": 13681 + }, + { + "epoch": 0.3207276264819471, + "grad_norm": 3.2074735164642334, + "learning_rate": 0.00015343133650646133, + "loss": 0.3148, + "step": 13682 + }, + { + "epoch": 0.3207510680567521, + "grad_norm": 2.870670795440674, + "learning_rate": 0.00015342510984790573, + "loss": 0.2632, + "step": 13683 + }, + { + "epoch": 0.32077450963155707, + "grad_norm": 4.423669338226318, + "learning_rate": 0.00015341888289946402, + "loss": 0.6847, + "step": 13684 + }, + { + "epoch": 0.32079795120636206, + "grad_norm": 1.5401726961135864, + "learning_rate": 0.00015341265566116993, + "loss": 0.1436, + "step": 13685 + }, + { + "epoch": 0.32082139278116706, + "grad_norm": 2.963277816772461, + "learning_rate": 0.00015340642813305727, + "loss": 1.7513, + "step": 13686 + }, + { + "epoch": 0.32084483435597205, + "grad_norm": 0.4359607398509979, + "learning_rate": 0.00015340020031515985, + "loss": 0.0299, + "step": 13687 + }, + { + "epoch": 0.32086827593077705, + "grad_norm": 0.8187798261642456, + "learning_rate": 0.00015339397220751145, + "loss": 0.097, + "step": 13688 + }, + { + "epoch": 0.32089171750558204, + "grad_norm": 3.1246719360351562, + "learning_rate": 0.00015338774381014583, + "loss": 0.3575, + "step": 13689 + }, + { + "epoch": 0.32091515908038704, + "grad_norm": 2.3302390575408936, + "learning_rate": 0.0001533815151230968, + "loss": 0.2667, + "step": 13690 + }, + { + "epoch": 0.32093860065519203, + "grad_norm": 3.7478837966918945, + "learning_rate": 0.0001533752861463982, + "loss": 2.1337, + "step": 13691 + }, + { + "epoch": 0.320962042229997, + "grad_norm": 3.438243865966797, + "learning_rate": 0.00015336905688008377, + "loss": 0.2391, + "step": 13692 + }, + { + "epoch": 0.320985483804802, + "grad_norm": 3.8449437618255615, + "learning_rate": 0.00015336282732418733, + "loss": 0.3971, + "step": 13693 + }, + { + "epoch": 0.321008925379607, + "grad_norm": 4.516076564788818, + "learning_rate": 0.00015335659747874272, + "loss": 0.4503, + "step": 13694 + }, + { + "epoch": 0.321032366954412, + "grad_norm": 0.518557071685791, + "learning_rate": 0.00015335036734378366, + "loss": 0.0747, + "step": 13695 + }, + { + "epoch": 0.321055808529217, + "grad_norm": 3.0497090816497803, + "learning_rate": 0.00015334413691934408, + "loss": 0.2985, + "step": 13696 + }, + { + "epoch": 0.321079250104022, + "grad_norm": 0.8324521780014038, + "learning_rate": 0.00015333790620545768, + "loss": 0.1376, + "step": 13697 + }, + { + "epoch": 0.321102691678827, + "grad_norm": 3.1105127334594727, + "learning_rate": 0.0001533316752021583, + "loss": 0.349, + "step": 13698 + }, + { + "epoch": 0.321126133253632, + "grad_norm": 3.868670701980591, + "learning_rate": 0.00015332544390947973, + "loss": 0.4779, + "step": 13699 + }, + { + "epoch": 0.321149574828437, + "grad_norm": 4.000874042510986, + "learning_rate": 0.00015331921232745584, + "loss": 0.4264, + "step": 13700 + }, + { + "epoch": 0.321173016403242, + "grad_norm": 3.7559475898742676, + "learning_rate": 0.00015331298045612036, + "loss": 0.5168, + "step": 13701 + }, + { + "epoch": 0.32119645797804697, + "grad_norm": 0.22693507373332977, + "learning_rate": 0.0001533067482955072, + "loss": 0.0163, + "step": 13702 + }, + { + "epoch": 0.32121989955285196, + "grad_norm": 0.5121335387229919, + "learning_rate": 0.0001533005158456501, + "loss": 0.0486, + "step": 13703 + }, + { + "epoch": 0.32124334112765696, + "grad_norm": 3.7193989753723145, + "learning_rate": 0.00015329428310658288, + "loss": 0.5085, + "step": 13704 + }, + { + "epoch": 0.32126678270246195, + "grad_norm": 3.5687453746795654, + "learning_rate": 0.00015328805007833938, + "loss": 0.349, + "step": 13705 + }, + { + "epoch": 0.32129022427726694, + "grad_norm": 3.4500439167022705, + "learning_rate": 0.00015328181676095348, + "loss": 0.3289, + "step": 13706 + }, + { + "epoch": 0.32131366585207194, + "grad_norm": 2.459489107131958, + "learning_rate": 0.0001532755831544589, + "loss": 0.2567, + "step": 13707 + }, + { + "epoch": 0.32133710742687693, + "grad_norm": 2.406045436859131, + "learning_rate": 0.0001532693492588895, + "loss": 1.4088, + "step": 13708 + }, + { + "epoch": 0.3213605490016819, + "grad_norm": 3.860001802444458, + "learning_rate": 0.00015326311507427915, + "loss": 0.2016, + "step": 13709 + }, + { + "epoch": 0.3213839905764869, + "grad_norm": 1.4608739614486694, + "learning_rate": 0.00015325688060066162, + "loss": 0.178, + "step": 13710 + }, + { + "epoch": 0.3214074321512919, + "grad_norm": 3.8137810230255127, + "learning_rate": 0.00015325064583807075, + "loss": 0.2674, + "step": 13711 + }, + { + "epoch": 0.3214308737260969, + "grad_norm": 2.484714984893799, + "learning_rate": 0.0001532444107865404, + "loss": 0.4405, + "step": 13712 + }, + { + "epoch": 0.3214543153009019, + "grad_norm": 3.5442469120025635, + "learning_rate": 0.00015323817544610438, + "loss": 0.2486, + "step": 13713 + }, + { + "epoch": 0.3214777568757069, + "grad_norm": 2.724297285079956, + "learning_rate": 0.0001532319398167965, + "loss": 0.3592, + "step": 13714 + }, + { + "epoch": 0.3215011984505119, + "grad_norm": 3.2866575717926025, + "learning_rate": 0.0001532257038986506, + "loss": 0.16, + "step": 13715 + }, + { + "epoch": 0.3215246400253169, + "grad_norm": 3.597195863723755, + "learning_rate": 0.00015321946769170056, + "loss": 0.2904, + "step": 13716 + }, + { + "epoch": 0.3215480816001219, + "grad_norm": 2.6759724617004395, + "learning_rate": 0.00015321323119598025, + "loss": 0.2758, + "step": 13717 + }, + { + "epoch": 0.3215715231749269, + "grad_norm": 3.8159666061401367, + "learning_rate": 0.00015320699441152336, + "loss": 0.4361, + "step": 13718 + }, + { + "epoch": 0.32159496474973187, + "grad_norm": 3.063232421875, + "learning_rate": 0.00015320075733836385, + "loss": 0.3819, + "step": 13719 + }, + { + "epoch": 0.32161840632453687, + "grad_norm": 4.22384786605835, + "learning_rate": 0.00015319451997653557, + "loss": 0.4394, + "step": 13720 + }, + { + "epoch": 0.32164184789934186, + "grad_norm": 2.767850160598755, + "learning_rate": 0.00015318828232607228, + "loss": 0.1227, + "step": 13721 + }, + { + "epoch": 0.32166528947414685, + "grad_norm": 3.176283836364746, + "learning_rate": 0.0001531820443870079, + "loss": 0.2965, + "step": 13722 + }, + { + "epoch": 0.32168873104895185, + "grad_norm": 1.9430012702941895, + "learning_rate": 0.00015317580615937626, + "loss": 0.1608, + "step": 13723 + }, + { + "epoch": 0.32171217262375684, + "grad_norm": 2.656405448913574, + "learning_rate": 0.00015316956764321117, + "loss": 1.6264, + "step": 13724 + }, + { + "epoch": 0.32173561419856184, + "grad_norm": 5.4120378494262695, + "learning_rate": 0.00015316332883854653, + "loss": 0.7127, + "step": 13725 + }, + { + "epoch": 0.32175905577336683, + "grad_norm": 1.608506441116333, + "learning_rate": 0.0001531570897454162, + "loss": 0.1571, + "step": 13726 + }, + { + "epoch": 0.3217824973481718, + "grad_norm": 4.273404598236084, + "learning_rate": 0.00015315085036385394, + "loss": 0.5272, + "step": 13727 + }, + { + "epoch": 0.3218059389229768, + "grad_norm": 2.8301923274993896, + "learning_rate": 0.00015314461069389373, + "loss": 0.2228, + "step": 13728 + }, + { + "epoch": 0.3218293804977818, + "grad_norm": 0.925371527671814, + "learning_rate": 0.00015313837073556934, + "loss": 0.0856, + "step": 13729 + }, + { + "epoch": 0.32185282207258686, + "grad_norm": 3.325427293777466, + "learning_rate": 0.00015313213048891467, + "loss": 0.3311, + "step": 13730 + }, + { + "epoch": 0.32187626364739186, + "grad_norm": 5.836112022399902, + "learning_rate": 0.00015312588995396355, + "loss": 0.7301, + "step": 13731 + }, + { + "epoch": 0.32189970522219685, + "grad_norm": 4.847346305847168, + "learning_rate": 0.00015311964913074987, + "loss": 0.8719, + "step": 13732 + }, + { + "epoch": 0.32192314679700185, + "grad_norm": 1.5317565202713013, + "learning_rate": 0.00015311340801930746, + "loss": 0.1176, + "step": 13733 + }, + { + "epoch": 0.32194658837180684, + "grad_norm": 3.568277597427368, + "learning_rate": 0.00015310716661967022, + "loss": 0.3069, + "step": 13734 + }, + { + "epoch": 0.32197002994661184, + "grad_norm": 3.21389102935791, + "learning_rate": 0.000153100924931872, + "loss": 0.2438, + "step": 13735 + }, + { + "epoch": 0.32199347152141683, + "grad_norm": 2.585123062133789, + "learning_rate": 0.0001530946829559467, + "loss": 0.1306, + "step": 13736 + }, + { + "epoch": 0.3220169130962218, + "grad_norm": 0.9275563359260559, + "learning_rate": 0.00015308844069192814, + "loss": 0.1197, + "step": 13737 + }, + { + "epoch": 0.3220403546710268, + "grad_norm": 2.077883243560791, + "learning_rate": 0.0001530821981398502, + "loss": 0.2543, + "step": 13738 + }, + { + "epoch": 0.3220637962458318, + "grad_norm": 3.607086420059204, + "learning_rate": 0.00015307595529974677, + "loss": 0.4315, + "step": 13739 + }, + { + "epoch": 0.3220872378206368, + "grad_norm": 5.628911972045898, + "learning_rate": 0.00015306971217165166, + "loss": 0.3054, + "step": 13740 + }, + { + "epoch": 0.3221106793954418, + "grad_norm": 1.7466238737106323, + "learning_rate": 0.00015306346875559884, + "loss": 0.1089, + "step": 13741 + }, + { + "epoch": 0.3221341209702468, + "grad_norm": 4.028595447540283, + "learning_rate": 0.00015305722505162218, + "loss": 0.47, + "step": 13742 + }, + { + "epoch": 0.3221575625450518, + "grad_norm": 4.427788734436035, + "learning_rate": 0.00015305098105975547, + "loss": 0.2634, + "step": 13743 + }, + { + "epoch": 0.3221810041198568, + "grad_norm": 3.388731002807617, + "learning_rate": 0.00015304473678003268, + "loss": 0.2678, + "step": 13744 + }, + { + "epoch": 0.3222044456946618, + "grad_norm": 3.0450050830841064, + "learning_rate": 0.00015303849221248763, + "loss": 0.1762, + "step": 13745 + }, + { + "epoch": 0.3222278872694668, + "grad_norm": 1.7664704322814941, + "learning_rate": 0.00015303224735715426, + "loss": 0.2464, + "step": 13746 + }, + { + "epoch": 0.32225132884427177, + "grad_norm": 1.3031269311904907, + "learning_rate": 0.00015302600221406637, + "loss": 0.1684, + "step": 13747 + }, + { + "epoch": 0.32227477041907676, + "grad_norm": 1.6922324895858765, + "learning_rate": 0.00015301975678325795, + "loss": 0.1825, + "step": 13748 + }, + { + "epoch": 0.32229821199388176, + "grad_norm": 2.7667651176452637, + "learning_rate": 0.00015301351106476285, + "loss": 0.2884, + "step": 13749 + }, + { + "epoch": 0.32232165356868675, + "grad_norm": 0.9272781610488892, + "learning_rate": 0.0001530072650586149, + "loss": 0.0935, + "step": 13750 + }, + { + "epoch": 0.32234509514349174, + "grad_norm": 1.3117314577102661, + "learning_rate": 0.0001530010187648481, + "loss": 0.2182, + "step": 13751 + }, + { + "epoch": 0.32236853671829674, + "grad_norm": 2.627753257751465, + "learning_rate": 0.00015299477218349626, + "loss": 0.3326, + "step": 13752 + }, + { + "epoch": 0.32239197829310173, + "grad_norm": 4.905147552490234, + "learning_rate": 0.00015298852531459324, + "loss": 0.5487, + "step": 13753 + }, + { + "epoch": 0.3224154198679067, + "grad_norm": 3.795074939727783, + "learning_rate": 0.00015298227815817306, + "loss": 0.4736, + "step": 13754 + }, + { + "epoch": 0.3224388614427117, + "grad_norm": 2.8979787826538086, + "learning_rate": 0.00015297603071426954, + "loss": 0.4368, + "step": 13755 + }, + { + "epoch": 0.3224623030175167, + "grad_norm": 3.9676620960235596, + "learning_rate": 0.00015296978298291655, + "loss": 0.4834, + "step": 13756 + }, + { + "epoch": 0.3224857445923217, + "grad_norm": 3.597630739212036, + "learning_rate": 0.00015296353496414806, + "loss": 0.3961, + "step": 13757 + }, + { + "epoch": 0.3225091861671267, + "grad_norm": 2.585117816925049, + "learning_rate": 0.00015295728665799794, + "loss": 0.2554, + "step": 13758 + }, + { + "epoch": 0.3225326277419317, + "grad_norm": 3.8169939517974854, + "learning_rate": 0.00015295103806450006, + "loss": 0.487, + "step": 13759 + }, + { + "epoch": 0.3225560693167367, + "grad_norm": 0.7476332783699036, + "learning_rate": 0.0001529447891836884, + "loss": 0.0689, + "step": 13760 + }, + { + "epoch": 0.3225795108915417, + "grad_norm": 0.996198296546936, + "learning_rate": 0.0001529385400155968, + "loss": 0.091, + "step": 13761 + }, + { + "epoch": 0.3226029524663467, + "grad_norm": 1.4506371021270752, + "learning_rate": 0.00015293229056025922, + "loss": 0.2123, + "step": 13762 + }, + { + "epoch": 0.3226263940411517, + "grad_norm": 3.2564258575439453, + "learning_rate": 0.00015292604081770952, + "loss": 0.4113, + "step": 13763 + }, + { + "epoch": 0.32264983561595667, + "grad_norm": 4.13185977935791, + "learning_rate": 0.00015291979078798162, + "loss": 0.3041, + "step": 13764 + }, + { + "epoch": 0.32267327719076166, + "grad_norm": 2.1056900024414062, + "learning_rate": 0.00015291354047110948, + "loss": 0.1481, + "step": 13765 + }, + { + "epoch": 0.32269671876556666, + "grad_norm": 3.2496109008789062, + "learning_rate": 0.00015290728986712697, + "loss": 0.4981, + "step": 13766 + }, + { + "epoch": 0.32272016034037165, + "grad_norm": 6.349450588226318, + "learning_rate": 0.00015290103897606802, + "loss": 0.8008, + "step": 13767 + }, + { + "epoch": 0.32274360191517665, + "grad_norm": 3.113950729370117, + "learning_rate": 0.0001528947877979665, + "loss": 0.3265, + "step": 13768 + }, + { + "epoch": 0.32276704348998164, + "grad_norm": 4.922294616699219, + "learning_rate": 0.00015288853633285644, + "loss": 0.536, + "step": 13769 + }, + { + "epoch": 0.32279048506478664, + "grad_norm": 2.948791980743408, + "learning_rate": 0.00015288228458077167, + "loss": 0.4881, + "step": 13770 + }, + { + "epoch": 0.32281392663959163, + "grad_norm": 0.981738805770874, + "learning_rate": 0.00015287603254174613, + "loss": 0.1278, + "step": 13771 + }, + { + "epoch": 0.3228373682143966, + "grad_norm": 3.1729116439819336, + "learning_rate": 0.00015286978021581375, + "loss": 0.3792, + "step": 13772 + }, + { + "epoch": 0.3228608097892016, + "grad_norm": 0.8051655888557434, + "learning_rate": 0.00015286352760300847, + "loss": 0.0463, + "step": 13773 + }, + { + "epoch": 0.3228842513640066, + "grad_norm": 2.6150729656219482, + "learning_rate": 0.00015285727470336423, + "loss": 0.2814, + "step": 13774 + }, + { + "epoch": 0.3229076929388116, + "grad_norm": 1.2083863019943237, + "learning_rate": 0.0001528510215169149, + "loss": 0.119, + "step": 13775 + }, + { + "epoch": 0.3229311345136166, + "grad_norm": 1.3295049667358398, + "learning_rate": 0.00015284476804369441, + "loss": 0.1543, + "step": 13776 + }, + { + "epoch": 0.3229545760884216, + "grad_norm": 3.9802820682525635, + "learning_rate": 0.00015283851428373675, + "loss": 0.5866, + "step": 13777 + }, + { + "epoch": 0.3229780176632266, + "grad_norm": 3.0363714694976807, + "learning_rate": 0.00015283226023707583, + "loss": 1.7319, + "step": 13778 + }, + { + "epoch": 0.3230014592380316, + "grad_norm": 1.3898594379425049, + "learning_rate": 0.0001528260059037456, + "loss": 0.1928, + "step": 13779 + }, + { + "epoch": 0.3230249008128366, + "grad_norm": 3.0923826694488525, + "learning_rate": 0.00015281975128377995, + "loss": 0.443, + "step": 13780 + }, + { + "epoch": 0.32304834238764163, + "grad_norm": 2.6641838550567627, + "learning_rate": 0.00015281349637721284, + "loss": 0.247, + "step": 13781 + }, + { + "epoch": 0.3230717839624466, + "grad_norm": 3.4117257595062256, + "learning_rate": 0.00015280724118407822, + "loss": 0.4498, + "step": 13782 + }, + { + "epoch": 0.3230952255372516, + "grad_norm": 2.481968879699707, + "learning_rate": 0.00015280098570441002, + "loss": 0.1919, + "step": 13783 + }, + { + "epoch": 0.3231186671120566, + "grad_norm": 5.652230262756348, + "learning_rate": 0.00015279472993824222, + "loss": 0.6739, + "step": 13784 + }, + { + "epoch": 0.3231421086868616, + "grad_norm": 3.851905107498169, + "learning_rate": 0.00015278847388560868, + "loss": 0.4432, + "step": 13785 + }, + { + "epoch": 0.3231655502616666, + "grad_norm": 1.9300978183746338, + "learning_rate": 0.00015278221754654343, + "loss": 0.2897, + "step": 13786 + }, + { + "epoch": 0.3231889918364716, + "grad_norm": 0.751023530960083, + "learning_rate": 0.00015277596092108038, + "loss": 0.0826, + "step": 13787 + }, + { + "epoch": 0.3232124334112766, + "grad_norm": 3.4170548915863037, + "learning_rate": 0.00015276970400925346, + "loss": 1.59, + "step": 13788 + }, + { + "epoch": 0.3232358749860816, + "grad_norm": 3.098013401031494, + "learning_rate": 0.00015276344681109665, + "loss": 1.021, + "step": 13789 + }, + { + "epoch": 0.3232593165608866, + "grad_norm": 3.227059841156006, + "learning_rate": 0.0001527571893266439, + "loss": 0.3173, + "step": 13790 + }, + { + "epoch": 0.3232827581356916, + "grad_norm": 3.233159065246582, + "learning_rate": 0.00015275093155592916, + "loss": 0.2276, + "step": 13791 + }, + { + "epoch": 0.32330619971049657, + "grad_norm": 2.3025782108306885, + "learning_rate": 0.0001527446734989864, + "loss": 0.1521, + "step": 13792 + }, + { + "epoch": 0.32332964128530156, + "grad_norm": 4.159168243408203, + "learning_rate": 0.00015273841515584956, + "loss": 0.5895, + "step": 13793 + }, + { + "epoch": 0.32335308286010656, + "grad_norm": 0.9747608304023743, + "learning_rate": 0.00015273215652655253, + "loss": 0.159, + "step": 13794 + }, + { + "epoch": 0.32337652443491155, + "grad_norm": 0.930255115032196, + "learning_rate": 0.0001527258976111294, + "loss": 0.0852, + "step": 13795 + }, + { + "epoch": 0.32339996600971654, + "grad_norm": 2.7926883697509766, + "learning_rate": 0.00015271963840961407, + "loss": 0.1527, + "step": 13796 + }, + { + "epoch": 0.32342340758452154, + "grad_norm": 3.198315382003784, + "learning_rate": 0.0001527133789220405, + "loss": 0.5706, + "step": 13797 + }, + { + "epoch": 0.32344684915932653, + "grad_norm": 2.725553274154663, + "learning_rate": 0.0001527071191484426, + "loss": 1.4912, + "step": 13798 + }, + { + "epoch": 0.3234702907341315, + "grad_norm": 2.330535888671875, + "learning_rate": 0.00015270085908885444, + "loss": 0.1441, + "step": 13799 + }, + { + "epoch": 0.3234937323089365, + "grad_norm": 0.4991436302661896, + "learning_rate": 0.0001526945987433099, + "loss": 0.0551, + "step": 13800 + }, + { + "epoch": 0.3235171738837415, + "grad_norm": 3.0660481452941895, + "learning_rate": 0.000152688338111843, + "loss": 0.2178, + "step": 13801 + }, + { + "epoch": 0.3235406154585465, + "grad_norm": 1.2315082550048828, + "learning_rate": 0.00015268207719448774, + "loss": 0.1461, + "step": 13802 + }, + { + "epoch": 0.3235640570333515, + "grad_norm": 1.2185311317443848, + "learning_rate": 0.000152675815991278, + "loss": 0.1418, + "step": 13803 + }, + { + "epoch": 0.3235874986081565, + "grad_norm": 3.7779083251953125, + "learning_rate": 0.0001526695545022478, + "loss": 0.6013, + "step": 13804 + }, + { + "epoch": 0.3236109401829615, + "grad_norm": 4.220991611480713, + "learning_rate": 0.00015266329272743115, + "loss": 0.4024, + "step": 13805 + }, + { + "epoch": 0.3236343817577665, + "grad_norm": 3.351454019546509, + "learning_rate": 0.00015265703066686195, + "loss": 0.2693, + "step": 13806 + }, + { + "epoch": 0.3236578233325715, + "grad_norm": 3.251302719116211, + "learning_rate": 0.00015265076832057423, + "loss": 0.2369, + "step": 13807 + }, + { + "epoch": 0.3236812649073765, + "grad_norm": 0.6294180154800415, + "learning_rate": 0.00015264450568860198, + "loss": 0.0516, + "step": 13808 + }, + { + "epoch": 0.32370470648218147, + "grad_norm": 4.438886642456055, + "learning_rate": 0.00015263824277097915, + "loss": 0.4223, + "step": 13809 + }, + { + "epoch": 0.32372814805698646, + "grad_norm": 3.6408939361572266, + "learning_rate": 0.00015263197956773974, + "loss": 0.2715, + "step": 13810 + }, + { + "epoch": 0.32375158963179146, + "grad_norm": 3.1364030838012695, + "learning_rate": 0.00015262571607891774, + "loss": 0.2766, + "step": 13811 + }, + { + "epoch": 0.32377503120659645, + "grad_norm": 0.6816038489341736, + "learning_rate": 0.00015261945230454708, + "loss": 0.0344, + "step": 13812 + }, + { + "epoch": 0.32379847278140145, + "grad_norm": 2.7698426246643066, + "learning_rate": 0.00015261318824466184, + "loss": 0.3218, + "step": 13813 + }, + { + "epoch": 0.32382191435620644, + "grad_norm": 1.7964725494384766, + "learning_rate": 0.00015260692389929594, + "loss": 0.1133, + "step": 13814 + }, + { + "epoch": 0.32384535593101144, + "grad_norm": 5.0137810707092285, + "learning_rate": 0.00015260065926848342, + "loss": 0.4978, + "step": 13815 + }, + { + "epoch": 0.32386879750581643, + "grad_norm": 10.699932098388672, + "learning_rate": 0.00015259439435225821, + "loss": 0.2877, + "step": 13816 + }, + { + "epoch": 0.3238922390806214, + "grad_norm": 6.1157917976379395, + "learning_rate": 0.00015258812915065433, + "loss": 0.6465, + "step": 13817 + }, + { + "epoch": 0.3239156806554264, + "grad_norm": 4.371198654174805, + "learning_rate": 0.00015258186366370584, + "loss": 0.4763, + "step": 13818 + }, + { + "epoch": 0.3239391222302314, + "grad_norm": 0.9529128074645996, + "learning_rate": 0.00015257559789144662, + "loss": 0.0486, + "step": 13819 + }, + { + "epoch": 0.3239625638050364, + "grad_norm": 6.599621772766113, + "learning_rate": 0.00015256933183391077, + "loss": 0.4236, + "step": 13820 + }, + { + "epoch": 0.3239860053798414, + "grad_norm": 3.1381757259368896, + "learning_rate": 0.00015256306549113222, + "loss": 0.178, + "step": 13821 + }, + { + "epoch": 0.3240094469546464, + "grad_norm": 3.5639302730560303, + "learning_rate": 0.000152556798863145, + "loss": 0.3791, + "step": 13822 + }, + { + "epoch": 0.3240328885294514, + "grad_norm": 3.930725336074829, + "learning_rate": 0.0001525505319499831, + "loss": 0.4273, + "step": 13823 + }, + { + "epoch": 0.3240563301042564, + "grad_norm": 4.661154270172119, + "learning_rate": 0.00015254426475168055, + "loss": 0.4033, + "step": 13824 + }, + { + "epoch": 0.3240797716790614, + "grad_norm": 1.60591459274292, + "learning_rate": 0.00015253799726827136, + "loss": 0.1751, + "step": 13825 + }, + { + "epoch": 0.3241032132538664, + "grad_norm": 13.988275527954102, + "learning_rate": 0.00015253172949978948, + "loss": 0.457, + "step": 13826 + }, + { + "epoch": 0.32412665482867137, + "grad_norm": 1.5284260511398315, + "learning_rate": 0.000152525461446269, + "loss": 0.0938, + "step": 13827 + }, + { + "epoch": 0.32415009640347636, + "grad_norm": 0.7947432994842529, + "learning_rate": 0.00015251919310774387, + "loss": 0.0712, + "step": 13828 + }, + { + "epoch": 0.32417353797828136, + "grad_norm": 2.2901113033294678, + "learning_rate": 0.00015251292448424813, + "loss": 0.2913, + "step": 13829 + }, + { + "epoch": 0.32419697955308635, + "grad_norm": 2.120466470718384, + "learning_rate": 0.00015250665557581578, + "loss": 0.1164, + "step": 13830 + }, + { + "epoch": 0.32422042112789135, + "grad_norm": 1.129356861114502, + "learning_rate": 0.00015250038638248083, + "loss": 0.1402, + "step": 13831 + }, + { + "epoch": 0.3242438627026964, + "grad_norm": 1.4003300666809082, + "learning_rate": 0.0001524941169042773, + "loss": 0.1254, + "step": 13832 + }, + { + "epoch": 0.3242673042775014, + "grad_norm": 4.008338451385498, + "learning_rate": 0.00015248784714123924, + "loss": 0.2379, + "step": 13833 + }, + { + "epoch": 0.3242907458523064, + "grad_norm": 1.3320164680480957, + "learning_rate": 0.00015248157709340065, + "loss": 0.1199, + "step": 13834 + }, + { + "epoch": 0.3243141874271114, + "grad_norm": 4.85890007019043, + "learning_rate": 0.00015247530676079554, + "loss": 0.4848, + "step": 13835 + }, + { + "epoch": 0.32433762900191637, + "grad_norm": 2.424010753631592, + "learning_rate": 0.00015246903614345793, + "loss": 0.2544, + "step": 13836 + }, + { + "epoch": 0.32436107057672137, + "grad_norm": 5.449301242828369, + "learning_rate": 0.00015246276524142184, + "loss": 0.7099, + "step": 13837 + }, + { + "epoch": 0.32438451215152636, + "grad_norm": 3.3905246257781982, + "learning_rate": 0.00015245649405472132, + "loss": 0.3277, + "step": 13838 + }, + { + "epoch": 0.32440795372633136, + "grad_norm": 1.603990077972412, + "learning_rate": 0.0001524502225833904, + "loss": 0.2454, + "step": 13839 + }, + { + "epoch": 0.32443139530113635, + "grad_norm": 2.8874518871307373, + "learning_rate": 0.00015244395082746311, + "loss": 0.1169, + "step": 13840 + }, + { + "epoch": 0.32445483687594134, + "grad_norm": 3.3403780460357666, + "learning_rate": 0.00015243767878697347, + "loss": 0.2298, + "step": 13841 + }, + { + "epoch": 0.32447827845074634, + "grad_norm": 4.0737996101379395, + "learning_rate": 0.00015243140646195547, + "loss": 0.3533, + "step": 13842 + }, + { + "epoch": 0.32450172002555133, + "grad_norm": 5.0775885581970215, + "learning_rate": 0.0001524251338524432, + "loss": 0.4693, + "step": 13843 + }, + { + "epoch": 0.3245251616003563, + "grad_norm": 4.753305435180664, + "learning_rate": 0.0001524188609584707, + "loss": 0.3939, + "step": 13844 + }, + { + "epoch": 0.3245486031751613, + "grad_norm": 4.858524799346924, + "learning_rate": 0.000152412587780072, + "loss": 0.6011, + "step": 13845 + }, + { + "epoch": 0.3245720447499663, + "grad_norm": 6.543729782104492, + "learning_rate": 0.00015240631431728108, + "loss": 0.2413, + "step": 13846 + }, + { + "epoch": 0.3245954863247713, + "grad_norm": 6.104820251464844, + "learning_rate": 0.00015240004057013207, + "loss": 0.5905, + "step": 13847 + }, + { + "epoch": 0.3246189278995763, + "grad_norm": 0.9222127795219421, + "learning_rate": 0.0001523937665386589, + "loss": 0.0989, + "step": 13848 + }, + { + "epoch": 0.3246423694743813, + "grad_norm": 3.980600118637085, + "learning_rate": 0.00015238749222289575, + "loss": 0.4507, + "step": 13849 + }, + { + "epoch": 0.3246658110491863, + "grad_norm": 1.7660402059555054, + "learning_rate": 0.00015238121762287655, + "loss": 0.2463, + "step": 13850 + }, + { + "epoch": 0.3246892526239913, + "grad_norm": 2.6076087951660156, + "learning_rate": 0.0001523749427386354, + "loss": 0.2182, + "step": 13851 + }, + { + "epoch": 0.3247126941987963, + "grad_norm": 4.34135103225708, + "learning_rate": 0.00015236866757020634, + "loss": 0.3592, + "step": 13852 + }, + { + "epoch": 0.3247361357736013, + "grad_norm": 2.304039716720581, + "learning_rate": 0.00015236239211762344, + "loss": 0.2529, + "step": 13853 + }, + { + "epoch": 0.32475957734840627, + "grad_norm": 6.435457229614258, + "learning_rate": 0.00015235611638092072, + "loss": 0.3914, + "step": 13854 + }, + { + "epoch": 0.32478301892321126, + "grad_norm": 1.180799126625061, + "learning_rate": 0.0001523498403601322, + "loss": 0.136, + "step": 13855 + }, + { + "epoch": 0.32480646049801626, + "grad_norm": 4.283947467803955, + "learning_rate": 0.00015234356405529202, + "loss": 0.1758, + "step": 13856 + }, + { + "epoch": 0.32482990207282125, + "grad_norm": 1.76919424533844, + "learning_rate": 0.00015233728746643418, + "loss": 0.2549, + "step": 13857 + }, + { + "epoch": 0.32485334364762625, + "grad_norm": 3.5061206817626953, + "learning_rate": 0.00015233101059359272, + "loss": 0.3432, + "step": 13858 + }, + { + "epoch": 0.32487678522243124, + "grad_norm": 5.506975173950195, + "learning_rate": 0.00015232473343680176, + "loss": 0.6546, + "step": 13859 + }, + { + "epoch": 0.32490022679723624, + "grad_norm": 1.3062039613723755, + "learning_rate": 0.00015231845599609532, + "loss": 0.1912, + "step": 13860 + }, + { + "epoch": 0.32492366837204123, + "grad_norm": 1.269080638885498, + "learning_rate": 0.00015231217827150743, + "loss": 0.2008, + "step": 13861 + }, + { + "epoch": 0.3249471099468462, + "grad_norm": 1.7425453662872314, + "learning_rate": 0.00015230590026307222, + "loss": 0.2327, + "step": 13862 + }, + { + "epoch": 0.3249705515216512, + "grad_norm": 3.686553716659546, + "learning_rate": 0.00015229962197082372, + "loss": 0.4528, + "step": 13863 + }, + { + "epoch": 0.3249939930964562, + "grad_norm": 2.2740275859832764, + "learning_rate": 0.00015229334339479598, + "loss": 0.3124, + "step": 13864 + }, + { + "epoch": 0.3250174346712612, + "grad_norm": 3.7668497562408447, + "learning_rate": 0.00015228706453502313, + "loss": 0.321, + "step": 13865 + }, + { + "epoch": 0.3250408762460662, + "grad_norm": 4.057259559631348, + "learning_rate": 0.0001522807853915392, + "loss": 0.4055, + "step": 13866 + }, + { + "epoch": 0.3250643178208712, + "grad_norm": 4.5684051513671875, + "learning_rate": 0.0001522745059643782, + "loss": 0.7099, + "step": 13867 + }, + { + "epoch": 0.3250877593956762, + "grad_norm": 4.6358323097229, + "learning_rate": 0.0001522682262535743, + "loss": 0.5325, + "step": 13868 + }, + { + "epoch": 0.3251112009704812, + "grad_norm": 1.4406682252883911, + "learning_rate": 0.00015226194625916155, + "loss": 0.277, + "step": 13869 + }, + { + "epoch": 0.3251346425452862, + "grad_norm": 0.48966652154922485, + "learning_rate": 0.00015225566598117397, + "loss": 0.0524, + "step": 13870 + }, + { + "epoch": 0.3251580841200912, + "grad_norm": 2.8991196155548096, + "learning_rate": 0.00015224938541964571, + "loss": 0.3095, + "step": 13871 + }, + { + "epoch": 0.32518152569489617, + "grad_norm": 1.2040010690689087, + "learning_rate": 0.0001522431045746108, + "loss": 0.1296, + "step": 13872 + }, + { + "epoch": 0.32520496726970116, + "grad_norm": 3.076206684112549, + "learning_rate": 0.00015223682344610336, + "loss": 0.3378, + "step": 13873 + }, + { + "epoch": 0.32522840884450616, + "grad_norm": 3.986833333969116, + "learning_rate": 0.0001522305420341574, + "loss": 0.4318, + "step": 13874 + }, + { + "epoch": 0.32525185041931115, + "grad_norm": 3.801870346069336, + "learning_rate": 0.0001522242603388071, + "loss": 0.5001, + "step": 13875 + }, + { + "epoch": 0.32527529199411614, + "grad_norm": 1.681816816329956, + "learning_rate": 0.00015221797836008648, + "loss": 0.1069, + "step": 13876 + }, + { + "epoch": 0.32529873356892114, + "grad_norm": 3.277693510055542, + "learning_rate": 0.00015221169609802964, + "loss": 0.3624, + "step": 13877 + }, + { + "epoch": 0.32532217514372613, + "grad_norm": 4.4493842124938965, + "learning_rate": 0.00015220541355267066, + "loss": 0.5109, + "step": 13878 + }, + { + "epoch": 0.3253456167185311, + "grad_norm": 4.6806745529174805, + "learning_rate": 0.00015219913072404362, + "loss": 0.3359, + "step": 13879 + }, + { + "epoch": 0.3253690582933361, + "grad_norm": 1.0732743740081787, + "learning_rate": 0.00015219284761218264, + "loss": 0.1145, + "step": 13880 + }, + { + "epoch": 0.3253924998681411, + "grad_norm": 0.5110541582107544, + "learning_rate": 0.00015218656421712182, + "loss": 0.041, + "step": 13881 + }, + { + "epoch": 0.3254159414429461, + "grad_norm": 3.693774461746216, + "learning_rate": 0.00015218028053889522, + "loss": 0.5265, + "step": 13882 + }, + { + "epoch": 0.3254393830177511, + "grad_norm": 3.158184289932251, + "learning_rate": 0.00015217399657753696, + "loss": 0.3327, + "step": 13883 + }, + { + "epoch": 0.32546282459255615, + "grad_norm": 5.217556476593018, + "learning_rate": 0.00015216771233308115, + "loss": 0.4111, + "step": 13884 + }, + { + "epoch": 0.32548626616736115, + "grad_norm": 3.3102262020111084, + "learning_rate": 0.00015216142780556183, + "loss": 0.4438, + "step": 13885 + }, + { + "epoch": 0.32550970774216614, + "grad_norm": 4.931548595428467, + "learning_rate": 0.0001521551429950131, + "loss": 0.4737, + "step": 13886 + }, + { + "epoch": 0.32553314931697114, + "grad_norm": 5.025437831878662, + "learning_rate": 0.00015214885790146916, + "loss": 0.9785, + "step": 13887 + }, + { + "epoch": 0.32555659089177613, + "grad_norm": 0.7894487977027893, + "learning_rate": 0.00015214257252496406, + "loss": 0.0544, + "step": 13888 + }, + { + "epoch": 0.3255800324665811, + "grad_norm": 1.2488365173339844, + "learning_rate": 0.00015213628686553188, + "loss": 0.1319, + "step": 13889 + }, + { + "epoch": 0.3256034740413861, + "grad_norm": 5.862491130828857, + "learning_rate": 0.00015213000092320674, + "loss": 0.2339, + "step": 13890 + }, + { + "epoch": 0.3256269156161911, + "grad_norm": 0.6040080189704895, + "learning_rate": 0.0001521237146980227, + "loss": 0.0431, + "step": 13891 + }, + { + "epoch": 0.3256503571909961, + "grad_norm": 4.055874347686768, + "learning_rate": 0.00015211742819001397, + "loss": 0.3481, + "step": 13892 + }, + { + "epoch": 0.3256737987658011, + "grad_norm": 4.045163631439209, + "learning_rate": 0.0001521111413992146, + "loss": 0.6276, + "step": 13893 + }, + { + "epoch": 0.3256972403406061, + "grad_norm": 6.058506965637207, + "learning_rate": 0.00015210485432565874, + "loss": 0.6354, + "step": 13894 + }, + { + "epoch": 0.3257206819154111, + "grad_norm": 1.3748831748962402, + "learning_rate": 0.00015209856696938044, + "loss": 0.1385, + "step": 13895 + }, + { + "epoch": 0.3257441234902161, + "grad_norm": 2.0841290950775146, + "learning_rate": 0.00015209227933041385, + "loss": 0.2647, + "step": 13896 + }, + { + "epoch": 0.3257675650650211, + "grad_norm": 3.232659339904785, + "learning_rate": 0.0001520859914087931, + "loss": 0.5004, + "step": 13897 + }, + { + "epoch": 0.3257910066398261, + "grad_norm": 2.255993127822876, + "learning_rate": 0.00015207970320455226, + "loss": 0.2644, + "step": 13898 + }, + { + "epoch": 0.32581444821463107, + "grad_norm": 2.4312548637390137, + "learning_rate": 0.0001520734147177255, + "loss": 0.2244, + "step": 13899 + }, + { + "epoch": 0.32583788978943606, + "grad_norm": 3.5122439861297607, + "learning_rate": 0.00015206712594834697, + "loss": 0.4179, + "step": 13900 + }, + { + "epoch": 0.32586133136424106, + "grad_norm": 4.837570667266846, + "learning_rate": 0.00015206083689645072, + "loss": 0.626, + "step": 13901 + }, + { + "epoch": 0.32588477293904605, + "grad_norm": 1.8568681478500366, + "learning_rate": 0.00015205454756207093, + "loss": 0.1279, + "step": 13902 + }, + { + "epoch": 0.32590821451385105, + "grad_norm": 4.037210941314697, + "learning_rate": 0.00015204825794524167, + "loss": 0.5567, + "step": 13903 + }, + { + "epoch": 0.32593165608865604, + "grad_norm": 7.247663974761963, + "learning_rate": 0.0001520419680459971, + "loss": 0.4525, + "step": 13904 + }, + { + "epoch": 0.32595509766346104, + "grad_norm": 5.08162260055542, + "learning_rate": 0.00015203567786437137, + "loss": 1.7689, + "step": 13905 + }, + { + "epoch": 0.32597853923826603, + "grad_norm": 3.5920462608337402, + "learning_rate": 0.00015202938740039858, + "loss": 0.2505, + "step": 13906 + }, + { + "epoch": 0.326001980813071, + "grad_norm": 2.7734785079956055, + "learning_rate": 0.00015202309665411287, + "loss": 1.5218, + "step": 13907 + }, + { + "epoch": 0.326025422387876, + "grad_norm": 2.680009126663208, + "learning_rate": 0.00015201680562554833, + "loss": 0.246, + "step": 13908 + }, + { + "epoch": 0.326048863962681, + "grad_norm": 1.5705024003982544, + "learning_rate": 0.00015201051431473924, + "loss": 0.1535, + "step": 13909 + }, + { + "epoch": 0.326072305537486, + "grad_norm": 2.271479368209839, + "learning_rate": 0.00015200422272171956, + "loss": 0.2714, + "step": 13910 + }, + { + "epoch": 0.326095747112291, + "grad_norm": 3.782212018966675, + "learning_rate": 0.00015199793084652352, + "loss": 0.5182, + "step": 13911 + }, + { + "epoch": 0.326119188687096, + "grad_norm": 0.4990271031856537, + "learning_rate": 0.00015199163868918525, + "loss": 0.0415, + "step": 13912 + }, + { + "epoch": 0.326142630261901, + "grad_norm": 2.4469778537750244, + "learning_rate": 0.00015198534624973888, + "loss": 0.2263, + "step": 13913 + }, + { + "epoch": 0.326166071836706, + "grad_norm": 2.1174604892730713, + "learning_rate": 0.00015197905352821856, + "loss": 0.2822, + "step": 13914 + }, + { + "epoch": 0.326189513411511, + "grad_norm": 1.2318572998046875, + "learning_rate": 0.00015197276052465844, + "loss": 0.1823, + "step": 13915 + }, + { + "epoch": 0.326212954986316, + "grad_norm": 2.192375421524048, + "learning_rate": 0.00015196646723909267, + "loss": 1.9413, + "step": 13916 + }, + { + "epoch": 0.32623639656112097, + "grad_norm": 3.0457041263580322, + "learning_rate": 0.00015196017367155538, + "loss": 0.3576, + "step": 13917 + }, + { + "epoch": 0.32625983813592596, + "grad_norm": 0.7780335545539856, + "learning_rate": 0.00015195387982208072, + "loss": 0.1048, + "step": 13918 + }, + { + "epoch": 0.32628327971073096, + "grad_norm": 0.882279634475708, + "learning_rate": 0.00015194758569070285, + "loss": 0.068, + "step": 13919 + }, + { + "epoch": 0.32630672128553595, + "grad_norm": 2.567540168762207, + "learning_rate": 0.00015194129127745593, + "loss": 0.2827, + "step": 13920 + }, + { + "epoch": 0.32633016286034094, + "grad_norm": 0.4146391749382019, + "learning_rate": 0.00015193499658237408, + "loss": 0.0442, + "step": 13921 + }, + { + "epoch": 0.32635360443514594, + "grad_norm": 6.463368892669678, + "learning_rate": 0.00015192870160549152, + "loss": 0.5541, + "step": 13922 + }, + { + "epoch": 0.32637704600995093, + "grad_norm": 4.064563751220703, + "learning_rate": 0.00015192240634684232, + "loss": 0.5875, + "step": 13923 + }, + { + "epoch": 0.3264004875847559, + "grad_norm": 1.3779069185256958, + "learning_rate": 0.0001519161108064607, + "loss": 0.104, + "step": 13924 + }, + { + "epoch": 0.3264239291595609, + "grad_norm": 0.5876948833465576, + "learning_rate": 0.00015190981498438083, + "loss": 0.064, + "step": 13925 + }, + { + "epoch": 0.3264473707343659, + "grad_norm": 3.8294248580932617, + "learning_rate": 0.0001519035188806368, + "loss": 1.4628, + "step": 13926 + }, + { + "epoch": 0.3264708123091709, + "grad_norm": 4.359628677368164, + "learning_rate": 0.00015189722249526284, + "loss": 0.8713, + "step": 13927 + }, + { + "epoch": 0.3264942538839759, + "grad_norm": 1.1324959993362427, + "learning_rate": 0.00015189092582829306, + "loss": 0.1441, + "step": 13928 + }, + { + "epoch": 0.3265176954587809, + "grad_norm": 4.851069927215576, + "learning_rate": 0.00015188462887976169, + "loss": 0.6604, + "step": 13929 + }, + { + "epoch": 0.3265411370335859, + "grad_norm": 2.916154384613037, + "learning_rate": 0.00015187833164970286, + "loss": 0.3587, + "step": 13930 + }, + { + "epoch": 0.3265645786083909, + "grad_norm": 2.7167294025421143, + "learning_rate": 0.00015187203413815075, + "loss": 0.3419, + "step": 13931 + }, + { + "epoch": 0.3265880201831959, + "grad_norm": 0.7777085900306702, + "learning_rate": 0.0001518657363451395, + "loss": 0.0607, + "step": 13932 + }, + { + "epoch": 0.3266114617580009, + "grad_norm": 3.641590118408203, + "learning_rate": 0.00015185943827070335, + "loss": 0.3132, + "step": 13933 + }, + { + "epoch": 0.32663490333280587, + "grad_norm": 4.063477993011475, + "learning_rate": 0.00015185313991487637, + "loss": 0.4243, + "step": 13934 + }, + { + "epoch": 0.3266583449076109, + "grad_norm": 4.463649749755859, + "learning_rate": 0.00015184684127769285, + "loss": 0.4949, + "step": 13935 + }, + { + "epoch": 0.3266817864824159, + "grad_norm": 3.110055923461914, + "learning_rate": 0.00015184054235918687, + "loss": 0.355, + "step": 13936 + }, + { + "epoch": 0.3267052280572209, + "grad_norm": 2.7741312980651855, + "learning_rate": 0.00015183424315939268, + "loss": 1.4258, + "step": 13937 + }, + { + "epoch": 0.3267286696320259, + "grad_norm": 0.726929247379303, + "learning_rate": 0.00015182794367834442, + "loss": 0.0755, + "step": 13938 + }, + { + "epoch": 0.3267521112068309, + "grad_norm": 1.2389384508132935, + "learning_rate": 0.00015182164391607628, + "loss": 0.1205, + "step": 13939 + }, + { + "epoch": 0.3267755527816359, + "grad_norm": 3.853469133377075, + "learning_rate": 0.00015181534387262243, + "loss": 0.2881, + "step": 13940 + }, + { + "epoch": 0.3267989943564409, + "grad_norm": 1.8703107833862305, + "learning_rate": 0.0001518090435480171, + "loss": 0.2041, + "step": 13941 + }, + { + "epoch": 0.3268224359312459, + "grad_norm": 2.2284634113311768, + "learning_rate": 0.00015180274294229442, + "loss": 0.2764, + "step": 13942 + }, + { + "epoch": 0.3268458775060509, + "grad_norm": 3.333174467086792, + "learning_rate": 0.00015179644205548856, + "loss": 0.4566, + "step": 13943 + }, + { + "epoch": 0.32686931908085587, + "grad_norm": 5.235617637634277, + "learning_rate": 0.00015179014088763384, + "loss": 0.4071, + "step": 13944 + }, + { + "epoch": 0.32689276065566086, + "grad_norm": 4.3085713386535645, + "learning_rate": 0.0001517838394387643, + "loss": 0.5001, + "step": 13945 + }, + { + "epoch": 0.32691620223046586, + "grad_norm": 3.4556338787078857, + "learning_rate": 0.0001517775377089142, + "loss": 0.3917, + "step": 13946 + }, + { + "epoch": 0.32693964380527085, + "grad_norm": 2.2407052516937256, + "learning_rate": 0.00015177123569811771, + "loss": 0.2595, + "step": 13947 + }, + { + "epoch": 0.32696308538007585, + "grad_norm": 3.560182809829712, + "learning_rate": 0.00015176493340640908, + "loss": 0.3161, + "step": 13948 + }, + { + "epoch": 0.32698652695488084, + "grad_norm": 4.011227607727051, + "learning_rate": 0.00015175863083382245, + "loss": 0.4493, + "step": 13949 + }, + { + "epoch": 0.32700996852968583, + "grad_norm": 6.560772895812988, + "learning_rate": 0.00015175232798039198, + "loss": 0.9496, + "step": 13950 + }, + { + "epoch": 0.32703341010449083, + "grad_norm": 2.304365634918213, + "learning_rate": 0.00015174602484615198, + "loss": 1.3945, + "step": 13951 + }, + { + "epoch": 0.3270568516792958, + "grad_norm": 3.6978511810302734, + "learning_rate": 0.00015173972143113658, + "loss": 0.3949, + "step": 13952 + }, + { + "epoch": 0.3270802932541008, + "grad_norm": 6.118159770965576, + "learning_rate": 0.00015173341773538, + "loss": 0.703, + "step": 13953 + }, + { + "epoch": 0.3271037348289058, + "grad_norm": 4.977223873138428, + "learning_rate": 0.00015172711375891645, + "loss": 0.6602, + "step": 13954 + }, + { + "epoch": 0.3271271764037108, + "grad_norm": 4.049535751342773, + "learning_rate": 0.00015172080950178013, + "loss": 0.5516, + "step": 13955 + }, + { + "epoch": 0.3271506179785158, + "grad_norm": 3.507117748260498, + "learning_rate": 0.0001517145049640052, + "loss": 0.3386, + "step": 13956 + }, + { + "epoch": 0.3271740595533208, + "grad_norm": 4.415619373321533, + "learning_rate": 0.00015170820014562595, + "loss": 0.3124, + "step": 13957 + }, + { + "epoch": 0.3271975011281258, + "grad_norm": 2.752565622329712, + "learning_rate": 0.00015170189504667655, + "loss": 0.2879, + "step": 13958 + }, + { + "epoch": 0.3272209427029308, + "grad_norm": 5.8954901695251465, + "learning_rate": 0.00015169558966719122, + "loss": 0.695, + "step": 13959 + }, + { + "epoch": 0.3272443842777358, + "grad_norm": 3.904287576675415, + "learning_rate": 0.0001516892840072041, + "loss": 0.494, + "step": 13960 + }, + { + "epoch": 0.3272678258525408, + "grad_norm": 2.214031219482422, + "learning_rate": 0.00015168297806674957, + "loss": 0.1256, + "step": 13961 + }, + { + "epoch": 0.32729126742734577, + "grad_norm": 3.276384115219116, + "learning_rate": 0.0001516766718458617, + "loss": 0.4048, + "step": 13962 + }, + { + "epoch": 0.32731470900215076, + "grad_norm": 2.5268852710723877, + "learning_rate": 0.00015167036534457476, + "loss": 0.223, + "step": 13963 + }, + { + "epoch": 0.32733815057695576, + "grad_norm": 4.603431701660156, + "learning_rate": 0.00015166405856292293, + "loss": 0.4324, + "step": 13964 + }, + { + "epoch": 0.32736159215176075, + "grad_norm": 0.8441375494003296, + "learning_rate": 0.00015165775150094053, + "loss": 0.075, + "step": 13965 + }, + { + "epoch": 0.32738503372656574, + "grad_norm": 3.397663116455078, + "learning_rate": 0.0001516514441586617, + "loss": 0.4228, + "step": 13966 + }, + { + "epoch": 0.32740847530137074, + "grad_norm": 3.0649161338806152, + "learning_rate": 0.00015164513653612068, + "loss": 0.3611, + "step": 13967 + }, + { + "epoch": 0.32743191687617573, + "grad_norm": 6.999721050262451, + "learning_rate": 0.00015163882863335166, + "loss": 0.6428, + "step": 13968 + }, + { + "epoch": 0.3274553584509807, + "grad_norm": 5.1591620445251465, + "learning_rate": 0.00015163252045038894, + "loss": 0.4464, + "step": 13969 + }, + { + "epoch": 0.3274788000257857, + "grad_norm": 4.164259433746338, + "learning_rate": 0.00015162621198726672, + "loss": 0.593, + "step": 13970 + }, + { + "epoch": 0.3275022416005907, + "grad_norm": 4.060969829559326, + "learning_rate": 0.00015161990324401918, + "loss": 0.4437, + "step": 13971 + }, + { + "epoch": 0.3275256831753957, + "grad_norm": 0.453186959028244, + "learning_rate": 0.00015161359422068062, + "loss": 0.0289, + "step": 13972 + }, + { + "epoch": 0.3275491247502007, + "grad_norm": 2.329835891723633, + "learning_rate": 0.00015160728491728524, + "loss": 0.2239, + "step": 13973 + }, + { + "epoch": 0.3275725663250057, + "grad_norm": 1.828062653541565, + "learning_rate": 0.00015160097533386727, + "loss": 0.1357, + "step": 13974 + }, + { + "epoch": 0.3275960078998107, + "grad_norm": 0.9496250152587891, + "learning_rate": 0.00015159466547046098, + "loss": 0.0798, + "step": 13975 + }, + { + "epoch": 0.3276194494746157, + "grad_norm": 4.5101213455200195, + "learning_rate": 0.00015158835532710061, + "loss": 0.4159, + "step": 13976 + }, + { + "epoch": 0.3276428910494207, + "grad_norm": 1.4610295295715332, + "learning_rate": 0.0001515820449038203, + "loss": 0.1471, + "step": 13977 + }, + { + "epoch": 0.3276663326242257, + "grad_norm": 0.4865866005420685, + "learning_rate": 0.00015157573420065444, + "loss": 0.0594, + "step": 13978 + }, + { + "epoch": 0.32768977419903067, + "grad_norm": 6.719733238220215, + "learning_rate": 0.00015156942321763715, + "loss": 0.481, + "step": 13979 + }, + { + "epoch": 0.32771321577383566, + "grad_norm": 0.412058562040329, + "learning_rate": 0.00015156311195480273, + "loss": 0.0435, + "step": 13980 + }, + { + "epoch": 0.32773665734864066, + "grad_norm": 3.1326231956481934, + "learning_rate": 0.0001515568004121854, + "loss": 1.5813, + "step": 13981 + }, + { + "epoch": 0.32776009892344565, + "grad_norm": 3.146165132522583, + "learning_rate": 0.00015155048858981945, + "loss": 0.384, + "step": 13982 + }, + { + "epoch": 0.32778354049825065, + "grad_norm": 3.225019931793213, + "learning_rate": 0.00015154417648773907, + "loss": 0.3343, + "step": 13983 + }, + { + "epoch": 0.32780698207305564, + "grad_norm": 2.8188161849975586, + "learning_rate": 0.00015153786410597857, + "loss": 0.3049, + "step": 13984 + }, + { + "epoch": 0.32783042364786064, + "grad_norm": 4.581936836242676, + "learning_rate": 0.00015153155144457213, + "loss": 0.5302, + "step": 13985 + }, + { + "epoch": 0.3278538652226657, + "grad_norm": 1.0255751609802246, + "learning_rate": 0.00015152523850355407, + "loss": 0.0962, + "step": 13986 + }, + { + "epoch": 0.3278773067974707, + "grad_norm": 1.485632061958313, + "learning_rate": 0.00015151892528295863, + "loss": 0.1141, + "step": 13987 + }, + { + "epoch": 0.3279007483722757, + "grad_norm": 1.5469801425933838, + "learning_rate": 0.00015151261178282004, + "loss": 0.2719, + "step": 13988 + }, + { + "epoch": 0.32792418994708067, + "grad_norm": 1.1755110025405884, + "learning_rate": 0.00015150629800317255, + "loss": 0.124, + "step": 13989 + }, + { + "epoch": 0.32794763152188566, + "grad_norm": 0.8123197555541992, + "learning_rate": 0.00015149998394405048, + "loss": 0.1104, + "step": 13990 + }, + { + "epoch": 0.32797107309669066, + "grad_norm": 6.32973051071167, + "learning_rate": 0.00015149366960548802, + "loss": 0.5018, + "step": 13991 + }, + { + "epoch": 0.32799451467149565, + "grad_norm": 4.008848190307617, + "learning_rate": 0.00015148735498751945, + "loss": 0.3825, + "step": 13992 + }, + { + "epoch": 0.32801795624630065, + "grad_norm": 2.79327392578125, + "learning_rate": 0.00015148104009017908, + "loss": 0.5419, + "step": 13993 + }, + { + "epoch": 0.32804139782110564, + "grad_norm": 2.039555788040161, + "learning_rate": 0.00015147472491350113, + "loss": 0.2454, + "step": 13994 + }, + { + "epoch": 0.32806483939591063, + "grad_norm": 3.5792572498321533, + "learning_rate": 0.00015146840945751984, + "loss": 0.3895, + "step": 13995 + }, + { + "epoch": 0.32808828097071563, + "grad_norm": 2.0650904178619385, + "learning_rate": 0.00015146209372226955, + "loss": 0.2163, + "step": 13996 + }, + { + "epoch": 0.3281117225455206, + "grad_norm": 2.7659006118774414, + "learning_rate": 0.00015145577770778448, + "loss": 0.2823, + "step": 13997 + }, + { + "epoch": 0.3281351641203256, + "grad_norm": 3.318108558654785, + "learning_rate": 0.0001514494614140989, + "loss": 2.0263, + "step": 13998 + }, + { + "epoch": 0.3281586056951306, + "grad_norm": 3.4570653438568115, + "learning_rate": 0.00015144314484124714, + "loss": 0.5732, + "step": 13999 + }, + { + "epoch": 0.3281820472699356, + "grad_norm": 5.106639862060547, + "learning_rate": 0.00015143682798926338, + "loss": 0.6033, + "step": 14000 + }, + { + "epoch": 0.3282054888447406, + "grad_norm": 2.0842368602752686, + "learning_rate": 0.00015143051085818193, + "loss": 1.0644, + "step": 14001 + }, + { + "epoch": 0.3282289304195456, + "grad_norm": 2.6420722007751465, + "learning_rate": 0.00015142419344803712, + "loss": 0.2587, + "step": 14002 + }, + { + "epoch": 0.3282523719943506, + "grad_norm": 5.735295295715332, + "learning_rate": 0.0001514178757588632, + "loss": 0.7827, + "step": 14003 + }, + { + "epoch": 0.3282758135691556, + "grad_norm": 5.258493423461914, + "learning_rate": 0.00015141155779069442, + "loss": 0.2556, + "step": 14004 + }, + { + "epoch": 0.3282992551439606, + "grad_norm": 5.3416314125061035, + "learning_rate": 0.0001514052395435651, + "loss": 0.5647, + "step": 14005 + }, + { + "epoch": 0.32832269671876557, + "grad_norm": 3.469557285308838, + "learning_rate": 0.0001513989210175095, + "loss": 1.3947, + "step": 14006 + }, + { + "epoch": 0.32834613829357057, + "grad_norm": 3.721405029296875, + "learning_rate": 0.00015139260221256188, + "loss": 0.3107, + "step": 14007 + }, + { + "epoch": 0.32836957986837556, + "grad_norm": 4.313159465789795, + "learning_rate": 0.0001513862831287566, + "loss": 0.4907, + "step": 14008 + }, + { + "epoch": 0.32839302144318055, + "grad_norm": 2.8305811882019043, + "learning_rate": 0.00015137996376612787, + "loss": 0.2539, + "step": 14009 + }, + { + "epoch": 0.32841646301798555, + "grad_norm": 5.313089370727539, + "learning_rate": 0.00015137364412471003, + "loss": 0.3455, + "step": 14010 + }, + { + "epoch": 0.32843990459279054, + "grad_norm": 3.008512258529663, + "learning_rate": 0.00015136732420453734, + "loss": 0.2199, + "step": 14011 + }, + { + "epoch": 0.32846334616759554, + "grad_norm": 3.1250693798065186, + "learning_rate": 0.00015136100400564413, + "loss": 1.9525, + "step": 14012 + }, + { + "epoch": 0.32848678774240053, + "grad_norm": 3.1688296794891357, + "learning_rate": 0.00015135468352806464, + "loss": 0.3795, + "step": 14013 + }, + { + "epoch": 0.3285102293172055, + "grad_norm": 1.6318351030349731, + "learning_rate": 0.0001513483627718332, + "loss": 0.1784, + "step": 14014 + }, + { + "epoch": 0.3285336708920105, + "grad_norm": 3.0588061809539795, + "learning_rate": 0.0001513420417369841, + "loss": 0.3356, + "step": 14015 + }, + { + "epoch": 0.3285571124668155, + "grad_norm": 5.117209434509277, + "learning_rate": 0.00015133572042355163, + "loss": 0.5273, + "step": 14016 + }, + { + "epoch": 0.3285805540416205, + "grad_norm": 2.5111303329467773, + "learning_rate": 0.0001513293988315701, + "loss": 1.5792, + "step": 14017 + }, + { + "epoch": 0.3286039956164255, + "grad_norm": 1.5346405506134033, + "learning_rate": 0.00015132307696107383, + "loss": 0.2104, + "step": 14018 + }, + { + "epoch": 0.3286274371912305, + "grad_norm": 4.802087783813477, + "learning_rate": 0.00015131675481209708, + "loss": 0.4907, + "step": 14019 + }, + { + "epoch": 0.3286508787660355, + "grad_norm": 1.2369722127914429, + "learning_rate": 0.00015131043238467417, + "loss": 0.219, + "step": 14020 + }, + { + "epoch": 0.3286743203408405, + "grad_norm": 4.15054988861084, + "learning_rate": 0.00015130410967883944, + "loss": 0.425, + "step": 14021 + }, + { + "epoch": 0.3286977619156455, + "grad_norm": 0.8227483630180359, + "learning_rate": 0.00015129778669462716, + "loss": 0.0672, + "step": 14022 + }, + { + "epoch": 0.3287212034904505, + "grad_norm": 7.7953715324401855, + "learning_rate": 0.00015129146343207165, + "loss": 0.7713, + "step": 14023 + }, + { + "epoch": 0.32874464506525547, + "grad_norm": 3.483618974685669, + "learning_rate": 0.00015128513989120718, + "loss": 0.168, + "step": 14024 + }, + { + "epoch": 0.32876808664006046, + "grad_norm": 4.163700580596924, + "learning_rate": 0.00015127881607206813, + "loss": 0.5551, + "step": 14025 + }, + { + "epoch": 0.32879152821486546, + "grad_norm": 3.307406187057495, + "learning_rate": 0.00015127249197468876, + "loss": 0.492, + "step": 14026 + }, + { + "epoch": 0.32881496978967045, + "grad_norm": 0.6301215291023254, + "learning_rate": 0.00015126616759910343, + "loss": 0.0834, + "step": 14027 + }, + { + "epoch": 0.32883841136447545, + "grad_norm": 2.8275375366210938, + "learning_rate": 0.00015125984294534644, + "loss": 0.3411, + "step": 14028 + }, + { + "epoch": 0.32886185293928044, + "grad_norm": 3.7911722660064697, + "learning_rate": 0.00015125351801345208, + "loss": 0.4838, + "step": 14029 + }, + { + "epoch": 0.32888529451408544, + "grad_norm": 4.837886810302734, + "learning_rate": 0.00015124719280345473, + "loss": 0.4028, + "step": 14030 + }, + { + "epoch": 0.32890873608889043, + "grad_norm": 3.7278850078582764, + "learning_rate": 0.0001512408673153886, + "loss": 0.4628, + "step": 14031 + }, + { + "epoch": 0.3289321776636954, + "grad_norm": 1.247462511062622, + "learning_rate": 0.00015123454154928814, + "loss": 0.1456, + "step": 14032 + }, + { + "epoch": 0.3289556192385004, + "grad_norm": 3.377774238586426, + "learning_rate": 0.00015122821550518758, + "loss": 0.3951, + "step": 14033 + }, + { + "epoch": 0.3289790608133054, + "grad_norm": 3.9947690963745117, + "learning_rate": 0.00015122188918312133, + "loss": 0.3455, + "step": 14034 + }, + { + "epoch": 0.3290025023881104, + "grad_norm": 2.5912137031555176, + "learning_rate": 0.00015121556258312362, + "loss": 0.2579, + "step": 14035 + }, + { + "epoch": 0.3290259439629154, + "grad_norm": 2.802488327026367, + "learning_rate": 0.00015120923570522888, + "loss": 0.2499, + "step": 14036 + }, + { + "epoch": 0.32904938553772045, + "grad_norm": 3.597374439239502, + "learning_rate": 0.00015120290854947133, + "loss": 0.3865, + "step": 14037 + }, + { + "epoch": 0.32907282711252545, + "grad_norm": 1.4052038192749023, + "learning_rate": 0.0001511965811158854, + "loss": 0.1812, + "step": 14038 + }, + { + "epoch": 0.32909626868733044, + "grad_norm": 5.38033390045166, + "learning_rate": 0.00015119025340450533, + "loss": 0.7715, + "step": 14039 + }, + { + "epoch": 0.32911971026213543, + "grad_norm": 5.375890731811523, + "learning_rate": 0.00015118392541536556, + "loss": 0.5926, + "step": 14040 + }, + { + "epoch": 0.32914315183694043, + "grad_norm": 1.4485691785812378, + "learning_rate": 0.00015117759714850036, + "loss": 0.1563, + "step": 14041 + }, + { + "epoch": 0.3291665934117454, + "grad_norm": 3.7531957626342773, + "learning_rate": 0.00015117126860394407, + "loss": 0.4354, + "step": 14042 + }, + { + "epoch": 0.3291900349865504, + "grad_norm": 0.6731376051902771, + "learning_rate": 0.000151164939781731, + "loss": 0.0528, + "step": 14043 + }, + { + "epoch": 0.3292134765613554, + "grad_norm": 3.401759386062622, + "learning_rate": 0.00015115861068189557, + "loss": 0.2788, + "step": 14044 + }, + { + "epoch": 0.3292369181361604, + "grad_norm": 6.010556697845459, + "learning_rate": 0.00015115228130447209, + "loss": 0.5741, + "step": 14045 + }, + { + "epoch": 0.3292603597109654, + "grad_norm": 4.2221245765686035, + "learning_rate": 0.0001511459516494949, + "loss": 0.416, + "step": 14046 + }, + { + "epoch": 0.3292838012857704, + "grad_norm": 1.1879420280456543, + "learning_rate": 0.0001511396217169983, + "loss": 0.092, + "step": 14047 + }, + { + "epoch": 0.3293072428605754, + "grad_norm": 2.325056552886963, + "learning_rate": 0.0001511332915070167, + "loss": 0.234, + "step": 14048 + }, + { + "epoch": 0.3293306844353804, + "grad_norm": 4.699248313903809, + "learning_rate": 0.0001511269610195844, + "loss": 0.5034, + "step": 14049 + }, + { + "epoch": 0.3293541260101854, + "grad_norm": 4.0418829917907715, + "learning_rate": 0.0001511206302547358, + "loss": 0.4272, + "step": 14050 + }, + { + "epoch": 0.32937756758499037, + "grad_norm": 0.5498937964439392, + "learning_rate": 0.00015111429921250524, + "loss": 0.0508, + "step": 14051 + }, + { + "epoch": 0.32940100915979537, + "grad_norm": 1.0997939109802246, + "learning_rate": 0.00015110796789292698, + "loss": 0.1103, + "step": 14052 + }, + { + "epoch": 0.32942445073460036, + "grad_norm": 3.686858654022217, + "learning_rate": 0.00015110163629603553, + "loss": 0.3831, + "step": 14053 + }, + { + "epoch": 0.32944789230940535, + "grad_norm": 3.4175233840942383, + "learning_rate": 0.00015109530442186512, + "loss": 0.4108, + "step": 14054 + }, + { + "epoch": 0.32947133388421035, + "grad_norm": 2.479045867919922, + "learning_rate": 0.00015108897227045015, + "loss": 0.215, + "step": 14055 + }, + { + "epoch": 0.32949477545901534, + "grad_norm": 5.5799880027771, + "learning_rate": 0.00015108263984182498, + "loss": 0.4659, + "step": 14056 + }, + { + "epoch": 0.32951821703382034, + "grad_norm": 3.029386043548584, + "learning_rate": 0.00015107630713602399, + "loss": 0.2214, + "step": 14057 + }, + { + "epoch": 0.32954165860862533, + "grad_norm": 3.0234670639038086, + "learning_rate": 0.00015106997415308154, + "loss": 0.2376, + "step": 14058 + }, + { + "epoch": 0.3295651001834303, + "grad_norm": 3.753403663635254, + "learning_rate": 0.00015106364089303192, + "loss": 0.2507, + "step": 14059 + }, + { + "epoch": 0.3295885417582353, + "grad_norm": 3.2715649604797363, + "learning_rate": 0.0001510573073559096, + "loss": 0.3925, + "step": 14060 + }, + { + "epoch": 0.3296119833330403, + "grad_norm": 1.0100420713424683, + "learning_rate": 0.00015105097354174886, + "loss": 0.0985, + "step": 14061 + }, + { + "epoch": 0.3296354249078453, + "grad_norm": 2.4797849655151367, + "learning_rate": 0.0001510446394505841, + "loss": 0.2624, + "step": 14062 + }, + { + "epoch": 0.3296588664826503, + "grad_norm": 4.161871433258057, + "learning_rate": 0.0001510383050824497, + "loss": 0.3463, + "step": 14063 + }, + { + "epoch": 0.3296823080574553, + "grad_norm": 2.72216796875, + "learning_rate": 0.00015103197043738002, + "loss": 0.2594, + "step": 14064 + }, + { + "epoch": 0.3297057496322603, + "grad_norm": 3.608149528503418, + "learning_rate": 0.00015102563551540945, + "loss": 0.4582, + "step": 14065 + }, + { + "epoch": 0.3297291912070653, + "grad_norm": 5.535455226898193, + "learning_rate": 0.00015101930031657235, + "loss": 0.4871, + "step": 14066 + }, + { + "epoch": 0.3297526327818703, + "grad_norm": 4.281422138214111, + "learning_rate": 0.00015101296484090307, + "loss": 0.4986, + "step": 14067 + }, + { + "epoch": 0.3297760743566753, + "grad_norm": 5.728602409362793, + "learning_rate": 0.000151006629088436, + "loss": 0.6762, + "step": 14068 + }, + { + "epoch": 0.32979951593148027, + "grad_norm": 4.598238945007324, + "learning_rate": 0.00015100029305920555, + "loss": 0.4501, + "step": 14069 + }, + { + "epoch": 0.32982295750628526, + "grad_norm": 4.902762413024902, + "learning_rate": 0.00015099395675324608, + "loss": 0.4881, + "step": 14070 + }, + { + "epoch": 0.32984639908109026, + "grad_norm": 4.81184196472168, + "learning_rate": 0.00015098762017059193, + "loss": 0.5551, + "step": 14071 + }, + { + "epoch": 0.32986984065589525, + "grad_norm": 6.668074131011963, + "learning_rate": 0.00015098128331127758, + "loss": 0.8078, + "step": 14072 + }, + { + "epoch": 0.32989328223070025, + "grad_norm": 0.8835123777389526, + "learning_rate": 0.0001509749461753373, + "loss": 0.0684, + "step": 14073 + }, + { + "epoch": 0.32991672380550524, + "grad_norm": 4.8264617919921875, + "learning_rate": 0.00015096860876280554, + "loss": 0.4818, + "step": 14074 + }, + { + "epoch": 0.32994016538031024, + "grad_norm": 0.7624520063400269, + "learning_rate": 0.00015096227107371668, + "loss": 0.0493, + "step": 14075 + }, + { + "epoch": 0.32996360695511523, + "grad_norm": 1.5600993633270264, + "learning_rate": 0.0001509559331081051, + "loss": 0.186, + "step": 14076 + }, + { + "epoch": 0.3299870485299202, + "grad_norm": 1.3832826614379883, + "learning_rate": 0.00015094959486600522, + "loss": 0.1376, + "step": 14077 + }, + { + "epoch": 0.3300104901047252, + "grad_norm": 4.2431206703186035, + "learning_rate": 0.00015094325634745136, + "loss": 0.2927, + "step": 14078 + }, + { + "epoch": 0.3300339316795302, + "grad_norm": 4.16940450668335, + "learning_rate": 0.00015093691755247801, + "loss": 0.4262, + "step": 14079 + }, + { + "epoch": 0.3300573732543352, + "grad_norm": 4.4169392585754395, + "learning_rate": 0.00015093057848111947, + "loss": 0.2668, + "step": 14080 + }, + { + "epoch": 0.3300808148291402, + "grad_norm": 4.182379722595215, + "learning_rate": 0.0001509242391334102, + "loss": 0.313, + "step": 14081 + }, + { + "epoch": 0.3301042564039452, + "grad_norm": 1.7023789882659912, + "learning_rate": 0.00015091789950938457, + "loss": 0.1506, + "step": 14082 + }, + { + "epoch": 0.3301276979787502, + "grad_norm": 4.654563903808594, + "learning_rate": 0.00015091155960907698, + "loss": 0.5727, + "step": 14083 + }, + { + "epoch": 0.3301511395535552, + "grad_norm": 2.4082870483398438, + "learning_rate": 0.00015090521943252188, + "loss": 0.2648, + "step": 14084 + }, + { + "epoch": 0.3301745811283602, + "grad_norm": 1.4925899505615234, + "learning_rate": 0.0001508988789797536, + "loss": 0.1899, + "step": 14085 + }, + { + "epoch": 0.3301980227031652, + "grad_norm": 6.461333274841309, + "learning_rate": 0.00015089253825080655, + "loss": 0.551, + "step": 14086 + }, + { + "epoch": 0.33022146427797017, + "grad_norm": 4.6517109870910645, + "learning_rate": 0.00015088619724571513, + "loss": 0.3177, + "step": 14087 + }, + { + "epoch": 0.3302449058527752, + "grad_norm": 5.4635210037231445, + "learning_rate": 0.00015087985596451384, + "loss": 0.7274, + "step": 14088 + }, + { + "epoch": 0.3302683474275802, + "grad_norm": 3.3273231983184814, + "learning_rate": 0.000150873514407237, + "loss": 0.528, + "step": 14089 + }, + { + "epoch": 0.3302917890023852, + "grad_norm": 1.3589733839035034, + "learning_rate": 0.00015086717257391903, + "loss": 0.1788, + "step": 14090 + }, + { + "epoch": 0.3303152305771902, + "grad_norm": 3.415440082550049, + "learning_rate": 0.0001508608304645944, + "loss": 0.3446, + "step": 14091 + }, + { + "epoch": 0.3303386721519952, + "grad_norm": 2.429986000061035, + "learning_rate": 0.00015085448807929737, + "loss": 0.3124, + "step": 14092 + }, + { + "epoch": 0.3303621137268002, + "grad_norm": 5.016232013702393, + "learning_rate": 0.00015084814541806252, + "loss": 0.4791, + "step": 14093 + }, + { + "epoch": 0.3303855553016052, + "grad_norm": 1.0066109895706177, + "learning_rate": 0.00015084180248092421, + "loss": 0.0774, + "step": 14094 + }, + { + "epoch": 0.3304089968764102, + "grad_norm": 2.8949620723724365, + "learning_rate": 0.00015083545926791687, + "loss": 0.2654, + "step": 14095 + }, + { + "epoch": 0.33043243845121517, + "grad_norm": 1.9469125270843506, + "learning_rate": 0.00015082911577907483, + "loss": 0.2074, + "step": 14096 + }, + { + "epoch": 0.33045588002602017, + "grad_norm": 3.6226706504821777, + "learning_rate": 0.00015082277201443263, + "loss": 0.5455, + "step": 14097 + }, + { + "epoch": 0.33047932160082516, + "grad_norm": 2.023082971572876, + "learning_rate": 0.00015081642797402464, + "loss": 1.0861, + "step": 14098 + }, + { + "epoch": 0.33050276317563015, + "grad_norm": 6.649123668670654, + "learning_rate": 0.00015081008365788528, + "loss": 0.6041, + "step": 14099 + }, + { + "epoch": 0.33052620475043515, + "grad_norm": 1.5425065755844116, + "learning_rate": 0.00015080373906604897, + "loss": 0.1213, + "step": 14100 + }, + { + "epoch": 0.33054964632524014, + "grad_norm": 4.4298996925354, + "learning_rate": 0.00015079739419855014, + "loss": 0.8479, + "step": 14101 + }, + { + "epoch": 0.33057308790004514, + "grad_norm": 1.5975607633590698, + "learning_rate": 0.00015079104905542322, + "loss": 0.0936, + "step": 14102 + }, + { + "epoch": 0.33059652947485013, + "grad_norm": 1.6996080875396729, + "learning_rate": 0.00015078470363670264, + "loss": 0.1591, + "step": 14103 + }, + { + "epoch": 0.3306199710496551, + "grad_norm": 3.073396682739258, + "learning_rate": 0.0001507783579424228, + "loss": 0.282, + "step": 14104 + }, + { + "epoch": 0.3306434126244601, + "grad_norm": 0.7405877709388733, + "learning_rate": 0.00015077201197261824, + "loss": 0.0514, + "step": 14105 + }, + { + "epoch": 0.3306668541992651, + "grad_norm": 4.754056453704834, + "learning_rate": 0.00015076566572732325, + "loss": 0.7287, + "step": 14106 + }, + { + "epoch": 0.3306902957740701, + "grad_norm": 2.8256590366363525, + "learning_rate": 0.00015075931920657235, + "loss": 0.2853, + "step": 14107 + }, + { + "epoch": 0.3307137373488751, + "grad_norm": 4.613710880279541, + "learning_rate": 0.0001507529724104, + "loss": 0.2315, + "step": 14108 + }, + { + "epoch": 0.3307371789236801, + "grad_norm": 1.0958815813064575, + "learning_rate": 0.00015074662533884052, + "loss": 0.0605, + "step": 14109 + }, + { + "epoch": 0.3307606204984851, + "grad_norm": 4.19525671005249, + "learning_rate": 0.00015074027799192845, + "loss": 0.4947, + "step": 14110 + }, + { + "epoch": 0.3307840620732901, + "grad_norm": 2.536675453186035, + "learning_rate": 0.00015073393036969823, + "loss": 0.3688, + "step": 14111 + }, + { + "epoch": 0.3308075036480951, + "grad_norm": 4.114478588104248, + "learning_rate": 0.00015072758247218425, + "loss": 0.4109, + "step": 14112 + }, + { + "epoch": 0.3308309452229001, + "grad_norm": 5.171223163604736, + "learning_rate": 0.000150721234299421, + "loss": 0.7597, + "step": 14113 + }, + { + "epoch": 0.33085438679770507, + "grad_norm": 3.4327971935272217, + "learning_rate": 0.00015071488585144288, + "loss": 0.3035, + "step": 14114 + }, + { + "epoch": 0.33087782837251006, + "grad_norm": 3.287358283996582, + "learning_rate": 0.00015070853712828436, + "loss": 0.423, + "step": 14115 + }, + { + "epoch": 0.33090126994731506, + "grad_norm": 2.607693910598755, + "learning_rate": 0.00015070218812997995, + "loss": 1.9315, + "step": 14116 + }, + { + "epoch": 0.33092471152212005, + "grad_norm": 2.7312049865722656, + "learning_rate": 0.00015069583885656397, + "loss": 0.3659, + "step": 14117 + }, + { + "epoch": 0.33094815309692505, + "grad_norm": 1.8177741765975952, + "learning_rate": 0.000150689489308071, + "loss": 0.9119, + "step": 14118 + }, + { + "epoch": 0.33097159467173004, + "grad_norm": 3.885756254196167, + "learning_rate": 0.0001506831394845354, + "loss": 0.4108, + "step": 14119 + }, + { + "epoch": 0.33099503624653503, + "grad_norm": 2.4570939540863037, + "learning_rate": 0.0001506767893859917, + "loss": 0.148, + "step": 14120 + }, + { + "epoch": 0.33101847782134003, + "grad_norm": 3.8193798065185547, + "learning_rate": 0.00015067043901247426, + "loss": 0.3002, + "step": 14121 + }, + { + "epoch": 0.331041919396145, + "grad_norm": 2.924504041671753, + "learning_rate": 0.0001506640883640176, + "loss": 0.2187, + "step": 14122 + }, + { + "epoch": 0.33106536097095, + "grad_norm": 3.7634997367858887, + "learning_rate": 0.0001506577374406562, + "loss": 0.333, + "step": 14123 + }, + { + "epoch": 0.331088802545755, + "grad_norm": 2.2275619506835938, + "learning_rate": 0.0001506513862424245, + "loss": 0.1715, + "step": 14124 + }, + { + "epoch": 0.33111224412056, + "grad_norm": 1.0680512189865112, + "learning_rate": 0.00015064503476935694, + "loss": 0.0938, + "step": 14125 + }, + { + "epoch": 0.331135685695365, + "grad_norm": 1.9895075559616089, + "learning_rate": 0.000150638683021488, + "loss": 0.1783, + "step": 14126 + }, + { + "epoch": 0.33115912727017, + "grad_norm": 1.188490867614746, + "learning_rate": 0.00015063233099885218, + "loss": 0.0916, + "step": 14127 + }, + { + "epoch": 0.331182568844975, + "grad_norm": 3.7493183612823486, + "learning_rate": 0.00015062597870148383, + "loss": 0.4253, + "step": 14128 + }, + { + "epoch": 0.33120601041978, + "grad_norm": 3.4701685905456543, + "learning_rate": 0.00015061962612941752, + "loss": 0.4155, + "step": 14129 + }, + { + "epoch": 0.331229451994585, + "grad_norm": 1.4048528671264648, + "learning_rate": 0.0001506132732826877, + "loss": 0.2481, + "step": 14130 + }, + { + "epoch": 0.33125289356939, + "grad_norm": 1.1502310037612915, + "learning_rate": 0.00015060692016132885, + "loss": 0.1786, + "step": 14131 + }, + { + "epoch": 0.33127633514419497, + "grad_norm": 2.0162646770477295, + "learning_rate": 0.00015060056676537544, + "loss": 0.1892, + "step": 14132 + }, + { + "epoch": 0.33129977671899996, + "grad_norm": 3.292924165725708, + "learning_rate": 0.00015059421309486192, + "loss": 0.3912, + "step": 14133 + }, + { + "epoch": 0.33132321829380496, + "grad_norm": 3.315258741378784, + "learning_rate": 0.00015058785914982275, + "loss": 0.2155, + "step": 14134 + }, + { + "epoch": 0.33134665986860995, + "grad_norm": 1.3819748163223267, + "learning_rate": 0.00015058150493029246, + "loss": 0.0668, + "step": 14135 + }, + { + "epoch": 0.33137010144341494, + "grad_norm": 3.565995454788208, + "learning_rate": 0.0001505751504363055, + "loss": 0.4778, + "step": 14136 + }, + { + "epoch": 0.33139354301821994, + "grad_norm": 3.828158140182495, + "learning_rate": 0.00015056879566789636, + "loss": 0.4404, + "step": 14137 + }, + { + "epoch": 0.33141698459302493, + "grad_norm": 8.773964881896973, + "learning_rate": 0.0001505624406250995, + "loss": 0.7584, + "step": 14138 + }, + { + "epoch": 0.33144042616783, + "grad_norm": 3.9478700160980225, + "learning_rate": 0.0001505560853079494, + "loss": 0.3824, + "step": 14139 + }, + { + "epoch": 0.331463867742635, + "grad_norm": 3.656473398208618, + "learning_rate": 0.00015054972971648058, + "loss": 0.3561, + "step": 14140 + }, + { + "epoch": 0.33148730931743997, + "grad_norm": 2.778373956680298, + "learning_rate": 0.0001505433738507275, + "loss": 0.1704, + "step": 14141 + }, + { + "epoch": 0.33151075089224497, + "grad_norm": 3.3889267444610596, + "learning_rate": 0.0001505370177107247, + "loss": 0.4625, + "step": 14142 + }, + { + "epoch": 0.33153419246704996, + "grad_norm": 4.175991058349609, + "learning_rate": 0.00015053066129650655, + "loss": 0.5133, + "step": 14143 + }, + { + "epoch": 0.33155763404185495, + "grad_norm": 6.056220054626465, + "learning_rate": 0.00015052430460810762, + "loss": 0.3358, + "step": 14144 + }, + { + "epoch": 0.33158107561665995, + "grad_norm": 1.1292585134506226, + "learning_rate": 0.00015051794764556244, + "loss": 0.1076, + "step": 14145 + }, + { + "epoch": 0.33160451719146494, + "grad_norm": 0.9117586612701416, + "learning_rate": 0.00015051159040890542, + "loss": 0.1075, + "step": 14146 + }, + { + "epoch": 0.33162795876626994, + "grad_norm": 1.3269487619400024, + "learning_rate": 0.00015050523289817107, + "loss": 0.6547, + "step": 14147 + }, + { + "epoch": 0.33165140034107493, + "grad_norm": 4.566107749938965, + "learning_rate": 0.00015049887511339394, + "loss": 0.2816, + "step": 14148 + }, + { + "epoch": 0.3316748419158799, + "grad_norm": 1.5531693696975708, + "learning_rate": 0.0001504925170546085, + "loss": 0.1559, + "step": 14149 + }, + { + "epoch": 0.3316982834906849, + "grad_norm": 3.4190006256103516, + "learning_rate": 0.00015048615872184922, + "loss": 0.558, + "step": 14150 + }, + { + "epoch": 0.3317217250654899, + "grad_norm": 3.2648167610168457, + "learning_rate": 0.00015047980011515061, + "loss": 0.4616, + "step": 14151 + }, + { + "epoch": 0.3317451666402949, + "grad_norm": 0.9363737106323242, + "learning_rate": 0.00015047344123454725, + "loss": 0.0788, + "step": 14152 + }, + { + "epoch": 0.3317686082150999, + "grad_norm": 4.546518802642822, + "learning_rate": 0.0001504670820800735, + "loss": 0.4157, + "step": 14153 + }, + { + "epoch": 0.3317920497899049, + "grad_norm": 3.5345213413238525, + "learning_rate": 0.000150460722651764, + "loss": 0.4862, + "step": 14154 + }, + { + "epoch": 0.3318154913647099, + "grad_norm": 2.077693462371826, + "learning_rate": 0.0001504543629496532, + "loss": 0.2762, + "step": 14155 + }, + { + "epoch": 0.3318389329395149, + "grad_norm": 9.997164726257324, + "learning_rate": 0.00015044800297377554, + "loss": 0.4563, + "step": 14156 + }, + { + "epoch": 0.3318623745143199, + "grad_norm": 2.2243032455444336, + "learning_rate": 0.00015044164272416566, + "loss": 0.1773, + "step": 14157 + }, + { + "epoch": 0.3318858160891249, + "grad_norm": 4.064181804656982, + "learning_rate": 0.00015043528220085798, + "loss": 0.4136, + "step": 14158 + }, + { + "epoch": 0.33190925766392987, + "grad_norm": 3.4805331230163574, + "learning_rate": 0.00015042892140388706, + "loss": 0.2781, + "step": 14159 + }, + { + "epoch": 0.33193269923873486, + "grad_norm": 2.6259732246398926, + "learning_rate": 0.00015042256033328738, + "loss": 0.2099, + "step": 14160 + }, + { + "epoch": 0.33195614081353986, + "grad_norm": 6.0331573486328125, + "learning_rate": 0.0001504161989890935, + "loss": 0.7815, + "step": 14161 + }, + { + "epoch": 0.33197958238834485, + "grad_norm": 1.1895335912704468, + "learning_rate": 0.0001504098373713399, + "loss": 0.2036, + "step": 14162 + }, + { + "epoch": 0.33200302396314985, + "grad_norm": 3.922527551651001, + "learning_rate": 0.00015040347548006108, + "loss": 0.3752, + "step": 14163 + }, + { + "epoch": 0.33202646553795484, + "grad_norm": 3.7589964866638184, + "learning_rate": 0.0001503971133152916, + "loss": 0.3492, + "step": 14164 + }, + { + "epoch": 0.33204990711275983, + "grad_norm": 2.4362642765045166, + "learning_rate": 0.00015039075087706593, + "loss": 0.2364, + "step": 14165 + }, + { + "epoch": 0.33207334868756483, + "grad_norm": 1.0074125528335571, + "learning_rate": 0.00015038438816541863, + "loss": 0.0801, + "step": 14166 + }, + { + "epoch": 0.3320967902623698, + "grad_norm": 2.878478527069092, + "learning_rate": 0.00015037802518038428, + "loss": 0.1756, + "step": 14167 + }, + { + "epoch": 0.3321202318371748, + "grad_norm": 1.0953102111816406, + "learning_rate": 0.00015037166192199733, + "loss": 0.1397, + "step": 14168 + }, + { + "epoch": 0.3321436734119798, + "grad_norm": 2.8307459354400635, + "learning_rate": 0.0001503652983902923, + "loss": 0.2876, + "step": 14169 + }, + { + "epoch": 0.3321671149867848, + "grad_norm": 3.526859998703003, + "learning_rate": 0.00015035893458530377, + "loss": 0.2814, + "step": 14170 + }, + { + "epoch": 0.3321905565615898, + "grad_norm": 0.8493608236312866, + "learning_rate": 0.00015035257050706623, + "loss": 0.1047, + "step": 14171 + }, + { + "epoch": 0.3322139981363948, + "grad_norm": 5.132297515869141, + "learning_rate": 0.00015034620615561424, + "loss": 0.6252, + "step": 14172 + }, + { + "epoch": 0.3322374397111998, + "grad_norm": 0.9648556709289551, + "learning_rate": 0.00015033984153098233, + "loss": 0.107, + "step": 14173 + }, + { + "epoch": 0.3322608812860048, + "grad_norm": 3.547049045562744, + "learning_rate": 0.00015033347663320502, + "loss": 0.3375, + "step": 14174 + }, + { + "epoch": 0.3322843228608098, + "grad_norm": 3.7364556789398193, + "learning_rate": 0.00015032711146231682, + "loss": 1.6964, + "step": 14175 + }, + { + "epoch": 0.33230776443561477, + "grad_norm": 1.4250547885894775, + "learning_rate": 0.0001503207460183523, + "loss": 0.1573, + "step": 14176 + }, + { + "epoch": 0.33233120601041977, + "grad_norm": 3.5971035957336426, + "learning_rate": 0.00015031438030134605, + "loss": 0.2228, + "step": 14177 + }, + { + "epoch": 0.33235464758522476, + "grad_norm": 3.1961934566497803, + "learning_rate": 0.0001503080143113325, + "loss": 0.3434, + "step": 14178 + }, + { + "epoch": 0.33237808916002975, + "grad_norm": 3.413160562515259, + "learning_rate": 0.00015030164804834625, + "loss": 0.2606, + "step": 14179 + }, + { + "epoch": 0.33240153073483475, + "grad_norm": 3.6385884284973145, + "learning_rate": 0.0001502952815124219, + "loss": 0.2524, + "step": 14180 + }, + { + "epoch": 0.33242497230963974, + "grad_norm": 5.1580491065979, + "learning_rate": 0.0001502889147035939, + "loss": 0.3875, + "step": 14181 + }, + { + "epoch": 0.33244841388444474, + "grad_norm": 6.556309223175049, + "learning_rate": 0.00015028254762189684, + "loss": 0.5037, + "step": 14182 + }, + { + "epoch": 0.33247185545924973, + "grad_norm": 1.8359390497207642, + "learning_rate": 0.00015027618026736524, + "loss": 0.1972, + "step": 14183 + }, + { + "epoch": 0.3324952970340547, + "grad_norm": 0.6315476298332214, + "learning_rate": 0.0001502698126400337, + "loss": 0.0761, + "step": 14184 + }, + { + "epoch": 0.3325187386088597, + "grad_norm": 2.873967409133911, + "learning_rate": 0.00015026344473993673, + "loss": 0.4035, + "step": 14185 + }, + { + "epoch": 0.3325421801836647, + "grad_norm": 3.6951727867126465, + "learning_rate": 0.0001502570765671089, + "loss": 0.4014, + "step": 14186 + }, + { + "epoch": 0.3325656217584697, + "grad_norm": 1.3172216415405273, + "learning_rate": 0.00015025070812158477, + "loss": 0.1611, + "step": 14187 + }, + { + "epoch": 0.3325890633332747, + "grad_norm": 5.3418869972229, + "learning_rate": 0.00015024433940339887, + "loss": 0.635, + "step": 14188 + }, + { + "epoch": 0.3326125049080797, + "grad_norm": 1.533146619796753, + "learning_rate": 0.00015023797041258577, + "loss": 0.8002, + "step": 14189 + }, + { + "epoch": 0.33263594648288475, + "grad_norm": 1.8816168308258057, + "learning_rate": 0.00015023160114918004, + "loss": 0.2623, + "step": 14190 + }, + { + "epoch": 0.33265938805768974, + "grad_norm": 1.8232065439224243, + "learning_rate": 0.00015022523161321622, + "loss": 1.595, + "step": 14191 + }, + { + "epoch": 0.33268282963249474, + "grad_norm": 3.220742702484131, + "learning_rate": 0.00015021886180472888, + "loss": 0.325, + "step": 14192 + }, + { + "epoch": 0.33270627120729973, + "grad_norm": 1.6275125741958618, + "learning_rate": 0.00015021249172375264, + "loss": 1.0468, + "step": 14193 + }, + { + "epoch": 0.3327297127821047, + "grad_norm": 3.625065803527832, + "learning_rate": 0.00015020612137032192, + "loss": 0.2142, + "step": 14194 + }, + { + "epoch": 0.3327531543569097, + "grad_norm": 4.547733783721924, + "learning_rate": 0.00015019975074447142, + "loss": 0.3941, + "step": 14195 + }, + { + "epoch": 0.3327765959317147, + "grad_norm": 1.397308588027954, + "learning_rate": 0.00015019337984623565, + "loss": 0.2203, + "step": 14196 + }, + { + "epoch": 0.3328000375065197, + "grad_norm": 1.6314398050308228, + "learning_rate": 0.0001501870086756492, + "loss": 0.2637, + "step": 14197 + }, + { + "epoch": 0.3328234790813247, + "grad_norm": 2.7518327236175537, + "learning_rate": 0.00015018063723274663, + "loss": 0.2366, + "step": 14198 + }, + { + "epoch": 0.3328469206561297, + "grad_norm": 3.476249933242798, + "learning_rate": 0.00015017426551756252, + "loss": 0.4851, + "step": 14199 + }, + { + "epoch": 0.3328703622309347, + "grad_norm": 5.0155720710754395, + "learning_rate": 0.00015016789353013144, + "loss": 0.566, + "step": 14200 + }, + { + "epoch": 0.3328938038057397, + "grad_norm": 2.858131170272827, + "learning_rate": 0.0001501615212704879, + "loss": 0.1819, + "step": 14201 + }, + { + "epoch": 0.3329172453805447, + "grad_norm": 1.7045599222183228, + "learning_rate": 0.00015015514873866658, + "loss": 0.2814, + "step": 14202 + }, + { + "epoch": 0.3329406869553497, + "grad_norm": 4.246191024780273, + "learning_rate": 0.000150148775934702, + "loss": 0.6004, + "step": 14203 + }, + { + "epoch": 0.33296412853015467, + "grad_norm": 1.0177154541015625, + "learning_rate": 0.00015014240285862878, + "loss": 0.0875, + "step": 14204 + }, + { + "epoch": 0.33298757010495966, + "grad_norm": 3.141651153564453, + "learning_rate": 0.00015013602951048145, + "loss": 0.2265, + "step": 14205 + }, + { + "epoch": 0.33301101167976466, + "grad_norm": 1.4564945697784424, + "learning_rate": 0.00015012965589029462, + "loss": 0.1608, + "step": 14206 + }, + { + "epoch": 0.33303445325456965, + "grad_norm": 2.5441648960113525, + "learning_rate": 0.00015012328199810285, + "loss": 0.2136, + "step": 14207 + }, + { + "epoch": 0.33305789482937465, + "grad_norm": 2.4502480030059814, + "learning_rate": 0.00015011690783394075, + "loss": 0.3096, + "step": 14208 + }, + { + "epoch": 0.33308133640417964, + "grad_norm": 2.570791006088257, + "learning_rate": 0.0001501105333978429, + "loss": 0.2596, + "step": 14209 + }, + { + "epoch": 0.33310477797898463, + "grad_norm": 2.8828067779541016, + "learning_rate": 0.0001501041586898439, + "loss": 0.2395, + "step": 14210 + }, + { + "epoch": 0.33312821955378963, + "grad_norm": 3.4537177085876465, + "learning_rate": 0.0001500977837099783, + "loss": 0.2493, + "step": 14211 + }, + { + "epoch": 0.3331516611285946, + "grad_norm": 4.413823127746582, + "learning_rate": 0.00015009140845828074, + "loss": 0.3835, + "step": 14212 + }, + { + "epoch": 0.3331751027033996, + "grad_norm": 2.7783076763153076, + "learning_rate": 0.00015008503293478578, + "loss": 0.3695, + "step": 14213 + }, + { + "epoch": 0.3331985442782046, + "grad_norm": 3.7550430297851562, + "learning_rate": 0.000150078657139528, + "loss": 0.3364, + "step": 14214 + }, + { + "epoch": 0.3332219858530096, + "grad_norm": 2.5505897998809814, + "learning_rate": 0.00015007228107254207, + "loss": 0.2211, + "step": 14215 + }, + { + "epoch": 0.3332454274278146, + "grad_norm": 2.747236490249634, + "learning_rate": 0.00015006590473386252, + "loss": 0.2683, + "step": 14216 + }, + { + "epoch": 0.3332688690026196, + "grad_norm": 1.974272608757019, + "learning_rate": 0.00015005952812352394, + "loss": 0.2395, + "step": 14217 + }, + { + "epoch": 0.3332923105774246, + "grad_norm": 2.5093600749969482, + "learning_rate": 0.00015005315124156098, + "loss": 0.3587, + "step": 14218 + }, + { + "epoch": 0.3333157521522296, + "grad_norm": 3.031402349472046, + "learning_rate": 0.00015004677408800822, + "loss": 0.3646, + "step": 14219 + }, + { + "epoch": 0.3333391937270346, + "grad_norm": 4.588280200958252, + "learning_rate": 0.00015004039666290022, + "loss": 0.4658, + "step": 14220 + }, + { + "epoch": 0.33336263530183957, + "grad_norm": 1.535696268081665, + "learning_rate": 0.00015003401896627162, + "loss": 0.235, + "step": 14221 + }, + { + "epoch": 0.33338607687664457, + "grad_norm": 2.44588041305542, + "learning_rate": 0.00015002764099815704, + "loss": 0.1608, + "step": 14222 + }, + { + "epoch": 0.33340951845144956, + "grad_norm": 0.6230125427246094, + "learning_rate": 0.00015002126275859108, + "loss": 0.0534, + "step": 14223 + }, + { + "epoch": 0.33343296002625455, + "grad_norm": 2.656200885772705, + "learning_rate": 0.00015001488424760834, + "loss": 0.2618, + "step": 14224 + }, + { + "epoch": 0.33345640160105955, + "grad_norm": 3.2791588306427, + "learning_rate": 0.00015000850546524345, + "loss": 0.2997, + "step": 14225 + }, + { + "epoch": 0.33347984317586454, + "grad_norm": 2.9255013465881348, + "learning_rate": 0.00015000212641153097, + "loss": 0.3812, + "step": 14226 + }, + { + "epoch": 0.33350328475066954, + "grad_norm": 3.892634153366089, + "learning_rate": 0.00014999574708650558, + "loss": 0.4989, + "step": 14227 + }, + { + "epoch": 0.33352672632547453, + "grad_norm": 4.07551908493042, + "learning_rate": 0.00014998936749020182, + "loss": 0.6509, + "step": 14228 + }, + { + "epoch": 0.3335501679002795, + "grad_norm": 4.044968128204346, + "learning_rate": 0.00014998298762265437, + "loss": 0.4107, + "step": 14229 + }, + { + "epoch": 0.3335736094750845, + "grad_norm": 2.899324893951416, + "learning_rate": 0.00014997660748389784, + "loss": 0.1229, + "step": 14230 + }, + { + "epoch": 0.3335970510498895, + "grad_norm": 3.2983808517456055, + "learning_rate": 0.0001499702270739668, + "loss": 0.505, + "step": 14231 + }, + { + "epoch": 0.3336204926246945, + "grad_norm": 5.778072834014893, + "learning_rate": 0.0001499638463928959, + "loss": 0.4171, + "step": 14232 + }, + { + "epoch": 0.3336439341994995, + "grad_norm": 4.820833683013916, + "learning_rate": 0.00014995746544071978, + "loss": 0.8893, + "step": 14233 + }, + { + "epoch": 0.3336673757743045, + "grad_norm": 3.699469566345215, + "learning_rate": 0.00014995108421747306, + "loss": 0.5734, + "step": 14234 + }, + { + "epoch": 0.3336908173491095, + "grad_norm": 2.7576935291290283, + "learning_rate": 0.00014994470272319035, + "loss": 0.4008, + "step": 14235 + }, + { + "epoch": 0.3337142589239145, + "grad_norm": 2.4566662311553955, + "learning_rate": 0.00014993832095790626, + "loss": 0.2587, + "step": 14236 + }, + { + "epoch": 0.3337377004987195, + "grad_norm": 0.76109778881073, + "learning_rate": 0.00014993193892165544, + "loss": 0.0821, + "step": 14237 + }, + { + "epoch": 0.3337611420735245, + "grad_norm": 5.279946804046631, + "learning_rate": 0.00014992555661447253, + "loss": 0.7569, + "step": 14238 + }, + { + "epoch": 0.33378458364832947, + "grad_norm": 4.983426094055176, + "learning_rate": 0.00014991917403639214, + "loss": 0.4287, + "step": 14239 + }, + { + "epoch": 0.33380802522313446, + "grad_norm": 0.5134122967720032, + "learning_rate": 0.0001499127911874489, + "loss": 0.0503, + "step": 14240 + }, + { + "epoch": 0.33383146679793946, + "grad_norm": 4.468264102935791, + "learning_rate": 0.00014990640806767747, + "loss": 0.2987, + "step": 14241 + }, + { + "epoch": 0.3338549083727445, + "grad_norm": 5.26512336730957, + "learning_rate": 0.00014990002467711244, + "loss": 0.5715, + "step": 14242 + }, + { + "epoch": 0.3338783499475495, + "grad_norm": 2.4562597274780273, + "learning_rate": 0.0001498936410157885, + "loss": 0.3142, + "step": 14243 + }, + { + "epoch": 0.3339017915223545, + "grad_norm": 9.245198249816895, + "learning_rate": 0.00014988725708374022, + "loss": 0.4089, + "step": 14244 + }, + { + "epoch": 0.3339252330971595, + "grad_norm": 2.821434497833252, + "learning_rate": 0.0001498808728810023, + "loss": 0.2869, + "step": 14245 + }, + { + "epoch": 0.3339486746719645, + "grad_norm": 3.161644458770752, + "learning_rate": 0.00014987448840760937, + "loss": 0.3471, + "step": 14246 + }, + { + "epoch": 0.3339721162467695, + "grad_norm": 3.16864275932312, + "learning_rate": 0.0001498681036635961, + "loss": 0.3104, + "step": 14247 + }, + { + "epoch": 0.3339955578215745, + "grad_norm": 1.2588319778442383, + "learning_rate": 0.00014986171864899702, + "loss": 0.1521, + "step": 14248 + }, + { + "epoch": 0.33401899939637947, + "grad_norm": 6.276431560516357, + "learning_rate": 0.00014985533336384692, + "loss": 0.6732, + "step": 14249 + }, + { + "epoch": 0.33404244097118446, + "grad_norm": 2.766186237335205, + "learning_rate": 0.00014984894780818034, + "loss": 0.2519, + "step": 14250 + }, + { + "epoch": 0.33406588254598946, + "grad_norm": 0.8252761363983154, + "learning_rate": 0.000149842561982032, + "loss": 0.1113, + "step": 14251 + }, + { + "epoch": 0.33408932412079445, + "grad_norm": 4.083439350128174, + "learning_rate": 0.00014983617588543648, + "loss": 0.4762, + "step": 14252 + }, + { + "epoch": 0.33411276569559945, + "grad_norm": 3.3316433429718018, + "learning_rate": 0.0001498297895184285, + "loss": 0.4319, + "step": 14253 + }, + { + "epoch": 0.33413620727040444, + "grad_norm": 3.749788522720337, + "learning_rate": 0.00014982340288104267, + "loss": 0.4026, + "step": 14254 + }, + { + "epoch": 0.33415964884520943, + "grad_norm": 2.1981539726257324, + "learning_rate": 0.00014981701597331363, + "loss": 0.2345, + "step": 14255 + }, + { + "epoch": 0.33418309042001443, + "grad_norm": 4.455482482910156, + "learning_rate": 0.00014981062879527608, + "loss": 0.2798, + "step": 14256 + }, + { + "epoch": 0.3342065319948194, + "grad_norm": 0.9222329258918762, + "learning_rate": 0.00014980424134696468, + "loss": 0.0689, + "step": 14257 + }, + { + "epoch": 0.3342299735696244, + "grad_norm": 0.968916118144989, + "learning_rate": 0.000149797853628414, + "loss": 0.1106, + "step": 14258 + }, + { + "epoch": 0.3342534151444294, + "grad_norm": 5.220627307891846, + "learning_rate": 0.00014979146563965883, + "loss": 0.6793, + "step": 14259 + }, + { + "epoch": 0.3342768567192344, + "grad_norm": 0.757452666759491, + "learning_rate": 0.00014978507738073378, + "loss": 0.0825, + "step": 14260 + }, + { + "epoch": 0.3343002982940394, + "grad_norm": 3.926219940185547, + "learning_rate": 0.00014977868885167344, + "loss": 0.3194, + "step": 14261 + }, + { + "epoch": 0.3343237398688444, + "grad_norm": 4.707557201385498, + "learning_rate": 0.00014977230005251257, + "loss": 0.4806, + "step": 14262 + }, + { + "epoch": 0.3343471814436494, + "grad_norm": 4.397871494293213, + "learning_rate": 0.0001497659109832858, + "loss": 0.471, + "step": 14263 + }, + { + "epoch": 0.3343706230184544, + "grad_norm": 3.091919422149658, + "learning_rate": 0.0001497595216440278, + "loss": 0.5411, + "step": 14264 + }, + { + "epoch": 0.3343940645932594, + "grad_norm": 1.5614248514175415, + "learning_rate": 0.00014975313203477323, + "loss": 0.0838, + "step": 14265 + }, + { + "epoch": 0.33441750616806437, + "grad_norm": 1.624885082244873, + "learning_rate": 0.00014974674215555676, + "loss": 1.1936, + "step": 14266 + }, + { + "epoch": 0.33444094774286937, + "grad_norm": 3.5779170989990234, + "learning_rate": 0.00014974035200641307, + "loss": 0.4805, + "step": 14267 + }, + { + "epoch": 0.33446438931767436, + "grad_norm": 2.7704761028289795, + "learning_rate": 0.00014973396158737682, + "loss": 1.478, + "step": 14268 + }, + { + "epoch": 0.33448783089247935, + "grad_norm": 4.398416996002197, + "learning_rate": 0.00014972757089848273, + "loss": 0.5086, + "step": 14269 + }, + { + "epoch": 0.33451127246728435, + "grad_norm": 1.5470694303512573, + "learning_rate": 0.0001497211799397654, + "loss": 0.2096, + "step": 14270 + }, + { + "epoch": 0.33453471404208934, + "grad_norm": 4.346372604370117, + "learning_rate": 0.0001497147887112596, + "loss": 0.3648, + "step": 14271 + }, + { + "epoch": 0.33455815561689434, + "grad_norm": 0.8993420004844666, + "learning_rate": 0.00014970839721299993, + "loss": 0.1573, + "step": 14272 + }, + { + "epoch": 0.33458159719169933, + "grad_norm": 3.0269267559051514, + "learning_rate": 0.0001497020054450211, + "loss": 0.3262, + "step": 14273 + }, + { + "epoch": 0.3346050387665043, + "grad_norm": 3.2599315643310547, + "learning_rate": 0.00014969561340735777, + "loss": 0.4251, + "step": 14274 + }, + { + "epoch": 0.3346284803413093, + "grad_norm": 0.3159758746623993, + "learning_rate": 0.00014968922110004468, + "loss": 0.0264, + "step": 14275 + }, + { + "epoch": 0.3346519219161143, + "grad_norm": 3.915679931640625, + "learning_rate": 0.00014968282852311645, + "loss": 0.4768, + "step": 14276 + }, + { + "epoch": 0.3346753634909193, + "grad_norm": 5.228361129760742, + "learning_rate": 0.00014967643567660784, + "loss": 0.5468, + "step": 14277 + }, + { + "epoch": 0.3346988050657243, + "grad_norm": 2.5340824127197266, + "learning_rate": 0.00014967004256055347, + "loss": 0.2249, + "step": 14278 + }, + { + "epoch": 0.3347222466405293, + "grad_norm": 1.3126246929168701, + "learning_rate": 0.00014966364917498804, + "loss": 0.1062, + "step": 14279 + }, + { + "epoch": 0.3347456882153343, + "grad_norm": 4.87766695022583, + "learning_rate": 0.00014965725551994625, + "loss": 0.4272, + "step": 14280 + }, + { + "epoch": 0.3347691297901393, + "grad_norm": 1.2328059673309326, + "learning_rate": 0.0001496508615954628, + "loss": 0.0927, + "step": 14281 + }, + { + "epoch": 0.3347925713649443, + "grad_norm": 4.2027130126953125, + "learning_rate": 0.00014964446740157238, + "loss": 0.6716, + "step": 14282 + }, + { + "epoch": 0.3348160129397493, + "grad_norm": 1.2353242635726929, + "learning_rate": 0.0001496380729383097, + "loss": 0.0997, + "step": 14283 + }, + { + "epoch": 0.33483945451455427, + "grad_norm": 0.48796841502189636, + "learning_rate": 0.0001496316782057094, + "loss": 0.0544, + "step": 14284 + }, + { + "epoch": 0.33486289608935926, + "grad_norm": 2.345560312271118, + "learning_rate": 0.00014962528320380626, + "loss": 0.2154, + "step": 14285 + }, + { + "epoch": 0.33488633766416426, + "grad_norm": 5.521158218383789, + "learning_rate": 0.00014961888793263493, + "loss": 0.5749, + "step": 14286 + }, + { + "epoch": 0.33490977923896925, + "grad_norm": 5.455990791320801, + "learning_rate": 0.0001496124923922301, + "loss": 0.6889, + "step": 14287 + }, + { + "epoch": 0.33493322081377425, + "grad_norm": 2.8745603561401367, + "learning_rate": 0.00014960609658262652, + "loss": 0.2799, + "step": 14288 + }, + { + "epoch": 0.33495666238857924, + "grad_norm": 6.259800434112549, + "learning_rate": 0.00014959970050385885, + "loss": 0.4169, + "step": 14289 + }, + { + "epoch": 0.33498010396338423, + "grad_norm": 1.4146186113357544, + "learning_rate": 0.0001495933041559618, + "loss": 0.1167, + "step": 14290 + }, + { + "epoch": 0.33500354553818923, + "grad_norm": 3.4141721725463867, + "learning_rate": 0.0001495869075389701, + "loss": 0.3358, + "step": 14291 + }, + { + "epoch": 0.3350269871129942, + "grad_norm": 3.390976667404175, + "learning_rate": 0.00014958051065291846, + "loss": 0.3973, + "step": 14292 + }, + { + "epoch": 0.3350504286877993, + "grad_norm": 2.724114418029785, + "learning_rate": 0.00014957411349784155, + "loss": 0.3546, + "step": 14293 + }, + { + "epoch": 0.33507387026260427, + "grad_norm": 3.4071340560913086, + "learning_rate": 0.00014956771607377411, + "loss": 0.3094, + "step": 14294 + }, + { + "epoch": 0.33509731183740926, + "grad_norm": 3.3575921058654785, + "learning_rate": 0.00014956131838075084, + "loss": 0.3717, + "step": 14295 + }, + { + "epoch": 0.33512075341221426, + "grad_norm": 2.3513705730438232, + "learning_rate": 0.00014955492041880648, + "loss": 0.293, + "step": 14296 + }, + { + "epoch": 0.33514419498701925, + "grad_norm": 2.5971128940582275, + "learning_rate": 0.00014954852218797572, + "loss": 0.2633, + "step": 14297 + }, + { + "epoch": 0.33516763656182424, + "grad_norm": 0.4842431843280792, + "learning_rate": 0.00014954212368829332, + "loss": 0.0393, + "step": 14298 + }, + { + "epoch": 0.33519107813662924, + "grad_norm": 3.2148778438568115, + "learning_rate": 0.0001495357249197939, + "loss": 0.3509, + "step": 14299 + }, + { + "epoch": 0.33521451971143423, + "grad_norm": 5.901909351348877, + "learning_rate": 0.0001495293258825123, + "loss": 0.7044, + "step": 14300 + }, + { + "epoch": 0.3352379612862392, + "grad_norm": 1.1862292289733887, + "learning_rate": 0.00014952292657648315, + "loss": 0.142, + "step": 14301 + }, + { + "epoch": 0.3352614028610442, + "grad_norm": 0.49580052495002747, + "learning_rate": 0.00014951652700174125, + "loss": 0.0621, + "step": 14302 + }, + { + "epoch": 0.3352848444358492, + "grad_norm": 2.190225601196289, + "learning_rate": 0.00014951012715832122, + "loss": 0.1893, + "step": 14303 + }, + { + "epoch": 0.3353082860106542, + "grad_norm": 1.1402961015701294, + "learning_rate": 0.0001495037270462579, + "loss": 0.085, + "step": 14304 + }, + { + "epoch": 0.3353317275854592, + "grad_norm": 2.7721288204193115, + "learning_rate": 0.00014949732666558595, + "loss": 0.3973, + "step": 14305 + }, + { + "epoch": 0.3353551691602642, + "grad_norm": 4.250739574432373, + "learning_rate": 0.00014949092601634009, + "loss": 0.7675, + "step": 14306 + }, + { + "epoch": 0.3353786107350692, + "grad_norm": 3.141528844833374, + "learning_rate": 0.00014948452509855507, + "loss": 0.3881, + "step": 14307 + }, + { + "epoch": 0.3354020523098742, + "grad_norm": 1.5639641284942627, + "learning_rate": 0.00014947812391226568, + "loss": 0.3147, + "step": 14308 + }, + { + "epoch": 0.3354254938846792, + "grad_norm": 6.997717380523682, + "learning_rate": 0.00014947172245750656, + "loss": 0.5739, + "step": 14309 + }, + { + "epoch": 0.3354489354594842, + "grad_norm": 0.9163528084754944, + "learning_rate": 0.0001494653207343125, + "loss": 0.1027, + "step": 14310 + }, + { + "epoch": 0.33547237703428917, + "grad_norm": 1.9098269939422607, + "learning_rate": 0.0001494589187427182, + "loss": 0.971, + "step": 14311 + }, + { + "epoch": 0.33549581860909417, + "grad_norm": 1.5387322902679443, + "learning_rate": 0.00014945251648275842, + "loss": 0.0615, + "step": 14312 + }, + { + "epoch": 0.33551926018389916, + "grad_norm": 3.2395341396331787, + "learning_rate": 0.0001494461139544679, + "loss": 0.3844, + "step": 14313 + }, + { + "epoch": 0.33554270175870415, + "grad_norm": 2.757110595703125, + "learning_rate": 0.0001494397111578814, + "loss": 0.4292, + "step": 14314 + }, + { + "epoch": 0.33556614333350915, + "grad_norm": 3.9176130294799805, + "learning_rate": 0.0001494333080930336, + "loss": 0.6116, + "step": 14315 + }, + { + "epoch": 0.33558958490831414, + "grad_norm": 4.3473992347717285, + "learning_rate": 0.00014942690475995928, + "loss": 0.5065, + "step": 14316 + }, + { + "epoch": 0.33561302648311914, + "grad_norm": 1.8011950254440308, + "learning_rate": 0.0001494205011586932, + "loss": 0.18, + "step": 14317 + }, + { + "epoch": 0.33563646805792413, + "grad_norm": 1.081383466720581, + "learning_rate": 0.0001494140972892701, + "loss": 0.0918, + "step": 14318 + }, + { + "epoch": 0.3356599096327291, + "grad_norm": 4.397697925567627, + "learning_rate": 0.00014940769315172469, + "loss": 0.4569, + "step": 14319 + }, + { + "epoch": 0.3356833512075341, + "grad_norm": 3.145789384841919, + "learning_rate": 0.00014940128874609178, + "loss": 0.381, + "step": 14320 + }, + { + "epoch": 0.3357067927823391, + "grad_norm": 3.938448667526245, + "learning_rate": 0.00014939488407240606, + "loss": 0.5348, + "step": 14321 + }, + { + "epoch": 0.3357302343571441, + "grad_norm": 2.204334020614624, + "learning_rate": 0.00014938847913070234, + "loss": 0.2368, + "step": 14322 + }, + { + "epoch": 0.3357536759319491, + "grad_norm": 3.0693888664245605, + "learning_rate": 0.00014938207392101532, + "loss": 0.3064, + "step": 14323 + }, + { + "epoch": 0.3357771175067541, + "grad_norm": 3.4175541400909424, + "learning_rate": 0.0001493756684433798, + "loss": 0.3526, + "step": 14324 + }, + { + "epoch": 0.3358005590815591, + "grad_norm": 2.6861350536346436, + "learning_rate": 0.0001493692626978305, + "loss": 0.2835, + "step": 14325 + }, + { + "epoch": 0.3358240006563641, + "grad_norm": 2.641085624694824, + "learning_rate": 0.00014936285668440218, + "loss": 0.2668, + "step": 14326 + }, + { + "epoch": 0.3358474422311691, + "grad_norm": 3.6858668327331543, + "learning_rate": 0.00014935645040312965, + "loss": 0.3676, + "step": 14327 + }, + { + "epoch": 0.3358708838059741, + "grad_norm": 3.6520354747772217, + "learning_rate": 0.0001493500438540476, + "loss": 1.6219, + "step": 14328 + }, + { + "epoch": 0.33589432538077907, + "grad_norm": 2.9409549236297607, + "learning_rate": 0.00014934363703719082, + "loss": 0.3327, + "step": 14329 + }, + { + "epoch": 0.33591776695558406, + "grad_norm": 3.1848483085632324, + "learning_rate": 0.00014933722995259409, + "loss": 0.3475, + "step": 14330 + }, + { + "epoch": 0.33594120853038906, + "grad_norm": 3.817826986312866, + "learning_rate": 0.00014933082260029217, + "loss": 2.06, + "step": 14331 + }, + { + "epoch": 0.33596465010519405, + "grad_norm": 3.759331464767456, + "learning_rate": 0.00014932441498031982, + "loss": 0.4143, + "step": 14332 + }, + { + "epoch": 0.33598809167999905, + "grad_norm": 1.1063542366027832, + "learning_rate": 0.0001493180070927118, + "loss": 0.0976, + "step": 14333 + }, + { + "epoch": 0.33601153325480404, + "grad_norm": 4.328507900238037, + "learning_rate": 0.00014931159893750287, + "loss": 0.2809, + "step": 14334 + }, + { + "epoch": 0.33603497482960903, + "grad_norm": 3.9544286727905273, + "learning_rate": 0.00014930519051472782, + "loss": 0.2588, + "step": 14335 + }, + { + "epoch": 0.33605841640441403, + "grad_norm": 4.579391956329346, + "learning_rate": 0.00014929878182442144, + "loss": 0.3697, + "step": 14336 + }, + { + "epoch": 0.336081857979219, + "grad_norm": 2.716094970703125, + "learning_rate": 0.0001492923728666185, + "loss": 0.2266, + "step": 14337 + }, + { + "epoch": 0.336105299554024, + "grad_norm": 2.745115041732788, + "learning_rate": 0.00014928596364135373, + "loss": 0.3326, + "step": 14338 + }, + { + "epoch": 0.336128741128829, + "grad_norm": 4.5205841064453125, + "learning_rate": 0.00014927955414866195, + "loss": 0.379, + "step": 14339 + }, + { + "epoch": 0.336152182703634, + "grad_norm": 4.669552803039551, + "learning_rate": 0.0001492731443885779, + "loss": 0.4906, + "step": 14340 + }, + { + "epoch": 0.336175624278439, + "grad_norm": 3.432642698287964, + "learning_rate": 0.00014926673436113638, + "loss": 0.4599, + "step": 14341 + }, + { + "epoch": 0.336199065853244, + "grad_norm": 3.8699145317077637, + "learning_rate": 0.0001492603240663722, + "loss": 1.7375, + "step": 14342 + }, + { + "epoch": 0.336222507428049, + "grad_norm": 0.7161343097686768, + "learning_rate": 0.0001492539135043201, + "loss": 0.1079, + "step": 14343 + }, + { + "epoch": 0.33624594900285404, + "grad_norm": 5.0007500648498535, + "learning_rate": 0.0001492475026750149, + "loss": 0.5655, + "step": 14344 + }, + { + "epoch": 0.33626939057765903, + "grad_norm": 2.3748302459716797, + "learning_rate": 0.0001492410915784914, + "loss": 1.7, + "step": 14345 + }, + { + "epoch": 0.336292832152464, + "grad_norm": 3.5014705657958984, + "learning_rate": 0.00014923468021478428, + "loss": 0.3587, + "step": 14346 + }, + { + "epoch": 0.336316273727269, + "grad_norm": 1.8439631462097168, + "learning_rate": 0.0001492282685839284, + "loss": 0.1129, + "step": 14347 + }, + { + "epoch": 0.336339715302074, + "grad_norm": 2.60585355758667, + "learning_rate": 0.0001492218566859586, + "loss": 0.2283, + "step": 14348 + }, + { + "epoch": 0.336363156876879, + "grad_norm": 1.099320888519287, + "learning_rate": 0.0001492154445209096, + "loss": 0.089, + "step": 14349 + }, + { + "epoch": 0.336386598451684, + "grad_norm": 4.2136383056640625, + "learning_rate": 0.0001492090320888162, + "loss": 0.2223, + "step": 14350 + }, + { + "epoch": 0.336410040026489, + "grad_norm": 5.036344051361084, + "learning_rate": 0.00014920261938971324, + "loss": 0.8173, + "step": 14351 + }, + { + "epoch": 0.336433481601294, + "grad_norm": 2.698662042617798, + "learning_rate": 0.00014919620642363547, + "loss": 1.9871, + "step": 14352 + }, + { + "epoch": 0.336456923176099, + "grad_norm": 4.709146976470947, + "learning_rate": 0.0001491897931906177, + "loss": 0.2086, + "step": 14353 + }, + { + "epoch": 0.336480364750904, + "grad_norm": 0.9796769618988037, + "learning_rate": 0.0001491833796906947, + "loss": 0.12, + "step": 14354 + }, + { + "epoch": 0.336503806325709, + "grad_norm": 1.2090915441513062, + "learning_rate": 0.00014917696592390135, + "loss": 0.151, + "step": 14355 + }, + { + "epoch": 0.33652724790051397, + "grad_norm": 2.7618296146392822, + "learning_rate": 0.0001491705518902724, + "loss": 0.3227, + "step": 14356 + }, + { + "epoch": 0.33655068947531896, + "grad_norm": 1.6251122951507568, + "learning_rate": 0.0001491641375898426, + "loss": 0.1886, + "step": 14357 + }, + { + "epoch": 0.33657413105012396, + "grad_norm": 4.826887607574463, + "learning_rate": 0.00014915772302264685, + "loss": 0.476, + "step": 14358 + }, + { + "epoch": 0.33659757262492895, + "grad_norm": 5.152050971984863, + "learning_rate": 0.0001491513081887199, + "loss": 0.5037, + "step": 14359 + }, + { + "epoch": 0.33662101419973395, + "grad_norm": 4.934535980224609, + "learning_rate": 0.00014914489308809654, + "loss": 0.3677, + "step": 14360 + }, + { + "epoch": 0.33664445577453894, + "grad_norm": 4.997623920440674, + "learning_rate": 0.00014913847772081166, + "loss": 0.4222, + "step": 14361 + }, + { + "epoch": 0.33666789734934394, + "grad_norm": 1.0436478853225708, + "learning_rate": 0.0001491320620869, + "loss": 0.085, + "step": 14362 + }, + { + "epoch": 0.33669133892414893, + "grad_norm": 4.227132797241211, + "learning_rate": 0.00014912564618639636, + "loss": 0.5078, + "step": 14363 + }, + { + "epoch": 0.3367147804989539, + "grad_norm": 4.0198774337768555, + "learning_rate": 0.00014911923001933563, + "loss": 0.2637, + "step": 14364 + }, + { + "epoch": 0.3367382220737589, + "grad_norm": 5.795548915863037, + "learning_rate": 0.00014911281358575255, + "loss": 0.96, + "step": 14365 + }, + { + "epoch": 0.3367616636485639, + "grad_norm": 1.7171539068222046, + "learning_rate": 0.00014910639688568196, + "loss": 0.0393, + "step": 14366 + }, + { + "epoch": 0.3367851052233689, + "grad_norm": 2.7820656299591064, + "learning_rate": 0.0001490999799191587, + "loss": 0.466, + "step": 14367 + }, + { + "epoch": 0.3368085467981739, + "grad_norm": 1.3511958122253418, + "learning_rate": 0.00014909356268621754, + "loss": 0.1342, + "step": 14368 + }, + { + "epoch": 0.3368319883729789, + "grad_norm": 0.32071301341056824, + "learning_rate": 0.00014908714518689336, + "loss": 0.0287, + "step": 14369 + }, + { + "epoch": 0.3368554299477839, + "grad_norm": 3.2582333087921143, + "learning_rate": 0.00014908072742122092, + "loss": 0.3006, + "step": 14370 + }, + { + "epoch": 0.3368788715225889, + "grad_norm": 0.2907772362232208, + "learning_rate": 0.00014907430938923511, + "loss": 0.026, + "step": 14371 + }, + { + "epoch": 0.3369023130973939, + "grad_norm": 1.4940916299819946, + "learning_rate": 0.00014906789109097067, + "loss": 0.1475, + "step": 14372 + }, + { + "epoch": 0.3369257546721989, + "grad_norm": 2.5177981853485107, + "learning_rate": 0.00014906147252646249, + "loss": 0.3112, + "step": 14373 + }, + { + "epoch": 0.33694919624700387, + "grad_norm": 0.5788042545318604, + "learning_rate": 0.0001490550536957454, + "loss": 0.042, + "step": 14374 + }, + { + "epoch": 0.33697263782180886, + "grad_norm": 2.779141902923584, + "learning_rate": 0.00014904863459885422, + "loss": 1.3082, + "step": 14375 + }, + { + "epoch": 0.33699607939661386, + "grad_norm": 0.9823120832443237, + "learning_rate": 0.00014904221523582374, + "loss": 0.0781, + "step": 14376 + }, + { + "epoch": 0.33701952097141885, + "grad_norm": 3.285111665725708, + "learning_rate": 0.00014903579560668885, + "loss": 0.3096, + "step": 14377 + }, + { + "epoch": 0.33704296254622385, + "grad_norm": 2.626936674118042, + "learning_rate": 0.0001490293757114843, + "loss": 0.2584, + "step": 14378 + }, + { + "epoch": 0.33706640412102884, + "grad_norm": 2.152686357498169, + "learning_rate": 0.00014902295555024498, + "loss": 0.2835, + "step": 14379 + }, + { + "epoch": 0.33708984569583383, + "grad_norm": 2.294487714767456, + "learning_rate": 0.00014901653512300577, + "loss": 0.1281, + "step": 14380 + }, + { + "epoch": 0.33711328727063883, + "grad_norm": 3.116238594055176, + "learning_rate": 0.00014901011442980147, + "loss": 0.2878, + "step": 14381 + }, + { + "epoch": 0.3371367288454438, + "grad_norm": 3.0163869857788086, + "learning_rate": 0.00014900369347066686, + "loss": 0.3396, + "step": 14382 + }, + { + "epoch": 0.3371601704202488, + "grad_norm": 3.607067823410034, + "learning_rate": 0.0001489972722456369, + "loss": 0.4686, + "step": 14383 + }, + { + "epoch": 0.3371836119950538, + "grad_norm": 1.0214734077453613, + "learning_rate": 0.0001489908507547463, + "loss": 0.062, + "step": 14384 + }, + { + "epoch": 0.3372070535698588, + "grad_norm": 3.2120840549468994, + "learning_rate": 0.00014898442899802998, + "loss": 0.2743, + "step": 14385 + }, + { + "epoch": 0.3372304951446638, + "grad_norm": 3.614995002746582, + "learning_rate": 0.00014897800697552278, + "loss": 0.4872, + "step": 14386 + }, + { + "epoch": 0.3372539367194688, + "grad_norm": 4.441131591796875, + "learning_rate": 0.00014897158468725957, + "loss": 0.645, + "step": 14387 + }, + { + "epoch": 0.3372773782942738, + "grad_norm": 2.4545207023620605, + "learning_rate": 0.0001489651621332751, + "loss": 0.1485, + "step": 14388 + }, + { + "epoch": 0.3373008198690788, + "grad_norm": 5.4420013427734375, + "learning_rate": 0.00014895873931360432, + "loss": 0.2875, + "step": 14389 + }, + { + "epoch": 0.3373242614438838, + "grad_norm": 1.612512469291687, + "learning_rate": 0.00014895231622828202, + "loss": 0.2394, + "step": 14390 + }, + { + "epoch": 0.33734770301868877, + "grad_norm": 1.6217548847198486, + "learning_rate": 0.0001489458928773431, + "loss": 0.1353, + "step": 14391 + }, + { + "epoch": 0.33737114459349377, + "grad_norm": 0.9253165125846863, + "learning_rate": 0.00014893946926082237, + "loss": 0.1061, + "step": 14392 + }, + { + "epoch": 0.33739458616829876, + "grad_norm": 3.6543407440185547, + "learning_rate": 0.00014893304537875476, + "loss": 0.3873, + "step": 14393 + }, + { + "epoch": 0.33741802774310375, + "grad_norm": 1.2258628606796265, + "learning_rate": 0.00014892662123117503, + "loss": 0.1327, + "step": 14394 + }, + { + "epoch": 0.3374414693179088, + "grad_norm": 1.0294910669326782, + "learning_rate": 0.00014892019681811807, + "loss": 0.0903, + "step": 14395 + }, + { + "epoch": 0.3374649108927138, + "grad_norm": 5.093254566192627, + "learning_rate": 0.0001489137721396187, + "loss": 0.4856, + "step": 14396 + }, + { + "epoch": 0.3374883524675188, + "grad_norm": 2.97981858253479, + "learning_rate": 0.00014890734719571192, + "loss": 0.3161, + "step": 14397 + }, + { + "epoch": 0.3375117940423238, + "grad_norm": 4.384069919586182, + "learning_rate": 0.00014890092198643245, + "loss": 0.3484, + "step": 14398 + }, + { + "epoch": 0.3375352356171288, + "grad_norm": 1.9916880130767822, + "learning_rate": 0.00014889449651181522, + "loss": 0.2384, + "step": 14399 + }, + { + "epoch": 0.3375586771919338, + "grad_norm": 1.6069300174713135, + "learning_rate": 0.00014888807077189506, + "loss": 0.1449, + "step": 14400 + }, + { + "epoch": 0.33758211876673877, + "grad_norm": 3.1609292030334473, + "learning_rate": 0.00014888164476670686, + "loss": 0.3051, + "step": 14401 + }, + { + "epoch": 0.33760556034154376, + "grad_norm": 3.537142038345337, + "learning_rate": 0.00014887521849628546, + "loss": 0.5472, + "step": 14402 + }, + { + "epoch": 0.33762900191634876, + "grad_norm": 3.4803056716918945, + "learning_rate": 0.0001488687919606658, + "loss": 0.3449, + "step": 14403 + }, + { + "epoch": 0.33765244349115375, + "grad_norm": 0.8655948042869568, + "learning_rate": 0.00014886236515988265, + "loss": 0.0986, + "step": 14404 + }, + { + "epoch": 0.33767588506595875, + "grad_norm": 3.260082483291626, + "learning_rate": 0.00014885593809397098, + "loss": 0.3076, + "step": 14405 + }, + { + "epoch": 0.33769932664076374, + "grad_norm": 2.8945000171661377, + "learning_rate": 0.00014884951076296558, + "loss": 0.3294, + "step": 14406 + }, + { + "epoch": 0.33772276821556874, + "grad_norm": 4.215839862823486, + "learning_rate": 0.00014884308316690138, + "loss": 0.4852, + "step": 14407 + }, + { + "epoch": 0.33774620979037373, + "grad_norm": 1.7277836799621582, + "learning_rate": 0.0001488366553058132, + "loss": 0.2131, + "step": 14408 + }, + { + "epoch": 0.3377696513651787, + "grad_norm": 2.109889507293701, + "learning_rate": 0.000148830227179736, + "loss": 0.2049, + "step": 14409 + }, + { + "epoch": 0.3377930929399837, + "grad_norm": 2.148805618286133, + "learning_rate": 0.00014882379878870464, + "loss": 1.2932, + "step": 14410 + }, + { + "epoch": 0.3378165345147887, + "grad_norm": 3.4922585487365723, + "learning_rate": 0.0001488173701327539, + "loss": 0.3197, + "step": 14411 + }, + { + "epoch": 0.3378399760895937, + "grad_norm": 3.705780029296875, + "learning_rate": 0.0001488109412119188, + "loss": 0.4206, + "step": 14412 + }, + { + "epoch": 0.3378634176643987, + "grad_norm": 3.8709726333618164, + "learning_rate": 0.00014880451202623412, + "loss": 0.465, + "step": 14413 + }, + { + "epoch": 0.3378868592392037, + "grad_norm": 2.5335910320281982, + "learning_rate": 0.00014879808257573478, + "loss": 0.3034, + "step": 14414 + }, + { + "epoch": 0.3379103008140087, + "grad_norm": 0.941680371761322, + "learning_rate": 0.0001487916528604557, + "loss": 0.1, + "step": 14415 + }, + { + "epoch": 0.3379337423888137, + "grad_norm": 5.995855331420898, + "learning_rate": 0.00014878522288043174, + "loss": 0.8139, + "step": 14416 + }, + { + "epoch": 0.3379571839636187, + "grad_norm": 1.1400965452194214, + "learning_rate": 0.00014877879263569776, + "loss": 0.1565, + "step": 14417 + }, + { + "epoch": 0.3379806255384237, + "grad_norm": 0.9284151196479797, + "learning_rate": 0.0001487723621262887, + "loss": 0.0824, + "step": 14418 + }, + { + "epoch": 0.33800406711322867, + "grad_norm": 1.8395134210586548, + "learning_rate": 0.00014876593135223945, + "loss": 0.2087, + "step": 14419 + }, + { + "epoch": 0.33802750868803366, + "grad_norm": 5.512516021728516, + "learning_rate": 0.00014875950031358485, + "loss": 0.5936, + "step": 14420 + }, + { + "epoch": 0.33805095026283866, + "grad_norm": 2.1285715103149414, + "learning_rate": 0.00014875306901035985, + "loss": 0.2366, + "step": 14421 + }, + { + "epoch": 0.33807439183764365, + "grad_norm": 0.6340988278388977, + "learning_rate": 0.0001487466374425993, + "loss": 0.0947, + "step": 14422 + }, + { + "epoch": 0.33809783341244865, + "grad_norm": 4.601064682006836, + "learning_rate": 0.00014874020561033815, + "loss": 0.1641, + "step": 14423 + }, + { + "epoch": 0.33812127498725364, + "grad_norm": 2.6335487365722656, + "learning_rate": 0.00014873377351361128, + "loss": 0.2536, + "step": 14424 + }, + { + "epoch": 0.33814471656205863, + "grad_norm": 4.797380447387695, + "learning_rate": 0.0001487273411524536, + "loss": 0.7563, + "step": 14425 + }, + { + "epoch": 0.33816815813686363, + "grad_norm": 4.073995113372803, + "learning_rate": 0.00014872090852689995, + "loss": 0.636, + "step": 14426 + }, + { + "epoch": 0.3381915997116686, + "grad_norm": 4.138184070587158, + "learning_rate": 0.0001487144756369853, + "loss": 0.2718, + "step": 14427 + }, + { + "epoch": 0.3382150412864736, + "grad_norm": 2.781806468963623, + "learning_rate": 0.00014870804248274454, + "loss": 0.3736, + "step": 14428 + }, + { + "epoch": 0.3382384828612786, + "grad_norm": 3.9540915489196777, + "learning_rate": 0.00014870160906421258, + "loss": 0.3914, + "step": 14429 + }, + { + "epoch": 0.3382619244360836, + "grad_norm": 5.061028957366943, + "learning_rate": 0.0001486951753814243, + "loss": 0.5325, + "step": 14430 + }, + { + "epoch": 0.3382853660108886, + "grad_norm": 4.579951286315918, + "learning_rate": 0.00014868874143441463, + "loss": 0.3508, + "step": 14431 + }, + { + "epoch": 0.3383088075856936, + "grad_norm": 1.4984115362167358, + "learning_rate": 0.0001486823072232185, + "loss": 0.2228, + "step": 14432 + }, + { + "epoch": 0.3383322491604986, + "grad_norm": 6.215976715087891, + "learning_rate": 0.00014867587274787077, + "loss": 0.3603, + "step": 14433 + }, + { + "epoch": 0.3383556907353036, + "grad_norm": 2.935373306274414, + "learning_rate": 0.00014866943800840644, + "loss": 0.202, + "step": 14434 + }, + { + "epoch": 0.3383791323101086, + "grad_norm": 4.052692413330078, + "learning_rate": 0.0001486630030048603, + "loss": 0.3607, + "step": 14435 + }, + { + "epoch": 0.33840257388491357, + "grad_norm": 2.7718682289123535, + "learning_rate": 0.00014865656773726738, + "loss": 1.6337, + "step": 14436 + }, + { + "epoch": 0.33842601545971857, + "grad_norm": 2.410752773284912, + "learning_rate": 0.00014865013220566255, + "loss": 0.2478, + "step": 14437 + }, + { + "epoch": 0.33844945703452356, + "grad_norm": 1.4491909742355347, + "learning_rate": 0.00014864369641008076, + "loss": 0.0959, + "step": 14438 + }, + { + "epoch": 0.33847289860932855, + "grad_norm": 2.1413941383361816, + "learning_rate": 0.00014863726035055686, + "loss": 0.3165, + "step": 14439 + }, + { + "epoch": 0.33849634018413355, + "grad_norm": 3.577056646347046, + "learning_rate": 0.00014863082402712582, + "loss": 1.2848, + "step": 14440 + }, + { + "epoch": 0.33851978175893854, + "grad_norm": 4.309433937072754, + "learning_rate": 0.0001486243874398226, + "loss": 0.4271, + "step": 14441 + }, + { + "epoch": 0.33854322333374354, + "grad_norm": 2.0785348415374756, + "learning_rate": 0.00014861795058868207, + "loss": 0.1882, + "step": 14442 + }, + { + "epoch": 0.33856666490854853, + "grad_norm": 1.922697901725769, + "learning_rate": 0.00014861151347373919, + "loss": 0.1582, + "step": 14443 + }, + { + "epoch": 0.3385901064833535, + "grad_norm": 6.464416027069092, + "learning_rate": 0.00014860507609502883, + "loss": 0.6685, + "step": 14444 + }, + { + "epoch": 0.3386135480581585, + "grad_norm": 2.6629419326782227, + "learning_rate": 0.00014859863845258597, + "loss": 0.2191, + "step": 14445 + }, + { + "epoch": 0.33863698963296357, + "grad_norm": 4.516000747680664, + "learning_rate": 0.00014859220054644554, + "loss": 0.3979, + "step": 14446 + }, + { + "epoch": 0.33866043120776856, + "grad_norm": 2.3363094329833984, + "learning_rate": 0.0001485857623766425, + "loss": 0.2773, + "step": 14447 + }, + { + "epoch": 0.33868387278257356, + "grad_norm": 1.5753942728042603, + "learning_rate": 0.0001485793239432117, + "loss": 0.5188, + "step": 14448 + }, + { + "epoch": 0.33870731435737855, + "grad_norm": 0.6564486026763916, + "learning_rate": 0.00014857288524618812, + "loss": 0.0478, + "step": 14449 + }, + { + "epoch": 0.33873075593218355, + "grad_norm": 2.9078195095062256, + "learning_rate": 0.0001485664462856067, + "loss": 0.2476, + "step": 14450 + }, + { + "epoch": 0.33875419750698854, + "grad_norm": 0.7199223637580872, + "learning_rate": 0.00014856000706150243, + "loss": 0.0677, + "step": 14451 + }, + { + "epoch": 0.33877763908179354, + "grad_norm": 4.819441795349121, + "learning_rate": 0.00014855356757391012, + "loss": 0.6794, + "step": 14452 + }, + { + "epoch": 0.33880108065659853, + "grad_norm": 3.6068456172943115, + "learning_rate": 0.00014854712782286483, + "loss": 0.3899, + "step": 14453 + }, + { + "epoch": 0.3388245222314035, + "grad_norm": 2.885561466217041, + "learning_rate": 0.00014854068780840147, + "loss": 0.1973, + "step": 14454 + }, + { + "epoch": 0.3388479638062085, + "grad_norm": 4.776673316955566, + "learning_rate": 0.00014853424753055494, + "loss": 0.4537, + "step": 14455 + }, + { + "epoch": 0.3388714053810135, + "grad_norm": 0.6289877891540527, + "learning_rate": 0.00014852780698936026, + "loss": 0.0624, + "step": 14456 + }, + { + "epoch": 0.3388948469558185, + "grad_norm": 3.7960808277130127, + "learning_rate": 0.0001485213661848523, + "loss": 0.3307, + "step": 14457 + }, + { + "epoch": 0.3389182885306235, + "grad_norm": 4.167929649353027, + "learning_rate": 0.00014851492511706605, + "loss": 0.3222, + "step": 14458 + }, + { + "epoch": 0.3389417301054285, + "grad_norm": 3.1045186519622803, + "learning_rate": 0.00014850848378603645, + "loss": 0.2174, + "step": 14459 + }, + { + "epoch": 0.3389651716802335, + "grad_norm": 2.3649799823760986, + "learning_rate": 0.00014850204219179846, + "loss": 0.1329, + "step": 14460 + }, + { + "epoch": 0.3389886132550385, + "grad_norm": 2.9426093101501465, + "learning_rate": 0.000148495600334387, + "loss": 0.263, + "step": 14461 + }, + { + "epoch": 0.3390120548298435, + "grad_norm": 1.2504891157150269, + "learning_rate": 0.0001484891582138371, + "loss": 0.1026, + "step": 14462 + }, + { + "epoch": 0.3390354964046485, + "grad_norm": 3.5005385875701904, + "learning_rate": 0.00014848271583018364, + "loss": 0.386, + "step": 14463 + }, + { + "epoch": 0.33905893797945347, + "grad_norm": 4.716573715209961, + "learning_rate": 0.0001484762731834616, + "loss": 0.69, + "step": 14464 + }, + { + "epoch": 0.33908237955425846, + "grad_norm": 4.818842887878418, + "learning_rate": 0.0001484698302737059, + "loss": 0.5639, + "step": 14465 + }, + { + "epoch": 0.33910582112906346, + "grad_norm": 1.2584505081176758, + "learning_rate": 0.0001484633871009516, + "loss": 0.1134, + "step": 14466 + }, + { + "epoch": 0.33912926270386845, + "grad_norm": 2.888063907623291, + "learning_rate": 0.00014845694366523356, + "loss": 0.2933, + "step": 14467 + }, + { + "epoch": 0.33915270427867344, + "grad_norm": 4.506526947021484, + "learning_rate": 0.00014845049996658682, + "loss": 0.324, + "step": 14468 + }, + { + "epoch": 0.33917614585347844, + "grad_norm": 6.2738423347473145, + "learning_rate": 0.00014844405600504624, + "loss": 0.5486, + "step": 14469 + }, + { + "epoch": 0.33919958742828343, + "grad_norm": 3.901578426361084, + "learning_rate": 0.00014843761178064694, + "loss": 0.2262, + "step": 14470 + }, + { + "epoch": 0.3392230290030884, + "grad_norm": 6.097157955169678, + "learning_rate": 0.00014843116729342372, + "loss": 0.5002, + "step": 14471 + }, + { + "epoch": 0.3392464705778934, + "grad_norm": 1.1600934267044067, + "learning_rate": 0.00014842472254341163, + "loss": 0.0931, + "step": 14472 + }, + { + "epoch": 0.3392699121526984, + "grad_norm": 2.879929780960083, + "learning_rate": 0.00014841827753064567, + "loss": 0.394, + "step": 14473 + }, + { + "epoch": 0.3392933537275034, + "grad_norm": 3.901787757873535, + "learning_rate": 0.00014841183225516074, + "loss": 0.4084, + "step": 14474 + }, + { + "epoch": 0.3393167953023084, + "grad_norm": 0.8789548873901367, + "learning_rate": 0.0001484053867169919, + "loss": 0.057, + "step": 14475 + }, + { + "epoch": 0.3393402368771134, + "grad_norm": 3.6157824993133545, + "learning_rate": 0.00014839894091617405, + "loss": 0.3134, + "step": 14476 + }, + { + "epoch": 0.3393636784519184, + "grad_norm": 0.8926867246627808, + "learning_rate": 0.00014839249485274213, + "loss": 0.0835, + "step": 14477 + }, + { + "epoch": 0.3393871200267234, + "grad_norm": 1.0287951231002808, + "learning_rate": 0.00014838604852673122, + "loss": 0.1296, + "step": 14478 + }, + { + "epoch": 0.3394105616015284, + "grad_norm": 2.1345813274383545, + "learning_rate": 0.00014837960193817625, + "loss": 0.1419, + "step": 14479 + }, + { + "epoch": 0.3394340031763334, + "grad_norm": 4.426779270172119, + "learning_rate": 0.0001483731550871122, + "loss": 0.3252, + "step": 14480 + }, + { + "epoch": 0.33945744475113837, + "grad_norm": 0.9162821769714355, + "learning_rate": 0.00014836670797357403, + "loss": 0.1852, + "step": 14481 + }, + { + "epoch": 0.33948088632594337, + "grad_norm": 2.974313974380493, + "learning_rate": 0.00014836026059759675, + "loss": 0.1888, + "step": 14482 + }, + { + "epoch": 0.33950432790074836, + "grad_norm": 1.0451983213424683, + "learning_rate": 0.00014835381295921538, + "loss": 0.1203, + "step": 14483 + }, + { + "epoch": 0.33952776947555335, + "grad_norm": 1.036730170249939, + "learning_rate": 0.0001483473650584648, + "loss": 0.1192, + "step": 14484 + }, + { + "epoch": 0.33955121105035835, + "grad_norm": 4.013701915740967, + "learning_rate": 0.0001483409168953801, + "loss": 0.3302, + "step": 14485 + }, + { + "epoch": 0.33957465262516334, + "grad_norm": 3.638592481613159, + "learning_rate": 0.00014833446846999621, + "loss": 0.2462, + "step": 14486 + }, + { + "epoch": 0.33959809419996834, + "grad_norm": 5.211601734161377, + "learning_rate": 0.0001483280197823481, + "loss": 0.5785, + "step": 14487 + }, + { + "epoch": 0.33962153577477333, + "grad_norm": 6.924477577209473, + "learning_rate": 0.00014832157083247084, + "loss": 0.6825, + "step": 14488 + }, + { + "epoch": 0.3396449773495783, + "grad_norm": 3.531913995742798, + "learning_rate": 0.00014831512162039937, + "loss": 0.3404, + "step": 14489 + }, + { + "epoch": 0.3396684189243833, + "grad_norm": 2.9961509704589844, + "learning_rate": 0.00014830867214616868, + "loss": 0.2497, + "step": 14490 + }, + { + "epoch": 0.3396918604991883, + "grad_norm": 0.7380627393722534, + "learning_rate": 0.0001483022224098138, + "loss": 0.0699, + "step": 14491 + }, + { + "epoch": 0.3397153020739933, + "grad_norm": 0.9456809759140015, + "learning_rate": 0.0001482957724113697, + "loss": 0.1011, + "step": 14492 + }, + { + "epoch": 0.3397387436487983, + "grad_norm": 1.5492395162582397, + "learning_rate": 0.00014828932215087134, + "loss": 0.1384, + "step": 14493 + }, + { + "epoch": 0.3397621852236033, + "grad_norm": 2.1028525829315186, + "learning_rate": 0.0001482828716283538, + "loss": 0.0649, + "step": 14494 + }, + { + "epoch": 0.3397856267984083, + "grad_norm": 4.101510524749756, + "learning_rate": 0.00014827642084385204, + "loss": 0.5846, + "step": 14495 + }, + { + "epoch": 0.3398090683732133, + "grad_norm": 5.0985918045043945, + "learning_rate": 0.00014826996979740105, + "loss": 0.5001, + "step": 14496 + }, + { + "epoch": 0.33983250994801834, + "grad_norm": 2.5554912090301514, + "learning_rate": 0.00014826351848903584, + "loss": 0.5634, + "step": 14497 + }, + { + "epoch": 0.33985595152282333, + "grad_norm": 2.1046929359436035, + "learning_rate": 0.00014825706691879144, + "loss": 0.195, + "step": 14498 + }, + { + "epoch": 0.3398793930976283, + "grad_norm": 7.602779388427734, + "learning_rate": 0.00014825061508670282, + "loss": 0.7619, + "step": 14499 + }, + { + "epoch": 0.3399028346724333, + "grad_norm": 1.5206178426742554, + "learning_rate": 0.000148244162992805, + "loss": 0.2217, + "step": 14500 + }, + { + "epoch": 0.3399262762472383, + "grad_norm": 3.474240779876709, + "learning_rate": 0.00014823771063713303, + "loss": 0.1789, + "step": 14501 + }, + { + "epoch": 0.3399497178220433, + "grad_norm": 4.9424943923950195, + "learning_rate": 0.00014823125801972185, + "loss": 0.6796, + "step": 14502 + }, + { + "epoch": 0.3399731593968483, + "grad_norm": 2.4979796409606934, + "learning_rate": 0.00014822480514060655, + "loss": 0.2133, + "step": 14503 + }, + { + "epoch": 0.3399966009716533, + "grad_norm": 2.3696043491363525, + "learning_rate": 0.00014821835199982207, + "loss": 0.2817, + "step": 14504 + }, + { + "epoch": 0.3400200425464583, + "grad_norm": 4.341365814208984, + "learning_rate": 0.00014821189859740344, + "loss": 0.5412, + "step": 14505 + }, + { + "epoch": 0.3400434841212633, + "grad_norm": 2.423555850982666, + "learning_rate": 0.0001482054449333857, + "loss": 0.1456, + "step": 14506 + }, + { + "epoch": 0.3400669256960683, + "grad_norm": 3.116821527481079, + "learning_rate": 0.0001481989910078039, + "loss": 0.2913, + "step": 14507 + }, + { + "epoch": 0.3400903672708733, + "grad_norm": 2.9745278358459473, + "learning_rate": 0.000148192536820693, + "loss": 1.4558, + "step": 14508 + }, + { + "epoch": 0.34011380884567827, + "grad_norm": 2.0648257732391357, + "learning_rate": 0.000148186082372088, + "loss": 0.3138, + "step": 14509 + }, + { + "epoch": 0.34013725042048326, + "grad_norm": 10.983488082885742, + "learning_rate": 0.00014817962766202402, + "loss": 0.657, + "step": 14510 + }, + { + "epoch": 0.34016069199528826, + "grad_norm": 2.9089601039886475, + "learning_rate": 0.00014817317269053603, + "loss": 0.2799, + "step": 14511 + }, + { + "epoch": 0.34018413357009325, + "grad_norm": 1.1310642957687378, + "learning_rate": 0.000148166717457659, + "loss": 0.1117, + "step": 14512 + }, + { + "epoch": 0.34020757514489824, + "grad_norm": 2.5333125591278076, + "learning_rate": 0.00014816026196342802, + "loss": 0.3098, + "step": 14513 + }, + { + "epoch": 0.34023101671970324, + "grad_norm": 7.393587112426758, + "learning_rate": 0.00014815380620787814, + "loss": 0.7107, + "step": 14514 + }, + { + "epoch": 0.34025445829450823, + "grad_norm": 3.584986448287964, + "learning_rate": 0.0001481473501910443, + "loss": 0.5141, + "step": 14515 + }, + { + "epoch": 0.3402778998693132, + "grad_norm": 2.296626567840576, + "learning_rate": 0.0001481408939129616, + "loss": 1.4609, + "step": 14516 + }, + { + "epoch": 0.3403013414441182, + "grad_norm": 4.837625503540039, + "learning_rate": 0.0001481344373736651, + "loss": 0.5268, + "step": 14517 + }, + { + "epoch": 0.3403247830189232, + "grad_norm": 3.1312289237976074, + "learning_rate": 0.00014812798057318974, + "loss": 0.4023, + "step": 14518 + }, + { + "epoch": 0.3403482245937282, + "grad_norm": 2.1938588619232178, + "learning_rate": 0.00014812152351157063, + "loss": 0.3333, + "step": 14519 + }, + { + "epoch": 0.3403716661685332, + "grad_norm": 2.664313793182373, + "learning_rate": 0.00014811506618884278, + "loss": 0.1887, + "step": 14520 + }, + { + "epoch": 0.3403951077433382, + "grad_norm": 0.9976897239685059, + "learning_rate": 0.0001481086086050412, + "loss": 0.0379, + "step": 14521 + }, + { + "epoch": 0.3404185493181432, + "grad_norm": 1.5202866792678833, + "learning_rate": 0.000148102150760201, + "loss": 0.1998, + "step": 14522 + }, + { + "epoch": 0.3404419908929482, + "grad_norm": 4.0875935554504395, + "learning_rate": 0.00014809569265435715, + "loss": 0.5124, + "step": 14523 + }, + { + "epoch": 0.3404654324677532, + "grad_norm": 2.192211151123047, + "learning_rate": 0.0001480892342875447, + "loss": 0.1663, + "step": 14524 + }, + { + "epoch": 0.3404888740425582, + "grad_norm": 4.873201847076416, + "learning_rate": 0.00014808277565979874, + "loss": 0.5658, + "step": 14525 + }, + { + "epoch": 0.34051231561736317, + "grad_norm": 4.867598056793213, + "learning_rate": 0.0001480763167711543, + "loss": 0.5067, + "step": 14526 + }, + { + "epoch": 0.34053575719216816, + "grad_norm": 3.599717378616333, + "learning_rate": 0.0001480698576216464, + "loss": 0.2305, + "step": 14527 + }, + { + "epoch": 0.34055919876697316, + "grad_norm": 0.4411122798919678, + "learning_rate": 0.00014806339821131007, + "loss": 0.0394, + "step": 14528 + }, + { + "epoch": 0.34058264034177815, + "grad_norm": 1.1042054891586304, + "learning_rate": 0.00014805693854018042, + "loss": 0.0752, + "step": 14529 + }, + { + "epoch": 0.34060608191658315, + "grad_norm": 3.9751574993133545, + "learning_rate": 0.00014805047860829248, + "loss": 0.5277, + "step": 14530 + }, + { + "epoch": 0.34062952349138814, + "grad_norm": 4.103702545166016, + "learning_rate": 0.0001480440184156813, + "loss": 0.4403, + "step": 14531 + }, + { + "epoch": 0.34065296506619314, + "grad_norm": 1.3084847927093506, + "learning_rate": 0.00014803755796238187, + "loss": 0.1768, + "step": 14532 + }, + { + "epoch": 0.34067640664099813, + "grad_norm": 2.234811544418335, + "learning_rate": 0.00014803109724842935, + "loss": 0.23, + "step": 14533 + }, + { + "epoch": 0.3406998482158031, + "grad_norm": 4.49425745010376, + "learning_rate": 0.0001480246362738587, + "loss": 0.499, + "step": 14534 + }, + { + "epoch": 0.3407232897906081, + "grad_norm": 0.9244576096534729, + "learning_rate": 0.00014801817503870508, + "loss": 0.0678, + "step": 14535 + }, + { + "epoch": 0.3407467313654131, + "grad_norm": 3.200711250305176, + "learning_rate": 0.00014801171354300347, + "loss": 0.3046, + "step": 14536 + }, + { + "epoch": 0.3407701729402181, + "grad_norm": 0.5752992630004883, + "learning_rate": 0.00014800525178678895, + "loss": 0.0527, + "step": 14537 + }, + { + "epoch": 0.3407936145150231, + "grad_norm": 1.31924307346344, + "learning_rate": 0.00014799878977009659, + "loss": 0.145, + "step": 14538 + }, + { + "epoch": 0.3408170560898281, + "grad_norm": 3.7323336601257324, + "learning_rate": 0.00014799232749296144, + "loss": 0.366, + "step": 14539 + }, + { + "epoch": 0.3408404976646331, + "grad_norm": 4.348148345947266, + "learning_rate": 0.00014798586495541857, + "loss": 0.5599, + "step": 14540 + }, + { + "epoch": 0.3408639392394381, + "grad_norm": 4.9180426597595215, + "learning_rate": 0.00014797940215750303, + "loss": 0.4071, + "step": 14541 + }, + { + "epoch": 0.3408873808142431, + "grad_norm": 5.571138858795166, + "learning_rate": 0.00014797293909924993, + "loss": 0.8874, + "step": 14542 + }, + { + "epoch": 0.3409108223890481, + "grad_norm": 4.135592460632324, + "learning_rate": 0.0001479664757806943, + "loss": 0.3338, + "step": 14543 + }, + { + "epoch": 0.34093426396385307, + "grad_norm": 5.274873733520508, + "learning_rate": 0.00014796001220187123, + "loss": 0.4889, + "step": 14544 + }, + { + "epoch": 0.34095770553865806, + "grad_norm": 3.4900431632995605, + "learning_rate": 0.00014795354836281578, + "loss": 0.3152, + "step": 14545 + }, + { + "epoch": 0.34098114711346306, + "grad_norm": 4.65242338180542, + "learning_rate": 0.00014794708426356306, + "loss": 0.5364, + "step": 14546 + }, + { + "epoch": 0.34100458868826805, + "grad_norm": 3.1861891746520996, + "learning_rate": 0.00014794061990414806, + "loss": 0.3195, + "step": 14547 + }, + { + "epoch": 0.3410280302630731, + "grad_norm": 2.9771149158477783, + "learning_rate": 0.00014793415528460596, + "loss": 0.2575, + "step": 14548 + }, + { + "epoch": 0.3410514718378781, + "grad_norm": 1.4145123958587646, + "learning_rate": 0.00014792769040497175, + "loss": 0.1014, + "step": 14549 + }, + { + "epoch": 0.3410749134126831, + "grad_norm": 3.4543697834014893, + "learning_rate": 0.00014792122526528057, + "loss": 0.6913, + "step": 14550 + }, + { + "epoch": 0.3410983549874881, + "grad_norm": 2.683889627456665, + "learning_rate": 0.00014791475986556743, + "loss": 0.38, + "step": 14551 + }, + { + "epoch": 0.3411217965622931, + "grad_norm": 1.579905390739441, + "learning_rate": 0.00014790829420586752, + "loss": 0.121, + "step": 14552 + }, + { + "epoch": 0.3411452381370981, + "grad_norm": 3.3434293270111084, + "learning_rate": 0.0001479018282862158, + "loss": 0.3028, + "step": 14553 + }, + { + "epoch": 0.34116867971190307, + "grad_norm": 3.2997772693634033, + "learning_rate": 0.00014789536210664742, + "loss": 0.4092, + "step": 14554 + }, + { + "epoch": 0.34119212128670806, + "grad_norm": 2.9066710472106934, + "learning_rate": 0.0001478888956671975, + "loss": 1.635, + "step": 14555 + }, + { + "epoch": 0.34121556286151306, + "grad_norm": 4.68048095703125, + "learning_rate": 0.00014788242896790105, + "loss": 0.3488, + "step": 14556 + }, + { + "epoch": 0.34123900443631805, + "grad_norm": 2.106403350830078, + "learning_rate": 0.00014787596200879319, + "loss": 0.2241, + "step": 14557 + }, + { + "epoch": 0.34126244601112304, + "grad_norm": 3.038943290710449, + "learning_rate": 0.00014786949478990904, + "loss": 0.4492, + "step": 14558 + }, + { + "epoch": 0.34128588758592804, + "grad_norm": 0.9713904857635498, + "learning_rate": 0.00014786302731128366, + "loss": 0.1344, + "step": 14559 + }, + { + "epoch": 0.34130932916073303, + "grad_norm": 1.1863837242126465, + "learning_rate": 0.00014785655957295213, + "loss": 0.1518, + "step": 14560 + }, + { + "epoch": 0.341332770735538, + "grad_norm": 0.796309769153595, + "learning_rate": 0.00014785009157494957, + "loss": 0.0601, + "step": 14561 + }, + { + "epoch": 0.341356212310343, + "grad_norm": 2.6862106323242188, + "learning_rate": 0.0001478436233173111, + "loss": 0.3609, + "step": 14562 + }, + { + "epoch": 0.341379653885148, + "grad_norm": 5.372072219848633, + "learning_rate": 0.00014783715480007174, + "loss": 0.388, + "step": 14563 + }, + { + "epoch": 0.341403095459953, + "grad_norm": 2.013258934020996, + "learning_rate": 0.00014783068602326665, + "loss": 1.0864, + "step": 14564 + }, + { + "epoch": 0.341426537034758, + "grad_norm": 2.9475834369659424, + "learning_rate": 0.00014782421698693093, + "loss": 0.4191, + "step": 14565 + }, + { + "epoch": 0.341449978609563, + "grad_norm": 3.6164450645446777, + "learning_rate": 0.0001478177476910996, + "loss": 0.4936, + "step": 14566 + }, + { + "epoch": 0.341473420184368, + "grad_norm": 3.848487138748169, + "learning_rate": 0.0001478112781358079, + "loss": 0.3476, + "step": 14567 + }, + { + "epoch": 0.341496861759173, + "grad_norm": 3.4721107482910156, + "learning_rate": 0.00014780480832109083, + "loss": 0.212, + "step": 14568 + }, + { + "epoch": 0.341520303333978, + "grad_norm": 3.682769775390625, + "learning_rate": 0.00014779833824698355, + "loss": 0.3231, + "step": 14569 + }, + { + "epoch": 0.341543744908783, + "grad_norm": 6.20226526260376, + "learning_rate": 0.00014779186791352112, + "loss": 0.7845, + "step": 14570 + }, + { + "epoch": 0.34156718648358797, + "grad_norm": 3.6707253456115723, + "learning_rate": 0.00014778539732073867, + "loss": 0.3425, + "step": 14571 + }, + { + "epoch": 0.34159062805839296, + "grad_norm": 3.552618980407715, + "learning_rate": 0.0001477789264686713, + "loss": 1.5518, + "step": 14572 + }, + { + "epoch": 0.34161406963319796, + "grad_norm": 3.999427556991577, + "learning_rate": 0.00014777245535735413, + "loss": 0.5212, + "step": 14573 + }, + { + "epoch": 0.34163751120800295, + "grad_norm": 5.601866245269775, + "learning_rate": 0.0001477659839868223, + "loss": 0.4289, + "step": 14574 + }, + { + "epoch": 0.34166095278280795, + "grad_norm": 3.595567226409912, + "learning_rate": 0.00014775951235711089, + "loss": 0.3825, + "step": 14575 + }, + { + "epoch": 0.34168439435761294, + "grad_norm": 2.384467601776123, + "learning_rate": 0.000147753040468255, + "loss": 1.6494, + "step": 14576 + }, + { + "epoch": 0.34170783593241794, + "grad_norm": 1.1386833190917969, + "learning_rate": 0.0001477465683202898, + "loss": 0.1394, + "step": 14577 + }, + { + "epoch": 0.34173127750722293, + "grad_norm": 2.153463125228882, + "learning_rate": 0.00014774009591325034, + "loss": 0.2096, + "step": 14578 + }, + { + "epoch": 0.3417547190820279, + "grad_norm": 1.9635788202285767, + "learning_rate": 0.00014773362324717182, + "loss": 0.154, + "step": 14579 + }, + { + "epoch": 0.3417781606568329, + "grad_norm": 4.147975444793701, + "learning_rate": 0.0001477271503220893, + "loss": 0.5246, + "step": 14580 + }, + { + "epoch": 0.3418016022316379, + "grad_norm": 4.398092269897461, + "learning_rate": 0.00014772067713803794, + "loss": 0.4518, + "step": 14581 + }, + { + "epoch": 0.3418250438064429, + "grad_norm": 2.5672671794891357, + "learning_rate": 0.0001477142036950528, + "loss": 0.2108, + "step": 14582 + }, + { + "epoch": 0.3418484853812479, + "grad_norm": 3.9274628162384033, + "learning_rate": 0.00014770772999316912, + "loss": 0.4364, + "step": 14583 + }, + { + "epoch": 0.3418719269560529, + "grad_norm": 2.860243320465088, + "learning_rate": 0.00014770125603242188, + "loss": 0.3235, + "step": 14584 + }, + { + "epoch": 0.3418953685308579, + "grad_norm": 2.362102746963501, + "learning_rate": 0.00014769478181284634, + "loss": 0.345, + "step": 14585 + }, + { + "epoch": 0.3419188101056629, + "grad_norm": 0.8576627373695374, + "learning_rate": 0.00014768830733447751, + "loss": 0.1052, + "step": 14586 + }, + { + "epoch": 0.3419422516804679, + "grad_norm": 1.5231267213821411, + "learning_rate": 0.00014768183259735064, + "loss": 0.1919, + "step": 14587 + }, + { + "epoch": 0.3419656932552729, + "grad_norm": 3.991889476776123, + "learning_rate": 0.00014767535760150082, + "loss": 0.3722, + "step": 14588 + }, + { + "epoch": 0.34198913483007787, + "grad_norm": 4.120297908782959, + "learning_rate": 0.0001476688823469631, + "loss": 0.3684, + "step": 14589 + }, + { + "epoch": 0.34201257640488286, + "grad_norm": 0.8837612271308899, + "learning_rate": 0.00014766240683377274, + "loss": 0.1499, + "step": 14590 + }, + { + "epoch": 0.34203601797968786, + "grad_norm": 2.8259618282318115, + "learning_rate": 0.00014765593106196479, + "loss": 0.3908, + "step": 14591 + }, + { + "epoch": 0.34205945955449285, + "grad_norm": 2.6740036010742188, + "learning_rate": 0.0001476494550315744, + "loss": 0.3143, + "step": 14592 + }, + { + "epoch": 0.34208290112929784, + "grad_norm": 4.298288822174072, + "learning_rate": 0.00014764297874263676, + "loss": 0.4963, + "step": 14593 + }, + { + "epoch": 0.34210634270410284, + "grad_norm": 2.028900384902954, + "learning_rate": 0.000147636502195187, + "loss": 0.0986, + "step": 14594 + }, + { + "epoch": 0.34212978427890783, + "grad_norm": 2.307386875152588, + "learning_rate": 0.00014763002538926021, + "loss": 0.3748, + "step": 14595 + }, + { + "epoch": 0.34215322585371283, + "grad_norm": 4.604424953460693, + "learning_rate": 0.0001476235483248916, + "loss": 0.82, + "step": 14596 + }, + { + "epoch": 0.3421766674285178, + "grad_norm": 3.7185556888580322, + "learning_rate": 0.00014761707100211622, + "loss": 0.2151, + "step": 14597 + }, + { + "epoch": 0.3422001090033228, + "grad_norm": 1.8334492444992065, + "learning_rate": 0.00014761059342096927, + "loss": 0.1534, + "step": 14598 + }, + { + "epoch": 0.3422235505781278, + "grad_norm": 1.4205610752105713, + "learning_rate": 0.00014760411558148594, + "loss": 0.1194, + "step": 14599 + }, + { + "epoch": 0.34224699215293286, + "grad_norm": 4.957028865814209, + "learning_rate": 0.00014759763748370134, + "loss": 0.6265, + "step": 14600 + }, + { + "epoch": 0.34227043372773786, + "grad_norm": 6.097722053527832, + "learning_rate": 0.0001475911591276506, + "loss": 0.4251, + "step": 14601 + }, + { + "epoch": 0.34229387530254285, + "grad_norm": 2.672367811203003, + "learning_rate": 0.0001475846805133689, + "loss": 0.174, + "step": 14602 + }, + { + "epoch": 0.34231731687734784, + "grad_norm": 3.103938102722168, + "learning_rate": 0.0001475782016408914, + "loss": 0.1968, + "step": 14603 + }, + { + "epoch": 0.34234075845215284, + "grad_norm": 1.9619654417037964, + "learning_rate": 0.00014757172251025325, + "loss": 0.2327, + "step": 14604 + }, + { + "epoch": 0.34236420002695783, + "grad_norm": 0.7996669411659241, + "learning_rate": 0.00014756524312148955, + "loss": 0.0767, + "step": 14605 + }, + { + "epoch": 0.3423876416017628, + "grad_norm": 3.726057767868042, + "learning_rate": 0.00014755876347463553, + "loss": 0.5425, + "step": 14606 + }, + { + "epoch": 0.3424110831765678, + "grad_norm": 1.123177170753479, + "learning_rate": 0.00014755228356972633, + "loss": 0.1319, + "step": 14607 + }, + { + "epoch": 0.3424345247513728, + "grad_norm": 3.6787772178649902, + "learning_rate": 0.0001475458034067971, + "loss": 0.4229, + "step": 14608 + }, + { + "epoch": 0.3424579663261778, + "grad_norm": 3.451801061630249, + "learning_rate": 0.00014753932298588296, + "loss": 0.437, + "step": 14609 + }, + { + "epoch": 0.3424814079009828, + "grad_norm": 4.700337886810303, + "learning_rate": 0.0001475328423070192, + "loss": 0.7381, + "step": 14610 + }, + { + "epoch": 0.3425048494757878, + "grad_norm": 1.3631649017333984, + "learning_rate": 0.00014752636137024082, + "loss": 0.1564, + "step": 14611 + }, + { + "epoch": 0.3425282910505928, + "grad_norm": 3.3393306732177734, + "learning_rate": 0.00014751988017558312, + "loss": 2.1141, + "step": 14612 + }, + { + "epoch": 0.3425517326253978, + "grad_norm": 5.8683929443359375, + "learning_rate": 0.00014751339872308119, + "loss": 0.6849, + "step": 14613 + }, + { + "epoch": 0.3425751742002028, + "grad_norm": 3.9977567195892334, + "learning_rate": 0.0001475069170127702, + "loss": 0.5164, + "step": 14614 + }, + { + "epoch": 0.3425986157750078, + "grad_norm": 4.555577754974365, + "learning_rate": 0.00014750043504468537, + "loss": 0.65, + "step": 14615 + }, + { + "epoch": 0.34262205734981277, + "grad_norm": 2.8324761390686035, + "learning_rate": 0.00014749395281886187, + "loss": 0.255, + "step": 14616 + }, + { + "epoch": 0.34264549892461776, + "grad_norm": 2.937554121017456, + "learning_rate": 0.0001474874703353348, + "loss": 0.2873, + "step": 14617 + }, + { + "epoch": 0.34266894049942276, + "grad_norm": 1.874928593635559, + "learning_rate": 0.0001474809875941394, + "loss": 0.2321, + "step": 14618 + }, + { + "epoch": 0.34269238207422775, + "grad_norm": 3.418553590774536, + "learning_rate": 0.00014747450459531085, + "loss": 0.4841, + "step": 14619 + }, + { + "epoch": 0.34271582364903275, + "grad_norm": 3.3726694583892822, + "learning_rate": 0.00014746802133888426, + "loss": 0.4805, + "step": 14620 + }, + { + "epoch": 0.34273926522383774, + "grad_norm": 4.440742492675781, + "learning_rate": 0.00014746153782489488, + "loss": 0.3426, + "step": 14621 + }, + { + "epoch": 0.34276270679864274, + "grad_norm": 2.7436978816986084, + "learning_rate": 0.00014745505405337785, + "loss": 0.2288, + "step": 14622 + }, + { + "epoch": 0.34278614837344773, + "grad_norm": 3.2837471961975098, + "learning_rate": 0.0001474485700243684, + "loss": 0.4061, + "step": 14623 + }, + { + "epoch": 0.3428095899482527, + "grad_norm": 3.1713268756866455, + "learning_rate": 0.0001474420857379016, + "loss": 0.4158, + "step": 14624 + }, + { + "epoch": 0.3428330315230577, + "grad_norm": 4.253812313079834, + "learning_rate": 0.00014743560119401277, + "loss": 0.5358, + "step": 14625 + }, + { + "epoch": 0.3428564730978627, + "grad_norm": 2.7796630859375, + "learning_rate": 0.00014742911639273703, + "loss": 0.2022, + "step": 14626 + }, + { + "epoch": 0.3428799146726677, + "grad_norm": 0.758663535118103, + "learning_rate": 0.00014742263133410953, + "loss": 0.1033, + "step": 14627 + }, + { + "epoch": 0.3429033562474727, + "grad_norm": 3.6124887466430664, + "learning_rate": 0.0001474161460181655, + "loss": 0.3499, + "step": 14628 + }, + { + "epoch": 0.3429267978222777, + "grad_norm": 5.273801326751709, + "learning_rate": 0.00014740966044494018, + "loss": 0.7722, + "step": 14629 + }, + { + "epoch": 0.3429502393970827, + "grad_norm": 1.977663278579712, + "learning_rate": 0.00014740317461446866, + "loss": 0.211, + "step": 14630 + }, + { + "epoch": 0.3429736809718877, + "grad_norm": 1.7681821584701538, + "learning_rate": 0.00014739668852678623, + "loss": 0.7555, + "step": 14631 + }, + { + "epoch": 0.3429971225466927, + "grad_norm": 3.7878997325897217, + "learning_rate": 0.000147390202181928, + "loss": 0.4965, + "step": 14632 + }, + { + "epoch": 0.3430205641214977, + "grad_norm": 6.168734550476074, + "learning_rate": 0.0001473837155799292, + "loss": 0.7838, + "step": 14633 + }, + { + "epoch": 0.34304400569630267, + "grad_norm": 2.250062942504883, + "learning_rate": 0.00014737722872082504, + "loss": 0.2679, + "step": 14634 + }, + { + "epoch": 0.34306744727110766, + "grad_norm": 4.0039191246032715, + "learning_rate": 0.00014737074160465073, + "loss": 0.4109, + "step": 14635 + }, + { + "epoch": 0.34309088884591266, + "grad_norm": 3.1029281616210938, + "learning_rate": 0.00014736425423144138, + "loss": 1.5297, + "step": 14636 + }, + { + "epoch": 0.34311433042071765, + "grad_norm": 5.885574817657471, + "learning_rate": 0.00014735776660123231, + "loss": 0.7113, + "step": 14637 + }, + { + "epoch": 0.34313777199552264, + "grad_norm": 1.4446754455566406, + "learning_rate": 0.00014735127871405866, + "loss": 0.0481, + "step": 14638 + }, + { + "epoch": 0.34316121357032764, + "grad_norm": 1.4503602981567383, + "learning_rate": 0.0001473447905699556, + "loss": 0.1919, + "step": 14639 + }, + { + "epoch": 0.34318465514513263, + "grad_norm": 5.244884014129639, + "learning_rate": 0.0001473383021689584, + "loss": 0.4494, + "step": 14640 + }, + { + "epoch": 0.3432080967199376, + "grad_norm": 0.9527187347412109, + "learning_rate": 0.00014733181351110224, + "loss": 0.1158, + "step": 14641 + }, + { + "epoch": 0.3432315382947426, + "grad_norm": 0.8708509206771851, + "learning_rate": 0.00014732532459642235, + "loss": 0.074, + "step": 14642 + }, + { + "epoch": 0.3432549798695476, + "grad_norm": 4.825735569000244, + "learning_rate": 0.0001473188354249539, + "loss": 0.5889, + "step": 14643 + }, + { + "epoch": 0.3432784214443526, + "grad_norm": 1.7675424814224243, + "learning_rate": 0.0001473123459967321, + "loss": 0.0529, + "step": 14644 + }, + { + "epoch": 0.3433018630191576, + "grad_norm": 3.5019516944885254, + "learning_rate": 0.0001473058563117922, + "loss": 0.4448, + "step": 14645 + }, + { + "epoch": 0.3433253045939626, + "grad_norm": 3.6012446880340576, + "learning_rate": 0.00014729936637016938, + "loss": 0.2981, + "step": 14646 + }, + { + "epoch": 0.3433487461687676, + "grad_norm": 2.2478206157684326, + "learning_rate": 0.0001472928761718989, + "loss": 1.223, + "step": 14647 + }, + { + "epoch": 0.3433721877435726, + "grad_norm": 2.7072017192840576, + "learning_rate": 0.00014728638571701594, + "loss": 0.2979, + "step": 14648 + }, + { + "epoch": 0.3433956293183776, + "grad_norm": 4.230391502380371, + "learning_rate": 0.0001472798950055557, + "loss": 0.3782, + "step": 14649 + }, + { + "epoch": 0.3434190708931826, + "grad_norm": 1.5454014539718628, + "learning_rate": 0.00014727340403755343, + "loss": 0.1937, + "step": 14650 + }, + { + "epoch": 0.3434425124679876, + "grad_norm": 2.5117902755737305, + "learning_rate": 0.00014726691281304432, + "loss": 0.2264, + "step": 14651 + }, + { + "epoch": 0.3434659540427926, + "grad_norm": 4.591550350189209, + "learning_rate": 0.00014726042133206364, + "loss": 0.6117, + "step": 14652 + }, + { + "epoch": 0.3434893956175976, + "grad_norm": 1.2451869249343872, + "learning_rate": 0.00014725392959464656, + "loss": 0.1537, + "step": 14653 + }, + { + "epoch": 0.3435128371924026, + "grad_norm": 2.993056535720825, + "learning_rate": 0.00014724743760082836, + "loss": 0.3081, + "step": 14654 + }, + { + "epoch": 0.3435362787672076, + "grad_norm": 3.796858310699463, + "learning_rate": 0.0001472409453506442, + "loss": 0.4361, + "step": 14655 + }, + { + "epoch": 0.3435597203420126, + "grad_norm": 2.1197268962860107, + "learning_rate": 0.00014723445284412938, + "loss": 0.1031, + "step": 14656 + }, + { + "epoch": 0.3435831619168176, + "grad_norm": 1.3856440782546997, + "learning_rate": 0.00014722796008131907, + "loss": 0.1161, + "step": 14657 + }, + { + "epoch": 0.3436066034916226, + "grad_norm": 4.828527450561523, + "learning_rate": 0.00014722146706224853, + "loss": 0.4868, + "step": 14658 + }, + { + "epoch": 0.3436300450664276, + "grad_norm": 2.508434295654297, + "learning_rate": 0.00014721497378695296, + "loss": 0.2375, + "step": 14659 + }, + { + "epoch": 0.3436534866412326, + "grad_norm": 0.7000135779380798, + "learning_rate": 0.00014720848025546763, + "loss": 0.0635, + "step": 14660 + }, + { + "epoch": 0.34367692821603757, + "grad_norm": 1.0094335079193115, + "learning_rate": 0.00014720198646782777, + "loss": 0.0801, + "step": 14661 + }, + { + "epoch": 0.34370036979084256, + "grad_norm": 3.711235761642456, + "learning_rate": 0.0001471954924240686, + "loss": 0.3996, + "step": 14662 + }, + { + "epoch": 0.34372381136564756, + "grad_norm": 2.9258968830108643, + "learning_rate": 0.00014718899812422533, + "loss": 1.4761, + "step": 14663 + }, + { + "epoch": 0.34374725294045255, + "grad_norm": 2.659247636795044, + "learning_rate": 0.00014718250356833326, + "loss": 1.3262, + "step": 14664 + }, + { + "epoch": 0.34377069451525755, + "grad_norm": 4.292162895202637, + "learning_rate": 0.00014717600875642758, + "loss": 0.6145, + "step": 14665 + }, + { + "epoch": 0.34379413609006254, + "grad_norm": 2.2786662578582764, + "learning_rate": 0.00014716951368854358, + "loss": 0.2, + "step": 14666 + }, + { + "epoch": 0.34381757766486754, + "grad_norm": 4.733072280883789, + "learning_rate": 0.00014716301836471645, + "loss": 0.4156, + "step": 14667 + }, + { + "epoch": 0.34384101923967253, + "grad_norm": 4.525859355926514, + "learning_rate": 0.00014715652278498147, + "loss": 0.5382, + "step": 14668 + }, + { + "epoch": 0.3438644608144775, + "grad_norm": 4.240720748901367, + "learning_rate": 0.00014715002694937386, + "loss": 0.3805, + "step": 14669 + }, + { + "epoch": 0.3438879023892825, + "grad_norm": 1.7005270719528198, + "learning_rate": 0.00014714353085792887, + "loss": 0.732, + "step": 14670 + }, + { + "epoch": 0.3439113439640875, + "grad_norm": 0.7932493090629578, + "learning_rate": 0.00014713703451068176, + "loss": 0.1287, + "step": 14671 + }, + { + "epoch": 0.3439347855388925, + "grad_norm": 2.870591163635254, + "learning_rate": 0.00014713053790766777, + "loss": 0.1706, + "step": 14672 + }, + { + "epoch": 0.3439582271136975, + "grad_norm": 4.092987537384033, + "learning_rate": 0.00014712404104892218, + "loss": 0.3691, + "step": 14673 + }, + { + "epoch": 0.3439816686885025, + "grad_norm": 0.988477349281311, + "learning_rate": 0.00014711754393448022, + "loss": 0.1493, + "step": 14674 + }, + { + "epoch": 0.3440051102633075, + "grad_norm": 2.8057336807250977, + "learning_rate": 0.00014711104656437712, + "loss": 0.2873, + "step": 14675 + }, + { + "epoch": 0.3440285518381125, + "grad_norm": 0.8689825534820557, + "learning_rate": 0.00014710454893864815, + "loss": 0.0752, + "step": 14676 + }, + { + "epoch": 0.3440519934129175, + "grad_norm": 3.872565507888794, + "learning_rate": 0.00014709805105732858, + "loss": 0.3528, + "step": 14677 + }, + { + "epoch": 0.3440754349877225, + "grad_norm": 2.157083749771118, + "learning_rate": 0.0001470915529204537, + "loss": 0.3041, + "step": 14678 + }, + { + "epoch": 0.34409887656252747, + "grad_norm": 1.1885136365890503, + "learning_rate": 0.0001470850545280587, + "loss": 0.1001, + "step": 14679 + }, + { + "epoch": 0.34412231813733246, + "grad_norm": 2.5480411052703857, + "learning_rate": 0.00014707855588017886, + "loss": 1.4947, + "step": 14680 + }, + { + "epoch": 0.34414575971213746, + "grad_norm": 4.431869029998779, + "learning_rate": 0.00014707205697684944, + "loss": 0.5861, + "step": 14681 + }, + { + "epoch": 0.34416920128694245, + "grad_norm": 1.9005286693572998, + "learning_rate": 0.00014706555781810574, + "loss": 1.835, + "step": 14682 + }, + { + "epoch": 0.34419264286174744, + "grad_norm": 1.1748149394989014, + "learning_rate": 0.000147059058403983, + "loss": 0.1837, + "step": 14683 + }, + { + "epoch": 0.34421608443655244, + "grad_norm": 1.228407382965088, + "learning_rate": 0.00014705255873451647, + "loss": 0.1399, + "step": 14684 + }, + { + "epoch": 0.34423952601135743, + "grad_norm": 4.2467851638793945, + "learning_rate": 0.00014704605880974146, + "loss": 0.4801, + "step": 14685 + }, + { + "epoch": 0.3442629675861624, + "grad_norm": 1.4281456470489502, + "learning_rate": 0.0001470395586296932, + "loss": 0.1411, + "step": 14686 + }, + { + "epoch": 0.3442864091609674, + "grad_norm": 3.4507806301116943, + "learning_rate": 0.00014703305819440695, + "loss": 0.2367, + "step": 14687 + }, + { + "epoch": 0.3443098507357724, + "grad_norm": 0.8458910584449768, + "learning_rate": 0.00014702655750391803, + "loss": 0.0914, + "step": 14688 + }, + { + "epoch": 0.3443332923105774, + "grad_norm": 0.6192181706428528, + "learning_rate": 0.00014702005655826167, + "loss": 0.0489, + "step": 14689 + }, + { + "epoch": 0.3443567338853824, + "grad_norm": 0.9591554403305054, + "learning_rate": 0.00014701355535747315, + "loss": 0.1188, + "step": 14690 + }, + { + "epoch": 0.3443801754601874, + "grad_norm": 3.8685812950134277, + "learning_rate": 0.0001470070539015878, + "loss": 0.4099, + "step": 14691 + }, + { + "epoch": 0.3444036170349924, + "grad_norm": 5.7585978507995605, + "learning_rate": 0.00014700055219064084, + "loss": 0.501, + "step": 14692 + }, + { + "epoch": 0.3444270586097974, + "grad_norm": 3.7021100521087646, + "learning_rate": 0.00014699405022466753, + "loss": 0.4854, + "step": 14693 + }, + { + "epoch": 0.3444505001846024, + "grad_norm": 2.5266940593719482, + "learning_rate": 0.00014698754800370321, + "loss": 0.224, + "step": 14694 + }, + { + "epoch": 0.3444739417594074, + "grad_norm": 2.250948429107666, + "learning_rate": 0.00014698104552778314, + "loss": 0.6654, + "step": 14695 + }, + { + "epoch": 0.34449738333421237, + "grad_norm": 1.906369686126709, + "learning_rate": 0.00014697454279694259, + "loss": 0.1878, + "step": 14696 + }, + { + "epoch": 0.34452082490901736, + "grad_norm": 4.49439001083374, + "learning_rate": 0.00014696803981121683, + "loss": 0.4594, + "step": 14697 + }, + { + "epoch": 0.34454426648382236, + "grad_norm": 2.3860175609588623, + "learning_rate": 0.0001469615365706412, + "loss": 1.4015, + "step": 14698 + }, + { + "epoch": 0.34456770805862735, + "grad_norm": 2.5958898067474365, + "learning_rate": 0.00014695503307525094, + "loss": 0.1968, + "step": 14699 + }, + { + "epoch": 0.34459114963343235, + "grad_norm": 0.6363241672515869, + "learning_rate": 0.00014694852932508131, + "loss": 0.0499, + "step": 14700 + }, + { + "epoch": 0.34461459120823734, + "grad_norm": 0.6159413456916809, + "learning_rate": 0.00014694202532016768, + "loss": 0.0533, + "step": 14701 + }, + { + "epoch": 0.3446380327830424, + "grad_norm": 0.8135128021240234, + "learning_rate": 0.0001469355210605453, + "loss": 0.0513, + "step": 14702 + }, + { + "epoch": 0.3446614743578474, + "grad_norm": 3.275449514389038, + "learning_rate": 0.00014692901654624944, + "loss": 0.3564, + "step": 14703 + }, + { + "epoch": 0.3446849159326524, + "grad_norm": 2.728604316711426, + "learning_rate": 0.00014692251177731545, + "loss": 0.343, + "step": 14704 + }, + { + "epoch": 0.3447083575074574, + "grad_norm": 3.9683878421783447, + "learning_rate": 0.00014691600675377858, + "loss": 0.5275, + "step": 14705 + }, + { + "epoch": 0.34473179908226237, + "grad_norm": 1.5921075344085693, + "learning_rate": 0.00014690950147567414, + "loss": 0.2951, + "step": 14706 + }, + { + "epoch": 0.34475524065706736, + "grad_norm": 3.5276927947998047, + "learning_rate": 0.00014690299594303737, + "loss": 0.392, + "step": 14707 + }, + { + "epoch": 0.34477868223187236, + "grad_norm": 5.937534809112549, + "learning_rate": 0.0001468964901559037, + "loss": 0.5573, + "step": 14708 + }, + { + "epoch": 0.34480212380667735, + "grad_norm": 2.6803548336029053, + "learning_rate": 0.0001468899841143083, + "loss": 0.2493, + "step": 14709 + }, + { + "epoch": 0.34482556538148235, + "grad_norm": 2.878650665283203, + "learning_rate": 0.0001468834778182866, + "loss": 0.2823, + "step": 14710 + }, + { + "epoch": 0.34484900695628734, + "grad_norm": 2.444240093231201, + "learning_rate": 0.0001468769712678738, + "loss": 0.1809, + "step": 14711 + }, + { + "epoch": 0.34487244853109233, + "grad_norm": 1.349616527557373, + "learning_rate": 0.0001468704644631052, + "loss": 0.1907, + "step": 14712 + }, + { + "epoch": 0.34489589010589733, + "grad_norm": 3.7108755111694336, + "learning_rate": 0.00014686395740401615, + "loss": 0.6204, + "step": 14713 + }, + { + "epoch": 0.3449193316807023, + "grad_norm": 5.224969387054443, + "learning_rate": 0.00014685745009064197, + "loss": 0.5328, + "step": 14714 + }, + { + "epoch": 0.3449427732555073, + "grad_norm": 2.380641222000122, + "learning_rate": 0.00014685094252301796, + "loss": 0.1779, + "step": 14715 + }, + { + "epoch": 0.3449662148303123, + "grad_norm": 4.174582004547119, + "learning_rate": 0.00014684443470117938, + "loss": 0.502, + "step": 14716 + }, + { + "epoch": 0.3449896564051173, + "grad_norm": 1.6287175416946411, + "learning_rate": 0.00014683792662516161, + "loss": 0.0984, + "step": 14717 + }, + { + "epoch": 0.3450130979799223, + "grad_norm": 2.3896117210388184, + "learning_rate": 0.00014683141829499995, + "loss": 0.2374, + "step": 14718 + }, + { + "epoch": 0.3450365395547273, + "grad_norm": 4.113864421844482, + "learning_rate": 0.00014682490971072964, + "loss": 0.4121, + "step": 14719 + }, + { + "epoch": 0.3450599811295323, + "grad_norm": 4.108596324920654, + "learning_rate": 0.0001468184008723861, + "loss": 0.3474, + "step": 14720 + }, + { + "epoch": 0.3450834227043373, + "grad_norm": 3.911947011947632, + "learning_rate": 0.00014681189178000458, + "loss": 0.3984, + "step": 14721 + }, + { + "epoch": 0.3451068642791423, + "grad_norm": 3.371962308883667, + "learning_rate": 0.00014680538243362043, + "loss": 0.3717, + "step": 14722 + }, + { + "epoch": 0.3451303058539473, + "grad_norm": 4.223438262939453, + "learning_rate": 0.00014679887283326897, + "loss": 0.5794, + "step": 14723 + }, + { + "epoch": 0.34515374742875227, + "grad_norm": 2.656162977218628, + "learning_rate": 0.0001467923629789855, + "loss": 0.19, + "step": 14724 + }, + { + "epoch": 0.34517718900355726, + "grad_norm": 4.61085844039917, + "learning_rate": 0.00014678585287080534, + "loss": 0.3452, + "step": 14725 + }, + { + "epoch": 0.34520063057836226, + "grad_norm": 1.9857168197631836, + "learning_rate": 0.00014677934250876382, + "loss": 1.3729, + "step": 14726 + }, + { + "epoch": 0.34522407215316725, + "grad_norm": 5.148944854736328, + "learning_rate": 0.0001467728318928963, + "loss": 0.4451, + "step": 14727 + }, + { + "epoch": 0.34524751372797224, + "grad_norm": 3.673372745513916, + "learning_rate": 0.00014676632102323805, + "loss": 0.414, + "step": 14728 + }, + { + "epoch": 0.34527095530277724, + "grad_norm": 1.2977641820907593, + "learning_rate": 0.00014675980989982445, + "loss": 0.235, + "step": 14729 + }, + { + "epoch": 0.34529439687758223, + "grad_norm": 0.9136316776275635, + "learning_rate": 0.0001467532985226908, + "loss": 0.106, + "step": 14730 + }, + { + "epoch": 0.3453178384523872, + "grad_norm": 1.0358788967132568, + "learning_rate": 0.00014674678689187242, + "loss": 0.1478, + "step": 14731 + }, + { + "epoch": 0.3453412800271922, + "grad_norm": 0.9807405471801758, + "learning_rate": 0.00014674027500740467, + "loss": 0.1238, + "step": 14732 + }, + { + "epoch": 0.3453647216019972, + "grad_norm": 3.4783935546875, + "learning_rate": 0.0001467337628693229, + "loss": 0.2895, + "step": 14733 + }, + { + "epoch": 0.3453881631768022, + "grad_norm": 0.34780508279800415, + "learning_rate": 0.0001467272504776624, + "loss": 0.027, + "step": 14734 + }, + { + "epoch": 0.3454116047516072, + "grad_norm": 3.330928325653076, + "learning_rate": 0.00014672073783245852, + "loss": 0.3117, + "step": 14735 + }, + { + "epoch": 0.3454350463264122, + "grad_norm": 4.047062397003174, + "learning_rate": 0.00014671422493374661, + "loss": 0.5448, + "step": 14736 + }, + { + "epoch": 0.3454584879012172, + "grad_norm": 2.475860118865967, + "learning_rate": 0.00014670771178156198, + "loss": 0.3674, + "step": 14737 + }, + { + "epoch": 0.3454819294760222, + "grad_norm": 1.0677894353866577, + "learning_rate": 0.00014670119837594, + "loss": 0.1186, + "step": 14738 + }, + { + "epoch": 0.3455053710508272, + "grad_norm": 3.135146379470825, + "learning_rate": 0.00014669468471691606, + "loss": 0.2974, + "step": 14739 + }, + { + "epoch": 0.3455288126256322, + "grad_norm": 3.524341106414795, + "learning_rate": 0.0001466881708045254, + "loss": 1.9071, + "step": 14740 + }, + { + "epoch": 0.34555225420043717, + "grad_norm": 3.5652055740356445, + "learning_rate": 0.0001466816566388034, + "loss": 0.4606, + "step": 14741 + }, + { + "epoch": 0.34557569577524216, + "grad_norm": 3.042940378189087, + "learning_rate": 0.00014667514221978546, + "loss": 0.3925, + "step": 14742 + }, + { + "epoch": 0.34559913735004716, + "grad_norm": 1.572847843170166, + "learning_rate": 0.00014666862754750682, + "loss": 0.1496, + "step": 14743 + }, + { + "epoch": 0.34562257892485215, + "grad_norm": 1.0680338144302368, + "learning_rate": 0.00014666211262200296, + "loss": 0.1454, + "step": 14744 + }, + { + "epoch": 0.34564602049965715, + "grad_norm": 2.8196656703948975, + "learning_rate": 0.0001466555974433091, + "loss": 0.3205, + "step": 14745 + }, + { + "epoch": 0.34566946207446214, + "grad_norm": 3.23195743560791, + "learning_rate": 0.00014664908201146074, + "loss": 0.4519, + "step": 14746 + }, + { + "epoch": 0.34569290364926714, + "grad_norm": 3.3298747539520264, + "learning_rate": 0.0001466425663264931, + "loss": 0.2477, + "step": 14747 + }, + { + "epoch": 0.34571634522407213, + "grad_norm": 1.1767292022705078, + "learning_rate": 0.00014663605038844158, + "loss": 0.0949, + "step": 14748 + }, + { + "epoch": 0.3457397867988771, + "grad_norm": 0.9666867256164551, + "learning_rate": 0.00014662953419734156, + "loss": 0.1515, + "step": 14749 + }, + { + "epoch": 0.3457632283736821, + "grad_norm": 2.780150890350342, + "learning_rate": 0.00014662301775322838, + "loss": 0.1575, + "step": 14750 + }, + { + "epoch": 0.3457866699484871, + "grad_norm": 4.300987720489502, + "learning_rate": 0.00014661650105613736, + "loss": 0.5378, + "step": 14751 + }, + { + "epoch": 0.3458101115232921, + "grad_norm": 4.248114109039307, + "learning_rate": 0.00014660998410610394, + "loss": 0.4296, + "step": 14752 + }, + { + "epoch": 0.34583355309809716, + "grad_norm": 2.8314037322998047, + "learning_rate": 0.0001466034669031634, + "loss": 0.3811, + "step": 14753 + }, + { + "epoch": 0.34585699467290215, + "grad_norm": 2.9533584117889404, + "learning_rate": 0.00014659694944735115, + "loss": 1.0661, + "step": 14754 + }, + { + "epoch": 0.34588043624770715, + "grad_norm": 5.347227096557617, + "learning_rate": 0.00014659043173870253, + "loss": 0.5502, + "step": 14755 + }, + { + "epoch": 0.34590387782251214, + "grad_norm": 1.3859646320343018, + "learning_rate": 0.00014658391377725294, + "loss": 0.1153, + "step": 14756 + }, + { + "epoch": 0.34592731939731713, + "grad_norm": 0.9293445348739624, + "learning_rate": 0.00014657739556303768, + "loss": 0.0783, + "step": 14757 + }, + { + "epoch": 0.34595076097212213, + "grad_norm": 3.7351393699645996, + "learning_rate": 0.0001465708770960922, + "loss": 0.3096, + "step": 14758 + }, + { + "epoch": 0.3459742025469271, + "grad_norm": 3.4317901134490967, + "learning_rate": 0.00014656435837645185, + "loss": 0.5749, + "step": 14759 + }, + { + "epoch": 0.3459976441217321, + "grad_norm": 2.6491684913635254, + "learning_rate": 0.00014655783940415194, + "loss": 0.2883, + "step": 14760 + }, + { + "epoch": 0.3460210856965371, + "grad_norm": 3.5072216987609863, + "learning_rate": 0.0001465513201792279, + "loss": 0.1469, + "step": 14761 + }, + { + "epoch": 0.3460445272713421, + "grad_norm": 2.3601551055908203, + "learning_rate": 0.0001465448007017151, + "loss": 0.2565, + "step": 14762 + }, + { + "epoch": 0.3460679688461471, + "grad_norm": 2.050306797027588, + "learning_rate": 0.0001465382809716489, + "loss": 1.1357, + "step": 14763 + }, + { + "epoch": 0.3460914104209521, + "grad_norm": 0.6678047180175781, + "learning_rate": 0.00014653176098906466, + "loss": 0.0673, + "step": 14764 + }, + { + "epoch": 0.3461148519957571, + "grad_norm": 0.9160787463188171, + "learning_rate": 0.00014652524075399782, + "loss": 0.0903, + "step": 14765 + }, + { + "epoch": 0.3461382935705621, + "grad_norm": 3.339221715927124, + "learning_rate": 0.00014651872026648366, + "loss": 0.2306, + "step": 14766 + }, + { + "epoch": 0.3461617351453671, + "grad_norm": 4.296627521514893, + "learning_rate": 0.0001465121995265576, + "loss": 0.3812, + "step": 14767 + }, + { + "epoch": 0.34618517672017207, + "grad_norm": 3.6897103786468506, + "learning_rate": 0.00014650567853425513, + "loss": 0.211, + "step": 14768 + }, + { + "epoch": 0.34620861829497707, + "grad_norm": 2.412457227706909, + "learning_rate": 0.00014649915728961148, + "loss": 0.1173, + "step": 14769 + }, + { + "epoch": 0.34623205986978206, + "grad_norm": 4.724392890930176, + "learning_rate": 0.00014649263579266208, + "loss": 0.5899, + "step": 14770 + }, + { + "epoch": 0.34625550144458705, + "grad_norm": 2.0425398349761963, + "learning_rate": 0.00014648611404344237, + "loss": 0.1764, + "step": 14771 + }, + { + "epoch": 0.34627894301939205, + "grad_norm": 2.3981311321258545, + "learning_rate": 0.00014647959204198768, + "loss": 0.2517, + "step": 14772 + }, + { + "epoch": 0.34630238459419704, + "grad_norm": 3.788067579269409, + "learning_rate": 0.0001464730697883334, + "loss": 1.4368, + "step": 14773 + }, + { + "epoch": 0.34632582616900204, + "grad_norm": 3.990859270095825, + "learning_rate": 0.00014646654728251498, + "loss": 0.2524, + "step": 14774 + }, + { + "epoch": 0.34634926774380703, + "grad_norm": 1.7512269020080566, + "learning_rate": 0.00014646002452456776, + "loss": 0.1596, + "step": 14775 + }, + { + "epoch": 0.346372709318612, + "grad_norm": 1.9016218185424805, + "learning_rate": 0.0001464535015145271, + "loss": 0.1266, + "step": 14776 + }, + { + "epoch": 0.346396150893417, + "grad_norm": 2.165177822113037, + "learning_rate": 0.00014644697825242844, + "loss": 1.1139, + "step": 14777 + }, + { + "epoch": 0.346419592468222, + "grad_norm": 2.182591676712036, + "learning_rate": 0.00014644045473830722, + "loss": 0.1363, + "step": 14778 + }, + { + "epoch": 0.346443034043027, + "grad_norm": 3.242340087890625, + "learning_rate": 0.00014643393097219876, + "loss": 0.4303, + "step": 14779 + }, + { + "epoch": 0.346466475617832, + "grad_norm": 3.1742849349975586, + "learning_rate": 0.00014642740695413846, + "loss": 0.3812, + "step": 14780 + }, + { + "epoch": 0.346489917192637, + "grad_norm": 2.8513238430023193, + "learning_rate": 0.00014642088268416176, + "loss": 0.2785, + "step": 14781 + }, + { + "epoch": 0.346513358767442, + "grad_norm": 6.928471565246582, + "learning_rate": 0.00014641435816230405, + "loss": 0.5835, + "step": 14782 + }, + { + "epoch": 0.346536800342247, + "grad_norm": 3.2913930416107178, + "learning_rate": 0.00014640783338860073, + "loss": 0.3576, + "step": 14783 + }, + { + "epoch": 0.346560241917052, + "grad_norm": 1.9747954607009888, + "learning_rate": 0.0001464013083630872, + "loss": 0.3228, + "step": 14784 + }, + { + "epoch": 0.346583683491857, + "grad_norm": 2.3113837242126465, + "learning_rate": 0.00014639478308579886, + "loss": 0.2295, + "step": 14785 + }, + { + "epoch": 0.34660712506666197, + "grad_norm": 4.485049247741699, + "learning_rate": 0.0001463882575567711, + "loss": 0.7196, + "step": 14786 + }, + { + "epoch": 0.34663056664146696, + "grad_norm": 1.5859856605529785, + "learning_rate": 0.0001463817317760394, + "loss": 0.1152, + "step": 14787 + }, + { + "epoch": 0.34665400821627196, + "grad_norm": 2.744680881500244, + "learning_rate": 0.00014637520574363913, + "loss": 0.1713, + "step": 14788 + }, + { + "epoch": 0.34667744979107695, + "grad_norm": 3.1427249908447266, + "learning_rate": 0.00014636867945960562, + "loss": 1.446, + "step": 14789 + }, + { + "epoch": 0.34670089136588195, + "grad_norm": 3.579580307006836, + "learning_rate": 0.0001463621529239744, + "loss": 0.3595, + "step": 14790 + }, + { + "epoch": 0.34672433294068694, + "grad_norm": 3.215836524963379, + "learning_rate": 0.00014635562613678083, + "loss": 0.3175, + "step": 14791 + }, + { + "epoch": 0.34674777451549194, + "grad_norm": 4.2397966384887695, + "learning_rate": 0.0001463490990980603, + "loss": 0.4831, + "step": 14792 + }, + { + "epoch": 0.34677121609029693, + "grad_norm": 4.675337791442871, + "learning_rate": 0.0001463425718078483, + "loss": 0.4338, + "step": 14793 + }, + { + "epoch": 0.3467946576651019, + "grad_norm": 2.215956211090088, + "learning_rate": 0.00014633604426618016, + "loss": 0.117, + "step": 14794 + }, + { + "epoch": 0.3468180992399069, + "grad_norm": 3.855570077896118, + "learning_rate": 0.00014632951647309137, + "loss": 0.4317, + "step": 14795 + }, + { + "epoch": 0.3468415408147119, + "grad_norm": 2.88832426071167, + "learning_rate": 0.0001463229884286173, + "loss": 0.3482, + "step": 14796 + }, + { + "epoch": 0.3468649823895169, + "grad_norm": 0.6862848997116089, + "learning_rate": 0.0001463164601327934, + "loss": 0.1081, + "step": 14797 + }, + { + "epoch": 0.3468884239643219, + "grad_norm": 3.3366124629974365, + "learning_rate": 0.0001463099315856551, + "loss": 0.4184, + "step": 14798 + }, + { + "epoch": 0.3469118655391269, + "grad_norm": 2.814972162246704, + "learning_rate": 0.00014630340278723778, + "loss": 1.4835, + "step": 14799 + }, + { + "epoch": 0.3469353071139319, + "grad_norm": 5.956217288970947, + "learning_rate": 0.00014629687373757692, + "loss": 0.4951, + "step": 14800 + }, + { + "epoch": 0.3469587486887369, + "grad_norm": 2.8417274951934814, + "learning_rate": 0.0001462903444367079, + "loss": 0.3395, + "step": 14801 + }, + { + "epoch": 0.3469821902635419, + "grad_norm": 2.781736373901367, + "learning_rate": 0.0001462838148846662, + "loss": 0.3495, + "step": 14802 + }, + { + "epoch": 0.3470056318383469, + "grad_norm": 2.4788413047790527, + "learning_rate": 0.0001462772850814872, + "loss": 0.3863, + "step": 14803 + }, + { + "epoch": 0.3470290734131519, + "grad_norm": 5.0716047286987305, + "learning_rate": 0.00014627075502720636, + "loss": 0.333, + "step": 14804 + }, + { + "epoch": 0.3470525149879569, + "grad_norm": 3.925490617752075, + "learning_rate": 0.00014626422472185908, + "loss": 0.2486, + "step": 14805 + }, + { + "epoch": 0.3470759565627619, + "grad_norm": 3.1639015674591064, + "learning_rate": 0.00014625769416548083, + "loss": 0.3355, + "step": 14806 + }, + { + "epoch": 0.3470993981375669, + "grad_norm": 3.2848451137542725, + "learning_rate": 0.00014625116335810704, + "loss": 0.1979, + "step": 14807 + }, + { + "epoch": 0.3471228397123719, + "grad_norm": 3.251037359237671, + "learning_rate": 0.00014624463229977313, + "loss": 1.9343, + "step": 14808 + }, + { + "epoch": 0.3471462812871769, + "grad_norm": 2.470231533050537, + "learning_rate": 0.00014623810099051454, + "loss": 1.5517, + "step": 14809 + }, + { + "epoch": 0.3471697228619819, + "grad_norm": 2.78303861618042, + "learning_rate": 0.0001462315694303667, + "loss": 0.194, + "step": 14810 + }, + { + "epoch": 0.3471931644367869, + "grad_norm": 2.3324005603790283, + "learning_rate": 0.00014622503761936508, + "loss": 0.2026, + "step": 14811 + }, + { + "epoch": 0.3472166060115919, + "grad_norm": 1.2722843885421753, + "learning_rate": 0.00014621850555754512, + "loss": 0.1264, + "step": 14812 + }, + { + "epoch": 0.34724004758639687, + "grad_norm": 1.3405894041061401, + "learning_rate": 0.00014621197324494225, + "loss": 0.172, + "step": 14813 + }, + { + "epoch": 0.34726348916120187, + "grad_norm": 0.9187163710594177, + "learning_rate": 0.00014620544068159189, + "loss": 0.0974, + "step": 14814 + }, + { + "epoch": 0.34728693073600686, + "grad_norm": 7.944971084594727, + "learning_rate": 0.0001461989078675295, + "loss": 0.4351, + "step": 14815 + }, + { + "epoch": 0.34731037231081185, + "grad_norm": 4.239300727844238, + "learning_rate": 0.00014619237480279057, + "loss": 0.5417, + "step": 14816 + }, + { + "epoch": 0.34733381388561685, + "grad_norm": 4.838801860809326, + "learning_rate": 0.00014618584148741052, + "loss": 0.3929, + "step": 14817 + }, + { + "epoch": 0.34735725546042184, + "grad_norm": 4.663344860076904, + "learning_rate": 0.00014617930792142477, + "loss": 0.5538, + "step": 14818 + }, + { + "epoch": 0.34738069703522684, + "grad_norm": 0.8659026622772217, + "learning_rate": 0.0001461727741048688, + "loss": 0.0724, + "step": 14819 + }, + { + "epoch": 0.34740413861003183, + "grad_norm": 4.532744884490967, + "learning_rate": 0.00014616624003777805, + "loss": 0.4995, + "step": 14820 + }, + { + "epoch": 0.3474275801848368, + "grad_norm": 4.14804744720459, + "learning_rate": 0.000146159705720188, + "loss": 0.2629, + "step": 14821 + }, + { + "epoch": 0.3474510217596418, + "grad_norm": 2.0247161388397217, + "learning_rate": 0.0001461531711521341, + "loss": 1.5055, + "step": 14822 + }, + { + "epoch": 0.3474744633344468, + "grad_norm": 2.678642511367798, + "learning_rate": 0.00014614663633365178, + "loss": 0.2682, + "step": 14823 + }, + { + "epoch": 0.3474979049092518, + "grad_norm": 2.8764634132385254, + "learning_rate": 0.00014614010126477652, + "loss": 0.2913, + "step": 14824 + }, + { + "epoch": 0.3475213464840568, + "grad_norm": 4.468255996704102, + "learning_rate": 0.00014613356594554374, + "loss": 0.5286, + "step": 14825 + }, + { + "epoch": 0.3475447880588618, + "grad_norm": 1.5818184614181519, + "learning_rate": 0.00014612703037598902, + "loss": 0.175, + "step": 14826 + }, + { + "epoch": 0.3475682296336668, + "grad_norm": 3.52496600151062, + "learning_rate": 0.00014612049455614765, + "loss": 0.2955, + "step": 14827 + }, + { + "epoch": 0.3475916712084718, + "grad_norm": 2.7238776683807373, + "learning_rate": 0.0001461139584860552, + "loss": 0.2837, + "step": 14828 + }, + { + "epoch": 0.3476151127832768, + "grad_norm": 3.570455551147461, + "learning_rate": 0.0001461074221657471, + "loss": 0.4047, + "step": 14829 + }, + { + "epoch": 0.3476385543580818, + "grad_norm": 8.04306697845459, + "learning_rate": 0.00014610088559525886, + "loss": 0.7115, + "step": 14830 + }, + { + "epoch": 0.34766199593288677, + "grad_norm": 3.819878339767456, + "learning_rate": 0.00014609434877462592, + "loss": 0.4077, + "step": 14831 + }, + { + "epoch": 0.34768543750769176, + "grad_norm": 3.852152109146118, + "learning_rate": 0.00014608781170388375, + "loss": 0.2845, + "step": 14832 + }, + { + "epoch": 0.34770887908249676, + "grad_norm": 1.0244674682617188, + "learning_rate": 0.0001460812743830678, + "loss": 0.1022, + "step": 14833 + }, + { + "epoch": 0.34773232065730175, + "grad_norm": 4.305421829223633, + "learning_rate": 0.00014607473681221356, + "loss": 0.3509, + "step": 14834 + }, + { + "epoch": 0.34775576223210675, + "grad_norm": 1.8775378465652466, + "learning_rate": 0.0001460681989913565, + "loss": 0.2252, + "step": 14835 + }, + { + "epoch": 0.34777920380691174, + "grad_norm": 2.395991086959839, + "learning_rate": 0.00014606166092053213, + "loss": 0.2009, + "step": 14836 + }, + { + "epoch": 0.34780264538171674, + "grad_norm": 2.1702721118927, + "learning_rate": 0.00014605512259977587, + "loss": 1.4294, + "step": 14837 + }, + { + "epoch": 0.34782608695652173, + "grad_norm": 3.2218472957611084, + "learning_rate": 0.00014604858402912323, + "loss": 0.3289, + "step": 14838 + }, + { + "epoch": 0.3478495285313267, + "grad_norm": 1.5167313814163208, + "learning_rate": 0.00014604204520860965, + "loss": 0.1288, + "step": 14839 + }, + { + "epoch": 0.3478729701061317, + "grad_norm": 3.4454848766326904, + "learning_rate": 0.00014603550613827065, + "loss": 0.3582, + "step": 14840 + }, + { + "epoch": 0.3478964116809367, + "grad_norm": 3.890650987625122, + "learning_rate": 0.00014602896681814172, + "loss": 0.5284, + "step": 14841 + }, + { + "epoch": 0.3479198532557417, + "grad_norm": 2.229602098464966, + "learning_rate": 0.0001460224272482583, + "loss": 0.2583, + "step": 14842 + }, + { + "epoch": 0.3479432948305467, + "grad_norm": 2.7521636486053467, + "learning_rate": 0.00014601588742865591, + "loss": 0.3362, + "step": 14843 + }, + { + "epoch": 0.3479667364053517, + "grad_norm": 3.788902521133423, + "learning_rate": 0.00014600934735937003, + "loss": 0.3965, + "step": 14844 + }, + { + "epoch": 0.3479901779801567, + "grad_norm": 3.420736312866211, + "learning_rate": 0.0001460028070404361, + "loss": 0.3332, + "step": 14845 + }, + { + "epoch": 0.3480136195549617, + "grad_norm": 3.286072015762329, + "learning_rate": 0.00014599626647188965, + "loss": 0.2267, + "step": 14846 + }, + { + "epoch": 0.3480370611297667, + "grad_norm": 3.1221890449523926, + "learning_rate": 0.00014598972565376618, + "loss": 2.0202, + "step": 14847 + }, + { + "epoch": 0.3480605027045717, + "grad_norm": 4.061990261077881, + "learning_rate": 0.00014598318458610118, + "loss": 0.2724, + "step": 14848 + }, + { + "epoch": 0.34808394427937667, + "grad_norm": 3.779623031616211, + "learning_rate": 0.0001459766432689301, + "loss": 0.4168, + "step": 14849 + }, + { + "epoch": 0.34810738585418166, + "grad_norm": 2.881777048110962, + "learning_rate": 0.00014597010170228848, + "loss": 0.2179, + "step": 14850 + }, + { + "epoch": 0.34813082742898666, + "grad_norm": 6.455265522003174, + "learning_rate": 0.0001459635598862118, + "loss": 0.7308, + "step": 14851 + }, + { + "epoch": 0.34815426900379165, + "grad_norm": 1.3286949396133423, + "learning_rate": 0.00014595701782073555, + "loss": 0.1453, + "step": 14852 + }, + { + "epoch": 0.34817771057859664, + "grad_norm": 2.5916757583618164, + "learning_rate": 0.00014595047550589524, + "loss": 0.3264, + "step": 14853 + }, + { + "epoch": 0.34820115215340164, + "grad_norm": 2.8116912841796875, + "learning_rate": 0.00014594393294172634, + "loss": 0.2766, + "step": 14854 + }, + { + "epoch": 0.3482245937282067, + "grad_norm": 3.12701416015625, + "learning_rate": 0.00014593739012826434, + "loss": 2.2391, + "step": 14855 + }, + { + "epoch": 0.3482480353030117, + "grad_norm": 2.7605090141296387, + "learning_rate": 0.0001459308470655448, + "loss": 0.1843, + "step": 14856 + }, + { + "epoch": 0.3482714768778167, + "grad_norm": 4.462523937225342, + "learning_rate": 0.0001459243037536032, + "loss": 0.6239, + "step": 14857 + }, + { + "epoch": 0.34829491845262167, + "grad_norm": 1.3305401802062988, + "learning_rate": 0.00014591776019247501, + "loss": 0.0669, + "step": 14858 + }, + { + "epoch": 0.34831836002742667, + "grad_norm": 1.591795563697815, + "learning_rate": 0.00014591121638219578, + "loss": 0.2242, + "step": 14859 + }, + { + "epoch": 0.34834180160223166, + "grad_norm": 2.6846792697906494, + "learning_rate": 0.000145904672322801, + "loss": 0.3269, + "step": 14860 + }, + { + "epoch": 0.34836524317703665, + "grad_norm": 3.335677146911621, + "learning_rate": 0.00014589812801432618, + "loss": 0.3406, + "step": 14861 + }, + { + "epoch": 0.34838868475184165, + "grad_norm": 4.846977710723877, + "learning_rate": 0.0001458915834568068, + "loss": 0.5485, + "step": 14862 + }, + { + "epoch": 0.34841212632664664, + "grad_norm": 1.5887396335601807, + "learning_rate": 0.00014588503865027846, + "loss": 0.1997, + "step": 14863 + }, + { + "epoch": 0.34843556790145164, + "grad_norm": 2.9720613956451416, + "learning_rate": 0.00014587849359477654, + "loss": 1.9744, + "step": 14864 + }, + { + "epoch": 0.34845900947625663, + "grad_norm": 4.14678430557251, + "learning_rate": 0.00014587194829033666, + "loss": 0.4337, + "step": 14865 + }, + { + "epoch": 0.3484824510510616, + "grad_norm": 2.818342685699463, + "learning_rate": 0.0001458654027369943, + "loss": 0.331, + "step": 14866 + }, + { + "epoch": 0.3485058926258666, + "grad_norm": 1.5382564067840576, + "learning_rate": 0.00014585885693478497, + "loss": 0.1169, + "step": 14867 + }, + { + "epoch": 0.3485293342006716, + "grad_norm": 1.1157896518707275, + "learning_rate": 0.00014585231088374418, + "loss": 0.1114, + "step": 14868 + }, + { + "epoch": 0.3485527757754766, + "grad_norm": 0.40160301327705383, + "learning_rate": 0.0001458457645839075, + "loss": 0.0331, + "step": 14869 + }, + { + "epoch": 0.3485762173502816, + "grad_norm": 2.2130064964294434, + "learning_rate": 0.00014583921803531036, + "loss": 1.3676, + "step": 14870 + }, + { + "epoch": 0.3485996589250866, + "grad_norm": 3.150501012802124, + "learning_rate": 0.00014583267123798835, + "loss": 0.3691, + "step": 14871 + }, + { + "epoch": 0.3486231004998916, + "grad_norm": 6.480461597442627, + "learning_rate": 0.00014582612419197698, + "loss": 0.499, + "step": 14872 + }, + { + "epoch": 0.3486465420746966, + "grad_norm": 4.330618381500244, + "learning_rate": 0.0001458195768973118, + "loss": 0.4475, + "step": 14873 + }, + { + "epoch": 0.3486699836495016, + "grad_norm": 4.844031810760498, + "learning_rate": 0.00014581302935402828, + "loss": 0.4386, + "step": 14874 + }, + { + "epoch": 0.3486934252243066, + "grad_norm": 4.809200286865234, + "learning_rate": 0.00014580648156216198, + "loss": 0.3631, + "step": 14875 + }, + { + "epoch": 0.34871686679911157, + "grad_norm": 5.053861141204834, + "learning_rate": 0.0001457999335217484, + "loss": 0.7492, + "step": 14876 + }, + { + "epoch": 0.34874030837391656, + "grad_norm": 8.04229736328125, + "learning_rate": 0.00014579338523282314, + "loss": 0.4077, + "step": 14877 + }, + { + "epoch": 0.34876374994872156, + "grad_norm": 3.694524049758911, + "learning_rate": 0.00014578683669542165, + "loss": 0.345, + "step": 14878 + }, + { + "epoch": 0.34878719152352655, + "grad_norm": 0.7425540685653687, + "learning_rate": 0.00014578028790957948, + "loss": 0.0851, + "step": 14879 + }, + { + "epoch": 0.34881063309833155, + "grad_norm": 3.646909236907959, + "learning_rate": 0.00014577373887533223, + "loss": 0.3539, + "step": 14880 + }, + { + "epoch": 0.34883407467313654, + "grad_norm": 3.4116830825805664, + "learning_rate": 0.00014576718959271535, + "loss": 0.4071, + "step": 14881 + }, + { + "epoch": 0.34885751624794153, + "grad_norm": 4.5151848793029785, + "learning_rate": 0.00014576064006176443, + "loss": 0.5187, + "step": 14882 + }, + { + "epoch": 0.34888095782274653, + "grad_norm": 4.889895439147949, + "learning_rate": 0.000145754090282515, + "loss": 0.4277, + "step": 14883 + }, + { + "epoch": 0.3489043993975515, + "grad_norm": 4.08361291885376, + "learning_rate": 0.00014574754025500254, + "loss": 0.3491, + "step": 14884 + }, + { + "epoch": 0.3489278409723565, + "grad_norm": 4.325130462646484, + "learning_rate": 0.00014574098997926268, + "loss": 0.5112, + "step": 14885 + }, + { + "epoch": 0.3489512825471615, + "grad_norm": 0.647213876247406, + "learning_rate": 0.00014573443945533094, + "loss": 0.0471, + "step": 14886 + }, + { + "epoch": 0.3489747241219665, + "grad_norm": 4.4251227378845215, + "learning_rate": 0.0001457278886832428, + "loss": 0.4451, + "step": 14887 + }, + { + "epoch": 0.3489981656967715, + "grad_norm": 4.34033727645874, + "learning_rate": 0.0001457213376630339, + "loss": 0.3863, + "step": 14888 + }, + { + "epoch": 0.3490216072715765, + "grad_norm": 0.9828884601593018, + "learning_rate": 0.00014571478639473969, + "loss": 0.1233, + "step": 14889 + }, + { + "epoch": 0.3490450488463815, + "grad_norm": 2.569336175918579, + "learning_rate": 0.0001457082348783958, + "loss": 1.0646, + "step": 14890 + }, + { + "epoch": 0.3490684904211865, + "grad_norm": 3.0135624408721924, + "learning_rate": 0.0001457016831140377, + "loss": 0.2458, + "step": 14891 + }, + { + "epoch": 0.3490919319959915, + "grad_norm": 2.8087158203125, + "learning_rate": 0.00014569513110170102, + "loss": 0.2899, + "step": 14892 + }, + { + "epoch": 0.3491153735707965, + "grad_norm": 3.0305697917938232, + "learning_rate": 0.0001456885788414213, + "loss": 0.2795, + "step": 14893 + }, + { + "epoch": 0.34913881514560147, + "grad_norm": 1.5464415550231934, + "learning_rate": 0.00014568202633323397, + "loss": 0.1744, + "step": 14894 + }, + { + "epoch": 0.34916225672040646, + "grad_norm": 1.3272242546081543, + "learning_rate": 0.00014567547357717474, + "loss": 0.2, + "step": 14895 + }, + { + "epoch": 0.34918569829521146, + "grad_norm": 3.02436900138855, + "learning_rate": 0.00014566892057327912, + "loss": 0.319, + "step": 14896 + }, + { + "epoch": 0.34920913987001645, + "grad_norm": 2.7181501388549805, + "learning_rate": 0.00014566236732158262, + "loss": 0.17, + "step": 14897 + }, + { + "epoch": 0.34923258144482144, + "grad_norm": 4.261940002441406, + "learning_rate": 0.00014565581382212085, + "loss": 0.4558, + "step": 14898 + }, + { + "epoch": 0.34925602301962644, + "grad_norm": 3.4235892295837402, + "learning_rate": 0.0001456492600749294, + "loss": 0.3507, + "step": 14899 + }, + { + "epoch": 0.34927946459443143, + "grad_norm": 3.054129123687744, + "learning_rate": 0.0001456427060800437, + "loss": 0.2481, + "step": 14900 + }, + { + "epoch": 0.3493029061692364, + "grad_norm": 2.3262240886688232, + "learning_rate": 0.0001456361518374994, + "loss": 0.1586, + "step": 14901 + }, + { + "epoch": 0.3493263477440414, + "grad_norm": 2.4088709354400635, + "learning_rate": 0.0001456295973473321, + "loss": 0.2399, + "step": 14902 + }, + { + "epoch": 0.3493497893188464, + "grad_norm": 4.435571193695068, + "learning_rate": 0.0001456230426095773, + "loss": 0.3845, + "step": 14903 + }, + { + "epoch": 0.3493732308936514, + "grad_norm": 2.9641857147216797, + "learning_rate": 0.00014561648762427058, + "loss": 0.2526, + "step": 14904 + }, + { + "epoch": 0.3493966724684564, + "grad_norm": 1.5112429857254028, + "learning_rate": 0.00014560993239144755, + "loss": 0.2472, + "step": 14905 + }, + { + "epoch": 0.34942011404326145, + "grad_norm": 4.261608123779297, + "learning_rate": 0.0001456033769111437, + "loss": 0.5171, + "step": 14906 + }, + { + "epoch": 0.34944355561806645, + "grad_norm": 3.721879005432129, + "learning_rate": 0.00014559682118339467, + "loss": 0.3657, + "step": 14907 + }, + { + "epoch": 0.34946699719287144, + "grad_norm": 2.6476736068725586, + "learning_rate": 0.000145590265208236, + "loss": 1.6393, + "step": 14908 + }, + { + "epoch": 0.34949043876767644, + "grad_norm": 2.7664742469787598, + "learning_rate": 0.00014558370898570328, + "loss": 0.1892, + "step": 14909 + }, + { + "epoch": 0.34951388034248143, + "grad_norm": 5.4367804527282715, + "learning_rate": 0.00014557715251583205, + "loss": 0.5621, + "step": 14910 + }, + { + "epoch": 0.3495373219172864, + "grad_norm": 3.477201223373413, + "learning_rate": 0.00014557059579865795, + "loss": 0.3743, + "step": 14911 + }, + { + "epoch": 0.3495607634920914, + "grad_norm": 2.6825644969940186, + "learning_rate": 0.0001455640388342165, + "loss": 0.3123, + "step": 14912 + }, + { + "epoch": 0.3495842050668964, + "grad_norm": 4.470137596130371, + "learning_rate": 0.00014555748162254324, + "loss": 0.3203, + "step": 14913 + }, + { + "epoch": 0.3496076466417014, + "grad_norm": 2.781681776046753, + "learning_rate": 0.00014555092416367385, + "loss": 0.1932, + "step": 14914 + }, + { + "epoch": 0.3496310882165064, + "grad_norm": 4.8786492347717285, + "learning_rate": 0.00014554436645764385, + "loss": 0.6435, + "step": 14915 + }, + { + "epoch": 0.3496545297913114, + "grad_norm": 1.5467811822891235, + "learning_rate": 0.00014553780850448886, + "loss": 0.1661, + "step": 14916 + }, + { + "epoch": 0.3496779713661164, + "grad_norm": 3.1338248252868652, + "learning_rate": 0.00014553125030424444, + "loss": 0.3677, + "step": 14917 + }, + { + "epoch": 0.3497014129409214, + "grad_norm": 1.4107279777526855, + "learning_rate": 0.00014552469185694618, + "loss": 0.1737, + "step": 14918 + }, + { + "epoch": 0.3497248545157264, + "grad_norm": 2.9705419540405273, + "learning_rate": 0.0001455181331626296, + "loss": 0.3492, + "step": 14919 + }, + { + "epoch": 0.3497482960905314, + "grad_norm": 1.5040156841278076, + "learning_rate": 0.00014551157422133042, + "loss": 0.1121, + "step": 14920 + }, + { + "epoch": 0.34977173766533637, + "grad_norm": 3.510082960128784, + "learning_rate": 0.00014550501503308414, + "loss": 0.4221, + "step": 14921 + }, + { + "epoch": 0.34979517924014136, + "grad_norm": 2.0909433364868164, + "learning_rate": 0.00014549845559792634, + "loss": 0.3141, + "step": 14922 + }, + { + "epoch": 0.34981862081494636, + "grad_norm": 4.553923606872559, + "learning_rate": 0.00014549189591589267, + "loss": 0.4618, + "step": 14923 + }, + { + "epoch": 0.34984206238975135, + "grad_norm": 4.240184307098389, + "learning_rate": 0.0001454853359870187, + "loss": 0.3175, + "step": 14924 + }, + { + "epoch": 0.34986550396455635, + "grad_norm": 3.4089853763580322, + "learning_rate": 0.00014547877581134, + "loss": 0.574, + "step": 14925 + }, + { + "epoch": 0.34988894553936134, + "grad_norm": 4.596423625946045, + "learning_rate": 0.00014547221538889217, + "loss": 0.47, + "step": 14926 + }, + { + "epoch": 0.34991238711416633, + "grad_norm": 1.5459526777267456, + "learning_rate": 0.00014546565471971085, + "loss": 0.1336, + "step": 14927 + }, + { + "epoch": 0.34993582868897133, + "grad_norm": 4.223186492919922, + "learning_rate": 0.00014545909380383164, + "loss": 0.5387, + "step": 14928 + }, + { + "epoch": 0.3499592702637763, + "grad_norm": 3.305961847305298, + "learning_rate": 0.00014545253264129003, + "loss": 1.801, + "step": 14929 + }, + { + "epoch": 0.3499827118385813, + "grad_norm": 3.6439709663391113, + "learning_rate": 0.00014544597123212176, + "loss": 0.3155, + "step": 14930 + }, + { + "epoch": 0.3500061534133863, + "grad_norm": 6.365681171417236, + "learning_rate": 0.00014543940957636233, + "loss": 0.8688, + "step": 14931 + }, + { + "epoch": 0.3500295949881913, + "grad_norm": 3.9871041774749756, + "learning_rate": 0.0001454328476740474, + "loss": 0.5051, + "step": 14932 + }, + { + "epoch": 0.3500530365629963, + "grad_norm": 1.389183521270752, + "learning_rate": 0.0001454262855252126, + "loss": 0.1116, + "step": 14933 + }, + { + "epoch": 0.3500764781378013, + "grad_norm": 3.236980438232422, + "learning_rate": 0.00014541972312989346, + "loss": 0.2286, + "step": 14934 + }, + { + "epoch": 0.3500999197126063, + "grad_norm": 3.9795796871185303, + "learning_rate": 0.0001454131604881256, + "loss": 0.6329, + "step": 14935 + }, + { + "epoch": 0.3501233612874113, + "grad_norm": 2.778207540512085, + "learning_rate": 0.0001454065975999447, + "loss": 0.2765, + "step": 14936 + }, + { + "epoch": 0.3501468028622163, + "grad_norm": 2.0687453746795654, + "learning_rate": 0.0001454000344653863, + "loss": 1.0359, + "step": 14937 + }, + { + "epoch": 0.35017024443702127, + "grad_norm": 4.242429733276367, + "learning_rate": 0.00014539347108448604, + "loss": 0.3455, + "step": 14938 + }, + { + "epoch": 0.35019368601182627, + "grad_norm": 3.018033504486084, + "learning_rate": 0.00014538690745727955, + "loss": 0.2602, + "step": 14939 + }, + { + "epoch": 0.35021712758663126, + "grad_norm": 1.5028600692749023, + "learning_rate": 0.0001453803435838024, + "loss": 0.896, + "step": 14940 + }, + { + "epoch": 0.35024056916143625, + "grad_norm": 2.367486000061035, + "learning_rate": 0.00014537377946409024, + "loss": 0.3183, + "step": 14941 + }, + { + "epoch": 0.35026401073624125, + "grad_norm": 0.4024040997028351, + "learning_rate": 0.00014536721509817867, + "loss": 0.0401, + "step": 14942 + }, + { + "epoch": 0.35028745231104624, + "grad_norm": 0.8594180345535278, + "learning_rate": 0.00014536065048610333, + "loss": 0.092, + "step": 14943 + }, + { + "epoch": 0.35031089388585124, + "grad_norm": 2.7300145626068115, + "learning_rate": 0.00014535408562789984, + "loss": 0.34, + "step": 14944 + }, + { + "epoch": 0.35033433546065623, + "grad_norm": 3.5617494583129883, + "learning_rate": 0.00014534752052360373, + "loss": 0.1882, + "step": 14945 + }, + { + "epoch": 0.3503577770354612, + "grad_norm": 1.475466251373291, + "learning_rate": 0.00014534095517325076, + "loss": 0.1978, + "step": 14946 + }, + { + "epoch": 0.3503812186102662, + "grad_norm": 3.10374116897583, + "learning_rate": 0.0001453343895768765, + "loss": 2.0463, + "step": 14947 + }, + { + "epoch": 0.3504046601850712, + "grad_norm": 3.3688900470733643, + "learning_rate": 0.00014532782373451655, + "loss": 0.3552, + "step": 14948 + }, + { + "epoch": 0.3504281017598762, + "grad_norm": 1.9820199012756348, + "learning_rate": 0.00014532125764620655, + "loss": 0.1461, + "step": 14949 + }, + { + "epoch": 0.3504515433346812, + "grad_norm": 1.8870446681976318, + "learning_rate": 0.00014531469131198214, + "loss": 0.2418, + "step": 14950 + }, + { + "epoch": 0.3504749849094862, + "grad_norm": 3.246018171310425, + "learning_rate": 0.0001453081247318789, + "loss": 0.29, + "step": 14951 + }, + { + "epoch": 0.3504984264842912, + "grad_norm": 1.1345491409301758, + "learning_rate": 0.00014530155790593255, + "loss": 0.1491, + "step": 14952 + }, + { + "epoch": 0.3505218680590962, + "grad_norm": 2.556446075439453, + "learning_rate": 0.0001452949908341787, + "loss": 0.2433, + "step": 14953 + }, + { + "epoch": 0.3505453096339012, + "grad_norm": 3.2657318115234375, + "learning_rate": 0.0001452884235166529, + "loss": 0.4437, + "step": 14954 + }, + { + "epoch": 0.3505687512087062, + "grad_norm": 4.2450971603393555, + "learning_rate": 0.00014528185595339083, + "loss": 0.5702, + "step": 14955 + }, + { + "epoch": 0.35059219278351117, + "grad_norm": 1.337766170501709, + "learning_rate": 0.00014527528814442818, + "loss": 0.1523, + "step": 14956 + }, + { + "epoch": 0.3506156343583162, + "grad_norm": 2.9261553287506104, + "learning_rate": 0.00014526872008980052, + "loss": 0.4018, + "step": 14957 + }, + { + "epoch": 0.3506390759331212, + "grad_norm": 1.7350139617919922, + "learning_rate": 0.00014526215178954353, + "loss": 0.3213, + "step": 14958 + }, + { + "epoch": 0.3506625175079262, + "grad_norm": 1.2299199104309082, + "learning_rate": 0.00014525558324369285, + "loss": 0.09, + "step": 14959 + }, + { + "epoch": 0.3506859590827312, + "grad_norm": 3.0848793983459473, + "learning_rate": 0.00014524901445228406, + "loss": 0.4405, + "step": 14960 + }, + { + "epoch": 0.3507094006575362, + "grad_norm": 2.3327016830444336, + "learning_rate": 0.00014524244541535284, + "loss": 0.3224, + "step": 14961 + }, + { + "epoch": 0.3507328422323412, + "grad_norm": 2.858682870864868, + "learning_rate": 0.00014523587613293488, + "loss": 0.2972, + "step": 14962 + }, + { + "epoch": 0.3507562838071462, + "grad_norm": 3.630873441696167, + "learning_rate": 0.00014522930660506575, + "loss": 0.5152, + "step": 14963 + }, + { + "epoch": 0.3507797253819512, + "grad_norm": 4.281020164489746, + "learning_rate": 0.00014522273683178114, + "loss": 0.4694, + "step": 14964 + }, + { + "epoch": 0.3508031669567562, + "grad_norm": 1.650346040725708, + "learning_rate": 0.0001452161668131167, + "loss": 0.1823, + "step": 14965 + }, + { + "epoch": 0.35082660853156117, + "grad_norm": 2.508711576461792, + "learning_rate": 0.00014520959654910806, + "loss": 0.264, + "step": 14966 + }, + { + "epoch": 0.35085005010636616, + "grad_norm": 3.11549711227417, + "learning_rate": 0.00014520302603979085, + "loss": 0.3722, + "step": 14967 + }, + { + "epoch": 0.35087349168117116, + "grad_norm": 1.104082465171814, + "learning_rate": 0.00014519645528520079, + "loss": 0.1019, + "step": 14968 + }, + { + "epoch": 0.35089693325597615, + "grad_norm": 3.4388444423675537, + "learning_rate": 0.0001451898842853735, + "loss": 0.4861, + "step": 14969 + }, + { + "epoch": 0.35092037483078115, + "grad_norm": 1.8320772647857666, + "learning_rate": 0.00014518331304034458, + "loss": 0.1513, + "step": 14970 + }, + { + "epoch": 0.35094381640558614, + "grad_norm": 0.7031598687171936, + "learning_rate": 0.00014517674155014977, + "loss": 0.1086, + "step": 14971 + }, + { + "epoch": 0.35096725798039113, + "grad_norm": 0.7320950627326965, + "learning_rate": 0.00014517016981482469, + "loss": 0.1222, + "step": 14972 + }, + { + "epoch": 0.35099069955519613, + "grad_norm": 2.7360644340515137, + "learning_rate": 0.000145163597834405, + "loss": 0.2761, + "step": 14973 + }, + { + "epoch": 0.3510141411300011, + "grad_norm": 2.6851513385772705, + "learning_rate": 0.00014515702560892634, + "loss": 1.5386, + "step": 14974 + }, + { + "epoch": 0.3510375827048061, + "grad_norm": 4.975789546966553, + "learning_rate": 0.00014515045313842438, + "loss": 0.61, + "step": 14975 + }, + { + "epoch": 0.3510610242796111, + "grad_norm": 4.12101411819458, + "learning_rate": 0.0001451438804229348, + "loss": 0.5581, + "step": 14976 + }, + { + "epoch": 0.3510844658544161, + "grad_norm": 3.082432270050049, + "learning_rate": 0.00014513730746249328, + "loss": 0.3909, + "step": 14977 + }, + { + "epoch": 0.3511079074292211, + "grad_norm": 2.708998203277588, + "learning_rate": 0.00014513073425713544, + "loss": 0.4325, + "step": 14978 + }, + { + "epoch": 0.3511313490040261, + "grad_norm": 1.6103301048278809, + "learning_rate": 0.00014512416080689696, + "loss": 0.2393, + "step": 14979 + }, + { + "epoch": 0.3511547905788311, + "grad_norm": 2.2835350036621094, + "learning_rate": 0.0001451175871118135, + "loss": 0.2719, + "step": 14980 + }, + { + "epoch": 0.3511782321536361, + "grad_norm": 3.3545355796813965, + "learning_rate": 0.0001451110131719208, + "loss": 1.4401, + "step": 14981 + }, + { + "epoch": 0.3512016737284411, + "grad_norm": 3.1990435123443604, + "learning_rate": 0.00014510443898725446, + "loss": 0.3361, + "step": 14982 + }, + { + "epoch": 0.35122511530324607, + "grad_norm": 4.7132487297058105, + "learning_rate": 0.00014509786455785013, + "loss": 0.3887, + "step": 14983 + }, + { + "epoch": 0.35124855687805107, + "grad_norm": 1.5249415636062622, + "learning_rate": 0.00014509128988374355, + "loss": 0.1864, + "step": 14984 + }, + { + "epoch": 0.35127199845285606, + "grad_norm": 3.5163347721099854, + "learning_rate": 0.00014508471496497036, + "loss": 0.4903, + "step": 14985 + }, + { + "epoch": 0.35129544002766105, + "grad_norm": 1.629551887512207, + "learning_rate": 0.0001450781398015662, + "loss": 0.1855, + "step": 14986 + }, + { + "epoch": 0.35131888160246605, + "grad_norm": 2.727430582046509, + "learning_rate": 0.00014507156439356683, + "loss": 0.2411, + "step": 14987 + }, + { + "epoch": 0.35134232317727104, + "grad_norm": 3.29524827003479, + "learning_rate": 0.00014506498874100786, + "loss": 0.3749, + "step": 14988 + }, + { + "epoch": 0.35136576475207604, + "grad_norm": 5.712284088134766, + "learning_rate": 0.000145058412843925, + "loss": 0.4659, + "step": 14989 + }, + { + "epoch": 0.35138920632688103, + "grad_norm": 4.570375919342041, + "learning_rate": 0.00014505183670235393, + "loss": 0.3347, + "step": 14990 + }, + { + "epoch": 0.351412647901686, + "grad_norm": 0.8854816555976868, + "learning_rate": 0.00014504526031633037, + "loss": 0.0638, + "step": 14991 + }, + { + "epoch": 0.351436089476491, + "grad_norm": 4.620330333709717, + "learning_rate": 0.0001450386836858899, + "loss": 0.4912, + "step": 14992 + }, + { + "epoch": 0.351459531051296, + "grad_norm": 6.67280387878418, + "learning_rate": 0.00014503210681106828, + "loss": 0.631, + "step": 14993 + }, + { + "epoch": 0.351482972626101, + "grad_norm": 4.731287479400635, + "learning_rate": 0.00014502552969190117, + "loss": 0.4573, + "step": 14994 + }, + { + "epoch": 0.351506414200906, + "grad_norm": 3.4146809577941895, + "learning_rate": 0.00014501895232842428, + "loss": 0.541, + "step": 14995 + }, + { + "epoch": 0.351529855775711, + "grad_norm": 4.092784404754639, + "learning_rate": 0.0001450123747206733, + "loss": 0.5618, + "step": 14996 + }, + { + "epoch": 0.351553297350516, + "grad_norm": 1.3712153434753418, + "learning_rate": 0.0001450057968686839, + "loss": 0.1979, + "step": 14997 + }, + { + "epoch": 0.351576738925321, + "grad_norm": 4.932800769805908, + "learning_rate": 0.00014499921877249178, + "loss": 0.7229, + "step": 14998 + }, + { + "epoch": 0.351600180500126, + "grad_norm": 3.3439226150512695, + "learning_rate": 0.00014499264043213263, + "loss": 0.3293, + "step": 14999 + }, + { + "epoch": 0.351623622074931, + "grad_norm": 3.1872403621673584, + "learning_rate": 0.00014498606184764219, + "loss": 0.3708, + "step": 15000 + }, + { + "epoch": 0.35164706364973597, + "grad_norm": 3.9913315773010254, + "learning_rate": 0.0001449794830190561, + "loss": 0.611, + "step": 15001 + }, + { + "epoch": 0.35167050522454096, + "grad_norm": 4.107431888580322, + "learning_rate": 0.00014497290394641001, + "loss": 0.3386, + "step": 15002 + }, + { + "epoch": 0.35169394679934596, + "grad_norm": 1.273182988166809, + "learning_rate": 0.00014496632462973974, + "loss": 0.1574, + "step": 15003 + }, + { + "epoch": 0.35171738837415095, + "grad_norm": 2.912263870239258, + "learning_rate": 0.0001449597450690809, + "loss": 1.6402, + "step": 15004 + }, + { + "epoch": 0.35174082994895595, + "grad_norm": 0.8347825407981873, + "learning_rate": 0.0001449531652644692, + "loss": 0.1012, + "step": 15005 + }, + { + "epoch": 0.35176427152376094, + "grad_norm": 3.8795464038848877, + "learning_rate": 0.0001449465852159404, + "loss": 0.6275, + "step": 15006 + }, + { + "epoch": 0.35178771309856594, + "grad_norm": 4.259031772613525, + "learning_rate": 0.00014494000492353018, + "loss": 0.3747, + "step": 15007 + }, + { + "epoch": 0.35181115467337093, + "grad_norm": 3.12038254737854, + "learning_rate": 0.00014493342438727419, + "loss": 0.3021, + "step": 15008 + }, + { + "epoch": 0.351834596248176, + "grad_norm": 2.4929256439208984, + "learning_rate": 0.00014492684360720818, + "loss": 0.2145, + "step": 15009 + }, + { + "epoch": 0.351858037822981, + "grad_norm": 3.724111318588257, + "learning_rate": 0.00014492026258336785, + "loss": 0.4346, + "step": 15010 + }, + { + "epoch": 0.35188147939778597, + "grad_norm": 1.199950098991394, + "learning_rate": 0.00014491368131578893, + "loss": 0.1031, + "step": 15011 + }, + { + "epoch": 0.35190492097259096, + "grad_norm": 2.9839158058166504, + "learning_rate": 0.0001449070998045071, + "loss": 0.2844, + "step": 15012 + }, + { + "epoch": 0.35192836254739596, + "grad_norm": 4.044597625732422, + "learning_rate": 0.00014490051804955808, + "loss": 0.2631, + "step": 15013 + }, + { + "epoch": 0.35195180412220095, + "grad_norm": 4.689559459686279, + "learning_rate": 0.0001448939360509776, + "loss": 0.6403, + "step": 15014 + }, + { + "epoch": 0.35197524569700595, + "grad_norm": 0.5909934639930725, + "learning_rate": 0.00014488735380880136, + "loss": 0.0766, + "step": 15015 + }, + { + "epoch": 0.35199868727181094, + "grad_norm": 3.524991989135742, + "learning_rate": 0.00014488077132306508, + "loss": 0.5362, + "step": 15016 + }, + { + "epoch": 0.35202212884661593, + "grad_norm": 1.3130505084991455, + "learning_rate": 0.00014487418859380447, + "loss": 0.1753, + "step": 15017 + }, + { + "epoch": 0.35204557042142093, + "grad_norm": 1.118957757949829, + "learning_rate": 0.00014486760562105523, + "loss": 0.1636, + "step": 15018 + }, + { + "epoch": 0.3520690119962259, + "grad_norm": 0.5804309248924255, + "learning_rate": 0.00014486102240485313, + "loss": 0.0747, + "step": 15019 + }, + { + "epoch": 0.3520924535710309, + "grad_norm": 1.233620047569275, + "learning_rate": 0.00014485443894523384, + "loss": 0.1103, + "step": 15020 + }, + { + "epoch": 0.3521158951458359, + "grad_norm": 4.5313849449157715, + "learning_rate": 0.00014484785524223313, + "loss": 0.4671, + "step": 15021 + }, + { + "epoch": 0.3521393367206409, + "grad_norm": 4.070285320281982, + "learning_rate": 0.00014484127129588665, + "loss": 0.4356, + "step": 15022 + }, + { + "epoch": 0.3521627782954459, + "grad_norm": 2.536320924758911, + "learning_rate": 0.00014483468710623022, + "loss": 0.1619, + "step": 15023 + }, + { + "epoch": 0.3521862198702509, + "grad_norm": 1.7115967273712158, + "learning_rate": 0.00014482810267329947, + "loss": 0.1454, + "step": 15024 + }, + { + "epoch": 0.3522096614450559, + "grad_norm": 3.2706105709075928, + "learning_rate": 0.00014482151799713022, + "loss": 0.4503, + "step": 15025 + }, + { + "epoch": 0.3522331030198609, + "grad_norm": 3.871178388595581, + "learning_rate": 0.00014481493307775812, + "loss": 0.4684, + "step": 15026 + }, + { + "epoch": 0.3522565445946659, + "grad_norm": 1.434229850769043, + "learning_rate": 0.00014480834791521895, + "loss": 0.158, + "step": 15027 + }, + { + "epoch": 0.35227998616947087, + "grad_norm": 4.332493305206299, + "learning_rate": 0.00014480176250954843, + "loss": 0.3916, + "step": 15028 + }, + { + "epoch": 0.35230342774427587, + "grad_norm": 3.8267111778259277, + "learning_rate": 0.00014479517686078228, + "loss": 0.4859, + "step": 15029 + }, + { + "epoch": 0.35232686931908086, + "grad_norm": 2.043126344680786, + "learning_rate": 0.00014478859096895622, + "loss": 0.1063, + "step": 15030 + }, + { + "epoch": 0.35235031089388585, + "grad_norm": 2.9690921306610107, + "learning_rate": 0.00014478200483410603, + "loss": 0.5, + "step": 15031 + }, + { + "epoch": 0.35237375246869085, + "grad_norm": 3.1710853576660156, + "learning_rate": 0.00014477541845626742, + "loss": 0.2174, + "step": 15032 + }, + { + "epoch": 0.35239719404349584, + "grad_norm": 0.2836735248565674, + "learning_rate": 0.0001447688318354761, + "loss": 0.0206, + "step": 15033 + }, + { + "epoch": 0.35242063561830084, + "grad_norm": 2.214062213897705, + "learning_rate": 0.00014476224497176788, + "loss": 0.1709, + "step": 15034 + }, + { + "epoch": 0.35244407719310583, + "grad_norm": 2.9319655895233154, + "learning_rate": 0.00014475565786517845, + "loss": 0.2887, + "step": 15035 + }, + { + "epoch": 0.3524675187679108, + "grad_norm": 4.400402545928955, + "learning_rate": 0.0001447490705157436, + "loss": 0.6588, + "step": 15036 + }, + { + "epoch": 0.3524909603427158, + "grad_norm": 3.4800937175750732, + "learning_rate": 0.00014474248292349896, + "loss": 0.5619, + "step": 15037 + }, + { + "epoch": 0.3525144019175208, + "grad_norm": 4.241909027099609, + "learning_rate": 0.0001447358950884804, + "loss": 1.9133, + "step": 15038 + }, + { + "epoch": 0.3525378434923258, + "grad_norm": 3.27769136428833, + "learning_rate": 0.0001447293070107236, + "loss": 1.9839, + "step": 15039 + }, + { + "epoch": 0.3525612850671308, + "grad_norm": 3.5341928005218506, + "learning_rate": 0.0001447227186902643, + "loss": 0.3392, + "step": 15040 + }, + { + "epoch": 0.3525847266419358, + "grad_norm": 2.9600963592529297, + "learning_rate": 0.0001447161301271383, + "loss": 0.4886, + "step": 15041 + }, + { + "epoch": 0.3526081682167408, + "grad_norm": 2.8856518268585205, + "learning_rate": 0.00014470954132138133, + "loss": 0.3562, + "step": 15042 + }, + { + "epoch": 0.3526316097915458, + "grad_norm": 3.308610439300537, + "learning_rate": 0.0001447029522730291, + "loss": 1.4306, + "step": 15043 + }, + { + "epoch": 0.3526550513663508, + "grad_norm": 7.769721984863281, + "learning_rate": 0.0001446963629821174, + "loss": 0.3853, + "step": 15044 + }, + { + "epoch": 0.3526784929411558, + "grad_norm": 0.9556517004966736, + "learning_rate": 0.000144689773448682, + "loss": 0.1299, + "step": 15045 + }, + { + "epoch": 0.35270193451596077, + "grad_norm": 1.592751383781433, + "learning_rate": 0.0001446831836727586, + "loss": 0.1972, + "step": 15046 + }, + { + "epoch": 0.35272537609076576, + "grad_norm": 1.230880618095398, + "learning_rate": 0.00014467659365438303, + "loss": 0.1168, + "step": 15047 + }, + { + "epoch": 0.35274881766557076, + "grad_norm": 4.136990547180176, + "learning_rate": 0.00014467000339359097, + "loss": 0.577, + "step": 15048 + }, + { + "epoch": 0.35277225924037575, + "grad_norm": 3.473931074142456, + "learning_rate": 0.00014466341289041826, + "loss": 0.2827, + "step": 15049 + }, + { + "epoch": 0.35279570081518075, + "grad_norm": 5.158565044403076, + "learning_rate": 0.00014465682214490059, + "loss": 1.0247, + "step": 15050 + }, + { + "epoch": 0.35281914238998574, + "grad_norm": 2.8404293060302734, + "learning_rate": 0.00014465023115707373, + "loss": 0.5649, + "step": 15051 + }, + { + "epoch": 0.35284258396479073, + "grad_norm": 1.3740133047103882, + "learning_rate": 0.00014464363992697346, + "loss": 0.2061, + "step": 15052 + }, + { + "epoch": 0.35286602553959573, + "grad_norm": 0.8664697408676147, + "learning_rate": 0.00014463704845463554, + "loss": 0.078, + "step": 15053 + }, + { + "epoch": 0.3528894671144007, + "grad_norm": 1.0069411993026733, + "learning_rate": 0.00014463045674009578, + "loss": 0.1414, + "step": 15054 + }, + { + "epoch": 0.3529129086892057, + "grad_norm": 2.5241799354553223, + "learning_rate": 0.0001446238647833899, + "loss": 0.305, + "step": 15055 + }, + { + "epoch": 0.3529363502640107, + "grad_norm": 1.211768627166748, + "learning_rate": 0.00014461727258455363, + "loss": 0.1428, + "step": 15056 + }, + { + "epoch": 0.3529597918388157, + "grad_norm": 0.7502564787864685, + "learning_rate": 0.00014461068014362286, + "loss": 0.1115, + "step": 15057 + }, + { + "epoch": 0.3529832334136207, + "grad_norm": 1.9917023181915283, + "learning_rate": 0.00014460408746063322, + "loss": 0.0984, + "step": 15058 + }, + { + "epoch": 0.3530066749884257, + "grad_norm": 3.2902536392211914, + "learning_rate": 0.00014459749453562054, + "loss": 0.3411, + "step": 15059 + }, + { + "epoch": 0.35303011656323074, + "grad_norm": 1.2176499366760254, + "learning_rate": 0.00014459090136862064, + "loss": 0.2298, + "step": 15060 + }, + { + "epoch": 0.35305355813803574, + "grad_norm": 3.9293174743652344, + "learning_rate": 0.00014458430795966925, + "loss": 0.3912, + "step": 15061 + }, + { + "epoch": 0.35307699971284073, + "grad_norm": 2.487966537475586, + "learning_rate": 0.00014457771430880214, + "loss": 0.2737, + "step": 15062 + }, + { + "epoch": 0.3531004412876457, + "grad_norm": 2.0768802165985107, + "learning_rate": 0.0001445711204160551, + "loss": 0.1673, + "step": 15063 + }, + { + "epoch": 0.3531238828624507, + "grad_norm": 2.7041540145874023, + "learning_rate": 0.00014456452628146393, + "loss": 0.2382, + "step": 15064 + }, + { + "epoch": 0.3531473244372557, + "grad_norm": 2.9289021492004395, + "learning_rate": 0.00014455793190506437, + "loss": 0.297, + "step": 15065 + }, + { + "epoch": 0.3531707660120607, + "grad_norm": 1.6013799905776978, + "learning_rate": 0.00014455133728689221, + "loss": 0.8851, + "step": 15066 + }, + { + "epoch": 0.3531942075868657, + "grad_norm": 4.189749717712402, + "learning_rate": 0.00014454474242698323, + "loss": 0.2189, + "step": 15067 + }, + { + "epoch": 0.3532176491616707, + "grad_norm": 3.4309167861938477, + "learning_rate": 0.00014453814732537327, + "loss": 1.851, + "step": 15068 + }, + { + "epoch": 0.3532410907364757, + "grad_norm": 4.903350353240967, + "learning_rate": 0.00014453155198209804, + "loss": 0.5355, + "step": 15069 + }, + { + "epoch": 0.3532645323112807, + "grad_norm": 1.5414509773254395, + "learning_rate": 0.00014452495639719336, + "loss": 0.15, + "step": 15070 + }, + { + "epoch": 0.3532879738860857, + "grad_norm": 1.263074278831482, + "learning_rate": 0.00014451836057069502, + "loss": 0.0863, + "step": 15071 + }, + { + "epoch": 0.3533114154608907, + "grad_norm": 3.0052852630615234, + "learning_rate": 0.00014451176450263877, + "loss": 0.2521, + "step": 15072 + }, + { + "epoch": 0.35333485703569567, + "grad_norm": 4.914721488952637, + "learning_rate": 0.00014450516819306049, + "loss": 0.3457, + "step": 15073 + }, + { + "epoch": 0.35335829861050067, + "grad_norm": 3.9803459644317627, + "learning_rate": 0.0001444985716419959, + "loss": 0.3326, + "step": 15074 + }, + { + "epoch": 0.35338174018530566, + "grad_norm": 1.0324455499649048, + "learning_rate": 0.00014449197484948077, + "loss": 0.1358, + "step": 15075 + }, + { + "epoch": 0.35340518176011065, + "grad_norm": 2.439196825027466, + "learning_rate": 0.00014448537781555098, + "loss": 1.9421, + "step": 15076 + }, + { + "epoch": 0.35342862333491565, + "grad_norm": 1.3486028909683228, + "learning_rate": 0.00014447878054024224, + "loss": 0.1618, + "step": 15077 + }, + { + "epoch": 0.35345206490972064, + "grad_norm": 4.088951587677002, + "learning_rate": 0.0001444721830235904, + "loss": 0.3595, + "step": 15078 + }, + { + "epoch": 0.35347550648452564, + "grad_norm": 4.940511703491211, + "learning_rate": 0.00014446558526563127, + "loss": 0.4741, + "step": 15079 + }, + { + "epoch": 0.35349894805933063, + "grad_norm": 3.2835545539855957, + "learning_rate": 0.0001444589872664006, + "loss": 2.0549, + "step": 15080 + }, + { + "epoch": 0.3535223896341356, + "grad_norm": 5.1928839683532715, + "learning_rate": 0.0001444523890259342, + "loss": 0.4795, + "step": 15081 + }, + { + "epoch": 0.3535458312089406, + "grad_norm": 1.8015481233596802, + "learning_rate": 0.00014444579054426795, + "loss": 1.3864, + "step": 15082 + }, + { + "epoch": 0.3535692727837456, + "grad_norm": 5.5037455558776855, + "learning_rate": 0.00014443919182143754, + "loss": 0.7708, + "step": 15083 + }, + { + "epoch": 0.3535927143585506, + "grad_norm": 5.016036033630371, + "learning_rate": 0.00014443259285747882, + "loss": 0.2514, + "step": 15084 + }, + { + "epoch": 0.3536161559333556, + "grad_norm": 2.5518856048583984, + "learning_rate": 0.00014442599365242762, + "loss": 0.1549, + "step": 15085 + }, + { + "epoch": 0.3536395975081606, + "grad_norm": 4.538981914520264, + "learning_rate": 0.00014441939420631973, + "loss": 0.2885, + "step": 15086 + }, + { + "epoch": 0.3536630390829656, + "grad_norm": 1.883566975593567, + "learning_rate": 0.00014441279451919096, + "loss": 0.3103, + "step": 15087 + }, + { + "epoch": 0.3536864806577706, + "grad_norm": 1.9234793186187744, + "learning_rate": 0.0001444061945910771, + "loss": 0.1783, + "step": 15088 + }, + { + "epoch": 0.3537099222325756, + "grad_norm": 1.6273928880691528, + "learning_rate": 0.000144399594422014, + "loss": 0.2017, + "step": 15089 + }, + { + "epoch": 0.3537333638073806, + "grad_norm": 6.078446865081787, + "learning_rate": 0.00014439299401203742, + "loss": 0.5927, + "step": 15090 + }, + { + "epoch": 0.35375680538218557, + "grad_norm": 0.4832656681537628, + "learning_rate": 0.00014438639336118324, + "loss": 0.0202, + "step": 15091 + }, + { + "epoch": 0.35378024695699056, + "grad_norm": 4.91266393661499, + "learning_rate": 0.00014437979246948723, + "loss": 0.5079, + "step": 15092 + }, + { + "epoch": 0.35380368853179556, + "grad_norm": 3.2325844764709473, + "learning_rate": 0.00014437319133698523, + "loss": 0.3573, + "step": 15093 + }, + { + "epoch": 0.35382713010660055, + "grad_norm": 4.23999547958374, + "learning_rate": 0.00014436658996371305, + "loss": 0.4148, + "step": 15094 + }, + { + "epoch": 0.35385057168140555, + "grad_norm": 3.9351584911346436, + "learning_rate": 0.00014435998834970646, + "loss": 0.3179, + "step": 15095 + }, + { + "epoch": 0.35387401325621054, + "grad_norm": 3.215864419937134, + "learning_rate": 0.00014435338649500137, + "loss": 0.5031, + "step": 15096 + }, + { + "epoch": 0.35389745483101553, + "grad_norm": 3.860611915588379, + "learning_rate": 0.00014434678439963355, + "loss": 0.3547, + "step": 15097 + }, + { + "epoch": 0.35392089640582053, + "grad_norm": 3.2853267192840576, + "learning_rate": 0.0001443401820636388, + "loss": 1.7969, + "step": 15098 + }, + { + "epoch": 0.3539443379806255, + "grad_norm": 1.122877836227417, + "learning_rate": 0.000144333579487053, + "loss": 0.0666, + "step": 15099 + }, + { + "epoch": 0.3539677795554305, + "grad_norm": 0.8057582378387451, + "learning_rate": 0.00014432697666991198, + "loss": 0.0838, + "step": 15100 + }, + { + "epoch": 0.3539912211302355, + "grad_norm": 2.2198233604431152, + "learning_rate": 0.0001443203736122515, + "loss": 0.2401, + "step": 15101 + }, + { + "epoch": 0.3540146627050405, + "grad_norm": 1.9575719833374023, + "learning_rate": 0.00014431377031410745, + "loss": 0.064, + "step": 15102 + }, + { + "epoch": 0.3540381042798455, + "grad_norm": 9.513169288635254, + "learning_rate": 0.0001443071667755156, + "loss": 0.6261, + "step": 15103 + }, + { + "epoch": 0.3540615458546505, + "grad_norm": 3.3753812313079834, + "learning_rate": 0.00014430056299651183, + "loss": 1.6307, + "step": 15104 + }, + { + "epoch": 0.3540849874294555, + "grad_norm": 4.2113213539123535, + "learning_rate": 0.000144293958977132, + "loss": 0.65, + "step": 15105 + }, + { + "epoch": 0.3541084290042605, + "grad_norm": 2.299668312072754, + "learning_rate": 0.00014428735471741185, + "loss": 0.1903, + "step": 15106 + }, + { + "epoch": 0.3541318705790655, + "grad_norm": 4.339334964752197, + "learning_rate": 0.00014428075021738728, + "loss": 0.1593, + "step": 15107 + }, + { + "epoch": 0.35415531215387047, + "grad_norm": 1.876550316810608, + "learning_rate": 0.00014427414547709414, + "loss": 0.2191, + "step": 15108 + }, + { + "epoch": 0.35417875372867547, + "grad_norm": 3.926530599594116, + "learning_rate": 0.00014426754049656824, + "loss": 0.563, + "step": 15109 + }, + { + "epoch": 0.35420219530348046, + "grad_norm": 2.0096895694732666, + "learning_rate": 0.0001442609352758454, + "loss": 0.1943, + "step": 15110 + }, + { + "epoch": 0.3542256368782855, + "grad_norm": 5.209651947021484, + "learning_rate": 0.0001442543298149615, + "loss": 0.7877, + "step": 15111 + }, + { + "epoch": 0.3542490784530905, + "grad_norm": 4.395588397979736, + "learning_rate": 0.00014424772411395236, + "loss": 1.8199, + "step": 15112 + }, + { + "epoch": 0.3542725200278955, + "grad_norm": 2.9285831451416016, + "learning_rate": 0.0001442411181728538, + "loss": 0.2689, + "step": 15113 + }, + { + "epoch": 0.3542959616027005, + "grad_norm": 2.8977737426757812, + "learning_rate": 0.0001442345119917017, + "loss": 0.1429, + "step": 15114 + }, + { + "epoch": 0.3543194031775055, + "grad_norm": 1.5845613479614258, + "learning_rate": 0.00014422790557053192, + "loss": 0.1429, + "step": 15115 + }, + { + "epoch": 0.3543428447523105, + "grad_norm": 2.8511338233947754, + "learning_rate": 0.00014422129890938023, + "loss": 0.369, + "step": 15116 + }, + { + "epoch": 0.3543662863271155, + "grad_norm": 5.621095180511475, + "learning_rate": 0.00014421469200828256, + "loss": 0.3541, + "step": 15117 + }, + { + "epoch": 0.35438972790192047, + "grad_norm": 3.263435125350952, + "learning_rate": 0.00014420808486727475, + "loss": 0.2748, + "step": 15118 + }, + { + "epoch": 0.35441316947672546, + "grad_norm": 5.08903694152832, + "learning_rate": 0.0001442014774863926, + "loss": 0.6891, + "step": 15119 + }, + { + "epoch": 0.35443661105153046, + "grad_norm": 0.7827412486076355, + "learning_rate": 0.00014419486986567197, + "loss": 0.0588, + "step": 15120 + }, + { + "epoch": 0.35446005262633545, + "grad_norm": 1.2017287015914917, + "learning_rate": 0.00014418826200514876, + "loss": 0.1284, + "step": 15121 + }, + { + "epoch": 0.35448349420114045, + "grad_norm": 2.6377339363098145, + "learning_rate": 0.00014418165390485882, + "loss": 0.2647, + "step": 15122 + }, + { + "epoch": 0.35450693577594544, + "grad_norm": 4.340078353881836, + "learning_rate": 0.00014417504556483795, + "loss": 0.4518, + "step": 15123 + }, + { + "epoch": 0.35453037735075044, + "grad_norm": 2.319042682647705, + "learning_rate": 0.00014416843698512204, + "loss": 0.2051, + "step": 15124 + }, + { + "epoch": 0.35455381892555543, + "grad_norm": 3.605485439300537, + "learning_rate": 0.00014416182816574696, + "loss": 0.3674, + "step": 15125 + }, + { + "epoch": 0.3545772605003604, + "grad_norm": 1.6309618949890137, + "learning_rate": 0.00014415521910674856, + "loss": 0.1942, + "step": 15126 + }, + { + "epoch": 0.3546007020751654, + "grad_norm": 2.9916021823883057, + "learning_rate": 0.0001441486098081627, + "loss": 0.3372, + "step": 15127 + }, + { + "epoch": 0.3546241436499704, + "grad_norm": 0.5697042942047119, + "learning_rate": 0.00014414200027002526, + "loss": 0.0677, + "step": 15128 + }, + { + "epoch": 0.3546475852247754, + "grad_norm": 2.462627410888672, + "learning_rate": 0.00014413539049237204, + "loss": 0.2494, + "step": 15129 + }, + { + "epoch": 0.3546710267995804, + "grad_norm": 4.746607303619385, + "learning_rate": 0.000144128780475239, + "loss": 0.5807, + "step": 15130 + }, + { + "epoch": 0.3546944683743854, + "grad_norm": 4.117047309875488, + "learning_rate": 0.00014412217021866192, + "loss": 0.5287, + "step": 15131 + }, + { + "epoch": 0.3547179099491904, + "grad_norm": 3.560781717300415, + "learning_rate": 0.0001441155597226767, + "loss": 0.1865, + "step": 15132 + }, + { + "epoch": 0.3547413515239954, + "grad_norm": 1.2243430614471436, + "learning_rate": 0.00014410894898731922, + "loss": 0.1147, + "step": 15133 + }, + { + "epoch": 0.3547647930988004, + "grad_norm": 3.2328011989593506, + "learning_rate": 0.00014410233801262536, + "loss": 0.3956, + "step": 15134 + }, + { + "epoch": 0.3547882346736054, + "grad_norm": 0.5572289824485779, + "learning_rate": 0.00014409572679863097, + "loss": 0.0508, + "step": 15135 + }, + { + "epoch": 0.35481167624841037, + "grad_norm": 4.1776556968688965, + "learning_rate": 0.00014408911534537192, + "loss": 0.2929, + "step": 15136 + }, + { + "epoch": 0.35483511782321536, + "grad_norm": 2.5157310962677, + "learning_rate": 0.00014408250365288412, + "loss": 0.1685, + "step": 15137 + }, + { + "epoch": 0.35485855939802036, + "grad_norm": 4.472541332244873, + "learning_rate": 0.00014407589172120338, + "loss": 0.4036, + "step": 15138 + }, + { + "epoch": 0.35488200097282535, + "grad_norm": 3.2340893745422363, + "learning_rate": 0.0001440692795503656, + "loss": 0.2805, + "step": 15139 + }, + { + "epoch": 0.35490544254763035, + "grad_norm": 3.1501543521881104, + "learning_rate": 0.0001440626671404067, + "loss": 0.2781, + "step": 15140 + }, + { + "epoch": 0.35492888412243534, + "grad_norm": 4.321723461151123, + "learning_rate": 0.00014405605449136253, + "loss": 0.6243, + "step": 15141 + }, + { + "epoch": 0.35495232569724033, + "grad_norm": 2.558532476425171, + "learning_rate": 0.00014404944160326894, + "loss": 0.2723, + "step": 15142 + }, + { + "epoch": 0.35497576727204533, + "grad_norm": 3.116818904876709, + "learning_rate": 0.00014404282847616187, + "loss": 0.3824, + "step": 15143 + }, + { + "epoch": 0.3549992088468503, + "grad_norm": 5.576885223388672, + "learning_rate": 0.00014403621511007713, + "loss": 0.6761, + "step": 15144 + }, + { + "epoch": 0.3550226504216553, + "grad_norm": 3.79709529876709, + "learning_rate": 0.00014402960150505068, + "loss": 0.6159, + "step": 15145 + }, + { + "epoch": 0.3550460919964603, + "grad_norm": 2.504828691482544, + "learning_rate": 0.0001440229876611184, + "loss": 0.1567, + "step": 15146 + }, + { + "epoch": 0.3550695335712653, + "grad_norm": 3.8908450603485107, + "learning_rate": 0.00014401637357831613, + "loss": 0.2937, + "step": 15147 + }, + { + "epoch": 0.3550929751460703, + "grad_norm": 0.4613003432750702, + "learning_rate": 0.00014400975925667975, + "loss": 0.0487, + "step": 15148 + }, + { + "epoch": 0.3551164167208753, + "grad_norm": 3.3627963066101074, + "learning_rate": 0.00014400314469624522, + "loss": 1.7469, + "step": 15149 + }, + { + "epoch": 0.3551398582956803, + "grad_norm": 6.436582088470459, + "learning_rate": 0.00014399652989704838, + "loss": 0.5954, + "step": 15150 + }, + { + "epoch": 0.3551632998704853, + "grad_norm": 1.859237790107727, + "learning_rate": 0.0001439899148591251, + "loss": 0.1858, + "step": 15151 + }, + { + "epoch": 0.3551867414452903, + "grad_norm": 5.921971797943115, + "learning_rate": 0.00014398329958251137, + "loss": 0.6536, + "step": 15152 + }, + { + "epoch": 0.35521018302009527, + "grad_norm": 9.5214204788208, + "learning_rate": 0.000143976684067243, + "loss": 0.3845, + "step": 15153 + }, + { + "epoch": 0.35523362459490027, + "grad_norm": 4.997982501983643, + "learning_rate": 0.00014397006831335583, + "loss": 0.4116, + "step": 15154 + }, + { + "epoch": 0.35525706616970526, + "grad_norm": 3.001732110977173, + "learning_rate": 0.00014396345232088592, + "loss": 0.4513, + "step": 15155 + }, + { + "epoch": 0.35528050774451025, + "grad_norm": 4.358073711395264, + "learning_rate": 0.00014395683608986904, + "loss": 0.5082, + "step": 15156 + }, + { + "epoch": 0.35530394931931525, + "grad_norm": 2.438154458999634, + "learning_rate": 0.00014395021962034118, + "loss": 0.313, + "step": 15157 + }, + { + "epoch": 0.35532739089412024, + "grad_norm": 1.8788161277770996, + "learning_rate": 0.00014394360291233814, + "loss": 0.1074, + "step": 15158 + }, + { + "epoch": 0.35535083246892524, + "grad_norm": 1.7182512283325195, + "learning_rate": 0.0001439369859658959, + "loss": 0.1242, + "step": 15159 + }, + { + "epoch": 0.35537427404373023, + "grad_norm": 2.686469316482544, + "learning_rate": 0.00014393036878105033, + "loss": 0.3567, + "step": 15160 + }, + { + "epoch": 0.3553977156185352, + "grad_norm": 4.347169876098633, + "learning_rate": 0.00014392375135783734, + "loss": 0.6161, + "step": 15161 + }, + { + "epoch": 0.3554211571933403, + "grad_norm": 4.804661750793457, + "learning_rate": 0.00014391713369629283, + "loss": 0.3246, + "step": 15162 + }, + { + "epoch": 0.35544459876814527, + "grad_norm": 0.5620885491371155, + "learning_rate": 0.00014391051579645275, + "loss": 0.0728, + "step": 15163 + }, + { + "epoch": 0.35546804034295026, + "grad_norm": 1.0377055406570435, + "learning_rate": 0.00014390389765835292, + "loss": 0.7907, + "step": 15164 + }, + { + "epoch": 0.35549148191775526, + "grad_norm": 5.3711395263671875, + "learning_rate": 0.00014389727928202937, + "loss": 0.6035, + "step": 15165 + }, + { + "epoch": 0.35551492349256025, + "grad_norm": 2.618243932723999, + "learning_rate": 0.00014389066066751793, + "loss": 0.1858, + "step": 15166 + }, + { + "epoch": 0.35553836506736525, + "grad_norm": 3.082082509994507, + "learning_rate": 0.0001438840418148545, + "loss": 0.2874, + "step": 15167 + }, + { + "epoch": 0.35556180664217024, + "grad_norm": 0.5446350574493408, + "learning_rate": 0.00014387742272407506, + "loss": 0.0332, + "step": 15168 + }, + { + "epoch": 0.35558524821697524, + "grad_norm": 2.5968031883239746, + "learning_rate": 0.00014387080339521547, + "loss": 1.7032, + "step": 15169 + }, + { + "epoch": 0.35560868979178023, + "grad_norm": 1.1248352527618408, + "learning_rate": 0.00014386418382831168, + "loss": 0.0984, + "step": 15170 + }, + { + "epoch": 0.3556321313665852, + "grad_norm": 4.50551700592041, + "learning_rate": 0.0001438575640233996, + "loss": 0.6375, + "step": 15171 + }, + { + "epoch": 0.3556555729413902, + "grad_norm": 3.0265491008758545, + "learning_rate": 0.00014385094398051515, + "loss": 0.2632, + "step": 15172 + }, + { + "epoch": 0.3556790145161952, + "grad_norm": 3.957087993621826, + "learning_rate": 0.00014384432369969423, + "loss": 0.5858, + "step": 15173 + }, + { + "epoch": 0.3557024560910002, + "grad_norm": 3.368814706802368, + "learning_rate": 0.00014383770318097277, + "loss": 0.3853, + "step": 15174 + }, + { + "epoch": 0.3557258976658052, + "grad_norm": 5.833426475524902, + "learning_rate": 0.0001438310824243867, + "loss": 0.4745, + "step": 15175 + }, + { + "epoch": 0.3557493392406102, + "grad_norm": 2.8294010162353516, + "learning_rate": 0.00014382446142997197, + "loss": 0.3318, + "step": 15176 + }, + { + "epoch": 0.3557727808154152, + "grad_norm": 3.6838886737823486, + "learning_rate": 0.00014381784019776447, + "loss": 0.4014, + "step": 15177 + }, + { + "epoch": 0.3557962223902202, + "grad_norm": 2.9444165229797363, + "learning_rate": 0.00014381121872780012, + "loss": 0.4359, + "step": 15178 + }, + { + "epoch": 0.3558196639650252, + "grad_norm": 1.8336167335510254, + "learning_rate": 0.00014380459702011488, + "loss": 1.2016, + "step": 15179 + }, + { + "epoch": 0.3558431055398302, + "grad_norm": 1.8440747261047363, + "learning_rate": 0.00014379797507474464, + "loss": 0.228, + "step": 15180 + }, + { + "epoch": 0.35586654711463517, + "grad_norm": 4.630331993103027, + "learning_rate": 0.00014379135289172538, + "loss": 0.3591, + "step": 15181 + }, + { + "epoch": 0.35588998868944016, + "grad_norm": 2.809138298034668, + "learning_rate": 0.00014378473047109303, + "loss": 0.2514, + "step": 15182 + }, + { + "epoch": 0.35591343026424516, + "grad_norm": 2.794273614883423, + "learning_rate": 0.00014377810781288346, + "loss": 0.3179, + "step": 15183 + }, + { + "epoch": 0.35593687183905015, + "grad_norm": 4.4072699546813965, + "learning_rate": 0.00014377148491713266, + "loss": 0.6531, + "step": 15184 + }, + { + "epoch": 0.35596031341385515, + "grad_norm": 1.36702561378479, + "learning_rate": 0.00014376486178387655, + "loss": 0.1259, + "step": 15185 + }, + { + "epoch": 0.35598375498866014, + "grad_norm": 1.0904943943023682, + "learning_rate": 0.00014375823841315108, + "loss": 0.1149, + "step": 15186 + }, + { + "epoch": 0.35600719656346513, + "grad_norm": 3.592930555343628, + "learning_rate": 0.00014375161480499214, + "loss": 0.3801, + "step": 15187 + }, + { + "epoch": 0.35603063813827013, + "grad_norm": 2.266216516494751, + "learning_rate": 0.00014374499095943578, + "loss": 0.1499, + "step": 15188 + }, + { + "epoch": 0.3560540797130751, + "grad_norm": 5.888956546783447, + "learning_rate": 0.0001437383668765178, + "loss": 0.7752, + "step": 15189 + }, + { + "epoch": 0.3560775212878801, + "grad_norm": 0.6149885058403015, + "learning_rate": 0.00014373174255627425, + "loss": 0.0676, + "step": 15190 + }, + { + "epoch": 0.3561009628626851, + "grad_norm": 3.060845136642456, + "learning_rate": 0.00014372511799874105, + "loss": 0.4151, + "step": 15191 + }, + { + "epoch": 0.3561244044374901, + "grad_norm": 5.2910614013671875, + "learning_rate": 0.00014371849320395409, + "loss": 0.6347, + "step": 15192 + }, + { + "epoch": 0.3561478460122951, + "grad_norm": 3.6830921173095703, + "learning_rate": 0.00014371186817194935, + "loss": 0.3117, + "step": 15193 + }, + { + "epoch": 0.3561712875871001, + "grad_norm": 1.8580960035324097, + "learning_rate": 0.0001437052429027628, + "loss": 0.2686, + "step": 15194 + }, + { + "epoch": 0.3561947291619051, + "grad_norm": 0.773856520652771, + "learning_rate": 0.0001436986173964304, + "loss": 0.0718, + "step": 15195 + }, + { + "epoch": 0.3562181707367101, + "grad_norm": 1.1571831703186035, + "learning_rate": 0.00014369199165298807, + "loss": 0.0871, + "step": 15196 + }, + { + "epoch": 0.3562416123115151, + "grad_norm": 3.5093274116516113, + "learning_rate": 0.00014368536567247174, + "loss": 0.4673, + "step": 15197 + }, + { + "epoch": 0.35626505388632007, + "grad_norm": 5.551442623138428, + "learning_rate": 0.00014367873945491738, + "loss": 0.6286, + "step": 15198 + }, + { + "epoch": 0.35628849546112507, + "grad_norm": 3.654196262359619, + "learning_rate": 0.00014367211300036096, + "loss": 0.2936, + "step": 15199 + }, + { + "epoch": 0.35631193703593006, + "grad_norm": 7.110778331756592, + "learning_rate": 0.00014366548630883843, + "loss": 0.6843, + "step": 15200 + }, + { + "epoch": 0.35633537861073505, + "grad_norm": 4.158756256103516, + "learning_rate": 0.00014365885938038577, + "loss": 0.3211, + "step": 15201 + }, + { + "epoch": 0.35635882018554005, + "grad_norm": 1.4955722093582153, + "learning_rate": 0.00014365223221503886, + "loss": 0.1608, + "step": 15202 + }, + { + "epoch": 0.35638226176034504, + "grad_norm": 3.5026206970214844, + "learning_rate": 0.00014364560481283377, + "loss": 1.8022, + "step": 15203 + }, + { + "epoch": 0.35640570333515004, + "grad_norm": 3.1675329208374023, + "learning_rate": 0.00014363897717380636, + "loss": 0.4599, + "step": 15204 + }, + { + "epoch": 0.35642914490995503, + "grad_norm": 3.1840858459472656, + "learning_rate": 0.00014363234929799264, + "loss": 0.2393, + "step": 15205 + }, + { + "epoch": 0.35645258648476, + "grad_norm": 4.097836494445801, + "learning_rate": 0.00014362572118542855, + "loss": 0.3533, + "step": 15206 + }, + { + "epoch": 0.356476028059565, + "grad_norm": 2.2501518726348877, + "learning_rate": 0.0001436190928361501, + "loss": 1.3535, + "step": 15207 + }, + { + "epoch": 0.35649946963437, + "grad_norm": 2.8566832542419434, + "learning_rate": 0.0001436124642501932, + "loss": 0.343, + "step": 15208 + }, + { + "epoch": 0.356522911209175, + "grad_norm": 3.962078094482422, + "learning_rate": 0.00014360583542759388, + "loss": 0.3891, + "step": 15209 + }, + { + "epoch": 0.35654635278398, + "grad_norm": 3.911620855331421, + "learning_rate": 0.00014359920636838807, + "loss": 0.5455, + "step": 15210 + }, + { + "epoch": 0.356569794358785, + "grad_norm": 3.260258436203003, + "learning_rate": 0.00014359257707261168, + "loss": 0.3109, + "step": 15211 + }, + { + "epoch": 0.35659323593359, + "grad_norm": 1.7693088054656982, + "learning_rate": 0.00014358594754030078, + "loss": 0.1447, + "step": 15212 + }, + { + "epoch": 0.35661667750839504, + "grad_norm": 1.282705307006836, + "learning_rate": 0.00014357931777149132, + "loss": 0.1356, + "step": 15213 + }, + { + "epoch": 0.35664011908320004, + "grad_norm": 3.507615089416504, + "learning_rate": 0.00014357268776621923, + "loss": 0.6385, + "step": 15214 + }, + { + "epoch": 0.35666356065800503, + "grad_norm": 3.4203648567199707, + "learning_rate": 0.00014356605752452052, + "loss": 0.39, + "step": 15215 + }, + { + "epoch": 0.35668700223281, + "grad_norm": 4.223201751708984, + "learning_rate": 0.00014355942704643118, + "loss": 0.4113, + "step": 15216 + }, + { + "epoch": 0.356710443807615, + "grad_norm": 0.7680564522743225, + "learning_rate": 0.00014355279633198714, + "loss": 0.0628, + "step": 15217 + }, + { + "epoch": 0.35673388538242, + "grad_norm": 2.858279228210449, + "learning_rate": 0.00014354616538122435, + "loss": 0.3121, + "step": 15218 + }, + { + "epoch": 0.356757326957225, + "grad_norm": 2.4885988235473633, + "learning_rate": 0.00014353953419417895, + "loss": 0.2431, + "step": 15219 + }, + { + "epoch": 0.35678076853203, + "grad_norm": 2.0267815589904785, + "learning_rate": 0.00014353290277088675, + "loss": 1.3883, + "step": 15220 + }, + { + "epoch": 0.356804210106835, + "grad_norm": 1.4163107872009277, + "learning_rate": 0.0001435262711113838, + "loss": 0.1262, + "step": 15221 + }, + { + "epoch": 0.35682765168164, + "grad_norm": 0.8162400722503662, + "learning_rate": 0.00014351963921570608, + "loss": 0.0679, + "step": 15222 + }, + { + "epoch": 0.356851093256445, + "grad_norm": 2.9582276344299316, + "learning_rate": 0.00014351300708388958, + "loss": 0.31, + "step": 15223 + }, + { + "epoch": 0.35687453483125, + "grad_norm": 0.4518873393535614, + "learning_rate": 0.00014350637471597025, + "loss": 0.0646, + "step": 15224 + }, + { + "epoch": 0.356897976406055, + "grad_norm": 3.3245718479156494, + "learning_rate": 0.00014349974211198417, + "loss": 0.3254, + "step": 15225 + }, + { + "epoch": 0.35692141798085997, + "grad_norm": 0.22245803475379944, + "learning_rate": 0.00014349310927196722, + "loss": 0.0179, + "step": 15226 + }, + { + "epoch": 0.35694485955566496, + "grad_norm": 0.7560730576515198, + "learning_rate": 0.00014348647619595543, + "loss": 0.0483, + "step": 15227 + }, + { + "epoch": 0.35696830113046996, + "grad_norm": 2.842287063598633, + "learning_rate": 0.00014347984288398484, + "loss": 0.2753, + "step": 15228 + }, + { + "epoch": 0.35699174270527495, + "grad_norm": 2.52036714553833, + "learning_rate": 0.00014347320933609138, + "loss": 0.1855, + "step": 15229 + }, + { + "epoch": 0.35701518428007994, + "grad_norm": 2.758878707885742, + "learning_rate": 0.00014346657555231105, + "loss": 0.324, + "step": 15230 + }, + { + "epoch": 0.35703862585488494, + "grad_norm": 1.9583401679992676, + "learning_rate": 0.00014345994153267986, + "loss": 0.1578, + "step": 15231 + }, + { + "epoch": 0.35706206742968993, + "grad_norm": 2.2202308177948, + "learning_rate": 0.0001434533072772338, + "loss": 0.123, + "step": 15232 + }, + { + "epoch": 0.3570855090044949, + "grad_norm": 1.7712376117706299, + "learning_rate": 0.0001434466727860089, + "loss": 1.3044, + "step": 15233 + }, + { + "epoch": 0.3571089505792999, + "grad_norm": 2.8509438037872314, + "learning_rate": 0.0001434400380590411, + "loss": 0.2051, + "step": 15234 + }, + { + "epoch": 0.3571323921541049, + "grad_norm": 3.306504964828491, + "learning_rate": 0.00014343340309636645, + "loss": 1.7762, + "step": 15235 + }, + { + "epoch": 0.3571558337289099, + "grad_norm": 1.0558724403381348, + "learning_rate": 0.00014342676789802096, + "loss": 0.0718, + "step": 15236 + }, + { + "epoch": 0.3571792753037149, + "grad_norm": 1.0347399711608887, + "learning_rate": 0.00014342013246404056, + "loss": 0.0847, + "step": 15237 + }, + { + "epoch": 0.3572027168785199, + "grad_norm": 5.152961730957031, + "learning_rate": 0.0001434134967944613, + "loss": 0.3152, + "step": 15238 + }, + { + "epoch": 0.3572261584533249, + "grad_norm": 0.892647385597229, + "learning_rate": 0.00014340686088931925, + "loss": 0.0999, + "step": 15239 + }, + { + "epoch": 0.3572496000281299, + "grad_norm": 1.774659514427185, + "learning_rate": 0.0001434002247486503, + "loss": 0.1693, + "step": 15240 + }, + { + "epoch": 0.3572730416029349, + "grad_norm": 4.7614970207214355, + "learning_rate": 0.0001433935883724905, + "loss": 0.462, + "step": 15241 + }, + { + "epoch": 0.3572964831777399, + "grad_norm": 5.618350028991699, + "learning_rate": 0.0001433869517608759, + "loss": 0.5162, + "step": 15242 + }, + { + "epoch": 0.35731992475254487, + "grad_norm": 1.2140469551086426, + "learning_rate": 0.00014338031491384245, + "loss": 0.118, + "step": 15243 + }, + { + "epoch": 0.35734336632734987, + "grad_norm": 3.1382083892822266, + "learning_rate": 0.0001433736778314262, + "loss": 0.3373, + "step": 15244 + }, + { + "epoch": 0.35736680790215486, + "grad_norm": 1.0203979015350342, + "learning_rate": 0.0001433670405136632, + "loss": 0.0463, + "step": 15245 + }, + { + "epoch": 0.35739024947695985, + "grad_norm": 4.132913589477539, + "learning_rate": 0.00014336040296058936, + "loss": 0.5461, + "step": 15246 + }, + { + "epoch": 0.35741369105176485, + "grad_norm": 4.146527290344238, + "learning_rate": 0.00014335376517224076, + "loss": 0.1749, + "step": 15247 + }, + { + "epoch": 0.35743713262656984, + "grad_norm": 3.7189555168151855, + "learning_rate": 0.00014334712714865347, + "loss": 0.59, + "step": 15248 + }, + { + "epoch": 0.35746057420137484, + "grad_norm": 4.165081024169922, + "learning_rate": 0.0001433404888898634, + "loss": 0.5014, + "step": 15249 + }, + { + "epoch": 0.35748401577617983, + "grad_norm": 5.3757429122924805, + "learning_rate": 0.0001433338503959066, + "loss": 0.8165, + "step": 15250 + }, + { + "epoch": 0.3575074573509848, + "grad_norm": 0.6273885369300842, + "learning_rate": 0.00014332721166681917, + "loss": 0.0338, + "step": 15251 + }, + { + "epoch": 0.3575308989257898, + "grad_norm": 1.490094542503357, + "learning_rate": 0.00014332057270263703, + "loss": 0.1197, + "step": 15252 + }, + { + "epoch": 0.3575543405005948, + "grad_norm": 2.397447109222412, + "learning_rate": 0.00014331393350339622, + "loss": 0.1505, + "step": 15253 + }, + { + "epoch": 0.3575777820753998, + "grad_norm": 5.353799343109131, + "learning_rate": 0.00014330729406913283, + "loss": 0.5125, + "step": 15254 + }, + { + "epoch": 0.3576012236502048, + "grad_norm": 4.823612689971924, + "learning_rate": 0.00014330065439988285, + "loss": 0.2631, + "step": 15255 + }, + { + "epoch": 0.3576246652250098, + "grad_norm": 3.800177812576294, + "learning_rate": 0.00014329401449568227, + "loss": 0.2998, + "step": 15256 + }, + { + "epoch": 0.3576481067998148, + "grad_norm": 0.4467701017856598, + "learning_rate": 0.00014328737435656717, + "loss": 0.0475, + "step": 15257 + }, + { + "epoch": 0.3576715483746198, + "grad_norm": 0.8288697600364685, + "learning_rate": 0.00014328073398257358, + "loss": 0.0475, + "step": 15258 + }, + { + "epoch": 0.3576949899494248, + "grad_norm": 3.832423448562622, + "learning_rate": 0.00014327409337373746, + "loss": 0.632, + "step": 15259 + }, + { + "epoch": 0.3577184315242298, + "grad_norm": 2.298456907272339, + "learning_rate": 0.00014326745253009492, + "loss": 0.2594, + "step": 15260 + }, + { + "epoch": 0.35774187309903477, + "grad_norm": 3.2658207416534424, + "learning_rate": 0.00014326081145168198, + "loss": 0.424, + "step": 15261 + }, + { + "epoch": 0.35776531467383976, + "grad_norm": 4.520688056945801, + "learning_rate": 0.00014325417013853463, + "loss": 0.4471, + "step": 15262 + }, + { + "epoch": 0.35778875624864476, + "grad_norm": 4.956229209899902, + "learning_rate": 0.00014324752859068895, + "loss": 0.3907, + "step": 15263 + }, + { + "epoch": 0.3578121978234498, + "grad_norm": 2.7930185794830322, + "learning_rate": 0.00014324088680818097, + "loss": 0.295, + "step": 15264 + }, + { + "epoch": 0.3578356393982548, + "grad_norm": 1.8859357833862305, + "learning_rate": 0.00014323424479104673, + "loss": 0.278, + "step": 15265 + }, + { + "epoch": 0.3578590809730598, + "grad_norm": 2.1603124141693115, + "learning_rate": 0.00014322760253932223, + "loss": 0.6867, + "step": 15266 + }, + { + "epoch": 0.3578825225478648, + "grad_norm": 3.491240978240967, + "learning_rate": 0.00014322096005304358, + "loss": 0.4622, + "step": 15267 + }, + { + "epoch": 0.3579059641226698, + "grad_norm": 6.468297481536865, + "learning_rate": 0.00014321431733224678, + "loss": 0.3664, + "step": 15268 + }, + { + "epoch": 0.3579294056974748, + "grad_norm": 3.1439528465270996, + "learning_rate": 0.00014320767437696786, + "loss": 1.7697, + "step": 15269 + }, + { + "epoch": 0.3579528472722798, + "grad_norm": 3.752032518386841, + "learning_rate": 0.0001432010311872429, + "loss": 0.2046, + "step": 15270 + }, + { + "epoch": 0.35797628884708477, + "grad_norm": 1.56609308719635, + "learning_rate": 0.0001431943877631079, + "loss": 0.1111, + "step": 15271 + }, + { + "epoch": 0.35799973042188976, + "grad_norm": 3.7115633487701416, + "learning_rate": 0.00014318774410459895, + "loss": 0.4334, + "step": 15272 + }, + { + "epoch": 0.35802317199669476, + "grad_norm": 5.374024868011475, + "learning_rate": 0.0001431811002117521, + "loss": 0.3962, + "step": 15273 + }, + { + "epoch": 0.35804661357149975, + "grad_norm": 4.12343168258667, + "learning_rate": 0.00014317445608460342, + "loss": 0.4147, + "step": 15274 + }, + { + "epoch": 0.35807005514630474, + "grad_norm": 4.615440845489502, + "learning_rate": 0.00014316781172318886, + "loss": 0.3275, + "step": 15275 + }, + { + "epoch": 0.35809349672110974, + "grad_norm": 2.2379872798919678, + "learning_rate": 0.00014316116712754457, + "loss": 0.1897, + "step": 15276 + }, + { + "epoch": 0.35811693829591473, + "grad_norm": 2.5266270637512207, + "learning_rate": 0.00014315452229770657, + "loss": 0.1806, + "step": 15277 + }, + { + "epoch": 0.3581403798707197, + "grad_norm": 2.6802191734313965, + "learning_rate": 0.00014314787723371093, + "loss": 0.2929, + "step": 15278 + }, + { + "epoch": 0.3581638214455247, + "grad_norm": 1.346923589706421, + "learning_rate": 0.00014314123193559367, + "loss": 0.15, + "step": 15279 + }, + { + "epoch": 0.3581872630203297, + "grad_norm": 3.507683277130127, + "learning_rate": 0.0001431345864033909, + "loss": 0.2376, + "step": 15280 + }, + { + "epoch": 0.3582107045951347, + "grad_norm": 3.0753862857818604, + "learning_rate": 0.00014312794063713863, + "loss": 0.2733, + "step": 15281 + }, + { + "epoch": 0.3582341461699397, + "grad_norm": 3.650320529937744, + "learning_rate": 0.00014312129463687296, + "loss": 0.3115, + "step": 15282 + }, + { + "epoch": 0.3582575877447447, + "grad_norm": 3.933373212814331, + "learning_rate": 0.00014311464840262994, + "loss": 0.3829, + "step": 15283 + }, + { + "epoch": 0.3582810293195497, + "grad_norm": 4.602670192718506, + "learning_rate": 0.0001431080019344456, + "loss": 0.5099, + "step": 15284 + }, + { + "epoch": 0.3583044708943547, + "grad_norm": 4.45597505569458, + "learning_rate": 0.00014310135523235603, + "loss": 0.4018, + "step": 15285 + }, + { + "epoch": 0.3583279124691597, + "grad_norm": 3.5018279552459717, + "learning_rate": 0.0001430947082963973, + "loss": 1.476, + "step": 15286 + }, + { + "epoch": 0.3583513540439647, + "grad_norm": 2.8162827491760254, + "learning_rate": 0.0001430880611266055, + "loss": 0.29, + "step": 15287 + }, + { + "epoch": 0.35837479561876967, + "grad_norm": 2.3754749298095703, + "learning_rate": 0.00014308141372301663, + "loss": 0.2285, + "step": 15288 + }, + { + "epoch": 0.35839823719357466, + "grad_norm": 2.997666358947754, + "learning_rate": 0.0001430747660856668, + "loss": 0.3629, + "step": 15289 + }, + { + "epoch": 0.35842167876837966, + "grad_norm": 2.947349786758423, + "learning_rate": 0.0001430681182145921, + "loss": 0.2618, + "step": 15290 + }, + { + "epoch": 0.35844512034318465, + "grad_norm": 3.6870317459106445, + "learning_rate": 0.00014306147010982855, + "loss": 0.6167, + "step": 15291 + }, + { + "epoch": 0.35846856191798965, + "grad_norm": 3.835400104522705, + "learning_rate": 0.0001430548217714123, + "loss": 0.2646, + "step": 15292 + }, + { + "epoch": 0.35849200349279464, + "grad_norm": 5.150649547576904, + "learning_rate": 0.00014304817319937935, + "loss": 0.4366, + "step": 15293 + }, + { + "epoch": 0.35851544506759964, + "grad_norm": 0.731773316860199, + "learning_rate": 0.0001430415243937658, + "loss": 0.1006, + "step": 15294 + }, + { + "epoch": 0.35853888664240463, + "grad_norm": 3.0696492195129395, + "learning_rate": 0.0001430348753546077, + "loss": 0.3438, + "step": 15295 + }, + { + "epoch": 0.3585623282172096, + "grad_norm": 1.5194822549819946, + "learning_rate": 0.00014302822608194119, + "loss": 0.1606, + "step": 15296 + }, + { + "epoch": 0.3585857697920146, + "grad_norm": 4.391671657562256, + "learning_rate": 0.00014302157657580233, + "loss": 0.4635, + "step": 15297 + }, + { + "epoch": 0.3586092113668196, + "grad_norm": 4.467856407165527, + "learning_rate": 0.00014301492683622716, + "loss": 0.3347, + "step": 15298 + }, + { + "epoch": 0.3586326529416246, + "grad_norm": 2.4625988006591797, + "learning_rate": 0.0001430082768632518, + "loss": 1.487, + "step": 15299 + }, + { + "epoch": 0.3586560945164296, + "grad_norm": 4.343918323516846, + "learning_rate": 0.0001430016266569123, + "loss": 0.5683, + "step": 15300 + }, + { + "epoch": 0.3586795360912346, + "grad_norm": 2.5982136726379395, + "learning_rate": 0.00014299497621724478, + "loss": 0.2554, + "step": 15301 + }, + { + "epoch": 0.3587029776660396, + "grad_norm": 5.767280101776123, + "learning_rate": 0.0001429883255442853, + "loss": 0.8993, + "step": 15302 + }, + { + "epoch": 0.3587264192408446, + "grad_norm": 0.9156640768051147, + "learning_rate": 0.00014298167463806997, + "loss": 0.121, + "step": 15303 + }, + { + "epoch": 0.3587498608156496, + "grad_norm": 4.8436760902404785, + "learning_rate": 0.00014297502349863486, + "loss": 0.6275, + "step": 15304 + }, + { + "epoch": 0.3587733023904546, + "grad_norm": 3.614647626876831, + "learning_rate": 0.00014296837212601608, + "loss": 0.265, + "step": 15305 + }, + { + "epoch": 0.35879674396525957, + "grad_norm": 2.6689226627349854, + "learning_rate": 0.0001429617205202497, + "loss": 0.2855, + "step": 15306 + }, + { + "epoch": 0.35882018554006456, + "grad_norm": 3.5801634788513184, + "learning_rate": 0.0001429550686813718, + "loss": 0.4709, + "step": 15307 + }, + { + "epoch": 0.35884362711486956, + "grad_norm": 3.8439409732818604, + "learning_rate": 0.00014294841660941846, + "loss": 0.4153, + "step": 15308 + }, + { + "epoch": 0.35886706868967455, + "grad_norm": 2.4441566467285156, + "learning_rate": 0.00014294176430442585, + "loss": 0.2997, + "step": 15309 + }, + { + "epoch": 0.35889051026447955, + "grad_norm": 1.3870418071746826, + "learning_rate": 0.00014293511176643003, + "loss": 0.157, + "step": 15310 + }, + { + "epoch": 0.35891395183928454, + "grad_norm": 3.4874930381774902, + "learning_rate": 0.00014292845899546705, + "loss": 0.3272, + "step": 15311 + }, + { + "epoch": 0.35893739341408953, + "grad_norm": 4.164764881134033, + "learning_rate": 0.00014292180599157308, + "loss": 0.3095, + "step": 15312 + }, + { + "epoch": 0.35896083498889453, + "grad_norm": 2.9876575469970703, + "learning_rate": 0.00014291515275478416, + "loss": 1.8904, + "step": 15313 + }, + { + "epoch": 0.3589842765636995, + "grad_norm": 3.0161025524139404, + "learning_rate": 0.0001429084992851364, + "loss": 0.9529, + "step": 15314 + }, + { + "epoch": 0.3590077181385046, + "grad_norm": 1.1437112092971802, + "learning_rate": 0.00014290184558266592, + "loss": 0.1242, + "step": 15315 + }, + { + "epoch": 0.35903115971330957, + "grad_norm": 5.075084209442139, + "learning_rate": 0.00014289519164740886, + "loss": 1.6857, + "step": 15316 + }, + { + "epoch": 0.35905460128811456, + "grad_norm": 2.6684670448303223, + "learning_rate": 0.00014288853747940124, + "loss": 0.324, + "step": 15317 + }, + { + "epoch": 0.35907804286291956, + "grad_norm": 0.5153065919876099, + "learning_rate": 0.00014288188307867925, + "loss": 0.0344, + "step": 15318 + }, + { + "epoch": 0.35910148443772455, + "grad_norm": 2.963669538497925, + "learning_rate": 0.00014287522844527893, + "loss": 1.7249, + "step": 15319 + }, + { + "epoch": 0.35912492601252954, + "grad_norm": 0.8996812105178833, + "learning_rate": 0.0001428685735792364, + "loss": 0.0725, + "step": 15320 + }, + { + "epoch": 0.35914836758733454, + "grad_norm": 0.6949593424797058, + "learning_rate": 0.0001428619184805878, + "loss": 0.0683, + "step": 15321 + }, + { + "epoch": 0.35917180916213953, + "grad_norm": 5.24788761138916, + "learning_rate": 0.00014285526314936924, + "loss": 0.4802, + "step": 15322 + }, + { + "epoch": 0.3591952507369445, + "grad_norm": 3.875213623046875, + "learning_rate": 0.0001428486075856168, + "loss": 0.2009, + "step": 15323 + }, + { + "epoch": 0.3592186923117495, + "grad_norm": 0.5965266823768616, + "learning_rate": 0.0001428419517893666, + "loss": 0.0297, + "step": 15324 + }, + { + "epoch": 0.3592421338865545, + "grad_norm": 1.1276267766952515, + "learning_rate": 0.00014283529576065483, + "loss": 0.124, + "step": 15325 + }, + { + "epoch": 0.3592655754613595, + "grad_norm": 3.9668703079223633, + "learning_rate": 0.00014282863949951742, + "loss": 1.3647, + "step": 15326 + }, + { + "epoch": 0.3592890170361645, + "grad_norm": 2.776357412338257, + "learning_rate": 0.0001428219830059907, + "loss": 1.3853, + "step": 15327 + }, + { + "epoch": 0.3593124586109695, + "grad_norm": 0.9593685269355774, + "learning_rate": 0.0001428153262801107, + "loss": 0.0816, + "step": 15328 + }, + { + "epoch": 0.3593359001857745, + "grad_norm": 2.8321845531463623, + "learning_rate": 0.0001428086693219135, + "loss": 0.5271, + "step": 15329 + }, + { + "epoch": 0.3593593417605795, + "grad_norm": 2.8924858570098877, + "learning_rate": 0.00014280201213143527, + "loss": 0.2702, + "step": 15330 + }, + { + "epoch": 0.3593827833353845, + "grad_norm": 2.902803421020508, + "learning_rate": 0.00014279535470871211, + "loss": 0.281, + "step": 15331 + }, + { + "epoch": 0.3594062249101895, + "grad_norm": 2.2618870735168457, + "learning_rate": 0.00014278869705378018, + "loss": 0.122, + "step": 15332 + }, + { + "epoch": 0.35942966648499447, + "grad_norm": 4.374307632446289, + "learning_rate": 0.00014278203916667553, + "loss": 0.4261, + "step": 15333 + }, + { + "epoch": 0.35945310805979946, + "grad_norm": 0.4555520713329315, + "learning_rate": 0.00014277538104743437, + "loss": 0.0204, + "step": 15334 + }, + { + "epoch": 0.35947654963460446, + "grad_norm": 3.488034963607788, + "learning_rate": 0.00014276872269609282, + "loss": 0.3831, + "step": 15335 + }, + { + "epoch": 0.35949999120940945, + "grad_norm": 3.7375948429107666, + "learning_rate": 0.00014276206411268692, + "loss": 1.4573, + "step": 15336 + }, + { + "epoch": 0.35952343278421445, + "grad_norm": 3.447014093399048, + "learning_rate": 0.00014275540529725287, + "loss": 0.2965, + "step": 15337 + }, + { + "epoch": 0.35954687435901944, + "grad_norm": 5.610274791717529, + "learning_rate": 0.00014274874624982677, + "loss": 0.6113, + "step": 15338 + }, + { + "epoch": 0.35957031593382444, + "grad_norm": 4.404144287109375, + "learning_rate": 0.00014274208697044477, + "loss": 0.2815, + "step": 15339 + }, + { + "epoch": 0.35959375750862943, + "grad_norm": 2.897088050842285, + "learning_rate": 0.00014273542745914307, + "loss": 0.2875, + "step": 15340 + }, + { + "epoch": 0.3596171990834344, + "grad_norm": 3.1162071228027344, + "learning_rate": 0.00014272876771595768, + "loss": 0.3788, + "step": 15341 + }, + { + "epoch": 0.3596406406582394, + "grad_norm": 1.1849374771118164, + "learning_rate": 0.00014272210774092482, + "loss": 0.1424, + "step": 15342 + }, + { + "epoch": 0.3596640822330444, + "grad_norm": 4.287553310394287, + "learning_rate": 0.0001427154475340806, + "loss": 0.2846, + "step": 15343 + }, + { + "epoch": 0.3596875238078494, + "grad_norm": 4.824264049530029, + "learning_rate": 0.00014270878709546114, + "loss": 0.3825, + "step": 15344 + }, + { + "epoch": 0.3597109653826544, + "grad_norm": 1.0445350408554077, + "learning_rate": 0.0001427021264251026, + "loss": 0.118, + "step": 15345 + }, + { + "epoch": 0.3597344069574594, + "grad_norm": 0.9520865678787231, + "learning_rate": 0.00014269546552304115, + "loss": 0.0827, + "step": 15346 + }, + { + "epoch": 0.3597578485322644, + "grad_norm": 1.3939759731292725, + "learning_rate": 0.0001426888043893129, + "loss": 0.1314, + "step": 15347 + }, + { + "epoch": 0.3597812901070694, + "grad_norm": 3.217381715774536, + "learning_rate": 0.00014268214302395397, + "loss": 0.4201, + "step": 15348 + }, + { + "epoch": 0.3598047316818744, + "grad_norm": 6.280221939086914, + "learning_rate": 0.00014267548142700054, + "loss": 0.4975, + "step": 15349 + }, + { + "epoch": 0.3598281732566794, + "grad_norm": 3.7485363483428955, + "learning_rate": 0.00014266881959848878, + "loss": 0.149, + "step": 15350 + }, + { + "epoch": 0.35985161483148437, + "grad_norm": 4.19219446182251, + "learning_rate": 0.00014266215753845477, + "loss": 0.3938, + "step": 15351 + }, + { + "epoch": 0.35987505640628936, + "grad_norm": 4.469082355499268, + "learning_rate": 0.00014265549524693468, + "loss": 0.3364, + "step": 15352 + }, + { + "epoch": 0.35989849798109436, + "grad_norm": 1.7668280601501465, + "learning_rate": 0.00014264883272396473, + "loss": 0.2273, + "step": 15353 + }, + { + "epoch": 0.35992193955589935, + "grad_norm": 3.0184009075164795, + "learning_rate": 0.000142642169969581, + "loss": 1.2579, + "step": 15354 + }, + { + "epoch": 0.35994538113070434, + "grad_norm": 2.5535316467285156, + "learning_rate": 0.00014263550698381962, + "loss": 0.2602, + "step": 15355 + }, + { + "epoch": 0.35996882270550934, + "grad_norm": 3.7697932720184326, + "learning_rate": 0.0001426288437667168, + "loss": 0.2471, + "step": 15356 + }, + { + "epoch": 0.35999226428031433, + "grad_norm": 3.612574338912964, + "learning_rate": 0.00014262218031830867, + "loss": 0.4748, + "step": 15357 + }, + { + "epoch": 0.36001570585511933, + "grad_norm": 3.5527243614196777, + "learning_rate": 0.0001426155166386314, + "loss": 0.3577, + "step": 15358 + }, + { + "epoch": 0.3600391474299243, + "grad_norm": 0.9187738299369812, + "learning_rate": 0.00014260885272772113, + "loss": 0.1564, + "step": 15359 + }, + { + "epoch": 0.3600625890047293, + "grad_norm": 3.3313417434692383, + "learning_rate": 0.00014260218858561406, + "loss": 0.4162, + "step": 15360 + }, + { + "epoch": 0.3600860305795343, + "grad_norm": 4.744118690490723, + "learning_rate": 0.00014259552421234625, + "loss": 0.6398, + "step": 15361 + }, + { + "epoch": 0.3601094721543393, + "grad_norm": 2.648375988006592, + "learning_rate": 0.000142588859607954, + "loss": 0.2681, + "step": 15362 + }, + { + "epoch": 0.3601329137291443, + "grad_norm": 3.24776554107666, + "learning_rate": 0.00014258219477247333, + "loss": 0.4227, + "step": 15363 + }, + { + "epoch": 0.3601563553039493, + "grad_norm": 1.5647094249725342, + "learning_rate": 0.00014257552970594052, + "loss": 0.0615, + "step": 15364 + }, + { + "epoch": 0.3601797968787543, + "grad_norm": 3.1922154426574707, + "learning_rate": 0.0001425688644083917, + "loss": 0.3282, + "step": 15365 + }, + { + "epoch": 0.3602032384535593, + "grad_norm": 3.8805618286132812, + "learning_rate": 0.00014256219887986299, + "loss": 0.4645, + "step": 15366 + }, + { + "epoch": 0.36022668002836433, + "grad_norm": 1.5328810214996338, + "learning_rate": 0.0001425555331203906, + "loss": 0.1399, + "step": 15367 + }, + { + "epoch": 0.3602501216031693, + "grad_norm": 2.568143606185913, + "learning_rate": 0.00014254886713001072, + "loss": 0.2865, + "step": 15368 + }, + { + "epoch": 0.3602735631779743, + "grad_norm": 3.678344964981079, + "learning_rate": 0.00014254220090875944, + "loss": 0.3205, + "step": 15369 + }, + { + "epoch": 0.3602970047527793, + "grad_norm": 4.182609558105469, + "learning_rate": 0.000142535534456673, + "loss": 0.4666, + "step": 15370 + }, + { + "epoch": 0.3603204463275843, + "grad_norm": 2.8045547008514404, + "learning_rate": 0.0001425288677737876, + "loss": 1.9854, + "step": 15371 + }, + { + "epoch": 0.3603438879023893, + "grad_norm": 5.90939474105835, + "learning_rate": 0.00014252220086013933, + "loss": 0.9229, + "step": 15372 + }, + { + "epoch": 0.3603673294771943, + "grad_norm": 1.0325732231140137, + "learning_rate": 0.00014251553371576441, + "loss": 0.1377, + "step": 15373 + }, + { + "epoch": 0.3603907710519993, + "grad_norm": 1.1187241077423096, + "learning_rate": 0.00014250886634069898, + "loss": 0.0875, + "step": 15374 + }, + { + "epoch": 0.3604142126268043, + "grad_norm": 2.258186101913452, + "learning_rate": 0.00014250219873497926, + "loss": 0.3795, + "step": 15375 + }, + { + "epoch": 0.3604376542016093, + "grad_norm": 3.628408432006836, + "learning_rate": 0.00014249553089864142, + "loss": 0.3795, + "step": 15376 + }, + { + "epoch": 0.3604610957764143, + "grad_norm": 4.079799652099609, + "learning_rate": 0.00014248886283172165, + "loss": 0.4736, + "step": 15377 + }, + { + "epoch": 0.36048453735121927, + "grad_norm": 0.6735917329788208, + "learning_rate": 0.00014248219453425612, + "loss": 0.0738, + "step": 15378 + }, + { + "epoch": 0.36050797892602426, + "grad_norm": 3.272662401199341, + "learning_rate": 0.000142475526006281, + "loss": 0.2845, + "step": 15379 + }, + { + "epoch": 0.36053142050082926, + "grad_norm": 3.257654905319214, + "learning_rate": 0.00014246885724783246, + "loss": 0.33, + "step": 15380 + }, + { + "epoch": 0.36055486207563425, + "grad_norm": 3.37742018699646, + "learning_rate": 0.0001424621882589467, + "loss": 0.3611, + "step": 15381 + }, + { + "epoch": 0.36057830365043925, + "grad_norm": 2.2192494869232178, + "learning_rate": 0.00014245551903965993, + "loss": 0.3074, + "step": 15382 + }, + { + "epoch": 0.36060174522524424, + "grad_norm": 4.196509838104248, + "learning_rate": 0.0001424488495900083, + "loss": 0.514, + "step": 15383 + }, + { + "epoch": 0.36062518680004924, + "grad_norm": 2.970001697540283, + "learning_rate": 0.00014244217991002807, + "loss": 0.6003, + "step": 15384 + }, + { + "epoch": 0.36064862837485423, + "grad_norm": 2.567490339279175, + "learning_rate": 0.00014243550999975535, + "loss": 0.2206, + "step": 15385 + }, + { + "epoch": 0.3606720699496592, + "grad_norm": 5.655520439147949, + "learning_rate": 0.00014242883985922635, + "loss": 0.6581, + "step": 15386 + }, + { + "epoch": 0.3606955115244642, + "grad_norm": 7.696215629577637, + "learning_rate": 0.00014242216948847728, + "loss": 0.5096, + "step": 15387 + }, + { + "epoch": 0.3607189530992692, + "grad_norm": 1.4359104633331299, + "learning_rate": 0.00014241549888754432, + "loss": 0.1615, + "step": 15388 + }, + { + "epoch": 0.3607423946740742, + "grad_norm": 3.5332746505737305, + "learning_rate": 0.00014240882805646368, + "loss": 1.7059, + "step": 15389 + }, + { + "epoch": 0.3607658362488792, + "grad_norm": 4.583108901977539, + "learning_rate": 0.00014240215699527154, + "loss": 0.2447, + "step": 15390 + }, + { + "epoch": 0.3607892778236842, + "grad_norm": 4.0891947746276855, + "learning_rate": 0.0001423954857040041, + "loss": 0.5048, + "step": 15391 + }, + { + "epoch": 0.3608127193984892, + "grad_norm": 2.719757318496704, + "learning_rate": 0.00014238881418269757, + "loss": 1.1335, + "step": 15392 + }, + { + "epoch": 0.3608361609732942, + "grad_norm": 0.9206036329269409, + "learning_rate": 0.00014238214243138814, + "loss": 0.4825, + "step": 15393 + }, + { + "epoch": 0.3608596025480992, + "grad_norm": 3.6053194999694824, + "learning_rate": 0.000142375470450112, + "loss": 0.2936, + "step": 15394 + }, + { + "epoch": 0.3608830441229042, + "grad_norm": 2.570803165435791, + "learning_rate": 0.0001423687982389054, + "loss": 0.2469, + "step": 15395 + }, + { + "epoch": 0.36090648569770917, + "grad_norm": 2.888139009475708, + "learning_rate": 0.0001423621257978045, + "loss": 0.2776, + "step": 15396 + }, + { + "epoch": 0.36092992727251416, + "grad_norm": 1.1237847805023193, + "learning_rate": 0.0001423554531268455, + "loss": 0.1068, + "step": 15397 + }, + { + "epoch": 0.36095336884731916, + "grad_norm": 2.2652270793914795, + "learning_rate": 0.00014234878022606465, + "loss": 0.1991, + "step": 15398 + }, + { + "epoch": 0.36097681042212415, + "grad_norm": 3.8030922412872314, + "learning_rate": 0.00014234210709549805, + "loss": 0.5951, + "step": 15399 + }, + { + "epoch": 0.36100025199692914, + "grad_norm": 1.7895002365112305, + "learning_rate": 0.00014233543373518205, + "loss": 1.3157, + "step": 15400 + }, + { + "epoch": 0.36102369357173414, + "grad_norm": 2.7420947551727295, + "learning_rate": 0.0001423287601451528, + "loss": 0.098, + "step": 15401 + }, + { + "epoch": 0.36104713514653913, + "grad_norm": 2.7262144088745117, + "learning_rate": 0.00014232208632544648, + "loss": 0.272, + "step": 15402 + }, + { + "epoch": 0.3610705767213441, + "grad_norm": 4.208614349365234, + "learning_rate": 0.00014231541227609934, + "loss": 0.3831, + "step": 15403 + }, + { + "epoch": 0.3610940182961491, + "grad_norm": 3.603248357772827, + "learning_rate": 0.0001423087379971476, + "loss": 0.6539, + "step": 15404 + }, + { + "epoch": 0.3611174598709541, + "grad_norm": 0.6973448395729065, + "learning_rate": 0.00014230206348862744, + "loss": 0.0628, + "step": 15405 + }, + { + "epoch": 0.3611409014457591, + "grad_norm": 1.9257677793502808, + "learning_rate": 0.0001422953887505751, + "loss": 0.186, + "step": 15406 + }, + { + "epoch": 0.3611643430205641, + "grad_norm": 0.7253926992416382, + "learning_rate": 0.0001422887137830268, + "loss": 0.0585, + "step": 15407 + }, + { + "epoch": 0.3611877845953691, + "grad_norm": 1.8138914108276367, + "learning_rate": 0.00014228203858601874, + "loss": 0.2126, + "step": 15408 + }, + { + "epoch": 0.3612112261701741, + "grad_norm": 2.498674154281616, + "learning_rate": 0.00014227536315958715, + "loss": 0.1939, + "step": 15409 + }, + { + "epoch": 0.3612346677449791, + "grad_norm": 2.691640615463257, + "learning_rate": 0.00014226868750376824, + "loss": 0.3674, + "step": 15410 + }, + { + "epoch": 0.3612581093197841, + "grad_norm": 1.195204257965088, + "learning_rate": 0.00014226201161859824, + "loss": 0.1206, + "step": 15411 + }, + { + "epoch": 0.3612815508945891, + "grad_norm": 2.261648654937744, + "learning_rate": 0.00014225533550411338, + "loss": 0.1684, + "step": 15412 + }, + { + "epoch": 0.36130499246939407, + "grad_norm": 3.6226806640625, + "learning_rate": 0.0001422486591603499, + "loss": 0.4327, + "step": 15413 + }, + { + "epoch": 0.36132843404419906, + "grad_norm": 1.141401767730713, + "learning_rate": 0.000142241982587344, + "loss": 0.1231, + "step": 15414 + }, + { + "epoch": 0.36135187561900406, + "grad_norm": 4.129227638244629, + "learning_rate": 0.0001422353057851319, + "loss": 0.3674, + "step": 15415 + }, + { + "epoch": 0.36137531719380905, + "grad_norm": 1.2333452701568604, + "learning_rate": 0.00014222862875374988, + "loss": 0.1334, + "step": 15416 + }, + { + "epoch": 0.36139875876861405, + "grad_norm": 1.1154595613479614, + "learning_rate": 0.0001422219514932341, + "loss": 0.0506, + "step": 15417 + }, + { + "epoch": 0.3614222003434191, + "grad_norm": 2.338559627532959, + "learning_rate": 0.0001422152740036208, + "loss": 0.3204, + "step": 15418 + }, + { + "epoch": 0.3614456419182241, + "grad_norm": 4.127712249755859, + "learning_rate": 0.00014220859628494628, + "loss": 0.4619, + "step": 15419 + }, + { + "epoch": 0.3614690834930291, + "grad_norm": 0.8103170394897461, + "learning_rate": 0.00014220191833724674, + "loss": 0.1452, + "step": 15420 + }, + { + "epoch": 0.3614925250678341, + "grad_norm": 2.464797019958496, + "learning_rate": 0.00014219524016055837, + "loss": 1.8048, + "step": 15421 + }, + { + "epoch": 0.3615159666426391, + "grad_norm": 1.699133038520813, + "learning_rate": 0.00014218856175491747, + "loss": 0.133, + "step": 15422 + }, + { + "epoch": 0.36153940821744407, + "grad_norm": 1.4800759553909302, + "learning_rate": 0.0001421818831203602, + "loss": 1.1179, + "step": 15423 + }, + { + "epoch": 0.36156284979224906, + "grad_norm": 0.7541555762290955, + "learning_rate": 0.0001421752042569229, + "loss": 0.0452, + "step": 15424 + }, + { + "epoch": 0.36158629136705406, + "grad_norm": 3.4057648181915283, + "learning_rate": 0.00014216852516464173, + "loss": 0.3167, + "step": 15425 + }, + { + "epoch": 0.36160973294185905, + "grad_norm": 3.3818130493164062, + "learning_rate": 0.00014216184584355296, + "loss": 0.4815, + "step": 15426 + }, + { + "epoch": 0.36163317451666405, + "grad_norm": 3.2508656978607178, + "learning_rate": 0.00014215516629369286, + "loss": 0.3619, + "step": 15427 + }, + { + "epoch": 0.36165661609146904, + "grad_norm": 1.3557593822479248, + "learning_rate": 0.00014214848651509759, + "loss": 0.1309, + "step": 15428 + }, + { + "epoch": 0.36168005766627404, + "grad_norm": 1.028057336807251, + "learning_rate": 0.00014214180650780347, + "loss": 0.1788, + "step": 15429 + }, + { + "epoch": 0.36170349924107903, + "grad_norm": 7.346075057983398, + "learning_rate": 0.0001421351262718467, + "loss": 0.6388, + "step": 15430 + }, + { + "epoch": 0.361726940815884, + "grad_norm": 2.575711488723755, + "learning_rate": 0.0001421284458072636, + "loss": 0.1867, + "step": 15431 + }, + { + "epoch": 0.361750382390689, + "grad_norm": 2.6362156867980957, + "learning_rate": 0.00014212176511409032, + "loss": 2.1232, + "step": 15432 + }, + { + "epoch": 0.361773823965494, + "grad_norm": 1.4674625396728516, + "learning_rate": 0.00014211508419236318, + "loss": 0.1219, + "step": 15433 + }, + { + "epoch": 0.361797265540299, + "grad_norm": 5.299696445465088, + "learning_rate": 0.0001421084030421184, + "loss": 0.1875, + "step": 15434 + }, + { + "epoch": 0.361820707115104, + "grad_norm": 3.8601856231689453, + "learning_rate": 0.00014210172166339227, + "loss": 0.5623, + "step": 15435 + }, + { + "epoch": 0.361844148689909, + "grad_norm": 3.6910130977630615, + "learning_rate": 0.000142095040056221, + "loss": 0.4801, + "step": 15436 + }, + { + "epoch": 0.361867590264714, + "grad_norm": 3.0637989044189453, + "learning_rate": 0.00014208835822064083, + "loss": 1.7613, + "step": 15437 + }, + { + "epoch": 0.361891031839519, + "grad_norm": 4.769352436065674, + "learning_rate": 0.00014208167615668805, + "loss": 0.3465, + "step": 15438 + }, + { + "epoch": 0.361914473414324, + "grad_norm": 2.780301809310913, + "learning_rate": 0.00014207499386439896, + "loss": 1.9846, + "step": 15439 + }, + { + "epoch": 0.361937914989129, + "grad_norm": 3.072944164276123, + "learning_rate": 0.00014206831134380972, + "loss": 0.2921, + "step": 15440 + }, + { + "epoch": 0.36196135656393397, + "grad_norm": 3.4202940464019775, + "learning_rate": 0.00014206162859495665, + "loss": 0.3648, + "step": 15441 + }, + { + "epoch": 0.36198479813873896, + "grad_norm": 4.852057933807373, + "learning_rate": 0.00014205494561787601, + "loss": 0.707, + "step": 15442 + }, + { + "epoch": 0.36200823971354396, + "grad_norm": 3.0256755352020264, + "learning_rate": 0.00014204826241260408, + "loss": 0.216, + "step": 15443 + }, + { + "epoch": 0.36203168128834895, + "grad_norm": 1.7925111055374146, + "learning_rate": 0.00014204157897917705, + "loss": 0.1901, + "step": 15444 + }, + { + "epoch": 0.36205512286315394, + "grad_norm": 1.6441495418548584, + "learning_rate": 0.00014203489531763126, + "loss": 0.2178, + "step": 15445 + }, + { + "epoch": 0.36207856443795894, + "grad_norm": 0.7056612372398376, + "learning_rate": 0.00014202821142800294, + "loss": 0.0557, + "step": 15446 + }, + { + "epoch": 0.36210200601276393, + "grad_norm": 4.98630428314209, + "learning_rate": 0.00014202152731032834, + "loss": 0.3747, + "step": 15447 + }, + { + "epoch": 0.3621254475875689, + "grad_norm": 1.6826503276824951, + "learning_rate": 0.00014201484296464377, + "loss": 1.0731, + "step": 15448 + }, + { + "epoch": 0.3621488891623739, + "grad_norm": 3.0131187438964844, + "learning_rate": 0.0001420081583909855, + "loss": 0.2081, + "step": 15449 + }, + { + "epoch": 0.3621723307371789, + "grad_norm": 1.3121930360794067, + "learning_rate": 0.00014200147358938973, + "loss": 0.1963, + "step": 15450 + }, + { + "epoch": 0.3621957723119839, + "grad_norm": 5.523985862731934, + "learning_rate": 0.00014199478855989284, + "loss": 0.5243, + "step": 15451 + }, + { + "epoch": 0.3622192138867889, + "grad_norm": 2.2065846920013428, + "learning_rate": 0.00014198810330253104, + "loss": 0.1925, + "step": 15452 + }, + { + "epoch": 0.3622426554615939, + "grad_norm": 5.160759449005127, + "learning_rate": 0.00014198141781734055, + "loss": 0.2097, + "step": 15453 + }, + { + "epoch": 0.3622660970363989, + "grad_norm": 1.66021728515625, + "learning_rate": 0.00014197473210435775, + "loss": 0.7921, + "step": 15454 + }, + { + "epoch": 0.3622895386112039, + "grad_norm": 3.988954544067383, + "learning_rate": 0.00014196804616361887, + "loss": 0.3792, + "step": 15455 + }, + { + "epoch": 0.3623129801860089, + "grad_norm": 2.896503210067749, + "learning_rate": 0.0001419613599951602, + "loss": 1.3524, + "step": 15456 + }, + { + "epoch": 0.3623364217608139, + "grad_norm": 1.3910590410232544, + "learning_rate": 0.000141954673599018, + "loss": 0.1342, + "step": 15457 + }, + { + "epoch": 0.36235986333561887, + "grad_norm": 2.8560538291931152, + "learning_rate": 0.00014194798697522858, + "loss": 0.2489, + "step": 15458 + }, + { + "epoch": 0.36238330491042386, + "grad_norm": 1.360153317451477, + "learning_rate": 0.0001419413001238282, + "loss": 0.196, + "step": 15459 + }, + { + "epoch": 0.36240674648522886, + "grad_norm": 4.082691192626953, + "learning_rate": 0.00014193461304485308, + "loss": 0.38, + "step": 15460 + }, + { + "epoch": 0.36243018806003385, + "grad_norm": 3.3649988174438477, + "learning_rate": 0.00014192792573833966, + "loss": 1.7396, + "step": 15461 + }, + { + "epoch": 0.36245362963483885, + "grad_norm": 4.154405117034912, + "learning_rate": 0.00014192123820432412, + "loss": 0.3728, + "step": 15462 + }, + { + "epoch": 0.36247707120964384, + "grad_norm": 4.022579669952393, + "learning_rate": 0.0001419145504428427, + "loss": 1.4385, + "step": 15463 + }, + { + "epoch": 0.36250051278444884, + "grad_norm": 3.876643657684326, + "learning_rate": 0.00014190786245393182, + "loss": 0.3084, + "step": 15464 + }, + { + "epoch": 0.36252395435925383, + "grad_norm": 1.725258469581604, + "learning_rate": 0.00014190117423762769, + "loss": 0.1542, + "step": 15465 + }, + { + "epoch": 0.3625473959340588, + "grad_norm": 2.9457809925079346, + "learning_rate": 0.00014189448579396657, + "loss": 0.1208, + "step": 15466 + }, + { + "epoch": 0.3625708375088638, + "grad_norm": 2.7409636974334717, + "learning_rate": 0.0001418877971229848, + "loss": 0.2063, + "step": 15467 + }, + { + "epoch": 0.3625942790836688, + "grad_norm": 3.5192055702209473, + "learning_rate": 0.00014188110822471872, + "loss": 0.3254, + "step": 15468 + }, + { + "epoch": 0.36261772065847386, + "grad_norm": 2.1516456604003906, + "learning_rate": 0.0001418744190992045, + "loss": 0.217, + "step": 15469 + }, + { + "epoch": 0.36264116223327886, + "grad_norm": 3.475856304168701, + "learning_rate": 0.00014186772974647855, + "loss": 1.5794, + "step": 15470 + }, + { + "epoch": 0.36266460380808385, + "grad_norm": 3.04370379447937, + "learning_rate": 0.0001418610401665771, + "loss": 0.4599, + "step": 15471 + }, + { + "epoch": 0.36268804538288885, + "grad_norm": 4.586609363555908, + "learning_rate": 0.00014185435035953648, + "loss": 0.4769, + "step": 15472 + }, + { + "epoch": 0.36271148695769384, + "grad_norm": 3.1060969829559326, + "learning_rate": 0.00014184766032539296, + "loss": 0.3201, + "step": 15473 + }, + { + "epoch": 0.36273492853249883, + "grad_norm": 0.4140775501728058, + "learning_rate": 0.00014184097006418288, + "loss": 0.042, + "step": 15474 + }, + { + "epoch": 0.36275837010730383, + "grad_norm": 3.5894615650177, + "learning_rate": 0.0001418342795759425, + "loss": 0.3298, + "step": 15475 + }, + { + "epoch": 0.3627818116821088, + "grad_norm": 4.147274971008301, + "learning_rate": 0.00014182758886070818, + "loss": 0.3829, + "step": 15476 + }, + { + "epoch": 0.3628052532569138, + "grad_norm": 2.909675121307373, + "learning_rate": 0.00014182089791851617, + "loss": 0.2718, + "step": 15477 + }, + { + "epoch": 0.3628286948317188, + "grad_norm": 3.6328964233398438, + "learning_rate": 0.0001418142067494028, + "loss": 0.3548, + "step": 15478 + }, + { + "epoch": 0.3628521364065238, + "grad_norm": 3.0588674545288086, + "learning_rate": 0.0001418075153534043, + "loss": 0.3032, + "step": 15479 + }, + { + "epoch": 0.3628755779813288, + "grad_norm": 5.307805061340332, + "learning_rate": 0.00014180082373055713, + "loss": 0.5249, + "step": 15480 + }, + { + "epoch": 0.3628990195561338, + "grad_norm": 5.199309349060059, + "learning_rate": 0.00014179413188089747, + "loss": 0.5369, + "step": 15481 + }, + { + "epoch": 0.3629224611309388, + "grad_norm": 4.626595497131348, + "learning_rate": 0.0001417874398044617, + "loss": 0.2992, + "step": 15482 + }, + { + "epoch": 0.3629459027057438, + "grad_norm": 3.0800106525421143, + "learning_rate": 0.0001417807475012861, + "loss": 0.4074, + "step": 15483 + }, + { + "epoch": 0.3629693442805488, + "grad_norm": 4.15625524520874, + "learning_rate": 0.00014177405497140696, + "loss": 0.3992, + "step": 15484 + }, + { + "epoch": 0.3629927858553538, + "grad_norm": 1.1538373231887817, + "learning_rate": 0.00014176736221486067, + "loss": 0.1142, + "step": 15485 + }, + { + "epoch": 0.36301622743015877, + "grad_norm": 3.8817994594573975, + "learning_rate": 0.00014176066923168347, + "loss": 0.5339, + "step": 15486 + }, + { + "epoch": 0.36303966900496376, + "grad_norm": 2.9289710521698, + "learning_rate": 0.00014175397602191173, + "loss": 1.7267, + "step": 15487 + }, + { + "epoch": 0.36306311057976876, + "grad_norm": 1.113006830215454, + "learning_rate": 0.0001417472825855817, + "loss": 0.1044, + "step": 15488 + }, + { + "epoch": 0.36308655215457375, + "grad_norm": 1.6978954076766968, + "learning_rate": 0.00014174058892272976, + "loss": 0.1408, + "step": 15489 + }, + { + "epoch": 0.36310999372937874, + "grad_norm": 2.836836814880371, + "learning_rate": 0.00014173389503339222, + "loss": 0.3728, + "step": 15490 + }, + { + "epoch": 0.36313343530418374, + "grad_norm": 2.8610191345214844, + "learning_rate": 0.00014172720091760538, + "loss": 0.4409, + "step": 15491 + }, + { + "epoch": 0.36315687687898873, + "grad_norm": 3.641695261001587, + "learning_rate": 0.00014172050657540558, + "loss": 0.3862, + "step": 15492 + }, + { + "epoch": 0.3631803184537937, + "grad_norm": 6.780318260192871, + "learning_rate": 0.00014171381200682917, + "loss": 0.5145, + "step": 15493 + }, + { + "epoch": 0.3632037600285987, + "grad_norm": 1.2671937942504883, + "learning_rate": 0.00014170711721191238, + "loss": 0.0855, + "step": 15494 + }, + { + "epoch": 0.3632272016034037, + "grad_norm": 3.124505043029785, + "learning_rate": 0.00014170042219069164, + "loss": 0.3667, + "step": 15495 + }, + { + "epoch": 0.3632506431782087, + "grad_norm": 1.8904461860656738, + "learning_rate": 0.00014169372694320322, + "loss": 1.2968, + "step": 15496 + }, + { + "epoch": 0.3632740847530137, + "grad_norm": 6.403974533081055, + "learning_rate": 0.0001416870314694835, + "loss": 0.5291, + "step": 15497 + }, + { + "epoch": 0.3632975263278187, + "grad_norm": 3.269115924835205, + "learning_rate": 0.0001416803357695687, + "loss": 0.5239, + "step": 15498 + }, + { + "epoch": 0.3633209679026237, + "grad_norm": 4.131180286407471, + "learning_rate": 0.0001416736398434953, + "loss": 0.4763, + "step": 15499 + }, + { + "epoch": 0.3633444094774287, + "grad_norm": 1.2707239389419556, + "learning_rate": 0.00014166694369129956, + "loss": 0.155, + "step": 15500 + }, + { + "epoch": 0.3633678510522337, + "grad_norm": 3.5279653072357178, + "learning_rate": 0.00014166024731301777, + "loss": 0.2757, + "step": 15501 + }, + { + "epoch": 0.3633912926270387, + "grad_norm": 4.668348789215088, + "learning_rate": 0.0001416535507086863, + "loss": 0.4022, + "step": 15502 + }, + { + "epoch": 0.36341473420184367, + "grad_norm": 3.2019574642181396, + "learning_rate": 0.00014164685387834153, + "loss": 1.5478, + "step": 15503 + }, + { + "epoch": 0.36343817577664866, + "grad_norm": 2.093071937561035, + "learning_rate": 0.00014164015682201973, + "loss": 0.1534, + "step": 15504 + }, + { + "epoch": 0.36346161735145366, + "grad_norm": 2.1444127559661865, + "learning_rate": 0.0001416334595397573, + "loss": 0.1816, + "step": 15505 + }, + { + "epoch": 0.36348505892625865, + "grad_norm": 2.9781947135925293, + "learning_rate": 0.00014162676203159053, + "loss": 0.3932, + "step": 15506 + }, + { + "epoch": 0.36350850050106365, + "grad_norm": 5.233457565307617, + "learning_rate": 0.0001416200642975558, + "loss": 0.6039, + "step": 15507 + }, + { + "epoch": 0.36353194207586864, + "grad_norm": 1.5898035764694214, + "learning_rate": 0.0001416133663376894, + "loss": 0.1498, + "step": 15508 + }, + { + "epoch": 0.36355538365067364, + "grad_norm": 1.7261401414871216, + "learning_rate": 0.0001416066681520277, + "loss": 0.1489, + "step": 15509 + }, + { + "epoch": 0.36357882522547863, + "grad_norm": 3.48152232170105, + "learning_rate": 0.00014159996974060708, + "loss": 0.2738, + "step": 15510 + }, + { + "epoch": 0.3636022668002836, + "grad_norm": 2.6536319255828857, + "learning_rate": 0.00014159327110346384, + "loss": 1.1452, + "step": 15511 + }, + { + "epoch": 0.3636257083750886, + "grad_norm": 2.787916898727417, + "learning_rate": 0.00014158657224063435, + "loss": 0.2571, + "step": 15512 + }, + { + "epoch": 0.3636491499498936, + "grad_norm": 0.5446735620498657, + "learning_rate": 0.00014157987315215493, + "loss": 0.0511, + "step": 15513 + }, + { + "epoch": 0.3636725915246986, + "grad_norm": 3.8042027950286865, + "learning_rate": 0.00014157317383806195, + "loss": 0.2581, + "step": 15514 + }, + { + "epoch": 0.3636960330995036, + "grad_norm": 2.8822245597839355, + "learning_rate": 0.00014156647429839177, + "loss": 0.3143, + "step": 15515 + }, + { + "epoch": 0.3637194746743086, + "grad_norm": 3.174600601196289, + "learning_rate": 0.00014155977453318073, + "loss": 0.2989, + "step": 15516 + }, + { + "epoch": 0.3637429162491136, + "grad_norm": 5.305489540100098, + "learning_rate": 0.00014155307454246518, + "loss": 0.38, + "step": 15517 + }, + { + "epoch": 0.3637663578239186, + "grad_norm": 0.9324700236320496, + "learning_rate": 0.0001415463743262815, + "loss": 0.0953, + "step": 15518 + }, + { + "epoch": 0.3637897993987236, + "grad_norm": 2.087665557861328, + "learning_rate": 0.000141539673884666, + "loss": 0.1774, + "step": 15519 + }, + { + "epoch": 0.36381324097352863, + "grad_norm": 2.9850869178771973, + "learning_rate": 0.00014153297321765505, + "loss": 0.3317, + "step": 15520 + }, + { + "epoch": 0.3638366825483336, + "grad_norm": 3.8163816928863525, + "learning_rate": 0.00014152627232528504, + "loss": 0.3954, + "step": 15521 + }, + { + "epoch": 0.3638601241231386, + "grad_norm": 5.4035964012146, + "learning_rate": 0.00014151957120759235, + "loss": 0.6978, + "step": 15522 + }, + { + "epoch": 0.3638835656979436, + "grad_norm": 0.4366593360900879, + "learning_rate": 0.00014151286986461322, + "loss": 0.0584, + "step": 15523 + }, + { + "epoch": 0.3639070072727486, + "grad_norm": 1.2044017314910889, + "learning_rate": 0.00014150616829638414, + "loss": 0.1142, + "step": 15524 + }, + { + "epoch": 0.3639304488475536, + "grad_norm": 4.2484025955200195, + "learning_rate": 0.00014149946650294143, + "loss": 0.3782, + "step": 15525 + }, + { + "epoch": 0.3639538904223586, + "grad_norm": 1.5038986206054688, + "learning_rate": 0.0001414927644843214, + "loss": 0.1048, + "step": 15526 + }, + { + "epoch": 0.3639773319971636, + "grad_norm": 2.469644784927368, + "learning_rate": 0.0001414860622405605, + "loss": 0.2283, + "step": 15527 + }, + { + "epoch": 0.3640007735719686, + "grad_norm": 3.072031021118164, + "learning_rate": 0.00014147935977169507, + "loss": 0.4534, + "step": 15528 + }, + { + "epoch": 0.3640242151467736, + "grad_norm": 5.222934722900391, + "learning_rate": 0.00014147265707776144, + "loss": 0.5715, + "step": 15529 + }, + { + "epoch": 0.36404765672157857, + "grad_norm": 2.3272438049316406, + "learning_rate": 0.00014146595415879602, + "loss": 1.3314, + "step": 15530 + }, + { + "epoch": 0.36407109829638357, + "grad_norm": 1.705858826637268, + "learning_rate": 0.00014145925101483517, + "loss": 1.1905, + "step": 15531 + }, + { + "epoch": 0.36409453987118856, + "grad_norm": 1.835817575454712, + "learning_rate": 0.00014145254764591524, + "loss": 0.1235, + "step": 15532 + }, + { + "epoch": 0.36411798144599355, + "grad_norm": 0.37686222791671753, + "learning_rate": 0.00014144584405207262, + "loss": 0.0225, + "step": 15533 + }, + { + "epoch": 0.36414142302079855, + "grad_norm": 0.47376665472984314, + "learning_rate": 0.0001414391402333437, + "loss": 0.0531, + "step": 15534 + }, + { + "epoch": 0.36416486459560354, + "grad_norm": 1.3236793279647827, + "learning_rate": 0.00014143243618976485, + "loss": 0.1232, + "step": 15535 + }, + { + "epoch": 0.36418830617040854, + "grad_norm": 2.4869210720062256, + "learning_rate": 0.00014142573192137242, + "loss": 1.0923, + "step": 15536 + }, + { + "epoch": 0.36421174774521353, + "grad_norm": 3.825183391571045, + "learning_rate": 0.00014141902742820283, + "loss": 0.3033, + "step": 15537 + }, + { + "epoch": 0.3642351893200185, + "grad_norm": 5.471068382263184, + "learning_rate": 0.0001414123227102924, + "loss": 0.746, + "step": 15538 + }, + { + "epoch": 0.3642586308948235, + "grad_norm": 4.295717239379883, + "learning_rate": 0.00014140561776767753, + "loss": 0.4184, + "step": 15539 + }, + { + "epoch": 0.3642820724696285, + "grad_norm": 3.691840171813965, + "learning_rate": 0.00014139891260039463, + "loss": 0.2272, + "step": 15540 + }, + { + "epoch": 0.3643055140444335, + "grad_norm": 0.8152182102203369, + "learning_rate": 0.0001413922072084801, + "loss": 0.1052, + "step": 15541 + }, + { + "epoch": 0.3643289556192385, + "grad_norm": 2.3327550888061523, + "learning_rate": 0.00014138550159197027, + "loss": 0.3739, + "step": 15542 + }, + { + "epoch": 0.3643523971940435, + "grad_norm": 3.6221377849578857, + "learning_rate": 0.00014137879575090153, + "loss": 0.3622, + "step": 15543 + }, + { + "epoch": 0.3643758387688485, + "grad_norm": 1.3145086765289307, + "learning_rate": 0.0001413720896853103, + "loss": 0.1228, + "step": 15544 + }, + { + "epoch": 0.3643992803436535, + "grad_norm": 4.516853332519531, + "learning_rate": 0.00014136538339523292, + "loss": 0.2785, + "step": 15545 + }, + { + "epoch": 0.3644227219184585, + "grad_norm": 3.5075886249542236, + "learning_rate": 0.00014135867688070582, + "loss": 0.5076, + "step": 15546 + }, + { + "epoch": 0.3644461634932635, + "grad_norm": 1.532554030418396, + "learning_rate": 0.00014135197014176538, + "loss": 0.0674, + "step": 15547 + }, + { + "epoch": 0.36446960506806847, + "grad_norm": 3.054513692855835, + "learning_rate": 0.000141345263178448, + "loss": 0.2368, + "step": 15548 + }, + { + "epoch": 0.36449304664287346, + "grad_norm": 3.3642125129699707, + "learning_rate": 0.00014133855599079005, + "loss": 0.3102, + "step": 15549 + }, + { + "epoch": 0.36451648821767846, + "grad_norm": 4.160514831542969, + "learning_rate": 0.00014133184857882794, + "loss": 2.6827, + "step": 15550 + }, + { + "epoch": 0.36453992979248345, + "grad_norm": 3.2280633449554443, + "learning_rate": 0.00014132514094259804, + "loss": 0.2639, + "step": 15551 + }, + { + "epoch": 0.36456337136728845, + "grad_norm": 3.7347257137298584, + "learning_rate": 0.00014131843308213675, + "loss": 0.3095, + "step": 15552 + }, + { + "epoch": 0.36458681294209344, + "grad_norm": 3.541965961456299, + "learning_rate": 0.0001413117249974805, + "loss": 0.314, + "step": 15553 + }, + { + "epoch": 0.36461025451689844, + "grad_norm": 1.0221278667449951, + "learning_rate": 0.00014130501668866568, + "loss": 0.0973, + "step": 15554 + }, + { + "epoch": 0.36463369609170343, + "grad_norm": 10.11745548248291, + "learning_rate": 0.00014129830815572868, + "loss": 0.6848, + "step": 15555 + }, + { + "epoch": 0.3646571376665084, + "grad_norm": 4.477929592132568, + "learning_rate": 0.0001412915993987059, + "loss": 0.5671, + "step": 15556 + }, + { + "epoch": 0.3646805792413134, + "grad_norm": 5.329981803894043, + "learning_rate": 0.00014128489041763368, + "loss": 0.8674, + "step": 15557 + }, + { + "epoch": 0.3647040208161184, + "grad_norm": 2.600037097930908, + "learning_rate": 0.00014127818121254855, + "loss": 0.3406, + "step": 15558 + }, + { + "epoch": 0.3647274623909234, + "grad_norm": 2.072537422180176, + "learning_rate": 0.00014127147178348684, + "loss": 0.205, + "step": 15559 + }, + { + "epoch": 0.3647509039657284, + "grad_norm": 0.5115452408790588, + "learning_rate": 0.00014126476213048494, + "loss": 0.021, + "step": 15560 + }, + { + "epoch": 0.3647743455405334, + "grad_norm": 3.6686182022094727, + "learning_rate": 0.0001412580522535793, + "loss": 0.4096, + "step": 15561 + }, + { + "epoch": 0.3647977871153384, + "grad_norm": 7.192116737365723, + "learning_rate": 0.00014125134215280628, + "loss": 0.7739, + "step": 15562 + }, + { + "epoch": 0.3648212286901434, + "grad_norm": 1.1978694200515747, + "learning_rate": 0.00014124463182820237, + "loss": 0.1209, + "step": 15563 + }, + { + "epoch": 0.3648446702649484, + "grad_norm": 2.1041133403778076, + "learning_rate": 0.00014123792127980389, + "loss": 0.1894, + "step": 15564 + }, + { + "epoch": 0.3648681118397534, + "grad_norm": 2.461428642272949, + "learning_rate": 0.00014123121050764727, + "loss": 0.2335, + "step": 15565 + }, + { + "epoch": 0.36489155341455837, + "grad_norm": 4.171176433563232, + "learning_rate": 0.00014122449951176899, + "loss": 0.4921, + "step": 15566 + }, + { + "epoch": 0.36491499498936336, + "grad_norm": 2.0921616554260254, + "learning_rate": 0.00014121778829220537, + "loss": 0.1476, + "step": 15567 + }, + { + "epoch": 0.36493843656416836, + "grad_norm": 4.0112409591674805, + "learning_rate": 0.0001412110768489929, + "loss": 0.5038, + "step": 15568 + }, + { + "epoch": 0.36496187813897335, + "grad_norm": 5.867026329040527, + "learning_rate": 0.00014120436518216798, + "loss": 0.1376, + "step": 15569 + }, + { + "epoch": 0.36498531971377834, + "grad_norm": 2.5924692153930664, + "learning_rate": 0.000141197653291767, + "loss": 0.2069, + "step": 15570 + }, + { + "epoch": 0.3650087612885834, + "grad_norm": 1.6461511850357056, + "learning_rate": 0.0001411909411778264, + "loss": 0.1636, + "step": 15571 + }, + { + "epoch": 0.3650322028633884, + "grad_norm": 0.40584397315979004, + "learning_rate": 0.00014118422884038258, + "loss": 0.0244, + "step": 15572 + }, + { + "epoch": 0.3650556444381934, + "grad_norm": 3.7302939891815186, + "learning_rate": 0.000141177516279472, + "loss": 0.3326, + "step": 15573 + }, + { + "epoch": 0.3650790860129984, + "grad_norm": 2.4337575435638428, + "learning_rate": 0.00014117080349513103, + "loss": 0.1609, + "step": 15574 + }, + { + "epoch": 0.36510252758780337, + "grad_norm": 4.036574363708496, + "learning_rate": 0.00014116409048739613, + "loss": 0.3756, + "step": 15575 + }, + { + "epoch": 0.36512596916260837, + "grad_norm": 1.3616043329238892, + "learning_rate": 0.00014115737725630374, + "loss": 0.1284, + "step": 15576 + }, + { + "epoch": 0.36514941073741336, + "grad_norm": 2.88346004486084, + "learning_rate": 0.00014115066380189023, + "loss": 0.2228, + "step": 15577 + }, + { + "epoch": 0.36517285231221835, + "grad_norm": 5.971189022064209, + "learning_rate": 0.0001411439501241921, + "loss": 0.4521, + "step": 15578 + }, + { + "epoch": 0.36519629388702335, + "grad_norm": 2.4361507892608643, + "learning_rate": 0.00014113723622324572, + "loss": 0.2603, + "step": 15579 + }, + { + "epoch": 0.36521973546182834, + "grad_norm": 2.4751434326171875, + "learning_rate": 0.00014113052209908754, + "loss": 1.5581, + "step": 15580 + }, + { + "epoch": 0.36524317703663334, + "grad_norm": 4.070769309997559, + "learning_rate": 0.000141123807751754, + "loss": 0.3556, + "step": 15581 + }, + { + "epoch": 0.36526661861143833, + "grad_norm": 3.999987840652466, + "learning_rate": 0.0001411170931812815, + "loss": 0.2423, + "step": 15582 + }, + { + "epoch": 0.3652900601862433, + "grad_norm": 4.55672550201416, + "learning_rate": 0.00014111037838770654, + "loss": 0.5028, + "step": 15583 + }, + { + "epoch": 0.3653135017610483, + "grad_norm": 1.8131835460662842, + "learning_rate": 0.00014110366337106544, + "loss": 0.1429, + "step": 15584 + }, + { + "epoch": 0.3653369433358533, + "grad_norm": 1.3194822072982788, + "learning_rate": 0.00014109694813139477, + "loss": 0.1334, + "step": 15585 + }, + { + "epoch": 0.3653603849106583, + "grad_norm": 3.0159952640533447, + "learning_rate": 0.00014109023266873087, + "loss": 0.2079, + "step": 15586 + }, + { + "epoch": 0.3653838264854633, + "grad_norm": 2.9384942054748535, + "learning_rate": 0.00014108351698311023, + "loss": 0.3868, + "step": 15587 + }, + { + "epoch": 0.3654072680602683, + "grad_norm": 2.530501127243042, + "learning_rate": 0.00014107680107456927, + "loss": 0.3043, + "step": 15588 + }, + { + "epoch": 0.3654307096350733, + "grad_norm": 5.086691856384277, + "learning_rate": 0.00014107008494314443, + "loss": 0.4224, + "step": 15589 + }, + { + "epoch": 0.3654541512098783, + "grad_norm": 3.119068145751953, + "learning_rate": 0.00014106336858887215, + "loss": 0.2583, + "step": 15590 + }, + { + "epoch": 0.3654775927846833, + "grad_norm": 2.085549831390381, + "learning_rate": 0.00014105665201178888, + "loss": 1.5198, + "step": 15591 + }, + { + "epoch": 0.3655010343594883, + "grad_norm": 3.4902124404907227, + "learning_rate": 0.0001410499352119311, + "loss": 0.4885, + "step": 15592 + }, + { + "epoch": 0.36552447593429327, + "grad_norm": 3.3173890113830566, + "learning_rate": 0.00014104321818933515, + "loss": 0.4064, + "step": 15593 + }, + { + "epoch": 0.36554791750909826, + "grad_norm": 2.7308075428009033, + "learning_rate": 0.00014103650094403756, + "loss": 0.2147, + "step": 15594 + }, + { + "epoch": 0.36557135908390326, + "grad_norm": 4.907716751098633, + "learning_rate": 0.0001410297834760748, + "loss": 0.2075, + "step": 15595 + }, + { + "epoch": 0.36559480065870825, + "grad_norm": 1.2220319509506226, + "learning_rate": 0.00014102306578548323, + "loss": 0.1743, + "step": 15596 + }, + { + "epoch": 0.36561824223351325, + "grad_norm": 4.926642894744873, + "learning_rate": 0.00014101634787229937, + "loss": 0.2776, + "step": 15597 + }, + { + "epoch": 0.36564168380831824, + "grad_norm": 1.7474355697631836, + "learning_rate": 0.00014100962973655968, + "loss": 0.1914, + "step": 15598 + }, + { + "epoch": 0.36566512538312324, + "grad_norm": 1.3170491456985474, + "learning_rate": 0.00014100291137830056, + "loss": 0.161, + "step": 15599 + }, + { + "epoch": 0.36568856695792823, + "grad_norm": 2.810243606567383, + "learning_rate": 0.0001409961927975585, + "loss": 0.4079, + "step": 15600 + }, + { + "epoch": 0.3657120085327332, + "grad_norm": 5.610962390899658, + "learning_rate": 0.00014098947399436993, + "loss": 0.6661, + "step": 15601 + }, + { + "epoch": 0.3657354501075382, + "grad_norm": 1.4907015562057495, + "learning_rate": 0.00014098275496877136, + "loss": 0.1949, + "step": 15602 + }, + { + "epoch": 0.3657588916823432, + "grad_norm": 3.590050220489502, + "learning_rate": 0.00014097603572079916, + "loss": 0.39, + "step": 15603 + }, + { + "epoch": 0.3657823332571482, + "grad_norm": 4.661703586578369, + "learning_rate": 0.0001409693162504899, + "loss": 0.6047, + "step": 15604 + }, + { + "epoch": 0.3658057748319532, + "grad_norm": 4.5015058517456055, + "learning_rate": 0.0001409625965578799, + "loss": 0.5604, + "step": 15605 + }, + { + "epoch": 0.3658292164067582, + "grad_norm": 1.029285192489624, + "learning_rate": 0.00014095587664300573, + "loss": 0.0685, + "step": 15606 + }, + { + "epoch": 0.3658526579815632, + "grad_norm": 0.5266308188438416, + "learning_rate": 0.00014094915650590383, + "loss": 0.0401, + "step": 15607 + }, + { + "epoch": 0.3658760995563682, + "grad_norm": 1.5732649564743042, + "learning_rate": 0.00014094243614661066, + "loss": 0.0764, + "step": 15608 + }, + { + "epoch": 0.3658995411311732, + "grad_norm": 5.013430595397949, + "learning_rate": 0.00014093571556516267, + "loss": 0.5302, + "step": 15609 + }, + { + "epoch": 0.3659229827059782, + "grad_norm": 4.012575149536133, + "learning_rate": 0.00014092899476159635, + "loss": 0.442, + "step": 15610 + }, + { + "epoch": 0.36594642428078317, + "grad_norm": 3.958378314971924, + "learning_rate": 0.00014092227373594816, + "loss": 0.3069, + "step": 15611 + }, + { + "epoch": 0.36596986585558816, + "grad_norm": 2.2766318321228027, + "learning_rate": 0.00014091555248825453, + "loss": 0.1893, + "step": 15612 + }, + { + "epoch": 0.36599330743039316, + "grad_norm": 0.7263758778572083, + "learning_rate": 0.00014090883101855196, + "loss": 0.1012, + "step": 15613 + }, + { + "epoch": 0.36601674900519815, + "grad_norm": 3.925891399383545, + "learning_rate": 0.00014090210932687697, + "loss": 0.3332, + "step": 15614 + }, + { + "epoch": 0.36604019058000314, + "grad_norm": 1.5498569011688232, + "learning_rate": 0.00014089538741326595, + "loss": 0.1945, + "step": 15615 + }, + { + "epoch": 0.36606363215480814, + "grad_norm": 2.866711139678955, + "learning_rate": 0.00014088866527775542, + "loss": 0.2403, + "step": 15616 + }, + { + "epoch": 0.36608707372961313, + "grad_norm": 3.2397375106811523, + "learning_rate": 0.00014088194292038187, + "loss": 0.3924, + "step": 15617 + }, + { + "epoch": 0.3661105153044181, + "grad_norm": 3.177999496459961, + "learning_rate": 0.00014087522034118168, + "loss": 0.5484, + "step": 15618 + }, + { + "epoch": 0.3661339568792231, + "grad_norm": 4.989599704742432, + "learning_rate": 0.0001408684975401914, + "loss": 0.4609, + "step": 15619 + }, + { + "epoch": 0.3661573984540281, + "grad_norm": 2.7055394649505615, + "learning_rate": 0.00014086177451744756, + "loss": 0.2378, + "step": 15620 + }, + { + "epoch": 0.3661808400288331, + "grad_norm": 1.6339993476867676, + "learning_rate": 0.00014085505127298657, + "loss": 0.2016, + "step": 15621 + }, + { + "epoch": 0.36620428160363816, + "grad_norm": 3.7474396228790283, + "learning_rate": 0.0001408483278068449, + "loss": 0.4161, + "step": 15622 + }, + { + "epoch": 0.36622772317844315, + "grad_norm": 4.365344047546387, + "learning_rate": 0.0001408416041190591, + "loss": 0.5198, + "step": 15623 + }, + { + "epoch": 0.36625116475324815, + "grad_norm": 0.9931511878967285, + "learning_rate": 0.00014083488020966554, + "loss": 0.1431, + "step": 15624 + }, + { + "epoch": 0.36627460632805314, + "grad_norm": 4.624195575714111, + "learning_rate": 0.0001408281560787008, + "loss": 0.2776, + "step": 15625 + }, + { + "epoch": 0.36629804790285814, + "grad_norm": 3.2834160327911377, + "learning_rate": 0.00014082143172620135, + "loss": 0.3465, + "step": 15626 + }, + { + "epoch": 0.36632148947766313, + "grad_norm": 2.869896650314331, + "learning_rate": 0.00014081470715220367, + "loss": 0.4169, + "step": 15627 + }, + { + "epoch": 0.3663449310524681, + "grad_norm": 2.254667043685913, + "learning_rate": 0.0001408079823567442, + "loss": 1.2148, + "step": 15628 + }, + { + "epoch": 0.3663683726272731, + "grad_norm": 3.466043710708618, + "learning_rate": 0.00014080125733985951, + "loss": 0.3333, + "step": 15629 + }, + { + "epoch": 0.3663918142020781, + "grad_norm": 3.5218396186828613, + "learning_rate": 0.000140794532101586, + "loss": 0.2323, + "step": 15630 + }, + { + "epoch": 0.3664152557768831, + "grad_norm": 0.7073655724525452, + "learning_rate": 0.00014078780664196025, + "loss": 0.0396, + "step": 15631 + }, + { + "epoch": 0.3664386973516881, + "grad_norm": 3.6429433822631836, + "learning_rate": 0.00014078108096101874, + "loss": 0.539, + "step": 15632 + }, + { + "epoch": 0.3664621389264931, + "grad_norm": 4.496579647064209, + "learning_rate": 0.00014077435505879793, + "loss": 0.5224, + "step": 15633 + }, + { + "epoch": 0.3664855805012981, + "grad_norm": 3.4320385456085205, + "learning_rate": 0.0001407676289353343, + "loss": 0.4182, + "step": 15634 + }, + { + "epoch": 0.3665090220761031, + "grad_norm": 0.9444901943206787, + "learning_rate": 0.00014076090259066437, + "loss": 0.0825, + "step": 15635 + }, + { + "epoch": 0.3665324636509081, + "grad_norm": 3.0433945655822754, + "learning_rate": 0.00014075417602482461, + "loss": 0.3158, + "step": 15636 + }, + { + "epoch": 0.3665559052257131, + "grad_norm": 0.5495057702064514, + "learning_rate": 0.0001407474492378516, + "loss": 0.0366, + "step": 15637 + }, + { + "epoch": 0.36657934680051807, + "grad_norm": 2.6240882873535156, + "learning_rate": 0.00014074072222978174, + "loss": 0.3065, + "step": 15638 + }, + { + "epoch": 0.36660278837532306, + "grad_norm": 4.987399101257324, + "learning_rate": 0.0001407339950006516, + "loss": 0.3971, + "step": 15639 + }, + { + "epoch": 0.36662622995012806, + "grad_norm": 3.2057693004608154, + "learning_rate": 0.00014072726755049768, + "loss": 0.3422, + "step": 15640 + }, + { + "epoch": 0.36664967152493305, + "grad_norm": 1.924849510192871, + "learning_rate": 0.0001407205398793564, + "loss": 0.1207, + "step": 15641 + }, + { + "epoch": 0.36667311309973805, + "grad_norm": 2.8736023902893066, + "learning_rate": 0.00014071381198726437, + "loss": 1.9572, + "step": 15642 + }, + { + "epoch": 0.36669655467454304, + "grad_norm": 2.910837411880493, + "learning_rate": 0.00014070708387425805, + "loss": 0.4058, + "step": 15643 + }, + { + "epoch": 0.36671999624934803, + "grad_norm": 3.829586982727051, + "learning_rate": 0.0001407003555403739, + "loss": 0.452, + "step": 15644 + }, + { + "epoch": 0.36674343782415303, + "grad_norm": 2.9964849948883057, + "learning_rate": 0.00014069362698564854, + "loss": 0.2685, + "step": 15645 + }, + { + "epoch": 0.366766879398958, + "grad_norm": 3.7715258598327637, + "learning_rate": 0.0001406868982101184, + "loss": 0.4957, + "step": 15646 + }, + { + "epoch": 0.366790320973763, + "grad_norm": 3.2581324577331543, + "learning_rate": 0.00014068016921382, + "loss": 0.4132, + "step": 15647 + }, + { + "epoch": 0.366813762548568, + "grad_norm": 0.5041139125823975, + "learning_rate": 0.00014067343999678983, + "loss": 0.0695, + "step": 15648 + }, + { + "epoch": 0.366837204123373, + "grad_norm": 4.689444065093994, + "learning_rate": 0.00014066671055906445, + "loss": 0.5608, + "step": 15649 + }, + { + "epoch": 0.366860645698178, + "grad_norm": 5.415050506591797, + "learning_rate": 0.00014065998090068038, + "loss": 0.4311, + "step": 15650 + }, + { + "epoch": 0.366884087272983, + "grad_norm": 2.856060266494751, + "learning_rate": 0.0001406532510216741, + "loss": 0.257, + "step": 15651 + }, + { + "epoch": 0.366907528847788, + "grad_norm": 4.3202314376831055, + "learning_rate": 0.00014064652092208213, + "loss": 0.4016, + "step": 15652 + }, + { + "epoch": 0.366930970422593, + "grad_norm": 3.6082043647766113, + "learning_rate": 0.000140639790601941, + "loss": 0.2911, + "step": 15653 + }, + { + "epoch": 0.366954411997398, + "grad_norm": 3.851018190383911, + "learning_rate": 0.0001406330600612872, + "loss": 0.4261, + "step": 15654 + }, + { + "epoch": 0.36697785357220297, + "grad_norm": 3.0288209915161133, + "learning_rate": 0.0001406263293001573, + "loss": 0.3046, + "step": 15655 + }, + { + "epoch": 0.36700129514700797, + "grad_norm": 0.5109379291534424, + "learning_rate": 0.0001406195983185878, + "loss": 0.0589, + "step": 15656 + }, + { + "epoch": 0.36702473672181296, + "grad_norm": 4.509860992431641, + "learning_rate": 0.0001406128671166152, + "loss": 0.6973, + "step": 15657 + }, + { + "epoch": 0.36704817829661796, + "grad_norm": 2.0796828269958496, + "learning_rate": 0.00014060613569427604, + "loss": 0.166, + "step": 15658 + }, + { + "epoch": 0.36707161987142295, + "grad_norm": 2.22094988822937, + "learning_rate": 0.00014059940405160686, + "loss": 1.7857, + "step": 15659 + }, + { + "epoch": 0.36709506144622794, + "grad_norm": 3.3929271697998047, + "learning_rate": 0.00014059267218864411, + "loss": 1.6131, + "step": 15660 + }, + { + "epoch": 0.36711850302103294, + "grad_norm": 1.6412148475646973, + "learning_rate": 0.00014058594010542445, + "loss": 0.1862, + "step": 15661 + }, + { + "epoch": 0.36714194459583793, + "grad_norm": 2.5477776527404785, + "learning_rate": 0.00014057920780198433, + "loss": 0.2806, + "step": 15662 + }, + { + "epoch": 0.3671653861706429, + "grad_norm": 3.987126111984253, + "learning_rate": 0.00014057247527836026, + "loss": 1.1896, + "step": 15663 + }, + { + "epoch": 0.3671888277454479, + "grad_norm": 3.7592520713806152, + "learning_rate": 0.00014056574253458882, + "loss": 0.2798, + "step": 15664 + }, + { + "epoch": 0.3672122693202529, + "grad_norm": 3.90079665184021, + "learning_rate": 0.00014055900957070654, + "loss": 0.4238, + "step": 15665 + }, + { + "epoch": 0.3672357108950579, + "grad_norm": 2.2204980850219727, + "learning_rate": 0.0001405522763867499, + "loss": 1.8965, + "step": 15666 + }, + { + "epoch": 0.3672591524698629, + "grad_norm": 4.049806594848633, + "learning_rate": 0.00014054554298275547, + "loss": 0.5587, + "step": 15667 + }, + { + "epoch": 0.3672825940446679, + "grad_norm": 4.1322150230407715, + "learning_rate": 0.0001405388093587598, + "loss": 0.5143, + "step": 15668 + }, + { + "epoch": 0.3673060356194729, + "grad_norm": 3.3041326999664307, + "learning_rate": 0.0001405320755147994, + "loss": 0.2916, + "step": 15669 + }, + { + "epoch": 0.3673294771942779, + "grad_norm": 0.7491956353187561, + "learning_rate": 0.0001405253414509108, + "loss": 0.0804, + "step": 15670 + }, + { + "epoch": 0.3673529187690829, + "grad_norm": 1.3397287130355835, + "learning_rate": 0.00014051860716713058, + "loss": 0.098, + "step": 15671 + }, + { + "epoch": 0.3673763603438879, + "grad_norm": 4.847414970397949, + "learning_rate": 0.00014051187266349527, + "loss": 0.3534, + "step": 15672 + }, + { + "epoch": 0.3673998019186929, + "grad_norm": 2.0705881118774414, + "learning_rate": 0.00014050513794004136, + "loss": 1.0433, + "step": 15673 + }, + { + "epoch": 0.3674232434934979, + "grad_norm": 4.159489631652832, + "learning_rate": 0.0001404984029968055, + "loss": 0.3539, + "step": 15674 + }, + { + "epoch": 0.3674466850683029, + "grad_norm": 1.33521568775177, + "learning_rate": 0.00014049166783382412, + "loss": 0.2325, + "step": 15675 + }, + { + "epoch": 0.3674701266431079, + "grad_norm": 2.0869333744049072, + "learning_rate": 0.0001404849324511338, + "loss": 0.2428, + "step": 15676 + }, + { + "epoch": 0.3674935682179129, + "grad_norm": 2.4999961853027344, + "learning_rate": 0.00014047819684877112, + "loss": 1.5045, + "step": 15677 + }, + { + "epoch": 0.3675170097927179, + "grad_norm": 3.663086414337158, + "learning_rate": 0.00014047146102677262, + "loss": 0.297, + "step": 15678 + }, + { + "epoch": 0.3675404513675229, + "grad_norm": 1.1346180438995361, + "learning_rate": 0.0001404647249851748, + "loss": 0.0884, + "step": 15679 + }, + { + "epoch": 0.3675638929423279, + "grad_norm": 3.4719138145446777, + "learning_rate": 0.00014045798872401426, + "loss": 0.2537, + "step": 15680 + }, + { + "epoch": 0.3675873345171329, + "grad_norm": 0.6846190094947815, + "learning_rate": 0.00014045125224332753, + "loss": 0.0573, + "step": 15681 + }, + { + "epoch": 0.3676107760919379, + "grad_norm": 0.8428004384040833, + "learning_rate": 0.00014044451554315118, + "loss": 0.1003, + "step": 15682 + }, + { + "epoch": 0.36763421766674287, + "grad_norm": 3.0333971977233887, + "learning_rate": 0.00014043777862352173, + "loss": 0.3111, + "step": 15683 + }, + { + "epoch": 0.36765765924154786, + "grad_norm": 0.8383353352546692, + "learning_rate": 0.0001404310414844758, + "loss": 0.1217, + "step": 15684 + }, + { + "epoch": 0.36768110081635286, + "grad_norm": 0.7824094295501709, + "learning_rate": 0.00014042430412604983, + "loss": 0.0972, + "step": 15685 + }, + { + "epoch": 0.36770454239115785, + "grad_norm": 4.490338325500488, + "learning_rate": 0.0001404175665482805, + "loss": 0.552, + "step": 15686 + }, + { + "epoch": 0.36772798396596285, + "grad_norm": 7.720902442932129, + "learning_rate": 0.00014041082875120432, + "loss": 0.7439, + "step": 15687 + }, + { + "epoch": 0.36775142554076784, + "grad_norm": 4.343069553375244, + "learning_rate": 0.00014040409073485784, + "loss": 0.649, + "step": 15688 + }, + { + "epoch": 0.36777486711557283, + "grad_norm": 2.510352849960327, + "learning_rate": 0.0001403973524992776, + "loss": 0.1838, + "step": 15689 + }, + { + "epoch": 0.36779830869037783, + "grad_norm": 3.7369017601013184, + "learning_rate": 0.00014039061404450025, + "loss": 0.4638, + "step": 15690 + }, + { + "epoch": 0.3678217502651828, + "grad_norm": 3.1260087490081787, + "learning_rate": 0.00014038387537056222, + "loss": 0.3532, + "step": 15691 + }, + { + "epoch": 0.3678451918399878, + "grad_norm": 3.110318422317505, + "learning_rate": 0.0001403771364775002, + "loss": 0.257, + "step": 15692 + }, + { + "epoch": 0.3678686334147928, + "grad_norm": 3.4949662685394287, + "learning_rate": 0.00014037039736535068, + "loss": 0.2404, + "step": 15693 + }, + { + "epoch": 0.3678920749895978, + "grad_norm": 1.7238165140151978, + "learning_rate": 0.0001403636580341503, + "loss": 0.0734, + "step": 15694 + }, + { + "epoch": 0.3679155165644028, + "grad_norm": 3.4112560749053955, + "learning_rate": 0.0001403569184839355, + "loss": 0.3671, + "step": 15695 + }, + { + "epoch": 0.3679389581392078, + "grad_norm": 4.443032264709473, + "learning_rate": 0.00014035017871474295, + "loss": 0.3426, + "step": 15696 + }, + { + "epoch": 0.3679623997140128, + "grad_norm": 2.2077739238739014, + "learning_rate": 0.0001403434387266092, + "loss": 0.1533, + "step": 15697 + }, + { + "epoch": 0.3679858412888178, + "grad_norm": 2.8090081214904785, + "learning_rate": 0.00014033669851957082, + "loss": 0.242, + "step": 15698 + }, + { + "epoch": 0.3680092828636228, + "grad_norm": 4.971368312835693, + "learning_rate": 0.00014032995809366437, + "loss": 0.8024, + "step": 15699 + }, + { + "epoch": 0.36803272443842777, + "grad_norm": 2.305650234222412, + "learning_rate": 0.00014032321744892644, + "loss": 0.2097, + "step": 15700 + }, + { + "epoch": 0.36805616601323277, + "grad_norm": 3.0250751972198486, + "learning_rate": 0.00014031647658539363, + "loss": 1.6205, + "step": 15701 + }, + { + "epoch": 0.36807960758803776, + "grad_norm": 3.6666767597198486, + "learning_rate": 0.00014030973550310243, + "loss": 0.7049, + "step": 15702 + }, + { + "epoch": 0.36810304916284275, + "grad_norm": 2.175417423248291, + "learning_rate": 0.0001403029942020895, + "loss": 0.1137, + "step": 15703 + }, + { + "epoch": 0.36812649073764775, + "grad_norm": 2.089146137237549, + "learning_rate": 0.00014029625268239138, + "loss": 0.1432, + "step": 15704 + }, + { + "epoch": 0.36814993231245274, + "grad_norm": 3.6120874881744385, + "learning_rate": 0.00014028951094404468, + "loss": 0.2608, + "step": 15705 + }, + { + "epoch": 0.36817337388725774, + "grad_norm": 4.029720306396484, + "learning_rate": 0.00014028276898708593, + "loss": 0.5591, + "step": 15706 + }, + { + "epoch": 0.36819681546206273, + "grad_norm": 3.739069700241089, + "learning_rate": 0.00014027602681155175, + "loss": 0.3954, + "step": 15707 + }, + { + "epoch": 0.3682202570368677, + "grad_norm": 3.4962241649627686, + "learning_rate": 0.00014026928441747874, + "loss": 0.4357, + "step": 15708 + }, + { + "epoch": 0.3682436986116727, + "grad_norm": 0.7618386149406433, + "learning_rate": 0.00014026254180490345, + "loss": 0.1063, + "step": 15709 + }, + { + "epoch": 0.3682671401864777, + "grad_norm": 1.1349477767944336, + "learning_rate": 0.00014025579897386248, + "loss": 0.1446, + "step": 15710 + }, + { + "epoch": 0.3682905817612827, + "grad_norm": 4.307143211364746, + "learning_rate": 0.0001402490559243924, + "loss": 0.4453, + "step": 15711 + }, + { + "epoch": 0.3683140233360877, + "grad_norm": 1.187818169593811, + "learning_rate": 0.00014024231265652982, + "loss": 0.1288, + "step": 15712 + }, + { + "epoch": 0.3683374649108927, + "grad_norm": 3.691227674484253, + "learning_rate": 0.00014023556917031133, + "loss": 0.3542, + "step": 15713 + }, + { + "epoch": 0.3683609064856977, + "grad_norm": 0.841223955154419, + "learning_rate": 0.0001402288254657735, + "loss": 0.0707, + "step": 15714 + }, + { + "epoch": 0.3683843480605027, + "grad_norm": 4.4238152503967285, + "learning_rate": 0.0001402220815429529, + "loss": 0.4391, + "step": 15715 + }, + { + "epoch": 0.3684077896353077, + "grad_norm": 2.2628631591796875, + "learning_rate": 0.0001402153374018862, + "loss": 1.5404, + "step": 15716 + }, + { + "epoch": 0.3684312312101127, + "grad_norm": 1.209719181060791, + "learning_rate": 0.0001402085930426099, + "loss": 0.132, + "step": 15717 + }, + { + "epoch": 0.36845467278491767, + "grad_norm": 2.1920371055603027, + "learning_rate": 0.00014020184846516071, + "loss": 0.2169, + "step": 15718 + }, + { + "epoch": 0.36847811435972266, + "grad_norm": 0.7707435488700867, + "learning_rate": 0.00014019510366957514, + "loss": 0.0948, + "step": 15719 + }, + { + "epoch": 0.36850155593452766, + "grad_norm": 1.521957516670227, + "learning_rate": 0.00014018835865588978, + "loss": 0.0685, + "step": 15720 + }, + { + "epoch": 0.36852499750933265, + "grad_norm": 4.330491065979004, + "learning_rate": 0.00014018161342414125, + "loss": 0.3463, + "step": 15721 + }, + { + "epoch": 0.36854843908413765, + "grad_norm": 4.025115966796875, + "learning_rate": 0.0001401748679743662, + "loss": 0.5827, + "step": 15722 + }, + { + "epoch": 0.36857188065894264, + "grad_norm": 2.9702322483062744, + "learning_rate": 0.00014016812230660116, + "loss": 1.6488, + "step": 15723 + }, + { + "epoch": 0.3685953222337477, + "grad_norm": 4.630373954772949, + "learning_rate": 0.00014016137642088275, + "loss": 0.3306, + "step": 15724 + }, + { + "epoch": 0.3686187638085527, + "grad_norm": 3.3653202056884766, + "learning_rate": 0.00014015463031724762, + "loss": 0.384, + "step": 15725 + }, + { + "epoch": 0.3686422053833577, + "grad_norm": 4.310866355895996, + "learning_rate": 0.0001401478839957323, + "loss": 0.4746, + "step": 15726 + }, + { + "epoch": 0.3686656469581627, + "grad_norm": 2.9032652378082275, + "learning_rate": 0.00014014113745637342, + "loss": 0.2304, + "step": 15727 + }, + { + "epoch": 0.36868908853296767, + "grad_norm": 5.1757121086120605, + "learning_rate": 0.00014013439069920762, + "loss": 0.3206, + "step": 15728 + }, + { + "epoch": 0.36871253010777266, + "grad_norm": 1.0153636932373047, + "learning_rate": 0.0001401276437242715, + "loss": 0.0881, + "step": 15729 + }, + { + "epoch": 0.36873597168257766, + "grad_norm": 1.4567240476608276, + "learning_rate": 0.00014012089653160162, + "loss": 0.1803, + "step": 15730 + }, + { + "epoch": 0.36875941325738265, + "grad_norm": 2.9631736278533936, + "learning_rate": 0.0001401141491212347, + "loss": 1.9466, + "step": 15731 + }, + { + "epoch": 0.36878285483218765, + "grad_norm": 0.3914017379283905, + "learning_rate": 0.00014010740149320722, + "loss": 0.0326, + "step": 15732 + }, + { + "epoch": 0.36880629640699264, + "grad_norm": 1.670737624168396, + "learning_rate": 0.00014010065364755583, + "loss": 0.2213, + "step": 15733 + }, + { + "epoch": 0.36882973798179763, + "grad_norm": 4.415823459625244, + "learning_rate": 0.0001400939055843172, + "loss": 0.3894, + "step": 15734 + }, + { + "epoch": 0.36885317955660263, + "grad_norm": 1.5225974321365356, + "learning_rate": 0.00014008715730352793, + "loss": 0.7233, + "step": 15735 + }, + { + "epoch": 0.3688766211314076, + "grad_norm": 3.513486862182617, + "learning_rate": 0.0001400804088052246, + "loss": 0.2003, + "step": 15736 + }, + { + "epoch": 0.3689000627062126, + "grad_norm": 4.44896936416626, + "learning_rate": 0.00014007366008944385, + "loss": 0.5258, + "step": 15737 + }, + { + "epoch": 0.3689235042810176, + "grad_norm": 1.8220802545547485, + "learning_rate": 0.00014006691115622227, + "loss": 0.2387, + "step": 15738 + }, + { + "epoch": 0.3689469458558226, + "grad_norm": 2.6952357292175293, + "learning_rate": 0.00014006016200559653, + "loss": 0.213, + "step": 15739 + }, + { + "epoch": 0.3689703874306276, + "grad_norm": 3.1247143745422363, + "learning_rate": 0.00014005341263760319, + "loss": 0.3298, + "step": 15740 + }, + { + "epoch": 0.3689938290054326, + "grad_norm": 0.7187239527702332, + "learning_rate": 0.00014004666305227894, + "loss": 0.0319, + "step": 15741 + }, + { + "epoch": 0.3690172705802376, + "grad_norm": 1.4851092100143433, + "learning_rate": 0.00014003991324966036, + "loss": 0.143, + "step": 15742 + }, + { + "epoch": 0.3690407121550426, + "grad_norm": 4.07187032699585, + "learning_rate": 0.0001400331632297841, + "loss": 0.4723, + "step": 15743 + }, + { + "epoch": 0.3690641537298476, + "grad_norm": 4.084564685821533, + "learning_rate": 0.00014002641299268675, + "loss": 0.6396, + "step": 15744 + }, + { + "epoch": 0.36908759530465257, + "grad_norm": 2.4984822273254395, + "learning_rate": 0.00014001966253840496, + "loss": 0.1153, + "step": 15745 + }, + { + "epoch": 0.36911103687945757, + "grad_norm": 5.361207008361816, + "learning_rate": 0.00014001291186697536, + "loss": 0.4979, + "step": 15746 + }, + { + "epoch": 0.36913447845426256, + "grad_norm": 6.111475944519043, + "learning_rate": 0.00014000616097843458, + "loss": 0.6535, + "step": 15747 + }, + { + "epoch": 0.36915792002906755, + "grad_norm": 3.600515842437744, + "learning_rate": 0.00013999940987281925, + "loss": 0.328, + "step": 15748 + }, + { + "epoch": 0.36918136160387255, + "grad_norm": 1.3902026414871216, + "learning_rate": 0.000139992658550166, + "loss": 0.1284, + "step": 15749 + }, + { + "epoch": 0.36920480317867754, + "grad_norm": 0.7991355657577515, + "learning_rate": 0.00013998590701051145, + "loss": 0.0988, + "step": 15750 + }, + { + "epoch": 0.36922824475348254, + "grad_norm": 4.942935943603516, + "learning_rate": 0.00013997915525389224, + "loss": 0.391, + "step": 15751 + }, + { + "epoch": 0.36925168632828753, + "grad_norm": 3.294459819793701, + "learning_rate": 0.00013997240328034497, + "loss": 0.4658, + "step": 15752 + }, + { + "epoch": 0.3692751279030925, + "grad_norm": 0.4810943007469177, + "learning_rate": 0.0001399656510899064, + "loss": 0.0448, + "step": 15753 + }, + { + "epoch": 0.3692985694778975, + "grad_norm": 6.753596782684326, + "learning_rate": 0.00013995889868261303, + "loss": 0.7925, + "step": 15754 + }, + { + "epoch": 0.3693220110527025, + "grad_norm": 1.1652307510375977, + "learning_rate": 0.00013995214605850155, + "loss": 0.1208, + "step": 15755 + }, + { + "epoch": 0.3693454526275075, + "grad_norm": 1.1475850343704224, + "learning_rate": 0.00013994539321760864, + "loss": 0.1502, + "step": 15756 + }, + { + "epoch": 0.3693688942023125, + "grad_norm": 3.6214630603790283, + "learning_rate": 0.0001399386401599709, + "loss": 0.2965, + "step": 15757 + }, + { + "epoch": 0.3693923357771175, + "grad_norm": 4.623840808868408, + "learning_rate": 0.00013993188688562493, + "loss": 0.4929, + "step": 15758 + }, + { + "epoch": 0.3694157773519225, + "grad_norm": 4.474966049194336, + "learning_rate": 0.00013992513339460745, + "loss": 0.3229, + "step": 15759 + }, + { + "epoch": 0.3694392189267275, + "grad_norm": 2.0175602436065674, + "learning_rate": 0.00013991837968695508, + "loss": 0.1882, + "step": 15760 + }, + { + "epoch": 0.3694626605015325, + "grad_norm": 1.4579848051071167, + "learning_rate": 0.0001399116257627044, + "loss": 0.1171, + "step": 15761 + }, + { + "epoch": 0.3694861020763375, + "grad_norm": 5.0059380531311035, + "learning_rate": 0.0001399048716218922, + "loss": 0.5246, + "step": 15762 + }, + { + "epoch": 0.36950954365114247, + "grad_norm": 3.4437880516052246, + "learning_rate": 0.000139898117264555, + "loss": 0.3286, + "step": 15763 + }, + { + "epoch": 0.36953298522594746, + "grad_norm": 1.7288691997528076, + "learning_rate": 0.00013989136269072948, + "loss": 0.2275, + "step": 15764 + }, + { + "epoch": 0.36955642680075246, + "grad_norm": 3.172536611557007, + "learning_rate": 0.0001398846079004523, + "loss": 0.3799, + "step": 15765 + }, + { + "epoch": 0.36957986837555745, + "grad_norm": 0.491347074508667, + "learning_rate": 0.00013987785289376018, + "loss": 0.0579, + "step": 15766 + }, + { + "epoch": 0.36960330995036245, + "grad_norm": 0.9123982191085815, + "learning_rate": 0.0001398710976706897, + "loss": 0.1132, + "step": 15767 + }, + { + "epoch": 0.36962675152516744, + "grad_norm": 1.9306594133377075, + "learning_rate": 0.00013986434223127746, + "loss": 0.2878, + "step": 15768 + }, + { + "epoch": 0.36965019309997244, + "grad_norm": 1.212950587272644, + "learning_rate": 0.00013985758657556022, + "loss": 0.1179, + "step": 15769 + }, + { + "epoch": 0.36967363467477743, + "grad_norm": 3.839416980743408, + "learning_rate": 0.0001398508307035746, + "loss": 0.4766, + "step": 15770 + }, + { + "epoch": 0.3696970762495824, + "grad_norm": 2.3096425533294678, + "learning_rate": 0.0001398440746153572, + "loss": 1.5506, + "step": 15771 + }, + { + "epoch": 0.3697205178243874, + "grad_norm": 3.571256399154663, + "learning_rate": 0.0001398373183109448, + "loss": 0.2984, + "step": 15772 + }, + { + "epoch": 0.3697439593991924, + "grad_norm": 1.8329523801803589, + "learning_rate": 0.00013983056179037398, + "loss": 0.2227, + "step": 15773 + }, + { + "epoch": 0.3697674009739974, + "grad_norm": 0.9222132563591003, + "learning_rate": 0.00013982380505368134, + "loss": 0.1354, + "step": 15774 + }, + { + "epoch": 0.3697908425488024, + "grad_norm": 4.483906269073486, + "learning_rate": 0.0001398170481009037, + "loss": 0.5388, + "step": 15775 + }, + { + "epoch": 0.36981428412360745, + "grad_norm": 2.428286075592041, + "learning_rate": 0.00013981029093207761, + "loss": 0.2787, + "step": 15776 + }, + { + "epoch": 0.36983772569841245, + "grad_norm": 0.7343394160270691, + "learning_rate": 0.00013980353354723976, + "loss": 0.0688, + "step": 15777 + }, + { + "epoch": 0.36986116727321744, + "grad_norm": 3.726778268814087, + "learning_rate": 0.0001397967759464268, + "loss": 0.4902, + "step": 15778 + }, + { + "epoch": 0.36988460884802243, + "grad_norm": 1.6816478967666626, + "learning_rate": 0.00013979001812967543, + "loss": 0.2192, + "step": 15779 + }, + { + "epoch": 0.36990805042282743, + "grad_norm": 4.459029674530029, + "learning_rate": 0.0001397832600970223, + "loss": 0.3016, + "step": 15780 + }, + { + "epoch": 0.3699314919976324, + "grad_norm": 2.815800189971924, + "learning_rate": 0.0001397765018485041, + "loss": 0.338, + "step": 15781 + }, + { + "epoch": 0.3699549335724374, + "grad_norm": 0.6678181290626526, + "learning_rate": 0.00013976974338415744, + "loss": 0.0626, + "step": 15782 + }, + { + "epoch": 0.3699783751472424, + "grad_norm": 1.8910744190216064, + "learning_rate": 0.00013976298470401907, + "loss": 0.7221, + "step": 15783 + }, + { + "epoch": 0.3700018167220474, + "grad_norm": 3.584643602371216, + "learning_rate": 0.0001397562258081256, + "loss": 0.279, + "step": 15784 + }, + { + "epoch": 0.3700252582968524, + "grad_norm": 3.1017353534698486, + "learning_rate": 0.00013974946669651376, + "loss": 0.3039, + "step": 15785 + }, + { + "epoch": 0.3700486998716574, + "grad_norm": 4.393222332000732, + "learning_rate": 0.00013974270736922018, + "loss": 0.3469, + "step": 15786 + }, + { + "epoch": 0.3700721414464624, + "grad_norm": 3.413771390914917, + "learning_rate": 0.00013973594782628156, + "loss": 0.4123, + "step": 15787 + }, + { + "epoch": 0.3700955830212674, + "grad_norm": 3.374131917953491, + "learning_rate": 0.00013972918806773457, + "loss": 1.5155, + "step": 15788 + }, + { + "epoch": 0.3701190245960724, + "grad_norm": 5.323971271514893, + "learning_rate": 0.0001397224280936159, + "loss": 0.3047, + "step": 15789 + }, + { + "epoch": 0.37014246617087737, + "grad_norm": 2.8944873809814453, + "learning_rate": 0.00013971566790396217, + "loss": 0.2804, + "step": 15790 + }, + { + "epoch": 0.37016590774568237, + "grad_norm": 2.199573040008545, + "learning_rate": 0.00013970890749881017, + "loss": 0.2583, + "step": 15791 + }, + { + "epoch": 0.37018934932048736, + "grad_norm": 3.5467658042907715, + "learning_rate": 0.0001397021468781965, + "loss": 0.4681, + "step": 15792 + }, + { + "epoch": 0.37021279089529235, + "grad_norm": 3.3443431854248047, + "learning_rate": 0.00013969538604215783, + "loss": 0.4463, + "step": 15793 + }, + { + "epoch": 0.37023623247009735, + "grad_norm": 3.563340187072754, + "learning_rate": 0.0001396886249907309, + "loss": 0.3041, + "step": 15794 + }, + { + "epoch": 0.37025967404490234, + "grad_norm": 3.1411242485046387, + "learning_rate": 0.0001396818637239524, + "loss": 0.2615, + "step": 15795 + }, + { + "epoch": 0.37028311561970734, + "grad_norm": 3.0556576251983643, + "learning_rate": 0.000139675102241859, + "loss": 1.9617, + "step": 15796 + }, + { + "epoch": 0.37030655719451233, + "grad_norm": 0.6418673992156982, + "learning_rate": 0.00013966834054448732, + "loss": 0.0503, + "step": 15797 + }, + { + "epoch": 0.3703299987693173, + "grad_norm": 4.899251461029053, + "learning_rate": 0.00013966157863187415, + "loss": 0.5565, + "step": 15798 + }, + { + "epoch": 0.3703534403441223, + "grad_norm": 3.873798370361328, + "learning_rate": 0.00013965481650405614, + "loss": 0.4942, + "step": 15799 + }, + { + "epoch": 0.3703768819189273, + "grad_norm": 2.928178071975708, + "learning_rate": 0.00013964805416106995, + "loss": 0.202, + "step": 15800 + }, + { + "epoch": 0.3704003234937323, + "grad_norm": 1.5822839736938477, + "learning_rate": 0.00013964129160295234, + "loss": 0.1665, + "step": 15801 + }, + { + "epoch": 0.3704237650685373, + "grad_norm": 3.2258238792419434, + "learning_rate": 0.00013963452882973995, + "loss": 1.0677, + "step": 15802 + }, + { + "epoch": 0.3704472066433423, + "grad_norm": 3.5520687103271484, + "learning_rate": 0.00013962776584146947, + "loss": 0.3474, + "step": 15803 + }, + { + "epoch": 0.3704706482181473, + "grad_norm": 2.586895704269409, + "learning_rate": 0.0001396210026381777, + "loss": 0.222, + "step": 15804 + }, + { + "epoch": 0.3704940897929523, + "grad_norm": 3.2104923725128174, + "learning_rate": 0.00013961423921990117, + "loss": 0.2525, + "step": 15805 + }, + { + "epoch": 0.3705175313677573, + "grad_norm": 3.951641321182251, + "learning_rate": 0.00013960747558667666, + "loss": 0.5202, + "step": 15806 + }, + { + "epoch": 0.3705409729425623, + "grad_norm": 3.039393901824951, + "learning_rate": 0.0001396007117385409, + "loss": 0.4882, + "step": 15807 + }, + { + "epoch": 0.37056441451736727, + "grad_norm": 1.091938853263855, + "learning_rate": 0.00013959394767553064, + "loss": 0.1642, + "step": 15808 + }, + { + "epoch": 0.37058785609217226, + "grad_norm": 2.2234694957733154, + "learning_rate": 0.0001395871833976824, + "loss": 0.2879, + "step": 15809 + }, + { + "epoch": 0.37061129766697726, + "grad_norm": 5.66638708114624, + "learning_rate": 0.00013958041890503305, + "loss": 0.7564, + "step": 15810 + }, + { + "epoch": 0.37063473924178225, + "grad_norm": 1.2358431816101074, + "learning_rate": 0.0001395736541976192, + "loss": 0.1107, + "step": 15811 + }, + { + "epoch": 0.37065818081658725, + "grad_norm": 3.0003795623779297, + "learning_rate": 0.0001395668892754776, + "loss": 0.8647, + "step": 15812 + }, + { + "epoch": 0.37068162239139224, + "grad_norm": 3.1160762310028076, + "learning_rate": 0.00013956012413864495, + "loss": 0.3092, + "step": 15813 + }, + { + "epoch": 0.37070506396619723, + "grad_norm": 4.371613025665283, + "learning_rate": 0.00013955335878715794, + "loss": 0.3539, + "step": 15814 + }, + { + "epoch": 0.37072850554100223, + "grad_norm": 2.37894344329834, + "learning_rate": 0.00013954659322105333, + "loss": 1.489, + "step": 15815 + }, + { + "epoch": 0.3707519471158072, + "grad_norm": 3.8398821353912354, + "learning_rate": 0.00013953982744036776, + "loss": 0.4474, + "step": 15816 + }, + { + "epoch": 0.3707753886906122, + "grad_norm": 5.933689594268799, + "learning_rate": 0.000139533061445138, + "loss": 0.5875, + "step": 15817 + }, + { + "epoch": 0.3707988302654172, + "grad_norm": 4.9803924560546875, + "learning_rate": 0.0001395262952354007, + "loss": 0.5357, + "step": 15818 + }, + { + "epoch": 0.3708222718402222, + "grad_norm": 3.8259706497192383, + "learning_rate": 0.00013951952881119262, + "loss": 0.3196, + "step": 15819 + }, + { + "epoch": 0.3708457134150272, + "grad_norm": 3.203711748123169, + "learning_rate": 0.00013951276217255047, + "loss": 0.3977, + "step": 15820 + }, + { + "epoch": 0.3708691549898322, + "grad_norm": 1.510341763496399, + "learning_rate": 0.00013950599531951097, + "loss": 0.2228, + "step": 15821 + }, + { + "epoch": 0.3708925965646372, + "grad_norm": 4.213919639587402, + "learning_rate": 0.00013949922825211083, + "loss": 0.2939, + "step": 15822 + }, + { + "epoch": 0.3709160381394422, + "grad_norm": 7.252641201019287, + "learning_rate": 0.00013949246097038677, + "loss": 0.3449, + "step": 15823 + }, + { + "epoch": 0.3709394797142472, + "grad_norm": 6.076499938964844, + "learning_rate": 0.00013948569347437554, + "loss": 0.5139, + "step": 15824 + }, + { + "epoch": 0.37096292128905217, + "grad_norm": 4.025646686553955, + "learning_rate": 0.00013947892576411373, + "loss": 0.5159, + "step": 15825 + }, + { + "epoch": 0.37098636286385717, + "grad_norm": 3.039551019668579, + "learning_rate": 0.00013947215783963822, + "loss": 0.3337, + "step": 15826 + }, + { + "epoch": 0.3710098044386622, + "grad_norm": 6.175354957580566, + "learning_rate": 0.0001394653897009857, + "loss": 0.557, + "step": 15827 + }, + { + "epoch": 0.3710332460134672, + "grad_norm": 0.7133832573890686, + "learning_rate": 0.0001394586213481928, + "loss": 0.073, + "step": 15828 + }, + { + "epoch": 0.3710566875882722, + "grad_norm": 3.9505274295806885, + "learning_rate": 0.00013945185278129636, + "loss": 0.3605, + "step": 15829 + }, + { + "epoch": 0.3710801291630772, + "grad_norm": 0.9328967332839966, + "learning_rate": 0.00013944508400033306, + "loss": 0.1183, + "step": 15830 + }, + { + "epoch": 0.3711035707378822, + "grad_norm": 2.0974881649017334, + "learning_rate": 0.0001394383150053396, + "loss": 0.176, + "step": 15831 + }, + { + "epoch": 0.3711270123126872, + "grad_norm": 4.946130752563477, + "learning_rate": 0.00013943154579635272, + "loss": 0.72, + "step": 15832 + }, + { + "epoch": 0.3711504538874922, + "grad_norm": 3.930288314819336, + "learning_rate": 0.0001394247763734092, + "loss": 0.473, + "step": 15833 + }, + { + "epoch": 0.3711738954622972, + "grad_norm": 2.4349005222320557, + "learning_rate": 0.00013941800673654571, + "loss": 0.1976, + "step": 15834 + }, + { + "epoch": 0.37119733703710217, + "grad_norm": 3.0774455070495605, + "learning_rate": 0.000139411236885799, + "loss": 0.283, + "step": 15835 + }, + { + "epoch": 0.37122077861190717, + "grad_norm": 3.831106185913086, + "learning_rate": 0.00013940446682120586, + "loss": 0.5159, + "step": 15836 + }, + { + "epoch": 0.37124422018671216, + "grad_norm": 2.9883861541748047, + "learning_rate": 0.0001393976965428029, + "loss": 0.3073, + "step": 15837 + }, + { + "epoch": 0.37126766176151715, + "grad_norm": 2.123612403869629, + "learning_rate": 0.00013939092605062697, + "loss": 0.9169, + "step": 15838 + }, + { + "epoch": 0.37129110333632215, + "grad_norm": 4.032456874847412, + "learning_rate": 0.00013938415534471479, + "loss": 0.6987, + "step": 15839 + }, + { + "epoch": 0.37131454491112714, + "grad_norm": 2.168673038482666, + "learning_rate": 0.00013937738442510307, + "loss": 0.1965, + "step": 15840 + }, + { + "epoch": 0.37133798648593214, + "grad_norm": 5.128718376159668, + "learning_rate": 0.00013937061329182852, + "loss": 0.4475, + "step": 15841 + }, + { + "epoch": 0.37136142806073713, + "grad_norm": 1.2704977989196777, + "learning_rate": 0.00013936384194492796, + "loss": 0.1158, + "step": 15842 + }, + { + "epoch": 0.3713848696355421, + "grad_norm": 4.582817554473877, + "learning_rate": 0.00013935707038443802, + "loss": 0.5049, + "step": 15843 + }, + { + "epoch": 0.3714083112103471, + "grad_norm": 1.0973789691925049, + "learning_rate": 0.00013935029861039558, + "loss": 0.1018, + "step": 15844 + }, + { + "epoch": 0.3714317527851521, + "grad_norm": 4.597095966339111, + "learning_rate": 0.00013934352662283727, + "loss": 0.547, + "step": 15845 + }, + { + "epoch": 0.3714551943599571, + "grad_norm": 3.376473903656006, + "learning_rate": 0.00013933675442179993, + "loss": 0.7861, + "step": 15846 + }, + { + "epoch": 0.3714786359347621, + "grad_norm": 2.150151014328003, + "learning_rate": 0.00013932998200732018, + "loss": 0.2926, + "step": 15847 + }, + { + "epoch": 0.3715020775095671, + "grad_norm": 4.223972320556641, + "learning_rate": 0.0001393232093794349, + "loss": 0.5579, + "step": 15848 + }, + { + "epoch": 0.3715255190843721, + "grad_norm": 3.238436698913574, + "learning_rate": 0.00013931643653818075, + "loss": 0.4286, + "step": 15849 + }, + { + "epoch": 0.3715489606591771, + "grad_norm": 3.3785061836242676, + "learning_rate": 0.0001393096634835945, + "loss": 0.2614, + "step": 15850 + }, + { + "epoch": 0.3715724022339821, + "grad_norm": 3.549471378326416, + "learning_rate": 0.00013930289021571294, + "loss": 0.5152, + "step": 15851 + }, + { + "epoch": 0.3715958438087871, + "grad_norm": 0.5408366918563843, + "learning_rate": 0.0001392961167345728, + "loss": 0.0674, + "step": 15852 + }, + { + "epoch": 0.37161928538359207, + "grad_norm": 4.009927749633789, + "learning_rate": 0.0001392893430402108, + "loss": 0.4675, + "step": 15853 + }, + { + "epoch": 0.37164272695839706, + "grad_norm": 4.429833889007568, + "learning_rate": 0.00013928256913266373, + "loss": 0.7294, + "step": 15854 + }, + { + "epoch": 0.37166616853320206, + "grad_norm": 5.100973129272461, + "learning_rate": 0.00013927579501196832, + "loss": 0.5637, + "step": 15855 + }, + { + "epoch": 0.37168961010800705, + "grad_norm": 3.4276680946350098, + "learning_rate": 0.00013926902067816136, + "loss": 0.2981, + "step": 15856 + }, + { + "epoch": 0.37171305168281205, + "grad_norm": 2.5037007331848145, + "learning_rate": 0.00013926224613127958, + "loss": 0.2039, + "step": 15857 + }, + { + "epoch": 0.37173649325761704, + "grad_norm": 2.323124408721924, + "learning_rate": 0.00013925547137135977, + "loss": 0.3569, + "step": 15858 + }, + { + "epoch": 0.37175993483242203, + "grad_norm": 2.5319087505340576, + "learning_rate": 0.00013924869639843866, + "loss": 0.2522, + "step": 15859 + }, + { + "epoch": 0.37178337640722703, + "grad_norm": 3.3213090896606445, + "learning_rate": 0.000139241921212553, + "loss": 0.4362, + "step": 15860 + }, + { + "epoch": 0.371806817982032, + "grad_norm": 0.4997048079967499, + "learning_rate": 0.00013923514581373959, + "loss": 0.0726, + "step": 15861 + }, + { + "epoch": 0.371830259556837, + "grad_norm": 4.482701778411865, + "learning_rate": 0.00013922837020203515, + "loss": 0.4132, + "step": 15862 + }, + { + "epoch": 0.371853701131642, + "grad_norm": 0.5699908137321472, + "learning_rate": 0.00013922159437747652, + "loss": 0.0682, + "step": 15863 + }, + { + "epoch": 0.371877142706447, + "grad_norm": 3.986478328704834, + "learning_rate": 0.00013921481834010038, + "loss": 0.3758, + "step": 15864 + }, + { + "epoch": 0.371900584281252, + "grad_norm": 4.338764190673828, + "learning_rate": 0.00013920804208994352, + "loss": 0.3946, + "step": 15865 + }, + { + "epoch": 0.371924025856057, + "grad_norm": 5.059213161468506, + "learning_rate": 0.00013920126562704275, + "loss": 0.3625, + "step": 15866 + }, + { + "epoch": 0.371947467430862, + "grad_norm": 3.277669906616211, + "learning_rate": 0.0001391944889514348, + "loss": 2.0932, + "step": 15867 + }, + { + "epoch": 0.371970909005667, + "grad_norm": 2.6976447105407715, + "learning_rate": 0.00013918771206315647, + "loss": 0.1762, + "step": 15868 + }, + { + "epoch": 0.371994350580472, + "grad_norm": 4.822678565979004, + "learning_rate": 0.00013918093496224449, + "loss": 0.3616, + "step": 15869 + }, + { + "epoch": 0.37201779215527697, + "grad_norm": 0.9705535769462585, + "learning_rate": 0.00013917415764873564, + "loss": 0.1165, + "step": 15870 + }, + { + "epoch": 0.37204123373008197, + "grad_norm": 3.634552478790283, + "learning_rate": 0.00013916738012266676, + "loss": 0.3469, + "step": 15871 + }, + { + "epoch": 0.37206467530488696, + "grad_norm": 1.7109423875808716, + "learning_rate": 0.00013916060238407452, + "loss": 0.0922, + "step": 15872 + }, + { + "epoch": 0.37208811687969195, + "grad_norm": 2.7055068016052246, + "learning_rate": 0.00013915382443299576, + "loss": 0.2946, + "step": 15873 + }, + { + "epoch": 0.37211155845449695, + "grad_norm": 2.0292909145355225, + "learning_rate": 0.0001391470462694673, + "loss": 0.1473, + "step": 15874 + }, + { + "epoch": 0.37213500002930194, + "grad_norm": 4.1132283210754395, + "learning_rate": 0.0001391402678935258, + "loss": 0.539, + "step": 15875 + }, + { + "epoch": 0.37215844160410694, + "grad_norm": 4.146591663360596, + "learning_rate": 0.00013913348930520814, + "loss": 0.462, + "step": 15876 + }, + { + "epoch": 0.37218188317891193, + "grad_norm": 3.973430871963501, + "learning_rate": 0.00013912671050455105, + "loss": 0.3595, + "step": 15877 + }, + { + "epoch": 0.372205324753717, + "grad_norm": 2.728374481201172, + "learning_rate": 0.00013911993149159137, + "loss": 0.2335, + "step": 15878 + }, + { + "epoch": 0.372228766328522, + "grad_norm": 2.073211193084717, + "learning_rate": 0.00013911315226636578, + "loss": 1.273, + "step": 15879 + }, + { + "epoch": 0.37225220790332697, + "grad_norm": 3.848519802093506, + "learning_rate": 0.00013910637282891117, + "loss": 0.332, + "step": 15880 + }, + { + "epoch": 0.37227564947813196, + "grad_norm": 3.2170093059539795, + "learning_rate": 0.00013909959317926426, + "loss": 0.2621, + "step": 15881 + }, + { + "epoch": 0.37229909105293696, + "grad_norm": 3.1588733196258545, + "learning_rate": 0.00013909281331746186, + "loss": 2.2317, + "step": 15882 + }, + { + "epoch": 0.37232253262774195, + "grad_norm": 1.8924647569656372, + "learning_rate": 0.00013908603324354078, + "loss": 0.2095, + "step": 15883 + }, + { + "epoch": 0.37234597420254695, + "grad_norm": 1.0564749240875244, + "learning_rate": 0.00013907925295753777, + "loss": 0.1233, + "step": 15884 + }, + { + "epoch": 0.37236941577735194, + "grad_norm": 1.3466918468475342, + "learning_rate": 0.00013907247245948962, + "loss": 0.084, + "step": 15885 + }, + { + "epoch": 0.37239285735215694, + "grad_norm": 4.531045913696289, + "learning_rate": 0.00013906569174943314, + "loss": 0.2097, + "step": 15886 + }, + { + "epoch": 0.37241629892696193, + "grad_norm": 3.7325289249420166, + "learning_rate": 0.00013905891082740515, + "loss": 0.4463, + "step": 15887 + }, + { + "epoch": 0.3724397405017669, + "grad_norm": 0.7217568159103394, + "learning_rate": 0.0001390521296934424, + "loss": 0.0885, + "step": 15888 + }, + { + "epoch": 0.3724631820765719, + "grad_norm": 3.468465805053711, + "learning_rate": 0.00013904534834758165, + "loss": 0.335, + "step": 15889 + }, + { + "epoch": 0.3724866236513769, + "grad_norm": 2.1643126010894775, + "learning_rate": 0.0001390385667898598, + "loss": 0.1027, + "step": 15890 + }, + { + "epoch": 0.3725100652261819, + "grad_norm": 3.134237766265869, + "learning_rate": 0.00013903178502031356, + "loss": 0.2055, + "step": 15891 + }, + { + "epoch": 0.3725335068009869, + "grad_norm": 3.9964568614959717, + "learning_rate": 0.00013902500303897974, + "loss": 0.4148, + "step": 15892 + }, + { + "epoch": 0.3725569483757919, + "grad_norm": 0.4376384913921356, + "learning_rate": 0.00013901822084589518, + "loss": 0.06, + "step": 15893 + }, + { + "epoch": 0.3725803899505969, + "grad_norm": 1.36359441280365, + "learning_rate": 0.00013901143844109666, + "loss": 0.1064, + "step": 15894 + }, + { + "epoch": 0.3726038315254019, + "grad_norm": 4.4314284324646, + "learning_rate": 0.00013900465582462095, + "loss": 0.4021, + "step": 15895 + }, + { + "epoch": 0.3726272731002069, + "grad_norm": 5.496298789978027, + "learning_rate": 0.00013899787299650492, + "loss": 0.3126, + "step": 15896 + }, + { + "epoch": 0.3726507146750119, + "grad_norm": 0.5019333362579346, + "learning_rate": 0.0001389910899567853, + "loss": 0.0373, + "step": 15897 + }, + { + "epoch": 0.37267415624981687, + "grad_norm": 2.922636032104492, + "learning_rate": 0.00013898430670549895, + "loss": 0.2744, + "step": 15898 + }, + { + "epoch": 0.37269759782462186, + "grad_norm": 1.5340840816497803, + "learning_rate": 0.00013897752324268264, + "loss": 0.211, + "step": 15899 + }, + { + "epoch": 0.37272103939942686, + "grad_norm": 3.285874366760254, + "learning_rate": 0.00013897073956837324, + "loss": 0.2862, + "step": 15900 + }, + { + "epoch": 0.37274448097423185, + "grad_norm": 4.6426239013671875, + "learning_rate": 0.00013896395568260746, + "loss": 0.4279, + "step": 15901 + }, + { + "epoch": 0.37276792254903685, + "grad_norm": 3.187614917755127, + "learning_rate": 0.00013895717158542219, + "loss": 0.3522, + "step": 15902 + }, + { + "epoch": 0.37279136412384184, + "grad_norm": 1.6728583574295044, + "learning_rate": 0.0001389503872768542, + "loss": 0.1546, + "step": 15903 + }, + { + "epoch": 0.37281480569864683, + "grad_norm": 1.6485811471939087, + "learning_rate": 0.00013894360275694032, + "loss": 1.1116, + "step": 15904 + }, + { + "epoch": 0.37283824727345183, + "grad_norm": 4.117175102233887, + "learning_rate": 0.00013893681802571735, + "loss": 0.2618, + "step": 15905 + }, + { + "epoch": 0.3728616888482568, + "grad_norm": 2.5697758197784424, + "learning_rate": 0.0001389300330832221, + "loss": 0.1685, + "step": 15906 + }, + { + "epoch": 0.3728851304230618, + "grad_norm": 6.428437232971191, + "learning_rate": 0.00013892324792949144, + "loss": 0.8583, + "step": 15907 + }, + { + "epoch": 0.3729085719978668, + "grad_norm": 2.9624438285827637, + "learning_rate": 0.0001389164625645621, + "loss": 0.2649, + "step": 15908 + }, + { + "epoch": 0.3729320135726718, + "grad_norm": 4.425642013549805, + "learning_rate": 0.00013890967698847092, + "loss": 0.494, + "step": 15909 + }, + { + "epoch": 0.3729554551474768, + "grad_norm": 5.252538681030273, + "learning_rate": 0.00013890289120125477, + "loss": 0.5065, + "step": 15910 + }, + { + "epoch": 0.3729788967222818, + "grad_norm": 2.09344482421875, + "learning_rate": 0.00013889610520295042, + "loss": 0.2347, + "step": 15911 + }, + { + "epoch": 0.3730023382970868, + "grad_norm": 1.1810890436172485, + "learning_rate": 0.00013888931899359475, + "loss": 0.1676, + "step": 15912 + }, + { + "epoch": 0.3730257798718918, + "grad_norm": 3.1965863704681396, + "learning_rate": 0.0001388825325732245, + "loss": 1.5684, + "step": 15913 + }, + { + "epoch": 0.3730492214466968, + "grad_norm": 3.235542058944702, + "learning_rate": 0.00013887574594187655, + "loss": 0.329, + "step": 15914 + }, + { + "epoch": 0.37307266302150177, + "grad_norm": 3.3179569244384766, + "learning_rate": 0.00013886895909958773, + "loss": 0.3688, + "step": 15915 + }, + { + "epoch": 0.37309610459630677, + "grad_norm": 2.6458511352539062, + "learning_rate": 0.0001388621720463948, + "loss": 0.3287, + "step": 15916 + }, + { + "epoch": 0.37311954617111176, + "grad_norm": 2.492521047592163, + "learning_rate": 0.00013885538478233467, + "loss": 0.3077, + "step": 15917 + }, + { + "epoch": 0.37314298774591675, + "grad_norm": 4.753476142883301, + "learning_rate": 0.0001388485973074441, + "loss": 0.8086, + "step": 15918 + }, + { + "epoch": 0.37316642932072175, + "grad_norm": 3.147160768508911, + "learning_rate": 0.00013884180962176, + "loss": 1.7619, + "step": 15919 + }, + { + "epoch": 0.37318987089552674, + "grad_norm": 1.012479543685913, + "learning_rate": 0.0001388350217253191, + "loss": 0.1313, + "step": 15920 + }, + { + "epoch": 0.37321331247033174, + "grad_norm": 3.228668689727783, + "learning_rate": 0.00013882823361815829, + "loss": 0.221, + "step": 15921 + }, + { + "epoch": 0.37323675404513673, + "grad_norm": 0.970066487789154, + "learning_rate": 0.0001388214453003144, + "loss": 0.0746, + "step": 15922 + }, + { + "epoch": 0.3732601956199417, + "grad_norm": 2.3008062839508057, + "learning_rate": 0.00013881465677182425, + "loss": 0.2666, + "step": 15923 + }, + { + "epoch": 0.3732836371947467, + "grad_norm": 0.7946169972419739, + "learning_rate": 0.00013880786803272468, + "loss": 0.0872, + "step": 15924 + }, + { + "epoch": 0.3733070787695517, + "grad_norm": 3.212399959564209, + "learning_rate": 0.00013880107908305253, + "loss": 0.251, + "step": 15925 + }, + { + "epoch": 0.3733305203443567, + "grad_norm": 1.91902494430542, + "learning_rate": 0.00013879428992284466, + "loss": 0.2803, + "step": 15926 + }, + { + "epoch": 0.3733539619191617, + "grad_norm": 2.7449326515197754, + "learning_rate": 0.00013878750055213782, + "loss": 0.3297, + "step": 15927 + }, + { + "epoch": 0.3733774034939667, + "grad_norm": 4.463451385498047, + "learning_rate": 0.00013878071097096897, + "loss": 0.4636, + "step": 15928 + }, + { + "epoch": 0.37340084506877175, + "grad_norm": 2.3604493141174316, + "learning_rate": 0.00013877392117937485, + "loss": 0.2496, + "step": 15929 + }, + { + "epoch": 0.37342428664357674, + "grad_norm": 4.0974225997924805, + "learning_rate": 0.00013876713117739236, + "loss": 0.3024, + "step": 15930 + }, + { + "epoch": 0.37344772821838174, + "grad_norm": 3.337899684906006, + "learning_rate": 0.00013876034096505833, + "loss": 0.5002, + "step": 15931 + }, + { + "epoch": 0.37347116979318673, + "grad_norm": 3.243459463119507, + "learning_rate": 0.00013875355054240963, + "loss": 0.6346, + "step": 15932 + }, + { + "epoch": 0.3734946113679917, + "grad_norm": 3.440322160720825, + "learning_rate": 0.00013874675990948303, + "loss": 0.5019, + "step": 15933 + }, + { + "epoch": 0.3735180529427967, + "grad_norm": 3.0964436531066895, + "learning_rate": 0.00013873996906631542, + "loss": 0.4029, + "step": 15934 + }, + { + "epoch": 0.3735414945176017, + "grad_norm": 0.9557746648788452, + "learning_rate": 0.00013873317801294367, + "loss": 0.0416, + "step": 15935 + }, + { + "epoch": 0.3735649360924067, + "grad_norm": 1.2863768339157104, + "learning_rate": 0.00013872638674940462, + "loss": 0.1112, + "step": 15936 + }, + { + "epoch": 0.3735883776672117, + "grad_norm": 3.4181721210479736, + "learning_rate": 0.00013871959527573507, + "loss": 0.3351, + "step": 15937 + }, + { + "epoch": 0.3736118192420167, + "grad_norm": 0.5742636322975159, + "learning_rate": 0.00013871280359197197, + "loss": 0.0484, + "step": 15938 + }, + { + "epoch": 0.3736352608168217, + "grad_norm": 3.407923460006714, + "learning_rate": 0.00013870601169815205, + "loss": 0.3894, + "step": 15939 + }, + { + "epoch": 0.3736587023916267, + "grad_norm": 4.9918718338012695, + "learning_rate": 0.0001386992195943122, + "loss": 0.4956, + "step": 15940 + }, + { + "epoch": 0.3736821439664317, + "grad_norm": 2.6127214431762695, + "learning_rate": 0.00013869242728048936, + "loss": 0.1978, + "step": 15941 + }, + { + "epoch": 0.3737055855412367, + "grad_norm": 0.9066374897956848, + "learning_rate": 0.0001386856347567203, + "loss": 0.1057, + "step": 15942 + }, + { + "epoch": 0.37372902711604167, + "grad_norm": 6.061785697937012, + "learning_rate": 0.00013867884202304188, + "loss": 0.5249, + "step": 15943 + }, + { + "epoch": 0.37375246869084666, + "grad_norm": 4.926314353942871, + "learning_rate": 0.00013867204907949102, + "loss": 0.4764, + "step": 15944 + }, + { + "epoch": 0.37377591026565166, + "grad_norm": 3.1667990684509277, + "learning_rate": 0.0001386652559261045, + "loss": 0.4472, + "step": 15945 + }, + { + "epoch": 0.37379935184045665, + "grad_norm": 3.2928435802459717, + "learning_rate": 0.0001386584625629192, + "loss": 0.2429, + "step": 15946 + }, + { + "epoch": 0.37382279341526164, + "grad_norm": 3.0862972736358643, + "learning_rate": 0.000138651668989972, + "loss": 0.2228, + "step": 15947 + }, + { + "epoch": 0.37384623499006664, + "grad_norm": 4.1606926918029785, + "learning_rate": 0.00013864487520729978, + "loss": 0.594, + "step": 15948 + }, + { + "epoch": 0.37386967656487163, + "grad_norm": 2.4518918991088867, + "learning_rate": 0.00013863808121493938, + "loss": 0.182, + "step": 15949 + }, + { + "epoch": 0.37389311813967663, + "grad_norm": 3.8195834159851074, + "learning_rate": 0.00013863128701292765, + "loss": 0.4226, + "step": 15950 + }, + { + "epoch": 0.3739165597144816, + "grad_norm": 2.7687735557556152, + "learning_rate": 0.00013862449260130147, + "loss": 0.1993, + "step": 15951 + }, + { + "epoch": 0.3739400012892866, + "grad_norm": 3.82887601852417, + "learning_rate": 0.00013861769798009771, + "loss": 0.2624, + "step": 15952 + }, + { + "epoch": 0.3739634428640916, + "grad_norm": 2.0416557788848877, + "learning_rate": 0.00013861090314935327, + "loss": 0.2423, + "step": 15953 + }, + { + "epoch": 0.3739868844388966, + "grad_norm": 5.931467533111572, + "learning_rate": 0.00013860410810910494, + "loss": 0.5688, + "step": 15954 + }, + { + "epoch": 0.3740103260137016, + "grad_norm": 2.554142951965332, + "learning_rate": 0.00013859731285938964, + "loss": 0.3205, + "step": 15955 + }, + { + "epoch": 0.3740337675885066, + "grad_norm": 2.523822069168091, + "learning_rate": 0.00013859051740024424, + "loss": 0.2773, + "step": 15956 + }, + { + "epoch": 0.3740572091633116, + "grad_norm": 1.233865737915039, + "learning_rate": 0.0001385837217317056, + "loss": 0.1262, + "step": 15957 + }, + { + "epoch": 0.3740806507381166, + "grad_norm": 4.342230319976807, + "learning_rate": 0.00013857692585381062, + "loss": 0.4827, + "step": 15958 + }, + { + "epoch": 0.3741040923129216, + "grad_norm": 1.9162417650222778, + "learning_rate": 0.00013857012976659615, + "loss": 0.2728, + "step": 15959 + }, + { + "epoch": 0.37412753388772657, + "grad_norm": 3.1871752738952637, + "learning_rate": 0.00013856333347009908, + "loss": 0.3374, + "step": 15960 + }, + { + "epoch": 0.37415097546253157, + "grad_norm": 1.8071707487106323, + "learning_rate": 0.00013855653696435627, + "loss": 0.1514, + "step": 15961 + }, + { + "epoch": 0.37417441703733656, + "grad_norm": 3.9941818714141846, + "learning_rate": 0.00013854974024940457, + "loss": 0.771, + "step": 15962 + }, + { + "epoch": 0.37419785861214155, + "grad_norm": 1.4670612812042236, + "learning_rate": 0.00013854294332528094, + "loss": 0.2184, + "step": 15963 + }, + { + "epoch": 0.37422130018694655, + "grad_norm": 3.7276694774627686, + "learning_rate": 0.0001385361461920222, + "loss": 0.5373, + "step": 15964 + }, + { + "epoch": 0.37424474176175154, + "grad_norm": 2.840961456298828, + "learning_rate": 0.00013852934884966525, + "loss": 0.3131, + "step": 15965 + }, + { + "epoch": 0.37426818333655654, + "grad_norm": 0.6981577277183533, + "learning_rate": 0.00013852255129824696, + "loss": 0.0579, + "step": 15966 + }, + { + "epoch": 0.37429162491136153, + "grad_norm": 0.7711860537528992, + "learning_rate": 0.00013851575353780423, + "loss": 0.0341, + "step": 15967 + }, + { + "epoch": 0.3743150664861665, + "grad_norm": 2.7029316425323486, + "learning_rate": 0.00013850895556837394, + "loss": 0.1885, + "step": 15968 + }, + { + "epoch": 0.3743385080609715, + "grad_norm": 3.7904717922210693, + "learning_rate": 0.000138502157389993, + "loss": 0.4392, + "step": 15969 + }, + { + "epoch": 0.3743619496357765, + "grad_norm": 5.412634372711182, + "learning_rate": 0.00013849535900269824, + "loss": 0.7723, + "step": 15970 + }, + { + "epoch": 0.3743853912105815, + "grad_norm": 0.32262179255485535, + "learning_rate": 0.00013848856040652658, + "loss": 0.0484, + "step": 15971 + }, + { + "epoch": 0.3744088327853865, + "grad_norm": 2.9638261795043945, + "learning_rate": 0.0001384817616015149, + "loss": 1.9216, + "step": 15972 + }, + { + "epoch": 0.3744322743601915, + "grad_norm": 3.5104691982269287, + "learning_rate": 0.00013847496258770014, + "loss": 0.3578, + "step": 15973 + }, + { + "epoch": 0.3744557159349965, + "grad_norm": 5.590705394744873, + "learning_rate": 0.0001384681633651191, + "loss": 0.5204, + "step": 15974 + }, + { + "epoch": 0.3744791575098015, + "grad_norm": 1.4145472049713135, + "learning_rate": 0.00013846136393380878, + "loss": 0.2061, + "step": 15975 + }, + { + "epoch": 0.3745025990846065, + "grad_norm": 1.3996672630310059, + "learning_rate": 0.00013845456429380598, + "loss": 0.1602, + "step": 15976 + }, + { + "epoch": 0.3745260406594115, + "grad_norm": 4.007279872894287, + "learning_rate": 0.00013844776444514766, + "loss": 0.587, + "step": 15977 + }, + { + "epoch": 0.37454948223421647, + "grad_norm": 3.235694408416748, + "learning_rate": 0.00013844096438787065, + "loss": 0.3943, + "step": 15978 + }, + { + "epoch": 0.37457292380902146, + "grad_norm": 3.0399832725524902, + "learning_rate": 0.0001384341641220119, + "loss": 0.1705, + "step": 15979 + }, + { + "epoch": 0.3745963653838265, + "grad_norm": 2.021348714828491, + "learning_rate": 0.0001384273636476083, + "loss": 0.2105, + "step": 15980 + }, + { + "epoch": 0.3746198069586315, + "grad_norm": 1.3923473358154297, + "learning_rate": 0.00013842056296469677, + "loss": 0.193, + "step": 15981 + }, + { + "epoch": 0.3746432485334365, + "grad_norm": 2.8791308403015137, + "learning_rate": 0.00013841376207331417, + "loss": 0.208, + "step": 15982 + }, + { + "epoch": 0.3746666901082415, + "grad_norm": 1.824251413345337, + "learning_rate": 0.00013840696097349738, + "loss": 0.1549, + "step": 15983 + }, + { + "epoch": 0.3746901316830465, + "grad_norm": 0.5512018203735352, + "learning_rate": 0.00013840015966528336, + "loss": 0.0457, + "step": 15984 + }, + { + "epoch": 0.3747135732578515, + "grad_norm": 4.096031188964844, + "learning_rate": 0.000138393358148709, + "loss": 0.2997, + "step": 15985 + }, + { + "epoch": 0.3747370148326565, + "grad_norm": 3.1274731159210205, + "learning_rate": 0.00013838655642381121, + "loss": 0.2987, + "step": 15986 + }, + { + "epoch": 0.3747604564074615, + "grad_norm": 1.1820175647735596, + "learning_rate": 0.00013837975449062687, + "loss": 0.1322, + "step": 15987 + }, + { + "epoch": 0.37478389798226647, + "grad_norm": 2.6589598655700684, + "learning_rate": 0.0001383729523491929, + "loss": 0.1233, + "step": 15988 + }, + { + "epoch": 0.37480733955707146, + "grad_norm": 0.5054351687431335, + "learning_rate": 0.00013836614999954624, + "loss": 0.0377, + "step": 15989 + }, + { + "epoch": 0.37483078113187646, + "grad_norm": 3.3927321434020996, + "learning_rate": 0.00013835934744172376, + "loss": 0.4268, + "step": 15990 + }, + { + "epoch": 0.37485422270668145, + "grad_norm": 5.061229228973389, + "learning_rate": 0.00013835254467576234, + "loss": 0.6733, + "step": 15991 + }, + { + "epoch": 0.37487766428148644, + "grad_norm": 3.3594326972961426, + "learning_rate": 0.000138345741701699, + "loss": 0.1792, + "step": 15992 + }, + { + "epoch": 0.37490110585629144, + "grad_norm": 3.6701455116271973, + "learning_rate": 0.00013833893851957055, + "loss": 0.4099, + "step": 15993 + }, + { + "epoch": 0.37492454743109643, + "grad_norm": 0.6128681898117065, + "learning_rate": 0.00013833213512941397, + "loss": 0.0371, + "step": 15994 + }, + { + "epoch": 0.3749479890059014, + "grad_norm": 3.3595058917999268, + "learning_rate": 0.0001383253315312661, + "loss": 0.3678, + "step": 15995 + }, + { + "epoch": 0.3749714305807064, + "grad_norm": 3.9676856994628906, + "learning_rate": 0.00013831852772516393, + "loss": 0.6455, + "step": 15996 + }, + { + "epoch": 0.3749948721555114, + "grad_norm": 1.8816903829574585, + "learning_rate": 0.00013831172371114433, + "loss": 0.1316, + "step": 15997 + }, + { + "epoch": 0.3750183137303164, + "grad_norm": 2.314685821533203, + "learning_rate": 0.0001383049194892443, + "loss": 0.2549, + "step": 15998 + }, + { + "epoch": 0.3750417553051214, + "grad_norm": 4.978537082672119, + "learning_rate": 0.00013829811505950066, + "loss": 0.5391, + "step": 15999 + }, + { + "epoch": 0.3750651968799264, + "grad_norm": 2.5492923259735107, + "learning_rate": 0.00013829131042195035, + "loss": 0.1275, + "step": 16000 + }, + { + "epoch": 0.3750886384547314, + "grad_norm": 8.977371215820312, + "learning_rate": 0.00013828450557663033, + "loss": 0.5369, + "step": 16001 + }, + { + "epoch": 0.3751120800295364, + "grad_norm": 1.9830541610717773, + "learning_rate": 0.0001382777005235775, + "loss": 0.1716, + "step": 16002 + }, + { + "epoch": 0.3751355216043414, + "grad_norm": 6.872176647186279, + "learning_rate": 0.00013827089526282878, + "loss": 0.7142, + "step": 16003 + }, + { + "epoch": 0.3751589631791464, + "grad_norm": 3.1751649379730225, + "learning_rate": 0.00013826408979442117, + "loss": 0.2626, + "step": 16004 + }, + { + "epoch": 0.37518240475395137, + "grad_norm": 2.5095152854919434, + "learning_rate": 0.00013825728411839147, + "loss": 0.2543, + "step": 16005 + }, + { + "epoch": 0.37520584632875636, + "grad_norm": 2.032654285430908, + "learning_rate": 0.0001382504782347767, + "loss": 0.1665, + "step": 16006 + }, + { + "epoch": 0.37522928790356136, + "grad_norm": 3.6436281204223633, + "learning_rate": 0.00013824367214361374, + "loss": 1.5517, + "step": 16007 + }, + { + "epoch": 0.37525272947836635, + "grad_norm": 3.4111709594726562, + "learning_rate": 0.00013823686584493955, + "loss": 0.3156, + "step": 16008 + }, + { + "epoch": 0.37527617105317135, + "grad_norm": 2.5571229457855225, + "learning_rate": 0.00013823005933879102, + "loss": 0.3043, + "step": 16009 + }, + { + "epoch": 0.37529961262797634, + "grad_norm": 3.0610806941986084, + "learning_rate": 0.00013822325262520516, + "loss": 0.2285, + "step": 16010 + }, + { + "epoch": 0.37532305420278134, + "grad_norm": 1.5790342092514038, + "learning_rate": 0.00013821644570421885, + "loss": 0.1504, + "step": 16011 + }, + { + "epoch": 0.37534649577758633, + "grad_norm": 3.1636884212493896, + "learning_rate": 0.00013820963857586903, + "loss": 0.3063, + "step": 16012 + }, + { + "epoch": 0.3753699373523913, + "grad_norm": 1.0575132369995117, + "learning_rate": 0.0001382028312401926, + "loss": 0.1301, + "step": 16013 + }, + { + "epoch": 0.3753933789271963, + "grad_norm": 2.3083927631378174, + "learning_rate": 0.00013819602369722657, + "loss": 0.1835, + "step": 16014 + }, + { + "epoch": 0.3754168205020013, + "grad_norm": 1.8578286170959473, + "learning_rate": 0.00013818921594700785, + "loss": 0.2218, + "step": 16015 + }, + { + "epoch": 0.3754402620768063, + "grad_norm": 0.9865984320640564, + "learning_rate": 0.00013818240798957335, + "loss": 0.0833, + "step": 16016 + }, + { + "epoch": 0.3754637036516113, + "grad_norm": 2.548541307449341, + "learning_rate": 0.00013817559982496006, + "loss": 0.2072, + "step": 16017 + }, + { + "epoch": 0.3754871452264163, + "grad_norm": 4.628004550933838, + "learning_rate": 0.00013816879145320485, + "loss": 0.4098, + "step": 16018 + }, + { + "epoch": 0.3755105868012213, + "grad_norm": 1.044771671295166, + "learning_rate": 0.0001381619828743447, + "loss": 0.1032, + "step": 16019 + }, + { + "epoch": 0.3755340283760263, + "grad_norm": 3.4984400272369385, + "learning_rate": 0.00013815517408841656, + "loss": 0.2804, + "step": 16020 + }, + { + "epoch": 0.3755574699508313, + "grad_norm": 4.400959014892578, + "learning_rate": 0.0001381483650954574, + "loss": 0.1535, + "step": 16021 + }, + { + "epoch": 0.3755809115256363, + "grad_norm": 4.3013153076171875, + "learning_rate": 0.00013814155589550413, + "loss": 0.4365, + "step": 16022 + }, + { + "epoch": 0.37560435310044127, + "grad_norm": 5.738515377044678, + "learning_rate": 0.00013813474648859372, + "loss": 0.3897, + "step": 16023 + }, + { + "epoch": 0.37562779467524626, + "grad_norm": 5.697710037231445, + "learning_rate": 0.00013812793687476309, + "loss": 0.5795, + "step": 16024 + }, + { + "epoch": 0.37565123625005126, + "grad_norm": 3.9819607734680176, + "learning_rate": 0.0001381211270540492, + "loss": 0.556, + "step": 16025 + }, + { + "epoch": 0.37567467782485625, + "grad_norm": 1.3485164642333984, + "learning_rate": 0.000138114317026489, + "loss": 0.1364, + "step": 16026 + }, + { + "epoch": 0.37569811939966125, + "grad_norm": 3.5880982875823975, + "learning_rate": 0.00013810750679211944, + "loss": 0.3859, + "step": 16027 + }, + { + "epoch": 0.37572156097446624, + "grad_norm": 1.2288224697113037, + "learning_rate": 0.00013810069635097747, + "loss": 0.1451, + "step": 16028 + }, + { + "epoch": 0.37574500254927123, + "grad_norm": 2.013871669769287, + "learning_rate": 0.00013809388570310008, + "loss": 0.2401, + "step": 16029 + }, + { + "epoch": 0.37576844412407623, + "grad_norm": 1.0524590015411377, + "learning_rate": 0.00013808707484852417, + "loss": 0.0835, + "step": 16030 + }, + { + "epoch": 0.3757918856988813, + "grad_norm": 5.474819183349609, + "learning_rate": 0.00013808026378728672, + "loss": 0.4225, + "step": 16031 + }, + { + "epoch": 0.3758153272736863, + "grad_norm": 0.7473617792129517, + "learning_rate": 0.0001380734525194247, + "loss": 0.093, + "step": 16032 + }, + { + "epoch": 0.37583876884849127, + "grad_norm": 3.719191551208496, + "learning_rate": 0.00013806664104497505, + "loss": 0.3804, + "step": 16033 + }, + { + "epoch": 0.37586221042329626, + "grad_norm": 3.8962595462799072, + "learning_rate": 0.00013805982936397474, + "loss": 0.2024, + "step": 16034 + }, + { + "epoch": 0.37588565199810126, + "grad_norm": 5.085841178894043, + "learning_rate": 0.00013805301747646072, + "loss": 0.5078, + "step": 16035 + }, + { + "epoch": 0.37590909357290625, + "grad_norm": 3.5675582885742188, + "learning_rate": 0.00013804620538246997, + "loss": 2.1432, + "step": 16036 + }, + { + "epoch": 0.37593253514771124, + "grad_norm": 4.643934726715088, + "learning_rate": 0.00013803939308203945, + "loss": 0.3785, + "step": 16037 + }, + { + "epoch": 0.37595597672251624, + "grad_norm": 2.167999744415283, + "learning_rate": 0.00013803258057520608, + "loss": 0.2104, + "step": 16038 + }, + { + "epoch": 0.37597941829732123, + "grad_norm": 3.072540521621704, + "learning_rate": 0.00013802576786200687, + "loss": 1.4053, + "step": 16039 + }, + { + "epoch": 0.3760028598721262, + "grad_norm": 3.500328540802002, + "learning_rate": 0.00013801895494247878, + "loss": 0.35, + "step": 16040 + }, + { + "epoch": 0.3760263014469312, + "grad_norm": 3.4987640380859375, + "learning_rate": 0.00013801214181665875, + "loss": 2.2595, + "step": 16041 + }, + { + "epoch": 0.3760497430217362, + "grad_norm": 6.869725704193115, + "learning_rate": 0.00013800532848458382, + "loss": 0.3116, + "step": 16042 + }, + { + "epoch": 0.3760731845965412, + "grad_norm": 5.31258487701416, + "learning_rate": 0.00013799851494629087, + "loss": 0.3233, + "step": 16043 + }, + { + "epoch": 0.3760966261713462, + "grad_norm": 4.1493988037109375, + "learning_rate": 0.0001379917012018169, + "loss": 0.2476, + "step": 16044 + }, + { + "epoch": 0.3761200677461512, + "grad_norm": 4.083460330963135, + "learning_rate": 0.0001379848872511989, + "loss": 0.3885, + "step": 16045 + }, + { + "epoch": 0.3761435093209562, + "grad_norm": 4.17771053314209, + "learning_rate": 0.00013797807309447386, + "loss": 0.6314, + "step": 16046 + }, + { + "epoch": 0.3761669508957612, + "grad_norm": 0.9697937369346619, + "learning_rate": 0.0001379712587316787, + "loss": 0.0917, + "step": 16047 + }, + { + "epoch": 0.3761903924705662, + "grad_norm": 2.636808156967163, + "learning_rate": 0.00013796444416285043, + "loss": 0.353, + "step": 16048 + }, + { + "epoch": 0.3762138340453712, + "grad_norm": 0.42460355162620544, + "learning_rate": 0.00013795762938802604, + "loss": 0.0516, + "step": 16049 + }, + { + "epoch": 0.37623727562017617, + "grad_norm": 1.034365177154541, + "learning_rate": 0.00013795081440724246, + "loss": 0.131, + "step": 16050 + }, + { + "epoch": 0.37626071719498116, + "grad_norm": 3.9961612224578857, + "learning_rate": 0.00013794399922053668, + "loss": 0.481, + "step": 16051 + }, + { + "epoch": 0.37628415876978616, + "grad_norm": 5.200538635253906, + "learning_rate": 0.00013793718382794574, + "loss": 0.6563, + "step": 16052 + }, + { + "epoch": 0.37630760034459115, + "grad_norm": 1.4843049049377441, + "learning_rate": 0.00013793036822950657, + "loss": 0.1711, + "step": 16053 + }, + { + "epoch": 0.37633104191939615, + "grad_norm": 2.6444289684295654, + "learning_rate": 0.0001379235524252561, + "loss": 0.3309, + "step": 16054 + }, + { + "epoch": 0.37635448349420114, + "grad_norm": 1.300883173942566, + "learning_rate": 0.00013791673641523142, + "loss": 0.2017, + "step": 16055 + }, + { + "epoch": 0.37637792506900614, + "grad_norm": 3.7846922874450684, + "learning_rate": 0.00013790992019946944, + "loss": 0.4946, + "step": 16056 + }, + { + "epoch": 0.37640136664381113, + "grad_norm": 3.6920547485351562, + "learning_rate": 0.00013790310377800714, + "loss": 1.5309, + "step": 16057 + }, + { + "epoch": 0.3764248082186161, + "grad_norm": 4.462801933288574, + "learning_rate": 0.0001378962871508816, + "loss": 0.2452, + "step": 16058 + }, + { + "epoch": 0.3764482497934211, + "grad_norm": 4.726358890533447, + "learning_rate": 0.0001378894703181297, + "loss": 0.5349, + "step": 16059 + }, + { + "epoch": 0.3764716913682261, + "grad_norm": 4.521806716918945, + "learning_rate": 0.00013788265327978845, + "loss": 0.3174, + "step": 16060 + }, + { + "epoch": 0.3764951329430311, + "grad_norm": 1.7268229722976685, + "learning_rate": 0.0001378758360358949, + "loss": 0.0954, + "step": 16061 + }, + { + "epoch": 0.3765185745178361, + "grad_norm": 4.072475910186768, + "learning_rate": 0.00013786901858648597, + "loss": 0.4869, + "step": 16062 + }, + { + "epoch": 0.3765420160926411, + "grad_norm": 3.973360776901245, + "learning_rate": 0.0001378622009315987, + "loss": 0.4118, + "step": 16063 + }, + { + "epoch": 0.3765654576674461, + "grad_norm": 3.7531425952911377, + "learning_rate": 0.00013785538307127003, + "loss": 0.2573, + "step": 16064 + }, + { + "epoch": 0.3765888992422511, + "grad_norm": 3.258234739303589, + "learning_rate": 0.00013784856500553702, + "loss": 0.2584, + "step": 16065 + }, + { + "epoch": 0.3766123408170561, + "grad_norm": 1.1324998140335083, + "learning_rate": 0.0001378417467344366, + "loss": 0.065, + "step": 16066 + }, + { + "epoch": 0.3766357823918611, + "grad_norm": 3.1914761066436768, + "learning_rate": 0.00013783492825800587, + "loss": 0.3093, + "step": 16067 + }, + { + "epoch": 0.37665922396666607, + "grad_norm": 1.3385075330734253, + "learning_rate": 0.00013782810957628166, + "loss": 0.1758, + "step": 16068 + }, + { + "epoch": 0.37668266554147106, + "grad_norm": 1.19801664352417, + "learning_rate": 0.00013782129068930113, + "loss": 0.6732, + "step": 16069 + }, + { + "epoch": 0.37670610711627606, + "grad_norm": 1.6436302661895752, + "learning_rate": 0.00013781447159710118, + "loss": 0.245, + "step": 16070 + }, + { + "epoch": 0.37672954869108105, + "grad_norm": 3.667839288711548, + "learning_rate": 0.00013780765229971887, + "loss": 0.34, + "step": 16071 + }, + { + "epoch": 0.37675299026588605, + "grad_norm": 3.0641562938690186, + "learning_rate": 0.00013780083279719117, + "loss": 0.3259, + "step": 16072 + }, + { + "epoch": 0.37677643184069104, + "grad_norm": 8.101805686950684, + "learning_rate": 0.00013779401308955505, + "loss": 0.6324, + "step": 16073 + }, + { + "epoch": 0.37679987341549603, + "grad_norm": 1.2668077945709229, + "learning_rate": 0.0001377871931768476, + "loss": 0.1274, + "step": 16074 + }, + { + "epoch": 0.37682331499030103, + "grad_norm": 2.39509916305542, + "learning_rate": 0.00013778037305910577, + "loss": 0.0719, + "step": 16075 + }, + { + "epoch": 0.376846756565106, + "grad_norm": 2.914156198501587, + "learning_rate": 0.00013777355273636655, + "loss": 1.6518, + "step": 16076 + }, + { + "epoch": 0.376870198139911, + "grad_norm": 3.0359275341033936, + "learning_rate": 0.00013776673220866698, + "loss": 0.4524, + "step": 16077 + }, + { + "epoch": 0.376893639714716, + "grad_norm": 3.9047467708587646, + "learning_rate": 0.00013775991147604408, + "loss": 0.5945, + "step": 16078 + }, + { + "epoch": 0.376917081289521, + "grad_norm": 4.137332916259766, + "learning_rate": 0.00013775309053853482, + "loss": 0.5049, + "step": 16079 + }, + { + "epoch": 0.376940522864326, + "grad_norm": 5.310647010803223, + "learning_rate": 0.00013774626939617622, + "loss": 0.522, + "step": 16080 + }, + { + "epoch": 0.376963964439131, + "grad_norm": 1.976316213607788, + "learning_rate": 0.00013773944804900532, + "loss": 1.8711, + "step": 16081 + }, + { + "epoch": 0.37698740601393604, + "grad_norm": 3.513272285461426, + "learning_rate": 0.00013773262649705912, + "loss": 0.3144, + "step": 16082 + }, + { + "epoch": 0.37701084758874104, + "grad_norm": 3.272339344024658, + "learning_rate": 0.0001377258047403746, + "loss": 1.8384, + "step": 16083 + }, + { + "epoch": 0.37703428916354603, + "grad_norm": 4.309386730194092, + "learning_rate": 0.00013771898277898883, + "loss": 0.2762, + "step": 16084 + }, + { + "epoch": 0.377057730738351, + "grad_norm": 3.9947099685668945, + "learning_rate": 0.00013771216061293876, + "loss": 0.7104, + "step": 16085 + }, + { + "epoch": 0.377081172313156, + "grad_norm": 0.5855878591537476, + "learning_rate": 0.00013770533824226147, + "loss": 0.0595, + "step": 16086 + }, + { + "epoch": 0.377104613887961, + "grad_norm": 5.153533458709717, + "learning_rate": 0.00013769851566699398, + "loss": 0.3244, + "step": 16087 + }, + { + "epoch": 0.377128055462766, + "grad_norm": 2.522426128387451, + "learning_rate": 0.00013769169288717326, + "loss": 0.2582, + "step": 16088 + }, + { + "epoch": 0.377151497037571, + "grad_norm": 3.3772451877593994, + "learning_rate": 0.00013768486990283632, + "loss": 0.4047, + "step": 16089 + }, + { + "epoch": 0.377174938612376, + "grad_norm": 3.4314441680908203, + "learning_rate": 0.00013767804671402026, + "loss": 0.4946, + "step": 16090 + }, + { + "epoch": 0.377198380187181, + "grad_norm": 4.725401401519775, + "learning_rate": 0.00013767122332076207, + "loss": 0.3968, + "step": 16091 + }, + { + "epoch": 0.377221821761986, + "grad_norm": 6.530608654022217, + "learning_rate": 0.0001376643997230987, + "loss": 0.4923, + "step": 16092 + }, + { + "epoch": 0.377245263336791, + "grad_norm": 1.0490590333938599, + "learning_rate": 0.0001376575759210673, + "loss": 0.0918, + "step": 16093 + }, + { + "epoch": 0.377268704911596, + "grad_norm": 4.887416362762451, + "learning_rate": 0.00013765075191470483, + "loss": 0.7196, + "step": 16094 + }, + { + "epoch": 0.37729214648640097, + "grad_norm": 0.4083094894886017, + "learning_rate": 0.0001376439277040483, + "loss": 0.024, + "step": 16095 + }, + { + "epoch": 0.37731558806120596, + "grad_norm": 3.6085731983184814, + "learning_rate": 0.00013763710328913478, + "loss": 0.3886, + "step": 16096 + }, + { + "epoch": 0.37733902963601096, + "grad_norm": 3.183910369873047, + "learning_rate": 0.00013763027867000127, + "loss": 2.0717, + "step": 16097 + }, + { + "epoch": 0.37736247121081595, + "grad_norm": 2.832213878631592, + "learning_rate": 0.0001376234538466848, + "loss": 0.3485, + "step": 16098 + }, + { + "epoch": 0.37738591278562095, + "grad_norm": 3.0936272144317627, + "learning_rate": 0.0001376166288192224, + "loss": 1.9616, + "step": 16099 + }, + { + "epoch": 0.37740935436042594, + "grad_norm": 3.9004130363464355, + "learning_rate": 0.00013760980358765114, + "loss": 0.2947, + "step": 16100 + }, + { + "epoch": 0.37743279593523094, + "grad_norm": 3.008040428161621, + "learning_rate": 0.00013760297815200803, + "loss": 0.2718, + "step": 16101 + }, + { + "epoch": 0.37745623751003593, + "grad_norm": 3.784498453140259, + "learning_rate": 0.00013759615251233008, + "loss": 0.5741, + "step": 16102 + }, + { + "epoch": 0.3774796790848409, + "grad_norm": 1.9346106052398682, + "learning_rate": 0.0001375893266686544, + "loss": 0.1012, + "step": 16103 + }, + { + "epoch": 0.3775031206596459, + "grad_norm": 0.5536900758743286, + "learning_rate": 0.0001375825006210179, + "loss": 0.0477, + "step": 16104 + }, + { + "epoch": 0.3775265622344509, + "grad_norm": 3.444000005722046, + "learning_rate": 0.00013757567436945773, + "loss": 0.3017, + "step": 16105 + }, + { + "epoch": 0.3775500038092559, + "grad_norm": 0.8325647711753845, + "learning_rate": 0.0001375688479140109, + "loss": 0.0584, + "step": 16106 + }, + { + "epoch": 0.3775734453840609, + "grad_norm": 8.410731315612793, + "learning_rate": 0.00013756202125471447, + "loss": 1.0598, + "step": 16107 + }, + { + "epoch": 0.3775968869588659, + "grad_norm": 0.5997397899627686, + "learning_rate": 0.0001375551943916054, + "loss": 0.0685, + "step": 16108 + }, + { + "epoch": 0.3776203285336709, + "grad_norm": 3.172914743423462, + "learning_rate": 0.00013754836732472083, + "loss": 0.3762, + "step": 16109 + }, + { + "epoch": 0.3776437701084759, + "grad_norm": 3.1707890033721924, + "learning_rate": 0.00013754154005409777, + "loss": 0.2206, + "step": 16110 + }, + { + "epoch": 0.3776672116832809, + "grad_norm": 3.7945053577423096, + "learning_rate": 0.0001375347125797732, + "loss": 2.017, + "step": 16111 + }, + { + "epoch": 0.3776906532580859, + "grad_norm": 2.317293643951416, + "learning_rate": 0.00013752788490178428, + "loss": 0.2793, + "step": 16112 + }, + { + "epoch": 0.37771409483289087, + "grad_norm": 4.245415210723877, + "learning_rate": 0.000137521057020168, + "loss": 0.4889, + "step": 16113 + }, + { + "epoch": 0.37773753640769586, + "grad_norm": 2.5903496742248535, + "learning_rate": 0.00013751422893496139, + "loss": 0.3193, + "step": 16114 + }, + { + "epoch": 0.37776097798250086, + "grad_norm": 3.960461139678955, + "learning_rate": 0.00013750740064620153, + "loss": 0.4448, + "step": 16115 + }, + { + "epoch": 0.37778441955730585, + "grad_norm": 0.8882575631141663, + "learning_rate": 0.00013750057215392546, + "loss": 0.0977, + "step": 16116 + }, + { + "epoch": 0.37780786113211084, + "grad_norm": 4.29555606842041, + "learning_rate": 0.00013749374345817022, + "loss": 0.2988, + "step": 16117 + }, + { + "epoch": 0.37783130270691584, + "grad_norm": 3.8154759407043457, + "learning_rate": 0.00013748691455897286, + "loss": 0.5018, + "step": 16118 + }, + { + "epoch": 0.37785474428172083, + "grad_norm": 3.190770387649536, + "learning_rate": 0.0001374800854563705, + "loss": 0.5445, + "step": 16119 + }, + { + "epoch": 0.37787818585652583, + "grad_norm": 1.19524085521698, + "learning_rate": 0.0001374732561504001, + "loss": 0.1544, + "step": 16120 + }, + { + "epoch": 0.3779016274313308, + "grad_norm": 3.1002767086029053, + "learning_rate": 0.00013746642664109877, + "loss": 0.2971, + "step": 16121 + }, + { + "epoch": 0.3779250690061358, + "grad_norm": 2.632688522338867, + "learning_rate": 0.00013745959692850352, + "loss": 0.2817, + "step": 16122 + }, + { + "epoch": 0.3779485105809408, + "grad_norm": 2.4809536933898926, + "learning_rate": 0.00013745276701265148, + "loss": 0.3097, + "step": 16123 + }, + { + "epoch": 0.3779719521557458, + "grad_norm": 0.8428764343261719, + "learning_rate": 0.00013744593689357968, + "loss": 0.0828, + "step": 16124 + }, + { + "epoch": 0.3779953937305508, + "grad_norm": 2.7827186584472656, + "learning_rate": 0.00013743910657132517, + "loss": 0.2437, + "step": 16125 + }, + { + "epoch": 0.3780188353053558, + "grad_norm": 3.3577985763549805, + "learning_rate": 0.00013743227604592502, + "loss": 0.3046, + "step": 16126 + }, + { + "epoch": 0.3780422768801608, + "grad_norm": 3.1264336109161377, + "learning_rate": 0.00013742544531741626, + "loss": 0.4315, + "step": 16127 + }, + { + "epoch": 0.3780657184549658, + "grad_norm": 0.9532961845397949, + "learning_rate": 0.00013741861438583598, + "loss": 0.1176, + "step": 16128 + }, + { + "epoch": 0.3780891600297708, + "grad_norm": 0.48432794213294983, + "learning_rate": 0.00013741178325122128, + "loss": 0.0327, + "step": 16129 + }, + { + "epoch": 0.37811260160457577, + "grad_norm": 3.167494773864746, + "learning_rate": 0.00013740495191360915, + "loss": 0.3474, + "step": 16130 + }, + { + "epoch": 0.37813604317938077, + "grad_norm": 3.130478620529175, + "learning_rate": 0.00013739812037303674, + "loss": 0.3301, + "step": 16131 + }, + { + "epoch": 0.37815948475418576, + "grad_norm": 3.610790491104126, + "learning_rate": 0.00013739128862954106, + "loss": 1.2557, + "step": 16132 + }, + { + "epoch": 0.37818292632899075, + "grad_norm": 3.345078229904175, + "learning_rate": 0.0001373844566831592, + "loss": 0.3947, + "step": 16133 + }, + { + "epoch": 0.3782063679037958, + "grad_norm": 3.7058558464050293, + "learning_rate": 0.00013737762453392823, + "loss": 0.357, + "step": 16134 + }, + { + "epoch": 0.3782298094786008, + "grad_norm": 0.7498524785041809, + "learning_rate": 0.00013737079218188522, + "loss": 0.0674, + "step": 16135 + }, + { + "epoch": 0.3782532510534058, + "grad_norm": 3.8914990425109863, + "learning_rate": 0.00013736395962706723, + "loss": 0.6023, + "step": 16136 + }, + { + "epoch": 0.3782766926282108, + "grad_norm": 1.1172573566436768, + "learning_rate": 0.00013735712686951135, + "loss": 0.1133, + "step": 16137 + }, + { + "epoch": 0.3783001342030158, + "grad_norm": 0.5032293200492859, + "learning_rate": 0.00013735029390925466, + "loss": 0.0585, + "step": 16138 + }, + { + "epoch": 0.3783235757778208, + "grad_norm": 3.7883830070495605, + "learning_rate": 0.0001373434607463342, + "loss": 0.5178, + "step": 16139 + }, + { + "epoch": 0.37834701735262577, + "grad_norm": 0.8840785622596741, + "learning_rate": 0.0001373366273807871, + "loss": 0.074, + "step": 16140 + }, + { + "epoch": 0.37837045892743076, + "grad_norm": 2.712737560272217, + "learning_rate": 0.00013732979381265038, + "loss": 1.559, + "step": 16141 + }, + { + "epoch": 0.37839390050223576, + "grad_norm": 3.363269090652466, + "learning_rate": 0.00013732296004196116, + "loss": 0.3313, + "step": 16142 + }, + { + "epoch": 0.37841734207704075, + "grad_norm": 1.258122205734253, + "learning_rate": 0.00013731612606875652, + "loss": 0.0572, + "step": 16143 + }, + { + "epoch": 0.37844078365184575, + "grad_norm": 2.4181911945343018, + "learning_rate": 0.00013730929189307352, + "loss": 0.2229, + "step": 16144 + }, + { + "epoch": 0.37846422522665074, + "grad_norm": 2.9182968139648438, + "learning_rate": 0.00013730245751494927, + "loss": 0.2985, + "step": 16145 + }, + { + "epoch": 0.37848766680145574, + "grad_norm": 5.473477840423584, + "learning_rate": 0.00013729562293442082, + "loss": 0.54, + "step": 16146 + }, + { + "epoch": 0.37851110837626073, + "grad_norm": 3.5585498809814453, + "learning_rate": 0.00013728878815152525, + "loss": 0.2883, + "step": 16147 + }, + { + "epoch": 0.3785345499510657, + "grad_norm": 5.148156642913818, + "learning_rate": 0.00013728195316629972, + "loss": 0.5959, + "step": 16148 + }, + { + "epoch": 0.3785579915258707, + "grad_norm": 4.181961536407471, + "learning_rate": 0.00013727511797878122, + "loss": 0.2905, + "step": 16149 + }, + { + "epoch": 0.3785814331006757, + "grad_norm": 4.165693283081055, + "learning_rate": 0.0001372682825890069, + "loss": 0.6048, + "step": 16150 + }, + { + "epoch": 0.3786048746754807, + "grad_norm": 2.9414215087890625, + "learning_rate": 0.00013726144699701385, + "loss": 0.2365, + "step": 16151 + }, + { + "epoch": 0.3786283162502857, + "grad_norm": 5.078858852386475, + "learning_rate": 0.0001372546112028391, + "loss": 0.3961, + "step": 16152 + }, + { + "epoch": 0.3786517578250907, + "grad_norm": 2.559701919555664, + "learning_rate": 0.0001372477752065198, + "loss": 0.1677, + "step": 16153 + }, + { + "epoch": 0.3786751993998957, + "grad_norm": 1.036666989326477, + "learning_rate": 0.00013724093900809305, + "loss": 0.0881, + "step": 16154 + }, + { + "epoch": 0.3786986409747007, + "grad_norm": 2.994293451309204, + "learning_rate": 0.0001372341026075959, + "loss": 0.3571, + "step": 16155 + }, + { + "epoch": 0.3787220825495057, + "grad_norm": 6.10048246383667, + "learning_rate": 0.00013722726600506543, + "loss": 0.6703, + "step": 16156 + }, + { + "epoch": 0.3787455241243107, + "grad_norm": 0.840155303478241, + "learning_rate": 0.00013722042920053882, + "loss": 0.1168, + "step": 16157 + }, + { + "epoch": 0.37876896569911567, + "grad_norm": 4.148875713348389, + "learning_rate": 0.0001372135921940531, + "loss": 0.2264, + "step": 16158 + }, + { + "epoch": 0.37879240727392066, + "grad_norm": 3.7177236080169678, + "learning_rate": 0.00013720675498564533, + "loss": 0.3606, + "step": 16159 + }, + { + "epoch": 0.37881584884872566, + "grad_norm": 2.1439549922943115, + "learning_rate": 0.0001371999175753527, + "loss": 0.2722, + "step": 16160 + }, + { + "epoch": 0.37883929042353065, + "grad_norm": 2.7422306537628174, + "learning_rate": 0.00013719307996321232, + "loss": 1.5408, + "step": 16161 + }, + { + "epoch": 0.37886273199833564, + "grad_norm": 3.313086748123169, + "learning_rate": 0.00013718624214926117, + "loss": 0.3814, + "step": 16162 + }, + { + "epoch": 0.37888617357314064, + "grad_norm": 1.3460781574249268, + "learning_rate": 0.00013717940413353645, + "loss": 0.0484, + "step": 16163 + }, + { + "epoch": 0.37890961514794563, + "grad_norm": 4.15170955657959, + "learning_rate": 0.00013717256591607525, + "loss": 0.3439, + "step": 16164 + }, + { + "epoch": 0.3789330567227506, + "grad_norm": 3.092529058456421, + "learning_rate": 0.00013716572749691464, + "loss": 0.2333, + "step": 16165 + }, + { + "epoch": 0.3789564982975556, + "grad_norm": 3.188952684402466, + "learning_rate": 0.00013715888887609175, + "loss": 0.4559, + "step": 16166 + }, + { + "epoch": 0.3789799398723606, + "grad_norm": 3.1966261863708496, + "learning_rate": 0.00013715205005364372, + "loss": 0.2701, + "step": 16167 + }, + { + "epoch": 0.3790033814471656, + "grad_norm": 2.51412034034729, + "learning_rate": 0.00013714521102960758, + "loss": 0.2585, + "step": 16168 + }, + { + "epoch": 0.3790268230219706, + "grad_norm": 4.929169178009033, + "learning_rate": 0.0001371383718040205, + "loss": 0.4612, + "step": 16169 + }, + { + "epoch": 0.3790502645967756, + "grad_norm": 1.3930455446243286, + "learning_rate": 0.0001371315323769196, + "loss": 0.158, + "step": 16170 + }, + { + "epoch": 0.3790737061715806, + "grad_norm": 1.6389856338500977, + "learning_rate": 0.00013712469274834187, + "loss": 0.1562, + "step": 16171 + }, + { + "epoch": 0.3790971477463856, + "grad_norm": 0.3893011808395386, + "learning_rate": 0.00013711785291832459, + "loss": 0.0233, + "step": 16172 + }, + { + "epoch": 0.3791205893211906, + "grad_norm": 1.49137544631958, + "learning_rate": 0.00013711101288690477, + "loss": 0.1592, + "step": 16173 + }, + { + "epoch": 0.3791440308959956, + "grad_norm": 3.0399887561798096, + "learning_rate": 0.00013710417265411958, + "loss": 0.2644, + "step": 16174 + }, + { + "epoch": 0.37916747247080057, + "grad_norm": 0.4253915250301361, + "learning_rate": 0.00013709733222000607, + "loss": 0.0456, + "step": 16175 + }, + { + "epoch": 0.37919091404560556, + "grad_norm": 2.6928014755249023, + "learning_rate": 0.0001370904915846014, + "loss": 0.319, + "step": 16176 + }, + { + "epoch": 0.37921435562041056, + "grad_norm": 3.860517740249634, + "learning_rate": 0.00013708365074794268, + "loss": 0.4373, + "step": 16177 + }, + { + "epoch": 0.37923779719521555, + "grad_norm": 5.678851127624512, + "learning_rate": 0.00013707680971006704, + "loss": 0.7872, + "step": 16178 + }, + { + "epoch": 0.37926123877002055, + "grad_norm": 3.7251298427581787, + "learning_rate": 0.00013706996847101158, + "loss": 0.4463, + "step": 16179 + }, + { + "epoch": 0.37928468034482554, + "grad_norm": 2.704376220703125, + "learning_rate": 0.00013706312703081343, + "loss": 0.2669, + "step": 16180 + }, + { + "epoch": 0.37930812191963054, + "grad_norm": 2.3489179611206055, + "learning_rate": 0.0001370562853895097, + "loss": 0.2581, + "step": 16181 + }, + { + "epoch": 0.37933156349443553, + "grad_norm": 1.0292956829071045, + "learning_rate": 0.00013704944354713756, + "loss": 0.0612, + "step": 16182 + }, + { + "epoch": 0.3793550050692405, + "grad_norm": 4.01752233505249, + "learning_rate": 0.00013704260150373404, + "loss": 0.3809, + "step": 16183 + }, + { + "epoch": 0.3793784466440455, + "grad_norm": 0.9961671233177185, + "learning_rate": 0.00013703575925933637, + "loss": 0.1005, + "step": 16184 + }, + { + "epoch": 0.37940188821885057, + "grad_norm": 1.6417431831359863, + "learning_rate": 0.0001370289168139816, + "loss": 0.1343, + "step": 16185 + }, + { + "epoch": 0.37942532979365556, + "grad_norm": 3.517228364944458, + "learning_rate": 0.0001370220741677069, + "loss": 0.3034, + "step": 16186 + }, + { + "epoch": 0.37944877136846056, + "grad_norm": 2.7096328735351562, + "learning_rate": 0.00013701523132054937, + "loss": 0.2248, + "step": 16187 + }, + { + "epoch": 0.37947221294326555, + "grad_norm": 2.5604357719421387, + "learning_rate": 0.0001370083882725462, + "loss": 0.299, + "step": 16188 + }, + { + "epoch": 0.37949565451807055, + "grad_norm": 1.6275705099105835, + "learning_rate": 0.00013700154502373443, + "loss": 0.1662, + "step": 16189 + }, + { + "epoch": 0.37951909609287554, + "grad_norm": 3.073169231414795, + "learning_rate": 0.00013699470157415127, + "loss": 0.3073, + "step": 16190 + }, + { + "epoch": 0.37954253766768054, + "grad_norm": 3.2281410694122314, + "learning_rate": 0.00013698785792383376, + "loss": 0.3737, + "step": 16191 + }, + { + "epoch": 0.37956597924248553, + "grad_norm": 3.5323472023010254, + "learning_rate": 0.00013698101407281916, + "loss": 0.238, + "step": 16192 + }, + { + "epoch": 0.3795894208172905, + "grad_norm": 0.9948902726173401, + "learning_rate": 0.0001369741700211445, + "loss": 0.0548, + "step": 16193 + }, + { + "epoch": 0.3796128623920955, + "grad_norm": 2.303232431411743, + "learning_rate": 0.00013696732576884698, + "loss": 0.1368, + "step": 16194 + }, + { + "epoch": 0.3796363039669005, + "grad_norm": 4.754240989685059, + "learning_rate": 0.0001369604813159637, + "loss": 0.5016, + "step": 16195 + }, + { + "epoch": 0.3796597455417055, + "grad_norm": 1.3417577743530273, + "learning_rate": 0.0001369536366625318, + "loss": 0.111, + "step": 16196 + }, + { + "epoch": 0.3796831871165105, + "grad_norm": 1.4114990234375, + "learning_rate": 0.0001369467918085884, + "loss": 0.1837, + "step": 16197 + }, + { + "epoch": 0.3797066286913155, + "grad_norm": 2.2072913646698, + "learning_rate": 0.00013693994675417073, + "loss": 0.287, + "step": 16198 + }, + { + "epoch": 0.3797300702661205, + "grad_norm": 1.3668386936187744, + "learning_rate": 0.00013693310149931584, + "loss": 0.1238, + "step": 16199 + }, + { + "epoch": 0.3797535118409255, + "grad_norm": 3.114915609359741, + "learning_rate": 0.0001369262560440609, + "loss": 0.3069, + "step": 16200 + }, + { + "epoch": 0.3797769534157305, + "grad_norm": 2.70528244972229, + "learning_rate": 0.00013691941038844303, + "loss": 1.7891, + "step": 16201 + }, + { + "epoch": 0.3798003949905355, + "grad_norm": 0.9874001741409302, + "learning_rate": 0.00013691256453249946, + "loss": 0.1527, + "step": 16202 + }, + { + "epoch": 0.37982383656534047, + "grad_norm": 1.721257209777832, + "learning_rate": 0.00013690571847626724, + "loss": 0.2266, + "step": 16203 + }, + { + "epoch": 0.37984727814014546, + "grad_norm": 4.441885948181152, + "learning_rate": 0.00013689887221978356, + "loss": 0.439, + "step": 16204 + }, + { + "epoch": 0.37987071971495046, + "grad_norm": 2.2007007598876953, + "learning_rate": 0.00013689202576308557, + "loss": 0.1865, + "step": 16205 + }, + { + "epoch": 0.37989416128975545, + "grad_norm": 2.5560057163238525, + "learning_rate": 0.0001368851791062104, + "loss": 1.5766, + "step": 16206 + }, + { + "epoch": 0.37991760286456044, + "grad_norm": 1.8157382011413574, + "learning_rate": 0.00013687833224919522, + "loss": 0.1929, + "step": 16207 + }, + { + "epoch": 0.37994104443936544, + "grad_norm": 1.2656285762786865, + "learning_rate": 0.00013687148519207716, + "loss": 0.0997, + "step": 16208 + }, + { + "epoch": 0.37996448601417043, + "grad_norm": 0.40471360087394714, + "learning_rate": 0.00013686463793489338, + "loss": 0.0371, + "step": 16209 + }, + { + "epoch": 0.3799879275889754, + "grad_norm": 1.551815152168274, + "learning_rate": 0.00013685779047768105, + "loss": 0.1811, + "step": 16210 + }, + { + "epoch": 0.3800113691637804, + "grad_norm": 4.139993190765381, + "learning_rate": 0.00013685094282047734, + "loss": 0.4534, + "step": 16211 + }, + { + "epoch": 0.3800348107385854, + "grad_norm": 2.122810125350952, + "learning_rate": 0.00013684409496331936, + "loss": 0.9601, + "step": 16212 + }, + { + "epoch": 0.3800582523133904, + "grad_norm": 3.9771454334259033, + "learning_rate": 0.00013683724690624426, + "loss": 0.3146, + "step": 16213 + }, + { + "epoch": 0.3800816938881954, + "grad_norm": 3.561983585357666, + "learning_rate": 0.00013683039864928923, + "loss": 0.3147, + "step": 16214 + }, + { + "epoch": 0.3801051354630004, + "grad_norm": 0.3862385153770447, + "learning_rate": 0.00013682355019249145, + "loss": 0.0389, + "step": 16215 + }, + { + "epoch": 0.3801285770378054, + "grad_norm": 2.8434903621673584, + "learning_rate": 0.000136816701535888, + "loss": 0.3066, + "step": 16216 + }, + { + "epoch": 0.3801520186126104, + "grad_norm": 3.862001419067383, + "learning_rate": 0.00013680985267951613, + "loss": 0.4566, + "step": 16217 + }, + { + "epoch": 0.3801754601874154, + "grad_norm": 0.9438421130180359, + "learning_rate": 0.00013680300362341297, + "loss": 0.1172, + "step": 16218 + }, + { + "epoch": 0.3801989017622204, + "grad_norm": 2.287632942199707, + "learning_rate": 0.00013679615436761565, + "loss": 1.1706, + "step": 16219 + }, + { + "epoch": 0.38022234333702537, + "grad_norm": 3.160667657852173, + "learning_rate": 0.00013678930491216136, + "loss": 0.3583, + "step": 16220 + }, + { + "epoch": 0.38024578491183036, + "grad_norm": 2.205021381378174, + "learning_rate": 0.0001367824552570873, + "loss": 0.962, + "step": 16221 + }, + { + "epoch": 0.38026922648663536, + "grad_norm": 2.2368669509887695, + "learning_rate": 0.00013677560540243058, + "loss": 1.4033, + "step": 16222 + }, + { + "epoch": 0.38029266806144035, + "grad_norm": 1.0361542701721191, + "learning_rate": 0.0001367687553482284, + "loss": 0.0929, + "step": 16223 + }, + { + "epoch": 0.38031610963624535, + "grad_norm": 2.7102737426757812, + "learning_rate": 0.00013676190509451794, + "loss": 0.2792, + "step": 16224 + }, + { + "epoch": 0.38033955121105034, + "grad_norm": 0.5402306914329529, + "learning_rate": 0.0001367550546413363, + "loss": 0.053, + "step": 16225 + }, + { + "epoch": 0.38036299278585534, + "grad_norm": 4.060301303863525, + "learning_rate": 0.00013674820398872073, + "loss": 0.5747, + "step": 16226 + }, + { + "epoch": 0.38038643436066033, + "grad_norm": 0.7781671285629272, + "learning_rate": 0.00013674135313670836, + "loss": 0.0714, + "step": 16227 + }, + { + "epoch": 0.3804098759354653, + "grad_norm": 2.717134475708008, + "learning_rate": 0.00013673450208533642, + "loss": 0.3138, + "step": 16228 + }, + { + "epoch": 0.3804333175102703, + "grad_norm": 3.8020503520965576, + "learning_rate": 0.000136727650834642, + "loss": 0.1759, + "step": 16229 + }, + { + "epoch": 0.3804567590850753, + "grad_norm": 3.0794389247894287, + "learning_rate": 0.0001367207993846623, + "loss": 0.2371, + "step": 16230 + }, + { + "epoch": 0.3804802006598803, + "grad_norm": 0.6801706552505493, + "learning_rate": 0.00013671394773543457, + "loss": 0.0994, + "step": 16231 + }, + { + "epoch": 0.3805036422346853, + "grad_norm": 3.5062999725341797, + "learning_rate": 0.00013670709588699584, + "loss": 0.4344, + "step": 16232 + }, + { + "epoch": 0.3805270838094903, + "grad_norm": 3.022423505783081, + "learning_rate": 0.00013670024383938343, + "loss": 0.3264, + "step": 16233 + }, + { + "epoch": 0.3805505253842953, + "grad_norm": 4.718576431274414, + "learning_rate": 0.00013669339159263445, + "loss": 0.683, + "step": 16234 + }, + { + "epoch": 0.3805739669591003, + "grad_norm": 2.3149213790893555, + "learning_rate": 0.00013668653914678612, + "loss": 0.2989, + "step": 16235 + }, + { + "epoch": 0.38059740853390533, + "grad_norm": 1.1895540952682495, + "learning_rate": 0.00013667968650187557, + "loss": 0.1313, + "step": 16236 + }, + { + "epoch": 0.38062085010871033, + "grad_norm": 2.3033337593078613, + "learning_rate": 0.00013667283365794003, + "loss": 0.2807, + "step": 16237 + }, + { + "epoch": 0.3806442916835153, + "grad_norm": 2.4819509983062744, + "learning_rate": 0.00013666598061501663, + "loss": 0.3219, + "step": 16238 + }, + { + "epoch": 0.3806677332583203, + "grad_norm": 2.949753522872925, + "learning_rate": 0.0001366591273731426, + "loss": 0.3498, + "step": 16239 + }, + { + "epoch": 0.3806911748331253, + "grad_norm": 0.6850598454475403, + "learning_rate": 0.00013665227393235514, + "loss": 0.0735, + "step": 16240 + }, + { + "epoch": 0.3807146164079303, + "grad_norm": 1.1819751262664795, + "learning_rate": 0.00013664542029269137, + "loss": 0.2302, + "step": 16241 + }, + { + "epoch": 0.3807380579827353, + "grad_norm": 1.188838005065918, + "learning_rate": 0.00013663856645418856, + "loss": 0.1252, + "step": 16242 + }, + { + "epoch": 0.3807614995575403, + "grad_norm": 1.1385846138000488, + "learning_rate": 0.00013663171241688383, + "loss": 0.0655, + "step": 16243 + }, + { + "epoch": 0.3807849411323453, + "grad_norm": 3.1051714420318604, + "learning_rate": 0.0001366248581808144, + "loss": 0.2678, + "step": 16244 + }, + { + "epoch": 0.3808083827071503, + "grad_norm": 2.424823045730591, + "learning_rate": 0.00013661800374601747, + "loss": 0.2344, + "step": 16245 + }, + { + "epoch": 0.3808318242819553, + "grad_norm": 1.5188956260681152, + "learning_rate": 0.00013661114911253024, + "loss": 0.1782, + "step": 16246 + }, + { + "epoch": 0.38085526585676027, + "grad_norm": 2.8988890647888184, + "learning_rate": 0.00013660429428038987, + "loss": 0.3102, + "step": 16247 + }, + { + "epoch": 0.38087870743156527, + "grad_norm": 3.587622880935669, + "learning_rate": 0.00013659743924963354, + "loss": 0.4105, + "step": 16248 + }, + { + "epoch": 0.38090214900637026, + "grad_norm": 4.65755033493042, + "learning_rate": 0.00013659058402029853, + "loss": 0.2787, + "step": 16249 + }, + { + "epoch": 0.38092559058117526, + "grad_norm": 0.9292618036270142, + "learning_rate": 0.00013658372859242195, + "loss": 0.0827, + "step": 16250 + }, + { + "epoch": 0.38094903215598025, + "grad_norm": 1.864116907119751, + "learning_rate": 0.00013657687296604102, + "loss": 0.2125, + "step": 16251 + }, + { + "epoch": 0.38097247373078524, + "grad_norm": 4.026446342468262, + "learning_rate": 0.00013657001714119298, + "loss": 0.428, + "step": 16252 + }, + { + "epoch": 0.38099591530559024, + "grad_norm": 0.7667132019996643, + "learning_rate": 0.000136563161117915, + "loss": 0.0653, + "step": 16253 + }, + { + "epoch": 0.38101935688039523, + "grad_norm": 0.8311643600463867, + "learning_rate": 0.00013655630489624424, + "loss": 0.0754, + "step": 16254 + }, + { + "epoch": 0.3810427984552002, + "grad_norm": 3.794553756713867, + "learning_rate": 0.000136549448476218, + "loss": 0.5337, + "step": 16255 + }, + { + "epoch": 0.3810662400300052, + "grad_norm": 3.2470223903656006, + "learning_rate": 0.00013654259185787336, + "loss": 1.7815, + "step": 16256 + }, + { + "epoch": 0.3810896816048102, + "grad_norm": 4.03599739074707, + "learning_rate": 0.00013653573504124761, + "loss": 0.3406, + "step": 16257 + }, + { + "epoch": 0.3811131231796152, + "grad_norm": 3.331852912902832, + "learning_rate": 0.00013652887802637794, + "loss": 0.4869, + "step": 16258 + }, + { + "epoch": 0.3811365647544202, + "grad_norm": 4.006659507751465, + "learning_rate": 0.00013652202081330155, + "loss": 0.4183, + "step": 16259 + }, + { + "epoch": 0.3811600063292252, + "grad_norm": 11.183976173400879, + "learning_rate": 0.00013651516340205567, + "loss": 0.5886, + "step": 16260 + }, + { + "epoch": 0.3811834479040302, + "grad_norm": 2.3553965091705322, + "learning_rate": 0.00013650830579267745, + "loss": 0.1726, + "step": 16261 + }, + { + "epoch": 0.3812068894788352, + "grad_norm": 1.0628327131271362, + "learning_rate": 0.00013650144798520416, + "loss": 0.0777, + "step": 16262 + }, + { + "epoch": 0.3812303310536402, + "grad_norm": 1.3565109968185425, + "learning_rate": 0.00013649458997967297, + "loss": 0.138, + "step": 16263 + }, + { + "epoch": 0.3812537726284452, + "grad_norm": 2.40055251121521, + "learning_rate": 0.00013648773177612114, + "loss": 0.1851, + "step": 16264 + }, + { + "epoch": 0.38127721420325017, + "grad_norm": 0.9940822720527649, + "learning_rate": 0.0001364808733745858, + "loss": 0.1112, + "step": 16265 + }, + { + "epoch": 0.38130065577805516, + "grad_norm": 5.055732727050781, + "learning_rate": 0.0001364740147751043, + "loss": 0.5197, + "step": 16266 + }, + { + "epoch": 0.38132409735286016, + "grad_norm": 3.0355911254882812, + "learning_rate": 0.0001364671559777137, + "loss": 0.198, + "step": 16267 + }, + { + "epoch": 0.38134753892766515, + "grad_norm": 3.9741013050079346, + "learning_rate": 0.00013646029698245128, + "loss": 0.5411, + "step": 16268 + }, + { + "epoch": 0.38137098050247015, + "grad_norm": 2.414351463317871, + "learning_rate": 0.00013645343778935428, + "loss": 0.3267, + "step": 16269 + }, + { + "epoch": 0.38139442207727514, + "grad_norm": 2.5583057403564453, + "learning_rate": 0.0001364465783984599, + "loss": 0.3591, + "step": 16270 + }, + { + "epoch": 0.38141786365208014, + "grad_norm": 4.13235330581665, + "learning_rate": 0.0001364397188098054, + "loss": 0.5719, + "step": 16271 + }, + { + "epoch": 0.38144130522688513, + "grad_norm": 2.8674631118774414, + "learning_rate": 0.00013643285902342794, + "loss": 0.4753, + "step": 16272 + }, + { + "epoch": 0.3814647468016901, + "grad_norm": 3.6816065311431885, + "learning_rate": 0.00013642599903936474, + "loss": 0.3882, + "step": 16273 + }, + { + "epoch": 0.3814881883764951, + "grad_norm": 2.2526676654815674, + "learning_rate": 0.00013641913885765306, + "loss": 0.2359, + "step": 16274 + }, + { + "epoch": 0.3815116299513001, + "grad_norm": 1.095738410949707, + "learning_rate": 0.00013641227847833008, + "loss": 0.1825, + "step": 16275 + }, + { + "epoch": 0.3815350715261051, + "grad_norm": 2.5552523136138916, + "learning_rate": 0.0001364054179014331, + "loss": 0.3258, + "step": 16276 + }, + { + "epoch": 0.3815585131009101, + "grad_norm": 3.51991868019104, + "learning_rate": 0.00013639855712699927, + "loss": 0.3945, + "step": 16277 + }, + { + "epoch": 0.3815819546757151, + "grad_norm": 2.5761899948120117, + "learning_rate": 0.00013639169615506586, + "loss": 0.233, + "step": 16278 + }, + { + "epoch": 0.3816053962505201, + "grad_norm": 5.359420299530029, + "learning_rate": 0.00013638483498567007, + "loss": 0.5163, + "step": 16279 + }, + { + "epoch": 0.3816288378253251, + "grad_norm": 1.4934138059616089, + "learning_rate": 0.00013637797361884914, + "loss": 0.1231, + "step": 16280 + }, + { + "epoch": 0.3816522794001301, + "grad_norm": 3.1085715293884277, + "learning_rate": 0.00013637111205464032, + "loss": 0.3949, + "step": 16281 + }, + { + "epoch": 0.3816757209749351, + "grad_norm": 2.8854546546936035, + "learning_rate": 0.00013636425029308081, + "loss": 1.4088, + "step": 16282 + }, + { + "epoch": 0.38169916254974007, + "grad_norm": 2.5834600925445557, + "learning_rate": 0.00013635738833420787, + "loss": 0.1755, + "step": 16283 + }, + { + "epoch": 0.38172260412454506, + "grad_norm": 4.170153617858887, + "learning_rate": 0.0001363505261780587, + "loss": 0.4398, + "step": 16284 + }, + { + "epoch": 0.38174604569935006, + "grad_norm": 4.970542907714844, + "learning_rate": 0.00013634366382467059, + "loss": 0.4631, + "step": 16285 + }, + { + "epoch": 0.38176948727415505, + "grad_norm": 0.5626896619796753, + "learning_rate": 0.0001363368012740807, + "loss": 0.0534, + "step": 16286 + }, + { + "epoch": 0.3817929288489601, + "grad_norm": 1.9908472299575806, + "learning_rate": 0.0001363299385263263, + "loss": 0.1723, + "step": 16287 + }, + { + "epoch": 0.3818163704237651, + "grad_norm": 2.891824960708618, + "learning_rate": 0.00013632307558144466, + "loss": 0.3556, + "step": 16288 + }, + { + "epoch": 0.3818398119985701, + "grad_norm": 1.9715734720230103, + "learning_rate": 0.00013631621243947296, + "loss": 0.2063, + "step": 16289 + }, + { + "epoch": 0.3818632535733751, + "grad_norm": 2.5786995887756348, + "learning_rate": 0.0001363093491004485, + "loss": 0.2991, + "step": 16290 + }, + { + "epoch": 0.3818866951481801, + "grad_norm": 4.41608190536499, + "learning_rate": 0.0001363024855644085, + "loss": 0.8763, + "step": 16291 + }, + { + "epoch": 0.38191013672298507, + "grad_norm": 1.352283239364624, + "learning_rate": 0.00013629562183139015, + "loss": 0.1419, + "step": 16292 + }, + { + "epoch": 0.38193357829779007, + "grad_norm": 1.3420735597610474, + "learning_rate": 0.00013628875790143075, + "loss": 0.175, + "step": 16293 + }, + { + "epoch": 0.38195701987259506, + "grad_norm": 3.9797520637512207, + "learning_rate": 0.00013628189377456752, + "loss": 1.774, + "step": 16294 + }, + { + "epoch": 0.38198046144740005, + "grad_norm": 1.602140188217163, + "learning_rate": 0.00013627502945083776, + "loss": 0.1991, + "step": 16295 + }, + { + "epoch": 0.38200390302220505, + "grad_norm": 3.222541332244873, + "learning_rate": 0.00013626816493027863, + "loss": 0.4152, + "step": 16296 + }, + { + "epoch": 0.38202734459701004, + "grad_norm": 3.4927005767822266, + "learning_rate": 0.00013626130021292742, + "loss": 0.3357, + "step": 16297 + }, + { + "epoch": 0.38205078617181504, + "grad_norm": 2.9401967525482178, + "learning_rate": 0.0001362544352988214, + "loss": 0.2415, + "step": 16298 + }, + { + "epoch": 0.38207422774662003, + "grad_norm": 3.510633707046509, + "learning_rate": 0.00013624757018799774, + "loss": 0.3834, + "step": 16299 + }, + { + "epoch": 0.382097669321425, + "grad_norm": 3.735450029373169, + "learning_rate": 0.0001362407048804938, + "loss": 0.265, + "step": 16300 + }, + { + "epoch": 0.38212111089623, + "grad_norm": 4.596969127655029, + "learning_rate": 0.00013623383937634678, + "loss": 0.4665, + "step": 16301 + }, + { + "epoch": 0.382144552471035, + "grad_norm": 3.273608446121216, + "learning_rate": 0.0001362269736755939, + "loss": 0.2283, + "step": 16302 + }, + { + "epoch": 0.38216799404584, + "grad_norm": 1.9183017015457153, + "learning_rate": 0.00013622010777827247, + "loss": 0.0668, + "step": 16303 + }, + { + "epoch": 0.382191435620645, + "grad_norm": 1.107288122177124, + "learning_rate": 0.0001362132416844197, + "loss": 0.132, + "step": 16304 + }, + { + "epoch": 0.38221487719545, + "grad_norm": 5.47471284866333, + "learning_rate": 0.0001362063753940728, + "loss": 0.5161, + "step": 16305 + }, + { + "epoch": 0.382238318770255, + "grad_norm": 3.757200002670288, + "learning_rate": 0.00013619950890726918, + "loss": 0.3675, + "step": 16306 + }, + { + "epoch": 0.38226176034506, + "grad_norm": 3.116359233856201, + "learning_rate": 0.00013619264222404598, + "loss": 0.1846, + "step": 16307 + }, + { + "epoch": 0.382285201919865, + "grad_norm": 2.7466306686401367, + "learning_rate": 0.00013618577534444046, + "loss": 0.2362, + "step": 16308 + }, + { + "epoch": 0.38230864349467, + "grad_norm": 2.5912411212921143, + "learning_rate": 0.00013617890826848994, + "loss": 0.2125, + "step": 16309 + }, + { + "epoch": 0.38233208506947497, + "grad_norm": 2.562983512878418, + "learning_rate": 0.00013617204099623163, + "loss": 0.1604, + "step": 16310 + }, + { + "epoch": 0.38235552664427996, + "grad_norm": 3.0063908100128174, + "learning_rate": 0.0001361651735277028, + "loss": 0.2268, + "step": 16311 + }, + { + "epoch": 0.38237896821908496, + "grad_norm": 3.1966753005981445, + "learning_rate": 0.00013615830586294074, + "loss": 0.431, + "step": 16312 + }, + { + "epoch": 0.38240240979388995, + "grad_norm": 3.129096031188965, + "learning_rate": 0.00013615143800198268, + "loss": 0.2515, + "step": 16313 + }, + { + "epoch": 0.38242585136869495, + "grad_norm": 1.0150099992752075, + "learning_rate": 0.00013614456994486592, + "loss": 0.12, + "step": 16314 + }, + { + "epoch": 0.38244929294349994, + "grad_norm": 2.2807345390319824, + "learning_rate": 0.0001361377016916277, + "loss": 0.1759, + "step": 16315 + }, + { + "epoch": 0.38247273451830494, + "grad_norm": 1.260035753250122, + "learning_rate": 0.0001361308332423053, + "loss": 0.0474, + "step": 16316 + }, + { + "epoch": 0.38249617609310993, + "grad_norm": 0.8503665924072266, + "learning_rate": 0.00013612396459693595, + "loss": 0.0556, + "step": 16317 + }, + { + "epoch": 0.3825196176679149, + "grad_norm": 3.5331335067749023, + "learning_rate": 0.00013611709575555696, + "loss": 0.3977, + "step": 16318 + }, + { + "epoch": 0.3825430592427199, + "grad_norm": 3.8194029331207275, + "learning_rate": 0.00013611022671820565, + "loss": 0.4029, + "step": 16319 + }, + { + "epoch": 0.3825665008175249, + "grad_norm": 2.2811357975006104, + "learning_rate": 0.00013610335748491918, + "loss": 1.5499, + "step": 16320 + }, + { + "epoch": 0.3825899423923299, + "grad_norm": 3.3227853775024414, + "learning_rate": 0.0001360964880557349, + "loss": 0.2501, + "step": 16321 + }, + { + "epoch": 0.3826133839671349, + "grad_norm": 3.2050282955169678, + "learning_rate": 0.00013608961843069003, + "loss": 0.2432, + "step": 16322 + }, + { + "epoch": 0.3826368255419399, + "grad_norm": 3.0526721477508545, + "learning_rate": 0.0001360827486098219, + "loss": 0.2701, + "step": 16323 + }, + { + "epoch": 0.3826602671167449, + "grad_norm": 2.085872173309326, + "learning_rate": 0.00013607587859316774, + "loss": 0.2506, + "step": 16324 + }, + { + "epoch": 0.3826837086915499, + "grad_norm": 4.164458751678467, + "learning_rate": 0.0001360690083807649, + "loss": 0.2668, + "step": 16325 + }, + { + "epoch": 0.3827071502663549, + "grad_norm": 1.8001580238342285, + "learning_rate": 0.00013606213797265058, + "loss": 0.1306, + "step": 16326 + }, + { + "epoch": 0.3827305918411599, + "grad_norm": 2.6156585216522217, + "learning_rate": 0.00013605526736886207, + "loss": 0.2731, + "step": 16327 + }, + { + "epoch": 0.38275403341596487, + "grad_norm": 0.9894265532493591, + "learning_rate": 0.00013604839656943665, + "loss": 0.0814, + "step": 16328 + }, + { + "epoch": 0.38277747499076986, + "grad_norm": 4.5742411613464355, + "learning_rate": 0.00013604152557441165, + "loss": 0.3169, + "step": 16329 + }, + { + "epoch": 0.38280091656557486, + "grad_norm": 2.3832364082336426, + "learning_rate": 0.00013603465438382433, + "loss": 0.3223, + "step": 16330 + }, + { + "epoch": 0.38282435814037985, + "grad_norm": 2.8243768215179443, + "learning_rate": 0.00013602778299771193, + "loss": 0.269, + "step": 16331 + }, + { + "epoch": 0.38284779971518484, + "grad_norm": 4.063199043273926, + "learning_rate": 0.0001360209114161118, + "loss": 0.6488, + "step": 16332 + }, + { + "epoch": 0.38287124128998984, + "grad_norm": 2.1998348236083984, + "learning_rate": 0.00013601403963906113, + "loss": 1.0027, + "step": 16333 + }, + { + "epoch": 0.38289468286479483, + "grad_norm": 4.717857360839844, + "learning_rate": 0.00013600716766659732, + "loss": 0.8017, + "step": 16334 + }, + { + "epoch": 0.3829181244395998, + "grad_norm": 1.1864203214645386, + "learning_rate": 0.0001360002954987576, + "loss": 0.0917, + "step": 16335 + }, + { + "epoch": 0.3829415660144048, + "grad_norm": 1.0529392957687378, + "learning_rate": 0.00013599342313557925, + "loss": 0.1098, + "step": 16336 + }, + { + "epoch": 0.3829650075892098, + "grad_norm": 1.437570333480835, + "learning_rate": 0.00013598655057709959, + "loss": 0.1706, + "step": 16337 + }, + { + "epoch": 0.38298844916401487, + "grad_norm": 6.635241508483887, + "learning_rate": 0.00013597967782335587, + "loss": 0.4184, + "step": 16338 + }, + { + "epoch": 0.38301189073881986, + "grad_norm": 1.9393317699432373, + "learning_rate": 0.00013597280487438545, + "loss": 0.7978, + "step": 16339 + }, + { + "epoch": 0.38303533231362485, + "grad_norm": 3.561107873916626, + "learning_rate": 0.00013596593173022554, + "loss": 0.4378, + "step": 16340 + }, + { + "epoch": 0.38305877388842985, + "grad_norm": 4.152292251586914, + "learning_rate": 0.00013595905839091347, + "loss": 0.3207, + "step": 16341 + }, + { + "epoch": 0.38308221546323484, + "grad_norm": 0.9842601418495178, + "learning_rate": 0.00013595218485648657, + "loss": 0.1474, + "step": 16342 + }, + { + "epoch": 0.38310565703803984, + "grad_norm": 2.0040736198425293, + "learning_rate": 0.0001359453111269821, + "loss": 0.1208, + "step": 16343 + }, + { + "epoch": 0.38312909861284483, + "grad_norm": 2.9786734580993652, + "learning_rate": 0.00013593843720243736, + "loss": 0.2866, + "step": 16344 + }, + { + "epoch": 0.3831525401876498, + "grad_norm": 5.577361106872559, + "learning_rate": 0.00013593156308288964, + "loss": 0.848, + "step": 16345 + }, + { + "epoch": 0.3831759817624548, + "grad_norm": 2.9079649448394775, + "learning_rate": 0.00013592468876837623, + "loss": 0.428, + "step": 16346 + }, + { + "epoch": 0.3831994233372598, + "grad_norm": 3.755981922149658, + "learning_rate": 0.00013591781425893444, + "loss": 0.3473, + "step": 16347 + }, + { + "epoch": 0.3832228649120648, + "grad_norm": 1.1952459812164307, + "learning_rate": 0.00013591093955460162, + "loss": 0.1266, + "step": 16348 + }, + { + "epoch": 0.3832463064868698, + "grad_norm": 4.8267903327941895, + "learning_rate": 0.00013590406465541503, + "loss": 0.5981, + "step": 16349 + }, + { + "epoch": 0.3832697480616748, + "grad_norm": 2.7891182899475098, + "learning_rate": 0.00013589718956141195, + "loss": 0.2508, + "step": 16350 + }, + { + "epoch": 0.3832931896364798, + "grad_norm": 2.6476902961730957, + "learning_rate": 0.00013589031427262973, + "loss": 0.177, + "step": 16351 + }, + { + "epoch": 0.3833166312112848, + "grad_norm": 0.5685988068580627, + "learning_rate": 0.00013588343878910566, + "loss": 0.0515, + "step": 16352 + }, + { + "epoch": 0.3833400727860898, + "grad_norm": 0.7451452612876892, + "learning_rate": 0.00013587656311087702, + "loss": 0.0624, + "step": 16353 + }, + { + "epoch": 0.3833635143608948, + "grad_norm": 3.694077253341675, + "learning_rate": 0.00013586968723798115, + "loss": 0.5802, + "step": 16354 + }, + { + "epoch": 0.38338695593569977, + "grad_norm": 4.233215808868408, + "learning_rate": 0.00013586281117045537, + "loss": 0.3529, + "step": 16355 + }, + { + "epoch": 0.38341039751050476, + "grad_norm": 4.618698596954346, + "learning_rate": 0.0001358559349083369, + "loss": 0.5405, + "step": 16356 + }, + { + "epoch": 0.38343383908530976, + "grad_norm": 2.827437162399292, + "learning_rate": 0.0001358490584516632, + "loss": 0.2587, + "step": 16357 + }, + { + "epoch": 0.38345728066011475, + "grad_norm": 2.578678607940674, + "learning_rate": 0.0001358421818004715, + "loss": 0.3752, + "step": 16358 + }, + { + "epoch": 0.38348072223491975, + "grad_norm": 3.3801681995391846, + "learning_rate": 0.00013583530495479905, + "loss": 0.4528, + "step": 16359 + }, + { + "epoch": 0.38350416380972474, + "grad_norm": 3.017542839050293, + "learning_rate": 0.00013582842791468324, + "loss": 1.5998, + "step": 16360 + }, + { + "epoch": 0.38352760538452974, + "grad_norm": 1.1843785047531128, + "learning_rate": 0.0001358215506801614, + "loss": 0.1313, + "step": 16361 + }, + { + "epoch": 0.38355104695933473, + "grad_norm": 0.6874520778656006, + "learning_rate": 0.0001358146732512708, + "loss": 0.0632, + "step": 16362 + }, + { + "epoch": 0.3835744885341397, + "grad_norm": 4.085703372955322, + "learning_rate": 0.00013580779562804882, + "loss": 1.6706, + "step": 16363 + }, + { + "epoch": 0.3835979301089447, + "grad_norm": 4.672416687011719, + "learning_rate": 0.0001358009178105327, + "loss": 0.57, + "step": 16364 + }, + { + "epoch": 0.3836213716837497, + "grad_norm": 0.6929853558540344, + "learning_rate": 0.00013579403979875978, + "loss": 0.0465, + "step": 16365 + }, + { + "epoch": 0.3836448132585547, + "grad_norm": 3.4063992500305176, + "learning_rate": 0.0001357871615927674, + "loss": 0.336, + "step": 16366 + }, + { + "epoch": 0.3836682548333597, + "grad_norm": 3.293985605239868, + "learning_rate": 0.00013578028319259292, + "loss": 0.5287, + "step": 16367 + }, + { + "epoch": 0.3836916964081647, + "grad_norm": 0.6362132430076599, + "learning_rate": 0.00013577340459827356, + "loss": 0.0606, + "step": 16368 + }, + { + "epoch": 0.3837151379829697, + "grad_norm": 3.9422607421875, + "learning_rate": 0.0001357665258098467, + "loss": 0.3697, + "step": 16369 + }, + { + "epoch": 0.3837385795577747, + "grad_norm": 3.0885169506073, + "learning_rate": 0.00013575964682734968, + "loss": 0.3417, + "step": 16370 + }, + { + "epoch": 0.3837620211325797, + "grad_norm": 0.5580852031707764, + "learning_rate": 0.00013575276765081983, + "loss": 0.0716, + "step": 16371 + }, + { + "epoch": 0.3837854627073847, + "grad_norm": 2.773806095123291, + "learning_rate": 0.00013574588828029445, + "loss": 0.32, + "step": 16372 + }, + { + "epoch": 0.38380890428218967, + "grad_norm": 4.602235794067383, + "learning_rate": 0.00013573900871581088, + "loss": 0.4366, + "step": 16373 + }, + { + "epoch": 0.38383234585699466, + "grad_norm": 3.1989388465881348, + "learning_rate": 0.00013573212895740645, + "loss": 0.1862, + "step": 16374 + }, + { + "epoch": 0.38385578743179966, + "grad_norm": 4.120454788208008, + "learning_rate": 0.00013572524900511845, + "loss": 0.4697, + "step": 16375 + }, + { + "epoch": 0.38387922900660465, + "grad_norm": 1.1618489027023315, + "learning_rate": 0.00013571836885898427, + "loss": 0.191, + "step": 16376 + }, + { + "epoch": 0.38390267058140964, + "grad_norm": 0.9477548599243164, + "learning_rate": 0.00013571148851904122, + "loss": 0.0637, + "step": 16377 + }, + { + "epoch": 0.38392611215621464, + "grad_norm": 3.657416820526123, + "learning_rate": 0.00013570460798532658, + "loss": 0.2917, + "step": 16378 + }, + { + "epoch": 0.38394955373101963, + "grad_norm": 1.889237880706787, + "learning_rate": 0.00013569772725787778, + "loss": 0.1407, + "step": 16379 + }, + { + "epoch": 0.3839729953058246, + "grad_norm": 1.6758313179016113, + "learning_rate": 0.0001356908463367321, + "loss": 0.2392, + "step": 16380 + }, + { + "epoch": 0.3839964368806296, + "grad_norm": 2.800539255142212, + "learning_rate": 0.00013568396522192688, + "loss": 1.743, + "step": 16381 + }, + { + "epoch": 0.3840198784554346, + "grad_norm": 4.9308600425720215, + "learning_rate": 0.00013567708391349948, + "loss": 0.4827, + "step": 16382 + }, + { + "epoch": 0.3840433200302396, + "grad_norm": 4.811861515045166, + "learning_rate": 0.0001356702024114872, + "loss": 0.4047, + "step": 16383 + }, + { + "epoch": 0.3840667616050446, + "grad_norm": 0.7548744678497314, + "learning_rate": 0.00013566332071592737, + "loss": 0.1228, + "step": 16384 + }, + { + "epoch": 0.3840902031798496, + "grad_norm": 3.2430508136749268, + "learning_rate": 0.0001356564388268574, + "loss": 0.343, + "step": 16385 + }, + { + "epoch": 0.3841136447546546, + "grad_norm": 4.875413417816162, + "learning_rate": 0.00013564955674431458, + "loss": 0.3921, + "step": 16386 + }, + { + "epoch": 0.3841370863294596, + "grad_norm": 3.376072406768799, + "learning_rate": 0.0001356426744683363, + "loss": 0.1624, + "step": 16387 + }, + { + "epoch": 0.3841605279042646, + "grad_norm": 0.6775119304656982, + "learning_rate": 0.0001356357919989598, + "loss": 0.0533, + "step": 16388 + }, + { + "epoch": 0.38418396947906963, + "grad_norm": 1.5108174085617065, + "learning_rate": 0.00013562890933622256, + "loss": 0.0915, + "step": 16389 + }, + { + "epoch": 0.3842074110538746, + "grad_norm": 3.179203748703003, + "learning_rate": 0.00013562202648016177, + "loss": 0.3232, + "step": 16390 + }, + { + "epoch": 0.3842308526286796, + "grad_norm": 2.6438355445861816, + "learning_rate": 0.00013561514343081492, + "loss": 0.2585, + "step": 16391 + }, + { + "epoch": 0.3842542942034846, + "grad_norm": 5.276031494140625, + "learning_rate": 0.00013560826018821928, + "loss": 0.7312, + "step": 16392 + }, + { + "epoch": 0.3842777357782896, + "grad_norm": 1.5407499074935913, + "learning_rate": 0.00013560137675241226, + "loss": 0.8488, + "step": 16393 + }, + { + "epoch": 0.3843011773530946, + "grad_norm": 4.398542881011963, + "learning_rate": 0.00013559449312343112, + "loss": 0.4468, + "step": 16394 + }, + { + "epoch": 0.3843246189278996, + "grad_norm": 4.385147571563721, + "learning_rate": 0.0001355876093013133, + "loss": 0.3205, + "step": 16395 + }, + { + "epoch": 0.3843480605027046, + "grad_norm": 5.809005260467529, + "learning_rate": 0.00013558072528609607, + "loss": 0.8192, + "step": 16396 + }, + { + "epoch": 0.3843715020775096, + "grad_norm": 1.0062453746795654, + "learning_rate": 0.00013557384107781682, + "loss": 0.1203, + "step": 16397 + }, + { + "epoch": 0.3843949436523146, + "grad_norm": 0.9550656080245972, + "learning_rate": 0.00013556695667651293, + "loss": 0.1662, + "step": 16398 + }, + { + "epoch": 0.3844183852271196, + "grad_norm": 2.40665340423584, + "learning_rate": 0.00013556007208222174, + "loss": 1.2291, + "step": 16399 + }, + { + "epoch": 0.38444182680192457, + "grad_norm": 3.800236701965332, + "learning_rate": 0.00013555318729498059, + "loss": 0.5163, + "step": 16400 + }, + { + "epoch": 0.38446526837672956, + "grad_norm": 1.3428528308868408, + "learning_rate": 0.00013554630231482685, + "loss": 0.1308, + "step": 16401 + }, + { + "epoch": 0.38448870995153456, + "grad_norm": 1.0974308252334595, + "learning_rate": 0.0001355394171417979, + "loss": 0.1427, + "step": 16402 + }, + { + "epoch": 0.38451215152633955, + "grad_norm": 1.4131373167037964, + "learning_rate": 0.00013553253177593105, + "loss": 0.1881, + "step": 16403 + }, + { + "epoch": 0.38453559310114455, + "grad_norm": 2.353717803955078, + "learning_rate": 0.00013552564621726365, + "loss": 0.2197, + "step": 16404 + }, + { + "epoch": 0.38455903467594954, + "grad_norm": 4.329951763153076, + "learning_rate": 0.00013551876046583313, + "loss": 0.5183, + "step": 16405 + }, + { + "epoch": 0.38458247625075453, + "grad_norm": 2.8857851028442383, + "learning_rate": 0.00013551187452167683, + "loss": 1.6342, + "step": 16406 + }, + { + "epoch": 0.38460591782555953, + "grad_norm": 0.8953196406364441, + "learning_rate": 0.00013550498838483206, + "loss": 0.0762, + "step": 16407 + }, + { + "epoch": 0.3846293594003645, + "grad_norm": 3.8499011993408203, + "learning_rate": 0.00013549810205533625, + "loss": 0.3854, + "step": 16408 + }, + { + "epoch": 0.3846528009751695, + "grad_norm": 3.9019081592559814, + "learning_rate": 0.00013549121553322674, + "loss": 0.3629, + "step": 16409 + }, + { + "epoch": 0.3846762425499745, + "grad_norm": 4.057740211486816, + "learning_rate": 0.00013548432881854087, + "loss": 1.7446, + "step": 16410 + }, + { + "epoch": 0.3846996841247795, + "grad_norm": 4.419573783874512, + "learning_rate": 0.00013547744191131608, + "loss": 0.4806, + "step": 16411 + }, + { + "epoch": 0.3847231256995845, + "grad_norm": 3.277789831161499, + "learning_rate": 0.00013547055481158968, + "loss": 0.3554, + "step": 16412 + }, + { + "epoch": 0.3847465672743895, + "grad_norm": 2.48433518409729, + "learning_rate": 0.00013546366751939902, + "loss": 0.2236, + "step": 16413 + }, + { + "epoch": 0.3847700088491945, + "grad_norm": 2.7350449562072754, + "learning_rate": 0.00013545678003478153, + "loss": 0.2371, + "step": 16414 + }, + { + "epoch": 0.3847934504239995, + "grad_norm": 3.5866007804870605, + "learning_rate": 0.00013544989235777456, + "loss": 0.3447, + "step": 16415 + }, + { + "epoch": 0.3848168919988045, + "grad_norm": 2.2776780128479004, + "learning_rate": 0.00013544300448841546, + "loss": 0.255, + "step": 16416 + }, + { + "epoch": 0.38484033357360947, + "grad_norm": 2.705155611038208, + "learning_rate": 0.00013543611642674163, + "loss": 0.3689, + "step": 16417 + }, + { + "epoch": 0.38486377514841447, + "grad_norm": 5.383171081542969, + "learning_rate": 0.00013542922817279046, + "loss": 0.6611, + "step": 16418 + }, + { + "epoch": 0.38488721672321946, + "grad_norm": 4.046764850616455, + "learning_rate": 0.00013542233972659926, + "loss": 0.3046, + "step": 16419 + }, + { + "epoch": 0.38491065829802446, + "grad_norm": 2.413944959640503, + "learning_rate": 0.00013541545108820545, + "loss": 1.2032, + "step": 16420 + }, + { + "epoch": 0.38493409987282945, + "grad_norm": 0.8014498353004456, + "learning_rate": 0.00013540856225764642, + "loss": 0.1113, + "step": 16421 + }, + { + "epoch": 0.38495754144763444, + "grad_norm": 4.355319023132324, + "learning_rate": 0.00013540167323495956, + "loss": 0.5938, + "step": 16422 + }, + { + "epoch": 0.38498098302243944, + "grad_norm": 0.8668460249900818, + "learning_rate": 0.00013539478402018216, + "loss": 0.1036, + "step": 16423 + }, + { + "epoch": 0.38500442459724443, + "grad_norm": 2.7448413372039795, + "learning_rate": 0.00013538789461335174, + "loss": 0.2306, + "step": 16424 + }, + { + "epoch": 0.3850278661720494, + "grad_norm": 3.655606508255005, + "learning_rate": 0.00013538100501450559, + "loss": 0.345, + "step": 16425 + }, + { + "epoch": 0.3850513077468544, + "grad_norm": 0.9544044137001038, + "learning_rate": 0.00013537411522368106, + "loss": 0.115, + "step": 16426 + }, + { + "epoch": 0.3850747493216594, + "grad_norm": 2.1790554523468018, + "learning_rate": 0.00013536722524091566, + "loss": 0.2976, + "step": 16427 + }, + { + "epoch": 0.3850981908964644, + "grad_norm": 1.1187093257904053, + "learning_rate": 0.00013536033506624664, + "loss": 0.1445, + "step": 16428 + }, + { + "epoch": 0.3851216324712694, + "grad_norm": 3.0596137046813965, + "learning_rate": 0.00013535344469971147, + "loss": 0.2958, + "step": 16429 + }, + { + "epoch": 0.3851450740460744, + "grad_norm": 3.0699307918548584, + "learning_rate": 0.00013534655414134753, + "loss": 0.3535, + "step": 16430 + }, + { + "epoch": 0.3851685156208794, + "grad_norm": 2.9767563343048096, + "learning_rate": 0.0001353396633911922, + "loss": 0.3414, + "step": 16431 + }, + { + "epoch": 0.3851919571956844, + "grad_norm": 3.1083550453186035, + "learning_rate": 0.00013533277244928284, + "loss": 0.4301, + "step": 16432 + }, + { + "epoch": 0.3852153987704894, + "grad_norm": 3.3514926433563232, + "learning_rate": 0.00013532588131565687, + "loss": 0.34, + "step": 16433 + }, + { + "epoch": 0.3852388403452944, + "grad_norm": 3.7910099029541016, + "learning_rate": 0.00013531898999035168, + "loss": 0.4393, + "step": 16434 + }, + { + "epoch": 0.38526228192009937, + "grad_norm": 4.13807487487793, + "learning_rate": 0.00013531209847340465, + "loss": 0.5963, + "step": 16435 + }, + { + "epoch": 0.38528572349490436, + "grad_norm": 0.81156325340271, + "learning_rate": 0.00013530520676485318, + "loss": 0.0935, + "step": 16436 + }, + { + "epoch": 0.38530916506970936, + "grad_norm": 3.0868852138519287, + "learning_rate": 0.00013529831486473469, + "loss": 0.3231, + "step": 16437 + }, + { + "epoch": 0.38533260664451435, + "grad_norm": 1.250374436378479, + "learning_rate": 0.0001352914227730865, + "loss": 0.1312, + "step": 16438 + }, + { + "epoch": 0.38535604821931935, + "grad_norm": 1.651450753211975, + "learning_rate": 0.00013528453048994608, + "loss": 0.1561, + "step": 16439 + }, + { + "epoch": 0.3853794897941244, + "grad_norm": 2.5560498237609863, + "learning_rate": 0.00013527763801535084, + "loss": 0.2563, + "step": 16440 + }, + { + "epoch": 0.3854029313689294, + "grad_norm": 3.9980082511901855, + "learning_rate": 0.0001352707453493381, + "loss": 0.3448, + "step": 16441 + }, + { + "epoch": 0.3854263729437344, + "grad_norm": 1.0649173259735107, + "learning_rate": 0.0001352638524919453, + "loss": 0.1152, + "step": 16442 + }, + { + "epoch": 0.3854498145185394, + "grad_norm": 2.1649045944213867, + "learning_rate": 0.00013525695944320988, + "loss": 1.229, + "step": 16443 + }, + { + "epoch": 0.3854732560933444, + "grad_norm": 2.629601001739502, + "learning_rate": 0.00013525006620316923, + "loss": 0.2398, + "step": 16444 + }, + { + "epoch": 0.38549669766814937, + "grad_norm": 3.3736791610717773, + "learning_rate": 0.00013524317277186065, + "loss": 0.2109, + "step": 16445 + }, + { + "epoch": 0.38552013924295436, + "grad_norm": 3.0886669158935547, + "learning_rate": 0.00013523627914932166, + "loss": 0.2535, + "step": 16446 + }, + { + "epoch": 0.38554358081775936, + "grad_norm": 0.8458542823791504, + "learning_rate": 0.00013522938533558967, + "loss": 0.1155, + "step": 16447 + }, + { + "epoch": 0.38556702239256435, + "grad_norm": 4.7477216720581055, + "learning_rate": 0.00013522249133070198, + "loss": 0.2081, + "step": 16448 + }, + { + "epoch": 0.38559046396736935, + "grad_norm": 1.3244565725326538, + "learning_rate": 0.0001352155971346961, + "loss": 0.1447, + "step": 16449 + }, + { + "epoch": 0.38561390554217434, + "grad_norm": 3.328871965408325, + "learning_rate": 0.0001352087027476094, + "loss": 0.397, + "step": 16450 + }, + { + "epoch": 0.38563734711697933, + "grad_norm": 0.9344071745872498, + "learning_rate": 0.00013520180816947925, + "loss": 0.0463, + "step": 16451 + }, + { + "epoch": 0.38566078869178433, + "grad_norm": 4.377514362335205, + "learning_rate": 0.00013519491340034315, + "loss": 0.4923, + "step": 16452 + }, + { + "epoch": 0.3856842302665893, + "grad_norm": 3.5870747566223145, + "learning_rate": 0.00013518801844023844, + "loss": 0.3656, + "step": 16453 + }, + { + "epoch": 0.3857076718413943, + "grad_norm": 4.286046504974365, + "learning_rate": 0.00013518112328920253, + "loss": 0.556, + "step": 16454 + }, + { + "epoch": 0.3857311134161993, + "grad_norm": 3.565284490585327, + "learning_rate": 0.00013517422794727288, + "loss": 0.3303, + "step": 16455 + }, + { + "epoch": 0.3857545549910043, + "grad_norm": 5.621853828430176, + "learning_rate": 0.0001351673324144869, + "loss": 0.4466, + "step": 16456 + }, + { + "epoch": 0.3857779965658093, + "grad_norm": 3.5373945236206055, + "learning_rate": 0.00013516043669088194, + "loss": 0.5025, + "step": 16457 + }, + { + "epoch": 0.3858014381406143, + "grad_norm": 1.3964265584945679, + "learning_rate": 0.00013515354077649547, + "loss": 0.1421, + "step": 16458 + }, + { + "epoch": 0.3858248797154193, + "grad_norm": 2.1230742931365967, + "learning_rate": 0.00013514664467136493, + "loss": 1.2817, + "step": 16459 + }, + { + "epoch": 0.3858483212902243, + "grad_norm": 1.4820033311843872, + "learning_rate": 0.0001351397483755277, + "loss": 0.1516, + "step": 16460 + }, + { + "epoch": 0.3858717628650293, + "grad_norm": 3.563136100769043, + "learning_rate": 0.00013513285188902118, + "loss": 0.3471, + "step": 16461 + }, + { + "epoch": 0.38589520443983427, + "grad_norm": 4.742204189300537, + "learning_rate": 0.00013512595521188284, + "loss": 0.3912, + "step": 16462 + }, + { + "epoch": 0.38591864601463927, + "grad_norm": 3.256707191467285, + "learning_rate": 0.00013511905834415007, + "loss": 1.531, + "step": 16463 + }, + { + "epoch": 0.38594208758944426, + "grad_norm": 3.5676348209381104, + "learning_rate": 0.0001351121612858603, + "loss": 0.2962, + "step": 16464 + }, + { + "epoch": 0.38596552916424925, + "grad_norm": 2.6097335815429688, + "learning_rate": 0.00013510526403705096, + "loss": 0.1896, + "step": 16465 + }, + { + "epoch": 0.38598897073905425, + "grad_norm": 2.8419313430786133, + "learning_rate": 0.0001350983665977595, + "loss": 0.3821, + "step": 16466 + }, + { + "epoch": 0.38601241231385924, + "grad_norm": 1.3851009607315063, + "learning_rate": 0.00013509146896802327, + "loss": 0.2304, + "step": 16467 + }, + { + "epoch": 0.38603585388866424, + "grad_norm": 3.6196911334991455, + "learning_rate": 0.00013508457114787977, + "loss": 0.435, + "step": 16468 + }, + { + "epoch": 0.38605929546346923, + "grad_norm": 2.220179796218872, + "learning_rate": 0.0001350776731373664, + "loss": 0.1879, + "step": 16469 + }, + { + "epoch": 0.3860827370382742, + "grad_norm": 4.42122220993042, + "learning_rate": 0.00013507077493652057, + "loss": 0.3477, + "step": 16470 + }, + { + "epoch": 0.3861061786130792, + "grad_norm": 6.0119781494140625, + "learning_rate": 0.00013506387654537973, + "loss": 0.489, + "step": 16471 + }, + { + "epoch": 0.3861296201878842, + "grad_norm": 5.170915603637695, + "learning_rate": 0.00013505697796398132, + "loss": 0.474, + "step": 16472 + }, + { + "epoch": 0.3861530617626892, + "grad_norm": 4.386557102203369, + "learning_rate": 0.00013505007919236277, + "loss": 0.4918, + "step": 16473 + }, + { + "epoch": 0.3861765033374942, + "grad_norm": 3.3698062896728516, + "learning_rate": 0.00013504318023056151, + "loss": 0.4167, + "step": 16474 + }, + { + "epoch": 0.3861999449122992, + "grad_norm": 0.9858154654502869, + "learning_rate": 0.00013503628107861495, + "loss": 0.0908, + "step": 16475 + }, + { + "epoch": 0.3862233864871042, + "grad_norm": 1.298831820487976, + "learning_rate": 0.00013502938173656055, + "loss": 0.1355, + "step": 16476 + }, + { + "epoch": 0.3862468280619092, + "grad_norm": 1.5687161684036255, + "learning_rate": 0.00013502248220443575, + "loss": 0.1071, + "step": 16477 + }, + { + "epoch": 0.3862702696367142, + "grad_norm": 1.8537577390670776, + "learning_rate": 0.00013501558248227795, + "loss": 0.1298, + "step": 16478 + }, + { + "epoch": 0.3862937112115192, + "grad_norm": 4.347967624664307, + "learning_rate": 0.00013500868257012465, + "loss": 0.4944, + "step": 16479 + }, + { + "epoch": 0.38631715278632417, + "grad_norm": 4.360519886016846, + "learning_rate": 0.00013500178246801322, + "loss": 0.309, + "step": 16480 + }, + { + "epoch": 0.38634059436112916, + "grad_norm": 4.710981845855713, + "learning_rate": 0.00013499488217598113, + "loss": 0.4412, + "step": 16481 + }, + { + "epoch": 0.38636403593593416, + "grad_norm": 5.785431385040283, + "learning_rate": 0.00013498798169406587, + "loss": 0.4308, + "step": 16482 + }, + { + "epoch": 0.38638747751073915, + "grad_norm": 4.140587329864502, + "learning_rate": 0.0001349810810223048, + "loss": 0.3473, + "step": 16483 + }, + { + "epoch": 0.38641091908554415, + "grad_norm": 5.928987979888916, + "learning_rate": 0.00013497418016073542, + "loss": 0.5941, + "step": 16484 + }, + { + "epoch": 0.38643436066034914, + "grad_norm": 3.9743175506591797, + "learning_rate": 0.00013496727910939515, + "loss": 0.492, + "step": 16485 + }, + { + "epoch": 0.38645780223515414, + "grad_norm": 2.2117443084716797, + "learning_rate": 0.00013496037786832143, + "loss": 0.1599, + "step": 16486 + }, + { + "epoch": 0.38648124380995913, + "grad_norm": 1.9295605421066284, + "learning_rate": 0.00013495347643755173, + "loss": 0.2261, + "step": 16487 + }, + { + "epoch": 0.3865046853847641, + "grad_norm": 2.195378303527832, + "learning_rate": 0.00013494657481712347, + "loss": 0.1861, + "step": 16488 + }, + { + "epoch": 0.3865281269595691, + "grad_norm": 2.917914867401123, + "learning_rate": 0.00013493967300707413, + "loss": 0.2327, + "step": 16489 + }, + { + "epoch": 0.3865515685343741, + "grad_norm": 0.8024557828903198, + "learning_rate": 0.00013493277100744112, + "loss": 0.0778, + "step": 16490 + }, + { + "epoch": 0.38657501010917916, + "grad_norm": 3.4793436527252197, + "learning_rate": 0.00013492586881826195, + "loss": 0.3304, + "step": 16491 + }, + { + "epoch": 0.38659845168398416, + "grad_norm": 2.441683053970337, + "learning_rate": 0.000134918966439574, + "loss": 0.2397, + "step": 16492 + }, + { + "epoch": 0.38662189325878915, + "grad_norm": 3.31260085105896, + "learning_rate": 0.00013491206387141475, + "loss": 0.1967, + "step": 16493 + }, + { + "epoch": 0.38664533483359415, + "grad_norm": 4.768126964569092, + "learning_rate": 0.0001349051611138217, + "loss": 0.8012, + "step": 16494 + }, + { + "epoch": 0.38666877640839914, + "grad_norm": 2.871687412261963, + "learning_rate": 0.00013489825816683223, + "loss": 0.2064, + "step": 16495 + }, + { + "epoch": 0.38669221798320413, + "grad_norm": 1.228818655014038, + "learning_rate": 0.00013489135503048382, + "loss": 0.0797, + "step": 16496 + }, + { + "epoch": 0.38671565955800913, + "grad_norm": 1.3423351049423218, + "learning_rate": 0.00013488445170481398, + "loss": 0.1856, + "step": 16497 + }, + { + "epoch": 0.3867391011328141, + "grad_norm": 5.19501256942749, + "learning_rate": 0.0001348775481898601, + "loss": 0.4391, + "step": 16498 + }, + { + "epoch": 0.3867625427076191, + "grad_norm": 1.4696385860443115, + "learning_rate": 0.00013487064448565965, + "loss": 1.4981, + "step": 16499 + }, + { + "epoch": 0.3867859842824241, + "grad_norm": 4.2951226234436035, + "learning_rate": 0.0001348637405922501, + "loss": 0.68, + "step": 16500 + }, + { + "epoch": 0.3868094258572291, + "grad_norm": 2.3379030227661133, + "learning_rate": 0.00013485683650966897, + "loss": 0.1499, + "step": 16501 + }, + { + "epoch": 0.3868328674320341, + "grad_norm": 3.564446449279785, + "learning_rate": 0.0001348499322379536, + "loss": 0.3468, + "step": 16502 + }, + { + "epoch": 0.3868563090068391, + "grad_norm": 0.7839107513427734, + "learning_rate": 0.00013484302777714156, + "loss": 0.0906, + "step": 16503 + }, + { + "epoch": 0.3868797505816441, + "grad_norm": 0.4458625018596649, + "learning_rate": 0.00013483612312727026, + "loss": 0.0402, + "step": 16504 + }, + { + "epoch": 0.3869031921564491, + "grad_norm": 3.168168067932129, + "learning_rate": 0.00013482921828837717, + "loss": 0.3983, + "step": 16505 + }, + { + "epoch": 0.3869266337312541, + "grad_norm": 3.73974871635437, + "learning_rate": 0.00013482231326049973, + "loss": 0.3989, + "step": 16506 + }, + { + "epoch": 0.38695007530605907, + "grad_norm": 4.174831867218018, + "learning_rate": 0.00013481540804367547, + "loss": 0.2369, + "step": 16507 + }, + { + "epoch": 0.38697351688086407, + "grad_norm": 2.1469452381134033, + "learning_rate": 0.00013480850263794183, + "loss": 1.3382, + "step": 16508 + }, + { + "epoch": 0.38699695845566906, + "grad_norm": 0.9320580363273621, + "learning_rate": 0.00013480159704333628, + "loss": 0.1327, + "step": 16509 + }, + { + "epoch": 0.38702040003047405, + "grad_norm": 1.9869880676269531, + "learning_rate": 0.0001347946912598963, + "loss": 0.1147, + "step": 16510 + }, + { + "epoch": 0.38704384160527905, + "grad_norm": 2.206242322921753, + "learning_rate": 0.00013478778528765932, + "loss": 1.814, + "step": 16511 + }, + { + "epoch": 0.38706728318008404, + "grad_norm": 2.602928876876831, + "learning_rate": 0.00013478087912666283, + "loss": 0.2997, + "step": 16512 + }, + { + "epoch": 0.38709072475488904, + "grad_norm": 1.2016311883926392, + "learning_rate": 0.00013477397277694432, + "loss": 0.0993, + "step": 16513 + }, + { + "epoch": 0.38711416632969403, + "grad_norm": 0.9630061984062195, + "learning_rate": 0.00013476706623854128, + "loss": 0.0821, + "step": 16514 + }, + { + "epoch": 0.387137607904499, + "grad_norm": 3.662186861038208, + "learning_rate": 0.00013476015951149114, + "loss": 0.421, + "step": 16515 + }, + { + "epoch": 0.387161049479304, + "grad_norm": 2.60929274559021, + "learning_rate": 0.0001347532525958314, + "loss": 0.1576, + "step": 16516 + }, + { + "epoch": 0.387184491054109, + "grad_norm": 3.1076438426971436, + "learning_rate": 0.00013474634549159956, + "loss": 0.2485, + "step": 16517 + }, + { + "epoch": 0.387207932628914, + "grad_norm": 1.1461836099624634, + "learning_rate": 0.00013473943819883303, + "loss": 0.1222, + "step": 16518 + }, + { + "epoch": 0.387231374203719, + "grad_norm": 4.090266227722168, + "learning_rate": 0.00013473253071756935, + "loss": 0.531, + "step": 16519 + }, + { + "epoch": 0.387254815778524, + "grad_norm": 2.719815969467163, + "learning_rate": 0.000134725623047846, + "loss": 0.3112, + "step": 16520 + }, + { + "epoch": 0.387278257353329, + "grad_norm": 3.022636651992798, + "learning_rate": 0.0001347187151897004, + "loss": 0.3336, + "step": 16521 + }, + { + "epoch": 0.387301698928134, + "grad_norm": 4.159023761749268, + "learning_rate": 0.00013471180714317012, + "loss": 2.2911, + "step": 16522 + }, + { + "epoch": 0.387325140502939, + "grad_norm": 2.849601984024048, + "learning_rate": 0.00013470489890829259, + "loss": 0.3015, + "step": 16523 + }, + { + "epoch": 0.387348582077744, + "grad_norm": 2.157785654067993, + "learning_rate": 0.00013469799048510529, + "loss": 0.2157, + "step": 16524 + }, + { + "epoch": 0.38737202365254897, + "grad_norm": 6.460100173950195, + "learning_rate": 0.0001346910818736457, + "loss": 0.8558, + "step": 16525 + }, + { + "epoch": 0.38739546522735396, + "grad_norm": 2.4663569927215576, + "learning_rate": 0.00013468417307395134, + "loss": 0.24, + "step": 16526 + }, + { + "epoch": 0.38741890680215896, + "grad_norm": 0.8643003106117249, + "learning_rate": 0.00013467726408605967, + "loss": 0.0984, + "step": 16527 + }, + { + "epoch": 0.38744234837696395, + "grad_norm": 3.7184743881225586, + "learning_rate": 0.0001346703549100082, + "loss": 0.3197, + "step": 16528 + }, + { + "epoch": 0.38746578995176895, + "grad_norm": 3.1085612773895264, + "learning_rate": 0.00013466344554583442, + "loss": 0.2747, + "step": 16529 + }, + { + "epoch": 0.38748923152657394, + "grad_norm": 2.2377097606658936, + "learning_rate": 0.0001346565359935758, + "loss": 0.2529, + "step": 16530 + }, + { + "epoch": 0.38751267310137893, + "grad_norm": 3.529998540878296, + "learning_rate": 0.00013464962625326984, + "loss": 0.3243, + "step": 16531 + }, + { + "epoch": 0.38753611467618393, + "grad_norm": 3.605832576751709, + "learning_rate": 0.00013464271632495405, + "loss": 0.3588, + "step": 16532 + }, + { + "epoch": 0.3875595562509889, + "grad_norm": 1.892082929611206, + "learning_rate": 0.0001346358062086659, + "loss": 0.1494, + "step": 16533 + }, + { + "epoch": 0.3875829978257939, + "grad_norm": 0.6474278569221497, + "learning_rate": 0.0001346288959044429, + "loss": 0.0428, + "step": 16534 + }, + { + "epoch": 0.3876064394005989, + "grad_norm": 4.003063678741455, + "learning_rate": 0.00013462198541232254, + "loss": 0.3853, + "step": 16535 + }, + { + "epoch": 0.3876298809754039, + "grad_norm": 1.5639283657073975, + "learning_rate": 0.00013461507473234226, + "loss": 0.1501, + "step": 16536 + }, + { + "epoch": 0.3876533225502089, + "grad_norm": 3.0521633625030518, + "learning_rate": 0.00013460816386453965, + "loss": 0.2855, + "step": 16537 + }, + { + "epoch": 0.3876767641250139, + "grad_norm": 2.879808187484741, + "learning_rate": 0.0001346012528089522, + "loss": 1.7434, + "step": 16538 + }, + { + "epoch": 0.3877002056998189, + "grad_norm": 4.169376850128174, + "learning_rate": 0.00013459434156561735, + "loss": 0.5944, + "step": 16539 + }, + { + "epoch": 0.3877236472746239, + "grad_norm": 6.235949993133545, + "learning_rate": 0.00013458743013457263, + "loss": 0.2897, + "step": 16540 + }, + { + "epoch": 0.3877470888494289, + "grad_norm": 3.2428386211395264, + "learning_rate": 0.00013458051851585553, + "loss": 0.3461, + "step": 16541 + }, + { + "epoch": 0.3877705304242339, + "grad_norm": 0.8800028562545776, + "learning_rate": 0.00013457360670950357, + "loss": 0.0784, + "step": 16542 + }, + { + "epoch": 0.3877939719990389, + "grad_norm": 2.77121639251709, + "learning_rate": 0.00013456669471555426, + "loss": 0.3248, + "step": 16543 + }, + { + "epoch": 0.3878174135738439, + "grad_norm": 1.1835708618164062, + "learning_rate": 0.00013455978253404507, + "loss": 0.1194, + "step": 16544 + }, + { + "epoch": 0.3878408551486489, + "grad_norm": 2.7605319023132324, + "learning_rate": 0.00013455287016501358, + "loss": 0.2237, + "step": 16545 + }, + { + "epoch": 0.3878642967234539, + "grad_norm": 2.9620449542999268, + "learning_rate": 0.00013454595760849718, + "loss": 1.8401, + "step": 16546 + }, + { + "epoch": 0.3878877382982589, + "grad_norm": 3.5372936725616455, + "learning_rate": 0.00013453904486453348, + "loss": 0.2625, + "step": 16547 + }, + { + "epoch": 0.3879111798730639, + "grad_norm": 2.8181371688842773, + "learning_rate": 0.00013453213193315993, + "loss": 1.2103, + "step": 16548 + }, + { + "epoch": 0.3879346214478689, + "grad_norm": 2.031115770339966, + "learning_rate": 0.0001345252188144141, + "loss": 0.3414, + "step": 16549 + }, + { + "epoch": 0.3879580630226739, + "grad_norm": 2.7494122982025146, + "learning_rate": 0.0001345183055083334, + "loss": 0.1903, + "step": 16550 + }, + { + "epoch": 0.3879815045974789, + "grad_norm": 1.0121703147888184, + "learning_rate": 0.00013451139201495546, + "loss": 0.1006, + "step": 16551 + }, + { + "epoch": 0.38800494617228387, + "grad_norm": 1.821759581565857, + "learning_rate": 0.0001345044783343177, + "loss": 0.1642, + "step": 16552 + }, + { + "epoch": 0.38802838774708887, + "grad_norm": 2.6367201805114746, + "learning_rate": 0.0001344975644664577, + "loss": 1.857, + "step": 16553 + }, + { + "epoch": 0.38805182932189386, + "grad_norm": 5.815637588500977, + "learning_rate": 0.0001344906504114129, + "loss": 0.6474, + "step": 16554 + }, + { + "epoch": 0.38807527089669885, + "grad_norm": 2.831897735595703, + "learning_rate": 0.00013448373616922093, + "loss": 1.4884, + "step": 16555 + }, + { + "epoch": 0.38809871247150385, + "grad_norm": 2.4904167652130127, + "learning_rate": 0.00013447682173991918, + "loss": 1.0035, + "step": 16556 + }, + { + "epoch": 0.38812215404630884, + "grad_norm": 5.301325798034668, + "learning_rate": 0.00013446990712354527, + "loss": 0.5606, + "step": 16557 + }, + { + "epoch": 0.38814559562111384, + "grad_norm": 3.2324838638305664, + "learning_rate": 0.00013446299232013667, + "loss": 0.1956, + "step": 16558 + }, + { + "epoch": 0.38816903719591883, + "grad_norm": 3.619643449783325, + "learning_rate": 0.00013445607732973084, + "loss": 0.2528, + "step": 16559 + }, + { + "epoch": 0.3881924787707238, + "grad_norm": 0.9405015707015991, + "learning_rate": 0.00013444916215236543, + "loss": 0.0997, + "step": 16560 + }, + { + "epoch": 0.3882159203455288, + "grad_norm": 3.3281452655792236, + "learning_rate": 0.00013444224678807787, + "loss": 0.3024, + "step": 16561 + }, + { + "epoch": 0.3882393619203338, + "grad_norm": 4.659650802612305, + "learning_rate": 0.00013443533123690574, + "loss": 0.7068, + "step": 16562 + }, + { + "epoch": 0.3882628034951388, + "grad_norm": 3.815734624862671, + "learning_rate": 0.00013442841549888648, + "loss": 0.3174, + "step": 16563 + }, + { + "epoch": 0.3882862450699438, + "grad_norm": 4.645941734313965, + "learning_rate": 0.00013442149957405773, + "loss": 0.4823, + "step": 16564 + }, + { + "epoch": 0.3883096866447488, + "grad_norm": 2.5294899940490723, + "learning_rate": 0.00013441458346245692, + "loss": 0.2422, + "step": 16565 + }, + { + "epoch": 0.3883331282195538, + "grad_norm": 3.6792702674865723, + "learning_rate": 0.0001344076671641216, + "loss": 0.3866, + "step": 16566 + }, + { + "epoch": 0.3883565697943588, + "grad_norm": 2.0963480472564697, + "learning_rate": 0.00013440075067908935, + "loss": 0.2473, + "step": 16567 + }, + { + "epoch": 0.3883800113691638, + "grad_norm": 0.7849412560462952, + "learning_rate": 0.00013439383400739766, + "loss": 0.0738, + "step": 16568 + }, + { + "epoch": 0.3884034529439688, + "grad_norm": 1.2446025609970093, + "learning_rate": 0.000134386917149084, + "loss": 0.1244, + "step": 16569 + }, + { + "epoch": 0.38842689451877377, + "grad_norm": 3.9668285846710205, + "learning_rate": 0.00013438000010418603, + "loss": 0.2653, + "step": 16570 + }, + { + "epoch": 0.38845033609357876, + "grad_norm": 3.366804599761963, + "learning_rate": 0.0001343730828727412, + "loss": 0.381, + "step": 16571 + }, + { + "epoch": 0.38847377766838376, + "grad_norm": 3.3199353218078613, + "learning_rate": 0.00013436616545478702, + "loss": 1.3731, + "step": 16572 + }, + { + "epoch": 0.38849721924318875, + "grad_norm": 3.0365161895751953, + "learning_rate": 0.00013435924785036108, + "loss": 1.6077, + "step": 16573 + }, + { + "epoch": 0.38852066081799375, + "grad_norm": 4.6299543380737305, + "learning_rate": 0.0001343523300595009, + "loss": 0.5307, + "step": 16574 + }, + { + "epoch": 0.38854410239279874, + "grad_norm": 3.7245025634765625, + "learning_rate": 0.000134345412082244, + "loss": 1.4934, + "step": 16575 + }, + { + "epoch": 0.38856754396760373, + "grad_norm": 3.6509273052215576, + "learning_rate": 0.00013433849391862794, + "loss": 0.5061, + "step": 16576 + }, + { + "epoch": 0.38859098554240873, + "grad_norm": 4.682189464569092, + "learning_rate": 0.00013433157556869024, + "loss": 0.4858, + "step": 16577 + }, + { + "epoch": 0.3886144271172137, + "grad_norm": 2.7506821155548096, + "learning_rate": 0.00013432465703246845, + "loss": 0.3184, + "step": 16578 + }, + { + "epoch": 0.3886378686920187, + "grad_norm": 3.525402307510376, + "learning_rate": 0.0001343177383100001, + "loss": 0.3402, + "step": 16579 + }, + { + "epoch": 0.3886613102668237, + "grad_norm": 3.2392284870147705, + "learning_rate": 0.00013431081940132276, + "loss": 0.2592, + "step": 16580 + }, + { + "epoch": 0.3886847518416287, + "grad_norm": 2.933790922164917, + "learning_rate": 0.00013430390030647394, + "loss": 1.7562, + "step": 16581 + }, + { + "epoch": 0.3887081934164337, + "grad_norm": 1.3083919286727905, + "learning_rate": 0.00013429698102549115, + "loss": 0.123, + "step": 16582 + }, + { + "epoch": 0.3887316349912387, + "grad_norm": 2.55429744720459, + "learning_rate": 0.000134290061558412, + "loss": 0.1514, + "step": 16583 + }, + { + "epoch": 0.3887550765660437, + "grad_norm": 4.469311237335205, + "learning_rate": 0.00013428314190527403, + "loss": 0.4939, + "step": 16584 + }, + { + "epoch": 0.3887785181408487, + "grad_norm": 5.174449443817139, + "learning_rate": 0.00013427622206611475, + "loss": 0.5436, + "step": 16585 + }, + { + "epoch": 0.3888019597156537, + "grad_norm": 1.36433744430542, + "learning_rate": 0.00013426930204097175, + "loss": 0.1804, + "step": 16586 + }, + { + "epoch": 0.38882540129045867, + "grad_norm": 3.583585739135742, + "learning_rate": 0.00013426238182988253, + "loss": 0.2889, + "step": 16587 + }, + { + "epoch": 0.38884884286526367, + "grad_norm": 3.1658623218536377, + "learning_rate": 0.00013425546143288467, + "loss": 0.3122, + "step": 16588 + }, + { + "epoch": 0.38887228444006866, + "grad_norm": 0.918036937713623, + "learning_rate": 0.0001342485408500157, + "loss": 0.1013, + "step": 16589 + }, + { + "epoch": 0.38889572601487365, + "grad_norm": 5.067862033843994, + "learning_rate": 0.0001342416200813132, + "loss": 0.6577, + "step": 16590 + }, + { + "epoch": 0.38891916758967865, + "grad_norm": 1.723414659500122, + "learning_rate": 0.0001342346991268147, + "loss": 0.1655, + "step": 16591 + }, + { + "epoch": 0.38894260916448364, + "grad_norm": 4.142942905426025, + "learning_rate": 0.00013422777798655777, + "loss": 0.4938, + "step": 16592 + }, + { + "epoch": 0.38896605073928864, + "grad_norm": 4.456716537475586, + "learning_rate": 0.00013422085666057996, + "loss": 0.2684, + "step": 16593 + }, + { + "epoch": 0.3889894923140937, + "grad_norm": 1.4388723373413086, + "learning_rate": 0.0001342139351489188, + "loss": 1.0826, + "step": 16594 + }, + { + "epoch": 0.3890129338888987, + "grad_norm": 2.574439525604248, + "learning_rate": 0.00013420701345161188, + "loss": 0.1647, + "step": 16595 + }, + { + "epoch": 0.3890363754637037, + "grad_norm": 0.840354859828949, + "learning_rate": 0.00013420009156869675, + "loss": 0.1069, + "step": 16596 + }, + { + "epoch": 0.38905981703850867, + "grad_norm": 4.2942376136779785, + "learning_rate": 0.00013419316950021094, + "loss": 0.2739, + "step": 16597 + }, + { + "epoch": 0.38908325861331367, + "grad_norm": 1.9229578971862793, + "learning_rate": 0.00013418624724619203, + "loss": 0.1708, + "step": 16598 + }, + { + "epoch": 0.38910670018811866, + "grad_norm": 1.2819322347640991, + "learning_rate": 0.0001341793248066776, + "loss": 0.0936, + "step": 16599 + }, + { + "epoch": 0.38913014176292365, + "grad_norm": 2.475964069366455, + "learning_rate": 0.0001341724021817052, + "loss": 1.8898, + "step": 16600 + }, + { + "epoch": 0.38915358333772865, + "grad_norm": 3.4555106163024902, + "learning_rate": 0.00013416547937131234, + "loss": 0.4027, + "step": 16601 + }, + { + "epoch": 0.38917702491253364, + "grad_norm": 1.2849550247192383, + "learning_rate": 0.00013415855637553668, + "loss": 0.1511, + "step": 16602 + }, + { + "epoch": 0.38920046648733864, + "grad_norm": 0.8804646134376526, + "learning_rate": 0.00013415163319441568, + "loss": 0.0813, + "step": 16603 + }, + { + "epoch": 0.38922390806214363, + "grad_norm": 1.0576026439666748, + "learning_rate": 0.00013414470982798697, + "loss": 0.0753, + "step": 16604 + }, + { + "epoch": 0.3892473496369486, + "grad_norm": 6.34456729888916, + "learning_rate": 0.00013413778627628816, + "loss": 0.5038, + "step": 16605 + }, + { + "epoch": 0.3892707912117536, + "grad_norm": 0.6870163679122925, + "learning_rate": 0.00013413086253935672, + "loss": 0.0691, + "step": 16606 + }, + { + "epoch": 0.3892942327865586, + "grad_norm": 3.5867278575897217, + "learning_rate": 0.00013412393861723026, + "loss": 0.3394, + "step": 16607 + }, + { + "epoch": 0.3893176743613636, + "grad_norm": 1.5924392938613892, + "learning_rate": 0.00013411701450994636, + "loss": 0.0346, + "step": 16608 + }, + { + "epoch": 0.3893411159361686, + "grad_norm": 1.5205415487289429, + "learning_rate": 0.00013411009021754256, + "loss": 0.1694, + "step": 16609 + }, + { + "epoch": 0.3893645575109736, + "grad_norm": 3.696971893310547, + "learning_rate": 0.00013410316574005644, + "loss": 0.2673, + "step": 16610 + }, + { + "epoch": 0.3893879990857786, + "grad_norm": 3.7568459510803223, + "learning_rate": 0.00013409624107752564, + "loss": 0.3504, + "step": 16611 + }, + { + "epoch": 0.3894114406605836, + "grad_norm": 0.770754337310791, + "learning_rate": 0.00013408931622998764, + "loss": 0.0607, + "step": 16612 + }, + { + "epoch": 0.3894348822353886, + "grad_norm": 4.006688594818115, + "learning_rate": 0.00013408239119748006, + "loss": 0.4729, + "step": 16613 + }, + { + "epoch": 0.3894583238101936, + "grad_norm": 1.1666526794433594, + "learning_rate": 0.00013407546598004046, + "loss": 0.1762, + "step": 16614 + }, + { + "epoch": 0.38948176538499857, + "grad_norm": 1.5018318891525269, + "learning_rate": 0.00013406854057770642, + "loss": 0.1406, + "step": 16615 + }, + { + "epoch": 0.38950520695980356, + "grad_norm": 7.208601951599121, + "learning_rate": 0.00013406161499051553, + "loss": 0.5485, + "step": 16616 + }, + { + "epoch": 0.38952864853460856, + "grad_norm": 0.8130049705505371, + "learning_rate": 0.00013405468921850533, + "loss": 0.0767, + "step": 16617 + }, + { + "epoch": 0.38955209010941355, + "grad_norm": 0.41564813256263733, + "learning_rate": 0.00013404776326171348, + "loss": 0.0394, + "step": 16618 + }, + { + "epoch": 0.38957553168421855, + "grad_norm": 1.8122128248214722, + "learning_rate": 0.0001340408371201775, + "loss": 0.2558, + "step": 16619 + }, + { + "epoch": 0.38959897325902354, + "grad_norm": 5.716441631317139, + "learning_rate": 0.00013403391079393494, + "loss": 0.4784, + "step": 16620 + }, + { + "epoch": 0.38962241483382853, + "grad_norm": 1.3347173929214478, + "learning_rate": 0.00013402698428302345, + "loss": 0.695, + "step": 16621 + }, + { + "epoch": 0.38964585640863353, + "grad_norm": 4.2075700759887695, + "learning_rate": 0.0001340200575874806, + "loss": 1.8972, + "step": 16622 + }, + { + "epoch": 0.3896692979834385, + "grad_norm": 3.3579277992248535, + "learning_rate": 0.00013401313070734395, + "loss": 0.3979, + "step": 16623 + }, + { + "epoch": 0.3896927395582435, + "grad_norm": 1.3320809602737427, + "learning_rate": 0.00013400620364265107, + "loss": 0.1664, + "step": 16624 + }, + { + "epoch": 0.3897161811330485, + "grad_norm": 4.826127529144287, + "learning_rate": 0.00013399927639343965, + "loss": 0.4501, + "step": 16625 + }, + { + "epoch": 0.3897396227078535, + "grad_norm": 7.937176704406738, + "learning_rate": 0.0001339923489597471, + "loss": 0.7409, + "step": 16626 + }, + { + "epoch": 0.3897630642826585, + "grad_norm": 2.5168673992156982, + "learning_rate": 0.00013398542134161114, + "loss": 1.4623, + "step": 16627 + }, + { + "epoch": 0.3897865058574635, + "grad_norm": 4.975948333740234, + "learning_rate": 0.00013397849353906938, + "loss": 0.368, + "step": 16628 + }, + { + "epoch": 0.3898099474322685, + "grad_norm": 3.8948771953582764, + "learning_rate": 0.0001339715655521593, + "loss": 0.2238, + "step": 16629 + }, + { + "epoch": 0.3898333890070735, + "grad_norm": 1.9240200519561768, + "learning_rate": 0.00013396463738091859, + "loss": 0.185, + "step": 16630 + }, + { + "epoch": 0.3898568305818785, + "grad_norm": 3.7578651905059814, + "learning_rate": 0.00013395770902538477, + "loss": 0.3866, + "step": 16631 + }, + { + "epoch": 0.38988027215668347, + "grad_norm": 6.002188682556152, + "learning_rate": 0.00013395078048559548, + "loss": 0.6167, + "step": 16632 + }, + { + "epoch": 0.38990371373148847, + "grad_norm": 3.2661478519439697, + "learning_rate": 0.0001339438517615883, + "loss": 0.3008, + "step": 16633 + }, + { + "epoch": 0.38992715530629346, + "grad_norm": 3.868368148803711, + "learning_rate": 0.00013393692285340082, + "loss": 0.4131, + "step": 16634 + }, + { + "epoch": 0.38995059688109845, + "grad_norm": 9.667672157287598, + "learning_rate": 0.00013392999376107066, + "loss": 0.7845, + "step": 16635 + }, + { + "epoch": 0.38997403845590345, + "grad_norm": 1.97532057762146, + "learning_rate": 0.0001339230644846354, + "loss": 0.2034, + "step": 16636 + }, + { + "epoch": 0.38999748003070844, + "grad_norm": 5.313431739807129, + "learning_rate": 0.00013391613502413262, + "loss": 0.6062, + "step": 16637 + }, + { + "epoch": 0.39002092160551344, + "grad_norm": 3.0773274898529053, + "learning_rate": 0.00013390920537959997, + "loss": 0.3186, + "step": 16638 + }, + { + "epoch": 0.39004436318031843, + "grad_norm": 5.493021011352539, + "learning_rate": 0.00013390227555107494, + "loss": 0.4648, + "step": 16639 + }, + { + "epoch": 0.3900678047551234, + "grad_norm": 3.638221263885498, + "learning_rate": 0.00013389534553859527, + "loss": 0.3744, + "step": 16640 + }, + { + "epoch": 0.3900912463299284, + "grad_norm": 3.373879909515381, + "learning_rate": 0.00013388841534219852, + "loss": 0.2985, + "step": 16641 + }, + { + "epoch": 0.3901146879047334, + "grad_norm": 3.701490879058838, + "learning_rate": 0.00013388148496192224, + "loss": 0.4227, + "step": 16642 + }, + { + "epoch": 0.3901381294795384, + "grad_norm": 4.273140907287598, + "learning_rate": 0.00013387455439780409, + "loss": 0.5407, + "step": 16643 + }, + { + "epoch": 0.3901615710543434, + "grad_norm": 3.6180994510650635, + "learning_rate": 0.00013386762364988165, + "loss": 0.3619, + "step": 16644 + }, + { + "epoch": 0.39018501262914845, + "grad_norm": 3.4189677238464355, + "learning_rate": 0.0001338606927181925, + "loss": 0.2469, + "step": 16645 + }, + { + "epoch": 0.39020845420395345, + "grad_norm": 7.665473937988281, + "learning_rate": 0.00013385376160277433, + "loss": 0.3315, + "step": 16646 + }, + { + "epoch": 0.39023189577875844, + "grad_norm": 3.0472490787506104, + "learning_rate": 0.00013384683030366468, + "loss": 0.3094, + "step": 16647 + }, + { + "epoch": 0.39025533735356344, + "grad_norm": 5.315479755401611, + "learning_rate": 0.00013383989882090113, + "loss": 0.4969, + "step": 16648 + }, + { + "epoch": 0.39027877892836843, + "grad_norm": 0.9344502091407776, + "learning_rate": 0.00013383296715452142, + "loss": 0.1046, + "step": 16649 + }, + { + "epoch": 0.3903022205031734, + "grad_norm": 0.31373608112335205, + "learning_rate": 0.00013382603530456304, + "loss": 0.0232, + "step": 16650 + }, + { + "epoch": 0.3903256620779784, + "grad_norm": 2.8932301998138428, + "learning_rate": 0.00013381910327106363, + "loss": 0.3756, + "step": 16651 + }, + { + "epoch": 0.3903491036527834, + "grad_norm": 3.7386789321899414, + "learning_rate": 0.0001338121710540608, + "loss": 0.3732, + "step": 16652 + }, + { + "epoch": 0.3903725452275884, + "grad_norm": 4.776594161987305, + "learning_rate": 0.00013380523865359222, + "loss": 0.6655, + "step": 16653 + }, + { + "epoch": 0.3903959868023934, + "grad_norm": 4.4841628074646, + "learning_rate": 0.00013379830606969546, + "loss": 0.5392, + "step": 16654 + }, + { + "epoch": 0.3904194283771984, + "grad_norm": 4.270510196685791, + "learning_rate": 0.00013379137330240813, + "loss": 0.179, + "step": 16655 + }, + { + "epoch": 0.3904428699520034, + "grad_norm": 2.824185371398926, + "learning_rate": 0.00013378444035176785, + "loss": 0.2485, + "step": 16656 + }, + { + "epoch": 0.3904663115268084, + "grad_norm": 3.1034553050994873, + "learning_rate": 0.00013377750721781223, + "loss": 0.2904, + "step": 16657 + }, + { + "epoch": 0.3904897531016134, + "grad_norm": 3.4979662895202637, + "learning_rate": 0.00013377057390057893, + "loss": 0.3764, + "step": 16658 + }, + { + "epoch": 0.3905131946764184, + "grad_norm": 1.7588341236114502, + "learning_rate": 0.00013376364040010557, + "loss": 0.1656, + "step": 16659 + }, + { + "epoch": 0.39053663625122337, + "grad_norm": 3.5767252445220947, + "learning_rate": 0.00013375670671642972, + "loss": 0.4289, + "step": 16660 + }, + { + "epoch": 0.39056007782602836, + "grad_norm": 2.8859734535217285, + "learning_rate": 0.00013374977284958902, + "loss": 0.3271, + "step": 16661 + }, + { + "epoch": 0.39058351940083336, + "grad_norm": 4.920441627502441, + "learning_rate": 0.00013374283879962113, + "loss": 0.5552, + "step": 16662 + }, + { + "epoch": 0.39060696097563835, + "grad_norm": 5.216928958892822, + "learning_rate": 0.00013373590456656364, + "loss": 0.6581, + "step": 16663 + }, + { + "epoch": 0.39063040255044335, + "grad_norm": 3.5080487728118896, + "learning_rate": 0.00013372897015045417, + "loss": 1.4093, + "step": 16664 + }, + { + "epoch": 0.39065384412524834, + "grad_norm": 3.3407206535339355, + "learning_rate": 0.00013372203555133037, + "loss": 0.2739, + "step": 16665 + }, + { + "epoch": 0.39067728570005333, + "grad_norm": 4.120659828186035, + "learning_rate": 0.00013371510076922986, + "loss": 0.5265, + "step": 16666 + }, + { + "epoch": 0.39070072727485833, + "grad_norm": 0.7949374914169312, + "learning_rate": 0.00013370816580419024, + "loss": 0.1105, + "step": 16667 + }, + { + "epoch": 0.3907241688496633, + "grad_norm": 4.237878799438477, + "learning_rate": 0.00013370123065624919, + "loss": 0.6314, + "step": 16668 + }, + { + "epoch": 0.3907476104244683, + "grad_norm": 3.5869076251983643, + "learning_rate": 0.00013369429532544434, + "loss": 0.3816, + "step": 16669 + }, + { + "epoch": 0.3907710519992733, + "grad_norm": 2.466273069381714, + "learning_rate": 0.00013368735981181327, + "loss": 1.604, + "step": 16670 + }, + { + "epoch": 0.3907944935740783, + "grad_norm": 0.7478917241096497, + "learning_rate": 0.0001336804241153936, + "loss": 0.0579, + "step": 16671 + }, + { + "epoch": 0.3908179351488833, + "grad_norm": 1.4851833581924438, + "learning_rate": 0.00013367348823622305, + "loss": 0.1386, + "step": 16672 + }, + { + "epoch": 0.3908413767236883, + "grad_norm": 0.9935420751571655, + "learning_rate": 0.00013366655217433922, + "loss": 0.1299, + "step": 16673 + }, + { + "epoch": 0.3908648182984933, + "grad_norm": 2.751725196838379, + "learning_rate": 0.00013365961592977967, + "loss": 0.3395, + "step": 16674 + }, + { + "epoch": 0.3908882598732983, + "grad_norm": 2.4511871337890625, + "learning_rate": 0.00013365267950258213, + "loss": 0.1574, + "step": 16675 + }, + { + "epoch": 0.3909117014481033, + "grad_norm": 4.239174842834473, + "learning_rate": 0.0001336457428927842, + "loss": 0.488, + "step": 16676 + }, + { + "epoch": 0.39093514302290827, + "grad_norm": 2.0685861110687256, + "learning_rate": 0.00013363880610042355, + "loss": 0.1652, + "step": 16677 + }, + { + "epoch": 0.39095858459771327, + "grad_norm": 3.807525157928467, + "learning_rate": 0.00013363186912553776, + "loss": 0.3104, + "step": 16678 + }, + { + "epoch": 0.39098202617251826, + "grad_norm": 4.5293402671813965, + "learning_rate": 0.00013362493196816453, + "loss": 0.2725, + "step": 16679 + }, + { + "epoch": 0.39100546774732325, + "grad_norm": 1.0199247598648071, + "learning_rate": 0.00013361799462834143, + "loss": 0.1024, + "step": 16680 + }, + { + "epoch": 0.39102890932212825, + "grad_norm": 2.714832067489624, + "learning_rate": 0.0001336110571061062, + "loss": 1.6774, + "step": 16681 + }, + { + "epoch": 0.39105235089693324, + "grad_norm": 2.643670082092285, + "learning_rate": 0.0001336041194014964, + "loss": 0.1796, + "step": 16682 + }, + { + "epoch": 0.39107579247173824, + "grad_norm": 5.015259742736816, + "learning_rate": 0.00013359718151454968, + "loss": 0.67, + "step": 16683 + }, + { + "epoch": 0.39109923404654323, + "grad_norm": 4.473841190338135, + "learning_rate": 0.00013359024344530377, + "loss": 0.5849, + "step": 16684 + }, + { + "epoch": 0.3911226756213482, + "grad_norm": 9.741410255432129, + "learning_rate": 0.0001335833051937962, + "loss": 0.5188, + "step": 16685 + }, + { + "epoch": 0.3911461171961532, + "grad_norm": 0.6974971890449524, + "learning_rate": 0.0001335763667600647, + "loss": 0.0398, + "step": 16686 + }, + { + "epoch": 0.3911695587709582, + "grad_norm": 4.5841779708862305, + "learning_rate": 0.00013356942814414687, + "loss": 0.4333, + "step": 16687 + }, + { + "epoch": 0.3911930003457632, + "grad_norm": 0.5750775933265686, + "learning_rate": 0.0001335624893460804, + "loss": 0.0826, + "step": 16688 + }, + { + "epoch": 0.3912164419205682, + "grad_norm": 3.51798415184021, + "learning_rate": 0.0001335555503659029, + "loss": 0.1322, + "step": 16689 + }, + { + "epoch": 0.3912398834953732, + "grad_norm": 3.1615140438079834, + "learning_rate": 0.00013354861120365205, + "loss": 0.34, + "step": 16690 + }, + { + "epoch": 0.3912633250701782, + "grad_norm": 0.8258885145187378, + "learning_rate": 0.0001335416718593655, + "loss": 0.0931, + "step": 16691 + }, + { + "epoch": 0.3912867666449832, + "grad_norm": 4.677842617034912, + "learning_rate": 0.00013353473233308088, + "loss": 0.6441, + "step": 16692 + }, + { + "epoch": 0.3913102082197882, + "grad_norm": 2.427055597305298, + "learning_rate": 0.00013352779262483587, + "loss": 0.2884, + "step": 16693 + }, + { + "epoch": 0.3913336497945932, + "grad_norm": 3.1873693466186523, + "learning_rate": 0.0001335208527346681, + "loss": 0.5297, + "step": 16694 + }, + { + "epoch": 0.39135709136939817, + "grad_norm": 2.9385666847229004, + "learning_rate": 0.00013351391266261525, + "loss": 0.2158, + "step": 16695 + }, + { + "epoch": 0.3913805329442032, + "grad_norm": 4.9192023277282715, + "learning_rate": 0.00013350697240871497, + "loss": 0.2985, + "step": 16696 + }, + { + "epoch": 0.3914039745190082, + "grad_norm": 1.1685131788253784, + "learning_rate": 0.0001335000319730049, + "loss": 0.0693, + "step": 16697 + }, + { + "epoch": 0.3914274160938132, + "grad_norm": 3.1675915718078613, + "learning_rate": 0.00013349309135552275, + "loss": 0.3129, + "step": 16698 + }, + { + "epoch": 0.3914508576686182, + "grad_norm": 2.8894572257995605, + "learning_rate": 0.0001334861505563061, + "loss": 0.2567, + "step": 16699 + }, + { + "epoch": 0.3914742992434232, + "grad_norm": 9.301445007324219, + "learning_rate": 0.00013347920957539267, + "loss": 0.6573, + "step": 16700 + }, + { + "epoch": 0.3914977408182282, + "grad_norm": 1.5596461296081543, + "learning_rate": 0.00013347226841282014, + "loss": 0.2022, + "step": 16701 + }, + { + "epoch": 0.3915211823930332, + "grad_norm": 2.4957096576690674, + "learning_rate": 0.0001334653270686261, + "loss": 0.3359, + "step": 16702 + }, + { + "epoch": 0.3915446239678382, + "grad_norm": 3.0465312004089355, + "learning_rate": 0.00013345838554284828, + "loss": 0.3655, + "step": 16703 + }, + { + "epoch": 0.3915680655426432, + "grad_norm": 2.157954216003418, + "learning_rate": 0.00013345144383552433, + "loss": 0.2392, + "step": 16704 + }, + { + "epoch": 0.39159150711744817, + "grad_norm": 3.776108980178833, + "learning_rate": 0.00013344450194669187, + "loss": 0.3213, + "step": 16705 + }, + { + "epoch": 0.39161494869225316, + "grad_norm": 3.5164685249328613, + "learning_rate": 0.0001334375598763886, + "loss": 0.3117, + "step": 16706 + }, + { + "epoch": 0.39163839026705816, + "grad_norm": 3.5269813537597656, + "learning_rate": 0.00013343061762465224, + "loss": 0.3823, + "step": 16707 + }, + { + "epoch": 0.39166183184186315, + "grad_norm": 4.957573890686035, + "learning_rate": 0.00013342367519152037, + "loss": 0.3934, + "step": 16708 + }, + { + "epoch": 0.39168527341666814, + "grad_norm": 3.4203906059265137, + "learning_rate": 0.0001334167325770307, + "loss": 0.3738, + "step": 16709 + }, + { + "epoch": 0.39170871499147314, + "grad_norm": 4.360598564147949, + "learning_rate": 0.00013340978978122091, + "loss": 0.5538, + "step": 16710 + }, + { + "epoch": 0.39173215656627813, + "grad_norm": 1.6152656078338623, + "learning_rate": 0.00013340284680412866, + "loss": 0.1721, + "step": 16711 + }, + { + "epoch": 0.39175559814108313, + "grad_norm": 3.1150240898132324, + "learning_rate": 0.0001333959036457916, + "loss": 0.168, + "step": 16712 + }, + { + "epoch": 0.3917790397158881, + "grad_norm": 5.24694299697876, + "learning_rate": 0.00013338896030624746, + "loss": 0.546, + "step": 16713 + }, + { + "epoch": 0.3918024812906931, + "grad_norm": 2.1349236965179443, + "learning_rate": 0.00013338201678553388, + "loss": 1.7221, + "step": 16714 + }, + { + "epoch": 0.3918259228654981, + "grad_norm": 1.0695066452026367, + "learning_rate": 0.00013337507308368852, + "loss": 0.1496, + "step": 16715 + }, + { + "epoch": 0.3918493644403031, + "grad_norm": 6.068994998931885, + "learning_rate": 0.00013336812920074909, + "loss": 0.6281, + "step": 16716 + }, + { + "epoch": 0.3918728060151081, + "grad_norm": 2.4456863403320312, + "learning_rate": 0.00013336118513675323, + "loss": 0.3687, + "step": 16717 + }, + { + "epoch": 0.3918962475899131, + "grad_norm": 0.6533973217010498, + "learning_rate": 0.00013335424089173866, + "loss": 0.0423, + "step": 16718 + }, + { + "epoch": 0.3919196891647181, + "grad_norm": 4.374701499938965, + "learning_rate": 0.00013334729646574304, + "loss": 0.3447, + "step": 16719 + }, + { + "epoch": 0.3919431307395231, + "grad_norm": 2.4641900062561035, + "learning_rate": 0.00013334035185880405, + "loss": 0.2625, + "step": 16720 + }, + { + "epoch": 0.3919665723143281, + "grad_norm": 3.035912036895752, + "learning_rate": 0.00013333340707095937, + "loss": 0.278, + "step": 16721 + }, + { + "epoch": 0.39199001388913307, + "grad_norm": 1.3199245929718018, + "learning_rate": 0.0001333264621022467, + "loss": 0.2124, + "step": 16722 + }, + { + "epoch": 0.39201345546393807, + "grad_norm": 0.731650173664093, + "learning_rate": 0.0001333195169527037, + "loss": 0.049, + "step": 16723 + }, + { + "epoch": 0.39203689703874306, + "grad_norm": 3.0708582401275635, + "learning_rate": 0.00013331257162236805, + "loss": 1.2253, + "step": 16724 + }, + { + "epoch": 0.39206033861354805, + "grad_norm": 6.662337303161621, + "learning_rate": 0.00013330562611127745, + "loss": 0.7564, + "step": 16725 + }, + { + "epoch": 0.39208378018835305, + "grad_norm": 3.8283681869506836, + "learning_rate": 0.0001332986804194696, + "loss": 0.4757, + "step": 16726 + }, + { + "epoch": 0.39210722176315804, + "grad_norm": 4.14474630355835, + "learning_rate": 0.0001332917345469822, + "loss": 0.5717, + "step": 16727 + }, + { + "epoch": 0.39213066333796304, + "grad_norm": 0.7365622520446777, + "learning_rate": 0.0001332847884938529, + "loss": 0.09, + "step": 16728 + }, + { + "epoch": 0.39215410491276803, + "grad_norm": 2.72861909866333, + "learning_rate": 0.00013327784226011937, + "loss": 1.4777, + "step": 16729 + }, + { + "epoch": 0.392177546487573, + "grad_norm": 4.468061923980713, + "learning_rate": 0.00013327089584581935, + "loss": 0.5358, + "step": 16730 + }, + { + "epoch": 0.392200988062378, + "grad_norm": 1.8305788040161133, + "learning_rate": 0.00013326394925099048, + "loss": 0.1724, + "step": 16731 + }, + { + "epoch": 0.392224429637183, + "grad_norm": 1.829046607017517, + "learning_rate": 0.00013325700247567054, + "loss": 0.1299, + "step": 16732 + }, + { + "epoch": 0.392247871211988, + "grad_norm": 3.5817246437072754, + "learning_rate": 0.00013325005551989715, + "loss": 0.3712, + "step": 16733 + }, + { + "epoch": 0.392271312786793, + "grad_norm": 0.3451342284679413, + "learning_rate": 0.00013324310838370803, + "loss": 0.0316, + "step": 16734 + }, + { + "epoch": 0.392294754361598, + "grad_norm": 5.4866623878479, + "learning_rate": 0.00013323616106714086, + "loss": 0.2661, + "step": 16735 + }, + { + "epoch": 0.392318195936403, + "grad_norm": 3.955615997314453, + "learning_rate": 0.00013322921357023334, + "loss": 0.7367, + "step": 16736 + }, + { + "epoch": 0.392341637511208, + "grad_norm": 1.7463620901107788, + "learning_rate": 0.0001332222658930232, + "loss": 0.1157, + "step": 16737 + }, + { + "epoch": 0.392365079086013, + "grad_norm": 1.3168448209762573, + "learning_rate": 0.00013321531803554807, + "loss": 0.1181, + "step": 16738 + }, + { + "epoch": 0.392388520660818, + "grad_norm": 0.9098019599914551, + "learning_rate": 0.0001332083699978457, + "loss": 0.3815, + "step": 16739 + }, + { + "epoch": 0.39241196223562297, + "grad_norm": 4.118349552154541, + "learning_rate": 0.00013320142177995382, + "loss": 0.42, + "step": 16740 + }, + { + "epoch": 0.39243540381042796, + "grad_norm": 3.4721314907073975, + "learning_rate": 0.00013319447338191006, + "loss": 0.4027, + "step": 16741 + }, + { + "epoch": 0.39245884538523296, + "grad_norm": 4.588522911071777, + "learning_rate": 0.00013318752480375213, + "loss": 0.4433, + "step": 16742 + }, + { + "epoch": 0.39248228696003795, + "grad_norm": 1.427626132965088, + "learning_rate": 0.00013318057604551781, + "loss": 0.1773, + "step": 16743 + }, + { + "epoch": 0.39250572853484295, + "grad_norm": 5.039010047912598, + "learning_rate": 0.0001331736271072447, + "loss": 0.3053, + "step": 16744 + }, + { + "epoch": 0.39252917010964794, + "grad_norm": 2.9050493240356445, + "learning_rate": 0.0001331666779889706, + "loss": 0.1943, + "step": 16745 + }, + { + "epoch": 0.39255261168445293, + "grad_norm": 2.5377352237701416, + "learning_rate": 0.00013315972869073318, + "loss": 1.421, + "step": 16746 + }, + { + "epoch": 0.392576053259258, + "grad_norm": 0.7775135636329651, + "learning_rate": 0.0001331527792125701, + "loss": 0.0723, + "step": 16747 + }, + { + "epoch": 0.392599494834063, + "grad_norm": 2.8888099193573, + "learning_rate": 0.00013314582955451914, + "loss": 0.3274, + "step": 16748 + }, + { + "epoch": 0.392622936408868, + "grad_norm": 1.9033788442611694, + "learning_rate": 0.00013313887971661795, + "loss": 1.1855, + "step": 16749 + }, + { + "epoch": 0.39264637798367297, + "grad_norm": 3.3327956199645996, + "learning_rate": 0.00013313192969890424, + "loss": 0.296, + "step": 16750 + }, + { + "epoch": 0.39266981955847796, + "grad_norm": 2.2024407386779785, + "learning_rate": 0.00013312497950141583, + "loss": 0.8234, + "step": 16751 + }, + { + "epoch": 0.39269326113328296, + "grad_norm": 0.8474900126457214, + "learning_rate": 0.00013311802912419028, + "loss": 0.0805, + "step": 16752 + }, + { + "epoch": 0.39271670270808795, + "grad_norm": 2.4949557781219482, + "learning_rate": 0.00013311107856726543, + "loss": 0.2069, + "step": 16753 + }, + { + "epoch": 0.39274014428289294, + "grad_norm": 3.4262242317199707, + "learning_rate": 0.0001331041278306789, + "loss": 0.4122, + "step": 16754 + }, + { + "epoch": 0.39276358585769794, + "grad_norm": 3.98106050491333, + "learning_rate": 0.00013309717691446844, + "loss": 0.3889, + "step": 16755 + }, + { + "epoch": 0.39278702743250293, + "grad_norm": 3.2246932983398438, + "learning_rate": 0.00013309022581867177, + "loss": 0.5058, + "step": 16756 + }, + { + "epoch": 0.3928104690073079, + "grad_norm": 4.863395690917969, + "learning_rate": 0.0001330832745433266, + "loss": 0.3055, + "step": 16757 + }, + { + "epoch": 0.3928339105821129, + "grad_norm": 3.219144344329834, + "learning_rate": 0.00013307632308847068, + "loss": 1.7771, + "step": 16758 + }, + { + "epoch": 0.3928573521569179, + "grad_norm": 2.792663097381592, + "learning_rate": 0.00013306937145414165, + "loss": 0.2676, + "step": 16759 + }, + { + "epoch": 0.3928807937317229, + "grad_norm": 3.2403690814971924, + "learning_rate": 0.00013306241964037733, + "loss": 1.9057, + "step": 16760 + }, + { + "epoch": 0.3929042353065279, + "grad_norm": 0.4201386570930481, + "learning_rate": 0.00013305546764721536, + "loss": 0.0119, + "step": 16761 + }, + { + "epoch": 0.3929276768813329, + "grad_norm": 2.9577417373657227, + "learning_rate": 0.00013304851547469353, + "loss": 0.1956, + "step": 16762 + }, + { + "epoch": 0.3929511184561379, + "grad_norm": 2.8708815574645996, + "learning_rate": 0.0001330415631228495, + "loss": 0.2898, + "step": 16763 + }, + { + "epoch": 0.3929745600309429, + "grad_norm": 5.438464164733887, + "learning_rate": 0.000133034610591721, + "loss": 0.3243, + "step": 16764 + }, + { + "epoch": 0.3929980016057479, + "grad_norm": 1.2476693391799927, + "learning_rate": 0.0001330276578813458, + "loss": 0.0832, + "step": 16765 + }, + { + "epoch": 0.3930214431805529, + "grad_norm": 0.32636114954948425, + "learning_rate": 0.0001330207049917616, + "loss": 0.0269, + "step": 16766 + }, + { + "epoch": 0.39304488475535787, + "grad_norm": 2.8672850131988525, + "learning_rate": 0.00013301375192300609, + "loss": 1.3055, + "step": 16767 + }, + { + "epoch": 0.39306832633016286, + "grad_norm": 2.7096188068389893, + "learning_rate": 0.00013300679867511708, + "loss": 0.307, + "step": 16768 + }, + { + "epoch": 0.39309176790496786, + "grad_norm": 1.053054690361023, + "learning_rate": 0.00013299984524813225, + "loss": 0.1439, + "step": 16769 + }, + { + "epoch": 0.39311520947977285, + "grad_norm": 2.1376988887786865, + "learning_rate": 0.00013299289164208934, + "loss": 1.18, + "step": 16770 + }, + { + "epoch": 0.39313865105457785, + "grad_norm": 0.5522685050964355, + "learning_rate": 0.00013298593785702605, + "loss": 0.0334, + "step": 16771 + }, + { + "epoch": 0.39316209262938284, + "grad_norm": 0.44092342257499695, + "learning_rate": 0.00013297898389298012, + "loss": 0.0503, + "step": 16772 + }, + { + "epoch": 0.39318553420418784, + "grad_norm": 5.702439308166504, + "learning_rate": 0.00013297202974998928, + "loss": 0.5946, + "step": 16773 + }, + { + "epoch": 0.39320897577899283, + "grad_norm": 0.9307424426078796, + "learning_rate": 0.00013296507542809134, + "loss": 0.1062, + "step": 16774 + }, + { + "epoch": 0.3932324173537978, + "grad_norm": 5.191793441772461, + "learning_rate": 0.00013295812092732392, + "loss": 0.3765, + "step": 16775 + }, + { + "epoch": 0.3932558589286028, + "grad_norm": 1.6219743490219116, + "learning_rate": 0.00013295116624772486, + "loss": 1.0253, + "step": 16776 + }, + { + "epoch": 0.3932793005034078, + "grad_norm": 5.951267719268799, + "learning_rate": 0.0001329442113893318, + "loss": 0.7226, + "step": 16777 + }, + { + "epoch": 0.3933027420782128, + "grad_norm": 2.4839091300964355, + "learning_rate": 0.00013293725635218256, + "loss": 0.2976, + "step": 16778 + }, + { + "epoch": 0.3933261836530178, + "grad_norm": 1.326872706413269, + "learning_rate": 0.0001329303011363148, + "loss": 0.1351, + "step": 16779 + }, + { + "epoch": 0.3933496252278228, + "grad_norm": 1.5494900941848755, + "learning_rate": 0.0001329233457417663, + "loss": 0.1376, + "step": 16780 + }, + { + "epoch": 0.3933730668026278, + "grad_norm": 2.276279926300049, + "learning_rate": 0.00013291639016857487, + "loss": 0.1833, + "step": 16781 + }, + { + "epoch": 0.3933965083774328, + "grad_norm": 0.8440219759941101, + "learning_rate": 0.00013290943441677811, + "loss": 0.0755, + "step": 16782 + }, + { + "epoch": 0.3934199499522378, + "grad_norm": 2.7986559867858887, + "learning_rate": 0.00013290247848641388, + "loss": 0.1432, + "step": 16783 + }, + { + "epoch": 0.3934433915270428, + "grad_norm": 1.3587019443511963, + "learning_rate": 0.00013289552237751986, + "loss": 0.1024, + "step": 16784 + }, + { + "epoch": 0.39346683310184777, + "grad_norm": 12.517472267150879, + "learning_rate": 0.00013288856609013379, + "loss": 0.4737, + "step": 16785 + }, + { + "epoch": 0.39349027467665276, + "grad_norm": 5.564125061035156, + "learning_rate": 0.00013288160962429343, + "loss": 0.4739, + "step": 16786 + }, + { + "epoch": 0.39351371625145776, + "grad_norm": 5.976822376251221, + "learning_rate": 0.00013287465298003658, + "loss": 0.2086, + "step": 16787 + }, + { + "epoch": 0.39353715782626275, + "grad_norm": 2.854793071746826, + "learning_rate": 0.0001328676961574009, + "loss": 0.284, + "step": 16788 + }, + { + "epoch": 0.39356059940106775, + "grad_norm": 1.2403579950332642, + "learning_rate": 0.0001328607391564242, + "loss": 0.1326, + "step": 16789 + }, + { + "epoch": 0.39358404097587274, + "grad_norm": 4.561711311340332, + "learning_rate": 0.0001328537819771442, + "loss": 0.6663, + "step": 16790 + }, + { + "epoch": 0.39360748255067773, + "grad_norm": 4.59283447265625, + "learning_rate": 0.00013284682461959862, + "loss": 0.3391, + "step": 16791 + }, + { + "epoch": 0.39363092412548273, + "grad_norm": 2.3817763328552246, + "learning_rate": 0.00013283986708382528, + "loss": 0.2478, + "step": 16792 + }, + { + "epoch": 0.3936543657002877, + "grad_norm": 3.1631174087524414, + "learning_rate": 0.0001328329093698619, + "loss": 0.1971, + "step": 16793 + }, + { + "epoch": 0.3936778072750927, + "grad_norm": 2.797006368637085, + "learning_rate": 0.00013282595147774622, + "loss": 0.1393, + "step": 16794 + }, + { + "epoch": 0.3937012488498977, + "grad_norm": 5.329184532165527, + "learning_rate": 0.000132818993407516, + "loss": 0.8628, + "step": 16795 + }, + { + "epoch": 0.3937246904247027, + "grad_norm": 2.5160715579986572, + "learning_rate": 0.00013281203515920904, + "loss": 0.2502, + "step": 16796 + }, + { + "epoch": 0.3937481319995077, + "grad_norm": 2.7295467853546143, + "learning_rate": 0.000132805076732863, + "loss": 0.3929, + "step": 16797 + }, + { + "epoch": 0.39377157357431275, + "grad_norm": 2.804389715194702, + "learning_rate": 0.0001327981181285157, + "loss": 0.1692, + "step": 16798 + }, + { + "epoch": 0.39379501514911774, + "grad_norm": 2.428048849105835, + "learning_rate": 0.00013279115934620493, + "loss": 1.6735, + "step": 16799 + }, + { + "epoch": 0.39381845672392274, + "grad_norm": 3.0114901065826416, + "learning_rate": 0.00013278420038596837, + "loss": 0.233, + "step": 16800 + }, + { + "epoch": 0.39384189829872773, + "grad_norm": 6.818001747131348, + "learning_rate": 0.0001327772412478438, + "loss": 0.9705, + "step": 16801 + }, + { + "epoch": 0.3938653398735327, + "grad_norm": 3.9910919666290283, + "learning_rate": 0.00013277028193186904, + "loss": 1.5024, + "step": 16802 + }, + { + "epoch": 0.3938887814483377, + "grad_norm": 1.1857304573059082, + "learning_rate": 0.00013276332243808178, + "loss": 0.1119, + "step": 16803 + }, + { + "epoch": 0.3939122230231427, + "grad_norm": 5.82062292098999, + "learning_rate": 0.00013275636276651983, + "loss": 0.3677, + "step": 16804 + }, + { + "epoch": 0.3939356645979477, + "grad_norm": 3.3731884956359863, + "learning_rate": 0.00013274940291722092, + "loss": 0.2242, + "step": 16805 + }, + { + "epoch": 0.3939591061727527, + "grad_norm": 2.672300338745117, + "learning_rate": 0.00013274244289022285, + "loss": 0.1868, + "step": 16806 + }, + { + "epoch": 0.3939825477475577, + "grad_norm": 1.028402328491211, + "learning_rate": 0.00013273548268556335, + "loss": 0.0861, + "step": 16807 + }, + { + "epoch": 0.3940059893223627, + "grad_norm": 3.3579158782958984, + "learning_rate": 0.0001327285223032802, + "loss": 0.3533, + "step": 16808 + }, + { + "epoch": 0.3940294308971677, + "grad_norm": 3.0552291870117188, + "learning_rate": 0.00013272156174341115, + "loss": 0.3479, + "step": 16809 + }, + { + "epoch": 0.3940528724719727, + "grad_norm": 4.315600872039795, + "learning_rate": 0.00013271460100599403, + "loss": 0.2225, + "step": 16810 + }, + { + "epoch": 0.3940763140467777, + "grad_norm": 2.7388460636138916, + "learning_rate": 0.0001327076400910665, + "loss": 0.2248, + "step": 16811 + }, + { + "epoch": 0.39409975562158267, + "grad_norm": 3.1000921726226807, + "learning_rate": 0.00013270067899866646, + "loss": 0.3417, + "step": 16812 + }, + { + "epoch": 0.39412319719638766, + "grad_norm": 5.175171375274658, + "learning_rate": 0.00013269371772883158, + "loss": 0.8846, + "step": 16813 + }, + { + "epoch": 0.39414663877119266, + "grad_norm": 4.22058629989624, + "learning_rate": 0.0001326867562815997, + "loss": 0.3974, + "step": 16814 + }, + { + "epoch": 0.39417008034599765, + "grad_norm": 2.2134969234466553, + "learning_rate": 0.00013267979465700854, + "loss": 1.5562, + "step": 16815 + }, + { + "epoch": 0.39419352192080265, + "grad_norm": 3.154881238937378, + "learning_rate": 0.0001326728328550959, + "loss": 0.2311, + "step": 16816 + }, + { + "epoch": 0.39421696349560764, + "grad_norm": 3.749422073364258, + "learning_rate": 0.00013266587087589952, + "loss": 0.2469, + "step": 16817 + }, + { + "epoch": 0.39424040507041264, + "grad_norm": 3.4924261569976807, + "learning_rate": 0.00013265890871945726, + "loss": 0.2133, + "step": 16818 + }, + { + "epoch": 0.39426384664521763, + "grad_norm": 3.40738582611084, + "learning_rate": 0.0001326519463858068, + "loss": 0.3654, + "step": 16819 + }, + { + "epoch": 0.3942872882200226, + "grad_norm": 0.8420665264129639, + "learning_rate": 0.00013264498387498597, + "loss": 0.1029, + "step": 16820 + }, + { + "epoch": 0.3943107297948276, + "grad_norm": 4.120183944702148, + "learning_rate": 0.00013263802118703254, + "loss": 0.4521, + "step": 16821 + }, + { + "epoch": 0.3943341713696326, + "grad_norm": 0.9544317722320557, + "learning_rate": 0.0001326310583219843, + "loss": 0.1041, + "step": 16822 + }, + { + "epoch": 0.3943576129444376, + "grad_norm": 3.783923864364624, + "learning_rate": 0.000132624095279879, + "loss": 0.3601, + "step": 16823 + }, + { + "epoch": 0.3943810545192426, + "grad_norm": 0.8909894227981567, + "learning_rate": 0.00013261713206075447, + "loss": 0.0953, + "step": 16824 + }, + { + "epoch": 0.3944044960940476, + "grad_norm": 3.584735631942749, + "learning_rate": 0.00013261016866464846, + "loss": 0.3378, + "step": 16825 + }, + { + "epoch": 0.3944279376688526, + "grad_norm": 3.5014631748199463, + "learning_rate": 0.00013260320509159875, + "loss": 0.4559, + "step": 16826 + }, + { + "epoch": 0.3944513792436576, + "grad_norm": 1.1290000677108765, + "learning_rate": 0.00013259624134164314, + "loss": 0.1085, + "step": 16827 + }, + { + "epoch": 0.3944748208184626, + "grad_norm": 0.6349131464958191, + "learning_rate": 0.0001325892774148194, + "loss": 0.0582, + "step": 16828 + }, + { + "epoch": 0.3944982623932676, + "grad_norm": 3.7864346504211426, + "learning_rate": 0.0001325823133111653, + "loss": 0.3458, + "step": 16829 + }, + { + "epoch": 0.39452170396807257, + "grad_norm": 2.4771416187286377, + "learning_rate": 0.0001325753490307187, + "loss": 0.3015, + "step": 16830 + }, + { + "epoch": 0.39454514554287756, + "grad_norm": 4.810871124267578, + "learning_rate": 0.00013256838457351733, + "loss": 0.2239, + "step": 16831 + }, + { + "epoch": 0.39456858711768256, + "grad_norm": 4.215750217437744, + "learning_rate": 0.00013256141993959894, + "loss": 0.3101, + "step": 16832 + }, + { + "epoch": 0.39459202869248755, + "grad_norm": 3.34505295753479, + "learning_rate": 0.0001325544551290014, + "loss": 0.3653, + "step": 16833 + }, + { + "epoch": 0.39461547026729255, + "grad_norm": 1.824116826057434, + "learning_rate": 0.0001325474901417625, + "loss": 0.0876, + "step": 16834 + }, + { + "epoch": 0.39463891184209754, + "grad_norm": 2.9521846771240234, + "learning_rate": 0.00013254052497791997, + "loss": 0.1299, + "step": 16835 + }, + { + "epoch": 0.39466235341690253, + "grad_norm": 1.945184350013733, + "learning_rate": 0.00013253355963751165, + "loss": 0.2048, + "step": 16836 + }, + { + "epoch": 0.39468579499170753, + "grad_norm": 0.7598783373832703, + "learning_rate": 0.00013252659412057533, + "loss": 0.0631, + "step": 16837 + }, + { + "epoch": 0.3947092365665125, + "grad_norm": 2.1380858421325684, + "learning_rate": 0.00013251962842714878, + "loss": 0.2311, + "step": 16838 + }, + { + "epoch": 0.3947326781413175, + "grad_norm": 1.2090861797332764, + "learning_rate": 0.0001325126625572698, + "loss": 0.0934, + "step": 16839 + }, + { + "epoch": 0.3947561197161225, + "grad_norm": 1.097917079925537, + "learning_rate": 0.0001325056965109762, + "loss": 0.0893, + "step": 16840 + }, + { + "epoch": 0.3947795612909275, + "grad_norm": 1.2745293378829956, + "learning_rate": 0.00013249873028830578, + "loss": 0.1232, + "step": 16841 + }, + { + "epoch": 0.3948030028657325, + "grad_norm": 3.197795867919922, + "learning_rate": 0.00013249176388929637, + "loss": 0.3026, + "step": 16842 + }, + { + "epoch": 0.3948264444405375, + "grad_norm": 3.762227773666382, + "learning_rate": 0.0001324847973139857, + "loss": 0.4079, + "step": 16843 + }, + { + "epoch": 0.3948498860153425, + "grad_norm": 3.684389114379883, + "learning_rate": 0.00013247783056241167, + "loss": 0.3077, + "step": 16844 + }, + { + "epoch": 0.3948733275901475, + "grad_norm": 0.523902177810669, + "learning_rate": 0.00013247086363461192, + "loss": 0.0487, + "step": 16845 + }, + { + "epoch": 0.3948967691649525, + "grad_norm": 1.5299886465072632, + "learning_rate": 0.00013246389653062438, + "loss": 0.1636, + "step": 16846 + }, + { + "epoch": 0.39492021073975747, + "grad_norm": 2.9561221599578857, + "learning_rate": 0.00013245692925048687, + "loss": 0.4186, + "step": 16847 + }, + { + "epoch": 0.39494365231456247, + "grad_norm": 4.151790142059326, + "learning_rate": 0.0001324499617942371, + "loss": 0.3551, + "step": 16848 + }, + { + "epoch": 0.3949670938893675, + "grad_norm": 5.309540271759033, + "learning_rate": 0.00013244299416191298, + "loss": 0.4714, + "step": 16849 + }, + { + "epoch": 0.3949905354641725, + "grad_norm": 3.352337598800659, + "learning_rate": 0.00013243602635355222, + "loss": 1.8828, + "step": 16850 + }, + { + "epoch": 0.3950139770389775, + "grad_norm": 2.7980356216430664, + "learning_rate": 0.00013242905836919266, + "loss": 0.4427, + "step": 16851 + }, + { + "epoch": 0.3950374186137825, + "grad_norm": 3.948503255844116, + "learning_rate": 0.00013242209020887212, + "loss": 0.321, + "step": 16852 + }, + { + "epoch": 0.3950608601885875, + "grad_norm": 2.529841423034668, + "learning_rate": 0.00013241512187262843, + "loss": 0.2299, + "step": 16853 + }, + { + "epoch": 0.3950843017633925, + "grad_norm": 4.6897430419921875, + "learning_rate": 0.00013240815336049939, + "loss": 0.5076, + "step": 16854 + }, + { + "epoch": 0.3951077433381975, + "grad_norm": 2.6217501163482666, + "learning_rate": 0.00013240118467252275, + "loss": 0.175, + "step": 16855 + }, + { + "epoch": 0.3951311849130025, + "grad_norm": 1.918053388595581, + "learning_rate": 0.00013239421580873642, + "loss": 1.8287, + "step": 16856 + }, + { + "epoch": 0.39515462648780747, + "grad_norm": 3.520704507827759, + "learning_rate": 0.00013238724676917815, + "loss": 0.2971, + "step": 16857 + }, + { + "epoch": 0.39517806806261246, + "grad_norm": 1.755565881729126, + "learning_rate": 0.00013238027755388572, + "loss": 0.2738, + "step": 16858 + }, + { + "epoch": 0.39520150963741746, + "grad_norm": 6.118006229400635, + "learning_rate": 0.00013237330816289706, + "loss": 0.6958, + "step": 16859 + }, + { + "epoch": 0.39522495121222245, + "grad_norm": 4.659603118896484, + "learning_rate": 0.0001323663385962499, + "loss": 0.413, + "step": 16860 + }, + { + "epoch": 0.39524839278702745, + "grad_norm": 4.935491561889648, + "learning_rate": 0.00013235936885398205, + "loss": 0.3874, + "step": 16861 + }, + { + "epoch": 0.39527183436183244, + "grad_norm": 0.7706207036972046, + "learning_rate": 0.00013235239893613139, + "loss": 0.1034, + "step": 16862 + }, + { + "epoch": 0.39529527593663744, + "grad_norm": 3.2119460105895996, + "learning_rate": 0.00013234542884273568, + "loss": 1.7639, + "step": 16863 + }, + { + "epoch": 0.39531871751144243, + "grad_norm": 4.011287689208984, + "learning_rate": 0.00013233845857383276, + "loss": 0.3593, + "step": 16864 + }, + { + "epoch": 0.3953421590862474, + "grad_norm": 1.3392200469970703, + "learning_rate": 0.00013233148812946048, + "loss": 0.1127, + "step": 16865 + }, + { + "epoch": 0.3953656006610524, + "grad_norm": 1.8470290899276733, + "learning_rate": 0.00013232451750965662, + "loss": 0.121, + "step": 16866 + }, + { + "epoch": 0.3953890422358574, + "grad_norm": 3.3582870960235596, + "learning_rate": 0.000132317546714459, + "loss": 0.3301, + "step": 16867 + }, + { + "epoch": 0.3954124838106624, + "grad_norm": 0.8871310353279114, + "learning_rate": 0.0001323105757439055, + "loss": 0.1204, + "step": 16868 + }, + { + "epoch": 0.3954359253854674, + "grad_norm": 1.9335068464279175, + "learning_rate": 0.0001323036045980339, + "loss": 0.3096, + "step": 16869 + }, + { + "epoch": 0.3954593669602724, + "grad_norm": 4.126404285430908, + "learning_rate": 0.00013229663327688203, + "loss": 0.4736, + "step": 16870 + }, + { + "epoch": 0.3954828085350774, + "grad_norm": 3.0460011959075928, + "learning_rate": 0.0001322896617804877, + "loss": 0.3644, + "step": 16871 + }, + { + "epoch": 0.3955062501098824, + "grad_norm": 2.688528060913086, + "learning_rate": 0.00013228269010888877, + "loss": 0.1837, + "step": 16872 + }, + { + "epoch": 0.3955296916846874, + "grad_norm": 3.4904096126556396, + "learning_rate": 0.0001322757182621231, + "loss": 0.3613, + "step": 16873 + }, + { + "epoch": 0.3955531332594924, + "grad_norm": 9.117985725402832, + "learning_rate": 0.0001322687462402284, + "loss": 0.5563, + "step": 16874 + }, + { + "epoch": 0.39557657483429737, + "grad_norm": 0.9481731057167053, + "learning_rate": 0.00013226177404324261, + "loss": 0.1183, + "step": 16875 + }, + { + "epoch": 0.39560001640910236, + "grad_norm": 2.416257858276367, + "learning_rate": 0.0001322548016712035, + "loss": 0.1956, + "step": 16876 + }, + { + "epoch": 0.39562345798390736, + "grad_norm": 3.6794252395629883, + "learning_rate": 0.00013224782912414897, + "loss": 0.5347, + "step": 16877 + }, + { + "epoch": 0.39564689955871235, + "grad_norm": 2.438319206237793, + "learning_rate": 0.0001322408564021168, + "loss": 0.221, + "step": 16878 + }, + { + "epoch": 0.39567034113351734, + "grad_norm": 0.8238228559494019, + "learning_rate": 0.00013223388350514485, + "loss": 0.1122, + "step": 16879 + }, + { + "epoch": 0.39569378270832234, + "grad_norm": 3.00658917427063, + "learning_rate": 0.00013222691043327092, + "loss": 0.2228, + "step": 16880 + }, + { + "epoch": 0.39571722428312733, + "grad_norm": 1.961898922920227, + "learning_rate": 0.00013221993718653288, + "loss": 0.2874, + "step": 16881 + }, + { + "epoch": 0.3957406658579323, + "grad_norm": 3.674098491668701, + "learning_rate": 0.00013221296376496853, + "loss": 0.5946, + "step": 16882 + }, + { + "epoch": 0.3957641074327373, + "grad_norm": 0.7837058901786804, + "learning_rate": 0.00013220599016861576, + "loss": 0.0623, + "step": 16883 + }, + { + "epoch": 0.3957875490075423, + "grad_norm": 2.7285139560699463, + "learning_rate": 0.0001321990163975124, + "loss": 1.0469, + "step": 16884 + }, + { + "epoch": 0.3958109905823473, + "grad_norm": 1.8576247692108154, + "learning_rate": 0.0001321920424516962, + "loss": 0.298, + "step": 16885 + }, + { + "epoch": 0.3958344321571523, + "grad_norm": 0.8642882108688354, + "learning_rate": 0.00013218506833120514, + "loss": 0.049, + "step": 16886 + }, + { + "epoch": 0.3958578737319573, + "grad_norm": 5.1247735023498535, + "learning_rate": 0.00013217809403607696, + "loss": 0.3561, + "step": 16887 + }, + { + "epoch": 0.3958813153067623, + "grad_norm": 2.760706663131714, + "learning_rate": 0.00013217111956634954, + "loss": 0.2904, + "step": 16888 + }, + { + "epoch": 0.3959047568815673, + "grad_norm": 2.132636308670044, + "learning_rate": 0.00013216414492206075, + "loss": 1.1851, + "step": 16889 + }, + { + "epoch": 0.3959281984563723, + "grad_norm": 3.3672337532043457, + "learning_rate": 0.00013215717010324837, + "loss": 0.4757, + "step": 16890 + }, + { + "epoch": 0.3959516400311773, + "grad_norm": 1.7190247774124146, + "learning_rate": 0.0001321501951099503, + "loss": 0.1819, + "step": 16891 + }, + { + "epoch": 0.39597508160598227, + "grad_norm": 2.9419426918029785, + "learning_rate": 0.00013214321994220438, + "loss": 0.2955, + "step": 16892 + }, + { + "epoch": 0.39599852318078727, + "grad_norm": 4.840956211090088, + "learning_rate": 0.0001321362446000484, + "loss": 0.5005, + "step": 16893 + }, + { + "epoch": 0.39602196475559226, + "grad_norm": 6.14555549621582, + "learning_rate": 0.00013212926908352029, + "loss": 0.4467, + "step": 16894 + }, + { + "epoch": 0.39604540633039725, + "grad_norm": 5.54951810836792, + "learning_rate": 0.00013212229339265786, + "loss": 0.5014, + "step": 16895 + }, + { + "epoch": 0.39606884790520225, + "grad_norm": 3.075983762741089, + "learning_rate": 0.00013211531752749896, + "loss": 0.2011, + "step": 16896 + }, + { + "epoch": 0.39609228948000724, + "grad_norm": 7.148394584655762, + "learning_rate": 0.0001321083414880814, + "loss": 0.3397, + "step": 16897 + }, + { + "epoch": 0.39611573105481224, + "grad_norm": 2.7625694274902344, + "learning_rate": 0.00013210136527444318, + "loss": 0.2932, + "step": 16898 + }, + { + "epoch": 0.39613917262961723, + "grad_norm": 2.319166898727417, + "learning_rate": 0.00013209438888662196, + "loss": 1.0904, + "step": 16899 + }, + { + "epoch": 0.3961626142044222, + "grad_norm": 2.074134111404419, + "learning_rate": 0.00013208741232465573, + "loss": 1.138, + "step": 16900 + }, + { + "epoch": 0.3961860557792273, + "grad_norm": 2.3981728553771973, + "learning_rate": 0.00013208043558858226, + "loss": 0.3122, + "step": 16901 + }, + { + "epoch": 0.39620949735403227, + "grad_norm": 3.8371036052703857, + "learning_rate": 0.00013207345867843946, + "loss": 0.5214, + "step": 16902 + }, + { + "epoch": 0.39623293892883726, + "grad_norm": 4.258230686187744, + "learning_rate": 0.00013206648159426516, + "loss": 0.2929, + "step": 16903 + }, + { + "epoch": 0.39625638050364226, + "grad_norm": 1.4605909585952759, + "learning_rate": 0.00013205950433609726, + "loss": 0.1833, + "step": 16904 + }, + { + "epoch": 0.39627982207844725, + "grad_norm": 3.121828079223633, + "learning_rate": 0.0001320525269039736, + "loss": 0.1939, + "step": 16905 + }, + { + "epoch": 0.39630326365325225, + "grad_norm": 6.007601261138916, + "learning_rate": 0.00013204554929793197, + "loss": 0.4954, + "step": 16906 + }, + { + "epoch": 0.39632670522805724, + "grad_norm": 2.6258506774902344, + "learning_rate": 0.0001320385715180103, + "loss": 0.2568, + "step": 16907 + }, + { + "epoch": 0.39635014680286224, + "grad_norm": 6.103431701660156, + "learning_rate": 0.00013203159356424646, + "loss": 0.8544, + "step": 16908 + }, + { + "epoch": 0.39637358837766723, + "grad_norm": 0.6900604367256165, + "learning_rate": 0.00013202461543667827, + "loss": 0.0846, + "step": 16909 + }, + { + "epoch": 0.3963970299524722, + "grad_norm": 3.765547752380371, + "learning_rate": 0.00013201763713534365, + "loss": 1.513, + "step": 16910 + }, + { + "epoch": 0.3964204715272772, + "grad_norm": 4.379896640777588, + "learning_rate": 0.00013201065866028043, + "loss": 0.3509, + "step": 16911 + }, + { + "epoch": 0.3964439131020822, + "grad_norm": 2.6214215755462646, + "learning_rate": 0.00013200368001152645, + "loss": 1.6731, + "step": 16912 + }, + { + "epoch": 0.3964673546768872, + "grad_norm": 0.9487466812133789, + "learning_rate": 0.00013199670118911963, + "loss": 0.1046, + "step": 16913 + }, + { + "epoch": 0.3964907962516922, + "grad_norm": 1.174813151359558, + "learning_rate": 0.0001319897221930978, + "loss": 0.0918, + "step": 16914 + }, + { + "epoch": 0.3965142378264972, + "grad_norm": 2.3994123935699463, + "learning_rate": 0.00013198274302349882, + "loss": 0.2488, + "step": 16915 + }, + { + "epoch": 0.3965376794013022, + "grad_norm": 2.3429696559906006, + "learning_rate": 0.0001319757636803606, + "loss": 1.5211, + "step": 16916 + }, + { + "epoch": 0.3965611209761072, + "grad_norm": 3.4274446964263916, + "learning_rate": 0.00013196878416372098, + "loss": 0.2311, + "step": 16917 + }, + { + "epoch": 0.3965845625509122, + "grad_norm": 0.5964416861534119, + "learning_rate": 0.00013196180447361781, + "loss": 0.0437, + "step": 16918 + }, + { + "epoch": 0.3966080041257172, + "grad_norm": 0.6204826235771179, + "learning_rate": 0.00013195482461008902, + "loss": 0.0404, + "step": 16919 + }, + { + "epoch": 0.39663144570052217, + "grad_norm": 3.4555907249450684, + "learning_rate": 0.00013194784457317245, + "loss": 0.4384, + "step": 16920 + }, + { + "epoch": 0.39665488727532716, + "grad_norm": 1.6051908731460571, + "learning_rate": 0.000131940864362906, + "loss": 0.1391, + "step": 16921 + }, + { + "epoch": 0.39667832885013216, + "grad_norm": 5.124591827392578, + "learning_rate": 0.00013193388397932747, + "loss": 0.3794, + "step": 16922 + }, + { + "epoch": 0.39670177042493715, + "grad_norm": 1.9535975456237793, + "learning_rate": 0.00013192690342247483, + "loss": 0.8304, + "step": 16923 + }, + { + "epoch": 0.39672521199974214, + "grad_norm": 3.963887929916382, + "learning_rate": 0.0001319199226923859, + "loss": 0.2966, + "step": 16924 + }, + { + "epoch": 0.39674865357454714, + "grad_norm": 3.040060520172119, + "learning_rate": 0.00013191294178909856, + "loss": 0.4259, + "step": 16925 + }, + { + "epoch": 0.39677209514935213, + "grad_norm": 2.250786781311035, + "learning_rate": 0.00013190596071265071, + "loss": 0.2126, + "step": 16926 + }, + { + "epoch": 0.3967955367241571, + "grad_norm": 5.817208766937256, + "learning_rate": 0.00013189897946308023, + "loss": 0.5917, + "step": 16927 + }, + { + "epoch": 0.3968189782989621, + "grad_norm": 4.038760662078857, + "learning_rate": 0.000131891998040425, + "loss": 0.372, + "step": 16928 + }, + { + "epoch": 0.3968424198737671, + "grad_norm": 2.8336057662963867, + "learning_rate": 0.00013188501644472288, + "loss": 0.3264, + "step": 16929 + }, + { + "epoch": 0.3968658614485721, + "grad_norm": 3.535006046295166, + "learning_rate": 0.00013187803467601175, + "loss": 0.3004, + "step": 16930 + }, + { + "epoch": 0.3968893030233771, + "grad_norm": 2.635643243789673, + "learning_rate": 0.0001318710527343295, + "loss": 0.2356, + "step": 16931 + }, + { + "epoch": 0.3969127445981821, + "grad_norm": 4.257837772369385, + "learning_rate": 0.00013186407061971405, + "loss": 1.7461, + "step": 16932 + }, + { + "epoch": 0.3969361861729871, + "grad_norm": 5.308584213256836, + "learning_rate": 0.00013185708833220324, + "loss": 0.5263, + "step": 16933 + }, + { + "epoch": 0.3969596277477921, + "grad_norm": 0.8217501640319824, + "learning_rate": 0.00013185010587183496, + "loss": 0.0706, + "step": 16934 + }, + { + "epoch": 0.3969830693225971, + "grad_norm": 3.190077304840088, + "learning_rate": 0.00013184312323864713, + "loss": 0.3624, + "step": 16935 + }, + { + "epoch": 0.3970065108974021, + "grad_norm": 2.384814739227295, + "learning_rate": 0.00013183614043267764, + "loss": 0.1621, + "step": 16936 + }, + { + "epoch": 0.39702995247220707, + "grad_norm": 1.6704089641571045, + "learning_rate": 0.00013182915745396434, + "loss": 0.1837, + "step": 16937 + }, + { + "epoch": 0.39705339404701206, + "grad_norm": 5.907252788543701, + "learning_rate": 0.0001318221743025451, + "loss": 0.4238, + "step": 16938 + }, + { + "epoch": 0.39707683562181706, + "grad_norm": 3.3814592361450195, + "learning_rate": 0.0001318151909784579, + "loss": 0.4355, + "step": 16939 + }, + { + "epoch": 0.39710027719662205, + "grad_norm": 2.9110124111175537, + "learning_rate": 0.00013180820748174055, + "loss": 0.1914, + "step": 16940 + }, + { + "epoch": 0.39712371877142705, + "grad_norm": 0.9410304427146912, + "learning_rate": 0.00013180122381243096, + "loss": 0.0519, + "step": 16941 + }, + { + "epoch": 0.39714716034623204, + "grad_norm": 2.9143588542938232, + "learning_rate": 0.0001317942399705671, + "loss": 0.2757, + "step": 16942 + }, + { + "epoch": 0.39717060192103704, + "grad_norm": 6.47818660736084, + "learning_rate": 0.0001317872559561867, + "loss": 0.4162, + "step": 16943 + }, + { + "epoch": 0.39719404349584203, + "grad_norm": 2.4582884311676025, + "learning_rate": 0.00013178027176932782, + "loss": 0.2082, + "step": 16944 + }, + { + "epoch": 0.397217485070647, + "grad_norm": 0.9976354837417603, + "learning_rate": 0.00013177328741002827, + "loss": 0.1221, + "step": 16945 + }, + { + "epoch": 0.397240926645452, + "grad_norm": 4.9612956047058105, + "learning_rate": 0.000131766302878326, + "loss": 0.6088, + "step": 16946 + }, + { + "epoch": 0.397264368220257, + "grad_norm": 3.5099892616271973, + "learning_rate": 0.00013175931817425887, + "loss": 0.3302, + "step": 16947 + }, + { + "epoch": 0.397287809795062, + "grad_norm": 2.4521901607513428, + "learning_rate": 0.00013175233329786477, + "loss": 0.221, + "step": 16948 + }, + { + "epoch": 0.397311251369867, + "grad_norm": 3.791276454925537, + "learning_rate": 0.00013174534824918161, + "loss": 0.3165, + "step": 16949 + }, + { + "epoch": 0.397334692944672, + "grad_norm": 5.420030117034912, + "learning_rate": 0.00013173836302824728, + "loss": 0.4057, + "step": 16950 + }, + { + "epoch": 0.397358134519477, + "grad_norm": 4.901146411895752, + "learning_rate": 0.00013173137763509976, + "loss": 0.3967, + "step": 16951 + }, + { + "epoch": 0.39738157609428204, + "grad_norm": 2.2009546756744385, + "learning_rate": 0.00013172439206977687, + "loss": 0.2543, + "step": 16952 + }, + { + "epoch": 0.39740501766908704, + "grad_norm": 4.001559734344482, + "learning_rate": 0.0001317174063323165, + "loss": 0.5471, + "step": 16953 + }, + { + "epoch": 0.39742845924389203, + "grad_norm": 5.930922985076904, + "learning_rate": 0.00013171042042275664, + "loss": 0.379, + "step": 16954 + }, + { + "epoch": 0.397451900818697, + "grad_norm": 4.941786289215088, + "learning_rate": 0.0001317034343411351, + "loss": 0.3788, + "step": 16955 + }, + { + "epoch": 0.397475342393502, + "grad_norm": 4.45538854598999, + "learning_rate": 0.00013169644808748988, + "loss": 0.3208, + "step": 16956 + }, + { + "epoch": 0.397498783968307, + "grad_norm": 4.018911361694336, + "learning_rate": 0.0001316894616618588, + "loss": 0.6432, + "step": 16957 + }, + { + "epoch": 0.397522225543112, + "grad_norm": 3.082935333251953, + "learning_rate": 0.00013168247506427987, + "loss": 0.1594, + "step": 16958 + }, + { + "epoch": 0.397545667117917, + "grad_norm": 3.945175886154175, + "learning_rate": 0.0001316754882947909, + "loss": 0.4248, + "step": 16959 + }, + { + "epoch": 0.397569108692722, + "grad_norm": 2.586911916732788, + "learning_rate": 0.00013166850135342984, + "loss": 0.2723, + "step": 16960 + }, + { + "epoch": 0.397592550267527, + "grad_norm": 4.068972587585449, + "learning_rate": 0.00013166151424023455, + "loss": 0.2078, + "step": 16961 + }, + { + "epoch": 0.397615991842332, + "grad_norm": 1.705377221107483, + "learning_rate": 0.00013165452695524307, + "loss": 0.0918, + "step": 16962 + }, + { + "epoch": 0.397639433417137, + "grad_norm": 4.859923839569092, + "learning_rate": 0.0001316475394984932, + "loss": 0.6932, + "step": 16963 + }, + { + "epoch": 0.397662874991942, + "grad_norm": 1.2199034690856934, + "learning_rate": 0.0001316405518700229, + "loss": 0.1513, + "step": 16964 + }, + { + "epoch": 0.39768631656674697, + "grad_norm": 5.271660804748535, + "learning_rate": 0.0001316335640698701, + "loss": 0.4097, + "step": 16965 + }, + { + "epoch": 0.39770975814155196, + "grad_norm": 0.8792717456817627, + "learning_rate": 0.00013162657609807268, + "loss": 0.0366, + "step": 16966 + }, + { + "epoch": 0.39773319971635696, + "grad_norm": 2.9390811920166016, + "learning_rate": 0.00013161958795466854, + "loss": 0.233, + "step": 16967 + }, + { + "epoch": 0.39775664129116195, + "grad_norm": 3.7449803352355957, + "learning_rate": 0.00013161259963969564, + "loss": 0.4594, + "step": 16968 + }, + { + "epoch": 0.39778008286596694, + "grad_norm": 2.666006326675415, + "learning_rate": 0.0001316056111531919, + "loss": 0.1597, + "step": 16969 + }, + { + "epoch": 0.39780352444077194, + "grad_norm": 2.8770229816436768, + "learning_rate": 0.00013159862249519522, + "loss": 0.2987, + "step": 16970 + }, + { + "epoch": 0.39782696601557693, + "grad_norm": 2.0482234954833984, + "learning_rate": 0.00013159163366574354, + "loss": 0.1573, + "step": 16971 + }, + { + "epoch": 0.3978504075903819, + "grad_norm": 0.7538205981254578, + "learning_rate": 0.00013158464466487475, + "loss": 0.0638, + "step": 16972 + }, + { + "epoch": 0.3978738491651869, + "grad_norm": 2.8010034561157227, + "learning_rate": 0.00013157765549262678, + "loss": 2.0038, + "step": 16973 + }, + { + "epoch": 0.3978972907399919, + "grad_norm": 1.1581543684005737, + "learning_rate": 0.00013157066614903756, + "loss": 0.043, + "step": 16974 + }, + { + "epoch": 0.3979207323147969, + "grad_norm": 5.978827476501465, + "learning_rate": 0.00013156367663414505, + "loss": 0.3223, + "step": 16975 + }, + { + "epoch": 0.3979441738896019, + "grad_norm": 4.560390949249268, + "learning_rate": 0.0001315566869479871, + "loss": 0.4565, + "step": 16976 + }, + { + "epoch": 0.3979676154644069, + "grad_norm": 0.2602404057979584, + "learning_rate": 0.00013154969709060172, + "loss": 0.0215, + "step": 16977 + }, + { + "epoch": 0.3979910570392119, + "grad_norm": 5.9596848487854, + "learning_rate": 0.0001315427070620268, + "loss": 0.4573, + "step": 16978 + }, + { + "epoch": 0.3980144986140169, + "grad_norm": 0.8047796487808228, + "learning_rate": 0.00013153571686230024, + "loss": 0.0501, + "step": 16979 + }, + { + "epoch": 0.3980379401888219, + "grad_norm": 2.59065842628479, + "learning_rate": 0.00013152872649145999, + "loss": 0.1804, + "step": 16980 + }, + { + "epoch": 0.3980613817636269, + "grad_norm": 3.856198310852051, + "learning_rate": 0.00013152173594954399, + "loss": 0.3657, + "step": 16981 + }, + { + "epoch": 0.39808482333843187, + "grad_norm": 9.60666275024414, + "learning_rate": 0.00013151474523659016, + "loss": 0.4185, + "step": 16982 + }, + { + "epoch": 0.39810826491323686, + "grad_norm": 0.4867907464504242, + "learning_rate": 0.00013150775435263645, + "loss": 0.031, + "step": 16983 + }, + { + "epoch": 0.39813170648804186, + "grad_norm": 3.480980634689331, + "learning_rate": 0.0001315007632977208, + "loss": 0.5027, + "step": 16984 + }, + { + "epoch": 0.39815514806284685, + "grad_norm": 1.6376359462738037, + "learning_rate": 0.00013149377207188106, + "loss": 0.1535, + "step": 16985 + }, + { + "epoch": 0.39817858963765185, + "grad_norm": 2.9533698558807373, + "learning_rate": 0.00013148678067515524, + "loss": 0.3682, + "step": 16986 + }, + { + "epoch": 0.39820203121245684, + "grad_norm": 8.726914405822754, + "learning_rate": 0.00013147978910758133, + "loss": 0.3175, + "step": 16987 + }, + { + "epoch": 0.39822547278726184, + "grad_norm": 4.444146156311035, + "learning_rate": 0.0001314727973691971, + "loss": 0.3759, + "step": 16988 + }, + { + "epoch": 0.39824891436206683, + "grad_norm": 2.4555342197418213, + "learning_rate": 0.00013146580546004067, + "loss": 0.1748, + "step": 16989 + }, + { + "epoch": 0.3982723559368718, + "grad_norm": 3.6192612648010254, + "learning_rate": 0.00013145881338014987, + "loss": 0.4739, + "step": 16990 + }, + { + "epoch": 0.3982957975116768, + "grad_norm": 4.682916164398193, + "learning_rate": 0.00013145182112956266, + "loss": 0.5259, + "step": 16991 + }, + { + "epoch": 0.3983192390864818, + "grad_norm": 4.295762538909912, + "learning_rate": 0.00013144482870831696, + "loss": 0.4755, + "step": 16992 + }, + { + "epoch": 0.3983426806612868, + "grad_norm": 1.9701423645019531, + "learning_rate": 0.00013143783611645077, + "loss": 0.1281, + "step": 16993 + }, + { + "epoch": 0.3983661222360918, + "grad_norm": 23.43644142150879, + "learning_rate": 0.000131430843354002, + "loss": 0.3336, + "step": 16994 + }, + { + "epoch": 0.3983895638108968, + "grad_norm": 3.388268232345581, + "learning_rate": 0.00013142385042100855, + "loss": 0.3213, + "step": 16995 + }, + { + "epoch": 0.3984130053857018, + "grad_norm": 3.9987294673919678, + "learning_rate": 0.00013141685731750847, + "loss": 0.338, + "step": 16996 + }, + { + "epoch": 0.3984364469605068, + "grad_norm": 1.9514122009277344, + "learning_rate": 0.0001314098640435396, + "loss": 0.1168, + "step": 16997 + }, + { + "epoch": 0.3984598885353118, + "grad_norm": 2.9333176612854004, + "learning_rate": 0.00013140287059913987, + "loss": 0.2853, + "step": 16998 + }, + { + "epoch": 0.3984833301101168, + "grad_norm": 3.7800381183624268, + "learning_rate": 0.00013139587698434734, + "loss": 0.3187, + "step": 16999 + }, + { + "epoch": 0.39850677168492177, + "grad_norm": 3.3255815505981445, + "learning_rate": 0.00013138888319919994, + "loss": 0.3358, + "step": 17000 + }, + { + "epoch": 0.39853021325972676, + "grad_norm": 2.4403181076049805, + "learning_rate": 0.0001313818892437355, + "loss": 0.2157, + "step": 17001 + }, + { + "epoch": 0.39855365483453176, + "grad_norm": 1.0127452611923218, + "learning_rate": 0.0001313748951179921, + "loss": 0.1025, + "step": 17002 + }, + { + "epoch": 0.3985770964093368, + "grad_norm": 3.601656198501587, + "learning_rate": 0.0001313679008220076, + "loss": 0.2967, + "step": 17003 + }, + { + "epoch": 0.3986005379841418, + "grad_norm": 1.3378660678863525, + "learning_rate": 0.00013136090635582, + "loss": 0.1061, + "step": 17004 + }, + { + "epoch": 0.3986239795589468, + "grad_norm": 3.412816047668457, + "learning_rate": 0.00013135391171946722, + "loss": 0.4617, + "step": 17005 + }, + { + "epoch": 0.3986474211337518, + "grad_norm": 3.6629042625427246, + "learning_rate": 0.00013134691691298728, + "loss": 0.281, + "step": 17006 + }, + { + "epoch": 0.3986708627085568, + "grad_norm": 4.017382621765137, + "learning_rate": 0.00013133992193641803, + "loss": 0.4186, + "step": 17007 + }, + { + "epoch": 0.3986943042833618, + "grad_norm": 5.070995330810547, + "learning_rate": 0.00013133292678979753, + "loss": 0.6158, + "step": 17008 + }, + { + "epoch": 0.39871774585816677, + "grad_norm": 1.574256181716919, + "learning_rate": 0.0001313259314731637, + "loss": 0.1834, + "step": 17009 + }, + { + "epoch": 0.39874118743297177, + "grad_norm": 0.5620881915092468, + "learning_rate": 0.00013131893598655444, + "loss": 0.0581, + "step": 17010 + }, + { + "epoch": 0.39876462900777676, + "grad_norm": 0.49104443192481995, + "learning_rate": 0.00013131194033000776, + "loss": 0.03, + "step": 17011 + }, + { + "epoch": 0.39878807058258176, + "grad_norm": 3.4571170806884766, + "learning_rate": 0.00013130494450356163, + "loss": 0.3212, + "step": 17012 + }, + { + "epoch": 0.39881151215738675, + "grad_norm": 3.508945941925049, + "learning_rate": 0.000131297948507254, + "loss": 0.421, + "step": 17013 + }, + { + "epoch": 0.39883495373219174, + "grad_norm": 4.701171875, + "learning_rate": 0.0001312909523411228, + "loss": 0.4871, + "step": 17014 + }, + { + "epoch": 0.39885839530699674, + "grad_norm": 5.681580066680908, + "learning_rate": 0.000131283956005206, + "loss": 0.4405, + "step": 17015 + }, + { + "epoch": 0.39888183688180173, + "grad_norm": 3.368011951446533, + "learning_rate": 0.0001312769594995416, + "loss": 0.432, + "step": 17016 + }, + { + "epoch": 0.3989052784566067, + "grad_norm": 4.517488956451416, + "learning_rate": 0.00013126996282416753, + "loss": 0.5778, + "step": 17017 + }, + { + "epoch": 0.3989287200314117, + "grad_norm": 3.1744778156280518, + "learning_rate": 0.00013126296597912174, + "loss": 0.3516, + "step": 17018 + }, + { + "epoch": 0.3989521616062167, + "grad_norm": 1.3704917430877686, + "learning_rate": 0.00013125596896444228, + "loss": 0.218, + "step": 17019 + }, + { + "epoch": 0.3989756031810217, + "grad_norm": 4.215203762054443, + "learning_rate": 0.00013124897178016698, + "loss": 0.4268, + "step": 17020 + }, + { + "epoch": 0.3989990447558267, + "grad_norm": 3.235196352005005, + "learning_rate": 0.00013124197442633396, + "loss": 0.4366, + "step": 17021 + }, + { + "epoch": 0.3990224863306317, + "grad_norm": 2.9893124103546143, + "learning_rate": 0.00013123497690298105, + "loss": 0.2227, + "step": 17022 + }, + { + "epoch": 0.3990459279054367, + "grad_norm": 2.2717673778533936, + "learning_rate": 0.0001312279792101463, + "loss": 1.6038, + "step": 17023 + }, + { + "epoch": 0.3990693694802417, + "grad_norm": 3.4130938053131104, + "learning_rate": 0.0001312209813478676, + "loss": 1.3538, + "step": 17024 + }, + { + "epoch": 0.3990928110550467, + "grad_norm": 1.4172403812408447, + "learning_rate": 0.00013121398331618307, + "loss": 0.2433, + "step": 17025 + }, + { + "epoch": 0.3991162526298517, + "grad_norm": 1.0559849739074707, + "learning_rate": 0.00013120698511513053, + "loss": 0.1089, + "step": 17026 + }, + { + "epoch": 0.39913969420465667, + "grad_norm": 4.196305274963379, + "learning_rate": 0.00013119998674474803, + "loss": 0.3096, + "step": 17027 + }, + { + "epoch": 0.39916313577946166, + "grad_norm": 3.2006447315216064, + "learning_rate": 0.00013119298820507352, + "loss": 0.2388, + "step": 17028 + }, + { + "epoch": 0.39918657735426666, + "grad_norm": 3.0242574214935303, + "learning_rate": 0.00013118598949614499, + "loss": 0.3056, + "step": 17029 + }, + { + "epoch": 0.39921001892907165, + "grad_norm": 2.719193696975708, + "learning_rate": 0.00013117899061800036, + "loss": 0.1749, + "step": 17030 + }, + { + "epoch": 0.39923346050387665, + "grad_norm": 1.394025444984436, + "learning_rate": 0.0001311719915706777, + "loss": 0.1627, + "step": 17031 + }, + { + "epoch": 0.39925690207868164, + "grad_norm": 0.7226278781890869, + "learning_rate": 0.00013116499235421493, + "loss": 0.0598, + "step": 17032 + }, + { + "epoch": 0.39928034365348664, + "grad_norm": 0.6765469312667847, + "learning_rate": 0.00013115799296865, + "loss": 0.0731, + "step": 17033 + }, + { + "epoch": 0.39930378522829163, + "grad_norm": 1.2916152477264404, + "learning_rate": 0.00013115099341402096, + "loss": 0.1363, + "step": 17034 + }, + { + "epoch": 0.3993272268030966, + "grad_norm": 2.248504400253296, + "learning_rate": 0.00013114399369036574, + "loss": 0.1407, + "step": 17035 + }, + { + "epoch": 0.3993506683779016, + "grad_norm": 3.153984785079956, + "learning_rate": 0.00013113699379772232, + "loss": 0.3684, + "step": 17036 + }, + { + "epoch": 0.3993741099527066, + "grad_norm": 4.460757732391357, + "learning_rate": 0.00013112999373612874, + "loss": 0.5664, + "step": 17037 + }, + { + "epoch": 0.3993975515275116, + "grad_norm": 3.8230206966400146, + "learning_rate": 0.0001311229935056229, + "loss": 0.4496, + "step": 17038 + }, + { + "epoch": 0.3994209931023166, + "grad_norm": 2.913492441177368, + "learning_rate": 0.00013111599310624284, + "loss": 1.1562, + "step": 17039 + }, + { + "epoch": 0.3994444346771216, + "grad_norm": 3.7298977375030518, + "learning_rate": 0.00013110899253802652, + "loss": 0.5139, + "step": 17040 + }, + { + "epoch": 0.3994678762519266, + "grad_norm": 4.846226215362549, + "learning_rate": 0.00013110199180101195, + "loss": 0.6125, + "step": 17041 + }, + { + "epoch": 0.3994913178267316, + "grad_norm": 3.021246910095215, + "learning_rate": 0.00013109499089523707, + "loss": 0.2242, + "step": 17042 + }, + { + "epoch": 0.3995147594015366, + "grad_norm": 2.459315299987793, + "learning_rate": 0.0001310879898207399, + "loss": 0.2157, + "step": 17043 + }, + { + "epoch": 0.3995382009763416, + "grad_norm": 4.47776985168457, + "learning_rate": 0.00013108098857755844, + "loss": 0.8655, + "step": 17044 + }, + { + "epoch": 0.39956164255114657, + "grad_norm": 3.6007494926452637, + "learning_rate": 0.00013107398716573064, + "loss": 0.4605, + "step": 17045 + }, + { + "epoch": 0.39958508412595156, + "grad_norm": 4.170523166656494, + "learning_rate": 0.0001310669855852945, + "loss": 0.4485, + "step": 17046 + }, + { + "epoch": 0.39960852570075656, + "grad_norm": 3.5553064346313477, + "learning_rate": 0.0001310599838362881, + "loss": 0.311, + "step": 17047 + }, + { + "epoch": 0.39963196727556155, + "grad_norm": 3.7666025161743164, + "learning_rate": 0.0001310529819187493, + "loss": 0.3981, + "step": 17048 + }, + { + "epoch": 0.39965540885036654, + "grad_norm": 3.573460578918457, + "learning_rate": 0.00013104597983271612, + "loss": 1.5711, + "step": 17049 + }, + { + "epoch": 0.39967885042517154, + "grad_norm": 3.287876605987549, + "learning_rate": 0.0001310389775782266, + "loss": 0.3002, + "step": 17050 + }, + { + "epoch": 0.39970229199997653, + "grad_norm": 0.5694274306297302, + "learning_rate": 0.00013103197515531876, + "loss": 0.0375, + "step": 17051 + }, + { + "epoch": 0.3997257335747815, + "grad_norm": 4.262608528137207, + "learning_rate": 0.00013102497256403046, + "loss": 0.1641, + "step": 17052 + }, + { + "epoch": 0.3997491751495865, + "grad_norm": 1.6098685264587402, + "learning_rate": 0.00013101796980439985, + "loss": 0.1515, + "step": 17053 + }, + { + "epoch": 0.39977261672439157, + "grad_norm": 3.468869209289551, + "learning_rate": 0.00013101096687646489, + "loss": 0.3948, + "step": 17054 + }, + { + "epoch": 0.39979605829919657, + "grad_norm": 3.853780508041382, + "learning_rate": 0.0001310039637802635, + "loss": 0.446, + "step": 17055 + }, + { + "epoch": 0.39981949987400156, + "grad_norm": 1.3128674030303955, + "learning_rate": 0.00013099696051583376, + "loss": 0.0677, + "step": 17056 + }, + { + "epoch": 0.39984294144880655, + "grad_norm": 5.05754280090332, + "learning_rate": 0.00013098995708321363, + "loss": 0.7841, + "step": 17057 + }, + { + "epoch": 0.39986638302361155, + "grad_norm": 2.061432123184204, + "learning_rate": 0.0001309829534824411, + "loss": 0.2281, + "step": 17058 + }, + { + "epoch": 0.39988982459841654, + "grad_norm": 2.365220785140991, + "learning_rate": 0.0001309759497135542, + "loss": 0.2678, + "step": 17059 + }, + { + "epoch": 0.39991326617322154, + "grad_norm": 3.7361233234405518, + "learning_rate": 0.00013096894577659094, + "loss": 0.2473, + "step": 17060 + }, + { + "epoch": 0.39993670774802653, + "grad_norm": 0.8294740319252014, + "learning_rate": 0.0001309619416715893, + "loss": 0.0777, + "step": 17061 + }, + { + "epoch": 0.3999601493228315, + "grad_norm": 3.320962905883789, + "learning_rate": 0.0001309549373985873, + "loss": 0.2854, + "step": 17062 + }, + { + "epoch": 0.3999835908976365, + "grad_norm": 5.213837623596191, + "learning_rate": 0.00013094793295762296, + "loss": 0.6311, + "step": 17063 + }, + { + "epoch": 0.4000070324724415, + "grad_norm": 1.8400071859359741, + "learning_rate": 0.00013094092834873425, + "loss": 0.181, + "step": 17064 + }, + { + "epoch": 0.4000304740472465, + "grad_norm": 3.0856575965881348, + "learning_rate": 0.00013093392357195916, + "loss": 0.4055, + "step": 17065 + }, + { + "epoch": 0.4000539156220515, + "grad_norm": 5.023142337799072, + "learning_rate": 0.00013092691862733578, + "loss": 0.4966, + "step": 17066 + }, + { + "epoch": 0.4000773571968565, + "grad_norm": 0.7129749655723572, + "learning_rate": 0.00013091991351490205, + "loss": 0.0893, + "step": 17067 + }, + { + "epoch": 0.4001007987716615, + "grad_norm": 1.5276025533676147, + "learning_rate": 0.00013091290823469598, + "loss": 0.0708, + "step": 17068 + }, + { + "epoch": 0.4001242403464665, + "grad_norm": 2.7603960037231445, + "learning_rate": 0.00013090590278675562, + "loss": 0.2761, + "step": 17069 + }, + { + "epoch": 0.4001476819212715, + "grad_norm": 0.6189796924591064, + "learning_rate": 0.00013089889717111898, + "loss": 0.0734, + "step": 17070 + }, + { + "epoch": 0.4001711234960765, + "grad_norm": 2.422558546066284, + "learning_rate": 0.00013089189138782402, + "loss": 0.2715, + "step": 17071 + }, + { + "epoch": 0.40019456507088147, + "grad_norm": 6.633198261260986, + "learning_rate": 0.0001308848854369088, + "loss": 0.6871, + "step": 17072 + }, + { + "epoch": 0.40021800664568646, + "grad_norm": 1.855206847190857, + "learning_rate": 0.00013087787931841133, + "loss": 0.0817, + "step": 17073 + }, + { + "epoch": 0.40024144822049146, + "grad_norm": 1.9147422313690186, + "learning_rate": 0.00013087087303236962, + "loss": 0.3044, + "step": 17074 + }, + { + "epoch": 0.40026488979529645, + "grad_norm": 0.41395241022109985, + "learning_rate": 0.0001308638665788217, + "loss": 0.0392, + "step": 17075 + }, + { + "epoch": 0.40028833137010145, + "grad_norm": 1.096280574798584, + "learning_rate": 0.00013085685995780554, + "loss": 0.0806, + "step": 17076 + }, + { + "epoch": 0.40031177294490644, + "grad_norm": 4.0526323318481445, + "learning_rate": 0.0001308498531693592, + "loss": 0.4516, + "step": 17077 + }, + { + "epoch": 0.40033521451971144, + "grad_norm": 2.449371099472046, + "learning_rate": 0.0001308428462135207, + "loss": 0.2214, + "step": 17078 + }, + { + "epoch": 0.40035865609451643, + "grad_norm": 2.657694101333618, + "learning_rate": 0.00013083583909032805, + "loss": 0.2176, + "step": 17079 + }, + { + "epoch": 0.4003820976693214, + "grad_norm": 3.2171127796173096, + "learning_rate": 0.00013082883179981928, + "loss": 1.6335, + "step": 17080 + }, + { + "epoch": 0.4004055392441264, + "grad_norm": 5.090291500091553, + "learning_rate": 0.00013082182434203237, + "loss": 1.7607, + "step": 17081 + }, + { + "epoch": 0.4004289808189314, + "grad_norm": 2.976057291030884, + "learning_rate": 0.0001308148167170054, + "loss": 0.2865, + "step": 17082 + }, + { + "epoch": 0.4004524223937364, + "grad_norm": 6.260116100311279, + "learning_rate": 0.00013080780892477635, + "loss": 0.6925, + "step": 17083 + }, + { + "epoch": 0.4004758639685414, + "grad_norm": 1.0478988885879517, + "learning_rate": 0.00013080080096538325, + "loss": 0.1435, + "step": 17084 + }, + { + "epoch": 0.4004993055433464, + "grad_norm": 1.4181325435638428, + "learning_rate": 0.00013079379283886417, + "loss": 0.1647, + "step": 17085 + }, + { + "epoch": 0.4005227471181514, + "grad_norm": 1.586413025856018, + "learning_rate": 0.0001307867845452571, + "loss": 0.2343, + "step": 17086 + }, + { + "epoch": 0.4005461886929564, + "grad_norm": 3.147629737854004, + "learning_rate": 0.00013077977608460007, + "loss": 0.4138, + "step": 17087 + }, + { + "epoch": 0.4005696302677614, + "grad_norm": 2.7909510135650635, + "learning_rate": 0.0001307727674569311, + "loss": 0.3184, + "step": 17088 + }, + { + "epoch": 0.4005930718425664, + "grad_norm": 5.736172676086426, + "learning_rate": 0.0001307657586622882, + "loss": 0.6892, + "step": 17089 + }, + { + "epoch": 0.40061651341737137, + "grad_norm": 2.904850959777832, + "learning_rate": 0.00013075874970070947, + "loss": 0.2885, + "step": 17090 + }, + { + "epoch": 0.40063995499217636, + "grad_norm": 0.47171035408973694, + "learning_rate": 0.0001307517405722329, + "loss": 0.0383, + "step": 17091 + }, + { + "epoch": 0.40066339656698136, + "grad_norm": 3.3766329288482666, + "learning_rate": 0.00013074473127689653, + "loss": 0.3509, + "step": 17092 + }, + { + "epoch": 0.40068683814178635, + "grad_norm": 1.1239311695098877, + "learning_rate": 0.00013073772181473834, + "loss": 0.1319, + "step": 17093 + }, + { + "epoch": 0.40071027971659134, + "grad_norm": 2.288252115249634, + "learning_rate": 0.00013073071218579644, + "loss": 0.1496, + "step": 17094 + }, + { + "epoch": 0.40073372129139634, + "grad_norm": 2.868248701095581, + "learning_rate": 0.0001307237023901088, + "loss": 0.3701, + "step": 17095 + }, + { + "epoch": 0.40075716286620133, + "grad_norm": 4.537121295928955, + "learning_rate": 0.00013071669242771354, + "loss": 0.6747, + "step": 17096 + }, + { + "epoch": 0.4007806044410063, + "grad_norm": 6.30523157119751, + "learning_rate": 0.00013070968229864858, + "loss": 0.7525, + "step": 17097 + }, + { + "epoch": 0.4008040460158113, + "grad_norm": 1.0705571174621582, + "learning_rate": 0.00013070267200295207, + "loss": 0.1468, + "step": 17098 + }, + { + "epoch": 0.4008274875906163, + "grad_norm": 6.465067386627197, + "learning_rate": 0.000130695661540662, + "loss": 0.5508, + "step": 17099 + }, + { + "epoch": 0.4008509291654213, + "grad_norm": 2.6182000637054443, + "learning_rate": 0.0001306886509118164, + "loss": 1.6553, + "step": 17100 + }, + { + "epoch": 0.4008743707402263, + "grad_norm": 2.1291182041168213, + "learning_rate": 0.00013068164011645328, + "loss": 0.1724, + "step": 17101 + }, + { + "epoch": 0.4008978123150313, + "grad_norm": 4.837392807006836, + "learning_rate": 0.00013067462915461076, + "loss": 0.6578, + "step": 17102 + }, + { + "epoch": 0.4009212538898363, + "grad_norm": 1.9800482988357544, + "learning_rate": 0.0001306676180263268, + "loss": 0.1274, + "step": 17103 + }, + { + "epoch": 0.4009446954646413, + "grad_norm": 5.0814619064331055, + "learning_rate": 0.0001306606067316395, + "loss": 0.6117, + "step": 17104 + }, + { + "epoch": 0.40096813703944634, + "grad_norm": 7.4738054275512695, + "learning_rate": 0.00013065359527058692, + "loss": 0.7071, + "step": 17105 + }, + { + "epoch": 0.40099157861425133, + "grad_norm": 2.67460036277771, + "learning_rate": 0.00013064658364320703, + "loss": 0.2461, + "step": 17106 + }, + { + "epoch": 0.4010150201890563, + "grad_norm": 1.0319859981536865, + "learning_rate": 0.00013063957184953793, + "loss": 0.1329, + "step": 17107 + }, + { + "epoch": 0.4010384617638613, + "grad_norm": 3.424172878265381, + "learning_rate": 0.00013063255988961765, + "loss": 0.2706, + "step": 17108 + }, + { + "epoch": 0.4010619033386663, + "grad_norm": 2.55816388130188, + "learning_rate": 0.00013062554776348425, + "loss": 0.2953, + "step": 17109 + }, + { + "epoch": 0.4010853449134713, + "grad_norm": 3.497087001800537, + "learning_rate": 0.00013061853547117574, + "loss": 0.4995, + "step": 17110 + }, + { + "epoch": 0.4011087864882763, + "grad_norm": 3.7159483432769775, + "learning_rate": 0.00013061152301273023, + "loss": 0.4355, + "step": 17111 + }, + { + "epoch": 0.4011322280630813, + "grad_norm": 3.7442526817321777, + "learning_rate": 0.00013060451038818572, + "loss": 0.3595, + "step": 17112 + }, + { + "epoch": 0.4011556696378863, + "grad_norm": 1.4087883234024048, + "learning_rate": 0.00013059749759758023, + "loss": 0.2697, + "step": 17113 + }, + { + "epoch": 0.4011791112126913, + "grad_norm": 2.3623344898223877, + "learning_rate": 0.00013059048464095194, + "loss": 1.7732, + "step": 17114 + }, + { + "epoch": 0.4012025527874963, + "grad_norm": 3.3614161014556885, + "learning_rate": 0.00013058347151833876, + "loss": 0.2607, + "step": 17115 + }, + { + "epoch": 0.4012259943623013, + "grad_norm": 2.9917359352111816, + "learning_rate": 0.00013057645822977883, + "loss": 0.2488, + "step": 17116 + }, + { + "epoch": 0.40124943593710627, + "grad_norm": 4.109788417816162, + "learning_rate": 0.00013056944477531017, + "loss": 0.3741, + "step": 17117 + }, + { + "epoch": 0.40127287751191126, + "grad_norm": 2.5731189250946045, + "learning_rate": 0.00013056243115497083, + "loss": 0.2247, + "step": 17118 + }, + { + "epoch": 0.40129631908671626, + "grad_norm": 3.2395808696746826, + "learning_rate": 0.00013055541736879892, + "loss": 1.3476, + "step": 17119 + }, + { + "epoch": 0.40131976066152125, + "grad_norm": 4.383813858032227, + "learning_rate": 0.0001305484034168324, + "loss": 0.4151, + "step": 17120 + }, + { + "epoch": 0.40134320223632625, + "grad_norm": 0.7286826968193054, + "learning_rate": 0.00013054138929910944, + "loss": 0.0495, + "step": 17121 + }, + { + "epoch": 0.40136664381113124, + "grad_norm": 3.9165806770324707, + "learning_rate": 0.000130534375015668, + "loss": 0.5849, + "step": 17122 + }, + { + "epoch": 0.40139008538593623, + "grad_norm": 3.1115217208862305, + "learning_rate": 0.00013052736056654623, + "loss": 0.2902, + "step": 17123 + }, + { + "epoch": 0.40141352696074123, + "grad_norm": 1.3804080486297607, + "learning_rate": 0.0001305203459517821, + "loss": 0.1656, + "step": 17124 + }, + { + "epoch": 0.4014369685355462, + "grad_norm": 3.3920111656188965, + "learning_rate": 0.0001305133311714137, + "loss": 0.5188, + "step": 17125 + }, + { + "epoch": 0.4014604101103512, + "grad_norm": 3.8248462677001953, + "learning_rate": 0.00013050631622547917, + "loss": 0.2801, + "step": 17126 + }, + { + "epoch": 0.4014838516851562, + "grad_norm": 1.134421467781067, + "learning_rate": 0.00013049930111401646, + "loss": 0.0831, + "step": 17127 + }, + { + "epoch": 0.4015072932599612, + "grad_norm": 4.005937576293945, + "learning_rate": 0.0001304922858370637, + "loss": 0.4893, + "step": 17128 + }, + { + "epoch": 0.4015307348347662, + "grad_norm": 0.8737419843673706, + "learning_rate": 0.00013048527039465896, + "loss": 0.1048, + "step": 17129 + }, + { + "epoch": 0.4015541764095712, + "grad_norm": 2.5686440467834473, + "learning_rate": 0.00013047825478684026, + "loss": 0.2959, + "step": 17130 + }, + { + "epoch": 0.4015776179843762, + "grad_norm": 1.4745233058929443, + "learning_rate": 0.00013047123901364569, + "loss": 0.2656, + "step": 17131 + }, + { + "epoch": 0.4016010595591812, + "grad_norm": 2.8478100299835205, + "learning_rate": 0.00013046422307511332, + "loss": 0.3718, + "step": 17132 + }, + { + "epoch": 0.4016245011339862, + "grad_norm": 1.5994752645492554, + "learning_rate": 0.00013045720697128123, + "loss": 0.1527, + "step": 17133 + }, + { + "epoch": 0.4016479427087912, + "grad_norm": 3.1227548122406006, + "learning_rate": 0.0001304501907021875, + "loss": 0.4656, + "step": 17134 + }, + { + "epoch": 0.40167138428359617, + "grad_norm": 0.9757754802703857, + "learning_rate": 0.00013044317426787015, + "loss": 0.0942, + "step": 17135 + }, + { + "epoch": 0.40169482585840116, + "grad_norm": 2.9016377925872803, + "learning_rate": 0.00013043615766836726, + "loss": 1.2203, + "step": 17136 + }, + { + "epoch": 0.40171826743320616, + "grad_norm": 0.5123389959335327, + "learning_rate": 0.00013042914090371697, + "loss": 0.0417, + "step": 17137 + }, + { + "epoch": 0.40174170900801115, + "grad_norm": 4.819086074829102, + "learning_rate": 0.00013042212397395726, + "loss": 0.4725, + "step": 17138 + }, + { + "epoch": 0.40176515058281614, + "grad_norm": 5.17697811126709, + "learning_rate": 0.00013041510687912628, + "loss": 0.5057, + "step": 17139 + }, + { + "epoch": 0.40178859215762114, + "grad_norm": 1.0152473449707031, + "learning_rate": 0.0001304080896192621, + "loss": 0.0943, + "step": 17140 + }, + { + "epoch": 0.40181203373242613, + "grad_norm": 3.3079707622528076, + "learning_rate": 0.0001304010721944027, + "loss": 0.4432, + "step": 17141 + }, + { + "epoch": 0.4018354753072311, + "grad_norm": 1.5665034055709839, + "learning_rate": 0.0001303940546045863, + "loss": 0.2067, + "step": 17142 + }, + { + "epoch": 0.4018589168820361, + "grad_norm": 1.8160665035247803, + "learning_rate": 0.00013038703684985087, + "loss": 0.2339, + "step": 17143 + }, + { + "epoch": 0.4018823584568411, + "grad_norm": 5.920079231262207, + "learning_rate": 0.00013038001893023448, + "loss": 0.752, + "step": 17144 + }, + { + "epoch": 0.4019058000316461, + "grad_norm": 2.883881092071533, + "learning_rate": 0.00013037300084577533, + "loss": 0.3244, + "step": 17145 + }, + { + "epoch": 0.4019292416064511, + "grad_norm": 1.1719964742660522, + "learning_rate": 0.0001303659825965114, + "loss": 0.0876, + "step": 17146 + }, + { + "epoch": 0.4019526831812561, + "grad_norm": 4.708256721496582, + "learning_rate": 0.00013035896418248078, + "loss": 0.5789, + "step": 17147 + }, + { + "epoch": 0.4019761247560611, + "grad_norm": 3.519477128982544, + "learning_rate": 0.00013035194560372156, + "loss": 0.3259, + "step": 17148 + }, + { + "epoch": 0.4019995663308661, + "grad_norm": 2.275918960571289, + "learning_rate": 0.00013034492686027187, + "loss": 1.5595, + "step": 17149 + }, + { + "epoch": 0.4020230079056711, + "grad_norm": 3.900806188583374, + "learning_rate": 0.0001303379079521697, + "loss": 0.6116, + "step": 17150 + }, + { + "epoch": 0.4020464494804761, + "grad_norm": 2.9036054611206055, + "learning_rate": 0.0001303308888794532, + "loss": 0.3944, + "step": 17151 + }, + { + "epoch": 0.40206989105528107, + "grad_norm": 3.107698917388916, + "learning_rate": 0.00013032386964216048, + "loss": 0.2821, + "step": 17152 + }, + { + "epoch": 0.40209333263008606, + "grad_norm": 4.674879550933838, + "learning_rate": 0.0001303168502403296, + "loss": 0.3356, + "step": 17153 + }, + { + "epoch": 0.40211677420489106, + "grad_norm": 0.7006890177726746, + "learning_rate": 0.00013030983067399858, + "loss": 0.0368, + "step": 17154 + }, + { + "epoch": 0.40214021577969605, + "grad_norm": 3.60227370262146, + "learning_rate": 0.0001303028109432056, + "loss": 0.3227, + "step": 17155 + }, + { + "epoch": 0.4021636573545011, + "grad_norm": 4.724165439605713, + "learning_rate": 0.0001302957910479887, + "loss": 1.7957, + "step": 17156 + }, + { + "epoch": 0.4021870989293061, + "grad_norm": 2.101491689682007, + "learning_rate": 0.00013028877098838598, + "loss": 1.168, + "step": 17157 + }, + { + "epoch": 0.4022105405041111, + "grad_norm": 2.4926350116729736, + "learning_rate": 0.00013028175076443559, + "loss": 0.1803, + "step": 17158 + }, + { + "epoch": 0.4022339820789161, + "grad_norm": 3.4214344024658203, + "learning_rate": 0.00013027473037617552, + "loss": 0.3255, + "step": 17159 + }, + { + "epoch": 0.4022574236537211, + "grad_norm": 1.282702922821045, + "learning_rate": 0.00013026770982364392, + "loss": 0.1603, + "step": 17160 + }, + { + "epoch": 0.4022808652285261, + "grad_norm": 3.171541213989258, + "learning_rate": 0.0001302606891068789, + "loss": 0.3228, + "step": 17161 + }, + { + "epoch": 0.40230430680333107, + "grad_norm": 4.774237155914307, + "learning_rate": 0.0001302536682259185, + "loss": 0.6349, + "step": 17162 + }, + { + "epoch": 0.40232774837813606, + "grad_norm": 2.513930082321167, + "learning_rate": 0.00013024664718080085, + "loss": 0.181, + "step": 17163 + }, + { + "epoch": 0.40235118995294106, + "grad_norm": 2.084857940673828, + "learning_rate": 0.00013023962597156404, + "loss": 0.2119, + "step": 17164 + }, + { + "epoch": 0.40237463152774605, + "grad_norm": 2.8173961639404297, + "learning_rate": 0.00013023260459824618, + "loss": 0.2492, + "step": 17165 + }, + { + "epoch": 0.40239807310255105, + "grad_norm": 6.51653528213501, + "learning_rate": 0.00013022558306088536, + "loss": 0.7108, + "step": 17166 + }, + { + "epoch": 0.40242151467735604, + "grad_norm": 2.7729732990264893, + "learning_rate": 0.00013021856135951967, + "loss": 0.2048, + "step": 17167 + }, + { + "epoch": 0.40244495625216103, + "grad_norm": 0.684952437877655, + "learning_rate": 0.0001302115394941872, + "loss": 0.0564, + "step": 17168 + }, + { + "epoch": 0.40246839782696603, + "grad_norm": 3.3957159519195557, + "learning_rate": 0.0001302045174649261, + "loss": 0.5778, + "step": 17169 + }, + { + "epoch": 0.402491839401771, + "grad_norm": 3.6952762603759766, + "learning_rate": 0.0001301974952717744, + "loss": 0.317, + "step": 17170 + }, + { + "epoch": 0.402515280976576, + "grad_norm": 1.8173094987869263, + "learning_rate": 0.00013019047291477028, + "loss": 0.2475, + "step": 17171 + }, + { + "epoch": 0.402538722551381, + "grad_norm": 3.932415246963501, + "learning_rate": 0.0001301834503939518, + "loss": 0.5953, + "step": 17172 + }, + { + "epoch": 0.402562164126186, + "grad_norm": 1.0349838733673096, + "learning_rate": 0.00013017642770935705, + "loss": 0.0623, + "step": 17173 + }, + { + "epoch": 0.402585605700991, + "grad_norm": 2.019526958465576, + "learning_rate": 0.00013016940486102415, + "loss": 0.1656, + "step": 17174 + }, + { + "epoch": 0.402609047275796, + "grad_norm": 1.5543851852416992, + "learning_rate": 0.00013016238184899122, + "loss": 0.1688, + "step": 17175 + }, + { + "epoch": 0.402632488850601, + "grad_norm": 1.5873308181762695, + "learning_rate": 0.00013015535867329638, + "loss": 1.0198, + "step": 17176 + }, + { + "epoch": 0.402655930425406, + "grad_norm": 1.7498571872711182, + "learning_rate": 0.0001301483353339777, + "loss": 0.1445, + "step": 17177 + }, + { + "epoch": 0.402679372000211, + "grad_norm": 3.4856696128845215, + "learning_rate": 0.00013014131183107332, + "loss": 0.3125, + "step": 17178 + }, + { + "epoch": 0.40270281357501597, + "grad_norm": 4.2270965576171875, + "learning_rate": 0.0001301342881646213, + "loss": 0.5022, + "step": 17179 + }, + { + "epoch": 0.40272625514982097, + "grad_norm": 3.4243738651275635, + "learning_rate": 0.00013012726433465977, + "loss": 0.2653, + "step": 17180 + }, + { + "epoch": 0.40274969672462596, + "grad_norm": 3.710308790206909, + "learning_rate": 0.0001301202403412269, + "loss": 0.3594, + "step": 17181 + }, + { + "epoch": 0.40277313829943096, + "grad_norm": 1.9719204902648926, + "learning_rate": 0.00013011321618436076, + "loss": 0.0873, + "step": 17182 + }, + { + "epoch": 0.40279657987423595, + "grad_norm": 2.7961204051971436, + "learning_rate": 0.00013010619186409948, + "loss": 0.3149, + "step": 17183 + }, + { + "epoch": 0.40282002144904094, + "grad_norm": 3.2535130977630615, + "learning_rate": 0.00013009916738048111, + "loss": 0.2896, + "step": 17184 + }, + { + "epoch": 0.40284346302384594, + "grad_norm": 1.9053235054016113, + "learning_rate": 0.00013009214273354382, + "loss": 0.2513, + "step": 17185 + }, + { + "epoch": 0.40286690459865093, + "grad_norm": 0.996379554271698, + "learning_rate": 0.00013008511792332573, + "loss": 0.1454, + "step": 17186 + }, + { + "epoch": 0.4028903461734559, + "grad_norm": 2.913341760635376, + "learning_rate": 0.00013007809294986492, + "loss": 0.1523, + "step": 17187 + }, + { + "epoch": 0.4029137877482609, + "grad_norm": 0.8074911236763, + "learning_rate": 0.00013007106781319958, + "loss": 0.0411, + "step": 17188 + }, + { + "epoch": 0.4029372293230659, + "grad_norm": 0.9178690910339355, + "learning_rate": 0.00013006404251336774, + "loss": 0.0804, + "step": 17189 + }, + { + "epoch": 0.4029606708978709, + "grad_norm": 2.597315549850464, + "learning_rate": 0.00013005701705040757, + "loss": 0.1788, + "step": 17190 + }, + { + "epoch": 0.4029841124726759, + "grad_norm": 0.9885777831077576, + "learning_rate": 0.00013004999142435722, + "loss": 0.0538, + "step": 17191 + }, + { + "epoch": 0.4030075540474809, + "grad_norm": 2.518000841140747, + "learning_rate": 0.0001300429656352547, + "loss": 1.4691, + "step": 17192 + }, + { + "epoch": 0.4030309956222859, + "grad_norm": 3.9602725505828857, + "learning_rate": 0.00013003593968313824, + "loss": 0.4233, + "step": 17193 + }, + { + "epoch": 0.4030544371970909, + "grad_norm": 3.4479289054870605, + "learning_rate": 0.00013002891356804594, + "loss": 0.3676, + "step": 17194 + }, + { + "epoch": 0.4030778787718959, + "grad_norm": 4.304924011230469, + "learning_rate": 0.00013002188729001587, + "loss": 0.3733, + "step": 17195 + }, + { + "epoch": 0.4031013203467009, + "grad_norm": 3.5662801265716553, + "learning_rate": 0.00013001486084908623, + "loss": 0.4416, + "step": 17196 + }, + { + "epoch": 0.40312476192150587, + "grad_norm": 4.025850296020508, + "learning_rate": 0.00013000783424529512, + "loss": 0.7363, + "step": 17197 + }, + { + "epoch": 0.40314820349631086, + "grad_norm": 0.3241097927093506, + "learning_rate": 0.00013000080747868062, + "loss": 0.0185, + "step": 17198 + }, + { + "epoch": 0.40317164507111586, + "grad_norm": 5.0417704582214355, + "learning_rate": 0.00012999378054928093, + "loss": 0.7861, + "step": 17199 + }, + { + "epoch": 0.40319508664592085, + "grad_norm": 3.1762638092041016, + "learning_rate": 0.00012998675345713413, + "loss": 0.247, + "step": 17200 + }, + { + "epoch": 0.40321852822072585, + "grad_norm": 3.9111859798431396, + "learning_rate": 0.00012997972620227833, + "loss": 0.2825, + "step": 17201 + }, + { + "epoch": 0.40324196979553084, + "grad_norm": 4.385280609130859, + "learning_rate": 0.00012997269878475175, + "loss": 0.3231, + "step": 17202 + }, + { + "epoch": 0.40326541137033584, + "grad_norm": 0.8080503940582275, + "learning_rate": 0.00012996567120459246, + "loss": 0.0883, + "step": 17203 + }, + { + "epoch": 0.40328885294514083, + "grad_norm": 2.3404996395111084, + "learning_rate": 0.00012995864346183856, + "loss": 0.203, + "step": 17204 + }, + { + "epoch": 0.4033122945199458, + "grad_norm": 2.3323707580566406, + "learning_rate": 0.00012995161555652822, + "loss": 0.1972, + "step": 17205 + }, + { + "epoch": 0.4033357360947508, + "grad_norm": 2.7154977321624756, + "learning_rate": 0.0001299445874886996, + "loss": 0.1984, + "step": 17206 + }, + { + "epoch": 0.40335917766955587, + "grad_norm": 10.990004539489746, + "learning_rate": 0.00012993755925839076, + "loss": 0.2035, + "step": 17207 + }, + { + "epoch": 0.40338261924436086, + "grad_norm": 1.3200931549072266, + "learning_rate": 0.0001299305308656399, + "loss": 0.1177, + "step": 17208 + }, + { + "epoch": 0.40340606081916586, + "grad_norm": 2.5137786865234375, + "learning_rate": 0.00012992350231048517, + "loss": 0.198, + "step": 17209 + }, + { + "epoch": 0.40342950239397085, + "grad_norm": 2.4062440395355225, + "learning_rate": 0.00012991647359296467, + "loss": 0.2221, + "step": 17210 + }, + { + "epoch": 0.40345294396877585, + "grad_norm": 6.240853786468506, + "learning_rate": 0.0001299094447131165, + "loss": 0.4881, + "step": 17211 + }, + { + "epoch": 0.40347638554358084, + "grad_norm": 3.35002064704895, + "learning_rate": 0.00012990241567097888, + "loss": 1.3655, + "step": 17212 + }, + { + "epoch": 0.40349982711838583, + "grad_norm": 3.143043041229248, + "learning_rate": 0.0001298953864665899, + "loss": 0.3599, + "step": 17213 + }, + { + "epoch": 0.40352326869319083, + "grad_norm": 3.9981367588043213, + "learning_rate": 0.00012988835709998772, + "loss": 1.9178, + "step": 17214 + }, + { + "epoch": 0.4035467102679958, + "grad_norm": 3.3754794597625732, + "learning_rate": 0.00012988132757121048, + "loss": 0.4428, + "step": 17215 + }, + { + "epoch": 0.4035701518428008, + "grad_norm": 0.6890917420387268, + "learning_rate": 0.0001298742978802963, + "loss": 0.0588, + "step": 17216 + }, + { + "epoch": 0.4035935934176058, + "grad_norm": 3.1151270866394043, + "learning_rate": 0.00012986726802728332, + "loss": 1.7969, + "step": 17217 + }, + { + "epoch": 0.4036170349924108, + "grad_norm": 1.4109601974487305, + "learning_rate": 0.00012986023801220972, + "loss": 0.1802, + "step": 17218 + }, + { + "epoch": 0.4036404765672158, + "grad_norm": 3.3275697231292725, + "learning_rate": 0.00012985320783511365, + "loss": 0.3519, + "step": 17219 + }, + { + "epoch": 0.4036639181420208, + "grad_norm": 3.409358501434326, + "learning_rate": 0.0001298461774960332, + "loss": 0.3129, + "step": 17220 + }, + { + "epoch": 0.4036873597168258, + "grad_norm": 1.5961204767227173, + "learning_rate": 0.0001298391469950066, + "loss": 0.1859, + "step": 17221 + }, + { + "epoch": 0.4037108012916308, + "grad_norm": 1.8557463884353638, + "learning_rate": 0.00012983211633207193, + "loss": 0.1348, + "step": 17222 + }, + { + "epoch": 0.4037342428664358, + "grad_norm": 2.087430477142334, + "learning_rate": 0.00012982508550726736, + "loss": 1.2967, + "step": 17223 + }, + { + "epoch": 0.40375768444124077, + "grad_norm": 2.431410789489746, + "learning_rate": 0.000129818054520631, + "loss": 0.3157, + "step": 17224 + }, + { + "epoch": 0.40378112601604577, + "grad_norm": 4.282179832458496, + "learning_rate": 0.0001298110233722011, + "loss": 0.3952, + "step": 17225 + }, + { + "epoch": 0.40380456759085076, + "grad_norm": 2.5367844104766846, + "learning_rate": 0.0001298039920620157, + "loss": 0.274, + "step": 17226 + }, + { + "epoch": 0.40382800916565575, + "grad_norm": 2.371350049972534, + "learning_rate": 0.000129796960590113, + "loss": 0.1767, + "step": 17227 + }, + { + "epoch": 0.40385145074046075, + "grad_norm": 0.9416778087615967, + "learning_rate": 0.0001297899289565312, + "loss": 0.0521, + "step": 17228 + }, + { + "epoch": 0.40387489231526574, + "grad_norm": 4.004603862762451, + "learning_rate": 0.00012978289716130832, + "loss": 0.2304, + "step": 17229 + }, + { + "epoch": 0.40389833389007074, + "grad_norm": 0.910325288772583, + "learning_rate": 0.00012977586520448268, + "loss": 0.0494, + "step": 17230 + }, + { + "epoch": 0.40392177546487573, + "grad_norm": 5.3831353187561035, + "learning_rate": 0.00012976883308609235, + "loss": 0.5217, + "step": 17231 + }, + { + "epoch": 0.4039452170396807, + "grad_norm": 3.240896224975586, + "learning_rate": 0.0001297618008061755, + "loss": 0.2549, + "step": 17232 + }, + { + "epoch": 0.4039686586144857, + "grad_norm": 2.627250909805298, + "learning_rate": 0.00012975476836477024, + "loss": 0.3268, + "step": 17233 + }, + { + "epoch": 0.4039921001892907, + "grad_norm": 2.7267515659332275, + "learning_rate": 0.00012974773576191482, + "loss": 0.2396, + "step": 17234 + }, + { + "epoch": 0.4040155417640957, + "grad_norm": 4.115359783172607, + "learning_rate": 0.0001297407029976473, + "loss": 0.4281, + "step": 17235 + }, + { + "epoch": 0.4040389833389007, + "grad_norm": 6.997797012329102, + "learning_rate": 0.0001297336700720059, + "loss": 0.9384, + "step": 17236 + }, + { + "epoch": 0.4040624249137057, + "grad_norm": 1.6416739225387573, + "learning_rate": 0.00012972663698502878, + "loss": 0.7631, + "step": 17237 + }, + { + "epoch": 0.4040858664885107, + "grad_norm": 1.3858063220977783, + "learning_rate": 0.00012971960373675412, + "loss": 0.0952, + "step": 17238 + }, + { + "epoch": 0.4041093080633157, + "grad_norm": 2.5999486446380615, + "learning_rate": 0.00012971257032722003, + "loss": 1.1466, + "step": 17239 + }, + { + "epoch": 0.4041327496381207, + "grad_norm": 2.3967695236206055, + "learning_rate": 0.00012970553675646472, + "loss": 0.3853, + "step": 17240 + }, + { + "epoch": 0.4041561912129257, + "grad_norm": 2.6124937534332275, + "learning_rate": 0.00012969850302452625, + "loss": 0.3924, + "step": 17241 + }, + { + "epoch": 0.40417963278773067, + "grad_norm": 2.6523218154907227, + "learning_rate": 0.00012969146913144298, + "loss": 1.6295, + "step": 17242 + }, + { + "epoch": 0.40420307436253566, + "grad_norm": 1.329727053642273, + "learning_rate": 0.0001296844350772529, + "loss": 0.167, + "step": 17243 + }, + { + "epoch": 0.40422651593734066, + "grad_norm": 3.6997454166412354, + "learning_rate": 0.00012967740086199426, + "loss": 0.4763, + "step": 17244 + }, + { + "epoch": 0.40424995751214565, + "grad_norm": 1.8637685775756836, + "learning_rate": 0.0001296703664857052, + "loss": 0.0782, + "step": 17245 + }, + { + "epoch": 0.40427339908695065, + "grad_norm": 3.5037317276000977, + "learning_rate": 0.00012966333194842388, + "loss": 0.2383, + "step": 17246 + }, + { + "epoch": 0.40429684066175564, + "grad_norm": 2.823281764984131, + "learning_rate": 0.00012965629725018851, + "loss": 0.2519, + "step": 17247 + }, + { + "epoch": 0.40432028223656064, + "grad_norm": 1.733809471130371, + "learning_rate": 0.00012964926239103722, + "loss": 0.1775, + "step": 17248 + }, + { + "epoch": 0.40434372381136563, + "grad_norm": 3.555079460144043, + "learning_rate": 0.0001296422273710082, + "loss": 0.4304, + "step": 17249 + }, + { + "epoch": 0.4043671653861706, + "grad_norm": 3.1305277347564697, + "learning_rate": 0.00012963519219013963, + "loss": 0.2427, + "step": 17250 + }, + { + "epoch": 0.4043906069609756, + "grad_norm": 2.8687775135040283, + "learning_rate": 0.00012962815684846972, + "loss": 0.2061, + "step": 17251 + }, + { + "epoch": 0.4044140485357806, + "grad_norm": 3.257185459136963, + "learning_rate": 0.00012962112134603653, + "loss": 0.372, + "step": 17252 + }, + { + "epoch": 0.4044374901105856, + "grad_norm": 4.638855934143066, + "learning_rate": 0.0001296140856828783, + "loss": 0.4521, + "step": 17253 + }, + { + "epoch": 0.4044609316853906, + "grad_norm": 2.9788944721221924, + "learning_rate": 0.00012960704985903325, + "loss": 0.3078, + "step": 17254 + }, + { + "epoch": 0.4044843732601956, + "grad_norm": 4.843672275543213, + "learning_rate": 0.00012960001387453952, + "loss": 0.5214, + "step": 17255 + }, + { + "epoch": 0.4045078148350006, + "grad_norm": 3.3444764614105225, + "learning_rate": 0.00012959297772943524, + "loss": 0.3834, + "step": 17256 + }, + { + "epoch": 0.4045312564098056, + "grad_norm": 3.262686252593994, + "learning_rate": 0.00012958594142375865, + "loss": 0.3777, + "step": 17257 + }, + { + "epoch": 0.4045546979846106, + "grad_norm": 3.623664379119873, + "learning_rate": 0.00012957890495754793, + "loss": 0.4902, + "step": 17258 + }, + { + "epoch": 0.40457813955941563, + "grad_norm": 0.9989569187164307, + "learning_rate": 0.00012957186833084121, + "loss": 0.1105, + "step": 17259 + }, + { + "epoch": 0.4046015811342206, + "grad_norm": 1.3162928819656372, + "learning_rate": 0.00012956483154367672, + "loss": 0.0374, + "step": 17260 + }, + { + "epoch": 0.4046250227090256, + "grad_norm": 3.356684923171997, + "learning_rate": 0.00012955779459609262, + "loss": 0.3951, + "step": 17261 + }, + { + "epoch": 0.4046484642838306, + "grad_norm": 5.569375038146973, + "learning_rate": 0.00012955075748812708, + "loss": 0.4862, + "step": 17262 + }, + { + "epoch": 0.4046719058586356, + "grad_norm": 3.570319890975952, + "learning_rate": 0.00012954372021981833, + "loss": 0.3887, + "step": 17263 + }, + { + "epoch": 0.4046953474334406, + "grad_norm": 1.582352876663208, + "learning_rate": 0.0001295366827912045, + "loss": 0.1634, + "step": 17264 + }, + { + "epoch": 0.4047187890082456, + "grad_norm": 1.262502670288086, + "learning_rate": 0.00012952964520232378, + "loss": 0.0874, + "step": 17265 + }, + { + "epoch": 0.4047422305830506, + "grad_norm": 3.933194875717163, + "learning_rate": 0.0001295226074532144, + "loss": 0.4942, + "step": 17266 + }, + { + "epoch": 0.4047656721578556, + "grad_norm": 3.5819709300994873, + "learning_rate": 0.00012951556954391452, + "loss": 0.3586, + "step": 17267 + }, + { + "epoch": 0.4047891137326606, + "grad_norm": 1.5266512632369995, + "learning_rate": 0.00012950853147446233, + "loss": 0.2163, + "step": 17268 + }, + { + "epoch": 0.40481255530746557, + "grad_norm": 0.957810640335083, + "learning_rate": 0.000129501493244896, + "loss": 0.1162, + "step": 17269 + }, + { + "epoch": 0.40483599688227057, + "grad_norm": 3.7700066566467285, + "learning_rate": 0.0001294944548552538, + "loss": 0.4315, + "step": 17270 + }, + { + "epoch": 0.40485943845707556, + "grad_norm": 1.7954000234603882, + "learning_rate": 0.0001294874163055738, + "loss": 1.5386, + "step": 17271 + }, + { + "epoch": 0.40488288003188055, + "grad_norm": 5.643893718719482, + "learning_rate": 0.00012948037759589424, + "loss": 0.8027, + "step": 17272 + }, + { + "epoch": 0.40490632160668555, + "grad_norm": 2.6714096069335938, + "learning_rate": 0.00012947333872625336, + "loss": 0.4277, + "step": 17273 + }, + { + "epoch": 0.40492976318149054, + "grad_norm": 5.432973861694336, + "learning_rate": 0.0001294662996966893, + "loss": 0.1196, + "step": 17274 + }, + { + "epoch": 0.40495320475629554, + "grad_norm": 0.7205430865287781, + "learning_rate": 0.00012945926050724027, + "loss": 0.0825, + "step": 17275 + }, + { + "epoch": 0.40497664633110053, + "grad_norm": 3.362877130508423, + "learning_rate": 0.00012945222115794448, + "loss": 0.3137, + "step": 17276 + }, + { + "epoch": 0.4050000879059055, + "grad_norm": 4.147922992706299, + "learning_rate": 0.0001294451816488401, + "loss": 0.2975, + "step": 17277 + }, + { + "epoch": 0.4050235294807105, + "grad_norm": 2.521336317062378, + "learning_rate": 0.0001294381419799653, + "loss": 0.1961, + "step": 17278 + }, + { + "epoch": 0.4050469710555155, + "grad_norm": 1.9722018241882324, + "learning_rate": 0.00012943110215135836, + "loss": 0.2899, + "step": 17279 + }, + { + "epoch": 0.4050704126303205, + "grad_norm": 2.54258131980896, + "learning_rate": 0.00012942406216305742, + "loss": 0.2499, + "step": 17280 + }, + { + "epoch": 0.4050938542051255, + "grad_norm": 3.5556817054748535, + "learning_rate": 0.00012941702201510068, + "loss": 0.2834, + "step": 17281 + }, + { + "epoch": 0.4051172957799305, + "grad_norm": 4.128847122192383, + "learning_rate": 0.0001294099817075264, + "loss": 0.3712, + "step": 17282 + }, + { + "epoch": 0.4051407373547355, + "grad_norm": 4.552219390869141, + "learning_rate": 0.00012940294124037268, + "loss": 0.6245, + "step": 17283 + }, + { + "epoch": 0.4051641789295405, + "grad_norm": 5.485067367553711, + "learning_rate": 0.00012939590061367774, + "loss": 0.7271, + "step": 17284 + }, + { + "epoch": 0.4051876205043455, + "grad_norm": 0.7153711318969727, + "learning_rate": 0.00012938885982747992, + "loss": 0.1087, + "step": 17285 + }, + { + "epoch": 0.4052110620791505, + "grad_norm": 3.7052700519561768, + "learning_rate": 0.00012938181888181723, + "loss": 0.3367, + "step": 17286 + }, + { + "epoch": 0.40523450365395547, + "grad_norm": 5.7487616539001465, + "learning_rate": 0.000129374777776728, + "loss": 0.3173, + "step": 17287 + }, + { + "epoch": 0.40525794522876046, + "grad_norm": 3.586113691329956, + "learning_rate": 0.0001293677365122504, + "loss": 0.4853, + "step": 17288 + }, + { + "epoch": 0.40528138680356546, + "grad_norm": 2.8524553775787354, + "learning_rate": 0.00012936069508842266, + "loss": 0.3889, + "step": 17289 + }, + { + "epoch": 0.40530482837837045, + "grad_norm": 3.8420045375823975, + "learning_rate": 0.00012935365350528292, + "loss": 0.3657, + "step": 17290 + }, + { + "epoch": 0.40532826995317545, + "grad_norm": 1.878206491470337, + "learning_rate": 0.00012934661176286944, + "loss": 0.1066, + "step": 17291 + }, + { + "epoch": 0.40535171152798044, + "grad_norm": 3.8909261226654053, + "learning_rate": 0.00012933956986122046, + "loss": 0.3244, + "step": 17292 + }, + { + "epoch": 0.40537515310278543, + "grad_norm": 2.061495542526245, + "learning_rate": 0.00012933252780037409, + "loss": 0.1632, + "step": 17293 + }, + { + "epoch": 0.40539859467759043, + "grad_norm": 1.9905346632003784, + "learning_rate": 0.00012932548558036862, + "loss": 0.2318, + "step": 17294 + }, + { + "epoch": 0.4054220362523954, + "grad_norm": 3.9043591022491455, + "learning_rate": 0.00012931844320124226, + "loss": 0.3203, + "step": 17295 + }, + { + "epoch": 0.4054454778272004, + "grad_norm": 3.675386905670166, + "learning_rate": 0.00012931140066303317, + "loss": 0.3375, + "step": 17296 + }, + { + "epoch": 0.4054689194020054, + "grad_norm": 1.9328311681747437, + "learning_rate": 0.0001293043579657796, + "loss": 0.1605, + "step": 17297 + }, + { + "epoch": 0.4054923609768104, + "grad_norm": 2.3490865230560303, + "learning_rate": 0.0001292973151095198, + "loss": 0.1624, + "step": 17298 + }, + { + "epoch": 0.4055158025516154, + "grad_norm": 3.37068247795105, + "learning_rate": 0.00012929027209429195, + "loss": 0.4127, + "step": 17299 + }, + { + "epoch": 0.4055392441264204, + "grad_norm": 1.717941403388977, + "learning_rate": 0.0001292832289201342, + "loss": 0.1184, + "step": 17300 + }, + { + "epoch": 0.4055626857012254, + "grad_norm": 0.9299479126930237, + "learning_rate": 0.00012927618558708484, + "loss": 0.0563, + "step": 17301 + }, + { + "epoch": 0.4055861272760304, + "grad_norm": 2.305220603942871, + "learning_rate": 0.0001292691420951821, + "loss": 0.2061, + "step": 17302 + }, + { + "epoch": 0.4056095688508354, + "grad_norm": 2.1136960983276367, + "learning_rate": 0.00012926209844446417, + "loss": 0.2411, + "step": 17303 + }, + { + "epoch": 0.4056330104256404, + "grad_norm": 3.0348174571990967, + "learning_rate": 0.00012925505463496926, + "loss": 0.2274, + "step": 17304 + }, + { + "epoch": 0.40565645200044537, + "grad_norm": 2.4984147548675537, + "learning_rate": 0.0001292480106667356, + "loss": 1.2646, + "step": 17305 + }, + { + "epoch": 0.40567989357525036, + "grad_norm": 4.1928791999816895, + "learning_rate": 0.00012924096653980143, + "loss": 0.3917, + "step": 17306 + }, + { + "epoch": 0.40570333515005536, + "grad_norm": 1.0229980945587158, + "learning_rate": 0.00012923392225420494, + "loss": 0.0995, + "step": 17307 + }, + { + "epoch": 0.40572677672486035, + "grad_norm": 1.854828953742981, + "learning_rate": 0.00012922687780998434, + "loss": 0.5555, + "step": 17308 + }, + { + "epoch": 0.40575021829966534, + "grad_norm": 0.9000149965286255, + "learning_rate": 0.00012921983320717792, + "loss": 0.122, + "step": 17309 + }, + { + "epoch": 0.4057736598744704, + "grad_norm": 2.9400739669799805, + "learning_rate": 0.00012921278844582382, + "loss": 0.3367, + "step": 17310 + }, + { + "epoch": 0.4057971014492754, + "grad_norm": 0.9738684296607971, + "learning_rate": 0.00012920574352596036, + "loss": 0.118, + "step": 17311 + }, + { + "epoch": 0.4058205430240804, + "grad_norm": 0.23694273829460144, + "learning_rate": 0.00012919869844762568, + "loss": 0.0166, + "step": 17312 + }, + { + "epoch": 0.4058439845988854, + "grad_norm": 5.23251485824585, + "learning_rate": 0.00012919165321085806, + "loss": 0.5368, + "step": 17313 + }, + { + "epoch": 0.40586742617369037, + "grad_norm": 0.543575644493103, + "learning_rate": 0.0001291846078156957, + "loss": 0.0435, + "step": 17314 + }, + { + "epoch": 0.40589086774849537, + "grad_norm": 3.209169387817383, + "learning_rate": 0.00012917756226217683, + "loss": 0.3417, + "step": 17315 + }, + { + "epoch": 0.40591430932330036, + "grad_norm": 0.32941165566444397, + "learning_rate": 0.0001291705165503397, + "loss": 0.0278, + "step": 17316 + }, + { + "epoch": 0.40593775089810535, + "grad_norm": 3.579810380935669, + "learning_rate": 0.0001291634706802225, + "loss": 0.3114, + "step": 17317 + }, + { + "epoch": 0.40596119247291035, + "grad_norm": 2.4227514266967773, + "learning_rate": 0.00012915642465186354, + "loss": 0.3397, + "step": 17318 + }, + { + "epoch": 0.40598463404771534, + "grad_norm": 3.129288911819458, + "learning_rate": 0.00012914937846530096, + "loss": 0.3952, + "step": 17319 + }, + { + "epoch": 0.40600807562252034, + "grad_norm": 3.0517892837524414, + "learning_rate": 0.00012914233212057302, + "loss": 0.3845, + "step": 17320 + }, + { + "epoch": 0.40603151719732533, + "grad_norm": 0.6253045201301575, + "learning_rate": 0.000129135285617718, + "loss": 0.0514, + "step": 17321 + }, + { + "epoch": 0.4060549587721303, + "grad_norm": 4.39675760269165, + "learning_rate": 0.0001291282389567741, + "loss": 0.5295, + "step": 17322 + }, + { + "epoch": 0.4060784003469353, + "grad_norm": 2.722792148590088, + "learning_rate": 0.00012912119213777956, + "loss": 0.2435, + "step": 17323 + }, + { + "epoch": 0.4061018419217403, + "grad_norm": 2.304645538330078, + "learning_rate": 0.00012911414516077261, + "loss": 1.9001, + "step": 17324 + }, + { + "epoch": 0.4061252834965453, + "grad_norm": 3.5966994762420654, + "learning_rate": 0.00012910709802579147, + "loss": 0.4273, + "step": 17325 + }, + { + "epoch": 0.4061487250713503, + "grad_norm": 4.6131367683410645, + "learning_rate": 0.00012910005073287438, + "loss": 0.3875, + "step": 17326 + }, + { + "epoch": 0.4061721666461553, + "grad_norm": 4.512929916381836, + "learning_rate": 0.00012909300328205965, + "loss": 0.4375, + "step": 17327 + }, + { + "epoch": 0.4061956082209603, + "grad_norm": 1.0111737251281738, + "learning_rate": 0.00012908595567338543, + "loss": 0.067, + "step": 17328 + }, + { + "epoch": 0.4062190497957653, + "grad_norm": 2.8663218021392822, + "learning_rate": 0.00012907890790689, + "loss": 0.3194, + "step": 17329 + }, + { + "epoch": 0.4062424913705703, + "grad_norm": 2.476280927658081, + "learning_rate": 0.0001290718599826116, + "loss": 0.2964, + "step": 17330 + }, + { + "epoch": 0.4062659329453753, + "grad_norm": 2.881855010986328, + "learning_rate": 0.00012906481190058848, + "loss": 0.3128, + "step": 17331 + }, + { + "epoch": 0.40628937452018027, + "grad_norm": 3.956855058670044, + "learning_rate": 0.00012905776366085885, + "loss": 0.5047, + "step": 17332 + }, + { + "epoch": 0.40631281609498526, + "grad_norm": 1.23832106590271, + "learning_rate": 0.00012905071526346103, + "loss": 0.0617, + "step": 17333 + }, + { + "epoch": 0.40633625766979026, + "grad_norm": 2.7356278896331787, + "learning_rate": 0.00012904366670843316, + "loss": 0.2894, + "step": 17334 + }, + { + "epoch": 0.40635969924459525, + "grad_norm": 1.3856438398361206, + "learning_rate": 0.00012903661799581356, + "loss": 0.148, + "step": 17335 + }, + { + "epoch": 0.40638314081940025, + "grad_norm": 4.553050518035889, + "learning_rate": 0.00012902956912564046, + "loss": 0.4236, + "step": 17336 + }, + { + "epoch": 0.40640658239420524, + "grad_norm": 3.7484662532806396, + "learning_rate": 0.00012902252009795208, + "loss": 0.3438, + "step": 17337 + }, + { + "epoch": 0.40643002396901023, + "grad_norm": 1.649093508720398, + "learning_rate": 0.00012901547091278667, + "loss": 0.1464, + "step": 17338 + }, + { + "epoch": 0.40645346554381523, + "grad_norm": 3.044645071029663, + "learning_rate": 0.00012900842157018255, + "loss": 0.2428, + "step": 17339 + }, + { + "epoch": 0.4064769071186202, + "grad_norm": 6.957577705383301, + "learning_rate": 0.0001290013720701779, + "loss": 0.779, + "step": 17340 + }, + { + "epoch": 0.4065003486934252, + "grad_norm": 2.1888530254364014, + "learning_rate": 0.00012899432241281097, + "loss": 0.1474, + "step": 17341 + }, + { + "epoch": 0.4065237902682302, + "grad_norm": 0.8117272853851318, + "learning_rate": 0.00012898727259812008, + "loss": 0.0756, + "step": 17342 + }, + { + "epoch": 0.4065472318430352, + "grad_norm": 3.7963707447052, + "learning_rate": 0.0001289802226261434, + "loss": 0.4154, + "step": 17343 + }, + { + "epoch": 0.4065706734178402, + "grad_norm": 0.9012610912322998, + "learning_rate": 0.0001289731724969192, + "loss": 0.0591, + "step": 17344 + }, + { + "epoch": 0.4065941149926452, + "grad_norm": 1.5373564958572388, + "learning_rate": 0.0001289661222104858, + "loss": 0.2001, + "step": 17345 + }, + { + "epoch": 0.4066175565674502, + "grad_norm": 1.1521456241607666, + "learning_rate": 0.00012895907176688136, + "loss": 0.1678, + "step": 17346 + }, + { + "epoch": 0.4066409981422552, + "grad_norm": 5.549492835998535, + "learning_rate": 0.00012895202116614422, + "loss": 0.2043, + "step": 17347 + }, + { + "epoch": 0.4066644397170602, + "grad_norm": 2.140223264694214, + "learning_rate": 0.00012894497040831258, + "loss": 0.9942, + "step": 17348 + }, + { + "epoch": 0.40668788129186517, + "grad_norm": 1.3432482481002808, + "learning_rate": 0.0001289379194934247, + "loss": 0.0816, + "step": 17349 + }, + { + "epoch": 0.40671132286667017, + "grad_norm": 4.5286664962768555, + "learning_rate": 0.00012893086842151886, + "loss": 0.2339, + "step": 17350 + }, + { + "epoch": 0.40673476444147516, + "grad_norm": 4.275692462921143, + "learning_rate": 0.00012892381719263333, + "loss": 0.5359, + "step": 17351 + }, + { + "epoch": 0.40675820601628015, + "grad_norm": 2.791795492172241, + "learning_rate": 0.00012891676580680637, + "loss": 0.3322, + "step": 17352 + }, + { + "epoch": 0.40678164759108515, + "grad_norm": 3.9568727016448975, + "learning_rate": 0.0001289097142640762, + "loss": 0.7083, + "step": 17353 + }, + { + "epoch": 0.40680508916589014, + "grad_norm": 2.031919240951538, + "learning_rate": 0.0001289026625644811, + "loss": 0.1511, + "step": 17354 + }, + { + "epoch": 0.40682853074069514, + "grad_norm": 0.7112885117530823, + "learning_rate": 0.00012889561070805936, + "loss": 0.0675, + "step": 17355 + }, + { + "epoch": 0.40685197231550013, + "grad_norm": 3.0723650455474854, + "learning_rate": 0.00012888855869484922, + "loss": 0.2874, + "step": 17356 + }, + { + "epoch": 0.4068754138903051, + "grad_norm": 7.10524320602417, + "learning_rate": 0.00012888150652488894, + "loss": 0.511, + "step": 17357 + }, + { + "epoch": 0.4068988554651101, + "grad_norm": 0.8111805319786072, + "learning_rate": 0.0001288744541982168, + "loss": 0.1117, + "step": 17358 + }, + { + "epoch": 0.4069222970399151, + "grad_norm": 2.8998167514801025, + "learning_rate": 0.0001288674017148711, + "loss": 1.5974, + "step": 17359 + }, + { + "epoch": 0.4069457386147201, + "grad_norm": 3.313129186630249, + "learning_rate": 0.00012886034907489, + "loss": 0.3831, + "step": 17360 + }, + { + "epoch": 0.40696918018952516, + "grad_norm": 2.0115268230438232, + "learning_rate": 0.00012885329627831187, + "loss": 0.2283, + "step": 17361 + }, + { + "epoch": 0.40699262176433015, + "grad_norm": 2.304202079772949, + "learning_rate": 0.00012884624332517493, + "loss": 0.1386, + "step": 17362 + }, + { + "epoch": 0.40701606333913515, + "grad_norm": 2.8561928272247314, + "learning_rate": 0.00012883919021551744, + "loss": 0.3343, + "step": 17363 + }, + { + "epoch": 0.40703950491394014, + "grad_norm": 4.389403820037842, + "learning_rate": 0.00012883213694937773, + "loss": 0.283, + "step": 17364 + }, + { + "epoch": 0.40706294648874514, + "grad_norm": 1.2928245067596436, + "learning_rate": 0.00012882508352679404, + "loss": 0.1543, + "step": 17365 + }, + { + "epoch": 0.40708638806355013, + "grad_norm": 5.6518235206604, + "learning_rate": 0.0001288180299478046, + "loss": 0.6236, + "step": 17366 + }, + { + "epoch": 0.4071098296383551, + "grad_norm": 1.3153254985809326, + "learning_rate": 0.00012881097621244775, + "loss": 0.1577, + "step": 17367 + }, + { + "epoch": 0.4071332712131601, + "grad_norm": 3.0688374042510986, + "learning_rate": 0.00012880392232076174, + "loss": 0.3714, + "step": 17368 + }, + { + "epoch": 0.4071567127879651, + "grad_norm": 0.9227372407913208, + "learning_rate": 0.00012879686827278478, + "loss": 0.0868, + "step": 17369 + }, + { + "epoch": 0.4071801543627701, + "grad_norm": 0.7667846083641052, + "learning_rate": 0.00012878981406855527, + "loss": 0.0828, + "step": 17370 + }, + { + "epoch": 0.4072035959375751, + "grad_norm": 2.6050331592559814, + "learning_rate": 0.0001287827597081114, + "loss": 0.2277, + "step": 17371 + }, + { + "epoch": 0.4072270375123801, + "grad_norm": 2.4388463497161865, + "learning_rate": 0.00012877570519149144, + "loss": 1.0446, + "step": 17372 + }, + { + "epoch": 0.4072504790871851, + "grad_norm": 1.6983741521835327, + "learning_rate": 0.0001287686505187337, + "loss": 0.7117, + "step": 17373 + }, + { + "epoch": 0.4072739206619901, + "grad_norm": 3.389044761657715, + "learning_rate": 0.0001287615956898765, + "loss": 0.3431, + "step": 17374 + }, + { + "epoch": 0.4072973622367951, + "grad_norm": 3.4866278171539307, + "learning_rate": 0.00012875454070495802, + "loss": 0.4056, + "step": 17375 + }, + { + "epoch": 0.4073208038116001, + "grad_norm": 4.028932094573975, + "learning_rate": 0.00012874748556401659, + "loss": 0.4494, + "step": 17376 + }, + { + "epoch": 0.40734424538640507, + "grad_norm": 2.6871979236602783, + "learning_rate": 0.00012874043026709053, + "loss": 0.2011, + "step": 17377 + }, + { + "epoch": 0.40736768696121006, + "grad_norm": 5.052716255187988, + "learning_rate": 0.0001287333748142181, + "loss": 0.4641, + "step": 17378 + }, + { + "epoch": 0.40739112853601506, + "grad_norm": 5.860998630523682, + "learning_rate": 0.00012872631920543754, + "loss": 0.5628, + "step": 17379 + }, + { + "epoch": 0.40741457011082005, + "grad_norm": 2.099048137664795, + "learning_rate": 0.00012871926344078718, + "loss": 1.2438, + "step": 17380 + }, + { + "epoch": 0.40743801168562505, + "grad_norm": 0.582752525806427, + "learning_rate": 0.00012871220752030526, + "loss": 0.0395, + "step": 17381 + }, + { + "epoch": 0.40746145326043004, + "grad_norm": 4.562426567077637, + "learning_rate": 0.00012870515144403012, + "loss": 0.5248, + "step": 17382 + }, + { + "epoch": 0.40748489483523503, + "grad_norm": 4.753899574279785, + "learning_rate": 0.000128698095212, + "loss": 0.302, + "step": 17383 + }, + { + "epoch": 0.40750833641004003, + "grad_norm": 4.556036472320557, + "learning_rate": 0.00012869103882425325, + "loss": 0.4682, + "step": 17384 + }, + { + "epoch": 0.407531777984845, + "grad_norm": 2.1801111698150635, + "learning_rate": 0.0001286839822808281, + "loss": 1.2726, + "step": 17385 + }, + { + "epoch": 0.40755521955965, + "grad_norm": 0.36065196990966797, + "learning_rate": 0.00012867692558176283, + "loss": 0.0156, + "step": 17386 + }, + { + "epoch": 0.407578661134455, + "grad_norm": 3.7629168033599854, + "learning_rate": 0.00012866986872709578, + "loss": 0.4806, + "step": 17387 + }, + { + "epoch": 0.40760210270926, + "grad_norm": 4.145245552062988, + "learning_rate": 0.0001286628117168652, + "loss": 0.5398, + "step": 17388 + }, + { + "epoch": 0.407625544284065, + "grad_norm": 1.764477252960205, + "learning_rate": 0.00012865575455110937, + "loss": 0.3074, + "step": 17389 + }, + { + "epoch": 0.40764898585887, + "grad_norm": 5.173034191131592, + "learning_rate": 0.00012864869722986668, + "loss": 0.6878, + "step": 17390 + }, + { + "epoch": 0.407672427433675, + "grad_norm": 2.626500368118286, + "learning_rate": 0.00012864163975317533, + "loss": 1.8701, + "step": 17391 + }, + { + "epoch": 0.40769586900848, + "grad_norm": 2.907191038131714, + "learning_rate": 0.00012863458212107362, + "loss": 0.2857, + "step": 17392 + }, + { + "epoch": 0.407719310583285, + "grad_norm": 1.327497124671936, + "learning_rate": 0.00012862752433359984, + "loss": 0.124, + "step": 17393 + }, + { + "epoch": 0.40774275215808997, + "grad_norm": 2.214522361755371, + "learning_rate": 0.00012862046639079236, + "loss": 0.2735, + "step": 17394 + }, + { + "epoch": 0.40776619373289497, + "grad_norm": 4.818826675415039, + "learning_rate": 0.00012861340829268937, + "loss": 0.4212, + "step": 17395 + }, + { + "epoch": 0.40778963530769996, + "grad_norm": 0.8982930183410645, + "learning_rate": 0.00012860635003932928, + "loss": 0.0619, + "step": 17396 + }, + { + "epoch": 0.40781307688250495, + "grad_norm": 1.214494228363037, + "learning_rate": 0.0001285992916307503, + "loss": 0.1326, + "step": 17397 + }, + { + "epoch": 0.40783651845730995, + "grad_norm": 1.1265596151351929, + "learning_rate": 0.00012859223306699075, + "loss": 0.1452, + "step": 17398 + }, + { + "epoch": 0.40785996003211494, + "grad_norm": 2.474792957305908, + "learning_rate": 0.00012858517434808893, + "loss": 0.2571, + "step": 17399 + }, + { + "epoch": 0.40788340160691994, + "grad_norm": 5.1085944175720215, + "learning_rate": 0.00012857811547408322, + "loss": 0.4133, + "step": 17400 + }, + { + "epoch": 0.40790684318172493, + "grad_norm": 3.273257255554199, + "learning_rate": 0.0001285710564450118, + "loss": 0.3778, + "step": 17401 + }, + { + "epoch": 0.4079302847565299, + "grad_norm": 4.646651744842529, + "learning_rate": 0.000128563997260913, + "loss": 0.662, + "step": 17402 + }, + { + "epoch": 0.4079537263313349, + "grad_norm": 1.5777102708816528, + "learning_rate": 0.0001285569379218252, + "loss": 0.1511, + "step": 17403 + }, + { + "epoch": 0.4079771679061399, + "grad_norm": 3.532222032546997, + "learning_rate": 0.00012854987842778664, + "loss": 0.3438, + "step": 17404 + }, + { + "epoch": 0.4080006094809449, + "grad_norm": 3.5945863723754883, + "learning_rate": 0.0001285428187788356, + "loss": 0.4988, + "step": 17405 + }, + { + "epoch": 0.4080240510557499, + "grad_norm": 3.3913073539733887, + "learning_rate": 0.00012853575897501043, + "loss": 0.2655, + "step": 17406 + }, + { + "epoch": 0.4080474926305549, + "grad_norm": 4.293335914611816, + "learning_rate": 0.00012852869901634948, + "loss": 0.4168, + "step": 17407 + }, + { + "epoch": 0.4080709342053599, + "grad_norm": 1.8689416646957397, + "learning_rate": 0.00012852163890289097, + "loss": 0.1161, + "step": 17408 + }, + { + "epoch": 0.4080943757801649, + "grad_norm": 3.9525094032287598, + "learning_rate": 0.00012851457863467325, + "loss": 0.4751, + "step": 17409 + }, + { + "epoch": 0.4081178173549699, + "grad_norm": 3.5974924564361572, + "learning_rate": 0.00012850751821173463, + "loss": 0.3212, + "step": 17410 + }, + { + "epoch": 0.4081412589297749, + "grad_norm": 0.624507486820221, + "learning_rate": 0.0001285004576341134, + "loss": 0.0528, + "step": 17411 + }, + { + "epoch": 0.4081647005045799, + "grad_norm": 2.4009690284729004, + "learning_rate": 0.0001284933969018479, + "loss": 0.2908, + "step": 17412 + }, + { + "epoch": 0.4081881420793849, + "grad_norm": 2.589538097381592, + "learning_rate": 0.00012848633601497643, + "loss": 1.5627, + "step": 17413 + }, + { + "epoch": 0.4082115836541899, + "grad_norm": 2.57448148727417, + "learning_rate": 0.00012847927497353727, + "loss": 0.3807, + "step": 17414 + }, + { + "epoch": 0.4082350252289949, + "grad_norm": 2.4746487140655518, + "learning_rate": 0.0001284722137775688, + "loss": 0.263, + "step": 17415 + }, + { + "epoch": 0.4082584668037999, + "grad_norm": 2.299774408340454, + "learning_rate": 0.0001284651524271093, + "loss": 0.1292, + "step": 17416 + }, + { + "epoch": 0.4082819083786049, + "grad_norm": 5.748205661773682, + "learning_rate": 0.00012845809092219704, + "loss": 0.6696, + "step": 17417 + }, + { + "epoch": 0.4083053499534099, + "grad_norm": 1.6206531524658203, + "learning_rate": 0.00012845102926287041, + "loss": 0.1495, + "step": 17418 + }, + { + "epoch": 0.4083287915282149, + "grad_norm": 0.8606131076812744, + "learning_rate": 0.00012844396744916766, + "loss": 0.0845, + "step": 17419 + }, + { + "epoch": 0.4083522331030199, + "grad_norm": 0.7890664935112, + "learning_rate": 0.0001284369054811272, + "loss": 0.0892, + "step": 17420 + }, + { + "epoch": 0.4083756746778249, + "grad_norm": 0.7874000072479248, + "learning_rate": 0.00012842984335878724, + "loss": 0.0764, + "step": 17421 + }, + { + "epoch": 0.40839911625262987, + "grad_norm": 3.465148687362671, + "learning_rate": 0.0001284227810821862, + "loss": 0.2923, + "step": 17422 + }, + { + "epoch": 0.40842255782743486, + "grad_norm": 2.698545217514038, + "learning_rate": 0.0001284157186513623, + "loss": 0.322, + "step": 17423 + }, + { + "epoch": 0.40844599940223986, + "grad_norm": 3.0230495929718018, + "learning_rate": 0.00012840865606635393, + "loss": 0.2386, + "step": 17424 + }, + { + "epoch": 0.40846944097704485, + "grad_norm": 3.310310125350952, + "learning_rate": 0.0001284015933271994, + "loss": 0.2022, + "step": 17425 + }, + { + "epoch": 0.40849288255184985, + "grad_norm": 1.5211305618286133, + "learning_rate": 0.00012839453043393705, + "loss": 0.1791, + "step": 17426 + }, + { + "epoch": 0.40851632412665484, + "grad_norm": 0.5600969791412354, + "learning_rate": 0.00012838746738660513, + "loss": 0.0696, + "step": 17427 + }, + { + "epoch": 0.40853976570145983, + "grad_norm": 0.35182273387908936, + "learning_rate": 0.00012838040418524207, + "loss": 0.0197, + "step": 17428 + }, + { + "epoch": 0.40856320727626483, + "grad_norm": 2.0027718544006348, + "learning_rate": 0.00012837334082988608, + "loss": 0.1929, + "step": 17429 + }, + { + "epoch": 0.4085866488510698, + "grad_norm": 10.087785720825195, + "learning_rate": 0.00012836627732057556, + "loss": 0.1242, + "step": 17430 + }, + { + "epoch": 0.4086100904258748, + "grad_norm": 4.374292850494385, + "learning_rate": 0.00012835921365734885, + "loss": 0.5446, + "step": 17431 + }, + { + "epoch": 0.4086335320006798, + "grad_norm": 0.331863671541214, + "learning_rate": 0.00012835214984024425, + "loss": 0.0423, + "step": 17432 + }, + { + "epoch": 0.4086569735754848, + "grad_norm": 3.1843860149383545, + "learning_rate": 0.00012834508586930007, + "loss": 1.7638, + "step": 17433 + }, + { + "epoch": 0.4086804151502898, + "grad_norm": 4.036632061004639, + "learning_rate": 0.00012833802174455465, + "loss": 0.2115, + "step": 17434 + }, + { + "epoch": 0.4087038567250948, + "grad_norm": 4.970232963562012, + "learning_rate": 0.00012833095746604632, + "loss": 0.6463, + "step": 17435 + }, + { + "epoch": 0.4087272982998998, + "grad_norm": 0.6033052206039429, + "learning_rate": 0.00012832389303381344, + "loss": 0.0747, + "step": 17436 + }, + { + "epoch": 0.4087507398747048, + "grad_norm": 3.8714332580566406, + "learning_rate": 0.0001283168284478943, + "loss": 0.2391, + "step": 17437 + }, + { + "epoch": 0.4087741814495098, + "grad_norm": 3.531231641769409, + "learning_rate": 0.00012830976370832727, + "loss": 0.2714, + "step": 17438 + }, + { + "epoch": 0.40879762302431477, + "grad_norm": 4.862822532653809, + "learning_rate": 0.0001283026988151507, + "loss": 0.6123, + "step": 17439 + }, + { + "epoch": 0.40882106459911977, + "grad_norm": 2.674520969390869, + "learning_rate": 0.00012829563376840283, + "loss": 0.2202, + "step": 17440 + }, + { + "epoch": 0.40884450617392476, + "grad_norm": 3.81190824508667, + "learning_rate": 0.00012828856856812208, + "loss": 0.3499, + "step": 17441 + }, + { + "epoch": 0.40886794774872975, + "grad_norm": 3.1256768703460693, + "learning_rate": 0.00012828150321434677, + "loss": 0.3232, + "step": 17442 + }, + { + "epoch": 0.40889138932353475, + "grad_norm": 3.26802396774292, + "learning_rate": 0.00012827443770711518, + "loss": 0.3773, + "step": 17443 + }, + { + "epoch": 0.40891483089833974, + "grad_norm": 2.5461244583129883, + "learning_rate": 0.00012826737204646576, + "loss": 0.31, + "step": 17444 + }, + { + "epoch": 0.40893827247314474, + "grad_norm": 0.36885538697242737, + "learning_rate": 0.00012826030623243676, + "loss": 0.0256, + "step": 17445 + }, + { + "epoch": 0.40896171404794973, + "grad_norm": 3.268544912338257, + "learning_rate": 0.00012825324026506654, + "loss": 0.1995, + "step": 17446 + }, + { + "epoch": 0.4089851556227547, + "grad_norm": 0.926665723323822, + "learning_rate": 0.00012824617414439345, + "loss": 0.1214, + "step": 17447 + }, + { + "epoch": 0.4090085971975597, + "grad_norm": 4.650047302246094, + "learning_rate": 0.0001282391078704558, + "loss": 0.4434, + "step": 17448 + }, + { + "epoch": 0.4090320387723647, + "grad_norm": 4.112199783325195, + "learning_rate": 0.00012823204144329198, + "loss": 0.4105, + "step": 17449 + }, + { + "epoch": 0.4090554803471697, + "grad_norm": 5.9830002784729, + "learning_rate": 0.00012822497486294032, + "loss": 0.4632, + "step": 17450 + }, + { + "epoch": 0.4090789219219747, + "grad_norm": 4.489085674285889, + "learning_rate": 0.00012821790812943916, + "loss": 0.4959, + "step": 17451 + }, + { + "epoch": 0.4091023634967797, + "grad_norm": 3.3038110733032227, + "learning_rate": 0.00012821084124282678, + "loss": 1.9608, + "step": 17452 + }, + { + "epoch": 0.4091258050715847, + "grad_norm": 4.240731239318848, + "learning_rate": 0.00012820377420314164, + "loss": 0.4042, + "step": 17453 + }, + { + "epoch": 0.4091492466463897, + "grad_norm": 4.986702919006348, + "learning_rate": 0.00012819670701042199, + "loss": 0.5008, + "step": 17454 + }, + { + "epoch": 0.4091726882211947, + "grad_norm": 2.7715296745300293, + "learning_rate": 0.00012818963966470626, + "loss": 0.2802, + "step": 17455 + }, + { + "epoch": 0.4091961297959997, + "grad_norm": 4.123091697692871, + "learning_rate": 0.00012818257216603272, + "loss": 0.5073, + "step": 17456 + }, + { + "epoch": 0.40921957137080467, + "grad_norm": 3.4024081230163574, + "learning_rate": 0.00012817550451443973, + "loss": 0.3416, + "step": 17457 + }, + { + "epoch": 0.40924301294560966, + "grad_norm": 3.055800199508667, + "learning_rate": 0.00012816843670996573, + "loss": 0.2763, + "step": 17458 + }, + { + "epoch": 0.40926645452041466, + "grad_norm": 1.3055509328842163, + "learning_rate": 0.00012816136875264892, + "loss": 0.1206, + "step": 17459 + }, + { + "epoch": 0.40928989609521965, + "grad_norm": 3.2019686698913574, + "learning_rate": 0.00012815430064252775, + "loss": 0.1004, + "step": 17460 + }, + { + "epoch": 0.40931333767002465, + "grad_norm": 5.092715263366699, + "learning_rate": 0.0001281472323796406, + "loss": 0.4127, + "step": 17461 + }, + { + "epoch": 0.40933677924482964, + "grad_norm": 4.3541154861450195, + "learning_rate": 0.00012814016396402574, + "loss": 0.2505, + "step": 17462 + }, + { + "epoch": 0.4093602208196347, + "grad_norm": 3.610330820083618, + "learning_rate": 0.00012813309539572157, + "loss": 0.4522, + "step": 17463 + }, + { + "epoch": 0.4093836623944397, + "grad_norm": 2.6986610889434814, + "learning_rate": 0.00012812602667476643, + "loss": 0.1602, + "step": 17464 + }, + { + "epoch": 0.4094071039692447, + "grad_norm": 2.656783103942871, + "learning_rate": 0.00012811895780119866, + "loss": 0.1546, + "step": 17465 + }, + { + "epoch": 0.4094305455440497, + "grad_norm": 2.982454776763916, + "learning_rate": 0.00012811188877505665, + "loss": 0.1707, + "step": 17466 + }, + { + "epoch": 0.40945398711885467, + "grad_norm": 3.599860906600952, + "learning_rate": 0.00012810481959637875, + "loss": 0.4776, + "step": 17467 + }, + { + "epoch": 0.40947742869365966, + "grad_norm": 2.430258274078369, + "learning_rate": 0.0001280977502652033, + "loss": 0.156, + "step": 17468 + }, + { + "epoch": 0.40950087026846466, + "grad_norm": 5.132901668548584, + "learning_rate": 0.00012809068078156868, + "loss": 0.6376, + "step": 17469 + }, + { + "epoch": 0.40952431184326965, + "grad_norm": 1.2211134433746338, + "learning_rate": 0.0001280836111455132, + "loss": 0.1216, + "step": 17470 + }, + { + "epoch": 0.40954775341807464, + "grad_norm": 1.0796796083450317, + "learning_rate": 0.0001280765413570753, + "loss": 0.1572, + "step": 17471 + }, + { + "epoch": 0.40957119499287964, + "grad_norm": 3.5686895847320557, + "learning_rate": 0.00012806947141629326, + "loss": 0.3054, + "step": 17472 + }, + { + "epoch": 0.40959463656768463, + "grad_norm": 8.056095123291016, + "learning_rate": 0.0001280624013232055, + "loss": 0.3431, + "step": 17473 + }, + { + "epoch": 0.40961807814248963, + "grad_norm": 1.0936533212661743, + "learning_rate": 0.00012805533107785038, + "loss": 0.0988, + "step": 17474 + }, + { + "epoch": 0.4096415197172946, + "grad_norm": 2.4039394855499268, + "learning_rate": 0.0001280482606802662, + "loss": 0.2548, + "step": 17475 + }, + { + "epoch": 0.4096649612920996, + "grad_norm": 2.8759090900421143, + "learning_rate": 0.00012804119013049142, + "loss": 0.258, + "step": 17476 + }, + { + "epoch": 0.4096884028669046, + "grad_norm": 6.553431034088135, + "learning_rate": 0.0001280341194285643, + "loss": 0.6249, + "step": 17477 + }, + { + "epoch": 0.4097118444417096, + "grad_norm": 1.9597605466842651, + "learning_rate": 0.00012802704857452327, + "loss": 0.1269, + "step": 17478 + }, + { + "epoch": 0.4097352860165146, + "grad_norm": 1.502606749534607, + "learning_rate": 0.00012801997756840673, + "loss": 0.1097, + "step": 17479 + }, + { + "epoch": 0.4097587275913196, + "grad_norm": 2.630845308303833, + "learning_rate": 0.00012801290641025297, + "loss": 0.161, + "step": 17480 + }, + { + "epoch": 0.4097821691661246, + "grad_norm": 1.6444445848464966, + "learning_rate": 0.0001280058351001004, + "loss": 0.228, + "step": 17481 + }, + { + "epoch": 0.4098056107409296, + "grad_norm": 1.6642614603042603, + "learning_rate": 0.00012799876363798736, + "loss": 0.2353, + "step": 17482 + }, + { + "epoch": 0.4098290523157346, + "grad_norm": 3.0720226764678955, + "learning_rate": 0.0001279916920239523, + "loss": 0.2495, + "step": 17483 + }, + { + "epoch": 0.40985249389053957, + "grad_norm": 2.439776659011841, + "learning_rate": 0.00012798462025803348, + "loss": 0.253, + "step": 17484 + }, + { + "epoch": 0.40987593546534457, + "grad_norm": 6.51364278793335, + "learning_rate": 0.0001279775483402693, + "loss": 0.8444, + "step": 17485 + }, + { + "epoch": 0.40989937704014956, + "grad_norm": 3.780522108078003, + "learning_rate": 0.00012797047627069824, + "loss": 0.4544, + "step": 17486 + }, + { + "epoch": 0.40992281861495455, + "grad_norm": 2.5488951206207275, + "learning_rate": 0.00012796340404935851, + "loss": 0.2113, + "step": 17487 + }, + { + "epoch": 0.40994626018975955, + "grad_norm": 2.5273382663726807, + "learning_rate": 0.0001279563316762886, + "loss": 0.1521, + "step": 17488 + }, + { + "epoch": 0.40996970176456454, + "grad_norm": 2.5778968334198, + "learning_rate": 0.00012794925915152688, + "loss": 0.2575, + "step": 17489 + }, + { + "epoch": 0.40999314333936954, + "grad_norm": 3.54785418510437, + "learning_rate": 0.0001279421864751116, + "loss": 0.4084, + "step": 17490 + }, + { + "epoch": 0.41001658491417453, + "grad_norm": 3.515580177307129, + "learning_rate": 0.0001279351136470813, + "loss": 0.3776, + "step": 17491 + }, + { + "epoch": 0.4100400264889795, + "grad_norm": 0.5728626251220703, + "learning_rate": 0.00012792804066747427, + "loss": 0.0633, + "step": 17492 + }, + { + "epoch": 0.4100634680637845, + "grad_norm": 3.3085594177246094, + "learning_rate": 0.00012792096753632895, + "loss": 0.3391, + "step": 17493 + }, + { + "epoch": 0.4100869096385895, + "grad_norm": 2.10251784324646, + "learning_rate": 0.00012791389425368362, + "loss": 1.5986, + "step": 17494 + }, + { + "epoch": 0.4101103512133945, + "grad_norm": 3.108186721801758, + "learning_rate": 0.00012790682081957676, + "loss": 0.2019, + "step": 17495 + }, + { + "epoch": 0.4101337927881995, + "grad_norm": 0.7146525382995605, + "learning_rate": 0.00012789974723404664, + "loss": 0.0802, + "step": 17496 + }, + { + "epoch": 0.4101572343630045, + "grad_norm": 3.8420279026031494, + "learning_rate": 0.00012789267349713174, + "loss": 0.5815, + "step": 17497 + }, + { + "epoch": 0.4101806759378095, + "grad_norm": 1.5059224367141724, + "learning_rate": 0.00012788559960887042, + "loss": 0.1348, + "step": 17498 + }, + { + "epoch": 0.4102041175126145, + "grad_norm": 4.754528045654297, + "learning_rate": 0.00012787852556930105, + "loss": 1.8964, + "step": 17499 + }, + { + "epoch": 0.4102275590874195, + "grad_norm": 3.0670676231384277, + "learning_rate": 0.000127871451378462, + "loss": 0.2113, + "step": 17500 + }, + { + "epoch": 0.4102510006622245, + "grad_norm": 3.4401185512542725, + "learning_rate": 0.00012786437703639169, + "loss": 0.3231, + "step": 17501 + }, + { + "epoch": 0.41027444223702947, + "grad_norm": 0.6472384333610535, + "learning_rate": 0.00012785730254312845, + "loss": 0.0391, + "step": 17502 + }, + { + "epoch": 0.41029788381183446, + "grad_norm": 2.9238922595977783, + "learning_rate": 0.00012785022789871077, + "loss": 0.2799, + "step": 17503 + }, + { + "epoch": 0.41032132538663946, + "grad_norm": 0.5392768979072571, + "learning_rate": 0.0001278431531031769, + "loss": 0.0748, + "step": 17504 + }, + { + "epoch": 0.41034476696144445, + "grad_norm": 3.257567882537842, + "learning_rate": 0.00012783607815656535, + "loss": 0.4256, + "step": 17505 + }, + { + "epoch": 0.41036820853624945, + "grad_norm": 0.7807869911193848, + "learning_rate": 0.00012782900305891444, + "loss": 0.0772, + "step": 17506 + }, + { + "epoch": 0.41039165011105444, + "grad_norm": 1.318638801574707, + "learning_rate": 0.0001278219278102626, + "loss": 0.148, + "step": 17507 + }, + { + "epoch": 0.41041509168585943, + "grad_norm": 6.320192813873291, + "learning_rate": 0.00012781485241064814, + "loss": 0.3959, + "step": 17508 + }, + { + "epoch": 0.41043853326066443, + "grad_norm": 2.824514150619507, + "learning_rate": 0.00012780777686010957, + "loss": 0.3066, + "step": 17509 + }, + { + "epoch": 0.4104619748354694, + "grad_norm": 2.174880266189575, + "learning_rate": 0.00012780070115868517, + "loss": 0.1374, + "step": 17510 + }, + { + "epoch": 0.4104854164102744, + "grad_norm": 1.9206866025924683, + "learning_rate": 0.0001277936253064134, + "loss": 0.213, + "step": 17511 + }, + { + "epoch": 0.4105088579850794, + "grad_norm": 7.788984775543213, + "learning_rate": 0.00012778654930333268, + "loss": 0.4507, + "step": 17512 + }, + { + "epoch": 0.4105322995598844, + "grad_norm": 1.139527440071106, + "learning_rate": 0.0001277794731494813, + "loss": 0.1084, + "step": 17513 + }, + { + "epoch": 0.41055574113468946, + "grad_norm": 1.0136500597000122, + "learning_rate": 0.00012777239684489775, + "loss": 0.1223, + "step": 17514 + }, + { + "epoch": 0.41057918270949445, + "grad_norm": 0.918675422668457, + "learning_rate": 0.00012776532038962039, + "loss": 0.1138, + "step": 17515 + }, + { + "epoch": 0.41060262428429944, + "grad_norm": 1.6145315170288086, + "learning_rate": 0.0001277582437836876, + "loss": 0.0827, + "step": 17516 + }, + { + "epoch": 0.41062606585910444, + "grad_norm": 2.0150985717773438, + "learning_rate": 0.00012775116702713784, + "loss": 1.6171, + "step": 17517 + }, + { + "epoch": 0.41064950743390943, + "grad_norm": 2.8930771350860596, + "learning_rate": 0.00012774409012000944, + "loss": 0.2564, + "step": 17518 + }, + { + "epoch": 0.4106729490087144, + "grad_norm": 0.8914780616760254, + "learning_rate": 0.00012773701306234081, + "loss": 0.1229, + "step": 17519 + }, + { + "epoch": 0.4106963905835194, + "grad_norm": 2.554102659225464, + "learning_rate": 0.0001277299358541704, + "loss": 0.3299, + "step": 17520 + }, + { + "epoch": 0.4107198321583244, + "grad_norm": 2.3925893306732178, + "learning_rate": 0.00012772285849553658, + "loss": 0.3131, + "step": 17521 + }, + { + "epoch": 0.4107432737331294, + "grad_norm": 4.3639397621154785, + "learning_rate": 0.00012771578098647775, + "loss": 0.4378, + "step": 17522 + }, + { + "epoch": 0.4107667153079344, + "grad_norm": 0.8765259385108948, + "learning_rate": 0.00012770870332703228, + "loss": 0.0464, + "step": 17523 + }, + { + "epoch": 0.4107901568827394, + "grad_norm": 3.259477376937866, + "learning_rate": 0.00012770162551723865, + "loss": 0.3264, + "step": 17524 + }, + { + "epoch": 0.4108135984575444, + "grad_norm": 3.450439691543579, + "learning_rate": 0.00012769454755713516, + "loss": 0.4132, + "step": 17525 + }, + { + "epoch": 0.4108370400323494, + "grad_norm": 3.7792415618896484, + "learning_rate": 0.00012768746944676033, + "loss": 0.2651, + "step": 17526 + }, + { + "epoch": 0.4108604816071544, + "grad_norm": 2.0360569953918457, + "learning_rate": 0.0001276803911861525, + "loss": 0.2393, + "step": 17527 + }, + { + "epoch": 0.4108839231819594, + "grad_norm": 3.860280990600586, + "learning_rate": 0.00012767331277535008, + "loss": 0.3169, + "step": 17528 + }, + { + "epoch": 0.41090736475676437, + "grad_norm": 4.570867538452148, + "learning_rate": 0.0001276662342143915, + "loss": 0.4083, + "step": 17529 + }, + { + "epoch": 0.41093080633156936, + "grad_norm": 1.4335263967514038, + "learning_rate": 0.00012765915550331516, + "loss": 0.1198, + "step": 17530 + }, + { + "epoch": 0.41095424790637436, + "grad_norm": 13.175980567932129, + "learning_rate": 0.00012765207664215946, + "loss": 0.5223, + "step": 17531 + }, + { + "epoch": 0.41097768948117935, + "grad_norm": 2.445969581604004, + "learning_rate": 0.0001276449976309628, + "loss": 0.2439, + "step": 17532 + }, + { + "epoch": 0.41100113105598435, + "grad_norm": 0.8961443901062012, + "learning_rate": 0.0001276379184697636, + "loss": 0.0605, + "step": 17533 + }, + { + "epoch": 0.41102457263078934, + "grad_norm": 4.478020668029785, + "learning_rate": 0.0001276308391586003, + "loss": 0.461, + "step": 17534 + }, + { + "epoch": 0.41104801420559434, + "grad_norm": 2.8741931915283203, + "learning_rate": 0.00012762375969751125, + "loss": 0.2477, + "step": 17535 + }, + { + "epoch": 0.41107145578039933, + "grad_norm": 3.5485498905181885, + "learning_rate": 0.00012761668008653496, + "loss": 0.4564, + "step": 17536 + }, + { + "epoch": 0.4110948973552043, + "grad_norm": 5.238780975341797, + "learning_rate": 0.00012760960032570976, + "loss": 0.3981, + "step": 17537 + }, + { + "epoch": 0.4111183389300093, + "grad_norm": 1.046796441078186, + "learning_rate": 0.0001276025204150741, + "loss": 0.074, + "step": 17538 + }, + { + "epoch": 0.4111417805048143, + "grad_norm": 2.0832338333129883, + "learning_rate": 0.00012759544035466637, + "loss": 0.1863, + "step": 17539 + }, + { + "epoch": 0.4111652220796193, + "grad_norm": 1.3578529357910156, + "learning_rate": 0.00012758836014452503, + "loss": 0.1521, + "step": 17540 + }, + { + "epoch": 0.4111886636544243, + "grad_norm": 3.0237789154052734, + "learning_rate": 0.00012758127978468846, + "loss": 0.2818, + "step": 17541 + }, + { + "epoch": 0.4112121052292293, + "grad_norm": 1.0141838788986206, + "learning_rate": 0.00012757419927519507, + "loss": 0.1512, + "step": 17542 + }, + { + "epoch": 0.4112355468040343, + "grad_norm": 2.314577341079712, + "learning_rate": 0.00012756711861608335, + "loss": 0.2873, + "step": 17543 + }, + { + "epoch": 0.4112589883788393, + "grad_norm": 2.6346943378448486, + "learning_rate": 0.0001275600378073916, + "loss": 0.2036, + "step": 17544 + }, + { + "epoch": 0.4112824299536443, + "grad_norm": 2.776962995529175, + "learning_rate": 0.00012755295684915833, + "loss": 0.2326, + "step": 17545 + }, + { + "epoch": 0.4113058715284493, + "grad_norm": 2.0003862380981445, + "learning_rate": 0.000127545875741422, + "loss": 1.4748, + "step": 17546 + }, + { + "epoch": 0.41132931310325427, + "grad_norm": 4.7273850440979, + "learning_rate": 0.00012753879448422094, + "loss": 0.7852, + "step": 17547 + }, + { + "epoch": 0.41135275467805926, + "grad_norm": 3.3788201808929443, + "learning_rate": 0.00012753171307759356, + "loss": 0.4459, + "step": 17548 + }, + { + "epoch": 0.41137619625286426, + "grad_norm": 3.5579164028167725, + "learning_rate": 0.0001275246315215784, + "loss": 0.2697, + "step": 17549 + }, + { + "epoch": 0.41139963782766925, + "grad_norm": 2.714270830154419, + "learning_rate": 0.00012751754981621378, + "loss": 0.3123, + "step": 17550 + }, + { + "epoch": 0.41142307940247425, + "grad_norm": 1.6962382793426514, + "learning_rate": 0.00012751046796153815, + "loss": 0.1743, + "step": 17551 + }, + { + "epoch": 0.41144652097727924, + "grad_norm": 3.855978488922119, + "learning_rate": 0.00012750338595758997, + "loss": 0.1808, + "step": 17552 + }, + { + "epoch": 0.41146996255208423, + "grad_norm": 3.4021694660186768, + "learning_rate": 0.00012749630380440766, + "loss": 0.2855, + "step": 17553 + }, + { + "epoch": 0.41149340412688923, + "grad_norm": 2.329826593399048, + "learning_rate": 0.00012748922150202958, + "loss": 0.2355, + "step": 17554 + }, + { + "epoch": 0.4115168457016942, + "grad_norm": 4.450825214385986, + "learning_rate": 0.00012748213905049423, + "loss": 0.4888, + "step": 17555 + }, + { + "epoch": 0.4115402872764992, + "grad_norm": 1.5529600381851196, + "learning_rate": 0.00012747505644984005, + "loss": 0.2005, + "step": 17556 + }, + { + "epoch": 0.4115637288513042, + "grad_norm": 6.2701311111450195, + "learning_rate": 0.00012746797370010544, + "loss": 0.9529, + "step": 17557 + }, + { + "epoch": 0.4115871704261092, + "grad_norm": 1.2042572498321533, + "learning_rate": 0.00012746089080132878, + "loss": 0.1363, + "step": 17558 + }, + { + "epoch": 0.4116106120009142, + "grad_norm": 0.8064252734184265, + "learning_rate": 0.00012745380775354862, + "loss": 0.0622, + "step": 17559 + }, + { + "epoch": 0.4116340535757192, + "grad_norm": 0.5178510546684265, + "learning_rate": 0.00012744672455680328, + "loss": 0.0394, + "step": 17560 + }, + { + "epoch": 0.4116574951505242, + "grad_norm": 2.116603374481201, + "learning_rate": 0.00012743964121113125, + "loss": 0.2399, + "step": 17561 + }, + { + "epoch": 0.4116809367253292, + "grad_norm": 3.0124716758728027, + "learning_rate": 0.000127432557716571, + "loss": 0.1712, + "step": 17562 + }, + { + "epoch": 0.4117043783001342, + "grad_norm": 1.5797226428985596, + "learning_rate": 0.00012742547407316085, + "loss": 0.1041, + "step": 17563 + }, + { + "epoch": 0.41172781987493917, + "grad_norm": 0.8413165211677551, + "learning_rate": 0.00012741839028093934, + "loss": 0.0933, + "step": 17564 + }, + { + "epoch": 0.4117512614497442, + "grad_norm": 3.3446521759033203, + "learning_rate": 0.00012741130633994488, + "loss": 0.2878, + "step": 17565 + }, + { + "epoch": 0.4117747030245492, + "grad_norm": 2.4119439125061035, + "learning_rate": 0.00012740422225021591, + "loss": 0.3057, + "step": 17566 + }, + { + "epoch": 0.4117981445993542, + "grad_norm": 3.2187001705169678, + "learning_rate": 0.00012739713801179084, + "loss": 0.282, + "step": 17567 + }, + { + "epoch": 0.4118215861741592, + "grad_norm": 2.1881706714630127, + "learning_rate": 0.00012739005362470812, + "loss": 0.2212, + "step": 17568 + }, + { + "epoch": 0.4118450277489642, + "grad_norm": 2.722350835800171, + "learning_rate": 0.00012738296908900622, + "loss": 1.5417, + "step": 17569 + }, + { + "epoch": 0.4118684693237692, + "grad_norm": 2.317405939102173, + "learning_rate": 0.00012737588440472355, + "loss": 0.2619, + "step": 17570 + }, + { + "epoch": 0.4118919108985742, + "grad_norm": 2.616075038909912, + "learning_rate": 0.00012736879957189856, + "loss": 1.3908, + "step": 17571 + }, + { + "epoch": 0.4119153524733792, + "grad_norm": 4.088318347930908, + "learning_rate": 0.0001273617145905697, + "loss": 0.4052, + "step": 17572 + }, + { + "epoch": 0.4119387940481842, + "grad_norm": 2.637594699859619, + "learning_rate": 0.0001273546294607754, + "loss": 0.2542, + "step": 17573 + }, + { + "epoch": 0.41196223562298917, + "grad_norm": 2.1053903102874756, + "learning_rate": 0.00012734754418255412, + "loss": 0.1155, + "step": 17574 + }, + { + "epoch": 0.41198567719779416, + "grad_norm": 3.5101845264434814, + "learning_rate": 0.0001273404587559443, + "loss": 0.2291, + "step": 17575 + }, + { + "epoch": 0.41200911877259916, + "grad_norm": 3.118084192276001, + "learning_rate": 0.00012733337318098436, + "loss": 0.2497, + "step": 17576 + }, + { + "epoch": 0.41203256034740415, + "grad_norm": 4.550370693206787, + "learning_rate": 0.0001273262874577128, + "loss": 0.3333, + "step": 17577 + }, + { + "epoch": 0.41205600192220915, + "grad_norm": 4.022693157196045, + "learning_rate": 0.000127319201586168, + "loss": 0.2975, + "step": 17578 + }, + { + "epoch": 0.41207944349701414, + "grad_norm": 5.9290995597839355, + "learning_rate": 0.00012731211556638847, + "loss": 0.5348, + "step": 17579 + }, + { + "epoch": 0.41210288507181914, + "grad_norm": 4.510249137878418, + "learning_rate": 0.00012730502939841263, + "loss": 0.4052, + "step": 17580 + }, + { + "epoch": 0.41212632664662413, + "grad_norm": 2.2722089290618896, + "learning_rate": 0.00012729794308227895, + "loss": 0.0956, + "step": 17581 + }, + { + "epoch": 0.4121497682214291, + "grad_norm": 2.811290740966797, + "learning_rate": 0.00012729085661802582, + "loss": 0.297, + "step": 17582 + }, + { + "epoch": 0.4121732097962341, + "grad_norm": 3.5733041763305664, + "learning_rate": 0.00012728377000569173, + "loss": 0.4134, + "step": 17583 + }, + { + "epoch": 0.4121966513710391, + "grad_norm": 0.9258620738983154, + "learning_rate": 0.00012727668324531518, + "loss": 0.0754, + "step": 17584 + }, + { + "epoch": 0.4122200929458441, + "grad_norm": 4.3866496086120605, + "learning_rate": 0.0001272695963369346, + "loss": 0.6139, + "step": 17585 + }, + { + "epoch": 0.4122435345206491, + "grad_norm": 3.85821795463562, + "learning_rate": 0.00012726250928058835, + "loss": 0.3601, + "step": 17586 + }, + { + "epoch": 0.4122669760954541, + "grad_norm": 3.131983518600464, + "learning_rate": 0.000127255422076315, + "loss": 0.2482, + "step": 17587 + }, + { + "epoch": 0.4122904176702591, + "grad_norm": 0.7370786070823669, + "learning_rate": 0.00012724833472415296, + "loss": 0.0589, + "step": 17588 + }, + { + "epoch": 0.4123138592450641, + "grad_norm": 2.9581737518310547, + "learning_rate": 0.00012724124722414068, + "loss": 1.5098, + "step": 17589 + }, + { + "epoch": 0.4123373008198691, + "grad_norm": 2.7998859882354736, + "learning_rate": 0.00012723415957631665, + "loss": 1.4701, + "step": 17590 + }, + { + "epoch": 0.4123607423946741, + "grad_norm": 4.0487799644470215, + "learning_rate": 0.0001272270717807193, + "loss": 0.3295, + "step": 17591 + }, + { + "epoch": 0.41238418396947907, + "grad_norm": 4.37554407119751, + "learning_rate": 0.00012721998383738708, + "loss": 0.269, + "step": 17592 + }, + { + "epoch": 0.41240762554428406, + "grad_norm": 0.6263686418533325, + "learning_rate": 0.00012721289574635844, + "loss": 0.0448, + "step": 17593 + }, + { + "epoch": 0.41243106711908906, + "grad_norm": 2.8658268451690674, + "learning_rate": 0.0001272058075076719, + "loss": 1.7805, + "step": 17594 + }, + { + "epoch": 0.41245450869389405, + "grad_norm": 2.746628761291504, + "learning_rate": 0.00012719871912136586, + "loss": 0.328, + "step": 17595 + }, + { + "epoch": 0.41247795026869905, + "grad_norm": 3.775813341140747, + "learning_rate": 0.00012719163058747882, + "loss": 0.4723, + "step": 17596 + }, + { + "epoch": 0.41250139184350404, + "grad_norm": 2.5517258644104004, + "learning_rate": 0.00012718454190604922, + "loss": 0.1233, + "step": 17597 + }, + { + "epoch": 0.41252483341830903, + "grad_norm": 2.3191027641296387, + "learning_rate": 0.00012717745307711553, + "loss": 1.5098, + "step": 17598 + }, + { + "epoch": 0.41254827499311403, + "grad_norm": 2.971006393432617, + "learning_rate": 0.00012717036410071623, + "loss": 0.3345, + "step": 17599 + }, + { + "epoch": 0.412571716567919, + "grad_norm": 3.7539165019989014, + "learning_rate": 0.00012716327497688974, + "loss": 0.2821, + "step": 17600 + }, + { + "epoch": 0.412595158142724, + "grad_norm": 0.9659140706062317, + "learning_rate": 0.0001271561857056746, + "loss": 0.0602, + "step": 17601 + }, + { + "epoch": 0.412618599717529, + "grad_norm": 8.330589294433594, + "learning_rate": 0.0001271490962871092, + "loss": 0.4296, + "step": 17602 + }, + { + "epoch": 0.412642041292334, + "grad_norm": 1.8525110483169556, + "learning_rate": 0.00012714200672123206, + "loss": 0.1709, + "step": 17603 + }, + { + "epoch": 0.412665482867139, + "grad_norm": 3.4587912559509277, + "learning_rate": 0.00012713491700808163, + "loss": 0.2878, + "step": 17604 + }, + { + "epoch": 0.412688924441944, + "grad_norm": 3.8847341537475586, + "learning_rate": 0.00012712782714769636, + "loss": 0.4164, + "step": 17605 + }, + { + "epoch": 0.412712366016749, + "grad_norm": 1.3123698234558105, + "learning_rate": 0.00012712073714011473, + "loss": 0.1789, + "step": 17606 + }, + { + "epoch": 0.412735807591554, + "grad_norm": 1.5915288925170898, + "learning_rate": 0.00012711364698537523, + "loss": 0.0834, + "step": 17607 + }, + { + "epoch": 0.412759249166359, + "grad_norm": 1.0884487628936768, + "learning_rate": 0.00012710655668351633, + "loss": 0.0887, + "step": 17608 + }, + { + "epoch": 0.41278269074116397, + "grad_norm": 3.5401244163513184, + "learning_rate": 0.00012709946623457648, + "loss": 0.2586, + "step": 17609 + }, + { + "epoch": 0.41280613231596897, + "grad_norm": 3.484501838684082, + "learning_rate": 0.0001270923756385942, + "loss": 0.3, + "step": 17610 + }, + { + "epoch": 0.41282957389077396, + "grad_norm": 0.429009348154068, + "learning_rate": 0.00012708528489560784, + "loss": 0.056, + "step": 17611 + }, + { + "epoch": 0.41285301546557895, + "grad_norm": 2.6571762561798096, + "learning_rate": 0.00012707819400565603, + "loss": 0.1938, + "step": 17612 + }, + { + "epoch": 0.41287645704038395, + "grad_norm": 5.402055740356445, + "learning_rate": 0.00012707110296877714, + "loss": 0.6272, + "step": 17613 + }, + { + "epoch": 0.41289989861518894, + "grad_norm": 2.4617044925689697, + "learning_rate": 0.00012706401178500974, + "loss": 0.3128, + "step": 17614 + }, + { + "epoch": 0.41292334018999394, + "grad_norm": 5.7712531089782715, + "learning_rate": 0.0001270569204543922, + "loss": 0.5736, + "step": 17615 + }, + { + "epoch": 0.412946781764799, + "grad_norm": 2.6062028408050537, + "learning_rate": 0.0001270498289769631, + "loss": 0.245, + "step": 17616 + }, + { + "epoch": 0.412970223339604, + "grad_norm": 3.6046502590179443, + "learning_rate": 0.0001270427373527608, + "loss": 0.4023, + "step": 17617 + }, + { + "epoch": 0.412993664914409, + "grad_norm": 3.7416915893554688, + "learning_rate": 0.00012703564558182387, + "loss": 0.3036, + "step": 17618 + }, + { + "epoch": 0.41301710648921397, + "grad_norm": 5.491835117340088, + "learning_rate": 0.0001270285536641908, + "loss": 0.6804, + "step": 17619 + }, + { + "epoch": 0.41304054806401896, + "grad_norm": 1.493691086769104, + "learning_rate": 0.00012702146159990002, + "loss": 0.1684, + "step": 17620 + }, + { + "epoch": 0.41306398963882396, + "grad_norm": 0.7731430530548096, + "learning_rate": 0.00012701436938899002, + "loss": 0.0591, + "step": 17621 + }, + { + "epoch": 0.41308743121362895, + "grad_norm": 3.1176555156707764, + "learning_rate": 0.0001270072770314993, + "loss": 0.4476, + "step": 17622 + }, + { + "epoch": 0.41311087278843395, + "grad_norm": 2.5619075298309326, + "learning_rate": 0.00012700018452746633, + "loss": 1.4909, + "step": 17623 + }, + { + "epoch": 0.41313431436323894, + "grad_norm": 2.89111065864563, + "learning_rate": 0.0001269930918769296, + "loss": 0.1148, + "step": 17624 + }, + { + "epoch": 0.41315775593804394, + "grad_norm": 4.773801803588867, + "learning_rate": 0.0001269859990799276, + "loss": 0.5114, + "step": 17625 + }, + { + "epoch": 0.41318119751284893, + "grad_norm": 0.5494440793991089, + "learning_rate": 0.00012697890613649882, + "loss": 0.0356, + "step": 17626 + }, + { + "epoch": 0.4132046390876539, + "grad_norm": 2.020651340484619, + "learning_rate": 0.0001269718130466817, + "loss": 0.2333, + "step": 17627 + }, + { + "epoch": 0.4132280806624589, + "grad_norm": 5.561739444732666, + "learning_rate": 0.00012696471981051482, + "loss": 0.3758, + "step": 17628 + }, + { + "epoch": 0.4132515222372639, + "grad_norm": 2.507397174835205, + "learning_rate": 0.00012695762642803656, + "loss": 0.2202, + "step": 17629 + }, + { + "epoch": 0.4132749638120689, + "grad_norm": 1.5341291427612305, + "learning_rate": 0.00012695053289928549, + "loss": 0.0941, + "step": 17630 + }, + { + "epoch": 0.4132984053868739, + "grad_norm": 0.9619807600975037, + "learning_rate": 0.00012694343922430005, + "loss": 0.0806, + "step": 17631 + }, + { + "epoch": 0.4133218469616789, + "grad_norm": 2.8745594024658203, + "learning_rate": 0.00012693634540311875, + "loss": 0.2826, + "step": 17632 + }, + { + "epoch": 0.4133452885364839, + "grad_norm": 1.2454794645309448, + "learning_rate": 0.00012692925143578015, + "loss": 0.1265, + "step": 17633 + }, + { + "epoch": 0.4133687301112889, + "grad_norm": 2.146641969680786, + "learning_rate": 0.00012692215732232258, + "loss": 0.2244, + "step": 17634 + }, + { + "epoch": 0.4133921716860939, + "grad_norm": 1.956679344177246, + "learning_rate": 0.0001269150630627847, + "loss": 0.1833, + "step": 17635 + }, + { + "epoch": 0.4134156132608989, + "grad_norm": 8.85722541809082, + "learning_rate": 0.00012690796865720487, + "loss": 0.9277, + "step": 17636 + }, + { + "epoch": 0.41343905483570387, + "grad_norm": 2.984419345855713, + "learning_rate": 0.00012690087410562165, + "loss": 0.3749, + "step": 17637 + }, + { + "epoch": 0.41346249641050886, + "grad_norm": 1.0120099782943726, + "learning_rate": 0.00012689377940807356, + "loss": 0.1, + "step": 17638 + }, + { + "epoch": 0.41348593798531386, + "grad_norm": 5.623795986175537, + "learning_rate": 0.0001268866845645991, + "loss": 0.5639, + "step": 17639 + }, + { + "epoch": 0.41350937956011885, + "grad_norm": 3.4125914573669434, + "learning_rate": 0.00012687958957523666, + "loss": 0.2347, + "step": 17640 + }, + { + "epoch": 0.41353282113492384, + "grad_norm": 5.201363563537598, + "learning_rate": 0.00012687249444002488, + "loss": 0.6851, + "step": 17641 + }, + { + "epoch": 0.41355626270972884, + "grad_norm": 2.5876693725585938, + "learning_rate": 0.00012686539915900214, + "loss": 1.1566, + "step": 17642 + }, + { + "epoch": 0.41357970428453383, + "grad_norm": 2.7953059673309326, + "learning_rate": 0.00012685830373220699, + "loss": 0.2878, + "step": 17643 + }, + { + "epoch": 0.4136031458593388, + "grad_norm": 1.9600119590759277, + "learning_rate": 0.00012685120815967795, + "loss": 0.2689, + "step": 17644 + }, + { + "epoch": 0.4136265874341438, + "grad_norm": 2.041314125061035, + "learning_rate": 0.0001268441124414535, + "loss": 0.1787, + "step": 17645 + }, + { + "epoch": 0.4136500290089488, + "grad_norm": 1.4655237197875977, + "learning_rate": 0.0001268370165775721, + "loss": 0.1609, + "step": 17646 + }, + { + "epoch": 0.4136734705837538, + "grad_norm": 5.36737585067749, + "learning_rate": 0.00012682992056807236, + "loss": 0.5492, + "step": 17647 + }, + { + "epoch": 0.4136969121585588, + "grad_norm": 1.1217552423477173, + "learning_rate": 0.00012682282441299267, + "loss": 0.0785, + "step": 17648 + }, + { + "epoch": 0.4137203537333638, + "grad_norm": 4.979503154754639, + "learning_rate": 0.00012681572811237162, + "loss": 0.2806, + "step": 17649 + }, + { + "epoch": 0.4137437953081688, + "grad_norm": 4.499761581420898, + "learning_rate": 0.00012680863166624764, + "loss": 0.4309, + "step": 17650 + }, + { + "epoch": 0.4137672368829738, + "grad_norm": 3.124660015106201, + "learning_rate": 0.00012680153507465928, + "loss": 1.4568, + "step": 17651 + }, + { + "epoch": 0.4137906784577788, + "grad_norm": 2.5546483993530273, + "learning_rate": 0.00012679443833764505, + "loss": 0.222, + "step": 17652 + }, + { + "epoch": 0.4138141200325838, + "grad_norm": 3.9847021102905273, + "learning_rate": 0.00012678734145524342, + "loss": 0.3207, + "step": 17653 + }, + { + "epoch": 0.41383756160738877, + "grad_norm": 2.7616539001464844, + "learning_rate": 0.00012678024442749292, + "loss": 0.3543, + "step": 17654 + }, + { + "epoch": 0.41386100318219377, + "grad_norm": 1.1816409826278687, + "learning_rate": 0.0001267731472544321, + "loss": 0.0963, + "step": 17655 + }, + { + "epoch": 0.41388444475699876, + "grad_norm": 1.956391453742981, + "learning_rate": 0.0001267660499360994, + "loss": 0.0732, + "step": 17656 + }, + { + "epoch": 0.41390788633180375, + "grad_norm": 5.225545883178711, + "learning_rate": 0.00012675895247253336, + "loss": 0.5046, + "step": 17657 + }, + { + "epoch": 0.41393132790660875, + "grad_norm": 3.0190470218658447, + "learning_rate": 0.0001267518548637725, + "loss": 1.3867, + "step": 17658 + }, + { + "epoch": 0.41395476948141374, + "grad_norm": 3.231393575668335, + "learning_rate": 0.0001267447571098553, + "loss": 0.2459, + "step": 17659 + }, + { + "epoch": 0.41397821105621874, + "grad_norm": 5.138014793395996, + "learning_rate": 0.00012673765921082032, + "loss": 0.1889, + "step": 17660 + }, + { + "epoch": 0.41400165263102373, + "grad_norm": 2.3372902870178223, + "learning_rate": 0.00012673056116670604, + "loss": 0.1329, + "step": 17661 + }, + { + "epoch": 0.4140250942058287, + "grad_norm": 4.822505474090576, + "learning_rate": 0.000126723462977551, + "loss": 0.2291, + "step": 17662 + }, + { + "epoch": 0.4140485357806337, + "grad_norm": 2.9482944011688232, + "learning_rate": 0.00012671636464339368, + "loss": 0.2908, + "step": 17663 + }, + { + "epoch": 0.4140719773554387, + "grad_norm": 3.731240749359131, + "learning_rate": 0.00012670926616427267, + "loss": 0.3668, + "step": 17664 + }, + { + "epoch": 0.4140954189302437, + "grad_norm": 2.9798409938812256, + "learning_rate": 0.00012670216754022634, + "loss": 0.2186, + "step": 17665 + }, + { + "epoch": 0.4141188605050487, + "grad_norm": 2.4051339626312256, + "learning_rate": 0.00012669506877129334, + "loss": 0.1109, + "step": 17666 + }, + { + "epoch": 0.4141423020798537, + "grad_norm": 1.368672251701355, + "learning_rate": 0.00012668796985751217, + "loss": 0.095, + "step": 17667 + }, + { + "epoch": 0.41416574365465875, + "grad_norm": 4.123008728027344, + "learning_rate": 0.0001266808707989213, + "loss": 0.3456, + "step": 17668 + }, + { + "epoch": 0.41418918522946374, + "grad_norm": 2.615774393081665, + "learning_rate": 0.00012667377159555925, + "loss": 0.2398, + "step": 17669 + }, + { + "epoch": 0.41421262680426874, + "grad_norm": 4.269869804382324, + "learning_rate": 0.00012666667224746461, + "loss": 0.465, + "step": 17670 + }, + { + "epoch": 0.41423606837907373, + "grad_norm": 5.722555160522461, + "learning_rate": 0.00012665957275467585, + "loss": 0.6518, + "step": 17671 + }, + { + "epoch": 0.4142595099538787, + "grad_norm": 2.572580575942993, + "learning_rate": 0.00012665247311723147, + "loss": 0.2863, + "step": 17672 + }, + { + "epoch": 0.4142829515286837, + "grad_norm": 2.7075111865997314, + "learning_rate": 0.00012664537333517006, + "loss": 0.2972, + "step": 17673 + }, + { + "epoch": 0.4143063931034887, + "grad_norm": 4.89901876449585, + "learning_rate": 0.00012663827340853007, + "loss": 0.2754, + "step": 17674 + }, + { + "epoch": 0.4143298346782937, + "grad_norm": 4.401688575744629, + "learning_rate": 0.00012663117333735005, + "loss": 0.4472, + "step": 17675 + }, + { + "epoch": 0.4143532762530987, + "grad_norm": 2.5381603240966797, + "learning_rate": 0.0001266240731216686, + "loss": 0.2426, + "step": 17676 + }, + { + "epoch": 0.4143767178279037, + "grad_norm": 5.22329568862915, + "learning_rate": 0.00012661697276152413, + "loss": 0.3753, + "step": 17677 + }, + { + "epoch": 0.4144001594027087, + "grad_norm": 4.405007839202881, + "learning_rate": 0.00012660987225695522, + "loss": 0.3896, + "step": 17678 + }, + { + "epoch": 0.4144236009775137, + "grad_norm": 2.2956020832061768, + "learning_rate": 0.0001266027716080004, + "loss": 1.3425, + "step": 17679 + }, + { + "epoch": 0.4144470425523187, + "grad_norm": 3.3690311908721924, + "learning_rate": 0.0001265956708146982, + "loss": 0.2187, + "step": 17680 + }, + { + "epoch": 0.4144704841271237, + "grad_norm": 3.2896742820739746, + "learning_rate": 0.00012658856987708713, + "loss": 1.6521, + "step": 17681 + }, + { + "epoch": 0.41449392570192867, + "grad_norm": 3.091627836227417, + "learning_rate": 0.00012658146879520574, + "loss": 0.2533, + "step": 17682 + }, + { + "epoch": 0.41451736727673366, + "grad_norm": 3.4879114627838135, + "learning_rate": 0.00012657436756909255, + "loss": 0.4262, + "step": 17683 + }, + { + "epoch": 0.41454080885153866, + "grad_norm": 3.197443723678589, + "learning_rate": 0.0001265672661987861, + "loss": 0.3178, + "step": 17684 + }, + { + "epoch": 0.41456425042634365, + "grad_norm": 1.741730809211731, + "learning_rate": 0.0001265601646843249, + "loss": 0.0883, + "step": 17685 + }, + { + "epoch": 0.41458769200114864, + "grad_norm": 4.602706432342529, + "learning_rate": 0.00012655306302574752, + "loss": 0.6189, + "step": 17686 + }, + { + "epoch": 0.41461113357595364, + "grad_norm": 4.9970316886901855, + "learning_rate": 0.00012654596122309248, + "loss": 0.6415, + "step": 17687 + }, + { + "epoch": 0.41463457515075863, + "grad_norm": 1.2503206729888916, + "learning_rate": 0.00012653885927639829, + "loss": 0.1476, + "step": 17688 + }, + { + "epoch": 0.4146580167255636, + "grad_norm": 2.448474884033203, + "learning_rate": 0.00012653175718570353, + "loss": 0.269, + "step": 17689 + }, + { + "epoch": 0.4146814583003686, + "grad_norm": 2.8449745178222656, + "learning_rate": 0.0001265246549510467, + "loss": 0.246, + "step": 17690 + }, + { + "epoch": 0.4147048998751736, + "grad_norm": 0.8383895754814148, + "learning_rate": 0.00012651755257246628, + "loss": 0.0613, + "step": 17691 + }, + { + "epoch": 0.4147283414499786, + "grad_norm": 3.1993584632873535, + "learning_rate": 0.00012651045005000096, + "loss": 0.2755, + "step": 17692 + }, + { + "epoch": 0.4147517830247836, + "grad_norm": 4.263916492462158, + "learning_rate": 0.00012650334738368915, + "loss": 0.4214, + "step": 17693 + }, + { + "epoch": 0.4147752245995886, + "grad_norm": 2.5261049270629883, + "learning_rate": 0.00012649624457356943, + "loss": 0.2319, + "step": 17694 + }, + { + "epoch": 0.4147986661743936, + "grad_norm": 4.196523666381836, + "learning_rate": 0.00012648914161968037, + "loss": 0.4546, + "step": 17695 + }, + { + "epoch": 0.4148221077491986, + "grad_norm": 1.4463541507720947, + "learning_rate": 0.00012648203852206046, + "loss": 0.1478, + "step": 17696 + }, + { + "epoch": 0.4148455493240036, + "grad_norm": 3.2109665870666504, + "learning_rate": 0.00012647493528074826, + "loss": 1.4017, + "step": 17697 + }, + { + "epoch": 0.4148689908988086, + "grad_norm": 5.81080436706543, + "learning_rate": 0.00012646783189578233, + "loss": 0.4215, + "step": 17698 + }, + { + "epoch": 0.41489243247361357, + "grad_norm": 4.355271816253662, + "learning_rate": 0.0001264607283672012, + "loss": 0.453, + "step": 17699 + }, + { + "epoch": 0.41491587404841856, + "grad_norm": 4.296247959136963, + "learning_rate": 0.0001264536246950434, + "loss": 0.2394, + "step": 17700 + }, + { + "epoch": 0.41493931562322356, + "grad_norm": 0.8253893852233887, + "learning_rate": 0.00012644652087934752, + "loss": 0.0574, + "step": 17701 + }, + { + "epoch": 0.41496275719802855, + "grad_norm": 3.428703784942627, + "learning_rate": 0.00012643941692015208, + "loss": 0.2264, + "step": 17702 + }, + { + "epoch": 0.41498619877283355, + "grad_norm": 3.4787018299102783, + "learning_rate": 0.00012643231281749557, + "loss": 0.4185, + "step": 17703 + }, + { + "epoch": 0.41500964034763854, + "grad_norm": 5.081615924835205, + "learning_rate": 0.0001264252085714166, + "loss": 0.5536, + "step": 17704 + }, + { + "epoch": 0.41503308192244354, + "grad_norm": 3.598416328430176, + "learning_rate": 0.00012641810418195372, + "loss": 0.4861, + "step": 17705 + }, + { + "epoch": 0.41505652349724853, + "grad_norm": 3.7648391723632812, + "learning_rate": 0.00012641099964914548, + "loss": 0.5115, + "step": 17706 + }, + { + "epoch": 0.4150799650720535, + "grad_norm": 2.5274672508239746, + "learning_rate": 0.00012640389497303037, + "loss": 1.5553, + "step": 17707 + }, + { + "epoch": 0.4151034066468585, + "grad_norm": 0.37926003336906433, + "learning_rate": 0.00012639679015364702, + "loss": 0.0283, + "step": 17708 + }, + { + "epoch": 0.4151268482216635, + "grad_norm": 0.6276325583457947, + "learning_rate": 0.00012638968519103393, + "loss": 0.0851, + "step": 17709 + }, + { + "epoch": 0.4151502897964685, + "grad_norm": 4.4484734535217285, + "learning_rate": 0.00012638258008522966, + "loss": 0.3154, + "step": 17710 + }, + { + "epoch": 0.4151737313712735, + "grad_norm": 3.538407802581787, + "learning_rate": 0.00012637547483627278, + "loss": 0.3394, + "step": 17711 + }, + { + "epoch": 0.4151971729460785, + "grad_norm": 2.6031460762023926, + "learning_rate": 0.00012636836944420184, + "loss": 1.4077, + "step": 17712 + }, + { + "epoch": 0.4152206145208835, + "grad_norm": 3.072587728500366, + "learning_rate": 0.00012636126390905535, + "loss": 1.7863, + "step": 17713 + }, + { + "epoch": 0.4152440560956885, + "grad_norm": 1.9068987369537354, + "learning_rate": 0.00012635415823087194, + "loss": 2.0775, + "step": 17714 + }, + { + "epoch": 0.4152674976704935, + "grad_norm": 1.2348462343215942, + "learning_rate": 0.0001263470524096901, + "loss": 0.1414, + "step": 17715 + }, + { + "epoch": 0.4152909392452985, + "grad_norm": 1.4605823755264282, + "learning_rate": 0.00012633994644554842, + "loss": 0.1156, + "step": 17716 + }, + { + "epoch": 0.41531438082010347, + "grad_norm": 3.021301507949829, + "learning_rate": 0.00012633284033848548, + "loss": 0.2477, + "step": 17717 + }, + { + "epoch": 0.41533782239490846, + "grad_norm": 3.118013620376587, + "learning_rate": 0.0001263257340885398, + "loss": 0.4349, + "step": 17718 + }, + { + "epoch": 0.4153612639697135, + "grad_norm": 4.665541648864746, + "learning_rate": 0.00012631862769574992, + "loss": 0.5823, + "step": 17719 + }, + { + "epoch": 0.4153847055445185, + "grad_norm": 3.5562744140625, + "learning_rate": 0.00012631152116015444, + "loss": 0.4154, + "step": 17720 + }, + { + "epoch": 0.4154081471193235, + "grad_norm": 2.8973002433776855, + "learning_rate": 0.0001263044144817919, + "loss": 1.5377, + "step": 17721 + }, + { + "epoch": 0.4154315886941285, + "grad_norm": 5.8338541984558105, + "learning_rate": 0.00012629730766070086, + "loss": 0.6576, + "step": 17722 + }, + { + "epoch": 0.4154550302689335, + "grad_norm": 1.1231197118759155, + "learning_rate": 0.0001262902006969199, + "loss": 0.1374, + "step": 17723 + }, + { + "epoch": 0.4154784718437385, + "grad_norm": 2.962244987487793, + "learning_rate": 0.00012628309359048758, + "loss": 0.41, + "step": 17724 + }, + { + "epoch": 0.4155019134185435, + "grad_norm": 5.620584011077881, + "learning_rate": 0.00012627598634144243, + "loss": 0.3953, + "step": 17725 + }, + { + "epoch": 0.4155253549933485, + "grad_norm": 2.5963900089263916, + "learning_rate": 0.00012626887894982307, + "loss": 0.1681, + "step": 17726 + }, + { + "epoch": 0.41554879656815347, + "grad_norm": 1.415956735610962, + "learning_rate": 0.000126261771415668, + "loss": 0.7403, + "step": 17727 + }, + { + "epoch": 0.41557223814295846, + "grad_norm": 1.7162461280822754, + "learning_rate": 0.00012625466373901583, + "loss": 0.2196, + "step": 17728 + }, + { + "epoch": 0.41559567971776346, + "grad_norm": 3.190159320831299, + "learning_rate": 0.00012624755591990515, + "loss": 0.2989, + "step": 17729 + }, + { + "epoch": 0.41561912129256845, + "grad_norm": 2.668344736099243, + "learning_rate": 0.00012624044795837445, + "loss": 0.1618, + "step": 17730 + }, + { + "epoch": 0.41564256286737344, + "grad_norm": 3.5385515689849854, + "learning_rate": 0.0001262333398544624, + "loss": 0.3828, + "step": 17731 + }, + { + "epoch": 0.41566600444217844, + "grad_norm": 0.9713113307952881, + "learning_rate": 0.00012622623160820745, + "loss": 0.0898, + "step": 17732 + }, + { + "epoch": 0.41568944601698343, + "grad_norm": 3.246669054031372, + "learning_rate": 0.00012621912321964825, + "loss": 0.4309, + "step": 17733 + }, + { + "epoch": 0.4157128875917884, + "grad_norm": 3.966108560562134, + "learning_rate": 0.00012621201468882335, + "loss": 0.4763, + "step": 17734 + }, + { + "epoch": 0.4157363291665934, + "grad_norm": 2.902672529220581, + "learning_rate": 0.00012620490601577132, + "loss": 0.3498, + "step": 17735 + }, + { + "epoch": 0.4157597707413984, + "grad_norm": 1.9362397193908691, + "learning_rate": 0.00012619779720053073, + "loss": 0.0826, + "step": 17736 + }, + { + "epoch": 0.4157832123162034, + "grad_norm": 0.8174089193344116, + "learning_rate": 0.00012619068824314017, + "loss": 0.1234, + "step": 17737 + }, + { + "epoch": 0.4158066538910084, + "grad_norm": 4.104186058044434, + "learning_rate": 0.0001261835791436382, + "loss": 0.667, + "step": 17738 + }, + { + "epoch": 0.4158300954658134, + "grad_norm": 5.857215881347656, + "learning_rate": 0.00012617646990206337, + "loss": 0.6431, + "step": 17739 + }, + { + "epoch": 0.4158535370406184, + "grad_norm": 1.908859133720398, + "learning_rate": 0.0001261693605184543, + "loss": 0.199, + "step": 17740 + }, + { + "epoch": 0.4158769786154234, + "grad_norm": 0.6514937281608582, + "learning_rate": 0.00012616225099284953, + "loss": 0.0696, + "step": 17741 + }, + { + "epoch": 0.4159004201902284, + "grad_norm": 0.7581610679626465, + "learning_rate": 0.00012615514132528766, + "loss": 0.0718, + "step": 17742 + }, + { + "epoch": 0.4159238617650334, + "grad_norm": 1.4568568468093872, + "learning_rate": 0.00012614803151580724, + "loss": 0.1711, + "step": 17743 + }, + { + "epoch": 0.41594730333983837, + "grad_norm": 1.0519386529922485, + "learning_rate": 0.00012614092156444691, + "loss": 0.0961, + "step": 17744 + }, + { + "epoch": 0.41597074491464336, + "grad_norm": 3.0811610221862793, + "learning_rate": 0.00012613381147124513, + "loss": 0.2134, + "step": 17745 + }, + { + "epoch": 0.41599418648944836, + "grad_norm": 4.100715160369873, + "learning_rate": 0.00012612670123624063, + "loss": 0.5702, + "step": 17746 + }, + { + "epoch": 0.41601762806425335, + "grad_norm": 2.1421759128570557, + "learning_rate": 0.0001261195908594719, + "loss": 0.191, + "step": 17747 + }, + { + "epoch": 0.41604106963905835, + "grad_norm": 3.294262170791626, + "learning_rate": 0.0001261124803409775, + "loss": 0.5023, + "step": 17748 + }, + { + "epoch": 0.41606451121386334, + "grad_norm": 2.848129987716675, + "learning_rate": 0.00012610536968079607, + "loss": 1.975, + "step": 17749 + }, + { + "epoch": 0.41608795278866834, + "grad_norm": 2.938915729522705, + "learning_rate": 0.00012609825887896617, + "loss": 0.3567, + "step": 17750 + }, + { + "epoch": 0.41611139436347333, + "grad_norm": 2.837801218032837, + "learning_rate": 0.00012609114793552637, + "loss": 0.3365, + "step": 17751 + }, + { + "epoch": 0.4161348359382783, + "grad_norm": 5.231606960296631, + "learning_rate": 0.00012608403685051526, + "loss": 0.5329, + "step": 17752 + }, + { + "epoch": 0.4161582775130833, + "grad_norm": 2.2585716247558594, + "learning_rate": 0.00012607692562397146, + "loss": 1.629, + "step": 17753 + }, + { + "epoch": 0.4161817190878883, + "grad_norm": 2.528228998184204, + "learning_rate": 0.00012606981425593351, + "loss": 0.2753, + "step": 17754 + }, + { + "epoch": 0.4162051606626933, + "grad_norm": 2.491499900817871, + "learning_rate": 0.00012606270274644006, + "loss": 0.3858, + "step": 17755 + }, + { + "epoch": 0.4162286022374983, + "grad_norm": 3.1608729362487793, + "learning_rate": 0.0001260555910955296, + "loss": 0.4875, + "step": 17756 + }, + { + "epoch": 0.4162520438123033, + "grad_norm": 3.0941002368927, + "learning_rate": 0.00012604847930324078, + "loss": 0.3084, + "step": 17757 + }, + { + "epoch": 0.4162754853871083, + "grad_norm": 0.8865477442741394, + "learning_rate": 0.00012604136736961218, + "loss": 0.0586, + "step": 17758 + }, + { + "epoch": 0.4162989269619133, + "grad_norm": 1.406933307647705, + "learning_rate": 0.00012603425529468241, + "loss": 0.1261, + "step": 17759 + }, + { + "epoch": 0.4163223685367183, + "grad_norm": 1.7472174167633057, + "learning_rate": 0.00012602714307849002, + "loss": 0.1697, + "step": 17760 + }, + { + "epoch": 0.4163458101115233, + "grad_norm": 2.4779393672943115, + "learning_rate": 0.00012602003072107362, + "loss": 1.7774, + "step": 17761 + }, + { + "epoch": 0.41636925168632827, + "grad_norm": 1.517518162727356, + "learning_rate": 0.0001260129182224718, + "loss": 0.183, + "step": 17762 + }, + { + "epoch": 0.41639269326113326, + "grad_norm": 2.3694324493408203, + "learning_rate": 0.00012600580558272317, + "loss": 1.5048, + "step": 17763 + }, + { + "epoch": 0.41641613483593826, + "grad_norm": 1.5788547992706299, + "learning_rate": 0.00012599869280186627, + "loss": 0.1216, + "step": 17764 + }, + { + "epoch": 0.41643957641074325, + "grad_norm": 4.021524906158447, + "learning_rate": 0.00012599157987993977, + "loss": 0.513, + "step": 17765 + }, + { + "epoch": 0.41646301798554825, + "grad_norm": 1.0264722108840942, + "learning_rate": 0.00012598446681698222, + "loss": 0.1384, + "step": 17766 + }, + { + "epoch": 0.41648645956035324, + "grad_norm": 2.5262625217437744, + "learning_rate": 0.0001259773536130322, + "loss": 1.7518, + "step": 17767 + }, + { + "epoch": 0.41650990113515823, + "grad_norm": 2.4043939113616943, + "learning_rate": 0.00012597024026812835, + "loss": 0.329, + "step": 17768 + }, + { + "epoch": 0.41653334270996323, + "grad_norm": 4.71379280090332, + "learning_rate": 0.00012596312678230925, + "loss": 0.5001, + "step": 17769 + }, + { + "epoch": 0.4165567842847683, + "grad_norm": 3.073221445083618, + "learning_rate": 0.00012595601315561346, + "loss": 0.2477, + "step": 17770 + }, + { + "epoch": 0.41658022585957327, + "grad_norm": 0.6241326928138733, + "learning_rate": 0.00012594889938807966, + "loss": 0.067, + "step": 17771 + }, + { + "epoch": 0.41660366743437827, + "grad_norm": 1.4625310897827148, + "learning_rate": 0.00012594178547974637, + "loss": 0.1686, + "step": 17772 + }, + { + "epoch": 0.41662710900918326, + "grad_norm": 5.497243404388428, + "learning_rate": 0.00012593467143065222, + "loss": 0.3121, + "step": 17773 + }, + { + "epoch": 0.41665055058398826, + "grad_norm": 2.166696548461914, + "learning_rate": 0.00012592755724083585, + "loss": 0.3432, + "step": 17774 + }, + { + "epoch": 0.41667399215879325, + "grad_norm": 3.2678699493408203, + "learning_rate": 0.0001259204429103358, + "loss": 0.3245, + "step": 17775 + }, + { + "epoch": 0.41669743373359824, + "grad_norm": 4.726624488830566, + "learning_rate": 0.0001259133284391907, + "loss": 0.5046, + "step": 17776 + }, + { + "epoch": 0.41672087530840324, + "grad_norm": 4.267308712005615, + "learning_rate": 0.00012590621382743914, + "loss": 0.3825, + "step": 17777 + }, + { + "epoch": 0.41674431688320823, + "grad_norm": 3.9785752296447754, + "learning_rate": 0.00012589909907511975, + "loss": 0.4673, + "step": 17778 + }, + { + "epoch": 0.4167677584580132, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.00012589198418227113, + "loss": 0.1603, + "step": 17779 + }, + { + "epoch": 0.4167912000328182, + "grad_norm": 0.38925644755363464, + "learning_rate": 0.00012588486914893185, + "loss": 0.0304, + "step": 17780 + }, + { + "epoch": 0.4168146416076232, + "grad_norm": 3.749725341796875, + "learning_rate": 0.00012587775397514056, + "loss": 1.709, + "step": 17781 + }, + { + "epoch": 0.4168380831824282, + "grad_norm": 2.3094563484191895, + "learning_rate": 0.00012587063866093584, + "loss": 1.4669, + "step": 17782 + }, + { + "epoch": 0.4168615247572332, + "grad_norm": 1.4430888891220093, + "learning_rate": 0.00012586352320635632, + "loss": 0.1247, + "step": 17783 + }, + { + "epoch": 0.4168849663320382, + "grad_norm": 2.7717738151550293, + "learning_rate": 0.00012585640761144057, + "loss": 0.2169, + "step": 17784 + }, + { + "epoch": 0.4169084079068432, + "grad_norm": 4.724583625793457, + "learning_rate": 0.00012584929187622724, + "loss": 0.515, + "step": 17785 + }, + { + "epoch": 0.4169318494816482, + "grad_norm": 2.596538782119751, + "learning_rate": 0.00012584217600075494, + "loss": 1.2396, + "step": 17786 + }, + { + "epoch": 0.4169552910564532, + "grad_norm": 3.1383111476898193, + "learning_rate": 0.00012583505998506228, + "loss": 0.2805, + "step": 17787 + }, + { + "epoch": 0.4169787326312582, + "grad_norm": 1.6028971672058105, + "learning_rate": 0.00012582794382918778, + "loss": 0.2283, + "step": 17788 + }, + { + "epoch": 0.41700217420606317, + "grad_norm": 4.2259392738342285, + "learning_rate": 0.00012582082753317023, + "loss": 2.5238, + "step": 17789 + }, + { + "epoch": 0.41702561578086816, + "grad_norm": 5.556058883666992, + "learning_rate": 0.00012581371109704806, + "loss": 0.4968, + "step": 17790 + }, + { + "epoch": 0.41704905735567316, + "grad_norm": 0.4896680414676666, + "learning_rate": 0.00012580659452086004, + "loss": 0.031, + "step": 17791 + }, + { + "epoch": 0.41707249893047815, + "grad_norm": 1.341856598854065, + "learning_rate": 0.00012579947780464465, + "loss": 0.1475, + "step": 17792 + }, + { + "epoch": 0.41709594050528315, + "grad_norm": 3.827755928039551, + "learning_rate": 0.0001257923609484406, + "loss": 0.4284, + "step": 17793 + }, + { + "epoch": 0.41711938208008814, + "grad_norm": 2.5935041904449463, + "learning_rate": 0.00012578524395228646, + "loss": 0.1431, + "step": 17794 + }, + { + "epoch": 0.41714282365489314, + "grad_norm": 3.468719244003296, + "learning_rate": 0.0001257781268162209, + "loss": 0.4574, + "step": 17795 + }, + { + "epoch": 0.41716626522969813, + "grad_norm": 4.8764328956604, + "learning_rate": 0.00012577100954028244, + "loss": 0.375, + "step": 17796 + }, + { + "epoch": 0.4171897068045031, + "grad_norm": 1.2569622993469238, + "learning_rate": 0.0001257638921245098, + "loss": 0.0893, + "step": 17797 + }, + { + "epoch": 0.4172131483793081, + "grad_norm": 2.693682909011841, + "learning_rate": 0.00012575677456894155, + "loss": 0.178, + "step": 17798 + }, + { + "epoch": 0.4172365899541131, + "grad_norm": 2.698394298553467, + "learning_rate": 0.0001257496568736163, + "loss": 1.2577, + "step": 17799 + }, + { + "epoch": 0.4172600315289181, + "grad_norm": 0.8713046312332153, + "learning_rate": 0.0001257425390385727, + "loss": 0.077, + "step": 17800 + }, + { + "epoch": 0.4172834731037231, + "grad_norm": 5.047590732574463, + "learning_rate": 0.00012573542106384937, + "loss": 0.5693, + "step": 17801 + }, + { + "epoch": 0.4173069146785281, + "grad_norm": 2.384197235107422, + "learning_rate": 0.0001257283029494849, + "loss": 1.8311, + "step": 17802 + }, + { + "epoch": 0.4173303562533331, + "grad_norm": 3.76619029045105, + "learning_rate": 0.00012572118469551794, + "loss": 0.3265, + "step": 17803 + }, + { + "epoch": 0.4173537978281381, + "grad_norm": 3.6085479259490967, + "learning_rate": 0.00012571406630198712, + "loss": 0.454, + "step": 17804 + }, + { + "epoch": 0.4173772394029431, + "grad_norm": 2.95293927192688, + "learning_rate": 0.00012570694776893105, + "loss": 0.3076, + "step": 17805 + }, + { + "epoch": 0.4174006809777481, + "grad_norm": 1.0856866836547852, + "learning_rate": 0.00012569982909638834, + "loss": 0.1035, + "step": 17806 + }, + { + "epoch": 0.41742412255255307, + "grad_norm": 2.418468952178955, + "learning_rate": 0.00012569271028439765, + "loss": 0.2732, + "step": 17807 + }, + { + "epoch": 0.41744756412735806, + "grad_norm": 2.167343854904175, + "learning_rate": 0.0001256855913329976, + "loss": 1.4927, + "step": 17808 + }, + { + "epoch": 0.41747100570216306, + "grad_norm": 2.3077991008758545, + "learning_rate": 0.0001256784722422268, + "loss": 0.176, + "step": 17809 + }, + { + "epoch": 0.41749444727696805, + "grad_norm": 9.788274765014648, + "learning_rate": 0.0001256713530121239, + "loss": 0.3576, + "step": 17810 + }, + { + "epoch": 0.41751788885177304, + "grad_norm": 0.9969102740287781, + "learning_rate": 0.00012566423364272747, + "loss": 0.0343, + "step": 17811 + }, + { + "epoch": 0.41754133042657804, + "grad_norm": 2.510998010635376, + "learning_rate": 0.00012565711413407623, + "loss": 0.1839, + "step": 17812 + }, + { + "epoch": 0.41756477200138303, + "grad_norm": 2.4791295528411865, + "learning_rate": 0.00012564999448620877, + "loss": 0.2982, + "step": 17813 + }, + { + "epoch": 0.417588213576188, + "grad_norm": 3.0013246536254883, + "learning_rate": 0.00012564287469916371, + "loss": 0.2492, + "step": 17814 + }, + { + "epoch": 0.417611655150993, + "grad_norm": 4.958730220794678, + "learning_rate": 0.00012563575477297968, + "loss": 0.3641, + "step": 17815 + }, + { + "epoch": 0.417635096725798, + "grad_norm": 4.147031784057617, + "learning_rate": 0.00012562863470769533, + "loss": 0.6378, + "step": 17816 + }, + { + "epoch": 0.417658538300603, + "grad_norm": 1.7418357133865356, + "learning_rate": 0.00012562151450334932, + "loss": 0.5658, + "step": 17817 + }, + { + "epoch": 0.417681979875408, + "grad_norm": 4.265190124511719, + "learning_rate": 0.0001256143941599802, + "loss": 0.5939, + "step": 17818 + }, + { + "epoch": 0.417705421450213, + "grad_norm": 1.216133713722229, + "learning_rate": 0.00012560727367762667, + "loss": 0.1365, + "step": 17819 + }, + { + "epoch": 0.417728863025018, + "grad_norm": 2.7805309295654297, + "learning_rate": 0.00012560015305632737, + "loss": 1.8106, + "step": 17820 + }, + { + "epoch": 0.41775230459982304, + "grad_norm": 9.398998260498047, + "learning_rate": 0.00012559303229612093, + "loss": 0.6361, + "step": 17821 + }, + { + "epoch": 0.41777574617462804, + "grad_norm": 4.506066799163818, + "learning_rate": 0.00012558591139704594, + "loss": 0.2799, + "step": 17822 + }, + { + "epoch": 0.41779918774943303, + "grad_norm": 4.0927886962890625, + "learning_rate": 0.0001255787903591411, + "loss": 0.3569, + "step": 17823 + }, + { + "epoch": 0.417822629324238, + "grad_norm": 7.946669578552246, + "learning_rate": 0.00012557166918244502, + "loss": 0.4181, + "step": 17824 + }, + { + "epoch": 0.417846070899043, + "grad_norm": 0.9715343713760376, + "learning_rate": 0.0001255645478669963, + "loss": 0.0922, + "step": 17825 + }, + { + "epoch": 0.417869512473848, + "grad_norm": 1.2824465036392212, + "learning_rate": 0.0001255574264128337, + "loss": 0.0984, + "step": 17826 + }, + { + "epoch": 0.417892954048653, + "grad_norm": 6.204959392547607, + "learning_rate": 0.00012555030481999576, + "loss": 0.4582, + "step": 17827 + }, + { + "epoch": 0.417916395623458, + "grad_norm": 4.14860725402832, + "learning_rate": 0.00012554318308852114, + "loss": 0.3449, + "step": 17828 + }, + { + "epoch": 0.417939837198263, + "grad_norm": 3.4176130294799805, + "learning_rate": 0.0001255360612184485, + "loss": 0.1671, + "step": 17829 + }, + { + "epoch": 0.417963278773068, + "grad_norm": 3.1602659225463867, + "learning_rate": 0.00012552893920981644, + "loss": 0.1695, + "step": 17830 + }, + { + "epoch": 0.417986720347873, + "grad_norm": 2.8182473182678223, + "learning_rate": 0.00012552181706266367, + "loss": 0.209, + "step": 17831 + }, + { + "epoch": 0.418010161922678, + "grad_norm": 1.5732362270355225, + "learning_rate": 0.0001255146947770288, + "loss": 0.1384, + "step": 17832 + }, + { + "epoch": 0.418033603497483, + "grad_norm": 3.928967237472534, + "learning_rate": 0.0001255075723529505, + "loss": 0.4882, + "step": 17833 + }, + { + "epoch": 0.41805704507228797, + "grad_norm": 4.231552600860596, + "learning_rate": 0.00012550044979046736, + "loss": 0.471, + "step": 17834 + }, + { + "epoch": 0.41808048664709296, + "grad_norm": 3.0889594554901123, + "learning_rate": 0.00012549332708961804, + "loss": 0.2277, + "step": 17835 + }, + { + "epoch": 0.41810392822189796, + "grad_norm": 7.606090545654297, + "learning_rate": 0.00012548620425044125, + "loss": 0.2606, + "step": 17836 + }, + { + "epoch": 0.41812736979670295, + "grad_norm": 2.189730167388916, + "learning_rate": 0.00012547908127297558, + "loss": 0.076, + "step": 17837 + }, + { + "epoch": 0.41815081137150795, + "grad_norm": 4.870401382446289, + "learning_rate": 0.00012547195815725974, + "loss": 0.3741, + "step": 17838 + }, + { + "epoch": 0.41817425294631294, + "grad_norm": 3.0149598121643066, + "learning_rate": 0.0001254648349033323, + "loss": 0.3197, + "step": 17839 + }, + { + "epoch": 0.41819769452111794, + "grad_norm": 2.3035552501678467, + "learning_rate": 0.00012545771151123195, + "loss": 0.2573, + "step": 17840 + }, + { + "epoch": 0.41822113609592293, + "grad_norm": 3.0265603065490723, + "learning_rate": 0.00012545058798099735, + "loss": 0.4008, + "step": 17841 + }, + { + "epoch": 0.4182445776707279, + "grad_norm": 3.844332456588745, + "learning_rate": 0.00012544346431266716, + "loss": 0.3705, + "step": 17842 + }, + { + "epoch": 0.4182680192455329, + "grad_norm": 0.6630290150642395, + "learning_rate": 0.00012543634050628, + "loss": 0.0444, + "step": 17843 + }, + { + "epoch": 0.4182914608203379, + "grad_norm": 3.7286288738250732, + "learning_rate": 0.0001254292165618745, + "loss": 0.282, + "step": 17844 + }, + { + "epoch": 0.4183149023951429, + "grad_norm": 0.9161208271980286, + "learning_rate": 0.00012542209247948943, + "loss": 0.1065, + "step": 17845 + }, + { + "epoch": 0.4183383439699479, + "grad_norm": 3.233319044113159, + "learning_rate": 0.00012541496825916335, + "loss": 0.2274, + "step": 17846 + }, + { + "epoch": 0.4183617855447529, + "grad_norm": 3.700213670730591, + "learning_rate": 0.0001254078439009349, + "loss": 0.4382, + "step": 17847 + }, + { + "epoch": 0.4183852271195579, + "grad_norm": 3.250025510787964, + "learning_rate": 0.00012540071940484284, + "loss": 0.3592, + "step": 17848 + }, + { + "epoch": 0.4184086686943629, + "grad_norm": 4.150752544403076, + "learning_rate": 0.00012539359477092572, + "loss": 0.6832, + "step": 17849 + }, + { + "epoch": 0.4184321102691679, + "grad_norm": 4.516594886779785, + "learning_rate": 0.00012538646999922222, + "loss": 0.422, + "step": 17850 + }, + { + "epoch": 0.4184555518439729, + "grad_norm": 1.0482492446899414, + "learning_rate": 0.00012537934508977106, + "loss": 0.0798, + "step": 17851 + }, + { + "epoch": 0.41847899341877787, + "grad_norm": 3.2402379512786865, + "learning_rate": 0.00012537222004261085, + "loss": 0.3119, + "step": 17852 + }, + { + "epoch": 0.41850243499358286, + "grad_norm": 3.154813766479492, + "learning_rate": 0.00012536509485778024, + "loss": 0.2984, + "step": 17853 + }, + { + "epoch": 0.41852587656838786, + "grad_norm": 6.5015363693237305, + "learning_rate": 0.00012535796953531793, + "loss": 0.3125, + "step": 17854 + }, + { + "epoch": 0.41854931814319285, + "grad_norm": 5.485360622406006, + "learning_rate": 0.00012535084407526258, + "loss": 0.2169, + "step": 17855 + }, + { + "epoch": 0.41857275971799784, + "grad_norm": 2.6697635650634766, + "learning_rate": 0.0001253437184776528, + "loss": 0.3125, + "step": 17856 + }, + { + "epoch": 0.41859620129280284, + "grad_norm": 3.0794007778167725, + "learning_rate": 0.00012533659274252732, + "loss": 0.1953, + "step": 17857 + }, + { + "epoch": 0.41861964286760783, + "grad_norm": 0.5030242800712585, + "learning_rate": 0.00012532946686992478, + "loss": 0.0195, + "step": 17858 + }, + { + "epoch": 0.4186430844424128, + "grad_norm": 2.3341503143310547, + "learning_rate": 0.0001253223408598838, + "loss": 0.2592, + "step": 17859 + }, + { + "epoch": 0.4186665260172178, + "grad_norm": 2.9595553874969482, + "learning_rate": 0.0001253152147124431, + "loss": 0.3846, + "step": 17860 + }, + { + "epoch": 0.4186899675920228, + "grad_norm": 4.2564697265625, + "learning_rate": 0.00012530808842764138, + "loss": 0.4136, + "step": 17861 + }, + { + "epoch": 0.4187134091668278, + "grad_norm": 3.6201794147491455, + "learning_rate": 0.00012530096200551722, + "loss": 0.3944, + "step": 17862 + }, + { + "epoch": 0.4187368507416328, + "grad_norm": 0.787097692489624, + "learning_rate": 0.00012529383544610933, + "loss": 0.0703, + "step": 17863 + }, + { + "epoch": 0.4187602923164378, + "grad_norm": 4.37881326675415, + "learning_rate": 0.00012528670874945637, + "loss": 0.3202, + "step": 17864 + }, + { + "epoch": 0.4187837338912428, + "grad_norm": 4.351399898529053, + "learning_rate": 0.00012527958191559707, + "loss": 0.3927, + "step": 17865 + }, + { + "epoch": 0.4188071754660478, + "grad_norm": 1.1823123693466187, + "learning_rate": 0.00012527245494456997, + "loss": 0.13, + "step": 17866 + }, + { + "epoch": 0.4188306170408528, + "grad_norm": 4.331855773925781, + "learning_rate": 0.00012526532783641386, + "loss": 0.2823, + "step": 17867 + }, + { + "epoch": 0.4188540586156578, + "grad_norm": 3.797992706298828, + "learning_rate": 0.00012525820059116737, + "loss": 0.2645, + "step": 17868 + }, + { + "epoch": 0.41887750019046277, + "grad_norm": 3.507789373397827, + "learning_rate": 0.00012525107320886914, + "loss": 1.796, + "step": 17869 + }, + { + "epoch": 0.41890094176526776, + "grad_norm": 4.219794273376465, + "learning_rate": 0.00012524394568955795, + "loss": 0.4599, + "step": 17870 + }, + { + "epoch": 0.41892438334007276, + "grad_norm": 4.125687122344971, + "learning_rate": 0.00012523681803327236, + "loss": 0.3007, + "step": 17871 + }, + { + "epoch": 0.4189478249148778, + "grad_norm": 1.7118322849273682, + "learning_rate": 0.00012522969024005107, + "loss": 0.188, + "step": 17872 + }, + { + "epoch": 0.4189712664896828, + "grad_norm": 2.4631247520446777, + "learning_rate": 0.00012522256230993278, + "loss": 1.4397, + "step": 17873 + }, + { + "epoch": 0.4189947080644878, + "grad_norm": 0.8537303805351257, + "learning_rate": 0.00012521543424295615, + "loss": 0.0799, + "step": 17874 + }, + { + "epoch": 0.4190181496392928, + "grad_norm": 2.8544960021972656, + "learning_rate": 0.0001252083060391599, + "loss": 0.2714, + "step": 17875 + }, + { + "epoch": 0.4190415912140978, + "grad_norm": 4.3132123947143555, + "learning_rate": 0.00012520117769858265, + "loss": 0.5531, + "step": 17876 + }, + { + "epoch": 0.4190650327889028, + "grad_norm": 2.404910087585449, + "learning_rate": 0.0001251940492212631, + "loss": 0.281, + "step": 17877 + }, + { + "epoch": 0.4190884743637078, + "grad_norm": 1.8114326000213623, + "learning_rate": 0.00012518692060723994, + "loss": 0.1907, + "step": 17878 + }, + { + "epoch": 0.41911191593851277, + "grad_norm": 10.313735008239746, + "learning_rate": 0.00012517979185655182, + "loss": 0.5924, + "step": 17879 + }, + { + "epoch": 0.41913535751331776, + "grad_norm": 4.148843765258789, + "learning_rate": 0.00012517266296923745, + "loss": 0.6194, + "step": 17880 + }, + { + "epoch": 0.41915879908812276, + "grad_norm": 1.8852839469909668, + "learning_rate": 0.00012516553394533554, + "loss": 1.6335, + "step": 17881 + }, + { + "epoch": 0.41918224066292775, + "grad_norm": 1.7278549671173096, + "learning_rate": 0.00012515840478488467, + "loss": 0.1105, + "step": 17882 + }, + { + "epoch": 0.41920568223773275, + "grad_norm": 4.860544204711914, + "learning_rate": 0.00012515127548792365, + "loss": 0.5338, + "step": 17883 + }, + { + "epoch": 0.41922912381253774, + "grad_norm": 6.253164291381836, + "learning_rate": 0.00012514414605449106, + "loss": 0.3869, + "step": 17884 + }, + { + "epoch": 0.41925256538734273, + "grad_norm": 4.020313262939453, + "learning_rate": 0.0001251370164846256, + "loss": 0.3484, + "step": 17885 + }, + { + "epoch": 0.41927600696214773, + "grad_norm": 0.9201660752296448, + "learning_rate": 0.00012512988677836604, + "loss": 0.1385, + "step": 17886 + }, + { + "epoch": 0.4192994485369527, + "grad_norm": 4.223028659820557, + "learning_rate": 0.00012512275693575103, + "loss": 0.2485, + "step": 17887 + }, + { + "epoch": 0.4193228901117577, + "grad_norm": 3.467710256576538, + "learning_rate": 0.00012511562695681917, + "loss": 0.3089, + "step": 17888 + }, + { + "epoch": 0.4193463316865627, + "grad_norm": 0.9655916094779968, + "learning_rate": 0.00012510849684160925, + "loss": 0.0765, + "step": 17889 + }, + { + "epoch": 0.4193697732613677, + "grad_norm": 3.3814241886138916, + "learning_rate": 0.0001251013665901599, + "loss": 0.3593, + "step": 17890 + }, + { + "epoch": 0.4193932148361727, + "grad_norm": 4.651219367980957, + "learning_rate": 0.00012509423620250983, + "loss": 0.2691, + "step": 17891 + }, + { + "epoch": 0.4194166564109777, + "grad_norm": 5.381046295166016, + "learning_rate": 0.00012508710567869772, + "loss": 0.4489, + "step": 17892 + }, + { + "epoch": 0.4194400979857827, + "grad_norm": 3.6916885375976562, + "learning_rate": 0.0001250799750187623, + "loss": 0.5555, + "step": 17893 + }, + { + "epoch": 0.4194635395605877, + "grad_norm": 3.372080087661743, + "learning_rate": 0.0001250728442227422, + "loss": 0.3015, + "step": 17894 + }, + { + "epoch": 0.4194869811353927, + "grad_norm": 1.8380595445632935, + "learning_rate": 0.00012506571329067617, + "loss": 0.2592, + "step": 17895 + }, + { + "epoch": 0.4195104227101977, + "grad_norm": 0.8958811163902283, + "learning_rate": 0.00012505858222260286, + "loss": 0.1143, + "step": 17896 + }, + { + "epoch": 0.41953386428500267, + "grad_norm": 4.148491382598877, + "learning_rate": 0.00012505145101856098, + "loss": 0.4063, + "step": 17897 + }, + { + "epoch": 0.41955730585980766, + "grad_norm": 2.6441686153411865, + "learning_rate": 0.00012504431967858917, + "loss": 0.2266, + "step": 17898 + }, + { + "epoch": 0.41958074743461266, + "grad_norm": 5.63945198059082, + "learning_rate": 0.0001250371882027263, + "loss": 0.4487, + "step": 17899 + }, + { + "epoch": 0.41960418900941765, + "grad_norm": 4.545249938964844, + "learning_rate": 0.00012503005659101084, + "loss": 0.4949, + "step": 17900 + }, + { + "epoch": 0.41962763058422264, + "grad_norm": 4.276548385620117, + "learning_rate": 0.0001250229248434816, + "loss": 0.8031, + "step": 17901 + }, + { + "epoch": 0.41965107215902764, + "grad_norm": 3.1920835971832275, + "learning_rate": 0.0001250157929601773, + "loss": 0.2078, + "step": 17902 + }, + { + "epoch": 0.41967451373383263, + "grad_norm": 0.7274578809738159, + "learning_rate": 0.00012500866094113659, + "loss": 0.0337, + "step": 17903 + }, + { + "epoch": 0.4196979553086376, + "grad_norm": 2.450928211212158, + "learning_rate": 0.00012500152878639818, + "loss": 0.2543, + "step": 17904 + }, + { + "epoch": 0.4197213968834426, + "grad_norm": 2.6151766777038574, + "learning_rate": 0.0001249943964960008, + "loss": 0.1803, + "step": 17905 + }, + { + "epoch": 0.4197448384582476, + "grad_norm": 4.850646018981934, + "learning_rate": 0.0001249872640699831, + "loss": 0.4796, + "step": 17906 + }, + { + "epoch": 0.4197682800330526, + "grad_norm": 0.8535318374633789, + "learning_rate": 0.00012498013150838376, + "loss": 0.1292, + "step": 17907 + }, + { + "epoch": 0.4197917216078576, + "grad_norm": 3.189626693725586, + "learning_rate": 0.00012497299881124158, + "loss": 0.2689, + "step": 17908 + }, + { + "epoch": 0.4198151631826626, + "grad_norm": 2.7778592109680176, + "learning_rate": 0.0001249658659785952, + "loss": 0.2858, + "step": 17909 + }, + { + "epoch": 0.4198386047574676, + "grad_norm": 2.0335750579833984, + "learning_rate": 0.00012495873301048328, + "loss": 0.1282, + "step": 17910 + }, + { + "epoch": 0.4198620463322726, + "grad_norm": 2.901420831680298, + "learning_rate": 0.00012495159990694464, + "loss": 0.1746, + "step": 17911 + }, + { + "epoch": 0.4198854879070776, + "grad_norm": 1.6709424257278442, + "learning_rate": 0.00012494446666801788, + "loss": 0.1954, + "step": 17912 + }, + { + "epoch": 0.4199089294818826, + "grad_norm": 0.8465816378593445, + "learning_rate": 0.00012493733329374177, + "loss": 0.0972, + "step": 17913 + }, + { + "epoch": 0.41993237105668757, + "grad_norm": 4.928277969360352, + "learning_rate": 0.00012493019978415497, + "loss": 0.5435, + "step": 17914 + }, + { + "epoch": 0.41995581263149256, + "grad_norm": 0.6507938504219055, + "learning_rate": 0.0001249230661392962, + "loss": 0.0444, + "step": 17915 + }, + { + "epoch": 0.41997925420629756, + "grad_norm": 6.069448947906494, + "learning_rate": 0.00012491593235920418, + "loss": 0.6005, + "step": 17916 + }, + { + "epoch": 0.42000269578110255, + "grad_norm": 1.096887230873108, + "learning_rate": 0.00012490879844391758, + "loss": 0.1255, + "step": 17917 + }, + { + "epoch": 0.42002613735590755, + "grad_norm": 2.6556332111358643, + "learning_rate": 0.0001249016643934752, + "loss": 1.7447, + "step": 17918 + }, + { + "epoch": 0.42004957893071254, + "grad_norm": 1.7300300598144531, + "learning_rate": 0.00012489453020791565, + "loss": 0.9847, + "step": 17919 + }, + { + "epoch": 0.42007302050551754, + "grad_norm": 4.919924259185791, + "learning_rate": 0.00012488739588727767, + "loss": 0.363, + "step": 17920 + }, + { + "epoch": 0.42009646208032253, + "grad_norm": 5.5419440269470215, + "learning_rate": 0.0001248802614316, + "loss": 0.7854, + "step": 17921 + }, + { + "epoch": 0.4201199036551275, + "grad_norm": 3.913163423538208, + "learning_rate": 0.00012487312684092132, + "loss": 0.3391, + "step": 17922 + }, + { + "epoch": 0.4201433452299326, + "grad_norm": 2.8478221893310547, + "learning_rate": 0.00012486599211528036, + "loss": 1.5536, + "step": 17923 + }, + { + "epoch": 0.42016678680473757, + "grad_norm": 3.849451780319214, + "learning_rate": 0.00012485885725471585, + "loss": 0.4917, + "step": 17924 + }, + { + "epoch": 0.42019022837954256, + "grad_norm": 4.226773738861084, + "learning_rate": 0.00012485172225926643, + "loss": 0.4394, + "step": 17925 + }, + { + "epoch": 0.42021366995434756, + "grad_norm": 2.280285120010376, + "learning_rate": 0.00012484458712897092, + "loss": 1.4775, + "step": 17926 + }, + { + "epoch": 0.42023711152915255, + "grad_norm": 2.3279175758361816, + "learning_rate": 0.00012483745186386795, + "loss": 1.6872, + "step": 17927 + }, + { + "epoch": 0.42026055310395755, + "grad_norm": 0.5765505433082581, + "learning_rate": 0.00012483031646399627, + "loss": 0.0437, + "step": 17928 + }, + { + "epoch": 0.42028399467876254, + "grad_norm": 4.212642192840576, + "learning_rate": 0.00012482318092939458, + "loss": 0.4285, + "step": 17929 + }, + { + "epoch": 0.42030743625356753, + "grad_norm": 3.9221010208129883, + "learning_rate": 0.00012481604526010163, + "loss": 0.5101, + "step": 17930 + }, + { + "epoch": 0.42033087782837253, + "grad_norm": 1.9632381200790405, + "learning_rate": 0.00012480890945615615, + "loss": 0.2237, + "step": 17931 + }, + { + "epoch": 0.4203543194031775, + "grad_norm": 1.0119519233703613, + "learning_rate": 0.0001248017735175968, + "loss": 0.0765, + "step": 17932 + }, + { + "epoch": 0.4203777609779825, + "grad_norm": 0.5578289031982422, + "learning_rate": 0.00012479463744446233, + "loss": 0.0719, + "step": 17933 + }, + { + "epoch": 0.4204012025527875, + "grad_norm": 3.53837513923645, + "learning_rate": 0.00012478750123679145, + "loss": 0.6354, + "step": 17934 + }, + { + "epoch": 0.4204246441275925, + "grad_norm": 3.020664691925049, + "learning_rate": 0.00012478036489462293, + "loss": 0.2547, + "step": 17935 + }, + { + "epoch": 0.4204480857023975, + "grad_norm": 3.2578792572021484, + "learning_rate": 0.0001247732284179954, + "loss": 0.3506, + "step": 17936 + }, + { + "epoch": 0.4204715272772025, + "grad_norm": 0.6326411366462708, + "learning_rate": 0.00012476609180694768, + "loss": 0.0658, + "step": 17937 + }, + { + "epoch": 0.4204949688520075, + "grad_norm": 1.2883648872375488, + "learning_rate": 0.00012475895506151844, + "loss": 0.1475, + "step": 17938 + }, + { + "epoch": 0.4205184104268125, + "grad_norm": 5.529178619384766, + "learning_rate": 0.0001247518181817464, + "loss": 0.5209, + "step": 17939 + }, + { + "epoch": 0.4205418520016175, + "grad_norm": 3.8715147972106934, + "learning_rate": 0.00012474468116767032, + "loss": 0.4884, + "step": 17940 + }, + { + "epoch": 0.42056529357642247, + "grad_norm": 5.416350841522217, + "learning_rate": 0.0001247375440193289, + "loss": 0.8876, + "step": 17941 + }, + { + "epoch": 0.42058873515122747, + "grad_norm": 0.9734994173049927, + "learning_rate": 0.00012473040673676084, + "loss": 0.0765, + "step": 17942 + }, + { + "epoch": 0.42061217672603246, + "grad_norm": 2.975445508956909, + "learning_rate": 0.00012472326932000494, + "loss": 0.2793, + "step": 17943 + }, + { + "epoch": 0.42063561830083745, + "grad_norm": 1.1661380529403687, + "learning_rate": 0.00012471613176909987, + "loss": 0.1326, + "step": 17944 + }, + { + "epoch": 0.42065905987564245, + "grad_norm": 1.1337741613388062, + "learning_rate": 0.00012470899408408435, + "loss": 0.139, + "step": 17945 + }, + { + "epoch": 0.42068250145044744, + "grad_norm": 3.6227447986602783, + "learning_rate": 0.00012470185626499712, + "loss": 0.2935, + "step": 17946 + }, + { + "epoch": 0.42070594302525244, + "grad_norm": 3.0184998512268066, + "learning_rate": 0.00012469471831187698, + "loss": 0.3347, + "step": 17947 + }, + { + "epoch": 0.42072938460005743, + "grad_norm": 2.633498430252075, + "learning_rate": 0.00012468758022476256, + "loss": 0.2488, + "step": 17948 + }, + { + "epoch": 0.4207528261748624, + "grad_norm": 2.4790055751800537, + "learning_rate": 0.00012468044200369266, + "loss": 0.2327, + "step": 17949 + }, + { + "epoch": 0.4207762677496674, + "grad_norm": 1.444591999053955, + "learning_rate": 0.000124673303648706, + "loss": 0.1193, + "step": 17950 + }, + { + "epoch": 0.4207997093244724, + "grad_norm": 1.2269947528839111, + "learning_rate": 0.00012466616515984126, + "loss": 0.14, + "step": 17951 + }, + { + "epoch": 0.4208231508992774, + "grad_norm": 0.7640067338943481, + "learning_rate": 0.00012465902653713723, + "loss": 0.1141, + "step": 17952 + }, + { + "epoch": 0.4208465924740824, + "grad_norm": 1.8717299699783325, + "learning_rate": 0.00012465188778063263, + "loss": 0.9655, + "step": 17953 + }, + { + "epoch": 0.4208700340488874, + "grad_norm": 1.2881031036376953, + "learning_rate": 0.00012464474889036623, + "loss": 0.1592, + "step": 17954 + }, + { + "epoch": 0.4208934756236924, + "grad_norm": 5.233586311340332, + "learning_rate": 0.0001246376098663767, + "loss": 0.5269, + "step": 17955 + }, + { + "epoch": 0.4209169171984974, + "grad_norm": 2.6972360610961914, + "learning_rate": 0.0001246304707087028, + "loss": 0.2012, + "step": 17956 + }, + { + "epoch": 0.4209403587733024, + "grad_norm": 3.098933696746826, + "learning_rate": 0.00012462333141738329, + "loss": 0.3079, + "step": 17957 + }, + { + "epoch": 0.4209638003481074, + "grad_norm": 2.744156837463379, + "learning_rate": 0.00012461619199245685, + "loss": 0.279, + "step": 17958 + }, + { + "epoch": 0.42098724192291237, + "grad_norm": 2.6237359046936035, + "learning_rate": 0.0001246090524339623, + "loss": 0.2631, + "step": 17959 + }, + { + "epoch": 0.42101068349771736, + "grad_norm": 3.4940738677978516, + "learning_rate": 0.00012460191274193833, + "loss": 0.245, + "step": 17960 + }, + { + "epoch": 0.42103412507252236, + "grad_norm": 1.8095213174819946, + "learning_rate": 0.0001245947729164237, + "loss": 0.2068, + "step": 17961 + }, + { + "epoch": 0.42105756664732735, + "grad_norm": 0.9030243158340454, + "learning_rate": 0.00012458763295745712, + "loss": 0.0709, + "step": 17962 + }, + { + "epoch": 0.42108100822213235, + "grad_norm": 4.037835121154785, + "learning_rate": 0.00012458049286507737, + "loss": 0.4403, + "step": 17963 + }, + { + "epoch": 0.42110444979693734, + "grad_norm": 3.296966075897217, + "learning_rate": 0.00012457335263932315, + "loss": 0.4819, + "step": 17964 + }, + { + "epoch": 0.42112789137174234, + "grad_norm": 3.4674363136291504, + "learning_rate": 0.00012456621228023324, + "loss": 0.402, + "step": 17965 + }, + { + "epoch": 0.42115133294654733, + "grad_norm": 3.0246522426605225, + "learning_rate": 0.0001245590717878464, + "loss": 0.2217, + "step": 17966 + }, + { + "epoch": 0.4211747745213523, + "grad_norm": 2.2846567630767822, + "learning_rate": 0.0001245519311622013, + "loss": 0.1259, + "step": 17967 + }, + { + "epoch": 0.4211982160961573, + "grad_norm": 3.820293664932251, + "learning_rate": 0.00012454479040333675, + "loss": 0.3883, + "step": 17968 + }, + { + "epoch": 0.4212216576709623, + "grad_norm": 2.8997442722320557, + "learning_rate": 0.0001245376495112915, + "loss": 0.1262, + "step": 17969 + }, + { + "epoch": 0.4212450992457673, + "grad_norm": 3.085643768310547, + "learning_rate": 0.00012453050848610424, + "loss": 0.3914, + "step": 17970 + }, + { + "epoch": 0.4212685408205723, + "grad_norm": 0.9007002711296082, + "learning_rate": 0.00012452336732781376, + "loss": 0.069, + "step": 17971 + }, + { + "epoch": 0.4212919823953773, + "grad_norm": 2.8199732303619385, + "learning_rate": 0.0001245162260364588, + "loss": 0.3555, + "step": 17972 + }, + { + "epoch": 0.4213154239701823, + "grad_norm": 4.605339527130127, + "learning_rate": 0.00012450908461207814, + "loss": 0.4676, + "step": 17973 + }, + { + "epoch": 0.42133886554498734, + "grad_norm": 3.63381290435791, + "learning_rate": 0.00012450194305471045, + "loss": 0.358, + "step": 17974 + }, + { + "epoch": 0.42136230711979233, + "grad_norm": 9.628496170043945, + "learning_rate": 0.00012449480136439455, + "loss": 1.186, + "step": 17975 + }, + { + "epoch": 0.42138574869459733, + "grad_norm": 2.1161136627197266, + "learning_rate": 0.00012448765954116916, + "loss": 0.1838, + "step": 17976 + }, + { + "epoch": 0.4214091902694023, + "grad_norm": 3.922804117202759, + "learning_rate": 0.00012448051758507305, + "loss": 0.5292, + "step": 17977 + }, + { + "epoch": 0.4214326318442073, + "grad_norm": 1.3886879682540894, + "learning_rate": 0.00012447337549614495, + "loss": 0.203, + "step": 17978 + }, + { + "epoch": 0.4214560734190123, + "grad_norm": 3.4175009727478027, + "learning_rate": 0.00012446623327442367, + "loss": 0.3502, + "step": 17979 + }, + { + "epoch": 0.4214795149938173, + "grad_norm": 2.0343408584594727, + "learning_rate": 0.00012445909091994786, + "loss": 0.1777, + "step": 17980 + }, + { + "epoch": 0.4215029565686223, + "grad_norm": 3.088581085205078, + "learning_rate": 0.00012445194843275638, + "loss": 0.4725, + "step": 17981 + }, + { + "epoch": 0.4215263981434273, + "grad_norm": 2.3207709789276123, + "learning_rate": 0.00012444480581288794, + "loss": 1.5122, + "step": 17982 + }, + { + "epoch": 0.4215498397182323, + "grad_norm": 2.6955487728118896, + "learning_rate": 0.00012443766306038126, + "loss": 0.3517, + "step": 17983 + }, + { + "epoch": 0.4215732812930373, + "grad_norm": 3.6128623485565186, + "learning_rate": 0.00012443052017527514, + "loss": 0.3526, + "step": 17984 + }, + { + "epoch": 0.4215967228678423, + "grad_norm": 1.6086318492889404, + "learning_rate": 0.00012442337715760834, + "loss": 0.1493, + "step": 17985 + }, + { + "epoch": 0.42162016444264727, + "grad_norm": 1.4810926914215088, + "learning_rate": 0.0001244162340074196, + "loss": 0.1716, + "step": 17986 + }, + { + "epoch": 0.42164360601745227, + "grad_norm": 2.9239754676818848, + "learning_rate": 0.00012440909072474772, + "loss": 0.2346, + "step": 17987 + }, + { + "epoch": 0.42166704759225726, + "grad_norm": 1.0994977951049805, + "learning_rate": 0.0001244019473096314, + "loss": 0.1526, + "step": 17988 + }, + { + "epoch": 0.42169048916706225, + "grad_norm": 5.51434326171875, + "learning_rate": 0.00012439480376210944, + "loss": 0.5454, + "step": 17989 + }, + { + "epoch": 0.42171393074186725, + "grad_norm": 3.623567819595337, + "learning_rate": 0.00012438766008222057, + "loss": 0.4552, + "step": 17990 + }, + { + "epoch": 0.42173737231667224, + "grad_norm": 1.1914082765579224, + "learning_rate": 0.0001243805162700036, + "loss": 0.0703, + "step": 17991 + }, + { + "epoch": 0.42176081389147724, + "grad_norm": 4.285167217254639, + "learning_rate": 0.00012437337232549725, + "loss": 0.402, + "step": 17992 + }, + { + "epoch": 0.42178425546628223, + "grad_norm": 5.4780497550964355, + "learning_rate": 0.0001243662282487403, + "loss": 0.5516, + "step": 17993 + }, + { + "epoch": 0.4218076970410872, + "grad_norm": 2.1745030879974365, + "learning_rate": 0.0001243590840397715, + "loss": 0.9331, + "step": 17994 + }, + { + "epoch": 0.4218311386158922, + "grad_norm": 2.105915069580078, + "learning_rate": 0.00012435193969862962, + "loss": 0.1726, + "step": 17995 + }, + { + "epoch": 0.4218545801906972, + "grad_norm": 2.043269634246826, + "learning_rate": 0.00012434479522535342, + "loss": 0.1402, + "step": 17996 + }, + { + "epoch": 0.4218780217655022, + "grad_norm": 0.6581394076347351, + "learning_rate": 0.00012433765061998172, + "loss": 0.0316, + "step": 17997 + }, + { + "epoch": 0.4219014633403072, + "grad_norm": 3.9339747428894043, + "learning_rate": 0.00012433050588255324, + "loss": 0.2836, + "step": 17998 + }, + { + "epoch": 0.4219249049151122, + "grad_norm": 3.7922677993774414, + "learning_rate": 0.00012432336101310672, + "loss": 0.408, + "step": 17999 + }, + { + "epoch": 0.4219483464899172, + "grad_norm": 2.346402168273926, + "learning_rate": 0.00012431621601168098, + "loss": 0.1338, + "step": 18000 + }, + { + "epoch": 0.4219717880647222, + "grad_norm": 3.4513051509857178, + "learning_rate": 0.00012430907087831474, + "loss": 0.4389, + "step": 18001 + }, + { + "epoch": 0.4219952296395272, + "grad_norm": 1.7624653577804565, + "learning_rate": 0.00012430192561304682, + "loss": 0.2026, + "step": 18002 + }, + { + "epoch": 0.4220186712143322, + "grad_norm": 2.4908130168914795, + "learning_rate": 0.00012429478021591596, + "loss": 0.246, + "step": 18003 + }, + { + "epoch": 0.42204211278913717, + "grad_norm": 2.3631064891815186, + "learning_rate": 0.00012428763468696094, + "loss": 1.5901, + "step": 18004 + }, + { + "epoch": 0.42206555436394216, + "grad_norm": 3.8065991401672363, + "learning_rate": 0.00012428048902622053, + "loss": 0.4395, + "step": 18005 + }, + { + "epoch": 0.42208899593874716, + "grad_norm": 0.7494539618492126, + "learning_rate": 0.00012427334323373353, + "loss": 0.0516, + "step": 18006 + }, + { + "epoch": 0.42211243751355215, + "grad_norm": 4.629289150238037, + "learning_rate": 0.00012426619730953865, + "loss": 0.4662, + "step": 18007 + }, + { + "epoch": 0.42213587908835715, + "grad_norm": 3.8659138679504395, + "learning_rate": 0.00012425905125367472, + "loss": 0.3436, + "step": 18008 + }, + { + "epoch": 0.42215932066316214, + "grad_norm": 3.652414083480835, + "learning_rate": 0.0001242519050661805, + "loss": 0.4674, + "step": 18009 + }, + { + "epoch": 0.42218276223796714, + "grad_norm": 2.919416904449463, + "learning_rate": 0.00012424475874709475, + "loss": 0.2743, + "step": 18010 + }, + { + "epoch": 0.42220620381277213, + "grad_norm": 0.8616435527801514, + "learning_rate": 0.00012423761229645626, + "loss": 0.0579, + "step": 18011 + }, + { + "epoch": 0.4222296453875771, + "grad_norm": 2.8249385356903076, + "learning_rate": 0.0001242304657143038, + "loss": 1.005, + "step": 18012 + }, + { + "epoch": 0.4222530869623821, + "grad_norm": 3.9294064044952393, + "learning_rate": 0.00012422331900067615, + "loss": 0.3187, + "step": 18013 + }, + { + "epoch": 0.4222765285371871, + "grad_norm": 3.8904614448547363, + "learning_rate": 0.00012421617215561213, + "loss": 0.3979, + "step": 18014 + }, + { + "epoch": 0.4222999701119921, + "grad_norm": 3.2434232234954834, + "learning_rate": 0.00012420902517915042, + "loss": 1.8467, + "step": 18015 + }, + { + "epoch": 0.4223234116867971, + "grad_norm": 4.3983869552612305, + "learning_rate": 0.0001242018780713299, + "loss": 0.3994, + "step": 18016 + }, + { + "epoch": 0.4223468532616021, + "grad_norm": 2.7513790130615234, + "learning_rate": 0.0001241947308321893, + "loss": 0.1872, + "step": 18017 + }, + { + "epoch": 0.4223702948364071, + "grad_norm": 1.0102077722549438, + "learning_rate": 0.0001241875834617674, + "loss": 0.1142, + "step": 18018 + }, + { + "epoch": 0.4223937364112121, + "grad_norm": 1.760597825050354, + "learning_rate": 0.00012418043596010297, + "loss": 0.1985, + "step": 18019 + }, + { + "epoch": 0.4224171779860171, + "grad_norm": 3.1497182846069336, + "learning_rate": 0.00012417328832723485, + "loss": 0.123, + "step": 18020 + }, + { + "epoch": 0.4224406195608221, + "grad_norm": 2.9114041328430176, + "learning_rate": 0.00012416614056320177, + "loss": 1.88, + "step": 18021 + }, + { + "epoch": 0.42246406113562707, + "grad_norm": 0.9775633215904236, + "learning_rate": 0.00012415899266804257, + "loss": 0.0913, + "step": 18022 + }, + { + "epoch": 0.42248750271043206, + "grad_norm": 1.0398818254470825, + "learning_rate": 0.00012415184464179596, + "loss": 0.1285, + "step": 18023 + }, + { + "epoch": 0.42251094428523706, + "grad_norm": 1.8130426406860352, + "learning_rate": 0.00012414469648450077, + "loss": 0.2128, + "step": 18024 + }, + { + "epoch": 0.42253438586004205, + "grad_norm": 2.43153977394104, + "learning_rate": 0.00012413754819619576, + "loss": 0.3339, + "step": 18025 + }, + { + "epoch": 0.4225578274348471, + "grad_norm": 4.981732368469238, + "learning_rate": 0.00012413039977691976, + "loss": 0.5143, + "step": 18026 + }, + { + "epoch": 0.4225812690096521, + "grad_norm": 2.8041436672210693, + "learning_rate": 0.00012412325122671154, + "loss": 1.4436, + "step": 18027 + }, + { + "epoch": 0.4226047105844571, + "grad_norm": 3.49122953414917, + "learning_rate": 0.00012411610254560985, + "loss": 0.3456, + "step": 18028 + }, + { + "epoch": 0.4226281521592621, + "grad_norm": 4.372109889984131, + "learning_rate": 0.00012410895373365355, + "loss": 0.3913, + "step": 18029 + }, + { + "epoch": 0.4226515937340671, + "grad_norm": 3.214470624923706, + "learning_rate": 0.00012410180479088137, + "loss": 1.981, + "step": 18030 + }, + { + "epoch": 0.42267503530887207, + "grad_norm": 3.8514773845672607, + "learning_rate": 0.00012409465571733207, + "loss": 0.3165, + "step": 18031 + }, + { + "epoch": 0.42269847688367707, + "grad_norm": 4.188338756561279, + "learning_rate": 0.00012408750651304455, + "loss": 0.2978, + "step": 18032 + }, + { + "epoch": 0.42272191845848206, + "grad_norm": 4.121883392333984, + "learning_rate": 0.00012408035717805752, + "loss": 0.4283, + "step": 18033 + }, + { + "epoch": 0.42274536003328705, + "grad_norm": 2.21223521232605, + "learning_rate": 0.0001240732077124098, + "loss": 0.1676, + "step": 18034 + }, + { + "epoch": 0.42276880160809205, + "grad_norm": 2.5296239852905273, + "learning_rate": 0.00012406605811614017, + "loss": 0.1523, + "step": 18035 + }, + { + "epoch": 0.42279224318289704, + "grad_norm": 2.2984488010406494, + "learning_rate": 0.00012405890838928745, + "loss": 0.1745, + "step": 18036 + }, + { + "epoch": 0.42281568475770204, + "grad_norm": 2.91275954246521, + "learning_rate": 0.00012405175853189042, + "loss": 0.3023, + "step": 18037 + }, + { + "epoch": 0.42283912633250703, + "grad_norm": 4.148429870605469, + "learning_rate": 0.00012404460854398785, + "loss": 0.5779, + "step": 18038 + }, + { + "epoch": 0.422862567907312, + "grad_norm": 2.7311034202575684, + "learning_rate": 0.00012403745842561858, + "loss": 1.7107, + "step": 18039 + }, + { + "epoch": 0.422886009482117, + "grad_norm": 3.604283332824707, + "learning_rate": 0.00012403030817682135, + "loss": 0.3424, + "step": 18040 + }, + { + "epoch": 0.422909451056922, + "grad_norm": 0.8809439539909363, + "learning_rate": 0.00012402315779763503, + "loss": 0.1113, + "step": 18041 + }, + { + "epoch": 0.422932892631727, + "grad_norm": 3.241302967071533, + "learning_rate": 0.00012401600728809836, + "loss": 0.4232, + "step": 18042 + }, + { + "epoch": 0.422956334206532, + "grad_norm": 3.832090139389038, + "learning_rate": 0.00012400885664825016, + "loss": 0.5883, + "step": 18043 + }, + { + "epoch": 0.422979775781337, + "grad_norm": 2.7471401691436768, + "learning_rate": 0.00012400170587812922, + "loss": 0.2079, + "step": 18044 + }, + { + "epoch": 0.423003217356142, + "grad_norm": 1.9737136363983154, + "learning_rate": 0.00012399455497777437, + "loss": 1.6936, + "step": 18045 + }, + { + "epoch": 0.423026658930947, + "grad_norm": 5.202999591827393, + "learning_rate": 0.0001239874039472244, + "loss": 0.6034, + "step": 18046 + }, + { + "epoch": 0.423050100505752, + "grad_norm": 3.6742756366729736, + "learning_rate": 0.00012398025278651805, + "loss": 0.4092, + "step": 18047 + }, + { + "epoch": 0.423073542080557, + "grad_norm": 4.245169162750244, + "learning_rate": 0.00012397310149569424, + "loss": 0.4189, + "step": 18048 + }, + { + "epoch": 0.42309698365536197, + "grad_norm": 4.286130428314209, + "learning_rate": 0.00012396595007479165, + "loss": 0.3812, + "step": 18049 + }, + { + "epoch": 0.42312042523016696, + "grad_norm": 3.524796724319458, + "learning_rate": 0.00012395879852384917, + "loss": 0.3436, + "step": 18050 + }, + { + "epoch": 0.42314386680497196, + "grad_norm": 1.1495437622070312, + "learning_rate": 0.00012395164684290557, + "loss": 0.1372, + "step": 18051 + }, + { + "epoch": 0.42316730837977695, + "grad_norm": 2.264280080795288, + "learning_rate": 0.00012394449503199968, + "loss": 0.1669, + "step": 18052 + }, + { + "epoch": 0.42319074995458195, + "grad_norm": 5.170156478881836, + "learning_rate": 0.00012393734309117024, + "loss": 0.3623, + "step": 18053 + }, + { + "epoch": 0.42321419152938694, + "grad_norm": 3.9896349906921387, + "learning_rate": 0.00012393019102045617, + "loss": 0.2614, + "step": 18054 + }, + { + "epoch": 0.42323763310419193, + "grad_norm": 4.5174784660339355, + "learning_rate": 0.00012392303881989615, + "loss": 0.6792, + "step": 18055 + }, + { + "epoch": 0.42326107467899693, + "grad_norm": 4.498547077178955, + "learning_rate": 0.00012391588648952908, + "loss": 0.4864, + "step": 18056 + }, + { + "epoch": 0.4232845162538019, + "grad_norm": 2.883627414703369, + "learning_rate": 0.0001239087340293937, + "loss": 0.1761, + "step": 18057 + }, + { + "epoch": 0.4233079578286069, + "grad_norm": 1.94358491897583, + "learning_rate": 0.00012390158143952892, + "loss": 1.6826, + "step": 18058 + }, + { + "epoch": 0.4233313994034119, + "grad_norm": 1.274976134300232, + "learning_rate": 0.00012389442871997344, + "loss": 0.146, + "step": 18059 + }, + { + "epoch": 0.4233548409782169, + "grad_norm": 3.6173417568206787, + "learning_rate": 0.00012388727587076613, + "loss": 0.1575, + "step": 18060 + }, + { + "epoch": 0.4233782825530219, + "grad_norm": 1.6583749055862427, + "learning_rate": 0.00012388012289194582, + "loss": 0.2323, + "step": 18061 + }, + { + "epoch": 0.4234017241278269, + "grad_norm": 3.9420688152313232, + "learning_rate": 0.00012387296978355123, + "loss": 0.5083, + "step": 18062 + }, + { + "epoch": 0.4234251657026319, + "grad_norm": 5.421494483947754, + "learning_rate": 0.00012386581654562126, + "loss": 0.4726, + "step": 18063 + }, + { + "epoch": 0.4234486072774369, + "grad_norm": 2.054076910018921, + "learning_rate": 0.00012385866317819472, + "loss": 0.2337, + "step": 18064 + }, + { + "epoch": 0.4234720488522419, + "grad_norm": 3.278468608856201, + "learning_rate": 0.0001238515096813104, + "loss": 0.2156, + "step": 18065 + }, + { + "epoch": 0.4234954904270469, + "grad_norm": 3.418302536010742, + "learning_rate": 0.00012384435605500708, + "loss": 0.4713, + "step": 18066 + }, + { + "epoch": 0.42351893200185187, + "grad_norm": 4.834647178649902, + "learning_rate": 0.00012383720229932368, + "loss": 0.6141, + "step": 18067 + }, + { + "epoch": 0.42354237357665686, + "grad_norm": 1.9375749826431274, + "learning_rate": 0.0001238300484142989, + "loss": 0.1459, + "step": 18068 + }, + { + "epoch": 0.42356581515146186, + "grad_norm": 2.42964243888855, + "learning_rate": 0.0001238228943999716, + "loss": 0.273, + "step": 18069 + }, + { + "epoch": 0.42358925672626685, + "grad_norm": 3.638524055480957, + "learning_rate": 0.0001238157402563806, + "loss": 0.412, + "step": 18070 + }, + { + "epoch": 0.42361269830107184, + "grad_norm": 2.7674174308776855, + "learning_rate": 0.00012380858598356478, + "loss": 0.215, + "step": 18071 + }, + { + "epoch": 0.42363613987587684, + "grad_norm": 3.5612857341766357, + "learning_rate": 0.00012380143158156285, + "loss": 0.3609, + "step": 18072 + }, + { + "epoch": 0.42365958145068183, + "grad_norm": 3.9421095848083496, + "learning_rate": 0.0001237942770504137, + "loss": 0.3533, + "step": 18073 + }, + { + "epoch": 0.4236830230254868, + "grad_norm": 2.0381720066070557, + "learning_rate": 0.00012378712239015613, + "loss": 0.0883, + "step": 18074 + }, + { + "epoch": 0.4237064646002918, + "grad_norm": 1.9500175714492798, + "learning_rate": 0.000123779967600829, + "loss": 0.2227, + "step": 18075 + }, + { + "epoch": 0.4237299061750968, + "grad_norm": 2.7496156692504883, + "learning_rate": 0.00012377281268247106, + "loss": 0.3212, + "step": 18076 + }, + { + "epoch": 0.42375334774990187, + "grad_norm": 3.989753007888794, + "learning_rate": 0.00012376565763512116, + "loss": 0.4164, + "step": 18077 + }, + { + "epoch": 0.42377678932470686, + "grad_norm": 0.34382057189941406, + "learning_rate": 0.00012375850245881814, + "loss": 0.0296, + "step": 18078 + }, + { + "epoch": 0.42380023089951185, + "grad_norm": 5.0921549797058105, + "learning_rate": 0.00012375134715360083, + "loss": 0.4736, + "step": 18079 + }, + { + "epoch": 0.42382367247431685, + "grad_norm": 3.2680349349975586, + "learning_rate": 0.00012374419171950804, + "loss": 0.6057, + "step": 18080 + }, + { + "epoch": 0.42384711404912184, + "grad_norm": 3.6913349628448486, + "learning_rate": 0.0001237370361565786, + "loss": 0.4679, + "step": 18081 + }, + { + "epoch": 0.42387055562392684, + "grad_norm": 3.2923483848571777, + "learning_rate": 0.00012372988046485136, + "loss": 0.2899, + "step": 18082 + }, + { + "epoch": 0.42389399719873183, + "grad_norm": 2.678734302520752, + "learning_rate": 0.00012372272464436507, + "loss": 0.2926, + "step": 18083 + }, + { + "epoch": 0.4239174387735368, + "grad_norm": 5.427066802978516, + "learning_rate": 0.00012371556869515866, + "loss": 0.578, + "step": 18084 + }, + { + "epoch": 0.4239408803483418, + "grad_norm": 1.894752860069275, + "learning_rate": 0.00012370841261727085, + "loss": 1.5894, + "step": 18085 + }, + { + "epoch": 0.4239643219231468, + "grad_norm": 2.096813678741455, + "learning_rate": 0.00012370125641074058, + "loss": 0.0782, + "step": 18086 + }, + { + "epoch": 0.4239877634979518, + "grad_norm": 4.261784553527832, + "learning_rate": 0.00012369410007560661, + "loss": 0.4125, + "step": 18087 + }, + { + "epoch": 0.4240112050727568, + "grad_norm": 0.7976709604263306, + "learning_rate": 0.00012368694361190776, + "loss": 0.0995, + "step": 18088 + }, + { + "epoch": 0.4240346466475618, + "grad_norm": 2.7371985912323, + "learning_rate": 0.00012367978701968292, + "loss": 0.2991, + "step": 18089 + }, + { + "epoch": 0.4240580882223668, + "grad_norm": 4.327144622802734, + "learning_rate": 0.00012367263029897088, + "loss": 0.3965, + "step": 18090 + }, + { + "epoch": 0.4240815297971718, + "grad_norm": 5.345952033996582, + "learning_rate": 0.0001236654734498105, + "loss": 0.7823, + "step": 18091 + }, + { + "epoch": 0.4241049713719768, + "grad_norm": 7.252579689025879, + "learning_rate": 0.00012365831647224054, + "loss": 0.2898, + "step": 18092 + }, + { + "epoch": 0.4241284129467818, + "grad_norm": 5.225094318389893, + "learning_rate": 0.00012365115936629993, + "loss": 0.5907, + "step": 18093 + }, + { + "epoch": 0.42415185452158677, + "grad_norm": 5.535244464874268, + "learning_rate": 0.00012364400213202748, + "loss": 0.8519, + "step": 18094 + }, + { + "epoch": 0.42417529609639176, + "grad_norm": 1.9649163484573364, + "learning_rate": 0.00012363684476946198, + "loss": 0.1991, + "step": 18095 + }, + { + "epoch": 0.42419873767119676, + "grad_norm": 4.058174133300781, + "learning_rate": 0.0001236296872786423, + "loss": 0.39, + "step": 18096 + }, + { + "epoch": 0.42422217924600175, + "grad_norm": 3.44744610786438, + "learning_rate": 0.0001236225296596073, + "loss": 0.4512, + "step": 18097 + }, + { + "epoch": 0.42424562082080675, + "grad_norm": 1.599999189376831, + "learning_rate": 0.00012361537191239575, + "loss": 0.1931, + "step": 18098 + }, + { + "epoch": 0.42426906239561174, + "grad_norm": 1.7734084129333496, + "learning_rate": 0.00012360821403704654, + "loss": 0.2093, + "step": 18099 + }, + { + "epoch": 0.42429250397041673, + "grad_norm": 2.0895986557006836, + "learning_rate": 0.0001236010560335985, + "loss": 1.3552, + "step": 18100 + }, + { + "epoch": 0.42431594554522173, + "grad_norm": 1.295044183731079, + "learning_rate": 0.00012359389790209048, + "loss": 0.1573, + "step": 18101 + }, + { + "epoch": 0.4243393871200267, + "grad_norm": 0.8882340788841248, + "learning_rate": 0.0001235867396425613, + "loss": 0.096, + "step": 18102 + }, + { + "epoch": 0.4243628286948317, + "grad_norm": 4.160126686096191, + "learning_rate": 0.0001235795812550498, + "loss": 0.4609, + "step": 18103 + }, + { + "epoch": 0.4243862702696367, + "grad_norm": 0.8552916646003723, + "learning_rate": 0.0001235724227395948, + "loss": 0.0861, + "step": 18104 + }, + { + "epoch": 0.4244097118444417, + "grad_norm": 4.4021220207214355, + "learning_rate": 0.00012356526409623524, + "loss": 0.3311, + "step": 18105 + }, + { + "epoch": 0.4244331534192467, + "grad_norm": 3.9423396587371826, + "learning_rate": 0.00012355810532500987, + "loss": 0.4043, + "step": 18106 + }, + { + "epoch": 0.4244565949940517, + "grad_norm": 3.5599260330200195, + "learning_rate": 0.00012355094642595754, + "loss": 0.3669, + "step": 18107 + }, + { + "epoch": 0.4244800365688567, + "grad_norm": 3.326505422592163, + "learning_rate": 0.00012354378739911712, + "loss": 0.3637, + "step": 18108 + }, + { + "epoch": 0.4245034781436617, + "grad_norm": 5.8533935546875, + "learning_rate": 0.00012353662824452746, + "loss": 0.4225, + "step": 18109 + }, + { + "epoch": 0.4245269197184667, + "grad_norm": 3.655693531036377, + "learning_rate": 0.00012352946896222736, + "loss": 0.3662, + "step": 18110 + }, + { + "epoch": 0.42455036129327167, + "grad_norm": 3.599186420440674, + "learning_rate": 0.0001235223095522557, + "loss": 0.3314, + "step": 18111 + }, + { + "epoch": 0.42457380286807667, + "grad_norm": 4.023182392120361, + "learning_rate": 0.00012351515001465138, + "loss": 0.2714, + "step": 18112 + }, + { + "epoch": 0.42459724444288166, + "grad_norm": 4.506155490875244, + "learning_rate": 0.00012350799034945316, + "loss": 0.3768, + "step": 18113 + }, + { + "epoch": 0.42462068601768665, + "grad_norm": 3.459442377090454, + "learning_rate": 0.0001235008305566999, + "loss": 0.2254, + "step": 18114 + }, + { + "epoch": 0.42464412759249165, + "grad_norm": 6.0961809158325195, + "learning_rate": 0.00012349367063643052, + "loss": 0.6175, + "step": 18115 + }, + { + "epoch": 0.42466756916729664, + "grad_norm": 1.4375083446502686, + "learning_rate": 0.0001234865105886838, + "loss": 0.1485, + "step": 18116 + }, + { + "epoch": 0.42469101074210164, + "grad_norm": 4.5815749168396, + "learning_rate": 0.00012347935041349858, + "loss": 0.5252, + "step": 18117 + }, + { + "epoch": 0.42471445231690663, + "grad_norm": 2.185774087905884, + "learning_rate": 0.0001234721901109138, + "loss": 1.5075, + "step": 18118 + }, + { + "epoch": 0.4247378938917116, + "grad_norm": 2.5659024715423584, + "learning_rate": 0.00012346502968096825, + "loss": 0.1861, + "step": 18119 + }, + { + "epoch": 0.4247613354665166, + "grad_norm": 1.3649473190307617, + "learning_rate": 0.00012345786912370074, + "loss": 0.1224, + "step": 18120 + }, + { + "epoch": 0.4247847770413216, + "grad_norm": 1.3011131286621094, + "learning_rate": 0.00012345070843915024, + "loss": 0.0895, + "step": 18121 + }, + { + "epoch": 0.4248082186161266, + "grad_norm": 2.02307391166687, + "learning_rate": 0.00012344354762735548, + "loss": 0.112, + "step": 18122 + }, + { + "epoch": 0.4248316601909316, + "grad_norm": 2.870814323425293, + "learning_rate": 0.00012343638668835537, + "loss": 0.1675, + "step": 18123 + }, + { + "epoch": 0.4248551017657366, + "grad_norm": 3.3489584922790527, + "learning_rate": 0.0001234292256221888, + "loss": 0.4376, + "step": 18124 + }, + { + "epoch": 0.4248785433405416, + "grad_norm": 3.852970600128174, + "learning_rate": 0.0001234220644288946, + "loss": 0.4518, + "step": 18125 + }, + { + "epoch": 0.4249019849153466, + "grad_norm": 4.537295341491699, + "learning_rate": 0.0001234149031085116, + "loss": 0.4674, + "step": 18126 + }, + { + "epoch": 0.4249254264901516, + "grad_norm": 3.072645902633667, + "learning_rate": 0.00012340774166107868, + "loss": 0.3881, + "step": 18127 + }, + { + "epoch": 0.42494886806495663, + "grad_norm": 4.0060248374938965, + "learning_rate": 0.0001234005800866347, + "loss": 0.1164, + "step": 18128 + }, + { + "epoch": 0.4249723096397616, + "grad_norm": 3.9928340911865234, + "learning_rate": 0.00012339341838521852, + "loss": 0.2382, + "step": 18129 + }, + { + "epoch": 0.4249957512145666, + "grad_norm": 3.554635524749756, + "learning_rate": 0.00012338625655686897, + "loss": 0.3823, + "step": 18130 + }, + { + "epoch": 0.4250191927893716, + "grad_norm": 2.9902591705322266, + "learning_rate": 0.00012337909460162497, + "loss": 0.5371, + "step": 18131 + }, + { + "epoch": 0.4250426343641766, + "grad_norm": 1.831616997718811, + "learning_rate": 0.00012337193251952533, + "loss": 1.2137, + "step": 18132 + }, + { + "epoch": 0.4250660759389816, + "grad_norm": 3.8872673511505127, + "learning_rate": 0.00012336477031060889, + "loss": 0.2914, + "step": 18133 + }, + { + "epoch": 0.4250895175137866, + "grad_norm": 1.6049622297286987, + "learning_rate": 0.0001233576079749146, + "loss": 0.1882, + "step": 18134 + }, + { + "epoch": 0.4251129590885916, + "grad_norm": 4.541254997253418, + "learning_rate": 0.00012335044551248125, + "loss": 0.4552, + "step": 18135 + }, + { + "epoch": 0.4251364006633966, + "grad_norm": 4.128963470458984, + "learning_rate": 0.00012334328292334771, + "loss": 1.557, + "step": 18136 + }, + { + "epoch": 0.4251598422382016, + "grad_norm": 2.5839006900787354, + "learning_rate": 0.0001233361202075529, + "loss": 1.4966, + "step": 18137 + }, + { + "epoch": 0.4251832838130066, + "grad_norm": 2.755553722381592, + "learning_rate": 0.00012332895736513562, + "loss": 0.1375, + "step": 18138 + }, + { + "epoch": 0.42520672538781157, + "grad_norm": 5.315330505371094, + "learning_rate": 0.00012332179439613475, + "loss": 0.5473, + "step": 18139 + }, + { + "epoch": 0.42523016696261656, + "grad_norm": 3.875744104385376, + "learning_rate": 0.0001233146313005892, + "loss": 0.312, + "step": 18140 + }, + { + "epoch": 0.42525360853742156, + "grad_norm": 3.384654998779297, + "learning_rate": 0.00012330746807853778, + "loss": 0.2974, + "step": 18141 + }, + { + "epoch": 0.42527705011222655, + "grad_norm": 3.0550687313079834, + "learning_rate": 0.0001233003047300194, + "loss": 0.3787, + "step": 18142 + }, + { + "epoch": 0.42530049168703155, + "grad_norm": 4.360583782196045, + "learning_rate": 0.0001232931412550729, + "loss": 0.4623, + "step": 18143 + }, + { + "epoch": 0.42532393326183654, + "grad_norm": 3.144019842147827, + "learning_rate": 0.00012328597765373718, + "loss": 0.3288, + "step": 18144 + }, + { + "epoch": 0.42534737483664153, + "grad_norm": 0.548405647277832, + "learning_rate": 0.00012327881392605106, + "loss": 0.0307, + "step": 18145 + }, + { + "epoch": 0.42537081641144653, + "grad_norm": 1.2034538984298706, + "learning_rate": 0.00012327165007205347, + "loss": 0.0777, + "step": 18146 + }, + { + "epoch": 0.4253942579862515, + "grad_norm": 6.261703968048096, + "learning_rate": 0.00012326448609178322, + "loss": 0.6333, + "step": 18147 + }, + { + "epoch": 0.4254176995610565, + "grad_norm": 0.5599477887153625, + "learning_rate": 0.00012325732198527925, + "loss": 0.0398, + "step": 18148 + }, + { + "epoch": 0.4254411411358615, + "grad_norm": 3.6282596588134766, + "learning_rate": 0.00012325015775258038, + "loss": 0.3394, + "step": 18149 + }, + { + "epoch": 0.4254645827106665, + "grad_norm": 2.9917142391204834, + "learning_rate": 0.00012324299339372548, + "loss": 0.227, + "step": 18150 + }, + { + "epoch": 0.4254880242854715, + "grad_norm": 7.306393146514893, + "learning_rate": 0.00012323582890875348, + "loss": 0.5475, + "step": 18151 + }, + { + "epoch": 0.4255114658602765, + "grad_norm": 1.3976445198059082, + "learning_rate": 0.0001232286642977032, + "loss": 0.1145, + "step": 18152 + }, + { + "epoch": 0.4255349074350815, + "grad_norm": 1.4791333675384521, + "learning_rate": 0.00012322149956061352, + "loss": 0.2221, + "step": 18153 + }, + { + "epoch": 0.4255583490098865, + "grad_norm": 2.627891778945923, + "learning_rate": 0.00012321433469752338, + "loss": 0.3046, + "step": 18154 + }, + { + "epoch": 0.4255817905846915, + "grad_norm": 3.7188236713409424, + "learning_rate": 0.00012320716970847157, + "loss": 0.4307, + "step": 18155 + }, + { + "epoch": 0.42560523215949647, + "grad_norm": 1.1962133646011353, + "learning_rate": 0.000123200004593497, + "loss": 0.1131, + "step": 18156 + }, + { + "epoch": 0.42562867373430147, + "grad_norm": 2.2152903079986572, + "learning_rate": 0.00012319283935263858, + "loss": 0.8527, + "step": 18157 + }, + { + "epoch": 0.42565211530910646, + "grad_norm": 4.162498950958252, + "learning_rate": 0.00012318567398593512, + "loss": 0.5775, + "step": 18158 + }, + { + "epoch": 0.42567555688391145, + "grad_norm": 1.3942327499389648, + "learning_rate": 0.00012317850849342555, + "loss": 0.1173, + "step": 18159 + }, + { + "epoch": 0.42569899845871645, + "grad_norm": 2.9540252685546875, + "learning_rate": 0.00012317134287514878, + "loss": 0.193, + "step": 18160 + }, + { + "epoch": 0.42572244003352144, + "grad_norm": 4.079094886779785, + "learning_rate": 0.00012316417713114364, + "loss": 0.3407, + "step": 18161 + }, + { + "epoch": 0.42574588160832644, + "grad_norm": 3.927713632583618, + "learning_rate": 0.000123157011261449, + "loss": 0.5614, + "step": 18162 + }, + { + "epoch": 0.42576932318313143, + "grad_norm": 4.19798469543457, + "learning_rate": 0.0001231498452661038, + "loss": 0.3999, + "step": 18163 + }, + { + "epoch": 0.4257927647579364, + "grad_norm": 2.631368398666382, + "learning_rate": 0.00012314267914514685, + "loss": 0.2517, + "step": 18164 + }, + { + "epoch": 0.4258162063327414, + "grad_norm": 5.2871012687683105, + "learning_rate": 0.0001231355128986171, + "loss": 0.5825, + "step": 18165 + }, + { + "epoch": 0.4258396479075464, + "grad_norm": 3.189349412918091, + "learning_rate": 0.00012312834652655337, + "loss": 0.2798, + "step": 18166 + }, + { + "epoch": 0.4258630894823514, + "grad_norm": 3.542132616043091, + "learning_rate": 0.00012312118002899463, + "loss": 0.4242, + "step": 18167 + }, + { + "epoch": 0.4258865310571564, + "grad_norm": 2.2023112773895264, + "learning_rate": 0.00012311401340597968, + "loss": 0.1526, + "step": 18168 + }, + { + "epoch": 0.4259099726319614, + "grad_norm": 4.746866703033447, + "learning_rate": 0.00012310684665754748, + "loss": 0.2644, + "step": 18169 + }, + { + "epoch": 0.4259334142067664, + "grad_norm": 1.6670855283737183, + "learning_rate": 0.00012309967978373686, + "loss": 0.1526, + "step": 18170 + }, + { + "epoch": 0.4259568557815714, + "grad_norm": 4.0198540687561035, + "learning_rate": 0.00012309251278458668, + "loss": 0.3948, + "step": 18171 + }, + { + "epoch": 0.4259802973563764, + "grad_norm": 2.41945743560791, + "learning_rate": 0.00012308534566013595, + "loss": 0.3375, + "step": 18172 + }, + { + "epoch": 0.4260037389311814, + "grad_norm": 1.974199652671814, + "learning_rate": 0.00012307817841042347, + "loss": 0.2594, + "step": 18173 + }, + { + "epoch": 0.42602718050598637, + "grad_norm": 4.879997730255127, + "learning_rate": 0.0001230710110354881, + "loss": 0.5594, + "step": 18174 + }, + { + "epoch": 0.42605062208079136, + "grad_norm": 2.145493507385254, + "learning_rate": 0.00012306384353536883, + "loss": 1.3422, + "step": 18175 + }, + { + "epoch": 0.42607406365559636, + "grad_norm": 4.106379985809326, + "learning_rate": 0.0001230566759101045, + "loss": 0.3241, + "step": 18176 + }, + { + "epoch": 0.42609750523040135, + "grad_norm": 5.929948329925537, + "learning_rate": 0.00012304950815973393, + "loss": 0.5254, + "step": 18177 + }, + { + "epoch": 0.42612094680520635, + "grad_norm": 2.8371975421905518, + "learning_rate": 0.00012304234028429612, + "loss": 0.225, + "step": 18178 + }, + { + "epoch": 0.4261443883800114, + "grad_norm": 3.935616970062256, + "learning_rate": 0.00012303517228382996, + "loss": 0.2893, + "step": 18179 + }, + { + "epoch": 0.4261678299548164, + "grad_norm": 4.830079078674316, + "learning_rate": 0.00012302800415837426, + "loss": 0.5642, + "step": 18180 + }, + { + "epoch": 0.4261912715296214, + "grad_norm": 2.1469433307647705, + "learning_rate": 0.00012302083590796798, + "loss": 0.2042, + "step": 18181 + }, + { + "epoch": 0.4262147131044264, + "grad_norm": 3.910571813583374, + "learning_rate": 0.00012301366753265, + "loss": 0.4544, + "step": 18182 + }, + { + "epoch": 0.4262381546792314, + "grad_norm": 1.4388225078582764, + "learning_rate": 0.0001230064990324592, + "loss": 0.1576, + "step": 18183 + }, + { + "epoch": 0.42626159625403637, + "grad_norm": 4.602324485778809, + "learning_rate": 0.00012299933040743448, + "loss": 0.3831, + "step": 18184 + }, + { + "epoch": 0.42628503782884136, + "grad_norm": 0.5152525901794434, + "learning_rate": 0.00012299216165761478, + "loss": 0.0774, + "step": 18185 + }, + { + "epoch": 0.42630847940364636, + "grad_norm": 2.8686554431915283, + "learning_rate": 0.00012298499278303894, + "loss": 1.2248, + "step": 18186 + }, + { + "epoch": 0.42633192097845135, + "grad_norm": 2.4114015102386475, + "learning_rate": 0.0001229778237837459, + "loss": 0.1723, + "step": 18187 + }, + { + "epoch": 0.42635536255325635, + "grad_norm": 2.416724443435669, + "learning_rate": 0.00012297065465977452, + "loss": 1.9276, + "step": 18188 + }, + { + "epoch": 0.42637880412806134, + "grad_norm": 0.743619978427887, + "learning_rate": 0.00012296348541116373, + "loss": 0.0477, + "step": 18189 + }, + { + "epoch": 0.42640224570286633, + "grad_norm": 2.2871615886688232, + "learning_rate": 0.00012295631603795242, + "loss": 0.3399, + "step": 18190 + }, + { + "epoch": 0.42642568727767133, + "grad_norm": 3.862112045288086, + "learning_rate": 0.0001229491465401795, + "loss": 0.5303, + "step": 18191 + }, + { + "epoch": 0.4264491288524763, + "grad_norm": 2.8853180408477783, + "learning_rate": 0.00012294197691788388, + "loss": 0.377, + "step": 18192 + }, + { + "epoch": 0.4264725704272813, + "grad_norm": 1.2604680061340332, + "learning_rate": 0.00012293480717110442, + "loss": 0.6077, + "step": 18193 + }, + { + "epoch": 0.4264960120020863, + "grad_norm": 4.690886497497559, + "learning_rate": 0.00012292763729988007, + "loss": 0.5581, + "step": 18194 + }, + { + "epoch": 0.4265194535768913, + "grad_norm": 2.1774957180023193, + "learning_rate": 0.0001229204673042497, + "loss": 1.5327, + "step": 18195 + }, + { + "epoch": 0.4265428951516963, + "grad_norm": 1.0975710153579712, + "learning_rate": 0.00012291329718425224, + "loss": 0.1277, + "step": 18196 + }, + { + "epoch": 0.4265663367265013, + "grad_norm": 3.0293266773223877, + "learning_rate": 0.0001229061269399266, + "loss": 0.2442, + "step": 18197 + }, + { + "epoch": 0.4265897783013063, + "grad_norm": 4.090175151824951, + "learning_rate": 0.00012289895657131166, + "loss": 0.2745, + "step": 18198 + }, + { + "epoch": 0.4266132198761113, + "grad_norm": 1.0426205396652222, + "learning_rate": 0.00012289178607844634, + "loss": 0.0403, + "step": 18199 + }, + { + "epoch": 0.4266366614509163, + "grad_norm": 4.386704921722412, + "learning_rate": 0.00012288461546136953, + "loss": 0.486, + "step": 18200 + }, + { + "epoch": 0.42666010302572127, + "grad_norm": 1.5333822965621948, + "learning_rate": 0.00012287744472012015, + "loss": 0.1393, + "step": 18201 + }, + { + "epoch": 0.42668354460052627, + "grad_norm": 0.9842867851257324, + "learning_rate": 0.00012287027385473714, + "loss": 0.1284, + "step": 18202 + }, + { + "epoch": 0.42670698617533126, + "grad_norm": 2.977766513824463, + "learning_rate": 0.00012286310286525935, + "loss": 0.2177, + "step": 18203 + }, + { + "epoch": 0.42673042775013625, + "grad_norm": 1.0440140962600708, + "learning_rate": 0.00012285593175172577, + "loss": 0.1342, + "step": 18204 + }, + { + "epoch": 0.42675386932494125, + "grad_norm": 2.1554622650146484, + "learning_rate": 0.0001228487605141752, + "loss": 0.1874, + "step": 18205 + }, + { + "epoch": 0.42677731089974624, + "grad_norm": 5.030059337615967, + "learning_rate": 0.00012284158915264664, + "loss": 0.6339, + "step": 18206 + }, + { + "epoch": 0.42680075247455124, + "grad_norm": 2.600193500518799, + "learning_rate": 0.00012283441766717897, + "loss": 0.2383, + "step": 18207 + }, + { + "epoch": 0.42682419404935623, + "grad_norm": 2.8554437160491943, + "learning_rate": 0.00012282724605781114, + "loss": 0.3395, + "step": 18208 + }, + { + "epoch": 0.4268476356241612, + "grad_norm": 2.5921597480773926, + "learning_rate": 0.000122820074324582, + "loss": 0.2869, + "step": 18209 + }, + { + "epoch": 0.4268710771989662, + "grad_norm": 1.4664677381515503, + "learning_rate": 0.0001228129024675305, + "loss": 0.1861, + "step": 18210 + }, + { + "epoch": 0.4268945187737712, + "grad_norm": 4.621686935424805, + "learning_rate": 0.00012280573048669555, + "loss": 0.3011, + "step": 18211 + }, + { + "epoch": 0.4269179603485762, + "grad_norm": 5.107275485992432, + "learning_rate": 0.00012279855838211604, + "loss": 0.3177, + "step": 18212 + }, + { + "epoch": 0.4269414019233812, + "grad_norm": 0.7699370384216309, + "learning_rate": 0.00012279138615383097, + "loss": 0.0924, + "step": 18213 + }, + { + "epoch": 0.4269648434981862, + "grad_norm": 4.215707778930664, + "learning_rate": 0.00012278421380187914, + "loss": 0.5454, + "step": 18214 + }, + { + "epoch": 0.4269882850729912, + "grad_norm": 2.8097803592681885, + "learning_rate": 0.00012277704132629955, + "loss": 0.2075, + "step": 18215 + }, + { + "epoch": 0.4270117266477962, + "grad_norm": 2.8768763542175293, + "learning_rate": 0.0001227698687271311, + "loss": 0.4074, + "step": 18216 + }, + { + "epoch": 0.4270351682226012, + "grad_norm": 0.767684817314148, + "learning_rate": 0.00012276269600441271, + "loss": 0.0758, + "step": 18217 + }, + { + "epoch": 0.4270586097974062, + "grad_norm": 2.0787856578826904, + "learning_rate": 0.00012275552315818325, + "loss": 0.1514, + "step": 18218 + }, + { + "epoch": 0.42708205137221117, + "grad_norm": 2.103219985961914, + "learning_rate": 0.00012274835018848173, + "loss": 0.2893, + "step": 18219 + }, + { + "epoch": 0.42710549294701616, + "grad_norm": 2.7818617820739746, + "learning_rate": 0.00012274117709534698, + "loss": 0.286, + "step": 18220 + }, + { + "epoch": 0.42712893452182116, + "grad_norm": 3.3887834548950195, + "learning_rate": 0.000122734003878818, + "loss": 0.2849, + "step": 18221 + }, + { + "epoch": 0.42715237609662615, + "grad_norm": 0.3164624571800232, + "learning_rate": 0.00012272683053893364, + "loss": 0.0253, + "step": 18222 + }, + { + "epoch": 0.42717581767143115, + "grad_norm": 3.0909855365753174, + "learning_rate": 0.00012271965707573288, + "loss": 0.2008, + "step": 18223 + }, + { + "epoch": 0.42719925924623614, + "grad_norm": 4.598025321960449, + "learning_rate": 0.00012271248348925463, + "loss": 0.3478, + "step": 18224 + }, + { + "epoch": 0.42722270082104113, + "grad_norm": 2.081747531890869, + "learning_rate": 0.00012270530977953777, + "loss": 0.1352, + "step": 18225 + }, + { + "epoch": 0.42724614239584613, + "grad_norm": 2.401326894760132, + "learning_rate": 0.0001226981359466213, + "loss": 0.3556, + "step": 18226 + }, + { + "epoch": 0.4272695839706511, + "grad_norm": 2.6471457481384277, + "learning_rate": 0.0001226909619905441, + "loss": 1.5132, + "step": 18227 + }, + { + "epoch": 0.4272930255454561, + "grad_norm": 0.8884686827659607, + "learning_rate": 0.00012268378791134506, + "loss": 0.0872, + "step": 18228 + }, + { + "epoch": 0.4273164671202611, + "grad_norm": 3.217757225036621, + "learning_rate": 0.0001226766137090632, + "loss": 1.2407, + "step": 18229 + }, + { + "epoch": 0.42733990869506616, + "grad_norm": 3.276296377182007, + "learning_rate": 0.0001226694393837374, + "loss": 0.2617, + "step": 18230 + }, + { + "epoch": 0.42736335026987116, + "grad_norm": 3.803891658782959, + "learning_rate": 0.00012266226493540657, + "loss": 0.2974, + "step": 18231 + }, + { + "epoch": 0.42738679184467615, + "grad_norm": 0.23967301845550537, + "learning_rate": 0.00012265509036410964, + "loss": 0.0128, + "step": 18232 + }, + { + "epoch": 0.42741023341948114, + "grad_norm": 1.5530425310134888, + "learning_rate": 0.00012264791566988557, + "loss": 0.1593, + "step": 18233 + }, + { + "epoch": 0.42743367499428614, + "grad_norm": 1.7892605066299438, + "learning_rate": 0.00012264074085277323, + "loss": 0.1811, + "step": 18234 + }, + { + "epoch": 0.42745711656909113, + "grad_norm": 1.9213838577270508, + "learning_rate": 0.00012263356591281162, + "loss": 0.1434, + "step": 18235 + }, + { + "epoch": 0.4274805581438961, + "grad_norm": 3.663468599319458, + "learning_rate": 0.00012262639085003967, + "loss": 0.2842, + "step": 18236 + }, + { + "epoch": 0.4275039997187011, + "grad_norm": 4.589759349822998, + "learning_rate": 0.00012261921566449624, + "loss": 0.4491, + "step": 18237 + }, + { + "epoch": 0.4275274412935061, + "grad_norm": 1.6022891998291016, + "learning_rate": 0.00012261204035622037, + "loss": 0.2275, + "step": 18238 + }, + { + "epoch": 0.4275508828683111, + "grad_norm": 1.5780807733535767, + "learning_rate": 0.0001226048649252509, + "loss": 0.1987, + "step": 18239 + }, + { + "epoch": 0.4275743244431161, + "grad_norm": 3.2684059143066406, + "learning_rate": 0.00012259768937162682, + "loss": 1.8582, + "step": 18240 + }, + { + "epoch": 0.4275977660179211, + "grad_norm": 3.8694753646850586, + "learning_rate": 0.00012259051369538699, + "loss": 0.4516, + "step": 18241 + }, + { + "epoch": 0.4276212075927261, + "grad_norm": 2.77850079536438, + "learning_rate": 0.00012258333789657043, + "loss": 1.5142, + "step": 18242 + }, + { + "epoch": 0.4276446491675311, + "grad_norm": 2.9154129028320312, + "learning_rate": 0.00012257616197521606, + "loss": 0.1665, + "step": 18243 + }, + { + "epoch": 0.4276680907423361, + "grad_norm": 0.62068772315979, + "learning_rate": 0.00012256898593136274, + "loss": 0.0322, + "step": 18244 + }, + { + "epoch": 0.4276915323171411, + "grad_norm": 2.5485782623291016, + "learning_rate": 0.00012256180976504954, + "loss": 0.2424, + "step": 18245 + }, + { + "epoch": 0.42771497389194607, + "grad_norm": 4.383578300476074, + "learning_rate": 0.0001225546334763153, + "loss": 0.2657, + "step": 18246 + }, + { + "epoch": 0.42773841546675107, + "grad_norm": 3.5347700119018555, + "learning_rate": 0.00012254745706519898, + "loss": 0.498, + "step": 18247 + }, + { + "epoch": 0.42776185704155606, + "grad_norm": 2.844771385192871, + "learning_rate": 0.00012254028053173952, + "loss": 0.3043, + "step": 18248 + }, + { + "epoch": 0.42778529861636105, + "grad_norm": 3.019418716430664, + "learning_rate": 0.00012253310387597588, + "loss": 1.6703, + "step": 18249 + }, + { + "epoch": 0.42780874019116605, + "grad_norm": 5.661525726318359, + "learning_rate": 0.00012252592709794698, + "loss": 0.53, + "step": 18250 + }, + { + "epoch": 0.42783218176597104, + "grad_norm": 3.0748867988586426, + "learning_rate": 0.00012251875019769174, + "loss": 1.9951, + "step": 18251 + }, + { + "epoch": 0.42785562334077604, + "grad_norm": 0.7504345774650574, + "learning_rate": 0.00012251157317524916, + "loss": 0.0524, + "step": 18252 + }, + { + "epoch": 0.42787906491558103, + "grad_norm": 2.816000461578369, + "learning_rate": 0.00012250439603065814, + "loss": 0.3627, + "step": 18253 + }, + { + "epoch": 0.427902506490386, + "grad_norm": 2.602508544921875, + "learning_rate": 0.0001224972187639576, + "loss": 0.2673, + "step": 18254 + }, + { + "epoch": 0.427925948065191, + "grad_norm": 1.0541965961456299, + "learning_rate": 0.0001224900413751866, + "loss": 0.1634, + "step": 18255 + }, + { + "epoch": 0.427949389639996, + "grad_norm": 2.083767890930176, + "learning_rate": 0.00012248286386438393, + "loss": 0.2282, + "step": 18256 + }, + { + "epoch": 0.427972831214801, + "grad_norm": 3.487016201019287, + "learning_rate": 0.00012247568623158863, + "loss": 0.4039, + "step": 18257 + }, + { + "epoch": 0.427996272789606, + "grad_norm": 4.4983906745910645, + "learning_rate": 0.00012246850847683963, + "loss": 0.5386, + "step": 18258 + }, + { + "epoch": 0.428019714364411, + "grad_norm": 4.816309452056885, + "learning_rate": 0.00012246133060017588, + "loss": 0.4796, + "step": 18259 + }, + { + "epoch": 0.428043155939216, + "grad_norm": 2.2279250621795654, + "learning_rate": 0.00012245415260163627, + "loss": 0.256, + "step": 18260 + }, + { + "epoch": 0.428066597514021, + "grad_norm": 1.7434749603271484, + "learning_rate": 0.00012244697448125986, + "loss": 0.2366, + "step": 18261 + }, + { + "epoch": 0.428090039088826, + "grad_norm": 3.642888069152832, + "learning_rate": 0.0001224397962390855, + "loss": 0.1923, + "step": 18262 + }, + { + "epoch": 0.428113480663631, + "grad_norm": 2.4730803966522217, + "learning_rate": 0.00012243261787515217, + "loss": 0.1778, + "step": 18263 + }, + { + "epoch": 0.42813692223843597, + "grad_norm": 3.0903403759002686, + "learning_rate": 0.00012242543938949886, + "loss": 0.2438, + "step": 18264 + }, + { + "epoch": 0.42816036381324096, + "grad_norm": 2.7104299068450928, + "learning_rate": 0.00012241826078216448, + "loss": 0.2537, + "step": 18265 + }, + { + "epoch": 0.42818380538804596, + "grad_norm": 5.16874361038208, + "learning_rate": 0.00012241108205318795, + "loss": 0.7395, + "step": 18266 + }, + { + "epoch": 0.42820724696285095, + "grad_norm": 2.984511137008667, + "learning_rate": 0.00012240390320260829, + "loss": 0.2979, + "step": 18267 + }, + { + "epoch": 0.42823068853765595, + "grad_norm": 2.066962957382202, + "learning_rate": 0.0001223967242304644, + "loss": 0.1821, + "step": 18268 + }, + { + "epoch": 0.42825413011246094, + "grad_norm": 1.8561053276062012, + "learning_rate": 0.00012238954513679528, + "loss": 0.8307, + "step": 18269 + }, + { + "epoch": 0.42827757168726593, + "grad_norm": 7.2993483543396, + "learning_rate": 0.00012238236592163984, + "loss": 0.6793, + "step": 18270 + }, + { + "epoch": 0.42830101326207093, + "grad_norm": 0.6340397596359253, + "learning_rate": 0.00012237518658503706, + "loss": 0.0578, + "step": 18271 + }, + { + "epoch": 0.4283244548368759, + "grad_norm": 4.244011878967285, + "learning_rate": 0.00012236800712702586, + "loss": 0.4726, + "step": 18272 + }, + { + "epoch": 0.4283478964116809, + "grad_norm": 5.168327331542969, + "learning_rate": 0.00012236082754764525, + "loss": 0.4282, + "step": 18273 + }, + { + "epoch": 0.4283713379864859, + "grad_norm": 3.2817304134368896, + "learning_rate": 0.00012235364784693418, + "loss": 0.3238, + "step": 18274 + }, + { + "epoch": 0.4283947795612909, + "grad_norm": 3.510897397994995, + "learning_rate": 0.00012234646802493157, + "loss": 0.6249, + "step": 18275 + }, + { + "epoch": 0.4284182211360959, + "grad_norm": 4.236361980438232, + "learning_rate": 0.0001223392880816764, + "loss": 0.3034, + "step": 18276 + }, + { + "epoch": 0.4284416627109009, + "grad_norm": 2.895397424697876, + "learning_rate": 0.0001223321080172076, + "loss": 0.3159, + "step": 18277 + }, + { + "epoch": 0.4284651042857059, + "grad_norm": 4.493800163269043, + "learning_rate": 0.0001223249278315642, + "loss": 0.4493, + "step": 18278 + }, + { + "epoch": 0.4284885458605109, + "grad_norm": 1.3114256858825684, + "learning_rate": 0.00012231774752478506, + "loss": 0.1093, + "step": 18279 + }, + { + "epoch": 0.4285119874353159, + "grad_norm": 3.765895366668701, + "learning_rate": 0.0001223105670969092, + "loss": 0.2941, + "step": 18280 + }, + { + "epoch": 0.4285354290101209, + "grad_norm": 2.669110059738159, + "learning_rate": 0.00012230338654797563, + "loss": 0.2503, + "step": 18281 + }, + { + "epoch": 0.4285588705849259, + "grad_norm": 1.7551393508911133, + "learning_rate": 0.00012229620587802322, + "loss": 0.1057, + "step": 18282 + }, + { + "epoch": 0.4285823121597309, + "grad_norm": 4.045108795166016, + "learning_rate": 0.000122289025087091, + "loss": 0.4411, + "step": 18283 + }, + { + "epoch": 0.4286057537345359, + "grad_norm": 1.4988282918930054, + "learning_rate": 0.00012228184417521786, + "loss": 0.2065, + "step": 18284 + }, + { + "epoch": 0.4286291953093409, + "grad_norm": 7.867419242858887, + "learning_rate": 0.00012227466314244283, + "loss": 0.5326, + "step": 18285 + }, + { + "epoch": 0.4286526368841459, + "grad_norm": 4.293398857116699, + "learning_rate": 0.0001222674819888048, + "loss": 0.5745, + "step": 18286 + }, + { + "epoch": 0.4286760784589509, + "grad_norm": 4.715767860412598, + "learning_rate": 0.00012226030071434287, + "loss": 0.412, + "step": 18287 + }, + { + "epoch": 0.4286995200337559, + "grad_norm": 2.350249767303467, + "learning_rate": 0.0001222531193190959, + "loss": 0.169, + "step": 18288 + }, + { + "epoch": 0.4287229616085609, + "grad_norm": 3.4836862087249756, + "learning_rate": 0.00012224593780310285, + "loss": 0.2729, + "step": 18289 + }, + { + "epoch": 0.4287464031833659, + "grad_norm": 2.799846887588501, + "learning_rate": 0.00012223875616640276, + "loss": 0.4641, + "step": 18290 + }, + { + "epoch": 0.42876984475817087, + "grad_norm": 4.172308444976807, + "learning_rate": 0.00012223157440903449, + "loss": 0.431, + "step": 18291 + }, + { + "epoch": 0.42879328633297586, + "grad_norm": 4.662920951843262, + "learning_rate": 0.00012222439253103712, + "loss": 0.7075, + "step": 18292 + }, + { + "epoch": 0.42881672790778086, + "grad_norm": 1.3077541589736938, + "learning_rate": 0.00012221721053244953, + "loss": 0.5374, + "step": 18293 + }, + { + "epoch": 0.42884016948258585, + "grad_norm": 3.1593856811523438, + "learning_rate": 0.0001222100284133108, + "loss": 0.3063, + "step": 18294 + }, + { + "epoch": 0.42886361105739085, + "grad_norm": 1.3690308332443237, + "learning_rate": 0.00012220284617365978, + "loss": 0.1511, + "step": 18295 + }, + { + "epoch": 0.42888705263219584, + "grad_norm": 3.3422834873199463, + "learning_rate": 0.00012219566381353552, + "loss": 0.2569, + "step": 18296 + }, + { + "epoch": 0.42891049420700084, + "grad_norm": 5.244882106781006, + "learning_rate": 0.00012218848133297694, + "loss": 0.3312, + "step": 18297 + }, + { + "epoch": 0.42893393578180583, + "grad_norm": 3.5125458240509033, + "learning_rate": 0.00012218129873202305, + "loss": 0.3732, + "step": 18298 + }, + { + "epoch": 0.4289573773566108, + "grad_norm": 3.212043046951294, + "learning_rate": 0.00012217411601071277, + "loss": 0.3297, + "step": 18299 + }, + { + "epoch": 0.4289808189314158, + "grad_norm": 4.102993011474609, + "learning_rate": 0.00012216693316908516, + "loss": 0.4637, + "step": 18300 + }, + { + "epoch": 0.4290042605062208, + "grad_norm": 2.835198163986206, + "learning_rate": 0.00012215975020717911, + "loss": 0.2902, + "step": 18301 + }, + { + "epoch": 0.4290277020810258, + "grad_norm": 1.1777101755142212, + "learning_rate": 0.00012215256712503367, + "loss": 0.0769, + "step": 18302 + }, + { + "epoch": 0.4290511436558308, + "grad_norm": 2.298677921295166, + "learning_rate": 0.00012214538392268777, + "loss": 1.2157, + "step": 18303 + }, + { + "epoch": 0.4290745852306358, + "grad_norm": 2.3718743324279785, + "learning_rate": 0.00012213820060018036, + "loss": 0.1968, + "step": 18304 + }, + { + "epoch": 0.4290980268054408, + "grad_norm": 1.814428687095642, + "learning_rate": 0.00012213101715755049, + "loss": 0.1745, + "step": 18305 + }, + { + "epoch": 0.4291214683802458, + "grad_norm": 2.9892961978912354, + "learning_rate": 0.0001221238335948371, + "loss": 0.2645, + "step": 18306 + }, + { + "epoch": 0.4291449099550508, + "grad_norm": 3.1151602268218994, + "learning_rate": 0.00012211664991207917, + "loss": 0.2762, + "step": 18307 + }, + { + "epoch": 0.4291683515298558, + "grad_norm": 3.718351125717163, + "learning_rate": 0.00012210946610931563, + "loss": 0.2303, + "step": 18308 + }, + { + "epoch": 0.42919179310466077, + "grad_norm": 0.48483389616012573, + "learning_rate": 0.00012210228218658556, + "loss": 0.0403, + "step": 18309 + }, + { + "epoch": 0.42921523467946576, + "grad_norm": 3.681489944458008, + "learning_rate": 0.00012209509814392784, + "loss": 0.5559, + "step": 18310 + }, + { + "epoch": 0.42923867625427076, + "grad_norm": 1.2611373662948608, + "learning_rate": 0.00012208791398138152, + "loss": 0.1407, + "step": 18311 + }, + { + "epoch": 0.42926211782907575, + "grad_norm": 8.287766456604004, + "learning_rate": 0.00012208072969898556, + "loss": 0.4479, + "step": 18312 + }, + { + "epoch": 0.42928555940388075, + "grad_norm": 3.2581353187561035, + "learning_rate": 0.00012207354529677894, + "loss": 0.365, + "step": 18313 + }, + { + "epoch": 0.42930900097868574, + "grad_norm": 2.6833226680755615, + "learning_rate": 0.00012206636077480062, + "loss": 0.1472, + "step": 18314 + }, + { + "epoch": 0.42933244255349073, + "grad_norm": 5.9575066566467285, + "learning_rate": 0.00012205917613308963, + "loss": 0.2529, + "step": 18315 + }, + { + "epoch": 0.42935588412829573, + "grad_norm": 4.222165584564209, + "learning_rate": 0.00012205199137168495, + "loss": 0.6396, + "step": 18316 + }, + { + "epoch": 0.4293793257031007, + "grad_norm": 2.8121585845947266, + "learning_rate": 0.00012204480649062549, + "loss": 0.2797, + "step": 18317 + }, + { + "epoch": 0.4294027672779057, + "grad_norm": 3.303900957107544, + "learning_rate": 0.00012203762148995031, + "loss": 0.4176, + "step": 18318 + }, + { + "epoch": 0.4294262088527107, + "grad_norm": 3.833840847015381, + "learning_rate": 0.0001220304363696984, + "loss": 0.2786, + "step": 18319 + }, + { + "epoch": 0.4294496504275157, + "grad_norm": 3.4895780086517334, + "learning_rate": 0.00012202325112990869, + "loss": 0.2389, + "step": 18320 + }, + { + "epoch": 0.4294730920023207, + "grad_norm": 2.1675548553466797, + "learning_rate": 0.00012201606577062024, + "loss": 1.6093, + "step": 18321 + }, + { + "epoch": 0.4294965335771257, + "grad_norm": 3.619020938873291, + "learning_rate": 0.00012200888029187196, + "loss": 0.3337, + "step": 18322 + }, + { + "epoch": 0.4295199751519307, + "grad_norm": 4.341684818267822, + "learning_rate": 0.00012200169469370289, + "loss": 0.5099, + "step": 18323 + }, + { + "epoch": 0.4295434167267357, + "grad_norm": 1.1236320734024048, + "learning_rate": 0.00012199450897615201, + "loss": 0.0876, + "step": 18324 + }, + { + "epoch": 0.4295668583015407, + "grad_norm": 5.520229339599609, + "learning_rate": 0.00012198732313925831, + "loss": 0.4614, + "step": 18325 + }, + { + "epoch": 0.42959029987634567, + "grad_norm": 1.3927725553512573, + "learning_rate": 0.00012198013718306076, + "loss": 0.1696, + "step": 18326 + }, + { + "epoch": 0.42961374145115067, + "grad_norm": 2.9965975284576416, + "learning_rate": 0.00012197295110759839, + "loss": 0.3691, + "step": 18327 + }, + { + "epoch": 0.42963718302595566, + "grad_norm": 3.370584011077881, + "learning_rate": 0.00012196576491291017, + "loss": 0.2577, + "step": 18328 + }, + { + "epoch": 0.42966062460076065, + "grad_norm": 2.901076555252075, + "learning_rate": 0.00012195857859903506, + "loss": 0.2316, + "step": 18329 + }, + { + "epoch": 0.42968406617556565, + "grad_norm": 2.329277992248535, + "learning_rate": 0.00012195139216601208, + "loss": 0.2858, + "step": 18330 + }, + { + "epoch": 0.42970750775037064, + "grad_norm": 2.8871335983276367, + "learning_rate": 0.00012194420561388028, + "loss": 0.2462, + "step": 18331 + }, + { + "epoch": 0.4297309493251757, + "grad_norm": 0.573585033416748, + "learning_rate": 0.00012193701894267857, + "loss": 0.0547, + "step": 18332 + }, + { + "epoch": 0.4297543908999807, + "grad_norm": 2.444551706314087, + "learning_rate": 0.00012192983215244598, + "loss": 0.1863, + "step": 18333 + }, + { + "epoch": 0.4297778324747857, + "grad_norm": 4.20174503326416, + "learning_rate": 0.00012192264524322153, + "loss": 0.5631, + "step": 18334 + }, + { + "epoch": 0.4298012740495907, + "grad_norm": 3.05338716506958, + "learning_rate": 0.00012191545821504413, + "loss": 0.3184, + "step": 18335 + }, + { + "epoch": 0.42982471562439567, + "grad_norm": 1.4642127752304077, + "learning_rate": 0.00012190827106795289, + "loss": 0.1471, + "step": 18336 + }, + { + "epoch": 0.42984815719920066, + "grad_norm": 3.4526758193969727, + "learning_rate": 0.00012190108380198675, + "loss": 0.3366, + "step": 18337 + }, + { + "epoch": 0.42987159877400566, + "grad_norm": 1.924721360206604, + "learning_rate": 0.0001218938964171847, + "loss": 0.1245, + "step": 18338 + }, + { + "epoch": 0.42989504034881065, + "grad_norm": 2.4656171798706055, + "learning_rate": 0.00012188670891358576, + "loss": 0.1264, + "step": 18339 + }, + { + "epoch": 0.42991848192361565, + "grad_norm": 7.5762410163879395, + "learning_rate": 0.00012187952129122892, + "loss": 0.7219, + "step": 18340 + }, + { + "epoch": 0.42994192349842064, + "grad_norm": 2.587364673614502, + "learning_rate": 0.00012187233355015317, + "loss": 1.9021, + "step": 18341 + }, + { + "epoch": 0.42996536507322564, + "grad_norm": 1.141445279121399, + "learning_rate": 0.00012186514569039756, + "loss": 0.1221, + "step": 18342 + }, + { + "epoch": 0.42998880664803063, + "grad_norm": 1.1406506299972534, + "learning_rate": 0.00012185795771200101, + "loss": 0.1725, + "step": 18343 + }, + { + "epoch": 0.4300122482228356, + "grad_norm": 2.605217218399048, + "learning_rate": 0.0001218507696150026, + "loss": 0.363, + "step": 18344 + }, + { + "epoch": 0.4300356897976406, + "grad_norm": 2.332916021347046, + "learning_rate": 0.00012184358139944128, + "loss": 0.2189, + "step": 18345 + }, + { + "epoch": 0.4300591313724456, + "grad_norm": 1.709631085395813, + "learning_rate": 0.00012183639306535608, + "loss": 0.2024, + "step": 18346 + }, + { + "epoch": 0.4300825729472506, + "grad_norm": 2.5648205280303955, + "learning_rate": 0.00012182920461278597, + "loss": 0.1757, + "step": 18347 + }, + { + "epoch": 0.4301060145220556, + "grad_norm": 0.910692572593689, + "learning_rate": 0.00012182201604177003, + "loss": 0.0865, + "step": 18348 + }, + { + "epoch": 0.4301294560968606, + "grad_norm": 1.4417394399642944, + "learning_rate": 0.00012181482735234719, + "loss": 0.1204, + "step": 18349 + }, + { + "epoch": 0.4301528976716656, + "grad_norm": 3.4553351402282715, + "learning_rate": 0.0001218076385445565, + "loss": 0.2145, + "step": 18350 + }, + { + "epoch": 0.4301763392464706, + "grad_norm": 0.6291057467460632, + "learning_rate": 0.00012180044961843696, + "loss": 0.0502, + "step": 18351 + }, + { + "epoch": 0.4301997808212756, + "grad_norm": 0.6130902171134949, + "learning_rate": 0.00012179326057402752, + "loss": 0.0623, + "step": 18352 + }, + { + "epoch": 0.4302232223960806, + "grad_norm": 2.3002917766571045, + "learning_rate": 0.00012178607141136725, + "loss": 0.1934, + "step": 18353 + }, + { + "epoch": 0.43024666397088557, + "grad_norm": 4.520608425140381, + "learning_rate": 0.00012177888213049518, + "loss": 0.707, + "step": 18354 + }, + { + "epoch": 0.43027010554569056, + "grad_norm": 0.6920592784881592, + "learning_rate": 0.00012177169273145024, + "loss": 0.0456, + "step": 18355 + }, + { + "epoch": 0.43029354712049556, + "grad_norm": 3.850438356399536, + "learning_rate": 0.0001217645032142715, + "loss": 0.3631, + "step": 18356 + }, + { + "epoch": 0.43031698869530055, + "grad_norm": 2.9131274223327637, + "learning_rate": 0.00012175731357899798, + "loss": 0.4048, + "step": 18357 + }, + { + "epoch": 0.43034043027010555, + "grad_norm": 3.6762962341308594, + "learning_rate": 0.00012175012382566862, + "loss": 0.4335, + "step": 18358 + }, + { + "epoch": 0.43036387184491054, + "grad_norm": 1.3478904962539673, + "learning_rate": 0.00012174293395432248, + "loss": 0.1238, + "step": 18359 + }, + { + "epoch": 0.43038731341971553, + "grad_norm": 2.954274892807007, + "learning_rate": 0.0001217357439649986, + "loss": 0.3528, + "step": 18360 + }, + { + "epoch": 0.43041075499452053, + "grad_norm": 2.4825735092163086, + "learning_rate": 0.00012172855385773596, + "loss": 0.2659, + "step": 18361 + }, + { + "epoch": 0.4304341965693255, + "grad_norm": 1.4885635375976562, + "learning_rate": 0.00012172136363257356, + "loss": 0.1094, + "step": 18362 + }, + { + "epoch": 0.4304576381441305, + "grad_norm": 2.88224458694458, + "learning_rate": 0.00012171417328955044, + "loss": 0.3387, + "step": 18363 + }, + { + "epoch": 0.4304810797189355, + "grad_norm": 2.599177837371826, + "learning_rate": 0.00012170698282870559, + "loss": 0.1608, + "step": 18364 + }, + { + "epoch": 0.4305045212937405, + "grad_norm": 3.531496524810791, + "learning_rate": 0.00012169979225007804, + "loss": 0.3086, + "step": 18365 + }, + { + "epoch": 0.4305279628685455, + "grad_norm": 2.4418327808380127, + "learning_rate": 0.00012169260155370683, + "loss": 0.1878, + "step": 18366 + }, + { + "epoch": 0.4305514044433505, + "grad_norm": 1.3118339776992798, + "learning_rate": 0.00012168541073963096, + "loss": 0.1869, + "step": 18367 + }, + { + "epoch": 0.4305748460181555, + "grad_norm": 0.9381282329559326, + "learning_rate": 0.0001216782198078894, + "loss": 0.0877, + "step": 18368 + }, + { + "epoch": 0.4305982875929605, + "grad_norm": 0.41717448830604553, + "learning_rate": 0.00012167102875852126, + "loss": 0.0641, + "step": 18369 + }, + { + "epoch": 0.4306217291677655, + "grad_norm": 3.3720510005950928, + "learning_rate": 0.00012166383759156547, + "loss": 0.4129, + "step": 18370 + }, + { + "epoch": 0.43064517074257047, + "grad_norm": 3.6986148357391357, + "learning_rate": 0.00012165664630706111, + "loss": 0.197, + "step": 18371 + }, + { + "epoch": 0.43066861231737547, + "grad_norm": 2.544584035873413, + "learning_rate": 0.00012164945490504713, + "loss": 0.1713, + "step": 18372 + }, + { + "epoch": 0.43069205389218046, + "grad_norm": 4.936112403869629, + "learning_rate": 0.00012164226338556267, + "loss": 0.47, + "step": 18373 + }, + { + "epoch": 0.43071549546698545, + "grad_norm": 3.036149263381958, + "learning_rate": 0.00012163507174864664, + "loss": 0.3097, + "step": 18374 + }, + { + "epoch": 0.43073893704179045, + "grad_norm": 4.885524749755859, + "learning_rate": 0.0001216278799943381, + "loss": 0.67, + "step": 18375 + }, + { + "epoch": 0.43076237861659544, + "grad_norm": 4.559704780578613, + "learning_rate": 0.00012162068812267611, + "loss": 0.4609, + "step": 18376 + }, + { + "epoch": 0.43078582019140044, + "grad_norm": 1.3260618448257446, + "learning_rate": 0.00012161349613369961, + "loss": 0.1715, + "step": 18377 + }, + { + "epoch": 0.43080926176620543, + "grad_norm": 2.3118996620178223, + "learning_rate": 0.00012160630402744768, + "loss": 0.9337, + "step": 18378 + }, + { + "epoch": 0.4308327033410104, + "grad_norm": 1.301438331604004, + "learning_rate": 0.00012159911180395937, + "loss": 0.1362, + "step": 18379 + }, + { + "epoch": 0.4308561449158154, + "grad_norm": 2.6806111335754395, + "learning_rate": 0.00012159191946327366, + "loss": 0.1942, + "step": 18380 + }, + { + "epoch": 0.4308795864906204, + "grad_norm": 1.8945249319076538, + "learning_rate": 0.00012158472700542958, + "loss": 0.1457, + "step": 18381 + }, + { + "epoch": 0.4309030280654254, + "grad_norm": 5.610650539398193, + "learning_rate": 0.00012157753443046618, + "loss": 0.7363, + "step": 18382 + }, + { + "epoch": 0.43092646964023046, + "grad_norm": 5.276165962219238, + "learning_rate": 0.00012157034173842243, + "loss": 0.6758, + "step": 18383 + }, + { + "epoch": 0.43094991121503545, + "grad_norm": 0.8348500728607178, + "learning_rate": 0.00012156314892933742, + "loss": 0.1122, + "step": 18384 + }, + { + "epoch": 0.43097335278984045, + "grad_norm": 3.631016731262207, + "learning_rate": 0.00012155595600325018, + "loss": 0.3139, + "step": 18385 + }, + { + "epoch": 0.43099679436464544, + "grad_norm": 1.9363462924957275, + "learning_rate": 0.00012154876296019971, + "loss": 1.247, + "step": 18386 + }, + { + "epoch": 0.43102023593945044, + "grad_norm": 0.6423811912536621, + "learning_rate": 0.00012154156980022502, + "loss": 0.0829, + "step": 18387 + }, + { + "epoch": 0.43104367751425543, + "grad_norm": 2.310523509979248, + "learning_rate": 0.00012153437652336519, + "loss": 0.1893, + "step": 18388 + }, + { + "epoch": 0.4310671190890604, + "grad_norm": 2.8313064575195312, + "learning_rate": 0.00012152718312965925, + "loss": 1.5461, + "step": 18389 + }, + { + "epoch": 0.4310905606638654, + "grad_norm": 1.8054368495941162, + "learning_rate": 0.00012151998961914612, + "loss": 0.099, + "step": 18390 + }, + { + "epoch": 0.4311140022386704, + "grad_norm": 2.9219889640808105, + "learning_rate": 0.000121512795991865, + "loss": 0.5551, + "step": 18391 + }, + { + "epoch": 0.4311374438134754, + "grad_norm": 2.983755111694336, + "learning_rate": 0.00012150560224785485, + "loss": 0.5333, + "step": 18392 + }, + { + "epoch": 0.4311608853882804, + "grad_norm": 4.371601581573486, + "learning_rate": 0.00012149840838715465, + "loss": 0.3354, + "step": 18393 + }, + { + "epoch": 0.4311843269630854, + "grad_norm": 2.964689254760742, + "learning_rate": 0.00012149121440980353, + "loss": 0.4183, + "step": 18394 + }, + { + "epoch": 0.4312077685378904, + "grad_norm": 1.7083033323287964, + "learning_rate": 0.00012148402031584046, + "loss": 0.1738, + "step": 18395 + }, + { + "epoch": 0.4312312101126954, + "grad_norm": 0.6512355804443359, + "learning_rate": 0.00012147682610530447, + "loss": 0.048, + "step": 18396 + }, + { + "epoch": 0.4312546516875004, + "grad_norm": 2.4808766841888428, + "learning_rate": 0.00012146963177823463, + "loss": 0.2303, + "step": 18397 + }, + { + "epoch": 0.4312780932623054, + "grad_norm": 3.3408405780792236, + "learning_rate": 0.00012146243733466995, + "loss": 0.3849, + "step": 18398 + }, + { + "epoch": 0.43130153483711037, + "grad_norm": 3.038123369216919, + "learning_rate": 0.00012145524277464953, + "loss": 0.19, + "step": 18399 + }, + { + "epoch": 0.43132497641191536, + "grad_norm": 3.1306607723236084, + "learning_rate": 0.00012144804809821231, + "loss": 0.3455, + "step": 18400 + }, + { + "epoch": 0.43134841798672036, + "grad_norm": 3.6919965744018555, + "learning_rate": 0.0001214408533053974, + "loss": 0.2665, + "step": 18401 + }, + { + "epoch": 0.43137185956152535, + "grad_norm": 4.422970771789551, + "learning_rate": 0.00012143365839624379, + "loss": 0.3427, + "step": 18402 + }, + { + "epoch": 0.43139530113633034, + "grad_norm": 2.0748109817504883, + "learning_rate": 0.00012142646337079055, + "loss": 0.1543, + "step": 18403 + }, + { + "epoch": 0.43141874271113534, + "grad_norm": 3.16322660446167, + "learning_rate": 0.00012141926822907676, + "loss": 0.4053, + "step": 18404 + }, + { + "epoch": 0.43144218428594033, + "grad_norm": 2.81441068649292, + "learning_rate": 0.00012141207297114141, + "loss": 0.3021, + "step": 18405 + }, + { + "epoch": 0.4314656258607453, + "grad_norm": 4.431305408477783, + "learning_rate": 0.00012140487759702353, + "loss": 0.4278, + "step": 18406 + }, + { + "epoch": 0.4314890674355503, + "grad_norm": 1.4713926315307617, + "learning_rate": 0.00012139768210676219, + "loss": 0.1206, + "step": 18407 + }, + { + "epoch": 0.4315125090103553, + "grad_norm": 3.4134504795074463, + "learning_rate": 0.0001213904865003964, + "loss": 0.2419, + "step": 18408 + }, + { + "epoch": 0.4315359505851603, + "grad_norm": 3.2683026790618896, + "learning_rate": 0.00012138329077796524, + "loss": 0.1934, + "step": 18409 + }, + { + "epoch": 0.4315593921599653, + "grad_norm": 3.154554843902588, + "learning_rate": 0.00012137609493950776, + "loss": 1.6177, + "step": 18410 + }, + { + "epoch": 0.4315828337347703, + "grad_norm": 5.485916614532471, + "learning_rate": 0.00012136889898506298, + "loss": 0.4042, + "step": 18411 + }, + { + "epoch": 0.4316062753095753, + "grad_norm": 1.0363337993621826, + "learning_rate": 0.00012136170291466994, + "loss": 0.0884, + "step": 18412 + }, + { + "epoch": 0.4316297168843803, + "grad_norm": 2.673901081085205, + "learning_rate": 0.00012135450672836774, + "loss": 0.2233, + "step": 18413 + }, + { + "epoch": 0.4316531584591853, + "grad_norm": 2.047860622406006, + "learning_rate": 0.00012134731042619534, + "loss": 0.1374, + "step": 18414 + }, + { + "epoch": 0.4316766000339903, + "grad_norm": 2.6572375297546387, + "learning_rate": 0.00012134011400819187, + "loss": 0.3999, + "step": 18415 + }, + { + "epoch": 0.43170004160879527, + "grad_norm": 5.553116321563721, + "learning_rate": 0.00012133291747439632, + "loss": 0.4418, + "step": 18416 + }, + { + "epoch": 0.43172348318360027, + "grad_norm": 2.3209309577941895, + "learning_rate": 0.00012132572082484776, + "loss": 0.3396, + "step": 18417 + }, + { + "epoch": 0.43174692475840526, + "grad_norm": 0.4946429133415222, + "learning_rate": 0.00012131852405958526, + "loss": 0.0405, + "step": 18418 + }, + { + "epoch": 0.43177036633321025, + "grad_norm": 5.724100589752197, + "learning_rate": 0.00012131132717864781, + "loss": 0.5211, + "step": 18419 + }, + { + "epoch": 0.43179380790801525, + "grad_norm": 1.5145528316497803, + "learning_rate": 0.00012130413018207451, + "loss": 0.1483, + "step": 18420 + }, + { + "epoch": 0.43181724948282024, + "grad_norm": 7.271914482116699, + "learning_rate": 0.00012129693306990443, + "loss": 0.4124, + "step": 18421 + }, + { + "epoch": 0.43184069105762524, + "grad_norm": 5.13018798828125, + "learning_rate": 0.00012128973584217658, + "loss": 0.4883, + "step": 18422 + }, + { + "epoch": 0.43186413263243023, + "grad_norm": 0.49309808015823364, + "learning_rate": 0.00012128253849893001, + "loss": 0.0389, + "step": 18423 + }, + { + "epoch": 0.4318875742072352, + "grad_norm": 0.7285884022712708, + "learning_rate": 0.00012127534104020382, + "loss": 0.0662, + "step": 18424 + }, + { + "epoch": 0.4319110157820402, + "grad_norm": 1.2174354791641235, + "learning_rate": 0.00012126814346603699, + "loss": 0.0989, + "step": 18425 + }, + { + "epoch": 0.4319344573568452, + "grad_norm": 2.86535382270813, + "learning_rate": 0.00012126094577646863, + "loss": 0.1613, + "step": 18426 + }, + { + "epoch": 0.4319578989316502, + "grad_norm": 1.4247550964355469, + "learning_rate": 0.00012125374797153779, + "loss": 0.1816, + "step": 18427 + }, + { + "epoch": 0.4319813405064552, + "grad_norm": 0.7007861733436584, + "learning_rate": 0.00012124655005128349, + "loss": 0.0643, + "step": 18428 + }, + { + "epoch": 0.4320047820812602, + "grad_norm": 1.3503355979919434, + "learning_rate": 0.00012123935201574485, + "loss": 0.1915, + "step": 18429 + }, + { + "epoch": 0.4320282236560652, + "grad_norm": 4.116066932678223, + "learning_rate": 0.00012123215386496087, + "loss": 0.4372, + "step": 18430 + }, + { + "epoch": 0.4320516652308702, + "grad_norm": 5.2750701904296875, + "learning_rate": 0.00012122495559897063, + "loss": 0.4675, + "step": 18431 + }, + { + "epoch": 0.4320751068056752, + "grad_norm": 2.2563555240631104, + "learning_rate": 0.00012121775721781317, + "loss": 0.3098, + "step": 18432 + }, + { + "epoch": 0.4320985483804802, + "grad_norm": 3.1736795902252197, + "learning_rate": 0.00012121055872152758, + "loss": 0.281, + "step": 18433 + }, + { + "epoch": 0.43212198995528517, + "grad_norm": 3.0308802127838135, + "learning_rate": 0.00012120336011015289, + "loss": 0.2796, + "step": 18434 + }, + { + "epoch": 0.4321454315300902, + "grad_norm": 1.250530481338501, + "learning_rate": 0.00012119616138372818, + "loss": 0.1665, + "step": 18435 + }, + { + "epoch": 0.4321688731048952, + "grad_norm": 2.9170260429382324, + "learning_rate": 0.00012118896254229253, + "loss": 0.2166, + "step": 18436 + }, + { + "epoch": 0.4321923146797002, + "grad_norm": 3.4110817909240723, + "learning_rate": 0.00012118176358588493, + "loss": 0.4193, + "step": 18437 + }, + { + "epoch": 0.4322157562545052, + "grad_norm": 1.1211997270584106, + "learning_rate": 0.00012117456451454447, + "loss": 0.1372, + "step": 18438 + }, + { + "epoch": 0.4322391978293102, + "grad_norm": 2.7483277320861816, + "learning_rate": 0.00012116736532831027, + "loss": 1.7652, + "step": 18439 + }, + { + "epoch": 0.4322626394041152, + "grad_norm": 1.046633005142212, + "learning_rate": 0.00012116016602722136, + "loss": 0.1032, + "step": 18440 + }, + { + "epoch": 0.4322860809789202, + "grad_norm": 0.992576539516449, + "learning_rate": 0.00012115296661131674, + "loss": 0.1005, + "step": 18441 + }, + { + "epoch": 0.4323095225537252, + "grad_norm": 5.530698776245117, + "learning_rate": 0.00012114576708063559, + "loss": 0.3735, + "step": 18442 + }, + { + "epoch": 0.4323329641285302, + "grad_norm": 2.648355484008789, + "learning_rate": 0.00012113856743521687, + "loss": 1.6373, + "step": 18443 + }, + { + "epoch": 0.43235640570333517, + "grad_norm": 2.9523074626922607, + "learning_rate": 0.00012113136767509972, + "loss": 0.239, + "step": 18444 + }, + { + "epoch": 0.43237984727814016, + "grad_norm": 0.9004402756690979, + "learning_rate": 0.00012112416780032312, + "loss": 0.1121, + "step": 18445 + }, + { + "epoch": 0.43240328885294516, + "grad_norm": 1.1300121545791626, + "learning_rate": 0.00012111696781092625, + "loss": 0.118, + "step": 18446 + }, + { + "epoch": 0.43242673042775015, + "grad_norm": 1.0101959705352783, + "learning_rate": 0.00012110976770694808, + "loss": 0.1343, + "step": 18447 + }, + { + "epoch": 0.43245017200255514, + "grad_norm": 2.294137477874756, + "learning_rate": 0.00012110256748842774, + "loss": 0.2459, + "step": 18448 + }, + { + "epoch": 0.43247361357736014, + "grad_norm": 1.4101229906082153, + "learning_rate": 0.00012109536715540428, + "loss": 0.1114, + "step": 18449 + }, + { + "epoch": 0.43249705515216513, + "grad_norm": 4.579897403717041, + "learning_rate": 0.00012108816670791676, + "loss": 0.4865, + "step": 18450 + }, + { + "epoch": 0.4325204967269701, + "grad_norm": 2.8144261837005615, + "learning_rate": 0.00012108096614600421, + "loss": 1.8512, + "step": 18451 + }, + { + "epoch": 0.4325439383017751, + "grad_norm": 5.092236518859863, + "learning_rate": 0.00012107376546970581, + "loss": 0.4366, + "step": 18452 + }, + { + "epoch": 0.4325673798765801, + "grad_norm": 3.853999137878418, + "learning_rate": 0.00012106656467906055, + "loss": 0.3788, + "step": 18453 + }, + { + "epoch": 0.4325908214513851, + "grad_norm": 3.584622621536255, + "learning_rate": 0.0001210593637741075, + "loss": 0.5584, + "step": 18454 + }, + { + "epoch": 0.4326142630261901, + "grad_norm": 2.8913259506225586, + "learning_rate": 0.00012105216275488576, + "loss": 1.3398, + "step": 18455 + }, + { + "epoch": 0.4326377046009951, + "grad_norm": 4.948477268218994, + "learning_rate": 0.00012104496162143442, + "loss": 0.7778, + "step": 18456 + }, + { + "epoch": 0.4326611461758001, + "grad_norm": 3.1361238956451416, + "learning_rate": 0.00012103776037379244, + "loss": 0.2659, + "step": 18457 + }, + { + "epoch": 0.4326845877506051, + "grad_norm": 3.4298927783966064, + "learning_rate": 0.00012103055901199907, + "loss": 0.272, + "step": 18458 + }, + { + "epoch": 0.4327080293254101, + "grad_norm": 9.293742179870605, + "learning_rate": 0.00012102335753609326, + "loss": 0.3051, + "step": 18459 + }, + { + "epoch": 0.4327314709002151, + "grad_norm": 3.9637460708618164, + "learning_rate": 0.0001210161559461141, + "loss": 0.4344, + "step": 18460 + }, + { + "epoch": 0.43275491247502007, + "grad_norm": 3.1507070064544678, + "learning_rate": 0.00012100895424210072, + "loss": 1.7633, + "step": 18461 + }, + { + "epoch": 0.43277835404982506, + "grad_norm": 3.748805522918701, + "learning_rate": 0.00012100175242409218, + "loss": 0.2525, + "step": 18462 + }, + { + "epoch": 0.43280179562463006, + "grad_norm": 1.4292689561843872, + "learning_rate": 0.0001209945504921275, + "loss": 0.1176, + "step": 18463 + }, + { + "epoch": 0.43282523719943505, + "grad_norm": 5.455677032470703, + "learning_rate": 0.0001209873484462458, + "loss": 0.3959, + "step": 18464 + }, + { + "epoch": 0.43284867877424005, + "grad_norm": 2.2015438079833984, + "learning_rate": 0.00012098014628648618, + "loss": 0.2288, + "step": 18465 + }, + { + "epoch": 0.43287212034904504, + "grad_norm": 2.970785617828369, + "learning_rate": 0.00012097294401288768, + "loss": 0.205, + "step": 18466 + }, + { + "epoch": 0.43289556192385004, + "grad_norm": 2.708278179168701, + "learning_rate": 0.00012096574162548942, + "loss": 0.168, + "step": 18467 + }, + { + "epoch": 0.43291900349865503, + "grad_norm": 3.352177619934082, + "learning_rate": 0.00012095853912433044, + "loss": 0.2701, + "step": 18468 + }, + { + "epoch": 0.43294244507346, + "grad_norm": 1.1385695934295654, + "learning_rate": 0.00012095133650944984, + "loss": 0.1343, + "step": 18469 + }, + { + "epoch": 0.432965886648265, + "grad_norm": 3.5840132236480713, + "learning_rate": 0.00012094413378088666, + "loss": 0.4458, + "step": 18470 + }, + { + "epoch": 0.43298932822307, + "grad_norm": 4.002110481262207, + "learning_rate": 0.00012093693093868008, + "loss": 0.5481, + "step": 18471 + }, + { + "epoch": 0.433012769797875, + "grad_norm": 4.008200168609619, + "learning_rate": 0.00012092972798286913, + "loss": 1.7857, + "step": 18472 + }, + { + "epoch": 0.43303621137268, + "grad_norm": 3.834280252456665, + "learning_rate": 0.00012092252491349283, + "loss": 1.1724, + "step": 18473 + }, + { + "epoch": 0.433059652947485, + "grad_norm": 0.6147708296775818, + "learning_rate": 0.00012091532173059037, + "loss": 0.0596, + "step": 18474 + }, + { + "epoch": 0.43308309452229, + "grad_norm": 3.0386407375335693, + "learning_rate": 0.00012090811843420074, + "loss": 1.6672, + "step": 18475 + }, + { + "epoch": 0.433106536097095, + "grad_norm": 3.288424491882324, + "learning_rate": 0.0001209009150243631, + "loss": 0.3677, + "step": 18476 + }, + { + "epoch": 0.4331299776719, + "grad_norm": 6.072203636169434, + "learning_rate": 0.0001208937115011165, + "loss": 0.8578, + "step": 18477 + }, + { + "epoch": 0.433153419246705, + "grad_norm": 1.8857591152191162, + "learning_rate": 0.00012088650786450007, + "loss": 0.1302, + "step": 18478 + }, + { + "epoch": 0.43317686082150997, + "grad_norm": 3.144422769546509, + "learning_rate": 0.00012087930411455281, + "loss": 0.2672, + "step": 18479 + }, + { + "epoch": 0.43320030239631496, + "grad_norm": 0.802754819393158, + "learning_rate": 0.00012087210025131387, + "loss": 0.0633, + "step": 18480 + }, + { + "epoch": 0.43322374397111996, + "grad_norm": 0.9118160009384155, + "learning_rate": 0.00012086489627482233, + "loss": 0.0664, + "step": 18481 + }, + { + "epoch": 0.43324718554592495, + "grad_norm": 0.45765936374664307, + "learning_rate": 0.00012085769218511729, + "loss": 0.0222, + "step": 18482 + }, + { + "epoch": 0.43327062712072995, + "grad_norm": 0.7027485966682434, + "learning_rate": 0.00012085048798223781, + "loss": 0.0916, + "step": 18483 + }, + { + "epoch": 0.43329406869553494, + "grad_norm": 1.458832025527954, + "learning_rate": 0.00012084328366622302, + "loss": 0.1632, + "step": 18484 + }, + { + "epoch": 0.43331751027033993, + "grad_norm": 2.309715747833252, + "learning_rate": 0.00012083607923711198, + "loss": 0.2875, + "step": 18485 + }, + { + "epoch": 0.433340951845145, + "grad_norm": 3.7869699001312256, + "learning_rate": 0.00012082887469494377, + "loss": 0.2436, + "step": 18486 + }, + { + "epoch": 0.43336439341995, + "grad_norm": 2.5562400817871094, + "learning_rate": 0.00012082167003975751, + "loss": 0.2607, + "step": 18487 + }, + { + "epoch": 0.433387834994755, + "grad_norm": 3.1529345512390137, + "learning_rate": 0.00012081446527159228, + "loss": 0.3612, + "step": 18488 + }, + { + "epoch": 0.43341127656955997, + "grad_norm": 1.1746606826782227, + "learning_rate": 0.00012080726039048718, + "loss": 0.0544, + "step": 18489 + }, + { + "epoch": 0.43343471814436496, + "grad_norm": 3.080418109893799, + "learning_rate": 0.00012080005539648129, + "loss": 0.1707, + "step": 18490 + }, + { + "epoch": 0.43345815971916996, + "grad_norm": 2.722409963607788, + "learning_rate": 0.00012079285028961376, + "loss": 0.2808, + "step": 18491 + }, + { + "epoch": 0.43348160129397495, + "grad_norm": 1.2955461740493774, + "learning_rate": 0.00012078564506992359, + "loss": 0.1375, + "step": 18492 + }, + { + "epoch": 0.43350504286877994, + "grad_norm": 4.001700401306152, + "learning_rate": 0.0001207784397374499, + "loss": 0.3765, + "step": 18493 + }, + { + "epoch": 0.43352848444358494, + "grad_norm": 4.072098731994629, + "learning_rate": 0.00012077123429223187, + "loss": 0.5913, + "step": 18494 + }, + { + "epoch": 0.43355192601838993, + "grad_norm": 2.643587827682495, + "learning_rate": 0.00012076402873430852, + "loss": 0.2312, + "step": 18495 + }, + { + "epoch": 0.4335753675931949, + "grad_norm": 1.6924442052841187, + "learning_rate": 0.00012075682306371897, + "loss": 0.1428, + "step": 18496 + }, + { + "epoch": 0.4335988091679999, + "grad_norm": 3.4139440059661865, + "learning_rate": 0.00012074961728050231, + "loss": 0.3755, + "step": 18497 + }, + { + "epoch": 0.4336222507428049, + "grad_norm": 1.9304064512252808, + "learning_rate": 0.00012074241138469765, + "loss": 0.11, + "step": 18498 + }, + { + "epoch": 0.4336456923176099, + "grad_norm": 3.3539023399353027, + "learning_rate": 0.00012073520537634407, + "loss": 0.3968, + "step": 18499 + }, + { + "epoch": 0.4336691338924149, + "grad_norm": 4.048435688018799, + "learning_rate": 0.00012072799925548069, + "loss": 0.5689, + "step": 18500 + }, + { + "epoch": 0.4336925754672199, + "grad_norm": 2.2448642253875732, + "learning_rate": 0.00012072079302214659, + "loss": 0.2074, + "step": 18501 + }, + { + "epoch": 0.4337160170420249, + "grad_norm": 3.7312002182006836, + "learning_rate": 0.00012071358667638089, + "loss": 0.3785, + "step": 18502 + }, + { + "epoch": 0.4337394586168299, + "grad_norm": 3.339817523956299, + "learning_rate": 0.00012070638021822272, + "loss": 0.4325, + "step": 18503 + }, + { + "epoch": 0.4337629001916349, + "grad_norm": 4.645453453063965, + "learning_rate": 0.00012069917364771111, + "loss": 0.4278, + "step": 18504 + }, + { + "epoch": 0.4337863417664399, + "grad_norm": 3.9816806316375732, + "learning_rate": 0.00012069196696488518, + "loss": 0.4535, + "step": 18505 + }, + { + "epoch": 0.43380978334124487, + "grad_norm": 3.443765163421631, + "learning_rate": 0.00012068476016978412, + "loss": 0.2961, + "step": 18506 + }, + { + "epoch": 0.43383322491604986, + "grad_norm": 4.264932632446289, + "learning_rate": 0.00012067755326244693, + "loss": 0.4887, + "step": 18507 + }, + { + "epoch": 0.43385666649085486, + "grad_norm": 1.0598775148391724, + "learning_rate": 0.00012067034624291275, + "loss": 0.0591, + "step": 18508 + }, + { + "epoch": 0.43388010806565985, + "grad_norm": 2.65089750289917, + "learning_rate": 0.00012066313911122072, + "loss": 0.1952, + "step": 18509 + }, + { + "epoch": 0.43390354964046485, + "grad_norm": 4.965702533721924, + "learning_rate": 0.00012065593186740991, + "loss": 0.5611, + "step": 18510 + }, + { + "epoch": 0.43392699121526984, + "grad_norm": 1.1417397260665894, + "learning_rate": 0.00012064872451151938, + "loss": 0.1203, + "step": 18511 + }, + { + "epoch": 0.43395043279007484, + "grad_norm": 3.948812484741211, + "learning_rate": 0.00012064151704358835, + "loss": 0.3967, + "step": 18512 + }, + { + "epoch": 0.43397387436487983, + "grad_norm": 1.2151367664337158, + "learning_rate": 0.00012063430946365586, + "loss": 0.1069, + "step": 18513 + }, + { + "epoch": 0.4339973159396848, + "grad_norm": 2.546926498413086, + "learning_rate": 0.000120627101771761, + "loss": 0.2581, + "step": 18514 + }, + { + "epoch": 0.4340207575144898, + "grad_norm": 5.108981609344482, + "learning_rate": 0.00012061989396794293, + "loss": 0.4895, + "step": 18515 + }, + { + "epoch": 0.4340441990892948, + "grad_norm": 3.8575539588928223, + "learning_rate": 0.00012061268605224074, + "loss": 0.4086, + "step": 18516 + }, + { + "epoch": 0.4340676406640998, + "grad_norm": 3.6081178188323975, + "learning_rate": 0.0001206054780246935, + "loss": 0.3587, + "step": 18517 + }, + { + "epoch": 0.4340910822389048, + "grad_norm": 2.493427276611328, + "learning_rate": 0.00012059826988534038, + "loss": 0.1158, + "step": 18518 + }, + { + "epoch": 0.4341145238137098, + "grad_norm": 5.31174898147583, + "learning_rate": 0.00012059106163422047, + "loss": 0.8337, + "step": 18519 + }, + { + "epoch": 0.4341379653885148, + "grad_norm": 3.48476243019104, + "learning_rate": 0.00012058385327137288, + "loss": 0.5672, + "step": 18520 + }, + { + "epoch": 0.4341614069633198, + "grad_norm": 3.405740976333618, + "learning_rate": 0.00012057664479683673, + "loss": 0.2665, + "step": 18521 + }, + { + "epoch": 0.4341848485381248, + "grad_norm": 1.0669629573822021, + "learning_rate": 0.00012056943621065113, + "loss": 0.1183, + "step": 18522 + }, + { + "epoch": 0.4342082901129298, + "grad_norm": 2.815769910812378, + "learning_rate": 0.00012056222751285517, + "loss": 0.2139, + "step": 18523 + }, + { + "epoch": 0.43423173168773477, + "grad_norm": 2.5401713848114014, + "learning_rate": 0.00012055501870348798, + "loss": 0.2497, + "step": 18524 + }, + { + "epoch": 0.43425517326253976, + "grad_norm": 4.150313854217529, + "learning_rate": 0.0001205478097825887, + "loss": 0.315, + "step": 18525 + }, + { + "epoch": 0.43427861483734476, + "grad_norm": 1.3553014993667603, + "learning_rate": 0.00012054060075019644, + "loss": 0.1508, + "step": 18526 + }, + { + "epoch": 0.43430205641214975, + "grad_norm": 3.397184371948242, + "learning_rate": 0.00012053339160635029, + "loss": 0.2311, + "step": 18527 + }, + { + "epoch": 0.43432549798695474, + "grad_norm": 5.416618347167969, + "learning_rate": 0.00012052618235108937, + "loss": 0.2914, + "step": 18528 + }, + { + "epoch": 0.43434893956175974, + "grad_norm": 4.798102855682373, + "learning_rate": 0.00012051897298445283, + "loss": 0.4586, + "step": 18529 + }, + { + "epoch": 0.43437238113656473, + "grad_norm": 2.206468105316162, + "learning_rate": 0.0001205117635064797, + "loss": 0.2758, + "step": 18530 + }, + { + "epoch": 0.43439582271136973, + "grad_norm": 1.533262014389038, + "learning_rate": 0.00012050455391720923, + "loss": 0.2051, + "step": 18531 + }, + { + "epoch": 0.4344192642861747, + "grad_norm": 2.9381966590881348, + "learning_rate": 0.00012049734421668046, + "loss": 0.3514, + "step": 18532 + }, + { + "epoch": 0.4344427058609797, + "grad_norm": 0.7091790437698364, + "learning_rate": 0.0001204901344049325, + "loss": 0.0752, + "step": 18533 + }, + { + "epoch": 0.4344661474357847, + "grad_norm": 3.7171096801757812, + "learning_rate": 0.00012048292448200453, + "loss": 0.401, + "step": 18534 + }, + { + "epoch": 0.4344895890105897, + "grad_norm": 3.0428948402404785, + "learning_rate": 0.00012047571444793564, + "loss": 0.3271, + "step": 18535 + }, + { + "epoch": 0.4345130305853947, + "grad_norm": 4.080360412597656, + "learning_rate": 0.00012046850430276493, + "loss": 0.3346, + "step": 18536 + }, + { + "epoch": 0.43453647216019975, + "grad_norm": 4.784063339233398, + "learning_rate": 0.00012046129404653152, + "loss": 0.654, + "step": 18537 + }, + { + "epoch": 0.43455991373500474, + "grad_norm": 2.8006255626678467, + "learning_rate": 0.00012045408367927458, + "loss": 0.2053, + "step": 18538 + }, + { + "epoch": 0.43458335530980974, + "grad_norm": 2.662574529647827, + "learning_rate": 0.00012044687320103319, + "loss": 1.3666, + "step": 18539 + }, + { + "epoch": 0.43460679688461473, + "grad_norm": 0.5707448124885559, + "learning_rate": 0.00012043966261184651, + "loss": 0.0443, + "step": 18540 + }, + { + "epoch": 0.4346302384594197, + "grad_norm": 3.813217878341675, + "learning_rate": 0.00012043245191175365, + "loss": 0.4379, + "step": 18541 + }, + { + "epoch": 0.4346536800342247, + "grad_norm": 2.0565953254699707, + "learning_rate": 0.0001204252411007937, + "loss": 0.1585, + "step": 18542 + }, + { + "epoch": 0.4346771216090297, + "grad_norm": 1.3413900136947632, + "learning_rate": 0.00012041803017900583, + "loss": 0.1451, + "step": 18543 + }, + { + "epoch": 0.4347005631838347, + "grad_norm": 4.8125152587890625, + "learning_rate": 0.00012041081914642918, + "loss": 0.4351, + "step": 18544 + }, + { + "epoch": 0.4347240047586397, + "grad_norm": 1.192152976989746, + "learning_rate": 0.00012040360800310281, + "loss": 0.0839, + "step": 18545 + }, + { + "epoch": 0.4347474463334447, + "grad_norm": 2.035874128341675, + "learning_rate": 0.00012039639674906591, + "loss": 0.1036, + "step": 18546 + }, + { + "epoch": 0.4347708879082497, + "grad_norm": 2.1179797649383545, + "learning_rate": 0.00012038918538435759, + "loss": 0.1715, + "step": 18547 + }, + { + "epoch": 0.4347943294830547, + "grad_norm": 3.010847568511963, + "learning_rate": 0.00012038197390901698, + "loss": 0.2048, + "step": 18548 + }, + { + "epoch": 0.4348177710578597, + "grad_norm": 3.506368637084961, + "learning_rate": 0.00012037476232308318, + "loss": 0.4318, + "step": 18549 + }, + { + "epoch": 0.4348412126326647, + "grad_norm": 3.562730550765991, + "learning_rate": 0.00012036755062659537, + "loss": 0.2278, + "step": 18550 + }, + { + "epoch": 0.43486465420746967, + "grad_norm": 1.839311957359314, + "learning_rate": 0.00012036033881959268, + "loss": 0.131, + "step": 18551 + }, + { + "epoch": 0.43488809578227466, + "grad_norm": 2.2471821308135986, + "learning_rate": 0.00012035312690211417, + "loss": 0.2649, + "step": 18552 + }, + { + "epoch": 0.43491153735707966, + "grad_norm": 4.3064117431640625, + "learning_rate": 0.00012034591487419906, + "loss": 0.3198, + "step": 18553 + }, + { + "epoch": 0.43493497893188465, + "grad_norm": 3.6258392333984375, + "learning_rate": 0.00012033870273588643, + "loss": 0.3791, + "step": 18554 + }, + { + "epoch": 0.43495842050668965, + "grad_norm": 4.668197154998779, + "learning_rate": 0.00012033149048721544, + "loss": 0.4333, + "step": 18555 + }, + { + "epoch": 0.43498186208149464, + "grad_norm": 6.469217300415039, + "learning_rate": 0.00012032427812822517, + "loss": 0.4808, + "step": 18556 + }, + { + "epoch": 0.43500530365629964, + "grad_norm": 0.9580244421958923, + "learning_rate": 0.00012031706565895484, + "loss": 0.0877, + "step": 18557 + }, + { + "epoch": 0.43502874523110463, + "grad_norm": 2.687150716781616, + "learning_rate": 0.00012030985307944351, + "loss": 1.451, + "step": 18558 + }, + { + "epoch": 0.4350521868059096, + "grad_norm": 6.020147323608398, + "learning_rate": 0.00012030264038973035, + "loss": 1.1449, + "step": 18559 + }, + { + "epoch": 0.4350756283807146, + "grad_norm": 4.174825668334961, + "learning_rate": 0.0001202954275898545, + "loss": 0.3028, + "step": 18560 + }, + { + "epoch": 0.4350990699555196, + "grad_norm": 3.1939587593078613, + "learning_rate": 0.00012028821467985512, + "loss": 1.8077, + "step": 18561 + }, + { + "epoch": 0.4351225115303246, + "grad_norm": 3.582085132598877, + "learning_rate": 0.00012028100165977125, + "loss": 0.4413, + "step": 18562 + }, + { + "epoch": 0.4351459531051296, + "grad_norm": 3.329237699508667, + "learning_rate": 0.00012027378852964215, + "loss": 0.3737, + "step": 18563 + }, + { + "epoch": 0.4351693946799346, + "grad_norm": 2.957026958465576, + "learning_rate": 0.00012026657528950691, + "loss": 1.3759, + "step": 18564 + }, + { + "epoch": 0.4351928362547396, + "grad_norm": 2.200927495956421, + "learning_rate": 0.00012025936193940461, + "loss": 0.2026, + "step": 18565 + }, + { + "epoch": 0.4352162778295446, + "grad_norm": 6.869544982910156, + "learning_rate": 0.00012025214847937447, + "loss": 0.293, + "step": 18566 + }, + { + "epoch": 0.4352397194043496, + "grad_norm": 3.638204574584961, + "learning_rate": 0.00012024493490945561, + "loss": 0.7022, + "step": 18567 + }, + { + "epoch": 0.4352631609791546, + "grad_norm": 1.2822933197021484, + "learning_rate": 0.00012023772122968715, + "loss": 0.0735, + "step": 18568 + }, + { + "epoch": 0.43528660255395957, + "grad_norm": 5.995347499847412, + "learning_rate": 0.00012023050744010826, + "loss": 0.6082, + "step": 18569 + }, + { + "epoch": 0.43531004412876456, + "grad_norm": 1.24712336063385, + "learning_rate": 0.00012022329354075807, + "loss": 0.0904, + "step": 18570 + }, + { + "epoch": 0.43533348570356956, + "grad_norm": 3.365811586380005, + "learning_rate": 0.00012021607953167571, + "loss": 0.3295, + "step": 18571 + }, + { + "epoch": 0.43535692727837455, + "grad_norm": 0.9653152823448181, + "learning_rate": 0.0001202088654129003, + "loss": 0.1112, + "step": 18572 + }, + { + "epoch": 0.43538036885317954, + "grad_norm": 3.2913525104522705, + "learning_rate": 0.00012020165118447109, + "loss": 0.2816, + "step": 18573 + }, + { + "epoch": 0.43540381042798454, + "grad_norm": 4.848526477813721, + "learning_rate": 0.00012019443684642712, + "loss": 0.3732, + "step": 18574 + }, + { + "epoch": 0.43542725200278953, + "grad_norm": 3.08349347114563, + "learning_rate": 0.00012018722239880755, + "loss": 0.2737, + "step": 18575 + }, + { + "epoch": 0.4354506935775945, + "grad_norm": 4.552346706390381, + "learning_rate": 0.00012018000784165158, + "loss": 0.4893, + "step": 18576 + }, + { + "epoch": 0.4354741351523995, + "grad_norm": 2.5857412815093994, + "learning_rate": 0.00012017279317499827, + "loss": 0.1279, + "step": 18577 + }, + { + "epoch": 0.4354975767272045, + "grad_norm": 1.1157735586166382, + "learning_rate": 0.00012016557839888682, + "loss": 0.1043, + "step": 18578 + }, + { + "epoch": 0.4355210183020095, + "grad_norm": 2.5880916118621826, + "learning_rate": 0.00012015836351335643, + "loss": 0.1564, + "step": 18579 + }, + { + "epoch": 0.4355444598768145, + "grad_norm": 1.643069863319397, + "learning_rate": 0.00012015114851844618, + "loss": 0.1194, + "step": 18580 + }, + { + "epoch": 0.4355679014516195, + "grad_norm": 2.0450048446655273, + "learning_rate": 0.00012014393341419518, + "loss": 0.1195, + "step": 18581 + }, + { + "epoch": 0.4355913430264245, + "grad_norm": 1.752493143081665, + "learning_rate": 0.00012013671820064269, + "loss": 0.158, + "step": 18582 + }, + { + "epoch": 0.4356147846012295, + "grad_norm": 2.7292230129241943, + "learning_rate": 0.00012012950287782779, + "loss": 0.2684, + "step": 18583 + }, + { + "epoch": 0.4356382261760345, + "grad_norm": 3.117427110671997, + "learning_rate": 0.0001201222874457896, + "loss": 0.2392, + "step": 18584 + }, + { + "epoch": 0.4356616677508395, + "grad_norm": 4.664453983306885, + "learning_rate": 0.00012011507190456734, + "loss": 0.4429, + "step": 18585 + }, + { + "epoch": 0.43568510932564447, + "grad_norm": 3.378272533416748, + "learning_rate": 0.00012010785625420013, + "loss": 0.355, + "step": 18586 + }, + { + "epoch": 0.43570855090044947, + "grad_norm": 3.4082911014556885, + "learning_rate": 0.00012010064049472714, + "loss": 0.2761, + "step": 18587 + }, + { + "epoch": 0.4357319924752545, + "grad_norm": 0.9913591146469116, + "learning_rate": 0.00012009342462618749, + "loss": 0.1332, + "step": 18588 + }, + { + "epoch": 0.4357554340500595, + "grad_norm": 0.535233736038208, + "learning_rate": 0.0001200862086486204, + "loss": 0.0311, + "step": 18589 + }, + { + "epoch": 0.4357788756248645, + "grad_norm": 3.4487762451171875, + "learning_rate": 0.00012007899256206491, + "loss": 0.3738, + "step": 18590 + }, + { + "epoch": 0.4358023171996695, + "grad_norm": 3.669363021850586, + "learning_rate": 0.00012007177636656026, + "loss": 0.3195, + "step": 18591 + }, + { + "epoch": 0.4358257587744745, + "grad_norm": 3.9083194732666016, + "learning_rate": 0.00012006456006214562, + "loss": 0.5274, + "step": 18592 + }, + { + "epoch": 0.4358492003492795, + "grad_norm": 3.933603048324585, + "learning_rate": 0.00012005734364886009, + "loss": 0.3216, + "step": 18593 + }, + { + "epoch": 0.4358726419240845, + "grad_norm": 2.5518786907196045, + "learning_rate": 0.00012005012712674284, + "loss": 0.2162, + "step": 18594 + }, + { + "epoch": 0.4358960834988895, + "grad_norm": 4.513363361358643, + "learning_rate": 0.00012004291049583305, + "loss": 0.3545, + "step": 18595 + }, + { + "epoch": 0.43591952507369447, + "grad_norm": 2.0302977561950684, + "learning_rate": 0.00012003569375616985, + "loss": 0.1644, + "step": 18596 + }, + { + "epoch": 0.43594296664849946, + "grad_norm": 1.485941767692566, + "learning_rate": 0.00012002847690779242, + "loss": 0.1256, + "step": 18597 + }, + { + "epoch": 0.43596640822330446, + "grad_norm": 3.106783866882324, + "learning_rate": 0.00012002125995073991, + "loss": 0.2764, + "step": 18598 + }, + { + "epoch": 0.43598984979810945, + "grad_norm": 2.3366644382476807, + "learning_rate": 0.00012001404288505151, + "loss": 0.2786, + "step": 18599 + }, + { + "epoch": 0.43601329137291445, + "grad_norm": 0.7916409969329834, + "learning_rate": 0.00012000682571076631, + "loss": 0.0559, + "step": 18600 + }, + { + "epoch": 0.43603673294771944, + "grad_norm": 2.967733383178711, + "learning_rate": 0.00011999960842792355, + "loss": 1.9878, + "step": 18601 + }, + { + "epoch": 0.43606017452252444, + "grad_norm": 3.31787371635437, + "learning_rate": 0.00011999239103656233, + "loss": 0.2629, + "step": 18602 + }, + { + "epoch": 0.43608361609732943, + "grad_norm": 3.8861050605773926, + "learning_rate": 0.0001199851735367218, + "loss": 0.5362, + "step": 18603 + }, + { + "epoch": 0.4361070576721344, + "grad_norm": 3.2158102989196777, + "learning_rate": 0.00011997795592844122, + "loss": 0.316, + "step": 18604 + }, + { + "epoch": 0.4361304992469394, + "grad_norm": 3.8404150009155273, + "learning_rate": 0.00011997073821175966, + "loss": 0.3216, + "step": 18605 + }, + { + "epoch": 0.4361539408217444, + "grad_norm": 1.1415072679519653, + "learning_rate": 0.00011996352038671629, + "loss": 0.0905, + "step": 18606 + }, + { + "epoch": 0.4361773823965494, + "grad_norm": 3.391193151473999, + "learning_rate": 0.00011995630245335032, + "loss": 0.4071, + "step": 18607 + }, + { + "epoch": 0.4362008239713544, + "grad_norm": 1.6795927286148071, + "learning_rate": 0.00011994908441170092, + "loss": 0.1846, + "step": 18608 + }, + { + "epoch": 0.4362242655461594, + "grad_norm": 4.388903617858887, + "learning_rate": 0.0001199418662618072, + "loss": 0.3528, + "step": 18609 + }, + { + "epoch": 0.4362477071209644, + "grad_norm": 3.3049206733703613, + "learning_rate": 0.00011993464800370833, + "loss": 0.323, + "step": 18610 + }, + { + "epoch": 0.4362711486957694, + "grad_norm": 1.8923643827438354, + "learning_rate": 0.00011992742963744352, + "loss": 1.1227, + "step": 18611 + }, + { + "epoch": 0.4362945902705744, + "grad_norm": 3.2850637435913086, + "learning_rate": 0.00011992021116305192, + "loss": 0.3934, + "step": 18612 + }, + { + "epoch": 0.4363180318453794, + "grad_norm": 3.171255350112915, + "learning_rate": 0.00011991299258057269, + "loss": 0.2852, + "step": 18613 + }, + { + "epoch": 0.43634147342018437, + "grad_norm": 0.48549142479896545, + "learning_rate": 0.00011990577389004501, + "loss": 0.0353, + "step": 18614 + }, + { + "epoch": 0.43636491499498936, + "grad_norm": 2.384962320327759, + "learning_rate": 0.00011989855509150802, + "loss": 1.6593, + "step": 18615 + }, + { + "epoch": 0.43638835656979436, + "grad_norm": 2.246297597885132, + "learning_rate": 0.00011989133618500092, + "loss": 0.2202, + "step": 18616 + }, + { + "epoch": 0.43641179814459935, + "grad_norm": 3.2445929050445557, + "learning_rate": 0.00011988411717056289, + "loss": 0.2777, + "step": 18617 + }, + { + "epoch": 0.43643523971940434, + "grad_norm": 2.3620941638946533, + "learning_rate": 0.00011987689804823308, + "loss": 0.233, + "step": 18618 + }, + { + "epoch": 0.43645868129420934, + "grad_norm": 4.677980422973633, + "learning_rate": 0.00011986967881805063, + "loss": 0.3395, + "step": 18619 + }, + { + "epoch": 0.43648212286901433, + "grad_norm": 4.110520839691162, + "learning_rate": 0.00011986245948005477, + "loss": 0.3012, + "step": 18620 + }, + { + "epoch": 0.4365055644438193, + "grad_norm": 4.374446868896484, + "learning_rate": 0.00011985524003428463, + "loss": 0.4142, + "step": 18621 + }, + { + "epoch": 0.4365290060186243, + "grad_norm": 4.2116851806640625, + "learning_rate": 0.00011984802048077941, + "loss": 0.2874, + "step": 18622 + }, + { + "epoch": 0.4365524475934293, + "grad_norm": 2.899129629135132, + "learning_rate": 0.00011984080081957827, + "loss": 0.1662, + "step": 18623 + }, + { + "epoch": 0.4365758891682343, + "grad_norm": 3.887531042098999, + "learning_rate": 0.0001198335810507204, + "loss": 0.3978, + "step": 18624 + }, + { + "epoch": 0.4365993307430393, + "grad_norm": 5.062718391418457, + "learning_rate": 0.00011982636117424495, + "loss": 0.4356, + "step": 18625 + }, + { + "epoch": 0.4366227723178443, + "grad_norm": 2.1933627128601074, + "learning_rate": 0.0001198191411901911, + "loss": 0.1661, + "step": 18626 + }, + { + "epoch": 0.4366462138926493, + "grad_norm": 3.1373066902160645, + "learning_rate": 0.00011981192109859805, + "loss": 0.4003, + "step": 18627 + }, + { + "epoch": 0.4366696554674543, + "grad_norm": 2.7158305644989014, + "learning_rate": 0.00011980470089950495, + "loss": 0.2559, + "step": 18628 + }, + { + "epoch": 0.4366930970422593, + "grad_norm": 1.5517752170562744, + "learning_rate": 0.00011979748059295097, + "loss": 0.2188, + "step": 18629 + }, + { + "epoch": 0.4367165386170643, + "grad_norm": 2.790048122406006, + "learning_rate": 0.00011979026017897532, + "loss": 1.3258, + "step": 18630 + }, + { + "epoch": 0.43673998019186927, + "grad_norm": 5.423187732696533, + "learning_rate": 0.00011978303965761718, + "loss": 0.5162, + "step": 18631 + }, + { + "epoch": 0.43676342176667426, + "grad_norm": 4.9961700439453125, + "learning_rate": 0.00011977581902891568, + "loss": 0.6559, + "step": 18632 + }, + { + "epoch": 0.43678686334147926, + "grad_norm": 5.287592887878418, + "learning_rate": 0.00011976859829291003, + "loss": 0.4647, + "step": 18633 + }, + { + "epoch": 0.43681030491628425, + "grad_norm": 4.118725776672363, + "learning_rate": 0.00011976137744963942, + "loss": 0.2445, + "step": 18634 + }, + { + "epoch": 0.43683374649108925, + "grad_norm": 2.6103293895721436, + "learning_rate": 0.00011975415649914304, + "loss": 0.2227, + "step": 18635 + }, + { + "epoch": 0.43685718806589424, + "grad_norm": 3.726694345474243, + "learning_rate": 0.00011974693544146003, + "loss": 0.2648, + "step": 18636 + }, + { + "epoch": 0.43688062964069924, + "grad_norm": 3.0177743434906006, + "learning_rate": 0.0001197397142766296, + "loss": 0.2423, + "step": 18637 + }, + { + "epoch": 0.43690407121550423, + "grad_norm": 2.972872734069824, + "learning_rate": 0.00011973249300469092, + "loss": 0.325, + "step": 18638 + }, + { + "epoch": 0.4369275127903093, + "grad_norm": 3.5424444675445557, + "learning_rate": 0.00011972527162568316, + "loss": 0.5782, + "step": 18639 + }, + { + "epoch": 0.4369509543651143, + "grad_norm": 3.629946708679199, + "learning_rate": 0.00011971805013964555, + "loss": 0.2916, + "step": 18640 + }, + { + "epoch": 0.43697439593991927, + "grad_norm": 1.11056649684906, + "learning_rate": 0.00011971082854661721, + "loss": 0.1126, + "step": 18641 + }, + { + "epoch": 0.43699783751472426, + "grad_norm": 3.9045464992523193, + "learning_rate": 0.00011970360684663742, + "loss": 0.244, + "step": 18642 + }, + { + "epoch": 0.43702127908952926, + "grad_norm": 3.1146087646484375, + "learning_rate": 0.00011969638503974527, + "loss": 0.3235, + "step": 18643 + }, + { + "epoch": 0.43704472066433425, + "grad_norm": 2.004371404647827, + "learning_rate": 0.00011968916312597998, + "loss": 0.1526, + "step": 18644 + }, + { + "epoch": 0.43706816223913925, + "grad_norm": 1.2574739456176758, + "learning_rate": 0.00011968194110538071, + "loss": 0.0894, + "step": 18645 + }, + { + "epoch": 0.43709160381394424, + "grad_norm": 4.919438362121582, + "learning_rate": 0.00011967471897798672, + "loss": 0.6229, + "step": 18646 + }, + { + "epoch": 0.43711504538874923, + "grad_norm": 2.8728342056274414, + "learning_rate": 0.00011966749674383715, + "loss": 0.2397, + "step": 18647 + }, + { + "epoch": 0.43713848696355423, + "grad_norm": 3.799172878265381, + "learning_rate": 0.00011966027440297116, + "loss": 0.3324, + "step": 18648 + }, + { + "epoch": 0.4371619285383592, + "grad_norm": 3.1747255325317383, + "learning_rate": 0.000119653051955428, + "loss": 0.2917, + "step": 18649 + }, + { + "epoch": 0.4371853701131642, + "grad_norm": 3.267791986465454, + "learning_rate": 0.00011964582940124682, + "loss": 0.4413, + "step": 18650 + }, + { + "epoch": 0.4372088116879692, + "grad_norm": 1.1513628959655762, + "learning_rate": 0.00011963860674046677, + "loss": 0.0762, + "step": 18651 + }, + { + "epoch": 0.4372322532627742, + "grad_norm": 3.706148624420166, + "learning_rate": 0.00011963138397312715, + "loss": 0.3131, + "step": 18652 + }, + { + "epoch": 0.4372556948375792, + "grad_norm": 0.9451831579208374, + "learning_rate": 0.00011962416109926706, + "loss": 0.0389, + "step": 18653 + }, + { + "epoch": 0.4372791364123842, + "grad_norm": 3.3999977111816406, + "learning_rate": 0.00011961693811892571, + "loss": 0.335, + "step": 18654 + }, + { + "epoch": 0.4373025779871892, + "grad_norm": 3.506352186203003, + "learning_rate": 0.00011960971503214233, + "loss": 0.3809, + "step": 18655 + }, + { + "epoch": 0.4373260195619942, + "grad_norm": 4.126288890838623, + "learning_rate": 0.00011960249183895609, + "loss": 0.4744, + "step": 18656 + }, + { + "epoch": 0.4373494611367992, + "grad_norm": 4.9658589363098145, + "learning_rate": 0.00011959526853940612, + "loss": 0.5174, + "step": 18657 + }, + { + "epoch": 0.4373729027116042, + "grad_norm": 1.0859895944595337, + "learning_rate": 0.00011958804513353173, + "loss": 0.128, + "step": 18658 + }, + { + "epoch": 0.43739634428640917, + "grad_norm": 3.457787036895752, + "learning_rate": 0.00011958082162137203, + "loss": 0.3372, + "step": 18659 + }, + { + "epoch": 0.43741978586121416, + "grad_norm": 1.809916615486145, + "learning_rate": 0.00011957359800296624, + "loss": 0.183, + "step": 18660 + }, + { + "epoch": 0.43744322743601916, + "grad_norm": 4.164453983306885, + "learning_rate": 0.00011956637427835358, + "loss": 0.3472, + "step": 18661 + }, + { + "epoch": 0.43746666901082415, + "grad_norm": 1.371394157409668, + "learning_rate": 0.00011955915044757322, + "loss": 0.1534, + "step": 18662 + }, + { + "epoch": 0.43749011058562914, + "grad_norm": 1.8615522384643555, + "learning_rate": 0.00011955192651066434, + "loss": 0.1762, + "step": 18663 + }, + { + "epoch": 0.43751355216043414, + "grad_norm": 3.3984506130218506, + "learning_rate": 0.00011954470246766615, + "loss": 0.3318, + "step": 18664 + }, + { + "epoch": 0.43753699373523913, + "grad_norm": 1.9371808767318726, + "learning_rate": 0.00011953747831861789, + "loss": 0.248, + "step": 18665 + }, + { + "epoch": 0.4375604353100441, + "grad_norm": 3.4789047241210938, + "learning_rate": 0.00011953025406355871, + "loss": 0.5472, + "step": 18666 + }, + { + "epoch": 0.4375838768848491, + "grad_norm": 1.4328713417053223, + "learning_rate": 0.0001195230297025278, + "loss": 0.1368, + "step": 18667 + }, + { + "epoch": 0.4376073184596541, + "grad_norm": 1.4627149105072021, + "learning_rate": 0.00011951580523556441, + "loss": 0.1647, + "step": 18668 + }, + { + "epoch": 0.4376307600344591, + "grad_norm": 0.8240583539009094, + "learning_rate": 0.00011950858066270769, + "loss": 0.0506, + "step": 18669 + }, + { + "epoch": 0.4376542016092641, + "grad_norm": 2.726228713989258, + "learning_rate": 0.00011950135598399686, + "loss": 0.2645, + "step": 18670 + }, + { + "epoch": 0.4376776431840691, + "grad_norm": 2.949713706970215, + "learning_rate": 0.00011949413119947115, + "loss": 0.2742, + "step": 18671 + }, + { + "epoch": 0.4377010847588741, + "grad_norm": 4.3494133949279785, + "learning_rate": 0.00011948690630916975, + "loss": 0.274, + "step": 18672 + }, + { + "epoch": 0.4377245263336791, + "grad_norm": 3.5729315280914307, + "learning_rate": 0.00011947968131313179, + "loss": 0.4417, + "step": 18673 + }, + { + "epoch": 0.4377479679084841, + "grad_norm": 3.443863868713379, + "learning_rate": 0.00011947245621139658, + "loss": 0.3578, + "step": 18674 + }, + { + "epoch": 0.4377714094832891, + "grad_norm": 0.9954337477684021, + "learning_rate": 0.00011946523100400326, + "loss": 0.1357, + "step": 18675 + }, + { + "epoch": 0.43779485105809407, + "grad_norm": 5.966698169708252, + "learning_rate": 0.00011945800569099102, + "loss": 0.5389, + "step": 18676 + }, + { + "epoch": 0.43781829263289906, + "grad_norm": 2.1628594398498535, + "learning_rate": 0.00011945078027239913, + "loss": 0.1551, + "step": 18677 + }, + { + "epoch": 0.43784173420770406, + "grad_norm": 1.0416265726089478, + "learning_rate": 0.00011944355474826676, + "loss": 0.1053, + "step": 18678 + }, + { + "epoch": 0.43786517578250905, + "grad_norm": 3.0643839836120605, + "learning_rate": 0.00011943632911863309, + "loss": 0.2605, + "step": 18679 + }, + { + "epoch": 0.43788861735731405, + "grad_norm": 3.5510618686676025, + "learning_rate": 0.00011942910338353739, + "loss": 1.229, + "step": 18680 + }, + { + "epoch": 0.43791205893211904, + "grad_norm": 2.076219320297241, + "learning_rate": 0.00011942187754301881, + "loss": 0.3248, + "step": 18681 + }, + { + "epoch": 0.43793550050692404, + "grad_norm": 3.2996768951416016, + "learning_rate": 0.00011941465159711656, + "loss": 0.3131, + "step": 18682 + }, + { + "epoch": 0.43795894208172903, + "grad_norm": 3.0947437286376953, + "learning_rate": 0.00011940742554586985, + "loss": 0.2078, + "step": 18683 + }, + { + "epoch": 0.437982383656534, + "grad_norm": 1.6747527122497559, + "learning_rate": 0.00011940019938931795, + "loss": 0.1741, + "step": 18684 + }, + { + "epoch": 0.438005825231339, + "grad_norm": 0.23488949239253998, + "learning_rate": 0.00011939297312750002, + "loss": 0.0317, + "step": 18685 + }, + { + "epoch": 0.438029266806144, + "grad_norm": 4.646968841552734, + "learning_rate": 0.00011938574676045524, + "loss": 0.4184, + "step": 18686 + }, + { + "epoch": 0.438052708380949, + "grad_norm": 3.9210357666015625, + "learning_rate": 0.00011937852028822287, + "loss": 0.295, + "step": 18687 + }, + { + "epoch": 0.438076149955754, + "grad_norm": 4.351168632507324, + "learning_rate": 0.0001193712937108421, + "loss": 0.2974, + "step": 18688 + }, + { + "epoch": 0.438099591530559, + "grad_norm": 3.5363869667053223, + "learning_rate": 0.00011936406702835214, + "loss": 0.3189, + "step": 18689 + }, + { + "epoch": 0.43812303310536405, + "grad_norm": 3.1446919441223145, + "learning_rate": 0.00011935684024079223, + "loss": 0.3769, + "step": 18690 + }, + { + "epoch": 0.43814647468016904, + "grad_norm": 2.691389799118042, + "learning_rate": 0.00011934961334820156, + "loss": 0.2597, + "step": 18691 + }, + { + "epoch": 0.43816991625497403, + "grad_norm": 1.0375947952270508, + "learning_rate": 0.00011934238635061931, + "loss": 0.127, + "step": 18692 + }, + { + "epoch": 0.43819335782977903, + "grad_norm": 3.5921437740325928, + "learning_rate": 0.00011933515924808477, + "loss": 0.5709, + "step": 18693 + }, + { + "epoch": 0.438216799404584, + "grad_norm": 2.712575674057007, + "learning_rate": 0.00011932793204063708, + "loss": 0.1948, + "step": 18694 + }, + { + "epoch": 0.438240240979389, + "grad_norm": 2.568913698196411, + "learning_rate": 0.00011932070472831547, + "loss": 0.2817, + "step": 18695 + }, + { + "epoch": 0.438263682554194, + "grad_norm": 3.8612403869628906, + "learning_rate": 0.00011931347731115923, + "loss": 0.5269, + "step": 18696 + }, + { + "epoch": 0.438287124128999, + "grad_norm": 1.7039612531661987, + "learning_rate": 0.00011930624978920748, + "loss": 0.6404, + "step": 18697 + }, + { + "epoch": 0.438310565703804, + "grad_norm": 2.353712797164917, + "learning_rate": 0.00011929902216249948, + "loss": 0.181, + "step": 18698 + }, + { + "epoch": 0.438334007278609, + "grad_norm": 1.0145201683044434, + "learning_rate": 0.00011929179443107445, + "loss": 0.0997, + "step": 18699 + }, + { + "epoch": 0.438357448853414, + "grad_norm": 2.912247657775879, + "learning_rate": 0.0001192845665949716, + "loss": 0.2014, + "step": 18700 + }, + { + "epoch": 0.438380890428219, + "grad_norm": 3.316612720489502, + "learning_rate": 0.00011927733865423016, + "loss": 0.3259, + "step": 18701 + }, + { + "epoch": 0.438404332003024, + "grad_norm": 3.906458616256714, + "learning_rate": 0.0001192701106088893, + "loss": 0.4977, + "step": 18702 + }, + { + "epoch": 0.43842777357782897, + "grad_norm": 2.390791416168213, + "learning_rate": 0.00011926288245898834, + "loss": 1.5899, + "step": 18703 + }, + { + "epoch": 0.43845121515263397, + "grad_norm": 2.115339756011963, + "learning_rate": 0.00011925565420456641, + "loss": 0.2894, + "step": 18704 + }, + { + "epoch": 0.43847465672743896, + "grad_norm": 2.8660497665405273, + "learning_rate": 0.00011924842584566274, + "loss": 0.5413, + "step": 18705 + }, + { + "epoch": 0.43849809830224395, + "grad_norm": 3.5603764057159424, + "learning_rate": 0.00011924119738231657, + "loss": 0.4761, + "step": 18706 + }, + { + "epoch": 0.43852153987704895, + "grad_norm": 2.1782705783843994, + "learning_rate": 0.00011923396881456715, + "loss": 0.1213, + "step": 18707 + }, + { + "epoch": 0.43854498145185394, + "grad_norm": 2.3440301418304443, + "learning_rate": 0.00011922674014245365, + "loss": 0.2599, + "step": 18708 + }, + { + "epoch": 0.43856842302665894, + "grad_norm": 4.516836643218994, + "learning_rate": 0.00011921951136601533, + "loss": 0.2706, + "step": 18709 + }, + { + "epoch": 0.43859186460146393, + "grad_norm": 3.400726556777954, + "learning_rate": 0.00011921228248529142, + "loss": 0.3361, + "step": 18710 + }, + { + "epoch": 0.4386153061762689, + "grad_norm": 1.6322754621505737, + "learning_rate": 0.00011920505350032107, + "loss": 0.1385, + "step": 18711 + }, + { + "epoch": 0.4386387477510739, + "grad_norm": 1.2206459045410156, + "learning_rate": 0.00011919782441114358, + "loss": 0.1256, + "step": 18712 + }, + { + "epoch": 0.4386621893258789, + "grad_norm": 1.1161580085754395, + "learning_rate": 0.00011919059521779819, + "loss": 0.1063, + "step": 18713 + }, + { + "epoch": 0.4386856309006839, + "grad_norm": 1.1347867250442505, + "learning_rate": 0.00011918336592032404, + "loss": 0.1324, + "step": 18714 + }, + { + "epoch": 0.4387090724754889, + "grad_norm": 4.397532939910889, + "learning_rate": 0.00011917613651876045, + "loss": 0.5351, + "step": 18715 + }, + { + "epoch": 0.4387325140502939, + "grad_norm": 2.4698970317840576, + "learning_rate": 0.0001191689070131466, + "loss": 0.1527, + "step": 18716 + }, + { + "epoch": 0.4387559556250989, + "grad_norm": 6.317165851593018, + "learning_rate": 0.00011916167740352171, + "loss": 1.1195, + "step": 18717 + }, + { + "epoch": 0.4387793971999039, + "grad_norm": 2.8565609455108643, + "learning_rate": 0.00011915444768992501, + "loss": 0.2939, + "step": 18718 + }, + { + "epoch": 0.4388028387747089, + "grad_norm": 3.1083943843841553, + "learning_rate": 0.00011914721787239576, + "loss": 0.1538, + "step": 18719 + }, + { + "epoch": 0.4388262803495139, + "grad_norm": 5.248408794403076, + "learning_rate": 0.00011913998795097316, + "loss": 0.5333, + "step": 18720 + }, + { + "epoch": 0.43884972192431887, + "grad_norm": 1.6801282167434692, + "learning_rate": 0.00011913275792569643, + "loss": 0.6409, + "step": 18721 + }, + { + "epoch": 0.43887316349912386, + "grad_norm": 2.9268157482147217, + "learning_rate": 0.00011912552779660484, + "loss": 0.2482, + "step": 18722 + }, + { + "epoch": 0.43889660507392886, + "grad_norm": 2.6305508613586426, + "learning_rate": 0.0001191182975637376, + "loss": 0.3389, + "step": 18723 + }, + { + "epoch": 0.43892004664873385, + "grad_norm": 3.0152804851531982, + "learning_rate": 0.0001191110672271339, + "loss": 1.7898, + "step": 18724 + }, + { + "epoch": 0.43894348822353885, + "grad_norm": 3.531079053878784, + "learning_rate": 0.00011910383678683305, + "loss": 0.3645, + "step": 18725 + }, + { + "epoch": 0.43896692979834384, + "grad_norm": 2.17075777053833, + "learning_rate": 0.00011909660624287425, + "loss": 0.1648, + "step": 18726 + }, + { + "epoch": 0.43899037137314884, + "grad_norm": 0.8501767516136169, + "learning_rate": 0.0001190893755952967, + "loss": 0.0601, + "step": 18727 + }, + { + "epoch": 0.43901381294795383, + "grad_norm": 5.549977779388428, + "learning_rate": 0.00011908214484413969, + "loss": 0.5395, + "step": 18728 + }, + { + "epoch": 0.4390372545227588, + "grad_norm": 1.9449455738067627, + "learning_rate": 0.00011907491398944242, + "loss": 0.2152, + "step": 18729 + }, + { + "epoch": 0.4390606960975638, + "grad_norm": 2.372018575668335, + "learning_rate": 0.00011906768303124409, + "loss": 0.1534, + "step": 18730 + }, + { + "epoch": 0.4390841376723688, + "grad_norm": 3.9812281131744385, + "learning_rate": 0.000119060451969584, + "loss": 0.5358, + "step": 18731 + }, + { + "epoch": 0.4391075792471738, + "grad_norm": 2.1309547424316406, + "learning_rate": 0.00011905322080450139, + "loss": 0.3038, + "step": 18732 + }, + { + "epoch": 0.4391310208219788, + "grad_norm": 3.4847514629364014, + "learning_rate": 0.00011904598953603542, + "loss": 0.4109, + "step": 18733 + }, + { + "epoch": 0.4391544623967838, + "grad_norm": 1.2230621576309204, + "learning_rate": 0.00011903875816422543, + "loss": 0.1606, + "step": 18734 + }, + { + "epoch": 0.4391779039715888, + "grad_norm": 1.0264177322387695, + "learning_rate": 0.00011903152668911057, + "loss": 0.0523, + "step": 18735 + }, + { + "epoch": 0.4392013455463938, + "grad_norm": 0.833346426486969, + "learning_rate": 0.0001190242951107301, + "loss": 0.0872, + "step": 18736 + }, + { + "epoch": 0.4392247871211988, + "grad_norm": 0.7963765859603882, + "learning_rate": 0.00011901706342912328, + "loss": 0.0623, + "step": 18737 + }, + { + "epoch": 0.4392482286960038, + "grad_norm": 3.597184181213379, + "learning_rate": 0.00011900983164432936, + "loss": 0.1855, + "step": 18738 + }, + { + "epoch": 0.43927167027080877, + "grad_norm": 1.175999641418457, + "learning_rate": 0.00011900259975638755, + "loss": 0.1041, + "step": 18739 + }, + { + "epoch": 0.43929511184561376, + "grad_norm": 0.8340128064155579, + "learning_rate": 0.00011899536776533709, + "loss": 0.0811, + "step": 18740 + }, + { + "epoch": 0.4393185534204188, + "grad_norm": 1.1153327226638794, + "learning_rate": 0.00011898813567121723, + "loss": 0.064, + "step": 18741 + }, + { + "epoch": 0.4393419949952238, + "grad_norm": 2.597395181655884, + "learning_rate": 0.00011898090347406721, + "loss": 1.6333, + "step": 18742 + }, + { + "epoch": 0.4393654365700288, + "grad_norm": 3.219130754470825, + "learning_rate": 0.00011897367117392627, + "loss": 0.3972, + "step": 18743 + }, + { + "epoch": 0.4393888781448338, + "grad_norm": 0.5304232835769653, + "learning_rate": 0.00011896643877083367, + "loss": 0.053, + "step": 18744 + }, + { + "epoch": 0.4394123197196388, + "grad_norm": 3.753390312194824, + "learning_rate": 0.00011895920626482863, + "loss": 0.2066, + "step": 18745 + }, + { + "epoch": 0.4394357612944438, + "grad_norm": 1.9911621809005737, + "learning_rate": 0.00011895197365595041, + "loss": 0.2403, + "step": 18746 + }, + { + "epoch": 0.4394592028692488, + "grad_norm": 4.406522750854492, + "learning_rate": 0.00011894474094423826, + "loss": 0.2862, + "step": 18747 + }, + { + "epoch": 0.43948264444405377, + "grad_norm": 0.6535826921463013, + "learning_rate": 0.0001189375081297314, + "loss": 0.0395, + "step": 18748 + }, + { + "epoch": 0.43950608601885877, + "grad_norm": 3.1881749629974365, + "learning_rate": 0.0001189302752124691, + "loss": 0.4282, + "step": 18749 + }, + { + "epoch": 0.43952952759366376, + "grad_norm": 3.6359806060791016, + "learning_rate": 0.00011892304219249055, + "loss": 0.401, + "step": 18750 + }, + { + "epoch": 0.43955296916846875, + "grad_norm": 4.243947982788086, + "learning_rate": 0.00011891580906983508, + "loss": 0.4569, + "step": 18751 + }, + { + "epoch": 0.43957641074327375, + "grad_norm": 3.2776243686676025, + "learning_rate": 0.0001189085758445419, + "loss": 0.2674, + "step": 18752 + }, + { + "epoch": 0.43959985231807874, + "grad_norm": 1.973413348197937, + "learning_rate": 0.00011890134251665026, + "loss": 0.2016, + "step": 18753 + }, + { + "epoch": 0.43962329389288374, + "grad_norm": 2.892542600631714, + "learning_rate": 0.00011889410908619937, + "loss": 0.1608, + "step": 18754 + }, + { + "epoch": 0.43964673546768873, + "grad_norm": 3.1265063285827637, + "learning_rate": 0.00011888687555322855, + "loss": 0.339, + "step": 18755 + }, + { + "epoch": 0.4396701770424937, + "grad_norm": 2.8887712955474854, + "learning_rate": 0.00011887964191777699, + "loss": 0.2627, + "step": 18756 + }, + { + "epoch": 0.4396936186172987, + "grad_norm": 1.1694679260253906, + "learning_rate": 0.00011887240817988396, + "loss": 0.1176, + "step": 18757 + }, + { + "epoch": 0.4397170601921037, + "grad_norm": 4.8662004470825195, + "learning_rate": 0.00011886517433958875, + "loss": 0.4701, + "step": 18758 + }, + { + "epoch": 0.4397405017669087, + "grad_norm": 2.883910655975342, + "learning_rate": 0.00011885794039693052, + "loss": 0.2678, + "step": 18759 + }, + { + "epoch": 0.4397639433417137, + "grad_norm": 5.216654300689697, + "learning_rate": 0.00011885070635194861, + "loss": 1.0029, + "step": 18760 + }, + { + "epoch": 0.4397873849165187, + "grad_norm": 0.9310591816902161, + "learning_rate": 0.00011884347220468222, + "loss": 0.0832, + "step": 18761 + }, + { + "epoch": 0.4398108264913237, + "grad_norm": 1.6448357105255127, + "learning_rate": 0.00011883623795517062, + "loss": 0.0511, + "step": 18762 + }, + { + "epoch": 0.4398342680661287, + "grad_norm": 3.7800679206848145, + "learning_rate": 0.00011882900360345308, + "loss": 0.3333, + "step": 18763 + }, + { + "epoch": 0.4398577096409337, + "grad_norm": 3.607154369354248, + "learning_rate": 0.00011882176914956884, + "loss": 0.3489, + "step": 18764 + }, + { + "epoch": 0.4398811512157387, + "grad_norm": 4.2982683181762695, + "learning_rate": 0.00011881453459355713, + "loss": 0.4211, + "step": 18765 + }, + { + "epoch": 0.43990459279054367, + "grad_norm": 2.4312655925750732, + "learning_rate": 0.00011880729993545725, + "loss": 0.2782, + "step": 18766 + }, + { + "epoch": 0.43992803436534866, + "grad_norm": 5.032980442047119, + "learning_rate": 0.00011880006517530839, + "loss": 0.6612, + "step": 18767 + }, + { + "epoch": 0.43995147594015366, + "grad_norm": 2.540677309036255, + "learning_rate": 0.0001187928303131499, + "loss": 1.3884, + "step": 18768 + }, + { + "epoch": 0.43997491751495865, + "grad_norm": 1.3636858463287354, + "learning_rate": 0.00011878559534902095, + "loss": 0.1648, + "step": 18769 + }, + { + "epoch": 0.43999835908976365, + "grad_norm": 1.7479087114334106, + "learning_rate": 0.00011877836028296083, + "loss": 0.1589, + "step": 18770 + }, + { + "epoch": 0.44002180066456864, + "grad_norm": 1.136970043182373, + "learning_rate": 0.00011877112511500881, + "loss": 0.1152, + "step": 18771 + }, + { + "epoch": 0.44004524223937364, + "grad_norm": 4.692764759063721, + "learning_rate": 0.00011876388984520414, + "loss": 0.545, + "step": 18772 + }, + { + "epoch": 0.44006868381417863, + "grad_norm": 3.2287256717681885, + "learning_rate": 0.00011875665447358606, + "loss": 0.3079, + "step": 18773 + }, + { + "epoch": 0.4400921253889836, + "grad_norm": 3.1022706031799316, + "learning_rate": 0.00011874941900019387, + "loss": 0.3642, + "step": 18774 + }, + { + "epoch": 0.4401155669637886, + "grad_norm": 2.1257851123809814, + "learning_rate": 0.00011874218342506678, + "loss": 0.2793, + "step": 18775 + }, + { + "epoch": 0.4401390085385936, + "grad_norm": 1.0536140203475952, + "learning_rate": 0.00011873494774824412, + "loss": 0.1267, + "step": 18776 + }, + { + "epoch": 0.4401624501133986, + "grad_norm": 2.2669994831085205, + "learning_rate": 0.00011872771196976509, + "loss": 0.1615, + "step": 18777 + }, + { + "epoch": 0.4401858916882036, + "grad_norm": 2.315178632736206, + "learning_rate": 0.00011872047608966892, + "loss": 0.1912, + "step": 18778 + }, + { + "epoch": 0.4402093332630086, + "grad_norm": 2.2963192462921143, + "learning_rate": 0.00011871324010799496, + "loss": 0.1673, + "step": 18779 + }, + { + "epoch": 0.4402327748378136, + "grad_norm": 0.8435307145118713, + "learning_rate": 0.00011870600402478242, + "loss": 0.0649, + "step": 18780 + }, + { + "epoch": 0.4402562164126186, + "grad_norm": 0.971409022808075, + "learning_rate": 0.00011869876784007059, + "loss": 0.1169, + "step": 18781 + }, + { + "epoch": 0.4402796579874236, + "grad_norm": 0.8620990514755249, + "learning_rate": 0.00011869153155389872, + "loss": 0.0724, + "step": 18782 + }, + { + "epoch": 0.4403030995622286, + "grad_norm": 2.8511171340942383, + "learning_rate": 0.00011868429516630609, + "loss": 0.4343, + "step": 18783 + }, + { + "epoch": 0.44032654113703357, + "grad_norm": 3.1302499771118164, + "learning_rate": 0.0001186770586773319, + "loss": 0.2856, + "step": 18784 + }, + { + "epoch": 0.44034998271183856, + "grad_norm": 2.977297782897949, + "learning_rate": 0.0001186698220870155, + "loss": 0.2217, + "step": 18785 + }, + { + "epoch": 0.44037342428664356, + "grad_norm": 3.7043402194976807, + "learning_rate": 0.00011866258539539612, + "loss": 0.3297, + "step": 18786 + }, + { + "epoch": 0.44039686586144855, + "grad_norm": 2.413884162902832, + "learning_rate": 0.000118655348602513, + "loss": 0.2643, + "step": 18787 + }, + { + "epoch": 0.44042030743625354, + "grad_norm": 1.9635143280029297, + "learning_rate": 0.00011864811170840546, + "loss": 0.1608, + "step": 18788 + }, + { + "epoch": 0.44044374901105854, + "grad_norm": 4.471302509307861, + "learning_rate": 0.00011864087471311274, + "loss": 0.6042, + "step": 18789 + }, + { + "epoch": 0.44046719058586353, + "grad_norm": 1.5932425260543823, + "learning_rate": 0.0001186336376166741, + "loss": 0.066, + "step": 18790 + }, + { + "epoch": 0.4404906321606685, + "grad_norm": 4.829171657562256, + "learning_rate": 0.0001186264004191288, + "loss": 0.6597, + "step": 18791 + }, + { + "epoch": 0.4405140737354735, + "grad_norm": 2.6993002891540527, + "learning_rate": 0.00011861916312051617, + "loss": 0.3022, + "step": 18792 + }, + { + "epoch": 0.44053751531027857, + "grad_norm": 2.7979114055633545, + "learning_rate": 0.00011861192572087542, + "loss": 0.1564, + "step": 18793 + }, + { + "epoch": 0.44056095688508357, + "grad_norm": 1.2028357982635498, + "learning_rate": 0.00011860468822024584, + "loss": 0.1266, + "step": 18794 + }, + { + "epoch": 0.44058439845988856, + "grad_norm": 4.087827682495117, + "learning_rate": 0.00011859745061866669, + "loss": 0.1662, + "step": 18795 + }, + { + "epoch": 0.44060784003469355, + "grad_norm": 1.8810327053070068, + "learning_rate": 0.00011859021291617727, + "loss": 0.1551, + "step": 18796 + }, + { + "epoch": 0.44063128160949855, + "grad_norm": 3.836113691329956, + "learning_rate": 0.00011858297511281677, + "loss": 0.2785, + "step": 18797 + }, + { + "epoch": 0.44065472318430354, + "grad_norm": 0.7680836319923401, + "learning_rate": 0.0001185757372086246, + "loss": 0.1294, + "step": 18798 + }, + { + "epoch": 0.44067816475910854, + "grad_norm": 4.361044883728027, + "learning_rate": 0.00011856849920363993, + "loss": 0.3306, + "step": 18799 + }, + { + "epoch": 0.44070160633391353, + "grad_norm": 3.6126575469970703, + "learning_rate": 0.00011856126109790204, + "loss": 0.3881, + "step": 18800 + }, + { + "epoch": 0.4407250479087185, + "grad_norm": 2.6247246265411377, + "learning_rate": 0.00011855402289145025, + "loss": 0.2722, + "step": 18801 + }, + { + "epoch": 0.4407484894835235, + "grad_norm": 5.229251384735107, + "learning_rate": 0.00011854678458432382, + "loss": 0.6913, + "step": 18802 + }, + { + "epoch": 0.4407719310583285, + "grad_norm": 3.807213306427002, + "learning_rate": 0.000118539546176562, + "loss": 0.1907, + "step": 18803 + }, + { + "epoch": 0.4407953726331335, + "grad_norm": 3.1033613681793213, + "learning_rate": 0.00011853230766820407, + "loss": 0.3803, + "step": 18804 + }, + { + "epoch": 0.4408188142079385, + "grad_norm": 0.690156102180481, + "learning_rate": 0.00011852506905928936, + "loss": 0.0716, + "step": 18805 + }, + { + "epoch": 0.4408422557827435, + "grad_norm": 2.3799853324890137, + "learning_rate": 0.00011851783034985705, + "loss": 0.3205, + "step": 18806 + }, + { + "epoch": 0.4408656973575485, + "grad_norm": 2.3606255054473877, + "learning_rate": 0.0001185105915399465, + "loss": 0.2366, + "step": 18807 + }, + { + "epoch": 0.4408891389323535, + "grad_norm": 0.4044630527496338, + "learning_rate": 0.00011850335262959698, + "loss": 0.0315, + "step": 18808 + }, + { + "epoch": 0.4409125805071585, + "grad_norm": 4.042853355407715, + "learning_rate": 0.00011849611361884769, + "loss": 0.2775, + "step": 18809 + }, + { + "epoch": 0.4409360220819635, + "grad_norm": 2.7022862434387207, + "learning_rate": 0.000118488874507738, + "loss": 0.3039, + "step": 18810 + }, + { + "epoch": 0.44095946365676847, + "grad_norm": 2.1978118419647217, + "learning_rate": 0.00011848163529630716, + "loss": 0.2336, + "step": 18811 + }, + { + "epoch": 0.44098290523157346, + "grad_norm": 2.323456287384033, + "learning_rate": 0.00011847439598459447, + "loss": 0.229, + "step": 18812 + }, + { + "epoch": 0.44100634680637846, + "grad_norm": 6.082571983337402, + "learning_rate": 0.00011846715657263913, + "loss": 0.77, + "step": 18813 + }, + { + "epoch": 0.44102978838118345, + "grad_norm": 5.4818196296691895, + "learning_rate": 0.00011845991706048054, + "loss": 0.5477, + "step": 18814 + }, + { + "epoch": 0.44105322995598845, + "grad_norm": 2.6803202629089355, + "learning_rate": 0.00011845267744815787, + "loss": 1.6761, + "step": 18815 + }, + { + "epoch": 0.44107667153079344, + "grad_norm": 0.7985587120056152, + "learning_rate": 0.00011844543773571047, + "loss": 0.0976, + "step": 18816 + }, + { + "epoch": 0.44110011310559843, + "grad_norm": 4.7919206619262695, + "learning_rate": 0.00011843819792317762, + "loss": 0.6773, + "step": 18817 + }, + { + "epoch": 0.44112355468040343, + "grad_norm": 2.5874457359313965, + "learning_rate": 0.0001184309580105986, + "loss": 0.1792, + "step": 18818 + }, + { + "epoch": 0.4411469962552084, + "grad_norm": 1.7353090047836304, + "learning_rate": 0.00011842371799801264, + "loss": 0.1994, + "step": 18819 + }, + { + "epoch": 0.4411704378300134, + "grad_norm": 2.088074207305908, + "learning_rate": 0.00011841647788545911, + "loss": 0.1844, + "step": 18820 + }, + { + "epoch": 0.4411938794048184, + "grad_norm": 1.8199994564056396, + "learning_rate": 0.00011840923767297723, + "loss": 1.0709, + "step": 18821 + }, + { + "epoch": 0.4412173209796234, + "grad_norm": 2.5245437622070312, + "learning_rate": 0.00011840199736060633, + "loss": 0.1432, + "step": 18822 + }, + { + "epoch": 0.4412407625544284, + "grad_norm": 3.5602834224700928, + "learning_rate": 0.00011839475694838564, + "loss": 0.3294, + "step": 18823 + }, + { + "epoch": 0.4412642041292334, + "grad_norm": 2.032541275024414, + "learning_rate": 0.00011838751643635453, + "loss": 0.2439, + "step": 18824 + }, + { + "epoch": 0.4412876457040384, + "grad_norm": 2.160043716430664, + "learning_rate": 0.0001183802758245522, + "loss": 0.0742, + "step": 18825 + }, + { + "epoch": 0.4413110872788434, + "grad_norm": 5.225127220153809, + "learning_rate": 0.00011837303511301799, + "loss": 0.5018, + "step": 18826 + }, + { + "epoch": 0.4413345288536484, + "grad_norm": 4.797053337097168, + "learning_rate": 0.00011836579430179115, + "loss": 0.4301, + "step": 18827 + }, + { + "epoch": 0.44135797042845337, + "grad_norm": 1.3441407680511475, + "learning_rate": 0.00011835855339091104, + "loss": 0.1783, + "step": 18828 + }, + { + "epoch": 0.44138141200325837, + "grad_norm": 1.2707408666610718, + "learning_rate": 0.00011835131238041685, + "loss": 0.1915, + "step": 18829 + }, + { + "epoch": 0.44140485357806336, + "grad_norm": 1.8913516998291016, + "learning_rate": 0.00011834407127034799, + "loss": 0.1992, + "step": 18830 + }, + { + "epoch": 0.44142829515286836, + "grad_norm": 1.7515077590942383, + "learning_rate": 0.00011833683006074365, + "loss": 0.0999, + "step": 18831 + }, + { + "epoch": 0.44145173672767335, + "grad_norm": 2.2665815353393555, + "learning_rate": 0.00011832958875164313, + "loss": 0.3302, + "step": 18832 + }, + { + "epoch": 0.44147517830247834, + "grad_norm": 3.5714924335479736, + "learning_rate": 0.00011832234734308578, + "loss": 0.2734, + "step": 18833 + }, + { + "epoch": 0.44149861987728334, + "grad_norm": 4.6063032150268555, + "learning_rate": 0.00011831510583511082, + "loss": 0.4442, + "step": 18834 + }, + { + "epoch": 0.44152206145208833, + "grad_norm": 3.299509286880493, + "learning_rate": 0.00011830786422775761, + "loss": 0.3821, + "step": 18835 + }, + { + "epoch": 0.4415455030268933, + "grad_norm": 2.541858196258545, + "learning_rate": 0.00011830062252106542, + "loss": 0.3268, + "step": 18836 + }, + { + "epoch": 0.4415689446016983, + "grad_norm": 1.2293306589126587, + "learning_rate": 0.00011829338071507355, + "loss": 0.1003, + "step": 18837 + }, + { + "epoch": 0.4415923861765033, + "grad_norm": 3.1208581924438477, + "learning_rate": 0.00011828613880982122, + "loss": 0.1717, + "step": 18838 + }, + { + "epoch": 0.4416158277513083, + "grad_norm": 2.8744828701019287, + "learning_rate": 0.00011827889680534785, + "loss": 0.2636, + "step": 18839 + }, + { + "epoch": 0.4416392693261133, + "grad_norm": 3.2165379524230957, + "learning_rate": 0.00011827165470169263, + "loss": 0.3752, + "step": 18840 + }, + { + "epoch": 0.4416627109009183, + "grad_norm": 3.1682217121124268, + "learning_rate": 0.00011826441249889492, + "loss": 1.5263, + "step": 18841 + }, + { + "epoch": 0.4416861524757233, + "grad_norm": 1.7718515396118164, + "learning_rate": 0.00011825717019699398, + "loss": 0.1274, + "step": 18842 + }, + { + "epoch": 0.4417095940505283, + "grad_norm": 4.048906326293945, + "learning_rate": 0.00011824992779602914, + "loss": 0.4291, + "step": 18843 + }, + { + "epoch": 0.44173303562533334, + "grad_norm": 0.9112381339073181, + "learning_rate": 0.00011824268529603968, + "loss": 0.073, + "step": 18844 + }, + { + "epoch": 0.44175647720013833, + "grad_norm": 3.065441608428955, + "learning_rate": 0.00011823544269706487, + "loss": 0.2112, + "step": 18845 + }, + { + "epoch": 0.4417799187749433, + "grad_norm": 1.065629243850708, + "learning_rate": 0.00011822819999914406, + "loss": 0.0807, + "step": 18846 + }, + { + "epoch": 0.4418033603497483, + "grad_norm": 3.6389338970184326, + "learning_rate": 0.00011822095720231652, + "loss": 0.3217, + "step": 18847 + }, + { + "epoch": 0.4418268019245533, + "grad_norm": 0.6339200735092163, + "learning_rate": 0.00011821371430662153, + "loss": 0.0353, + "step": 18848 + }, + { + "epoch": 0.4418502434993583, + "grad_norm": 1.1002261638641357, + "learning_rate": 0.00011820647131209845, + "loss": 0.078, + "step": 18849 + }, + { + "epoch": 0.4418736850741633, + "grad_norm": 4.5608391761779785, + "learning_rate": 0.00011819922821878655, + "loss": 0.6393, + "step": 18850 + }, + { + "epoch": 0.4418971266489683, + "grad_norm": 3.879425525665283, + "learning_rate": 0.00011819198502672508, + "loss": 0.3583, + "step": 18851 + }, + { + "epoch": 0.4419205682237733, + "grad_norm": 3.9213545322418213, + "learning_rate": 0.0001181847417359534, + "loss": 0.4514, + "step": 18852 + }, + { + "epoch": 0.4419440097985783, + "grad_norm": 2.6652472019195557, + "learning_rate": 0.00011817749834651084, + "loss": 0.2391, + "step": 18853 + }, + { + "epoch": 0.4419674513733833, + "grad_norm": 2.9112002849578857, + "learning_rate": 0.0001181702548584366, + "loss": 0.2234, + "step": 18854 + }, + { + "epoch": 0.4419908929481883, + "grad_norm": 4.539226531982422, + "learning_rate": 0.0001181630112717701, + "loss": 0.433, + "step": 18855 + }, + { + "epoch": 0.44201433452299327, + "grad_norm": 3.947225332260132, + "learning_rate": 0.00011815576758655058, + "loss": 0.5073, + "step": 18856 + }, + { + "epoch": 0.44203777609779826, + "grad_norm": 3.020958185195923, + "learning_rate": 0.00011814852380281733, + "loss": 0.3001, + "step": 18857 + }, + { + "epoch": 0.44206121767260326, + "grad_norm": 2.5157411098480225, + "learning_rate": 0.0001181412799206097, + "loss": 0.2587, + "step": 18858 + }, + { + "epoch": 0.44208465924740825, + "grad_norm": 1.1977289915084839, + "learning_rate": 0.00011813403593996696, + "loss": 0.0586, + "step": 18859 + }, + { + "epoch": 0.44210810082221325, + "grad_norm": 4.441188812255859, + "learning_rate": 0.00011812679186092847, + "loss": 0.3685, + "step": 18860 + }, + { + "epoch": 0.44213154239701824, + "grad_norm": 3.9277355670928955, + "learning_rate": 0.00011811954768353347, + "loss": 0.4212, + "step": 18861 + }, + { + "epoch": 0.44215498397182323, + "grad_norm": 2.8807671070098877, + "learning_rate": 0.00011811230340782128, + "loss": 0.3209, + "step": 18862 + }, + { + "epoch": 0.44217842554662823, + "grad_norm": 3.76628041267395, + "learning_rate": 0.00011810505903383124, + "loss": 0.3969, + "step": 18863 + }, + { + "epoch": 0.4422018671214332, + "grad_norm": 2.598116159439087, + "learning_rate": 0.00011809781456160262, + "loss": 0.1241, + "step": 18864 + }, + { + "epoch": 0.4422253086962382, + "grad_norm": 0.9993415474891663, + "learning_rate": 0.00011809056999117479, + "loss": 0.1235, + "step": 18865 + }, + { + "epoch": 0.4422487502710432, + "grad_norm": 3.95135498046875, + "learning_rate": 0.00011808332532258699, + "loss": 0.5516, + "step": 18866 + }, + { + "epoch": 0.4422721918458482, + "grad_norm": 4.0510149002075195, + "learning_rate": 0.00011807608055587855, + "loss": 0.4111, + "step": 18867 + }, + { + "epoch": 0.4422956334206532, + "grad_norm": 2.8625340461730957, + "learning_rate": 0.00011806883569108882, + "loss": 0.1991, + "step": 18868 + }, + { + "epoch": 0.4423190749954582, + "grad_norm": 4.637742042541504, + "learning_rate": 0.00011806159072825706, + "loss": 0.4887, + "step": 18869 + }, + { + "epoch": 0.4423425165702632, + "grad_norm": 0.952307939529419, + "learning_rate": 0.00011805434566742257, + "loss": 0.127, + "step": 18870 + }, + { + "epoch": 0.4423659581450682, + "grad_norm": 1.0895922183990479, + "learning_rate": 0.00011804710050862476, + "loss": 0.1014, + "step": 18871 + }, + { + "epoch": 0.4423893997198732, + "grad_norm": 1.0844881534576416, + "learning_rate": 0.00011803985525190284, + "loss": 0.0827, + "step": 18872 + }, + { + "epoch": 0.44241284129467817, + "grad_norm": 3.147960662841797, + "learning_rate": 0.00011803260989729614, + "loss": 0.3958, + "step": 18873 + }, + { + "epoch": 0.44243628286948317, + "grad_norm": 3.1413347721099854, + "learning_rate": 0.00011802536444484402, + "loss": 0.2538, + "step": 18874 + }, + { + "epoch": 0.44245972444428816, + "grad_norm": 2.320380926132202, + "learning_rate": 0.00011801811889458578, + "loss": 0.1746, + "step": 18875 + }, + { + "epoch": 0.44248316601909315, + "grad_norm": 8.068000793457031, + "learning_rate": 0.00011801087324656068, + "loss": 0.5283, + "step": 18876 + }, + { + "epoch": 0.44250660759389815, + "grad_norm": 0.5042237043380737, + "learning_rate": 0.00011800362750080807, + "loss": 0.0396, + "step": 18877 + }, + { + "epoch": 0.44253004916870314, + "grad_norm": 3.0167078971862793, + "learning_rate": 0.0001179963816573673, + "loss": 0.2305, + "step": 18878 + }, + { + "epoch": 0.44255349074350814, + "grad_norm": 3.7432262897491455, + "learning_rate": 0.00011798913571627768, + "loss": 0.3378, + "step": 18879 + }, + { + "epoch": 0.44257693231831313, + "grad_norm": 6.56703519821167, + "learning_rate": 0.00011798188967757844, + "loss": 0.4604, + "step": 18880 + }, + { + "epoch": 0.4426003738931181, + "grad_norm": 3.7194888591766357, + "learning_rate": 0.00011797464354130902, + "loss": 0.4029, + "step": 18881 + }, + { + "epoch": 0.4426238154679231, + "grad_norm": 1.9607027769088745, + "learning_rate": 0.00011796739730750863, + "loss": 0.2045, + "step": 18882 + }, + { + "epoch": 0.4426472570427281, + "grad_norm": 3.4347496032714844, + "learning_rate": 0.00011796015097621667, + "loss": 0.2707, + "step": 18883 + }, + { + "epoch": 0.4426706986175331, + "grad_norm": 2.4463369846343994, + "learning_rate": 0.0001179529045474724, + "loss": 0.2008, + "step": 18884 + }, + { + "epoch": 0.4426941401923381, + "grad_norm": 4.972349166870117, + "learning_rate": 0.00011794565802131521, + "loss": 0.6824, + "step": 18885 + }, + { + "epoch": 0.4427175817671431, + "grad_norm": 2.4160585403442383, + "learning_rate": 0.00011793841139778432, + "loss": 0.2676, + "step": 18886 + }, + { + "epoch": 0.4427410233419481, + "grad_norm": 3.164621591567993, + "learning_rate": 0.00011793116467691913, + "loss": 0.2338, + "step": 18887 + }, + { + "epoch": 0.4427644649167531, + "grad_norm": 3.358227014541626, + "learning_rate": 0.00011792391785875893, + "loss": 0.2684, + "step": 18888 + }, + { + "epoch": 0.4427879064915581, + "grad_norm": 3.8433785438537598, + "learning_rate": 0.00011791667094334307, + "loss": 0.3508, + "step": 18889 + }, + { + "epoch": 0.4428113480663631, + "grad_norm": 0.48187798261642456, + "learning_rate": 0.00011790942393071083, + "loss": 0.0473, + "step": 18890 + }, + { + "epoch": 0.44283478964116807, + "grad_norm": 3.630288600921631, + "learning_rate": 0.00011790217682090157, + "loss": 0.3596, + "step": 18891 + }, + { + "epoch": 0.44285823121597306, + "grad_norm": 2.798194408416748, + "learning_rate": 0.00011789492961395456, + "loss": 0.1645, + "step": 18892 + }, + { + "epoch": 0.44288167279077806, + "grad_norm": 3.8280465602874756, + "learning_rate": 0.0001178876823099092, + "loss": 0.564, + "step": 18893 + }, + { + "epoch": 0.44290511436558305, + "grad_norm": 3.6166555881500244, + "learning_rate": 0.00011788043490880473, + "loss": 0.4235, + "step": 18894 + }, + { + "epoch": 0.4429285559403881, + "grad_norm": 1.3290317058563232, + "learning_rate": 0.00011787318741068056, + "loss": 0.0976, + "step": 18895 + }, + { + "epoch": 0.4429519975151931, + "grad_norm": 3.761159658432007, + "learning_rate": 0.00011786593981557595, + "loss": 0.3439, + "step": 18896 + }, + { + "epoch": 0.4429754390899981, + "grad_norm": 5.836744785308838, + "learning_rate": 0.00011785869212353026, + "loss": 0.4861, + "step": 18897 + }, + { + "epoch": 0.4429988806648031, + "grad_norm": 3.0607118606567383, + "learning_rate": 0.00011785144433458282, + "loss": 0.1132, + "step": 18898 + }, + { + "epoch": 0.4430223222396081, + "grad_norm": 0.5933123230934143, + "learning_rate": 0.00011784419644877289, + "loss": 0.0833, + "step": 18899 + }, + { + "epoch": 0.4430457638144131, + "grad_norm": 4.121816158294678, + "learning_rate": 0.00011783694846613986, + "loss": 0.4225, + "step": 18900 + }, + { + "epoch": 0.44306920538921807, + "grad_norm": 4.337246894836426, + "learning_rate": 0.00011782970038672308, + "loss": 0.2459, + "step": 18901 + }, + { + "epoch": 0.44309264696402306, + "grad_norm": 1.4661850929260254, + "learning_rate": 0.00011782245221056183, + "loss": 0.1016, + "step": 18902 + }, + { + "epoch": 0.44311608853882806, + "grad_norm": 0.2181936800479889, + "learning_rate": 0.00011781520393769545, + "loss": 0.0158, + "step": 18903 + }, + { + "epoch": 0.44313953011363305, + "grad_norm": 3.4951274394989014, + "learning_rate": 0.00011780795556816328, + "loss": 0.3671, + "step": 18904 + }, + { + "epoch": 0.44316297168843805, + "grad_norm": 2.5764288902282715, + "learning_rate": 0.00011780070710200464, + "loss": 0.2705, + "step": 18905 + }, + { + "epoch": 0.44318641326324304, + "grad_norm": 4.845753192901611, + "learning_rate": 0.00011779345853925885, + "loss": 0.2928, + "step": 18906 + }, + { + "epoch": 0.44320985483804803, + "grad_norm": 4.69016170501709, + "learning_rate": 0.00011778620987996525, + "loss": 0.3263, + "step": 18907 + }, + { + "epoch": 0.44323329641285303, + "grad_norm": 3.011178731918335, + "learning_rate": 0.00011777896112416318, + "loss": 0.3648, + "step": 18908 + }, + { + "epoch": 0.443256737987658, + "grad_norm": 1.8953118324279785, + "learning_rate": 0.00011777171227189199, + "loss": 0.1681, + "step": 18909 + }, + { + "epoch": 0.443280179562463, + "grad_norm": 1.251494288444519, + "learning_rate": 0.00011776446332319098, + "loss": 0.1651, + "step": 18910 + }, + { + "epoch": 0.443303621137268, + "grad_norm": 0.7134552001953125, + "learning_rate": 0.00011775721427809947, + "loss": 0.0641, + "step": 18911 + }, + { + "epoch": 0.443327062712073, + "grad_norm": 1.1506651639938354, + "learning_rate": 0.00011774996513665683, + "loss": 0.0912, + "step": 18912 + }, + { + "epoch": 0.443350504286878, + "grad_norm": 4.8813958168029785, + "learning_rate": 0.00011774271589890237, + "loss": 0.7606, + "step": 18913 + }, + { + "epoch": 0.443373945861683, + "grad_norm": 3.275665521621704, + "learning_rate": 0.00011773546656487546, + "loss": 0.3666, + "step": 18914 + }, + { + "epoch": 0.443397387436488, + "grad_norm": 3.7515642642974854, + "learning_rate": 0.00011772821713461539, + "loss": 0.3203, + "step": 18915 + }, + { + "epoch": 0.443420829011293, + "grad_norm": 1.242607593536377, + "learning_rate": 0.00011772096760816154, + "loss": 0.123, + "step": 18916 + }, + { + "epoch": 0.443444270586098, + "grad_norm": 2.2909951210021973, + "learning_rate": 0.00011771371798555319, + "loss": 0.1902, + "step": 18917 + }, + { + "epoch": 0.44346771216090297, + "grad_norm": 3.3722074031829834, + "learning_rate": 0.00011770646826682972, + "loss": 0.1964, + "step": 18918 + }, + { + "epoch": 0.44349115373570797, + "grad_norm": 5.376132011413574, + "learning_rate": 0.00011769921845203044, + "loss": 0.3737, + "step": 18919 + }, + { + "epoch": 0.44351459531051296, + "grad_norm": 2.028191328048706, + "learning_rate": 0.00011769196854119474, + "loss": 0.764, + "step": 18920 + }, + { + "epoch": 0.44353803688531795, + "grad_norm": 3.7286882400512695, + "learning_rate": 0.0001176847185343619, + "loss": 0.3916, + "step": 18921 + }, + { + "epoch": 0.44356147846012295, + "grad_norm": 2.69934344291687, + "learning_rate": 0.00011767746843157127, + "loss": 0.1935, + "step": 18922 + }, + { + "epoch": 0.44358492003492794, + "grad_norm": 3.035428524017334, + "learning_rate": 0.00011767021823286222, + "loss": 0.3862, + "step": 18923 + }, + { + "epoch": 0.44360836160973294, + "grad_norm": 1.6010066270828247, + "learning_rate": 0.00011766296793827405, + "loss": 0.5362, + "step": 18924 + }, + { + "epoch": 0.44363180318453793, + "grad_norm": 4.736315727233887, + "learning_rate": 0.00011765571754784612, + "loss": 0.5148, + "step": 18925 + }, + { + "epoch": 0.4436552447593429, + "grad_norm": 5.522245407104492, + "learning_rate": 0.00011764846706161778, + "loss": 0.5348, + "step": 18926 + }, + { + "epoch": 0.4436786863341479, + "grad_norm": 3.7790889739990234, + "learning_rate": 0.00011764121647962835, + "loss": 0.366, + "step": 18927 + }, + { + "epoch": 0.4437021279089529, + "grad_norm": 5.125478267669678, + "learning_rate": 0.00011763396580191722, + "loss": 0.6342, + "step": 18928 + }, + { + "epoch": 0.4437255694837579, + "grad_norm": 2.805964946746826, + "learning_rate": 0.00011762671502852367, + "loss": 0.2683, + "step": 18929 + }, + { + "epoch": 0.4437490110585629, + "grad_norm": 0.8327692747116089, + "learning_rate": 0.00011761946415948706, + "loss": 0.047, + "step": 18930 + }, + { + "epoch": 0.4437724526333679, + "grad_norm": 3.6313726902008057, + "learning_rate": 0.00011761221319484674, + "loss": 0.2867, + "step": 18931 + }, + { + "epoch": 0.4437958942081729, + "grad_norm": 3.5519137382507324, + "learning_rate": 0.00011760496213464207, + "loss": 0.494, + "step": 18932 + }, + { + "epoch": 0.4438193357829779, + "grad_norm": 1.3602120876312256, + "learning_rate": 0.00011759771097891239, + "loss": 0.1434, + "step": 18933 + }, + { + "epoch": 0.4438427773577829, + "grad_norm": 5.332483768463135, + "learning_rate": 0.000117590459727697, + "loss": 0.5918, + "step": 18934 + }, + { + "epoch": 0.4438662189325879, + "grad_norm": 3.4668385982513428, + "learning_rate": 0.00011758320838103532, + "loss": 0.2503, + "step": 18935 + }, + { + "epoch": 0.44388966050739287, + "grad_norm": 2.8147943019866943, + "learning_rate": 0.00011757595693896666, + "loss": 1.1006, + "step": 18936 + }, + { + "epoch": 0.44391310208219786, + "grad_norm": 4.133835315704346, + "learning_rate": 0.00011756870540153029, + "loss": 0.2782, + "step": 18937 + }, + { + "epoch": 0.44393654365700286, + "grad_norm": 2.8407044410705566, + "learning_rate": 0.00011756145376876571, + "loss": 0.1484, + "step": 18938 + }, + { + "epoch": 0.44395998523180785, + "grad_norm": 2.1912283897399902, + "learning_rate": 0.00011755420204071217, + "loss": 0.1178, + "step": 18939 + }, + { + "epoch": 0.44398342680661285, + "grad_norm": 2.8644275665283203, + "learning_rate": 0.000117546950217409, + "loss": 0.167, + "step": 18940 + }, + { + "epoch": 0.44400686838141784, + "grad_norm": 2.908405065536499, + "learning_rate": 0.00011753969829889564, + "loss": 0.298, + "step": 18941 + }, + { + "epoch": 0.44403030995622284, + "grad_norm": 3.294192314147949, + "learning_rate": 0.00011753244628521137, + "loss": 0.3817, + "step": 18942 + }, + { + "epoch": 0.44405375153102783, + "grad_norm": 2.5652081966400146, + "learning_rate": 0.00011752519417639549, + "loss": 0.183, + "step": 18943 + }, + { + "epoch": 0.4440771931058328, + "grad_norm": 1.353602409362793, + "learning_rate": 0.00011751794197248749, + "loss": 0.1492, + "step": 18944 + }, + { + "epoch": 0.4441006346806378, + "grad_norm": 2.4394114017486572, + "learning_rate": 0.00011751068967352661, + "loss": 0.1976, + "step": 18945 + }, + { + "epoch": 0.44412407625544287, + "grad_norm": 4.233973503112793, + "learning_rate": 0.00011750343727955224, + "loss": 0.4626, + "step": 18946 + }, + { + "epoch": 0.44414751783024786, + "grad_norm": 3.3275327682495117, + "learning_rate": 0.00011749618479060373, + "loss": 0.372, + "step": 18947 + }, + { + "epoch": 0.44417095940505286, + "grad_norm": 1.4717371463775635, + "learning_rate": 0.00011748893220672044, + "loss": 0.1382, + "step": 18948 + }, + { + "epoch": 0.44419440097985785, + "grad_norm": 1.8441179990768433, + "learning_rate": 0.0001174816795279417, + "loss": 0.2119, + "step": 18949 + }, + { + "epoch": 0.44421784255466285, + "grad_norm": 0.9090055227279663, + "learning_rate": 0.00011747442675430687, + "loss": 0.1491, + "step": 18950 + }, + { + "epoch": 0.44424128412946784, + "grad_norm": 1.4916287660598755, + "learning_rate": 0.0001174671738858553, + "loss": 0.0989, + "step": 18951 + }, + { + "epoch": 0.44426472570427283, + "grad_norm": 0.3790673315525055, + "learning_rate": 0.0001174599209226264, + "loss": 0.0347, + "step": 18952 + }, + { + "epoch": 0.44428816727907783, + "grad_norm": 1.907882809638977, + "learning_rate": 0.00011745266786465942, + "loss": 0.2131, + "step": 18953 + }, + { + "epoch": 0.4443116088538828, + "grad_norm": 4.196138858795166, + "learning_rate": 0.00011744541471199382, + "loss": 0.2944, + "step": 18954 + }, + { + "epoch": 0.4443350504286878, + "grad_norm": 2.841111898422241, + "learning_rate": 0.00011743816146466889, + "loss": 0.2324, + "step": 18955 + }, + { + "epoch": 0.4443584920034928, + "grad_norm": 3.6626555919647217, + "learning_rate": 0.000117430908122724, + "loss": 0.2727, + "step": 18956 + }, + { + "epoch": 0.4443819335782978, + "grad_norm": 2.411853313446045, + "learning_rate": 0.00011742365468619851, + "loss": 0.1043, + "step": 18957 + }, + { + "epoch": 0.4444053751531028, + "grad_norm": 2.594045877456665, + "learning_rate": 0.00011741640115513179, + "loss": 0.2615, + "step": 18958 + }, + { + "epoch": 0.4444288167279078, + "grad_norm": 2.333451509475708, + "learning_rate": 0.00011740914752956318, + "loss": 0.2151, + "step": 18959 + }, + { + "epoch": 0.4444522583027128, + "grad_norm": 3.354665994644165, + "learning_rate": 0.00011740189380953205, + "loss": 0.3606, + "step": 18960 + }, + { + "epoch": 0.4444756998775178, + "grad_norm": 1.0488115549087524, + "learning_rate": 0.00011739463999507776, + "loss": 0.1433, + "step": 18961 + }, + { + "epoch": 0.4444991414523228, + "grad_norm": 3.2805440425872803, + "learning_rate": 0.00011738738608623965, + "loss": 1.4985, + "step": 18962 + }, + { + "epoch": 0.44452258302712777, + "grad_norm": 0.8019036054611206, + "learning_rate": 0.00011738013208305712, + "loss": 0.0987, + "step": 18963 + }, + { + "epoch": 0.44454602460193277, + "grad_norm": 3.996664524078369, + "learning_rate": 0.00011737287798556949, + "loss": 0.6111, + "step": 18964 + }, + { + "epoch": 0.44456946617673776, + "grad_norm": 2.2098512649536133, + "learning_rate": 0.00011736562379381612, + "loss": 1.4635, + "step": 18965 + }, + { + "epoch": 0.44459290775154275, + "grad_norm": 4.5587615966796875, + "learning_rate": 0.00011735836950783643, + "loss": 0.682, + "step": 18966 + }, + { + "epoch": 0.44461634932634775, + "grad_norm": 0.789432942867279, + "learning_rate": 0.0001173511151276697, + "loss": 0.0682, + "step": 18967 + }, + { + "epoch": 0.44463979090115274, + "grad_norm": 4.159919738769531, + "learning_rate": 0.00011734386065335534, + "loss": 0.1964, + "step": 18968 + }, + { + "epoch": 0.44466323247595774, + "grad_norm": 3.0258960723876953, + "learning_rate": 0.0001173366060849327, + "loss": 1.747, + "step": 18969 + }, + { + "epoch": 0.44468667405076273, + "grad_norm": 1.5292296409606934, + "learning_rate": 0.00011732935142244117, + "loss": 0.1621, + "step": 18970 + }, + { + "epoch": 0.4447101156255677, + "grad_norm": 5.07357120513916, + "learning_rate": 0.00011732209666592009, + "loss": 0.778, + "step": 18971 + }, + { + "epoch": 0.4447335572003727, + "grad_norm": 4.031445503234863, + "learning_rate": 0.0001173148418154088, + "loss": 0.2569, + "step": 18972 + }, + { + "epoch": 0.4447569987751777, + "grad_norm": 4.169162273406982, + "learning_rate": 0.00011730758687094669, + "loss": 0.3183, + "step": 18973 + }, + { + "epoch": 0.4447804403499827, + "grad_norm": 2.37557053565979, + "learning_rate": 0.00011730033183257314, + "loss": 0.3044, + "step": 18974 + }, + { + "epoch": 0.4448038819247877, + "grad_norm": 4.014041900634766, + "learning_rate": 0.0001172930767003275, + "loss": 0.4108, + "step": 18975 + }, + { + "epoch": 0.4448273234995927, + "grad_norm": 4.65183162689209, + "learning_rate": 0.00011728582147424915, + "loss": 0.3546, + "step": 18976 + }, + { + "epoch": 0.4448507650743977, + "grad_norm": 4.171074867248535, + "learning_rate": 0.00011727856615437747, + "loss": 0.3094, + "step": 18977 + }, + { + "epoch": 0.4448742066492027, + "grad_norm": 0.27759045362472534, + "learning_rate": 0.00011727131074075176, + "loss": 0.0285, + "step": 18978 + }, + { + "epoch": 0.4448976482240077, + "grad_norm": 4.946386814117432, + "learning_rate": 0.00011726405523341141, + "loss": 0.4366, + "step": 18979 + }, + { + "epoch": 0.4449210897988127, + "grad_norm": 1.6656701564788818, + "learning_rate": 0.00011725679963239589, + "loss": 0.3089, + "step": 18980 + }, + { + "epoch": 0.44494453137361767, + "grad_norm": 3.7899370193481445, + "learning_rate": 0.00011724954393774443, + "loss": 0.3877, + "step": 18981 + }, + { + "epoch": 0.44496797294842266, + "grad_norm": 3.649988889694214, + "learning_rate": 0.00011724228814949646, + "loss": 0.4321, + "step": 18982 + }, + { + "epoch": 0.44499141452322766, + "grad_norm": 2.983898401260376, + "learning_rate": 0.00011723503226769139, + "loss": 1.8077, + "step": 18983 + }, + { + "epoch": 0.44501485609803265, + "grad_norm": 1.3123677968978882, + "learning_rate": 0.00011722777629236852, + "loss": 0.0656, + "step": 18984 + }, + { + "epoch": 0.44503829767283765, + "grad_norm": 3.37042498588562, + "learning_rate": 0.00011722052022356726, + "loss": 0.323, + "step": 18985 + }, + { + "epoch": 0.44506173924764264, + "grad_norm": 4.100193500518799, + "learning_rate": 0.00011721326406132698, + "loss": 0.2569, + "step": 18986 + }, + { + "epoch": 0.44508518082244763, + "grad_norm": 3.824822187423706, + "learning_rate": 0.00011720600780568705, + "loss": 0.3403, + "step": 18987 + }, + { + "epoch": 0.44510862239725263, + "grad_norm": 3.4137823581695557, + "learning_rate": 0.00011719875145668679, + "loss": 0.2276, + "step": 18988 + }, + { + "epoch": 0.4451320639720576, + "grad_norm": 2.610684394836426, + "learning_rate": 0.00011719149501436568, + "loss": 0.2189, + "step": 18989 + }, + { + "epoch": 0.4451555055468626, + "grad_norm": 4.198744297027588, + "learning_rate": 0.00011718423847876303, + "loss": 0.5449, + "step": 18990 + }, + { + "epoch": 0.4451789471216676, + "grad_norm": 1.0707484483718872, + "learning_rate": 0.00011717698184991819, + "loss": 0.1109, + "step": 18991 + }, + { + "epoch": 0.4452023886964726, + "grad_norm": 4.801026821136475, + "learning_rate": 0.00011716972512787054, + "loss": 0.5362, + "step": 18992 + }, + { + "epoch": 0.4452258302712776, + "grad_norm": 0.6022970080375671, + "learning_rate": 0.00011716246831265954, + "loss": 0.0543, + "step": 18993 + }, + { + "epoch": 0.4452492718460826, + "grad_norm": 2.659605026245117, + "learning_rate": 0.00011715521140432446, + "loss": 0.2576, + "step": 18994 + }, + { + "epoch": 0.4452727134208876, + "grad_norm": 3.604010581970215, + "learning_rate": 0.00011714795440290475, + "loss": 0.3672, + "step": 18995 + }, + { + "epoch": 0.4452961549956926, + "grad_norm": 1.8402647972106934, + "learning_rate": 0.00011714069730843976, + "loss": 0.0975, + "step": 18996 + }, + { + "epoch": 0.44531959657049763, + "grad_norm": 4.242069244384766, + "learning_rate": 0.00011713344012096884, + "loss": 0.2909, + "step": 18997 + }, + { + "epoch": 0.4453430381453026, + "grad_norm": 0.6493940949440002, + "learning_rate": 0.00011712618284053139, + "loss": 0.0578, + "step": 18998 + }, + { + "epoch": 0.4453664797201076, + "grad_norm": 3.4887290000915527, + "learning_rate": 0.00011711892546716681, + "loss": 0.4031, + "step": 18999 + }, + { + "epoch": 0.4453899212949126, + "grad_norm": 1.388521432876587, + "learning_rate": 0.00011711166800091441, + "loss": 0.1472, + "step": 19000 + }, + { + "epoch": 0.4454133628697176, + "grad_norm": 1.6929571628570557, + "learning_rate": 0.00011710441044181368, + "loss": 0.1604, + "step": 19001 + }, + { + "epoch": 0.4454368044445226, + "grad_norm": 4.711705684661865, + "learning_rate": 0.00011709715278990395, + "loss": 0.3856, + "step": 19002 + }, + { + "epoch": 0.4454602460193276, + "grad_norm": 3.589526653289795, + "learning_rate": 0.00011708989504522454, + "loss": 0.512, + "step": 19003 + }, + { + "epoch": 0.4454836875941326, + "grad_norm": 2.2828612327575684, + "learning_rate": 0.00011708263720781488, + "loss": 1.3164, + "step": 19004 + }, + { + "epoch": 0.4455071291689376, + "grad_norm": 4.592358112335205, + "learning_rate": 0.00011707537927771436, + "loss": 0.5448, + "step": 19005 + }, + { + "epoch": 0.4455305707437426, + "grad_norm": 3.00972318649292, + "learning_rate": 0.00011706812125496238, + "loss": 0.4158, + "step": 19006 + }, + { + "epoch": 0.4455540123185476, + "grad_norm": 3.5766537189483643, + "learning_rate": 0.00011706086313959825, + "loss": 0.4601, + "step": 19007 + }, + { + "epoch": 0.44557745389335257, + "grad_norm": 4.4196391105651855, + "learning_rate": 0.00011705360493166142, + "loss": 0.3637, + "step": 19008 + }, + { + "epoch": 0.44560089546815757, + "grad_norm": 1.2952414751052856, + "learning_rate": 0.00011704634663119127, + "loss": 0.1251, + "step": 19009 + }, + { + "epoch": 0.44562433704296256, + "grad_norm": 1.2540127038955688, + "learning_rate": 0.00011703908823822711, + "loss": 0.1438, + "step": 19010 + }, + { + "epoch": 0.44564777861776755, + "grad_norm": 1.1625690460205078, + "learning_rate": 0.00011703182975280843, + "loss": 0.1111, + "step": 19011 + }, + { + "epoch": 0.44567122019257255, + "grad_norm": 7.847521781921387, + "learning_rate": 0.00011702457117497454, + "loss": 0.9901, + "step": 19012 + }, + { + "epoch": 0.44569466176737754, + "grad_norm": 3.981004476547241, + "learning_rate": 0.00011701731250476484, + "loss": 0.3593, + "step": 19013 + }, + { + "epoch": 0.44571810334218254, + "grad_norm": 2.6193785667419434, + "learning_rate": 0.00011701005374221874, + "loss": 0.1578, + "step": 19014 + }, + { + "epoch": 0.44574154491698753, + "grad_norm": 1.2943147420883179, + "learning_rate": 0.00011700279488737562, + "loss": 0.1484, + "step": 19015 + }, + { + "epoch": 0.4457649864917925, + "grad_norm": 3.9520986080169678, + "learning_rate": 0.00011699553594027481, + "loss": 0.3965, + "step": 19016 + }, + { + "epoch": 0.4457884280665975, + "grad_norm": 0.6807709336280823, + "learning_rate": 0.00011698827690095578, + "loss": 0.1034, + "step": 19017 + }, + { + "epoch": 0.4458118696414025, + "grad_norm": 1.155656337738037, + "learning_rate": 0.00011698101776945789, + "loss": 0.1281, + "step": 19018 + }, + { + "epoch": 0.4458353112162075, + "grad_norm": 2.491307258605957, + "learning_rate": 0.00011697375854582048, + "loss": 0.2112, + "step": 19019 + }, + { + "epoch": 0.4458587527910125, + "grad_norm": 2.710592746734619, + "learning_rate": 0.00011696649923008303, + "loss": 0.255, + "step": 19020 + }, + { + "epoch": 0.4458821943658175, + "grad_norm": 2.0678653717041016, + "learning_rate": 0.00011695923982228488, + "loss": 0.2014, + "step": 19021 + }, + { + "epoch": 0.4459056359406225, + "grad_norm": 4.136247158050537, + "learning_rate": 0.00011695198032246537, + "loss": 0.4084, + "step": 19022 + }, + { + "epoch": 0.4459290775154275, + "grad_norm": 2.874825954437256, + "learning_rate": 0.00011694472073066394, + "loss": 1.814, + "step": 19023 + }, + { + "epoch": 0.4459525190902325, + "grad_norm": 0.5144577622413635, + "learning_rate": 0.00011693746104692001, + "loss": 0.0363, + "step": 19024 + }, + { + "epoch": 0.4459759606650375, + "grad_norm": 0.6697942614555359, + "learning_rate": 0.00011693020127127294, + "loss": 0.0678, + "step": 19025 + }, + { + "epoch": 0.44599940223984247, + "grad_norm": 4.0978827476501465, + "learning_rate": 0.0001169229414037621, + "loss": 1.7022, + "step": 19026 + }, + { + "epoch": 0.44602284381464746, + "grad_norm": 4.10374641418457, + "learning_rate": 0.00011691568144442693, + "loss": 0.4234, + "step": 19027 + }, + { + "epoch": 0.44604628538945246, + "grad_norm": 2.8543412685394287, + "learning_rate": 0.00011690842139330675, + "loss": 0.2226, + "step": 19028 + }, + { + "epoch": 0.44606972696425745, + "grad_norm": 2.859605073928833, + "learning_rate": 0.00011690116125044104, + "loss": 1.3219, + "step": 19029 + }, + { + "epoch": 0.44609316853906245, + "grad_norm": 3.789041519165039, + "learning_rate": 0.00011689390101586914, + "loss": 0.5117, + "step": 19030 + }, + { + "epoch": 0.44611661011386744, + "grad_norm": 4.676056385040283, + "learning_rate": 0.00011688664068963049, + "loss": 0.3137, + "step": 19031 + }, + { + "epoch": 0.44614005168867243, + "grad_norm": 3.451141834259033, + "learning_rate": 0.00011687938027176441, + "loss": 0.3698, + "step": 19032 + }, + { + "epoch": 0.44616349326347743, + "grad_norm": 3.094693660736084, + "learning_rate": 0.00011687211976231037, + "loss": 0.4123, + "step": 19033 + }, + { + "epoch": 0.4461869348382824, + "grad_norm": 2.3028790950775146, + "learning_rate": 0.00011686485916130769, + "loss": 0.1575, + "step": 19034 + }, + { + "epoch": 0.4462103764130874, + "grad_norm": 4.893887996673584, + "learning_rate": 0.00011685759846879585, + "loss": 0.3445, + "step": 19035 + }, + { + "epoch": 0.4462338179878924, + "grad_norm": 2.90087628364563, + "learning_rate": 0.00011685033768481419, + "loss": 0.1731, + "step": 19036 + }, + { + "epoch": 0.4462572595626974, + "grad_norm": 3.630258798599243, + "learning_rate": 0.00011684307680940215, + "loss": 0.5143, + "step": 19037 + }, + { + "epoch": 0.4462807011375024, + "grad_norm": 3.1674695014953613, + "learning_rate": 0.00011683581584259906, + "loss": 0.3492, + "step": 19038 + }, + { + "epoch": 0.4463041427123074, + "grad_norm": 3.1872658729553223, + "learning_rate": 0.00011682855478444439, + "loss": 0.4595, + "step": 19039 + }, + { + "epoch": 0.4463275842871124, + "grad_norm": 3.1031136512756348, + "learning_rate": 0.0001168212936349775, + "loss": 0.3309, + "step": 19040 + }, + { + "epoch": 0.4463510258619174, + "grad_norm": 2.07839298248291, + "learning_rate": 0.0001168140323942378, + "loss": 0.2461, + "step": 19041 + }, + { + "epoch": 0.4463744674367224, + "grad_norm": 2.3151071071624756, + "learning_rate": 0.00011680677106226468, + "loss": 0.1737, + "step": 19042 + }, + { + "epoch": 0.44639790901152737, + "grad_norm": 3.3590285778045654, + "learning_rate": 0.00011679950963909756, + "loss": 0.3305, + "step": 19043 + }, + { + "epoch": 0.44642135058633237, + "grad_norm": 3.0897598266601562, + "learning_rate": 0.00011679224812477583, + "loss": 0.3391, + "step": 19044 + }, + { + "epoch": 0.44644479216113736, + "grad_norm": 4.2750935554504395, + "learning_rate": 0.00011678498651933887, + "loss": 0.6045, + "step": 19045 + }, + { + "epoch": 0.44646823373594235, + "grad_norm": 3.010509729385376, + "learning_rate": 0.00011677772482282609, + "loss": 0.3182, + "step": 19046 + }, + { + "epoch": 0.44649167531074735, + "grad_norm": 3.7555487155914307, + "learning_rate": 0.00011677046303527695, + "loss": 0.3839, + "step": 19047 + }, + { + "epoch": 0.4465151168855524, + "grad_norm": 1.194880485534668, + "learning_rate": 0.00011676320115673076, + "loss": 0.149, + "step": 19048 + }, + { + "epoch": 0.4465385584603574, + "grad_norm": 2.1098239421844482, + "learning_rate": 0.000116755939187227, + "loss": 1.7534, + "step": 19049 + }, + { + "epoch": 0.4465620000351624, + "grad_norm": 2.5644912719726562, + "learning_rate": 0.00011674867712680504, + "loss": 0.2408, + "step": 19050 + }, + { + "epoch": 0.4465854416099674, + "grad_norm": 0.4092073142528534, + "learning_rate": 0.00011674141497550426, + "loss": 0.0489, + "step": 19051 + }, + { + "epoch": 0.4466088831847724, + "grad_norm": 3.9195797443389893, + "learning_rate": 0.0001167341527333641, + "loss": 0.5265, + "step": 19052 + }, + { + "epoch": 0.44663232475957737, + "grad_norm": 2.685755491256714, + "learning_rate": 0.000116726890400424, + "loss": 0.2726, + "step": 19053 + }, + { + "epoch": 0.44665576633438236, + "grad_norm": 2.9616448879241943, + "learning_rate": 0.0001167196279767233, + "loss": 0.4278, + "step": 19054 + }, + { + "epoch": 0.44667920790918736, + "grad_norm": 0.6563368439674377, + "learning_rate": 0.0001167123654623014, + "loss": 0.0466, + "step": 19055 + }, + { + "epoch": 0.44670264948399235, + "grad_norm": 2.2467997074127197, + "learning_rate": 0.00011670510285719776, + "loss": 0.2872, + "step": 19056 + }, + { + "epoch": 0.44672609105879735, + "grad_norm": 0.5227845907211304, + "learning_rate": 0.00011669784016145173, + "loss": 0.0535, + "step": 19057 + }, + { + "epoch": 0.44674953263360234, + "grad_norm": 3.404921770095825, + "learning_rate": 0.00011669057737510279, + "loss": 0.443, + "step": 19058 + }, + { + "epoch": 0.44677297420840734, + "grad_norm": 3.041219472885132, + "learning_rate": 0.0001166833144981903, + "loss": 0.2543, + "step": 19059 + }, + { + "epoch": 0.44679641578321233, + "grad_norm": 2.934807300567627, + "learning_rate": 0.00011667605153075367, + "loss": 0.3358, + "step": 19060 + }, + { + "epoch": 0.4468198573580173, + "grad_norm": 3.589881420135498, + "learning_rate": 0.00011666878847283231, + "loss": 0.4794, + "step": 19061 + }, + { + "epoch": 0.4468432989328223, + "grad_norm": 2.985335350036621, + "learning_rate": 0.00011666152532446564, + "loss": 0.2301, + "step": 19062 + }, + { + "epoch": 0.4468667405076273, + "grad_norm": 2.1932759284973145, + "learning_rate": 0.00011665426208569307, + "loss": 0.3317, + "step": 19063 + }, + { + "epoch": 0.4468901820824323, + "grad_norm": 3.7606775760650635, + "learning_rate": 0.000116646998756554, + "loss": 0.2209, + "step": 19064 + }, + { + "epoch": 0.4469136236572373, + "grad_norm": 0.811804473400116, + "learning_rate": 0.00011663973533708783, + "loss": 0.1311, + "step": 19065 + }, + { + "epoch": 0.4469370652320423, + "grad_norm": 7.214784145355225, + "learning_rate": 0.00011663247182733403, + "loss": 1.1625, + "step": 19066 + }, + { + "epoch": 0.4469605068068473, + "grad_norm": 3.2111716270446777, + "learning_rate": 0.00011662520822733195, + "loss": 0.2788, + "step": 19067 + }, + { + "epoch": 0.4469839483816523, + "grad_norm": 6.599725246429443, + "learning_rate": 0.000116617944537121, + "loss": 0.4458, + "step": 19068 + }, + { + "epoch": 0.4470073899564573, + "grad_norm": 3.139756679534912, + "learning_rate": 0.00011661068075674065, + "loss": 0.3361, + "step": 19069 + }, + { + "epoch": 0.4470308315312623, + "grad_norm": 3.328056812286377, + "learning_rate": 0.00011660341688623027, + "loss": 1.1837, + "step": 19070 + }, + { + "epoch": 0.44705427310606727, + "grad_norm": 3.4281113147735596, + "learning_rate": 0.00011659615292562924, + "loss": 0.4002, + "step": 19071 + }, + { + "epoch": 0.44707771468087226, + "grad_norm": 2.541188955307007, + "learning_rate": 0.00011658888887497707, + "loss": 0.3667, + "step": 19072 + }, + { + "epoch": 0.44710115625567726, + "grad_norm": 2.744387149810791, + "learning_rate": 0.00011658162473431314, + "loss": 0.2803, + "step": 19073 + }, + { + "epoch": 0.44712459783048225, + "grad_norm": 1.848451852798462, + "learning_rate": 0.00011657436050367678, + "loss": 0.1863, + "step": 19074 + }, + { + "epoch": 0.44714803940528725, + "grad_norm": 1.3598524332046509, + "learning_rate": 0.00011656709618310752, + "loss": 0.1542, + "step": 19075 + }, + { + "epoch": 0.44717148098009224, + "grad_norm": 3.252824544906616, + "learning_rate": 0.00011655983177264472, + "loss": 0.4853, + "step": 19076 + }, + { + "epoch": 0.44719492255489723, + "grad_norm": 0.9915542006492615, + "learning_rate": 0.00011655256727232779, + "loss": 0.1382, + "step": 19077 + }, + { + "epoch": 0.44721836412970223, + "grad_norm": 0.5704331398010254, + "learning_rate": 0.00011654530268219619, + "loss": 0.0662, + "step": 19078 + }, + { + "epoch": 0.4472418057045072, + "grad_norm": 3.470686435699463, + "learning_rate": 0.0001165380380022893, + "loss": 0.3862, + "step": 19079 + }, + { + "epoch": 0.4472652472793122, + "grad_norm": 2.908536672592163, + "learning_rate": 0.00011653077323264656, + "loss": 0.1873, + "step": 19080 + }, + { + "epoch": 0.4472886888541172, + "grad_norm": 1.0202720165252686, + "learning_rate": 0.0001165235083733074, + "loss": 0.0724, + "step": 19081 + }, + { + "epoch": 0.4473121304289222, + "grad_norm": 1.197100281715393, + "learning_rate": 0.00011651624342431118, + "loss": 0.1652, + "step": 19082 + }, + { + "epoch": 0.4473355720037272, + "grad_norm": 2.882204294204712, + "learning_rate": 0.00011650897838569735, + "loss": 0.2729, + "step": 19083 + }, + { + "epoch": 0.4473590135785322, + "grad_norm": 1.2579065561294556, + "learning_rate": 0.00011650171325750536, + "loss": 0.1381, + "step": 19084 + }, + { + "epoch": 0.4473824551533372, + "grad_norm": 4.832923889160156, + "learning_rate": 0.00011649444803977463, + "loss": 0.5216, + "step": 19085 + }, + { + "epoch": 0.4474058967281422, + "grad_norm": 5.825641632080078, + "learning_rate": 0.00011648718273254453, + "loss": 0.4982, + "step": 19086 + }, + { + "epoch": 0.4474293383029472, + "grad_norm": 6.195641994476318, + "learning_rate": 0.00011647991733585453, + "loss": 0.7814, + "step": 19087 + }, + { + "epoch": 0.44745277987775217, + "grad_norm": 6.3322296142578125, + "learning_rate": 0.00011647265184974406, + "loss": 0.5443, + "step": 19088 + }, + { + "epoch": 0.44747622145255717, + "grad_norm": 1.572197437286377, + "learning_rate": 0.00011646538627425246, + "loss": 0.1649, + "step": 19089 + }, + { + "epoch": 0.44749966302736216, + "grad_norm": 3.560286045074463, + "learning_rate": 0.00011645812060941924, + "loss": 0.3813, + "step": 19090 + }, + { + "epoch": 0.44752310460216715, + "grad_norm": 0.9298381805419922, + "learning_rate": 0.0001164508548552838, + "loss": 0.0718, + "step": 19091 + }, + { + "epoch": 0.44754654617697215, + "grad_norm": 2.3271970748901367, + "learning_rate": 0.00011644358901188554, + "loss": 0.2585, + "step": 19092 + }, + { + "epoch": 0.44756998775177714, + "grad_norm": 2.245544910430908, + "learning_rate": 0.00011643632307926391, + "loss": 0.3539, + "step": 19093 + }, + { + "epoch": 0.44759342932658214, + "grad_norm": 0.994094729423523, + "learning_rate": 0.00011642905705745835, + "loss": 0.1785, + "step": 19094 + }, + { + "epoch": 0.44761687090138713, + "grad_norm": 2.57543683052063, + "learning_rate": 0.00011642179094650824, + "loss": 0.2073, + "step": 19095 + }, + { + "epoch": 0.4476403124761921, + "grad_norm": 2.6717851161956787, + "learning_rate": 0.00011641452474645302, + "loss": 0.3721, + "step": 19096 + }, + { + "epoch": 0.4476637540509971, + "grad_norm": 3.036947250366211, + "learning_rate": 0.00011640725845733215, + "loss": 0.1721, + "step": 19097 + }, + { + "epoch": 0.4476871956258021, + "grad_norm": 3.3235955238342285, + "learning_rate": 0.00011639999207918506, + "loss": 0.5385, + "step": 19098 + }, + { + "epoch": 0.44771063720060716, + "grad_norm": 0.9816124439239502, + "learning_rate": 0.00011639272561205112, + "loss": 0.0847, + "step": 19099 + }, + { + "epoch": 0.44773407877541216, + "grad_norm": 3.397456645965576, + "learning_rate": 0.00011638545905596981, + "loss": 0.3584, + "step": 19100 + }, + { + "epoch": 0.44775752035021715, + "grad_norm": 1.6975942850112915, + "learning_rate": 0.00011637819241098049, + "loss": 0.1903, + "step": 19101 + }, + { + "epoch": 0.44778096192502215, + "grad_norm": 3.1019444465637207, + "learning_rate": 0.00011637092567712267, + "loss": 0.3257, + "step": 19102 + }, + { + "epoch": 0.44780440349982714, + "grad_norm": 4.020751953125, + "learning_rate": 0.00011636365885443574, + "loss": 0.5491, + "step": 19103 + }, + { + "epoch": 0.44782784507463214, + "grad_norm": 3.218780279159546, + "learning_rate": 0.00011635639194295915, + "loss": 0.1932, + "step": 19104 + }, + { + "epoch": 0.44785128664943713, + "grad_norm": 4.548428058624268, + "learning_rate": 0.00011634912494273231, + "loss": 0.5301, + "step": 19105 + }, + { + "epoch": 0.4478747282242421, + "grad_norm": 2.98687744140625, + "learning_rate": 0.00011634185785379468, + "loss": 0.4615, + "step": 19106 + }, + { + "epoch": 0.4478981697990471, + "grad_norm": 2.797334909439087, + "learning_rate": 0.00011633459067618564, + "loss": 0.2902, + "step": 19107 + }, + { + "epoch": 0.4479216113738521, + "grad_norm": 1.2899491786956787, + "learning_rate": 0.00011632732340994466, + "loss": 0.1312, + "step": 19108 + }, + { + "epoch": 0.4479450529486571, + "grad_norm": 2.534747838973999, + "learning_rate": 0.00011632005605511117, + "loss": 1.5701, + "step": 19109 + }, + { + "epoch": 0.4479684945234621, + "grad_norm": 1.783400058746338, + "learning_rate": 0.00011631278861172458, + "loss": 0.2029, + "step": 19110 + }, + { + "epoch": 0.4479919360982671, + "grad_norm": 1.062567114830017, + "learning_rate": 0.00011630552107982435, + "loss": 0.1904, + "step": 19111 + }, + { + "epoch": 0.4480153776730721, + "grad_norm": 2.388930559158325, + "learning_rate": 0.00011629825345944991, + "loss": 0.1692, + "step": 19112 + }, + { + "epoch": 0.4480388192478771, + "grad_norm": 1.5189217329025269, + "learning_rate": 0.00011629098575064067, + "loss": 0.1595, + "step": 19113 + }, + { + "epoch": 0.4480622608226821, + "grad_norm": 4.390067100524902, + "learning_rate": 0.00011628371795343611, + "loss": 0.3663, + "step": 19114 + }, + { + "epoch": 0.4480857023974871, + "grad_norm": 4.95844030380249, + "learning_rate": 0.00011627645006787562, + "loss": 0.4455, + "step": 19115 + }, + { + "epoch": 0.44810914397229207, + "grad_norm": 4.06093168258667, + "learning_rate": 0.00011626918209399866, + "loss": 0.3362, + "step": 19116 + }, + { + "epoch": 0.44813258554709706, + "grad_norm": 0.6877337098121643, + "learning_rate": 0.00011626191403184468, + "loss": 0.0584, + "step": 19117 + }, + { + "epoch": 0.44815602712190206, + "grad_norm": 3.433708667755127, + "learning_rate": 0.00011625464588145307, + "loss": 0.4707, + "step": 19118 + }, + { + "epoch": 0.44817946869670705, + "grad_norm": 3.793525457382202, + "learning_rate": 0.00011624737764286328, + "loss": 0.3812, + "step": 19119 + }, + { + "epoch": 0.44820291027151204, + "grad_norm": 3.8006317615509033, + "learning_rate": 0.00011624010931611479, + "loss": 0.3314, + "step": 19120 + }, + { + "epoch": 0.44822635184631704, + "grad_norm": 1.804794192314148, + "learning_rate": 0.00011623284090124703, + "loss": 0.1534, + "step": 19121 + }, + { + "epoch": 0.44824979342112203, + "grad_norm": 1.585745096206665, + "learning_rate": 0.00011622557239829939, + "loss": 0.2572, + "step": 19122 + }, + { + "epoch": 0.44827323499592703, + "grad_norm": 1.2615457773208618, + "learning_rate": 0.00011621830380731133, + "loss": 0.0749, + "step": 19123 + }, + { + "epoch": 0.448296676570732, + "grad_norm": 2.7113075256347656, + "learning_rate": 0.00011621103512832232, + "loss": 0.2514, + "step": 19124 + }, + { + "epoch": 0.448320118145537, + "grad_norm": 2.8948822021484375, + "learning_rate": 0.00011620376636137175, + "loss": 0.1641, + "step": 19125 + }, + { + "epoch": 0.448343559720342, + "grad_norm": 2.610699415206909, + "learning_rate": 0.00011619649750649913, + "loss": 0.1347, + "step": 19126 + }, + { + "epoch": 0.448367001295147, + "grad_norm": 1.9279829263687134, + "learning_rate": 0.00011618922856374384, + "loss": 0.2627, + "step": 19127 + }, + { + "epoch": 0.448390442869952, + "grad_norm": 4.2683258056640625, + "learning_rate": 0.0001161819595331453, + "loss": 0.3719, + "step": 19128 + }, + { + "epoch": 0.448413884444757, + "grad_norm": 1.0663832426071167, + "learning_rate": 0.00011617469041474305, + "loss": 0.0999, + "step": 19129 + }, + { + "epoch": 0.448437326019562, + "grad_norm": 4.698899269104004, + "learning_rate": 0.00011616742120857644, + "loss": 0.5075, + "step": 19130 + }, + { + "epoch": 0.448460767594367, + "grad_norm": 0.8341051340103149, + "learning_rate": 0.00011616015191468495, + "loss": 0.1445, + "step": 19131 + }, + { + "epoch": 0.448484209169172, + "grad_norm": 1.9500036239624023, + "learning_rate": 0.00011615288253310803, + "loss": 0.0713, + "step": 19132 + }, + { + "epoch": 0.44850765074397697, + "grad_norm": 4.200876235961914, + "learning_rate": 0.00011614561306388513, + "loss": 0.3006, + "step": 19133 + }, + { + "epoch": 0.44853109231878197, + "grad_norm": 1.4623219966888428, + "learning_rate": 0.00011613834350705566, + "loss": 0.1764, + "step": 19134 + }, + { + "epoch": 0.44855453389358696, + "grad_norm": 2.388735771179199, + "learning_rate": 0.00011613107386265912, + "loss": 1.755, + "step": 19135 + }, + { + "epoch": 0.44857797546839195, + "grad_norm": 2.4343152046203613, + "learning_rate": 0.00011612380413073489, + "loss": 0.1555, + "step": 19136 + }, + { + "epoch": 0.44860141704319695, + "grad_norm": 2.213792324066162, + "learning_rate": 0.00011611653431132244, + "loss": 0.188, + "step": 19137 + }, + { + "epoch": 0.44862485861800194, + "grad_norm": 0.7843689918518066, + "learning_rate": 0.0001161092644044612, + "loss": 0.0711, + "step": 19138 + }, + { + "epoch": 0.44864830019280694, + "grad_norm": 5.895503997802734, + "learning_rate": 0.00011610199441019067, + "loss": 0.5486, + "step": 19139 + }, + { + "epoch": 0.44867174176761193, + "grad_norm": 4.584117889404297, + "learning_rate": 0.00011609472432855027, + "loss": 0.5138, + "step": 19140 + }, + { + "epoch": 0.4486951833424169, + "grad_norm": 0.7241494655609131, + "learning_rate": 0.00011608745415957944, + "loss": 0.0736, + "step": 19141 + }, + { + "epoch": 0.4487186249172219, + "grad_norm": 3.304591655731201, + "learning_rate": 0.00011608018390331762, + "loss": 0.2433, + "step": 19142 + }, + { + "epoch": 0.4487420664920269, + "grad_norm": 2.6187212467193604, + "learning_rate": 0.00011607291355980427, + "loss": 1.2844, + "step": 19143 + }, + { + "epoch": 0.4487655080668319, + "grad_norm": 2.2200207710266113, + "learning_rate": 0.00011606564312907884, + "loss": 0.2579, + "step": 19144 + }, + { + "epoch": 0.4487889496416369, + "grad_norm": 1.007218837738037, + "learning_rate": 0.0001160583726111808, + "loss": 0.0822, + "step": 19145 + }, + { + "epoch": 0.4488123912164419, + "grad_norm": 4.204614162445068, + "learning_rate": 0.00011605110200614957, + "loss": 0.319, + "step": 19146 + }, + { + "epoch": 0.4488358327912469, + "grad_norm": 3.4705395698547363, + "learning_rate": 0.00011604383131402459, + "loss": 0.2762, + "step": 19147 + }, + { + "epoch": 0.4488592743660519, + "grad_norm": 0.7324798703193665, + "learning_rate": 0.00011603656053484534, + "loss": 0.0824, + "step": 19148 + }, + { + "epoch": 0.4488827159408569, + "grad_norm": 2.9742774963378906, + "learning_rate": 0.00011602928966865127, + "loss": 0.276, + "step": 19149 + }, + { + "epoch": 0.4489061575156619, + "grad_norm": 4.5833024978637695, + "learning_rate": 0.0001160220187154818, + "loss": 0.4351, + "step": 19150 + }, + { + "epoch": 0.4489295990904669, + "grad_norm": 2.906057596206665, + "learning_rate": 0.00011601474767537644, + "loss": 0.2743, + "step": 19151 + }, + { + "epoch": 0.4489530406652719, + "grad_norm": 2.8812222480773926, + "learning_rate": 0.0001160074765483746, + "loss": 0.2347, + "step": 19152 + }, + { + "epoch": 0.4489764822400769, + "grad_norm": 3.600637197494507, + "learning_rate": 0.00011600020533451572, + "loss": 0.2955, + "step": 19153 + }, + { + "epoch": 0.4489999238148819, + "grad_norm": 2.7056591510772705, + "learning_rate": 0.00011599293403383931, + "loss": 0.3403, + "step": 19154 + }, + { + "epoch": 0.4490233653896869, + "grad_norm": 2.7849888801574707, + "learning_rate": 0.00011598566264638478, + "loss": 1.5721, + "step": 19155 + }, + { + "epoch": 0.4490468069644919, + "grad_norm": 1.0813871622085571, + "learning_rate": 0.00011597839117219154, + "loss": 0.0816, + "step": 19156 + }, + { + "epoch": 0.4490702485392969, + "grad_norm": 3.860314130783081, + "learning_rate": 0.00011597111961129916, + "loss": 0.4509, + "step": 19157 + }, + { + "epoch": 0.4490936901141019, + "grad_norm": 2.7068874835968018, + "learning_rate": 0.00011596384796374702, + "loss": 0.3794, + "step": 19158 + }, + { + "epoch": 0.4491171316889069, + "grad_norm": 1.5829455852508545, + "learning_rate": 0.0001159565762295746, + "loss": 0.0988, + "step": 19159 + }, + { + "epoch": 0.4491405732637119, + "grad_norm": 4.036509037017822, + "learning_rate": 0.00011594930440882135, + "loss": 0.3115, + "step": 19160 + }, + { + "epoch": 0.44916401483851687, + "grad_norm": 4.466240406036377, + "learning_rate": 0.00011594203250152671, + "loss": 0.7622, + "step": 19161 + }, + { + "epoch": 0.44918745641332186, + "grad_norm": 0.8880118131637573, + "learning_rate": 0.00011593476050773016, + "loss": 0.0715, + "step": 19162 + }, + { + "epoch": 0.44921089798812686, + "grad_norm": 2.332249641418457, + "learning_rate": 0.00011592748842747115, + "loss": 0.2092, + "step": 19163 + }, + { + "epoch": 0.44923433956293185, + "grad_norm": 3.6354668140411377, + "learning_rate": 0.00011592021626078914, + "loss": 0.3404, + "step": 19164 + }, + { + "epoch": 0.44925778113773684, + "grad_norm": 0.5299221873283386, + "learning_rate": 0.0001159129440077236, + "loss": 0.0565, + "step": 19165 + }, + { + "epoch": 0.44928122271254184, + "grad_norm": 0.9305745959281921, + "learning_rate": 0.00011590567166831397, + "loss": 0.4262, + "step": 19166 + }, + { + "epoch": 0.44930466428734683, + "grad_norm": 3.2434728145599365, + "learning_rate": 0.00011589839924259974, + "loss": 0.3897, + "step": 19167 + }, + { + "epoch": 0.4493281058621518, + "grad_norm": 3.9268975257873535, + "learning_rate": 0.00011589112673062031, + "loss": 0.4416, + "step": 19168 + }, + { + "epoch": 0.4493515474369568, + "grad_norm": 2.324432611465454, + "learning_rate": 0.00011588385413241519, + "loss": 0.1582, + "step": 19169 + }, + { + "epoch": 0.4493749890117618, + "grad_norm": 2.0300683975219727, + "learning_rate": 0.00011587658144802383, + "loss": 1.0438, + "step": 19170 + }, + { + "epoch": 0.4493984305865668, + "grad_norm": 4.016324520111084, + "learning_rate": 0.00011586930867748571, + "loss": 0.5318, + "step": 19171 + }, + { + "epoch": 0.4494218721613718, + "grad_norm": 3.0210769176483154, + "learning_rate": 0.00011586203582084025, + "loss": 0.3248, + "step": 19172 + }, + { + "epoch": 0.4494453137361768, + "grad_norm": 0.6783554553985596, + "learning_rate": 0.00011585476287812698, + "loss": 0.0571, + "step": 19173 + }, + { + "epoch": 0.4494687553109818, + "grad_norm": 2.8977346420288086, + "learning_rate": 0.00011584748984938527, + "loss": 0.3463, + "step": 19174 + }, + { + "epoch": 0.4494921968857868, + "grad_norm": 3.4669599533081055, + "learning_rate": 0.00011584021673465467, + "loss": 0.3014, + "step": 19175 + }, + { + "epoch": 0.4495156384605918, + "grad_norm": 0.9956746101379395, + "learning_rate": 0.00011583294353397458, + "loss": 0.1361, + "step": 19176 + }, + { + "epoch": 0.4495390800353968, + "grad_norm": 3.440206289291382, + "learning_rate": 0.00011582567024738453, + "loss": 0.4952, + "step": 19177 + }, + { + "epoch": 0.44956252161020177, + "grad_norm": 3.212143898010254, + "learning_rate": 0.00011581839687492391, + "loss": 0.5004, + "step": 19178 + }, + { + "epoch": 0.44958596318500677, + "grad_norm": 2.783109664916992, + "learning_rate": 0.00011581112341663225, + "loss": 1.8412, + "step": 19179 + }, + { + "epoch": 0.44960940475981176, + "grad_norm": 3.0714962482452393, + "learning_rate": 0.00011580384987254898, + "loss": 0.239, + "step": 19180 + }, + { + "epoch": 0.44963284633461675, + "grad_norm": 2.963217258453369, + "learning_rate": 0.00011579657624271358, + "loss": 0.3889, + "step": 19181 + }, + { + "epoch": 0.44965628790942175, + "grad_norm": 3.221910238265991, + "learning_rate": 0.00011578930252716551, + "loss": 0.388, + "step": 19182 + }, + { + "epoch": 0.44967972948422674, + "grad_norm": 1.9488624334335327, + "learning_rate": 0.00011578202872594424, + "loss": 0.1913, + "step": 19183 + }, + { + "epoch": 0.44970317105903174, + "grad_norm": 2.481390953063965, + "learning_rate": 0.00011577475483908926, + "loss": 0.2659, + "step": 19184 + }, + { + "epoch": 0.44972661263383673, + "grad_norm": 2.952183485031128, + "learning_rate": 0.00011576748086663998, + "loss": 0.3978, + "step": 19185 + }, + { + "epoch": 0.4497500542086417, + "grad_norm": 2.8248860836029053, + "learning_rate": 0.0001157602068086359, + "loss": 0.2086, + "step": 19186 + }, + { + "epoch": 0.4497734957834467, + "grad_norm": 2.725740671157837, + "learning_rate": 0.00011575293266511654, + "loss": 1.2604, + "step": 19187 + }, + { + "epoch": 0.4497969373582517, + "grad_norm": 3.20908784866333, + "learning_rate": 0.00011574565843612128, + "loss": 0.342, + "step": 19188 + }, + { + "epoch": 0.4498203789330567, + "grad_norm": 2.5440261363983154, + "learning_rate": 0.00011573838412168967, + "loss": 0.1843, + "step": 19189 + }, + { + "epoch": 0.4498438205078617, + "grad_norm": 1.6316962242126465, + "learning_rate": 0.00011573110972186115, + "loss": 1.5475, + "step": 19190 + }, + { + "epoch": 0.4498672620826667, + "grad_norm": 3.1900861263275146, + "learning_rate": 0.00011572383523667515, + "loss": 0.4596, + "step": 19191 + }, + { + "epoch": 0.4498907036574717, + "grad_norm": 4.000682830810547, + "learning_rate": 0.00011571656066617118, + "loss": 0.3442, + "step": 19192 + }, + { + "epoch": 0.4499141452322767, + "grad_norm": 3.8621089458465576, + "learning_rate": 0.00011570928601038873, + "loss": 0.2403, + "step": 19193 + }, + { + "epoch": 0.4499375868070817, + "grad_norm": 3.2465364933013916, + "learning_rate": 0.00011570201126936723, + "loss": 0.3792, + "step": 19194 + }, + { + "epoch": 0.4499610283818867, + "grad_norm": 3.2217295169830322, + "learning_rate": 0.00011569473644314622, + "loss": 1.7682, + "step": 19195 + }, + { + "epoch": 0.44998446995669167, + "grad_norm": 4.817076206207275, + "learning_rate": 0.00011568746153176511, + "loss": 0.4009, + "step": 19196 + }, + { + "epoch": 0.45000791153149666, + "grad_norm": 3.5384910106658936, + "learning_rate": 0.00011568018653526338, + "loss": 0.3513, + "step": 19197 + }, + { + "epoch": 0.45003135310630166, + "grad_norm": 1.4152215719223022, + "learning_rate": 0.00011567291145368051, + "loss": 0.183, + "step": 19198 + }, + { + "epoch": 0.45005479468110665, + "grad_norm": 3.7803502082824707, + "learning_rate": 0.000115665636287056, + "loss": 0.368, + "step": 19199 + }, + { + "epoch": 0.45007823625591165, + "grad_norm": 2.968597888946533, + "learning_rate": 0.0001156583610354293, + "loss": 0.2788, + "step": 19200 + }, + { + "epoch": 0.45010167783071664, + "grad_norm": 3.2852213382720947, + "learning_rate": 0.00011565108569883989, + "loss": 0.2285, + "step": 19201 + }, + { + "epoch": 0.4501251194055217, + "grad_norm": 3.268965244293213, + "learning_rate": 0.00011564381027732727, + "loss": 1.5946, + "step": 19202 + }, + { + "epoch": 0.4501485609803267, + "grad_norm": 3.182166814804077, + "learning_rate": 0.00011563653477093088, + "loss": 0.3263, + "step": 19203 + }, + { + "epoch": 0.4501720025551317, + "grad_norm": 5.038103103637695, + "learning_rate": 0.00011562925917969019, + "loss": 0.4764, + "step": 19204 + }, + { + "epoch": 0.4501954441299367, + "grad_norm": 1.280915379524231, + "learning_rate": 0.00011562198350364473, + "loss": 0.1947, + "step": 19205 + }, + { + "epoch": 0.45021888570474167, + "grad_norm": 3.410325765609741, + "learning_rate": 0.00011561470774283394, + "loss": 0.4744, + "step": 19206 + }, + { + "epoch": 0.45024232727954666, + "grad_norm": 5.141796112060547, + "learning_rate": 0.00011560743189729728, + "loss": 0.5246, + "step": 19207 + }, + { + "epoch": 0.45026576885435166, + "grad_norm": 3.150463342666626, + "learning_rate": 0.0001156001559670743, + "loss": 1.7297, + "step": 19208 + }, + { + "epoch": 0.45028921042915665, + "grad_norm": 3.4476866722106934, + "learning_rate": 0.00011559287995220442, + "loss": 0.0698, + "step": 19209 + }, + { + "epoch": 0.45031265200396164, + "grad_norm": 3.3246943950653076, + "learning_rate": 0.00011558560385272712, + "loss": 0.2594, + "step": 19210 + }, + { + "epoch": 0.45033609357876664, + "grad_norm": 0.7909008264541626, + "learning_rate": 0.00011557832766868191, + "loss": 0.0608, + "step": 19211 + }, + { + "epoch": 0.45035953515357163, + "grad_norm": 0.792496383190155, + "learning_rate": 0.00011557105140010827, + "loss": 0.0524, + "step": 19212 + }, + { + "epoch": 0.4503829767283766, + "grad_norm": 0.7349861264228821, + "learning_rate": 0.00011556377504704563, + "loss": 0.0761, + "step": 19213 + }, + { + "epoch": 0.4504064183031816, + "grad_norm": 3.4436817169189453, + "learning_rate": 0.00011555649860953352, + "loss": 0.3072, + "step": 19214 + }, + { + "epoch": 0.4504298598779866, + "grad_norm": 2.446336507797241, + "learning_rate": 0.00011554922208761146, + "loss": 0.2777, + "step": 19215 + }, + { + "epoch": 0.4504533014527916, + "grad_norm": 4.095585823059082, + "learning_rate": 0.00011554194548131883, + "loss": 0.474, + "step": 19216 + }, + { + "epoch": 0.4504767430275966, + "grad_norm": 0.8169788122177124, + "learning_rate": 0.00011553466879069514, + "loss": 0.037, + "step": 19217 + }, + { + "epoch": 0.4505001846024016, + "grad_norm": 2.3954129219055176, + "learning_rate": 0.00011552739201577995, + "loss": 0.2562, + "step": 19218 + }, + { + "epoch": 0.4505236261772066, + "grad_norm": 3.763542652130127, + "learning_rate": 0.00011552011515661271, + "loss": 0.2983, + "step": 19219 + }, + { + "epoch": 0.4505470677520116, + "grad_norm": 1.0550644397735596, + "learning_rate": 0.00011551283821323285, + "loss": 0.074, + "step": 19220 + }, + { + "epoch": 0.4505705093268166, + "grad_norm": 1.4708499908447266, + "learning_rate": 0.00011550556118567992, + "loss": 0.2158, + "step": 19221 + }, + { + "epoch": 0.4505939509016216, + "grad_norm": 3.7701635360717773, + "learning_rate": 0.00011549828407399337, + "loss": 0.4503, + "step": 19222 + }, + { + "epoch": 0.45061739247642657, + "grad_norm": 1.0026435852050781, + "learning_rate": 0.00011549100687821266, + "loss": 0.1, + "step": 19223 + }, + { + "epoch": 0.45064083405123156, + "grad_norm": 3.0958032608032227, + "learning_rate": 0.00011548372959837737, + "loss": 0.2907, + "step": 19224 + }, + { + "epoch": 0.45066427562603656, + "grad_norm": 4.877163410186768, + "learning_rate": 0.0001154764522345269, + "loss": 0.5197, + "step": 19225 + }, + { + "epoch": 0.45068771720084155, + "grad_norm": 1.1295740604400635, + "learning_rate": 0.00011546917478670077, + "loss": 0.0897, + "step": 19226 + }, + { + "epoch": 0.45071115877564655, + "grad_norm": 1.9894987344741821, + "learning_rate": 0.00011546189725493845, + "loss": 0.2797, + "step": 19227 + }, + { + "epoch": 0.45073460035045154, + "grad_norm": 2.9053356647491455, + "learning_rate": 0.00011545461963927946, + "loss": 0.2677, + "step": 19228 + }, + { + "epoch": 0.45075804192525654, + "grad_norm": 4.086818695068359, + "learning_rate": 0.00011544734193976323, + "loss": 0.4766, + "step": 19229 + }, + { + "epoch": 0.45078148350006153, + "grad_norm": 3.3636083602905273, + "learning_rate": 0.00011544006415642932, + "loss": 0.3243, + "step": 19230 + }, + { + "epoch": 0.4508049250748665, + "grad_norm": 2.014047145843506, + "learning_rate": 0.0001154327862893172, + "loss": 0.2762, + "step": 19231 + }, + { + "epoch": 0.4508283666496715, + "grad_norm": 3.7614786624908447, + "learning_rate": 0.00011542550833846633, + "loss": 0.3821, + "step": 19232 + }, + { + "epoch": 0.4508518082244765, + "grad_norm": 5.052582263946533, + "learning_rate": 0.00011541823030391621, + "loss": 0.5552, + "step": 19233 + }, + { + "epoch": 0.4508752497992815, + "grad_norm": 0.9956457018852234, + "learning_rate": 0.00011541095218570636, + "loss": 0.0871, + "step": 19234 + }, + { + "epoch": 0.4508986913740865, + "grad_norm": 1.6803566217422485, + "learning_rate": 0.00011540367398387624, + "loss": 0.2114, + "step": 19235 + }, + { + "epoch": 0.4509221329488915, + "grad_norm": 4.634322166442871, + "learning_rate": 0.00011539639569846532, + "loss": 0.509, + "step": 19236 + }, + { + "epoch": 0.4509455745236965, + "grad_norm": 2.61784291267395, + "learning_rate": 0.00011538911732951319, + "loss": 1.4281, + "step": 19237 + }, + { + "epoch": 0.4509690160985015, + "grad_norm": 2.8199462890625, + "learning_rate": 0.00011538183887705925, + "loss": 0.2589, + "step": 19238 + }, + { + "epoch": 0.4509924576733065, + "grad_norm": 0.465398371219635, + "learning_rate": 0.00011537456034114298, + "loss": 0.022, + "step": 19239 + }, + { + "epoch": 0.4510158992481115, + "grad_norm": 3.5182111263275146, + "learning_rate": 0.00011536728172180396, + "loss": 0.3657, + "step": 19240 + }, + { + "epoch": 0.45103934082291647, + "grad_norm": 4.953571319580078, + "learning_rate": 0.00011536000301908161, + "loss": 0.4288, + "step": 19241 + }, + { + "epoch": 0.45106278239772146, + "grad_norm": 3.970712423324585, + "learning_rate": 0.00011535272423301547, + "loss": 0.3408, + "step": 19242 + }, + { + "epoch": 0.45108622397252646, + "grad_norm": 2.750291347503662, + "learning_rate": 0.000115345445363645, + "loss": 0.213, + "step": 19243 + }, + { + "epoch": 0.45110966554733145, + "grad_norm": 3.9823930263519287, + "learning_rate": 0.00011533816641100974, + "loss": 0.3563, + "step": 19244 + }, + { + "epoch": 0.45113310712213645, + "grad_norm": 3.3070297241210938, + "learning_rate": 0.00011533088737514912, + "loss": 0.3362, + "step": 19245 + }, + { + "epoch": 0.45115654869694144, + "grad_norm": 3.1810646057128906, + "learning_rate": 0.0001153236082561027, + "loss": 0.3687, + "step": 19246 + }, + { + "epoch": 0.45117999027174643, + "grad_norm": 1.9888657331466675, + "learning_rate": 0.00011531632905390994, + "loss": 0.1011, + "step": 19247 + }, + { + "epoch": 0.45120343184655143, + "grad_norm": 3.2583885192871094, + "learning_rate": 0.00011530904976861034, + "loss": 0.4386, + "step": 19248 + }, + { + "epoch": 0.4512268734213564, + "grad_norm": 3.696183204650879, + "learning_rate": 0.00011530177040024344, + "loss": 0.3032, + "step": 19249 + }, + { + "epoch": 0.4512503149961614, + "grad_norm": 2.5184431076049805, + "learning_rate": 0.00011529449094884867, + "loss": 0.1726, + "step": 19250 + }, + { + "epoch": 0.4512737565709664, + "grad_norm": 3.405330181121826, + "learning_rate": 0.00011528721141446557, + "loss": 0.3951, + "step": 19251 + }, + { + "epoch": 0.4512971981457714, + "grad_norm": 2.4410479068756104, + "learning_rate": 0.00011527993179713365, + "loss": 0.1658, + "step": 19252 + }, + { + "epoch": 0.45132063972057646, + "grad_norm": 1.5803943872451782, + "learning_rate": 0.00011527265209689236, + "loss": 0.1125, + "step": 19253 + }, + { + "epoch": 0.45134408129538145, + "grad_norm": 1.031496286392212, + "learning_rate": 0.00011526537231378125, + "loss": 0.103, + "step": 19254 + }, + { + "epoch": 0.45136752287018644, + "grad_norm": 4.02825927734375, + "learning_rate": 0.00011525809244783979, + "loss": 0.6033, + "step": 19255 + }, + { + "epoch": 0.45139096444499144, + "grad_norm": 4.071589946746826, + "learning_rate": 0.0001152508124991075, + "loss": 0.5814, + "step": 19256 + }, + { + "epoch": 0.45141440601979643, + "grad_norm": 1.307744026184082, + "learning_rate": 0.00011524353246762389, + "loss": 0.1672, + "step": 19257 + }, + { + "epoch": 0.4514378475946014, + "grad_norm": 4.39321756362915, + "learning_rate": 0.0001152362523534284, + "loss": 0.2585, + "step": 19258 + }, + { + "epoch": 0.4514612891694064, + "grad_norm": 2.518005847930908, + "learning_rate": 0.00011522897215656062, + "loss": 0.2587, + "step": 19259 + }, + { + "epoch": 0.4514847307442114, + "grad_norm": 4.451715469360352, + "learning_rate": 0.00011522169187705998, + "loss": 0.3301, + "step": 19260 + }, + { + "epoch": 0.4515081723190164, + "grad_norm": 2.622462749481201, + "learning_rate": 0.00011521441151496601, + "loss": 0.2713, + "step": 19261 + }, + { + "epoch": 0.4515316138938214, + "grad_norm": 3.203782320022583, + "learning_rate": 0.00011520713107031824, + "loss": 0.2286, + "step": 19262 + }, + { + "epoch": 0.4515550554686264, + "grad_norm": 3.8282723426818848, + "learning_rate": 0.00011519985054315613, + "loss": 0.3029, + "step": 19263 + }, + { + "epoch": 0.4515784970434314, + "grad_norm": 1.0286250114440918, + "learning_rate": 0.0001151925699335192, + "loss": 0.0594, + "step": 19264 + }, + { + "epoch": 0.4516019386182364, + "grad_norm": 6.275177478790283, + "learning_rate": 0.00011518528924144696, + "loss": 0.732, + "step": 19265 + }, + { + "epoch": 0.4516253801930414, + "grad_norm": 3.607335090637207, + "learning_rate": 0.00011517800846697892, + "loss": 1.9265, + "step": 19266 + }, + { + "epoch": 0.4516488217678464, + "grad_norm": 3.203131675720215, + "learning_rate": 0.00011517072761015458, + "loss": 0.3771, + "step": 19267 + }, + { + "epoch": 0.45167226334265137, + "grad_norm": 2.8398308753967285, + "learning_rate": 0.00011516344667101346, + "loss": 0.3264, + "step": 19268 + }, + { + "epoch": 0.45169570491745636, + "grad_norm": 1.998356819152832, + "learning_rate": 0.00011515616564959505, + "loss": 0.2293, + "step": 19269 + }, + { + "epoch": 0.45171914649226136, + "grad_norm": 4.363178730010986, + "learning_rate": 0.00011514888454593884, + "loss": 0.3658, + "step": 19270 + }, + { + "epoch": 0.45174258806706635, + "grad_norm": 2.878059148788452, + "learning_rate": 0.00011514160336008435, + "loss": 0.2208, + "step": 19271 + }, + { + "epoch": 0.45176602964187135, + "grad_norm": 0.8355652093887329, + "learning_rate": 0.00011513432209207111, + "loss": 0.0826, + "step": 19272 + }, + { + "epoch": 0.45178947121667634, + "grad_norm": 4.326432704925537, + "learning_rate": 0.00011512704074193863, + "loss": 0.5546, + "step": 19273 + }, + { + "epoch": 0.45181291279148134, + "grad_norm": 2.6457769870758057, + "learning_rate": 0.00011511975930972636, + "loss": 0.2374, + "step": 19274 + }, + { + "epoch": 0.45183635436628633, + "grad_norm": 5.33619499206543, + "learning_rate": 0.0001151124777954739, + "loss": 0.4535, + "step": 19275 + }, + { + "epoch": 0.4518597959410913, + "grad_norm": 1.0538356304168701, + "learning_rate": 0.00011510519619922069, + "loss": 0.1117, + "step": 19276 + }, + { + "epoch": 0.4518832375158963, + "grad_norm": 3.4656975269317627, + "learning_rate": 0.00011509791452100622, + "loss": 0.3284, + "step": 19277 + }, + { + "epoch": 0.4519066790907013, + "grad_norm": 1.1103485822677612, + "learning_rate": 0.00011509063276087009, + "loss": 0.1166, + "step": 19278 + }, + { + "epoch": 0.4519301206655063, + "grad_norm": 2.856536865234375, + "learning_rate": 0.00011508335091885174, + "loss": 0.2799, + "step": 19279 + }, + { + "epoch": 0.4519535622403113, + "grad_norm": 0.7650512456893921, + "learning_rate": 0.0001150760689949907, + "loss": 0.0824, + "step": 19280 + }, + { + "epoch": 0.4519770038151163, + "grad_norm": 2.272761821746826, + "learning_rate": 0.00011506878698932651, + "loss": 1.3646, + "step": 19281 + }, + { + "epoch": 0.4520004453899213, + "grad_norm": 4.354363441467285, + "learning_rate": 0.00011506150490189867, + "loss": 0.5252, + "step": 19282 + }, + { + "epoch": 0.4520238869647263, + "grad_norm": 3.2116496562957764, + "learning_rate": 0.00011505422273274665, + "loss": 0.4551, + "step": 19283 + }, + { + "epoch": 0.4520473285395313, + "grad_norm": 3.679198741912842, + "learning_rate": 0.00011504694048190998, + "loss": 0.4716, + "step": 19284 + }, + { + "epoch": 0.4520707701143363, + "grad_norm": 2.293619394302368, + "learning_rate": 0.00011503965814942821, + "loss": 0.2613, + "step": 19285 + }, + { + "epoch": 0.45209421168914127, + "grad_norm": 0.5268093943595886, + "learning_rate": 0.00011503237573534082, + "loss": 0.0555, + "step": 19286 + }, + { + "epoch": 0.45211765326394626, + "grad_norm": 4.194250583648682, + "learning_rate": 0.00011502509323968735, + "loss": 0.6561, + "step": 19287 + }, + { + "epoch": 0.45214109483875126, + "grad_norm": 2.4493720531463623, + "learning_rate": 0.00011501781066250732, + "loss": 0.1222, + "step": 19288 + }, + { + "epoch": 0.45216453641355625, + "grad_norm": 3.506720781326294, + "learning_rate": 0.0001150105280038402, + "loss": 0.2713, + "step": 19289 + }, + { + "epoch": 0.45218797798836124, + "grad_norm": 2.579442024230957, + "learning_rate": 0.00011500324526372552, + "loss": 0.1559, + "step": 19290 + }, + { + "epoch": 0.45221141956316624, + "grad_norm": 1.7738854885101318, + "learning_rate": 0.00011499596244220283, + "loss": 1.0319, + "step": 19291 + }, + { + "epoch": 0.45223486113797123, + "grad_norm": 0.7493136525154114, + "learning_rate": 0.00011498867953931165, + "loss": 0.069, + "step": 19292 + }, + { + "epoch": 0.45225830271277623, + "grad_norm": 10.654891967773438, + "learning_rate": 0.00011498139655509143, + "loss": 0.6404, + "step": 19293 + }, + { + "epoch": 0.4522817442875812, + "grad_norm": 1.8614968061447144, + "learning_rate": 0.00011497411348958176, + "loss": 0.1215, + "step": 19294 + }, + { + "epoch": 0.4523051858623862, + "grad_norm": 0.5476803779602051, + "learning_rate": 0.00011496683034282214, + "loss": 0.0548, + "step": 19295 + }, + { + "epoch": 0.4523286274371912, + "grad_norm": 3.2952632904052734, + "learning_rate": 0.00011495954711485202, + "loss": 0.3314, + "step": 19296 + }, + { + "epoch": 0.4523520690119962, + "grad_norm": 3.5293426513671875, + "learning_rate": 0.000114952263805711, + "loss": 0.3076, + "step": 19297 + }, + { + "epoch": 0.4523755105868012, + "grad_norm": 3.86372709274292, + "learning_rate": 0.00011494498041543862, + "loss": 0.5694, + "step": 19298 + }, + { + "epoch": 0.4523989521616062, + "grad_norm": 0.6242660284042358, + "learning_rate": 0.0001149376969440743, + "loss": 0.0652, + "step": 19299 + }, + { + "epoch": 0.4524223937364112, + "grad_norm": 1.0337533950805664, + "learning_rate": 0.00011493041339165767, + "loss": 0.1381, + "step": 19300 + }, + { + "epoch": 0.4524458353112162, + "grad_norm": 3.103978395462036, + "learning_rate": 0.00011492312975822817, + "loss": 0.3093, + "step": 19301 + }, + { + "epoch": 0.4524692768860212, + "grad_norm": 4.419283390045166, + "learning_rate": 0.00011491584604382532, + "loss": 0.5375, + "step": 19302 + }, + { + "epoch": 0.45249271846082617, + "grad_norm": 1.6936455965042114, + "learning_rate": 0.0001149085622484887, + "loss": 0.0878, + "step": 19303 + }, + { + "epoch": 0.4525161600356312, + "grad_norm": 2.4175519943237305, + "learning_rate": 0.0001149012783722578, + "loss": 0.2241, + "step": 19304 + }, + { + "epoch": 0.4525396016104362, + "grad_norm": 6.534682750701904, + "learning_rate": 0.00011489399441517213, + "loss": 0.6955, + "step": 19305 + }, + { + "epoch": 0.4525630431852412, + "grad_norm": 0.1629633605480194, + "learning_rate": 0.00011488671037727126, + "loss": 0.0072, + "step": 19306 + }, + { + "epoch": 0.4525864847600462, + "grad_norm": 1.472771406173706, + "learning_rate": 0.00011487942625859467, + "loss": 0.1394, + "step": 19307 + }, + { + "epoch": 0.4526099263348512, + "grad_norm": 1.5778048038482666, + "learning_rate": 0.00011487214205918186, + "loss": 0.1475, + "step": 19308 + }, + { + "epoch": 0.4526333679096562, + "grad_norm": 0.9098759889602661, + "learning_rate": 0.00011486485777907241, + "loss": 0.0858, + "step": 19309 + }, + { + "epoch": 0.4526568094844612, + "grad_norm": 1.0820313692092896, + "learning_rate": 0.00011485757341830585, + "loss": 0.0954, + "step": 19310 + }, + { + "epoch": 0.4526802510592662, + "grad_norm": 0.5654562711715698, + "learning_rate": 0.00011485028897692167, + "loss": 0.0304, + "step": 19311 + }, + { + "epoch": 0.4527036926340712, + "grad_norm": 2.698005437850952, + "learning_rate": 0.00011484300445495936, + "loss": 0.2384, + "step": 19312 + }, + { + "epoch": 0.45272713420887617, + "grad_norm": 3.012794017791748, + "learning_rate": 0.00011483571985245856, + "loss": 0.2004, + "step": 19313 + }, + { + "epoch": 0.45275057578368116, + "grad_norm": 6.3224263191223145, + "learning_rate": 0.00011482843516945867, + "loss": 0.6296, + "step": 19314 + }, + { + "epoch": 0.45277401735848616, + "grad_norm": 4.513423442840576, + "learning_rate": 0.00011482115040599928, + "loss": 0.5615, + "step": 19315 + }, + { + "epoch": 0.45279745893329115, + "grad_norm": 1.101186990737915, + "learning_rate": 0.00011481386556211993, + "loss": 0.0918, + "step": 19316 + }, + { + "epoch": 0.45282090050809615, + "grad_norm": 2.3915328979492188, + "learning_rate": 0.00011480658063786014, + "loss": 0.1425, + "step": 19317 + }, + { + "epoch": 0.45284434208290114, + "grad_norm": 1.5995616912841797, + "learning_rate": 0.00011479929563325939, + "loss": 0.1597, + "step": 19318 + }, + { + "epoch": 0.45286778365770614, + "grad_norm": 2.7209954261779785, + "learning_rate": 0.00011479201054835728, + "loss": 0.3448, + "step": 19319 + }, + { + "epoch": 0.45289122523251113, + "grad_norm": 1.2660280466079712, + "learning_rate": 0.0001147847253831933, + "loss": 0.1045, + "step": 19320 + }, + { + "epoch": 0.4529146668073161, + "grad_norm": 2.177260160446167, + "learning_rate": 0.00011477744013780695, + "loss": 0.2146, + "step": 19321 + }, + { + "epoch": 0.4529381083821211, + "grad_norm": 3.2175769805908203, + "learning_rate": 0.00011477015481223783, + "loss": 0.3663, + "step": 19322 + }, + { + "epoch": 0.4529615499569261, + "grad_norm": 2.8381800651550293, + "learning_rate": 0.00011476286940652543, + "loss": 0.3332, + "step": 19323 + }, + { + "epoch": 0.4529849915317311, + "grad_norm": 1.3704851865768433, + "learning_rate": 0.00011475558392070927, + "loss": 0.0818, + "step": 19324 + }, + { + "epoch": 0.4530084331065361, + "grad_norm": 1.3258745670318604, + "learning_rate": 0.00011474829835482893, + "loss": 0.2231, + "step": 19325 + }, + { + "epoch": 0.4530318746813411, + "grad_norm": 1.2229994535446167, + "learning_rate": 0.00011474101270892386, + "loss": 0.0973, + "step": 19326 + }, + { + "epoch": 0.4530553162561461, + "grad_norm": 3.095548391342163, + "learning_rate": 0.00011473372698303368, + "loss": 0.2226, + "step": 19327 + }, + { + "epoch": 0.4530787578309511, + "grad_norm": 2.026700735092163, + "learning_rate": 0.00011472644117719787, + "loss": 0.9844, + "step": 19328 + }, + { + "epoch": 0.4531021994057561, + "grad_norm": 4.950118541717529, + "learning_rate": 0.00011471915529145599, + "loss": 0.3773, + "step": 19329 + }, + { + "epoch": 0.4531256409805611, + "grad_norm": 1.5295612812042236, + "learning_rate": 0.00011471186932584755, + "loss": 0.2568, + "step": 19330 + }, + { + "epoch": 0.45314908255536607, + "grad_norm": 1.1972774267196655, + "learning_rate": 0.00011470458328041207, + "loss": 0.159, + "step": 19331 + }, + { + "epoch": 0.45317252413017106, + "grad_norm": 4.045089244842529, + "learning_rate": 0.0001146972971551891, + "loss": 0.3548, + "step": 19332 + }, + { + "epoch": 0.45319596570497606, + "grad_norm": 1.264719009399414, + "learning_rate": 0.00011469001095021823, + "loss": 0.1601, + "step": 19333 + }, + { + "epoch": 0.45321940727978105, + "grad_norm": 4.446267604827881, + "learning_rate": 0.0001146827246655389, + "loss": 0.3656, + "step": 19334 + }, + { + "epoch": 0.45324284885458604, + "grad_norm": 2.9587783813476562, + "learning_rate": 0.00011467543830119075, + "loss": 0.3018, + "step": 19335 + }, + { + "epoch": 0.45326629042939104, + "grad_norm": 3.545548439025879, + "learning_rate": 0.00011466815185721323, + "loss": 0.3583, + "step": 19336 + }, + { + "epoch": 0.45328973200419603, + "grad_norm": 4.451587200164795, + "learning_rate": 0.00011466086533364589, + "loss": 0.5412, + "step": 19337 + }, + { + "epoch": 0.453313173579001, + "grad_norm": 3.05218505859375, + "learning_rate": 0.00011465357873052829, + "loss": 0.3589, + "step": 19338 + }, + { + "epoch": 0.453336615153806, + "grad_norm": 4.018267631530762, + "learning_rate": 0.00011464629204789996, + "loss": 0.3427, + "step": 19339 + }, + { + "epoch": 0.453360056728611, + "grad_norm": 4.031534671783447, + "learning_rate": 0.00011463900528580047, + "loss": 0.4221, + "step": 19340 + }, + { + "epoch": 0.453383498303416, + "grad_norm": 4.3589701652526855, + "learning_rate": 0.00011463171844426928, + "loss": 0.4461, + "step": 19341 + }, + { + "epoch": 0.453406939878221, + "grad_norm": 3.214101552963257, + "learning_rate": 0.000114624431523346, + "loss": 0.3973, + "step": 19342 + }, + { + "epoch": 0.453430381453026, + "grad_norm": 3.5625648498535156, + "learning_rate": 0.00011461714452307012, + "loss": 0.186, + "step": 19343 + }, + { + "epoch": 0.453453823027831, + "grad_norm": 2.2790679931640625, + "learning_rate": 0.00011460985744348121, + "loss": 0.0922, + "step": 19344 + }, + { + "epoch": 0.453477264602636, + "grad_norm": 2.094548463821411, + "learning_rate": 0.00011460257028461883, + "loss": 1.4041, + "step": 19345 + }, + { + "epoch": 0.453500706177441, + "grad_norm": 1.1375296115875244, + "learning_rate": 0.0001145952830465225, + "loss": 0.1122, + "step": 19346 + }, + { + "epoch": 0.453524147752246, + "grad_norm": 0.4883379340171814, + "learning_rate": 0.00011458799572923171, + "loss": 0.0412, + "step": 19347 + }, + { + "epoch": 0.45354758932705097, + "grad_norm": 1.6126497983932495, + "learning_rate": 0.00011458070833278607, + "loss": 0.1434, + "step": 19348 + }, + { + "epoch": 0.45357103090185596, + "grad_norm": 0.7671038508415222, + "learning_rate": 0.00011457342085722509, + "loss": 0.0498, + "step": 19349 + }, + { + "epoch": 0.45359447247666096, + "grad_norm": 4.022693157196045, + "learning_rate": 0.0001145661333025883, + "loss": 0.3312, + "step": 19350 + }, + { + "epoch": 0.45361791405146595, + "grad_norm": 0.8960660696029663, + "learning_rate": 0.00011455884566891528, + "loss": 0.072, + "step": 19351 + }, + { + "epoch": 0.45364135562627095, + "grad_norm": 5.422720432281494, + "learning_rate": 0.00011455155795624557, + "loss": 0.6894, + "step": 19352 + }, + { + "epoch": 0.45366479720107594, + "grad_norm": 1.2134993076324463, + "learning_rate": 0.00011454427016461866, + "loss": 0.1353, + "step": 19353 + }, + { + "epoch": 0.45368823877588094, + "grad_norm": 1.1945621967315674, + "learning_rate": 0.00011453698229407417, + "loss": 0.101, + "step": 19354 + }, + { + "epoch": 0.453711680350686, + "grad_norm": 5.852933883666992, + "learning_rate": 0.0001145296943446516, + "loss": 0.4805, + "step": 19355 + }, + { + "epoch": 0.453735121925491, + "grad_norm": 3.7051658630371094, + "learning_rate": 0.00011452240631639048, + "loss": 0.2328, + "step": 19356 + }, + { + "epoch": 0.453758563500296, + "grad_norm": 1.8832948207855225, + "learning_rate": 0.00011451511820933037, + "loss": 0.1652, + "step": 19357 + }, + { + "epoch": 0.45378200507510097, + "grad_norm": 6.479060649871826, + "learning_rate": 0.00011450783002351084, + "loss": 1.0124, + "step": 19358 + }, + { + "epoch": 0.45380544664990596, + "grad_norm": 2.113250970840454, + "learning_rate": 0.00011450054175897143, + "loss": 0.1434, + "step": 19359 + }, + { + "epoch": 0.45382888822471096, + "grad_norm": 3.7179508209228516, + "learning_rate": 0.00011449325341575163, + "loss": 0.2328, + "step": 19360 + }, + { + "epoch": 0.45385232979951595, + "grad_norm": 0.8487473726272583, + "learning_rate": 0.00011448596499389105, + "loss": 0.082, + "step": 19361 + }, + { + "epoch": 0.45387577137432095, + "grad_norm": 0.786825954914093, + "learning_rate": 0.00011447867649342922, + "loss": 0.1051, + "step": 19362 + }, + { + "epoch": 0.45389921294912594, + "grad_norm": 1.5237157344818115, + "learning_rate": 0.00011447138791440568, + "loss": 0.1664, + "step": 19363 + }, + { + "epoch": 0.45392265452393094, + "grad_norm": 3.591963052749634, + "learning_rate": 0.00011446409925686, + "loss": 0.2207, + "step": 19364 + }, + { + "epoch": 0.45394609609873593, + "grad_norm": 2.9686217308044434, + "learning_rate": 0.0001144568105208317, + "loss": 2.0015, + "step": 19365 + }, + { + "epoch": 0.4539695376735409, + "grad_norm": 1.3972253799438477, + "learning_rate": 0.00011444952170636032, + "loss": 0.564, + "step": 19366 + }, + { + "epoch": 0.4539929792483459, + "grad_norm": 4.266989231109619, + "learning_rate": 0.00011444223281348545, + "loss": 0.1949, + "step": 19367 + }, + { + "epoch": 0.4540164208231509, + "grad_norm": 6.727077007293701, + "learning_rate": 0.00011443494384224665, + "loss": 0.4414, + "step": 19368 + }, + { + "epoch": 0.4540398623979559, + "grad_norm": 3.604177713394165, + "learning_rate": 0.00011442765479268335, + "loss": 0.3285, + "step": 19369 + }, + { + "epoch": 0.4540633039727609, + "grad_norm": 1.1912952661514282, + "learning_rate": 0.00011442036566483525, + "loss": 0.1337, + "step": 19370 + }, + { + "epoch": 0.4540867455475659, + "grad_norm": 3.9469993114471436, + "learning_rate": 0.00011441307645874183, + "loss": 0.3966, + "step": 19371 + }, + { + "epoch": 0.4541101871223709, + "grad_norm": 3.8242428302764893, + "learning_rate": 0.00011440578717444265, + "loss": 0.7757, + "step": 19372 + }, + { + "epoch": 0.4541336286971759, + "grad_norm": 0.7216225266456604, + "learning_rate": 0.0001143984978119773, + "loss": 0.0547, + "step": 19373 + }, + { + "epoch": 0.4541570702719809, + "grad_norm": 4.148489475250244, + "learning_rate": 0.00011439120837138526, + "loss": 0.3349, + "step": 19374 + }, + { + "epoch": 0.4541805118467859, + "grad_norm": 3.442441940307617, + "learning_rate": 0.00011438391885270611, + "loss": 0.4118, + "step": 19375 + }, + { + "epoch": 0.45420395342159087, + "grad_norm": 3.3048038482666016, + "learning_rate": 0.00011437662925597941, + "loss": 1.4501, + "step": 19376 + }, + { + "epoch": 0.45422739499639586, + "grad_norm": 2.9139719009399414, + "learning_rate": 0.00011436933958124474, + "loss": 1.3957, + "step": 19377 + }, + { + "epoch": 0.45425083657120086, + "grad_norm": 3.1987674236297607, + "learning_rate": 0.00011436204982854164, + "loss": 0.2461, + "step": 19378 + }, + { + "epoch": 0.45427427814600585, + "grad_norm": 3.8730416297912598, + "learning_rate": 0.00011435475999790961, + "loss": 0.4523, + "step": 19379 + }, + { + "epoch": 0.45429771972081084, + "grad_norm": 0.3384299576282501, + "learning_rate": 0.00011434747008938828, + "loss": 0.0318, + "step": 19380 + }, + { + "epoch": 0.45432116129561584, + "grad_norm": 2.4297218322753906, + "learning_rate": 0.00011434018010301717, + "loss": 1.1384, + "step": 19381 + }, + { + "epoch": 0.45434460287042083, + "grad_norm": 2.0051803588867188, + "learning_rate": 0.00011433289003883581, + "loss": 0.2237, + "step": 19382 + }, + { + "epoch": 0.4543680444452258, + "grad_norm": 3.127467393875122, + "learning_rate": 0.00011432559989688382, + "loss": 0.2256, + "step": 19383 + }, + { + "epoch": 0.4543914860200308, + "grad_norm": 0.261141836643219, + "learning_rate": 0.00011431830967720074, + "loss": 0.0222, + "step": 19384 + }, + { + "epoch": 0.4544149275948358, + "grad_norm": 1.8271207809448242, + "learning_rate": 0.00011431101937982607, + "loss": 0.2192, + "step": 19385 + }, + { + "epoch": 0.4544383691696408, + "grad_norm": 1.8484492301940918, + "learning_rate": 0.00011430372900479942, + "loss": 0.0483, + "step": 19386 + }, + { + "epoch": 0.4544618107444458, + "grad_norm": 2.381229877471924, + "learning_rate": 0.00011429643855216032, + "loss": 0.2657, + "step": 19387 + }, + { + "epoch": 0.4544852523192508, + "grad_norm": 3.3569490909576416, + "learning_rate": 0.00011428914802194834, + "loss": 1.4186, + "step": 19388 + }, + { + "epoch": 0.4545086938940558, + "grad_norm": 2.869485378265381, + "learning_rate": 0.00011428185741420307, + "loss": 0.2164, + "step": 19389 + }, + { + "epoch": 0.4545321354688608, + "grad_norm": 2.1950325965881348, + "learning_rate": 0.00011427456672896403, + "loss": 0.1032, + "step": 19390 + }, + { + "epoch": 0.4545555770436658, + "grad_norm": 3.059288740158081, + "learning_rate": 0.00011426727596627079, + "loss": 0.2028, + "step": 19391 + }, + { + "epoch": 0.4545790186184708, + "grad_norm": 3.782416343688965, + "learning_rate": 0.0001142599851261629, + "loss": 0.2687, + "step": 19392 + }, + { + "epoch": 0.45460246019327577, + "grad_norm": 3.3524701595306396, + "learning_rate": 0.00011425269420867992, + "loss": 1.4061, + "step": 19393 + }, + { + "epoch": 0.45462590176808076, + "grad_norm": 5.3324198722839355, + "learning_rate": 0.00011424540321386146, + "loss": 0.4999, + "step": 19394 + }, + { + "epoch": 0.45464934334288576, + "grad_norm": 0.7796926498413086, + "learning_rate": 0.000114238112141747, + "loss": 0.0716, + "step": 19395 + }, + { + "epoch": 0.45467278491769075, + "grad_norm": 3.305710554122925, + "learning_rate": 0.00011423082099237617, + "loss": 0.3822, + "step": 19396 + }, + { + "epoch": 0.45469622649249575, + "grad_norm": 2.8217008113861084, + "learning_rate": 0.0001142235297657885, + "loss": 1.4489, + "step": 19397 + }, + { + "epoch": 0.45471966806730074, + "grad_norm": 4.405259609222412, + "learning_rate": 0.00011421623846202354, + "loss": 0.3159, + "step": 19398 + }, + { + "epoch": 0.45474310964210574, + "grad_norm": 1.8725183010101318, + "learning_rate": 0.00011420894708112089, + "loss": 1.209, + "step": 19399 + }, + { + "epoch": 0.45476655121691073, + "grad_norm": 0.5986860990524292, + "learning_rate": 0.00011420165562312008, + "loss": 0.0486, + "step": 19400 + }, + { + "epoch": 0.4547899927917157, + "grad_norm": 3.3174352645874023, + "learning_rate": 0.00011419436408806068, + "loss": 0.3617, + "step": 19401 + }, + { + "epoch": 0.4548134343665207, + "grad_norm": 2.557263135910034, + "learning_rate": 0.0001141870724759823, + "loss": 0.1741, + "step": 19402 + }, + { + "epoch": 0.4548368759413257, + "grad_norm": 2.9948456287384033, + "learning_rate": 0.00011417978078692445, + "loss": 0.2974, + "step": 19403 + }, + { + "epoch": 0.4548603175161307, + "grad_norm": 2.9703259468078613, + "learning_rate": 0.0001141724890209267, + "loss": 0.3825, + "step": 19404 + }, + { + "epoch": 0.4548837590909357, + "grad_norm": 4.871339797973633, + "learning_rate": 0.0001141651971780286, + "loss": 0.2841, + "step": 19405 + }, + { + "epoch": 0.45490720066574075, + "grad_norm": 1.5662312507629395, + "learning_rate": 0.00011415790525826979, + "loss": 0.1157, + "step": 19406 + }, + { + "epoch": 0.45493064224054575, + "grad_norm": 1.8326760530471802, + "learning_rate": 0.00011415061326168976, + "loss": 0.0982, + "step": 19407 + }, + { + "epoch": 0.45495408381535074, + "grad_norm": 3.7635233402252197, + "learning_rate": 0.00011414332118832811, + "loss": 0.2474, + "step": 19408 + }, + { + "epoch": 0.45497752539015573, + "grad_norm": 3.0831449031829834, + "learning_rate": 0.00011413602903822443, + "loss": 0.4948, + "step": 19409 + }, + { + "epoch": 0.45500096696496073, + "grad_norm": 3.859830379486084, + "learning_rate": 0.0001141287368114182, + "loss": 0.2735, + "step": 19410 + }, + { + "epoch": 0.4550244085397657, + "grad_norm": 3.8533122539520264, + "learning_rate": 0.00011412144450794909, + "loss": 0.2305, + "step": 19411 + }, + { + "epoch": 0.4550478501145707, + "grad_norm": 1.6182256937026978, + "learning_rate": 0.00011411415212785662, + "loss": 0.2795, + "step": 19412 + }, + { + "epoch": 0.4550712916893757, + "grad_norm": 6.8639631271362305, + "learning_rate": 0.00011410685967118036, + "loss": 0.4382, + "step": 19413 + }, + { + "epoch": 0.4550947332641807, + "grad_norm": 2.2449722290039062, + "learning_rate": 0.00011409956713795988, + "loss": 0.1664, + "step": 19414 + }, + { + "epoch": 0.4551181748389857, + "grad_norm": 8.833070755004883, + "learning_rate": 0.00011409227452823477, + "loss": 0.5054, + "step": 19415 + }, + { + "epoch": 0.4551416164137907, + "grad_norm": 2.6537818908691406, + "learning_rate": 0.00011408498184204457, + "loss": 0.3753, + "step": 19416 + }, + { + "epoch": 0.4551650579885957, + "grad_norm": 4.7381792068481445, + "learning_rate": 0.00011407768907942882, + "loss": 0.5146, + "step": 19417 + }, + { + "epoch": 0.4551884995634007, + "grad_norm": 5.474999904632568, + "learning_rate": 0.00011407039624042719, + "loss": 0.5181, + "step": 19418 + }, + { + "epoch": 0.4552119411382057, + "grad_norm": 2.731821060180664, + "learning_rate": 0.00011406310332507919, + "loss": 0.2041, + "step": 19419 + }, + { + "epoch": 0.4552353827130107, + "grad_norm": 5.622439861297607, + "learning_rate": 0.00011405581033342436, + "loss": 0.4413, + "step": 19420 + }, + { + "epoch": 0.45525882428781567, + "grad_norm": 3.414661407470703, + "learning_rate": 0.00011404851726550234, + "loss": 0.3336, + "step": 19421 + }, + { + "epoch": 0.45528226586262066, + "grad_norm": 1.5601298809051514, + "learning_rate": 0.00011404122412135268, + "loss": 0.1345, + "step": 19422 + }, + { + "epoch": 0.45530570743742566, + "grad_norm": 4.666260242462158, + "learning_rate": 0.00011403393090101488, + "loss": 0.5195, + "step": 19423 + }, + { + "epoch": 0.45532914901223065, + "grad_norm": 1.7335747480392456, + "learning_rate": 0.00011402663760452863, + "loss": 0.1721, + "step": 19424 + }, + { + "epoch": 0.45535259058703564, + "grad_norm": 2.5984556674957275, + "learning_rate": 0.00011401934423193345, + "loss": 0.2683, + "step": 19425 + }, + { + "epoch": 0.45537603216184064, + "grad_norm": 2.1599085330963135, + "learning_rate": 0.00011401205078326889, + "loss": 0.1379, + "step": 19426 + }, + { + "epoch": 0.45539947373664563, + "grad_norm": 3.238053560256958, + "learning_rate": 0.00011400475725857457, + "loss": 1.8719, + "step": 19427 + }, + { + "epoch": 0.4554229153114506, + "grad_norm": 3.5160562992095947, + "learning_rate": 0.00011399746365789006, + "loss": 2.1725, + "step": 19428 + }, + { + "epoch": 0.4554463568862556, + "grad_norm": 4.965295314788818, + "learning_rate": 0.00011399016998125487, + "loss": 0.3757, + "step": 19429 + }, + { + "epoch": 0.4554697984610606, + "grad_norm": 2.35866641998291, + "learning_rate": 0.00011398287622870864, + "loss": 0.1762, + "step": 19430 + }, + { + "epoch": 0.4554932400358656, + "grad_norm": 2.8754971027374268, + "learning_rate": 0.00011397558240029096, + "loss": 0.2832, + "step": 19431 + }, + { + "epoch": 0.4555166816106706, + "grad_norm": 3.1553797721862793, + "learning_rate": 0.00011396828849604135, + "loss": 0.3562, + "step": 19432 + }, + { + "epoch": 0.4555401231854756, + "grad_norm": 3.294917583465576, + "learning_rate": 0.00011396099451599942, + "loss": 0.2441, + "step": 19433 + }, + { + "epoch": 0.4555635647602806, + "grad_norm": 1.1344382762908936, + "learning_rate": 0.00011395370046020474, + "loss": 0.1645, + "step": 19434 + }, + { + "epoch": 0.4555870063350856, + "grad_norm": 1.3691917657852173, + "learning_rate": 0.00011394640632869688, + "loss": 0.4573, + "step": 19435 + }, + { + "epoch": 0.4556104479098906, + "grad_norm": 3.648996591567993, + "learning_rate": 0.00011393911212151542, + "loss": 0.4508, + "step": 19436 + }, + { + "epoch": 0.4556338894846956, + "grad_norm": 2.623375415802002, + "learning_rate": 0.00011393181783869998, + "loss": 0.196, + "step": 19437 + }, + { + "epoch": 0.45565733105950057, + "grad_norm": 3.332933187484741, + "learning_rate": 0.0001139245234802901, + "loss": 0.3262, + "step": 19438 + }, + { + "epoch": 0.45568077263430556, + "grad_norm": 3.432591438293457, + "learning_rate": 0.00011391722904632534, + "loss": 0.2955, + "step": 19439 + }, + { + "epoch": 0.45570421420911056, + "grad_norm": 3.036008358001709, + "learning_rate": 0.00011390993453684533, + "loss": 0.321, + "step": 19440 + }, + { + "epoch": 0.45572765578391555, + "grad_norm": 4.830687522888184, + "learning_rate": 0.00011390263995188958, + "loss": 0.5587, + "step": 19441 + }, + { + "epoch": 0.45575109735872055, + "grad_norm": 1.6057740449905396, + "learning_rate": 0.00011389534529149775, + "loss": 0.203, + "step": 19442 + }, + { + "epoch": 0.45577453893352554, + "grad_norm": 5.18429708480835, + "learning_rate": 0.00011388805055570939, + "loss": 0.3332, + "step": 19443 + }, + { + "epoch": 0.45579798050833054, + "grad_norm": 3.710608959197998, + "learning_rate": 0.00011388075574456407, + "loss": 0.3435, + "step": 19444 + }, + { + "epoch": 0.45582142208313553, + "grad_norm": 7.807027816772461, + "learning_rate": 0.00011387346085810135, + "loss": 0.7806, + "step": 19445 + }, + { + "epoch": 0.4558448636579405, + "grad_norm": 0.6033555865287781, + "learning_rate": 0.0001138661658963609, + "loss": 0.0854, + "step": 19446 + }, + { + "epoch": 0.4558683052327455, + "grad_norm": 4.895906925201416, + "learning_rate": 0.0001138588708593822, + "loss": 0.3622, + "step": 19447 + }, + { + "epoch": 0.4558917468075505, + "grad_norm": 3.351766347885132, + "learning_rate": 0.00011385157574720488, + "loss": 0.3492, + "step": 19448 + }, + { + "epoch": 0.4559151883823555, + "grad_norm": 2.7766828536987305, + "learning_rate": 0.00011384428055986853, + "loss": 1.639, + "step": 19449 + }, + { + "epoch": 0.4559386299571605, + "grad_norm": 1.992648720741272, + "learning_rate": 0.00011383698529741274, + "loss": 0.1092, + "step": 19450 + }, + { + "epoch": 0.4559620715319655, + "grad_norm": 3.7454581260681152, + "learning_rate": 0.00011382968995987706, + "loss": 0.3612, + "step": 19451 + }, + { + "epoch": 0.4559855131067705, + "grad_norm": 2.237833261489868, + "learning_rate": 0.00011382239454730108, + "loss": 0.2109, + "step": 19452 + }, + { + "epoch": 0.4560089546815755, + "grad_norm": 4.703036308288574, + "learning_rate": 0.00011381509905972443, + "loss": 0.3335, + "step": 19453 + }, + { + "epoch": 0.4560323962563805, + "grad_norm": 3.4824607372283936, + "learning_rate": 0.00011380780349718662, + "loss": 0.3121, + "step": 19454 + }, + { + "epoch": 0.4560558378311855, + "grad_norm": 0.9309441447257996, + "learning_rate": 0.00011380050785972731, + "loss": 0.1014, + "step": 19455 + }, + { + "epoch": 0.45607927940599047, + "grad_norm": 3.3276326656341553, + "learning_rate": 0.00011379321214738607, + "loss": 0.4158, + "step": 19456 + }, + { + "epoch": 0.4561027209807955, + "grad_norm": 4.311075687408447, + "learning_rate": 0.00011378591636020246, + "loss": 0.4004, + "step": 19457 + }, + { + "epoch": 0.4561261625556005, + "grad_norm": 2.9421963691711426, + "learning_rate": 0.00011377862049821608, + "loss": 0.1826, + "step": 19458 + }, + { + "epoch": 0.4561496041304055, + "grad_norm": 4.385960578918457, + "learning_rate": 0.00011377132456146651, + "loss": 0.5194, + "step": 19459 + }, + { + "epoch": 0.4561730457052105, + "grad_norm": 4.640927314758301, + "learning_rate": 0.00011376402854999335, + "loss": 0.4377, + "step": 19460 + }, + { + "epoch": 0.4561964872800155, + "grad_norm": 0.8064950704574585, + "learning_rate": 0.00011375673246383618, + "loss": 0.0835, + "step": 19461 + }, + { + "epoch": 0.4562199288548205, + "grad_norm": 2.9256234169006348, + "learning_rate": 0.0001137494363030346, + "loss": 0.3226, + "step": 19462 + }, + { + "epoch": 0.4562433704296255, + "grad_norm": 4.261623859405518, + "learning_rate": 0.0001137421400676282, + "loss": 0.5657, + "step": 19463 + }, + { + "epoch": 0.4562668120044305, + "grad_norm": 3.560422420501709, + "learning_rate": 0.00011373484375765652, + "loss": 0.3766, + "step": 19464 + }, + { + "epoch": 0.45629025357923547, + "grad_norm": 3.864722728729248, + "learning_rate": 0.00011372754737315925, + "loss": 0.6457, + "step": 19465 + }, + { + "epoch": 0.45631369515404047, + "grad_norm": 2.3587236404418945, + "learning_rate": 0.00011372025091417589, + "loss": 0.2663, + "step": 19466 + }, + { + "epoch": 0.45633713672884546, + "grad_norm": 8.003158569335938, + "learning_rate": 0.00011371295438074607, + "loss": 0.4158, + "step": 19467 + }, + { + "epoch": 0.45636057830365045, + "grad_norm": 2.3282666206359863, + "learning_rate": 0.00011370565777290935, + "loss": 0.1627, + "step": 19468 + }, + { + "epoch": 0.45638401987845545, + "grad_norm": 3.370370864868164, + "learning_rate": 0.0001136983610907054, + "loss": 0.2938, + "step": 19469 + }, + { + "epoch": 0.45640746145326044, + "grad_norm": 2.614816665649414, + "learning_rate": 0.00011369106433417371, + "loss": 0.1623, + "step": 19470 + }, + { + "epoch": 0.45643090302806544, + "grad_norm": 3.311685562133789, + "learning_rate": 0.00011368376750335394, + "loss": 0.1132, + "step": 19471 + }, + { + "epoch": 0.45645434460287043, + "grad_norm": 1.0515546798706055, + "learning_rate": 0.00011367647059828563, + "loss": 0.1076, + "step": 19472 + }, + { + "epoch": 0.4564777861776754, + "grad_norm": 3.849447250366211, + "learning_rate": 0.00011366917361900846, + "loss": 2.1973, + "step": 19473 + }, + { + "epoch": 0.4565012277524804, + "grad_norm": 2.269559860229492, + "learning_rate": 0.00011366187656556192, + "loss": 0.2444, + "step": 19474 + }, + { + "epoch": 0.4565246693272854, + "grad_norm": 2.776684284210205, + "learning_rate": 0.0001136545794379857, + "loss": 0.2759, + "step": 19475 + }, + { + "epoch": 0.4565481109020904, + "grad_norm": 4.87249231338501, + "learning_rate": 0.00011364728223631934, + "loss": 0.4376, + "step": 19476 + }, + { + "epoch": 0.4565715524768954, + "grad_norm": 3.738083600997925, + "learning_rate": 0.00011363998496060242, + "loss": 0.382, + "step": 19477 + }, + { + "epoch": 0.4565949940517004, + "grad_norm": 3.0974009037017822, + "learning_rate": 0.00011363268761087453, + "loss": 0.326, + "step": 19478 + }, + { + "epoch": 0.4566184356265054, + "grad_norm": 3.2899580001831055, + "learning_rate": 0.00011362539018717535, + "loss": 0.2421, + "step": 19479 + }, + { + "epoch": 0.4566418772013104, + "grad_norm": 1.0189639329910278, + "learning_rate": 0.0001136180926895444, + "loss": 0.1132, + "step": 19480 + }, + { + "epoch": 0.4566653187761154, + "grad_norm": 0.406292587518692, + "learning_rate": 0.00011361079511802129, + "loss": 0.0354, + "step": 19481 + }, + { + "epoch": 0.4566887603509204, + "grad_norm": 3.4906959533691406, + "learning_rate": 0.00011360349747264565, + "loss": 0.3773, + "step": 19482 + }, + { + "epoch": 0.45671220192572537, + "grad_norm": 3.0676991939544678, + "learning_rate": 0.00011359619975345701, + "loss": 0.2824, + "step": 19483 + }, + { + "epoch": 0.45673564350053036, + "grad_norm": 3.427612781524658, + "learning_rate": 0.000113588901960495, + "loss": 0.4345, + "step": 19484 + }, + { + "epoch": 0.45675908507533536, + "grad_norm": 3.270834445953369, + "learning_rate": 0.00011358160409379927, + "loss": 0.3071, + "step": 19485 + }, + { + "epoch": 0.45678252665014035, + "grad_norm": 2.7301647663116455, + "learning_rate": 0.00011357430615340938, + "loss": 0.3063, + "step": 19486 + }, + { + "epoch": 0.45680596822494535, + "grad_norm": 3.7245874404907227, + "learning_rate": 0.00011356700813936489, + "loss": 0.2783, + "step": 19487 + }, + { + "epoch": 0.45682940979975034, + "grad_norm": 4.538973808288574, + "learning_rate": 0.00011355971005170545, + "loss": 0.343, + "step": 19488 + }, + { + "epoch": 0.45685285137455534, + "grad_norm": 2.711630344390869, + "learning_rate": 0.00011355241189047064, + "loss": 0.1851, + "step": 19489 + }, + { + "epoch": 0.45687629294936033, + "grad_norm": 0.8761885166168213, + "learning_rate": 0.00011354511365570002, + "loss": 0.0732, + "step": 19490 + }, + { + "epoch": 0.4568997345241653, + "grad_norm": 0.7840362191200256, + "learning_rate": 0.00011353781534743328, + "loss": 0.0927, + "step": 19491 + }, + { + "epoch": 0.4569231760989703, + "grad_norm": 2.7510933876037598, + "learning_rate": 0.00011353051696570998, + "loss": 0.3962, + "step": 19492 + }, + { + "epoch": 0.4569466176737753, + "grad_norm": 2.625136137008667, + "learning_rate": 0.00011352321851056968, + "loss": 0.1733, + "step": 19493 + }, + { + "epoch": 0.4569700592485803, + "grad_norm": 1.1693943738937378, + "learning_rate": 0.00011351591998205205, + "loss": 0.1049, + "step": 19494 + }, + { + "epoch": 0.4569935008233853, + "grad_norm": 1.815354347229004, + "learning_rate": 0.00011350862138019664, + "loss": 0.2084, + "step": 19495 + }, + { + "epoch": 0.4570169423981903, + "grad_norm": 3.50178861618042, + "learning_rate": 0.00011350132270504305, + "loss": 0.4249, + "step": 19496 + }, + { + "epoch": 0.4570403839729953, + "grad_norm": 1.0399479866027832, + "learning_rate": 0.00011349402395663093, + "loss": 0.0469, + "step": 19497 + }, + { + "epoch": 0.4570638255478003, + "grad_norm": 3.844459295272827, + "learning_rate": 0.00011348672513499985, + "loss": 0.4054, + "step": 19498 + }, + { + "epoch": 0.4570872671226053, + "grad_norm": 3.2341623306274414, + "learning_rate": 0.00011347942624018941, + "loss": 0.4162, + "step": 19499 + }, + { + "epoch": 0.4571107086974103, + "grad_norm": 2.8187038898468018, + "learning_rate": 0.00011347212727223926, + "loss": 0.1976, + "step": 19500 + }, + { + "epoch": 0.45713415027221527, + "grad_norm": 4.063335418701172, + "learning_rate": 0.00011346482823118894, + "loss": 0.3943, + "step": 19501 + }, + { + "epoch": 0.45715759184702026, + "grad_norm": 0.6186267137527466, + "learning_rate": 0.00011345752911707808, + "loss": 0.0294, + "step": 19502 + }, + { + "epoch": 0.45718103342182526, + "grad_norm": 0.4435047209262848, + "learning_rate": 0.00011345022992994629, + "loss": 0.0532, + "step": 19503 + }, + { + "epoch": 0.45720447499663025, + "grad_norm": 3.3789968490600586, + "learning_rate": 0.00011344293066983318, + "loss": 0.3689, + "step": 19504 + }, + { + "epoch": 0.45722791657143524, + "grad_norm": 1.27802574634552, + "learning_rate": 0.00011343563133677838, + "loss": 0.0876, + "step": 19505 + }, + { + "epoch": 0.45725135814624024, + "grad_norm": 5.1622419357299805, + "learning_rate": 0.00011342833193082141, + "loss": 0.4783, + "step": 19506 + }, + { + "epoch": 0.45727479972104523, + "grad_norm": 4.796516418457031, + "learning_rate": 0.00011342103245200198, + "loss": 0.4062, + "step": 19507 + }, + { + "epoch": 0.4572982412958503, + "grad_norm": 0.9758027791976929, + "learning_rate": 0.00011341373290035962, + "loss": 0.1033, + "step": 19508 + }, + { + "epoch": 0.4573216828706553, + "grad_norm": 4.122453212738037, + "learning_rate": 0.00011340643327593396, + "loss": 0.3454, + "step": 19509 + }, + { + "epoch": 0.45734512444546027, + "grad_norm": 3.892017126083374, + "learning_rate": 0.00011339913357876464, + "loss": 0.4125, + "step": 19510 + }, + { + "epoch": 0.45736856602026527, + "grad_norm": 3.0123074054718018, + "learning_rate": 0.00011339183380889125, + "loss": 0.262, + "step": 19511 + }, + { + "epoch": 0.45739200759507026, + "grad_norm": 3.2915172576904297, + "learning_rate": 0.00011338453396635338, + "loss": 0.4181, + "step": 19512 + }, + { + "epoch": 0.45741544916987525, + "grad_norm": 2.4923603534698486, + "learning_rate": 0.00011337723405119067, + "loss": 0.1718, + "step": 19513 + }, + { + "epoch": 0.45743889074468025, + "grad_norm": 1.0016086101531982, + "learning_rate": 0.00011336993406344267, + "loss": 0.0599, + "step": 19514 + }, + { + "epoch": 0.45746233231948524, + "grad_norm": 0.8194146752357483, + "learning_rate": 0.00011336263400314908, + "loss": 0.0729, + "step": 19515 + }, + { + "epoch": 0.45748577389429024, + "grad_norm": 3.001951217651367, + "learning_rate": 0.00011335533387034942, + "loss": 0.3116, + "step": 19516 + }, + { + "epoch": 0.45750921546909523, + "grad_norm": 0.6298375129699707, + "learning_rate": 0.00011334803366508338, + "loss": 0.0627, + "step": 19517 + }, + { + "epoch": 0.4575326570439002, + "grad_norm": 3.006176233291626, + "learning_rate": 0.00011334073338739051, + "loss": 0.2141, + "step": 19518 + }, + { + "epoch": 0.4575560986187052, + "grad_norm": 6.120827674865723, + "learning_rate": 0.00011333343303731046, + "loss": 0.4808, + "step": 19519 + }, + { + "epoch": 0.4575795401935102, + "grad_norm": 7.054782867431641, + "learning_rate": 0.00011332613261488281, + "loss": 0.3417, + "step": 19520 + }, + { + "epoch": 0.4576029817683152, + "grad_norm": 1.2633565664291382, + "learning_rate": 0.00011331883212014722, + "loss": 0.6041, + "step": 19521 + }, + { + "epoch": 0.4576264233431202, + "grad_norm": 5.348971366882324, + "learning_rate": 0.00011331153155314324, + "loss": 0.2498, + "step": 19522 + }, + { + "epoch": 0.4576498649179252, + "grad_norm": 1.9361345767974854, + "learning_rate": 0.00011330423091391054, + "loss": 0.9426, + "step": 19523 + }, + { + "epoch": 0.4576733064927302, + "grad_norm": 3.338548421859741, + "learning_rate": 0.0001132969302024887, + "loss": 0.294, + "step": 19524 + }, + { + "epoch": 0.4576967480675352, + "grad_norm": 1.617626428604126, + "learning_rate": 0.00011328962941891734, + "loss": 0.1465, + "step": 19525 + }, + { + "epoch": 0.4577201896423402, + "grad_norm": 2.8249950408935547, + "learning_rate": 0.00011328232856323608, + "loss": 0.2439, + "step": 19526 + }, + { + "epoch": 0.4577436312171452, + "grad_norm": 0.9585378170013428, + "learning_rate": 0.00011327502763548452, + "loss": 0.1555, + "step": 19527 + }, + { + "epoch": 0.45776707279195017, + "grad_norm": 0.6918816566467285, + "learning_rate": 0.00011326772663570229, + "loss": 0.0387, + "step": 19528 + }, + { + "epoch": 0.45779051436675516, + "grad_norm": 3.155134439468384, + "learning_rate": 0.00011326042556392902, + "loss": 0.295, + "step": 19529 + }, + { + "epoch": 0.45781395594156016, + "grad_norm": 1.2965155839920044, + "learning_rate": 0.0001132531244202043, + "loss": 0.1632, + "step": 19530 + }, + { + "epoch": 0.45783739751636515, + "grad_norm": 2.388437271118164, + "learning_rate": 0.00011324582320456774, + "loss": 0.1745, + "step": 19531 + }, + { + "epoch": 0.45786083909117015, + "grad_norm": 2.3148691654205322, + "learning_rate": 0.000113238521917059, + "loss": 1.5489, + "step": 19532 + }, + { + "epoch": 0.45788428066597514, + "grad_norm": 1.2306469678878784, + "learning_rate": 0.00011323122055771763, + "loss": 0.1221, + "step": 19533 + }, + { + "epoch": 0.45790772224078014, + "grad_norm": 3.697638511657715, + "learning_rate": 0.00011322391912658332, + "loss": 0.155, + "step": 19534 + }, + { + "epoch": 0.45793116381558513, + "grad_norm": 1.22174072265625, + "learning_rate": 0.00011321661762369562, + "loss": 0.1312, + "step": 19535 + }, + { + "epoch": 0.4579546053903901, + "grad_norm": 0.7490472793579102, + "learning_rate": 0.00011320931604909421, + "loss": 0.0696, + "step": 19536 + }, + { + "epoch": 0.4579780469651951, + "grad_norm": 2.460825204849243, + "learning_rate": 0.00011320201440281865, + "loss": 0.2403, + "step": 19537 + }, + { + "epoch": 0.4580014885400001, + "grad_norm": 0.9129469394683838, + "learning_rate": 0.00011319471268490862, + "loss": 0.0639, + "step": 19538 + }, + { + "epoch": 0.4580249301148051, + "grad_norm": 2.098590612411499, + "learning_rate": 0.00011318741089540368, + "loss": 1.1193, + "step": 19539 + }, + { + "epoch": 0.4580483716896101, + "grad_norm": 2.9208853244781494, + "learning_rate": 0.00011318010903434351, + "loss": 0.2455, + "step": 19540 + }, + { + "epoch": 0.4580718132644151, + "grad_norm": 3.9533262252807617, + "learning_rate": 0.00011317280710176767, + "loss": 0.3086, + "step": 19541 + }, + { + "epoch": 0.4580952548392201, + "grad_norm": 2.5351808071136475, + "learning_rate": 0.00011316550509771584, + "loss": 0.3007, + "step": 19542 + }, + { + "epoch": 0.4581186964140251, + "grad_norm": 2.7898762226104736, + "learning_rate": 0.0001131582030222276, + "loss": 0.2719, + "step": 19543 + }, + { + "epoch": 0.4581421379888301, + "grad_norm": 2.4500277042388916, + "learning_rate": 0.00011315090087534254, + "loss": 0.2039, + "step": 19544 + }, + { + "epoch": 0.4581655795636351, + "grad_norm": 2.591552257537842, + "learning_rate": 0.00011314359865710033, + "loss": 0.1727, + "step": 19545 + }, + { + "epoch": 0.45818902113844007, + "grad_norm": 4.353245735168457, + "learning_rate": 0.00011313629636754064, + "loss": 0.3814, + "step": 19546 + }, + { + "epoch": 0.45821246271324506, + "grad_norm": 3.18318772315979, + "learning_rate": 0.00011312899400670298, + "loss": 0.1447, + "step": 19547 + }, + { + "epoch": 0.45823590428805006, + "grad_norm": 4.332388401031494, + "learning_rate": 0.00011312169157462706, + "loss": 0.5122, + "step": 19548 + }, + { + "epoch": 0.45825934586285505, + "grad_norm": 3.841197967529297, + "learning_rate": 0.00011311438907135248, + "loss": 0.3779, + "step": 19549 + }, + { + "epoch": 0.45828278743766004, + "grad_norm": 3.339789628982544, + "learning_rate": 0.00011310708649691881, + "loss": 0.3772, + "step": 19550 + }, + { + "epoch": 0.45830622901246504, + "grad_norm": 3.1725146770477295, + "learning_rate": 0.00011309978385136574, + "loss": 0.2926, + "step": 19551 + }, + { + "epoch": 0.45832967058727003, + "grad_norm": 1.4030611515045166, + "learning_rate": 0.00011309248113473289, + "loss": 0.1292, + "step": 19552 + }, + { + "epoch": 0.458353112162075, + "grad_norm": 0.7919996976852417, + "learning_rate": 0.00011308517834705989, + "loss": 0.1035, + "step": 19553 + }, + { + "epoch": 0.45837655373688, + "grad_norm": 2.6334731578826904, + "learning_rate": 0.00011307787548838631, + "loss": 1.7389, + "step": 19554 + }, + { + "epoch": 0.458399995311685, + "grad_norm": 2.872727394104004, + "learning_rate": 0.00011307057255875182, + "loss": 0.2216, + "step": 19555 + }, + { + "epoch": 0.45842343688649, + "grad_norm": 0.9692941308021545, + "learning_rate": 0.00011306326955819602, + "loss": 0.1056, + "step": 19556 + }, + { + "epoch": 0.458446878461295, + "grad_norm": 4.123860836029053, + "learning_rate": 0.00011305596648675855, + "loss": 0.447, + "step": 19557 + }, + { + "epoch": 0.4584703200361, + "grad_norm": 1.5000360012054443, + "learning_rate": 0.00011304866334447908, + "loss": 0.1319, + "step": 19558 + }, + { + "epoch": 0.458493761610905, + "grad_norm": 5.6767354011535645, + "learning_rate": 0.00011304136013139716, + "loss": 0.621, + "step": 19559 + }, + { + "epoch": 0.45851720318571004, + "grad_norm": 2.289342164993286, + "learning_rate": 0.00011303405684755247, + "loss": 0.1362, + "step": 19560 + }, + { + "epoch": 0.45854064476051504, + "grad_norm": 2.942768096923828, + "learning_rate": 0.00011302675349298461, + "loss": 1.4081, + "step": 19561 + }, + { + "epoch": 0.45856408633532003, + "grad_norm": 2.95414662361145, + "learning_rate": 0.00011301945006773324, + "loss": 0.3141, + "step": 19562 + }, + { + "epoch": 0.458587527910125, + "grad_norm": 1.8726792335510254, + "learning_rate": 0.00011301214657183791, + "loss": 0.1983, + "step": 19563 + }, + { + "epoch": 0.45861096948493, + "grad_norm": 2.7323062419891357, + "learning_rate": 0.00011300484300533835, + "loss": 0.3262, + "step": 19564 + }, + { + "epoch": 0.458634411059735, + "grad_norm": 4.574333190917969, + "learning_rate": 0.00011299753936827418, + "loss": 0.3055, + "step": 19565 + }, + { + "epoch": 0.45865785263454, + "grad_norm": 3.669494867324829, + "learning_rate": 0.00011299023566068494, + "loss": 0.3365, + "step": 19566 + }, + { + "epoch": 0.458681294209345, + "grad_norm": 3.1645009517669678, + "learning_rate": 0.00011298293188261035, + "loss": 0.2673, + "step": 19567 + }, + { + "epoch": 0.45870473578415, + "grad_norm": 1.7229288816452026, + "learning_rate": 0.00011297562803409, + "loss": 0.2392, + "step": 19568 + }, + { + "epoch": 0.458728177358955, + "grad_norm": 2.882493019104004, + "learning_rate": 0.00011296832411516352, + "loss": 0.1742, + "step": 19569 + }, + { + "epoch": 0.45875161893376, + "grad_norm": 1.4888681173324585, + "learning_rate": 0.00011296102012587053, + "loss": 0.1157, + "step": 19570 + }, + { + "epoch": 0.458775060508565, + "grad_norm": 0.8901292681694031, + "learning_rate": 0.00011295371606625072, + "loss": 0.1026, + "step": 19571 + }, + { + "epoch": 0.45879850208337, + "grad_norm": 3.919945001602173, + "learning_rate": 0.00011294641193634364, + "loss": 0.4335, + "step": 19572 + }, + { + "epoch": 0.45882194365817497, + "grad_norm": 3.0778470039367676, + "learning_rate": 0.00011293910773618899, + "loss": 0.1666, + "step": 19573 + }, + { + "epoch": 0.45884538523297996, + "grad_norm": 1.7089093923568726, + "learning_rate": 0.00011293180346582639, + "loss": 0.1474, + "step": 19574 + }, + { + "epoch": 0.45886882680778496, + "grad_norm": 1.7217451333999634, + "learning_rate": 0.00011292449912529543, + "loss": 0.2033, + "step": 19575 + }, + { + "epoch": 0.45889226838258995, + "grad_norm": 3.6828176975250244, + "learning_rate": 0.00011291719471463578, + "loss": 0.3962, + "step": 19576 + }, + { + "epoch": 0.45891570995739495, + "grad_norm": 3.199815511703491, + "learning_rate": 0.0001129098902338871, + "loss": 0.2103, + "step": 19577 + }, + { + "epoch": 0.45893915153219994, + "grad_norm": 3.9943788051605225, + "learning_rate": 0.00011290258568308895, + "loss": 0.3302, + "step": 19578 + }, + { + "epoch": 0.45896259310700493, + "grad_norm": 3.6923203468322754, + "learning_rate": 0.00011289528106228102, + "loss": 0.2134, + "step": 19579 + }, + { + "epoch": 0.45898603468180993, + "grad_norm": 2.6961281299591064, + "learning_rate": 0.00011288797637150297, + "loss": 0.2439, + "step": 19580 + }, + { + "epoch": 0.4590094762566149, + "grad_norm": 3.5461385250091553, + "learning_rate": 0.00011288067161079434, + "loss": 0.6378, + "step": 19581 + }, + { + "epoch": 0.4590329178314199, + "grad_norm": 3.062617778778076, + "learning_rate": 0.00011287336678019484, + "loss": 0.4571, + "step": 19582 + }, + { + "epoch": 0.4590563594062249, + "grad_norm": 2.342817544937134, + "learning_rate": 0.00011286606187974411, + "loss": 0.1391, + "step": 19583 + }, + { + "epoch": 0.4590798009810299, + "grad_norm": 3.0516741275787354, + "learning_rate": 0.00011285875690948178, + "loss": 1.3435, + "step": 19584 + }, + { + "epoch": 0.4591032425558349, + "grad_norm": 5.658666610717773, + "learning_rate": 0.00011285145186944744, + "loss": 0.6626, + "step": 19585 + }, + { + "epoch": 0.4591266841306399, + "grad_norm": 0.6655724048614502, + "learning_rate": 0.00011284414675968078, + "loss": 0.1008, + "step": 19586 + }, + { + "epoch": 0.4591501257054449, + "grad_norm": 3.3905837535858154, + "learning_rate": 0.0001128368415802214, + "loss": 0.4611, + "step": 19587 + }, + { + "epoch": 0.4591735672802499, + "grad_norm": 2.0205938816070557, + "learning_rate": 0.00011282953633110897, + "loss": 0.2222, + "step": 19588 + }, + { + "epoch": 0.4591970088550549, + "grad_norm": 5.828976631164551, + "learning_rate": 0.00011282223101238312, + "loss": 0.4359, + "step": 19589 + }, + { + "epoch": 0.45922045042985987, + "grad_norm": 3.48476243019104, + "learning_rate": 0.00011281492562408347, + "loss": 0.18, + "step": 19590 + }, + { + "epoch": 0.45924389200466487, + "grad_norm": 1.0100442171096802, + "learning_rate": 0.00011280762016624966, + "loss": 0.1138, + "step": 19591 + }, + { + "epoch": 0.45926733357946986, + "grad_norm": 3.9382755756378174, + "learning_rate": 0.00011280031463892138, + "loss": 0.2703, + "step": 19592 + }, + { + "epoch": 0.45929077515427486, + "grad_norm": 3.3504562377929688, + "learning_rate": 0.00011279300904213822, + "loss": 0.1754, + "step": 19593 + }, + { + "epoch": 0.45931421672907985, + "grad_norm": 1.7716996669769287, + "learning_rate": 0.00011278570337593983, + "loss": 0.2627, + "step": 19594 + }, + { + "epoch": 0.45933765830388484, + "grad_norm": 2.0484461784362793, + "learning_rate": 0.00011277839764036584, + "loss": 0.1504, + "step": 19595 + }, + { + "epoch": 0.45936109987868984, + "grad_norm": 3.1135458946228027, + "learning_rate": 0.00011277109183545593, + "loss": 0.2951, + "step": 19596 + }, + { + "epoch": 0.45938454145349483, + "grad_norm": 5.446203708648682, + "learning_rate": 0.0001127637859612497, + "loss": 0.757, + "step": 19597 + }, + { + "epoch": 0.4594079830282998, + "grad_norm": 4.427317142486572, + "learning_rate": 0.0001127564800177868, + "loss": 0.4219, + "step": 19598 + }, + { + "epoch": 0.4594314246031048, + "grad_norm": 3.3320529460906982, + "learning_rate": 0.0001127491740051069, + "loss": 0.4645, + "step": 19599 + }, + { + "epoch": 0.4594548661779098, + "grad_norm": 0.8668488264083862, + "learning_rate": 0.00011274186792324961, + "loss": 0.1042, + "step": 19600 + }, + { + "epoch": 0.4594783077527148, + "grad_norm": 5.376486301422119, + "learning_rate": 0.00011273456177225459, + "loss": 0.4417, + "step": 19601 + }, + { + "epoch": 0.4595017493275198, + "grad_norm": 3.330487012863159, + "learning_rate": 0.0001127272555521615, + "loss": 0.2957, + "step": 19602 + }, + { + "epoch": 0.4595251909023248, + "grad_norm": 3.6750245094299316, + "learning_rate": 0.00011271994926300996, + "loss": 0.4003, + "step": 19603 + }, + { + "epoch": 0.4595486324771298, + "grad_norm": 2.7387163639068604, + "learning_rate": 0.00011271264290483958, + "loss": 1.3485, + "step": 19604 + }, + { + "epoch": 0.4595720740519348, + "grad_norm": 0.9621925354003906, + "learning_rate": 0.00011270533647769008, + "loss": 0.0628, + "step": 19605 + }, + { + "epoch": 0.4595955156267398, + "grad_norm": 3.3965888023376465, + "learning_rate": 0.00011269802998160103, + "loss": 0.4324, + "step": 19606 + }, + { + "epoch": 0.4596189572015448, + "grad_norm": 1.5004087686538696, + "learning_rate": 0.00011269072341661216, + "loss": 0.2101, + "step": 19607 + }, + { + "epoch": 0.45964239877634977, + "grad_norm": 2.1859865188598633, + "learning_rate": 0.00011268341678276302, + "loss": 1.4525, + "step": 19608 + }, + { + "epoch": 0.45966584035115476, + "grad_norm": 0.9098352193832397, + "learning_rate": 0.00011267611008009336, + "loss": 0.0365, + "step": 19609 + }, + { + "epoch": 0.45968928192595976, + "grad_norm": 2.8787403106689453, + "learning_rate": 0.00011266880330864271, + "loss": 0.3363, + "step": 19610 + }, + { + "epoch": 0.4597127235007648, + "grad_norm": 4.21389627456665, + "learning_rate": 0.00011266149646845081, + "loss": 0.5319, + "step": 19611 + }, + { + "epoch": 0.4597361650755698, + "grad_norm": 2.656580924987793, + "learning_rate": 0.00011265418955955727, + "loss": 0.1962, + "step": 19612 + }, + { + "epoch": 0.4597596066503748, + "grad_norm": 4.105010986328125, + "learning_rate": 0.00011264688258200175, + "loss": 0.4273, + "step": 19613 + }, + { + "epoch": 0.4597830482251798, + "grad_norm": 3.5012786388397217, + "learning_rate": 0.00011263957553582388, + "loss": 0.3896, + "step": 19614 + }, + { + "epoch": 0.4598064897999848, + "grad_norm": 2.9852700233459473, + "learning_rate": 0.00011263226842106332, + "loss": 0.3613, + "step": 19615 + }, + { + "epoch": 0.4598299313747898, + "grad_norm": 1.2279949188232422, + "learning_rate": 0.00011262496123775972, + "loss": 0.1702, + "step": 19616 + }, + { + "epoch": 0.4598533729495948, + "grad_norm": 3.822678327560425, + "learning_rate": 0.0001126176539859527, + "loss": 0.2632, + "step": 19617 + }, + { + "epoch": 0.45987681452439977, + "grad_norm": 3.442601203918457, + "learning_rate": 0.00011261034666568195, + "loss": 0.4424, + "step": 19618 + }, + { + "epoch": 0.45990025609920476, + "grad_norm": 3.4083120822906494, + "learning_rate": 0.00011260303927698712, + "loss": 0.2586, + "step": 19619 + }, + { + "epoch": 0.45992369767400976, + "grad_norm": 3.1993606090545654, + "learning_rate": 0.00011259573181990782, + "loss": 0.2065, + "step": 19620 + }, + { + "epoch": 0.45994713924881475, + "grad_norm": 2.327197790145874, + "learning_rate": 0.00011258842429448376, + "loss": 0.1256, + "step": 19621 + }, + { + "epoch": 0.45997058082361975, + "grad_norm": 3.301696538925171, + "learning_rate": 0.00011258111670075453, + "loss": 0.3384, + "step": 19622 + }, + { + "epoch": 0.45999402239842474, + "grad_norm": 2.310072660446167, + "learning_rate": 0.00011257380903875981, + "loss": 0.2043, + "step": 19623 + }, + { + "epoch": 0.46001746397322973, + "grad_norm": 4.5249481201171875, + "learning_rate": 0.00011256650130853923, + "loss": 0.2655, + "step": 19624 + }, + { + "epoch": 0.46004090554803473, + "grad_norm": 5.111729621887207, + "learning_rate": 0.00011255919351013248, + "loss": 0.3187, + "step": 19625 + }, + { + "epoch": 0.4600643471228397, + "grad_norm": 1.7782820463180542, + "learning_rate": 0.00011255188564357921, + "loss": 0.1436, + "step": 19626 + }, + { + "epoch": 0.4600877886976447, + "grad_norm": 4.475370407104492, + "learning_rate": 0.00011254457770891905, + "loss": 0.4524, + "step": 19627 + }, + { + "epoch": 0.4601112302724497, + "grad_norm": 3.7739315032958984, + "learning_rate": 0.00011253726970619162, + "loss": 0.5303, + "step": 19628 + }, + { + "epoch": 0.4601346718472547, + "grad_norm": 3.3802812099456787, + "learning_rate": 0.00011252996163543665, + "loss": 0.3103, + "step": 19629 + }, + { + "epoch": 0.4601581134220597, + "grad_norm": 4.535321235656738, + "learning_rate": 0.00011252265349669373, + "loss": 0.3744, + "step": 19630 + }, + { + "epoch": 0.4601815549968647, + "grad_norm": 3.9162778854370117, + "learning_rate": 0.00011251534529000257, + "loss": 0.3754, + "step": 19631 + }, + { + "epoch": 0.4602049965716697, + "grad_norm": 0.7276287078857422, + "learning_rate": 0.00011250803701540277, + "loss": 0.0605, + "step": 19632 + }, + { + "epoch": 0.4602284381464747, + "grad_norm": 2.891512632369995, + "learning_rate": 0.00011250072867293401, + "loss": 0.2512, + "step": 19633 + }, + { + "epoch": 0.4602518797212797, + "grad_norm": 3.4500679969787598, + "learning_rate": 0.00011249342026263598, + "loss": 0.3004, + "step": 19634 + }, + { + "epoch": 0.46027532129608467, + "grad_norm": 3.464047431945801, + "learning_rate": 0.00011248611178454829, + "loss": 0.3378, + "step": 19635 + }, + { + "epoch": 0.46029876287088967, + "grad_norm": 1.256828784942627, + "learning_rate": 0.00011247880323871053, + "loss": 0.4648, + "step": 19636 + }, + { + "epoch": 0.46032220444569466, + "grad_norm": 3.958089828491211, + "learning_rate": 0.00011247149462516251, + "loss": 0.3399, + "step": 19637 + }, + { + "epoch": 0.46034564602049965, + "grad_norm": 4.07022762298584, + "learning_rate": 0.00011246418594394381, + "loss": 0.3438, + "step": 19638 + }, + { + "epoch": 0.46036908759530465, + "grad_norm": 0.4756706953048706, + "learning_rate": 0.00011245687719509405, + "loss": 0.0371, + "step": 19639 + }, + { + "epoch": 0.46039252917010964, + "grad_norm": 0.6332361102104187, + "learning_rate": 0.00011244956837865297, + "loss": 0.0412, + "step": 19640 + }, + { + "epoch": 0.46041597074491464, + "grad_norm": 1.099361777305603, + "learning_rate": 0.00011244225949466017, + "loss": 0.6904, + "step": 19641 + }, + { + "epoch": 0.46043941231971963, + "grad_norm": 3.703540325164795, + "learning_rate": 0.0001124349505431553, + "loss": 0.4154, + "step": 19642 + }, + { + "epoch": 0.4604628538945246, + "grad_norm": 1.5872395038604736, + "learning_rate": 0.00011242764152417802, + "loss": 0.1473, + "step": 19643 + }, + { + "epoch": 0.4604862954693296, + "grad_norm": 3.771576166152954, + "learning_rate": 0.00011242033243776804, + "loss": 0.4199, + "step": 19644 + }, + { + "epoch": 0.4605097370441346, + "grad_norm": 3.1651663780212402, + "learning_rate": 0.00011241302328396499, + "loss": 0.264, + "step": 19645 + }, + { + "epoch": 0.4605331786189396, + "grad_norm": 3.066840648651123, + "learning_rate": 0.00011240571406280851, + "loss": 0.3817, + "step": 19646 + }, + { + "epoch": 0.4605566201937446, + "grad_norm": 1.2575591802597046, + "learning_rate": 0.00011239840477433829, + "loss": 0.1517, + "step": 19647 + }, + { + "epoch": 0.4605800617685496, + "grad_norm": 0.9341135025024414, + "learning_rate": 0.00011239109541859397, + "loss": 0.0739, + "step": 19648 + }, + { + "epoch": 0.4606035033433546, + "grad_norm": 2.68165922164917, + "learning_rate": 0.00011238378599561519, + "loss": 0.2963, + "step": 19649 + }, + { + "epoch": 0.4606269449181596, + "grad_norm": 4.261504173278809, + "learning_rate": 0.00011237647650544167, + "loss": 0.2887, + "step": 19650 + }, + { + "epoch": 0.4606503864929646, + "grad_norm": 4.384532928466797, + "learning_rate": 0.00011236916694811306, + "loss": 0.5544, + "step": 19651 + }, + { + "epoch": 0.4606738280677696, + "grad_norm": 1.889136791229248, + "learning_rate": 0.00011236185732366896, + "loss": 0.1952, + "step": 19652 + }, + { + "epoch": 0.46069726964257457, + "grad_norm": 2.266922950744629, + "learning_rate": 0.0001123545476321491, + "loss": 0.2297, + "step": 19653 + }, + { + "epoch": 0.46072071121737956, + "grad_norm": 2.720512628555298, + "learning_rate": 0.00011234723787359311, + "loss": 0.152, + "step": 19654 + }, + { + "epoch": 0.46074415279218456, + "grad_norm": 4.855443954467773, + "learning_rate": 0.00011233992804804063, + "loss": 0.6102, + "step": 19655 + }, + { + "epoch": 0.46076759436698955, + "grad_norm": 1.9979535341262817, + "learning_rate": 0.0001123326181555314, + "loss": 0.1279, + "step": 19656 + }, + { + "epoch": 0.46079103594179455, + "grad_norm": 4.827664375305176, + "learning_rate": 0.00011232530819610502, + "loss": 0.5876, + "step": 19657 + }, + { + "epoch": 0.46081447751659954, + "grad_norm": 1.8094288110733032, + "learning_rate": 0.00011231799816980117, + "loss": 0.2261, + "step": 19658 + }, + { + "epoch": 0.46083791909140454, + "grad_norm": 3.834249973297119, + "learning_rate": 0.00011231068807665955, + "loss": 0.4827, + "step": 19659 + }, + { + "epoch": 0.46086136066620953, + "grad_norm": 0.6742256283760071, + "learning_rate": 0.00011230337791671974, + "loss": 0.0854, + "step": 19660 + }, + { + "epoch": 0.4608848022410145, + "grad_norm": 3.1247591972351074, + "learning_rate": 0.00011229606769002147, + "loss": 0.1948, + "step": 19661 + }, + { + "epoch": 0.4609082438158196, + "grad_norm": 2.943138360977173, + "learning_rate": 0.00011228875739660438, + "loss": 1.7223, + "step": 19662 + }, + { + "epoch": 0.46093168539062457, + "grad_norm": 4.19885778427124, + "learning_rate": 0.00011228144703650818, + "loss": 0.3275, + "step": 19663 + }, + { + "epoch": 0.46095512696542956, + "grad_norm": 3.1109743118286133, + "learning_rate": 0.0001122741366097725, + "loss": 1.6609, + "step": 19664 + }, + { + "epoch": 0.46097856854023456, + "grad_norm": 0.858981728553772, + "learning_rate": 0.00011226682611643698, + "loss": 0.0756, + "step": 19665 + }, + { + "epoch": 0.46100201011503955, + "grad_norm": 2.395007848739624, + "learning_rate": 0.00011225951555654133, + "loss": 0.223, + "step": 19666 + }, + { + "epoch": 0.46102545168984455, + "grad_norm": 2.443293571472168, + "learning_rate": 0.0001122522049301252, + "loss": 0.2798, + "step": 19667 + }, + { + "epoch": 0.46104889326464954, + "grad_norm": 1.8069257736206055, + "learning_rate": 0.00011224489423722826, + "loss": 0.0915, + "step": 19668 + }, + { + "epoch": 0.46107233483945453, + "grad_norm": 4.0164666175842285, + "learning_rate": 0.00011223758347789019, + "loss": 0.3084, + "step": 19669 + }, + { + "epoch": 0.46109577641425953, + "grad_norm": 0.9169342517852783, + "learning_rate": 0.00011223027265215067, + "loss": 0.0596, + "step": 19670 + }, + { + "epoch": 0.4611192179890645, + "grad_norm": 2.2610702514648438, + "learning_rate": 0.0001122229617600493, + "loss": 0.1076, + "step": 19671 + }, + { + "epoch": 0.4611426595638695, + "grad_norm": 1.8891065120697021, + "learning_rate": 0.00011221565080162582, + "loss": 0.5906, + "step": 19672 + }, + { + "epoch": 0.4611661011386745, + "grad_norm": 3.0733537673950195, + "learning_rate": 0.00011220833977691986, + "loss": 0.3065, + "step": 19673 + }, + { + "epoch": 0.4611895427134795, + "grad_norm": 4.093722343444824, + "learning_rate": 0.00011220102868597111, + "loss": 0.3054, + "step": 19674 + }, + { + "epoch": 0.4612129842882845, + "grad_norm": 2.91762638092041, + "learning_rate": 0.00011219371752881927, + "loss": 0.1895, + "step": 19675 + }, + { + "epoch": 0.4612364258630895, + "grad_norm": 5.525267601013184, + "learning_rate": 0.00011218640630550394, + "loss": 0.5644, + "step": 19676 + }, + { + "epoch": 0.4612598674378945, + "grad_norm": 3.7038943767547607, + "learning_rate": 0.00011217909501606482, + "loss": 0.567, + "step": 19677 + }, + { + "epoch": 0.4612833090126995, + "grad_norm": 3.021524667739868, + "learning_rate": 0.00011217178366054162, + "loss": 0.3937, + "step": 19678 + }, + { + "epoch": 0.4613067505875045, + "grad_norm": 1.052045226097107, + "learning_rate": 0.00011216447223897395, + "loss": 0.0776, + "step": 19679 + }, + { + "epoch": 0.46133019216230947, + "grad_norm": 3.1585283279418945, + "learning_rate": 0.00011215716075140155, + "loss": 0.2038, + "step": 19680 + }, + { + "epoch": 0.46135363373711447, + "grad_norm": 2.3529303073883057, + "learning_rate": 0.000112149849197864, + "loss": 0.9611, + "step": 19681 + }, + { + "epoch": 0.46137707531191946, + "grad_norm": 3.3760173320770264, + "learning_rate": 0.00011214253757840108, + "loss": 0.2853, + "step": 19682 + }, + { + "epoch": 0.46140051688672445, + "grad_norm": 1.537053108215332, + "learning_rate": 0.00011213522589305239, + "loss": 0.1344, + "step": 19683 + }, + { + "epoch": 0.46142395846152945, + "grad_norm": 3.2729876041412354, + "learning_rate": 0.0001121279141418576, + "loss": 0.3438, + "step": 19684 + }, + { + "epoch": 0.46144740003633444, + "grad_norm": 0.7434825897216797, + "learning_rate": 0.00011212060232485642, + "loss": 0.0645, + "step": 19685 + }, + { + "epoch": 0.46147084161113944, + "grad_norm": 1.644364356994629, + "learning_rate": 0.00011211329044208851, + "loss": 0.1171, + "step": 19686 + }, + { + "epoch": 0.46149428318594443, + "grad_norm": 2.872299909591675, + "learning_rate": 0.00011210597849359355, + "loss": 0.3065, + "step": 19687 + }, + { + "epoch": 0.4615177247607494, + "grad_norm": 2.416083335876465, + "learning_rate": 0.00011209866647941121, + "loss": 0.1049, + "step": 19688 + }, + { + "epoch": 0.4615411663355544, + "grad_norm": 3.712683916091919, + "learning_rate": 0.00011209135439958118, + "loss": 0.2863, + "step": 19689 + }, + { + "epoch": 0.4615646079103594, + "grad_norm": 3.819873094558716, + "learning_rate": 0.0001120840422541431, + "loss": 0.3279, + "step": 19690 + }, + { + "epoch": 0.4615880494851644, + "grad_norm": 2.7727668285369873, + "learning_rate": 0.00011207673004313666, + "loss": 0.1511, + "step": 19691 + }, + { + "epoch": 0.4616114910599694, + "grad_norm": 2.705493688583374, + "learning_rate": 0.00011206941776660159, + "loss": 0.1385, + "step": 19692 + }, + { + "epoch": 0.4616349326347744, + "grad_norm": 0.9838907718658447, + "learning_rate": 0.00011206210542457745, + "loss": 0.0535, + "step": 19693 + }, + { + "epoch": 0.4616583742095794, + "grad_norm": 2.8896679878234863, + "learning_rate": 0.00011205479301710405, + "loss": 0.2594, + "step": 19694 + }, + { + "epoch": 0.4616818157843844, + "grad_norm": 1.2507028579711914, + "learning_rate": 0.00011204748054422099, + "loss": 0.1068, + "step": 19695 + }, + { + "epoch": 0.4617052573591894, + "grad_norm": 2.8545844554901123, + "learning_rate": 0.00011204016800596792, + "loss": 0.3731, + "step": 19696 + }, + { + "epoch": 0.4617286989339944, + "grad_norm": 3.3131775856018066, + "learning_rate": 0.0001120328554023846, + "loss": 1.4636, + "step": 19697 + }, + { + "epoch": 0.46175214050879937, + "grad_norm": 2.2977378368377686, + "learning_rate": 0.00011202554273351065, + "loss": 0.142, + "step": 19698 + }, + { + "epoch": 0.46177558208360436, + "grad_norm": 3.8462300300598145, + "learning_rate": 0.0001120182299993858, + "loss": 0.3, + "step": 19699 + }, + { + "epoch": 0.46179902365840936, + "grad_norm": 1.0351312160491943, + "learning_rate": 0.00011201091720004965, + "loss": 0.0946, + "step": 19700 + }, + { + "epoch": 0.46182246523321435, + "grad_norm": 3.5710608959198, + "learning_rate": 0.00011200360433554195, + "loss": 0.2895, + "step": 19701 + }, + { + "epoch": 0.46184590680801935, + "grad_norm": 3.671354293823242, + "learning_rate": 0.00011199629140590238, + "loss": 0.4469, + "step": 19702 + }, + { + "epoch": 0.46186934838282434, + "grad_norm": 0.8278492093086243, + "learning_rate": 0.00011198897841117054, + "loss": 0.059, + "step": 19703 + }, + { + "epoch": 0.46189278995762933, + "grad_norm": 2.731107234954834, + "learning_rate": 0.0001119816653513862, + "loss": 0.2235, + "step": 19704 + }, + { + "epoch": 0.46191623153243433, + "grad_norm": 0.2977452576160431, + "learning_rate": 0.00011197435222658901, + "loss": 0.0153, + "step": 19705 + }, + { + "epoch": 0.4619396731072393, + "grad_norm": 4.737465858459473, + "learning_rate": 0.00011196703903681863, + "loss": 0.3865, + "step": 19706 + }, + { + "epoch": 0.4619631146820443, + "grad_norm": 4.163116455078125, + "learning_rate": 0.00011195972578211478, + "loss": 0.3978, + "step": 19707 + }, + { + "epoch": 0.4619865562568493, + "grad_norm": 0.8009359240531921, + "learning_rate": 0.0001119524124625171, + "loss": 0.0597, + "step": 19708 + }, + { + "epoch": 0.4620099978316543, + "grad_norm": 2.2166824340820312, + "learning_rate": 0.00011194509907806528, + "loss": 0.1906, + "step": 19709 + }, + { + "epoch": 0.4620334394064593, + "grad_norm": 1.3782004117965698, + "learning_rate": 0.00011193778562879905, + "loss": 0.0846, + "step": 19710 + }, + { + "epoch": 0.4620568809812643, + "grad_norm": 1.4269317388534546, + "learning_rate": 0.00011193047211475806, + "loss": 0.615, + "step": 19711 + }, + { + "epoch": 0.4620803225560693, + "grad_norm": 4.1343207359313965, + "learning_rate": 0.00011192315853598197, + "loss": 0.4672, + "step": 19712 + }, + { + "epoch": 0.46210376413087434, + "grad_norm": 4.427908420562744, + "learning_rate": 0.00011191584489251053, + "loss": 0.3389, + "step": 19713 + }, + { + "epoch": 0.46212720570567933, + "grad_norm": 3.219040632247925, + "learning_rate": 0.00011190853118438333, + "loss": 0.3465, + "step": 19714 + }, + { + "epoch": 0.46215064728048433, + "grad_norm": 2.398440361022949, + "learning_rate": 0.00011190121741164012, + "loss": 0.2237, + "step": 19715 + }, + { + "epoch": 0.4621740888552893, + "grad_norm": 3.540513515472412, + "learning_rate": 0.00011189390357432057, + "loss": 1.6805, + "step": 19716 + }, + { + "epoch": 0.4621975304300943, + "grad_norm": 0.5579545497894287, + "learning_rate": 0.00011188658967246439, + "loss": 0.0657, + "step": 19717 + }, + { + "epoch": 0.4622209720048993, + "grad_norm": 3.4578888416290283, + "learning_rate": 0.00011187927570611122, + "loss": 0.2859, + "step": 19718 + }, + { + "epoch": 0.4622444135797043, + "grad_norm": 2.5614445209503174, + "learning_rate": 0.00011187196167530077, + "loss": 0.2581, + "step": 19719 + }, + { + "epoch": 0.4622678551545093, + "grad_norm": 5.317439079284668, + "learning_rate": 0.00011186464758007271, + "loss": 0.3514, + "step": 19720 + }, + { + "epoch": 0.4622912967293143, + "grad_norm": 2.730417013168335, + "learning_rate": 0.00011185733342046674, + "loss": 0.3571, + "step": 19721 + }, + { + "epoch": 0.4623147383041193, + "grad_norm": 1.0905330181121826, + "learning_rate": 0.00011185001919652254, + "loss": 0.1358, + "step": 19722 + }, + { + "epoch": 0.4623381798789243, + "grad_norm": 4.435669422149658, + "learning_rate": 0.00011184270490827982, + "loss": 0.2523, + "step": 19723 + }, + { + "epoch": 0.4623616214537293, + "grad_norm": 1.6699036359786987, + "learning_rate": 0.00011183539055577826, + "loss": 0.0818, + "step": 19724 + }, + { + "epoch": 0.46238506302853427, + "grad_norm": 0.6882646679878235, + "learning_rate": 0.0001118280761390575, + "loss": 0.088, + "step": 19725 + }, + { + "epoch": 0.46240850460333927, + "grad_norm": 3.5545713901519775, + "learning_rate": 0.00011182076165815732, + "loss": 0.2923, + "step": 19726 + }, + { + "epoch": 0.46243194617814426, + "grad_norm": 2.1456973552703857, + "learning_rate": 0.0001118134471131173, + "loss": 0.2717, + "step": 19727 + }, + { + "epoch": 0.46245538775294925, + "grad_norm": 0.31291958689689636, + "learning_rate": 0.0001118061325039772, + "loss": 0.021, + "step": 19728 + }, + { + "epoch": 0.46247882932775425, + "grad_norm": 2.1474127769470215, + "learning_rate": 0.00011179881783077671, + "loss": 0.2412, + "step": 19729 + }, + { + "epoch": 0.46250227090255924, + "grad_norm": 3.6340606212615967, + "learning_rate": 0.0001117915030935555, + "loss": 0.2833, + "step": 19730 + }, + { + "epoch": 0.46252571247736424, + "grad_norm": 4.254147529602051, + "learning_rate": 0.00011178418829235324, + "loss": 0.463, + "step": 19731 + }, + { + "epoch": 0.46254915405216923, + "grad_norm": 1.0037620067596436, + "learning_rate": 0.00011177687342720969, + "loss": 0.0628, + "step": 19732 + }, + { + "epoch": 0.4625725956269742, + "grad_norm": 2.4441416263580322, + "learning_rate": 0.00011176955849816444, + "loss": 0.2707, + "step": 19733 + }, + { + "epoch": 0.4625960372017792, + "grad_norm": 1.995373249053955, + "learning_rate": 0.00011176224350525728, + "loss": 1.3805, + "step": 19734 + }, + { + "epoch": 0.4626194787765842, + "grad_norm": 3.539595365524292, + "learning_rate": 0.00011175492844852782, + "loss": 0.2452, + "step": 19735 + }, + { + "epoch": 0.4626429203513892, + "grad_norm": 0.750554621219635, + "learning_rate": 0.0001117476133280158, + "loss": 0.0664, + "step": 19736 + }, + { + "epoch": 0.4626663619261942, + "grad_norm": 0.7119802236557007, + "learning_rate": 0.00011174029814376092, + "loss": 0.0914, + "step": 19737 + }, + { + "epoch": 0.4626898035009992, + "grad_norm": 4.576741695404053, + "learning_rate": 0.00011173298289580283, + "loss": 0.2056, + "step": 19738 + }, + { + "epoch": 0.4627132450758042, + "grad_norm": 3.916110038757324, + "learning_rate": 0.00011172566758418122, + "loss": 0.5029, + "step": 19739 + }, + { + "epoch": 0.4627366866506092, + "grad_norm": 0.8736225366592407, + "learning_rate": 0.00011171835220893585, + "loss": 0.0676, + "step": 19740 + }, + { + "epoch": 0.4627601282254142, + "grad_norm": 3.415916681289673, + "learning_rate": 0.00011171103677010634, + "loss": 0.3096, + "step": 19741 + }, + { + "epoch": 0.4627835698002192, + "grad_norm": 2.473773956298828, + "learning_rate": 0.00011170372126773245, + "loss": 0.2226, + "step": 19742 + }, + { + "epoch": 0.46280701137502417, + "grad_norm": 4.477961540222168, + "learning_rate": 0.00011169640570185382, + "loss": 0.469, + "step": 19743 + }, + { + "epoch": 0.46283045294982916, + "grad_norm": 3.374915838241577, + "learning_rate": 0.00011168909007251015, + "loss": 0.4423, + "step": 19744 + }, + { + "epoch": 0.46285389452463416, + "grad_norm": 6.253993034362793, + "learning_rate": 0.00011168177437974116, + "loss": 0.1627, + "step": 19745 + }, + { + "epoch": 0.46287733609943915, + "grad_norm": 1.907569408416748, + "learning_rate": 0.00011167445862358657, + "loss": 0.1072, + "step": 19746 + }, + { + "epoch": 0.46290077767424415, + "grad_norm": 2.8625967502593994, + "learning_rate": 0.000111667142804086, + "loss": 1.6999, + "step": 19747 + }, + { + "epoch": 0.46292421924904914, + "grad_norm": 3.583136558532715, + "learning_rate": 0.00011165982692127918, + "loss": 0.2649, + "step": 19748 + }, + { + "epoch": 0.46294766082385413, + "grad_norm": 3.7386345863342285, + "learning_rate": 0.00011165251097520584, + "loss": 1.9267, + "step": 19749 + }, + { + "epoch": 0.46297110239865913, + "grad_norm": 3.163588047027588, + "learning_rate": 0.00011164519496590562, + "loss": 0.2782, + "step": 19750 + }, + { + "epoch": 0.4629945439734641, + "grad_norm": 4.5555291175842285, + "learning_rate": 0.00011163787889341825, + "loss": 0.5748, + "step": 19751 + }, + { + "epoch": 0.4630179855482691, + "grad_norm": 4.056031703948975, + "learning_rate": 0.00011163056275778343, + "loss": 0.6342, + "step": 19752 + }, + { + "epoch": 0.4630414271230741, + "grad_norm": 2.7167181968688965, + "learning_rate": 0.00011162324655904086, + "loss": 0.2393, + "step": 19753 + }, + { + "epoch": 0.4630648686978791, + "grad_norm": 4.713411808013916, + "learning_rate": 0.00011161593029723021, + "loss": 0.4952, + "step": 19754 + }, + { + "epoch": 0.4630883102726841, + "grad_norm": 5.30246114730835, + "learning_rate": 0.00011160861397239122, + "loss": 0.296, + "step": 19755 + }, + { + "epoch": 0.4631117518474891, + "grad_norm": 2.334444522857666, + "learning_rate": 0.00011160129758456355, + "loss": 0.284, + "step": 19756 + }, + { + "epoch": 0.4631351934222941, + "grad_norm": 2.8461437225341797, + "learning_rate": 0.0001115939811337869, + "loss": 0.3645, + "step": 19757 + }, + { + "epoch": 0.4631586349970991, + "grad_norm": 3.343832015991211, + "learning_rate": 0.00011158666462010096, + "loss": 0.2233, + "step": 19758 + }, + { + "epoch": 0.4631820765719041, + "grad_norm": 1.3180961608886719, + "learning_rate": 0.0001115793480435455, + "loss": 0.0828, + "step": 19759 + }, + { + "epoch": 0.46320551814670907, + "grad_norm": 1.2754744291305542, + "learning_rate": 0.00011157203140416015, + "loss": 0.1565, + "step": 19760 + }, + { + "epoch": 0.46322895972151407, + "grad_norm": 3.2030789852142334, + "learning_rate": 0.00011156471470198464, + "loss": 0.2954, + "step": 19761 + }, + { + "epoch": 0.46325240129631906, + "grad_norm": 0.43355244398117065, + "learning_rate": 0.00011155739793705869, + "loss": 0.024, + "step": 19762 + }, + { + "epoch": 0.46327584287112406, + "grad_norm": 2.4551262855529785, + "learning_rate": 0.00011155008110942191, + "loss": 0.2215, + "step": 19763 + }, + { + "epoch": 0.4632992844459291, + "grad_norm": 0.6178876757621765, + "learning_rate": 0.0001115427642191141, + "loss": 0.0319, + "step": 19764 + }, + { + "epoch": 0.4633227260207341, + "grad_norm": 3.8961331844329834, + "learning_rate": 0.00011153544726617492, + "loss": 0.2732, + "step": 19765 + }, + { + "epoch": 0.4633461675955391, + "grad_norm": 2.529719591140747, + "learning_rate": 0.00011152813025064409, + "loss": 0.2411, + "step": 19766 + }, + { + "epoch": 0.4633696091703441, + "grad_norm": 3.3868374824523926, + "learning_rate": 0.00011152081317256129, + "loss": 0.414, + "step": 19767 + }, + { + "epoch": 0.4633930507451491, + "grad_norm": 3.0431957244873047, + "learning_rate": 0.00011151349603196626, + "loss": 0.2965, + "step": 19768 + }, + { + "epoch": 0.4634164923199541, + "grad_norm": 1.2852516174316406, + "learning_rate": 0.00011150617882889864, + "loss": 0.1689, + "step": 19769 + }, + { + "epoch": 0.46343993389475907, + "grad_norm": 2.7818238735198975, + "learning_rate": 0.00011149886156339818, + "loss": 0.2745, + "step": 19770 + }, + { + "epoch": 0.46346337546956407, + "grad_norm": 4.052012920379639, + "learning_rate": 0.00011149154423550458, + "loss": 0.4096, + "step": 19771 + }, + { + "epoch": 0.46348681704436906, + "grad_norm": 3.3377022743225098, + "learning_rate": 0.00011148422684525755, + "loss": 0.3916, + "step": 19772 + }, + { + "epoch": 0.46351025861917405, + "grad_norm": 2.857897996902466, + "learning_rate": 0.00011147690939269675, + "loss": 0.2333, + "step": 19773 + }, + { + "epoch": 0.46353370019397905, + "grad_norm": 1.039874792098999, + "learning_rate": 0.00011146959187786195, + "loss": 0.0946, + "step": 19774 + }, + { + "epoch": 0.46355714176878404, + "grad_norm": 3.242835521697998, + "learning_rate": 0.00011146227430079284, + "loss": 1.497, + "step": 19775 + }, + { + "epoch": 0.46358058334358904, + "grad_norm": 2.999746084213257, + "learning_rate": 0.00011145495666152903, + "loss": 0.248, + "step": 19776 + }, + { + "epoch": 0.46360402491839403, + "grad_norm": 3.240797758102417, + "learning_rate": 0.00011144763896011037, + "loss": 0.3214, + "step": 19777 + }, + { + "epoch": 0.463627466493199, + "grad_norm": 4.021653175354004, + "learning_rate": 0.00011144032119657647, + "loss": 0.3752, + "step": 19778 + }, + { + "epoch": 0.463650908068004, + "grad_norm": 3.096855640411377, + "learning_rate": 0.00011143300337096707, + "loss": 0.3161, + "step": 19779 + }, + { + "epoch": 0.463674349642809, + "grad_norm": 1.958057165145874, + "learning_rate": 0.00011142568548332189, + "loss": 1.4785, + "step": 19780 + }, + { + "epoch": 0.463697791217614, + "grad_norm": 3.390477180480957, + "learning_rate": 0.0001114183675336806, + "loss": 0.2079, + "step": 19781 + }, + { + "epoch": 0.463721232792419, + "grad_norm": 4.259123802185059, + "learning_rate": 0.00011141104952208292, + "loss": 0.4332, + "step": 19782 + }, + { + "epoch": 0.463744674367224, + "grad_norm": 2.3912997245788574, + "learning_rate": 0.00011140373144856857, + "loss": 1.4404, + "step": 19783 + }, + { + "epoch": 0.463768115942029, + "grad_norm": 3.673520803451538, + "learning_rate": 0.00011139641331317728, + "loss": 0.2116, + "step": 19784 + }, + { + "epoch": 0.463791557516834, + "grad_norm": 4.1827874183654785, + "learning_rate": 0.00011138909511594869, + "loss": 0.4494, + "step": 19785 + }, + { + "epoch": 0.463814999091639, + "grad_norm": 3.840082883834839, + "learning_rate": 0.00011138177685692256, + "loss": 0.4675, + "step": 19786 + }, + { + "epoch": 0.463838440666444, + "grad_norm": 0.7270359992980957, + "learning_rate": 0.00011137445853613863, + "loss": 0.0677, + "step": 19787 + }, + { + "epoch": 0.46386188224124897, + "grad_norm": 5.547219753265381, + "learning_rate": 0.00011136714015363652, + "loss": 0.355, + "step": 19788 + }, + { + "epoch": 0.46388532381605396, + "grad_norm": 1.824617624282837, + "learning_rate": 0.000111359821709456, + "loss": 0.1372, + "step": 19789 + }, + { + "epoch": 0.46390876539085896, + "grad_norm": 1.1757739782333374, + "learning_rate": 0.00011135250320363677, + "loss": 0.0505, + "step": 19790 + }, + { + "epoch": 0.46393220696566395, + "grad_norm": 3.4147279262542725, + "learning_rate": 0.00011134518463621857, + "loss": 1.6165, + "step": 19791 + }, + { + "epoch": 0.46395564854046895, + "grad_norm": 2.138549327850342, + "learning_rate": 0.00011133786600724104, + "loss": 0.1692, + "step": 19792 + }, + { + "epoch": 0.46397909011527394, + "grad_norm": 3.4818837642669678, + "learning_rate": 0.00011133054731674395, + "loss": 0.1961, + "step": 19793 + }, + { + "epoch": 0.46400253169007893, + "grad_norm": 2.7261571884155273, + "learning_rate": 0.00011132322856476696, + "loss": 0.242, + "step": 19794 + }, + { + "epoch": 0.46402597326488393, + "grad_norm": 3.1643030643463135, + "learning_rate": 0.00011131590975134983, + "loss": 0.3422, + "step": 19795 + }, + { + "epoch": 0.4640494148396889, + "grad_norm": 2.478963613510132, + "learning_rate": 0.00011130859087653227, + "loss": 0.2252, + "step": 19796 + }, + { + "epoch": 0.4640728564144939, + "grad_norm": 1.2808901071548462, + "learning_rate": 0.00011130127194035398, + "loss": 0.1113, + "step": 19797 + }, + { + "epoch": 0.4640962979892989, + "grad_norm": 5.579983711242676, + "learning_rate": 0.00011129395294285465, + "loss": 0.3033, + "step": 19798 + }, + { + "epoch": 0.4641197395641039, + "grad_norm": 3.297361135482788, + "learning_rate": 0.00011128663388407403, + "loss": 0.2011, + "step": 19799 + }, + { + "epoch": 0.4641431811389089, + "grad_norm": 2.073496103286743, + "learning_rate": 0.0001112793147640518, + "loss": 0.1137, + "step": 19800 + }, + { + "epoch": 0.4641666227137139, + "grad_norm": 4.373892307281494, + "learning_rate": 0.0001112719955828277, + "loss": 0.3207, + "step": 19801 + }, + { + "epoch": 0.4641900642885189, + "grad_norm": 2.6897096633911133, + "learning_rate": 0.00011126467634044143, + "loss": 0.2989, + "step": 19802 + }, + { + "epoch": 0.4642135058633239, + "grad_norm": 3.311314105987549, + "learning_rate": 0.00011125735703693272, + "loss": 0.4073, + "step": 19803 + }, + { + "epoch": 0.4642369474381289, + "grad_norm": 2.8364460468292236, + "learning_rate": 0.00011125003767234126, + "loss": 0.2892, + "step": 19804 + }, + { + "epoch": 0.46426038901293387, + "grad_norm": 0.9408908486366272, + "learning_rate": 0.0001112427182467068, + "loss": 0.1139, + "step": 19805 + }, + { + "epoch": 0.46428383058773887, + "grad_norm": 3.902559757232666, + "learning_rate": 0.000111235398760069, + "loss": 0.3287, + "step": 19806 + }, + { + "epoch": 0.46430727216254386, + "grad_norm": 0.7628236413002014, + "learning_rate": 0.00011122807921246767, + "loss": 0.0629, + "step": 19807 + }, + { + "epoch": 0.46433071373734885, + "grad_norm": 2.372077703475952, + "learning_rate": 0.0001112207596039424, + "loss": 0.2736, + "step": 19808 + }, + { + "epoch": 0.46435415531215385, + "grad_norm": 1.8619376420974731, + "learning_rate": 0.00011121343993453304, + "loss": 1.2484, + "step": 19809 + }, + { + "epoch": 0.46437759688695884, + "grad_norm": 7.026789665222168, + "learning_rate": 0.00011120612020427921, + "loss": 0.5662, + "step": 19810 + }, + { + "epoch": 0.46440103846176384, + "grad_norm": 2.3313913345336914, + "learning_rate": 0.00011119880041322064, + "loss": 1.7913, + "step": 19811 + }, + { + "epoch": 0.46442448003656883, + "grad_norm": 2.847877264022827, + "learning_rate": 0.00011119148056139706, + "loss": 0.3202, + "step": 19812 + }, + { + "epoch": 0.4644479216113738, + "grad_norm": 2.634896755218506, + "learning_rate": 0.0001111841606488482, + "loss": 0.1343, + "step": 19813 + }, + { + "epoch": 0.4644713631861788, + "grad_norm": 1.1729600429534912, + "learning_rate": 0.00011117684067561378, + "loss": 0.1206, + "step": 19814 + }, + { + "epoch": 0.46449480476098387, + "grad_norm": 3.531996011734009, + "learning_rate": 0.0001111695206417335, + "loss": 0.3561, + "step": 19815 + }, + { + "epoch": 0.46451824633578886, + "grad_norm": 2.3583931922912598, + "learning_rate": 0.00011116220054724711, + "loss": 0.1925, + "step": 19816 + }, + { + "epoch": 0.46454168791059386, + "grad_norm": 3.0660862922668457, + "learning_rate": 0.00011115488039219429, + "loss": 0.2809, + "step": 19817 + }, + { + "epoch": 0.46456512948539885, + "grad_norm": 0.9380584359169006, + "learning_rate": 0.00011114756017661474, + "loss": 0.0726, + "step": 19818 + }, + { + "epoch": 0.46458857106020385, + "grad_norm": 2.9852418899536133, + "learning_rate": 0.00011114023990054825, + "loss": 0.3724, + "step": 19819 + }, + { + "epoch": 0.46461201263500884, + "grad_norm": 0.7704975008964539, + "learning_rate": 0.00011113291956403453, + "loss": 0.0571, + "step": 19820 + }, + { + "epoch": 0.46463545420981384, + "grad_norm": 2.969684600830078, + "learning_rate": 0.00011112559916711323, + "loss": 0.2323, + "step": 19821 + }, + { + "epoch": 0.46465889578461883, + "grad_norm": 3.849357843399048, + "learning_rate": 0.00011111827870982414, + "loss": 0.4644, + "step": 19822 + }, + { + "epoch": 0.4646823373594238, + "grad_norm": 0.5986884832382202, + "learning_rate": 0.00011111095819220693, + "loss": 0.0387, + "step": 19823 + }, + { + "epoch": 0.4647057789342288, + "grad_norm": 3.741399049758911, + "learning_rate": 0.00011110363761430136, + "loss": 0.4637, + "step": 19824 + }, + { + "epoch": 0.4647292205090338, + "grad_norm": 1.1725784540176392, + "learning_rate": 0.00011109631697614717, + "loss": 0.0789, + "step": 19825 + }, + { + "epoch": 0.4647526620838388, + "grad_norm": 2.341175079345703, + "learning_rate": 0.00011108899627778404, + "loss": 0.3175, + "step": 19826 + }, + { + "epoch": 0.4647761036586438, + "grad_norm": 3.2178966999053955, + "learning_rate": 0.00011108167551925168, + "loss": 1.8844, + "step": 19827 + }, + { + "epoch": 0.4647995452334488, + "grad_norm": 1.1882497072219849, + "learning_rate": 0.00011107435470058986, + "loss": 0.109, + "step": 19828 + }, + { + "epoch": 0.4648229868082538, + "grad_norm": 3.386178970336914, + "learning_rate": 0.0001110670338218383, + "loss": 0.3083, + "step": 19829 + }, + { + "epoch": 0.4648464283830588, + "grad_norm": 2.1859283447265625, + "learning_rate": 0.00011105971288303663, + "loss": 0.2077, + "step": 19830 + }, + { + "epoch": 0.4648698699578638, + "grad_norm": 7.615441799163818, + "learning_rate": 0.00011105239188422472, + "loss": 0.3378, + "step": 19831 + }, + { + "epoch": 0.4648933115326688, + "grad_norm": 5.064464092254639, + "learning_rate": 0.0001110450708254422, + "loss": 0.6853, + "step": 19832 + }, + { + "epoch": 0.46491675310747377, + "grad_norm": 2.5852508544921875, + "learning_rate": 0.0001110377497067288, + "loss": 0.4003, + "step": 19833 + }, + { + "epoch": 0.46494019468227876, + "grad_norm": 4.744956970214844, + "learning_rate": 0.00011103042852812427, + "loss": 0.4128, + "step": 19834 + }, + { + "epoch": 0.46496363625708376, + "grad_norm": 2.603343963623047, + "learning_rate": 0.00011102310728966834, + "loss": 1.4335, + "step": 19835 + }, + { + "epoch": 0.46498707783188875, + "grad_norm": 2.1165075302124023, + "learning_rate": 0.00011101578599140067, + "loss": 1.6648, + "step": 19836 + }, + { + "epoch": 0.46501051940669375, + "grad_norm": 3.7233903408050537, + "learning_rate": 0.00011100846463336107, + "loss": 0.2248, + "step": 19837 + }, + { + "epoch": 0.46503396098149874, + "grad_norm": 2.9994139671325684, + "learning_rate": 0.00011100114321558925, + "loss": 1.432, + "step": 19838 + }, + { + "epoch": 0.46505740255630373, + "grad_norm": 0.9495536684989929, + "learning_rate": 0.00011099382173812492, + "loss": 0.0718, + "step": 19839 + }, + { + "epoch": 0.46508084413110873, + "grad_norm": 3.467021942138672, + "learning_rate": 0.00011098650020100775, + "loss": 0.153, + "step": 19840 + }, + { + "epoch": 0.4651042857059137, + "grad_norm": 1.723100185394287, + "learning_rate": 0.00011097917860427758, + "loss": 0.1583, + "step": 19841 + }, + { + "epoch": 0.4651277272807187, + "grad_norm": 1.2617449760437012, + "learning_rate": 0.00011097185694797403, + "loss": 0.6386, + "step": 19842 + }, + { + "epoch": 0.4651511688555237, + "grad_norm": 2.7980315685272217, + "learning_rate": 0.00011096453523213688, + "loss": 0.2144, + "step": 19843 + }, + { + "epoch": 0.4651746104303287, + "grad_norm": 6.218197345733643, + "learning_rate": 0.00011095721345680588, + "loss": 0.5055, + "step": 19844 + }, + { + "epoch": 0.4651980520051337, + "grad_norm": 4.181087970733643, + "learning_rate": 0.00011094989162202075, + "loss": 0.2259, + "step": 19845 + }, + { + "epoch": 0.4652214935799387, + "grad_norm": 3.891845226287842, + "learning_rate": 0.00011094256972782115, + "loss": 0.1642, + "step": 19846 + }, + { + "epoch": 0.4652449351547437, + "grad_norm": 2.1267924308776855, + "learning_rate": 0.00011093524777424689, + "loss": 0.2289, + "step": 19847 + }, + { + "epoch": 0.4652683767295487, + "grad_norm": 4.56112003326416, + "learning_rate": 0.00011092792576133766, + "loss": 0.3985, + "step": 19848 + }, + { + "epoch": 0.4652918183043537, + "grad_norm": 2.917361259460449, + "learning_rate": 0.00011092060368913318, + "loss": 0.28, + "step": 19849 + }, + { + "epoch": 0.46531525987915867, + "grad_norm": 4.1539506912231445, + "learning_rate": 0.00011091328155767323, + "loss": 0.4209, + "step": 19850 + }, + { + "epoch": 0.46533870145396367, + "grad_norm": 3.8906800746917725, + "learning_rate": 0.00011090595936699747, + "loss": 0.4268, + "step": 19851 + }, + { + "epoch": 0.46536214302876866, + "grad_norm": 2.4807701110839844, + "learning_rate": 0.0001108986371171457, + "loss": 0.1658, + "step": 19852 + }, + { + "epoch": 0.46538558460357365, + "grad_norm": 3.837885618209839, + "learning_rate": 0.00011089131480815762, + "loss": 0.4101, + "step": 19853 + }, + { + "epoch": 0.46540902617837865, + "grad_norm": 3.0561113357543945, + "learning_rate": 0.00011088399244007295, + "loss": 0.2515, + "step": 19854 + }, + { + "epoch": 0.46543246775318364, + "grad_norm": 1.503147840499878, + "learning_rate": 0.00011087667001293142, + "loss": 0.1319, + "step": 19855 + }, + { + "epoch": 0.46545590932798864, + "grad_norm": 0.64434814453125, + "learning_rate": 0.00011086934752677277, + "loss": 0.122, + "step": 19856 + }, + { + "epoch": 0.46547935090279363, + "grad_norm": 1.569795846939087, + "learning_rate": 0.00011086202498163676, + "loss": 0.0958, + "step": 19857 + }, + { + "epoch": 0.4655027924775986, + "grad_norm": 0.7044201493263245, + "learning_rate": 0.0001108547023775631, + "loss": 0.0399, + "step": 19858 + }, + { + "epoch": 0.4655262340524036, + "grad_norm": 1.0593045949935913, + "learning_rate": 0.00011084737971459147, + "loss": 0.1577, + "step": 19859 + }, + { + "epoch": 0.4655496756272086, + "grad_norm": 3.9260482788085938, + "learning_rate": 0.0001108400569927617, + "loss": 0.6202, + "step": 19860 + }, + { + "epoch": 0.4655731172020136, + "grad_norm": 3.5966715812683105, + "learning_rate": 0.00011083273421211347, + "loss": 0.4283, + "step": 19861 + }, + { + "epoch": 0.4655965587768186, + "grad_norm": 1.9963661432266235, + "learning_rate": 0.0001108254113726865, + "loss": 0.6075, + "step": 19862 + }, + { + "epoch": 0.4656200003516236, + "grad_norm": 2.425306797027588, + "learning_rate": 0.00011081808847452055, + "loss": 0.1734, + "step": 19863 + }, + { + "epoch": 0.4656434419264286, + "grad_norm": 2.9521265029907227, + "learning_rate": 0.00011081076551765537, + "loss": 0.3222, + "step": 19864 + }, + { + "epoch": 0.4656668835012336, + "grad_norm": 1.9126973152160645, + "learning_rate": 0.00011080344250213064, + "loss": 0.1998, + "step": 19865 + }, + { + "epoch": 0.46569032507603864, + "grad_norm": 1.7198220491409302, + "learning_rate": 0.00011079611942798615, + "loss": 1.6372, + "step": 19866 + }, + { + "epoch": 0.46571376665084363, + "grad_norm": 3.6099438667297363, + "learning_rate": 0.00011078879629526159, + "loss": 0.4281, + "step": 19867 + }, + { + "epoch": 0.4657372082256486, + "grad_norm": 1.9023436307907104, + "learning_rate": 0.00011078147310399672, + "loss": 0.1584, + "step": 19868 + }, + { + "epoch": 0.4657606498004536, + "grad_norm": 3.6704375743865967, + "learning_rate": 0.00011077414985423129, + "loss": 0.476, + "step": 19869 + }, + { + "epoch": 0.4657840913752586, + "grad_norm": 1.0585676431655884, + "learning_rate": 0.00011076682654600504, + "loss": 0.1067, + "step": 19870 + }, + { + "epoch": 0.4658075329500636, + "grad_norm": 1.2065879106521606, + "learning_rate": 0.00011075950317935764, + "loss": 0.1125, + "step": 19871 + }, + { + "epoch": 0.4658309745248686, + "grad_norm": 0.9147006273269653, + "learning_rate": 0.00011075217975432892, + "loss": 0.0553, + "step": 19872 + }, + { + "epoch": 0.4658544160996736, + "grad_norm": 3.87418532371521, + "learning_rate": 0.00011074485627095853, + "loss": 0.1941, + "step": 19873 + }, + { + "epoch": 0.4658778576744786, + "grad_norm": 2.809485673904419, + "learning_rate": 0.00011073753272928627, + "loss": 0.2734, + "step": 19874 + }, + { + "epoch": 0.4659012992492836, + "grad_norm": 3.5277929306030273, + "learning_rate": 0.00011073020912935183, + "loss": 0.3728, + "step": 19875 + }, + { + "epoch": 0.4659247408240886, + "grad_norm": 1.8636125326156616, + "learning_rate": 0.000110722885471195, + "loss": 0.1971, + "step": 19876 + }, + { + "epoch": 0.4659481823988936, + "grad_norm": 1.5763126611709595, + "learning_rate": 0.00011071556175485551, + "loss": 0.1293, + "step": 19877 + }, + { + "epoch": 0.46597162397369857, + "grad_norm": 3.9933953285217285, + "learning_rate": 0.00011070823798037304, + "loss": 0.4201, + "step": 19878 + }, + { + "epoch": 0.46599506554850356, + "grad_norm": 2.067686080932617, + "learning_rate": 0.00011070091414778738, + "loss": 0.223, + "step": 19879 + }, + { + "epoch": 0.46601850712330856, + "grad_norm": 3.1983706951141357, + "learning_rate": 0.00011069359025713828, + "loss": 0.3332, + "step": 19880 + }, + { + "epoch": 0.46604194869811355, + "grad_norm": 5.331818580627441, + "learning_rate": 0.00011068626630846541, + "loss": 0.6266, + "step": 19881 + }, + { + "epoch": 0.46606539027291854, + "grad_norm": 3.396096706390381, + "learning_rate": 0.00011067894230180862, + "loss": 0.3514, + "step": 19882 + }, + { + "epoch": 0.46608883184772354, + "grad_norm": 1.564538836479187, + "learning_rate": 0.00011067161823720755, + "loss": 0.1619, + "step": 19883 + }, + { + "epoch": 0.46611227342252853, + "grad_norm": 2.825568914413452, + "learning_rate": 0.00011066429411470198, + "loss": 0.1912, + "step": 19884 + }, + { + "epoch": 0.46613571499733353, + "grad_norm": 1.6270800828933716, + "learning_rate": 0.00011065696993433166, + "loss": 0.221, + "step": 19885 + }, + { + "epoch": 0.4661591565721385, + "grad_norm": 2.3778982162475586, + "learning_rate": 0.00011064964569613633, + "loss": 0.1964, + "step": 19886 + }, + { + "epoch": 0.4661825981469435, + "grad_norm": 4.596756458282471, + "learning_rate": 0.00011064232140015569, + "loss": 0.4037, + "step": 19887 + }, + { + "epoch": 0.4662060397217485, + "grad_norm": 0.8861894011497498, + "learning_rate": 0.00011063499704642956, + "loss": 0.1035, + "step": 19888 + }, + { + "epoch": 0.4662294812965535, + "grad_norm": 2.6426095962524414, + "learning_rate": 0.00011062767263499762, + "loss": 0.325, + "step": 19889 + }, + { + "epoch": 0.4662529228713585, + "grad_norm": 0.5169786214828491, + "learning_rate": 0.00011062034816589962, + "loss": 0.0464, + "step": 19890 + }, + { + "epoch": 0.4662763644461635, + "grad_norm": 4.835724353790283, + "learning_rate": 0.00011061302363917531, + "loss": 0.7399, + "step": 19891 + }, + { + "epoch": 0.4662998060209685, + "grad_norm": 4.357682704925537, + "learning_rate": 0.00011060569905486444, + "loss": 0.5051, + "step": 19892 + }, + { + "epoch": 0.4663232475957735, + "grad_norm": 4.213548183441162, + "learning_rate": 0.00011059837441300678, + "loss": 0.4537, + "step": 19893 + }, + { + "epoch": 0.4663466891705785, + "grad_norm": 1.9728052616119385, + "learning_rate": 0.000110591049713642, + "loss": 0.1362, + "step": 19894 + }, + { + "epoch": 0.46637013074538347, + "grad_norm": 1.0017117261886597, + "learning_rate": 0.00011058372495680991, + "loss": 0.1289, + "step": 19895 + }, + { + "epoch": 0.46639357232018847, + "grad_norm": 2.1942827701568604, + "learning_rate": 0.00011057640014255019, + "loss": 0.2706, + "step": 19896 + }, + { + "epoch": 0.46641701389499346, + "grad_norm": 3.0993459224700928, + "learning_rate": 0.00011056907527090266, + "loss": 0.3076, + "step": 19897 + }, + { + "epoch": 0.46644045546979845, + "grad_norm": 3.79318904876709, + "learning_rate": 0.00011056175034190703, + "loss": 0.2281, + "step": 19898 + }, + { + "epoch": 0.46646389704460345, + "grad_norm": 2.315995216369629, + "learning_rate": 0.00011055442535560305, + "loss": 1.5569, + "step": 19899 + }, + { + "epoch": 0.46648733861940844, + "grad_norm": 1.9727511405944824, + "learning_rate": 0.00011054710031203046, + "loss": 0.1454, + "step": 19900 + }, + { + "epoch": 0.46651078019421344, + "grad_norm": 2.6834776401519775, + "learning_rate": 0.000110539775211229, + "loss": 1.6837, + "step": 19901 + }, + { + "epoch": 0.46653422176901843, + "grad_norm": 0.9122545719146729, + "learning_rate": 0.00011053245005323844, + "loss": 0.1283, + "step": 19902 + }, + { + "epoch": 0.4665576633438234, + "grad_norm": 0.2122945338487625, + "learning_rate": 0.00011052512483809847, + "loss": 0.0073, + "step": 19903 + }, + { + "epoch": 0.4665811049186284, + "grad_norm": 6.676165580749512, + "learning_rate": 0.00011051779956584891, + "loss": 0.3237, + "step": 19904 + }, + { + "epoch": 0.4666045464934334, + "grad_norm": 1.5117452144622803, + "learning_rate": 0.00011051047423652947, + "loss": 0.1449, + "step": 19905 + }, + { + "epoch": 0.4666279880682384, + "grad_norm": 1.3079265356063843, + "learning_rate": 0.0001105031488501799, + "loss": 0.2278, + "step": 19906 + }, + { + "epoch": 0.4666514296430434, + "grad_norm": 0.8379679322242737, + "learning_rate": 0.00011049582340683992, + "loss": 0.0645, + "step": 19907 + }, + { + "epoch": 0.4666748712178484, + "grad_norm": 2.733335018157959, + "learning_rate": 0.00011048849790654935, + "loss": 0.2561, + "step": 19908 + }, + { + "epoch": 0.4666983127926534, + "grad_norm": 0.9920181632041931, + "learning_rate": 0.00011048117234934788, + "loss": 0.1101, + "step": 19909 + }, + { + "epoch": 0.4667217543674584, + "grad_norm": 1.2130799293518066, + "learning_rate": 0.00011047384673527525, + "loss": 0.083, + "step": 19910 + }, + { + "epoch": 0.4667451959422634, + "grad_norm": 4.304091930389404, + "learning_rate": 0.00011046652106437128, + "loss": 0.2335, + "step": 19911 + }, + { + "epoch": 0.4667686375170684, + "grad_norm": 4.714005470275879, + "learning_rate": 0.00011045919533667566, + "loss": 0.5883, + "step": 19912 + }, + { + "epoch": 0.46679207909187337, + "grad_norm": 3.924278497695923, + "learning_rate": 0.00011045186955222813, + "loss": 0.4205, + "step": 19913 + }, + { + "epoch": 0.46681552066667836, + "grad_norm": 3.907667636871338, + "learning_rate": 0.00011044454371106847, + "loss": 0.6296, + "step": 19914 + }, + { + "epoch": 0.46683896224148336, + "grad_norm": 1.0007516145706177, + "learning_rate": 0.00011043721781323641, + "loss": 0.0709, + "step": 19915 + }, + { + "epoch": 0.46686240381628835, + "grad_norm": 3.0813164710998535, + "learning_rate": 0.0001104298918587717, + "loss": 0.2104, + "step": 19916 + }, + { + "epoch": 0.46688584539109335, + "grad_norm": 3.4426662921905518, + "learning_rate": 0.00011042256584771415, + "loss": 0.2487, + "step": 19917 + }, + { + "epoch": 0.4669092869658984, + "grad_norm": 3.5995936393737793, + "learning_rate": 0.00011041523978010347, + "loss": 0.3293, + "step": 19918 + }, + { + "epoch": 0.4669327285407034, + "grad_norm": 3.2672150135040283, + "learning_rate": 0.00011040791365597935, + "loss": 0.2277, + "step": 19919 + }, + { + "epoch": 0.4669561701155084, + "grad_norm": 2.3422207832336426, + "learning_rate": 0.00011040058747538164, + "loss": 0.1662, + "step": 19920 + }, + { + "epoch": 0.4669796116903134, + "grad_norm": 2.4280998706817627, + "learning_rate": 0.00011039326123835006, + "loss": 0.2602, + "step": 19921 + }, + { + "epoch": 0.4670030532651184, + "grad_norm": 2.7118706703186035, + "learning_rate": 0.00011038593494492429, + "loss": 0.2211, + "step": 19922 + }, + { + "epoch": 0.46702649483992337, + "grad_norm": 3.244609832763672, + "learning_rate": 0.00011037860859514419, + "loss": 0.2923, + "step": 19923 + }, + { + "epoch": 0.46704993641472836, + "grad_norm": 3.408036947250366, + "learning_rate": 0.00011037128218904949, + "loss": 0.3592, + "step": 19924 + }, + { + "epoch": 0.46707337798953336, + "grad_norm": 2.031907081604004, + "learning_rate": 0.00011036395572667989, + "loss": 0.1919, + "step": 19925 + }, + { + "epoch": 0.46709681956433835, + "grad_norm": 4.376325607299805, + "learning_rate": 0.00011035662920807517, + "loss": 1.8448, + "step": 19926 + }, + { + "epoch": 0.46712026113914334, + "grad_norm": 3.177293062210083, + "learning_rate": 0.00011034930263327513, + "loss": 0.2729, + "step": 19927 + }, + { + "epoch": 0.46714370271394834, + "grad_norm": 4.052722930908203, + "learning_rate": 0.00011034197600231943, + "loss": 0.2594, + "step": 19928 + }, + { + "epoch": 0.46716714428875333, + "grad_norm": 1.111509084701538, + "learning_rate": 0.00011033464931524788, + "loss": 0.0927, + "step": 19929 + }, + { + "epoch": 0.4671905858635583, + "grad_norm": 3.7099335193634033, + "learning_rate": 0.00011032732257210026, + "loss": 0.3773, + "step": 19930 + }, + { + "epoch": 0.4672140274383633, + "grad_norm": 1.5756759643554688, + "learning_rate": 0.0001103199957729163, + "loss": 0.2217, + "step": 19931 + }, + { + "epoch": 0.4672374690131683, + "grad_norm": 2.6187095642089844, + "learning_rate": 0.00011031266891773573, + "loss": 0.2364, + "step": 19932 + }, + { + "epoch": 0.4672609105879733, + "grad_norm": 2.8167996406555176, + "learning_rate": 0.00011030534200659834, + "loss": 0.2216, + "step": 19933 + }, + { + "epoch": 0.4672843521627783, + "grad_norm": 3.70835280418396, + "learning_rate": 0.00011029801503954386, + "loss": 0.3646, + "step": 19934 + }, + { + "epoch": 0.4673077937375833, + "grad_norm": 2.588440179824829, + "learning_rate": 0.00011029068801661208, + "loss": 0.2759, + "step": 19935 + }, + { + "epoch": 0.4673312353123883, + "grad_norm": 3.4864606857299805, + "learning_rate": 0.00011028336093784273, + "loss": 0.391, + "step": 19936 + }, + { + "epoch": 0.4673546768871933, + "grad_norm": 3.904155731201172, + "learning_rate": 0.00011027603380327561, + "loss": 0.4061, + "step": 19937 + }, + { + "epoch": 0.4673781184619983, + "grad_norm": 0.7298229336738586, + "learning_rate": 0.00011026870661295038, + "loss": 0.0856, + "step": 19938 + }, + { + "epoch": 0.4674015600368033, + "grad_norm": 1.8590519428253174, + "learning_rate": 0.0001102613793669069, + "loss": 0.0994, + "step": 19939 + }, + { + "epoch": 0.46742500161160827, + "grad_norm": 2.6389942169189453, + "learning_rate": 0.00011025405206518486, + "loss": 0.2392, + "step": 19940 + }, + { + "epoch": 0.46744844318641326, + "grad_norm": 1.6921625137329102, + "learning_rate": 0.00011024672470782404, + "loss": 0.1186, + "step": 19941 + }, + { + "epoch": 0.46747188476121826, + "grad_norm": 5.186007976531982, + "learning_rate": 0.00011023939729486423, + "loss": 0.2772, + "step": 19942 + }, + { + "epoch": 0.46749532633602325, + "grad_norm": 4.237432479858398, + "learning_rate": 0.00011023206982634517, + "loss": 0.2938, + "step": 19943 + }, + { + "epoch": 0.46751876791082825, + "grad_norm": 2.7839434146881104, + "learning_rate": 0.0001102247423023066, + "loss": 0.3039, + "step": 19944 + }, + { + "epoch": 0.46754220948563324, + "grad_norm": 2.8331334590911865, + "learning_rate": 0.0001102174147227883, + "loss": 0.4071, + "step": 19945 + }, + { + "epoch": 0.46756565106043824, + "grad_norm": 1.40580153465271, + "learning_rate": 0.00011021008708782998, + "loss": 0.1009, + "step": 19946 + }, + { + "epoch": 0.46758909263524323, + "grad_norm": 0.6301449537277222, + "learning_rate": 0.00011020275939747149, + "loss": 0.0412, + "step": 19947 + }, + { + "epoch": 0.4676125342100482, + "grad_norm": 3.329667329788208, + "learning_rate": 0.0001101954316517525, + "loss": 0.2135, + "step": 19948 + }, + { + "epoch": 0.4676359757848532, + "grad_norm": 2.8400416374206543, + "learning_rate": 0.00011018810385071286, + "loss": 0.2458, + "step": 19949 + }, + { + "epoch": 0.4676594173596582, + "grad_norm": 5.2042975425720215, + "learning_rate": 0.00011018077599439224, + "loss": 0.2638, + "step": 19950 + }, + { + "epoch": 0.4676828589344632, + "grad_norm": 2.4748315811157227, + "learning_rate": 0.00011017344808283046, + "loss": 0.2981, + "step": 19951 + }, + { + "epoch": 0.4677063005092682, + "grad_norm": 4.055984973907471, + "learning_rate": 0.00011016612011606724, + "loss": 0.3402, + "step": 19952 + }, + { + "epoch": 0.4677297420840732, + "grad_norm": 3.0214807987213135, + "learning_rate": 0.00011015879209414239, + "loss": 0.2525, + "step": 19953 + }, + { + "epoch": 0.4677531836588782, + "grad_norm": 3.1170384883880615, + "learning_rate": 0.00011015146401709564, + "loss": 0.2884, + "step": 19954 + }, + { + "epoch": 0.4677766252336832, + "grad_norm": 2.8189172744750977, + "learning_rate": 0.00011014413588496677, + "loss": 0.305, + "step": 19955 + }, + { + "epoch": 0.4678000668084882, + "grad_norm": 3.994778633117676, + "learning_rate": 0.00011013680769779556, + "loss": 0.4042, + "step": 19956 + }, + { + "epoch": 0.4678235083832932, + "grad_norm": 5.243259906768799, + "learning_rate": 0.00011012947945562169, + "loss": 0.5041, + "step": 19957 + }, + { + "epoch": 0.46784694995809817, + "grad_norm": 1.457922101020813, + "learning_rate": 0.000110122151158485, + "loss": 0.139, + "step": 19958 + }, + { + "epoch": 0.46787039153290316, + "grad_norm": 5.749773979187012, + "learning_rate": 0.00011011482280642524, + "loss": 0.4118, + "step": 19959 + }, + { + "epoch": 0.46789383310770816, + "grad_norm": 0.9428538680076599, + "learning_rate": 0.00011010749439948216, + "loss": 0.0508, + "step": 19960 + }, + { + "epoch": 0.46791727468251315, + "grad_norm": 3.4080255031585693, + "learning_rate": 0.00011010016593769554, + "loss": 0.2862, + "step": 19961 + }, + { + "epoch": 0.46794071625731815, + "grad_norm": 1.9714903831481934, + "learning_rate": 0.00011009283742110515, + "loss": 0.1453, + "step": 19962 + }, + { + "epoch": 0.46796415783212314, + "grad_norm": 2.8311588764190674, + "learning_rate": 0.00011008550884975071, + "loss": 0.1652, + "step": 19963 + }, + { + "epoch": 0.46798759940692813, + "grad_norm": 4.365966320037842, + "learning_rate": 0.00011007818022367202, + "loss": 0.3239, + "step": 19964 + }, + { + "epoch": 0.46801104098173313, + "grad_norm": 2.5598866939544678, + "learning_rate": 0.00011007085154290885, + "loss": 1.7373, + "step": 19965 + }, + { + "epoch": 0.4680344825565381, + "grad_norm": 5.030275821685791, + "learning_rate": 0.00011006352280750098, + "loss": 0.4687, + "step": 19966 + }, + { + "epoch": 0.4680579241313431, + "grad_norm": 1.5883114337921143, + "learning_rate": 0.0001100561940174881, + "loss": 0.2136, + "step": 19967 + }, + { + "epoch": 0.4680813657061481, + "grad_norm": 2.9241530895233154, + "learning_rate": 0.00011004886517291008, + "loss": 0.351, + "step": 19968 + }, + { + "epoch": 0.46810480728095316, + "grad_norm": 2.415997266769409, + "learning_rate": 0.00011004153627380662, + "loss": 1.0025, + "step": 19969 + }, + { + "epoch": 0.46812824885575816, + "grad_norm": 2.8649022579193115, + "learning_rate": 0.00011003420732021745, + "loss": 0.2623, + "step": 19970 + }, + { + "epoch": 0.46815169043056315, + "grad_norm": 4.546813488006592, + "learning_rate": 0.00011002687831218246, + "loss": 0.5158, + "step": 19971 + }, + { + "epoch": 0.46817513200536814, + "grad_norm": 2.903742790222168, + "learning_rate": 0.00011001954924974132, + "loss": 0.4022, + "step": 19972 + }, + { + "epoch": 0.46819857358017314, + "grad_norm": 1.0041884183883667, + "learning_rate": 0.00011001222013293382, + "loss": 0.1138, + "step": 19973 + }, + { + "epoch": 0.46822201515497813, + "grad_norm": 1.3391342163085938, + "learning_rate": 0.00011000489096179974, + "loss": 0.0816, + "step": 19974 + }, + { + "epoch": 0.4682454567297831, + "grad_norm": 0.9563319087028503, + "learning_rate": 0.00010999756173637885, + "loss": 0.1406, + "step": 19975 + }, + { + "epoch": 0.4682688983045881, + "grad_norm": 1.1645196676254272, + "learning_rate": 0.00010999023245671087, + "loss": 0.1126, + "step": 19976 + }, + { + "epoch": 0.4682923398793931, + "grad_norm": 2.4462790489196777, + "learning_rate": 0.00010998290312283564, + "loss": 0.1727, + "step": 19977 + }, + { + "epoch": 0.4683157814541981, + "grad_norm": 3.170938491821289, + "learning_rate": 0.00010997557373479289, + "loss": 0.4287, + "step": 19978 + }, + { + "epoch": 0.4683392230290031, + "grad_norm": 1.0972027778625488, + "learning_rate": 0.00010996824429262237, + "loss": 0.1182, + "step": 19979 + }, + { + "epoch": 0.4683626646038081, + "grad_norm": 2.949780225753784, + "learning_rate": 0.00010996091479636393, + "loss": 0.1334, + "step": 19980 + }, + { + "epoch": 0.4683861061786131, + "grad_norm": 2.314483880996704, + "learning_rate": 0.00010995358524605726, + "loss": 0.1675, + "step": 19981 + }, + { + "epoch": 0.4684095477534181, + "grad_norm": 2.599813938140869, + "learning_rate": 0.00010994625564174214, + "loss": 0.3545, + "step": 19982 + }, + { + "epoch": 0.4684329893282231, + "grad_norm": 2.778559923171997, + "learning_rate": 0.00010993892598345836, + "loss": 0.2903, + "step": 19983 + }, + { + "epoch": 0.4684564309030281, + "grad_norm": 4.550291061401367, + "learning_rate": 0.00010993159627124572, + "loss": 0.5307, + "step": 19984 + }, + { + "epoch": 0.46847987247783307, + "grad_norm": 4.933663368225098, + "learning_rate": 0.00010992426650514394, + "loss": 0.4319, + "step": 19985 + }, + { + "epoch": 0.46850331405263806, + "grad_norm": 4.021340370178223, + "learning_rate": 0.00010991693668519278, + "loss": 0.5021, + "step": 19986 + }, + { + "epoch": 0.46852675562744306, + "grad_norm": 2.977982759475708, + "learning_rate": 0.0001099096068114321, + "loss": 0.2667, + "step": 19987 + }, + { + "epoch": 0.46855019720224805, + "grad_norm": 17.950199127197266, + "learning_rate": 0.0001099022768839016, + "loss": 0.6134, + "step": 19988 + }, + { + "epoch": 0.46857363877705305, + "grad_norm": 2.075904369354248, + "learning_rate": 0.00010989494690264102, + "loss": 0.2677, + "step": 19989 + }, + { + "epoch": 0.46859708035185804, + "grad_norm": 2.6118967533111572, + "learning_rate": 0.00010988761686769023, + "loss": 0.1986, + "step": 19990 + }, + { + "epoch": 0.46862052192666304, + "grad_norm": 2.606020212173462, + "learning_rate": 0.00010988028677908895, + "loss": 0.1391, + "step": 19991 + }, + { + "epoch": 0.46864396350146803, + "grad_norm": 2.639838695526123, + "learning_rate": 0.00010987295663687691, + "loss": 0.2269, + "step": 19992 + }, + { + "epoch": 0.468667405076273, + "grad_norm": 1.8107093572616577, + "learning_rate": 0.00010986562644109398, + "loss": 0.1317, + "step": 19993 + }, + { + "epoch": 0.468690846651078, + "grad_norm": 2.4637393951416016, + "learning_rate": 0.00010985829619177987, + "loss": 0.1562, + "step": 19994 + }, + { + "epoch": 0.468714288225883, + "grad_norm": 5.263649940490723, + "learning_rate": 0.00010985096588897437, + "loss": 0.3104, + "step": 19995 + }, + { + "epoch": 0.468737729800688, + "grad_norm": 2.319521903991699, + "learning_rate": 0.00010984363553271722, + "loss": 0.3698, + "step": 19996 + }, + { + "epoch": 0.468761171375493, + "grad_norm": 1.4675077199935913, + "learning_rate": 0.00010983630512304828, + "loss": 0.2004, + "step": 19997 + }, + { + "epoch": 0.468784612950298, + "grad_norm": 2.885878086090088, + "learning_rate": 0.00010982897466000724, + "loss": 0.2623, + "step": 19998 + }, + { + "epoch": 0.468808054525103, + "grad_norm": 2.0996830463409424, + "learning_rate": 0.00010982164414363389, + "loss": 0.2163, + "step": 19999 + }, + { + "epoch": 0.468831496099908, + "grad_norm": 4.040693283081055, + "learning_rate": 0.00010981431357396808, + "loss": 0.4426, + "step": 20000 + }, + { + "epoch": 0.468854937674713, + "grad_norm": 0.3472028970718384, + "learning_rate": 0.00010980698295104948, + "loss": 0.037, + "step": 20001 + }, + { + "epoch": 0.468878379249518, + "grad_norm": 2.90608549118042, + "learning_rate": 0.00010979965227491791, + "loss": 1.5933, + "step": 20002 + }, + { + "epoch": 0.46890182082432297, + "grad_norm": 1.0854074954986572, + "learning_rate": 0.00010979232154561318, + "loss": 0.1157, + "step": 20003 + }, + { + "epoch": 0.46892526239912796, + "grad_norm": 3.535964012145996, + "learning_rate": 0.00010978499076317504, + "loss": 0.3783, + "step": 20004 + }, + { + "epoch": 0.46894870397393296, + "grad_norm": 1.5553334951400757, + "learning_rate": 0.00010977765992764324, + "loss": 0.1266, + "step": 20005 + }, + { + "epoch": 0.46897214554873795, + "grad_norm": 5.426249027252197, + "learning_rate": 0.0001097703290390576, + "loss": 0.3382, + "step": 20006 + }, + { + "epoch": 0.46899558712354295, + "grad_norm": 3.313338041305542, + "learning_rate": 0.00010976299809745787, + "loss": 0.3213, + "step": 20007 + }, + { + "epoch": 0.46901902869834794, + "grad_norm": 3.743353843688965, + "learning_rate": 0.00010975566710288383, + "loss": 0.3681, + "step": 20008 + }, + { + "epoch": 0.46904247027315293, + "grad_norm": 2.5135271549224854, + "learning_rate": 0.0001097483360553753, + "loss": 1.8909, + "step": 20009 + }, + { + "epoch": 0.46906591184795793, + "grad_norm": 2.6105854511260986, + "learning_rate": 0.000109741004954972, + "loss": 0.2954, + "step": 20010 + }, + { + "epoch": 0.4690893534227629, + "grad_norm": 2.3437652587890625, + "learning_rate": 0.00010973367380171371, + "loss": 1.5033, + "step": 20011 + }, + { + "epoch": 0.4691127949975679, + "grad_norm": 4.277329444885254, + "learning_rate": 0.00010972634259564028, + "loss": 0.5519, + "step": 20012 + }, + { + "epoch": 0.4691362365723729, + "grad_norm": 0.7102829813957214, + "learning_rate": 0.00010971901133679141, + "loss": 0.0566, + "step": 20013 + }, + { + "epoch": 0.4691596781471779, + "grad_norm": 2.9239256381988525, + "learning_rate": 0.00010971168002520692, + "loss": 0.2424, + "step": 20014 + }, + { + "epoch": 0.4691831197219829, + "grad_norm": 3.0653340816497803, + "learning_rate": 0.00010970434866092658, + "loss": 0.3611, + "step": 20015 + }, + { + "epoch": 0.4692065612967879, + "grad_norm": 2.7365164756774902, + "learning_rate": 0.00010969701724399017, + "loss": 1.2653, + "step": 20016 + }, + { + "epoch": 0.4692300028715929, + "grad_norm": 4.049787521362305, + "learning_rate": 0.00010968968577443748, + "loss": 0.4792, + "step": 20017 + }, + { + "epoch": 0.4692534444463979, + "grad_norm": 2.751810312271118, + "learning_rate": 0.00010968235425230829, + "loss": 0.1037, + "step": 20018 + }, + { + "epoch": 0.4692768860212029, + "grad_norm": 2.514198064804077, + "learning_rate": 0.00010967502267764234, + "loss": 0.3284, + "step": 20019 + }, + { + "epoch": 0.4693003275960079, + "grad_norm": 2.8993332386016846, + "learning_rate": 0.00010966769105047947, + "loss": 0.3482, + "step": 20020 + }, + { + "epoch": 0.4693237691708129, + "grad_norm": 1.4578293561935425, + "learning_rate": 0.00010966035937085944, + "loss": 0.1151, + "step": 20021 + }, + { + "epoch": 0.4693472107456179, + "grad_norm": 3.0967586040496826, + "learning_rate": 0.00010965302763882202, + "loss": 0.1791, + "step": 20022 + }, + { + "epoch": 0.4693706523204229, + "grad_norm": 3.6861870288848877, + "learning_rate": 0.000109645695854407, + "loss": 0.3362, + "step": 20023 + }, + { + "epoch": 0.4693940938952279, + "grad_norm": 0.3371734619140625, + "learning_rate": 0.00010963836401765418, + "loss": 0.035, + "step": 20024 + }, + { + "epoch": 0.4694175354700329, + "grad_norm": 4.294587135314941, + "learning_rate": 0.00010963103212860328, + "loss": 0.4122, + "step": 20025 + }, + { + "epoch": 0.4694409770448379, + "grad_norm": 2.6594605445861816, + "learning_rate": 0.00010962370018729418, + "loss": 0.1716, + "step": 20026 + }, + { + "epoch": 0.4694644186196429, + "grad_norm": 2.708810329437256, + "learning_rate": 0.00010961636819376657, + "loss": 1.8561, + "step": 20027 + }, + { + "epoch": 0.4694878601944479, + "grad_norm": 2.959118127822876, + "learning_rate": 0.00010960903614806031, + "loss": 1.6137, + "step": 20028 + }, + { + "epoch": 0.4695113017692529, + "grad_norm": 2.5140960216522217, + "learning_rate": 0.00010960170405021515, + "loss": 1.7215, + "step": 20029 + }, + { + "epoch": 0.46953474334405787, + "grad_norm": 4.805915355682373, + "learning_rate": 0.00010959437190027084, + "loss": 0.5102, + "step": 20030 + }, + { + "epoch": 0.46955818491886286, + "grad_norm": 1.3728504180908203, + "learning_rate": 0.00010958703969826722, + "loss": 0.0857, + "step": 20031 + }, + { + "epoch": 0.46958162649366786, + "grad_norm": 3.7795634269714355, + "learning_rate": 0.00010957970744424406, + "loss": 0.5403, + "step": 20032 + }, + { + "epoch": 0.46960506806847285, + "grad_norm": 2.9851648807525635, + "learning_rate": 0.00010957237513824112, + "loss": 0.3505, + "step": 20033 + }, + { + "epoch": 0.46962850964327785, + "grad_norm": 2.9491381645202637, + "learning_rate": 0.00010956504278029822, + "loss": 0.3211, + "step": 20034 + }, + { + "epoch": 0.46965195121808284, + "grad_norm": 6.236459255218506, + "learning_rate": 0.00010955771037045515, + "loss": 0.5117, + "step": 20035 + }, + { + "epoch": 0.46967539279288784, + "grad_norm": 3.3404219150543213, + "learning_rate": 0.00010955037790875162, + "loss": 0.2417, + "step": 20036 + }, + { + "epoch": 0.46969883436769283, + "grad_norm": 3.177351713180542, + "learning_rate": 0.00010954304539522749, + "loss": 0.3848, + "step": 20037 + }, + { + "epoch": 0.4697222759424978, + "grad_norm": 2.1258323192596436, + "learning_rate": 0.00010953571282992254, + "loss": 0.1814, + "step": 20038 + }, + { + "epoch": 0.4697457175173028, + "grad_norm": 2.939805030822754, + "learning_rate": 0.00010952838021287655, + "loss": 0.3264, + "step": 20039 + }, + { + "epoch": 0.4697691590921078, + "grad_norm": 6.544392108917236, + "learning_rate": 0.00010952104754412929, + "loss": 0.5851, + "step": 20040 + }, + { + "epoch": 0.4697926006669128, + "grad_norm": 4.720094203948975, + "learning_rate": 0.00010951371482372057, + "loss": 0.4818, + "step": 20041 + }, + { + "epoch": 0.4698160422417178, + "grad_norm": 4.061634063720703, + "learning_rate": 0.00010950638205169018, + "loss": 0.3445, + "step": 20042 + }, + { + "epoch": 0.4698394838165228, + "grad_norm": 3.268664836883545, + "learning_rate": 0.00010949904922807781, + "loss": 1.7512, + "step": 20043 + }, + { + "epoch": 0.4698629253913278, + "grad_norm": 2.793126344680786, + "learning_rate": 0.00010949171635292342, + "loss": 1.1895, + "step": 20044 + }, + { + "epoch": 0.4698863669661328, + "grad_norm": 3.5853185653686523, + "learning_rate": 0.00010948438342626667, + "loss": 0.4499, + "step": 20045 + }, + { + "epoch": 0.4699098085409378, + "grad_norm": 5.10582971572876, + "learning_rate": 0.0001094770504481474, + "loss": 0.64, + "step": 20046 + }, + { + "epoch": 0.4699332501157428, + "grad_norm": 2.0636227130889893, + "learning_rate": 0.00010946971741860539, + "loss": 0.1188, + "step": 20047 + }, + { + "epoch": 0.46995669169054777, + "grad_norm": 3.0459675788879395, + "learning_rate": 0.00010946238433768043, + "loss": 0.2881, + "step": 20048 + }, + { + "epoch": 0.46998013326535276, + "grad_norm": 1.6431934833526611, + "learning_rate": 0.00010945505120541231, + "loss": 0.1237, + "step": 20049 + }, + { + "epoch": 0.47000357484015776, + "grad_norm": 6.2267279624938965, + "learning_rate": 0.00010944771802184077, + "loss": 0.7119, + "step": 20050 + }, + { + "epoch": 0.47002701641496275, + "grad_norm": 2.7020463943481445, + "learning_rate": 0.00010944038478700571, + "loss": 0.179, + "step": 20051 + }, + { + "epoch": 0.47005045798976774, + "grad_norm": 2.1766505241394043, + "learning_rate": 0.00010943305150094683, + "loss": 0.1777, + "step": 20052 + }, + { + "epoch": 0.47007389956457274, + "grad_norm": 1.654548168182373, + "learning_rate": 0.00010942571816370395, + "loss": 0.1054, + "step": 20053 + }, + { + "epoch": 0.47009734113937773, + "grad_norm": 4.727837562561035, + "learning_rate": 0.00010941838477531685, + "loss": 0.2428, + "step": 20054 + }, + { + "epoch": 0.47012078271418273, + "grad_norm": 3.6773338317871094, + "learning_rate": 0.00010941105133582535, + "loss": 0.4533, + "step": 20055 + }, + { + "epoch": 0.4701442242889877, + "grad_norm": 4.553866863250732, + "learning_rate": 0.00010940371784526917, + "loss": 0.526, + "step": 20056 + }, + { + "epoch": 0.4701676658637927, + "grad_norm": 2.8331642150878906, + "learning_rate": 0.00010939638430368822, + "loss": 0.287, + "step": 20057 + }, + { + "epoch": 0.4701911074385977, + "grad_norm": 0.8519133925437927, + "learning_rate": 0.0001093890507111222, + "loss": 0.061, + "step": 20058 + }, + { + "epoch": 0.4702145490134027, + "grad_norm": 2.2293570041656494, + "learning_rate": 0.00010938171706761088, + "loss": 0.184, + "step": 20059 + }, + { + "epoch": 0.4702379905882077, + "grad_norm": 3.298776149749756, + "learning_rate": 0.00010937438337319417, + "loss": 0.3083, + "step": 20060 + }, + { + "epoch": 0.4702614321630127, + "grad_norm": 1.4567769765853882, + "learning_rate": 0.00010936704962791175, + "loss": 0.9672, + "step": 20061 + }, + { + "epoch": 0.4702848737378177, + "grad_norm": 1.908799648284912, + "learning_rate": 0.00010935971583180347, + "loss": 0.1396, + "step": 20062 + }, + { + "epoch": 0.4703083153126227, + "grad_norm": 4.870742321014404, + "learning_rate": 0.00010935238198490911, + "loss": 0.408, + "step": 20063 + }, + { + "epoch": 0.4703317568874277, + "grad_norm": 2.148189067840576, + "learning_rate": 0.00010934504808726847, + "loss": 1.6924, + "step": 20064 + }, + { + "epoch": 0.47035519846223267, + "grad_norm": 4.602982997894287, + "learning_rate": 0.00010933771413892131, + "loss": 0.6871, + "step": 20065 + }, + { + "epoch": 0.47037864003703767, + "grad_norm": 5.1785664558410645, + "learning_rate": 0.00010933038013990747, + "loss": 0.6844, + "step": 20066 + }, + { + "epoch": 0.47040208161184266, + "grad_norm": 4.555720329284668, + "learning_rate": 0.00010932304609026673, + "loss": 0.3792, + "step": 20067 + }, + { + "epoch": 0.47042552318664765, + "grad_norm": 2.587368965148926, + "learning_rate": 0.00010931571199003885, + "loss": 1.1466, + "step": 20068 + }, + { + "epoch": 0.47044896476145265, + "grad_norm": 3.039165496826172, + "learning_rate": 0.00010930837783926367, + "loss": 0.3516, + "step": 20069 + }, + { + "epoch": 0.47047240633625764, + "grad_norm": 5.024366855621338, + "learning_rate": 0.000109301043637981, + "loss": 0.4221, + "step": 20070 + }, + { + "epoch": 0.4704958479110627, + "grad_norm": 2.4348981380462646, + "learning_rate": 0.00010929370938623056, + "loss": 0.2993, + "step": 20071 + }, + { + "epoch": 0.4705192894858677, + "grad_norm": 3.2612533569335938, + "learning_rate": 0.00010928637508405222, + "loss": 0.2257, + "step": 20072 + }, + { + "epoch": 0.4705427310606727, + "grad_norm": 3.0773565769195557, + "learning_rate": 0.00010927904073148574, + "loss": 0.3073, + "step": 20073 + }, + { + "epoch": 0.4705661726354777, + "grad_norm": 2.512162208557129, + "learning_rate": 0.00010927170632857091, + "loss": 1.5455, + "step": 20074 + }, + { + "epoch": 0.47058961421028267, + "grad_norm": 3.3554866313934326, + "learning_rate": 0.00010926437187534755, + "loss": 0.2081, + "step": 20075 + }, + { + "epoch": 0.47061305578508766, + "grad_norm": 4.946415424346924, + "learning_rate": 0.00010925703737185546, + "loss": 0.4462, + "step": 20076 + }, + { + "epoch": 0.47063649735989266, + "grad_norm": 1.8414636850357056, + "learning_rate": 0.00010924970281813443, + "loss": 0.1726, + "step": 20077 + }, + { + "epoch": 0.47065993893469765, + "grad_norm": 3.833632707595825, + "learning_rate": 0.00010924236821422422, + "loss": 0.42, + "step": 20078 + }, + { + "epoch": 0.47068338050950265, + "grad_norm": 3.9296817779541016, + "learning_rate": 0.00010923503356016469, + "loss": 0.3703, + "step": 20079 + }, + { + "epoch": 0.47070682208430764, + "grad_norm": 4.293691635131836, + "learning_rate": 0.00010922769885599559, + "loss": 0.4063, + "step": 20080 + }, + { + "epoch": 0.47073026365911264, + "grad_norm": 2.4038538932800293, + "learning_rate": 0.00010922036410175673, + "loss": 0.1762, + "step": 20081 + }, + { + "epoch": 0.47075370523391763, + "grad_norm": 1.3532459735870361, + "learning_rate": 0.00010921302929748795, + "loss": 0.0977, + "step": 20082 + }, + { + "epoch": 0.4707771468087226, + "grad_norm": 3.1006879806518555, + "learning_rate": 0.00010920569444322902, + "loss": 0.2358, + "step": 20083 + }, + { + "epoch": 0.4708005883835276, + "grad_norm": 2.845884323120117, + "learning_rate": 0.00010919835953901967, + "loss": 0.1801, + "step": 20084 + }, + { + "epoch": 0.4708240299583326, + "grad_norm": 0.8768733739852905, + "learning_rate": 0.00010919102458489982, + "loss": 0.0849, + "step": 20085 + }, + { + "epoch": 0.4708474715331376, + "grad_norm": 3.858919143676758, + "learning_rate": 0.00010918368958090918, + "loss": 0.3935, + "step": 20086 + }, + { + "epoch": 0.4708709131079426, + "grad_norm": 1.995020866394043, + "learning_rate": 0.0001091763545270876, + "loss": 0.2224, + "step": 20087 + }, + { + "epoch": 0.4708943546827476, + "grad_norm": 1.582694411277771, + "learning_rate": 0.00010916901942347484, + "loss": 0.2214, + "step": 20088 + }, + { + "epoch": 0.4709177962575526, + "grad_norm": 1.2899411916732788, + "learning_rate": 0.00010916168427011075, + "loss": 0.2081, + "step": 20089 + }, + { + "epoch": 0.4709412378323576, + "grad_norm": 3.466856002807617, + "learning_rate": 0.00010915434906703508, + "loss": 0.2668, + "step": 20090 + }, + { + "epoch": 0.4709646794071626, + "grad_norm": 4.430616855621338, + "learning_rate": 0.00010914701381428769, + "loss": 0.335, + "step": 20091 + }, + { + "epoch": 0.4709881209819676, + "grad_norm": 1.7302272319793701, + "learning_rate": 0.00010913967851190829, + "loss": 0.2509, + "step": 20092 + }, + { + "epoch": 0.47101156255677257, + "grad_norm": 1.2928680181503296, + "learning_rate": 0.00010913234315993678, + "loss": 0.1761, + "step": 20093 + }, + { + "epoch": 0.47103500413157756, + "grad_norm": 1.686040997505188, + "learning_rate": 0.00010912500775841293, + "loss": 1.1315, + "step": 20094 + }, + { + "epoch": 0.47105844570638256, + "grad_norm": 2.3467540740966797, + "learning_rate": 0.00010911767230737651, + "loss": 0.2169, + "step": 20095 + }, + { + "epoch": 0.47108188728118755, + "grad_norm": 4.1041083335876465, + "learning_rate": 0.00010911033680686735, + "loss": 0.5251, + "step": 20096 + }, + { + "epoch": 0.47110532885599254, + "grad_norm": 3.42832088470459, + "learning_rate": 0.00010910300125692523, + "loss": 0.3805, + "step": 20097 + }, + { + "epoch": 0.47112877043079754, + "grad_norm": 3.0440094470977783, + "learning_rate": 0.00010909566565758997, + "loss": 0.2228, + "step": 20098 + }, + { + "epoch": 0.47115221200560253, + "grad_norm": 2.4368908405303955, + "learning_rate": 0.0001090883300089014, + "loss": 1.2972, + "step": 20099 + }, + { + "epoch": 0.4711756535804075, + "grad_norm": 1.887497901916504, + "learning_rate": 0.00010908099431089926, + "loss": 0.2968, + "step": 20100 + }, + { + "epoch": 0.4711990951552125, + "grad_norm": 2.9624595642089844, + "learning_rate": 0.00010907365856362341, + "loss": 0.1814, + "step": 20101 + }, + { + "epoch": 0.4712225367300175, + "grad_norm": 2.372955560684204, + "learning_rate": 0.00010906632276711364, + "loss": 1.7653, + "step": 20102 + }, + { + "epoch": 0.4712459783048225, + "grad_norm": 1.9705471992492676, + "learning_rate": 0.00010905898692140972, + "loss": 0.2151, + "step": 20103 + }, + { + "epoch": 0.4712694198796275, + "grad_norm": 2.805316686630249, + "learning_rate": 0.0001090516510265515, + "loss": 0.3376, + "step": 20104 + }, + { + "epoch": 0.4712928614544325, + "grad_norm": 1.929543375968933, + "learning_rate": 0.00010904431508257877, + "loss": 0.243, + "step": 20105 + }, + { + "epoch": 0.4713163030292375, + "grad_norm": 2.4894585609436035, + "learning_rate": 0.00010903697908953134, + "loss": 1.7537, + "step": 20106 + }, + { + "epoch": 0.4713397446040425, + "grad_norm": 2.106372117996216, + "learning_rate": 0.000109029643047449, + "loss": 0.1209, + "step": 20107 + }, + { + "epoch": 0.4713631861788475, + "grad_norm": 4.0696702003479, + "learning_rate": 0.00010902230695637156, + "loss": 0.2955, + "step": 20108 + }, + { + "epoch": 0.4713866277536525, + "grad_norm": 6.4138007164001465, + "learning_rate": 0.00010901497081633883, + "loss": 0.2578, + "step": 20109 + }, + { + "epoch": 0.47141006932845747, + "grad_norm": 1.1665676832199097, + "learning_rate": 0.00010900763462739059, + "loss": 0.1121, + "step": 20110 + }, + { + "epoch": 0.47143351090326246, + "grad_norm": 1.3550375699996948, + "learning_rate": 0.00010900029838956671, + "loss": 0.1467, + "step": 20111 + }, + { + "epoch": 0.47145695247806746, + "grad_norm": 2.5068752765655518, + "learning_rate": 0.00010899296210290695, + "loss": 0.228, + "step": 20112 + }, + { + "epoch": 0.47148039405287245, + "grad_norm": 2.9686577320098877, + "learning_rate": 0.00010898562576745111, + "loss": 1.9317, + "step": 20113 + }, + { + "epoch": 0.47150383562767745, + "grad_norm": 1.2045202255249023, + "learning_rate": 0.00010897828938323904, + "loss": 0.224, + "step": 20114 + }, + { + "epoch": 0.47152727720248244, + "grad_norm": 2.3430116176605225, + "learning_rate": 0.0001089709529503105, + "loss": 0.159, + "step": 20115 + }, + { + "epoch": 0.47155071877728744, + "grad_norm": 3.9309983253479004, + "learning_rate": 0.00010896361646870527, + "loss": 0.4014, + "step": 20116 + }, + { + "epoch": 0.47157416035209243, + "grad_norm": 3.424325704574585, + "learning_rate": 0.00010895627993846327, + "loss": 0.4274, + "step": 20117 + }, + { + "epoch": 0.4715976019268974, + "grad_norm": 1.6066778898239136, + "learning_rate": 0.00010894894335962425, + "loss": 0.1276, + "step": 20118 + }, + { + "epoch": 0.4716210435017024, + "grad_norm": 2.709815502166748, + "learning_rate": 0.00010894160673222798, + "loss": 0.1933, + "step": 20119 + }, + { + "epoch": 0.4716444850765074, + "grad_norm": 1.7815951108932495, + "learning_rate": 0.00010893427005631432, + "loss": 0.988, + "step": 20120 + }, + { + "epoch": 0.4716679266513124, + "grad_norm": 2.681284189224243, + "learning_rate": 0.00010892693333192303, + "loss": 0.215, + "step": 20121 + }, + { + "epoch": 0.47169136822611746, + "grad_norm": 2.550504684448242, + "learning_rate": 0.00010891959655909396, + "loss": 0.1944, + "step": 20122 + }, + { + "epoch": 0.47171480980092245, + "grad_norm": 2.5564284324645996, + "learning_rate": 0.00010891225973786689, + "loss": 1.8499, + "step": 20123 + }, + { + "epoch": 0.47173825137572745, + "grad_norm": 1.1136213541030884, + "learning_rate": 0.00010890492286828169, + "loss": 0.1098, + "step": 20124 + }, + { + "epoch": 0.47176169295053244, + "grad_norm": 3.2525827884674072, + "learning_rate": 0.00010889758595037813, + "loss": 0.3266, + "step": 20125 + }, + { + "epoch": 0.47178513452533744, + "grad_norm": 5.7830939292907715, + "learning_rate": 0.00010889024898419597, + "loss": 0.8217, + "step": 20126 + }, + { + "epoch": 0.47180857610014243, + "grad_norm": 2.5110747814178467, + "learning_rate": 0.00010888291196977512, + "loss": 1.7388, + "step": 20127 + }, + { + "epoch": 0.4718320176749474, + "grad_norm": 0.7381631135940552, + "learning_rate": 0.00010887557490715529, + "loss": 0.0552, + "step": 20128 + }, + { + "epoch": 0.4718554592497524, + "grad_norm": 1.090166449546814, + "learning_rate": 0.00010886823779637638, + "loss": 0.1498, + "step": 20129 + }, + { + "epoch": 0.4718789008245574, + "grad_norm": 0.8843032121658325, + "learning_rate": 0.00010886090063747815, + "loss": 0.0612, + "step": 20130 + }, + { + "epoch": 0.4719023423993624, + "grad_norm": 5.116535186767578, + "learning_rate": 0.00010885356343050044, + "loss": 0.5183, + "step": 20131 + }, + { + "epoch": 0.4719257839741674, + "grad_norm": 2.1061246395111084, + "learning_rate": 0.00010884622617548301, + "loss": 0.1317, + "step": 20132 + }, + { + "epoch": 0.4719492255489724, + "grad_norm": 2.740138292312622, + "learning_rate": 0.00010883888887246573, + "loss": 0.2488, + "step": 20133 + }, + { + "epoch": 0.4719726671237774, + "grad_norm": 3.2701127529144287, + "learning_rate": 0.00010883155152148838, + "loss": 0.2279, + "step": 20134 + }, + { + "epoch": 0.4719961086985824, + "grad_norm": 0.6666979193687439, + "learning_rate": 0.00010882421412259078, + "loss": 0.0693, + "step": 20135 + }, + { + "epoch": 0.4720195502733874, + "grad_norm": 3.1722302436828613, + "learning_rate": 0.00010881687667581278, + "loss": 0.2747, + "step": 20136 + }, + { + "epoch": 0.4720429918481924, + "grad_norm": 2.546130657196045, + "learning_rate": 0.00010880953918119414, + "loss": 0.2487, + "step": 20137 + }, + { + "epoch": 0.47206643342299737, + "grad_norm": 14.458314895629883, + "learning_rate": 0.00010880220163877468, + "loss": 0.1383, + "step": 20138 + }, + { + "epoch": 0.47208987499780236, + "grad_norm": 3.0173139572143555, + "learning_rate": 0.00010879486404859425, + "loss": 0.2664, + "step": 20139 + }, + { + "epoch": 0.47211331657260736, + "grad_norm": 4.6491241455078125, + "learning_rate": 0.00010878752641069262, + "loss": 0.6768, + "step": 20140 + }, + { + "epoch": 0.47213675814741235, + "grad_norm": 2.812884569168091, + "learning_rate": 0.00010878018872510965, + "loss": 1.776, + "step": 20141 + }, + { + "epoch": 0.47216019972221734, + "grad_norm": 2.9699184894561768, + "learning_rate": 0.00010877285099188508, + "loss": 0.2971, + "step": 20142 + }, + { + "epoch": 0.47218364129702234, + "grad_norm": 2.7567014694213867, + "learning_rate": 0.00010876551321105881, + "loss": 0.3898, + "step": 20143 + }, + { + "epoch": 0.47220708287182733, + "grad_norm": 1.772686243057251, + "learning_rate": 0.00010875817538267063, + "loss": 0.1082, + "step": 20144 + }, + { + "epoch": 0.4722305244466323, + "grad_norm": 0.8491365909576416, + "learning_rate": 0.00010875083750676031, + "loss": 0.0683, + "step": 20145 + }, + { + "epoch": 0.4722539660214373, + "grad_norm": 1.1351487636566162, + "learning_rate": 0.00010874349958336774, + "loss": 0.1325, + "step": 20146 + }, + { + "epoch": 0.4722774075962423, + "grad_norm": 0.49934500455856323, + "learning_rate": 0.00010873616161253267, + "loss": 0.0352, + "step": 20147 + }, + { + "epoch": 0.4723008491710473, + "grad_norm": 2.4994726181030273, + "learning_rate": 0.00010872882359429493, + "loss": 0.1724, + "step": 20148 + }, + { + "epoch": 0.4723242907458523, + "grad_norm": 3.9101736545562744, + "learning_rate": 0.00010872148552869436, + "loss": 0.4372, + "step": 20149 + }, + { + "epoch": 0.4723477323206573, + "grad_norm": 1.5886212587356567, + "learning_rate": 0.00010871414741577077, + "loss": 1.083, + "step": 20150 + }, + { + "epoch": 0.4723711738954623, + "grad_norm": 3.2199923992156982, + "learning_rate": 0.00010870680925556395, + "loss": 1.935, + "step": 20151 + }, + { + "epoch": 0.4723946154702673, + "grad_norm": 3.9941799640655518, + "learning_rate": 0.00010869947104811377, + "loss": 0.3237, + "step": 20152 + }, + { + "epoch": 0.4724180570450723, + "grad_norm": 3.5502519607543945, + "learning_rate": 0.00010869213279346, + "loss": 0.2013, + "step": 20153 + }, + { + "epoch": 0.4724414986198773, + "grad_norm": 4.137287139892578, + "learning_rate": 0.00010868479449164246, + "loss": 0.4887, + "step": 20154 + }, + { + "epoch": 0.47246494019468227, + "grad_norm": 2.657217502593994, + "learning_rate": 0.00010867745614270099, + "loss": 0.2302, + "step": 20155 + }, + { + "epoch": 0.47248838176948726, + "grad_norm": 3.2659366130828857, + "learning_rate": 0.00010867011774667543, + "loss": 0.3993, + "step": 20156 + }, + { + "epoch": 0.47251182334429226, + "grad_norm": 0.8493834733963013, + "learning_rate": 0.00010866277930360553, + "loss": 0.08, + "step": 20157 + }, + { + "epoch": 0.47253526491909725, + "grad_norm": 3.857330322265625, + "learning_rate": 0.00010865544081353117, + "loss": 0.4455, + "step": 20158 + }, + { + "epoch": 0.47255870649390225, + "grad_norm": 2.700434684753418, + "learning_rate": 0.00010864810227649212, + "loss": 0.1512, + "step": 20159 + }, + { + "epoch": 0.47258214806870724, + "grad_norm": 2.1867125034332275, + "learning_rate": 0.00010864076369252825, + "loss": 0.2204, + "step": 20160 + }, + { + "epoch": 0.47260558964351224, + "grad_norm": 5.943007469177246, + "learning_rate": 0.00010863342506167931, + "loss": 0.2949, + "step": 20161 + }, + { + "epoch": 0.47262903121831723, + "grad_norm": 2.27417254447937, + "learning_rate": 0.00010862608638398521, + "loss": 0.2515, + "step": 20162 + }, + { + "epoch": 0.4726524727931222, + "grad_norm": 1.0096250772476196, + "learning_rate": 0.00010861874765948572, + "loss": 0.0838, + "step": 20163 + }, + { + "epoch": 0.4726759143679272, + "grad_norm": 1.4307228326797485, + "learning_rate": 0.00010861140888822064, + "loss": 0.1517, + "step": 20164 + }, + { + "epoch": 0.4726993559427322, + "grad_norm": 3.821202278137207, + "learning_rate": 0.00010860407007022979, + "loss": 0.3571, + "step": 20165 + }, + { + "epoch": 0.4727227975175372, + "grad_norm": 3.7225759029388428, + "learning_rate": 0.00010859673120555304, + "loss": 0.2998, + "step": 20166 + }, + { + "epoch": 0.4727462390923422, + "grad_norm": 6.041740417480469, + "learning_rate": 0.0001085893922942302, + "loss": 0.343, + "step": 20167 + }, + { + "epoch": 0.4727696806671472, + "grad_norm": 3.118743658065796, + "learning_rate": 0.00010858205333630107, + "loss": 0.434, + "step": 20168 + }, + { + "epoch": 0.4727931222419522, + "grad_norm": 2.4886314868927, + "learning_rate": 0.00010857471433180548, + "loss": 0.2819, + "step": 20169 + }, + { + "epoch": 0.4728165638167572, + "grad_norm": 3.512787103652954, + "learning_rate": 0.00010856737528078322, + "loss": 0.2552, + "step": 20170 + }, + { + "epoch": 0.4728400053915622, + "grad_norm": 3.9730443954467773, + "learning_rate": 0.00010856003618327415, + "loss": 0.2612, + "step": 20171 + }, + { + "epoch": 0.4728634469663672, + "grad_norm": 2.665194272994995, + "learning_rate": 0.00010855269703931812, + "loss": 0.1945, + "step": 20172 + }, + { + "epoch": 0.4728868885411722, + "grad_norm": 2.6591217517852783, + "learning_rate": 0.00010854535784895488, + "loss": 0.349, + "step": 20173 + }, + { + "epoch": 0.4729103301159772, + "grad_norm": 3.3890395164489746, + "learning_rate": 0.00010853801861222431, + "loss": 0.3799, + "step": 20174 + }, + { + "epoch": 0.4729337716907822, + "grad_norm": 3.488532781600952, + "learning_rate": 0.00010853067932916621, + "loss": 0.416, + "step": 20175 + }, + { + "epoch": 0.4729572132655872, + "grad_norm": 4.067234992980957, + "learning_rate": 0.00010852333999982039, + "loss": 0.2411, + "step": 20176 + }, + { + "epoch": 0.4729806548403922, + "grad_norm": 2.583202838897705, + "learning_rate": 0.00010851600062422667, + "loss": 0.3242, + "step": 20177 + }, + { + "epoch": 0.4730040964151972, + "grad_norm": 1.4976125955581665, + "learning_rate": 0.00010850866120242491, + "loss": 0.123, + "step": 20178 + }, + { + "epoch": 0.4730275379900022, + "grad_norm": 3.7239813804626465, + "learning_rate": 0.00010850132173445492, + "loss": 0.3303, + "step": 20179 + }, + { + "epoch": 0.4730509795648072, + "grad_norm": 4.195439338684082, + "learning_rate": 0.00010849398222035651, + "loss": 0.4357, + "step": 20180 + }, + { + "epoch": 0.4730744211396122, + "grad_norm": 4.888631343841553, + "learning_rate": 0.00010848664266016955, + "loss": 0.3453, + "step": 20181 + }, + { + "epoch": 0.47309786271441717, + "grad_norm": 1.5485697984695435, + "learning_rate": 0.00010847930305393378, + "loss": 0.6916, + "step": 20182 + }, + { + "epoch": 0.47312130428922217, + "grad_norm": 4.501913547515869, + "learning_rate": 0.00010847196340168908, + "loss": 0.433, + "step": 20183 + }, + { + "epoch": 0.47314474586402716, + "grad_norm": 3.892460584640503, + "learning_rate": 0.00010846462370347527, + "loss": 0.5338, + "step": 20184 + }, + { + "epoch": 0.47316818743883216, + "grad_norm": 3.5658459663391113, + "learning_rate": 0.00010845728395933221, + "loss": 0.4445, + "step": 20185 + }, + { + "epoch": 0.47319162901363715, + "grad_norm": 4.064870357513428, + "learning_rate": 0.00010844994416929966, + "loss": 0.43, + "step": 20186 + }, + { + "epoch": 0.47321507058844214, + "grad_norm": 1.5315072536468506, + "learning_rate": 0.00010844260433341749, + "loss": 0.1733, + "step": 20187 + }, + { + "epoch": 0.47323851216324714, + "grad_norm": 2.9490792751312256, + "learning_rate": 0.00010843526445172549, + "loss": 0.1698, + "step": 20188 + }, + { + "epoch": 0.47326195373805213, + "grad_norm": 4.430788993835449, + "learning_rate": 0.00010842792452426349, + "loss": 0.6812, + "step": 20189 + }, + { + "epoch": 0.4732853953128571, + "grad_norm": 3.3135874271392822, + "learning_rate": 0.00010842058455107137, + "loss": 0.2556, + "step": 20190 + }, + { + "epoch": 0.4733088368876621, + "grad_norm": 2.13254451751709, + "learning_rate": 0.00010841324453218892, + "loss": 1.4135, + "step": 20191 + }, + { + "epoch": 0.4733322784624671, + "grad_norm": 3.7656519412994385, + "learning_rate": 0.00010840590446765595, + "loss": 0.9957, + "step": 20192 + }, + { + "epoch": 0.4733557200372721, + "grad_norm": 2.0165717601776123, + "learning_rate": 0.00010839856435751234, + "loss": 0.1236, + "step": 20193 + }, + { + "epoch": 0.4733791616120771, + "grad_norm": 3.179154396057129, + "learning_rate": 0.00010839122420179789, + "loss": 0.2254, + "step": 20194 + }, + { + "epoch": 0.4734026031868821, + "grad_norm": 1.7159674167633057, + "learning_rate": 0.00010838388400055237, + "loss": 0.1064, + "step": 20195 + }, + { + "epoch": 0.4734260447616871, + "grad_norm": 2.0965404510498047, + "learning_rate": 0.00010837654375381565, + "loss": 0.2488, + "step": 20196 + }, + { + "epoch": 0.4734494863364921, + "grad_norm": 2.900404691696167, + "learning_rate": 0.00010836920346162762, + "loss": 0.2454, + "step": 20197 + }, + { + "epoch": 0.4734729279112971, + "grad_norm": 2.607706308364868, + "learning_rate": 0.00010836186312402805, + "loss": 0.3177, + "step": 20198 + }, + { + "epoch": 0.4734963694861021, + "grad_norm": 4.017857074737549, + "learning_rate": 0.00010835452274105675, + "loss": 0.4443, + "step": 20199 + }, + { + "epoch": 0.47351981106090707, + "grad_norm": 1.7849746942520142, + "learning_rate": 0.00010834718231275361, + "loss": 0.1772, + "step": 20200 + }, + { + "epoch": 0.47354325263571206, + "grad_norm": 2.2110681533813477, + "learning_rate": 0.00010833984183915837, + "loss": 0.1307, + "step": 20201 + }, + { + "epoch": 0.47356669421051706, + "grad_norm": 2.784830331802368, + "learning_rate": 0.00010833250132031094, + "loss": 1.6003, + "step": 20202 + }, + { + "epoch": 0.47359013578532205, + "grad_norm": 2.8596577644348145, + "learning_rate": 0.00010832516075625114, + "loss": 0.3076, + "step": 20203 + }, + { + "epoch": 0.47361357736012705, + "grad_norm": 2.6314573287963867, + "learning_rate": 0.0001083178201470188, + "loss": 0.1976, + "step": 20204 + }, + { + "epoch": 0.47363701893493204, + "grad_norm": 3.4025778770446777, + "learning_rate": 0.0001083104794926537, + "loss": 0.3804, + "step": 20205 + }, + { + "epoch": 0.47366046050973704, + "grad_norm": 3.2198309898376465, + "learning_rate": 0.00010830313879319571, + "loss": 0.4859, + "step": 20206 + }, + { + "epoch": 0.47368390208454203, + "grad_norm": 5.097441673278809, + "learning_rate": 0.00010829579804868464, + "loss": 0.6841, + "step": 20207 + }, + { + "epoch": 0.473707343659347, + "grad_norm": 2.7431955337524414, + "learning_rate": 0.00010828845725916035, + "loss": 0.3184, + "step": 20208 + }, + { + "epoch": 0.473730785234152, + "grad_norm": 2.833681583404541, + "learning_rate": 0.00010828111642466268, + "loss": 0.2734, + "step": 20209 + }, + { + "epoch": 0.473754226808957, + "grad_norm": 3.9557271003723145, + "learning_rate": 0.00010827377554523141, + "loss": 0.2814, + "step": 20210 + }, + { + "epoch": 0.473777668383762, + "grad_norm": 3.6010472774505615, + "learning_rate": 0.00010826643462090643, + "loss": 0.2139, + "step": 20211 + }, + { + "epoch": 0.473801109958567, + "grad_norm": 2.575252056121826, + "learning_rate": 0.00010825909365172751, + "loss": 0.315, + "step": 20212 + }, + { + "epoch": 0.473824551533372, + "grad_norm": 3.00103759765625, + "learning_rate": 0.00010825175263773453, + "loss": 0.3779, + "step": 20213 + }, + { + "epoch": 0.473847993108177, + "grad_norm": 4.405328273773193, + "learning_rate": 0.00010824441157896732, + "loss": 0.3843, + "step": 20214 + }, + { + "epoch": 0.473871434682982, + "grad_norm": 2.8628008365631104, + "learning_rate": 0.00010823707047546567, + "loss": 0.2466, + "step": 20215 + }, + { + "epoch": 0.473894876257787, + "grad_norm": 1.5258166790008545, + "learning_rate": 0.00010822972932726947, + "loss": 0.25, + "step": 20216 + }, + { + "epoch": 0.473918317832592, + "grad_norm": 3.9256746768951416, + "learning_rate": 0.00010822238813441853, + "loss": 0.3404, + "step": 20217 + }, + { + "epoch": 0.47394175940739697, + "grad_norm": 3.551154613494873, + "learning_rate": 0.00010821504689695268, + "loss": 0.5552, + "step": 20218 + }, + { + "epoch": 0.47396520098220196, + "grad_norm": 3.8034706115722656, + "learning_rate": 0.00010820770561491176, + "loss": 0.2803, + "step": 20219 + }, + { + "epoch": 0.47398864255700696, + "grad_norm": 1.766782283782959, + "learning_rate": 0.00010820036428833555, + "loss": 0.0797, + "step": 20220 + }, + { + "epoch": 0.47401208413181195, + "grad_norm": 3.172417163848877, + "learning_rate": 0.00010819302291726397, + "loss": 0.3479, + "step": 20221 + }, + { + "epoch": 0.47403552570661694, + "grad_norm": 3.2319769859313965, + "learning_rate": 0.00010818568150173682, + "loss": 0.3461, + "step": 20222 + }, + { + "epoch": 0.47405896728142194, + "grad_norm": 3.065551280975342, + "learning_rate": 0.00010817834004179395, + "loss": 0.3153, + "step": 20223 + }, + { + "epoch": 0.474082408856227, + "grad_norm": 1.563174843788147, + "learning_rate": 0.00010817099853747515, + "loss": 0.1682, + "step": 20224 + }, + { + "epoch": 0.474105850431032, + "grad_norm": 3.3417203426361084, + "learning_rate": 0.0001081636569888203, + "loss": 0.2947, + "step": 20225 + }, + { + "epoch": 0.474129292005837, + "grad_norm": 3.3169970512390137, + "learning_rate": 0.00010815631539586917, + "loss": 0.3264, + "step": 20226 + }, + { + "epoch": 0.47415273358064197, + "grad_norm": 7.517910003662109, + "learning_rate": 0.00010814897375866166, + "loss": 0.7186, + "step": 20227 + }, + { + "epoch": 0.47417617515544697, + "grad_norm": 3.128657579421997, + "learning_rate": 0.00010814163207723762, + "loss": 0.3603, + "step": 20228 + }, + { + "epoch": 0.47419961673025196, + "grad_norm": 2.387493371963501, + "learning_rate": 0.00010813429035163686, + "loss": 0.1311, + "step": 20229 + }, + { + "epoch": 0.47422305830505695, + "grad_norm": 2.4011359214782715, + "learning_rate": 0.00010812694858189917, + "loss": 1.5989, + "step": 20230 + }, + { + "epoch": 0.47424649987986195, + "grad_norm": 1.955330491065979, + "learning_rate": 0.00010811960676806448, + "loss": 0.1681, + "step": 20231 + }, + { + "epoch": 0.47426994145466694, + "grad_norm": 1.8030778169631958, + "learning_rate": 0.00010811226491017254, + "loss": 0.1061, + "step": 20232 + }, + { + "epoch": 0.47429338302947194, + "grad_norm": 4.503469944000244, + "learning_rate": 0.00010810492300826323, + "loss": 0.4155, + "step": 20233 + }, + { + "epoch": 0.47431682460427693, + "grad_norm": 8.58820629119873, + "learning_rate": 0.00010809758106237637, + "loss": 0.493, + "step": 20234 + }, + { + "epoch": 0.4743402661790819, + "grad_norm": 5.041870594024658, + "learning_rate": 0.00010809023907255183, + "loss": 0.4021, + "step": 20235 + }, + { + "epoch": 0.4743637077538869, + "grad_norm": 6.461726188659668, + "learning_rate": 0.00010808289703882943, + "loss": 0.4818, + "step": 20236 + }, + { + "epoch": 0.4743871493286919, + "grad_norm": 1.4562417268753052, + "learning_rate": 0.00010807555496124896, + "loss": 0.113, + "step": 20237 + }, + { + "epoch": 0.4744105909034969, + "grad_norm": 1.4662946462631226, + "learning_rate": 0.00010806821283985034, + "loss": 0.1153, + "step": 20238 + }, + { + "epoch": 0.4744340324783019, + "grad_norm": 4.165161609649658, + "learning_rate": 0.00010806087067467337, + "loss": 0.4952, + "step": 20239 + }, + { + "epoch": 0.4744574740531069, + "grad_norm": 0.7684903144836426, + "learning_rate": 0.00010805352846575785, + "loss": 0.0533, + "step": 20240 + }, + { + "epoch": 0.4744809156279119, + "grad_norm": 2.336451530456543, + "learning_rate": 0.0001080461862131437, + "loss": 0.2155, + "step": 20241 + }, + { + "epoch": 0.4745043572027169, + "grad_norm": 5.258605480194092, + "learning_rate": 0.00010803884391687071, + "loss": 0.5381, + "step": 20242 + }, + { + "epoch": 0.4745277987775219, + "grad_norm": 3.3437671661376953, + "learning_rate": 0.0001080315015769787, + "loss": 0.3262, + "step": 20243 + }, + { + "epoch": 0.4745512403523269, + "grad_norm": 4.518928527832031, + "learning_rate": 0.00010802415919350755, + "loss": 0.4429, + "step": 20244 + }, + { + "epoch": 0.47457468192713187, + "grad_norm": 2.3495166301727295, + "learning_rate": 0.0001080168167664971, + "loss": 0.2224, + "step": 20245 + }, + { + "epoch": 0.47459812350193686, + "grad_norm": 3.275090217590332, + "learning_rate": 0.00010800947429598719, + "loss": 0.2574, + "step": 20246 + }, + { + "epoch": 0.47462156507674186, + "grad_norm": 3.2471656799316406, + "learning_rate": 0.00010800213178201762, + "loss": 0.3731, + "step": 20247 + }, + { + "epoch": 0.47464500665154685, + "grad_norm": 0.8644886612892151, + "learning_rate": 0.00010799478922462827, + "loss": 0.5125, + "step": 20248 + }, + { + "epoch": 0.47466844822635185, + "grad_norm": 3.7063865661621094, + "learning_rate": 0.00010798744662385896, + "loss": 0.5232, + "step": 20249 + }, + { + "epoch": 0.47469188980115684, + "grad_norm": 4.7919721603393555, + "learning_rate": 0.00010798010397974954, + "loss": 0.3903, + "step": 20250 + }, + { + "epoch": 0.47471533137596184, + "grad_norm": 3.3340606689453125, + "learning_rate": 0.00010797276129233984, + "loss": 1.6845, + "step": 20251 + }, + { + "epoch": 0.47473877295076683, + "grad_norm": 3.569754123687744, + "learning_rate": 0.00010796541856166975, + "loss": 0.2254, + "step": 20252 + }, + { + "epoch": 0.4747622145255718, + "grad_norm": 1.64529287815094, + "learning_rate": 0.00010795807578777903, + "loss": 0.1397, + "step": 20253 + }, + { + "epoch": 0.4747856561003768, + "grad_norm": 1.7624644041061401, + "learning_rate": 0.00010795073297070762, + "loss": 0.1468, + "step": 20254 + }, + { + "epoch": 0.4748090976751818, + "grad_norm": 2.667471408843994, + "learning_rate": 0.00010794339011049528, + "loss": 0.204, + "step": 20255 + }, + { + "epoch": 0.4748325392499868, + "grad_norm": 4.343811511993408, + "learning_rate": 0.00010793604720718184, + "loss": 0.4541, + "step": 20256 + }, + { + "epoch": 0.4748559808247918, + "grad_norm": 1.8194801807403564, + "learning_rate": 0.00010792870426080725, + "loss": 0.1929, + "step": 20257 + }, + { + "epoch": 0.4748794223995968, + "grad_norm": 2.6466994285583496, + "learning_rate": 0.00010792136127141128, + "loss": 0.2764, + "step": 20258 + }, + { + "epoch": 0.4749028639744018, + "grad_norm": 1.4523367881774902, + "learning_rate": 0.00010791401823903375, + "loss": 0.1518, + "step": 20259 + }, + { + "epoch": 0.4749263055492068, + "grad_norm": 2.518037796020508, + "learning_rate": 0.00010790667516371455, + "loss": 0.422, + "step": 20260 + }, + { + "epoch": 0.4749497471240118, + "grad_norm": 3.434075117111206, + "learning_rate": 0.00010789933204549352, + "loss": 0.3615, + "step": 20261 + }, + { + "epoch": 0.4749731886988168, + "grad_norm": 0.6795756816864014, + "learning_rate": 0.00010789198888441048, + "loss": 0.0815, + "step": 20262 + }, + { + "epoch": 0.47499663027362177, + "grad_norm": 1.6526668071746826, + "learning_rate": 0.00010788464568050526, + "loss": 0.1837, + "step": 20263 + }, + { + "epoch": 0.47502007184842676, + "grad_norm": 3.0869204998016357, + "learning_rate": 0.00010787730243381777, + "loss": 0.1701, + "step": 20264 + }, + { + "epoch": 0.47504351342323176, + "grad_norm": 6.7855916023254395, + "learning_rate": 0.0001078699591443878, + "loss": 0.6398, + "step": 20265 + }, + { + "epoch": 0.47506695499803675, + "grad_norm": 3.4924874305725098, + "learning_rate": 0.00010786261581225523, + "loss": 1.3327, + "step": 20266 + }, + { + "epoch": 0.47509039657284174, + "grad_norm": 3.1175405979156494, + "learning_rate": 0.00010785527243745987, + "loss": 0.254, + "step": 20267 + }, + { + "epoch": 0.47511383814764674, + "grad_norm": 3.1087429523468018, + "learning_rate": 0.00010784792902004157, + "loss": 0.153, + "step": 20268 + }, + { + "epoch": 0.47513727972245173, + "grad_norm": 8.465348243713379, + "learning_rate": 0.00010784058556004018, + "loss": 0.2207, + "step": 20269 + }, + { + "epoch": 0.4751607212972567, + "grad_norm": 4.862687587738037, + "learning_rate": 0.00010783324205749558, + "loss": 0.6257, + "step": 20270 + }, + { + "epoch": 0.4751841628720617, + "grad_norm": 0.9927590489387512, + "learning_rate": 0.0001078258985124476, + "loss": 0.0897, + "step": 20271 + }, + { + "epoch": 0.4752076044468667, + "grad_norm": 2.78234601020813, + "learning_rate": 0.00010781855492493602, + "loss": 1.0441, + "step": 20272 + }, + { + "epoch": 0.4752310460216717, + "grad_norm": 2.661358118057251, + "learning_rate": 0.0001078112112950008, + "loss": 0.3666, + "step": 20273 + }, + { + "epoch": 0.4752544875964767, + "grad_norm": 1.7981518507003784, + "learning_rate": 0.00010780386762268169, + "loss": 0.1759, + "step": 20274 + }, + { + "epoch": 0.47527792917128175, + "grad_norm": 4.28582239151001, + "learning_rate": 0.00010779652390801858, + "loss": 0.5997, + "step": 20275 + }, + { + "epoch": 0.47530137074608675, + "grad_norm": 0.9226189851760864, + "learning_rate": 0.00010778918015105131, + "loss": 0.1115, + "step": 20276 + }, + { + "epoch": 0.47532481232089174, + "grad_norm": 3.401280164718628, + "learning_rate": 0.00010778183635181975, + "loss": 0.2861, + "step": 20277 + }, + { + "epoch": 0.47534825389569674, + "grad_norm": 1.048214077949524, + "learning_rate": 0.00010777449251036371, + "loss": 0.0931, + "step": 20278 + }, + { + "epoch": 0.47537169547050173, + "grad_norm": 3.8607685565948486, + "learning_rate": 0.00010776714862672306, + "loss": 0.3541, + "step": 20279 + }, + { + "epoch": 0.4753951370453067, + "grad_norm": 3.6027700901031494, + "learning_rate": 0.00010775980470093762, + "loss": 0.3554, + "step": 20280 + }, + { + "epoch": 0.4754185786201117, + "grad_norm": 5.063754081726074, + "learning_rate": 0.0001077524607330473, + "loss": 0.6291, + "step": 20281 + }, + { + "epoch": 0.4754420201949167, + "grad_norm": 2.435929298400879, + "learning_rate": 0.00010774511672309189, + "loss": 0.3067, + "step": 20282 + }, + { + "epoch": 0.4754654617697217, + "grad_norm": 4.097647190093994, + "learning_rate": 0.00010773777267111127, + "loss": 0.5083, + "step": 20283 + }, + { + "epoch": 0.4754889033445267, + "grad_norm": 1.9330806732177734, + "learning_rate": 0.00010773042857714526, + "loss": 0.147, + "step": 20284 + }, + { + "epoch": 0.4755123449193317, + "grad_norm": 1.3759409189224243, + "learning_rate": 0.00010772308444123376, + "loss": 0.1394, + "step": 20285 + }, + { + "epoch": 0.4755357864941367, + "grad_norm": 3.380305051803589, + "learning_rate": 0.00010771574026341656, + "loss": 0.4531, + "step": 20286 + }, + { + "epoch": 0.4755592280689417, + "grad_norm": 0.45265448093414307, + "learning_rate": 0.00010770839604373353, + "loss": 0.0333, + "step": 20287 + }, + { + "epoch": 0.4755826696437467, + "grad_norm": 4.394320487976074, + "learning_rate": 0.00010770105178222454, + "loss": 0.1715, + "step": 20288 + }, + { + "epoch": 0.4756061112185517, + "grad_norm": 0.5722806453704834, + "learning_rate": 0.00010769370747892942, + "loss": 0.0571, + "step": 20289 + }, + { + "epoch": 0.47562955279335667, + "grad_norm": 1.0143723487854004, + "learning_rate": 0.00010768636313388804, + "loss": 0.1682, + "step": 20290 + }, + { + "epoch": 0.47565299436816166, + "grad_norm": 2.9912381172180176, + "learning_rate": 0.00010767901874714022, + "loss": 0.1973, + "step": 20291 + }, + { + "epoch": 0.47567643594296666, + "grad_norm": 1.6432442665100098, + "learning_rate": 0.00010767167431872582, + "loss": 0.2128, + "step": 20292 + }, + { + "epoch": 0.47569987751777165, + "grad_norm": 2.2012341022491455, + "learning_rate": 0.00010766432984868473, + "loss": 0.1217, + "step": 20293 + }, + { + "epoch": 0.47572331909257665, + "grad_norm": 4.354483127593994, + "learning_rate": 0.00010765698533705675, + "loss": 0.5154, + "step": 20294 + }, + { + "epoch": 0.47574676066738164, + "grad_norm": 3.0596983432769775, + "learning_rate": 0.00010764964078388178, + "loss": 0.2291, + "step": 20295 + }, + { + "epoch": 0.47577020224218664, + "grad_norm": 3.0188357830047607, + "learning_rate": 0.00010764229618919963, + "loss": 0.324, + "step": 20296 + }, + { + "epoch": 0.47579364381699163, + "grad_norm": 2.992255449295044, + "learning_rate": 0.00010763495155305015, + "loss": 0.287, + "step": 20297 + }, + { + "epoch": 0.4758170853917966, + "grad_norm": 5.856387138366699, + "learning_rate": 0.00010762760687547322, + "loss": 0.6439, + "step": 20298 + }, + { + "epoch": 0.4758405269666016, + "grad_norm": 1.2286324501037598, + "learning_rate": 0.00010762026215650866, + "loss": 0.0971, + "step": 20299 + }, + { + "epoch": 0.4758639685414066, + "grad_norm": 3.5057573318481445, + "learning_rate": 0.00010761291739619636, + "loss": 0.2911, + "step": 20300 + }, + { + "epoch": 0.4758874101162116, + "grad_norm": 0.6959079504013062, + "learning_rate": 0.00010760557259457615, + "loss": 0.1044, + "step": 20301 + }, + { + "epoch": 0.4759108516910166, + "grad_norm": 2.025444746017456, + "learning_rate": 0.0001075982277516879, + "loss": 0.1569, + "step": 20302 + }, + { + "epoch": 0.4759342932658216, + "grad_norm": 2.501004695892334, + "learning_rate": 0.00010759088286757147, + "loss": 0.2877, + "step": 20303 + }, + { + "epoch": 0.4759577348406266, + "grad_norm": 3.067591667175293, + "learning_rate": 0.00010758353794226668, + "loss": 0.2288, + "step": 20304 + }, + { + "epoch": 0.4759811764154316, + "grad_norm": 4.202296733856201, + "learning_rate": 0.00010757619297581339, + "loss": 0.3662, + "step": 20305 + }, + { + "epoch": 0.4760046179902366, + "grad_norm": 3.5552797317504883, + "learning_rate": 0.00010756884796825147, + "loss": 0.3839, + "step": 20306 + }, + { + "epoch": 0.4760280595650416, + "grad_norm": 3.4633777141571045, + "learning_rate": 0.00010756150291962075, + "loss": 0.4153, + "step": 20307 + }, + { + "epoch": 0.47605150113984657, + "grad_norm": 4.331472396850586, + "learning_rate": 0.00010755415782996112, + "loss": 0.3482, + "step": 20308 + }, + { + "epoch": 0.47607494271465156, + "grad_norm": 0.36186668276786804, + "learning_rate": 0.00010754681269931243, + "loss": 0.0363, + "step": 20309 + }, + { + "epoch": 0.47609838428945656, + "grad_norm": 3.189744472503662, + "learning_rate": 0.00010753946752771452, + "loss": 0.1871, + "step": 20310 + }, + { + "epoch": 0.47612182586426155, + "grad_norm": 1.892258882522583, + "learning_rate": 0.00010753212231520721, + "loss": 0.2135, + "step": 20311 + }, + { + "epoch": 0.47614526743906654, + "grad_norm": 2.087984800338745, + "learning_rate": 0.00010752477706183045, + "loss": 0.2051, + "step": 20312 + }, + { + "epoch": 0.47616870901387154, + "grad_norm": 0.717806875705719, + "learning_rate": 0.00010751743176762397, + "loss": 0.0444, + "step": 20313 + }, + { + "epoch": 0.47619215058867653, + "grad_norm": 3.181797742843628, + "learning_rate": 0.00010751008643262774, + "loss": 0.3136, + "step": 20314 + }, + { + "epoch": 0.4762155921634815, + "grad_norm": 4.050534248352051, + "learning_rate": 0.00010750274105688158, + "loss": 0.4696, + "step": 20315 + }, + { + "epoch": 0.4762390337382865, + "grad_norm": 1.2981147766113281, + "learning_rate": 0.0001074953956404253, + "loss": 0.1668, + "step": 20316 + }, + { + "epoch": 0.4762624753130915, + "grad_norm": 5.704726219177246, + "learning_rate": 0.0001074880501832988, + "loss": 0.4517, + "step": 20317 + }, + { + "epoch": 0.4762859168878965, + "grad_norm": 6.479120254516602, + "learning_rate": 0.00010748070468554194, + "loss": 0.5214, + "step": 20318 + }, + { + "epoch": 0.4763093584627015, + "grad_norm": 0.6365209221839905, + "learning_rate": 0.00010747335914719459, + "loss": 0.0959, + "step": 20319 + }, + { + "epoch": 0.4763328000375065, + "grad_norm": 3.5888352394104004, + "learning_rate": 0.00010746601356829653, + "loss": 0.32, + "step": 20320 + }, + { + "epoch": 0.4763562416123115, + "grad_norm": 2.2311511039733887, + "learning_rate": 0.00010745866794888773, + "loss": 0.2466, + "step": 20321 + }, + { + "epoch": 0.4763796831871165, + "grad_norm": 2.5639798641204834, + "learning_rate": 0.00010745132228900794, + "loss": 0.2783, + "step": 20322 + }, + { + "epoch": 0.4764031247619215, + "grad_norm": 2.271929979324341, + "learning_rate": 0.00010744397658869707, + "loss": 0.1541, + "step": 20323 + }, + { + "epoch": 0.4764265663367265, + "grad_norm": 3.487083911895752, + "learning_rate": 0.00010743663084799498, + "loss": 0.3451, + "step": 20324 + }, + { + "epoch": 0.47645000791153147, + "grad_norm": 5.295649528503418, + "learning_rate": 0.00010742928506694155, + "loss": 0.3277, + "step": 20325 + }, + { + "epoch": 0.47647344948633646, + "grad_norm": 1.0711815357208252, + "learning_rate": 0.00010742193924557658, + "loss": 0.1583, + "step": 20326 + }, + { + "epoch": 0.4764968910611415, + "grad_norm": 1.891869306564331, + "learning_rate": 0.00010741459338393998, + "loss": 0.0756, + "step": 20327 + }, + { + "epoch": 0.4765203326359465, + "grad_norm": 1.4816648960113525, + "learning_rate": 0.00010740724748207156, + "loss": 0.2138, + "step": 20328 + }, + { + "epoch": 0.4765437742107515, + "grad_norm": 2.5288748741149902, + "learning_rate": 0.0001073999015400112, + "loss": 0.1805, + "step": 20329 + }, + { + "epoch": 0.4765672157855565, + "grad_norm": 1.9363797903060913, + "learning_rate": 0.00010739255555779878, + "loss": 0.1585, + "step": 20330 + }, + { + "epoch": 0.4765906573603615, + "grad_norm": 2.5620386600494385, + "learning_rate": 0.00010738520953547418, + "loss": 0.4006, + "step": 20331 + }, + { + "epoch": 0.4766140989351665, + "grad_norm": 3.12921404838562, + "learning_rate": 0.00010737786347307719, + "loss": 0.2452, + "step": 20332 + }, + { + "epoch": 0.4766375405099715, + "grad_norm": 3.4756433963775635, + "learning_rate": 0.00010737051737064772, + "loss": 0.3473, + "step": 20333 + }, + { + "epoch": 0.4766609820847765, + "grad_norm": 0.9505228996276855, + "learning_rate": 0.00010736317122822562, + "loss": 0.0729, + "step": 20334 + }, + { + "epoch": 0.47668442365958147, + "grad_norm": 2.6880712509155273, + "learning_rate": 0.00010735582504585074, + "loss": 1.315, + "step": 20335 + }, + { + "epoch": 0.47670786523438646, + "grad_norm": 3.7879395484924316, + "learning_rate": 0.00010734847882356291, + "loss": 0.4394, + "step": 20336 + }, + { + "epoch": 0.47673130680919146, + "grad_norm": 3.5780136585235596, + "learning_rate": 0.00010734113256140208, + "loss": 0.2233, + "step": 20337 + }, + { + "epoch": 0.47675474838399645, + "grad_norm": 1.998497486114502, + "learning_rate": 0.00010733378625940801, + "loss": 0.1908, + "step": 20338 + }, + { + "epoch": 0.47677818995880145, + "grad_norm": 2.3510637283325195, + "learning_rate": 0.00010732643991762066, + "loss": 0.3351, + "step": 20339 + }, + { + "epoch": 0.47680163153360644, + "grad_norm": 3.2237772941589355, + "learning_rate": 0.00010731909353607984, + "loss": 0.382, + "step": 20340 + }, + { + "epoch": 0.47682507310841143, + "grad_norm": 3.4449548721313477, + "learning_rate": 0.00010731174711482537, + "loss": 1.9952, + "step": 20341 + }, + { + "epoch": 0.47684851468321643, + "grad_norm": 1.984481930732727, + "learning_rate": 0.00010730440065389716, + "loss": 0.1749, + "step": 20342 + }, + { + "epoch": 0.4768719562580214, + "grad_norm": 2.470618486404419, + "learning_rate": 0.00010729705415333508, + "loss": 1.2737, + "step": 20343 + }, + { + "epoch": 0.4768953978328264, + "grad_norm": 1.6499165296554565, + "learning_rate": 0.000107289707613179, + "loss": 0.211, + "step": 20344 + }, + { + "epoch": 0.4769188394076314, + "grad_norm": 2.7718505859375, + "learning_rate": 0.00010728236103346874, + "loss": 0.2468, + "step": 20345 + }, + { + "epoch": 0.4769422809824364, + "grad_norm": 4.671642303466797, + "learning_rate": 0.00010727501441424419, + "loss": 0.7051, + "step": 20346 + }, + { + "epoch": 0.4769657225572414, + "grad_norm": 4.5234456062316895, + "learning_rate": 0.0001072676677555452, + "loss": 0.4718, + "step": 20347 + }, + { + "epoch": 0.4769891641320464, + "grad_norm": 1.6311143636703491, + "learning_rate": 0.00010726032105741161, + "loss": 0.1821, + "step": 20348 + }, + { + "epoch": 0.4770126057068514, + "grad_norm": 1.600526213645935, + "learning_rate": 0.00010725297431988337, + "loss": 0.1445, + "step": 20349 + }, + { + "epoch": 0.4770360472816564, + "grad_norm": 5.093624591827393, + "learning_rate": 0.00010724562754300027, + "loss": 0.3455, + "step": 20350 + }, + { + "epoch": 0.4770594888564614, + "grad_norm": 2.4200849533081055, + "learning_rate": 0.00010723828072680218, + "loss": 0.1827, + "step": 20351 + }, + { + "epoch": 0.47708293043126637, + "grad_norm": 4.026987552642822, + "learning_rate": 0.00010723093387132898, + "loss": 0.2706, + "step": 20352 + }, + { + "epoch": 0.47710637200607137, + "grad_norm": 1.536981225013733, + "learning_rate": 0.00010722358697662053, + "loss": 0.132, + "step": 20353 + }, + { + "epoch": 0.47712981358087636, + "grad_norm": 2.640416145324707, + "learning_rate": 0.00010721624004271671, + "loss": 1.7587, + "step": 20354 + }, + { + "epoch": 0.47715325515568136, + "grad_norm": 6.329278469085693, + "learning_rate": 0.00010720889306965734, + "loss": 0.59, + "step": 20355 + }, + { + "epoch": 0.47717669673048635, + "grad_norm": 2.57802414894104, + "learning_rate": 0.00010720154605748233, + "loss": 1.7326, + "step": 20356 + }, + { + "epoch": 0.47720013830529134, + "grad_norm": 4.63844633102417, + "learning_rate": 0.00010719419900623153, + "loss": 0.6213, + "step": 20357 + }, + { + "epoch": 0.47722357988009634, + "grad_norm": 1.027905821800232, + "learning_rate": 0.0001071868519159448, + "loss": 0.0545, + "step": 20358 + }, + { + "epoch": 0.47724702145490133, + "grad_norm": 1.9545077085494995, + "learning_rate": 0.00010717950478666201, + "loss": 0.2108, + "step": 20359 + }, + { + "epoch": 0.4772704630297063, + "grad_norm": 4.115689277648926, + "learning_rate": 0.00010717215761842303, + "loss": 0.5298, + "step": 20360 + }, + { + "epoch": 0.4772939046045113, + "grad_norm": 3.2894229888916016, + "learning_rate": 0.0001071648104112677, + "loss": 0.2842, + "step": 20361 + }, + { + "epoch": 0.4773173461793163, + "grad_norm": 0.8825594186782837, + "learning_rate": 0.00010715746316523593, + "loss": 0.0873, + "step": 20362 + }, + { + "epoch": 0.4773407877541213, + "grad_norm": 1.3625417947769165, + "learning_rate": 0.00010715011588036756, + "loss": 0.0917, + "step": 20363 + }, + { + "epoch": 0.4773642293289263, + "grad_norm": 1.3216283321380615, + "learning_rate": 0.00010714276855670245, + "loss": 0.1854, + "step": 20364 + }, + { + "epoch": 0.4773876709037313, + "grad_norm": 3.5148580074310303, + "learning_rate": 0.00010713542119428049, + "loss": 0.3349, + "step": 20365 + }, + { + "epoch": 0.4774111124785363, + "grad_norm": 5.196868896484375, + "learning_rate": 0.00010712807379314154, + "loss": 0.6834, + "step": 20366 + }, + { + "epoch": 0.4774345540533413, + "grad_norm": 1.7683842182159424, + "learning_rate": 0.00010712072635332542, + "loss": 0.1458, + "step": 20367 + }, + { + "epoch": 0.4774579956281463, + "grad_norm": 2.218818187713623, + "learning_rate": 0.00010711337887487207, + "loss": 0.2068, + "step": 20368 + }, + { + "epoch": 0.4774814372029513, + "grad_norm": 4.541383743286133, + "learning_rate": 0.00010710603135782133, + "loss": 0.3215, + "step": 20369 + }, + { + "epoch": 0.47750487877775627, + "grad_norm": 0.6002811193466187, + "learning_rate": 0.00010709868380221306, + "loss": 0.0644, + "step": 20370 + }, + { + "epoch": 0.47752832035256126, + "grad_norm": 3.6427273750305176, + "learning_rate": 0.00010709133620808714, + "loss": 2.0352, + "step": 20371 + }, + { + "epoch": 0.47755176192736626, + "grad_norm": 3.327972888946533, + "learning_rate": 0.0001070839885754834, + "loss": 0.2149, + "step": 20372 + }, + { + "epoch": 0.47757520350217125, + "grad_norm": 3.1533432006835938, + "learning_rate": 0.00010707664090444178, + "loss": 1.6868, + "step": 20373 + }, + { + "epoch": 0.47759864507697625, + "grad_norm": 2.218787908554077, + "learning_rate": 0.00010706929319500208, + "loss": 0.2606, + "step": 20374 + }, + { + "epoch": 0.47762208665178124, + "grad_norm": 1.7488658428192139, + "learning_rate": 0.0001070619454472042, + "loss": 0.1174, + "step": 20375 + }, + { + "epoch": 0.47764552822658624, + "grad_norm": 3.1308722496032715, + "learning_rate": 0.00010705459766108801, + "loss": 0.3242, + "step": 20376 + }, + { + "epoch": 0.47766896980139123, + "grad_norm": 2.0566141605377197, + "learning_rate": 0.00010704724983669339, + "loss": 0.1516, + "step": 20377 + }, + { + "epoch": 0.4776924113761963, + "grad_norm": 4.159459114074707, + "learning_rate": 0.00010703990197406016, + "loss": 0.2266, + "step": 20378 + }, + { + "epoch": 0.4777158529510013, + "grad_norm": 3.123363733291626, + "learning_rate": 0.00010703255407322826, + "loss": 0.4073, + "step": 20379 + }, + { + "epoch": 0.47773929452580627, + "grad_norm": 0.698110818862915, + "learning_rate": 0.00010702520613423747, + "loss": 0.0534, + "step": 20380 + }, + { + "epoch": 0.47776273610061126, + "grad_norm": 4.144158840179443, + "learning_rate": 0.00010701785815712777, + "loss": 1.5476, + "step": 20381 + }, + { + "epoch": 0.47778617767541626, + "grad_norm": 0.8933610320091248, + "learning_rate": 0.00010701051014193897, + "loss": 0.0762, + "step": 20382 + }, + { + "epoch": 0.47780961925022125, + "grad_norm": 2.7869937419891357, + "learning_rate": 0.00010700316208871091, + "loss": 0.2284, + "step": 20383 + }, + { + "epoch": 0.47783306082502625, + "grad_norm": 2.752466917037964, + "learning_rate": 0.00010699581399748351, + "loss": 0.3444, + "step": 20384 + }, + { + "epoch": 0.47785650239983124, + "grad_norm": 1.1080583333969116, + "learning_rate": 0.00010698846586829663, + "loss": 0.1971, + "step": 20385 + }, + { + "epoch": 0.47787994397463623, + "grad_norm": 3.238816976547241, + "learning_rate": 0.00010698111770119013, + "loss": 0.2308, + "step": 20386 + }, + { + "epoch": 0.47790338554944123, + "grad_norm": 3.021124839782715, + "learning_rate": 0.0001069737694962039, + "loss": 0.2943, + "step": 20387 + }, + { + "epoch": 0.4779268271242462, + "grad_norm": 2.9730780124664307, + "learning_rate": 0.00010696642125337781, + "loss": 0.1455, + "step": 20388 + }, + { + "epoch": 0.4779502686990512, + "grad_norm": 2.7653982639312744, + "learning_rate": 0.00010695907297275169, + "loss": 0.1826, + "step": 20389 + }, + { + "epoch": 0.4779737102738562, + "grad_norm": 2.060422658920288, + "learning_rate": 0.00010695172465436547, + "loss": 0.1084, + "step": 20390 + }, + { + "epoch": 0.4779971518486612, + "grad_norm": 0.6013646721839905, + "learning_rate": 0.000106944376298259, + "loss": 0.0559, + "step": 20391 + }, + { + "epoch": 0.4780205934234662, + "grad_norm": 4.691611289978027, + "learning_rate": 0.00010693702790447216, + "loss": 0.6691, + "step": 20392 + }, + { + "epoch": 0.4780440349982712, + "grad_norm": 3.143477439880371, + "learning_rate": 0.00010692967947304477, + "loss": 0.3061, + "step": 20393 + }, + { + "epoch": 0.4780674765730762, + "grad_norm": 3.2607011795043945, + "learning_rate": 0.00010692233100401676, + "loss": 0.3361, + "step": 20394 + }, + { + "epoch": 0.4780909181478812, + "grad_norm": 2.9012606143951416, + "learning_rate": 0.00010691498249742799, + "loss": 0.3037, + "step": 20395 + }, + { + "epoch": 0.4781143597226862, + "grad_norm": 1.9657918214797974, + "learning_rate": 0.00010690763395331833, + "loss": 0.187, + "step": 20396 + }, + { + "epoch": 0.47813780129749117, + "grad_norm": 2.529454469680786, + "learning_rate": 0.00010690028537172767, + "loss": 0.1974, + "step": 20397 + }, + { + "epoch": 0.47816124287229617, + "grad_norm": 0.7712857723236084, + "learning_rate": 0.00010689293675269584, + "loss": 0.112, + "step": 20398 + }, + { + "epoch": 0.47818468444710116, + "grad_norm": 1.9715797901153564, + "learning_rate": 0.00010688558809626277, + "loss": 0.1907, + "step": 20399 + }, + { + "epoch": 0.47820812602190615, + "grad_norm": 3.0913264751434326, + "learning_rate": 0.00010687823940246828, + "loss": 1.7817, + "step": 20400 + }, + { + "epoch": 0.47823156759671115, + "grad_norm": 2.562180757522583, + "learning_rate": 0.0001068708906713523, + "loss": 0.2111, + "step": 20401 + }, + { + "epoch": 0.47825500917151614, + "grad_norm": 3.775054454803467, + "learning_rate": 0.00010686354190295462, + "loss": 0.4776, + "step": 20402 + }, + { + "epoch": 0.47827845074632114, + "grad_norm": 3.6137964725494385, + "learning_rate": 0.00010685619309731523, + "loss": 0.2664, + "step": 20403 + }, + { + "epoch": 0.47830189232112613, + "grad_norm": 0.7655256986618042, + "learning_rate": 0.00010684884425447393, + "loss": 0.0565, + "step": 20404 + }, + { + "epoch": 0.4783253338959311, + "grad_norm": 2.647211790084839, + "learning_rate": 0.00010684149537447057, + "loss": 1.6243, + "step": 20405 + }, + { + "epoch": 0.4783487754707361, + "grad_norm": 1.4999741315841675, + "learning_rate": 0.00010683414645734511, + "loss": 0.1242, + "step": 20406 + }, + { + "epoch": 0.4783722170455411, + "grad_norm": 2.771055221557617, + "learning_rate": 0.00010682679750313735, + "loss": 0.1527, + "step": 20407 + }, + { + "epoch": 0.4783956586203461, + "grad_norm": 2.3020431995391846, + "learning_rate": 0.00010681944851188721, + "loss": 0.236, + "step": 20408 + }, + { + "epoch": 0.4784191001951511, + "grad_norm": 5.069129467010498, + "learning_rate": 0.00010681209948363453, + "loss": 0.7062, + "step": 20409 + }, + { + "epoch": 0.4784425417699561, + "grad_norm": 3.459695816040039, + "learning_rate": 0.00010680475041841922, + "loss": 0.4601, + "step": 20410 + }, + { + "epoch": 0.4784659833447611, + "grad_norm": 2.7159204483032227, + "learning_rate": 0.00010679740131628118, + "loss": 1.643, + "step": 20411 + }, + { + "epoch": 0.4784894249195661, + "grad_norm": 2.635730266571045, + "learning_rate": 0.00010679005217726019, + "loss": 0.2766, + "step": 20412 + }, + { + "epoch": 0.4785128664943711, + "grad_norm": 2.8808741569519043, + "learning_rate": 0.00010678270300139621, + "loss": 0.2358, + "step": 20413 + }, + { + "epoch": 0.4785363080691761, + "grad_norm": 2.457977056503296, + "learning_rate": 0.00010677535378872909, + "loss": 0.2507, + "step": 20414 + }, + { + "epoch": 0.47855974964398107, + "grad_norm": 1.4280693531036377, + "learning_rate": 0.0001067680045392987, + "loss": 0.1516, + "step": 20415 + }, + { + "epoch": 0.47858319121878606, + "grad_norm": 1.3678847551345825, + "learning_rate": 0.00010676065525314496, + "loss": 0.0901, + "step": 20416 + }, + { + "epoch": 0.47860663279359106, + "grad_norm": 3.0453178882598877, + "learning_rate": 0.0001067533059303077, + "loss": 0.4493, + "step": 20417 + }, + { + "epoch": 0.47863007436839605, + "grad_norm": 3.9633686542510986, + "learning_rate": 0.00010674595657082678, + "loss": 0.3226, + "step": 20418 + }, + { + "epoch": 0.47865351594320105, + "grad_norm": 4.810579776763916, + "learning_rate": 0.00010673860717474216, + "loss": 0.6423, + "step": 20419 + }, + { + "epoch": 0.47867695751800604, + "grad_norm": 2.4909486770629883, + "learning_rate": 0.00010673125774209362, + "loss": 0.2043, + "step": 20420 + }, + { + "epoch": 0.47870039909281104, + "grad_norm": 5.33693265914917, + "learning_rate": 0.0001067239082729211, + "loss": 0.3271, + "step": 20421 + }, + { + "epoch": 0.47872384066761603, + "grad_norm": 1.2052289247512817, + "learning_rate": 0.0001067165587672645, + "loss": 0.0815, + "step": 20422 + }, + { + "epoch": 0.478747282242421, + "grad_norm": 0.8258589506149292, + "learning_rate": 0.00010670920922516363, + "loss": 0.0968, + "step": 20423 + }, + { + "epoch": 0.478770723817226, + "grad_norm": 0.6856569051742554, + "learning_rate": 0.0001067018596466584, + "loss": 0.1127, + "step": 20424 + }, + { + "epoch": 0.478794165392031, + "grad_norm": 2.2303214073181152, + "learning_rate": 0.00010669451003178871, + "loss": 0.2271, + "step": 20425 + }, + { + "epoch": 0.478817606966836, + "grad_norm": 0.5756930112838745, + "learning_rate": 0.00010668716038059442, + "loss": 0.0163, + "step": 20426 + }, + { + "epoch": 0.478841048541641, + "grad_norm": 1.5842478275299072, + "learning_rate": 0.0001066798106931154, + "loss": 0.1049, + "step": 20427 + }, + { + "epoch": 0.478864490116446, + "grad_norm": 4.982247352600098, + "learning_rate": 0.00010667246096939152, + "loss": 0.3584, + "step": 20428 + }, + { + "epoch": 0.47888793169125105, + "grad_norm": 2.206986665725708, + "learning_rate": 0.00010666511120946272, + "loss": 1.2667, + "step": 20429 + }, + { + "epoch": 0.47891137326605604, + "grad_norm": 3.668194055557251, + "learning_rate": 0.00010665776141336883, + "loss": 0.3577, + "step": 20430 + }, + { + "epoch": 0.47893481484086103, + "grad_norm": 4.454441547393799, + "learning_rate": 0.0001066504115811497, + "loss": 0.3773, + "step": 20431 + }, + { + "epoch": 0.47895825641566603, + "grad_norm": 3.160780191421509, + "learning_rate": 0.00010664306171284529, + "loss": 0.218, + "step": 20432 + }, + { + "epoch": 0.478981697990471, + "grad_norm": 1.2472699880599976, + "learning_rate": 0.00010663571180849543, + "loss": 0.0746, + "step": 20433 + }, + { + "epoch": 0.479005139565276, + "grad_norm": 5.005685329437256, + "learning_rate": 0.00010662836186814001, + "loss": 0.9257, + "step": 20434 + }, + { + "epoch": 0.479028581140081, + "grad_norm": 4.017049789428711, + "learning_rate": 0.0001066210118918189, + "loss": 0.5033, + "step": 20435 + }, + { + "epoch": 0.479052022714886, + "grad_norm": 2.4896395206451416, + "learning_rate": 0.00010661366187957203, + "loss": 0.1827, + "step": 20436 + }, + { + "epoch": 0.479075464289691, + "grad_norm": 2.913757085800171, + "learning_rate": 0.00010660631183143919, + "loss": 0.3408, + "step": 20437 + }, + { + "epoch": 0.479098905864496, + "grad_norm": 4.195137023925781, + "learning_rate": 0.00010659896174746034, + "loss": 0.5544, + "step": 20438 + }, + { + "epoch": 0.479122347439301, + "grad_norm": 2.0399253368377686, + "learning_rate": 0.00010659161162767534, + "loss": 0.1569, + "step": 20439 + }, + { + "epoch": 0.479145789014106, + "grad_norm": 3.510887861251831, + "learning_rate": 0.00010658426147212408, + "loss": 0.4985, + "step": 20440 + }, + { + "epoch": 0.479169230588911, + "grad_norm": 1.0697931051254272, + "learning_rate": 0.00010657691128084644, + "loss": 0.0927, + "step": 20441 + }, + { + "epoch": 0.47919267216371597, + "grad_norm": 2.872983455657959, + "learning_rate": 0.0001065695610538823, + "loss": 0.3667, + "step": 20442 + }, + { + "epoch": 0.47921611373852097, + "grad_norm": 2.3173840045928955, + "learning_rate": 0.00010656221079127148, + "loss": 0.1711, + "step": 20443 + }, + { + "epoch": 0.47923955531332596, + "grad_norm": 4.357158660888672, + "learning_rate": 0.00010655486049305396, + "loss": 0.6545, + "step": 20444 + }, + { + "epoch": 0.47926299688813095, + "grad_norm": 2.376732349395752, + "learning_rate": 0.00010654751015926959, + "loss": 0.2936, + "step": 20445 + }, + { + "epoch": 0.47928643846293595, + "grad_norm": 4.413963317871094, + "learning_rate": 0.00010654015978995824, + "loss": 0.4378, + "step": 20446 + }, + { + "epoch": 0.47930988003774094, + "grad_norm": 1.9911442995071411, + "learning_rate": 0.0001065328093851598, + "loss": 0.1059, + "step": 20447 + }, + { + "epoch": 0.47933332161254594, + "grad_norm": 3.2676663398742676, + "learning_rate": 0.00010652545894491415, + "loss": 0.269, + "step": 20448 + }, + { + "epoch": 0.47935676318735093, + "grad_norm": 3.286079168319702, + "learning_rate": 0.00010651810846926117, + "loss": 0.4015, + "step": 20449 + }, + { + "epoch": 0.4793802047621559, + "grad_norm": 0.8418652415275574, + "learning_rate": 0.00010651075795824075, + "loss": 0.0845, + "step": 20450 + }, + { + "epoch": 0.4794036463369609, + "grad_norm": 3.8317458629608154, + "learning_rate": 0.00010650340741189278, + "loss": 0.3595, + "step": 20451 + }, + { + "epoch": 0.4794270879117659, + "grad_norm": 1.9450352191925049, + "learning_rate": 0.00010649605683025714, + "loss": 0.1914, + "step": 20452 + }, + { + "epoch": 0.4794505294865709, + "grad_norm": 3.044715404510498, + "learning_rate": 0.0001064887062133737, + "loss": 0.4032, + "step": 20453 + }, + { + "epoch": 0.4794739710613759, + "grad_norm": 3.67236590385437, + "learning_rate": 0.00010648135556128236, + "loss": 0.383, + "step": 20454 + }, + { + "epoch": 0.4794974126361809, + "grad_norm": 3.334677219390869, + "learning_rate": 0.00010647400487402303, + "loss": 0.2711, + "step": 20455 + }, + { + "epoch": 0.4795208542109859, + "grad_norm": 0.5894796252250671, + "learning_rate": 0.00010646665415163552, + "loss": 0.0334, + "step": 20456 + }, + { + "epoch": 0.4795442957857909, + "grad_norm": 2.6586861610412598, + "learning_rate": 0.00010645930339415979, + "loss": 0.1825, + "step": 20457 + }, + { + "epoch": 0.4795677373605959, + "grad_norm": 0.41684308648109436, + "learning_rate": 0.00010645195260163572, + "loss": 0.0334, + "step": 20458 + }, + { + "epoch": 0.4795911789354009, + "grad_norm": 1.3742687702178955, + "learning_rate": 0.00010644460177410314, + "loss": 0.7462, + "step": 20459 + }, + { + "epoch": 0.47961462051020587, + "grad_norm": 5.271636486053467, + "learning_rate": 0.00010643725091160197, + "loss": 0.4429, + "step": 20460 + }, + { + "epoch": 0.47963806208501086, + "grad_norm": 2.715663433074951, + "learning_rate": 0.00010642990001417212, + "loss": 0.3552, + "step": 20461 + }, + { + "epoch": 0.47966150365981586, + "grad_norm": 0.8258357048034668, + "learning_rate": 0.00010642254908185342, + "loss": 0.1281, + "step": 20462 + }, + { + "epoch": 0.47968494523462085, + "grad_norm": 5.986030578613281, + "learning_rate": 0.0001064151981146858, + "loss": 0.6686, + "step": 20463 + }, + { + "epoch": 0.47970838680942585, + "grad_norm": 0.8235607743263245, + "learning_rate": 0.00010640784711270918, + "loss": 0.0894, + "step": 20464 + }, + { + "epoch": 0.47973182838423084, + "grad_norm": 6.211448669433594, + "learning_rate": 0.00010640049607596336, + "loss": 0.4937, + "step": 20465 + }, + { + "epoch": 0.47975526995903583, + "grad_norm": 3.0232484340667725, + "learning_rate": 0.00010639314500448824, + "loss": 0.3723, + "step": 20466 + }, + { + "epoch": 0.47977871153384083, + "grad_norm": 1.303349256515503, + "learning_rate": 0.0001063857938983238, + "loss": 0.0973, + "step": 20467 + }, + { + "epoch": 0.4798021531086458, + "grad_norm": 4.217250823974609, + "learning_rate": 0.00010637844275750983, + "loss": 0.3631, + "step": 20468 + }, + { + "epoch": 0.4798255946834508, + "grad_norm": 1.839416742324829, + "learning_rate": 0.00010637109158208622, + "loss": 1.4452, + "step": 20469 + }, + { + "epoch": 0.4798490362582558, + "grad_norm": 2.029984712600708, + "learning_rate": 0.00010636374037209294, + "loss": 0.2138, + "step": 20470 + }, + { + "epoch": 0.4798724778330608, + "grad_norm": 3.205366373062134, + "learning_rate": 0.0001063563891275698, + "loss": 1.5749, + "step": 20471 + }, + { + "epoch": 0.4798959194078658, + "grad_norm": 3.209743022918701, + "learning_rate": 0.0001063490378485567, + "loss": 0.3878, + "step": 20472 + }, + { + "epoch": 0.4799193609826708, + "grad_norm": 0.875396192073822, + "learning_rate": 0.00010634168653509358, + "loss": 0.0515, + "step": 20473 + }, + { + "epoch": 0.4799428025574758, + "grad_norm": 2.9815664291381836, + "learning_rate": 0.00010633433518722027, + "loss": 0.2615, + "step": 20474 + }, + { + "epoch": 0.4799662441322808, + "grad_norm": 1.203119158744812, + "learning_rate": 0.00010632698380497665, + "loss": 0.0699, + "step": 20475 + }, + { + "epoch": 0.4799896857070858, + "grad_norm": 3.9183459281921387, + "learning_rate": 0.00010631963238840268, + "loss": 0.3448, + "step": 20476 + }, + { + "epoch": 0.4800131272818908, + "grad_norm": 5.614096164703369, + "learning_rate": 0.0001063122809375382, + "loss": 0.2376, + "step": 20477 + }, + { + "epoch": 0.48003656885669577, + "grad_norm": 3.193469762802124, + "learning_rate": 0.0001063049294524231, + "loss": 0.1657, + "step": 20478 + }, + { + "epoch": 0.48006001043150076, + "grad_norm": 2.1739182472229004, + "learning_rate": 0.00010629757793309727, + "loss": 0.2065, + "step": 20479 + }, + { + "epoch": 0.4800834520063058, + "grad_norm": 3.1580848693847656, + "learning_rate": 0.00010629022637960063, + "loss": 0.2335, + "step": 20480 + }, + { + "epoch": 0.4801068935811108, + "grad_norm": 3.1913254261016846, + "learning_rate": 0.00010628287479197301, + "loss": 0.3406, + "step": 20481 + }, + { + "epoch": 0.4801303351559158, + "grad_norm": 2.9142162799835205, + "learning_rate": 0.00010627552317025433, + "loss": 0.2474, + "step": 20482 + }, + { + "epoch": 0.4801537767307208, + "grad_norm": 2.0952539443969727, + "learning_rate": 0.00010626817151448452, + "loss": 0.1595, + "step": 20483 + }, + { + "epoch": 0.4801772183055258, + "grad_norm": 0.743602991104126, + "learning_rate": 0.00010626081982470343, + "loss": 0.0476, + "step": 20484 + }, + { + "epoch": 0.4802006598803308, + "grad_norm": 2.1834161281585693, + "learning_rate": 0.00010625346810095093, + "loss": 0.1689, + "step": 20485 + }, + { + "epoch": 0.4802241014551358, + "grad_norm": 3.476794481277466, + "learning_rate": 0.00010624611634326698, + "loss": 1.5271, + "step": 20486 + }, + { + "epoch": 0.48024754302994077, + "grad_norm": 1.3555423021316528, + "learning_rate": 0.00010623876455169137, + "loss": 0.1081, + "step": 20487 + }, + { + "epoch": 0.48027098460474577, + "grad_norm": 4.133632659912109, + "learning_rate": 0.00010623141272626407, + "loss": 0.6322, + "step": 20488 + }, + { + "epoch": 0.48029442617955076, + "grad_norm": 3.4922330379486084, + "learning_rate": 0.00010622406086702497, + "loss": 0.4392, + "step": 20489 + }, + { + "epoch": 0.48031786775435575, + "grad_norm": 4.220193386077881, + "learning_rate": 0.00010621670897401395, + "loss": 0.5685, + "step": 20490 + }, + { + "epoch": 0.48034130932916075, + "grad_norm": 1.921938180923462, + "learning_rate": 0.00010620935704727084, + "loss": 0.1497, + "step": 20491 + }, + { + "epoch": 0.48036475090396574, + "grad_norm": 2.230239152908325, + "learning_rate": 0.00010620200508683564, + "loss": 0.2341, + "step": 20492 + }, + { + "epoch": 0.48038819247877074, + "grad_norm": 3.2504055500030518, + "learning_rate": 0.00010619465309274814, + "loss": 0.297, + "step": 20493 + }, + { + "epoch": 0.48041163405357573, + "grad_norm": 2.4058260917663574, + "learning_rate": 0.00010618730106504829, + "loss": 0.3061, + "step": 20494 + }, + { + "epoch": 0.4804350756283807, + "grad_norm": 4.219265937805176, + "learning_rate": 0.000106179949003776, + "loss": 0.4114, + "step": 20495 + }, + { + "epoch": 0.4804585172031857, + "grad_norm": 4.885930061340332, + "learning_rate": 0.0001061725969089711, + "loss": 0.2919, + "step": 20496 + }, + { + "epoch": 0.4804819587779907, + "grad_norm": 2.5399010181427, + "learning_rate": 0.00010616524478067351, + "loss": 0.2944, + "step": 20497 + }, + { + "epoch": 0.4805054003527957, + "grad_norm": 1.496278166770935, + "learning_rate": 0.00010615789261892317, + "loss": 0.1407, + "step": 20498 + }, + { + "epoch": 0.4805288419276007, + "grad_norm": 3.3390848636627197, + "learning_rate": 0.00010615054042375989, + "loss": 0.3574, + "step": 20499 + }, + { + "epoch": 0.4805522835024057, + "grad_norm": 1.790201187133789, + "learning_rate": 0.00010614318819522364, + "loss": 0.1165, + "step": 20500 + }, + { + "epoch": 0.4805757250772107, + "grad_norm": 4.150493621826172, + "learning_rate": 0.00010613583593335423, + "loss": 0.3005, + "step": 20501 + }, + { + "epoch": 0.4805991666520157, + "grad_norm": 7.2060933113098145, + "learning_rate": 0.00010612848363819165, + "loss": 0.6405, + "step": 20502 + }, + { + "epoch": 0.4806226082268207, + "grad_norm": 2.1435434818267822, + "learning_rate": 0.00010612113130977575, + "loss": 0.1337, + "step": 20503 + }, + { + "epoch": 0.4806460498016257, + "grad_norm": 2.703213930130005, + "learning_rate": 0.00010611377894814639, + "loss": 0.2501, + "step": 20504 + }, + { + "epoch": 0.48066949137643067, + "grad_norm": 3.7254717350006104, + "learning_rate": 0.00010610642655334346, + "loss": 0.4661, + "step": 20505 + }, + { + "epoch": 0.48069293295123566, + "grad_norm": 3.483898401260376, + "learning_rate": 0.00010609907412540696, + "loss": 0.3882, + "step": 20506 + }, + { + "epoch": 0.48071637452604066, + "grad_norm": 3.3807828426361084, + "learning_rate": 0.00010609172166437667, + "loss": 0.2138, + "step": 20507 + }, + { + "epoch": 0.48073981610084565, + "grad_norm": 4.910810947418213, + "learning_rate": 0.00010608436917029254, + "loss": 0.4832, + "step": 20508 + }, + { + "epoch": 0.48076325767565065, + "grad_norm": 2.1650421619415283, + "learning_rate": 0.00010607701664319445, + "loss": 0.2656, + "step": 20509 + }, + { + "epoch": 0.48078669925045564, + "grad_norm": 7.234731197357178, + "learning_rate": 0.00010606966408312229, + "loss": 0.5586, + "step": 20510 + }, + { + "epoch": 0.48081014082526063, + "grad_norm": 2.9303767681121826, + "learning_rate": 0.00010606231149011596, + "loss": 0.2689, + "step": 20511 + }, + { + "epoch": 0.48083358240006563, + "grad_norm": 5.066576957702637, + "learning_rate": 0.00010605495886421538, + "loss": 0.3949, + "step": 20512 + }, + { + "epoch": 0.4808570239748706, + "grad_norm": 2.323430299758911, + "learning_rate": 0.00010604760620546039, + "loss": 0.1787, + "step": 20513 + }, + { + "epoch": 0.4808804655496756, + "grad_norm": 5.729613304138184, + "learning_rate": 0.00010604025351389096, + "loss": 0.4488, + "step": 20514 + }, + { + "epoch": 0.4809039071244806, + "grad_norm": 1.8743138313293457, + "learning_rate": 0.00010603290078954691, + "loss": 0.2161, + "step": 20515 + }, + { + "epoch": 0.4809273486992856, + "grad_norm": 3.222964286804199, + "learning_rate": 0.00010602554803246818, + "loss": 0.4384, + "step": 20516 + }, + { + "epoch": 0.4809507902740906, + "grad_norm": 1.4282227754592896, + "learning_rate": 0.00010601819524269465, + "loss": 0.8089, + "step": 20517 + }, + { + "epoch": 0.4809742318488956, + "grad_norm": 1.8816813230514526, + "learning_rate": 0.00010601084242026625, + "loss": 0.1472, + "step": 20518 + }, + { + "epoch": 0.4809976734237006, + "grad_norm": 2.1006197929382324, + "learning_rate": 0.00010600348956522284, + "loss": 0.9402, + "step": 20519 + }, + { + "epoch": 0.4810211149985056, + "grad_norm": 2.8016960620880127, + "learning_rate": 0.00010599613667760429, + "loss": 0.3146, + "step": 20520 + }, + { + "epoch": 0.4810445565733106, + "grad_norm": 1.335279107093811, + "learning_rate": 0.00010598878375745058, + "loss": 0.0723, + "step": 20521 + }, + { + "epoch": 0.48106799814811557, + "grad_norm": 3.7092299461364746, + "learning_rate": 0.00010598143080480157, + "loss": 0.3746, + "step": 20522 + }, + { + "epoch": 0.48109143972292057, + "grad_norm": 3.0130081176757812, + "learning_rate": 0.00010597407781969706, + "loss": 0.2144, + "step": 20523 + }, + { + "epoch": 0.48111488129772556, + "grad_norm": 5.983561038970947, + "learning_rate": 0.0001059667248021771, + "loss": 0.2306, + "step": 20524 + }, + { + "epoch": 0.48113832287253055, + "grad_norm": 1.8856379985809326, + "learning_rate": 0.00010595937175228154, + "loss": 0.2331, + "step": 20525 + }, + { + "epoch": 0.48116176444733555, + "grad_norm": 1.3902679681777954, + "learning_rate": 0.00010595201867005025, + "loss": 0.1491, + "step": 20526 + }, + { + "epoch": 0.48118520602214054, + "grad_norm": 2.9808716773986816, + "learning_rate": 0.00010594466555552314, + "loss": 0.2255, + "step": 20527 + }, + { + "epoch": 0.48120864759694554, + "grad_norm": 5.099009990692139, + "learning_rate": 0.0001059373124087401, + "loss": 0.3379, + "step": 20528 + }, + { + "epoch": 0.48123208917175053, + "grad_norm": 3.037760019302368, + "learning_rate": 0.00010592995922974103, + "loss": 0.3694, + "step": 20529 + }, + { + "epoch": 0.4812555307465555, + "grad_norm": 1.6026794910430908, + "learning_rate": 0.00010592260601856583, + "loss": 0.1637, + "step": 20530 + }, + { + "epoch": 0.4812789723213606, + "grad_norm": 1.087776780128479, + "learning_rate": 0.00010591525277525443, + "loss": 0.1273, + "step": 20531 + }, + { + "epoch": 0.48130241389616557, + "grad_norm": 3.4870965480804443, + "learning_rate": 0.00010590789949984665, + "loss": 0.3664, + "step": 20532 + }, + { + "epoch": 0.48132585547097056, + "grad_norm": 2.0805516242980957, + "learning_rate": 0.0001059005461923825, + "loss": 0.1447, + "step": 20533 + }, + { + "epoch": 0.48134929704577556, + "grad_norm": 4.293622016906738, + "learning_rate": 0.0001058931928529018, + "loss": 0.3929, + "step": 20534 + }, + { + "epoch": 0.48137273862058055, + "grad_norm": 2.078943967819214, + "learning_rate": 0.00010588583948144446, + "loss": 0.1733, + "step": 20535 + }, + { + "epoch": 0.48139618019538555, + "grad_norm": 0.5915123820304871, + "learning_rate": 0.00010587848607805037, + "loss": 0.0551, + "step": 20536 + }, + { + "epoch": 0.48141962177019054, + "grad_norm": 1.3061833381652832, + "learning_rate": 0.00010587113264275949, + "loss": 0.934, + "step": 20537 + }, + { + "epoch": 0.48144306334499554, + "grad_norm": 0.796501874923706, + "learning_rate": 0.00010586377917561167, + "loss": 0.0845, + "step": 20538 + }, + { + "epoch": 0.48146650491980053, + "grad_norm": 4.83420991897583, + "learning_rate": 0.00010585642567664682, + "loss": 0.5365, + "step": 20539 + }, + { + "epoch": 0.4814899464946055, + "grad_norm": 5.799541473388672, + "learning_rate": 0.00010584907214590485, + "loss": 0.9398, + "step": 20540 + }, + { + "epoch": 0.4815133880694105, + "grad_norm": 3.7544636726379395, + "learning_rate": 0.00010584171858342564, + "loss": 0.5884, + "step": 20541 + }, + { + "epoch": 0.4815368296442155, + "grad_norm": 0.420780748128891, + "learning_rate": 0.00010583436498924906, + "loss": 0.0394, + "step": 20542 + }, + { + "epoch": 0.4815602712190205, + "grad_norm": 3.668727159500122, + "learning_rate": 0.00010582701136341508, + "loss": 0.4367, + "step": 20543 + }, + { + "epoch": 0.4815837127938255, + "grad_norm": 3.2813878059387207, + "learning_rate": 0.00010581965770596361, + "loss": 0.2244, + "step": 20544 + }, + { + "epoch": 0.4816071543686305, + "grad_norm": 3.24021053314209, + "learning_rate": 0.00010581230401693447, + "loss": 0.4385, + "step": 20545 + }, + { + "epoch": 0.4816305959434355, + "grad_norm": 3.4912495613098145, + "learning_rate": 0.00010580495029636762, + "loss": 0.3531, + "step": 20546 + }, + { + "epoch": 0.4816540375182405, + "grad_norm": 3.2883496284484863, + "learning_rate": 0.00010579759654430295, + "loss": 0.2607, + "step": 20547 + }, + { + "epoch": 0.4816774790930455, + "grad_norm": 2.447338581085205, + "learning_rate": 0.00010579024276078034, + "loss": 0.3068, + "step": 20548 + }, + { + "epoch": 0.4817009206678505, + "grad_norm": 1.1437913179397583, + "learning_rate": 0.00010578288894583972, + "loss": 0.0983, + "step": 20549 + }, + { + "epoch": 0.48172436224265547, + "grad_norm": 4.328986644744873, + "learning_rate": 0.00010577553509952098, + "loss": 0.6819, + "step": 20550 + }, + { + "epoch": 0.48174780381746046, + "grad_norm": 2.107494354248047, + "learning_rate": 0.00010576818122186402, + "loss": 0.0689, + "step": 20551 + }, + { + "epoch": 0.48177124539226546, + "grad_norm": 3.796344757080078, + "learning_rate": 0.00010576082731290878, + "loss": 0.3475, + "step": 20552 + }, + { + "epoch": 0.48179468696707045, + "grad_norm": 3.6591436862945557, + "learning_rate": 0.0001057534733726951, + "loss": 0.4719, + "step": 20553 + }, + { + "epoch": 0.48181812854187545, + "grad_norm": 4.87919807434082, + "learning_rate": 0.00010574611940126291, + "loss": 0.4589, + "step": 20554 + }, + { + "epoch": 0.48184157011668044, + "grad_norm": 3.0123302936553955, + "learning_rate": 0.0001057387653986521, + "loss": 0.2697, + "step": 20555 + }, + { + "epoch": 0.48186501169148543, + "grad_norm": 2.381990909576416, + "learning_rate": 0.00010573141136490261, + "loss": 0.2426, + "step": 20556 + }, + { + "epoch": 0.48188845326629043, + "grad_norm": 3.0095877647399902, + "learning_rate": 0.00010572405730005434, + "loss": 0.4254, + "step": 20557 + }, + { + "epoch": 0.4819118948410954, + "grad_norm": 1.1292085647583008, + "learning_rate": 0.00010571670320414712, + "loss": 0.1573, + "step": 20558 + }, + { + "epoch": 0.4819353364159004, + "grad_norm": 1.1316975355148315, + "learning_rate": 0.00010570934907722096, + "loss": 0.1142, + "step": 20559 + }, + { + "epoch": 0.4819587779907054, + "grad_norm": 2.406688690185547, + "learning_rate": 0.0001057019949193157, + "loss": 1.1875, + "step": 20560 + }, + { + "epoch": 0.4819822195655104, + "grad_norm": 7.495935916900635, + "learning_rate": 0.00010569464073047121, + "loss": 0.5809, + "step": 20561 + }, + { + "epoch": 0.4820056611403154, + "grad_norm": 2.513321876525879, + "learning_rate": 0.0001056872865107275, + "loss": 1.476, + "step": 20562 + }, + { + "epoch": 0.4820291027151204, + "grad_norm": 1.485248327255249, + "learning_rate": 0.0001056799322601244, + "loss": 0.0971, + "step": 20563 + }, + { + "epoch": 0.4820525442899254, + "grad_norm": 3.3147287368774414, + "learning_rate": 0.0001056725779787018, + "loss": 0.3353, + "step": 20564 + }, + { + "epoch": 0.4820759858647304, + "grad_norm": 3.8263185024261475, + "learning_rate": 0.00010566522366649967, + "loss": 0.2507, + "step": 20565 + }, + { + "epoch": 0.4820994274395354, + "grad_norm": 4.340149402618408, + "learning_rate": 0.00010565786932355784, + "loss": 0.1537, + "step": 20566 + }, + { + "epoch": 0.48212286901434037, + "grad_norm": 4.103359699249268, + "learning_rate": 0.00010565051494991627, + "loss": 0.3094, + "step": 20567 + }, + { + "epoch": 0.48214631058914537, + "grad_norm": 1.2662919759750366, + "learning_rate": 0.00010564316054561483, + "loss": 0.1594, + "step": 20568 + }, + { + "epoch": 0.48216975216395036, + "grad_norm": 1.2785269021987915, + "learning_rate": 0.00010563580611069348, + "loss": 0.1237, + "step": 20569 + }, + { + "epoch": 0.48219319373875535, + "grad_norm": 5.316470623016357, + "learning_rate": 0.00010562845164519206, + "loss": 0.4606, + "step": 20570 + }, + { + "epoch": 0.48221663531356035, + "grad_norm": 2.522993564605713, + "learning_rate": 0.00010562109714915051, + "loss": 1.5305, + "step": 20571 + }, + { + "epoch": 0.48224007688836534, + "grad_norm": 1.673467755317688, + "learning_rate": 0.00010561374262260872, + "loss": 0.1291, + "step": 20572 + }, + { + "epoch": 0.48226351846317034, + "grad_norm": 2.8575310707092285, + "learning_rate": 0.0001056063880656066, + "loss": 1.3508, + "step": 20573 + }, + { + "epoch": 0.48228696003797533, + "grad_norm": 1.0359820127487183, + "learning_rate": 0.00010559903347818407, + "loss": 0.0714, + "step": 20574 + }, + { + "epoch": 0.4823104016127803, + "grad_norm": 0.48168379068374634, + "learning_rate": 0.00010559167886038104, + "loss": 0.033, + "step": 20575 + }, + { + "epoch": 0.4823338431875853, + "grad_norm": 2.8590500354766846, + "learning_rate": 0.0001055843242122374, + "loss": 0.2329, + "step": 20576 + }, + { + "epoch": 0.4823572847623903, + "grad_norm": 2.2272651195526123, + "learning_rate": 0.00010557696953379306, + "loss": 0.2742, + "step": 20577 + }, + { + "epoch": 0.4823807263371953, + "grad_norm": 2.029675006866455, + "learning_rate": 0.00010556961482508789, + "loss": 0.2338, + "step": 20578 + }, + { + "epoch": 0.4824041679120003, + "grad_norm": 1.05936598777771, + "learning_rate": 0.00010556226008616188, + "loss": 0.0427, + "step": 20579 + }, + { + "epoch": 0.4824276094868053, + "grad_norm": 4.58821964263916, + "learning_rate": 0.00010555490531705486, + "loss": 0.3497, + "step": 20580 + }, + { + "epoch": 0.4824510510616103, + "grad_norm": 4.8889689445495605, + "learning_rate": 0.00010554755051780678, + "loss": 0.2944, + "step": 20581 + }, + { + "epoch": 0.48247449263641534, + "grad_norm": 4.043626308441162, + "learning_rate": 0.00010554019568845757, + "loss": 0.323, + "step": 20582 + }, + { + "epoch": 0.48249793421122034, + "grad_norm": 5.9043965339660645, + "learning_rate": 0.00010553284082904705, + "loss": 0.6255, + "step": 20583 + }, + { + "epoch": 0.48252137578602533, + "grad_norm": 4.266034126281738, + "learning_rate": 0.00010552548593961517, + "loss": 0.249, + "step": 20584 + }, + { + "epoch": 0.4825448173608303, + "grad_norm": 3.450563669204712, + "learning_rate": 0.00010551813102020189, + "loss": 0.2369, + "step": 20585 + }, + { + "epoch": 0.4825682589356353, + "grad_norm": 3.338728427886963, + "learning_rate": 0.00010551077607084707, + "loss": 0.3433, + "step": 20586 + }, + { + "epoch": 0.4825917005104403, + "grad_norm": 2.837123394012451, + "learning_rate": 0.00010550342109159062, + "loss": 0.2657, + "step": 20587 + }, + { + "epoch": 0.4826151420852453, + "grad_norm": 3.7453386783599854, + "learning_rate": 0.00010549606608247244, + "loss": 0.2574, + "step": 20588 + }, + { + "epoch": 0.4826385836600503, + "grad_norm": 3.0631117820739746, + "learning_rate": 0.00010548871104353246, + "loss": 0.2334, + "step": 20589 + }, + { + "epoch": 0.4826620252348553, + "grad_norm": 4.051828861236572, + "learning_rate": 0.00010548135597481057, + "loss": 0.4011, + "step": 20590 + }, + { + "epoch": 0.4826854668096603, + "grad_norm": 3.168184757232666, + "learning_rate": 0.00010547400087634668, + "loss": 0.2909, + "step": 20591 + }, + { + "epoch": 0.4827089083844653, + "grad_norm": 4.231868743896484, + "learning_rate": 0.00010546664574818075, + "loss": 0.7033, + "step": 20592 + }, + { + "epoch": 0.4827323499592703, + "grad_norm": 2.4934885501861572, + "learning_rate": 0.00010545929059035262, + "loss": 0.2349, + "step": 20593 + }, + { + "epoch": 0.4827557915340753, + "grad_norm": 4.459707736968994, + "learning_rate": 0.00010545193540290223, + "loss": 0.5831, + "step": 20594 + }, + { + "epoch": 0.48277923310888027, + "grad_norm": 2.7230021953582764, + "learning_rate": 0.0001054445801858695, + "loss": 1.4934, + "step": 20595 + }, + { + "epoch": 0.48280267468368526, + "grad_norm": 5.9046630859375, + "learning_rate": 0.00010543722493929427, + "loss": 0.8534, + "step": 20596 + }, + { + "epoch": 0.48282611625849026, + "grad_norm": 1.5756360292434692, + "learning_rate": 0.00010542986966321655, + "loss": 0.7936, + "step": 20597 + }, + { + "epoch": 0.48284955783329525, + "grad_norm": 5.238986492156982, + "learning_rate": 0.00010542251435767621, + "loss": 0.2067, + "step": 20598 + }, + { + "epoch": 0.48287299940810025, + "grad_norm": 4.7036237716674805, + "learning_rate": 0.00010541515902271312, + "loss": 0.565, + "step": 20599 + }, + { + "epoch": 0.48289644098290524, + "grad_norm": 4.987478256225586, + "learning_rate": 0.00010540780365836726, + "loss": 0.5229, + "step": 20600 + }, + { + "epoch": 0.48291988255771023, + "grad_norm": 2.7219300270080566, + "learning_rate": 0.0001054004482646785, + "loss": 0.3295, + "step": 20601 + }, + { + "epoch": 0.48294332413251523, + "grad_norm": 3.843095064163208, + "learning_rate": 0.00010539309284168673, + "loss": 0.4421, + "step": 20602 + }, + { + "epoch": 0.4829667657073202, + "grad_norm": 0.9710829854011536, + "learning_rate": 0.0001053857373894319, + "loss": 0.0829, + "step": 20603 + }, + { + "epoch": 0.4829902072821252, + "grad_norm": 3.2284393310546875, + "learning_rate": 0.0001053783819079539, + "loss": 0.3648, + "step": 20604 + }, + { + "epoch": 0.4830136488569302, + "grad_norm": 2.228870391845703, + "learning_rate": 0.00010537102639729268, + "loss": 2.1455, + "step": 20605 + }, + { + "epoch": 0.4830370904317352, + "grad_norm": 2.341390371322632, + "learning_rate": 0.00010536367085748808, + "loss": 0.2404, + "step": 20606 + }, + { + "epoch": 0.4830605320065402, + "grad_norm": 1.3515363931655884, + "learning_rate": 0.0001053563152885801, + "loss": 0.2057, + "step": 20607 + }, + { + "epoch": 0.4830839735813452, + "grad_norm": 3.1527462005615234, + "learning_rate": 0.00010534895969060854, + "loss": 0.3531, + "step": 20608 + }, + { + "epoch": 0.4831074151561502, + "grad_norm": 1.880509376525879, + "learning_rate": 0.0001053416040636134, + "loss": 0.2255, + "step": 20609 + }, + { + "epoch": 0.4831308567309552, + "grad_norm": 3.180264949798584, + "learning_rate": 0.0001053342484076346, + "loss": 0.322, + "step": 20610 + }, + { + "epoch": 0.4831542983057602, + "grad_norm": 3.8101844787597656, + "learning_rate": 0.00010532689272271198, + "loss": 0.2785, + "step": 20611 + }, + { + "epoch": 0.48317773988056517, + "grad_norm": 3.9195632934570312, + "learning_rate": 0.0001053195370088855, + "loss": 0.2556, + "step": 20612 + }, + { + "epoch": 0.48320118145537017, + "grad_norm": 3.3514652252197266, + "learning_rate": 0.00010531218126619506, + "loss": 0.2561, + "step": 20613 + }, + { + "epoch": 0.48322462303017516, + "grad_norm": 3.5633251667022705, + "learning_rate": 0.00010530482549468062, + "loss": 0.2295, + "step": 20614 + }, + { + "epoch": 0.48324806460498015, + "grad_norm": 2.839850902557373, + "learning_rate": 0.00010529746969438197, + "loss": 0.1826, + "step": 20615 + }, + { + "epoch": 0.48327150617978515, + "grad_norm": 0.7569644451141357, + "learning_rate": 0.00010529011386533913, + "loss": 0.085, + "step": 20616 + }, + { + "epoch": 0.48329494775459014, + "grad_norm": 2.641268014907837, + "learning_rate": 0.000105282758007592, + "loss": 1.72, + "step": 20617 + }, + { + "epoch": 0.48331838932939514, + "grad_norm": 2.3804478645324707, + "learning_rate": 0.00010527540212118045, + "loss": 0.2281, + "step": 20618 + }, + { + "epoch": 0.48334183090420013, + "grad_norm": 1.4983640909194946, + "learning_rate": 0.00010526804620614446, + "loss": 0.1972, + "step": 20619 + }, + { + "epoch": 0.4833652724790051, + "grad_norm": 3.434889554977417, + "learning_rate": 0.0001052606902625239, + "loss": 0.3795, + "step": 20620 + }, + { + "epoch": 0.4833887140538101, + "grad_norm": 2.893359422683716, + "learning_rate": 0.00010525333429035864, + "loss": 0.2541, + "step": 20621 + }, + { + "epoch": 0.4834121556286151, + "grad_norm": 0.9590392112731934, + "learning_rate": 0.00010524597828968865, + "loss": 0.0864, + "step": 20622 + }, + { + "epoch": 0.4834355972034201, + "grad_norm": 4.90975284576416, + "learning_rate": 0.00010523862226055387, + "loss": 0.613, + "step": 20623 + }, + { + "epoch": 0.4834590387782251, + "grad_norm": 4.4325642585754395, + "learning_rate": 0.00010523126620299417, + "loss": 0.4989, + "step": 20624 + }, + { + "epoch": 0.4834824803530301, + "grad_norm": 1.0640790462493896, + "learning_rate": 0.00010522391011704946, + "loss": 0.1206, + "step": 20625 + }, + { + "epoch": 0.4835059219278351, + "grad_norm": 3.8149454593658447, + "learning_rate": 0.00010521655400275967, + "loss": 0.2894, + "step": 20626 + }, + { + "epoch": 0.4835293635026401, + "grad_norm": 0.42617717385292053, + "learning_rate": 0.0001052091978601647, + "loss": 0.0258, + "step": 20627 + }, + { + "epoch": 0.4835528050774451, + "grad_norm": 0.3864450752735138, + "learning_rate": 0.00010520184168930447, + "loss": 0.0416, + "step": 20628 + }, + { + "epoch": 0.4835762466522501, + "grad_norm": 3.482276678085327, + "learning_rate": 0.00010519448549021893, + "loss": 0.2964, + "step": 20629 + }, + { + "epoch": 0.48359968822705507, + "grad_norm": 5.3323469161987305, + "learning_rate": 0.00010518712926294796, + "loss": 0.5087, + "step": 20630 + }, + { + "epoch": 0.48362312980186006, + "grad_norm": 2.9968459606170654, + "learning_rate": 0.00010517977300753146, + "loss": 1.53, + "step": 20631 + }, + { + "epoch": 0.48364657137666506, + "grad_norm": 4.20368766784668, + "learning_rate": 0.00010517241672400939, + "loss": 0.5369, + "step": 20632 + }, + { + "epoch": 0.4836700129514701, + "grad_norm": 3.308351993560791, + "learning_rate": 0.00010516506041242162, + "loss": 0.2472, + "step": 20633 + }, + { + "epoch": 0.4836934545262751, + "grad_norm": 1.7602530717849731, + "learning_rate": 0.0001051577040728081, + "loss": 1.0773, + "step": 20634 + }, + { + "epoch": 0.4837168961010801, + "grad_norm": 1.0570801496505737, + "learning_rate": 0.00010515034770520874, + "loss": 0.135, + "step": 20635 + }, + { + "epoch": 0.4837403376758851, + "grad_norm": 2.2295939922332764, + "learning_rate": 0.00010514299130966346, + "loss": 0.1311, + "step": 20636 + }, + { + "epoch": 0.4837637792506901, + "grad_norm": 4.281958103179932, + "learning_rate": 0.00010513563488621211, + "loss": 0.3989, + "step": 20637 + }, + { + "epoch": 0.4837872208254951, + "grad_norm": 3.9019434452056885, + "learning_rate": 0.0001051282784348947, + "loss": 0.3139, + "step": 20638 + }, + { + "epoch": 0.4838106624003001, + "grad_norm": 4.302525520324707, + "learning_rate": 0.00010512092195575109, + "loss": 0.3627, + "step": 20639 + }, + { + "epoch": 0.48383410397510507, + "grad_norm": 2.4165191650390625, + "learning_rate": 0.00010511356544882125, + "loss": 0.2168, + "step": 20640 + }, + { + "epoch": 0.48385754554991006, + "grad_norm": 3.094391345977783, + "learning_rate": 0.00010510620891414502, + "loss": 0.2267, + "step": 20641 + }, + { + "epoch": 0.48388098712471506, + "grad_norm": 4.840278148651123, + "learning_rate": 0.00010509885235176237, + "loss": 0.3904, + "step": 20642 + }, + { + "epoch": 0.48390442869952005, + "grad_norm": 2.35614275932312, + "learning_rate": 0.00010509149576171322, + "loss": 0.2608, + "step": 20643 + }, + { + "epoch": 0.48392787027432504, + "grad_norm": 2.072409152984619, + "learning_rate": 0.00010508413914403745, + "loss": 0.1994, + "step": 20644 + }, + { + "epoch": 0.48395131184913004, + "grad_norm": 0.4131857752799988, + "learning_rate": 0.000105076782498775, + "loss": 0.0507, + "step": 20645 + }, + { + "epoch": 0.48397475342393503, + "grad_norm": 1.0625261068344116, + "learning_rate": 0.00010506942582596579, + "loss": 0.054, + "step": 20646 + }, + { + "epoch": 0.48399819499874003, + "grad_norm": 3.957082748413086, + "learning_rate": 0.00010506206912564973, + "loss": 0.4354, + "step": 20647 + }, + { + "epoch": 0.484021636573545, + "grad_norm": 3.1599526405334473, + "learning_rate": 0.00010505471239786675, + "loss": 0.3675, + "step": 20648 + }, + { + "epoch": 0.48404507814835, + "grad_norm": 3.167069911956787, + "learning_rate": 0.00010504735564265675, + "loss": 0.3676, + "step": 20649 + }, + { + "epoch": 0.484068519723155, + "grad_norm": 2.9690229892730713, + "learning_rate": 0.00010503999886005967, + "loss": 0.367, + "step": 20650 + }, + { + "epoch": 0.48409196129796, + "grad_norm": 3.87804913520813, + "learning_rate": 0.00010503264205011538, + "loss": 0.3845, + "step": 20651 + }, + { + "epoch": 0.484115402872765, + "grad_norm": 4.829841613769531, + "learning_rate": 0.00010502528521286386, + "loss": 0.4955, + "step": 20652 + }, + { + "epoch": 0.48413884444757, + "grad_norm": 0.9470479488372803, + "learning_rate": 0.00010501792834834499, + "loss": 0.0345, + "step": 20653 + }, + { + "epoch": 0.484162286022375, + "grad_norm": 2.3126251697540283, + "learning_rate": 0.00010501057145659872, + "loss": 0.2764, + "step": 20654 + }, + { + "epoch": 0.48418572759718, + "grad_norm": 3.2478675842285156, + "learning_rate": 0.00010500321453766495, + "loss": 0.3197, + "step": 20655 + }, + { + "epoch": 0.484209169171985, + "grad_norm": 2.582404375076294, + "learning_rate": 0.00010499585759158357, + "loss": 0.1699, + "step": 20656 + }, + { + "epoch": 0.48423261074678997, + "grad_norm": 4.185058116912842, + "learning_rate": 0.00010498850061839451, + "loss": 0.3059, + "step": 20657 + }, + { + "epoch": 0.48425605232159497, + "grad_norm": 3.6074330806732178, + "learning_rate": 0.00010498114361813776, + "loss": 0.3279, + "step": 20658 + }, + { + "epoch": 0.48427949389639996, + "grad_norm": 1.1706384420394897, + "learning_rate": 0.00010497378659085315, + "loss": 0.0572, + "step": 20659 + }, + { + "epoch": 0.48430293547120495, + "grad_norm": 2.665994167327881, + "learning_rate": 0.00010496642953658065, + "loss": 0.2117, + "step": 20660 + }, + { + "epoch": 0.48432637704600995, + "grad_norm": 0.8533589839935303, + "learning_rate": 0.00010495907245536016, + "loss": 0.1108, + "step": 20661 + }, + { + "epoch": 0.48434981862081494, + "grad_norm": 3.246790647506714, + "learning_rate": 0.00010495171534723158, + "loss": 0.3256, + "step": 20662 + }, + { + "epoch": 0.48437326019561994, + "grad_norm": 1.148765206336975, + "learning_rate": 0.00010494435821223488, + "loss": 0.1564, + "step": 20663 + }, + { + "epoch": 0.48439670177042493, + "grad_norm": 3.39819598197937, + "learning_rate": 0.00010493700105040994, + "loss": 0.3889, + "step": 20664 + }, + { + "epoch": 0.4844201433452299, + "grad_norm": 3.5258753299713135, + "learning_rate": 0.00010492964386179671, + "loss": 0.2025, + "step": 20665 + }, + { + "epoch": 0.4844435849200349, + "grad_norm": 4.989241123199463, + "learning_rate": 0.00010492228664643506, + "loss": 0.3047, + "step": 20666 + }, + { + "epoch": 0.4844670264948399, + "grad_norm": 2.2401492595672607, + "learning_rate": 0.00010491492940436498, + "loss": 0.0794, + "step": 20667 + }, + { + "epoch": 0.4844904680696449, + "grad_norm": 3.4114720821380615, + "learning_rate": 0.00010490757213562637, + "loss": 0.3018, + "step": 20668 + }, + { + "epoch": 0.4845139096444499, + "grad_norm": 3.3847479820251465, + "learning_rate": 0.00010490021484025907, + "loss": 0.1728, + "step": 20669 + }, + { + "epoch": 0.4845373512192549, + "grad_norm": 1.1928884983062744, + "learning_rate": 0.00010489285751830311, + "loss": 0.0736, + "step": 20670 + }, + { + "epoch": 0.4845607927940599, + "grad_norm": 3.9626717567443848, + "learning_rate": 0.00010488550016979837, + "loss": 0.3857, + "step": 20671 + }, + { + "epoch": 0.4845842343688649, + "grad_norm": 1.9738346338272095, + "learning_rate": 0.00010487814279478477, + "loss": 0.1818, + "step": 20672 + }, + { + "epoch": 0.4846076759436699, + "grad_norm": 2.2902894020080566, + "learning_rate": 0.00010487078539330222, + "loss": 0.2548, + "step": 20673 + }, + { + "epoch": 0.4846311175184749, + "grad_norm": 1.7565611600875854, + "learning_rate": 0.00010486342796539065, + "loss": 0.2063, + "step": 20674 + }, + { + "epoch": 0.48465455909327987, + "grad_norm": 2.6718461513519287, + "learning_rate": 0.00010485607051108998, + "loss": 0.3423, + "step": 20675 + }, + { + "epoch": 0.48467800066808486, + "grad_norm": 1.8497204780578613, + "learning_rate": 0.00010484871303044012, + "loss": 1.9781, + "step": 20676 + }, + { + "epoch": 0.48470144224288986, + "grad_norm": 4.723905563354492, + "learning_rate": 0.00010484135552348104, + "loss": 0.5505, + "step": 20677 + }, + { + "epoch": 0.48472488381769485, + "grad_norm": 0.8508719205856323, + "learning_rate": 0.00010483399799025264, + "loss": 0.069, + "step": 20678 + }, + { + "epoch": 0.48474832539249985, + "grad_norm": 4.962258338928223, + "learning_rate": 0.0001048266404307948, + "loss": 0.5595, + "step": 20679 + }, + { + "epoch": 0.48477176696730484, + "grad_norm": 4.082103252410889, + "learning_rate": 0.00010481928284514748, + "loss": 0.5282, + "step": 20680 + }, + { + "epoch": 0.48479520854210983, + "grad_norm": 1.4545058012008667, + "learning_rate": 0.0001048119252333506, + "loss": 0.1004, + "step": 20681 + }, + { + "epoch": 0.48481865011691483, + "grad_norm": 3.1801106929779053, + "learning_rate": 0.00010480456759544406, + "loss": 0.3735, + "step": 20682 + }, + { + "epoch": 0.4848420916917198, + "grad_norm": 5.288829803466797, + "learning_rate": 0.00010479720993146782, + "loss": 0.52, + "step": 20683 + }, + { + "epoch": 0.4848655332665248, + "grad_norm": 3.7321531772613525, + "learning_rate": 0.00010478985224146178, + "loss": 0.2221, + "step": 20684 + }, + { + "epoch": 0.48488897484132987, + "grad_norm": 1.1010850667953491, + "learning_rate": 0.00010478249452546587, + "loss": 0.1337, + "step": 20685 + }, + { + "epoch": 0.48491241641613486, + "grad_norm": 1.5793460607528687, + "learning_rate": 0.00010477513678352, + "loss": 0.7565, + "step": 20686 + }, + { + "epoch": 0.48493585799093986, + "grad_norm": 2.8564841747283936, + "learning_rate": 0.00010476777901566412, + "loss": 0.2084, + "step": 20687 + }, + { + "epoch": 0.48495929956574485, + "grad_norm": 3.0899717807769775, + "learning_rate": 0.0001047604212219381, + "loss": 0.3002, + "step": 20688 + }, + { + "epoch": 0.48498274114054984, + "grad_norm": 2.977454423904419, + "learning_rate": 0.00010475306340238193, + "loss": 0.3558, + "step": 20689 + }, + { + "epoch": 0.48500618271535484, + "grad_norm": 3.0759336948394775, + "learning_rate": 0.00010474570555703551, + "loss": 0.2243, + "step": 20690 + }, + { + "epoch": 0.48502962429015983, + "grad_norm": 3.3413503170013428, + "learning_rate": 0.00010473834768593875, + "loss": 0.3077, + "step": 20691 + }, + { + "epoch": 0.4850530658649648, + "grad_norm": 2.904571771621704, + "learning_rate": 0.00010473098978913158, + "loss": 0.2829, + "step": 20692 + }, + { + "epoch": 0.4850765074397698, + "grad_norm": 4.024960517883301, + "learning_rate": 0.00010472363186665395, + "loss": 0.4923, + "step": 20693 + }, + { + "epoch": 0.4850999490145748, + "grad_norm": 1.164374828338623, + "learning_rate": 0.00010471627391854573, + "loss": 0.1344, + "step": 20694 + }, + { + "epoch": 0.4851233905893798, + "grad_norm": 2.430145025253296, + "learning_rate": 0.00010470891594484687, + "loss": 0.2879, + "step": 20695 + }, + { + "epoch": 0.4851468321641848, + "grad_norm": 3.6031227111816406, + "learning_rate": 0.00010470155794559733, + "loss": 0.2556, + "step": 20696 + }, + { + "epoch": 0.4851702737389898, + "grad_norm": 0.3307746648788452, + "learning_rate": 0.000104694199920837, + "loss": 0.0239, + "step": 20697 + }, + { + "epoch": 0.4851937153137948, + "grad_norm": 3.7640342712402344, + "learning_rate": 0.00010468684187060577, + "loss": 0.3072, + "step": 20698 + }, + { + "epoch": 0.4852171568885998, + "grad_norm": 3.693343162536621, + "learning_rate": 0.00010467948379494366, + "loss": 1.828, + "step": 20699 + }, + { + "epoch": 0.4852405984634048, + "grad_norm": 0.7855455279350281, + "learning_rate": 0.0001046721256938905, + "loss": 0.0569, + "step": 20700 + }, + { + "epoch": 0.4852640400382098, + "grad_norm": 4.418436050415039, + "learning_rate": 0.00010466476756748624, + "loss": 0.6288, + "step": 20701 + }, + { + "epoch": 0.48528748161301477, + "grad_norm": 4.144776344299316, + "learning_rate": 0.00010465740941577087, + "loss": 2.5401, + "step": 20702 + }, + { + "epoch": 0.48531092318781976, + "grad_norm": 3.349740982055664, + "learning_rate": 0.00010465005123878423, + "loss": 0.3751, + "step": 20703 + }, + { + "epoch": 0.48533436476262476, + "grad_norm": 1.5775569677352905, + "learning_rate": 0.0001046426930365663, + "loss": 0.1895, + "step": 20704 + }, + { + "epoch": 0.48535780633742975, + "grad_norm": 4.878164768218994, + "learning_rate": 0.00010463533480915699, + "loss": 0.4214, + "step": 20705 + }, + { + "epoch": 0.48538124791223475, + "grad_norm": 3.429216146469116, + "learning_rate": 0.0001046279765565962, + "loss": 0.3374, + "step": 20706 + }, + { + "epoch": 0.48540468948703974, + "grad_norm": 1.2957998514175415, + "learning_rate": 0.00010462061827892389, + "loss": 0.0974, + "step": 20707 + }, + { + "epoch": 0.48542813106184474, + "grad_norm": 3.607379674911499, + "learning_rate": 0.00010461325997617998, + "loss": 0.1488, + "step": 20708 + }, + { + "epoch": 0.48545157263664973, + "grad_norm": 2.132317543029785, + "learning_rate": 0.00010460590164840441, + "loss": 0.0896, + "step": 20709 + }, + { + "epoch": 0.4854750142114547, + "grad_norm": 3.4711873531341553, + "learning_rate": 0.00010459854329563705, + "loss": 0.2188, + "step": 20710 + }, + { + "epoch": 0.4854984557862597, + "grad_norm": 2.5701005458831787, + "learning_rate": 0.00010459118491791788, + "loss": 0.3594, + "step": 20711 + }, + { + "epoch": 0.4855218973610647, + "grad_norm": 3.693782329559326, + "learning_rate": 0.00010458382651528682, + "loss": 0.2571, + "step": 20712 + }, + { + "epoch": 0.4855453389358697, + "grad_norm": 1.326961636543274, + "learning_rate": 0.00010457646808778378, + "loss": 0.1631, + "step": 20713 + }, + { + "epoch": 0.4855687805106747, + "grad_norm": 1.3262598514556885, + "learning_rate": 0.00010456910963544868, + "loss": 0.1539, + "step": 20714 + }, + { + "epoch": 0.4855922220854797, + "grad_norm": 2.278289318084717, + "learning_rate": 0.0001045617511583215, + "loss": 0.1812, + "step": 20715 + }, + { + "epoch": 0.4856156636602847, + "grad_norm": 0.8613080978393555, + "learning_rate": 0.00010455439265644211, + "loss": 0.0712, + "step": 20716 + }, + { + "epoch": 0.4856391052350897, + "grad_norm": 2.4580202102661133, + "learning_rate": 0.00010454703412985045, + "loss": 0.1581, + "step": 20717 + }, + { + "epoch": 0.4856625468098947, + "grad_norm": 2.2018909454345703, + "learning_rate": 0.00010453967557858646, + "loss": 0.1298, + "step": 20718 + }, + { + "epoch": 0.4856859883846997, + "grad_norm": 4.630178451538086, + "learning_rate": 0.00010453231700269008, + "loss": 0.6371, + "step": 20719 + }, + { + "epoch": 0.48570942995950467, + "grad_norm": 1.6930327415466309, + "learning_rate": 0.00010452495840220118, + "loss": 0.1558, + "step": 20720 + }, + { + "epoch": 0.48573287153430966, + "grad_norm": 0.5758791565895081, + "learning_rate": 0.00010451759977715978, + "loss": 0.0774, + "step": 20721 + }, + { + "epoch": 0.48575631310911466, + "grad_norm": 0.5457995533943176, + "learning_rate": 0.00010451024112760573, + "loss": 0.0589, + "step": 20722 + }, + { + "epoch": 0.48577975468391965, + "grad_norm": 1.4734758138656616, + "learning_rate": 0.00010450288245357897, + "loss": 0.1304, + "step": 20723 + }, + { + "epoch": 0.48580319625872465, + "grad_norm": 2.5809388160705566, + "learning_rate": 0.00010449552375511946, + "loss": 0.2726, + "step": 20724 + }, + { + "epoch": 0.48582663783352964, + "grad_norm": 1.7910399436950684, + "learning_rate": 0.0001044881650322671, + "loss": 0.1402, + "step": 20725 + }, + { + "epoch": 0.48585007940833463, + "grad_norm": 2.4564435482025146, + "learning_rate": 0.00010448080628506184, + "loss": 1.5774, + "step": 20726 + }, + { + "epoch": 0.48587352098313963, + "grad_norm": 1.6447298526763916, + "learning_rate": 0.00010447344751354362, + "loss": 0.1778, + "step": 20727 + }, + { + "epoch": 0.4858969625579446, + "grad_norm": 3.5409374237060547, + "learning_rate": 0.00010446608871775234, + "loss": 1.743, + "step": 20728 + }, + { + "epoch": 0.4859204041327496, + "grad_norm": 0.9867662191390991, + "learning_rate": 0.00010445872989772792, + "loss": 0.0662, + "step": 20729 + }, + { + "epoch": 0.4859438457075546, + "grad_norm": 2.662003755569458, + "learning_rate": 0.00010445137105351028, + "loss": 1.4136, + "step": 20730 + }, + { + "epoch": 0.4859672872823596, + "grad_norm": 2.3624608516693115, + "learning_rate": 0.00010444401218513942, + "loss": 0.2118, + "step": 20731 + }, + { + "epoch": 0.4859907288571646, + "grad_norm": 2.736173629760742, + "learning_rate": 0.00010443665329265522, + "loss": 0.1775, + "step": 20732 + }, + { + "epoch": 0.4860141704319696, + "grad_norm": 4.406921863555908, + "learning_rate": 0.00010442929437609761, + "loss": 0.3385, + "step": 20733 + }, + { + "epoch": 0.4860376120067746, + "grad_norm": 4.115255832672119, + "learning_rate": 0.00010442193543550652, + "loss": 1.2841, + "step": 20734 + }, + { + "epoch": 0.4860610535815796, + "grad_norm": 1.591029167175293, + "learning_rate": 0.0001044145764709219, + "loss": 0.2318, + "step": 20735 + }, + { + "epoch": 0.48608449515638463, + "grad_norm": 2.6647415161132812, + "learning_rate": 0.0001044072174823836, + "loss": 0.2206, + "step": 20736 + }, + { + "epoch": 0.4861079367311896, + "grad_norm": 1.3495972156524658, + "learning_rate": 0.00010439985846993167, + "loss": 0.1179, + "step": 20737 + }, + { + "epoch": 0.4861313783059946, + "grad_norm": 3.682183265686035, + "learning_rate": 0.000104392499433606, + "loss": 0.3716, + "step": 20738 + }, + { + "epoch": 0.4861548198807996, + "grad_norm": 0.7229151725769043, + "learning_rate": 0.00010438514037344647, + "loss": 0.0475, + "step": 20739 + }, + { + "epoch": 0.4861782614556046, + "grad_norm": 4.188364505767822, + "learning_rate": 0.00010437778128949307, + "loss": 0.5039, + "step": 20740 + }, + { + "epoch": 0.4862017030304096, + "grad_norm": 3.677381992340088, + "learning_rate": 0.0001043704221817857, + "loss": 0.3143, + "step": 20741 + }, + { + "epoch": 0.4862251446052146, + "grad_norm": 4.239202499389648, + "learning_rate": 0.00010436306305036428, + "loss": 0.4436, + "step": 20742 + }, + { + "epoch": 0.4862485861800196, + "grad_norm": 3.995023488998413, + "learning_rate": 0.00010435570389526875, + "loss": 0.4357, + "step": 20743 + }, + { + "epoch": 0.4862720277548246, + "grad_norm": 0.8078953623771667, + "learning_rate": 0.00010434834471653907, + "loss": 0.1174, + "step": 20744 + }, + { + "epoch": 0.4862954693296296, + "grad_norm": 3.3329017162323, + "learning_rate": 0.00010434098551421514, + "loss": 0.5005, + "step": 20745 + }, + { + "epoch": 0.4863189109044346, + "grad_norm": 2.4494428634643555, + "learning_rate": 0.0001043336262883369, + "loss": 0.1437, + "step": 20746 + }, + { + "epoch": 0.48634235247923957, + "grad_norm": 3.1345033645629883, + "learning_rate": 0.0001043262670389443, + "loss": 0.2839, + "step": 20747 + }, + { + "epoch": 0.48636579405404456, + "grad_norm": 2.5046589374542236, + "learning_rate": 0.00010431890776607722, + "loss": 0.2089, + "step": 20748 + }, + { + "epoch": 0.48638923562884956, + "grad_norm": 3.7720625400543213, + "learning_rate": 0.00010431154846977564, + "loss": 0.4988, + "step": 20749 + }, + { + "epoch": 0.48641267720365455, + "grad_norm": 2.329730987548828, + "learning_rate": 0.00010430418915007949, + "loss": 1.8237, + "step": 20750 + }, + { + "epoch": 0.48643611877845955, + "grad_norm": 4.848628997802734, + "learning_rate": 0.00010429682980702868, + "loss": 0.3088, + "step": 20751 + }, + { + "epoch": 0.48645956035326454, + "grad_norm": 3.65763521194458, + "learning_rate": 0.00010428947044066315, + "loss": 1.6846, + "step": 20752 + }, + { + "epoch": 0.48648300192806954, + "grad_norm": 2.07126522064209, + "learning_rate": 0.00010428211105102283, + "loss": 0.1657, + "step": 20753 + }, + { + "epoch": 0.48650644350287453, + "grad_norm": 2.2067172527313232, + "learning_rate": 0.00010427475163814766, + "loss": 0.2706, + "step": 20754 + }, + { + "epoch": 0.4865298850776795, + "grad_norm": 0.5702735781669617, + "learning_rate": 0.00010426739220207755, + "loss": 0.0387, + "step": 20755 + }, + { + "epoch": 0.4865533266524845, + "grad_norm": 3.3868496417999268, + "learning_rate": 0.00010426003274285248, + "loss": 0.2173, + "step": 20756 + }, + { + "epoch": 0.4865767682272895, + "grad_norm": 3.3553903102874756, + "learning_rate": 0.00010425267326051234, + "loss": 0.2783, + "step": 20757 + }, + { + "epoch": 0.4866002098020945, + "grad_norm": 4.29572868347168, + "learning_rate": 0.00010424531375509707, + "loss": 0.4428, + "step": 20758 + }, + { + "epoch": 0.4866236513768995, + "grad_norm": 3.9598898887634277, + "learning_rate": 0.00010423795422664662, + "loss": 1.5447, + "step": 20759 + }, + { + "epoch": 0.4866470929517045, + "grad_norm": 3.578646659851074, + "learning_rate": 0.00010423059467520091, + "loss": 0.2975, + "step": 20760 + }, + { + "epoch": 0.4866705345265095, + "grad_norm": 0.9052550196647644, + "learning_rate": 0.00010422323510079985, + "loss": 0.0726, + "step": 20761 + }, + { + "epoch": 0.4866939761013145, + "grad_norm": 3.899653196334839, + "learning_rate": 0.0001042158755034834, + "loss": 0.2868, + "step": 20762 + }, + { + "epoch": 0.4867174176761195, + "grad_norm": 3.9465060234069824, + "learning_rate": 0.00010420851588329152, + "loss": 0.3532, + "step": 20763 + }, + { + "epoch": 0.4867408592509245, + "grad_norm": 3.805391311645508, + "learning_rate": 0.0001042011562402641, + "loss": 0.5591, + "step": 20764 + }, + { + "epoch": 0.48676430082572947, + "grad_norm": 3.5826964378356934, + "learning_rate": 0.00010419379657444108, + "loss": 0.4078, + "step": 20765 + }, + { + "epoch": 0.48678774240053446, + "grad_norm": 3.670734167098999, + "learning_rate": 0.00010418643688586243, + "loss": 0.3809, + "step": 20766 + }, + { + "epoch": 0.48681118397533946, + "grad_norm": 3.7812340259552, + "learning_rate": 0.00010417907717456803, + "loss": 0.619, + "step": 20767 + }, + { + "epoch": 0.48683462555014445, + "grad_norm": 1.1658340692520142, + "learning_rate": 0.00010417171744059782, + "loss": 0.0854, + "step": 20768 + }, + { + "epoch": 0.48685806712494945, + "grad_norm": 3.0878279209136963, + "learning_rate": 0.0001041643576839918, + "loss": 0.379, + "step": 20769 + }, + { + "epoch": 0.48688150869975444, + "grad_norm": 1.7527003288269043, + "learning_rate": 0.00010415699790478983, + "loss": 0.1609, + "step": 20770 + }, + { + "epoch": 0.48690495027455943, + "grad_norm": 0.9141643643379211, + "learning_rate": 0.00010414963810303189, + "loss": 0.091, + "step": 20771 + }, + { + "epoch": 0.48692839184936443, + "grad_norm": 4.50454568862915, + "learning_rate": 0.00010414227827875789, + "loss": 0.2532, + "step": 20772 + }, + { + "epoch": 0.4869518334241694, + "grad_norm": 4.889155864715576, + "learning_rate": 0.00010413491843200775, + "loss": 0.7962, + "step": 20773 + }, + { + "epoch": 0.4869752749989744, + "grad_norm": 2.286454439163208, + "learning_rate": 0.00010412755856282143, + "loss": 1.4537, + "step": 20774 + }, + { + "epoch": 0.4869987165737794, + "grad_norm": 1.406922459602356, + "learning_rate": 0.00010412019867123888, + "loss": 0.1371, + "step": 20775 + }, + { + "epoch": 0.4870221581485844, + "grad_norm": 4.254957675933838, + "learning_rate": 0.00010411283875730002, + "loss": 1.845, + "step": 20776 + }, + { + "epoch": 0.4870455997233894, + "grad_norm": 3.5918819904327393, + "learning_rate": 0.00010410547882104478, + "loss": 0.4889, + "step": 20777 + }, + { + "epoch": 0.4870690412981944, + "grad_norm": 2.548483371734619, + "learning_rate": 0.00010409811886251309, + "loss": 1.1737, + "step": 20778 + }, + { + "epoch": 0.4870924828729994, + "grad_norm": 3.263601064682007, + "learning_rate": 0.0001040907588817449, + "loss": 0.2896, + "step": 20779 + }, + { + "epoch": 0.4871159244478044, + "grad_norm": 2.4881856441497803, + "learning_rate": 0.00010408339887878012, + "loss": 0.2489, + "step": 20780 + }, + { + "epoch": 0.4871393660226094, + "grad_norm": 4.122557163238525, + "learning_rate": 0.0001040760388536587, + "loss": 0.4387, + "step": 20781 + }, + { + "epoch": 0.48716280759741437, + "grad_norm": 2.1672394275665283, + "learning_rate": 0.0001040686788064206, + "loss": 1.3534, + "step": 20782 + }, + { + "epoch": 0.48718624917221937, + "grad_norm": 1.8606257438659668, + "learning_rate": 0.00010406131873710571, + "loss": 0.2336, + "step": 20783 + }, + { + "epoch": 0.48720969074702436, + "grad_norm": 4.235834121704102, + "learning_rate": 0.00010405395864575402, + "loss": 0.3887, + "step": 20784 + }, + { + "epoch": 0.48723313232182935, + "grad_norm": 3.8299076557159424, + "learning_rate": 0.00010404659853240541, + "loss": 0.3178, + "step": 20785 + }, + { + "epoch": 0.48725657389663435, + "grad_norm": 0.39939773082733154, + "learning_rate": 0.00010403923839709986, + "loss": 0.0212, + "step": 20786 + }, + { + "epoch": 0.4872800154714394, + "grad_norm": 3.440751314163208, + "learning_rate": 0.00010403187823987727, + "loss": 0.5068, + "step": 20787 + }, + { + "epoch": 0.4873034570462444, + "grad_norm": 5.508290767669678, + "learning_rate": 0.00010402451806077762, + "loss": 0.7724, + "step": 20788 + }, + { + "epoch": 0.4873268986210494, + "grad_norm": 5.965757369995117, + "learning_rate": 0.00010401715785984083, + "loss": 0.3037, + "step": 20789 + }, + { + "epoch": 0.4873503401958544, + "grad_norm": 1.738433837890625, + "learning_rate": 0.00010400979763710678, + "loss": 0.2652, + "step": 20790 + }, + { + "epoch": 0.4873737817706594, + "grad_norm": 3.850377321243286, + "learning_rate": 0.00010400243739261548, + "loss": 0.7234, + "step": 20791 + }, + { + "epoch": 0.48739722334546437, + "grad_norm": 4.9877166748046875, + "learning_rate": 0.00010399507712640686, + "loss": 0.6044, + "step": 20792 + }, + { + "epoch": 0.48742066492026936, + "grad_norm": 3.3871545791625977, + "learning_rate": 0.0001039877168385208, + "loss": 0.3415, + "step": 20793 + }, + { + "epoch": 0.48744410649507436, + "grad_norm": 1.9394664764404297, + "learning_rate": 0.00010398035652899731, + "loss": 0.18, + "step": 20794 + }, + { + "epoch": 0.48746754806987935, + "grad_norm": 3.9537317752838135, + "learning_rate": 0.0001039729961978763, + "loss": 0.3181, + "step": 20795 + }, + { + "epoch": 0.48749098964468435, + "grad_norm": 6.282483100891113, + "learning_rate": 0.00010396563584519767, + "loss": 0.6768, + "step": 20796 + }, + { + "epoch": 0.48751443121948934, + "grad_norm": 3.415199041366577, + "learning_rate": 0.00010395827547100141, + "loss": 0.3951, + "step": 20797 + }, + { + "epoch": 0.48753787279429434, + "grad_norm": 2.6094841957092285, + "learning_rate": 0.00010395091507532744, + "loss": 0.2321, + "step": 20798 + }, + { + "epoch": 0.48756131436909933, + "grad_norm": 1.2048333883285522, + "learning_rate": 0.00010394355465821566, + "loss": 0.6714, + "step": 20799 + }, + { + "epoch": 0.4875847559439043, + "grad_norm": 1.1940711736679077, + "learning_rate": 0.00010393619421970609, + "loss": 0.0568, + "step": 20800 + }, + { + "epoch": 0.4876081975187093, + "grad_norm": 4.2925801277160645, + "learning_rate": 0.0001039288337598386, + "loss": 0.6306, + "step": 20801 + }, + { + "epoch": 0.4876316390935143, + "grad_norm": 4.489193439483643, + "learning_rate": 0.00010392147327865314, + "loss": 0.1955, + "step": 20802 + }, + { + "epoch": 0.4876550806683193, + "grad_norm": 3.0836946964263916, + "learning_rate": 0.00010391411277618963, + "loss": 0.4151, + "step": 20803 + }, + { + "epoch": 0.4876785222431243, + "grad_norm": 5.2521538734436035, + "learning_rate": 0.00010390675225248807, + "loss": 0.7173, + "step": 20804 + }, + { + "epoch": 0.4877019638179293, + "grad_norm": 2.6521871089935303, + "learning_rate": 0.00010389939170758838, + "loss": 1.1946, + "step": 20805 + }, + { + "epoch": 0.4877254053927343, + "grad_norm": 3.0289313793182373, + "learning_rate": 0.00010389203114153044, + "loss": 0.4609, + "step": 20806 + }, + { + "epoch": 0.4877488469675393, + "grad_norm": 0.6336272954940796, + "learning_rate": 0.00010388467055435424, + "loss": 0.0752, + "step": 20807 + }, + { + "epoch": 0.4877722885423443, + "grad_norm": 3.2261643409729004, + "learning_rate": 0.00010387730994609972, + "loss": 0.3062, + "step": 20808 + }, + { + "epoch": 0.4877957301171493, + "grad_norm": 2.647348403930664, + "learning_rate": 0.00010386994931680678, + "loss": 0.2136, + "step": 20809 + }, + { + "epoch": 0.48781917169195427, + "grad_norm": 3.9801950454711914, + "learning_rate": 0.00010386258866651541, + "loss": 0.3183, + "step": 20810 + }, + { + "epoch": 0.48784261326675926, + "grad_norm": 1.9246467351913452, + "learning_rate": 0.00010385522799526554, + "loss": 0.2467, + "step": 20811 + }, + { + "epoch": 0.48786605484156426, + "grad_norm": 3.348228693008423, + "learning_rate": 0.00010384786730309705, + "loss": 0.2846, + "step": 20812 + }, + { + "epoch": 0.48788949641636925, + "grad_norm": 3.5047552585601807, + "learning_rate": 0.00010384050659004997, + "loss": 0.614, + "step": 20813 + }, + { + "epoch": 0.48791293799117424, + "grad_norm": 2.8675081729888916, + "learning_rate": 0.00010383314585616419, + "loss": 0.2321, + "step": 20814 + }, + { + "epoch": 0.48793637956597924, + "grad_norm": 2.6781904697418213, + "learning_rate": 0.00010382578510147963, + "loss": 0.329, + "step": 20815 + }, + { + "epoch": 0.48795982114078423, + "grad_norm": 0.3596414625644684, + "learning_rate": 0.00010381842432603624, + "loss": 0.0214, + "step": 20816 + }, + { + "epoch": 0.4879832627155892, + "grad_norm": 4.27178430557251, + "learning_rate": 0.000103811063529874, + "loss": 0.5171, + "step": 20817 + }, + { + "epoch": 0.4880067042903942, + "grad_norm": 1.9362138509750366, + "learning_rate": 0.00010380370271303282, + "loss": 0.2348, + "step": 20818 + }, + { + "epoch": 0.4880301458651992, + "grad_norm": 1.0171034336090088, + "learning_rate": 0.00010379634187555263, + "loss": 0.0958, + "step": 20819 + }, + { + "epoch": 0.4880535874400042, + "grad_norm": 2.8619682788848877, + "learning_rate": 0.00010378898101747341, + "loss": 0.2362, + "step": 20820 + }, + { + "epoch": 0.4880770290148092, + "grad_norm": 2.52946400642395, + "learning_rate": 0.00010378162013883503, + "loss": 1.2616, + "step": 20821 + }, + { + "epoch": 0.4881004705896142, + "grad_norm": 2.803007125854492, + "learning_rate": 0.00010377425923967747, + "loss": 0.3186, + "step": 20822 + }, + { + "epoch": 0.4881239121644192, + "grad_norm": 0.7536377310752869, + "learning_rate": 0.00010376689832004072, + "loss": 0.0563, + "step": 20823 + }, + { + "epoch": 0.4881473537392242, + "grad_norm": 1.0302189588546753, + "learning_rate": 0.00010375953737996464, + "loss": 0.0698, + "step": 20824 + }, + { + "epoch": 0.4881707953140292, + "grad_norm": 3.8749334812164307, + "learning_rate": 0.00010375217641948923, + "loss": 0.4048, + "step": 20825 + }, + { + "epoch": 0.4881942368888342, + "grad_norm": 5.220578670501709, + "learning_rate": 0.0001037448154386544, + "loss": 0.6417, + "step": 20826 + }, + { + "epoch": 0.48821767846363917, + "grad_norm": 2.731663703918457, + "learning_rate": 0.00010373745443750007, + "loss": 0.2434, + "step": 20827 + }, + { + "epoch": 0.48824112003844417, + "grad_norm": 2.6732826232910156, + "learning_rate": 0.00010373009341606621, + "loss": 0.2944, + "step": 20828 + }, + { + "epoch": 0.48826456161324916, + "grad_norm": 2.018493175506592, + "learning_rate": 0.00010372273237439279, + "loss": 1.3498, + "step": 20829 + }, + { + "epoch": 0.48828800318805415, + "grad_norm": 1.4387648105621338, + "learning_rate": 0.00010371537131251969, + "loss": 0.076, + "step": 20830 + }, + { + "epoch": 0.48831144476285915, + "grad_norm": 1.9210048913955688, + "learning_rate": 0.00010370801023048689, + "loss": 1.1691, + "step": 20831 + }, + { + "epoch": 0.48833488633766414, + "grad_norm": 2.8118741512298584, + "learning_rate": 0.00010370064912833432, + "loss": 0.3153, + "step": 20832 + }, + { + "epoch": 0.48835832791246914, + "grad_norm": 4.118387222290039, + "learning_rate": 0.00010369328800610192, + "loss": 0.4222, + "step": 20833 + }, + { + "epoch": 0.48838176948727413, + "grad_norm": 2.2420713901519775, + "learning_rate": 0.00010368592686382966, + "loss": 1.0477, + "step": 20834 + }, + { + "epoch": 0.4884052110620791, + "grad_norm": 2.5109009742736816, + "learning_rate": 0.00010367856570155744, + "loss": 0.1295, + "step": 20835 + }, + { + "epoch": 0.4884286526368841, + "grad_norm": 3.2544052600860596, + "learning_rate": 0.00010367120451932521, + "loss": 0.4214, + "step": 20836 + }, + { + "epoch": 0.4884520942116891, + "grad_norm": 0.8788211345672607, + "learning_rate": 0.00010366384331717292, + "loss": 0.0771, + "step": 20837 + }, + { + "epoch": 0.48847553578649416, + "grad_norm": 4.405072212219238, + "learning_rate": 0.00010365648209514054, + "loss": 0.517, + "step": 20838 + }, + { + "epoch": 0.48849897736129916, + "grad_norm": 0.35307446122169495, + "learning_rate": 0.000103649120853268, + "loss": 0.0293, + "step": 20839 + }, + { + "epoch": 0.48852241893610415, + "grad_norm": 3.196587562561035, + "learning_rate": 0.00010364175959159517, + "loss": 0.3164, + "step": 20840 + }, + { + "epoch": 0.48854586051090915, + "grad_norm": 0.8848536610603333, + "learning_rate": 0.00010363439831016207, + "loss": 0.091, + "step": 20841 + }, + { + "epoch": 0.48856930208571414, + "grad_norm": 1.4226710796356201, + "learning_rate": 0.00010362703700900863, + "loss": 0.0813, + "step": 20842 + }, + { + "epoch": 0.48859274366051914, + "grad_norm": 0.8925696015357971, + "learning_rate": 0.00010361967568817479, + "loss": 0.0927, + "step": 20843 + }, + { + "epoch": 0.48861618523532413, + "grad_norm": 2.5225882530212402, + "learning_rate": 0.00010361231434770047, + "loss": 0.3346, + "step": 20844 + }, + { + "epoch": 0.4886396268101291, + "grad_norm": 1.8361040353775024, + "learning_rate": 0.00010360495298762564, + "loss": 0.1469, + "step": 20845 + }, + { + "epoch": 0.4886630683849341, + "grad_norm": 3.5871963500976562, + "learning_rate": 0.00010359759160799024, + "loss": 0.2239, + "step": 20846 + }, + { + "epoch": 0.4886865099597391, + "grad_norm": 0.5959751605987549, + "learning_rate": 0.00010359023020883418, + "loss": 0.0571, + "step": 20847 + }, + { + "epoch": 0.4887099515345441, + "grad_norm": 2.183349609375, + "learning_rate": 0.00010358286879019747, + "loss": 0.2, + "step": 20848 + }, + { + "epoch": 0.4887333931093491, + "grad_norm": 1.4860702753067017, + "learning_rate": 0.00010357550735212, + "loss": 0.1317, + "step": 20849 + }, + { + "epoch": 0.4887568346841541, + "grad_norm": 3.9888529777526855, + "learning_rate": 0.00010356814589464173, + "loss": 0.3466, + "step": 20850 + }, + { + "epoch": 0.4887802762589591, + "grad_norm": 2.7427818775177, + "learning_rate": 0.0001035607844178026, + "loss": 0.4059, + "step": 20851 + }, + { + "epoch": 0.4888037178337641, + "grad_norm": 3.832031488418579, + "learning_rate": 0.00010355342292164253, + "loss": 0.5605, + "step": 20852 + }, + { + "epoch": 0.4888271594085691, + "grad_norm": 1.2278239727020264, + "learning_rate": 0.00010354606140620151, + "loss": 0.1525, + "step": 20853 + }, + { + "epoch": 0.4888506009833741, + "grad_norm": 2.750227689743042, + "learning_rate": 0.00010353869987151945, + "loss": 0.364, + "step": 20854 + }, + { + "epoch": 0.48887404255817907, + "grad_norm": 2.760861396789551, + "learning_rate": 0.00010353133831763633, + "loss": 0.2294, + "step": 20855 + }, + { + "epoch": 0.48889748413298406, + "grad_norm": 2.121894359588623, + "learning_rate": 0.00010352397674459202, + "loss": 0.2045, + "step": 20856 + }, + { + "epoch": 0.48892092570778906, + "grad_norm": 1.5549589395523071, + "learning_rate": 0.00010351661515242657, + "loss": 0.8961, + "step": 20857 + }, + { + "epoch": 0.48894436728259405, + "grad_norm": 2.7588441371917725, + "learning_rate": 0.00010350925354117982, + "loss": 0.4118, + "step": 20858 + }, + { + "epoch": 0.48896780885739904, + "grad_norm": 0.550007700920105, + "learning_rate": 0.00010350189191089179, + "loss": 0.0448, + "step": 20859 + }, + { + "epoch": 0.48899125043220404, + "grad_norm": 0.5560225248336792, + "learning_rate": 0.00010349453026160239, + "loss": 0.0529, + "step": 20860 + }, + { + "epoch": 0.48901469200700903, + "grad_norm": 0.7374075651168823, + "learning_rate": 0.00010348716859335157, + "loss": 0.0327, + "step": 20861 + }, + { + "epoch": 0.489038133581814, + "grad_norm": 1.9352108240127563, + "learning_rate": 0.00010347980690617929, + "loss": 0.0813, + "step": 20862 + }, + { + "epoch": 0.489061575156619, + "grad_norm": 2.2153372764587402, + "learning_rate": 0.00010347244520012545, + "loss": 0.1473, + "step": 20863 + }, + { + "epoch": 0.489085016731424, + "grad_norm": 3.435314893722534, + "learning_rate": 0.00010346508347523001, + "loss": 0.3768, + "step": 20864 + }, + { + "epoch": 0.489108458306229, + "grad_norm": 0.9369084239006042, + "learning_rate": 0.00010345772173153296, + "loss": 0.0965, + "step": 20865 + }, + { + "epoch": 0.489131899881034, + "grad_norm": 2.909550666809082, + "learning_rate": 0.00010345035996907422, + "loss": 0.2752, + "step": 20866 + }, + { + "epoch": 0.489155341455839, + "grad_norm": 2.1223649978637695, + "learning_rate": 0.00010344299818789371, + "loss": 0.1438, + "step": 20867 + }, + { + "epoch": 0.489178783030644, + "grad_norm": 3.289106607437134, + "learning_rate": 0.00010343563638803144, + "loss": 0.2158, + "step": 20868 + }, + { + "epoch": 0.489202224605449, + "grad_norm": 5.081658840179443, + "learning_rate": 0.00010342827456952726, + "loss": 0.5579, + "step": 20869 + }, + { + "epoch": 0.489225666180254, + "grad_norm": 3.518812656402588, + "learning_rate": 0.00010342091273242116, + "loss": 0.3945, + "step": 20870 + }, + { + "epoch": 0.489249107755059, + "grad_norm": 1.126861572265625, + "learning_rate": 0.00010341355087675313, + "loss": 0.1137, + "step": 20871 + }, + { + "epoch": 0.48927254932986397, + "grad_norm": 4.206203460693359, + "learning_rate": 0.00010340618900256306, + "loss": 0.4778, + "step": 20872 + }, + { + "epoch": 0.48929599090466896, + "grad_norm": 6.927947521209717, + "learning_rate": 0.0001033988271098909, + "loss": 0.4496, + "step": 20873 + }, + { + "epoch": 0.48931943247947396, + "grad_norm": 4.049510955810547, + "learning_rate": 0.00010339146519877665, + "loss": 0.5317, + "step": 20874 + }, + { + "epoch": 0.48934287405427895, + "grad_norm": 5.111794948577881, + "learning_rate": 0.00010338410326926018, + "loss": 0.3472, + "step": 20875 + }, + { + "epoch": 0.48936631562908395, + "grad_norm": 0.6882324814796448, + "learning_rate": 0.00010337674132138146, + "loss": 0.0279, + "step": 20876 + }, + { + "epoch": 0.48938975720388894, + "grad_norm": 6.145678520202637, + "learning_rate": 0.00010336937935518047, + "loss": 0.4006, + "step": 20877 + }, + { + "epoch": 0.48941319877869394, + "grad_norm": 4.2556352615356445, + "learning_rate": 0.00010336201737069716, + "loss": 0.4725, + "step": 20878 + }, + { + "epoch": 0.48943664035349893, + "grad_norm": 3.87376070022583, + "learning_rate": 0.00010335465536797137, + "loss": 0.2745, + "step": 20879 + }, + { + "epoch": 0.4894600819283039, + "grad_norm": 3.524075508117676, + "learning_rate": 0.0001033472933470432, + "loss": 0.2752, + "step": 20880 + }, + { + "epoch": 0.4894835235031089, + "grad_norm": 3.356703042984009, + "learning_rate": 0.0001033399313079525, + "loss": 0.5028, + "step": 20881 + }, + { + "epoch": 0.4895069650779139, + "grad_norm": 3.229736804962158, + "learning_rate": 0.0001033325692507392, + "loss": 0.3203, + "step": 20882 + }, + { + "epoch": 0.4895304066527189, + "grad_norm": 2.649237871170044, + "learning_rate": 0.00010332520717544334, + "loss": 0.3064, + "step": 20883 + }, + { + "epoch": 0.4895538482275239, + "grad_norm": 2.9423186779022217, + "learning_rate": 0.0001033178450821048, + "loss": 1.592, + "step": 20884 + }, + { + "epoch": 0.4895772898023289, + "grad_norm": 4.246830940246582, + "learning_rate": 0.0001033104829707635, + "loss": 0.4425, + "step": 20885 + }, + { + "epoch": 0.4896007313771339, + "grad_norm": 0.9507192373275757, + "learning_rate": 0.00010330312084145948, + "loss": 0.0953, + "step": 20886 + }, + { + "epoch": 0.4896241729519389, + "grad_norm": 0.3253350257873535, + "learning_rate": 0.00010329575869423262, + "loss": 0.0198, + "step": 20887 + }, + { + "epoch": 0.4896476145267439, + "grad_norm": 2.326998472213745, + "learning_rate": 0.00010328839652912286, + "loss": 0.1587, + "step": 20888 + }, + { + "epoch": 0.48967105610154893, + "grad_norm": 1.6748816967010498, + "learning_rate": 0.00010328103434617016, + "loss": 0.2836, + "step": 20889 + }, + { + "epoch": 0.4896944976763539, + "grad_norm": 1.266892671585083, + "learning_rate": 0.00010327367214541451, + "loss": 0.1547, + "step": 20890 + }, + { + "epoch": 0.4897179392511589, + "grad_norm": 1.7669134140014648, + "learning_rate": 0.0001032663099268958, + "loss": 1.6529, + "step": 20891 + }, + { + "epoch": 0.4897413808259639, + "grad_norm": 2.7340738773345947, + "learning_rate": 0.00010325894769065401, + "loss": 1.1766, + "step": 20892 + }, + { + "epoch": 0.4897648224007689, + "grad_norm": 3.270089864730835, + "learning_rate": 0.00010325158543672909, + "loss": 0.4005, + "step": 20893 + }, + { + "epoch": 0.4897882639755739, + "grad_norm": 2.657912254333496, + "learning_rate": 0.00010324422316516094, + "loss": 1.4094, + "step": 20894 + }, + { + "epoch": 0.4898117055503789, + "grad_norm": 3.4174535274505615, + "learning_rate": 0.00010323686087598957, + "loss": 0.2842, + "step": 20895 + }, + { + "epoch": 0.4898351471251839, + "grad_norm": 0.8205329775810242, + "learning_rate": 0.0001032294985692549, + "loss": 0.1028, + "step": 20896 + }, + { + "epoch": 0.4898585886999889, + "grad_norm": 1.6083887815475464, + "learning_rate": 0.00010322213624499687, + "loss": 0.7443, + "step": 20897 + }, + { + "epoch": 0.4898820302747939, + "grad_norm": 2.285559892654419, + "learning_rate": 0.00010321477390325545, + "loss": 0.1309, + "step": 20898 + }, + { + "epoch": 0.4899054718495989, + "grad_norm": 1.4407731294631958, + "learning_rate": 0.00010320741154407058, + "loss": 0.1081, + "step": 20899 + }, + { + "epoch": 0.48992891342440387, + "grad_norm": 2.5693154335021973, + "learning_rate": 0.00010320004916748218, + "loss": 0.3597, + "step": 20900 + }, + { + "epoch": 0.48995235499920886, + "grad_norm": 0.6440458297729492, + "learning_rate": 0.00010319268677353023, + "loss": 0.0565, + "step": 20901 + }, + { + "epoch": 0.48997579657401386, + "grad_norm": 3.2986202239990234, + "learning_rate": 0.0001031853243622547, + "loss": 0.342, + "step": 20902 + }, + { + "epoch": 0.48999923814881885, + "grad_norm": 1.9734392166137695, + "learning_rate": 0.00010317796193369549, + "loss": 0.4916, + "step": 20903 + }, + { + "epoch": 0.49002267972362384, + "grad_norm": 1.3442951440811157, + "learning_rate": 0.00010317059948789256, + "loss": 0.1705, + "step": 20904 + }, + { + "epoch": 0.49004612129842884, + "grad_norm": 0.7955862879753113, + "learning_rate": 0.00010316323702488589, + "loss": 0.0699, + "step": 20905 + }, + { + "epoch": 0.49006956287323383, + "grad_norm": 3.493314266204834, + "learning_rate": 0.00010315587454471538, + "loss": 1.2998, + "step": 20906 + }, + { + "epoch": 0.4900930044480388, + "grad_norm": 0.861384391784668, + "learning_rate": 0.00010314851204742105, + "loss": 0.1011, + "step": 20907 + }, + { + "epoch": 0.4901164460228438, + "grad_norm": 3.3987679481506348, + "learning_rate": 0.00010314114953304275, + "loss": 0.5098, + "step": 20908 + }, + { + "epoch": 0.4901398875976488, + "grad_norm": 2.426257371902466, + "learning_rate": 0.00010313378700162053, + "loss": 0.2537, + "step": 20909 + }, + { + "epoch": 0.4901633291724538, + "grad_norm": 7.894789695739746, + "learning_rate": 0.00010312642445319429, + "loss": 0.85, + "step": 20910 + }, + { + "epoch": 0.4901867707472588, + "grad_norm": 2.312544822692871, + "learning_rate": 0.00010311906188780396, + "loss": 0.8398, + "step": 20911 + }, + { + "epoch": 0.4902102123220638, + "grad_norm": 3.3520169258117676, + "learning_rate": 0.00010311169930548951, + "loss": 0.4411, + "step": 20912 + }, + { + "epoch": 0.4902336538968688, + "grad_norm": 1.1825034618377686, + "learning_rate": 0.0001031043367062909, + "loss": 0.1325, + "step": 20913 + }, + { + "epoch": 0.4902570954716738, + "grad_norm": 2.4420166015625, + "learning_rate": 0.00010309697409024805, + "loss": 1.3074, + "step": 20914 + }, + { + "epoch": 0.4902805370464788, + "grad_norm": 2.4519565105438232, + "learning_rate": 0.00010308961145740098, + "loss": 0.2143, + "step": 20915 + }, + { + "epoch": 0.4903039786212838, + "grad_norm": 1.8327959775924683, + "learning_rate": 0.00010308224880778958, + "loss": 0.1821, + "step": 20916 + }, + { + "epoch": 0.49032742019608877, + "grad_norm": 3.777512788772583, + "learning_rate": 0.00010307488614145378, + "loss": 0.2623, + "step": 20917 + }, + { + "epoch": 0.49035086177089376, + "grad_norm": 1.141527771949768, + "learning_rate": 0.0001030675234584336, + "loss": 0.117, + "step": 20918 + }, + { + "epoch": 0.49037430334569876, + "grad_norm": 3.837955951690674, + "learning_rate": 0.00010306016075876892, + "loss": 0.5698, + "step": 20919 + }, + { + "epoch": 0.49039774492050375, + "grad_norm": 3.841830253601074, + "learning_rate": 0.00010305279804249972, + "loss": 0.403, + "step": 20920 + }, + { + "epoch": 0.49042118649530875, + "grad_norm": 3.0741782188415527, + "learning_rate": 0.00010304543530966597, + "loss": 0.2531, + "step": 20921 + }, + { + "epoch": 0.49044462807011374, + "grad_norm": 3.304792642593384, + "learning_rate": 0.0001030380725603076, + "loss": 0.3007, + "step": 20922 + }, + { + "epoch": 0.49046806964491874, + "grad_norm": 2.844697952270508, + "learning_rate": 0.00010303070979446456, + "loss": 0.1661, + "step": 20923 + }, + { + "epoch": 0.49049151121972373, + "grad_norm": 3.4405109882354736, + "learning_rate": 0.00010302334701217681, + "loss": 0.3194, + "step": 20924 + }, + { + "epoch": 0.4905149527945287, + "grad_norm": 3.4382214546203613, + "learning_rate": 0.00010301598421348429, + "loss": 0.3206, + "step": 20925 + }, + { + "epoch": 0.4905383943693337, + "grad_norm": 1.9226081371307373, + "learning_rate": 0.00010300862139842695, + "loss": 0.9262, + "step": 20926 + }, + { + "epoch": 0.4905618359441387, + "grad_norm": 5.009076118469238, + "learning_rate": 0.00010300125856704474, + "loss": 0.4927, + "step": 20927 + }, + { + "epoch": 0.4905852775189437, + "grad_norm": 3.327866315841675, + "learning_rate": 0.00010299389571937762, + "loss": 0.3393, + "step": 20928 + }, + { + "epoch": 0.4906087190937487, + "grad_norm": 3.498934507369995, + "learning_rate": 0.00010298653285546555, + "loss": 0.1795, + "step": 20929 + }, + { + "epoch": 0.4906321606685537, + "grad_norm": 4.682991981506348, + "learning_rate": 0.00010297916997534847, + "loss": 0.4606, + "step": 20930 + }, + { + "epoch": 0.4906556022433587, + "grad_norm": 3.5739643573760986, + "learning_rate": 0.00010297180707906631, + "loss": 0.346, + "step": 20931 + }, + { + "epoch": 0.4906790438181637, + "grad_norm": 4.655478477478027, + "learning_rate": 0.00010296444416665904, + "loss": 0.5992, + "step": 20932 + }, + { + "epoch": 0.4907024853929687, + "grad_norm": 3.5427486896514893, + "learning_rate": 0.00010295708123816664, + "loss": 0.3792, + "step": 20933 + }, + { + "epoch": 0.4907259269677737, + "grad_norm": 4.409211158752441, + "learning_rate": 0.00010294971829362902, + "loss": 0.3891, + "step": 20934 + }, + { + "epoch": 0.49074936854257867, + "grad_norm": 1.643512487411499, + "learning_rate": 0.00010294235533308616, + "loss": 0.1581, + "step": 20935 + }, + { + "epoch": 0.49077281011738366, + "grad_norm": 1.5685977935791016, + "learning_rate": 0.00010293499235657795, + "loss": 0.1207, + "step": 20936 + }, + { + "epoch": 0.49079625169218866, + "grad_norm": 1.9454891681671143, + "learning_rate": 0.00010292762936414442, + "loss": 0.1932, + "step": 20937 + }, + { + "epoch": 0.49081969326699365, + "grad_norm": 0.8618748188018799, + "learning_rate": 0.0001029202663558255, + "loss": 0.067, + "step": 20938 + }, + { + "epoch": 0.49084313484179865, + "grad_norm": 1.6540671586990356, + "learning_rate": 0.00010291290333166112, + "loss": 0.1402, + "step": 20939 + }, + { + "epoch": 0.4908665764166037, + "grad_norm": 3.295686960220337, + "learning_rate": 0.00010290554029169126, + "loss": 0.4073, + "step": 20940 + }, + { + "epoch": 0.4908900179914087, + "grad_norm": 1.9284461736679077, + "learning_rate": 0.00010289817723595585, + "loss": 0.1504, + "step": 20941 + }, + { + "epoch": 0.4909134595662137, + "grad_norm": 0.4771822690963745, + "learning_rate": 0.00010289081416449484, + "loss": 0.0208, + "step": 20942 + }, + { + "epoch": 0.4909369011410187, + "grad_norm": 1.842765212059021, + "learning_rate": 0.00010288345107734819, + "loss": 0.191, + "step": 20943 + }, + { + "epoch": 0.49096034271582367, + "grad_norm": 2.8956775665283203, + "learning_rate": 0.00010287608797455586, + "loss": 0.3013, + "step": 20944 + }, + { + "epoch": 0.49098378429062867, + "grad_norm": 2.323509454727173, + "learning_rate": 0.0001028687248561578, + "loss": 0.2975, + "step": 20945 + }, + { + "epoch": 0.49100722586543366, + "grad_norm": 0.9679815173149109, + "learning_rate": 0.00010286136172219394, + "loss": 0.0924, + "step": 20946 + }, + { + "epoch": 0.49103066744023866, + "grad_norm": 3.00457501411438, + "learning_rate": 0.00010285399857270427, + "loss": 0.3923, + "step": 20947 + }, + { + "epoch": 0.49105410901504365, + "grad_norm": 4.111043930053711, + "learning_rate": 0.00010284663540772874, + "loss": 0.4869, + "step": 20948 + }, + { + "epoch": 0.49107755058984864, + "grad_norm": 1.477341651916504, + "learning_rate": 0.00010283927222730722, + "loss": 0.1342, + "step": 20949 + }, + { + "epoch": 0.49110099216465364, + "grad_norm": 1.5652230978012085, + "learning_rate": 0.00010283190903147979, + "loss": 0.1201, + "step": 20950 + }, + { + "epoch": 0.49112443373945863, + "grad_norm": 0.9728596210479736, + "learning_rate": 0.00010282454582028634, + "loss": 0.1367, + "step": 20951 + }, + { + "epoch": 0.4911478753142636, + "grad_norm": 2.955753803253174, + "learning_rate": 0.00010281718259376681, + "loss": 0.2227, + "step": 20952 + }, + { + "epoch": 0.4911713168890686, + "grad_norm": 2.4224419593811035, + "learning_rate": 0.00010280981935196117, + "loss": 0.247, + "step": 20953 + }, + { + "epoch": 0.4911947584638736, + "grad_norm": 3.5328736305236816, + "learning_rate": 0.00010280245609490937, + "loss": 1.696, + "step": 20954 + }, + { + "epoch": 0.4912182000386786, + "grad_norm": 6.195456027984619, + "learning_rate": 0.00010279509282265132, + "loss": 0.5586, + "step": 20955 + }, + { + "epoch": 0.4912416416134836, + "grad_norm": 4.728081703186035, + "learning_rate": 0.00010278772953522709, + "loss": 0.55, + "step": 20956 + }, + { + "epoch": 0.4912650831882886, + "grad_norm": 3.6244356632232666, + "learning_rate": 0.00010278036623267652, + "loss": 0.5304, + "step": 20957 + }, + { + "epoch": 0.4912885247630936, + "grad_norm": 2.283947229385376, + "learning_rate": 0.00010277300291503962, + "loss": 0.1584, + "step": 20958 + }, + { + "epoch": 0.4913119663378986, + "grad_norm": 3.129741668701172, + "learning_rate": 0.00010276563958235634, + "loss": 0.3247, + "step": 20959 + }, + { + "epoch": 0.4913354079127036, + "grad_norm": 2.0140979290008545, + "learning_rate": 0.0001027582762346666, + "loss": 0.0908, + "step": 20960 + }, + { + "epoch": 0.4913588494875086, + "grad_norm": 2.4987733364105225, + "learning_rate": 0.00010275091287201037, + "loss": 0.2319, + "step": 20961 + }, + { + "epoch": 0.49138229106231357, + "grad_norm": 2.9799227714538574, + "learning_rate": 0.00010274354949442761, + "loss": 1.8434, + "step": 20962 + }, + { + "epoch": 0.49140573263711856, + "grad_norm": 5.238487243652344, + "learning_rate": 0.0001027361861019583, + "loss": 0.3237, + "step": 20963 + }, + { + "epoch": 0.49142917421192356, + "grad_norm": 2.479320526123047, + "learning_rate": 0.00010272882269464237, + "loss": 0.202, + "step": 20964 + }, + { + "epoch": 0.49145261578672855, + "grad_norm": 3.256035327911377, + "learning_rate": 0.00010272145927251975, + "loss": 0.3001, + "step": 20965 + }, + { + "epoch": 0.49147605736153355, + "grad_norm": 3.93503999710083, + "learning_rate": 0.00010271409583563043, + "loss": 0.3223, + "step": 20966 + }, + { + "epoch": 0.49149949893633854, + "grad_norm": 3.3906166553497314, + "learning_rate": 0.00010270673238401432, + "loss": 0.3133, + "step": 20967 + }, + { + "epoch": 0.49152294051114354, + "grad_norm": 1.0737814903259277, + "learning_rate": 0.00010269936891771142, + "loss": 0.1239, + "step": 20968 + }, + { + "epoch": 0.49154638208594853, + "grad_norm": 3.9318673610687256, + "learning_rate": 0.00010269200543676168, + "loss": 0.4046, + "step": 20969 + }, + { + "epoch": 0.4915698236607535, + "grad_norm": 3.1419034004211426, + "learning_rate": 0.00010268464194120504, + "loss": 0.3242, + "step": 20970 + }, + { + "epoch": 0.4915932652355585, + "grad_norm": 4.172131538391113, + "learning_rate": 0.00010267727843108143, + "loss": 0.26, + "step": 20971 + }, + { + "epoch": 0.4916167068103635, + "grad_norm": 4.3241753578186035, + "learning_rate": 0.00010266991490643088, + "loss": 0.4177, + "step": 20972 + }, + { + "epoch": 0.4916401483851685, + "grad_norm": 0.6209138035774231, + "learning_rate": 0.00010266255136729324, + "loss": 0.0545, + "step": 20973 + }, + { + "epoch": 0.4916635899599735, + "grad_norm": 2.3615293502807617, + "learning_rate": 0.00010265518781370855, + "loss": 1.6823, + "step": 20974 + }, + { + "epoch": 0.4916870315347785, + "grad_norm": 5.35239839553833, + "learning_rate": 0.00010264782424571675, + "loss": 0.5474, + "step": 20975 + }, + { + "epoch": 0.4917104731095835, + "grad_norm": 2.9896011352539062, + "learning_rate": 0.00010264046066335777, + "loss": 0.3245, + "step": 20976 + }, + { + "epoch": 0.4917339146843885, + "grad_norm": 1.350032091140747, + "learning_rate": 0.00010263309706667157, + "loss": 0.1159, + "step": 20977 + }, + { + "epoch": 0.4917573562591935, + "grad_norm": 0.6842332482337952, + "learning_rate": 0.00010262573345569813, + "loss": 0.0655, + "step": 20978 + }, + { + "epoch": 0.4917807978339985, + "grad_norm": 3.5483851432800293, + "learning_rate": 0.00010261836983047734, + "loss": 0.2965, + "step": 20979 + }, + { + "epoch": 0.49180423940880347, + "grad_norm": 2.9535071849823, + "learning_rate": 0.00010261100619104926, + "loss": 1.764, + "step": 20980 + }, + { + "epoch": 0.49182768098360846, + "grad_norm": 3.0893166065216064, + "learning_rate": 0.00010260364253745376, + "loss": 0.287, + "step": 20981 + }, + { + "epoch": 0.49185112255841346, + "grad_norm": 2.4554107189178467, + "learning_rate": 0.00010259627886973082, + "loss": 1.4609, + "step": 20982 + }, + { + "epoch": 0.49187456413321845, + "grad_norm": 3.8138515949249268, + "learning_rate": 0.00010258891518792042, + "loss": 0.4785, + "step": 20983 + }, + { + "epoch": 0.49189800570802344, + "grad_norm": 5.209067344665527, + "learning_rate": 0.00010258155149206245, + "loss": 0.7815, + "step": 20984 + }, + { + "epoch": 0.49192144728282844, + "grad_norm": 3.5271260738372803, + "learning_rate": 0.00010257418778219693, + "loss": 0.2517, + "step": 20985 + }, + { + "epoch": 0.49194488885763343, + "grad_norm": 2.9064395427703857, + "learning_rate": 0.00010256682405836382, + "loss": 1.807, + "step": 20986 + }, + { + "epoch": 0.4919683304324384, + "grad_norm": 3.147172212600708, + "learning_rate": 0.00010255946032060302, + "loss": 1.8499, + "step": 20987 + }, + { + "epoch": 0.4919917720072434, + "grad_norm": 0.9594841599464417, + "learning_rate": 0.00010255209656895452, + "loss": 0.107, + "step": 20988 + }, + { + "epoch": 0.4920152135820484, + "grad_norm": 2.3045849800109863, + "learning_rate": 0.00010254473280345828, + "loss": 0.186, + "step": 20989 + }, + { + "epoch": 0.4920386551568534, + "grad_norm": 0.3471790552139282, + "learning_rate": 0.00010253736902415425, + "loss": 0.0195, + "step": 20990 + }, + { + "epoch": 0.49206209673165846, + "grad_norm": 2.292959690093994, + "learning_rate": 0.00010253000523108238, + "loss": 1.3647, + "step": 20991 + }, + { + "epoch": 0.49208553830646345, + "grad_norm": 0.8076607584953308, + "learning_rate": 0.00010252264142428263, + "loss": 0.12, + "step": 20992 + }, + { + "epoch": 0.49210897988126845, + "grad_norm": 2.2315752506256104, + "learning_rate": 0.00010251527760379494, + "loss": 0.1419, + "step": 20993 + }, + { + "epoch": 0.49213242145607344, + "grad_norm": 0.946811854839325, + "learning_rate": 0.00010250791376965932, + "loss": 0.0528, + "step": 20994 + }, + { + "epoch": 0.49215586303087844, + "grad_norm": 2.4009644985198975, + "learning_rate": 0.00010250054992191567, + "loss": 0.3027, + "step": 20995 + }, + { + "epoch": 0.49217930460568343, + "grad_norm": 1.0552324056625366, + "learning_rate": 0.00010249318606060396, + "loss": 0.2174, + "step": 20996 + }, + { + "epoch": 0.4922027461804884, + "grad_norm": 3.4169368743896484, + "learning_rate": 0.00010248582218576417, + "loss": 0.3312, + "step": 20997 + }, + { + "epoch": 0.4922261877552934, + "grad_norm": 1.085472822189331, + "learning_rate": 0.0001024784582974362, + "loss": 0.0875, + "step": 20998 + }, + { + "epoch": 0.4922496293300984, + "grad_norm": 4.621455669403076, + "learning_rate": 0.0001024710943956601, + "loss": 0.4434, + "step": 20999 + }, + { + "epoch": 0.4922730709049034, + "grad_norm": 2.5924885272979736, + "learning_rate": 0.00010246373048047573, + "loss": 0.3264, + "step": 21000 + }, + { + "epoch": 0.4922965124797084, + "grad_norm": 3.275822639465332, + "learning_rate": 0.00010245636655192313, + "loss": 0.3111, + "step": 21001 + }, + { + "epoch": 0.4923199540545134, + "grad_norm": 2.684095621109009, + "learning_rate": 0.00010244900261004218, + "loss": 0.2365, + "step": 21002 + }, + { + "epoch": 0.4923433956293184, + "grad_norm": 2.1376919746398926, + "learning_rate": 0.00010244163865487289, + "loss": 0.2237, + "step": 21003 + }, + { + "epoch": 0.4923668372041234, + "grad_norm": 5.5654616355896, + "learning_rate": 0.00010243427468645517, + "loss": 0.4576, + "step": 21004 + }, + { + "epoch": 0.4923902787789284, + "grad_norm": 1.779762864112854, + "learning_rate": 0.00010242691070482904, + "loss": 0.2132, + "step": 21005 + }, + { + "epoch": 0.4924137203537334, + "grad_norm": 3.665823221206665, + "learning_rate": 0.0001024195467100344, + "loss": 0.2239, + "step": 21006 + }, + { + "epoch": 0.49243716192853837, + "grad_norm": 1.6122872829437256, + "learning_rate": 0.00010241218270211126, + "loss": 0.1112, + "step": 21007 + }, + { + "epoch": 0.49246060350334336, + "grad_norm": 0.7394933104515076, + "learning_rate": 0.00010240481868109955, + "loss": 0.1109, + "step": 21008 + }, + { + "epoch": 0.49248404507814836, + "grad_norm": 2.8450944423675537, + "learning_rate": 0.00010239745464703918, + "loss": 1.3466, + "step": 21009 + }, + { + "epoch": 0.49250748665295335, + "grad_norm": 4.725423812866211, + "learning_rate": 0.00010239009059997018, + "loss": 0.3962, + "step": 21010 + }, + { + "epoch": 0.49253092822775835, + "grad_norm": 3.449345350265503, + "learning_rate": 0.0001023827265399325, + "loss": 0.2197, + "step": 21011 + }, + { + "epoch": 0.49255436980256334, + "grad_norm": 3.9423537254333496, + "learning_rate": 0.00010237536246696606, + "loss": 0.2664, + "step": 21012 + }, + { + "epoch": 0.49257781137736834, + "grad_norm": 1.0609818696975708, + "learning_rate": 0.00010236799838111085, + "loss": 0.1322, + "step": 21013 + }, + { + "epoch": 0.49260125295217333, + "grad_norm": 1.9962514638900757, + "learning_rate": 0.0001023606342824068, + "loss": 0.2179, + "step": 21014 + }, + { + "epoch": 0.4926246945269783, + "grad_norm": 3.341240406036377, + "learning_rate": 0.00010235327017089389, + "loss": 0.4614, + "step": 21015 + }, + { + "epoch": 0.4926481361017833, + "grad_norm": 2.9063565731048584, + "learning_rate": 0.00010234590604661206, + "loss": 0.376, + "step": 21016 + }, + { + "epoch": 0.4926715776765883, + "grad_norm": 3.5283617973327637, + "learning_rate": 0.00010233854190960128, + "loss": 1.8138, + "step": 21017 + }, + { + "epoch": 0.4926950192513933, + "grad_norm": 3.5271694660186768, + "learning_rate": 0.00010233117775990153, + "loss": 0.4419, + "step": 21018 + }, + { + "epoch": 0.4927184608261983, + "grad_norm": 2.235452651977539, + "learning_rate": 0.00010232381359755272, + "loss": 0.2121, + "step": 21019 + }, + { + "epoch": 0.4927419024010033, + "grad_norm": 3.2653915882110596, + "learning_rate": 0.00010231644942259484, + "loss": 0.2375, + "step": 21020 + }, + { + "epoch": 0.4927653439758083, + "grad_norm": 3.5334062576293945, + "learning_rate": 0.00010230908523506784, + "loss": 0.3004, + "step": 21021 + }, + { + "epoch": 0.4927887855506133, + "grad_norm": 4.768547058105469, + "learning_rate": 0.00010230172103501165, + "loss": 0.2823, + "step": 21022 + }, + { + "epoch": 0.4928122271254183, + "grad_norm": 1.5677244663238525, + "learning_rate": 0.00010229435682246629, + "loss": 0.115, + "step": 21023 + }, + { + "epoch": 0.4928356687002233, + "grad_norm": 0.30288636684417725, + "learning_rate": 0.0001022869925974717, + "loss": 0.0465, + "step": 21024 + }, + { + "epoch": 0.49285911027502827, + "grad_norm": 3.9933371543884277, + "learning_rate": 0.00010227962836006777, + "loss": 0.4168, + "step": 21025 + }, + { + "epoch": 0.49288255184983326, + "grad_norm": 3.564784288406372, + "learning_rate": 0.00010227226411029456, + "loss": 0.3459, + "step": 21026 + }, + { + "epoch": 0.49290599342463826, + "grad_norm": 0.7609792351722717, + "learning_rate": 0.00010226489984819198, + "loss": 0.0588, + "step": 21027 + }, + { + "epoch": 0.49292943499944325, + "grad_norm": 1.4353545904159546, + "learning_rate": 0.00010225753557379996, + "loss": 0.0497, + "step": 21028 + }, + { + "epoch": 0.49295287657424824, + "grad_norm": 3.5217626094818115, + "learning_rate": 0.00010225017128715851, + "loss": 0.282, + "step": 21029 + }, + { + "epoch": 0.49297631814905324, + "grad_norm": 2.54593825340271, + "learning_rate": 0.00010224280698830755, + "loss": 0.3175, + "step": 21030 + }, + { + "epoch": 0.49299975972385823, + "grad_norm": 3.2638189792633057, + "learning_rate": 0.00010223544267728707, + "loss": 0.4762, + "step": 21031 + }, + { + "epoch": 0.4930232012986632, + "grad_norm": 3.5344231128692627, + "learning_rate": 0.00010222807835413702, + "loss": 0.3302, + "step": 21032 + }, + { + "epoch": 0.4930466428734682, + "grad_norm": 3.1933839321136475, + "learning_rate": 0.00010222071401889735, + "loss": 0.1745, + "step": 21033 + }, + { + "epoch": 0.4930700844482732, + "grad_norm": 3.972452402114868, + "learning_rate": 0.000102213349671608, + "loss": 0.2229, + "step": 21034 + }, + { + "epoch": 0.4930935260230782, + "grad_norm": 3.162562370300293, + "learning_rate": 0.00010220598531230899, + "loss": 1.2888, + "step": 21035 + }, + { + "epoch": 0.4931169675978832, + "grad_norm": 1.6175189018249512, + "learning_rate": 0.00010219862094104023, + "loss": 0.2666, + "step": 21036 + }, + { + "epoch": 0.4931404091726882, + "grad_norm": 4.743586540222168, + "learning_rate": 0.00010219125655784167, + "loss": 0.5346, + "step": 21037 + }, + { + "epoch": 0.4931638507474932, + "grad_norm": 5.162312030792236, + "learning_rate": 0.00010218389216275331, + "loss": 0.2731, + "step": 21038 + }, + { + "epoch": 0.4931872923222982, + "grad_norm": 3.31516695022583, + "learning_rate": 0.0001021765277558151, + "loss": 0.2951, + "step": 21039 + }, + { + "epoch": 0.4932107338971032, + "grad_norm": 2.7645866870880127, + "learning_rate": 0.00010216916333706697, + "loss": 0.2787, + "step": 21040 + }, + { + "epoch": 0.4932341754719082, + "grad_norm": 2.1180877685546875, + "learning_rate": 0.00010216179890654887, + "loss": 0.1624, + "step": 21041 + }, + { + "epoch": 0.4932576170467132, + "grad_norm": 2.924362897872925, + "learning_rate": 0.00010215443446430084, + "loss": 0.3204, + "step": 21042 + }, + { + "epoch": 0.4932810586215182, + "grad_norm": 0.47919660806655884, + "learning_rate": 0.00010214707001036279, + "loss": 0.0531, + "step": 21043 + }, + { + "epoch": 0.4933045001963232, + "grad_norm": 1.6493325233459473, + "learning_rate": 0.00010213970554477463, + "loss": 0.1245, + "step": 21044 + }, + { + "epoch": 0.4933279417711282, + "grad_norm": 2.02022123336792, + "learning_rate": 0.00010213234106757641, + "loss": 1.2618, + "step": 21045 + }, + { + "epoch": 0.4933513833459332, + "grad_norm": 0.5394209027290344, + "learning_rate": 0.00010212497657880801, + "loss": 0.0301, + "step": 21046 + }, + { + "epoch": 0.4933748249207382, + "grad_norm": 3.2761666774749756, + "learning_rate": 0.00010211761207850947, + "loss": 2.1275, + "step": 21047 + }, + { + "epoch": 0.4933982664955432, + "grad_norm": 3.0129008293151855, + "learning_rate": 0.00010211024756672067, + "loss": 0.3305, + "step": 21048 + }, + { + "epoch": 0.4934217080703482, + "grad_norm": 2.9039220809936523, + "learning_rate": 0.00010210288304348163, + "loss": 0.3211, + "step": 21049 + }, + { + "epoch": 0.4934451496451532, + "grad_norm": 2.199369192123413, + "learning_rate": 0.00010209551850883228, + "loss": 0.1827, + "step": 21050 + }, + { + "epoch": 0.4934685912199582, + "grad_norm": 2.992309331893921, + "learning_rate": 0.0001020881539628126, + "loss": 0.2962, + "step": 21051 + }, + { + "epoch": 0.49349203279476317, + "grad_norm": 2.277817964553833, + "learning_rate": 0.0001020807894054625, + "loss": 0.2275, + "step": 21052 + }, + { + "epoch": 0.49351547436956816, + "grad_norm": 14.901799201965332, + "learning_rate": 0.00010207342483682202, + "loss": 0.7558, + "step": 21053 + }, + { + "epoch": 0.49353891594437316, + "grad_norm": 5.6681694984436035, + "learning_rate": 0.00010206606025693106, + "loss": 1.8349, + "step": 21054 + }, + { + "epoch": 0.49356235751917815, + "grad_norm": 3.054582357406616, + "learning_rate": 0.00010205869566582962, + "loss": 0.3116, + "step": 21055 + }, + { + "epoch": 0.49358579909398315, + "grad_norm": 1.8812233209609985, + "learning_rate": 0.00010205133106355762, + "loss": 1.2351, + "step": 21056 + }, + { + "epoch": 0.49360924066878814, + "grad_norm": 0.5491081476211548, + "learning_rate": 0.00010204396645015504, + "loss": 0.0412, + "step": 21057 + }, + { + "epoch": 0.49363268224359313, + "grad_norm": 1.9086923599243164, + "learning_rate": 0.00010203660182566184, + "loss": 0.1151, + "step": 21058 + }, + { + "epoch": 0.49365612381839813, + "grad_norm": 1.309719443321228, + "learning_rate": 0.00010202923719011799, + "loss": 0.1636, + "step": 21059 + }, + { + "epoch": 0.4936795653932031, + "grad_norm": 3.4297361373901367, + "learning_rate": 0.00010202187254356341, + "loss": 0.2394, + "step": 21060 + }, + { + "epoch": 0.4937030069680081, + "grad_norm": 6.51106071472168, + "learning_rate": 0.00010201450788603815, + "loss": 0.3979, + "step": 21061 + }, + { + "epoch": 0.4937264485428131, + "grad_norm": 4.37130069732666, + "learning_rate": 0.00010200714321758207, + "loss": 0.7735, + "step": 21062 + }, + { + "epoch": 0.4937498901176181, + "grad_norm": 3.881819248199463, + "learning_rate": 0.00010199977853823519, + "loss": 0.455, + "step": 21063 + }, + { + "epoch": 0.4937733316924231, + "grad_norm": 3.3025569915771484, + "learning_rate": 0.00010199241384803743, + "loss": 0.3562, + "step": 21064 + }, + { + "epoch": 0.4937967732672281, + "grad_norm": 3.127354621887207, + "learning_rate": 0.00010198504914702882, + "loss": 0.4174, + "step": 21065 + }, + { + "epoch": 0.4938202148420331, + "grad_norm": 4.062175273895264, + "learning_rate": 0.00010197768443524927, + "loss": 0.4569, + "step": 21066 + }, + { + "epoch": 0.4938436564168381, + "grad_norm": 2.781128406524658, + "learning_rate": 0.00010197031971273872, + "loss": 0.141, + "step": 21067 + }, + { + "epoch": 0.4938670979916431, + "grad_norm": 2.4093103408813477, + "learning_rate": 0.00010196295497953718, + "loss": 0.1894, + "step": 21068 + }, + { + "epoch": 0.4938905395664481, + "grad_norm": 0.7503939270973206, + "learning_rate": 0.00010195559023568457, + "loss": 0.0536, + "step": 21069 + }, + { + "epoch": 0.49391398114125307, + "grad_norm": 3.3775341510772705, + "learning_rate": 0.00010194822548122089, + "loss": 0.3968, + "step": 21070 + }, + { + "epoch": 0.49393742271605806, + "grad_norm": 2.1373519897460938, + "learning_rate": 0.00010194086071618607, + "loss": 0.2443, + "step": 21071 + }, + { + "epoch": 0.49396086429086306, + "grad_norm": 3.878258228302002, + "learning_rate": 0.00010193349594062012, + "loss": 0.2847, + "step": 21072 + }, + { + "epoch": 0.49398430586566805, + "grad_norm": 1.9206241369247437, + "learning_rate": 0.00010192613115456293, + "loss": 1.302, + "step": 21073 + }, + { + "epoch": 0.49400774744047304, + "grad_norm": 1.230704426765442, + "learning_rate": 0.00010191876635805453, + "loss": 0.1335, + "step": 21074 + }, + { + "epoch": 0.49403118901527804, + "grad_norm": 2.013103485107422, + "learning_rate": 0.00010191140155113484, + "loss": 0.1749, + "step": 21075 + }, + { + "epoch": 0.49405463059008303, + "grad_norm": 0.6341842412948608, + "learning_rate": 0.00010190403673384381, + "loss": 0.039, + "step": 21076 + }, + { + "epoch": 0.494078072164888, + "grad_norm": 2.578878164291382, + "learning_rate": 0.0001018966719062214, + "loss": 0.1675, + "step": 21077 + }, + { + "epoch": 0.494101513739693, + "grad_norm": 1.6283135414123535, + "learning_rate": 0.00010188930706830766, + "loss": 0.1117, + "step": 21078 + }, + { + "epoch": 0.494124955314498, + "grad_norm": 4.003758430480957, + "learning_rate": 0.00010188194222014244, + "loss": 0.2855, + "step": 21079 + }, + { + "epoch": 0.494148396889303, + "grad_norm": 1.1521085500717163, + "learning_rate": 0.00010187457736176577, + "loss": 0.125, + "step": 21080 + }, + { + "epoch": 0.494171838464108, + "grad_norm": 4.733184337615967, + "learning_rate": 0.00010186721249321761, + "loss": 0.3848, + "step": 21081 + }, + { + "epoch": 0.494195280038913, + "grad_norm": 3.704352378845215, + "learning_rate": 0.00010185984761453786, + "loss": 0.2505, + "step": 21082 + }, + { + "epoch": 0.494218721613718, + "grad_norm": 0.3923285901546478, + "learning_rate": 0.00010185248272576655, + "loss": 0.0314, + "step": 21083 + }, + { + "epoch": 0.494242163188523, + "grad_norm": 1.8799045085906982, + "learning_rate": 0.00010184511782694362, + "loss": 0.2214, + "step": 21084 + }, + { + "epoch": 0.494265604763328, + "grad_norm": 3.722709894180298, + "learning_rate": 0.00010183775291810902, + "loss": 0.2582, + "step": 21085 + }, + { + "epoch": 0.494289046338133, + "grad_norm": 2.8427181243896484, + "learning_rate": 0.00010183038799930271, + "loss": 0.2633, + "step": 21086 + }, + { + "epoch": 0.49431248791293797, + "grad_norm": 3.5432565212249756, + "learning_rate": 0.00010182302307056468, + "loss": 1.3928, + "step": 21087 + }, + { + "epoch": 0.49433592948774296, + "grad_norm": 4.9636311531066895, + "learning_rate": 0.00010181565813193485, + "loss": 0.4435, + "step": 21088 + }, + { + "epoch": 0.49435937106254796, + "grad_norm": 3.2463173866271973, + "learning_rate": 0.00010180829318345322, + "loss": 1.7175, + "step": 21089 + }, + { + "epoch": 0.49438281263735295, + "grad_norm": 3.6684811115264893, + "learning_rate": 0.00010180092822515976, + "loss": 0.3262, + "step": 21090 + }, + { + "epoch": 0.49440625421215795, + "grad_norm": 4.504047870635986, + "learning_rate": 0.00010179356325709441, + "loss": 0.4806, + "step": 21091 + }, + { + "epoch": 0.49442969578696294, + "grad_norm": 1.174654245376587, + "learning_rate": 0.00010178619827929712, + "loss": 0.1418, + "step": 21092 + }, + { + "epoch": 0.49445313736176794, + "grad_norm": 5.299684524536133, + "learning_rate": 0.00010177883329180788, + "loss": 0.668, + "step": 21093 + }, + { + "epoch": 0.494476578936573, + "grad_norm": 4.313796043395996, + "learning_rate": 0.00010177146829466666, + "loss": 0.421, + "step": 21094 + }, + { + "epoch": 0.494500020511378, + "grad_norm": 3.28778076171875, + "learning_rate": 0.00010176410328791332, + "loss": 0.2755, + "step": 21095 + }, + { + "epoch": 0.494523462086183, + "grad_norm": 1.7246612310409546, + "learning_rate": 0.00010175673827158798, + "loss": 0.2885, + "step": 21096 + }, + { + "epoch": 0.49454690366098797, + "grad_norm": 0.6469531059265137, + "learning_rate": 0.00010174937324573052, + "loss": 0.0308, + "step": 21097 + }, + { + "epoch": 0.49457034523579296, + "grad_norm": 1.7061967849731445, + "learning_rate": 0.00010174200821038092, + "loss": 0.171, + "step": 21098 + }, + { + "epoch": 0.49459378681059796, + "grad_norm": 2.807269334793091, + "learning_rate": 0.0001017346431655791, + "loss": 0.3073, + "step": 21099 + }, + { + "epoch": 0.49461722838540295, + "grad_norm": 2.5393595695495605, + "learning_rate": 0.00010172727811136508, + "loss": 0.2799, + "step": 21100 + }, + { + "epoch": 0.49464066996020795, + "grad_norm": 2.080033540725708, + "learning_rate": 0.00010171991304777879, + "loss": 0.1889, + "step": 21101 + }, + { + "epoch": 0.49466411153501294, + "grad_norm": 4.381195068359375, + "learning_rate": 0.00010171254797486021, + "loss": 0.544, + "step": 21102 + }, + { + "epoch": 0.49468755310981793, + "grad_norm": 0.9128807187080383, + "learning_rate": 0.0001017051828926493, + "loss": 0.0827, + "step": 21103 + }, + { + "epoch": 0.49471099468462293, + "grad_norm": 1.2638884782791138, + "learning_rate": 0.00010169781780118604, + "loss": 0.0953, + "step": 21104 + }, + { + "epoch": 0.4947344362594279, + "grad_norm": 2.423909902572632, + "learning_rate": 0.00010169045270051032, + "loss": 1.5962, + "step": 21105 + }, + { + "epoch": 0.4947578778342329, + "grad_norm": 1.415088415145874, + "learning_rate": 0.0001016830875906622, + "loss": 0.1369, + "step": 21106 + }, + { + "epoch": 0.4947813194090379, + "grad_norm": 3.199866533279419, + "learning_rate": 0.00010167572247168158, + "loss": 0.4428, + "step": 21107 + }, + { + "epoch": 0.4948047609838429, + "grad_norm": 2.701072931289673, + "learning_rate": 0.00010166835734360843, + "loss": 0.1472, + "step": 21108 + }, + { + "epoch": 0.4948282025586479, + "grad_norm": 4.516227722167969, + "learning_rate": 0.00010166099220648276, + "loss": 0.5297, + "step": 21109 + }, + { + "epoch": 0.4948516441334529, + "grad_norm": 2.4642350673675537, + "learning_rate": 0.0001016536270603445, + "loss": 0.1625, + "step": 21110 + }, + { + "epoch": 0.4948750857082579, + "grad_norm": 1.7617918252944946, + "learning_rate": 0.00010164626190523357, + "loss": 0.127, + "step": 21111 + }, + { + "epoch": 0.4948985272830629, + "grad_norm": 2.921935796737671, + "learning_rate": 0.00010163889674119, + "loss": 0.2965, + "step": 21112 + }, + { + "epoch": 0.4949219688578679, + "grad_norm": 2.8006603717803955, + "learning_rate": 0.00010163153156825373, + "loss": 0.2916, + "step": 21113 + }, + { + "epoch": 0.49494541043267287, + "grad_norm": 3.210697889328003, + "learning_rate": 0.00010162416638646472, + "loss": 1.9146, + "step": 21114 + }, + { + "epoch": 0.49496885200747787, + "grad_norm": 4.214261054992676, + "learning_rate": 0.00010161680119586295, + "loss": 0.2837, + "step": 21115 + }, + { + "epoch": 0.49499229358228286, + "grad_norm": 3.4162254333496094, + "learning_rate": 0.00010160943599648838, + "loss": 0.4472, + "step": 21116 + }, + { + "epoch": 0.49501573515708786, + "grad_norm": 3.529984951019287, + "learning_rate": 0.00010160207078838093, + "loss": 0.3093, + "step": 21117 + }, + { + "epoch": 0.49503917673189285, + "grad_norm": 0.6130467057228088, + "learning_rate": 0.00010159470557158064, + "loss": 0.0509, + "step": 21118 + }, + { + "epoch": 0.49506261830669784, + "grad_norm": 2.8040661811828613, + "learning_rate": 0.00010158734034612741, + "loss": 0.2958, + "step": 21119 + }, + { + "epoch": 0.49508605988150284, + "grad_norm": 1.33268404006958, + "learning_rate": 0.00010157997511206123, + "loss": 0.0664, + "step": 21120 + }, + { + "epoch": 0.49510950145630783, + "grad_norm": 1.3600616455078125, + "learning_rate": 0.00010157260986942204, + "loss": 0.1794, + "step": 21121 + }, + { + "epoch": 0.4951329430311128, + "grad_norm": 0.9083639979362488, + "learning_rate": 0.00010156524461824986, + "loss": 0.0607, + "step": 21122 + }, + { + "epoch": 0.4951563846059178, + "grad_norm": 3.092668056488037, + "learning_rate": 0.00010155787935858458, + "loss": 0.2539, + "step": 21123 + }, + { + "epoch": 0.4951798261807228, + "grad_norm": 3.2169487476348877, + "learning_rate": 0.00010155051409046623, + "loss": 0.3139, + "step": 21124 + }, + { + "epoch": 0.4952032677555278, + "grad_norm": 3.1828112602233887, + "learning_rate": 0.00010154314881393474, + "loss": 1.4947, + "step": 21125 + }, + { + "epoch": 0.4952267093303328, + "grad_norm": 2.3399300575256348, + "learning_rate": 0.00010153578352903008, + "loss": 0.2595, + "step": 21126 + }, + { + "epoch": 0.4952501509051378, + "grad_norm": 3.5061042308807373, + "learning_rate": 0.00010152841823579223, + "loss": 0.1981, + "step": 21127 + }, + { + "epoch": 0.4952735924799428, + "grad_norm": 5.371285438537598, + "learning_rate": 0.00010152105293426113, + "loss": 0.244, + "step": 21128 + }, + { + "epoch": 0.4952970340547478, + "grad_norm": 1.8080165386199951, + "learning_rate": 0.00010151368762447676, + "loss": 0.0647, + "step": 21129 + }, + { + "epoch": 0.4953204756295528, + "grad_norm": 3.322988748550415, + "learning_rate": 0.00010150632230647905, + "loss": 0.2378, + "step": 21130 + }, + { + "epoch": 0.4953439172043578, + "grad_norm": 2.3207149505615234, + "learning_rate": 0.000101498956980308, + "loss": 0.1879, + "step": 21131 + }, + { + "epoch": 0.49536735877916277, + "grad_norm": 2.7830049991607666, + "learning_rate": 0.00010149159164600359, + "loss": 0.2536, + "step": 21132 + }, + { + "epoch": 0.49539080035396776, + "grad_norm": 3.5788381099700928, + "learning_rate": 0.00010148422630360574, + "loss": 0.2149, + "step": 21133 + }, + { + "epoch": 0.49541424192877276, + "grad_norm": 4.582805156707764, + "learning_rate": 0.00010147686095315447, + "loss": 0.4274, + "step": 21134 + }, + { + "epoch": 0.49543768350357775, + "grad_norm": 2.58685564994812, + "learning_rate": 0.0001014694955946897, + "loss": 0.203, + "step": 21135 + }, + { + "epoch": 0.49546112507838275, + "grad_norm": 4.550820827484131, + "learning_rate": 0.0001014621302282514, + "loss": 0.3165, + "step": 21136 + }, + { + "epoch": 0.49548456665318774, + "grad_norm": 2.222121477127075, + "learning_rate": 0.0001014547648538795, + "loss": 0.212, + "step": 21137 + }, + { + "epoch": 0.49550800822799274, + "grad_norm": 4.460404872894287, + "learning_rate": 0.00010144739947161405, + "loss": 0.4007, + "step": 21138 + }, + { + "epoch": 0.49553144980279773, + "grad_norm": 2.321878433227539, + "learning_rate": 0.00010144003408149498, + "loss": 0.2751, + "step": 21139 + }, + { + "epoch": 0.4955548913776027, + "grad_norm": 5.337826251983643, + "learning_rate": 0.0001014326686835622, + "loss": 0.7036, + "step": 21140 + }, + { + "epoch": 0.4955783329524077, + "grad_norm": 4.287667751312256, + "learning_rate": 0.00010142530327785578, + "loss": 0.449, + "step": 21141 + }, + { + "epoch": 0.4956017745272127, + "grad_norm": 4.568943023681641, + "learning_rate": 0.00010141793786441556, + "loss": 0.4838, + "step": 21142 + }, + { + "epoch": 0.4956252161020177, + "grad_norm": 3.471170663833618, + "learning_rate": 0.0001014105724432816, + "loss": 0.3351, + "step": 21143 + }, + { + "epoch": 0.4956486576768227, + "grad_norm": 4.751067161560059, + "learning_rate": 0.00010140320701449384, + "loss": 0.42, + "step": 21144 + }, + { + "epoch": 0.49567209925162775, + "grad_norm": 2.3821909427642822, + "learning_rate": 0.00010139584157809224, + "loss": 0.2476, + "step": 21145 + }, + { + "epoch": 0.49569554082643275, + "grad_norm": 0.45548155903816223, + "learning_rate": 0.00010138847613411675, + "loss": 0.0442, + "step": 21146 + }, + { + "epoch": 0.49571898240123774, + "grad_norm": 5.487277507781982, + "learning_rate": 0.00010138111068260736, + "loss": 0.697, + "step": 21147 + }, + { + "epoch": 0.49574242397604273, + "grad_norm": 2.508232593536377, + "learning_rate": 0.00010137374522360404, + "loss": 0.1744, + "step": 21148 + }, + { + "epoch": 0.49576586555084773, + "grad_norm": 5.481882572174072, + "learning_rate": 0.00010136637975714669, + "loss": 0.5027, + "step": 21149 + }, + { + "epoch": 0.4957893071256527, + "grad_norm": 2.473780870437622, + "learning_rate": 0.00010135901428327534, + "loss": 1.6585, + "step": 21150 + }, + { + "epoch": 0.4958127487004577, + "grad_norm": 2.8392043113708496, + "learning_rate": 0.00010135164880202996, + "loss": 0.0593, + "step": 21151 + }, + { + "epoch": 0.4958361902752627, + "grad_norm": 4.025027751922607, + "learning_rate": 0.0001013442833134505, + "loss": 0.4638, + "step": 21152 + }, + { + "epoch": 0.4958596318500677, + "grad_norm": 2.253039836883545, + "learning_rate": 0.0001013369178175769, + "loss": 1.4677, + "step": 21153 + }, + { + "epoch": 0.4958830734248727, + "grad_norm": 3.2033944129943848, + "learning_rate": 0.00010132955231444919, + "loss": 0.2916, + "step": 21154 + }, + { + "epoch": 0.4959065149996777, + "grad_norm": 1.2976912260055542, + "learning_rate": 0.00010132218680410723, + "loss": 0.1307, + "step": 21155 + }, + { + "epoch": 0.4959299565744827, + "grad_norm": 0.6101981401443481, + "learning_rate": 0.00010131482128659106, + "loss": 0.0275, + "step": 21156 + }, + { + "epoch": 0.4959533981492877, + "grad_norm": 1.023355484008789, + "learning_rate": 0.00010130745576194068, + "loss": 0.8028, + "step": 21157 + }, + { + "epoch": 0.4959768397240927, + "grad_norm": 2.9675614833831787, + "learning_rate": 0.00010130009023019598, + "loss": 1.6971, + "step": 21158 + }, + { + "epoch": 0.49600028129889767, + "grad_norm": 3.222360372543335, + "learning_rate": 0.00010129272469139694, + "loss": 0.4463, + "step": 21159 + }, + { + "epoch": 0.49602372287370267, + "grad_norm": 2.6344799995422363, + "learning_rate": 0.00010128535914558357, + "loss": 0.1574, + "step": 21160 + }, + { + "epoch": 0.49604716444850766, + "grad_norm": 3.0049500465393066, + "learning_rate": 0.00010127799359279577, + "loss": 0.3737, + "step": 21161 + }, + { + "epoch": 0.49607060602331265, + "grad_norm": 2.5373682975769043, + "learning_rate": 0.00010127062803307354, + "loss": 1.4319, + "step": 21162 + }, + { + "epoch": 0.49609404759811765, + "grad_norm": 1.1938343048095703, + "learning_rate": 0.00010126326246645688, + "loss": 0.0517, + "step": 21163 + }, + { + "epoch": 0.49611748917292264, + "grad_norm": 3.5489704608917236, + "learning_rate": 0.00010125589689298571, + "loss": 0.3792, + "step": 21164 + }, + { + "epoch": 0.49614093074772764, + "grad_norm": 6.506783962249756, + "learning_rate": 0.0001012485313127, + "loss": 0.726, + "step": 21165 + }, + { + "epoch": 0.49616437232253263, + "grad_norm": 1.4867465496063232, + "learning_rate": 0.00010124116572563973, + "loss": 0.6501, + "step": 21166 + }, + { + "epoch": 0.4961878138973376, + "grad_norm": 1.3591339588165283, + "learning_rate": 0.00010123380013184487, + "loss": 0.1277, + "step": 21167 + }, + { + "epoch": 0.4962112554721426, + "grad_norm": 0.5786293148994446, + "learning_rate": 0.00010122643453135534, + "loss": 0.0372, + "step": 21168 + }, + { + "epoch": 0.4962346970469476, + "grad_norm": 2.497828722000122, + "learning_rate": 0.00010121906892421118, + "loss": 0.2622, + "step": 21169 + }, + { + "epoch": 0.4962581386217526, + "grad_norm": 3.3727667331695557, + "learning_rate": 0.00010121170331045231, + "loss": 0.3348, + "step": 21170 + }, + { + "epoch": 0.4962815801965576, + "grad_norm": 1.8827247619628906, + "learning_rate": 0.00010120433769011868, + "loss": 0.1203, + "step": 21171 + }, + { + "epoch": 0.4963050217713626, + "grad_norm": 1.0575367212295532, + "learning_rate": 0.00010119697206325031, + "loss": 0.1324, + "step": 21172 + }, + { + "epoch": 0.4963284633461676, + "grad_norm": 3.431049108505249, + "learning_rate": 0.00010118960642988712, + "loss": 0.2761, + "step": 21173 + }, + { + "epoch": 0.4963519049209726, + "grad_norm": 3.59785795211792, + "learning_rate": 0.0001011822407900691, + "loss": 0.3446, + "step": 21174 + }, + { + "epoch": 0.4963753464957776, + "grad_norm": 1.5298802852630615, + "learning_rate": 0.00010117487514383619, + "loss": 0.0974, + "step": 21175 + }, + { + "epoch": 0.4963987880705826, + "grad_norm": 5.060024261474609, + "learning_rate": 0.0001011675094912284, + "loss": 0.2146, + "step": 21176 + }, + { + "epoch": 0.49642222964538757, + "grad_norm": 3.430382490158081, + "learning_rate": 0.00010116014383228567, + "loss": 0.2385, + "step": 21177 + }, + { + "epoch": 0.49644567122019256, + "grad_norm": 2.463665246963501, + "learning_rate": 0.00010115277816704794, + "loss": 0.2917, + "step": 21178 + }, + { + "epoch": 0.49646911279499756, + "grad_norm": 2.2607014179229736, + "learning_rate": 0.00010114541249555523, + "loss": 0.0789, + "step": 21179 + }, + { + "epoch": 0.49649255436980255, + "grad_norm": 2.097160816192627, + "learning_rate": 0.00010113804681784744, + "loss": 0.9385, + "step": 21180 + }, + { + "epoch": 0.49651599594460755, + "grad_norm": 2.825620412826538, + "learning_rate": 0.0001011306811339646, + "loss": 0.2617, + "step": 21181 + }, + { + "epoch": 0.49653943751941254, + "grad_norm": 4.395534038543701, + "learning_rate": 0.00010112331544394667, + "loss": 0.2759, + "step": 21182 + }, + { + "epoch": 0.49656287909421754, + "grad_norm": 4.081974029541016, + "learning_rate": 0.00010111594974783361, + "loss": 0.4847, + "step": 21183 + }, + { + "epoch": 0.49658632066902253, + "grad_norm": 3.2760989665985107, + "learning_rate": 0.00010110858404566533, + "loss": 1.6489, + "step": 21184 + }, + { + "epoch": 0.4966097622438275, + "grad_norm": 2.1144204139709473, + "learning_rate": 0.00010110121833748188, + "loss": 0.2114, + "step": 21185 + }, + { + "epoch": 0.4966332038186325, + "grad_norm": 2.749671220779419, + "learning_rate": 0.00010109385262332314, + "loss": 0.3422, + "step": 21186 + }, + { + "epoch": 0.4966566453934375, + "grad_norm": 0.7727717757225037, + "learning_rate": 0.00010108648690322915, + "loss": 0.0291, + "step": 21187 + }, + { + "epoch": 0.4966800869682425, + "grad_norm": 3.792029857635498, + "learning_rate": 0.00010107912117723987, + "loss": 0.3512, + "step": 21188 + }, + { + "epoch": 0.4967035285430475, + "grad_norm": 2.834461212158203, + "learning_rate": 0.00010107175544539524, + "loss": 0.3827, + "step": 21189 + }, + { + "epoch": 0.4967269701178525, + "grad_norm": 3.362464189529419, + "learning_rate": 0.00010106438970773522, + "loss": 0.3253, + "step": 21190 + }, + { + "epoch": 0.4967504116926575, + "grad_norm": 3.2172586917877197, + "learning_rate": 0.00010105702396429981, + "loss": 0.4926, + "step": 21191 + }, + { + "epoch": 0.4967738532674625, + "grad_norm": 4.029160022735596, + "learning_rate": 0.00010104965821512895, + "loss": 0.4749, + "step": 21192 + }, + { + "epoch": 0.4967972948422675, + "grad_norm": 1.5653537511825562, + "learning_rate": 0.00010104229246026262, + "loss": 0.1364, + "step": 21193 + }, + { + "epoch": 0.4968207364170725, + "grad_norm": 2.843140125274658, + "learning_rate": 0.00010103492669974077, + "loss": 1.2013, + "step": 21194 + }, + { + "epoch": 0.49684417799187747, + "grad_norm": 1.9557188749313354, + "learning_rate": 0.00010102756093360339, + "loss": 0.2496, + "step": 21195 + }, + { + "epoch": 0.4968676195666825, + "grad_norm": 4.022300720214844, + "learning_rate": 0.00010102019516189042, + "loss": 0.405, + "step": 21196 + }, + { + "epoch": 0.4968910611414875, + "grad_norm": 2.589276075363159, + "learning_rate": 0.00010101282938464187, + "loss": 0.192, + "step": 21197 + }, + { + "epoch": 0.4969145027162925, + "grad_norm": 2.775404691696167, + "learning_rate": 0.00010100546360189765, + "loss": 0.2095, + "step": 21198 + }, + { + "epoch": 0.4969379442910975, + "grad_norm": 2.475977897644043, + "learning_rate": 0.00010099809781369777, + "loss": 0.2712, + "step": 21199 + }, + { + "epoch": 0.4969613858659025, + "grad_norm": 2.3926024436950684, + "learning_rate": 0.00010099073202008218, + "loss": 1.6757, + "step": 21200 + }, + { + "epoch": 0.4969848274407075, + "grad_norm": 2.833810806274414, + "learning_rate": 0.00010098336622109085, + "loss": 0.222, + "step": 21201 + }, + { + "epoch": 0.4970082690155125, + "grad_norm": 2.623870611190796, + "learning_rate": 0.00010097600041676375, + "loss": 0.3413, + "step": 21202 + }, + { + "epoch": 0.4970317105903175, + "grad_norm": 3.1047208309173584, + "learning_rate": 0.00010096863460714086, + "loss": 0.1652, + "step": 21203 + }, + { + "epoch": 0.49705515216512247, + "grad_norm": 1.8896621465682983, + "learning_rate": 0.00010096126879226209, + "loss": 0.2529, + "step": 21204 + }, + { + "epoch": 0.49707859373992747, + "grad_norm": 4.427923202514648, + "learning_rate": 0.00010095390297216749, + "loss": 0.4216, + "step": 21205 + }, + { + "epoch": 0.49710203531473246, + "grad_norm": 0.7593860030174255, + "learning_rate": 0.00010094653714689694, + "loss": 0.0827, + "step": 21206 + }, + { + "epoch": 0.49712547688953745, + "grad_norm": 2.7372355461120605, + "learning_rate": 0.00010093917131649051, + "loss": 0.1711, + "step": 21207 + }, + { + "epoch": 0.49714891846434245, + "grad_norm": 3.9138808250427246, + "learning_rate": 0.00010093180548098809, + "loss": 0.3385, + "step": 21208 + }, + { + "epoch": 0.49717236003914744, + "grad_norm": 3.6260807514190674, + "learning_rate": 0.00010092443964042964, + "loss": 0.2343, + "step": 21209 + }, + { + "epoch": 0.49719580161395244, + "grad_norm": 0.9929012656211853, + "learning_rate": 0.00010091707379485515, + "loss": 0.1024, + "step": 21210 + }, + { + "epoch": 0.49721924318875743, + "grad_norm": 3.862032890319824, + "learning_rate": 0.00010090970794430464, + "loss": 0.4156, + "step": 21211 + }, + { + "epoch": 0.4972426847635624, + "grad_norm": 0.9108270406723022, + "learning_rate": 0.00010090234208881801, + "loss": 0.0836, + "step": 21212 + }, + { + "epoch": 0.4972661263383674, + "grad_norm": 1.290794014930725, + "learning_rate": 0.00010089497622843522, + "loss": 0.1468, + "step": 21213 + }, + { + "epoch": 0.4972895679131724, + "grad_norm": 3.3232429027557373, + "learning_rate": 0.00010088761036319632, + "loss": 0.3946, + "step": 21214 + }, + { + "epoch": 0.4973130094879774, + "grad_norm": 2.986842632293701, + "learning_rate": 0.0001008802444931412, + "loss": 0.2734, + "step": 21215 + }, + { + "epoch": 0.4973364510627824, + "grad_norm": 2.9639856815338135, + "learning_rate": 0.00010087287861830979, + "loss": 0.3264, + "step": 21216 + }, + { + "epoch": 0.4973598926375874, + "grad_norm": 3.058734893798828, + "learning_rate": 0.00010086551273874216, + "loss": 1.2653, + "step": 21217 + }, + { + "epoch": 0.4973833342123924, + "grad_norm": 3.538358688354492, + "learning_rate": 0.00010085814685447825, + "loss": 0.2916, + "step": 21218 + }, + { + "epoch": 0.4974067757871974, + "grad_norm": 3.529508590698242, + "learning_rate": 0.00010085078096555798, + "loss": 0.3696, + "step": 21219 + }, + { + "epoch": 0.4974302173620024, + "grad_norm": 3.934537649154663, + "learning_rate": 0.00010084341507202137, + "loss": 0.3299, + "step": 21220 + }, + { + "epoch": 0.4974536589368074, + "grad_norm": 4.573088645935059, + "learning_rate": 0.00010083604917390839, + "loss": 0.4327, + "step": 21221 + }, + { + "epoch": 0.49747710051161237, + "grad_norm": 4.386691093444824, + "learning_rate": 0.00010082868327125894, + "loss": 0.1739, + "step": 21222 + }, + { + "epoch": 0.49750054208641736, + "grad_norm": 1.0109829902648926, + "learning_rate": 0.00010082131736411302, + "loss": 0.1132, + "step": 21223 + }, + { + "epoch": 0.49752398366122236, + "grad_norm": 0.9044709801673889, + "learning_rate": 0.00010081395145251065, + "loss": 0.0708, + "step": 21224 + }, + { + "epoch": 0.49754742523602735, + "grad_norm": 2.4577739238739014, + "learning_rate": 0.00010080658553649174, + "loss": 0.2962, + "step": 21225 + }, + { + "epoch": 0.49757086681083235, + "grad_norm": 3.0585150718688965, + "learning_rate": 0.00010079921961609629, + "loss": 0.2475, + "step": 21226 + }, + { + "epoch": 0.49759430838563734, + "grad_norm": 1.1664276123046875, + "learning_rate": 0.00010079185369136424, + "loss": 0.0776, + "step": 21227 + }, + { + "epoch": 0.49761774996044233, + "grad_norm": 2.482424736022949, + "learning_rate": 0.00010078448776233556, + "loss": 0.1338, + "step": 21228 + }, + { + "epoch": 0.49764119153524733, + "grad_norm": 2.546781301498413, + "learning_rate": 0.00010077712182905021, + "loss": 0.2332, + "step": 21229 + }, + { + "epoch": 0.4976646331100523, + "grad_norm": 2.2776007652282715, + "learning_rate": 0.0001007697558915482, + "loss": 0.3946, + "step": 21230 + }, + { + "epoch": 0.4976880746848573, + "grad_norm": 1.8246550559997559, + "learning_rate": 0.00010076238994986948, + "loss": 0.0987, + "step": 21231 + }, + { + "epoch": 0.4977115162596623, + "grad_norm": 2.8315865993499756, + "learning_rate": 0.00010075502400405399, + "loss": 0.3128, + "step": 21232 + }, + { + "epoch": 0.4977349578344673, + "grad_norm": 3.764038562774658, + "learning_rate": 0.00010074765805414175, + "loss": 0.2522, + "step": 21233 + }, + { + "epoch": 0.4977583994092723, + "grad_norm": 3.715836763381958, + "learning_rate": 0.00010074029210017268, + "loss": 0.3033, + "step": 21234 + }, + { + "epoch": 0.4977818409840773, + "grad_norm": 1.4898934364318848, + "learning_rate": 0.00010073292614218674, + "loss": 0.1344, + "step": 21235 + }, + { + "epoch": 0.4978052825588823, + "grad_norm": 3.8335299491882324, + "learning_rate": 0.00010072556018022394, + "loss": 0.3529, + "step": 21236 + }, + { + "epoch": 0.4978287241336873, + "grad_norm": 1.3280991315841675, + "learning_rate": 0.00010071819421432425, + "loss": 0.1186, + "step": 21237 + }, + { + "epoch": 0.4978521657084923, + "grad_norm": 3.3701839447021484, + "learning_rate": 0.0001007108282445276, + "loss": 0.2977, + "step": 21238 + }, + { + "epoch": 0.4978756072832973, + "grad_norm": 2.984492063522339, + "learning_rate": 0.00010070346227087397, + "loss": 0.3028, + "step": 21239 + }, + { + "epoch": 0.49789904885810227, + "grad_norm": 3.8327012062072754, + "learning_rate": 0.00010069609629340336, + "loss": 0.5989, + "step": 21240 + }, + { + "epoch": 0.49792249043290726, + "grad_norm": 3.268336057662964, + "learning_rate": 0.00010068873031215566, + "loss": 0.5378, + "step": 21241 + }, + { + "epoch": 0.49794593200771226, + "grad_norm": 4.783053398132324, + "learning_rate": 0.00010068136432717095, + "loss": 0.2999, + "step": 21242 + }, + { + "epoch": 0.49796937358251725, + "grad_norm": 3.1582024097442627, + "learning_rate": 0.00010067399833848911, + "loss": 0.3495, + "step": 21243 + }, + { + "epoch": 0.49799281515732224, + "grad_norm": 4.787650108337402, + "learning_rate": 0.00010066663234615012, + "loss": 0.5335, + "step": 21244 + }, + { + "epoch": 0.49801625673212724, + "grad_norm": 0.9456210136413574, + "learning_rate": 0.000100659266350194, + "loss": 0.0877, + "step": 21245 + }, + { + "epoch": 0.49803969830693223, + "grad_norm": 5.615272045135498, + "learning_rate": 0.00010065190035066068, + "loss": 0.3269, + "step": 21246 + }, + { + "epoch": 0.4980631398817373, + "grad_norm": 3.6616506576538086, + "learning_rate": 0.0001006445343475901, + "loss": 0.3352, + "step": 21247 + }, + { + "epoch": 0.4980865814565423, + "grad_norm": 4.086386680603027, + "learning_rate": 0.00010063716834102225, + "loss": 0.3087, + "step": 21248 + }, + { + "epoch": 0.49811002303134727, + "grad_norm": 2.7702436447143555, + "learning_rate": 0.00010062980233099715, + "loss": 0.2881, + "step": 21249 + }, + { + "epoch": 0.49813346460615227, + "grad_norm": 3.1840388774871826, + "learning_rate": 0.0001006224363175547, + "loss": 0.3693, + "step": 21250 + }, + { + "epoch": 0.49815690618095726, + "grad_norm": 3.7342844009399414, + "learning_rate": 0.00010061507030073489, + "loss": 0.4323, + "step": 21251 + }, + { + "epoch": 0.49818034775576225, + "grad_norm": 2.754997730255127, + "learning_rate": 0.00010060770428057771, + "loss": 1.3101, + "step": 21252 + }, + { + "epoch": 0.49820378933056725, + "grad_norm": 3.311537027359009, + "learning_rate": 0.0001006003382571231, + "loss": 0.3181, + "step": 21253 + }, + { + "epoch": 0.49822723090537224, + "grad_norm": 3.4526960849761963, + "learning_rate": 0.00010059297223041102, + "loss": 0.3409, + "step": 21254 + }, + { + "epoch": 0.49825067248017724, + "grad_norm": 2.997955799102783, + "learning_rate": 0.00010058560620048147, + "loss": 0.154, + "step": 21255 + }, + { + "epoch": 0.49827411405498223, + "grad_norm": 0.47519463300704956, + "learning_rate": 0.00010057824016737443, + "loss": 0.0412, + "step": 21256 + }, + { + "epoch": 0.4982975556297872, + "grad_norm": 1.5469005107879639, + "learning_rate": 0.0001005708741311298, + "loss": 0.1189, + "step": 21257 + }, + { + "epoch": 0.4983209972045922, + "grad_norm": 2.485339403152466, + "learning_rate": 0.0001005635080917876, + "loss": 0.2097, + "step": 21258 + }, + { + "epoch": 0.4983444387793972, + "grad_norm": 3.967806577682495, + "learning_rate": 0.00010055614204938781, + "loss": 0.4033, + "step": 21259 + }, + { + "epoch": 0.4983678803542022, + "grad_norm": 5.530417442321777, + "learning_rate": 0.00010054877600397035, + "loss": 0.4136, + "step": 21260 + }, + { + "epoch": 0.4983913219290072, + "grad_norm": 4.5046868324279785, + "learning_rate": 0.00010054140995557524, + "loss": 0.6386, + "step": 21261 + }, + { + "epoch": 0.4984147635038122, + "grad_norm": 3.860154151916504, + "learning_rate": 0.00010053404390424243, + "loss": 0.48, + "step": 21262 + }, + { + "epoch": 0.4984382050786172, + "grad_norm": 1.9639567136764526, + "learning_rate": 0.00010052667785001186, + "loss": 0.1932, + "step": 21263 + }, + { + "epoch": 0.4984616466534222, + "grad_norm": 1.5169899463653564, + "learning_rate": 0.00010051931179292353, + "loss": 0.1433, + "step": 21264 + }, + { + "epoch": 0.4984850882282272, + "grad_norm": 1.5506211519241333, + "learning_rate": 0.00010051194573301738, + "loss": 0.8384, + "step": 21265 + }, + { + "epoch": 0.4985085298030322, + "grad_norm": 3.2644169330596924, + "learning_rate": 0.00010050457967033345, + "loss": 0.1903, + "step": 21266 + }, + { + "epoch": 0.49853197137783717, + "grad_norm": 1.5558342933654785, + "learning_rate": 0.0001004972136049116, + "loss": 0.22, + "step": 21267 + }, + { + "epoch": 0.49855541295264216, + "grad_norm": 2.2964675426483154, + "learning_rate": 0.0001004898475367919, + "loss": 0.2777, + "step": 21268 + }, + { + "epoch": 0.49857885452744716, + "grad_norm": 2.5463123321533203, + "learning_rate": 0.00010048248146601426, + "loss": 0.2005, + "step": 21269 + }, + { + "epoch": 0.49860229610225215, + "grad_norm": 5.019364833831787, + "learning_rate": 0.00010047511539261864, + "loss": 0.7475, + "step": 21270 + }, + { + "epoch": 0.49862573767705715, + "grad_norm": 2.9687914848327637, + "learning_rate": 0.00010046774931664504, + "loss": 0.2149, + "step": 21271 + }, + { + "epoch": 0.49864917925186214, + "grad_norm": 1.4724998474121094, + "learning_rate": 0.00010046038323813343, + "loss": 0.1403, + "step": 21272 + }, + { + "epoch": 0.49867262082666713, + "grad_norm": 3.9601900577545166, + "learning_rate": 0.00010045301715712377, + "loss": 0.4735, + "step": 21273 + }, + { + "epoch": 0.49869606240147213, + "grad_norm": 0.52092045545578, + "learning_rate": 0.00010044565107365603, + "loss": 0.0356, + "step": 21274 + }, + { + "epoch": 0.4987195039762771, + "grad_norm": 0.4208599627017975, + "learning_rate": 0.00010043828498777015, + "loss": 0.0507, + "step": 21275 + }, + { + "epoch": 0.4987429455510821, + "grad_norm": 2.9897334575653076, + "learning_rate": 0.00010043091889950613, + "loss": 0.1733, + "step": 21276 + }, + { + "epoch": 0.4987663871258871, + "grad_norm": 3.9127259254455566, + "learning_rate": 0.00010042355280890392, + "loss": 0.3146, + "step": 21277 + }, + { + "epoch": 0.4987898287006921, + "grad_norm": 1.9531382322311401, + "learning_rate": 0.00010041618671600355, + "loss": 0.2155, + "step": 21278 + }, + { + "epoch": 0.4988132702754971, + "grad_norm": 0.6942286491394043, + "learning_rate": 0.0001004088206208449, + "loss": 0.0538, + "step": 21279 + }, + { + "epoch": 0.4988367118503021, + "grad_norm": 3.231900453567505, + "learning_rate": 0.00010040145452346798, + "loss": 0.3574, + "step": 21280 + }, + { + "epoch": 0.4988601534251071, + "grad_norm": 3.5689609050750732, + "learning_rate": 0.00010039408842391278, + "loss": 0.4043, + "step": 21281 + }, + { + "epoch": 0.4988835949999121, + "grad_norm": 3.004333019256592, + "learning_rate": 0.00010038672232221922, + "loss": 0.2068, + "step": 21282 + }, + { + "epoch": 0.4989070365747171, + "grad_norm": 1.5479122400283813, + "learning_rate": 0.0001003793562184273, + "loss": 0.2115, + "step": 21283 + }, + { + "epoch": 0.49893047814952207, + "grad_norm": 2.450505256652832, + "learning_rate": 0.00010037199011257699, + "loss": 0.1817, + "step": 21284 + }, + { + "epoch": 0.49895391972432707, + "grad_norm": 1.5222219228744507, + "learning_rate": 0.00010036462400470825, + "loss": 0.1276, + "step": 21285 + }, + { + "epoch": 0.49897736129913206, + "grad_norm": 4.101844310760498, + "learning_rate": 0.00010035725789486104, + "loss": 0.4973, + "step": 21286 + }, + { + "epoch": 0.49900080287393705, + "grad_norm": 4.065077304840088, + "learning_rate": 0.00010034989178307535, + "loss": 0.3564, + "step": 21287 + }, + { + "epoch": 0.49902424444874205, + "grad_norm": 3.476147174835205, + "learning_rate": 0.00010034252566939114, + "loss": 0.2889, + "step": 21288 + }, + { + "epoch": 0.49904768602354704, + "grad_norm": 3.713719606399536, + "learning_rate": 0.00010033515955384833, + "loss": 0.4773, + "step": 21289 + }, + { + "epoch": 0.49907112759835204, + "grad_norm": 1.704952359199524, + "learning_rate": 0.00010032779343648697, + "loss": 1.0034, + "step": 21290 + }, + { + "epoch": 0.49909456917315703, + "grad_norm": 2.1042582988739014, + "learning_rate": 0.00010032042731734701, + "loss": 0.1063, + "step": 21291 + }, + { + "epoch": 0.499118010747962, + "grad_norm": 1.5069636106491089, + "learning_rate": 0.00010031306119646837, + "loss": 0.8011, + "step": 21292 + }, + { + "epoch": 0.499141452322767, + "grad_norm": 2.6771392822265625, + "learning_rate": 0.0001003056950738911, + "loss": 1.5162, + "step": 21293 + }, + { + "epoch": 0.499164893897572, + "grad_norm": 1.0805644989013672, + "learning_rate": 0.0001002983289496551, + "loss": 0.1226, + "step": 21294 + }, + { + "epoch": 0.499188335472377, + "grad_norm": 3.798799753189087, + "learning_rate": 0.00010029096282380033, + "loss": 0.4775, + "step": 21295 + }, + { + "epoch": 0.499211777047182, + "grad_norm": 2.006241798400879, + "learning_rate": 0.00010028359669636677, + "loss": 1.5069, + "step": 21296 + }, + { + "epoch": 0.499235218621987, + "grad_norm": 2.6893532276153564, + "learning_rate": 0.00010027623056739448, + "loss": 1.4951, + "step": 21297 + }, + { + "epoch": 0.49925866019679205, + "grad_norm": 3.884953260421753, + "learning_rate": 0.0001002688644369233, + "loss": 0.2561, + "step": 21298 + }, + { + "epoch": 0.49928210177159704, + "grad_norm": 3.5568034648895264, + "learning_rate": 0.00010026149830499328, + "loss": 0.3224, + "step": 21299 + }, + { + "epoch": 0.49930554334640204, + "grad_norm": 1.3170993328094482, + "learning_rate": 0.00010025413217164436, + "loss": 0.1272, + "step": 21300 + }, + { + "epoch": 0.49932898492120703, + "grad_norm": 1.1892281770706177, + "learning_rate": 0.00010024676603691649, + "loss": 0.1313, + "step": 21301 + }, + { + "epoch": 0.499352426496012, + "grad_norm": 3.111100196838379, + "learning_rate": 0.00010023939990084967, + "loss": 0.3366, + "step": 21302 + }, + { + "epoch": 0.499375868070817, + "grad_norm": 1.8822816610336304, + "learning_rate": 0.00010023203376348387, + "loss": 0.1009, + "step": 21303 + }, + { + "epoch": 0.499399309645622, + "grad_norm": 2.0280354022979736, + "learning_rate": 0.00010022466762485905, + "loss": 0.2247, + "step": 21304 + }, + { + "epoch": 0.499422751220427, + "grad_norm": 3.548043966293335, + "learning_rate": 0.00010021730148501515, + "loss": 0.4279, + "step": 21305 + }, + { + "epoch": 0.499446192795232, + "grad_norm": 4.414106845855713, + "learning_rate": 0.00010020993534399222, + "loss": 0.5969, + "step": 21306 + }, + { + "epoch": 0.499469634370037, + "grad_norm": 1.5248600244522095, + "learning_rate": 0.00010020256920183016, + "loss": 0.1289, + "step": 21307 + }, + { + "epoch": 0.499493075944842, + "grad_norm": 3.239039421081543, + "learning_rate": 0.00010019520305856889, + "loss": 0.2488, + "step": 21308 + }, + { + "epoch": 0.499516517519647, + "grad_norm": 2.5319244861602783, + "learning_rate": 0.0001001878369142485, + "loss": 0.2845, + "step": 21309 + }, + { + "epoch": 0.499539959094452, + "grad_norm": 3.457866668701172, + "learning_rate": 0.0001001804707689089, + "loss": 0.301, + "step": 21310 + }, + { + "epoch": 0.499563400669257, + "grad_norm": 1.062696099281311, + "learning_rate": 0.00010017310462259004, + "loss": 0.1125, + "step": 21311 + }, + { + "epoch": 0.49958684224406197, + "grad_norm": 4.445420742034912, + "learning_rate": 0.00010016573847533192, + "loss": 0.4731, + "step": 21312 + }, + { + "epoch": 0.49961028381886696, + "grad_norm": 5.207907676696777, + "learning_rate": 0.00010015837232717452, + "loss": 0.6144, + "step": 21313 + }, + { + "epoch": 0.49963372539367196, + "grad_norm": 1.793948769569397, + "learning_rate": 0.00010015100617815775, + "loss": 0.1336, + "step": 21314 + }, + { + "epoch": 0.49965716696847695, + "grad_norm": 2.6954896450042725, + "learning_rate": 0.00010014364002832162, + "loss": 0.3064, + "step": 21315 + }, + { + "epoch": 0.49968060854328195, + "grad_norm": 10.503077507019043, + "learning_rate": 0.00010013627387770612, + "loss": 0.2917, + "step": 21316 + }, + { + "epoch": 0.49970405011808694, + "grad_norm": 2.8688244819641113, + "learning_rate": 0.00010012890772635117, + "loss": 1.8419, + "step": 21317 + }, + { + "epoch": 0.49972749169289193, + "grad_norm": 1.6889958381652832, + "learning_rate": 0.00010012154157429679, + "loss": 0.1542, + "step": 21318 + }, + { + "epoch": 0.49975093326769693, + "grad_norm": 2.20125150680542, + "learning_rate": 0.0001001141754215829, + "loss": 0.1875, + "step": 21319 + }, + { + "epoch": 0.4997743748425019, + "grad_norm": 4.092122554779053, + "learning_rate": 0.00010010680926824949, + "loss": 0.4487, + "step": 21320 + }, + { + "epoch": 0.4997978164173069, + "grad_norm": 1.3267416954040527, + "learning_rate": 0.00010009944311433652, + "loss": 0.1755, + "step": 21321 + }, + { + "epoch": 0.4998212579921119, + "grad_norm": 8.217644691467285, + "learning_rate": 0.00010009207695988401, + "loss": 0.6142, + "step": 21322 + }, + { + "epoch": 0.4998446995669169, + "grad_norm": 4.9622907638549805, + "learning_rate": 0.00010008471080493187, + "loss": 0.4994, + "step": 21323 + }, + { + "epoch": 0.4998681411417219, + "grad_norm": 0.6150484681129456, + "learning_rate": 0.00010007734464952008, + "loss": 0.0588, + "step": 21324 + }, + { + "epoch": 0.4998915827165269, + "grad_norm": 4.736047267913818, + "learning_rate": 0.00010006997849368862, + "loss": 0.3792, + "step": 21325 + }, + { + "epoch": 0.4999150242913319, + "grad_norm": 2.7175865173339844, + "learning_rate": 0.00010006261233747743, + "loss": 0.1542, + "step": 21326 + }, + { + "epoch": 0.4999384658661369, + "grad_norm": 4.420478343963623, + "learning_rate": 0.00010005524618092652, + "loss": 0.4165, + "step": 21327 + }, + { + "epoch": 0.4999619074409419, + "grad_norm": 1.0065487623214722, + "learning_rate": 0.00010004788002407585, + "loss": 0.1336, + "step": 21328 + }, + { + "epoch": 0.49998534901574687, + "grad_norm": 2.758286476135254, + "learning_rate": 0.0001000405138669654, + "loss": 0.2157, + "step": 21329 + }, + { + "epoch": 0.5000087905905519, + "grad_norm": 3.8770980834960938, + "learning_rate": 0.00010003314770963507, + "loss": 0.4541, + "step": 21330 + }, + { + "epoch": 0.5000322321653569, + "grad_norm": 3.464383125305176, + "learning_rate": 0.00010002578155212492, + "loss": 0.2205, + "step": 21331 + }, + { + "epoch": 0.5000556737401619, + "grad_norm": 1.862094759941101, + "learning_rate": 0.00010001841539447487, + "loss": 0.1415, + "step": 21332 + }, + { + "epoch": 0.5000791153149668, + "grad_norm": 0.8415228724479675, + "learning_rate": 0.0001000110492367249, + "loss": 0.0408, + "step": 21333 + }, + { + "epoch": 0.5001025568897719, + "grad_norm": 3.393789529800415, + "learning_rate": 0.00010000368307891496, + "loss": 0.2343, + "step": 21334 + }, + { + "epoch": 0.5001259984645768, + "grad_norm": 1.9455926418304443, + "learning_rate": 9.999631692108506e-05, + "loss": 0.7601, + "step": 21335 + }, + { + "epoch": 0.5001494400393819, + "grad_norm": 0.5840550661087036, + "learning_rate": 9.998895076327512e-05, + "loss": 0.0727, + "step": 21336 + }, + { + "epoch": 0.5001728816141868, + "grad_norm": 2.1833393573760986, + "learning_rate": 9.998158460552515e-05, + "loss": 0.2507, + "step": 21337 + }, + { + "epoch": 0.5001963231889919, + "grad_norm": 2.0856738090515137, + "learning_rate": 9.99742184478751e-05, + "loss": 0.1501, + "step": 21338 + }, + { + "epoch": 0.5002197647637968, + "grad_norm": 0.5260106921195984, + "learning_rate": 9.996685229036495e-05, + "loss": 0.0432, + "step": 21339 + }, + { + "epoch": 0.5002432063386019, + "grad_norm": 3.685887575149536, + "learning_rate": 9.995948613303465e-05, + "loss": 0.3168, + "step": 21340 + }, + { + "epoch": 0.5002666479134068, + "grad_norm": 2.338989496231079, + "learning_rate": 9.995211997592414e-05, + "loss": 0.1826, + "step": 21341 + }, + { + "epoch": 0.5002900894882119, + "grad_norm": 4.891212463378906, + "learning_rate": 9.994475381907349e-05, + "loss": 0.3324, + "step": 21342 + }, + { + "epoch": 0.5003135310630168, + "grad_norm": 3.875333547592163, + "learning_rate": 9.993738766252257e-05, + "loss": 0.3403, + "step": 21343 + }, + { + "epoch": 0.5003369726378218, + "grad_norm": 1.9234989881515503, + "learning_rate": 9.993002150631141e-05, + "loss": 0.2457, + "step": 21344 + }, + { + "epoch": 0.5003604142126268, + "grad_norm": 1.3223220109939575, + "learning_rate": 9.992265535047995e-05, + "loss": 0.2575, + "step": 21345 + }, + { + "epoch": 0.5003838557874318, + "grad_norm": 3.7688753604888916, + "learning_rate": 9.991528919506816e-05, + "loss": 0.54, + "step": 21346 + }, + { + "epoch": 0.5004072973622368, + "grad_norm": 2.4447638988494873, + "learning_rate": 9.990792304011601e-05, + "loss": 1.5281, + "step": 21347 + }, + { + "epoch": 0.5004307389370418, + "grad_norm": 3.5211215019226074, + "learning_rate": 9.99005568856635e-05, + "loss": 0.4695, + "step": 21348 + }, + { + "epoch": 0.5004541805118468, + "grad_norm": 4.492210388183594, + "learning_rate": 9.989319073175053e-05, + "loss": 0.4472, + "step": 21349 + }, + { + "epoch": 0.5004776220866518, + "grad_norm": 1.6712981462478638, + "learning_rate": 9.988582457841714e-05, + "loss": 0.2168, + "step": 21350 + }, + { + "epoch": 0.5005010636614567, + "grad_norm": 2.4700520038604736, + "learning_rate": 9.987845842570326e-05, + "loss": 0.2983, + "step": 21351 + }, + { + "epoch": 0.5005245052362618, + "grad_norm": 1.7197654247283936, + "learning_rate": 9.987109227364885e-05, + "loss": 0.2466, + "step": 21352 + }, + { + "epoch": 0.5005479468110667, + "grad_norm": 3.617279052734375, + "learning_rate": 9.986372612229389e-05, + "loss": 0.5641, + "step": 21353 + }, + { + "epoch": 0.5005713883858718, + "grad_norm": 2.9563827514648438, + "learning_rate": 9.985635997167837e-05, + "loss": 0.2659, + "step": 21354 + }, + { + "epoch": 0.5005948299606767, + "grad_norm": 4.565617561340332, + "learning_rate": 9.984899382184227e-05, + "loss": 0.4749, + "step": 21355 + }, + { + "epoch": 0.5006182715354818, + "grad_norm": 2.032003879547119, + "learning_rate": 9.984162767282551e-05, + "loss": 0.8366, + "step": 21356 + }, + { + "epoch": 0.5006417131102867, + "grad_norm": 0.29108983278274536, + "learning_rate": 9.98342615246681e-05, + "loss": 0.0094, + "step": 21357 + }, + { + "epoch": 0.5006651546850918, + "grad_norm": 4.239631175994873, + "learning_rate": 9.982689537741e-05, + "loss": 0.5777, + "step": 21358 + }, + { + "epoch": 0.5006885962598967, + "grad_norm": 0.7614045739173889, + "learning_rate": 9.981952923109112e-05, + "loss": 0.0655, + "step": 21359 + }, + { + "epoch": 0.5007120378347018, + "grad_norm": 3.0968260765075684, + "learning_rate": 9.981216308575151e-05, + "loss": 0.1876, + "step": 21360 + }, + { + "epoch": 0.5007354794095067, + "grad_norm": 2.284541368484497, + "learning_rate": 9.980479694143114e-05, + "loss": 0.1978, + "step": 21361 + }, + { + "epoch": 0.5007589209843117, + "grad_norm": 2.4582319259643555, + "learning_rate": 9.97974307981699e-05, + "loss": 0.2934, + "step": 21362 + }, + { + "epoch": 0.5007823625591167, + "grad_norm": 3.086643695831299, + "learning_rate": 9.979006465600783e-05, + "loss": 0.28, + "step": 21363 + }, + { + "epoch": 0.5008058041339217, + "grad_norm": 2.4997763633728027, + "learning_rate": 9.978269851498487e-05, + "loss": 1.2899, + "step": 21364 + }, + { + "epoch": 0.5008292457087267, + "grad_norm": 2.4358692169189453, + "learning_rate": 9.977533237514096e-05, + "loss": 0.0987, + "step": 21365 + }, + { + "epoch": 0.5008526872835317, + "grad_norm": 2.0292484760284424, + "learning_rate": 9.976796623651613e-05, + "loss": 0.1694, + "step": 21366 + }, + { + "epoch": 0.5008761288583367, + "grad_norm": 2.575035333633423, + "learning_rate": 9.976060009915034e-05, + "loss": 0.2741, + "step": 21367 + }, + { + "epoch": 0.5008995704331417, + "grad_norm": 2.395461320877075, + "learning_rate": 9.975323396308353e-05, + "loss": 0.158, + "step": 21368 + }, + { + "epoch": 0.5009230120079466, + "grad_norm": 4.103787899017334, + "learning_rate": 9.974586782835566e-05, + "loss": 0.2514, + "step": 21369 + }, + { + "epoch": 0.5009464535827517, + "grad_norm": 2.7757441997528076, + "learning_rate": 9.973850169500675e-05, + "loss": 0.2049, + "step": 21370 + }, + { + "epoch": 0.5009698951575566, + "grad_norm": 0.5157278180122375, + "learning_rate": 9.973113556307673e-05, + "loss": 0.0545, + "step": 21371 + }, + { + "epoch": 0.5009933367323617, + "grad_norm": 4.07223653793335, + "learning_rate": 9.972376943260555e-05, + "loss": 0.4401, + "step": 21372 + }, + { + "epoch": 0.5010167783071666, + "grad_norm": 3.973466396331787, + "learning_rate": 9.971640330363324e-05, + "loss": 0.2543, + "step": 21373 + }, + { + "epoch": 0.5010402198819717, + "grad_norm": 4.00796365737915, + "learning_rate": 9.970903717619969e-05, + "loss": 0.3391, + "step": 21374 + }, + { + "epoch": 0.5010636614567767, + "grad_norm": 2.5452568531036377, + "learning_rate": 9.970167105034496e-05, + "loss": 1.5773, + "step": 21375 + }, + { + "epoch": 0.5010871030315817, + "grad_norm": 2.860950231552124, + "learning_rate": 9.969430492610896e-05, + "loss": 1.5172, + "step": 21376 + }, + { + "epoch": 0.5011105446063867, + "grad_norm": 3.8030807971954346, + "learning_rate": 9.968693880353165e-05, + "loss": 0.5126, + "step": 21377 + }, + { + "epoch": 0.5011339861811916, + "grad_norm": 5.000507831573486, + "learning_rate": 9.967957268265299e-05, + "loss": 0.4367, + "step": 21378 + }, + { + "epoch": 0.5011574277559967, + "grad_norm": 3.9714505672454834, + "learning_rate": 9.967220656351303e-05, + "loss": 0.2784, + "step": 21379 + }, + { + "epoch": 0.5011808693308016, + "grad_norm": 1.2424606084823608, + "learning_rate": 9.966484044615168e-05, + "loss": 0.1327, + "step": 21380 + }, + { + "epoch": 0.5012043109056067, + "grad_norm": 8.08551025390625, + "learning_rate": 9.965747433060889e-05, + "loss": 0.6648, + "step": 21381 + }, + { + "epoch": 0.5012277524804116, + "grad_norm": 2.76139760017395, + "learning_rate": 9.965010821692469e-05, + "loss": 0.2863, + "step": 21382 + }, + { + "epoch": 0.5012511940552167, + "grad_norm": 4.562452793121338, + "learning_rate": 9.9642742105139e-05, + "loss": 0.3598, + "step": 21383 + }, + { + "epoch": 0.5012746356300216, + "grad_norm": 3.8477859497070312, + "learning_rate": 9.963537599529178e-05, + "loss": 0.3283, + "step": 21384 + }, + { + "epoch": 0.5012980772048267, + "grad_norm": 0.6307849884033203, + "learning_rate": 9.962800988742305e-05, + "loss": 0.0543, + "step": 21385 + }, + { + "epoch": 0.5013215187796316, + "grad_norm": 4.0777587890625, + "learning_rate": 9.962064378157274e-05, + "loss": 0.2633, + "step": 21386 + }, + { + "epoch": 0.5013449603544367, + "grad_norm": 0.4343324899673462, + "learning_rate": 9.961327767778082e-05, + "loss": 0.0253, + "step": 21387 + }, + { + "epoch": 0.5013684019292416, + "grad_norm": 2.4978015422821045, + "learning_rate": 9.960591157608727e-05, + "loss": 0.2548, + "step": 21388 + }, + { + "epoch": 0.5013918435040466, + "grad_norm": 2.8648288249969482, + "learning_rate": 9.959854547653206e-05, + "loss": 0.2564, + "step": 21389 + }, + { + "epoch": 0.5014152850788516, + "grad_norm": 2.3665454387664795, + "learning_rate": 9.959117937915511e-05, + "loss": 0.2696, + "step": 21390 + }, + { + "epoch": 0.5014387266536566, + "grad_norm": 3.4728946685791016, + "learning_rate": 9.958381328399646e-05, + "loss": 0.2247, + "step": 21391 + }, + { + "epoch": 0.5014621682284616, + "grad_norm": 5.155776500701904, + "learning_rate": 9.957644719109607e-05, + "loss": 0.5415, + "step": 21392 + }, + { + "epoch": 0.5014856098032666, + "grad_norm": 1.9889274835586548, + "learning_rate": 9.956908110049388e-05, + "loss": 0.2095, + "step": 21393 + }, + { + "epoch": 0.5015090513780716, + "grad_norm": 8.103718757629395, + "learning_rate": 9.956171501222986e-05, + "loss": 0.3144, + "step": 21394 + }, + { + "epoch": 0.5015324929528766, + "grad_norm": 3.75604248046875, + "learning_rate": 9.955434892634401e-05, + "loss": 0.6081, + "step": 21395 + }, + { + "epoch": 0.5015559345276815, + "grad_norm": 3.532968282699585, + "learning_rate": 9.954698284287627e-05, + "loss": 0.3714, + "step": 21396 + }, + { + "epoch": 0.5015793761024866, + "grad_norm": 3.4002532958984375, + "learning_rate": 9.953961676186658e-05, + "loss": 0.4816, + "step": 21397 + }, + { + "epoch": 0.5016028176772915, + "grad_norm": 3.077965259552002, + "learning_rate": 9.953225068335498e-05, + "loss": 0.1873, + "step": 21398 + }, + { + "epoch": 0.5016262592520966, + "grad_norm": 1.4529880285263062, + "learning_rate": 9.95248846073814e-05, + "loss": 0.0863, + "step": 21399 + }, + { + "epoch": 0.5016497008269015, + "grad_norm": 3.54276967048645, + "learning_rate": 9.951751853398579e-05, + "loss": 0.4123, + "step": 21400 + }, + { + "epoch": 0.5016731424017066, + "grad_norm": 5.082706451416016, + "learning_rate": 9.951015246320816e-05, + "loss": 0.5587, + "step": 21401 + }, + { + "epoch": 0.5016965839765115, + "grad_norm": 4.262426376342773, + "learning_rate": 9.950278639508838e-05, + "loss": 0.4134, + "step": 21402 + }, + { + "epoch": 0.5017200255513166, + "grad_norm": 1.1868181228637695, + "learning_rate": 9.949542032966658e-05, + "loss": 0.1748, + "step": 21403 + }, + { + "epoch": 0.5017434671261215, + "grad_norm": 4.158404350280762, + "learning_rate": 9.94880542669826e-05, + "loss": 0.2408, + "step": 21404 + }, + { + "epoch": 0.5017669087009266, + "grad_norm": 2.627211332321167, + "learning_rate": 9.94806882070765e-05, + "loss": 1.8537, + "step": 21405 + }, + { + "epoch": 0.5017903502757315, + "grad_norm": 3.205965042114258, + "learning_rate": 9.947332214998816e-05, + "loss": 0.4506, + "step": 21406 + }, + { + "epoch": 0.5018137918505365, + "grad_norm": 2.094764232635498, + "learning_rate": 9.946595609575759e-05, + "loss": 0.2108, + "step": 21407 + }, + { + "epoch": 0.5018372334253415, + "grad_norm": 2.2593533992767334, + "learning_rate": 9.945859004442479e-05, + "loss": 0.2068, + "step": 21408 + }, + { + "epoch": 0.5018606750001465, + "grad_norm": 3.6740360260009766, + "learning_rate": 9.945122399602967e-05, + "loss": 0.2007, + "step": 21409 + }, + { + "epoch": 0.5018841165749515, + "grad_norm": 2.4500184059143066, + "learning_rate": 9.944385795061222e-05, + "loss": 0.255, + "step": 21410 + }, + { + "epoch": 0.5019075581497565, + "grad_norm": 4.218637466430664, + "learning_rate": 9.943649190821242e-05, + "loss": 0.2935, + "step": 21411 + }, + { + "epoch": 0.5019309997245615, + "grad_norm": 1.5792219638824463, + "learning_rate": 9.942912586887022e-05, + "loss": 0.0888, + "step": 21412 + }, + { + "epoch": 0.5019544412993665, + "grad_norm": 1.557827353477478, + "learning_rate": 9.942175983262563e-05, + "loss": 1.0034, + "step": 21413 + }, + { + "epoch": 0.5019778828741714, + "grad_norm": 4.5440144538879395, + "learning_rate": 9.941439379951853e-05, + "loss": 0.4747, + "step": 21414 + }, + { + "epoch": 0.5020013244489765, + "grad_norm": 0.9623569250106812, + "learning_rate": 9.940702776958899e-05, + "loss": 0.052, + "step": 21415 + }, + { + "epoch": 0.5020247660237814, + "grad_norm": 3.921769380569458, + "learning_rate": 9.939966174287691e-05, + "loss": 0.4425, + "step": 21416 + }, + { + "epoch": 0.5020482075985865, + "grad_norm": 3.112635612487793, + "learning_rate": 9.939229571942231e-05, + "loss": 0.3039, + "step": 21417 + }, + { + "epoch": 0.5020716491733914, + "grad_norm": 3.6876742839813232, + "learning_rate": 9.938492969926513e-05, + "loss": 0.5956, + "step": 21418 + }, + { + "epoch": 0.5020950907481965, + "grad_norm": 4.396071910858154, + "learning_rate": 9.937756368244532e-05, + "loss": 0.4955, + "step": 21419 + }, + { + "epoch": 0.5021185323230014, + "grad_norm": 1.9951916933059692, + "learning_rate": 9.937019766900289e-05, + "loss": 0.8687, + "step": 21420 + }, + { + "epoch": 0.5021419738978065, + "grad_norm": 2.2668964862823486, + "learning_rate": 9.936283165897777e-05, + "loss": 0.1973, + "step": 21421 + }, + { + "epoch": 0.5021654154726114, + "grad_norm": 2.1910300254821777, + "learning_rate": 9.935546565240993e-05, + "loss": 0.1547, + "step": 21422 + }, + { + "epoch": 0.5021888570474164, + "grad_norm": 3.4029486179351807, + "learning_rate": 9.934809964933938e-05, + "loss": 0.2445, + "step": 21423 + }, + { + "epoch": 0.5022122986222214, + "grad_norm": 1.2859328985214233, + "learning_rate": 9.934073364980605e-05, + "loss": 0.1245, + "step": 21424 + }, + { + "epoch": 0.5022357401970264, + "grad_norm": 4.638745307922363, + "learning_rate": 9.93333676538499e-05, + "loss": 0.4362, + "step": 21425 + }, + { + "epoch": 0.5022591817718315, + "grad_norm": 1.7070043087005615, + "learning_rate": 9.93260016615109e-05, + "loss": 0.1518, + "step": 21426 + }, + { + "epoch": 0.5022826233466364, + "grad_norm": 2.8185040950775146, + "learning_rate": 9.931863567282906e-05, + "loss": 0.1731, + "step": 21427 + }, + { + "epoch": 0.5023060649214415, + "grad_norm": 1.8666225671768188, + "learning_rate": 9.931126968784433e-05, + "loss": 0.0943, + "step": 21428 + }, + { + "epoch": 0.5023295064962464, + "grad_norm": 2.296689510345459, + "learning_rate": 9.930390370659665e-05, + "loss": 1.3056, + "step": 21429 + }, + { + "epoch": 0.5023529480710515, + "grad_norm": 0.7233662605285645, + "learning_rate": 9.929653772912604e-05, + "loss": 0.0382, + "step": 21430 + }, + { + "epoch": 0.5023763896458564, + "grad_norm": 1.7023093700408936, + "learning_rate": 9.928917175547244e-05, + "loss": 0.1552, + "step": 21431 + }, + { + "epoch": 0.5023998312206615, + "grad_norm": 2.0009992122650146, + "learning_rate": 9.928180578567577e-05, + "loss": 1.3214, + "step": 21432 + }, + { + "epoch": 0.5024232727954664, + "grad_norm": 1.9453712701797485, + "learning_rate": 9.927443981977608e-05, + "loss": 0.1807, + "step": 21433 + }, + { + "epoch": 0.5024467143702714, + "grad_norm": 5.754812240600586, + "learning_rate": 9.926707385781328e-05, + "loss": 2.1398, + "step": 21434 + }, + { + "epoch": 0.5024701559450764, + "grad_norm": 1.841915488243103, + "learning_rate": 9.925970789982736e-05, + "loss": 0.1124, + "step": 21435 + }, + { + "epoch": 0.5024935975198814, + "grad_norm": 2.8337972164154053, + "learning_rate": 9.92523419458583e-05, + "loss": 0.2425, + "step": 21436 + }, + { + "epoch": 0.5025170390946864, + "grad_norm": 3.1334128379821777, + "learning_rate": 9.924497599594605e-05, + "loss": 0.3161, + "step": 21437 + }, + { + "epoch": 0.5025404806694914, + "grad_norm": 1.032364010810852, + "learning_rate": 9.923761005013054e-05, + "loss": 0.0982, + "step": 21438 + }, + { + "epoch": 0.5025639222442964, + "grad_norm": 4.344804763793945, + "learning_rate": 9.923024410845179e-05, + "loss": 0.3496, + "step": 21439 + }, + { + "epoch": 0.5025873638191014, + "grad_norm": 2.8635640144348145, + "learning_rate": 9.92228781709498e-05, + "loss": 0.2756, + "step": 21440 + }, + { + "epoch": 0.5026108053939063, + "grad_norm": 2.3607499599456787, + "learning_rate": 9.921551223766448e-05, + "loss": 0.2179, + "step": 21441 + }, + { + "epoch": 0.5026342469687114, + "grad_norm": 4.108090400695801, + "learning_rate": 9.920814630863578e-05, + "loss": 0.3913, + "step": 21442 + }, + { + "epoch": 0.5026576885435163, + "grad_norm": 0.5624514222145081, + "learning_rate": 9.920078038390375e-05, + "loss": 0.0553, + "step": 21443 + }, + { + "epoch": 0.5026811301183214, + "grad_norm": 3.6082301139831543, + "learning_rate": 9.91934144635083e-05, + "loss": 0.2443, + "step": 21444 + }, + { + "epoch": 0.5027045716931263, + "grad_norm": 1.389589786529541, + "learning_rate": 9.918604854748937e-05, + "loss": 0.8404, + "step": 21445 + }, + { + "epoch": 0.5027280132679314, + "grad_norm": 3.396756172180176, + "learning_rate": 9.9178682635887e-05, + "loss": 0.2628, + "step": 21446 + }, + { + "epoch": 0.5027514548427363, + "grad_norm": 0.9690312743186951, + "learning_rate": 9.91713167287411e-05, + "loss": 0.1172, + "step": 21447 + }, + { + "epoch": 0.5027748964175414, + "grad_norm": 3.9395956993103027, + "learning_rate": 9.916395082609166e-05, + "loss": 0.1476, + "step": 21448 + }, + { + "epoch": 0.5027983379923463, + "grad_norm": 2.919572591781616, + "learning_rate": 9.915658492797866e-05, + "loss": 0.3489, + "step": 21449 + }, + { + "epoch": 0.5028217795671514, + "grad_norm": 3.917825698852539, + "learning_rate": 9.914921903444201e-05, + "loss": 0.4425, + "step": 21450 + }, + { + "epoch": 0.5028452211419563, + "grad_norm": 2.9488115310668945, + "learning_rate": 9.914185314552176e-05, + "loss": 0.3017, + "step": 21451 + }, + { + "epoch": 0.5028686627167613, + "grad_norm": 3.0724685192108154, + "learning_rate": 9.913448726125783e-05, + "loss": 1.4157, + "step": 21452 + }, + { + "epoch": 0.5028921042915663, + "grad_norm": 2.459395170211792, + "learning_rate": 9.912712138169021e-05, + "loss": 1.4433, + "step": 21453 + }, + { + "epoch": 0.5029155458663713, + "grad_norm": 2.282890796661377, + "learning_rate": 9.911975550685884e-05, + "loss": 0.9275, + "step": 21454 + }, + { + "epoch": 0.5029389874411763, + "grad_norm": 5.294217109680176, + "learning_rate": 9.911238963680372e-05, + "loss": 0.6048, + "step": 21455 + }, + { + "epoch": 0.5029624290159813, + "grad_norm": 0.4540344178676605, + "learning_rate": 9.910502377156479e-05, + "loss": 0.0395, + "step": 21456 + }, + { + "epoch": 0.5029858705907863, + "grad_norm": 3.6322240829467773, + "learning_rate": 9.909765791118201e-05, + "loss": 0.2986, + "step": 21457 + }, + { + "epoch": 0.5030093121655913, + "grad_norm": 0.6727731227874756, + "learning_rate": 9.909029205569537e-05, + "loss": 0.0803, + "step": 21458 + }, + { + "epoch": 0.5030327537403962, + "grad_norm": 0.435702919960022, + "learning_rate": 9.908292620514486e-05, + "loss": 0.0479, + "step": 21459 + }, + { + "epoch": 0.5030561953152013, + "grad_norm": 3.213419198989868, + "learning_rate": 9.907556035957039e-05, + "loss": 0.385, + "step": 21460 + }, + { + "epoch": 0.5030796368900062, + "grad_norm": 2.5869388580322266, + "learning_rate": 9.906819451901196e-05, + "loss": 0.1289, + "step": 21461 + }, + { + "epoch": 0.5031030784648113, + "grad_norm": 2.729030132293701, + "learning_rate": 9.906082868350954e-05, + "loss": 0.2135, + "step": 21462 + }, + { + "epoch": 0.5031265200396162, + "grad_norm": 3.4686357975006104, + "learning_rate": 9.905346285310306e-05, + "loss": 0.3865, + "step": 21463 + }, + { + "epoch": 0.5031499616144213, + "grad_norm": 5.551064491271973, + "learning_rate": 9.904609702783252e-05, + "loss": 0.4401, + "step": 21464 + }, + { + "epoch": 0.5031734031892262, + "grad_norm": 5.577131271362305, + "learning_rate": 9.903873120773791e-05, + "loss": 0.3653, + "step": 21465 + }, + { + "epoch": 0.5031968447640313, + "grad_norm": 0.8114488124847412, + "learning_rate": 9.903136539285918e-05, + "loss": 0.1152, + "step": 21466 + }, + { + "epoch": 0.5032202863388362, + "grad_norm": 3.048241376876831, + "learning_rate": 9.902399958323626e-05, + "loss": 0.2519, + "step": 21467 + }, + { + "epoch": 0.5032437279136412, + "grad_norm": 2.6419754028320312, + "learning_rate": 9.901663377890917e-05, + "loss": 0.2668, + "step": 21468 + }, + { + "epoch": 0.5032671694884462, + "grad_norm": 3.4244306087493896, + "learning_rate": 9.900926797991785e-05, + "loss": 0.3249, + "step": 21469 + }, + { + "epoch": 0.5032906110632512, + "grad_norm": 1.5849536657333374, + "learning_rate": 9.900190218630225e-05, + "loss": 0.2393, + "step": 21470 + }, + { + "epoch": 0.5033140526380562, + "grad_norm": 1.0144904851913452, + "learning_rate": 9.899453639810238e-05, + "loss": 0.0722, + "step": 21471 + }, + { + "epoch": 0.5033374942128612, + "grad_norm": 2.9496116638183594, + "learning_rate": 9.898717061535818e-05, + "loss": 0.2828, + "step": 21472 + }, + { + "epoch": 0.5033609357876662, + "grad_norm": 2.272021532058716, + "learning_rate": 9.89798048381096e-05, + "loss": 0.242, + "step": 21473 + }, + { + "epoch": 0.5033843773624712, + "grad_norm": 2.674006938934326, + "learning_rate": 9.897243906639666e-05, + "loss": 0.1624, + "step": 21474 + }, + { + "epoch": 0.5034078189372762, + "grad_norm": 5.595478534698486, + "learning_rate": 9.896507330025924e-05, + "loss": 0.4819, + "step": 21475 + }, + { + "epoch": 0.5034312605120812, + "grad_norm": 1.1823546886444092, + "learning_rate": 9.895770753973739e-05, + "loss": 0.0907, + "step": 21476 + }, + { + "epoch": 0.5034547020868863, + "grad_norm": 4.80626916885376, + "learning_rate": 9.895034178487106e-05, + "loss": 0.3989, + "step": 21477 + }, + { + "epoch": 0.5034781436616912, + "grad_norm": 1.1784157752990723, + "learning_rate": 9.894297603570021e-05, + "loss": 0.1341, + "step": 21478 + }, + { + "epoch": 0.5035015852364962, + "grad_norm": 3.5755443572998047, + "learning_rate": 9.89356102922648e-05, + "loss": 0.2782, + "step": 21479 + }, + { + "epoch": 0.5035250268113012, + "grad_norm": 2.776320219039917, + "learning_rate": 9.892824455460478e-05, + "loss": 0.234, + "step": 21480 + }, + { + "epoch": 0.5035484683861062, + "grad_norm": 1.513292908668518, + "learning_rate": 9.892087882276015e-05, + "loss": 0.0515, + "step": 21481 + }, + { + "epoch": 0.5035719099609112, + "grad_norm": 2.3005940914154053, + "learning_rate": 9.891351309677087e-05, + "loss": 1.328, + "step": 21482 + }, + { + "epoch": 0.5035953515357162, + "grad_norm": 3.1414897441864014, + "learning_rate": 9.890614737667687e-05, + "loss": 0.2427, + "step": 21483 + }, + { + "epoch": 0.5036187931105212, + "grad_norm": 1.4617449045181274, + "learning_rate": 9.889878166251819e-05, + "loss": 0.0842, + "step": 21484 + }, + { + "epoch": 0.5036422346853262, + "grad_norm": 8.019043922424316, + "learning_rate": 9.88914159543347e-05, + "loss": 0.4124, + "step": 21485 + }, + { + "epoch": 0.5036656762601311, + "grad_norm": 2.9890296459198, + "learning_rate": 9.888405025216645e-05, + "loss": 0.2472, + "step": 21486 + }, + { + "epoch": 0.5036891178349362, + "grad_norm": 4.643558502197266, + "learning_rate": 9.887668455605333e-05, + "loss": 0.2153, + "step": 21487 + }, + { + "epoch": 0.5037125594097411, + "grad_norm": 1.4578269720077515, + "learning_rate": 9.88693188660354e-05, + "loss": 0.0697, + "step": 21488 + }, + { + "epoch": 0.5037360009845462, + "grad_norm": 1.3707325458526611, + "learning_rate": 9.886195318215257e-05, + "loss": 0.1256, + "step": 21489 + }, + { + "epoch": 0.5037594425593511, + "grad_norm": 1.4526416063308716, + "learning_rate": 9.885458750444479e-05, + "loss": 0.0715, + "step": 21490 + }, + { + "epoch": 0.5037828841341562, + "grad_norm": 4.183679580688477, + "learning_rate": 9.884722183295208e-05, + "loss": 2.2584, + "step": 21491 + }, + { + "epoch": 0.5038063257089611, + "grad_norm": 2.544985294342041, + "learning_rate": 9.883985616771437e-05, + "loss": 0.2152, + "step": 21492 + }, + { + "epoch": 0.5038297672837662, + "grad_norm": 3.2182457447052, + "learning_rate": 9.883249050877163e-05, + "loss": 1.8235, + "step": 21493 + }, + { + "epoch": 0.5038532088585711, + "grad_norm": 2.758126735687256, + "learning_rate": 9.882512485616384e-05, + "loss": 0.2308, + "step": 21494 + }, + { + "epoch": 0.5038766504333761, + "grad_norm": 3.9462594985961914, + "learning_rate": 9.881775920993092e-05, + "loss": 0.3232, + "step": 21495 + }, + { + "epoch": 0.5039000920081811, + "grad_norm": 1.727856993675232, + "learning_rate": 9.881039357011291e-05, + "loss": 0.0871, + "step": 21496 + }, + { + "epoch": 0.5039235335829861, + "grad_norm": 0.7361617684364319, + "learning_rate": 9.880302793674974e-05, + "loss": 0.0708, + "step": 21497 + }, + { + "epoch": 0.5039469751577911, + "grad_norm": 3.1295673847198486, + "learning_rate": 9.879566230988134e-05, + "loss": 1.9548, + "step": 21498 + }, + { + "epoch": 0.5039704167325961, + "grad_norm": 2.3860700130462646, + "learning_rate": 9.87882966895477e-05, + "loss": 0.2302, + "step": 21499 + }, + { + "epoch": 0.5039938583074011, + "grad_norm": 2.650144100189209, + "learning_rate": 9.878093107578883e-05, + "loss": 0.2615, + "step": 21500 + }, + { + "epoch": 0.5040172998822061, + "grad_norm": 1.7517353296279907, + "learning_rate": 9.877356546864467e-05, + "loss": 0.1225, + "step": 21501 + }, + { + "epoch": 0.504040741457011, + "grad_norm": 3.9320826530456543, + "learning_rate": 9.876619986815516e-05, + "loss": 1.3224, + "step": 21502 + }, + { + "epoch": 0.5040641830318161, + "grad_norm": 2.5247092247009277, + "learning_rate": 9.875883427436028e-05, + "loss": 0.2263, + "step": 21503 + }, + { + "epoch": 0.504087624606621, + "grad_norm": 4.064460277557373, + "learning_rate": 9.875146868730002e-05, + "loss": 0.4316, + "step": 21504 + }, + { + "epoch": 0.5041110661814261, + "grad_norm": 5.8230414390563965, + "learning_rate": 9.87441031070143e-05, + "loss": 0.2964, + "step": 21505 + }, + { + "epoch": 0.504134507756231, + "grad_norm": 5.139202117919922, + "learning_rate": 9.873673753354315e-05, + "loss": 0.4182, + "step": 21506 + }, + { + "epoch": 0.5041579493310361, + "grad_norm": 2.549852132797241, + "learning_rate": 9.872937196692648e-05, + "loss": 0.1795, + "step": 21507 + }, + { + "epoch": 0.504181390905841, + "grad_norm": 4.053520202636719, + "learning_rate": 9.872200640720424e-05, + "loss": 0.2738, + "step": 21508 + }, + { + "epoch": 0.5042048324806461, + "grad_norm": 1.945417046546936, + "learning_rate": 9.871464085441647e-05, + "loss": 0.1387, + "step": 21509 + }, + { + "epoch": 0.504228274055451, + "grad_norm": 5.922982215881348, + "learning_rate": 9.870727530860309e-05, + "loss": 0.516, + "step": 21510 + }, + { + "epoch": 0.5042517156302561, + "grad_norm": 0.5347540974617004, + "learning_rate": 9.869990976980403e-05, + "loss": 0.062, + "step": 21511 + }, + { + "epoch": 0.504275157205061, + "grad_norm": 4.0393500328063965, + "learning_rate": 9.869254423805932e-05, + "loss": 0.4647, + "step": 21512 + }, + { + "epoch": 0.504298598779866, + "grad_norm": 3.521897077560425, + "learning_rate": 9.868517871340892e-05, + "loss": 0.238, + "step": 21513 + }, + { + "epoch": 0.504322040354671, + "grad_norm": 1.9705355167388916, + "learning_rate": 9.867781319589278e-05, + "loss": 0.2518, + "step": 21514 + }, + { + "epoch": 0.504345481929476, + "grad_norm": 2.302932024002075, + "learning_rate": 9.867044768555084e-05, + "loss": 0.1148, + "step": 21515 + }, + { + "epoch": 0.504368923504281, + "grad_norm": 0.6480237245559692, + "learning_rate": 9.866308218242311e-05, + "loss": 0.0537, + "step": 21516 + }, + { + "epoch": 0.504392365079086, + "grad_norm": 2.8569176197052, + "learning_rate": 9.865571668654953e-05, + "loss": 0.2368, + "step": 21517 + }, + { + "epoch": 0.504415806653891, + "grad_norm": 4.62930154800415, + "learning_rate": 9.864835119797005e-05, + "loss": 0.4579, + "step": 21518 + }, + { + "epoch": 0.504439248228696, + "grad_norm": 4.581308841705322, + "learning_rate": 9.864098571672468e-05, + "loss": 0.5258, + "step": 21519 + }, + { + "epoch": 0.504462689803501, + "grad_norm": 1.3822932243347168, + "learning_rate": 9.863362024285335e-05, + "loss": 0.0512, + "step": 21520 + }, + { + "epoch": 0.504486131378306, + "grad_norm": 1.926741600036621, + "learning_rate": 9.862625477639603e-05, + "loss": 0.0985, + "step": 21521 + }, + { + "epoch": 0.5045095729531109, + "grad_norm": 3.1677489280700684, + "learning_rate": 9.861888931739269e-05, + "loss": 1.5431, + "step": 21522 + }, + { + "epoch": 0.504533014527916, + "grad_norm": 2.387833595275879, + "learning_rate": 9.861152386588326e-05, + "loss": 0.182, + "step": 21523 + }, + { + "epoch": 0.5045564561027209, + "grad_norm": 3.6081104278564453, + "learning_rate": 9.860415842190779e-05, + "loss": 0.4501, + "step": 21524 + }, + { + "epoch": 0.504579897677526, + "grad_norm": 2.433750629425049, + "learning_rate": 9.859679298550617e-05, + "loss": 0.0987, + "step": 21525 + }, + { + "epoch": 0.5046033392523309, + "grad_norm": 4.790223121643066, + "learning_rate": 9.858942755671841e-05, + "loss": 0.3318, + "step": 21526 + }, + { + "epoch": 0.504626780827136, + "grad_norm": 0.7095457911491394, + "learning_rate": 9.858206213558445e-05, + "loss": 0.0722, + "step": 21527 + }, + { + "epoch": 0.504650222401941, + "grad_norm": 0.8323898911476135, + "learning_rate": 9.857469672214426e-05, + "loss": 0.088, + "step": 21528 + }, + { + "epoch": 0.504673663976746, + "grad_norm": 5.095459461212158, + "learning_rate": 9.85673313164378e-05, + "loss": 0.2532, + "step": 21529 + }, + { + "epoch": 0.504697105551551, + "grad_norm": 4.590455532073975, + "learning_rate": 9.855996591850504e-05, + "loss": 0.329, + "step": 21530 + }, + { + "epoch": 0.504720547126356, + "grad_norm": 2.925825834274292, + "learning_rate": 9.855260052838596e-05, + "loss": 0.2142, + "step": 21531 + }, + { + "epoch": 0.504743988701161, + "grad_norm": 5.316653251647949, + "learning_rate": 9.854523514612051e-05, + "loss": 0.2316, + "step": 21532 + }, + { + "epoch": 0.5047674302759659, + "grad_norm": 4.317012310028076, + "learning_rate": 9.853786977174864e-05, + "loss": 0.2493, + "step": 21533 + }, + { + "epoch": 0.504790871850771, + "grad_norm": 1.0846410989761353, + "learning_rate": 9.853050440531035e-05, + "loss": 0.168, + "step": 21534 + }, + { + "epoch": 0.5048143134255759, + "grad_norm": 3.056788206100464, + "learning_rate": 9.852313904684558e-05, + "loss": 0.22, + "step": 21535 + }, + { + "epoch": 0.504837755000381, + "grad_norm": 4.1628594398498535, + "learning_rate": 9.851577369639424e-05, + "loss": 0.2875, + "step": 21536 + }, + { + "epoch": 0.5048611965751859, + "grad_norm": 2.063316583633423, + "learning_rate": 9.85084083539964e-05, + "loss": 1.3713, + "step": 21537 + }, + { + "epoch": 0.504884638149991, + "grad_norm": 1.8009511232376099, + "learning_rate": 9.850104301969199e-05, + "loss": 0.0369, + "step": 21538 + }, + { + "epoch": 0.5049080797247959, + "grad_norm": 4.391161918640137, + "learning_rate": 9.849367769352096e-05, + "loss": 0.3223, + "step": 21539 + }, + { + "epoch": 0.504931521299601, + "grad_norm": 8.678637504577637, + "learning_rate": 9.848631237552325e-05, + "loss": 1.0145, + "step": 21540 + }, + { + "epoch": 0.5049549628744059, + "grad_norm": 3.8116302490234375, + "learning_rate": 9.847894706573889e-05, + "loss": 0.4007, + "step": 21541 + }, + { + "epoch": 0.5049784044492109, + "grad_norm": 4.107347011566162, + "learning_rate": 9.84715817642078e-05, + "loss": 0.2012, + "step": 21542 + }, + { + "epoch": 0.5050018460240159, + "grad_norm": 3.683929204940796, + "learning_rate": 9.846421647096993e-05, + "loss": 0.5476, + "step": 21543 + }, + { + "epoch": 0.5050252875988209, + "grad_norm": 1.4577057361602783, + "learning_rate": 9.845685118606528e-05, + "loss": 0.1237, + "step": 21544 + }, + { + "epoch": 0.5050487291736259, + "grad_norm": 4.203850746154785, + "learning_rate": 9.84494859095338e-05, + "loss": 0.3245, + "step": 21545 + }, + { + "epoch": 0.5050721707484309, + "grad_norm": 0.9398291707038879, + "learning_rate": 9.844212064141544e-05, + "loss": 0.0658, + "step": 21546 + }, + { + "epoch": 0.5050956123232359, + "grad_norm": 3.117506504058838, + "learning_rate": 9.843475538175019e-05, + "loss": 1.9657, + "step": 21547 + }, + { + "epoch": 0.5051190538980409, + "grad_norm": 3.987555980682373, + "learning_rate": 9.842739013057796e-05, + "loss": 0.406, + "step": 21548 + }, + { + "epoch": 0.5051424954728458, + "grad_norm": 1.3263232707977295, + "learning_rate": 9.84200248879388e-05, + "loss": 0.127, + "step": 21549 + }, + { + "epoch": 0.5051659370476509, + "grad_norm": 4.299317836761475, + "learning_rate": 9.84126596538726e-05, + "loss": 0.3081, + "step": 21550 + }, + { + "epoch": 0.5051893786224558, + "grad_norm": 2.2872543334960938, + "learning_rate": 9.840529442841939e-05, + "loss": 0.1903, + "step": 21551 + }, + { + "epoch": 0.5052128201972609, + "grad_norm": 2.912144184112549, + "learning_rate": 9.83979292116191e-05, + "loss": 0.3653, + "step": 21552 + }, + { + "epoch": 0.5052362617720658, + "grad_norm": 5.357004165649414, + "learning_rate": 9.839056400351164e-05, + "loss": 0.6244, + "step": 21553 + }, + { + "epoch": 0.5052597033468709, + "grad_norm": 2.355327844619751, + "learning_rate": 9.838319880413706e-05, + "loss": 0.2641, + "step": 21554 + }, + { + "epoch": 0.5052831449216758, + "grad_norm": 2.3930160999298096, + "learning_rate": 9.83758336135353e-05, + "loss": 0.157, + "step": 21555 + }, + { + "epoch": 0.5053065864964809, + "grad_norm": 3.509605646133423, + "learning_rate": 9.836846843174629e-05, + "loss": 0.4238, + "step": 21556 + }, + { + "epoch": 0.5053300280712858, + "grad_norm": 2.9552366733551025, + "learning_rate": 9.836110325881003e-05, + "loss": 0.252, + "step": 21557 + }, + { + "epoch": 0.5053534696460908, + "grad_norm": 2.220247268676758, + "learning_rate": 9.835373809476648e-05, + "loss": 0.2385, + "step": 21558 + }, + { + "epoch": 0.5053769112208958, + "grad_norm": 3.7115073204040527, + "learning_rate": 9.834637293965557e-05, + "loss": 0.2034, + "step": 21559 + }, + { + "epoch": 0.5054003527957008, + "grad_norm": 0.5106372237205505, + "learning_rate": 9.833900779351725e-05, + "loss": 0.0491, + "step": 21560 + }, + { + "epoch": 0.5054237943705058, + "grad_norm": 3.257024049758911, + "learning_rate": 9.833164265639157e-05, + "loss": 0.3142, + "step": 21561 + }, + { + "epoch": 0.5054472359453108, + "grad_norm": 0.5885415077209473, + "learning_rate": 9.832427752831844e-05, + "loss": 0.0445, + "step": 21562 + }, + { + "epoch": 0.5054706775201158, + "grad_norm": 4.462916851043701, + "learning_rate": 9.831691240933782e-05, + "loss": 0.5269, + "step": 21563 + }, + { + "epoch": 0.5054941190949208, + "grad_norm": 3.143200397491455, + "learning_rate": 9.830954729948969e-05, + "loss": 2.0752, + "step": 21564 + }, + { + "epoch": 0.5055175606697258, + "grad_norm": 3.9725828170776367, + "learning_rate": 9.8302182198814e-05, + "loss": 0.2683, + "step": 21565 + }, + { + "epoch": 0.5055410022445308, + "grad_norm": 3.447988748550415, + "learning_rate": 9.829481710735073e-05, + "loss": 0.2573, + "step": 21566 + }, + { + "epoch": 0.5055644438193357, + "grad_norm": 2.350888252258301, + "learning_rate": 9.828745202513982e-05, + "loss": 0.1844, + "step": 21567 + }, + { + "epoch": 0.5055878853941408, + "grad_norm": 2.5637435913085938, + "learning_rate": 9.828008695222122e-05, + "loss": 0.4265, + "step": 21568 + }, + { + "epoch": 0.5056113269689457, + "grad_norm": 4.785125732421875, + "learning_rate": 9.827272188863494e-05, + "loss": 0.3769, + "step": 21569 + }, + { + "epoch": 0.5056347685437508, + "grad_norm": 4.044583320617676, + "learning_rate": 9.826535683442093e-05, + "loss": 0.2449, + "step": 21570 + }, + { + "epoch": 0.5056582101185557, + "grad_norm": 1.7215514183044434, + "learning_rate": 9.825799178961912e-05, + "loss": 0.1681, + "step": 21571 + }, + { + "epoch": 0.5056816516933608, + "grad_norm": 3.658421039581299, + "learning_rate": 9.825062675426947e-05, + "loss": 0.2079, + "step": 21572 + }, + { + "epoch": 0.5057050932681657, + "grad_norm": 0.906937301158905, + "learning_rate": 9.824326172841202e-05, + "loss": 0.0643, + "step": 21573 + }, + { + "epoch": 0.5057285348429708, + "grad_norm": 3.776449203491211, + "learning_rate": 9.823589671208666e-05, + "loss": 0.4427, + "step": 21574 + }, + { + "epoch": 0.5057519764177757, + "grad_norm": 3.1649787425994873, + "learning_rate": 9.822853170533338e-05, + "loss": 0.2871, + "step": 21575 + }, + { + "epoch": 0.5057754179925807, + "grad_norm": 4.661312580108643, + "learning_rate": 9.822116670819214e-05, + "loss": 0.4456, + "step": 21576 + }, + { + "epoch": 0.5057988595673857, + "grad_norm": 4.711620330810547, + "learning_rate": 9.82138017207029e-05, + "loss": 0.6374, + "step": 21577 + }, + { + "epoch": 0.5058223011421907, + "grad_norm": 2.931105136871338, + "learning_rate": 9.820643674290561e-05, + "loss": 0.208, + "step": 21578 + }, + { + "epoch": 0.5058457427169958, + "grad_norm": 1.3969858884811401, + "learning_rate": 9.819907177484026e-05, + "loss": 0.1473, + "step": 21579 + }, + { + "epoch": 0.5058691842918007, + "grad_norm": 3.8630948066711426, + "learning_rate": 9.819170681654679e-05, + "loss": 0.2654, + "step": 21580 + }, + { + "epoch": 0.5058926258666058, + "grad_norm": 2.247079372406006, + "learning_rate": 9.818434186806515e-05, + "loss": 0.1095, + "step": 21581 + }, + { + "epoch": 0.5059160674414107, + "grad_norm": 0.5389949083328247, + "learning_rate": 9.817697692943535e-05, + "loss": 0.0427, + "step": 21582 + }, + { + "epoch": 0.5059395090162158, + "grad_norm": 2.959656000137329, + "learning_rate": 9.816961200069733e-05, + "loss": 0.2237, + "step": 21583 + }, + { + "epoch": 0.5059629505910207, + "grad_norm": 1.6059470176696777, + "learning_rate": 9.8162247081891e-05, + "loss": 0.155, + "step": 21584 + }, + { + "epoch": 0.5059863921658257, + "grad_norm": 0.9717779755592346, + "learning_rate": 9.815488217305639e-05, + "loss": 0.1326, + "step": 21585 + }, + { + "epoch": 0.5060098337406307, + "grad_norm": 3.433669328689575, + "learning_rate": 9.814751727423346e-05, + "loss": 0.3097, + "step": 21586 + }, + { + "epoch": 0.5060332753154357, + "grad_norm": 2.5994601249694824, + "learning_rate": 9.814015238546215e-05, + "loss": 0.1995, + "step": 21587 + }, + { + "epoch": 0.5060567168902407, + "grad_norm": 3.1463968753814697, + "learning_rate": 9.81327875067824e-05, + "loss": 0.3464, + "step": 21588 + }, + { + "epoch": 0.5060801584650457, + "grad_norm": 3.1183226108551025, + "learning_rate": 9.812542263823424e-05, + "loss": 0.3083, + "step": 21589 + }, + { + "epoch": 0.5061036000398507, + "grad_norm": 2.3390157222747803, + "learning_rate": 9.811805777985757e-05, + "loss": 0.0988, + "step": 21590 + }, + { + "epoch": 0.5061270416146557, + "grad_norm": 0.611786425113678, + "learning_rate": 9.811069293169237e-05, + "loss": 0.104, + "step": 21591 + }, + { + "epoch": 0.5061504831894607, + "grad_norm": 4.441116809844971, + "learning_rate": 9.81033280937786e-05, + "loss": 0.3684, + "step": 21592 + }, + { + "epoch": 0.5061739247642657, + "grad_norm": 3.7568068504333496, + "learning_rate": 9.809596326615624e-05, + "loss": 0.5203, + "step": 21593 + }, + { + "epoch": 0.5061973663390706, + "grad_norm": 2.8757011890411377, + "learning_rate": 9.808859844886521e-05, + "loss": 0.2173, + "step": 21594 + }, + { + "epoch": 0.5062208079138757, + "grad_norm": 1.8255314826965332, + "learning_rate": 9.808123364194552e-05, + "loss": 0.1937, + "step": 21595 + }, + { + "epoch": 0.5062442494886806, + "grad_norm": 2.4218649864196777, + "learning_rate": 9.807386884543706e-05, + "loss": 0.2443, + "step": 21596 + }, + { + "epoch": 0.5062676910634857, + "grad_norm": 0.9582502841949463, + "learning_rate": 9.806650405937991e-05, + "loss": 0.0873, + "step": 21597 + }, + { + "epoch": 0.5062911326382906, + "grad_norm": 1.9354948997497559, + "learning_rate": 9.805913928381391e-05, + "loss": 1.1821, + "step": 21598 + }, + { + "epoch": 0.5063145742130957, + "grad_norm": 3.970961809158325, + "learning_rate": 9.805177451877912e-05, + "loss": 0.1674, + "step": 21599 + }, + { + "epoch": 0.5063380157879006, + "grad_norm": 3.1485178470611572, + "learning_rate": 9.804440976431545e-05, + "loss": 0.2264, + "step": 21600 + }, + { + "epoch": 0.5063614573627057, + "grad_norm": 4.7148942947387695, + "learning_rate": 9.803704502046285e-05, + "loss": 0.506, + "step": 21601 + }, + { + "epoch": 0.5063848989375106, + "grad_norm": 2.627244710922241, + "learning_rate": 9.80296802872613e-05, + "loss": 1.8595, + "step": 21602 + }, + { + "epoch": 0.5064083405123156, + "grad_norm": 5.1476149559021, + "learning_rate": 9.802231556475077e-05, + "loss": 0.6418, + "step": 21603 + }, + { + "epoch": 0.5064317820871206, + "grad_norm": 3.4798505306243896, + "learning_rate": 9.801495085297122e-05, + "loss": 0.3346, + "step": 21604 + }, + { + "epoch": 0.5064552236619256, + "grad_norm": 4.302474021911621, + "learning_rate": 9.800758615196259e-05, + "loss": 0.3977, + "step": 21605 + }, + { + "epoch": 0.5064786652367306, + "grad_norm": 3.943392753601074, + "learning_rate": 9.800022146176484e-05, + "loss": 0.4627, + "step": 21606 + }, + { + "epoch": 0.5065021068115356, + "grad_norm": 2.6855194568634033, + "learning_rate": 9.799285678241797e-05, + "loss": 0.2761, + "step": 21607 + }, + { + "epoch": 0.5065255483863406, + "grad_norm": 0.6947006583213806, + "learning_rate": 9.79854921139619e-05, + "loss": 0.0396, + "step": 21608 + }, + { + "epoch": 0.5065489899611456, + "grad_norm": 2.4718899726867676, + "learning_rate": 9.797812745643659e-05, + "loss": 0.2106, + "step": 21609 + }, + { + "epoch": 0.5065724315359506, + "grad_norm": 0.962317168712616, + "learning_rate": 9.797076280988202e-05, + "loss": 0.0617, + "step": 21610 + }, + { + "epoch": 0.5065958731107556, + "grad_norm": 0.9995372295379639, + "learning_rate": 9.796339817433817e-05, + "loss": 0.0973, + "step": 21611 + }, + { + "epoch": 0.5066193146855605, + "grad_norm": 3.6766042709350586, + "learning_rate": 9.795603354984499e-05, + "loss": 0.4156, + "step": 21612 + }, + { + "epoch": 0.5066427562603656, + "grad_norm": 3.07499623298645, + "learning_rate": 9.79486689364424e-05, + "loss": 0.2117, + "step": 21613 + }, + { + "epoch": 0.5066661978351705, + "grad_norm": 3.5881049633026123, + "learning_rate": 9.79413043341704e-05, + "loss": 0.4323, + "step": 21614 + }, + { + "epoch": 0.5066896394099756, + "grad_norm": 1.381295919418335, + "learning_rate": 9.793393974306896e-05, + "loss": 0.1231, + "step": 21615 + }, + { + "epoch": 0.5067130809847805, + "grad_norm": 2.3852202892303467, + "learning_rate": 9.792657516317799e-05, + "loss": 1.2946, + "step": 21616 + }, + { + "epoch": 0.5067365225595856, + "grad_norm": 2.4273295402526855, + "learning_rate": 9.791921059453752e-05, + "loss": 1.7655, + "step": 21617 + }, + { + "epoch": 0.5067599641343905, + "grad_norm": 4.548512935638428, + "learning_rate": 9.791184603718744e-05, + "loss": 0.4602, + "step": 21618 + }, + { + "epoch": 0.5067834057091956, + "grad_norm": 2.593266010284424, + "learning_rate": 9.790448149116774e-05, + "loss": 0.2191, + "step": 21619 + }, + { + "epoch": 0.5068068472840005, + "grad_norm": 0.7846821546554565, + "learning_rate": 9.78971169565184e-05, + "loss": 0.0412, + "step": 21620 + }, + { + "epoch": 0.5068302888588055, + "grad_norm": 3.249659538269043, + "learning_rate": 9.788975243327932e-05, + "loss": 0.2679, + "step": 21621 + }, + { + "epoch": 0.5068537304336105, + "grad_norm": 2.0306055545806885, + "learning_rate": 9.788238792149055e-05, + "loss": 0.1826, + "step": 21622 + }, + { + "epoch": 0.5068771720084155, + "grad_norm": 2.986112356185913, + "learning_rate": 9.787502342119198e-05, + "loss": 0.2044, + "step": 21623 + }, + { + "epoch": 0.5069006135832205, + "grad_norm": 3.3837778568267822, + "learning_rate": 9.786765893242362e-05, + "loss": 0.2327, + "step": 21624 + }, + { + "epoch": 0.5069240551580255, + "grad_norm": 2.4511630535125732, + "learning_rate": 9.786029445522538e-05, + "loss": 0.1981, + "step": 21625 + }, + { + "epoch": 0.5069474967328305, + "grad_norm": 2.811431646347046, + "learning_rate": 9.785292998963724e-05, + "loss": 0.3475, + "step": 21626 + }, + { + "epoch": 0.5069709383076355, + "grad_norm": 2.071843385696411, + "learning_rate": 9.784556553569919e-05, + "loss": 0.1885, + "step": 21627 + }, + { + "epoch": 0.5069943798824404, + "grad_norm": 2.804288387298584, + "learning_rate": 9.783820109345114e-05, + "loss": 0.1309, + "step": 21628 + }, + { + "epoch": 0.5070178214572455, + "grad_norm": 1.4732303619384766, + "learning_rate": 9.783083666293307e-05, + "loss": 0.8243, + "step": 21629 + }, + { + "epoch": 0.5070412630320504, + "grad_norm": 2.8447611331939697, + "learning_rate": 9.782347224418496e-05, + "loss": 0.2861, + "step": 21630 + }, + { + "epoch": 0.5070647046068555, + "grad_norm": 1.9575949907302856, + "learning_rate": 9.781610783724674e-05, + "loss": 0.1114, + "step": 21631 + }, + { + "epoch": 0.5070881461816605, + "grad_norm": 2.753357172012329, + "learning_rate": 9.780874344215836e-05, + "loss": 0.2069, + "step": 21632 + }, + { + "epoch": 0.5071115877564655, + "grad_norm": 3.1022300720214844, + "learning_rate": 9.780137905895978e-05, + "loss": 0.2955, + "step": 21633 + }, + { + "epoch": 0.5071350293312705, + "grad_norm": 3.3400232791900635, + "learning_rate": 9.779401468769102e-05, + "loss": 0.1037, + "step": 21634 + }, + { + "epoch": 0.5071584709060755, + "grad_norm": 2.1901283264160156, + "learning_rate": 9.778665032839201e-05, + "loss": 1.8477, + "step": 21635 + }, + { + "epoch": 0.5071819124808805, + "grad_norm": 2.4178285598754883, + "learning_rate": 9.777928598110266e-05, + "loss": 0.1671, + "step": 21636 + }, + { + "epoch": 0.5072053540556855, + "grad_norm": 1.4743090867996216, + "learning_rate": 9.7771921645863e-05, + "loss": 0.1924, + "step": 21637 + }, + { + "epoch": 0.5072287956304905, + "grad_norm": 4.906419277191162, + "learning_rate": 9.776455732271296e-05, + "loss": 0.4986, + "step": 21638 + }, + { + "epoch": 0.5072522372052954, + "grad_norm": 2.9912209510803223, + "learning_rate": 9.775719301169246e-05, + "loss": 0.2736, + "step": 21639 + }, + { + "epoch": 0.5072756787801005, + "grad_norm": 2.7855639457702637, + "learning_rate": 9.774982871284153e-05, + "loss": 0.2615, + "step": 21640 + }, + { + "epoch": 0.5072991203549054, + "grad_norm": 2.480278730392456, + "learning_rate": 9.774246442620005e-05, + "loss": 0.2499, + "step": 21641 + }, + { + "epoch": 0.5073225619297105, + "grad_norm": 3.1136364936828613, + "learning_rate": 9.773510015180807e-05, + "loss": 0.2084, + "step": 21642 + }, + { + "epoch": 0.5073460035045154, + "grad_norm": 1.7913583517074585, + "learning_rate": 9.772773588970548e-05, + "loss": 0.1556, + "step": 21643 + }, + { + "epoch": 0.5073694450793205, + "grad_norm": 1.947792649269104, + "learning_rate": 9.772037163993224e-05, + "loss": 0.1695, + "step": 21644 + }, + { + "epoch": 0.5073928866541254, + "grad_norm": 1.7846441268920898, + "learning_rate": 9.771300740252833e-05, + "loss": 0.1194, + "step": 21645 + }, + { + "epoch": 0.5074163282289305, + "grad_norm": 4.471668720245361, + "learning_rate": 9.770564317753371e-05, + "loss": 0.4127, + "step": 21646 + }, + { + "epoch": 0.5074397698037354, + "grad_norm": 3.5633108615875244, + "learning_rate": 9.769827896498836e-05, + "loss": 0.3637, + "step": 21647 + }, + { + "epoch": 0.5074632113785404, + "grad_norm": 1.778025507926941, + "learning_rate": 9.769091476493217e-05, + "loss": 0.0882, + "step": 21648 + }, + { + "epoch": 0.5074866529533454, + "grad_norm": 11.46816349029541, + "learning_rate": 9.768355057740518e-05, + "loss": 0.7428, + "step": 21649 + }, + { + "epoch": 0.5075100945281504, + "grad_norm": 2.8987483978271484, + "learning_rate": 9.767618640244732e-05, + "loss": 0.3074, + "step": 21650 + }, + { + "epoch": 0.5075335361029554, + "grad_norm": 3.686415672302246, + "learning_rate": 9.766882224009849e-05, + "loss": 0.4351, + "step": 21651 + }, + { + "epoch": 0.5075569776777604, + "grad_norm": 3.584355115890503, + "learning_rate": 9.766145809039874e-05, + "loss": 0.2907, + "step": 21652 + }, + { + "epoch": 0.5075804192525654, + "grad_norm": 4.211704730987549, + "learning_rate": 9.765409395338797e-05, + "loss": 0.4307, + "step": 21653 + }, + { + "epoch": 0.5076038608273704, + "grad_norm": 3.306321144104004, + "learning_rate": 9.764672982910613e-05, + "loss": 0.3494, + "step": 21654 + }, + { + "epoch": 0.5076273024021754, + "grad_norm": 2.8141729831695557, + "learning_rate": 9.763936571759323e-05, + "loss": 0.255, + "step": 21655 + }, + { + "epoch": 0.5076507439769804, + "grad_norm": 3.7013015747070312, + "learning_rate": 9.76320016188892e-05, + "loss": 0.5078, + "step": 21656 + }, + { + "epoch": 0.5076741855517853, + "grad_norm": 1.926303505897522, + "learning_rate": 9.762463753303395e-05, + "loss": 0.0892, + "step": 21657 + }, + { + "epoch": 0.5076976271265904, + "grad_norm": 2.1794989109039307, + "learning_rate": 9.761727346006751e-05, + "loss": 0.211, + "step": 21658 + }, + { + "epoch": 0.5077210687013953, + "grad_norm": 3.735346555709839, + "learning_rate": 9.760990940002982e-05, + "loss": 0.2787, + "step": 21659 + }, + { + "epoch": 0.5077445102762004, + "grad_norm": 5.065316677093506, + "learning_rate": 9.760254535296083e-05, + "loss": 0.3869, + "step": 21660 + }, + { + "epoch": 0.5077679518510053, + "grad_norm": 2.275861978530884, + "learning_rate": 9.759518131890049e-05, + "loss": 0.2241, + "step": 21661 + }, + { + "epoch": 0.5077913934258104, + "grad_norm": 3.192188262939453, + "learning_rate": 9.758781729788876e-05, + "loss": 0.4381, + "step": 21662 + }, + { + "epoch": 0.5078148350006153, + "grad_norm": 0.9437404274940491, + "learning_rate": 9.758045328996562e-05, + "loss": 0.0651, + "step": 21663 + }, + { + "epoch": 0.5078382765754204, + "grad_norm": 1.379431962966919, + "learning_rate": 9.757308929517098e-05, + "loss": 0.1601, + "step": 21664 + }, + { + "epoch": 0.5078617181502253, + "grad_norm": 4.850366115570068, + "learning_rate": 9.756572531354485e-05, + "loss": 0.457, + "step": 21665 + }, + { + "epoch": 0.5078851597250303, + "grad_norm": 2.3814189434051514, + "learning_rate": 9.755836134512716e-05, + "loss": 0.1032, + "step": 21666 + }, + { + "epoch": 0.5079086012998353, + "grad_norm": 3.6535942554473877, + "learning_rate": 9.755099738995787e-05, + "loss": 0.3308, + "step": 21667 + }, + { + "epoch": 0.5079320428746403, + "grad_norm": 0.9292622804641724, + "learning_rate": 9.754363344807694e-05, + "loss": 0.0583, + "step": 21668 + }, + { + "epoch": 0.5079554844494453, + "grad_norm": 2.169360876083374, + "learning_rate": 9.753626951952428e-05, + "loss": 0.3039, + "step": 21669 + }, + { + "epoch": 0.5079789260242503, + "grad_norm": 4.342981338500977, + "learning_rate": 9.752890560433992e-05, + "loss": 0.2435, + "step": 21670 + }, + { + "epoch": 0.5080023675990553, + "grad_norm": 3.732642889022827, + "learning_rate": 9.752154170256379e-05, + "loss": 0.4845, + "step": 21671 + }, + { + "epoch": 0.5080258091738603, + "grad_norm": 4.369710922241211, + "learning_rate": 9.751417781423586e-05, + "loss": 0.6771, + "step": 21672 + }, + { + "epoch": 0.5080492507486652, + "grad_norm": 2.197268009185791, + "learning_rate": 9.750681393939606e-05, + "loss": 0.2122, + "step": 21673 + }, + { + "epoch": 0.5080726923234703, + "grad_norm": 1.1018109321594238, + "learning_rate": 9.749945007808435e-05, + "loss": 0.0788, + "step": 21674 + }, + { + "epoch": 0.5080961338982752, + "grad_norm": 3.6172280311584473, + "learning_rate": 9.74920862303407e-05, + "loss": 0.414, + "step": 21675 + }, + { + "epoch": 0.5081195754730803, + "grad_norm": 5.903236389160156, + "learning_rate": 9.748472239620507e-05, + "loss": 0.6649, + "step": 21676 + }, + { + "epoch": 0.5081430170478852, + "grad_norm": 2.8353397846221924, + "learning_rate": 9.747735857571739e-05, + "loss": 0.325, + "step": 21677 + }, + { + "epoch": 0.5081664586226903, + "grad_norm": 3.5994882583618164, + "learning_rate": 9.746999476891766e-05, + "loss": 0.3436, + "step": 21678 + }, + { + "epoch": 0.5081899001974952, + "grad_norm": 1.1218653917312622, + "learning_rate": 9.746263097584579e-05, + "loss": 0.15, + "step": 21679 + }, + { + "epoch": 0.5082133417723003, + "grad_norm": 3.138659715652466, + "learning_rate": 9.745526719654176e-05, + "loss": 0.3698, + "step": 21680 + }, + { + "epoch": 0.5082367833471052, + "grad_norm": 3.58293080329895, + "learning_rate": 9.744790343104547e-05, + "loss": 0.4304, + "step": 21681 + }, + { + "epoch": 0.5082602249219103, + "grad_norm": 2.9029223918914795, + "learning_rate": 9.744053967939699e-05, + "loss": 1.5562, + "step": 21682 + }, + { + "epoch": 0.5082836664967153, + "grad_norm": 1.0115137100219727, + "learning_rate": 9.74331759416362e-05, + "loss": 0.0613, + "step": 21683 + }, + { + "epoch": 0.5083071080715202, + "grad_norm": 4.002623558044434, + "learning_rate": 9.742581221780307e-05, + "loss": 0.483, + "step": 21684 + }, + { + "epoch": 0.5083305496463253, + "grad_norm": 2.032339096069336, + "learning_rate": 9.741844850793757e-05, + "loss": 0.7953, + "step": 21685 + }, + { + "epoch": 0.5083539912211302, + "grad_norm": 2.0949034690856934, + "learning_rate": 9.741108481207961e-05, + "loss": 0.1745, + "step": 21686 + }, + { + "epoch": 0.5083774327959353, + "grad_norm": 4.3999223709106445, + "learning_rate": 9.74037211302692e-05, + "loss": 0.4436, + "step": 21687 + }, + { + "epoch": 0.5084008743707402, + "grad_norm": 3.389904260635376, + "learning_rate": 9.739635746254627e-05, + "loss": 0.2607, + "step": 21688 + }, + { + "epoch": 0.5084243159455453, + "grad_norm": 2.508899211883545, + "learning_rate": 9.738899380895077e-05, + "loss": 0.1413, + "step": 21689 + }, + { + "epoch": 0.5084477575203502, + "grad_norm": 1.5803536176681519, + "learning_rate": 9.738163016952267e-05, + "loss": 0.15, + "step": 21690 + }, + { + "epoch": 0.5084711990951553, + "grad_norm": 1.9353049993515015, + "learning_rate": 9.737426654430192e-05, + "loss": 0.1256, + "step": 21691 + }, + { + "epoch": 0.5084946406699602, + "grad_norm": 2.320103883743286, + "learning_rate": 9.736690293332845e-05, + "loss": 0.3467, + "step": 21692 + }, + { + "epoch": 0.5085180822447652, + "grad_norm": 4.50710391998291, + "learning_rate": 9.735953933664227e-05, + "loss": 0.4067, + "step": 21693 + }, + { + "epoch": 0.5085415238195702, + "grad_norm": 1.0003533363342285, + "learning_rate": 9.735217575428326e-05, + "loss": 0.0743, + "step": 21694 + }, + { + "epoch": 0.5085649653943752, + "grad_norm": 2.732811212539673, + "learning_rate": 9.734481218629146e-05, + "loss": 1.5788, + "step": 21695 + }, + { + "epoch": 0.5085884069691802, + "grad_norm": 0.8360646367073059, + "learning_rate": 9.733744863270676e-05, + "loss": 0.0899, + "step": 21696 + }, + { + "epoch": 0.5086118485439852, + "grad_norm": 2.761246681213379, + "learning_rate": 9.733008509356916e-05, + "loss": 0.1964, + "step": 21697 + }, + { + "epoch": 0.5086352901187902, + "grad_norm": 0.999330461025238, + "learning_rate": 9.732272156891859e-05, + "loss": 0.0988, + "step": 21698 + }, + { + "epoch": 0.5086587316935952, + "grad_norm": 3.596309185028076, + "learning_rate": 9.731535805879499e-05, + "loss": 0.3462, + "step": 21699 + }, + { + "epoch": 0.5086821732684002, + "grad_norm": 4.454980373382568, + "learning_rate": 9.730799456323835e-05, + "loss": 0.4353, + "step": 21700 + }, + { + "epoch": 0.5087056148432052, + "grad_norm": 3.7916150093078613, + "learning_rate": 9.730063108228862e-05, + "loss": 0.3754, + "step": 21701 + }, + { + "epoch": 0.5087290564180101, + "grad_norm": 1.5636627674102783, + "learning_rate": 9.72932676159857e-05, + "loss": 0.0737, + "step": 21702 + }, + { + "epoch": 0.5087524979928152, + "grad_norm": 2.888469934463501, + "learning_rate": 9.728590416436962e-05, + "loss": 0.353, + "step": 21703 + }, + { + "epoch": 0.5087759395676201, + "grad_norm": 0.7809990048408508, + "learning_rate": 9.72785407274803e-05, + "loss": 0.0652, + "step": 21704 + }, + { + "epoch": 0.5087993811424252, + "grad_norm": 1.1294538974761963, + "learning_rate": 9.727117730535768e-05, + "loss": 0.0715, + "step": 21705 + }, + { + "epoch": 0.5088228227172301, + "grad_norm": 2.405125141143799, + "learning_rate": 9.726381389804169e-05, + "loss": 0.273, + "step": 21706 + }, + { + "epoch": 0.5088462642920352, + "grad_norm": 0.8363164067268372, + "learning_rate": 9.725645050557239e-05, + "loss": 0.1015, + "step": 21707 + }, + { + "epoch": 0.5088697058668401, + "grad_norm": 4.116907596588135, + "learning_rate": 9.724908712798964e-05, + "loss": 0.2117, + "step": 21708 + }, + { + "epoch": 0.5088931474416452, + "grad_norm": 2.3190603256225586, + "learning_rate": 9.724172376533342e-05, + "loss": 0.1539, + "step": 21709 + }, + { + "epoch": 0.5089165890164501, + "grad_norm": 2.6566128730773926, + "learning_rate": 9.723436041764369e-05, + "loss": 0.2929, + "step": 21710 + }, + { + "epoch": 0.5089400305912551, + "grad_norm": 4.156472206115723, + "learning_rate": 9.722699708496041e-05, + "loss": 0.2544, + "step": 21711 + }, + { + "epoch": 0.5089634721660601, + "grad_norm": 0.6422979235649109, + "learning_rate": 9.72196337673235e-05, + "loss": 0.0453, + "step": 21712 + }, + { + "epoch": 0.5089869137408651, + "grad_norm": 3.6551826000213623, + "learning_rate": 9.721227046477294e-05, + "loss": 0.2758, + "step": 21713 + }, + { + "epoch": 0.5090103553156701, + "grad_norm": 2.4574081897735596, + "learning_rate": 9.720490717734869e-05, + "loss": 0.2772, + "step": 21714 + }, + { + "epoch": 0.5090337968904751, + "grad_norm": 1.1708292961120605, + "learning_rate": 9.719754390509067e-05, + "loss": 0.0932, + "step": 21715 + }, + { + "epoch": 0.5090572384652801, + "grad_norm": 2.5997583866119385, + "learning_rate": 9.719018064803888e-05, + "loss": 0.21, + "step": 21716 + }, + { + "epoch": 0.5090806800400851, + "grad_norm": 0.2898847162723541, + "learning_rate": 9.718281740623323e-05, + "loss": 0.0173, + "step": 21717 + }, + { + "epoch": 0.50910412161489, + "grad_norm": 2.125770092010498, + "learning_rate": 9.717545417971369e-05, + "loss": 0.2375, + "step": 21718 + }, + { + "epoch": 0.5091275631896951, + "grad_norm": 2.758563756942749, + "learning_rate": 9.716809096852021e-05, + "loss": 0.2411, + "step": 21719 + }, + { + "epoch": 0.5091510047645, + "grad_norm": 4.676153659820557, + "learning_rate": 9.716072777269278e-05, + "loss": 0.5524, + "step": 21720 + }, + { + "epoch": 0.5091744463393051, + "grad_norm": 2.8689496517181396, + "learning_rate": 9.715336459227129e-05, + "loss": 0.225, + "step": 21721 + }, + { + "epoch": 0.50919788791411, + "grad_norm": 1.613960862159729, + "learning_rate": 9.714600142729574e-05, + "loss": 0.1703, + "step": 21722 + }, + { + "epoch": 0.5092213294889151, + "grad_norm": 2.7341973781585693, + "learning_rate": 9.713863827780609e-05, + "loss": 0.2559, + "step": 21723 + }, + { + "epoch": 0.50924477106372, + "grad_norm": 3.9756674766540527, + "learning_rate": 9.713127514384221e-05, + "loss": 0.4191, + "step": 21724 + }, + { + "epoch": 0.5092682126385251, + "grad_norm": 0.39380162954330444, + "learning_rate": 9.712391202544416e-05, + "loss": 0.0213, + "step": 21725 + }, + { + "epoch": 0.50929165421333, + "grad_norm": 5.053986549377441, + "learning_rate": 9.711654892265184e-05, + "loss": 0.2462, + "step": 21726 + }, + { + "epoch": 0.509315095788135, + "grad_norm": 2.209012985229492, + "learning_rate": 9.71091858355052e-05, + "loss": 0.2217, + "step": 21727 + }, + { + "epoch": 0.50933853736294, + "grad_norm": 3.8869502544403076, + "learning_rate": 9.71018227640442e-05, + "loss": 0.4028, + "step": 21728 + }, + { + "epoch": 0.509361978937745, + "grad_norm": 3.1975436210632324, + "learning_rate": 9.709445970830878e-05, + "loss": 0.3618, + "step": 21729 + }, + { + "epoch": 0.50938542051255, + "grad_norm": 1.7739031314849854, + "learning_rate": 9.708709666833889e-05, + "loss": 0.2814, + "step": 21730 + }, + { + "epoch": 0.509408862087355, + "grad_norm": 3.1877129077911377, + "learning_rate": 9.70797336441745e-05, + "loss": 0.3458, + "step": 21731 + }, + { + "epoch": 0.50943230366216, + "grad_norm": 3.442694902420044, + "learning_rate": 9.707237063585558e-05, + "loss": 0.4967, + "step": 21732 + }, + { + "epoch": 0.509455745236965, + "grad_norm": 3.283433437347412, + "learning_rate": 9.706500764342206e-05, + "loss": 0.2156, + "step": 21733 + }, + { + "epoch": 0.5094791868117701, + "grad_norm": 2.5851657390594482, + "learning_rate": 9.705764466691388e-05, + "loss": 0.3422, + "step": 21734 + }, + { + "epoch": 0.509502628386575, + "grad_norm": 4.451934337615967, + "learning_rate": 9.7050281706371e-05, + "loss": 0.4354, + "step": 21735 + }, + { + "epoch": 0.5095260699613801, + "grad_norm": 1.4296568632125854, + "learning_rate": 9.704291876183339e-05, + "loss": 0.1192, + "step": 21736 + }, + { + "epoch": 0.509549511536185, + "grad_norm": 3.437957525253296, + "learning_rate": 9.703555583334096e-05, + "loss": 0.3396, + "step": 21737 + }, + { + "epoch": 0.50957295311099, + "grad_norm": 3.471428871154785, + "learning_rate": 9.702819292093372e-05, + "loss": 0.4868, + "step": 21738 + }, + { + "epoch": 0.509596394685795, + "grad_norm": 4.159058570861816, + "learning_rate": 9.702083002465158e-05, + "loss": 0.3506, + "step": 21739 + }, + { + "epoch": 0.5096198362606, + "grad_norm": 2.1816678047180176, + "learning_rate": 9.701346714453448e-05, + "loss": 0.2496, + "step": 21740 + }, + { + "epoch": 0.509643277835405, + "grad_norm": 3.8041446208953857, + "learning_rate": 9.700610428062241e-05, + "loss": 0.2784, + "step": 21741 + }, + { + "epoch": 0.50966671941021, + "grad_norm": 2.592581033706665, + "learning_rate": 9.699874143295527e-05, + "loss": 0.265, + "step": 21742 + }, + { + "epoch": 0.509690160985015, + "grad_norm": 1.543563961982727, + "learning_rate": 9.699137860157307e-05, + "loss": 1.2373, + "step": 21743 + }, + { + "epoch": 0.50971360255982, + "grad_norm": 3.4309024810791016, + "learning_rate": 9.698401578651572e-05, + "loss": 0.3577, + "step": 21744 + }, + { + "epoch": 0.509737044134625, + "grad_norm": 4.828775405883789, + "learning_rate": 9.697665298782321e-05, + "loss": 0.5865, + "step": 21745 + }, + { + "epoch": 0.50976048570943, + "grad_norm": 0.9834780097007751, + "learning_rate": 9.696929020553546e-05, + "loss": 0.0978, + "step": 21746 + }, + { + "epoch": 0.5097839272842349, + "grad_norm": 4.413672924041748, + "learning_rate": 9.696192743969241e-05, + "loss": 0.5023, + "step": 21747 + }, + { + "epoch": 0.50980736885904, + "grad_norm": 3.805650472640991, + "learning_rate": 9.695456469033405e-05, + "loss": 0.3253, + "step": 21748 + }, + { + "epoch": 0.5098308104338449, + "grad_norm": 1.0715237855911255, + "learning_rate": 9.694720195750032e-05, + "loss": 0.091, + "step": 21749 + }, + { + "epoch": 0.50985425200865, + "grad_norm": 2.396063804626465, + "learning_rate": 9.693983924123112e-05, + "loss": 0.212, + "step": 21750 + }, + { + "epoch": 0.5098776935834549, + "grad_norm": 2.5462474822998047, + "learning_rate": 9.693247654156645e-05, + "loss": 0.1517, + "step": 21751 + }, + { + "epoch": 0.50990113515826, + "grad_norm": 4.842738628387451, + "learning_rate": 9.692511385854626e-05, + "loss": 0.2199, + "step": 21752 + }, + { + "epoch": 0.5099245767330649, + "grad_norm": 3.6637892723083496, + "learning_rate": 9.691775119221047e-05, + "loss": 0.2468, + "step": 21753 + }, + { + "epoch": 0.50994801830787, + "grad_norm": 3.444061040878296, + "learning_rate": 9.691038854259902e-05, + "loss": 0.2247, + "step": 21754 + }, + { + "epoch": 0.5099714598826749, + "grad_norm": 2.633451461791992, + "learning_rate": 9.690302590975193e-05, + "loss": 0.2162, + "step": 21755 + }, + { + "epoch": 0.5099949014574799, + "grad_norm": 4.0815749168396, + "learning_rate": 9.689566329370913e-05, + "loss": 0.5324, + "step": 21756 + }, + { + "epoch": 0.5100183430322849, + "grad_norm": 2.661851644515991, + "learning_rate": 9.688830069451051e-05, + "loss": 1.8575, + "step": 21757 + }, + { + "epoch": 0.5100417846070899, + "grad_norm": 2.4284400939941406, + "learning_rate": 9.688093811219607e-05, + "loss": 0.0895, + "step": 21758 + }, + { + "epoch": 0.5100652261818949, + "grad_norm": 2.997271776199341, + "learning_rate": 9.687357554680574e-05, + "loss": 0.221, + "step": 21759 + }, + { + "epoch": 0.5100886677566999, + "grad_norm": 3.8385767936706543, + "learning_rate": 9.686621299837949e-05, + "loss": 0.2562, + "step": 21760 + }, + { + "epoch": 0.5101121093315049, + "grad_norm": 2.748995065689087, + "learning_rate": 9.685885046695726e-05, + "loss": 0.3659, + "step": 21761 + }, + { + "epoch": 0.5101355509063099, + "grad_norm": 2.706002712249756, + "learning_rate": 9.685148795257898e-05, + "loss": 0.1539, + "step": 21762 + }, + { + "epoch": 0.5101589924811148, + "grad_norm": 0.529954195022583, + "learning_rate": 9.684412545528464e-05, + "loss": 0.0517, + "step": 21763 + }, + { + "epoch": 0.5101824340559199, + "grad_norm": 2.4331226348876953, + "learning_rate": 9.683676297511415e-05, + "loss": 0.1555, + "step": 21764 + }, + { + "epoch": 0.5102058756307248, + "grad_norm": 3.2377166748046875, + "learning_rate": 9.682940051210746e-05, + "loss": 0.1737, + "step": 21765 + }, + { + "epoch": 0.5102293172055299, + "grad_norm": 1.2253378629684448, + "learning_rate": 9.682203806630456e-05, + "loss": 0.1312, + "step": 21766 + }, + { + "epoch": 0.5102527587803348, + "grad_norm": 1.2353599071502686, + "learning_rate": 9.681467563774531e-05, + "loss": 0.1322, + "step": 21767 + }, + { + "epoch": 0.5102762003551399, + "grad_norm": 3.565366506576538, + "learning_rate": 9.680731322646977e-05, + "loss": 0.2172, + "step": 21768 + }, + { + "epoch": 0.5102996419299448, + "grad_norm": 2.454007863998413, + "learning_rate": 9.679995083251782e-05, + "loss": 0.2523, + "step": 21769 + }, + { + "epoch": 0.5103230835047499, + "grad_norm": 2.2658047676086426, + "learning_rate": 9.679258845592944e-05, + "loss": 0.0926, + "step": 21770 + }, + { + "epoch": 0.5103465250795548, + "grad_norm": 5.020618915557861, + "learning_rate": 9.678522609674457e-05, + "loss": 0.6599, + "step": 21771 + }, + { + "epoch": 0.5103699666543599, + "grad_norm": 2.192413806915283, + "learning_rate": 9.677786375500314e-05, + "loss": 0.27, + "step": 21772 + }, + { + "epoch": 0.5103934082291648, + "grad_norm": 4.171939849853516, + "learning_rate": 9.677050143074513e-05, + "loss": 0.3362, + "step": 21773 + }, + { + "epoch": 0.5104168498039698, + "grad_norm": 1.0331425666809082, + "learning_rate": 9.676313912401046e-05, + "loss": 0.0833, + "step": 21774 + }, + { + "epoch": 0.5104402913787748, + "grad_norm": 4.857456207275391, + "learning_rate": 9.675577683483907e-05, + "loss": 0.5058, + "step": 21775 + }, + { + "epoch": 0.5104637329535798, + "grad_norm": 3.355994701385498, + "learning_rate": 9.674841456327096e-05, + "loss": 0.2453, + "step": 21776 + }, + { + "epoch": 0.5104871745283848, + "grad_norm": 2.1559598445892334, + "learning_rate": 9.674105230934604e-05, + "loss": 0.1102, + "step": 21777 + }, + { + "epoch": 0.5105106161031898, + "grad_norm": 2.8411500453948975, + "learning_rate": 9.673369007310423e-05, + "loss": 0.4153, + "step": 21778 + }, + { + "epoch": 0.5105340576779948, + "grad_norm": 3.5884859561920166, + "learning_rate": 9.67263278545855e-05, + "loss": 0.3897, + "step": 21779 + }, + { + "epoch": 0.5105574992527998, + "grad_norm": 1.8918627500534058, + "learning_rate": 9.671896565382984e-05, + "loss": 0.1731, + "step": 21780 + }, + { + "epoch": 0.5105809408276047, + "grad_norm": 3.097795009613037, + "learning_rate": 9.671160347087717e-05, + "loss": 0.2854, + "step": 21781 + }, + { + "epoch": 0.5106043824024098, + "grad_norm": 3.946074962615967, + "learning_rate": 9.67042413057674e-05, + "loss": 0.3518, + "step": 21782 + }, + { + "epoch": 0.5106278239772147, + "grad_norm": 2.9414992332458496, + "learning_rate": 9.669687915854055e-05, + "loss": 0.1181, + "step": 21783 + }, + { + "epoch": 0.5106512655520198, + "grad_norm": 4.647970199584961, + "learning_rate": 9.66895170292365e-05, + "loss": 0.5541, + "step": 21784 + }, + { + "epoch": 0.5106747071268248, + "grad_norm": 1.2374271154403687, + "learning_rate": 9.668215491789522e-05, + "loss": 0.1374, + "step": 21785 + }, + { + "epoch": 0.5106981487016298, + "grad_norm": 3.2196013927459717, + "learning_rate": 9.667479282455669e-05, + "loss": 0.1778, + "step": 21786 + }, + { + "epoch": 0.5107215902764348, + "grad_norm": 8.279208183288574, + "learning_rate": 9.666743074926082e-05, + "loss": 0.6886, + "step": 21787 + }, + { + "epoch": 0.5107450318512398, + "grad_norm": 4.971981048583984, + "learning_rate": 9.666006869204755e-05, + "loss": 0.2744, + "step": 21788 + }, + { + "epoch": 0.5107684734260448, + "grad_norm": 1.5146640539169312, + "learning_rate": 9.665270665295685e-05, + "loss": 0.1889, + "step": 21789 + }, + { + "epoch": 0.5107919150008497, + "grad_norm": 2.814267635345459, + "learning_rate": 9.664534463202864e-05, + "loss": 0.1789, + "step": 21790 + }, + { + "epoch": 0.5108153565756548, + "grad_norm": 3.262409210205078, + "learning_rate": 9.663798262930288e-05, + "loss": 0.335, + "step": 21791 + }, + { + "epoch": 0.5108387981504597, + "grad_norm": 2.240870714187622, + "learning_rate": 9.663062064481953e-05, + "loss": 0.2659, + "step": 21792 + }, + { + "epoch": 0.5108622397252648, + "grad_norm": 2.031827449798584, + "learning_rate": 9.662325867861855e-05, + "loss": 0.2272, + "step": 21793 + }, + { + "epoch": 0.5108856813000697, + "grad_norm": 2.9627468585968018, + "learning_rate": 9.661589673073986e-05, + "loss": 0.2723, + "step": 21794 + }, + { + "epoch": 0.5109091228748748, + "grad_norm": 0.8751662373542786, + "learning_rate": 9.660853480122337e-05, + "loss": 0.102, + "step": 21795 + }, + { + "epoch": 0.5109325644496797, + "grad_norm": 3.0331742763519287, + "learning_rate": 9.660117289010911e-05, + "loss": 0.4442, + "step": 21796 + }, + { + "epoch": 0.5109560060244848, + "grad_norm": 3.48281192779541, + "learning_rate": 9.659381099743695e-05, + "loss": 0.429, + "step": 21797 + }, + { + "epoch": 0.5109794475992897, + "grad_norm": 0.6471133828163147, + "learning_rate": 9.65864491232469e-05, + "loss": 0.0972, + "step": 21798 + }, + { + "epoch": 0.5110028891740948, + "grad_norm": 1.9783289432525635, + "learning_rate": 9.657908726757885e-05, + "loss": 0.247, + "step": 21799 + }, + { + "epoch": 0.5110263307488997, + "grad_norm": 0.39300471544265747, + "learning_rate": 9.657172543047275e-05, + "loss": 0.0238, + "step": 21800 + }, + { + "epoch": 0.5110497723237047, + "grad_norm": 2.9784462451934814, + "learning_rate": 9.656436361196861e-05, + "loss": 0.1719, + "step": 21801 + }, + { + "epoch": 0.5110732138985097, + "grad_norm": 8.607665061950684, + "learning_rate": 9.655700181210632e-05, + "loss": 1.5333, + "step": 21802 + }, + { + "epoch": 0.5110966554733147, + "grad_norm": 5.307067394256592, + "learning_rate": 9.654964003092579e-05, + "loss": 0.4918, + "step": 21803 + }, + { + "epoch": 0.5111200970481197, + "grad_norm": 3.0341222286224365, + "learning_rate": 9.654227826846702e-05, + "loss": 0.3134, + "step": 21804 + }, + { + "epoch": 0.5111435386229247, + "grad_norm": 2.9256975650787354, + "learning_rate": 9.653491652476999e-05, + "loss": 0.2288, + "step": 21805 + }, + { + "epoch": 0.5111669801977297, + "grad_norm": 2.4991300106048584, + "learning_rate": 9.652755479987459e-05, + "loss": 0.1845, + "step": 21806 + }, + { + "epoch": 0.5111904217725347, + "grad_norm": 2.706023931503296, + "learning_rate": 9.652019309382075e-05, + "loss": 0.356, + "step": 21807 + }, + { + "epoch": 0.5112138633473396, + "grad_norm": 3.1712393760681152, + "learning_rate": 9.651283140664846e-05, + "loss": 0.2604, + "step": 21808 + }, + { + "epoch": 0.5112373049221447, + "grad_norm": 2.074035406112671, + "learning_rate": 9.650546973839765e-05, + "loss": 0.1769, + "step": 21809 + }, + { + "epoch": 0.5112607464969496, + "grad_norm": 3.6637375354766846, + "learning_rate": 9.649810808910823e-05, + "loss": 0.4978, + "step": 21810 + }, + { + "epoch": 0.5112841880717547, + "grad_norm": 1.4953340291976929, + "learning_rate": 9.64907464588202e-05, + "loss": 0.1129, + "step": 21811 + }, + { + "epoch": 0.5113076296465596, + "grad_norm": 3.061147689819336, + "learning_rate": 9.648338484757349e-05, + "loss": 0.2886, + "step": 21812 + }, + { + "epoch": 0.5113310712213647, + "grad_norm": 1.7069028615951538, + "learning_rate": 9.647602325540799e-05, + "loss": 0.1087, + "step": 21813 + }, + { + "epoch": 0.5113545127961696, + "grad_norm": 4.024786472320557, + "learning_rate": 9.646866168236373e-05, + "loss": 0.5111, + "step": 21814 + }, + { + "epoch": 0.5113779543709747, + "grad_norm": 3.1900417804718018, + "learning_rate": 9.646130012848055e-05, + "loss": 2.0071, + "step": 21815 + }, + { + "epoch": 0.5114013959457796, + "grad_norm": 0.6831266283988953, + "learning_rate": 9.645393859379851e-05, + "loss": 0.0488, + "step": 21816 + }, + { + "epoch": 0.5114248375205847, + "grad_norm": 1.4760503768920898, + "learning_rate": 9.644657707835748e-05, + "loss": 0.1766, + "step": 21817 + }, + { + "epoch": 0.5114482790953896, + "grad_norm": 1.7801425457000732, + "learning_rate": 9.643921558219743e-05, + "loss": 0.1272, + "step": 21818 + }, + { + "epoch": 0.5114717206701946, + "grad_norm": 0.8018986582756042, + "learning_rate": 9.64318541053583e-05, + "loss": 0.0501, + "step": 21819 + }, + { + "epoch": 0.5114951622449996, + "grad_norm": 3.602512836456299, + "learning_rate": 9.642449264788001e-05, + "loss": 0.3809, + "step": 21820 + }, + { + "epoch": 0.5115186038198046, + "grad_norm": 3.2881646156311035, + "learning_rate": 9.641713120980255e-05, + "loss": 0.3347, + "step": 21821 + }, + { + "epoch": 0.5115420453946096, + "grad_norm": 3.0724685192108154, + "learning_rate": 9.640976979116583e-05, + "loss": 0.4938, + "step": 21822 + }, + { + "epoch": 0.5115654869694146, + "grad_norm": 3.092546224594116, + "learning_rate": 9.640240839200977e-05, + "loss": 0.1915, + "step": 21823 + }, + { + "epoch": 0.5115889285442196, + "grad_norm": 4.609537601470947, + "learning_rate": 9.639504701237438e-05, + "loss": 0.5174, + "step": 21824 + }, + { + "epoch": 0.5116123701190246, + "grad_norm": 2.29339599609375, + "learning_rate": 9.638768565229957e-05, + "loss": 0.214, + "step": 21825 + }, + { + "epoch": 0.5116358116938295, + "grad_norm": 2.2004692554473877, + "learning_rate": 9.638032431182526e-05, + "loss": 0.1518, + "step": 21826 + }, + { + "epoch": 0.5116592532686346, + "grad_norm": 0.33478444814682007, + "learning_rate": 9.637296299099138e-05, + "loss": 0.0383, + "step": 21827 + }, + { + "epoch": 0.5116826948434395, + "grad_norm": 2.9094631671905518, + "learning_rate": 9.636560168983794e-05, + "loss": 1.2546, + "step": 21828 + }, + { + "epoch": 0.5117061364182446, + "grad_norm": 2.588367462158203, + "learning_rate": 9.635824040840487e-05, + "loss": 0.2242, + "step": 21829 + }, + { + "epoch": 0.5117295779930495, + "grad_norm": 3.058468818664551, + "learning_rate": 9.635087914673204e-05, + "loss": 0.1956, + "step": 21830 + }, + { + "epoch": 0.5117530195678546, + "grad_norm": 1.7329888343811035, + "learning_rate": 9.634351790485948e-05, + "loss": 0.1486, + "step": 21831 + }, + { + "epoch": 0.5117764611426595, + "grad_norm": 1.1823865175247192, + "learning_rate": 9.63361566828271e-05, + "loss": 0.065, + "step": 21832 + }, + { + "epoch": 0.5117999027174646, + "grad_norm": 3.534858465194702, + "learning_rate": 9.632879548067481e-05, + "loss": 0.4926, + "step": 21833 + }, + { + "epoch": 0.5118233442922695, + "grad_norm": 1.6490848064422607, + "learning_rate": 9.632143429844259e-05, + "loss": 0.2271, + "step": 21834 + }, + { + "epoch": 0.5118467858670745, + "grad_norm": 1.7872706651687622, + "learning_rate": 9.631407313617036e-05, + "loss": 0.1004, + "step": 21835 + }, + { + "epoch": 0.5118702274418796, + "grad_norm": 3.563108205795288, + "learning_rate": 9.63067119938981e-05, + "loss": 0.3671, + "step": 21836 + }, + { + "epoch": 0.5118936690166845, + "grad_norm": 0.4479895234107971, + "learning_rate": 9.629935087166571e-05, + "loss": 0.0455, + "step": 21837 + }, + { + "epoch": 0.5119171105914896, + "grad_norm": 2.7184741497039795, + "learning_rate": 9.629198976951314e-05, + "loss": 0.2534, + "step": 21838 + }, + { + "epoch": 0.5119405521662945, + "grad_norm": 1.69025456905365, + "learning_rate": 9.628462868748035e-05, + "loss": 0.1121, + "step": 21839 + }, + { + "epoch": 0.5119639937410996, + "grad_norm": 1.6994200944900513, + "learning_rate": 9.627726762560722e-05, + "loss": 1.7088, + "step": 21840 + }, + { + "epoch": 0.5119874353159045, + "grad_norm": 1.1322365999221802, + "learning_rate": 9.62699065839338e-05, + "loss": 0.0702, + "step": 21841 + }, + { + "epoch": 0.5120108768907096, + "grad_norm": 2.2329893112182617, + "learning_rate": 9.626254556249994e-05, + "loss": 0.1866, + "step": 21842 + }, + { + "epoch": 0.5120343184655145, + "grad_norm": 2.0784971714019775, + "learning_rate": 9.625518456134564e-05, + "loss": 1.0352, + "step": 21843 + }, + { + "epoch": 0.5120577600403196, + "grad_norm": 2.1264054775238037, + "learning_rate": 9.624782358051081e-05, + "loss": 0.1576, + "step": 21844 + }, + { + "epoch": 0.5120812016151245, + "grad_norm": 3.163930892944336, + "learning_rate": 9.624046262003537e-05, + "loss": 2.0821, + "step": 21845 + }, + { + "epoch": 0.5121046431899295, + "grad_norm": 3.666095018386841, + "learning_rate": 9.62331016799593e-05, + "loss": 0.2095, + "step": 21846 + }, + { + "epoch": 0.5121280847647345, + "grad_norm": 1.9144750833511353, + "learning_rate": 9.622574076032254e-05, + "loss": 0.1293, + "step": 21847 + }, + { + "epoch": 0.5121515263395395, + "grad_norm": 1.575276255607605, + "learning_rate": 9.6218379861165e-05, + "loss": 0.1289, + "step": 21848 + }, + { + "epoch": 0.5121749679143445, + "grad_norm": 3.5830953121185303, + "learning_rate": 9.621101898252664e-05, + "loss": 0.3429, + "step": 21849 + }, + { + "epoch": 0.5121984094891495, + "grad_norm": 2.781109571456909, + "learning_rate": 9.620365812444741e-05, + "loss": 1.3847, + "step": 21850 + }, + { + "epoch": 0.5122218510639545, + "grad_norm": 4.279513359069824, + "learning_rate": 9.619629728696722e-05, + "loss": 0.382, + "step": 21851 + }, + { + "epoch": 0.5122452926387595, + "grad_norm": 3.6991026401519775, + "learning_rate": 9.6188936470126e-05, + "loss": 0.2086, + "step": 21852 + }, + { + "epoch": 0.5122687342135644, + "grad_norm": 3.721607208251953, + "learning_rate": 9.618157567396376e-05, + "loss": 0.38, + "step": 21853 + }, + { + "epoch": 0.5122921757883695, + "grad_norm": 5.418306827545166, + "learning_rate": 9.61742148985204e-05, + "loss": 0.2725, + "step": 21854 + }, + { + "epoch": 0.5123156173631744, + "grad_norm": 2.2767465114593506, + "learning_rate": 9.616685414383583e-05, + "loss": 0.2159, + "step": 21855 + }, + { + "epoch": 0.5123390589379795, + "grad_norm": 2.793919324874878, + "learning_rate": 9.615949340995005e-05, + "loss": 0.322, + "step": 21856 + }, + { + "epoch": 0.5123625005127844, + "grad_norm": 3.750152826309204, + "learning_rate": 9.615213269690296e-05, + "loss": 0.3033, + "step": 21857 + }, + { + "epoch": 0.5123859420875895, + "grad_norm": 0.36678504943847656, + "learning_rate": 9.614477200473448e-05, + "loss": 0.0206, + "step": 21858 + }, + { + "epoch": 0.5124093836623944, + "grad_norm": 3.333735704421997, + "learning_rate": 9.613741133348461e-05, + "loss": 0.2004, + "step": 21859 + }, + { + "epoch": 0.5124328252371995, + "grad_norm": 1.8873356580734253, + "learning_rate": 9.613005068319324e-05, + "loss": 0.1522, + "step": 21860 + }, + { + "epoch": 0.5124562668120044, + "grad_norm": 0.6502966284751892, + "learning_rate": 9.61226900539003e-05, + "loss": 0.059, + "step": 21861 + }, + { + "epoch": 0.5124797083868095, + "grad_norm": 5.772049903869629, + "learning_rate": 9.61153294456458e-05, + "loss": 0.1578, + "step": 21862 + }, + { + "epoch": 0.5125031499616144, + "grad_norm": 3.067046642303467, + "learning_rate": 9.610796885846961e-05, + "loss": 1.5968, + "step": 21863 + }, + { + "epoch": 0.5125265915364194, + "grad_norm": 3.758852243423462, + "learning_rate": 9.610060829241166e-05, + "loss": 0.1908, + "step": 21864 + }, + { + "epoch": 0.5125500331112244, + "grad_norm": 1.4611150026321411, + "learning_rate": 9.609324774751192e-05, + "loss": 0.1376, + "step": 21865 + }, + { + "epoch": 0.5125734746860294, + "grad_norm": 2.3306057453155518, + "learning_rate": 9.608588722381037e-05, + "loss": 0.2491, + "step": 21866 + }, + { + "epoch": 0.5125969162608344, + "grad_norm": 1.6474049091339111, + "learning_rate": 9.60785267213469e-05, + "loss": 0.1398, + "step": 21867 + }, + { + "epoch": 0.5126203578356394, + "grad_norm": 2.834791660308838, + "learning_rate": 9.607116624016143e-05, + "loss": 0.2227, + "step": 21868 + }, + { + "epoch": 0.5126437994104444, + "grad_norm": 3.1489968299865723, + "learning_rate": 9.606380578029395e-05, + "loss": 0.2116, + "step": 21869 + }, + { + "epoch": 0.5126672409852494, + "grad_norm": 3.0976126194000244, + "learning_rate": 9.605644534178437e-05, + "loss": 0.4932, + "step": 21870 + }, + { + "epoch": 0.5126906825600543, + "grad_norm": 3.2925047874450684, + "learning_rate": 9.604908492467259e-05, + "loss": 0.2558, + "step": 21871 + }, + { + "epoch": 0.5127141241348594, + "grad_norm": 0.852817714214325, + "learning_rate": 9.604172452899862e-05, + "loss": 0.0523, + "step": 21872 + }, + { + "epoch": 0.5127375657096643, + "grad_norm": 1.6697983741760254, + "learning_rate": 9.603436415480235e-05, + "loss": 0.1673, + "step": 21873 + }, + { + "epoch": 0.5127610072844694, + "grad_norm": 0.35584118962287903, + "learning_rate": 9.602700380212374e-05, + "loss": 0.0462, + "step": 21874 + }, + { + "epoch": 0.5127844488592743, + "grad_norm": 2.2018985748291016, + "learning_rate": 9.601964347100272e-05, + "loss": 0.2588, + "step": 21875 + }, + { + "epoch": 0.5128078904340794, + "grad_norm": 1.56086266040802, + "learning_rate": 9.601228316147918e-05, + "loss": 0.1541, + "step": 21876 + }, + { + "epoch": 0.5128313320088843, + "grad_norm": 1.1328495740890503, + "learning_rate": 9.600492287359314e-05, + "loss": 0.126, + "step": 21877 + }, + { + "epoch": 0.5128547735836894, + "grad_norm": 4.1296610832214355, + "learning_rate": 9.599756260738452e-05, + "loss": 0.4489, + "step": 21878 + }, + { + "epoch": 0.5128782151584943, + "grad_norm": 3.7506704330444336, + "learning_rate": 9.599020236289323e-05, + "loss": 0.4279, + "step": 21879 + }, + { + "epoch": 0.5129016567332993, + "grad_norm": 3.5924365520477295, + "learning_rate": 9.59828421401592e-05, + "loss": 0.3379, + "step": 21880 + }, + { + "epoch": 0.5129250983081043, + "grad_norm": 1.1332119703292847, + "learning_rate": 9.59754819392224e-05, + "loss": 0.0881, + "step": 21881 + }, + { + "epoch": 0.5129485398829093, + "grad_norm": 2.7825517654418945, + "learning_rate": 9.596812176012275e-05, + "loss": 0.2666, + "step": 21882 + }, + { + "epoch": 0.5129719814577143, + "grad_norm": 2.7614355087280273, + "learning_rate": 9.596076160290017e-05, + "loss": 0.1405, + "step": 21883 + }, + { + "epoch": 0.5129954230325193, + "grad_norm": 4.1594672203063965, + "learning_rate": 9.595340146759462e-05, + "loss": 0.3273, + "step": 21884 + }, + { + "epoch": 0.5130188646073243, + "grad_norm": 4.518322467803955, + "learning_rate": 9.594604135424603e-05, + "loss": 0.1859, + "step": 21885 + }, + { + "epoch": 0.5130423061821293, + "grad_norm": 1.7955490350723267, + "learning_rate": 9.593868126289431e-05, + "loss": 0.1015, + "step": 21886 + }, + { + "epoch": 0.5130657477569344, + "grad_norm": 3.7147765159606934, + "learning_rate": 9.593132119357945e-05, + "loss": 0.5134, + "step": 21887 + }, + { + "epoch": 0.5130891893317393, + "grad_norm": 3.9584286212921143, + "learning_rate": 9.59239611463413e-05, + "loss": 0.2069, + "step": 21888 + }, + { + "epoch": 0.5131126309065444, + "grad_norm": 4.212172508239746, + "learning_rate": 9.591660112121991e-05, + "loss": 0.3905, + "step": 21889 + }, + { + "epoch": 0.5131360724813493, + "grad_norm": 1.5239614248275757, + "learning_rate": 9.590924111825512e-05, + "loss": 0.1561, + "step": 21890 + }, + { + "epoch": 0.5131595140561543, + "grad_norm": 4.107086658477783, + "learning_rate": 9.590188113748692e-05, + "loss": 0.3356, + "step": 21891 + }, + { + "epoch": 0.5131829556309593, + "grad_norm": 2.629565954208374, + "learning_rate": 9.589452117895526e-05, + "loss": 1.8968, + "step": 21892 + }, + { + "epoch": 0.5132063972057643, + "grad_norm": 3.515075445175171, + "learning_rate": 9.588716124269999e-05, + "loss": 0.3308, + "step": 21893 + }, + { + "epoch": 0.5132298387805693, + "grad_norm": 2.072032928466797, + "learning_rate": 9.587980132876113e-05, + "loss": 0.158, + "step": 21894 + }, + { + "epoch": 0.5132532803553743, + "grad_norm": 1.747840166091919, + "learning_rate": 9.587244143717859e-05, + "loss": 0.0746, + "step": 21895 + }, + { + "epoch": 0.5132767219301793, + "grad_norm": 0.4230935275554657, + "learning_rate": 9.586508156799227e-05, + "loss": 0.027, + "step": 21896 + }, + { + "epoch": 0.5133001635049843, + "grad_norm": 5.0552287101745605, + "learning_rate": 9.585772172124216e-05, + "loss": 0.3337, + "step": 21897 + }, + { + "epoch": 0.5133236050797892, + "grad_norm": 3.050462007522583, + "learning_rate": 9.585036189696816e-05, + "loss": 0.2661, + "step": 21898 + }, + { + "epoch": 0.5133470466545943, + "grad_norm": 3.2985618114471436, + "learning_rate": 9.58430020952102e-05, + "loss": 0.2389, + "step": 21899 + }, + { + "epoch": 0.5133704882293992, + "grad_norm": 0.9650824069976807, + "learning_rate": 9.583564231600821e-05, + "loss": 0.1312, + "step": 21900 + }, + { + "epoch": 0.5133939298042043, + "grad_norm": 0.8229609131813049, + "learning_rate": 9.582828255940218e-05, + "loss": 0.0671, + "step": 21901 + }, + { + "epoch": 0.5134173713790092, + "grad_norm": 3.3690969944000244, + "learning_rate": 9.5820922825432e-05, + "loss": 0.2527, + "step": 21902 + }, + { + "epoch": 0.5134408129538143, + "grad_norm": 3.1453685760498047, + "learning_rate": 9.58135631141376e-05, + "loss": 1.7971, + "step": 21903 + }, + { + "epoch": 0.5134642545286192, + "grad_norm": 5.283724308013916, + "learning_rate": 9.580620342555893e-05, + "loss": 0.3528, + "step": 21904 + }, + { + "epoch": 0.5134876961034243, + "grad_norm": 2.7458302974700928, + "learning_rate": 9.579884375973593e-05, + "loss": 0.3005, + "step": 21905 + }, + { + "epoch": 0.5135111376782292, + "grad_norm": 4.612495422363281, + "learning_rate": 9.57914841167085e-05, + "loss": 0.4445, + "step": 21906 + }, + { + "epoch": 0.5135345792530343, + "grad_norm": 3.6116185188293457, + "learning_rate": 9.578412449651661e-05, + "loss": 0.2289, + "step": 21907 + }, + { + "epoch": 0.5135580208278392, + "grad_norm": 2.8648648262023926, + "learning_rate": 9.577676489920017e-05, + "loss": 0.237, + "step": 21908 + }, + { + "epoch": 0.5135814624026442, + "grad_norm": 1.3920536041259766, + "learning_rate": 9.576940532479914e-05, + "loss": 0.1825, + "step": 21909 + }, + { + "epoch": 0.5136049039774492, + "grad_norm": 3.498295307159424, + "learning_rate": 9.576204577335343e-05, + "loss": 0.328, + "step": 21910 + }, + { + "epoch": 0.5136283455522542, + "grad_norm": 4.723217487335205, + "learning_rate": 9.575468624490296e-05, + "loss": 0.6305, + "step": 21911 + }, + { + "epoch": 0.5136517871270592, + "grad_norm": 3.7780561447143555, + "learning_rate": 9.574732673948767e-05, + "loss": 0.3973, + "step": 21912 + }, + { + "epoch": 0.5136752287018642, + "grad_norm": 0.8845635652542114, + "learning_rate": 9.573996725714752e-05, + "loss": 0.0563, + "step": 21913 + }, + { + "epoch": 0.5136986702766692, + "grad_norm": 3.1529674530029297, + "learning_rate": 9.573260779792245e-05, + "loss": 0.3784, + "step": 21914 + }, + { + "epoch": 0.5137221118514742, + "grad_norm": 0.6835020780563354, + "learning_rate": 9.572524836185235e-05, + "loss": 0.0499, + "step": 21915 + }, + { + "epoch": 0.5137455534262791, + "grad_norm": 3.486058473587036, + "learning_rate": 9.571788894897719e-05, + "loss": 0.3686, + "step": 21916 + }, + { + "epoch": 0.5137689950010842, + "grad_norm": 2.2605412006378174, + "learning_rate": 9.571052955933687e-05, + "loss": 0.1065, + "step": 21917 + }, + { + "epoch": 0.5137924365758891, + "grad_norm": 1.545462965965271, + "learning_rate": 9.570317019297133e-05, + "loss": 0.146, + "step": 21918 + }, + { + "epoch": 0.5138158781506942, + "grad_norm": 4.275519847869873, + "learning_rate": 9.569581084992053e-05, + "loss": 0.4629, + "step": 21919 + }, + { + "epoch": 0.5138393197254991, + "grad_norm": 1.6308284997940063, + "learning_rate": 9.568845153022439e-05, + "loss": 0.0687, + "step": 21920 + }, + { + "epoch": 0.5138627613003042, + "grad_norm": 1.526688575744629, + "learning_rate": 9.568109223392279e-05, + "loss": 0.223, + "step": 21921 + }, + { + "epoch": 0.5138862028751091, + "grad_norm": 6.078229904174805, + "learning_rate": 9.567373296105574e-05, + "loss": 0.5917, + "step": 21922 + }, + { + "epoch": 0.5139096444499142, + "grad_norm": 2.345266342163086, + "learning_rate": 9.566637371166314e-05, + "loss": 0.1442, + "step": 21923 + }, + { + "epoch": 0.5139330860247191, + "grad_norm": 3.1105854511260986, + "learning_rate": 9.565901448578489e-05, + "loss": 0.2822, + "step": 21924 + }, + { + "epoch": 0.5139565275995241, + "grad_norm": 5.279313564300537, + "learning_rate": 9.565165528346093e-05, + "loss": 0.3721, + "step": 21925 + }, + { + "epoch": 0.5139799691743291, + "grad_norm": 4.892674446105957, + "learning_rate": 9.564429610473125e-05, + "loss": 0.2768, + "step": 21926 + }, + { + "epoch": 0.5140034107491341, + "grad_norm": 2.3734002113342285, + "learning_rate": 9.563693694963575e-05, + "loss": 1.6393, + "step": 21927 + }, + { + "epoch": 0.5140268523239391, + "grad_norm": 3.610548973083496, + "learning_rate": 9.562957781821433e-05, + "loss": 0.366, + "step": 21928 + }, + { + "epoch": 0.5140502938987441, + "grad_norm": 4.565196990966797, + "learning_rate": 9.562221871050697e-05, + "loss": 0.644, + "step": 21929 + }, + { + "epoch": 0.5140737354735491, + "grad_norm": 4.797708034515381, + "learning_rate": 9.561485962655356e-05, + "loss": 0.2851, + "step": 21930 + }, + { + "epoch": 0.5140971770483541, + "grad_norm": 4.652833461761475, + "learning_rate": 9.560750056639402e-05, + "loss": 0.6074, + "step": 21931 + }, + { + "epoch": 0.514120618623159, + "grad_norm": 0.41996175050735474, + "learning_rate": 9.560014153006834e-05, + "loss": 0.0163, + "step": 21932 + }, + { + "epoch": 0.5141440601979641, + "grad_norm": 1.2572965621948242, + "learning_rate": 9.559278251761641e-05, + "loss": 0.8644, + "step": 21933 + }, + { + "epoch": 0.514167501772769, + "grad_norm": 2.925387144088745, + "learning_rate": 9.558542352907816e-05, + "loss": 1.4547, + "step": 21934 + }, + { + "epoch": 0.5141909433475741, + "grad_norm": 2.604372262954712, + "learning_rate": 9.557806456449352e-05, + "loss": 0.2776, + "step": 21935 + }, + { + "epoch": 0.514214384922379, + "grad_norm": 3.753671646118164, + "learning_rate": 9.557070562390244e-05, + "loss": 0.3205, + "step": 21936 + }, + { + "epoch": 0.5142378264971841, + "grad_norm": 2.537806510925293, + "learning_rate": 9.556334670734479e-05, + "loss": 1.5411, + "step": 21937 + }, + { + "epoch": 0.5142612680719891, + "grad_norm": 4.114151477813721, + "learning_rate": 9.555598781486058e-05, + "loss": 0.5784, + "step": 21938 + }, + { + "epoch": 0.5142847096467941, + "grad_norm": 1.169853925704956, + "learning_rate": 9.554862894648971e-05, + "loss": 0.1276, + "step": 21939 + }, + { + "epoch": 0.5143081512215991, + "grad_norm": 3.1537587642669678, + "learning_rate": 9.554127010227212e-05, + "loss": 0.155, + "step": 21940 + }, + { + "epoch": 0.5143315927964041, + "grad_norm": 0.5777316689491272, + "learning_rate": 9.55339112822477e-05, + "loss": 0.0582, + "step": 21941 + }, + { + "epoch": 0.5143550343712091, + "grad_norm": 3.808908462524414, + "learning_rate": 9.552655248645641e-05, + "loss": 0.3688, + "step": 21942 + }, + { + "epoch": 0.514378475946014, + "grad_norm": 0.7885637879371643, + "learning_rate": 9.551919371493818e-05, + "loss": 0.0716, + "step": 21943 + }, + { + "epoch": 0.5144019175208191, + "grad_norm": 2.336635112762451, + "learning_rate": 9.551183496773291e-05, + "loss": 0.1594, + "step": 21944 + }, + { + "epoch": 0.514425359095624, + "grad_norm": 1.343941569328308, + "learning_rate": 9.550447624488057e-05, + "loss": 0.1, + "step": 21945 + }, + { + "epoch": 0.5144488006704291, + "grad_norm": 3.415503740310669, + "learning_rate": 9.549711754642104e-05, + "loss": 0.3092, + "step": 21946 + }, + { + "epoch": 0.514472242245234, + "grad_norm": 0.6554749011993408, + "learning_rate": 9.54897588723943e-05, + "loss": 0.0607, + "step": 21947 + }, + { + "epoch": 0.5144956838200391, + "grad_norm": 0.48249033093452454, + "learning_rate": 9.548240022284028e-05, + "loss": 0.0589, + "step": 21948 + }, + { + "epoch": 0.514519125394844, + "grad_norm": 0.4070843458175659, + "learning_rate": 9.54750415977988e-05, + "loss": 0.0304, + "step": 21949 + }, + { + "epoch": 0.5145425669696491, + "grad_norm": 2.124483585357666, + "learning_rate": 9.546768299730995e-05, + "loss": 0.2626, + "step": 21950 + }, + { + "epoch": 0.514566008544454, + "grad_norm": 3.399599075317383, + "learning_rate": 9.546032442141353e-05, + "loss": 0.2648, + "step": 21951 + }, + { + "epoch": 0.514589450119259, + "grad_norm": 1.8080494403839111, + "learning_rate": 9.545296587014957e-05, + "loss": 0.9672, + "step": 21952 + }, + { + "epoch": 0.514612891694064, + "grad_norm": 3.346524238586426, + "learning_rate": 9.54456073435579e-05, + "loss": 0.3503, + "step": 21953 + }, + { + "epoch": 0.514636333268869, + "grad_norm": 0.5703389048576355, + "learning_rate": 9.543824884167852e-05, + "loss": 0.0651, + "step": 21954 + }, + { + "epoch": 0.514659774843674, + "grad_norm": 0.5950151085853577, + "learning_rate": 9.543089036455133e-05, + "loss": 0.0428, + "step": 21955 + }, + { + "epoch": 0.514683216418479, + "grad_norm": 0.9419391751289368, + "learning_rate": 9.542353191221623e-05, + "loss": 0.0604, + "step": 21956 + }, + { + "epoch": 0.514706657993284, + "grad_norm": 4.815430164337158, + "learning_rate": 9.541617348471322e-05, + "loss": 0.4736, + "step": 21957 + }, + { + "epoch": 0.514730099568089, + "grad_norm": 3.377877950668335, + "learning_rate": 9.540881508208216e-05, + "loss": 0.3907, + "step": 21958 + }, + { + "epoch": 0.514753541142894, + "grad_norm": 1.5176883935928345, + "learning_rate": 9.540145670436298e-05, + "loss": 0.1393, + "step": 21959 + }, + { + "epoch": 0.514776982717699, + "grad_norm": 1.9194000959396362, + "learning_rate": 9.539409835159564e-05, + "loss": 1.5388, + "step": 21960 + }, + { + "epoch": 0.5148004242925039, + "grad_norm": 3.5779457092285156, + "learning_rate": 9.538674002382003e-05, + "loss": 0.3189, + "step": 21961 + }, + { + "epoch": 0.514823865867309, + "grad_norm": 3.920079231262207, + "learning_rate": 9.537938172107612e-05, + "loss": 0.544, + "step": 21962 + }, + { + "epoch": 0.5148473074421139, + "grad_norm": 0.980311930179596, + "learning_rate": 9.537202344340379e-05, + "loss": 0.1475, + "step": 21963 + }, + { + "epoch": 0.514870749016919, + "grad_norm": 3.4908671379089355, + "learning_rate": 9.536466519084303e-05, + "loss": 0.2822, + "step": 21964 + }, + { + "epoch": 0.5148941905917239, + "grad_norm": 1.2299680709838867, + "learning_rate": 9.535730696343372e-05, + "loss": 0.1033, + "step": 21965 + }, + { + "epoch": 0.514917632166529, + "grad_norm": 1.2761627435684204, + "learning_rate": 9.534994876121578e-05, + "loss": 0.1135, + "step": 21966 + }, + { + "epoch": 0.5149410737413339, + "grad_norm": 1.0123165845870972, + "learning_rate": 9.534259058422915e-05, + "loss": 0.1427, + "step": 21967 + }, + { + "epoch": 0.514964515316139, + "grad_norm": 3.042614221572876, + "learning_rate": 9.533523243251377e-05, + "loss": 0.3168, + "step": 21968 + }, + { + "epoch": 0.5149879568909439, + "grad_norm": 5.691685199737549, + "learning_rate": 9.532787430610952e-05, + "loss": 0.265, + "step": 21969 + }, + { + "epoch": 0.515011398465749, + "grad_norm": 3.795320749282837, + "learning_rate": 9.532051620505639e-05, + "loss": 0.3171, + "step": 21970 + }, + { + "epoch": 0.5150348400405539, + "grad_norm": 5.9677348136901855, + "learning_rate": 9.531315812939426e-05, + "loss": 0.2862, + "step": 21971 + }, + { + "epoch": 0.5150582816153589, + "grad_norm": 3.655447483062744, + "learning_rate": 9.530580007916305e-05, + "loss": 0.2586, + "step": 21972 + }, + { + "epoch": 0.5150817231901639, + "grad_norm": 3.3884618282318115, + "learning_rate": 9.529844205440267e-05, + "loss": 0.3897, + "step": 21973 + }, + { + "epoch": 0.5151051647649689, + "grad_norm": 6.779545783996582, + "learning_rate": 9.529108405515313e-05, + "loss": 0.8687, + "step": 21974 + }, + { + "epoch": 0.5151286063397739, + "grad_norm": 1.5183175802230835, + "learning_rate": 9.52837260814543e-05, + "loss": 0.1558, + "step": 21975 + }, + { + "epoch": 0.5151520479145789, + "grad_norm": 3.416396141052246, + "learning_rate": 9.527636813334607e-05, + "loss": 0.3308, + "step": 21976 + }, + { + "epoch": 0.5151754894893839, + "grad_norm": 3.0539021492004395, + "learning_rate": 9.526901021086843e-05, + "loss": 0.298, + "step": 21977 + }, + { + "epoch": 0.5151989310641889, + "grad_norm": 2.2098307609558105, + "learning_rate": 9.526165231406128e-05, + "loss": 1.3547, + "step": 21978 + }, + { + "epoch": 0.5152223726389938, + "grad_norm": 2.870309352874756, + "learning_rate": 9.52542944429645e-05, + "loss": 0.3198, + "step": 21979 + }, + { + "epoch": 0.5152458142137989, + "grad_norm": 0.6520829200744629, + "learning_rate": 9.524693659761808e-05, + "loss": 0.0691, + "step": 21980 + }, + { + "epoch": 0.5152692557886038, + "grad_norm": 1.7017977237701416, + "learning_rate": 9.523957877806192e-05, + "loss": 0.191, + "step": 21981 + }, + { + "epoch": 0.5152926973634089, + "grad_norm": 0.45428550243377686, + "learning_rate": 9.523222098433592e-05, + "loss": 0.0366, + "step": 21982 + }, + { + "epoch": 0.5153161389382138, + "grad_norm": 2.006007194519043, + "learning_rate": 9.522486321648005e-05, + "loss": 0.1668, + "step": 21983 + }, + { + "epoch": 0.5153395805130189, + "grad_norm": 1.363167405128479, + "learning_rate": 9.521750547453416e-05, + "loss": 0.0512, + "step": 21984 + }, + { + "epoch": 0.5153630220878238, + "grad_norm": 2.6858015060424805, + "learning_rate": 9.521014775853824e-05, + "loss": 0.1662, + "step": 21985 + }, + { + "epoch": 0.5153864636626289, + "grad_norm": 4.8372344970703125, + "learning_rate": 9.520279006853219e-05, + "loss": 0.281, + "step": 21986 + }, + { + "epoch": 0.5154099052374338, + "grad_norm": 1.005486249923706, + "learning_rate": 9.519543240455595e-05, + "loss": 0.0479, + "step": 21987 + }, + { + "epoch": 0.5154333468122388, + "grad_norm": 2.972083806991577, + "learning_rate": 9.518807476664944e-05, + "loss": 0.3395, + "step": 21988 + }, + { + "epoch": 0.5154567883870439, + "grad_norm": 3.0370066165924072, + "learning_rate": 9.518071715485255e-05, + "loss": 0.2195, + "step": 21989 + }, + { + "epoch": 0.5154802299618488, + "grad_norm": 2.7172021865844727, + "learning_rate": 9.517335956920524e-05, + "loss": 0.4682, + "step": 21990 + }, + { + "epoch": 0.5155036715366539, + "grad_norm": 2.5320441722869873, + "learning_rate": 9.516600200974739e-05, + "loss": 0.1497, + "step": 21991 + }, + { + "epoch": 0.5155271131114588, + "grad_norm": 3.006044864654541, + "learning_rate": 9.515864447651897e-05, + "loss": 0.3159, + "step": 21992 + }, + { + "epoch": 0.5155505546862639, + "grad_norm": 0.6190634965896606, + "learning_rate": 9.515128696955989e-05, + "loss": 0.054, + "step": 21993 + }, + { + "epoch": 0.5155739962610688, + "grad_norm": 2.4245858192443848, + "learning_rate": 9.514392948891004e-05, + "loss": 0.1824, + "step": 21994 + }, + { + "epoch": 0.5155974378358739, + "grad_norm": 5.404728889465332, + "learning_rate": 9.513657203460938e-05, + "loss": 0.6386, + "step": 21995 + }, + { + "epoch": 0.5156208794106788, + "grad_norm": 2.5727591514587402, + "learning_rate": 9.512921460669783e-05, + "loss": 0.2387, + "step": 21996 + }, + { + "epoch": 0.5156443209854839, + "grad_norm": 0.5861360430717468, + "learning_rate": 9.512185720521527e-05, + "loss": 0.0597, + "step": 21997 + }, + { + "epoch": 0.5156677625602888, + "grad_norm": 1.4415773153305054, + "learning_rate": 9.511449983020163e-05, + "loss": 0.1534, + "step": 21998 + }, + { + "epoch": 0.5156912041350938, + "grad_norm": 4.958859443664551, + "learning_rate": 9.510714248169689e-05, + "loss": 0.4863, + "step": 21999 + }, + { + "epoch": 0.5157146457098988, + "grad_norm": 0.8926721215248108, + "learning_rate": 9.509978515974093e-05, + "loss": 0.0949, + "step": 22000 + }, + { + "epoch": 0.5157380872847038, + "grad_norm": 1.7921034097671509, + "learning_rate": 9.509242786437365e-05, + "loss": 0.1972, + "step": 22001 + }, + { + "epoch": 0.5157615288595088, + "grad_norm": 2.6863925457000732, + "learning_rate": 9.508507059563503e-05, + "loss": 0.4655, + "step": 22002 + }, + { + "epoch": 0.5157849704343138, + "grad_norm": 3.380424976348877, + "learning_rate": 9.507771335356495e-05, + "loss": 0.327, + "step": 22003 + }, + { + "epoch": 0.5158084120091188, + "grad_norm": 3.9470598697662354, + "learning_rate": 9.507035613820332e-05, + "loss": 0.4086, + "step": 22004 + }, + { + "epoch": 0.5158318535839238, + "grad_norm": 2.439014434814453, + "learning_rate": 9.506299894959007e-05, + "loss": 0.2835, + "step": 22005 + }, + { + "epoch": 0.5158552951587287, + "grad_norm": 2.7368152141571045, + "learning_rate": 9.505564178776516e-05, + "loss": 0.1557, + "step": 22006 + }, + { + "epoch": 0.5158787367335338, + "grad_norm": 2.807716131210327, + "learning_rate": 9.504828465276844e-05, + "loss": 0.2706, + "step": 22007 + }, + { + "epoch": 0.5159021783083387, + "grad_norm": 3.1040656566619873, + "learning_rate": 9.504092754463989e-05, + "loss": 0.4438, + "step": 22008 + }, + { + "epoch": 0.5159256198831438, + "grad_norm": 0.6787630319595337, + "learning_rate": 9.50335704634194e-05, + "loss": 0.0783, + "step": 22009 + }, + { + "epoch": 0.5159490614579487, + "grad_norm": 2.49662446975708, + "learning_rate": 9.502621340914686e-05, + "loss": 0.1488, + "step": 22010 + }, + { + "epoch": 0.5159725030327538, + "grad_norm": 2.910301446914673, + "learning_rate": 9.501885638186224e-05, + "loss": 0.2318, + "step": 22011 + }, + { + "epoch": 0.5159959446075587, + "grad_norm": 2.523083448410034, + "learning_rate": 9.501149938160549e-05, + "loss": 0.1634, + "step": 22012 + }, + { + "epoch": 0.5160193861823638, + "grad_norm": 3.0245063304901123, + "learning_rate": 9.500414240841647e-05, + "loss": 0.2394, + "step": 22013 + }, + { + "epoch": 0.5160428277571687, + "grad_norm": 1.4604566097259521, + "learning_rate": 9.499678546233508e-05, + "loss": 0.133, + "step": 22014 + }, + { + "epoch": 0.5160662693319737, + "grad_norm": 1.9755240678787231, + "learning_rate": 9.498942854340131e-05, + "loss": 0.24, + "step": 22015 + }, + { + "epoch": 0.5160897109067787, + "grad_norm": 0.7922939658164978, + "learning_rate": 9.498207165165503e-05, + "loss": 0.0766, + "step": 22016 + }, + { + "epoch": 0.5161131524815837, + "grad_norm": 2.9607958793640137, + "learning_rate": 9.497471478713615e-05, + "loss": 1.5356, + "step": 22017 + }, + { + "epoch": 0.5161365940563887, + "grad_norm": 3.929363489151001, + "learning_rate": 9.496735794988465e-05, + "loss": 0.3077, + "step": 22018 + }, + { + "epoch": 0.5161600356311937, + "grad_norm": 2.4137017726898193, + "learning_rate": 9.49600011399404e-05, + "loss": 0.1731, + "step": 22019 + }, + { + "epoch": 0.5161834772059987, + "grad_norm": 3.0993740558624268, + "learning_rate": 9.495264435734327e-05, + "loss": 0.231, + "step": 22020 + }, + { + "epoch": 0.5162069187808037, + "grad_norm": 3.844550371170044, + "learning_rate": 9.494528760213329e-05, + "loss": 0.4404, + "step": 22021 + }, + { + "epoch": 0.5162303603556087, + "grad_norm": 3.1018800735473633, + "learning_rate": 9.493793087435027e-05, + "loss": 0.3013, + "step": 22022 + }, + { + "epoch": 0.5162538019304137, + "grad_norm": 1.1181873083114624, + "learning_rate": 9.493057417403422e-05, + "loss": 0.1001, + "step": 22023 + }, + { + "epoch": 0.5162772435052186, + "grad_norm": 2.2414512634277344, + "learning_rate": 9.492321750122501e-05, + "loss": 0.2252, + "step": 22024 + }, + { + "epoch": 0.5163006850800237, + "grad_norm": 3.621964454650879, + "learning_rate": 9.491586085596257e-05, + "loss": 0.2111, + "step": 22025 + }, + { + "epoch": 0.5163241266548286, + "grad_norm": 4.664168357849121, + "learning_rate": 9.49085042382868e-05, + "loss": 0.7625, + "step": 22026 + }, + { + "epoch": 0.5163475682296337, + "grad_norm": 6.176337242126465, + "learning_rate": 9.490114764823765e-05, + "loss": 0.7046, + "step": 22027 + }, + { + "epoch": 0.5163710098044386, + "grad_norm": 3.4859654903411865, + "learning_rate": 9.4893791085855e-05, + "loss": 1.6753, + "step": 22028 + }, + { + "epoch": 0.5163944513792437, + "grad_norm": 2.7636992931365967, + "learning_rate": 9.488643455117879e-05, + "loss": 0.3274, + "step": 22029 + }, + { + "epoch": 0.5164178929540486, + "grad_norm": 2.088441848754883, + "learning_rate": 9.487907804424893e-05, + "loss": 0.1471, + "step": 22030 + }, + { + "epoch": 0.5164413345288537, + "grad_norm": 1.2158340215682983, + "learning_rate": 9.487172156510533e-05, + "loss": 0.1029, + "step": 22031 + }, + { + "epoch": 0.5164647761036586, + "grad_norm": 3.5152084827423096, + "learning_rate": 9.486436511378791e-05, + "loss": 0.2352, + "step": 22032 + }, + { + "epoch": 0.5164882176784636, + "grad_norm": 5.199075222015381, + "learning_rate": 9.485700869033661e-05, + "loss": 0.9177, + "step": 22033 + }, + { + "epoch": 0.5165116592532686, + "grad_norm": 0.9270088076591492, + "learning_rate": 9.484965229479127e-05, + "loss": 0.0573, + "step": 22034 + }, + { + "epoch": 0.5165351008280736, + "grad_norm": 2.265699863433838, + "learning_rate": 9.48422959271919e-05, + "loss": 0.2282, + "step": 22035 + }, + { + "epoch": 0.5165585424028786, + "grad_norm": 2.372098207473755, + "learning_rate": 9.483493958757838e-05, + "loss": 0.1161, + "step": 22036 + }, + { + "epoch": 0.5165819839776836, + "grad_norm": 1.1559702157974243, + "learning_rate": 9.482758327599062e-05, + "loss": 0.0325, + "step": 22037 + }, + { + "epoch": 0.5166054255524886, + "grad_norm": 3.7704672813415527, + "learning_rate": 9.482022699246855e-05, + "loss": 0.4009, + "step": 22038 + }, + { + "epoch": 0.5166288671272936, + "grad_norm": 3.3475797176361084, + "learning_rate": 9.481287073705205e-05, + "loss": 0.3717, + "step": 22039 + }, + { + "epoch": 0.5166523087020987, + "grad_norm": 2.601473331451416, + "learning_rate": 9.480551450978109e-05, + "loss": 1.1702, + "step": 22040 + }, + { + "epoch": 0.5166757502769036, + "grad_norm": 0.7083301544189453, + "learning_rate": 9.479815831069555e-05, + "loss": 0.0395, + "step": 22041 + }, + { + "epoch": 0.5166991918517087, + "grad_norm": 4.607386112213135, + "learning_rate": 9.479080213983533e-05, + "loss": 0.4823, + "step": 22042 + }, + { + "epoch": 0.5167226334265136, + "grad_norm": 2.85605788230896, + "learning_rate": 9.478344599724038e-05, + "loss": 1.7023, + "step": 22043 + }, + { + "epoch": 0.5167460750013186, + "grad_norm": 3.127103090286255, + "learning_rate": 9.477608988295059e-05, + "loss": 0.4193, + "step": 22044 + }, + { + "epoch": 0.5167695165761236, + "grad_norm": 3.755023717880249, + "learning_rate": 9.476873379700588e-05, + "loss": 0.3985, + "step": 22045 + }, + { + "epoch": 0.5167929581509286, + "grad_norm": 2.800307273864746, + "learning_rate": 9.476137773944614e-05, + "loss": 0.3794, + "step": 22046 + }, + { + "epoch": 0.5168163997257336, + "grad_norm": 3.657461166381836, + "learning_rate": 9.475402171031133e-05, + "loss": 0.4862, + "step": 22047 + }, + { + "epoch": 0.5168398413005386, + "grad_norm": 6.020986557006836, + "learning_rate": 9.474666570964137e-05, + "loss": 0.4527, + "step": 22048 + }, + { + "epoch": 0.5168632828753436, + "grad_norm": 0.7353612184524536, + "learning_rate": 9.473930973747613e-05, + "loss": 0.0512, + "step": 22049 + }, + { + "epoch": 0.5168867244501486, + "grad_norm": 0.7835706472396851, + "learning_rate": 9.473195379385557e-05, + "loss": 0.0629, + "step": 22050 + }, + { + "epoch": 0.5169101660249535, + "grad_norm": 0.7351489067077637, + "learning_rate": 9.472459787881957e-05, + "loss": 0.3187, + "step": 22051 + }, + { + "epoch": 0.5169336075997586, + "grad_norm": 2.811828136444092, + "learning_rate": 9.471724199240801e-05, + "loss": 0.2749, + "step": 22052 + }, + { + "epoch": 0.5169570491745635, + "grad_norm": 3.3472611904144287, + "learning_rate": 9.470988613466089e-05, + "loss": 0.2097, + "step": 22053 + }, + { + "epoch": 0.5169804907493686, + "grad_norm": 0.7028676867485046, + "learning_rate": 9.470253030561806e-05, + "loss": 0.044, + "step": 22054 + }, + { + "epoch": 0.5170039323241735, + "grad_norm": 4.039394855499268, + "learning_rate": 9.469517450531944e-05, + "loss": 0.4708, + "step": 22055 + }, + { + "epoch": 0.5170273738989786, + "grad_norm": 0.8647756576538086, + "learning_rate": 9.468781873380497e-05, + "loss": 0.1025, + "step": 22056 + }, + { + "epoch": 0.5170508154737835, + "grad_norm": 3.552130937576294, + "learning_rate": 9.468046299111454e-05, + "loss": 0.3218, + "step": 22057 + }, + { + "epoch": 0.5170742570485886, + "grad_norm": 2.8001868724823, + "learning_rate": 9.467310727728803e-05, + "loss": 1.525, + "step": 22058 + }, + { + "epoch": 0.5170976986233935, + "grad_norm": 1.2258107662200928, + "learning_rate": 9.466575159236541e-05, + "loss": 0.1833, + "step": 22059 + }, + { + "epoch": 0.5171211401981985, + "grad_norm": 2.1561121940612793, + "learning_rate": 9.46583959363866e-05, + "loss": 0.2842, + "step": 22060 + }, + { + "epoch": 0.5171445817730035, + "grad_norm": 4.930257797241211, + "learning_rate": 9.465104030939147e-05, + "loss": 0.368, + "step": 22061 + }, + { + "epoch": 0.5171680233478085, + "grad_norm": 3.5927014350891113, + "learning_rate": 9.464368471141994e-05, + "loss": 0.3869, + "step": 22062 + }, + { + "epoch": 0.5171914649226135, + "grad_norm": 2.5801424980163574, + "learning_rate": 9.463632914251193e-05, + "loss": 0.2098, + "step": 22063 + }, + { + "epoch": 0.5172149064974185, + "grad_norm": 2.5648441314697266, + "learning_rate": 9.462897360270734e-05, + "loss": 0.2134, + "step": 22064 + }, + { + "epoch": 0.5172383480722235, + "grad_norm": 2.5906107425689697, + "learning_rate": 9.46216180920461e-05, + "loss": 1.6141, + "step": 22065 + }, + { + "epoch": 0.5172617896470285, + "grad_norm": 3.97225284576416, + "learning_rate": 9.461426261056812e-05, + "loss": 0.5358, + "step": 22066 + }, + { + "epoch": 0.5172852312218335, + "grad_norm": 2.48404598236084, + "learning_rate": 9.460690715831328e-05, + "loss": 0.1486, + "step": 22067 + }, + { + "epoch": 0.5173086727966385, + "grad_norm": 4.92503547668457, + "learning_rate": 9.459955173532154e-05, + "loss": 0.503, + "step": 22068 + }, + { + "epoch": 0.5173321143714434, + "grad_norm": 2.054560422897339, + "learning_rate": 9.459219634163279e-05, + "loss": 0.2133, + "step": 22069 + }, + { + "epoch": 0.5173555559462485, + "grad_norm": 1.9623526334762573, + "learning_rate": 9.45848409772869e-05, + "loss": 0.0969, + "step": 22070 + }, + { + "epoch": 0.5173789975210534, + "grad_norm": 2.117875337600708, + "learning_rate": 9.45774856423238e-05, + "loss": 0.2093, + "step": 22071 + }, + { + "epoch": 0.5174024390958585, + "grad_norm": 0.23475989699363708, + "learning_rate": 9.457013033678346e-05, + "loss": 0.0133, + "step": 22072 + }, + { + "epoch": 0.5174258806706634, + "grad_norm": 4.94104528427124, + "learning_rate": 9.456277506070573e-05, + "loss": 0.4875, + "step": 22073 + }, + { + "epoch": 0.5174493222454685, + "grad_norm": 3.5002355575561523, + "learning_rate": 9.455541981413053e-05, + "loss": 0.2652, + "step": 22074 + }, + { + "epoch": 0.5174727638202734, + "grad_norm": 1.8389347791671753, + "learning_rate": 9.45480645970978e-05, + "loss": 0.2768, + "step": 22075 + }, + { + "epoch": 0.5174962053950785, + "grad_norm": 4.8335700035095215, + "learning_rate": 9.454070940964741e-05, + "loss": 0.4621, + "step": 22076 + }, + { + "epoch": 0.5175196469698834, + "grad_norm": 2.870598316192627, + "learning_rate": 9.453335425181926e-05, + "loss": 0.2179, + "step": 22077 + }, + { + "epoch": 0.5175430885446884, + "grad_norm": 3.0352489948272705, + "learning_rate": 9.452599912365333e-05, + "loss": 0.3053, + "step": 22078 + }, + { + "epoch": 0.5175665301194934, + "grad_norm": 3.2616190910339355, + "learning_rate": 9.451864402518947e-05, + "loss": 0.3522, + "step": 22079 + }, + { + "epoch": 0.5175899716942984, + "grad_norm": 4.143038272857666, + "learning_rate": 9.451128895646758e-05, + "loss": 0.3998, + "step": 22080 + }, + { + "epoch": 0.5176134132691034, + "grad_norm": 2.829091787338257, + "learning_rate": 9.45039339175276e-05, + "loss": 0.3524, + "step": 22081 + }, + { + "epoch": 0.5176368548439084, + "grad_norm": 4.627856731414795, + "learning_rate": 9.449657890840942e-05, + "loss": 0.4787, + "step": 22082 + }, + { + "epoch": 0.5176602964187134, + "grad_norm": 4.174689769744873, + "learning_rate": 9.448922392915295e-05, + "loss": 0.4279, + "step": 22083 + }, + { + "epoch": 0.5176837379935184, + "grad_norm": 0.7596393823623657, + "learning_rate": 9.44818689797981e-05, + "loss": 0.0442, + "step": 22084 + }, + { + "epoch": 0.5177071795683234, + "grad_norm": 1.6631582975387573, + "learning_rate": 9.447451406038483e-05, + "loss": 0.1214, + "step": 22085 + }, + { + "epoch": 0.5177306211431284, + "grad_norm": 1.437326192855835, + "learning_rate": 9.446715917095298e-05, + "loss": 0.1356, + "step": 22086 + }, + { + "epoch": 0.5177540627179333, + "grad_norm": 8.769331932067871, + "learning_rate": 9.445980431154247e-05, + "loss": 0.2422, + "step": 22087 + }, + { + "epoch": 0.5177775042927384, + "grad_norm": 3.545888662338257, + "learning_rate": 9.445244948219323e-05, + "loss": 0.6886, + "step": 22088 + }, + { + "epoch": 0.5178009458675433, + "grad_norm": 4.079319000244141, + "learning_rate": 9.444509468294515e-05, + "loss": 0.3146, + "step": 22089 + }, + { + "epoch": 0.5178243874423484, + "grad_norm": 1.740508794784546, + "learning_rate": 9.443773991383814e-05, + "loss": 0.1812, + "step": 22090 + }, + { + "epoch": 0.5178478290171534, + "grad_norm": 3.456897258758545, + "learning_rate": 9.443038517491212e-05, + "loss": 0.1996, + "step": 22091 + }, + { + "epoch": 0.5178712705919584, + "grad_norm": 3.383999824523926, + "learning_rate": 9.442303046620699e-05, + "loss": 0.278, + "step": 22092 + }, + { + "epoch": 0.5178947121667634, + "grad_norm": 2.8007233142852783, + "learning_rate": 9.441567578776263e-05, + "loss": 0.2305, + "step": 22093 + }, + { + "epoch": 0.5179181537415684, + "grad_norm": 0.5586198568344116, + "learning_rate": 9.440832113961899e-05, + "loss": 0.0565, + "step": 22094 + }, + { + "epoch": 0.5179415953163734, + "grad_norm": 2.8476874828338623, + "learning_rate": 9.440096652181593e-05, + "loss": 0.334, + "step": 22095 + }, + { + "epoch": 0.5179650368911783, + "grad_norm": 2.325345993041992, + "learning_rate": 9.43936119343934e-05, + "loss": 0.1774, + "step": 22096 + }, + { + "epoch": 0.5179884784659834, + "grad_norm": 0.2942754030227661, + "learning_rate": 9.438625737739129e-05, + "loss": 0.0213, + "step": 22097 + }, + { + "epoch": 0.5180119200407883, + "grad_norm": 0.8758678436279297, + "learning_rate": 9.437890285084951e-05, + "loss": 0.0581, + "step": 22098 + }, + { + "epoch": 0.5180353616155934, + "grad_norm": 3.3392269611358643, + "learning_rate": 9.437154835480797e-05, + "loss": 0.3832, + "step": 22099 + }, + { + "epoch": 0.5180588031903983, + "grad_norm": 3.098308563232422, + "learning_rate": 9.436419388930654e-05, + "loss": 0.3486, + "step": 22100 + }, + { + "epoch": 0.5180822447652034, + "grad_norm": 3.149024248123169, + "learning_rate": 9.435683945438518e-05, + "loss": 0.2845, + "step": 22101 + }, + { + "epoch": 0.5181056863400083, + "grad_norm": 3.933832883834839, + "learning_rate": 9.434948505008374e-05, + "loss": 0.4244, + "step": 22102 + }, + { + "epoch": 0.5181291279148134, + "grad_norm": 2.6586105823516846, + "learning_rate": 9.434213067644219e-05, + "loss": 1.5918, + "step": 22103 + }, + { + "epoch": 0.5181525694896183, + "grad_norm": 27.618194580078125, + "learning_rate": 9.433477633350038e-05, + "loss": 0.2165, + "step": 22104 + }, + { + "epoch": 0.5181760110644233, + "grad_norm": 4.318343162536621, + "learning_rate": 9.432742202129822e-05, + "loss": 0.2019, + "step": 22105 + }, + { + "epoch": 0.5181994526392283, + "grad_norm": 3.622480869293213, + "learning_rate": 9.432006773987564e-05, + "loss": 0.3521, + "step": 22106 + }, + { + "epoch": 0.5182228942140333, + "grad_norm": 4.504343509674072, + "learning_rate": 9.431271348927249e-05, + "loss": 0.4337, + "step": 22107 + }, + { + "epoch": 0.5182463357888383, + "grad_norm": 2.2876455783843994, + "learning_rate": 9.430535926952879e-05, + "loss": 0.2651, + "step": 22108 + }, + { + "epoch": 0.5182697773636433, + "grad_norm": 1.026265025138855, + "learning_rate": 9.429800508068431e-05, + "loss": 0.0969, + "step": 22109 + }, + { + "epoch": 0.5182932189384483, + "grad_norm": 3.651339530944824, + "learning_rate": 9.429065092277905e-05, + "loss": 0.4, + "step": 22110 + }, + { + "epoch": 0.5183166605132533, + "grad_norm": 1.1940323114395142, + "learning_rate": 9.428329679585289e-05, + "loss": 0.1024, + "step": 22111 + }, + { + "epoch": 0.5183401020880583, + "grad_norm": 1.4751474857330322, + "learning_rate": 9.427594269994568e-05, + "loss": 0.0888, + "step": 22112 + }, + { + "epoch": 0.5183635436628633, + "grad_norm": 2.298335552215576, + "learning_rate": 9.42685886350974e-05, + "loss": 0.1816, + "step": 22113 + }, + { + "epoch": 0.5183869852376682, + "grad_norm": 5.45119047164917, + "learning_rate": 9.426123460134792e-05, + "loss": 0.8824, + "step": 22114 + }, + { + "epoch": 0.5184104268124733, + "grad_norm": 2.758417844772339, + "learning_rate": 9.425388059873713e-05, + "loss": 1.3957, + "step": 22115 + }, + { + "epoch": 0.5184338683872782, + "grad_norm": 5.1664605140686035, + "learning_rate": 9.424652662730495e-05, + "loss": 0.5218, + "step": 22116 + }, + { + "epoch": 0.5184573099620833, + "grad_norm": 5.679549217224121, + "learning_rate": 9.423917268709129e-05, + "loss": 0.4671, + "step": 22117 + }, + { + "epoch": 0.5184807515368882, + "grad_norm": 3.0280814170837402, + "learning_rate": 9.4231818778136e-05, + "loss": 0.2285, + "step": 22118 + }, + { + "epoch": 0.5185041931116933, + "grad_norm": 2.070408582687378, + "learning_rate": 9.422446490047902e-05, + "loss": 0.1228, + "step": 22119 + }, + { + "epoch": 0.5185276346864982, + "grad_norm": 2.763345718383789, + "learning_rate": 9.421711105416029e-05, + "loss": 1.4606, + "step": 22120 + }, + { + "epoch": 0.5185510762613033, + "grad_norm": 1.948904275894165, + "learning_rate": 9.420975723921967e-05, + "loss": 0.2457, + "step": 22121 + }, + { + "epoch": 0.5185745178361082, + "grad_norm": 1.667417049407959, + "learning_rate": 9.420240345569706e-05, + "loss": 1.1794, + "step": 22122 + }, + { + "epoch": 0.5185979594109132, + "grad_norm": 3.2443020343780518, + "learning_rate": 9.419504970363239e-05, + "loss": 0.1816, + "step": 22123 + }, + { + "epoch": 0.5186214009857182, + "grad_norm": 1.4669984579086304, + "learning_rate": 9.418769598306556e-05, + "loss": 0.1433, + "step": 22124 + }, + { + "epoch": 0.5186448425605232, + "grad_norm": 2.7754428386688232, + "learning_rate": 9.418034229403641e-05, + "loss": 0.2255, + "step": 22125 + }, + { + "epoch": 0.5186682841353282, + "grad_norm": 3.8608241081237793, + "learning_rate": 9.417298863658493e-05, + "loss": 0.5074, + "step": 22126 + }, + { + "epoch": 0.5186917257101332, + "grad_norm": 1.7436518669128418, + "learning_rate": 9.416563501075096e-05, + "loss": 0.1345, + "step": 22127 + }, + { + "epoch": 0.5187151672849382, + "grad_norm": 4.849158763885498, + "learning_rate": 9.415828141657441e-05, + "loss": 0.423, + "step": 22128 + }, + { + "epoch": 0.5187386088597432, + "grad_norm": 5.315583229064941, + "learning_rate": 9.41509278540952e-05, + "loss": 0.6803, + "step": 22129 + }, + { + "epoch": 0.5187620504345481, + "grad_norm": 2.671590805053711, + "learning_rate": 9.414357432335322e-05, + "loss": 0.2642, + "step": 22130 + }, + { + "epoch": 0.5187854920093532, + "grad_norm": 1.101311445236206, + "learning_rate": 9.413622082438834e-05, + "loss": 0.1019, + "step": 22131 + }, + { + "epoch": 0.5188089335841581, + "grad_norm": 2.7337045669555664, + "learning_rate": 9.41288673572405e-05, + "loss": 0.251, + "step": 22132 + }, + { + "epoch": 0.5188323751589632, + "grad_norm": 4.636272430419922, + "learning_rate": 9.412151392194961e-05, + "loss": 0.442, + "step": 22133 + }, + { + "epoch": 0.5188558167337681, + "grad_norm": 3.7505955696105957, + "learning_rate": 9.411416051855556e-05, + "loss": 0.1701, + "step": 22134 + }, + { + "epoch": 0.5188792583085732, + "grad_norm": 1.7525080442428589, + "learning_rate": 9.410680714709822e-05, + "loss": 0.1541, + "step": 22135 + }, + { + "epoch": 0.5189026998833781, + "grad_norm": 0.5729225277900696, + "learning_rate": 9.409945380761752e-05, + "loss": 0.044, + "step": 22136 + }, + { + "epoch": 0.5189261414581832, + "grad_norm": 2.883511543273926, + "learning_rate": 9.409210050015336e-05, + "loss": 0.2431, + "step": 22137 + }, + { + "epoch": 0.5189495830329881, + "grad_norm": 3.610497236251831, + "learning_rate": 9.408474722474561e-05, + "loss": 0.2793, + "step": 22138 + }, + { + "epoch": 0.5189730246077932, + "grad_norm": 0.7001115083694458, + "learning_rate": 9.407739398143419e-05, + "loss": 0.0269, + "step": 22139 + }, + { + "epoch": 0.5189964661825981, + "grad_norm": 3.8984854221343994, + "learning_rate": 9.4070040770259e-05, + "loss": 0.2432, + "step": 22140 + }, + { + "epoch": 0.5190199077574031, + "grad_norm": 0.4197228252887726, + "learning_rate": 9.406268759125995e-05, + "loss": 0.0377, + "step": 22141 + }, + { + "epoch": 0.5190433493322082, + "grad_norm": 2.376699209213257, + "learning_rate": 9.405533444447691e-05, + "loss": 0.1725, + "step": 22142 + }, + { + "epoch": 0.5190667909070131, + "grad_norm": 3.20458984375, + "learning_rate": 9.404798132994976e-05, + "loss": 0.219, + "step": 22143 + }, + { + "epoch": 0.5190902324818182, + "grad_norm": 4.084804058074951, + "learning_rate": 9.404062824771846e-05, + "loss": 0.2407, + "step": 22144 + }, + { + "epoch": 0.5191136740566231, + "grad_norm": 2.3505990505218506, + "learning_rate": 9.403327519782288e-05, + "loss": 1.2656, + "step": 22145 + }, + { + "epoch": 0.5191371156314282, + "grad_norm": 1.592130422592163, + "learning_rate": 9.402592218030293e-05, + "loss": 0.1302, + "step": 22146 + }, + { + "epoch": 0.5191605572062331, + "grad_norm": 3.647359848022461, + "learning_rate": 9.401856919519847e-05, + "loss": 0.6354, + "step": 22147 + }, + { + "epoch": 0.5191839987810382, + "grad_norm": 5.408194541931152, + "learning_rate": 9.401121624254945e-05, + "loss": 0.5453, + "step": 22148 + }, + { + "epoch": 0.5192074403558431, + "grad_norm": 1.9438358545303345, + "learning_rate": 9.400386332239572e-05, + "loss": 0.2748, + "step": 22149 + }, + { + "epoch": 0.5192308819306481, + "grad_norm": 0.7018835544586182, + "learning_rate": 9.39965104347772e-05, + "loss": 0.0565, + "step": 22150 + }, + { + "epoch": 0.5192543235054531, + "grad_norm": 3.6295366287231445, + "learning_rate": 9.398915757973379e-05, + "loss": 0.2344, + "step": 22151 + }, + { + "epoch": 0.5192777650802581, + "grad_norm": 1.08893620967865, + "learning_rate": 9.398180475730538e-05, + "loss": 0.0652, + "step": 22152 + }, + { + "epoch": 0.5193012066550631, + "grad_norm": 0.9875664114952087, + "learning_rate": 9.397445196753185e-05, + "loss": 0.0891, + "step": 22153 + }, + { + "epoch": 0.5193246482298681, + "grad_norm": 4.1820878982543945, + "learning_rate": 9.396709921045313e-05, + "loss": 0.5158, + "step": 22154 + }, + { + "epoch": 0.5193480898046731, + "grad_norm": 2.1873676776885986, + "learning_rate": 9.39597464861091e-05, + "loss": 1.2426, + "step": 22155 + }, + { + "epoch": 0.5193715313794781, + "grad_norm": 3.9468905925750732, + "learning_rate": 9.395239379453962e-05, + "loss": 0.329, + "step": 22156 + }, + { + "epoch": 0.519394972954283, + "grad_norm": 3.3357596397399902, + "learning_rate": 9.394504113578463e-05, + "loss": 0.3843, + "step": 22157 + }, + { + "epoch": 0.5194184145290881, + "grad_norm": 0.8784940242767334, + "learning_rate": 9.393768850988405e-05, + "loss": 0.0899, + "step": 22158 + }, + { + "epoch": 0.519441856103893, + "grad_norm": 2.6926727294921875, + "learning_rate": 9.393033591687773e-05, + "loss": 0.2491, + "step": 22159 + }, + { + "epoch": 0.5194652976786981, + "grad_norm": 1.9368106126785278, + "learning_rate": 9.392298335680556e-05, + "loss": 1.4123, + "step": 22160 + }, + { + "epoch": 0.519488739253503, + "grad_norm": 5.7452497482299805, + "learning_rate": 9.391563082970749e-05, + "loss": 0.3876, + "step": 22161 + }, + { + "epoch": 0.5195121808283081, + "grad_norm": 4.306260585784912, + "learning_rate": 9.390827833562337e-05, + "loss": 0.5558, + "step": 22162 + }, + { + "epoch": 0.519535622403113, + "grad_norm": 4.6793599128723145, + "learning_rate": 9.390092587459308e-05, + "loss": 0.6387, + "step": 22163 + }, + { + "epoch": 0.5195590639779181, + "grad_norm": 3.236845016479492, + "learning_rate": 9.389357344665655e-05, + "loss": 0.1799, + "step": 22164 + }, + { + "epoch": 0.519582505552723, + "grad_norm": 1.4153048992156982, + "learning_rate": 9.388622105185368e-05, + "loss": 0.1492, + "step": 22165 + }, + { + "epoch": 0.5196059471275281, + "grad_norm": 2.951101779937744, + "learning_rate": 9.38788686902243e-05, + "loss": 0.1894, + "step": 22166 + }, + { + "epoch": 0.519629388702333, + "grad_norm": 0.3748001158237457, + "learning_rate": 9.387151636180838e-05, + "loss": 0.0288, + "step": 22167 + }, + { + "epoch": 0.519652830277138, + "grad_norm": 2.9918205738067627, + "learning_rate": 9.386416406664575e-05, + "loss": 0.3149, + "step": 22168 + }, + { + "epoch": 0.519676271851943, + "grad_norm": 1.381913661956787, + "learning_rate": 9.385681180477639e-05, + "loss": 0.1361, + "step": 22169 + }, + { + "epoch": 0.519699713426748, + "grad_norm": 3.2425150871276855, + "learning_rate": 9.38494595762401e-05, + "loss": 0.4521, + "step": 22170 + }, + { + "epoch": 0.519723155001553, + "grad_norm": 0.408370703458786, + "learning_rate": 9.384210738107685e-05, + "loss": 0.0233, + "step": 22171 + }, + { + "epoch": 0.519746596576358, + "grad_norm": 3.144649028778076, + "learning_rate": 9.38347552193265e-05, + "loss": 0.205, + "step": 22172 + }, + { + "epoch": 0.519770038151163, + "grad_norm": 5.0315937995910645, + "learning_rate": 9.38274030910289e-05, + "loss": 0.4215, + "step": 22173 + }, + { + "epoch": 0.519793479725968, + "grad_norm": 0.960130512714386, + "learning_rate": 9.382005099622404e-05, + "loss": 0.1225, + "step": 22174 + }, + { + "epoch": 0.519816921300773, + "grad_norm": 1.1469331979751587, + "learning_rate": 9.381269893495173e-05, + "loss": 0.136, + "step": 22175 + }, + { + "epoch": 0.519840362875578, + "grad_norm": 4.705083847045898, + "learning_rate": 9.380534690725188e-05, + "loss": 0.1448, + "step": 22176 + }, + { + "epoch": 0.5198638044503829, + "grad_norm": 1.2691748142242432, + "learning_rate": 9.379799491316441e-05, + "loss": 0.1567, + "step": 22177 + }, + { + "epoch": 0.519887246025188, + "grad_norm": 2.6487462520599365, + "learning_rate": 9.379064295272917e-05, + "loss": 0.1994, + "step": 22178 + }, + { + "epoch": 0.5199106875999929, + "grad_norm": 4.0304365158081055, + "learning_rate": 9.378329102598611e-05, + "loss": 0.554, + "step": 22179 + }, + { + "epoch": 0.519934129174798, + "grad_norm": 2.6550707817077637, + "learning_rate": 9.377593913297502e-05, + "loss": 0.3737, + "step": 22180 + }, + { + "epoch": 0.5199575707496029, + "grad_norm": 1.3901721239089966, + "learning_rate": 9.376858727373592e-05, + "loss": 0.1069, + "step": 22181 + }, + { + "epoch": 0.519981012324408, + "grad_norm": 2.250689744949341, + "learning_rate": 9.376123544830862e-05, + "loss": 0.1797, + "step": 22182 + }, + { + "epoch": 0.5200044538992129, + "grad_norm": 3.191291093826294, + "learning_rate": 9.375388365673305e-05, + "loss": 0.2966, + "step": 22183 + }, + { + "epoch": 0.520027895474018, + "grad_norm": 0.3871791362762451, + "learning_rate": 9.374653189904908e-05, + "loss": 0.0134, + "step": 22184 + }, + { + "epoch": 0.5200513370488229, + "grad_norm": 3.9701623916625977, + "learning_rate": 9.373918017529658e-05, + "loss": 0.4324, + "step": 22185 + }, + { + "epoch": 0.5200747786236279, + "grad_norm": 3.4339189529418945, + "learning_rate": 9.373182848551549e-05, + "loss": 0.3912, + "step": 22186 + }, + { + "epoch": 0.5200982201984329, + "grad_norm": 2.3934545516967773, + "learning_rate": 9.372447682974568e-05, + "loss": 1.471, + "step": 22187 + }, + { + "epoch": 0.5201216617732379, + "grad_norm": 2.6517889499664307, + "learning_rate": 9.371712520802701e-05, + "loss": 0.2535, + "step": 22188 + }, + { + "epoch": 0.5201451033480429, + "grad_norm": 1.9281386137008667, + "learning_rate": 9.370977362039941e-05, + "loss": 0.1834, + "step": 22189 + }, + { + "epoch": 0.5201685449228479, + "grad_norm": 1.1916985511779785, + "learning_rate": 9.370242206690277e-05, + "loss": 0.1342, + "step": 22190 + }, + { + "epoch": 0.5201919864976529, + "grad_norm": 2.465916872024536, + "learning_rate": 9.369507054757694e-05, + "loss": 0.1216, + "step": 22191 + }, + { + "epoch": 0.5202154280724579, + "grad_norm": 2.54085111618042, + "learning_rate": 9.36877190624618e-05, + "loss": 0.2927, + "step": 22192 + }, + { + "epoch": 0.520238869647263, + "grad_norm": 4.361405849456787, + "learning_rate": 9.368036761159732e-05, + "loss": 0.5329, + "step": 22193 + }, + { + "epoch": 0.5202623112220679, + "grad_norm": 2.795646905899048, + "learning_rate": 9.367301619502335e-05, + "loss": 0.2132, + "step": 22194 + }, + { + "epoch": 0.520285752796873, + "grad_norm": 4.400868892669678, + "learning_rate": 9.366566481277974e-05, + "loss": 0.5224, + "step": 22195 + }, + { + "epoch": 0.5203091943716779, + "grad_norm": 3.306699752807617, + "learning_rate": 9.365831346490645e-05, + "loss": 0.312, + "step": 22196 + }, + { + "epoch": 0.5203326359464829, + "grad_norm": 3.474940061569214, + "learning_rate": 9.365096215144332e-05, + "loss": 0.3841, + "step": 22197 + }, + { + "epoch": 0.5203560775212879, + "grad_norm": 2.453476667404175, + "learning_rate": 9.364361087243022e-05, + "loss": 0.3436, + "step": 22198 + }, + { + "epoch": 0.5203795190960929, + "grad_norm": 1.1784298419952393, + "learning_rate": 9.36362596279071e-05, + "loss": 0.1091, + "step": 22199 + }, + { + "epoch": 0.5204029606708979, + "grad_norm": 2.084686040878296, + "learning_rate": 9.362890841791381e-05, + "loss": 0.2746, + "step": 22200 + }, + { + "epoch": 0.5204264022457029, + "grad_norm": 3.1231138706207275, + "learning_rate": 9.36215572424902e-05, + "loss": 0.3735, + "step": 22201 + }, + { + "epoch": 0.5204498438205079, + "grad_norm": 1.343219518661499, + "learning_rate": 9.361420610167625e-05, + "loss": 0.127, + "step": 22202 + }, + { + "epoch": 0.5204732853953129, + "grad_norm": 4.187291622161865, + "learning_rate": 9.360685499551178e-05, + "loss": 0.2775, + "step": 22203 + }, + { + "epoch": 0.5204967269701178, + "grad_norm": 2.208585262298584, + "learning_rate": 9.359950392403666e-05, + "loss": 0.2712, + "step": 22204 + }, + { + "epoch": 0.5205201685449229, + "grad_norm": 2.554891586303711, + "learning_rate": 9.359215288729083e-05, + "loss": 0.1623, + "step": 22205 + }, + { + "epoch": 0.5205436101197278, + "grad_norm": 4.309299468994141, + "learning_rate": 9.358480188531419e-05, + "loss": 0.3363, + "step": 22206 + }, + { + "epoch": 0.5205670516945329, + "grad_norm": 3.882519006729126, + "learning_rate": 9.357745091814659e-05, + "loss": 0.3591, + "step": 22207 + }, + { + "epoch": 0.5205904932693378, + "grad_norm": 5.299320220947266, + "learning_rate": 9.357009998582789e-05, + "loss": 0.6071, + "step": 22208 + }, + { + "epoch": 0.5206139348441429, + "grad_norm": 0.3632643222808838, + "learning_rate": 9.356274908839804e-05, + "loss": 0.0398, + "step": 22209 + }, + { + "epoch": 0.5206373764189478, + "grad_norm": 3.4513235092163086, + "learning_rate": 9.35553982258969e-05, + "loss": 0.377, + "step": 22210 + }, + { + "epoch": 0.5206608179937529, + "grad_norm": 2.980983018875122, + "learning_rate": 9.354804739836432e-05, + "loss": 0.2744, + "step": 22211 + }, + { + "epoch": 0.5206842595685578, + "grad_norm": 2.093000650405884, + "learning_rate": 9.354069660584023e-05, + "loss": 0.1323, + "step": 22212 + }, + { + "epoch": 0.5207077011433628, + "grad_norm": 1.6832736730575562, + "learning_rate": 9.353334584836449e-05, + "loss": 1.492, + "step": 22213 + }, + { + "epoch": 0.5207311427181678, + "grad_norm": 1.7706975936889648, + "learning_rate": 9.352599512597702e-05, + "loss": 1.0172, + "step": 22214 + }, + { + "epoch": 0.5207545842929728, + "grad_norm": 2.436276435852051, + "learning_rate": 9.351864443871768e-05, + "loss": 1.2466, + "step": 22215 + }, + { + "epoch": 0.5207780258677778, + "grad_norm": 3.5159108638763428, + "learning_rate": 9.351129378662631e-05, + "loss": 0.3936, + "step": 22216 + }, + { + "epoch": 0.5208014674425828, + "grad_norm": 1.951917052268982, + "learning_rate": 9.35039431697429e-05, + "loss": 0.1625, + "step": 22217 + }, + { + "epoch": 0.5208249090173878, + "grad_norm": 0.5899054408073425, + "learning_rate": 9.349659258810723e-05, + "loss": 0.0486, + "step": 22218 + }, + { + "epoch": 0.5208483505921928, + "grad_norm": 1.9255980253219604, + "learning_rate": 9.348924204175927e-05, + "loss": 0.3175, + "step": 22219 + }, + { + "epoch": 0.5208717921669977, + "grad_norm": 3.9904375076293945, + "learning_rate": 9.348189153073884e-05, + "loss": 0.3565, + "step": 22220 + }, + { + "epoch": 0.5208952337418028, + "grad_norm": 1.9952806234359741, + "learning_rate": 9.347454105508587e-05, + "loss": 0.1123, + "step": 22221 + }, + { + "epoch": 0.5209186753166077, + "grad_norm": 4.5819525718688965, + "learning_rate": 9.346719061484024e-05, + "loss": 0.2867, + "step": 22222 + }, + { + "epoch": 0.5209421168914128, + "grad_norm": 0.4267653822898865, + "learning_rate": 9.345984021004179e-05, + "loss": 0.0571, + "step": 22223 + }, + { + "epoch": 0.5209655584662177, + "grad_norm": 1.0391645431518555, + "learning_rate": 9.345248984073042e-05, + "loss": 0.0636, + "step": 22224 + }, + { + "epoch": 0.5209890000410228, + "grad_norm": 1.8143051862716675, + "learning_rate": 9.344513950694607e-05, + "loss": 0.0917, + "step": 22225 + }, + { + "epoch": 0.5210124416158277, + "grad_norm": 0.844369113445282, + "learning_rate": 9.343778920872853e-05, + "loss": 0.092, + "step": 22226 + }, + { + "epoch": 0.5210358831906328, + "grad_norm": 2.599773645401001, + "learning_rate": 9.343043894611775e-05, + "loss": 0.3392, + "step": 22227 + }, + { + "epoch": 0.5210593247654377, + "grad_norm": 4.665065765380859, + "learning_rate": 9.342308871915361e-05, + "loss": 0.496, + "step": 22228 + }, + { + "epoch": 0.5210827663402428, + "grad_norm": 0.8857790231704712, + "learning_rate": 9.341573852787591e-05, + "loss": 0.0881, + "step": 22229 + }, + { + "epoch": 0.5211062079150477, + "grad_norm": 2.930053472518921, + "learning_rate": 9.340838837232464e-05, + "loss": 1.486, + "step": 22230 + }, + { + "epoch": 0.5211296494898527, + "grad_norm": 2.116478681564331, + "learning_rate": 9.340103825253965e-05, + "loss": 1.5663, + "step": 22231 + }, + { + "epoch": 0.5211530910646577, + "grad_norm": 4.972635746002197, + "learning_rate": 9.339368816856082e-05, + "loss": 0.5618, + "step": 22232 + }, + { + "epoch": 0.5211765326394627, + "grad_norm": 5.121503829956055, + "learning_rate": 9.338633812042801e-05, + "loss": 0.3001, + "step": 22233 + }, + { + "epoch": 0.5211999742142677, + "grad_norm": 3.627902030944824, + "learning_rate": 9.337898810818112e-05, + "loss": 0.4085, + "step": 22234 + }, + { + "epoch": 0.5212234157890727, + "grad_norm": 3.323637008666992, + "learning_rate": 9.337163813186003e-05, + "loss": 0.2733, + "step": 22235 + }, + { + "epoch": 0.5212468573638777, + "grad_norm": 2.047607183456421, + "learning_rate": 9.33642881915046e-05, + "loss": 0.1178, + "step": 22236 + }, + { + "epoch": 0.5212702989386827, + "grad_norm": 3.7074928283691406, + "learning_rate": 9.335693828715475e-05, + "loss": 0.3161, + "step": 22237 + }, + { + "epoch": 0.5212937405134876, + "grad_norm": 4.2852864265441895, + "learning_rate": 9.334958841885034e-05, + "loss": 0.2697, + "step": 22238 + }, + { + "epoch": 0.5213171820882927, + "grad_norm": 3.761723279953003, + "learning_rate": 9.334223858663122e-05, + "loss": 0.4406, + "step": 22239 + }, + { + "epoch": 0.5213406236630976, + "grad_norm": 4.425027370452881, + "learning_rate": 9.333488879053733e-05, + "loss": 0.2428, + "step": 22240 + }, + { + "epoch": 0.5213640652379027, + "grad_norm": 4.465299606323242, + "learning_rate": 9.332753903060847e-05, + "loss": 0.6127, + "step": 22241 + }, + { + "epoch": 0.5213875068127076, + "grad_norm": 6.067410945892334, + "learning_rate": 9.332018930688462e-05, + "loss": 0.4018, + "step": 22242 + }, + { + "epoch": 0.5214109483875127, + "grad_norm": 1.7537834644317627, + "learning_rate": 9.33128396194056e-05, + "loss": 0.2433, + "step": 22243 + }, + { + "epoch": 0.5214343899623177, + "grad_norm": 1.579412817955017, + "learning_rate": 9.330548996821131e-05, + "loss": 0.2015, + "step": 22244 + }, + { + "epoch": 0.5214578315371227, + "grad_norm": 1.8178972005844116, + "learning_rate": 9.329814035334162e-05, + "loss": 0.0892, + "step": 22245 + }, + { + "epoch": 0.5214812731119277, + "grad_norm": 0.5854407548904419, + "learning_rate": 9.32907907748364e-05, + "loss": 0.0418, + "step": 22246 + }, + { + "epoch": 0.5215047146867327, + "grad_norm": 3.6036274433135986, + "learning_rate": 9.328344123273554e-05, + "loss": 0.3639, + "step": 22247 + }, + { + "epoch": 0.5215281562615377, + "grad_norm": 1.1526172161102295, + "learning_rate": 9.327609172707891e-05, + "loss": 0.0413, + "step": 22248 + }, + { + "epoch": 0.5215515978363426, + "grad_norm": 2.4260387420654297, + "learning_rate": 9.32687422579064e-05, + "loss": 1.2663, + "step": 22249 + }, + { + "epoch": 0.5215750394111477, + "grad_norm": 1.9991482496261597, + "learning_rate": 9.32613928252579e-05, + "loss": 0.2254, + "step": 22250 + }, + { + "epoch": 0.5215984809859526, + "grad_norm": 1.6272159814834595, + "learning_rate": 9.325404342917323e-05, + "loss": 0.183, + "step": 22251 + }, + { + "epoch": 0.5216219225607577, + "grad_norm": 2.1230521202087402, + "learning_rate": 9.324669406969235e-05, + "loss": 0.1912, + "step": 22252 + }, + { + "epoch": 0.5216453641355626, + "grad_norm": 4.515263557434082, + "learning_rate": 9.323934474685505e-05, + "loss": 0.4264, + "step": 22253 + }, + { + "epoch": 0.5216688057103677, + "grad_norm": 2.304287910461426, + "learning_rate": 9.32319954607013e-05, + "loss": 1.4341, + "step": 22254 + }, + { + "epoch": 0.5216922472851726, + "grad_norm": 3.8451244831085205, + "learning_rate": 9.322464621127094e-05, + "loss": 0.2886, + "step": 22255 + }, + { + "epoch": 0.5217156888599777, + "grad_norm": 2.5192453861236572, + "learning_rate": 9.321729699860381e-05, + "loss": 0.1335, + "step": 22256 + }, + { + "epoch": 0.5217391304347826, + "grad_norm": 2.2039971351623535, + "learning_rate": 9.320994782273984e-05, + "loss": 0.1229, + "step": 22257 + }, + { + "epoch": 0.5217625720095876, + "grad_norm": 1.1270817518234253, + "learning_rate": 9.320259868371886e-05, + "loss": 0.1119, + "step": 22258 + }, + { + "epoch": 0.5217860135843926, + "grad_norm": 4.18872594833374, + "learning_rate": 9.319524958158079e-05, + "loss": 0.3656, + "step": 22259 + }, + { + "epoch": 0.5218094551591976, + "grad_norm": 3.5309410095214844, + "learning_rate": 9.318790051636548e-05, + "loss": 0.253, + "step": 22260 + }, + { + "epoch": 0.5218328967340026, + "grad_norm": 1.2248154878616333, + "learning_rate": 9.318055148811281e-05, + "loss": 0.1128, + "step": 22261 + }, + { + "epoch": 0.5218563383088076, + "grad_norm": 2.492527961730957, + "learning_rate": 9.317320249686267e-05, + "loss": 1.0384, + "step": 22262 + }, + { + "epoch": 0.5218797798836126, + "grad_norm": 2.612272262573242, + "learning_rate": 9.316585354265494e-05, + "loss": 0.1958, + "step": 22263 + }, + { + "epoch": 0.5219032214584176, + "grad_norm": 2.7128140926361084, + "learning_rate": 9.315850462552944e-05, + "loss": 1.9922, + "step": 22264 + }, + { + "epoch": 0.5219266630332225, + "grad_norm": 3.2252514362335205, + "learning_rate": 9.315115574552607e-05, + "loss": 0.2289, + "step": 22265 + }, + { + "epoch": 0.5219501046080276, + "grad_norm": 2.5680296421051025, + "learning_rate": 9.314380690268478e-05, + "loss": 0.1474, + "step": 22266 + }, + { + "epoch": 0.5219735461828325, + "grad_norm": 4.381807327270508, + "learning_rate": 9.313645809704537e-05, + "loss": 0.3093, + "step": 22267 + }, + { + "epoch": 0.5219969877576376, + "grad_norm": 2.1129250526428223, + "learning_rate": 9.31291093286477e-05, + "loss": 0.1357, + "step": 22268 + }, + { + "epoch": 0.5220204293324425, + "grad_norm": 4.421219825744629, + "learning_rate": 9.312176059753173e-05, + "loss": 0.3792, + "step": 22269 + }, + { + "epoch": 0.5220438709072476, + "grad_norm": 1.7454837560653687, + "learning_rate": 9.311441190373726e-05, + "loss": 0.9501, + "step": 22270 + }, + { + "epoch": 0.5220673124820525, + "grad_norm": 4.2585978507995605, + "learning_rate": 9.310706324730416e-05, + "loss": 0.3915, + "step": 22271 + }, + { + "epoch": 0.5220907540568576, + "grad_norm": 0.44041943550109863, + "learning_rate": 9.309971462827235e-05, + "loss": 0.0476, + "step": 22272 + }, + { + "epoch": 0.5221141956316625, + "grad_norm": 3.160888910293579, + "learning_rate": 9.30923660466817e-05, + "loss": 1.0345, + "step": 22273 + }, + { + "epoch": 0.5221376372064676, + "grad_norm": 3.034571409225464, + "learning_rate": 9.308501750257203e-05, + "loss": 0.4033, + "step": 22274 + }, + { + "epoch": 0.5221610787812725, + "grad_norm": 4.088990688323975, + "learning_rate": 9.307766899598328e-05, + "loss": 0.2663, + "step": 22275 + }, + { + "epoch": 0.5221845203560775, + "grad_norm": 3.602412700653076, + "learning_rate": 9.307032052695528e-05, + "loss": 0.4591, + "step": 22276 + }, + { + "epoch": 0.5222079619308825, + "grad_norm": 7.066102027893066, + "learning_rate": 9.306297209552788e-05, + "loss": 0.4826, + "step": 22277 + }, + { + "epoch": 0.5222314035056875, + "grad_norm": 2.549769639968872, + "learning_rate": 9.3055623701741e-05, + "loss": 0.282, + "step": 22278 + }, + { + "epoch": 0.5222548450804925, + "grad_norm": 2.7956316471099854, + "learning_rate": 9.304827534563454e-05, + "loss": 0.2985, + "step": 22279 + }, + { + "epoch": 0.5222782866552975, + "grad_norm": 3.4893574714660645, + "learning_rate": 9.304092702724833e-05, + "loss": 0.3528, + "step": 22280 + }, + { + "epoch": 0.5223017282301025, + "grad_norm": 2.1721527576446533, + "learning_rate": 9.30335787466222e-05, + "loss": 0.2447, + "step": 22281 + }, + { + "epoch": 0.5223251698049075, + "grad_norm": 2.5697426795959473, + "learning_rate": 9.302623050379612e-05, + "loss": 0.2215, + "step": 22282 + }, + { + "epoch": 0.5223486113797124, + "grad_norm": 2.697016716003418, + "learning_rate": 9.30188822988099e-05, + "loss": 0.1874, + "step": 22283 + }, + { + "epoch": 0.5223720529545175, + "grad_norm": 3.5779240131378174, + "learning_rate": 9.301153413170339e-05, + "loss": 0.4839, + "step": 22284 + }, + { + "epoch": 0.5223954945293224, + "grad_norm": 3.277618646621704, + "learning_rate": 9.300418600251652e-05, + "loss": 0.2868, + "step": 22285 + }, + { + "epoch": 0.5224189361041275, + "grad_norm": 3.502385377883911, + "learning_rate": 9.299683791128914e-05, + "loss": 0.3745, + "step": 22286 + }, + { + "epoch": 0.5224423776789324, + "grad_norm": 3.714233875274658, + "learning_rate": 9.298948985806108e-05, + "loss": 0.1694, + "step": 22287 + }, + { + "epoch": 0.5224658192537375, + "grad_norm": 3.3108904361724854, + "learning_rate": 9.298214184287229e-05, + "loss": 0.3084, + "step": 22288 + }, + { + "epoch": 0.5224892608285424, + "grad_norm": 2.5352895259857178, + "learning_rate": 9.297479386576252e-05, + "loss": 0.1573, + "step": 22289 + }, + { + "epoch": 0.5225127024033475, + "grad_norm": 2.358356475830078, + "learning_rate": 9.296744592677178e-05, + "loss": 1.6775, + "step": 22290 + }, + { + "epoch": 0.5225361439781524, + "grad_norm": 0.4668775498867035, + "learning_rate": 9.296009802593984e-05, + "loss": 0.0404, + "step": 22291 + }, + { + "epoch": 0.5225595855529575, + "grad_norm": 3.571556329727173, + "learning_rate": 9.295275016330665e-05, + "loss": 0.349, + "step": 22292 + }, + { + "epoch": 0.5225830271277624, + "grad_norm": 1.844394326210022, + "learning_rate": 9.294540233891202e-05, + "loss": 0.1816, + "step": 22293 + }, + { + "epoch": 0.5226064687025674, + "grad_norm": 3.2731692790985107, + "learning_rate": 9.293805455279582e-05, + "loss": 0.424, + "step": 22294 + }, + { + "epoch": 0.5226299102773725, + "grad_norm": 4.039392948150635, + "learning_rate": 9.293070680499795e-05, + "loss": 0.4238, + "step": 22295 + }, + { + "epoch": 0.5226533518521774, + "grad_norm": 2.7166755199432373, + "learning_rate": 9.292335909555825e-05, + "loss": 1.3127, + "step": 22296 + }, + { + "epoch": 0.5226767934269825, + "grad_norm": 3.1146953105926514, + "learning_rate": 9.291601142451661e-05, + "loss": 0.6907, + "step": 22297 + }, + { + "epoch": 0.5227002350017874, + "grad_norm": 2.722036361694336, + "learning_rate": 9.29086637919129e-05, + "loss": 0.4142, + "step": 22298 + }, + { + "epoch": 0.5227236765765925, + "grad_norm": 3.514570951461792, + "learning_rate": 9.290131619778696e-05, + "loss": 0.3096, + "step": 22299 + }, + { + "epoch": 0.5227471181513974, + "grad_norm": 3.1071295738220215, + "learning_rate": 9.28939686421787e-05, + "loss": 0.2304, + "step": 22300 + }, + { + "epoch": 0.5227705597262025, + "grad_norm": 3.773607015609741, + "learning_rate": 9.288662112512797e-05, + "loss": 0.2637, + "step": 22301 + }, + { + "epoch": 0.5227940013010074, + "grad_norm": 4.546671390533447, + "learning_rate": 9.287927364667458e-05, + "loss": 0.3358, + "step": 22302 + }, + { + "epoch": 0.5228174428758124, + "grad_norm": 2.215747117996216, + "learning_rate": 9.287192620685847e-05, + "loss": 0.1433, + "step": 22303 + }, + { + "epoch": 0.5228408844506174, + "grad_norm": 1.2058165073394775, + "learning_rate": 9.286457880571952e-05, + "loss": 0.0746, + "step": 22304 + }, + { + "epoch": 0.5228643260254224, + "grad_norm": 4.230032920837402, + "learning_rate": 9.285723144329756e-05, + "loss": 0.5017, + "step": 22305 + }, + { + "epoch": 0.5228877676002274, + "grad_norm": 3.8772125244140625, + "learning_rate": 9.284988411963245e-05, + "loss": 0.4978, + "step": 22306 + }, + { + "epoch": 0.5229112091750324, + "grad_norm": 2.498713731765747, + "learning_rate": 9.284253683476409e-05, + "loss": 0.2232, + "step": 22307 + }, + { + "epoch": 0.5229346507498374, + "grad_norm": 0.8502787351608276, + "learning_rate": 9.283518958873232e-05, + "loss": 0.0726, + "step": 22308 + }, + { + "epoch": 0.5229580923246424, + "grad_norm": 1.844191074371338, + "learning_rate": 9.282784238157701e-05, + "loss": 0.2339, + "step": 22309 + }, + { + "epoch": 0.5229815338994473, + "grad_norm": 3.962498664855957, + "learning_rate": 9.282049521333802e-05, + "loss": 0.3032, + "step": 22310 + }, + { + "epoch": 0.5230049754742524, + "grad_norm": 4.179256916046143, + "learning_rate": 9.281314808405524e-05, + "loss": 0.4391, + "step": 22311 + }, + { + "epoch": 0.5230284170490573, + "grad_norm": 1.0636215209960938, + "learning_rate": 9.28058009937685e-05, + "loss": 0.1516, + "step": 22312 + }, + { + "epoch": 0.5230518586238624, + "grad_norm": 2.436720848083496, + "learning_rate": 9.27984539425177e-05, + "loss": 0.2244, + "step": 22313 + }, + { + "epoch": 0.5230753001986673, + "grad_norm": 1.3409059047698975, + "learning_rate": 9.279110693034266e-05, + "loss": 0.2002, + "step": 22314 + }, + { + "epoch": 0.5230987417734724, + "grad_norm": 0.6653322577476501, + "learning_rate": 9.278375995728331e-05, + "loss": 0.0535, + "step": 22315 + }, + { + "epoch": 0.5231221833482773, + "grad_norm": 0.8935208320617676, + "learning_rate": 9.277641302337948e-05, + "loss": 0.0748, + "step": 22316 + }, + { + "epoch": 0.5231456249230824, + "grad_norm": 3.7785167694091797, + "learning_rate": 9.276906612867103e-05, + "loss": 0.4521, + "step": 22317 + }, + { + "epoch": 0.5231690664978873, + "grad_norm": 3.291447401046753, + "learning_rate": 9.276171927319785e-05, + "loss": 0.4075, + "step": 22318 + }, + { + "epoch": 0.5231925080726924, + "grad_norm": 3.8151166439056396, + "learning_rate": 9.275437245699974e-05, + "loss": 0.5666, + "step": 22319 + }, + { + "epoch": 0.5232159496474973, + "grad_norm": 1.9615064859390259, + "learning_rate": 9.274702568011665e-05, + "loss": 0.3053, + "step": 22320 + }, + { + "epoch": 0.5232393912223023, + "grad_norm": 1.1996564865112305, + "learning_rate": 9.273967894258841e-05, + "loss": 0.1348, + "step": 22321 + }, + { + "epoch": 0.5232628327971073, + "grad_norm": 3.3501901626586914, + "learning_rate": 9.273233224445483e-05, + "loss": 0.2877, + "step": 22322 + }, + { + "epoch": 0.5232862743719123, + "grad_norm": 5.785887241363525, + "learning_rate": 9.272498558575585e-05, + "loss": 0.6484, + "step": 22323 + }, + { + "epoch": 0.5233097159467173, + "grad_norm": 2.9923818111419678, + "learning_rate": 9.271763896653131e-05, + "loss": 0.189, + "step": 22324 + }, + { + "epoch": 0.5233331575215223, + "grad_norm": 0.7725161910057068, + "learning_rate": 9.271029238682104e-05, + "loss": 0.0427, + "step": 22325 + }, + { + "epoch": 0.5233565990963273, + "grad_norm": 3.1436731815338135, + "learning_rate": 9.270294584666491e-05, + "loss": 0.2893, + "step": 22326 + }, + { + "epoch": 0.5233800406711323, + "grad_norm": 2.2120795249938965, + "learning_rate": 9.269559934610285e-05, + "loss": 0.1914, + "step": 22327 + }, + { + "epoch": 0.5234034822459372, + "grad_norm": 5.302031517028809, + "learning_rate": 9.268825288517466e-05, + "loss": 0.9737, + "step": 22328 + }, + { + "epoch": 0.5234269238207423, + "grad_norm": 1.684200644493103, + "learning_rate": 9.26809064639202e-05, + "loss": 0.1855, + "step": 22329 + }, + { + "epoch": 0.5234503653955472, + "grad_norm": 2.1493453979492188, + "learning_rate": 9.267356008237936e-05, + "loss": 0.1757, + "step": 22330 + }, + { + "epoch": 0.5234738069703523, + "grad_norm": 0.8051766753196716, + "learning_rate": 9.2666213740592e-05, + "loss": 0.073, + "step": 22331 + }, + { + "epoch": 0.5234972485451572, + "grad_norm": 4.510801792144775, + "learning_rate": 9.265886743859795e-05, + "loss": 0.2615, + "step": 22332 + }, + { + "epoch": 0.5235206901199623, + "grad_norm": 2.9164607524871826, + "learning_rate": 9.265152117643711e-05, + "loss": 0.1895, + "step": 22333 + }, + { + "epoch": 0.5235441316947672, + "grad_norm": 2.8982961177825928, + "learning_rate": 9.26441749541493e-05, + "loss": 0.4536, + "step": 22334 + }, + { + "epoch": 0.5235675732695723, + "grad_norm": 2.6903629302978516, + "learning_rate": 9.263682877177442e-05, + "loss": 0.1902, + "step": 22335 + }, + { + "epoch": 0.5235910148443772, + "grad_norm": 1.6318881511688232, + "learning_rate": 9.262948262935233e-05, + "loss": 0.1269, + "step": 22336 + }, + { + "epoch": 0.5236144564191823, + "grad_norm": 2.620295286178589, + "learning_rate": 9.262213652692285e-05, + "loss": 1.073, + "step": 22337 + }, + { + "epoch": 0.5236378979939872, + "grad_norm": 3.0881972312927246, + "learning_rate": 9.261479046452583e-05, + "loss": 0.2484, + "step": 22338 + }, + { + "epoch": 0.5236613395687922, + "grad_norm": 4.5561394691467285, + "learning_rate": 9.260744444220121e-05, + "loss": 0.3798, + "step": 22339 + }, + { + "epoch": 0.5236847811435972, + "grad_norm": 4.191370487213135, + "learning_rate": 9.260009845998881e-05, + "loss": 0.3333, + "step": 22340 + }, + { + "epoch": 0.5237082227184022, + "grad_norm": 2.3262903690338135, + "learning_rate": 9.259275251792845e-05, + "loss": 1.8644, + "step": 22341 + }, + { + "epoch": 0.5237316642932072, + "grad_norm": 3.2084357738494873, + "learning_rate": 9.258540661606006e-05, + "loss": 0.2378, + "step": 22342 + }, + { + "epoch": 0.5237551058680122, + "grad_norm": 1.0899436473846436, + "learning_rate": 9.257806075442346e-05, + "loss": 0.0908, + "step": 22343 + }, + { + "epoch": 0.5237785474428172, + "grad_norm": 3.4283406734466553, + "learning_rate": 9.257071493305848e-05, + "loss": 0.2771, + "step": 22344 + }, + { + "epoch": 0.5238019890176222, + "grad_norm": 0.7113908529281616, + "learning_rate": 9.256336915200504e-05, + "loss": 0.0623, + "step": 22345 + }, + { + "epoch": 0.5238254305924273, + "grad_norm": 3.5001888275146484, + "learning_rate": 9.255602341130297e-05, + "loss": 0.4015, + "step": 22346 + }, + { + "epoch": 0.5238488721672322, + "grad_norm": 4.8526482582092285, + "learning_rate": 9.25486777109921e-05, + "loss": 0.4684, + "step": 22347 + }, + { + "epoch": 0.5238723137420372, + "grad_norm": 1.1401697397232056, + "learning_rate": 9.254133205111233e-05, + "loss": 0.0786, + "step": 22348 + }, + { + "epoch": 0.5238957553168422, + "grad_norm": 0.9810397028923035, + "learning_rate": 9.253398643170351e-05, + "loss": 0.0896, + "step": 22349 + }, + { + "epoch": 0.5239191968916472, + "grad_norm": 1.2173004150390625, + "learning_rate": 9.252664085280545e-05, + "loss": 0.1652, + "step": 22350 + }, + { + "epoch": 0.5239426384664522, + "grad_norm": 4.411875247955322, + "learning_rate": 9.251929531445805e-05, + "loss": 0.3776, + "step": 22351 + }, + { + "epoch": 0.5239660800412572, + "grad_norm": 2.6926674842834473, + "learning_rate": 9.251194981670121e-05, + "loss": 0.3025, + "step": 22352 + }, + { + "epoch": 0.5239895216160622, + "grad_norm": 1.3244892358779907, + "learning_rate": 9.250460435957472e-05, + "loss": 0.1626, + "step": 22353 + }, + { + "epoch": 0.5240129631908672, + "grad_norm": 2.7840700149536133, + "learning_rate": 9.249725894311845e-05, + "loss": 0.3509, + "step": 22354 + }, + { + "epoch": 0.5240364047656721, + "grad_norm": 1.355897068977356, + "learning_rate": 9.248991356737228e-05, + "loss": 0.1355, + "step": 22355 + }, + { + "epoch": 0.5240598463404772, + "grad_norm": 2.912454128265381, + "learning_rate": 9.248256823237604e-05, + "loss": 0.1599, + "step": 22356 + }, + { + "epoch": 0.5240832879152821, + "grad_norm": 0.8445784449577332, + "learning_rate": 9.247522293816959e-05, + "loss": 0.101, + "step": 22357 + }, + { + "epoch": 0.5241067294900872, + "grad_norm": 3.804166793823242, + "learning_rate": 9.246787768479281e-05, + "loss": 0.403, + "step": 22358 + }, + { + "epoch": 0.5241301710648921, + "grad_norm": 2.471693515777588, + "learning_rate": 9.246053247228555e-05, + "loss": 0.164, + "step": 22359 + }, + { + "epoch": 0.5241536126396972, + "grad_norm": 1.8552829027175903, + "learning_rate": 9.24531873006876e-05, + "loss": 1.3242, + "step": 22360 + }, + { + "epoch": 0.5241770542145021, + "grad_norm": 3.967027187347412, + "learning_rate": 9.24458421700389e-05, + "loss": 0.2672, + "step": 22361 + }, + { + "epoch": 0.5242004957893072, + "grad_norm": 1.688359260559082, + "learning_rate": 9.243849708037924e-05, + "loss": 0.1159, + "step": 22362 + }, + { + "epoch": 0.5242239373641121, + "grad_norm": 4.565077781677246, + "learning_rate": 9.243115203174856e-05, + "loss": 0.1196, + "step": 22363 + }, + { + "epoch": 0.5242473789389172, + "grad_norm": 2.323312282562256, + "learning_rate": 9.242380702418662e-05, + "loss": 0.3057, + "step": 22364 + }, + { + "epoch": 0.5242708205137221, + "grad_norm": 3.169506788253784, + "learning_rate": 9.241646205773336e-05, + "loss": 0.3524, + "step": 22365 + }, + { + "epoch": 0.5242942620885271, + "grad_norm": 2.5057482719421387, + "learning_rate": 9.240911713242857e-05, + "loss": 0.1906, + "step": 22366 + }, + { + "epoch": 0.5243177036633321, + "grad_norm": 2.8805456161499023, + "learning_rate": 9.24017722483121e-05, + "loss": 0.4421, + "step": 22367 + }, + { + "epoch": 0.5243411452381371, + "grad_norm": 3.0124635696411133, + "learning_rate": 9.239442740542386e-05, + "loss": 0.2033, + "step": 22368 + }, + { + "epoch": 0.5243645868129421, + "grad_norm": 3.4743874073028564, + "learning_rate": 9.238708260380365e-05, + "loss": 0.3105, + "step": 22369 + }, + { + "epoch": 0.5243880283877471, + "grad_norm": 3.10326886177063, + "learning_rate": 9.237973784349136e-05, + "loss": 0.1621, + "step": 22370 + }, + { + "epoch": 0.5244114699625521, + "grad_norm": 3.4330320358276367, + "learning_rate": 9.237239312452682e-05, + "loss": 0.3229, + "step": 22371 + }, + { + "epoch": 0.5244349115373571, + "grad_norm": 5.035243988037109, + "learning_rate": 9.236504844694989e-05, + "loss": 0.4494, + "step": 22372 + }, + { + "epoch": 0.524458353112162, + "grad_norm": 3.557230234146118, + "learning_rate": 9.235770381080042e-05, + "loss": 2.0988, + "step": 22373 + }, + { + "epoch": 0.5244817946869671, + "grad_norm": 4.183475971221924, + "learning_rate": 9.235035921611823e-05, + "loss": 0.3762, + "step": 22374 + }, + { + "epoch": 0.524505236261772, + "grad_norm": 0.8057581782341003, + "learning_rate": 9.234301466294326e-05, + "loss": 0.0705, + "step": 22375 + }, + { + "epoch": 0.5245286778365771, + "grad_norm": 0.3091266453266144, + "learning_rate": 9.233567015131526e-05, + "loss": 0.033, + "step": 22376 + }, + { + "epoch": 0.524552119411382, + "grad_norm": 1.9523524045944214, + "learning_rate": 9.232832568127417e-05, + "loss": 0.128, + "step": 22377 + }, + { + "epoch": 0.5245755609861871, + "grad_norm": 4.539919853210449, + "learning_rate": 9.232098125285979e-05, + "loss": 0.3426, + "step": 22378 + }, + { + "epoch": 0.524599002560992, + "grad_norm": 1.1250653266906738, + "learning_rate": 9.231363686611198e-05, + "loss": 0.1519, + "step": 22379 + }, + { + "epoch": 0.5246224441357971, + "grad_norm": 1.4738802909851074, + "learning_rate": 9.230629252107059e-05, + "loss": 0.1599, + "step": 22380 + }, + { + "epoch": 0.524645885710602, + "grad_norm": 3.2591540813446045, + "learning_rate": 9.229894821777549e-05, + "loss": 0.5049, + "step": 22381 + }, + { + "epoch": 0.524669327285407, + "grad_norm": 5.143949508666992, + "learning_rate": 9.229160395626648e-05, + "loss": 0.3933, + "step": 22382 + }, + { + "epoch": 0.524692768860212, + "grad_norm": 4.033670902252197, + "learning_rate": 9.228425973658348e-05, + "loss": 0.3437, + "step": 22383 + }, + { + "epoch": 0.524716210435017, + "grad_norm": 2.037132501602173, + "learning_rate": 9.22769155587663e-05, + "loss": 0.1164, + "step": 22384 + }, + { + "epoch": 0.524739652009822, + "grad_norm": 2.077432632446289, + "learning_rate": 9.226957142285476e-05, + "loss": 0.1654, + "step": 22385 + }, + { + "epoch": 0.524763093584627, + "grad_norm": 2.5992205142974854, + "learning_rate": 9.226222732888878e-05, + "loss": 0.2439, + "step": 22386 + }, + { + "epoch": 0.524786535159432, + "grad_norm": 1.3022048473358154, + "learning_rate": 9.22548832769081e-05, + "loss": 0.1346, + "step": 22387 + }, + { + "epoch": 0.524809976734237, + "grad_norm": 3.541630268096924, + "learning_rate": 9.224753926695271e-05, + "loss": 0.408, + "step": 22388 + }, + { + "epoch": 0.524833418309042, + "grad_norm": 1.790791392326355, + "learning_rate": 9.224019529906236e-05, + "loss": 0.1422, + "step": 22389 + }, + { + "epoch": 0.524856859883847, + "grad_norm": 2.5813634395599365, + "learning_rate": 9.223285137327695e-05, + "loss": 0.2543, + "step": 22390 + }, + { + "epoch": 0.5248803014586519, + "grad_norm": 3.535861015319824, + "learning_rate": 9.222550748963633e-05, + "loss": 0.4001, + "step": 22391 + }, + { + "epoch": 0.524903743033457, + "grad_norm": 0.8376826643943787, + "learning_rate": 9.221816364818027e-05, + "loss": 0.0659, + "step": 22392 + }, + { + "epoch": 0.5249271846082619, + "grad_norm": 1.4698596000671387, + "learning_rate": 9.22108198489487e-05, + "loss": 0.081, + "step": 22393 + }, + { + "epoch": 0.524950626183067, + "grad_norm": 3.666069746017456, + "learning_rate": 9.220347609198146e-05, + "loss": 0.192, + "step": 22394 + }, + { + "epoch": 0.5249740677578719, + "grad_norm": 5.245151042938232, + "learning_rate": 9.219613237731833e-05, + "loss": 0.3798, + "step": 22395 + }, + { + "epoch": 0.524997509332677, + "grad_norm": 0.42001137137413025, + "learning_rate": 9.218878870499925e-05, + "loss": 0.0378, + "step": 22396 + }, + { + "epoch": 0.5250209509074819, + "grad_norm": 5.103188991546631, + "learning_rate": 9.2181445075064e-05, + "loss": 0.5302, + "step": 22397 + }, + { + "epoch": 0.525044392482287, + "grad_norm": 5.931938171386719, + "learning_rate": 9.217410148755245e-05, + "loss": 0.5287, + "step": 22398 + }, + { + "epoch": 0.525067834057092, + "grad_norm": 2.2919347286224365, + "learning_rate": 9.216675794250441e-05, + "loss": 0.1557, + "step": 22399 + }, + { + "epoch": 0.525091275631897, + "grad_norm": 4.574503421783447, + "learning_rate": 9.215941443995982e-05, + "loss": 0.5166, + "step": 22400 + }, + { + "epoch": 0.525114717206702, + "grad_norm": 3.1286332607269287, + "learning_rate": 9.215207097995845e-05, + "loss": 0.35, + "step": 22401 + }, + { + "epoch": 0.5251381587815069, + "grad_norm": 1.7583235502243042, + "learning_rate": 9.214472756254014e-05, + "loss": 0.1072, + "step": 22402 + }, + { + "epoch": 0.525161600356312, + "grad_norm": 2.5747478008270264, + "learning_rate": 9.21373841877448e-05, + "loss": 0.2367, + "step": 22403 + }, + { + "epoch": 0.5251850419311169, + "grad_norm": 4.476462364196777, + "learning_rate": 9.213004085561222e-05, + "loss": 0.4828, + "step": 22404 + }, + { + "epoch": 0.525208483505922, + "grad_norm": 3.4099838733673096, + "learning_rate": 9.212269756618225e-05, + "loss": 0.5505, + "step": 22405 + }, + { + "epoch": 0.5252319250807269, + "grad_norm": 4.8367533683776855, + "learning_rate": 9.211535431949475e-05, + "loss": 0.428, + "step": 22406 + }, + { + "epoch": 0.525255366655532, + "grad_norm": 0.7370365262031555, + "learning_rate": 9.210801111558955e-05, + "loss": 0.089, + "step": 22407 + }, + { + "epoch": 0.5252788082303369, + "grad_norm": 2.793976068496704, + "learning_rate": 9.210066795450651e-05, + "loss": 0.3099, + "step": 22408 + }, + { + "epoch": 0.525302249805142, + "grad_norm": 1.4887728691101074, + "learning_rate": 9.209332483628549e-05, + "loss": 0.1522, + "step": 22409 + }, + { + "epoch": 0.5253256913799469, + "grad_norm": 2.4529025554656982, + "learning_rate": 9.208598176096628e-05, + "loss": 0.2146, + "step": 22410 + }, + { + "epoch": 0.5253491329547519, + "grad_norm": 2.645897388458252, + "learning_rate": 9.207863872858875e-05, + "loss": 0.2, + "step": 22411 + }, + { + "epoch": 0.5253725745295569, + "grad_norm": 1.0828758478164673, + "learning_rate": 9.207129573919276e-05, + "loss": 0.0894, + "step": 22412 + }, + { + "epoch": 0.5253960161043619, + "grad_norm": 4.621785640716553, + "learning_rate": 9.206395279281814e-05, + "loss": 0.4847, + "step": 22413 + }, + { + "epoch": 0.5254194576791669, + "grad_norm": 7.056242942810059, + "learning_rate": 9.205660988950474e-05, + "loss": 0.5647, + "step": 22414 + }, + { + "epoch": 0.5254428992539719, + "grad_norm": 3.953181505203247, + "learning_rate": 9.204926702929241e-05, + "loss": 0.4635, + "step": 22415 + }, + { + "epoch": 0.5254663408287769, + "grad_norm": 2.924098491668701, + "learning_rate": 9.204192421222098e-05, + "loss": 0.2771, + "step": 22416 + }, + { + "epoch": 0.5254897824035819, + "grad_norm": 1.1906079053878784, + "learning_rate": 9.203458143833027e-05, + "loss": 0.1283, + "step": 22417 + }, + { + "epoch": 0.5255132239783868, + "grad_norm": 1.1209114789962769, + "learning_rate": 9.202723870766017e-05, + "loss": 0.0724, + "step": 22418 + }, + { + "epoch": 0.5255366655531919, + "grad_norm": 1.8252407312393188, + "learning_rate": 9.20198960202505e-05, + "loss": 0.1453, + "step": 22419 + }, + { + "epoch": 0.5255601071279968, + "grad_norm": 2.1812686920166016, + "learning_rate": 9.201255337614107e-05, + "loss": 1.3494, + "step": 22420 + }, + { + "epoch": 0.5255835487028019, + "grad_norm": 0.7187263369560242, + "learning_rate": 9.200521077537178e-05, + "loss": 0.0528, + "step": 22421 + }, + { + "epoch": 0.5256069902776068, + "grad_norm": 2.8493971824645996, + "learning_rate": 9.199786821798243e-05, + "loss": 0.3763, + "step": 22422 + }, + { + "epoch": 0.5256304318524119, + "grad_norm": 1.9583323001861572, + "learning_rate": 9.199052570401282e-05, + "loss": 0.217, + "step": 22423 + }, + { + "epoch": 0.5256538734272168, + "grad_norm": 6.477059841156006, + "learning_rate": 9.198318323350289e-05, + "loss": 0.8766, + "step": 22424 + }, + { + "epoch": 0.5256773150020219, + "grad_norm": 1.4734290838241577, + "learning_rate": 9.197584080649244e-05, + "loss": 0.0747, + "step": 22425 + }, + { + "epoch": 0.5257007565768268, + "grad_norm": 3.9067633152008057, + "learning_rate": 9.19684984230213e-05, + "loss": 0.3567, + "step": 22426 + }, + { + "epoch": 0.5257241981516318, + "grad_norm": 1.8698930740356445, + "learning_rate": 9.196115608312931e-05, + "loss": 0.7195, + "step": 22427 + }, + { + "epoch": 0.5257476397264368, + "grad_norm": 1.0163888931274414, + "learning_rate": 9.195381378685632e-05, + "loss": 0.1055, + "step": 22428 + }, + { + "epoch": 0.5257710813012418, + "grad_norm": 2.7592618465423584, + "learning_rate": 9.194647153424217e-05, + "loss": 0.3619, + "step": 22429 + }, + { + "epoch": 0.5257945228760468, + "grad_norm": 3.3568203449249268, + "learning_rate": 9.193912932532667e-05, + "loss": 0.3881, + "step": 22430 + }, + { + "epoch": 0.5258179644508518, + "grad_norm": 0.8219413757324219, + "learning_rate": 9.19317871601497e-05, + "loss": 0.0708, + "step": 22431 + }, + { + "epoch": 0.5258414060256568, + "grad_norm": 1.9869115352630615, + "learning_rate": 9.192444503875107e-05, + "loss": 1.411, + "step": 22432 + }, + { + "epoch": 0.5258648476004618, + "grad_norm": 2.4332478046417236, + "learning_rate": 9.191710296117062e-05, + "loss": 0.1475, + "step": 22433 + }, + { + "epoch": 0.5258882891752668, + "grad_norm": 2.6877031326293945, + "learning_rate": 9.190976092744821e-05, + "loss": 0.1402, + "step": 22434 + }, + { + "epoch": 0.5259117307500718, + "grad_norm": 4.926708221435547, + "learning_rate": 9.190241893762362e-05, + "loss": 0.467, + "step": 22435 + }, + { + "epoch": 0.5259351723248767, + "grad_norm": 3.22326397895813, + "learning_rate": 9.189507699173678e-05, + "loss": 0.1792, + "step": 22436 + }, + { + "epoch": 0.5259586138996818, + "grad_norm": 3.0661404132843018, + "learning_rate": 9.188773508982746e-05, + "loss": 0.2575, + "step": 22437 + }, + { + "epoch": 0.5259820554744867, + "grad_norm": 0.8211102485656738, + "learning_rate": 9.188039323193555e-05, + "loss": 0.0945, + "step": 22438 + }, + { + "epoch": 0.5260054970492918, + "grad_norm": 3.175043821334839, + "learning_rate": 9.187305141810084e-05, + "loss": 0.3157, + "step": 22439 + }, + { + "epoch": 0.5260289386240967, + "grad_norm": 0.4841384291648865, + "learning_rate": 9.186570964836316e-05, + "loss": 0.0491, + "step": 22440 + }, + { + "epoch": 0.5260523801989018, + "grad_norm": 2.1293365955352783, + "learning_rate": 9.185836792276239e-05, + "loss": 0.2415, + "step": 22441 + }, + { + "epoch": 0.5260758217737067, + "grad_norm": 3.84795880317688, + "learning_rate": 9.185102624133835e-05, + "loss": 0.3943, + "step": 22442 + }, + { + "epoch": 0.5260992633485118, + "grad_norm": 0.9705818295478821, + "learning_rate": 9.184368460413084e-05, + "loss": 0.0631, + "step": 22443 + }, + { + "epoch": 0.5261227049233167, + "grad_norm": 2.6123592853546143, + "learning_rate": 9.183634301117976e-05, + "loss": 1.5223, + "step": 22444 + }, + { + "epoch": 0.5261461464981217, + "grad_norm": 3.3597443103790283, + "learning_rate": 9.182900146252489e-05, + "loss": 0.3698, + "step": 22445 + }, + { + "epoch": 0.5261695880729267, + "grad_norm": 1.4783895015716553, + "learning_rate": 9.18216599582061e-05, + "loss": 0.1251, + "step": 22446 + }, + { + "epoch": 0.5261930296477317, + "grad_norm": 3.8552467823028564, + "learning_rate": 9.181431849826318e-05, + "loss": 0.2284, + "step": 22447 + }, + { + "epoch": 0.5262164712225367, + "grad_norm": 1.419105887413025, + "learning_rate": 9.180697708273603e-05, + "loss": 0.2429, + "step": 22448 + }, + { + "epoch": 0.5262399127973417, + "grad_norm": 2.236719846725464, + "learning_rate": 9.179963571166445e-05, + "loss": 0.225, + "step": 22449 + }, + { + "epoch": 0.5262633543721468, + "grad_norm": 2.714454174041748, + "learning_rate": 9.179229438508828e-05, + "loss": 0.252, + "step": 22450 + }, + { + "epoch": 0.5262867959469517, + "grad_norm": 4.819522857666016, + "learning_rate": 9.178495310304736e-05, + "loss": 0.5056, + "step": 22451 + }, + { + "epoch": 0.5263102375217568, + "grad_norm": 2.0840847492218018, + "learning_rate": 9.177761186558148e-05, + "loss": 0.1854, + "step": 22452 + }, + { + "epoch": 0.5263336790965617, + "grad_norm": 1.5519757270812988, + "learning_rate": 9.177027067273054e-05, + "loss": 0.1319, + "step": 22453 + }, + { + "epoch": 0.5263571206713668, + "grad_norm": 3.609435796737671, + "learning_rate": 9.176292952453435e-05, + "loss": 0.5394, + "step": 22454 + }, + { + "epoch": 0.5263805622461717, + "grad_norm": 1.8175562620162964, + "learning_rate": 9.175558842103272e-05, + "loss": 0.096, + "step": 22455 + }, + { + "epoch": 0.5264040038209767, + "grad_norm": 0.5400462746620178, + "learning_rate": 9.17482473622655e-05, + "loss": 0.0408, + "step": 22456 + }, + { + "epoch": 0.5264274453957817, + "grad_norm": 0.62742680311203, + "learning_rate": 9.174090634827253e-05, + "loss": 0.0542, + "step": 22457 + }, + { + "epoch": 0.5264508869705867, + "grad_norm": 2.5065581798553467, + "learning_rate": 9.173356537909361e-05, + "loss": 0.2656, + "step": 22458 + }, + { + "epoch": 0.5264743285453917, + "grad_norm": 2.649567127227783, + "learning_rate": 9.172622445476862e-05, + "loss": 0.3123, + "step": 22459 + }, + { + "epoch": 0.5264977701201967, + "grad_norm": 0.6439147591590881, + "learning_rate": 9.171888357533733e-05, + "loss": 0.0789, + "step": 22460 + }, + { + "epoch": 0.5265212116950017, + "grad_norm": 3.153409242630005, + "learning_rate": 9.171154274083965e-05, + "loss": 0.3404, + "step": 22461 + }, + { + "epoch": 0.5265446532698067, + "grad_norm": 2.353644371032715, + "learning_rate": 9.170420195131535e-05, + "loss": 0.294, + "step": 22462 + }, + { + "epoch": 0.5265680948446116, + "grad_norm": 0.6809727549552917, + "learning_rate": 9.169686120680431e-05, + "loss": 0.0274, + "step": 22463 + }, + { + "epoch": 0.5265915364194167, + "grad_norm": 3.130953788757324, + "learning_rate": 9.168952050734634e-05, + "loss": 0.3571, + "step": 22464 + }, + { + "epoch": 0.5266149779942216, + "grad_norm": 2.2793455123901367, + "learning_rate": 9.168217985298123e-05, + "loss": 0.1005, + "step": 22465 + }, + { + "epoch": 0.5266384195690267, + "grad_norm": 3.879091739654541, + "learning_rate": 9.167483924374888e-05, + "loss": 0.4117, + "step": 22466 + }, + { + "epoch": 0.5266618611438316, + "grad_norm": 3.635999917984009, + "learning_rate": 9.166749867968908e-05, + "loss": 0.3131, + "step": 22467 + }, + { + "epoch": 0.5266853027186367, + "grad_norm": 2.8892617225646973, + "learning_rate": 9.166015816084164e-05, + "loss": 0.1819, + "step": 22468 + }, + { + "epoch": 0.5267087442934416, + "grad_norm": 2.657465696334839, + "learning_rate": 9.165281768724644e-05, + "loss": 1.5673, + "step": 22469 + }, + { + "epoch": 0.5267321858682467, + "grad_norm": 4.628087043762207, + "learning_rate": 9.16454772589433e-05, + "loss": 0.2526, + "step": 22470 + }, + { + "epoch": 0.5267556274430516, + "grad_norm": 3.3835296630859375, + "learning_rate": 9.163813687597199e-05, + "loss": 0.3666, + "step": 22471 + }, + { + "epoch": 0.5267790690178566, + "grad_norm": 2.778681755065918, + "learning_rate": 9.163079653837237e-05, + "loss": 0.3792, + "step": 22472 + }, + { + "epoch": 0.5268025105926616, + "grad_norm": 3.2150745391845703, + "learning_rate": 9.162345624618433e-05, + "loss": 0.1744, + "step": 22473 + }, + { + "epoch": 0.5268259521674666, + "grad_norm": 3.8149852752685547, + "learning_rate": 9.161611599944766e-05, + "loss": 0.3208, + "step": 22474 + }, + { + "epoch": 0.5268493937422716, + "grad_norm": 7.153125286102295, + "learning_rate": 9.160877579820215e-05, + "loss": 0.5643, + "step": 22475 + }, + { + "epoch": 0.5268728353170766, + "grad_norm": 4.747644901275635, + "learning_rate": 9.16014356424877e-05, + "loss": 0.5718, + "step": 22476 + }, + { + "epoch": 0.5268962768918816, + "grad_norm": 3.464384078979492, + "learning_rate": 9.159409553234406e-05, + "loss": 0.3777, + "step": 22477 + }, + { + "epoch": 0.5269197184666866, + "grad_norm": 1.3064136505126953, + "learning_rate": 9.15867554678111e-05, + "loss": 0.1403, + "step": 22478 + }, + { + "epoch": 0.5269431600414916, + "grad_norm": 0.8504205942153931, + "learning_rate": 9.157941544892865e-05, + "loss": 0.0324, + "step": 22479 + }, + { + "epoch": 0.5269666016162966, + "grad_norm": 3.4823591709136963, + "learning_rate": 9.157207547573652e-05, + "loss": 1.7221, + "step": 22480 + }, + { + "epoch": 0.5269900431911015, + "grad_norm": 2.268378734588623, + "learning_rate": 9.156473554827454e-05, + "loss": 0.2392, + "step": 22481 + }, + { + "epoch": 0.5270134847659066, + "grad_norm": 4.558962821960449, + "learning_rate": 9.155739566658256e-05, + "loss": 0.3937, + "step": 22482 + }, + { + "epoch": 0.5270369263407115, + "grad_norm": 2.5591444969177246, + "learning_rate": 9.155005583070038e-05, + "loss": 1.4313, + "step": 22483 + }, + { + "epoch": 0.5270603679155166, + "grad_norm": 0.6257538199424744, + "learning_rate": 9.154271604066782e-05, + "loss": 0.0549, + "step": 22484 + }, + { + "epoch": 0.5270838094903215, + "grad_norm": 3.6328394412994385, + "learning_rate": 9.153537629652471e-05, + "loss": 0.4437, + "step": 22485 + }, + { + "epoch": 0.5271072510651266, + "grad_norm": 3.185213565826416, + "learning_rate": 9.152803659831092e-05, + "loss": 0.3981, + "step": 22486 + }, + { + "epoch": 0.5271306926399315, + "grad_norm": 1.627582311630249, + "learning_rate": 9.152069694606623e-05, + "loss": 0.1077, + "step": 22487 + }, + { + "epoch": 0.5271541342147366, + "grad_norm": 1.7074377536773682, + "learning_rate": 9.151335733983049e-05, + "loss": 0.2202, + "step": 22488 + }, + { + "epoch": 0.5271775757895415, + "grad_norm": 3.53047513961792, + "learning_rate": 9.150601777964351e-05, + "loss": 0.2993, + "step": 22489 + }, + { + "epoch": 0.5272010173643465, + "grad_norm": 5.786437511444092, + "learning_rate": 9.149867826554508e-05, + "loss": 0.5482, + "step": 22490 + }, + { + "epoch": 0.5272244589391515, + "grad_norm": 2.9226977825164795, + "learning_rate": 9.149133879757511e-05, + "loss": 0.1854, + "step": 22491 + }, + { + "epoch": 0.5272479005139565, + "grad_norm": 4.847297191619873, + "learning_rate": 9.148399937577336e-05, + "loss": 0.3695, + "step": 22492 + }, + { + "epoch": 0.5272713420887615, + "grad_norm": 1.9772865772247314, + "learning_rate": 9.147666000017965e-05, + "loss": 0.2207, + "step": 22493 + }, + { + "epoch": 0.5272947836635665, + "grad_norm": 3.4604735374450684, + "learning_rate": 9.146932067083384e-05, + "loss": 0.1914, + "step": 22494 + }, + { + "epoch": 0.5273182252383715, + "grad_norm": 3.344599485397339, + "learning_rate": 9.146198138777574e-05, + "loss": 0.3032, + "step": 22495 + }, + { + "epoch": 0.5273416668131765, + "grad_norm": 0.9784694910049438, + "learning_rate": 9.145464215104512e-05, + "loss": 0.0742, + "step": 22496 + }, + { + "epoch": 0.5273651083879815, + "grad_norm": 1.5754268169403076, + "learning_rate": 9.144730296068189e-05, + "loss": 0.2067, + "step": 22497 + }, + { + "epoch": 0.5273885499627865, + "grad_norm": 0.6922513246536255, + "learning_rate": 9.143996381672584e-05, + "loss": 0.0325, + "step": 22498 + }, + { + "epoch": 0.5274119915375914, + "grad_norm": 2.070862054824829, + "learning_rate": 9.143262471921679e-05, + "loss": 0.1463, + "step": 22499 + }, + { + "epoch": 0.5274354331123965, + "grad_norm": 8.013124465942383, + "learning_rate": 9.142528566819454e-05, + "loss": 0.4729, + "step": 22500 + }, + { + "epoch": 0.5274588746872015, + "grad_norm": 6.395820140838623, + "learning_rate": 9.141794666369895e-05, + "loss": 0.2953, + "step": 22501 + }, + { + "epoch": 0.5274823162620065, + "grad_norm": 2.375110626220703, + "learning_rate": 9.141060770576983e-05, + "loss": 1.7888, + "step": 22502 + }, + { + "epoch": 0.5275057578368115, + "grad_norm": 2.2364349365234375, + "learning_rate": 9.140326879444697e-05, + "loss": 0.2342, + "step": 22503 + }, + { + "epoch": 0.5275291994116165, + "grad_norm": 3.744532346725464, + "learning_rate": 9.139592992977023e-05, + "loss": 0.2291, + "step": 22504 + }, + { + "epoch": 0.5275526409864215, + "grad_norm": 0.9641980528831482, + "learning_rate": 9.138859111177942e-05, + "loss": 0.0928, + "step": 22505 + }, + { + "epoch": 0.5275760825612265, + "grad_norm": 4.008561611175537, + "learning_rate": 9.138125234051434e-05, + "loss": 0.3196, + "step": 22506 + }, + { + "epoch": 0.5275995241360315, + "grad_norm": 4.04383659362793, + "learning_rate": 9.137391361601484e-05, + "loss": 0.4865, + "step": 22507 + }, + { + "epoch": 0.5276229657108364, + "grad_norm": 2.893312931060791, + "learning_rate": 9.136657493832069e-05, + "loss": 0.3241, + "step": 22508 + }, + { + "epoch": 0.5276464072856415, + "grad_norm": 2.2980778217315674, + "learning_rate": 9.135923630747179e-05, + "loss": 0.2221, + "step": 22509 + }, + { + "epoch": 0.5276698488604464, + "grad_norm": 2.873027801513672, + "learning_rate": 9.135189772350789e-05, + "loss": 0.1954, + "step": 22510 + }, + { + "epoch": 0.5276932904352515, + "grad_norm": 4.161048412322998, + "learning_rate": 9.134455918646886e-05, + "loss": 0.2299, + "step": 22511 + }, + { + "epoch": 0.5277167320100564, + "grad_norm": 1.5177698135375977, + "learning_rate": 9.133722069639449e-05, + "loss": 0.1642, + "step": 22512 + }, + { + "epoch": 0.5277401735848615, + "grad_norm": 3.245863199234009, + "learning_rate": 9.132988225332459e-05, + "loss": 0.2376, + "step": 22513 + }, + { + "epoch": 0.5277636151596664, + "grad_norm": 3.099159002304077, + "learning_rate": 9.132254385729902e-05, + "loss": 0.3405, + "step": 22514 + }, + { + "epoch": 0.5277870567344715, + "grad_norm": 4.11952543258667, + "learning_rate": 9.131520550835755e-05, + "loss": 0.3834, + "step": 22515 + }, + { + "epoch": 0.5278104983092764, + "grad_norm": 4.435582160949707, + "learning_rate": 9.130786720654003e-05, + "loss": 0.4936, + "step": 22516 + }, + { + "epoch": 0.5278339398840814, + "grad_norm": 1.8662047386169434, + "learning_rate": 9.130052895188626e-05, + "loss": 1.3519, + "step": 22517 + }, + { + "epoch": 0.5278573814588864, + "grad_norm": 2.2149484157562256, + "learning_rate": 9.129319074443608e-05, + "loss": 0.1468, + "step": 22518 + }, + { + "epoch": 0.5278808230336914, + "grad_norm": 2.4676401615142822, + "learning_rate": 9.128585258422927e-05, + "loss": 0.2467, + "step": 22519 + }, + { + "epoch": 0.5279042646084964, + "grad_norm": 0.8804575204849243, + "learning_rate": 9.127851447130565e-05, + "loss": 0.1399, + "step": 22520 + }, + { + "epoch": 0.5279277061833014, + "grad_norm": 1.7786914110183716, + "learning_rate": 9.127117640570508e-05, + "loss": 0.9595, + "step": 22521 + }, + { + "epoch": 0.5279511477581064, + "grad_norm": 2.398885726928711, + "learning_rate": 9.126383838746736e-05, + "loss": 0.3094, + "step": 22522 + }, + { + "epoch": 0.5279745893329114, + "grad_norm": 3.540078639984131, + "learning_rate": 9.125650041663228e-05, + "loss": 0.1848, + "step": 22523 + }, + { + "epoch": 0.5279980309077164, + "grad_norm": 3.5181241035461426, + "learning_rate": 9.124916249323971e-05, + "loss": 0.2155, + "step": 22524 + }, + { + "epoch": 0.5280214724825214, + "grad_norm": 6.718722343444824, + "learning_rate": 9.124182461732938e-05, + "loss": 0.694, + "step": 22525 + }, + { + "epoch": 0.5280449140573263, + "grad_norm": 1.9099910259246826, + "learning_rate": 9.12344867889412e-05, + "loss": 0.2574, + "step": 22526 + }, + { + "epoch": 0.5280683556321314, + "grad_norm": 3.164940357208252, + "learning_rate": 9.122714900811493e-05, + "loss": 0.2517, + "step": 22527 + }, + { + "epoch": 0.5280917972069363, + "grad_norm": 4.05443811416626, + "learning_rate": 9.121981127489039e-05, + "loss": 0.3934, + "step": 22528 + }, + { + "epoch": 0.5281152387817414, + "grad_norm": 2.5787160396575928, + "learning_rate": 9.12124735893074e-05, + "loss": 0.2269, + "step": 22529 + }, + { + "epoch": 0.5281386803565463, + "grad_norm": 1.191441297531128, + "learning_rate": 9.120513595140579e-05, + "loss": 0.0765, + "step": 22530 + }, + { + "epoch": 0.5281621219313514, + "grad_norm": 0.5113551616668701, + "learning_rate": 9.119779836122534e-05, + "loss": 0.0648, + "step": 22531 + }, + { + "epoch": 0.5281855635061563, + "grad_norm": 3.5898563861846924, + "learning_rate": 9.11904608188059e-05, + "loss": 0.3692, + "step": 22532 + }, + { + "epoch": 0.5282090050809614, + "grad_norm": 2.0703423023223877, + "learning_rate": 9.118312332418722e-05, + "loss": 0.1753, + "step": 22533 + }, + { + "epoch": 0.5282324466557663, + "grad_norm": 2.3882758617401123, + "learning_rate": 9.11757858774092e-05, + "loss": 0.2375, + "step": 22534 + }, + { + "epoch": 0.5282558882305713, + "grad_norm": 5.265703201293945, + "learning_rate": 9.11684484785116e-05, + "loss": 0.4781, + "step": 22535 + }, + { + "epoch": 0.5282793298053763, + "grad_norm": 3.3237571716308594, + "learning_rate": 9.116111112753428e-05, + "loss": 0.2722, + "step": 22536 + }, + { + "epoch": 0.5283027713801813, + "grad_norm": 1.0753812789916992, + "learning_rate": 9.115377382451701e-05, + "loss": 0.0932, + "step": 22537 + }, + { + "epoch": 0.5283262129549863, + "grad_norm": 2.9743900299072266, + "learning_rate": 9.114643656949958e-05, + "loss": 0.3996, + "step": 22538 + }, + { + "epoch": 0.5283496545297913, + "grad_norm": 1.676896333694458, + "learning_rate": 9.113909936252187e-05, + "loss": 0.094, + "step": 22539 + }, + { + "epoch": 0.5283730961045963, + "grad_norm": 1.6065870523452759, + "learning_rate": 9.113176220362365e-05, + "loss": 0.191, + "step": 22540 + }, + { + "epoch": 0.5283965376794013, + "grad_norm": 2.504866600036621, + "learning_rate": 9.112442509284472e-05, + "loss": 0.2829, + "step": 22541 + }, + { + "epoch": 0.5284199792542063, + "grad_norm": 2.5712897777557373, + "learning_rate": 9.111708803022492e-05, + "loss": 0.2708, + "step": 22542 + }, + { + "epoch": 0.5284434208290113, + "grad_norm": 1.7587403059005737, + "learning_rate": 9.110975101580405e-05, + "loss": 0.1525, + "step": 22543 + }, + { + "epoch": 0.5284668624038162, + "grad_norm": 3.094304323196411, + "learning_rate": 9.110241404962192e-05, + "loss": 0.1939, + "step": 22544 + }, + { + "epoch": 0.5284903039786213, + "grad_norm": 3.875194549560547, + "learning_rate": 9.10950771317183e-05, + "loss": 0.3618, + "step": 22545 + }, + { + "epoch": 0.5285137455534262, + "grad_norm": 3.0967202186584473, + "learning_rate": 9.108774026213309e-05, + "loss": 0.135, + "step": 22546 + }, + { + "epoch": 0.5285371871282313, + "grad_norm": 2.588933229446411, + "learning_rate": 9.108040344090606e-05, + "loss": 0.2193, + "step": 22547 + }, + { + "epoch": 0.5285606287030362, + "grad_norm": 0.5687673091888428, + "learning_rate": 9.107306666807698e-05, + "loss": 0.0591, + "step": 22548 + }, + { + "epoch": 0.5285840702778413, + "grad_norm": 3.0557360649108887, + "learning_rate": 9.106572994368572e-05, + "loss": 0.1491, + "step": 22549 + }, + { + "epoch": 0.5286075118526462, + "grad_norm": 1.3274182081222534, + "learning_rate": 9.105839326777206e-05, + "loss": 0.1516, + "step": 22550 + }, + { + "epoch": 0.5286309534274513, + "grad_norm": 2.2009189128875732, + "learning_rate": 9.105105664037577e-05, + "loss": 0.1619, + "step": 22551 + }, + { + "epoch": 0.5286543950022563, + "grad_norm": 4.313630104064941, + "learning_rate": 9.104372006153674e-05, + "loss": 0.3578, + "step": 22552 + }, + { + "epoch": 0.5286778365770612, + "grad_norm": 3.2008748054504395, + "learning_rate": 9.103638353129474e-05, + "loss": 0.5129, + "step": 22553 + }, + { + "epoch": 0.5287012781518663, + "grad_norm": 1.3244277238845825, + "learning_rate": 9.102904704968954e-05, + "loss": 0.1647, + "step": 22554 + }, + { + "epoch": 0.5287247197266712, + "grad_norm": 1.7127795219421387, + "learning_rate": 9.102171061676101e-05, + "loss": 0.1719, + "step": 22555 + }, + { + "epoch": 0.5287481613014763, + "grad_norm": 3.8488192558288574, + "learning_rate": 9.10143742325489e-05, + "loss": 0.2313, + "step": 22556 + }, + { + "epoch": 0.5287716028762812, + "grad_norm": 4.267795562744141, + "learning_rate": 9.100703789709306e-05, + "loss": 0.3185, + "step": 22557 + }, + { + "epoch": 0.5287950444510863, + "grad_norm": 2.95861554145813, + "learning_rate": 9.099970161043329e-05, + "loss": 0.2854, + "step": 22558 + }, + { + "epoch": 0.5288184860258912, + "grad_norm": 1.8915876150131226, + "learning_rate": 9.09923653726094e-05, + "loss": 0.159, + "step": 22559 + }, + { + "epoch": 0.5288419276006963, + "grad_norm": 3.7972166538238525, + "learning_rate": 9.09850291836612e-05, + "loss": 0.3548, + "step": 22560 + }, + { + "epoch": 0.5288653691755012, + "grad_norm": 2.728668212890625, + "learning_rate": 9.097769304362846e-05, + "loss": 0.2239, + "step": 22561 + }, + { + "epoch": 0.5288888107503062, + "grad_norm": 4.758148193359375, + "learning_rate": 9.097035695255103e-05, + "loss": 0.3882, + "step": 22562 + }, + { + "epoch": 0.5289122523251112, + "grad_norm": 3.355630397796631, + "learning_rate": 9.096302091046867e-05, + "loss": 0.4253, + "step": 22563 + }, + { + "epoch": 0.5289356938999162, + "grad_norm": 1.1057944297790527, + "learning_rate": 9.095568491742125e-05, + "loss": 0.348, + "step": 22564 + }, + { + "epoch": 0.5289591354747212, + "grad_norm": 3.93652606010437, + "learning_rate": 9.094834897344851e-05, + "loss": 0.3383, + "step": 22565 + }, + { + "epoch": 0.5289825770495262, + "grad_norm": 3.55185604095459, + "learning_rate": 9.094101307859029e-05, + "loss": 0.4681, + "step": 22566 + }, + { + "epoch": 0.5290060186243312, + "grad_norm": 0.42118680477142334, + "learning_rate": 9.093367723288639e-05, + "loss": 0.0469, + "step": 22567 + }, + { + "epoch": 0.5290294601991362, + "grad_norm": 2.1827895641326904, + "learning_rate": 9.092634143637663e-05, + "loss": 0.2436, + "step": 22568 + }, + { + "epoch": 0.5290529017739412, + "grad_norm": 4.040435314178467, + "learning_rate": 9.091900568910075e-05, + "loss": 0.2215, + "step": 22569 + }, + { + "epoch": 0.5290763433487462, + "grad_norm": 4.165473461151123, + "learning_rate": 9.091166999109861e-05, + "loss": 0.7406, + "step": 22570 + }, + { + "epoch": 0.5290997849235511, + "grad_norm": 2.8433218002319336, + "learning_rate": 9.090433434241004e-05, + "loss": 0.2315, + "step": 22571 + }, + { + "epoch": 0.5291232264983562, + "grad_norm": 1.283959984779358, + "learning_rate": 9.089699874307479e-05, + "loss": 0.0454, + "step": 22572 + }, + { + "epoch": 0.5291466680731611, + "grad_norm": 3.5424177646636963, + "learning_rate": 9.088966319313268e-05, + "loss": 0.3979, + "step": 22573 + }, + { + "epoch": 0.5291701096479662, + "grad_norm": 1.0178242921829224, + "learning_rate": 9.088232769262351e-05, + "loss": 0.1146, + "step": 22574 + }, + { + "epoch": 0.5291935512227711, + "grad_norm": 2.9136972427368164, + "learning_rate": 9.087499224158711e-05, + "loss": 0.2596, + "step": 22575 + }, + { + "epoch": 0.5292169927975762, + "grad_norm": 1.161014199256897, + "learning_rate": 9.086765684006323e-05, + "loss": 0.1099, + "step": 22576 + }, + { + "epoch": 0.5292404343723811, + "grad_norm": 6.270291805267334, + "learning_rate": 9.086032148809172e-05, + "loss": 0.2221, + "step": 22577 + }, + { + "epoch": 0.5292638759471862, + "grad_norm": 1.7816098928451538, + "learning_rate": 9.085298618571236e-05, + "loss": 0.5329, + "step": 22578 + }, + { + "epoch": 0.5292873175219911, + "grad_norm": 0.9603404402732849, + "learning_rate": 9.084565093296493e-05, + "loss": 0.1058, + "step": 22579 + }, + { + "epoch": 0.5293107590967961, + "grad_norm": 2.9742088317871094, + "learning_rate": 9.08383157298893e-05, + "loss": 1.2375, + "step": 22580 + }, + { + "epoch": 0.5293342006716011, + "grad_norm": 1.553521990776062, + "learning_rate": 9.083098057652515e-05, + "loss": 0.1568, + "step": 22581 + }, + { + "epoch": 0.5293576422464061, + "grad_norm": 0.8548592329025269, + "learning_rate": 9.082364547291242e-05, + "loss": 0.0673, + "step": 22582 + }, + { + "epoch": 0.5293810838212111, + "grad_norm": 3.783029079437256, + "learning_rate": 9.081631041909083e-05, + "loss": 0.4765, + "step": 22583 + }, + { + "epoch": 0.5294045253960161, + "grad_norm": 0.8752313256263733, + "learning_rate": 9.08089754151002e-05, + "loss": 0.0769, + "step": 22584 + }, + { + "epoch": 0.5294279669708211, + "grad_norm": 3.1866073608398438, + "learning_rate": 9.080164046098034e-05, + "loss": 0.3108, + "step": 22585 + }, + { + "epoch": 0.5294514085456261, + "grad_norm": 4.43406867980957, + "learning_rate": 9.079430555677102e-05, + "loss": 0.3505, + "step": 22586 + }, + { + "epoch": 0.529474850120431, + "grad_norm": 1.0374743938446045, + "learning_rate": 9.078697070251207e-05, + "loss": 0.0513, + "step": 22587 + }, + { + "epoch": 0.5294982916952361, + "grad_norm": 1.1544888019561768, + "learning_rate": 9.077963589824329e-05, + "loss": 0.0885, + "step": 22588 + }, + { + "epoch": 0.529521733270041, + "grad_norm": 2.68796443939209, + "learning_rate": 9.077230114400442e-05, + "loss": 1.3615, + "step": 22589 + }, + { + "epoch": 0.5295451748448461, + "grad_norm": 1.7870421409606934, + "learning_rate": 9.076496643983535e-05, + "loss": 1.643, + "step": 22590 + }, + { + "epoch": 0.529568616419651, + "grad_norm": 1.415289044380188, + "learning_rate": 9.075763178577581e-05, + "loss": 0.1476, + "step": 22591 + }, + { + "epoch": 0.5295920579944561, + "grad_norm": 2.4629807472229004, + "learning_rate": 9.07502971818656e-05, + "loss": 0.3715, + "step": 22592 + }, + { + "epoch": 0.529615499569261, + "grad_norm": 2.3607077598571777, + "learning_rate": 9.074296262814453e-05, + "loss": 0.2415, + "step": 22593 + }, + { + "epoch": 0.5296389411440661, + "grad_norm": 4.310977935791016, + "learning_rate": 9.073562812465245e-05, + "loss": 0.373, + "step": 22594 + }, + { + "epoch": 0.529662382718871, + "grad_norm": 3.7549359798431396, + "learning_rate": 9.07282936714291e-05, + "loss": 0.3227, + "step": 22595 + }, + { + "epoch": 0.529685824293676, + "grad_norm": 3.185997724533081, + "learning_rate": 9.072095926851427e-05, + "loss": 0.2465, + "step": 22596 + }, + { + "epoch": 0.529709265868481, + "grad_norm": 2.541191577911377, + "learning_rate": 9.07136249159478e-05, + "loss": 0.4096, + "step": 22597 + }, + { + "epoch": 0.529732707443286, + "grad_norm": 2.7567648887634277, + "learning_rate": 9.070629061376947e-05, + "loss": 1.2052, + "step": 22598 + }, + { + "epoch": 0.529756149018091, + "grad_norm": 3.757498264312744, + "learning_rate": 9.069895636201904e-05, + "loss": 0.3362, + "step": 22599 + }, + { + "epoch": 0.529779590592896, + "grad_norm": 6.462819576263428, + "learning_rate": 9.069162216073636e-05, + "loss": 0.3648, + "step": 22600 + }, + { + "epoch": 0.529803032167701, + "grad_norm": 2.8772928714752197, + "learning_rate": 9.068428800996116e-05, + "loss": 0.1976, + "step": 22601 + }, + { + "epoch": 0.529826473742506, + "grad_norm": 4.656257152557373, + "learning_rate": 9.067695390973332e-05, + "loss": 0.3729, + "step": 22602 + }, + { + "epoch": 0.5298499153173111, + "grad_norm": 2.104879379272461, + "learning_rate": 9.066961986009257e-05, + "loss": 0.1897, + "step": 22603 + }, + { + "epoch": 0.529873356892116, + "grad_norm": 1.8345165252685547, + "learning_rate": 9.066228586107873e-05, + "loss": 0.1849, + "step": 22604 + }, + { + "epoch": 0.5298967984669211, + "grad_norm": 0.7582244277000427, + "learning_rate": 9.065495191273154e-05, + "loss": 0.0554, + "step": 22605 + }, + { + "epoch": 0.529920240041726, + "grad_norm": 2.972252368927002, + "learning_rate": 9.06476180150909e-05, + "loss": 0.2854, + "step": 22606 + }, + { + "epoch": 0.529943681616531, + "grad_norm": 3.8466477394104004, + "learning_rate": 9.064028416819654e-05, + "loss": 0.2524, + "step": 22607 + }, + { + "epoch": 0.529967123191336, + "grad_norm": 4.009194850921631, + "learning_rate": 9.063295037208826e-05, + "loss": 0.3609, + "step": 22608 + }, + { + "epoch": 0.529990564766141, + "grad_norm": 0.68752521276474, + "learning_rate": 9.062561662680586e-05, + "loss": 0.06, + "step": 22609 + }, + { + "epoch": 0.530014006340946, + "grad_norm": 3.2075412273406982, + "learning_rate": 9.061828293238912e-05, + "loss": 0.5036, + "step": 22610 + }, + { + "epoch": 0.530037447915751, + "grad_norm": 3.3014297485351562, + "learning_rate": 9.061094928887784e-05, + "loss": 0.467, + "step": 22611 + }, + { + "epoch": 0.530060889490556, + "grad_norm": 3.7734992504119873, + "learning_rate": 9.060361569631182e-05, + "loss": 0.0547, + "step": 22612 + }, + { + "epoch": 0.530084331065361, + "grad_norm": 3.854424238204956, + "learning_rate": 9.059628215473084e-05, + "loss": 2.0115, + "step": 22613 + }, + { + "epoch": 0.530107772640166, + "grad_norm": 1.747782588005066, + "learning_rate": 9.058894866417469e-05, + "loss": 0.2144, + "step": 22614 + }, + { + "epoch": 0.530131214214971, + "grad_norm": 1.197407603263855, + "learning_rate": 9.058161522468317e-05, + "loss": 0.1377, + "step": 22615 + }, + { + "epoch": 0.5301546557897759, + "grad_norm": 2.28842830657959, + "learning_rate": 9.05742818362961e-05, + "loss": 0.213, + "step": 22616 + }, + { + "epoch": 0.530178097364581, + "grad_norm": 2.855956554412842, + "learning_rate": 9.056694849905321e-05, + "loss": 0.2419, + "step": 22617 + }, + { + "epoch": 0.5302015389393859, + "grad_norm": 4.30440616607666, + "learning_rate": 9.05596152129943e-05, + "loss": 0.334, + "step": 22618 + }, + { + "epoch": 0.530224980514191, + "grad_norm": 0.6233290433883667, + "learning_rate": 9.055228197815922e-05, + "loss": 0.0606, + "step": 22619 + }, + { + "epoch": 0.5302484220889959, + "grad_norm": 3.325047492980957, + "learning_rate": 9.054494879458773e-05, + "loss": 0.1506, + "step": 22620 + }, + { + "epoch": 0.530271863663801, + "grad_norm": 4.424548625946045, + "learning_rate": 9.053761566231959e-05, + "loss": 0.2124, + "step": 22621 + }, + { + "epoch": 0.5302953052386059, + "grad_norm": 3.262303352355957, + "learning_rate": 9.053028258139463e-05, + "loss": 0.3271, + "step": 22622 + }, + { + "epoch": 0.530318746813411, + "grad_norm": 1.3750028610229492, + "learning_rate": 9.052294955185262e-05, + "loss": 0.0427, + "step": 22623 + }, + { + "epoch": 0.5303421883882159, + "grad_norm": 4.5751423835754395, + "learning_rate": 9.051561657373334e-05, + "loss": 0.6235, + "step": 22624 + }, + { + "epoch": 0.530365629963021, + "grad_norm": 3.198580265045166, + "learning_rate": 9.050828364707662e-05, + "loss": 0.162, + "step": 22625 + }, + { + "epoch": 0.5303890715378259, + "grad_norm": 3.1529147624969482, + "learning_rate": 9.05009507719222e-05, + "loss": 1.5839, + "step": 22626 + }, + { + "epoch": 0.5304125131126309, + "grad_norm": 2.7164864540100098, + "learning_rate": 9.049361794830989e-05, + "loss": 0.2253, + "step": 22627 + }, + { + "epoch": 0.5304359546874359, + "grad_norm": 3.0136122703552246, + "learning_rate": 9.048628517627948e-05, + "loss": 0.2477, + "step": 22628 + }, + { + "epoch": 0.5304593962622409, + "grad_norm": 2.7358877658843994, + "learning_rate": 9.047895245587076e-05, + "loss": 0.1855, + "step": 22629 + }, + { + "epoch": 0.5304828378370459, + "grad_norm": 1.962458848953247, + "learning_rate": 9.047161978712346e-05, + "loss": 0.7683, + "step": 22630 + }, + { + "epoch": 0.5305062794118509, + "grad_norm": 0.9298766851425171, + "learning_rate": 9.046428717007745e-05, + "loss": 0.0893, + "step": 22631 + }, + { + "epoch": 0.5305297209866559, + "grad_norm": 3.264291524887085, + "learning_rate": 9.04569546047725e-05, + "loss": 0.2448, + "step": 22632 + }, + { + "epoch": 0.5305531625614609, + "grad_norm": 4.825321674346924, + "learning_rate": 9.04496220912484e-05, + "loss": 0.382, + "step": 22633 + }, + { + "epoch": 0.5305766041362658, + "grad_norm": 1.6733638048171997, + "learning_rate": 9.044228962954489e-05, + "loss": 0.2366, + "step": 22634 + }, + { + "epoch": 0.5306000457110709, + "grad_norm": 3.1065664291381836, + "learning_rate": 9.04349572197018e-05, + "loss": 0.3075, + "step": 22635 + }, + { + "epoch": 0.5306234872858758, + "grad_norm": 3.7500202655792236, + "learning_rate": 9.04276248617589e-05, + "loss": 0.3551, + "step": 22636 + }, + { + "epoch": 0.5306469288606809, + "grad_norm": 2.9503772258758545, + "learning_rate": 9.042029255575595e-05, + "loss": 0.46, + "step": 22637 + }, + { + "epoch": 0.5306703704354858, + "grad_norm": 4.178165435791016, + "learning_rate": 9.04129603017328e-05, + "loss": 0.3623, + "step": 22638 + }, + { + "epoch": 0.5306938120102909, + "grad_norm": 5.106112957000732, + "learning_rate": 9.040562809972917e-05, + "loss": 0.3609, + "step": 22639 + }, + { + "epoch": 0.5307172535850958, + "grad_norm": 1.7886054515838623, + "learning_rate": 9.03982959497849e-05, + "loss": 0.1941, + "step": 22640 + }, + { + "epoch": 0.5307406951599009, + "grad_norm": 2.4023001194000244, + "learning_rate": 9.039096385193973e-05, + "loss": 0.1922, + "step": 22641 + }, + { + "epoch": 0.5307641367347058, + "grad_norm": 3.679558515548706, + "learning_rate": 9.038363180623342e-05, + "loss": 0.4676, + "step": 22642 + }, + { + "epoch": 0.5307875783095108, + "grad_norm": 3.124051094055176, + "learning_rate": 9.037629981270582e-05, + "loss": 0.1627, + "step": 22643 + }, + { + "epoch": 0.5308110198843158, + "grad_norm": 2.418210029602051, + "learning_rate": 9.036896787139672e-05, + "loss": 0.2453, + "step": 22644 + }, + { + "epoch": 0.5308344614591208, + "grad_norm": 3.693880796432495, + "learning_rate": 9.036163598234586e-05, + "loss": 0.3779, + "step": 22645 + }, + { + "epoch": 0.5308579030339258, + "grad_norm": 5.554117679595947, + "learning_rate": 9.035430414559301e-05, + "loss": 0.4188, + "step": 22646 + }, + { + "epoch": 0.5308813446087308, + "grad_norm": 3.6818113327026367, + "learning_rate": 9.034697236117799e-05, + "loss": 0.368, + "step": 22647 + }, + { + "epoch": 0.5309047861835358, + "grad_norm": 4.831816673278809, + "learning_rate": 9.033964062914059e-05, + "loss": 0.4978, + "step": 22648 + }, + { + "epoch": 0.5309282277583408, + "grad_norm": 2.1046979427337646, + "learning_rate": 9.033230894952054e-05, + "loss": 0.2918, + "step": 22649 + }, + { + "epoch": 0.5309516693331457, + "grad_norm": 2.7115137577056885, + "learning_rate": 9.032497732235767e-05, + "loss": 0.2776, + "step": 22650 + }, + { + "epoch": 0.5309751109079508, + "grad_norm": 3.032177209854126, + "learning_rate": 9.031764574769176e-05, + "loss": 0.2341, + "step": 22651 + }, + { + "epoch": 0.5309985524827557, + "grad_norm": 5.416610240936279, + "learning_rate": 9.031031422556254e-05, + "loss": 0.5928, + "step": 22652 + }, + { + "epoch": 0.5310219940575608, + "grad_norm": 2.1832005977630615, + "learning_rate": 9.030298275600986e-05, + "loss": 0.1554, + "step": 22653 + }, + { + "epoch": 0.5310454356323658, + "grad_norm": 3.981161594390869, + "learning_rate": 9.029565133907343e-05, + "loss": 0.3462, + "step": 22654 + }, + { + "epoch": 0.5310688772071708, + "grad_norm": 3.3094983100891113, + "learning_rate": 9.02883199747931e-05, + "loss": 0.4946, + "step": 22655 + }, + { + "epoch": 0.5310923187819758, + "grad_norm": 2.614790916442871, + "learning_rate": 9.02809886632086e-05, + "loss": 0.2544, + "step": 22656 + }, + { + "epoch": 0.5311157603567808, + "grad_norm": 3.1560935974121094, + "learning_rate": 9.027365740435974e-05, + "loss": 0.3329, + "step": 22657 + }, + { + "epoch": 0.5311392019315858, + "grad_norm": 2.322856903076172, + "learning_rate": 9.02663261982863e-05, + "loss": 1.5448, + "step": 22658 + }, + { + "epoch": 0.5311626435063908, + "grad_norm": 3.010223388671875, + "learning_rate": 9.025899504502802e-05, + "loss": 0.3178, + "step": 22659 + }, + { + "epoch": 0.5311860850811958, + "grad_norm": 3.1871137619018555, + "learning_rate": 9.025166394462474e-05, + "loss": 0.3181, + "step": 22660 + }, + { + "epoch": 0.5312095266560007, + "grad_norm": 3.2373788356781006, + "learning_rate": 9.02443328971162e-05, + "loss": 0.2824, + "step": 22661 + }, + { + "epoch": 0.5312329682308058, + "grad_norm": 2.5117177963256836, + "learning_rate": 9.023700190254217e-05, + "loss": 0.1893, + "step": 22662 + }, + { + "epoch": 0.5312564098056107, + "grad_norm": 0.7028478980064392, + "learning_rate": 9.022967096094245e-05, + "loss": 0.0646, + "step": 22663 + }, + { + "epoch": 0.5312798513804158, + "grad_norm": 5.197161674499512, + "learning_rate": 9.022234007235681e-05, + "loss": 0.3601, + "step": 22664 + }, + { + "epoch": 0.5313032929552207, + "grad_norm": 3.185469150543213, + "learning_rate": 9.021500923682501e-05, + "loss": 0.3062, + "step": 22665 + }, + { + "epoch": 0.5313267345300258, + "grad_norm": 3.3372225761413574, + "learning_rate": 9.020767845438683e-05, + "loss": 0.2166, + "step": 22666 + }, + { + "epoch": 0.5313501761048307, + "grad_norm": 3.526897430419922, + "learning_rate": 9.02003477250821e-05, + "loss": 0.4165, + "step": 22667 + }, + { + "epoch": 0.5313736176796358, + "grad_norm": 2.7547738552093506, + "learning_rate": 9.019301704895056e-05, + "loss": 0.1952, + "step": 22668 + }, + { + "epoch": 0.5313970592544407, + "grad_norm": 4.914731502532959, + "learning_rate": 9.018568642603196e-05, + "loss": 0.363, + "step": 22669 + }, + { + "epoch": 0.5314205008292457, + "grad_norm": 3.032175302505493, + "learning_rate": 9.017835585636612e-05, + "loss": 0.1159, + "step": 22670 + }, + { + "epoch": 0.5314439424040507, + "grad_norm": 2.043536901473999, + "learning_rate": 9.01710253399928e-05, + "loss": 0.162, + "step": 22671 + }, + { + "epoch": 0.5314673839788557, + "grad_norm": 3.6885972023010254, + "learning_rate": 9.016369487695175e-05, + "loss": 0.2052, + "step": 22672 + }, + { + "epoch": 0.5314908255536607, + "grad_norm": 1.449852705001831, + "learning_rate": 9.015636446728279e-05, + "loss": 0.1749, + "step": 22673 + }, + { + "epoch": 0.5315142671284657, + "grad_norm": 1.4726916551589966, + "learning_rate": 9.014903411102567e-05, + "loss": 0.1505, + "step": 22674 + }, + { + "epoch": 0.5315377087032707, + "grad_norm": 3.2344346046447754, + "learning_rate": 9.014170380822016e-05, + "loss": 0.3048, + "step": 22675 + }, + { + "epoch": 0.5315611502780757, + "grad_norm": 2.40995192527771, + "learning_rate": 9.013437355890606e-05, + "loss": 1.5709, + "step": 22676 + }, + { + "epoch": 0.5315845918528807, + "grad_norm": 1.9314261674880981, + "learning_rate": 9.01270433631231e-05, + "loss": 0.1858, + "step": 22677 + }, + { + "epoch": 0.5316080334276857, + "grad_norm": 4.645638465881348, + "learning_rate": 9.011971322091108e-05, + "loss": 0.2912, + "step": 22678 + }, + { + "epoch": 0.5316314750024906, + "grad_norm": 3.6767807006835938, + "learning_rate": 9.011238313230978e-05, + "loss": 0.5016, + "step": 22679 + }, + { + "epoch": 0.5316549165772957, + "grad_norm": 2.5731828212738037, + "learning_rate": 9.010505309735898e-05, + "loss": 0.2367, + "step": 22680 + }, + { + "epoch": 0.5316783581521006, + "grad_norm": 4.2174072265625, + "learning_rate": 9.009772311609843e-05, + "loss": 0.354, + "step": 22681 + }, + { + "epoch": 0.5317017997269057, + "grad_norm": 1.0106970071792603, + "learning_rate": 9.009039318856793e-05, + "loss": 0.1431, + "step": 22682 + }, + { + "epoch": 0.5317252413017106, + "grad_norm": 4.562479496002197, + "learning_rate": 9.008306331480723e-05, + "loss": 0.2755, + "step": 22683 + }, + { + "epoch": 0.5317486828765157, + "grad_norm": 2.2584619522094727, + "learning_rate": 9.007573349485609e-05, + "loss": 0.2037, + "step": 22684 + }, + { + "epoch": 0.5317721244513206, + "grad_norm": 2.3046202659606934, + "learning_rate": 9.006840372875432e-05, + "loss": 0.2336, + "step": 22685 + }, + { + "epoch": 0.5317955660261257, + "grad_norm": 1.4238700866699219, + "learning_rate": 9.006107401654167e-05, + "loss": 0.1766, + "step": 22686 + }, + { + "epoch": 0.5318190076009306, + "grad_norm": 2.899538516998291, + "learning_rate": 9.005374435825788e-05, + "loss": 0.236, + "step": 22687 + }, + { + "epoch": 0.5318424491757356, + "grad_norm": 2.249634265899658, + "learning_rate": 9.004641475394279e-05, + "loss": 0.124, + "step": 22688 + }, + { + "epoch": 0.5318658907505406, + "grad_norm": 2.6897270679473877, + "learning_rate": 9.003908520363614e-05, + "loss": 0.1159, + "step": 22689 + }, + { + "epoch": 0.5318893323253456, + "grad_norm": 3.0438601970672607, + "learning_rate": 9.003175570737765e-05, + "loss": 0.1952, + "step": 22690 + }, + { + "epoch": 0.5319127739001506, + "grad_norm": 3.347860813140869, + "learning_rate": 9.002442626520712e-05, + "loss": 0.2649, + "step": 22691 + }, + { + "epoch": 0.5319362154749556, + "grad_norm": 0.24176599085330963, + "learning_rate": 9.001709687716437e-05, + "loss": 0.0161, + "step": 22692 + }, + { + "epoch": 0.5319596570497606, + "grad_norm": 3.4387094974517822, + "learning_rate": 9.000976754328915e-05, + "loss": 0.161, + "step": 22693 + }, + { + "epoch": 0.5319830986245656, + "grad_norm": 2.562208414077759, + "learning_rate": 9.000243826362118e-05, + "loss": 1.9302, + "step": 22694 + }, + { + "epoch": 0.5320065401993705, + "grad_norm": 2.234225034713745, + "learning_rate": 8.999510903820028e-05, + "loss": 0.2064, + "step": 22695 + }, + { + "epoch": 0.5320299817741756, + "grad_norm": 1.1032061576843262, + "learning_rate": 8.99877798670662e-05, + "loss": 0.1413, + "step": 22696 + }, + { + "epoch": 0.5320534233489805, + "grad_norm": 1.0164005756378174, + "learning_rate": 8.998045075025871e-05, + "loss": 0.0835, + "step": 22697 + }, + { + "epoch": 0.5320768649237856, + "grad_norm": 0.5812474489212036, + "learning_rate": 8.997312168781758e-05, + "loss": 0.0685, + "step": 22698 + }, + { + "epoch": 0.5321003064985905, + "grad_norm": 0.6765310168266296, + "learning_rate": 8.996579267978255e-05, + "loss": 0.0599, + "step": 22699 + }, + { + "epoch": 0.5321237480733956, + "grad_norm": 3.031751871109009, + "learning_rate": 8.995846372619343e-05, + "loss": 0.2409, + "step": 22700 + }, + { + "epoch": 0.5321471896482005, + "grad_norm": 0.2945026457309723, + "learning_rate": 8.995113482708997e-05, + "loss": 0.0326, + "step": 22701 + }, + { + "epoch": 0.5321706312230056, + "grad_norm": 1.4065730571746826, + "learning_rate": 8.994380598251193e-05, + "loss": 0.1908, + "step": 22702 + }, + { + "epoch": 0.5321940727978105, + "grad_norm": 0.6683770418167114, + "learning_rate": 8.993647719249904e-05, + "loss": 0.0877, + "step": 22703 + }, + { + "epoch": 0.5322175143726156, + "grad_norm": 0.6796673536300659, + "learning_rate": 8.992914845709114e-05, + "loss": 0.0403, + "step": 22704 + }, + { + "epoch": 0.5322409559474206, + "grad_norm": 3.4917101860046387, + "learning_rate": 8.992181977632798e-05, + "loss": 0.2453, + "step": 22705 + }, + { + "epoch": 0.5322643975222255, + "grad_norm": 2.4667651653289795, + "learning_rate": 8.99144911502493e-05, + "loss": 0.1428, + "step": 22706 + }, + { + "epoch": 0.5322878390970306, + "grad_norm": 4.319096088409424, + "learning_rate": 8.990716257889486e-05, + "loss": 0.2335, + "step": 22707 + }, + { + "epoch": 0.5323112806718355, + "grad_norm": 0.5614820718765259, + "learning_rate": 8.989983406230447e-05, + "loss": 0.0417, + "step": 22708 + }, + { + "epoch": 0.5323347222466406, + "grad_norm": 2.097052812576294, + "learning_rate": 8.989250560051786e-05, + "loss": 0.257, + "step": 22709 + }, + { + "epoch": 0.5323581638214455, + "grad_norm": 0.5935290455818176, + "learning_rate": 8.988517719357478e-05, + "loss": 0.0567, + "step": 22710 + }, + { + "epoch": 0.5323816053962506, + "grad_norm": 2.96736741065979, + "learning_rate": 8.987784884151502e-05, + "loss": 0.2455, + "step": 22711 + }, + { + "epoch": 0.5324050469710555, + "grad_norm": 2.4215080738067627, + "learning_rate": 8.987052054437832e-05, + "loss": 0.1725, + "step": 22712 + }, + { + "epoch": 0.5324284885458606, + "grad_norm": 1.337780475616455, + "learning_rate": 8.986319230220449e-05, + "loss": 0.0622, + "step": 22713 + }, + { + "epoch": 0.5324519301206655, + "grad_norm": 1.3668060302734375, + "learning_rate": 8.985586411503326e-05, + "loss": 0.1046, + "step": 22714 + }, + { + "epoch": 0.5324753716954705, + "grad_norm": 0.9053287506103516, + "learning_rate": 8.984853598290435e-05, + "loss": 0.1166, + "step": 22715 + }, + { + "epoch": 0.5324988132702755, + "grad_norm": 3.2026922702789307, + "learning_rate": 8.984120790585762e-05, + "loss": 0.2128, + "step": 22716 + }, + { + "epoch": 0.5325222548450805, + "grad_norm": 1.1803960800170898, + "learning_rate": 8.983387988393275e-05, + "loss": 0.157, + "step": 22717 + }, + { + "epoch": 0.5325456964198855, + "grad_norm": 3.016369104385376, + "learning_rate": 8.982655191716956e-05, + "loss": 0.198, + "step": 22718 + }, + { + "epoch": 0.5325691379946905, + "grad_norm": 4.425835132598877, + "learning_rate": 8.981922400560777e-05, + "loss": 0.3725, + "step": 22719 + }, + { + "epoch": 0.5325925795694955, + "grad_norm": 2.179614305496216, + "learning_rate": 8.981189614928718e-05, + "loss": 0.2392, + "step": 22720 + }, + { + "epoch": 0.5326160211443005, + "grad_norm": 1.1670901775360107, + "learning_rate": 8.98045683482475e-05, + "loss": 0.0717, + "step": 22721 + }, + { + "epoch": 0.5326394627191054, + "grad_norm": 0.5413491725921631, + "learning_rate": 8.979724060252854e-05, + "loss": 0.0299, + "step": 22722 + }, + { + "epoch": 0.5326629042939105, + "grad_norm": 0.507341206073761, + "learning_rate": 8.978991291217003e-05, + "loss": 0.0505, + "step": 22723 + }, + { + "epoch": 0.5326863458687154, + "grad_norm": 3.952855348587036, + "learning_rate": 8.978258527721175e-05, + "loss": 0.4183, + "step": 22724 + }, + { + "epoch": 0.5327097874435205, + "grad_norm": 0.9609048366546631, + "learning_rate": 8.977525769769343e-05, + "loss": 0.0428, + "step": 22725 + }, + { + "epoch": 0.5327332290183254, + "grad_norm": 4.028496265411377, + "learning_rate": 8.976793017365487e-05, + "loss": 0.3253, + "step": 22726 + }, + { + "epoch": 0.5327566705931305, + "grad_norm": 4.186106204986572, + "learning_rate": 8.976060270513575e-05, + "loss": 0.453, + "step": 22727 + }, + { + "epoch": 0.5327801121679354, + "grad_norm": 4.477625370025635, + "learning_rate": 8.975327529217596e-05, + "loss": 0.4789, + "step": 22728 + }, + { + "epoch": 0.5328035537427405, + "grad_norm": 3.6370277404785156, + "learning_rate": 8.974594793481513e-05, + "loss": 0.399, + "step": 22729 + }, + { + "epoch": 0.5328269953175454, + "grad_norm": 3.9133288860321045, + "learning_rate": 8.973862063309312e-05, + "loss": 0.4547, + "step": 22730 + }, + { + "epoch": 0.5328504368923505, + "grad_norm": 2.6390321254730225, + "learning_rate": 8.973129338704964e-05, + "loss": 0.2345, + "step": 22731 + }, + { + "epoch": 0.5328738784671554, + "grad_norm": 1.7622334957122803, + "learning_rate": 8.972396619672443e-05, + "loss": 0.1422, + "step": 22732 + }, + { + "epoch": 0.5328973200419604, + "grad_norm": 0.6246902942657471, + "learning_rate": 8.971663906215728e-05, + "loss": 0.0461, + "step": 22733 + }, + { + "epoch": 0.5329207616167654, + "grad_norm": 2.0237045288085938, + "learning_rate": 8.970931198338794e-05, + "loss": 0.2073, + "step": 22734 + }, + { + "epoch": 0.5329442031915704, + "grad_norm": 0.8733507394790649, + "learning_rate": 8.970198496045616e-05, + "loss": 0.1095, + "step": 22735 + }, + { + "epoch": 0.5329676447663754, + "grad_norm": 1.824911117553711, + "learning_rate": 8.96946579934017e-05, + "loss": 0.2695, + "step": 22736 + }, + { + "epoch": 0.5329910863411804, + "grad_norm": 0.9560599327087402, + "learning_rate": 8.968733108226432e-05, + "loss": 0.0718, + "step": 22737 + }, + { + "epoch": 0.5330145279159854, + "grad_norm": 2.971994638442993, + "learning_rate": 8.968000422708375e-05, + "loss": 0.17, + "step": 22738 + }, + { + "epoch": 0.5330379694907904, + "grad_norm": 0.9741966128349304, + "learning_rate": 8.967267742789974e-05, + "loss": 0.0786, + "step": 22739 + }, + { + "epoch": 0.5330614110655953, + "grad_norm": 3.62418532371521, + "learning_rate": 8.966535068475211e-05, + "loss": 0.2023, + "step": 22740 + }, + { + "epoch": 0.5330848526404004, + "grad_norm": 2.105623960494995, + "learning_rate": 8.96580239976806e-05, + "loss": 0.1507, + "step": 22741 + }, + { + "epoch": 0.5331082942152053, + "grad_norm": 1.367241382598877, + "learning_rate": 8.965069736672491e-05, + "loss": 0.1421, + "step": 22742 + }, + { + "epoch": 0.5331317357900104, + "grad_norm": 3.5615174770355225, + "learning_rate": 8.964337079192484e-05, + "loss": 0.3149, + "step": 22743 + }, + { + "epoch": 0.5331551773648153, + "grad_norm": 4.231812000274658, + "learning_rate": 8.963604427332014e-05, + "loss": 0.27, + "step": 22744 + }, + { + "epoch": 0.5331786189396204, + "grad_norm": 5.649521350860596, + "learning_rate": 8.962871781095053e-05, + "loss": 0.5129, + "step": 22745 + }, + { + "epoch": 0.5332020605144253, + "grad_norm": 5.428858280181885, + "learning_rate": 8.962139140485582e-05, + "loss": 0.3634, + "step": 22746 + }, + { + "epoch": 0.5332255020892304, + "grad_norm": 1.8547178506851196, + "learning_rate": 8.961406505507572e-05, + "loss": 0.1569, + "step": 22747 + }, + { + "epoch": 0.5332489436640353, + "grad_norm": 1.339667558670044, + "learning_rate": 8.960673876165e-05, + "loss": 0.1926, + "step": 22748 + }, + { + "epoch": 0.5332723852388404, + "grad_norm": 0.9066116809844971, + "learning_rate": 8.95994125246184e-05, + "loss": 0.1138, + "step": 22749 + }, + { + "epoch": 0.5332958268136453, + "grad_norm": 0.6947355270385742, + "learning_rate": 8.959208634402067e-05, + "loss": 0.061, + "step": 22750 + }, + { + "epoch": 0.5333192683884503, + "grad_norm": 1.8509814739227295, + "learning_rate": 8.958476021989656e-05, + "loss": 0.0787, + "step": 22751 + }, + { + "epoch": 0.5333427099632553, + "grad_norm": 4.36359167098999, + "learning_rate": 8.957743415228584e-05, + "loss": 0.2281, + "step": 22752 + }, + { + "epoch": 0.5333661515380603, + "grad_norm": 1.3651695251464844, + "learning_rate": 8.957010814122828e-05, + "loss": 0.1058, + "step": 22753 + }, + { + "epoch": 0.5333895931128653, + "grad_norm": 3.671525478363037, + "learning_rate": 8.956278218676361e-05, + "loss": 1.5094, + "step": 22754 + }, + { + "epoch": 0.5334130346876703, + "grad_norm": 1.9824306964874268, + "learning_rate": 8.955545628893154e-05, + "loss": 0.2291, + "step": 22755 + }, + { + "epoch": 0.5334364762624754, + "grad_norm": 3.588413953781128, + "learning_rate": 8.95481304477719e-05, + "loss": 0.3896, + "step": 22756 + }, + { + "epoch": 0.5334599178372803, + "grad_norm": 2.867778778076172, + "learning_rate": 8.954080466332438e-05, + "loss": 1.5581, + "step": 22757 + }, + { + "epoch": 0.5334833594120854, + "grad_norm": 0.5375615358352661, + "learning_rate": 8.953347893562874e-05, + "loss": 0.053, + "step": 22758 + }, + { + "epoch": 0.5335068009868903, + "grad_norm": 0.5887987613677979, + "learning_rate": 8.952615326472476e-05, + "loss": 0.0529, + "step": 22759 + }, + { + "epoch": 0.5335302425616953, + "grad_norm": 3.8005964756011963, + "learning_rate": 8.951882765065215e-05, + "loss": 0.3402, + "step": 22760 + }, + { + "epoch": 0.5335536841365003, + "grad_norm": 3.477529525756836, + "learning_rate": 8.951150209345068e-05, + "loss": 0.4131, + "step": 22761 + }, + { + "epoch": 0.5335771257113053, + "grad_norm": 3.7113699913024902, + "learning_rate": 8.95041765931601e-05, + "loss": 1.7396, + "step": 22762 + }, + { + "epoch": 0.5336005672861103, + "grad_norm": 4.717876434326172, + "learning_rate": 8.949685114982014e-05, + "loss": 0.5644, + "step": 22763 + }, + { + "epoch": 0.5336240088609153, + "grad_norm": 5.482684135437012, + "learning_rate": 8.948952576347053e-05, + "loss": 0.4236, + "step": 22764 + }, + { + "epoch": 0.5336474504357203, + "grad_norm": 4.1582465171813965, + "learning_rate": 8.94822004341511e-05, + "loss": 0.3718, + "step": 22765 + }, + { + "epoch": 0.5336708920105253, + "grad_norm": 2.152590036392212, + "learning_rate": 8.947487516190154e-05, + "loss": 0.2924, + "step": 22766 + }, + { + "epoch": 0.5336943335853302, + "grad_norm": 0.5973043441772461, + "learning_rate": 8.946754994676158e-05, + "loss": 0.0419, + "step": 22767 + }, + { + "epoch": 0.5337177751601353, + "grad_norm": 2.591224193572998, + "learning_rate": 8.946022478877101e-05, + "loss": 0.2184, + "step": 22768 + }, + { + "epoch": 0.5337412167349402, + "grad_norm": 1.9100240468978882, + "learning_rate": 8.945289968796957e-05, + "loss": 0.2357, + "step": 22769 + }, + { + "epoch": 0.5337646583097453, + "grad_norm": 4.016902446746826, + "learning_rate": 8.944557464439696e-05, + "loss": 0.3998, + "step": 22770 + }, + { + "epoch": 0.5337880998845502, + "grad_norm": 3.6365432739257812, + "learning_rate": 8.943824965809299e-05, + "loss": 0.3434, + "step": 22771 + }, + { + "epoch": 0.5338115414593553, + "grad_norm": 3.925438165664673, + "learning_rate": 8.943092472909736e-05, + "loss": 0.4337, + "step": 22772 + }, + { + "epoch": 0.5338349830341602, + "grad_norm": 2.8173718452453613, + "learning_rate": 8.942359985744982e-05, + "loss": 0.3152, + "step": 22773 + }, + { + "epoch": 0.5338584246089653, + "grad_norm": 3.7445356845855713, + "learning_rate": 8.941627504319014e-05, + "loss": 0.3544, + "step": 22774 + }, + { + "epoch": 0.5338818661837702, + "grad_norm": 4.8449015617370605, + "learning_rate": 8.940895028635805e-05, + "loss": 0.6128, + "step": 22775 + }, + { + "epoch": 0.5339053077585753, + "grad_norm": 1.1835567951202393, + "learning_rate": 8.940162558699325e-05, + "loss": 0.1396, + "step": 22776 + }, + { + "epoch": 0.5339287493333802, + "grad_norm": 1.0492793321609497, + "learning_rate": 8.939430094513554e-05, + "loss": 0.0779, + "step": 22777 + }, + { + "epoch": 0.5339521909081852, + "grad_norm": 3.281754732131958, + "learning_rate": 8.93869763608247e-05, + "loss": 0.1875, + "step": 22778 + }, + { + "epoch": 0.5339756324829902, + "grad_norm": 3.953270196914673, + "learning_rate": 8.93796518341004e-05, + "loss": 0.4108, + "step": 22779 + }, + { + "epoch": 0.5339990740577952, + "grad_norm": 1.4735362529754639, + "learning_rate": 8.93723273650024e-05, + "loss": 0.2043, + "step": 22780 + }, + { + "epoch": 0.5340225156326002, + "grad_norm": 3.148801803588867, + "learning_rate": 8.936500295357046e-05, + "loss": 0.4046, + "step": 22781 + }, + { + "epoch": 0.5340459572074052, + "grad_norm": 3.526073455810547, + "learning_rate": 8.935767859984432e-05, + "loss": 0.2279, + "step": 22782 + }, + { + "epoch": 0.5340693987822102, + "grad_norm": 2.338324785232544, + "learning_rate": 8.935035430386369e-05, + "loss": 0.2625, + "step": 22783 + }, + { + "epoch": 0.5340928403570152, + "grad_norm": 0.8002071976661682, + "learning_rate": 8.934303006566837e-05, + "loss": 0.0687, + "step": 22784 + }, + { + "epoch": 0.5341162819318201, + "grad_norm": 1.18073570728302, + "learning_rate": 8.933570588529805e-05, + "loss": 0.0938, + "step": 22785 + }, + { + "epoch": 0.5341397235066252, + "grad_norm": 1.9776815176010132, + "learning_rate": 8.932838176279248e-05, + "loss": 0.2245, + "step": 22786 + }, + { + "epoch": 0.5341631650814301, + "grad_norm": 0.4416990280151367, + "learning_rate": 8.932105769819143e-05, + "loss": 0.027, + "step": 22787 + }, + { + "epoch": 0.5341866066562352, + "grad_norm": 3.149888038635254, + "learning_rate": 8.931373369153457e-05, + "loss": 0.3729, + "step": 22788 + }, + { + "epoch": 0.5342100482310401, + "grad_norm": 3.912717342376709, + "learning_rate": 8.930640974286175e-05, + "loss": 0.4333, + "step": 22789 + }, + { + "epoch": 0.5342334898058452, + "grad_norm": 0.6607580780982971, + "learning_rate": 8.929908585221263e-05, + "loss": 0.0769, + "step": 22790 + }, + { + "epoch": 0.5342569313806501, + "grad_norm": 1.435112476348877, + "learning_rate": 8.929176201962697e-05, + "loss": 0.6497, + "step": 22791 + }, + { + "epoch": 0.5342803729554552, + "grad_norm": 0.9613621234893799, + "learning_rate": 8.928443824514451e-05, + "loss": 0.0665, + "step": 22792 + }, + { + "epoch": 0.5343038145302601, + "grad_norm": 2.4965310096740723, + "learning_rate": 8.927711452880501e-05, + "loss": 0.4699, + "step": 22793 + }, + { + "epoch": 0.5343272561050652, + "grad_norm": 0.8609751462936401, + "learning_rate": 8.926979087064819e-05, + "loss": 0.0542, + "step": 22794 + }, + { + "epoch": 0.5343506976798701, + "grad_norm": 2.889906644821167, + "learning_rate": 8.926246727071374e-05, + "loss": 0.3148, + "step": 22795 + }, + { + "epoch": 0.5343741392546751, + "grad_norm": 4.852987289428711, + "learning_rate": 8.925514372904148e-05, + "loss": 0.4951, + "step": 22796 + }, + { + "epoch": 0.5343975808294801, + "grad_norm": 1.7691047191619873, + "learning_rate": 8.924782024567113e-05, + "loss": 0.1347, + "step": 22797 + }, + { + "epoch": 0.5344210224042851, + "grad_norm": 2.2710256576538086, + "learning_rate": 8.924049682064237e-05, + "loss": 0.2013, + "step": 22798 + }, + { + "epoch": 0.5344444639790901, + "grad_norm": 1.6144448518753052, + "learning_rate": 8.923317345399501e-05, + "loss": 0.1547, + "step": 22799 + }, + { + "epoch": 0.5344679055538951, + "grad_norm": 2.63476824760437, + "learning_rate": 8.92258501457687e-05, + "loss": 0.3354, + "step": 22800 + }, + { + "epoch": 0.5344913471287001, + "grad_norm": 0.9751724004745483, + "learning_rate": 8.921852689600328e-05, + "loss": 0.051, + "step": 22801 + }, + { + "epoch": 0.5345147887035051, + "grad_norm": 2.8109116554260254, + "learning_rate": 8.921120370473842e-05, + "loss": 0.2475, + "step": 22802 + }, + { + "epoch": 0.53453823027831, + "grad_norm": 0.9671927690505981, + "learning_rate": 8.920388057201387e-05, + "loss": 0.1295, + "step": 22803 + }, + { + "epoch": 0.5345616718531151, + "grad_norm": 2.296769380569458, + "learning_rate": 8.919655749786938e-05, + "loss": 0.1888, + "step": 22804 + }, + { + "epoch": 0.53458511342792, + "grad_norm": 0.7219014167785645, + "learning_rate": 8.918923448234465e-05, + "loss": 0.0538, + "step": 22805 + }, + { + "epoch": 0.5346085550027251, + "grad_norm": 3.213529586791992, + "learning_rate": 8.918191152547947e-05, + "loss": 0.2479, + "step": 22806 + }, + { + "epoch": 0.5346319965775301, + "grad_norm": 4.834479331970215, + "learning_rate": 8.917458862731354e-05, + "loss": 0.4345, + "step": 22807 + }, + { + "epoch": 0.5346554381523351, + "grad_norm": 2.4060611724853516, + "learning_rate": 8.916726578788657e-05, + "loss": 1.8988, + "step": 22808 + }, + { + "epoch": 0.5346788797271401, + "grad_norm": 0.9010484218597412, + "learning_rate": 8.915994300723833e-05, + "loss": 0.0711, + "step": 22809 + }, + { + "epoch": 0.5347023213019451, + "grad_norm": 2.044642448425293, + "learning_rate": 8.915262028540856e-05, + "loss": 0.1989, + "step": 22810 + }, + { + "epoch": 0.5347257628767501, + "grad_norm": 3.46974515914917, + "learning_rate": 8.914529762243695e-05, + "loss": 0.3075, + "step": 22811 + }, + { + "epoch": 0.534749204451555, + "grad_norm": 3.6714072227478027, + "learning_rate": 8.913797501836324e-05, + "loss": 0.2245, + "step": 22812 + }, + { + "epoch": 0.5347726460263601, + "grad_norm": 2.1575000286102295, + "learning_rate": 8.913065247322723e-05, + "loss": 0.211, + "step": 22813 + }, + { + "epoch": 0.534796087601165, + "grad_norm": 2.901045799255371, + "learning_rate": 8.91233299870686e-05, + "loss": 0.4663, + "step": 22814 + }, + { + "epoch": 0.5348195291759701, + "grad_norm": 2.699225664138794, + "learning_rate": 8.911600755992705e-05, + "loss": 0.2232, + "step": 22815 + }, + { + "epoch": 0.534842970750775, + "grad_norm": 1.4521547555923462, + "learning_rate": 8.910868519184241e-05, + "loss": 0.1635, + "step": 22816 + }, + { + "epoch": 0.5348664123255801, + "grad_norm": 1.4326856136322021, + "learning_rate": 8.910136288285433e-05, + "loss": 0.0644, + "step": 22817 + }, + { + "epoch": 0.534889853900385, + "grad_norm": 3.810148239135742, + "learning_rate": 8.909404063300253e-05, + "loss": 0.2726, + "step": 22818 + }, + { + "epoch": 0.5349132954751901, + "grad_norm": 2.672701120376587, + "learning_rate": 8.908671844232681e-05, + "loss": 0.2956, + "step": 22819 + }, + { + "epoch": 0.534936737049995, + "grad_norm": 3.950363874435425, + "learning_rate": 8.907939631086685e-05, + "loss": 0.3994, + "step": 22820 + }, + { + "epoch": 0.5349601786248, + "grad_norm": 2.4968721866607666, + "learning_rate": 8.907207423866237e-05, + "loss": 1.7942, + "step": 22821 + }, + { + "epoch": 0.534983620199605, + "grad_norm": 1.8891609907150269, + "learning_rate": 8.906475222575316e-05, + "loss": 0.1558, + "step": 22822 + }, + { + "epoch": 0.53500706177441, + "grad_norm": 3.5222578048706055, + "learning_rate": 8.90574302721789e-05, + "loss": 0.3961, + "step": 22823 + }, + { + "epoch": 0.535030503349215, + "grad_norm": 1.7244997024536133, + "learning_rate": 8.905010837797929e-05, + "loss": 0.2424, + "step": 22824 + }, + { + "epoch": 0.53505394492402, + "grad_norm": 3.064085006713867, + "learning_rate": 8.90427865431941e-05, + "loss": 0.2384, + "step": 22825 + }, + { + "epoch": 0.535077386498825, + "grad_norm": 1.2218793630599976, + "learning_rate": 8.903546476786312e-05, + "loss": 0.0677, + "step": 22826 + }, + { + "epoch": 0.53510082807363, + "grad_norm": 1.5488617420196533, + "learning_rate": 8.9028143052026e-05, + "loss": 0.0543, + "step": 22827 + }, + { + "epoch": 0.535124269648435, + "grad_norm": 3.111290216445923, + "learning_rate": 8.902082139572245e-05, + "loss": 0.1812, + "step": 22828 + }, + { + "epoch": 0.53514771122324, + "grad_norm": 2.888892650604248, + "learning_rate": 8.901349979899227e-05, + "loss": 0.3359, + "step": 22829 + }, + { + "epoch": 0.535171152798045, + "grad_norm": 2.956429958343506, + "learning_rate": 8.900617826187512e-05, + "loss": 0.3344, + "step": 22830 + }, + { + "epoch": 0.53519459437285, + "grad_norm": 4.49926233291626, + "learning_rate": 8.899885678441078e-05, + "loss": 0.5098, + "step": 22831 + }, + { + "epoch": 0.5352180359476549, + "grad_norm": 3.7557148933410645, + "learning_rate": 8.899153536663895e-05, + "loss": 0.2836, + "step": 22832 + }, + { + "epoch": 0.53524147752246, + "grad_norm": 1.7603017091751099, + "learning_rate": 8.898421400859934e-05, + "loss": 0.1012, + "step": 22833 + }, + { + "epoch": 0.5352649190972649, + "grad_norm": 3.85090708732605, + "learning_rate": 8.89768927103317e-05, + "loss": 0.3644, + "step": 22834 + }, + { + "epoch": 0.53528836067207, + "grad_norm": 0.5945640206336975, + "learning_rate": 8.896957147187577e-05, + "loss": 0.0522, + "step": 22835 + }, + { + "epoch": 0.5353118022468749, + "grad_norm": 2.9143636226654053, + "learning_rate": 8.896225029327121e-05, + "loss": 1.4164, + "step": 22836 + }, + { + "epoch": 0.53533524382168, + "grad_norm": 3.327223300933838, + "learning_rate": 8.895492917455781e-05, + "loss": 0.2859, + "step": 22837 + }, + { + "epoch": 0.5353586853964849, + "grad_norm": 1.1740491390228271, + "learning_rate": 8.894760811577529e-05, + "loss": 0.0996, + "step": 22838 + }, + { + "epoch": 0.53538212697129, + "grad_norm": 6.49878454208374, + "learning_rate": 8.894028711696335e-05, + "loss": 0.2098, + "step": 22839 + }, + { + "epoch": 0.5354055685460949, + "grad_norm": 2.6582963466644287, + "learning_rate": 8.893296617816172e-05, + "loss": 0.2206, + "step": 22840 + }, + { + "epoch": 0.5354290101208999, + "grad_norm": 4.764126777648926, + "learning_rate": 8.892564529941016e-05, + "loss": 0.2229, + "step": 22841 + }, + { + "epoch": 0.5354524516957049, + "grad_norm": 3.027416944503784, + "learning_rate": 8.891832448074833e-05, + "loss": 1.6549, + "step": 22842 + }, + { + "epoch": 0.5354758932705099, + "grad_norm": 3.6813085079193115, + "learning_rate": 8.891100372221598e-05, + "loss": 0.3162, + "step": 22843 + }, + { + "epoch": 0.5354993348453149, + "grad_norm": 2.1332592964172363, + "learning_rate": 8.890368302385286e-05, + "loss": 0.2084, + "step": 22844 + }, + { + "epoch": 0.5355227764201199, + "grad_norm": 3.0870471000671387, + "learning_rate": 8.889636238569865e-05, + "loss": 0.2666, + "step": 22845 + }, + { + "epoch": 0.5355462179949249, + "grad_norm": 1.465397834777832, + "learning_rate": 8.888904180779308e-05, + "loss": 0.1397, + "step": 22846 + }, + { + "epoch": 0.5355696595697299, + "grad_norm": 2.470439910888672, + "learning_rate": 8.888172129017591e-05, + "loss": 0.1452, + "step": 22847 + }, + { + "epoch": 0.5355931011445348, + "grad_norm": 4.496996879577637, + "learning_rate": 8.887440083288682e-05, + "loss": 0.5097, + "step": 22848 + }, + { + "epoch": 0.5356165427193399, + "grad_norm": 1.1348124742507935, + "learning_rate": 8.88670804359655e-05, + "loss": 0.1518, + "step": 22849 + }, + { + "epoch": 0.5356399842941448, + "grad_norm": 3.1470043659210205, + "learning_rate": 8.885976009945173e-05, + "loss": 0.1857, + "step": 22850 + }, + { + "epoch": 0.5356634258689499, + "grad_norm": 4.80931282043457, + "learning_rate": 8.885243982338525e-05, + "loss": 0.5185, + "step": 22851 + }, + { + "epoch": 0.5356868674437548, + "grad_norm": 1.0448143482208252, + "learning_rate": 8.884511960780575e-05, + "loss": 0.1317, + "step": 22852 + }, + { + "epoch": 0.5357103090185599, + "grad_norm": 4.074031352996826, + "learning_rate": 8.883779945275292e-05, + "loss": 0.2921, + "step": 22853 + }, + { + "epoch": 0.5357337505933648, + "grad_norm": 0.9113491177558899, + "learning_rate": 8.883047935826652e-05, + "loss": 0.1475, + "step": 22854 + }, + { + "epoch": 0.5357571921681699, + "grad_norm": 3.800849437713623, + "learning_rate": 8.882315932438626e-05, + "loss": 0.3483, + "step": 22855 + }, + { + "epoch": 0.5357806337429748, + "grad_norm": 2.40368390083313, + "learning_rate": 8.881583935115182e-05, + "loss": 0.2627, + "step": 22856 + }, + { + "epoch": 0.5358040753177798, + "grad_norm": 2.1014161109924316, + "learning_rate": 8.880851943860298e-05, + "loss": 0.3122, + "step": 22857 + }, + { + "epoch": 0.5358275168925849, + "grad_norm": 3.4805827140808105, + "learning_rate": 8.880119958677941e-05, + "loss": 0.2826, + "step": 22858 + }, + { + "epoch": 0.5358509584673898, + "grad_norm": 5.530999183654785, + "learning_rate": 8.879387979572084e-05, + "loss": 0.378, + "step": 22859 + }, + { + "epoch": 0.5358744000421949, + "grad_norm": 2.666794538497925, + "learning_rate": 8.878656006546701e-05, + "loss": 0.1612, + "step": 22860 + }, + { + "epoch": 0.5358978416169998, + "grad_norm": 1.693485140800476, + "learning_rate": 8.877924039605758e-05, + "loss": 0.093, + "step": 22861 + }, + { + "epoch": 0.5359212831918049, + "grad_norm": 0.9767745733261108, + "learning_rate": 8.877192078753236e-05, + "loss": 0.0421, + "step": 22862 + }, + { + "epoch": 0.5359447247666098, + "grad_norm": 1.7387831211090088, + "learning_rate": 8.876460123993098e-05, + "loss": 0.2324, + "step": 22863 + }, + { + "epoch": 0.5359681663414149, + "grad_norm": 4.029706001281738, + "learning_rate": 8.875728175329323e-05, + "loss": 0.4358, + "step": 22864 + }, + { + "epoch": 0.5359916079162198, + "grad_norm": 2.9083251953125, + "learning_rate": 8.874996232765875e-05, + "loss": 0.2164, + "step": 22865 + }, + { + "epoch": 0.5360150494910249, + "grad_norm": 4.058785915374756, + "learning_rate": 8.87426429630673e-05, + "loss": 0.4699, + "step": 22866 + }, + { + "epoch": 0.5360384910658298, + "grad_norm": 4.399571895599365, + "learning_rate": 8.873532365955858e-05, + "loss": 0.3397, + "step": 22867 + }, + { + "epoch": 0.5360619326406348, + "grad_norm": 1.57384192943573, + "learning_rate": 8.872800441717231e-05, + "loss": 0.0849, + "step": 22868 + }, + { + "epoch": 0.5360853742154398, + "grad_norm": 2.8690037727355957, + "learning_rate": 8.872068523594823e-05, + "loss": 0.3616, + "step": 22869 + }, + { + "epoch": 0.5361088157902448, + "grad_norm": 0.842850923538208, + "learning_rate": 8.871336611592602e-05, + "loss": 0.1014, + "step": 22870 + }, + { + "epoch": 0.5361322573650498, + "grad_norm": 2.268455982208252, + "learning_rate": 8.870604705714539e-05, + "loss": 0.1162, + "step": 22871 + }, + { + "epoch": 0.5361556989398548, + "grad_norm": 2.069671630859375, + "learning_rate": 8.869872805964607e-05, + "loss": 0.2211, + "step": 22872 + }, + { + "epoch": 0.5361791405146598, + "grad_norm": 3.832868814468384, + "learning_rate": 8.869140912346774e-05, + "loss": 0.2513, + "step": 22873 + }, + { + "epoch": 0.5362025820894648, + "grad_norm": 0.8567213416099548, + "learning_rate": 8.868409024865018e-05, + "loss": 0.1243, + "step": 22874 + }, + { + "epoch": 0.5362260236642697, + "grad_norm": 0.4190548360347748, + "learning_rate": 8.867677143523304e-05, + "loss": 0.0217, + "step": 22875 + }, + { + "epoch": 0.5362494652390748, + "grad_norm": 2.256809711456299, + "learning_rate": 8.866945268325609e-05, + "loss": 0.2242, + "step": 22876 + }, + { + "epoch": 0.5362729068138797, + "grad_norm": 3.961168050765991, + "learning_rate": 8.866213399275899e-05, + "loss": 0.4601, + "step": 22877 + }, + { + "epoch": 0.5362963483886848, + "grad_norm": 2.8325440883636475, + "learning_rate": 8.865481536378147e-05, + "loss": 0.2, + "step": 22878 + }, + { + "epoch": 0.5363197899634897, + "grad_norm": 2.4659128189086914, + "learning_rate": 8.864749679636324e-05, + "loss": 0.1628, + "step": 22879 + }, + { + "epoch": 0.5363432315382948, + "grad_norm": 1.6760252714157104, + "learning_rate": 8.864017829054402e-05, + "loss": 0.1092, + "step": 22880 + }, + { + "epoch": 0.5363666731130997, + "grad_norm": 0.3775373697280884, + "learning_rate": 8.863285984636349e-05, + "loss": 0.0245, + "step": 22881 + }, + { + "epoch": 0.5363901146879048, + "grad_norm": 4.729948997497559, + "learning_rate": 8.862554146386142e-05, + "loss": 0.3851, + "step": 22882 + }, + { + "epoch": 0.5364135562627097, + "grad_norm": 0.6689310073852539, + "learning_rate": 8.861822314307746e-05, + "loss": 0.0554, + "step": 22883 + }, + { + "epoch": 0.5364369978375148, + "grad_norm": 2.7363924980163574, + "learning_rate": 8.861090488405132e-05, + "loss": 0.2682, + "step": 22884 + }, + { + "epoch": 0.5364604394123197, + "grad_norm": 3.974442720413208, + "learning_rate": 8.860358668682273e-05, + "loss": 0.3227, + "step": 22885 + }, + { + "epoch": 0.5364838809871247, + "grad_norm": 3.2374460697174072, + "learning_rate": 8.859626855143142e-05, + "loss": 0.2793, + "step": 22886 + }, + { + "epoch": 0.5365073225619297, + "grad_norm": 2.7049403190612793, + "learning_rate": 8.858895047791709e-05, + "loss": 0.3637, + "step": 22887 + }, + { + "epoch": 0.5365307641367347, + "grad_norm": 0.5269322395324707, + "learning_rate": 8.858163246631941e-05, + "loss": 0.0548, + "step": 22888 + }, + { + "epoch": 0.5365542057115397, + "grad_norm": 3.0790421962738037, + "learning_rate": 8.857431451667813e-05, + "loss": 0.3061, + "step": 22889 + }, + { + "epoch": 0.5365776472863447, + "grad_norm": 3.300975799560547, + "learning_rate": 8.856699662903296e-05, + "loss": 2.1618, + "step": 22890 + }, + { + "epoch": 0.5366010888611497, + "grad_norm": 2.1000795364379883, + "learning_rate": 8.855967880342355e-05, + "loss": 0.1308, + "step": 22891 + }, + { + "epoch": 0.5366245304359547, + "grad_norm": 1.0135685205459595, + "learning_rate": 8.855236103988966e-05, + "loss": 0.0622, + "step": 22892 + }, + { + "epoch": 0.5366479720107596, + "grad_norm": 1.148722529411316, + "learning_rate": 8.8545043338471e-05, + "loss": 0.0958, + "step": 22893 + }, + { + "epoch": 0.5366714135855647, + "grad_norm": 3.3637547492980957, + "learning_rate": 8.853772569920723e-05, + "loss": 0.5032, + "step": 22894 + }, + { + "epoch": 0.5366948551603696, + "grad_norm": 3.82971453666687, + "learning_rate": 8.853040812213809e-05, + "loss": 0.3752, + "step": 22895 + }, + { + "epoch": 0.5367182967351747, + "grad_norm": 3.8173532485961914, + "learning_rate": 8.852309060730328e-05, + "loss": 0.439, + "step": 22896 + }, + { + "epoch": 0.5367417383099796, + "grad_norm": 2.025247812271118, + "learning_rate": 8.851577315474247e-05, + "loss": 0.1182, + "step": 22897 + }, + { + "epoch": 0.5367651798847847, + "grad_norm": 1.1150623559951782, + "learning_rate": 8.850845576449543e-05, + "loss": 0.1544, + "step": 22898 + }, + { + "epoch": 0.5367886214595896, + "grad_norm": 0.5334940552711487, + "learning_rate": 8.850113843660182e-05, + "loss": 0.0482, + "step": 22899 + }, + { + "epoch": 0.5368120630343947, + "grad_norm": 2.3449954986572266, + "learning_rate": 8.849382117110138e-05, + "loss": 0.1511, + "step": 22900 + }, + { + "epoch": 0.5368355046091996, + "grad_norm": 1.9862887859344482, + "learning_rate": 8.848650396803377e-05, + "loss": 0.2438, + "step": 22901 + }, + { + "epoch": 0.5368589461840046, + "grad_norm": 3.7029619216918945, + "learning_rate": 8.847918682743874e-05, + "loss": 0.3712, + "step": 22902 + }, + { + "epoch": 0.5368823877588096, + "grad_norm": 3.900136947631836, + "learning_rate": 8.847186974935594e-05, + "loss": 0.6384, + "step": 22903 + }, + { + "epoch": 0.5369058293336146, + "grad_norm": 2.662583351135254, + "learning_rate": 8.84645527338251e-05, + "loss": 0.2348, + "step": 22904 + }, + { + "epoch": 0.5369292709084196, + "grad_norm": 1.4123796224594116, + "learning_rate": 8.845723578088593e-05, + "loss": 0.1256, + "step": 22905 + }, + { + "epoch": 0.5369527124832246, + "grad_norm": 3.216817617416382, + "learning_rate": 8.844991889057811e-05, + "loss": 0.2066, + "step": 22906 + }, + { + "epoch": 0.5369761540580296, + "grad_norm": 2.9043498039245605, + "learning_rate": 8.844260206294138e-05, + "loss": 0.4295, + "step": 22907 + }, + { + "epoch": 0.5369995956328346, + "grad_norm": 2.9525697231292725, + "learning_rate": 8.84352852980154e-05, + "loss": 0.3201, + "step": 22908 + }, + { + "epoch": 0.5370230372076397, + "grad_norm": 2.4407355785369873, + "learning_rate": 8.842796859583984e-05, + "loss": 0.1497, + "step": 22909 + }, + { + "epoch": 0.5370464787824446, + "grad_norm": 2.6029109954833984, + "learning_rate": 8.842065195645452e-05, + "loss": 0.1508, + "step": 22910 + }, + { + "epoch": 0.5370699203572497, + "grad_norm": 2.445612907409668, + "learning_rate": 8.841333537989903e-05, + "loss": 0.1944, + "step": 22911 + }, + { + "epoch": 0.5370933619320546, + "grad_norm": 3.025226593017578, + "learning_rate": 8.840601886621314e-05, + "loss": 0.2791, + "step": 22912 + }, + { + "epoch": 0.5371168035068596, + "grad_norm": 1.2044488191604614, + "learning_rate": 8.839870241543647e-05, + "loss": 0.1711, + "step": 22913 + }, + { + "epoch": 0.5371402450816646, + "grad_norm": 1.6189236640930176, + "learning_rate": 8.839138602760882e-05, + "loss": 0.2067, + "step": 22914 + }, + { + "epoch": 0.5371636866564696, + "grad_norm": 2.317227602005005, + "learning_rate": 8.838406970276981e-05, + "loss": 0.1673, + "step": 22915 + }, + { + "epoch": 0.5371871282312746, + "grad_norm": 2.8177289962768555, + "learning_rate": 8.837675344095915e-05, + "loss": 0.29, + "step": 22916 + }, + { + "epoch": 0.5372105698060796, + "grad_norm": 3.6476995944976807, + "learning_rate": 8.836943724221658e-05, + "loss": 0.268, + "step": 22917 + }, + { + "epoch": 0.5372340113808846, + "grad_norm": 1.4687683582305908, + "learning_rate": 8.836212110658177e-05, + "loss": 1.4455, + "step": 22918 + }, + { + "epoch": 0.5372574529556896, + "grad_norm": 1.9697680473327637, + "learning_rate": 8.83548050340944e-05, + "loss": 0.1642, + "step": 22919 + }, + { + "epoch": 0.5372808945304945, + "grad_norm": 12.790617942810059, + "learning_rate": 8.834748902479421e-05, + "loss": 0.804, + "step": 22920 + }, + { + "epoch": 0.5373043361052996, + "grad_norm": 2.536302089691162, + "learning_rate": 8.834017307872085e-05, + "loss": 0.1965, + "step": 22921 + }, + { + "epoch": 0.5373277776801045, + "grad_norm": 4.362029552459717, + "learning_rate": 8.833285719591402e-05, + "loss": 0.325, + "step": 22922 + }, + { + "epoch": 0.5373512192549096, + "grad_norm": 3.1307785511016846, + "learning_rate": 8.832554137641344e-05, + "loss": 0.2453, + "step": 22923 + }, + { + "epoch": 0.5373746608297145, + "grad_norm": 2.279918909072876, + "learning_rate": 8.831822562025883e-05, + "loss": 0.2158, + "step": 22924 + }, + { + "epoch": 0.5373981024045196, + "grad_norm": 0.8434581756591797, + "learning_rate": 8.831090992748986e-05, + "loss": 0.0749, + "step": 22925 + }, + { + "epoch": 0.5374215439793245, + "grad_norm": 2.9981324672698975, + "learning_rate": 8.830359429814619e-05, + "loss": 0.719, + "step": 22926 + }, + { + "epoch": 0.5374449855541296, + "grad_norm": 3.0913829803466797, + "learning_rate": 8.829627873226757e-05, + "loss": 0.1936, + "step": 22927 + }, + { + "epoch": 0.5374684271289345, + "grad_norm": 3.555457592010498, + "learning_rate": 8.828896322989368e-05, + "loss": 0.4975, + "step": 22928 + }, + { + "epoch": 0.5374918687037396, + "grad_norm": 1.0847573280334473, + "learning_rate": 8.828164779106418e-05, + "loss": 0.095, + "step": 22929 + }, + { + "epoch": 0.5375153102785445, + "grad_norm": 4.203252792358398, + "learning_rate": 8.827433241581879e-05, + "loss": 0.2824, + "step": 22930 + }, + { + "epoch": 0.5375387518533495, + "grad_norm": 2.9444942474365234, + "learning_rate": 8.826701710419722e-05, + "loss": 0.0846, + "step": 22931 + }, + { + "epoch": 0.5375621934281545, + "grad_norm": 1.6008245944976807, + "learning_rate": 8.825970185623913e-05, + "loss": 0.1173, + "step": 22932 + }, + { + "epoch": 0.5375856350029595, + "grad_norm": 2.9942543506622314, + "learning_rate": 8.825238667198425e-05, + "loss": 0.3661, + "step": 22933 + }, + { + "epoch": 0.5376090765777645, + "grad_norm": 0.5717900991439819, + "learning_rate": 8.824507155147219e-05, + "loss": 0.0285, + "step": 22934 + }, + { + "epoch": 0.5376325181525695, + "grad_norm": 2.5115585327148438, + "learning_rate": 8.823775649474275e-05, + "loss": 0.1853, + "step": 22935 + }, + { + "epoch": 0.5376559597273745, + "grad_norm": 2.539482593536377, + "learning_rate": 8.823044150183555e-05, + "loss": 0.1397, + "step": 22936 + }, + { + "epoch": 0.5376794013021795, + "grad_norm": 2.405858278274536, + "learning_rate": 8.822312657279033e-05, + "loss": 0.2433, + "step": 22937 + }, + { + "epoch": 0.5377028428769844, + "grad_norm": 3.6772940158843994, + "learning_rate": 8.821581170764677e-05, + "loss": 1.5981, + "step": 22938 + }, + { + "epoch": 0.5377262844517895, + "grad_norm": 2.1948630809783936, + "learning_rate": 8.820849690644452e-05, + "loss": 0.7297, + "step": 22939 + }, + { + "epoch": 0.5377497260265944, + "grad_norm": 5.193979740142822, + "learning_rate": 8.82011821692233e-05, + "loss": 0.4479, + "step": 22940 + }, + { + "epoch": 0.5377731676013995, + "grad_norm": 1.2278035879135132, + "learning_rate": 8.819386749602283e-05, + "loss": 0.1124, + "step": 22941 + }, + { + "epoch": 0.5377966091762044, + "grad_norm": 2.0082848072052, + "learning_rate": 8.81865528868827e-05, + "loss": 0.1301, + "step": 22942 + }, + { + "epoch": 0.5378200507510095, + "grad_norm": 1.2309439182281494, + "learning_rate": 8.817923834184273e-05, + "loss": 0.1169, + "step": 22943 + }, + { + "epoch": 0.5378434923258144, + "grad_norm": 2.152987480163574, + "learning_rate": 8.817192386094251e-05, + "loss": 0.1605, + "step": 22944 + }, + { + "epoch": 0.5378669339006195, + "grad_norm": 4.717626094818115, + "learning_rate": 8.816460944422178e-05, + "loss": 0.3164, + "step": 22945 + }, + { + "epoch": 0.5378903754754244, + "grad_norm": 0.5915259122848511, + "learning_rate": 8.815729509172017e-05, + "loss": 0.041, + "step": 22946 + }, + { + "epoch": 0.5379138170502294, + "grad_norm": 2.394745111465454, + "learning_rate": 8.814998080347746e-05, + "loss": 0.2885, + "step": 22947 + }, + { + "epoch": 0.5379372586250344, + "grad_norm": 2.7189018726348877, + "learning_rate": 8.814266657953327e-05, + "loss": 0.3713, + "step": 22948 + }, + { + "epoch": 0.5379607001998394, + "grad_norm": 1.18766188621521, + "learning_rate": 8.81353524199273e-05, + "loss": 0.0527, + "step": 22949 + }, + { + "epoch": 0.5379841417746444, + "grad_norm": 2.1681132316589355, + "learning_rate": 8.812803832469927e-05, + "loss": 0.2254, + "step": 22950 + }, + { + "epoch": 0.5380075833494494, + "grad_norm": 0.4034351706504822, + "learning_rate": 8.81207242938888e-05, + "loss": 0.0319, + "step": 22951 + }, + { + "epoch": 0.5380310249242544, + "grad_norm": 1.2707180976867676, + "learning_rate": 8.811341032753564e-05, + "loss": 0.0909, + "step": 22952 + }, + { + "epoch": 0.5380544664990594, + "grad_norm": 5.090408802032471, + "learning_rate": 8.810609642567945e-05, + "loss": 0.4763, + "step": 22953 + }, + { + "epoch": 0.5380779080738644, + "grad_norm": 0.4352875053882599, + "learning_rate": 8.80987825883599e-05, + "loss": 0.0223, + "step": 22954 + }, + { + "epoch": 0.5381013496486694, + "grad_norm": 4.113614082336426, + "learning_rate": 8.809146881561669e-05, + "loss": 0.4438, + "step": 22955 + }, + { + "epoch": 0.5381247912234743, + "grad_norm": 3.099707841873169, + "learning_rate": 8.808415510748953e-05, + "loss": 0.3803, + "step": 22956 + }, + { + "epoch": 0.5381482327982794, + "grad_norm": 4.548293113708496, + "learning_rate": 8.807684146401804e-05, + "loss": 0.4666, + "step": 22957 + }, + { + "epoch": 0.5381716743730843, + "grad_norm": 4.855387210845947, + "learning_rate": 8.806952788524195e-05, + "loss": 0.326, + "step": 22958 + }, + { + "epoch": 0.5381951159478894, + "grad_norm": 4.350375652313232, + "learning_rate": 8.806221437120095e-05, + "loss": 0.4341, + "step": 22959 + }, + { + "epoch": 0.5382185575226944, + "grad_norm": 0.4307735562324524, + "learning_rate": 8.805490092193472e-05, + "loss": 0.0302, + "step": 22960 + }, + { + "epoch": 0.5382419990974994, + "grad_norm": 3.753471851348877, + "learning_rate": 8.804758753748292e-05, + "loss": 0.3038, + "step": 22961 + }, + { + "epoch": 0.5382654406723044, + "grad_norm": 2.7395877838134766, + "learning_rate": 8.804027421788526e-05, + "loss": 0.2929, + "step": 22962 + }, + { + "epoch": 0.5382888822471094, + "grad_norm": 2.286536693572998, + "learning_rate": 8.80329609631814e-05, + "loss": 0.1642, + "step": 22963 + }, + { + "epoch": 0.5383123238219144, + "grad_norm": 0.9820874929428101, + "learning_rate": 8.802564777341103e-05, + "loss": 0.0773, + "step": 22964 + }, + { + "epoch": 0.5383357653967193, + "grad_norm": 1.139703631401062, + "learning_rate": 8.801833464861383e-05, + "loss": 0.1203, + "step": 22965 + }, + { + "epoch": 0.5383592069715244, + "grad_norm": 0.9052955508232117, + "learning_rate": 8.80110215888295e-05, + "loss": 0.0603, + "step": 22966 + }, + { + "epoch": 0.5383826485463293, + "grad_norm": 1.724356770515442, + "learning_rate": 8.800370859409767e-05, + "loss": 0.136, + "step": 22967 + }, + { + "epoch": 0.5384060901211344, + "grad_norm": 6.474277019500732, + "learning_rate": 8.799639566445808e-05, + "loss": 0.3737, + "step": 22968 + }, + { + "epoch": 0.5384295316959393, + "grad_norm": 0.8308912515640259, + "learning_rate": 8.79890827999504e-05, + "loss": 0.0969, + "step": 22969 + }, + { + "epoch": 0.5384529732707444, + "grad_norm": 0.39536014199256897, + "learning_rate": 8.798177000061422e-05, + "loss": 0.0248, + "step": 22970 + }, + { + "epoch": 0.5384764148455493, + "grad_norm": 1.8617154359817505, + "learning_rate": 8.797445726648934e-05, + "loss": 0.2416, + "step": 22971 + }, + { + "epoch": 0.5384998564203544, + "grad_norm": 5.354311943054199, + "learning_rate": 8.796714459761541e-05, + "loss": 0.082, + "step": 22972 + }, + { + "epoch": 0.5385232979951593, + "grad_norm": 2.854027509689331, + "learning_rate": 8.795983199403209e-05, + "loss": 0.2508, + "step": 22973 + }, + { + "epoch": 0.5385467395699644, + "grad_norm": 3.844120740890503, + "learning_rate": 8.795251945577904e-05, + "loss": 0.2747, + "step": 22974 + }, + { + "epoch": 0.5385701811447693, + "grad_norm": 4.599196910858154, + "learning_rate": 8.794520698289598e-05, + "loss": 0.3058, + "step": 22975 + }, + { + "epoch": 0.5385936227195743, + "grad_norm": 2.9579927921295166, + "learning_rate": 8.793789457542256e-05, + "loss": 1.7097, + "step": 22976 + }, + { + "epoch": 0.5386170642943793, + "grad_norm": 0.6257365345954895, + "learning_rate": 8.793058223339845e-05, + "loss": 0.076, + "step": 22977 + }, + { + "epoch": 0.5386405058691843, + "grad_norm": 8.61191463470459, + "learning_rate": 8.792326995686335e-05, + "loss": 0.1852, + "step": 22978 + }, + { + "epoch": 0.5386639474439893, + "grad_norm": 5.093127250671387, + "learning_rate": 8.791595774585695e-05, + "loss": 0.3407, + "step": 22979 + }, + { + "epoch": 0.5386873890187943, + "grad_norm": 2.852325201034546, + "learning_rate": 8.790864560041885e-05, + "loss": 0.1438, + "step": 22980 + }, + { + "epoch": 0.5387108305935993, + "grad_norm": 3.564509868621826, + "learning_rate": 8.790133352058883e-05, + "loss": 0.4541, + "step": 22981 + }, + { + "epoch": 0.5387342721684043, + "grad_norm": 1.0959392786026, + "learning_rate": 8.789402150640646e-05, + "loss": 0.0551, + "step": 22982 + }, + { + "epoch": 0.5387577137432092, + "grad_norm": 2.987877130508423, + "learning_rate": 8.788670955791151e-05, + "loss": 0.2995, + "step": 22983 + }, + { + "epoch": 0.5387811553180143, + "grad_norm": 0.8206743001937866, + "learning_rate": 8.787939767514359e-05, + "loss": 0.0938, + "step": 22984 + }, + { + "epoch": 0.5388045968928192, + "grad_norm": 3.154346466064453, + "learning_rate": 8.787208585814242e-05, + "loss": 0.26, + "step": 22985 + }, + { + "epoch": 0.5388280384676243, + "grad_norm": 1.1603100299835205, + "learning_rate": 8.786477410694764e-05, + "loss": 0.1338, + "step": 22986 + }, + { + "epoch": 0.5388514800424292, + "grad_norm": 1.295651912689209, + "learning_rate": 8.785746242159896e-05, + "loss": 0.1288, + "step": 22987 + }, + { + "epoch": 0.5388749216172343, + "grad_norm": 1.2441786527633667, + "learning_rate": 8.785015080213602e-05, + "loss": 0.1425, + "step": 22988 + }, + { + "epoch": 0.5388983631920392, + "grad_norm": 2.100050210952759, + "learning_rate": 8.784283924859849e-05, + "loss": 0.1437, + "step": 22989 + }, + { + "epoch": 0.5389218047668443, + "grad_norm": 2.9266536235809326, + "learning_rate": 8.783552776102607e-05, + "loss": 1.5317, + "step": 22990 + }, + { + "epoch": 0.5389452463416492, + "grad_norm": 4.406258583068848, + "learning_rate": 8.782821633945842e-05, + "loss": 0.2174, + "step": 22991 + }, + { + "epoch": 0.5389686879164542, + "grad_norm": 1.4494799375534058, + "learning_rate": 8.78209049839352e-05, + "loss": 0.0457, + "step": 22992 + }, + { + "epoch": 0.5389921294912592, + "grad_norm": 2.2612719535827637, + "learning_rate": 8.781359369449609e-05, + "loss": 0.1076, + "step": 22993 + }, + { + "epoch": 0.5390155710660642, + "grad_norm": 3.749847173690796, + "learning_rate": 8.780628247118078e-05, + "loss": 0.2963, + "step": 22994 + }, + { + "epoch": 0.5390390126408692, + "grad_norm": 5.279480934143066, + "learning_rate": 8.779897131402888e-05, + "loss": 0.4405, + "step": 22995 + }, + { + "epoch": 0.5390624542156742, + "grad_norm": 4.315882205963135, + "learning_rate": 8.779166022308013e-05, + "loss": 0.3834, + "step": 22996 + }, + { + "epoch": 0.5390858957904792, + "grad_norm": 3.2980728149414062, + "learning_rate": 8.77843491983742e-05, + "loss": 0.2598, + "step": 22997 + }, + { + "epoch": 0.5391093373652842, + "grad_norm": 3.9061031341552734, + "learning_rate": 8.777703823995071e-05, + "loss": 0.3089, + "step": 22998 + }, + { + "epoch": 0.5391327789400892, + "grad_norm": 5.474940299987793, + "learning_rate": 8.776972734784937e-05, + "loss": 0.5691, + "step": 22999 + }, + { + "epoch": 0.5391562205148942, + "grad_norm": 2.655709981918335, + "learning_rate": 8.776241652210983e-05, + "loss": 1.69, + "step": 23000 + }, + { + "epoch": 0.5391796620896991, + "grad_norm": 2.902172803878784, + "learning_rate": 8.775510576277177e-05, + "loss": 0.2354, + "step": 23001 + }, + { + "epoch": 0.5392031036645042, + "grad_norm": 1.0712345838546753, + "learning_rate": 8.774779506987482e-05, + "loss": 0.0943, + "step": 23002 + }, + { + "epoch": 0.5392265452393091, + "grad_norm": 2.2652533054351807, + "learning_rate": 8.77404844434587e-05, + "loss": 0.1336, + "step": 23003 + }, + { + "epoch": 0.5392499868141142, + "grad_norm": 4.59879732131958, + "learning_rate": 8.773317388356306e-05, + "loss": 0.5167, + "step": 23004 + }, + { + "epoch": 0.5392734283889191, + "grad_norm": 2.82552433013916, + "learning_rate": 8.772586339022755e-05, + "loss": 0.2459, + "step": 23005 + }, + { + "epoch": 0.5392968699637242, + "grad_norm": 0.7519593834877014, + "learning_rate": 8.771855296349187e-05, + "loss": 0.0687, + "step": 23006 + }, + { + "epoch": 0.5393203115385291, + "grad_norm": 4.50750732421875, + "learning_rate": 8.77112426033956e-05, + "loss": 0.4861, + "step": 23007 + }, + { + "epoch": 0.5393437531133342, + "grad_norm": 3.22200345993042, + "learning_rate": 8.770393230997854e-05, + "loss": 0.2899, + "step": 23008 + }, + { + "epoch": 0.5393671946881391, + "grad_norm": 1.7226558923721313, + "learning_rate": 8.769662208328027e-05, + "loss": 0.0685, + "step": 23009 + }, + { + "epoch": 0.5393906362629441, + "grad_norm": 3.419447660446167, + "learning_rate": 8.768931192334049e-05, + "loss": 1.9608, + "step": 23010 + }, + { + "epoch": 0.5394140778377492, + "grad_norm": 3.6700189113616943, + "learning_rate": 8.768200183019884e-05, + "loss": 0.327, + "step": 23011 + }, + { + "epoch": 0.5394375194125541, + "grad_norm": 5.902551651000977, + "learning_rate": 8.767469180389499e-05, + "loss": 0.6665, + "step": 23012 + }, + { + "epoch": 0.5394609609873592, + "grad_norm": 4.7651567459106445, + "learning_rate": 8.766738184446861e-05, + "loss": 0.2329, + "step": 23013 + }, + { + "epoch": 0.5394844025621641, + "grad_norm": 2.875830888748169, + "learning_rate": 8.766007195195938e-05, + "loss": 0.2635, + "step": 23014 + }, + { + "epoch": 0.5395078441369692, + "grad_norm": 0.6278722286224365, + "learning_rate": 8.765276212640691e-05, + "loss": 0.0546, + "step": 23015 + }, + { + "epoch": 0.5395312857117741, + "grad_norm": 6.020153999328613, + "learning_rate": 8.764545236785095e-05, + "loss": 0.4776, + "step": 23016 + }, + { + "epoch": 0.5395547272865792, + "grad_norm": 5.163860321044922, + "learning_rate": 8.763814267633105e-05, + "loss": 0.5628, + "step": 23017 + }, + { + "epoch": 0.5395781688613841, + "grad_norm": 2.9676077365875244, + "learning_rate": 8.7630833051887e-05, + "loss": 0.1221, + "step": 23018 + }, + { + "epoch": 0.5396016104361891, + "grad_norm": 2.8714075088500977, + "learning_rate": 8.762352349455833e-05, + "loss": 0.2767, + "step": 23019 + }, + { + "epoch": 0.5396250520109941, + "grad_norm": 2.161088466644287, + "learning_rate": 8.76162140043848e-05, + "loss": 0.1179, + "step": 23020 + }, + { + "epoch": 0.5396484935857991, + "grad_norm": 1.1856147050857544, + "learning_rate": 8.760890458140607e-05, + "loss": 0.0453, + "step": 23021 + }, + { + "epoch": 0.5396719351606041, + "grad_norm": 2.492649793624878, + "learning_rate": 8.760159522566173e-05, + "loss": 0.1281, + "step": 23022 + }, + { + "epoch": 0.5396953767354091, + "grad_norm": 4.528236389160156, + "learning_rate": 8.759428593719153e-05, + "loss": 0.3174, + "step": 23023 + }, + { + "epoch": 0.5397188183102141, + "grad_norm": 3.71406626701355, + "learning_rate": 8.758697671603503e-05, + "loss": 0.1904, + "step": 23024 + }, + { + "epoch": 0.5397422598850191, + "grad_norm": 2.767076253890991, + "learning_rate": 8.757966756223198e-05, + "loss": 0.2452, + "step": 23025 + }, + { + "epoch": 0.539765701459824, + "grad_norm": 1.3362441062927246, + "learning_rate": 8.7572358475822e-05, + "loss": 0.6743, + "step": 23026 + }, + { + "epoch": 0.5397891430346291, + "grad_norm": 2.515198230743408, + "learning_rate": 8.756504945684474e-05, + "loss": 0.1673, + "step": 23027 + }, + { + "epoch": 0.539812584609434, + "grad_norm": 2.3488199710845947, + "learning_rate": 8.755774050533988e-05, + "loss": 0.1411, + "step": 23028 + }, + { + "epoch": 0.5398360261842391, + "grad_norm": 2.4789345264434814, + "learning_rate": 8.755043162134709e-05, + "loss": 0.1513, + "step": 23029 + }, + { + "epoch": 0.539859467759044, + "grad_norm": 5.226690769195557, + "learning_rate": 8.754312280490596e-05, + "loss": 0.6383, + "step": 23030 + }, + { + "epoch": 0.5398829093338491, + "grad_norm": 4.143743991851807, + "learning_rate": 8.753581405605619e-05, + "loss": 0.5271, + "step": 23031 + }, + { + "epoch": 0.539906350908654, + "grad_norm": 2.2907705307006836, + "learning_rate": 8.752850537483747e-05, + "loss": 0.2053, + "step": 23032 + }, + { + "epoch": 0.5399297924834591, + "grad_norm": 4.027316093444824, + "learning_rate": 8.752119676128945e-05, + "loss": 0.3137, + "step": 23033 + }, + { + "epoch": 0.539953234058264, + "grad_norm": 2.9257469177246094, + "learning_rate": 8.751388821545175e-05, + "loss": 1.2502, + "step": 23034 + }, + { + "epoch": 0.5399766756330691, + "grad_norm": 2.3231654167175293, + "learning_rate": 8.750657973736404e-05, + "loss": 0.122, + "step": 23035 + }, + { + "epoch": 0.540000117207874, + "grad_norm": 2.931986093521118, + "learning_rate": 8.7499271327066e-05, + "loss": 0.2212, + "step": 23036 + }, + { + "epoch": 0.540023558782679, + "grad_norm": 1.574257254600525, + "learning_rate": 8.749196298459725e-05, + "loss": 0.1745, + "step": 23037 + }, + { + "epoch": 0.540047000357484, + "grad_norm": 1.393065333366394, + "learning_rate": 8.748465470999746e-05, + "loss": 0.1629, + "step": 23038 + }, + { + "epoch": 0.540070441932289, + "grad_norm": 1.1728683710098267, + "learning_rate": 8.747734650330629e-05, + "loss": 0.0492, + "step": 23039 + }, + { + "epoch": 0.540093883507094, + "grad_norm": 2.6720495223999023, + "learning_rate": 8.747003836456337e-05, + "loss": 0.3265, + "step": 23040 + }, + { + "epoch": 0.540117325081899, + "grad_norm": 1.594508171081543, + "learning_rate": 8.74627302938084e-05, + "loss": 0.1341, + "step": 23041 + }, + { + "epoch": 0.540140766656704, + "grad_norm": 0.5330643653869629, + "learning_rate": 8.745542229108102e-05, + "loss": 0.0356, + "step": 23042 + }, + { + "epoch": 0.540164208231509, + "grad_norm": 1.7844839096069336, + "learning_rate": 8.744811435642081e-05, + "loss": 0.1465, + "step": 23043 + }, + { + "epoch": 0.540187649806314, + "grad_norm": 0.8635719418525696, + "learning_rate": 8.744080648986751e-05, + "loss": 0.049, + "step": 23044 + }, + { + "epoch": 0.540211091381119, + "grad_norm": 4.164755344390869, + "learning_rate": 8.743349869146077e-05, + "loss": 0.2022, + "step": 23045 + }, + { + "epoch": 0.5402345329559239, + "grad_norm": 2.9313156604766846, + "learning_rate": 8.742619096124021e-05, + "loss": 0.2014, + "step": 23046 + }, + { + "epoch": 0.540257974530729, + "grad_norm": 5.5618815422058105, + "learning_rate": 8.74188832992455e-05, + "loss": 0.3876, + "step": 23047 + }, + { + "epoch": 0.5402814161055339, + "grad_norm": 3.3820769786834717, + "learning_rate": 8.741157570551628e-05, + "loss": 0.242, + "step": 23048 + }, + { + "epoch": 0.540304857680339, + "grad_norm": 2.0652883052825928, + "learning_rate": 8.740426818009219e-05, + "loss": 0.0582, + "step": 23049 + }, + { + "epoch": 0.5403282992551439, + "grad_norm": 3.4458935260772705, + "learning_rate": 8.73969607230129e-05, + "loss": 0.2358, + "step": 23050 + }, + { + "epoch": 0.540351740829949, + "grad_norm": 1.5332759618759155, + "learning_rate": 8.738965333431808e-05, + "loss": 0.0487, + "step": 23051 + }, + { + "epoch": 0.5403751824047539, + "grad_norm": 3.2830145359039307, + "learning_rate": 8.738234601404733e-05, + "loss": 1.5294, + "step": 23052 + }, + { + "epoch": 0.540398623979559, + "grad_norm": 4.28328275680542, + "learning_rate": 8.737503876224032e-05, + "loss": 0.497, + "step": 23053 + }, + { + "epoch": 0.5404220655543639, + "grad_norm": 2.750471353530884, + "learning_rate": 8.736773157893673e-05, + "loss": 0.2223, + "step": 23054 + }, + { + "epoch": 0.540445507129169, + "grad_norm": 5.1603498458862305, + "learning_rate": 8.736042446417613e-05, + "loss": 0.1919, + "step": 23055 + }, + { + "epoch": 0.5404689487039739, + "grad_norm": 2.0878219604492188, + "learning_rate": 8.735311741799827e-05, + "loss": 0.2633, + "step": 23056 + }, + { + "epoch": 0.5404923902787789, + "grad_norm": 3.2917277812957764, + "learning_rate": 8.734581044044273e-05, + "loss": 0.2407, + "step": 23057 + }, + { + "epoch": 0.5405158318535839, + "grad_norm": 1.0341529846191406, + "learning_rate": 8.73385035315492e-05, + "loss": 0.0922, + "step": 23058 + }, + { + "epoch": 0.5405392734283889, + "grad_norm": 3.203626871109009, + "learning_rate": 8.733119669135731e-05, + "loss": 1.5159, + "step": 23059 + }, + { + "epoch": 0.5405627150031939, + "grad_norm": 2.8911778926849365, + "learning_rate": 8.732388991990668e-05, + "loss": 0.2634, + "step": 23060 + }, + { + "epoch": 0.5405861565779989, + "grad_norm": 3.295318126678467, + "learning_rate": 8.731658321723699e-05, + "loss": 0.467, + "step": 23061 + }, + { + "epoch": 0.540609598152804, + "grad_norm": 0.6359501481056213, + "learning_rate": 8.730927658338787e-05, + "loss": 0.0514, + "step": 23062 + }, + { + "epoch": 0.5406330397276089, + "grad_norm": 6.058008670806885, + "learning_rate": 8.730197001839898e-05, + "loss": 0.4953, + "step": 23063 + }, + { + "epoch": 0.540656481302414, + "grad_norm": 3.460622549057007, + "learning_rate": 8.729466352230997e-05, + "loss": 0.3493, + "step": 23064 + }, + { + "epoch": 0.5406799228772189, + "grad_norm": 3.6880013942718506, + "learning_rate": 8.728735709516044e-05, + "loss": 0.3041, + "step": 23065 + }, + { + "epoch": 0.5407033644520239, + "grad_norm": 2.4908268451690674, + "learning_rate": 8.72800507369901e-05, + "loss": 1.4862, + "step": 23066 + }, + { + "epoch": 0.5407268060268289, + "grad_norm": 4.254985809326172, + "learning_rate": 8.727274444783851e-05, + "loss": 0.3737, + "step": 23067 + }, + { + "epoch": 0.5407502476016339, + "grad_norm": 2.071227788925171, + "learning_rate": 8.72654382277454e-05, + "loss": 0.1058, + "step": 23068 + }, + { + "epoch": 0.5407736891764389, + "grad_norm": 2.809659719467163, + "learning_rate": 8.725813207675037e-05, + "loss": 0.2383, + "step": 23069 + }, + { + "epoch": 0.5407971307512439, + "grad_norm": 3.6416783332824707, + "learning_rate": 8.725082599489311e-05, + "loss": 1.9081, + "step": 23070 + }, + { + "epoch": 0.5408205723260489, + "grad_norm": 1.2911897897720337, + "learning_rate": 8.72435199822132e-05, + "loss": 0.129, + "step": 23071 + }, + { + "epoch": 0.5408440139008539, + "grad_norm": 1.1199193000793457, + "learning_rate": 8.723621403875031e-05, + "loss": 0.078, + "step": 23072 + }, + { + "epoch": 0.5408674554756588, + "grad_norm": 2.5806236267089844, + "learning_rate": 8.72289081645441e-05, + "loss": 0.2308, + "step": 23073 + }, + { + "epoch": 0.5408908970504639, + "grad_norm": 2.2912967205047607, + "learning_rate": 8.722160235963417e-05, + "loss": 0.2011, + "step": 23074 + }, + { + "epoch": 0.5409143386252688, + "grad_norm": 1.8459984064102173, + "learning_rate": 8.72142966240602e-05, + "loss": 0.1207, + "step": 23075 + }, + { + "epoch": 0.5409377802000739, + "grad_norm": 1.997912049293518, + "learning_rate": 8.720699095786181e-05, + "loss": 0.2144, + "step": 23076 + }, + { + "epoch": 0.5409612217748788, + "grad_norm": 1.1946355104446411, + "learning_rate": 8.719968536107866e-05, + "loss": 0.0289, + "step": 23077 + }, + { + "epoch": 0.5409846633496839, + "grad_norm": 5.066964149475098, + "learning_rate": 8.719237983375035e-05, + "loss": 0.4598, + "step": 23078 + }, + { + "epoch": 0.5410081049244888, + "grad_norm": 0.9669526219367981, + "learning_rate": 8.718507437591656e-05, + "loss": 0.0742, + "step": 23079 + }, + { + "epoch": 0.5410315464992939, + "grad_norm": 6.273025035858154, + "learning_rate": 8.717776898761689e-05, + "loss": 0.6622, + "step": 23080 + }, + { + "epoch": 0.5410549880740988, + "grad_norm": 4.320913314819336, + "learning_rate": 8.717046366889105e-05, + "loss": 0.4376, + "step": 23081 + }, + { + "epoch": 0.5410784296489038, + "grad_norm": 3.6809163093566895, + "learning_rate": 8.71631584197786e-05, + "loss": 0.2675, + "step": 23082 + }, + { + "epoch": 0.5411018712237088, + "grad_norm": 1.5419443845748901, + "learning_rate": 8.715585324031923e-05, + "loss": 0.0983, + "step": 23083 + }, + { + "epoch": 0.5411253127985138, + "grad_norm": 5.830020904541016, + "learning_rate": 8.714854813055258e-05, + "loss": 0.6444, + "step": 23084 + }, + { + "epoch": 0.5411487543733188, + "grad_norm": 3.3363466262817383, + "learning_rate": 8.714124309051824e-05, + "loss": 0.2085, + "step": 23085 + }, + { + "epoch": 0.5411721959481238, + "grad_norm": 1.439920425415039, + "learning_rate": 8.71339381202559e-05, + "loss": 0.1711, + "step": 23086 + }, + { + "epoch": 0.5411956375229288, + "grad_norm": 0.22772496938705444, + "learning_rate": 8.712663321980518e-05, + "loss": 0.0396, + "step": 23087 + }, + { + "epoch": 0.5412190790977338, + "grad_norm": 2.2582180500030518, + "learning_rate": 8.711932838920567e-05, + "loss": 0.1917, + "step": 23088 + }, + { + "epoch": 0.5412425206725388, + "grad_norm": 2.6555497646331787, + "learning_rate": 8.711202362849708e-05, + "loss": 0.2865, + "step": 23089 + }, + { + "epoch": 0.5412659622473438, + "grad_norm": 0.6052409410476685, + "learning_rate": 8.710471893771901e-05, + "loss": 0.0519, + "step": 23090 + }, + { + "epoch": 0.5412894038221487, + "grad_norm": 2.1654210090637207, + "learning_rate": 8.709741431691107e-05, + "loss": 0.9189, + "step": 23091 + }, + { + "epoch": 0.5413128453969538, + "grad_norm": 0.42231327295303345, + "learning_rate": 8.709010976611291e-05, + "loss": 0.0395, + "step": 23092 + }, + { + "epoch": 0.5413362869717587, + "grad_norm": 0.8104327917098999, + "learning_rate": 8.708280528536422e-05, + "loss": 0.0727, + "step": 23093 + }, + { + "epoch": 0.5413597285465638, + "grad_norm": 2.7110962867736816, + "learning_rate": 8.70755008747046e-05, + "loss": 0.2274, + "step": 23094 + }, + { + "epoch": 0.5413831701213687, + "grad_norm": 4.715309143066406, + "learning_rate": 8.706819653417363e-05, + "loss": 0.3342, + "step": 23095 + }, + { + "epoch": 0.5414066116961738, + "grad_norm": 0.7946643233299255, + "learning_rate": 8.706089226381102e-05, + "loss": 0.0657, + "step": 23096 + }, + { + "epoch": 0.5414300532709787, + "grad_norm": 2.4713146686553955, + "learning_rate": 8.705358806365639e-05, + "loss": 0.1174, + "step": 23097 + }, + { + "epoch": 0.5414534948457838, + "grad_norm": 1.9425231218338013, + "learning_rate": 8.704628393374932e-05, + "loss": 0.187, + "step": 23098 + }, + { + "epoch": 0.5414769364205887, + "grad_norm": 0.7670047283172607, + "learning_rate": 8.70389798741295e-05, + "loss": 0.0731, + "step": 23099 + }, + { + "epoch": 0.5415003779953937, + "grad_norm": 2.907407760620117, + "learning_rate": 8.703167588483652e-05, + "loss": 2.1158, + "step": 23100 + }, + { + "epoch": 0.5415238195701987, + "grad_norm": 0.7600564360618591, + "learning_rate": 8.702437196591004e-05, + "loss": 0.106, + "step": 23101 + }, + { + "epoch": 0.5415472611450037, + "grad_norm": 0.9113576412200928, + "learning_rate": 8.70170681173897e-05, + "loss": 0.0608, + "step": 23102 + }, + { + "epoch": 0.5415707027198087, + "grad_norm": 1.5357474088668823, + "learning_rate": 8.700976433931509e-05, + "loss": 0.1899, + "step": 23103 + }, + { + "epoch": 0.5415941442946137, + "grad_norm": 0.7985090613365173, + "learning_rate": 8.700246063172583e-05, + "loss": 0.0379, + "step": 23104 + }, + { + "epoch": 0.5416175858694187, + "grad_norm": 2.589002847671509, + "learning_rate": 8.699515699466163e-05, + "loss": 0.1962, + "step": 23105 + }, + { + "epoch": 0.5416410274442237, + "grad_norm": 2.5046920776367188, + "learning_rate": 8.698785342816208e-05, + "loss": 0.2434, + "step": 23106 + }, + { + "epoch": 0.5416644690190286, + "grad_norm": 2.6913983821868896, + "learning_rate": 8.698054993226678e-05, + "loss": 0.1806, + "step": 23107 + }, + { + "epoch": 0.5416879105938337, + "grad_norm": 2.4891860485076904, + "learning_rate": 8.69732465070154e-05, + "loss": 0.1592, + "step": 23108 + }, + { + "epoch": 0.5417113521686386, + "grad_norm": 3.2397215366363525, + "learning_rate": 8.696594315244756e-05, + "loss": 0.1799, + "step": 23109 + }, + { + "epoch": 0.5417347937434437, + "grad_norm": 3.73439884185791, + "learning_rate": 8.695863986860285e-05, + "loss": 0.2171, + "step": 23110 + }, + { + "epoch": 0.5417582353182486, + "grad_norm": 5.499408721923828, + "learning_rate": 8.695133665552095e-05, + "loss": 0.5929, + "step": 23111 + }, + { + "epoch": 0.5417816768930537, + "grad_norm": 2.814221143722534, + "learning_rate": 8.694403351324146e-05, + "loss": 1.4786, + "step": 23112 + }, + { + "epoch": 0.5418051184678587, + "grad_norm": 4.782937526702881, + "learning_rate": 8.693673044180399e-05, + "loss": 0.2396, + "step": 23113 + }, + { + "epoch": 0.5418285600426637, + "grad_norm": 1.1906923055648804, + "learning_rate": 8.692942744124823e-05, + "loss": 0.0696, + "step": 23114 + }, + { + "epoch": 0.5418520016174687, + "grad_norm": 2.9768624305725098, + "learning_rate": 8.692212451161374e-05, + "loss": 1.5653, + "step": 23115 + }, + { + "epoch": 0.5418754431922737, + "grad_norm": 2.4622037410736084, + "learning_rate": 8.691482165294013e-05, + "loss": 0.8823, + "step": 23116 + }, + { + "epoch": 0.5418988847670787, + "grad_norm": 2.3997654914855957, + "learning_rate": 8.690751886526709e-05, + "loss": 1.83, + "step": 23117 + }, + { + "epoch": 0.5419223263418836, + "grad_norm": 4.627529621124268, + "learning_rate": 8.690021614863425e-05, + "loss": 0.2927, + "step": 23118 + }, + { + "epoch": 0.5419457679166887, + "grad_norm": 4.356489658355713, + "learning_rate": 8.68929135030812e-05, + "loss": 0.2605, + "step": 23119 + }, + { + "epoch": 0.5419692094914936, + "grad_norm": 4.943859577178955, + "learning_rate": 8.688561092864755e-05, + "loss": 0.5879, + "step": 23120 + }, + { + "epoch": 0.5419926510662987, + "grad_norm": 3.568692207336426, + "learning_rate": 8.687830842537297e-05, + "loss": 0.2615, + "step": 23121 + }, + { + "epoch": 0.5420160926411036, + "grad_norm": 2.627234935760498, + "learning_rate": 8.687100599329705e-05, + "loss": 0.2256, + "step": 23122 + }, + { + "epoch": 0.5420395342159087, + "grad_norm": 3.2484867572784424, + "learning_rate": 8.68637036324594e-05, + "loss": 0.3393, + "step": 23123 + }, + { + "epoch": 0.5420629757907136, + "grad_norm": 4.15095853805542, + "learning_rate": 8.685640134289968e-05, + "loss": 0.3156, + "step": 23124 + }, + { + "epoch": 0.5420864173655187, + "grad_norm": 3.850775957107544, + "learning_rate": 8.684909912465749e-05, + "loss": 0.1838, + "step": 23125 + }, + { + "epoch": 0.5421098589403236, + "grad_norm": 0.6735920310020447, + "learning_rate": 8.684179697777246e-05, + "loss": 0.0372, + "step": 23126 + }, + { + "epoch": 0.5421333005151286, + "grad_norm": 3.8566272258758545, + "learning_rate": 8.683449490228421e-05, + "loss": 0.2357, + "step": 23127 + }, + { + "epoch": 0.5421567420899336, + "grad_norm": 2.8103535175323486, + "learning_rate": 8.682719289823234e-05, + "loss": 0.2943, + "step": 23128 + }, + { + "epoch": 0.5421801836647386, + "grad_norm": 1.1046632528305054, + "learning_rate": 8.681989096565651e-05, + "loss": 0.1133, + "step": 23129 + }, + { + "epoch": 0.5422036252395436, + "grad_norm": 8.081823348999023, + "learning_rate": 8.68125891045963e-05, + "loss": 0.6117, + "step": 23130 + }, + { + "epoch": 0.5422270668143486, + "grad_norm": 2.589293956756592, + "learning_rate": 8.68052873150914e-05, + "loss": 0.0866, + "step": 23131 + }, + { + "epoch": 0.5422505083891536, + "grad_norm": 1.1038330793380737, + "learning_rate": 8.679798559718136e-05, + "loss": 0.1197, + "step": 23132 + }, + { + "epoch": 0.5422739499639586, + "grad_norm": 3.0964386463165283, + "learning_rate": 8.679068395090581e-05, + "loss": 0.1654, + "step": 23133 + }, + { + "epoch": 0.5422973915387636, + "grad_norm": 3.5421431064605713, + "learning_rate": 8.67833823763044e-05, + "loss": 0.2652, + "step": 23134 + }, + { + "epoch": 0.5423208331135686, + "grad_norm": 3.408003568649292, + "learning_rate": 8.677608087341671e-05, + "loss": 0.2385, + "step": 23135 + }, + { + "epoch": 0.5423442746883735, + "grad_norm": 3.4774539470672607, + "learning_rate": 8.676877944228238e-05, + "loss": 0.2269, + "step": 23136 + }, + { + "epoch": 0.5423677162631786, + "grad_norm": 4.2188005447387695, + "learning_rate": 8.676147808294105e-05, + "loss": 0.3359, + "step": 23137 + }, + { + "epoch": 0.5423911578379835, + "grad_norm": 2.764333486557007, + "learning_rate": 8.675417679543229e-05, + "loss": 0.416, + "step": 23138 + }, + { + "epoch": 0.5424145994127886, + "grad_norm": 0.3663977384567261, + "learning_rate": 8.674687557979575e-05, + "loss": 0.0282, + "step": 23139 + }, + { + "epoch": 0.5424380409875935, + "grad_norm": 0.6972777843475342, + "learning_rate": 8.673957443607099e-05, + "loss": 0.0848, + "step": 23140 + }, + { + "epoch": 0.5424614825623986, + "grad_norm": 4.282898426055908, + "learning_rate": 8.67322733642977e-05, + "loss": 0.1794, + "step": 23141 + }, + { + "epoch": 0.5424849241372035, + "grad_norm": 0.7673207521438599, + "learning_rate": 8.672497236451548e-05, + "loss": 0.0335, + "step": 23142 + }, + { + "epoch": 0.5425083657120086, + "grad_norm": 4.731052875518799, + "learning_rate": 8.671767143676394e-05, + "loss": 0.3675, + "step": 23143 + }, + { + "epoch": 0.5425318072868135, + "grad_norm": 3.2899136543273926, + "learning_rate": 8.67103705810827e-05, + "loss": 0.2665, + "step": 23144 + }, + { + "epoch": 0.5425552488616185, + "grad_norm": 1.8811923265457153, + "learning_rate": 8.670306979751132e-05, + "loss": 1.0501, + "step": 23145 + }, + { + "epoch": 0.5425786904364235, + "grad_norm": 5.377945423126221, + "learning_rate": 8.669576908608949e-05, + "loss": 0.4196, + "step": 23146 + }, + { + "epoch": 0.5426021320112285, + "grad_norm": 1.0711288452148438, + "learning_rate": 8.668846844685678e-05, + "loss": 0.0726, + "step": 23147 + }, + { + "epoch": 0.5426255735860335, + "grad_norm": 3.4250051975250244, + "learning_rate": 8.66811678798528e-05, + "loss": 0.5956, + "step": 23148 + }, + { + "epoch": 0.5426490151608385, + "grad_norm": 3.561056613922119, + "learning_rate": 8.66738673851172e-05, + "loss": 0.2646, + "step": 23149 + }, + { + "epoch": 0.5426724567356435, + "grad_norm": 3.0147764682769775, + "learning_rate": 8.666656696268958e-05, + "loss": 0.1221, + "step": 23150 + }, + { + "epoch": 0.5426958983104485, + "grad_norm": 1.3598142862319946, + "learning_rate": 8.665926661260951e-05, + "loss": 0.1504, + "step": 23151 + }, + { + "epoch": 0.5427193398852534, + "grad_norm": 0.9088742136955261, + "learning_rate": 8.665196633491666e-05, + "loss": 0.038, + "step": 23152 + }, + { + "epoch": 0.5427427814600585, + "grad_norm": 4.036155700683594, + "learning_rate": 8.664466612965056e-05, + "loss": 0.5725, + "step": 23153 + }, + { + "epoch": 0.5427662230348634, + "grad_norm": 3.618772029876709, + "learning_rate": 8.663736599685094e-05, + "loss": 0.2865, + "step": 23154 + }, + { + "epoch": 0.5427896646096685, + "grad_norm": 2.620264768600464, + "learning_rate": 8.663006593655731e-05, + "loss": 0.3477, + "step": 23155 + }, + { + "epoch": 0.5428131061844734, + "grad_norm": 2.7582032680511475, + "learning_rate": 8.662276594880935e-05, + "loss": 0.2085, + "step": 23156 + }, + { + "epoch": 0.5428365477592785, + "grad_norm": 0.7844973206520081, + "learning_rate": 8.661546603364665e-05, + "loss": 0.0646, + "step": 23157 + }, + { + "epoch": 0.5428599893340834, + "grad_norm": 3.806513547897339, + "learning_rate": 8.660816619110876e-05, + "loss": 0.242, + "step": 23158 + }, + { + "epoch": 0.5428834309088885, + "grad_norm": 1.799408197402954, + "learning_rate": 8.660086642123537e-05, + "loss": 0.2011, + "step": 23159 + }, + { + "epoch": 0.5429068724836934, + "grad_norm": 3.9951250553131104, + "learning_rate": 8.659356672406606e-05, + "loss": 0.298, + "step": 23160 + }, + { + "epoch": 0.5429303140584985, + "grad_norm": 0.6924543976783752, + "learning_rate": 8.658626709964041e-05, + "loss": 0.0327, + "step": 23161 + }, + { + "epoch": 0.5429537556333034, + "grad_norm": 1.7564141750335693, + "learning_rate": 8.657896754799807e-05, + "loss": 0.2055, + "step": 23162 + }, + { + "epoch": 0.5429771972081084, + "grad_norm": 0.9990464448928833, + "learning_rate": 8.657166806917863e-05, + "loss": 0.0957, + "step": 23163 + }, + { + "epoch": 0.5430006387829134, + "grad_norm": 3.8169713020324707, + "learning_rate": 8.656436866322167e-05, + "loss": 0.3222, + "step": 23164 + }, + { + "epoch": 0.5430240803577184, + "grad_norm": 2.4650580883026123, + "learning_rate": 8.65570693301668e-05, + "loss": 0.2726, + "step": 23165 + }, + { + "epoch": 0.5430475219325235, + "grad_norm": 3.4994797706604004, + "learning_rate": 8.654977007005372e-05, + "loss": 0.2002, + "step": 23166 + }, + { + "epoch": 0.5430709635073284, + "grad_norm": 3.0764126777648926, + "learning_rate": 8.654247088292193e-05, + "loss": 0.3438, + "step": 23167 + }, + { + "epoch": 0.5430944050821335, + "grad_norm": 5.273629188537598, + "learning_rate": 8.653517176881107e-05, + "loss": 0.5626, + "step": 23168 + }, + { + "epoch": 0.5431178466569384, + "grad_norm": 0.6154733300209045, + "learning_rate": 8.652787272776077e-05, + "loss": 0.0327, + "step": 23169 + }, + { + "epoch": 0.5431412882317435, + "grad_norm": 2.6880624294281006, + "learning_rate": 8.652057375981061e-05, + "loss": 0.2646, + "step": 23170 + }, + { + "epoch": 0.5431647298065484, + "grad_norm": 4.053452491760254, + "learning_rate": 8.651327486500016e-05, + "loss": 0.3624, + "step": 23171 + }, + { + "epoch": 0.5431881713813534, + "grad_norm": 1.1652824878692627, + "learning_rate": 8.650597604336909e-05, + "loss": 0.106, + "step": 23172 + }, + { + "epoch": 0.5432116129561584, + "grad_norm": 2.9661998748779297, + "learning_rate": 8.649867729495696e-05, + "loss": 0.2178, + "step": 23173 + }, + { + "epoch": 0.5432350545309634, + "grad_norm": 1.8750556707382202, + "learning_rate": 8.649137861980339e-05, + "loss": 0.0838, + "step": 23174 + }, + { + "epoch": 0.5432584961057684, + "grad_norm": 0.41082635521888733, + "learning_rate": 8.6484080017948e-05, + "loss": 0.0281, + "step": 23175 + }, + { + "epoch": 0.5432819376805734, + "grad_norm": 2.8149056434631348, + "learning_rate": 8.647678148943034e-05, + "loss": 0.1894, + "step": 23176 + }, + { + "epoch": 0.5433053792553784, + "grad_norm": 3.3067941665649414, + "learning_rate": 8.646948303429006e-05, + "loss": 0.2644, + "step": 23177 + }, + { + "epoch": 0.5433288208301834, + "grad_norm": 1.408237099647522, + "learning_rate": 8.646218465256673e-05, + "loss": 0.132, + "step": 23178 + }, + { + "epoch": 0.5433522624049884, + "grad_norm": 5.135115623474121, + "learning_rate": 8.645488634429998e-05, + "loss": 0.8027, + "step": 23179 + }, + { + "epoch": 0.5433757039797934, + "grad_norm": 0.4397585093975067, + "learning_rate": 8.644758810952937e-05, + "loss": 0.0192, + "step": 23180 + }, + { + "epoch": 0.5433991455545983, + "grad_norm": 3.8765785694122314, + "learning_rate": 8.644028994829457e-05, + "loss": 0.1806, + "step": 23181 + }, + { + "epoch": 0.5434225871294034, + "grad_norm": 1.9948451519012451, + "learning_rate": 8.643299186063514e-05, + "loss": 1.5822, + "step": 23182 + }, + { + "epoch": 0.5434460287042083, + "grad_norm": 1.312089204788208, + "learning_rate": 8.642569384659066e-05, + "loss": 0.1106, + "step": 23183 + }, + { + "epoch": 0.5434694702790134, + "grad_norm": 3.8650598526000977, + "learning_rate": 8.641839590620074e-05, + "loss": 0.1951, + "step": 23184 + }, + { + "epoch": 0.5434929118538183, + "grad_norm": 3.7053492069244385, + "learning_rate": 8.6411098039505e-05, + "loss": 0.4821, + "step": 23185 + }, + { + "epoch": 0.5435163534286234, + "grad_norm": 0.46855074167251587, + "learning_rate": 8.640380024654301e-05, + "loss": 0.0122, + "step": 23186 + }, + { + "epoch": 0.5435397950034283, + "grad_norm": 3.65781307220459, + "learning_rate": 8.639650252735441e-05, + "loss": 0.3484, + "step": 23187 + }, + { + "epoch": 0.5435632365782334, + "grad_norm": 3.1655588150024414, + "learning_rate": 8.638920488197875e-05, + "loss": 0.2795, + "step": 23188 + }, + { + "epoch": 0.5435866781530383, + "grad_norm": 1.4693408012390137, + "learning_rate": 8.63819073104556e-05, + "loss": 0.1315, + "step": 23189 + }, + { + "epoch": 0.5436101197278433, + "grad_norm": 0.6650733947753906, + "learning_rate": 8.637460981282465e-05, + "loss": 0.0518, + "step": 23190 + }, + { + "epoch": 0.5436335613026483, + "grad_norm": 2.303560972213745, + "learning_rate": 8.636731238912545e-05, + "loss": 1.4677, + "step": 23191 + }, + { + "epoch": 0.5436570028774533, + "grad_norm": 0.9554063081741333, + "learning_rate": 8.636001503939762e-05, + "loss": 0.0725, + "step": 23192 + }, + { + "epoch": 0.5436804444522583, + "grad_norm": 4.437839984893799, + "learning_rate": 8.635271776368069e-05, + "loss": 0.2393, + "step": 23193 + }, + { + "epoch": 0.5437038860270633, + "grad_norm": 3.3692784309387207, + "learning_rate": 8.634542056201432e-05, + "loss": 0.3469, + "step": 23194 + }, + { + "epoch": 0.5437273276018683, + "grad_norm": 2.966169595718384, + "learning_rate": 8.633812343443808e-05, + "loss": 0.3344, + "step": 23195 + }, + { + "epoch": 0.5437507691766733, + "grad_norm": 3.1543996334075928, + "learning_rate": 8.633082638099157e-05, + "loss": 1.9026, + "step": 23196 + }, + { + "epoch": 0.5437742107514782, + "grad_norm": 4.787632942199707, + "learning_rate": 8.632352940171438e-05, + "loss": 0.528, + "step": 23197 + }, + { + "epoch": 0.5437976523262833, + "grad_norm": 2.212254762649536, + "learning_rate": 8.631623249664611e-05, + "loss": 0.1057, + "step": 23198 + }, + { + "epoch": 0.5438210939010882, + "grad_norm": 3.4265823364257812, + "learning_rate": 8.630893566582631e-05, + "loss": 0.4177, + "step": 23199 + }, + { + "epoch": 0.5438445354758933, + "grad_norm": 2.757397174835205, + "learning_rate": 8.630163890929466e-05, + "loss": 0.2328, + "step": 23200 + }, + { + "epoch": 0.5438679770506982, + "grad_norm": 4.6886138916015625, + "learning_rate": 8.629434222709064e-05, + "loss": 0.3846, + "step": 23201 + }, + { + "epoch": 0.5438914186255033, + "grad_norm": 1.1717753410339355, + "learning_rate": 8.628704561925395e-05, + "loss": 0.0976, + "step": 23202 + }, + { + "epoch": 0.5439148602003082, + "grad_norm": 2.2368416786193848, + "learning_rate": 8.627974908582411e-05, + "loss": 0.1882, + "step": 23203 + }, + { + "epoch": 0.5439383017751133, + "grad_norm": 1.7814103364944458, + "learning_rate": 8.627245262684077e-05, + "loss": 0.1343, + "step": 23204 + }, + { + "epoch": 0.5439617433499182, + "grad_norm": 2.3654255867004395, + "learning_rate": 8.626515624234349e-05, + "loss": 0.1852, + "step": 23205 + }, + { + "epoch": 0.5439851849247233, + "grad_norm": 2.6011409759521484, + "learning_rate": 8.625785993237184e-05, + "loss": 0.2574, + "step": 23206 + }, + { + "epoch": 0.5440086264995282, + "grad_norm": 6.4795637130737305, + "learning_rate": 8.625056369696543e-05, + "loss": 0.6965, + "step": 23207 + }, + { + "epoch": 0.5440320680743332, + "grad_norm": 3.710026502609253, + "learning_rate": 8.624326753616386e-05, + "loss": 0.2667, + "step": 23208 + }, + { + "epoch": 0.5440555096491382, + "grad_norm": 1.407680630683899, + "learning_rate": 8.623597145000668e-05, + "loss": 0.0782, + "step": 23209 + }, + { + "epoch": 0.5440789512239432, + "grad_norm": 0.48525238037109375, + "learning_rate": 8.622867543853353e-05, + "loss": 0.0608, + "step": 23210 + }, + { + "epoch": 0.5441023927987482, + "grad_norm": 4.918442249298096, + "learning_rate": 8.622137950178395e-05, + "loss": 0.6941, + "step": 23211 + }, + { + "epoch": 0.5441258343735532, + "grad_norm": 4.779699802398682, + "learning_rate": 8.621408363979759e-05, + "loss": 0.1232, + "step": 23212 + }, + { + "epoch": 0.5441492759483582, + "grad_norm": 0.6532396078109741, + "learning_rate": 8.620678785261394e-05, + "loss": 0.0716, + "step": 23213 + }, + { + "epoch": 0.5441727175231632, + "grad_norm": 2.016327381134033, + "learning_rate": 8.619949214027269e-05, + "loss": 1.463, + "step": 23214 + }, + { + "epoch": 0.5441961590979681, + "grad_norm": 3.0762436389923096, + "learning_rate": 8.619219650281339e-05, + "loss": 0.3673, + "step": 23215 + }, + { + "epoch": 0.5442196006727732, + "grad_norm": 3.50740647315979, + "learning_rate": 8.618490094027559e-05, + "loss": 0.3719, + "step": 23216 + }, + { + "epoch": 0.5442430422475782, + "grad_norm": 1.0225396156311035, + "learning_rate": 8.617760545269893e-05, + "loss": 0.0727, + "step": 23217 + }, + { + "epoch": 0.5442664838223832, + "grad_norm": 3.874880790710449, + "learning_rate": 8.617031004012295e-05, + "loss": 0.3778, + "step": 23218 + }, + { + "epoch": 0.5442899253971882, + "grad_norm": 0.6393011212348938, + "learning_rate": 8.61630147025873e-05, + "loss": 0.0652, + "step": 23219 + }, + { + "epoch": 0.5443133669719932, + "grad_norm": 2.161487579345703, + "learning_rate": 8.61557194401315e-05, + "loss": 1.9364, + "step": 23220 + }, + { + "epoch": 0.5443368085467982, + "grad_norm": 4.517380237579346, + "learning_rate": 8.614842425279513e-05, + "loss": 0.4756, + "step": 23221 + }, + { + "epoch": 0.5443602501216032, + "grad_norm": 2.1456634998321533, + "learning_rate": 8.614112914061782e-05, + "loss": 0.1936, + "step": 23222 + }, + { + "epoch": 0.5443836916964082, + "grad_norm": 2.2635080814361572, + "learning_rate": 8.613383410363915e-05, + "loss": 1.1909, + "step": 23223 + }, + { + "epoch": 0.5444071332712132, + "grad_norm": 2.6498730182647705, + "learning_rate": 8.612653914189866e-05, + "loss": 0.1626, + "step": 23224 + }, + { + "epoch": 0.5444305748460182, + "grad_norm": 2.927337169647217, + "learning_rate": 8.611924425543597e-05, + "loss": 0.4221, + "step": 23225 + }, + { + "epoch": 0.5444540164208231, + "grad_norm": 1.8407375812530518, + "learning_rate": 8.611194944429062e-05, + "loss": 0.2078, + "step": 23226 + }, + { + "epoch": 0.5444774579956282, + "grad_norm": 1.439581274986267, + "learning_rate": 8.610465470850226e-05, + "loss": 0.1511, + "step": 23227 + }, + { + "epoch": 0.5445008995704331, + "grad_norm": 2.418083906173706, + "learning_rate": 8.609736004811041e-05, + "loss": 1.212, + "step": 23228 + }, + { + "epoch": 0.5445243411452382, + "grad_norm": 2.6292853355407715, + "learning_rate": 8.60900654631547e-05, + "loss": 0.2073, + "step": 23229 + }, + { + "epoch": 0.5445477827200431, + "grad_norm": 3.7393720149993896, + "learning_rate": 8.608277095367469e-05, + "loss": 0.2227, + "step": 23230 + }, + { + "epoch": 0.5445712242948482, + "grad_norm": 2.0515859127044678, + "learning_rate": 8.607547651970993e-05, + "loss": 1.0517, + "step": 23231 + }, + { + "epoch": 0.5445946658696531, + "grad_norm": 3.491966485977173, + "learning_rate": 8.606818216130003e-05, + "loss": 0.3916, + "step": 23232 + }, + { + "epoch": 0.5446181074444582, + "grad_norm": 1.8036478757858276, + "learning_rate": 8.60608878784846e-05, + "loss": 0.9454, + "step": 23233 + }, + { + "epoch": 0.5446415490192631, + "grad_norm": 3.5383124351501465, + "learning_rate": 8.605359367130314e-05, + "loss": 0.27, + "step": 23234 + }, + { + "epoch": 0.5446649905940681, + "grad_norm": 1.1292438507080078, + "learning_rate": 8.604629953979531e-05, + "loss": 0.0649, + "step": 23235 + }, + { + "epoch": 0.5446884321688731, + "grad_norm": 2.5538177490234375, + "learning_rate": 8.603900548400063e-05, + "loss": 0.2404, + "step": 23236 + }, + { + "epoch": 0.5447118737436781, + "grad_norm": 1.0962979793548584, + "learning_rate": 8.603171150395869e-05, + "loss": 0.0949, + "step": 23237 + }, + { + "epoch": 0.5447353153184831, + "grad_norm": 2.4399616718292236, + "learning_rate": 8.602441759970905e-05, + "loss": 0.1842, + "step": 23238 + }, + { + "epoch": 0.5447587568932881, + "grad_norm": 3.9623305797576904, + "learning_rate": 8.601712377129137e-05, + "loss": 0.4338, + "step": 23239 + }, + { + "epoch": 0.5447821984680931, + "grad_norm": 3.502336025238037, + "learning_rate": 8.600983001874515e-05, + "loss": 0.2935, + "step": 23240 + }, + { + "epoch": 0.5448056400428981, + "grad_norm": 2.414201259613037, + "learning_rate": 8.600253634210997e-05, + "loss": 1.2306, + "step": 23241 + }, + { + "epoch": 0.544829081617703, + "grad_norm": 2.7211129665374756, + "learning_rate": 8.599524274142545e-05, + "loss": 0.4193, + "step": 23242 + }, + { + "epoch": 0.5448525231925081, + "grad_norm": 3.2319562435150146, + "learning_rate": 8.598794921673114e-05, + "loss": 0.2138, + "step": 23243 + }, + { + "epoch": 0.544875964767313, + "grad_norm": 2.19266676902771, + "learning_rate": 8.598065576806657e-05, + "loss": 0.1705, + "step": 23244 + }, + { + "epoch": 0.5448994063421181, + "grad_norm": 1.8803786039352417, + "learning_rate": 8.597336239547139e-05, + "loss": 0.1511, + "step": 23245 + }, + { + "epoch": 0.544922847916923, + "grad_norm": 2.9766178131103516, + "learning_rate": 8.596606909898514e-05, + "loss": 0.1896, + "step": 23246 + }, + { + "epoch": 0.5449462894917281, + "grad_norm": 3.123656749725342, + "learning_rate": 8.595877587864737e-05, + "loss": 0.3663, + "step": 23247 + }, + { + "epoch": 0.544969731066533, + "grad_norm": 1.7849584817886353, + "learning_rate": 8.59514827344977e-05, + "loss": 0.1999, + "step": 23248 + }, + { + "epoch": 0.5449931726413381, + "grad_norm": 3.5416152477264404, + "learning_rate": 8.594418966657567e-05, + "loss": 0.1514, + "step": 23249 + }, + { + "epoch": 0.545016614216143, + "grad_norm": 2.905099391937256, + "learning_rate": 8.593689667492084e-05, + "loss": 0.1859, + "step": 23250 + }, + { + "epoch": 0.545040055790948, + "grad_norm": 2.6754696369171143, + "learning_rate": 8.592960375957282e-05, + "loss": 0.097, + "step": 23251 + }, + { + "epoch": 0.545063497365753, + "grad_norm": 1.3818563222885132, + "learning_rate": 8.592231092057119e-05, + "loss": 0.0916, + "step": 23252 + }, + { + "epoch": 0.545086938940558, + "grad_norm": 4.540874481201172, + "learning_rate": 8.591501815795547e-05, + "loss": 0.3747, + "step": 23253 + }, + { + "epoch": 0.545110380515363, + "grad_norm": 1.924838900566101, + "learning_rate": 8.590772547176527e-05, + "loss": 0.1409, + "step": 23254 + }, + { + "epoch": 0.545133822090168, + "grad_norm": 2.3614754676818848, + "learning_rate": 8.590043286204014e-05, + "loss": 0.0889, + "step": 23255 + }, + { + "epoch": 0.545157263664973, + "grad_norm": 2.320852041244507, + "learning_rate": 8.589314032881965e-05, + "loss": 0.1441, + "step": 23256 + }, + { + "epoch": 0.545180705239778, + "grad_norm": 4.265422821044922, + "learning_rate": 8.588584787214341e-05, + "loss": 0.3531, + "step": 23257 + }, + { + "epoch": 0.545204146814583, + "grad_norm": 4.908668041229248, + "learning_rate": 8.587855549205094e-05, + "loss": 0.4167, + "step": 23258 + }, + { + "epoch": 0.545227588389388, + "grad_norm": 2.0503151416778564, + "learning_rate": 8.587126318858181e-05, + "loss": 0.1507, + "step": 23259 + }, + { + "epoch": 0.545251029964193, + "grad_norm": 3.7712318897247314, + "learning_rate": 8.586397096177562e-05, + "loss": 0.3001, + "step": 23260 + }, + { + "epoch": 0.545274471538998, + "grad_norm": 2.540146827697754, + "learning_rate": 8.585667881167192e-05, + "loss": 0.161, + "step": 23261 + }, + { + "epoch": 0.5452979131138029, + "grad_norm": 3.0340559482574463, + "learning_rate": 8.584938673831025e-05, + "loss": 0.175, + "step": 23262 + }, + { + "epoch": 0.545321354688608, + "grad_norm": 3.1622297763824463, + "learning_rate": 8.584209474173022e-05, + "loss": 0.2392, + "step": 23263 + }, + { + "epoch": 0.5453447962634129, + "grad_norm": 3.533607006072998, + "learning_rate": 8.58348028219714e-05, + "loss": 0.2184, + "step": 23264 + }, + { + "epoch": 0.545368237838218, + "grad_norm": 2.0700666904449463, + "learning_rate": 8.582751097907333e-05, + "loss": 0.1092, + "step": 23265 + }, + { + "epoch": 0.5453916794130229, + "grad_norm": 2.830000877380371, + "learning_rate": 8.582021921307556e-05, + "loss": 0.294, + "step": 23266 + }, + { + "epoch": 0.545415120987828, + "grad_norm": 2.729417324066162, + "learning_rate": 8.581292752401773e-05, + "loss": 0.0844, + "step": 23267 + }, + { + "epoch": 0.545438562562633, + "grad_norm": 3.1385045051574707, + "learning_rate": 8.580563591193933e-05, + "loss": 0.2739, + "step": 23268 + }, + { + "epoch": 0.545462004137438, + "grad_norm": 2.7924420833587646, + "learning_rate": 8.579834437687994e-05, + "loss": 0.2318, + "step": 23269 + }, + { + "epoch": 0.545485445712243, + "grad_norm": 3.7938339710235596, + "learning_rate": 8.579105291887915e-05, + "loss": 0.2101, + "step": 23270 + }, + { + "epoch": 0.5455088872870479, + "grad_norm": 4.585341930389404, + "learning_rate": 8.57837615379765e-05, + "loss": 0.4406, + "step": 23271 + }, + { + "epoch": 0.545532328861853, + "grad_norm": 0.9587781429290771, + "learning_rate": 8.577647023421154e-05, + "loss": 0.1009, + "step": 23272 + }, + { + "epoch": 0.5455557704366579, + "grad_norm": 4.261502265930176, + "learning_rate": 8.576917900762387e-05, + "loss": 0.343, + "step": 23273 + }, + { + "epoch": 0.545579212011463, + "grad_norm": 2.2432923316955566, + "learning_rate": 8.5761887858253e-05, + "loss": 1.4452, + "step": 23274 + }, + { + "epoch": 0.5456026535862679, + "grad_norm": 3.7363789081573486, + "learning_rate": 8.575459678613857e-05, + "loss": 0.3778, + "step": 23275 + }, + { + "epoch": 0.545626095161073, + "grad_norm": 3.601510524749756, + "learning_rate": 8.574730579132006e-05, + "loss": 0.2141, + "step": 23276 + }, + { + "epoch": 0.5456495367358779, + "grad_norm": 1.2755180597305298, + "learning_rate": 8.574001487383711e-05, + "loss": 0.1048, + "step": 23277 + }, + { + "epoch": 0.545672978310683, + "grad_norm": 1.307515025138855, + "learning_rate": 8.573272403372924e-05, + "loss": 0.6229, + "step": 23278 + }, + { + "epoch": 0.5456964198854879, + "grad_norm": 3.3419690132141113, + "learning_rate": 8.572543327103598e-05, + "loss": 0.2152, + "step": 23279 + }, + { + "epoch": 0.5457198614602929, + "grad_norm": 0.516363263130188, + "learning_rate": 8.571814258579695e-05, + "loss": 0.0381, + "step": 23280 + }, + { + "epoch": 0.5457433030350979, + "grad_norm": 3.1863551139831543, + "learning_rate": 8.571085197805167e-05, + "loss": 0.1616, + "step": 23281 + }, + { + "epoch": 0.5457667446099029, + "grad_norm": 0.6236674189567566, + "learning_rate": 8.570356144783969e-05, + "loss": 0.0448, + "step": 23282 + }, + { + "epoch": 0.5457901861847079, + "grad_norm": 4.504897117614746, + "learning_rate": 8.569627099520062e-05, + "loss": 0.4871, + "step": 23283 + }, + { + "epoch": 0.5458136277595129, + "grad_norm": 3.0654377937316895, + "learning_rate": 8.568898062017398e-05, + "loss": 0.428, + "step": 23284 + }, + { + "epoch": 0.5458370693343179, + "grad_norm": 4.071808338165283, + "learning_rate": 8.568169032279931e-05, + "loss": 0.58, + "step": 23285 + }, + { + "epoch": 0.5458605109091229, + "grad_norm": 4.022036552429199, + "learning_rate": 8.567440010311617e-05, + "loss": 0.4108, + "step": 23286 + }, + { + "epoch": 0.5458839524839278, + "grad_norm": 4.309418678283691, + "learning_rate": 8.566710996116419e-05, + "loss": 0.3333, + "step": 23287 + }, + { + "epoch": 0.5459073940587329, + "grad_norm": 2.7091708183288574, + "learning_rate": 8.565981989698287e-05, + "loss": 0.1498, + "step": 23288 + }, + { + "epoch": 0.5459308356335378, + "grad_norm": 2.7048914432525635, + "learning_rate": 8.565252991061173e-05, + "loss": 0.2296, + "step": 23289 + }, + { + "epoch": 0.5459542772083429, + "grad_norm": 5.307757377624512, + "learning_rate": 8.56452400020904e-05, + "loss": 0.3109, + "step": 23290 + }, + { + "epoch": 0.5459777187831478, + "grad_norm": 0.9992303252220154, + "learning_rate": 8.56379501714584e-05, + "loss": 0.0481, + "step": 23291 + }, + { + "epoch": 0.5460011603579529, + "grad_norm": 3.5816597938537598, + "learning_rate": 8.563066041875526e-05, + "loss": 0.5016, + "step": 23292 + }, + { + "epoch": 0.5460246019327578, + "grad_norm": 6.928346633911133, + "learning_rate": 8.56233707440206e-05, + "loss": 0.7757, + "step": 23293 + }, + { + "epoch": 0.5460480435075629, + "grad_norm": 1.91426420211792, + "learning_rate": 8.561608114729391e-05, + "loss": 1.2356, + "step": 23294 + }, + { + "epoch": 0.5460714850823678, + "grad_norm": 2.002763271331787, + "learning_rate": 8.560879162861478e-05, + "loss": 0.2427, + "step": 23295 + }, + { + "epoch": 0.5460949266571729, + "grad_norm": 0.4409441351890564, + "learning_rate": 8.560150218802276e-05, + "loss": 0.0368, + "step": 23296 + }, + { + "epoch": 0.5461183682319778, + "grad_norm": 3.48327898979187, + "learning_rate": 8.559421282555737e-05, + "loss": 0.1854, + "step": 23297 + }, + { + "epoch": 0.5461418098067828, + "grad_norm": 2.0737109184265137, + "learning_rate": 8.558692354125815e-05, + "loss": 1.8021, + "step": 23298 + }, + { + "epoch": 0.5461652513815878, + "grad_norm": 2.722130060195923, + "learning_rate": 8.557963433516475e-05, + "loss": 1.2835, + "step": 23299 + }, + { + "epoch": 0.5461886929563928, + "grad_norm": 0.9260198473930359, + "learning_rate": 8.557234520731666e-05, + "loss": 0.0762, + "step": 23300 + }, + { + "epoch": 0.5462121345311978, + "grad_norm": 1.5657261610031128, + "learning_rate": 8.556505615775339e-05, + "loss": 0.1163, + "step": 23301 + }, + { + "epoch": 0.5462355761060028, + "grad_norm": 4.278341770172119, + "learning_rate": 8.555776718651456e-05, + "loss": 0.5303, + "step": 23302 + }, + { + "epoch": 0.5462590176808078, + "grad_norm": 1.263253092765808, + "learning_rate": 8.555047829363969e-05, + "loss": 0.13, + "step": 23303 + }, + { + "epoch": 0.5462824592556128, + "grad_norm": 1.2165541648864746, + "learning_rate": 8.554318947916832e-05, + "loss": 0.1276, + "step": 23304 + }, + { + "epoch": 0.5463059008304177, + "grad_norm": 3.3728601932525635, + "learning_rate": 8.553590074314003e-05, + "loss": 0.205, + "step": 23305 + }, + { + "epoch": 0.5463293424052228, + "grad_norm": 4.231003761291504, + "learning_rate": 8.552861208559434e-05, + "loss": 0.2822, + "step": 23306 + }, + { + "epoch": 0.5463527839800277, + "grad_norm": 1.395871877670288, + "learning_rate": 8.55213235065708e-05, + "loss": 0.143, + "step": 23307 + }, + { + "epoch": 0.5463762255548328, + "grad_norm": 1.6801245212554932, + "learning_rate": 8.551403500610897e-05, + "loss": 0.1791, + "step": 23308 + }, + { + "epoch": 0.5463996671296377, + "grad_norm": 2.996304750442505, + "learning_rate": 8.550674658424841e-05, + "loss": 0.4267, + "step": 23309 + }, + { + "epoch": 0.5464231087044428, + "grad_norm": 1.7237132787704468, + "learning_rate": 8.549945824102862e-05, + "loss": 0.0476, + "step": 23310 + }, + { + "epoch": 0.5464465502792477, + "grad_norm": 2.535630464553833, + "learning_rate": 8.549216997648916e-05, + "loss": 1.0594, + "step": 23311 + }, + { + "epoch": 0.5464699918540528, + "grad_norm": 3.472200393676758, + "learning_rate": 8.548488179066963e-05, + "loss": 0.3362, + "step": 23312 + }, + { + "epoch": 0.5464934334288577, + "grad_norm": 0.5810859799385071, + "learning_rate": 8.547759368360955e-05, + "loss": 0.0759, + "step": 23313 + }, + { + "epoch": 0.5465168750036627, + "grad_norm": 3.6431849002838135, + "learning_rate": 8.547030565534841e-05, + "loss": 0.2276, + "step": 23314 + }, + { + "epoch": 0.5465403165784677, + "grad_norm": 2.8144984245300293, + "learning_rate": 8.546301770592586e-05, + "loss": 0.3287, + "step": 23315 + }, + { + "epoch": 0.5465637581532727, + "grad_norm": 1.355686068534851, + "learning_rate": 8.545572983538135e-05, + "loss": 0.0595, + "step": 23316 + }, + { + "epoch": 0.5465871997280777, + "grad_norm": 1.1113148927688599, + "learning_rate": 8.544844204375445e-05, + "loss": 0.0994, + "step": 23317 + }, + { + "epoch": 0.5466106413028827, + "grad_norm": 2.347169876098633, + "learning_rate": 8.544115433108473e-05, + "loss": 0.1629, + "step": 23318 + }, + { + "epoch": 0.5466340828776878, + "grad_norm": 3.248063564300537, + "learning_rate": 8.543386669741172e-05, + "loss": 1.5297, + "step": 23319 + }, + { + "epoch": 0.5466575244524927, + "grad_norm": 2.7238993644714355, + "learning_rate": 8.542657914277495e-05, + "loss": 1.8115, + "step": 23320 + }, + { + "epoch": 0.5466809660272978, + "grad_norm": 3.632686138153076, + "learning_rate": 8.541929166721397e-05, + "loss": 0.2888, + "step": 23321 + }, + { + "epoch": 0.5467044076021027, + "grad_norm": 2.274590492248535, + "learning_rate": 8.541200427076832e-05, + "loss": 0.1067, + "step": 23322 + }, + { + "epoch": 0.5467278491769078, + "grad_norm": 5.438400745391846, + "learning_rate": 8.540471695347754e-05, + "loss": 0.6093, + "step": 23323 + }, + { + "epoch": 0.5467512907517127, + "grad_norm": 4.142704963684082, + "learning_rate": 8.539742971538117e-05, + "loss": 0.4526, + "step": 23324 + }, + { + "epoch": 0.5467747323265177, + "grad_norm": 2.791360855102539, + "learning_rate": 8.539014255651878e-05, + "loss": 1.2798, + "step": 23325 + }, + { + "epoch": 0.5467981739013227, + "grad_norm": 4.691077709197998, + "learning_rate": 8.538285547692989e-05, + "loss": 0.3543, + "step": 23326 + }, + { + "epoch": 0.5468216154761277, + "grad_norm": 2.2732512950897217, + "learning_rate": 8.537556847665402e-05, + "loss": 0.148, + "step": 23327 + }, + { + "epoch": 0.5468450570509327, + "grad_norm": 2.3729724884033203, + "learning_rate": 8.536828155573074e-05, + "loss": 0.1164, + "step": 23328 + }, + { + "epoch": 0.5468684986257377, + "grad_norm": 1.1715742349624634, + "learning_rate": 8.536099471419955e-05, + "loss": 0.1244, + "step": 23329 + }, + { + "epoch": 0.5468919402005427, + "grad_norm": 3.7952804565429688, + "learning_rate": 8.535370795210005e-05, + "loss": 0.4063, + "step": 23330 + }, + { + "epoch": 0.5469153817753477, + "grad_norm": 2.313072443008423, + "learning_rate": 8.534642126947175e-05, + "loss": 0.148, + "step": 23331 + }, + { + "epoch": 0.5469388233501526, + "grad_norm": 2.7114813327789307, + "learning_rate": 8.533913466635414e-05, + "loss": 0.27, + "step": 23332 + }, + { + "epoch": 0.5469622649249577, + "grad_norm": 3.964127779006958, + "learning_rate": 8.533184814278682e-05, + "loss": 0.4336, + "step": 23333 + }, + { + "epoch": 0.5469857064997626, + "grad_norm": 0.48238271474838257, + "learning_rate": 8.53245616988093e-05, + "loss": 0.0384, + "step": 23334 + }, + { + "epoch": 0.5470091480745677, + "grad_norm": 2.7056353092193604, + "learning_rate": 8.531727533446109e-05, + "loss": 0.2554, + "step": 23335 + }, + { + "epoch": 0.5470325896493726, + "grad_norm": 3.330314874649048, + "learning_rate": 8.530998904978178e-05, + "loss": 0.2204, + "step": 23336 + }, + { + "epoch": 0.5470560312241777, + "grad_norm": 3.695756435394287, + "learning_rate": 8.53027028448109e-05, + "loss": 0.1744, + "step": 23337 + }, + { + "epoch": 0.5470794727989826, + "grad_norm": 1.2876518964767456, + "learning_rate": 8.529541671958796e-05, + "loss": 0.1129, + "step": 23338 + }, + { + "epoch": 0.5471029143737877, + "grad_norm": 2.3724095821380615, + "learning_rate": 8.528813067415249e-05, + "loss": 0.156, + "step": 23339 + }, + { + "epoch": 0.5471263559485926, + "grad_norm": 2.9917988777160645, + "learning_rate": 8.528084470854405e-05, + "loss": 0.2343, + "step": 23340 + }, + { + "epoch": 0.5471497975233977, + "grad_norm": 3.4807143211364746, + "learning_rate": 8.527355882280217e-05, + "loss": 0.3627, + "step": 23341 + }, + { + "epoch": 0.5471732390982026, + "grad_norm": 0.6694411039352417, + "learning_rate": 8.526627301696634e-05, + "loss": 0.045, + "step": 23342 + }, + { + "epoch": 0.5471966806730076, + "grad_norm": 1.884581208229065, + "learning_rate": 8.525898729107617e-05, + "loss": 0.1253, + "step": 23343 + }, + { + "epoch": 0.5472201222478126, + "grad_norm": 2.5288233757019043, + "learning_rate": 8.525170164517112e-05, + "loss": 0.3002, + "step": 23344 + }, + { + "epoch": 0.5472435638226176, + "grad_norm": 4.071664810180664, + "learning_rate": 8.524441607929074e-05, + "loss": 0.3871, + "step": 23345 + }, + { + "epoch": 0.5472670053974226, + "grad_norm": 3.41357421875, + "learning_rate": 8.523713059347462e-05, + "loss": 0.21, + "step": 23346 + }, + { + "epoch": 0.5472904469722276, + "grad_norm": 4.574234962463379, + "learning_rate": 8.522984518776216e-05, + "loss": 0.3035, + "step": 23347 + }, + { + "epoch": 0.5473138885470326, + "grad_norm": 4.071958065032959, + "learning_rate": 8.522255986219306e-05, + "loss": 0.5086, + "step": 23348 + }, + { + "epoch": 0.5473373301218376, + "grad_norm": 1.6818913221359253, + "learning_rate": 8.521527461680671e-05, + "loss": 0.1303, + "step": 23349 + }, + { + "epoch": 0.5473607716966425, + "grad_norm": 3.901618003845215, + "learning_rate": 8.520798945164274e-05, + "loss": 0.1738, + "step": 23350 + }, + { + "epoch": 0.5473842132714476, + "grad_norm": 1.5966264009475708, + "learning_rate": 8.520070436674063e-05, + "loss": 0.0984, + "step": 23351 + }, + { + "epoch": 0.5474076548462525, + "grad_norm": 0.775180995464325, + "learning_rate": 8.51934193621399e-05, + "loss": 0.0929, + "step": 23352 + }, + { + "epoch": 0.5474310964210576, + "grad_norm": 5.326017379760742, + "learning_rate": 8.518613443788008e-05, + "loss": 0.3051, + "step": 23353 + }, + { + "epoch": 0.5474545379958625, + "grad_norm": 5.407330513000488, + "learning_rate": 8.517884959400075e-05, + "loss": 0.4998, + "step": 23354 + }, + { + "epoch": 0.5474779795706676, + "grad_norm": 1.3249926567077637, + "learning_rate": 8.517156483054136e-05, + "loss": 0.2575, + "step": 23355 + }, + { + "epoch": 0.5475014211454725, + "grad_norm": 1.1830741167068481, + "learning_rate": 8.51642801475415e-05, + "loss": 0.1614, + "step": 23356 + }, + { + "epoch": 0.5475248627202776, + "grad_norm": 1.676851511001587, + "learning_rate": 8.515699554504066e-05, + "loss": 0.2193, + "step": 23357 + }, + { + "epoch": 0.5475483042950825, + "grad_norm": 7.906621932983398, + "learning_rate": 8.514971102307838e-05, + "loss": 0.5795, + "step": 23358 + }, + { + "epoch": 0.5475717458698875, + "grad_norm": 3.1914641857147217, + "learning_rate": 8.514242658169415e-05, + "loss": 0.2719, + "step": 23359 + }, + { + "epoch": 0.5475951874446925, + "grad_norm": 5.430370807647705, + "learning_rate": 8.513514222092757e-05, + "loss": 0.4567, + "step": 23360 + }, + { + "epoch": 0.5476186290194975, + "grad_norm": 3.15952467918396, + "learning_rate": 8.512785794081815e-05, + "loss": 0.4452, + "step": 23361 + }, + { + "epoch": 0.5476420705943025, + "grad_norm": 4.42260217666626, + "learning_rate": 8.512057374140535e-05, + "loss": 0.3875, + "step": 23362 + }, + { + "epoch": 0.5476655121691075, + "grad_norm": 3.7444512844085693, + "learning_rate": 8.511328962272876e-05, + "loss": 0.4716, + "step": 23363 + }, + { + "epoch": 0.5476889537439125, + "grad_norm": 4.033120632171631, + "learning_rate": 8.510600558482788e-05, + "loss": 0.3001, + "step": 23364 + }, + { + "epoch": 0.5477123953187175, + "grad_norm": 5.0383830070495605, + "learning_rate": 8.509872162774221e-05, + "loss": 0.5256, + "step": 23365 + }, + { + "epoch": 0.5477358368935225, + "grad_norm": 3.281590700149536, + "learning_rate": 8.509143775151133e-05, + "loss": 0.2312, + "step": 23366 + }, + { + "epoch": 0.5477592784683275, + "grad_norm": 1.3279072046279907, + "learning_rate": 8.50841539561747e-05, + "loss": 0.1215, + "step": 23367 + }, + { + "epoch": 0.5477827200431324, + "grad_norm": 3.8797786235809326, + "learning_rate": 8.507687024177188e-05, + "loss": 0.3003, + "step": 23368 + }, + { + "epoch": 0.5478061616179375, + "grad_norm": 3.1337039470672607, + "learning_rate": 8.50695866083424e-05, + "loss": 0.2045, + "step": 23369 + }, + { + "epoch": 0.5478296031927425, + "grad_norm": 0.5998138785362244, + "learning_rate": 8.506230305592573e-05, + "loss": 0.0464, + "step": 23370 + }, + { + "epoch": 0.5478530447675475, + "grad_norm": 4.843188285827637, + "learning_rate": 8.50550195845614e-05, + "loss": 0.8316, + "step": 23371 + }, + { + "epoch": 0.5478764863423525, + "grad_norm": 2.1980574131011963, + "learning_rate": 8.504773619428899e-05, + "loss": 0.1082, + "step": 23372 + }, + { + "epoch": 0.5478999279171575, + "grad_norm": 0.8424931764602661, + "learning_rate": 8.504045288514799e-05, + "loss": 0.0613, + "step": 23373 + }, + { + "epoch": 0.5479233694919625, + "grad_norm": 3.6642889976501465, + "learning_rate": 8.50331696571779e-05, + "loss": 0.1798, + "step": 23374 + }, + { + "epoch": 0.5479468110667675, + "grad_norm": 2.831892251968384, + "learning_rate": 8.502588651041826e-05, + "loss": 0.3207, + "step": 23375 + }, + { + "epoch": 0.5479702526415725, + "grad_norm": 2.962725877761841, + "learning_rate": 8.501860344490859e-05, + "loss": 0.1889, + "step": 23376 + }, + { + "epoch": 0.5479936942163774, + "grad_norm": 2.669065475463867, + "learning_rate": 8.501132046068838e-05, + "loss": 1.2061, + "step": 23377 + }, + { + "epoch": 0.5480171357911825, + "grad_norm": 1.298601746559143, + "learning_rate": 8.500403755779718e-05, + "loss": 0.2099, + "step": 23378 + }, + { + "epoch": 0.5480405773659874, + "grad_norm": 3.8868703842163086, + "learning_rate": 8.499675473627449e-05, + "loss": 0.4269, + "step": 23379 + }, + { + "epoch": 0.5480640189407925, + "grad_norm": 2.243098258972168, + "learning_rate": 8.498947199615983e-05, + "loss": 0.1862, + "step": 23380 + }, + { + "epoch": 0.5480874605155974, + "grad_norm": 2.955723762512207, + "learning_rate": 8.498218933749273e-05, + "loss": 0.3142, + "step": 23381 + }, + { + "epoch": 0.5481109020904025, + "grad_norm": 2.0167624950408936, + "learning_rate": 8.497490676031267e-05, + "loss": 0.1681, + "step": 23382 + }, + { + "epoch": 0.5481343436652074, + "grad_norm": 2.5221095085144043, + "learning_rate": 8.49676242646592e-05, + "loss": 0.2552, + "step": 23383 + }, + { + "epoch": 0.5481577852400125, + "grad_norm": 3.4508156776428223, + "learning_rate": 8.496034185057178e-05, + "loss": 0.2612, + "step": 23384 + }, + { + "epoch": 0.5481812268148174, + "grad_norm": 1.112074851989746, + "learning_rate": 8.495305951809003e-05, + "loss": 0.183, + "step": 23385 + }, + { + "epoch": 0.5482046683896225, + "grad_norm": 3.195645570755005, + "learning_rate": 8.494577726725339e-05, + "loss": 0.4985, + "step": 23386 + }, + { + "epoch": 0.5482281099644274, + "grad_norm": 2.2764060497283936, + "learning_rate": 8.493849509810136e-05, + "loss": 0.1738, + "step": 23387 + }, + { + "epoch": 0.5482515515392324, + "grad_norm": 2.117048978805542, + "learning_rate": 8.49312130106735e-05, + "loss": 0.2498, + "step": 23388 + }, + { + "epoch": 0.5482749931140374, + "grad_norm": 0.787772536277771, + "learning_rate": 8.492393100500932e-05, + "loss": 0.1047, + "step": 23389 + }, + { + "epoch": 0.5482984346888424, + "grad_norm": 2.6167995929718018, + "learning_rate": 8.491664908114827e-05, + "loss": 2.1827, + "step": 23390 + }, + { + "epoch": 0.5483218762636474, + "grad_norm": 1.5403151512145996, + "learning_rate": 8.490936723912994e-05, + "loss": 0.0954, + "step": 23391 + }, + { + "epoch": 0.5483453178384524, + "grad_norm": 0.6370131969451904, + "learning_rate": 8.49020854789938e-05, + "loss": 0.063, + "step": 23392 + }, + { + "epoch": 0.5483687594132574, + "grad_norm": 3.202439069747925, + "learning_rate": 8.489480380077937e-05, + "loss": 0.3671, + "step": 23393 + }, + { + "epoch": 0.5483922009880624, + "grad_norm": 3.0008273124694824, + "learning_rate": 8.488752220452616e-05, + "loss": 0.3368, + "step": 23394 + }, + { + "epoch": 0.5484156425628673, + "grad_norm": 2.786945104598999, + "learning_rate": 8.488024069027367e-05, + "loss": 0.1316, + "step": 23395 + }, + { + "epoch": 0.5484390841376724, + "grad_norm": 1.1294385194778442, + "learning_rate": 8.487295925806141e-05, + "loss": 0.6765, + "step": 23396 + }, + { + "epoch": 0.5484625257124773, + "grad_norm": 0.578577995300293, + "learning_rate": 8.486567790792888e-05, + "loss": 0.0286, + "step": 23397 + }, + { + "epoch": 0.5484859672872824, + "grad_norm": 2.8422834873199463, + "learning_rate": 8.485839663991566e-05, + "loss": 0.2434, + "step": 23398 + }, + { + "epoch": 0.5485094088620873, + "grad_norm": 4.072866916656494, + "learning_rate": 8.485111545406119e-05, + "loss": 0.5762, + "step": 23399 + }, + { + "epoch": 0.5485328504368924, + "grad_norm": 3.4639666080474854, + "learning_rate": 8.484383435040497e-05, + "loss": 0.3503, + "step": 23400 + }, + { + "epoch": 0.5485562920116973, + "grad_norm": 1.637885332107544, + "learning_rate": 8.483655332898656e-05, + "loss": 0.1562, + "step": 23401 + }, + { + "epoch": 0.5485797335865024, + "grad_norm": 0.9275709390640259, + "learning_rate": 8.482927238984543e-05, + "loss": 0.0904, + "step": 23402 + }, + { + "epoch": 0.5486031751613073, + "grad_norm": 0.8815475106239319, + "learning_rate": 8.482199153302109e-05, + "loss": 0.0873, + "step": 23403 + }, + { + "epoch": 0.5486266167361123, + "grad_norm": 1.5825762748718262, + "learning_rate": 8.481471075855306e-05, + "loss": 0.0798, + "step": 23404 + }, + { + "epoch": 0.5486500583109173, + "grad_norm": 2.644277811050415, + "learning_rate": 8.480743006648082e-05, + "loss": 0.1342, + "step": 23405 + }, + { + "epoch": 0.5486734998857223, + "grad_norm": 5.074216842651367, + "learning_rate": 8.48001494568439e-05, + "loss": 0.6006, + "step": 23406 + }, + { + "epoch": 0.5486969414605273, + "grad_norm": 3.1882247924804688, + "learning_rate": 8.479286892968181e-05, + "loss": 0.3416, + "step": 23407 + }, + { + "epoch": 0.5487203830353323, + "grad_norm": 3.2578814029693604, + "learning_rate": 8.4785588485034e-05, + "loss": 0.2371, + "step": 23408 + }, + { + "epoch": 0.5487438246101373, + "grad_norm": 4.15861701965332, + "learning_rate": 8.477830812294003e-05, + "loss": 0.3199, + "step": 23409 + }, + { + "epoch": 0.5487672661849423, + "grad_norm": 2.795194625854492, + "learning_rate": 8.477102784343939e-05, + "loss": 1.0713, + "step": 23410 + }, + { + "epoch": 0.5487907077597473, + "grad_norm": 6.322228908538818, + "learning_rate": 8.476374764657161e-05, + "loss": 0.5089, + "step": 23411 + }, + { + "epoch": 0.5488141493345523, + "grad_norm": 1.5027669668197632, + "learning_rate": 8.475646753237614e-05, + "loss": 0.0928, + "step": 23412 + }, + { + "epoch": 0.5488375909093572, + "grad_norm": 1.8190429210662842, + "learning_rate": 8.474918750089251e-05, + "loss": 0.1701, + "step": 23413 + }, + { + "epoch": 0.5488610324841623, + "grad_norm": 2.7214763164520264, + "learning_rate": 8.474190755216022e-05, + "loss": 0.2146, + "step": 23414 + }, + { + "epoch": 0.5488844740589672, + "grad_norm": 3.174417018890381, + "learning_rate": 8.473462768621878e-05, + "loss": 1.8039, + "step": 23415 + }, + { + "epoch": 0.5489079156337723, + "grad_norm": 1.973055124282837, + "learning_rate": 8.472734790310765e-05, + "loss": 0.1473, + "step": 23416 + }, + { + "epoch": 0.5489313572085772, + "grad_norm": 0.8592649698257446, + "learning_rate": 8.47200682028664e-05, + "loss": 0.1022, + "step": 23417 + }, + { + "epoch": 0.5489547987833823, + "grad_norm": 1.907969355583191, + "learning_rate": 8.471278858553445e-05, + "loss": 0.1107, + "step": 23418 + }, + { + "epoch": 0.5489782403581872, + "grad_norm": 4.8820013999938965, + "learning_rate": 8.470550905115137e-05, + "loss": 0.3606, + "step": 23419 + }, + { + "epoch": 0.5490016819329923, + "grad_norm": 3.000544786453247, + "learning_rate": 8.469822959975657e-05, + "loss": 0.201, + "step": 23420 + }, + { + "epoch": 0.5490251235077973, + "grad_norm": 7.452936172485352, + "learning_rate": 8.469095023138967e-05, + "loss": 0.3635, + "step": 23421 + }, + { + "epoch": 0.5490485650826022, + "grad_norm": 0.5660890340805054, + "learning_rate": 8.468367094609007e-05, + "loss": 0.0372, + "step": 23422 + }, + { + "epoch": 0.5490720066574073, + "grad_norm": 1.0595955848693848, + "learning_rate": 8.467639174389732e-05, + "loss": 0.1139, + "step": 23423 + }, + { + "epoch": 0.5490954482322122, + "grad_norm": 5.2202348709106445, + "learning_rate": 8.46691126248509e-05, + "loss": 0.4856, + "step": 23424 + }, + { + "epoch": 0.5491188898070173, + "grad_norm": 0.6855443716049194, + "learning_rate": 8.46618335889903e-05, + "loss": 0.0582, + "step": 23425 + }, + { + "epoch": 0.5491423313818222, + "grad_norm": 3.069864511489868, + "learning_rate": 8.465455463635502e-05, + "loss": 0.2485, + "step": 23426 + }, + { + "epoch": 0.5491657729566273, + "grad_norm": 4.340301513671875, + "learning_rate": 8.464727576698457e-05, + "loss": 0.316, + "step": 23427 + }, + { + "epoch": 0.5491892145314322, + "grad_norm": 2.2032933235168457, + "learning_rate": 8.463999698091841e-05, + "loss": 0.0869, + "step": 23428 + }, + { + "epoch": 0.5492126561062373, + "grad_norm": 2.8139407634735107, + "learning_rate": 8.463271827819609e-05, + "loss": 0.187, + "step": 23429 + }, + { + "epoch": 0.5492360976810422, + "grad_norm": 2.2412474155426025, + "learning_rate": 8.462543965885705e-05, + "loss": 0.1953, + "step": 23430 + }, + { + "epoch": 0.5492595392558473, + "grad_norm": 2.7234933376312256, + "learning_rate": 8.46181611229408e-05, + "loss": 0.2851, + "step": 23431 + }, + { + "epoch": 0.5492829808306522, + "grad_norm": 4.774564266204834, + "learning_rate": 8.461088267048682e-05, + "loss": 0.3793, + "step": 23432 + }, + { + "epoch": 0.5493064224054572, + "grad_norm": 2.3593661785125732, + "learning_rate": 8.460360430153466e-05, + "loss": 0.2321, + "step": 23433 + }, + { + "epoch": 0.5493298639802622, + "grad_norm": 1.0165423154830933, + "learning_rate": 8.45963260161238e-05, + "loss": 0.156, + "step": 23434 + }, + { + "epoch": 0.5493533055550672, + "grad_norm": 3.522947072982788, + "learning_rate": 8.458904781429366e-05, + "loss": 0.3171, + "step": 23435 + }, + { + "epoch": 0.5493767471298722, + "grad_norm": 4.922612190246582, + "learning_rate": 8.45817696960838e-05, + "loss": 0.6127, + "step": 23436 + }, + { + "epoch": 0.5494001887046772, + "grad_norm": 2.633969306945801, + "learning_rate": 8.45744916615337e-05, + "loss": 0.1647, + "step": 23437 + }, + { + "epoch": 0.5494236302794822, + "grad_norm": 2.752399206161499, + "learning_rate": 8.456721371068282e-05, + "loss": 1.8565, + "step": 23438 + }, + { + "epoch": 0.5494470718542872, + "grad_norm": 1.3248167037963867, + "learning_rate": 8.45599358435707e-05, + "loss": 0.2117, + "step": 23439 + }, + { + "epoch": 0.5494705134290921, + "grad_norm": 1.4629079103469849, + "learning_rate": 8.45526580602368e-05, + "loss": 0.1265, + "step": 23440 + }, + { + "epoch": 0.5494939550038972, + "grad_norm": 1.872572422027588, + "learning_rate": 8.454538036072058e-05, + "loss": 0.1348, + "step": 23441 + }, + { + "epoch": 0.5495173965787021, + "grad_norm": 4.667778491973877, + "learning_rate": 8.45381027450616e-05, + "loss": 0.3493, + "step": 23442 + }, + { + "epoch": 0.5495408381535072, + "grad_norm": 0.5933252573013306, + "learning_rate": 8.453082521329927e-05, + "loss": 0.0428, + "step": 23443 + }, + { + "epoch": 0.5495642797283121, + "grad_norm": 2.4733405113220215, + "learning_rate": 8.452354776547312e-05, + "loss": 0.2171, + "step": 23444 + }, + { + "epoch": 0.5495877213031172, + "grad_norm": 4.061178207397461, + "learning_rate": 8.451627040162264e-05, + "loss": 0.2052, + "step": 23445 + }, + { + "epoch": 0.5496111628779221, + "grad_norm": 3.84350323677063, + "learning_rate": 8.450899312178733e-05, + "loss": 0.2936, + "step": 23446 + }, + { + "epoch": 0.5496346044527272, + "grad_norm": 2.152689218521118, + "learning_rate": 8.450171592600664e-05, + "loss": 0.1524, + "step": 23447 + }, + { + "epoch": 0.5496580460275321, + "grad_norm": 6.747554302215576, + "learning_rate": 8.449443881432011e-05, + "loss": 0.5241, + "step": 23448 + }, + { + "epoch": 0.5496814876023371, + "grad_norm": 2.562988042831421, + "learning_rate": 8.448716178676716e-05, + "loss": 0.2182, + "step": 23449 + }, + { + "epoch": 0.5497049291771421, + "grad_norm": 3.009052276611328, + "learning_rate": 8.447988484338731e-05, + "loss": 0.3848, + "step": 23450 + }, + { + "epoch": 0.5497283707519471, + "grad_norm": 1.7816026210784912, + "learning_rate": 8.447260798422006e-05, + "loss": 0.9277, + "step": 23451 + }, + { + "epoch": 0.5497518123267521, + "grad_norm": 3.6307902336120605, + "learning_rate": 8.446533120930487e-05, + "loss": 0.3907, + "step": 23452 + }, + { + "epoch": 0.5497752539015571, + "grad_norm": 4.081002235412598, + "learning_rate": 8.44580545186812e-05, + "loss": 0.3269, + "step": 23453 + }, + { + "epoch": 0.5497986954763621, + "grad_norm": 2.9767889976501465, + "learning_rate": 8.44507779123886e-05, + "loss": 0.3161, + "step": 23454 + }, + { + "epoch": 0.5498221370511671, + "grad_norm": 6.582054615020752, + "learning_rate": 8.44435013904665e-05, + "loss": 0.597, + "step": 23455 + }, + { + "epoch": 0.549845578625972, + "grad_norm": 3.706298589706421, + "learning_rate": 8.44362249529544e-05, + "loss": 0.3472, + "step": 23456 + }, + { + "epoch": 0.5498690202007771, + "grad_norm": 0.8635045886039734, + "learning_rate": 8.442894859989174e-05, + "loss": 0.1189, + "step": 23457 + }, + { + "epoch": 0.549892461775582, + "grad_norm": 2.1310791969299316, + "learning_rate": 8.44216723313181e-05, + "loss": 0.1821, + "step": 23458 + }, + { + "epoch": 0.5499159033503871, + "grad_norm": 2.9467437267303467, + "learning_rate": 8.441439614727289e-05, + "loss": 0.1713, + "step": 23459 + }, + { + "epoch": 0.549939344925192, + "grad_norm": 2.608279228210449, + "learning_rate": 8.440712004779559e-05, + "loss": 1.0843, + "step": 23460 + }, + { + "epoch": 0.5499627864999971, + "grad_norm": 1.7419241666793823, + "learning_rate": 8.439984403292572e-05, + "loss": 0.1529, + "step": 23461 + }, + { + "epoch": 0.549986228074802, + "grad_norm": 1.0595545768737793, + "learning_rate": 8.439256810270273e-05, + "loss": 0.1116, + "step": 23462 + }, + { + "epoch": 0.5500096696496071, + "grad_norm": 1.2621963024139404, + "learning_rate": 8.438529225716609e-05, + "loss": 0.0935, + "step": 23463 + }, + { + "epoch": 0.550033111224412, + "grad_norm": 1.9494048357009888, + "learning_rate": 8.43780164963553e-05, + "loss": 0.1244, + "step": 23464 + }, + { + "epoch": 0.5500565527992171, + "grad_norm": 3.194467544555664, + "learning_rate": 8.437074082030985e-05, + "loss": 0.1918, + "step": 23465 + }, + { + "epoch": 0.550079994374022, + "grad_norm": 2.2782647609710693, + "learning_rate": 8.436346522906915e-05, + "loss": 0.2183, + "step": 23466 + }, + { + "epoch": 0.550103435948827, + "grad_norm": 3.7767276763916016, + "learning_rate": 8.435618972267278e-05, + "loss": 0.3508, + "step": 23467 + }, + { + "epoch": 0.550126877523632, + "grad_norm": 2.67221999168396, + "learning_rate": 8.434891430116016e-05, + "loss": 1.8719, + "step": 23468 + }, + { + "epoch": 0.550150319098437, + "grad_norm": 3.4482486248016357, + "learning_rate": 8.434163896457072e-05, + "loss": 0.4531, + "step": 23469 + }, + { + "epoch": 0.550173760673242, + "grad_norm": 0.510003924369812, + "learning_rate": 8.4334363712944e-05, + "loss": 0.0498, + "step": 23470 + }, + { + "epoch": 0.550197202248047, + "grad_norm": 2.5086495876312256, + "learning_rate": 8.43270885463195e-05, + "loss": 0.1986, + "step": 23471 + }, + { + "epoch": 0.5502206438228521, + "grad_norm": 3.5472824573516846, + "learning_rate": 8.431981346473666e-05, + "loss": 0.3203, + "step": 23472 + }, + { + "epoch": 0.550244085397657, + "grad_norm": 2.8006179332733154, + "learning_rate": 8.43125384682349e-05, + "loss": 0.1588, + "step": 23473 + }, + { + "epoch": 0.5502675269724621, + "grad_norm": 3.5125255584716797, + "learning_rate": 8.430526355685379e-05, + "loss": 0.4024, + "step": 23474 + }, + { + "epoch": 0.550290968547267, + "grad_norm": 2.148127794265747, + "learning_rate": 8.429798873063278e-05, + "loss": 0.1664, + "step": 23475 + }, + { + "epoch": 0.550314410122072, + "grad_norm": 1.2739847898483276, + "learning_rate": 8.429071398961128e-05, + "loss": 0.1468, + "step": 23476 + }, + { + "epoch": 0.550337851696877, + "grad_norm": 3.4029269218444824, + "learning_rate": 8.428343933382884e-05, + "loss": 0.3369, + "step": 23477 + }, + { + "epoch": 0.550361293271682, + "grad_norm": 2.662389039993286, + "learning_rate": 8.427616476332487e-05, + "loss": 0.1969, + "step": 23478 + }, + { + "epoch": 0.550384734846487, + "grad_norm": 0.8000360131263733, + "learning_rate": 8.42688902781389e-05, + "loss": 0.0821, + "step": 23479 + }, + { + "epoch": 0.550408176421292, + "grad_norm": 0.6843093633651733, + "learning_rate": 8.426161587831037e-05, + "loss": 0.0954, + "step": 23480 + }, + { + "epoch": 0.550431617996097, + "grad_norm": 2.8464274406433105, + "learning_rate": 8.425434156387871e-05, + "loss": 0.2807, + "step": 23481 + }, + { + "epoch": 0.550455059570902, + "grad_norm": 3.2592599391937256, + "learning_rate": 8.424706733488347e-05, + "loss": 0.419, + "step": 23482 + }, + { + "epoch": 0.550478501145707, + "grad_norm": 0.7604807019233704, + "learning_rate": 8.423979319136409e-05, + "loss": 0.0682, + "step": 23483 + }, + { + "epoch": 0.550501942720512, + "grad_norm": 0.8681419491767883, + "learning_rate": 8.423251913336004e-05, + "loss": 0.0974, + "step": 23484 + }, + { + "epoch": 0.5505253842953169, + "grad_norm": 4.473780632019043, + "learning_rate": 8.422524516091077e-05, + "loss": 0.3364, + "step": 23485 + }, + { + "epoch": 0.550548825870122, + "grad_norm": 0.6827797293663025, + "learning_rate": 8.421797127405577e-05, + "loss": 0.0716, + "step": 23486 + }, + { + "epoch": 0.5505722674449269, + "grad_norm": 1.148930549621582, + "learning_rate": 8.421069747283451e-05, + "loss": 0.1043, + "step": 23487 + }, + { + "epoch": 0.550595709019732, + "grad_norm": 1.5096865892410278, + "learning_rate": 8.420342375728644e-05, + "loss": 0.1635, + "step": 23488 + }, + { + "epoch": 0.5506191505945369, + "grad_norm": 2.6670610904693604, + "learning_rate": 8.419615012745104e-05, + "loss": 0.1859, + "step": 23489 + }, + { + "epoch": 0.550642592169342, + "grad_norm": 2.989180088043213, + "learning_rate": 8.41888765833678e-05, + "loss": 0.2201, + "step": 23490 + }, + { + "epoch": 0.5506660337441469, + "grad_norm": 1.8345956802368164, + "learning_rate": 8.41816031250761e-05, + "loss": 0.1517, + "step": 23491 + }, + { + "epoch": 0.550689475318952, + "grad_norm": 3.3206090927124023, + "learning_rate": 8.417432975261552e-05, + "loss": 1.7305, + "step": 23492 + }, + { + "epoch": 0.5507129168937569, + "grad_norm": 4.842867374420166, + "learning_rate": 8.416705646602541e-05, + "loss": 0.465, + "step": 23493 + }, + { + "epoch": 0.550736358468562, + "grad_norm": 2.782735586166382, + "learning_rate": 8.415978326534534e-05, + "loss": 0.3686, + "step": 23494 + }, + { + "epoch": 0.5507598000433669, + "grad_norm": 3.0162272453308105, + "learning_rate": 8.415251015061472e-05, + "loss": 0.2178, + "step": 23495 + }, + { + "epoch": 0.5507832416181719, + "grad_norm": 2.6218388080596924, + "learning_rate": 8.414523712187306e-05, + "loss": 1.2833, + "step": 23496 + }, + { + "epoch": 0.5508066831929769, + "grad_norm": 0.5066182017326355, + "learning_rate": 8.413796417915977e-05, + "loss": 0.058, + "step": 23497 + }, + { + "epoch": 0.5508301247677819, + "grad_norm": 3.4656875133514404, + "learning_rate": 8.413069132251431e-05, + "loss": 0.2503, + "step": 23498 + }, + { + "epoch": 0.5508535663425869, + "grad_norm": 2.0387654304504395, + "learning_rate": 8.412341855197618e-05, + "loss": 0.189, + "step": 23499 + }, + { + "epoch": 0.5508770079173919, + "grad_norm": 0.9757109880447388, + "learning_rate": 8.411614586758485e-05, + "loss": 0.0764, + "step": 23500 + }, + { + "epoch": 0.5509004494921969, + "grad_norm": 3.3517649173736572, + "learning_rate": 8.410887326937972e-05, + "loss": 0.4116, + "step": 23501 + }, + { + "epoch": 0.5509238910670019, + "grad_norm": 2.902074098587036, + "learning_rate": 8.410160075740033e-05, + "loss": 0.3076, + "step": 23502 + }, + { + "epoch": 0.5509473326418068, + "grad_norm": 3.033205270767212, + "learning_rate": 8.409432833168608e-05, + "loss": 0.3831, + "step": 23503 + }, + { + "epoch": 0.5509707742166119, + "grad_norm": 3.8318986892700195, + "learning_rate": 8.408705599227644e-05, + "loss": 0.3527, + "step": 23504 + }, + { + "epoch": 0.5509942157914168, + "grad_norm": 4.500415325164795, + "learning_rate": 8.407978373921087e-05, + "loss": 0.6121, + "step": 23505 + }, + { + "epoch": 0.5510176573662219, + "grad_norm": 4.412562847137451, + "learning_rate": 8.407251157252885e-05, + "loss": 0.1995, + "step": 23506 + }, + { + "epoch": 0.5510410989410268, + "grad_norm": 0.9735939502716064, + "learning_rate": 8.406523949226986e-05, + "loss": 0.0466, + "step": 23507 + }, + { + "epoch": 0.5510645405158319, + "grad_norm": 1.0711976289749146, + "learning_rate": 8.40579674984733e-05, + "loss": 0.0853, + "step": 23508 + }, + { + "epoch": 0.5510879820906368, + "grad_norm": 1.1872313022613525, + "learning_rate": 8.405069559117869e-05, + "loss": 0.1271, + "step": 23509 + }, + { + "epoch": 0.5511114236654419, + "grad_norm": 1.539540410041809, + "learning_rate": 8.404342377042543e-05, + "loss": 0.1599, + "step": 23510 + }, + { + "epoch": 0.5511348652402468, + "grad_norm": 2.858935594558716, + "learning_rate": 8.403615203625299e-05, + "loss": 0.2825, + "step": 23511 + }, + { + "epoch": 0.5511583068150518, + "grad_norm": 3.8640732765197754, + "learning_rate": 8.402888038870087e-05, + "loss": 0.5217, + "step": 23512 + }, + { + "epoch": 0.5511817483898568, + "grad_norm": 1.2891004085540771, + "learning_rate": 8.402160882780848e-05, + "loss": 0.152, + "step": 23513 + }, + { + "epoch": 0.5512051899646618, + "grad_norm": 3.2403030395507812, + "learning_rate": 8.401433735361528e-05, + "loss": 0.3666, + "step": 23514 + }, + { + "epoch": 0.5512286315394668, + "grad_norm": 0.9878960847854614, + "learning_rate": 8.400706596616075e-05, + "loss": 0.0736, + "step": 23515 + }, + { + "epoch": 0.5512520731142718, + "grad_norm": 2.380941390991211, + "learning_rate": 8.399979466548432e-05, + "loss": 0.3374, + "step": 23516 + }, + { + "epoch": 0.5512755146890768, + "grad_norm": 3.293215274810791, + "learning_rate": 8.399252345162542e-05, + "loss": 0.1722, + "step": 23517 + }, + { + "epoch": 0.5512989562638818, + "grad_norm": 1.09858238697052, + "learning_rate": 8.398525232462356e-05, + "loss": 0.1201, + "step": 23518 + }, + { + "epoch": 0.5513223978386868, + "grad_norm": 2.9114716053009033, + "learning_rate": 8.39779812845182e-05, + "loss": 0.2402, + "step": 23519 + }, + { + "epoch": 0.5513458394134918, + "grad_norm": 2.7851624488830566, + "learning_rate": 8.397071033134875e-05, + "loss": 0.2634, + "step": 23520 + }, + { + "epoch": 0.5513692809882967, + "grad_norm": 0.9689934253692627, + "learning_rate": 8.396343946515466e-05, + "loss": 0.0771, + "step": 23521 + }, + { + "epoch": 0.5513927225631018, + "grad_norm": 1.5080612897872925, + "learning_rate": 8.395616868597543e-05, + "loss": 0.0898, + "step": 23522 + }, + { + "epoch": 0.5514161641379068, + "grad_norm": 3.383164405822754, + "learning_rate": 8.394889799385045e-05, + "loss": 0.2006, + "step": 23523 + }, + { + "epoch": 0.5514396057127118, + "grad_norm": 2.2939257621765137, + "learning_rate": 8.394162738881923e-05, + "loss": 0.2285, + "step": 23524 + }, + { + "epoch": 0.5514630472875168, + "grad_norm": 2.145514488220215, + "learning_rate": 8.393435687092119e-05, + "loss": 0.2686, + "step": 23525 + }, + { + "epoch": 0.5514864888623218, + "grad_norm": 2.810959577560425, + "learning_rate": 8.392708644019575e-05, + "loss": 0.1853, + "step": 23526 + }, + { + "epoch": 0.5515099304371268, + "grad_norm": 3.481339931488037, + "learning_rate": 8.391981609668241e-05, + "loss": 0.3101, + "step": 23527 + }, + { + "epoch": 0.5515333720119318, + "grad_norm": 3.2417683601379395, + "learning_rate": 8.391254584042061e-05, + "loss": 0.1528, + "step": 23528 + }, + { + "epoch": 0.5515568135867368, + "grad_norm": 2.0750203132629395, + "learning_rate": 8.390527567144973e-05, + "loss": 0.0968, + "step": 23529 + }, + { + "epoch": 0.5515802551615417, + "grad_norm": 3.0622832775115967, + "learning_rate": 8.389800558980931e-05, + "loss": 0.3157, + "step": 23530 + }, + { + "epoch": 0.5516036967363468, + "grad_norm": 3.458212375640869, + "learning_rate": 8.38907355955388e-05, + "loss": 0.3786, + "step": 23531 + }, + { + "epoch": 0.5516271383111517, + "grad_norm": 2.3891232013702393, + "learning_rate": 8.38834656886776e-05, + "loss": 0.1516, + "step": 23532 + }, + { + "epoch": 0.5516505798859568, + "grad_norm": 3.8100428581237793, + "learning_rate": 8.387619586926514e-05, + "loss": 0.5028, + "step": 23533 + }, + { + "epoch": 0.5516740214607617, + "grad_norm": 0.9364830255508423, + "learning_rate": 8.386892613734092e-05, + "loss": 0.0606, + "step": 23534 + }, + { + "epoch": 0.5516974630355668, + "grad_norm": 2.2790613174438477, + "learning_rate": 8.386165649294436e-05, + "loss": 0.1747, + "step": 23535 + }, + { + "epoch": 0.5517209046103717, + "grad_norm": 3.869091749191284, + "learning_rate": 8.385438693611488e-05, + "loss": 0.3842, + "step": 23536 + }, + { + "epoch": 0.5517443461851768, + "grad_norm": 3.036360502243042, + "learning_rate": 8.384711746689198e-05, + "loss": 1.544, + "step": 23537 + }, + { + "epoch": 0.5517677877599817, + "grad_norm": 0.5143707990646362, + "learning_rate": 8.383984808531508e-05, + "loss": 0.0431, + "step": 23538 + }, + { + "epoch": 0.5517912293347867, + "grad_norm": 4.0637030601501465, + "learning_rate": 8.383257879142358e-05, + "loss": 0.5195, + "step": 23539 + }, + { + "epoch": 0.5518146709095917, + "grad_norm": 0.968189537525177, + "learning_rate": 8.3825309585257e-05, + "loss": 0.1175, + "step": 23540 + }, + { + "epoch": 0.5518381124843967, + "grad_norm": 1.586687684059143, + "learning_rate": 8.381804046685474e-05, + "loss": 0.1673, + "step": 23541 + }, + { + "epoch": 0.5518615540592017, + "grad_norm": 4.808603763580322, + "learning_rate": 8.38107714362562e-05, + "loss": 0.3667, + "step": 23542 + }, + { + "epoch": 0.5518849956340067, + "grad_norm": 1.1121351718902588, + "learning_rate": 8.380350249350088e-05, + "loss": 0.1369, + "step": 23543 + }, + { + "epoch": 0.5519084372088117, + "grad_norm": 3.1949076652526855, + "learning_rate": 8.379623363862826e-05, + "loss": 0.2891, + "step": 23544 + }, + { + "epoch": 0.5519318787836167, + "grad_norm": 4.583175182342529, + "learning_rate": 8.37889648716777e-05, + "loss": 0.4271, + "step": 23545 + }, + { + "epoch": 0.5519553203584217, + "grad_norm": 0.3328500986099243, + "learning_rate": 8.378169619268868e-05, + "loss": 0.0283, + "step": 23546 + }, + { + "epoch": 0.5519787619332267, + "grad_norm": 3.023873805999756, + "learning_rate": 8.377442760170064e-05, + "loss": 0.3299, + "step": 23547 + }, + { + "epoch": 0.5520022035080316, + "grad_norm": 5.585364818572998, + "learning_rate": 8.376715909875301e-05, + "loss": 0.6978, + "step": 23548 + }, + { + "epoch": 0.5520256450828367, + "grad_norm": 2.8397443294525146, + "learning_rate": 8.375989068388522e-05, + "loss": 0.3036, + "step": 23549 + }, + { + "epoch": 0.5520490866576416, + "grad_norm": 2.647803544998169, + "learning_rate": 8.375262235713674e-05, + "loss": 0.084, + "step": 23550 + }, + { + "epoch": 0.5520725282324467, + "grad_norm": 2.149775743484497, + "learning_rate": 8.374535411854698e-05, + "loss": 0.1196, + "step": 23551 + }, + { + "epoch": 0.5520959698072516, + "grad_norm": 2.3982245922088623, + "learning_rate": 8.373808596815535e-05, + "loss": 0.3319, + "step": 23552 + }, + { + "epoch": 0.5521194113820567, + "grad_norm": 3.1548547744750977, + "learning_rate": 8.373081790600136e-05, + "loss": 0.1587, + "step": 23553 + }, + { + "epoch": 0.5521428529568616, + "grad_norm": 2.3718719482421875, + "learning_rate": 8.372354993212437e-05, + "loss": 0.1891, + "step": 23554 + }, + { + "epoch": 0.5521662945316667, + "grad_norm": 2.588679552078247, + "learning_rate": 8.371628204656392e-05, + "loss": 1.3467, + "step": 23555 + }, + { + "epoch": 0.5521897361064716, + "grad_norm": 2.266794443130493, + "learning_rate": 8.370901424935934e-05, + "loss": 1.2031, + "step": 23556 + }, + { + "epoch": 0.5522131776812766, + "grad_norm": 1.7409991025924683, + "learning_rate": 8.370174654055011e-05, + "loss": 0.0856, + "step": 23557 + }, + { + "epoch": 0.5522366192560816, + "grad_norm": 0.5832388401031494, + "learning_rate": 8.369447892017568e-05, + "loss": 0.0645, + "step": 23558 + }, + { + "epoch": 0.5522600608308866, + "grad_norm": 1.2083189487457275, + "learning_rate": 8.368721138827543e-05, + "loss": 0.0989, + "step": 23559 + }, + { + "epoch": 0.5522835024056916, + "grad_norm": 3.072822093963623, + "learning_rate": 8.367994394488887e-05, + "loss": 0.2982, + "step": 23560 + }, + { + "epoch": 0.5523069439804966, + "grad_norm": 3.8906099796295166, + "learning_rate": 8.367267659005535e-05, + "loss": 0.5313, + "step": 23561 + }, + { + "epoch": 0.5523303855553016, + "grad_norm": 1.5046263933181763, + "learning_rate": 8.366540932381439e-05, + "loss": 0.1425, + "step": 23562 + }, + { + "epoch": 0.5523538271301066, + "grad_norm": 5.574397563934326, + "learning_rate": 8.365814214620537e-05, + "loss": 0.2957, + "step": 23563 + }, + { + "epoch": 0.5523772687049116, + "grad_norm": 3.402982711791992, + "learning_rate": 8.365087505726771e-05, + "loss": 0.3106, + "step": 23564 + }, + { + "epoch": 0.5524007102797166, + "grad_norm": 1.3722405433654785, + "learning_rate": 8.364360805704087e-05, + "loss": 0.1188, + "step": 23565 + }, + { + "epoch": 0.5524241518545215, + "grad_norm": 3.9246928691864014, + "learning_rate": 8.363634114556424e-05, + "loss": 0.4153, + "step": 23566 + }, + { + "epoch": 0.5524475934293266, + "grad_norm": 2.0441195964813232, + "learning_rate": 8.362907432287733e-05, + "loss": 0.1773, + "step": 23567 + }, + { + "epoch": 0.5524710350041315, + "grad_norm": 2.7081329822540283, + "learning_rate": 8.362180758901951e-05, + "loss": 0.1652, + "step": 23568 + }, + { + "epoch": 0.5524944765789366, + "grad_norm": 3.2175238132476807, + "learning_rate": 8.361454094403023e-05, + "loss": 0.3938, + "step": 23569 + }, + { + "epoch": 0.5525179181537415, + "grad_norm": 2.538691520690918, + "learning_rate": 8.360727438794891e-05, + "loss": 0.1139, + "step": 23570 + }, + { + "epoch": 0.5525413597285466, + "grad_norm": 2.2014596462249756, + "learning_rate": 8.360000792081496e-05, + "loss": 0.1541, + "step": 23571 + }, + { + "epoch": 0.5525648013033515, + "grad_norm": 3.576063871383667, + "learning_rate": 8.359274154266786e-05, + "loss": 0.3587, + "step": 23572 + }, + { + "epoch": 0.5525882428781566, + "grad_norm": 2.4619247913360596, + "learning_rate": 8.358547525354699e-05, + "loss": 0.1896, + "step": 23573 + }, + { + "epoch": 0.5526116844529616, + "grad_norm": 2.0731563568115234, + "learning_rate": 8.357820905349178e-05, + "loss": 0.1683, + "step": 23574 + }, + { + "epoch": 0.5526351260277665, + "grad_norm": 2.8989312648773193, + "learning_rate": 8.357094294254169e-05, + "loss": 0.1814, + "step": 23575 + }, + { + "epoch": 0.5526585676025716, + "grad_norm": 1.7282943725585938, + "learning_rate": 8.356367692073613e-05, + "loss": 0.1206, + "step": 23576 + }, + { + "epoch": 0.5526820091773765, + "grad_norm": 2.552344799041748, + "learning_rate": 8.355641098811448e-05, + "loss": 0.123, + "step": 23577 + }, + { + "epoch": 0.5527054507521816, + "grad_norm": 2.4376630783081055, + "learning_rate": 8.354914514471622e-05, + "loss": 0.0961, + "step": 23578 + }, + { + "epoch": 0.5527288923269865, + "grad_norm": 3.4379336833953857, + "learning_rate": 8.354187939058077e-05, + "loss": 0.2465, + "step": 23579 + }, + { + "epoch": 0.5527523339017916, + "grad_norm": 4.266254901885986, + "learning_rate": 8.353461372574755e-05, + "loss": 1.9437, + "step": 23580 + }, + { + "epoch": 0.5527757754765965, + "grad_norm": 2.732714891433716, + "learning_rate": 8.352734815025598e-05, + "loss": 0.2642, + "step": 23581 + }, + { + "epoch": 0.5527992170514016, + "grad_norm": 3.1171035766601562, + "learning_rate": 8.35200826641455e-05, + "loss": 0.2438, + "step": 23582 + }, + { + "epoch": 0.5528226586262065, + "grad_norm": 2.7445333003997803, + "learning_rate": 8.35128172674555e-05, + "loss": 0.2367, + "step": 23583 + }, + { + "epoch": 0.5528461002010115, + "grad_norm": 2.049910068511963, + "learning_rate": 8.350555196022539e-05, + "loss": 0.1074, + "step": 23584 + }, + { + "epoch": 0.5528695417758165, + "grad_norm": 0.8477792143821716, + "learning_rate": 8.349828674249466e-05, + "loss": 0.063, + "step": 23585 + }, + { + "epoch": 0.5528929833506215, + "grad_norm": 2.0949223041534424, + "learning_rate": 8.349102161430268e-05, + "loss": 0.2269, + "step": 23586 + }, + { + "epoch": 0.5529164249254265, + "grad_norm": 2.455414056777954, + "learning_rate": 8.348375657568886e-05, + "loss": 0.2343, + "step": 23587 + }, + { + "epoch": 0.5529398665002315, + "grad_norm": 15.072948455810547, + "learning_rate": 8.347649162669267e-05, + "loss": 0.2891, + "step": 23588 + }, + { + "epoch": 0.5529633080750365, + "grad_norm": 2.501857042312622, + "learning_rate": 8.34692267673535e-05, + "loss": 0.3001, + "step": 23589 + }, + { + "epoch": 0.5529867496498415, + "grad_norm": 0.49594759941101074, + "learning_rate": 8.34619619977107e-05, + "loss": 0.0405, + "step": 23590 + }, + { + "epoch": 0.5530101912246465, + "grad_norm": 3.368028402328491, + "learning_rate": 8.34546973178038e-05, + "loss": 0.1888, + "step": 23591 + }, + { + "epoch": 0.5530336327994515, + "grad_norm": 2.5861546993255615, + "learning_rate": 8.344743272767222e-05, + "loss": 1.2667, + "step": 23592 + }, + { + "epoch": 0.5530570743742564, + "grad_norm": 2.546351194381714, + "learning_rate": 8.34401682273553e-05, + "loss": 0.1842, + "step": 23593 + }, + { + "epoch": 0.5530805159490615, + "grad_norm": 4.4521002769470215, + "learning_rate": 8.34329038168925e-05, + "loss": 0.5171, + "step": 23594 + }, + { + "epoch": 0.5531039575238664, + "grad_norm": 1.9893306493759155, + "learning_rate": 8.342563949632324e-05, + "loss": 0.1386, + "step": 23595 + }, + { + "epoch": 0.5531273990986715, + "grad_norm": 0.7684982419013977, + "learning_rate": 8.34183752656869e-05, + "loss": 0.0674, + "step": 23596 + }, + { + "epoch": 0.5531508406734764, + "grad_norm": 2.700221538543701, + "learning_rate": 8.341111112502294e-05, + "loss": 0.3279, + "step": 23597 + }, + { + "epoch": 0.5531742822482815, + "grad_norm": 5.898282527923584, + "learning_rate": 8.340384707437077e-05, + "loss": 0.1846, + "step": 23598 + }, + { + "epoch": 0.5531977238230864, + "grad_norm": 0.8189725875854492, + "learning_rate": 8.339658311376977e-05, + "loss": 0.0792, + "step": 23599 + }, + { + "epoch": 0.5532211653978915, + "grad_norm": 5.581617832183838, + "learning_rate": 8.338931924325939e-05, + "loss": 0.3171, + "step": 23600 + }, + { + "epoch": 0.5532446069726964, + "grad_norm": 1.5894536972045898, + "learning_rate": 8.338205546287904e-05, + "loss": 0.0705, + "step": 23601 + }, + { + "epoch": 0.5532680485475014, + "grad_norm": 4.518916606903076, + "learning_rate": 8.337479177266806e-05, + "loss": 0.5345, + "step": 23602 + }, + { + "epoch": 0.5532914901223064, + "grad_norm": 5.936638832092285, + "learning_rate": 8.336752817266598e-05, + "loss": 0.6102, + "step": 23603 + }, + { + "epoch": 0.5533149316971114, + "grad_norm": 2.490664482116699, + "learning_rate": 8.336026466291215e-05, + "loss": 0.2693, + "step": 23604 + }, + { + "epoch": 0.5533383732719164, + "grad_norm": 2.5316269397735596, + "learning_rate": 8.335300124344603e-05, + "loss": 0.1257, + "step": 23605 + }, + { + "epoch": 0.5533618148467214, + "grad_norm": 4.561210632324219, + "learning_rate": 8.334573791430694e-05, + "loss": 0.4947, + "step": 23606 + }, + { + "epoch": 0.5533852564215264, + "grad_norm": 2.3985977172851562, + "learning_rate": 8.333847467553437e-05, + "loss": 0.1957, + "step": 23607 + }, + { + "epoch": 0.5534086979963314, + "grad_norm": 2.8302533626556396, + "learning_rate": 8.333121152716772e-05, + "loss": 1.1447, + "step": 23608 + }, + { + "epoch": 0.5534321395711364, + "grad_norm": 4.355576038360596, + "learning_rate": 8.332394846924635e-05, + "loss": 0.323, + "step": 23609 + }, + { + "epoch": 0.5534555811459414, + "grad_norm": 2.88173246383667, + "learning_rate": 8.331668550180974e-05, + "loss": 0.127, + "step": 23610 + }, + { + "epoch": 0.5534790227207463, + "grad_norm": 1.2448318004608154, + "learning_rate": 8.330942262489725e-05, + "loss": 0.1312, + "step": 23611 + }, + { + "epoch": 0.5535024642955514, + "grad_norm": 3.1975157260894775, + "learning_rate": 8.330215983854828e-05, + "loss": 0.2328, + "step": 23612 + }, + { + "epoch": 0.5535259058703563, + "grad_norm": 1.1383285522460938, + "learning_rate": 8.329489714280229e-05, + "loss": 0.0618, + "step": 23613 + }, + { + "epoch": 0.5535493474451614, + "grad_norm": 3.071380138397217, + "learning_rate": 8.328763453769866e-05, + "loss": 1.523, + "step": 23614 + }, + { + "epoch": 0.5535727890199663, + "grad_norm": 0.9217262864112854, + "learning_rate": 8.328037202327674e-05, + "loss": 0.1068, + "step": 23615 + }, + { + "epoch": 0.5535962305947714, + "grad_norm": 2.6597886085510254, + "learning_rate": 8.327310959957601e-05, + "loss": 0.0702, + "step": 23616 + }, + { + "epoch": 0.5536196721695763, + "grad_norm": 3.7542524337768555, + "learning_rate": 8.326584726663589e-05, + "loss": 0.254, + "step": 23617 + }, + { + "epoch": 0.5536431137443814, + "grad_norm": 3.1407828330993652, + "learning_rate": 8.325858502449575e-05, + "loss": 0.2118, + "step": 23618 + }, + { + "epoch": 0.5536665553191863, + "grad_norm": 2.958641529083252, + "learning_rate": 8.325132287319498e-05, + "loss": 0.1535, + "step": 23619 + }, + { + "epoch": 0.5536899968939913, + "grad_norm": 2.1411750316619873, + "learning_rate": 8.324406081277303e-05, + "loss": 0.9392, + "step": 23620 + }, + { + "epoch": 0.5537134384687963, + "grad_norm": 3.314117908477783, + "learning_rate": 8.323679884326927e-05, + "loss": 0.2415, + "step": 23621 + }, + { + "epoch": 0.5537368800436013, + "grad_norm": 2.37038254737854, + "learning_rate": 8.322953696472309e-05, + "loss": 1.5719, + "step": 23622 + }, + { + "epoch": 0.5537603216184063, + "grad_norm": 3.8253676891326904, + "learning_rate": 8.322227517717393e-05, + "loss": 0.2302, + "step": 23623 + }, + { + "epoch": 0.5537837631932113, + "grad_norm": 0.36508721113204956, + "learning_rate": 8.321501348066119e-05, + "loss": 0.0544, + "step": 23624 + }, + { + "epoch": 0.5538072047680164, + "grad_norm": 4.548277378082275, + "learning_rate": 8.320775187522422e-05, + "loss": 0.3408, + "step": 23625 + }, + { + "epoch": 0.5538306463428213, + "grad_norm": 1.0344878435134888, + "learning_rate": 8.32004903609025e-05, + "loss": 0.0482, + "step": 23626 + }, + { + "epoch": 0.5538540879176264, + "grad_norm": 0.7680381536483765, + "learning_rate": 8.319322893773533e-05, + "loss": 0.0344, + "step": 23627 + }, + { + "epoch": 0.5538775294924313, + "grad_norm": 3.3793771266937256, + "learning_rate": 8.318596760576222e-05, + "loss": 0.4533, + "step": 23628 + }, + { + "epoch": 0.5539009710672363, + "grad_norm": 0.9236978888511658, + "learning_rate": 8.317870636502251e-05, + "loss": 0.1083, + "step": 23629 + }, + { + "epoch": 0.5539244126420413, + "grad_norm": 4.269637584686279, + "learning_rate": 8.317144521555563e-05, + "loss": 0.2477, + "step": 23630 + }, + { + "epoch": 0.5539478542168463, + "grad_norm": 3.289621114730835, + "learning_rate": 8.316418415740096e-05, + "loss": 0.3412, + "step": 23631 + }, + { + "epoch": 0.5539712957916513, + "grad_norm": 1.5433577299118042, + "learning_rate": 8.315692319059788e-05, + "loss": 0.1367, + "step": 23632 + }, + { + "epoch": 0.5539947373664563, + "grad_norm": 3.237459897994995, + "learning_rate": 8.314966231518582e-05, + "loss": 0.1967, + "step": 23633 + }, + { + "epoch": 0.5540181789412613, + "grad_norm": 3.2135841846466064, + "learning_rate": 8.314240153120416e-05, + "loss": 0.2506, + "step": 23634 + }, + { + "epoch": 0.5540416205160663, + "grad_norm": 2.9931139945983887, + "learning_rate": 8.313514083869232e-05, + "loss": 0.3881, + "step": 23635 + }, + { + "epoch": 0.5540650620908713, + "grad_norm": 2.696727991104126, + "learning_rate": 8.312788023768968e-05, + "loss": 0.475, + "step": 23636 + }, + { + "epoch": 0.5540885036656763, + "grad_norm": 4.682214260101318, + "learning_rate": 8.312061972823561e-05, + "loss": 0.5385, + "step": 23637 + }, + { + "epoch": 0.5541119452404812, + "grad_norm": 3.6065847873687744, + "learning_rate": 8.311335931036956e-05, + "loss": 0.2624, + "step": 23638 + }, + { + "epoch": 0.5541353868152863, + "grad_norm": 0.4586665630340576, + "learning_rate": 8.310609898413086e-05, + "loss": 0.0391, + "step": 23639 + }, + { + "epoch": 0.5541588283900912, + "grad_norm": 1.2139302492141724, + "learning_rate": 8.309883874955897e-05, + "loss": 0.1145, + "step": 23640 + }, + { + "epoch": 0.5541822699648963, + "grad_norm": 2.6332008838653564, + "learning_rate": 8.309157860669323e-05, + "loss": 0.2882, + "step": 23641 + }, + { + "epoch": 0.5542057115397012, + "grad_norm": 2.3818728923797607, + "learning_rate": 8.308431855557311e-05, + "loss": 0.1967, + "step": 23642 + }, + { + "epoch": 0.5542291531145063, + "grad_norm": 4.481512069702148, + "learning_rate": 8.307705859623792e-05, + "loss": 0.2426, + "step": 23643 + }, + { + "epoch": 0.5542525946893112, + "grad_norm": 2.5475082397460938, + "learning_rate": 8.306979872872709e-05, + "loss": 0.2656, + "step": 23644 + }, + { + "epoch": 0.5542760362641163, + "grad_norm": 2.0858211517333984, + "learning_rate": 8.306253895308001e-05, + "loss": 0.2001, + "step": 23645 + }, + { + "epoch": 0.5542994778389212, + "grad_norm": 1.9284603595733643, + "learning_rate": 8.305527926933609e-05, + "loss": 0.1793, + "step": 23646 + }, + { + "epoch": 0.5543229194137262, + "grad_norm": 3.464259147644043, + "learning_rate": 8.304801967753466e-05, + "loss": 0.3753, + "step": 23647 + }, + { + "epoch": 0.5543463609885312, + "grad_norm": 3.4727258682250977, + "learning_rate": 8.304076017771519e-05, + "loss": 0.2582, + "step": 23648 + }, + { + "epoch": 0.5543698025633362, + "grad_norm": 3.2299916744232178, + "learning_rate": 8.303350076991702e-05, + "loss": 0.3012, + "step": 23649 + }, + { + "epoch": 0.5543932441381412, + "grad_norm": 0.5057944059371948, + "learning_rate": 8.302624145417953e-05, + "loss": 0.2469, + "step": 23650 + }, + { + "epoch": 0.5544166857129462, + "grad_norm": 4.741881847381592, + "learning_rate": 8.301898223054212e-05, + "loss": 0.4701, + "step": 23651 + }, + { + "epoch": 0.5544401272877512, + "grad_norm": 4.384498119354248, + "learning_rate": 8.301172309904422e-05, + "loss": 0.7236, + "step": 23652 + }, + { + "epoch": 0.5544635688625562, + "grad_norm": 4.373600006103516, + "learning_rate": 8.30044640597252e-05, + "loss": 0.2582, + "step": 23653 + }, + { + "epoch": 0.5544870104373611, + "grad_norm": 0.5337424278259277, + "learning_rate": 8.299720511262441e-05, + "loss": 0.0486, + "step": 23654 + }, + { + "epoch": 0.5545104520121662, + "grad_norm": 0.6116155385971069, + "learning_rate": 8.298994625778128e-05, + "loss": 0.0272, + "step": 23655 + }, + { + "epoch": 0.5545338935869711, + "grad_norm": 4.7687153816223145, + "learning_rate": 8.298268749523519e-05, + "loss": 0.5059, + "step": 23656 + }, + { + "epoch": 0.5545573351617762, + "grad_norm": 3.47294020652771, + "learning_rate": 8.297542882502548e-05, + "loss": 0.3178, + "step": 23657 + }, + { + "epoch": 0.5545807767365811, + "grad_norm": 0.8224114179611206, + "learning_rate": 8.296817024719161e-05, + "loss": 0.0656, + "step": 23658 + }, + { + "epoch": 0.5546042183113862, + "grad_norm": 4.468104362487793, + "learning_rate": 8.296091176177293e-05, + "loss": 0.2915, + "step": 23659 + }, + { + "epoch": 0.5546276598861911, + "grad_norm": 0.7639911770820618, + "learning_rate": 8.295365336880878e-05, + "loss": 0.0789, + "step": 23660 + }, + { + "epoch": 0.5546511014609962, + "grad_norm": 1.5353679656982422, + "learning_rate": 8.294639506833861e-05, + "loss": 1.3724, + "step": 23661 + }, + { + "epoch": 0.5546745430358011, + "grad_norm": 2.7449703216552734, + "learning_rate": 8.29391368604018e-05, + "loss": 0.1894, + "step": 23662 + }, + { + "epoch": 0.5546979846106062, + "grad_norm": 3.0631163120269775, + "learning_rate": 8.293187874503766e-05, + "loss": 0.212, + "step": 23663 + }, + { + "epoch": 0.5547214261854111, + "grad_norm": 0.6742442846298218, + "learning_rate": 8.292462072228563e-05, + "loss": 0.0387, + "step": 23664 + }, + { + "epoch": 0.5547448677602161, + "grad_norm": 0.900664210319519, + "learning_rate": 8.291736279218512e-05, + "loss": 0.1091, + "step": 23665 + }, + { + "epoch": 0.5547683093350211, + "grad_norm": 3.58425235748291, + "learning_rate": 8.291010495477549e-05, + "loss": 0.342, + "step": 23666 + }, + { + "epoch": 0.5547917509098261, + "grad_norm": 5.067742347717285, + "learning_rate": 8.290284721009609e-05, + "loss": 0.5177, + "step": 23667 + }, + { + "epoch": 0.5548151924846311, + "grad_norm": 0.893101692199707, + "learning_rate": 8.289558955818633e-05, + "loss": 0.0837, + "step": 23668 + }, + { + "epoch": 0.5548386340594361, + "grad_norm": 1.1908663511276245, + "learning_rate": 8.28883319990856e-05, + "loss": 0.1015, + "step": 23669 + }, + { + "epoch": 0.5548620756342411, + "grad_norm": 4.234071254730225, + "learning_rate": 8.288107453283323e-05, + "loss": 0.2441, + "step": 23670 + }, + { + "epoch": 0.5548855172090461, + "grad_norm": 3.2116801738739014, + "learning_rate": 8.287381715946864e-05, + "loss": 0.3651, + "step": 23671 + }, + { + "epoch": 0.554908958783851, + "grad_norm": 5.236876010894775, + "learning_rate": 8.286655987903119e-05, + "loss": 0.3006, + "step": 23672 + }, + { + "epoch": 0.5549324003586561, + "grad_norm": 3.868314027786255, + "learning_rate": 8.28593026915603e-05, + "loss": 0.5016, + "step": 23673 + }, + { + "epoch": 0.554955841933461, + "grad_norm": 3.429611921310425, + "learning_rate": 8.28520455970953e-05, + "loss": 0.23, + "step": 23674 + }, + { + "epoch": 0.5549792835082661, + "grad_norm": 2.687528133392334, + "learning_rate": 8.284478859567553e-05, + "loss": 0.2772, + "step": 23675 + }, + { + "epoch": 0.5550027250830711, + "grad_norm": 1.8733576536178589, + "learning_rate": 8.283753168734048e-05, + "loss": 0.1184, + "step": 23676 + }, + { + "epoch": 0.5550261666578761, + "grad_norm": 1.9844021797180176, + "learning_rate": 8.283027487212944e-05, + "loss": 0.1275, + "step": 23677 + }, + { + "epoch": 0.5550496082326811, + "grad_norm": 1.1325584650039673, + "learning_rate": 8.282301815008185e-05, + "loss": 0.0815, + "step": 23678 + }, + { + "epoch": 0.5550730498074861, + "grad_norm": 3.536956548690796, + "learning_rate": 8.2815761521237e-05, + "loss": 0.499, + "step": 23679 + }, + { + "epoch": 0.5550964913822911, + "grad_norm": 4.362584590911865, + "learning_rate": 8.280850498563433e-05, + "loss": 0.4163, + "step": 23680 + }, + { + "epoch": 0.555119932957096, + "grad_norm": 1.5411829948425293, + "learning_rate": 8.280124854331322e-05, + "loss": 0.1569, + "step": 23681 + }, + { + "epoch": 0.5551433745319011, + "grad_norm": 3.611273765563965, + "learning_rate": 8.279399219431298e-05, + "loss": 0.2234, + "step": 23682 + }, + { + "epoch": 0.555166816106706, + "grad_norm": 2.590049982070923, + "learning_rate": 8.278673593867304e-05, + "loss": 0.257, + "step": 23683 + }, + { + "epoch": 0.5551902576815111, + "grad_norm": 4.528890609741211, + "learning_rate": 8.277947977643276e-05, + "loss": 0.2757, + "step": 23684 + }, + { + "epoch": 0.555213699256316, + "grad_norm": 2.0875308513641357, + "learning_rate": 8.27722237076315e-05, + "loss": 0.2138, + "step": 23685 + }, + { + "epoch": 0.5552371408311211, + "grad_norm": 2.886068105697632, + "learning_rate": 8.276496773230864e-05, + "loss": 0.3124, + "step": 23686 + }, + { + "epoch": 0.555260582405926, + "grad_norm": 1.8260596990585327, + "learning_rate": 8.275771185050356e-05, + "loss": 0.1659, + "step": 23687 + }, + { + "epoch": 0.5552840239807311, + "grad_norm": 1.205262541770935, + "learning_rate": 8.275045606225558e-05, + "loss": 0.0631, + "step": 23688 + }, + { + "epoch": 0.555307465555536, + "grad_norm": 4.65833044052124, + "learning_rate": 8.274320036760412e-05, + "loss": 0.3483, + "step": 23689 + }, + { + "epoch": 0.5553309071303411, + "grad_norm": 2.639401435852051, + "learning_rate": 8.273594476658857e-05, + "loss": 0.1981, + "step": 23690 + }, + { + "epoch": 0.555354348705146, + "grad_norm": 4.897193431854248, + "learning_rate": 8.272868925924827e-05, + "loss": 0.6178, + "step": 23691 + }, + { + "epoch": 0.555377790279951, + "grad_norm": 2.0924429893493652, + "learning_rate": 8.272143384562256e-05, + "loss": 1.5281, + "step": 23692 + }, + { + "epoch": 0.555401231854756, + "grad_norm": 2.8862452507019043, + "learning_rate": 8.271417852575086e-05, + "loss": 0.211, + "step": 23693 + }, + { + "epoch": 0.555424673429561, + "grad_norm": 2.1786081790924072, + "learning_rate": 8.270692329967252e-05, + "loss": 1.5804, + "step": 23694 + }, + { + "epoch": 0.555448115004366, + "grad_norm": 2.2922520637512207, + "learning_rate": 8.269966816742687e-05, + "loss": 0.2769, + "step": 23695 + }, + { + "epoch": 0.555471556579171, + "grad_norm": 1.6465648412704468, + "learning_rate": 8.269241312905333e-05, + "loss": 0.1344, + "step": 23696 + }, + { + "epoch": 0.555494998153976, + "grad_norm": 3.065784215927124, + "learning_rate": 8.268515818459125e-05, + "loss": 0.2685, + "step": 23697 + }, + { + "epoch": 0.555518439728781, + "grad_norm": 3.5309898853302, + "learning_rate": 8.267790333407996e-05, + "loss": 0.4245, + "step": 23698 + }, + { + "epoch": 0.555541881303586, + "grad_norm": 1.5996479988098145, + "learning_rate": 8.267064857755888e-05, + "loss": 0.0679, + "step": 23699 + }, + { + "epoch": 0.555565322878391, + "grad_norm": 3.463695764541626, + "learning_rate": 8.26633939150673e-05, + "loss": 0.4496, + "step": 23700 + }, + { + "epoch": 0.5555887644531959, + "grad_norm": 3.0546767711639404, + "learning_rate": 8.265613934664467e-05, + "loss": 0.172, + "step": 23701 + }, + { + "epoch": 0.555612206028001, + "grad_norm": 3.1939194202423096, + "learning_rate": 8.264888487233032e-05, + "loss": 1.5734, + "step": 23702 + }, + { + "epoch": 0.5556356476028059, + "grad_norm": 4.716037750244141, + "learning_rate": 8.26416304921636e-05, + "loss": 0.3174, + "step": 23703 + }, + { + "epoch": 0.555659089177611, + "grad_norm": 5.124598979949951, + "learning_rate": 8.263437620618389e-05, + "loss": 0.5264, + "step": 23704 + }, + { + "epoch": 0.5556825307524159, + "grad_norm": 5.933039665222168, + "learning_rate": 8.262712201443052e-05, + "loss": 0.3557, + "step": 23705 + }, + { + "epoch": 0.555705972327221, + "grad_norm": 3.4870798587799072, + "learning_rate": 8.261986791694291e-05, + "loss": 1.5252, + "step": 23706 + }, + { + "epoch": 0.5557294139020259, + "grad_norm": 3.2139792442321777, + "learning_rate": 8.261261391376038e-05, + "loss": 0.1969, + "step": 23707 + }, + { + "epoch": 0.555752855476831, + "grad_norm": 3.5831172466278076, + "learning_rate": 8.260536000492226e-05, + "loss": 0.2075, + "step": 23708 + }, + { + "epoch": 0.5557762970516359, + "grad_norm": 3.5044190883636475, + "learning_rate": 8.259810619046797e-05, + "loss": 0.2747, + "step": 23709 + }, + { + "epoch": 0.5557997386264409, + "grad_norm": 2.842102527618408, + "learning_rate": 8.259085247043683e-05, + "loss": 0.2415, + "step": 23710 + }, + { + "epoch": 0.5558231802012459, + "grad_norm": 0.5688152313232422, + "learning_rate": 8.258359884486824e-05, + "loss": 0.048, + "step": 23711 + }, + { + "epoch": 0.5558466217760509, + "grad_norm": 3.6214263439178467, + "learning_rate": 8.257634531380148e-05, + "loss": 0.303, + "step": 23712 + }, + { + "epoch": 0.5558700633508559, + "grad_norm": 3.420621633529663, + "learning_rate": 8.256909187727601e-05, + "loss": 0.2102, + "step": 23713 + }, + { + "epoch": 0.5558935049256609, + "grad_norm": 4.394234657287598, + "learning_rate": 8.256183853533112e-05, + "loss": 0.6499, + "step": 23714 + }, + { + "epoch": 0.5559169465004659, + "grad_norm": 6.4671244621276855, + "learning_rate": 8.25545852880062e-05, + "loss": 0.4428, + "step": 23715 + }, + { + "epoch": 0.5559403880752709, + "grad_norm": 0.9072214961051941, + "learning_rate": 8.254733213534059e-05, + "loss": 0.1128, + "step": 23716 + }, + { + "epoch": 0.5559638296500758, + "grad_norm": 2.0315611362457275, + "learning_rate": 8.254007907737362e-05, + "loss": 0.1549, + "step": 23717 + }, + { + "epoch": 0.5559872712248809, + "grad_norm": 1.3383690118789673, + "learning_rate": 8.25328261141447e-05, + "loss": 0.1077, + "step": 23718 + }, + { + "epoch": 0.5560107127996858, + "grad_norm": 3.3957059383392334, + "learning_rate": 8.252557324569317e-05, + "loss": 0.2764, + "step": 23719 + }, + { + "epoch": 0.5560341543744909, + "grad_norm": 1.9725425243377686, + "learning_rate": 8.251832047205833e-05, + "loss": 0.1192, + "step": 23720 + }, + { + "epoch": 0.5560575959492958, + "grad_norm": 1.122715711593628, + "learning_rate": 8.25110677932796e-05, + "loss": 0.1143, + "step": 23721 + }, + { + "epoch": 0.5560810375241009, + "grad_norm": 3.1988868713378906, + "learning_rate": 8.25038152093963e-05, + "loss": 0.2887, + "step": 23722 + }, + { + "epoch": 0.5561044790989058, + "grad_norm": 1.1712974309921265, + "learning_rate": 8.249656272044778e-05, + "loss": 0.1579, + "step": 23723 + }, + { + "epoch": 0.5561279206737109, + "grad_norm": 2.567523241043091, + "learning_rate": 8.248931032647338e-05, + "loss": 1.9545, + "step": 23724 + }, + { + "epoch": 0.5561513622485158, + "grad_norm": 2.763453960418701, + "learning_rate": 8.248205802751252e-05, + "loss": 1.3368, + "step": 23725 + }, + { + "epoch": 0.5561748038233209, + "grad_norm": 0.8094480037689209, + "learning_rate": 8.247480582360451e-05, + "loss": 0.0726, + "step": 23726 + }, + { + "epoch": 0.5561982453981259, + "grad_norm": 0.9043419361114502, + "learning_rate": 8.246755371478867e-05, + "loss": 0.0975, + "step": 23727 + }, + { + "epoch": 0.5562216869729308, + "grad_norm": 1.2344403266906738, + "learning_rate": 8.246030170110439e-05, + "loss": 0.0524, + "step": 23728 + }, + { + "epoch": 0.5562451285477359, + "grad_norm": 3.831587076187134, + "learning_rate": 8.2453049782591e-05, + "loss": 0.2499, + "step": 23729 + }, + { + "epoch": 0.5562685701225408, + "grad_norm": 0.7992597222328186, + "learning_rate": 8.244579795928787e-05, + "loss": 0.0337, + "step": 23730 + }, + { + "epoch": 0.5562920116973459, + "grad_norm": 2.884371757507324, + "learning_rate": 8.243854623123433e-05, + "loss": 0.3196, + "step": 23731 + }, + { + "epoch": 0.5563154532721508, + "grad_norm": 2.2960305213928223, + "learning_rate": 8.243129459846972e-05, + "loss": 0.1916, + "step": 23732 + }, + { + "epoch": 0.5563388948469559, + "grad_norm": 3.369502305984497, + "learning_rate": 8.24240430610334e-05, + "loss": 0.1615, + "step": 23733 + }, + { + "epoch": 0.5563623364217608, + "grad_norm": 1.470619797706604, + "learning_rate": 8.241679161896472e-05, + "loss": 0.788, + "step": 23734 + }, + { + "epoch": 0.5563857779965659, + "grad_norm": 5.091094493865967, + "learning_rate": 8.240954027230303e-05, + "loss": 0.4996, + "step": 23735 + }, + { + "epoch": 0.5564092195713708, + "grad_norm": 4.851541042327881, + "learning_rate": 8.240228902108764e-05, + "loss": 0.3494, + "step": 23736 + }, + { + "epoch": 0.5564326611461758, + "grad_norm": 3.431608200073242, + "learning_rate": 8.239503786535793e-05, + "loss": 0.2655, + "step": 23737 + }, + { + "epoch": 0.5564561027209808, + "grad_norm": 3.2082231044769287, + "learning_rate": 8.238778680515327e-05, + "loss": 0.2933, + "step": 23738 + }, + { + "epoch": 0.5564795442957858, + "grad_norm": 1.489335536956787, + "learning_rate": 8.238053584051296e-05, + "loss": 0.2126, + "step": 23739 + }, + { + "epoch": 0.5565029858705908, + "grad_norm": 2.632073163986206, + "learning_rate": 8.237328497147635e-05, + "loss": 0.2794, + "step": 23740 + }, + { + "epoch": 0.5565264274453958, + "grad_norm": 3.144501209259033, + "learning_rate": 8.23660341980828e-05, + "loss": 0.2582, + "step": 23741 + }, + { + "epoch": 0.5565498690202008, + "grad_norm": 3.0451085567474365, + "learning_rate": 8.235878352037166e-05, + "loss": 0.3213, + "step": 23742 + }, + { + "epoch": 0.5565733105950058, + "grad_norm": 3.9560976028442383, + "learning_rate": 8.235153293838223e-05, + "loss": 0.452, + "step": 23743 + }, + { + "epoch": 0.5565967521698107, + "grad_norm": 1.6844552755355835, + "learning_rate": 8.234428245215389e-05, + "loss": 0.1991, + "step": 23744 + }, + { + "epoch": 0.5566201937446158, + "grad_norm": 0.6311696171760559, + "learning_rate": 8.233703206172598e-05, + "loss": 0.0599, + "step": 23745 + }, + { + "epoch": 0.5566436353194207, + "grad_norm": 3.6676387786865234, + "learning_rate": 8.232978176713782e-05, + "loss": 0.3923, + "step": 23746 + }, + { + "epoch": 0.5566670768942258, + "grad_norm": 4.255706310272217, + "learning_rate": 8.232253156842876e-05, + "loss": 0.3998, + "step": 23747 + }, + { + "epoch": 0.5566905184690307, + "grad_norm": 2.299229860305786, + "learning_rate": 8.231528146563811e-05, + "loss": 0.2202, + "step": 23748 + }, + { + "epoch": 0.5567139600438358, + "grad_norm": 2.8678324222564697, + "learning_rate": 8.230803145880527e-05, + "loss": 0.2992, + "step": 23749 + }, + { + "epoch": 0.5567374016186407, + "grad_norm": 2.541776418685913, + "learning_rate": 8.230078154796955e-05, + "loss": 0.2255, + "step": 23750 + }, + { + "epoch": 0.5567608431934458, + "grad_norm": 2.4217660427093506, + "learning_rate": 8.229353173317029e-05, + "loss": 0.1639, + "step": 23751 + }, + { + "epoch": 0.5567842847682507, + "grad_norm": 2.7177021503448486, + "learning_rate": 8.228628201444682e-05, + "loss": 0.161, + "step": 23752 + }, + { + "epoch": 0.5568077263430558, + "grad_norm": 2.3926897048950195, + "learning_rate": 8.227903239183848e-05, + "loss": 0.1724, + "step": 23753 + }, + { + "epoch": 0.5568311679178607, + "grad_norm": 3.163165330886841, + "learning_rate": 8.227178286538463e-05, + "loss": 0.1637, + "step": 23754 + }, + { + "epoch": 0.5568546094926657, + "grad_norm": 1.9822858572006226, + "learning_rate": 8.226453343512456e-05, + "loss": 1.5067, + "step": 23755 + }, + { + "epoch": 0.5568780510674707, + "grad_norm": 0.48750877380371094, + "learning_rate": 8.225728410109764e-05, + "loss": 0.0505, + "step": 23756 + }, + { + "epoch": 0.5569014926422757, + "grad_norm": 4.05214786529541, + "learning_rate": 8.22500348633432e-05, + "loss": 0.3692, + "step": 23757 + }, + { + "epoch": 0.5569249342170807, + "grad_norm": 3.0796117782592773, + "learning_rate": 8.224278572190055e-05, + "loss": 0.2074, + "step": 23758 + }, + { + "epoch": 0.5569483757918857, + "grad_norm": 0.4708680808544159, + "learning_rate": 8.223553667680906e-05, + "loss": 0.0459, + "step": 23759 + }, + { + "epoch": 0.5569718173666907, + "grad_norm": 3.8136253356933594, + "learning_rate": 8.222828772810802e-05, + "loss": 0.2579, + "step": 23760 + }, + { + "epoch": 0.5569952589414957, + "grad_norm": 1.814652919769287, + "learning_rate": 8.222103887583682e-05, + "loss": 0.1662, + "step": 23761 + }, + { + "epoch": 0.5570187005163006, + "grad_norm": 1.5239667892456055, + "learning_rate": 8.221379012003474e-05, + "loss": 0.0988, + "step": 23762 + }, + { + "epoch": 0.5570421420911057, + "grad_norm": 2.6290318965911865, + "learning_rate": 8.220654146074116e-05, + "loss": 0.1718, + "step": 23763 + }, + { + "epoch": 0.5570655836659106, + "grad_norm": 0.6860957145690918, + "learning_rate": 8.21992928979954e-05, + "loss": 0.0343, + "step": 23764 + }, + { + "epoch": 0.5570890252407157, + "grad_norm": 4.718218803405762, + "learning_rate": 8.219204443183674e-05, + "loss": 0.555, + "step": 23765 + }, + { + "epoch": 0.5571124668155206, + "grad_norm": 3.5370359420776367, + "learning_rate": 8.218479606230457e-05, + "loss": 0.3118, + "step": 23766 + }, + { + "epoch": 0.5571359083903257, + "grad_norm": 0.7171716094017029, + "learning_rate": 8.217754778943821e-05, + "loss": 0.0565, + "step": 23767 + }, + { + "epoch": 0.5571593499651306, + "grad_norm": 2.1745426654815674, + "learning_rate": 8.217029961327693e-05, + "loss": 0.1093, + "step": 23768 + }, + { + "epoch": 0.5571827915399357, + "grad_norm": 4.4765119552612305, + "learning_rate": 8.216305153386015e-05, + "loss": 0.361, + "step": 23769 + }, + { + "epoch": 0.5572062331147406, + "grad_norm": 5.055786609649658, + "learning_rate": 8.215580355122713e-05, + "loss": 0.4006, + "step": 23770 + }, + { + "epoch": 0.5572296746895457, + "grad_norm": 3.2519567012786865, + "learning_rate": 8.214855566541724e-05, + "loss": 0.4139, + "step": 23771 + }, + { + "epoch": 0.5572531162643506, + "grad_norm": 1.0630338191986084, + "learning_rate": 8.214130787646977e-05, + "loss": 0.0483, + "step": 23772 + }, + { + "epoch": 0.5572765578391556, + "grad_norm": 3.908109188079834, + "learning_rate": 8.213406018442405e-05, + "loss": 0.4035, + "step": 23773 + }, + { + "epoch": 0.5572999994139606, + "grad_norm": 2.3441758155822754, + "learning_rate": 8.212681258931945e-05, + "loss": 0.1952, + "step": 23774 + }, + { + "epoch": 0.5573234409887656, + "grad_norm": 2.353337287902832, + "learning_rate": 8.211956509119525e-05, + "loss": 1.7862, + "step": 23775 + }, + { + "epoch": 0.5573468825635706, + "grad_norm": 0.5663390755653381, + "learning_rate": 8.211231769009082e-05, + "loss": 0.0518, + "step": 23776 + }, + { + "epoch": 0.5573703241383756, + "grad_norm": 3.982759952545166, + "learning_rate": 8.210507038604545e-05, + "loss": 0.339, + "step": 23777 + }, + { + "epoch": 0.5573937657131807, + "grad_norm": 4.664148330688477, + "learning_rate": 8.209782317909844e-05, + "loss": 0.6768, + "step": 23778 + }, + { + "epoch": 0.5574172072879856, + "grad_norm": 4.2998809814453125, + "learning_rate": 8.209057606928918e-05, + "loss": 0.3904, + "step": 23779 + }, + { + "epoch": 0.5574406488627907, + "grad_norm": 3.4692564010620117, + "learning_rate": 8.208332905665697e-05, + "loss": 0.2554, + "step": 23780 + }, + { + "epoch": 0.5574640904375956, + "grad_norm": 2.575514078140259, + "learning_rate": 8.207608214124108e-05, + "loss": 1.5732, + "step": 23781 + }, + { + "epoch": 0.5574875320124006, + "grad_norm": 2.1760997772216797, + "learning_rate": 8.206883532308089e-05, + "loss": 0.1428, + "step": 23782 + }, + { + "epoch": 0.5575109735872056, + "grad_norm": 3.1515417098999023, + "learning_rate": 8.206158860221569e-05, + "loss": 0.2769, + "step": 23783 + }, + { + "epoch": 0.5575344151620106, + "grad_norm": 1.4840648174285889, + "learning_rate": 8.205434197868484e-05, + "loss": 0.138, + "step": 23784 + }, + { + "epoch": 0.5575578567368156, + "grad_norm": 2.8412113189697266, + "learning_rate": 8.204709545252759e-05, + "loss": 0.2153, + "step": 23785 + }, + { + "epoch": 0.5575812983116206, + "grad_norm": 1.0836338996887207, + "learning_rate": 8.203984902378334e-05, + "loss": 0.0617, + "step": 23786 + }, + { + "epoch": 0.5576047398864256, + "grad_norm": 0.36780646443367004, + "learning_rate": 8.203260269249139e-05, + "loss": 0.0486, + "step": 23787 + }, + { + "epoch": 0.5576281814612306, + "grad_norm": 4.66978120803833, + "learning_rate": 8.2025356458691e-05, + "loss": 0.3771, + "step": 23788 + }, + { + "epoch": 0.5576516230360355, + "grad_norm": 1.4205960035324097, + "learning_rate": 8.201811032242157e-05, + "loss": 0.1784, + "step": 23789 + }, + { + "epoch": 0.5576750646108406, + "grad_norm": 2.8102829456329346, + "learning_rate": 8.201086428372236e-05, + "loss": 0.1397, + "step": 23790 + }, + { + "epoch": 0.5576985061856455, + "grad_norm": 3.4415476322174072, + "learning_rate": 8.200361834263272e-05, + "loss": 0.311, + "step": 23791 + }, + { + "epoch": 0.5577219477604506, + "grad_norm": 3.394787549972534, + "learning_rate": 8.199637249919194e-05, + "loss": 0.2053, + "step": 23792 + }, + { + "epoch": 0.5577453893352555, + "grad_norm": 3.7747175693511963, + "learning_rate": 8.198912675343935e-05, + "loss": 0.3897, + "step": 23793 + }, + { + "epoch": 0.5577688309100606, + "grad_norm": 1.5842170715332031, + "learning_rate": 8.198188110541427e-05, + "loss": 0.9664, + "step": 23794 + }, + { + "epoch": 0.5577922724848655, + "grad_norm": 3.6260881423950195, + "learning_rate": 8.197463555515603e-05, + "loss": 0.4417, + "step": 23795 + }, + { + "epoch": 0.5578157140596706, + "grad_norm": 3.1509816646575928, + "learning_rate": 8.196739010270388e-05, + "loss": 1.5413, + "step": 23796 + }, + { + "epoch": 0.5578391556344755, + "grad_norm": 2.4899556636810303, + "learning_rate": 8.196014474809717e-05, + "loss": 1.9279, + "step": 23797 + }, + { + "epoch": 0.5578625972092806, + "grad_norm": 2.9872567653656006, + "learning_rate": 8.195289949137525e-05, + "loss": 0.2551, + "step": 23798 + }, + { + "epoch": 0.5578860387840855, + "grad_norm": 0.8190516829490662, + "learning_rate": 8.194565433257741e-05, + "loss": 0.0482, + "step": 23799 + }, + { + "epoch": 0.5579094803588905, + "grad_norm": 0.8601893186569214, + "learning_rate": 8.193840927174295e-05, + "loss": 0.0923, + "step": 23800 + }, + { + "epoch": 0.5579329219336955, + "grad_norm": 3.1371874809265137, + "learning_rate": 8.19311643089112e-05, + "loss": 0.3145, + "step": 23801 + }, + { + "epoch": 0.5579563635085005, + "grad_norm": 2.601854085922241, + "learning_rate": 8.192391944412146e-05, + "loss": 0.1757, + "step": 23802 + }, + { + "epoch": 0.5579798050833055, + "grad_norm": 0.22886136174201965, + "learning_rate": 8.191667467741302e-05, + "loss": 0.0193, + "step": 23803 + }, + { + "epoch": 0.5580032466581105, + "grad_norm": 2.271682024002075, + "learning_rate": 8.190943000882524e-05, + "loss": 0.1333, + "step": 23804 + }, + { + "epoch": 0.5580266882329155, + "grad_norm": 1.754576563835144, + "learning_rate": 8.19021854383974e-05, + "loss": 0.2038, + "step": 23805 + }, + { + "epoch": 0.5580501298077205, + "grad_norm": 2.3654847145080566, + "learning_rate": 8.189494096616879e-05, + "loss": 0.301, + "step": 23806 + }, + { + "epoch": 0.5580735713825254, + "grad_norm": 1.3911428451538086, + "learning_rate": 8.188769659217875e-05, + "loss": 0.1708, + "step": 23807 + }, + { + "epoch": 0.5580970129573305, + "grad_norm": 2.4431581497192383, + "learning_rate": 8.188045231646658e-05, + "loss": 0.2661, + "step": 23808 + }, + { + "epoch": 0.5581204545321354, + "grad_norm": 3.104973554611206, + "learning_rate": 8.187320813907155e-05, + "loss": 0.2029, + "step": 23809 + }, + { + "epoch": 0.5581438961069405, + "grad_norm": 0.7607009410858154, + "learning_rate": 8.186596406003302e-05, + "loss": 0.1112, + "step": 23810 + }, + { + "epoch": 0.5581673376817454, + "grad_norm": 3.399508237838745, + "learning_rate": 8.185872007939031e-05, + "loss": 0.3683, + "step": 23811 + }, + { + "epoch": 0.5581907792565505, + "grad_norm": 0.39596548676490784, + "learning_rate": 8.185147619718268e-05, + "loss": 0.0315, + "step": 23812 + }, + { + "epoch": 0.5582142208313554, + "grad_norm": 2.1132004261016846, + "learning_rate": 8.184423241344943e-05, + "loss": 0.1326, + "step": 23813 + }, + { + "epoch": 0.5582376624061605, + "grad_norm": 1.5958958864212036, + "learning_rate": 8.183698872822991e-05, + "loss": 0.1238, + "step": 23814 + }, + { + "epoch": 0.5582611039809654, + "grad_norm": 2.2796168327331543, + "learning_rate": 8.182974514156341e-05, + "loss": 0.149, + "step": 23815 + }, + { + "epoch": 0.5582845455557705, + "grad_norm": 1.778787612915039, + "learning_rate": 8.182250165348919e-05, + "loss": 0.108, + "step": 23816 + }, + { + "epoch": 0.5583079871305754, + "grad_norm": 1.5757067203521729, + "learning_rate": 8.181525826404661e-05, + "loss": 0.1414, + "step": 23817 + }, + { + "epoch": 0.5583314287053804, + "grad_norm": 2.669140100479126, + "learning_rate": 8.180801497327495e-05, + "loss": 0.35, + "step": 23818 + }, + { + "epoch": 0.5583548702801854, + "grad_norm": 2.226752758026123, + "learning_rate": 8.18007717812135e-05, + "loss": 0.1747, + "step": 23819 + }, + { + "epoch": 0.5583783118549904, + "grad_norm": 3.3021609783172607, + "learning_rate": 8.179352868790158e-05, + "loss": 0.4219, + "step": 23820 + }, + { + "epoch": 0.5584017534297954, + "grad_norm": 0.44367465376853943, + "learning_rate": 8.178628569337847e-05, + "loss": 0.0351, + "step": 23821 + }, + { + "epoch": 0.5584251950046004, + "grad_norm": 2.609551429748535, + "learning_rate": 8.177904279768349e-05, + "loss": 0.2675, + "step": 23822 + }, + { + "epoch": 0.5584486365794054, + "grad_norm": 3.311293125152588, + "learning_rate": 8.177180000085594e-05, + "loss": 0.1149, + "step": 23823 + }, + { + "epoch": 0.5584720781542104, + "grad_norm": 3.064603328704834, + "learning_rate": 8.176455730293514e-05, + "loss": 0.1794, + "step": 23824 + }, + { + "epoch": 0.5584955197290153, + "grad_norm": 3.079650402069092, + "learning_rate": 8.175731470396036e-05, + "loss": 0.2109, + "step": 23825 + }, + { + "epoch": 0.5585189613038204, + "grad_norm": 4.518723011016846, + "learning_rate": 8.175007220397087e-05, + "loss": 0.4718, + "step": 23826 + }, + { + "epoch": 0.5585424028786253, + "grad_norm": 3.282244920730591, + "learning_rate": 8.174282980300603e-05, + "loss": 0.3134, + "step": 23827 + }, + { + "epoch": 0.5585658444534304, + "grad_norm": 0.7936842441558838, + "learning_rate": 8.173558750110509e-05, + "loss": 0.0512, + "step": 23828 + }, + { + "epoch": 0.5585892860282354, + "grad_norm": 1.9404245615005493, + "learning_rate": 8.172834529830739e-05, + "loss": 0.1868, + "step": 23829 + }, + { + "epoch": 0.5586127276030404, + "grad_norm": 3.975358247756958, + "learning_rate": 8.172110319465219e-05, + "loss": 0.347, + "step": 23830 + }, + { + "epoch": 0.5586361691778454, + "grad_norm": 0.693609893321991, + "learning_rate": 8.171386119017879e-05, + "loss": 0.0637, + "step": 23831 + }, + { + "epoch": 0.5586596107526504, + "grad_norm": 3.930988311767578, + "learning_rate": 8.170661928492651e-05, + "loss": 0.4795, + "step": 23832 + }, + { + "epoch": 0.5586830523274554, + "grad_norm": 4.420318126678467, + "learning_rate": 8.169937747893458e-05, + "loss": 0.3449, + "step": 23833 + }, + { + "epoch": 0.5587064939022603, + "grad_norm": 1.2029101848602295, + "learning_rate": 8.16921357722424e-05, + "loss": 0.106, + "step": 23834 + }, + { + "epoch": 0.5587299354770654, + "grad_norm": 2.954523801803589, + "learning_rate": 8.168489416488916e-05, + "loss": 0.1926, + "step": 23835 + }, + { + "epoch": 0.5587533770518703, + "grad_norm": 2.2857227325439453, + "learning_rate": 8.167765265691424e-05, + "loss": 0.2465, + "step": 23836 + }, + { + "epoch": 0.5587768186266754, + "grad_norm": 2.4467334747314453, + "learning_rate": 8.167041124835688e-05, + "loss": 0.3144, + "step": 23837 + }, + { + "epoch": 0.5588002602014803, + "grad_norm": 1.450255274772644, + "learning_rate": 8.166316993925637e-05, + "loss": 0.1574, + "step": 23838 + }, + { + "epoch": 0.5588237017762854, + "grad_norm": 4.868868827819824, + "learning_rate": 8.165592872965204e-05, + "loss": 0.3256, + "step": 23839 + }, + { + "epoch": 0.5588471433510903, + "grad_norm": 2.767005681991577, + "learning_rate": 8.164868761958315e-05, + "loss": 0.2767, + "step": 23840 + }, + { + "epoch": 0.5588705849258954, + "grad_norm": 1.0335322618484497, + "learning_rate": 8.164144660908898e-05, + "loss": 0.1204, + "step": 23841 + }, + { + "epoch": 0.5588940265007003, + "grad_norm": 1.469642996788025, + "learning_rate": 8.163420569820886e-05, + "loss": 0.1007, + "step": 23842 + }, + { + "epoch": 0.5589174680755054, + "grad_norm": 0.22919391095638275, + "learning_rate": 8.162696488698205e-05, + "loss": 0.013, + "step": 23843 + }, + { + "epoch": 0.5589409096503103, + "grad_norm": 2.662060022354126, + "learning_rate": 8.161972417544783e-05, + "loss": 1.291, + "step": 23844 + }, + { + "epoch": 0.5589643512251153, + "grad_norm": 2.5512406826019287, + "learning_rate": 8.161248356364552e-05, + "loss": 0.2011, + "step": 23845 + }, + { + "epoch": 0.5589877927999203, + "grad_norm": 2.948495388031006, + "learning_rate": 8.160524305161436e-05, + "loss": 0.1251, + "step": 23846 + }, + { + "epoch": 0.5590112343747253, + "grad_norm": 2.421462297439575, + "learning_rate": 8.15980026393937e-05, + "loss": 2.0077, + "step": 23847 + }, + { + "epoch": 0.5590346759495303, + "grad_norm": 4.352869987487793, + "learning_rate": 8.159076232702277e-05, + "loss": 0.4263, + "step": 23848 + }, + { + "epoch": 0.5590581175243353, + "grad_norm": 4.05947732925415, + "learning_rate": 8.15835221145409e-05, + "loss": 0.3021, + "step": 23849 + }, + { + "epoch": 0.5590815590991403, + "grad_norm": 3.5470314025878906, + "learning_rate": 8.157628200198737e-05, + "loss": 0.1912, + "step": 23850 + }, + { + "epoch": 0.5591050006739453, + "grad_norm": 1.4099395275115967, + "learning_rate": 8.156904198940144e-05, + "loss": 0.1495, + "step": 23851 + }, + { + "epoch": 0.5591284422487502, + "grad_norm": 1.8376318216323853, + "learning_rate": 8.15618020768224e-05, + "loss": 0.2253, + "step": 23852 + }, + { + "epoch": 0.5591518838235553, + "grad_norm": 3.475536346435547, + "learning_rate": 8.155456226428954e-05, + "loss": 0.3449, + "step": 23853 + }, + { + "epoch": 0.5591753253983602, + "grad_norm": 2.3975610733032227, + "learning_rate": 8.154732255184214e-05, + "loss": 0.2123, + "step": 23854 + }, + { + "epoch": 0.5591987669731653, + "grad_norm": 4.80748987197876, + "learning_rate": 8.154008293951952e-05, + "loss": 0.6076, + "step": 23855 + }, + { + "epoch": 0.5592222085479702, + "grad_norm": 2.7616093158721924, + "learning_rate": 8.15328434273609e-05, + "loss": 0.0926, + "step": 23856 + }, + { + "epoch": 0.5592456501227753, + "grad_norm": 4.606760501861572, + "learning_rate": 8.152560401540559e-05, + "loss": 0.2768, + "step": 23857 + }, + { + "epoch": 0.5592690916975802, + "grad_norm": 2.752319812774658, + "learning_rate": 8.151836470369282e-05, + "loss": 0.2544, + "step": 23858 + }, + { + "epoch": 0.5592925332723853, + "grad_norm": 1.1756489276885986, + "learning_rate": 8.151112549226201e-05, + "loss": 0.1, + "step": 23859 + }, + { + "epoch": 0.5593159748471902, + "grad_norm": 5.259437561035156, + "learning_rate": 8.150388638115232e-05, + "loss": 0.3655, + "step": 23860 + }, + { + "epoch": 0.5593394164219953, + "grad_norm": 3.1532704830169678, + "learning_rate": 8.149664737040306e-05, + "loss": 0.3459, + "step": 23861 + }, + { + "epoch": 0.5593628579968002, + "grad_norm": 2.4173309803009033, + "learning_rate": 8.148940846005353e-05, + "loss": 0.2494, + "step": 23862 + }, + { + "epoch": 0.5593862995716052, + "grad_norm": 2.3498120307922363, + "learning_rate": 8.148216965014298e-05, + "loss": 0.2041, + "step": 23863 + }, + { + "epoch": 0.5594097411464102, + "grad_norm": 0.7495884299278259, + "learning_rate": 8.147493094071068e-05, + "loss": 0.0727, + "step": 23864 + }, + { + "epoch": 0.5594331827212152, + "grad_norm": 2.201958179473877, + "learning_rate": 8.146769233179595e-05, + "loss": 0.1319, + "step": 23865 + }, + { + "epoch": 0.5594566242960202, + "grad_norm": 3.7560365200042725, + "learning_rate": 8.146045382343803e-05, + "loss": 0.4734, + "step": 23866 + }, + { + "epoch": 0.5594800658708252, + "grad_norm": 9.394489288330078, + "learning_rate": 8.14532154156762e-05, + "loss": 0.406, + "step": 23867 + }, + { + "epoch": 0.5595035074456302, + "grad_norm": 2.9822123050689697, + "learning_rate": 8.144597710854978e-05, + "loss": 0.1415, + "step": 23868 + }, + { + "epoch": 0.5595269490204352, + "grad_norm": 3.183232069015503, + "learning_rate": 8.143873890209797e-05, + "loss": 0.2079, + "step": 23869 + }, + { + "epoch": 0.5595503905952401, + "grad_norm": 3.0053675174713135, + "learning_rate": 8.143150079636006e-05, + "loss": 1.8761, + "step": 23870 + }, + { + "epoch": 0.5595738321700452, + "grad_norm": 0.9790275692939758, + "learning_rate": 8.142426279137541e-05, + "loss": 0.0978, + "step": 23871 + }, + { + "epoch": 0.5595972737448501, + "grad_norm": 1.5865370035171509, + "learning_rate": 8.141702488718321e-05, + "loss": 0.1727, + "step": 23872 + }, + { + "epoch": 0.5596207153196552, + "grad_norm": 0.26468542218208313, + "learning_rate": 8.140978708382275e-05, + "loss": 0.0195, + "step": 23873 + }, + { + "epoch": 0.5596441568944601, + "grad_norm": 0.777741551399231, + "learning_rate": 8.140254938133333e-05, + "loss": 0.0974, + "step": 23874 + }, + { + "epoch": 0.5596675984692652, + "grad_norm": 2.505632162094116, + "learning_rate": 8.13953117797542e-05, + "loss": 0.2541, + "step": 23875 + }, + { + "epoch": 0.5596910400440701, + "grad_norm": 1.6363706588745117, + "learning_rate": 8.13880742791246e-05, + "loss": 0.124, + "step": 23876 + }, + { + "epoch": 0.5597144816188752, + "grad_norm": 2.456218719482422, + "learning_rate": 8.138083687948385e-05, + "loss": 0.2501, + "step": 23877 + }, + { + "epoch": 0.5597379231936801, + "grad_norm": 5.430382251739502, + "learning_rate": 8.137359958087121e-05, + "loss": 0.3014, + "step": 23878 + }, + { + "epoch": 0.5597613647684851, + "grad_norm": 1.7164937257766724, + "learning_rate": 8.136636238332592e-05, + "loss": 1.3878, + "step": 23879 + }, + { + "epoch": 0.5597848063432902, + "grad_norm": 2.525651216506958, + "learning_rate": 8.135912528688729e-05, + "loss": 1.8789, + "step": 23880 + }, + { + "epoch": 0.5598082479180951, + "grad_norm": 2.429497718811035, + "learning_rate": 8.135188829159458e-05, + "loss": 0.2186, + "step": 23881 + }, + { + "epoch": 0.5598316894929002, + "grad_norm": 0.8566527366638184, + "learning_rate": 8.1344651397487e-05, + "loss": 0.0695, + "step": 23882 + }, + { + "epoch": 0.5598551310677051, + "grad_norm": 1.974945068359375, + "learning_rate": 8.133741460460389e-05, + "loss": 0.2633, + "step": 23883 + }, + { + "epoch": 0.5598785726425102, + "grad_norm": 0.7833136320114136, + "learning_rate": 8.133017791298451e-05, + "loss": 0.0866, + "step": 23884 + }, + { + "epoch": 0.5599020142173151, + "grad_norm": 0.286942720413208, + "learning_rate": 8.13229413226681e-05, + "loss": 0.0142, + "step": 23885 + }, + { + "epoch": 0.5599254557921202, + "grad_norm": 2.536868095397949, + "learning_rate": 8.131570483369394e-05, + "loss": 0.2077, + "step": 23886 + }, + { + "epoch": 0.5599488973669251, + "grad_norm": 1.9705665111541748, + "learning_rate": 8.13084684461013e-05, + "loss": 0.1944, + "step": 23887 + }, + { + "epoch": 0.5599723389417302, + "grad_norm": 0.4709986448287964, + "learning_rate": 8.130123215992944e-05, + "loss": 0.0247, + "step": 23888 + }, + { + "epoch": 0.5599957805165351, + "grad_norm": 2.8750855922698975, + "learning_rate": 8.129399597521759e-05, + "loss": 0.3424, + "step": 23889 + }, + { + "epoch": 0.5600192220913401, + "grad_norm": 4.996434688568115, + "learning_rate": 8.128675989200507e-05, + "loss": 0.2443, + "step": 23890 + }, + { + "epoch": 0.5600426636661451, + "grad_norm": 3.5585217475891113, + "learning_rate": 8.127952391033111e-05, + "loss": 0.2152, + "step": 23891 + }, + { + "epoch": 0.5600661052409501, + "grad_norm": 1.6894556283950806, + "learning_rate": 8.127228803023496e-05, + "loss": 0.1041, + "step": 23892 + }, + { + "epoch": 0.5600895468157551, + "grad_norm": 2.8268237113952637, + "learning_rate": 8.126505225175593e-05, + "loss": 0.1397, + "step": 23893 + }, + { + "epoch": 0.5601129883905601, + "grad_norm": 4.56774377822876, + "learning_rate": 8.125781657493321e-05, + "loss": 0.4636, + "step": 23894 + }, + { + "epoch": 0.5601364299653651, + "grad_norm": 3.706458330154419, + "learning_rate": 8.125058099980615e-05, + "loss": 0.2941, + "step": 23895 + }, + { + "epoch": 0.5601598715401701, + "grad_norm": 3.272298574447632, + "learning_rate": 8.124334552641393e-05, + "loss": 0.2602, + "step": 23896 + }, + { + "epoch": 0.560183313114975, + "grad_norm": 0.722972571849823, + "learning_rate": 8.123611015479588e-05, + "loss": 0.0736, + "step": 23897 + }, + { + "epoch": 0.5602067546897801, + "grad_norm": 1.9496697187423706, + "learning_rate": 8.122887488499121e-05, + "loss": 0.12, + "step": 23898 + }, + { + "epoch": 0.560230196264585, + "grad_norm": 2.1409661769866943, + "learning_rate": 8.122163971703918e-05, + "loss": 0.1425, + "step": 23899 + }, + { + "epoch": 0.5602536378393901, + "grad_norm": 1.149216890335083, + "learning_rate": 8.121440465097908e-05, + "loss": 0.054, + "step": 23900 + }, + { + "epoch": 0.560277079414195, + "grad_norm": 1.3405061960220337, + "learning_rate": 8.120716968685013e-05, + "loss": 0.1625, + "step": 23901 + }, + { + "epoch": 0.5603005209890001, + "grad_norm": 0.9225391149520874, + "learning_rate": 8.119993482469162e-05, + "loss": 0.0745, + "step": 23902 + }, + { + "epoch": 0.560323962563805, + "grad_norm": 2.4820494651794434, + "learning_rate": 8.11927000645428e-05, + "loss": 0.3719, + "step": 23903 + }, + { + "epoch": 0.5603474041386101, + "grad_norm": 2.5992581844329834, + "learning_rate": 8.118546540644289e-05, + "loss": 0.1392, + "step": 23904 + }, + { + "epoch": 0.560370845713415, + "grad_norm": 2.9902775287628174, + "learning_rate": 8.11782308504312e-05, + "loss": 1.4139, + "step": 23905 + }, + { + "epoch": 0.56039428728822, + "grad_norm": 3.6403446197509766, + "learning_rate": 8.117099639654693e-05, + "loss": 0.26, + "step": 23906 + }, + { + "epoch": 0.560417728863025, + "grad_norm": 2.4773173332214355, + "learning_rate": 8.116376204482939e-05, + "loss": 0.2095, + "step": 23907 + }, + { + "epoch": 0.56044117043783, + "grad_norm": 3.7122769355773926, + "learning_rate": 8.115652779531778e-05, + "loss": 0.4062, + "step": 23908 + }, + { + "epoch": 0.560464612012635, + "grad_norm": 2.033890962600708, + "learning_rate": 8.114929364805141e-05, + "loss": 0.2162, + "step": 23909 + }, + { + "epoch": 0.56048805358744, + "grad_norm": 3.502797842025757, + "learning_rate": 8.11420596030695e-05, + "loss": 1.715, + "step": 23910 + }, + { + "epoch": 0.560511495162245, + "grad_norm": 1.2342246770858765, + "learning_rate": 8.113482566041128e-05, + "loss": 0.0724, + "step": 23911 + }, + { + "epoch": 0.56053493673705, + "grad_norm": 3.74794340133667, + "learning_rate": 8.112759182011605e-05, + "loss": 0.3849, + "step": 23912 + }, + { + "epoch": 0.560558378311855, + "grad_norm": 0.34768638014793396, + "learning_rate": 8.112035808222305e-05, + "loss": 0.014, + "step": 23913 + }, + { + "epoch": 0.56058181988666, + "grad_norm": 0.5557196736335754, + "learning_rate": 8.111312444677149e-05, + "loss": 0.0187, + "step": 23914 + }, + { + "epoch": 0.5606052614614649, + "grad_norm": 4.073975563049316, + "learning_rate": 8.110589091380065e-05, + "loss": 0.4958, + "step": 23915 + }, + { + "epoch": 0.56062870303627, + "grad_norm": 2.900604724884033, + "learning_rate": 8.109865748334979e-05, + "loss": 0.2673, + "step": 23916 + }, + { + "epoch": 0.5606521446110749, + "grad_norm": 1.2140027284622192, + "learning_rate": 8.109142415545813e-05, + "loss": 0.1926, + "step": 23917 + }, + { + "epoch": 0.56067558618588, + "grad_norm": 1.8245407342910767, + "learning_rate": 8.10841909301649e-05, + "loss": 0.1673, + "step": 23918 + }, + { + "epoch": 0.5606990277606849, + "grad_norm": 4.662491321563721, + "learning_rate": 8.107695780750945e-05, + "loss": 0.4103, + "step": 23919 + }, + { + "epoch": 0.56072246933549, + "grad_norm": 1.204969048500061, + "learning_rate": 8.106972478753094e-05, + "loss": 0.0932, + "step": 23920 + }, + { + "epoch": 0.5607459109102949, + "grad_norm": 3.5134971141815186, + "learning_rate": 8.106249187026861e-05, + "loss": 0.1234, + "step": 23921 + }, + { + "epoch": 0.5607693524851, + "grad_norm": 2.3302266597747803, + "learning_rate": 8.105525905576176e-05, + "loss": 0.2531, + "step": 23922 + }, + { + "epoch": 0.5607927940599049, + "grad_norm": 5.117525577545166, + "learning_rate": 8.104802634404961e-05, + "loss": 0.7161, + "step": 23923 + }, + { + "epoch": 0.56081623563471, + "grad_norm": 2.662158727645874, + "learning_rate": 8.104079373517138e-05, + "loss": 0.1988, + "step": 23924 + }, + { + "epoch": 0.5608396772095149, + "grad_norm": 3.029110908508301, + "learning_rate": 8.103356122916635e-05, + "loss": 0.3269, + "step": 23925 + }, + { + "epoch": 0.5608631187843199, + "grad_norm": 1.5583999156951904, + "learning_rate": 8.102632882607376e-05, + "loss": 0.1838, + "step": 23926 + }, + { + "epoch": 0.5608865603591249, + "grad_norm": 1.8640820980072021, + "learning_rate": 8.101909652593283e-05, + "loss": 0.08, + "step": 23927 + }, + { + "epoch": 0.5609100019339299, + "grad_norm": 3.2423999309539795, + "learning_rate": 8.101186432878282e-05, + "loss": 0.2214, + "step": 23928 + }, + { + "epoch": 0.5609334435087349, + "grad_norm": 4.244804382324219, + "learning_rate": 8.100463223466296e-05, + "loss": 0.3366, + "step": 23929 + }, + { + "epoch": 0.5609568850835399, + "grad_norm": 1.1404740810394287, + "learning_rate": 8.09974002436125e-05, + "loss": 0.159, + "step": 23930 + }, + { + "epoch": 0.5609803266583449, + "grad_norm": 2.6548030376434326, + "learning_rate": 8.099016835567065e-05, + "loss": 0.2237, + "step": 23931 + }, + { + "epoch": 0.5610037682331499, + "grad_norm": 2.745333194732666, + "learning_rate": 8.098293657087673e-05, + "loss": 0.2012, + "step": 23932 + }, + { + "epoch": 0.561027209807955, + "grad_norm": 2.4333455562591553, + "learning_rate": 8.097570488926991e-05, + "loss": 0.161, + "step": 23933 + }, + { + "epoch": 0.5610506513827599, + "grad_norm": 1.2858083248138428, + "learning_rate": 8.096847331088945e-05, + "loss": 0.0977, + "step": 23934 + }, + { + "epoch": 0.5610740929575649, + "grad_norm": 3.042797088623047, + "learning_rate": 8.09612418357746e-05, + "loss": 0.2245, + "step": 23935 + }, + { + "epoch": 0.5610975345323699, + "grad_norm": 1.3582618236541748, + "learning_rate": 8.095401046396459e-05, + "loss": 0.1379, + "step": 23936 + }, + { + "epoch": 0.5611209761071749, + "grad_norm": 0.7254873514175415, + "learning_rate": 8.094677919549864e-05, + "loss": 0.06, + "step": 23937 + }, + { + "epoch": 0.5611444176819799, + "grad_norm": 2.5604677200317383, + "learning_rate": 8.093954803041602e-05, + "loss": 1.7388, + "step": 23938 + }, + { + "epoch": 0.5611678592567849, + "grad_norm": 4.545814037322998, + "learning_rate": 8.093231696875592e-05, + "loss": 0.236, + "step": 23939 + }, + { + "epoch": 0.5611913008315899, + "grad_norm": 2.73159122467041, + "learning_rate": 8.092508601055763e-05, + "loss": 0.2297, + "step": 23940 + }, + { + "epoch": 0.5612147424063949, + "grad_norm": 0.8185508251190186, + "learning_rate": 8.091785515586036e-05, + "loss": 0.0797, + "step": 23941 + }, + { + "epoch": 0.5612381839811998, + "grad_norm": 2.5931265354156494, + "learning_rate": 8.091062440470333e-05, + "loss": 0.253, + "step": 23942 + }, + { + "epoch": 0.5612616255560049, + "grad_norm": 4.117819786071777, + "learning_rate": 8.090339375712578e-05, + "loss": 0.2355, + "step": 23943 + }, + { + "epoch": 0.5612850671308098, + "grad_norm": 1.5596363544464111, + "learning_rate": 8.089616321316696e-05, + "loss": 0.0617, + "step": 23944 + }, + { + "epoch": 0.5613085087056149, + "grad_norm": 3.9424479007720947, + "learning_rate": 8.08889327728661e-05, + "loss": 0.1562, + "step": 23945 + }, + { + "epoch": 0.5613319502804198, + "grad_norm": 4.035557270050049, + "learning_rate": 8.088170243626242e-05, + "loss": 0.3368, + "step": 23946 + }, + { + "epoch": 0.5613553918552249, + "grad_norm": 3.0109920501708984, + "learning_rate": 8.087447220339519e-05, + "loss": 0.2704, + "step": 23947 + }, + { + "epoch": 0.5613788334300298, + "grad_norm": 3.3503081798553467, + "learning_rate": 8.086724207430358e-05, + "loss": 0.2817, + "step": 23948 + }, + { + "epoch": 0.5614022750048349, + "grad_norm": 1.4003870487213135, + "learning_rate": 8.086001204902686e-05, + "loss": 0.114, + "step": 23949 + }, + { + "epoch": 0.5614257165796398, + "grad_norm": 3.4064695835113525, + "learning_rate": 8.085278212760426e-05, + "loss": 0.3916, + "step": 23950 + }, + { + "epoch": 0.5614491581544448, + "grad_norm": 4.178652286529541, + "learning_rate": 8.084555231007502e-05, + "loss": 0.2829, + "step": 23951 + }, + { + "epoch": 0.5614725997292498, + "grad_norm": 2.748039722442627, + "learning_rate": 8.083832259647833e-05, + "loss": 0.0847, + "step": 23952 + }, + { + "epoch": 0.5614960413040548, + "grad_norm": 3.2382287979125977, + "learning_rate": 8.083109298685345e-05, + "loss": 0.2779, + "step": 23953 + }, + { + "epoch": 0.5615194828788598, + "grad_norm": 2.2618932723999023, + "learning_rate": 8.082386348123958e-05, + "loss": 0.1512, + "step": 23954 + }, + { + "epoch": 0.5615429244536648, + "grad_norm": 6.303078651428223, + "learning_rate": 8.081663407967595e-05, + "loss": 0.559, + "step": 23955 + }, + { + "epoch": 0.5615663660284698, + "grad_norm": 0.3565097153186798, + "learning_rate": 8.080940478220181e-05, + "loss": 0.0251, + "step": 23956 + }, + { + "epoch": 0.5615898076032748, + "grad_norm": 0.58465176820755, + "learning_rate": 8.08021755888564e-05, + "loss": 0.0684, + "step": 23957 + }, + { + "epoch": 0.5616132491780798, + "grad_norm": 0.4895346760749817, + "learning_rate": 8.079494649967895e-05, + "loss": 0.0283, + "step": 23958 + }, + { + "epoch": 0.5616366907528848, + "grad_norm": 3.0051510334014893, + "learning_rate": 8.078771751470862e-05, + "loss": 1.3988, + "step": 23959 + }, + { + "epoch": 0.5616601323276897, + "grad_norm": 2.688767433166504, + "learning_rate": 8.07804886339847e-05, + "loss": 0.2775, + "step": 23960 + }, + { + "epoch": 0.5616835739024948, + "grad_norm": 1.6637651920318604, + "learning_rate": 8.077325985754638e-05, + "loss": 0.1844, + "step": 23961 + }, + { + "epoch": 0.5617070154772997, + "grad_norm": 3.8613712787628174, + "learning_rate": 8.076603118543288e-05, + "loss": 0.4213, + "step": 23962 + }, + { + "epoch": 0.5617304570521048, + "grad_norm": 1.8968698978424072, + "learning_rate": 8.075880261768344e-05, + "loss": 1.0616, + "step": 23963 + }, + { + "epoch": 0.5617538986269097, + "grad_norm": 1.1738396883010864, + "learning_rate": 8.07515741543373e-05, + "loss": 0.112, + "step": 23964 + }, + { + "epoch": 0.5617773402017148, + "grad_norm": 3.0290040969848633, + "learning_rate": 8.074434579543364e-05, + "loss": 1.5179, + "step": 23965 + }, + { + "epoch": 0.5618007817765197, + "grad_norm": 1.5865398645401, + "learning_rate": 8.073711754101171e-05, + "loss": 0.1345, + "step": 23966 + }, + { + "epoch": 0.5618242233513248, + "grad_norm": 2.2147645950317383, + "learning_rate": 8.072988939111068e-05, + "loss": 0.0908, + "step": 23967 + }, + { + "epoch": 0.5618476649261297, + "grad_norm": 2.9620790481567383, + "learning_rate": 8.072266134576987e-05, + "loss": 0.172, + "step": 23968 + }, + { + "epoch": 0.5618711065009347, + "grad_norm": 5.483488082885742, + "learning_rate": 8.07154334050284e-05, + "loss": 0.9988, + "step": 23969 + }, + { + "epoch": 0.5618945480757397, + "grad_norm": 3.486335039138794, + "learning_rate": 8.070820556892556e-05, + "loss": 0.2509, + "step": 23970 + }, + { + "epoch": 0.5619179896505447, + "grad_norm": 1.647040605545044, + "learning_rate": 8.070097783750053e-05, + "loss": 0.0749, + "step": 23971 + }, + { + "epoch": 0.5619414312253497, + "grad_norm": 3.485252618789673, + "learning_rate": 8.069375021079253e-05, + "loss": 0.371, + "step": 23972 + }, + { + "epoch": 0.5619648728001547, + "grad_norm": 2.500999927520752, + "learning_rate": 8.068652268884081e-05, + "loss": 0.171, + "step": 23973 + }, + { + "epoch": 0.5619883143749597, + "grad_norm": 3.4015419483184814, + "learning_rate": 8.067929527168454e-05, + "loss": 0.3553, + "step": 23974 + }, + { + "epoch": 0.5620117559497647, + "grad_norm": 0.9342315196990967, + "learning_rate": 8.067206795936295e-05, + "loss": 0.0304, + "step": 23975 + }, + { + "epoch": 0.5620351975245697, + "grad_norm": 3.4112236499786377, + "learning_rate": 8.066484075191528e-05, + "loss": 0.3398, + "step": 23976 + }, + { + "epoch": 0.5620586390993747, + "grad_norm": 0.8167096972465515, + "learning_rate": 8.065761364938071e-05, + "loss": 0.0322, + "step": 23977 + }, + { + "epoch": 0.5620820806741796, + "grad_norm": 1.1238685846328735, + "learning_rate": 8.065038665179849e-05, + "loss": 0.0892, + "step": 23978 + }, + { + "epoch": 0.5621055222489847, + "grad_norm": 2.2300844192504883, + "learning_rate": 8.064315975920778e-05, + "loss": 0.1014, + "step": 23979 + }, + { + "epoch": 0.5621289638237896, + "grad_norm": 0.42270970344543457, + "learning_rate": 8.063593297164787e-05, + "loss": 0.0409, + "step": 23980 + }, + { + "epoch": 0.5621524053985947, + "grad_norm": 1.1143364906311035, + "learning_rate": 8.062870628915792e-05, + "loss": 0.1132, + "step": 23981 + }, + { + "epoch": 0.5621758469733996, + "grad_norm": 2.9399619102478027, + "learning_rate": 8.062147971177714e-05, + "loss": 0.1915, + "step": 23982 + }, + { + "epoch": 0.5621992885482047, + "grad_norm": 3.2185816764831543, + "learning_rate": 8.061425323954478e-05, + "loss": 0.2321, + "step": 23983 + }, + { + "epoch": 0.5622227301230097, + "grad_norm": 1.0047153234481812, + "learning_rate": 8.06070268725e-05, + "loss": 0.1051, + "step": 23984 + }, + { + "epoch": 0.5622461716978147, + "grad_norm": 2.1523592472076416, + "learning_rate": 8.059980061068206e-05, + "loss": 0.2411, + "step": 23985 + }, + { + "epoch": 0.5622696132726197, + "grad_norm": 3.841320753097534, + "learning_rate": 8.059257445413016e-05, + "loss": 0.2877, + "step": 23986 + }, + { + "epoch": 0.5622930548474246, + "grad_norm": 2.8090908527374268, + "learning_rate": 8.058534840288346e-05, + "loss": 0.2662, + "step": 23987 + }, + { + "epoch": 0.5623164964222297, + "grad_norm": 3.3698346614837646, + "learning_rate": 8.057812245698123e-05, + "loss": 0.2699, + "step": 23988 + }, + { + "epoch": 0.5623399379970346, + "grad_norm": 2.9801762104034424, + "learning_rate": 8.057089661646267e-05, + "loss": 0.2373, + "step": 23989 + }, + { + "epoch": 0.5623633795718397, + "grad_norm": 2.6025686264038086, + "learning_rate": 8.056367088136692e-05, + "loss": 0.1472, + "step": 23990 + }, + { + "epoch": 0.5623868211466446, + "grad_norm": 3.3340160846710205, + "learning_rate": 8.055644525173323e-05, + "loss": 0.1524, + "step": 23991 + }, + { + "epoch": 0.5624102627214497, + "grad_norm": 4.174121856689453, + "learning_rate": 8.054921972760088e-05, + "loss": 0.4374, + "step": 23992 + }, + { + "epoch": 0.5624337042962546, + "grad_norm": 2.3954737186431885, + "learning_rate": 8.054199430900899e-05, + "loss": 0.0853, + "step": 23993 + }, + { + "epoch": 0.5624571458710597, + "grad_norm": 3.7342417240142822, + "learning_rate": 8.053476899599676e-05, + "loss": 0.725, + "step": 23994 + }, + { + "epoch": 0.5624805874458646, + "grad_norm": 0.7927690744400024, + "learning_rate": 8.052754378860344e-05, + "loss": 0.0777, + "step": 23995 + }, + { + "epoch": 0.5625040290206696, + "grad_norm": 4.443732738494873, + "learning_rate": 8.052031868686822e-05, + "loss": 0.4252, + "step": 23996 + }, + { + "epoch": 0.5625274705954746, + "grad_norm": 1.6842457056045532, + "learning_rate": 8.051309369083029e-05, + "loss": 0.7356, + "step": 23997 + }, + { + "epoch": 0.5625509121702796, + "grad_norm": 1.7132346630096436, + "learning_rate": 8.050586880052888e-05, + "loss": 0.1312, + "step": 23998 + }, + { + "epoch": 0.5625743537450846, + "grad_norm": 3.114593029022217, + "learning_rate": 8.049864401600315e-05, + "loss": 1.3878, + "step": 23999 + }, + { + "epoch": 0.5625977953198896, + "grad_norm": 3.4718921184539795, + "learning_rate": 8.049141933729233e-05, + "loss": 0.4165, + "step": 24000 + }, + { + "epoch": 0.5626212368946946, + "grad_norm": 3.173353910446167, + "learning_rate": 8.048419476443563e-05, + "loss": 0.2448, + "step": 24001 + }, + { + "epoch": 0.5626446784694996, + "grad_norm": 4.919990062713623, + "learning_rate": 8.047697029747225e-05, + "loss": 0.836, + "step": 24002 + }, + { + "epoch": 0.5626681200443046, + "grad_norm": 2.859560012817383, + "learning_rate": 8.046974593644134e-05, + "loss": 0.2407, + "step": 24003 + }, + { + "epoch": 0.5626915616191096, + "grad_norm": 1.9347732067108154, + "learning_rate": 8.046252168138212e-05, + "loss": 0.201, + "step": 24004 + }, + { + "epoch": 0.5627150031939145, + "grad_norm": 3.6197733879089355, + "learning_rate": 8.045529753233385e-05, + "loss": 0.3825, + "step": 24005 + }, + { + "epoch": 0.5627384447687196, + "grad_norm": 0.5678427815437317, + "learning_rate": 8.044807348933568e-05, + "loss": 0.0482, + "step": 24006 + }, + { + "epoch": 0.5627618863435245, + "grad_norm": 2.8748526573181152, + "learning_rate": 8.04408495524268e-05, + "loss": 0.3379, + "step": 24007 + }, + { + "epoch": 0.5627853279183296, + "grad_norm": 0.6265013813972473, + "learning_rate": 8.043362572164643e-05, + "loss": 0.0745, + "step": 24008 + }, + { + "epoch": 0.5628087694931345, + "grad_norm": 1.1189793348312378, + "learning_rate": 8.042640199703377e-05, + "loss": 0.0435, + "step": 24009 + }, + { + "epoch": 0.5628322110679396, + "grad_norm": 2.5502991676330566, + "learning_rate": 8.041917837862798e-05, + "loss": 0.2278, + "step": 24010 + }, + { + "epoch": 0.5628556526427445, + "grad_norm": 2.273728847503662, + "learning_rate": 8.04119548664683e-05, + "loss": 0.0781, + "step": 24011 + }, + { + "epoch": 0.5628790942175496, + "grad_norm": 1.1753824949264526, + "learning_rate": 8.04047314605939e-05, + "loss": 0.106, + "step": 24012 + }, + { + "epoch": 0.5629025357923545, + "grad_norm": 4.128522872924805, + "learning_rate": 8.039750816104396e-05, + "loss": 0.3704, + "step": 24013 + }, + { + "epoch": 0.5629259773671595, + "grad_norm": 0.845976710319519, + "learning_rate": 8.039028496785771e-05, + "loss": 0.0513, + "step": 24014 + }, + { + "epoch": 0.5629494189419645, + "grad_norm": 1.4127763509750366, + "learning_rate": 8.03830618810743e-05, + "loss": 0.1771, + "step": 24015 + }, + { + "epoch": 0.5629728605167695, + "grad_norm": 3.9918019771575928, + "learning_rate": 8.037583890073295e-05, + "loss": 0.349, + "step": 24016 + }, + { + "epoch": 0.5629963020915745, + "grad_norm": 3.1536245346069336, + "learning_rate": 8.036861602687286e-05, + "loss": 0.2009, + "step": 24017 + }, + { + "epoch": 0.5630197436663795, + "grad_norm": 3.480705738067627, + "learning_rate": 8.036139325953323e-05, + "loss": 0.4045, + "step": 24018 + }, + { + "epoch": 0.5630431852411845, + "grad_norm": 2.6628589630126953, + "learning_rate": 8.03541705987532e-05, + "loss": 0.1368, + "step": 24019 + }, + { + "epoch": 0.5630666268159895, + "grad_norm": 3.6289727687835693, + "learning_rate": 8.034694804457203e-05, + "loss": 0.3581, + "step": 24020 + }, + { + "epoch": 0.5630900683907945, + "grad_norm": 0.6084108352661133, + "learning_rate": 8.033972559702886e-05, + "loss": 0.0413, + "step": 24021 + }, + { + "epoch": 0.5631135099655995, + "grad_norm": 0.9768022298812866, + "learning_rate": 8.033250325616287e-05, + "loss": 0.0762, + "step": 24022 + }, + { + "epoch": 0.5631369515404044, + "grad_norm": 2.283679246902466, + "learning_rate": 8.03252810220133e-05, + "loss": 0.1632, + "step": 24023 + }, + { + "epoch": 0.5631603931152095, + "grad_norm": 3.635840654373169, + "learning_rate": 8.03180588946193e-05, + "loss": 0.2935, + "step": 24024 + }, + { + "epoch": 0.5631838346900144, + "grad_norm": 3.7820048332214355, + "learning_rate": 8.031083687402006e-05, + "loss": 0.4192, + "step": 24025 + }, + { + "epoch": 0.5632072762648195, + "grad_norm": 2.943465232849121, + "learning_rate": 8.030361496025476e-05, + "loss": 0.2823, + "step": 24026 + }, + { + "epoch": 0.5632307178396244, + "grad_norm": 3.0832626819610596, + "learning_rate": 8.029639315336264e-05, + "loss": 1.6939, + "step": 24027 + }, + { + "epoch": 0.5632541594144295, + "grad_norm": 4.135133743286133, + "learning_rate": 8.028917145338277e-05, + "loss": 0.3391, + "step": 24028 + }, + { + "epoch": 0.5632776009892344, + "grad_norm": 1.7332617044448853, + "learning_rate": 8.028194986035446e-05, + "loss": 0.1007, + "step": 24029 + }, + { + "epoch": 0.5633010425640395, + "grad_norm": 3.342893600463867, + "learning_rate": 8.027472837431685e-05, + "loss": 0.306, + "step": 24030 + }, + { + "epoch": 0.5633244841388444, + "grad_norm": 4.1874871253967285, + "learning_rate": 8.026750699530912e-05, + "loss": 0.3771, + "step": 24031 + }, + { + "epoch": 0.5633479257136494, + "grad_norm": 1.1984285116195679, + "learning_rate": 8.026028572337043e-05, + "loss": 0.116, + "step": 24032 + }, + { + "epoch": 0.5633713672884544, + "grad_norm": 0.9663557410240173, + "learning_rate": 8.025306455854e-05, + "loss": 0.0627, + "step": 24033 + }, + { + "epoch": 0.5633948088632594, + "grad_norm": 1.2667531967163086, + "learning_rate": 8.0245843500857e-05, + "loss": 0.1135, + "step": 24034 + }, + { + "epoch": 0.5634182504380645, + "grad_norm": 0.7072350382804871, + "learning_rate": 8.023862255036059e-05, + "loss": 0.0958, + "step": 24035 + }, + { + "epoch": 0.5634416920128694, + "grad_norm": 4.6306304931640625, + "learning_rate": 8.023140170709e-05, + "loss": 0.3594, + "step": 24036 + }, + { + "epoch": 0.5634651335876745, + "grad_norm": 2.9283411502838135, + "learning_rate": 8.022418097108436e-05, + "loss": 0.2509, + "step": 24037 + }, + { + "epoch": 0.5634885751624794, + "grad_norm": 3.675532102584839, + "learning_rate": 8.021696034238287e-05, + "loss": 0.318, + "step": 24038 + }, + { + "epoch": 0.5635120167372845, + "grad_norm": 2.2845685482025146, + "learning_rate": 8.02097398210247e-05, + "loss": 0.2747, + "step": 24039 + }, + { + "epoch": 0.5635354583120894, + "grad_norm": 0.5924220085144043, + "learning_rate": 8.020251940704903e-05, + "loss": 0.0461, + "step": 24040 + }, + { + "epoch": 0.5635588998868944, + "grad_norm": 0.6074898838996887, + "learning_rate": 8.019529910049506e-05, + "loss": 0.0262, + "step": 24041 + }, + { + "epoch": 0.5635823414616994, + "grad_norm": 1.5048178434371948, + "learning_rate": 8.018807890140196e-05, + "loss": 0.1446, + "step": 24042 + }, + { + "epoch": 0.5636057830365044, + "grad_norm": 1.1710587739944458, + "learning_rate": 8.018085880980891e-05, + "loss": 0.0781, + "step": 24043 + }, + { + "epoch": 0.5636292246113094, + "grad_norm": 0.8547073602676392, + "learning_rate": 8.017363882575508e-05, + "loss": 0.0494, + "step": 24044 + }, + { + "epoch": 0.5636526661861144, + "grad_norm": 1.9236565828323364, + "learning_rate": 8.016641894927962e-05, + "loss": 0.1657, + "step": 24045 + }, + { + "epoch": 0.5636761077609194, + "grad_norm": 3.2241480350494385, + "learning_rate": 8.015919918042174e-05, + "loss": 0.3125, + "step": 24046 + }, + { + "epoch": 0.5636995493357244, + "grad_norm": 0.911304771900177, + "learning_rate": 8.015197951922062e-05, + "loss": 0.0778, + "step": 24047 + }, + { + "epoch": 0.5637229909105294, + "grad_norm": 3.3213064670562744, + "learning_rate": 8.014475996571538e-05, + "loss": 0.4111, + "step": 24048 + }, + { + "epoch": 0.5637464324853344, + "grad_norm": 2.331416606903076, + "learning_rate": 8.013754051994526e-05, + "loss": 0.1986, + "step": 24049 + }, + { + "epoch": 0.5637698740601393, + "grad_norm": 2.0886456966400146, + "learning_rate": 8.013032118194941e-05, + "loss": 0.1575, + "step": 24050 + }, + { + "epoch": 0.5637933156349444, + "grad_norm": 3.42004132270813, + "learning_rate": 8.012310195176697e-05, + "loss": 0.3535, + "step": 24051 + }, + { + "epoch": 0.5638167572097493, + "grad_norm": 2.355752468109131, + "learning_rate": 8.011588282943712e-05, + "loss": 0.2007, + "step": 24052 + }, + { + "epoch": 0.5638401987845544, + "grad_norm": 0.5753120183944702, + "learning_rate": 8.010866381499908e-05, + "loss": 0.0503, + "step": 24053 + }, + { + "epoch": 0.5638636403593593, + "grad_norm": 2.336799144744873, + "learning_rate": 8.010144490849199e-05, + "loss": 0.2406, + "step": 24054 + }, + { + "epoch": 0.5638870819341644, + "grad_norm": 3.2101197242736816, + "learning_rate": 8.0094226109955e-05, + "loss": 0.2749, + "step": 24055 + }, + { + "epoch": 0.5639105235089693, + "grad_norm": 2.6795103549957275, + "learning_rate": 8.008700741942733e-05, + "loss": 0.341, + "step": 24056 + }, + { + "epoch": 0.5639339650837744, + "grad_norm": 3.4600346088409424, + "learning_rate": 8.007978883694809e-05, + "loss": 0.2297, + "step": 24057 + }, + { + "epoch": 0.5639574066585793, + "grad_norm": 4.182016372680664, + "learning_rate": 8.00725703625565e-05, + "loss": 0.3202, + "step": 24058 + }, + { + "epoch": 0.5639808482333843, + "grad_norm": 1.7380555868148804, + "learning_rate": 8.00653519962917e-05, + "loss": 0.1233, + "step": 24059 + }, + { + "epoch": 0.5640042898081893, + "grad_norm": 2.6684229373931885, + "learning_rate": 8.005813373819284e-05, + "loss": 0.1572, + "step": 24060 + }, + { + "epoch": 0.5640277313829943, + "grad_norm": 2.3350374698638916, + "learning_rate": 8.005091558829913e-05, + "loss": 1.6079, + "step": 24061 + }, + { + "epoch": 0.5640511729577993, + "grad_norm": 0.36248064041137695, + "learning_rate": 8.00436975466497e-05, + "loss": 0.023, + "step": 24062 + }, + { + "epoch": 0.5640746145326043, + "grad_norm": 3.617968797683716, + "learning_rate": 8.003647961328373e-05, + "loss": 0.3558, + "step": 24063 + }, + { + "epoch": 0.5640980561074093, + "grad_norm": 4.215408802032471, + "learning_rate": 8.002926178824035e-05, + "loss": 0.4296, + "step": 24064 + }, + { + "epoch": 0.5641214976822143, + "grad_norm": 0.9921948909759521, + "learning_rate": 8.00220440715588e-05, + "loss": 0.1036, + "step": 24065 + }, + { + "epoch": 0.5641449392570193, + "grad_norm": 1.069344162940979, + "learning_rate": 8.001482646327821e-05, + "loss": 0.0703, + "step": 24066 + }, + { + "epoch": 0.5641683808318243, + "grad_norm": 4.701587200164795, + "learning_rate": 8.000760896343768e-05, + "loss": 0.2618, + "step": 24067 + }, + { + "epoch": 0.5641918224066292, + "grad_norm": 3.4403908252716064, + "learning_rate": 8.000039157207649e-05, + "loss": 0.4134, + "step": 24068 + }, + { + "epoch": 0.5642152639814343, + "grad_norm": 5.409668445587158, + "learning_rate": 7.999317428923371e-05, + "loss": 0.4708, + "step": 24069 + }, + { + "epoch": 0.5642387055562392, + "grad_norm": 3.1899542808532715, + "learning_rate": 7.998595711494852e-05, + "loss": 0.1649, + "step": 24070 + }, + { + "epoch": 0.5642621471310443, + "grad_norm": 1.1940109729766846, + "learning_rate": 7.99787400492601e-05, + "loss": 0.0508, + "step": 24071 + }, + { + "epoch": 0.5642855887058492, + "grad_norm": 3.424283981323242, + "learning_rate": 7.99715230922076e-05, + "loss": 0.3015, + "step": 24072 + }, + { + "epoch": 0.5643090302806543, + "grad_norm": 0.8688659071922302, + "learning_rate": 7.996430624383016e-05, + "loss": 0.0729, + "step": 24073 + }, + { + "epoch": 0.5643324718554592, + "grad_norm": 4.298670291900635, + "learning_rate": 7.995708950416698e-05, + "loss": 0.2781, + "step": 24074 + }, + { + "epoch": 0.5643559134302643, + "grad_norm": 3.653346061706543, + "learning_rate": 7.99498728732572e-05, + "loss": 0.4021, + "step": 24075 + }, + { + "epoch": 0.5643793550050692, + "grad_norm": 2.530205011367798, + "learning_rate": 7.994265635113996e-05, + "loss": 0.2441, + "step": 24076 + }, + { + "epoch": 0.5644027965798742, + "grad_norm": 0.6726983189582825, + "learning_rate": 7.993543993785439e-05, + "loss": 0.0516, + "step": 24077 + }, + { + "epoch": 0.5644262381546792, + "grad_norm": 1.7834573984146118, + "learning_rate": 7.992822363343975e-05, + "loss": 0.117, + "step": 24078 + }, + { + "epoch": 0.5644496797294842, + "grad_norm": 2.7478888034820557, + "learning_rate": 7.992100743793511e-05, + "loss": 0.2395, + "step": 24079 + }, + { + "epoch": 0.5644731213042892, + "grad_norm": 3.168142080307007, + "learning_rate": 7.991379135137964e-05, + "loss": 0.2604, + "step": 24080 + }, + { + "epoch": 0.5644965628790942, + "grad_norm": 1.357120156288147, + "learning_rate": 7.990657537381252e-05, + "loss": 1.2208, + "step": 24081 + }, + { + "epoch": 0.5645200044538992, + "grad_norm": 2.4467198848724365, + "learning_rate": 7.98993595052729e-05, + "loss": 0.1452, + "step": 24082 + }, + { + "epoch": 0.5645434460287042, + "grad_norm": 1.3266253471374512, + "learning_rate": 7.989214374579989e-05, + "loss": 0.156, + "step": 24083 + }, + { + "epoch": 0.5645668876035091, + "grad_norm": 4.285297393798828, + "learning_rate": 7.988492809543268e-05, + "loss": 0.3008, + "step": 24084 + }, + { + "epoch": 0.5645903291783142, + "grad_norm": 2.5458245277404785, + "learning_rate": 7.987771255421044e-05, + "loss": 0.1547, + "step": 24085 + }, + { + "epoch": 0.5646137707531192, + "grad_norm": 0.9103150963783264, + "learning_rate": 7.987049712217226e-05, + "loss": 0.0862, + "step": 24086 + }, + { + "epoch": 0.5646372123279242, + "grad_norm": 4.189099311828613, + "learning_rate": 7.986328179935736e-05, + "loss": 0.3773, + "step": 24087 + }, + { + "epoch": 0.5646606539027292, + "grad_norm": 1.4593397378921509, + "learning_rate": 7.985606658580485e-05, + "loss": 0.1096, + "step": 24088 + }, + { + "epoch": 0.5646840954775342, + "grad_norm": 2.2165579795837402, + "learning_rate": 7.984885148155386e-05, + "loss": 0.1783, + "step": 24089 + }, + { + "epoch": 0.5647075370523392, + "grad_norm": 1.1767598390579224, + "learning_rate": 7.984163648664358e-05, + "loss": 0.1134, + "step": 24090 + }, + { + "epoch": 0.5647309786271442, + "grad_norm": 0.9853153228759766, + "learning_rate": 7.983442160111316e-05, + "loss": 0.0713, + "step": 24091 + }, + { + "epoch": 0.5647544202019492, + "grad_norm": 1.0955218076705933, + "learning_rate": 7.982720682500175e-05, + "loss": 0.1532, + "step": 24092 + }, + { + "epoch": 0.5647778617767542, + "grad_norm": 1.1155112981796265, + "learning_rate": 7.981999215834846e-05, + "loss": 0.0863, + "step": 24093 + }, + { + "epoch": 0.5648013033515592, + "grad_norm": 3.7761306762695312, + "learning_rate": 7.981277760119248e-05, + "loss": 0.4059, + "step": 24094 + }, + { + "epoch": 0.5648247449263641, + "grad_norm": 3.7285428047180176, + "learning_rate": 7.98055631535729e-05, + "loss": 0.4077, + "step": 24095 + }, + { + "epoch": 0.5648481865011692, + "grad_norm": 3.7498788833618164, + "learning_rate": 7.979834881552895e-05, + "loss": 0.3489, + "step": 24096 + }, + { + "epoch": 0.5648716280759741, + "grad_norm": 3.7162106037139893, + "learning_rate": 7.979113458709971e-05, + "loss": 2.6207, + "step": 24097 + }, + { + "epoch": 0.5648950696507792, + "grad_norm": 2.172137498855591, + "learning_rate": 7.978392046832433e-05, + "loss": 0.1682, + "step": 24098 + }, + { + "epoch": 0.5649185112255841, + "grad_norm": 0.23935586214065552, + "learning_rate": 7.977670645924197e-05, + "loss": 0.0192, + "step": 24099 + }, + { + "epoch": 0.5649419528003892, + "grad_norm": 3.2360994815826416, + "learning_rate": 7.976949255989177e-05, + "loss": 1.6879, + "step": 24100 + }, + { + "epoch": 0.5649653943751941, + "grad_norm": 3.1843490600585938, + "learning_rate": 7.976227877031285e-05, + "loss": 1.4705, + "step": 24101 + }, + { + "epoch": 0.5649888359499992, + "grad_norm": 1.7597694396972656, + "learning_rate": 7.975506509054439e-05, + "loss": 0.0989, + "step": 24102 + }, + { + "epoch": 0.5650122775248041, + "grad_norm": 3.294940233230591, + "learning_rate": 7.974785152062554e-05, + "loss": 0.2768, + "step": 24103 + }, + { + "epoch": 0.5650357190996091, + "grad_norm": 1.1051177978515625, + "learning_rate": 7.97406380605954e-05, + "loss": 0.0605, + "step": 24104 + }, + { + "epoch": 0.5650591606744141, + "grad_norm": 3.551227569580078, + "learning_rate": 7.973342471049312e-05, + "loss": 0.1847, + "step": 24105 + }, + { + "epoch": 0.5650826022492191, + "grad_norm": 0.5692724585533142, + "learning_rate": 7.972621147035786e-05, + "loss": 0.0309, + "step": 24106 + }, + { + "epoch": 0.5651060438240241, + "grad_norm": 4.4089837074279785, + "learning_rate": 7.971899834022876e-05, + "loss": 0.4701, + "step": 24107 + }, + { + "epoch": 0.5651294853988291, + "grad_norm": 2.775630235671997, + "learning_rate": 7.971178532014492e-05, + "loss": 0.1614, + "step": 24108 + }, + { + "epoch": 0.5651529269736341, + "grad_norm": 2.266164779663086, + "learning_rate": 7.970457241014552e-05, + "loss": 0.1873, + "step": 24109 + }, + { + "epoch": 0.5651763685484391, + "grad_norm": 1.0087069272994995, + "learning_rate": 7.969735961026967e-05, + "loss": 0.1254, + "step": 24110 + }, + { + "epoch": 0.565199810123244, + "grad_norm": 1.7338534593582153, + "learning_rate": 7.969014692055652e-05, + "loss": 0.1364, + "step": 24111 + }, + { + "epoch": 0.5652232516980491, + "grad_norm": 4.412939071655273, + "learning_rate": 7.968293434104521e-05, + "loss": 0.4296, + "step": 24112 + }, + { + "epoch": 0.565246693272854, + "grad_norm": 3.307241678237915, + "learning_rate": 7.967572187177484e-05, + "loss": 0.2004, + "step": 24113 + }, + { + "epoch": 0.5652701348476591, + "grad_norm": 2.467113971710205, + "learning_rate": 7.96685095127846e-05, + "loss": 0.1187, + "step": 24114 + }, + { + "epoch": 0.565293576422464, + "grad_norm": 3.8215808868408203, + "learning_rate": 7.966129726411357e-05, + "loss": 0.3949, + "step": 24115 + }, + { + "epoch": 0.5653170179972691, + "grad_norm": 1.9881418943405151, + "learning_rate": 7.965408512580096e-05, + "loss": 0.2151, + "step": 24116 + }, + { + "epoch": 0.565340459572074, + "grad_norm": 2.6664106845855713, + "learning_rate": 7.964687309788585e-05, + "loss": 0.1995, + "step": 24117 + }, + { + "epoch": 0.5653639011468791, + "grad_norm": 2.5954785346984863, + "learning_rate": 7.963966118040734e-05, + "loss": 0.1609, + "step": 24118 + }, + { + "epoch": 0.565387342721684, + "grad_norm": 1.6249299049377441, + "learning_rate": 7.963244937340464e-05, + "loss": 0.105, + "step": 24119 + }, + { + "epoch": 0.565410784296489, + "grad_norm": 2.491532325744629, + "learning_rate": 7.962523767691684e-05, + "loss": 0.2647, + "step": 24120 + }, + { + "epoch": 0.565434225871294, + "grad_norm": 2.668678045272827, + "learning_rate": 7.961802609098305e-05, + "loss": 0.2338, + "step": 24121 + }, + { + "epoch": 0.565457667446099, + "grad_norm": 4.180294513702393, + "learning_rate": 7.961081461564244e-05, + "loss": 0.301, + "step": 24122 + }, + { + "epoch": 0.565481109020904, + "grad_norm": 1.930647850036621, + "learning_rate": 7.960360325093413e-05, + "loss": 0.1655, + "step": 24123 + }, + { + "epoch": 0.565504550595709, + "grad_norm": 2.399915933609009, + "learning_rate": 7.959639199689721e-05, + "loss": 0.2664, + "step": 24124 + }, + { + "epoch": 0.565527992170514, + "grad_norm": 0.7821868062019348, + "learning_rate": 7.958918085357083e-05, + "loss": 0.0632, + "step": 24125 + }, + { + "epoch": 0.565551433745319, + "grad_norm": 2.933441638946533, + "learning_rate": 7.958196982099418e-05, + "loss": 0.2867, + "step": 24126 + }, + { + "epoch": 0.565574875320124, + "grad_norm": 2.5439510345458984, + "learning_rate": 7.957475889920632e-05, + "loss": 0.1445, + "step": 24127 + }, + { + "epoch": 0.565598316894929, + "grad_norm": 5.593344688415527, + "learning_rate": 7.956754808824638e-05, + "loss": 0.4482, + "step": 24128 + }, + { + "epoch": 0.565621758469734, + "grad_norm": 2.531876564025879, + "learning_rate": 7.956033738815351e-05, + "loss": 0.2339, + "step": 24129 + }, + { + "epoch": 0.565645200044539, + "grad_norm": 3.436777353286743, + "learning_rate": 7.955312679896682e-05, + "loss": 0.3178, + "step": 24130 + }, + { + "epoch": 0.5656686416193439, + "grad_norm": 3.09795880317688, + "learning_rate": 7.954591632072543e-05, + "loss": 0.2918, + "step": 24131 + }, + { + "epoch": 0.565692083194149, + "grad_norm": 1.8869125843048096, + "learning_rate": 7.95387059534685e-05, + "loss": 0.2214, + "step": 24132 + }, + { + "epoch": 0.5657155247689539, + "grad_norm": 3.563411235809326, + "learning_rate": 7.95314956972351e-05, + "loss": 0.3204, + "step": 24133 + }, + { + "epoch": 0.565738966343759, + "grad_norm": 1.3227300643920898, + "learning_rate": 7.95242855520644e-05, + "loss": 0.6199, + "step": 24134 + }, + { + "epoch": 0.5657624079185639, + "grad_norm": 3.4046013355255127, + "learning_rate": 7.95170755179955e-05, + "loss": 0.1639, + "step": 24135 + }, + { + "epoch": 0.565785849493369, + "grad_norm": 3.3571419715881348, + "learning_rate": 7.95098655950675e-05, + "loss": 2.0231, + "step": 24136 + }, + { + "epoch": 0.565809291068174, + "grad_norm": 3.6048831939697266, + "learning_rate": 7.950265578331955e-05, + "loss": 0.2981, + "step": 24137 + }, + { + "epoch": 0.565832732642979, + "grad_norm": 0.8624767065048218, + "learning_rate": 7.949544608279077e-05, + "loss": 0.087, + "step": 24138 + }, + { + "epoch": 0.565856174217784, + "grad_norm": 2.926278829574585, + "learning_rate": 7.948823649352029e-05, + "loss": 0.2204, + "step": 24139 + }, + { + "epoch": 0.5658796157925889, + "grad_norm": 2.7335379123687744, + "learning_rate": 7.94810270155472e-05, + "loss": 0.2456, + "step": 24140 + }, + { + "epoch": 0.565903057367394, + "grad_norm": 0.7292208671569824, + "learning_rate": 7.947381764891066e-05, + "loss": 0.0775, + "step": 24141 + }, + { + "epoch": 0.5659264989421989, + "grad_norm": 2.332545042037964, + "learning_rate": 7.946660839364975e-05, + "loss": 0.1512, + "step": 24142 + }, + { + "epoch": 0.565949940517004, + "grad_norm": 4.6651763916015625, + "learning_rate": 7.945939924980357e-05, + "loss": 0.4326, + "step": 24143 + }, + { + "epoch": 0.5659733820918089, + "grad_norm": 0.2677665650844574, + "learning_rate": 7.945219021741132e-05, + "loss": 0.0178, + "step": 24144 + }, + { + "epoch": 0.565996823666614, + "grad_norm": 4.8660430908203125, + "learning_rate": 7.944498129651203e-05, + "loss": 0.3255, + "step": 24145 + }, + { + "epoch": 0.5660202652414189, + "grad_norm": 3.351534843444824, + "learning_rate": 7.943777248714485e-05, + "loss": 0.2997, + "step": 24146 + }, + { + "epoch": 0.566043706816224, + "grad_norm": 0.8056332468986511, + "learning_rate": 7.943056378934891e-05, + "loss": 0.0745, + "step": 24147 + }, + { + "epoch": 0.5660671483910289, + "grad_norm": 2.9725193977355957, + "learning_rate": 7.942335520316331e-05, + "loss": 0.1764, + "step": 24148 + }, + { + "epoch": 0.566090589965834, + "grad_norm": 4.148820877075195, + "learning_rate": 7.941614672862712e-05, + "loss": 0.3354, + "step": 24149 + }, + { + "epoch": 0.5661140315406389, + "grad_norm": 6.42699670791626, + "learning_rate": 7.940893836577951e-05, + "loss": 0.6285, + "step": 24150 + }, + { + "epoch": 0.5661374731154439, + "grad_norm": 3.416562080383301, + "learning_rate": 7.940173011465963e-05, + "loss": 1.8775, + "step": 24151 + }, + { + "epoch": 0.5661609146902489, + "grad_norm": 4.070358753204346, + "learning_rate": 7.939452197530652e-05, + "loss": 0.2265, + "step": 24152 + }, + { + "epoch": 0.5661843562650539, + "grad_norm": 4.033573150634766, + "learning_rate": 7.938731394775927e-05, + "loss": 0.3534, + "step": 24153 + }, + { + "epoch": 0.5662077978398589, + "grad_norm": 6.14363956451416, + "learning_rate": 7.938010603205709e-05, + "loss": 0.9535, + "step": 24154 + }, + { + "epoch": 0.5662312394146639, + "grad_norm": 4.0831170082092285, + "learning_rate": 7.937289822823902e-05, + "loss": 0.1764, + "step": 24155 + }, + { + "epoch": 0.5662546809894689, + "grad_norm": 3.3158369064331055, + "learning_rate": 7.936569053634417e-05, + "loss": 0.3347, + "step": 24156 + }, + { + "epoch": 0.5662781225642739, + "grad_norm": 2.611081123352051, + "learning_rate": 7.935848295641169e-05, + "loss": 1.9437, + "step": 24157 + }, + { + "epoch": 0.5663015641390788, + "grad_norm": 1.2946540117263794, + "learning_rate": 7.935127548848064e-05, + "loss": 0.0635, + "step": 24158 + }, + { + "epoch": 0.5663250057138839, + "grad_norm": 4.098653793334961, + "learning_rate": 7.934406813259014e-05, + "loss": 0.3291, + "step": 24159 + }, + { + "epoch": 0.5663484472886888, + "grad_norm": 4.596480369567871, + "learning_rate": 7.933686088877933e-05, + "loss": 0.5004, + "step": 24160 + }, + { + "epoch": 0.5663718888634939, + "grad_norm": 4.960780620574951, + "learning_rate": 7.93296537570873e-05, + "loss": 0.5042, + "step": 24161 + }, + { + "epoch": 0.5663953304382988, + "grad_norm": 2.821129083633423, + "learning_rate": 7.932244673755309e-05, + "loss": 0.245, + "step": 24162 + }, + { + "epoch": 0.5664187720131039, + "grad_norm": 6.097233772277832, + "learning_rate": 7.931523983021589e-05, + "loss": 0.4676, + "step": 24163 + }, + { + "epoch": 0.5664422135879088, + "grad_norm": 4.559014797210693, + "learning_rate": 7.930803303511481e-05, + "loss": 0.3257, + "step": 24164 + }, + { + "epoch": 0.5664656551627139, + "grad_norm": 1.3051592111587524, + "learning_rate": 7.930082635228893e-05, + "loss": 0.0979, + "step": 24165 + }, + { + "epoch": 0.5664890967375188, + "grad_norm": 2.3106842041015625, + "learning_rate": 7.929361978177732e-05, + "loss": 0.153, + "step": 24166 + }, + { + "epoch": 0.5665125383123238, + "grad_norm": 3.8375635147094727, + "learning_rate": 7.928641332361912e-05, + "loss": 0.2262, + "step": 24167 + }, + { + "epoch": 0.5665359798871288, + "grad_norm": 0.7957830429077148, + "learning_rate": 7.927920697785345e-05, + "loss": 0.0913, + "step": 24168 + }, + { + "epoch": 0.5665594214619338, + "grad_norm": 2.743079662322998, + "learning_rate": 7.927200074451935e-05, + "loss": 0.255, + "step": 24169 + }, + { + "epoch": 0.5665828630367388, + "grad_norm": 0.9869435429573059, + "learning_rate": 7.926479462365597e-05, + "loss": 0.103, + "step": 24170 + }, + { + "epoch": 0.5666063046115438, + "grad_norm": 0.9256467819213867, + "learning_rate": 7.925758861530238e-05, + "loss": 0.0934, + "step": 24171 + }, + { + "epoch": 0.5666297461863488, + "grad_norm": 2.647873640060425, + "learning_rate": 7.925038271949772e-05, + "loss": 0.1795, + "step": 24172 + }, + { + "epoch": 0.5666531877611538, + "grad_norm": 5.041677951812744, + "learning_rate": 7.924317693628107e-05, + "loss": 0.4178, + "step": 24173 + }, + { + "epoch": 0.5666766293359587, + "grad_norm": 4.783102035522461, + "learning_rate": 7.923597126569149e-05, + "loss": 0.3324, + "step": 24174 + }, + { + "epoch": 0.5667000709107638, + "grad_norm": 4.842617988586426, + "learning_rate": 7.922876570776813e-05, + "loss": 0.3691, + "step": 24175 + }, + { + "epoch": 0.5667235124855687, + "grad_norm": 3.1798765659332275, + "learning_rate": 7.922156026255009e-05, + "loss": 0.2644, + "step": 24176 + }, + { + "epoch": 0.5667469540603738, + "grad_norm": 1.7721452713012695, + "learning_rate": 7.921435493007645e-05, + "loss": 0.1801, + "step": 24177 + }, + { + "epoch": 0.5667703956351787, + "grad_norm": 0.7236594557762146, + "learning_rate": 7.920714971038628e-05, + "loss": 0.0655, + "step": 24178 + }, + { + "epoch": 0.5667938372099838, + "grad_norm": 5.2442193031311035, + "learning_rate": 7.919994460351872e-05, + "loss": 0.5473, + "step": 24179 + }, + { + "epoch": 0.5668172787847887, + "grad_norm": 0.5112236142158508, + "learning_rate": 7.919273960951285e-05, + "loss": 0.0271, + "step": 24180 + }, + { + "epoch": 0.5668407203595938, + "grad_norm": 3.0870273113250732, + "learning_rate": 7.918553472840773e-05, + "loss": 0.2338, + "step": 24181 + }, + { + "epoch": 0.5668641619343987, + "grad_norm": 1.120262861251831, + "learning_rate": 7.917832996024253e-05, + "loss": 0.0977, + "step": 24182 + }, + { + "epoch": 0.5668876035092038, + "grad_norm": 1.2020916938781738, + "learning_rate": 7.917112530505627e-05, + "loss": 0.1213, + "step": 24183 + }, + { + "epoch": 0.5669110450840087, + "grad_norm": 1.2548236846923828, + "learning_rate": 7.916392076288805e-05, + "loss": 0.1235, + "step": 24184 + }, + { + "epoch": 0.5669344866588137, + "grad_norm": 12.484148979187012, + "learning_rate": 7.915671633377702e-05, + "loss": 0.5414, + "step": 24185 + }, + { + "epoch": 0.5669579282336187, + "grad_norm": 2.7539379596710205, + "learning_rate": 7.914951201776217e-05, + "loss": 0.285, + "step": 24186 + }, + { + "epoch": 0.5669813698084237, + "grad_norm": 3.427359104156494, + "learning_rate": 7.914230781488272e-05, + "loss": 0.2657, + "step": 24187 + }, + { + "epoch": 0.5670048113832288, + "grad_norm": 4.409172058105469, + "learning_rate": 7.913510372517766e-05, + "loss": 0.4164, + "step": 24188 + }, + { + "epoch": 0.5670282529580337, + "grad_norm": 2.3561742305755615, + "learning_rate": 7.912789974868614e-05, + "loss": 1.8439, + "step": 24189 + }, + { + "epoch": 0.5670516945328388, + "grad_norm": 0.5413212180137634, + "learning_rate": 7.912069588544722e-05, + "loss": 0.0278, + "step": 24190 + }, + { + "epoch": 0.5670751361076437, + "grad_norm": 3.3783628940582275, + "learning_rate": 7.911349213549997e-05, + "loss": 0.2956, + "step": 24191 + }, + { + "epoch": 0.5670985776824488, + "grad_norm": 4.711256504058838, + "learning_rate": 7.910628849888351e-05, + "loss": 0.5012, + "step": 24192 + }, + { + "epoch": 0.5671220192572537, + "grad_norm": 2.469068765640259, + "learning_rate": 7.909908497563693e-05, + "loss": 1.1317, + "step": 24193 + }, + { + "epoch": 0.5671454608320587, + "grad_norm": 2.4999351501464844, + "learning_rate": 7.909188156579927e-05, + "loss": 0.2603, + "step": 24194 + }, + { + "epoch": 0.5671689024068637, + "grad_norm": 2.7076985836029053, + "learning_rate": 7.908467826940969e-05, + "loss": 0.219, + "step": 24195 + }, + { + "epoch": 0.5671923439816687, + "grad_norm": 2.40741229057312, + "learning_rate": 7.907747508650722e-05, + "loss": 0.1351, + "step": 24196 + }, + { + "epoch": 0.5672157855564737, + "grad_norm": 3.5614936351776123, + "learning_rate": 7.907027201713093e-05, + "loss": 0.3278, + "step": 24197 + }, + { + "epoch": 0.5672392271312787, + "grad_norm": 1.9027220010757446, + "learning_rate": 7.906306906131992e-05, + "loss": 0.2457, + "step": 24198 + }, + { + "epoch": 0.5672626687060837, + "grad_norm": 4.77105188369751, + "learning_rate": 7.905586621911333e-05, + "loss": 0.4684, + "step": 24199 + }, + { + "epoch": 0.5672861102808887, + "grad_norm": 3.885314702987671, + "learning_rate": 7.904866349055019e-05, + "loss": 0.1533, + "step": 24200 + }, + { + "epoch": 0.5673095518556937, + "grad_norm": 4.1777143478393555, + "learning_rate": 7.904146087566958e-05, + "loss": 0.4283, + "step": 24201 + }, + { + "epoch": 0.5673329934304987, + "grad_norm": 5.074306488037109, + "learning_rate": 7.90342583745106e-05, + "loss": 0.5883, + "step": 24202 + }, + { + "epoch": 0.5673564350053036, + "grad_norm": 7.756374835968018, + "learning_rate": 7.902705598711234e-05, + "loss": 0.2417, + "step": 24203 + }, + { + "epoch": 0.5673798765801087, + "grad_norm": 2.794743776321411, + "learning_rate": 7.901985371351383e-05, + "loss": 0.128, + "step": 24204 + }, + { + "epoch": 0.5674033181549136, + "grad_norm": 3.2268779277801514, + "learning_rate": 7.901265155375421e-05, + "loss": 0.2736, + "step": 24205 + }, + { + "epoch": 0.5674267597297187, + "grad_norm": 4.520478248596191, + "learning_rate": 7.900544950787254e-05, + "loss": 0.2476, + "step": 24206 + }, + { + "epoch": 0.5674502013045236, + "grad_norm": 3.098858118057251, + "learning_rate": 7.899824757590787e-05, + "loss": 0.3598, + "step": 24207 + }, + { + "epoch": 0.5674736428793287, + "grad_norm": 3.7934069633483887, + "learning_rate": 7.89910457578993e-05, + "loss": 1.9685, + "step": 24208 + }, + { + "epoch": 0.5674970844541336, + "grad_norm": 0.6913261413574219, + "learning_rate": 7.898384405388592e-05, + "loss": 0.0436, + "step": 24209 + }, + { + "epoch": 0.5675205260289387, + "grad_norm": 3.577279567718506, + "learning_rate": 7.897664246390677e-05, + "loss": 0.2971, + "step": 24210 + }, + { + "epoch": 0.5675439676037436, + "grad_norm": 4.5775628089904785, + "learning_rate": 7.896944098800094e-05, + "loss": 0.437, + "step": 24211 + }, + { + "epoch": 0.5675674091785486, + "grad_norm": 3.7348620891571045, + "learning_rate": 7.896223962620754e-05, + "loss": 1.705, + "step": 24212 + }, + { + "epoch": 0.5675908507533536, + "grad_norm": 6.428890705108643, + "learning_rate": 7.895503837856562e-05, + "loss": 0.2323, + "step": 24213 + }, + { + "epoch": 0.5676142923281586, + "grad_norm": 0.6232393980026245, + "learning_rate": 7.894783724511425e-05, + "loss": 0.0342, + "step": 24214 + }, + { + "epoch": 0.5676377339029636, + "grad_norm": 4.142514228820801, + "learning_rate": 7.894063622589253e-05, + "loss": 0.1655, + "step": 24215 + }, + { + "epoch": 0.5676611754777686, + "grad_norm": 0.699935257434845, + "learning_rate": 7.893343532093948e-05, + "loss": 0.0164, + "step": 24216 + }, + { + "epoch": 0.5676846170525736, + "grad_norm": 2.690929889678955, + "learning_rate": 7.89262345302942e-05, + "loss": 0.1398, + "step": 24217 + }, + { + "epoch": 0.5677080586273786, + "grad_norm": 3.1274187564849854, + "learning_rate": 7.89190338539958e-05, + "loss": 0.3588, + "step": 24218 + }, + { + "epoch": 0.5677315002021835, + "grad_norm": 3.1032915115356445, + "learning_rate": 7.891183329208328e-05, + "loss": 0.2036, + "step": 24219 + }, + { + "epoch": 0.5677549417769886, + "grad_norm": 1.1708831787109375, + "learning_rate": 7.890463284459575e-05, + "loss": 0.1376, + "step": 24220 + }, + { + "epoch": 0.5677783833517935, + "grad_norm": 4.853874206542969, + "learning_rate": 7.889743251157229e-05, + "loss": 0.4089, + "step": 24221 + }, + { + "epoch": 0.5678018249265986, + "grad_norm": 6.0704851150512695, + "learning_rate": 7.889023229305191e-05, + "loss": 0.5543, + "step": 24222 + }, + { + "epoch": 0.5678252665014035, + "grad_norm": 3.469996690750122, + "learning_rate": 7.888303218907376e-05, + "loss": 0.5155, + "step": 24223 + }, + { + "epoch": 0.5678487080762086, + "grad_norm": 3.123933792114258, + "learning_rate": 7.887583219967687e-05, + "loss": 1.6141, + "step": 24224 + }, + { + "epoch": 0.5678721496510135, + "grad_norm": 3.260111093521118, + "learning_rate": 7.886863232490032e-05, + "loss": 0.4933, + "step": 24225 + }, + { + "epoch": 0.5678955912258186, + "grad_norm": 3.3101143836975098, + "learning_rate": 7.886143256478314e-05, + "loss": 0.1534, + "step": 24226 + }, + { + "epoch": 0.5679190328006235, + "grad_norm": 2.8630919456481934, + "learning_rate": 7.885423291936445e-05, + "loss": 0.1888, + "step": 24227 + }, + { + "epoch": 0.5679424743754286, + "grad_norm": 0.7863369584083557, + "learning_rate": 7.884703338868327e-05, + "loss": 0.0574, + "step": 24228 + }, + { + "epoch": 0.5679659159502335, + "grad_norm": 3.0640106201171875, + "learning_rate": 7.883983397277868e-05, + "loss": 0.3464, + "step": 24229 + }, + { + "epoch": 0.5679893575250385, + "grad_norm": 2.760476589202881, + "learning_rate": 7.883263467168975e-05, + "loss": 0.2961, + "step": 24230 + }, + { + "epoch": 0.5680127990998435, + "grad_norm": 2.7451529502868652, + "learning_rate": 7.882543548545555e-05, + "loss": 0.2449, + "step": 24231 + }, + { + "epoch": 0.5680362406746485, + "grad_norm": 1.5145213603973389, + "learning_rate": 7.88182364141151e-05, + "loss": 0.0746, + "step": 24232 + }, + { + "epoch": 0.5680596822494535, + "grad_norm": 4.904965400695801, + "learning_rate": 7.881103745770752e-05, + "loss": 0.3981, + "step": 24233 + }, + { + "epoch": 0.5680831238242585, + "grad_norm": 1.4432638883590698, + "learning_rate": 7.880383861627185e-05, + "loss": 0.0602, + "step": 24234 + }, + { + "epoch": 0.5681065653990635, + "grad_norm": 2.674725294113159, + "learning_rate": 7.879663988984712e-05, + "loss": 0.2516, + "step": 24235 + }, + { + "epoch": 0.5681300069738685, + "grad_norm": 2.91888689994812, + "learning_rate": 7.878944127847242e-05, + "loss": 1.7857, + "step": 24236 + }, + { + "epoch": 0.5681534485486734, + "grad_norm": 1.2278975248336792, + "learning_rate": 7.878224278218683e-05, + "loss": 0.1645, + "step": 24237 + }, + { + "epoch": 0.5681768901234785, + "grad_norm": 2.506303310394287, + "learning_rate": 7.877504440102939e-05, + "loss": 0.2485, + "step": 24238 + }, + { + "epoch": 0.5682003316982835, + "grad_norm": 2.4646339416503906, + "learning_rate": 7.876784613503914e-05, + "loss": 0.1434, + "step": 24239 + }, + { + "epoch": 0.5682237732730885, + "grad_norm": 3.295001983642578, + "learning_rate": 7.876064798425517e-05, + "loss": 0.3659, + "step": 24240 + }, + { + "epoch": 0.5682472148478935, + "grad_norm": 7.729194641113281, + "learning_rate": 7.875344994871652e-05, + "loss": 0.262, + "step": 24241 + }, + { + "epoch": 0.5682706564226985, + "grad_norm": 0.7340376973152161, + "learning_rate": 7.874625202846224e-05, + "loss": 0.0576, + "step": 24242 + }, + { + "epoch": 0.5682940979975035, + "grad_norm": 3.6008126735687256, + "learning_rate": 7.873905422353139e-05, + "loss": 0.2583, + "step": 24243 + }, + { + "epoch": 0.5683175395723085, + "grad_norm": 2.729996919631958, + "learning_rate": 7.873185653396303e-05, + "loss": 0.3597, + "step": 24244 + }, + { + "epoch": 0.5683409811471135, + "grad_norm": 4.649001598358154, + "learning_rate": 7.872465895979622e-05, + "loss": 0.3252, + "step": 24245 + }, + { + "epoch": 0.5683644227219184, + "grad_norm": 3.3644649982452393, + "learning_rate": 7.871746150107002e-05, + "loss": 0.2618, + "step": 24246 + }, + { + "epoch": 0.5683878642967235, + "grad_norm": 3.0846219062805176, + "learning_rate": 7.871026415782343e-05, + "loss": 0.2421, + "step": 24247 + }, + { + "epoch": 0.5684113058715284, + "grad_norm": 0.7462123036384583, + "learning_rate": 7.870306693009559e-05, + "loss": 0.0343, + "step": 24248 + }, + { + "epoch": 0.5684347474463335, + "grad_norm": 3.1886255741119385, + "learning_rate": 7.869586981792547e-05, + "loss": 0.2931, + "step": 24249 + }, + { + "epoch": 0.5684581890211384, + "grad_norm": 3.275485038757324, + "learning_rate": 7.86886728213522e-05, + "loss": 0.2593, + "step": 24250 + }, + { + "epoch": 0.5684816305959435, + "grad_norm": 2.9550933837890625, + "learning_rate": 7.868147594041477e-05, + "loss": 0.2116, + "step": 24251 + }, + { + "epoch": 0.5685050721707484, + "grad_norm": 0.8952300548553467, + "learning_rate": 7.867427917515226e-05, + "loss": 0.0408, + "step": 24252 + }, + { + "epoch": 0.5685285137455535, + "grad_norm": 3.2836344242095947, + "learning_rate": 7.866708252560371e-05, + "loss": 0.244, + "step": 24253 + }, + { + "epoch": 0.5685519553203584, + "grad_norm": 2.4731764793395996, + "learning_rate": 7.865988599180815e-05, + "loss": 0.1796, + "step": 24254 + }, + { + "epoch": 0.5685753968951635, + "grad_norm": 2.2249321937561035, + "learning_rate": 7.865268957380467e-05, + "loss": 0.2326, + "step": 24255 + }, + { + "epoch": 0.5685988384699684, + "grad_norm": 4.2363481521606445, + "learning_rate": 7.864549327163231e-05, + "loss": 0.1312, + "step": 24256 + }, + { + "epoch": 0.5686222800447734, + "grad_norm": 3.2820701599121094, + "learning_rate": 7.863829708533007e-05, + "loss": 0.3203, + "step": 24257 + }, + { + "epoch": 0.5686457216195784, + "grad_norm": 1.8401904106140137, + "learning_rate": 7.863110101493704e-05, + "loss": 0.6491, + "step": 24258 + }, + { + "epoch": 0.5686691631943834, + "grad_norm": 2.6744749546051025, + "learning_rate": 7.862390506049225e-05, + "loss": 0.1189, + "step": 24259 + }, + { + "epoch": 0.5686926047691884, + "grad_norm": 0.8040401339530945, + "learning_rate": 7.861670922203477e-05, + "loss": 0.0468, + "step": 24260 + }, + { + "epoch": 0.5687160463439934, + "grad_norm": 3.828507900238037, + "learning_rate": 7.86095134996036e-05, + "loss": 0.2796, + "step": 24261 + }, + { + "epoch": 0.5687394879187984, + "grad_norm": 1.8497257232666016, + "learning_rate": 7.860231789323785e-05, + "loss": 0.1301, + "step": 24262 + }, + { + "epoch": 0.5687629294936034, + "grad_norm": 0.6503018140792847, + "learning_rate": 7.859512240297651e-05, + "loss": 0.0304, + "step": 24263 + }, + { + "epoch": 0.5687863710684083, + "grad_norm": 3.5364365577697754, + "learning_rate": 7.858792702885862e-05, + "loss": 0.2221, + "step": 24264 + }, + { + "epoch": 0.5688098126432134, + "grad_norm": 1.3303302526474, + "learning_rate": 7.858073177092327e-05, + "loss": 0.1433, + "step": 24265 + }, + { + "epoch": 0.5688332542180183, + "grad_norm": 2.2109405994415283, + "learning_rate": 7.857353662920946e-05, + "loss": 0.2341, + "step": 24266 + }, + { + "epoch": 0.5688566957928234, + "grad_norm": 2.2375550270080566, + "learning_rate": 7.856634160375622e-05, + "loss": 0.2115, + "step": 24267 + }, + { + "epoch": 0.5688801373676283, + "grad_norm": 3.512030839920044, + "learning_rate": 7.855914669460265e-05, + "loss": 0.34, + "step": 24268 + }, + { + "epoch": 0.5689035789424334, + "grad_norm": 3.002178430557251, + "learning_rate": 7.855195190178774e-05, + "loss": 1.5979, + "step": 24269 + }, + { + "epoch": 0.5689270205172383, + "grad_norm": 2.168154239654541, + "learning_rate": 7.854475722535052e-05, + "loss": 0.143, + "step": 24270 + }, + { + "epoch": 0.5689504620920434, + "grad_norm": 2.2251288890838623, + "learning_rate": 7.853756266533003e-05, + "loss": 1.1025, + "step": 24271 + }, + { + "epoch": 0.5689739036668483, + "grad_norm": 4.650207042694092, + "learning_rate": 7.853036822176538e-05, + "loss": 0.4978, + "step": 24272 + }, + { + "epoch": 0.5689973452416534, + "grad_norm": 0.9807513356208801, + "learning_rate": 7.852317389469555e-05, + "loss": 0.1047, + "step": 24273 + }, + { + "epoch": 0.5690207868164583, + "grad_norm": 1.564731240272522, + "learning_rate": 7.851597968415956e-05, + "loss": 1.3161, + "step": 24274 + }, + { + "epoch": 0.5690442283912633, + "grad_norm": 1.1670794486999512, + "learning_rate": 7.850878559019649e-05, + "loss": 0.1469, + "step": 24275 + }, + { + "epoch": 0.5690676699660683, + "grad_norm": 0.4058593511581421, + "learning_rate": 7.850159161284536e-05, + "loss": 0.0222, + "step": 24276 + }, + { + "epoch": 0.5690911115408733, + "grad_norm": 2.8643431663513184, + "learning_rate": 7.849439775214518e-05, + "loss": 0.2047, + "step": 24277 + }, + { + "epoch": 0.5691145531156783, + "grad_norm": 2.912593126296997, + "learning_rate": 7.8487204008135e-05, + "loss": 0.2286, + "step": 24278 + }, + { + "epoch": 0.5691379946904833, + "grad_norm": 3.5080628395080566, + "learning_rate": 7.848001038085389e-05, + "loss": 0.3106, + "step": 24279 + }, + { + "epoch": 0.5691614362652883, + "grad_norm": 5.306736469268799, + "learning_rate": 7.847281687034081e-05, + "loss": 0.2974, + "step": 24280 + }, + { + "epoch": 0.5691848778400933, + "grad_norm": 3.442816972732544, + "learning_rate": 7.846562347663484e-05, + "loss": 0.2762, + "step": 24281 + }, + { + "epoch": 0.5692083194148982, + "grad_norm": 0.86316978931427, + "learning_rate": 7.8458430199775e-05, + "loss": 0.0869, + "step": 24282 + }, + { + "epoch": 0.5692317609897033, + "grad_norm": 1.9003486633300781, + "learning_rate": 7.845123703980032e-05, + "loss": 0.2073, + "step": 24283 + }, + { + "epoch": 0.5692552025645082, + "grad_norm": 0.4090195596218109, + "learning_rate": 7.844404399674983e-05, + "loss": 0.0391, + "step": 24284 + }, + { + "epoch": 0.5692786441393133, + "grad_norm": 0.47185853123664856, + "learning_rate": 7.843685107066257e-05, + "loss": 0.0489, + "step": 24285 + }, + { + "epoch": 0.5693020857141182, + "grad_norm": 1.0626249313354492, + "learning_rate": 7.842965826157759e-05, + "loss": 0.1139, + "step": 24286 + }, + { + "epoch": 0.5693255272889233, + "grad_norm": 1.2510182857513428, + "learning_rate": 7.842246556953386e-05, + "loss": 0.0382, + "step": 24287 + }, + { + "epoch": 0.5693489688637282, + "grad_norm": 2.311342239379883, + "learning_rate": 7.841527299457046e-05, + "loss": 0.2219, + "step": 24288 + }, + { + "epoch": 0.5693724104385333, + "grad_norm": 2.053966522216797, + "learning_rate": 7.840808053672636e-05, + "loss": 0.1417, + "step": 24289 + }, + { + "epoch": 0.5693958520133383, + "grad_norm": 0.39547380805015564, + "learning_rate": 7.840088819604065e-05, + "loss": 0.0449, + "step": 24290 + }, + { + "epoch": 0.5694192935881432, + "grad_norm": 3.4402430057525635, + "learning_rate": 7.839369597255234e-05, + "loss": 0.2493, + "step": 24291 + }, + { + "epoch": 0.5694427351629483, + "grad_norm": 0.6481066942214966, + "learning_rate": 7.838650386630041e-05, + "loss": 0.0641, + "step": 24292 + }, + { + "epoch": 0.5694661767377532, + "grad_norm": 2.2941155433654785, + "learning_rate": 7.837931187732395e-05, + "loss": 0.2657, + "step": 24293 + }, + { + "epoch": 0.5694896183125583, + "grad_norm": 4.402869701385498, + "learning_rate": 7.837212000566194e-05, + "loss": 0.5189, + "step": 24294 + }, + { + "epoch": 0.5695130598873632, + "grad_norm": 2.7525172233581543, + "learning_rate": 7.836492825135337e-05, + "loss": 0.2862, + "step": 24295 + }, + { + "epoch": 0.5695365014621683, + "grad_norm": 2.9568400382995605, + "learning_rate": 7.835773661443734e-05, + "loss": 0.2851, + "step": 24296 + }, + { + "epoch": 0.5695599430369732, + "grad_norm": 0.5139586925506592, + "learning_rate": 7.835054509495285e-05, + "loss": 0.0333, + "step": 24297 + }, + { + "epoch": 0.5695833846117783, + "grad_norm": 4.308618545532227, + "learning_rate": 7.834335369293893e-05, + "loss": 0.3968, + "step": 24298 + }, + { + "epoch": 0.5696068261865832, + "grad_norm": 2.709955930709839, + "learning_rate": 7.833616240843454e-05, + "loss": 0.1656, + "step": 24299 + }, + { + "epoch": 0.5696302677613883, + "grad_norm": 4.004452228546143, + "learning_rate": 7.832897124147878e-05, + "loss": 0.4621, + "step": 24300 + }, + { + "epoch": 0.5696537093361932, + "grad_norm": 2.5456061363220215, + "learning_rate": 7.832178019211062e-05, + "loss": 0.1234, + "step": 24301 + }, + { + "epoch": 0.5696771509109982, + "grad_norm": 3.19464111328125, + "learning_rate": 7.831458926036908e-05, + "loss": 0.592, + "step": 24302 + }, + { + "epoch": 0.5697005924858032, + "grad_norm": 0.4983896017074585, + "learning_rate": 7.830739844629319e-05, + "loss": 0.0536, + "step": 24303 + }, + { + "epoch": 0.5697240340606082, + "grad_norm": 2.8134310245513916, + "learning_rate": 7.830020774992198e-05, + "loss": 1.7587, + "step": 24304 + }, + { + "epoch": 0.5697474756354132, + "grad_norm": 2.5256385803222656, + "learning_rate": 7.829301717129442e-05, + "loss": 0.213, + "step": 24305 + }, + { + "epoch": 0.5697709172102182, + "grad_norm": 3.7859244346618652, + "learning_rate": 7.82858267104496e-05, + "loss": 0.3149, + "step": 24306 + }, + { + "epoch": 0.5697943587850232, + "grad_norm": 3.9204418659210205, + "learning_rate": 7.827863636742649e-05, + "loss": 0.3206, + "step": 24307 + }, + { + "epoch": 0.5698178003598282, + "grad_norm": 3.1331911087036133, + "learning_rate": 7.827144614226407e-05, + "loss": 0.3487, + "step": 24308 + }, + { + "epoch": 0.5698412419346331, + "grad_norm": 2.440131187438965, + "learning_rate": 7.82642560350014e-05, + "loss": 0.226, + "step": 24309 + }, + { + "epoch": 0.5698646835094382, + "grad_norm": 6.750133037567139, + "learning_rate": 7.82570660456775e-05, + "loss": 0.5017, + "step": 24310 + }, + { + "epoch": 0.5698881250842431, + "grad_norm": 3.780580997467041, + "learning_rate": 7.824987617433139e-05, + "loss": 0.2661, + "step": 24311 + }, + { + "epoch": 0.5699115666590482, + "grad_norm": 2.7966461181640625, + "learning_rate": 7.824268642100206e-05, + "loss": 0.2739, + "step": 24312 + }, + { + "epoch": 0.5699350082338531, + "grad_norm": 4.422506809234619, + "learning_rate": 7.82354967857285e-05, + "loss": 0.3535, + "step": 24313 + }, + { + "epoch": 0.5699584498086582, + "grad_norm": 3.900966167449951, + "learning_rate": 7.822830726854978e-05, + "loss": 0.4949, + "step": 24314 + }, + { + "epoch": 0.5699818913834631, + "grad_norm": 1.7645232677459717, + "learning_rate": 7.822111786950486e-05, + "loss": 0.163, + "step": 24315 + }, + { + "epoch": 0.5700053329582682, + "grad_norm": 2.5836634635925293, + "learning_rate": 7.821392858863276e-05, + "loss": 0.2341, + "step": 24316 + }, + { + "epoch": 0.5700287745330731, + "grad_norm": 2.2426419258117676, + "learning_rate": 7.820673942597252e-05, + "loss": 0.0951, + "step": 24317 + }, + { + "epoch": 0.5700522161078782, + "grad_norm": 0.9087803959846497, + "learning_rate": 7.819955038156311e-05, + "loss": 0.1498, + "step": 24318 + }, + { + "epoch": 0.5700756576826831, + "grad_norm": 0.8037276864051819, + "learning_rate": 7.819236145544355e-05, + "loss": 0.0732, + "step": 24319 + }, + { + "epoch": 0.5700990992574881, + "grad_norm": 3.1982977390289307, + "learning_rate": 7.818517264765282e-05, + "loss": 0.1315, + "step": 24320 + }, + { + "epoch": 0.5701225408322931, + "grad_norm": 7.518835544586182, + "learning_rate": 7.817798395822999e-05, + "loss": 0.4755, + "step": 24321 + }, + { + "epoch": 0.5701459824070981, + "grad_norm": 3.4893598556518555, + "learning_rate": 7.817079538721401e-05, + "loss": 1.1219, + "step": 24322 + }, + { + "epoch": 0.5701694239819031, + "grad_norm": 2.2236907482147217, + "learning_rate": 7.816360693464394e-05, + "loss": 0.1621, + "step": 24323 + }, + { + "epoch": 0.5701928655567081, + "grad_norm": 2.432497262954712, + "learning_rate": 7.815641860055876e-05, + "loss": 0.2299, + "step": 24324 + }, + { + "epoch": 0.5702163071315131, + "grad_norm": 2.0965728759765625, + "learning_rate": 7.814923038499743e-05, + "loss": 1.4472, + "step": 24325 + }, + { + "epoch": 0.5702397487063181, + "grad_norm": 1.1399412155151367, + "learning_rate": 7.814204228799901e-05, + "loss": 0.1405, + "step": 24326 + }, + { + "epoch": 0.570263190281123, + "grad_norm": 10.602252006530762, + "learning_rate": 7.813485430960248e-05, + "loss": 0.8386, + "step": 24327 + }, + { + "epoch": 0.5702866318559281, + "grad_norm": 0.8985846638679504, + "learning_rate": 7.812766644984685e-05, + "loss": 0.0988, + "step": 24328 + }, + { + "epoch": 0.570310073430733, + "grad_norm": 2.1105198860168457, + "learning_rate": 7.812047870877113e-05, + "loss": 0.1213, + "step": 24329 + }, + { + "epoch": 0.5703335150055381, + "grad_norm": 1.9649251699447632, + "learning_rate": 7.811329108641428e-05, + "loss": 1.3066, + "step": 24330 + }, + { + "epoch": 0.570356956580343, + "grad_norm": 4.973324775695801, + "learning_rate": 7.810610358281534e-05, + "loss": 0.1581, + "step": 24331 + }, + { + "epoch": 0.5703803981551481, + "grad_norm": 3.256870746612549, + "learning_rate": 7.809891619801326e-05, + "loss": 0.2435, + "step": 24332 + }, + { + "epoch": 0.570403839729953, + "grad_norm": 2.1677498817443848, + "learning_rate": 7.809172893204712e-05, + "loss": 0.1009, + "step": 24333 + }, + { + "epoch": 0.5704272813047581, + "grad_norm": 3.0664989948272705, + "learning_rate": 7.808454178495585e-05, + "loss": 0.2608, + "step": 24334 + }, + { + "epoch": 0.570450722879563, + "grad_norm": 0.9359325170516968, + "learning_rate": 7.807735475677851e-05, + "loss": 0.0799, + "step": 24335 + }, + { + "epoch": 0.570474164454368, + "grad_norm": 3.82681941986084, + "learning_rate": 7.807016784755404e-05, + "loss": 0.2544, + "step": 24336 + }, + { + "epoch": 0.570497606029173, + "grad_norm": 5.360743045806885, + "learning_rate": 7.806298105732145e-05, + "loss": 0.6308, + "step": 24337 + }, + { + "epoch": 0.570521047603978, + "grad_norm": 3.952307939529419, + "learning_rate": 7.805579438611975e-05, + "loss": 0.2701, + "step": 24338 + }, + { + "epoch": 0.570544489178783, + "grad_norm": 1.0659079551696777, + "learning_rate": 7.804860783398794e-05, + "loss": 0.0739, + "step": 24339 + }, + { + "epoch": 0.570567930753588, + "grad_norm": 3.4050891399383545, + "learning_rate": 7.804142140096498e-05, + "loss": 0.282, + "step": 24340 + }, + { + "epoch": 0.5705913723283931, + "grad_norm": 3.6949520111083984, + "learning_rate": 7.80342350870899e-05, + "loss": 0.2353, + "step": 24341 + }, + { + "epoch": 0.570614813903198, + "grad_norm": 2.9756999015808105, + "learning_rate": 7.802704889240166e-05, + "loss": 2.1043, + "step": 24342 + }, + { + "epoch": 0.5706382554780031, + "grad_norm": 3.732895612716675, + "learning_rate": 7.801986281693927e-05, + "loss": 0.3611, + "step": 24343 + }, + { + "epoch": 0.570661697052808, + "grad_norm": 3.0845720767974854, + "learning_rate": 7.80126768607417e-05, + "loss": 0.4769, + "step": 24344 + }, + { + "epoch": 0.570685138627613, + "grad_norm": 0.6253615021705627, + "learning_rate": 7.8005491023848e-05, + "loss": 0.0468, + "step": 24345 + }, + { + "epoch": 0.570708580202418, + "grad_norm": 2.723438024520874, + "learning_rate": 7.799830530629712e-05, + "loss": 0.3678, + "step": 24346 + }, + { + "epoch": 0.570732021777223, + "grad_norm": 3.0282607078552246, + "learning_rate": 7.799111970812805e-05, + "loss": 0.1756, + "step": 24347 + }, + { + "epoch": 0.570755463352028, + "grad_norm": 2.9093210697174072, + "learning_rate": 7.79839342293798e-05, + "loss": 0.1941, + "step": 24348 + }, + { + "epoch": 0.570778904926833, + "grad_norm": 3.9071786403656006, + "learning_rate": 7.797674887009133e-05, + "loss": 0.3538, + "step": 24349 + }, + { + "epoch": 0.570802346501638, + "grad_norm": 3.7584941387176514, + "learning_rate": 7.796956363030162e-05, + "loss": 0.2452, + "step": 24350 + }, + { + "epoch": 0.570825788076443, + "grad_norm": 2.12724232673645, + "learning_rate": 7.79623785100497e-05, + "loss": 0.2311, + "step": 24351 + }, + { + "epoch": 0.570849229651248, + "grad_norm": 2.5565013885498047, + "learning_rate": 7.795519350937455e-05, + "loss": 0.2352, + "step": 24352 + }, + { + "epoch": 0.570872671226053, + "grad_norm": 1.6252366304397583, + "learning_rate": 7.794800862831511e-05, + "loss": 1.0542, + "step": 24353 + }, + { + "epoch": 0.570896112800858, + "grad_norm": 1.2226200103759766, + "learning_rate": 7.79408238669104e-05, + "loss": 0.0746, + "step": 24354 + }, + { + "epoch": 0.570919554375663, + "grad_norm": 2.4910998344421387, + "learning_rate": 7.793363922519941e-05, + "loss": 0.2678, + "step": 24355 + }, + { + "epoch": 0.5709429959504679, + "grad_norm": 4.237127780914307, + "learning_rate": 7.792645470322109e-05, + "loss": 0.2053, + "step": 24356 + }, + { + "epoch": 0.570966437525273, + "grad_norm": 4.020205497741699, + "learning_rate": 7.791927030101445e-05, + "loss": 0.4153, + "step": 24357 + }, + { + "epoch": 0.5709898791000779, + "grad_norm": 3.5663838386535645, + "learning_rate": 7.791208601861849e-05, + "loss": 0.2038, + "step": 24358 + }, + { + "epoch": 0.571013320674883, + "grad_norm": 4.3592915534973145, + "learning_rate": 7.790490185607218e-05, + "loss": 0.5509, + "step": 24359 + }, + { + "epoch": 0.5710367622496879, + "grad_norm": 1.8417173624038696, + "learning_rate": 7.789771781341446e-05, + "loss": 0.2608, + "step": 24360 + }, + { + "epoch": 0.571060203824493, + "grad_norm": 0.5647303462028503, + "learning_rate": 7.789053389068438e-05, + "loss": 0.0479, + "step": 24361 + }, + { + "epoch": 0.5710836453992979, + "grad_norm": 2.7474024295806885, + "learning_rate": 7.788335008792086e-05, + "loss": 0.2295, + "step": 24362 + }, + { + "epoch": 0.571107086974103, + "grad_norm": 2.856072187423706, + "learning_rate": 7.787616640516292e-05, + "loss": 0.3115, + "step": 24363 + }, + { + "epoch": 0.5711305285489079, + "grad_norm": 3.3238019943237305, + "learning_rate": 7.786898284244952e-05, + "loss": 0.2642, + "step": 24364 + }, + { + "epoch": 0.5711539701237129, + "grad_norm": 3.314974069595337, + "learning_rate": 7.786179939981963e-05, + "loss": 0.4567, + "step": 24365 + }, + { + "epoch": 0.5711774116985179, + "grad_norm": 0.7593262195587158, + "learning_rate": 7.785461607731226e-05, + "loss": 0.036, + "step": 24366 + }, + { + "epoch": 0.5712008532733229, + "grad_norm": 2.5079872608184814, + "learning_rate": 7.784743287496637e-05, + "loss": 1.7101, + "step": 24367 + }, + { + "epoch": 0.5712242948481279, + "grad_norm": 1.2223361730575562, + "learning_rate": 7.784024979282087e-05, + "loss": 1.4896, + "step": 24368 + }, + { + "epoch": 0.5712477364229329, + "grad_norm": 2.150928258895874, + "learning_rate": 7.783306683091485e-05, + "loss": 0.1935, + "step": 24369 + }, + { + "epoch": 0.5712711779977379, + "grad_norm": 2.3015780448913574, + "learning_rate": 7.782588398928723e-05, + "loss": 0.1395, + "step": 24370 + }, + { + "epoch": 0.5712946195725429, + "grad_norm": 2.2759623527526855, + "learning_rate": 7.781870126797699e-05, + "loss": 1.0131, + "step": 24371 + }, + { + "epoch": 0.5713180611473478, + "grad_norm": 1.5587689876556396, + "learning_rate": 7.781151866702307e-05, + "loss": 0.1528, + "step": 24372 + }, + { + "epoch": 0.5713415027221529, + "grad_norm": 3.0628364086151123, + "learning_rate": 7.780433618646452e-05, + "loss": 0.1713, + "step": 24373 + }, + { + "epoch": 0.5713649442969578, + "grad_norm": 2.969278335571289, + "learning_rate": 7.779715382634026e-05, + "loss": 0.3052, + "step": 24374 + }, + { + "epoch": 0.5713883858717629, + "grad_norm": 3.231806755065918, + "learning_rate": 7.778997158668923e-05, + "loss": 0.288, + "step": 24375 + }, + { + "epoch": 0.5714118274465678, + "grad_norm": 3.551248550415039, + "learning_rate": 7.778278946755046e-05, + "loss": 0.2909, + "step": 24376 + }, + { + "epoch": 0.5714352690213729, + "grad_norm": 1.609153389930725, + "learning_rate": 7.777560746896292e-05, + "loss": 1.0105, + "step": 24377 + }, + { + "epoch": 0.5714587105961778, + "grad_norm": 8.334385871887207, + "learning_rate": 7.776842559096552e-05, + "loss": 0.2678, + "step": 24378 + }, + { + "epoch": 0.5714821521709829, + "grad_norm": 5.017575263977051, + "learning_rate": 7.776124383359729e-05, + "loss": 0.3358, + "step": 24379 + }, + { + "epoch": 0.5715055937457878, + "grad_norm": 0.625739574432373, + "learning_rate": 7.775406219689714e-05, + "loss": 0.0611, + "step": 24380 + }, + { + "epoch": 0.5715290353205928, + "grad_norm": 2.252807378768921, + "learning_rate": 7.774688068090412e-05, + "loss": 0.3023, + "step": 24381 + }, + { + "epoch": 0.5715524768953978, + "grad_norm": 2.9333248138427734, + "learning_rate": 7.773969928565713e-05, + "loss": 0.258, + "step": 24382 + }, + { + "epoch": 0.5715759184702028, + "grad_norm": 3.495435953140259, + "learning_rate": 7.773251801119517e-05, + "loss": 0.2971, + "step": 24383 + }, + { + "epoch": 0.5715993600450078, + "grad_norm": 1.6081827878952026, + "learning_rate": 7.77253368575572e-05, + "loss": 0.1877, + "step": 24384 + }, + { + "epoch": 0.5716228016198128, + "grad_norm": 2.5845537185668945, + "learning_rate": 7.771815582478214e-05, + "loss": 1.2769, + "step": 24385 + }, + { + "epoch": 0.5716462431946178, + "grad_norm": 3.076549530029297, + "learning_rate": 7.771097491290904e-05, + "loss": 0.242, + "step": 24386 + }, + { + "epoch": 0.5716696847694228, + "grad_norm": 3.1320760250091553, + "learning_rate": 7.77037941219768e-05, + "loss": 0.2753, + "step": 24387 + }, + { + "epoch": 0.5716931263442278, + "grad_norm": 3.4116482734680176, + "learning_rate": 7.769661345202439e-05, + "loss": 0.3588, + "step": 24388 + }, + { + "epoch": 0.5717165679190328, + "grad_norm": 1.8689348697662354, + "learning_rate": 7.76894329030908e-05, + "loss": 0.1703, + "step": 24389 + }, + { + "epoch": 0.5717400094938377, + "grad_norm": 1.858870029449463, + "learning_rate": 7.768225247521496e-05, + "loss": 0.1237, + "step": 24390 + }, + { + "epoch": 0.5717634510686428, + "grad_norm": 3.3539674282073975, + "learning_rate": 7.767507216843585e-05, + "loss": 0.2248, + "step": 24391 + }, + { + "epoch": 0.5717868926434478, + "grad_norm": 2.8775382041931152, + "learning_rate": 7.766789198279242e-05, + "loss": 0.1998, + "step": 24392 + }, + { + "epoch": 0.5718103342182528, + "grad_norm": 3.4088025093078613, + "learning_rate": 7.766071191832362e-05, + "loss": 0.2596, + "step": 24393 + }, + { + "epoch": 0.5718337757930578, + "grad_norm": 1.2556933164596558, + "learning_rate": 7.765353197506844e-05, + "loss": 0.0778, + "step": 24394 + }, + { + "epoch": 0.5718572173678628, + "grad_norm": 2.935274600982666, + "learning_rate": 7.764635215306584e-05, + "loss": 0.1687, + "step": 24395 + }, + { + "epoch": 0.5718806589426678, + "grad_norm": 0.8167514801025391, + "learning_rate": 7.763917245235474e-05, + "loss": 0.0726, + "step": 24396 + }, + { + "epoch": 0.5719041005174728, + "grad_norm": 3.4206812381744385, + "learning_rate": 7.763199287297416e-05, + "loss": 0.4074, + "step": 24397 + }, + { + "epoch": 0.5719275420922778, + "grad_norm": 1.9717445373535156, + "learning_rate": 7.762481341496298e-05, + "loss": 0.1502, + "step": 24398 + }, + { + "epoch": 0.5719509836670827, + "grad_norm": 0.5163123607635498, + "learning_rate": 7.76176340783602e-05, + "loss": 0.0519, + "step": 24399 + }, + { + "epoch": 0.5719744252418878, + "grad_norm": 2.2783193588256836, + "learning_rate": 7.761045486320476e-05, + "loss": 0.1217, + "step": 24400 + }, + { + "epoch": 0.5719978668166927, + "grad_norm": 2.347266435623169, + "learning_rate": 7.760327576953562e-05, + "loss": 0.1912, + "step": 24401 + }, + { + "epoch": 0.5720213083914978, + "grad_norm": 3.6886179447174072, + "learning_rate": 7.759609679739175e-05, + "loss": 0.3092, + "step": 24402 + }, + { + "epoch": 0.5720447499663027, + "grad_norm": 2.85109281539917, + "learning_rate": 7.758891794681207e-05, + "loss": 0.2252, + "step": 24403 + }, + { + "epoch": 0.5720681915411078, + "grad_norm": 1.4065556526184082, + "learning_rate": 7.758173921783557e-05, + "loss": 0.1389, + "step": 24404 + }, + { + "epoch": 0.5720916331159127, + "grad_norm": 3.4194469451904297, + "learning_rate": 7.757456061050115e-05, + "loss": 0.6914, + "step": 24405 + }, + { + "epoch": 0.5721150746907178, + "grad_norm": 0.6477677822113037, + "learning_rate": 7.756738212484782e-05, + "loss": 0.0301, + "step": 24406 + }, + { + "epoch": 0.5721385162655227, + "grad_norm": 2.6898722648620605, + "learning_rate": 7.75602037609145e-05, + "loss": 0.3003, + "step": 24407 + }, + { + "epoch": 0.5721619578403278, + "grad_norm": 2.5122616291046143, + "learning_rate": 7.755302551874015e-05, + "loss": 1.7508, + "step": 24408 + }, + { + "epoch": 0.5721853994151327, + "grad_norm": 4.166747093200684, + "learning_rate": 7.754584739836372e-05, + "loss": 0.3903, + "step": 24409 + }, + { + "epoch": 0.5722088409899377, + "grad_norm": 0.469619482755661, + "learning_rate": 7.753866939982415e-05, + "loss": 0.0573, + "step": 24410 + }, + { + "epoch": 0.5722322825647427, + "grad_norm": 1.6314157247543335, + "learning_rate": 7.753149152316039e-05, + "loss": 0.1463, + "step": 24411 + }, + { + "epoch": 0.5722557241395477, + "grad_norm": 1.5976678133010864, + "learning_rate": 7.752431376841141e-05, + "loss": 0.0977, + "step": 24412 + }, + { + "epoch": 0.5722791657143527, + "grad_norm": 5.051811695098877, + "learning_rate": 7.75171361356161e-05, + "loss": 0.5321, + "step": 24413 + }, + { + "epoch": 0.5723026072891577, + "grad_norm": 6.461174488067627, + "learning_rate": 7.750995862481346e-05, + "loss": 0.5935, + "step": 24414 + }, + { + "epoch": 0.5723260488639627, + "grad_norm": 2.7308292388916016, + "learning_rate": 7.750278123604241e-05, + "loss": 0.3153, + "step": 24415 + }, + { + "epoch": 0.5723494904387677, + "grad_norm": 2.3826301097869873, + "learning_rate": 7.74956039693419e-05, + "loss": 0.1233, + "step": 24416 + }, + { + "epoch": 0.5723729320135726, + "grad_norm": 1.6496915817260742, + "learning_rate": 7.748842682475085e-05, + "loss": 0.083, + "step": 24417 + }, + { + "epoch": 0.5723963735883777, + "grad_norm": 0.5840827822685242, + "learning_rate": 7.748124980230826e-05, + "loss": 0.0477, + "step": 24418 + }, + { + "epoch": 0.5724198151631826, + "grad_norm": 4.471336364746094, + "learning_rate": 7.747407290205306e-05, + "loss": 0.3872, + "step": 24419 + }, + { + "epoch": 0.5724432567379877, + "grad_norm": 2.178246021270752, + "learning_rate": 7.746689612402414e-05, + "loss": 0.1045, + "step": 24420 + }, + { + "epoch": 0.5724666983127926, + "grad_norm": 1.658408522605896, + "learning_rate": 7.74597194682605e-05, + "loss": 0.076, + "step": 24421 + }, + { + "epoch": 0.5724901398875977, + "grad_norm": 3.302032947540283, + "learning_rate": 7.745254293480106e-05, + "loss": 1.4695, + "step": 24422 + }, + { + "epoch": 0.5725135814624026, + "grad_norm": 3.9877233505249023, + "learning_rate": 7.744536652368472e-05, + "loss": 0.41, + "step": 24423 + }, + { + "epoch": 0.5725370230372077, + "grad_norm": 1.3796660900115967, + "learning_rate": 7.743819023495049e-05, + "loss": 0.1671, + "step": 24424 + }, + { + "epoch": 0.5725604646120126, + "grad_norm": 3.094310998916626, + "learning_rate": 7.743101406863726e-05, + "loss": 0.432, + "step": 24425 + }, + { + "epoch": 0.5725839061868176, + "grad_norm": 3.09379506111145, + "learning_rate": 7.742383802478397e-05, + "loss": 0.4062, + "step": 24426 + }, + { + "epoch": 0.5726073477616226, + "grad_norm": 2.582798480987549, + "learning_rate": 7.741666210342961e-05, + "loss": 0.2477, + "step": 24427 + }, + { + "epoch": 0.5726307893364276, + "grad_norm": 2.8867990970611572, + "learning_rate": 7.740948630461304e-05, + "loss": 0.1095, + "step": 24428 + }, + { + "epoch": 0.5726542309112326, + "grad_norm": 2.611494302749634, + "learning_rate": 7.740231062837322e-05, + "loss": 1.39, + "step": 24429 + }, + { + "epoch": 0.5726776724860376, + "grad_norm": 4.70471715927124, + "learning_rate": 7.73951350747491e-05, + "loss": 0.4125, + "step": 24430 + }, + { + "epoch": 0.5727011140608426, + "grad_norm": 5.618725776672363, + "learning_rate": 7.738795964377965e-05, + "loss": 0.3789, + "step": 24431 + }, + { + "epoch": 0.5727245556356476, + "grad_norm": 2.4434776306152344, + "learning_rate": 7.738078433550375e-05, + "loss": 0.1934, + "step": 24432 + }, + { + "epoch": 0.5727479972104526, + "grad_norm": 2.9242913722991943, + "learning_rate": 7.737360914996035e-05, + "loss": 0.1708, + "step": 24433 + }, + { + "epoch": 0.5727714387852576, + "grad_norm": 5.575986862182617, + "learning_rate": 7.736643408718837e-05, + "loss": 0.3892, + "step": 24434 + }, + { + "epoch": 0.5727948803600625, + "grad_norm": 3.9632668495178223, + "learning_rate": 7.735925914722679e-05, + "loss": 0.3176, + "step": 24435 + }, + { + "epoch": 0.5728183219348676, + "grad_norm": 3.595055341720581, + "learning_rate": 7.735208433011447e-05, + "loss": 0.4106, + "step": 24436 + }, + { + "epoch": 0.5728417635096725, + "grad_norm": 1.7268441915512085, + "learning_rate": 7.734490963589041e-05, + "loss": 0.245, + "step": 24437 + }, + { + "epoch": 0.5728652050844776, + "grad_norm": 1.646533727645874, + "learning_rate": 7.733773506459347e-05, + "loss": 0.1144, + "step": 24438 + }, + { + "epoch": 0.5728886466592825, + "grad_norm": 7.983530044555664, + "learning_rate": 7.733056061626265e-05, + "loss": 0.5944, + "step": 24439 + }, + { + "epoch": 0.5729120882340876, + "grad_norm": 2.975630283355713, + "learning_rate": 7.732338629093682e-05, + "loss": 0.1978, + "step": 24440 + }, + { + "epoch": 0.5729355298088925, + "grad_norm": 3.2887966632843018, + "learning_rate": 7.731621208865492e-05, + "loss": 0.3972, + "step": 24441 + }, + { + "epoch": 0.5729589713836976, + "grad_norm": 4.653741836547852, + "learning_rate": 7.730903800945592e-05, + "loss": 0.5345, + "step": 24442 + }, + { + "epoch": 0.5729824129585026, + "grad_norm": 1.7946497201919556, + "learning_rate": 7.730186405337872e-05, + "loss": 0.1079, + "step": 24443 + }, + { + "epoch": 0.5730058545333075, + "grad_norm": 4.432745456695557, + "learning_rate": 7.729469022046223e-05, + "loss": 0.4787, + "step": 24444 + }, + { + "epoch": 0.5730292961081126, + "grad_norm": 3.071150302886963, + "learning_rate": 7.72875165107454e-05, + "loss": 0.148, + "step": 24445 + }, + { + "epoch": 0.5730527376829175, + "grad_norm": 0.33995380997657776, + "learning_rate": 7.728034292426714e-05, + "loss": 0.02, + "step": 24446 + }, + { + "epoch": 0.5730761792577226, + "grad_norm": 4.770660400390625, + "learning_rate": 7.727316946106639e-05, + "loss": 0.3891, + "step": 24447 + }, + { + "epoch": 0.5730996208325275, + "grad_norm": 3.0022430419921875, + "learning_rate": 7.726599612118204e-05, + "loss": 0.3793, + "step": 24448 + }, + { + "epoch": 0.5731230624073326, + "grad_norm": 3.254753351211548, + "learning_rate": 7.725882290465305e-05, + "loss": 0.3126, + "step": 24449 + }, + { + "epoch": 0.5731465039821375, + "grad_norm": 0.9149224162101746, + "learning_rate": 7.725164981151832e-05, + "loss": 0.0818, + "step": 24450 + }, + { + "epoch": 0.5731699455569426, + "grad_norm": 3.1174569129943848, + "learning_rate": 7.724447684181677e-05, + "loss": 0.3289, + "step": 24451 + }, + { + "epoch": 0.5731933871317475, + "grad_norm": 1.027891755104065, + "learning_rate": 7.723730399558735e-05, + "loss": 0.0983, + "step": 24452 + }, + { + "epoch": 0.5732168287065526, + "grad_norm": 0.45254406332969666, + "learning_rate": 7.723013127286891e-05, + "loss": 0.0257, + "step": 24453 + }, + { + "epoch": 0.5732402702813575, + "grad_norm": 6.127050399780273, + "learning_rate": 7.722295867370047e-05, + "loss": 0.3155, + "step": 24454 + }, + { + "epoch": 0.5732637118561625, + "grad_norm": 0.4185376763343811, + "learning_rate": 7.721578619812086e-05, + "loss": 0.0362, + "step": 24455 + }, + { + "epoch": 0.5732871534309675, + "grad_norm": 2.8929028511047363, + "learning_rate": 7.720861384616907e-05, + "loss": 0.3177, + "step": 24456 + }, + { + "epoch": 0.5733105950057725, + "grad_norm": 2.693419933319092, + "learning_rate": 7.720144161788397e-05, + "loss": 0.3128, + "step": 24457 + }, + { + "epoch": 0.5733340365805775, + "grad_norm": 2.775500774383545, + "learning_rate": 7.719426951330448e-05, + "loss": 0.2714, + "step": 24458 + }, + { + "epoch": 0.5733574781553825, + "grad_norm": 2.4750070571899414, + "learning_rate": 7.718709753246953e-05, + "loss": 0.281, + "step": 24459 + }, + { + "epoch": 0.5733809197301875, + "grad_norm": 0.92817223072052, + "learning_rate": 7.717992567541805e-05, + "loss": 0.1051, + "step": 24460 + }, + { + "epoch": 0.5734043613049925, + "grad_norm": 3.5381669998168945, + "learning_rate": 7.71727539421889e-05, + "loss": 0.321, + "step": 24461 + }, + { + "epoch": 0.5734278028797974, + "grad_norm": 3.522020101547241, + "learning_rate": 7.716558233282105e-05, + "loss": 0.2922, + "step": 24462 + }, + { + "epoch": 0.5734512444546025, + "grad_norm": 0.6290668845176697, + "learning_rate": 7.715841084735341e-05, + "loss": 0.0426, + "step": 24463 + }, + { + "epoch": 0.5734746860294074, + "grad_norm": 4.623945713043213, + "learning_rate": 7.715123948582483e-05, + "loss": 0.1764, + "step": 24464 + }, + { + "epoch": 0.5734981276042125, + "grad_norm": 2.5263636112213135, + "learning_rate": 7.714406824827429e-05, + "loss": 0.2318, + "step": 24465 + }, + { + "epoch": 0.5735215691790174, + "grad_norm": 2.3568637371063232, + "learning_rate": 7.713689713474064e-05, + "loss": 0.1122, + "step": 24466 + }, + { + "epoch": 0.5735450107538225, + "grad_norm": 4.737544536590576, + "learning_rate": 7.712972614526288e-05, + "loss": 0.5393, + "step": 24467 + }, + { + "epoch": 0.5735684523286274, + "grad_norm": 2.5667691230773926, + "learning_rate": 7.712255527987984e-05, + "loss": 0.2977, + "step": 24468 + }, + { + "epoch": 0.5735918939034325, + "grad_norm": 2.8629190921783447, + "learning_rate": 7.711538453863048e-05, + "loss": 0.3053, + "step": 24469 + }, + { + "epoch": 0.5736153354782374, + "grad_norm": 3.958794593811035, + "learning_rate": 7.71082139215537e-05, + "loss": 0.3744, + "step": 24470 + }, + { + "epoch": 0.5736387770530424, + "grad_norm": 0.6137873530387878, + "learning_rate": 7.710104342868836e-05, + "loss": 0.1119, + "step": 24471 + }, + { + "epoch": 0.5736622186278474, + "grad_norm": 2.42350435256958, + "learning_rate": 7.709387306007343e-05, + "loss": 0.2976, + "step": 24472 + }, + { + "epoch": 0.5736856602026524, + "grad_norm": 1.6053810119628906, + "learning_rate": 7.708670281574778e-05, + "loss": 0.1588, + "step": 24473 + }, + { + "epoch": 0.5737091017774574, + "grad_norm": 3.4956114292144775, + "learning_rate": 7.707953269575032e-05, + "loss": 0.3519, + "step": 24474 + }, + { + "epoch": 0.5737325433522624, + "grad_norm": 2.6938648223876953, + "learning_rate": 7.707236270011996e-05, + "loss": 0.2354, + "step": 24475 + }, + { + "epoch": 0.5737559849270674, + "grad_norm": 3.6232030391693115, + "learning_rate": 7.70651928288956e-05, + "loss": 0.4158, + "step": 24476 + }, + { + "epoch": 0.5737794265018724, + "grad_norm": 1.3857580423355103, + "learning_rate": 7.705802308211615e-05, + "loss": 0.0699, + "step": 24477 + }, + { + "epoch": 0.5738028680766774, + "grad_norm": 3.7747466564178467, + "learning_rate": 7.705085345982049e-05, + "loss": 0.2328, + "step": 24478 + }, + { + "epoch": 0.5738263096514824, + "grad_norm": 4.885790824890137, + "learning_rate": 7.704368396204758e-05, + "loss": 0.3589, + "step": 24479 + }, + { + "epoch": 0.5738497512262873, + "grad_norm": 3.550034999847412, + "learning_rate": 7.703651458883628e-05, + "loss": 0.1578, + "step": 24480 + }, + { + "epoch": 0.5738731928010924, + "grad_norm": 3.0814390182495117, + "learning_rate": 7.702934534022549e-05, + "loss": 0.1936, + "step": 24481 + }, + { + "epoch": 0.5738966343758973, + "grad_norm": 3.089153528213501, + "learning_rate": 7.702217621625414e-05, + "loss": 0.2912, + "step": 24482 + }, + { + "epoch": 0.5739200759507024, + "grad_norm": 2.7769389152526855, + "learning_rate": 7.701500721696107e-05, + "loss": 1.0037, + "step": 24483 + }, + { + "epoch": 0.5739435175255073, + "grad_norm": 4.906437397003174, + "learning_rate": 7.700783834238525e-05, + "loss": 0.4648, + "step": 24484 + }, + { + "epoch": 0.5739669591003124, + "grad_norm": 2.0890727043151855, + "learning_rate": 7.700066959256553e-05, + "loss": 0.2588, + "step": 24485 + }, + { + "epoch": 0.5739904006751173, + "grad_norm": 2.6697728633880615, + "learning_rate": 7.699350096754082e-05, + "loss": 0.2571, + "step": 24486 + }, + { + "epoch": 0.5740138422499224, + "grad_norm": 2.029358148574829, + "learning_rate": 7.698633246735005e-05, + "loss": 1.2532, + "step": 24487 + }, + { + "epoch": 0.5740372838247273, + "grad_norm": 2.004798412322998, + "learning_rate": 7.697916409203206e-05, + "loss": 0.1561, + "step": 24488 + }, + { + "epoch": 0.5740607253995323, + "grad_norm": 2.2838382720947266, + "learning_rate": 7.697199584162576e-05, + "loss": 1.5722, + "step": 24489 + }, + { + "epoch": 0.5740841669743373, + "grad_norm": 3.4917073249816895, + "learning_rate": 7.696482771617005e-05, + "loss": 0.3711, + "step": 24490 + }, + { + "epoch": 0.5741076085491423, + "grad_norm": 4.190942764282227, + "learning_rate": 7.695765971570386e-05, + "loss": 0.4758, + "step": 24491 + }, + { + "epoch": 0.5741310501239473, + "grad_norm": 5.655314922332764, + "learning_rate": 7.695049184026607e-05, + "loss": 0.395, + "step": 24492 + }, + { + "epoch": 0.5741544916987523, + "grad_norm": 2.974900960922241, + "learning_rate": 7.694332408989553e-05, + "loss": 0.2788, + "step": 24493 + }, + { + "epoch": 0.5741779332735574, + "grad_norm": 1.3849984407424927, + "learning_rate": 7.693615646463119e-05, + "loss": 0.1238, + "step": 24494 + }, + { + "epoch": 0.5742013748483623, + "grad_norm": 4.8130950927734375, + "learning_rate": 7.69289889645119e-05, + "loss": 0.5944, + "step": 24495 + }, + { + "epoch": 0.5742248164231674, + "grad_norm": 2.1981141567230225, + "learning_rate": 7.692182158957655e-05, + "loss": 0.1274, + "step": 24496 + }, + { + "epoch": 0.5742482579979723, + "grad_norm": 1.31827974319458, + "learning_rate": 7.691465433986407e-05, + "loss": 0.167, + "step": 24497 + }, + { + "epoch": 0.5742716995727773, + "grad_norm": 3.548126459121704, + "learning_rate": 7.690748721541332e-05, + "loss": 0.3503, + "step": 24498 + }, + { + "epoch": 0.5742951411475823, + "grad_norm": 0.4575003683567047, + "learning_rate": 7.690032021626319e-05, + "loss": 0.0314, + "step": 24499 + }, + { + "epoch": 0.5743185827223873, + "grad_norm": 2.3003156185150146, + "learning_rate": 7.689315334245257e-05, + "loss": 0.2124, + "step": 24500 + }, + { + "epoch": 0.5743420242971923, + "grad_norm": 1.109509825706482, + "learning_rate": 7.688598659402035e-05, + "loss": 0.0761, + "step": 24501 + }, + { + "epoch": 0.5743654658719973, + "grad_norm": 2.423238515853882, + "learning_rate": 7.687881997100539e-05, + "loss": 0.3743, + "step": 24502 + }, + { + "epoch": 0.5743889074468023, + "grad_norm": 1.3726329803466797, + "learning_rate": 7.687165347344661e-05, + "loss": 0.2351, + "step": 24503 + }, + { + "epoch": 0.5744123490216073, + "grad_norm": 2.704094648361206, + "learning_rate": 7.686448710138293e-05, + "loss": 0.1902, + "step": 24504 + }, + { + "epoch": 0.5744357905964123, + "grad_norm": 0.8621909618377686, + "learning_rate": 7.685732085485317e-05, + "loss": 0.0667, + "step": 24505 + }, + { + "epoch": 0.5744592321712173, + "grad_norm": 3.243441104888916, + "learning_rate": 7.685015473389623e-05, + "loss": 0.3769, + "step": 24506 + }, + { + "epoch": 0.5744826737460222, + "grad_norm": 2.712890625, + "learning_rate": 7.684298873855102e-05, + "loss": 0.1729, + "step": 24507 + }, + { + "epoch": 0.5745061153208273, + "grad_norm": 2.41078782081604, + "learning_rate": 7.68358228688564e-05, + "loss": 0.2613, + "step": 24508 + }, + { + "epoch": 0.5745295568956322, + "grad_norm": 4.084625720977783, + "learning_rate": 7.682865712485123e-05, + "loss": 0.3197, + "step": 24509 + }, + { + "epoch": 0.5745529984704373, + "grad_norm": 1.114065170288086, + "learning_rate": 7.682149150657446e-05, + "loss": 0.1264, + "step": 24510 + }, + { + "epoch": 0.5745764400452422, + "grad_norm": 2.6163644790649414, + "learning_rate": 7.68143260140649e-05, + "loss": 0.2321, + "step": 24511 + }, + { + "epoch": 0.5745998816200473, + "grad_norm": 4.4603271484375, + "learning_rate": 7.680716064736146e-05, + "loss": 0.5167, + "step": 24512 + }, + { + "epoch": 0.5746233231948522, + "grad_norm": 2.36885404586792, + "learning_rate": 7.679999540650303e-05, + "loss": 0.2548, + "step": 24513 + }, + { + "epoch": 0.5746467647696573, + "grad_norm": 1.5312328338623047, + "learning_rate": 7.679283029152844e-05, + "loss": 0.1095, + "step": 24514 + }, + { + "epoch": 0.5746702063444622, + "grad_norm": 1.3990570306777954, + "learning_rate": 7.678566530247665e-05, + "loss": 0.9905, + "step": 24515 + }, + { + "epoch": 0.5746936479192672, + "grad_norm": 3.579796552658081, + "learning_rate": 7.677850043938645e-05, + "loss": 0.2577, + "step": 24516 + }, + { + "epoch": 0.5747170894940722, + "grad_norm": 2.8149566650390625, + "learning_rate": 7.677133570229681e-05, + "loss": 0.384, + "step": 24517 + }, + { + "epoch": 0.5747405310688772, + "grad_norm": 3.074242115020752, + "learning_rate": 7.676417109124653e-05, + "loss": 0.3387, + "step": 24518 + }, + { + "epoch": 0.5747639726436822, + "grad_norm": 0.9277353286743164, + "learning_rate": 7.675700660627451e-05, + "loss": 0.1288, + "step": 24519 + }, + { + "epoch": 0.5747874142184872, + "grad_norm": 9.259199142456055, + "learning_rate": 7.674984224741966e-05, + "loss": 0.5468, + "step": 24520 + }, + { + "epoch": 0.5748108557932922, + "grad_norm": 5.257399082183838, + "learning_rate": 7.674267801472077e-05, + "loss": 0.5268, + "step": 24521 + }, + { + "epoch": 0.5748342973680972, + "grad_norm": 1.0554693937301636, + "learning_rate": 7.673551390821679e-05, + "loss": 0.0978, + "step": 24522 + }, + { + "epoch": 0.5748577389429022, + "grad_norm": 3.664186954498291, + "learning_rate": 7.672834992794658e-05, + "loss": 0.3377, + "step": 24523 + }, + { + "epoch": 0.5748811805177072, + "grad_norm": 2.9151201248168945, + "learning_rate": 7.672118607394896e-05, + "loss": 0.2594, + "step": 24524 + }, + { + "epoch": 0.5749046220925121, + "grad_norm": 4.509986877441406, + "learning_rate": 7.671402234626286e-05, + "loss": 0.3578, + "step": 24525 + }, + { + "epoch": 0.5749280636673172, + "grad_norm": 0.7261716723442078, + "learning_rate": 7.670685874492709e-05, + "loss": 0.0527, + "step": 24526 + }, + { + "epoch": 0.5749515052421221, + "grad_norm": 4.072798252105713, + "learning_rate": 7.66996952699806e-05, + "loss": 0.4686, + "step": 24527 + }, + { + "epoch": 0.5749749468169272, + "grad_norm": 1.1302242279052734, + "learning_rate": 7.66925319214622e-05, + "loss": 0.1195, + "step": 24528 + }, + { + "epoch": 0.5749983883917321, + "grad_norm": 2.7085397243499756, + "learning_rate": 7.668536869941081e-05, + "loss": 1.4329, + "step": 24529 + }, + { + "epoch": 0.5750218299665372, + "grad_norm": 0.3815423846244812, + "learning_rate": 7.667820560386525e-05, + "loss": 0.0387, + "step": 24530 + }, + { + "epoch": 0.5750452715413421, + "grad_norm": 0.4573143720626831, + "learning_rate": 7.667104263486439e-05, + "loss": 0.0415, + "step": 24531 + }, + { + "epoch": 0.5750687131161472, + "grad_norm": 2.192882776260376, + "learning_rate": 7.666387979244713e-05, + "loss": 1.7034, + "step": 24532 + }, + { + "epoch": 0.5750921546909521, + "grad_norm": 3.174376964569092, + "learning_rate": 7.665671707665231e-05, + "loss": 0.3975, + "step": 24533 + }, + { + "epoch": 0.5751155962657571, + "grad_norm": 1.3886191844940186, + "learning_rate": 7.664955448751878e-05, + "loss": 0.2192, + "step": 24534 + }, + { + "epoch": 0.5751390378405621, + "grad_norm": 2.7788889408111572, + "learning_rate": 7.664239202508544e-05, + "loss": 0.3176, + "step": 24535 + }, + { + "epoch": 0.5751624794153671, + "grad_norm": 4.4988532066345215, + "learning_rate": 7.663522968939114e-05, + "loss": 0.3259, + "step": 24536 + }, + { + "epoch": 0.5751859209901721, + "grad_norm": 2.7404446601867676, + "learning_rate": 7.662806748047473e-05, + "loss": 0.1859, + "step": 24537 + }, + { + "epoch": 0.5752093625649771, + "grad_norm": 2.683945417404175, + "learning_rate": 7.662090539837508e-05, + "loss": 0.1906, + "step": 24538 + }, + { + "epoch": 0.5752328041397821, + "grad_norm": 3.276298761367798, + "learning_rate": 7.661374344313104e-05, + "loss": 0.245, + "step": 24539 + }, + { + "epoch": 0.5752562457145871, + "grad_norm": 2.469844341278076, + "learning_rate": 7.660658161478151e-05, + "loss": 0.1722, + "step": 24540 + }, + { + "epoch": 0.575279687289392, + "grad_norm": 2.5252649784088135, + "learning_rate": 7.659941991336531e-05, + "loss": 0.3585, + "step": 24541 + }, + { + "epoch": 0.5753031288641971, + "grad_norm": 4.0308074951171875, + "learning_rate": 7.659225833892134e-05, + "loss": 0.5361, + "step": 24542 + }, + { + "epoch": 0.575326570439002, + "grad_norm": 2.995030403137207, + "learning_rate": 7.658509689148843e-05, + "loss": 0.3646, + "step": 24543 + }, + { + "epoch": 0.5753500120138071, + "grad_norm": 1.1345213651657104, + "learning_rate": 7.657793557110542e-05, + "loss": 0.0965, + "step": 24544 + }, + { + "epoch": 0.5753734535886121, + "grad_norm": 1.4638272523880005, + "learning_rate": 7.657077437781122e-05, + "loss": 0.1229, + "step": 24545 + }, + { + "epoch": 0.5753968951634171, + "grad_norm": 1.7746479511260986, + "learning_rate": 7.656361331164464e-05, + "loss": 0.1725, + "step": 24546 + }, + { + "epoch": 0.5754203367382221, + "grad_norm": 3.0784480571746826, + "learning_rate": 7.655645237264456e-05, + "loss": 0.2497, + "step": 24547 + }, + { + "epoch": 0.5754437783130271, + "grad_norm": 1.5249751806259155, + "learning_rate": 7.654929156084982e-05, + "loss": 0.1181, + "step": 24548 + }, + { + "epoch": 0.5754672198878321, + "grad_norm": 3.298299789428711, + "learning_rate": 7.654213087629927e-05, + "loss": 0.513, + "step": 24549 + }, + { + "epoch": 0.575490661462637, + "grad_norm": 3.7494750022888184, + "learning_rate": 7.653497031903181e-05, + "loss": 0.2481, + "step": 24550 + }, + { + "epoch": 0.5755141030374421, + "grad_norm": 3.0355234146118164, + "learning_rate": 7.652780988908621e-05, + "loss": 0.2507, + "step": 24551 + }, + { + "epoch": 0.575537544612247, + "grad_norm": 0.8512369394302368, + "learning_rate": 7.652064958650142e-05, + "loss": 0.0414, + "step": 24552 + }, + { + "epoch": 0.5755609861870521, + "grad_norm": 1.9959096908569336, + "learning_rate": 7.651348941131625e-05, + "loss": 0.1356, + "step": 24553 + }, + { + "epoch": 0.575584427761857, + "grad_norm": 4.223534107208252, + "learning_rate": 7.65063293635695e-05, + "loss": 0.3983, + "step": 24554 + }, + { + "epoch": 0.5756078693366621, + "grad_norm": 2.771472930908203, + "learning_rate": 7.649916944330012e-05, + "loss": 0.1872, + "step": 24555 + }, + { + "epoch": 0.575631310911467, + "grad_norm": 5.3520379066467285, + "learning_rate": 7.649200965054687e-05, + "loss": 0.2261, + "step": 24556 + }, + { + "epoch": 0.5756547524862721, + "grad_norm": 1.1251964569091797, + "learning_rate": 7.648484998534866e-05, + "loss": 0.134, + "step": 24557 + }, + { + "epoch": 0.575678194061077, + "grad_norm": 4.511096477508545, + "learning_rate": 7.647769044774433e-05, + "loss": 0.3051, + "step": 24558 + }, + { + "epoch": 0.5757016356358821, + "grad_norm": 1.2086466550827026, + "learning_rate": 7.647053103777267e-05, + "loss": 0.1005, + "step": 24559 + }, + { + "epoch": 0.575725077210687, + "grad_norm": 1.6570831537246704, + "learning_rate": 7.64633717554726e-05, + "loss": 0.1931, + "step": 24560 + }, + { + "epoch": 0.575748518785492, + "grad_norm": 2.864579677581787, + "learning_rate": 7.645621260088293e-05, + "loss": 0.2259, + "step": 24561 + }, + { + "epoch": 0.575771960360297, + "grad_norm": 1.1550171375274658, + "learning_rate": 7.64490535740425e-05, + "loss": 0.1497, + "step": 24562 + }, + { + "epoch": 0.575795401935102, + "grad_norm": 2.5343985557556152, + "learning_rate": 7.644189467499015e-05, + "loss": 1.5105, + "step": 24563 + }, + { + "epoch": 0.575818843509907, + "grad_norm": 3.3770363330841064, + "learning_rate": 7.643473590376477e-05, + "loss": 0.4553, + "step": 24564 + }, + { + "epoch": 0.575842285084712, + "grad_norm": 3.17386531829834, + "learning_rate": 7.642757726040519e-05, + "loss": 0.3373, + "step": 24565 + }, + { + "epoch": 0.575865726659517, + "grad_norm": 1.5957674980163574, + "learning_rate": 7.64204187449502e-05, + "loss": 0.1093, + "step": 24566 + }, + { + "epoch": 0.575889168234322, + "grad_norm": 3.1489744186401367, + "learning_rate": 7.641326035743872e-05, + "loss": 0.2057, + "step": 24567 + }, + { + "epoch": 0.575912609809127, + "grad_norm": 2.9893670082092285, + "learning_rate": 7.640610209790955e-05, + "loss": 1.5029, + "step": 24568 + }, + { + "epoch": 0.575936051383932, + "grad_norm": 2.4067599773406982, + "learning_rate": 7.63989439664015e-05, + "loss": 0.1741, + "step": 24569 + }, + { + "epoch": 0.5759594929587369, + "grad_norm": 2.998121976852417, + "learning_rate": 7.639178596295347e-05, + "loss": 0.2822, + "step": 24570 + }, + { + "epoch": 0.575982934533542, + "grad_norm": 1.865946650505066, + "learning_rate": 7.638462808760428e-05, + "loss": 0.172, + "step": 24571 + }, + { + "epoch": 0.5760063761083469, + "grad_norm": 2.5835378170013428, + "learning_rate": 7.637747034039275e-05, + "loss": 0.2726, + "step": 24572 + }, + { + "epoch": 0.576029817683152, + "grad_norm": 0.7837176322937012, + "learning_rate": 7.637031272135772e-05, + "loss": 0.0562, + "step": 24573 + }, + { + "epoch": 0.5760532592579569, + "grad_norm": 0.5615963339805603, + "learning_rate": 7.636315523053806e-05, + "loss": 0.0506, + "step": 24574 + }, + { + "epoch": 0.576076700832762, + "grad_norm": 3.1233842372894287, + "learning_rate": 7.635599786797253e-05, + "loss": 1.6025, + "step": 24575 + }, + { + "epoch": 0.5761001424075669, + "grad_norm": 1.8023492097854614, + "learning_rate": 7.634884063370007e-05, + "loss": 0.1539, + "step": 24576 + }, + { + "epoch": 0.576123583982372, + "grad_norm": 6.929950714111328, + "learning_rate": 7.634168352775945e-05, + "loss": 0.4616, + "step": 24577 + }, + { + "epoch": 0.5761470255571769, + "grad_norm": 2.6547372341156006, + "learning_rate": 7.633452655018955e-05, + "loss": 1.2674, + "step": 24578 + }, + { + "epoch": 0.576170467131982, + "grad_norm": 2.595968723297119, + "learning_rate": 7.632736970102914e-05, + "loss": 0.1823, + "step": 24579 + }, + { + "epoch": 0.5761939087067869, + "grad_norm": 3.185950517654419, + "learning_rate": 7.632021298031711e-05, + "loss": 0.303, + "step": 24580 + }, + { + "epoch": 0.5762173502815919, + "grad_norm": 3.487295627593994, + "learning_rate": 7.631305638809226e-05, + "loss": 0.3804, + "step": 24581 + }, + { + "epoch": 0.5762407918563969, + "grad_norm": 4.9452948570251465, + "learning_rate": 7.630589992439342e-05, + "loss": 0.3359, + "step": 24582 + }, + { + "epoch": 0.5762642334312019, + "grad_norm": 0.4113912582397461, + "learning_rate": 7.629874358925946e-05, + "loss": 0.0363, + "step": 24583 + }, + { + "epoch": 0.5762876750060069, + "grad_norm": 4.396946430206299, + "learning_rate": 7.629158738272919e-05, + "loss": 0.3496, + "step": 24584 + }, + { + "epoch": 0.5763111165808119, + "grad_norm": 3.445509910583496, + "learning_rate": 7.628443130484139e-05, + "loss": 0.2852, + "step": 24585 + }, + { + "epoch": 0.5763345581556168, + "grad_norm": 1.0119041204452515, + "learning_rate": 7.627727535563496e-05, + "loss": 0.0749, + "step": 24586 + }, + { + "epoch": 0.5763579997304219, + "grad_norm": 2.9125096797943115, + "learning_rate": 7.627011953514867e-05, + "loss": 0.5566, + "step": 24587 + }, + { + "epoch": 0.5763814413052268, + "grad_norm": 3.470202684402466, + "learning_rate": 7.626296384342141e-05, + "loss": 1.3317, + "step": 24588 + }, + { + "epoch": 0.5764048828800319, + "grad_norm": 3.378399133682251, + "learning_rate": 7.625580828049194e-05, + "loss": 0.4528, + "step": 24589 + }, + { + "epoch": 0.5764283244548368, + "grad_norm": 3.5163838863372803, + "learning_rate": 7.624865284639917e-05, + "loss": 0.2192, + "step": 24590 + }, + { + "epoch": 0.5764517660296419, + "grad_norm": 2.203204870223999, + "learning_rate": 7.624149754118187e-05, + "loss": 1.512, + "step": 24591 + }, + { + "epoch": 0.5764752076044468, + "grad_norm": 3.404998302459717, + "learning_rate": 7.623434236487885e-05, + "loss": 0.3743, + "step": 24592 + }, + { + "epoch": 0.5764986491792519, + "grad_norm": 4.066580772399902, + "learning_rate": 7.622718731752897e-05, + "loss": 0.4121, + "step": 24593 + }, + { + "epoch": 0.5765220907540568, + "grad_norm": 3.626695156097412, + "learning_rate": 7.622003239917103e-05, + "loss": 0.2638, + "step": 24594 + }, + { + "epoch": 0.5765455323288619, + "grad_norm": 1.176235318183899, + "learning_rate": 7.621287760984388e-05, + "loss": 0.0664, + "step": 24595 + }, + { + "epoch": 0.5765689739036669, + "grad_norm": 2.58184814453125, + "learning_rate": 7.620572294958634e-05, + "loss": 0.2346, + "step": 24596 + }, + { + "epoch": 0.5765924154784718, + "grad_norm": 2.5777320861816406, + "learning_rate": 7.619856841843717e-05, + "loss": 0.2826, + "step": 24597 + }, + { + "epoch": 0.5766158570532769, + "grad_norm": 4.560110569000244, + "learning_rate": 7.619141401643527e-05, + "loss": 0.3406, + "step": 24598 + }, + { + "epoch": 0.5766392986280818, + "grad_norm": 1.5511503219604492, + "learning_rate": 7.618425974361938e-05, + "loss": 0.1366, + "step": 24599 + }, + { + "epoch": 0.5766627402028869, + "grad_norm": 2.677053928375244, + "learning_rate": 7.617710560002842e-05, + "loss": 1.4539, + "step": 24600 + }, + { + "epoch": 0.5766861817776918, + "grad_norm": 2.140462636947632, + "learning_rate": 7.616995158570113e-05, + "loss": 0.1968, + "step": 24601 + }, + { + "epoch": 0.5767096233524969, + "grad_norm": 4.407143592834473, + "learning_rate": 7.616279770067635e-05, + "loss": 0.4257, + "step": 24602 + }, + { + "epoch": 0.5767330649273018, + "grad_norm": 1.242441177368164, + "learning_rate": 7.615564394499293e-05, + "loss": 0.1414, + "step": 24603 + }, + { + "epoch": 0.5767565065021069, + "grad_norm": 3.7672743797302246, + "learning_rate": 7.614849031868962e-05, + "loss": 0.3046, + "step": 24604 + }, + { + "epoch": 0.5767799480769118, + "grad_norm": 1.9503427743911743, + "learning_rate": 7.61413368218053e-05, + "loss": 0.4761, + "step": 24605 + }, + { + "epoch": 0.5768033896517168, + "grad_norm": 1.8050739765167236, + "learning_rate": 7.613418345437876e-05, + "loss": 0.1481, + "step": 24606 + }, + { + "epoch": 0.5768268312265218, + "grad_norm": 3.078355073928833, + "learning_rate": 7.612703021644878e-05, + "loss": 0.2692, + "step": 24607 + }, + { + "epoch": 0.5768502728013268, + "grad_norm": 1.5435967445373535, + "learning_rate": 7.611987710805424e-05, + "loss": 0.1555, + "step": 24608 + }, + { + "epoch": 0.5768737143761318, + "grad_norm": 3.7353882789611816, + "learning_rate": 7.61127241292339e-05, + "loss": 0.3724, + "step": 24609 + }, + { + "epoch": 0.5768971559509368, + "grad_norm": 1.5017894506454468, + "learning_rate": 7.610557128002658e-05, + "loss": 0.1835, + "step": 24610 + }, + { + "epoch": 0.5769205975257418, + "grad_norm": 7.0848708152771, + "learning_rate": 7.609841856047109e-05, + "loss": 0.7291, + "step": 24611 + }, + { + "epoch": 0.5769440391005468, + "grad_norm": 2.236882209777832, + "learning_rate": 7.609126597060628e-05, + "loss": 1.2041, + "step": 24612 + }, + { + "epoch": 0.5769674806753518, + "grad_norm": 0.8787895441055298, + "learning_rate": 7.608411351047093e-05, + "loss": 0.0668, + "step": 24613 + }, + { + "epoch": 0.5769909222501568, + "grad_norm": 4.229663372039795, + "learning_rate": 7.607696118010386e-05, + "loss": 0.5212, + "step": 24614 + }, + { + "epoch": 0.5770143638249617, + "grad_norm": 0.7499812245368958, + "learning_rate": 7.606980897954387e-05, + "loss": 0.0579, + "step": 24615 + }, + { + "epoch": 0.5770378053997668, + "grad_norm": 3.3609867095947266, + "learning_rate": 7.606265690882976e-05, + "loss": 0.4126, + "step": 24616 + }, + { + "epoch": 0.5770612469745717, + "grad_norm": 3.1571595668792725, + "learning_rate": 7.605550496800035e-05, + "loss": 0.2556, + "step": 24617 + }, + { + "epoch": 0.5770846885493768, + "grad_norm": 3.3657889366149902, + "learning_rate": 7.604835315709444e-05, + "loss": 0.4751, + "step": 24618 + }, + { + "epoch": 0.5771081301241817, + "grad_norm": 3.1228463649749756, + "learning_rate": 7.604120147615085e-05, + "loss": 0.2155, + "step": 24619 + }, + { + "epoch": 0.5771315716989868, + "grad_norm": 1.394768238067627, + "learning_rate": 7.603404992520836e-05, + "loss": 0.1303, + "step": 24620 + }, + { + "epoch": 0.5771550132737917, + "grad_norm": 2.3673813343048096, + "learning_rate": 7.60268985043058e-05, + "loss": 0.2983, + "step": 24621 + }, + { + "epoch": 0.5771784548485968, + "grad_norm": 1.7097078561782837, + "learning_rate": 7.601974721348197e-05, + "loss": 0.1555, + "step": 24622 + }, + { + "epoch": 0.5772018964234017, + "grad_norm": 3.3808038234710693, + "learning_rate": 7.601259605277565e-05, + "loss": 0.2032, + "step": 24623 + }, + { + "epoch": 0.5772253379982067, + "grad_norm": 1.2034999132156372, + "learning_rate": 7.600544502222561e-05, + "loss": 0.119, + "step": 24624 + }, + { + "epoch": 0.5772487795730117, + "grad_norm": 1.1017751693725586, + "learning_rate": 7.599829412187077e-05, + "loss": 0.1527, + "step": 24625 + }, + { + "epoch": 0.5772722211478167, + "grad_norm": 3.1419968605041504, + "learning_rate": 7.599114335174986e-05, + "loss": 0.1743, + "step": 24626 + }, + { + "epoch": 0.5772956627226217, + "grad_norm": 3.3318636417388916, + "learning_rate": 7.598399271190165e-05, + "loss": 0.2239, + "step": 24627 + }, + { + "epoch": 0.5773191042974267, + "grad_norm": 4.319878578186035, + "learning_rate": 7.5976842202365e-05, + "loss": 0.3157, + "step": 24628 + }, + { + "epoch": 0.5773425458722317, + "grad_norm": 2.8280205726623535, + "learning_rate": 7.596969182317866e-05, + "loss": 0.2478, + "step": 24629 + }, + { + "epoch": 0.5773659874470367, + "grad_norm": 2.810054063796997, + "learning_rate": 7.596254157438145e-05, + "loss": 0.3618, + "step": 24630 + }, + { + "epoch": 0.5773894290218416, + "grad_norm": 2.8123905658721924, + "learning_rate": 7.595539145601218e-05, + "loss": 0.2018, + "step": 24631 + }, + { + "epoch": 0.5774128705966467, + "grad_norm": 3.5765926837921143, + "learning_rate": 7.59482414681096e-05, + "loss": 0.4181, + "step": 24632 + }, + { + "epoch": 0.5774363121714516, + "grad_norm": 3.374052047729492, + "learning_rate": 7.594109161071258e-05, + "loss": 0.1282, + "step": 24633 + }, + { + "epoch": 0.5774597537462567, + "grad_norm": 1.4528850317001343, + "learning_rate": 7.593394188385987e-05, + "loss": 0.1322, + "step": 24634 + }, + { + "epoch": 0.5774831953210616, + "grad_norm": 2.275306463241577, + "learning_rate": 7.592679228759022e-05, + "loss": 0.2683, + "step": 24635 + }, + { + "epoch": 0.5775066368958667, + "grad_norm": 1.1039090156555176, + "learning_rate": 7.591964282194247e-05, + "loss": 0.0446, + "step": 24636 + }, + { + "epoch": 0.5775300784706716, + "grad_norm": 1.3762191534042358, + "learning_rate": 7.591249348695546e-05, + "loss": 0.8491, + "step": 24637 + }, + { + "epoch": 0.5775535200454767, + "grad_norm": 0.5393273830413818, + "learning_rate": 7.590534428266794e-05, + "loss": 0.0339, + "step": 24638 + }, + { + "epoch": 0.5775769616202816, + "grad_norm": 0.58859783411026, + "learning_rate": 7.589819520911866e-05, + "loss": 0.0398, + "step": 24639 + }, + { + "epoch": 0.5776004031950867, + "grad_norm": 3.466658115386963, + "learning_rate": 7.589104626634649e-05, + "loss": 0.3202, + "step": 24640 + }, + { + "epoch": 0.5776238447698916, + "grad_norm": 5.207369327545166, + "learning_rate": 7.588389745439017e-05, + "loss": 0.3361, + "step": 24641 + }, + { + "epoch": 0.5776472863446966, + "grad_norm": 2.112172842025757, + "learning_rate": 7.587674877328849e-05, + "loss": 1.3615, + "step": 24642 + }, + { + "epoch": 0.5776707279195016, + "grad_norm": 0.5535882115364075, + "learning_rate": 7.586960022308026e-05, + "loss": 0.038, + "step": 24643 + }, + { + "epoch": 0.5776941694943066, + "grad_norm": 4.177436351776123, + "learning_rate": 7.586245180380428e-05, + "loss": 0.2951, + "step": 24644 + }, + { + "epoch": 0.5777176110691116, + "grad_norm": 1.6413052082061768, + "learning_rate": 7.585530351549926e-05, + "loss": 0.1264, + "step": 24645 + }, + { + "epoch": 0.5777410526439166, + "grad_norm": 0.9864651560783386, + "learning_rate": 7.584815535820408e-05, + "loss": 0.1191, + "step": 24646 + }, + { + "epoch": 0.5777644942187217, + "grad_norm": 4.633902072906494, + "learning_rate": 7.584100733195749e-05, + "loss": 0.2457, + "step": 24647 + }, + { + "epoch": 0.5777879357935266, + "grad_norm": 3.9827773571014404, + "learning_rate": 7.583385943679822e-05, + "loss": 0.3831, + "step": 24648 + }, + { + "epoch": 0.5778113773683317, + "grad_norm": 0.675726592540741, + "learning_rate": 7.582671167276514e-05, + "loss": 0.0454, + "step": 24649 + }, + { + "epoch": 0.5778348189431366, + "grad_norm": 1.859920859336853, + "learning_rate": 7.581956403989702e-05, + "loss": 0.1313, + "step": 24650 + }, + { + "epoch": 0.5778582605179416, + "grad_norm": 3.987787961959839, + "learning_rate": 7.581241653823263e-05, + "loss": 0.2968, + "step": 24651 + }, + { + "epoch": 0.5778817020927466, + "grad_norm": 5.219505310058594, + "learning_rate": 7.580526916781072e-05, + "loss": 0.5361, + "step": 24652 + }, + { + "epoch": 0.5779051436675516, + "grad_norm": 2.0264151096343994, + "learning_rate": 7.579812192867013e-05, + "loss": 0.1944, + "step": 24653 + }, + { + "epoch": 0.5779285852423566, + "grad_norm": 3.8386917114257812, + "learning_rate": 7.57909748208496e-05, + "loss": 0.5111, + "step": 24654 + }, + { + "epoch": 0.5779520268171616, + "grad_norm": 3.168391704559326, + "learning_rate": 7.57838278443879e-05, + "loss": 0.3222, + "step": 24655 + }, + { + "epoch": 0.5779754683919666, + "grad_norm": 3.431641101837158, + "learning_rate": 7.577668099932387e-05, + "loss": 0.2974, + "step": 24656 + }, + { + "epoch": 0.5779989099667716, + "grad_norm": 2.6927878856658936, + "learning_rate": 7.576953428569624e-05, + "loss": 0.1315, + "step": 24657 + }, + { + "epoch": 0.5780223515415766, + "grad_norm": 0.5587016344070435, + "learning_rate": 7.576238770354377e-05, + "loss": 0.0348, + "step": 24658 + }, + { + "epoch": 0.5780457931163816, + "grad_norm": 3.3854641914367676, + "learning_rate": 7.575524125290529e-05, + "loss": 0.275, + "step": 24659 + }, + { + "epoch": 0.5780692346911865, + "grad_norm": 4.2048163414001465, + "learning_rate": 7.574809493381951e-05, + "loss": 0.3751, + "step": 24660 + }, + { + "epoch": 0.5780926762659916, + "grad_norm": 1.2363914251327515, + "learning_rate": 7.57409487463253e-05, + "loss": 0.0727, + "step": 24661 + }, + { + "epoch": 0.5781161178407965, + "grad_norm": 1.0309420824050903, + "learning_rate": 7.573380269046136e-05, + "loss": 0.0985, + "step": 24662 + }, + { + "epoch": 0.5781395594156016, + "grad_norm": 5.0450239181518555, + "learning_rate": 7.57266567662665e-05, + "loss": 0.3222, + "step": 24663 + }, + { + "epoch": 0.5781630009904065, + "grad_norm": 2.0045053958892822, + "learning_rate": 7.571951097377948e-05, + "loss": 0.1948, + "step": 24664 + }, + { + "epoch": 0.5781864425652116, + "grad_norm": 0.2845708727836609, + "learning_rate": 7.571236531303907e-05, + "loss": 0.0178, + "step": 24665 + }, + { + "epoch": 0.5782098841400165, + "grad_norm": 2.287472724914551, + "learning_rate": 7.570521978408407e-05, + "loss": 0.2163, + "step": 24666 + }, + { + "epoch": 0.5782333257148216, + "grad_norm": 1.3687655925750732, + "learning_rate": 7.569807438695323e-05, + "loss": 0.1429, + "step": 24667 + }, + { + "epoch": 0.5782567672896265, + "grad_norm": 3.4402003288269043, + "learning_rate": 7.569092912168529e-05, + "loss": 0.3853, + "step": 24668 + }, + { + "epoch": 0.5782802088644315, + "grad_norm": 2.911090612411499, + "learning_rate": 7.568378398831908e-05, + "loss": 0.2197, + "step": 24669 + }, + { + "epoch": 0.5783036504392365, + "grad_norm": 2.3560192584991455, + "learning_rate": 7.567663898689332e-05, + "loss": 0.1877, + "step": 24670 + }, + { + "epoch": 0.5783270920140415, + "grad_norm": 0.6749258041381836, + "learning_rate": 7.566949411744681e-05, + "loss": 0.0469, + "step": 24671 + }, + { + "epoch": 0.5783505335888465, + "grad_norm": 2.5314419269561768, + "learning_rate": 7.566234938001828e-05, + "loss": 0.2463, + "step": 24672 + }, + { + "epoch": 0.5783739751636515, + "grad_norm": 3.3481719493865967, + "learning_rate": 7.565520477464657e-05, + "loss": 0.2863, + "step": 24673 + }, + { + "epoch": 0.5783974167384565, + "grad_norm": 3.9520132541656494, + "learning_rate": 7.564806030137037e-05, + "loss": 0.3606, + "step": 24674 + }, + { + "epoch": 0.5784208583132615, + "grad_norm": 2.378164529800415, + "learning_rate": 7.564091596022851e-05, + "loss": 0.1505, + "step": 24675 + }, + { + "epoch": 0.5784442998880664, + "grad_norm": 2.06961727142334, + "learning_rate": 7.563377175125973e-05, + "loss": 0.1825, + "step": 24676 + }, + { + "epoch": 0.5784677414628715, + "grad_norm": 2.743076801300049, + "learning_rate": 7.562662767450276e-05, + "loss": 0.2349, + "step": 24677 + }, + { + "epoch": 0.5784911830376764, + "grad_norm": 3.077540636062622, + "learning_rate": 7.561948372999642e-05, + "loss": 0.2843, + "step": 24678 + }, + { + "epoch": 0.5785146246124815, + "grad_norm": 1.3804495334625244, + "learning_rate": 7.561233991777945e-05, + "loss": 0.1541, + "step": 24679 + }, + { + "epoch": 0.5785380661872864, + "grad_norm": 0.8081443905830383, + "learning_rate": 7.560519623789058e-05, + "loss": 0.0603, + "step": 24680 + }, + { + "epoch": 0.5785615077620915, + "grad_norm": 1.0486892461776733, + "learning_rate": 7.559805269036862e-05, + "loss": 0.0316, + "step": 24681 + }, + { + "epoch": 0.5785849493368964, + "grad_norm": 3.6064839363098145, + "learning_rate": 7.559090927525231e-05, + "loss": 0.3784, + "step": 24682 + }, + { + "epoch": 0.5786083909117015, + "grad_norm": 0.7746391892433167, + "learning_rate": 7.558376599258041e-05, + "loss": 0.079, + "step": 24683 + }, + { + "epoch": 0.5786318324865064, + "grad_norm": 3.8890204429626465, + "learning_rate": 7.557662284239164e-05, + "loss": 0.4051, + "step": 24684 + }, + { + "epoch": 0.5786552740613115, + "grad_norm": 3.775299549102783, + "learning_rate": 7.556947982472487e-05, + "loss": 0.2263, + "step": 24685 + }, + { + "epoch": 0.5786787156361164, + "grad_norm": 3.677011251449585, + "learning_rate": 7.556233693961876e-05, + "loss": 0.2361, + "step": 24686 + }, + { + "epoch": 0.5787021572109214, + "grad_norm": 5.175505638122559, + "learning_rate": 7.555519418711209e-05, + "loss": 0.2626, + "step": 24687 + }, + { + "epoch": 0.5787255987857264, + "grad_norm": 3.078437089920044, + "learning_rate": 7.554805156724366e-05, + "loss": 0.3359, + "step": 24688 + }, + { + "epoch": 0.5787490403605314, + "grad_norm": 1.0205062627792358, + "learning_rate": 7.554090908005216e-05, + "loss": 0.1057, + "step": 24689 + }, + { + "epoch": 0.5787724819353364, + "grad_norm": 2.645552158355713, + "learning_rate": 7.553376672557636e-05, + "loss": 0.152, + "step": 24690 + }, + { + "epoch": 0.5787959235101414, + "grad_norm": 2.8447773456573486, + "learning_rate": 7.552662450385506e-05, + "loss": 0.2336, + "step": 24691 + }, + { + "epoch": 0.5788193650849464, + "grad_norm": 0.4040132462978363, + "learning_rate": 7.551948241492698e-05, + "loss": 0.0402, + "step": 24692 + }, + { + "epoch": 0.5788428066597514, + "grad_norm": 4.031641960144043, + "learning_rate": 7.551234045883086e-05, + "loss": 0.2957, + "step": 24693 + }, + { + "epoch": 0.5788662482345563, + "grad_norm": 5.235466480255127, + "learning_rate": 7.550519863560548e-05, + "loss": 0.3087, + "step": 24694 + }, + { + "epoch": 0.5788896898093614, + "grad_norm": 2.9780545234680176, + "learning_rate": 7.54980569452896e-05, + "loss": 0.2895, + "step": 24695 + }, + { + "epoch": 0.5789131313841663, + "grad_norm": 4.413323879241943, + "learning_rate": 7.549091538792191e-05, + "loss": 0.5054, + "step": 24696 + }, + { + "epoch": 0.5789365729589714, + "grad_norm": 4.086065769195557, + "learning_rate": 7.548377396354119e-05, + "loss": 0.4273, + "step": 24697 + }, + { + "epoch": 0.5789600145337763, + "grad_norm": 1.221685767173767, + "learning_rate": 7.547663267218624e-05, + "loss": 0.0962, + "step": 24698 + }, + { + "epoch": 0.5789834561085814, + "grad_norm": 4.471377372741699, + "learning_rate": 7.546949151389578e-05, + "loss": 0.4544, + "step": 24699 + }, + { + "epoch": 0.5790068976833864, + "grad_norm": 0.39668765664100647, + "learning_rate": 7.546235048870852e-05, + "loss": 0.0173, + "step": 24700 + }, + { + "epoch": 0.5790303392581914, + "grad_norm": 2.9657113552093506, + "learning_rate": 7.545520959666326e-05, + "loss": 0.2075, + "step": 24701 + }, + { + "epoch": 0.5790537808329964, + "grad_norm": 4.09439754486084, + "learning_rate": 7.544806883779873e-05, + "loss": 0.3587, + "step": 24702 + }, + { + "epoch": 0.5790772224078014, + "grad_norm": 2.4160149097442627, + "learning_rate": 7.544092821215363e-05, + "loss": 0.1584, + "step": 24703 + }, + { + "epoch": 0.5791006639826064, + "grad_norm": 4.77292537689209, + "learning_rate": 7.543378771976678e-05, + "loss": 0.8157, + "step": 24704 + }, + { + "epoch": 0.5791241055574113, + "grad_norm": 2.5678603649139404, + "learning_rate": 7.542664736067686e-05, + "loss": 0.2609, + "step": 24705 + }, + { + "epoch": 0.5791475471322164, + "grad_norm": 2.4741950035095215, + "learning_rate": 7.541950713492267e-05, + "loss": 0.2192, + "step": 24706 + }, + { + "epoch": 0.5791709887070213, + "grad_norm": 0.7909097671508789, + "learning_rate": 7.541236704254291e-05, + "loss": 0.0675, + "step": 24707 + }, + { + "epoch": 0.5791944302818264, + "grad_norm": 2.9173483848571777, + "learning_rate": 7.540522708357634e-05, + "loss": 0.4447, + "step": 24708 + }, + { + "epoch": 0.5792178718566313, + "grad_norm": 3.3342630863189697, + "learning_rate": 7.539808725806168e-05, + "loss": 0.1949, + "step": 24709 + }, + { + "epoch": 0.5792413134314364, + "grad_norm": 0.6694237589836121, + "learning_rate": 7.53909475660377e-05, + "loss": 0.0684, + "step": 24710 + }, + { + "epoch": 0.5792647550062413, + "grad_norm": 2.95609712600708, + "learning_rate": 7.538380800754315e-05, + "loss": 0.4307, + "step": 24711 + }, + { + "epoch": 0.5792881965810464, + "grad_norm": 6.315615177154541, + "learning_rate": 7.537666858261674e-05, + "loss": 0.5578, + "step": 24712 + }, + { + "epoch": 0.5793116381558513, + "grad_norm": 3.35791277885437, + "learning_rate": 7.536952929129722e-05, + "loss": 0.3922, + "step": 24713 + }, + { + "epoch": 0.5793350797306563, + "grad_norm": 3.7752180099487305, + "learning_rate": 7.536239013362334e-05, + "loss": 0.3576, + "step": 24714 + }, + { + "epoch": 0.5793585213054613, + "grad_norm": 2.163970708847046, + "learning_rate": 7.535525110963379e-05, + "loss": 1.5286, + "step": 24715 + }, + { + "epoch": 0.5793819628802663, + "grad_norm": 2.357069969177246, + "learning_rate": 7.534811221936737e-05, + "loss": 0.2172, + "step": 24716 + }, + { + "epoch": 0.5794054044550713, + "grad_norm": 1.5030995607376099, + "learning_rate": 7.534097346286279e-05, + "loss": 0.0918, + "step": 24717 + }, + { + "epoch": 0.5794288460298763, + "grad_norm": 2.4757542610168457, + "learning_rate": 7.533383484015876e-05, + "loss": 0.2254, + "step": 24718 + }, + { + "epoch": 0.5794522876046813, + "grad_norm": 2.9215033054351807, + "learning_rate": 7.532669635129404e-05, + "loss": 0.3483, + "step": 24719 + }, + { + "epoch": 0.5794757291794863, + "grad_norm": 3.562734603881836, + "learning_rate": 7.531955799630738e-05, + "loss": 0.5222, + "step": 24720 + }, + { + "epoch": 0.5794991707542912, + "grad_norm": 1.107506513595581, + "learning_rate": 7.531241977523744e-05, + "loss": 0.1255, + "step": 24721 + }, + { + "epoch": 0.5795226123290963, + "grad_norm": 1.0464544296264648, + "learning_rate": 7.530528168812303e-05, + "loss": 0.0696, + "step": 24722 + }, + { + "epoch": 0.5795460539039012, + "grad_norm": 0.9561673402786255, + "learning_rate": 7.529814373500287e-05, + "loss": 0.0641, + "step": 24723 + }, + { + "epoch": 0.5795694954787063, + "grad_norm": 4.089236736297607, + "learning_rate": 7.529100591591567e-05, + "loss": 0.3611, + "step": 24724 + }, + { + "epoch": 0.5795929370535112, + "grad_norm": 3.938770294189453, + "learning_rate": 7.528386823090017e-05, + "loss": 0.3382, + "step": 24725 + }, + { + "epoch": 0.5796163786283163, + "grad_norm": 2.41025447845459, + "learning_rate": 7.52767306799951e-05, + "loss": 0.287, + "step": 24726 + }, + { + "epoch": 0.5796398202031212, + "grad_norm": 2.4971721172332764, + "learning_rate": 7.52695932632392e-05, + "loss": 1.3277, + "step": 24727 + }, + { + "epoch": 0.5796632617779263, + "grad_norm": 2.401394844055176, + "learning_rate": 7.526245598067115e-05, + "loss": 0.1669, + "step": 24728 + }, + { + "epoch": 0.5796867033527312, + "grad_norm": 0.36699721217155457, + "learning_rate": 7.525531883232973e-05, + "loss": 0.0334, + "step": 24729 + }, + { + "epoch": 0.5797101449275363, + "grad_norm": 1.7104482650756836, + "learning_rate": 7.524818181825364e-05, + "loss": 1.0819, + "step": 24730 + }, + { + "epoch": 0.5797335865023412, + "grad_norm": 1.817075252532959, + "learning_rate": 7.524104493848159e-05, + "loss": 0.0913, + "step": 24731 + }, + { + "epoch": 0.5797570280771462, + "grad_norm": 4.065634727478027, + "learning_rate": 7.523390819305235e-05, + "loss": 0.2813, + "step": 24732 + }, + { + "epoch": 0.5797804696519512, + "grad_norm": 3.1242828369140625, + "learning_rate": 7.522677158200459e-05, + "loss": 0.155, + "step": 24733 + }, + { + "epoch": 0.5798039112267562, + "grad_norm": 3.866579055786133, + "learning_rate": 7.52196351053771e-05, + "loss": 0.2672, + "step": 24734 + }, + { + "epoch": 0.5798273528015612, + "grad_norm": 0.9478734731674194, + "learning_rate": 7.521249876320853e-05, + "loss": 0.0901, + "step": 24735 + }, + { + "epoch": 0.5798507943763662, + "grad_norm": 3.6003241539001465, + "learning_rate": 7.520536255553767e-05, + "loss": 0.1324, + "step": 24736 + }, + { + "epoch": 0.5798742359511712, + "grad_norm": 2.326991081237793, + "learning_rate": 7.519822648240322e-05, + "loss": 1.9334, + "step": 24737 + }, + { + "epoch": 0.5798976775259762, + "grad_norm": 2.3875458240509033, + "learning_rate": 7.519109054384387e-05, + "loss": 0.2507, + "step": 24738 + }, + { + "epoch": 0.5799211191007811, + "grad_norm": 4.928410053253174, + "learning_rate": 7.518395473989836e-05, + "loss": 0.3734, + "step": 24739 + }, + { + "epoch": 0.5799445606755862, + "grad_norm": 2.6378674507141113, + "learning_rate": 7.517681907060543e-05, + "loss": 0.2038, + "step": 24740 + }, + { + "epoch": 0.5799680022503911, + "grad_norm": 2.93745756149292, + "learning_rate": 7.516968353600376e-05, + "loss": 0.3714, + "step": 24741 + }, + { + "epoch": 0.5799914438251962, + "grad_norm": 2.9613869190216064, + "learning_rate": 7.516254813613209e-05, + "loss": 0.146, + "step": 24742 + }, + { + "epoch": 0.5800148854000011, + "grad_norm": 2.5671498775482178, + "learning_rate": 7.515541287102912e-05, + "loss": 0.2063, + "step": 24743 + }, + { + "epoch": 0.5800383269748062, + "grad_norm": 2.9713332653045654, + "learning_rate": 7.514827774073359e-05, + "loss": 0.3373, + "step": 24744 + }, + { + "epoch": 0.5800617685496111, + "grad_norm": 2.0396568775177, + "learning_rate": 7.514114274528417e-05, + "loss": 0.0979, + "step": 24745 + }, + { + "epoch": 0.5800852101244162, + "grad_norm": 2.8075549602508545, + "learning_rate": 7.513400788471965e-05, + "loss": 0.1972, + "step": 24746 + }, + { + "epoch": 0.5801086516992211, + "grad_norm": 2.2246804237365723, + "learning_rate": 7.512687315907869e-05, + "loss": 0.179, + "step": 24747 + }, + { + "epoch": 0.5801320932740262, + "grad_norm": 2.808537483215332, + "learning_rate": 7.511973856840001e-05, + "loss": 0.2847, + "step": 24748 + }, + { + "epoch": 0.5801555348488311, + "grad_norm": 0.38557782769203186, + "learning_rate": 7.511260411272235e-05, + "loss": 0.0103, + "step": 24749 + }, + { + "epoch": 0.5801789764236361, + "grad_norm": 0.7669500708580017, + "learning_rate": 7.510546979208436e-05, + "loss": 0.0525, + "step": 24750 + }, + { + "epoch": 0.5802024179984412, + "grad_norm": 0.7587719559669495, + "learning_rate": 7.509833560652483e-05, + "loss": 0.0415, + "step": 24751 + }, + { + "epoch": 0.5802258595732461, + "grad_norm": 7.748475074768066, + "learning_rate": 7.509120155608244e-05, + "loss": 0.9695, + "step": 24752 + }, + { + "epoch": 0.5802493011480512, + "grad_norm": 4.920602798461914, + "learning_rate": 7.508406764079585e-05, + "loss": 0.8666, + "step": 24753 + }, + { + "epoch": 0.5802727427228561, + "grad_norm": 2.670729637145996, + "learning_rate": 7.507693386070384e-05, + "loss": 0.2541, + "step": 24754 + }, + { + "epoch": 0.5802961842976612, + "grad_norm": 1.6754511594772339, + "learning_rate": 7.50698002158451e-05, + "loss": 0.1421, + "step": 24755 + }, + { + "epoch": 0.5803196258724661, + "grad_norm": 5.321581840515137, + "learning_rate": 7.506266670625828e-05, + "loss": 0.5453, + "step": 24756 + }, + { + "epoch": 0.5803430674472712, + "grad_norm": 2.4670016765594482, + "learning_rate": 7.505553333198211e-05, + "loss": 0.2036, + "step": 24757 + }, + { + "epoch": 0.5803665090220761, + "grad_norm": 2.5257415771484375, + "learning_rate": 7.504840009305537e-05, + "loss": 0.2748, + "step": 24758 + }, + { + "epoch": 0.5803899505968811, + "grad_norm": 2.5287282466888428, + "learning_rate": 7.50412669895167e-05, + "loss": 0.2967, + "step": 24759 + }, + { + "epoch": 0.5804133921716861, + "grad_norm": 3.1420395374298096, + "learning_rate": 7.503413402140483e-05, + "loss": 0.2528, + "step": 24760 + }, + { + "epoch": 0.5804368337464911, + "grad_norm": 2.50582218170166, + "learning_rate": 7.502700118875844e-05, + "loss": 1.522, + "step": 24761 + }, + { + "epoch": 0.5804602753212961, + "grad_norm": 2.1373002529144287, + "learning_rate": 7.501986849161625e-05, + "loss": 0.1186, + "step": 24762 + }, + { + "epoch": 0.5804837168961011, + "grad_norm": 3.283827066421509, + "learning_rate": 7.501273593001694e-05, + "loss": 0.1728, + "step": 24763 + }, + { + "epoch": 0.5805071584709061, + "grad_norm": 3.314497232437134, + "learning_rate": 7.500560350399925e-05, + "loss": 0.2288, + "step": 24764 + }, + { + "epoch": 0.5805306000457111, + "grad_norm": 4.704927921295166, + "learning_rate": 7.499847121360185e-05, + "loss": 0.3026, + "step": 24765 + }, + { + "epoch": 0.580554041620516, + "grad_norm": 3.3557872772216797, + "learning_rate": 7.499133905886342e-05, + "loss": 0.215, + "step": 24766 + }, + { + "epoch": 0.5805774831953211, + "grad_norm": 4.320333480834961, + "learning_rate": 7.498420703982274e-05, + "loss": 0.5199, + "step": 24767 + }, + { + "epoch": 0.580600924770126, + "grad_norm": 3.63104248046875, + "learning_rate": 7.497707515651843e-05, + "loss": 1.9056, + "step": 24768 + }, + { + "epoch": 0.5806243663449311, + "grad_norm": 2.245551586151123, + "learning_rate": 7.49699434089892e-05, + "loss": 0.27, + "step": 24769 + }, + { + "epoch": 0.580647807919736, + "grad_norm": 3.711366653442383, + "learning_rate": 7.496281179727374e-05, + "loss": 0.2796, + "step": 24770 + }, + { + "epoch": 0.5806712494945411, + "grad_norm": 2.6912965774536133, + "learning_rate": 7.495568032141081e-05, + "loss": 0.2509, + "step": 24771 + }, + { + "epoch": 0.580694691069346, + "grad_norm": 3.2036166191101074, + "learning_rate": 7.494854898143906e-05, + "loss": 0.2487, + "step": 24772 + }, + { + "epoch": 0.5807181326441511, + "grad_norm": 1.9774798154830933, + "learning_rate": 7.494141777739716e-05, + "loss": 0.1802, + "step": 24773 + }, + { + "epoch": 0.580741574218956, + "grad_norm": 2.503694534301758, + "learning_rate": 7.493428670932387e-05, + "loss": 1.4761, + "step": 24774 + }, + { + "epoch": 0.580765015793761, + "grad_norm": 0.5947322249412537, + "learning_rate": 7.492715577725783e-05, + "loss": 0.0466, + "step": 24775 + }, + { + "epoch": 0.580788457368566, + "grad_norm": 2.5152387619018555, + "learning_rate": 7.492002498123774e-05, + "loss": 1.961, + "step": 24776 + }, + { + "epoch": 0.580811898943371, + "grad_norm": 3.1813435554504395, + "learning_rate": 7.49128943213023e-05, + "loss": 0.2204, + "step": 24777 + }, + { + "epoch": 0.580835340518176, + "grad_norm": 3.3602657318115234, + "learning_rate": 7.490576379749021e-05, + "loss": 0.1938, + "step": 24778 + }, + { + "epoch": 0.580858782092981, + "grad_norm": 2.053088665008545, + "learning_rate": 7.489863340984014e-05, + "loss": 0.2089, + "step": 24779 + }, + { + "epoch": 0.580882223667786, + "grad_norm": 2.0276267528533936, + "learning_rate": 7.48915031583908e-05, + "loss": 0.2787, + "step": 24780 + }, + { + "epoch": 0.580905665242591, + "grad_norm": 3.8482866287231445, + "learning_rate": 7.488437304318086e-05, + "loss": 0.4528, + "step": 24781 + }, + { + "epoch": 0.580929106817396, + "grad_norm": 0.6661332845687866, + "learning_rate": 7.487724306424902e-05, + "loss": 0.0616, + "step": 24782 + }, + { + "epoch": 0.580952548392201, + "grad_norm": 0.7844173312187195, + "learning_rate": 7.487011322163395e-05, + "loss": 0.066, + "step": 24783 + }, + { + "epoch": 0.580975989967006, + "grad_norm": 2.851659059524536, + "learning_rate": 7.486298351537437e-05, + "loss": 0.3373, + "step": 24784 + }, + { + "epoch": 0.580999431541811, + "grad_norm": 2.32379150390625, + "learning_rate": 7.485585394550896e-05, + "loss": 1.7034, + "step": 24785 + }, + { + "epoch": 0.5810228731166159, + "grad_norm": 0.47342801094055176, + "learning_rate": 7.484872451207639e-05, + "loss": 0.0386, + "step": 24786 + }, + { + "epoch": 0.581046314691421, + "grad_norm": 3.136134386062622, + "learning_rate": 7.484159521511535e-05, + "loss": 0.279, + "step": 24787 + }, + { + "epoch": 0.5810697562662259, + "grad_norm": 2.005375385284424, + "learning_rate": 7.483446605466449e-05, + "loss": 0.1007, + "step": 24788 + }, + { + "epoch": 0.581093197841031, + "grad_norm": 1.3504459857940674, + "learning_rate": 7.482733703076257e-05, + "loss": 0.1604, + "step": 24789 + }, + { + "epoch": 0.5811166394158359, + "grad_norm": 2.817706346511841, + "learning_rate": 7.482020814344822e-05, + "loss": 0.2263, + "step": 24790 + }, + { + "epoch": 0.581140080990641, + "grad_norm": 3.581315279006958, + "learning_rate": 7.481307939276008e-05, + "loss": 0.2354, + "step": 24791 + }, + { + "epoch": 0.5811635225654459, + "grad_norm": 2.8884637355804443, + "learning_rate": 7.480595077873694e-05, + "loss": 0.1671, + "step": 24792 + }, + { + "epoch": 0.581186964140251, + "grad_norm": 3.0868496894836426, + "learning_rate": 7.47988223014174e-05, + "loss": 0.3256, + "step": 24793 + }, + { + "epoch": 0.5812104057150559, + "grad_norm": 3.31654953956604, + "learning_rate": 7.479169396084011e-05, + "loss": 0.4046, + "step": 24794 + }, + { + "epoch": 0.5812338472898609, + "grad_norm": 2.028430700302124, + "learning_rate": 7.478456575704383e-05, + "loss": 1.1272, + "step": 24795 + }, + { + "epoch": 0.5812572888646659, + "grad_norm": 2.3670568466186523, + "learning_rate": 7.477743769006722e-05, + "loss": 1.9825, + "step": 24796 + }, + { + "epoch": 0.5812807304394709, + "grad_norm": 0.7045678496360779, + "learning_rate": 7.477030975994894e-05, + "loss": 0.0699, + "step": 24797 + }, + { + "epoch": 0.5813041720142759, + "grad_norm": 0.9725903272628784, + "learning_rate": 7.476318196672767e-05, + "loss": 0.1084, + "step": 24798 + }, + { + "epoch": 0.5813276135890809, + "grad_norm": 2.1708321571350098, + "learning_rate": 7.475605431044208e-05, + "loss": 0.1721, + "step": 24799 + }, + { + "epoch": 0.5813510551638859, + "grad_norm": 4.005659103393555, + "learning_rate": 7.474892679113086e-05, + "loss": 0.2909, + "step": 24800 + }, + { + "epoch": 0.5813744967386909, + "grad_norm": 3.5614326000213623, + "learning_rate": 7.474179940883265e-05, + "loss": 0.2632, + "step": 24801 + }, + { + "epoch": 0.581397938313496, + "grad_norm": 2.5350399017333984, + "learning_rate": 7.473467216358616e-05, + "loss": 0.2433, + "step": 24802 + }, + { + "epoch": 0.5814213798883009, + "grad_norm": 0.9144695997238159, + "learning_rate": 7.472754505543008e-05, + "loss": 0.0786, + "step": 24803 + }, + { + "epoch": 0.5814448214631059, + "grad_norm": 2.834963083267212, + "learning_rate": 7.472041808440298e-05, + "loss": 0.3261, + "step": 24804 + }, + { + "epoch": 0.5814682630379109, + "grad_norm": 3.222719430923462, + "learning_rate": 7.471329125054366e-05, + "loss": 1.5793, + "step": 24805 + }, + { + "epoch": 0.5814917046127159, + "grad_norm": 1.0547442436218262, + "learning_rate": 7.470616455389068e-05, + "loss": 0.082, + "step": 24806 + }, + { + "epoch": 0.5815151461875209, + "grad_norm": 2.197648525238037, + "learning_rate": 7.469903799448282e-05, + "loss": 1.2049, + "step": 24807 + }, + { + "epoch": 0.5815385877623259, + "grad_norm": 0.7696723341941833, + "learning_rate": 7.469191157235865e-05, + "loss": 0.0403, + "step": 24808 + }, + { + "epoch": 0.5815620293371309, + "grad_norm": 2.553168773651123, + "learning_rate": 7.46847852875569e-05, + "loss": 0.1535, + "step": 24809 + }, + { + "epoch": 0.5815854709119359, + "grad_norm": 4.85386848449707, + "learning_rate": 7.467765914011622e-05, + "loss": 0.3636, + "step": 24810 + }, + { + "epoch": 0.5816089124867408, + "grad_norm": 1.2097868919372559, + "learning_rate": 7.467053313007526e-05, + "loss": 0.076, + "step": 24811 + }, + { + "epoch": 0.5816323540615459, + "grad_norm": 3.2722461223602295, + "learning_rate": 7.466340725747271e-05, + "loss": 1.127, + "step": 24812 + }, + { + "epoch": 0.5816557956363508, + "grad_norm": 1.9576610326766968, + "learning_rate": 7.465628152234722e-05, + "loss": 0.1857, + "step": 24813 + }, + { + "epoch": 0.5816792372111559, + "grad_norm": 2.24196195602417, + "learning_rate": 7.464915592473746e-05, + "loss": 1.5487, + "step": 24814 + }, + { + "epoch": 0.5817026787859608, + "grad_norm": 5.335123538970947, + "learning_rate": 7.46420304646821e-05, + "loss": 0.2961, + "step": 24815 + }, + { + "epoch": 0.5817261203607659, + "grad_norm": 0.6778619289398193, + "learning_rate": 7.46349051422198e-05, + "loss": 0.0364, + "step": 24816 + }, + { + "epoch": 0.5817495619355708, + "grad_norm": 3.0393998622894287, + "learning_rate": 7.462777995738919e-05, + "loss": 0.1491, + "step": 24817 + }, + { + "epoch": 0.5817730035103759, + "grad_norm": 0.9265560507774353, + "learning_rate": 7.462065491022894e-05, + "loss": 0.0551, + "step": 24818 + }, + { + "epoch": 0.5817964450851808, + "grad_norm": 3.445556163787842, + "learning_rate": 7.461353000077777e-05, + "loss": 0.242, + "step": 24819 + }, + { + "epoch": 0.5818198866599859, + "grad_norm": 1.719291090965271, + "learning_rate": 7.460640522907432e-05, + "loss": 0.2415, + "step": 24820 + }, + { + "epoch": 0.5818433282347908, + "grad_norm": 4.198101043701172, + "learning_rate": 7.45992805951572e-05, + "loss": 0.1733, + "step": 24821 + }, + { + "epoch": 0.5818667698095958, + "grad_norm": 2.6352181434631348, + "learning_rate": 7.45921560990651e-05, + "loss": 0.3249, + "step": 24822 + }, + { + "epoch": 0.5818902113844008, + "grad_norm": 3.1025593280792236, + "learning_rate": 7.458503174083666e-05, + "loss": 0.2706, + "step": 24823 + }, + { + "epoch": 0.5819136529592058, + "grad_norm": 1.8179912567138672, + "learning_rate": 7.457790752051059e-05, + "loss": 0.0776, + "step": 24824 + }, + { + "epoch": 0.5819370945340108, + "grad_norm": 2.307224750518799, + "learning_rate": 7.457078343812549e-05, + "loss": 0.1743, + "step": 24825 + }, + { + "epoch": 0.5819605361088158, + "grad_norm": 1.3606128692626953, + "learning_rate": 7.456365949372003e-05, + "loss": 0.1649, + "step": 24826 + }, + { + "epoch": 0.5819839776836208, + "grad_norm": 2.9207770824432373, + "learning_rate": 7.455653568733288e-05, + "loss": 0.3458, + "step": 24827 + }, + { + "epoch": 0.5820074192584258, + "grad_norm": 1.2125060558319092, + "learning_rate": 7.454941201900268e-05, + "loss": 0.0661, + "step": 24828 + }, + { + "epoch": 0.5820308608332307, + "grad_norm": 3.523674488067627, + "learning_rate": 7.454228848876808e-05, + "loss": 0.2886, + "step": 24829 + }, + { + "epoch": 0.5820543024080358, + "grad_norm": 3.0135562419891357, + "learning_rate": 7.453516509666769e-05, + "loss": 0.1697, + "step": 24830 + }, + { + "epoch": 0.5820777439828407, + "grad_norm": 0.7678976655006409, + "learning_rate": 7.452804184274027e-05, + "loss": 0.0862, + "step": 24831 + }, + { + "epoch": 0.5821011855576458, + "grad_norm": 1.2105157375335693, + "learning_rate": 7.452091872702441e-05, + "loss": 0.1289, + "step": 24832 + }, + { + "epoch": 0.5821246271324507, + "grad_norm": 4.488314628601074, + "learning_rate": 7.451379574955874e-05, + "loss": 0.2521, + "step": 24833 + }, + { + "epoch": 0.5821480687072558, + "grad_norm": 2.3370959758758545, + "learning_rate": 7.450667291038196e-05, + "loss": 0.1376, + "step": 24834 + }, + { + "epoch": 0.5821715102820607, + "grad_norm": 4.051397323608398, + "learning_rate": 7.449955020953267e-05, + "loss": 0.4208, + "step": 24835 + }, + { + "epoch": 0.5821949518568658, + "grad_norm": 3.1844873428344727, + "learning_rate": 7.449242764704953e-05, + "loss": 0.2661, + "step": 24836 + }, + { + "epoch": 0.5822183934316707, + "grad_norm": 0.9700027108192444, + "learning_rate": 7.448530522297121e-05, + "loss": 0.0981, + "step": 24837 + }, + { + "epoch": 0.5822418350064757, + "grad_norm": 3.2010891437530518, + "learning_rate": 7.447818293733635e-05, + "loss": 0.2723, + "step": 24838 + }, + { + "epoch": 0.5822652765812807, + "grad_norm": 3.6234002113342285, + "learning_rate": 7.447106079018357e-05, + "loss": 0.1779, + "step": 24839 + }, + { + "epoch": 0.5822887181560857, + "grad_norm": 5.045915126800537, + "learning_rate": 7.446393878155153e-05, + "loss": 0.4719, + "step": 24840 + }, + { + "epoch": 0.5823121597308907, + "grad_norm": 0.8866539597511292, + "learning_rate": 7.44568169114789e-05, + "loss": 0.0832, + "step": 24841 + }, + { + "epoch": 0.5823356013056957, + "grad_norm": 2.376392126083374, + "learning_rate": 7.444969518000424e-05, + "loss": 0.212, + "step": 24842 + }, + { + "epoch": 0.5823590428805007, + "grad_norm": 2.697864532470703, + "learning_rate": 7.444257358716629e-05, + "loss": 0.1858, + "step": 24843 + }, + { + "epoch": 0.5823824844553057, + "grad_norm": 6.3628740310668945, + "learning_rate": 7.443545213300367e-05, + "loss": 0.5954, + "step": 24844 + }, + { + "epoch": 0.5824059260301107, + "grad_norm": 3.833756923675537, + "learning_rate": 7.4428330817555e-05, + "loss": 0.5107, + "step": 24845 + }, + { + "epoch": 0.5824293676049157, + "grad_norm": 2.447075605392456, + "learning_rate": 7.44212096408589e-05, + "loss": 0.1625, + "step": 24846 + }, + { + "epoch": 0.5824528091797206, + "grad_norm": 0.2964465916156769, + "learning_rate": 7.441408860295407e-05, + "loss": 0.0154, + "step": 24847 + }, + { + "epoch": 0.5824762507545257, + "grad_norm": 3.731879711151123, + "learning_rate": 7.440696770387911e-05, + "loss": 0.3544, + "step": 24848 + }, + { + "epoch": 0.5824996923293306, + "grad_norm": 2.708430051803589, + "learning_rate": 7.439984694367264e-05, + "loss": 0.1577, + "step": 24849 + }, + { + "epoch": 0.5825231339041357, + "grad_norm": 4.3332319259643555, + "learning_rate": 7.439272632237334e-05, + "loss": 0.3659, + "step": 24850 + }, + { + "epoch": 0.5825465754789406, + "grad_norm": 3.8418161869049072, + "learning_rate": 7.438560584001983e-05, + "loss": 0.3986, + "step": 24851 + }, + { + "epoch": 0.5825700170537457, + "grad_norm": 4.0155768394470215, + "learning_rate": 7.437848549665072e-05, + "loss": 0.4416, + "step": 24852 + }, + { + "epoch": 0.5825934586285507, + "grad_norm": 2.762714385986328, + "learning_rate": 7.43713652923047e-05, + "loss": 0.2006, + "step": 24853 + }, + { + "epoch": 0.5826169002033557, + "grad_norm": 1.114754557609558, + "learning_rate": 7.436424522702035e-05, + "loss": 0.0891, + "step": 24854 + }, + { + "epoch": 0.5826403417781607, + "grad_norm": 2.937084197998047, + "learning_rate": 7.43571253008363e-05, + "loss": 0.1868, + "step": 24855 + }, + { + "epoch": 0.5826637833529656, + "grad_norm": 2.6077022552490234, + "learning_rate": 7.435000551379124e-05, + "loss": 0.328, + "step": 24856 + }, + { + "epoch": 0.5826872249277707, + "grad_norm": 2.0190482139587402, + "learning_rate": 7.434288586592378e-05, + "loss": 0.1232, + "step": 24857 + }, + { + "epoch": 0.5827106665025756, + "grad_norm": 4.177587509155273, + "learning_rate": 7.433576635727254e-05, + "loss": 0.3984, + "step": 24858 + }, + { + "epoch": 0.5827341080773807, + "grad_norm": 0.4205853343009949, + "learning_rate": 7.432864698787611e-05, + "loss": 0.0377, + "step": 24859 + }, + { + "epoch": 0.5827575496521856, + "grad_norm": 0.6088464856147766, + "learning_rate": 7.432152775777323e-05, + "loss": 0.0607, + "step": 24860 + }, + { + "epoch": 0.5827809912269907, + "grad_norm": 2.4596853256225586, + "learning_rate": 7.431440866700241e-05, + "loss": 1.3592, + "step": 24861 + }, + { + "epoch": 0.5828044328017956, + "grad_norm": 3.7170932292938232, + "learning_rate": 7.430728971560237e-05, + "loss": 0.5365, + "step": 24862 + }, + { + "epoch": 0.5828278743766007, + "grad_norm": 3.116126298904419, + "learning_rate": 7.430017090361167e-05, + "loss": 0.3871, + "step": 24863 + }, + { + "epoch": 0.5828513159514056, + "grad_norm": 2.6685242652893066, + "learning_rate": 7.429305223106897e-05, + "loss": 0.3903, + "step": 24864 + }, + { + "epoch": 0.5828747575262107, + "grad_norm": 1.2652702331542969, + "learning_rate": 7.42859336980129e-05, + "loss": 0.0988, + "step": 24865 + }, + { + "epoch": 0.5828981991010156, + "grad_norm": 5.281042575836182, + "learning_rate": 7.42788153044821e-05, + "loss": 0.6032, + "step": 24866 + }, + { + "epoch": 0.5829216406758206, + "grad_norm": 1.8359330892562866, + "learning_rate": 7.427169705051511e-05, + "loss": 1.4582, + "step": 24867 + }, + { + "epoch": 0.5829450822506256, + "grad_norm": 3.764413356781006, + "learning_rate": 7.426457893615064e-05, + "loss": 1.2707, + "step": 24868 + }, + { + "epoch": 0.5829685238254306, + "grad_norm": 2.5350630283355713, + "learning_rate": 7.42574609614273e-05, + "loss": 1.0208, + "step": 24869 + }, + { + "epoch": 0.5829919654002356, + "grad_norm": 2.8092966079711914, + "learning_rate": 7.425034312638371e-05, + "loss": 1.7526, + "step": 24870 + }, + { + "epoch": 0.5830154069750406, + "grad_norm": 3.4172329902648926, + "learning_rate": 7.424322543105846e-05, + "loss": 0.3844, + "step": 24871 + }, + { + "epoch": 0.5830388485498456, + "grad_norm": 2.268465042114258, + "learning_rate": 7.423610787549022e-05, + "loss": 0.1669, + "step": 24872 + }, + { + "epoch": 0.5830622901246506, + "grad_norm": 2.7453396320343018, + "learning_rate": 7.422899045971759e-05, + "loss": 0.2501, + "step": 24873 + }, + { + "epoch": 0.5830857316994555, + "grad_norm": 2.4920260906219482, + "learning_rate": 7.422187318377913e-05, + "loss": 0.1558, + "step": 24874 + }, + { + "epoch": 0.5831091732742606, + "grad_norm": 4.1361260414123535, + "learning_rate": 7.421475604771355e-05, + "loss": 0.2936, + "step": 24875 + }, + { + "epoch": 0.5831326148490655, + "grad_norm": 4.897640228271484, + "learning_rate": 7.420763905155944e-05, + "loss": 0.3967, + "step": 24876 + }, + { + "epoch": 0.5831560564238706, + "grad_norm": 2.6721644401550293, + "learning_rate": 7.420052219535537e-05, + "loss": 0.0955, + "step": 24877 + }, + { + "epoch": 0.5831794979986755, + "grad_norm": 1.3508120775222778, + "learning_rate": 7.419340547914002e-05, + "loss": 0.0836, + "step": 24878 + }, + { + "epoch": 0.5832029395734806, + "grad_norm": 1.0424878597259521, + "learning_rate": 7.418628890295193e-05, + "loss": 0.0863, + "step": 24879 + }, + { + "epoch": 0.5832263811482855, + "grad_norm": 3.0728750228881836, + "learning_rate": 7.417917246682981e-05, + "loss": 0.2339, + "step": 24880 + }, + { + "epoch": 0.5832498227230906, + "grad_norm": 0.4759027659893036, + "learning_rate": 7.41720561708122e-05, + "loss": 0.0382, + "step": 24881 + }, + { + "epoch": 0.5832732642978955, + "grad_norm": 3.9797732830047607, + "learning_rate": 7.416494001493777e-05, + "loss": 0.2783, + "step": 24882 + }, + { + "epoch": 0.5832967058727005, + "grad_norm": 3.638744592666626, + "learning_rate": 7.415782399924509e-05, + "loss": 0.4348, + "step": 24883 + }, + { + "epoch": 0.5833201474475055, + "grad_norm": 2.7849621772766113, + "learning_rate": 7.415070812377277e-05, + "loss": 0.2315, + "step": 24884 + }, + { + "epoch": 0.5833435890223105, + "grad_norm": 1.2695696353912354, + "learning_rate": 7.414359238855944e-05, + "loss": 0.0675, + "step": 24885 + }, + { + "epoch": 0.5833670305971155, + "grad_norm": 2.2683956623077393, + "learning_rate": 7.413647679364373e-05, + "loss": 0.1254, + "step": 24886 + }, + { + "epoch": 0.5833904721719205, + "grad_norm": 2.3419454097747803, + "learning_rate": 7.412936133906419e-05, + "loss": 0.1929, + "step": 24887 + }, + { + "epoch": 0.5834139137467255, + "grad_norm": 1.717881441116333, + "learning_rate": 7.412224602485948e-05, + "loss": 0.1396, + "step": 24888 + }, + { + "epoch": 0.5834373553215305, + "grad_norm": 7.618199348449707, + "learning_rate": 7.41151308510682e-05, + "loss": 0.5681, + "step": 24889 + }, + { + "epoch": 0.5834607968963355, + "grad_norm": 2.4832050800323486, + "learning_rate": 7.41080158177289e-05, + "loss": 0.195, + "step": 24890 + }, + { + "epoch": 0.5834842384711405, + "grad_norm": 1.186407446861267, + "learning_rate": 7.410090092488024e-05, + "loss": 0.0535, + "step": 24891 + }, + { + "epoch": 0.5835076800459454, + "grad_norm": 1.3891459703445435, + "learning_rate": 7.409378617256086e-05, + "loss": 0.0952, + "step": 24892 + }, + { + "epoch": 0.5835311216207505, + "grad_norm": 5.156900405883789, + "learning_rate": 7.408667156080933e-05, + "loss": 0.2795, + "step": 24893 + }, + { + "epoch": 0.5835545631955554, + "grad_norm": 3.1372952461242676, + "learning_rate": 7.407955708966421e-05, + "loss": 0.1885, + "step": 24894 + }, + { + "epoch": 0.5835780047703605, + "grad_norm": 0.6812229156494141, + "learning_rate": 7.407244275916418e-05, + "loss": 0.0758, + "step": 24895 + }, + { + "epoch": 0.5836014463451654, + "grad_norm": 2.2431557178497314, + "learning_rate": 7.406532856934779e-05, + "loss": 0.1761, + "step": 24896 + }, + { + "epoch": 0.5836248879199705, + "grad_norm": 0.6498695015907288, + "learning_rate": 7.405821452025364e-05, + "loss": 0.0604, + "step": 24897 + }, + { + "epoch": 0.5836483294947754, + "grad_norm": 1.519058346748352, + "learning_rate": 7.405110061192038e-05, + "loss": 0.1461, + "step": 24898 + }, + { + "epoch": 0.5836717710695805, + "grad_norm": 4.052705764770508, + "learning_rate": 7.404398684438653e-05, + "loss": 0.2825, + "step": 24899 + }, + { + "epoch": 0.5836952126443854, + "grad_norm": 0.2754136919975281, + "learning_rate": 7.403687321769079e-05, + "loss": 0.0132, + "step": 24900 + }, + { + "epoch": 0.5837186542191904, + "grad_norm": 2.5381157398223877, + "learning_rate": 7.40297597318717e-05, + "loss": 0.3053, + "step": 24901 + }, + { + "epoch": 0.5837420957939954, + "grad_norm": 3.6329691410064697, + "learning_rate": 7.402264638696783e-05, + "loss": 0.2933, + "step": 24902 + }, + { + "epoch": 0.5837655373688004, + "grad_norm": 2.331148386001587, + "learning_rate": 7.40155331830178e-05, + "loss": 0.1403, + "step": 24903 + }, + { + "epoch": 0.5837889789436055, + "grad_norm": 2.7035999298095703, + "learning_rate": 7.400842012006024e-05, + "loss": 0.1596, + "step": 24904 + }, + { + "epoch": 0.5838124205184104, + "grad_norm": 2.222442865371704, + "learning_rate": 7.400130719813374e-05, + "loss": 0.1683, + "step": 24905 + }, + { + "epoch": 0.5838358620932155, + "grad_norm": 1.8327648639678955, + "learning_rate": 7.399419441727686e-05, + "loss": 0.1085, + "step": 24906 + }, + { + "epoch": 0.5838593036680204, + "grad_norm": 2.8643195629119873, + "learning_rate": 7.398708177752823e-05, + "loss": 0.1427, + "step": 24907 + }, + { + "epoch": 0.5838827452428255, + "grad_norm": 3.268693208694458, + "learning_rate": 7.397996927892641e-05, + "loss": 0.228, + "step": 24908 + }, + { + "epoch": 0.5839061868176304, + "grad_norm": 4.726062297821045, + "learning_rate": 7.397285692151e-05, + "loss": 0.5608, + "step": 24909 + }, + { + "epoch": 0.5839296283924355, + "grad_norm": 1.5076075792312622, + "learning_rate": 7.396574470531764e-05, + "loss": 1.1081, + "step": 24910 + }, + { + "epoch": 0.5839530699672404, + "grad_norm": 3.843031644821167, + "learning_rate": 7.395863263038785e-05, + "loss": 0.3048, + "step": 24911 + }, + { + "epoch": 0.5839765115420454, + "grad_norm": 3.51110577583313, + "learning_rate": 7.395152069675924e-05, + "loss": 0.2909, + "step": 24912 + }, + { + "epoch": 0.5839999531168504, + "grad_norm": 2.955190420150757, + "learning_rate": 7.394440890447043e-05, + "loss": 0.3052, + "step": 24913 + }, + { + "epoch": 0.5840233946916554, + "grad_norm": 1.0699901580810547, + "learning_rate": 7.393729725356e-05, + "loss": 0.0929, + "step": 24914 + }, + { + "epoch": 0.5840468362664604, + "grad_norm": 2.773386001586914, + "learning_rate": 7.393018574406648e-05, + "loss": 0.2155, + "step": 24915 + }, + { + "epoch": 0.5840702778412654, + "grad_norm": 0.7684455513954163, + "learning_rate": 7.392307437602852e-05, + "loss": 0.0582, + "step": 24916 + }, + { + "epoch": 0.5840937194160704, + "grad_norm": 3.6825149059295654, + "learning_rate": 7.391596314948473e-05, + "loss": 0.3551, + "step": 24917 + }, + { + "epoch": 0.5841171609908754, + "grad_norm": 0.605870246887207, + "learning_rate": 7.390885206447365e-05, + "loss": 0.0464, + "step": 24918 + }, + { + "epoch": 0.5841406025656803, + "grad_norm": 2.884813070297241, + "learning_rate": 7.390174112103386e-05, + "loss": 0.2171, + "step": 24919 + }, + { + "epoch": 0.5841640441404854, + "grad_norm": 3.36761474609375, + "learning_rate": 7.389463031920396e-05, + "loss": 0.3071, + "step": 24920 + }, + { + "epoch": 0.5841874857152903, + "grad_norm": 0.7070145010948181, + "learning_rate": 7.388751965902252e-05, + "loss": 0.0927, + "step": 24921 + }, + { + "epoch": 0.5842109272900954, + "grad_norm": 3.7194366455078125, + "learning_rate": 7.388040914052814e-05, + "loss": 0.2061, + "step": 24922 + }, + { + "epoch": 0.5842343688649003, + "grad_norm": 0.5108891129493713, + "learning_rate": 7.387329876375941e-05, + "loss": 0.0426, + "step": 24923 + }, + { + "epoch": 0.5842578104397054, + "grad_norm": 2.7474477291107178, + "learning_rate": 7.386618852875489e-05, + "loss": 0.2614, + "step": 24924 + }, + { + "epoch": 0.5842812520145103, + "grad_norm": 3.03277850151062, + "learning_rate": 7.385907843555314e-05, + "loss": 0.1891, + "step": 24925 + }, + { + "epoch": 0.5843046935893154, + "grad_norm": 2.2130126953125, + "learning_rate": 7.385196848419278e-05, + "loss": 0.2141, + "step": 24926 + }, + { + "epoch": 0.5843281351641203, + "grad_norm": 2.9883315563201904, + "learning_rate": 7.38448586747124e-05, + "loss": 0.1868, + "step": 24927 + }, + { + "epoch": 0.5843515767389253, + "grad_norm": 2.951505422592163, + "learning_rate": 7.383774900715048e-05, + "loss": 1.5319, + "step": 24928 + }, + { + "epoch": 0.5843750183137303, + "grad_norm": 12.074477195739746, + "learning_rate": 7.383063948154572e-05, + "loss": 0.3339, + "step": 24929 + }, + { + "epoch": 0.5843984598885353, + "grad_norm": 3.944498300552368, + "learning_rate": 7.382353009793664e-05, + "loss": 0.301, + "step": 24930 + }, + { + "epoch": 0.5844219014633403, + "grad_norm": 0.4346645176410675, + "learning_rate": 7.381642085636183e-05, + "loss": 0.0363, + "step": 24931 + }, + { + "epoch": 0.5844453430381453, + "grad_norm": 1.8242173194885254, + "learning_rate": 7.380931175685984e-05, + "loss": 0.1253, + "step": 24932 + }, + { + "epoch": 0.5844687846129503, + "grad_norm": 2.9098446369171143, + "learning_rate": 7.380220279946928e-05, + "loss": 0.163, + "step": 24933 + }, + { + "epoch": 0.5844922261877553, + "grad_norm": 4.921793460845947, + "learning_rate": 7.379509398422872e-05, + "loss": 0.3098, + "step": 24934 + }, + { + "epoch": 0.5845156677625603, + "grad_norm": 2.645519733428955, + "learning_rate": 7.378798531117667e-05, + "loss": 0.1847, + "step": 24935 + }, + { + "epoch": 0.5845391093373653, + "grad_norm": 1.272096872329712, + "learning_rate": 7.378087678035179e-05, + "loss": 0.1341, + "step": 24936 + }, + { + "epoch": 0.5845625509121702, + "grad_norm": 12.880807876586914, + "learning_rate": 7.377376839179258e-05, + "loss": 0.6304, + "step": 24937 + }, + { + "epoch": 0.5845859924869753, + "grad_norm": 3.6012492179870605, + "learning_rate": 7.376666014553765e-05, + "loss": 0.485, + "step": 24938 + }, + { + "epoch": 0.5846094340617802, + "grad_norm": 1.9011080265045166, + "learning_rate": 7.375955204162558e-05, + "loss": 0.3237, + "step": 24939 + }, + { + "epoch": 0.5846328756365853, + "grad_norm": 3.2590010166168213, + "learning_rate": 7.375244408009486e-05, + "loss": 0.4604, + "step": 24940 + }, + { + "epoch": 0.5846563172113902, + "grad_norm": 2.2419588565826416, + "learning_rate": 7.374533626098414e-05, + "loss": 0.1646, + "step": 24941 + }, + { + "epoch": 0.5846797587861953, + "grad_norm": 2.823276996612549, + "learning_rate": 7.3738228584332e-05, + "loss": 0.2143, + "step": 24942 + }, + { + "epoch": 0.5847032003610002, + "grad_norm": 5.188289165496826, + "learning_rate": 7.373112105017695e-05, + "loss": 0.3562, + "step": 24943 + }, + { + "epoch": 0.5847266419358053, + "grad_norm": 0.8095693588256836, + "learning_rate": 7.372401365855756e-05, + "loss": 0.0741, + "step": 24944 + }, + { + "epoch": 0.5847500835106102, + "grad_norm": 2.161935329437256, + "learning_rate": 7.371690640951245e-05, + "loss": 0.1109, + "step": 24945 + }, + { + "epoch": 0.5847735250854152, + "grad_norm": 2.8937950134277344, + "learning_rate": 7.370979930308013e-05, + "loss": 0.2599, + "step": 24946 + }, + { + "epoch": 0.5847969666602202, + "grad_norm": 2.85117506980896, + "learning_rate": 7.370269233929917e-05, + "loss": 1.3444, + "step": 24947 + }, + { + "epoch": 0.5848204082350252, + "grad_norm": 6.866672039031982, + "learning_rate": 7.369558551820814e-05, + "loss": 0.4258, + "step": 24948 + }, + { + "epoch": 0.5848438498098302, + "grad_norm": 1.9602267742156982, + "learning_rate": 7.368847883984561e-05, + "loss": 0.1415, + "step": 24949 + }, + { + "epoch": 0.5848672913846352, + "grad_norm": 3.4830574989318848, + "learning_rate": 7.368137230425012e-05, + "loss": 0.2228, + "step": 24950 + }, + { + "epoch": 0.5848907329594402, + "grad_norm": 2.210141658782959, + "learning_rate": 7.367426591146025e-05, + "loss": 0.1609, + "step": 24951 + }, + { + "epoch": 0.5849141745342452, + "grad_norm": 3.6139519214630127, + "learning_rate": 7.366715966151452e-05, + "loss": 1.582, + "step": 24952 + }, + { + "epoch": 0.5849376161090502, + "grad_norm": 1.4456455707550049, + "learning_rate": 7.366005355445157e-05, + "loss": 0.1032, + "step": 24953 + }, + { + "epoch": 0.5849610576838552, + "grad_norm": 2.0590157508850098, + "learning_rate": 7.365294759030988e-05, + "loss": 0.1127, + "step": 24954 + }, + { + "epoch": 0.5849844992586603, + "grad_norm": 2.4583253860473633, + "learning_rate": 7.364584176912807e-05, + "loss": 1.4915, + "step": 24955 + }, + { + "epoch": 0.5850079408334652, + "grad_norm": 2.7554314136505127, + "learning_rate": 7.363873609094464e-05, + "loss": 0.1689, + "step": 24956 + }, + { + "epoch": 0.5850313824082702, + "grad_norm": 3.205125331878662, + "learning_rate": 7.363163055579818e-05, + "loss": 0.3372, + "step": 24957 + }, + { + "epoch": 0.5850548239830752, + "grad_norm": 2.7818660736083984, + "learning_rate": 7.362452516372724e-05, + "loss": 0.2563, + "step": 24958 + }, + { + "epoch": 0.5850782655578802, + "grad_norm": 2.991577625274658, + "learning_rate": 7.361741991477036e-05, + "loss": 1.2081, + "step": 24959 + }, + { + "epoch": 0.5851017071326852, + "grad_norm": 1.9035698175430298, + "learning_rate": 7.36103148089661e-05, + "loss": 0.1563, + "step": 24960 + }, + { + "epoch": 0.5851251487074902, + "grad_norm": 2.296555519104004, + "learning_rate": 7.3603209846353e-05, + "loss": 0.1463, + "step": 24961 + }, + { + "epoch": 0.5851485902822952, + "grad_norm": 1.0320653915405273, + "learning_rate": 7.359610502696967e-05, + "loss": 0.1265, + "step": 24962 + }, + { + "epoch": 0.5851720318571002, + "grad_norm": 2.8943915367126465, + "learning_rate": 7.358900035085456e-05, + "loss": 0.4159, + "step": 24963 + }, + { + "epoch": 0.5851954734319051, + "grad_norm": 2.8678483963012695, + "learning_rate": 7.358189581804627e-05, + "loss": 0.3097, + "step": 24964 + }, + { + "epoch": 0.5852189150067102, + "grad_norm": 2.708220958709717, + "learning_rate": 7.35747914285834e-05, + "loss": 0.624, + "step": 24965 + }, + { + "epoch": 0.5852423565815151, + "grad_norm": 2.233029365539551, + "learning_rate": 7.356768718250446e-05, + "loss": 0.1138, + "step": 24966 + }, + { + "epoch": 0.5852657981563202, + "grad_norm": 2.461811065673828, + "learning_rate": 7.356058307984796e-05, + "loss": 0.2004, + "step": 24967 + }, + { + "epoch": 0.5852892397311251, + "grad_norm": 3.1796298027038574, + "learning_rate": 7.35534791206525e-05, + "loss": 0.2947, + "step": 24968 + }, + { + "epoch": 0.5853126813059302, + "grad_norm": 3.047173023223877, + "learning_rate": 7.35463753049566e-05, + "loss": 0.216, + "step": 24969 + }, + { + "epoch": 0.5853361228807351, + "grad_norm": 1.1634769439697266, + "learning_rate": 7.353927163279882e-05, + "loss": 0.2108, + "step": 24970 + }, + { + "epoch": 0.5853595644555402, + "grad_norm": 2.311901569366455, + "learning_rate": 7.35321681042177e-05, + "loss": 0.2017, + "step": 24971 + }, + { + "epoch": 0.5853830060303451, + "grad_norm": 2.7030317783355713, + "learning_rate": 7.352506471925177e-05, + "loss": 0.1387, + "step": 24972 + }, + { + "epoch": 0.5854064476051501, + "grad_norm": 1.5386712551116943, + "learning_rate": 7.351796147793956e-05, + "loss": 0.1317, + "step": 24973 + }, + { + "epoch": 0.5854298891799551, + "grad_norm": 1.458483099937439, + "learning_rate": 7.351085838031968e-05, + "loss": 0.0897, + "step": 24974 + }, + { + "epoch": 0.5854533307547601, + "grad_norm": 4.2064385414123535, + "learning_rate": 7.350375542643059e-05, + "loss": 0.4208, + "step": 24975 + }, + { + "epoch": 0.5854767723295651, + "grad_norm": 3.3104538917541504, + "learning_rate": 7.349665261631086e-05, + "loss": 0.3098, + "step": 24976 + }, + { + "epoch": 0.5855002139043701, + "grad_norm": 2.5113043785095215, + "learning_rate": 7.348954994999906e-05, + "loss": 0.1654, + "step": 24977 + }, + { + "epoch": 0.5855236554791751, + "grad_norm": 2.4500560760498047, + "learning_rate": 7.348244742753372e-05, + "loss": 0.1033, + "step": 24978 + }, + { + "epoch": 0.5855470970539801, + "grad_norm": 0.7456445693969727, + "learning_rate": 7.347534504895334e-05, + "loss": 0.0549, + "step": 24979 + }, + { + "epoch": 0.585570538628785, + "grad_norm": 3.1758782863616943, + "learning_rate": 7.346824281429651e-05, + "loss": 0.3351, + "step": 24980 + }, + { + "epoch": 0.5855939802035901, + "grad_norm": 1.633766531944275, + "learning_rate": 7.346114072360172e-05, + "loss": 0.1738, + "step": 24981 + }, + { + "epoch": 0.585617421778395, + "grad_norm": 5.044683456420898, + "learning_rate": 7.345403877690753e-05, + "loss": 0.418, + "step": 24982 + }, + { + "epoch": 0.5856408633532001, + "grad_norm": 1.1308372020721436, + "learning_rate": 7.344693697425249e-05, + "loss": 0.094, + "step": 24983 + }, + { + "epoch": 0.585664304928005, + "grad_norm": 0.6029471755027771, + "learning_rate": 7.343983531567511e-05, + "loss": 0.0227, + "step": 24984 + }, + { + "epoch": 0.5856877465028101, + "grad_norm": 3.282525062561035, + "learning_rate": 7.343273380121391e-05, + "loss": 0.3946, + "step": 24985 + }, + { + "epoch": 0.585711188077615, + "grad_norm": 4.8340678215026855, + "learning_rate": 7.342563243090747e-05, + "loss": 0.4912, + "step": 24986 + }, + { + "epoch": 0.5857346296524201, + "grad_norm": 3.7959165573120117, + "learning_rate": 7.34185312047943e-05, + "loss": 1.6468, + "step": 24987 + }, + { + "epoch": 0.585758071227225, + "grad_norm": 3.8066749572753906, + "learning_rate": 7.341143012291288e-05, + "loss": 0.2921, + "step": 24988 + }, + { + "epoch": 0.5857815128020301, + "grad_norm": 3.318127393722534, + "learning_rate": 7.34043291853018e-05, + "loss": 0.312, + "step": 24989 + }, + { + "epoch": 0.585804954376835, + "grad_norm": 3.4894819259643555, + "learning_rate": 7.339722839199961e-05, + "loss": 0.4072, + "step": 24990 + }, + { + "epoch": 0.58582839595164, + "grad_norm": 1.2293416261672974, + "learning_rate": 7.33901277430448e-05, + "loss": 0.0909, + "step": 24991 + }, + { + "epoch": 0.585851837526445, + "grad_norm": 2.5120651721954346, + "learning_rate": 7.338302723847588e-05, + "loss": 0.1578, + "step": 24992 + }, + { + "epoch": 0.58587527910125, + "grad_norm": 0.8827155232429504, + "learning_rate": 7.337592687833143e-05, + "loss": 0.0842, + "step": 24993 + }, + { + "epoch": 0.585898720676055, + "grad_norm": 2.4689829349517822, + "learning_rate": 7.336882666264995e-05, + "loss": 0.2462, + "step": 24994 + }, + { + "epoch": 0.58592216225086, + "grad_norm": 3.7655086517333984, + "learning_rate": 7.336172659146994e-05, + "loss": 0.3543, + "step": 24995 + }, + { + "epoch": 0.585945603825665, + "grad_norm": 3.018557071685791, + "learning_rate": 7.335462666482998e-05, + "loss": 0.189, + "step": 24996 + }, + { + "epoch": 0.58596904540047, + "grad_norm": 1.9427472352981567, + "learning_rate": 7.334752688276856e-05, + "loss": 1.3732, + "step": 24997 + }, + { + "epoch": 0.585992486975275, + "grad_norm": 2.7335264682769775, + "learning_rate": 7.334042724532419e-05, + "loss": 0.2929, + "step": 24998 + }, + { + "epoch": 0.58601592855008, + "grad_norm": 4.352538585662842, + "learning_rate": 7.333332775253542e-05, + "loss": 0.4159, + "step": 24999 + }, + { + "epoch": 0.5860393701248849, + "grad_norm": 2.4530298709869385, + "learning_rate": 7.332622840444077e-05, + "loss": 0.2559, + "step": 25000 + }, + { + "epoch": 0.58606281169969, + "grad_norm": 3.5619566440582275, + "learning_rate": 7.331912920107872e-05, + "loss": 0.325, + "step": 25001 + }, + { + "epoch": 0.5860862532744949, + "grad_norm": 0.426328182220459, + "learning_rate": 7.331203014248784e-05, + "loss": 0.0253, + "step": 25002 + }, + { + "epoch": 0.5861096948493, + "grad_norm": 2.2242352962493896, + "learning_rate": 7.330493122870666e-05, + "loss": 1.3563, + "step": 25003 + }, + { + "epoch": 0.5861331364241049, + "grad_norm": 0.8311837911605835, + "learning_rate": 7.329783245977367e-05, + "loss": 0.0778, + "step": 25004 + }, + { + "epoch": 0.58615657799891, + "grad_norm": 1.703224539756775, + "learning_rate": 7.329073383572737e-05, + "loss": 0.1518, + "step": 25005 + }, + { + "epoch": 0.586180019573715, + "grad_norm": 3.538341760635376, + "learning_rate": 7.328363535660631e-05, + "loss": 0.3032, + "step": 25006 + }, + { + "epoch": 0.58620346114852, + "grad_norm": 3.9171574115753174, + "learning_rate": 7.327653702244902e-05, + "loss": 0.4262, + "step": 25007 + }, + { + "epoch": 0.586226902723325, + "grad_norm": 3.18851900100708, + "learning_rate": 7.326943883329396e-05, + "loss": 0.2426, + "step": 25008 + }, + { + "epoch": 0.5862503442981299, + "grad_norm": 2.107588529586792, + "learning_rate": 7.326234078917969e-05, + "loss": 0.1602, + "step": 25009 + }, + { + "epoch": 0.586273785872935, + "grad_norm": 2.5686838626861572, + "learning_rate": 7.32552428901447e-05, + "loss": 0.2707, + "step": 25010 + }, + { + "epoch": 0.5862972274477399, + "grad_norm": 3.542881965637207, + "learning_rate": 7.324814513622753e-05, + "loss": 0.284, + "step": 25011 + }, + { + "epoch": 0.586320669022545, + "grad_norm": 4.873616695404053, + "learning_rate": 7.324104752746668e-05, + "loss": 0.5279, + "step": 25012 + }, + { + "epoch": 0.5863441105973499, + "grad_norm": 0.6217198371887207, + "learning_rate": 7.323395006390062e-05, + "loss": 0.0157, + "step": 25013 + }, + { + "epoch": 0.586367552172155, + "grad_norm": 2.01932430267334, + "learning_rate": 7.322685274556793e-05, + "loss": 1.4629, + "step": 25014 + }, + { + "epoch": 0.5863909937469599, + "grad_norm": 1.8811417818069458, + "learning_rate": 7.321975557250706e-05, + "loss": 0.1723, + "step": 25015 + }, + { + "epoch": 0.586414435321765, + "grad_norm": 1.3437045812606812, + "learning_rate": 7.32126585447566e-05, + "loss": 0.1314, + "step": 25016 + }, + { + "epoch": 0.5864378768965699, + "grad_norm": 0.5553520917892456, + "learning_rate": 7.320556166235497e-05, + "loss": 0.0482, + "step": 25017 + }, + { + "epoch": 0.586461318471375, + "grad_norm": 2.6419882774353027, + "learning_rate": 7.319846492534073e-05, + "loss": 0.2794, + "step": 25018 + }, + { + "epoch": 0.5864847600461799, + "grad_norm": 2.0867509841918945, + "learning_rate": 7.319136833375239e-05, + "loss": 0.1591, + "step": 25019 + }, + { + "epoch": 0.5865082016209849, + "grad_norm": 1.9438486099243164, + "learning_rate": 7.31842718876284e-05, + "loss": 0.2092, + "step": 25020 + }, + { + "epoch": 0.5865316431957899, + "grad_norm": 1.8465899229049683, + "learning_rate": 7.317717558700734e-05, + "loss": 0.1006, + "step": 25021 + }, + { + "epoch": 0.5865550847705949, + "grad_norm": 3.74579119682312, + "learning_rate": 7.317007943192768e-05, + "loss": 0.2606, + "step": 25022 + }, + { + "epoch": 0.5865785263453999, + "grad_norm": 3.1882777214050293, + "learning_rate": 7.316298342242788e-05, + "loss": 0.3648, + "step": 25023 + }, + { + "epoch": 0.5866019679202049, + "grad_norm": 2.367114305496216, + "learning_rate": 7.315588755854654e-05, + "loss": 0.3033, + "step": 25024 + }, + { + "epoch": 0.5866254094950099, + "grad_norm": 1.6105797290802002, + "learning_rate": 7.314879184032204e-05, + "loss": 0.7985, + "step": 25025 + }, + { + "epoch": 0.5866488510698149, + "grad_norm": 3.2411069869995117, + "learning_rate": 7.314169626779301e-05, + "loss": 0.4364, + "step": 25026 + }, + { + "epoch": 0.5866722926446198, + "grad_norm": 1.4489269256591797, + "learning_rate": 7.313460084099785e-05, + "loss": 0.0903, + "step": 25027 + }, + { + "epoch": 0.5866957342194249, + "grad_norm": 1.201610803604126, + "learning_rate": 7.312750555997516e-05, + "loss": 0.1774, + "step": 25028 + }, + { + "epoch": 0.5867191757942298, + "grad_norm": 0.9856187701225281, + "learning_rate": 7.312041042476334e-05, + "loss": 0.0882, + "step": 25029 + }, + { + "epoch": 0.5867426173690349, + "grad_norm": 3.5072832107543945, + "learning_rate": 7.311331543540093e-05, + "loss": 0.3177, + "step": 25030 + }, + { + "epoch": 0.5867660589438398, + "grad_norm": 4.190835475921631, + "learning_rate": 7.310622059192644e-05, + "loss": 0.1979, + "step": 25031 + }, + { + "epoch": 0.5867895005186449, + "grad_norm": 4.258406162261963, + "learning_rate": 7.309912589437835e-05, + "loss": 0.2564, + "step": 25032 + }, + { + "epoch": 0.5868129420934498, + "grad_norm": 2.4537112712860107, + "learning_rate": 7.309203134279515e-05, + "loss": 0.2073, + "step": 25033 + }, + { + "epoch": 0.5868363836682549, + "grad_norm": 3.320354700088501, + "learning_rate": 7.308493693721536e-05, + "loss": 0.3674, + "step": 25034 + }, + { + "epoch": 0.5868598252430598, + "grad_norm": 3.525841474533081, + "learning_rate": 7.307784267767745e-05, + "loss": 0.3756, + "step": 25035 + }, + { + "epoch": 0.5868832668178648, + "grad_norm": 3.846971035003662, + "learning_rate": 7.307074856421992e-05, + "loss": 0.36, + "step": 25036 + }, + { + "epoch": 0.5869067083926698, + "grad_norm": 4.0728349685668945, + "learning_rate": 7.306365459688123e-05, + "loss": 0.2927, + "step": 25037 + }, + { + "epoch": 0.5869301499674748, + "grad_norm": 3.4319865703582764, + "learning_rate": 7.305656077569995e-05, + "loss": 0.3643, + "step": 25038 + }, + { + "epoch": 0.5869535915422798, + "grad_norm": 3.368516683578491, + "learning_rate": 7.304946710071454e-05, + "loss": 0.457, + "step": 25039 + }, + { + "epoch": 0.5869770331170848, + "grad_norm": 1.3581129312515259, + "learning_rate": 7.304237357196344e-05, + "loss": 0.1401, + "step": 25040 + }, + { + "epoch": 0.5870004746918898, + "grad_norm": 0.570421576499939, + "learning_rate": 7.303528018948523e-05, + "loss": 0.0529, + "step": 25041 + }, + { + "epoch": 0.5870239162666948, + "grad_norm": 1.0250920057296753, + "learning_rate": 7.302818695331832e-05, + "loss": 0.0774, + "step": 25042 + }, + { + "epoch": 0.5870473578414998, + "grad_norm": 2.395188808441162, + "learning_rate": 7.30210938635012e-05, + "loss": 0.2288, + "step": 25043 + }, + { + "epoch": 0.5870707994163048, + "grad_norm": 2.0087671279907227, + "learning_rate": 7.301400092007242e-05, + "loss": 1.0022, + "step": 25044 + }, + { + "epoch": 0.5870942409911097, + "grad_norm": 4.071774482727051, + "learning_rate": 7.300690812307043e-05, + "loss": 0.4604, + "step": 25045 + }, + { + "epoch": 0.5871176825659148, + "grad_norm": 2.2685434818267822, + "learning_rate": 7.29998154725337e-05, + "loss": 1.9596, + "step": 25046 + }, + { + "epoch": 0.5871411241407197, + "grad_norm": 2.7552671432495117, + "learning_rate": 7.299272296850075e-05, + "loss": 0.1014, + "step": 25047 + }, + { + "epoch": 0.5871645657155248, + "grad_norm": 2.8017420768737793, + "learning_rate": 7.298563061101e-05, + "loss": 0.1747, + "step": 25048 + }, + { + "epoch": 0.5871880072903297, + "grad_norm": 3.5812034606933594, + "learning_rate": 7.297853840010001e-05, + "loss": 0.4115, + "step": 25049 + }, + { + "epoch": 0.5872114488651348, + "grad_norm": 0.5703878402709961, + "learning_rate": 7.297144633580922e-05, + "loss": 0.0397, + "step": 25050 + }, + { + "epoch": 0.5872348904399397, + "grad_norm": 5.7606964111328125, + "learning_rate": 7.296435441817612e-05, + "loss": 0.4006, + "step": 25051 + }, + { + "epoch": 0.5872583320147448, + "grad_norm": 1.6828984022140503, + "learning_rate": 7.295726264723922e-05, + "loss": 0.1502, + "step": 25052 + }, + { + "epoch": 0.5872817735895497, + "grad_norm": 2.815356492996216, + "learning_rate": 7.295017102303694e-05, + "loss": 0.2024, + "step": 25053 + }, + { + "epoch": 0.5873052151643547, + "grad_norm": 1.7138298749923706, + "learning_rate": 7.294307954560782e-05, + "loss": 0.1485, + "step": 25054 + }, + { + "epoch": 0.5873286567391597, + "grad_norm": 3.13633131980896, + "learning_rate": 7.293598821499029e-05, + "loss": 0.48, + "step": 25055 + }, + { + "epoch": 0.5873520983139647, + "grad_norm": 3.6263465881347656, + "learning_rate": 7.292889703122287e-05, + "loss": 0.3034, + "step": 25056 + }, + { + "epoch": 0.5873755398887698, + "grad_norm": 4.709010601043701, + "learning_rate": 7.292180599434401e-05, + "loss": 0.3782, + "step": 25057 + }, + { + "epoch": 0.5873989814635747, + "grad_norm": 3.6948153972625732, + "learning_rate": 7.291471510439218e-05, + "loss": 0.3593, + "step": 25058 + }, + { + "epoch": 0.5874224230383798, + "grad_norm": 2.0916104316711426, + "learning_rate": 7.290762436140588e-05, + "loss": 0.0856, + "step": 25059 + }, + { + "epoch": 0.5874458646131847, + "grad_norm": 0.36798202991485596, + "learning_rate": 7.290053376542357e-05, + "loss": 0.0205, + "step": 25060 + }, + { + "epoch": 0.5874693061879898, + "grad_norm": 1.8729902505874634, + "learning_rate": 7.289344331648368e-05, + "loss": 0.074, + "step": 25061 + }, + { + "epoch": 0.5874927477627947, + "grad_norm": 0.9417272210121155, + "learning_rate": 7.288635301462477e-05, + "loss": 0.0882, + "step": 25062 + }, + { + "epoch": 0.5875161893375997, + "grad_norm": 4.676126956939697, + "learning_rate": 7.287926285988528e-05, + "loss": 0.4704, + "step": 25063 + }, + { + "epoch": 0.5875396309124047, + "grad_norm": 3.840143918991089, + "learning_rate": 7.287217285230367e-05, + "loss": 0.2815, + "step": 25064 + }, + { + "epoch": 0.5875630724872097, + "grad_norm": 1.852920651435852, + "learning_rate": 7.286508299191839e-05, + "loss": 0.2144, + "step": 25065 + }, + { + "epoch": 0.5875865140620147, + "grad_norm": 0.6887340545654297, + "learning_rate": 7.285799327876796e-05, + "loss": 0.0335, + "step": 25066 + }, + { + "epoch": 0.5876099556368197, + "grad_norm": 1.060551404953003, + "learning_rate": 7.285090371289083e-05, + "loss": 0.0863, + "step": 25067 + }, + { + "epoch": 0.5876333972116247, + "grad_norm": 2.394813299179077, + "learning_rate": 7.284381429432542e-05, + "loss": 0.1442, + "step": 25068 + }, + { + "epoch": 0.5876568387864297, + "grad_norm": 1.0115405321121216, + "learning_rate": 7.283672502311027e-05, + "loss": 0.0925, + "step": 25069 + }, + { + "epoch": 0.5876802803612347, + "grad_norm": 3.0617294311523438, + "learning_rate": 7.28296358992838e-05, + "loss": 0.1181, + "step": 25070 + }, + { + "epoch": 0.5877037219360397, + "grad_norm": 2.9396374225616455, + "learning_rate": 7.282254692288448e-05, + "loss": 0.2494, + "step": 25071 + }, + { + "epoch": 0.5877271635108446, + "grad_norm": 2.889249801635742, + "learning_rate": 7.281545809395081e-05, + "loss": 0.3229, + "step": 25072 + }, + { + "epoch": 0.5877506050856497, + "grad_norm": 1.8373359441757202, + "learning_rate": 7.280836941252119e-05, + "loss": 0.191, + "step": 25073 + }, + { + "epoch": 0.5877740466604546, + "grad_norm": 3.4186601638793945, + "learning_rate": 7.280128087863414e-05, + "loss": 0.3115, + "step": 25074 + }, + { + "epoch": 0.5877974882352597, + "grad_norm": 3.5844223499298096, + "learning_rate": 7.27941924923281e-05, + "loss": 0.3243, + "step": 25075 + }, + { + "epoch": 0.5878209298100646, + "grad_norm": 2.810868978500366, + "learning_rate": 7.278710425364157e-05, + "loss": 0.2671, + "step": 25076 + }, + { + "epoch": 0.5878443713848697, + "grad_norm": 2.365015745162964, + "learning_rate": 7.278001616261296e-05, + "loss": 1.5441, + "step": 25077 + }, + { + "epoch": 0.5878678129596746, + "grad_norm": 4.85864782333374, + "learning_rate": 7.277292821928072e-05, + "loss": 0.4883, + "step": 25078 + }, + { + "epoch": 0.5878912545344797, + "grad_norm": 4.384846210479736, + "learning_rate": 7.276584042368338e-05, + "loss": 0.2349, + "step": 25079 + }, + { + "epoch": 0.5879146961092846, + "grad_norm": 3.237812042236328, + "learning_rate": 7.275875277585934e-05, + "loss": 0.3511, + "step": 25080 + }, + { + "epoch": 0.5879381376840896, + "grad_norm": 0.6376365423202515, + "learning_rate": 7.275166527584706e-05, + "loss": 0.0761, + "step": 25081 + }, + { + "epoch": 0.5879615792588946, + "grad_norm": 0.8095986843109131, + "learning_rate": 7.274457792368503e-05, + "loss": 0.0465, + "step": 25082 + }, + { + "epoch": 0.5879850208336996, + "grad_norm": 3.7030980587005615, + "learning_rate": 7.273749071941167e-05, + "loss": 0.2463, + "step": 25083 + }, + { + "epoch": 0.5880084624085046, + "grad_norm": 1.1083626747131348, + "learning_rate": 7.273040366306545e-05, + "loss": 0.1141, + "step": 25084 + }, + { + "epoch": 0.5880319039833096, + "grad_norm": 0.2639613747596741, + "learning_rate": 7.272331675468485e-05, + "loss": 0.0209, + "step": 25085 + }, + { + "epoch": 0.5880553455581146, + "grad_norm": 2.7556540966033936, + "learning_rate": 7.271622999430825e-05, + "loss": 0.2097, + "step": 25086 + }, + { + "epoch": 0.5880787871329196, + "grad_norm": 1.6974903345108032, + "learning_rate": 7.270914338197419e-05, + "loss": 0.2302, + "step": 25087 + }, + { + "epoch": 0.5881022287077246, + "grad_norm": 3.20478892326355, + "learning_rate": 7.270205691772107e-05, + "loss": 0.5012, + "step": 25088 + }, + { + "epoch": 0.5881256702825296, + "grad_norm": 5.054197788238525, + "learning_rate": 7.269497060158739e-05, + "loss": 0.5069, + "step": 25089 + }, + { + "epoch": 0.5881491118573345, + "grad_norm": 2.849853277206421, + "learning_rate": 7.268788443361156e-05, + "loss": 0.2339, + "step": 25090 + }, + { + "epoch": 0.5881725534321396, + "grad_norm": 3.157529354095459, + "learning_rate": 7.2680798413832e-05, + "loss": 0.347, + "step": 25091 + }, + { + "epoch": 0.5881959950069445, + "grad_norm": 4.8996124267578125, + "learning_rate": 7.267371254228723e-05, + "loss": 0.6146, + "step": 25092 + }, + { + "epoch": 0.5882194365817496, + "grad_norm": 1.4936894178390503, + "learning_rate": 7.266662681901564e-05, + "loss": 0.1931, + "step": 25093 + }, + { + "epoch": 0.5882428781565545, + "grad_norm": 4.199818134307861, + "learning_rate": 7.265954124405573e-05, + "loss": 0.4637, + "step": 25094 + }, + { + "epoch": 0.5882663197313596, + "grad_norm": 3.33101749420166, + "learning_rate": 7.26524558174459e-05, + "loss": 0.2939, + "step": 25095 + }, + { + "epoch": 0.5882897613061645, + "grad_norm": 2.8802568912506104, + "learning_rate": 7.264537053922462e-05, + "loss": 0.1361, + "step": 25096 + }, + { + "epoch": 0.5883132028809696, + "grad_norm": 3.5170156955718994, + "learning_rate": 7.263828540943033e-05, + "loss": 0.2413, + "step": 25097 + }, + { + "epoch": 0.5883366444557745, + "grad_norm": 2.580028533935547, + "learning_rate": 7.263120042810143e-05, + "loss": 0.3139, + "step": 25098 + }, + { + "epoch": 0.5883600860305795, + "grad_norm": 3.3006467819213867, + "learning_rate": 7.262411559527646e-05, + "loss": 0.1883, + "step": 25099 + }, + { + "epoch": 0.5883835276053845, + "grad_norm": 0.9225574135780334, + "learning_rate": 7.261703091099378e-05, + "loss": 0.0772, + "step": 25100 + }, + { + "epoch": 0.5884069691801895, + "grad_norm": 2.034288167953491, + "learning_rate": 7.260994637529188e-05, + "loss": 0.2411, + "step": 25101 + }, + { + "epoch": 0.5884304107549945, + "grad_norm": 3.919708728790283, + "learning_rate": 7.260286198820918e-05, + "loss": 0.2336, + "step": 25102 + }, + { + "epoch": 0.5884538523297995, + "grad_norm": 4.0982184410095215, + "learning_rate": 7.259577774978411e-05, + "loss": 0.4004, + "step": 25103 + }, + { + "epoch": 0.5884772939046045, + "grad_norm": 2.1814002990722656, + "learning_rate": 7.258869366005514e-05, + "loss": 0.1777, + "step": 25104 + }, + { + "epoch": 0.5885007354794095, + "grad_norm": 5.494659900665283, + "learning_rate": 7.258160971906067e-05, + "loss": 0.8877, + "step": 25105 + }, + { + "epoch": 0.5885241770542144, + "grad_norm": 4.185049533843994, + "learning_rate": 7.257452592683915e-05, + "loss": 0.4817, + "step": 25106 + }, + { + "epoch": 0.5885476186290195, + "grad_norm": 4.325591087341309, + "learning_rate": 7.256744228342905e-05, + "loss": 0.3233, + "step": 25107 + }, + { + "epoch": 0.5885710602038245, + "grad_norm": 3.8418798446655273, + "learning_rate": 7.256035878886878e-05, + "loss": 0.3668, + "step": 25108 + }, + { + "epoch": 0.5885945017786295, + "grad_norm": 1.5028983354568481, + "learning_rate": 7.255327544319674e-05, + "loss": 1.3041, + "step": 25109 + }, + { + "epoch": 0.5886179433534345, + "grad_norm": 3.347639799118042, + "learning_rate": 7.254619224645139e-05, + "loss": 0.1258, + "step": 25110 + }, + { + "epoch": 0.5886413849282395, + "grad_norm": 2.165578842163086, + "learning_rate": 7.25391091986712e-05, + "loss": 1.2399, + "step": 25111 + }, + { + "epoch": 0.5886648265030445, + "grad_norm": 1.7467983961105347, + "learning_rate": 7.25320262998946e-05, + "loss": 1.6101, + "step": 25112 + }, + { + "epoch": 0.5886882680778495, + "grad_norm": 2.0536959171295166, + "learning_rate": 7.252494355015995e-05, + "loss": 0.1972, + "step": 25113 + }, + { + "epoch": 0.5887117096526545, + "grad_norm": 3.889540433883667, + "learning_rate": 7.251786094950576e-05, + "loss": 0.2959, + "step": 25114 + }, + { + "epoch": 0.5887351512274595, + "grad_norm": 1.4378479719161987, + "learning_rate": 7.251077849797043e-05, + "loss": 0.1575, + "step": 25115 + }, + { + "epoch": 0.5887585928022645, + "grad_norm": 1.905975580215454, + "learning_rate": 7.250369619559238e-05, + "loss": 0.2022, + "step": 25116 + }, + { + "epoch": 0.5887820343770694, + "grad_norm": 3.210155963897705, + "learning_rate": 7.249661404241006e-05, + "loss": 0.2042, + "step": 25117 + }, + { + "epoch": 0.5888054759518745, + "grad_norm": 3.6626012325286865, + "learning_rate": 7.248953203846187e-05, + "loss": 0.3459, + "step": 25118 + }, + { + "epoch": 0.5888289175266794, + "grad_norm": 2.971062421798706, + "learning_rate": 7.248245018378626e-05, + "loss": 0.1812, + "step": 25119 + }, + { + "epoch": 0.5888523591014845, + "grad_norm": 1.0782006978988647, + "learning_rate": 7.247536847842164e-05, + "loss": 0.0777, + "step": 25120 + }, + { + "epoch": 0.5888758006762894, + "grad_norm": 2.569411039352417, + "learning_rate": 7.246828692240647e-05, + "loss": 0.0996, + "step": 25121 + }, + { + "epoch": 0.5888992422510945, + "grad_norm": 5.061971187591553, + "learning_rate": 7.24612055157791e-05, + "loss": 0.3543, + "step": 25122 + }, + { + "epoch": 0.5889226838258994, + "grad_norm": 1.524237036705017, + "learning_rate": 7.245412425857802e-05, + "loss": 0.1138, + "step": 25123 + }, + { + "epoch": 0.5889461254007045, + "grad_norm": 0.9339098334312439, + "learning_rate": 7.244704315084166e-05, + "loss": 0.3193, + "step": 25124 + }, + { + "epoch": 0.5889695669755094, + "grad_norm": 3.889857530593872, + "learning_rate": 7.243996219260841e-05, + "loss": 0.2825, + "step": 25125 + }, + { + "epoch": 0.5889930085503144, + "grad_norm": 4.719790458679199, + "learning_rate": 7.243288138391668e-05, + "loss": 0.4078, + "step": 25126 + }, + { + "epoch": 0.5890164501251194, + "grad_norm": 0.7989391684532166, + "learning_rate": 7.242580072480494e-05, + "loss": 0.0295, + "step": 25127 + }, + { + "epoch": 0.5890398916999244, + "grad_norm": 2.0120604038238525, + "learning_rate": 7.241872021531157e-05, + "loss": 0.1918, + "step": 25128 + }, + { + "epoch": 0.5890633332747294, + "grad_norm": 1.3485972881317139, + "learning_rate": 7.2411639855475e-05, + "loss": 0.098, + "step": 25129 + }, + { + "epoch": 0.5890867748495344, + "grad_norm": 2.7520484924316406, + "learning_rate": 7.240455964533366e-05, + "loss": 0.1374, + "step": 25130 + }, + { + "epoch": 0.5891102164243394, + "grad_norm": 0.22172315418720245, + "learning_rate": 7.239747958492594e-05, + "loss": 0.0147, + "step": 25131 + }, + { + "epoch": 0.5891336579991444, + "grad_norm": 2.8095338344573975, + "learning_rate": 7.239039967429026e-05, + "loss": 0.1201, + "step": 25132 + }, + { + "epoch": 0.5891570995739493, + "grad_norm": 0.7409512996673584, + "learning_rate": 7.238331991346508e-05, + "loss": 0.0347, + "step": 25133 + }, + { + "epoch": 0.5891805411487544, + "grad_norm": 2.145230293273926, + "learning_rate": 7.237624030248872e-05, + "loss": 0.2152, + "step": 25134 + }, + { + "epoch": 0.5892039827235593, + "grad_norm": 6.000970840454102, + "learning_rate": 7.236916084139971e-05, + "loss": 0.5012, + "step": 25135 + }, + { + "epoch": 0.5892274242983644, + "grad_norm": 4.1261820793151855, + "learning_rate": 7.23620815302364e-05, + "loss": 0.3494, + "step": 25136 + }, + { + "epoch": 0.5892508658731693, + "grad_norm": 5.421205997467041, + "learning_rate": 7.235500236903723e-05, + "loss": 0.3385, + "step": 25137 + }, + { + "epoch": 0.5892743074479744, + "grad_norm": 2.692406415939331, + "learning_rate": 7.234792335784055e-05, + "loss": 0.2364, + "step": 25138 + }, + { + "epoch": 0.5892977490227793, + "grad_norm": 3.573194742202759, + "learning_rate": 7.234084449668485e-05, + "loss": 0.2567, + "step": 25139 + }, + { + "epoch": 0.5893211905975844, + "grad_norm": 2.9617831707000732, + "learning_rate": 7.233376578560852e-05, + "loss": 0.3328, + "step": 25140 + }, + { + "epoch": 0.5893446321723893, + "grad_norm": 1.941123604774475, + "learning_rate": 7.232668722464992e-05, + "loss": 0.1486, + "step": 25141 + }, + { + "epoch": 0.5893680737471944, + "grad_norm": 1.4662985801696777, + "learning_rate": 7.231960881384752e-05, + "loss": 0.083, + "step": 25142 + }, + { + "epoch": 0.5893915153219993, + "grad_norm": 1.1563504934310913, + "learning_rate": 7.23125305532397e-05, + "loss": 0.0383, + "step": 25143 + }, + { + "epoch": 0.5894149568968043, + "grad_norm": 2.3360755443573, + "learning_rate": 7.230545244286484e-05, + "loss": 0.2632, + "step": 25144 + }, + { + "epoch": 0.5894383984716093, + "grad_norm": 0.5137950778007507, + "learning_rate": 7.22983744827614e-05, + "loss": 0.0698, + "step": 25145 + }, + { + "epoch": 0.5894618400464143, + "grad_norm": 5.194354057312012, + "learning_rate": 7.229129667296772e-05, + "loss": 0.5497, + "step": 25146 + }, + { + "epoch": 0.5894852816212193, + "grad_norm": 0.8479472994804382, + "learning_rate": 7.228421901352227e-05, + "loss": 0.0834, + "step": 25147 + }, + { + "epoch": 0.5895087231960243, + "grad_norm": 0.9195650815963745, + "learning_rate": 7.227714150446343e-05, + "loss": 0.0769, + "step": 25148 + }, + { + "epoch": 0.5895321647708293, + "grad_norm": 4.399597644805908, + "learning_rate": 7.22700641458296e-05, + "loss": 0.2613, + "step": 25149 + }, + { + "epoch": 0.5895556063456343, + "grad_norm": 1.1534732580184937, + "learning_rate": 7.226298693765918e-05, + "loss": 0.098, + "step": 25150 + }, + { + "epoch": 0.5895790479204392, + "grad_norm": 2.779566526412964, + "learning_rate": 7.225590987999057e-05, + "loss": 0.3409, + "step": 25151 + }, + { + "epoch": 0.5896024894952443, + "grad_norm": 3.364201545715332, + "learning_rate": 7.224883297286218e-05, + "loss": 0.3205, + "step": 25152 + }, + { + "epoch": 0.5896259310700492, + "grad_norm": 0.7966633439064026, + "learning_rate": 7.224175621631242e-05, + "loss": 0.0381, + "step": 25153 + }, + { + "epoch": 0.5896493726448543, + "grad_norm": 2.968017101287842, + "learning_rate": 7.223467961037962e-05, + "loss": 0.3602, + "step": 25154 + }, + { + "epoch": 0.5896728142196592, + "grad_norm": 3.3851566314697266, + "learning_rate": 7.222760315510229e-05, + "loss": 0.305, + "step": 25155 + }, + { + "epoch": 0.5896962557944643, + "grad_norm": 4.030223846435547, + "learning_rate": 7.222052685051873e-05, + "loss": 0.2661, + "step": 25156 + }, + { + "epoch": 0.5897196973692692, + "grad_norm": 3.521322250366211, + "learning_rate": 7.221345069666736e-05, + "loss": 0.3492, + "step": 25157 + }, + { + "epoch": 0.5897431389440743, + "grad_norm": 1.0820308923721313, + "learning_rate": 7.220637469358662e-05, + "loss": 0.0755, + "step": 25158 + }, + { + "epoch": 0.5897665805188793, + "grad_norm": 4.842183589935303, + "learning_rate": 7.219929884131482e-05, + "loss": 0.5965, + "step": 25159 + }, + { + "epoch": 0.5897900220936843, + "grad_norm": 4.211267471313477, + "learning_rate": 7.219222313989047e-05, + "loss": 0.4095, + "step": 25160 + }, + { + "epoch": 0.5898134636684893, + "grad_norm": 2.6054673194885254, + "learning_rate": 7.218514758935186e-05, + "loss": 0.2025, + "step": 25161 + }, + { + "epoch": 0.5898369052432942, + "grad_norm": 2.3436174392700195, + "learning_rate": 7.217807218973745e-05, + "loss": 0.1115, + "step": 25162 + }, + { + "epoch": 0.5898603468180993, + "grad_norm": 4.259163856506348, + "learning_rate": 7.217099694108557e-05, + "loss": 0.4906, + "step": 25163 + }, + { + "epoch": 0.5898837883929042, + "grad_norm": 0.9704297184944153, + "learning_rate": 7.216392184343466e-05, + "loss": 0.0506, + "step": 25164 + }, + { + "epoch": 0.5899072299677093, + "grad_norm": 3.2243289947509766, + "learning_rate": 7.21568468968231e-05, + "loss": 1.3475, + "step": 25165 + }, + { + "epoch": 0.5899306715425142, + "grad_norm": 1.5473805665969849, + "learning_rate": 7.214977210128927e-05, + "loss": 0.2488, + "step": 25166 + }, + { + "epoch": 0.5899541131173193, + "grad_norm": 2.34452223777771, + "learning_rate": 7.214269745687154e-05, + "loss": 0.123, + "step": 25167 + }, + { + "epoch": 0.5899775546921242, + "grad_norm": 3.740950107574463, + "learning_rate": 7.213562296360834e-05, + "loss": 0.3647, + "step": 25168 + }, + { + "epoch": 0.5900009962669293, + "grad_norm": 2.749075174331665, + "learning_rate": 7.212854862153802e-05, + "loss": 0.1634, + "step": 25169 + }, + { + "epoch": 0.5900244378417342, + "grad_norm": 0.7056984305381775, + "learning_rate": 7.212147443069899e-05, + "loss": 0.0929, + "step": 25170 + }, + { + "epoch": 0.5900478794165392, + "grad_norm": 3.696019411087036, + "learning_rate": 7.211440039112959e-05, + "loss": 0.5167, + "step": 25171 + }, + { + "epoch": 0.5900713209913442, + "grad_norm": 4.236106872558594, + "learning_rate": 7.210732650286827e-05, + "loss": 0.3782, + "step": 25172 + }, + { + "epoch": 0.5900947625661492, + "grad_norm": 0.8478764295578003, + "learning_rate": 7.210025276595335e-05, + "loss": 0.0741, + "step": 25173 + }, + { + "epoch": 0.5901182041409542, + "grad_norm": 2.527683973312378, + "learning_rate": 7.209317918042328e-05, + "loss": 0.0979, + "step": 25174 + }, + { + "epoch": 0.5901416457157592, + "grad_norm": 2.8979198932647705, + "learning_rate": 7.208610574631641e-05, + "loss": 0.1235, + "step": 25175 + }, + { + "epoch": 0.5901650872905642, + "grad_norm": 4.097794532775879, + "learning_rate": 7.207903246367109e-05, + "loss": 0.3427, + "step": 25176 + }, + { + "epoch": 0.5901885288653692, + "grad_norm": 0.5317398905754089, + "learning_rate": 7.207195933252574e-05, + "loss": 0.0819, + "step": 25177 + }, + { + "epoch": 0.5902119704401741, + "grad_norm": 2.3041155338287354, + "learning_rate": 7.206488635291872e-05, + "loss": 0.1723, + "step": 25178 + }, + { + "epoch": 0.5902354120149792, + "grad_norm": 1.1145098209381104, + "learning_rate": 7.205781352488841e-05, + "loss": 0.1155, + "step": 25179 + }, + { + "epoch": 0.5902588535897841, + "grad_norm": 3.291829824447632, + "learning_rate": 7.205074084847319e-05, + "loss": 0.2081, + "step": 25180 + }, + { + "epoch": 0.5902822951645892, + "grad_norm": 0.8761309385299683, + "learning_rate": 7.204366832371144e-05, + "loss": 0.1022, + "step": 25181 + }, + { + "epoch": 0.5903057367393941, + "grad_norm": 4.465076446533203, + "learning_rate": 7.203659595064152e-05, + "loss": 0.4988, + "step": 25182 + }, + { + "epoch": 0.5903291783141992, + "grad_norm": 2.624927043914795, + "learning_rate": 7.202952372930179e-05, + "loss": 0.2043, + "step": 25183 + }, + { + "epoch": 0.5903526198890041, + "grad_norm": 3.453822612762451, + "learning_rate": 7.202245165973069e-05, + "loss": 0.3261, + "step": 25184 + }, + { + "epoch": 0.5903760614638092, + "grad_norm": 1.9807027578353882, + "learning_rate": 7.201537974196654e-05, + "loss": 0.1688, + "step": 25185 + }, + { + "epoch": 0.5903995030386141, + "grad_norm": 2.3194339275360107, + "learning_rate": 7.200830797604775e-05, + "loss": 0.2152, + "step": 25186 + }, + { + "epoch": 0.5904229446134192, + "grad_norm": 1.721191167831421, + "learning_rate": 7.200123636201263e-05, + "loss": 0.8043, + "step": 25187 + }, + { + "epoch": 0.5904463861882241, + "grad_norm": 4.494699001312256, + "learning_rate": 7.199416489989963e-05, + "loss": 0.3487, + "step": 25188 + }, + { + "epoch": 0.5904698277630291, + "grad_norm": 1.6200242042541504, + "learning_rate": 7.198709358974705e-05, + "loss": 0.1109, + "step": 25189 + }, + { + "epoch": 0.5904932693378341, + "grad_norm": 3.215268135070801, + "learning_rate": 7.19800224315933e-05, + "loss": 1.444, + "step": 25190 + }, + { + "epoch": 0.5905167109126391, + "grad_norm": 2.3276078701019287, + "learning_rate": 7.197295142547675e-05, + "loss": 0.1348, + "step": 25191 + }, + { + "epoch": 0.5905401524874441, + "grad_norm": 2.5916829109191895, + "learning_rate": 7.196588057143572e-05, + "loss": 0.7421, + "step": 25192 + }, + { + "epoch": 0.5905635940622491, + "grad_norm": 3.521801710128784, + "learning_rate": 7.195880986950863e-05, + "loss": 0.2244, + "step": 25193 + }, + { + "epoch": 0.5905870356370541, + "grad_norm": 0.45439413189888, + "learning_rate": 7.195173931973384e-05, + "loss": 0.0311, + "step": 25194 + }, + { + "epoch": 0.5906104772118591, + "grad_norm": 1.8698745965957642, + "learning_rate": 7.194466892214965e-05, + "loss": 1.2873, + "step": 25195 + }, + { + "epoch": 0.590633918786664, + "grad_norm": 3.309366464614868, + "learning_rate": 7.19375986767945e-05, + "loss": 1.2079, + "step": 25196 + }, + { + "epoch": 0.5906573603614691, + "grad_norm": 3.6333794593811035, + "learning_rate": 7.193052858370675e-05, + "loss": 0.3211, + "step": 25197 + }, + { + "epoch": 0.590680801936274, + "grad_norm": 3.5408449172973633, + "learning_rate": 7.192345864292474e-05, + "loss": 0.1895, + "step": 25198 + }, + { + "epoch": 0.5907042435110791, + "grad_norm": 3.3317034244537354, + "learning_rate": 7.19163888544868e-05, + "loss": 0.2275, + "step": 25199 + }, + { + "epoch": 0.590727685085884, + "grad_norm": 0.9760609865188599, + "learning_rate": 7.190931921843136e-05, + "loss": 0.0878, + "step": 25200 + }, + { + "epoch": 0.5907511266606891, + "grad_norm": 3.817753791809082, + "learning_rate": 7.190224973479673e-05, + "loss": 0.3956, + "step": 25201 + }, + { + "epoch": 0.590774568235494, + "grad_norm": 3.7798802852630615, + "learning_rate": 7.189518040362127e-05, + "loss": 0.2954, + "step": 25202 + }, + { + "epoch": 0.5907980098102991, + "grad_norm": 1.234032154083252, + "learning_rate": 7.188811122494337e-05, + "loss": 0.0975, + "step": 25203 + }, + { + "epoch": 0.590821451385104, + "grad_norm": 2.257418155670166, + "learning_rate": 7.188104219880136e-05, + "loss": 0.2874, + "step": 25204 + }, + { + "epoch": 0.590844892959909, + "grad_norm": 1.614392638206482, + "learning_rate": 7.187397332523361e-05, + "loss": 0.1592, + "step": 25205 + }, + { + "epoch": 0.590868334534714, + "grad_norm": 4.146676540374756, + "learning_rate": 7.186690460427847e-05, + "loss": 0.4684, + "step": 25206 + }, + { + "epoch": 0.590891776109519, + "grad_norm": 2.4213764667510986, + "learning_rate": 7.185983603597429e-05, + "loss": 0.2836, + "step": 25207 + }, + { + "epoch": 0.590915217684324, + "grad_norm": 2.20491099357605, + "learning_rate": 7.185276762035942e-05, + "loss": 0.1609, + "step": 25208 + }, + { + "epoch": 0.590938659259129, + "grad_norm": 0.4937579929828644, + "learning_rate": 7.184569935747224e-05, + "loss": 0.0594, + "step": 25209 + }, + { + "epoch": 0.5909621008339341, + "grad_norm": 3.416961908340454, + "learning_rate": 7.18386312473511e-05, + "loss": 0.3252, + "step": 25210 + }, + { + "epoch": 0.590985542408739, + "grad_norm": 3.129626989364624, + "learning_rate": 7.183156329003431e-05, + "loss": 0.2912, + "step": 25211 + }, + { + "epoch": 0.5910089839835441, + "grad_norm": 3.125588893890381, + "learning_rate": 7.182449548556028e-05, + "loss": 0.2704, + "step": 25212 + }, + { + "epoch": 0.591032425558349, + "grad_norm": 1.301615834236145, + "learning_rate": 7.181742783396732e-05, + "loss": 0.0831, + "step": 25213 + }, + { + "epoch": 0.5910558671331541, + "grad_norm": 3.331498861312866, + "learning_rate": 7.181036033529378e-05, + "loss": 0.2281, + "step": 25214 + }, + { + "epoch": 0.591079308707959, + "grad_norm": 0.7871124744415283, + "learning_rate": 7.180329298957804e-05, + "loss": 0.0473, + "step": 25215 + }, + { + "epoch": 0.591102750282764, + "grad_norm": 4.597546577453613, + "learning_rate": 7.179622579685841e-05, + "loss": 0.3396, + "step": 25216 + }, + { + "epoch": 0.591126191857569, + "grad_norm": 2.7177810668945312, + "learning_rate": 7.178915875717323e-05, + "loss": 0.2415, + "step": 25217 + }, + { + "epoch": 0.591149633432374, + "grad_norm": 2.370316505432129, + "learning_rate": 7.178209187056091e-05, + "loss": 0.243, + "step": 25218 + }, + { + "epoch": 0.591173075007179, + "grad_norm": 1.2307674884796143, + "learning_rate": 7.17750251370597e-05, + "loss": 0.0795, + "step": 25219 + }, + { + "epoch": 0.591196516581984, + "grad_norm": 3.5060722827911377, + "learning_rate": 7.176795855670803e-05, + "loss": 0.3881, + "step": 25220 + }, + { + "epoch": 0.591219958156789, + "grad_norm": 0.6997515559196472, + "learning_rate": 7.176089212954419e-05, + "loss": 0.0396, + "step": 25221 + }, + { + "epoch": 0.591243399731594, + "grad_norm": 3.272949457168579, + "learning_rate": 7.175382585560657e-05, + "loss": 0.3013, + "step": 25222 + }, + { + "epoch": 0.591266841306399, + "grad_norm": 2.428446054458618, + "learning_rate": 7.17467597349335e-05, + "loss": 1.2111, + "step": 25223 + }, + { + "epoch": 0.591290282881204, + "grad_norm": 2.0188357830047607, + "learning_rate": 7.173969376756327e-05, + "loss": 0.1376, + "step": 25224 + }, + { + "epoch": 0.5913137244560089, + "grad_norm": 1.7907798290252686, + "learning_rate": 7.173262795353427e-05, + "loss": 0.1708, + "step": 25225 + }, + { + "epoch": 0.591337166030814, + "grad_norm": 1.7425017356872559, + "learning_rate": 7.172556229288483e-05, + "loss": 0.1494, + "step": 25226 + }, + { + "epoch": 0.5913606076056189, + "grad_norm": 3.5249388217926025, + "learning_rate": 7.171849678565327e-05, + "loss": 0.2983, + "step": 25227 + }, + { + "epoch": 0.591384049180424, + "grad_norm": 2.4327032566070557, + "learning_rate": 7.171143143187795e-05, + "loss": 0.1861, + "step": 25228 + }, + { + "epoch": 0.5914074907552289, + "grad_norm": 4.243174076080322, + "learning_rate": 7.170436623159722e-05, + "loss": 0.5355, + "step": 25229 + }, + { + "epoch": 0.591430932330034, + "grad_norm": 0.9635024666786194, + "learning_rate": 7.169730118484936e-05, + "loss": 0.1096, + "step": 25230 + }, + { + "epoch": 0.5914543739048389, + "grad_norm": 3.3065431118011475, + "learning_rate": 7.169023629167275e-05, + "loss": 1.697, + "step": 25231 + }, + { + "epoch": 0.591477815479644, + "grad_norm": 5.893087387084961, + "learning_rate": 7.168317155210569e-05, + "loss": 0.6738, + "step": 25232 + }, + { + "epoch": 0.5915012570544489, + "grad_norm": 1.8843402862548828, + "learning_rate": 7.167610696618657e-05, + "loss": 0.1571, + "step": 25233 + }, + { + "epoch": 0.5915246986292539, + "grad_norm": 0.961747407913208, + "learning_rate": 7.166904253395368e-05, + "loss": 0.0767, + "step": 25234 + }, + { + "epoch": 0.5915481402040589, + "grad_norm": 1.191076397895813, + "learning_rate": 7.166197825544538e-05, + "loss": 0.1177, + "step": 25235 + }, + { + "epoch": 0.5915715817788639, + "grad_norm": 5.873642444610596, + "learning_rate": 7.165491413069997e-05, + "loss": 0.2212, + "step": 25236 + }, + { + "epoch": 0.5915950233536689, + "grad_norm": 3.895078182220459, + "learning_rate": 7.164785015975579e-05, + "loss": 0.3736, + "step": 25237 + }, + { + "epoch": 0.5916184649284739, + "grad_norm": 2.6471633911132812, + "learning_rate": 7.164078634265116e-05, + "loss": 1.5414, + "step": 25238 + }, + { + "epoch": 0.5916419065032789, + "grad_norm": 2.1803042888641357, + "learning_rate": 7.163372267942445e-05, + "loss": 0.2439, + "step": 25239 + }, + { + "epoch": 0.5916653480780839, + "grad_norm": 2.4379122257232666, + "learning_rate": 7.162665917011393e-05, + "loss": 0.2324, + "step": 25240 + }, + { + "epoch": 0.5916887896528888, + "grad_norm": 1.6579302549362183, + "learning_rate": 7.161959581475798e-05, + "loss": 0.8032, + "step": 25241 + }, + { + "epoch": 0.5917122312276939, + "grad_norm": 1.7865009307861328, + "learning_rate": 7.16125326133949e-05, + "loss": 1.0248, + "step": 25242 + }, + { + "epoch": 0.5917356728024988, + "grad_norm": 2.455576181411743, + "learning_rate": 7.1605469566063e-05, + "loss": 2.1787, + "step": 25243 + }, + { + "epoch": 0.5917591143773039, + "grad_norm": 3.7242677211761475, + "learning_rate": 7.159840667280059e-05, + "loss": 0.3944, + "step": 25244 + }, + { + "epoch": 0.5917825559521088, + "grad_norm": 1.5522171258926392, + "learning_rate": 7.159134393364606e-05, + "loss": 0.1256, + "step": 25245 + }, + { + "epoch": 0.5918059975269139, + "grad_norm": 3.5810577869415283, + "learning_rate": 7.158428134863769e-05, + "loss": 0.4198, + "step": 25246 + }, + { + "epoch": 0.5918294391017188, + "grad_norm": 1.1776198148727417, + "learning_rate": 7.157721891781383e-05, + "loss": 0.1239, + "step": 25247 + }, + { + "epoch": 0.5918528806765239, + "grad_norm": 4.442065715789795, + "learning_rate": 7.157015664121277e-05, + "loss": 0.3942, + "step": 25248 + }, + { + "epoch": 0.5918763222513288, + "grad_norm": 2.10713791847229, + "learning_rate": 7.156309451887282e-05, + "loss": 0.1709, + "step": 25249 + }, + { + "epoch": 0.5918997638261339, + "grad_norm": 0.750002384185791, + "learning_rate": 7.155603255083234e-05, + "loss": 0.0586, + "step": 25250 + }, + { + "epoch": 0.5919232054009388, + "grad_norm": 4.2023138999938965, + "learning_rate": 7.154897073712962e-05, + "loss": 0.2872, + "step": 25251 + }, + { + "epoch": 0.5919466469757438, + "grad_norm": 3.0309078693389893, + "learning_rate": 7.154190907780298e-05, + "loss": 0.2597, + "step": 25252 + }, + { + "epoch": 0.5919700885505488, + "grad_norm": 0.869499146938324, + "learning_rate": 7.153484757289075e-05, + "loss": 0.1046, + "step": 25253 + }, + { + "epoch": 0.5919935301253538, + "grad_norm": 4.16797399520874, + "learning_rate": 7.152778622243124e-05, + "loss": 0.4657, + "step": 25254 + }, + { + "epoch": 0.5920169717001588, + "grad_norm": 1.580392837524414, + "learning_rate": 7.152072502646276e-05, + "loss": 0.0385, + "step": 25255 + }, + { + "epoch": 0.5920404132749638, + "grad_norm": 2.168379068374634, + "learning_rate": 7.151366398502358e-05, + "loss": 0.1762, + "step": 25256 + }, + { + "epoch": 0.5920638548497688, + "grad_norm": 0.8923681974411011, + "learning_rate": 7.150660309815211e-05, + "loss": 0.048, + "step": 25257 + }, + { + "epoch": 0.5920872964245738, + "grad_norm": 1.5423706769943237, + "learning_rate": 7.149954236588662e-05, + "loss": 0.1703, + "step": 25258 + }, + { + "epoch": 0.5921107379993787, + "grad_norm": 1.0161372423171997, + "learning_rate": 7.149248178826538e-05, + "loss": 0.0949, + "step": 25259 + }, + { + "epoch": 0.5921341795741838, + "grad_norm": 0.38354936242103577, + "learning_rate": 7.148542136532677e-05, + "loss": 0.0182, + "step": 25260 + }, + { + "epoch": 0.5921576211489888, + "grad_norm": 2.3287453651428223, + "learning_rate": 7.147836109710906e-05, + "loss": 0.232, + "step": 25261 + }, + { + "epoch": 0.5921810627237938, + "grad_norm": 2.3349549770355225, + "learning_rate": 7.147130098365054e-05, + "loss": 0.1517, + "step": 25262 + }, + { + "epoch": 0.5922045042985988, + "grad_norm": 1.2786568403244019, + "learning_rate": 7.146424102498956e-05, + "loss": 1.5313, + "step": 25263 + }, + { + "epoch": 0.5922279458734038, + "grad_norm": 2.9807095527648926, + "learning_rate": 7.145718122116442e-05, + "loss": 0.1936, + "step": 25264 + }, + { + "epoch": 0.5922513874482088, + "grad_norm": 1.8366395235061646, + "learning_rate": 7.14501215722134e-05, + "loss": 0.1, + "step": 25265 + }, + { + "epoch": 0.5922748290230138, + "grad_norm": 1.813050389289856, + "learning_rate": 7.144306207817484e-05, + "loss": 0.1061, + "step": 25266 + }, + { + "epoch": 0.5922982705978188, + "grad_norm": 6.240150451660156, + "learning_rate": 7.143600273908703e-05, + "loss": 0.6531, + "step": 25267 + }, + { + "epoch": 0.5923217121726237, + "grad_norm": 4.260559558868408, + "learning_rate": 7.142894355498823e-05, + "loss": 0.3732, + "step": 25268 + }, + { + "epoch": 0.5923451537474288, + "grad_norm": 2.173701763153076, + "learning_rate": 7.14218845259168e-05, + "loss": 2.0013, + "step": 25269 + }, + { + "epoch": 0.5923685953222337, + "grad_norm": 0.831721842288971, + "learning_rate": 7.141482565191105e-05, + "loss": 0.0913, + "step": 25270 + }, + { + "epoch": 0.5923920368970388, + "grad_norm": 2.1834282875061035, + "learning_rate": 7.140776693300926e-05, + "loss": 0.2701, + "step": 25271 + }, + { + "epoch": 0.5924154784718437, + "grad_norm": 0.36384451389312744, + "learning_rate": 7.140070836924972e-05, + "loss": 0.0177, + "step": 25272 + }, + { + "epoch": 0.5924389200466488, + "grad_norm": 1.0692760944366455, + "learning_rate": 7.139364996067075e-05, + "loss": 0.1228, + "step": 25273 + }, + { + "epoch": 0.5924623616214537, + "grad_norm": 2.7339425086975098, + "learning_rate": 7.138659170731064e-05, + "loss": 0.2352, + "step": 25274 + }, + { + "epoch": 0.5924858031962588, + "grad_norm": 4.175636291503906, + "learning_rate": 7.137953360920768e-05, + "loss": 0.3682, + "step": 25275 + }, + { + "epoch": 0.5925092447710637, + "grad_norm": 3.3456122875213623, + "learning_rate": 7.137247566640017e-05, + "loss": 1.2464, + "step": 25276 + }, + { + "epoch": 0.5925326863458688, + "grad_norm": 4.883561611175537, + "learning_rate": 7.136541787892643e-05, + "loss": 0.4358, + "step": 25277 + }, + { + "epoch": 0.5925561279206737, + "grad_norm": 2.365567445755005, + "learning_rate": 7.135836024682472e-05, + "loss": 1.3946, + "step": 25278 + }, + { + "epoch": 0.5925795694954787, + "grad_norm": 2.70121169090271, + "learning_rate": 7.135130277013336e-05, + "loss": 0.2197, + "step": 25279 + }, + { + "epoch": 0.5926030110702837, + "grad_norm": 2.1924312114715576, + "learning_rate": 7.134424544889061e-05, + "loss": 0.1775, + "step": 25280 + }, + { + "epoch": 0.5926264526450887, + "grad_norm": 1.7567763328552246, + "learning_rate": 7.133718828313482e-05, + "loss": 0.191, + "step": 25281 + }, + { + "epoch": 0.5926498942198937, + "grad_norm": 1.7112383842468262, + "learning_rate": 7.133013127290425e-05, + "loss": 0.1977, + "step": 25282 + }, + { + "epoch": 0.5926733357946987, + "grad_norm": 0.6649146676063538, + "learning_rate": 7.13230744182372e-05, + "loss": 0.0298, + "step": 25283 + }, + { + "epoch": 0.5926967773695037, + "grad_norm": 2.0149505138397217, + "learning_rate": 7.131601771917193e-05, + "loss": 0.1324, + "step": 25284 + }, + { + "epoch": 0.5927202189443087, + "grad_norm": 1.7010623216629028, + "learning_rate": 7.130896117574678e-05, + "loss": 0.2039, + "step": 25285 + }, + { + "epoch": 0.5927436605191136, + "grad_norm": 0.7614791393280029, + "learning_rate": 7.1301904788e-05, + "loss": 0.0661, + "step": 25286 + }, + { + "epoch": 0.5927671020939187, + "grad_norm": 1.8906044960021973, + "learning_rate": 7.12948485559699e-05, + "loss": 0.1362, + "step": 25287 + }, + { + "epoch": 0.5927905436687236, + "grad_norm": 2.8633322715759277, + "learning_rate": 7.128779247969477e-05, + "loss": 0.1358, + "step": 25288 + }, + { + "epoch": 0.5928139852435287, + "grad_norm": 2.3484108448028564, + "learning_rate": 7.128073655921287e-05, + "loss": 0.0986, + "step": 25289 + }, + { + "epoch": 0.5928374268183336, + "grad_norm": 2.326509714126587, + "learning_rate": 7.12736807945625e-05, + "loss": 0.1216, + "step": 25290 + }, + { + "epoch": 0.5928608683931387, + "grad_norm": 3.5680034160614014, + "learning_rate": 7.126662518578195e-05, + "loss": 0.4734, + "step": 25291 + }, + { + "epoch": 0.5928843099679436, + "grad_norm": 3.1212735176086426, + "learning_rate": 7.125956973290945e-05, + "loss": 0.1, + "step": 25292 + }, + { + "epoch": 0.5929077515427487, + "grad_norm": 0.8523810505867004, + "learning_rate": 7.12525144359834e-05, + "loss": 0.0645, + "step": 25293 + }, + { + "epoch": 0.5929311931175536, + "grad_norm": 0.5584491491317749, + "learning_rate": 7.124545929504198e-05, + "loss": 0.0306, + "step": 25294 + }, + { + "epoch": 0.5929546346923587, + "grad_norm": 1.1594139337539673, + "learning_rate": 7.123840431012353e-05, + "loss": 0.0824, + "step": 25295 + }, + { + "epoch": 0.5929780762671636, + "grad_norm": 3.535771608352661, + "learning_rate": 7.123134948126631e-05, + "loss": 0.45, + "step": 25296 + }, + { + "epoch": 0.5930015178419686, + "grad_norm": 3.150928497314453, + "learning_rate": 7.122429480850856e-05, + "loss": 0.3363, + "step": 25297 + }, + { + "epoch": 0.5930249594167736, + "grad_norm": 1.7276989221572876, + "learning_rate": 7.121724029188863e-05, + "loss": 0.1862, + "step": 25298 + }, + { + "epoch": 0.5930484009915786, + "grad_norm": 1.6371642351150513, + "learning_rate": 7.121018593144478e-05, + "loss": 0.1581, + "step": 25299 + }, + { + "epoch": 0.5930718425663836, + "grad_norm": 0.6046463847160339, + "learning_rate": 7.120313172721522e-05, + "loss": 0.0517, + "step": 25300 + }, + { + "epoch": 0.5930952841411886, + "grad_norm": 2.43269419670105, + "learning_rate": 7.11960776792383e-05, + "loss": 1.5406, + "step": 25301 + }, + { + "epoch": 0.5931187257159936, + "grad_norm": 3.5744106769561768, + "learning_rate": 7.11890237875523e-05, + "loss": 0.2857, + "step": 25302 + }, + { + "epoch": 0.5931421672907986, + "grad_norm": 2.9192960262298584, + "learning_rate": 7.118197005219543e-05, + "loss": 0.2461, + "step": 25303 + }, + { + "epoch": 0.5931656088656035, + "grad_norm": 2.5485854148864746, + "learning_rate": 7.117491647320597e-05, + "loss": 0.3119, + "step": 25304 + }, + { + "epoch": 0.5931890504404086, + "grad_norm": 7.2377028465271, + "learning_rate": 7.116786305062228e-05, + "loss": 0.3053, + "step": 25305 + }, + { + "epoch": 0.5932124920152135, + "grad_norm": 4.099189281463623, + "learning_rate": 7.116080978448257e-05, + "loss": 0.581, + "step": 25306 + }, + { + "epoch": 0.5932359335900186, + "grad_norm": 2.5683815479278564, + "learning_rate": 7.11537566748251e-05, + "loss": 0.2001, + "step": 25307 + }, + { + "epoch": 0.5932593751648235, + "grad_norm": 2.583505868911743, + "learning_rate": 7.114670372168817e-05, + "loss": 0.1614, + "step": 25308 + }, + { + "epoch": 0.5932828167396286, + "grad_norm": 4.435747146606445, + "learning_rate": 7.113965092511003e-05, + "loss": 0.3724, + "step": 25309 + }, + { + "epoch": 0.5933062583144335, + "grad_norm": 3.814417600631714, + "learning_rate": 7.113259828512895e-05, + "loss": 0.4969, + "step": 25310 + }, + { + "epoch": 0.5933296998892386, + "grad_norm": 1.7252713441848755, + "learning_rate": 7.112554580178321e-05, + "loss": 0.14, + "step": 25311 + }, + { + "epoch": 0.5933531414640436, + "grad_norm": 1.1143853664398193, + "learning_rate": 7.111849347511108e-05, + "loss": 0.0962, + "step": 25312 + }, + { + "epoch": 0.5933765830388485, + "grad_norm": 1.5929027795791626, + "learning_rate": 7.111144130515079e-05, + "loss": 0.1145, + "step": 25313 + }, + { + "epoch": 0.5934000246136536, + "grad_norm": 2.0838420391082764, + "learning_rate": 7.110438929194067e-05, + "loss": 0.147, + "step": 25314 + }, + { + "epoch": 0.5934234661884585, + "grad_norm": 1.0242060422897339, + "learning_rate": 7.109733743551894e-05, + "loss": 0.1053, + "step": 25315 + }, + { + "epoch": 0.5934469077632636, + "grad_norm": 4.360939025878906, + "learning_rate": 7.109028573592384e-05, + "loss": 0.1632, + "step": 25316 + }, + { + "epoch": 0.5934703493380685, + "grad_norm": 2.604095458984375, + "learning_rate": 7.108323419319364e-05, + "loss": 0.3957, + "step": 25317 + }, + { + "epoch": 0.5934937909128736, + "grad_norm": 3.7205684185028076, + "learning_rate": 7.107618280736668e-05, + "loss": 1.4067, + "step": 25318 + }, + { + "epoch": 0.5935172324876785, + "grad_norm": 3.0186593532562256, + "learning_rate": 7.106913157848115e-05, + "loss": 0.0807, + "step": 25319 + }, + { + "epoch": 0.5935406740624836, + "grad_norm": 0.585180401802063, + "learning_rate": 7.106208050657532e-05, + "loss": 0.0812, + "step": 25320 + }, + { + "epoch": 0.5935641156372885, + "grad_norm": 4.1917033195495605, + "learning_rate": 7.105502959168746e-05, + "loss": 2.1359, + "step": 25321 + }, + { + "epoch": 0.5935875572120936, + "grad_norm": 1.5037785768508911, + "learning_rate": 7.10479788338558e-05, + "loss": 0.1487, + "step": 25322 + }, + { + "epoch": 0.5936109987868985, + "grad_norm": 1.1585235595703125, + "learning_rate": 7.104092823311865e-05, + "loss": 0.0926, + "step": 25323 + }, + { + "epoch": 0.5936344403617035, + "grad_norm": 1.6911581754684448, + "learning_rate": 7.103387778951425e-05, + "loss": 0.1446, + "step": 25324 + }, + { + "epoch": 0.5936578819365085, + "grad_norm": 1.114446997642517, + "learning_rate": 7.102682750308081e-05, + "loss": 0.1237, + "step": 25325 + }, + { + "epoch": 0.5936813235113135, + "grad_norm": 2.4386961460113525, + "learning_rate": 7.101977737385663e-05, + "loss": 1.7944, + "step": 25326 + }, + { + "epoch": 0.5937047650861185, + "grad_norm": 2.529283285140991, + "learning_rate": 7.101272740187996e-05, + "loss": 0.2357, + "step": 25327 + }, + { + "epoch": 0.5937282066609235, + "grad_norm": 3.421151638031006, + "learning_rate": 7.100567758718904e-05, + "loss": 0.47, + "step": 25328 + }, + { + "epoch": 0.5937516482357285, + "grad_norm": 1.3693275451660156, + "learning_rate": 7.09986279298221e-05, + "loss": 0.16, + "step": 25329 + }, + { + "epoch": 0.5937750898105335, + "grad_norm": 1.5995279550552368, + "learning_rate": 7.099157842981746e-05, + "loss": 0.0771, + "step": 25330 + }, + { + "epoch": 0.5937985313853384, + "grad_norm": 1.7585290670394897, + "learning_rate": 7.098452908721333e-05, + "loss": 0.2346, + "step": 25331 + }, + { + "epoch": 0.5938219729601435, + "grad_norm": 2.563894510269165, + "learning_rate": 7.097747990204793e-05, + "loss": 0.2006, + "step": 25332 + }, + { + "epoch": 0.5938454145349484, + "grad_norm": 6.611295700073242, + "learning_rate": 7.097043087435957e-05, + "loss": 0.6353, + "step": 25333 + }, + { + "epoch": 0.5938688561097535, + "grad_norm": 0.7005134224891663, + "learning_rate": 7.096338200418648e-05, + "loss": 0.0809, + "step": 25334 + }, + { + "epoch": 0.5938922976845584, + "grad_norm": 3.6485583782196045, + "learning_rate": 7.095633329156686e-05, + "loss": 0.3193, + "step": 25335 + }, + { + "epoch": 0.5939157392593635, + "grad_norm": 0.7621651887893677, + "learning_rate": 7.094928473653901e-05, + "loss": 0.0837, + "step": 25336 + }, + { + "epoch": 0.5939391808341684, + "grad_norm": 3.2994015216827393, + "learning_rate": 7.094223633914117e-05, + "loss": 0.2907, + "step": 25337 + }, + { + "epoch": 0.5939626224089735, + "grad_norm": 2.6608431339263916, + "learning_rate": 7.093518809941156e-05, + "loss": 0.2542, + "step": 25338 + }, + { + "epoch": 0.5939860639837784, + "grad_norm": 1.9945011138916016, + "learning_rate": 7.092814001738844e-05, + "loss": 0.2024, + "step": 25339 + }, + { + "epoch": 0.5940095055585835, + "grad_norm": 1.2155253887176514, + "learning_rate": 7.092109209311004e-05, + "loss": 0.1103, + "step": 25340 + }, + { + "epoch": 0.5940329471333884, + "grad_norm": 0.7441766262054443, + "learning_rate": 7.09140443266146e-05, + "loss": 0.0379, + "step": 25341 + }, + { + "epoch": 0.5940563887081934, + "grad_norm": 0.5531450510025024, + "learning_rate": 7.090699671794036e-05, + "loss": 0.0584, + "step": 25342 + }, + { + "epoch": 0.5940798302829984, + "grad_norm": 3.0753936767578125, + "learning_rate": 7.089994926712563e-05, + "loss": 0.4826, + "step": 25343 + }, + { + "epoch": 0.5941032718578034, + "grad_norm": 2.7308928966522217, + "learning_rate": 7.089290197420856e-05, + "loss": 0.2065, + "step": 25344 + }, + { + "epoch": 0.5941267134326084, + "grad_norm": 4.593348979949951, + "learning_rate": 7.088585483922742e-05, + "loss": 0.5341, + "step": 25345 + }, + { + "epoch": 0.5941501550074134, + "grad_norm": 0.9916401505470276, + "learning_rate": 7.087880786222047e-05, + "loss": 0.0956, + "step": 25346 + }, + { + "epoch": 0.5941735965822184, + "grad_norm": 2.4724841117858887, + "learning_rate": 7.087176104322592e-05, + "loss": 0.3262, + "step": 25347 + }, + { + "epoch": 0.5941970381570234, + "grad_norm": 4.900595188140869, + "learning_rate": 7.086471438228201e-05, + "loss": 0.8869, + "step": 25348 + }, + { + "epoch": 0.5942204797318283, + "grad_norm": 1.1072282791137695, + "learning_rate": 7.085766787942698e-05, + "loss": 0.0888, + "step": 25349 + }, + { + "epoch": 0.5942439213066334, + "grad_norm": 3.1983697414398193, + "learning_rate": 7.085062153469908e-05, + "loss": 0.53, + "step": 25350 + }, + { + "epoch": 0.5942673628814383, + "grad_norm": 2.254209280014038, + "learning_rate": 7.084357534813651e-05, + "loss": 0.1258, + "step": 25351 + }, + { + "epoch": 0.5942908044562434, + "grad_norm": 2.242316484451294, + "learning_rate": 7.083652931977751e-05, + "loss": 0.2345, + "step": 25352 + }, + { + "epoch": 0.5943142460310483, + "grad_norm": 2.705167770385742, + "learning_rate": 7.08294834496603e-05, + "loss": 0.2359, + "step": 25353 + }, + { + "epoch": 0.5943376876058534, + "grad_norm": 2.879927635192871, + "learning_rate": 7.082243773782318e-05, + "loss": 1.3653, + "step": 25354 + }, + { + "epoch": 0.5943611291806583, + "grad_norm": 2.305241823196411, + "learning_rate": 7.081539218430431e-05, + "loss": 0.1384, + "step": 25355 + }, + { + "epoch": 0.5943845707554634, + "grad_norm": 2.6126387119293213, + "learning_rate": 7.080834678914196e-05, + "loss": 0.2092, + "step": 25356 + }, + { + "epoch": 0.5944080123302683, + "grad_norm": 3.7581706047058105, + "learning_rate": 7.080130155237434e-05, + "loss": 0.3311, + "step": 25357 + }, + { + "epoch": 0.5944314539050733, + "grad_norm": 2.4790332317352295, + "learning_rate": 7.079425647403965e-05, + "loss": 0.2636, + "step": 25358 + }, + { + "epoch": 0.5944548954798783, + "grad_norm": 2.8683698177337646, + "learning_rate": 7.078721155417618e-05, + "loss": 1.3029, + "step": 25359 + }, + { + "epoch": 0.5944783370546833, + "grad_norm": 1.8204965591430664, + "learning_rate": 7.07801667928221e-05, + "loss": 0.0714, + "step": 25360 + }, + { + "epoch": 0.5945017786294883, + "grad_norm": 2.153553009033203, + "learning_rate": 7.077312219001567e-05, + "loss": 0.1174, + "step": 25361 + }, + { + "epoch": 0.5945252202042933, + "grad_norm": 3.3337018489837646, + "learning_rate": 7.07660777457951e-05, + "loss": 0.2739, + "step": 25362 + }, + { + "epoch": 0.5945486617790984, + "grad_norm": 1.3002985715866089, + "learning_rate": 7.07590334601986e-05, + "loss": 0.1441, + "step": 25363 + }, + { + "epoch": 0.5945721033539033, + "grad_norm": 2.402296304702759, + "learning_rate": 7.075198933326443e-05, + "loss": 1.3173, + "step": 25364 + }, + { + "epoch": 0.5945955449287084, + "grad_norm": 1.440144419670105, + "learning_rate": 7.074494536503075e-05, + "loss": 0.1913, + "step": 25365 + }, + { + "epoch": 0.5946189865035133, + "grad_norm": 1.8363932371139526, + "learning_rate": 7.073790155553584e-05, + "loss": 0.1357, + "step": 25366 + }, + { + "epoch": 0.5946424280783184, + "grad_norm": 1.694455862045288, + "learning_rate": 7.07308579048179e-05, + "loss": 0.1369, + "step": 25367 + }, + { + "epoch": 0.5946658696531233, + "grad_norm": 4.229879856109619, + "learning_rate": 7.072381441291515e-05, + "loss": 0.3479, + "step": 25368 + }, + { + "epoch": 0.5946893112279283, + "grad_norm": 0.5461647510528564, + "learning_rate": 7.071677107986583e-05, + "loss": 0.0322, + "step": 25369 + }, + { + "epoch": 0.5947127528027333, + "grad_norm": 2.2408628463745117, + "learning_rate": 7.070972790570808e-05, + "loss": 1.0314, + "step": 25370 + }, + { + "epoch": 0.5947361943775383, + "grad_norm": 0.8711369037628174, + "learning_rate": 7.070268489048021e-05, + "loss": 0.0739, + "step": 25371 + }, + { + "epoch": 0.5947596359523433, + "grad_norm": 0.8214542865753174, + "learning_rate": 7.069564203422041e-05, + "loss": 0.075, + "step": 25372 + }, + { + "epoch": 0.5947830775271483, + "grad_norm": 2.0031068325042725, + "learning_rate": 7.068859933696683e-05, + "loss": 0.1426, + "step": 25373 + }, + { + "epoch": 0.5948065191019533, + "grad_norm": 2.2778968811035156, + "learning_rate": 7.068155679875779e-05, + "loss": 0.091, + "step": 25374 + }, + { + "epoch": 0.5948299606767583, + "grad_norm": 0.19926129281520844, + "learning_rate": 7.067451441963143e-05, + "loss": 0.0135, + "step": 25375 + }, + { + "epoch": 0.5948534022515632, + "grad_norm": 2.8664679527282715, + "learning_rate": 7.066747219962594e-05, + "loss": 0.172, + "step": 25376 + }, + { + "epoch": 0.5948768438263683, + "grad_norm": 0.9685268402099609, + "learning_rate": 7.066043013877957e-05, + "loss": 0.1259, + "step": 25377 + }, + { + "epoch": 0.5949002854011732, + "grad_norm": 0.9343639612197876, + "learning_rate": 7.065338823713057e-05, + "loss": 0.5159, + "step": 25378 + }, + { + "epoch": 0.5949237269759783, + "grad_norm": 3.188450336456299, + "learning_rate": 7.06463464947171e-05, + "loss": 0.3025, + "step": 25379 + }, + { + "epoch": 0.5949471685507832, + "grad_norm": 2.408371925354004, + "learning_rate": 7.063930491157736e-05, + "loss": 0.1576, + "step": 25380 + }, + { + "epoch": 0.5949706101255883, + "grad_norm": 1.1171081066131592, + "learning_rate": 7.06322634877496e-05, + "loss": 0.123, + "step": 25381 + }, + { + "epoch": 0.5949940517003932, + "grad_norm": 3.313741683959961, + "learning_rate": 7.062522222327202e-05, + "loss": 0.2602, + "step": 25382 + }, + { + "epoch": 0.5950174932751983, + "grad_norm": 2.3308663368225098, + "learning_rate": 7.061818111818277e-05, + "loss": 0.3492, + "step": 25383 + }, + { + "epoch": 0.5950409348500032, + "grad_norm": 1.5987318754196167, + "learning_rate": 7.061114017252014e-05, + "loss": 0.1207, + "step": 25384 + }, + { + "epoch": 0.5950643764248083, + "grad_norm": 2.3653218746185303, + "learning_rate": 7.060409938632227e-05, + "loss": 0.2189, + "step": 25385 + }, + { + "epoch": 0.5950878179996132, + "grad_norm": 2.953862190246582, + "learning_rate": 7.059705875962736e-05, + "loss": 2.0703, + "step": 25386 + }, + { + "epoch": 0.5951112595744182, + "grad_norm": 3.4536311626434326, + "learning_rate": 7.059001829247367e-05, + "loss": 0.2259, + "step": 25387 + }, + { + "epoch": 0.5951347011492232, + "grad_norm": 3.030742883682251, + "learning_rate": 7.058297798489935e-05, + "loss": 0.1509, + "step": 25388 + }, + { + "epoch": 0.5951581427240282, + "grad_norm": 1.9925743341445923, + "learning_rate": 7.057593783694261e-05, + "loss": 0.1176, + "step": 25389 + }, + { + "epoch": 0.5951815842988332, + "grad_norm": 3.8687448501586914, + "learning_rate": 7.056889784864163e-05, + "loss": 0.3592, + "step": 25390 + }, + { + "epoch": 0.5952050258736382, + "grad_norm": 3.197582244873047, + "learning_rate": 7.05618580200347e-05, + "loss": 0.1753, + "step": 25391 + }, + { + "epoch": 0.5952284674484432, + "grad_norm": 2.9856009483337402, + "learning_rate": 7.055481835115994e-05, + "loss": 0.3127, + "step": 25392 + }, + { + "epoch": 0.5952519090232482, + "grad_norm": 1.180975079536438, + "learning_rate": 7.054777884205555e-05, + "loss": 0.0348, + "step": 25393 + }, + { + "epoch": 0.5952753505980531, + "grad_norm": 2.9577975273132324, + "learning_rate": 7.054073949275975e-05, + "loss": 0.2214, + "step": 25394 + }, + { + "epoch": 0.5952987921728582, + "grad_norm": 2.637601137161255, + "learning_rate": 7.053370030331073e-05, + "loss": 0.1385, + "step": 25395 + }, + { + "epoch": 0.5953222337476631, + "grad_norm": 2.518219470977783, + "learning_rate": 7.052666127374665e-05, + "loss": 0.2577, + "step": 25396 + }, + { + "epoch": 0.5953456753224682, + "grad_norm": 2.916308641433716, + "learning_rate": 7.051962240410577e-05, + "loss": 0.2447, + "step": 25397 + }, + { + "epoch": 0.5953691168972731, + "grad_norm": 4.799435615539551, + "learning_rate": 7.051258369442623e-05, + "loss": 0.5586, + "step": 25398 + }, + { + "epoch": 0.5953925584720782, + "grad_norm": 2.8371777534484863, + "learning_rate": 7.050554514474626e-05, + "loss": 0.2622, + "step": 25399 + }, + { + "epoch": 0.5954160000468831, + "grad_norm": 3.5834083557128906, + "learning_rate": 7.049850675510401e-05, + "loss": 0.1753, + "step": 25400 + }, + { + "epoch": 0.5954394416216882, + "grad_norm": 2.007642984390259, + "learning_rate": 7.04914685255377e-05, + "loss": 0.1994, + "step": 25401 + }, + { + "epoch": 0.5954628831964931, + "grad_norm": 2.896392583847046, + "learning_rate": 7.048443045608546e-05, + "loss": 0.3564, + "step": 25402 + }, + { + "epoch": 0.5954863247712981, + "grad_norm": 3.930812120437622, + "learning_rate": 7.04773925467856e-05, + "loss": 0.3816, + "step": 25403 + }, + { + "epoch": 0.5955097663461031, + "grad_norm": 0.6804596185684204, + "learning_rate": 7.047035479767623e-05, + "loss": 0.0689, + "step": 25404 + }, + { + "epoch": 0.5955332079209081, + "grad_norm": 0.695330798625946, + "learning_rate": 7.04633172087955e-05, + "loss": 0.0323, + "step": 25405 + }, + { + "epoch": 0.5955566494957131, + "grad_norm": 1.1764192581176758, + "learning_rate": 7.04562797801817e-05, + "loss": 0.0649, + "step": 25406 + }, + { + "epoch": 0.5955800910705181, + "grad_norm": 2.313065767288208, + "learning_rate": 7.044924251187294e-05, + "loss": 1.9689, + "step": 25407 + }, + { + "epoch": 0.5956035326453231, + "grad_norm": 1.1828017234802246, + "learning_rate": 7.04422054039074e-05, + "loss": 0.1078, + "step": 25408 + }, + { + "epoch": 0.5956269742201281, + "grad_norm": 0.9510473608970642, + "learning_rate": 7.043516845632331e-05, + "loss": 0.0804, + "step": 25409 + }, + { + "epoch": 0.595650415794933, + "grad_norm": 3.6329703330993652, + "learning_rate": 7.042813166915882e-05, + "loss": 0.4493, + "step": 25410 + }, + { + "epoch": 0.5956738573697381, + "grad_norm": 2.17718243598938, + "learning_rate": 7.042109504245211e-05, + "loss": 0.1537, + "step": 25411 + }, + { + "epoch": 0.595697298944543, + "grad_norm": 0.7007050514221191, + "learning_rate": 7.041405857624137e-05, + "loss": 0.0282, + "step": 25412 + }, + { + "epoch": 0.5957207405193481, + "grad_norm": 2.971390962600708, + "learning_rate": 7.040702227056479e-05, + "loss": 0.2241, + "step": 25413 + }, + { + "epoch": 0.595744182094153, + "grad_norm": 2.873745918273926, + "learning_rate": 7.03999861254605e-05, + "loss": 0.1769, + "step": 25414 + }, + { + "epoch": 0.5957676236689581, + "grad_norm": 6.6365742683410645, + "learning_rate": 7.039295014096675e-05, + "loss": 0.4718, + "step": 25415 + }, + { + "epoch": 0.5957910652437631, + "grad_norm": 1.738206386566162, + "learning_rate": 7.038591431712169e-05, + "loss": 0.111, + "step": 25416 + }, + { + "epoch": 0.5958145068185681, + "grad_norm": 3.538928508758545, + "learning_rate": 7.037887865396348e-05, + "loss": 0.327, + "step": 25417 + }, + { + "epoch": 0.5958379483933731, + "grad_norm": 0.35525262355804443, + "learning_rate": 7.037184315153032e-05, + "loss": 0.0379, + "step": 25418 + }, + { + "epoch": 0.5958613899681781, + "grad_norm": 3.816617012023926, + "learning_rate": 7.036480780986036e-05, + "loss": 0.4469, + "step": 25419 + }, + { + "epoch": 0.5958848315429831, + "grad_norm": 2.8881142139434814, + "learning_rate": 7.035777262899181e-05, + "loss": 0.3581, + "step": 25420 + }, + { + "epoch": 0.595908273117788, + "grad_norm": 4.261312007904053, + "learning_rate": 7.035073760896279e-05, + "loss": 0.1906, + "step": 25421 + }, + { + "epoch": 0.5959317146925931, + "grad_norm": 3.374697685241699, + "learning_rate": 7.034370274981152e-05, + "loss": 0.2691, + "step": 25422 + }, + { + "epoch": 0.595955156267398, + "grad_norm": 2.497903347015381, + "learning_rate": 7.033666805157616e-05, + "loss": 0.1932, + "step": 25423 + }, + { + "epoch": 0.5959785978422031, + "grad_norm": 1.553440809249878, + "learning_rate": 7.032963351429484e-05, + "loss": 0.1388, + "step": 25424 + }, + { + "epoch": 0.596002039417008, + "grad_norm": 2.8763537406921387, + "learning_rate": 7.032259913800579e-05, + "loss": 0.1661, + "step": 25425 + }, + { + "epoch": 0.5960254809918131, + "grad_norm": 1.5337938070297241, + "learning_rate": 7.031556492274711e-05, + "loss": 0.1095, + "step": 25426 + }, + { + "epoch": 0.596048922566618, + "grad_norm": 2.9163832664489746, + "learning_rate": 7.030853086855706e-05, + "loss": 0.1861, + "step": 25427 + }, + { + "epoch": 0.5960723641414231, + "grad_norm": 3.184964895248413, + "learning_rate": 7.030149697547373e-05, + "loss": 0.1538, + "step": 25428 + }, + { + "epoch": 0.596095805716228, + "grad_norm": 1.5138388872146606, + "learning_rate": 7.029446324353533e-05, + "loss": 0.1182, + "step": 25429 + }, + { + "epoch": 0.596119247291033, + "grad_norm": 2.3318746089935303, + "learning_rate": 7.028742967278001e-05, + "loss": 0.3067, + "step": 25430 + }, + { + "epoch": 0.596142688865838, + "grad_norm": 2.7061798572540283, + "learning_rate": 7.02803962632459e-05, + "loss": 0.2827, + "step": 25431 + }, + { + "epoch": 0.596166130440643, + "grad_norm": 2.9714808464050293, + "learning_rate": 7.027336301497123e-05, + "loss": 1.8067, + "step": 25432 + }, + { + "epoch": 0.596189572015448, + "grad_norm": 5.116937160491943, + "learning_rate": 7.026632992799411e-05, + "loss": 0.6251, + "step": 25433 + }, + { + "epoch": 0.596213013590253, + "grad_norm": 2.422585964202881, + "learning_rate": 7.025929700235272e-05, + "loss": 0.2559, + "step": 25434 + }, + { + "epoch": 0.596236455165058, + "grad_norm": 0.860004723072052, + "learning_rate": 7.025226423808525e-05, + "loss": 0.0708, + "step": 25435 + }, + { + "epoch": 0.596259896739863, + "grad_norm": 1.5006946325302124, + "learning_rate": 7.024523163522978e-05, + "loss": 0.0726, + "step": 25436 + }, + { + "epoch": 0.596283338314668, + "grad_norm": 0.5346481800079346, + "learning_rate": 7.023819919382455e-05, + "loss": 0.0436, + "step": 25437 + }, + { + "epoch": 0.596306779889473, + "grad_norm": 2.8136415481567383, + "learning_rate": 7.023116691390765e-05, + "loss": 1.9516, + "step": 25438 + }, + { + "epoch": 0.5963302214642779, + "grad_norm": 3.3851470947265625, + "learning_rate": 7.022413479551733e-05, + "loss": 0.4681, + "step": 25439 + }, + { + "epoch": 0.596353663039083, + "grad_norm": 3.009805917739868, + "learning_rate": 7.021710283869165e-05, + "loss": 0.3101, + "step": 25440 + }, + { + "epoch": 0.5963771046138879, + "grad_norm": 2.705209493637085, + "learning_rate": 7.021007104346884e-05, + "loss": 0.2954, + "step": 25441 + }, + { + "epoch": 0.596400546188693, + "grad_norm": 3.8594915866851807, + "learning_rate": 7.020303940988702e-05, + "loss": 0.3146, + "step": 25442 + }, + { + "epoch": 0.5964239877634979, + "grad_norm": 7.086894989013672, + "learning_rate": 7.019600793798433e-05, + "loss": 0.4861, + "step": 25443 + }, + { + "epoch": 0.596447429338303, + "grad_norm": 4.50667142868042, + "learning_rate": 7.018897662779895e-05, + "loss": 0.4514, + "step": 25444 + }, + { + "epoch": 0.5964708709131079, + "grad_norm": 2.0006368160247803, + "learning_rate": 7.018194547936903e-05, + "loss": 0.1798, + "step": 25445 + }, + { + "epoch": 0.596494312487913, + "grad_norm": 6.86552095413208, + "learning_rate": 7.017491449273268e-05, + "loss": 0.5932, + "step": 25446 + }, + { + "epoch": 0.5965177540627179, + "grad_norm": 2.7024056911468506, + "learning_rate": 7.016788366792812e-05, + "loss": 0.1873, + "step": 25447 + }, + { + "epoch": 0.596541195637523, + "grad_norm": 1.3938043117523193, + "learning_rate": 7.016085300499344e-05, + "loss": 0.1414, + "step": 25448 + }, + { + "epoch": 0.5965646372123279, + "grad_norm": 1.414553165435791, + "learning_rate": 7.015382250396681e-05, + "loss": 0.1581, + "step": 25449 + }, + { + "epoch": 0.5965880787871329, + "grad_norm": 2.9893505573272705, + "learning_rate": 7.014679216488634e-05, + "loss": 0.1965, + "step": 25450 + }, + { + "epoch": 0.5966115203619379, + "grad_norm": 3.246772527694702, + "learning_rate": 7.013976198779026e-05, + "loss": 0.3718, + "step": 25451 + }, + { + "epoch": 0.5966349619367429, + "grad_norm": 4.1682000160217285, + "learning_rate": 7.013273197271668e-05, + "loss": 0.2802, + "step": 25452 + }, + { + "epoch": 0.5966584035115479, + "grad_norm": 1.7447162866592407, + "learning_rate": 7.012570211970372e-05, + "loss": 0.1477, + "step": 25453 + }, + { + "epoch": 0.5966818450863529, + "grad_norm": 2.059699296951294, + "learning_rate": 7.011867242878955e-05, + "loss": 0.2308, + "step": 25454 + }, + { + "epoch": 0.5967052866611579, + "grad_norm": 1.6914113759994507, + "learning_rate": 7.011164290001231e-05, + "loss": 0.1502, + "step": 25455 + }, + { + "epoch": 0.5967287282359629, + "grad_norm": 0.7186012268066406, + "learning_rate": 7.010461353341012e-05, + "loss": 0.0192, + "step": 25456 + }, + { + "epoch": 0.5967521698107678, + "grad_norm": 4.938406467437744, + "learning_rate": 7.009758432902114e-05, + "loss": 0.5475, + "step": 25457 + }, + { + "epoch": 0.5967756113855729, + "grad_norm": 3.229309320449829, + "learning_rate": 7.009055528688353e-05, + "loss": 0.3008, + "step": 25458 + }, + { + "epoch": 0.5967990529603778, + "grad_norm": 3.197903633117676, + "learning_rate": 7.008352640703537e-05, + "loss": 0.2208, + "step": 25459 + }, + { + "epoch": 0.5968224945351829, + "grad_norm": 2.813444137573242, + "learning_rate": 7.007649768951487e-05, + "loss": 1.6709, + "step": 25460 + }, + { + "epoch": 0.5968459361099878, + "grad_norm": 0.9442526698112488, + "learning_rate": 7.006946913436013e-05, + "loss": 0.0431, + "step": 25461 + }, + { + "epoch": 0.5968693776847929, + "grad_norm": 2.10756254196167, + "learning_rate": 7.006244074160926e-05, + "loss": 0.0867, + "step": 25462 + }, + { + "epoch": 0.5968928192595978, + "grad_norm": 2.1500120162963867, + "learning_rate": 7.00554125113004e-05, + "loss": 0.2115, + "step": 25463 + }, + { + "epoch": 0.5969162608344029, + "grad_norm": 2.3916537761688232, + "learning_rate": 7.00483844434718e-05, + "loss": 0.1798, + "step": 25464 + }, + { + "epoch": 0.5969397024092078, + "grad_norm": 3.2225019931793213, + "learning_rate": 7.004135653816148e-05, + "loss": 1.6138, + "step": 25465 + }, + { + "epoch": 0.5969631439840128, + "grad_norm": 0.4479483962059021, + "learning_rate": 7.003432879540757e-05, + "loss": 0.0217, + "step": 25466 + }, + { + "epoch": 0.5969865855588179, + "grad_norm": 0.7284439206123352, + "learning_rate": 7.002730121524827e-05, + "loss": 0.0574, + "step": 25467 + }, + { + "epoch": 0.5970100271336228, + "grad_norm": 3.3401601314544678, + "learning_rate": 7.002027379772167e-05, + "loss": 0.3158, + "step": 25468 + }, + { + "epoch": 0.5970334687084279, + "grad_norm": 2.988457441329956, + "learning_rate": 7.00132465428659e-05, + "loss": 0.2108, + "step": 25469 + }, + { + "epoch": 0.5970569102832328, + "grad_norm": 1.0584216117858887, + "learning_rate": 7.00062194507191e-05, + "loss": 0.053, + "step": 25470 + }, + { + "epoch": 0.5970803518580379, + "grad_norm": 3.5373098850250244, + "learning_rate": 6.99991925213194e-05, + "loss": 0.3775, + "step": 25471 + }, + { + "epoch": 0.5971037934328428, + "grad_norm": 0.6409738659858704, + "learning_rate": 6.999216575470491e-05, + "loss": 0.0484, + "step": 25472 + }, + { + "epoch": 0.5971272350076479, + "grad_norm": 3.021650552749634, + "learning_rate": 6.99851391509138e-05, + "loss": 0.1975, + "step": 25473 + }, + { + "epoch": 0.5971506765824528, + "grad_norm": 3.601979970932007, + "learning_rate": 6.997811270998413e-05, + "loss": 0.3048, + "step": 25474 + }, + { + "epoch": 0.5971741181572578, + "grad_norm": 0.8005527853965759, + "learning_rate": 6.99710864319541e-05, + "loss": 0.0231, + "step": 25475 + }, + { + "epoch": 0.5971975597320628, + "grad_norm": 2.3109188079833984, + "learning_rate": 6.996406031686176e-05, + "loss": 0.2174, + "step": 25476 + }, + { + "epoch": 0.5972210013068678, + "grad_norm": 1.0297948122024536, + "learning_rate": 6.995703436474531e-05, + "loss": 0.0514, + "step": 25477 + }, + { + "epoch": 0.5972444428816728, + "grad_norm": 3.0896949768066406, + "learning_rate": 6.99500085756428e-05, + "loss": 0.2957, + "step": 25478 + }, + { + "epoch": 0.5972678844564778, + "grad_norm": 2.57562255859375, + "learning_rate": 6.994298294959242e-05, + "loss": 0.146, + "step": 25479 + }, + { + "epoch": 0.5972913260312828, + "grad_norm": 3.5535049438476562, + "learning_rate": 6.993595748663227e-05, + "loss": 0.3163, + "step": 25480 + }, + { + "epoch": 0.5973147676060878, + "grad_norm": 3.048323154449463, + "learning_rate": 6.992893218680043e-05, + "loss": 0.3114, + "step": 25481 + }, + { + "epoch": 0.5973382091808928, + "grad_norm": 2.618210792541504, + "learning_rate": 6.992190705013509e-05, + "loss": 0.3253, + "step": 25482 + }, + { + "epoch": 0.5973616507556978, + "grad_norm": 6.333125114440918, + "learning_rate": 6.99148820766743e-05, + "loss": 0.4262, + "step": 25483 + }, + { + "epoch": 0.5973850923305027, + "grad_norm": 2.0781660079956055, + "learning_rate": 6.99078572664562e-05, + "loss": 0.1165, + "step": 25484 + }, + { + "epoch": 0.5974085339053078, + "grad_norm": 2.778611660003662, + "learning_rate": 6.990083261951892e-05, + "loss": 0.2478, + "step": 25485 + }, + { + "epoch": 0.5974319754801127, + "grad_norm": 1.5962964296340942, + "learning_rate": 6.989380813590058e-05, + "loss": 0.1896, + "step": 25486 + }, + { + "epoch": 0.5974554170549178, + "grad_norm": 3.8417115211486816, + "learning_rate": 6.988678381563924e-05, + "loss": 0.4306, + "step": 25487 + }, + { + "epoch": 0.5974788586297227, + "grad_norm": 3.495758056640625, + "learning_rate": 6.987975965877309e-05, + "loss": 0.3798, + "step": 25488 + }, + { + "epoch": 0.5975023002045278, + "grad_norm": 0.8596968054771423, + "learning_rate": 6.987273566534021e-05, + "loss": 0.0802, + "step": 25489 + }, + { + "epoch": 0.5975257417793327, + "grad_norm": 2.4472949504852295, + "learning_rate": 6.986571183537872e-05, + "loss": 0.16, + "step": 25490 + }, + { + "epoch": 0.5975491833541378, + "grad_norm": 2.271057367324829, + "learning_rate": 6.985868816892672e-05, + "loss": 0.1757, + "step": 25491 + }, + { + "epoch": 0.5975726249289427, + "grad_norm": 4.076443195343018, + "learning_rate": 6.985166466602233e-05, + "loss": 0.528, + "step": 25492 + }, + { + "epoch": 0.5975960665037477, + "grad_norm": 2.2095510959625244, + "learning_rate": 6.984464132670365e-05, + "loss": 0.1554, + "step": 25493 + }, + { + "epoch": 0.5976195080785527, + "grad_norm": 1.7303071022033691, + "learning_rate": 6.983761815100878e-05, + "loss": 0.1313, + "step": 25494 + }, + { + "epoch": 0.5976429496533577, + "grad_norm": 1.847062110900879, + "learning_rate": 6.983059513897588e-05, + "loss": 0.155, + "step": 25495 + }, + { + "epoch": 0.5976663912281627, + "grad_norm": 2.201504707336426, + "learning_rate": 6.982357229064299e-05, + "loss": 0.1597, + "step": 25496 + }, + { + "epoch": 0.5976898328029677, + "grad_norm": 1.9885704517364502, + "learning_rate": 6.981654960604824e-05, + "loss": 1.0867, + "step": 25497 + }, + { + "epoch": 0.5977132743777727, + "grad_norm": 4.475475311279297, + "learning_rate": 6.980952708522976e-05, + "loss": 0.4842, + "step": 25498 + }, + { + "epoch": 0.5977367159525777, + "grad_norm": 2.177351713180542, + "learning_rate": 6.980250472822558e-05, + "loss": 0.1197, + "step": 25499 + }, + { + "epoch": 0.5977601575273827, + "grad_norm": 1.0567866563796997, + "learning_rate": 6.979548253507391e-05, + "loss": 0.4452, + "step": 25500 + }, + { + "epoch": 0.5977835991021877, + "grad_norm": 2.2440881729125977, + "learning_rate": 6.978846050581281e-05, + "loss": 0.1197, + "step": 25501 + }, + { + "epoch": 0.5978070406769926, + "grad_norm": 1.1394925117492676, + "learning_rate": 6.978143864048034e-05, + "loss": 0.0674, + "step": 25502 + }, + { + "epoch": 0.5978304822517977, + "grad_norm": 3.8024089336395264, + "learning_rate": 6.977441693911468e-05, + "loss": 0.5854, + "step": 25503 + }, + { + "epoch": 0.5978539238266026, + "grad_norm": 5.281238555908203, + "learning_rate": 6.976739540175383e-05, + "loss": 0.3581, + "step": 25504 + }, + { + "epoch": 0.5978773654014077, + "grad_norm": 0.9766117334365845, + "learning_rate": 6.976037402843599e-05, + "loss": 0.1088, + "step": 25505 + }, + { + "epoch": 0.5979008069762126, + "grad_norm": 1.453476905822754, + "learning_rate": 6.97533528191992e-05, + "loss": 0.9127, + "step": 25506 + }, + { + "epoch": 0.5979242485510177, + "grad_norm": 3.3326241970062256, + "learning_rate": 6.974633177408153e-05, + "loss": 1.2101, + "step": 25507 + }, + { + "epoch": 0.5979476901258226, + "grad_norm": 0.8370782732963562, + "learning_rate": 6.973931089312116e-05, + "loss": 0.0303, + "step": 25508 + }, + { + "epoch": 0.5979711317006277, + "grad_norm": 0.9375919103622437, + "learning_rate": 6.973229017635609e-05, + "loss": 0.074, + "step": 25509 + }, + { + "epoch": 0.5979945732754326, + "grad_norm": 4.8207902908325195, + "learning_rate": 6.97252696238245e-05, + "loss": 0.54, + "step": 25510 + }, + { + "epoch": 0.5980180148502376, + "grad_norm": 4.71159553527832, + "learning_rate": 6.971824923556442e-05, + "loss": 0.5527, + "step": 25511 + }, + { + "epoch": 0.5980414564250426, + "grad_norm": 3.2531635761260986, + "learning_rate": 6.9711229011614e-05, + "loss": 0.279, + "step": 25512 + }, + { + "epoch": 0.5980648979998476, + "grad_norm": 2.494399070739746, + "learning_rate": 6.97042089520113e-05, + "loss": 0.2772, + "step": 25513 + }, + { + "epoch": 0.5980883395746526, + "grad_norm": 2.754333972930908, + "learning_rate": 6.969718905679441e-05, + "loss": 0.187, + "step": 25514 + }, + { + "epoch": 0.5981117811494576, + "grad_norm": 0.9304050803184509, + "learning_rate": 6.969016932600143e-05, + "loss": 0.052, + "step": 25515 + }, + { + "epoch": 0.5981352227242626, + "grad_norm": 2.062574863433838, + "learning_rate": 6.968314975967043e-05, + "loss": 0.155, + "step": 25516 + }, + { + "epoch": 0.5981586642990676, + "grad_norm": 5.432929515838623, + "learning_rate": 6.967613035783953e-05, + "loss": 0.4996, + "step": 25517 + }, + { + "epoch": 0.5981821058738727, + "grad_norm": 4.3655500411987305, + "learning_rate": 6.96691111205468e-05, + "loss": 0.4615, + "step": 25518 + }, + { + "epoch": 0.5982055474486776, + "grad_norm": 3.5253748893737793, + "learning_rate": 6.966209204783031e-05, + "loss": 0.3696, + "step": 25519 + }, + { + "epoch": 0.5982289890234826, + "grad_norm": 1.4100147485733032, + "learning_rate": 6.965507313972818e-05, + "loss": 0.0607, + "step": 25520 + }, + { + "epoch": 0.5982524305982876, + "grad_norm": 1.382972240447998, + "learning_rate": 6.964805439627846e-05, + "loss": 0.0625, + "step": 25521 + }, + { + "epoch": 0.5982758721730926, + "grad_norm": 0.6051512360572815, + "learning_rate": 6.964103581751924e-05, + "loss": 0.0492, + "step": 25522 + }, + { + "epoch": 0.5982993137478976, + "grad_norm": 3.401864767074585, + "learning_rate": 6.963401740348861e-05, + "loss": 0.304, + "step": 25523 + }, + { + "epoch": 0.5983227553227026, + "grad_norm": 2.59074330329895, + "learning_rate": 6.962699915422468e-05, + "loss": 0.311, + "step": 25524 + }, + { + "epoch": 0.5983461968975076, + "grad_norm": 2.240488290786743, + "learning_rate": 6.96199810697655e-05, + "loss": 0.1807, + "step": 25525 + }, + { + "epoch": 0.5983696384723126, + "grad_norm": 3.1804425716400146, + "learning_rate": 6.961296315014915e-05, + "loss": 0.3402, + "step": 25526 + }, + { + "epoch": 0.5983930800471176, + "grad_norm": 0.9402193427085876, + "learning_rate": 6.960594539541374e-05, + "loss": 0.0549, + "step": 25527 + }, + { + "epoch": 0.5984165216219226, + "grad_norm": 1.743398904800415, + "learning_rate": 6.959892780559732e-05, + "loss": 0.1821, + "step": 25528 + }, + { + "epoch": 0.5984399631967275, + "grad_norm": 1.7728955745697021, + "learning_rate": 6.959191038073794e-05, + "loss": 1.7154, + "step": 25529 + }, + { + "epoch": 0.5984634047715326, + "grad_norm": 1.3923842906951904, + "learning_rate": 6.958489312087374e-05, + "loss": 0.1094, + "step": 25530 + }, + { + "epoch": 0.5984868463463375, + "grad_norm": 1.8902921676635742, + "learning_rate": 6.957787602604275e-05, + "loss": 1.2278, + "step": 25531 + }, + { + "epoch": 0.5985102879211426, + "grad_norm": 3.5068044662475586, + "learning_rate": 6.957085909628307e-05, + "loss": 0.2632, + "step": 25532 + }, + { + "epoch": 0.5985337294959475, + "grad_norm": 3.7490127086639404, + "learning_rate": 6.956384233163276e-05, + "loss": 0.2692, + "step": 25533 + }, + { + "epoch": 0.5985571710707526, + "grad_norm": 2.796682119369507, + "learning_rate": 6.955682573212991e-05, + "loss": 1.697, + "step": 25534 + }, + { + "epoch": 0.5985806126455575, + "grad_norm": 3.632838010787964, + "learning_rate": 6.954980929781253e-05, + "loss": 0.4388, + "step": 25535 + }, + { + "epoch": 0.5986040542203626, + "grad_norm": 0.8438617587089539, + "learning_rate": 6.954279302871876e-05, + "loss": 0.0528, + "step": 25536 + }, + { + "epoch": 0.5986274957951675, + "grad_norm": 0.45759981870651245, + "learning_rate": 6.953577692488668e-05, + "loss": 0.056, + "step": 25537 + }, + { + "epoch": 0.5986509373699725, + "grad_norm": 0.8025227785110474, + "learning_rate": 6.952876098635432e-05, + "loss": 0.0791, + "step": 25538 + }, + { + "epoch": 0.5986743789447775, + "grad_norm": 3.0823848247528076, + "learning_rate": 6.952174521315975e-05, + "loss": 0.2217, + "step": 25539 + }, + { + "epoch": 0.5986978205195825, + "grad_norm": 1.7645692825317383, + "learning_rate": 6.951472960534107e-05, + "loss": 0.1095, + "step": 25540 + }, + { + "epoch": 0.5987212620943875, + "grad_norm": 2.4706718921661377, + "learning_rate": 6.950771416293632e-05, + "loss": 0.2318, + "step": 25541 + }, + { + "epoch": 0.5987447036691925, + "grad_norm": 4.201397895812988, + "learning_rate": 6.950069888598355e-05, + "loss": 0.2521, + "step": 25542 + }, + { + "epoch": 0.5987681452439975, + "grad_norm": 2.374467611312866, + "learning_rate": 6.949368377452087e-05, + "loss": 0.2285, + "step": 25543 + }, + { + "epoch": 0.5987915868188025, + "grad_norm": 2.4852280616760254, + "learning_rate": 6.948666882858632e-05, + "loss": 0.2425, + "step": 25544 + }, + { + "epoch": 0.5988150283936075, + "grad_norm": 3.322503089904785, + "learning_rate": 6.947965404821792e-05, + "loss": 0.2254, + "step": 25545 + }, + { + "epoch": 0.5988384699684125, + "grad_norm": 0.4384288787841797, + "learning_rate": 6.947263943345382e-05, + "loss": 0.0355, + "step": 25546 + }, + { + "epoch": 0.5988619115432174, + "grad_norm": 2.560023784637451, + "learning_rate": 6.946562498433201e-05, + "loss": 0.2376, + "step": 25547 + }, + { + "epoch": 0.5988853531180225, + "grad_norm": 3.2110352516174316, + "learning_rate": 6.945861070089059e-05, + "loss": 0.3746, + "step": 25548 + }, + { + "epoch": 0.5989087946928274, + "grad_norm": 3.471169948577881, + "learning_rate": 6.945159658316757e-05, + "loss": 0.2873, + "step": 25549 + }, + { + "epoch": 0.5989322362676325, + "grad_norm": 3.424177885055542, + "learning_rate": 6.94445826312011e-05, + "loss": 0.1914, + "step": 25550 + }, + { + "epoch": 0.5989556778424374, + "grad_norm": 3.0157506465911865, + "learning_rate": 6.943756884502918e-05, + "loss": 0.1451, + "step": 25551 + }, + { + "epoch": 0.5989791194172425, + "grad_norm": 2.8037478923797607, + "learning_rate": 6.943055522468984e-05, + "loss": 0.2616, + "step": 25552 + }, + { + "epoch": 0.5990025609920474, + "grad_norm": 1.4184553623199463, + "learning_rate": 6.942354177022119e-05, + "loss": 0.6629, + "step": 25553 + }, + { + "epoch": 0.5990260025668525, + "grad_norm": 2.464887857437134, + "learning_rate": 6.941652848166125e-05, + "loss": 0.1876, + "step": 25554 + }, + { + "epoch": 0.5990494441416574, + "grad_norm": 0.5991261601448059, + "learning_rate": 6.940951535904809e-05, + "loss": 0.0241, + "step": 25555 + }, + { + "epoch": 0.5990728857164624, + "grad_norm": 1.8252869844436646, + "learning_rate": 6.940250240241978e-05, + "loss": 0.1697, + "step": 25556 + }, + { + "epoch": 0.5990963272912674, + "grad_norm": 3.7705085277557373, + "learning_rate": 6.939548961181433e-05, + "loss": 0.4502, + "step": 25557 + }, + { + "epoch": 0.5991197688660724, + "grad_norm": 2.146192789077759, + "learning_rate": 6.938847698726981e-05, + "loss": 0.1775, + "step": 25558 + }, + { + "epoch": 0.5991432104408774, + "grad_norm": 3.5406172275543213, + "learning_rate": 6.93814645288243e-05, + "loss": 0.1389, + "step": 25559 + }, + { + "epoch": 0.5991666520156824, + "grad_norm": 3.169621229171753, + "learning_rate": 6.937445223651578e-05, + "loss": 0.4615, + "step": 25560 + }, + { + "epoch": 0.5991900935904874, + "grad_norm": 4.2451324462890625, + "learning_rate": 6.936744011038234e-05, + "loss": 0.4535, + "step": 25561 + }, + { + "epoch": 0.5992135351652924, + "grad_norm": 1.636709451675415, + "learning_rate": 6.936042815046208e-05, + "loss": 0.1309, + "step": 25562 + }, + { + "epoch": 0.5992369767400973, + "grad_norm": 1.9011399745941162, + "learning_rate": 6.935341635679298e-05, + "loss": 0.2449, + "step": 25563 + }, + { + "epoch": 0.5992604183149024, + "grad_norm": 1.2558270692825317, + "learning_rate": 6.93464047294131e-05, + "loss": 0.1845, + "step": 25564 + }, + { + "epoch": 0.5992838598897073, + "grad_norm": 0.3584883511066437, + "learning_rate": 6.93393932683605e-05, + "loss": 0.0371, + "step": 25565 + }, + { + "epoch": 0.5993073014645124, + "grad_norm": 1.5884521007537842, + "learning_rate": 6.933238197367322e-05, + "loss": 0.0664, + "step": 25566 + }, + { + "epoch": 0.5993307430393173, + "grad_norm": 4.0788421630859375, + "learning_rate": 6.932537084538926e-05, + "loss": 0.4798, + "step": 25567 + }, + { + "epoch": 0.5993541846141224, + "grad_norm": 3.1586477756500244, + "learning_rate": 6.931835988354674e-05, + "loss": 0.3928, + "step": 25568 + }, + { + "epoch": 0.5993776261889274, + "grad_norm": 2.5712108612060547, + "learning_rate": 6.931134908818367e-05, + "loss": 0.1597, + "step": 25569 + }, + { + "epoch": 0.5994010677637324, + "grad_norm": 0.4963364601135254, + "learning_rate": 6.930433845933804e-05, + "loss": 0.0352, + "step": 25570 + }, + { + "epoch": 0.5994245093385374, + "grad_norm": 0.533673882484436, + "learning_rate": 6.929732799704796e-05, + "loss": 0.0246, + "step": 25571 + }, + { + "epoch": 0.5994479509133424, + "grad_norm": 2.4075942039489746, + "learning_rate": 6.929031770135139e-05, + "loss": 0.1096, + "step": 25572 + }, + { + "epoch": 0.5994713924881474, + "grad_norm": 4.166572093963623, + "learning_rate": 6.928330757228648e-05, + "loss": 0.3253, + "step": 25573 + }, + { + "epoch": 0.5994948340629523, + "grad_norm": 0.36897844076156616, + "learning_rate": 6.927629760989118e-05, + "loss": 0.0363, + "step": 25574 + }, + { + "epoch": 0.5995182756377574, + "grad_norm": 5.64450216293335, + "learning_rate": 6.926928781420357e-05, + "loss": 0.4744, + "step": 25575 + }, + { + "epoch": 0.5995417172125623, + "grad_norm": 3.70638370513916, + "learning_rate": 6.926227818526168e-05, + "loss": 0.348, + "step": 25576 + }, + { + "epoch": 0.5995651587873674, + "grad_norm": 1.8107647895812988, + "learning_rate": 6.92552687231035e-05, + "loss": 1.13, + "step": 25577 + }, + { + "epoch": 0.5995886003621723, + "grad_norm": 3.2674155235290527, + "learning_rate": 6.924825942776711e-05, + "loss": 0.4491, + "step": 25578 + }, + { + "epoch": 0.5996120419369774, + "grad_norm": 1.8295443058013916, + "learning_rate": 6.924125029929055e-05, + "loss": 0.1438, + "step": 25579 + }, + { + "epoch": 0.5996354835117823, + "grad_norm": 5.539737224578857, + "learning_rate": 6.92342413377118e-05, + "loss": 0.2599, + "step": 25580 + }, + { + "epoch": 0.5996589250865874, + "grad_norm": 3.5195698738098145, + "learning_rate": 6.922723254306894e-05, + "loss": 0.2054, + "step": 25581 + }, + { + "epoch": 0.5996823666613923, + "grad_norm": 2.8540291786193848, + "learning_rate": 6.922022391539998e-05, + "loss": 0.3394, + "step": 25582 + }, + { + "epoch": 0.5997058082361973, + "grad_norm": 1.8456428050994873, + "learning_rate": 6.921321545474294e-05, + "loss": 0.1486, + "step": 25583 + }, + { + "epoch": 0.5997292498110023, + "grad_norm": 0.6275972127914429, + "learning_rate": 6.920620716113582e-05, + "loss": 0.0531, + "step": 25584 + }, + { + "epoch": 0.5997526913858073, + "grad_norm": 4.233526229858398, + "learning_rate": 6.919919903461675e-05, + "loss": 0.4307, + "step": 25585 + }, + { + "epoch": 0.5997761329606123, + "grad_norm": 1.4432779550552368, + "learning_rate": 6.919219107522367e-05, + "loss": 0.0919, + "step": 25586 + }, + { + "epoch": 0.5997995745354173, + "grad_norm": 2.3351869583129883, + "learning_rate": 6.918518328299463e-05, + "loss": 0.2404, + "step": 25587 + }, + { + "epoch": 0.5998230161102223, + "grad_norm": 0.9669783711433411, + "learning_rate": 6.917817565796764e-05, + "loss": 0.0564, + "step": 25588 + }, + { + "epoch": 0.5998464576850273, + "grad_norm": 4.68617582321167, + "learning_rate": 6.917116820018075e-05, + "loss": 0.4764, + "step": 25589 + }, + { + "epoch": 0.5998698992598323, + "grad_norm": 1.0702701807022095, + "learning_rate": 6.916416090967197e-05, + "loss": 0.0706, + "step": 25590 + }, + { + "epoch": 0.5998933408346373, + "grad_norm": 6.903135776519775, + "learning_rate": 6.915715378647933e-05, + "loss": 0.3638, + "step": 25591 + }, + { + "epoch": 0.5999167824094422, + "grad_norm": 2.997474431991577, + "learning_rate": 6.91501468306408e-05, + "loss": 0.1991, + "step": 25592 + }, + { + "epoch": 0.5999402239842473, + "grad_norm": 4.900352954864502, + "learning_rate": 6.914314004219449e-05, + "loss": 0.4476, + "step": 25593 + }, + { + "epoch": 0.5999636655590522, + "grad_norm": 3.970970392227173, + "learning_rate": 6.913613342117834e-05, + "loss": 0.4727, + "step": 25594 + }, + { + "epoch": 0.5999871071338573, + "grad_norm": 3.690424680709839, + "learning_rate": 6.912912696763041e-05, + "loss": 0.4109, + "step": 25595 + }, + { + "epoch": 0.6000105487086622, + "grad_norm": 1.4905478954315186, + "learning_rate": 6.912212068158866e-05, + "loss": 0.0699, + "step": 25596 + }, + { + "epoch": 0.6000339902834673, + "grad_norm": 3.214345932006836, + "learning_rate": 6.911511456309118e-05, + "loss": 0.2424, + "step": 25597 + }, + { + "epoch": 0.6000574318582722, + "grad_norm": 4.139660835266113, + "learning_rate": 6.910810861217599e-05, + "loss": 0.4587, + "step": 25598 + }, + { + "epoch": 0.6000808734330773, + "grad_norm": 1.20802903175354, + "learning_rate": 6.910110282888103e-05, + "loss": 0.1396, + "step": 25599 + }, + { + "epoch": 0.6001043150078822, + "grad_norm": 2.5529861450195312, + "learning_rate": 6.909409721324438e-05, + "loss": 0.2078, + "step": 25600 + }, + { + "epoch": 0.6001277565826872, + "grad_norm": 1.377916693687439, + "learning_rate": 6.908709176530404e-05, + "loss": 0.1628, + "step": 25601 + }, + { + "epoch": 0.6001511981574922, + "grad_norm": 3.2416839599609375, + "learning_rate": 6.908008648509798e-05, + "loss": 0.3491, + "step": 25602 + }, + { + "epoch": 0.6001746397322972, + "grad_norm": 2.1636171340942383, + "learning_rate": 6.907308137266426e-05, + "loss": 0.1522, + "step": 25603 + }, + { + "epoch": 0.6001980813071022, + "grad_norm": 2.6356923580169678, + "learning_rate": 6.906607642804086e-05, + "loss": 0.2243, + "step": 25604 + }, + { + "epoch": 0.6002215228819072, + "grad_norm": 1.0034536123275757, + "learning_rate": 6.905907165126579e-05, + "loss": 0.0793, + "step": 25605 + }, + { + "epoch": 0.6002449644567122, + "grad_norm": 1.7425850629806519, + "learning_rate": 6.905206704237709e-05, + "loss": 0.0636, + "step": 25606 + }, + { + "epoch": 0.6002684060315172, + "grad_norm": 2.1043293476104736, + "learning_rate": 6.904506260141274e-05, + "loss": 0.1663, + "step": 25607 + }, + { + "epoch": 0.6002918476063221, + "grad_norm": 1.4744495153427124, + "learning_rate": 6.90380583284107e-05, + "loss": 0.1362, + "step": 25608 + }, + { + "epoch": 0.6003152891811272, + "grad_norm": 2.281688928604126, + "learning_rate": 6.903105422340906e-05, + "loss": 1.4191, + "step": 25609 + }, + { + "epoch": 0.6003387307559321, + "grad_norm": 3.369974374771118, + "learning_rate": 6.90240502864458e-05, + "loss": 0.3263, + "step": 25610 + }, + { + "epoch": 0.6003621723307372, + "grad_norm": 3.653191566467285, + "learning_rate": 6.901704651755892e-05, + "loss": 0.3254, + "step": 25611 + }, + { + "epoch": 0.6003856139055421, + "grad_norm": 1.729611873626709, + "learning_rate": 6.901004291678641e-05, + "loss": 0.1272, + "step": 25612 + }, + { + "epoch": 0.6004090554803472, + "grad_norm": 1.5536309480667114, + "learning_rate": 6.900303948416629e-05, + "loss": 0.1482, + "step": 25613 + }, + { + "epoch": 0.6004324970551521, + "grad_norm": 3.3401265144348145, + "learning_rate": 6.899603621973653e-05, + "loss": 0.3516, + "step": 25614 + }, + { + "epoch": 0.6004559386299572, + "grad_norm": 2.2756824493408203, + "learning_rate": 6.898903312353515e-05, + "loss": 0.1768, + "step": 25615 + }, + { + "epoch": 0.6004793802047621, + "grad_norm": 0.7332435250282288, + "learning_rate": 6.898203019560017e-05, + "loss": 0.0475, + "step": 25616 + }, + { + "epoch": 0.6005028217795672, + "grad_norm": 3.6833815574645996, + "learning_rate": 6.897502743596955e-05, + "loss": 0.4236, + "step": 25617 + }, + { + "epoch": 0.6005262633543721, + "grad_norm": 1.0553711652755737, + "learning_rate": 6.89680248446813e-05, + "loss": 0.0859, + "step": 25618 + }, + { + "epoch": 0.6005497049291771, + "grad_norm": 2.425276517868042, + "learning_rate": 6.896102242177343e-05, + "loss": 0.2418, + "step": 25619 + }, + { + "epoch": 0.6005731465039822, + "grad_norm": 1.1837716102600098, + "learning_rate": 6.895402016728393e-05, + "loss": 0.0956, + "step": 25620 + }, + { + "epoch": 0.6005965880787871, + "grad_norm": 2.379761219024658, + "learning_rate": 6.894701808125075e-05, + "loss": 0.2019, + "step": 25621 + }, + { + "epoch": 0.6006200296535922, + "grad_norm": 1.3505775928497314, + "learning_rate": 6.894001616371193e-05, + "loss": 0.1154, + "step": 25622 + }, + { + "epoch": 0.6006434712283971, + "grad_norm": 2.890852689743042, + "learning_rate": 6.893301441470548e-05, + "loss": 0.2277, + "step": 25623 + }, + { + "epoch": 0.6006669128032022, + "grad_norm": 3.396543502807617, + "learning_rate": 6.892601283426938e-05, + "loss": 0.304, + "step": 25624 + }, + { + "epoch": 0.6006903543780071, + "grad_norm": 2.245088577270508, + "learning_rate": 6.891901142244159e-05, + "loss": 1.5082, + "step": 25625 + }, + { + "epoch": 0.6007137959528122, + "grad_norm": 0.6511101126670837, + "learning_rate": 6.891201017926012e-05, + "loss": 0.0671, + "step": 25626 + }, + { + "epoch": 0.6007372375276171, + "grad_norm": 1.6364116668701172, + "learning_rate": 6.890500910476294e-05, + "loss": 0.0997, + "step": 25627 + }, + { + "epoch": 0.6007606791024221, + "grad_norm": 3.467838764190674, + "learning_rate": 6.889800819898808e-05, + "loss": 0.2447, + "step": 25628 + }, + { + "epoch": 0.6007841206772271, + "grad_norm": 3.770970582962036, + "learning_rate": 6.88910074619735e-05, + "loss": 0.3178, + "step": 25629 + }, + { + "epoch": 0.6008075622520321, + "grad_norm": 0.8752634525299072, + "learning_rate": 6.888400689375718e-05, + "loss": 0.0745, + "step": 25630 + }, + { + "epoch": 0.6008310038268371, + "grad_norm": 2.1732516288757324, + "learning_rate": 6.887700649437711e-05, + "loss": 0.4926, + "step": 25631 + }, + { + "epoch": 0.6008544454016421, + "grad_norm": 1.0985088348388672, + "learning_rate": 6.88700062638713e-05, + "loss": 0.0875, + "step": 25632 + }, + { + "epoch": 0.6008778869764471, + "grad_norm": 2.0539145469665527, + "learning_rate": 6.886300620227767e-05, + "loss": 1.5315, + "step": 25633 + }, + { + "epoch": 0.6009013285512521, + "grad_norm": 1.0649884939193726, + "learning_rate": 6.885600630963424e-05, + "loss": 0.0885, + "step": 25634 + }, + { + "epoch": 0.600924770126057, + "grad_norm": 5.023349761962891, + "learning_rate": 6.884900658597905e-05, + "loss": 0.4425, + "step": 25635 + }, + { + "epoch": 0.6009482117008621, + "grad_norm": 3.938697576522827, + "learning_rate": 6.884200703135e-05, + "loss": 0.1882, + "step": 25636 + }, + { + "epoch": 0.600971653275667, + "grad_norm": 2.0486319065093994, + "learning_rate": 6.883500764578509e-05, + "loss": 0.2892, + "step": 25637 + }, + { + "epoch": 0.6009950948504721, + "grad_norm": 1.9676294326782227, + "learning_rate": 6.882800842932232e-05, + "loss": 0.1895, + "step": 25638 + }, + { + "epoch": 0.601018536425277, + "grad_norm": 5.5416717529296875, + "learning_rate": 6.882100938199965e-05, + "loss": 0.5016, + "step": 25639 + }, + { + "epoch": 0.6010419780000821, + "grad_norm": 3.476200819015503, + "learning_rate": 6.881401050385505e-05, + "loss": 0.2669, + "step": 25640 + }, + { + "epoch": 0.601065419574887, + "grad_norm": 3.443645715713501, + "learning_rate": 6.88070117949265e-05, + "loss": 0.3513, + "step": 25641 + }, + { + "epoch": 0.6010888611496921, + "grad_norm": 1.6610846519470215, + "learning_rate": 6.8800013255252e-05, + "loss": 0.1504, + "step": 25642 + }, + { + "epoch": 0.601112302724497, + "grad_norm": 2.3939287662506104, + "learning_rate": 6.879301488486949e-05, + "loss": 0.8751, + "step": 25643 + }, + { + "epoch": 0.601135744299302, + "grad_norm": 2.9121506214141846, + "learning_rate": 6.878601668381698e-05, + "loss": 0.2348, + "step": 25644 + }, + { + "epoch": 0.601159185874107, + "grad_norm": 4.081214904785156, + "learning_rate": 6.877901865213237e-05, + "loss": 0.4636, + "step": 25645 + }, + { + "epoch": 0.601182627448912, + "grad_norm": 3.565709352493286, + "learning_rate": 6.877202078985374e-05, + "loss": 0.323, + "step": 25646 + }, + { + "epoch": 0.601206069023717, + "grad_norm": 2.638723611831665, + "learning_rate": 6.876502309701895e-05, + "loss": 0.1985, + "step": 25647 + }, + { + "epoch": 0.601229510598522, + "grad_norm": 3.230562210083008, + "learning_rate": 6.875802557366608e-05, + "loss": 0.3284, + "step": 25648 + }, + { + "epoch": 0.601252952173327, + "grad_norm": 2.990722179412842, + "learning_rate": 6.875102821983301e-05, + "loss": 0.2057, + "step": 25649 + }, + { + "epoch": 0.601276393748132, + "grad_norm": 4.258155345916748, + "learning_rate": 6.874403103555774e-05, + "loss": 0.4252, + "step": 25650 + }, + { + "epoch": 0.601299835322937, + "grad_norm": 2.9627041816711426, + "learning_rate": 6.873703402087825e-05, + "loss": 1.5749, + "step": 25651 + }, + { + "epoch": 0.601323276897742, + "grad_norm": 2.2372219562530518, + "learning_rate": 6.87300371758325e-05, + "loss": 0.1956, + "step": 25652 + }, + { + "epoch": 0.601346718472547, + "grad_norm": 3.4516639709472656, + "learning_rate": 6.872304050045841e-05, + "loss": 0.2265, + "step": 25653 + }, + { + "epoch": 0.601370160047352, + "grad_norm": 1.8409053087234497, + "learning_rate": 6.871604399479402e-05, + "loss": 0.1192, + "step": 25654 + }, + { + "epoch": 0.6013936016221569, + "grad_norm": 3.132458448410034, + "learning_rate": 6.870904765887725e-05, + "loss": 0.3278, + "step": 25655 + }, + { + "epoch": 0.601417043196962, + "grad_norm": 2.071653366088867, + "learning_rate": 6.870205149274604e-05, + "loss": 1.4065, + "step": 25656 + }, + { + "epoch": 0.6014404847717669, + "grad_norm": 3.065988302230835, + "learning_rate": 6.869505549643837e-05, + "loss": 0.1977, + "step": 25657 + }, + { + "epoch": 0.601463926346572, + "grad_norm": 3.535595655441284, + "learning_rate": 6.868805966999223e-05, + "loss": 0.4353, + "step": 25658 + }, + { + "epoch": 0.6014873679213769, + "grad_norm": 0.6740946769714355, + "learning_rate": 6.868106401344559e-05, + "loss": 0.0688, + "step": 25659 + }, + { + "epoch": 0.601510809496182, + "grad_norm": 3.246727466583252, + "learning_rate": 6.867406852683633e-05, + "loss": 0.1863, + "step": 25660 + }, + { + "epoch": 0.6015342510709869, + "grad_norm": 3.1198062896728516, + "learning_rate": 6.866707321020248e-05, + "loss": 0.1758, + "step": 25661 + }, + { + "epoch": 0.601557692645792, + "grad_norm": 1.727034091949463, + "learning_rate": 6.866007806358196e-05, + "loss": 0.1419, + "step": 25662 + }, + { + "epoch": 0.6015811342205969, + "grad_norm": 1.4441403150558472, + "learning_rate": 6.865308308701274e-05, + "loss": 0.2248, + "step": 25663 + }, + { + "epoch": 0.6016045757954019, + "grad_norm": 1.0790969133377075, + "learning_rate": 6.864608828053279e-05, + "loss": 0.0765, + "step": 25664 + }, + { + "epoch": 0.6016280173702069, + "grad_norm": 1.5457570552825928, + "learning_rate": 6.863909364418002e-05, + "loss": 0.1445, + "step": 25665 + }, + { + "epoch": 0.6016514589450119, + "grad_norm": 2.9649412631988525, + "learning_rate": 6.863209917799244e-05, + "loss": 1.8767, + "step": 25666 + }, + { + "epoch": 0.6016749005198169, + "grad_norm": 0.7679578065872192, + "learning_rate": 6.862510488200795e-05, + "loss": 0.0526, + "step": 25667 + }, + { + "epoch": 0.6016983420946219, + "grad_norm": 1.975937843322754, + "learning_rate": 6.861811075626452e-05, + "loss": 0.0686, + "step": 25668 + }, + { + "epoch": 0.6017217836694269, + "grad_norm": 4.386346340179443, + "learning_rate": 6.86111168008001e-05, + "loss": 0.4306, + "step": 25669 + }, + { + "epoch": 0.6017452252442319, + "grad_norm": 2.2522244453430176, + "learning_rate": 6.860412301565265e-05, + "loss": 0.2061, + "step": 25670 + }, + { + "epoch": 0.601768666819037, + "grad_norm": 1.1684311628341675, + "learning_rate": 6.859712940086011e-05, + "loss": 0.1473, + "step": 25671 + }, + { + "epoch": 0.6017921083938419, + "grad_norm": 3.629225254058838, + "learning_rate": 6.859013595646044e-05, + "loss": 0.2792, + "step": 25672 + }, + { + "epoch": 0.601815549968647, + "grad_norm": 2.8782589435577393, + "learning_rate": 6.858314268249157e-05, + "loss": 0.2051, + "step": 25673 + }, + { + "epoch": 0.6018389915434519, + "grad_norm": 4.92269229888916, + "learning_rate": 6.857614957899146e-05, + "loss": 0.683, + "step": 25674 + }, + { + "epoch": 0.6018624331182569, + "grad_norm": 3.2008168697357178, + "learning_rate": 6.856915664599803e-05, + "loss": 0.2125, + "step": 25675 + }, + { + "epoch": 0.6018858746930619, + "grad_norm": 1.3778692483901978, + "learning_rate": 6.856216388354925e-05, + "loss": 0.7345, + "step": 25676 + }, + { + "epoch": 0.6019093162678669, + "grad_norm": 1.4284621477127075, + "learning_rate": 6.855517129168305e-05, + "loss": 0.0869, + "step": 25677 + }, + { + "epoch": 0.6019327578426719, + "grad_norm": 6.105733394622803, + "learning_rate": 6.854817887043736e-05, + "loss": 0.4306, + "step": 25678 + }, + { + "epoch": 0.6019561994174769, + "grad_norm": 0.71174556016922, + "learning_rate": 6.854118661985017e-05, + "loss": 0.1086, + "step": 25679 + }, + { + "epoch": 0.6019796409922819, + "grad_norm": 1.701919674873352, + "learning_rate": 6.853419453995937e-05, + "loss": 0.1773, + "step": 25680 + }, + { + "epoch": 0.6020030825670869, + "grad_norm": 2.382399320602417, + "learning_rate": 6.852720263080287e-05, + "loss": 0.2224, + "step": 25681 + }, + { + "epoch": 0.6020265241418918, + "grad_norm": 1.8205097913742065, + "learning_rate": 6.85202108924187e-05, + "loss": 0.1572, + "step": 25682 + }, + { + "epoch": 0.6020499657166969, + "grad_norm": 0.46218782663345337, + "learning_rate": 6.851321932484474e-05, + "loss": 0.0316, + "step": 25683 + }, + { + "epoch": 0.6020734072915018, + "grad_norm": 0.7728284001350403, + "learning_rate": 6.850622792811896e-05, + "loss": 0.0444, + "step": 25684 + }, + { + "epoch": 0.6020968488663069, + "grad_norm": 0.8587519526481628, + "learning_rate": 6.849923670227925e-05, + "loss": 0.0866, + "step": 25685 + }, + { + "epoch": 0.6021202904411118, + "grad_norm": 3.216634750366211, + "learning_rate": 6.849224564736357e-05, + "loss": 0.2954, + "step": 25686 + }, + { + "epoch": 0.6021437320159169, + "grad_norm": 2.5271778106689453, + "learning_rate": 6.848525476340986e-05, + "loss": 0.1619, + "step": 25687 + }, + { + "epoch": 0.6021671735907218, + "grad_norm": 0.7010467648506165, + "learning_rate": 6.847826405045604e-05, + "loss": 0.0586, + "step": 25688 + }, + { + "epoch": 0.6021906151655269, + "grad_norm": 1.0239113569259644, + "learning_rate": 6.847127350854005e-05, + "loss": 0.1199, + "step": 25689 + }, + { + "epoch": 0.6022140567403318, + "grad_norm": 0.28778910636901855, + "learning_rate": 6.846428313769981e-05, + "loss": 0.0185, + "step": 25690 + }, + { + "epoch": 0.6022374983151368, + "grad_norm": 1.339352011680603, + "learning_rate": 6.845729293797325e-05, + "loss": 0.1436, + "step": 25691 + }, + { + "epoch": 0.6022609398899418, + "grad_norm": 1.821197748184204, + "learning_rate": 6.845030290939831e-05, + "loss": 0.1494, + "step": 25692 + }, + { + "epoch": 0.6022843814647468, + "grad_norm": 3.0401556491851807, + "learning_rate": 6.844331305201292e-05, + "loss": 0.2802, + "step": 25693 + }, + { + "epoch": 0.6023078230395518, + "grad_norm": 1.5109245777130127, + "learning_rate": 6.843632336585497e-05, + "loss": 0.1413, + "step": 25694 + }, + { + "epoch": 0.6023312646143568, + "grad_norm": 0.898079514503479, + "learning_rate": 6.842933385096244e-05, + "loss": 0.4562, + "step": 25695 + }, + { + "epoch": 0.6023547061891618, + "grad_norm": 2.5877535343170166, + "learning_rate": 6.842234450737323e-05, + "loss": 0.1565, + "step": 25696 + }, + { + "epoch": 0.6023781477639668, + "grad_norm": 2.997121572494507, + "learning_rate": 6.841535533512528e-05, + "loss": 0.2907, + "step": 25697 + }, + { + "epoch": 0.6024015893387717, + "grad_norm": 3.2110140323638916, + "learning_rate": 6.84083663342565e-05, + "loss": 0.206, + "step": 25698 + }, + { + "epoch": 0.6024250309135768, + "grad_norm": 4.513759613037109, + "learning_rate": 6.84013775048048e-05, + "loss": 0.3956, + "step": 25699 + }, + { + "epoch": 0.6024484724883817, + "grad_norm": 0.7177605032920837, + "learning_rate": 6.839438884680813e-05, + "loss": 0.0739, + "step": 25700 + }, + { + "epoch": 0.6024719140631868, + "grad_norm": 3.399083375930786, + "learning_rate": 6.838740036030437e-05, + "loss": 0.2747, + "step": 25701 + }, + { + "epoch": 0.6024953556379917, + "grad_norm": 3.2346792221069336, + "learning_rate": 6.838041204533148e-05, + "loss": 1.66, + "step": 25702 + }, + { + "epoch": 0.6025187972127968, + "grad_norm": 2.3510169982910156, + "learning_rate": 6.837342390192736e-05, + "loss": 0.1043, + "step": 25703 + }, + { + "epoch": 0.6025422387876017, + "grad_norm": 2.791898727416992, + "learning_rate": 6.836643593012993e-05, + "loss": 0.4352, + "step": 25704 + }, + { + "epoch": 0.6025656803624068, + "grad_norm": 1.1179261207580566, + "learning_rate": 6.835944812997712e-05, + "loss": 0.1128, + "step": 25705 + }, + { + "epoch": 0.6025891219372117, + "grad_norm": 0.5819125771522522, + "learning_rate": 6.835246050150679e-05, + "loss": 0.0318, + "step": 25706 + }, + { + "epoch": 0.6026125635120168, + "grad_norm": 2.999784231185913, + "learning_rate": 6.834547304475691e-05, + "loss": 0.2803, + "step": 25707 + }, + { + "epoch": 0.6026360050868217, + "grad_norm": 1.3041250705718994, + "learning_rate": 6.833848575976543e-05, + "loss": 0.0978, + "step": 25708 + }, + { + "epoch": 0.6026594466616267, + "grad_norm": 3.7667806148529053, + "learning_rate": 6.833149864657021e-05, + "loss": 0.3454, + "step": 25709 + }, + { + "epoch": 0.6026828882364317, + "grad_norm": 1.5779063701629639, + "learning_rate": 6.832451170520912e-05, + "loss": 0.1678, + "step": 25710 + }, + { + "epoch": 0.6027063298112367, + "grad_norm": 0.7707092761993408, + "learning_rate": 6.831752493572017e-05, + "loss": 0.0657, + "step": 25711 + }, + { + "epoch": 0.6027297713860417, + "grad_norm": 0.8695182800292969, + "learning_rate": 6.831053833814121e-05, + "loss": 0.0531, + "step": 25712 + }, + { + "epoch": 0.6027532129608467, + "grad_norm": 3.0387985706329346, + "learning_rate": 6.830355191251015e-05, + "loss": 1.5114, + "step": 25713 + }, + { + "epoch": 0.6027766545356517, + "grad_norm": 3.522064208984375, + "learning_rate": 6.829656565886491e-05, + "loss": 0.4464, + "step": 25714 + }, + { + "epoch": 0.6028000961104567, + "grad_norm": 2.3298747539520264, + "learning_rate": 6.82895795772434e-05, + "loss": 1.3761, + "step": 25715 + }, + { + "epoch": 0.6028235376852616, + "grad_norm": 3.8141579627990723, + "learning_rate": 6.828259366768352e-05, + "loss": 0.3347, + "step": 25716 + }, + { + "epoch": 0.6028469792600667, + "grad_norm": 3.8783986568450928, + "learning_rate": 6.827560793022319e-05, + "loss": 0.3238, + "step": 25717 + }, + { + "epoch": 0.6028704208348716, + "grad_norm": 3.4434988498687744, + "learning_rate": 6.826862236490025e-05, + "loss": 0.3659, + "step": 25718 + }, + { + "epoch": 0.6028938624096767, + "grad_norm": 4.481888771057129, + "learning_rate": 6.82616369717527e-05, + "loss": 0.4512, + "step": 25719 + }, + { + "epoch": 0.6029173039844816, + "grad_norm": 3.6158933639526367, + "learning_rate": 6.82546517508184e-05, + "loss": 0.267, + "step": 25720 + }, + { + "epoch": 0.6029407455592867, + "grad_norm": 2.3834588527679443, + "learning_rate": 6.824766670213524e-05, + "loss": 0.2342, + "step": 25721 + }, + { + "epoch": 0.6029641871340917, + "grad_norm": 6.355679512023926, + "learning_rate": 6.824068182574117e-05, + "loss": 0.3215, + "step": 25722 + }, + { + "epoch": 0.6029876287088967, + "grad_norm": 4.1591901779174805, + "learning_rate": 6.823369712167402e-05, + "loss": 0.3514, + "step": 25723 + }, + { + "epoch": 0.6030110702837017, + "grad_norm": 2.3022520542144775, + "learning_rate": 6.822671258997174e-05, + "loss": 0.2375, + "step": 25724 + }, + { + "epoch": 0.6030345118585066, + "grad_norm": 1.3877869844436646, + "learning_rate": 6.82197282306722e-05, + "loss": 0.1326, + "step": 25725 + }, + { + "epoch": 0.6030579534333117, + "grad_norm": 1.691104531288147, + "learning_rate": 6.82127440438133e-05, + "loss": 0.191, + "step": 25726 + }, + { + "epoch": 0.6030813950081166, + "grad_norm": 2.186443567276001, + "learning_rate": 6.820576002943297e-05, + "loss": 0.1972, + "step": 25727 + }, + { + "epoch": 0.6031048365829217, + "grad_norm": 2.0734033584594727, + "learning_rate": 6.819877618756907e-05, + "loss": 0.1402, + "step": 25728 + }, + { + "epoch": 0.6031282781577266, + "grad_norm": 3.9791650772094727, + "learning_rate": 6.819179251825949e-05, + "loss": 0.4695, + "step": 25729 + }, + { + "epoch": 0.6031517197325317, + "grad_norm": 3.535388946533203, + "learning_rate": 6.818480902154212e-05, + "loss": 0.3337, + "step": 25730 + }, + { + "epoch": 0.6031751613073366, + "grad_norm": 3.156414031982422, + "learning_rate": 6.81778256974549e-05, + "loss": 0.3427, + "step": 25731 + }, + { + "epoch": 0.6031986028821417, + "grad_norm": 3.6153945922851562, + "learning_rate": 6.81708425460357e-05, + "loss": 0.3859, + "step": 25732 + }, + { + "epoch": 0.6032220444569466, + "grad_norm": 2.707350969314575, + "learning_rate": 6.816385956732238e-05, + "loss": 0.1699, + "step": 25733 + }, + { + "epoch": 0.6032454860317517, + "grad_norm": 1.996289849281311, + "learning_rate": 6.815687676135287e-05, + "loss": 0.1101, + "step": 25734 + }, + { + "epoch": 0.6032689276065566, + "grad_norm": 2.501237154006958, + "learning_rate": 6.814989412816506e-05, + "loss": 0.2556, + "step": 25735 + }, + { + "epoch": 0.6032923691813616, + "grad_norm": 3.498183250427246, + "learning_rate": 6.814291166779678e-05, + "loss": 0.1686, + "step": 25736 + }, + { + "epoch": 0.6033158107561666, + "grad_norm": 2.136432409286499, + "learning_rate": 6.813592938028598e-05, + "loss": 1.2319, + "step": 25737 + }, + { + "epoch": 0.6033392523309716, + "grad_norm": 2.690396547317505, + "learning_rate": 6.812894726567053e-05, + "loss": 0.1088, + "step": 25738 + }, + { + "epoch": 0.6033626939057766, + "grad_norm": 2.9734675884246826, + "learning_rate": 6.812196532398829e-05, + "loss": 0.2417, + "step": 25739 + }, + { + "epoch": 0.6033861354805816, + "grad_norm": 3.5616698265075684, + "learning_rate": 6.811498355527717e-05, + "loss": 0.3642, + "step": 25740 + }, + { + "epoch": 0.6034095770553866, + "grad_norm": 1.833797812461853, + "learning_rate": 6.810800195957505e-05, + "loss": 0.2883, + "step": 25741 + }, + { + "epoch": 0.6034330186301916, + "grad_norm": 6.206674575805664, + "learning_rate": 6.810102053691978e-05, + "loss": 0.7922, + "step": 25742 + }, + { + "epoch": 0.6034564602049965, + "grad_norm": 3.308720350265503, + "learning_rate": 6.809403928734928e-05, + "loss": 0.189, + "step": 25743 + }, + { + "epoch": 0.6034799017798016, + "grad_norm": 2.923604965209961, + "learning_rate": 6.808705821090145e-05, + "loss": 0.2096, + "step": 25744 + }, + { + "epoch": 0.6035033433546065, + "grad_norm": 0.5299196839332581, + "learning_rate": 6.808007730761411e-05, + "loss": 0.0685, + "step": 25745 + }, + { + "epoch": 0.6035267849294116, + "grad_norm": 3.68416690826416, + "learning_rate": 6.807309657752519e-05, + "loss": 0.3358, + "step": 25746 + }, + { + "epoch": 0.6035502265042165, + "grad_norm": 3.264500379562378, + "learning_rate": 6.806611602067254e-05, + "loss": 0.1743, + "step": 25747 + }, + { + "epoch": 0.6035736680790216, + "grad_norm": 2.9582488536834717, + "learning_rate": 6.805913563709402e-05, + "loss": 0.2288, + "step": 25748 + }, + { + "epoch": 0.6035971096538265, + "grad_norm": 3.231837034225464, + "learning_rate": 6.805215542682757e-05, + "loss": 0.2087, + "step": 25749 + }, + { + "epoch": 0.6036205512286316, + "grad_norm": 3.401394844055176, + "learning_rate": 6.8045175389911e-05, + "loss": 0.1706, + "step": 25750 + }, + { + "epoch": 0.6036439928034365, + "grad_norm": 2.3469064235687256, + "learning_rate": 6.80381955263822e-05, + "loss": 0.2366, + "step": 25751 + }, + { + "epoch": 0.6036674343782416, + "grad_norm": 3.4962918758392334, + "learning_rate": 6.803121583627907e-05, + "loss": 0.651, + "step": 25752 + }, + { + "epoch": 0.6036908759530465, + "grad_norm": 3.4771342277526855, + "learning_rate": 6.802423631963946e-05, + "loss": 0.321, + "step": 25753 + }, + { + "epoch": 0.6037143175278515, + "grad_norm": 2.2547171115875244, + "learning_rate": 6.801725697650119e-05, + "loss": 0.3458, + "step": 25754 + }, + { + "epoch": 0.6037377591026565, + "grad_norm": 0.6080065369606018, + "learning_rate": 6.80102778069022e-05, + "loss": 0.0444, + "step": 25755 + }, + { + "epoch": 0.6037612006774615, + "grad_norm": 2.30403208732605, + "learning_rate": 6.800329881088038e-05, + "loss": 0.1773, + "step": 25756 + }, + { + "epoch": 0.6037846422522665, + "grad_norm": 3.1056618690490723, + "learning_rate": 6.799631998847356e-05, + "loss": 0.2008, + "step": 25757 + }, + { + "epoch": 0.6038080838270715, + "grad_norm": 3.83231520652771, + "learning_rate": 6.798934133971958e-05, + "loss": 0.3314, + "step": 25758 + }, + { + "epoch": 0.6038315254018765, + "grad_norm": 0.7337249517440796, + "learning_rate": 6.798236286465636e-05, + "loss": 0.0421, + "step": 25759 + }, + { + "epoch": 0.6038549669766815, + "grad_norm": 2.824303388595581, + "learning_rate": 6.797538456332174e-05, + "loss": 0.2885, + "step": 25760 + }, + { + "epoch": 0.6038784085514864, + "grad_norm": 4.799665451049805, + "learning_rate": 6.796840643575355e-05, + "loss": 0.5306, + "step": 25761 + }, + { + "epoch": 0.6039018501262915, + "grad_norm": 3.0582480430603027, + "learning_rate": 6.796142848198973e-05, + "loss": 0.2188, + "step": 25762 + }, + { + "epoch": 0.6039252917010964, + "grad_norm": 5.4103169441223145, + "learning_rate": 6.795445070206807e-05, + "loss": 0.4556, + "step": 25763 + }, + { + "epoch": 0.6039487332759015, + "grad_norm": 1.1148035526275635, + "learning_rate": 6.794747309602647e-05, + "loss": 0.0726, + "step": 25764 + }, + { + "epoch": 0.6039721748507064, + "grad_norm": 2.823591947555542, + "learning_rate": 6.794049566390279e-05, + "loss": 0.2103, + "step": 25765 + }, + { + "epoch": 0.6039956164255115, + "grad_norm": 3.1182901859283447, + "learning_rate": 6.793351840573483e-05, + "loss": 0.3767, + "step": 25766 + }, + { + "epoch": 0.6040190580003164, + "grad_norm": 2.790822982788086, + "learning_rate": 6.792654132156056e-05, + "loss": 0.2128, + "step": 25767 + }, + { + "epoch": 0.6040424995751215, + "grad_norm": 2.341121196746826, + "learning_rate": 6.791956441141776e-05, + "loss": 1.5265, + "step": 25768 + }, + { + "epoch": 0.6040659411499264, + "grad_norm": 2.9392874240875244, + "learning_rate": 6.791258767534432e-05, + "loss": 0.3746, + "step": 25769 + }, + { + "epoch": 0.6040893827247314, + "grad_norm": 0.7372092008590698, + "learning_rate": 6.790561111337808e-05, + "loss": 0.0479, + "step": 25770 + }, + { + "epoch": 0.6041128242995364, + "grad_norm": 3.3300728797912598, + "learning_rate": 6.789863472555687e-05, + "loss": 0.268, + "step": 25771 + }, + { + "epoch": 0.6041362658743414, + "grad_norm": 2.4065284729003906, + "learning_rate": 6.789165851191858e-05, + "loss": 0.1776, + "step": 25772 + }, + { + "epoch": 0.6041597074491465, + "grad_norm": 1.0383893251419067, + "learning_rate": 6.788468247250108e-05, + "loss": 0.0567, + "step": 25773 + }, + { + "epoch": 0.6041831490239514, + "grad_norm": 0.8872694373130798, + "learning_rate": 6.787770660734216e-05, + "loss": 0.0935, + "step": 25774 + }, + { + "epoch": 0.6042065905987565, + "grad_norm": 1.4013298749923706, + "learning_rate": 6.787073091647974e-05, + "loss": 0.1044, + "step": 25775 + }, + { + "epoch": 0.6042300321735614, + "grad_norm": 0.7964940667152405, + "learning_rate": 6.786375539995161e-05, + "loss": 0.0881, + "step": 25776 + }, + { + "epoch": 0.6042534737483665, + "grad_norm": 2.439244031906128, + "learning_rate": 6.785678005779567e-05, + "loss": 0.218, + "step": 25777 + }, + { + "epoch": 0.6042769153231714, + "grad_norm": 3.3308053016662598, + "learning_rate": 6.784980489004975e-05, + "loss": 0.1865, + "step": 25778 + }, + { + "epoch": 0.6043003568979765, + "grad_norm": 2.3615944385528564, + "learning_rate": 6.784282989675164e-05, + "loss": 0.2822, + "step": 25779 + }, + { + "epoch": 0.6043237984727814, + "grad_norm": 1.5210745334625244, + "learning_rate": 6.783585507793929e-05, + "loss": 0.0921, + "step": 25780 + }, + { + "epoch": 0.6043472400475864, + "grad_norm": 1.0431569814682007, + "learning_rate": 6.782888043365045e-05, + "loss": 0.0933, + "step": 25781 + }, + { + "epoch": 0.6043706816223914, + "grad_norm": 1.8609182834625244, + "learning_rate": 6.782190596392306e-05, + "loss": 0.1447, + "step": 25782 + }, + { + "epoch": 0.6043941231971964, + "grad_norm": 2.496669292449951, + "learning_rate": 6.781493166879487e-05, + "loss": 0.1705, + "step": 25783 + }, + { + "epoch": 0.6044175647720014, + "grad_norm": 3.5126430988311768, + "learning_rate": 6.78079575483038e-05, + "loss": 0.2852, + "step": 25784 + }, + { + "epoch": 0.6044410063468064, + "grad_norm": 3.2048637866973877, + "learning_rate": 6.780098360248766e-05, + "loss": 0.1117, + "step": 25785 + }, + { + "epoch": 0.6044644479216114, + "grad_norm": 3.173851251602173, + "learning_rate": 6.779400983138426e-05, + "loss": 0.3067, + "step": 25786 + }, + { + "epoch": 0.6044878894964164, + "grad_norm": 2.2055869102478027, + "learning_rate": 6.778703623503149e-05, + "loss": 0.9822, + "step": 25787 + }, + { + "epoch": 0.6045113310712213, + "grad_norm": 2.612863540649414, + "learning_rate": 6.778006281346717e-05, + "loss": 0.3518, + "step": 25788 + }, + { + "epoch": 0.6045347726460264, + "grad_norm": 2.584578275680542, + "learning_rate": 6.777308956672912e-05, + "loss": 0.2674, + "step": 25789 + }, + { + "epoch": 0.6045582142208313, + "grad_norm": 2.5031962394714355, + "learning_rate": 6.77661164948552e-05, + "loss": 0.3688, + "step": 25790 + }, + { + "epoch": 0.6045816557956364, + "grad_norm": 2.486511468887329, + "learning_rate": 6.77591435978832e-05, + "loss": 0.2005, + "step": 25791 + }, + { + "epoch": 0.6046050973704413, + "grad_norm": 3.0851128101348877, + "learning_rate": 6.775217087585104e-05, + "loss": 0.394, + "step": 25792 + }, + { + "epoch": 0.6046285389452464, + "grad_norm": 3.3445117473602295, + "learning_rate": 6.774519832879648e-05, + "loss": 0.3925, + "step": 25793 + }, + { + "epoch": 0.6046519805200513, + "grad_norm": 3.64982008934021, + "learning_rate": 6.773822595675741e-05, + "loss": 0.4963, + "step": 25794 + }, + { + "epoch": 0.6046754220948564, + "grad_norm": 3.0910708904266357, + "learning_rate": 6.773125375977162e-05, + "loss": 0.2942, + "step": 25795 + }, + { + "epoch": 0.6046988636696613, + "grad_norm": 2.89715313911438, + "learning_rate": 6.772428173787694e-05, + "loss": 0.384, + "step": 25796 + }, + { + "epoch": 0.6047223052444664, + "grad_norm": 0.7716100811958313, + "learning_rate": 6.771730989111124e-05, + "loss": 0.0605, + "step": 25797 + }, + { + "epoch": 0.6047457468192713, + "grad_norm": 1.5761417150497437, + "learning_rate": 6.771033821951232e-05, + "loss": 0.1151, + "step": 25798 + }, + { + "epoch": 0.6047691883940763, + "grad_norm": 0.730085551738739, + "learning_rate": 6.7703366723118e-05, + "loss": 0.0662, + "step": 25799 + }, + { + "epoch": 0.6047926299688813, + "grad_norm": 10.110665321350098, + "learning_rate": 6.769639540196613e-05, + "loss": 0.265, + "step": 25800 + }, + { + "epoch": 0.6048160715436863, + "grad_norm": 1.4754749536514282, + "learning_rate": 6.768942425609454e-05, + "loss": 0.988, + "step": 25801 + }, + { + "epoch": 0.6048395131184913, + "grad_norm": 3.1606364250183105, + "learning_rate": 6.768245328554101e-05, + "loss": 0.1797, + "step": 25802 + }, + { + "epoch": 0.6048629546932963, + "grad_norm": 3.488503932952881, + "learning_rate": 6.767548249034339e-05, + "loss": 0.1682, + "step": 25803 + }, + { + "epoch": 0.6048863962681013, + "grad_norm": 1.7502707242965698, + "learning_rate": 6.766851187053953e-05, + "loss": 0.087, + "step": 25804 + }, + { + "epoch": 0.6049098378429063, + "grad_norm": 2.1628470420837402, + "learning_rate": 6.766154142616725e-05, + "loss": 0.1935, + "step": 25805 + }, + { + "epoch": 0.6049332794177112, + "grad_norm": 3.439617156982422, + "learning_rate": 6.765457115726432e-05, + "loss": 1.1665, + "step": 25806 + }, + { + "epoch": 0.6049567209925163, + "grad_norm": 3.496014356613159, + "learning_rate": 6.764760106386864e-05, + "loss": 0.5041, + "step": 25807 + }, + { + "epoch": 0.6049801625673212, + "grad_norm": 2.8118176460266113, + "learning_rate": 6.764063114601797e-05, + "loss": 0.301, + "step": 25808 + }, + { + "epoch": 0.6050036041421263, + "grad_norm": 4.558779716491699, + "learning_rate": 6.763366140375014e-05, + "loss": 0.6806, + "step": 25809 + }, + { + "epoch": 0.6050270457169312, + "grad_norm": 2.224581003189087, + "learning_rate": 6.762669183710298e-05, + "loss": 0.139, + "step": 25810 + }, + { + "epoch": 0.6050504872917363, + "grad_norm": 0.4982497990131378, + "learning_rate": 6.76197224461143e-05, + "loss": 0.028, + "step": 25811 + }, + { + "epoch": 0.6050739288665412, + "grad_norm": 3.36673641204834, + "learning_rate": 6.76127532308219e-05, + "loss": 0.3145, + "step": 25812 + }, + { + "epoch": 0.6050973704413463, + "grad_norm": 2.198485851287842, + "learning_rate": 6.760578419126362e-05, + "loss": 0.2588, + "step": 25813 + }, + { + "epoch": 0.6051208120161512, + "grad_norm": 0.5138484835624695, + "learning_rate": 6.759881532747727e-05, + "loss": 0.0334, + "step": 25814 + }, + { + "epoch": 0.6051442535909562, + "grad_norm": 3.346419095993042, + "learning_rate": 6.759184663950065e-05, + "loss": 0.3988, + "step": 25815 + }, + { + "epoch": 0.6051676951657612, + "grad_norm": 4.013106822967529, + "learning_rate": 6.758487812737157e-05, + "loss": 0.3161, + "step": 25816 + }, + { + "epoch": 0.6051911367405662, + "grad_norm": 3.275614023208618, + "learning_rate": 6.757790979112789e-05, + "loss": 0.4121, + "step": 25817 + }, + { + "epoch": 0.6052145783153712, + "grad_norm": 2.6866848468780518, + "learning_rate": 6.757094163080736e-05, + "loss": 0.3313, + "step": 25818 + }, + { + "epoch": 0.6052380198901762, + "grad_norm": 0.9282033443450928, + "learning_rate": 6.75639736464478e-05, + "loss": 0.0785, + "step": 25819 + }, + { + "epoch": 0.6052614614649812, + "grad_norm": 3.6147687435150146, + "learning_rate": 6.755700583808706e-05, + "loss": 0.2614, + "step": 25820 + }, + { + "epoch": 0.6052849030397862, + "grad_norm": 2.4436089992523193, + "learning_rate": 6.755003820576288e-05, + "loss": 0.2876, + "step": 25821 + }, + { + "epoch": 0.6053083446145912, + "grad_norm": 2.4832303524017334, + "learning_rate": 6.754307074951315e-05, + "loss": 1.4503, + "step": 25822 + }, + { + "epoch": 0.6053317861893962, + "grad_norm": 2.4483516216278076, + "learning_rate": 6.753610346937562e-05, + "loss": 0.214, + "step": 25823 + }, + { + "epoch": 0.6053552277642013, + "grad_norm": 2.7934677600860596, + "learning_rate": 6.75291363653881e-05, + "loss": 0.2869, + "step": 25824 + }, + { + "epoch": 0.6053786693390062, + "grad_norm": 4.184495449066162, + "learning_rate": 6.75221694375884e-05, + "loss": 1.8651, + "step": 25825 + }, + { + "epoch": 0.6054021109138112, + "grad_norm": 2.8360941410064697, + "learning_rate": 6.751520268601433e-05, + "loss": 0.3066, + "step": 25826 + }, + { + "epoch": 0.6054255524886162, + "grad_norm": 1.8062433004379272, + "learning_rate": 6.750823611070364e-05, + "loss": 0.1206, + "step": 25827 + }, + { + "epoch": 0.6054489940634212, + "grad_norm": 2.136695146560669, + "learning_rate": 6.75012697116942e-05, + "loss": 0.1546, + "step": 25828 + }, + { + "epoch": 0.6054724356382262, + "grad_norm": 2.2448980808258057, + "learning_rate": 6.749430348902379e-05, + "loss": 1.7496, + "step": 25829 + }, + { + "epoch": 0.6054958772130312, + "grad_norm": 2.3553714752197266, + "learning_rate": 6.748733744273021e-05, + "loss": 0.1606, + "step": 25830 + }, + { + "epoch": 0.6055193187878362, + "grad_norm": 2.1237809658050537, + "learning_rate": 6.748037157285123e-05, + "loss": 0.114, + "step": 25831 + }, + { + "epoch": 0.6055427603626412, + "grad_norm": 2.4649696350097656, + "learning_rate": 6.74734058794247e-05, + "loss": 0.305, + "step": 25832 + }, + { + "epoch": 0.6055662019374461, + "grad_norm": 3.9560635089874268, + "learning_rate": 6.746644036248838e-05, + "loss": 0.4372, + "step": 25833 + }, + { + "epoch": 0.6055896435122512, + "grad_norm": 3.04666805267334, + "learning_rate": 6.745947502208004e-05, + "loss": 0.1899, + "step": 25834 + }, + { + "epoch": 0.6056130850870561, + "grad_norm": 2.383237838745117, + "learning_rate": 6.745250985823753e-05, + "loss": 0.2232, + "step": 25835 + }, + { + "epoch": 0.6056365266618612, + "grad_norm": 2.9752161502838135, + "learning_rate": 6.744554487099862e-05, + "loss": 0.2123, + "step": 25836 + }, + { + "epoch": 0.6056599682366661, + "grad_norm": 3.09120774269104, + "learning_rate": 6.743858006040107e-05, + "loss": 0.4251, + "step": 25837 + }, + { + "epoch": 0.6056834098114712, + "grad_norm": 4.543549060821533, + "learning_rate": 6.743161542648274e-05, + "loss": 0.5381, + "step": 25838 + }, + { + "epoch": 0.6057068513862761, + "grad_norm": 0.8355132341384888, + "learning_rate": 6.742465096928132e-05, + "loss": 0.057, + "step": 25839 + }, + { + "epoch": 0.6057302929610812, + "grad_norm": 4.66876745223999, + "learning_rate": 6.74176866888347e-05, + "loss": 0.4767, + "step": 25840 + }, + { + "epoch": 0.6057537345358861, + "grad_norm": 2.8615198135375977, + "learning_rate": 6.741072258518062e-05, + "loss": 0.1911, + "step": 25841 + }, + { + "epoch": 0.6057771761106912, + "grad_norm": 0.5622727870941162, + "learning_rate": 6.740375865835689e-05, + "loss": 0.0629, + "step": 25842 + }, + { + "epoch": 0.6058006176854961, + "grad_norm": 3.730713367462158, + "learning_rate": 6.739679490840127e-05, + "loss": 0.2945, + "step": 25843 + }, + { + "epoch": 0.6058240592603011, + "grad_norm": 0.17264053225517273, + "learning_rate": 6.738983133535155e-05, + "loss": 0.0154, + "step": 25844 + }, + { + "epoch": 0.6058475008351061, + "grad_norm": 1.3741071224212646, + "learning_rate": 6.738286793924554e-05, + "loss": 0.1174, + "step": 25845 + }, + { + "epoch": 0.6058709424099111, + "grad_norm": 3.340897560119629, + "learning_rate": 6.7375904720121e-05, + "loss": 0.3797, + "step": 25846 + }, + { + "epoch": 0.6058943839847161, + "grad_norm": 2.159184217453003, + "learning_rate": 6.736894167801571e-05, + "loss": 0.1803, + "step": 25847 + }, + { + "epoch": 0.6059178255595211, + "grad_norm": 1.8471969366073608, + "learning_rate": 6.736197881296748e-05, + "loss": 0.1281, + "step": 25848 + }, + { + "epoch": 0.6059412671343261, + "grad_norm": 2.4309186935424805, + "learning_rate": 6.735501612501407e-05, + "loss": 1.5215, + "step": 25849 + }, + { + "epoch": 0.6059647087091311, + "grad_norm": 0.2834518551826477, + "learning_rate": 6.734805361419322e-05, + "loss": 0.0198, + "step": 25850 + }, + { + "epoch": 0.605988150283936, + "grad_norm": 4.05013370513916, + "learning_rate": 6.73410912805428e-05, + "loss": 0.4013, + "step": 25851 + }, + { + "epoch": 0.6060115918587411, + "grad_norm": 1.4142402410507202, + "learning_rate": 6.733412912410048e-05, + "loss": 0.1122, + "step": 25852 + }, + { + "epoch": 0.606035033433546, + "grad_norm": 1.17308509349823, + "learning_rate": 6.732716714490413e-05, + "loss": 0.0843, + "step": 25853 + }, + { + "epoch": 0.6060584750083511, + "grad_norm": 3.5826423168182373, + "learning_rate": 6.732020534299148e-05, + "loss": 0.3128, + "step": 25854 + }, + { + "epoch": 0.606081916583156, + "grad_norm": 2.36668062210083, + "learning_rate": 6.731324371840033e-05, + "loss": 0.1996, + "step": 25855 + }, + { + "epoch": 0.6061053581579611, + "grad_norm": 2.968327760696411, + "learning_rate": 6.730628227116844e-05, + "loss": 0.2265, + "step": 25856 + }, + { + "epoch": 0.606128799732766, + "grad_norm": 1.2337605953216553, + "learning_rate": 6.729932100133355e-05, + "loss": 0.2033, + "step": 25857 + }, + { + "epoch": 0.6061522413075711, + "grad_norm": 4.1333184242248535, + "learning_rate": 6.72923599089335e-05, + "loss": 0.4016, + "step": 25858 + }, + { + "epoch": 0.606175682882376, + "grad_norm": 2.414828062057495, + "learning_rate": 6.7285398994006e-05, + "loss": 0.185, + "step": 25859 + }, + { + "epoch": 0.606199124457181, + "grad_norm": 3.052222967147827, + "learning_rate": 6.727843825658886e-05, + "loss": 0.1918, + "step": 25860 + }, + { + "epoch": 0.606222566031986, + "grad_norm": 2.36607027053833, + "learning_rate": 6.727147769671984e-05, + "loss": 0.1434, + "step": 25861 + }, + { + "epoch": 0.606246007606791, + "grad_norm": 3.9869468212127686, + "learning_rate": 6.726451731443669e-05, + "loss": 0.4615, + "step": 25862 + }, + { + "epoch": 0.606269449181596, + "grad_norm": 3.73823881149292, + "learning_rate": 6.72575571097772e-05, + "loss": 0.3025, + "step": 25863 + }, + { + "epoch": 0.606292890756401, + "grad_norm": 2.3385908603668213, + "learning_rate": 6.725059708277907e-05, + "loss": 0.222, + "step": 25864 + }, + { + "epoch": 0.606316332331206, + "grad_norm": 2.4691689014434814, + "learning_rate": 6.724363723348018e-05, + "loss": 0.0574, + "step": 25865 + }, + { + "epoch": 0.606339773906011, + "grad_norm": 1.435122013092041, + "learning_rate": 6.723667756191821e-05, + "loss": 0.1774, + "step": 25866 + }, + { + "epoch": 0.606363215480816, + "grad_norm": 3.1832528114318848, + "learning_rate": 6.722971806813097e-05, + "loss": 0.3008, + "step": 25867 + }, + { + "epoch": 0.606386657055621, + "grad_norm": 2.702253580093384, + "learning_rate": 6.722275875215621e-05, + "loss": 0.1834, + "step": 25868 + }, + { + "epoch": 0.6064100986304259, + "grad_norm": 2.9222159385681152, + "learning_rate": 6.721579961403166e-05, + "loss": 0.2122, + "step": 25869 + }, + { + "epoch": 0.606433540205231, + "grad_norm": 1.5716876983642578, + "learning_rate": 6.720884065379511e-05, + "loss": 0.1495, + "step": 25870 + }, + { + "epoch": 0.6064569817800359, + "grad_norm": 2.7855064868927, + "learning_rate": 6.720188187148433e-05, + "loss": 0.1448, + "step": 25871 + }, + { + "epoch": 0.606480423354841, + "grad_norm": 3.545933246612549, + "learning_rate": 6.719492326713703e-05, + "loss": 0.2701, + "step": 25872 + }, + { + "epoch": 0.6065038649296459, + "grad_norm": 3.4025683403015137, + "learning_rate": 6.718796484079102e-05, + "loss": 0.2671, + "step": 25873 + }, + { + "epoch": 0.606527306504451, + "grad_norm": 3.160011053085327, + "learning_rate": 6.718100659248402e-05, + "loss": 0.2899, + "step": 25874 + }, + { + "epoch": 0.606550748079256, + "grad_norm": 2.667701482772827, + "learning_rate": 6.71740485222538e-05, + "loss": 0.2483, + "step": 25875 + }, + { + "epoch": 0.606574189654061, + "grad_norm": 3.562717914581299, + "learning_rate": 6.71670906301381e-05, + "loss": 0.367, + "step": 25876 + }, + { + "epoch": 0.606597631228866, + "grad_norm": 5.022034168243408, + "learning_rate": 6.716013291617472e-05, + "loss": 0.4009, + "step": 25877 + }, + { + "epoch": 0.606621072803671, + "grad_norm": 2.2964603900909424, + "learning_rate": 6.715317538040138e-05, + "loss": 0.1531, + "step": 25878 + }, + { + "epoch": 0.606644514378476, + "grad_norm": 2.419158697128296, + "learning_rate": 6.714621802285583e-05, + "loss": 0.2417, + "step": 25879 + }, + { + "epoch": 0.6066679559532809, + "grad_norm": 1.1904770135879517, + "learning_rate": 6.713926084357582e-05, + "loss": 0.1249, + "step": 25880 + }, + { + "epoch": 0.606691397528086, + "grad_norm": 1.8885678052902222, + "learning_rate": 6.713230384259912e-05, + "loss": 0.1306, + "step": 25881 + }, + { + "epoch": 0.6067148391028909, + "grad_norm": 3.336245059967041, + "learning_rate": 6.712534701996344e-05, + "loss": 0.1803, + "step": 25882 + }, + { + "epoch": 0.606738280677696, + "grad_norm": 2.565258026123047, + "learning_rate": 6.711839037570658e-05, + "loss": 0.1638, + "step": 25883 + }, + { + "epoch": 0.6067617222525009, + "grad_norm": 2.866800546646118, + "learning_rate": 6.711143390986625e-05, + "loss": 0.2934, + "step": 25884 + }, + { + "epoch": 0.606785163827306, + "grad_norm": 1.3456101417541504, + "learning_rate": 6.710447762248018e-05, + "loss": 0.0786, + "step": 25885 + }, + { + "epoch": 0.6068086054021109, + "grad_norm": 1.3854351043701172, + "learning_rate": 6.709752151358617e-05, + "loss": 0.1008, + "step": 25886 + }, + { + "epoch": 0.606832046976916, + "grad_norm": 3.194059371948242, + "learning_rate": 6.709056558322192e-05, + "loss": 0.2096, + "step": 25887 + }, + { + "epoch": 0.6068554885517209, + "grad_norm": 2.4002602100372314, + "learning_rate": 6.708360983142515e-05, + "loss": 0.2369, + "step": 25888 + }, + { + "epoch": 0.6068789301265259, + "grad_norm": 1.7386802434921265, + "learning_rate": 6.707665425823366e-05, + "loss": 0.0946, + "step": 25889 + }, + { + "epoch": 0.6069023717013309, + "grad_norm": 1.8218833208084106, + "learning_rate": 6.706969886368519e-05, + "loss": 0.1144, + "step": 25890 + }, + { + "epoch": 0.6069258132761359, + "grad_norm": 4.678318500518799, + "learning_rate": 6.706274364781747e-05, + "loss": 0.3443, + "step": 25891 + }, + { + "epoch": 0.6069492548509409, + "grad_norm": 2.8082704544067383, + "learning_rate": 6.70557886106682e-05, + "loss": 0.22, + "step": 25892 + }, + { + "epoch": 0.6069726964257459, + "grad_norm": 2.9467928409576416, + "learning_rate": 6.704883375227517e-05, + "loss": 0.2086, + "step": 25893 + }, + { + "epoch": 0.6069961380005509, + "grad_norm": 4.47515344619751, + "learning_rate": 6.70418790726761e-05, + "loss": 0.3916, + "step": 25894 + }, + { + "epoch": 0.6070195795753559, + "grad_norm": 2.8566648960113525, + "learning_rate": 6.70349245719087e-05, + "loss": 0.1634, + "step": 25895 + }, + { + "epoch": 0.6070430211501608, + "grad_norm": 2.2900497913360596, + "learning_rate": 6.702797025001073e-05, + "loss": 0.138, + "step": 25896 + }, + { + "epoch": 0.6070664627249659, + "grad_norm": 2.6550214290618896, + "learning_rate": 6.702101610701991e-05, + "loss": 1.6826, + "step": 25897 + }, + { + "epoch": 0.6070899042997708, + "grad_norm": 4.402494430541992, + "learning_rate": 6.701406214297401e-05, + "loss": 0.3503, + "step": 25898 + }, + { + "epoch": 0.6071133458745759, + "grad_norm": 2.6954398155212402, + "learning_rate": 6.700710835791073e-05, + "loss": 0.1127, + "step": 25899 + }, + { + "epoch": 0.6071367874493808, + "grad_norm": 0.4304526150226593, + "learning_rate": 6.700015475186776e-05, + "loss": 0.03, + "step": 25900 + }, + { + "epoch": 0.6071602290241859, + "grad_norm": 3.771306276321411, + "learning_rate": 6.699320132488291e-05, + "loss": 0.3489, + "step": 25901 + }, + { + "epoch": 0.6071836705989908, + "grad_norm": 1.9013898372650146, + "learning_rate": 6.69862480769939e-05, + "loss": 0.0926, + "step": 25902 + }, + { + "epoch": 0.6072071121737959, + "grad_norm": 2.8173019886016846, + "learning_rate": 6.697929500823843e-05, + "loss": 0.2496, + "step": 25903 + }, + { + "epoch": 0.6072305537486008, + "grad_norm": 3.224740505218506, + "learning_rate": 6.697234211865422e-05, + "loss": 0.2934, + "step": 25904 + }, + { + "epoch": 0.6072539953234058, + "grad_norm": 2.3259034156799316, + "learning_rate": 6.696538940827901e-05, + "loss": 0.1396, + "step": 25905 + }, + { + "epoch": 0.6072774368982108, + "grad_norm": 2.7554264068603516, + "learning_rate": 6.695843687715055e-05, + "loss": 0.1775, + "step": 25906 + }, + { + "epoch": 0.6073008784730158, + "grad_norm": 2.9078547954559326, + "learning_rate": 6.69514845253065e-05, + "loss": 0.3531, + "step": 25907 + }, + { + "epoch": 0.6073243200478208, + "grad_norm": 0.5686345100402832, + "learning_rate": 6.694453235278466e-05, + "loss": 0.048, + "step": 25908 + }, + { + "epoch": 0.6073477616226258, + "grad_norm": 1.3030380010604858, + "learning_rate": 6.693758035962272e-05, + "loss": 0.0621, + "step": 25909 + }, + { + "epoch": 0.6073712031974308, + "grad_norm": 2.026951551437378, + "learning_rate": 6.693062854585837e-05, + "loss": 0.1442, + "step": 25910 + }, + { + "epoch": 0.6073946447722358, + "grad_norm": 2.3969883918762207, + "learning_rate": 6.692367691152938e-05, + "loss": 1.3965, + "step": 25911 + }, + { + "epoch": 0.6074180863470408, + "grad_norm": 1.5204561948776245, + "learning_rate": 6.69167254566734e-05, + "loss": 0.0683, + "step": 25912 + }, + { + "epoch": 0.6074415279218458, + "grad_norm": 1.884413242340088, + "learning_rate": 6.690977418132825e-05, + "loss": 0.1546, + "step": 25913 + }, + { + "epoch": 0.6074649694966507, + "grad_norm": 0.5708409547805786, + "learning_rate": 6.690282308553158e-05, + "loss": 0.0129, + "step": 25914 + }, + { + "epoch": 0.6074884110714558, + "grad_norm": 1.2366491556167603, + "learning_rate": 6.689587216932114e-05, + "loss": 0.1587, + "step": 25915 + }, + { + "epoch": 0.6075118526462607, + "grad_norm": 0.6958323121070862, + "learning_rate": 6.688892143273462e-05, + "loss": 0.0691, + "step": 25916 + }, + { + "epoch": 0.6075352942210658, + "grad_norm": 0.7281233668327332, + "learning_rate": 6.688197087580972e-05, + "loss": 0.056, + "step": 25917 + }, + { + "epoch": 0.6075587357958707, + "grad_norm": 3.608137369155884, + "learning_rate": 6.68750204985842e-05, + "loss": 0.2675, + "step": 25918 + }, + { + "epoch": 0.6075821773706758, + "grad_norm": 3.192922592163086, + "learning_rate": 6.686807030109577e-05, + "loss": 0.4319, + "step": 25919 + }, + { + "epoch": 0.6076056189454807, + "grad_norm": 2.742208957672119, + "learning_rate": 6.686112028338207e-05, + "loss": 0.1859, + "step": 25920 + }, + { + "epoch": 0.6076290605202858, + "grad_norm": 3.906869411468506, + "learning_rate": 6.685417044548091e-05, + "loss": 0.2526, + "step": 25921 + }, + { + "epoch": 0.6076525020950907, + "grad_norm": 3.4869744777679443, + "learning_rate": 6.684722078742994e-05, + "loss": 1.8972, + "step": 25922 + }, + { + "epoch": 0.6076759436698957, + "grad_norm": 1.8671178817749023, + "learning_rate": 6.684027130926687e-05, + "loss": 0.1048, + "step": 25923 + }, + { + "epoch": 0.6076993852447007, + "grad_norm": 1.0215582847595215, + "learning_rate": 6.683332201102943e-05, + "loss": 0.1018, + "step": 25924 + }, + { + "epoch": 0.6077228268195057, + "grad_norm": 2.345921754837036, + "learning_rate": 6.682637289275529e-05, + "loss": 1.0643, + "step": 25925 + }, + { + "epoch": 0.6077462683943108, + "grad_norm": 3.4649789333343506, + "learning_rate": 6.681942395448222e-05, + "loss": 0.3418, + "step": 25926 + }, + { + "epoch": 0.6077697099691157, + "grad_norm": 4.842194080352783, + "learning_rate": 6.681247519624785e-05, + "loss": 0.3301, + "step": 25927 + }, + { + "epoch": 0.6077931515439208, + "grad_norm": 1.4173998832702637, + "learning_rate": 6.680552661808996e-05, + "loss": 0.0917, + "step": 25928 + }, + { + "epoch": 0.6078165931187257, + "grad_norm": 4.553996562957764, + "learning_rate": 6.679857822004623e-05, + "loss": 0.2238, + "step": 25929 + }, + { + "epoch": 0.6078400346935308, + "grad_norm": 1.3097079992294312, + "learning_rate": 6.679163000215429e-05, + "loss": 0.1115, + "step": 25930 + }, + { + "epoch": 0.6078634762683357, + "grad_norm": 2.977403163909912, + "learning_rate": 6.678468196445195e-05, + "loss": 0.3176, + "step": 25931 + }, + { + "epoch": 0.6078869178431408, + "grad_norm": 1.5346440076828003, + "learning_rate": 6.677773410697683e-05, + "loss": 0.1792, + "step": 25932 + }, + { + "epoch": 0.6079103594179457, + "grad_norm": 0.34316304326057434, + "learning_rate": 6.677078642976669e-05, + "loss": 0.0401, + "step": 25933 + }, + { + "epoch": 0.6079338009927507, + "grad_norm": 2.169675827026367, + "learning_rate": 6.676383893285917e-05, + "loss": 0.1114, + "step": 25934 + }, + { + "epoch": 0.6079572425675557, + "grad_norm": 2.85526442527771, + "learning_rate": 6.6756891616292e-05, + "loss": 1.2263, + "step": 25935 + }, + { + "epoch": 0.6079806841423607, + "grad_norm": 2.4701766967773438, + "learning_rate": 6.674994448010287e-05, + "loss": 0.2629, + "step": 25936 + }, + { + "epoch": 0.6080041257171657, + "grad_norm": 3.8634586334228516, + "learning_rate": 6.674299752432946e-05, + "loss": 0.2963, + "step": 25937 + }, + { + "epoch": 0.6080275672919707, + "grad_norm": 3.02449893951416, + "learning_rate": 6.67360507490095e-05, + "loss": 0.2173, + "step": 25938 + }, + { + "epoch": 0.6080510088667757, + "grad_norm": 3.49904465675354, + "learning_rate": 6.672910415418066e-05, + "loss": 0.1687, + "step": 25939 + }, + { + "epoch": 0.6080744504415807, + "grad_norm": 0.3313174843788147, + "learning_rate": 6.672215773988064e-05, + "loss": 0.0212, + "step": 25940 + }, + { + "epoch": 0.6080978920163856, + "grad_norm": 4.984338283538818, + "learning_rate": 6.671521150614715e-05, + "loss": 0.5268, + "step": 25941 + }, + { + "epoch": 0.6081213335911907, + "grad_norm": 3.631657600402832, + "learning_rate": 6.670826545301783e-05, + "loss": 0.2662, + "step": 25942 + }, + { + "epoch": 0.6081447751659956, + "grad_norm": 3.405971050262451, + "learning_rate": 6.670131958053041e-05, + "loss": 0.3017, + "step": 25943 + }, + { + "epoch": 0.6081682167408007, + "grad_norm": 1.3710684776306152, + "learning_rate": 6.669437388872256e-05, + "loss": 0.1338, + "step": 25944 + }, + { + "epoch": 0.6081916583156056, + "grad_norm": 3.4398014545440674, + "learning_rate": 6.668742837763197e-05, + "loss": 0.2487, + "step": 25945 + }, + { + "epoch": 0.6082150998904107, + "grad_norm": 3.327237367630005, + "learning_rate": 6.668048304729634e-05, + "loss": 0.2736, + "step": 25946 + }, + { + "epoch": 0.6082385414652156, + "grad_norm": 3.3700156211853027, + "learning_rate": 6.667353789775335e-05, + "loss": 0.1998, + "step": 25947 + }, + { + "epoch": 0.6082619830400207, + "grad_norm": 1.362470269203186, + "learning_rate": 6.666659292904065e-05, + "loss": 0.1421, + "step": 25948 + }, + { + "epoch": 0.6082854246148256, + "grad_norm": 0.509120523929596, + "learning_rate": 6.665964814119595e-05, + "loss": 0.0708, + "step": 25949 + }, + { + "epoch": 0.6083088661896306, + "grad_norm": 1.921307921409607, + "learning_rate": 6.665270353425697e-05, + "loss": 0.1167, + "step": 25950 + }, + { + "epoch": 0.6083323077644356, + "grad_norm": 1.1930408477783203, + "learning_rate": 6.664575910826137e-05, + "loss": 0.108, + "step": 25951 + }, + { + "epoch": 0.6083557493392406, + "grad_norm": 1.0381017923355103, + "learning_rate": 6.663881486324677e-05, + "loss": 0.0786, + "step": 25952 + }, + { + "epoch": 0.6083791909140456, + "grad_norm": 0.5589065551757812, + "learning_rate": 6.663187079925094e-05, + "loss": 0.0394, + "step": 25953 + }, + { + "epoch": 0.6084026324888506, + "grad_norm": 2.5969796180725098, + "learning_rate": 6.662492691631152e-05, + "loss": 0.2013, + "step": 25954 + }, + { + "epoch": 0.6084260740636556, + "grad_norm": 2.297966241836548, + "learning_rate": 6.661798321446615e-05, + "loss": 0.2326, + "step": 25955 + }, + { + "epoch": 0.6084495156384606, + "grad_norm": 5.2585906982421875, + "learning_rate": 6.661103969375257e-05, + "loss": 0.5946, + "step": 25956 + }, + { + "epoch": 0.6084729572132656, + "grad_norm": 3.145502805709839, + "learning_rate": 6.660409635420844e-05, + "loss": 0.3024, + "step": 25957 + }, + { + "epoch": 0.6084963987880706, + "grad_norm": 1.372063159942627, + "learning_rate": 6.659715319587138e-05, + "loss": 0.2073, + "step": 25958 + }, + { + "epoch": 0.6085198403628755, + "grad_norm": 1.818503499031067, + "learning_rate": 6.659021021877914e-05, + "loss": 0.1721, + "step": 25959 + }, + { + "epoch": 0.6085432819376806, + "grad_norm": 2.242936134338379, + "learning_rate": 6.658326742296934e-05, + "loss": 1.1294, + "step": 25960 + }, + { + "epoch": 0.6085667235124855, + "grad_norm": 2.4955101013183594, + "learning_rate": 6.657632480847965e-05, + "loss": 0.1601, + "step": 25961 + }, + { + "epoch": 0.6085901650872906, + "grad_norm": 2.531731367111206, + "learning_rate": 6.656938237534778e-05, + "loss": 0.3015, + "step": 25962 + }, + { + "epoch": 0.6086136066620955, + "grad_norm": 3.809027910232544, + "learning_rate": 6.656244012361139e-05, + "loss": 0.5757, + "step": 25963 + }, + { + "epoch": 0.6086370482369006, + "grad_norm": 1.5157291889190674, + "learning_rate": 6.655549805330815e-05, + "loss": 0.1633, + "step": 25964 + }, + { + "epoch": 0.6086604898117055, + "grad_norm": 5.318216323852539, + "learning_rate": 6.65485561644757e-05, + "loss": 0.4658, + "step": 25965 + }, + { + "epoch": 0.6086839313865106, + "grad_norm": 3.060776948928833, + "learning_rate": 6.654161445715173e-05, + "loss": 1.489, + "step": 25966 + }, + { + "epoch": 0.6087073729613155, + "grad_norm": 1.2890201807022095, + "learning_rate": 6.653467293137392e-05, + "loss": 0.108, + "step": 25967 + }, + { + "epoch": 0.6087308145361205, + "grad_norm": 2.409442901611328, + "learning_rate": 6.652773158717988e-05, + "loss": 0.1456, + "step": 25968 + }, + { + "epoch": 0.6087542561109255, + "grad_norm": 1.460829257965088, + "learning_rate": 6.652079042460734e-05, + "loss": 0.1304, + "step": 25969 + }, + { + "epoch": 0.6087776976857305, + "grad_norm": 3.548013687133789, + "learning_rate": 6.65138494436939e-05, + "loss": 0.326, + "step": 25970 + }, + { + "epoch": 0.6088011392605355, + "grad_norm": 2.131269693374634, + "learning_rate": 6.650690864447729e-05, + "loss": 0.211, + "step": 25971 + }, + { + "epoch": 0.6088245808353405, + "grad_norm": 3.305823802947998, + "learning_rate": 6.649996802699513e-05, + "loss": 0.326, + "step": 25972 + }, + { + "epoch": 0.6088480224101455, + "grad_norm": 1.6023898124694824, + "learning_rate": 6.649302759128504e-05, + "loss": 0.1331, + "step": 25973 + }, + { + "epoch": 0.6088714639849505, + "grad_norm": 3.660893678665161, + "learning_rate": 6.648608733738477e-05, + "loss": 0.3546, + "step": 25974 + }, + { + "epoch": 0.6088949055597555, + "grad_norm": 3.9415860176086426, + "learning_rate": 6.647914726533191e-05, + "loss": 0.2997, + "step": 25975 + }, + { + "epoch": 0.6089183471345605, + "grad_norm": 2.6952428817749023, + "learning_rate": 6.647220737516416e-05, + "loss": 0.2748, + "step": 25976 + }, + { + "epoch": 0.6089417887093656, + "grad_norm": 2.1760334968566895, + "learning_rate": 6.646526766691914e-05, + "loss": 0.1481, + "step": 25977 + }, + { + "epoch": 0.6089652302841705, + "grad_norm": 3.716780662536621, + "learning_rate": 6.645832814063453e-05, + "loss": 0.2779, + "step": 25978 + }, + { + "epoch": 0.6089886718589755, + "grad_norm": 3.863307476043701, + "learning_rate": 6.645138879634799e-05, + "loss": 0.2044, + "step": 25979 + }, + { + "epoch": 0.6090121134337805, + "grad_norm": 3.260739326477051, + "learning_rate": 6.644444963409712e-05, + "loss": 0.18, + "step": 25980 + }, + { + "epoch": 0.6090355550085855, + "grad_norm": 4.61034631729126, + "learning_rate": 6.643751065391964e-05, + "loss": 0.3853, + "step": 25981 + }, + { + "epoch": 0.6090589965833905, + "grad_norm": 1.0494177341461182, + "learning_rate": 6.643057185585317e-05, + "loss": 0.1359, + "step": 25982 + }, + { + "epoch": 0.6090824381581955, + "grad_norm": 3.8332619667053223, + "learning_rate": 6.642363323993534e-05, + "loss": 0.5739, + "step": 25983 + }, + { + "epoch": 0.6091058797330005, + "grad_norm": 2.184870719909668, + "learning_rate": 6.641669480620383e-05, + "loss": 1.1386, + "step": 25984 + }, + { + "epoch": 0.6091293213078055, + "grad_norm": 3.432264804840088, + "learning_rate": 6.640975655469626e-05, + "loss": 0.2923, + "step": 25985 + }, + { + "epoch": 0.6091527628826104, + "grad_norm": 2.964621067047119, + "learning_rate": 6.640281848545032e-05, + "loss": 0.2166, + "step": 25986 + }, + { + "epoch": 0.6091762044574155, + "grad_norm": 3.38421893119812, + "learning_rate": 6.639588059850361e-05, + "loss": 0.3935, + "step": 25987 + }, + { + "epoch": 0.6091996460322204, + "grad_norm": 5.8074421882629395, + "learning_rate": 6.638894289389382e-05, + "loss": 0.172, + "step": 25988 + }, + { + "epoch": 0.6092230876070255, + "grad_norm": 1.1563327312469482, + "learning_rate": 6.638200537165858e-05, + "loss": 0.1024, + "step": 25989 + }, + { + "epoch": 0.6092465291818304, + "grad_norm": 3.090749502182007, + "learning_rate": 6.63750680318355e-05, + "loss": 0.3233, + "step": 25990 + }, + { + "epoch": 0.6092699707566355, + "grad_norm": 0.4277782738208771, + "learning_rate": 6.636813087446225e-05, + "loss": 0.0288, + "step": 25991 + }, + { + "epoch": 0.6092934123314404, + "grad_norm": 5.740579605102539, + "learning_rate": 6.636119389957649e-05, + "loss": 0.2947, + "step": 25992 + }, + { + "epoch": 0.6093168539062455, + "grad_norm": 2.9891154766082764, + "learning_rate": 6.635425710721582e-05, + "loss": 0.3742, + "step": 25993 + }, + { + "epoch": 0.6093402954810504, + "grad_norm": 0.8096688985824585, + "learning_rate": 6.634732049741789e-05, + "loss": 0.0998, + "step": 25994 + }, + { + "epoch": 0.6093637370558554, + "grad_norm": 5.63199520111084, + "learning_rate": 6.634038407022036e-05, + "loss": 0.5073, + "step": 25995 + }, + { + "epoch": 0.6093871786306604, + "grad_norm": 3.4398651123046875, + "learning_rate": 6.633344782566085e-05, + "loss": 0.2741, + "step": 25996 + }, + { + "epoch": 0.6094106202054654, + "grad_norm": 3.6759605407714844, + "learning_rate": 6.632651176377696e-05, + "loss": 0.4385, + "step": 25997 + }, + { + "epoch": 0.6094340617802704, + "grad_norm": 14.729351997375488, + "learning_rate": 6.63195758846064e-05, + "loss": 0.5675, + "step": 25998 + }, + { + "epoch": 0.6094575033550754, + "grad_norm": 2.7527737617492676, + "learning_rate": 6.631264018818678e-05, + "loss": 0.2527, + "step": 25999 + }, + { + "epoch": 0.6094809449298804, + "grad_norm": 2.0568487644195557, + "learning_rate": 6.630570467455568e-05, + "loss": 1.0971, + "step": 26000 + }, + { + "epoch": 0.6095043865046854, + "grad_norm": 2.8980252742767334, + "learning_rate": 6.629876934375081e-05, + "loss": 1.1972, + "step": 26001 + }, + { + "epoch": 0.6095278280794904, + "grad_norm": 0.4738779664039612, + "learning_rate": 6.629183419580978e-05, + "loss": 0.0151, + "step": 26002 + }, + { + "epoch": 0.6095512696542954, + "grad_norm": 1.7617571353912354, + "learning_rate": 6.628489923077016e-05, + "loss": 0.0832, + "step": 26003 + }, + { + "epoch": 0.6095747112291003, + "grad_norm": 3.0033481121063232, + "learning_rate": 6.627796444866965e-05, + "loss": 0.2563, + "step": 26004 + }, + { + "epoch": 0.6095981528039054, + "grad_norm": 3.379257917404175, + "learning_rate": 6.627102984954588e-05, + "loss": 0.3676, + "step": 26005 + }, + { + "epoch": 0.6096215943787103, + "grad_norm": 3.226022720336914, + "learning_rate": 6.62640954334364e-05, + "loss": 0.3286, + "step": 26006 + }, + { + "epoch": 0.6096450359535154, + "grad_norm": 1.9550057649612427, + "learning_rate": 6.62571612003789e-05, + "loss": 0.2418, + "step": 26007 + }, + { + "epoch": 0.6096684775283203, + "grad_norm": 1.3359260559082031, + "learning_rate": 6.6250227150411e-05, + "loss": 0.0549, + "step": 26008 + }, + { + "epoch": 0.6096919191031254, + "grad_norm": 0.5311167240142822, + "learning_rate": 6.624329328357032e-05, + "loss": 0.0514, + "step": 26009 + }, + { + "epoch": 0.6097153606779303, + "grad_norm": 2.561913251876831, + "learning_rate": 6.623635959989445e-05, + "loss": 0.1982, + "step": 26010 + }, + { + "epoch": 0.6097388022527354, + "grad_norm": 3.6547203063964844, + "learning_rate": 6.622942609942107e-05, + "loss": 0.2983, + "step": 26011 + }, + { + "epoch": 0.6097622438275403, + "grad_norm": 1.3225809335708618, + "learning_rate": 6.622249278218777e-05, + "loss": 0.1387, + "step": 26012 + }, + { + "epoch": 0.6097856854023453, + "grad_norm": 1.1388853788375854, + "learning_rate": 6.621555964823218e-05, + "loss": 0.0981, + "step": 26013 + }, + { + "epoch": 0.6098091269771503, + "grad_norm": 3.6970863342285156, + "learning_rate": 6.620862669759192e-05, + "loss": 0.3934, + "step": 26014 + }, + { + "epoch": 0.6098325685519553, + "grad_norm": 0.7823604345321655, + "learning_rate": 6.620169393030456e-05, + "loss": 0.0844, + "step": 26015 + }, + { + "epoch": 0.6098560101267603, + "grad_norm": 0.6633498668670654, + "learning_rate": 6.61947613464078e-05, + "loss": 0.0687, + "step": 26016 + }, + { + "epoch": 0.6098794517015653, + "grad_norm": 1.738445520401001, + "learning_rate": 6.618782894593923e-05, + "loss": 0.0615, + "step": 26017 + }, + { + "epoch": 0.6099028932763703, + "grad_norm": 1.915158987045288, + "learning_rate": 6.618089672893639e-05, + "loss": 0.0971, + "step": 26018 + }, + { + "epoch": 0.6099263348511753, + "grad_norm": 0.7737309336662292, + "learning_rate": 6.617396469543701e-05, + "loss": 0.0651, + "step": 26019 + }, + { + "epoch": 0.6099497764259803, + "grad_norm": 3.0702736377716064, + "learning_rate": 6.616703284547863e-05, + "loss": 0.2399, + "step": 26020 + }, + { + "epoch": 0.6099732180007853, + "grad_norm": 1.8555852174758911, + "learning_rate": 6.616010117909888e-05, + "loss": 0.2144, + "step": 26021 + }, + { + "epoch": 0.6099966595755902, + "grad_norm": 2.4665560722351074, + "learning_rate": 6.615316969633534e-05, + "loss": 0.2767, + "step": 26022 + }, + { + "epoch": 0.6100201011503953, + "grad_norm": 0.9802274107933044, + "learning_rate": 6.61462383972257e-05, + "loss": 0.1101, + "step": 26023 + }, + { + "epoch": 0.6100435427252002, + "grad_norm": 2.5126760005950928, + "learning_rate": 6.61393072818075e-05, + "loss": 0.1713, + "step": 26024 + }, + { + "epoch": 0.6100669843000053, + "grad_norm": 4.099970817565918, + "learning_rate": 6.613237635011837e-05, + "loss": 0.2818, + "step": 26025 + }, + { + "epoch": 0.6100904258748102, + "grad_norm": 3.6772449016571045, + "learning_rate": 6.612544560219594e-05, + "loss": 0.4656, + "step": 26026 + }, + { + "epoch": 0.6101138674496153, + "grad_norm": 2.4955124855041504, + "learning_rate": 6.611851503807779e-05, + "loss": 0.2645, + "step": 26027 + }, + { + "epoch": 0.6101373090244203, + "grad_norm": 5.105742931365967, + "learning_rate": 6.61115846578015e-05, + "loss": 0.6142, + "step": 26028 + }, + { + "epoch": 0.6101607505992253, + "grad_norm": 2.1537630558013916, + "learning_rate": 6.610465446140475e-05, + "loss": 0.1954, + "step": 26029 + }, + { + "epoch": 0.6101841921740303, + "grad_norm": 1.4386299848556519, + "learning_rate": 6.609772444892508e-05, + "loss": 0.0701, + "step": 26030 + }, + { + "epoch": 0.6102076337488352, + "grad_norm": 2.4460973739624023, + "learning_rate": 6.609079462040009e-05, + "loss": 1.1671, + "step": 26031 + }, + { + "epoch": 0.6102310753236403, + "grad_norm": 2.7770168781280518, + "learning_rate": 6.608386497586743e-05, + "loss": 1.551, + "step": 26032 + }, + { + "epoch": 0.6102545168984452, + "grad_norm": 5.3479533195495605, + "learning_rate": 6.607693551536466e-05, + "loss": 0.4879, + "step": 26033 + }, + { + "epoch": 0.6102779584732503, + "grad_norm": 3.3138232231140137, + "learning_rate": 6.607000623892936e-05, + "loss": 0.2294, + "step": 26034 + }, + { + "epoch": 0.6103014000480552, + "grad_norm": 4.629650115966797, + "learning_rate": 6.606307714659918e-05, + "loss": 0.3885, + "step": 26035 + }, + { + "epoch": 0.6103248416228603, + "grad_norm": 1.773358702659607, + "learning_rate": 6.60561482384117e-05, + "loss": 0.1935, + "step": 26036 + }, + { + "epoch": 0.6103482831976652, + "grad_norm": 2.358456611633301, + "learning_rate": 6.604921951440454e-05, + "loss": 0.2555, + "step": 26037 + }, + { + "epoch": 0.6103717247724703, + "grad_norm": 2.8678386211395264, + "learning_rate": 6.604229097461524e-05, + "loss": 0.1383, + "step": 26038 + }, + { + "epoch": 0.6103951663472752, + "grad_norm": 1.7210198640823364, + "learning_rate": 6.603536261908145e-05, + "loss": 0.1959, + "step": 26039 + }, + { + "epoch": 0.6104186079220802, + "grad_norm": 1.898645043373108, + "learning_rate": 6.602843444784072e-05, + "loss": 0.1333, + "step": 26040 + }, + { + "epoch": 0.6104420494968852, + "grad_norm": 4.703577041625977, + "learning_rate": 6.602150646093066e-05, + "loss": 0.3477, + "step": 26041 + }, + { + "epoch": 0.6104654910716902, + "grad_norm": 0.9369081854820251, + "learning_rate": 6.601457865838887e-05, + "loss": 0.1192, + "step": 26042 + }, + { + "epoch": 0.6104889326464952, + "grad_norm": 5.484609603881836, + "learning_rate": 6.600765104025294e-05, + "loss": 0.4406, + "step": 26043 + }, + { + "epoch": 0.6105123742213002, + "grad_norm": 0.6114612221717834, + "learning_rate": 6.600072360656042e-05, + "loss": 0.0206, + "step": 26044 + }, + { + "epoch": 0.6105358157961052, + "grad_norm": 2.8319191932678223, + "learning_rate": 6.599379635734896e-05, + "loss": 0.2325, + "step": 26045 + }, + { + "epoch": 0.6105592573709102, + "grad_norm": 4.0146613121032715, + "learning_rate": 6.598686929265608e-05, + "loss": 0.2974, + "step": 26046 + }, + { + "epoch": 0.6105826989457152, + "grad_norm": 0.9890760779380798, + "learning_rate": 6.597994241251943e-05, + "loss": 0.1128, + "step": 26047 + }, + { + "epoch": 0.6106061405205202, + "grad_norm": 3.407552480697632, + "learning_rate": 6.597301571697655e-05, + "loss": 0.314, + "step": 26048 + }, + { + "epoch": 0.6106295820953251, + "grad_norm": 6.680484294891357, + "learning_rate": 6.596608920606506e-05, + "loss": 0.2202, + "step": 26049 + }, + { + "epoch": 0.6106530236701302, + "grad_norm": 1.5494707822799683, + "learning_rate": 6.595916287982253e-05, + "loss": 0.8905, + "step": 26050 + }, + { + "epoch": 0.6106764652449351, + "grad_norm": 2.738959789276123, + "learning_rate": 6.595223673828655e-05, + "loss": 0.2415, + "step": 26051 + }, + { + "epoch": 0.6106999068197402, + "grad_norm": 1.794482707977295, + "learning_rate": 6.594531078149468e-05, + "loss": 1.0147, + "step": 26052 + }, + { + "epoch": 0.6107233483945451, + "grad_norm": 2.958571434020996, + "learning_rate": 6.593838500948449e-05, + "loss": 0.1445, + "step": 26053 + }, + { + "epoch": 0.6107467899693502, + "grad_norm": 3.067883253097534, + "learning_rate": 6.593145942229361e-05, + "loss": 0.3058, + "step": 26054 + }, + { + "epoch": 0.6107702315441551, + "grad_norm": 2.0073482990264893, + "learning_rate": 6.592453401995959e-05, + "loss": 0.1234, + "step": 26055 + }, + { + "epoch": 0.6107936731189602, + "grad_norm": 3.152678966522217, + "learning_rate": 6.591760880251998e-05, + "loss": 1.5777, + "step": 26056 + }, + { + "epoch": 0.6108171146937651, + "grad_norm": 3.8947486877441406, + "learning_rate": 6.591068377001241e-05, + "loss": 0.3138, + "step": 26057 + }, + { + "epoch": 0.6108405562685701, + "grad_norm": 2.7782557010650635, + "learning_rate": 6.590375892247438e-05, + "loss": 0.2731, + "step": 26058 + }, + { + "epoch": 0.6108639978433751, + "grad_norm": 2.0520901679992676, + "learning_rate": 6.589683425994355e-05, + "loss": 0.177, + "step": 26059 + }, + { + "epoch": 0.6108874394181801, + "grad_norm": 0.8768372535705566, + "learning_rate": 6.588990978245745e-05, + "loss": 0.0648, + "step": 26060 + }, + { + "epoch": 0.6109108809929851, + "grad_norm": 2.4999289512634277, + "learning_rate": 6.588298549005367e-05, + "loss": 0.185, + "step": 26061 + }, + { + "epoch": 0.6109343225677901, + "grad_norm": 1.9828461408615112, + "learning_rate": 6.587606138276979e-05, + "loss": 0.1258, + "step": 26062 + }, + { + "epoch": 0.6109577641425951, + "grad_norm": 1.3615353107452393, + "learning_rate": 6.586913746064332e-05, + "loss": 0.182, + "step": 26063 + }, + { + "epoch": 0.6109812057174001, + "grad_norm": 2.9993155002593994, + "learning_rate": 6.586221372371188e-05, + "loss": 0.2927, + "step": 26064 + }, + { + "epoch": 0.611004647292205, + "grad_norm": 3.776930332183838, + "learning_rate": 6.585529017201304e-05, + "loss": 0.3048, + "step": 26065 + }, + { + "epoch": 0.6110280888670101, + "grad_norm": 2.60430908203125, + "learning_rate": 6.584836680558434e-05, + "loss": 0.1755, + "step": 26066 + }, + { + "epoch": 0.611051530441815, + "grad_norm": 2.839780330657959, + "learning_rate": 6.584144362446337e-05, + "loss": 0.273, + "step": 26067 + }, + { + "epoch": 0.6110749720166201, + "grad_norm": 3.2304744720458984, + "learning_rate": 6.58345206286877e-05, + "loss": 0.2254, + "step": 26068 + }, + { + "epoch": 0.611098413591425, + "grad_norm": 3.104397773742676, + "learning_rate": 6.582759781829486e-05, + "loss": 1.5706, + "step": 26069 + }, + { + "epoch": 0.6111218551662301, + "grad_norm": 2.815002679824829, + "learning_rate": 6.58206751933224e-05, + "loss": 0.1887, + "step": 26070 + }, + { + "epoch": 0.611145296741035, + "grad_norm": 2.624382257461548, + "learning_rate": 6.581375275380798e-05, + "loss": 0.2287, + "step": 26071 + }, + { + "epoch": 0.6111687383158401, + "grad_norm": 2.7742810249328613, + "learning_rate": 6.580683049978908e-05, + "loss": 0.2501, + "step": 26072 + }, + { + "epoch": 0.611192179890645, + "grad_norm": 2.0292696952819824, + "learning_rate": 6.579990843130328e-05, + "loss": 0.1989, + "step": 26073 + }, + { + "epoch": 0.61121562146545, + "grad_norm": 2.613469123840332, + "learning_rate": 6.579298654838814e-05, + "loss": 0.1093, + "step": 26074 + }, + { + "epoch": 0.611239063040255, + "grad_norm": 2.4980320930480957, + "learning_rate": 6.578606485108123e-05, + "loss": 0.298, + "step": 26075 + }, + { + "epoch": 0.61126250461506, + "grad_norm": 5.917455673217773, + "learning_rate": 6.577914333942005e-05, + "loss": 0.5033, + "step": 26076 + }, + { + "epoch": 0.611285946189865, + "grad_norm": 2.8069379329681396, + "learning_rate": 6.577222201344224e-05, + "loss": 0.2082, + "step": 26077 + }, + { + "epoch": 0.61130938776467, + "grad_norm": 0.3948637545108795, + "learning_rate": 6.576530087318533e-05, + "loss": 0.0486, + "step": 26078 + }, + { + "epoch": 0.6113328293394751, + "grad_norm": 1.3709263801574707, + "learning_rate": 6.57583799186868e-05, + "loss": 0.1229, + "step": 26079 + }, + { + "epoch": 0.61135627091428, + "grad_norm": 1.23460054397583, + "learning_rate": 6.575145914998432e-05, + "loss": 0.0698, + "step": 26080 + }, + { + "epoch": 0.6113797124890851, + "grad_norm": 4.258566856384277, + "learning_rate": 6.574453856711538e-05, + "loss": 0.3238, + "step": 26081 + }, + { + "epoch": 0.61140315406389, + "grad_norm": 2.2092068195343018, + "learning_rate": 6.57376181701175e-05, + "loss": 1.4937, + "step": 26082 + }, + { + "epoch": 0.6114265956386951, + "grad_norm": 2.698538064956665, + "learning_rate": 6.573069795902827e-05, + "loss": 1.1926, + "step": 26083 + }, + { + "epoch": 0.6114500372135, + "grad_norm": 1.5745211839675903, + "learning_rate": 6.572377793388525e-05, + "loss": 0.0705, + "step": 26084 + }, + { + "epoch": 0.611473478788305, + "grad_norm": 5.071122646331787, + "learning_rate": 6.571685809472599e-05, + "loss": 0.3343, + "step": 26085 + }, + { + "epoch": 0.61149692036311, + "grad_norm": 3.343332529067993, + "learning_rate": 6.570993844158799e-05, + "loss": 0.2469, + "step": 26086 + }, + { + "epoch": 0.611520361937915, + "grad_norm": 2.6620917320251465, + "learning_rate": 6.570301897450887e-05, + "loss": 0.2245, + "step": 26087 + }, + { + "epoch": 0.61154380351272, + "grad_norm": 0.6988462805747986, + "learning_rate": 6.56960996935261e-05, + "loss": 0.0561, + "step": 26088 + }, + { + "epoch": 0.611567245087525, + "grad_norm": 1.8398357629776, + "learning_rate": 6.568918059867728e-05, + "loss": 0.1272, + "step": 26089 + }, + { + "epoch": 0.61159068666233, + "grad_norm": 3.2466349601745605, + "learning_rate": 6.568226168999992e-05, + "loss": 0.1812, + "step": 26090 + }, + { + "epoch": 0.611614128237135, + "grad_norm": 0.2696491479873657, + "learning_rate": 6.567534296753157e-05, + "loss": 0.0269, + "step": 26091 + }, + { + "epoch": 0.61163756981194, + "grad_norm": 3.7657933235168457, + "learning_rate": 6.566842443130978e-05, + "loss": 0.2718, + "step": 26092 + }, + { + "epoch": 0.611661011386745, + "grad_norm": 2.752871513366699, + "learning_rate": 6.566150608137209e-05, + "loss": 0.2199, + "step": 26093 + }, + { + "epoch": 0.6116844529615499, + "grad_norm": 0.7675556540489197, + "learning_rate": 6.565458791775603e-05, + "loss": 0.0511, + "step": 26094 + }, + { + "epoch": 0.611707894536355, + "grad_norm": 0.9863309860229492, + "learning_rate": 6.564766994049909e-05, + "loss": 0.0774, + "step": 26095 + }, + { + "epoch": 0.6117313361111599, + "grad_norm": 2.3183834552764893, + "learning_rate": 6.564075214963892e-05, + "loss": 0.0926, + "step": 26096 + }, + { + "epoch": 0.611754777685965, + "grad_norm": 1.808395504951477, + "learning_rate": 6.563383454521299e-05, + "loss": 0.1457, + "step": 26097 + }, + { + "epoch": 0.6117782192607699, + "grad_norm": 3.438153028488159, + "learning_rate": 6.562691712725882e-05, + "loss": 0.3543, + "step": 26098 + }, + { + "epoch": 0.611801660835575, + "grad_norm": 3.389460563659668, + "learning_rate": 6.561999989581398e-05, + "loss": 0.281, + "step": 26099 + }, + { + "epoch": 0.6118251024103799, + "grad_norm": 2.3293426036834717, + "learning_rate": 6.5613082850916e-05, + "loss": 0.2046, + "step": 26100 + }, + { + "epoch": 0.611848543985185, + "grad_norm": 5.927170753479004, + "learning_rate": 6.560616599260236e-05, + "loss": 0.6908, + "step": 26101 + }, + { + "epoch": 0.6118719855599899, + "grad_norm": 3.7126731872558594, + "learning_rate": 6.559924932091066e-05, + "loss": 0.2064, + "step": 26102 + }, + { + "epoch": 0.611895427134795, + "grad_norm": 0.693878710269928, + "learning_rate": 6.559233283587841e-05, + "loss": 0.0688, + "step": 26103 + }, + { + "epoch": 0.6119188687095999, + "grad_norm": 3.7731387615203857, + "learning_rate": 6.55854165375431e-05, + "loss": 0.457, + "step": 26104 + }, + { + "epoch": 0.6119423102844049, + "grad_norm": 2.408597707748413, + "learning_rate": 6.557850042594232e-05, + "loss": 1.8757, + "step": 26105 + }, + { + "epoch": 0.6119657518592099, + "grad_norm": 1.0456690788269043, + "learning_rate": 6.557158450111354e-05, + "loss": 0.0701, + "step": 26106 + }, + { + "epoch": 0.6119891934340149, + "grad_norm": 3.228687047958374, + "learning_rate": 6.556466876309428e-05, + "loss": 0.2945, + "step": 26107 + }, + { + "epoch": 0.6120126350088199, + "grad_norm": 1.8566391468048096, + "learning_rate": 6.555775321192212e-05, + "loss": 0.0641, + "step": 26108 + }, + { + "epoch": 0.6120360765836249, + "grad_norm": 4.708924770355225, + "learning_rate": 6.555083784763458e-05, + "loss": 0.3427, + "step": 26109 + }, + { + "epoch": 0.6120595181584298, + "grad_norm": 0.5944734811782837, + "learning_rate": 6.554392267026916e-05, + "loss": 0.0711, + "step": 26110 + }, + { + "epoch": 0.6120829597332349, + "grad_norm": 3.431942939758301, + "learning_rate": 6.553700767986337e-05, + "loss": 0.2209, + "step": 26111 + }, + { + "epoch": 0.6121064013080398, + "grad_norm": 3.365314483642578, + "learning_rate": 6.553009287645475e-05, + "loss": 0.1551, + "step": 26112 + }, + { + "epoch": 0.6121298428828449, + "grad_norm": 3.6332356929779053, + "learning_rate": 6.552317826008083e-05, + "loss": 0.3549, + "step": 26113 + }, + { + "epoch": 0.6121532844576498, + "grad_norm": 0.724147379398346, + "learning_rate": 6.551626383077908e-05, + "loss": 0.0703, + "step": 26114 + }, + { + "epoch": 0.6121767260324549, + "grad_norm": 2.752213954925537, + "learning_rate": 6.55093495885871e-05, + "loss": 0.1334, + "step": 26115 + }, + { + "epoch": 0.6122001676072598, + "grad_norm": 0.6789582967758179, + "learning_rate": 6.550243553354233e-05, + "loss": 0.086, + "step": 26116 + }, + { + "epoch": 0.6122236091820649, + "grad_norm": 2.8501408100128174, + "learning_rate": 6.549552166568231e-05, + "loss": 0.3106, + "step": 26117 + }, + { + "epoch": 0.6122470507568698, + "grad_norm": 3.4750618934631348, + "learning_rate": 6.548860798504459e-05, + "loss": 0.4618, + "step": 26118 + }, + { + "epoch": 0.6122704923316749, + "grad_norm": 1.508594036102295, + "learning_rate": 6.548169449166659e-05, + "loss": 0.205, + "step": 26119 + }, + { + "epoch": 0.6122939339064798, + "grad_norm": 1.2156620025634766, + "learning_rate": 6.547478118558595e-05, + "loss": 0.1039, + "step": 26120 + }, + { + "epoch": 0.6123173754812848, + "grad_norm": 2.54762864112854, + "learning_rate": 6.546786806684006e-05, + "loss": 0.1568, + "step": 26121 + }, + { + "epoch": 0.6123408170560898, + "grad_norm": 3.1522951126098633, + "learning_rate": 6.546095513546654e-05, + "loss": 0.2205, + "step": 26122 + }, + { + "epoch": 0.6123642586308948, + "grad_norm": 0.5052512288093567, + "learning_rate": 6.545404239150284e-05, + "loss": 0.0474, + "step": 26123 + }, + { + "epoch": 0.6123877002056998, + "grad_norm": 2.608531951904297, + "learning_rate": 6.544712983498647e-05, + "loss": 0.2444, + "step": 26124 + }, + { + "epoch": 0.6124111417805048, + "grad_norm": 3.8015785217285156, + "learning_rate": 6.544021746595495e-05, + "loss": 0.3682, + "step": 26125 + }, + { + "epoch": 0.6124345833553098, + "grad_norm": 3.313426971435547, + "learning_rate": 6.543330528444576e-05, + "loss": 0.32, + "step": 26126 + }, + { + "epoch": 0.6124580249301148, + "grad_norm": 4.152525901794434, + "learning_rate": 6.542639329049645e-05, + "loss": 0.5116, + "step": 26127 + }, + { + "epoch": 0.6124814665049197, + "grad_norm": 1.17949378490448, + "learning_rate": 6.541948148414451e-05, + "loss": 0.1094, + "step": 26128 + }, + { + "epoch": 0.6125049080797248, + "grad_norm": 0.5798752307891846, + "learning_rate": 6.54125698654274e-05, + "loss": 0.0409, + "step": 26129 + }, + { + "epoch": 0.6125283496545298, + "grad_norm": 3.1593358516693115, + "learning_rate": 6.540565843438269e-05, + "loss": 1.6657, + "step": 26130 + }, + { + "epoch": 0.6125517912293348, + "grad_norm": 1.8537732362747192, + "learning_rate": 6.539874719104782e-05, + "loss": 0.1012, + "step": 26131 + }, + { + "epoch": 0.6125752328041398, + "grad_norm": 3.5094263553619385, + "learning_rate": 6.539183613546033e-05, + "loss": 0.2348, + "step": 26132 + }, + { + "epoch": 0.6125986743789448, + "grad_norm": 5.008612155914307, + "learning_rate": 6.538492526765771e-05, + "loss": 0.2703, + "step": 26133 + }, + { + "epoch": 0.6126221159537498, + "grad_norm": 3.0417520999908447, + "learning_rate": 6.53780145876775e-05, + "loss": 0.9736, + "step": 26134 + }, + { + "epoch": 0.6126455575285548, + "grad_norm": 4.486577987670898, + "learning_rate": 6.537110409555712e-05, + "loss": 0.3983, + "step": 26135 + }, + { + "epoch": 0.6126689991033598, + "grad_norm": 0.9240790605545044, + "learning_rate": 6.536419379133409e-05, + "loss": 0.1428, + "step": 26136 + }, + { + "epoch": 0.6126924406781648, + "grad_norm": 2.831897497177124, + "learning_rate": 6.535728367504597e-05, + "loss": 0.2858, + "step": 26137 + }, + { + "epoch": 0.6127158822529698, + "grad_norm": 2.3056228160858154, + "learning_rate": 6.535037374673017e-05, + "loss": 0.1822, + "step": 26138 + }, + { + "epoch": 0.6127393238277747, + "grad_norm": 2.462538242340088, + "learning_rate": 6.53434640064242e-05, + "loss": 0.1756, + "step": 26139 + }, + { + "epoch": 0.6127627654025798, + "grad_norm": 1.3849116563796997, + "learning_rate": 6.53365544541656e-05, + "loss": 0.0938, + "step": 26140 + }, + { + "epoch": 0.6127862069773847, + "grad_norm": 4.691567897796631, + "learning_rate": 6.532964508999182e-05, + "loss": 0.4029, + "step": 26141 + }, + { + "epoch": 0.6128096485521898, + "grad_norm": 2.7462880611419678, + "learning_rate": 6.532273591394035e-05, + "loss": 0.2121, + "step": 26142 + }, + { + "epoch": 0.6128330901269947, + "grad_norm": 2.0776052474975586, + "learning_rate": 6.531582692604865e-05, + "loss": 0.1557, + "step": 26143 + }, + { + "epoch": 0.6128565317017998, + "grad_norm": 3.4707961082458496, + "learning_rate": 6.530891812635431e-05, + "loss": 0.3086, + "step": 26144 + }, + { + "epoch": 0.6128799732766047, + "grad_norm": 2.5804243087768555, + "learning_rate": 6.530200951489475e-05, + "loss": 0.1775, + "step": 26145 + }, + { + "epoch": 0.6129034148514098, + "grad_norm": 4.913031101226807, + "learning_rate": 6.529510109170745e-05, + "loss": 0.6951, + "step": 26146 + }, + { + "epoch": 0.6129268564262147, + "grad_norm": 0.9924241900444031, + "learning_rate": 6.528819285682991e-05, + "loss": 0.0229, + "step": 26147 + }, + { + "epoch": 0.6129502980010197, + "grad_norm": 0.7021187543869019, + "learning_rate": 6.528128481029963e-05, + "loss": 0.0384, + "step": 26148 + }, + { + "epoch": 0.6129737395758247, + "grad_norm": 1.5638220310211182, + "learning_rate": 6.527437695215403e-05, + "loss": 0.7328, + "step": 26149 + }, + { + "epoch": 0.6129971811506297, + "grad_norm": 1.7761800289154053, + "learning_rate": 6.526746928243069e-05, + "loss": 0.1403, + "step": 26150 + }, + { + "epoch": 0.6130206227254347, + "grad_norm": 1.7385772466659546, + "learning_rate": 6.5260561801167e-05, + "loss": 0.1279, + "step": 26151 + }, + { + "epoch": 0.6130440643002397, + "grad_norm": 1.699958324432373, + "learning_rate": 6.525365450840049e-05, + "loss": 0.1111, + "step": 26152 + }, + { + "epoch": 0.6130675058750447, + "grad_norm": 4.560554027557373, + "learning_rate": 6.524674740416863e-05, + "loss": 0.2236, + "step": 26153 + }, + { + "epoch": 0.6130909474498497, + "grad_norm": 1.8882317543029785, + "learning_rate": 6.523984048850891e-05, + "loss": 0.2028, + "step": 26154 + }, + { + "epoch": 0.6131143890246546, + "grad_norm": 2.7794034481048584, + "learning_rate": 6.523293376145875e-05, + "loss": 0.1769, + "step": 26155 + }, + { + "epoch": 0.6131378305994597, + "grad_norm": 0.7784814834594727, + "learning_rate": 6.522602722305567e-05, + "loss": 0.098, + "step": 26156 + }, + { + "epoch": 0.6131612721742646, + "grad_norm": 0.4808652997016907, + "learning_rate": 6.521912087333716e-05, + "loss": 0.0598, + "step": 26157 + }, + { + "epoch": 0.6131847137490697, + "grad_norm": 3.4091947078704834, + "learning_rate": 6.521221471234072e-05, + "loss": 0.4179, + "step": 26158 + }, + { + "epoch": 0.6132081553238746, + "grad_norm": 0.8782337307929993, + "learning_rate": 6.520530874010374e-05, + "loss": 0.0694, + "step": 26159 + }, + { + "epoch": 0.6132315968986797, + "grad_norm": 4.105627059936523, + "learning_rate": 6.519840295666373e-05, + "loss": 0.3657, + "step": 26160 + }, + { + "epoch": 0.6132550384734846, + "grad_norm": 2.642913341522217, + "learning_rate": 6.519149736205819e-05, + "loss": 0.2194, + "step": 26161 + }, + { + "epoch": 0.6132784800482897, + "grad_norm": 0.7051386833190918, + "learning_rate": 6.518459195632454e-05, + "loss": 0.0266, + "step": 26162 + }, + { + "epoch": 0.6133019216230946, + "grad_norm": 1.0878758430480957, + "learning_rate": 6.517768673950028e-05, + "loss": 0.0939, + "step": 26163 + }, + { + "epoch": 0.6133253631978997, + "grad_norm": 0.9590646028518677, + "learning_rate": 6.517078171162287e-05, + "loss": 0.0806, + "step": 26164 + }, + { + "epoch": 0.6133488047727046, + "grad_norm": 2.4164867401123047, + "learning_rate": 6.51638768727298e-05, + "loss": 0.1295, + "step": 26165 + }, + { + "epoch": 0.6133722463475096, + "grad_norm": 4.561418056488037, + "learning_rate": 6.515697222285849e-05, + "loss": 0.49, + "step": 26166 + }, + { + "epoch": 0.6133956879223146, + "grad_norm": 1.3328516483306885, + "learning_rate": 6.515006776204642e-05, + "loss": 0.1315, + "step": 26167 + }, + { + "epoch": 0.6134191294971196, + "grad_norm": 3.7008919715881348, + "learning_rate": 6.514316349033105e-05, + "loss": 0.3511, + "step": 26168 + }, + { + "epoch": 0.6134425710719246, + "grad_norm": 2.8929500579833984, + "learning_rate": 6.513625940774987e-05, + "loss": 0.1766, + "step": 26169 + }, + { + "epoch": 0.6134660126467296, + "grad_norm": 2.749004364013672, + "learning_rate": 6.512935551434036e-05, + "loss": 0.2497, + "step": 26170 + }, + { + "epoch": 0.6134894542215346, + "grad_norm": 2.583345651626587, + "learning_rate": 6.512245181013992e-05, + "loss": 0.2048, + "step": 26171 + }, + { + "epoch": 0.6135128957963396, + "grad_norm": 2.1393730640411377, + "learning_rate": 6.511554829518604e-05, + "loss": 0.1053, + "step": 26172 + }, + { + "epoch": 0.6135363373711445, + "grad_norm": 2.00136661529541, + "learning_rate": 6.510864496951619e-05, + "loss": 0.1646, + "step": 26173 + }, + { + "epoch": 0.6135597789459496, + "grad_norm": 1.1235498189926147, + "learning_rate": 6.51017418331678e-05, + "loss": 0.0915, + "step": 26174 + }, + { + "epoch": 0.6135832205207545, + "grad_norm": 2.41642165184021, + "learning_rate": 6.509483888617833e-05, + "loss": 0.2567, + "step": 26175 + }, + { + "epoch": 0.6136066620955596, + "grad_norm": 2.277080535888672, + "learning_rate": 6.508793612858528e-05, + "loss": 0.1492, + "step": 26176 + }, + { + "epoch": 0.6136301036703645, + "grad_norm": 0.844016432762146, + "learning_rate": 6.508103356042603e-05, + "loss": 0.0841, + "step": 26177 + }, + { + "epoch": 0.6136535452451696, + "grad_norm": 3.0620005130767822, + "learning_rate": 6.50741311817381e-05, + "loss": 0.1646, + "step": 26178 + }, + { + "epoch": 0.6136769868199745, + "grad_norm": 2.6080245971679688, + "learning_rate": 6.506722899255892e-05, + "loss": 0.2289, + "step": 26179 + }, + { + "epoch": 0.6137004283947796, + "grad_norm": 2.50632381439209, + "learning_rate": 6.506032699292588e-05, + "loss": 0.3726, + "step": 26180 + }, + { + "epoch": 0.6137238699695845, + "grad_norm": 4.109777450561523, + "learning_rate": 6.505342518287653e-05, + "loss": 0.4513, + "step": 26181 + }, + { + "epoch": 0.6137473115443896, + "grad_norm": 5.215284824371338, + "learning_rate": 6.504652356244829e-05, + "loss": 0.3635, + "step": 26182 + }, + { + "epoch": 0.6137707531191946, + "grad_norm": 2.84303617477417, + "learning_rate": 6.50396221316786e-05, + "loss": 0.1631, + "step": 26183 + }, + { + "epoch": 0.6137941946939995, + "grad_norm": 1.7750465869903564, + "learning_rate": 6.503272089060486e-05, + "loss": 0.1639, + "step": 26184 + }, + { + "epoch": 0.6138176362688046, + "grad_norm": 4.447951793670654, + "learning_rate": 6.50258198392646e-05, + "loss": 0.5517, + "step": 26185 + }, + { + "epoch": 0.6138410778436095, + "grad_norm": 0.5858228206634521, + "learning_rate": 6.501891897769523e-05, + "loss": 0.0379, + "step": 26186 + }, + { + "epoch": 0.6138645194184146, + "grad_norm": 0.3630009889602661, + "learning_rate": 6.501201830593415e-05, + "loss": 0.023, + "step": 26187 + }, + { + "epoch": 0.6138879609932195, + "grad_norm": 3.3807272911071777, + "learning_rate": 6.500511782401888e-05, + "loss": 0.2519, + "step": 26188 + }, + { + "epoch": 0.6139114025680246, + "grad_norm": 2.310732126235962, + "learning_rate": 6.499821753198683e-05, + "loss": 0.2429, + "step": 26189 + }, + { + "epoch": 0.6139348441428295, + "grad_norm": 1.4957400560379028, + "learning_rate": 6.499131742987538e-05, + "loss": 0.0839, + "step": 26190 + }, + { + "epoch": 0.6139582857176346, + "grad_norm": 0.9554202556610107, + "learning_rate": 6.498441751772208e-05, + "loss": 0.086, + "step": 26191 + }, + { + "epoch": 0.6139817272924395, + "grad_norm": 3.950178861618042, + "learning_rate": 6.497751779556426e-05, + "loss": 0.6149, + "step": 26192 + }, + { + "epoch": 0.6140051688672445, + "grad_norm": 2.796522617340088, + "learning_rate": 6.497061826343946e-05, + "loss": 0.1622, + "step": 26193 + }, + { + "epoch": 0.6140286104420495, + "grad_norm": 1.2962180376052856, + "learning_rate": 6.496371892138506e-05, + "loss": 0.085, + "step": 26194 + }, + { + "epoch": 0.6140520520168545, + "grad_norm": 0.8122990727424622, + "learning_rate": 6.495681976943851e-05, + "loss": 0.0606, + "step": 26195 + }, + { + "epoch": 0.6140754935916595, + "grad_norm": 2.978168487548828, + "learning_rate": 6.494992080763724e-05, + "loss": 0.2697, + "step": 26196 + }, + { + "epoch": 0.6140989351664645, + "grad_norm": 3.0124988555908203, + "learning_rate": 6.494302203601867e-05, + "loss": 0.134, + "step": 26197 + }, + { + "epoch": 0.6141223767412695, + "grad_norm": 0.6790000200271606, + "learning_rate": 6.493612345462028e-05, + "loss": 0.0732, + "step": 26198 + }, + { + "epoch": 0.6141458183160745, + "grad_norm": 2.548257350921631, + "learning_rate": 6.492922506347946e-05, + "loss": 1.5791, + "step": 26199 + }, + { + "epoch": 0.6141692598908794, + "grad_norm": 1.0677506923675537, + "learning_rate": 6.492232686263363e-05, + "loss": 0.0996, + "step": 26200 + }, + { + "epoch": 0.6141927014656845, + "grad_norm": 2.342611312866211, + "learning_rate": 6.491542885212026e-05, + "loss": 0.1009, + "step": 26201 + }, + { + "epoch": 0.6142161430404894, + "grad_norm": 1.6217350959777832, + "learning_rate": 6.490853103197674e-05, + "loss": 0.1288, + "step": 26202 + }, + { + "epoch": 0.6142395846152945, + "grad_norm": 5.0207295417785645, + "learning_rate": 6.490163340224055e-05, + "loss": 0.2024, + "step": 26203 + }, + { + "epoch": 0.6142630261900994, + "grad_norm": 1.9718303680419922, + "learning_rate": 6.489473596294902e-05, + "loss": 0.1473, + "step": 26204 + }, + { + "epoch": 0.6142864677649045, + "grad_norm": 3.1557810306549072, + "learning_rate": 6.48878387141397e-05, + "loss": 0.1783, + "step": 26205 + }, + { + "epoch": 0.6143099093397094, + "grad_norm": 2.6830811500549316, + "learning_rate": 6.488094165584992e-05, + "loss": 0.2777, + "step": 26206 + }, + { + "epoch": 0.6143333509145145, + "grad_norm": 0.8599252700805664, + "learning_rate": 6.487404478811717e-05, + "loss": 0.0748, + "step": 26207 + }, + { + "epoch": 0.6143567924893194, + "grad_norm": 1.0079107284545898, + "learning_rate": 6.486714811097883e-05, + "loss": 0.068, + "step": 26208 + }, + { + "epoch": 0.6143802340641245, + "grad_norm": 2.691555976867676, + "learning_rate": 6.486025162447233e-05, + "loss": 1.5498, + "step": 26209 + }, + { + "epoch": 0.6144036756389294, + "grad_norm": 3.6111319065093994, + "learning_rate": 6.48533553286351e-05, + "loss": 0.247, + "step": 26210 + }, + { + "epoch": 0.6144271172137344, + "grad_norm": 2.9975461959838867, + "learning_rate": 6.484645922350454e-05, + "loss": 1.6402, + "step": 26211 + }, + { + "epoch": 0.6144505587885394, + "grad_norm": 2.657618522644043, + "learning_rate": 6.483956330911808e-05, + "loss": 0.3386, + "step": 26212 + }, + { + "epoch": 0.6144740003633444, + "grad_norm": 3.6851465702056885, + "learning_rate": 6.483266758551315e-05, + "loss": 0.1692, + "step": 26213 + }, + { + "epoch": 0.6144974419381494, + "grad_norm": 3.382880210876465, + "learning_rate": 6.482577205272715e-05, + "loss": 0.3963, + "step": 26214 + }, + { + "epoch": 0.6145208835129544, + "grad_norm": 0.8454932570457458, + "learning_rate": 6.48188767107975e-05, + "loss": 0.0751, + "step": 26215 + }, + { + "epoch": 0.6145443250877594, + "grad_norm": 1.0817118883132935, + "learning_rate": 6.481198155976157e-05, + "loss": 0.0577, + "step": 26216 + }, + { + "epoch": 0.6145677666625644, + "grad_norm": 0.7881677746772766, + "learning_rate": 6.480508659965688e-05, + "loss": 0.0644, + "step": 26217 + }, + { + "epoch": 0.6145912082373693, + "grad_norm": 0.7268195748329163, + "learning_rate": 6.479819183052076e-05, + "loss": 0.0561, + "step": 26218 + }, + { + "epoch": 0.6146146498121744, + "grad_norm": 3.177656650543213, + "learning_rate": 6.479129725239061e-05, + "loss": 0.2043, + "step": 26219 + }, + { + "epoch": 0.6146380913869793, + "grad_norm": 2.2529923915863037, + "learning_rate": 6.478440286530392e-05, + "loss": 0.179, + "step": 26220 + }, + { + "epoch": 0.6146615329617844, + "grad_norm": 0.6860307455062866, + "learning_rate": 6.477750866929804e-05, + "loss": 0.0754, + "step": 26221 + }, + { + "epoch": 0.6146849745365893, + "grad_norm": 0.3652111291885376, + "learning_rate": 6.477061466441037e-05, + "loss": 0.0304, + "step": 26222 + }, + { + "epoch": 0.6147084161113944, + "grad_norm": 2.8761067390441895, + "learning_rate": 6.476372085067835e-05, + "loss": 0.2846, + "step": 26223 + }, + { + "epoch": 0.6147318576861993, + "grad_norm": 1.8698177337646484, + "learning_rate": 6.475682722813938e-05, + "loss": 0.122, + "step": 26224 + }, + { + "epoch": 0.6147552992610044, + "grad_norm": 0.7129626274108887, + "learning_rate": 6.474993379683083e-05, + "loss": 0.0795, + "step": 26225 + }, + { + "epoch": 0.6147787408358093, + "grad_norm": 0.9054706692695618, + "learning_rate": 6.474304055679014e-05, + "loss": 0.0742, + "step": 26226 + }, + { + "epoch": 0.6148021824106144, + "grad_norm": 2.4485526084899902, + "learning_rate": 6.473614750805472e-05, + "loss": 0.2207, + "step": 26227 + }, + { + "epoch": 0.6148256239854193, + "grad_norm": 2.682427406311035, + "learning_rate": 6.472925465066191e-05, + "loss": 0.1945, + "step": 26228 + }, + { + "epoch": 0.6148490655602243, + "grad_norm": 2.001363754272461, + "learning_rate": 6.472236198464919e-05, + "loss": 0.1527, + "step": 26229 + }, + { + "epoch": 0.6148725071350293, + "grad_norm": 3.35659122467041, + "learning_rate": 6.471546951005392e-05, + "loss": 0.257, + "step": 26230 + }, + { + "epoch": 0.6148959487098343, + "grad_norm": 1.3041313886642456, + "learning_rate": 6.470857722691352e-05, + "loss": 0.115, + "step": 26231 + }, + { + "epoch": 0.6149193902846393, + "grad_norm": 2.5944385528564453, + "learning_rate": 6.470168513526535e-05, + "loss": 0.1441, + "step": 26232 + }, + { + "epoch": 0.6149428318594443, + "grad_norm": 3.009424924850464, + "learning_rate": 6.469479323514684e-05, + "loss": 0.1978, + "step": 26233 + }, + { + "epoch": 0.6149662734342494, + "grad_norm": 1.630314588546753, + "learning_rate": 6.468790152659539e-05, + "loss": 0.1062, + "step": 26234 + }, + { + "epoch": 0.6149897150090543, + "grad_norm": 2.6668996810913086, + "learning_rate": 6.468101000964834e-05, + "loss": 0.1384, + "step": 26235 + }, + { + "epoch": 0.6150131565838594, + "grad_norm": 1.903954267501831, + "learning_rate": 6.467411868434315e-05, + "loss": 0.0567, + "step": 26236 + }, + { + "epoch": 0.6150365981586643, + "grad_norm": 2.8699798583984375, + "learning_rate": 6.466722755071717e-05, + "loss": 0.4509, + "step": 26237 + }, + { + "epoch": 0.6150600397334693, + "grad_norm": 2.1692488193511963, + "learning_rate": 6.466033660880784e-05, + "loss": 0.2444, + "step": 26238 + }, + { + "epoch": 0.6150834813082743, + "grad_norm": 3.5553205013275146, + "learning_rate": 6.46534458586525e-05, + "loss": 0.1982, + "step": 26239 + }, + { + "epoch": 0.6151069228830793, + "grad_norm": 1.9175468683242798, + "learning_rate": 6.464655530028853e-05, + "loss": 0.1977, + "step": 26240 + }, + { + "epoch": 0.6151303644578843, + "grad_norm": 2.5653092861175537, + "learning_rate": 6.463966493375335e-05, + "loss": 0.4379, + "step": 26241 + }, + { + "epoch": 0.6151538060326893, + "grad_norm": 2.5584287643432617, + "learning_rate": 6.463277475908436e-05, + "loss": 0.1418, + "step": 26242 + }, + { + "epoch": 0.6151772476074943, + "grad_norm": 1.857008695602417, + "learning_rate": 6.462588477631893e-05, + "loss": 0.1869, + "step": 26243 + }, + { + "epoch": 0.6152006891822993, + "grad_norm": 2.1068806648254395, + "learning_rate": 6.461899498549442e-05, + "loss": 0.1973, + "step": 26244 + }, + { + "epoch": 0.6152241307571042, + "grad_norm": 1.3592007160186768, + "learning_rate": 6.461210538664829e-05, + "loss": 0.0999, + "step": 26245 + }, + { + "epoch": 0.6152475723319093, + "grad_norm": 0.9470353722572327, + "learning_rate": 6.460521597981783e-05, + "loss": 0.0748, + "step": 26246 + }, + { + "epoch": 0.6152710139067142, + "grad_norm": 3.0882129669189453, + "learning_rate": 6.459832676504048e-05, + "loss": 0.2022, + "step": 26247 + }, + { + "epoch": 0.6152944554815193, + "grad_norm": 2.9682040214538574, + "learning_rate": 6.459143774235359e-05, + "loss": 0.2051, + "step": 26248 + }, + { + "epoch": 0.6153178970563242, + "grad_norm": 3.6550028324127197, + "learning_rate": 6.458454891179458e-05, + "loss": 0.2014, + "step": 26249 + }, + { + "epoch": 0.6153413386311293, + "grad_norm": 2.1749348640441895, + "learning_rate": 6.457766027340078e-05, + "loss": 0.18, + "step": 26250 + }, + { + "epoch": 0.6153647802059342, + "grad_norm": 3.9877097606658936, + "learning_rate": 6.457077182720958e-05, + "loss": 0.3022, + "step": 26251 + }, + { + "epoch": 0.6153882217807393, + "grad_norm": 2.1779916286468506, + "learning_rate": 6.456388357325841e-05, + "loss": 0.1512, + "step": 26252 + }, + { + "epoch": 0.6154116633555442, + "grad_norm": 3.281533718109131, + "learning_rate": 6.455699551158454e-05, + "loss": 0.5543, + "step": 26253 + }, + { + "epoch": 0.6154351049303493, + "grad_norm": 0.7622866034507751, + "learning_rate": 6.455010764222544e-05, + "loss": 0.0475, + "step": 26254 + }, + { + "epoch": 0.6154585465051542, + "grad_norm": 1.6699320077896118, + "learning_rate": 6.454321996521847e-05, + "loss": 0.1932, + "step": 26255 + }, + { + "epoch": 0.6154819880799592, + "grad_norm": 3.8271496295928955, + "learning_rate": 6.453633248060099e-05, + "loss": 0.1903, + "step": 26256 + }, + { + "epoch": 0.6155054296547642, + "grad_norm": 2.3681936264038086, + "learning_rate": 6.452944518841032e-05, + "loss": 1.1246, + "step": 26257 + }, + { + "epoch": 0.6155288712295692, + "grad_norm": 1.1117942333221436, + "learning_rate": 6.452255808868394e-05, + "loss": 0.1304, + "step": 26258 + }, + { + "epoch": 0.6155523128043742, + "grad_norm": 3.5488076210021973, + "learning_rate": 6.451567118145914e-05, + "loss": 0.2168, + "step": 26259 + }, + { + "epoch": 0.6155757543791792, + "grad_norm": 0.5252986550331116, + "learning_rate": 6.450878446677328e-05, + "loss": 0.0236, + "step": 26260 + }, + { + "epoch": 0.6155991959539842, + "grad_norm": 2.8968255519866943, + "learning_rate": 6.450189794466377e-05, + "loss": 0.2204, + "step": 26261 + }, + { + "epoch": 0.6156226375287892, + "grad_norm": 1.9612812995910645, + "learning_rate": 6.449501161516797e-05, + "loss": 0.1183, + "step": 26262 + }, + { + "epoch": 0.6156460791035941, + "grad_norm": 2.3045809268951416, + "learning_rate": 6.448812547832322e-05, + "loss": 0.1884, + "step": 26263 + }, + { + "epoch": 0.6156695206783992, + "grad_norm": 0.4194652736186981, + "learning_rate": 6.448123953416691e-05, + "loss": 0.0469, + "step": 26264 + }, + { + "epoch": 0.6156929622532041, + "grad_norm": 3.128715991973877, + "learning_rate": 6.447435378273634e-05, + "loss": 0.2858, + "step": 26265 + }, + { + "epoch": 0.6157164038280092, + "grad_norm": 2.2784669399261475, + "learning_rate": 6.446746822406899e-05, + "loss": 0.1382, + "step": 26266 + }, + { + "epoch": 0.6157398454028141, + "grad_norm": 2.358368396759033, + "learning_rate": 6.446058285820212e-05, + "loss": 0.1796, + "step": 26267 + }, + { + "epoch": 0.6157632869776192, + "grad_norm": 4.262556552886963, + "learning_rate": 6.445369768517316e-05, + "loss": 0.2841, + "step": 26268 + }, + { + "epoch": 0.6157867285524241, + "grad_norm": 5.381286144256592, + "learning_rate": 6.444681270501942e-05, + "loss": 0.2784, + "step": 26269 + }, + { + "epoch": 0.6158101701272292, + "grad_norm": 1.2249727249145508, + "learning_rate": 6.443992791777827e-05, + "loss": 0.1335, + "step": 26270 + }, + { + "epoch": 0.6158336117020341, + "grad_norm": 3.084120750427246, + "learning_rate": 6.443304332348708e-05, + "loss": 0.3109, + "step": 26271 + }, + { + "epoch": 0.6158570532768392, + "grad_norm": 3.187037229537964, + "learning_rate": 6.44261589221832e-05, + "loss": 0.3709, + "step": 26272 + }, + { + "epoch": 0.6158804948516441, + "grad_norm": 1.840789794921875, + "learning_rate": 6.441927471390396e-05, + "loss": 0.2449, + "step": 26273 + }, + { + "epoch": 0.6159039364264491, + "grad_norm": 3.30757474899292, + "learning_rate": 6.441239069868676e-05, + "loss": 0.1539, + "step": 26274 + }, + { + "epoch": 0.6159273780012541, + "grad_norm": 4.239408016204834, + "learning_rate": 6.440550687656889e-05, + "loss": 0.3655, + "step": 26275 + }, + { + "epoch": 0.6159508195760591, + "grad_norm": 1.3353641033172607, + "learning_rate": 6.439862324758779e-05, + "loss": 0.1165, + "step": 26276 + }, + { + "epoch": 0.6159742611508641, + "grad_norm": 2.7689712047576904, + "learning_rate": 6.43917398117807e-05, + "loss": 1.059, + "step": 26277 + }, + { + "epoch": 0.6159977027256691, + "grad_norm": 3.4172658920288086, + "learning_rate": 6.438485656918509e-05, + "loss": 1.8725, + "step": 26278 + }, + { + "epoch": 0.616021144300474, + "grad_norm": 1.1133559942245483, + "learning_rate": 6.437797351983824e-05, + "loss": 0.065, + "step": 26279 + }, + { + "epoch": 0.6160445858752791, + "grad_norm": 2.20391583442688, + "learning_rate": 6.43710906637775e-05, + "loss": 0.1517, + "step": 26280 + }, + { + "epoch": 0.616068027450084, + "grad_norm": 2.3284900188446045, + "learning_rate": 6.436420800104022e-05, + "loss": 0.1835, + "step": 26281 + }, + { + "epoch": 0.6160914690248891, + "grad_norm": 2.8758842945098877, + "learning_rate": 6.435732553166374e-05, + "loss": 1.7286, + "step": 26282 + }, + { + "epoch": 0.616114910599694, + "grad_norm": 1.0784566402435303, + "learning_rate": 6.435044325568543e-05, + "loss": 0.074, + "step": 26283 + }, + { + "epoch": 0.6161383521744991, + "grad_norm": 3.383223056793213, + "learning_rate": 6.434356117314261e-05, + "loss": 1.0429, + "step": 26284 + }, + { + "epoch": 0.6161617937493041, + "grad_norm": 4.315315246582031, + "learning_rate": 6.433667928407264e-05, + "loss": 0.2828, + "step": 26285 + }, + { + "epoch": 0.6161852353241091, + "grad_norm": 0.7964097857475281, + "learning_rate": 6.432979758851285e-05, + "loss": 0.0634, + "step": 26286 + }, + { + "epoch": 0.6162086768989141, + "grad_norm": 2.500293254852295, + "learning_rate": 6.432291608650059e-05, + "loss": 1.7267, + "step": 26287 + }, + { + "epoch": 0.6162321184737191, + "grad_norm": 1.4534635543823242, + "learning_rate": 6.431603477807315e-05, + "loss": 0.1304, + "step": 26288 + }, + { + "epoch": 0.6162555600485241, + "grad_norm": 3.138720989227295, + "learning_rate": 6.430915366326791e-05, + "loss": 0.3837, + "step": 26289 + }, + { + "epoch": 0.616279001623329, + "grad_norm": 3.239887237548828, + "learning_rate": 6.430227274212223e-05, + "loss": 0.2332, + "step": 26290 + }, + { + "epoch": 0.6163024431981341, + "grad_norm": 3.2486836910247803, + "learning_rate": 6.429539201467343e-05, + "loss": 0.2321, + "step": 26291 + }, + { + "epoch": 0.616325884772939, + "grad_norm": 0.5614252090454102, + "learning_rate": 6.428851148095882e-05, + "loss": 0.0473, + "step": 26292 + }, + { + "epoch": 0.6163493263477441, + "grad_norm": 4.184835910797119, + "learning_rate": 6.428163114101577e-05, + "loss": 0.385, + "step": 26293 + }, + { + "epoch": 0.616372767922549, + "grad_norm": 5.292683124542236, + "learning_rate": 6.427475099488159e-05, + "loss": 0.6869, + "step": 26294 + }, + { + "epoch": 0.6163962094973541, + "grad_norm": 1.984068512916565, + "learning_rate": 6.42678710425936e-05, + "loss": 0.1319, + "step": 26295 + }, + { + "epoch": 0.616419651072159, + "grad_norm": 2.5976781845092773, + "learning_rate": 6.426099128418916e-05, + "loss": 0.2893, + "step": 26296 + }, + { + "epoch": 0.6164430926469641, + "grad_norm": 3.9595444202423096, + "learning_rate": 6.425411171970559e-05, + "loss": 0.33, + "step": 26297 + }, + { + "epoch": 0.616466534221769, + "grad_norm": 4.135586738586426, + "learning_rate": 6.424723234918019e-05, + "loss": 0.3133, + "step": 26298 + }, + { + "epoch": 0.616489975796574, + "grad_norm": 2.1545722484588623, + "learning_rate": 6.424035317265034e-05, + "loss": 0.1167, + "step": 26299 + }, + { + "epoch": 0.616513417371379, + "grad_norm": 1.4315060377120972, + "learning_rate": 6.423347419015334e-05, + "loss": 0.1732, + "step": 26300 + }, + { + "epoch": 0.616536858946184, + "grad_norm": 3.508305311203003, + "learning_rate": 6.422659540172646e-05, + "loss": 0.2679, + "step": 26301 + }, + { + "epoch": 0.616560300520989, + "grad_norm": 2.8709280490875244, + "learning_rate": 6.421971680740711e-05, + "loss": 0.2731, + "step": 26302 + }, + { + "epoch": 0.616583742095794, + "grad_norm": 1.0456392765045166, + "learning_rate": 6.42128384072326e-05, + "loss": 0.0731, + "step": 26303 + }, + { + "epoch": 0.616607183670599, + "grad_norm": 2.5063493251800537, + "learning_rate": 6.420596020124023e-05, + "loss": 0.1674, + "step": 26304 + }, + { + "epoch": 0.616630625245404, + "grad_norm": 3.5735697746276855, + "learning_rate": 6.419908218946732e-05, + "loss": 0.2104, + "step": 26305 + }, + { + "epoch": 0.616654066820209, + "grad_norm": 1.7561538219451904, + "learning_rate": 6.419220437195122e-05, + "loss": 0.1788, + "step": 26306 + }, + { + "epoch": 0.616677508395014, + "grad_norm": 2.0369021892547607, + "learning_rate": 6.418532674872921e-05, + "loss": 0.2912, + "step": 26307 + }, + { + "epoch": 0.616700949969819, + "grad_norm": 0.9544631838798523, + "learning_rate": 6.417844931983861e-05, + "loss": 0.083, + "step": 26308 + }, + { + "epoch": 0.616724391544624, + "grad_norm": 1.7962077856063843, + "learning_rate": 6.417157208531676e-05, + "loss": 0.1079, + "step": 26309 + }, + { + "epoch": 0.6167478331194289, + "grad_norm": 3.5710296630859375, + "learning_rate": 6.416469504520099e-05, + "loss": 0.3426, + "step": 26310 + }, + { + "epoch": 0.616771274694234, + "grad_norm": 4.158603668212891, + "learning_rate": 6.415781819952856e-05, + "loss": 0.374, + "step": 26311 + }, + { + "epoch": 0.6167947162690389, + "grad_norm": 4.317155838012695, + "learning_rate": 6.415094154833684e-05, + "loss": 0.4188, + "step": 26312 + }, + { + "epoch": 0.616818157843844, + "grad_norm": 3.6642227172851562, + "learning_rate": 6.414406509166309e-05, + "loss": 0.2195, + "step": 26313 + }, + { + "epoch": 0.6168415994186489, + "grad_norm": 0.2692887783050537, + "learning_rate": 6.413718882954466e-05, + "loss": 0.0229, + "step": 26314 + }, + { + "epoch": 0.616865040993454, + "grad_norm": 2.725921392440796, + "learning_rate": 6.413031276201885e-05, + "loss": 0.2429, + "step": 26315 + }, + { + "epoch": 0.6168884825682589, + "grad_norm": 3.0262248516082764, + "learning_rate": 6.412343688912299e-05, + "loss": 0.2669, + "step": 26316 + }, + { + "epoch": 0.616911924143064, + "grad_norm": 1.059496521949768, + "learning_rate": 6.411656121089437e-05, + "loss": 0.1149, + "step": 26317 + }, + { + "epoch": 0.6169353657178689, + "grad_norm": 1.2787331342697144, + "learning_rate": 6.410968572737028e-05, + "loss": 0.1192, + "step": 26318 + }, + { + "epoch": 0.6169588072926739, + "grad_norm": 3.100964307785034, + "learning_rate": 6.410281043858806e-05, + "loss": 0.2779, + "step": 26319 + }, + { + "epoch": 0.6169822488674789, + "grad_norm": 4.602634906768799, + "learning_rate": 6.409593534458499e-05, + "loss": 0.8078, + "step": 26320 + }, + { + "epoch": 0.6170056904422839, + "grad_norm": 3.259472131729126, + "learning_rate": 6.408906044539839e-05, + "loss": 0.2785, + "step": 26321 + }, + { + "epoch": 0.6170291320170889, + "grad_norm": 2.0054242610931396, + "learning_rate": 6.408218574106557e-05, + "loss": 0.2225, + "step": 26322 + }, + { + "epoch": 0.6170525735918939, + "grad_norm": 0.8463411927223206, + "learning_rate": 6.40753112316238e-05, + "loss": 0.0645, + "step": 26323 + }, + { + "epoch": 0.6170760151666989, + "grad_norm": 4.786448955535889, + "learning_rate": 6.406843691711043e-05, + "loss": 0.399, + "step": 26324 + }, + { + "epoch": 0.6170994567415039, + "grad_norm": 1.784018874168396, + "learning_rate": 6.40615627975627e-05, + "loss": 0.1739, + "step": 26325 + }, + { + "epoch": 0.6171228983163088, + "grad_norm": 3.7841036319732666, + "learning_rate": 6.405468887301792e-05, + "loss": 0.3619, + "step": 26326 + }, + { + "epoch": 0.6171463398911139, + "grad_norm": 2.8707447052001953, + "learning_rate": 6.404781514351343e-05, + "loss": 0.238, + "step": 26327 + }, + { + "epoch": 0.6171697814659188, + "grad_norm": 3.8605918884277344, + "learning_rate": 6.404094160908651e-05, + "loss": 0.2802, + "step": 26328 + }, + { + "epoch": 0.6171932230407239, + "grad_norm": 2.0710623264312744, + "learning_rate": 6.403406826977447e-05, + "loss": 0.1495, + "step": 26329 + }, + { + "epoch": 0.6172166646155288, + "grad_norm": 3.6881563663482666, + "learning_rate": 6.402719512561457e-05, + "loss": 0.2412, + "step": 26330 + }, + { + "epoch": 0.6172401061903339, + "grad_norm": 3.940707206726074, + "learning_rate": 6.402032217664413e-05, + "loss": 0.231, + "step": 26331 + }, + { + "epoch": 0.6172635477651388, + "grad_norm": 0.34896165132522583, + "learning_rate": 6.401344942290044e-05, + "loss": 0.0327, + "step": 26332 + }, + { + "epoch": 0.6172869893399439, + "grad_norm": 0.9724031686782837, + "learning_rate": 6.400657686442077e-05, + "loss": 0.1115, + "step": 26333 + }, + { + "epoch": 0.6173104309147488, + "grad_norm": 4.304942607879639, + "learning_rate": 6.399970450124243e-05, + "loss": 0.2339, + "step": 26334 + }, + { + "epoch": 0.6173338724895538, + "grad_norm": 2.489839553833008, + "learning_rate": 6.399283233340271e-05, + "loss": 0.2201, + "step": 26335 + }, + { + "epoch": 0.6173573140643589, + "grad_norm": 2.6099982261657715, + "learning_rate": 6.398596036093888e-05, + "loss": 0.2197, + "step": 26336 + }, + { + "epoch": 0.6173807556391638, + "grad_norm": 1.077046275138855, + "learning_rate": 6.397908858388826e-05, + "loss": 0.106, + "step": 26337 + }, + { + "epoch": 0.6174041972139689, + "grad_norm": 3.0294055938720703, + "learning_rate": 6.397221700228808e-05, + "loss": 0.1804, + "step": 26338 + }, + { + "epoch": 0.6174276387887738, + "grad_norm": 3.2322897911071777, + "learning_rate": 6.39653456161757e-05, + "loss": 0.1724, + "step": 26339 + }, + { + "epoch": 0.6174510803635789, + "grad_norm": 3.0351145267486572, + "learning_rate": 6.395847442558836e-05, + "loss": 1.5912, + "step": 26340 + }, + { + "epoch": 0.6174745219383838, + "grad_norm": 0.9738667607307434, + "learning_rate": 6.395160343056334e-05, + "loss": 0.0815, + "step": 26341 + }, + { + "epoch": 0.6174979635131889, + "grad_norm": 2.4235551357269287, + "learning_rate": 6.394473263113796e-05, + "loss": 0.1199, + "step": 26342 + }, + { + "epoch": 0.6175214050879938, + "grad_norm": 0.6150152087211609, + "learning_rate": 6.393786202734945e-05, + "loss": 0.063, + "step": 26343 + }, + { + "epoch": 0.6175448466627989, + "grad_norm": 3.102689504623413, + "learning_rate": 6.393099161923514e-05, + "loss": 0.1687, + "step": 26344 + }, + { + "epoch": 0.6175682882376038, + "grad_norm": 0.4466313123703003, + "learning_rate": 6.392412140683227e-05, + "loss": 0.027, + "step": 26345 + }, + { + "epoch": 0.6175917298124088, + "grad_norm": 2.1590945720672607, + "learning_rate": 6.391725139017813e-05, + "loss": 0.2085, + "step": 26346 + }, + { + "epoch": 0.6176151713872138, + "grad_norm": 2.141751289367676, + "learning_rate": 6.391038156930999e-05, + "loss": 0.2083, + "step": 26347 + }, + { + "epoch": 0.6176386129620188, + "grad_norm": 5.112919330596924, + "learning_rate": 6.390351194426517e-05, + "loss": 0.599, + "step": 26348 + }, + { + "epoch": 0.6176620545368238, + "grad_norm": 3.6969244480133057, + "learning_rate": 6.389664251508085e-05, + "loss": 0.2402, + "step": 26349 + }, + { + "epoch": 0.6176854961116288, + "grad_norm": 2.0932483673095703, + "learning_rate": 6.388977328179438e-05, + "loss": 1.0813, + "step": 26350 + }, + { + "epoch": 0.6177089376864338, + "grad_norm": 2.2464685440063477, + "learning_rate": 6.388290424444303e-05, + "loss": 1.2722, + "step": 26351 + }, + { + "epoch": 0.6177323792612388, + "grad_norm": 0.5424966812133789, + "learning_rate": 6.387603540306406e-05, + "loss": 0.0278, + "step": 26352 + }, + { + "epoch": 0.6177558208360437, + "grad_norm": 4.320620536804199, + "learning_rate": 6.386916675769472e-05, + "loss": 0.348, + "step": 26353 + }, + { + "epoch": 0.6177792624108488, + "grad_norm": 2.3211565017700195, + "learning_rate": 6.386229830837233e-05, + "loss": 0.1765, + "step": 26354 + }, + { + "epoch": 0.6178027039856537, + "grad_norm": 1.5007890462875366, + "learning_rate": 6.385543005513409e-05, + "loss": 0.185, + "step": 26355 + }, + { + "epoch": 0.6178261455604588, + "grad_norm": 4.428146839141846, + "learning_rate": 6.384856199801733e-05, + "loss": 0.4076, + "step": 26356 + }, + { + "epoch": 0.6178495871352637, + "grad_norm": 3.8588221073150635, + "learning_rate": 6.384169413705929e-05, + "loss": 0.478, + "step": 26357 + }, + { + "epoch": 0.6178730287100688, + "grad_norm": 3.531261920928955, + "learning_rate": 6.383482647229721e-05, + "loss": 0.313, + "step": 26358 + }, + { + "epoch": 0.6178964702848737, + "grad_norm": 1.587679386138916, + "learning_rate": 6.38279590037684e-05, + "loss": 0.1348, + "step": 26359 + }, + { + "epoch": 0.6179199118596788, + "grad_norm": 2.299691915512085, + "learning_rate": 6.38210917315101e-05, + "loss": 0.1769, + "step": 26360 + }, + { + "epoch": 0.6179433534344837, + "grad_norm": 3.918374538421631, + "learning_rate": 6.381422465555955e-05, + "loss": 0.415, + "step": 26361 + }, + { + "epoch": 0.6179667950092887, + "grad_norm": 0.8906726837158203, + "learning_rate": 6.380735777595404e-05, + "loss": 0.0612, + "step": 26362 + }, + { + "epoch": 0.6179902365840937, + "grad_norm": 3.4480695724487305, + "learning_rate": 6.380049109273083e-05, + "loss": 0.2442, + "step": 26363 + }, + { + "epoch": 0.6180136781588987, + "grad_norm": 0.7301601767539978, + "learning_rate": 6.379362460592719e-05, + "loss": 0.106, + "step": 26364 + }, + { + "epoch": 0.6180371197337037, + "grad_norm": 0.4690130650997162, + "learning_rate": 6.378675831558035e-05, + "loss": 0.0299, + "step": 26365 + }, + { + "epoch": 0.6180605613085087, + "grad_norm": 3.1581106185913086, + "learning_rate": 6.377989222172757e-05, + "loss": 0.2788, + "step": 26366 + }, + { + "epoch": 0.6180840028833137, + "grad_norm": 2.3844311237335205, + "learning_rate": 6.377302632440614e-05, + "loss": 1.6812, + "step": 26367 + }, + { + "epoch": 0.6181074444581187, + "grad_norm": 4.569007873535156, + "learning_rate": 6.376616062365325e-05, + "loss": 0.6237, + "step": 26368 + }, + { + "epoch": 0.6181308860329237, + "grad_norm": 2.3844048976898193, + "learning_rate": 6.375929511950622e-05, + "loss": 0.2412, + "step": 26369 + }, + { + "epoch": 0.6181543276077287, + "grad_norm": 2.5950613021850586, + "learning_rate": 6.375242981200227e-05, + "loss": 0.2009, + "step": 26370 + }, + { + "epoch": 0.6181777691825336, + "grad_norm": 0.22004462778568268, + "learning_rate": 6.374556470117863e-05, + "loss": 0.0253, + "step": 26371 + }, + { + "epoch": 0.6182012107573387, + "grad_norm": 5.329720497131348, + "learning_rate": 6.373869978707261e-05, + "loss": 0.2667, + "step": 26372 + }, + { + "epoch": 0.6182246523321436, + "grad_norm": 3.023655652999878, + "learning_rate": 6.373183506972142e-05, + "loss": 0.2316, + "step": 26373 + }, + { + "epoch": 0.6182480939069487, + "grad_norm": 2.673459529876709, + "learning_rate": 6.372497054916227e-05, + "loss": 0.2178, + "step": 26374 + }, + { + "epoch": 0.6182715354817536, + "grad_norm": 2.7953712940216064, + "learning_rate": 6.371810622543246e-05, + "loss": 0.24, + "step": 26375 + }, + { + "epoch": 0.6182949770565587, + "grad_norm": 0.5818561315536499, + "learning_rate": 6.371124209856927e-05, + "loss": 0.0656, + "step": 26376 + }, + { + "epoch": 0.6183184186313636, + "grad_norm": 1.5234090089797974, + "learning_rate": 6.370437816860988e-05, + "loss": 0.1713, + "step": 26377 + }, + { + "epoch": 0.6183418602061687, + "grad_norm": 4.1686625480651855, + "learning_rate": 6.369751443559153e-05, + "loss": 0.3446, + "step": 26378 + }, + { + "epoch": 0.6183653017809736, + "grad_norm": 1.2006946802139282, + "learning_rate": 6.369065089955152e-05, + "loss": 0.0707, + "step": 26379 + }, + { + "epoch": 0.6183887433557786, + "grad_norm": 3.450883150100708, + "learning_rate": 6.368378756052707e-05, + "loss": 0.2635, + "step": 26380 + }, + { + "epoch": 0.6184121849305836, + "grad_norm": 3.10430645942688, + "learning_rate": 6.367692441855537e-05, + "loss": 0.2606, + "step": 26381 + }, + { + "epoch": 0.6184356265053886, + "grad_norm": 1.4478497505187988, + "learning_rate": 6.367006147367373e-05, + "loss": 0.0996, + "step": 26382 + }, + { + "epoch": 0.6184590680801936, + "grad_norm": 3.5342559814453125, + "learning_rate": 6.366319872591935e-05, + "loss": 0.3155, + "step": 26383 + }, + { + "epoch": 0.6184825096549986, + "grad_norm": 4.006580829620361, + "learning_rate": 6.365633617532945e-05, + "loss": 0.5136, + "step": 26384 + }, + { + "epoch": 0.6185059512298036, + "grad_norm": 1.9613643884658813, + "learning_rate": 6.364947382194133e-05, + "loss": 1.2919, + "step": 26385 + }, + { + "epoch": 0.6185293928046086, + "grad_norm": 2.184486150741577, + "learning_rate": 6.364261166579214e-05, + "loss": 0.1241, + "step": 26386 + }, + { + "epoch": 0.6185528343794137, + "grad_norm": 2.845078229904175, + "learning_rate": 6.36357497069192e-05, + "loss": 1.1635, + "step": 26387 + }, + { + "epoch": 0.6185762759542186, + "grad_norm": 2.7082831859588623, + "learning_rate": 6.362888794535968e-05, + "loss": 1.412, + "step": 26388 + }, + { + "epoch": 0.6185997175290237, + "grad_norm": 4.371167182922363, + "learning_rate": 6.362202638115087e-05, + "loss": 0.5403, + "step": 26389 + }, + { + "epoch": 0.6186231591038286, + "grad_norm": 2.1915907859802246, + "learning_rate": 6.361516501432995e-05, + "loss": 1.1867, + "step": 26390 + }, + { + "epoch": 0.6186466006786336, + "grad_norm": 2.8075695037841797, + "learning_rate": 6.360830384493416e-05, + "loss": 1.7821, + "step": 26391 + }, + { + "epoch": 0.6186700422534386, + "grad_norm": 4.631058216094971, + "learning_rate": 6.360144287300075e-05, + "loss": 0.2186, + "step": 26392 + }, + { + "epoch": 0.6186934838282436, + "grad_norm": 9.426875114440918, + "learning_rate": 6.359458209856693e-05, + "loss": 0.4512, + "step": 26393 + }, + { + "epoch": 0.6187169254030486, + "grad_norm": 0.621260941028595, + "learning_rate": 6.358772152166992e-05, + "loss": 0.0504, + "step": 26394 + }, + { + "epoch": 0.6187403669778536, + "grad_norm": 2.8416199684143066, + "learning_rate": 6.358086114234698e-05, + "loss": 0.2998, + "step": 26395 + }, + { + "epoch": 0.6187638085526586, + "grad_norm": 5.7978668212890625, + "learning_rate": 6.357400096063528e-05, + "loss": 0.4436, + "step": 26396 + }, + { + "epoch": 0.6187872501274636, + "grad_norm": 1.2165465354919434, + "learning_rate": 6.356714097657211e-05, + "loss": 0.1488, + "step": 26397 + }, + { + "epoch": 0.6188106917022685, + "grad_norm": 3.709602117538452, + "learning_rate": 6.356028119019466e-05, + "loss": 0.3198, + "step": 26398 + }, + { + "epoch": 0.6188341332770736, + "grad_norm": 2.9472649097442627, + "learning_rate": 6.355342160154009e-05, + "loss": 0.2379, + "step": 26399 + }, + { + "epoch": 0.6188575748518785, + "grad_norm": 2.1422061920166016, + "learning_rate": 6.35465622106457e-05, + "loss": 0.2564, + "step": 26400 + }, + { + "epoch": 0.6188810164266836, + "grad_norm": 2.415541172027588, + "learning_rate": 6.353970301754872e-05, + "loss": 0.1259, + "step": 26401 + }, + { + "epoch": 0.6189044580014885, + "grad_norm": 1.3659946918487549, + "learning_rate": 6.353284402228635e-05, + "loss": 0.1366, + "step": 26402 + }, + { + "epoch": 0.6189278995762936, + "grad_norm": 1.9407557249069214, + "learning_rate": 6.352598522489575e-05, + "loss": 1.5448, + "step": 26403 + }, + { + "epoch": 0.6189513411510985, + "grad_norm": 2.820067882537842, + "learning_rate": 6.351912662541419e-05, + "loss": 0.211, + "step": 26404 + }, + { + "epoch": 0.6189747827259036, + "grad_norm": 3.2581675052642822, + "learning_rate": 6.35122682238789e-05, + "loss": 0.3577, + "step": 26405 + }, + { + "epoch": 0.6189982243007085, + "grad_norm": 2.801607608795166, + "learning_rate": 6.350541002032704e-05, + "loss": 0.1856, + "step": 26406 + }, + { + "epoch": 0.6190216658755135, + "grad_norm": 1.6627384424209595, + "learning_rate": 6.349855201479586e-05, + "loss": 0.1161, + "step": 26407 + }, + { + "epoch": 0.6190451074503185, + "grad_norm": 2.6007871627807617, + "learning_rate": 6.349169420732259e-05, + "loss": 0.264, + "step": 26408 + }, + { + "epoch": 0.6190685490251235, + "grad_norm": 0.2977376878261566, + "learning_rate": 6.348483659794438e-05, + "loss": 0.0132, + "step": 26409 + }, + { + "epoch": 0.6190919905999285, + "grad_norm": 0.6481781005859375, + "learning_rate": 6.347797918669848e-05, + "loss": 0.047, + "step": 26410 + }, + { + "epoch": 0.6191154321747335, + "grad_norm": 2.7958080768585205, + "learning_rate": 6.347112197362206e-05, + "loss": 0.183, + "step": 26411 + }, + { + "epoch": 0.6191388737495385, + "grad_norm": 2.8170924186706543, + "learning_rate": 6.34642649587524e-05, + "loss": 0.2091, + "step": 26412 + }, + { + "epoch": 0.6191623153243435, + "grad_norm": 1.9206310510635376, + "learning_rate": 6.345740814212665e-05, + "loss": 0.1703, + "step": 26413 + }, + { + "epoch": 0.6191857568991485, + "grad_norm": 4.263484954833984, + "learning_rate": 6.345055152378206e-05, + "loss": 0.3739, + "step": 26414 + }, + { + "epoch": 0.6192091984739535, + "grad_norm": 1.108530044555664, + "learning_rate": 6.344369510375579e-05, + "loss": 0.0611, + "step": 26415 + }, + { + "epoch": 0.6192326400487584, + "grad_norm": 1.2876331806182861, + "learning_rate": 6.343683888208502e-05, + "loss": 0.7824, + "step": 26416 + }, + { + "epoch": 0.6192560816235635, + "grad_norm": 2.8700077533721924, + "learning_rate": 6.342998285880704e-05, + "loss": 0.2377, + "step": 26417 + }, + { + "epoch": 0.6192795231983684, + "grad_norm": 4.032572269439697, + "learning_rate": 6.3423127033959e-05, + "loss": 0.3632, + "step": 26418 + }, + { + "epoch": 0.6193029647731735, + "grad_norm": 1.7151681184768677, + "learning_rate": 6.341627140757807e-05, + "loss": 0.1664, + "step": 26419 + }, + { + "epoch": 0.6193264063479784, + "grad_norm": 2.021852731704712, + "learning_rate": 6.340941597970152e-05, + "loss": 0.1716, + "step": 26420 + }, + { + "epoch": 0.6193498479227835, + "grad_norm": 5.6404314041137695, + "learning_rate": 6.340256075036648e-05, + "loss": 0.6556, + "step": 26421 + }, + { + "epoch": 0.6193732894975884, + "grad_norm": 0.647169828414917, + "learning_rate": 6.339570571961016e-05, + "loss": 0.0503, + "step": 26422 + }, + { + "epoch": 0.6193967310723935, + "grad_norm": 3.6017563343048096, + "learning_rate": 6.338885088746976e-05, + "loss": 0.3306, + "step": 26423 + }, + { + "epoch": 0.6194201726471984, + "grad_norm": 3.331644296646118, + "learning_rate": 6.338199625398252e-05, + "loss": 0.5763, + "step": 26424 + }, + { + "epoch": 0.6194436142220034, + "grad_norm": 2.091500997543335, + "learning_rate": 6.33751418191856e-05, + "loss": 0.2126, + "step": 26425 + }, + { + "epoch": 0.6194670557968084, + "grad_norm": 1.7257269620895386, + "learning_rate": 6.336828758311617e-05, + "loss": 0.9233, + "step": 26426 + }, + { + "epoch": 0.6194904973716134, + "grad_norm": 1.0597513914108276, + "learning_rate": 6.336143354581145e-05, + "loss": 0.0636, + "step": 26427 + }, + { + "epoch": 0.6195139389464184, + "grad_norm": 1.066770315170288, + "learning_rate": 6.335457970730865e-05, + "loss": 0.1396, + "step": 26428 + }, + { + "epoch": 0.6195373805212234, + "grad_norm": 2.7976491451263428, + "learning_rate": 6.334772606764489e-05, + "loss": 0.2138, + "step": 26429 + }, + { + "epoch": 0.6195608220960284, + "grad_norm": 3.7833914756774902, + "learning_rate": 6.334087262685741e-05, + "loss": 0.231, + "step": 26430 + }, + { + "epoch": 0.6195842636708334, + "grad_norm": 1.6709798574447632, + "learning_rate": 6.333401938498339e-05, + "loss": 0.1444, + "step": 26431 + }, + { + "epoch": 0.6196077052456384, + "grad_norm": 3.2254438400268555, + "learning_rate": 6.332716634206001e-05, + "loss": 0.3325, + "step": 26432 + }, + { + "epoch": 0.6196311468204434, + "grad_norm": 3.08972430229187, + "learning_rate": 6.332031349812448e-05, + "loss": 0.1439, + "step": 26433 + }, + { + "epoch": 0.6196545883952483, + "grad_norm": 1.510011911392212, + "learning_rate": 6.331346085321393e-05, + "loss": 0.0864, + "step": 26434 + }, + { + "epoch": 0.6196780299700534, + "grad_norm": 2.8778419494628906, + "learning_rate": 6.330660840736556e-05, + "loss": 0.1823, + "step": 26435 + }, + { + "epoch": 0.6197014715448583, + "grad_norm": 1.1835707426071167, + "learning_rate": 6.329975616061658e-05, + "loss": 0.0738, + "step": 26436 + }, + { + "epoch": 0.6197249131196634, + "grad_norm": 1.9850800037384033, + "learning_rate": 6.329290411300417e-05, + "loss": 0.1537, + "step": 26437 + }, + { + "epoch": 0.6197483546944684, + "grad_norm": 2.464754343032837, + "learning_rate": 6.328605226456548e-05, + "loss": 1.5475, + "step": 26438 + }, + { + "epoch": 0.6197717962692734, + "grad_norm": 3.4422953128814697, + "learning_rate": 6.327920061533771e-05, + "loss": 0.3152, + "step": 26439 + }, + { + "epoch": 0.6197952378440784, + "grad_norm": 3.096345901489258, + "learning_rate": 6.327234916535803e-05, + "loss": 0.2107, + "step": 26440 + }, + { + "epoch": 0.6198186794188834, + "grad_norm": 2.399765729904175, + "learning_rate": 6.326549791466361e-05, + "loss": 1.8789, + "step": 26441 + }, + { + "epoch": 0.6198421209936884, + "grad_norm": 4.680772304534912, + "learning_rate": 6.325864686329163e-05, + "loss": 0.3957, + "step": 26442 + }, + { + "epoch": 0.6198655625684933, + "grad_norm": 3.074538230895996, + "learning_rate": 6.325179601127928e-05, + "loss": 0.1237, + "step": 26443 + }, + { + "epoch": 0.6198890041432984, + "grad_norm": 4.201650619506836, + "learning_rate": 6.32449453586637e-05, + "loss": 0.2468, + "step": 26444 + }, + { + "epoch": 0.6199124457181033, + "grad_norm": 3.358354091644287, + "learning_rate": 6.323809490548211e-05, + "loss": 0.3886, + "step": 26445 + }, + { + "epoch": 0.6199358872929084, + "grad_norm": 1.97011137008667, + "learning_rate": 6.323124465177164e-05, + "loss": 0.1766, + "step": 26446 + }, + { + "epoch": 0.6199593288677133, + "grad_norm": 2.7342352867126465, + "learning_rate": 6.322439459756941e-05, + "loss": 1.3548, + "step": 26447 + }, + { + "epoch": 0.6199827704425184, + "grad_norm": 2.676708459854126, + "learning_rate": 6.32175447429127e-05, + "loss": 0.1994, + "step": 26448 + }, + { + "epoch": 0.6200062120173233, + "grad_norm": 2.6735241413116455, + "learning_rate": 6.321069508783863e-05, + "loss": 0.1097, + "step": 26449 + }, + { + "epoch": 0.6200296535921284, + "grad_norm": 3.2474963665008545, + "learning_rate": 6.320384563238436e-05, + "loss": 0.1609, + "step": 26450 + }, + { + "epoch": 0.6200530951669333, + "grad_norm": 3.3656909465789795, + "learning_rate": 6.319699637658705e-05, + "loss": 0.1431, + "step": 26451 + }, + { + "epoch": 0.6200765367417383, + "grad_norm": 7.0774712562561035, + "learning_rate": 6.319014732048388e-05, + "loss": 0.6076, + "step": 26452 + }, + { + "epoch": 0.6200999783165433, + "grad_norm": 2.3110694885253906, + "learning_rate": 6.318329846411202e-05, + "loss": 1.5589, + "step": 26453 + }, + { + "epoch": 0.6201234198913483, + "grad_norm": 0.7932669520378113, + "learning_rate": 6.317644980750859e-05, + "loss": 0.0905, + "step": 26454 + }, + { + "epoch": 0.6201468614661533, + "grad_norm": 3.8928844928741455, + "learning_rate": 6.31696013507108e-05, + "loss": 0.4118, + "step": 26455 + }, + { + "epoch": 0.6201703030409583, + "grad_norm": 2.424285650253296, + "learning_rate": 6.316275309375579e-05, + "loss": 0.1859, + "step": 26456 + }, + { + "epoch": 0.6201937446157633, + "grad_norm": 3.346229076385498, + "learning_rate": 6.315590503668069e-05, + "loss": 0.2989, + "step": 26457 + }, + { + "epoch": 0.6202171861905683, + "grad_norm": 1.5810045003890991, + "learning_rate": 6.314905717952271e-05, + "loss": 0.1767, + "step": 26458 + }, + { + "epoch": 0.6202406277653733, + "grad_norm": 2.380808115005493, + "learning_rate": 6.314220952231895e-05, + "loss": 0.2008, + "step": 26459 + }, + { + "epoch": 0.6202640693401783, + "grad_norm": 2.412818670272827, + "learning_rate": 6.313536206510661e-05, + "loss": 0.2453, + "step": 26460 + }, + { + "epoch": 0.6202875109149832, + "grad_norm": 3.594085693359375, + "learning_rate": 6.312851480792285e-05, + "loss": 0.3226, + "step": 26461 + }, + { + "epoch": 0.6203109524897883, + "grad_norm": 2.2227773666381836, + "learning_rate": 6.31216677508048e-05, + "loss": 0.1951, + "step": 26462 + }, + { + "epoch": 0.6203343940645932, + "grad_norm": 2.891169548034668, + "learning_rate": 6.311482089378962e-05, + "loss": 0.349, + "step": 26463 + }, + { + "epoch": 0.6203578356393983, + "grad_norm": 2.547867774963379, + "learning_rate": 6.310797423691445e-05, + "loss": 1.8261, + "step": 26464 + }, + { + "epoch": 0.6203812772142032, + "grad_norm": 1.1334975957870483, + "learning_rate": 6.310112778021647e-05, + "loss": 0.0698, + "step": 26465 + }, + { + "epoch": 0.6204047187890083, + "grad_norm": 2.090346574783325, + "learning_rate": 6.30942815237328e-05, + "loss": 0.2101, + "step": 26466 + }, + { + "epoch": 0.6204281603638132, + "grad_norm": 0.4627007246017456, + "learning_rate": 6.308743546750057e-05, + "loss": 0.0677, + "step": 26467 + }, + { + "epoch": 0.6204516019386183, + "grad_norm": 3.9007506370544434, + "learning_rate": 6.308058961155698e-05, + "loss": 1.4612, + "step": 26468 + }, + { + "epoch": 0.6204750435134232, + "grad_norm": 2.319134473800659, + "learning_rate": 6.307374395593914e-05, + "loss": 0.1242, + "step": 26469 + }, + { + "epoch": 0.6204984850882282, + "grad_norm": 0.5612279176712036, + "learning_rate": 6.306689850068421e-05, + "loss": 0.0556, + "step": 26470 + }, + { + "epoch": 0.6205219266630332, + "grad_norm": 0.8592501282691956, + "learning_rate": 6.306005324582933e-05, + "loss": 0.0878, + "step": 26471 + }, + { + "epoch": 0.6205453682378382, + "grad_norm": 2.0235183238983154, + "learning_rate": 6.30532081914116e-05, + "loss": 0.1523, + "step": 26472 + }, + { + "epoch": 0.6205688098126432, + "grad_norm": 2.403825521469116, + "learning_rate": 6.304636333746823e-05, + "loss": 0.1335, + "step": 26473 + }, + { + "epoch": 0.6205922513874482, + "grad_norm": 4.153733253479004, + "learning_rate": 6.303951868403633e-05, + "loss": 0.2029, + "step": 26474 + }, + { + "epoch": 0.6206156929622532, + "grad_norm": 3.0748090744018555, + "learning_rate": 6.303267423115305e-05, + "loss": 0.168, + "step": 26475 + }, + { + "epoch": 0.6206391345370582, + "grad_norm": 5.656301975250244, + "learning_rate": 6.302582997885551e-05, + "loss": 0.4948, + "step": 26476 + }, + { + "epoch": 0.6206625761118632, + "grad_norm": 3.799743413925171, + "learning_rate": 6.301898592718086e-05, + "loss": 0.3368, + "step": 26477 + }, + { + "epoch": 0.6206860176866682, + "grad_norm": 4.76701021194458, + "learning_rate": 6.301214207616623e-05, + "loss": 0.4884, + "step": 26478 + }, + { + "epoch": 0.6207094592614731, + "grad_norm": 3.194540023803711, + "learning_rate": 6.300529842584877e-05, + "loss": 0.1981, + "step": 26479 + }, + { + "epoch": 0.6207329008362782, + "grad_norm": 1.202046275138855, + "learning_rate": 6.29984549762656e-05, + "loss": 0.0365, + "step": 26480 + }, + { + "epoch": 0.6207563424110831, + "grad_norm": 3.650952100753784, + "learning_rate": 6.299161172745384e-05, + "loss": 0.4008, + "step": 26481 + }, + { + "epoch": 0.6207797839858882, + "grad_norm": 2.038593292236328, + "learning_rate": 6.298476867945062e-05, + "loss": 0.1713, + "step": 26482 + }, + { + "epoch": 0.6208032255606931, + "grad_norm": 1.978258490562439, + "learning_rate": 6.297792583229312e-05, + "loss": 1.3654, + "step": 26483 + }, + { + "epoch": 0.6208266671354982, + "grad_norm": 4.62600040435791, + "learning_rate": 6.297108318601838e-05, + "loss": 0.4576, + "step": 26484 + }, + { + "epoch": 0.6208501087103031, + "grad_norm": 3.243924617767334, + "learning_rate": 6.296424074066365e-05, + "loss": 1.422, + "step": 26485 + }, + { + "epoch": 0.6208735502851082, + "grad_norm": 4.065007209777832, + "learning_rate": 6.295739849626595e-05, + "loss": 0.3988, + "step": 26486 + }, + { + "epoch": 0.6208969918599131, + "grad_norm": 3.383470058441162, + "learning_rate": 6.295055645286248e-05, + "loss": 0.377, + "step": 26487 + }, + { + "epoch": 0.6209204334347181, + "grad_norm": 4.271601676940918, + "learning_rate": 6.294371461049031e-05, + "loss": 0.3322, + "step": 26488 + }, + { + "epoch": 0.6209438750095232, + "grad_norm": 3.8964481353759766, + "learning_rate": 6.293687296918657e-05, + "loss": 0.2779, + "step": 26489 + }, + { + "epoch": 0.6209673165843281, + "grad_norm": 3.3845267295837402, + "learning_rate": 6.293003152898844e-05, + "loss": 0.4619, + "step": 26490 + }, + { + "epoch": 0.6209907581591332, + "grad_norm": 1.508589744567871, + "learning_rate": 6.2923190289933e-05, + "loss": 0.0855, + "step": 26491 + }, + { + "epoch": 0.6210141997339381, + "grad_norm": 2.742755174636841, + "learning_rate": 6.291634925205733e-05, + "loss": 0.2307, + "step": 26492 + }, + { + "epoch": 0.6210376413087432, + "grad_norm": 1.40371572971344, + "learning_rate": 6.290950841539864e-05, + "loss": 0.1534, + "step": 26493 + }, + { + "epoch": 0.6210610828835481, + "grad_norm": 2.037992000579834, + "learning_rate": 6.290266777999398e-05, + "loss": 0.1302, + "step": 26494 + }, + { + "epoch": 0.6210845244583532, + "grad_norm": 2.4052865505218506, + "learning_rate": 6.289582734588046e-05, + "loss": 0.1382, + "step": 26495 + }, + { + "epoch": 0.6211079660331581, + "grad_norm": 1.1481938362121582, + "learning_rate": 6.288898711309523e-05, + "loss": 0.1429, + "step": 26496 + }, + { + "epoch": 0.6211314076079631, + "grad_norm": 3.3445847034454346, + "learning_rate": 6.288214708167542e-05, + "loss": 0.0773, + "step": 26497 + }, + { + "epoch": 0.6211548491827681, + "grad_norm": 1.9698760509490967, + "learning_rate": 6.287530725165813e-05, + "loss": 0.1972, + "step": 26498 + }, + { + "epoch": 0.6211782907575731, + "grad_norm": 2.7692296504974365, + "learning_rate": 6.286846762308046e-05, + "loss": 0.1783, + "step": 26499 + }, + { + "epoch": 0.6212017323323781, + "grad_norm": 4.941635608673096, + "learning_rate": 6.286162819597952e-05, + "loss": 0.7086, + "step": 26500 + }, + { + "epoch": 0.6212251739071831, + "grad_norm": 1.9882947206497192, + "learning_rate": 6.285478897039245e-05, + "loss": 0.1159, + "step": 26501 + }, + { + "epoch": 0.6212486154819881, + "grad_norm": 3.025912046432495, + "learning_rate": 6.284794994635631e-05, + "loss": 0.198, + "step": 26502 + }, + { + "epoch": 0.6212720570567931, + "grad_norm": 0.37410280108451843, + "learning_rate": 6.284111112390827e-05, + "loss": 0.0136, + "step": 26503 + }, + { + "epoch": 0.621295498631598, + "grad_norm": 1.2912236452102661, + "learning_rate": 6.28342725030854e-05, + "loss": 0.1551, + "step": 26504 + }, + { + "epoch": 0.6213189402064031, + "grad_norm": 2.328202724456787, + "learning_rate": 6.282743408392478e-05, + "loss": 1.957, + "step": 26505 + }, + { + "epoch": 0.621342381781208, + "grad_norm": 1.4411836862564087, + "learning_rate": 6.282059586646358e-05, + "loss": 0.0822, + "step": 26506 + }, + { + "epoch": 0.6213658233560131, + "grad_norm": 0.6724788546562195, + "learning_rate": 6.281375785073885e-05, + "loss": 0.048, + "step": 26507 + }, + { + "epoch": 0.621389264930818, + "grad_norm": 2.752126932144165, + "learning_rate": 6.280692003678772e-05, + "loss": 0.2411, + "step": 26508 + }, + { + "epoch": 0.6214127065056231, + "grad_norm": 3.0276834964752197, + "learning_rate": 6.280008242464728e-05, + "loss": 0.1924, + "step": 26509 + }, + { + "epoch": 0.621436148080428, + "grad_norm": 1.9927650690078735, + "learning_rate": 6.279324501435466e-05, + "loss": 1.1424, + "step": 26510 + }, + { + "epoch": 0.6214595896552331, + "grad_norm": 1.3171327114105225, + "learning_rate": 6.278640780594693e-05, + "loss": 0.1891, + "step": 26511 + }, + { + "epoch": 0.621483031230038, + "grad_norm": 3.329437255859375, + "learning_rate": 6.277957079946121e-05, + "loss": 1.2176, + "step": 26512 + }, + { + "epoch": 0.6215064728048431, + "grad_norm": 2.836106538772583, + "learning_rate": 6.277273399493458e-05, + "loss": 0.2029, + "step": 26513 + }, + { + "epoch": 0.621529914379648, + "grad_norm": 1.4136221408843994, + "learning_rate": 6.276589739240413e-05, + "loss": 0.1695, + "step": 26514 + }, + { + "epoch": 0.621553355954453, + "grad_norm": 0.9067323803901672, + "learning_rate": 6.275906099190699e-05, + "loss": 0.0617, + "step": 26515 + }, + { + "epoch": 0.621576797529258, + "grad_norm": 3.576392650604248, + "learning_rate": 6.275222479348022e-05, + "loss": 0.2927, + "step": 26516 + }, + { + "epoch": 0.621600239104063, + "grad_norm": 2.49466609954834, + "learning_rate": 6.274538879716091e-05, + "loss": 0.256, + "step": 26517 + }, + { + "epoch": 0.621623680678868, + "grad_norm": 0.9693692326545715, + "learning_rate": 6.273855300298618e-05, + "loss": 0.095, + "step": 26518 + }, + { + "epoch": 0.621647122253673, + "grad_norm": 2.6355388164520264, + "learning_rate": 6.273171741099313e-05, + "loss": 0.1637, + "step": 26519 + }, + { + "epoch": 0.621670563828478, + "grad_norm": 4.464707851409912, + "learning_rate": 6.272488202121878e-05, + "loss": 0.47, + "step": 26520 + }, + { + "epoch": 0.621694005403283, + "grad_norm": 0.7106619477272034, + "learning_rate": 6.271804683370028e-05, + "loss": 0.0553, + "step": 26521 + }, + { + "epoch": 0.621717446978088, + "grad_norm": 1.2972451448440552, + "learning_rate": 6.271121184847473e-05, + "loss": 0.185, + "step": 26522 + }, + { + "epoch": 0.621740888552893, + "grad_norm": 0.8944860696792603, + "learning_rate": 6.27043770655792e-05, + "loss": 0.1102, + "step": 26523 + }, + { + "epoch": 0.6217643301276979, + "grad_norm": 1.6902257204055786, + "learning_rate": 6.269754248505074e-05, + "loss": 0.1747, + "step": 26524 + }, + { + "epoch": 0.621787771702503, + "grad_norm": 0.9585304260253906, + "learning_rate": 6.269070810692649e-05, + "loss": 0.0919, + "step": 26525 + }, + { + "epoch": 0.6218112132773079, + "grad_norm": 3.447648048400879, + "learning_rate": 6.26838739312435e-05, + "loss": 0.3185, + "step": 26526 + }, + { + "epoch": 0.621834654852113, + "grad_norm": 0.8673545718193054, + "learning_rate": 6.267703995803884e-05, + "loss": 0.0801, + "step": 26527 + }, + { + "epoch": 0.6218580964269179, + "grad_norm": 7.530107021331787, + "learning_rate": 6.267020618734965e-05, + "loss": 0.2363, + "step": 26528 + }, + { + "epoch": 0.621881538001723, + "grad_norm": 2.7120001316070557, + "learning_rate": 6.266337261921294e-05, + "loss": 0.1626, + "step": 26529 + }, + { + "epoch": 0.6219049795765279, + "grad_norm": 3.1605730056762695, + "learning_rate": 6.265653925366581e-05, + "loss": 0.1714, + "step": 26530 + }, + { + "epoch": 0.621928421151333, + "grad_norm": 2.955472707748413, + "learning_rate": 6.264970609074539e-05, + "loss": 0.1746, + "step": 26531 + }, + { + "epoch": 0.6219518627261379, + "grad_norm": 6.579541206359863, + "learning_rate": 6.264287313048865e-05, + "loss": 0.24, + "step": 26532 + }, + { + "epoch": 0.6219753043009429, + "grad_norm": 0.8341567516326904, + "learning_rate": 6.263604037293278e-05, + "loss": 0.0858, + "step": 26533 + }, + { + "epoch": 0.6219987458757479, + "grad_norm": 3.2763845920562744, + "learning_rate": 6.26292078181148e-05, + "loss": 0.1978, + "step": 26534 + }, + { + "epoch": 0.6220221874505529, + "grad_norm": 0.7135927081108093, + "learning_rate": 6.262237546607178e-05, + "loss": 0.039, + "step": 26535 + }, + { + "epoch": 0.6220456290253579, + "grad_norm": 1.576828956604004, + "learning_rate": 6.261554331684081e-05, + "loss": 0.085, + "step": 26536 + }, + { + "epoch": 0.6220690706001629, + "grad_norm": 2.165850877761841, + "learning_rate": 6.260871137045894e-05, + "loss": 0.1251, + "step": 26537 + }, + { + "epoch": 0.6220925121749679, + "grad_norm": 0.9359259009361267, + "learning_rate": 6.260187962696326e-05, + "loss": 0.0974, + "step": 26538 + }, + { + "epoch": 0.6221159537497729, + "grad_norm": 2.0545973777770996, + "learning_rate": 6.259504808639085e-05, + "loss": 0.2354, + "step": 26539 + }, + { + "epoch": 0.622139395324578, + "grad_norm": 0.5192754864692688, + "learning_rate": 6.258821674877875e-05, + "loss": 0.0198, + "step": 26540 + }, + { + "epoch": 0.6221628368993829, + "grad_norm": 1.9866787195205688, + "learning_rate": 6.258138561416403e-05, + "loss": 0.1855, + "step": 26541 + }, + { + "epoch": 0.622186278474188, + "grad_norm": 2.681189775466919, + "learning_rate": 6.257455468258379e-05, + "loss": 0.1617, + "step": 26542 + }, + { + "epoch": 0.6222097200489929, + "grad_norm": 2.8075461387634277, + "learning_rate": 6.256772395407503e-05, + "loss": 1.3307, + "step": 26543 + }, + { + "epoch": 0.6222331616237979, + "grad_norm": 2.1955409049987793, + "learning_rate": 6.256089342867488e-05, + "loss": 1.5982, + "step": 26544 + }, + { + "epoch": 0.6222566031986029, + "grad_norm": 3.5864264965057373, + "learning_rate": 6.255406310642033e-05, + "loss": 0.3192, + "step": 26545 + }, + { + "epoch": 0.6222800447734079, + "grad_norm": 3.4000306129455566, + "learning_rate": 6.254723298734853e-05, + "loss": 0.2133, + "step": 26546 + }, + { + "epoch": 0.6223034863482129, + "grad_norm": 3.276339292526245, + "learning_rate": 6.254040307149646e-05, + "loss": 0.3663, + "step": 26547 + }, + { + "epoch": 0.6223269279230179, + "grad_norm": 1.2283180952072144, + "learning_rate": 6.253357335890127e-05, + "loss": 0.0723, + "step": 26548 + }, + { + "epoch": 0.6223503694978229, + "grad_norm": 0.9523839950561523, + "learning_rate": 6.252674384959992e-05, + "loss": 0.0603, + "step": 26549 + }, + { + "epoch": 0.6223738110726279, + "grad_norm": 1.9137094020843506, + "learning_rate": 6.251991454362955e-05, + "loss": 1.3729, + "step": 26550 + }, + { + "epoch": 0.6223972526474328, + "grad_norm": 1.6932263374328613, + "learning_rate": 6.251308544102717e-05, + "loss": 0.1673, + "step": 26551 + }, + { + "epoch": 0.6224206942222379, + "grad_norm": 2.787454128265381, + "learning_rate": 6.25062565418298e-05, + "loss": 0.2897, + "step": 26552 + }, + { + "epoch": 0.6224441357970428, + "grad_norm": 3.2708849906921387, + "learning_rate": 6.249942784607458e-05, + "loss": 0.2485, + "step": 26553 + }, + { + "epoch": 0.6224675773718479, + "grad_norm": 2.3910019397735596, + "learning_rate": 6.249259935379852e-05, + "loss": 0.1798, + "step": 26554 + }, + { + "epoch": 0.6224910189466528, + "grad_norm": 3.4518990516662598, + "learning_rate": 6.248577106503864e-05, + "loss": 0.2899, + "step": 26555 + }, + { + "epoch": 0.6225144605214579, + "grad_norm": 0.9430679678916931, + "learning_rate": 6.247894297983203e-05, + "loss": 0.1098, + "step": 26556 + }, + { + "epoch": 0.6225379020962628, + "grad_norm": 3.973581314086914, + "learning_rate": 6.247211509821572e-05, + "loss": 0.2769, + "step": 26557 + }, + { + "epoch": 0.6225613436710679, + "grad_norm": 1.0770987272262573, + "learning_rate": 6.24652874202268e-05, + "loss": 0.0831, + "step": 26558 + }, + { + "epoch": 0.6225847852458728, + "grad_norm": 1.764184594154358, + "learning_rate": 6.245845994590225e-05, + "loss": 0.1284, + "step": 26559 + }, + { + "epoch": 0.6226082268206778, + "grad_norm": 3.7245047092437744, + "learning_rate": 6.245163267527918e-05, + "loss": 1.8789, + "step": 26560 + }, + { + "epoch": 0.6226316683954828, + "grad_norm": 1.894370436668396, + "learning_rate": 6.244480560839461e-05, + "loss": 1.4811, + "step": 26561 + }, + { + "epoch": 0.6226551099702878, + "grad_norm": 4.2197265625, + "learning_rate": 6.243797874528557e-05, + "loss": 0.3005, + "step": 26562 + }, + { + "epoch": 0.6226785515450928, + "grad_norm": 2.891634941101074, + "learning_rate": 6.24311520859891e-05, + "loss": 0.23, + "step": 26563 + }, + { + "epoch": 0.6227019931198978, + "grad_norm": 2.9869344234466553, + "learning_rate": 6.242432563054229e-05, + "loss": 0.2574, + "step": 26564 + }, + { + "epoch": 0.6227254346947028, + "grad_norm": 3.102447509765625, + "learning_rate": 6.241749937898212e-05, + "loss": 0.3381, + "step": 26565 + }, + { + "epoch": 0.6227488762695078, + "grad_norm": 5.895184516906738, + "learning_rate": 6.241067333134567e-05, + "loss": 0.2477, + "step": 26566 + }, + { + "epoch": 0.6227723178443128, + "grad_norm": 3.513104200363159, + "learning_rate": 6.240384748766996e-05, + "loss": 0.3423, + "step": 26567 + }, + { + "epoch": 0.6227957594191178, + "grad_norm": 2.190120220184326, + "learning_rate": 6.239702184799201e-05, + "loss": 0.1463, + "step": 26568 + }, + { + "epoch": 0.6228192009939227, + "grad_norm": 1.1460515260696411, + "learning_rate": 6.239019641234886e-05, + "loss": 0.0934, + "step": 26569 + }, + { + "epoch": 0.6228426425687278, + "grad_norm": 2.7961931228637695, + "learning_rate": 6.238337118077759e-05, + "loss": 0.2011, + "step": 26570 + }, + { + "epoch": 0.6228660841435327, + "grad_norm": 1.8920046091079712, + "learning_rate": 6.237654615331523e-05, + "loss": 0.1778, + "step": 26571 + }, + { + "epoch": 0.6228895257183378, + "grad_norm": 2.520287275314331, + "learning_rate": 6.236972132999874e-05, + "loss": 0.1866, + "step": 26572 + }, + { + "epoch": 0.6229129672931427, + "grad_norm": 2.605194330215454, + "learning_rate": 6.236289671086524e-05, + "loss": 0.1875, + "step": 26573 + }, + { + "epoch": 0.6229364088679478, + "grad_norm": 4.068037986755371, + "learning_rate": 6.235607229595173e-05, + "loss": 0.4777, + "step": 26574 + }, + { + "epoch": 0.6229598504427527, + "grad_norm": 1.348593831062317, + "learning_rate": 6.234924808529518e-05, + "loss": 0.0975, + "step": 26575 + }, + { + "epoch": 0.6229832920175578, + "grad_norm": 3.186915397644043, + "learning_rate": 6.234242407893272e-05, + "loss": 0.2551, + "step": 26576 + }, + { + "epoch": 0.6230067335923627, + "grad_norm": 0.6277115345001221, + "learning_rate": 6.233560027690131e-05, + "loss": 0.0313, + "step": 26577 + }, + { + "epoch": 0.6230301751671677, + "grad_norm": 3.6718223094940186, + "learning_rate": 6.232877667923796e-05, + "loss": 0.0719, + "step": 26578 + }, + { + "epoch": 0.6230536167419727, + "grad_norm": 1.2565311193466187, + "learning_rate": 6.232195328597977e-05, + "loss": 0.0953, + "step": 26579 + }, + { + "epoch": 0.6230770583167777, + "grad_norm": 1.654235601425171, + "learning_rate": 6.231513009716369e-05, + "loss": 0.1755, + "step": 26580 + }, + { + "epoch": 0.6231004998915827, + "grad_norm": 2.7678134441375732, + "learning_rate": 6.230830711282677e-05, + "loss": 0.2817, + "step": 26581 + }, + { + "epoch": 0.6231239414663877, + "grad_norm": 2.6080048084259033, + "learning_rate": 6.230148433300605e-05, + "loss": 0.0859, + "step": 26582 + }, + { + "epoch": 0.6231473830411927, + "grad_norm": 5.5063042640686035, + "learning_rate": 6.229466175773853e-05, + "loss": 0.4629, + "step": 26583 + }, + { + "epoch": 0.6231708246159977, + "grad_norm": 2.3550302982330322, + "learning_rate": 6.228783938706125e-05, + "loss": 0.136, + "step": 26584 + }, + { + "epoch": 0.6231942661908026, + "grad_norm": 2.020855665206909, + "learning_rate": 6.22810172210112e-05, + "loss": 0.0833, + "step": 26585 + }, + { + "epoch": 0.6232177077656077, + "grad_norm": 4.114763259887695, + "learning_rate": 6.227419525962544e-05, + "loss": 0.2641, + "step": 26586 + }, + { + "epoch": 0.6232411493404126, + "grad_norm": 2.3011059761047363, + "learning_rate": 6.22673735029409e-05, + "loss": 0.223, + "step": 26587 + }, + { + "epoch": 0.6232645909152177, + "grad_norm": 2.408480405807495, + "learning_rate": 6.22605519509947e-05, + "loss": 0.1862, + "step": 26588 + }, + { + "epoch": 0.6232880324900226, + "grad_norm": 3.307452917098999, + "learning_rate": 6.22537306038238e-05, + "loss": 0.2899, + "step": 26589 + }, + { + "epoch": 0.6233114740648277, + "grad_norm": 3.3852410316467285, + "learning_rate": 6.22469094614652e-05, + "loss": 0.3131, + "step": 26590 + }, + { + "epoch": 0.6233349156396327, + "grad_norm": 2.274399757385254, + "learning_rate": 6.224008852395595e-05, + "loss": 1.4156, + "step": 26591 + }, + { + "epoch": 0.6233583572144377, + "grad_norm": 2.45314884185791, + "learning_rate": 6.223326779133305e-05, + "loss": 0.1774, + "step": 26592 + }, + { + "epoch": 0.6233817987892427, + "grad_norm": 2.854952096939087, + "learning_rate": 6.222644726363346e-05, + "loss": 1.0917, + "step": 26593 + }, + { + "epoch": 0.6234052403640477, + "grad_norm": 4.81610107421875, + "learning_rate": 6.221962694089424e-05, + "loss": 0.2204, + "step": 26594 + }, + { + "epoch": 0.6234286819388527, + "grad_norm": 2.981382131576538, + "learning_rate": 6.22128068231524e-05, + "loss": 1.2731, + "step": 26595 + }, + { + "epoch": 0.6234521235136576, + "grad_norm": 3.398787498474121, + "learning_rate": 6.220598691044495e-05, + "loss": 0.2785, + "step": 26596 + }, + { + "epoch": 0.6234755650884627, + "grad_norm": 1.4855499267578125, + "learning_rate": 6.219916720280885e-05, + "loss": 0.147, + "step": 26597 + }, + { + "epoch": 0.6234990066632676, + "grad_norm": 2.985804557800293, + "learning_rate": 6.219234770028115e-05, + "loss": 0.2381, + "step": 26598 + }, + { + "epoch": 0.6235224482380727, + "grad_norm": 0.6653736233711243, + "learning_rate": 6.218552840289884e-05, + "loss": 0.0665, + "step": 26599 + }, + { + "epoch": 0.6235458898128776, + "grad_norm": 2.9995548725128174, + "learning_rate": 6.21787093106989e-05, + "loss": 0.2599, + "step": 26600 + }, + { + "epoch": 0.6235693313876827, + "grad_norm": 3.3090732097625732, + "learning_rate": 6.217189042371834e-05, + "loss": 0.2092, + "step": 26601 + }, + { + "epoch": 0.6235927729624876, + "grad_norm": 1.8948626518249512, + "learning_rate": 6.21650717419942e-05, + "loss": 1.1319, + "step": 26602 + }, + { + "epoch": 0.6236162145372927, + "grad_norm": 2.579340934753418, + "learning_rate": 6.21582532655634e-05, + "loss": 0.1683, + "step": 26603 + }, + { + "epoch": 0.6236396561120976, + "grad_norm": 2.0805256366729736, + "learning_rate": 6.215143499446301e-05, + "loss": 0.1388, + "step": 26604 + }, + { + "epoch": 0.6236630976869026, + "grad_norm": 2.8384649753570557, + "learning_rate": 6.214461692872996e-05, + "loss": 1.7953, + "step": 26605 + }, + { + "epoch": 0.6236865392617076, + "grad_norm": 1.8973867893218994, + "learning_rate": 6.213779906840134e-05, + "loss": 0.2058, + "step": 26606 + }, + { + "epoch": 0.6237099808365126, + "grad_norm": 2.0144944190979004, + "learning_rate": 6.213098141351403e-05, + "loss": 0.2242, + "step": 26607 + }, + { + "epoch": 0.6237334224113176, + "grad_norm": 0.35404732823371887, + "learning_rate": 6.212416396410511e-05, + "loss": 0.0274, + "step": 26608 + }, + { + "epoch": 0.6237568639861226, + "grad_norm": 1.3214328289031982, + "learning_rate": 6.211734672021157e-05, + "loss": 0.0963, + "step": 26609 + }, + { + "epoch": 0.6237803055609276, + "grad_norm": 2.786480665206909, + "learning_rate": 6.211052968187032e-05, + "loss": 0.171, + "step": 26610 + }, + { + "epoch": 0.6238037471357326, + "grad_norm": 2.6356022357940674, + "learning_rate": 6.210371284911845e-05, + "loss": 0.3142, + "step": 26611 + }, + { + "epoch": 0.6238271887105376, + "grad_norm": 4.200261116027832, + "learning_rate": 6.209689622199287e-05, + "loss": 0.3506, + "step": 26612 + }, + { + "epoch": 0.6238506302853426, + "grad_norm": 0.21884500980377197, + "learning_rate": 6.209007980053059e-05, + "loss": 0.0142, + "step": 26613 + }, + { + "epoch": 0.6238740718601475, + "grad_norm": 3.514893054962158, + "learning_rate": 6.208326358476863e-05, + "loss": 0.1282, + "step": 26614 + }, + { + "epoch": 0.6238975134349526, + "grad_norm": 4.519376754760742, + "learning_rate": 6.207644757474393e-05, + "loss": 0.338, + "step": 26615 + }, + { + "epoch": 0.6239209550097575, + "grad_norm": 2.377310037612915, + "learning_rate": 6.20696317704935e-05, + "loss": 0.074, + "step": 26616 + }, + { + "epoch": 0.6239443965845626, + "grad_norm": 3.1254382133483887, + "learning_rate": 6.206281617205427e-05, + "loss": 0.303, + "step": 26617 + }, + { + "epoch": 0.6239678381593675, + "grad_norm": 0.6482910513877869, + "learning_rate": 6.20560007794633e-05, + "loss": 0.0413, + "step": 26618 + }, + { + "epoch": 0.6239912797341726, + "grad_norm": 2.7003467082977295, + "learning_rate": 6.204918559275757e-05, + "loss": 0.234, + "step": 26619 + }, + { + "epoch": 0.6240147213089775, + "grad_norm": 3.2757389545440674, + "learning_rate": 6.204237061197397e-05, + "loss": 0.2298, + "step": 26620 + }, + { + "epoch": 0.6240381628837826, + "grad_norm": 3.3186659812927246, + "learning_rate": 6.203555583714957e-05, + "loss": 0.2245, + "step": 26621 + }, + { + "epoch": 0.6240616044585875, + "grad_norm": 0.5627231597900391, + "learning_rate": 6.202874126832133e-05, + "loss": 0.0171, + "step": 26622 + }, + { + "epoch": 0.6240850460333925, + "grad_norm": 2.1449272632598877, + "learning_rate": 6.202192690552615e-05, + "loss": 0.1925, + "step": 26623 + }, + { + "epoch": 0.6241084876081975, + "grad_norm": 2.161487340927124, + "learning_rate": 6.201511274880111e-05, + "loss": 1.6559, + "step": 26624 + }, + { + "epoch": 0.6241319291830025, + "grad_norm": 3.928898811340332, + "learning_rate": 6.20082987981831e-05, + "loss": 0.4891, + "step": 26625 + }, + { + "epoch": 0.6241553707578075, + "grad_norm": 1.3760449886322021, + "learning_rate": 6.200148505370916e-05, + "loss": 0.167, + "step": 26626 + }, + { + "epoch": 0.6241788123326125, + "grad_norm": 0.768990695476532, + "learning_rate": 6.199467151541622e-05, + "loss": 0.1156, + "step": 26627 + }, + { + "epoch": 0.6242022539074175, + "grad_norm": 2.6245276927948, + "learning_rate": 6.198785818334125e-05, + "loss": 0.13, + "step": 26628 + }, + { + "epoch": 0.6242256954822225, + "grad_norm": 4.18078088760376, + "learning_rate": 6.198104505752126e-05, + "loss": 0.5033, + "step": 26629 + }, + { + "epoch": 0.6242491370570274, + "grad_norm": 2.8333964347839355, + "learning_rate": 6.197423213799314e-05, + "loss": 0.1415, + "step": 26630 + }, + { + "epoch": 0.6242725786318325, + "grad_norm": 2.5898866653442383, + "learning_rate": 6.196741942479394e-05, + "loss": 0.2507, + "step": 26631 + }, + { + "epoch": 0.6242960202066374, + "grad_norm": 4.197666645050049, + "learning_rate": 6.196060691796057e-05, + "loss": 0.3171, + "step": 26632 + }, + { + "epoch": 0.6243194617814425, + "grad_norm": 2.007167339324951, + "learning_rate": 6.195379461753004e-05, + "loss": 0.1728, + "step": 26633 + }, + { + "epoch": 0.6243429033562474, + "grad_norm": 1.6762778759002686, + "learning_rate": 6.19469825235393e-05, + "loss": 0.1064, + "step": 26634 + }, + { + "epoch": 0.6243663449310525, + "grad_norm": 1.2979234457015991, + "learning_rate": 6.194017063602527e-05, + "loss": 0.1354, + "step": 26635 + }, + { + "epoch": 0.6243897865058574, + "grad_norm": 4.304891586303711, + "learning_rate": 6.193335895502497e-05, + "loss": 0.2851, + "step": 26636 + }, + { + "epoch": 0.6244132280806625, + "grad_norm": 2.36405348777771, + "learning_rate": 6.192654748057534e-05, + "loss": 0.2076, + "step": 26637 + }, + { + "epoch": 0.6244366696554674, + "grad_norm": 4.115187644958496, + "learning_rate": 6.19197362127133e-05, + "loss": 0.4699, + "step": 26638 + }, + { + "epoch": 0.6244601112302725, + "grad_norm": 4.237993240356445, + "learning_rate": 6.191292515147587e-05, + "loss": 0.3826, + "step": 26639 + }, + { + "epoch": 0.6244835528050774, + "grad_norm": 3.72572660446167, + "learning_rate": 6.190611429689996e-05, + "loss": 0.1639, + "step": 26640 + }, + { + "epoch": 0.6245069943798824, + "grad_norm": 0.5534528493881226, + "learning_rate": 6.189930364902255e-05, + "loss": 0.0427, + "step": 26641 + }, + { + "epoch": 0.6245304359546875, + "grad_norm": 3.7312469482421875, + "learning_rate": 6.189249320788057e-05, + "loss": 0.4129, + "step": 26642 + }, + { + "epoch": 0.6245538775294924, + "grad_norm": 2.490501880645752, + "learning_rate": 6.188568297351103e-05, + "loss": 1.3469, + "step": 26643 + }, + { + "epoch": 0.6245773191042975, + "grad_norm": 3.9767262935638428, + "learning_rate": 6.187887294595083e-05, + "loss": 0.2405, + "step": 26644 + }, + { + "epoch": 0.6246007606791024, + "grad_norm": 3.383168935775757, + "learning_rate": 6.187206312523694e-05, + "loss": 0.3076, + "step": 26645 + }, + { + "epoch": 0.6246242022539075, + "grad_norm": 3.327049493789673, + "learning_rate": 6.18652535114063e-05, + "loss": 0.3285, + "step": 26646 + }, + { + "epoch": 0.6246476438287124, + "grad_norm": 1.9566587209701538, + "learning_rate": 6.18584441044959e-05, + "loss": 0.173, + "step": 26647 + }, + { + "epoch": 0.6246710854035175, + "grad_norm": 0.8585577011108398, + "learning_rate": 6.185163490454261e-05, + "loss": 0.1058, + "step": 26648 + }, + { + "epoch": 0.6246945269783224, + "grad_norm": 0.7544001340866089, + "learning_rate": 6.184482591158345e-05, + "loss": 0.0753, + "step": 26649 + }, + { + "epoch": 0.6247179685531274, + "grad_norm": 2.354736328125, + "learning_rate": 6.183801712565533e-05, + "loss": 0.248, + "step": 26650 + }, + { + "epoch": 0.6247414101279324, + "grad_norm": 1.4813989400863647, + "learning_rate": 6.183120854679519e-05, + "loss": 0.1843, + "step": 26651 + }, + { + "epoch": 0.6247648517027374, + "grad_norm": 3.397416830062866, + "learning_rate": 6.182440017504e-05, + "loss": 1.6129, + "step": 26652 + }, + { + "epoch": 0.6247882932775424, + "grad_norm": 0.6005620360374451, + "learning_rate": 6.181759201042669e-05, + "loss": 0.0203, + "step": 26653 + }, + { + "epoch": 0.6248117348523474, + "grad_norm": 0.7661812901496887, + "learning_rate": 6.181078405299216e-05, + "loss": 0.1012, + "step": 26654 + }, + { + "epoch": 0.6248351764271524, + "grad_norm": 2.216327667236328, + "learning_rate": 6.180397630277343e-05, + "loss": 0.2475, + "step": 26655 + }, + { + "epoch": 0.6248586180019574, + "grad_norm": 0.9179754257202148, + "learning_rate": 6.179716875980739e-05, + "loss": 0.0954, + "step": 26656 + }, + { + "epoch": 0.6248820595767623, + "grad_norm": 3.4925575256347656, + "learning_rate": 6.1790361424131e-05, + "loss": 0.5104, + "step": 26657 + }, + { + "epoch": 0.6249055011515674, + "grad_norm": 4.315399169921875, + "learning_rate": 6.178355429578115e-05, + "loss": 0.504, + "step": 26658 + }, + { + "epoch": 0.6249289427263723, + "grad_norm": 3.5890285968780518, + "learning_rate": 6.177674737479485e-05, + "loss": 0.2964, + "step": 26659 + }, + { + "epoch": 0.6249523843011774, + "grad_norm": 0.571833610534668, + "learning_rate": 6.176994066120898e-05, + "loss": 0.0624, + "step": 26660 + }, + { + "epoch": 0.6249758258759823, + "grad_norm": 3.294248342514038, + "learning_rate": 6.176313415506047e-05, + "loss": 0.2803, + "step": 26661 + }, + { + "epoch": 0.6249992674507874, + "grad_norm": 2.969195604324341, + "learning_rate": 6.17563278563863e-05, + "loss": 0.3532, + "step": 26662 + }, + { + "epoch": 0.6250227090255923, + "grad_norm": 2.5175278186798096, + "learning_rate": 6.174952176522333e-05, + "loss": 0.9806, + "step": 26663 + }, + { + "epoch": 0.6250461506003974, + "grad_norm": 2.169102668762207, + "learning_rate": 6.174271588160856e-05, + "loss": 0.1661, + "step": 26664 + }, + { + "epoch": 0.6250695921752023, + "grad_norm": 3.4569592475891113, + "learning_rate": 6.173591020557888e-05, + "loss": 0.3392, + "step": 26665 + }, + { + "epoch": 0.6250930337500074, + "grad_norm": 2.5406644344329834, + "learning_rate": 6.17291047371712e-05, + "loss": 0.1772, + "step": 26666 + }, + { + "epoch": 0.6251164753248123, + "grad_norm": 2.4488673210144043, + "learning_rate": 6.172229947642249e-05, + "loss": 0.2957, + "step": 26667 + }, + { + "epoch": 0.6251399168996173, + "grad_norm": 2.975324869155884, + "learning_rate": 6.171549442336968e-05, + "loss": 0.2595, + "step": 26668 + }, + { + "epoch": 0.6251633584744223, + "grad_norm": 4.322503089904785, + "learning_rate": 6.170868957804968e-05, + "loss": 0.465, + "step": 26669 + }, + { + "epoch": 0.6251868000492273, + "grad_norm": 2.8012287616729736, + "learning_rate": 6.170188494049938e-05, + "loss": 1.8153, + "step": 26670 + }, + { + "epoch": 0.6252102416240323, + "grad_norm": 3.866065740585327, + "learning_rate": 6.169508051075575e-05, + "loss": 0.2195, + "step": 26671 + }, + { + "epoch": 0.6252336831988373, + "grad_norm": 3.2189555168151855, + "learning_rate": 6.168827628885567e-05, + "loss": 0.2623, + "step": 26672 + }, + { + "epoch": 0.6252571247736423, + "grad_norm": 3.060541868209839, + "learning_rate": 6.16814722748361e-05, + "loss": 1.5307, + "step": 26673 + }, + { + "epoch": 0.6252805663484473, + "grad_norm": 3.712155342102051, + "learning_rate": 6.167466846873393e-05, + "loss": 0.4838, + "step": 26674 + }, + { + "epoch": 0.6253040079232522, + "grad_norm": 1.1370762586593628, + "learning_rate": 6.16678648705861e-05, + "loss": 0.0617, + "step": 26675 + }, + { + "epoch": 0.6253274494980573, + "grad_norm": 1.6870959997177124, + "learning_rate": 6.166106148042948e-05, + "loss": 0.0732, + "step": 26676 + }, + { + "epoch": 0.6253508910728622, + "grad_norm": 0.4157399535179138, + "learning_rate": 6.165425829830105e-05, + "loss": 0.0403, + "step": 26677 + }, + { + "epoch": 0.6253743326476673, + "grad_norm": 3.2088470458984375, + "learning_rate": 6.164745532423764e-05, + "loss": 0.3471, + "step": 26678 + }, + { + "epoch": 0.6253977742224722, + "grad_norm": 3.157776355743408, + "learning_rate": 6.164065255827627e-05, + "loss": 0.2525, + "step": 26679 + }, + { + "epoch": 0.6254212157972773, + "grad_norm": 1.766910433769226, + "learning_rate": 6.163385000045377e-05, + "loss": 0.0968, + "step": 26680 + }, + { + "epoch": 0.6254446573720822, + "grad_norm": 1.9344156980514526, + "learning_rate": 6.162704765080709e-05, + "loss": 0.1136, + "step": 26681 + }, + { + "epoch": 0.6254680989468873, + "grad_norm": 3.42991042137146, + "learning_rate": 6.162024550937315e-05, + "loss": 0.2153, + "step": 26682 + }, + { + "epoch": 0.6254915405216922, + "grad_norm": 4.48040771484375, + "learning_rate": 6.16134435761888e-05, + "loss": 0.4434, + "step": 26683 + }, + { + "epoch": 0.6255149820964973, + "grad_norm": 2.2704975605010986, + "learning_rate": 6.160664185129101e-05, + "loss": 0.2567, + "step": 26684 + }, + { + "epoch": 0.6255384236713022, + "grad_norm": 1.1975440979003906, + "learning_rate": 6.159984033471667e-05, + "loss": 0.0497, + "step": 26685 + }, + { + "epoch": 0.6255618652461072, + "grad_norm": 2.427945375442505, + "learning_rate": 6.159303902650264e-05, + "loss": 0.1407, + "step": 26686 + }, + { + "epoch": 0.6255853068209122, + "grad_norm": 3.7115375995635986, + "learning_rate": 6.158623792668588e-05, + "loss": 0.3618, + "step": 26687 + }, + { + "epoch": 0.6256087483957172, + "grad_norm": 4.964528560638428, + "learning_rate": 6.157943703530329e-05, + "loss": 0.4144, + "step": 26688 + }, + { + "epoch": 0.6256321899705222, + "grad_norm": 1.592988133430481, + "learning_rate": 6.157263635239172e-05, + "loss": 0.1011, + "step": 26689 + }, + { + "epoch": 0.6256556315453272, + "grad_norm": 2.083876132965088, + "learning_rate": 6.15658358779881e-05, + "loss": 0.2818, + "step": 26690 + }, + { + "epoch": 0.6256790731201322, + "grad_norm": 2.4404118061065674, + "learning_rate": 6.155903561212937e-05, + "loss": 0.1885, + "step": 26691 + }, + { + "epoch": 0.6257025146949372, + "grad_norm": 2.696155071258545, + "learning_rate": 6.15522355548524e-05, + "loss": 1.4429, + "step": 26692 + }, + { + "epoch": 0.6257259562697423, + "grad_norm": 3.5603585243225098, + "learning_rate": 6.154543570619403e-05, + "loss": 0.222, + "step": 26693 + }, + { + "epoch": 0.6257493978445472, + "grad_norm": 3.2068402767181396, + "learning_rate": 6.153863606619126e-05, + "loss": 0.2629, + "step": 26694 + }, + { + "epoch": 0.6257728394193522, + "grad_norm": 2.379483699798584, + "learning_rate": 6.153183663488091e-05, + "loss": 0.214, + "step": 26695 + }, + { + "epoch": 0.6257962809941572, + "grad_norm": 3.698396682739258, + "learning_rate": 6.152503741229989e-05, + "loss": 0.5332, + "step": 26696 + }, + { + "epoch": 0.6258197225689622, + "grad_norm": 1.6773302555084229, + "learning_rate": 6.151823839848512e-05, + "loss": 0.1018, + "step": 26697 + }, + { + "epoch": 0.6258431641437672, + "grad_norm": 1.183493733406067, + "learning_rate": 6.151143959347343e-05, + "loss": 0.1597, + "step": 26698 + }, + { + "epoch": 0.6258666057185722, + "grad_norm": 2.0656211376190186, + "learning_rate": 6.15046409973018e-05, + "loss": 0.1237, + "step": 26699 + }, + { + "epoch": 0.6258900472933772, + "grad_norm": 4.905429840087891, + "learning_rate": 6.149784261000705e-05, + "loss": 0.5114, + "step": 26700 + }, + { + "epoch": 0.6259134888681822, + "grad_norm": 2.3233211040496826, + "learning_rate": 6.149104443162607e-05, + "loss": 0.1908, + "step": 26701 + }, + { + "epoch": 0.6259369304429871, + "grad_norm": 2.3645613193511963, + "learning_rate": 6.14842464621958e-05, + "loss": 0.165, + "step": 26702 + }, + { + "epoch": 0.6259603720177922, + "grad_norm": 1.9419760704040527, + "learning_rate": 6.147744870175303e-05, + "loss": 0.0939, + "step": 26703 + }, + { + "epoch": 0.6259838135925971, + "grad_norm": 3.100041389465332, + "learning_rate": 6.147065115033476e-05, + "loss": 0.4411, + "step": 26704 + }, + { + "epoch": 0.6260072551674022, + "grad_norm": 0.6302490234375, + "learning_rate": 6.146385380797781e-05, + "loss": 0.0447, + "step": 26705 + }, + { + "epoch": 0.6260306967422071, + "grad_norm": 0.8049345016479492, + "learning_rate": 6.145705667471908e-05, + "loss": 0.0402, + "step": 26706 + }, + { + "epoch": 0.6260541383170122, + "grad_norm": 1.5397005081176758, + "learning_rate": 6.145025975059544e-05, + "loss": 0.1365, + "step": 26707 + }, + { + "epoch": 0.6260775798918171, + "grad_norm": 1.8992528915405273, + "learning_rate": 6.144346303564376e-05, + "loss": 0.1228, + "step": 26708 + }, + { + "epoch": 0.6261010214666222, + "grad_norm": 4.182703018188477, + "learning_rate": 6.143666652990096e-05, + "loss": 0.6517, + "step": 26709 + }, + { + "epoch": 0.6261244630414271, + "grad_norm": 3.7494871616363525, + "learning_rate": 6.142987023340389e-05, + "loss": 0.1863, + "step": 26710 + }, + { + "epoch": 0.6261479046162322, + "grad_norm": 1.485180139541626, + "learning_rate": 6.14230741461894e-05, + "loss": 0.1689, + "step": 26711 + }, + { + "epoch": 0.6261713461910371, + "grad_norm": 2.858905792236328, + "learning_rate": 6.141627826829442e-05, + "loss": 0.4073, + "step": 26712 + }, + { + "epoch": 0.6261947877658421, + "grad_norm": 3.6032538414001465, + "learning_rate": 6.14094825997558e-05, + "loss": 0.3418, + "step": 26713 + }, + { + "epoch": 0.6262182293406471, + "grad_norm": 2.076669931411743, + "learning_rate": 6.14026871406104e-05, + "loss": 0.207, + "step": 26714 + }, + { + "epoch": 0.6262416709154521, + "grad_norm": 2.6930577754974365, + "learning_rate": 6.139589189089506e-05, + "loss": 0.186, + "step": 26715 + }, + { + "epoch": 0.6262651124902571, + "grad_norm": 4.922543525695801, + "learning_rate": 6.138909685064676e-05, + "loss": 0.6257, + "step": 26716 + }, + { + "epoch": 0.6262885540650621, + "grad_norm": 0.6326761245727539, + "learning_rate": 6.13823020199023e-05, + "loss": 0.102, + "step": 26717 + }, + { + "epoch": 0.6263119956398671, + "grad_norm": 3.3328676223754883, + "learning_rate": 6.137550739869852e-05, + "loss": 0.9457, + "step": 26718 + }, + { + "epoch": 0.6263354372146721, + "grad_norm": 4.465683937072754, + "learning_rate": 6.136871298707236e-05, + "loss": 0.7067, + "step": 26719 + }, + { + "epoch": 0.626358878789477, + "grad_norm": 2.550724506378174, + "learning_rate": 6.136191878506066e-05, + "loss": 0.1923, + "step": 26720 + }, + { + "epoch": 0.6263823203642821, + "grad_norm": 3.0956337451934814, + "learning_rate": 6.135512479270022e-05, + "loss": 0.3191, + "step": 26721 + }, + { + "epoch": 0.626405761939087, + "grad_norm": 0.9794039726257324, + "learning_rate": 6.134833101002802e-05, + "loss": 0.1025, + "step": 26722 + }, + { + "epoch": 0.6264292035138921, + "grad_norm": 1.2084167003631592, + "learning_rate": 6.134153743708084e-05, + "loss": 0.1524, + "step": 26723 + }, + { + "epoch": 0.626452645088697, + "grad_norm": 1.6812055110931396, + "learning_rate": 6.133474407389555e-05, + "loss": 0.0923, + "step": 26724 + }, + { + "epoch": 0.6264760866635021, + "grad_norm": 3.6414904594421387, + "learning_rate": 6.132795092050902e-05, + "loss": 0.4836, + "step": 26725 + }, + { + "epoch": 0.626499528238307, + "grad_norm": 3.451482057571411, + "learning_rate": 6.132115797695815e-05, + "loss": 0.4772, + "step": 26726 + }, + { + "epoch": 0.6265229698131121, + "grad_norm": 2.9344654083251953, + "learning_rate": 6.131436524327972e-05, + "loss": 0.2665, + "step": 26727 + }, + { + "epoch": 0.626546411387917, + "grad_norm": 3.453620672225952, + "learning_rate": 6.130757271951064e-05, + "loss": 0.1078, + "step": 26728 + }, + { + "epoch": 0.626569852962722, + "grad_norm": 1.9793963432312012, + "learning_rate": 6.130078040568778e-05, + "loss": 0.1685, + "step": 26729 + }, + { + "epoch": 0.626593294537527, + "grad_norm": 3.8304286003112793, + "learning_rate": 6.129398830184798e-05, + "loss": 0.2445, + "step": 26730 + }, + { + "epoch": 0.626616736112332, + "grad_norm": 3.7958855628967285, + "learning_rate": 6.128719640802806e-05, + "loss": 0.2761, + "step": 26731 + }, + { + "epoch": 0.626640177687137, + "grad_norm": 2.2400100231170654, + "learning_rate": 6.128040472426492e-05, + "loss": 0.0728, + "step": 26732 + }, + { + "epoch": 0.626663619261942, + "grad_norm": 5.179642677307129, + "learning_rate": 6.127361325059541e-05, + "loss": 0.3899, + "step": 26733 + }, + { + "epoch": 0.626687060836747, + "grad_norm": 2.2438015937805176, + "learning_rate": 6.126682198705633e-05, + "loss": 0.1437, + "step": 26734 + }, + { + "epoch": 0.626710502411552, + "grad_norm": 3.943455457687378, + "learning_rate": 6.126003093368459e-05, + "loss": 0.4218, + "step": 26735 + }, + { + "epoch": 0.626733943986357, + "grad_norm": 1.414581537246704, + "learning_rate": 6.125324009051698e-05, + "loss": 0.0705, + "step": 26736 + }, + { + "epoch": 0.626757385561162, + "grad_norm": 1.8829413652420044, + "learning_rate": 6.124644945759041e-05, + "loss": 0.2202, + "step": 26737 + }, + { + "epoch": 0.626780827135967, + "grad_norm": 4.301530838012695, + "learning_rate": 6.123965903494169e-05, + "loss": 0.403, + "step": 26738 + }, + { + "epoch": 0.626804268710772, + "grad_norm": 0.8930493593215942, + "learning_rate": 6.123286882260763e-05, + "loss": 0.0504, + "step": 26739 + }, + { + "epoch": 0.6268277102855769, + "grad_norm": 2.4781675338745117, + "learning_rate": 6.122607882062516e-05, + "loss": 0.3442, + "step": 26740 + }, + { + "epoch": 0.626851151860382, + "grad_norm": 2.970149040222168, + "learning_rate": 6.121928902903104e-05, + "loss": 0.2322, + "step": 26741 + }, + { + "epoch": 0.6268745934351869, + "grad_norm": 2.1452698707580566, + "learning_rate": 6.121249944786219e-05, + "loss": 0.2043, + "step": 26742 + }, + { + "epoch": 0.626898035009992, + "grad_norm": 3.4752044677734375, + "learning_rate": 6.120571007715537e-05, + "loss": 0.2752, + "step": 26743 + }, + { + "epoch": 0.626921476584797, + "grad_norm": 3.3942909240722656, + "learning_rate": 6.119892091694748e-05, + "loss": 1.6819, + "step": 26744 + }, + { + "epoch": 0.626944918159602, + "grad_norm": 2.0491223335266113, + "learning_rate": 6.119213196727534e-05, + "loss": 0.1427, + "step": 26745 + }, + { + "epoch": 0.626968359734407, + "grad_norm": 3.284388303756714, + "learning_rate": 6.118534322817577e-05, + "loss": 0.1599, + "step": 26746 + }, + { + "epoch": 0.626991801309212, + "grad_norm": 0.9687182903289795, + "learning_rate": 6.117855469968563e-05, + "loss": 0.0549, + "step": 26747 + }, + { + "epoch": 0.627015242884017, + "grad_norm": 1.4170678853988647, + "learning_rate": 6.117176638184174e-05, + "loss": 0.0449, + "step": 26748 + }, + { + "epoch": 0.6270386844588219, + "grad_norm": 2.34346342086792, + "learning_rate": 6.116497827468093e-05, + "loss": 1.418, + "step": 26749 + }, + { + "epoch": 0.627062126033627, + "grad_norm": 3.7742207050323486, + "learning_rate": 6.115819037824005e-05, + "loss": 0.1465, + "step": 26750 + }, + { + "epoch": 0.6270855676084319, + "grad_norm": 1.6409908533096313, + "learning_rate": 6.11514026925559e-05, + "loss": 1.0264, + "step": 26751 + }, + { + "epoch": 0.627109009183237, + "grad_norm": 1.042323350906372, + "learning_rate": 6.114461521766534e-05, + "loss": 0.0415, + "step": 26752 + }, + { + "epoch": 0.6271324507580419, + "grad_norm": 2.712869167327881, + "learning_rate": 6.113782795360519e-05, + "loss": 0.3546, + "step": 26753 + }, + { + "epoch": 0.627155892332847, + "grad_norm": 2.7401299476623535, + "learning_rate": 6.113104090041231e-05, + "loss": 0.9908, + "step": 26754 + }, + { + "epoch": 0.6271793339076519, + "grad_norm": 2.5226168632507324, + "learning_rate": 6.112425405812348e-05, + "loss": 0.1821, + "step": 26755 + }, + { + "epoch": 0.627202775482457, + "grad_norm": 0.35021737217903137, + "learning_rate": 6.111746742677551e-05, + "loss": 0.0146, + "step": 26756 + }, + { + "epoch": 0.6272262170572619, + "grad_norm": 4.124192237854004, + "learning_rate": 6.11106810064053e-05, + "loss": 0.4738, + "step": 26757 + }, + { + "epoch": 0.6272496586320669, + "grad_norm": 3.1187713146209717, + "learning_rate": 6.11038947970496e-05, + "loss": 2.0176, + "step": 26758 + }, + { + "epoch": 0.6272731002068719, + "grad_norm": 0.3444940149784088, + "learning_rate": 6.109710879874526e-05, + "loss": 0.0285, + "step": 26759 + }, + { + "epoch": 0.6272965417816769, + "grad_norm": 0.8473650813102722, + "learning_rate": 6.10903230115291e-05, + "loss": 0.2685, + "step": 26760 + }, + { + "epoch": 0.6273199833564819, + "grad_norm": 2.3874242305755615, + "learning_rate": 6.108353743543796e-05, + "loss": 0.2028, + "step": 26761 + }, + { + "epoch": 0.6273434249312869, + "grad_norm": 2.177529811859131, + "learning_rate": 6.107675207050862e-05, + "loss": 0.0812, + "step": 26762 + }, + { + "epoch": 0.6273668665060919, + "grad_norm": 0.9677981734275818, + "learning_rate": 6.10699669167779e-05, + "loss": 0.1067, + "step": 26763 + }, + { + "epoch": 0.6273903080808969, + "grad_norm": 0.6202557682991028, + "learning_rate": 6.106318197428267e-05, + "loss": 0.0701, + "step": 26764 + }, + { + "epoch": 0.6274137496557018, + "grad_norm": 0.6860631704330444, + "learning_rate": 6.105639724305971e-05, + "loss": 0.0298, + "step": 26765 + }, + { + "epoch": 0.6274371912305069, + "grad_norm": 3.212783098220825, + "learning_rate": 6.104961272314581e-05, + "loss": 0.3704, + "step": 26766 + }, + { + "epoch": 0.6274606328053118, + "grad_norm": 5.631601810455322, + "learning_rate": 6.104282841457782e-05, + "loss": 0.423, + "step": 26767 + }, + { + "epoch": 0.6274840743801169, + "grad_norm": 3.160090684890747, + "learning_rate": 6.103604431739256e-05, + "loss": 0.1909, + "step": 26768 + }, + { + "epoch": 0.6275075159549218, + "grad_norm": 5.405327320098877, + "learning_rate": 6.102926043162678e-05, + "loss": 0.3101, + "step": 26769 + }, + { + "epoch": 0.6275309575297269, + "grad_norm": 2.0685300827026367, + "learning_rate": 6.102247675731736e-05, + "loss": 0.1649, + "step": 26770 + }, + { + "epoch": 0.6275543991045318, + "grad_norm": 3.284531593322754, + "learning_rate": 6.101569329450108e-05, + "loss": 0.4251, + "step": 26771 + }, + { + "epoch": 0.6275778406793369, + "grad_norm": 2.587435245513916, + "learning_rate": 6.1008910043214705e-05, + "loss": 0.2536, + "step": 26772 + }, + { + "epoch": 0.6276012822541418, + "grad_norm": 2.8943183422088623, + "learning_rate": 6.100212700349511e-05, + "loss": 0.1777, + "step": 26773 + }, + { + "epoch": 0.6276247238289469, + "grad_norm": 3.173055410385132, + "learning_rate": 6.0995344175379055e-05, + "loss": 0.252, + "step": 26774 + }, + { + "epoch": 0.6276481654037518, + "grad_norm": 3.9305343627929688, + "learning_rate": 6.098856155890338e-05, + "loss": 0.2275, + "step": 26775 + }, + { + "epoch": 0.6276716069785568, + "grad_norm": 2.708338499069214, + "learning_rate": 6.098177915410481e-05, + "loss": 0.192, + "step": 26776 + }, + { + "epoch": 0.6276950485533618, + "grad_norm": 1.6087497472763062, + "learning_rate": 6.097499696102027e-05, + "loss": 0.118, + "step": 26777 + }, + { + "epoch": 0.6277184901281668, + "grad_norm": 5.555713653564453, + "learning_rate": 6.096821497968648e-05, + "loss": 0.3909, + "step": 26778 + }, + { + "epoch": 0.6277419317029718, + "grad_norm": 2.501190185546875, + "learning_rate": 6.096143321014023e-05, + "loss": 0.1768, + "step": 26779 + }, + { + "epoch": 0.6277653732777768, + "grad_norm": 3.8614912033081055, + "learning_rate": 6.095465165241836e-05, + "loss": 0.4186, + "step": 26780 + }, + { + "epoch": 0.6277888148525818, + "grad_norm": 0.6251235604286194, + "learning_rate": 6.094787030655764e-05, + "loss": 0.0646, + "step": 26781 + }, + { + "epoch": 0.6278122564273868, + "grad_norm": 1.7682615518569946, + "learning_rate": 6.094108917259489e-05, + "loss": 0.0925, + "step": 26782 + }, + { + "epoch": 0.6278356980021917, + "grad_norm": 3.8240773677825928, + "learning_rate": 6.093430825056687e-05, + "loss": 0.4882, + "step": 26783 + }, + { + "epoch": 0.6278591395769968, + "grad_norm": 1.628903865814209, + "learning_rate": 6.092752754051039e-05, + "loss": 0.0843, + "step": 26784 + }, + { + "epoch": 0.6278825811518017, + "grad_norm": 1.2875947952270508, + "learning_rate": 6.0920747042462266e-05, + "loss": 0.1432, + "step": 26785 + }, + { + "epoch": 0.6279060227266068, + "grad_norm": 3.2072973251342773, + "learning_rate": 6.091396675645926e-05, + "loss": 0.4535, + "step": 26786 + }, + { + "epoch": 0.6279294643014117, + "grad_norm": 2.239539384841919, + "learning_rate": 6.090718668253815e-05, + "loss": 0.146, + "step": 26787 + }, + { + "epoch": 0.6279529058762168, + "grad_norm": 2.1687958240509033, + "learning_rate": 6.0900406820735726e-05, + "loss": 0.1576, + "step": 26788 + }, + { + "epoch": 0.6279763474510217, + "grad_norm": 2.596857786178589, + "learning_rate": 6.089362717108884e-05, + "loss": 0.1826, + "step": 26789 + }, + { + "epoch": 0.6279997890258268, + "grad_norm": 3.471294403076172, + "learning_rate": 6.088684773363422e-05, + "loss": 0.3417, + "step": 26790 + }, + { + "epoch": 0.6280232306006317, + "grad_norm": 2.890929937362671, + "learning_rate": 6.088006850840865e-05, + "loss": 0.2896, + "step": 26791 + }, + { + "epoch": 0.6280466721754367, + "grad_norm": 1.5711115598678589, + "learning_rate": 6.0873289495448946e-05, + "loss": 0.0789, + "step": 26792 + }, + { + "epoch": 0.6280701137502417, + "grad_norm": 3.6233513355255127, + "learning_rate": 6.086651069479188e-05, + "loss": 0.4584, + "step": 26793 + }, + { + "epoch": 0.6280935553250467, + "grad_norm": 3.372394323348999, + "learning_rate": 6.08597321064742e-05, + "loss": 0.2654, + "step": 26794 + }, + { + "epoch": 0.6281169968998518, + "grad_norm": 2.6630797386169434, + "learning_rate": 6.085295373053275e-05, + "loss": 1.5169, + "step": 26795 + }, + { + "epoch": 0.6281404384746567, + "grad_norm": 1.0810127258300781, + "learning_rate": 6.0846175567004246e-05, + "loss": 0.0544, + "step": 26796 + }, + { + "epoch": 0.6281638800494618, + "grad_norm": 3.7177958488464355, + "learning_rate": 6.08393976159255e-05, + "loss": 0.3727, + "step": 26797 + }, + { + "epoch": 0.6281873216242667, + "grad_norm": 0.7511110305786133, + "learning_rate": 6.0832619877333284e-05, + "loss": 0.0824, + "step": 26798 + }, + { + "epoch": 0.6282107631990718, + "grad_norm": 3.6999571323394775, + "learning_rate": 6.0825842351264385e-05, + "loss": 0.1265, + "step": 26799 + }, + { + "epoch": 0.6282342047738767, + "grad_norm": 4.635721206665039, + "learning_rate": 6.081906503775553e-05, + "loss": 0.3938, + "step": 26800 + }, + { + "epoch": 0.6282576463486818, + "grad_norm": 2.506054401397705, + "learning_rate": 6.081228793684355e-05, + "loss": 0.1539, + "step": 26801 + }, + { + "epoch": 0.6282810879234867, + "grad_norm": 2.4635496139526367, + "learning_rate": 6.08055110485652e-05, + "loss": 1.4551, + "step": 26802 + }, + { + "epoch": 0.6283045294982917, + "grad_norm": 5.820284843444824, + "learning_rate": 6.0798734372957264e-05, + "loss": 0.4399, + "step": 26803 + }, + { + "epoch": 0.6283279710730967, + "grad_norm": 1.2428061962127686, + "learning_rate": 6.0791957910056473e-05, + "loss": 0.1168, + "step": 26804 + }, + { + "epoch": 0.6283514126479017, + "grad_norm": 2.432310104370117, + "learning_rate": 6.0785181659899635e-05, + "loss": 0.2122, + "step": 26805 + }, + { + "epoch": 0.6283748542227067, + "grad_norm": 2.6293931007385254, + "learning_rate": 6.0778405622523525e-05, + "loss": 0.2053, + "step": 26806 + }, + { + "epoch": 0.6283982957975117, + "grad_norm": 1.7905129194259644, + "learning_rate": 6.077162979796485e-05, + "loss": 0.1612, + "step": 26807 + }, + { + "epoch": 0.6284217373723167, + "grad_norm": 1.9857854843139648, + "learning_rate": 6.076485418626044e-05, + "loss": 0.134, + "step": 26808 + }, + { + "epoch": 0.6284451789471217, + "grad_norm": 4.76678991317749, + "learning_rate": 6.0758078787447036e-05, + "loss": 0.5022, + "step": 26809 + }, + { + "epoch": 0.6284686205219266, + "grad_norm": 2.2930619716644287, + "learning_rate": 6.075130360156138e-05, + "loss": 0.0962, + "step": 26810 + }, + { + "epoch": 0.6284920620967317, + "grad_norm": 4.307422637939453, + "learning_rate": 6.0744528628640265e-05, + "loss": 0.362, + "step": 26811 + }, + { + "epoch": 0.6285155036715366, + "grad_norm": 3.8539164066314697, + "learning_rate": 6.073775386872042e-05, + "loss": 0.2877, + "step": 26812 + }, + { + "epoch": 0.6285389452463417, + "grad_norm": 3.5128846168518066, + "learning_rate": 6.0730979321838646e-05, + "loss": 0.2903, + "step": 26813 + }, + { + "epoch": 0.6285623868211466, + "grad_norm": 2.4020893573760986, + "learning_rate": 6.072420498803167e-05, + "loss": 1.5694, + "step": 26814 + }, + { + "epoch": 0.6285858283959517, + "grad_norm": 0.9626052975654602, + "learning_rate": 6.071743086733629e-05, + "loss": 0.0588, + "step": 26815 + }, + { + "epoch": 0.6286092699707566, + "grad_norm": 3.601109504699707, + "learning_rate": 6.0710656959789215e-05, + "loss": 0.2001, + "step": 26816 + }, + { + "epoch": 0.6286327115455617, + "grad_norm": 1.0696229934692383, + "learning_rate": 6.070388326542722e-05, + "loss": 0.1237, + "step": 26817 + }, + { + "epoch": 0.6286561531203666, + "grad_norm": 3.9973769187927246, + "learning_rate": 6.0697109784287074e-05, + "loss": 0.3424, + "step": 26818 + }, + { + "epoch": 0.6286795946951717, + "grad_norm": 2.8459246158599854, + "learning_rate": 6.06903365164055e-05, + "loss": 0.2613, + "step": 26819 + }, + { + "epoch": 0.6287030362699766, + "grad_norm": 1.0463578701019287, + "learning_rate": 6.068356346181927e-05, + "loss": 0.1216, + "step": 26820 + }, + { + "epoch": 0.6287264778447816, + "grad_norm": 0.7394016981124878, + "learning_rate": 6.067679062056515e-05, + "loss": 0.0411, + "step": 26821 + }, + { + "epoch": 0.6287499194195866, + "grad_norm": 3.655644655227661, + "learning_rate": 6.067001799267984e-05, + "loss": 0.3652, + "step": 26822 + }, + { + "epoch": 0.6287733609943916, + "grad_norm": 3.5503623485565186, + "learning_rate": 6.0663245578200135e-05, + "loss": 0.2539, + "step": 26823 + }, + { + "epoch": 0.6287968025691966, + "grad_norm": 1.9457494020462036, + "learning_rate": 6.065647337716273e-05, + "loss": 0.232, + "step": 26824 + }, + { + "epoch": 0.6288202441440016, + "grad_norm": 2.952392578125, + "learning_rate": 6.064970138960444e-05, + "loss": 0.2337, + "step": 26825 + }, + { + "epoch": 0.6288436857188066, + "grad_norm": 3.6709108352661133, + "learning_rate": 6.064292961556196e-05, + "loss": 0.3269, + "step": 26826 + }, + { + "epoch": 0.6288671272936116, + "grad_norm": 2.3742191791534424, + "learning_rate": 6.0636158055072075e-05, + "loss": 0.3068, + "step": 26827 + }, + { + "epoch": 0.6288905688684165, + "grad_norm": 0.8205605149269104, + "learning_rate": 6.062938670817151e-05, + "loss": 0.0661, + "step": 26828 + }, + { + "epoch": 0.6289140104432216, + "grad_norm": 1.4788837432861328, + "learning_rate": 6.062261557489697e-05, + "loss": 0.8283, + "step": 26829 + }, + { + "epoch": 0.6289374520180265, + "grad_norm": 1.5031907558441162, + "learning_rate": 6.061584465528524e-05, + "loss": 0.1018, + "step": 26830 + }, + { + "epoch": 0.6289608935928316, + "grad_norm": 0.885431170463562, + "learning_rate": 6.0609073949373055e-05, + "loss": 0.0759, + "step": 26831 + }, + { + "epoch": 0.6289843351676365, + "grad_norm": 1.158325433731079, + "learning_rate": 6.060230345719711e-05, + "loss": 0.1405, + "step": 26832 + }, + { + "epoch": 0.6290077767424416, + "grad_norm": 3.694087505340576, + "learning_rate": 6.0595533178794205e-05, + "loss": 0.2789, + "step": 26833 + }, + { + "epoch": 0.6290312183172465, + "grad_norm": 2.598937511444092, + "learning_rate": 6.058876311420103e-05, + "loss": 0.2911, + "step": 26834 + }, + { + "epoch": 0.6290546598920516, + "grad_norm": 0.48286521434783936, + "learning_rate": 6.058199326345433e-05, + "loss": 0.0362, + "step": 26835 + }, + { + "epoch": 0.6290781014668565, + "grad_norm": 4.224769592285156, + "learning_rate": 6.057522362659082e-05, + "loss": 0.3073, + "step": 26836 + }, + { + "epoch": 0.6291015430416615, + "grad_norm": 6.970907211303711, + "learning_rate": 6.0568454203647295e-05, + "loss": 0.4394, + "step": 26837 + }, + { + "epoch": 0.6291249846164665, + "grad_norm": 2.169065237045288, + "learning_rate": 6.056168499466044e-05, + "loss": 0.0793, + "step": 26838 + }, + { + "epoch": 0.6291484261912715, + "grad_norm": 2.1993510723114014, + "learning_rate": 6.055491599966698e-05, + "loss": 0.1818, + "step": 26839 + }, + { + "epoch": 0.6291718677660765, + "grad_norm": 3.8213260173797607, + "learning_rate": 6.054814721870367e-05, + "loss": 0.4701, + "step": 26840 + }, + { + "epoch": 0.6291953093408815, + "grad_norm": 4.346488952636719, + "learning_rate": 6.0541378651807224e-05, + "loss": 0.5922, + "step": 26841 + }, + { + "epoch": 0.6292187509156865, + "grad_norm": 3.2883715629577637, + "learning_rate": 6.053461029901435e-05, + "loss": 0.1929, + "step": 26842 + }, + { + "epoch": 0.6292421924904915, + "grad_norm": 3.5868186950683594, + "learning_rate": 6.0527842160361804e-05, + "loss": 0.2239, + "step": 26843 + }, + { + "epoch": 0.6292656340652965, + "grad_norm": 3.723437786102295, + "learning_rate": 6.052107423588629e-05, + "loss": 0.3145, + "step": 26844 + }, + { + "epoch": 0.6292890756401015, + "grad_norm": 3.4295268058776855, + "learning_rate": 6.0514306525624534e-05, + "loss": 0.2606, + "step": 26845 + }, + { + "epoch": 0.6293125172149066, + "grad_norm": 5.666292667388916, + "learning_rate": 6.050753902961327e-05, + "loss": 0.579, + "step": 26846 + }, + { + "epoch": 0.6293359587897115, + "grad_norm": 0.6958510279655457, + "learning_rate": 6.050077174788922e-05, + "loss": 0.07, + "step": 26847 + }, + { + "epoch": 0.6293594003645165, + "grad_norm": 3.7562222480773926, + "learning_rate": 6.0494004680489036e-05, + "loss": 0.4742, + "step": 26848 + }, + { + "epoch": 0.6293828419393215, + "grad_norm": 3.5284581184387207, + "learning_rate": 6.048723782744952e-05, + "loss": 0.2857, + "step": 26849 + }, + { + "epoch": 0.6294062835141265, + "grad_norm": 3.687129497528076, + "learning_rate": 6.048047118880738e-05, + "loss": 0.2766, + "step": 26850 + }, + { + "epoch": 0.6294297250889315, + "grad_norm": 2.860748529434204, + "learning_rate": 6.047370476459933e-05, + "loss": 0.3078, + "step": 26851 + }, + { + "epoch": 0.6294531666637365, + "grad_norm": 3.5887749195098877, + "learning_rate": 6.046693855486203e-05, + "loss": 0.3187, + "step": 26852 + }, + { + "epoch": 0.6294766082385415, + "grad_norm": 2.2103614807128906, + "learning_rate": 6.046017255963228e-05, + "loss": 0.0721, + "step": 26853 + }, + { + "epoch": 0.6295000498133465, + "grad_norm": 2.712900161743164, + "learning_rate": 6.04534067789467e-05, + "loss": 0.1274, + "step": 26854 + }, + { + "epoch": 0.6295234913881514, + "grad_norm": 3.374983549118042, + "learning_rate": 6.044664121284207e-05, + "loss": 0.4074, + "step": 26855 + }, + { + "epoch": 0.6295469329629565, + "grad_norm": 5.097738742828369, + "learning_rate": 6.043987586135509e-05, + "loss": 0.4632, + "step": 26856 + }, + { + "epoch": 0.6295703745377614, + "grad_norm": 3.111389636993408, + "learning_rate": 6.043311072452242e-05, + "loss": 0.3085, + "step": 26857 + }, + { + "epoch": 0.6295938161125665, + "grad_norm": 3.028926372528076, + "learning_rate": 6.0426345802380824e-05, + "loss": 0.3071, + "step": 26858 + }, + { + "epoch": 0.6296172576873714, + "grad_norm": 2.221557855606079, + "learning_rate": 6.0419581094967004e-05, + "loss": 0.1479, + "step": 26859 + }, + { + "epoch": 0.6296406992621765, + "grad_norm": 2.874748706817627, + "learning_rate": 6.041281660231762e-05, + "loss": 0.3095, + "step": 26860 + }, + { + "epoch": 0.6296641408369814, + "grad_norm": 2.668335199356079, + "learning_rate": 6.040605232446939e-05, + "loss": 0.413, + "step": 26861 + }, + { + "epoch": 0.6296875824117865, + "grad_norm": 2.682339906692505, + "learning_rate": 6.0399288261459064e-05, + "loss": 0.3973, + "step": 26862 + }, + { + "epoch": 0.6297110239865914, + "grad_norm": 2.808504343032837, + "learning_rate": 6.039252441332333e-05, + "loss": 0.204, + "step": 26863 + }, + { + "epoch": 0.6297344655613965, + "grad_norm": 2.6793317794799805, + "learning_rate": 6.038576078009883e-05, + "loss": 1.3074, + "step": 26864 + }, + { + "epoch": 0.6297579071362014, + "grad_norm": 1.0670585632324219, + "learning_rate": 6.037899736182235e-05, + "loss": 0.1438, + "step": 26865 + }, + { + "epoch": 0.6297813487110064, + "grad_norm": 0.6468392014503479, + "learning_rate": 6.037223415853053e-05, + "loss": 0.0438, + "step": 26866 + }, + { + "epoch": 0.6298047902858114, + "grad_norm": 1.3076132535934448, + "learning_rate": 6.036547117026006e-05, + "loss": 0.1453, + "step": 26867 + }, + { + "epoch": 0.6298282318606164, + "grad_norm": 3.257455587387085, + "learning_rate": 6.0358708397047694e-05, + "loss": 0.1854, + "step": 26868 + }, + { + "epoch": 0.6298516734354214, + "grad_norm": 1.1463459730148315, + "learning_rate": 6.035194583893007e-05, + "loss": 0.0447, + "step": 26869 + }, + { + "epoch": 0.6298751150102264, + "grad_norm": 0.6409338712692261, + "learning_rate": 6.034518349594389e-05, + "loss": 0.0481, + "step": 26870 + }, + { + "epoch": 0.6298985565850314, + "grad_norm": 0.8638325333595276, + "learning_rate": 6.033842136812589e-05, + "loss": 0.0496, + "step": 26871 + }, + { + "epoch": 0.6299219981598364, + "grad_norm": 3.229512929916382, + "learning_rate": 6.0331659455512715e-05, + "loss": 0.3159, + "step": 26872 + }, + { + "epoch": 0.6299454397346413, + "grad_norm": 3.787012815475464, + "learning_rate": 6.032489775814104e-05, + "loss": 0.2552, + "step": 26873 + }, + { + "epoch": 0.6299688813094464, + "grad_norm": 1.4420565366744995, + "learning_rate": 6.031813627604761e-05, + "loss": 0.9942, + "step": 26874 + }, + { + "epoch": 0.6299923228842513, + "grad_norm": 3.2678377628326416, + "learning_rate": 6.0311375009269086e-05, + "loss": 0.2328, + "step": 26875 + }, + { + "epoch": 0.6300157644590564, + "grad_norm": 1.2586976289749146, + "learning_rate": 6.0304613957842184e-05, + "loss": 0.1004, + "step": 26876 + }, + { + "epoch": 0.6300392060338613, + "grad_norm": 3.100179672241211, + "learning_rate": 6.029785312180353e-05, + "loss": 0.3086, + "step": 26877 + }, + { + "epoch": 0.6300626476086664, + "grad_norm": 1.2129286527633667, + "learning_rate": 6.029109250118986e-05, + "loss": 0.1373, + "step": 26878 + }, + { + "epoch": 0.6300860891834713, + "grad_norm": 1.972857117652893, + "learning_rate": 6.0284332096037845e-05, + "loss": 0.2763, + "step": 26879 + }, + { + "epoch": 0.6301095307582764, + "grad_norm": 2.946727752685547, + "learning_rate": 6.0277571906384134e-05, + "loss": 0.1298, + "step": 26880 + }, + { + "epoch": 0.6301329723330813, + "grad_norm": 2.308668375015259, + "learning_rate": 6.027081193226546e-05, + "loss": 0.0908, + "step": 26881 + }, + { + "epoch": 0.6301564139078863, + "grad_norm": 2.8567326068878174, + "learning_rate": 6.0264052173718475e-05, + "loss": 0.2957, + "step": 26882 + }, + { + "epoch": 0.6301798554826913, + "grad_norm": 2.49977707862854, + "learning_rate": 6.025729263077984e-05, + "loss": 1.0398, + "step": 26883 + }, + { + "epoch": 0.6302032970574963, + "grad_norm": 4.408058166503906, + "learning_rate": 6.025053330348628e-05, + "loss": 0.3671, + "step": 26884 + }, + { + "epoch": 0.6302267386323013, + "grad_norm": 1.7313685417175293, + "learning_rate": 6.024377419187439e-05, + "loss": 0.0605, + "step": 26885 + }, + { + "epoch": 0.6302501802071063, + "grad_norm": 1.340896487236023, + "learning_rate": 6.0237015295980946e-05, + "loss": 0.2044, + "step": 26886 + }, + { + "epoch": 0.6302736217819113, + "grad_norm": 4.967189788818359, + "learning_rate": 6.023025661584256e-05, + "loss": 0.436, + "step": 26887 + }, + { + "epoch": 0.6302970633567163, + "grad_norm": 1.3426876068115234, + "learning_rate": 6.022349815149593e-05, + "loss": 0.0844, + "step": 26888 + }, + { + "epoch": 0.6303205049315213, + "grad_norm": 2.7955574989318848, + "learning_rate": 6.021673990297773e-05, + "loss": 0.1924, + "step": 26889 + }, + { + "epoch": 0.6303439465063263, + "grad_norm": 3.0201351642608643, + "learning_rate": 6.020998187032458e-05, + "loss": 0.2851, + "step": 26890 + }, + { + "epoch": 0.6303673880811312, + "grad_norm": 1.0973167419433594, + "learning_rate": 6.020322405357323e-05, + "loss": 0.1155, + "step": 26891 + }, + { + "epoch": 0.6303908296559363, + "grad_norm": 3.9413185119628906, + "learning_rate": 6.019646645276027e-05, + "loss": 0.3264, + "step": 26892 + }, + { + "epoch": 0.6304142712307412, + "grad_norm": 4.625504016876221, + "learning_rate": 6.018970906792243e-05, + "loss": 0.7265, + "step": 26893 + }, + { + "epoch": 0.6304377128055463, + "grad_norm": 1.256897211074829, + "learning_rate": 6.018295189909634e-05, + "loss": 0.0688, + "step": 26894 + }, + { + "epoch": 0.6304611543803512, + "grad_norm": 1.95228111743927, + "learning_rate": 6.017619494631865e-05, + "loss": 0.0718, + "step": 26895 + }, + { + "epoch": 0.6304845959551563, + "grad_norm": 2.6720693111419678, + "learning_rate": 6.016943820962608e-05, + "loss": 0.2474, + "step": 26896 + }, + { + "epoch": 0.6305080375299613, + "grad_norm": 3.355323314666748, + "learning_rate": 6.016268168905521e-05, + "loss": 0.2469, + "step": 26897 + }, + { + "epoch": 0.6305314791047663, + "grad_norm": 2.9826018810272217, + "learning_rate": 6.0155925384642784e-05, + "loss": 0.2171, + "step": 26898 + }, + { + "epoch": 0.6305549206795713, + "grad_norm": 2.431061267852783, + "learning_rate": 6.014916929642541e-05, + "loss": 0.1968, + "step": 26899 + }, + { + "epoch": 0.6305783622543762, + "grad_norm": 0.655881941318512, + "learning_rate": 6.014241342443979e-05, + "loss": 0.0563, + "step": 26900 + }, + { + "epoch": 0.6306018038291813, + "grad_norm": 1.07097589969635, + "learning_rate": 6.013565776872255e-05, + "loss": 0.1307, + "step": 26901 + }, + { + "epoch": 0.6306252454039862, + "grad_norm": 2.5480566024780273, + "learning_rate": 6.012890232931034e-05, + "loss": 0.156, + "step": 26902 + }, + { + "epoch": 0.6306486869787913, + "grad_norm": 1.2477216720581055, + "learning_rate": 6.012214710623983e-05, + "loss": 0.1333, + "step": 26903 + }, + { + "epoch": 0.6306721285535962, + "grad_norm": 2.5485618114471436, + "learning_rate": 6.0115392099547685e-05, + "loss": 0.1811, + "step": 26904 + }, + { + "epoch": 0.6306955701284013, + "grad_norm": 2.2010488510131836, + "learning_rate": 6.010863730927052e-05, + "loss": 0.1599, + "step": 26905 + }, + { + "epoch": 0.6307190117032062, + "grad_norm": 3.3105525970458984, + "learning_rate": 6.0101882735445036e-05, + "loss": 0.2425, + "step": 26906 + }, + { + "epoch": 0.6307424532780113, + "grad_norm": 0.905205249786377, + "learning_rate": 6.009512837810785e-05, + "loss": 0.0548, + "step": 26907 + }, + { + "epoch": 0.6307658948528162, + "grad_norm": 2.8673272132873535, + "learning_rate": 6.008837423729561e-05, + "loss": 0.2106, + "step": 26908 + }, + { + "epoch": 0.6307893364276213, + "grad_norm": 3.434731960296631, + "learning_rate": 6.008162031304494e-05, + "loss": 0.3669, + "step": 26909 + }, + { + "epoch": 0.6308127780024262, + "grad_norm": 3.489555835723877, + "learning_rate": 6.0074866605392564e-05, + "loss": 0.2238, + "step": 26910 + }, + { + "epoch": 0.6308362195772312, + "grad_norm": 2.6655755043029785, + "learning_rate": 6.006811311437509e-05, + "loss": 0.2753, + "step": 26911 + }, + { + "epoch": 0.6308596611520362, + "grad_norm": 3.641963005065918, + "learning_rate": 6.006135984002914e-05, + "loss": 0.401, + "step": 26912 + }, + { + "epoch": 0.6308831027268412, + "grad_norm": 3.6038167476654053, + "learning_rate": 6.0054606782391386e-05, + "loss": 0.2745, + "step": 26913 + }, + { + "epoch": 0.6309065443016462, + "grad_norm": 3.056246280670166, + "learning_rate": 6.0047853941498456e-05, + "loss": 0.1832, + "step": 26914 + }, + { + "epoch": 0.6309299858764512, + "grad_norm": 4.351804256439209, + "learning_rate": 6.0041101317386985e-05, + "loss": 0.2634, + "step": 26915 + }, + { + "epoch": 0.6309534274512562, + "grad_norm": 0.39956486225128174, + "learning_rate": 6.003434891009364e-05, + "loss": 0.0337, + "step": 26916 + }, + { + "epoch": 0.6309768690260612, + "grad_norm": 2.541334390640259, + "learning_rate": 6.002759671965503e-05, + "loss": 0.1361, + "step": 26917 + }, + { + "epoch": 0.6310003106008661, + "grad_norm": 1.7096996307373047, + "learning_rate": 6.00208447461078e-05, + "loss": 1.2842, + "step": 26918 + }, + { + "epoch": 0.6310237521756712, + "grad_norm": 1.2846108675003052, + "learning_rate": 6.00140929894886e-05, + "loss": 0.1726, + "step": 26919 + }, + { + "epoch": 0.6310471937504761, + "grad_norm": 0.944460391998291, + "learning_rate": 6.000734144983407e-05, + "loss": 0.0693, + "step": 26920 + }, + { + "epoch": 0.6310706353252812, + "grad_norm": 3.2197368144989014, + "learning_rate": 6.0000590127180766e-05, + "loss": 0.2585, + "step": 26921 + }, + { + "epoch": 0.6310940769000861, + "grad_norm": 1.838154911994934, + "learning_rate": 5.999383902156541e-05, + "loss": 0.138, + "step": 26922 + }, + { + "epoch": 0.6311175184748912, + "grad_norm": 1.0370835065841675, + "learning_rate": 5.998708813302465e-05, + "loss": 0.1044, + "step": 26923 + }, + { + "epoch": 0.6311409600496961, + "grad_norm": 0.8043288588523865, + "learning_rate": 5.998033746159506e-05, + "loss": 0.0666, + "step": 26924 + }, + { + "epoch": 0.6311644016245012, + "grad_norm": 2.5793774127960205, + "learning_rate": 5.997358700731326e-05, + "loss": 0.1295, + "step": 26925 + }, + { + "epoch": 0.6311878431993061, + "grad_norm": 2.589010238647461, + "learning_rate": 5.996683677021593e-05, + "loss": 0.1738, + "step": 26926 + }, + { + "epoch": 0.6312112847741111, + "grad_norm": 1.9249721765518188, + "learning_rate": 5.996008675033966e-05, + "loss": 0.1764, + "step": 26927 + }, + { + "epoch": 0.6312347263489161, + "grad_norm": 2.8094305992126465, + "learning_rate": 5.9953336947721074e-05, + "loss": 0.317, + "step": 26928 + }, + { + "epoch": 0.6312581679237211, + "grad_norm": 1.6779948472976685, + "learning_rate": 5.994658736239682e-05, + "loss": 0.1182, + "step": 26929 + }, + { + "epoch": 0.6312816094985261, + "grad_norm": 0.45743444561958313, + "learning_rate": 5.99398379944035e-05, + "loss": 0.0237, + "step": 26930 + }, + { + "epoch": 0.6313050510733311, + "grad_norm": 0.6542527079582214, + "learning_rate": 5.993308884377775e-05, + "loss": 0.0636, + "step": 26931 + }, + { + "epoch": 0.6313284926481361, + "grad_norm": 0.21630258858203888, + "learning_rate": 5.9926339910556206e-05, + "loss": 0.0176, + "step": 26932 + }, + { + "epoch": 0.6313519342229411, + "grad_norm": 0.8525700569152832, + "learning_rate": 5.991959119477544e-05, + "loss": 0.0434, + "step": 26933 + }, + { + "epoch": 0.631375375797746, + "grad_norm": 2.926835060119629, + "learning_rate": 5.991284269647207e-05, + "loss": 0.3302, + "step": 26934 + }, + { + "epoch": 0.6313988173725511, + "grad_norm": 2.042738437652588, + "learning_rate": 5.990609441568279e-05, + "loss": 0.2187, + "step": 26935 + }, + { + "epoch": 0.631422258947356, + "grad_norm": 4.207162857055664, + "learning_rate": 5.9899346352444165e-05, + "loss": 0.4988, + "step": 26936 + }, + { + "epoch": 0.6314457005221611, + "grad_norm": 4.969411373138428, + "learning_rate": 5.989259850679281e-05, + "loss": 0.4765, + "step": 26937 + }, + { + "epoch": 0.631469142096966, + "grad_norm": 2.6610524654388428, + "learning_rate": 5.988585087876535e-05, + "loss": 0.3384, + "step": 26938 + }, + { + "epoch": 0.6314925836717711, + "grad_norm": 0.7592484951019287, + "learning_rate": 5.987910346839839e-05, + "loss": 0.0345, + "step": 26939 + }, + { + "epoch": 0.631516025246576, + "grad_norm": 4.111253261566162, + "learning_rate": 5.987235627572852e-05, + "loss": 0.2987, + "step": 26940 + }, + { + "epoch": 0.6315394668213811, + "grad_norm": 3.304123878479004, + "learning_rate": 5.98656093007924e-05, + "loss": 0.3272, + "step": 26941 + }, + { + "epoch": 0.631562908396186, + "grad_norm": 3.628171682357788, + "learning_rate": 5.985886254362661e-05, + "loss": 0.2769, + "step": 26942 + }, + { + "epoch": 0.6315863499709911, + "grad_norm": 1.7780104875564575, + "learning_rate": 5.9852116004267745e-05, + "loss": 0.1391, + "step": 26943 + }, + { + "epoch": 0.631609791545796, + "grad_norm": 1.0661535263061523, + "learning_rate": 5.984536968275245e-05, + "loss": 0.0507, + "step": 26944 + }, + { + "epoch": 0.631633233120601, + "grad_norm": 2.2991862297058105, + "learning_rate": 5.983862357911729e-05, + "loss": 0.1447, + "step": 26945 + }, + { + "epoch": 0.631656674695406, + "grad_norm": 1.8575233221054077, + "learning_rate": 5.9831877693398866e-05, + "loss": 0.1989, + "step": 26946 + }, + { + "epoch": 0.631680116270211, + "grad_norm": 3.7150449752807617, + "learning_rate": 5.982513202563381e-05, + "loss": 0.47, + "step": 26947 + }, + { + "epoch": 0.631703557845016, + "grad_norm": 1.1688445806503296, + "learning_rate": 5.981838657585874e-05, + "loss": 0.0816, + "step": 26948 + }, + { + "epoch": 0.631726999419821, + "grad_norm": 2.797071695327759, + "learning_rate": 5.9811641344110246e-05, + "loss": 0.1349, + "step": 26949 + }, + { + "epoch": 0.6317504409946261, + "grad_norm": 0.4203127920627594, + "learning_rate": 5.9804896330424886e-05, + "loss": 0.0174, + "step": 26950 + }, + { + "epoch": 0.631773882569431, + "grad_norm": 2.433985471725464, + "learning_rate": 5.979815153483932e-05, + "loss": 0.0992, + "step": 26951 + }, + { + "epoch": 0.6317973241442361, + "grad_norm": 1.075085163116455, + "learning_rate": 5.979140695739009e-05, + "loss": 0.0885, + "step": 26952 + }, + { + "epoch": 0.631820765719041, + "grad_norm": 0.4588584303855896, + "learning_rate": 5.9784662598113816e-05, + "loss": 0.0205, + "step": 26953 + }, + { + "epoch": 0.631844207293846, + "grad_norm": 0.8385533690452576, + "learning_rate": 5.9777918457047125e-05, + "loss": 0.0502, + "step": 26954 + }, + { + "epoch": 0.631867648868651, + "grad_norm": 2.994208574295044, + "learning_rate": 5.9771174534226556e-05, + "loss": 0.2992, + "step": 26955 + }, + { + "epoch": 0.631891090443456, + "grad_norm": 2.285747766494751, + "learning_rate": 5.976443082968871e-05, + "loss": 0.1811, + "step": 26956 + }, + { + "epoch": 0.631914532018261, + "grad_norm": 4.05813455581665, + "learning_rate": 5.975768734347022e-05, + "loss": 0.4461, + "step": 26957 + }, + { + "epoch": 0.631937973593066, + "grad_norm": 2.4789111614227295, + "learning_rate": 5.9750944075607595e-05, + "loss": 0.1647, + "step": 26958 + }, + { + "epoch": 0.631961415167871, + "grad_norm": 1.4007972478866577, + "learning_rate": 5.974420102613754e-05, + "loss": 0.117, + "step": 26959 + }, + { + "epoch": 0.631984856742676, + "grad_norm": 2.7041823863983154, + "learning_rate": 5.973745819509655e-05, + "loss": 0.2075, + "step": 26960 + }, + { + "epoch": 0.632008298317481, + "grad_norm": 3.541828155517578, + "learning_rate": 5.9730715582521256e-05, + "loss": 0.3053, + "step": 26961 + }, + { + "epoch": 0.632031739892286, + "grad_norm": 2.710723400115967, + "learning_rate": 5.972397318844825e-05, + "loss": 0.234, + "step": 26962 + }, + { + "epoch": 0.6320551814670909, + "grad_norm": 1.292155385017395, + "learning_rate": 5.9717231012914064e-05, + "loss": 0.1409, + "step": 26963 + }, + { + "epoch": 0.632078623041896, + "grad_norm": 4.8899102210998535, + "learning_rate": 5.9710489055955356e-05, + "loss": 0.2828, + "step": 26964 + }, + { + "epoch": 0.6321020646167009, + "grad_norm": 6.413466453552246, + "learning_rate": 5.9703747317608635e-05, + "loss": 0.52, + "step": 26965 + }, + { + "epoch": 0.632125506191506, + "grad_norm": 4.631537914276123, + "learning_rate": 5.969700579791052e-05, + "loss": 0.3846, + "step": 26966 + }, + { + "epoch": 0.6321489477663109, + "grad_norm": 3.1502881050109863, + "learning_rate": 5.9690264496897596e-05, + "loss": 0.2638, + "step": 26967 + }, + { + "epoch": 0.632172389341116, + "grad_norm": 1.7834203243255615, + "learning_rate": 5.9683523414606414e-05, + "loss": 0.2551, + "step": 26968 + }, + { + "epoch": 0.6321958309159209, + "grad_norm": 2.3975250720977783, + "learning_rate": 5.9676782551073586e-05, + "loss": 1.8494, + "step": 26969 + }, + { + "epoch": 0.632219272490726, + "grad_norm": 1.1980832815170288, + "learning_rate": 5.967004190633563e-05, + "loss": 0.1541, + "step": 26970 + }, + { + "epoch": 0.6322427140655309, + "grad_norm": 2.5844337940216064, + "learning_rate": 5.966330148042919e-05, + "loss": 0.1733, + "step": 26971 + }, + { + "epoch": 0.632266155640336, + "grad_norm": 2.7335007190704346, + "learning_rate": 5.96565612733908e-05, + "loss": 0.2301, + "step": 26972 + }, + { + "epoch": 0.6322895972151409, + "grad_norm": 1.0386487245559692, + "learning_rate": 5.964982128525707e-05, + "loss": 0.097, + "step": 26973 + }, + { + "epoch": 0.6323130387899459, + "grad_norm": 0.8998867273330688, + "learning_rate": 5.9643081516064526e-05, + "loss": 0.0795, + "step": 26974 + }, + { + "epoch": 0.6323364803647509, + "grad_norm": 3.0935120582580566, + "learning_rate": 5.963634196584975e-05, + "loss": 0.2522, + "step": 26975 + }, + { + "epoch": 0.6323599219395559, + "grad_norm": 0.6483485698699951, + "learning_rate": 5.962960263464933e-05, + "loss": 0.0553, + "step": 26976 + }, + { + "epoch": 0.6323833635143609, + "grad_norm": 4.659386157989502, + "learning_rate": 5.9622863522499816e-05, + "loss": 0.3483, + "step": 26977 + }, + { + "epoch": 0.6324068050891659, + "grad_norm": 4.249722957611084, + "learning_rate": 5.961612462943778e-05, + "loss": 0.3777, + "step": 26978 + }, + { + "epoch": 0.6324302466639709, + "grad_norm": 3.3195199966430664, + "learning_rate": 5.960938595549978e-05, + "loss": 0.3554, + "step": 26979 + }, + { + "epoch": 0.6324536882387759, + "grad_norm": 0.9798465967178345, + "learning_rate": 5.960264750072242e-05, + "loss": 0.1434, + "step": 26980 + }, + { + "epoch": 0.6324771298135808, + "grad_norm": 2.5058577060699463, + "learning_rate": 5.959590926514219e-05, + "loss": 0.1347, + "step": 26981 + }, + { + "epoch": 0.6325005713883859, + "grad_norm": 2.2109386920928955, + "learning_rate": 5.9589171248795683e-05, + "loss": 0.2589, + "step": 26982 + }, + { + "epoch": 0.6325240129631908, + "grad_norm": 4.180381774902344, + "learning_rate": 5.95824334517195e-05, + "loss": 0.1615, + "step": 26983 + }, + { + "epoch": 0.6325474545379959, + "grad_norm": 1.0467169284820557, + "learning_rate": 5.957569587395016e-05, + "loss": 0.0496, + "step": 26984 + }, + { + "epoch": 0.6325708961128008, + "grad_norm": 0.8373894095420837, + "learning_rate": 5.956895851552423e-05, + "loss": 0.0889, + "step": 26985 + }, + { + "epoch": 0.6325943376876059, + "grad_norm": 2.322798252105713, + "learning_rate": 5.9562221376478275e-05, + "loss": 0.2258, + "step": 26986 + }, + { + "epoch": 0.6326177792624108, + "grad_norm": 3.853403091430664, + "learning_rate": 5.955548445684885e-05, + "loss": 0.2524, + "step": 26987 + }, + { + "epoch": 0.6326412208372159, + "grad_norm": 1.0270103216171265, + "learning_rate": 5.9548747756672485e-05, + "loss": 0.0976, + "step": 26988 + }, + { + "epoch": 0.6326646624120208, + "grad_norm": 1.8942497968673706, + "learning_rate": 5.9542011275985774e-05, + "loss": 0.193, + "step": 26989 + }, + { + "epoch": 0.6326881039868258, + "grad_norm": 3.4434711933135986, + "learning_rate": 5.953527501482524e-05, + "loss": 0.2892, + "step": 26990 + }, + { + "epoch": 0.6327115455616308, + "grad_norm": 1.2545804977416992, + "learning_rate": 5.9528538973227434e-05, + "loss": 0.0895, + "step": 26991 + }, + { + "epoch": 0.6327349871364358, + "grad_norm": 0.8958832621574402, + "learning_rate": 5.9521803151228915e-05, + "loss": 0.1061, + "step": 26992 + }, + { + "epoch": 0.6327584287112408, + "grad_norm": 0.613788902759552, + "learning_rate": 5.951506754886624e-05, + "loss": 0.0562, + "step": 26993 + }, + { + "epoch": 0.6327818702860458, + "grad_norm": 2.437326192855835, + "learning_rate": 5.9508332166175906e-05, + "loss": 1.5808, + "step": 26994 + }, + { + "epoch": 0.6328053118608508, + "grad_norm": 2.1649794578552246, + "learning_rate": 5.950159700319452e-05, + "loss": 0.1054, + "step": 26995 + }, + { + "epoch": 0.6328287534356558, + "grad_norm": 0.9015418291091919, + "learning_rate": 5.9494862059958624e-05, + "loss": 0.109, + "step": 26996 + }, + { + "epoch": 0.6328521950104607, + "grad_norm": 1.7845752239227295, + "learning_rate": 5.948812733650475e-05, + "loss": 0.1466, + "step": 26997 + }, + { + "epoch": 0.6328756365852658, + "grad_norm": 0.9562547206878662, + "learning_rate": 5.948139283286942e-05, + "loss": 0.1043, + "step": 26998 + }, + { + "epoch": 0.6328990781600707, + "grad_norm": 4.082275867462158, + "learning_rate": 5.94746585490892e-05, + "loss": 0.4252, + "step": 26999 + }, + { + "epoch": 0.6329225197348758, + "grad_norm": 3.8595054149627686, + "learning_rate": 5.9467924485200644e-05, + "loss": 0.3466, + "step": 27000 + }, + { + "epoch": 0.6329459613096808, + "grad_norm": 3.7393951416015625, + "learning_rate": 5.9461190641240226e-05, + "loss": 0.3015, + "step": 27001 + }, + { + "epoch": 0.6329694028844858, + "grad_norm": 6.628475189208984, + "learning_rate": 5.945445701724456e-05, + "loss": 0.2283, + "step": 27002 + }, + { + "epoch": 0.6329928444592908, + "grad_norm": 2.7757956981658936, + "learning_rate": 5.9447723613250125e-05, + "loss": 0.2092, + "step": 27003 + }, + { + "epoch": 0.6330162860340958, + "grad_norm": 3.859722852706909, + "learning_rate": 5.94409904292935e-05, + "loss": 0.5182, + "step": 27004 + }, + { + "epoch": 0.6330397276089008, + "grad_norm": 2.4593005180358887, + "learning_rate": 5.94342574654112e-05, + "loss": 0.1449, + "step": 27005 + }, + { + "epoch": 0.6330631691837058, + "grad_norm": 4.459479331970215, + "learning_rate": 5.942752472163975e-05, + "loss": 0.3425, + "step": 27006 + }, + { + "epoch": 0.6330866107585108, + "grad_norm": 4.6780314445495605, + "learning_rate": 5.9420792198015686e-05, + "loss": 0.5, + "step": 27007 + }, + { + "epoch": 0.6331100523333157, + "grad_norm": 3.505311965942383, + "learning_rate": 5.941405989457555e-05, + "loss": 0.3879, + "step": 27008 + }, + { + "epoch": 0.6331334939081208, + "grad_norm": 4.2625579833984375, + "learning_rate": 5.9407327811355876e-05, + "loss": 1.5032, + "step": 27009 + }, + { + "epoch": 0.6331569354829257, + "grad_norm": 0.7831044793128967, + "learning_rate": 5.940059594839317e-05, + "loss": 0.0295, + "step": 27010 + }, + { + "epoch": 0.6331803770577308, + "grad_norm": 0.563803493976593, + "learning_rate": 5.939386430572399e-05, + "loss": 0.0202, + "step": 27011 + }, + { + "epoch": 0.6332038186325357, + "grad_norm": 0.790781557559967, + "learning_rate": 5.9387132883384844e-05, + "loss": 0.0385, + "step": 27012 + }, + { + "epoch": 0.6332272602073408, + "grad_norm": 0.6251846551895142, + "learning_rate": 5.938040168141224e-05, + "loss": 0.068, + "step": 27013 + }, + { + "epoch": 0.6332507017821457, + "grad_norm": 0.8912739157676697, + "learning_rate": 5.937367069984273e-05, + "loss": 0.077, + "step": 27014 + }, + { + "epoch": 0.6332741433569508, + "grad_norm": 1.5650222301483154, + "learning_rate": 5.936693993871283e-05, + "loss": 1.0353, + "step": 27015 + }, + { + "epoch": 0.6332975849317557, + "grad_norm": 4.386260986328125, + "learning_rate": 5.936020939805904e-05, + "loss": 0.3539, + "step": 27016 + }, + { + "epoch": 0.6333210265065607, + "grad_norm": 3.892838478088379, + "learning_rate": 5.9353479077917906e-05, + "loss": 0.6428, + "step": 27017 + }, + { + "epoch": 0.6333444680813657, + "grad_norm": 2.4841549396514893, + "learning_rate": 5.934674897832595e-05, + "loss": 0.872, + "step": 27018 + }, + { + "epoch": 0.6333679096561707, + "grad_norm": 1.4941961765289307, + "learning_rate": 5.934001909931963e-05, + "loss": 0.0564, + "step": 27019 + }, + { + "epoch": 0.6333913512309757, + "grad_norm": 2.2744035720825195, + "learning_rate": 5.933328944093552e-05, + "loss": 0.1339, + "step": 27020 + }, + { + "epoch": 0.6334147928057807, + "grad_norm": 2.017359495162964, + "learning_rate": 5.9326560003210166e-05, + "loss": 0.1139, + "step": 27021 + }, + { + "epoch": 0.6334382343805857, + "grad_norm": 1.6343600749969482, + "learning_rate": 5.931983078618003e-05, + "loss": 0.1091, + "step": 27022 + }, + { + "epoch": 0.6334616759553907, + "grad_norm": 4.561960697174072, + "learning_rate": 5.931310178988162e-05, + "loss": 0.2196, + "step": 27023 + }, + { + "epoch": 0.6334851175301957, + "grad_norm": 2.5480306148529053, + "learning_rate": 5.930637301435148e-05, + "loss": 0.3287, + "step": 27024 + }, + { + "epoch": 0.6335085591050007, + "grad_norm": 0.3661433160305023, + "learning_rate": 5.92996444596261e-05, + "loss": 0.0264, + "step": 27025 + }, + { + "epoch": 0.6335320006798056, + "grad_norm": 3.2469120025634766, + "learning_rate": 5.929291612574198e-05, + "loss": 0.304, + "step": 27026 + }, + { + "epoch": 0.6335554422546107, + "grad_norm": 3.486621141433716, + "learning_rate": 5.9286188012735664e-05, + "loss": 0.3791, + "step": 27027 + }, + { + "epoch": 0.6335788838294156, + "grad_norm": 1.929734468460083, + "learning_rate": 5.9279460120643625e-05, + "loss": 0.1292, + "step": 27028 + }, + { + "epoch": 0.6336023254042207, + "grad_norm": 2.3285984992980957, + "learning_rate": 5.9272732449502374e-05, + "loss": 0.1477, + "step": 27029 + }, + { + "epoch": 0.6336257669790256, + "grad_norm": 2.6032145023345947, + "learning_rate": 5.9266004999348434e-05, + "loss": 0.139, + "step": 27030 + }, + { + "epoch": 0.6336492085538307, + "grad_norm": 4.926976203918457, + "learning_rate": 5.925927777021826e-05, + "loss": 0.3195, + "step": 27031 + }, + { + "epoch": 0.6336726501286356, + "grad_norm": 7.188705921173096, + "learning_rate": 5.925255076214843e-05, + "loss": 1.0458, + "step": 27032 + }, + { + "epoch": 0.6336960917034407, + "grad_norm": 3.185661792755127, + "learning_rate": 5.9245823975175375e-05, + "loss": 0.2619, + "step": 27033 + }, + { + "epoch": 0.6337195332782456, + "grad_norm": 1.7792518138885498, + "learning_rate": 5.923909740933566e-05, + "loss": 0.167, + "step": 27034 + }, + { + "epoch": 0.6337429748530506, + "grad_norm": 0.8899329900741577, + "learning_rate": 5.923237106466575e-05, + "loss": 0.1105, + "step": 27035 + }, + { + "epoch": 0.6337664164278556, + "grad_norm": 5.432460784912109, + "learning_rate": 5.9225644941202104e-05, + "loss": 0.422, + "step": 27036 + }, + { + "epoch": 0.6337898580026606, + "grad_norm": 7.432644844055176, + "learning_rate": 5.921891903898129e-05, + "loss": 0.2663, + "step": 27037 + }, + { + "epoch": 0.6338132995774656, + "grad_norm": 0.4004817605018616, + "learning_rate": 5.921219335803976e-05, + "loss": 0.0308, + "step": 27038 + }, + { + "epoch": 0.6338367411522706, + "grad_norm": 2.620934247970581, + "learning_rate": 5.920546789841399e-05, + "loss": 0.163, + "step": 27039 + }, + { + "epoch": 0.6338601827270756, + "grad_norm": 3.187995195388794, + "learning_rate": 5.9198742660140536e-05, + "loss": 0.1597, + "step": 27040 + }, + { + "epoch": 0.6338836243018806, + "grad_norm": 2.411586046218872, + "learning_rate": 5.919201764325581e-05, + "loss": 1.4677, + "step": 27041 + }, + { + "epoch": 0.6339070658766855, + "grad_norm": 5.043181419372559, + "learning_rate": 5.918529284779639e-05, + "loss": 0.3021, + "step": 27042 + }, + { + "epoch": 0.6339305074514906, + "grad_norm": 4.153552055358887, + "learning_rate": 5.9178568273798666e-05, + "loss": 0.2528, + "step": 27043 + }, + { + "epoch": 0.6339539490262955, + "grad_norm": 4.040579795837402, + "learning_rate": 5.917184392129921e-05, + "loss": 0.2015, + "step": 27044 + }, + { + "epoch": 0.6339773906011006, + "grad_norm": 3.5629231929779053, + "learning_rate": 5.916511979033448e-05, + "loss": 0.2664, + "step": 27045 + }, + { + "epoch": 0.6340008321759055, + "grad_norm": 0.5420300960540771, + "learning_rate": 5.915839588094094e-05, + "loss": 0.035, + "step": 27046 + }, + { + "epoch": 0.6340242737507106, + "grad_norm": 1.5624873638153076, + "learning_rate": 5.915167219315512e-05, + "loss": 0.1837, + "step": 27047 + }, + { + "epoch": 0.6340477153255155, + "grad_norm": 3.1726725101470947, + "learning_rate": 5.914494872701345e-05, + "loss": 0.1837, + "step": 27048 + }, + { + "epoch": 0.6340711569003206, + "grad_norm": 2.2868340015411377, + "learning_rate": 5.913822548255246e-05, + "loss": 0.3532, + "step": 27049 + }, + { + "epoch": 0.6340945984751255, + "grad_norm": 2.59692120552063, + "learning_rate": 5.9131502459808596e-05, + "loss": 0.0771, + "step": 27050 + }, + { + "epoch": 0.6341180400499306, + "grad_norm": 4.550843715667725, + "learning_rate": 5.912477965881833e-05, + "loss": 0.2901, + "step": 27051 + }, + { + "epoch": 0.6341414816247356, + "grad_norm": 2.4699604511260986, + "learning_rate": 5.911805707961819e-05, + "loss": 0.1818, + "step": 27052 + }, + { + "epoch": 0.6341649231995405, + "grad_norm": 2.1653060913085938, + "learning_rate": 5.9111334722244615e-05, + "loss": 0.2332, + "step": 27053 + }, + { + "epoch": 0.6341883647743456, + "grad_norm": 1.9375956058502197, + "learning_rate": 5.9104612586734074e-05, + "loss": 0.1592, + "step": 27054 + }, + { + "epoch": 0.6342118063491505, + "grad_norm": 2.5687179565429688, + "learning_rate": 5.909789067312304e-05, + "loss": 0.2301, + "step": 27055 + }, + { + "epoch": 0.6342352479239556, + "grad_norm": 0.4768073558807373, + "learning_rate": 5.9091168981448016e-05, + "loss": 0.0301, + "step": 27056 + }, + { + "epoch": 0.6342586894987605, + "grad_norm": 2.920409679412842, + "learning_rate": 5.908444751174548e-05, + "loss": 0.2332, + "step": 27057 + }, + { + "epoch": 0.6342821310735656, + "grad_norm": 0.6164277195930481, + "learning_rate": 5.907772626405186e-05, + "loss": 0.0594, + "step": 27058 + }, + { + "epoch": 0.6343055726483705, + "grad_norm": 2.4389233589172363, + "learning_rate": 5.907100523840366e-05, + "loss": 0.1133, + "step": 27059 + }, + { + "epoch": 0.6343290142231756, + "grad_norm": 4.473981857299805, + "learning_rate": 5.906428443483734e-05, + "loss": 0.2909, + "step": 27060 + }, + { + "epoch": 0.6343524557979805, + "grad_norm": 1.270998477935791, + "learning_rate": 5.905756385338934e-05, + "loss": 0.8345, + "step": 27061 + }, + { + "epoch": 0.6343758973727855, + "grad_norm": 7.78815221786499, + "learning_rate": 5.905084349409618e-05, + "loss": 0.0924, + "step": 27062 + }, + { + "epoch": 0.6343993389475905, + "grad_norm": 4.193188667297363, + "learning_rate": 5.9044123356994296e-05, + "loss": 0.4633, + "step": 27063 + }, + { + "epoch": 0.6344227805223955, + "grad_norm": 0.8186548948287964, + "learning_rate": 5.903740344212012e-05, + "loss": 0.0623, + "step": 27064 + }, + { + "epoch": 0.6344462220972005, + "grad_norm": 1.50901198387146, + "learning_rate": 5.903068374951017e-05, + "loss": 0.0993, + "step": 27065 + }, + { + "epoch": 0.6344696636720055, + "grad_norm": 1.779273509979248, + "learning_rate": 5.902396427920088e-05, + "loss": 0.1007, + "step": 27066 + }, + { + "epoch": 0.6344931052468105, + "grad_norm": 2.282264232635498, + "learning_rate": 5.901724503122867e-05, + "loss": 0.2115, + "step": 27067 + }, + { + "epoch": 0.6345165468216155, + "grad_norm": 5.555544853210449, + "learning_rate": 5.901052600563006e-05, + "loss": 0.2287, + "step": 27068 + }, + { + "epoch": 0.6345399883964205, + "grad_norm": 1.0904165506362915, + "learning_rate": 5.9003807202441516e-05, + "loss": 0.0593, + "step": 27069 + }, + { + "epoch": 0.6345634299712255, + "grad_norm": 3.77132511138916, + "learning_rate": 5.899708862169946e-05, + "loss": 0.2591, + "step": 27070 + }, + { + "epoch": 0.6345868715460304, + "grad_norm": 3.2216951847076416, + "learning_rate": 5.899037026344034e-05, + "loss": 0.2799, + "step": 27071 + }, + { + "epoch": 0.6346103131208355, + "grad_norm": 3.1057586669921875, + "learning_rate": 5.8983652127700636e-05, + "loss": 0.2195, + "step": 27072 + }, + { + "epoch": 0.6346337546956404, + "grad_norm": 3.3567488193511963, + "learning_rate": 5.89769342145168e-05, + "loss": 1.759, + "step": 27073 + }, + { + "epoch": 0.6346571962704455, + "grad_norm": 0.8012617826461792, + "learning_rate": 5.897021652392525e-05, + "loss": 0.0464, + "step": 27074 + }, + { + "epoch": 0.6346806378452504, + "grad_norm": 2.32425856590271, + "learning_rate": 5.896349905596247e-05, + "loss": 0.1364, + "step": 27075 + }, + { + "epoch": 0.6347040794200555, + "grad_norm": 4.056572914123535, + "learning_rate": 5.895678181066491e-05, + "loss": 0.2809, + "step": 27076 + }, + { + "epoch": 0.6347275209948604, + "grad_norm": 1.0104138851165771, + "learning_rate": 5.895006478806897e-05, + "loss": 0.1387, + "step": 27077 + }, + { + "epoch": 0.6347509625696655, + "grad_norm": 3.743157386779785, + "learning_rate": 5.8943347988211165e-05, + "loss": 0.2821, + "step": 27078 + }, + { + "epoch": 0.6347744041444704, + "grad_norm": 4.40519905090332, + "learning_rate": 5.8936631411127864e-05, + "loss": 0.4182, + "step": 27079 + }, + { + "epoch": 0.6347978457192754, + "grad_norm": 4.821981906890869, + "learning_rate": 5.892991505685559e-05, + "loss": 0.4756, + "step": 27080 + }, + { + "epoch": 0.6348212872940804, + "grad_norm": 3.7455196380615234, + "learning_rate": 5.8923198925430746e-05, + "loss": 0.4397, + "step": 27081 + }, + { + "epoch": 0.6348447288688854, + "grad_norm": 0.8850146532058716, + "learning_rate": 5.891648301688979e-05, + "loss": 0.0731, + "step": 27082 + }, + { + "epoch": 0.6348681704436904, + "grad_norm": 4.656692981719971, + "learning_rate": 5.890976733126915e-05, + "loss": 0.3403, + "step": 27083 + }, + { + "epoch": 0.6348916120184954, + "grad_norm": 3.6675844192504883, + "learning_rate": 5.890305186860525e-05, + "loss": 0.2939, + "step": 27084 + }, + { + "epoch": 0.6349150535933004, + "grad_norm": 3.126546621322632, + "learning_rate": 5.8896336628934566e-05, + "loss": 0.214, + "step": 27085 + }, + { + "epoch": 0.6349384951681054, + "grad_norm": 3.007350206375122, + "learning_rate": 5.88896216122935e-05, + "loss": 0.2478, + "step": 27086 + }, + { + "epoch": 0.6349619367429103, + "grad_norm": 2.4899446964263916, + "learning_rate": 5.888290681871852e-05, + "loss": 0.1286, + "step": 27087 + }, + { + "epoch": 0.6349853783177154, + "grad_norm": 2.9913525581359863, + "learning_rate": 5.887619224824604e-05, + "loss": 1.591, + "step": 27088 + }, + { + "epoch": 0.6350088198925203, + "grad_norm": 0.942100465297699, + "learning_rate": 5.886947790091248e-05, + "loss": 0.0978, + "step": 27089 + }, + { + "epoch": 0.6350322614673254, + "grad_norm": 3.079967975616455, + "learning_rate": 5.8862763776754306e-05, + "loss": 0.212, + "step": 27090 + }, + { + "epoch": 0.6350557030421303, + "grad_norm": 2.9289331436157227, + "learning_rate": 5.885604987580794e-05, + "loss": 0.3904, + "step": 27091 + }, + { + "epoch": 0.6350791446169354, + "grad_norm": 2.063751459121704, + "learning_rate": 5.884933619810976e-05, + "loss": 0.2048, + "step": 27092 + }, + { + "epoch": 0.6351025861917403, + "grad_norm": 2.2818403244018555, + "learning_rate": 5.884262274369626e-05, + "loss": 0.1675, + "step": 27093 + }, + { + "epoch": 0.6351260277665454, + "grad_norm": 1.319741129875183, + "learning_rate": 5.883590951260387e-05, + "loss": 0.0932, + "step": 27094 + }, + { + "epoch": 0.6351494693413503, + "grad_norm": 1.0960668325424194, + "learning_rate": 5.882919650486898e-05, + "loss": 0.1114, + "step": 27095 + }, + { + "epoch": 0.6351729109161554, + "grad_norm": 0.7313358783721924, + "learning_rate": 5.882248372052802e-05, + "loss": 0.0498, + "step": 27096 + }, + { + "epoch": 0.6351963524909603, + "grad_norm": 5.021644592285156, + "learning_rate": 5.881577115961744e-05, + "loss": 0.361, + "step": 27097 + }, + { + "epoch": 0.6352197940657653, + "grad_norm": 2.2405567169189453, + "learning_rate": 5.880905882217363e-05, + "loss": 0.1873, + "step": 27098 + }, + { + "epoch": 0.6352432356405703, + "grad_norm": 2.247668743133545, + "learning_rate": 5.880234670823302e-05, + "loss": 0.1372, + "step": 27099 + }, + { + "epoch": 0.6352666772153753, + "grad_norm": 1.2093981504440308, + "learning_rate": 5.879563481783206e-05, + "loss": 0.1203, + "step": 27100 + }, + { + "epoch": 0.6352901187901803, + "grad_norm": 0.677513599395752, + "learning_rate": 5.878892315100713e-05, + "loss": 0.0567, + "step": 27101 + }, + { + "epoch": 0.6353135603649853, + "grad_norm": 4.290207386016846, + "learning_rate": 5.878221170779464e-05, + "loss": 0.5905, + "step": 27102 + }, + { + "epoch": 0.6353370019397904, + "grad_norm": 0.913069486618042, + "learning_rate": 5.877550048823106e-05, + "loss": 0.0325, + "step": 27103 + }, + { + "epoch": 0.6353604435145953, + "grad_norm": 3.845176935195923, + "learning_rate": 5.876878949235273e-05, + "loss": 1.7424, + "step": 27104 + }, + { + "epoch": 0.6353838850894004, + "grad_norm": 4.245578289031982, + "learning_rate": 5.8762078720196144e-05, + "loss": 0.4164, + "step": 27105 + }, + { + "epoch": 0.6354073266642053, + "grad_norm": 3.308581829071045, + "learning_rate": 5.875536817179765e-05, + "loss": 0.2653, + "step": 27106 + }, + { + "epoch": 0.6354307682390103, + "grad_norm": 2.7585771083831787, + "learning_rate": 5.8748657847193725e-05, + "loss": 0.266, + "step": 27107 + }, + { + "epoch": 0.6354542098138153, + "grad_norm": 2.4859836101531982, + "learning_rate": 5.874194774642074e-05, + "loss": 0.2003, + "step": 27108 + }, + { + "epoch": 0.6354776513886203, + "grad_norm": 4.017473220825195, + "learning_rate": 5.873523786951507e-05, + "loss": 0.4069, + "step": 27109 + }, + { + "epoch": 0.6355010929634253, + "grad_norm": 3.928516387939453, + "learning_rate": 5.8728528216513204e-05, + "loss": 0.3544, + "step": 27110 + }, + { + "epoch": 0.6355245345382303, + "grad_norm": 3.613276958465576, + "learning_rate": 5.872181878745149e-05, + "loss": 0.1506, + "step": 27111 + }, + { + "epoch": 0.6355479761130353, + "grad_norm": 1.2918601036071777, + "learning_rate": 5.871510958236632e-05, + "loss": 0.1483, + "step": 27112 + }, + { + "epoch": 0.6355714176878403, + "grad_norm": 2.174060344696045, + "learning_rate": 5.870840060129417e-05, + "loss": 1.035, + "step": 27113 + }, + { + "epoch": 0.6355948592626453, + "grad_norm": 1.0054563283920288, + "learning_rate": 5.870169184427138e-05, + "loss": 0.0876, + "step": 27114 + }, + { + "epoch": 0.6356183008374503, + "grad_norm": 2.6507742404937744, + "learning_rate": 5.869498331133436e-05, + "loss": 0.2722, + "step": 27115 + }, + { + "epoch": 0.6356417424122552, + "grad_norm": 2.5342533588409424, + "learning_rate": 5.86882750025195e-05, + "loss": 0.1783, + "step": 27116 + }, + { + "epoch": 0.6356651839870603, + "grad_norm": 2.76857852935791, + "learning_rate": 5.868156691786325e-05, + "loss": 1.8629, + "step": 27117 + }, + { + "epoch": 0.6356886255618652, + "grad_norm": 1.8949486017227173, + "learning_rate": 5.8674859057402e-05, + "loss": 0.1567, + "step": 27118 + }, + { + "epoch": 0.6357120671366703, + "grad_norm": 3.554381847381592, + "learning_rate": 5.866815142117209e-05, + "loss": 0.3439, + "step": 27119 + }, + { + "epoch": 0.6357355087114752, + "grad_norm": 1.1372517347335815, + "learning_rate": 5.866144400920997e-05, + "loss": 0.1736, + "step": 27120 + }, + { + "epoch": 0.6357589502862803, + "grad_norm": 4.2676591873168945, + "learning_rate": 5.8654736821552026e-05, + "loss": 0.245, + "step": 27121 + }, + { + "epoch": 0.6357823918610852, + "grad_norm": 6.641439914703369, + "learning_rate": 5.864802985823463e-05, + "loss": 0.4599, + "step": 27122 + }, + { + "epoch": 0.6358058334358903, + "grad_norm": 3.082247018814087, + "learning_rate": 5.86413231192942e-05, + "loss": 0.2737, + "step": 27123 + }, + { + "epoch": 0.6358292750106952, + "grad_norm": 4.620151042938232, + "learning_rate": 5.863461660476709e-05, + "loss": 0.4461, + "step": 27124 + }, + { + "epoch": 0.6358527165855002, + "grad_norm": 2.580547332763672, + "learning_rate": 5.862791031468974e-05, + "loss": 0.1204, + "step": 27125 + }, + { + "epoch": 0.6358761581603052, + "grad_norm": 2.6578030586242676, + "learning_rate": 5.86212042490985e-05, + "loss": 0.1797, + "step": 27126 + }, + { + "epoch": 0.6358995997351102, + "grad_norm": 1.1588047742843628, + "learning_rate": 5.8614498408029764e-05, + "loss": 0.1603, + "step": 27127 + }, + { + "epoch": 0.6359230413099152, + "grad_norm": 2.2769017219543457, + "learning_rate": 5.860779279151989e-05, + "loss": 0.2478, + "step": 27128 + }, + { + "epoch": 0.6359464828847202, + "grad_norm": 3.1935436725616455, + "learning_rate": 5.8601087399605345e-05, + "loss": 0.2028, + "step": 27129 + }, + { + "epoch": 0.6359699244595252, + "grad_norm": 6.087676525115967, + "learning_rate": 5.859438223232246e-05, + "loss": 0.338, + "step": 27130 + }, + { + "epoch": 0.6359933660343302, + "grad_norm": 3.142549991607666, + "learning_rate": 5.858767728970761e-05, + "loss": 0.2094, + "step": 27131 + }, + { + "epoch": 0.6360168076091351, + "grad_norm": 3.3790504932403564, + "learning_rate": 5.85809725717972e-05, + "loss": 0.2288, + "step": 27132 + }, + { + "epoch": 0.6360402491839402, + "grad_norm": 2.8918774127960205, + "learning_rate": 5.85742680786276e-05, + "loss": 0.3376, + "step": 27133 + }, + { + "epoch": 0.6360636907587451, + "grad_norm": 3.2409331798553467, + "learning_rate": 5.856756381023516e-05, + "loss": 0.3387, + "step": 27134 + }, + { + "epoch": 0.6360871323335502, + "grad_norm": 3.3501040935516357, + "learning_rate": 5.856085976665631e-05, + "loss": 0.1862, + "step": 27135 + }, + { + "epoch": 0.6361105739083551, + "grad_norm": 3.407111883163452, + "learning_rate": 5.855415594792739e-05, + "loss": 2.0704, + "step": 27136 + }, + { + "epoch": 0.6361340154831602, + "grad_norm": 3.299966335296631, + "learning_rate": 5.854745235408479e-05, + "loss": 0.2437, + "step": 27137 + }, + { + "epoch": 0.6361574570579651, + "grad_norm": 3.2298128604888916, + "learning_rate": 5.8540748985164876e-05, + "loss": 0.3118, + "step": 27138 + }, + { + "epoch": 0.6361808986327702, + "grad_norm": 2.0598702430725098, + "learning_rate": 5.8534045841204035e-05, + "loss": 0.1744, + "step": 27139 + }, + { + "epoch": 0.6362043402075751, + "grad_norm": 1.6091598272323608, + "learning_rate": 5.852734292223857e-05, + "loss": 0.2017, + "step": 27140 + }, + { + "epoch": 0.6362277817823802, + "grad_norm": 1.6372759342193604, + "learning_rate": 5.8520640228304944e-05, + "loss": 0.131, + "step": 27141 + }, + { + "epoch": 0.6362512233571851, + "grad_norm": 2.7850027084350586, + "learning_rate": 5.85139377594395e-05, + "loss": 0.1268, + "step": 27142 + }, + { + "epoch": 0.6362746649319901, + "grad_norm": 0.23321805894374847, + "learning_rate": 5.85072355156786e-05, + "loss": 0.0425, + "step": 27143 + }, + { + "epoch": 0.6362981065067951, + "grad_norm": 2.5278196334838867, + "learning_rate": 5.8500533497058595e-05, + "loss": 0.2901, + "step": 27144 + }, + { + "epoch": 0.6363215480816001, + "grad_norm": 3.073244094848633, + "learning_rate": 5.849383170361589e-05, + "loss": 0.256, + "step": 27145 + }, + { + "epoch": 0.6363449896564051, + "grad_norm": 3.4434173107147217, + "learning_rate": 5.8487130135386805e-05, + "loss": 0.3685, + "step": 27146 + }, + { + "epoch": 0.6363684312312101, + "grad_norm": 2.5863797664642334, + "learning_rate": 5.848042879240769e-05, + "loss": 1.5905, + "step": 27147 + }, + { + "epoch": 0.6363918728060151, + "grad_norm": 3.6817996501922607, + "learning_rate": 5.847372767471497e-05, + "loss": 0.2223, + "step": 27148 + }, + { + "epoch": 0.6364153143808201, + "grad_norm": 1.6484092473983765, + "learning_rate": 5.846702678234497e-05, + "loss": 0.1513, + "step": 27149 + }, + { + "epoch": 0.636438755955625, + "grad_norm": 3.4060566425323486, + "learning_rate": 5.846032611533403e-05, + "loss": 0.3827, + "step": 27150 + }, + { + "epoch": 0.6364621975304301, + "grad_norm": 0.6233469247817993, + "learning_rate": 5.845362567371855e-05, + "loss": 0.0495, + "step": 27151 + }, + { + "epoch": 0.636485639105235, + "grad_norm": 1.316825032234192, + "learning_rate": 5.844692545753483e-05, + "loss": 0.1248, + "step": 27152 + }, + { + "epoch": 0.6365090806800401, + "grad_norm": 2.2956314086914062, + "learning_rate": 5.844022546681929e-05, + "loss": 0.1216, + "step": 27153 + }, + { + "epoch": 0.6365325222548451, + "grad_norm": 3.554426670074463, + "learning_rate": 5.8433525701608236e-05, + "loss": 0.3096, + "step": 27154 + }, + { + "epoch": 0.6365559638296501, + "grad_norm": 2.5492353439331055, + "learning_rate": 5.842682616193806e-05, + "loss": 1.6684, + "step": 27155 + }, + { + "epoch": 0.6365794054044551, + "grad_norm": 0.7081993818283081, + "learning_rate": 5.84201268478451e-05, + "loss": 0.0601, + "step": 27156 + }, + { + "epoch": 0.6366028469792601, + "grad_norm": 2.7600314617156982, + "learning_rate": 5.841342775936568e-05, + "loss": 0.4007, + "step": 27157 + }, + { + "epoch": 0.6366262885540651, + "grad_norm": 3.3751611709594727, + "learning_rate": 5.840672889653619e-05, + "loss": 0.3404, + "step": 27158 + }, + { + "epoch": 0.63664973012887, + "grad_norm": 1.7972517013549805, + "learning_rate": 5.8400030259392936e-05, + "loss": 0.0547, + "step": 27159 + }, + { + "epoch": 0.6366731717036751, + "grad_norm": 3.4231715202331543, + "learning_rate": 5.839333184797231e-05, + "loss": 0.124, + "step": 27160 + }, + { + "epoch": 0.63669661327848, + "grad_norm": 1.4492069482803345, + "learning_rate": 5.838663366231063e-05, + "loss": 0.1199, + "step": 27161 + }, + { + "epoch": 0.6367200548532851, + "grad_norm": 2.9107472896575928, + "learning_rate": 5.8379935702444256e-05, + "loss": 0.4462, + "step": 27162 + }, + { + "epoch": 0.63674349642809, + "grad_norm": 4.242940902709961, + "learning_rate": 5.837323796840952e-05, + "loss": 0.5373, + "step": 27163 + }, + { + "epoch": 0.6367669380028951, + "grad_norm": 2.8537204265594482, + "learning_rate": 5.836654046024271e-05, + "loss": 0.2011, + "step": 27164 + }, + { + "epoch": 0.6367903795777, + "grad_norm": 2.2224438190460205, + "learning_rate": 5.8359843177980244e-05, + "loss": 0.1794, + "step": 27165 + }, + { + "epoch": 0.6368138211525051, + "grad_norm": 6.1755805015563965, + "learning_rate": 5.835314612165846e-05, + "loss": 0.8521, + "step": 27166 + }, + { + "epoch": 0.63683726272731, + "grad_norm": 2.957200527191162, + "learning_rate": 5.834644929131369e-05, + "loss": 0.2244, + "step": 27167 + }, + { + "epoch": 0.636860704302115, + "grad_norm": 1.5822573900222778, + "learning_rate": 5.833975268698224e-05, + "loss": 0.105, + "step": 27168 + }, + { + "epoch": 0.63688414587692, + "grad_norm": 0.4322719871997833, + "learning_rate": 5.8333056308700485e-05, + "loss": 0.0393, + "step": 27169 + }, + { + "epoch": 0.636907587451725, + "grad_norm": 3.256232738494873, + "learning_rate": 5.832636015650472e-05, + "loss": 0.2308, + "step": 27170 + }, + { + "epoch": 0.63693102902653, + "grad_norm": 2.7146966457366943, + "learning_rate": 5.831966423043127e-05, + "loss": 0.239, + "step": 27171 + }, + { + "epoch": 0.636954470601335, + "grad_norm": 3.0342788696289062, + "learning_rate": 5.831296853051653e-05, + "loss": 1.6444, + "step": 27172 + }, + { + "epoch": 0.63697791217614, + "grad_norm": 1.7532413005828857, + "learning_rate": 5.830627305679679e-05, + "loss": 0.1654, + "step": 27173 + }, + { + "epoch": 0.637001353750945, + "grad_norm": 3.2151904106140137, + "learning_rate": 5.8299577809308394e-05, + "loss": 1.5553, + "step": 27174 + }, + { + "epoch": 0.63702479532575, + "grad_norm": 4.113795280456543, + "learning_rate": 5.8292882788087656e-05, + "loss": 0.2811, + "step": 27175 + }, + { + "epoch": 0.637048236900555, + "grad_norm": 3.3395183086395264, + "learning_rate": 5.828618799317087e-05, + "loss": 0.2529, + "step": 27176 + }, + { + "epoch": 0.63707167847536, + "grad_norm": 1.5519758462905884, + "learning_rate": 5.827949342459443e-05, + "loss": 0.6498, + "step": 27177 + }, + { + "epoch": 0.637095120050165, + "grad_norm": 2.585014820098877, + "learning_rate": 5.82727990823946e-05, + "loss": 0.1929, + "step": 27178 + }, + { + "epoch": 0.6371185616249699, + "grad_norm": 2.9350674152374268, + "learning_rate": 5.826610496660778e-05, + "loss": 0.2884, + "step": 27179 + }, + { + "epoch": 0.637142003199775, + "grad_norm": 3.413578987121582, + "learning_rate": 5.8259411077270245e-05, + "loss": 0.3097, + "step": 27180 + }, + { + "epoch": 0.6371654447745799, + "grad_norm": 1.184827446937561, + "learning_rate": 5.8252717414418315e-05, + "loss": 0.1673, + "step": 27181 + }, + { + "epoch": 0.637188886349385, + "grad_norm": 2.5989229679107666, + "learning_rate": 5.824602397808832e-05, + "loss": 0.1676, + "step": 27182 + }, + { + "epoch": 0.6372123279241899, + "grad_norm": 1.8419163227081299, + "learning_rate": 5.8239330768316534e-05, + "loss": 0.1824, + "step": 27183 + }, + { + "epoch": 0.637235769498995, + "grad_norm": 2.5701231956481934, + "learning_rate": 5.8232637785139345e-05, + "loss": 0.2906, + "step": 27184 + }, + { + "epoch": 0.6372592110737999, + "grad_norm": 3.1634531021118164, + "learning_rate": 5.8225945028593045e-05, + "loss": 0.2879, + "step": 27185 + }, + { + "epoch": 0.637282652648605, + "grad_norm": 0.7821442484855652, + "learning_rate": 5.821925249871394e-05, + "loss": 0.0513, + "step": 27186 + }, + { + "epoch": 0.6373060942234099, + "grad_norm": 2.050840377807617, + "learning_rate": 5.8212560195538345e-05, + "loss": 0.0924, + "step": 27187 + }, + { + "epoch": 0.6373295357982149, + "grad_norm": 2.9344968795776367, + "learning_rate": 5.820586811910258e-05, + "loss": 0.2451, + "step": 27188 + }, + { + "epoch": 0.6373529773730199, + "grad_norm": 1.9852731227874756, + "learning_rate": 5.819917626944289e-05, + "loss": 0.1148, + "step": 27189 + }, + { + "epoch": 0.6373764189478249, + "grad_norm": 3.5088016986846924, + "learning_rate": 5.819248464659566e-05, + "loss": 0.4508, + "step": 27190 + }, + { + "epoch": 0.6373998605226299, + "grad_norm": 1.7649632692337036, + "learning_rate": 5.8185793250597234e-05, + "loss": 0.1426, + "step": 27191 + }, + { + "epoch": 0.6374233020974349, + "grad_norm": 0.4686235189437866, + "learning_rate": 5.817910208148385e-05, + "loss": 0.0326, + "step": 27192 + }, + { + "epoch": 0.6374467436722399, + "grad_norm": 3.9428133964538574, + "learning_rate": 5.817241113929184e-05, + "loss": 0.2878, + "step": 27193 + }, + { + "epoch": 0.6374701852470449, + "grad_norm": 0.46604984998703003, + "learning_rate": 5.81657204240575e-05, + "loss": 0.0239, + "step": 27194 + }, + { + "epoch": 0.6374936268218498, + "grad_norm": 2.975053310394287, + "learning_rate": 5.815902993581714e-05, + "loss": 0.2314, + "step": 27195 + }, + { + "epoch": 0.6375170683966549, + "grad_norm": 2.2715721130371094, + "learning_rate": 5.8152339674607026e-05, + "loss": 0.1829, + "step": 27196 + }, + { + "epoch": 0.6375405099714598, + "grad_norm": 2.463174343109131, + "learning_rate": 5.814564964046354e-05, + "loss": 0.2974, + "step": 27197 + }, + { + "epoch": 0.6375639515462649, + "grad_norm": 4.920764923095703, + "learning_rate": 5.8138959833422925e-05, + "loss": 0.5007, + "step": 27198 + }, + { + "epoch": 0.6375873931210698, + "grad_norm": 2.219675064086914, + "learning_rate": 5.8132270253521485e-05, + "loss": 0.1555, + "step": 27199 + }, + { + "epoch": 0.6376108346958749, + "grad_norm": 3.238955497741699, + "learning_rate": 5.812558090079554e-05, + "loss": 0.3563, + "step": 27200 + }, + { + "epoch": 0.6376342762706798, + "grad_norm": 3.4278478622436523, + "learning_rate": 5.8118891775281314e-05, + "loss": 0.262, + "step": 27201 + }, + { + "epoch": 0.6376577178454849, + "grad_norm": 2.376554012298584, + "learning_rate": 5.81122028770152e-05, + "loss": 0.1685, + "step": 27202 + }, + { + "epoch": 0.6376811594202898, + "grad_norm": 3.205871105194092, + "learning_rate": 5.810551420603342e-05, + "loss": 0.1627, + "step": 27203 + }, + { + "epoch": 0.6377046009950949, + "grad_norm": 0.5406495332717896, + "learning_rate": 5.809882576237233e-05, + "loss": 0.0637, + "step": 27204 + }, + { + "epoch": 0.6377280425698999, + "grad_norm": 3.811972141265869, + "learning_rate": 5.8092137546068196e-05, + "loss": 0.3458, + "step": 27205 + }, + { + "epoch": 0.6377514841447048, + "grad_norm": 2.4501237869262695, + "learning_rate": 5.80854495571573e-05, + "loss": 0.1257, + "step": 27206 + }, + { + "epoch": 0.6377749257195099, + "grad_norm": 3.214339017868042, + "learning_rate": 5.807876179567593e-05, + "loss": 0.3615, + "step": 27207 + }, + { + "epoch": 0.6377983672943148, + "grad_norm": 3.1781280040740967, + "learning_rate": 5.8072074261660395e-05, + "loss": 0.2161, + "step": 27208 + }, + { + "epoch": 0.6378218088691199, + "grad_norm": 1.182380199432373, + "learning_rate": 5.8065386955146894e-05, + "loss": 0.0918, + "step": 27209 + }, + { + "epoch": 0.6378452504439248, + "grad_norm": 2.100083827972412, + "learning_rate": 5.8058699876171854e-05, + "loss": 0.2413, + "step": 27210 + }, + { + "epoch": 0.6378686920187299, + "grad_norm": 6.26738166809082, + "learning_rate": 5.805201302477146e-05, + "loss": 0.4701, + "step": 27211 + }, + { + "epoch": 0.6378921335935348, + "grad_norm": 4.308437824249268, + "learning_rate": 5.804532640098204e-05, + "loss": 0.4444, + "step": 27212 + }, + { + "epoch": 0.6379155751683399, + "grad_norm": 2.0379462242126465, + "learning_rate": 5.8038640004839804e-05, + "loss": 0.1306, + "step": 27213 + }, + { + "epoch": 0.6379390167431448, + "grad_norm": 2.9849534034729004, + "learning_rate": 5.8031953836381136e-05, + "loss": 0.4386, + "step": 27214 + }, + { + "epoch": 0.6379624583179498, + "grad_norm": 1.922615647315979, + "learning_rate": 5.802526789564227e-05, + "loss": 1.4831, + "step": 27215 + }, + { + "epoch": 0.6379858998927548, + "grad_norm": 4.040355205535889, + "learning_rate": 5.801858218265944e-05, + "loss": 0.2969, + "step": 27216 + }, + { + "epoch": 0.6380093414675598, + "grad_norm": 3.517965078353882, + "learning_rate": 5.8011896697469e-05, + "loss": 0.2865, + "step": 27217 + }, + { + "epoch": 0.6380327830423648, + "grad_norm": 3.293861150741577, + "learning_rate": 5.80052114401072e-05, + "loss": 0.3688, + "step": 27218 + }, + { + "epoch": 0.6380562246171698, + "grad_norm": 0.8980490565299988, + "learning_rate": 5.7998526410610275e-05, + "loss": 0.0877, + "step": 27219 + }, + { + "epoch": 0.6380796661919748, + "grad_norm": 2.969540596008301, + "learning_rate": 5.799184160901455e-05, + "loss": 0.3375, + "step": 27220 + }, + { + "epoch": 0.6381031077667798, + "grad_norm": 5.033304214477539, + "learning_rate": 5.798515703535623e-05, + "loss": 0.35, + "step": 27221 + }, + { + "epoch": 0.6381265493415847, + "grad_norm": 1.6505494117736816, + "learning_rate": 5.797847268967167e-05, + "loss": 0.1244, + "step": 27222 + }, + { + "epoch": 0.6381499909163898, + "grad_norm": 3.3052966594696045, + "learning_rate": 5.79717885719971e-05, + "loss": 0.131, + "step": 27223 + }, + { + "epoch": 0.6381734324911947, + "grad_norm": 2.6444480419158936, + "learning_rate": 5.7965104682368776e-05, + "loss": 0.2294, + "step": 27224 + }, + { + "epoch": 0.6381968740659998, + "grad_norm": 3.143414258956909, + "learning_rate": 5.795842102082295e-05, + "loss": 0.3096, + "step": 27225 + }, + { + "epoch": 0.6382203156408047, + "grad_norm": 3.6866629123687744, + "learning_rate": 5.795173758739595e-05, + "loss": 0.4135, + "step": 27226 + }, + { + "epoch": 0.6382437572156098, + "grad_norm": 0.7315050363540649, + "learning_rate": 5.7945054382123996e-05, + "loss": 0.0748, + "step": 27227 + }, + { + "epoch": 0.6382671987904147, + "grad_norm": 3.12048077583313, + "learning_rate": 5.7938371405043323e-05, + "loss": 0.2777, + "step": 27228 + }, + { + "epoch": 0.6382906403652198, + "grad_norm": 2.6107733249664307, + "learning_rate": 5.793168865619028e-05, + "loss": 0.1809, + "step": 27229 + }, + { + "epoch": 0.6383140819400247, + "grad_norm": 2.764957904815674, + "learning_rate": 5.792500613560106e-05, + "loss": 0.3266, + "step": 27230 + }, + { + "epoch": 0.6383375235148298, + "grad_norm": 1.6642781496047974, + "learning_rate": 5.791832384331195e-05, + "loss": 0.0999, + "step": 27231 + }, + { + "epoch": 0.6383609650896347, + "grad_norm": 0.6488891243934631, + "learning_rate": 5.791164177935919e-05, + "loss": 0.0599, + "step": 27232 + }, + { + "epoch": 0.6383844066644397, + "grad_norm": 2.9641125202178955, + "learning_rate": 5.7904959943779055e-05, + "loss": 0.2878, + "step": 27233 + }, + { + "epoch": 0.6384078482392447, + "grad_norm": 3.632565975189209, + "learning_rate": 5.789827833660775e-05, + "loss": 0.2449, + "step": 27234 + }, + { + "epoch": 0.6384312898140497, + "grad_norm": 2.7921082973480225, + "learning_rate": 5.789159695788161e-05, + "loss": 0.1402, + "step": 27235 + }, + { + "epoch": 0.6384547313888547, + "grad_norm": 2.1782419681549072, + "learning_rate": 5.788491580763684e-05, + "loss": 0.1429, + "step": 27236 + }, + { + "epoch": 0.6384781729636597, + "grad_norm": 1.1230684518814087, + "learning_rate": 5.787823488590971e-05, + "loss": 0.1041, + "step": 27237 + }, + { + "epoch": 0.6385016145384647, + "grad_norm": 3.990651845932007, + "learning_rate": 5.787155419273642e-05, + "loss": 0.3882, + "step": 27238 + }, + { + "epoch": 0.6385250561132697, + "grad_norm": 1.1549526453018188, + "learning_rate": 5.7864873728153305e-05, + "loss": 0.0895, + "step": 27239 + }, + { + "epoch": 0.6385484976880746, + "grad_norm": 2.6159744262695312, + "learning_rate": 5.785819349219656e-05, + "loss": 0.2149, + "step": 27240 + }, + { + "epoch": 0.6385719392628797, + "grad_norm": 1.4287190437316895, + "learning_rate": 5.785151348490241e-05, + "loss": 0.1224, + "step": 27241 + }, + { + "epoch": 0.6385953808376846, + "grad_norm": 2.156100273132324, + "learning_rate": 5.784483370630718e-05, + "loss": 0.1277, + "step": 27242 + }, + { + "epoch": 0.6386188224124897, + "grad_norm": 0.798685610294342, + "learning_rate": 5.783815415644705e-05, + "loss": 0.0918, + "step": 27243 + }, + { + "epoch": 0.6386422639872946, + "grad_norm": 3.6778452396392822, + "learning_rate": 5.783147483535829e-05, + "loss": 0.3084, + "step": 27244 + }, + { + "epoch": 0.6386657055620997, + "grad_norm": 3.581772804260254, + "learning_rate": 5.782479574307713e-05, + "loss": 0.2985, + "step": 27245 + }, + { + "epoch": 0.6386891471369046, + "grad_norm": 3.6949939727783203, + "learning_rate": 5.7818116879639824e-05, + "loss": 0.2844, + "step": 27246 + }, + { + "epoch": 0.6387125887117097, + "grad_norm": 3.137244701385498, + "learning_rate": 5.781143824508255e-05, + "loss": 0.2268, + "step": 27247 + }, + { + "epoch": 0.6387360302865146, + "grad_norm": 3.5254194736480713, + "learning_rate": 5.780475983944165e-05, + "loss": 0.3344, + "step": 27248 + }, + { + "epoch": 0.6387594718613196, + "grad_norm": 1.8348323106765747, + "learning_rate": 5.77980816627533e-05, + "loss": 0.1337, + "step": 27249 + }, + { + "epoch": 0.6387829134361246, + "grad_norm": 4.4709320068359375, + "learning_rate": 5.779140371505371e-05, + "loss": 0.3593, + "step": 27250 + }, + { + "epoch": 0.6388063550109296, + "grad_norm": 2.608579635620117, + "learning_rate": 5.7784725996379185e-05, + "loss": 0.131, + "step": 27251 + }, + { + "epoch": 0.6388297965857346, + "grad_norm": 3.143636703491211, + "learning_rate": 5.7778048506765934e-05, + "loss": 0.2867, + "step": 27252 + }, + { + "epoch": 0.6388532381605396, + "grad_norm": 1.4294192790985107, + "learning_rate": 5.777137124625016e-05, + "loss": 0.0894, + "step": 27253 + }, + { + "epoch": 0.6388766797353446, + "grad_norm": 3.075080633163452, + "learning_rate": 5.77646942148681e-05, + "loss": 1.7552, + "step": 27254 + }, + { + "epoch": 0.6389001213101496, + "grad_norm": 3.3813364505767822, + "learning_rate": 5.7758017412656005e-05, + "loss": 0.1038, + "step": 27255 + }, + { + "epoch": 0.6389235628849547, + "grad_norm": 2.951683282852173, + "learning_rate": 5.775134083965013e-05, + "loss": 0.2426, + "step": 27256 + }, + { + "epoch": 0.6389470044597596, + "grad_norm": 4.0384440422058105, + "learning_rate": 5.774466449588665e-05, + "loss": 0.3964, + "step": 27257 + }, + { + "epoch": 0.6389704460345647, + "grad_norm": 2.058471202850342, + "learning_rate": 5.7737988381401795e-05, + "loss": 1.2123, + "step": 27258 + }, + { + "epoch": 0.6389938876093696, + "grad_norm": 1.2545346021652222, + "learning_rate": 5.773131249623178e-05, + "loss": 0.1782, + "step": 27259 + }, + { + "epoch": 0.6390173291841746, + "grad_norm": 5.108903408050537, + "learning_rate": 5.7724636840412895e-05, + "loss": 0.6012, + "step": 27260 + }, + { + "epoch": 0.6390407707589796, + "grad_norm": 2.891425371170044, + "learning_rate": 5.7717961413981314e-05, + "loss": 0.1248, + "step": 27261 + }, + { + "epoch": 0.6390642123337846, + "grad_norm": 2.747424364089966, + "learning_rate": 5.7711286216973215e-05, + "loss": 0.113, + "step": 27262 + }, + { + "epoch": 0.6390876539085896, + "grad_norm": 2.3003907203674316, + "learning_rate": 5.770461124942491e-05, + "loss": 0.2089, + "step": 27263 + }, + { + "epoch": 0.6391110954833946, + "grad_norm": 3.2250378131866455, + "learning_rate": 5.769793651137259e-05, + "loss": 0.3754, + "step": 27264 + }, + { + "epoch": 0.6391345370581996, + "grad_norm": 5.3795366287231445, + "learning_rate": 5.769126200285243e-05, + "loss": 0.3435, + "step": 27265 + }, + { + "epoch": 0.6391579786330046, + "grad_norm": 2.8709981441497803, + "learning_rate": 5.768458772390064e-05, + "loss": 0.1216, + "step": 27266 + }, + { + "epoch": 0.6391814202078095, + "grad_norm": 2.28222918510437, + "learning_rate": 5.767791367455352e-05, + "loss": 0.2031, + "step": 27267 + }, + { + "epoch": 0.6392048617826146, + "grad_norm": 2.666008710861206, + "learning_rate": 5.767123985484722e-05, + "loss": 0.1581, + "step": 27268 + }, + { + "epoch": 0.6392283033574195, + "grad_norm": 2.5892322063446045, + "learning_rate": 5.766456626481796e-05, + "loss": 0.2408, + "step": 27269 + }, + { + "epoch": 0.6392517449322246, + "grad_norm": 3.004470109939575, + "learning_rate": 5.7657892904501965e-05, + "loss": 0.168, + "step": 27270 + }, + { + "epoch": 0.6392751865070295, + "grad_norm": 2.658477783203125, + "learning_rate": 5.765121977393541e-05, + "loss": 0.1403, + "step": 27271 + }, + { + "epoch": 0.6392986280818346, + "grad_norm": 1.9780699014663696, + "learning_rate": 5.764454687315453e-05, + "loss": 1.2568, + "step": 27272 + }, + { + "epoch": 0.6393220696566395, + "grad_norm": 6.790922164916992, + "learning_rate": 5.763787420219553e-05, + "loss": 0.3626, + "step": 27273 + }, + { + "epoch": 0.6393455112314446, + "grad_norm": 2.3866162300109863, + "learning_rate": 5.763120176109459e-05, + "loss": 0.1986, + "step": 27274 + }, + { + "epoch": 0.6393689528062495, + "grad_norm": 2.6697897911071777, + "learning_rate": 5.762452954988797e-05, + "loss": 0.2481, + "step": 27275 + }, + { + "epoch": 0.6393923943810546, + "grad_norm": 1.5254405736923218, + "learning_rate": 5.761785756861187e-05, + "loss": 0.0992, + "step": 27276 + }, + { + "epoch": 0.6394158359558595, + "grad_norm": 3.25996470451355, + "learning_rate": 5.761118581730245e-05, + "loss": 0.2712, + "step": 27277 + }, + { + "epoch": 0.6394392775306645, + "grad_norm": 1.8690048456192017, + "learning_rate": 5.7604514295995926e-05, + "loss": 0.0978, + "step": 27278 + }, + { + "epoch": 0.6394627191054695, + "grad_norm": 3.0993998050689697, + "learning_rate": 5.759784300472846e-05, + "loss": 0.2395, + "step": 27279 + }, + { + "epoch": 0.6394861606802745, + "grad_norm": 2.9542412757873535, + "learning_rate": 5.759117194353634e-05, + "loss": 0.2736, + "step": 27280 + }, + { + "epoch": 0.6395096022550795, + "grad_norm": 3.7485156059265137, + "learning_rate": 5.75845011124557e-05, + "loss": 0.4088, + "step": 27281 + }, + { + "epoch": 0.6395330438298845, + "grad_norm": 3.490266799926758, + "learning_rate": 5.757783051152275e-05, + "loss": 0.235, + "step": 27282 + }, + { + "epoch": 0.6395564854046895, + "grad_norm": 3.318519353866577, + "learning_rate": 5.7571160140773683e-05, + "loss": 0.2625, + "step": 27283 + }, + { + "epoch": 0.6395799269794945, + "grad_norm": 2.3766512870788574, + "learning_rate": 5.756449000024471e-05, + "loss": 0.1459, + "step": 27284 + }, + { + "epoch": 0.6396033685542994, + "grad_norm": 1.8635189533233643, + "learning_rate": 5.7557820089971956e-05, + "loss": 0.2152, + "step": 27285 + }, + { + "epoch": 0.6396268101291045, + "grad_norm": 3.5884060859680176, + "learning_rate": 5.7551150409991663e-05, + "loss": 0.3383, + "step": 27286 + }, + { + "epoch": 0.6396502517039094, + "grad_norm": 3.08962345123291, + "learning_rate": 5.7544480960340067e-05, + "loss": 0.3329, + "step": 27287 + }, + { + "epoch": 0.6396736932787145, + "grad_norm": 0.8966231942176819, + "learning_rate": 5.75378117410533e-05, + "loss": 0.0629, + "step": 27288 + }, + { + "epoch": 0.6396971348535194, + "grad_norm": 2.734691858291626, + "learning_rate": 5.753114275216758e-05, + "loss": 0.1007, + "step": 27289 + }, + { + "epoch": 0.6397205764283245, + "grad_norm": 3.4591355323791504, + "learning_rate": 5.752447399371905e-05, + "loss": 0.3074, + "step": 27290 + }, + { + "epoch": 0.6397440180031294, + "grad_norm": 4.43780517578125, + "learning_rate": 5.7517805465743946e-05, + "loss": 0.2768, + "step": 27291 + }, + { + "epoch": 0.6397674595779345, + "grad_norm": 2.8083112239837646, + "learning_rate": 5.751113716827836e-05, + "loss": 0.2392, + "step": 27292 + }, + { + "epoch": 0.6397909011527394, + "grad_norm": 2.2992284297943115, + "learning_rate": 5.7504469101358585e-05, + "loss": 0.2174, + "step": 27293 + }, + { + "epoch": 0.6398143427275444, + "grad_norm": 3.195096969604492, + "learning_rate": 5.749780126502076e-05, + "loss": 0.3258, + "step": 27294 + }, + { + "epoch": 0.6398377843023494, + "grad_norm": 2.51751708984375, + "learning_rate": 5.749113365930106e-05, + "loss": 0.3304, + "step": 27295 + }, + { + "epoch": 0.6398612258771544, + "grad_norm": 2.015805959701538, + "learning_rate": 5.748446628423566e-05, + "loss": 0.1512, + "step": 27296 + }, + { + "epoch": 0.6398846674519594, + "grad_norm": 1.4139500856399536, + "learning_rate": 5.74777991398607e-05, + "loss": 0.0988, + "step": 27297 + }, + { + "epoch": 0.6399081090267644, + "grad_norm": 0.9691592454910278, + "learning_rate": 5.747113222621244e-05, + "loss": 0.1019, + "step": 27298 + }, + { + "epoch": 0.6399315506015694, + "grad_norm": 1.8911148309707642, + "learning_rate": 5.7464465543326976e-05, + "loss": 0.0607, + "step": 27299 + }, + { + "epoch": 0.6399549921763744, + "grad_norm": 2.7347793579101562, + "learning_rate": 5.745779909124056e-05, + "loss": 1.3605, + "step": 27300 + }, + { + "epoch": 0.6399784337511794, + "grad_norm": 0.6280021071434021, + "learning_rate": 5.745113286998931e-05, + "loss": 0.0599, + "step": 27301 + }, + { + "epoch": 0.6400018753259844, + "grad_norm": 0.8876166939735413, + "learning_rate": 5.744446687960942e-05, + "loss": 0.0263, + "step": 27302 + }, + { + "epoch": 0.6400253169007893, + "grad_norm": 3.5794267654418945, + "learning_rate": 5.743780112013704e-05, + "loss": 0.3501, + "step": 27303 + }, + { + "epoch": 0.6400487584755944, + "grad_norm": 4.98282527923584, + "learning_rate": 5.743113559160831e-05, + "loss": 0.4092, + "step": 27304 + }, + { + "epoch": 0.6400722000503993, + "grad_norm": 3.354005813598633, + "learning_rate": 5.7424470294059485e-05, + "loss": 0.3009, + "step": 27305 + }, + { + "epoch": 0.6400956416252044, + "grad_norm": 0.5760796666145325, + "learning_rate": 5.7417805227526675e-05, + "loss": 0.045, + "step": 27306 + }, + { + "epoch": 0.6401190832000094, + "grad_norm": 1.6866168975830078, + "learning_rate": 5.741114039204605e-05, + "loss": 0.2379, + "step": 27307 + }, + { + "epoch": 0.6401425247748144, + "grad_norm": 3.091723680496216, + "learning_rate": 5.740447578765378e-05, + "loss": 0.2604, + "step": 27308 + }, + { + "epoch": 0.6401659663496194, + "grad_norm": 3.0710160732269287, + "learning_rate": 5.739781141438602e-05, + "loss": 0.2846, + "step": 27309 + }, + { + "epoch": 0.6401894079244244, + "grad_norm": 1.9282557964324951, + "learning_rate": 5.739114727227889e-05, + "loss": 0.1225, + "step": 27310 + }, + { + "epoch": 0.6402128494992294, + "grad_norm": 1.0953736305236816, + "learning_rate": 5.738448336136859e-05, + "loss": 0.0811, + "step": 27311 + }, + { + "epoch": 0.6402362910740343, + "grad_norm": 1.9249813556671143, + "learning_rate": 5.737781968169133e-05, + "loss": 0.1287, + "step": 27312 + }, + { + "epoch": 0.6402597326488394, + "grad_norm": 2.0985121726989746, + "learning_rate": 5.737115623328322e-05, + "loss": 1.402, + "step": 27313 + }, + { + "epoch": 0.6402831742236443, + "grad_norm": 2.7336068153381348, + "learning_rate": 5.7364493016180406e-05, + "loss": 0.0822, + "step": 27314 + }, + { + "epoch": 0.6403066157984494, + "grad_norm": 1.6302635669708252, + "learning_rate": 5.735783003041906e-05, + "loss": 0.1817, + "step": 27315 + }, + { + "epoch": 0.6403300573732543, + "grad_norm": 1.026554822921753, + "learning_rate": 5.735116727603531e-05, + "loss": 0.0938, + "step": 27316 + }, + { + "epoch": 0.6403534989480594, + "grad_norm": 3.3248729705810547, + "learning_rate": 5.73445047530653e-05, + "loss": 0.2309, + "step": 27317 + }, + { + "epoch": 0.6403769405228643, + "grad_norm": 2.7733259201049805, + "learning_rate": 5.7337842461545244e-05, + "loss": 0.2207, + "step": 27318 + }, + { + "epoch": 0.6404003820976694, + "grad_norm": 2.432701587677002, + "learning_rate": 5.7331180401511265e-05, + "loss": 0.137, + "step": 27319 + }, + { + "epoch": 0.6404238236724743, + "grad_norm": 3.9341702461242676, + "learning_rate": 5.732451857299947e-05, + "loss": 0.3449, + "step": 27320 + }, + { + "epoch": 0.6404472652472794, + "grad_norm": 3.8535282611846924, + "learning_rate": 5.731785697604607e-05, + "loss": 0.3565, + "step": 27321 + }, + { + "epoch": 0.6404707068220843, + "grad_norm": 3.310600757598877, + "learning_rate": 5.731119561068716e-05, + "loss": 0.2846, + "step": 27322 + }, + { + "epoch": 0.6404941483968893, + "grad_norm": 1.6009446382522583, + "learning_rate": 5.730453447695887e-05, + "loss": 0.5624, + "step": 27323 + }, + { + "epoch": 0.6405175899716943, + "grad_norm": 1.9096615314483643, + "learning_rate": 5.7297873574897374e-05, + "loss": 0.1788, + "step": 27324 + }, + { + "epoch": 0.6405410315464993, + "grad_norm": 2.8320932388305664, + "learning_rate": 5.729121290453887e-05, + "loss": 0.3972, + "step": 27325 + }, + { + "epoch": 0.6405644731213043, + "grad_norm": 2.6768438816070557, + "learning_rate": 5.728455246591943e-05, + "loss": 0.2336, + "step": 27326 + }, + { + "epoch": 0.6405879146961093, + "grad_norm": 1.1005803346633911, + "learning_rate": 5.7277892259075206e-05, + "loss": 0.0731, + "step": 27327 + }, + { + "epoch": 0.6406113562709143, + "grad_norm": 0.6628221869468689, + "learning_rate": 5.727123228404235e-05, + "loss": 0.0511, + "step": 27328 + }, + { + "epoch": 0.6406347978457193, + "grad_norm": 4.419778823852539, + "learning_rate": 5.726457254085699e-05, + "loss": 0.4031, + "step": 27329 + }, + { + "epoch": 0.6406582394205242, + "grad_norm": 4.275484085083008, + "learning_rate": 5.725791302955521e-05, + "loss": 0.4105, + "step": 27330 + }, + { + "epoch": 0.6406816809953293, + "grad_norm": 1.3713901042938232, + "learning_rate": 5.725125375017325e-05, + "loss": 0.8543, + "step": 27331 + }, + { + "epoch": 0.6407051225701342, + "grad_norm": 3.003326177597046, + "learning_rate": 5.724459470274718e-05, + "loss": 0.1629, + "step": 27332 + }, + { + "epoch": 0.6407285641449393, + "grad_norm": 0.4852352738380432, + "learning_rate": 5.7237935887313144e-05, + "loss": 0.0494, + "step": 27333 + }, + { + "epoch": 0.6407520057197442, + "grad_norm": 3.157799005508423, + "learning_rate": 5.723127730390727e-05, + "loss": 0.38, + "step": 27334 + }, + { + "epoch": 0.6407754472945493, + "grad_norm": 2.9948720932006836, + "learning_rate": 5.7224618952565646e-05, + "loss": 0.259, + "step": 27335 + }, + { + "epoch": 0.6407988888693542, + "grad_norm": 2.5862843990325928, + "learning_rate": 5.721796083332445e-05, + "loss": 1.4184, + "step": 27336 + }, + { + "epoch": 0.6408223304441593, + "grad_norm": 2.4879508018493652, + "learning_rate": 5.721130294621984e-05, + "loss": 0.1593, + "step": 27337 + }, + { + "epoch": 0.6408457720189642, + "grad_norm": 0.26704034209251404, + "learning_rate": 5.720464529128788e-05, + "loss": 0.0131, + "step": 27338 + }, + { + "epoch": 0.6408692135937692, + "grad_norm": 1.052676796913147, + "learning_rate": 5.719798786856474e-05, + "loss": 0.1273, + "step": 27339 + }, + { + "epoch": 0.6408926551685742, + "grad_norm": 2.8627493381500244, + "learning_rate": 5.719133067808652e-05, + "loss": 0.352, + "step": 27340 + }, + { + "epoch": 0.6409160967433792, + "grad_norm": 3.185624599456787, + "learning_rate": 5.718467371988934e-05, + "loss": 0.1887, + "step": 27341 + }, + { + "epoch": 0.6409395383181842, + "grad_norm": 1.2115100622177124, + "learning_rate": 5.717801699400929e-05, + "loss": 0.1099, + "step": 27342 + }, + { + "epoch": 0.6409629798929892, + "grad_norm": 1.3243331909179688, + "learning_rate": 5.717136050048255e-05, + "loss": 0.1039, + "step": 27343 + }, + { + "epoch": 0.6409864214677942, + "grad_norm": 3.860239028930664, + "learning_rate": 5.7164704239345235e-05, + "loss": 0.3641, + "step": 27344 + }, + { + "epoch": 0.6410098630425992, + "grad_norm": 4.3082451820373535, + "learning_rate": 5.715804821063342e-05, + "loss": 0.3608, + "step": 27345 + }, + { + "epoch": 0.6410333046174042, + "grad_norm": 1.1788820028305054, + "learning_rate": 5.7151392414383244e-05, + "loss": 0.0661, + "step": 27346 + }, + { + "epoch": 0.6410567461922092, + "grad_norm": 4.167501449584961, + "learning_rate": 5.714473685063078e-05, + "loss": 0.7326, + "step": 27347 + }, + { + "epoch": 0.6410801877670141, + "grad_norm": 2.579861640930176, + "learning_rate": 5.713808151941222e-05, + "loss": 0.2777, + "step": 27348 + }, + { + "epoch": 0.6411036293418192, + "grad_norm": 1.658476710319519, + "learning_rate": 5.713142642076359e-05, + "loss": 0.1556, + "step": 27349 + }, + { + "epoch": 0.6411270709166241, + "grad_norm": 0.47493892908096313, + "learning_rate": 5.712477155472109e-05, + "loss": 0.035, + "step": 27350 + }, + { + "epoch": 0.6411505124914292, + "grad_norm": 0.5507429242134094, + "learning_rate": 5.711811692132079e-05, + "loss": 0.0434, + "step": 27351 + }, + { + "epoch": 0.6411739540662341, + "grad_norm": 1.068203330039978, + "learning_rate": 5.711146252059878e-05, + "loss": 0.0984, + "step": 27352 + }, + { + "epoch": 0.6411973956410392, + "grad_norm": 2.331040382385254, + "learning_rate": 5.710480835259119e-05, + "loss": 0.2592, + "step": 27353 + }, + { + "epoch": 0.6412208372158441, + "grad_norm": 2.1941661834716797, + "learning_rate": 5.7098154417334105e-05, + "loss": 0.1747, + "step": 27354 + }, + { + "epoch": 0.6412442787906492, + "grad_norm": 0.8406171798706055, + "learning_rate": 5.709150071486361e-05, + "loss": 0.0649, + "step": 27355 + }, + { + "epoch": 0.6412677203654541, + "grad_norm": 4.013798713684082, + "learning_rate": 5.708484724521588e-05, + "loss": 0.2956, + "step": 27356 + }, + { + "epoch": 0.6412911619402591, + "grad_norm": 3.8299901485443115, + "learning_rate": 5.707819400842698e-05, + "loss": 0.2706, + "step": 27357 + }, + { + "epoch": 0.6413146035150642, + "grad_norm": 1.3517568111419678, + "learning_rate": 5.707154100453299e-05, + "loss": 0.244, + "step": 27358 + }, + { + "epoch": 0.6413380450898691, + "grad_norm": 3.534630060195923, + "learning_rate": 5.706488823356999e-05, + "loss": 0.2909, + "step": 27359 + }, + { + "epoch": 0.6413614866646742, + "grad_norm": 2.238771677017212, + "learning_rate": 5.705823569557416e-05, + "loss": 0.2431, + "step": 27360 + }, + { + "epoch": 0.6413849282394791, + "grad_norm": 3.740025520324707, + "learning_rate": 5.705158339058155e-05, + "loss": 0.3696, + "step": 27361 + }, + { + "epoch": 0.6414083698142842, + "grad_norm": 3.646448850631714, + "learning_rate": 5.704493131862821e-05, + "loss": 0.2839, + "step": 27362 + }, + { + "epoch": 0.6414318113890891, + "grad_norm": 1.1984431743621826, + "learning_rate": 5.7038279479750335e-05, + "loss": 0.155, + "step": 27363 + }, + { + "epoch": 0.6414552529638942, + "grad_norm": 2.508418321609497, + "learning_rate": 5.703162787398395e-05, + "loss": 0.1782, + "step": 27364 + }, + { + "epoch": 0.6414786945386991, + "grad_norm": 2.943972587585449, + "learning_rate": 5.7024976501365156e-05, + "loss": 0.4455, + "step": 27365 + }, + { + "epoch": 0.6415021361135042, + "grad_norm": 3.4443421363830566, + "learning_rate": 5.701832536193006e-05, + "loss": 0.3901, + "step": 27366 + }, + { + "epoch": 0.6415255776883091, + "grad_norm": 2.0319912433624268, + "learning_rate": 5.7011674455714734e-05, + "loss": 1.4611, + "step": 27367 + }, + { + "epoch": 0.6415490192631141, + "grad_norm": 2.4084606170654297, + "learning_rate": 5.700502378275523e-05, + "loss": 0.3581, + "step": 27368 + }, + { + "epoch": 0.6415724608379191, + "grad_norm": 2.5234155654907227, + "learning_rate": 5.699837334308773e-05, + "loss": 0.0986, + "step": 27369 + }, + { + "epoch": 0.6415959024127241, + "grad_norm": 1.649888277053833, + "learning_rate": 5.699172313674825e-05, + "loss": 0.1473, + "step": 27370 + }, + { + "epoch": 0.6416193439875291, + "grad_norm": 1.1462788581848145, + "learning_rate": 5.698507316377285e-05, + "loss": 0.1074, + "step": 27371 + }, + { + "epoch": 0.6416427855623341, + "grad_norm": 3.617518424987793, + "learning_rate": 5.6978423424197704e-05, + "loss": 0.3922, + "step": 27372 + }, + { + "epoch": 0.6416662271371391, + "grad_norm": 3.8935699462890625, + "learning_rate": 5.697177391805882e-05, + "loss": 0.2249, + "step": 27373 + }, + { + "epoch": 0.6416896687119441, + "grad_norm": 2.9970316886901855, + "learning_rate": 5.6965124645392275e-05, + "loss": 0.2016, + "step": 27374 + }, + { + "epoch": 0.641713110286749, + "grad_norm": 2.1409566402435303, + "learning_rate": 5.695847560623422e-05, + "loss": 0.1808, + "step": 27375 + }, + { + "epoch": 0.6417365518615541, + "grad_norm": 7.165311813354492, + "learning_rate": 5.6951826800620675e-05, + "loss": 0.3067, + "step": 27376 + }, + { + "epoch": 0.641759993436359, + "grad_norm": 2.951890707015991, + "learning_rate": 5.6945178228587736e-05, + "loss": 0.2003, + "step": 27377 + }, + { + "epoch": 0.6417834350111641, + "grad_norm": 3.5502431392669678, + "learning_rate": 5.6938529890171464e-05, + "loss": 0.3281, + "step": 27378 + }, + { + "epoch": 0.641806876585969, + "grad_norm": 4.37681245803833, + "learning_rate": 5.693188178540794e-05, + "loss": 0.2986, + "step": 27379 + }, + { + "epoch": 0.6418303181607741, + "grad_norm": 3.0297904014587402, + "learning_rate": 5.6925233914333197e-05, + "loss": 0.2716, + "step": 27380 + }, + { + "epoch": 0.641853759735579, + "grad_norm": 3.194941520690918, + "learning_rate": 5.691858627698339e-05, + "loss": 0.2608, + "step": 27381 + }, + { + "epoch": 0.6418772013103841, + "grad_norm": 1.0336586236953735, + "learning_rate": 5.691193887339454e-05, + "loss": 0.099, + "step": 27382 + }, + { + "epoch": 0.641900642885189, + "grad_norm": 0.6217254400253296, + "learning_rate": 5.690529170360269e-05, + "loss": 0.0136, + "step": 27383 + }, + { + "epoch": 0.641924084459994, + "grad_norm": 0.4372245669364929, + "learning_rate": 5.689864476764397e-05, + "loss": 0.0263, + "step": 27384 + }, + { + "epoch": 0.641947526034799, + "grad_norm": 2.2791666984558105, + "learning_rate": 5.689199806555442e-05, + "loss": 0.2943, + "step": 27385 + }, + { + "epoch": 0.641970967609604, + "grad_norm": 4.679315567016602, + "learning_rate": 5.6885351597370096e-05, + "loss": 0.4135, + "step": 27386 + }, + { + "epoch": 0.641994409184409, + "grad_norm": 3.4258370399475098, + "learning_rate": 5.6878705363127024e-05, + "loss": 0.2779, + "step": 27387 + }, + { + "epoch": 0.642017850759214, + "grad_norm": 3.0920071601867676, + "learning_rate": 5.6872059362861374e-05, + "loss": 0.2034, + "step": 27388 + }, + { + "epoch": 0.642041292334019, + "grad_norm": 0.792696475982666, + "learning_rate": 5.686541359660911e-05, + "loss": 0.0513, + "step": 27389 + }, + { + "epoch": 0.642064733908824, + "grad_norm": 3.0982840061187744, + "learning_rate": 5.685876806440634e-05, + "loss": 0.3626, + "step": 27390 + }, + { + "epoch": 0.642088175483629, + "grad_norm": 0.38665324449539185, + "learning_rate": 5.685212276628911e-05, + "loss": 0.0333, + "step": 27391 + }, + { + "epoch": 0.642111617058434, + "grad_norm": 3.067831039428711, + "learning_rate": 5.684547770229347e-05, + "loss": 0.1859, + "step": 27392 + }, + { + "epoch": 0.6421350586332389, + "grad_norm": 0.9529029130935669, + "learning_rate": 5.683883287245544e-05, + "loss": 0.0927, + "step": 27393 + }, + { + "epoch": 0.642158500208044, + "grad_norm": 2.8722479343414307, + "learning_rate": 5.683218827681116e-05, + "loss": 0.3257, + "step": 27394 + }, + { + "epoch": 0.6421819417828489, + "grad_norm": 0.46102175116539, + "learning_rate": 5.682554391539664e-05, + "loss": 0.0572, + "step": 27395 + }, + { + "epoch": 0.642205383357654, + "grad_norm": 2.529539108276367, + "learning_rate": 5.6818899788247895e-05, + "loss": 0.1283, + "step": 27396 + }, + { + "epoch": 0.6422288249324589, + "grad_norm": 2.91634464263916, + "learning_rate": 5.6812255895401044e-05, + "loss": 0.2218, + "step": 27397 + }, + { + "epoch": 0.642252266507264, + "grad_norm": 2.7728161811828613, + "learning_rate": 5.680561223689211e-05, + "loss": 0.2929, + "step": 27398 + }, + { + "epoch": 0.6422757080820689, + "grad_norm": 4.579987049102783, + "learning_rate": 5.6798968812757144e-05, + "loss": 0.5495, + "step": 27399 + }, + { + "epoch": 0.642299149656874, + "grad_norm": 3.439997673034668, + "learning_rate": 5.679232562303215e-05, + "loss": 0.457, + "step": 27400 + }, + { + "epoch": 0.6423225912316789, + "grad_norm": 1.3120523691177368, + "learning_rate": 5.6785682667753245e-05, + "loss": 0.1123, + "step": 27401 + }, + { + "epoch": 0.642346032806484, + "grad_norm": 3.1085500717163086, + "learning_rate": 5.677903994695645e-05, + "loss": 0.2206, + "step": 27402 + }, + { + "epoch": 0.6423694743812889, + "grad_norm": 2.6555917263031006, + "learning_rate": 5.67723974606778e-05, + "loss": 0.3434, + "step": 27403 + }, + { + "epoch": 0.6423929159560939, + "grad_norm": 3.124145030975342, + "learning_rate": 5.676575520895332e-05, + "loss": 0.2911, + "step": 27404 + }, + { + "epoch": 0.6424163575308989, + "grad_norm": 3.404038667678833, + "learning_rate": 5.675911319181903e-05, + "loss": 0.2241, + "step": 27405 + }, + { + "epoch": 0.6424397991057039, + "grad_norm": 1.1577876806259155, + "learning_rate": 5.675247140931106e-05, + "loss": 0.104, + "step": 27406 + }, + { + "epoch": 0.6424632406805089, + "grad_norm": 3.337050199508667, + "learning_rate": 5.674582986146539e-05, + "loss": 0.2546, + "step": 27407 + }, + { + "epoch": 0.6424866822553139, + "grad_norm": 2.375917673110962, + "learning_rate": 5.673918854831803e-05, + "loss": 0.0967, + "step": 27408 + }, + { + "epoch": 0.642510123830119, + "grad_norm": 2.1396000385284424, + "learning_rate": 5.673254746990508e-05, + "loss": 0.1138, + "step": 27409 + }, + { + "epoch": 0.6425335654049239, + "grad_norm": 3.9053971767425537, + "learning_rate": 5.6725906626262546e-05, + "loss": 0.2258, + "step": 27410 + }, + { + "epoch": 0.642557006979729, + "grad_norm": 0.5038068890571594, + "learning_rate": 5.671926601742647e-05, + "loss": 0.032, + "step": 27411 + }, + { + "epoch": 0.6425804485545339, + "grad_norm": 3.2075257301330566, + "learning_rate": 5.671262564343283e-05, + "loss": 0.2332, + "step": 27412 + }, + { + "epoch": 0.6426038901293389, + "grad_norm": 3.126615285873413, + "learning_rate": 5.6705985504317726e-05, + "loss": 0.3384, + "step": 27413 + }, + { + "epoch": 0.6426273317041439, + "grad_norm": 2.2041728496551514, + "learning_rate": 5.669934560011717e-05, + "loss": 0.1718, + "step": 27414 + }, + { + "epoch": 0.6426507732789489, + "grad_norm": 0.8799924254417419, + "learning_rate": 5.669270593086719e-05, + "loss": 0.0956, + "step": 27415 + }, + { + "epoch": 0.6426742148537539, + "grad_norm": 2.191659450531006, + "learning_rate": 5.6686066496603795e-05, + "loss": 0.1419, + "step": 27416 + }, + { + "epoch": 0.6426976564285589, + "grad_norm": 0.2372211515903473, + "learning_rate": 5.667942729736302e-05, + "loss": 0.0107, + "step": 27417 + }, + { + "epoch": 0.6427210980033639, + "grad_norm": 1.9873172044754028, + "learning_rate": 5.667278833318086e-05, + "loss": 0.1147, + "step": 27418 + }, + { + "epoch": 0.6427445395781689, + "grad_norm": 3.150352954864502, + "learning_rate": 5.66661496040934e-05, + "loss": 0.2226, + "step": 27419 + }, + { + "epoch": 0.6427679811529738, + "grad_norm": 0.5435411334037781, + "learning_rate": 5.6659511110136596e-05, + "loss": 0.0568, + "step": 27420 + }, + { + "epoch": 0.6427914227277789, + "grad_norm": 3.1586570739746094, + "learning_rate": 5.665287285134655e-05, + "loss": 0.2581, + "step": 27421 + }, + { + "epoch": 0.6428148643025838, + "grad_norm": 0.31438979506492615, + "learning_rate": 5.664623482775923e-05, + "loss": 0.0214, + "step": 27422 + }, + { + "epoch": 0.6428383058773889, + "grad_norm": 0.5646962523460388, + "learning_rate": 5.663959703941065e-05, + "loss": 0.0226, + "step": 27423 + }, + { + "epoch": 0.6428617474521938, + "grad_norm": 2.8168036937713623, + "learning_rate": 5.663295948633685e-05, + "loss": 1.3884, + "step": 27424 + }, + { + "epoch": 0.6428851890269989, + "grad_norm": 0.789295494556427, + "learning_rate": 5.662632216857378e-05, + "loss": 0.0989, + "step": 27425 + }, + { + "epoch": 0.6429086306018038, + "grad_norm": 0.7827643156051636, + "learning_rate": 5.661968508615754e-05, + "loss": 0.0732, + "step": 27426 + }, + { + "epoch": 0.6429320721766089, + "grad_norm": 2.1243903636932373, + "learning_rate": 5.661304823912412e-05, + "loss": 0.1449, + "step": 27427 + }, + { + "epoch": 0.6429555137514138, + "grad_norm": 2.2560348510742188, + "learning_rate": 5.660641162750952e-05, + "loss": 0.1791, + "step": 27428 + }, + { + "epoch": 0.6429789553262188, + "grad_norm": 2.217752456665039, + "learning_rate": 5.659977525134974e-05, + "loss": 0.1652, + "step": 27429 + }, + { + "epoch": 0.6430023969010238, + "grad_norm": 1.8097584247589111, + "learning_rate": 5.659313911068082e-05, + "loss": 0.2298, + "step": 27430 + }, + { + "epoch": 0.6430258384758288, + "grad_norm": 0.4307430386543274, + "learning_rate": 5.658650320553869e-05, + "loss": 0.027, + "step": 27431 + }, + { + "epoch": 0.6430492800506338, + "grad_norm": 3.6276276111602783, + "learning_rate": 5.657986753595942e-05, + "loss": 1.918, + "step": 27432 + }, + { + "epoch": 0.6430727216254388, + "grad_norm": 4.073271751403809, + "learning_rate": 5.657323210197904e-05, + "loss": 0.3243, + "step": 27433 + }, + { + "epoch": 0.6430961632002438, + "grad_norm": 5.913954734802246, + "learning_rate": 5.656659690363354e-05, + "loss": 0.5634, + "step": 27434 + }, + { + "epoch": 0.6431196047750488, + "grad_norm": 1.9365932941436768, + "learning_rate": 5.655996194095889e-05, + "loss": 0.1151, + "step": 27435 + }, + { + "epoch": 0.6431430463498538, + "grad_norm": 4.937531471252441, + "learning_rate": 5.6553327213991124e-05, + "loss": 0.4347, + "step": 27436 + }, + { + "epoch": 0.6431664879246588, + "grad_norm": 1.393210768699646, + "learning_rate": 5.6546692722766224e-05, + "loss": 0.1231, + "step": 27437 + }, + { + "epoch": 0.6431899294994637, + "grad_norm": 0.37112435698509216, + "learning_rate": 5.654005846732015e-05, + "loss": 0.0259, + "step": 27438 + }, + { + "epoch": 0.6432133710742688, + "grad_norm": 3.112492561340332, + "learning_rate": 5.6533424447688967e-05, + "loss": 0.2443, + "step": 27439 + }, + { + "epoch": 0.6432368126490737, + "grad_norm": 2.0677318572998047, + "learning_rate": 5.652679066390866e-05, + "loss": 0.2542, + "step": 27440 + }, + { + "epoch": 0.6432602542238788, + "grad_norm": 3.1665735244750977, + "learning_rate": 5.6520157116015206e-05, + "loss": 0.2487, + "step": 27441 + }, + { + "epoch": 0.6432836957986837, + "grad_norm": 1.7406195402145386, + "learning_rate": 5.65135238040446e-05, + "loss": 0.1892, + "step": 27442 + }, + { + "epoch": 0.6433071373734888, + "grad_norm": 2.786315679550171, + "learning_rate": 5.650689072803279e-05, + "loss": 0.2456, + "step": 27443 + }, + { + "epoch": 0.6433305789482937, + "grad_norm": 3.3650200366973877, + "learning_rate": 5.650025788801587e-05, + "loss": 0.2993, + "step": 27444 + }, + { + "epoch": 0.6433540205230988, + "grad_norm": 1.4958323240280151, + "learning_rate": 5.649362528402973e-05, + "loss": 0.0857, + "step": 27445 + }, + { + "epoch": 0.6433774620979037, + "grad_norm": 3.32202410697937, + "learning_rate": 5.648699291611042e-05, + "loss": 0.4454, + "step": 27446 + }, + { + "epoch": 0.6434009036727087, + "grad_norm": 1.4743467569351196, + "learning_rate": 5.648036078429393e-05, + "loss": 0.0568, + "step": 27447 + }, + { + "epoch": 0.6434243452475137, + "grad_norm": 3.436363458633423, + "learning_rate": 5.647372888861623e-05, + "loss": 2.0072, + "step": 27448 + }, + { + "epoch": 0.6434477868223187, + "grad_norm": 2.3356025218963623, + "learning_rate": 5.64670972291133e-05, + "loss": 0.2258, + "step": 27449 + }, + { + "epoch": 0.6434712283971237, + "grad_norm": 5.782802104949951, + "learning_rate": 5.646046580582107e-05, + "loss": 0.3282, + "step": 27450 + }, + { + "epoch": 0.6434946699719287, + "grad_norm": 0.7680416703224182, + "learning_rate": 5.6453834618775624e-05, + "loss": 0.0835, + "step": 27451 + }, + { + "epoch": 0.6435181115467337, + "grad_norm": 3.8128700256347656, + "learning_rate": 5.6447203668012905e-05, + "loss": 0.2807, + "step": 27452 + }, + { + "epoch": 0.6435415531215387, + "grad_norm": 2.5027084350585938, + "learning_rate": 5.644057295356887e-05, + "loss": 0.1667, + "step": 27453 + }, + { + "epoch": 0.6435649946963437, + "grad_norm": 1.6871609687805176, + "learning_rate": 5.6433942475479516e-05, + "loss": 1.7389, + "step": 27454 + }, + { + "epoch": 0.6435884362711487, + "grad_norm": 1.7115700244903564, + "learning_rate": 5.6427312233780816e-05, + "loss": 0.7819, + "step": 27455 + }, + { + "epoch": 0.6436118778459536, + "grad_norm": 3.9210784435272217, + "learning_rate": 5.6420682228508704e-05, + "loss": 0.6886, + "step": 27456 + }, + { + "epoch": 0.6436353194207587, + "grad_norm": 5.174879550933838, + "learning_rate": 5.6414052459699195e-05, + "loss": 0.2194, + "step": 27457 + }, + { + "epoch": 0.6436587609955636, + "grad_norm": 0.4853108525276184, + "learning_rate": 5.64074229273883e-05, + "loss": 0.0462, + "step": 27458 + }, + { + "epoch": 0.6436822025703687, + "grad_norm": 3.755399703979492, + "learning_rate": 5.640079363161197e-05, + "loss": 0.3222, + "step": 27459 + }, + { + "epoch": 0.6437056441451737, + "grad_norm": 1.9213297367095947, + "learning_rate": 5.639416457240614e-05, + "loss": 0.1422, + "step": 27460 + }, + { + "epoch": 0.6437290857199787, + "grad_norm": 2.497758388519287, + "learning_rate": 5.638753574980681e-05, + "loss": 0.2466, + "step": 27461 + }, + { + "epoch": 0.6437525272947837, + "grad_norm": 1.614453673362732, + "learning_rate": 5.638090716384994e-05, + "loss": 0.1337, + "step": 27462 + }, + { + "epoch": 0.6437759688695887, + "grad_norm": 1.7102383375167847, + "learning_rate": 5.6374278814571446e-05, + "loss": 1.2343, + "step": 27463 + }, + { + "epoch": 0.6437994104443937, + "grad_norm": 2.1303298473358154, + "learning_rate": 5.636765070200738e-05, + "loss": 0.1511, + "step": 27464 + }, + { + "epoch": 0.6438228520191986, + "grad_norm": 0.7336826920509338, + "learning_rate": 5.6361022826193665e-05, + "loss": 0.0687, + "step": 27465 + }, + { + "epoch": 0.6438462935940037, + "grad_norm": 2.8969502449035645, + "learning_rate": 5.635439518716628e-05, + "loss": 0.1212, + "step": 27466 + }, + { + "epoch": 0.6438697351688086, + "grad_norm": 0.5626105070114136, + "learning_rate": 5.634776778496117e-05, + "loss": 0.0706, + "step": 27467 + }, + { + "epoch": 0.6438931767436137, + "grad_norm": 2.8383967876434326, + "learning_rate": 5.6341140619614295e-05, + "loss": 0.2609, + "step": 27468 + }, + { + "epoch": 0.6439166183184186, + "grad_norm": 2.3290555477142334, + "learning_rate": 5.633451369116158e-05, + "loss": 1.5106, + "step": 27469 + }, + { + "epoch": 0.6439400598932237, + "grad_norm": 0.8312028050422668, + "learning_rate": 5.6327886999639026e-05, + "loss": 0.0668, + "step": 27470 + }, + { + "epoch": 0.6439635014680286, + "grad_norm": 2.414566993713379, + "learning_rate": 5.632126054508262e-05, + "loss": 0.1529, + "step": 27471 + }, + { + "epoch": 0.6439869430428337, + "grad_norm": 2.1805906295776367, + "learning_rate": 5.631463432752828e-05, + "loss": 0.1344, + "step": 27472 + }, + { + "epoch": 0.6440103846176386, + "grad_norm": 3.2653353214263916, + "learning_rate": 5.6308008347011976e-05, + "loss": 0.196, + "step": 27473 + }, + { + "epoch": 0.6440338261924436, + "grad_norm": 0.9965779781341553, + "learning_rate": 5.630138260356963e-05, + "loss": 0.5346, + "step": 27474 + }, + { + "epoch": 0.6440572677672486, + "grad_norm": 2.054525375366211, + "learning_rate": 5.629475709723722e-05, + "loss": 0.2068, + "step": 27475 + }, + { + "epoch": 0.6440807093420536, + "grad_norm": 0.5522551536560059, + "learning_rate": 5.628813182805064e-05, + "loss": 0.044, + "step": 27476 + }, + { + "epoch": 0.6441041509168586, + "grad_norm": 1.199280023574829, + "learning_rate": 5.628150679604594e-05, + "loss": 0.0791, + "step": 27477 + }, + { + "epoch": 0.6441275924916636, + "grad_norm": 1.238887071609497, + "learning_rate": 5.627488200125901e-05, + "loss": 0.0523, + "step": 27478 + }, + { + "epoch": 0.6441510340664686, + "grad_norm": 0.4872601628303528, + "learning_rate": 5.626825744372579e-05, + "loss": 0.0191, + "step": 27479 + }, + { + "epoch": 0.6441744756412736, + "grad_norm": 3.460315704345703, + "learning_rate": 5.626163312348224e-05, + "loss": 0.3264, + "step": 27480 + }, + { + "epoch": 0.6441979172160786, + "grad_norm": 2.562549114227295, + "learning_rate": 5.625500904056425e-05, + "loss": 0.3228, + "step": 27481 + }, + { + "epoch": 0.6442213587908836, + "grad_norm": 0.5113455057144165, + "learning_rate": 5.6248385195007856e-05, + "loss": 0.0269, + "step": 27482 + }, + { + "epoch": 0.6442448003656885, + "grad_norm": 3.4530141353607178, + "learning_rate": 5.624176158684893e-05, + "loss": 0.2961, + "step": 27483 + }, + { + "epoch": 0.6442682419404936, + "grad_norm": 3.1535706520080566, + "learning_rate": 5.623513821612345e-05, + "loss": 0.4275, + "step": 27484 + }, + { + "epoch": 0.6442916835152985, + "grad_norm": 2.354489326477051, + "learning_rate": 5.622851508286735e-05, + "loss": 0.1582, + "step": 27485 + }, + { + "epoch": 0.6443151250901036, + "grad_norm": 3.1331138610839844, + "learning_rate": 5.6221892187116574e-05, + "loss": 0.2075, + "step": 27486 + }, + { + "epoch": 0.6443385666649085, + "grad_norm": 4.660004615783691, + "learning_rate": 5.621526952890702e-05, + "loss": 0.2916, + "step": 27487 + }, + { + "epoch": 0.6443620082397136, + "grad_norm": 2.7093310356140137, + "learning_rate": 5.620864710827463e-05, + "loss": 0.1841, + "step": 27488 + }, + { + "epoch": 0.6443854498145185, + "grad_norm": 2.0980563163757324, + "learning_rate": 5.620202492525537e-05, + "loss": 0.1256, + "step": 27489 + }, + { + "epoch": 0.6444088913893236, + "grad_norm": 3.432650566101074, + "learning_rate": 5.6195402979885145e-05, + "loss": 0.3473, + "step": 27490 + }, + { + "epoch": 0.6444323329641285, + "grad_norm": 2.6504015922546387, + "learning_rate": 5.618878127219992e-05, + "loss": 1.2776, + "step": 27491 + }, + { + "epoch": 0.6444557745389335, + "grad_norm": 4.303489685058594, + "learning_rate": 5.618215980223559e-05, + "loss": 0.3695, + "step": 27492 + }, + { + "epoch": 0.6444792161137385, + "grad_norm": 4.538620471954346, + "learning_rate": 5.6175538570028056e-05, + "loss": 0.4014, + "step": 27493 + }, + { + "epoch": 0.6445026576885435, + "grad_norm": 2.6856987476348877, + "learning_rate": 5.616891757561331e-05, + "loss": 0.2668, + "step": 27494 + }, + { + "epoch": 0.6445260992633485, + "grad_norm": 2.329894542694092, + "learning_rate": 5.6162296819027224e-05, + "loss": 0.2011, + "step": 27495 + }, + { + "epoch": 0.6445495408381535, + "grad_norm": 2.7507741451263428, + "learning_rate": 5.615567630030578e-05, + "loss": 0.1983, + "step": 27496 + }, + { + "epoch": 0.6445729824129585, + "grad_norm": 3.606792688369751, + "learning_rate": 5.614905601948488e-05, + "loss": 0.2894, + "step": 27497 + }, + { + "epoch": 0.6445964239877635, + "grad_norm": 0.5260579586029053, + "learning_rate": 5.6142435976600415e-05, + "loss": 0.0473, + "step": 27498 + }, + { + "epoch": 0.6446198655625685, + "grad_norm": 4.48136568069458, + "learning_rate": 5.613581617168835e-05, + "loss": 0.1255, + "step": 27499 + }, + { + "epoch": 0.6446433071373735, + "grad_norm": 2.4511055946350098, + "learning_rate": 5.6129196604784554e-05, + "loss": 0.2196, + "step": 27500 + }, + { + "epoch": 0.6446667487121784, + "grad_norm": 3.295941114425659, + "learning_rate": 5.612257727592495e-05, + "loss": 0.2157, + "step": 27501 + }, + { + "epoch": 0.6446901902869835, + "grad_norm": 2.5273070335388184, + "learning_rate": 5.61159581851455e-05, + "loss": 0.0858, + "step": 27502 + }, + { + "epoch": 0.6447136318617884, + "grad_norm": 4.380373001098633, + "learning_rate": 5.6109339332482104e-05, + "loss": 0.3371, + "step": 27503 + }, + { + "epoch": 0.6447370734365935, + "grad_norm": 1.3572994470596313, + "learning_rate": 5.610272071797067e-05, + "loss": 0.1282, + "step": 27504 + }, + { + "epoch": 0.6447605150113984, + "grad_norm": 4.543000221252441, + "learning_rate": 5.609610234164706e-05, + "loss": 0.5493, + "step": 27505 + }, + { + "epoch": 0.6447839565862035, + "grad_norm": 1.9776690006256104, + "learning_rate": 5.608948420354729e-05, + "loss": 0.1341, + "step": 27506 + }, + { + "epoch": 0.6448073981610084, + "grad_norm": 2.968245506286621, + "learning_rate": 5.6082866303707206e-05, + "loss": 0.2693, + "step": 27507 + }, + { + "epoch": 0.6448308397358135, + "grad_norm": 2.738070249557495, + "learning_rate": 5.6076248642162675e-05, + "loss": 0.1355, + "step": 27508 + }, + { + "epoch": 0.6448542813106184, + "grad_norm": 1.339901089668274, + "learning_rate": 5.606963121894969e-05, + "loss": 0.0659, + "step": 27509 + }, + { + "epoch": 0.6448777228854234, + "grad_norm": 1.4912326335906982, + "learning_rate": 5.6063014034104135e-05, + "loss": 0.0863, + "step": 27510 + }, + { + "epoch": 0.6449011644602285, + "grad_norm": 4.007905006408691, + "learning_rate": 5.605639708766189e-05, + "loss": 0.28, + "step": 27511 + }, + { + "epoch": 0.6449246060350334, + "grad_norm": 0.8243100643157959, + "learning_rate": 5.604978037965888e-05, + "loss": 0.0654, + "step": 27512 + }, + { + "epoch": 0.6449480476098385, + "grad_norm": 2.3553614616394043, + "learning_rate": 5.604316391013099e-05, + "loss": 1.6566, + "step": 27513 + }, + { + "epoch": 0.6449714891846434, + "grad_norm": 3.21850323677063, + "learning_rate": 5.603654767911409e-05, + "loss": 0.3082, + "step": 27514 + }, + { + "epoch": 0.6449949307594485, + "grad_norm": 1.0345045328140259, + "learning_rate": 5.6029931686644164e-05, + "loss": 0.0729, + "step": 27515 + }, + { + "epoch": 0.6450183723342534, + "grad_norm": 3.9260082244873047, + "learning_rate": 5.6023315932757056e-05, + "loss": 0.4478, + "step": 27516 + }, + { + "epoch": 0.6450418139090585, + "grad_norm": 2.91672682762146, + "learning_rate": 5.6016700417488643e-05, + "loss": 0.2581, + "step": 27517 + }, + { + "epoch": 0.6450652554838634, + "grad_norm": 2.7213759422302246, + "learning_rate": 5.601008514087489e-05, + "loss": 0.3762, + "step": 27518 + }, + { + "epoch": 0.6450886970586684, + "grad_norm": 2.357789993286133, + "learning_rate": 5.600347010295164e-05, + "loss": 0.2774, + "step": 27519 + }, + { + "epoch": 0.6451121386334734, + "grad_norm": 3.094254493713379, + "learning_rate": 5.59968553037548e-05, + "loss": 0.3135, + "step": 27520 + }, + { + "epoch": 0.6451355802082784, + "grad_norm": 4.356751441955566, + "learning_rate": 5.5990240743320235e-05, + "loss": 0.3372, + "step": 27521 + }, + { + "epoch": 0.6451590217830834, + "grad_norm": 1.601650595664978, + "learning_rate": 5.598362642168389e-05, + "loss": 0.1865, + "step": 27522 + }, + { + "epoch": 0.6451824633578884, + "grad_norm": 4.343074798583984, + "learning_rate": 5.5977012338881615e-05, + "loss": 0.2556, + "step": 27523 + }, + { + "epoch": 0.6452059049326934, + "grad_norm": 2.802900552749634, + "learning_rate": 5.597039849494933e-05, + "loss": 0.2243, + "step": 27524 + }, + { + "epoch": 0.6452293465074984, + "grad_norm": 3.2704720497131348, + "learning_rate": 5.5963784889922885e-05, + "loss": 0.2402, + "step": 27525 + }, + { + "epoch": 0.6452527880823034, + "grad_norm": 1.3024548292160034, + "learning_rate": 5.595717152383815e-05, + "loss": 0.0752, + "step": 27526 + }, + { + "epoch": 0.6452762296571084, + "grad_norm": 4.044860363006592, + "learning_rate": 5.595055839673109e-05, + "loss": 0.318, + "step": 27527 + }, + { + "epoch": 0.6452996712319133, + "grad_norm": 0.5102373957633972, + "learning_rate": 5.594394550863752e-05, + "loss": 0.0323, + "step": 27528 + }, + { + "epoch": 0.6453231128067184, + "grad_norm": 0.9286488890647888, + "learning_rate": 5.5937332859593306e-05, + "loss": 0.0358, + "step": 27529 + }, + { + "epoch": 0.6453465543815233, + "grad_norm": 4.278995037078857, + "learning_rate": 5.5930720449634414e-05, + "loss": 0.4386, + "step": 27530 + }, + { + "epoch": 0.6453699959563284, + "grad_norm": 1.895589828491211, + "learning_rate": 5.5924108278796663e-05, + "loss": 0.2, + "step": 27531 + }, + { + "epoch": 0.6453934375311333, + "grad_norm": 2.543091297149658, + "learning_rate": 5.591749634711594e-05, + "loss": 0.2469, + "step": 27532 + }, + { + "epoch": 0.6454168791059384, + "grad_norm": 3.7553887367248535, + "learning_rate": 5.591088465462807e-05, + "loss": 0.2319, + "step": 27533 + }, + { + "epoch": 0.6454403206807433, + "grad_norm": 2.8901350498199463, + "learning_rate": 5.590427320136904e-05, + "loss": 0.2094, + "step": 27534 + }, + { + "epoch": 0.6454637622555484, + "grad_norm": 6.593051910400391, + "learning_rate": 5.589766198737465e-05, + "loss": 0.6286, + "step": 27535 + }, + { + "epoch": 0.6454872038303533, + "grad_norm": 1.2240999937057495, + "learning_rate": 5.58910510126808e-05, + "loss": 0.1339, + "step": 27536 + }, + { + "epoch": 0.6455106454051583, + "grad_norm": 0.7226303219795227, + "learning_rate": 5.588444027732332e-05, + "loss": 0.0602, + "step": 27537 + }, + { + "epoch": 0.6455340869799633, + "grad_norm": 1.2189854383468628, + "learning_rate": 5.587782978133813e-05, + "loss": 0.1047, + "step": 27538 + }, + { + "epoch": 0.6455575285547683, + "grad_norm": 1.2834789752960205, + "learning_rate": 5.5871219524761034e-05, + "loss": 0.0827, + "step": 27539 + }, + { + "epoch": 0.6455809701295733, + "grad_norm": 1.1513683795928955, + "learning_rate": 5.5864609507627975e-05, + "loss": 0.0979, + "step": 27540 + }, + { + "epoch": 0.6456044117043783, + "grad_norm": 2.6672651767730713, + "learning_rate": 5.585799972997474e-05, + "loss": 0.1166, + "step": 27541 + }, + { + "epoch": 0.6456278532791833, + "grad_norm": 4.328014373779297, + "learning_rate": 5.5851390191837294e-05, + "loss": 0.3269, + "step": 27542 + }, + { + "epoch": 0.6456512948539883, + "grad_norm": 3.3693504333496094, + "learning_rate": 5.5844780893251445e-05, + "loss": 0.2547, + "step": 27543 + }, + { + "epoch": 0.6456747364287933, + "grad_norm": 1.957624077796936, + "learning_rate": 5.583817183425306e-05, + "loss": 0.0849, + "step": 27544 + }, + { + "epoch": 0.6456981780035983, + "grad_norm": 4.037158012390137, + "learning_rate": 5.583156301487797e-05, + "loss": 1.5112, + "step": 27545 + }, + { + "epoch": 0.6457216195784032, + "grad_norm": 2.90803599357605, + "learning_rate": 5.582495443516205e-05, + "loss": 0.251, + "step": 27546 + }, + { + "epoch": 0.6457450611532083, + "grad_norm": 0.6890577673912048, + "learning_rate": 5.581834609514121e-05, + "loss": 0.0545, + "step": 27547 + }, + { + "epoch": 0.6457685027280132, + "grad_norm": 2.4734907150268555, + "learning_rate": 5.581173799485124e-05, + "loss": 0.2047, + "step": 27548 + }, + { + "epoch": 0.6457919443028183, + "grad_norm": 3.293217420578003, + "learning_rate": 5.580513013432804e-05, + "loss": 0.2802, + "step": 27549 + }, + { + "epoch": 0.6458153858776232, + "grad_norm": 0.7182093262672424, + "learning_rate": 5.5798522513607445e-05, + "loss": 0.0648, + "step": 27550 + }, + { + "epoch": 0.6458388274524283, + "grad_norm": 0.6477944254875183, + "learning_rate": 5.579191513272532e-05, + "loss": 0.0338, + "step": 27551 + }, + { + "epoch": 0.6458622690272332, + "grad_norm": 7.02012825012207, + "learning_rate": 5.578530799171745e-05, + "loss": 0.8203, + "step": 27552 + }, + { + "epoch": 0.6458857106020383, + "grad_norm": 4.203930377960205, + "learning_rate": 5.577870109061978e-05, + "loss": 0.4332, + "step": 27553 + }, + { + "epoch": 0.6459091521768432, + "grad_norm": 1.7548061609268188, + "learning_rate": 5.577209442946809e-05, + "loss": 0.1342, + "step": 27554 + }, + { + "epoch": 0.6459325937516482, + "grad_norm": 3.0578179359436035, + "learning_rate": 5.576548800829831e-05, + "loss": 0.4256, + "step": 27555 + }, + { + "epoch": 0.6459560353264532, + "grad_norm": 1.0509746074676514, + "learning_rate": 5.575888182714621e-05, + "loss": 0.0719, + "step": 27556 + }, + { + "epoch": 0.6459794769012582, + "grad_norm": 3.2434372901916504, + "learning_rate": 5.575227588604768e-05, + "loss": 1.8786, + "step": 27557 + }, + { + "epoch": 0.6460029184760632, + "grad_norm": 2.7087488174438477, + "learning_rate": 5.574567018503854e-05, + "loss": 0.2626, + "step": 27558 + }, + { + "epoch": 0.6460263600508682, + "grad_norm": 2.9710912704467773, + "learning_rate": 5.57390647241546e-05, + "loss": 0.2352, + "step": 27559 + }, + { + "epoch": 0.6460498016256732, + "grad_norm": 3.034114122390747, + "learning_rate": 5.573245950343178e-05, + "loss": 0.2248, + "step": 27560 + }, + { + "epoch": 0.6460732432004782, + "grad_norm": 2.4107229709625244, + "learning_rate": 5.572585452290587e-05, + "loss": 0.1356, + "step": 27561 + }, + { + "epoch": 0.6460966847752833, + "grad_norm": 1.6651726961135864, + "learning_rate": 5.571924978261274e-05, + "loss": 1.4344, + "step": 27562 + }, + { + "epoch": 0.6461201263500882, + "grad_norm": 3.0693721771240234, + "learning_rate": 5.5712645282588184e-05, + "loss": 0.3754, + "step": 27563 + }, + { + "epoch": 0.6461435679248932, + "grad_norm": 1.0452359914779663, + "learning_rate": 5.570604102286803e-05, + "loss": 0.0897, + "step": 27564 + }, + { + "epoch": 0.6461670094996982, + "grad_norm": 3.1979875564575195, + "learning_rate": 5.5699437003488175e-05, + "loss": 0.232, + "step": 27565 + }, + { + "epoch": 0.6461904510745032, + "grad_norm": 2.5186474323272705, + "learning_rate": 5.569283322448438e-05, + "loss": 1.0241, + "step": 27566 + }, + { + "epoch": 0.6462138926493082, + "grad_norm": 1.815561294555664, + "learning_rate": 5.568622968589257e-05, + "loss": 0.1029, + "step": 27567 + }, + { + "epoch": 0.6462373342241132, + "grad_norm": 0.8079357147216797, + "learning_rate": 5.567962638774851e-05, + "loss": 0.0628, + "step": 27568 + }, + { + "epoch": 0.6462607757989182, + "grad_norm": 1.2034120559692383, + "learning_rate": 5.567302333008806e-05, + "loss": 0.0826, + "step": 27569 + }, + { + "epoch": 0.6462842173737232, + "grad_norm": 6.305290222167969, + "learning_rate": 5.566642051294701e-05, + "loss": 0.2905, + "step": 27570 + }, + { + "epoch": 0.6463076589485282, + "grad_norm": 3.9591434001922607, + "learning_rate": 5.565981793636119e-05, + "loss": 0.3469, + "step": 27571 + }, + { + "epoch": 0.6463311005233332, + "grad_norm": 2.6258962154388428, + "learning_rate": 5.565321560036647e-05, + "loss": 0.1914, + "step": 27572 + }, + { + "epoch": 0.6463545420981381, + "grad_norm": 0.8930313587188721, + "learning_rate": 5.564661350499864e-05, + "loss": 0.0547, + "step": 27573 + }, + { + "epoch": 0.6463779836729432, + "grad_norm": 2.454935073852539, + "learning_rate": 5.564001165029355e-05, + "loss": 0.2349, + "step": 27574 + }, + { + "epoch": 0.6464014252477481, + "grad_norm": 0.9909329414367676, + "learning_rate": 5.5633410036287016e-05, + "loss": 0.1009, + "step": 27575 + }, + { + "epoch": 0.6464248668225532, + "grad_norm": 5.971994400024414, + "learning_rate": 5.562680866301482e-05, + "loss": 0.3261, + "step": 27576 + }, + { + "epoch": 0.6464483083973581, + "grad_norm": 1.572942852973938, + "learning_rate": 5.562020753051278e-05, + "loss": 0.8345, + "step": 27577 + }, + { + "epoch": 0.6464717499721632, + "grad_norm": 2.0071890354156494, + "learning_rate": 5.561360663881674e-05, + "loss": 0.189, + "step": 27578 + }, + { + "epoch": 0.6464951915469681, + "grad_norm": 3.6133642196655273, + "learning_rate": 5.560700598796256e-05, + "loss": 0.1439, + "step": 27579 + }, + { + "epoch": 0.6465186331217732, + "grad_norm": 4.915867328643799, + "learning_rate": 5.560040557798603e-05, + "loss": 0.5359, + "step": 27580 + }, + { + "epoch": 0.6465420746965781, + "grad_norm": 3.349066734313965, + "learning_rate": 5.559380540892292e-05, + "loss": 0.4054, + "step": 27581 + }, + { + "epoch": 0.6465655162713831, + "grad_norm": 1.9804140329360962, + "learning_rate": 5.558720548080908e-05, + "loss": 0.109, + "step": 27582 + }, + { + "epoch": 0.6465889578461881, + "grad_norm": 2.1987757682800293, + "learning_rate": 5.5580605793680315e-05, + "loss": 0.1522, + "step": 27583 + }, + { + "epoch": 0.6466123994209931, + "grad_norm": 0.8666771650314331, + "learning_rate": 5.557400634757239e-05, + "loss": 0.0979, + "step": 27584 + }, + { + "epoch": 0.6466358409957981, + "grad_norm": 5.743947505950928, + "learning_rate": 5.55674071425212e-05, + "loss": 0.5732, + "step": 27585 + }, + { + "epoch": 0.6466592825706031, + "grad_norm": 2.3521642684936523, + "learning_rate": 5.5560808178562494e-05, + "loss": 0.1846, + "step": 27586 + }, + { + "epoch": 0.6466827241454081, + "grad_norm": 4.585079193115234, + "learning_rate": 5.555420945573211e-05, + "loss": 0.5229, + "step": 27587 + }, + { + "epoch": 0.6467061657202131, + "grad_norm": 0.40330177545547485, + "learning_rate": 5.554761097406582e-05, + "loss": 0.0399, + "step": 27588 + }, + { + "epoch": 0.646729607295018, + "grad_norm": 2.226419687271118, + "learning_rate": 5.554101273359945e-05, + "loss": 1.265, + "step": 27589 + }, + { + "epoch": 0.6467530488698231, + "grad_norm": 0.66065913438797, + "learning_rate": 5.5534414734368756e-05, + "loss": 0.0712, + "step": 27590 + }, + { + "epoch": 0.646776490444628, + "grad_norm": 4.437702655792236, + "learning_rate": 5.5527816976409584e-05, + "loss": 0.2788, + "step": 27591 + }, + { + "epoch": 0.6467999320194331, + "grad_norm": 1.3663465976715088, + "learning_rate": 5.552121945975776e-05, + "loss": 0.2086, + "step": 27592 + }, + { + "epoch": 0.646823373594238, + "grad_norm": 1.445682406425476, + "learning_rate": 5.551462218444905e-05, + "loss": 0.0921, + "step": 27593 + }, + { + "epoch": 0.6468468151690431, + "grad_norm": 4.314206123352051, + "learning_rate": 5.550802515051925e-05, + "loss": 0.4405, + "step": 27594 + }, + { + "epoch": 0.646870256743848, + "grad_norm": 2.6640026569366455, + "learning_rate": 5.550142835800416e-05, + "loss": 0.2315, + "step": 27595 + }, + { + "epoch": 0.6468936983186531, + "grad_norm": 2.5854690074920654, + "learning_rate": 5.5494831806939565e-05, + "loss": 0.1348, + "step": 27596 + }, + { + "epoch": 0.646917139893458, + "grad_norm": 1.2916476726531982, + "learning_rate": 5.548823549736122e-05, + "loss": 0.1218, + "step": 27597 + }, + { + "epoch": 0.646940581468263, + "grad_norm": 3.8341166973114014, + "learning_rate": 5.548163942930501e-05, + "loss": 0.4048, + "step": 27598 + }, + { + "epoch": 0.646964023043068, + "grad_norm": 2.2228076457977295, + "learning_rate": 5.5475043602806675e-05, + "loss": 0.1995, + "step": 27599 + }, + { + "epoch": 0.646987464617873, + "grad_norm": 3.891077756881714, + "learning_rate": 5.546844801790201e-05, + "loss": 0.2626, + "step": 27600 + }, + { + "epoch": 0.647010906192678, + "grad_norm": 3.2903635501861572, + "learning_rate": 5.546185267462679e-05, + "loss": 0.2788, + "step": 27601 + }, + { + "epoch": 0.647034347767483, + "grad_norm": 2.5659992694854736, + "learning_rate": 5.545525757301677e-05, + "loss": 0.2189, + "step": 27602 + }, + { + "epoch": 0.647057789342288, + "grad_norm": 2.3200032711029053, + "learning_rate": 5.544866271310782e-05, + "loss": 0.1472, + "step": 27603 + }, + { + "epoch": 0.647081230917093, + "grad_norm": 1.6554789543151855, + "learning_rate": 5.544206809493564e-05, + "loss": 0.1623, + "step": 27604 + }, + { + "epoch": 0.647104672491898, + "grad_norm": 2.1947500705718994, + "learning_rate": 5.543547371853608e-05, + "loss": 0.0857, + "step": 27605 + }, + { + "epoch": 0.647128114066703, + "grad_norm": 1.4672906398773193, + "learning_rate": 5.54288795839449e-05, + "loss": 0.15, + "step": 27606 + }, + { + "epoch": 0.647151555641508, + "grad_norm": 1.3431274890899658, + "learning_rate": 5.542228569119787e-05, + "loss": 0.153, + "step": 27607 + }, + { + "epoch": 0.647174997216313, + "grad_norm": 1.174765706062317, + "learning_rate": 5.541569204033079e-05, + "loss": 0.1016, + "step": 27608 + }, + { + "epoch": 0.6471984387911179, + "grad_norm": 1.6145070791244507, + "learning_rate": 5.5409098631379354e-05, + "loss": 0.2314, + "step": 27609 + }, + { + "epoch": 0.647221880365923, + "grad_norm": 0.691375732421875, + "learning_rate": 5.540250546437946e-05, + "loss": 0.0225, + "step": 27610 + }, + { + "epoch": 0.6472453219407279, + "grad_norm": 4.702880382537842, + "learning_rate": 5.539591253936681e-05, + "loss": 0.3073, + "step": 27611 + }, + { + "epoch": 0.647268763515533, + "grad_norm": 3.1055800914764404, + "learning_rate": 5.53893198563772e-05, + "loss": 0.1958, + "step": 27612 + }, + { + "epoch": 0.647292205090338, + "grad_norm": 3.145362138748169, + "learning_rate": 5.538272741544639e-05, + "loss": 0.3584, + "step": 27613 + }, + { + "epoch": 0.647315646665143, + "grad_norm": 3.2540509700775146, + "learning_rate": 5.5376135216610125e-05, + "loss": 0.2576, + "step": 27614 + }, + { + "epoch": 0.647339088239948, + "grad_norm": 1.2787832021713257, + "learning_rate": 5.5369543259904243e-05, + "loss": 0.1339, + "step": 27615 + }, + { + "epoch": 0.647362529814753, + "grad_norm": 0.8577843308448792, + "learning_rate": 5.536295154536443e-05, + "loss": 0.0919, + "step": 27616 + }, + { + "epoch": 0.647385971389558, + "grad_norm": 1.9622458219528198, + "learning_rate": 5.5356360073026534e-05, + "loss": 0.1438, + "step": 27617 + }, + { + "epoch": 0.6474094129643629, + "grad_norm": 0.9836806058883667, + "learning_rate": 5.534976884292628e-05, + "loss": 0.0828, + "step": 27618 + }, + { + "epoch": 0.647432854539168, + "grad_norm": 4.054267883300781, + "learning_rate": 5.534317785509946e-05, + "loss": 0.3875, + "step": 27619 + }, + { + "epoch": 0.6474562961139729, + "grad_norm": 3.3291335105895996, + "learning_rate": 5.533658710958178e-05, + "loss": 1.5953, + "step": 27620 + }, + { + "epoch": 0.647479737688778, + "grad_norm": 3.166949987411499, + "learning_rate": 5.532999660640905e-05, + "loss": 0.2631, + "step": 27621 + }, + { + "epoch": 0.6475031792635829, + "grad_norm": 3.0780208110809326, + "learning_rate": 5.532340634561699e-05, + "loss": 0.2409, + "step": 27622 + }, + { + "epoch": 0.647526620838388, + "grad_norm": 0.7252013683319092, + "learning_rate": 5.53168163272414e-05, + "loss": 0.0855, + "step": 27623 + }, + { + "epoch": 0.6475500624131929, + "grad_norm": 1.3276575803756714, + "learning_rate": 5.5310226551318033e-05, + "loss": 0.0818, + "step": 27624 + }, + { + "epoch": 0.647573503987998, + "grad_norm": 3.983412027359009, + "learning_rate": 5.5303637017882625e-05, + "loss": 0.2426, + "step": 27625 + }, + { + "epoch": 0.6475969455628029, + "grad_norm": 0.6436700224876404, + "learning_rate": 5.529704772697095e-05, + "loss": 0.1428, + "step": 27626 + }, + { + "epoch": 0.647620387137608, + "grad_norm": 5.256618499755859, + "learning_rate": 5.529045867861871e-05, + "loss": 0.3439, + "step": 27627 + }, + { + "epoch": 0.6476438287124129, + "grad_norm": 2.4305942058563232, + "learning_rate": 5.5283869872861716e-05, + "loss": 1.4838, + "step": 27628 + }, + { + "epoch": 0.6476672702872179, + "grad_norm": 1.1554838418960571, + "learning_rate": 5.5277281309735685e-05, + "loss": 0.8808, + "step": 27629 + }, + { + "epoch": 0.6476907118620229, + "grad_norm": 4.273703575134277, + "learning_rate": 5.527069298927642e-05, + "loss": 0.2507, + "step": 27630 + }, + { + "epoch": 0.6477141534368279, + "grad_norm": 0.9295450448989868, + "learning_rate": 5.5264104911519635e-05, + "loss": 0.0814, + "step": 27631 + }, + { + "epoch": 0.6477375950116329, + "grad_norm": 2.5844554901123047, + "learning_rate": 5.5257517076501065e-05, + "loss": 0.1038, + "step": 27632 + }, + { + "epoch": 0.6477610365864379, + "grad_norm": 3.4762330055236816, + "learning_rate": 5.5250929484256463e-05, + "loss": 0.2241, + "step": 27633 + }, + { + "epoch": 0.6477844781612428, + "grad_norm": 3.2782928943634033, + "learning_rate": 5.524434213482158e-05, + "loss": 0.237, + "step": 27634 + }, + { + "epoch": 0.6478079197360479, + "grad_norm": 3.888707160949707, + "learning_rate": 5.523775502823212e-05, + "loss": 0.2669, + "step": 27635 + }, + { + "epoch": 0.6478313613108528, + "grad_norm": 3.054490804672241, + "learning_rate": 5.523116816452389e-05, + "loss": 0.4632, + "step": 27636 + }, + { + "epoch": 0.6478548028856579, + "grad_norm": 3.694524049758911, + "learning_rate": 5.5224581543732615e-05, + "loss": 0.3539, + "step": 27637 + }, + { + "epoch": 0.6478782444604628, + "grad_norm": 0.8447960019111633, + "learning_rate": 5.521799516589401e-05, + "loss": 0.0431, + "step": 27638 + }, + { + "epoch": 0.6479016860352679, + "grad_norm": 4.196222305297852, + "learning_rate": 5.521140903104378e-05, + "loss": 0.2899, + "step": 27639 + }, + { + "epoch": 0.6479251276100728, + "grad_norm": 2.1834394931793213, + "learning_rate": 5.520482313921775e-05, + "loss": 0.3228, + "step": 27640 + }, + { + "epoch": 0.6479485691848779, + "grad_norm": 0.5460634827613831, + "learning_rate": 5.519823749045156e-05, + "loss": 0.0376, + "step": 27641 + }, + { + "epoch": 0.6479720107596828, + "grad_norm": 3.154407501220703, + "learning_rate": 5.519165208478104e-05, + "loss": 0.3134, + "step": 27642 + }, + { + "epoch": 0.6479954523344879, + "grad_norm": 3.981323003768921, + "learning_rate": 5.518506692224188e-05, + "loss": 0.3195, + "step": 27643 + }, + { + "epoch": 0.6480188939092928, + "grad_norm": 4.0494184494018555, + "learning_rate": 5.51784820028698e-05, + "loss": 0.272, + "step": 27644 + }, + { + "epoch": 0.6480423354840978, + "grad_norm": 3.4411556720733643, + "learning_rate": 5.517189732670054e-05, + "loss": 0.3456, + "step": 27645 + }, + { + "epoch": 0.6480657770589028, + "grad_norm": 3.3483991622924805, + "learning_rate": 5.5165312893769824e-05, + "loss": 0.3925, + "step": 27646 + }, + { + "epoch": 0.6480892186337078, + "grad_norm": 1.4327162504196167, + "learning_rate": 5.5158728704113336e-05, + "loss": 0.1448, + "step": 27647 + }, + { + "epoch": 0.6481126602085128, + "grad_norm": 2.0314455032348633, + "learning_rate": 5.51521447577669e-05, + "loss": 0.2063, + "step": 27648 + }, + { + "epoch": 0.6481361017833178, + "grad_norm": 1.8840574026107788, + "learning_rate": 5.514556105476618e-05, + "loss": 0.1297, + "step": 27649 + }, + { + "epoch": 0.6481595433581228, + "grad_norm": 4.360628128051758, + "learning_rate": 5.513897759514691e-05, + "loss": 0.4618, + "step": 27650 + }, + { + "epoch": 0.6481829849329278, + "grad_norm": 2.342292547225952, + "learning_rate": 5.513239437894477e-05, + "loss": 0.1026, + "step": 27651 + }, + { + "epoch": 0.6482064265077327, + "grad_norm": 2.1915929317474365, + "learning_rate": 5.512581140619555e-05, + "loss": 1.6703, + "step": 27652 + }, + { + "epoch": 0.6482298680825378, + "grad_norm": 2.892719268798828, + "learning_rate": 5.511922867693495e-05, + "loss": 0.2115, + "step": 27653 + }, + { + "epoch": 0.6482533096573427, + "grad_norm": 3.748645782470703, + "learning_rate": 5.5112646191198624e-05, + "loss": 0.3673, + "step": 27654 + }, + { + "epoch": 0.6482767512321478, + "grad_norm": 5.46120548248291, + "learning_rate": 5.51060639490224e-05, + "loss": 0.1708, + "step": 27655 + }, + { + "epoch": 0.6483001928069527, + "grad_norm": 2.2252941131591797, + "learning_rate": 5.5099481950441926e-05, + "loss": 0.1146, + "step": 27656 + }, + { + "epoch": 0.6483236343817578, + "grad_norm": 1.331707239151001, + "learning_rate": 5.5092900195492916e-05, + "loss": 0.1294, + "step": 27657 + }, + { + "epoch": 0.6483470759565627, + "grad_norm": 0.5931494235992432, + "learning_rate": 5.50863186842111e-05, + "loss": 0.0723, + "step": 27658 + }, + { + "epoch": 0.6483705175313678, + "grad_norm": 0.7998698353767395, + "learning_rate": 5.507973741663219e-05, + "loss": 0.0426, + "step": 27659 + }, + { + "epoch": 0.6483939591061727, + "grad_norm": 1.293980360031128, + "learning_rate": 5.507315639279184e-05, + "loss": 0.1018, + "step": 27660 + }, + { + "epoch": 0.6484174006809778, + "grad_norm": 3.510270595550537, + "learning_rate": 5.5066575612725844e-05, + "loss": 0.205, + "step": 27661 + }, + { + "epoch": 0.6484408422557827, + "grad_norm": 0.9639936089515686, + "learning_rate": 5.5059995076469875e-05, + "loss": 0.0798, + "step": 27662 + }, + { + "epoch": 0.6484642838305877, + "grad_norm": 2.7125446796417236, + "learning_rate": 5.50534147840596e-05, + "loss": 0.2758, + "step": 27663 + }, + { + "epoch": 0.6484877254053928, + "grad_norm": 1.0283126831054688, + "learning_rate": 5.504683473553079e-05, + "loss": 0.0452, + "step": 27664 + }, + { + "epoch": 0.6485111669801977, + "grad_norm": 3.1728594303131104, + "learning_rate": 5.5040254930919136e-05, + "loss": 0.3303, + "step": 27665 + }, + { + "epoch": 0.6485346085550028, + "grad_norm": 1.5293411016464233, + "learning_rate": 5.503367537026031e-05, + "loss": 0.078, + "step": 27666 + }, + { + "epoch": 0.6485580501298077, + "grad_norm": 3.6939170360565186, + "learning_rate": 5.5027096053589975e-05, + "loss": 0.2986, + "step": 27667 + }, + { + "epoch": 0.6485814917046128, + "grad_norm": 3.3410139083862305, + "learning_rate": 5.502051698094395e-05, + "loss": 0.2135, + "step": 27668 + }, + { + "epoch": 0.6486049332794177, + "grad_norm": 5.9340925216674805, + "learning_rate": 5.5013938152357845e-05, + "loss": 0.347, + "step": 27669 + }, + { + "epoch": 0.6486283748542228, + "grad_norm": 1.757304310798645, + "learning_rate": 5.500735956786738e-05, + "loss": 0.1383, + "step": 27670 + }, + { + "epoch": 0.6486518164290277, + "grad_norm": 4.818680286407471, + "learning_rate": 5.5000781227508244e-05, + "loss": 0.2743, + "step": 27671 + }, + { + "epoch": 0.6486752580038327, + "grad_norm": 1.987221121788025, + "learning_rate": 5.499420313131614e-05, + "loss": 0.0965, + "step": 27672 + }, + { + "epoch": 0.6486986995786377, + "grad_norm": 2.2621984481811523, + "learning_rate": 5.49876252793267e-05, + "loss": 0.1339, + "step": 27673 + }, + { + "epoch": 0.6487221411534427, + "grad_norm": 2.238530397415161, + "learning_rate": 5.498104767157573e-05, + "loss": 0.1408, + "step": 27674 + }, + { + "epoch": 0.6487455827282477, + "grad_norm": 0.3318816125392914, + "learning_rate": 5.4974470308098814e-05, + "loss": 0.024, + "step": 27675 + }, + { + "epoch": 0.6487690243030527, + "grad_norm": 3.086719274520874, + "learning_rate": 5.496789318893174e-05, + "loss": 0.2059, + "step": 27676 + }, + { + "epoch": 0.6487924658778577, + "grad_norm": 1.6100094318389893, + "learning_rate": 5.496131631411012e-05, + "loss": 0.0978, + "step": 27677 + }, + { + "epoch": 0.6488159074526627, + "grad_norm": 1.771628975868225, + "learning_rate": 5.4954739683669685e-05, + "loss": 0.0873, + "step": 27678 + }, + { + "epoch": 0.6488393490274676, + "grad_norm": 1.0273391008377075, + "learning_rate": 5.494816329764605e-05, + "loss": 0.0802, + "step": 27679 + }, + { + "epoch": 0.6488627906022727, + "grad_norm": 5.243627548217773, + "learning_rate": 5.494158715607499e-05, + "loss": 0.8241, + "step": 27680 + }, + { + "epoch": 0.6488862321770776, + "grad_norm": 0.7471398115158081, + "learning_rate": 5.493501125899214e-05, + "loss": 0.048, + "step": 27681 + }, + { + "epoch": 0.6489096737518827, + "grad_norm": 2.88659930229187, + "learning_rate": 5.4928435606433194e-05, + "loss": 0.265, + "step": 27682 + }, + { + "epoch": 0.6489331153266876, + "grad_norm": 2.5912024974823, + "learning_rate": 5.4921860198433816e-05, + "loss": 0.1704, + "step": 27683 + }, + { + "epoch": 0.6489565569014927, + "grad_norm": 1.0983023643493652, + "learning_rate": 5.4915285035029696e-05, + "loss": 0.11, + "step": 27684 + }, + { + "epoch": 0.6489799984762976, + "grad_norm": 0.6387540698051453, + "learning_rate": 5.490871011625647e-05, + "loss": 0.0749, + "step": 27685 + }, + { + "epoch": 0.6490034400511027, + "grad_norm": 3.4653968811035156, + "learning_rate": 5.4902135442149884e-05, + "loss": 0.3093, + "step": 27686 + }, + { + "epoch": 0.6490268816259076, + "grad_norm": 4.166308879852295, + "learning_rate": 5.4895561012745545e-05, + "loss": 0.5188, + "step": 27687 + }, + { + "epoch": 0.6490503232007127, + "grad_norm": 0.6747072339057922, + "learning_rate": 5.4888986828079193e-05, + "loss": 0.0474, + "step": 27688 + }, + { + "epoch": 0.6490737647755176, + "grad_norm": 2.3726165294647217, + "learning_rate": 5.488241288818647e-05, + "loss": 0.1714, + "step": 27689 + }, + { + "epoch": 0.6490972063503226, + "grad_norm": 2.0315396785736084, + "learning_rate": 5.487583919310305e-05, + "loss": 0.1312, + "step": 27690 + }, + { + "epoch": 0.6491206479251276, + "grad_norm": 0.7968350648880005, + "learning_rate": 5.4869265742864596e-05, + "loss": 0.118, + "step": 27691 + }, + { + "epoch": 0.6491440894999326, + "grad_norm": 2.8201842308044434, + "learning_rate": 5.486269253750673e-05, + "loss": 1.6597, + "step": 27692 + }, + { + "epoch": 0.6491675310747376, + "grad_norm": 3.5590972900390625, + "learning_rate": 5.48561195770652e-05, + "loss": 0.3262, + "step": 27693 + }, + { + "epoch": 0.6491909726495426, + "grad_norm": 3.456315040588379, + "learning_rate": 5.484954686157564e-05, + "loss": 0.2386, + "step": 27694 + }, + { + "epoch": 0.6492144142243476, + "grad_norm": 2.7702903747558594, + "learning_rate": 5.484297439107371e-05, + "loss": 0.2139, + "step": 27695 + }, + { + "epoch": 0.6492378557991526, + "grad_norm": 3.1798346042633057, + "learning_rate": 5.483640216559506e-05, + "loss": 0.4694, + "step": 27696 + }, + { + "epoch": 0.6492612973739575, + "grad_norm": 2.523615598678589, + "learning_rate": 5.482983018517536e-05, + "loss": 0.258, + "step": 27697 + }, + { + "epoch": 0.6492847389487626, + "grad_norm": 2.58406925201416, + "learning_rate": 5.4823258449850246e-05, + "loss": 0.2113, + "step": 27698 + }, + { + "epoch": 0.6493081805235675, + "grad_norm": 4.0393452644348145, + "learning_rate": 5.4816686959655425e-05, + "loss": 0.4549, + "step": 27699 + }, + { + "epoch": 0.6493316220983726, + "grad_norm": 3.379401206970215, + "learning_rate": 5.481011571462651e-05, + "loss": 1.5106, + "step": 27700 + }, + { + "epoch": 0.6493550636731775, + "grad_norm": 3.8412506580352783, + "learning_rate": 5.4803544714799203e-05, + "loss": 0.279, + "step": 27701 + }, + { + "epoch": 0.6493785052479826, + "grad_norm": 4.656585216522217, + "learning_rate": 5.479697396020915e-05, + "loss": 0.45, + "step": 27702 + }, + { + "epoch": 0.6494019468227875, + "grad_norm": 2.595456838607788, + "learning_rate": 5.4790403450891967e-05, + "loss": 0.1573, + "step": 27703 + }, + { + "epoch": 0.6494253883975926, + "grad_norm": 2.648561954498291, + "learning_rate": 5.478383318688334e-05, + "loss": 0.1728, + "step": 27704 + }, + { + "epoch": 0.6494488299723975, + "grad_norm": 1.5539748668670654, + "learning_rate": 5.477726316821886e-05, + "loss": 0.0904, + "step": 27705 + }, + { + "epoch": 0.6494722715472026, + "grad_norm": 3.6531412601470947, + "learning_rate": 5.4770693394934256e-05, + "loss": 0.2424, + "step": 27706 + }, + { + "epoch": 0.6494957131220075, + "grad_norm": 3.4577081203460693, + "learning_rate": 5.476412386706515e-05, + "loss": 0.2426, + "step": 27707 + }, + { + "epoch": 0.6495191546968125, + "grad_norm": 2.712010145187378, + "learning_rate": 5.4757554584647185e-05, + "loss": 0.2508, + "step": 27708 + }, + { + "epoch": 0.6495425962716175, + "grad_norm": 3.023855686187744, + "learning_rate": 5.4750985547716e-05, + "loss": 0.2794, + "step": 27709 + }, + { + "epoch": 0.6495660378464225, + "grad_norm": 1.1447654962539673, + "learning_rate": 5.474441675630723e-05, + "loss": 0.11, + "step": 27710 + }, + { + "epoch": 0.6495894794212275, + "grad_norm": 0.625518798828125, + "learning_rate": 5.473784821045649e-05, + "loss": 0.0402, + "step": 27711 + }, + { + "epoch": 0.6496129209960325, + "grad_norm": 1.3375848531723022, + "learning_rate": 5.4731279910199465e-05, + "loss": 0.1544, + "step": 27712 + }, + { + "epoch": 0.6496363625708375, + "grad_norm": 1.2818357944488525, + "learning_rate": 5.472471185557182e-05, + "loss": 0.1055, + "step": 27713 + }, + { + "epoch": 0.6496598041456425, + "grad_norm": 2.27227783203125, + "learning_rate": 5.4718144046609154e-05, + "loss": 0.1834, + "step": 27714 + }, + { + "epoch": 0.6496832457204474, + "grad_norm": 1.1401467323303223, + "learning_rate": 5.471157648334713e-05, + "loss": 0.1219, + "step": 27715 + }, + { + "epoch": 0.6497066872952525, + "grad_norm": 3.7171967029571533, + "learning_rate": 5.470500916582135e-05, + "loss": 0.4597, + "step": 27716 + }, + { + "epoch": 0.6497301288700575, + "grad_norm": 1.9904444217681885, + "learning_rate": 5.4698442094067426e-05, + "loss": 0.2368, + "step": 27717 + }, + { + "epoch": 0.6497535704448625, + "grad_norm": 2.536590099334717, + "learning_rate": 5.469187526812107e-05, + "loss": 0.2651, + "step": 27718 + }, + { + "epoch": 0.6497770120196675, + "grad_norm": 1.6809544563293457, + "learning_rate": 5.468530868801789e-05, + "loss": 0.1249, + "step": 27719 + }, + { + "epoch": 0.6498004535944725, + "grad_norm": 1.430391550064087, + "learning_rate": 5.4678742353793475e-05, + "loss": 0.0986, + "step": 27720 + }, + { + "epoch": 0.6498238951692775, + "grad_norm": 1.5495634078979492, + "learning_rate": 5.4672176265483486e-05, + "loss": 0.1114, + "step": 27721 + }, + { + "epoch": 0.6498473367440825, + "grad_norm": 2.566136121749878, + "learning_rate": 5.4665610423123546e-05, + "loss": 0.2664, + "step": 27722 + }, + { + "epoch": 0.6498707783188875, + "grad_norm": 2.602504253387451, + "learning_rate": 5.465904482674924e-05, + "loss": 0.1507, + "step": 27723 + }, + { + "epoch": 0.6498942198936924, + "grad_norm": 2.907790422439575, + "learning_rate": 5.4652479476396234e-05, + "loss": 0.2584, + "step": 27724 + }, + { + "epoch": 0.6499176614684975, + "grad_norm": 1.6540124416351318, + "learning_rate": 5.464591437210018e-05, + "loss": 0.1495, + "step": 27725 + }, + { + "epoch": 0.6499411030433024, + "grad_norm": 2.177095413208008, + "learning_rate": 5.463934951389669e-05, + "loss": 0.2082, + "step": 27726 + }, + { + "epoch": 0.6499645446181075, + "grad_norm": 2.4102649688720703, + "learning_rate": 5.4632784901821335e-05, + "loss": 0.144, + "step": 27727 + }, + { + "epoch": 0.6499879861929124, + "grad_norm": 1.6220098733901978, + "learning_rate": 5.462622053590978e-05, + "loss": 1.1274, + "step": 27728 + }, + { + "epoch": 0.6500114277677175, + "grad_norm": 0.7475029230117798, + "learning_rate": 5.461965641619764e-05, + "loss": 0.0893, + "step": 27729 + }, + { + "epoch": 0.6500348693425224, + "grad_norm": 4.01336669921875, + "learning_rate": 5.461309254272047e-05, + "loss": 0.2238, + "step": 27730 + }, + { + "epoch": 0.6500583109173275, + "grad_norm": 1.064282774925232, + "learning_rate": 5.4606528915513965e-05, + "loss": 0.102, + "step": 27731 + }, + { + "epoch": 0.6500817524921324, + "grad_norm": 1.828817367553711, + "learning_rate": 5.459996553461372e-05, + "loss": 0.17, + "step": 27732 + }, + { + "epoch": 0.6501051940669375, + "grad_norm": 2.663943290710449, + "learning_rate": 5.459340240005534e-05, + "loss": 0.2762, + "step": 27733 + }, + { + "epoch": 0.6501286356417424, + "grad_norm": 3.7399964332580566, + "learning_rate": 5.4586839511874435e-05, + "loss": 0.2973, + "step": 27734 + }, + { + "epoch": 0.6501520772165474, + "grad_norm": 2.371878147125244, + "learning_rate": 5.45802768701066e-05, + "loss": 1.2154, + "step": 27735 + }, + { + "epoch": 0.6501755187913524, + "grad_norm": 3.463071346282959, + "learning_rate": 5.457371447478745e-05, + "loss": 0.2698, + "step": 27736 + }, + { + "epoch": 0.6501989603661574, + "grad_norm": 1.8193058967590332, + "learning_rate": 5.456715232595257e-05, + "loss": 0.8767, + "step": 27737 + }, + { + "epoch": 0.6502224019409624, + "grad_norm": 1.133012056350708, + "learning_rate": 5.456059042363766e-05, + "loss": 0.1129, + "step": 27738 + }, + { + "epoch": 0.6502458435157674, + "grad_norm": 0.9114055633544922, + "learning_rate": 5.455402876787827e-05, + "loss": 0.0778, + "step": 27739 + }, + { + "epoch": 0.6502692850905724, + "grad_norm": 1.333725929260254, + "learning_rate": 5.454746735870998e-05, + "loss": 0.0489, + "step": 27740 + }, + { + "epoch": 0.6502927266653774, + "grad_norm": 0.37774208188056946, + "learning_rate": 5.454090619616843e-05, + "loss": 0.028, + "step": 27741 + }, + { + "epoch": 0.6503161682401823, + "grad_norm": 4.480640888214111, + "learning_rate": 5.453434528028918e-05, + "loss": 0.3585, + "step": 27742 + }, + { + "epoch": 0.6503396098149874, + "grad_norm": 2.4421091079711914, + "learning_rate": 5.452778461110781e-05, + "loss": 0.2571, + "step": 27743 + }, + { + "epoch": 0.6503630513897923, + "grad_norm": 1.744948148727417, + "learning_rate": 5.4521224188660016e-05, + "loss": 0.0516, + "step": 27744 + }, + { + "epoch": 0.6503864929645974, + "grad_norm": 3.824971914291382, + "learning_rate": 5.451466401298133e-05, + "loss": 0.3599, + "step": 27745 + }, + { + "epoch": 0.6504099345394023, + "grad_norm": 1.9706377983093262, + "learning_rate": 5.450810408410736e-05, + "loss": 0.1128, + "step": 27746 + }, + { + "epoch": 0.6504333761142074, + "grad_norm": 5.544763088226318, + "learning_rate": 5.450154440207369e-05, + "loss": 0.4498, + "step": 27747 + }, + { + "epoch": 0.6504568176890123, + "grad_norm": 1.4325432777404785, + "learning_rate": 5.449498496691589e-05, + "loss": 0.1096, + "step": 27748 + }, + { + "epoch": 0.6504802592638174, + "grad_norm": 1.474294900894165, + "learning_rate": 5.4488425778669616e-05, + "loss": 0.1871, + "step": 27749 + }, + { + "epoch": 0.6505037008386223, + "grad_norm": 4.75718879699707, + "learning_rate": 5.448186683737038e-05, + "loss": 0.3718, + "step": 27750 + }, + { + "epoch": 0.6505271424134274, + "grad_norm": 3.1093578338623047, + "learning_rate": 5.447530814305385e-05, + "loss": 1.701, + "step": 27751 + }, + { + "epoch": 0.6505505839882323, + "grad_norm": 4.221881866455078, + "learning_rate": 5.446874969575558e-05, + "loss": 0.362, + "step": 27752 + }, + { + "epoch": 0.6505740255630373, + "grad_norm": 2.314908981323242, + "learning_rate": 5.446219149551116e-05, + "loss": 0.3473, + "step": 27753 + }, + { + "epoch": 0.6505974671378423, + "grad_norm": 1.218446969985962, + "learning_rate": 5.445563354235616e-05, + "loss": 0.1126, + "step": 27754 + }, + { + "epoch": 0.6506209087126473, + "grad_norm": 3.076352834701538, + "learning_rate": 5.4449075836326144e-05, + "loss": 0.398, + "step": 27755 + }, + { + "epoch": 0.6506443502874523, + "grad_norm": 4.281950950622559, + "learning_rate": 5.444251837745676e-05, + "loss": 0.3444, + "step": 27756 + }, + { + "epoch": 0.6506677918622573, + "grad_norm": 1.9044171571731567, + "learning_rate": 5.443596116578356e-05, + "loss": 0.1895, + "step": 27757 + }, + { + "epoch": 0.6506912334370623, + "grad_norm": 2.8313825130462646, + "learning_rate": 5.4429404201342105e-05, + "loss": 0.1446, + "step": 27758 + }, + { + "epoch": 0.6507146750118673, + "grad_norm": 4.10049295425415, + "learning_rate": 5.4422847484167995e-05, + "loss": 0.2799, + "step": 27759 + }, + { + "epoch": 0.6507381165866722, + "grad_norm": 4.301251411437988, + "learning_rate": 5.441629101429674e-05, + "loss": 0.3148, + "step": 27760 + }, + { + "epoch": 0.6507615581614773, + "grad_norm": 0.938020646572113, + "learning_rate": 5.440973479176402e-05, + "loss": 0.0967, + "step": 27761 + }, + { + "epoch": 0.6507849997362822, + "grad_norm": 3.7758872509002686, + "learning_rate": 5.440317881660533e-05, + "loss": 0.3429, + "step": 27762 + }, + { + "epoch": 0.6508084413110873, + "grad_norm": 1.638022780418396, + "learning_rate": 5.4396623088856294e-05, + "loss": 1.2137, + "step": 27763 + }, + { + "epoch": 0.6508318828858922, + "grad_norm": 3.221662998199463, + "learning_rate": 5.439006760855248e-05, + "loss": 0.3601, + "step": 27764 + }, + { + "epoch": 0.6508553244606973, + "grad_norm": 3.3901305198669434, + "learning_rate": 5.438351237572943e-05, + "loss": 0.3766, + "step": 27765 + }, + { + "epoch": 0.6508787660355022, + "grad_norm": 3.150956869125366, + "learning_rate": 5.437695739042272e-05, + "loss": 1.0909, + "step": 27766 + }, + { + "epoch": 0.6509022076103073, + "grad_norm": 2.0869762897491455, + "learning_rate": 5.437040265266794e-05, + "loss": 0.1629, + "step": 27767 + }, + { + "epoch": 0.6509256491851123, + "grad_norm": 4.236637592315674, + "learning_rate": 5.4363848162500595e-05, + "loss": 0.4946, + "step": 27768 + }, + { + "epoch": 0.6509490907599172, + "grad_norm": 3.9141299724578857, + "learning_rate": 5.435729391995632e-05, + "loss": 0.2213, + "step": 27769 + }, + { + "epoch": 0.6509725323347223, + "grad_norm": 2.5015625953674316, + "learning_rate": 5.435073992507067e-05, + "loss": 0.2707, + "step": 27770 + }, + { + "epoch": 0.6509959739095272, + "grad_norm": 3.6524181365966797, + "learning_rate": 5.434418617787918e-05, + "loss": 0.2685, + "step": 27771 + }, + { + "epoch": 0.6510194154843323, + "grad_norm": 0.6799060702323914, + "learning_rate": 5.433763267841737e-05, + "loss": 0.0474, + "step": 27772 + }, + { + "epoch": 0.6510428570591372, + "grad_norm": 1.3129637241363525, + "learning_rate": 5.4331079426720896e-05, + "loss": 0.1228, + "step": 27773 + }, + { + "epoch": 0.6510662986339423, + "grad_norm": 2.033506155014038, + "learning_rate": 5.432452642282527e-05, + "loss": 0.1471, + "step": 27774 + }, + { + "epoch": 0.6510897402087472, + "grad_norm": 3.0115745067596436, + "learning_rate": 5.431797366676601e-05, + "loss": 0.1746, + "step": 27775 + }, + { + "epoch": 0.6511131817835523, + "grad_norm": 1.6553348302841187, + "learning_rate": 5.4311421158578744e-05, + "loss": 0.136, + "step": 27776 + }, + { + "epoch": 0.6511366233583572, + "grad_norm": 2.7135255336761475, + "learning_rate": 5.4304868898299e-05, + "loss": 0.1673, + "step": 27777 + }, + { + "epoch": 0.6511600649331623, + "grad_norm": 3.926945924758911, + "learning_rate": 5.429831688596232e-05, + "loss": 0.3202, + "step": 27778 + }, + { + "epoch": 0.6511835065079672, + "grad_norm": 1.8065497875213623, + "learning_rate": 5.4291765121604256e-05, + "loss": 0.1698, + "step": 27779 + }, + { + "epoch": 0.6512069480827722, + "grad_norm": 1.7917898893356323, + "learning_rate": 5.428521360526036e-05, + "loss": 0.0583, + "step": 27780 + }, + { + "epoch": 0.6512303896575772, + "grad_norm": 2.8887033462524414, + "learning_rate": 5.427866233696613e-05, + "loss": 1.5169, + "step": 27781 + }, + { + "epoch": 0.6512538312323822, + "grad_norm": 2.903944253921509, + "learning_rate": 5.4272111316757215e-05, + "loss": 0.1745, + "step": 27782 + }, + { + "epoch": 0.6512772728071872, + "grad_norm": 4.827025890350342, + "learning_rate": 5.42655605446691e-05, + "loss": 0.4012, + "step": 27783 + }, + { + "epoch": 0.6513007143819922, + "grad_norm": 1.2792625427246094, + "learning_rate": 5.425901002073734e-05, + "loss": 0.1231, + "step": 27784 + }, + { + "epoch": 0.6513241559567972, + "grad_norm": 4.606050968170166, + "learning_rate": 5.425245974499745e-05, + "loss": 0.3091, + "step": 27785 + }, + { + "epoch": 0.6513475975316022, + "grad_norm": 2.234595537185669, + "learning_rate": 5.4245909717485045e-05, + "loss": 1.2958, + "step": 27786 + }, + { + "epoch": 0.6513710391064071, + "grad_norm": 2.1330690383911133, + "learning_rate": 5.42393599382356e-05, + "loss": 0.1535, + "step": 27787 + }, + { + "epoch": 0.6513944806812122, + "grad_norm": 0.7871618270874023, + "learning_rate": 5.4232810407284653e-05, + "loss": 0.0561, + "step": 27788 + }, + { + "epoch": 0.6514179222560171, + "grad_norm": 1.5998525619506836, + "learning_rate": 5.4226261124667786e-05, + "loss": 0.0731, + "step": 27789 + }, + { + "epoch": 0.6514413638308222, + "grad_norm": 3.574554920196533, + "learning_rate": 5.4219712090420514e-05, + "loss": 0.5152, + "step": 27790 + }, + { + "epoch": 0.6514648054056271, + "grad_norm": 2.6929452419281006, + "learning_rate": 5.421316330457839e-05, + "loss": 0.3207, + "step": 27791 + }, + { + "epoch": 0.6514882469804322, + "grad_norm": 1.3011587858200073, + "learning_rate": 5.420661476717692e-05, + "loss": 0.0912, + "step": 27792 + }, + { + "epoch": 0.6515116885552371, + "grad_norm": 1.430223822593689, + "learning_rate": 5.42000664782516e-05, + "loss": 0.1601, + "step": 27793 + }, + { + "epoch": 0.6515351301300422, + "grad_norm": 2.0883610248565674, + "learning_rate": 5.419351843783805e-05, + "loss": 0.1086, + "step": 27794 + }, + { + "epoch": 0.6515585717048471, + "grad_norm": 4.918107986450195, + "learning_rate": 5.418697064597176e-05, + "loss": 0.3039, + "step": 27795 + }, + { + "epoch": 0.6515820132796522, + "grad_norm": 2.6096699237823486, + "learning_rate": 5.418042310268825e-05, + "loss": 0.2709, + "step": 27796 + }, + { + "epoch": 0.6516054548544571, + "grad_norm": 2.8287270069122314, + "learning_rate": 5.417387580802302e-05, + "loss": 0.3207, + "step": 27797 + }, + { + "epoch": 0.6516288964292621, + "grad_norm": 5.206225395202637, + "learning_rate": 5.416732876201167e-05, + "loss": 0.6134, + "step": 27798 + }, + { + "epoch": 0.6516523380040671, + "grad_norm": 0.794127881526947, + "learning_rate": 5.416078196468968e-05, + "loss": 0.0563, + "step": 27799 + }, + { + "epoch": 0.6516757795788721, + "grad_norm": 2.164163112640381, + "learning_rate": 5.415423541609253e-05, + "loss": 0.1151, + "step": 27800 + }, + { + "epoch": 0.6516992211536771, + "grad_norm": 1.740251898765564, + "learning_rate": 5.414768911625583e-05, + "loss": 0.1122, + "step": 27801 + }, + { + "epoch": 0.6517226627284821, + "grad_norm": 4.076727390289307, + "learning_rate": 5.414114306521505e-05, + "loss": 0.1827, + "step": 27802 + }, + { + "epoch": 0.651746104303287, + "grad_norm": 2.648423671722412, + "learning_rate": 5.4134597263005736e-05, + "loss": 0.1035, + "step": 27803 + }, + { + "epoch": 0.6517695458780921, + "grad_norm": 3.185100793838501, + "learning_rate": 5.4128051709663374e-05, + "loss": 0.1238, + "step": 27804 + }, + { + "epoch": 0.651792987452897, + "grad_norm": 4.005465507507324, + "learning_rate": 5.4121506405223486e-05, + "loss": 0.5502, + "step": 27805 + }, + { + "epoch": 0.6518164290277021, + "grad_norm": 2.1523633003234863, + "learning_rate": 5.4114961349721574e-05, + "loss": 1.524, + "step": 27806 + }, + { + "epoch": 0.651839870602507, + "grad_norm": 3.1596078872680664, + "learning_rate": 5.4108416543193205e-05, + "loss": 0.22, + "step": 27807 + }, + { + "epoch": 0.6518633121773121, + "grad_norm": 1.575574517250061, + "learning_rate": 5.410187198567385e-05, + "loss": 0.1157, + "step": 27808 + }, + { + "epoch": 0.651886753752117, + "grad_norm": 3.187671422958374, + "learning_rate": 5.4095327677199e-05, + "loss": 1.4372, + "step": 27809 + }, + { + "epoch": 0.6519101953269221, + "grad_norm": 1.8498766422271729, + "learning_rate": 5.4088783617804225e-05, + "loss": 0.1484, + "step": 27810 + }, + { + "epoch": 0.651933636901727, + "grad_norm": 0.48599952459335327, + "learning_rate": 5.4082239807525e-05, + "loss": 0.0667, + "step": 27811 + }, + { + "epoch": 0.6519570784765321, + "grad_norm": 2.977409839630127, + "learning_rate": 5.407569624639683e-05, + "loss": 0.296, + "step": 27812 + }, + { + "epoch": 0.651980520051337, + "grad_norm": 2.3809783458709717, + "learning_rate": 5.40691529344552e-05, + "loss": 0.2755, + "step": 27813 + }, + { + "epoch": 0.652003961626142, + "grad_norm": 0.8028793334960938, + "learning_rate": 5.406260987173566e-05, + "loss": 0.0591, + "step": 27814 + }, + { + "epoch": 0.652027403200947, + "grad_norm": 2.7650692462921143, + "learning_rate": 5.405606705827371e-05, + "loss": 0.1862, + "step": 27815 + }, + { + "epoch": 0.652050844775752, + "grad_norm": 1.9900929927825928, + "learning_rate": 5.404952449410481e-05, + "loss": 0.2215, + "step": 27816 + }, + { + "epoch": 0.652074286350557, + "grad_norm": 2.962362766265869, + "learning_rate": 5.404298217926449e-05, + "loss": 0.2565, + "step": 27817 + }, + { + "epoch": 0.652097727925362, + "grad_norm": 2.779633045196533, + "learning_rate": 5.4036440113788244e-05, + "loss": 0.2492, + "step": 27818 + }, + { + "epoch": 0.6521211695001671, + "grad_norm": 1.6641446352005005, + "learning_rate": 5.4029898297711526e-05, + "loss": 0.342, + "step": 27819 + }, + { + "epoch": 0.652144611074972, + "grad_norm": 2.5498714447021484, + "learning_rate": 5.402335673106991e-05, + "loss": 0.238, + "step": 27820 + }, + { + "epoch": 0.6521680526497771, + "grad_norm": 2.4365103244781494, + "learning_rate": 5.4016815413898816e-05, + "loss": 0.239, + "step": 27821 + }, + { + "epoch": 0.652191494224582, + "grad_norm": 0.6922123432159424, + "learning_rate": 5.40102743462338e-05, + "loss": 0.035, + "step": 27822 + }, + { + "epoch": 0.652214935799387, + "grad_norm": 0.562909722328186, + "learning_rate": 5.400373352811035e-05, + "loss": 0.0333, + "step": 27823 + }, + { + "epoch": 0.652238377374192, + "grad_norm": 2.493807554244995, + "learning_rate": 5.399719295956393e-05, + "loss": 0.2142, + "step": 27824 + }, + { + "epoch": 0.652261818948997, + "grad_norm": 0.5843040943145752, + "learning_rate": 5.3990652640630025e-05, + "loss": 0.0321, + "step": 27825 + }, + { + "epoch": 0.652285260523802, + "grad_norm": 1.3110759258270264, + "learning_rate": 5.39841125713441e-05, + "loss": 0.1231, + "step": 27826 + }, + { + "epoch": 0.652308702098607, + "grad_norm": 1.2222005128860474, + "learning_rate": 5.397757275174171e-05, + "loss": 0.1407, + "step": 27827 + }, + { + "epoch": 0.652332143673412, + "grad_norm": 0.6724620461463928, + "learning_rate": 5.397103318185831e-05, + "loss": 0.0898, + "step": 27828 + }, + { + "epoch": 0.652355585248217, + "grad_norm": 2.433257818222046, + "learning_rate": 5.396449386172937e-05, + "loss": 0.1651, + "step": 27829 + }, + { + "epoch": 0.652379026823022, + "grad_norm": 3.224489688873291, + "learning_rate": 5.3957954791390386e-05, + "loss": 0.2436, + "step": 27830 + }, + { + "epoch": 0.652402468397827, + "grad_norm": 1.7718803882598877, + "learning_rate": 5.3951415970876795e-05, + "loss": 0.0566, + "step": 27831 + }, + { + "epoch": 0.652425909972632, + "grad_norm": 3.5524563789367676, + "learning_rate": 5.394487740022416e-05, + "loss": 0.3576, + "step": 27832 + }, + { + "epoch": 0.652449351547437, + "grad_norm": 1.7773901224136353, + "learning_rate": 5.393833907946786e-05, + "loss": 0.0517, + "step": 27833 + }, + { + "epoch": 0.6524727931222419, + "grad_norm": 4.8442182540893555, + "learning_rate": 5.393180100864348e-05, + "loss": 0.6695, + "step": 27834 + }, + { + "epoch": 0.652496234697047, + "grad_norm": 1.116585612297058, + "learning_rate": 5.3925263187786434e-05, + "loss": 0.0802, + "step": 27835 + }, + { + "epoch": 0.6525196762718519, + "grad_norm": 1.7815381288528442, + "learning_rate": 5.391872561693221e-05, + "loss": 0.1962, + "step": 27836 + }, + { + "epoch": 0.652543117846657, + "grad_norm": 3.523973226547241, + "learning_rate": 5.3912188296116283e-05, + "loss": 0.1737, + "step": 27837 + }, + { + "epoch": 0.6525665594214619, + "grad_norm": 2.347867250442505, + "learning_rate": 5.3905651225374075e-05, + "loss": 0.1712, + "step": 27838 + }, + { + "epoch": 0.652590000996267, + "grad_norm": 0.7736392021179199, + "learning_rate": 5.389911440474113e-05, + "loss": 0.0715, + "step": 27839 + }, + { + "epoch": 0.6526134425710719, + "grad_norm": 2.8694190979003906, + "learning_rate": 5.38925778342529e-05, + "loss": 0.1998, + "step": 27840 + }, + { + "epoch": 0.652636884145877, + "grad_norm": 0.9926239848136902, + "learning_rate": 5.388604151394483e-05, + "loss": 0.0822, + "step": 27841 + }, + { + "epoch": 0.6526603257206819, + "grad_norm": 0.5100039839744568, + "learning_rate": 5.3879505443852405e-05, + "loss": 0.0486, + "step": 27842 + }, + { + "epoch": 0.6526837672954869, + "grad_norm": 2.9705123901367188, + "learning_rate": 5.387296962401107e-05, + "loss": 0.2767, + "step": 27843 + }, + { + "epoch": 0.6527072088702919, + "grad_norm": 1.5113089084625244, + "learning_rate": 5.386643405445626e-05, + "loss": 0.0975, + "step": 27844 + }, + { + "epoch": 0.6527306504450969, + "grad_norm": 4.244465351104736, + "learning_rate": 5.385989873522348e-05, + "loss": 0.3126, + "step": 27845 + }, + { + "epoch": 0.6527540920199019, + "grad_norm": 4.169153690338135, + "learning_rate": 5.385336366634822e-05, + "loss": 0.5427, + "step": 27846 + }, + { + "epoch": 0.6527775335947069, + "grad_norm": 1.9838228225708008, + "learning_rate": 5.384682884786592e-05, + "loss": 0.124, + "step": 27847 + }, + { + "epoch": 0.6528009751695119, + "grad_norm": 1.0613281726837158, + "learning_rate": 5.384029427981201e-05, + "loss": 0.0753, + "step": 27848 + }, + { + "epoch": 0.6528244167443169, + "grad_norm": 2.9710476398468018, + "learning_rate": 5.383375996222196e-05, + "loss": 0.1885, + "step": 27849 + }, + { + "epoch": 0.6528478583191218, + "grad_norm": 0.5440493226051331, + "learning_rate": 5.382722589513124e-05, + "loss": 0.0582, + "step": 27850 + }, + { + "epoch": 0.6528712998939269, + "grad_norm": 0.6833273768424988, + "learning_rate": 5.3820692078575244e-05, + "loss": 0.0221, + "step": 27851 + }, + { + "epoch": 0.6528947414687318, + "grad_norm": 1.2668694257736206, + "learning_rate": 5.3814158512589505e-05, + "loss": 0.1627, + "step": 27852 + }, + { + "epoch": 0.6529181830435369, + "grad_norm": 2.636176586151123, + "learning_rate": 5.3807625197209455e-05, + "loss": 0.2242, + "step": 27853 + }, + { + "epoch": 0.6529416246183418, + "grad_norm": 2.746258020401001, + "learning_rate": 5.380109213247051e-05, + "loss": 0.206, + "step": 27854 + }, + { + "epoch": 0.6529650661931469, + "grad_norm": 3.8710780143737793, + "learning_rate": 5.379455931840816e-05, + "loss": 0.5253, + "step": 27855 + }, + { + "epoch": 0.6529885077679518, + "grad_norm": 1.9465818405151367, + "learning_rate": 5.3788026755057805e-05, + "loss": 0.2588, + "step": 27856 + }, + { + "epoch": 0.6530119493427569, + "grad_norm": 2.919138193130493, + "learning_rate": 5.378149444245491e-05, + "loss": 0.2703, + "step": 27857 + }, + { + "epoch": 0.6530353909175618, + "grad_norm": 3.0906424522399902, + "learning_rate": 5.37749623806349e-05, + "loss": 0.2127, + "step": 27858 + }, + { + "epoch": 0.6530588324923668, + "grad_norm": 1.0193941593170166, + "learning_rate": 5.376843056963329e-05, + "loss": 0.1437, + "step": 27859 + }, + { + "epoch": 0.6530822740671718, + "grad_norm": 0.8675416707992554, + "learning_rate": 5.376189900948547e-05, + "loss": 0.0733, + "step": 27860 + }, + { + "epoch": 0.6531057156419768, + "grad_norm": 3.968017101287842, + "learning_rate": 5.375536770022689e-05, + "loss": 0.1968, + "step": 27861 + }, + { + "epoch": 0.6531291572167818, + "grad_norm": 4.984671592712402, + "learning_rate": 5.374883664189299e-05, + "loss": 0.2985, + "step": 27862 + }, + { + "epoch": 0.6531525987915868, + "grad_norm": 2.997049331665039, + "learning_rate": 5.3742305834519203e-05, + "loss": 0.1744, + "step": 27863 + }, + { + "epoch": 0.6531760403663918, + "grad_norm": 2.293146848678589, + "learning_rate": 5.373577527814092e-05, + "loss": 0.2197, + "step": 27864 + }, + { + "epoch": 0.6531994819411968, + "grad_norm": 0.46798911690711975, + "learning_rate": 5.372924497279366e-05, + "loss": 0.0308, + "step": 27865 + }, + { + "epoch": 0.6532229235160018, + "grad_norm": 4.099368572235107, + "learning_rate": 5.3722714918512816e-05, + "loss": 0.4324, + "step": 27866 + }, + { + "epoch": 0.6532463650908068, + "grad_norm": 1.6530274152755737, + "learning_rate": 5.371618511533384e-05, + "loss": 0.14, + "step": 27867 + }, + { + "epoch": 0.6532698066656117, + "grad_norm": 2.0134153366088867, + "learning_rate": 5.3709655563292125e-05, + "loss": 0.1461, + "step": 27868 + }, + { + "epoch": 0.6532932482404168, + "grad_norm": 3.8836135864257812, + "learning_rate": 5.370312626242309e-05, + "loss": 0.2425, + "step": 27869 + }, + { + "epoch": 0.6533166898152218, + "grad_norm": 3.9688994884490967, + "learning_rate": 5.369659721276224e-05, + "loss": 0.5691, + "step": 27870 + }, + { + "epoch": 0.6533401313900268, + "grad_norm": 2.878425121307373, + "learning_rate": 5.36900684143449e-05, + "loss": 0.1686, + "step": 27871 + }, + { + "epoch": 0.6533635729648318, + "grad_norm": 4.2119951248168945, + "learning_rate": 5.36835398672066e-05, + "loss": 0.3108, + "step": 27872 + }, + { + "epoch": 0.6533870145396368, + "grad_norm": 4.187289714813232, + "learning_rate": 5.367701157138271e-05, + "loss": 0.4105, + "step": 27873 + }, + { + "epoch": 0.6534104561144418, + "grad_norm": 1.9362765550613403, + "learning_rate": 5.367048352690866e-05, + "loss": 0.9674, + "step": 27874 + }, + { + "epoch": 0.6534338976892468, + "grad_norm": 4.454595565795898, + "learning_rate": 5.366395573381987e-05, + "loss": 0.1244, + "step": 27875 + }, + { + "epoch": 0.6534573392640518, + "grad_norm": 4.01686429977417, + "learning_rate": 5.365742819215172e-05, + "loss": 0.2464, + "step": 27876 + }, + { + "epoch": 0.6534807808388567, + "grad_norm": 4.9121012687683105, + "learning_rate": 5.36509009019397e-05, + "loss": 0.3586, + "step": 27877 + }, + { + "epoch": 0.6535042224136618, + "grad_norm": 1.2423419952392578, + "learning_rate": 5.364437386321921e-05, + "loss": 0.0803, + "step": 27878 + }, + { + "epoch": 0.6535276639884667, + "grad_norm": 1.9576133489608765, + "learning_rate": 5.363784707602564e-05, + "loss": 0.2533, + "step": 27879 + }, + { + "epoch": 0.6535511055632718, + "grad_norm": 2.2790091037750244, + "learning_rate": 5.3631320540394414e-05, + "loss": 0.2339, + "step": 27880 + }, + { + "epoch": 0.6535745471380767, + "grad_norm": 3.162412643432617, + "learning_rate": 5.362479425636096e-05, + "loss": 0.1802, + "step": 27881 + }, + { + "epoch": 0.6535979887128818, + "grad_norm": 3.9795892238616943, + "learning_rate": 5.361826822396061e-05, + "loss": 0.2137, + "step": 27882 + }, + { + "epoch": 0.6536214302876867, + "grad_norm": 2.7474513053894043, + "learning_rate": 5.3611742443228863e-05, + "loss": 0.1908, + "step": 27883 + }, + { + "epoch": 0.6536448718624918, + "grad_norm": 2.619967460632324, + "learning_rate": 5.360521691420114e-05, + "loss": 0.1288, + "step": 27884 + }, + { + "epoch": 0.6536683134372967, + "grad_norm": 2.1892757415771484, + "learning_rate": 5.359869163691281e-05, + "loss": 0.1628, + "step": 27885 + }, + { + "epoch": 0.6536917550121017, + "grad_norm": 1.229909896850586, + "learning_rate": 5.3592166611399296e-05, + "loss": 0.2202, + "step": 27886 + }, + { + "epoch": 0.6537151965869067, + "grad_norm": 1.285658359527588, + "learning_rate": 5.358564183769598e-05, + "loss": 0.1515, + "step": 27887 + }, + { + "epoch": 0.6537386381617117, + "grad_norm": 3.981127977371216, + "learning_rate": 5.357911731583827e-05, + "loss": 0.308, + "step": 27888 + }, + { + "epoch": 0.6537620797365167, + "grad_norm": 3.648169994354248, + "learning_rate": 5.357259304586155e-05, + "loss": 0.2511, + "step": 27889 + }, + { + "epoch": 0.6537855213113217, + "grad_norm": 3.0338094234466553, + "learning_rate": 5.356606902780128e-05, + "loss": 0.2456, + "step": 27890 + }, + { + "epoch": 0.6538089628861267, + "grad_norm": 0.2298762947320938, + "learning_rate": 5.3559545261692826e-05, + "loss": 0.0103, + "step": 27891 + }, + { + "epoch": 0.6538324044609317, + "grad_norm": 1.2070088386535645, + "learning_rate": 5.355302174757157e-05, + "loss": 0.0905, + "step": 27892 + }, + { + "epoch": 0.6538558460357367, + "grad_norm": 5.734522819519043, + "learning_rate": 5.354649848547295e-05, + "loss": 0.7269, + "step": 27893 + }, + { + "epoch": 0.6538792876105417, + "grad_norm": 2.02335786819458, + "learning_rate": 5.353997547543228e-05, + "loss": 1.8215, + "step": 27894 + }, + { + "epoch": 0.6539027291853466, + "grad_norm": 3.3019320964813232, + "learning_rate": 5.353345271748505e-05, + "loss": 0.4599, + "step": 27895 + }, + { + "epoch": 0.6539261707601517, + "grad_norm": 0.9925239086151123, + "learning_rate": 5.352693021166658e-05, + "loss": 0.0835, + "step": 27896 + }, + { + "epoch": 0.6539496123349566, + "grad_norm": 2.6625845432281494, + "learning_rate": 5.352040795801232e-05, + "loss": 0.21, + "step": 27897 + }, + { + "epoch": 0.6539730539097617, + "grad_norm": 2.388381004333496, + "learning_rate": 5.351388595655764e-05, + "loss": 0.1932, + "step": 27898 + }, + { + "epoch": 0.6539964954845666, + "grad_norm": 2.1687729358673096, + "learning_rate": 5.350736420733793e-05, + "loss": 0.1209, + "step": 27899 + }, + { + "epoch": 0.6540199370593717, + "grad_norm": 0.7821508049964905, + "learning_rate": 5.3500842710388565e-05, + "loss": 0.0763, + "step": 27900 + }, + { + "epoch": 0.6540433786341766, + "grad_norm": 5.304374694824219, + "learning_rate": 5.349432146574493e-05, + "loss": 0.7369, + "step": 27901 + }, + { + "epoch": 0.6540668202089817, + "grad_norm": 0.8752128481864929, + "learning_rate": 5.348780047344237e-05, + "loss": 0.0595, + "step": 27902 + }, + { + "epoch": 0.6540902617837866, + "grad_norm": 1.575617790222168, + "learning_rate": 5.3481279733516365e-05, + "loss": 0.1995, + "step": 27903 + }, + { + "epoch": 0.6541137033585916, + "grad_norm": 1.9667658805847168, + "learning_rate": 5.347475924600224e-05, + "loss": 0.178, + "step": 27904 + }, + { + "epoch": 0.6541371449333966, + "grad_norm": 3.03182315826416, + "learning_rate": 5.346823901093538e-05, + "loss": 0.2281, + "step": 27905 + }, + { + "epoch": 0.6541605865082016, + "grad_norm": 1.973607063293457, + "learning_rate": 5.346171902835112e-05, + "loss": 0.2868, + "step": 27906 + }, + { + "epoch": 0.6541840280830066, + "grad_norm": 2.5304923057556152, + "learning_rate": 5.345519929828492e-05, + "loss": 0.2566, + "step": 27907 + }, + { + "epoch": 0.6542074696578116, + "grad_norm": 2.720402956008911, + "learning_rate": 5.344867982077212e-05, + "loss": 0.1599, + "step": 27908 + }, + { + "epoch": 0.6542309112326166, + "grad_norm": 2.8246521949768066, + "learning_rate": 5.3442160595848056e-05, + "loss": 0.226, + "step": 27909 + }, + { + "epoch": 0.6542543528074216, + "grad_norm": 1.1436522006988525, + "learning_rate": 5.343564162354817e-05, + "loss": 0.1007, + "step": 27910 + }, + { + "epoch": 0.6542777943822266, + "grad_norm": 3.5311717987060547, + "learning_rate": 5.34291229039078e-05, + "loss": 0.3137, + "step": 27911 + }, + { + "epoch": 0.6543012359570316, + "grad_norm": 2.3093318939208984, + "learning_rate": 5.342260443696232e-05, + "loss": 0.2787, + "step": 27912 + }, + { + "epoch": 0.6543246775318365, + "grad_norm": 1.81916344165802, + "learning_rate": 5.341608622274711e-05, + "loss": 0.0943, + "step": 27913 + }, + { + "epoch": 0.6543481191066416, + "grad_norm": 2.472440242767334, + "learning_rate": 5.3409568261297484e-05, + "loss": 0.3276, + "step": 27914 + }, + { + "epoch": 0.6543715606814465, + "grad_norm": 3.229522228240967, + "learning_rate": 5.340305055264887e-05, + "loss": 0.2878, + "step": 27915 + }, + { + "epoch": 0.6543950022562516, + "grad_norm": 0.9826372861862183, + "learning_rate": 5.3396533096836635e-05, + "loss": 0.1455, + "step": 27916 + }, + { + "epoch": 0.6544184438310565, + "grad_norm": 1.4182857275009155, + "learning_rate": 5.339001589389612e-05, + "loss": 0.1594, + "step": 27917 + }, + { + "epoch": 0.6544418854058616, + "grad_norm": 1.8403778076171875, + "learning_rate": 5.338349894386264e-05, + "loss": 0.1579, + "step": 27918 + }, + { + "epoch": 0.6544653269806665, + "grad_norm": 1.7684189081192017, + "learning_rate": 5.3376982246771656e-05, + "loss": 0.0836, + "step": 27919 + }, + { + "epoch": 0.6544887685554716, + "grad_norm": 0.6627650856971741, + "learning_rate": 5.337046580265847e-05, + "loss": 0.0405, + "step": 27920 + }, + { + "epoch": 0.6545122101302766, + "grad_norm": 0.7718677520751953, + "learning_rate": 5.3363949611558404e-05, + "loss": 0.0707, + "step": 27921 + }, + { + "epoch": 0.6545356517050815, + "grad_norm": 1.0182093381881714, + "learning_rate": 5.335743367350689e-05, + "loss": 0.1136, + "step": 27922 + }, + { + "epoch": 0.6545590932798866, + "grad_norm": 3.6683154106140137, + "learning_rate": 5.3350917988539285e-05, + "loss": 0.2896, + "step": 27923 + }, + { + "epoch": 0.6545825348546915, + "grad_norm": 2.3947784900665283, + "learning_rate": 5.3344402556690886e-05, + "loss": 0.1359, + "step": 27924 + }, + { + "epoch": 0.6546059764294966, + "grad_norm": 3.0532054901123047, + "learning_rate": 5.333788737799707e-05, + "loss": 0.248, + "step": 27925 + }, + { + "epoch": 0.6546294180043015, + "grad_norm": 2.6762964725494385, + "learning_rate": 5.33313724524932e-05, + "loss": 0.1795, + "step": 27926 + }, + { + "epoch": 0.6546528595791066, + "grad_norm": 0.9927926659584045, + "learning_rate": 5.332485778021457e-05, + "loss": 0.0756, + "step": 27927 + }, + { + "epoch": 0.6546763011539115, + "grad_norm": 5.429215431213379, + "learning_rate": 5.331834336119661e-05, + "loss": 0.1793, + "step": 27928 + }, + { + "epoch": 0.6546997427287166, + "grad_norm": 3.0345962047576904, + "learning_rate": 5.331182919547464e-05, + "loss": 0.2056, + "step": 27929 + }, + { + "epoch": 0.6547231843035215, + "grad_norm": 0.848393976688385, + "learning_rate": 5.3305315283084e-05, + "loss": 0.0495, + "step": 27930 + }, + { + "epoch": 0.6547466258783265, + "grad_norm": 1.190537452697754, + "learning_rate": 5.329880162405998e-05, + "loss": 0.0276, + "step": 27931 + }, + { + "epoch": 0.6547700674531315, + "grad_norm": 2.1855013370513916, + "learning_rate": 5.329228821843801e-05, + "loss": 0.1943, + "step": 27932 + }, + { + "epoch": 0.6547935090279365, + "grad_norm": 2.379587173461914, + "learning_rate": 5.3285775066253415e-05, + "loss": 0.3183, + "step": 27933 + }, + { + "epoch": 0.6548169506027415, + "grad_norm": 1.5014503002166748, + "learning_rate": 5.3279262167541476e-05, + "loss": 0.1378, + "step": 27934 + }, + { + "epoch": 0.6548403921775465, + "grad_norm": 8.449999809265137, + "learning_rate": 5.327274952233761e-05, + "loss": 0.2795, + "step": 27935 + }, + { + "epoch": 0.6548638337523515, + "grad_norm": 4.589609622955322, + "learning_rate": 5.326623713067711e-05, + "loss": 0.3588, + "step": 27936 + }, + { + "epoch": 0.6548872753271565, + "grad_norm": 3.0230870246887207, + "learning_rate": 5.3259724992595336e-05, + "loss": 0.2866, + "step": 27937 + }, + { + "epoch": 0.6549107169019615, + "grad_norm": 3.0315096378326416, + "learning_rate": 5.325321310812761e-05, + "loss": 0.366, + "step": 27938 + }, + { + "epoch": 0.6549341584767665, + "grad_norm": 3.933919906616211, + "learning_rate": 5.3246701477309255e-05, + "loss": 0.3329, + "step": 27939 + }, + { + "epoch": 0.6549576000515714, + "grad_norm": 3.985053777694702, + "learning_rate": 5.324019010017557e-05, + "loss": 0.5799, + "step": 27940 + }, + { + "epoch": 0.6549810416263765, + "grad_norm": 0.6293239593505859, + "learning_rate": 5.323367897676197e-05, + "loss": 0.0517, + "step": 27941 + }, + { + "epoch": 0.6550044832011814, + "grad_norm": 5.690484046936035, + "learning_rate": 5.3227168107103755e-05, + "loss": 0.4192, + "step": 27942 + }, + { + "epoch": 0.6550279247759865, + "grad_norm": 0.42141348123550415, + "learning_rate": 5.322065749123618e-05, + "loss": 0.025, + "step": 27943 + }, + { + "epoch": 0.6550513663507914, + "grad_norm": 3.4916746616363525, + "learning_rate": 5.321414712919468e-05, + "loss": 0.308, + "step": 27944 + }, + { + "epoch": 0.6550748079255965, + "grad_norm": 1.2836257219314575, + "learning_rate": 5.3207637021014545e-05, + "loss": 0.0572, + "step": 27945 + }, + { + "epoch": 0.6550982495004014, + "grad_norm": 2.3348917961120605, + "learning_rate": 5.320112716673108e-05, + "loss": 0.2319, + "step": 27946 + }, + { + "epoch": 0.6551216910752065, + "grad_norm": 3.779301404953003, + "learning_rate": 5.319461756637958e-05, + "loss": 0.2717, + "step": 27947 + }, + { + "epoch": 0.6551451326500114, + "grad_norm": 5.016529083251953, + "learning_rate": 5.3188108219995424e-05, + "loss": 0.4208, + "step": 27948 + }, + { + "epoch": 0.6551685742248164, + "grad_norm": 3.1019203662872314, + "learning_rate": 5.3181599127613926e-05, + "loss": 0.2844, + "step": 27949 + }, + { + "epoch": 0.6551920157996214, + "grad_norm": 1.4247432947158813, + "learning_rate": 5.317509028927038e-05, + "loss": 0.2396, + "step": 27950 + }, + { + "epoch": 0.6552154573744264, + "grad_norm": 1.2691919803619385, + "learning_rate": 5.316858170500011e-05, + "loss": 0.0973, + "step": 27951 + }, + { + "epoch": 0.6552388989492314, + "grad_norm": 2.7751803398132324, + "learning_rate": 5.3162073374838395e-05, + "loss": 1.3022, + "step": 27952 + }, + { + "epoch": 0.6552623405240364, + "grad_norm": 2.5036394596099854, + "learning_rate": 5.3155565298820634e-05, + "loss": 1.8865, + "step": 27953 + }, + { + "epoch": 0.6552857820988414, + "grad_norm": 7.395594120025635, + "learning_rate": 5.314905747698208e-05, + "loss": 0.339, + "step": 27954 + }, + { + "epoch": 0.6553092236736464, + "grad_norm": 1.3309975862503052, + "learning_rate": 5.3142549909358034e-05, + "loss": 0.1119, + "step": 27955 + }, + { + "epoch": 0.6553326652484514, + "grad_norm": 1.6364922523498535, + "learning_rate": 5.313604259598385e-05, + "loss": 0.1473, + "step": 27956 + }, + { + "epoch": 0.6553561068232564, + "grad_norm": 1.6904349327087402, + "learning_rate": 5.3129535536894825e-05, + "loss": 0.107, + "step": 27957 + }, + { + "epoch": 0.6553795483980613, + "grad_norm": 3.0580248832702637, + "learning_rate": 5.3123028732126256e-05, + "loss": 0.2076, + "step": 27958 + }, + { + "epoch": 0.6554029899728664, + "grad_norm": 3.0560877323150635, + "learning_rate": 5.3116522181713415e-05, + "loss": 0.2315, + "step": 27959 + }, + { + "epoch": 0.6554264315476713, + "grad_norm": 3.2167763710021973, + "learning_rate": 5.3110015885691686e-05, + "loss": 0.3121, + "step": 27960 + }, + { + "epoch": 0.6554498731224764, + "grad_norm": 1.1909946203231812, + "learning_rate": 5.310350984409632e-05, + "loss": 0.0408, + "step": 27961 + }, + { + "epoch": 0.6554733146972813, + "grad_norm": 2.4050323963165283, + "learning_rate": 5.309700405696263e-05, + "loss": 0.1928, + "step": 27962 + }, + { + "epoch": 0.6554967562720864, + "grad_norm": 0.2931895852088928, + "learning_rate": 5.309049852432593e-05, + "loss": 0.0252, + "step": 27963 + }, + { + "epoch": 0.6555201978468913, + "grad_norm": 1.0059047937393188, + "learning_rate": 5.308399324622149e-05, + "loss": 0.0933, + "step": 27964 + }, + { + "epoch": 0.6555436394216964, + "grad_norm": 5.473376750946045, + "learning_rate": 5.307748822268457e-05, + "loss": 0.3713, + "step": 27965 + }, + { + "epoch": 0.6555670809965013, + "grad_norm": 3.0789551734924316, + "learning_rate": 5.307098345375058e-05, + "loss": 0.3445, + "step": 27966 + }, + { + "epoch": 0.6555905225713063, + "grad_norm": 3.8555026054382324, + "learning_rate": 5.30644789394547e-05, + "loss": 0.3257, + "step": 27967 + }, + { + "epoch": 0.6556139641461113, + "grad_norm": 3.57775616645813, + "learning_rate": 5.3057974679832324e-05, + "loss": 0.2606, + "step": 27968 + }, + { + "epoch": 0.6556374057209163, + "grad_norm": 3.0367372035980225, + "learning_rate": 5.305147067491869e-05, + "loss": 0.24, + "step": 27969 + }, + { + "epoch": 0.6556608472957213, + "grad_norm": 3.276798963546753, + "learning_rate": 5.3044966924749106e-05, + "loss": 0.2487, + "step": 27970 + }, + { + "epoch": 0.6556842888705263, + "grad_norm": 2.9308509826660156, + "learning_rate": 5.3038463429358856e-05, + "loss": 0.338, + "step": 27971 + }, + { + "epoch": 0.6557077304453314, + "grad_norm": 2.312870502471924, + "learning_rate": 5.3031960188783166e-05, + "loss": 0.2816, + "step": 27972 + }, + { + "epoch": 0.6557311720201363, + "grad_norm": 0.6573295593261719, + "learning_rate": 5.3025457203057424e-05, + "loss": 0.0702, + "step": 27973 + }, + { + "epoch": 0.6557546135949414, + "grad_norm": 3.1836962699890137, + "learning_rate": 5.3018954472216886e-05, + "loss": 0.3383, + "step": 27974 + }, + { + "epoch": 0.6557780551697463, + "grad_norm": 3.280390501022339, + "learning_rate": 5.301245199629681e-05, + "loss": 0.3871, + "step": 27975 + }, + { + "epoch": 0.6558014967445513, + "grad_norm": 2.9340662956237793, + "learning_rate": 5.30059497753325e-05, + "loss": 0.2193, + "step": 27976 + }, + { + "epoch": 0.6558249383193563, + "grad_norm": 1.7759339809417725, + "learning_rate": 5.2999447809359214e-05, + "loss": 0.1797, + "step": 27977 + }, + { + "epoch": 0.6558483798941613, + "grad_norm": 0.3691716492176056, + "learning_rate": 5.299294609841221e-05, + "loss": 0.0278, + "step": 27978 + }, + { + "epoch": 0.6558718214689663, + "grad_norm": 1.8723405599594116, + "learning_rate": 5.298644464252682e-05, + "loss": 0.1139, + "step": 27979 + }, + { + "epoch": 0.6558952630437713, + "grad_norm": 4.780988693237305, + "learning_rate": 5.297994344173832e-05, + "loss": 0.5649, + "step": 27980 + }, + { + "epoch": 0.6559187046185763, + "grad_norm": 2.3349475860595703, + "learning_rate": 5.297344249608198e-05, + "loss": 0.2527, + "step": 27981 + }, + { + "epoch": 0.6559421461933813, + "grad_norm": 3.0347163677215576, + "learning_rate": 5.296694180559306e-05, + "loss": 1.1021, + "step": 27982 + }, + { + "epoch": 0.6559655877681863, + "grad_norm": 2.6907947063446045, + "learning_rate": 5.296044137030685e-05, + "loss": 0.2114, + "step": 27983 + }, + { + "epoch": 0.6559890293429913, + "grad_norm": 3.4483540058135986, + "learning_rate": 5.295394119025855e-05, + "loss": 0.2509, + "step": 27984 + }, + { + "epoch": 0.6560124709177962, + "grad_norm": 3.6590652465820312, + "learning_rate": 5.294744126548353e-05, + "loss": 0.2669, + "step": 27985 + }, + { + "epoch": 0.6560359124926013, + "grad_norm": 2.9703872203826904, + "learning_rate": 5.294094159601702e-05, + "loss": 0.3085, + "step": 27986 + }, + { + "epoch": 0.6560593540674062, + "grad_norm": 2.638427734375, + "learning_rate": 5.293444218189428e-05, + "loss": 0.1964, + "step": 27987 + }, + { + "epoch": 0.6560827956422113, + "grad_norm": 0.7280782461166382, + "learning_rate": 5.2927943023150586e-05, + "loss": 0.041, + "step": 27988 + }, + { + "epoch": 0.6561062372170162, + "grad_norm": 5.2610626220703125, + "learning_rate": 5.2921444119821204e-05, + "loss": 0.5517, + "step": 27989 + }, + { + "epoch": 0.6561296787918213, + "grad_norm": 3.4590365886688232, + "learning_rate": 5.2914945471941334e-05, + "loss": 0.251, + "step": 27990 + }, + { + "epoch": 0.6561531203666262, + "grad_norm": 6.057846546173096, + "learning_rate": 5.29084470795463e-05, + "loss": 0.3092, + "step": 27991 + }, + { + "epoch": 0.6561765619414313, + "grad_norm": 2.3822576999664307, + "learning_rate": 5.29019489426714e-05, + "loss": 1.006, + "step": 27992 + }, + { + "epoch": 0.6562000035162362, + "grad_norm": 3.3739445209503174, + "learning_rate": 5.289545106135185e-05, + "loss": 0.218, + "step": 27993 + }, + { + "epoch": 0.6562234450910412, + "grad_norm": 0.3936115801334381, + "learning_rate": 5.28889534356229e-05, + "loss": 0.0463, + "step": 27994 + }, + { + "epoch": 0.6562468866658462, + "grad_norm": 2.329301357269287, + "learning_rate": 5.288245606551983e-05, + "loss": 1.0798, + "step": 27995 + }, + { + "epoch": 0.6562703282406512, + "grad_norm": 2.269078016281128, + "learning_rate": 5.287595895107785e-05, + "loss": 0.0836, + "step": 27996 + }, + { + "epoch": 0.6562937698154562, + "grad_norm": 2.8356032371520996, + "learning_rate": 5.286946209233221e-05, + "loss": 0.2309, + "step": 27997 + }, + { + "epoch": 0.6563172113902612, + "grad_norm": 1.9576143026351929, + "learning_rate": 5.286296548931825e-05, + "loss": 0.2043, + "step": 27998 + }, + { + "epoch": 0.6563406529650662, + "grad_norm": 0.6768832802772522, + "learning_rate": 5.285646914207116e-05, + "loss": 0.0722, + "step": 27999 + }, + { + "epoch": 0.6563640945398712, + "grad_norm": 2.9577600955963135, + "learning_rate": 5.284997305062618e-05, + "loss": 0.1337, + "step": 28000 + }, + { + "epoch": 0.6563875361146762, + "grad_norm": 6.367641448974609, + "learning_rate": 5.284347721501859e-05, + "loss": 0.4708, + "step": 28001 + }, + { + "epoch": 0.6564109776894812, + "grad_norm": 3.4748620986938477, + "learning_rate": 5.283698163528361e-05, + "loss": 0.3815, + "step": 28002 + }, + { + "epoch": 0.6564344192642861, + "grad_norm": 1.1807985305786133, + "learning_rate": 5.283048631145645e-05, + "loss": 0.0888, + "step": 28003 + }, + { + "epoch": 0.6564578608390912, + "grad_norm": 2.8631887435913086, + "learning_rate": 5.2823991243572404e-05, + "loss": 0.3453, + "step": 28004 + }, + { + "epoch": 0.6564813024138961, + "grad_norm": 3.328939437866211, + "learning_rate": 5.2817496431666736e-05, + "loss": 0.4179, + "step": 28005 + }, + { + "epoch": 0.6565047439887012, + "grad_norm": 2.6958811283111572, + "learning_rate": 5.281100187577468e-05, + "loss": 1.4828, + "step": 28006 + }, + { + "epoch": 0.6565281855635061, + "grad_norm": 0.12312542647123337, + "learning_rate": 5.2804507575931424e-05, + "loss": 0.0042, + "step": 28007 + }, + { + "epoch": 0.6565516271383112, + "grad_norm": 1.6798758506774902, + "learning_rate": 5.279801353217226e-05, + "loss": 0.1279, + "step": 28008 + }, + { + "epoch": 0.6565750687131161, + "grad_norm": 1.9048385620117188, + "learning_rate": 5.27915197445324e-05, + "loss": 0.2213, + "step": 28009 + }, + { + "epoch": 0.6565985102879212, + "grad_norm": 0.9169635772705078, + "learning_rate": 5.278502621304704e-05, + "loss": 0.0959, + "step": 28010 + }, + { + "epoch": 0.6566219518627261, + "grad_norm": 3.7808964252471924, + "learning_rate": 5.27785329377515e-05, + "loss": 1.0411, + "step": 28011 + }, + { + "epoch": 0.6566453934375311, + "grad_norm": 3.079580068588257, + "learning_rate": 5.277203991868096e-05, + "loss": 0.2519, + "step": 28012 + }, + { + "epoch": 0.6566688350123361, + "grad_norm": 2.312596559524536, + "learning_rate": 5.2765547155870654e-05, + "loss": 0.2993, + "step": 28013 + }, + { + "epoch": 0.6566922765871411, + "grad_norm": 2.6167140007019043, + "learning_rate": 5.2759054649355836e-05, + "loss": 0.1929, + "step": 28014 + }, + { + "epoch": 0.6567157181619461, + "grad_norm": 3.9406323432922363, + "learning_rate": 5.27525623991717e-05, + "loss": 0.4449, + "step": 28015 + }, + { + "epoch": 0.6567391597367511, + "grad_norm": 0.21354587376117706, + "learning_rate": 5.2746070405353456e-05, + "loss": 0.0061, + "step": 28016 + }, + { + "epoch": 0.6567626013115561, + "grad_norm": 2.3881585597991943, + "learning_rate": 5.2739578667936354e-05, + "loss": 0.0705, + "step": 28017 + }, + { + "epoch": 0.6567860428863611, + "grad_norm": 2.7522192001342773, + "learning_rate": 5.273308718695568e-05, + "loss": 0.1672, + "step": 28018 + }, + { + "epoch": 0.656809484461166, + "grad_norm": 1.5278438329696655, + "learning_rate": 5.27265959624466e-05, + "loss": 0.0983, + "step": 28019 + }, + { + "epoch": 0.6568329260359711, + "grad_norm": 3.5457310676574707, + "learning_rate": 5.272010499444433e-05, + "loss": 0.2927, + "step": 28020 + }, + { + "epoch": 0.656856367610776, + "grad_norm": 3.0462732315063477, + "learning_rate": 5.27136142829841e-05, + "loss": 0.2279, + "step": 28021 + }, + { + "epoch": 0.6568798091855811, + "grad_norm": 2.917919158935547, + "learning_rate": 5.27071238281011e-05, + "loss": 1.4764, + "step": 28022 + }, + { + "epoch": 0.6569032507603861, + "grad_norm": 2.010376453399658, + "learning_rate": 5.270063362983061e-05, + "loss": 0.1541, + "step": 28023 + }, + { + "epoch": 0.6569266923351911, + "grad_norm": 4.016635894775391, + "learning_rate": 5.269414368820782e-05, + "loss": 0.3497, + "step": 28024 + }, + { + "epoch": 0.6569501339099961, + "grad_norm": 1.5131077766418457, + "learning_rate": 5.2687654003267915e-05, + "loss": 0.8878, + "step": 28025 + }, + { + "epoch": 0.6569735754848011, + "grad_norm": 2.964078187942505, + "learning_rate": 5.268116457504615e-05, + "loss": 0.2388, + "step": 28026 + }, + { + "epoch": 0.6569970170596061, + "grad_norm": 3.665177345275879, + "learning_rate": 5.267467540357771e-05, + "loss": 0.268, + "step": 28027 + }, + { + "epoch": 0.657020458634411, + "grad_norm": 2.3542490005493164, + "learning_rate": 5.2668186488897775e-05, + "loss": 0.1173, + "step": 28028 + }, + { + "epoch": 0.6570439002092161, + "grad_norm": 2.2601828575134277, + "learning_rate": 5.266169783104158e-05, + "loss": 0.1842, + "step": 28029 + }, + { + "epoch": 0.657067341784021, + "grad_norm": 2.3272697925567627, + "learning_rate": 5.2655209430044396e-05, + "loss": 0.0924, + "step": 28030 + }, + { + "epoch": 0.6570907833588261, + "grad_norm": 1.7483248710632324, + "learning_rate": 5.264872128594137e-05, + "loss": 0.1119, + "step": 28031 + }, + { + "epoch": 0.657114224933631, + "grad_norm": 0.7019217014312744, + "learning_rate": 5.264223339876773e-05, + "loss": 0.0487, + "step": 28032 + }, + { + "epoch": 0.6571376665084361, + "grad_norm": 2.3705623149871826, + "learning_rate": 5.263574576855863e-05, + "loss": 0.1277, + "step": 28033 + }, + { + "epoch": 0.657161108083241, + "grad_norm": 1.1006149053573608, + "learning_rate": 5.2629258395349334e-05, + "loss": 0.0854, + "step": 28034 + }, + { + "epoch": 0.6571845496580461, + "grad_norm": 3.262322425842285, + "learning_rate": 5.2622771279174966e-05, + "loss": 0.3146, + "step": 28035 + }, + { + "epoch": 0.657207991232851, + "grad_norm": 1.6006264686584473, + "learning_rate": 5.261628442007081e-05, + "loss": 0.0941, + "step": 28036 + }, + { + "epoch": 0.6572314328076561, + "grad_norm": 1.8261831998825073, + "learning_rate": 5.260979781807203e-05, + "loss": 0.0688, + "step": 28037 + }, + { + "epoch": 0.657254874382461, + "grad_norm": 3.665074110031128, + "learning_rate": 5.260331147321381e-05, + "loss": 0.1712, + "step": 28038 + }, + { + "epoch": 0.657278315957266, + "grad_norm": 2.699043035507202, + "learning_rate": 5.259682538553137e-05, + "loss": 0.1624, + "step": 28039 + }, + { + "epoch": 0.657301757532071, + "grad_norm": 1.646606683731079, + "learning_rate": 5.2590339555059833e-05, + "loss": 0.5957, + "step": 28040 + }, + { + "epoch": 0.657325199106876, + "grad_norm": 0.40095874667167664, + "learning_rate": 5.2583853981834495e-05, + "loss": 0.0212, + "step": 28041 + }, + { + "epoch": 0.657348640681681, + "grad_norm": 1.931139588356018, + "learning_rate": 5.257736866589047e-05, + "loss": 1.4168, + "step": 28042 + }, + { + "epoch": 0.657372082256486, + "grad_norm": 1.4491287469863892, + "learning_rate": 5.2570883607263e-05, + "loss": 0.1513, + "step": 28043 + }, + { + "epoch": 0.657395523831291, + "grad_norm": 2.609818935394287, + "learning_rate": 5.2564398805987244e-05, + "loss": 1.4383, + "step": 28044 + }, + { + "epoch": 0.657418965406096, + "grad_norm": 3.5119292736053467, + "learning_rate": 5.255791426209842e-05, + "loss": 0.3097, + "step": 28045 + }, + { + "epoch": 0.657442406980901, + "grad_norm": 2.8511242866516113, + "learning_rate": 5.2551429975631663e-05, + "loss": 0.1967, + "step": 28046 + }, + { + "epoch": 0.657465848555706, + "grad_norm": 0.6490343809127808, + "learning_rate": 5.2544945946622184e-05, + "loss": 0.047, + "step": 28047 + }, + { + "epoch": 0.6574892901305109, + "grad_norm": 0.588927686214447, + "learning_rate": 5.253846217510513e-05, + "loss": 0.0366, + "step": 28048 + }, + { + "epoch": 0.657512731705316, + "grad_norm": 2.5190811157226562, + "learning_rate": 5.2531978661115745e-05, + "loss": 0.3053, + "step": 28049 + }, + { + "epoch": 0.6575361732801209, + "grad_norm": 3.3649184703826904, + "learning_rate": 5.2525495404689184e-05, + "loss": 0.1914, + "step": 28050 + }, + { + "epoch": 0.657559614854926, + "grad_norm": 2.2964940071105957, + "learning_rate": 5.251901240586064e-05, + "loss": 0.1476, + "step": 28051 + }, + { + "epoch": 0.6575830564297309, + "grad_norm": 2.6117300987243652, + "learning_rate": 5.25125296646652e-05, + "loss": 0.1744, + "step": 28052 + }, + { + "epoch": 0.657606498004536, + "grad_norm": 0.9035806655883789, + "learning_rate": 5.250604718113816e-05, + "loss": 0.0661, + "step": 28053 + }, + { + "epoch": 0.6576299395793409, + "grad_norm": 1.144331932067871, + "learning_rate": 5.249956495531464e-05, + "loss": 0.0856, + "step": 28054 + }, + { + "epoch": 0.657653381154146, + "grad_norm": 1.9674407243728638, + "learning_rate": 5.249308298722978e-05, + "loss": 0.1545, + "step": 28055 + }, + { + "epoch": 0.6576768227289509, + "grad_norm": 1.7432771921157837, + "learning_rate": 5.2486601276918824e-05, + "loss": 0.0541, + "step": 28056 + }, + { + "epoch": 0.6577002643037559, + "grad_norm": 1.2232894897460938, + "learning_rate": 5.2480119824416915e-05, + "loss": 0.133, + "step": 28057 + }, + { + "epoch": 0.6577237058785609, + "grad_norm": 1.790092945098877, + "learning_rate": 5.24736386297592e-05, + "loss": 0.2012, + "step": 28058 + }, + { + "epoch": 0.6577471474533659, + "grad_norm": 2.5547983646392822, + "learning_rate": 5.246715769298087e-05, + "loss": 0.2369, + "step": 28059 + }, + { + "epoch": 0.6577705890281709, + "grad_norm": 1.4245903491973877, + "learning_rate": 5.246067701411703e-05, + "loss": 0.1002, + "step": 28060 + }, + { + "epoch": 0.6577940306029759, + "grad_norm": 3.1658244132995605, + "learning_rate": 5.245419659320294e-05, + "loss": 0.2979, + "step": 28061 + }, + { + "epoch": 0.6578174721777809, + "grad_norm": 2.695819854736328, + "learning_rate": 5.244771643027371e-05, + "loss": 0.1889, + "step": 28062 + }, + { + "epoch": 0.6578409137525859, + "grad_norm": 0.8442542552947998, + "learning_rate": 5.24412365253645e-05, + "loss": 0.0452, + "step": 28063 + }, + { + "epoch": 0.6578643553273908, + "grad_norm": 4.685684680938721, + "learning_rate": 5.243475687851045e-05, + "loss": 0.2383, + "step": 28064 + }, + { + "epoch": 0.6578877969021959, + "grad_norm": 3.572239637374878, + "learning_rate": 5.242827748974679e-05, + "loss": 0.2611, + "step": 28065 + }, + { + "epoch": 0.6579112384770008, + "grad_norm": 1.6344146728515625, + "learning_rate": 5.242179835910863e-05, + "loss": 0.046, + "step": 28066 + }, + { + "epoch": 0.6579346800518059, + "grad_norm": 0.8112280368804932, + "learning_rate": 5.241531948663108e-05, + "loss": 0.1139, + "step": 28067 + }, + { + "epoch": 0.6579581216266108, + "grad_norm": 0.5691813826560974, + "learning_rate": 5.2408840872349395e-05, + "loss": 0.0598, + "step": 28068 + }, + { + "epoch": 0.6579815632014159, + "grad_norm": 1.0041145086288452, + "learning_rate": 5.2402362516298675e-05, + "loss": 0.1036, + "step": 28069 + }, + { + "epoch": 0.6580050047762208, + "grad_norm": 0.5496953129768372, + "learning_rate": 5.2395884418514074e-05, + "loss": 0.0531, + "step": 28070 + }, + { + "epoch": 0.6580284463510259, + "grad_norm": 3.5798609256744385, + "learning_rate": 5.238940657903074e-05, + "loss": 0.2671, + "step": 28071 + }, + { + "epoch": 0.6580518879258308, + "grad_norm": 2.85265851020813, + "learning_rate": 5.238292899788383e-05, + "loss": 0.2328, + "step": 28072 + }, + { + "epoch": 0.6580753295006359, + "grad_norm": 2.12770414352417, + "learning_rate": 5.237645167510845e-05, + "loss": 0.0873, + "step": 28073 + }, + { + "epoch": 0.6580987710754409, + "grad_norm": 2.9926724433898926, + "learning_rate": 5.23699746107398e-05, + "loss": 0.2428, + "step": 28074 + }, + { + "epoch": 0.6581222126502458, + "grad_norm": 1.9252121448516846, + "learning_rate": 5.236349780481303e-05, + "loss": 0.1244, + "step": 28075 + }, + { + "epoch": 0.6581456542250509, + "grad_norm": 0.9592763781547546, + "learning_rate": 5.2357021257363213e-05, + "loss": 0.0485, + "step": 28076 + }, + { + "epoch": 0.6581690957998558, + "grad_norm": 1.7260308265686035, + "learning_rate": 5.2350544968425576e-05, + "loss": 0.1932, + "step": 28077 + }, + { + "epoch": 0.6581925373746609, + "grad_norm": 2.4370229244232178, + "learning_rate": 5.234406893803523e-05, + "loss": 1.8817, + "step": 28078 + }, + { + "epoch": 0.6582159789494658, + "grad_norm": 1.4810354709625244, + "learning_rate": 5.233759316622731e-05, + "loss": 0.1151, + "step": 28079 + }, + { + "epoch": 0.6582394205242709, + "grad_norm": 3.738173007965088, + "learning_rate": 5.233111765303689e-05, + "loss": 0.1878, + "step": 28080 + }, + { + "epoch": 0.6582628620990758, + "grad_norm": 3.822683334350586, + "learning_rate": 5.232464239849921e-05, + "loss": 0.4473, + "step": 28081 + }, + { + "epoch": 0.6582863036738809, + "grad_norm": 1.0509368181228638, + "learning_rate": 5.231816740264936e-05, + "loss": 0.1054, + "step": 28082 + }, + { + "epoch": 0.6583097452486858, + "grad_norm": 4.072659015655518, + "learning_rate": 5.2311692665522495e-05, + "loss": 0.2099, + "step": 28083 + }, + { + "epoch": 0.6583331868234908, + "grad_norm": 2.829272747039795, + "learning_rate": 5.23052181871537e-05, + "loss": 0.2286, + "step": 28084 + }, + { + "epoch": 0.6583566283982958, + "grad_norm": 2.3076071739196777, + "learning_rate": 5.229874396757815e-05, + "loss": 1.8612, + "step": 28085 + }, + { + "epoch": 0.6583800699731008, + "grad_norm": 3.1401281356811523, + "learning_rate": 5.229227000683091e-05, + "loss": 0.2742, + "step": 28086 + }, + { + "epoch": 0.6584035115479058, + "grad_norm": 1.9455668926239014, + "learning_rate": 5.22857963049472e-05, + "loss": 0.0846, + "step": 28087 + }, + { + "epoch": 0.6584269531227108, + "grad_norm": 3.6669986248016357, + "learning_rate": 5.2279322861962085e-05, + "loss": 0.4927, + "step": 28088 + }, + { + "epoch": 0.6584503946975158, + "grad_norm": 0.3256487548351288, + "learning_rate": 5.227284967791068e-05, + "loss": 0.0292, + "step": 28089 + }, + { + "epoch": 0.6584738362723208, + "grad_norm": 1.890960693359375, + "learning_rate": 5.226637675282817e-05, + "loss": 0.1681, + "step": 28090 + }, + { + "epoch": 0.6584972778471258, + "grad_norm": 2.4085264205932617, + "learning_rate": 5.2259904086749654e-05, + "loss": 0.1303, + "step": 28091 + }, + { + "epoch": 0.6585207194219308, + "grad_norm": 6.997681140899658, + "learning_rate": 5.225343167971023e-05, + "loss": 0.4064, + "step": 28092 + }, + { + "epoch": 0.6585441609967357, + "grad_norm": 1.4136557579040527, + "learning_rate": 5.224695953174499e-05, + "loss": 0.0939, + "step": 28093 + }, + { + "epoch": 0.6585676025715408, + "grad_norm": 3.7706472873687744, + "learning_rate": 5.224048764288914e-05, + "loss": 0.3553, + "step": 28094 + }, + { + "epoch": 0.6585910441463457, + "grad_norm": 3.8078079223632812, + "learning_rate": 5.2234016013177724e-05, + "loss": 0.3867, + "step": 28095 + }, + { + "epoch": 0.6586144857211508, + "grad_norm": 2.6219193935394287, + "learning_rate": 5.22275446426459e-05, + "loss": 1.0442, + "step": 28096 + }, + { + "epoch": 0.6586379272959557, + "grad_norm": 1.2528011798858643, + "learning_rate": 5.2221073531328745e-05, + "loss": 0.0978, + "step": 28097 + }, + { + "epoch": 0.6586613688707608, + "grad_norm": 0.7168949842453003, + "learning_rate": 5.221460267926136e-05, + "loss": 0.0747, + "step": 28098 + }, + { + "epoch": 0.6586848104455657, + "grad_norm": 1.8254873752593994, + "learning_rate": 5.220813208647892e-05, + "loss": 1.2428, + "step": 28099 + }, + { + "epoch": 0.6587082520203708, + "grad_norm": 2.208343267440796, + "learning_rate": 5.2201661753016504e-05, + "loss": 1.6237, + "step": 28100 + }, + { + "epoch": 0.6587316935951757, + "grad_norm": 2.9872710704803467, + "learning_rate": 5.219519167890916e-05, + "loss": 0.3018, + "step": 28101 + }, + { + "epoch": 0.6587551351699807, + "grad_norm": 1.9424488544464111, + "learning_rate": 5.2188721864192115e-05, + "loss": 0.4809, + "step": 28102 + }, + { + "epoch": 0.6587785767447857, + "grad_norm": 3.343162775039673, + "learning_rate": 5.218225230890039e-05, + "loss": 0.2567, + "step": 28103 + }, + { + "epoch": 0.6588020183195907, + "grad_norm": 3.7465152740478516, + "learning_rate": 5.217578301306911e-05, + "loss": 0.1061, + "step": 28104 + }, + { + "epoch": 0.6588254598943957, + "grad_norm": 1.5733892917633057, + "learning_rate": 5.216931397673335e-05, + "loss": 0.0991, + "step": 28105 + }, + { + "epoch": 0.6588489014692007, + "grad_norm": 2.8759007453918457, + "learning_rate": 5.216284519992827e-05, + "loss": 0.1485, + "step": 28106 + }, + { + "epoch": 0.6588723430440057, + "grad_norm": 3.412973403930664, + "learning_rate": 5.215637668268893e-05, + "loss": 0.3882, + "step": 28107 + }, + { + "epoch": 0.6588957846188107, + "grad_norm": 1.4712848663330078, + "learning_rate": 5.214990842505044e-05, + "loss": 0.0621, + "step": 28108 + }, + { + "epoch": 0.6589192261936156, + "grad_norm": 2.7371063232421875, + "learning_rate": 5.21434404270479e-05, + "loss": 0.2074, + "step": 28109 + }, + { + "epoch": 0.6589426677684207, + "grad_norm": 4.374042987823486, + "learning_rate": 5.213697268871639e-05, + "loss": 0.6457, + "step": 28110 + }, + { + "epoch": 0.6589661093432256, + "grad_norm": 3.9076621532440186, + "learning_rate": 5.2130505210090977e-05, + "loss": 0.1934, + "step": 28111 + }, + { + "epoch": 0.6589895509180307, + "grad_norm": 0.7006300687789917, + "learning_rate": 5.212403799120682e-05, + "loss": 0.0683, + "step": 28112 + }, + { + "epoch": 0.6590129924928356, + "grad_norm": 1.7264920473098755, + "learning_rate": 5.211757103209894e-05, + "loss": 1.2349, + "step": 28113 + }, + { + "epoch": 0.6590364340676407, + "grad_norm": 1.043351650238037, + "learning_rate": 5.2111104332802505e-05, + "loss": 0.1165, + "step": 28114 + }, + { + "epoch": 0.6590598756424456, + "grad_norm": 1.0921692848205566, + "learning_rate": 5.210463789335258e-05, + "loss": 0.0527, + "step": 28115 + }, + { + "epoch": 0.6590833172172507, + "grad_norm": 2.448908567428589, + "learning_rate": 5.209817171378422e-05, + "loss": 0.1524, + "step": 28116 + }, + { + "epoch": 0.6591067587920556, + "grad_norm": 2.94819974899292, + "learning_rate": 5.2091705794132536e-05, + "loss": 0.2077, + "step": 28117 + }, + { + "epoch": 0.6591302003668607, + "grad_norm": 3.166785955429077, + "learning_rate": 5.208524013443257e-05, + "loss": 0.405, + "step": 28118 + }, + { + "epoch": 0.6591536419416656, + "grad_norm": 3.7856662273406982, + "learning_rate": 5.207877473471946e-05, + "loss": 0.4122, + "step": 28119 + }, + { + "epoch": 0.6591770835164706, + "grad_norm": 2.5718839168548584, + "learning_rate": 5.207230959502827e-05, + "loss": 0.2412, + "step": 28120 + }, + { + "epoch": 0.6592005250912756, + "grad_norm": 1.2085169553756714, + "learning_rate": 5.2065844715394086e-05, + "loss": 0.093, + "step": 28121 + }, + { + "epoch": 0.6592239666660806, + "grad_norm": 0.7929154634475708, + "learning_rate": 5.205938009585196e-05, + "loss": 0.0498, + "step": 28122 + }, + { + "epoch": 0.6592474082408856, + "grad_norm": 0.8557501435279846, + "learning_rate": 5.2052915736437e-05, + "loss": 0.0263, + "step": 28123 + }, + { + "epoch": 0.6592708498156906, + "grad_norm": 0.9629720449447632, + "learning_rate": 5.2046451637184224e-05, + "loss": 0.056, + "step": 28124 + }, + { + "epoch": 0.6592942913904957, + "grad_norm": 2.533362627029419, + "learning_rate": 5.2039987798128756e-05, + "loss": 0.22, + "step": 28125 + }, + { + "epoch": 0.6593177329653006, + "grad_norm": 2.837515354156494, + "learning_rate": 5.2033524219305695e-05, + "loss": 1.5896, + "step": 28126 + }, + { + "epoch": 0.6593411745401057, + "grad_norm": 2.2781736850738525, + "learning_rate": 5.202706090075008e-05, + "loss": 1.4526, + "step": 28127 + }, + { + "epoch": 0.6593646161149106, + "grad_norm": 3.2589547634124756, + "learning_rate": 5.202059784249698e-05, + "loss": 0.137, + "step": 28128 + }, + { + "epoch": 0.6593880576897156, + "grad_norm": 1.561389684677124, + "learning_rate": 5.201413504458147e-05, + "loss": 0.1809, + "step": 28129 + }, + { + "epoch": 0.6594114992645206, + "grad_norm": 3.0803062915802, + "learning_rate": 5.200767250703861e-05, + "loss": 0.2813, + "step": 28130 + }, + { + "epoch": 0.6594349408393256, + "grad_norm": 3.7415804862976074, + "learning_rate": 5.200121022990343e-05, + "loss": 0.1822, + "step": 28131 + }, + { + "epoch": 0.6594583824141306, + "grad_norm": 3.469656467437744, + "learning_rate": 5.1994748213211064e-05, + "loss": 0.4097, + "step": 28132 + }, + { + "epoch": 0.6594818239889356, + "grad_norm": 0.5961861610412598, + "learning_rate": 5.198828645699656e-05, + "loss": 0.0202, + "step": 28133 + }, + { + "epoch": 0.6595052655637406, + "grad_norm": 4.0557684898376465, + "learning_rate": 5.198182496129496e-05, + "loss": 0.4218, + "step": 28134 + }, + { + "epoch": 0.6595287071385456, + "grad_norm": 2.0004632472991943, + "learning_rate": 5.197536372614131e-05, + "loss": 1.0682, + "step": 28135 + }, + { + "epoch": 0.6595521487133506, + "grad_norm": 1.819338083267212, + "learning_rate": 5.196890275157067e-05, + "loss": 0.1347, + "step": 28136 + }, + { + "epoch": 0.6595755902881556, + "grad_norm": 4.002955436706543, + "learning_rate": 5.196244203761814e-05, + "loss": 0.3824, + "step": 28137 + }, + { + "epoch": 0.6595990318629605, + "grad_norm": 1.4510712623596191, + "learning_rate": 5.1955981584318716e-05, + "loss": 0.7348, + "step": 28138 + }, + { + "epoch": 0.6596224734377656, + "grad_norm": 0.7708653807640076, + "learning_rate": 5.1949521391707526e-05, + "loss": 0.0896, + "step": 28139 + }, + { + "epoch": 0.6596459150125705, + "grad_norm": 0.8466702103614807, + "learning_rate": 5.1943061459819575e-05, + "loss": 0.0273, + "step": 28140 + }, + { + "epoch": 0.6596693565873756, + "grad_norm": 0.6178320646286011, + "learning_rate": 5.193660178868993e-05, + "loss": 0.0613, + "step": 28141 + }, + { + "epoch": 0.6596927981621805, + "grad_norm": 3.0236077308654785, + "learning_rate": 5.193014237835364e-05, + "loss": 0.2467, + "step": 28142 + }, + { + "epoch": 0.6597162397369856, + "grad_norm": 2.0282740592956543, + "learning_rate": 5.192368322884571e-05, + "loss": 0.1946, + "step": 28143 + }, + { + "epoch": 0.6597396813117905, + "grad_norm": 3.272963285446167, + "learning_rate": 5.191722434020125e-05, + "loss": 0.1872, + "step": 28144 + }, + { + "epoch": 0.6597631228865956, + "grad_norm": 3.092423677444458, + "learning_rate": 5.19107657124553e-05, + "loss": 0.2975, + "step": 28145 + }, + { + "epoch": 0.6597865644614005, + "grad_norm": 2.6792497634887695, + "learning_rate": 5.190430734564289e-05, + "loss": 0.156, + "step": 28146 + }, + { + "epoch": 0.6598100060362055, + "grad_norm": 2.0912582874298096, + "learning_rate": 5.1897849239799056e-05, + "loss": 0.1625, + "step": 28147 + }, + { + "epoch": 0.6598334476110105, + "grad_norm": 0.5702987313270569, + "learning_rate": 5.1891391394958844e-05, + "loss": 0.0546, + "step": 28148 + }, + { + "epoch": 0.6598568891858155, + "grad_norm": 2.5213782787323, + "learning_rate": 5.1884933811157245e-05, + "loss": 1.3669, + "step": 28149 + }, + { + "epoch": 0.6598803307606205, + "grad_norm": 2.5982725620269775, + "learning_rate": 5.187847648842936e-05, + "loss": 0.3511, + "step": 28150 + }, + { + "epoch": 0.6599037723354255, + "grad_norm": 1.760087013244629, + "learning_rate": 5.1872019426810245e-05, + "loss": 0.1989, + "step": 28151 + }, + { + "epoch": 0.6599272139102305, + "grad_norm": 2.815284490585327, + "learning_rate": 5.186556262633492e-05, + "loss": 0.2109, + "step": 28152 + }, + { + "epoch": 0.6599506554850355, + "grad_norm": 3.6638081073760986, + "learning_rate": 5.185910608703839e-05, + "loss": 0.2658, + "step": 28153 + }, + { + "epoch": 0.6599740970598404, + "grad_norm": 3.7818377017974854, + "learning_rate": 5.185264980895571e-05, + "loss": 0.1768, + "step": 28154 + }, + { + "epoch": 0.6599975386346455, + "grad_norm": 0.9244245886802673, + "learning_rate": 5.184619379212191e-05, + "loss": 0.0608, + "step": 28155 + }, + { + "epoch": 0.6600209802094504, + "grad_norm": 0.4784632921218872, + "learning_rate": 5.183973803657197e-05, + "loss": 0.0375, + "step": 28156 + }, + { + "epoch": 0.6600444217842555, + "grad_norm": 1.9468250274658203, + "learning_rate": 5.183328254234101e-05, + "loss": 0.0996, + "step": 28157 + }, + { + "epoch": 0.6600678633590604, + "grad_norm": 3.220212936401367, + "learning_rate": 5.182682730946402e-05, + "loss": 0.3384, + "step": 28158 + }, + { + "epoch": 0.6600913049338655, + "grad_norm": 4.009212017059326, + "learning_rate": 5.182037233797601e-05, + "loss": 0.3249, + "step": 28159 + }, + { + "epoch": 0.6601147465086704, + "grad_norm": 2.898587465286255, + "learning_rate": 5.181391762791201e-05, + "loss": 0.3202, + "step": 28160 + }, + { + "epoch": 0.6601381880834755, + "grad_norm": 1.5748928785324097, + "learning_rate": 5.180746317930705e-05, + "loss": 0.1223, + "step": 28161 + }, + { + "epoch": 0.6601616296582804, + "grad_norm": 2.410156726837158, + "learning_rate": 5.180100899219612e-05, + "loss": 0.1533, + "step": 28162 + }, + { + "epoch": 0.6601850712330855, + "grad_norm": 1.8914519548416138, + "learning_rate": 5.179455506661426e-05, + "loss": 0.1378, + "step": 28163 + }, + { + "epoch": 0.6602085128078904, + "grad_norm": 0.9680304527282715, + "learning_rate": 5.1788101402596544e-05, + "loss": 0.0732, + "step": 28164 + }, + { + "epoch": 0.6602319543826954, + "grad_norm": 2.66843843460083, + "learning_rate": 5.178164800017795e-05, + "loss": 0.3253, + "step": 28165 + }, + { + "epoch": 0.6602553959575004, + "grad_norm": 3.249211072921753, + "learning_rate": 5.1775194859393484e-05, + "loss": 0.2018, + "step": 28166 + }, + { + "epoch": 0.6602788375323054, + "grad_norm": 3.522146463394165, + "learning_rate": 5.1768741980278165e-05, + "loss": 0.2423, + "step": 28167 + }, + { + "epoch": 0.6603022791071104, + "grad_norm": 0.9465515613555908, + "learning_rate": 5.176228936286701e-05, + "loss": 0.0903, + "step": 28168 + }, + { + "epoch": 0.6603257206819154, + "grad_norm": 0.8960002660751343, + "learning_rate": 5.1755837007194994e-05, + "loss": 0.0463, + "step": 28169 + }, + { + "epoch": 0.6603491622567204, + "grad_norm": 0.710445761680603, + "learning_rate": 5.1749384913297194e-05, + "loss": 0.0427, + "step": 28170 + }, + { + "epoch": 0.6603726038315254, + "grad_norm": 2.3369944095611572, + "learning_rate": 5.174293308120859e-05, + "loss": 0.0995, + "step": 28171 + }, + { + "epoch": 0.6603960454063303, + "grad_norm": 2.2961649894714355, + "learning_rate": 5.17364815109642e-05, + "loss": 0.2975, + "step": 28172 + }, + { + "epoch": 0.6604194869811354, + "grad_norm": 4.397153854370117, + "learning_rate": 5.173003020259901e-05, + "loss": 0.2483, + "step": 28173 + }, + { + "epoch": 0.6604429285559403, + "grad_norm": 2.7962677478790283, + "learning_rate": 5.1723579156147986e-05, + "loss": 0.4114, + "step": 28174 + }, + { + "epoch": 0.6604663701307454, + "grad_norm": 1.2819322347640991, + "learning_rate": 5.1717128371646226e-05, + "loss": 0.2004, + "step": 28175 + }, + { + "epoch": 0.6604898117055504, + "grad_norm": 2.8293259143829346, + "learning_rate": 5.171067784912864e-05, + "loss": 0.0884, + "step": 28176 + }, + { + "epoch": 0.6605132532803554, + "grad_norm": 3.74051833152771, + "learning_rate": 5.1704227588630326e-05, + "loss": 0.0793, + "step": 28177 + }, + { + "epoch": 0.6605366948551604, + "grad_norm": 0.9238530993461609, + "learning_rate": 5.1697777590186214e-05, + "loss": 0.0545, + "step": 28178 + }, + { + "epoch": 0.6605601364299654, + "grad_norm": 2.32209849357605, + "learning_rate": 5.169132785383133e-05, + "loss": 0.1145, + "step": 28179 + }, + { + "epoch": 0.6605835780047704, + "grad_norm": 1.3372457027435303, + "learning_rate": 5.1684878379600665e-05, + "loss": 0.102, + "step": 28180 + }, + { + "epoch": 0.6606070195795753, + "grad_norm": 4.081599235534668, + "learning_rate": 5.167842916752916e-05, + "loss": 0.1889, + "step": 28181 + }, + { + "epoch": 0.6606304611543804, + "grad_norm": 1.168635606765747, + "learning_rate": 5.167198021765189e-05, + "loss": 0.1388, + "step": 28182 + }, + { + "epoch": 0.6606539027291853, + "grad_norm": 0.5146289467811584, + "learning_rate": 5.166553153000383e-05, + "loss": 0.0774, + "step": 28183 + }, + { + "epoch": 0.6606773443039904, + "grad_norm": 2.9645698070526123, + "learning_rate": 5.1659083104619955e-05, + "loss": 0.319, + "step": 28184 + }, + { + "epoch": 0.6607007858787953, + "grad_norm": 1.7011516094207764, + "learning_rate": 5.165263494153524e-05, + "loss": 0.0818, + "step": 28185 + }, + { + "epoch": 0.6607242274536004, + "grad_norm": 3.270465850830078, + "learning_rate": 5.1646187040784655e-05, + "loss": 0.1706, + "step": 28186 + }, + { + "epoch": 0.6607476690284053, + "grad_norm": 3.05819034576416, + "learning_rate": 5.1639739402403255e-05, + "loss": 0.2438, + "step": 28187 + }, + { + "epoch": 0.6607711106032104, + "grad_norm": 2.325256586074829, + "learning_rate": 5.163329202642595e-05, + "loss": 0.1815, + "step": 28188 + }, + { + "epoch": 0.6607945521780153, + "grad_norm": 1.3959641456604004, + "learning_rate": 5.1626844912887806e-05, + "loss": 0.1024, + "step": 28189 + }, + { + "epoch": 0.6608179937528204, + "grad_norm": 3.647002696990967, + "learning_rate": 5.162039806182376e-05, + "loss": 0.7611, + "step": 28190 + }, + { + "epoch": 0.6608414353276253, + "grad_norm": 1.5797195434570312, + "learning_rate": 5.1613951473268794e-05, + "loss": 0.8177, + "step": 28191 + }, + { + "epoch": 0.6608648769024303, + "grad_norm": 1.470414161682129, + "learning_rate": 5.160750514725788e-05, + "loss": 0.1866, + "step": 28192 + }, + { + "epoch": 0.6608883184772353, + "grad_norm": 2.104188919067383, + "learning_rate": 5.160105908382602e-05, + "loss": 0.1795, + "step": 28193 + }, + { + "epoch": 0.6609117600520403, + "grad_norm": 1.599626064300537, + "learning_rate": 5.159461328300812e-05, + "loss": 0.1524, + "step": 28194 + }, + { + "epoch": 0.6609352016268453, + "grad_norm": 2.0943639278411865, + "learning_rate": 5.1588167744839254e-05, + "loss": 0.148, + "step": 28195 + }, + { + "epoch": 0.6609586432016503, + "grad_norm": 1.8101170063018799, + "learning_rate": 5.1581722469354354e-05, + "loss": 0.1701, + "step": 28196 + }, + { + "epoch": 0.6609820847764553, + "grad_norm": 1.0041210651397705, + "learning_rate": 5.157527745658839e-05, + "loss": 0.1236, + "step": 28197 + }, + { + "epoch": 0.6610055263512603, + "grad_norm": 3.6212456226348877, + "learning_rate": 5.1568832706576285e-05, + "loss": 0.2413, + "step": 28198 + }, + { + "epoch": 0.6610289679260652, + "grad_norm": 1.9389005899429321, + "learning_rate": 5.156238821935311e-05, + "loss": 0.0883, + "step": 28199 + }, + { + "epoch": 0.6610524095008703, + "grad_norm": 1.1463549137115479, + "learning_rate": 5.1555943994953756e-05, + "loss": 0.0948, + "step": 28200 + }, + { + "epoch": 0.6610758510756752, + "grad_norm": 1.1904667615890503, + "learning_rate": 5.154950003341319e-05, + "loss": 0.1132, + "step": 28201 + }, + { + "epoch": 0.6610992926504803, + "grad_norm": 4.103997230529785, + "learning_rate": 5.154305633476644e-05, + "loss": 0.3577, + "step": 28202 + }, + { + "epoch": 0.6611227342252852, + "grad_norm": 3.7823381423950195, + "learning_rate": 5.153661289904842e-05, + "loss": 0.4686, + "step": 28203 + }, + { + "epoch": 0.6611461758000903, + "grad_norm": 3.2364113330841064, + "learning_rate": 5.15301697262941e-05, + "loss": 0.1501, + "step": 28204 + }, + { + "epoch": 0.6611696173748952, + "grad_norm": 3.0622987747192383, + "learning_rate": 5.1523726816538454e-05, + "loss": 0.3085, + "step": 28205 + }, + { + "epoch": 0.6611930589497003, + "grad_norm": 2.4310264587402344, + "learning_rate": 5.15172841698164e-05, + "loss": 0.1922, + "step": 28206 + }, + { + "epoch": 0.6612165005245052, + "grad_norm": 3.1922173500061035, + "learning_rate": 5.1510841786162924e-05, + "loss": 0.3739, + "step": 28207 + }, + { + "epoch": 0.6612399420993103, + "grad_norm": 1.73800528049469, + "learning_rate": 5.1504399665613e-05, + "loss": 0.0886, + "step": 28208 + }, + { + "epoch": 0.6612633836741152, + "grad_norm": 2.4253458976745605, + "learning_rate": 5.149795780820157e-05, + "loss": 0.1674, + "step": 28209 + }, + { + "epoch": 0.6612868252489202, + "grad_norm": 4.7786688804626465, + "learning_rate": 5.1491516213963555e-05, + "loss": 0.3592, + "step": 28210 + }, + { + "epoch": 0.6613102668237252, + "grad_norm": 2.887873411178589, + "learning_rate": 5.148507488293397e-05, + "loss": 0.1517, + "step": 28211 + }, + { + "epoch": 0.6613337083985302, + "grad_norm": 4.7344207763671875, + "learning_rate": 5.147863381514772e-05, + "loss": 0.4388, + "step": 28212 + }, + { + "epoch": 0.6613571499733352, + "grad_norm": 4.3013105392456055, + "learning_rate": 5.147219301063979e-05, + "loss": 0.4298, + "step": 28213 + }, + { + "epoch": 0.6613805915481402, + "grad_norm": 4.727625846862793, + "learning_rate": 5.146575246944505e-05, + "loss": 0.3191, + "step": 28214 + }, + { + "epoch": 0.6614040331229452, + "grad_norm": 2.950413703918457, + "learning_rate": 5.1459312191598544e-05, + "loss": 0.2217, + "step": 28215 + }, + { + "epoch": 0.6614274746977502, + "grad_norm": 2.8004236221313477, + "learning_rate": 5.145287217713518e-05, + "loss": 0.2096, + "step": 28216 + }, + { + "epoch": 0.6614509162725551, + "grad_norm": 3.972391128540039, + "learning_rate": 5.14464324260899e-05, + "loss": 0.3611, + "step": 28217 + }, + { + "epoch": 0.6614743578473602, + "grad_norm": 3.012335777282715, + "learning_rate": 5.143999293849764e-05, + "loss": 0.0924, + "step": 28218 + }, + { + "epoch": 0.6614977994221651, + "grad_norm": 0.6246747374534607, + "learning_rate": 5.143355371439329e-05, + "loss": 0.0243, + "step": 28219 + }, + { + "epoch": 0.6615212409969702, + "grad_norm": 3.4424657821655273, + "learning_rate": 5.14271147538119e-05, + "loss": 0.3765, + "step": 28220 + }, + { + "epoch": 0.6615446825717751, + "grad_norm": 3.9781646728515625, + "learning_rate": 5.142067605678834e-05, + "loss": 0.1875, + "step": 28221 + }, + { + "epoch": 0.6615681241465802, + "grad_norm": 4.701913356781006, + "learning_rate": 5.1414237623357534e-05, + "loss": 0.4316, + "step": 28222 + }, + { + "epoch": 0.6615915657213851, + "grad_norm": 2.992856740951538, + "learning_rate": 5.140779945355446e-05, + "loss": 0.5326, + "step": 28223 + }, + { + "epoch": 0.6616150072961902, + "grad_norm": 2.669369697570801, + "learning_rate": 5.140136154741404e-05, + "loss": 0.135, + "step": 28224 + }, + { + "epoch": 0.6616384488709951, + "grad_norm": 2.156785488128662, + "learning_rate": 5.13949239049712e-05, + "loss": 0.2337, + "step": 28225 + }, + { + "epoch": 0.6616618904458001, + "grad_norm": 2.032055139541626, + "learning_rate": 5.1388486526260824e-05, + "loss": 0.1843, + "step": 28226 + }, + { + "epoch": 0.6616853320206052, + "grad_norm": 2.559682607650757, + "learning_rate": 5.138204941131794e-05, + "loss": 0.2253, + "step": 28227 + }, + { + "epoch": 0.6617087735954101, + "grad_norm": 1.686932921409607, + "learning_rate": 5.137561256017742e-05, + "loss": 0.0864, + "step": 28228 + }, + { + "epoch": 0.6617322151702152, + "grad_norm": 3.2864503860473633, + "learning_rate": 5.136917597287419e-05, + "loss": 0.2175, + "step": 28229 + }, + { + "epoch": 0.6617556567450201, + "grad_norm": 1.384142518043518, + "learning_rate": 5.136273964944317e-05, + "loss": 1.6737, + "step": 28230 + }, + { + "epoch": 0.6617790983198252, + "grad_norm": 1.7845183610916138, + "learning_rate": 5.13563035899193e-05, + "loss": 0.1507, + "step": 28231 + }, + { + "epoch": 0.6618025398946301, + "grad_norm": 3.1895618438720703, + "learning_rate": 5.134986779433746e-05, + "loss": 0.3945, + "step": 28232 + }, + { + "epoch": 0.6618259814694352, + "grad_norm": 0.39544662833213806, + "learning_rate": 5.134343226273264e-05, + "loss": 0.0275, + "step": 28233 + }, + { + "epoch": 0.6618494230442401, + "grad_norm": 2.190981388092041, + "learning_rate": 5.1336996995139676e-05, + "loss": 0.1873, + "step": 28234 + }, + { + "epoch": 0.6618728646190452, + "grad_norm": 0.35315731167793274, + "learning_rate": 5.1330561991593586e-05, + "loss": 0.0206, + "step": 28235 + }, + { + "epoch": 0.6618963061938501, + "grad_norm": 2.3462331295013428, + "learning_rate": 5.132412725212923e-05, + "loss": 0.2629, + "step": 28236 + }, + { + "epoch": 0.6619197477686551, + "grad_norm": 2.3182501792907715, + "learning_rate": 5.1317692776781535e-05, + "loss": 0.1584, + "step": 28237 + }, + { + "epoch": 0.6619431893434601, + "grad_norm": 4.666834354400635, + "learning_rate": 5.131125856558541e-05, + "loss": 0.3703, + "step": 28238 + }, + { + "epoch": 0.6619666309182651, + "grad_norm": 2.208277702331543, + "learning_rate": 5.130482461857571e-05, + "loss": 0.1444, + "step": 28239 + }, + { + "epoch": 0.6619900724930701, + "grad_norm": 6.011932373046875, + "learning_rate": 5.1298390935787454e-05, + "loss": 0.3079, + "step": 28240 + }, + { + "epoch": 0.6620135140678751, + "grad_norm": 2.239097833633423, + "learning_rate": 5.129195751725549e-05, + "loss": 1.639, + "step": 28241 + }, + { + "epoch": 0.6620369556426801, + "grad_norm": 1.6400445699691772, + "learning_rate": 5.1285524363014734e-05, + "loss": 0.0988, + "step": 28242 + }, + { + "epoch": 0.6620603972174851, + "grad_norm": 2.8479433059692383, + "learning_rate": 5.127909147310009e-05, + "loss": 0.3167, + "step": 28243 + }, + { + "epoch": 0.66208383879229, + "grad_norm": 2.5109636783599854, + "learning_rate": 5.127265884754647e-05, + "loss": 0.1202, + "step": 28244 + }, + { + "epoch": 0.6621072803670951, + "grad_norm": 5.18613862991333, + "learning_rate": 5.1266226486388746e-05, + "loss": 0.7164, + "step": 28245 + }, + { + "epoch": 0.6621307219419, + "grad_norm": 2.382296562194824, + "learning_rate": 5.1259794389661865e-05, + "loss": 0.1246, + "step": 28246 + }, + { + "epoch": 0.6621541635167051, + "grad_norm": 2.1361849308013916, + "learning_rate": 5.1253362557400686e-05, + "loss": 0.2246, + "step": 28247 + }, + { + "epoch": 0.66217760509151, + "grad_norm": 2.5043327808380127, + "learning_rate": 5.1246930989640174e-05, + "loss": 0.3554, + "step": 28248 + }, + { + "epoch": 0.6622010466663151, + "grad_norm": 3.10786771774292, + "learning_rate": 5.124049968641518e-05, + "loss": 0.1805, + "step": 28249 + }, + { + "epoch": 0.66222448824112, + "grad_norm": 4.014123916625977, + "learning_rate": 5.123406864776059e-05, + "loss": 0.3605, + "step": 28250 + }, + { + "epoch": 0.6622479298159251, + "grad_norm": 0.3798697888851166, + "learning_rate": 5.122763787371134e-05, + "loss": 0.0247, + "step": 28251 + }, + { + "epoch": 0.66227137139073, + "grad_norm": 3.422424554824829, + "learning_rate": 5.1221207364302246e-05, + "loss": 0.123, + "step": 28252 + }, + { + "epoch": 0.662294812965535, + "grad_norm": 2.7939984798431396, + "learning_rate": 5.121477711956829e-05, + "loss": 0.1621, + "step": 28253 + }, + { + "epoch": 0.66231825454034, + "grad_norm": 3.1748268604278564, + "learning_rate": 5.120834713954432e-05, + "loss": 0.287, + "step": 28254 + }, + { + "epoch": 0.662341696115145, + "grad_norm": 2.120260000228882, + "learning_rate": 5.1201917424265246e-05, + "loss": 0.2001, + "step": 28255 + }, + { + "epoch": 0.66236513768995, + "grad_norm": 1.2063570022583008, + "learning_rate": 5.1195487973765935e-05, + "loss": 0.0968, + "step": 28256 + }, + { + "epoch": 0.662388579264755, + "grad_norm": 1.782667875289917, + "learning_rate": 5.1189058788081236e-05, + "loss": 0.1546, + "step": 28257 + }, + { + "epoch": 0.66241202083956, + "grad_norm": 1.0713030099868774, + "learning_rate": 5.1182629867246114e-05, + "loss": 0.1128, + "step": 28258 + }, + { + "epoch": 0.662435462414365, + "grad_norm": 0.637229323387146, + "learning_rate": 5.117620121129538e-05, + "loss": 0.0429, + "step": 28259 + }, + { + "epoch": 0.66245890398917, + "grad_norm": 2.2910938262939453, + "learning_rate": 5.1169772820263985e-05, + "loss": 0.192, + "step": 28260 + }, + { + "epoch": 0.662482345563975, + "grad_norm": 1.1231927871704102, + "learning_rate": 5.116334469418678e-05, + "loss": 0.7946, + "step": 28261 + }, + { + "epoch": 0.66250578713878, + "grad_norm": 3.4425270557403564, + "learning_rate": 5.115691683309865e-05, + "loss": 0.2904, + "step": 28262 + }, + { + "epoch": 0.662529228713585, + "grad_norm": 3.6606130599975586, + "learning_rate": 5.1150489237034446e-05, + "loss": 0.3343, + "step": 28263 + }, + { + "epoch": 0.6625526702883899, + "grad_norm": 1.9574992656707764, + "learning_rate": 5.1144061906029026e-05, + "loss": 0.1448, + "step": 28264 + }, + { + "epoch": 0.662576111863195, + "grad_norm": 0.8368389010429382, + "learning_rate": 5.1137634840117346e-05, + "loss": 0.0831, + "step": 28265 + }, + { + "epoch": 0.6625995534379999, + "grad_norm": 1.270577311515808, + "learning_rate": 5.113120803933422e-05, + "loss": 0.1704, + "step": 28266 + }, + { + "epoch": 0.662622995012805, + "grad_norm": 4.541891574859619, + "learning_rate": 5.1124781503714546e-05, + "loss": 0.5326, + "step": 28267 + }, + { + "epoch": 0.6626464365876099, + "grad_norm": 3.7175681591033936, + "learning_rate": 5.111835523329318e-05, + "loss": 0.42, + "step": 28268 + }, + { + "epoch": 0.662669878162415, + "grad_norm": 3.343836545944214, + "learning_rate": 5.111192922810499e-05, + "loss": 0.3849, + "step": 28269 + }, + { + "epoch": 0.6626933197372199, + "grad_norm": 4.39510440826416, + "learning_rate": 5.1105503488184805e-05, + "loss": 0.2296, + "step": 28270 + }, + { + "epoch": 0.662716761312025, + "grad_norm": 5.34102725982666, + "learning_rate": 5.109907801356752e-05, + "loss": 0.4639, + "step": 28271 + }, + { + "epoch": 0.6627402028868299, + "grad_norm": 0.8757540583610535, + "learning_rate": 5.1092652804288075e-05, + "loss": 0.0786, + "step": 28272 + }, + { + "epoch": 0.6627636444616349, + "grad_norm": 2.341991901397705, + "learning_rate": 5.1086227860381264e-05, + "loss": 0.1994, + "step": 28273 + }, + { + "epoch": 0.6627870860364399, + "grad_norm": 0.9278847575187683, + "learning_rate": 5.1079803181881955e-05, + "loss": 0.0543, + "step": 28274 + }, + { + "epoch": 0.6628105276112449, + "grad_norm": 1.521475911140442, + "learning_rate": 5.107337876882501e-05, + "loss": 0.0571, + "step": 28275 + }, + { + "epoch": 0.6628339691860499, + "grad_norm": 1.1420550346374512, + "learning_rate": 5.106695462124529e-05, + "loss": 0.1053, + "step": 28276 + }, + { + "epoch": 0.6628574107608549, + "grad_norm": 5.147447109222412, + "learning_rate": 5.1060530739177605e-05, + "loss": 0.3636, + "step": 28277 + }, + { + "epoch": 0.66288085233566, + "grad_norm": 1.7807008028030396, + "learning_rate": 5.1054107122656905e-05, + "loss": 0.1278, + "step": 28278 + }, + { + "epoch": 0.6629042939104649, + "grad_norm": 2.37080454826355, + "learning_rate": 5.104768377171798e-05, + "loss": 0.202, + "step": 28279 + }, + { + "epoch": 0.66292773548527, + "grad_norm": 3.8902907371520996, + "learning_rate": 5.104126068639572e-05, + "loss": 0.2412, + "step": 28280 + }, + { + "epoch": 0.6629511770600749, + "grad_norm": 0.7269631624221802, + "learning_rate": 5.103483786672494e-05, + "loss": 0.0946, + "step": 28281 + }, + { + "epoch": 0.6629746186348799, + "grad_norm": 3.5980615615844727, + "learning_rate": 5.102841531274052e-05, + "loss": 0.2122, + "step": 28282 + }, + { + "epoch": 0.6629980602096849, + "grad_norm": 4.027834892272949, + "learning_rate": 5.1021993024477234e-05, + "loss": 0.3471, + "step": 28283 + }, + { + "epoch": 0.6630215017844899, + "grad_norm": 3.0525319576263428, + "learning_rate": 5.101557100197001e-05, + "loss": 0.2395, + "step": 28284 + }, + { + "epoch": 0.6630449433592949, + "grad_norm": 5.642755031585693, + "learning_rate": 5.100914924525371e-05, + "loss": 0.5146, + "step": 28285 + }, + { + "epoch": 0.6630683849340999, + "grad_norm": 3.190944194793701, + "learning_rate": 5.1002727754363144e-05, + "loss": 0.2066, + "step": 28286 + }, + { + "epoch": 0.6630918265089049, + "grad_norm": 2.415699005126953, + "learning_rate": 5.099630652933315e-05, + "loss": 1.5242, + "step": 28287 + }, + { + "epoch": 0.6631152680837099, + "grad_norm": 2.45320725440979, + "learning_rate": 5.098988557019857e-05, + "loss": 0.2267, + "step": 28288 + }, + { + "epoch": 0.6631387096585148, + "grad_norm": 4.174241065979004, + "learning_rate": 5.0983464876994256e-05, + "loss": 0.2796, + "step": 28289 + }, + { + "epoch": 0.6631621512333199, + "grad_norm": 2.875688314437866, + "learning_rate": 5.097704444975501e-05, + "loss": 0.2225, + "step": 28290 + }, + { + "epoch": 0.6631855928081248, + "grad_norm": 3.839012622833252, + "learning_rate": 5.0970624288515734e-05, + "loss": 0.3815, + "step": 28291 + }, + { + "epoch": 0.6632090343829299, + "grad_norm": 2.893841505050659, + "learning_rate": 5.0964204393311225e-05, + "loss": 0.3262, + "step": 28292 + }, + { + "epoch": 0.6632324759577348, + "grad_norm": 2.499901056289673, + "learning_rate": 5.095778476417632e-05, + "loss": 1.6378, + "step": 28293 + }, + { + "epoch": 0.6632559175325399, + "grad_norm": 2.1624085903167725, + "learning_rate": 5.0951365401145845e-05, + "loss": 0.1934, + "step": 28294 + }, + { + "epoch": 0.6632793591073448, + "grad_norm": 0.8815268278121948, + "learning_rate": 5.0944946304254614e-05, + "loss": 0.0634, + "step": 28295 + }, + { + "epoch": 0.6633028006821499, + "grad_norm": 0.898033857345581, + "learning_rate": 5.093852747353749e-05, + "loss": 0.0736, + "step": 28296 + }, + { + "epoch": 0.6633262422569548, + "grad_norm": 4.19094705581665, + "learning_rate": 5.093210890902933e-05, + "loss": 0.2283, + "step": 28297 + }, + { + "epoch": 0.6633496838317599, + "grad_norm": 2.919264078140259, + "learning_rate": 5.092569061076491e-05, + "loss": 0.2596, + "step": 28298 + }, + { + "epoch": 0.6633731254065648, + "grad_norm": 1.1518278121948242, + "learning_rate": 5.0919272578779086e-05, + "loss": 0.0577, + "step": 28299 + }, + { + "epoch": 0.6633965669813698, + "grad_norm": 3.1323447227478027, + "learning_rate": 5.091285481310668e-05, + "loss": 1.9322, + "step": 28300 + }, + { + "epoch": 0.6634200085561748, + "grad_norm": 4.17408561706543, + "learning_rate": 5.090643731378249e-05, + "loss": 0.4826, + "step": 28301 + }, + { + "epoch": 0.6634434501309798, + "grad_norm": 1.2833256721496582, + "learning_rate": 5.090002008084132e-05, + "loss": 0.0403, + "step": 28302 + }, + { + "epoch": 0.6634668917057848, + "grad_norm": 0.6273460388183594, + "learning_rate": 5.089360311431806e-05, + "loss": 0.0398, + "step": 28303 + }, + { + "epoch": 0.6634903332805898, + "grad_norm": 3.359754800796509, + "learning_rate": 5.088718641424748e-05, + "loss": 0.3684, + "step": 28304 + }, + { + "epoch": 0.6635137748553948, + "grad_norm": 3.079286813735962, + "learning_rate": 5.08807699806644e-05, + "loss": 0.4645, + "step": 28305 + }, + { + "epoch": 0.6635372164301998, + "grad_norm": 1.033774495124817, + "learning_rate": 5.087435381360366e-05, + "loss": 0.0284, + "step": 28306 + }, + { + "epoch": 0.6635606580050047, + "grad_norm": 1.8404463529586792, + "learning_rate": 5.0867937913100015e-05, + "loss": 0.8688, + "step": 28307 + }, + { + "epoch": 0.6635840995798098, + "grad_norm": 3.9110400676727295, + "learning_rate": 5.086152227918837e-05, + "loss": 0.3546, + "step": 28308 + }, + { + "epoch": 0.6636075411546147, + "grad_norm": 2.503410577774048, + "learning_rate": 5.085510691190343e-05, + "loss": 0.083, + "step": 28309 + }, + { + "epoch": 0.6636309827294198, + "grad_norm": 4.56255578994751, + "learning_rate": 5.0848691811280115e-05, + "loss": 0.4539, + "step": 28310 + }, + { + "epoch": 0.6636544243042247, + "grad_norm": 2.8042590618133545, + "learning_rate": 5.084227697735317e-05, + "loss": 0.1227, + "step": 28311 + }, + { + "epoch": 0.6636778658790298, + "grad_norm": 4.264058589935303, + "learning_rate": 5.083586241015741e-05, + "loss": 0.4662, + "step": 28312 + }, + { + "epoch": 0.6637013074538347, + "grad_norm": 2.1313679218292236, + "learning_rate": 5.082944810972765e-05, + "loss": 0.0653, + "step": 28313 + }, + { + "epoch": 0.6637247490286398, + "grad_norm": 1.2368903160095215, + "learning_rate": 5.0823034076098684e-05, + "loss": 0.095, + "step": 28314 + }, + { + "epoch": 0.6637481906034447, + "grad_norm": 2.0914642810821533, + "learning_rate": 5.0816620309305274e-05, + "loss": 0.166, + "step": 28315 + }, + { + "epoch": 0.6637716321782497, + "grad_norm": 2.5835154056549072, + "learning_rate": 5.081020680938231e-05, + "loss": 0.1323, + "step": 28316 + }, + { + "epoch": 0.6637950737530547, + "grad_norm": 0.665620744228363, + "learning_rate": 5.0803793576364555e-05, + "loss": 0.0384, + "step": 28317 + }, + { + "epoch": 0.6638185153278597, + "grad_norm": 1.8760873079299927, + "learning_rate": 5.079738061028679e-05, + "loss": 0.181, + "step": 28318 + }, + { + "epoch": 0.6638419569026647, + "grad_norm": 2.891042470932007, + "learning_rate": 5.0790967911183826e-05, + "loss": 0.2333, + "step": 28319 + }, + { + "epoch": 0.6638653984774697, + "grad_norm": 2.64621901512146, + "learning_rate": 5.078455547909041e-05, + "loss": 0.1402, + "step": 28320 + }, + { + "epoch": 0.6638888400522747, + "grad_norm": 2.6976864337921143, + "learning_rate": 5.077814331404143e-05, + "loss": 0.2353, + "step": 28321 + }, + { + "epoch": 0.6639122816270797, + "grad_norm": 3.217763900756836, + "learning_rate": 5.077173141607157e-05, + "loss": 0.2897, + "step": 28322 + }, + { + "epoch": 0.6639357232018847, + "grad_norm": 3.16839861869812, + "learning_rate": 5.076531978521573e-05, + "loss": 0.2318, + "step": 28323 + }, + { + "epoch": 0.6639591647766897, + "grad_norm": 2.349750280380249, + "learning_rate": 5.0758908421508655e-05, + "loss": 0.251, + "step": 28324 + }, + { + "epoch": 0.6639826063514946, + "grad_norm": 1.0793956518173218, + "learning_rate": 5.0752497324985106e-05, + "loss": 0.0589, + "step": 28325 + }, + { + "epoch": 0.6640060479262997, + "grad_norm": 2.824489116668701, + "learning_rate": 5.074608649567991e-05, + "loss": 0.1587, + "step": 28326 + }, + { + "epoch": 0.6640294895011046, + "grad_norm": 0.9816007018089294, + "learning_rate": 5.0739675933627785e-05, + "loss": 0.0773, + "step": 28327 + }, + { + "epoch": 0.6640529310759097, + "grad_norm": 1.1535605192184448, + "learning_rate": 5.073326563886361e-05, + "loss": 0.1065, + "step": 28328 + }, + { + "epoch": 0.6640763726507147, + "grad_norm": 3.67158842086792, + "learning_rate": 5.072685561142212e-05, + "loss": 0.2688, + "step": 28329 + }, + { + "epoch": 0.6640998142255197, + "grad_norm": 3.779181480407715, + "learning_rate": 5.0720445851338104e-05, + "loss": 0.233, + "step": 28330 + }, + { + "epoch": 0.6641232558003247, + "grad_norm": 0.47501450777053833, + "learning_rate": 5.071403635864631e-05, + "loss": 0.0354, + "step": 28331 + }, + { + "epoch": 0.6641466973751297, + "grad_norm": 2.155797004699707, + "learning_rate": 5.070762713338153e-05, + "loss": 0.0979, + "step": 28332 + }, + { + "epoch": 0.6641701389499347, + "grad_norm": 0.9342397451400757, + "learning_rate": 5.070121817557858e-05, + "loss": 0.0863, + "step": 28333 + }, + { + "epoch": 0.6641935805247396, + "grad_norm": 3.362729072570801, + "learning_rate": 5.069480948527217e-05, + "loss": 0.1491, + "step": 28334 + }, + { + "epoch": 0.6642170220995447, + "grad_norm": 2.5528697967529297, + "learning_rate": 5.068840106249712e-05, + "loss": 0.2202, + "step": 28335 + }, + { + "epoch": 0.6642404636743496, + "grad_norm": 5.136819362640381, + "learning_rate": 5.0681992907288234e-05, + "loss": 0.3389, + "step": 28336 + }, + { + "epoch": 0.6642639052491547, + "grad_norm": 7.680203914642334, + "learning_rate": 5.067558501968021e-05, + "loss": 0.6567, + "step": 28337 + }, + { + "epoch": 0.6642873468239596, + "grad_norm": 0.9506316781044006, + "learning_rate": 5.0669177399707865e-05, + "loss": 0.0559, + "step": 28338 + }, + { + "epoch": 0.6643107883987647, + "grad_norm": 2.070315361022949, + "learning_rate": 5.066277004740595e-05, + "loss": 0.1601, + "step": 28339 + }, + { + "epoch": 0.6643342299735696, + "grad_norm": 3.5449509620666504, + "learning_rate": 5.065636296280919e-05, + "loss": 0.2698, + "step": 28340 + }, + { + "epoch": 0.6643576715483747, + "grad_norm": 1.4764000177383423, + "learning_rate": 5.064995614595243e-05, + "loss": 0.0951, + "step": 28341 + }, + { + "epoch": 0.6643811131231796, + "grad_norm": 3.6452419757843018, + "learning_rate": 5.06435495968704e-05, + "loss": 0.2912, + "step": 28342 + }, + { + "epoch": 0.6644045546979847, + "grad_norm": 0.9538125395774841, + "learning_rate": 5.063714331559786e-05, + "loss": 0.0562, + "step": 28343 + }, + { + "epoch": 0.6644279962727896, + "grad_norm": 2.9386115074157715, + "learning_rate": 5.0630737302169515e-05, + "loss": 0.2847, + "step": 28344 + }, + { + "epoch": 0.6644514378475946, + "grad_norm": 0.5301110744476318, + "learning_rate": 5.062433155662023e-05, + "loss": 0.0387, + "step": 28345 + }, + { + "epoch": 0.6644748794223996, + "grad_norm": 5.313811779022217, + "learning_rate": 5.0617926078984704e-05, + "loss": 0.4845, + "step": 28346 + }, + { + "epoch": 0.6644983209972046, + "grad_norm": 3.3412415981292725, + "learning_rate": 5.061152086929766e-05, + "loss": 0.1567, + "step": 28347 + }, + { + "epoch": 0.6645217625720096, + "grad_norm": 2.0398287773132324, + "learning_rate": 5.060511592759394e-05, + "loss": 0.137, + "step": 28348 + }, + { + "epoch": 0.6645452041468146, + "grad_norm": 3.323021411895752, + "learning_rate": 5.059871125390824e-05, + "loss": 0.2693, + "step": 28349 + }, + { + "epoch": 0.6645686457216196, + "grad_norm": 0.9689208269119263, + "learning_rate": 5.059230684827532e-05, + "loss": 0.0406, + "step": 28350 + }, + { + "epoch": 0.6645920872964246, + "grad_norm": 3.622556686401367, + "learning_rate": 5.0585902710729935e-05, + "loss": 0.3511, + "step": 28351 + }, + { + "epoch": 0.6646155288712295, + "grad_norm": 3.530336618423462, + "learning_rate": 5.0579498841306835e-05, + "loss": 0.3822, + "step": 28352 + }, + { + "epoch": 0.6646389704460346, + "grad_norm": 1.1799217462539673, + "learning_rate": 5.0573095240040726e-05, + "loss": 0.1285, + "step": 28353 + }, + { + "epoch": 0.6646624120208395, + "grad_norm": 3.0185465812683105, + "learning_rate": 5.056669190696644e-05, + "loss": 0.3705, + "step": 28354 + }, + { + "epoch": 0.6646858535956446, + "grad_norm": 2.155280828475952, + "learning_rate": 5.056028884211865e-05, + "loss": 0.2838, + "step": 28355 + }, + { + "epoch": 0.6647092951704495, + "grad_norm": 2.7939531803131104, + "learning_rate": 5.0553886045532106e-05, + "loss": 0.1855, + "step": 28356 + }, + { + "epoch": 0.6647327367452546, + "grad_norm": 2.440624952316284, + "learning_rate": 5.054748351724159e-05, + "loss": 0.1263, + "step": 28357 + }, + { + "epoch": 0.6647561783200595, + "grad_norm": 4.090879440307617, + "learning_rate": 5.054108125728183e-05, + "loss": 0.3886, + "step": 28358 + }, + { + "epoch": 0.6647796198948646, + "grad_norm": 2.749093770980835, + "learning_rate": 5.0534679265687535e-05, + "loss": 0.1675, + "step": 28359 + }, + { + "epoch": 0.6648030614696695, + "grad_norm": 4.772921562194824, + "learning_rate": 5.052827754249344e-05, + "loss": 0.4798, + "step": 28360 + }, + { + "epoch": 0.6648265030444745, + "grad_norm": 4.773016929626465, + "learning_rate": 5.0521876087734334e-05, + "loss": 0.3608, + "step": 28361 + }, + { + "epoch": 0.6648499446192795, + "grad_norm": 0.9116702675819397, + "learning_rate": 5.0515474901444924e-05, + "loss": 0.0847, + "step": 28362 + }, + { + "epoch": 0.6648733861940845, + "grad_norm": 2.9989593029022217, + "learning_rate": 5.050907398365994e-05, + "loss": 0.205, + "step": 28363 + }, + { + "epoch": 0.6648968277688895, + "grad_norm": 0.4268876314163208, + "learning_rate": 5.050267333441411e-05, + "loss": 0.0294, + "step": 28364 + }, + { + "epoch": 0.6649202693436945, + "grad_norm": 0.7356321215629578, + "learning_rate": 5.049627295374213e-05, + "loss": 0.0659, + "step": 28365 + }, + { + "epoch": 0.6649437109184995, + "grad_norm": 2.7378978729248047, + "learning_rate": 5.048987284167879e-05, + "loss": 1.9651, + "step": 28366 + }, + { + "epoch": 0.6649671524933045, + "grad_norm": 3.566810369491577, + "learning_rate": 5.048347299825881e-05, + "loss": 0.1494, + "step": 28367 + }, + { + "epoch": 0.6649905940681095, + "grad_norm": 3.343799114227295, + "learning_rate": 5.0477073423516844e-05, + "loss": 0.2845, + "step": 28368 + }, + { + "epoch": 0.6650140356429145, + "grad_norm": 3.4699501991271973, + "learning_rate": 5.047067411748772e-05, + "loss": 0.2229, + "step": 28369 + }, + { + "epoch": 0.6650374772177194, + "grad_norm": 4.143496513366699, + "learning_rate": 5.04642750802061e-05, + "loss": 0.2778, + "step": 28370 + }, + { + "epoch": 0.6650609187925245, + "grad_norm": 3.534356117248535, + "learning_rate": 5.045787631170673e-05, + "loss": 0.3546, + "step": 28371 + }, + { + "epoch": 0.6650843603673294, + "grad_norm": 2.4583773612976074, + "learning_rate": 5.045147781202426e-05, + "loss": 0.2277, + "step": 28372 + }, + { + "epoch": 0.6651078019421345, + "grad_norm": 2.419442653656006, + "learning_rate": 5.044507958119352e-05, + "loss": 0.2405, + "step": 28373 + }, + { + "epoch": 0.6651312435169394, + "grad_norm": 2.9032604694366455, + "learning_rate": 5.043868161924916e-05, + "loss": 0.1559, + "step": 28374 + }, + { + "epoch": 0.6651546850917445, + "grad_norm": 0.9117690324783325, + "learning_rate": 5.043228392622592e-05, + "loss": 0.0545, + "step": 28375 + }, + { + "epoch": 0.6651781266665494, + "grad_norm": 2.4306623935699463, + "learning_rate": 5.042588650215849e-05, + "loss": 0.3477, + "step": 28376 + }, + { + "epoch": 0.6652015682413545, + "grad_norm": 1.637446403503418, + "learning_rate": 5.041948934708161e-05, + "loss": 0.0724, + "step": 28377 + }, + { + "epoch": 0.6652250098161594, + "grad_norm": 2.249821662902832, + "learning_rate": 5.041309246102992e-05, + "loss": 0.1287, + "step": 28378 + }, + { + "epoch": 0.6652484513909644, + "grad_norm": 0.41510656476020813, + "learning_rate": 5.040669584403822e-05, + "loss": 0.0368, + "step": 28379 + }, + { + "epoch": 0.6652718929657695, + "grad_norm": 0.7023733854293823, + "learning_rate": 5.0400299496141156e-05, + "loss": 0.0719, + "step": 28380 + }, + { + "epoch": 0.6652953345405744, + "grad_norm": 2.487272024154663, + "learning_rate": 5.0393903417373486e-05, + "loss": 0.1733, + "step": 28381 + }, + { + "epoch": 0.6653187761153795, + "grad_norm": 3.1573495864868164, + "learning_rate": 5.0387507607769915e-05, + "loss": 0.3121, + "step": 28382 + }, + { + "epoch": 0.6653422176901844, + "grad_norm": 3.25243878364563, + "learning_rate": 5.03811120673651e-05, + "loss": 0.1374, + "step": 28383 + }, + { + "epoch": 0.6653656592649895, + "grad_norm": 2.5888609886169434, + "learning_rate": 5.037471679619377e-05, + "loss": 0.5148, + "step": 28384 + }, + { + "epoch": 0.6653891008397944, + "grad_norm": 3.1204423904418945, + "learning_rate": 5.0368321794290585e-05, + "loss": 0.1571, + "step": 28385 + }, + { + "epoch": 0.6654125424145995, + "grad_norm": 2.7938880920410156, + "learning_rate": 5.0361927061690314e-05, + "loss": 0.1336, + "step": 28386 + }, + { + "epoch": 0.6654359839894044, + "grad_norm": 2.719733715057373, + "learning_rate": 5.035553259842765e-05, + "loss": 0.2505, + "step": 28387 + }, + { + "epoch": 0.6654594255642095, + "grad_norm": 1.5206685066223145, + "learning_rate": 5.034913840453723e-05, + "loss": 0.1027, + "step": 28388 + }, + { + "epoch": 0.6654828671390144, + "grad_norm": 5.1006364822387695, + "learning_rate": 5.0342744480053804e-05, + "loss": 0.3934, + "step": 28389 + }, + { + "epoch": 0.6655063087138194, + "grad_norm": 4.7629218101501465, + "learning_rate": 5.0336350825012027e-05, + "loss": 0.2793, + "step": 28390 + }, + { + "epoch": 0.6655297502886244, + "grad_norm": 0.24616892635822296, + "learning_rate": 5.032995743944657e-05, + "loss": 0.0187, + "step": 28391 + }, + { + "epoch": 0.6655531918634294, + "grad_norm": 3.5245401859283447, + "learning_rate": 5.03235643233922e-05, + "loss": 0.3508, + "step": 28392 + }, + { + "epoch": 0.6655766334382344, + "grad_norm": 0.37614917755126953, + "learning_rate": 5.031717147688353e-05, + "loss": 0.0293, + "step": 28393 + }, + { + "epoch": 0.6656000750130394, + "grad_norm": 2.9768643379211426, + "learning_rate": 5.031077889995532e-05, + "loss": 1.5658, + "step": 28394 + }, + { + "epoch": 0.6656235165878444, + "grad_norm": 4.0407609939575195, + "learning_rate": 5.030438659264223e-05, + "loss": 0.3366, + "step": 28395 + }, + { + "epoch": 0.6656469581626494, + "grad_norm": 2.4500386714935303, + "learning_rate": 5.029799455497893e-05, + "loss": 0.2252, + "step": 28396 + }, + { + "epoch": 0.6656703997374543, + "grad_norm": 1.780328392982483, + "learning_rate": 5.0291602787000114e-05, + "loss": 0.1619, + "step": 28397 + }, + { + "epoch": 0.6656938413122594, + "grad_norm": 4.387721538543701, + "learning_rate": 5.028521128874042e-05, + "loss": 0.275, + "step": 28398 + }, + { + "epoch": 0.6657172828870643, + "grad_norm": 0.3046056926250458, + "learning_rate": 5.02788200602346e-05, + "loss": 0.0235, + "step": 28399 + }, + { + "epoch": 0.6657407244618694, + "grad_norm": 0.2189294397830963, + "learning_rate": 5.02724291015173e-05, + "loss": 0.0217, + "step": 28400 + }, + { + "epoch": 0.6657641660366743, + "grad_norm": 0.7334334850311279, + "learning_rate": 5.02660384126232e-05, + "loss": 0.0517, + "step": 28401 + }, + { + "epoch": 0.6657876076114794, + "grad_norm": 2.4860079288482666, + "learning_rate": 5.025964799358698e-05, + "loss": 0.2053, + "step": 28402 + }, + { + "epoch": 0.6658110491862843, + "grad_norm": 0.9187993407249451, + "learning_rate": 5.025325784444326e-05, + "loss": 0.0773, + "step": 28403 + }, + { + "epoch": 0.6658344907610894, + "grad_norm": 1.503702998161316, + "learning_rate": 5.02468679652268e-05, + "loss": 0.1533, + "step": 28404 + }, + { + "epoch": 0.6658579323358943, + "grad_norm": 2.645399570465088, + "learning_rate": 5.024047835597221e-05, + "loss": 0.1059, + "step": 28405 + }, + { + "epoch": 0.6658813739106993, + "grad_norm": 4.787595272064209, + "learning_rate": 5.02340890167142e-05, + "loss": 0.4169, + "step": 28406 + }, + { + "epoch": 0.6659048154855043, + "grad_norm": 1.921549916267395, + "learning_rate": 5.022769994748744e-05, + "loss": 0.1429, + "step": 28407 + }, + { + "epoch": 0.6659282570603093, + "grad_norm": 1.0791703462600708, + "learning_rate": 5.022131114832658e-05, + "loss": 0.0698, + "step": 28408 + }, + { + "epoch": 0.6659516986351143, + "grad_norm": 3.868428945541382, + "learning_rate": 5.021492261926627e-05, + "loss": 0.3667, + "step": 28409 + }, + { + "epoch": 0.6659751402099193, + "grad_norm": 3.4191360473632812, + "learning_rate": 5.0208534360341163e-05, + "loss": 0.3114, + "step": 28410 + }, + { + "epoch": 0.6659985817847243, + "grad_norm": 4.566728115081787, + "learning_rate": 5.020214637158598e-05, + "loss": 0.4109, + "step": 28411 + }, + { + "epoch": 0.6660220233595293, + "grad_norm": 0.9398028254508972, + "learning_rate": 5.0195758653035355e-05, + "loss": 0.0872, + "step": 28412 + }, + { + "epoch": 0.6660454649343343, + "grad_norm": 2.8787167072296143, + "learning_rate": 5.018937120472393e-05, + "loss": 0.198, + "step": 28413 + }, + { + "epoch": 0.6660689065091393, + "grad_norm": 3.1174633502960205, + "learning_rate": 5.0182984026686396e-05, + "loss": 0.2385, + "step": 28414 + }, + { + "epoch": 0.6660923480839442, + "grad_norm": 0.5202060341835022, + "learning_rate": 5.017659711895739e-05, + "loss": 0.0205, + "step": 28415 + }, + { + "epoch": 0.6661157896587493, + "grad_norm": 2.011765480041504, + "learning_rate": 5.017021048157153e-05, + "loss": 0.151, + "step": 28416 + }, + { + "epoch": 0.6661392312335542, + "grad_norm": 3.8930160999298096, + "learning_rate": 5.0163824114563505e-05, + "loss": 0.3911, + "step": 28417 + }, + { + "epoch": 0.6661626728083593, + "grad_norm": 3.0186502933502197, + "learning_rate": 5.0157438017968014e-05, + "loss": 0.2122, + "step": 28418 + }, + { + "epoch": 0.6661861143831642, + "grad_norm": 2.620821714401245, + "learning_rate": 5.015105219181966e-05, + "loss": 0.2568, + "step": 28419 + }, + { + "epoch": 0.6662095559579693, + "grad_norm": 3.27811336517334, + "learning_rate": 5.0144666636153105e-05, + "loss": 0.2939, + "step": 28420 + }, + { + "epoch": 0.6662329975327742, + "grad_norm": 2.8787436485290527, + "learning_rate": 5.013828135100298e-05, + "loss": 0.1519, + "step": 28421 + }, + { + "epoch": 0.6662564391075793, + "grad_norm": 0.8104360699653625, + "learning_rate": 5.013189633640396e-05, + "loss": 0.066, + "step": 28422 + }, + { + "epoch": 0.6662798806823842, + "grad_norm": 2.198120355606079, + "learning_rate": 5.0125511592390615e-05, + "loss": 0.1446, + "step": 28423 + }, + { + "epoch": 0.6663033222571892, + "grad_norm": 1.5030577182769775, + "learning_rate": 5.011912711899769e-05, + "loss": 0.0358, + "step": 28424 + }, + { + "epoch": 0.6663267638319942, + "grad_norm": 1.0528539419174194, + "learning_rate": 5.011274291625979e-05, + "loss": 0.0877, + "step": 28425 + }, + { + "epoch": 0.6663502054067992, + "grad_norm": 3.230539083480835, + "learning_rate": 5.010635898421154e-05, + "loss": 0.259, + "step": 28426 + }, + { + "epoch": 0.6663736469816042, + "grad_norm": 2.224074125289917, + "learning_rate": 5.009997532288761e-05, + "loss": 0.2564, + "step": 28427 + }, + { + "epoch": 0.6663970885564092, + "grad_norm": 2.162182331085205, + "learning_rate": 5.009359193232259e-05, + "loss": 0.1515, + "step": 28428 + }, + { + "epoch": 0.6664205301312142, + "grad_norm": 2.397589683532715, + "learning_rate": 5.008720881255112e-05, + "loss": 0.2359, + "step": 28429 + }, + { + "epoch": 0.6664439717060192, + "grad_norm": 0.36680275201797485, + "learning_rate": 5.008082596360786e-05, + "loss": 0.0405, + "step": 28430 + }, + { + "epoch": 0.6664674132808243, + "grad_norm": 3.2895686626434326, + "learning_rate": 5.0074443385527467e-05, + "loss": 0.3046, + "step": 28431 + }, + { + "epoch": 0.6664908548556292, + "grad_norm": 2.1591684818267822, + "learning_rate": 5.006806107834457e-05, + "loss": 0.2321, + "step": 28432 + }, + { + "epoch": 0.6665142964304342, + "grad_norm": 2.3049306869506836, + "learning_rate": 5.0061679042093754e-05, + "loss": 0.1515, + "step": 28433 + }, + { + "epoch": 0.6665377380052392, + "grad_norm": 2.523303747177124, + "learning_rate": 5.005529727680969e-05, + "loss": 0.1863, + "step": 28434 + }, + { + "epoch": 0.6665611795800442, + "grad_norm": 3.5555131435394287, + "learning_rate": 5.0048915782526986e-05, + "loss": 0.3551, + "step": 28435 + }, + { + "epoch": 0.6665846211548492, + "grad_norm": 3.6247105598449707, + "learning_rate": 5.0042534559280216e-05, + "loss": 0.4582, + "step": 28436 + }, + { + "epoch": 0.6666080627296542, + "grad_norm": 0.3850175440311432, + "learning_rate": 5.003615360710411e-05, + "loss": 0.0262, + "step": 28437 + }, + { + "epoch": 0.6666315043044592, + "grad_norm": 1.859259009361267, + "learning_rate": 5.0029772926033234e-05, + "loss": 0.1295, + "step": 28438 + }, + { + "epoch": 0.6666549458792642, + "grad_norm": 0.45999279618263245, + "learning_rate": 5.002339251610222e-05, + "loss": 0.0186, + "step": 28439 + }, + { + "epoch": 0.6666783874540692, + "grad_norm": 4.519641399383545, + "learning_rate": 5.001701237734568e-05, + "loss": 0.1585, + "step": 28440 + }, + { + "epoch": 0.6667018290288742, + "grad_norm": 2.0954842567443848, + "learning_rate": 5.00106325097982e-05, + "loss": 1.7275, + "step": 28441 + }, + { + "epoch": 0.6667252706036791, + "grad_norm": 2.766502857208252, + "learning_rate": 5.0004252913494466e-05, + "loss": 0.2407, + "step": 28442 + }, + { + "epoch": 0.6667487121784842, + "grad_norm": 4.934020042419434, + "learning_rate": 4.999787358846902e-05, + "loss": 0.523, + "step": 28443 + }, + { + "epoch": 0.6667721537532891, + "grad_norm": 2.3319122791290283, + "learning_rate": 4.9991494534756566e-05, + "loss": 0.1302, + "step": 28444 + }, + { + "epoch": 0.6667955953280942, + "grad_norm": 0.8195972442626953, + "learning_rate": 4.9985115752391664e-05, + "loss": 0.065, + "step": 28445 + }, + { + "epoch": 0.6668190369028991, + "grad_norm": 3.889828681945801, + "learning_rate": 4.9978737241408934e-05, + "loss": 0.3169, + "step": 28446 + }, + { + "epoch": 0.6668424784777042, + "grad_norm": 2.8546385765075684, + "learning_rate": 4.9972359001842985e-05, + "loss": 0.1436, + "step": 28447 + }, + { + "epoch": 0.6668659200525091, + "grad_norm": 0.9122024774551392, + "learning_rate": 4.996598103372837e-05, + "loss": 0.0609, + "step": 28448 + }, + { + "epoch": 0.6668893616273142, + "grad_norm": 1.4740415811538696, + "learning_rate": 4.9959603337099804e-05, + "loss": 0.0957, + "step": 28449 + }, + { + "epoch": 0.6669128032021191, + "grad_norm": 1.8811613321304321, + "learning_rate": 4.995322591199183e-05, + "loss": 0.1583, + "step": 28450 + }, + { + "epoch": 0.6669362447769241, + "grad_norm": 2.1947638988494873, + "learning_rate": 4.994684875843906e-05, + "loss": 0.219, + "step": 28451 + }, + { + "epoch": 0.6669596863517291, + "grad_norm": 2.6650617122650146, + "learning_rate": 4.994047187647609e-05, + "loss": 0.2456, + "step": 28452 + }, + { + "epoch": 0.6669831279265341, + "grad_norm": 0.842848539352417, + "learning_rate": 4.9934095266137494e-05, + "loss": 0.0481, + "step": 28453 + }, + { + "epoch": 0.6670065695013391, + "grad_norm": 0.9402142763137817, + "learning_rate": 4.992771892745794e-05, + "loss": 0.0858, + "step": 28454 + }, + { + "epoch": 0.6670300110761441, + "grad_norm": 1.8175855875015259, + "learning_rate": 4.992134286047195e-05, + "loss": 1.4422, + "step": 28455 + }, + { + "epoch": 0.6670534526509491, + "grad_norm": 1.179412841796875, + "learning_rate": 4.9914967065214214e-05, + "loss": 0.159, + "step": 28456 + }, + { + "epoch": 0.6670768942257541, + "grad_norm": 3.4522716999053955, + "learning_rate": 4.9908591541719265e-05, + "loss": 0.3703, + "step": 28457 + }, + { + "epoch": 0.667100335800559, + "grad_norm": 3.3623030185699463, + "learning_rate": 4.990221629002171e-05, + "loss": 0.2649, + "step": 28458 + }, + { + "epoch": 0.6671237773753641, + "grad_norm": 1.5147392749786377, + "learning_rate": 4.989584131015613e-05, + "loss": 0.066, + "step": 28459 + }, + { + "epoch": 0.667147218950169, + "grad_norm": 2.5439035892486572, + "learning_rate": 4.988946660215713e-05, + "loss": 0.1578, + "step": 28460 + }, + { + "epoch": 0.6671706605249741, + "grad_norm": 3.0866997241973877, + "learning_rate": 4.9883092166059266e-05, + "loss": 0.2961, + "step": 28461 + }, + { + "epoch": 0.667194102099779, + "grad_norm": 2.63801646232605, + "learning_rate": 4.987671800189717e-05, + "loss": 0.1986, + "step": 28462 + }, + { + "epoch": 0.6672175436745841, + "grad_norm": 4.441329002380371, + "learning_rate": 4.987034410970543e-05, + "loss": 0.4437, + "step": 28463 + }, + { + "epoch": 0.667240985249389, + "grad_norm": 3.575526714324951, + "learning_rate": 4.9863970489518596e-05, + "loss": 0.4423, + "step": 28464 + }, + { + "epoch": 0.6672644268241941, + "grad_norm": 2.659071207046509, + "learning_rate": 4.985759714137125e-05, + "loss": 0.1923, + "step": 28465 + }, + { + "epoch": 0.667287868398999, + "grad_norm": 3.4604268074035645, + "learning_rate": 4.985122406529801e-05, + "loss": 0.2575, + "step": 28466 + }, + { + "epoch": 0.6673113099738041, + "grad_norm": 2.902355909347534, + "learning_rate": 4.984485126133345e-05, + "loss": 0.3829, + "step": 28467 + }, + { + "epoch": 0.667334751548609, + "grad_norm": 3.318617105484009, + "learning_rate": 4.983847872951209e-05, + "loss": 0.2054, + "step": 28468 + }, + { + "epoch": 0.667358193123414, + "grad_norm": 1.6146397590637207, + "learning_rate": 4.98321064698686e-05, + "loss": 0.0592, + "step": 28469 + }, + { + "epoch": 0.667381634698219, + "grad_norm": 0.7201565504074097, + "learning_rate": 4.9825734482437515e-05, + "loss": 0.0694, + "step": 28470 + }, + { + "epoch": 0.667405076273024, + "grad_norm": 3.945359468460083, + "learning_rate": 4.9819362767253396e-05, + "loss": 0.2764, + "step": 28471 + }, + { + "epoch": 0.667428517847829, + "grad_norm": 2.4953150749206543, + "learning_rate": 4.981299132435083e-05, + "loss": 0.7341, + "step": 28472 + }, + { + "epoch": 0.667451959422634, + "grad_norm": 2.345205545425415, + "learning_rate": 4.9806620153764384e-05, + "loss": 0.2152, + "step": 28473 + }, + { + "epoch": 0.667475400997439, + "grad_norm": 1.696709156036377, + "learning_rate": 4.980024925552859e-05, + "loss": 0.6813, + "step": 28474 + }, + { + "epoch": 0.667498842572244, + "grad_norm": 3.8299570083618164, + "learning_rate": 4.979387862967809e-05, + "loss": 0.2933, + "step": 28475 + }, + { + "epoch": 0.667522284147049, + "grad_norm": 3.2984509468078613, + "learning_rate": 4.978750827624742e-05, + "loss": 0.1681, + "step": 28476 + }, + { + "epoch": 0.667545725721854, + "grad_norm": 2.2597532272338867, + "learning_rate": 4.9781138195271134e-05, + "loss": 0.2145, + "step": 28477 + }, + { + "epoch": 0.6675691672966589, + "grad_norm": 2.842708110809326, + "learning_rate": 4.9774768386783774e-05, + "loss": 0.2587, + "step": 28478 + }, + { + "epoch": 0.667592608871464, + "grad_norm": 1.4281407594680786, + "learning_rate": 4.9768398850819975e-05, + "loss": 0.1509, + "step": 28479 + }, + { + "epoch": 0.6676160504462689, + "grad_norm": 2.309360980987549, + "learning_rate": 4.976202958741425e-05, + "loss": 0.1304, + "step": 28480 + }, + { + "epoch": 0.667639492021074, + "grad_norm": 3.721278429031372, + "learning_rate": 4.975566059660113e-05, + "loss": 0.4514, + "step": 28481 + }, + { + "epoch": 0.6676629335958789, + "grad_norm": 1.262803554534912, + "learning_rate": 4.974929187841525e-05, + "loss": 0.0754, + "step": 28482 + }, + { + "epoch": 0.667686375170684, + "grad_norm": 0.4491646885871887, + "learning_rate": 4.974292343289112e-05, + "loss": 0.0346, + "step": 28483 + }, + { + "epoch": 0.667709816745489, + "grad_norm": 2.0931556224823, + "learning_rate": 4.97365552600633e-05, + "loss": 0.3013, + "step": 28484 + }, + { + "epoch": 0.667733258320294, + "grad_norm": 3.046501874923706, + "learning_rate": 4.973018735996634e-05, + "loss": 0.2495, + "step": 28485 + }, + { + "epoch": 0.667756699895099, + "grad_norm": 2.7232985496520996, + "learning_rate": 4.972381973263477e-05, + "loss": 0.1452, + "step": 28486 + }, + { + "epoch": 0.6677801414699039, + "grad_norm": 3.0777316093444824, + "learning_rate": 4.971745237810319e-05, + "loss": 0.1664, + "step": 28487 + }, + { + "epoch": 0.667803583044709, + "grad_norm": 0.9268031120300293, + "learning_rate": 4.971108529640615e-05, + "loss": 0.0634, + "step": 28488 + }, + { + "epoch": 0.6678270246195139, + "grad_norm": 3.535832166671753, + "learning_rate": 4.970471848757815e-05, + "loss": 0.2045, + "step": 28489 + }, + { + "epoch": 0.667850466194319, + "grad_norm": 3.0859663486480713, + "learning_rate": 4.9698351951653745e-05, + "loss": 0.4366, + "step": 28490 + }, + { + "epoch": 0.6678739077691239, + "grad_norm": 2.3762192726135254, + "learning_rate": 4.969198568866752e-05, + "loss": 0.164, + "step": 28491 + }, + { + "epoch": 0.667897349343929, + "grad_norm": 2.327195882797241, + "learning_rate": 4.9685619698653994e-05, + "loss": 0.1388, + "step": 28492 + }, + { + "epoch": 0.6679207909187339, + "grad_norm": 5.056883335113525, + "learning_rate": 4.967925398164769e-05, + "loss": 0.5931, + "step": 28493 + }, + { + "epoch": 0.667944232493539, + "grad_norm": 4.123617172241211, + "learning_rate": 4.967288853768319e-05, + "loss": 0.4223, + "step": 28494 + }, + { + "epoch": 0.6679676740683439, + "grad_norm": 1.5044459104537964, + "learning_rate": 4.966652336679501e-05, + "loss": 0.1207, + "step": 28495 + }, + { + "epoch": 0.667991115643149, + "grad_norm": 3.1127982139587402, + "learning_rate": 4.96601584690177e-05, + "loss": 0.0897, + "step": 28496 + }, + { + "epoch": 0.6680145572179539, + "grad_norm": 5.798924446105957, + "learning_rate": 4.965379384438579e-05, + "loss": 0.5835, + "step": 28497 + }, + { + "epoch": 0.6680379987927589, + "grad_norm": 2.3696346282958984, + "learning_rate": 4.964742949293379e-05, + "loss": 0.2141, + "step": 28498 + }, + { + "epoch": 0.6680614403675639, + "grad_norm": 2.585545539855957, + "learning_rate": 4.9641065414696244e-05, + "loss": 0.1098, + "step": 28499 + }, + { + "epoch": 0.6680848819423689, + "grad_norm": 3.4707658290863037, + "learning_rate": 4.963470160970771e-05, + "loss": 0.2485, + "step": 28500 + }, + { + "epoch": 0.6681083235171739, + "grad_norm": 3.2011966705322266, + "learning_rate": 4.96283380780027e-05, + "loss": 0.3386, + "step": 28501 + }, + { + "epoch": 0.6681317650919789, + "grad_norm": 2.8552069664001465, + "learning_rate": 4.962197481961571e-05, + "loss": 0.2042, + "step": 28502 + }, + { + "epoch": 0.6681552066667839, + "grad_norm": 1.595898151397705, + "learning_rate": 4.961561183458134e-05, + "loss": 0.1527, + "step": 28503 + }, + { + "epoch": 0.6681786482415889, + "grad_norm": 3.8095805644989014, + "learning_rate": 4.9609249122934086e-05, + "loss": 0.4485, + "step": 28504 + }, + { + "epoch": 0.6682020898163938, + "grad_norm": 2.7189176082611084, + "learning_rate": 4.960288668470846e-05, + "loss": 0.1686, + "step": 28505 + }, + { + "epoch": 0.6682255313911989, + "grad_norm": 4.010475158691406, + "learning_rate": 4.9596524519938934e-05, + "loss": 0.2838, + "step": 28506 + }, + { + "epoch": 0.6682489729660038, + "grad_norm": 1.0693060159683228, + "learning_rate": 4.959016262866014e-05, + "loss": 0.1353, + "step": 28507 + }, + { + "epoch": 0.6682724145408089, + "grad_norm": 2.3049633502960205, + "learning_rate": 4.9583801010906526e-05, + "loss": 1.6741, + "step": 28508 + }, + { + "epoch": 0.6682958561156138, + "grad_norm": 2.2577810287475586, + "learning_rate": 4.957743966671265e-05, + "loss": 0.1401, + "step": 28509 + }, + { + "epoch": 0.6683192976904189, + "grad_norm": 2.640144109725952, + "learning_rate": 4.957107859611297e-05, + "loss": 0.0734, + "step": 28510 + }, + { + "epoch": 0.6683427392652238, + "grad_norm": 0.5606049299240112, + "learning_rate": 4.956471779914206e-05, + "loss": 0.0893, + "step": 28511 + }, + { + "epoch": 0.6683661808400289, + "grad_norm": 1.835747241973877, + "learning_rate": 4.955835727583437e-05, + "loss": 1.114, + "step": 28512 + }, + { + "epoch": 0.6683896224148338, + "grad_norm": 4.2835001945495605, + "learning_rate": 4.955199702622447e-05, + "loss": 0.632, + "step": 28513 + }, + { + "epoch": 0.6684130639896388, + "grad_norm": 3.1703057289123535, + "learning_rate": 4.9545637050346835e-05, + "loss": 0.9999, + "step": 28514 + }, + { + "epoch": 0.6684365055644438, + "grad_norm": 3.437425374984741, + "learning_rate": 4.9539277348236026e-05, + "loss": 0.3644, + "step": 28515 + }, + { + "epoch": 0.6684599471392488, + "grad_norm": 1.2350075244903564, + "learning_rate": 4.953291791992652e-05, + "loss": 0.1073, + "step": 28516 + }, + { + "epoch": 0.6684833887140538, + "grad_norm": 0.4255327582359314, + "learning_rate": 4.9526558765452805e-05, + "loss": 0.0333, + "step": 28517 + }, + { + "epoch": 0.6685068302888588, + "grad_norm": 3.4152729511260986, + "learning_rate": 4.952019988484941e-05, + "loss": 0.1417, + "step": 28518 + }, + { + "epoch": 0.6685302718636638, + "grad_norm": 1.8475995063781738, + "learning_rate": 4.9513841278150786e-05, + "loss": 1.2521, + "step": 28519 + }, + { + "epoch": 0.6685537134384688, + "grad_norm": 2.7734146118164062, + "learning_rate": 4.950748294539153e-05, + "loss": 0.3703, + "step": 28520 + }, + { + "epoch": 0.6685771550132737, + "grad_norm": 4.0838446617126465, + "learning_rate": 4.950112488660609e-05, + "loss": 0.4466, + "step": 28521 + }, + { + "epoch": 0.6686005965880788, + "grad_norm": 2.681725025177002, + "learning_rate": 4.9494767101828955e-05, + "loss": 0.2374, + "step": 28522 + }, + { + "epoch": 0.6686240381628837, + "grad_norm": 2.846872091293335, + "learning_rate": 4.948840959109464e-05, + "loss": 0.2794, + "step": 28523 + }, + { + "epoch": 0.6686474797376888, + "grad_norm": 0.5615993738174438, + "learning_rate": 4.948205235443759e-05, + "loss": 0.0265, + "step": 28524 + }, + { + "epoch": 0.6686709213124937, + "grad_norm": 2.840818166732788, + "learning_rate": 4.9475695391892394e-05, + "loss": 0.2325, + "step": 28525 + }, + { + "epoch": 0.6686943628872988, + "grad_norm": 0.597296416759491, + "learning_rate": 4.946933870349345e-05, + "loss": 0.042, + "step": 28526 + }, + { + "epoch": 0.6687178044621037, + "grad_norm": 1.7369886636734009, + "learning_rate": 4.946298228927534e-05, + "loss": 0.2016, + "step": 28527 + }, + { + "epoch": 0.6687412460369088, + "grad_norm": 3.5762007236480713, + "learning_rate": 4.94566261492725e-05, + "loss": 0.2443, + "step": 28528 + }, + { + "epoch": 0.6687646876117137, + "grad_norm": 3.1581292152404785, + "learning_rate": 4.945027028351943e-05, + "loss": 0.2082, + "step": 28529 + }, + { + "epoch": 0.6687881291865188, + "grad_norm": 3.973884105682373, + "learning_rate": 4.944391469205062e-05, + "loss": 0.2597, + "step": 28530 + }, + { + "epoch": 0.6688115707613237, + "grad_norm": 0.5813103914260864, + "learning_rate": 4.943755937490051e-05, + "loss": 0.0375, + "step": 28531 + }, + { + "epoch": 0.6688350123361287, + "grad_norm": 3.3017380237579346, + "learning_rate": 4.9431204332103654e-05, + "loss": 0.2576, + "step": 28532 + }, + { + "epoch": 0.6688584539109337, + "grad_norm": 3.0700440406799316, + "learning_rate": 4.942484956369452e-05, + "loss": 0.3072, + "step": 28533 + }, + { + "epoch": 0.6688818954857387, + "grad_norm": 4.994998931884766, + "learning_rate": 4.941849506970756e-05, + "loss": 0.6708, + "step": 28534 + }, + { + "epoch": 0.6689053370605438, + "grad_norm": 2.125112533569336, + "learning_rate": 4.941214085017728e-05, + "loss": 0.1066, + "step": 28535 + }, + { + "epoch": 0.6689287786353487, + "grad_norm": 1.5957924127578735, + "learning_rate": 4.940578690513814e-05, + "loss": 0.1544, + "step": 28536 + }, + { + "epoch": 0.6689522202101538, + "grad_norm": 0.8127283453941345, + "learning_rate": 4.939943323462458e-05, + "loss": 0.0992, + "step": 28537 + }, + { + "epoch": 0.6689756617849587, + "grad_norm": 0.837485134601593, + "learning_rate": 4.939307983867113e-05, + "loss": 0.1038, + "step": 28538 + }, + { + "epoch": 0.6689991033597638, + "grad_norm": 0.7612943053245544, + "learning_rate": 4.938672671731228e-05, + "loss": 0.0384, + "step": 28539 + }, + { + "epoch": 0.6690225449345687, + "grad_norm": 0.8382937908172607, + "learning_rate": 4.9380373870582474e-05, + "loss": 0.0758, + "step": 28540 + }, + { + "epoch": 0.6690459865093737, + "grad_norm": 0.4059826135635376, + "learning_rate": 4.937402129851619e-05, + "loss": 0.0286, + "step": 28541 + }, + { + "epoch": 0.6690694280841787, + "grad_norm": 2.8493435382843018, + "learning_rate": 4.9367669001147886e-05, + "loss": 1.1599, + "step": 28542 + }, + { + "epoch": 0.6690928696589837, + "grad_norm": 3.6024906635284424, + "learning_rate": 4.936131697851204e-05, + "loss": 0.3777, + "step": 28543 + }, + { + "epoch": 0.6691163112337887, + "grad_norm": 2.6526904106140137, + "learning_rate": 4.935496523064307e-05, + "loss": 0.13, + "step": 28544 + }, + { + "epoch": 0.6691397528085937, + "grad_norm": 0.9454807043075562, + "learning_rate": 4.9348613757575513e-05, + "loss": 0.1194, + "step": 28545 + }, + { + "epoch": 0.6691631943833987, + "grad_norm": 0.30466631054878235, + "learning_rate": 4.934226255934381e-05, + "loss": 0.0302, + "step": 28546 + }, + { + "epoch": 0.6691866359582037, + "grad_norm": 5.253718376159668, + "learning_rate": 4.933591163598241e-05, + "loss": 0.3906, + "step": 28547 + }, + { + "epoch": 0.6692100775330087, + "grad_norm": 3.3290746212005615, + "learning_rate": 4.932956098752578e-05, + "loss": 0.2157, + "step": 28548 + }, + { + "epoch": 0.6692335191078137, + "grad_norm": 3.897094964981079, + "learning_rate": 4.932321061400837e-05, + "loss": 0.3535, + "step": 28549 + }, + { + "epoch": 0.6692569606826186, + "grad_norm": 3.5014052391052246, + "learning_rate": 4.9316860515464623e-05, + "loss": 0.3301, + "step": 28550 + }, + { + "epoch": 0.6692804022574237, + "grad_norm": 2.840468168258667, + "learning_rate": 4.9310510691929e-05, + "loss": 0.2566, + "step": 28551 + }, + { + "epoch": 0.6693038438322286, + "grad_norm": 1.7338789701461792, + "learning_rate": 4.9304161143436025e-05, + "loss": 0.1241, + "step": 28552 + }, + { + "epoch": 0.6693272854070337, + "grad_norm": 1.3118876218795776, + "learning_rate": 4.9297811870020084e-05, + "loss": 0.0558, + "step": 28553 + }, + { + "epoch": 0.6693507269818386, + "grad_norm": 2.207913875579834, + "learning_rate": 4.9291462871715647e-05, + "loss": 0.1262, + "step": 28554 + }, + { + "epoch": 0.6693741685566437, + "grad_norm": 2.3441426753997803, + "learning_rate": 4.9285114148557156e-05, + "loss": 0.0648, + "step": 28555 + }, + { + "epoch": 0.6693976101314486, + "grad_norm": 3.335566997528076, + "learning_rate": 4.927876570057906e-05, + "loss": 0.3224, + "step": 28556 + }, + { + "epoch": 0.6694210517062537, + "grad_norm": 2.992107391357422, + "learning_rate": 4.927241752781576e-05, + "loss": 0.2488, + "step": 28557 + }, + { + "epoch": 0.6694444932810586, + "grad_norm": 6.120455741882324, + "learning_rate": 4.92660696303018e-05, + "loss": 1.0285, + "step": 28558 + }, + { + "epoch": 0.6694679348558636, + "grad_norm": 4.266399383544922, + "learning_rate": 4.925972200807157e-05, + "loss": 0.2544, + "step": 28559 + }, + { + "epoch": 0.6694913764306686, + "grad_norm": 3.840156316757202, + "learning_rate": 4.925337466115952e-05, + "loss": 0.3679, + "step": 28560 + }, + { + "epoch": 0.6695148180054736, + "grad_norm": 0.6938939094543457, + "learning_rate": 4.9247027589600084e-05, + "loss": 0.0777, + "step": 28561 + }, + { + "epoch": 0.6695382595802786, + "grad_norm": 0.84137362241745, + "learning_rate": 4.9240680793427664e-05, + "loss": 0.0378, + "step": 28562 + }, + { + "epoch": 0.6695617011550836, + "grad_norm": 2.255255699157715, + "learning_rate": 4.923433427267673e-05, + "loss": 0.2647, + "step": 28563 + }, + { + "epoch": 0.6695851427298886, + "grad_norm": 2.550708532333374, + "learning_rate": 4.9227988027381774e-05, + "loss": 0.147, + "step": 28564 + }, + { + "epoch": 0.6696085843046936, + "grad_norm": 1.0486557483673096, + "learning_rate": 4.922164205757718e-05, + "loss": 0.1358, + "step": 28565 + }, + { + "epoch": 0.6696320258794985, + "grad_norm": 2.9881463050842285, + "learning_rate": 4.921529636329738e-05, + "loss": 0.1933, + "step": 28566 + }, + { + "epoch": 0.6696554674543036, + "grad_norm": 2.230682373046875, + "learning_rate": 4.9208950944576813e-05, + "loss": 1.2778, + "step": 28567 + }, + { + "epoch": 0.6696789090291085, + "grad_norm": 2.226165771484375, + "learning_rate": 4.9202605801449906e-05, + "loss": 0.1441, + "step": 28568 + }, + { + "epoch": 0.6697023506039136, + "grad_norm": 3.4597299098968506, + "learning_rate": 4.919626093395104e-05, + "loss": 0.3408, + "step": 28569 + }, + { + "epoch": 0.6697257921787185, + "grad_norm": 2.421259880065918, + "learning_rate": 4.918991634211475e-05, + "loss": 0.1978, + "step": 28570 + }, + { + "epoch": 0.6697492337535236, + "grad_norm": 2.8335671424865723, + "learning_rate": 4.918357202597539e-05, + "loss": 0.2526, + "step": 28571 + }, + { + "epoch": 0.6697726753283285, + "grad_norm": 2.9589035511016846, + "learning_rate": 4.9177227985567396e-05, + "loss": 0.3129, + "step": 28572 + }, + { + "epoch": 0.6697961169031336, + "grad_norm": 5.344758033752441, + "learning_rate": 4.91708842209252e-05, + "loss": 0.5481, + "step": 28573 + }, + { + "epoch": 0.6698195584779385, + "grad_norm": 3.4013893604278564, + "learning_rate": 4.91645407320832e-05, + "loss": 0.2962, + "step": 28574 + }, + { + "epoch": 0.6698430000527436, + "grad_norm": 2.5773322582244873, + "learning_rate": 4.91581975190758e-05, + "loss": 1.5275, + "step": 28575 + }, + { + "epoch": 0.6698664416275485, + "grad_norm": 4.498460292816162, + "learning_rate": 4.915185458193746e-05, + "loss": 0.3392, + "step": 28576 + }, + { + "epoch": 0.6698898832023535, + "grad_norm": 3.481562614440918, + "learning_rate": 4.914551192070262e-05, + "loss": 0.1991, + "step": 28577 + }, + { + "epoch": 0.6699133247771585, + "grad_norm": 1.185151219367981, + "learning_rate": 4.913916953540565e-05, + "loss": 0.1188, + "step": 28578 + }, + { + "epoch": 0.6699367663519635, + "grad_norm": 0.5544365048408508, + "learning_rate": 4.913282742608098e-05, + "loss": 0.0397, + "step": 28579 + }, + { + "epoch": 0.6699602079267685, + "grad_norm": 3.283156156539917, + "learning_rate": 4.912648559276304e-05, + "loss": 0.3408, + "step": 28580 + }, + { + "epoch": 0.6699836495015735, + "grad_norm": 2.855337381362915, + "learning_rate": 4.912014403548619e-05, + "loss": 0.2354, + "step": 28581 + }, + { + "epoch": 0.6700070910763785, + "grad_norm": 0.3097623288631439, + "learning_rate": 4.9113802754284855e-05, + "loss": 0.0152, + "step": 28582 + }, + { + "epoch": 0.6700305326511835, + "grad_norm": 0.5486602783203125, + "learning_rate": 4.910746174919348e-05, + "loss": 0.044, + "step": 28583 + }, + { + "epoch": 0.6700539742259884, + "grad_norm": 3.947547674179077, + "learning_rate": 4.910112102024645e-05, + "loss": 0.2979, + "step": 28584 + }, + { + "epoch": 0.6700774158007935, + "grad_norm": 3.8882198333740234, + "learning_rate": 4.909478056747818e-05, + "loss": 0.2969, + "step": 28585 + }, + { + "epoch": 0.6701008573755985, + "grad_norm": 2.6603176593780518, + "learning_rate": 4.908844039092305e-05, + "loss": 0.1088, + "step": 28586 + }, + { + "epoch": 0.6701242989504035, + "grad_norm": 1.1885762214660645, + "learning_rate": 4.9082100490615444e-05, + "loss": 0.0696, + "step": 28587 + }, + { + "epoch": 0.6701477405252085, + "grad_norm": 3.335787773132324, + "learning_rate": 4.9075760866589826e-05, + "loss": 0.2286, + "step": 28588 + }, + { + "epoch": 0.6701711821000135, + "grad_norm": 1.7324055433273315, + "learning_rate": 4.9069421518880524e-05, + "loss": 0.1583, + "step": 28589 + }, + { + "epoch": 0.6701946236748185, + "grad_norm": 2.7653074264526367, + "learning_rate": 4.9063082447522016e-05, + "loss": 0.22, + "step": 28590 + }, + { + "epoch": 0.6702180652496235, + "grad_norm": 2.999372720718384, + "learning_rate": 4.9056743652548645e-05, + "loss": 1.3048, + "step": 28591 + }, + { + "epoch": 0.6702415068244285, + "grad_norm": 2.0494446754455566, + "learning_rate": 4.905040513399481e-05, + "loss": 0.0979, + "step": 28592 + }, + { + "epoch": 0.6702649483992335, + "grad_norm": 4.721913814544678, + "learning_rate": 4.904406689189492e-05, + "loss": 0.3285, + "step": 28593 + }, + { + "epoch": 0.6702883899740385, + "grad_norm": 2.7611868381500244, + "learning_rate": 4.9037728926283365e-05, + "loss": 0.2255, + "step": 28594 + }, + { + "epoch": 0.6703118315488434, + "grad_norm": 3.153385877609253, + "learning_rate": 4.903139123719448e-05, + "loss": 0.4098, + "step": 28595 + }, + { + "epoch": 0.6703352731236485, + "grad_norm": 1.9542229175567627, + "learning_rate": 4.902505382466272e-05, + "loss": 0.1089, + "step": 28596 + }, + { + "epoch": 0.6703587146984534, + "grad_norm": 2.250887393951416, + "learning_rate": 4.901871668872248e-05, + "loss": 0.2248, + "step": 28597 + }, + { + "epoch": 0.6703821562732585, + "grad_norm": 1.4755264520645142, + "learning_rate": 4.90123798294081e-05, + "loss": 0.1349, + "step": 28598 + }, + { + "epoch": 0.6704055978480634, + "grad_norm": 2.2145798206329346, + "learning_rate": 4.900604324675394e-05, + "loss": 0.2073, + "step": 28599 + }, + { + "epoch": 0.6704290394228685, + "grad_norm": 2.119356155395508, + "learning_rate": 4.8999706940794474e-05, + "loss": 0.1227, + "step": 28600 + }, + { + "epoch": 0.6704524809976734, + "grad_norm": 2.66882586479187, + "learning_rate": 4.8993370911563985e-05, + "loss": 0.1297, + "step": 28601 + }, + { + "epoch": 0.6704759225724785, + "grad_norm": 3.376089572906494, + "learning_rate": 4.898703515909694e-05, + "loss": 1.8731, + "step": 28602 + }, + { + "epoch": 0.6704993641472834, + "grad_norm": 2.3574278354644775, + "learning_rate": 4.898069968342768e-05, + "loss": 0.1933, + "step": 28603 + }, + { + "epoch": 0.6705228057220884, + "grad_norm": 2.0981650352478027, + "learning_rate": 4.8974364484590576e-05, + "loss": 0.1372, + "step": 28604 + }, + { + "epoch": 0.6705462472968934, + "grad_norm": 3.397495985031128, + "learning_rate": 4.896802956262e-05, + "loss": 0.4523, + "step": 28605 + }, + { + "epoch": 0.6705696888716984, + "grad_norm": 2.7518908977508545, + "learning_rate": 4.896169491755033e-05, + "loss": 0.2393, + "step": 28606 + }, + { + "epoch": 0.6705931304465034, + "grad_norm": 2.7280144691467285, + "learning_rate": 4.8955360549415906e-05, + "loss": 0.2463, + "step": 28607 + }, + { + "epoch": 0.6706165720213084, + "grad_norm": 0.8895553350448608, + "learning_rate": 4.8949026458251167e-05, + "loss": 0.0573, + "step": 28608 + }, + { + "epoch": 0.6706400135961134, + "grad_norm": 4.37359619140625, + "learning_rate": 4.8942692644090446e-05, + "loss": 0.2946, + "step": 28609 + }, + { + "epoch": 0.6706634551709184, + "grad_norm": 1.8417519330978394, + "learning_rate": 4.893635910696811e-05, + "loss": 0.2676, + "step": 28610 + }, + { + "epoch": 0.6706868967457233, + "grad_norm": 3.228571891784668, + "learning_rate": 4.893002584691848e-05, + "loss": 0.306, + "step": 28611 + }, + { + "epoch": 0.6707103383205284, + "grad_norm": 2.0499370098114014, + "learning_rate": 4.892369286397601e-05, + "loss": 0.1888, + "step": 28612 + }, + { + "epoch": 0.6707337798953333, + "grad_norm": 2.438274621963501, + "learning_rate": 4.891736015817503e-05, + "loss": 0.2429, + "step": 28613 + }, + { + "epoch": 0.6707572214701384, + "grad_norm": 2.6119894981384277, + "learning_rate": 4.8911027729549844e-05, + "loss": 0.2465, + "step": 28614 + }, + { + "epoch": 0.6707806630449433, + "grad_norm": 1.8231817483901978, + "learning_rate": 4.8904695578134883e-05, + "loss": 0.1015, + "step": 28615 + }, + { + "epoch": 0.6708041046197484, + "grad_norm": 3.7239480018615723, + "learning_rate": 4.88983637039645e-05, + "loss": 0.5419, + "step": 28616 + }, + { + "epoch": 0.6708275461945533, + "grad_norm": 1.7203911542892456, + "learning_rate": 4.8892032107073014e-05, + "loss": 0.0935, + "step": 28617 + }, + { + "epoch": 0.6708509877693584, + "grad_norm": 0.7487347722053528, + "learning_rate": 4.888570078749482e-05, + "loss": 0.0574, + "step": 28618 + }, + { + "epoch": 0.6708744293441633, + "grad_norm": 2.7673821449279785, + "learning_rate": 4.8879369745264236e-05, + "loss": 0.1857, + "step": 28619 + }, + { + "epoch": 0.6708978709189684, + "grad_norm": 2.709625005722046, + "learning_rate": 4.88730389804156e-05, + "loss": 0.22, + "step": 28620 + }, + { + "epoch": 0.6709213124937733, + "grad_norm": 1.0728998184204102, + "learning_rate": 4.886670849298331e-05, + "loss": 0.1101, + "step": 28621 + }, + { + "epoch": 0.6709447540685783, + "grad_norm": 2.728598117828369, + "learning_rate": 4.8860378283001714e-05, + "loss": 0.2186, + "step": 28622 + }, + { + "epoch": 0.6709681956433833, + "grad_norm": 2.6754772663116455, + "learning_rate": 4.885404835050514e-05, + "loss": 0.3423, + "step": 28623 + }, + { + "epoch": 0.6709916372181883, + "grad_norm": 1.2496161460876465, + "learning_rate": 4.884771869552791e-05, + "loss": 0.1079, + "step": 28624 + }, + { + "epoch": 0.6710150787929933, + "grad_norm": 1.2751290798187256, + "learning_rate": 4.884138931810442e-05, + "loss": 0.0935, + "step": 28625 + }, + { + "epoch": 0.6710385203677983, + "grad_norm": 3.447671890258789, + "learning_rate": 4.8835060218269e-05, + "loss": 0.3007, + "step": 28626 + }, + { + "epoch": 0.6710619619426033, + "grad_norm": 3.557325839996338, + "learning_rate": 4.882873139605594e-05, + "loss": 0.4462, + "step": 28627 + }, + { + "epoch": 0.6710854035174083, + "grad_norm": 3.936056137084961, + "learning_rate": 4.882240285149966e-05, + "loss": 0.2595, + "step": 28628 + }, + { + "epoch": 0.6711088450922132, + "grad_norm": 3.855823040008545, + "learning_rate": 4.8816074584634463e-05, + "loss": 0.4119, + "step": 28629 + }, + { + "epoch": 0.6711322866670183, + "grad_norm": 2.0366885662078857, + "learning_rate": 4.8809746595494676e-05, + "loss": 0.2124, + "step": 28630 + }, + { + "epoch": 0.6711557282418232, + "grad_norm": 2.894932985305786, + "learning_rate": 4.880341888411466e-05, + "loss": 0.3327, + "step": 28631 + }, + { + "epoch": 0.6711791698166283, + "grad_norm": 0.5199347138404846, + "learning_rate": 4.8797091450528676e-05, + "loss": 0.0478, + "step": 28632 + }, + { + "epoch": 0.6712026113914332, + "grad_norm": 3.5071582794189453, + "learning_rate": 4.879076429477116e-05, + "loss": 1.369, + "step": 28633 + }, + { + "epoch": 0.6712260529662383, + "grad_norm": 2.7034244537353516, + "learning_rate": 4.878443741687641e-05, + "loss": 0.2347, + "step": 28634 + }, + { + "epoch": 0.6712494945410432, + "grad_norm": 4.183754920959473, + "learning_rate": 4.877811081687872e-05, + "loss": 0.2077, + "step": 28635 + }, + { + "epoch": 0.6712729361158483, + "grad_norm": 0.27335864305496216, + "learning_rate": 4.877178449481241e-05, + "loss": 0.0173, + "step": 28636 + }, + { + "epoch": 0.6712963776906533, + "grad_norm": 3.8311960697174072, + "learning_rate": 4.8765458450711874e-05, + "loss": 1.5128, + "step": 28637 + }, + { + "epoch": 0.6713198192654583, + "grad_norm": 3.2168209552764893, + "learning_rate": 4.875913268461141e-05, + "loss": 0.2226, + "step": 28638 + }, + { + "epoch": 0.6713432608402633, + "grad_norm": 2.3190083503723145, + "learning_rate": 4.875280719654529e-05, + "loss": 0.1216, + "step": 28639 + }, + { + "epoch": 0.6713667024150682, + "grad_norm": 1.2666162252426147, + "learning_rate": 4.874648198654792e-05, + "loss": 0.0933, + "step": 28640 + }, + { + "epoch": 0.6713901439898733, + "grad_norm": 3.5933680534362793, + "learning_rate": 4.874015705465358e-05, + "loss": 0.4322, + "step": 28641 + }, + { + "epoch": 0.6714135855646782, + "grad_norm": 1.6079548597335815, + "learning_rate": 4.8733832400896586e-05, + "loss": 0.1078, + "step": 28642 + }, + { + "epoch": 0.6714370271394833, + "grad_norm": 0.8632850646972656, + "learning_rate": 4.8727508025311255e-05, + "loss": 0.0772, + "step": 28643 + }, + { + "epoch": 0.6714604687142882, + "grad_norm": 0.7727105617523193, + "learning_rate": 4.8721183927931916e-05, + "loss": 0.0651, + "step": 28644 + }, + { + "epoch": 0.6714839102890933, + "grad_norm": 2.650303363800049, + "learning_rate": 4.8714860108792824e-05, + "loss": 0.1546, + "step": 28645 + }, + { + "epoch": 0.6715073518638982, + "grad_norm": 1.7094957828521729, + "learning_rate": 4.870853656792839e-05, + "loss": 0.1449, + "step": 28646 + }, + { + "epoch": 0.6715307934387033, + "grad_norm": 1.9779162406921387, + "learning_rate": 4.870221330537288e-05, + "loss": 0.1989, + "step": 28647 + }, + { + "epoch": 0.6715542350135082, + "grad_norm": 2.4006834030151367, + "learning_rate": 4.869589032116057e-05, + "loss": 0.1902, + "step": 28648 + }, + { + "epoch": 0.6715776765883132, + "grad_norm": 1.9419960975646973, + "learning_rate": 4.868956761532583e-05, + "loss": 0.9393, + "step": 28649 + }, + { + "epoch": 0.6716011181631182, + "grad_norm": 2.970209836959839, + "learning_rate": 4.8683245187902934e-05, + "loss": 0.2134, + "step": 28650 + }, + { + "epoch": 0.6716245597379232, + "grad_norm": 2.0659806728363037, + "learning_rate": 4.86769230389262e-05, + "loss": 0.1734, + "step": 28651 + }, + { + "epoch": 0.6716480013127282, + "grad_norm": 3.45804762840271, + "learning_rate": 4.867060116842988e-05, + "loss": 1.5384, + "step": 28652 + }, + { + "epoch": 0.6716714428875332, + "grad_norm": 1.0489777326583862, + "learning_rate": 4.866427957644838e-05, + "loss": 0.1113, + "step": 28653 + }, + { + "epoch": 0.6716948844623382, + "grad_norm": 4.143621444702148, + "learning_rate": 4.8657958263015926e-05, + "loss": 0.4238, + "step": 28654 + }, + { + "epoch": 0.6717183260371432, + "grad_norm": 2.7618520259857178, + "learning_rate": 4.865163722816684e-05, + "loss": 0.3649, + "step": 28655 + }, + { + "epoch": 0.6717417676119481, + "grad_norm": 0.8367321491241455, + "learning_rate": 4.86453164719354e-05, + "loss": 0.1209, + "step": 28656 + }, + { + "epoch": 0.6717652091867532, + "grad_norm": 3.0281670093536377, + "learning_rate": 4.863899599435593e-05, + "loss": 0.2775, + "step": 28657 + }, + { + "epoch": 0.6717886507615581, + "grad_norm": 0.33605825901031494, + "learning_rate": 4.8632675795462665e-05, + "loss": 0.042, + "step": 28658 + }, + { + "epoch": 0.6718120923363632, + "grad_norm": 2.750453472137451, + "learning_rate": 4.8626355875289996e-05, + "loss": 0.1649, + "step": 28659 + }, + { + "epoch": 0.6718355339111681, + "grad_norm": 2.3259057998657227, + "learning_rate": 4.862003623387212e-05, + "loss": 0.0971, + "step": 28660 + }, + { + "epoch": 0.6718589754859732, + "grad_norm": 3.2282536029815674, + "learning_rate": 4.861371687124341e-05, + "loss": 0.1837, + "step": 28661 + }, + { + "epoch": 0.6718824170607781, + "grad_norm": 2.001791000366211, + "learning_rate": 4.860739778743811e-05, + "loss": 1.2613, + "step": 28662 + }, + { + "epoch": 0.6719058586355832, + "grad_norm": 4.616305828094482, + "learning_rate": 4.8601078982490535e-05, + "loss": 0.2115, + "step": 28663 + }, + { + "epoch": 0.6719293002103881, + "grad_norm": 0.46807917952537537, + "learning_rate": 4.859476045643493e-05, + "loss": 0.0426, + "step": 28664 + }, + { + "epoch": 0.6719527417851932, + "grad_norm": 1.3545777797698975, + "learning_rate": 4.8588442209305574e-05, + "loss": 0.2461, + "step": 28665 + }, + { + "epoch": 0.6719761833599981, + "grad_norm": 3.034709930419922, + "learning_rate": 4.8582124241136805e-05, + "loss": 0.3341, + "step": 28666 + }, + { + "epoch": 0.6719996249348031, + "grad_norm": 2.9274001121520996, + "learning_rate": 4.857580655196289e-05, + "loss": 0.231, + "step": 28667 + }, + { + "epoch": 0.6720230665096081, + "grad_norm": 3.0977416038513184, + "learning_rate": 4.856948914181808e-05, + "loss": 0.3103, + "step": 28668 + }, + { + "epoch": 0.6720465080844131, + "grad_norm": 0.8164501786231995, + "learning_rate": 4.856317201073667e-05, + "loss": 0.0785, + "step": 28669 + }, + { + "epoch": 0.6720699496592181, + "grad_norm": 1.2996269464492798, + "learning_rate": 4.85568551587529e-05, + "loss": 0.1025, + "step": 28670 + }, + { + "epoch": 0.6720933912340231, + "grad_norm": 2.6819965839385986, + "learning_rate": 4.855053858590112e-05, + "loss": 0.2663, + "step": 28671 + }, + { + "epoch": 0.6721168328088281, + "grad_norm": 3.8820056915283203, + "learning_rate": 4.854422229221553e-05, + "loss": 0.2772, + "step": 28672 + }, + { + "epoch": 0.6721402743836331, + "grad_norm": 1.3448235988616943, + "learning_rate": 4.853790627773046e-05, + "loss": 0.0995, + "step": 28673 + }, + { + "epoch": 0.672163715958438, + "grad_norm": 1.2126498222351074, + "learning_rate": 4.8531590542480165e-05, + "loss": 1.3483, + "step": 28674 + }, + { + "epoch": 0.6721871575332431, + "grad_norm": 3.6506643295288086, + "learning_rate": 4.852527508649891e-05, + "loss": 0.2849, + "step": 28675 + }, + { + "epoch": 0.672210599108048, + "grad_norm": 0.7099581956863403, + "learning_rate": 4.851895990982096e-05, + "loss": 0.0704, + "step": 28676 + }, + { + "epoch": 0.6722340406828531, + "grad_norm": 3.3225293159484863, + "learning_rate": 4.851264501248054e-05, + "loss": 0.2493, + "step": 28677 + }, + { + "epoch": 0.672257482257658, + "grad_norm": 2.925163745880127, + "learning_rate": 4.8506330394512e-05, + "loss": 0.225, + "step": 28678 + }, + { + "epoch": 0.6722809238324631, + "grad_norm": 3.5026979446411133, + "learning_rate": 4.850001605594955e-05, + "loss": 0.3033, + "step": 28679 + }, + { + "epoch": 0.672304365407268, + "grad_norm": 2.0175509452819824, + "learning_rate": 4.849370199682747e-05, + "loss": 1.6793, + "step": 28680 + }, + { + "epoch": 0.6723278069820731, + "grad_norm": 3.2547056674957275, + "learning_rate": 4.8487388217180006e-05, + "loss": 0.2563, + "step": 28681 + }, + { + "epoch": 0.672351248556878, + "grad_norm": 0.723621666431427, + "learning_rate": 4.848107471704142e-05, + "loss": 0.0679, + "step": 28682 + }, + { + "epoch": 0.672374690131683, + "grad_norm": 0.9908585548400879, + "learning_rate": 4.847476149644594e-05, + "loss": 0.0885, + "step": 28683 + }, + { + "epoch": 0.672398131706488, + "grad_norm": 2.785618782043457, + "learning_rate": 4.8468448555427845e-05, + "loss": 1.2663, + "step": 28684 + }, + { + "epoch": 0.672421573281293, + "grad_norm": 2.4544174671173096, + "learning_rate": 4.846213589402144e-05, + "loss": 1.7038, + "step": 28685 + }, + { + "epoch": 0.672445014856098, + "grad_norm": 3.052760362625122, + "learning_rate": 4.8455823512260936e-05, + "loss": 0.3496, + "step": 28686 + }, + { + "epoch": 0.672468456430903, + "grad_norm": 2.145716428756714, + "learning_rate": 4.8449511410180585e-05, + "loss": 0.071, + "step": 28687 + }, + { + "epoch": 0.6724918980057081, + "grad_norm": 1.9754774570465088, + "learning_rate": 4.844319958781462e-05, + "loss": 0.1702, + "step": 28688 + }, + { + "epoch": 0.672515339580513, + "grad_norm": 1.4070147275924683, + "learning_rate": 4.8436888045197316e-05, + "loss": 0.0764, + "step": 28689 + }, + { + "epoch": 0.6725387811553181, + "grad_norm": 3.5572869777679443, + "learning_rate": 4.843057678236287e-05, + "loss": 0.2084, + "step": 28690 + }, + { + "epoch": 0.672562222730123, + "grad_norm": 4.044216632843018, + "learning_rate": 4.842426579934559e-05, + "loss": 0.2092, + "step": 28691 + }, + { + "epoch": 0.672585664304928, + "grad_norm": 3.1854493618011475, + "learning_rate": 4.841795509617971e-05, + "loss": 0.4042, + "step": 28692 + }, + { + "epoch": 0.672609105879733, + "grad_norm": 0.6736868023872375, + "learning_rate": 4.841164467289945e-05, + "loss": 0.0757, + "step": 28693 + }, + { + "epoch": 0.672632547454538, + "grad_norm": 2.9397244453430176, + "learning_rate": 4.840533452953905e-05, + "loss": 0.1817, + "step": 28694 + }, + { + "epoch": 0.672655989029343, + "grad_norm": 2.2981863021850586, + "learning_rate": 4.8399024666132765e-05, + "loss": 0.1633, + "step": 28695 + }, + { + "epoch": 0.672679430604148, + "grad_norm": 3.4589078426361084, + "learning_rate": 4.8392715082714776e-05, + "loss": 1.5478, + "step": 28696 + }, + { + "epoch": 0.672702872178953, + "grad_norm": 2.4690613746643066, + "learning_rate": 4.838640577931938e-05, + "loss": 1.8364, + "step": 28697 + }, + { + "epoch": 0.672726313753758, + "grad_norm": 0.2601863741874695, + "learning_rate": 4.838009675598082e-05, + "loss": 0.0129, + "step": 28698 + }, + { + "epoch": 0.672749755328563, + "grad_norm": 2.760281562805176, + "learning_rate": 4.837378801273331e-05, + "loss": 0.2203, + "step": 28699 + }, + { + "epoch": 0.672773196903368, + "grad_norm": 6.814544200897217, + "learning_rate": 4.836747954961108e-05, + "loss": 0.815, + "step": 28700 + }, + { + "epoch": 0.672796638478173, + "grad_norm": 4.006589412689209, + "learning_rate": 4.836117136664836e-05, + "loss": 0.355, + "step": 28701 + }, + { + "epoch": 0.672820080052978, + "grad_norm": 0.2699829638004303, + "learning_rate": 4.835486346387939e-05, + "loss": 0.0124, + "step": 28702 + }, + { + "epoch": 0.6728435216277829, + "grad_norm": 3.200007200241089, + "learning_rate": 4.8348555841338314e-05, + "loss": 0.1606, + "step": 28703 + }, + { + "epoch": 0.672866963202588, + "grad_norm": 1.3001192808151245, + "learning_rate": 4.8342248499059484e-05, + "loss": 0.0906, + "step": 28704 + }, + { + "epoch": 0.6728904047773929, + "grad_norm": 4.212231159210205, + "learning_rate": 4.833594143707707e-05, + "loss": 0.3822, + "step": 28705 + }, + { + "epoch": 0.672913846352198, + "grad_norm": 2.2872116565704346, + "learning_rate": 4.8329634655425284e-05, + "loss": 0.1371, + "step": 28706 + }, + { + "epoch": 0.6729372879270029, + "grad_norm": 3.273393392562866, + "learning_rate": 4.832332815413835e-05, + "loss": 0.3218, + "step": 28707 + }, + { + "epoch": 0.672960729501808, + "grad_norm": 1.1803460121154785, + "learning_rate": 4.831702193325045e-05, + "loss": 0.1713, + "step": 28708 + }, + { + "epoch": 0.6729841710766129, + "grad_norm": 3.540050745010376, + "learning_rate": 4.8310715992795885e-05, + "loss": 0.4393, + "step": 28709 + }, + { + "epoch": 0.673007612651418, + "grad_norm": 3.091015100479126, + "learning_rate": 4.830441033280879e-05, + "loss": 0.2755, + "step": 28710 + }, + { + "epoch": 0.6730310542262229, + "grad_norm": 0.7476053833961487, + "learning_rate": 4.8298104953323455e-05, + "loss": 0.0625, + "step": 28711 + }, + { + "epoch": 0.6730544958010279, + "grad_norm": 2.5349135398864746, + "learning_rate": 4.829179985437407e-05, + "loss": 0.2111, + "step": 28712 + }, + { + "epoch": 0.6730779373758329, + "grad_norm": 1.4000672101974487, + "learning_rate": 4.828549503599481e-05, + "loss": 0.099, + "step": 28713 + }, + { + "epoch": 0.6731013789506379, + "grad_norm": 0.4129062592983246, + "learning_rate": 4.827919049821992e-05, + "loss": 0.0326, + "step": 28714 + }, + { + "epoch": 0.6731248205254429, + "grad_norm": 0.6192899942398071, + "learning_rate": 4.827288624108356e-05, + "loss": 0.0397, + "step": 28715 + }, + { + "epoch": 0.6731482621002479, + "grad_norm": 1.0257319211959839, + "learning_rate": 4.8266582264620005e-05, + "loss": 0.0975, + "step": 28716 + }, + { + "epoch": 0.6731717036750529, + "grad_norm": 3.2158126831054688, + "learning_rate": 4.826027856886344e-05, + "loss": 0.2959, + "step": 28717 + }, + { + "epoch": 0.6731951452498579, + "grad_norm": 0.9493342041969299, + "learning_rate": 4.825397515384804e-05, + "loss": 0.0559, + "step": 28718 + }, + { + "epoch": 0.6732185868246628, + "grad_norm": 4.283229827880859, + "learning_rate": 4.824767201960804e-05, + "loss": 0.4123, + "step": 28719 + }, + { + "epoch": 0.6732420283994679, + "grad_norm": 1.7231533527374268, + "learning_rate": 4.8241369166177616e-05, + "loss": 0.1507, + "step": 28720 + }, + { + "epoch": 0.6732654699742728, + "grad_norm": 2.0002620220184326, + "learning_rate": 4.823506659359096e-05, + "loss": 0.1703, + "step": 28721 + }, + { + "epoch": 0.6732889115490779, + "grad_norm": 0.543109118938446, + "learning_rate": 4.822876430188227e-05, + "loss": 0.0376, + "step": 28722 + }, + { + "epoch": 0.6733123531238828, + "grad_norm": 2.35630202293396, + "learning_rate": 4.8222462291085804e-05, + "loss": 0.1766, + "step": 28723 + }, + { + "epoch": 0.6733357946986879, + "grad_norm": 0.28174924850463867, + "learning_rate": 4.8216160561235715e-05, + "loss": 0.0318, + "step": 28724 + }, + { + "epoch": 0.6733592362734928, + "grad_norm": 1.098055362701416, + "learning_rate": 4.820985911236619e-05, + "loss": 0.0477, + "step": 28725 + }, + { + "epoch": 0.6733826778482979, + "grad_norm": 3.6678223609924316, + "learning_rate": 4.820355794451143e-05, + "loss": 0.281, + "step": 28726 + }, + { + "epoch": 0.6734061194231028, + "grad_norm": 3.2604217529296875, + "learning_rate": 4.8197257057705626e-05, + "loss": 0.2118, + "step": 28727 + }, + { + "epoch": 0.6734295609979079, + "grad_norm": 3.581420660018921, + "learning_rate": 4.819095645198291e-05, + "loss": 0.4319, + "step": 28728 + }, + { + "epoch": 0.6734530025727128, + "grad_norm": 4.617359638214111, + "learning_rate": 4.818465612737758e-05, + "loss": 0.4549, + "step": 28729 + }, + { + "epoch": 0.6734764441475178, + "grad_norm": 3.1917836666107178, + "learning_rate": 4.8178356083923746e-05, + "loss": 0.2646, + "step": 28730 + }, + { + "epoch": 0.6734998857223228, + "grad_norm": 2.5245189666748047, + "learning_rate": 4.817205632165561e-05, + "loss": 1.6434, + "step": 28731 + }, + { + "epoch": 0.6735233272971278, + "grad_norm": 3.038602352142334, + "learning_rate": 4.8165756840607354e-05, + "loss": 1.5719, + "step": 28732 + }, + { + "epoch": 0.6735467688719328, + "grad_norm": 2.882847309112549, + "learning_rate": 4.8159457640813135e-05, + "loss": 0.2823, + "step": 28733 + }, + { + "epoch": 0.6735702104467378, + "grad_norm": 3.102895975112915, + "learning_rate": 4.815315872230718e-05, + "loss": 0.3964, + "step": 28734 + }, + { + "epoch": 0.6735936520215428, + "grad_norm": 1.2418068647384644, + "learning_rate": 4.81468600851236e-05, + "loss": 0.1459, + "step": 28735 + }, + { + "epoch": 0.6736170935963478, + "grad_norm": 2.399339437484741, + "learning_rate": 4.814056172929667e-05, + "loss": 0.145, + "step": 28736 + }, + { + "epoch": 0.6736405351711527, + "grad_norm": 2.3762476444244385, + "learning_rate": 4.81342636548605e-05, + "loss": 0.9882, + "step": 28737 + }, + { + "epoch": 0.6736639767459578, + "grad_norm": 3.2303760051727295, + "learning_rate": 4.8127965861849276e-05, + "loss": 0.194, + "step": 28738 + }, + { + "epoch": 0.6736874183207628, + "grad_norm": 2.436662435531616, + "learning_rate": 4.8121668350297166e-05, + "loss": 0.1748, + "step": 28739 + }, + { + "epoch": 0.6737108598955678, + "grad_norm": 0.3246423304080963, + "learning_rate": 4.811537112023835e-05, + "loss": 0.0321, + "step": 28740 + }, + { + "epoch": 0.6737343014703728, + "grad_norm": 2.7079532146453857, + "learning_rate": 4.810907417170693e-05, + "loss": 1.9556, + "step": 28741 + }, + { + "epoch": 0.6737577430451778, + "grad_norm": 2.0833778381347656, + "learning_rate": 4.8102777504737196e-05, + "loss": 0.9954, + "step": 28742 + }, + { + "epoch": 0.6737811846199828, + "grad_norm": 1.0555227994918823, + "learning_rate": 4.809648111936324e-05, + "loss": 0.0567, + "step": 28743 + }, + { + "epoch": 0.6738046261947878, + "grad_norm": 2.7941951751708984, + "learning_rate": 4.8090185015619235e-05, + "loss": 0.2335, + "step": 28744 + }, + { + "epoch": 0.6738280677695928, + "grad_norm": 1.978804111480713, + "learning_rate": 4.808388919353931e-05, + "loss": 1.2459, + "step": 28745 + }, + { + "epoch": 0.6738515093443977, + "grad_norm": 4.194635391235352, + "learning_rate": 4.8077593653157707e-05, + "loss": 0.337, + "step": 28746 + }, + { + "epoch": 0.6738749509192028, + "grad_norm": 3.638434410095215, + "learning_rate": 4.8071298394508525e-05, + "loss": 0.3156, + "step": 28747 + }, + { + "epoch": 0.6738983924940077, + "grad_norm": 5.750690460205078, + "learning_rate": 4.806500341762591e-05, + "loss": 0.5835, + "step": 28748 + }, + { + "epoch": 0.6739218340688128, + "grad_norm": 0.2008754462003708, + "learning_rate": 4.805870872254409e-05, + "loss": 0.0165, + "step": 28749 + }, + { + "epoch": 0.6739452756436177, + "grad_norm": 3.3409218788146973, + "learning_rate": 4.805241430929717e-05, + "loss": 0.2547, + "step": 28750 + }, + { + "epoch": 0.6739687172184228, + "grad_norm": 1.5638943910598755, + "learning_rate": 4.8046120177919307e-05, + "loss": 1.1633, + "step": 28751 + }, + { + "epoch": 0.6739921587932277, + "grad_norm": 0.9660997986793518, + "learning_rate": 4.803982632844466e-05, + "loss": 0.0598, + "step": 28752 + }, + { + "epoch": 0.6740156003680328, + "grad_norm": 3.701786518096924, + "learning_rate": 4.803353276090734e-05, + "loss": 2.0508, + "step": 28753 + }, + { + "epoch": 0.6740390419428377, + "grad_norm": 1.5530033111572266, + "learning_rate": 4.802723947534157e-05, + "loss": 0.1314, + "step": 28754 + }, + { + "epoch": 0.6740624835176428, + "grad_norm": 4.74641752243042, + "learning_rate": 4.802094647178146e-05, + "loss": 0.4991, + "step": 28755 + }, + { + "epoch": 0.6740859250924477, + "grad_norm": 4.005130290985107, + "learning_rate": 4.801465375026115e-05, + "loss": 0.329, + "step": 28756 + }, + { + "epoch": 0.6741093666672527, + "grad_norm": 2.1997737884521484, + "learning_rate": 4.8008361310814755e-05, + "loss": 1.3844, + "step": 28757 + }, + { + "epoch": 0.6741328082420577, + "grad_norm": 1.1566569805145264, + "learning_rate": 4.800206915347649e-05, + "loss": 0.0705, + "step": 28758 + }, + { + "epoch": 0.6741562498168627, + "grad_norm": 2.548652172088623, + "learning_rate": 4.799577727828047e-05, + "loss": 0.1464, + "step": 28759 + }, + { + "epoch": 0.6741796913916677, + "grad_norm": 1.943406343460083, + "learning_rate": 4.7989485685260785e-05, + "loss": 0.1277, + "step": 28760 + }, + { + "epoch": 0.6742031329664727, + "grad_norm": 2.299147605895996, + "learning_rate": 4.7983194374451645e-05, + "loss": 0.1816, + "step": 28761 + }, + { + "epoch": 0.6742265745412777, + "grad_norm": 2.0113935470581055, + "learning_rate": 4.797690334588716e-05, + "loss": 0.1382, + "step": 28762 + }, + { + "epoch": 0.6742500161160827, + "grad_norm": 1.455094814300537, + "learning_rate": 4.797061259960145e-05, + "loss": 0.1388, + "step": 28763 + }, + { + "epoch": 0.6742734576908876, + "grad_norm": 2.734844446182251, + "learning_rate": 4.796432213562867e-05, + "loss": 0.3235, + "step": 28764 + }, + { + "epoch": 0.6742968992656927, + "grad_norm": 3.914846658706665, + "learning_rate": 4.795803195400294e-05, + "loss": 0.4786, + "step": 28765 + }, + { + "epoch": 0.6743203408404976, + "grad_norm": 4.686838626861572, + "learning_rate": 4.795174205475834e-05, + "loss": 0.2524, + "step": 28766 + }, + { + "epoch": 0.6743437824153027, + "grad_norm": 2.2828545570373535, + "learning_rate": 4.7945452437929104e-05, + "loss": 0.1356, + "step": 28767 + }, + { + "epoch": 0.6743672239901076, + "grad_norm": 1.5980801582336426, + "learning_rate": 4.79391631035493e-05, + "loss": 0.168, + "step": 28768 + }, + { + "epoch": 0.6743906655649127, + "grad_norm": 1.2154667377471924, + "learning_rate": 4.793287405165303e-05, + "loss": 0.1128, + "step": 28769 + }, + { + "epoch": 0.6744141071397176, + "grad_norm": 1.6652756929397583, + "learning_rate": 4.7926585282274484e-05, + "loss": 0.0949, + "step": 28770 + }, + { + "epoch": 0.6744375487145227, + "grad_norm": 0.3964783251285553, + "learning_rate": 4.792029679544775e-05, + "loss": 0.0151, + "step": 28771 + }, + { + "epoch": 0.6744609902893276, + "grad_norm": 3.3966145515441895, + "learning_rate": 4.791400859120695e-05, + "loss": 0.1635, + "step": 28772 + }, + { + "epoch": 0.6744844318641326, + "grad_norm": 4.051290988922119, + "learning_rate": 4.790772066958616e-05, + "loss": 0.2588, + "step": 28773 + }, + { + "epoch": 0.6745078734389376, + "grad_norm": 2.313044786453247, + "learning_rate": 4.790143303061958e-05, + "loss": 1.6094, + "step": 28774 + }, + { + "epoch": 0.6745313150137426, + "grad_norm": 2.6628754138946533, + "learning_rate": 4.7895145674341305e-05, + "loss": 0.2109, + "step": 28775 + }, + { + "epoch": 0.6745547565885476, + "grad_norm": 1.0469577312469482, + "learning_rate": 4.788885860078541e-05, + "loss": 0.1018, + "step": 28776 + }, + { + "epoch": 0.6745781981633526, + "grad_norm": 2.2136948108673096, + "learning_rate": 4.788257180998605e-05, + "loss": 0.1967, + "step": 28777 + }, + { + "epoch": 0.6746016397381576, + "grad_norm": 2.515071392059326, + "learning_rate": 4.787628530197732e-05, + "loss": 0.258, + "step": 28778 + }, + { + "epoch": 0.6746250813129626, + "grad_norm": 2.1371853351593018, + "learning_rate": 4.786999907679329e-05, + "loss": 0.2162, + "step": 28779 + }, + { + "epoch": 0.6746485228877676, + "grad_norm": 1.7902724742889404, + "learning_rate": 4.786371313446816e-05, + "loss": 1.3253, + "step": 28780 + }, + { + "epoch": 0.6746719644625726, + "grad_norm": 1.076192021369934, + "learning_rate": 4.785742747503598e-05, + "loss": 0.0772, + "step": 28781 + }, + { + "epoch": 0.6746954060373775, + "grad_norm": 0.7541303634643555, + "learning_rate": 4.7851142098530825e-05, + "loss": 0.0477, + "step": 28782 + }, + { + "epoch": 0.6747188476121826, + "grad_norm": 2.547028064727783, + "learning_rate": 4.784485700498688e-05, + "loss": 0.2605, + "step": 28783 + }, + { + "epoch": 0.6747422891869875, + "grad_norm": 2.3231618404388428, + "learning_rate": 4.7838572194438205e-05, + "loss": 0.2156, + "step": 28784 + }, + { + "epoch": 0.6747657307617926, + "grad_norm": 2.073667287826538, + "learning_rate": 4.7832287666918896e-05, + "loss": 0.1008, + "step": 28785 + }, + { + "epoch": 0.6747891723365975, + "grad_norm": 2.8743629455566406, + "learning_rate": 4.7826003422463036e-05, + "loss": 0.2945, + "step": 28786 + }, + { + "epoch": 0.6748126139114026, + "grad_norm": 4.182389259338379, + "learning_rate": 4.7819719461104785e-05, + "loss": 0.4985, + "step": 28787 + }, + { + "epoch": 0.6748360554862075, + "grad_norm": 1.7310270071029663, + "learning_rate": 4.7813435782878205e-05, + "loss": 0.1166, + "step": 28788 + }, + { + "epoch": 0.6748594970610126, + "grad_norm": 1.5969074964523315, + "learning_rate": 4.7807152387817376e-05, + "loss": 0.1223, + "step": 28789 + }, + { + "epoch": 0.6748829386358176, + "grad_norm": 3.7543413639068604, + "learning_rate": 4.7800869275956416e-05, + "loss": 0.4324, + "step": 28790 + }, + { + "epoch": 0.6749063802106225, + "grad_norm": 4.623332500457764, + "learning_rate": 4.779458644732937e-05, + "loss": 0.3071, + "step": 28791 + }, + { + "epoch": 0.6749298217854276, + "grad_norm": 2.1072397232055664, + "learning_rate": 4.77883039019704e-05, + "loss": 1.2614, + "step": 28792 + }, + { + "epoch": 0.6749532633602325, + "grad_norm": 2.80849027633667, + "learning_rate": 4.7782021639913566e-05, + "loss": 0.2324, + "step": 28793 + }, + { + "epoch": 0.6749767049350376, + "grad_norm": 2.6032209396362305, + "learning_rate": 4.7775739661192906e-05, + "loss": 0.1746, + "step": 28794 + }, + { + "epoch": 0.6750001465098425, + "grad_norm": 3.4557418823242188, + "learning_rate": 4.776945796584259e-05, + "loss": 0.3153, + "step": 28795 + }, + { + "epoch": 0.6750235880846476, + "grad_norm": 2.478071451187134, + "learning_rate": 4.776317655389667e-05, + "loss": 0.1661, + "step": 28796 + }, + { + "epoch": 0.6750470296594525, + "grad_norm": 3.3221001625061035, + "learning_rate": 4.775689542538921e-05, + "loss": 0.334, + "step": 28797 + }, + { + "epoch": 0.6750704712342576, + "grad_norm": 1.6724801063537598, + "learning_rate": 4.775061458035428e-05, + "loss": 0.1273, + "step": 28798 + }, + { + "epoch": 0.6750939128090625, + "grad_norm": 1.406427025794983, + "learning_rate": 4.774433401882602e-05, + "loss": 0.1257, + "step": 28799 + }, + { + "epoch": 0.6751173543838676, + "grad_norm": 2.435314178466797, + "learning_rate": 4.7738053740838476e-05, + "loss": 0.1494, + "step": 28800 + }, + { + "epoch": 0.6751407959586725, + "grad_norm": 3.3438498973846436, + "learning_rate": 4.773177374642571e-05, + "loss": 0.2511, + "step": 28801 + }, + { + "epoch": 0.6751642375334775, + "grad_norm": 3.4203078746795654, + "learning_rate": 4.772549403562182e-05, + "loss": 0.3716, + "step": 28802 + }, + { + "epoch": 0.6751876791082825, + "grad_norm": 4.2231764793396, + "learning_rate": 4.771921460846086e-05, + "loss": 0.3144, + "step": 28803 + }, + { + "epoch": 0.6752111206830875, + "grad_norm": 2.8351995944976807, + "learning_rate": 4.7712935464976884e-05, + "loss": 0.3845, + "step": 28804 + }, + { + "epoch": 0.6752345622578925, + "grad_norm": 2.9571287631988525, + "learning_rate": 4.770665660520401e-05, + "loss": 1.1222, + "step": 28805 + }, + { + "epoch": 0.6752580038326975, + "grad_norm": 0.8909620642662048, + "learning_rate": 4.7700378029176275e-05, + "loss": 0.0724, + "step": 28806 + }, + { + "epoch": 0.6752814454075025, + "grad_norm": 1.7976425886154175, + "learning_rate": 4.769409973692778e-05, + "loss": 0.188, + "step": 28807 + }, + { + "epoch": 0.6753048869823075, + "grad_norm": 2.2788209915161133, + "learning_rate": 4.7687821728492575e-05, + "loss": 0.1502, + "step": 28808 + }, + { + "epoch": 0.6753283285571124, + "grad_norm": 2.7713630199432373, + "learning_rate": 4.768154400390472e-05, + "loss": 0.413, + "step": 28809 + }, + { + "epoch": 0.6753517701319175, + "grad_norm": 2.6249072551727295, + "learning_rate": 4.7675266563198286e-05, + "loss": 0.1435, + "step": 28810 + }, + { + "epoch": 0.6753752117067224, + "grad_norm": 0.6425501108169556, + "learning_rate": 4.7668989406407286e-05, + "loss": 0.0634, + "step": 28811 + }, + { + "epoch": 0.6753986532815275, + "grad_norm": 2.564894437789917, + "learning_rate": 4.766271253356585e-05, + "loss": 0.2277, + "step": 28812 + }, + { + "epoch": 0.6754220948563324, + "grad_norm": 2.3700203895568848, + "learning_rate": 4.765643594470801e-05, + "loss": 0.215, + "step": 28813 + }, + { + "epoch": 0.6754455364311375, + "grad_norm": 3.8654308319091797, + "learning_rate": 4.765015963986782e-05, + "loss": 0.5149, + "step": 28814 + }, + { + "epoch": 0.6754689780059424, + "grad_norm": 2.6182568073272705, + "learning_rate": 4.7643883619079345e-05, + "loss": 0.1654, + "step": 28815 + }, + { + "epoch": 0.6754924195807475, + "grad_norm": 3.2571706771850586, + "learning_rate": 4.763760788237661e-05, + "loss": 0.1487, + "step": 28816 + }, + { + "epoch": 0.6755158611555524, + "grad_norm": 3.3670334815979004, + "learning_rate": 4.763133242979367e-05, + "loss": 0.1641, + "step": 28817 + }, + { + "epoch": 0.6755393027303574, + "grad_norm": 2.658205032348633, + "learning_rate": 4.762505726136458e-05, + "loss": 0.1081, + "step": 28818 + }, + { + "epoch": 0.6755627443051624, + "grad_norm": 2.8408660888671875, + "learning_rate": 4.761878237712345e-05, + "loss": 0.1101, + "step": 28819 + }, + { + "epoch": 0.6755861858799674, + "grad_norm": 2.8987550735473633, + "learning_rate": 4.7612507777104277e-05, + "loss": 0.3062, + "step": 28820 + }, + { + "epoch": 0.6756096274547724, + "grad_norm": 2.656163215637207, + "learning_rate": 4.76062334613411e-05, + "loss": 1.4956, + "step": 28821 + }, + { + "epoch": 0.6756330690295774, + "grad_norm": 2.5020761489868164, + "learning_rate": 4.759995942986798e-05, + "loss": 0.237, + "step": 28822 + }, + { + "epoch": 0.6756565106043824, + "grad_norm": 0.47159332036972046, + "learning_rate": 4.759368568271896e-05, + "loss": 0.0332, + "step": 28823 + }, + { + "epoch": 0.6756799521791874, + "grad_norm": 3.4785590171813965, + "learning_rate": 4.758741221992803e-05, + "loss": 0.439, + "step": 28824 + }, + { + "epoch": 0.6757033937539924, + "grad_norm": 1.6229296922683716, + "learning_rate": 4.758113904152931e-05, + "loss": 0.2315, + "step": 28825 + }, + { + "epoch": 0.6757268353287974, + "grad_norm": 0.7205690145492554, + "learning_rate": 4.757486614755681e-05, + "loss": 0.0333, + "step": 28826 + }, + { + "epoch": 0.6757502769036023, + "grad_norm": 1.8853942155838013, + "learning_rate": 4.756859353804456e-05, + "loss": 0.058, + "step": 28827 + }, + { + "epoch": 0.6757737184784074, + "grad_norm": 1.1010226011276245, + "learning_rate": 4.7562321213026585e-05, + "loss": 0.0958, + "step": 28828 + }, + { + "epoch": 0.6757971600532123, + "grad_norm": 1.5669151544570923, + "learning_rate": 4.7556049172536895e-05, + "loss": 0.1242, + "step": 28829 + }, + { + "epoch": 0.6758206016280174, + "grad_norm": 1.9709116220474243, + "learning_rate": 4.754977741660961e-05, + "loss": 0.2104, + "step": 28830 + }, + { + "epoch": 0.6758440432028223, + "grad_norm": 4.168389320373535, + "learning_rate": 4.7543505945278656e-05, + "loss": 0.4582, + "step": 28831 + }, + { + "epoch": 0.6758674847776274, + "grad_norm": 3.8402795791625977, + "learning_rate": 4.753723475857815e-05, + "loss": 0.2547, + "step": 28832 + }, + { + "epoch": 0.6758909263524323, + "grad_norm": 1.944196105003357, + "learning_rate": 4.7530963856542086e-05, + "loss": 0.1277, + "step": 28833 + }, + { + "epoch": 0.6759143679272374, + "grad_norm": 1.5502921342849731, + "learning_rate": 4.75246932392045e-05, + "loss": 0.0865, + "step": 28834 + }, + { + "epoch": 0.6759378095020423, + "grad_norm": 7.167579174041748, + "learning_rate": 4.751842290659939e-05, + "loss": 0.4688, + "step": 28835 + }, + { + "epoch": 0.6759612510768473, + "grad_norm": 3.00639009475708, + "learning_rate": 4.751215285876076e-05, + "loss": 0.265, + "step": 28836 + }, + { + "epoch": 0.6759846926516523, + "grad_norm": 2.83612322807312, + "learning_rate": 4.75058830957227e-05, + "loss": 0.2809, + "step": 28837 + }, + { + "epoch": 0.6760081342264573, + "grad_norm": 1.6630529165267944, + "learning_rate": 4.749961361751919e-05, + "loss": 0.1439, + "step": 28838 + }, + { + "epoch": 0.6760315758012623, + "grad_norm": 4.257303237915039, + "learning_rate": 4.749334442418426e-05, + "loss": 0.2698, + "step": 28839 + }, + { + "epoch": 0.6760550173760673, + "grad_norm": 2.328951120376587, + "learning_rate": 4.7487075515751924e-05, + "loss": 0.1216, + "step": 28840 + }, + { + "epoch": 0.6760784589508724, + "grad_norm": 0.5595044493675232, + "learning_rate": 4.748080689225618e-05, + "loss": 0.0247, + "step": 28841 + }, + { + "epoch": 0.6761019005256773, + "grad_norm": 0.6600719094276428, + "learning_rate": 4.7474538553731016e-05, + "loss": 0.0607, + "step": 28842 + }, + { + "epoch": 0.6761253421004824, + "grad_norm": 0.9807417988777161, + "learning_rate": 4.74682705002105e-05, + "loss": 0.044, + "step": 28843 + }, + { + "epoch": 0.6761487836752873, + "grad_norm": 0.6105033755302429, + "learning_rate": 4.746200273172865e-05, + "loss": 0.0896, + "step": 28844 + }, + { + "epoch": 0.6761722252500924, + "grad_norm": 0.3779311180114746, + "learning_rate": 4.7455735248319455e-05, + "loss": 0.0179, + "step": 28845 + }, + { + "epoch": 0.6761956668248973, + "grad_norm": 2.5232226848602295, + "learning_rate": 4.744946805001691e-05, + "loss": 0.226, + "step": 28846 + }, + { + "epoch": 0.6762191083997023, + "grad_norm": 4.0254435539245605, + "learning_rate": 4.744320113685504e-05, + "loss": 0.265, + "step": 28847 + }, + { + "epoch": 0.6762425499745073, + "grad_norm": 1.453165888786316, + "learning_rate": 4.7436934508867836e-05, + "loss": 0.0571, + "step": 28848 + }, + { + "epoch": 0.6762659915493123, + "grad_norm": 0.7941531538963318, + "learning_rate": 4.7430668166089255e-05, + "loss": 0.0291, + "step": 28849 + }, + { + "epoch": 0.6762894331241173, + "grad_norm": 4.872342586517334, + "learning_rate": 4.74244021085534e-05, + "loss": 0.3813, + "step": 28850 + }, + { + "epoch": 0.6763128746989223, + "grad_norm": 2.688539981842041, + "learning_rate": 4.74181363362942e-05, + "loss": 0.0925, + "step": 28851 + }, + { + "epoch": 0.6763363162737273, + "grad_norm": 3.5266191959381104, + "learning_rate": 4.741187084934568e-05, + "loss": 0.1989, + "step": 28852 + }, + { + "epoch": 0.6763597578485323, + "grad_norm": 3.4292056560516357, + "learning_rate": 4.7405605647741835e-05, + "loss": 0.3156, + "step": 28853 + }, + { + "epoch": 0.6763831994233372, + "grad_norm": 1.5160093307495117, + "learning_rate": 4.739934073151664e-05, + "loss": 0.1702, + "step": 28854 + }, + { + "epoch": 0.6764066409981423, + "grad_norm": 1.9781147241592407, + "learning_rate": 4.739307610070407e-05, + "loss": 0.1894, + "step": 28855 + }, + { + "epoch": 0.6764300825729472, + "grad_norm": 3.094801664352417, + "learning_rate": 4.738681175533814e-05, + "loss": 0.249, + "step": 28856 + }, + { + "epoch": 0.6764535241477523, + "grad_norm": 4.914846420288086, + "learning_rate": 4.7380547695452905e-05, + "loss": 0.2792, + "step": 28857 + }, + { + "epoch": 0.6764769657225572, + "grad_norm": 3.902219295501709, + "learning_rate": 4.737428392108229e-05, + "loss": 0.3851, + "step": 28858 + }, + { + "epoch": 0.6765004072973623, + "grad_norm": 3.1189396381378174, + "learning_rate": 4.7368020432260286e-05, + "loss": 0.3073, + "step": 28859 + }, + { + "epoch": 0.6765238488721672, + "grad_norm": 2.6868278980255127, + "learning_rate": 4.736175722902088e-05, + "loss": 0.2309, + "step": 28860 + }, + { + "epoch": 0.6765472904469723, + "grad_norm": 3.744014263153076, + "learning_rate": 4.7355494311398064e-05, + "loss": 0.3261, + "step": 28861 + }, + { + "epoch": 0.6765707320217772, + "grad_norm": 5.126159191131592, + "learning_rate": 4.7349231679425776e-05, + "loss": 0.6537, + "step": 28862 + }, + { + "epoch": 0.6765941735965822, + "grad_norm": 0.21526460349559784, + "learning_rate": 4.734296933313808e-05, + "loss": 0.0231, + "step": 28863 + }, + { + "epoch": 0.6766176151713872, + "grad_norm": 3.0171902179718018, + "learning_rate": 4.73367072725689e-05, + "loss": 0.222, + "step": 28864 + }, + { + "epoch": 0.6766410567461922, + "grad_norm": 0.8985026478767395, + "learning_rate": 4.7330445497752243e-05, + "loss": 0.0921, + "step": 28865 + }, + { + "epoch": 0.6766644983209972, + "grad_norm": 1.458085298538208, + "learning_rate": 4.732418400872206e-05, + "loss": 0.1064, + "step": 28866 + }, + { + "epoch": 0.6766879398958022, + "grad_norm": 1.8668272495269775, + "learning_rate": 4.7317922805512304e-05, + "loss": 0.0972, + "step": 28867 + }, + { + "epoch": 0.6767113814706072, + "grad_norm": 3.3509538173675537, + "learning_rate": 4.731166188815701e-05, + "loss": 0.5091, + "step": 28868 + }, + { + "epoch": 0.6767348230454122, + "grad_norm": 2.0261240005493164, + "learning_rate": 4.730540125669009e-05, + "loss": 0.1581, + "step": 28869 + }, + { + "epoch": 0.6767582646202172, + "grad_norm": 3.409306526184082, + "learning_rate": 4.729914091114558e-05, + "loss": 0.3531, + "step": 28870 + }, + { + "epoch": 0.6767817061950222, + "grad_norm": 4.87249755859375, + "learning_rate": 4.7292880851557405e-05, + "loss": 0.2274, + "step": 28871 + }, + { + "epoch": 0.6768051477698271, + "grad_norm": 0.5959218144416809, + "learning_rate": 4.728662107795956e-05, + "loss": 0.0341, + "step": 28872 + }, + { + "epoch": 0.6768285893446322, + "grad_norm": 3.4119863510131836, + "learning_rate": 4.728036159038597e-05, + "loss": 0.3819, + "step": 28873 + }, + { + "epoch": 0.6768520309194371, + "grad_norm": 5.748258590698242, + "learning_rate": 4.72741023888706e-05, + "loss": 0.2123, + "step": 28874 + }, + { + "epoch": 0.6768754724942422, + "grad_norm": 1.0956398248672485, + "learning_rate": 4.726784347344746e-05, + "loss": 0.11, + "step": 28875 + }, + { + "epoch": 0.6768989140690471, + "grad_norm": 2.8105251789093018, + "learning_rate": 4.7261584844150485e-05, + "loss": 0.412, + "step": 28876 + }, + { + "epoch": 0.6769223556438522, + "grad_norm": 1.811165690422058, + "learning_rate": 4.725532650101363e-05, + "loss": 0.1045, + "step": 28877 + }, + { + "epoch": 0.6769457972186571, + "grad_norm": 2.571335554122925, + "learning_rate": 4.724906844407087e-05, + "loss": 0.2032, + "step": 28878 + }, + { + "epoch": 0.6769692387934622, + "grad_norm": 3.3238818645477295, + "learning_rate": 4.7242810673356094e-05, + "loss": 0.2933, + "step": 28879 + }, + { + "epoch": 0.6769926803682671, + "grad_norm": 0.6914038062095642, + "learning_rate": 4.723655318890336e-05, + "loss": 0.058, + "step": 28880 + }, + { + "epoch": 0.6770161219430721, + "grad_norm": 3.026750087738037, + "learning_rate": 4.7230295990746534e-05, + "loss": 0.1956, + "step": 28881 + }, + { + "epoch": 0.6770395635178771, + "grad_norm": 2.5329689979553223, + "learning_rate": 4.722403907891963e-05, + "loss": 0.4289, + "step": 28882 + }, + { + "epoch": 0.6770630050926821, + "grad_norm": 2.9811744689941406, + "learning_rate": 4.721778245345659e-05, + "loss": 0.1309, + "step": 28883 + }, + { + "epoch": 0.6770864466674871, + "grad_norm": 3.135356903076172, + "learning_rate": 4.721152611439134e-05, + "loss": 0.1726, + "step": 28884 + }, + { + "epoch": 0.6771098882422921, + "grad_norm": 2.3136658668518066, + "learning_rate": 4.7205270061757834e-05, + "loss": 0.1656, + "step": 28885 + }, + { + "epoch": 0.6771333298170971, + "grad_norm": 3.4367175102233887, + "learning_rate": 4.7199014295590014e-05, + "loss": 0.5266, + "step": 28886 + }, + { + "epoch": 0.6771567713919021, + "grad_norm": 2.8004395961761475, + "learning_rate": 4.71927588159218e-05, + "loss": 0.4222, + "step": 28887 + }, + { + "epoch": 0.677180212966707, + "grad_norm": 1.7602251768112183, + "learning_rate": 4.718650362278718e-05, + "loss": 0.1951, + "step": 28888 + }, + { + "epoch": 0.6772036545415121, + "grad_norm": 0.8207140564918518, + "learning_rate": 4.71802487162201e-05, + "loss": 0.0634, + "step": 28889 + }, + { + "epoch": 0.677227096116317, + "grad_norm": 3.6443047523498535, + "learning_rate": 4.717399409625446e-05, + "loss": 0.2411, + "step": 28890 + }, + { + "epoch": 0.6772505376911221, + "grad_norm": 3.7166998386383057, + "learning_rate": 4.716773976292418e-05, + "loss": 0.3617, + "step": 28891 + }, + { + "epoch": 0.6772739792659271, + "grad_norm": 2.2918965816497803, + "learning_rate": 4.716148571626326e-05, + "loss": 0.3439, + "step": 28892 + }, + { + "epoch": 0.6772974208407321, + "grad_norm": 0.7515818476676941, + "learning_rate": 4.7155231956305615e-05, + "loss": 0.0617, + "step": 28893 + }, + { + "epoch": 0.6773208624155371, + "grad_norm": 2.03671932220459, + "learning_rate": 4.714897848308513e-05, + "loss": 0.066, + "step": 28894 + }, + { + "epoch": 0.6773443039903421, + "grad_norm": 4.0206217765808105, + "learning_rate": 4.7142725296635804e-05, + "loss": 1.5579, + "step": 28895 + }, + { + "epoch": 0.6773677455651471, + "grad_norm": 3.5515384674072266, + "learning_rate": 4.713647239699154e-05, + "loss": 0.1447, + "step": 28896 + }, + { + "epoch": 0.6773911871399521, + "grad_norm": 1.158331274986267, + "learning_rate": 4.713021978418626e-05, + "loss": 0.1361, + "step": 28897 + }, + { + "epoch": 0.6774146287147571, + "grad_norm": 1.0383727550506592, + "learning_rate": 4.712396745825389e-05, + "loss": 0.1272, + "step": 28898 + }, + { + "epoch": 0.677438070289562, + "grad_norm": 2.6771984100341797, + "learning_rate": 4.711771541922837e-05, + "loss": 0.4294, + "step": 28899 + }, + { + "epoch": 0.6774615118643671, + "grad_norm": 3.6052470207214355, + "learning_rate": 4.7111463667143576e-05, + "loss": 0.2468, + "step": 28900 + }, + { + "epoch": 0.677484953439172, + "grad_norm": 4.615069389343262, + "learning_rate": 4.71052122020335e-05, + "loss": 0.448, + "step": 28901 + }, + { + "epoch": 0.6775083950139771, + "grad_norm": 2.745112419128418, + "learning_rate": 4.709896102393202e-05, + "loss": 0.3682, + "step": 28902 + }, + { + "epoch": 0.677531836588782, + "grad_norm": 1.8382372856140137, + "learning_rate": 4.709271013287304e-05, + "loss": 1.1976, + "step": 28903 + }, + { + "epoch": 0.6775552781635871, + "grad_norm": 0.5292385220527649, + "learning_rate": 4.7086459528890535e-05, + "loss": 0.0295, + "step": 28904 + }, + { + "epoch": 0.677578719738392, + "grad_norm": 3.548793077468872, + "learning_rate": 4.70802092120184e-05, + "loss": 0.2024, + "step": 28905 + }, + { + "epoch": 0.6776021613131971, + "grad_norm": 1.0512962341308594, + "learning_rate": 4.707395918229048e-05, + "loss": 0.1066, + "step": 28906 + }, + { + "epoch": 0.677625602888002, + "grad_norm": 2.81074857711792, + "learning_rate": 4.70677094397408e-05, + "loss": 0.2517, + "step": 28907 + }, + { + "epoch": 0.677649044462807, + "grad_norm": 0.5547386407852173, + "learning_rate": 4.706145998440321e-05, + "loss": 0.0418, + "step": 28908 + }, + { + "epoch": 0.677672486037612, + "grad_norm": 3.357006549835205, + "learning_rate": 4.705521081631162e-05, + "loss": 0.3509, + "step": 28909 + }, + { + "epoch": 0.677695927612417, + "grad_norm": 2.805619716644287, + "learning_rate": 4.704896193549996e-05, + "loss": 0.2261, + "step": 28910 + }, + { + "epoch": 0.677719369187222, + "grad_norm": 2.7206737995147705, + "learning_rate": 4.704271334200211e-05, + "loss": 0.1151, + "step": 28911 + }, + { + "epoch": 0.677742810762027, + "grad_norm": 2.9900712966918945, + "learning_rate": 4.7036465035851964e-05, + "loss": 0.2672, + "step": 28912 + }, + { + "epoch": 0.677766252336832, + "grad_norm": 0.7540246844291687, + "learning_rate": 4.7030217017083464e-05, + "loss": 0.0565, + "step": 28913 + }, + { + "epoch": 0.677789693911637, + "grad_norm": 0.6120431423187256, + "learning_rate": 4.702396928573051e-05, + "loss": 0.0481, + "step": 28914 + }, + { + "epoch": 0.677813135486442, + "grad_norm": 2.232840061187744, + "learning_rate": 4.701772184182695e-05, + "loss": 0.1493, + "step": 28915 + }, + { + "epoch": 0.677836577061247, + "grad_norm": 1.0937741994857788, + "learning_rate": 4.701147468540675e-05, + "loss": 0.0903, + "step": 28916 + }, + { + "epoch": 0.6778600186360519, + "grad_norm": 2.77163028717041, + "learning_rate": 4.700522781650378e-05, + "loss": 0.1071, + "step": 28917 + }, + { + "epoch": 0.677883460210857, + "grad_norm": 1.4922147989273071, + "learning_rate": 4.699898123515194e-05, + "loss": 0.1245, + "step": 28918 + }, + { + "epoch": 0.6779069017856619, + "grad_norm": 2.196906805038452, + "learning_rate": 4.699273494138507e-05, + "loss": 0.2507, + "step": 28919 + }, + { + "epoch": 0.677930343360467, + "grad_norm": 3.117013454437256, + "learning_rate": 4.698648893523716e-05, + "loss": 0.4172, + "step": 28920 + }, + { + "epoch": 0.6779537849352719, + "grad_norm": 0.44819188117980957, + "learning_rate": 4.698024321674206e-05, + "loss": 0.021, + "step": 28921 + }, + { + "epoch": 0.677977226510077, + "grad_norm": 0.647228479385376, + "learning_rate": 4.6973997785933635e-05, + "loss": 0.042, + "step": 28922 + }, + { + "epoch": 0.6780006680848819, + "grad_norm": 3.4861502647399902, + "learning_rate": 4.6967752642845786e-05, + "loss": 0.4266, + "step": 28923 + }, + { + "epoch": 0.678024109659687, + "grad_norm": 5.916617393493652, + "learning_rate": 4.6961507787512416e-05, + "loss": 0.3357, + "step": 28924 + }, + { + "epoch": 0.6780475512344919, + "grad_norm": 2.230207681655884, + "learning_rate": 4.695526321996735e-05, + "loss": 0.385, + "step": 28925 + }, + { + "epoch": 0.678070992809297, + "grad_norm": 1.0509952306747437, + "learning_rate": 4.694901894024456e-05, + "loss": 0.1, + "step": 28926 + }, + { + "epoch": 0.6780944343841019, + "grad_norm": 1.3467446565628052, + "learning_rate": 4.694277494837784e-05, + "loss": 0.1174, + "step": 28927 + }, + { + "epoch": 0.6781178759589069, + "grad_norm": 3.8209004402160645, + "learning_rate": 4.6936531244401164e-05, + "loss": 0.3271, + "step": 28928 + }, + { + "epoch": 0.6781413175337119, + "grad_norm": 3.0912530422210693, + "learning_rate": 4.693028782834834e-05, + "loss": 0.2153, + "step": 28929 + }, + { + "epoch": 0.6781647591085169, + "grad_norm": 3.392056941986084, + "learning_rate": 4.692404470025329e-05, + "loss": 0.3129, + "step": 28930 + }, + { + "epoch": 0.6781882006833219, + "grad_norm": 0.2151823788881302, + "learning_rate": 4.691780186014986e-05, + "loss": 0.0052, + "step": 28931 + }, + { + "epoch": 0.6782116422581269, + "grad_norm": 3.225062370300293, + "learning_rate": 4.691155930807188e-05, + "loss": 0.4958, + "step": 28932 + }, + { + "epoch": 0.6782350838329319, + "grad_norm": 2.285301446914673, + "learning_rate": 4.690531704405332e-05, + "loss": 1.3251, + "step": 28933 + }, + { + "epoch": 0.6782585254077369, + "grad_norm": 0.8559070825576782, + "learning_rate": 4.6899075068127996e-05, + "loss": 0.0831, + "step": 28934 + }, + { + "epoch": 0.6782819669825418, + "grad_norm": 4.039603233337402, + "learning_rate": 4.689283338032979e-05, + "loss": 0.3644, + "step": 28935 + }, + { + "epoch": 0.6783054085573469, + "grad_norm": 3.13191819190979, + "learning_rate": 4.6886591980692566e-05, + "loss": 0.1604, + "step": 28936 + }, + { + "epoch": 0.6783288501321518, + "grad_norm": 3.433441638946533, + "learning_rate": 4.688035086925018e-05, + "loss": 0.2903, + "step": 28937 + }, + { + "epoch": 0.6783522917069569, + "grad_norm": 6.19972038269043, + "learning_rate": 4.687411004603647e-05, + "loss": 0.6256, + "step": 28938 + }, + { + "epoch": 0.6783757332817618, + "grad_norm": 3.0280725955963135, + "learning_rate": 4.686786951108536e-05, + "loss": 0.2124, + "step": 28939 + }, + { + "epoch": 0.6783991748565669, + "grad_norm": 0.401418000459671, + "learning_rate": 4.686162926443066e-05, + "loss": 0.0475, + "step": 28940 + }, + { + "epoch": 0.6784226164313718, + "grad_norm": 4.055541038513184, + "learning_rate": 4.685538930610629e-05, + "loss": 0.2663, + "step": 28941 + }, + { + "epoch": 0.6784460580061769, + "grad_norm": 3.647291421890259, + "learning_rate": 4.684914963614606e-05, + "loss": 0.3062, + "step": 28942 + }, + { + "epoch": 0.6784694995809819, + "grad_norm": 3.1520166397094727, + "learning_rate": 4.684291025458385e-05, + "loss": 0.3408, + "step": 28943 + }, + { + "epoch": 0.6784929411557868, + "grad_norm": 3.483059883117676, + "learning_rate": 4.683667116145347e-05, + "loss": 0.2211, + "step": 28944 + }, + { + "epoch": 0.6785163827305919, + "grad_norm": 2.497215986251831, + "learning_rate": 4.6830432356788837e-05, + "loss": 0.2104, + "step": 28945 + }, + { + "epoch": 0.6785398243053968, + "grad_norm": 0.8609344363212585, + "learning_rate": 4.682419384062377e-05, + "loss": 0.0836, + "step": 28946 + }, + { + "epoch": 0.6785632658802019, + "grad_norm": 3.914783000946045, + "learning_rate": 4.681795561299213e-05, + "loss": 0.3019, + "step": 28947 + }, + { + "epoch": 0.6785867074550068, + "grad_norm": 3.230207681655884, + "learning_rate": 4.6811717673927755e-05, + "loss": 0.2813, + "step": 28948 + }, + { + "epoch": 0.6786101490298119, + "grad_norm": 0.7601866722106934, + "learning_rate": 4.680548002346449e-05, + "loss": 0.0666, + "step": 28949 + }, + { + "epoch": 0.6786335906046168, + "grad_norm": 2.6681463718414307, + "learning_rate": 4.679924266163616e-05, + "loss": 0.2819, + "step": 28950 + }, + { + "epoch": 0.6786570321794219, + "grad_norm": 2.0211353302001953, + "learning_rate": 4.679300558847666e-05, + "loss": 0.1272, + "step": 28951 + }, + { + "epoch": 0.6786804737542268, + "grad_norm": 4.395824432373047, + "learning_rate": 4.678676880401978e-05, + "loss": 0.3165, + "step": 28952 + }, + { + "epoch": 0.6787039153290318, + "grad_norm": 3.6462607383728027, + "learning_rate": 4.678053230829942e-05, + "loss": 0.3464, + "step": 28953 + }, + { + "epoch": 0.6787273569038368, + "grad_norm": 3.278304100036621, + "learning_rate": 4.677429610134939e-05, + "loss": 1.2866, + "step": 28954 + }, + { + "epoch": 0.6787507984786418, + "grad_norm": 2.239392042160034, + "learning_rate": 4.676806018320353e-05, + "loss": 0.2246, + "step": 28955 + }, + { + "epoch": 0.6787742400534468, + "grad_norm": 2.9597091674804688, + "learning_rate": 4.676182455389567e-05, + "loss": 0.3601, + "step": 28956 + }, + { + "epoch": 0.6787976816282518, + "grad_norm": 0.8793622851371765, + "learning_rate": 4.675558921345961e-05, + "loss": 0.0936, + "step": 28957 + }, + { + "epoch": 0.6788211232030568, + "grad_norm": 2.8819661140441895, + "learning_rate": 4.674935416192925e-05, + "loss": 0.328, + "step": 28958 + }, + { + "epoch": 0.6788445647778618, + "grad_norm": 4.649065017700195, + "learning_rate": 4.6743119399338394e-05, + "loss": 0.1824, + "step": 28959 + }, + { + "epoch": 0.6788680063526668, + "grad_norm": 1.99522864818573, + "learning_rate": 4.673688492572088e-05, + "loss": 0.1238, + "step": 28960 + }, + { + "epoch": 0.6788914479274718, + "grad_norm": 1.6848453283309937, + "learning_rate": 4.6730650741110515e-05, + "loss": 0.1225, + "step": 28961 + }, + { + "epoch": 0.6789148895022767, + "grad_norm": 1.3019165992736816, + "learning_rate": 4.672441684554114e-05, + "loss": 0.1462, + "step": 28962 + }, + { + "epoch": 0.6789383310770818, + "grad_norm": 1.3532520532608032, + "learning_rate": 4.6718183239046545e-05, + "loss": 0.1964, + "step": 28963 + }, + { + "epoch": 0.6789617726518867, + "grad_norm": 3.805145263671875, + "learning_rate": 4.671194992166058e-05, + "loss": 0.3257, + "step": 28964 + }, + { + "epoch": 0.6789852142266918, + "grad_norm": 1.3079984188079834, + "learning_rate": 4.670571689341711e-05, + "loss": 0.0639, + "step": 28965 + }, + { + "epoch": 0.6790086558014967, + "grad_norm": 3.5436503887176514, + "learning_rate": 4.669948415434993e-05, + "loss": 0.424, + "step": 28966 + }, + { + "epoch": 0.6790320973763018, + "grad_norm": 0.6553099751472473, + "learning_rate": 4.6693251704492835e-05, + "loss": 0.0639, + "step": 28967 + }, + { + "epoch": 0.6790555389511067, + "grad_norm": 4.153056621551514, + "learning_rate": 4.668701954387965e-05, + "loss": 0.3011, + "step": 28968 + }, + { + "epoch": 0.6790789805259118, + "grad_norm": 2.960516929626465, + "learning_rate": 4.6680787672544214e-05, + "loss": 0.3199, + "step": 28969 + }, + { + "epoch": 0.6791024221007167, + "grad_norm": 4.06766414642334, + "learning_rate": 4.667455609052027e-05, + "loss": 0.3615, + "step": 28970 + }, + { + "epoch": 0.6791258636755217, + "grad_norm": 2.8462138175964355, + "learning_rate": 4.666832479784172e-05, + "loss": 0.1642, + "step": 28971 + }, + { + "epoch": 0.6791493052503267, + "grad_norm": 1.3678396940231323, + "learning_rate": 4.666209379454235e-05, + "loss": 0.1969, + "step": 28972 + }, + { + "epoch": 0.6791727468251317, + "grad_norm": 3.504483699798584, + "learning_rate": 4.665586308065596e-05, + "loss": 0.3269, + "step": 28973 + }, + { + "epoch": 0.6791961883999367, + "grad_norm": 3.6790153980255127, + "learning_rate": 4.6649632656216356e-05, + "loss": 0.2626, + "step": 28974 + }, + { + "epoch": 0.6792196299747417, + "grad_norm": 2.5835344791412354, + "learning_rate": 4.6643402521257306e-05, + "loss": 0.1784, + "step": 28975 + }, + { + "epoch": 0.6792430715495467, + "grad_norm": 0.9717334508895874, + "learning_rate": 4.663717267581269e-05, + "loss": 0.0925, + "step": 28976 + }, + { + "epoch": 0.6792665131243517, + "grad_norm": 3.6004672050476074, + "learning_rate": 4.6630943119916235e-05, + "loss": 0.2801, + "step": 28977 + }, + { + "epoch": 0.6792899546991567, + "grad_norm": 3.7532920837402344, + "learning_rate": 4.662471385360182e-05, + "loss": 0.3309, + "step": 28978 + }, + { + "epoch": 0.6793133962739617, + "grad_norm": 3.9504148960113525, + "learning_rate": 4.6618484876903216e-05, + "loss": 0.3075, + "step": 28979 + }, + { + "epoch": 0.6793368378487666, + "grad_norm": 2.2366719245910645, + "learning_rate": 4.661225618985421e-05, + "loss": 1.3004, + "step": 28980 + }, + { + "epoch": 0.6793602794235717, + "grad_norm": 0.9578567147254944, + "learning_rate": 4.660602779248861e-05, + "loss": 0.0694, + "step": 28981 + }, + { + "epoch": 0.6793837209983766, + "grad_norm": 3.427173376083374, + "learning_rate": 4.659979968484015e-05, + "loss": 0.3535, + "step": 28982 + }, + { + "epoch": 0.6794071625731817, + "grad_norm": 2.494589328765869, + "learning_rate": 4.659357186694273e-05, + "loss": 0.1594, + "step": 28983 + }, + { + "epoch": 0.6794306041479866, + "grad_norm": 4.957276821136475, + "learning_rate": 4.6587344338830086e-05, + "loss": 0.4363, + "step": 28984 + }, + { + "epoch": 0.6794540457227917, + "grad_norm": 2.2297894954681396, + "learning_rate": 4.658111710053601e-05, + "loss": 0.1897, + "step": 28985 + }, + { + "epoch": 0.6794774872975966, + "grad_norm": 3.5519726276397705, + "learning_rate": 4.65748901520943e-05, + "loss": 0.4506, + "step": 28986 + }, + { + "epoch": 0.6795009288724017, + "grad_norm": 0.8567319512367249, + "learning_rate": 4.6568663493538735e-05, + "loss": 0.0458, + "step": 28987 + }, + { + "epoch": 0.6795243704472066, + "grad_norm": 1.4647892713546753, + "learning_rate": 4.6562437124903066e-05, + "loss": 0.2026, + "step": 28988 + }, + { + "epoch": 0.6795478120220116, + "grad_norm": 3.2881360054016113, + "learning_rate": 4.655621104622112e-05, + "loss": 0.3526, + "step": 28989 + }, + { + "epoch": 0.6795712535968166, + "grad_norm": 2.2625434398651123, + "learning_rate": 4.654998525752672e-05, + "loss": 0.136, + "step": 28990 + }, + { + "epoch": 0.6795946951716216, + "grad_norm": 6.883440017700195, + "learning_rate": 4.6543759758853586e-05, + "loss": 0.1355, + "step": 28991 + }, + { + "epoch": 0.6796181367464266, + "grad_norm": 2.017235040664673, + "learning_rate": 4.653753455023553e-05, + "loss": 0.0862, + "step": 28992 + }, + { + "epoch": 0.6796415783212316, + "grad_norm": 1.4295213222503662, + "learning_rate": 4.653130963170631e-05, + "loss": 0.6511, + "step": 28993 + }, + { + "epoch": 0.6796650198960367, + "grad_norm": 1.2872135639190674, + "learning_rate": 4.652508500329969e-05, + "loss": 0.1443, + "step": 28994 + }, + { + "epoch": 0.6796884614708416, + "grad_norm": 1.9386727809906006, + "learning_rate": 4.6518860665049447e-05, + "loss": 1.103, + "step": 28995 + }, + { + "epoch": 0.6797119030456467, + "grad_norm": 2.217125415802002, + "learning_rate": 4.65126366169894e-05, + "loss": 1.0929, + "step": 28996 + }, + { + "epoch": 0.6797353446204516, + "grad_norm": 2.7355105876922607, + "learning_rate": 4.650641285915329e-05, + "loss": 0.2443, + "step": 28997 + }, + { + "epoch": 0.6797587861952566, + "grad_norm": 4.575040340423584, + "learning_rate": 4.650018939157489e-05, + "loss": 0.5072, + "step": 28998 + }, + { + "epoch": 0.6797822277700616, + "grad_norm": 1.078248381614685, + "learning_rate": 4.6493966214287964e-05, + "loss": 0.0876, + "step": 28999 + }, + { + "epoch": 0.6798056693448666, + "grad_norm": 3.5609803199768066, + "learning_rate": 4.648774332732625e-05, + "loss": 0.3474, + "step": 29000 + }, + { + "epoch": 0.6798291109196716, + "grad_norm": 0.3570948839187622, + "learning_rate": 4.648152073072358e-05, + "loss": 0.0261, + "step": 29001 + }, + { + "epoch": 0.6798525524944766, + "grad_norm": 1.5756595134735107, + "learning_rate": 4.647529842451365e-05, + "loss": 0.2006, + "step": 29002 + }, + { + "epoch": 0.6798759940692816, + "grad_norm": 0.7451487183570862, + "learning_rate": 4.646907640873028e-05, + "loss": 0.0641, + "step": 29003 + }, + { + "epoch": 0.6798994356440866, + "grad_norm": 0.7801825404167175, + "learning_rate": 4.646285468340722e-05, + "loss": 0.085, + "step": 29004 + }, + { + "epoch": 0.6799228772188916, + "grad_norm": 3.67724347114563, + "learning_rate": 4.6456633248578205e-05, + "loss": 0.2229, + "step": 29005 + }, + { + "epoch": 0.6799463187936966, + "grad_norm": 2.4276673793792725, + "learning_rate": 4.645041210427701e-05, + "loss": 0.2127, + "step": 29006 + }, + { + "epoch": 0.6799697603685015, + "grad_norm": 0.9387832283973694, + "learning_rate": 4.6444191250537384e-05, + "loss": 0.1173, + "step": 29007 + }, + { + "epoch": 0.6799932019433066, + "grad_norm": 2.9013214111328125, + "learning_rate": 4.643797068739304e-05, + "loss": 1.5484, + "step": 29008 + }, + { + "epoch": 0.6800166435181115, + "grad_norm": 3.2372236251831055, + "learning_rate": 4.6431750414877816e-05, + "loss": 0.189, + "step": 29009 + }, + { + "epoch": 0.6800400850929166, + "grad_norm": 3.1287660598754883, + "learning_rate": 4.642553043302542e-05, + "loss": 0.1547, + "step": 29010 + }, + { + "epoch": 0.6800635266677215, + "grad_norm": 2.585017681121826, + "learning_rate": 4.6419310741869604e-05, + "loss": 0.1287, + "step": 29011 + }, + { + "epoch": 0.6800869682425266, + "grad_norm": 1.3020845651626587, + "learning_rate": 4.6413091341444104e-05, + "loss": 0.0593, + "step": 29012 + }, + { + "epoch": 0.6801104098173315, + "grad_norm": 4.749641418457031, + "learning_rate": 4.640687223178265e-05, + "loss": 0.3568, + "step": 29013 + }, + { + "epoch": 0.6801338513921366, + "grad_norm": 0.8957303762435913, + "learning_rate": 4.640065341291905e-05, + "loss": 0.122, + "step": 29014 + }, + { + "epoch": 0.6801572929669415, + "grad_norm": 2.4444386959075928, + "learning_rate": 4.6394434884886976e-05, + "loss": 0.1231, + "step": 29015 + }, + { + "epoch": 0.6801807345417465, + "grad_norm": 1.4196304082870483, + "learning_rate": 4.638821664772024e-05, + "loss": 0.0923, + "step": 29016 + }, + { + "epoch": 0.6802041761165515, + "grad_norm": 0.6132407784461975, + "learning_rate": 4.6381998701452546e-05, + "loss": 0.0885, + "step": 29017 + }, + { + "epoch": 0.6802276176913565, + "grad_norm": 1.0834842920303345, + "learning_rate": 4.6375781046117636e-05, + "loss": 0.0476, + "step": 29018 + }, + { + "epoch": 0.6802510592661615, + "grad_norm": 2.04244327545166, + "learning_rate": 4.636956368174925e-05, + "loss": 1.0328, + "step": 29019 + }, + { + "epoch": 0.6802745008409665, + "grad_norm": 0.36965519189834595, + "learning_rate": 4.636334660838108e-05, + "loss": 0.0309, + "step": 29020 + }, + { + "epoch": 0.6802979424157715, + "grad_norm": 2.4598493576049805, + "learning_rate": 4.635712982604693e-05, + "loss": 0.2515, + "step": 29021 + }, + { + "epoch": 0.6803213839905765, + "grad_norm": 1.2694700956344604, + "learning_rate": 4.635091333478051e-05, + "loss": 0.1402, + "step": 29022 + }, + { + "epoch": 0.6803448255653815, + "grad_norm": 1.8881657123565674, + "learning_rate": 4.634469713461553e-05, + "loss": 0.1791, + "step": 29023 + }, + { + "epoch": 0.6803682671401865, + "grad_norm": 2.8857595920562744, + "learning_rate": 4.633848122558575e-05, + "loss": 0.1914, + "step": 29024 + }, + { + "epoch": 0.6803917087149914, + "grad_norm": 3.6331913471221924, + "learning_rate": 4.6332265607724836e-05, + "loss": 0.245, + "step": 29025 + }, + { + "epoch": 0.6804151502897965, + "grad_norm": 0.9697263836860657, + "learning_rate": 4.632605028106659e-05, + "loss": 0.1756, + "step": 29026 + }, + { + "epoch": 0.6804385918646014, + "grad_norm": 1.6226463317871094, + "learning_rate": 4.631983524564467e-05, + "loss": 0.2237, + "step": 29027 + }, + { + "epoch": 0.6804620334394065, + "grad_norm": 0.8511794209480286, + "learning_rate": 4.6313620501492875e-05, + "loss": 0.0576, + "step": 29028 + }, + { + "epoch": 0.6804854750142114, + "grad_norm": 4.121524333953857, + "learning_rate": 4.630740604864487e-05, + "loss": 0.2357, + "step": 29029 + }, + { + "epoch": 0.6805089165890165, + "grad_norm": 2.989333152770996, + "learning_rate": 4.630119188713441e-05, + "loss": 0.2197, + "step": 29030 + }, + { + "epoch": 0.6805323581638214, + "grad_norm": 2.3655526638031006, + "learning_rate": 4.6294978016995184e-05, + "loss": 0.1413, + "step": 29031 + }, + { + "epoch": 0.6805557997386265, + "grad_norm": 0.5218403339385986, + "learning_rate": 4.628876443826091e-05, + "loss": 0.036, + "step": 29032 + }, + { + "epoch": 0.6805792413134314, + "grad_norm": 4.535827159881592, + "learning_rate": 4.628255115096528e-05, + "loss": 0.205, + "step": 29033 + }, + { + "epoch": 0.6806026828882364, + "grad_norm": 0.7026483416557312, + "learning_rate": 4.6276338155142074e-05, + "loss": 0.0503, + "step": 29034 + }, + { + "epoch": 0.6806261244630414, + "grad_norm": 3.1238059997558594, + "learning_rate": 4.6270125450824976e-05, + "loss": 0.3308, + "step": 29035 + }, + { + "epoch": 0.6806495660378464, + "grad_norm": 2.5167646408081055, + "learning_rate": 4.626391303804768e-05, + "loss": 0.3036, + "step": 29036 + }, + { + "epoch": 0.6806730076126514, + "grad_norm": 1.543581485748291, + "learning_rate": 4.625770091684386e-05, + "loss": 0.1045, + "step": 29037 + }, + { + "epoch": 0.6806964491874564, + "grad_norm": 2.6043200492858887, + "learning_rate": 4.625148908724731e-05, + "loss": 0.1373, + "step": 29038 + }, + { + "epoch": 0.6807198907622614, + "grad_norm": 3.2371344566345215, + "learning_rate": 4.624527754929169e-05, + "loss": 0.2711, + "step": 29039 + }, + { + "epoch": 0.6807433323370664, + "grad_norm": 5.285795211791992, + "learning_rate": 4.623906630301067e-05, + "loss": 0.6815, + "step": 29040 + }, + { + "epoch": 0.6807667739118713, + "grad_norm": 2.313885450363159, + "learning_rate": 4.623285534843802e-05, + "loss": 0.15, + "step": 29041 + }, + { + "epoch": 0.6807902154866764, + "grad_norm": 3.785093307495117, + "learning_rate": 4.622664468560741e-05, + "loss": 0.5263, + "step": 29042 + }, + { + "epoch": 0.6808136570614813, + "grad_norm": 2.1129496097564697, + "learning_rate": 4.622043431455254e-05, + "loss": 0.1959, + "step": 29043 + }, + { + "epoch": 0.6808370986362864, + "grad_norm": 2.7554707527160645, + "learning_rate": 4.6214224235307116e-05, + "loss": 0.1843, + "step": 29044 + }, + { + "epoch": 0.6808605402110914, + "grad_norm": 0.8810194134712219, + "learning_rate": 4.6208014447904816e-05, + "loss": 0.0889, + "step": 29045 + }, + { + "epoch": 0.6808839817858964, + "grad_norm": 4.608108997344971, + "learning_rate": 4.620180495237929e-05, + "loss": 0.3045, + "step": 29046 + }, + { + "epoch": 0.6809074233607014, + "grad_norm": 3.95080304145813, + "learning_rate": 4.619559574876434e-05, + "loss": 0.4076, + "step": 29047 + }, + { + "epoch": 0.6809308649355064, + "grad_norm": 4.419467449188232, + "learning_rate": 4.6189386837093595e-05, + "loss": 0.2316, + "step": 29048 + }, + { + "epoch": 0.6809543065103114, + "grad_norm": 2.1012964248657227, + "learning_rate": 4.6183178217400705e-05, + "loss": 0.0971, + "step": 29049 + }, + { + "epoch": 0.6809777480851164, + "grad_norm": 2.5048625469207764, + "learning_rate": 4.617696988971945e-05, + "loss": 1.8598, + "step": 29050 + }, + { + "epoch": 0.6810011896599214, + "grad_norm": 2.755767583847046, + "learning_rate": 4.617076185408347e-05, + "loss": 1.6926, + "step": 29051 + }, + { + "epoch": 0.6810246312347263, + "grad_norm": 3.1983118057250977, + "learning_rate": 4.616455411052644e-05, + "loss": 1.5762, + "step": 29052 + }, + { + "epoch": 0.6810480728095314, + "grad_norm": 0.9577866792678833, + "learning_rate": 4.615834665908202e-05, + "loss": 0.0721, + "step": 29053 + }, + { + "epoch": 0.6810715143843363, + "grad_norm": 2.8294997215270996, + "learning_rate": 4.6152139499783965e-05, + "loss": 0.3367, + "step": 29054 + }, + { + "epoch": 0.6810949559591414, + "grad_norm": 2.559674024581909, + "learning_rate": 4.614593263266591e-05, + "loss": 0.315, + "step": 29055 + }, + { + "epoch": 0.6811183975339463, + "grad_norm": 1.779158592224121, + "learning_rate": 4.613972605776155e-05, + "loss": 0.1173, + "step": 29056 + }, + { + "epoch": 0.6811418391087514, + "grad_norm": 2.3591208457946777, + "learning_rate": 4.6133519775104536e-05, + "loss": 0.2397, + "step": 29057 + }, + { + "epoch": 0.6811652806835563, + "grad_norm": 2.3767335414886475, + "learning_rate": 4.6127313784728534e-05, + "loss": 0.2142, + "step": 29058 + }, + { + "epoch": 0.6811887222583614, + "grad_norm": 2.7452268600463867, + "learning_rate": 4.6121108086667276e-05, + "loss": 0.3348, + "step": 29059 + }, + { + "epoch": 0.6812121638331663, + "grad_norm": 2.399583101272583, + "learning_rate": 4.61149026809544e-05, + "loss": 0.1581, + "step": 29060 + }, + { + "epoch": 0.6812356054079713, + "grad_norm": 3.380436658859253, + "learning_rate": 4.610869756762354e-05, + "loss": 1.9852, + "step": 29061 + }, + { + "epoch": 0.6812590469827763, + "grad_norm": 2.6435866355895996, + "learning_rate": 4.610249274670844e-05, + "loss": 0.1986, + "step": 29062 + }, + { + "epoch": 0.6812824885575813, + "grad_norm": 2.324737310409546, + "learning_rate": 4.609628821824272e-05, + "loss": 0.1538, + "step": 29063 + }, + { + "epoch": 0.6813059301323863, + "grad_norm": 3.715409755706787, + "learning_rate": 4.609008398226006e-05, + "loss": 0.4208, + "step": 29064 + }, + { + "epoch": 0.6813293717071913, + "grad_norm": 2.597902297973633, + "learning_rate": 4.608388003879409e-05, + "loss": 0.1444, + "step": 29065 + }, + { + "epoch": 0.6813528132819963, + "grad_norm": 3.6361441612243652, + "learning_rate": 4.6077676387878533e-05, + "loss": 0.3908, + "step": 29066 + }, + { + "epoch": 0.6813762548568013, + "grad_norm": 2.458228349685669, + "learning_rate": 4.607147302954703e-05, + "loss": 0.1276, + "step": 29067 + }, + { + "epoch": 0.6813996964316062, + "grad_norm": 0.30161985754966736, + "learning_rate": 4.6065269963833223e-05, + "loss": 0.0143, + "step": 29068 + }, + { + "epoch": 0.6814231380064113, + "grad_norm": 2.227670192718506, + "learning_rate": 4.605906719077078e-05, + "loss": 0.1713, + "step": 29069 + }, + { + "epoch": 0.6814465795812162, + "grad_norm": 3.032606840133667, + "learning_rate": 4.605286471039335e-05, + "loss": 0.2585, + "step": 29070 + }, + { + "epoch": 0.6814700211560213, + "grad_norm": 2.5764482021331787, + "learning_rate": 4.604666252273456e-05, + "loss": 0.3898, + "step": 29071 + }, + { + "epoch": 0.6814934627308262, + "grad_norm": 3.1706790924072266, + "learning_rate": 4.604046062782813e-05, + "loss": 0.2576, + "step": 29072 + }, + { + "epoch": 0.6815169043056313, + "grad_norm": 1.3689521551132202, + "learning_rate": 4.603425902570765e-05, + "loss": 0.1098, + "step": 29073 + }, + { + "epoch": 0.6815403458804362, + "grad_norm": 2.554783344268799, + "learning_rate": 4.602805771640682e-05, + "loss": 0.1802, + "step": 29074 + }, + { + "epoch": 0.6815637874552413, + "grad_norm": 0.8246085047721863, + "learning_rate": 4.6021856699959276e-05, + "loss": 0.0499, + "step": 29075 + }, + { + "epoch": 0.6815872290300462, + "grad_norm": 2.1207895278930664, + "learning_rate": 4.601565597639867e-05, + "loss": 0.0871, + "step": 29076 + }, + { + "epoch": 0.6816106706048513, + "grad_norm": 1.008449912071228, + "learning_rate": 4.6009455545758606e-05, + "loss": 0.1033, + "step": 29077 + }, + { + "epoch": 0.6816341121796562, + "grad_norm": 1.6719377040863037, + "learning_rate": 4.600325540807273e-05, + "loss": 0.1155, + "step": 29078 + }, + { + "epoch": 0.6816575537544612, + "grad_norm": 1.583816647529602, + "learning_rate": 4.599705556337475e-05, + "loss": 0.1368, + "step": 29079 + }, + { + "epoch": 0.6816809953292662, + "grad_norm": 3.9175424575805664, + "learning_rate": 4.599085601169825e-05, + "loss": 0.4307, + "step": 29080 + }, + { + "epoch": 0.6817044369040712, + "grad_norm": 3.271998882293701, + "learning_rate": 4.5984656753076895e-05, + "loss": 0.1588, + "step": 29081 + }, + { + "epoch": 0.6817278784788762, + "grad_norm": 0.5290845632553101, + "learning_rate": 4.59784577875443e-05, + "loss": 0.0701, + "step": 29082 + }, + { + "epoch": 0.6817513200536812, + "grad_norm": 4.134457111358643, + "learning_rate": 4.5972259115134134e-05, + "loss": 0.3028, + "step": 29083 + }, + { + "epoch": 0.6817747616284862, + "grad_norm": 0.807530403137207, + "learning_rate": 4.5966060735879955e-05, + "loss": 0.0705, + "step": 29084 + }, + { + "epoch": 0.6817982032032912, + "grad_norm": 3.0563101768493652, + "learning_rate": 4.595986264981546e-05, + "loss": 0.3056, + "step": 29085 + }, + { + "epoch": 0.6818216447780961, + "grad_norm": 6.062608242034912, + "learning_rate": 4.595366485697429e-05, + "loss": 0.763, + "step": 29086 + }, + { + "epoch": 0.6818450863529012, + "grad_norm": 1.0500816106796265, + "learning_rate": 4.5947467357390065e-05, + "loss": 0.0723, + "step": 29087 + }, + { + "epoch": 0.6818685279277061, + "grad_norm": 0.8617952466011047, + "learning_rate": 4.5941270151096405e-05, + "loss": 0.0878, + "step": 29088 + }, + { + "epoch": 0.6818919695025112, + "grad_norm": 3.117804527282715, + "learning_rate": 4.593507323812693e-05, + "loss": 0.1803, + "step": 29089 + }, + { + "epoch": 0.6819154110773161, + "grad_norm": 5.6167073249816895, + "learning_rate": 4.592887661851527e-05, + "loss": 0.3755, + "step": 29090 + }, + { + "epoch": 0.6819388526521212, + "grad_norm": 0.5077204704284668, + "learning_rate": 4.592268029229501e-05, + "loss": 0.0251, + "step": 29091 + }, + { + "epoch": 0.6819622942269261, + "grad_norm": 1.4400715827941895, + "learning_rate": 4.5916484259499835e-05, + "loss": 0.1611, + "step": 29092 + }, + { + "epoch": 0.6819857358017312, + "grad_norm": 2.6701791286468506, + "learning_rate": 4.591028852016333e-05, + "loss": 0.1992, + "step": 29093 + }, + { + "epoch": 0.6820091773765361, + "grad_norm": 3.3992040157318115, + "learning_rate": 4.5904093074319146e-05, + "loss": 0.2181, + "step": 29094 + }, + { + "epoch": 0.6820326189513412, + "grad_norm": 2.493603229522705, + "learning_rate": 4.589789792200086e-05, + "loss": 0.2285, + "step": 29095 + }, + { + "epoch": 0.6820560605261462, + "grad_norm": 2.805264472961426, + "learning_rate": 4.589170306324206e-05, + "loss": 0.2038, + "step": 29096 + }, + { + "epoch": 0.6820795021009511, + "grad_norm": 2.11079740524292, + "learning_rate": 4.588550849807645e-05, + "loss": 0.172, + "step": 29097 + }, + { + "epoch": 0.6821029436757562, + "grad_norm": 3.3802390098571777, + "learning_rate": 4.587931422653754e-05, + "loss": 0.2173, + "step": 29098 + }, + { + "epoch": 0.6821263852505611, + "grad_norm": 10.203042984008789, + "learning_rate": 4.587312024865904e-05, + "loss": 0.5176, + "step": 29099 + }, + { + "epoch": 0.6821498268253662, + "grad_norm": 0.9230262637138367, + "learning_rate": 4.586692656447451e-05, + "loss": 0.1002, + "step": 29100 + }, + { + "epoch": 0.6821732684001711, + "grad_norm": 3.6444883346557617, + "learning_rate": 4.5860733174017555e-05, + "loss": 0.1861, + "step": 29101 + }, + { + "epoch": 0.6821967099749762, + "grad_norm": 2.0333943367004395, + "learning_rate": 4.58545400773218e-05, + "loss": 0.1723, + "step": 29102 + }, + { + "epoch": 0.6822201515497811, + "grad_norm": 2.044466018676758, + "learning_rate": 4.584834727442078e-05, + "loss": 0.237, + "step": 29103 + }, + { + "epoch": 0.6822435931245862, + "grad_norm": 4.790694713592529, + "learning_rate": 4.584215476534819e-05, + "loss": 0.441, + "step": 29104 + }, + { + "epoch": 0.6822670346993911, + "grad_norm": 2.087836980819702, + "learning_rate": 4.583596255013759e-05, + "loss": 0.1472, + "step": 29105 + }, + { + "epoch": 0.6822904762741961, + "grad_norm": 3.5433120727539062, + "learning_rate": 4.582977062882258e-05, + "loss": 0.2861, + "step": 29106 + }, + { + "epoch": 0.6823139178490011, + "grad_norm": 3.523463487625122, + "learning_rate": 4.582357900143675e-05, + "loss": 0.5845, + "step": 29107 + }, + { + "epoch": 0.6823373594238061, + "grad_norm": 2.946335792541504, + "learning_rate": 4.5817387668013724e-05, + "loss": 0.219, + "step": 29108 + }, + { + "epoch": 0.6823608009986111, + "grad_norm": 0.8312664031982422, + "learning_rate": 4.581119662858703e-05, + "loss": 0.1004, + "step": 29109 + }, + { + "epoch": 0.6823842425734161, + "grad_norm": 2.9187827110290527, + "learning_rate": 4.5805005883190297e-05, + "loss": 0.17, + "step": 29110 + }, + { + "epoch": 0.6824076841482211, + "grad_norm": 1.1074479818344116, + "learning_rate": 4.579881543185717e-05, + "loss": 0.1058, + "step": 29111 + }, + { + "epoch": 0.6824311257230261, + "grad_norm": 0.4850355088710785, + "learning_rate": 4.5792625274621195e-05, + "loss": 0.0199, + "step": 29112 + }, + { + "epoch": 0.682454567297831, + "grad_norm": 2.6474101543426514, + "learning_rate": 4.578643541151596e-05, + "loss": 0.2302, + "step": 29113 + }, + { + "epoch": 0.6824780088726361, + "grad_norm": 1.083579182624817, + "learning_rate": 4.5780245842575056e-05, + "loss": 0.1185, + "step": 29114 + }, + { + "epoch": 0.682501450447441, + "grad_norm": 3.3005905151367188, + "learning_rate": 4.577405656783205e-05, + "loss": 0.2596, + "step": 29115 + }, + { + "epoch": 0.6825248920222461, + "grad_norm": 3.321733236312866, + "learning_rate": 4.576786758732051e-05, + "loss": 0.3053, + "step": 29116 + }, + { + "epoch": 0.682548333597051, + "grad_norm": 1.2875456809997559, + "learning_rate": 4.576167890107408e-05, + "loss": 0.0786, + "step": 29117 + }, + { + "epoch": 0.6825717751718561, + "grad_norm": 4.82401180267334, + "learning_rate": 4.575549050912631e-05, + "loss": 0.352, + "step": 29118 + }, + { + "epoch": 0.682595216746661, + "grad_norm": 0.707899272441864, + "learning_rate": 4.574930241151076e-05, + "loss": 0.0516, + "step": 29119 + }, + { + "epoch": 0.6826186583214661, + "grad_norm": 1.100647211074829, + "learning_rate": 4.574311460826103e-05, + "loss": 0.0854, + "step": 29120 + }, + { + "epoch": 0.682642099896271, + "grad_norm": 2.8953285217285156, + "learning_rate": 4.5736927099410687e-05, + "loss": 0.1651, + "step": 29121 + }, + { + "epoch": 0.682665541471076, + "grad_norm": 0.6595351099967957, + "learning_rate": 4.573073988499327e-05, + "loss": 0.053, + "step": 29122 + }, + { + "epoch": 0.682688983045881, + "grad_norm": 2.9548301696777344, + "learning_rate": 4.572455296504238e-05, + "loss": 0.2118, + "step": 29123 + }, + { + "epoch": 0.682712424620686, + "grad_norm": 3.919508934020996, + "learning_rate": 4.571836633959162e-05, + "loss": 0.3738, + "step": 29124 + }, + { + "epoch": 0.682735866195491, + "grad_norm": 2.820474863052368, + "learning_rate": 4.571218000867453e-05, + "loss": 0.3294, + "step": 29125 + }, + { + "epoch": 0.682759307770296, + "grad_norm": 2.4254984855651855, + "learning_rate": 4.570599397232469e-05, + "loss": 0.1766, + "step": 29126 + }, + { + "epoch": 0.682782749345101, + "grad_norm": 3.118861675262451, + "learning_rate": 4.5699808230575644e-05, + "loss": 0.2636, + "step": 29127 + }, + { + "epoch": 0.682806190919906, + "grad_norm": 3.0376830101013184, + "learning_rate": 4.5693622783460976e-05, + "loss": 0.332, + "step": 29128 + }, + { + "epoch": 0.682829632494711, + "grad_norm": 1.7772481441497803, + "learning_rate": 4.5687437631014196e-05, + "loss": 0.2086, + "step": 29129 + }, + { + "epoch": 0.682853074069516, + "grad_norm": 3.3072187900543213, + "learning_rate": 4.568125277326895e-05, + "loss": 0.2765, + "step": 29130 + }, + { + "epoch": 0.682876515644321, + "grad_norm": 1.474686861038208, + "learning_rate": 4.567506821025874e-05, + "loss": 0.1413, + "step": 29131 + }, + { + "epoch": 0.682899957219126, + "grad_norm": 1.93781578540802, + "learning_rate": 4.566888394201715e-05, + "loss": 0.1068, + "step": 29132 + }, + { + "epoch": 0.6829233987939309, + "grad_norm": 3.4086525440216064, + "learning_rate": 4.566269996857771e-05, + "loss": 0.2907, + "step": 29133 + }, + { + "epoch": 0.682946840368736, + "grad_norm": 1.3905878067016602, + "learning_rate": 4.565651628997396e-05, + "loss": 0.0962, + "step": 29134 + }, + { + "epoch": 0.6829702819435409, + "grad_norm": 4.952513217926025, + "learning_rate": 4.565033290623951e-05, + "loss": 0.3913, + "step": 29135 + }, + { + "epoch": 0.682993723518346, + "grad_norm": 1.1444401741027832, + "learning_rate": 4.5644149817407854e-05, + "loss": 0.0907, + "step": 29136 + }, + { + "epoch": 0.6830171650931509, + "grad_norm": 2.622936964035034, + "learning_rate": 4.563796702351261e-05, + "loss": 0.2576, + "step": 29137 + }, + { + "epoch": 0.683040606667956, + "grad_norm": 1.9653828144073486, + "learning_rate": 4.563178452458727e-05, + "loss": 0.1165, + "step": 29138 + }, + { + "epoch": 0.6830640482427609, + "grad_norm": 0.8755510449409485, + "learning_rate": 4.562560232066541e-05, + "loss": 0.0943, + "step": 29139 + }, + { + "epoch": 0.683087489817566, + "grad_norm": 1.138553261756897, + "learning_rate": 4.5619420411780555e-05, + "loss": 0.1209, + "step": 29140 + }, + { + "epoch": 0.6831109313923709, + "grad_norm": 2.5580220222473145, + "learning_rate": 4.561323879796621e-05, + "loss": 1.5679, + "step": 29141 + }, + { + "epoch": 0.6831343729671759, + "grad_norm": 2.700444459915161, + "learning_rate": 4.560705747925601e-05, + "loss": 0.2465, + "step": 29142 + }, + { + "epoch": 0.6831578145419809, + "grad_norm": 3.7611782550811768, + "learning_rate": 4.560087645568344e-05, + "loss": 0.5642, + "step": 29143 + }, + { + "epoch": 0.6831812561167859, + "grad_norm": 3.1593382358551025, + "learning_rate": 4.5594695727282043e-05, + "loss": 0.2708, + "step": 29144 + }, + { + "epoch": 0.6832046976915909, + "grad_norm": 1.8263297080993652, + "learning_rate": 4.5588515294085353e-05, + "loss": 0.1042, + "step": 29145 + }, + { + "epoch": 0.6832281392663959, + "grad_norm": 4.0864577293396, + "learning_rate": 4.5582335156126876e-05, + "loss": 0.3871, + "step": 29146 + }, + { + "epoch": 0.683251580841201, + "grad_norm": 0.9149084687232971, + "learning_rate": 4.5576155313440215e-05, + "loss": 0.072, + "step": 29147 + }, + { + "epoch": 0.6832750224160059, + "grad_norm": 2.6179051399230957, + "learning_rate": 4.5569975766058824e-05, + "loss": 0.1877, + "step": 29148 + }, + { + "epoch": 0.683298463990811, + "grad_norm": 4.8548583984375, + "learning_rate": 4.556379651401632e-05, + "loss": 0.3567, + "step": 29149 + }, + { + "epoch": 0.6833219055656159, + "grad_norm": 3.7563223838806152, + "learning_rate": 4.555761755734619e-05, + "loss": 0.338, + "step": 29150 + }, + { + "epoch": 0.683345347140421, + "grad_norm": 0.8179722428321838, + "learning_rate": 4.555143889608195e-05, + "loss": 0.0748, + "step": 29151 + }, + { + "epoch": 0.6833687887152259, + "grad_norm": 0.4257781207561493, + "learning_rate": 4.5545260530257136e-05, + "loss": 0.0295, + "step": 29152 + }, + { + "epoch": 0.6833922302900309, + "grad_norm": 3.1242318153381348, + "learning_rate": 4.553908245990527e-05, + "loss": 1.5164, + "step": 29153 + }, + { + "epoch": 0.6834156718648359, + "grad_norm": 3.7015397548675537, + "learning_rate": 4.553290468505983e-05, + "loss": 0.3912, + "step": 29154 + }, + { + "epoch": 0.6834391134396409, + "grad_norm": 2.7212891578674316, + "learning_rate": 4.5526727205754426e-05, + "loss": 0.1951, + "step": 29155 + }, + { + "epoch": 0.6834625550144459, + "grad_norm": 1.8616585731506348, + "learning_rate": 4.5520550022022534e-05, + "loss": 1.5543, + "step": 29156 + }, + { + "epoch": 0.6834859965892509, + "grad_norm": 2.8669071197509766, + "learning_rate": 4.551437313389766e-05, + "loss": 0.3196, + "step": 29157 + }, + { + "epoch": 0.6835094381640558, + "grad_norm": 1.5070773363113403, + "learning_rate": 4.55081965414133e-05, + "loss": 0.1623, + "step": 29158 + }, + { + "epoch": 0.6835328797388609, + "grad_norm": 1.977891206741333, + "learning_rate": 4.550202024460303e-05, + "loss": 0.0656, + "step": 29159 + }, + { + "epoch": 0.6835563213136658, + "grad_norm": 0.704846203327179, + "learning_rate": 4.549584424350033e-05, + "loss": 0.0564, + "step": 29160 + }, + { + "epoch": 0.6835797628884709, + "grad_norm": 2.0354502201080322, + "learning_rate": 4.548966853813867e-05, + "loss": 0.2547, + "step": 29161 + }, + { + "epoch": 0.6836032044632758, + "grad_norm": 2.0092804431915283, + "learning_rate": 4.5483493128551656e-05, + "loss": 0.1397, + "step": 29162 + }, + { + "epoch": 0.6836266460380809, + "grad_norm": 4.196861267089844, + "learning_rate": 4.547731801477273e-05, + "loss": 0.2427, + "step": 29163 + }, + { + "epoch": 0.6836500876128858, + "grad_norm": 3.692566156387329, + "learning_rate": 4.547114319683541e-05, + "loss": 0.4985, + "step": 29164 + }, + { + "epoch": 0.6836735291876909, + "grad_norm": 1.3555989265441895, + "learning_rate": 4.5464968674773204e-05, + "loss": 0.1509, + "step": 29165 + }, + { + "epoch": 0.6836969707624958, + "grad_norm": 0.983601987361908, + "learning_rate": 4.54587944486196e-05, + "loss": 0.1083, + "step": 29166 + }, + { + "epoch": 0.6837204123373009, + "grad_norm": 3.2770586013793945, + "learning_rate": 4.5452620518408084e-05, + "loss": 0.198, + "step": 29167 + }, + { + "epoch": 0.6837438539121058, + "grad_norm": 3.001183032989502, + "learning_rate": 4.544644688417221e-05, + "loss": 0.3294, + "step": 29168 + }, + { + "epoch": 0.6837672954869108, + "grad_norm": 3.5849292278289795, + "learning_rate": 4.544027354594547e-05, + "loss": 0.312, + "step": 29169 + }, + { + "epoch": 0.6837907370617158, + "grad_norm": 2.6376607418060303, + "learning_rate": 4.5434100503761325e-05, + "loss": 0.1856, + "step": 29170 + }, + { + "epoch": 0.6838141786365208, + "grad_norm": 1.893707036972046, + "learning_rate": 4.5427927757653244e-05, + "loss": 0.1297, + "step": 29171 + }, + { + "epoch": 0.6838376202113258, + "grad_norm": 1.9543302059173584, + "learning_rate": 4.5421755307654804e-05, + "loss": 1.4878, + "step": 29172 + }, + { + "epoch": 0.6838610617861308, + "grad_norm": 0.9716572761535645, + "learning_rate": 4.5415583153799444e-05, + "loss": 0.0776, + "step": 29173 + }, + { + "epoch": 0.6838845033609358, + "grad_norm": 2.5443625450134277, + "learning_rate": 4.540941129612064e-05, + "loss": 0.2995, + "step": 29174 + }, + { + "epoch": 0.6839079449357408, + "grad_norm": 1.6928558349609375, + "learning_rate": 4.540323973465195e-05, + "loss": 0.1501, + "step": 29175 + }, + { + "epoch": 0.6839313865105457, + "grad_norm": 2.5832910537719727, + "learning_rate": 4.5397068469426796e-05, + "loss": 1.4501, + "step": 29176 + }, + { + "epoch": 0.6839548280853508, + "grad_norm": 1.8946270942687988, + "learning_rate": 4.5390897500478705e-05, + "loss": 1.6419, + "step": 29177 + }, + { + "epoch": 0.6839782696601557, + "grad_norm": 3.0112950801849365, + "learning_rate": 4.5384726827841116e-05, + "loss": 0.2656, + "step": 29178 + }, + { + "epoch": 0.6840017112349608, + "grad_norm": 4.085930824279785, + "learning_rate": 4.537855645154752e-05, + "loss": 0.4717, + "step": 29179 + }, + { + "epoch": 0.6840251528097657, + "grad_norm": 4.817037582397461, + "learning_rate": 4.537238637163144e-05, + "loss": 0.2773, + "step": 29180 + }, + { + "epoch": 0.6840485943845708, + "grad_norm": 0.35857874155044556, + "learning_rate": 4.536621658812632e-05, + "loss": 0.0218, + "step": 29181 + }, + { + "epoch": 0.6840720359593757, + "grad_norm": 4.052372455596924, + "learning_rate": 4.536004710106565e-05, + "loss": 0.2509, + "step": 29182 + }, + { + "epoch": 0.6840954775341808, + "grad_norm": 2.7218246459960938, + "learning_rate": 4.535387791048287e-05, + "loss": 0.2279, + "step": 29183 + }, + { + "epoch": 0.6841189191089857, + "grad_norm": 3.8234667778015137, + "learning_rate": 4.534770901641152e-05, + "loss": 0.5485, + "step": 29184 + }, + { + "epoch": 0.6841423606837908, + "grad_norm": 2.6719863414764404, + "learning_rate": 4.534154041888503e-05, + "loss": 0.1781, + "step": 29185 + }, + { + "epoch": 0.6841658022585957, + "grad_norm": 3.082099437713623, + "learning_rate": 4.5335372117936846e-05, + "loss": 0.1728, + "step": 29186 + }, + { + "epoch": 0.6841892438334007, + "grad_norm": 0.8937799334526062, + "learning_rate": 4.532920411360051e-05, + "loss": 0.1072, + "step": 29187 + }, + { + "epoch": 0.6842126854082057, + "grad_norm": 3.3290998935699463, + "learning_rate": 4.532303640590945e-05, + "loss": 0.2509, + "step": 29188 + }, + { + "epoch": 0.6842361269830107, + "grad_norm": 2.7624902725219727, + "learning_rate": 4.531686899489713e-05, + "loss": 0.2075, + "step": 29189 + }, + { + "epoch": 0.6842595685578157, + "grad_norm": 2.692044496536255, + "learning_rate": 4.5310701880597017e-05, + "loss": 1.731, + "step": 29190 + }, + { + "epoch": 0.6842830101326207, + "grad_norm": 1.956889271736145, + "learning_rate": 4.530453506304258e-05, + "loss": 0.076, + "step": 29191 + }, + { + "epoch": 0.6843064517074257, + "grad_norm": 2.571166753768921, + "learning_rate": 4.529836854226722e-05, + "loss": 0.2598, + "step": 29192 + }, + { + "epoch": 0.6843298932822307, + "grad_norm": 2.536557912826538, + "learning_rate": 4.52922023183045e-05, + "loss": 0.1931, + "step": 29193 + }, + { + "epoch": 0.6843533348570356, + "grad_norm": 1.8031375408172607, + "learning_rate": 4.528603639118781e-05, + "loss": 0.1318, + "step": 29194 + }, + { + "epoch": 0.6843767764318407, + "grad_norm": 2.0583064556121826, + "learning_rate": 4.527987076095061e-05, + "loss": 0.1008, + "step": 29195 + }, + { + "epoch": 0.6844002180066456, + "grad_norm": 3.0382189750671387, + "learning_rate": 4.52737054276264e-05, + "loss": 0.2213, + "step": 29196 + }, + { + "epoch": 0.6844236595814507, + "grad_norm": 0.9170458912849426, + "learning_rate": 4.526754039124861e-05, + "loss": 0.061, + "step": 29197 + }, + { + "epoch": 0.6844471011562556, + "grad_norm": 2.282193183898926, + "learning_rate": 4.526137565185067e-05, + "loss": 0.2037, + "step": 29198 + }, + { + "epoch": 0.6844705427310607, + "grad_norm": 1.4400941133499146, + "learning_rate": 4.525521120946602e-05, + "loss": 0.1056, + "step": 29199 + }, + { + "epoch": 0.6844939843058657, + "grad_norm": 0.8693697452545166, + "learning_rate": 4.524904706412816e-05, + "loss": 0.0374, + "step": 29200 + }, + { + "epoch": 0.6845174258806707, + "grad_norm": 2.6869406700134277, + "learning_rate": 4.5242883215870514e-05, + "loss": 0.4405, + "step": 29201 + }, + { + "epoch": 0.6845408674554757, + "grad_norm": 2.249972343444824, + "learning_rate": 4.5236719664726526e-05, + "loss": 1.4292, + "step": 29202 + }, + { + "epoch": 0.6845643090302806, + "grad_norm": 3.2948570251464844, + "learning_rate": 4.5230556410729644e-05, + "loss": 0.2318, + "step": 29203 + }, + { + "epoch": 0.6845877506050857, + "grad_norm": 1.3274489641189575, + "learning_rate": 4.522439345391328e-05, + "loss": 0.0692, + "step": 29204 + }, + { + "epoch": 0.6846111921798906, + "grad_norm": 2.162484645843506, + "learning_rate": 4.521823079431088e-05, + "loss": 1.4514, + "step": 29205 + }, + { + "epoch": 0.6846346337546957, + "grad_norm": 2.7336244583129883, + "learning_rate": 4.521206843195594e-05, + "loss": 0.113, + "step": 29206 + }, + { + "epoch": 0.6846580753295006, + "grad_norm": 4.420712947845459, + "learning_rate": 4.5205906366881815e-05, + "loss": 0.5123, + "step": 29207 + }, + { + "epoch": 0.6846815169043057, + "grad_norm": 2.8393845558166504, + "learning_rate": 4.519974459912201e-05, + "loss": 0.3067, + "step": 29208 + }, + { + "epoch": 0.6847049584791106, + "grad_norm": 6.85927677154541, + "learning_rate": 4.519358312870994e-05, + "loss": 0.5133, + "step": 29209 + }, + { + "epoch": 0.6847284000539157, + "grad_norm": 2.2793421745300293, + "learning_rate": 4.518742195567902e-05, + "loss": 0.1982, + "step": 29210 + }, + { + "epoch": 0.6847518416287206, + "grad_norm": 1.1290689706802368, + "learning_rate": 4.5181261080062666e-05, + "loss": 0.1008, + "step": 29211 + }, + { + "epoch": 0.6847752832035257, + "grad_norm": 0.7770822644233704, + "learning_rate": 4.5175100501894364e-05, + "loss": 0.0641, + "step": 29212 + }, + { + "epoch": 0.6847987247783306, + "grad_norm": 3.7552192211151123, + "learning_rate": 4.51689402212075e-05, + "loss": 0.2682, + "step": 29213 + }, + { + "epoch": 0.6848221663531356, + "grad_norm": 1.8607043027877808, + "learning_rate": 4.5162780238035516e-05, + "loss": 0.1969, + "step": 29214 + }, + { + "epoch": 0.6848456079279406, + "grad_norm": 2.233593463897705, + "learning_rate": 4.515662055241183e-05, + "loss": 0.1971, + "step": 29215 + }, + { + "epoch": 0.6848690495027456, + "grad_norm": 2.21805477142334, + "learning_rate": 4.515046116436986e-05, + "loss": 0.232, + "step": 29216 + }, + { + "epoch": 0.6848924910775506, + "grad_norm": 5.064181804656982, + "learning_rate": 4.514430207394299e-05, + "loss": 0.2499, + "step": 29217 + }, + { + "epoch": 0.6849159326523556, + "grad_norm": 2.1830382347106934, + "learning_rate": 4.513814328116471e-05, + "loss": 0.1511, + "step": 29218 + }, + { + "epoch": 0.6849393742271606, + "grad_norm": 1.902445673942566, + "learning_rate": 4.5131984786068385e-05, + "loss": 0.1132, + "step": 29219 + }, + { + "epoch": 0.6849628158019656, + "grad_norm": 3.1644976139068604, + "learning_rate": 4.5125826588687484e-05, + "loss": 0.1962, + "step": 29220 + }, + { + "epoch": 0.6849862573767705, + "grad_norm": 3.172818422317505, + "learning_rate": 4.5119668689055394e-05, + "loss": 0.253, + "step": 29221 + }, + { + "epoch": 0.6850096989515756, + "grad_norm": 2.211697578430176, + "learning_rate": 4.5113511087205516e-05, + "loss": 1.4376, + "step": 29222 + }, + { + "epoch": 0.6850331405263805, + "grad_norm": 3.5051350593566895, + "learning_rate": 4.510735378317128e-05, + "loss": 0.2991, + "step": 29223 + }, + { + "epoch": 0.6850565821011856, + "grad_norm": 2.884289264678955, + "learning_rate": 4.510119677698604e-05, + "loss": 0.2651, + "step": 29224 + }, + { + "epoch": 0.6850800236759905, + "grad_norm": 3.2352964878082275, + "learning_rate": 4.509504006868329e-05, + "loss": 0.256, + "step": 29225 + }, + { + "epoch": 0.6851034652507956, + "grad_norm": 0.812831461429596, + "learning_rate": 4.508888365829641e-05, + "loss": 0.0358, + "step": 29226 + }, + { + "epoch": 0.6851269068256005, + "grad_norm": 8.350471496582031, + "learning_rate": 4.508272754585877e-05, + "loss": 0.3138, + "step": 29227 + }, + { + "epoch": 0.6851503484004056, + "grad_norm": 3.0583043098449707, + "learning_rate": 4.5076571731403804e-05, + "loss": 0.4122, + "step": 29228 + }, + { + "epoch": 0.6851737899752105, + "grad_norm": 1.201551079750061, + "learning_rate": 4.5070416214964905e-05, + "loss": 0.1491, + "step": 29229 + }, + { + "epoch": 0.6851972315500156, + "grad_norm": 1.74831223487854, + "learning_rate": 4.5064260996575434e-05, + "loss": 0.0956, + "step": 29230 + }, + { + "epoch": 0.6852206731248205, + "grad_norm": 3.4649906158447266, + "learning_rate": 4.505810607626884e-05, + "loss": 0.4285, + "step": 29231 + }, + { + "epoch": 0.6852441146996255, + "grad_norm": 2.291198492050171, + "learning_rate": 4.505195145407852e-05, + "loss": 0.1529, + "step": 29232 + }, + { + "epoch": 0.6852675562744305, + "grad_norm": 1.8430083990097046, + "learning_rate": 4.504579713003787e-05, + "loss": 0.1987, + "step": 29233 + }, + { + "epoch": 0.6852909978492355, + "grad_norm": 2.188884735107422, + "learning_rate": 4.5039643104180274e-05, + "loss": 0.165, + "step": 29234 + }, + { + "epoch": 0.6853144394240405, + "grad_norm": 1.9408961534500122, + "learning_rate": 4.50334893765391e-05, + "loss": 0.0793, + "step": 29235 + }, + { + "epoch": 0.6853378809988455, + "grad_norm": 4.656046390533447, + "learning_rate": 4.5027335947147773e-05, + "loss": 0.2837, + "step": 29236 + }, + { + "epoch": 0.6853613225736505, + "grad_norm": 3.302151679992676, + "learning_rate": 4.502118281603963e-05, + "loss": 0.2862, + "step": 29237 + }, + { + "epoch": 0.6853847641484555, + "grad_norm": 5.417627811431885, + "learning_rate": 4.501502998324814e-05, + "loss": 0.3259, + "step": 29238 + }, + { + "epoch": 0.6854082057232604, + "grad_norm": 1.6638374328613281, + "learning_rate": 4.500887744880663e-05, + "loss": 0.0925, + "step": 29239 + }, + { + "epoch": 0.6854316472980655, + "grad_norm": 1.9758507013320923, + "learning_rate": 4.5002725212748497e-05, + "loss": 0.0931, + "step": 29240 + }, + { + "epoch": 0.6854550888728704, + "grad_norm": 0.3231811821460724, + "learning_rate": 4.4996573275107135e-05, + "loss": 0.0273, + "step": 29241 + }, + { + "epoch": 0.6854785304476755, + "grad_norm": 4.171866416931152, + "learning_rate": 4.49904216359159e-05, + "loss": 0.4046, + "step": 29242 + }, + { + "epoch": 0.6855019720224804, + "grad_norm": 2.2737977504730225, + "learning_rate": 4.498427029520815e-05, + "loss": 0.1684, + "step": 29243 + }, + { + "epoch": 0.6855254135972855, + "grad_norm": 1.74850594997406, + "learning_rate": 4.49781192530173e-05, + "loss": 1.3449, + "step": 29244 + }, + { + "epoch": 0.6855488551720904, + "grad_norm": 0.22086572647094727, + "learning_rate": 4.497196850937676e-05, + "loss": 0.0237, + "step": 29245 + }, + { + "epoch": 0.6855722967468955, + "grad_norm": 2.372832775115967, + "learning_rate": 4.4965818064319865e-05, + "loss": 1.8684, + "step": 29246 + }, + { + "epoch": 0.6855957383217004, + "grad_norm": 3.9799137115478516, + "learning_rate": 4.495966791787999e-05, + "loss": 0.198, + "step": 29247 + }, + { + "epoch": 0.6856191798965054, + "grad_norm": 3.10522198677063, + "learning_rate": 4.4953518070090504e-05, + "loss": 0.3165, + "step": 29248 + }, + { + "epoch": 0.6856426214713104, + "grad_norm": 3.3376352787017822, + "learning_rate": 4.494736852098473e-05, + "loss": 0.1855, + "step": 29249 + }, + { + "epoch": 0.6856660630461154, + "grad_norm": 0.5856285691261292, + "learning_rate": 4.4941219270596135e-05, + "loss": 0.05, + "step": 29250 + }, + { + "epoch": 0.6856895046209205, + "grad_norm": 1.178605556488037, + "learning_rate": 4.493507031895803e-05, + "loss": 0.0822, + "step": 29251 + }, + { + "epoch": 0.6857129461957254, + "grad_norm": 1.8567627668380737, + "learning_rate": 4.4928921666103776e-05, + "loss": 0.1339, + "step": 29252 + }, + { + "epoch": 0.6857363877705305, + "grad_norm": 0.8330439329147339, + "learning_rate": 4.4922773312066734e-05, + "loss": 0.0662, + "step": 29253 + }, + { + "epoch": 0.6857598293453354, + "grad_norm": 2.8411200046539307, + "learning_rate": 4.491662525688028e-05, + "loss": 0.3206, + "step": 29254 + }, + { + "epoch": 0.6857832709201405, + "grad_norm": 2.417358636856079, + "learning_rate": 4.491047750057773e-05, + "loss": 1.3008, + "step": 29255 + }, + { + "epoch": 0.6858067124949454, + "grad_norm": 0.22910399734973907, + "learning_rate": 4.490433004319248e-05, + "loss": 0.0123, + "step": 29256 + }, + { + "epoch": 0.6858301540697505, + "grad_norm": 3.9110770225524902, + "learning_rate": 4.4898182884757925e-05, + "loss": 0.3291, + "step": 29257 + }, + { + "epoch": 0.6858535956445554, + "grad_norm": 3.7811155319213867, + "learning_rate": 4.4892036025307374e-05, + "loss": 0.3132, + "step": 29258 + }, + { + "epoch": 0.6858770372193604, + "grad_norm": 0.36753612756729126, + "learning_rate": 4.488588946487418e-05, + "loss": 0.0251, + "step": 29259 + }, + { + "epoch": 0.6859004787941654, + "grad_norm": 2.169638156890869, + "learning_rate": 4.48797432034917e-05, + "loss": 0.1063, + "step": 29260 + }, + { + "epoch": 0.6859239203689704, + "grad_norm": 0.32776331901550293, + "learning_rate": 4.487359724119329e-05, + "loss": 0.0348, + "step": 29261 + }, + { + "epoch": 0.6859473619437754, + "grad_norm": 1.413474678993225, + "learning_rate": 4.4867451578012244e-05, + "loss": 0.1703, + "step": 29262 + }, + { + "epoch": 0.6859708035185804, + "grad_norm": 2.5105855464935303, + "learning_rate": 4.4861306213982e-05, + "loss": 0.1609, + "step": 29263 + }, + { + "epoch": 0.6859942450933854, + "grad_norm": 1.1097261905670166, + "learning_rate": 4.485516114913586e-05, + "loss": 0.0795, + "step": 29264 + }, + { + "epoch": 0.6860176866681904, + "grad_norm": 2.6295506954193115, + "learning_rate": 4.484901638350717e-05, + "loss": 0.287, + "step": 29265 + }, + { + "epoch": 0.6860411282429953, + "grad_norm": 2.883760690689087, + "learning_rate": 4.484287191712926e-05, + "loss": 0.213, + "step": 29266 + }, + { + "epoch": 0.6860645698178004, + "grad_norm": 3.509363889694214, + "learning_rate": 4.483672775003548e-05, + "loss": 0.3814, + "step": 29267 + }, + { + "epoch": 0.6860880113926053, + "grad_norm": 1.2565832138061523, + "learning_rate": 4.483058388225913e-05, + "loss": 0.1192, + "step": 29268 + }, + { + "epoch": 0.6861114529674104, + "grad_norm": 2.778752565383911, + "learning_rate": 4.482444031383358e-05, + "loss": 0.2033, + "step": 29269 + }, + { + "epoch": 0.6861348945422153, + "grad_norm": 2.8589649200439453, + "learning_rate": 4.481829704479221e-05, + "loss": 1.4764, + "step": 29270 + }, + { + "epoch": 0.6861583361170204, + "grad_norm": 0.6195129156112671, + "learning_rate": 4.481215407516831e-05, + "loss": 0.0537, + "step": 29271 + }, + { + "epoch": 0.6861817776918253, + "grad_norm": 3.896825075149536, + "learning_rate": 4.480601140499522e-05, + "loss": 0.3714, + "step": 29272 + }, + { + "epoch": 0.6862052192666304, + "grad_norm": 0.9620806574821472, + "learning_rate": 4.479986903430625e-05, + "loss": 0.0802, + "step": 29273 + }, + { + "epoch": 0.6862286608414353, + "grad_norm": 2.401205062866211, + "learning_rate": 4.479372696313475e-05, + "loss": 0.2253, + "step": 29274 + }, + { + "epoch": 0.6862521024162404, + "grad_norm": 2.2559964656829834, + "learning_rate": 4.4787585191514004e-05, + "loss": 0.1722, + "step": 29275 + }, + { + "epoch": 0.6862755439910453, + "grad_norm": 3.929903984069824, + "learning_rate": 4.478144371947741e-05, + "loss": 0.3183, + "step": 29276 + }, + { + "epoch": 0.6862989855658503, + "grad_norm": 2.1473424434661865, + "learning_rate": 4.4775302547058265e-05, + "loss": 0.1056, + "step": 29277 + }, + { + "epoch": 0.6863224271406553, + "grad_norm": 2.608818292617798, + "learning_rate": 4.476916167428986e-05, + "loss": 0.2227, + "step": 29278 + }, + { + "epoch": 0.6863458687154603, + "grad_norm": 2.1972432136535645, + "learning_rate": 4.476302110120555e-05, + "loss": 0.2044, + "step": 29279 + }, + { + "epoch": 0.6863693102902653, + "grad_norm": 6.394810199737549, + "learning_rate": 4.47568808278386e-05, + "loss": 0.4559, + "step": 29280 + }, + { + "epoch": 0.6863927518650703, + "grad_norm": 1.9406187534332275, + "learning_rate": 4.475074085422241e-05, + "loss": 0.2041, + "step": 29281 + }, + { + "epoch": 0.6864161934398753, + "grad_norm": 1.5703891515731812, + "learning_rate": 4.474460118039021e-05, + "loss": 0.0968, + "step": 29282 + }, + { + "epoch": 0.6864396350146803, + "grad_norm": 3.427098512649536, + "learning_rate": 4.47384618063754e-05, + "loss": 0.326, + "step": 29283 + }, + { + "epoch": 0.6864630765894852, + "grad_norm": 1.2644506692886353, + "learning_rate": 4.473232273221124e-05, + "loss": 0.1354, + "step": 29284 + }, + { + "epoch": 0.6864865181642903, + "grad_norm": 3.113809823989868, + "learning_rate": 4.472618395793107e-05, + "loss": 0.3973, + "step": 29285 + }, + { + "epoch": 0.6865099597390952, + "grad_norm": 1.0450503826141357, + "learning_rate": 4.4720045483568164e-05, + "loss": 0.1363, + "step": 29286 + }, + { + "epoch": 0.6865334013139003, + "grad_norm": 1.3035894632339478, + "learning_rate": 4.471390730915581e-05, + "loss": 0.1196, + "step": 29287 + }, + { + "epoch": 0.6865568428887052, + "grad_norm": 3.8287768363952637, + "learning_rate": 4.470776943472739e-05, + "loss": 0.2729, + "step": 29288 + }, + { + "epoch": 0.6865802844635103, + "grad_norm": 2.284336566925049, + "learning_rate": 4.470163186031616e-05, + "loss": 0.3385, + "step": 29289 + }, + { + "epoch": 0.6866037260383152, + "grad_norm": 1.5304068326950073, + "learning_rate": 4.469549458595545e-05, + "loss": 0.0862, + "step": 29290 + }, + { + "epoch": 0.6866271676131203, + "grad_norm": 0.12433375418186188, + "learning_rate": 4.4689357611678516e-05, + "loss": 0.003, + "step": 29291 + }, + { + "epoch": 0.6866506091879252, + "grad_norm": 0.15179859101772308, + "learning_rate": 4.4683220937518666e-05, + "loss": 0.0112, + "step": 29292 + }, + { + "epoch": 0.6866740507627302, + "grad_norm": 2.9315078258514404, + "learning_rate": 4.467708456350924e-05, + "loss": 0.3162, + "step": 29293 + }, + { + "epoch": 0.6866974923375352, + "grad_norm": 0.6458041071891785, + "learning_rate": 4.467094848968347e-05, + "loss": 0.0445, + "step": 29294 + }, + { + "epoch": 0.6867209339123402, + "grad_norm": 1.4018679857254028, + "learning_rate": 4.4664812716074734e-05, + "loss": 0.1234, + "step": 29295 + }, + { + "epoch": 0.6867443754871452, + "grad_norm": 2.9375030994415283, + "learning_rate": 4.4658677242716285e-05, + "loss": 0.1534, + "step": 29296 + }, + { + "epoch": 0.6867678170619502, + "grad_norm": 2.453622341156006, + "learning_rate": 4.4652542069641404e-05, + "loss": 0.2349, + "step": 29297 + }, + { + "epoch": 0.6867912586367552, + "grad_norm": 0.9948170185089111, + "learning_rate": 4.4646407196883375e-05, + "loss": 0.0581, + "step": 29298 + }, + { + "epoch": 0.6868147002115602, + "grad_norm": 4.494681358337402, + "learning_rate": 4.464027262447551e-05, + "loss": 0.4649, + "step": 29299 + }, + { + "epoch": 0.6868381417863652, + "grad_norm": 1.5060458183288574, + "learning_rate": 4.463413835245104e-05, + "loss": 0.1194, + "step": 29300 + }, + { + "epoch": 0.6868615833611702, + "grad_norm": 1.537209153175354, + "learning_rate": 4.462800438084334e-05, + "loss": 0.0957, + "step": 29301 + }, + { + "epoch": 0.6868850249359753, + "grad_norm": 2.930079221725464, + "learning_rate": 4.462187070968563e-05, + "loss": 0.1566, + "step": 29302 + }, + { + "epoch": 0.6869084665107802, + "grad_norm": 1.0936609506607056, + "learning_rate": 4.461573733901121e-05, + "loss": 0.0657, + "step": 29303 + }, + { + "epoch": 0.6869319080855852, + "grad_norm": 3.1682446002960205, + "learning_rate": 4.4609604268853324e-05, + "loss": 0.33, + "step": 29304 + }, + { + "epoch": 0.6869553496603902, + "grad_norm": 1.3756684064865112, + "learning_rate": 4.460347149924532e-05, + "loss": 0.1109, + "step": 29305 + }, + { + "epoch": 0.6869787912351952, + "grad_norm": 5.783763408660889, + "learning_rate": 4.459733903022043e-05, + "loss": 0.8934, + "step": 29306 + }, + { + "epoch": 0.6870022328100002, + "grad_norm": 0.9855126142501831, + "learning_rate": 4.459120686181191e-05, + "loss": 0.1063, + "step": 29307 + }, + { + "epoch": 0.6870256743848052, + "grad_norm": 3.0014259815216064, + "learning_rate": 4.458507499405309e-05, + "loss": 0.3168, + "step": 29308 + }, + { + "epoch": 0.6870491159596102, + "grad_norm": 2.4443235397338867, + "learning_rate": 4.4578943426977205e-05, + "loss": 0.2712, + "step": 29309 + }, + { + "epoch": 0.6870725575344152, + "grad_norm": 1.2021722793579102, + "learning_rate": 4.457281216061754e-05, + "loss": 0.1285, + "step": 29310 + }, + { + "epoch": 0.6870959991092201, + "grad_norm": 3.5041120052337646, + "learning_rate": 4.456668119500736e-05, + "loss": 0.5298, + "step": 29311 + }, + { + "epoch": 0.6871194406840252, + "grad_norm": 2.7595694065093994, + "learning_rate": 4.456055053017991e-05, + "loss": 0.1826, + "step": 29312 + }, + { + "epoch": 0.6871428822588301, + "grad_norm": 3.092574119567871, + "learning_rate": 4.455442016616844e-05, + "loss": 0.2404, + "step": 29313 + }, + { + "epoch": 0.6871663238336352, + "grad_norm": 1.4181569814682007, + "learning_rate": 4.4548290103006295e-05, + "loss": 0.0777, + "step": 29314 + }, + { + "epoch": 0.6871897654084401, + "grad_norm": 0.5863091945648193, + "learning_rate": 4.454216034072668e-05, + "loss": 0.0751, + "step": 29315 + }, + { + "epoch": 0.6872132069832452, + "grad_norm": 2.7134509086608887, + "learning_rate": 4.453603087936282e-05, + "loss": 0.1483, + "step": 29316 + }, + { + "epoch": 0.6872366485580501, + "grad_norm": 0.8412912487983704, + "learning_rate": 4.452990171894805e-05, + "loss": 0.0848, + "step": 29317 + }, + { + "epoch": 0.6872600901328552, + "grad_norm": 2.3870222568511963, + "learning_rate": 4.45237728595156e-05, + "loss": 0.219, + "step": 29318 + }, + { + "epoch": 0.6872835317076601, + "grad_norm": 3.136284351348877, + "learning_rate": 4.451764430109873e-05, + "loss": 1.3143, + "step": 29319 + }, + { + "epoch": 0.6873069732824652, + "grad_norm": 3.216844081878662, + "learning_rate": 4.451151604373063e-05, + "loss": 0.2236, + "step": 29320 + }, + { + "epoch": 0.6873304148572701, + "grad_norm": 2.5225586891174316, + "learning_rate": 4.4505388087444635e-05, + "loss": 0.2279, + "step": 29321 + }, + { + "epoch": 0.6873538564320751, + "grad_norm": 2.7895328998565674, + "learning_rate": 4.449926043227398e-05, + "loss": 0.1474, + "step": 29322 + }, + { + "epoch": 0.6873772980068801, + "grad_norm": 2.238481283187866, + "learning_rate": 4.449313307825189e-05, + "loss": 0.1503, + "step": 29323 + }, + { + "epoch": 0.6874007395816851, + "grad_norm": 0.2402791529893875, + "learning_rate": 4.4487006025411625e-05, + "loss": 0.0206, + "step": 29324 + }, + { + "epoch": 0.6874241811564901, + "grad_norm": 2.3303706645965576, + "learning_rate": 4.448087927378638e-05, + "loss": 0.1074, + "step": 29325 + }, + { + "epoch": 0.6874476227312951, + "grad_norm": 1.753639817237854, + "learning_rate": 4.4474752823409495e-05, + "loss": 1.2148, + "step": 29326 + }, + { + "epoch": 0.6874710643061, + "grad_norm": 1.0159085988998413, + "learning_rate": 4.446862667431415e-05, + "loss": 0.0771, + "step": 29327 + }, + { + "epoch": 0.6874945058809051, + "grad_norm": 3.1092259883880615, + "learning_rate": 4.44625008265336e-05, + "loss": 0.2348, + "step": 29328 + }, + { + "epoch": 0.68751794745571, + "grad_norm": 0.8087366819381714, + "learning_rate": 4.445637528010106e-05, + "loss": 0.0472, + "step": 29329 + }, + { + "epoch": 0.6875413890305151, + "grad_norm": 3.1955618858337402, + "learning_rate": 4.445025003504981e-05, + "loss": 0.3712, + "step": 29330 + }, + { + "epoch": 0.68756483060532, + "grad_norm": 0.5345845222473145, + "learning_rate": 4.444412509141307e-05, + "loss": 0.0621, + "step": 29331 + }, + { + "epoch": 0.6875882721801251, + "grad_norm": 3.1779279708862305, + "learning_rate": 4.4438000449224036e-05, + "loss": 0.1477, + "step": 29332 + }, + { + "epoch": 0.68761171375493, + "grad_norm": 4.055814266204834, + "learning_rate": 4.443187610851601e-05, + "loss": 0.3989, + "step": 29333 + }, + { + "epoch": 0.6876351553297351, + "grad_norm": 3.7363486289978027, + "learning_rate": 4.442575206932218e-05, + "loss": 0.1404, + "step": 29334 + }, + { + "epoch": 0.68765859690454, + "grad_norm": 2.209074020385742, + "learning_rate": 4.441962833167579e-05, + "loss": 0.1365, + "step": 29335 + }, + { + "epoch": 0.6876820384793451, + "grad_norm": 2.523064374923706, + "learning_rate": 4.441350489561006e-05, + "loss": 0.2561, + "step": 29336 + }, + { + "epoch": 0.68770548005415, + "grad_norm": 3.738133192062378, + "learning_rate": 4.4407381761158205e-05, + "loss": 0.4132, + "step": 29337 + }, + { + "epoch": 0.687728921628955, + "grad_norm": 0.4628670811653137, + "learning_rate": 4.4401258928353426e-05, + "loss": 0.0434, + "step": 29338 + }, + { + "epoch": 0.68775236320376, + "grad_norm": 3.3353588581085205, + "learning_rate": 4.439513639722902e-05, + "loss": 0.1759, + "step": 29339 + }, + { + "epoch": 0.687775804778565, + "grad_norm": 0.4907189905643463, + "learning_rate": 4.4389014167818175e-05, + "loss": 0.0702, + "step": 29340 + }, + { + "epoch": 0.68779924635337, + "grad_norm": 2.285024881362915, + "learning_rate": 4.438289224015405e-05, + "loss": 0.2422, + "step": 29341 + }, + { + "epoch": 0.687822687928175, + "grad_norm": 5.260673999786377, + "learning_rate": 4.437677061426997e-05, + "loss": 0.4235, + "step": 29342 + }, + { + "epoch": 0.68784612950298, + "grad_norm": 0.3074061870574951, + "learning_rate": 4.4370649290199084e-05, + "loss": 0.0169, + "step": 29343 + }, + { + "epoch": 0.687869571077785, + "grad_norm": 2.474968433380127, + "learning_rate": 4.436452826797462e-05, + "loss": 0.2556, + "step": 29344 + }, + { + "epoch": 0.68789301265259, + "grad_norm": 2.743729591369629, + "learning_rate": 4.4358407547629756e-05, + "loss": 0.2194, + "step": 29345 + }, + { + "epoch": 0.687916454227395, + "grad_norm": 0.3429698348045349, + "learning_rate": 4.4352287129197776e-05, + "loss": 0.054, + "step": 29346 + }, + { + "epoch": 0.6879398958021999, + "grad_norm": 2.367905855178833, + "learning_rate": 4.4346167012711846e-05, + "loss": 0.2274, + "step": 29347 + }, + { + "epoch": 0.687963337377005, + "grad_norm": 2.683028221130371, + "learning_rate": 4.434004719820518e-05, + "loss": 0.281, + "step": 29348 + }, + { + "epoch": 0.6879867789518099, + "grad_norm": 4.221334934234619, + "learning_rate": 4.433392768571098e-05, + "loss": 0.4126, + "step": 29349 + }, + { + "epoch": 0.688010220526615, + "grad_norm": 0.4719026982784271, + "learning_rate": 4.4327808475262456e-05, + "loss": 0.034, + "step": 29350 + }, + { + "epoch": 0.6880336621014199, + "grad_norm": 1.0873796939849854, + "learning_rate": 4.4321689566892774e-05, + "loss": 0.0865, + "step": 29351 + }, + { + "epoch": 0.688057103676225, + "grad_norm": 1.1408119201660156, + "learning_rate": 4.431557096063521e-05, + "loss": 0.1143, + "step": 29352 + }, + { + "epoch": 0.68808054525103, + "grad_norm": 2.619398593902588, + "learning_rate": 4.430945265652288e-05, + "loss": 1.6585, + "step": 29353 + }, + { + "epoch": 0.688103986825835, + "grad_norm": 2.6232893466949463, + "learning_rate": 4.430333465458907e-05, + "loss": 0.2049, + "step": 29354 + }, + { + "epoch": 0.68812742840064, + "grad_norm": 0.36353087425231934, + "learning_rate": 4.429721695486693e-05, + "loss": 0.0426, + "step": 29355 + }, + { + "epoch": 0.688150869975445, + "grad_norm": 2.122097969055176, + "learning_rate": 4.429109955738965e-05, + "loss": 0.2002, + "step": 29356 + }, + { + "epoch": 0.68817431155025, + "grad_norm": 4.375586032867432, + "learning_rate": 4.428498246219043e-05, + "loss": 0.2578, + "step": 29357 + }, + { + "epoch": 0.6881977531250549, + "grad_norm": 4.431764602661133, + "learning_rate": 4.427886566930243e-05, + "loss": 0.3708, + "step": 29358 + }, + { + "epoch": 0.68822119469986, + "grad_norm": 4.015775680541992, + "learning_rate": 4.427274917875891e-05, + "loss": 0.457, + "step": 29359 + }, + { + "epoch": 0.6882446362746649, + "grad_norm": 5.020622253417969, + "learning_rate": 4.426663299059301e-05, + "loss": 0.3486, + "step": 29360 + }, + { + "epoch": 0.68826807784947, + "grad_norm": 0.35847708582878113, + "learning_rate": 4.426051710483794e-05, + "loss": 0.0203, + "step": 29361 + }, + { + "epoch": 0.6882915194242749, + "grad_norm": 3.4190263748168945, + "learning_rate": 4.425440152152687e-05, + "loss": 0.2604, + "step": 29362 + }, + { + "epoch": 0.68831496099908, + "grad_norm": 2.079817771911621, + "learning_rate": 4.424828624069294e-05, + "loss": 0.1838, + "step": 29363 + }, + { + "epoch": 0.6883384025738849, + "grad_norm": 10.980125427246094, + "learning_rate": 4.424217126236941e-05, + "loss": 0.1883, + "step": 29364 + }, + { + "epoch": 0.68836184414869, + "grad_norm": 1.5546534061431885, + "learning_rate": 4.4236056586589406e-05, + "loss": 0.0696, + "step": 29365 + }, + { + "epoch": 0.6883852857234949, + "grad_norm": 2.7179436683654785, + "learning_rate": 4.4229942213386156e-05, + "loss": 1.4261, + "step": 29366 + }, + { + "epoch": 0.6884087272982999, + "grad_norm": 0.6119775176048279, + "learning_rate": 4.42238281427928e-05, + "loss": 0.0311, + "step": 29367 + }, + { + "epoch": 0.6884321688731049, + "grad_norm": 1.8063749074935913, + "learning_rate": 4.4217714374842525e-05, + "loss": 1.0241, + "step": 29368 + }, + { + "epoch": 0.6884556104479099, + "grad_norm": 2.632004976272583, + "learning_rate": 4.4211600909568504e-05, + "loss": 0.2932, + "step": 29369 + }, + { + "epoch": 0.6884790520227149, + "grad_norm": 2.370875835418701, + "learning_rate": 4.420548774700387e-05, + "loss": 1.7913, + "step": 29370 + }, + { + "epoch": 0.6885024935975199, + "grad_norm": 3.7638015747070312, + "learning_rate": 4.419937488718185e-05, + "loss": 0.1936, + "step": 29371 + }, + { + "epoch": 0.6885259351723249, + "grad_norm": 3.317204713821411, + "learning_rate": 4.419326233013561e-05, + "loss": 0.2959, + "step": 29372 + }, + { + "epoch": 0.6885493767471299, + "grad_norm": 4.728260517120361, + "learning_rate": 4.4187150075898285e-05, + "loss": 0.3792, + "step": 29373 + }, + { + "epoch": 0.6885728183219348, + "grad_norm": 2.273297071456909, + "learning_rate": 4.418103812450306e-05, + "loss": 1.4023, + "step": 29374 + }, + { + "epoch": 0.6885962598967399, + "grad_norm": 2.3151333332061768, + "learning_rate": 4.417492647598309e-05, + "loss": 0.222, + "step": 29375 + }, + { + "epoch": 0.6886197014715448, + "grad_norm": 2.794541120529175, + "learning_rate": 4.41688151303715e-05, + "loss": 0.313, + "step": 29376 + }, + { + "epoch": 0.6886431430463499, + "grad_norm": 2.2352538108825684, + "learning_rate": 4.416270408770149e-05, + "loss": 0.1068, + "step": 29377 + }, + { + "epoch": 0.6886665846211548, + "grad_norm": 3.4644393920898438, + "learning_rate": 4.415659334800626e-05, + "loss": 0.2999, + "step": 29378 + }, + { + "epoch": 0.6886900261959599, + "grad_norm": 1.8894741535186768, + "learning_rate": 4.4150482911318914e-05, + "loss": 0.1369, + "step": 29379 + }, + { + "epoch": 0.6887134677707648, + "grad_norm": 0.7319827675819397, + "learning_rate": 4.4144372777672625e-05, + "loss": 0.0664, + "step": 29380 + }, + { + "epoch": 0.6887369093455699, + "grad_norm": 2.3463916778564453, + "learning_rate": 4.413826294710054e-05, + "loss": 0.1144, + "step": 29381 + }, + { + "epoch": 0.6887603509203748, + "grad_norm": 4.70914888381958, + "learning_rate": 4.413215341963581e-05, + "loss": 0.416, + "step": 29382 + }, + { + "epoch": 0.6887837924951798, + "grad_norm": 2.2052996158599854, + "learning_rate": 4.4126044195311545e-05, + "loss": 1.6918, + "step": 29383 + }, + { + "epoch": 0.6888072340699848, + "grad_norm": 2.1648879051208496, + "learning_rate": 4.411993527416097e-05, + "loss": 0.158, + "step": 29384 + }, + { + "epoch": 0.6888306756447898, + "grad_norm": 3.8354756832122803, + "learning_rate": 4.41138266562172e-05, + "loss": 0.3774, + "step": 29385 + }, + { + "epoch": 0.6888541172195948, + "grad_norm": 2.941298723220825, + "learning_rate": 4.410771834151339e-05, + "loss": 1.7958, + "step": 29386 + }, + { + "epoch": 0.6888775587943998, + "grad_norm": 2.1109490394592285, + "learning_rate": 4.410161033008267e-05, + "loss": 0.2952, + "step": 29387 + }, + { + "epoch": 0.6889010003692048, + "grad_norm": 2.0038368701934814, + "learning_rate": 4.409550262195817e-05, + "loss": 0.1465, + "step": 29388 + }, + { + "epoch": 0.6889244419440098, + "grad_norm": 1.7603329420089722, + "learning_rate": 4.408939521717301e-05, + "loss": 0.1391, + "step": 29389 + }, + { + "epoch": 0.6889478835188148, + "grad_norm": 4.437039852142334, + "learning_rate": 4.408328811576038e-05, + "loss": 0.481, + "step": 29390 + }, + { + "epoch": 0.6889713250936198, + "grad_norm": 1.902606725692749, + "learning_rate": 4.4077181317753426e-05, + "loss": 1.3753, + "step": 29391 + }, + { + "epoch": 0.6889947666684247, + "grad_norm": 1.909494400024414, + "learning_rate": 4.407107482318526e-05, + "loss": 0.1283, + "step": 29392 + }, + { + "epoch": 0.6890182082432298, + "grad_norm": 3.120055675506592, + "learning_rate": 4.406496863208901e-05, + "loss": 0.2021, + "step": 29393 + }, + { + "epoch": 0.6890416498180347, + "grad_norm": 0.32283690571784973, + "learning_rate": 4.405886274449782e-05, + "loss": 0.031, + "step": 29394 + }, + { + "epoch": 0.6890650913928398, + "grad_norm": 2.17242169380188, + "learning_rate": 4.40527571604448e-05, + "loss": 0.1193, + "step": 29395 + }, + { + "epoch": 0.6890885329676447, + "grad_norm": 1.9374269247055054, + "learning_rate": 4.404665187996307e-05, + "loss": 0.1463, + "step": 29396 + }, + { + "epoch": 0.6891119745424498, + "grad_norm": 3.55208158493042, + "learning_rate": 4.404054690308581e-05, + "loss": 0.1821, + "step": 29397 + }, + { + "epoch": 0.6891354161172547, + "grad_norm": 2.6662967205047607, + "learning_rate": 4.403444222984612e-05, + "loss": 0.2426, + "step": 29398 + }, + { + "epoch": 0.6891588576920598, + "grad_norm": 2.4604127407073975, + "learning_rate": 4.402833786027712e-05, + "loss": 0.1838, + "step": 29399 + }, + { + "epoch": 0.6891822992668647, + "grad_norm": 3.9483773708343506, + "learning_rate": 4.402223379441194e-05, + "loss": 0.1586, + "step": 29400 + }, + { + "epoch": 0.6892057408416697, + "grad_norm": 0.4297148883342743, + "learning_rate": 4.401613003228364e-05, + "loss": 0.0394, + "step": 29401 + }, + { + "epoch": 0.6892291824164747, + "grad_norm": 0.5034379959106445, + "learning_rate": 4.4010026573925434e-05, + "loss": 0.0619, + "step": 29402 + }, + { + "epoch": 0.6892526239912797, + "grad_norm": 4.825687885284424, + "learning_rate": 4.400392341937037e-05, + "loss": 0.4897, + "step": 29403 + }, + { + "epoch": 0.6892760655660848, + "grad_norm": 2.8633642196655273, + "learning_rate": 4.399782056865163e-05, + "loss": 0.3278, + "step": 29404 + }, + { + "epoch": 0.6892995071408897, + "grad_norm": 3.9854156970977783, + "learning_rate": 4.399171802180227e-05, + "loss": 0.2608, + "step": 29405 + }, + { + "epoch": 0.6893229487156948, + "grad_norm": 3.2315659523010254, + "learning_rate": 4.398561577885545e-05, + "loss": 0.4169, + "step": 29406 + }, + { + "epoch": 0.6893463902904997, + "grad_norm": 2.8393003940582275, + "learning_rate": 4.397951383984423e-05, + "loss": 0.1839, + "step": 29407 + }, + { + "epoch": 0.6893698318653048, + "grad_norm": 1.6078354120254517, + "learning_rate": 4.3973412204801726e-05, + "loss": 0.1484, + "step": 29408 + }, + { + "epoch": 0.6893932734401097, + "grad_norm": 2.58959698677063, + "learning_rate": 4.396731087376108e-05, + "loss": 0.2574, + "step": 29409 + }, + { + "epoch": 0.6894167150149147, + "grad_norm": 0.572808563709259, + "learning_rate": 4.396120984675539e-05, + "loss": 0.0287, + "step": 29410 + }, + { + "epoch": 0.6894401565897197, + "grad_norm": 3.278031826019287, + "learning_rate": 4.395510912381776e-05, + "loss": 0.1895, + "step": 29411 + }, + { + "epoch": 0.6894635981645247, + "grad_norm": 5.036642074584961, + "learning_rate": 4.394900870498128e-05, + "loss": 0.4463, + "step": 29412 + }, + { + "epoch": 0.6894870397393297, + "grad_norm": 3.0252761840820312, + "learning_rate": 4.3942908590279044e-05, + "loss": 0.1398, + "step": 29413 + }, + { + "epoch": 0.6895104813141347, + "grad_norm": 2.8709137439727783, + "learning_rate": 4.393680877974413e-05, + "loss": 1.3307, + "step": 29414 + }, + { + "epoch": 0.6895339228889397, + "grad_norm": 0.6059035658836365, + "learning_rate": 4.393070927340967e-05, + "loss": 0.0379, + "step": 29415 + }, + { + "epoch": 0.6895573644637447, + "grad_norm": 1.9096256494522095, + "learning_rate": 4.3924610071308794e-05, + "loss": 0.1763, + "step": 29416 + }, + { + "epoch": 0.6895808060385497, + "grad_norm": 2.7272374629974365, + "learning_rate": 4.391851117347456e-05, + "loss": 0.3055, + "step": 29417 + }, + { + "epoch": 0.6896042476133547, + "grad_norm": 2.4021854400634766, + "learning_rate": 4.391241257994006e-05, + "loss": 0.2124, + "step": 29418 + }, + { + "epoch": 0.6896276891881596, + "grad_norm": 1.2664576768875122, + "learning_rate": 4.3906314290738396e-05, + "loss": 0.0628, + "step": 29419 + }, + { + "epoch": 0.6896511307629647, + "grad_norm": 2.013455629348755, + "learning_rate": 4.390021630590264e-05, + "loss": 1.1707, + "step": 29420 + }, + { + "epoch": 0.6896745723377696, + "grad_norm": 2.0675313472747803, + "learning_rate": 4.389411862546585e-05, + "loss": 1.0451, + "step": 29421 + }, + { + "epoch": 0.6896980139125747, + "grad_norm": 2.038309097290039, + "learning_rate": 4.3888021249461186e-05, + "loss": 0.1893, + "step": 29422 + }, + { + "epoch": 0.6897214554873796, + "grad_norm": 2.3388190269470215, + "learning_rate": 4.388192417792169e-05, + "loss": 0.3077, + "step": 29423 + }, + { + "epoch": 0.6897448970621847, + "grad_norm": 0.5226054191589355, + "learning_rate": 4.387582741088047e-05, + "loss": 0.0443, + "step": 29424 + }, + { + "epoch": 0.6897683386369896, + "grad_norm": 2.844909429550171, + "learning_rate": 4.386973094837058e-05, + "loss": 0.2867, + "step": 29425 + }, + { + "epoch": 0.6897917802117947, + "grad_norm": 1.5689902305603027, + "learning_rate": 4.386363479042507e-05, + "loss": 0.1951, + "step": 29426 + }, + { + "epoch": 0.6898152217865996, + "grad_norm": 2.6468725204467773, + "learning_rate": 4.385753893707709e-05, + "loss": 0.2867, + "step": 29427 + }, + { + "epoch": 0.6898386633614046, + "grad_norm": 0.8014426231384277, + "learning_rate": 4.385144338835965e-05, + "loss": 0.0545, + "step": 29428 + }, + { + "epoch": 0.6898621049362096, + "grad_norm": 2.575925588607788, + "learning_rate": 4.384534814430589e-05, + "loss": 1.678, + "step": 29429 + }, + { + "epoch": 0.6898855465110146, + "grad_norm": 0.7105898857116699, + "learning_rate": 4.383925320494885e-05, + "loss": 0.0445, + "step": 29430 + }, + { + "epoch": 0.6899089880858196, + "grad_norm": 3.351696729660034, + "learning_rate": 4.3833158570321606e-05, + "loss": 0.346, + "step": 29431 + }, + { + "epoch": 0.6899324296606246, + "grad_norm": 2.044454574584961, + "learning_rate": 4.382706424045722e-05, + "loss": 1.264, + "step": 29432 + }, + { + "epoch": 0.6899558712354296, + "grad_norm": 3.4995951652526855, + "learning_rate": 4.3820970215388766e-05, + "loss": 0.3112, + "step": 29433 + }, + { + "epoch": 0.6899793128102346, + "grad_norm": 3.8806304931640625, + "learning_rate": 4.381487649514926e-05, + "loss": 0.2028, + "step": 29434 + }, + { + "epoch": 0.6900027543850396, + "grad_norm": 0.6611467003822327, + "learning_rate": 4.380878307977186e-05, + "loss": 0.0659, + "step": 29435 + }, + { + "epoch": 0.6900261959598446, + "grad_norm": 0.7352832555770874, + "learning_rate": 4.380268996928958e-05, + "loss": 0.0545, + "step": 29436 + }, + { + "epoch": 0.6900496375346495, + "grad_norm": 1.6668498516082764, + "learning_rate": 4.3796597163735486e-05, + "loss": 0.1612, + "step": 29437 + }, + { + "epoch": 0.6900730791094546, + "grad_norm": 0.9335618019104004, + "learning_rate": 4.3790504663142594e-05, + "loss": 0.0955, + "step": 29438 + }, + { + "epoch": 0.6900965206842595, + "grad_norm": 1.732951045036316, + "learning_rate": 4.3784412467544054e-05, + "loss": 0.1841, + "step": 29439 + }, + { + "epoch": 0.6901199622590646, + "grad_norm": 0.9149571061134338, + "learning_rate": 4.3778320576972865e-05, + "loss": 0.0597, + "step": 29440 + }, + { + "epoch": 0.6901434038338695, + "grad_norm": 0.9746533632278442, + "learning_rate": 4.377222899146205e-05, + "loss": 0.11, + "step": 29441 + }, + { + "epoch": 0.6901668454086746, + "grad_norm": 3.7569022178649902, + "learning_rate": 4.3766137711044744e-05, + "loss": 0.3023, + "step": 29442 + }, + { + "epoch": 0.6901902869834795, + "grad_norm": 1.5910282135009766, + "learning_rate": 4.376004673575395e-05, + "loss": 0.0676, + "step": 29443 + }, + { + "epoch": 0.6902137285582846, + "grad_norm": 1.6199010610580444, + "learning_rate": 4.375395606562273e-05, + "loss": 0.173, + "step": 29444 + }, + { + "epoch": 0.6902371701330895, + "grad_norm": 1.9202309846878052, + "learning_rate": 4.374786570068411e-05, + "loss": 0.184, + "step": 29445 + }, + { + "epoch": 0.6902606117078945, + "grad_norm": 1.738608479499817, + "learning_rate": 4.374177564097113e-05, + "loss": 0.0895, + "step": 29446 + }, + { + "epoch": 0.6902840532826995, + "grad_norm": 2.848440170288086, + "learning_rate": 4.373568588651689e-05, + "loss": 0.3464, + "step": 29447 + }, + { + "epoch": 0.6903074948575045, + "grad_norm": 3.8992137908935547, + "learning_rate": 4.37295964373544e-05, + "loss": 0.2655, + "step": 29448 + }, + { + "epoch": 0.6903309364323095, + "grad_norm": 2.4831149578094482, + "learning_rate": 4.3723507293516704e-05, + "loss": 0.1483, + "step": 29449 + }, + { + "epoch": 0.6903543780071145, + "grad_norm": 3.790585517883301, + "learning_rate": 4.3717418455036804e-05, + "loss": 0.2685, + "step": 29450 + }, + { + "epoch": 0.6903778195819195, + "grad_norm": 2.6958167552948, + "learning_rate": 4.371132992194781e-05, + "loss": 1.6654, + "step": 29451 + }, + { + "epoch": 0.6904012611567245, + "grad_norm": 2.3788444995880127, + "learning_rate": 4.3705241694282714e-05, + "loss": 0.1417, + "step": 29452 + }, + { + "epoch": 0.6904247027315294, + "grad_norm": 0.8620824813842773, + "learning_rate": 4.3699153772074544e-05, + "loss": 0.0759, + "step": 29453 + }, + { + "epoch": 0.6904481443063345, + "grad_norm": 2.928234100341797, + "learning_rate": 4.369306615535636e-05, + "loss": 0.1033, + "step": 29454 + }, + { + "epoch": 0.6904715858811395, + "grad_norm": 3.912540912628174, + "learning_rate": 4.36869788441612e-05, + "loss": 0.4092, + "step": 29455 + }, + { + "epoch": 0.6904950274559445, + "grad_norm": 2.59424090385437, + "learning_rate": 4.368089183852207e-05, + "loss": 0.2104, + "step": 29456 + }, + { + "epoch": 0.6905184690307495, + "grad_norm": 3.390812397003174, + "learning_rate": 4.367480513847201e-05, + "loss": 0.3766, + "step": 29457 + }, + { + "epoch": 0.6905419106055545, + "grad_norm": 3.423945903778076, + "learning_rate": 4.366871874404402e-05, + "loss": 0.3242, + "step": 29458 + }, + { + "epoch": 0.6905653521803595, + "grad_norm": 1.2132699489593506, + "learning_rate": 4.366263265527113e-05, + "loss": 0.1227, + "step": 29459 + }, + { + "epoch": 0.6905887937551645, + "grad_norm": 2.029738187789917, + "learning_rate": 4.365654687218641e-05, + "loss": 0.2283, + "step": 29460 + }, + { + "epoch": 0.6906122353299695, + "grad_norm": 2.118155002593994, + "learning_rate": 4.365046139482286e-05, + "loss": 0.2266, + "step": 29461 + }, + { + "epoch": 0.6906356769047745, + "grad_norm": 2.546950340270996, + "learning_rate": 4.364437622321345e-05, + "loss": 1.1033, + "step": 29462 + }, + { + "epoch": 0.6906591184795795, + "grad_norm": 5.514673233032227, + "learning_rate": 4.363829135739126e-05, + "loss": 0.3641, + "step": 29463 + }, + { + "epoch": 0.6906825600543844, + "grad_norm": 3.0179994106292725, + "learning_rate": 4.36322067973893e-05, + "loss": 0.3027, + "step": 29464 + }, + { + "epoch": 0.6907060016291895, + "grad_norm": 2.4060425758361816, + "learning_rate": 4.362612254324057e-05, + "loss": 0.2281, + "step": 29465 + }, + { + "epoch": 0.6907294432039944, + "grad_norm": 2.816408395767212, + "learning_rate": 4.362003859497805e-05, + "loss": 0.2108, + "step": 29466 + }, + { + "epoch": 0.6907528847787995, + "grad_norm": 2.069685220718384, + "learning_rate": 4.3613954952634815e-05, + "loss": 0.1155, + "step": 29467 + }, + { + "epoch": 0.6907763263536044, + "grad_norm": 2.7216904163360596, + "learning_rate": 4.3607871616243846e-05, + "loss": 0.1165, + "step": 29468 + }, + { + "epoch": 0.6907997679284095, + "grad_norm": 1.3401944637298584, + "learning_rate": 4.3601788585838155e-05, + "loss": 0.2052, + "step": 29469 + }, + { + "epoch": 0.6908232095032144, + "grad_norm": 0.8034856915473938, + "learning_rate": 4.3595705861450756e-05, + "loss": 0.0712, + "step": 29470 + }, + { + "epoch": 0.6908466510780195, + "grad_norm": 2.109686851501465, + "learning_rate": 4.3589623443114624e-05, + "loss": 0.0482, + "step": 29471 + }, + { + "epoch": 0.6908700926528244, + "grad_norm": 0.26927128434181213, + "learning_rate": 4.358354133086275e-05, + "loss": 0.0245, + "step": 29472 + }, + { + "epoch": 0.6908935342276294, + "grad_norm": 4.1626386642456055, + "learning_rate": 4.3577459524728205e-05, + "loss": 0.3935, + "step": 29473 + }, + { + "epoch": 0.6909169758024344, + "grad_norm": 4.275102138519287, + "learning_rate": 4.357137802474396e-05, + "loss": 0.4318, + "step": 29474 + }, + { + "epoch": 0.6909404173772394, + "grad_norm": 0.4606214463710785, + "learning_rate": 4.356529683094296e-05, + "loss": 0.0381, + "step": 29475 + }, + { + "epoch": 0.6909638589520444, + "grad_norm": 4.089170932769775, + "learning_rate": 4.3559215943358275e-05, + "loss": 0.1859, + "step": 29476 + }, + { + "epoch": 0.6909873005268494, + "grad_norm": 0.7299501299858093, + "learning_rate": 4.355313536202288e-05, + "loss": 0.0759, + "step": 29477 + }, + { + "epoch": 0.6910107421016544, + "grad_norm": 3.0606443881988525, + "learning_rate": 4.354705508696976e-05, + "loss": 0.2224, + "step": 29478 + }, + { + "epoch": 0.6910341836764594, + "grad_norm": 1.2589598894119263, + "learning_rate": 4.354097511823186e-05, + "loss": 0.1055, + "step": 29479 + }, + { + "epoch": 0.6910576252512644, + "grad_norm": 5.051412582397461, + "learning_rate": 4.3534895455842275e-05, + "loss": 0.3602, + "step": 29480 + }, + { + "epoch": 0.6910810668260694, + "grad_norm": 0.8624789714813232, + "learning_rate": 4.352881609983391e-05, + "loss": 0.0851, + "step": 29481 + }, + { + "epoch": 0.6911045084008743, + "grad_norm": 2.548421859741211, + "learning_rate": 4.352273705023979e-05, + "loss": 0.1492, + "step": 29482 + }, + { + "epoch": 0.6911279499756794, + "grad_norm": 2.4151384830474854, + "learning_rate": 4.351665830709288e-05, + "loss": 0.2034, + "step": 29483 + }, + { + "epoch": 0.6911513915504843, + "grad_norm": 3.599370002746582, + "learning_rate": 4.3510579870426146e-05, + "loss": 0.386, + "step": 29484 + }, + { + "epoch": 0.6911748331252894, + "grad_norm": 4.888034343719482, + "learning_rate": 4.3504501740272616e-05, + "loss": 0.3305, + "step": 29485 + }, + { + "epoch": 0.6911982747000943, + "grad_norm": 3.202158212661743, + "learning_rate": 4.349842391666526e-05, + "loss": 0.2878, + "step": 29486 + }, + { + "epoch": 0.6912217162748994, + "grad_norm": 1.7287509441375732, + "learning_rate": 4.3492346399637005e-05, + "loss": 0.0584, + "step": 29487 + }, + { + "epoch": 0.6912451578497043, + "grad_norm": 1.9152907133102417, + "learning_rate": 4.3486269189220905e-05, + "loss": 0.1765, + "step": 29488 + }, + { + "epoch": 0.6912685994245094, + "grad_norm": 1.363355040550232, + "learning_rate": 4.3480192285449896e-05, + "loss": 0.0783, + "step": 29489 + }, + { + "epoch": 0.6912920409993143, + "grad_norm": 2.680000066757202, + "learning_rate": 4.347411568835694e-05, + "loss": 1.4603, + "step": 29490 + }, + { + "epoch": 0.6913154825741193, + "grad_norm": 3.082805871963501, + "learning_rate": 4.3468039397975e-05, + "loss": 1.679, + "step": 29491 + }, + { + "epoch": 0.6913389241489243, + "grad_norm": 3.39359450340271, + "learning_rate": 4.346196341433711e-05, + "loss": 0.3698, + "step": 29492 + }, + { + "epoch": 0.6913623657237293, + "grad_norm": 0.7669510841369629, + "learning_rate": 4.345588773747618e-05, + "loss": 0.1004, + "step": 29493 + }, + { + "epoch": 0.6913858072985343, + "grad_norm": 1.8757655620574951, + "learning_rate": 4.3449812367425205e-05, + "loss": 0.1158, + "step": 29494 + }, + { + "epoch": 0.6914092488733393, + "grad_norm": 1.4231228828430176, + "learning_rate": 4.344373730421713e-05, + "loss": 0.1453, + "step": 29495 + }, + { + "epoch": 0.6914326904481443, + "grad_norm": 1.7369556427001953, + "learning_rate": 4.343766254788493e-05, + "loss": 0.2312, + "step": 29496 + }, + { + "epoch": 0.6914561320229493, + "grad_norm": 2.2554659843444824, + "learning_rate": 4.343158809846153e-05, + "loss": 0.2205, + "step": 29497 + }, + { + "epoch": 0.6914795735977542, + "grad_norm": 2.184689521789551, + "learning_rate": 4.3425513955979946e-05, + "loss": 0.1534, + "step": 29498 + }, + { + "epoch": 0.6915030151725593, + "grad_norm": 4.120988845825195, + "learning_rate": 4.3419440120473085e-05, + "loss": 0.3118, + "step": 29499 + }, + { + "epoch": 0.6915264567473642, + "grad_norm": 3.0243847370147705, + "learning_rate": 4.341336659197398e-05, + "loss": 0.112, + "step": 29500 + }, + { + "epoch": 0.6915498983221693, + "grad_norm": 1.2459484338760376, + "learning_rate": 4.3407293370515536e-05, + "loss": 0.1018, + "step": 29501 + }, + { + "epoch": 0.6915733398969742, + "grad_norm": 0.7594524621963501, + "learning_rate": 4.34012204561307e-05, + "loss": 0.0423, + "step": 29502 + }, + { + "epoch": 0.6915967814717793, + "grad_norm": 3.2095935344696045, + "learning_rate": 4.339514784885244e-05, + "loss": 0.2258, + "step": 29503 + }, + { + "epoch": 0.6916202230465842, + "grad_norm": 1.3662694692611694, + "learning_rate": 4.3389075548713655e-05, + "loss": 0.1517, + "step": 29504 + }, + { + "epoch": 0.6916436646213893, + "grad_norm": 2.9068684577941895, + "learning_rate": 4.338300355574738e-05, + "loss": 0.234, + "step": 29505 + }, + { + "epoch": 0.6916671061961943, + "grad_norm": 3.1335394382476807, + "learning_rate": 4.337693186998652e-05, + "loss": 0.209, + "step": 29506 + }, + { + "epoch": 0.6916905477709993, + "grad_norm": 2.148637294769287, + "learning_rate": 4.337086049146402e-05, + "loss": 0.2299, + "step": 29507 + }, + { + "epoch": 0.6917139893458043, + "grad_norm": 1.703261375427246, + "learning_rate": 4.336478942021283e-05, + "loss": 0.1093, + "step": 29508 + }, + { + "epoch": 0.6917374309206092, + "grad_norm": 1.945312738418579, + "learning_rate": 4.335871865626587e-05, + "loss": 0.1277, + "step": 29509 + }, + { + "epoch": 0.6917608724954143, + "grad_norm": 5.468430995941162, + "learning_rate": 4.335264819965607e-05, + "loss": 0.2881, + "step": 29510 + }, + { + "epoch": 0.6917843140702192, + "grad_norm": 1.5961393117904663, + "learning_rate": 4.3346578050416386e-05, + "loss": 0.2079, + "step": 29511 + }, + { + "epoch": 0.6918077556450243, + "grad_norm": 3.9948530197143555, + "learning_rate": 4.33405082085798e-05, + "loss": 0.3021, + "step": 29512 + }, + { + "epoch": 0.6918311972198292, + "grad_norm": 1.3171449899673462, + "learning_rate": 4.3334438674179225e-05, + "loss": 0.0476, + "step": 29513 + }, + { + "epoch": 0.6918546387946343, + "grad_norm": 0.8974043726921082, + "learning_rate": 4.3328369447247564e-05, + "loss": 0.1073, + "step": 29514 + }, + { + "epoch": 0.6918780803694392, + "grad_norm": 3.610175609588623, + "learning_rate": 4.332230052781776e-05, + "loss": 0.2761, + "step": 29515 + }, + { + "epoch": 0.6919015219442443, + "grad_norm": 1.0820082426071167, + "learning_rate": 4.3316231915922755e-05, + "loss": 0.0644, + "step": 29516 + }, + { + "epoch": 0.6919249635190492, + "grad_norm": 0.9253113865852356, + "learning_rate": 4.331016361159543e-05, + "loss": 0.1143, + "step": 29517 + }, + { + "epoch": 0.6919484050938542, + "grad_norm": 0.7490279078483582, + "learning_rate": 4.330409561486879e-05, + "loss": 0.0271, + "step": 29518 + }, + { + "epoch": 0.6919718466686592, + "grad_norm": 1.302828073501587, + "learning_rate": 4.329802792577572e-05, + "loss": 0.0531, + "step": 29519 + }, + { + "epoch": 0.6919952882434642, + "grad_norm": 2.901561975479126, + "learning_rate": 4.329196054434914e-05, + "loss": 0.2939, + "step": 29520 + }, + { + "epoch": 0.6920187298182692, + "grad_norm": 2.2753679752349854, + "learning_rate": 4.328589347062199e-05, + "loss": 0.2194, + "step": 29521 + }, + { + "epoch": 0.6920421713930742, + "grad_norm": 3.589625120162964, + "learning_rate": 4.3279826704627135e-05, + "loss": 0.2912, + "step": 29522 + }, + { + "epoch": 0.6920656129678792, + "grad_norm": 0.311493456363678, + "learning_rate": 4.327376024639753e-05, + "loss": 0.0263, + "step": 29523 + }, + { + "epoch": 0.6920890545426842, + "grad_norm": 1.0804463624954224, + "learning_rate": 4.326769409596614e-05, + "loss": 0.0981, + "step": 29524 + }, + { + "epoch": 0.6921124961174892, + "grad_norm": 3.095620632171631, + "learning_rate": 4.3261628253365826e-05, + "loss": 0.1955, + "step": 29525 + }, + { + "epoch": 0.6921359376922942, + "grad_norm": 2.5512478351593018, + "learning_rate": 4.325556271862952e-05, + "loss": 0.2531, + "step": 29526 + }, + { + "epoch": 0.6921593792670991, + "grad_norm": 4.176750659942627, + "learning_rate": 4.3249497491790126e-05, + "loss": 0.3472, + "step": 29527 + }, + { + "epoch": 0.6921828208419042, + "grad_norm": 1.277087926864624, + "learning_rate": 4.324343257288055e-05, + "loss": 0.0542, + "step": 29528 + }, + { + "epoch": 0.6922062624167091, + "grad_norm": 1.4760500192642212, + "learning_rate": 4.3237367961933675e-05, + "loss": 0.1434, + "step": 29529 + }, + { + "epoch": 0.6922297039915142, + "grad_norm": 3.916322708129883, + "learning_rate": 4.323130365898247e-05, + "loss": 0.3397, + "step": 29530 + }, + { + "epoch": 0.6922531455663191, + "grad_norm": 2.225597620010376, + "learning_rate": 4.3225239664059805e-05, + "loss": 0.1647, + "step": 29531 + }, + { + "epoch": 0.6922765871411242, + "grad_norm": 1.8905428647994995, + "learning_rate": 4.321917597719859e-05, + "loss": 0.0888, + "step": 29532 + }, + { + "epoch": 0.6923000287159291, + "grad_norm": 3.0499157905578613, + "learning_rate": 4.3213112598431726e-05, + "loss": 0.2388, + "step": 29533 + }, + { + "epoch": 0.6923234702907342, + "grad_norm": 2.1781904697418213, + "learning_rate": 4.32070495277921e-05, + "loss": 0.2518, + "step": 29534 + }, + { + "epoch": 0.6923469118655391, + "grad_norm": 0.6696667671203613, + "learning_rate": 4.320098676531259e-05, + "loss": 0.0651, + "step": 29535 + }, + { + "epoch": 0.6923703534403441, + "grad_norm": 0.5584152340888977, + "learning_rate": 4.319492431102612e-05, + "loss": 0.0215, + "step": 29536 + }, + { + "epoch": 0.6923937950151491, + "grad_norm": 3.3791606426239014, + "learning_rate": 4.318886216496563e-05, + "loss": 0.3587, + "step": 29537 + }, + { + "epoch": 0.6924172365899541, + "grad_norm": 0.5321511030197144, + "learning_rate": 4.3182800327163965e-05, + "loss": 0.0366, + "step": 29538 + }, + { + "epoch": 0.6924406781647591, + "grad_norm": 4.430440902709961, + "learning_rate": 4.317673879765402e-05, + "loss": 0.2997, + "step": 29539 + }, + { + "epoch": 0.6924641197395641, + "grad_norm": 4.016301155090332, + "learning_rate": 4.3170677576468696e-05, + "loss": 0.1667, + "step": 29540 + }, + { + "epoch": 0.6924875613143691, + "grad_norm": 2.6793718338012695, + "learning_rate": 4.3164616663640875e-05, + "loss": 0.225, + "step": 29541 + }, + { + "epoch": 0.6925110028891741, + "grad_norm": 3.3612732887268066, + "learning_rate": 4.3158556059203405e-05, + "loss": 0.2201, + "step": 29542 + }, + { + "epoch": 0.692534444463979, + "grad_norm": 3.013113260269165, + "learning_rate": 4.315249576318924e-05, + "loss": 0.3102, + "step": 29543 + }, + { + "epoch": 0.6925578860387841, + "grad_norm": 0.7148602604866028, + "learning_rate": 4.3146435775631234e-05, + "loss": 0.0657, + "step": 29544 + }, + { + "epoch": 0.692581327613589, + "grad_norm": 2.7876970767974854, + "learning_rate": 4.314037609656226e-05, + "loss": 1.5935, + "step": 29545 + }, + { + "epoch": 0.6926047691883941, + "grad_norm": 1.0201237201690674, + "learning_rate": 4.313431672601521e-05, + "loss": 0.1207, + "step": 29546 + }, + { + "epoch": 0.692628210763199, + "grad_norm": 0.899649977684021, + "learning_rate": 4.312825766402292e-05, + "loss": 0.068, + "step": 29547 + }, + { + "epoch": 0.6926516523380041, + "grad_norm": 2.5405406951904297, + "learning_rate": 4.312219891061834e-05, + "loss": 0.2017, + "step": 29548 + }, + { + "epoch": 0.692675093912809, + "grad_norm": 2.637742757797241, + "learning_rate": 4.311614046583427e-05, + "loss": 0.2179, + "step": 29549 + }, + { + "epoch": 0.6926985354876141, + "grad_norm": 2.7354907989501953, + "learning_rate": 4.311008232970366e-05, + "loss": 0.1449, + "step": 29550 + }, + { + "epoch": 0.692721977062419, + "grad_norm": 1.0930062532424927, + "learning_rate": 4.310402450225933e-05, + "loss": 0.1022, + "step": 29551 + }, + { + "epoch": 0.692745418637224, + "grad_norm": 2.4787514209747314, + "learning_rate": 4.309796698353418e-05, + "loss": 0.2245, + "step": 29552 + }, + { + "epoch": 0.692768860212029, + "grad_norm": 3.6206908226013184, + "learning_rate": 4.3091909773561056e-05, + "loss": 0.2849, + "step": 29553 + }, + { + "epoch": 0.692792301786834, + "grad_norm": 0.2508280575275421, + "learning_rate": 4.308585287237279e-05, + "loss": 0.02, + "step": 29554 + }, + { + "epoch": 0.692815743361639, + "grad_norm": 2.9826180934906006, + "learning_rate": 4.307979628000233e-05, + "loss": 0.2329, + "step": 29555 + }, + { + "epoch": 0.692839184936444, + "grad_norm": 2.5528640747070312, + "learning_rate": 4.3073739996482486e-05, + "loss": 1.0808, + "step": 29556 + }, + { + "epoch": 0.6928626265112491, + "grad_norm": 0.5995289087295532, + "learning_rate": 4.3067684021846136e-05, + "loss": 0.0309, + "step": 29557 + }, + { + "epoch": 0.692886068086054, + "grad_norm": 3.1134445667266846, + "learning_rate": 4.3061628356126124e-05, + "loss": 0.2342, + "step": 29558 + }, + { + "epoch": 0.6929095096608591, + "grad_norm": 2.453070640563965, + "learning_rate": 4.3055572999355334e-05, + "loss": 0.0793, + "step": 29559 + }, + { + "epoch": 0.692932951235664, + "grad_norm": 1.5027058124542236, + "learning_rate": 4.304951795156655e-05, + "loss": 0.1766, + "step": 29560 + }, + { + "epoch": 0.6929563928104691, + "grad_norm": 0.6924612522125244, + "learning_rate": 4.30434632127927e-05, + "loss": 0.0527, + "step": 29561 + }, + { + "epoch": 0.692979834385274, + "grad_norm": 2.401873826980591, + "learning_rate": 4.303740878306664e-05, + "loss": 0.1756, + "step": 29562 + }, + { + "epoch": 0.693003275960079, + "grad_norm": 0.4654177129268646, + "learning_rate": 4.303135466242122e-05, + "loss": 0.0229, + "step": 29563 + }, + { + "epoch": 0.693026717534884, + "grad_norm": 2.8503177165985107, + "learning_rate": 4.302530085088926e-05, + "loss": 0.2871, + "step": 29564 + }, + { + "epoch": 0.693050159109689, + "grad_norm": 3.2022626399993896, + "learning_rate": 4.301924734850362e-05, + "loss": 0.2228, + "step": 29565 + }, + { + "epoch": 0.693073600684494, + "grad_norm": 1.3331581354141235, + "learning_rate": 4.3013194155297155e-05, + "loss": 0.0992, + "step": 29566 + }, + { + "epoch": 0.693097042259299, + "grad_norm": 3.0128591060638428, + "learning_rate": 4.3007141271302656e-05, + "loss": 0.3333, + "step": 29567 + }, + { + "epoch": 0.693120483834104, + "grad_norm": 0.9225651621818542, + "learning_rate": 4.300108869655305e-05, + "loss": 0.0692, + "step": 29568 + }, + { + "epoch": 0.693143925408909, + "grad_norm": 1.8894301652908325, + "learning_rate": 4.299503643108114e-05, + "loss": 1.9008, + "step": 29569 + }, + { + "epoch": 0.693167366983714, + "grad_norm": 2.9002275466918945, + "learning_rate": 4.298898447491976e-05, + "loss": 0.1684, + "step": 29570 + }, + { + "epoch": 0.693190808558519, + "grad_norm": 0.6194493770599365, + "learning_rate": 4.298293282810175e-05, + "loss": 0.046, + "step": 29571 + }, + { + "epoch": 0.6932142501333239, + "grad_norm": 3.57956600189209, + "learning_rate": 4.297688149065993e-05, + "loss": 0.3041, + "step": 29572 + }, + { + "epoch": 0.693237691708129, + "grad_norm": 2.4344987869262695, + "learning_rate": 4.297083046262718e-05, + "loss": 0.1286, + "step": 29573 + }, + { + "epoch": 0.6932611332829339, + "grad_norm": 1.0245082378387451, + "learning_rate": 4.296477974403626e-05, + "loss": 0.1507, + "step": 29574 + }, + { + "epoch": 0.693284574857739, + "grad_norm": 5.131880283355713, + "learning_rate": 4.29587293349201e-05, + "loss": 0.4175, + "step": 29575 + }, + { + "epoch": 0.6933080164325439, + "grad_norm": 3.2242469787597656, + "learning_rate": 4.2952679235311474e-05, + "loss": 1.3633, + "step": 29576 + }, + { + "epoch": 0.693331458007349, + "grad_norm": 0.8005566596984863, + "learning_rate": 4.294662944524323e-05, + "loss": 0.0754, + "step": 29577 + }, + { + "epoch": 0.6933548995821539, + "grad_norm": 1.282263994216919, + "learning_rate": 4.294057996474815e-05, + "loss": 0.0731, + "step": 29578 + }, + { + "epoch": 0.693378341156959, + "grad_norm": 1.9013149738311768, + "learning_rate": 4.293453079385911e-05, + "loss": 0.1396, + "step": 29579 + }, + { + "epoch": 0.6934017827317639, + "grad_norm": 5.908544063568115, + "learning_rate": 4.2928481932608864e-05, + "loss": 0.4368, + "step": 29580 + }, + { + "epoch": 0.6934252243065689, + "grad_norm": 2.0751779079437256, + "learning_rate": 4.2922433381030315e-05, + "loss": 0.1422, + "step": 29581 + }, + { + "epoch": 0.6934486658813739, + "grad_norm": 1.9748115539550781, + "learning_rate": 4.291638513915626e-05, + "loss": 0.1652, + "step": 29582 + }, + { + "epoch": 0.6934721074561789, + "grad_norm": 2.186539649963379, + "learning_rate": 4.291033720701949e-05, + "loss": 1.0382, + "step": 29583 + }, + { + "epoch": 0.6934955490309839, + "grad_norm": 3.370729923248291, + "learning_rate": 4.29042895846528e-05, + "loss": 0.4037, + "step": 29584 + }, + { + "epoch": 0.6935189906057889, + "grad_norm": 2.315704822540283, + "learning_rate": 4.2898242272089084e-05, + "loss": 1.4582, + "step": 29585 + }, + { + "epoch": 0.6935424321805939, + "grad_norm": 4.228597640991211, + "learning_rate": 4.2892195269361114e-05, + "loss": 0.3519, + "step": 29586 + }, + { + "epoch": 0.6935658737553989, + "grad_norm": 2.3050973415374756, + "learning_rate": 4.288614857650165e-05, + "loss": 0.1577, + "step": 29587 + }, + { + "epoch": 0.6935893153302038, + "grad_norm": 4.298289775848389, + "learning_rate": 4.2880102193543595e-05, + "loss": 0.425, + "step": 29588 + }, + { + "epoch": 0.6936127569050089, + "grad_norm": 4.329049587249756, + "learning_rate": 4.28740561205197e-05, + "loss": 0.4183, + "step": 29589 + }, + { + "epoch": 0.6936361984798138, + "grad_norm": 3.4447975158691406, + "learning_rate": 4.2868010357462795e-05, + "loss": 0.1866, + "step": 29590 + }, + { + "epoch": 0.6936596400546189, + "grad_norm": 1.6136722564697266, + "learning_rate": 4.286196490440567e-05, + "loss": 0.1332, + "step": 29591 + }, + { + "epoch": 0.6936830816294238, + "grad_norm": 1.63275945186615, + "learning_rate": 4.2855919761381094e-05, + "loss": 0.6567, + "step": 29592 + }, + { + "epoch": 0.6937065232042289, + "grad_norm": 3.3803927898406982, + "learning_rate": 4.284987492842194e-05, + "loss": 0.2982, + "step": 29593 + }, + { + "epoch": 0.6937299647790338, + "grad_norm": 1.3320996761322021, + "learning_rate": 4.284383040556097e-05, + "loss": 0.1225, + "step": 29594 + }, + { + "epoch": 0.6937534063538389, + "grad_norm": 1.7658590078353882, + "learning_rate": 4.2837786192830974e-05, + "loss": 0.107, + "step": 29595 + }, + { + "epoch": 0.6937768479286438, + "grad_norm": 3.1337649822235107, + "learning_rate": 4.2831742290264733e-05, + "loss": 0.3142, + "step": 29596 + }, + { + "epoch": 0.6938002895034489, + "grad_norm": 4.223437309265137, + "learning_rate": 4.28256986978951e-05, + "loss": 0.4261, + "step": 29597 + }, + { + "epoch": 0.6938237310782538, + "grad_norm": 1.9833709001541138, + "learning_rate": 4.2819655415754836e-05, + "loss": 0.1817, + "step": 29598 + }, + { + "epoch": 0.6938471726530588, + "grad_norm": 2.4904794692993164, + "learning_rate": 4.2813612443876685e-05, + "loss": 0.1637, + "step": 29599 + }, + { + "epoch": 0.6938706142278638, + "grad_norm": 3.156202793121338, + "learning_rate": 4.280756978229352e-05, + "loss": 0.7244, + "step": 29600 + }, + { + "epoch": 0.6938940558026688, + "grad_norm": 3.465261459350586, + "learning_rate": 4.280152743103809e-05, + "loss": 0.3117, + "step": 29601 + }, + { + "epoch": 0.6939174973774738, + "grad_norm": 3.2640323638916016, + "learning_rate": 4.2795485390143185e-05, + "loss": 0.2229, + "step": 29602 + }, + { + "epoch": 0.6939409389522788, + "grad_norm": 3.8013014793395996, + "learning_rate": 4.278944365964158e-05, + "loss": 0.2789, + "step": 29603 + }, + { + "epoch": 0.6939643805270838, + "grad_norm": 1.2204509973526, + "learning_rate": 4.278340223956607e-05, + "loss": 0.0801, + "step": 29604 + }, + { + "epoch": 0.6939878221018888, + "grad_norm": 4.242551326751709, + "learning_rate": 4.277736112994939e-05, + "loss": 0.3772, + "step": 29605 + }, + { + "epoch": 0.6940112636766937, + "grad_norm": 2.4089505672454834, + "learning_rate": 4.277132033082438e-05, + "loss": 0.2045, + "step": 29606 + }, + { + "epoch": 0.6940347052514988, + "grad_norm": 2.329813003540039, + "learning_rate": 4.2765279842223804e-05, + "loss": 0.1647, + "step": 29607 + }, + { + "epoch": 0.6940581468263038, + "grad_norm": 2.56178879737854, + "learning_rate": 4.2759239664180396e-05, + "loss": 0.3242, + "step": 29608 + }, + { + "epoch": 0.6940815884011088, + "grad_norm": 3.9838292598724365, + "learning_rate": 4.2753199796726996e-05, + "loss": 0.4972, + "step": 29609 + }, + { + "epoch": 0.6941050299759138, + "grad_norm": 4.65011739730835, + "learning_rate": 4.2747160239896346e-05, + "loss": 0.447, + "step": 29610 + }, + { + "epoch": 0.6941284715507188, + "grad_norm": 1.2183438539505005, + "learning_rate": 4.2741120993721216e-05, + "loss": 0.0867, + "step": 29611 + }, + { + "epoch": 0.6941519131255238, + "grad_norm": 4.9244513511657715, + "learning_rate": 4.273508205823433e-05, + "loss": 0.285, + "step": 29612 + }, + { + "epoch": 0.6941753547003288, + "grad_norm": 1.6678069829940796, + "learning_rate": 4.272904343346854e-05, + "loss": 0.2768, + "step": 29613 + }, + { + "epoch": 0.6941987962751338, + "grad_norm": 1.7776851654052734, + "learning_rate": 4.272300511945657e-05, + "loss": 0.1607, + "step": 29614 + }, + { + "epoch": 0.6942222378499388, + "grad_norm": 2.0614144802093506, + "learning_rate": 4.271696711623119e-05, + "loss": 0.1956, + "step": 29615 + }, + { + "epoch": 0.6942456794247438, + "grad_norm": 3.1470396518707275, + "learning_rate": 4.271092942382514e-05, + "loss": 0.4817, + "step": 29616 + }, + { + "epoch": 0.6942691209995487, + "grad_norm": 2.1788218021392822, + "learning_rate": 4.2704892042271216e-05, + "loss": 0.1725, + "step": 29617 + }, + { + "epoch": 0.6942925625743538, + "grad_norm": 5.984906196594238, + "learning_rate": 4.269885497160211e-05, + "loss": 0.7897, + "step": 29618 + }, + { + "epoch": 0.6943160041491587, + "grad_norm": 3.729688882827759, + "learning_rate": 4.269281821185067e-05, + "loss": 0.3694, + "step": 29619 + }, + { + "epoch": 0.6943394457239638, + "grad_norm": 0.2464877814054489, + "learning_rate": 4.268678176304958e-05, + "loss": 0.0155, + "step": 29620 + }, + { + "epoch": 0.6943628872987687, + "grad_norm": 3.113070249557495, + "learning_rate": 4.2680745625231653e-05, + "loss": 0.1953, + "step": 29621 + }, + { + "epoch": 0.6943863288735738, + "grad_norm": 1.1237678527832031, + "learning_rate": 4.267470979842961e-05, + "loss": 0.1049, + "step": 29622 + }, + { + "epoch": 0.6944097704483787, + "grad_norm": 3.4621241092681885, + "learning_rate": 4.26686742826762e-05, + "loss": 0.2547, + "step": 29623 + }, + { + "epoch": 0.6944332120231838, + "grad_norm": 1.0710771083831787, + "learning_rate": 4.2662639078004186e-05, + "loss": 0.0617, + "step": 29624 + }, + { + "epoch": 0.6944566535979887, + "grad_norm": 3.724795341491699, + "learning_rate": 4.265660418444627e-05, + "loss": 0.3128, + "step": 29625 + }, + { + "epoch": 0.6944800951727937, + "grad_norm": 0.8488214612007141, + "learning_rate": 4.2650569602035264e-05, + "loss": 0.0451, + "step": 29626 + }, + { + "epoch": 0.6945035367475987, + "grad_norm": 3.323580265045166, + "learning_rate": 4.264453533080387e-05, + "loss": 0.2892, + "step": 29627 + }, + { + "epoch": 0.6945269783224037, + "grad_norm": 1.8503680229187012, + "learning_rate": 4.263850137078484e-05, + "loss": 0.1586, + "step": 29628 + }, + { + "epoch": 0.6945504198972087, + "grad_norm": 2.3817813396453857, + "learning_rate": 4.263246772201093e-05, + "loss": 0.2158, + "step": 29629 + }, + { + "epoch": 0.6945738614720137, + "grad_norm": 3.6262223720550537, + "learning_rate": 4.262643438451482e-05, + "loss": 0.3667, + "step": 29630 + }, + { + "epoch": 0.6945973030468187, + "grad_norm": 0.9324565529823303, + "learning_rate": 4.262040135832932e-05, + "loss": 0.0964, + "step": 29631 + }, + { + "epoch": 0.6946207446216237, + "grad_norm": 3.05364727973938, + "learning_rate": 4.2614368643487144e-05, + "loss": 0.2702, + "step": 29632 + }, + { + "epoch": 0.6946441861964286, + "grad_norm": 4.664748191833496, + "learning_rate": 4.260833624002098e-05, + "loss": 0.4425, + "step": 29633 + }, + { + "epoch": 0.6946676277712337, + "grad_norm": 1.2643940448760986, + "learning_rate": 4.260230414796363e-05, + "loss": 0.0927, + "step": 29634 + }, + { + "epoch": 0.6946910693460386, + "grad_norm": 2.789121389389038, + "learning_rate": 4.259627236734779e-05, + "loss": 0.2818, + "step": 29635 + }, + { + "epoch": 0.6947145109208437, + "grad_norm": 2.282750368118286, + "learning_rate": 4.25902408982062e-05, + "loss": 0.2136, + "step": 29636 + }, + { + "epoch": 0.6947379524956486, + "grad_norm": 0.3938525319099426, + "learning_rate": 4.2584209740571534e-05, + "loss": 0.0303, + "step": 29637 + }, + { + "epoch": 0.6947613940704537, + "grad_norm": 2.3075475692749023, + "learning_rate": 4.2578178894476596e-05, + "loss": 0.0991, + "step": 29638 + }, + { + "epoch": 0.6947848356452586, + "grad_norm": 2.0083160400390625, + "learning_rate": 4.257214835995407e-05, + "loss": 1.222, + "step": 29639 + }, + { + "epoch": 0.6948082772200637, + "grad_norm": 1.4149757623672485, + "learning_rate": 4.256611813703669e-05, + "loss": 0.1231, + "step": 29640 + }, + { + "epoch": 0.6948317187948686, + "grad_norm": 3.001033306121826, + "learning_rate": 4.2560088225757164e-05, + "loss": 0.1825, + "step": 29641 + }, + { + "epoch": 0.6948551603696737, + "grad_norm": 1.9284223318099976, + "learning_rate": 4.255405862614822e-05, + "loss": 0.114, + "step": 29642 + }, + { + "epoch": 0.6948786019444786, + "grad_norm": 2.811901092529297, + "learning_rate": 4.254802933824252e-05, + "loss": 0.332, + "step": 29643 + }, + { + "epoch": 0.6949020435192836, + "grad_norm": 0.8678759932518005, + "learning_rate": 4.254200036207287e-05, + "loss": 0.0684, + "step": 29644 + }, + { + "epoch": 0.6949254850940886, + "grad_norm": 4.34553337097168, + "learning_rate": 4.253597169767191e-05, + "loss": 0.3617, + "step": 29645 + }, + { + "epoch": 0.6949489266688936, + "grad_norm": 1.311264991760254, + "learning_rate": 4.252994334507242e-05, + "loss": 0.1213, + "step": 29646 + }, + { + "epoch": 0.6949723682436986, + "grad_norm": 2.3536715507507324, + "learning_rate": 4.252391530430707e-05, + "loss": 0.1333, + "step": 29647 + }, + { + "epoch": 0.6949958098185036, + "grad_norm": 0.9353229403495789, + "learning_rate": 4.251788757540857e-05, + "loss": 0.0766, + "step": 29648 + }, + { + "epoch": 0.6950192513933086, + "grad_norm": 2.764686107635498, + "learning_rate": 4.2511860158409634e-05, + "loss": 0.2126, + "step": 29649 + }, + { + "epoch": 0.6950426929681136, + "grad_norm": 1.1174410581588745, + "learning_rate": 4.2505833053342915e-05, + "loss": 0.1037, + "step": 29650 + }, + { + "epoch": 0.6950661345429185, + "grad_norm": 4.585814476013184, + "learning_rate": 4.24998062602412e-05, + "loss": 0.2747, + "step": 29651 + }, + { + "epoch": 0.6950895761177236, + "grad_norm": 0.882371723651886, + "learning_rate": 4.2493779779137164e-05, + "loss": 0.0987, + "step": 29652 + }, + { + "epoch": 0.6951130176925285, + "grad_norm": 0.5209161639213562, + "learning_rate": 4.248775361006349e-05, + "loss": 0.0577, + "step": 29653 + }, + { + "epoch": 0.6951364592673336, + "grad_norm": 3.958709478378296, + "learning_rate": 4.248172775305288e-05, + "loss": 0.1732, + "step": 29654 + }, + { + "epoch": 0.6951599008421385, + "grad_norm": 2.4568448066711426, + "learning_rate": 4.2475702208138034e-05, + "loss": 0.2221, + "step": 29655 + }, + { + "epoch": 0.6951833424169436, + "grad_norm": 1.2807197570800781, + "learning_rate": 4.2469676975351614e-05, + "loss": 0.1411, + "step": 29656 + }, + { + "epoch": 0.6952067839917485, + "grad_norm": 2.4460577964782715, + "learning_rate": 4.246365205472633e-05, + "loss": 1.5068, + "step": 29657 + }, + { + "epoch": 0.6952302255665536, + "grad_norm": 2.546541452407837, + "learning_rate": 4.2457627446294937e-05, + "loss": 0.1567, + "step": 29658 + }, + { + "epoch": 0.6952536671413586, + "grad_norm": 3.3666698932647705, + "learning_rate": 4.245160315009007e-05, + "loss": 0.26, + "step": 29659 + }, + { + "epoch": 0.6952771087161635, + "grad_norm": 2.4627630710601807, + "learning_rate": 4.244557916614444e-05, + "loss": 0.2964, + "step": 29660 + }, + { + "epoch": 0.6953005502909686, + "grad_norm": 3.560319423675537, + "learning_rate": 4.243955549449069e-05, + "loss": 0.3669, + "step": 29661 + }, + { + "epoch": 0.6953239918657735, + "grad_norm": 1.396634817123413, + "learning_rate": 4.2433532135161536e-05, + "loss": 0.186, + "step": 29662 + }, + { + "epoch": 0.6953474334405786, + "grad_norm": 2.427736759185791, + "learning_rate": 4.242750908818962e-05, + "loss": 0.1523, + "step": 29663 + }, + { + "epoch": 0.6953708750153835, + "grad_norm": 2.8569560050964355, + "learning_rate": 4.242148635360771e-05, + "loss": 0.2732, + "step": 29664 + }, + { + "epoch": 0.6953943165901886, + "grad_norm": 6.907983303070068, + "learning_rate": 4.241546393144842e-05, + "loss": 0.6399, + "step": 29665 + }, + { + "epoch": 0.6954177581649935, + "grad_norm": 3.3701186180114746, + "learning_rate": 4.2409441821744446e-05, + "loss": 0.2329, + "step": 29666 + }, + { + "epoch": 0.6954411997397986, + "grad_norm": 3.234131336212158, + "learning_rate": 4.240342002452845e-05, + "loss": 0.265, + "step": 29667 + }, + { + "epoch": 0.6954646413146035, + "grad_norm": 1.7388945817947388, + "learning_rate": 4.239739853983309e-05, + "loss": 0.1535, + "step": 29668 + }, + { + "epoch": 0.6954880828894086, + "grad_norm": 1.2862591743469238, + "learning_rate": 4.23913773676911e-05, + "loss": 0.1107, + "step": 29669 + }, + { + "epoch": 0.6955115244642135, + "grad_norm": 3.003044605255127, + "learning_rate": 4.2385356508135074e-05, + "loss": 0.2546, + "step": 29670 + }, + { + "epoch": 0.6955349660390185, + "grad_norm": 4.329277515411377, + "learning_rate": 4.2379335961197774e-05, + "loss": 0.3807, + "step": 29671 + }, + { + "epoch": 0.6955584076138235, + "grad_norm": 3.044503688812256, + "learning_rate": 4.23733157269118e-05, + "loss": 0.3127, + "step": 29672 + }, + { + "epoch": 0.6955818491886285, + "grad_norm": 3.320476531982422, + "learning_rate": 4.236729580530985e-05, + "loss": 0.1669, + "step": 29673 + }, + { + "epoch": 0.6956052907634335, + "grad_norm": 3.100644588470459, + "learning_rate": 4.2361276196424574e-05, + "loss": 0.2276, + "step": 29674 + }, + { + "epoch": 0.6956287323382385, + "grad_norm": 3.219837188720703, + "learning_rate": 4.235525690028859e-05, + "loss": 0.2526, + "step": 29675 + }, + { + "epoch": 0.6956521739130435, + "grad_norm": 1.2058002948760986, + "learning_rate": 4.234923791693464e-05, + "loss": 0.0809, + "step": 29676 + }, + { + "epoch": 0.6956756154878485, + "grad_norm": 4.750325679779053, + "learning_rate": 4.2343219246395335e-05, + "loss": 0.42, + "step": 29677 + }, + { + "epoch": 0.6956990570626534, + "grad_norm": 1.4801783561706543, + "learning_rate": 4.2337200888703355e-05, + "loss": 0.8458, + "step": 29678 + }, + { + "epoch": 0.6957224986374585, + "grad_norm": 3.1313281059265137, + "learning_rate": 4.2331182843891346e-05, + "loss": 0.1773, + "step": 29679 + }, + { + "epoch": 0.6957459402122634, + "grad_norm": 4.031566619873047, + "learning_rate": 4.232516511199196e-05, + "loss": 0.3091, + "step": 29680 + }, + { + "epoch": 0.6957693817870685, + "grad_norm": 1.3545386791229248, + "learning_rate": 4.2319147693037805e-05, + "loss": 0.0476, + "step": 29681 + }, + { + "epoch": 0.6957928233618734, + "grad_norm": 3.7859325408935547, + "learning_rate": 4.231313058706159e-05, + "loss": 0.2532, + "step": 29682 + }, + { + "epoch": 0.6958162649366785, + "grad_norm": 3.6214773654937744, + "learning_rate": 4.230711379409598e-05, + "loss": 0.3695, + "step": 29683 + }, + { + "epoch": 0.6958397065114834, + "grad_norm": 0.5823973417282104, + "learning_rate": 4.2301097314173586e-05, + "loss": 0.0884, + "step": 29684 + }, + { + "epoch": 0.6958631480862885, + "grad_norm": 1.1037794351577759, + "learning_rate": 4.229508114732708e-05, + "loss": 0.1289, + "step": 29685 + }, + { + "epoch": 0.6958865896610934, + "grad_norm": 2.773643732070923, + "learning_rate": 4.228906529358907e-05, + "loss": 0.1993, + "step": 29686 + }, + { + "epoch": 0.6959100312358985, + "grad_norm": 5.096100807189941, + "learning_rate": 4.228304975299221e-05, + "loss": 0.3043, + "step": 29687 + }, + { + "epoch": 0.6959334728107034, + "grad_norm": 2.420424699783325, + "learning_rate": 4.2277034525569126e-05, + "loss": 0.1415, + "step": 29688 + }, + { + "epoch": 0.6959569143855084, + "grad_norm": 2.1895852088928223, + "learning_rate": 4.22710196113525e-05, + "loss": 0.1715, + "step": 29689 + }, + { + "epoch": 0.6959803559603134, + "grad_norm": 1.2784159183502197, + "learning_rate": 4.226500501037495e-05, + "loss": 0.1429, + "step": 29690 + }, + { + "epoch": 0.6960037975351184, + "grad_norm": 2.4103105068206787, + "learning_rate": 4.22589907226691e-05, + "loss": 0.2325, + "step": 29691 + }, + { + "epoch": 0.6960272391099234, + "grad_norm": 2.5794641971588135, + "learning_rate": 4.225297674826759e-05, + "loss": 0.1395, + "step": 29692 + }, + { + "epoch": 0.6960506806847284, + "grad_norm": 3.7654430866241455, + "learning_rate": 4.224696308720302e-05, + "loss": 0.3315, + "step": 29693 + }, + { + "epoch": 0.6960741222595334, + "grad_norm": 4.574285507202148, + "learning_rate": 4.224094973950808e-05, + "loss": 0.7221, + "step": 29694 + }, + { + "epoch": 0.6960975638343384, + "grad_norm": 4.2497429847717285, + "learning_rate": 4.223493670521534e-05, + "loss": 0.4964, + "step": 29695 + }, + { + "epoch": 0.6961210054091433, + "grad_norm": 0.852124035358429, + "learning_rate": 4.222892398435751e-05, + "loss": 0.0622, + "step": 29696 + }, + { + "epoch": 0.6961444469839484, + "grad_norm": 1.793139100074768, + "learning_rate": 4.222291157696714e-05, + "loss": 0.1102, + "step": 29697 + }, + { + "epoch": 0.6961678885587533, + "grad_norm": 2.8542838096618652, + "learning_rate": 4.2216899483076886e-05, + "loss": 0.1821, + "step": 29698 + }, + { + "epoch": 0.6961913301335584, + "grad_norm": 0.998649537563324, + "learning_rate": 4.221088770271936e-05, + "loss": 0.0836, + "step": 29699 + }, + { + "epoch": 0.6962147717083633, + "grad_norm": 3.498429298400879, + "learning_rate": 4.2204876235927184e-05, + "loss": 0.2939, + "step": 29700 + }, + { + "epoch": 0.6962382132831684, + "grad_norm": 1.1784058809280396, + "learning_rate": 4.219886508273293e-05, + "loss": 0.1959, + "step": 29701 + }, + { + "epoch": 0.6962616548579733, + "grad_norm": 4.801083087921143, + "learning_rate": 4.21928542431693e-05, + "loss": 0.43, + "step": 29702 + }, + { + "epoch": 0.6962850964327784, + "grad_norm": 1.5926483869552612, + "learning_rate": 4.2186843717268874e-05, + "loss": 0.0861, + "step": 29703 + }, + { + "epoch": 0.6963085380075833, + "grad_norm": 2.5374677181243896, + "learning_rate": 4.218083350506426e-05, + "loss": 0.2022, + "step": 29704 + }, + { + "epoch": 0.6963319795823883, + "grad_norm": 2.0017051696777344, + "learning_rate": 4.2174823606588054e-05, + "loss": 0.1981, + "step": 29705 + }, + { + "epoch": 0.6963554211571933, + "grad_norm": 3.841052770614624, + "learning_rate": 4.216881402187286e-05, + "loss": 0.3559, + "step": 29706 + }, + { + "epoch": 0.6963788627319983, + "grad_norm": 2.379728317260742, + "learning_rate": 4.216280475095134e-05, + "loss": 0.1336, + "step": 29707 + }, + { + "epoch": 0.6964023043068033, + "grad_norm": 2.2300357818603516, + "learning_rate": 4.215679579385603e-05, + "loss": 0.1492, + "step": 29708 + }, + { + "epoch": 0.6964257458816083, + "grad_norm": 4.5320258140563965, + "learning_rate": 4.215078715061961e-05, + "loss": 0.1714, + "step": 29709 + }, + { + "epoch": 0.6964491874564134, + "grad_norm": 3.39058518409729, + "learning_rate": 4.214477882127463e-05, + "loss": 0.2681, + "step": 29710 + }, + { + "epoch": 0.6964726290312183, + "grad_norm": 2.900956869125366, + "learning_rate": 4.213877080585373e-05, + "loss": 0.2978, + "step": 29711 + }, + { + "epoch": 0.6964960706060234, + "grad_norm": 2.9358901977539062, + "learning_rate": 4.2132763104389475e-05, + "loss": 0.2578, + "step": 29712 + }, + { + "epoch": 0.6965195121808283, + "grad_norm": 0.9774271845817566, + "learning_rate": 4.2126755716914444e-05, + "loss": 0.0924, + "step": 29713 + }, + { + "epoch": 0.6965429537556334, + "grad_norm": 1.918555498123169, + "learning_rate": 4.2120748643461293e-05, + "loss": 0.141, + "step": 29714 + }, + { + "epoch": 0.6965663953304383, + "grad_norm": 0.48464059829711914, + "learning_rate": 4.211474188406259e-05, + "loss": 0.0415, + "step": 29715 + }, + { + "epoch": 0.6965898369052433, + "grad_norm": 2.0777668952941895, + "learning_rate": 4.2108735438750925e-05, + "loss": 1.2381, + "step": 29716 + }, + { + "epoch": 0.6966132784800483, + "grad_norm": 3.3991732597351074, + "learning_rate": 4.2102729307558895e-05, + "loss": 0.2008, + "step": 29717 + }, + { + "epoch": 0.6966367200548533, + "grad_norm": 2.696174383163452, + "learning_rate": 4.209672349051904e-05, + "loss": 0.1798, + "step": 29718 + }, + { + "epoch": 0.6966601616296583, + "grad_norm": 3.0284054279327393, + "learning_rate": 4.209071798766403e-05, + "loss": 0.2387, + "step": 29719 + }, + { + "epoch": 0.6966836032044633, + "grad_norm": 2.9557576179504395, + "learning_rate": 4.208471279902637e-05, + "loss": 0.1174, + "step": 29720 + }, + { + "epoch": 0.6967070447792683, + "grad_norm": 3.1458284854888916, + "learning_rate": 4.2078707924638736e-05, + "loss": 0.2187, + "step": 29721 + }, + { + "epoch": 0.6967304863540733, + "grad_norm": 0.5110946297645569, + "learning_rate": 4.207270336453365e-05, + "loss": 0.0471, + "step": 29722 + }, + { + "epoch": 0.6967539279288782, + "grad_norm": 3.8356714248657227, + "learning_rate": 4.206669911874371e-05, + "loss": 0.3838, + "step": 29723 + }, + { + "epoch": 0.6967773695036833, + "grad_norm": 1.9681339263916016, + "learning_rate": 4.206069518730149e-05, + "loss": 0.1708, + "step": 29724 + }, + { + "epoch": 0.6968008110784882, + "grad_norm": 1.2507643699645996, + "learning_rate": 4.2054691570239566e-05, + "loss": 0.1162, + "step": 29725 + }, + { + "epoch": 0.6968242526532933, + "grad_norm": 1.7426550388336182, + "learning_rate": 4.204868826759048e-05, + "loss": 0.0754, + "step": 29726 + }, + { + "epoch": 0.6968476942280982, + "grad_norm": 2.278717517852783, + "learning_rate": 4.2042685279386874e-05, + "loss": 0.1795, + "step": 29727 + }, + { + "epoch": 0.6968711358029033, + "grad_norm": 2.0209619998931885, + "learning_rate": 4.2036682605661294e-05, + "loss": 0.1114, + "step": 29728 + }, + { + "epoch": 0.6968945773777082, + "grad_norm": 3.6715927124023438, + "learning_rate": 4.20306802464463e-05, + "loss": 0.2782, + "step": 29729 + }, + { + "epoch": 0.6969180189525133, + "grad_norm": 2.6694154739379883, + "learning_rate": 4.2024678201774426e-05, + "loss": 0.1438, + "step": 29730 + }, + { + "epoch": 0.6969414605273182, + "grad_norm": 0.7538463473320007, + "learning_rate": 4.201867647167831e-05, + "loss": 0.0753, + "step": 29731 + }, + { + "epoch": 0.6969649021021233, + "grad_norm": 0.6486510038375854, + "learning_rate": 4.201267505619049e-05, + "loss": 0.0516, + "step": 29732 + }, + { + "epoch": 0.6969883436769282, + "grad_norm": 0.7087084650993347, + "learning_rate": 4.200667395534349e-05, + "loss": 0.045, + "step": 29733 + }, + { + "epoch": 0.6970117852517332, + "grad_norm": 0.8890050649642944, + "learning_rate": 4.200067316916995e-05, + "loss": 0.0684, + "step": 29734 + }, + { + "epoch": 0.6970352268265382, + "grad_norm": 1.2663195133209229, + "learning_rate": 4.1994672697702383e-05, + "loss": 0.1052, + "step": 29735 + }, + { + "epoch": 0.6970586684013432, + "grad_norm": 2.1101555824279785, + "learning_rate": 4.198867254097335e-05, + "loss": 1.1823, + "step": 29736 + }, + { + "epoch": 0.6970821099761482, + "grad_norm": 0.5297476053237915, + "learning_rate": 4.1982672699015414e-05, + "loss": 0.021, + "step": 29737 + }, + { + "epoch": 0.6971055515509532, + "grad_norm": 2.1239945888519287, + "learning_rate": 4.197667317186112e-05, + "loss": 0.2527, + "step": 29738 + }, + { + "epoch": 0.6971289931257582, + "grad_norm": 1.7371385097503662, + "learning_rate": 4.1970673959543006e-05, + "loss": 0.2928, + "step": 29739 + }, + { + "epoch": 0.6971524347005632, + "grad_norm": 2.112884759902954, + "learning_rate": 4.196467506209368e-05, + "loss": 1.5088, + "step": 29740 + }, + { + "epoch": 0.6971758762753681, + "grad_norm": 2.331660509109497, + "learning_rate": 4.195867647954565e-05, + "loss": 0.1877, + "step": 29741 + }, + { + "epoch": 0.6971993178501732, + "grad_norm": 2.8198530673980713, + "learning_rate": 4.1952678211931454e-05, + "loss": 0.3223, + "step": 29742 + }, + { + "epoch": 0.6972227594249781, + "grad_norm": 2.729161024093628, + "learning_rate": 4.194668025928368e-05, + "loss": 1.9558, + "step": 29743 + }, + { + "epoch": 0.6972462009997832, + "grad_norm": 1.7218122482299805, + "learning_rate": 4.194068262163485e-05, + "loss": 0.1495, + "step": 29744 + }, + { + "epoch": 0.6972696425745881, + "grad_norm": 1.984317660331726, + "learning_rate": 4.193468529901752e-05, + "loss": 0.1793, + "step": 29745 + }, + { + "epoch": 0.6972930841493932, + "grad_norm": 0.5492504239082336, + "learning_rate": 4.192868829146417e-05, + "loss": 0.0499, + "step": 29746 + }, + { + "epoch": 0.6973165257241981, + "grad_norm": 4.020596981048584, + "learning_rate": 4.1922691599007436e-05, + "loss": 0.3913, + "step": 29747 + }, + { + "epoch": 0.6973399672990032, + "grad_norm": 3.3477442264556885, + "learning_rate": 4.19166952216798e-05, + "loss": 0.5189, + "step": 29748 + }, + { + "epoch": 0.6973634088738081, + "grad_norm": 3.4292280673980713, + "learning_rate": 4.191069915951382e-05, + "loss": 0.2481, + "step": 29749 + }, + { + "epoch": 0.6973868504486131, + "grad_norm": 4.296149253845215, + "learning_rate": 4.1904703412542024e-05, + "loss": 0.2907, + "step": 29750 + }, + { + "epoch": 0.6974102920234181, + "grad_norm": 3.030734062194824, + "learning_rate": 4.189870798079689e-05, + "loss": 0.3351, + "step": 29751 + }, + { + "epoch": 0.6974337335982231, + "grad_norm": 0.22484472393989563, + "learning_rate": 4.189271286431104e-05, + "loss": 0.014, + "step": 29752 + }, + { + "epoch": 0.6974571751730281, + "grad_norm": 1.9828033447265625, + "learning_rate": 4.188671806311697e-05, + "loss": 0.1625, + "step": 29753 + }, + { + "epoch": 0.6974806167478331, + "grad_norm": 1.4747273921966553, + "learning_rate": 4.188072357724716e-05, + "loss": 0.0786, + "step": 29754 + }, + { + "epoch": 0.6975040583226381, + "grad_norm": 2.3027193546295166, + "learning_rate": 4.1874729406734223e-05, + "loss": 1.1027, + "step": 29755 + }, + { + "epoch": 0.6975274998974431, + "grad_norm": 2.6785967350006104, + "learning_rate": 4.186873555161063e-05, + "loss": 0.328, + "step": 29756 + }, + { + "epoch": 0.697550941472248, + "grad_norm": 2.2023088932037354, + "learning_rate": 4.186274201190892e-05, + "loss": 0.1767, + "step": 29757 + }, + { + "epoch": 0.6975743830470531, + "grad_norm": 2.5203168392181396, + "learning_rate": 4.185674878766156e-05, + "loss": 0.1893, + "step": 29758 + }, + { + "epoch": 0.697597824621858, + "grad_norm": 1.9563888311386108, + "learning_rate": 4.185075587890116e-05, + "loss": 0.165, + "step": 29759 + }, + { + "epoch": 0.6976212661966631, + "grad_norm": 3.752746105194092, + "learning_rate": 4.1844763285660196e-05, + "loss": 0.3679, + "step": 29760 + }, + { + "epoch": 0.6976447077714681, + "grad_norm": 3.438961982727051, + "learning_rate": 4.183877100797117e-05, + "loss": 0.2676, + "step": 29761 + }, + { + "epoch": 0.6976681493462731, + "grad_norm": 1.5012516975402832, + "learning_rate": 4.183277904586661e-05, + "loss": 0.0796, + "step": 29762 + }, + { + "epoch": 0.6976915909210781, + "grad_norm": 4.200178623199463, + "learning_rate": 4.182678739937903e-05, + "loss": 0.5667, + "step": 29763 + }, + { + "epoch": 0.6977150324958831, + "grad_norm": 3.521395206451416, + "learning_rate": 4.1820796068540903e-05, + "loss": 1.311, + "step": 29764 + }, + { + "epoch": 0.6977384740706881, + "grad_norm": 0.9069419503211975, + "learning_rate": 4.1814805053384806e-05, + "loss": 0.0918, + "step": 29765 + }, + { + "epoch": 0.6977619156454931, + "grad_norm": 1.2565768957138062, + "learning_rate": 4.180881435394318e-05, + "loss": 0.1284, + "step": 29766 + }, + { + "epoch": 0.6977853572202981, + "grad_norm": 3.1892027854919434, + "learning_rate": 4.18028239702486e-05, + "loss": 0.2388, + "step": 29767 + }, + { + "epoch": 0.697808798795103, + "grad_norm": 3.3781750202178955, + "learning_rate": 4.179683390233353e-05, + "loss": 0.2722, + "step": 29768 + }, + { + "epoch": 0.6978322403699081, + "grad_norm": 2.29402494430542, + "learning_rate": 4.179084415023048e-05, + "loss": 0.1437, + "step": 29769 + }, + { + "epoch": 0.697855681944713, + "grad_norm": 3.166111469268799, + "learning_rate": 4.1784854713971946e-05, + "loss": 0.3654, + "step": 29770 + }, + { + "epoch": 0.6978791235195181, + "grad_norm": 3.8488876819610596, + "learning_rate": 4.177886559359039e-05, + "loss": 0.4535, + "step": 29771 + }, + { + "epoch": 0.697902565094323, + "grad_norm": 1.9464123249053955, + "learning_rate": 4.177287678911837e-05, + "loss": 0.0848, + "step": 29772 + }, + { + "epoch": 0.6979260066691281, + "grad_norm": 0.6546407341957092, + "learning_rate": 4.1766888300588366e-05, + "loss": 0.0282, + "step": 29773 + }, + { + "epoch": 0.697949448243933, + "grad_norm": 3.2115671634674072, + "learning_rate": 4.1760900128032874e-05, + "loss": 0.213, + "step": 29774 + }, + { + "epoch": 0.6979728898187381, + "grad_norm": 2.740384578704834, + "learning_rate": 4.175491227148437e-05, + "loss": 0.2228, + "step": 29775 + }, + { + "epoch": 0.697996331393543, + "grad_norm": 4.203597068786621, + "learning_rate": 4.174892473097535e-05, + "loss": 0.1848, + "step": 29776 + }, + { + "epoch": 0.698019772968348, + "grad_norm": 0.6633142828941345, + "learning_rate": 4.174293750653826e-05, + "loss": 0.0523, + "step": 29777 + }, + { + "epoch": 0.698043214543153, + "grad_norm": 4.92379903793335, + "learning_rate": 4.173695059820563e-05, + "loss": 0.2779, + "step": 29778 + }, + { + "epoch": 0.698066656117958, + "grad_norm": 2.955338954925537, + "learning_rate": 4.173096400600999e-05, + "loss": 0.2626, + "step": 29779 + }, + { + "epoch": 0.698090097692763, + "grad_norm": 3.0966508388519287, + "learning_rate": 4.172497772998376e-05, + "loss": 0.217, + "step": 29780 + }, + { + "epoch": 0.698113539267568, + "grad_norm": 1.9831087589263916, + "learning_rate": 4.1718991770159456e-05, + "loss": 0.9581, + "step": 29781 + }, + { + "epoch": 0.698136980842373, + "grad_norm": 3.283224582672119, + "learning_rate": 4.171300612656953e-05, + "loss": 0.287, + "step": 29782 + }, + { + "epoch": 0.698160422417178, + "grad_norm": 0.7634434700012207, + "learning_rate": 4.170702079924648e-05, + "loss": 0.0957, + "step": 29783 + }, + { + "epoch": 0.698183863991983, + "grad_norm": 1.7816189527511597, + "learning_rate": 4.170103578822273e-05, + "loss": 0.7877, + "step": 29784 + }, + { + "epoch": 0.698207305566788, + "grad_norm": 1.5250592231750488, + "learning_rate": 4.169505109353085e-05, + "loss": 0.071, + "step": 29785 + }, + { + "epoch": 0.698230747141593, + "grad_norm": 1.5877717733383179, + "learning_rate": 4.1689066715203254e-05, + "loss": 0.1383, + "step": 29786 + }, + { + "epoch": 0.698254188716398, + "grad_norm": 3.145465850830078, + "learning_rate": 4.168308265327241e-05, + "loss": 0.2495, + "step": 29787 + }, + { + "epoch": 0.6982776302912029, + "grad_norm": 0.7923652529716492, + "learning_rate": 4.1677098907770816e-05, + "loss": 0.075, + "step": 29788 + }, + { + "epoch": 0.698301071866008, + "grad_norm": 4.02098274230957, + "learning_rate": 4.167111547873088e-05, + "loss": 0.3058, + "step": 29789 + }, + { + "epoch": 0.6983245134408129, + "grad_norm": 8.18656063079834, + "learning_rate": 4.1665132366185146e-05, + "loss": 1.0049, + "step": 29790 + }, + { + "epoch": 0.698347955015618, + "grad_norm": 1.3524881601333618, + "learning_rate": 4.1659149570166015e-05, + "loss": 0.0876, + "step": 29791 + }, + { + "epoch": 0.6983713965904229, + "grad_norm": 3.3680801391601562, + "learning_rate": 4.1653167090706e-05, + "loss": 0.3503, + "step": 29792 + }, + { + "epoch": 0.698394838165228, + "grad_norm": 1.4592236280441284, + "learning_rate": 4.164718492783756e-05, + "loss": 0.1097, + "step": 29793 + }, + { + "epoch": 0.6984182797400329, + "grad_norm": 2.9246208667755127, + "learning_rate": 4.164120308159312e-05, + "loss": 0.3816, + "step": 29794 + }, + { + "epoch": 0.698441721314838, + "grad_norm": 2.2804548740386963, + "learning_rate": 4.163522155200515e-05, + "loss": 1.5057, + "step": 29795 + }, + { + "epoch": 0.6984651628896429, + "grad_norm": 2.11061954498291, + "learning_rate": 4.162924033910609e-05, + "loss": 0.1771, + "step": 29796 + }, + { + "epoch": 0.6984886044644479, + "grad_norm": 3.1580095291137695, + "learning_rate": 4.162325944292844e-05, + "loss": 0.2126, + "step": 29797 + }, + { + "epoch": 0.6985120460392529, + "grad_norm": 0.7309976816177368, + "learning_rate": 4.1617278863504626e-05, + "loss": 0.0452, + "step": 29798 + }, + { + "epoch": 0.6985354876140579, + "grad_norm": 0.6231219172477722, + "learning_rate": 4.16112986008671e-05, + "loss": 0.0627, + "step": 29799 + }, + { + "epoch": 0.6985589291888629, + "grad_norm": 1.2938581705093384, + "learning_rate": 4.160531865504831e-05, + "loss": 0.1523, + "step": 29800 + }, + { + "epoch": 0.6985823707636679, + "grad_norm": 1.828611135482788, + "learning_rate": 4.159933902608071e-05, + "loss": 0.1864, + "step": 29801 + }, + { + "epoch": 0.6986058123384729, + "grad_norm": 3.540776252746582, + "learning_rate": 4.159335971399668e-05, + "loss": 0.2729, + "step": 29802 + }, + { + "epoch": 0.6986292539132779, + "grad_norm": 2.8073856830596924, + "learning_rate": 4.158738071882874e-05, + "loss": 0.3432, + "step": 29803 + }, + { + "epoch": 0.6986526954880828, + "grad_norm": 3.3575360774993896, + "learning_rate": 4.1581402040609355e-05, + "loss": 0.288, + "step": 29804 + }, + { + "epoch": 0.6986761370628879, + "grad_norm": 2.2258293628692627, + "learning_rate": 4.157542367937092e-05, + "loss": 0.9719, + "step": 29805 + }, + { + "epoch": 0.6986995786376928, + "grad_norm": 3.192389726638794, + "learning_rate": 4.1569445635145865e-05, + "loss": 0.2875, + "step": 29806 + }, + { + "epoch": 0.6987230202124979, + "grad_norm": 3.695295572280884, + "learning_rate": 4.156346790796666e-05, + "loss": 0.3583, + "step": 29807 + }, + { + "epoch": 0.6987464617873028, + "grad_norm": 2.7344136238098145, + "learning_rate": 4.1557490497865715e-05, + "loss": 0.1744, + "step": 29808 + }, + { + "epoch": 0.6987699033621079, + "grad_norm": 1.4559109210968018, + "learning_rate": 4.155151340487543e-05, + "loss": 0.0749, + "step": 29809 + }, + { + "epoch": 0.6987933449369128, + "grad_norm": 1.2363425493240356, + "learning_rate": 4.154553662902831e-05, + "loss": 0.1859, + "step": 29810 + }, + { + "epoch": 0.6988167865117179, + "grad_norm": 3.8573310375213623, + "learning_rate": 4.153956017035676e-05, + "loss": 0.3604, + "step": 29811 + }, + { + "epoch": 0.6988402280865229, + "grad_norm": 1.6860734224319458, + "learning_rate": 4.153358402889319e-05, + "loss": 0.0824, + "step": 29812 + }, + { + "epoch": 0.6988636696613278, + "grad_norm": 2.409473180770874, + "learning_rate": 4.1527608204670056e-05, + "loss": 0.2099, + "step": 29813 + }, + { + "epoch": 0.6988871112361329, + "grad_norm": 2.1056244373321533, + "learning_rate": 4.152163269771975e-05, + "loss": 0.0896, + "step": 29814 + }, + { + "epoch": 0.6989105528109378, + "grad_norm": 0.5139236450195312, + "learning_rate": 4.151565750807467e-05, + "loss": 0.0525, + "step": 29815 + }, + { + "epoch": 0.6989339943857429, + "grad_norm": 3.676112174987793, + "learning_rate": 4.1509682635767274e-05, + "loss": 0.1924, + "step": 29816 + }, + { + "epoch": 0.6989574359605478, + "grad_norm": 2.908264398574829, + "learning_rate": 4.150370808083004e-05, + "loss": 0.1517, + "step": 29817 + }, + { + "epoch": 0.6989808775353529, + "grad_norm": 0.8475441932678223, + "learning_rate": 4.1497733843295314e-05, + "loss": 0.0514, + "step": 29818 + }, + { + "epoch": 0.6990043191101578, + "grad_norm": 0.6836283802986145, + "learning_rate": 4.149175992319554e-05, + "loss": 0.063, + "step": 29819 + }, + { + "epoch": 0.6990277606849629, + "grad_norm": 3.872957468032837, + "learning_rate": 4.148578632056311e-05, + "loss": 0.322, + "step": 29820 + }, + { + "epoch": 0.6990512022597678, + "grad_norm": 4.404240131378174, + "learning_rate": 4.147981303543046e-05, + "loss": 0.3994, + "step": 29821 + }, + { + "epoch": 0.6990746438345729, + "grad_norm": 1.1754101514816284, + "learning_rate": 4.147384006782995e-05, + "loss": 0.0579, + "step": 29822 + }, + { + "epoch": 0.6990980854093778, + "grad_norm": 3.9957988262176514, + "learning_rate": 4.146786741779406e-05, + "loss": 0.2431, + "step": 29823 + }, + { + "epoch": 0.6991215269841828, + "grad_norm": 1.051533579826355, + "learning_rate": 4.1461895085355175e-05, + "loss": 1.1834, + "step": 29824 + }, + { + "epoch": 0.6991449685589878, + "grad_norm": 0.44759073853492737, + "learning_rate": 4.145592307054569e-05, + "loss": 0.0484, + "step": 29825 + }, + { + "epoch": 0.6991684101337928, + "grad_norm": 1.1067967414855957, + "learning_rate": 4.144995137339801e-05, + "loss": 0.0806, + "step": 29826 + }, + { + "epoch": 0.6991918517085978, + "grad_norm": 2.930137872695923, + "learning_rate": 4.1443979993944504e-05, + "loss": 0.1952, + "step": 29827 + }, + { + "epoch": 0.6992152932834028, + "grad_norm": 3.9373104572296143, + "learning_rate": 4.143800893221761e-05, + "loss": 0.2186, + "step": 29828 + }, + { + "epoch": 0.6992387348582078, + "grad_norm": 0.5185988545417786, + "learning_rate": 4.143203818824977e-05, + "loss": 0.0413, + "step": 29829 + }, + { + "epoch": 0.6992621764330128, + "grad_norm": 2.9332480430603027, + "learning_rate": 4.142606776207333e-05, + "loss": 0.2093, + "step": 29830 + }, + { + "epoch": 0.6992856180078177, + "grad_norm": 2.9143741130828857, + "learning_rate": 4.142009765372069e-05, + "loss": 0.1689, + "step": 29831 + }, + { + "epoch": 0.6993090595826228, + "grad_norm": 1.4071012735366821, + "learning_rate": 4.141412786322425e-05, + "loss": 0.1467, + "step": 29832 + }, + { + "epoch": 0.6993325011574277, + "grad_norm": 3.69404935836792, + "learning_rate": 4.1408158390616406e-05, + "loss": 0.3077, + "step": 29833 + }, + { + "epoch": 0.6993559427322328, + "grad_norm": 3.4899964332580566, + "learning_rate": 4.14021892359295e-05, + "loss": 0.2739, + "step": 29834 + }, + { + "epoch": 0.6993793843070377, + "grad_norm": 2.2747550010681152, + "learning_rate": 4.139622039919599e-05, + "loss": 0.1389, + "step": 29835 + }, + { + "epoch": 0.6994028258818428, + "grad_norm": 0.3832278549671173, + "learning_rate": 4.139025188044825e-05, + "loss": 0.0289, + "step": 29836 + }, + { + "epoch": 0.6994262674566477, + "grad_norm": 2.6877803802490234, + "learning_rate": 4.138428367971865e-05, + "loss": 0.2276, + "step": 29837 + }, + { + "epoch": 0.6994497090314528, + "grad_norm": 2.432725429534912, + "learning_rate": 4.137831579703957e-05, + "loss": 1.5469, + "step": 29838 + }, + { + "epoch": 0.6994731506062577, + "grad_norm": 2.9871370792388916, + "learning_rate": 4.137234823244337e-05, + "loss": 0.1665, + "step": 29839 + }, + { + "epoch": 0.6994965921810627, + "grad_norm": 2.4560706615448, + "learning_rate": 4.136638098596248e-05, + "loss": 1.7706, + "step": 29840 + }, + { + "epoch": 0.6995200337558677, + "grad_norm": 1.8815046548843384, + "learning_rate": 4.1360414057629224e-05, + "loss": 0.086, + "step": 29841 + }, + { + "epoch": 0.6995434753306727, + "grad_norm": 2.816964626312256, + "learning_rate": 4.1354447447476054e-05, + "loss": 0.0526, + "step": 29842 + }, + { + "epoch": 0.6995669169054777, + "grad_norm": 2.235163688659668, + "learning_rate": 4.1348481155535294e-05, + "loss": 0.1257, + "step": 29843 + }, + { + "epoch": 0.6995903584802827, + "grad_norm": 3.588636636734009, + "learning_rate": 4.1342515181839326e-05, + "loss": 0.4033, + "step": 29844 + }, + { + "epoch": 0.6996138000550877, + "grad_norm": 0.5439311265945435, + "learning_rate": 4.133654952642051e-05, + "loss": 0.0465, + "step": 29845 + }, + { + "epoch": 0.6996372416298927, + "grad_norm": 1.6295697689056396, + "learning_rate": 4.133058418931124e-05, + "loss": 0.0702, + "step": 29846 + }, + { + "epoch": 0.6996606832046977, + "grad_norm": 2.908458948135376, + "learning_rate": 4.132461917054382e-05, + "loss": 0.4425, + "step": 29847 + }, + { + "epoch": 0.6996841247795027, + "grad_norm": 0.6973700523376465, + "learning_rate": 4.131865447015072e-05, + "loss": 0.0521, + "step": 29848 + }, + { + "epoch": 0.6997075663543076, + "grad_norm": 2.770437717437744, + "learning_rate": 4.131269008816424e-05, + "loss": 0.2333, + "step": 29849 + }, + { + "epoch": 0.6997310079291127, + "grad_norm": 1.9744161367416382, + "learning_rate": 4.130672602461675e-05, + "loss": 1.28, + "step": 29850 + }, + { + "epoch": 0.6997544495039176, + "grad_norm": 1.7030384540557861, + "learning_rate": 4.1300762279540583e-05, + "loss": 0.1318, + "step": 29851 + }, + { + "epoch": 0.6997778910787227, + "grad_norm": 2.1304633617401123, + "learning_rate": 4.1294798852968166e-05, + "loss": 0.1443, + "step": 29852 + }, + { + "epoch": 0.6998013326535276, + "grad_norm": 2.9871017932891846, + "learning_rate": 4.128883574493183e-05, + "loss": 0.2508, + "step": 29853 + }, + { + "epoch": 0.6998247742283327, + "grad_norm": 3.9698150157928467, + "learning_rate": 4.128287295546387e-05, + "loss": 0.1924, + "step": 29854 + }, + { + "epoch": 0.6998482158031376, + "grad_norm": 4.342004299163818, + "learning_rate": 4.127691048459674e-05, + "loss": 0.5078, + "step": 29855 + }, + { + "epoch": 0.6998716573779427, + "grad_norm": 3.070570707321167, + "learning_rate": 4.127094833236275e-05, + "loss": 0.2279, + "step": 29856 + }, + { + "epoch": 0.6998950989527476, + "grad_norm": 2.0875048637390137, + "learning_rate": 4.126498649879424e-05, + "loss": 0.1935, + "step": 29857 + }, + { + "epoch": 0.6999185405275526, + "grad_norm": 3.1745896339416504, + "learning_rate": 4.125902498392355e-05, + "loss": 0.3959, + "step": 29858 + }, + { + "epoch": 0.6999419821023576, + "grad_norm": 2.3879823684692383, + "learning_rate": 4.125306378778302e-05, + "loss": 0.1847, + "step": 29859 + }, + { + "epoch": 0.6999654236771626, + "grad_norm": 2.344712734222412, + "learning_rate": 4.124710291040506e-05, + "loss": 0.197, + "step": 29860 + }, + { + "epoch": 0.6999888652519676, + "grad_norm": 0.897651195526123, + "learning_rate": 4.124114235182195e-05, + "loss": 0.4917, + "step": 29861 + }, + { + "epoch": 0.7000123068267726, + "grad_norm": 3.0937395095825195, + "learning_rate": 4.1235182112066076e-05, + "loss": 0.1241, + "step": 29862 + }, + { + "epoch": 0.7000357484015777, + "grad_norm": 2.739309310913086, + "learning_rate": 4.1229222191169734e-05, + "loss": 1.3336, + "step": 29863 + }, + { + "epoch": 0.7000591899763826, + "grad_norm": 3.3345391750335693, + "learning_rate": 4.122326258916527e-05, + "loss": 0.2786, + "step": 29864 + }, + { + "epoch": 0.7000826315511877, + "grad_norm": 0.9861012101173401, + "learning_rate": 4.121730330608506e-05, + "loss": 0.0607, + "step": 29865 + }, + { + "epoch": 0.7001060731259926, + "grad_norm": 0.9752846956253052, + "learning_rate": 4.121134434196139e-05, + "loss": 0.0979, + "step": 29866 + }, + { + "epoch": 0.7001295147007977, + "grad_norm": 4.559325695037842, + "learning_rate": 4.1205385696826646e-05, + "loss": 0.331, + "step": 29867 + }, + { + "epoch": 0.7001529562756026, + "grad_norm": 3.49320387840271, + "learning_rate": 4.119942737071313e-05, + "loss": 0.1688, + "step": 29868 + }, + { + "epoch": 0.7001763978504076, + "grad_norm": 2.836571216583252, + "learning_rate": 4.119346936365318e-05, + "loss": 0.2, + "step": 29869 + }, + { + "epoch": 0.7001998394252126, + "grad_norm": 2.0777688026428223, + "learning_rate": 4.118751167567911e-05, + "loss": 1.1072, + "step": 29870 + }, + { + "epoch": 0.7002232810000176, + "grad_norm": 1.6282103061676025, + "learning_rate": 4.1181554306823265e-05, + "loss": 0.0976, + "step": 29871 + }, + { + "epoch": 0.7002467225748226, + "grad_norm": 5.13287878036499, + "learning_rate": 4.1175597257117926e-05, + "loss": 0.4428, + "step": 29872 + }, + { + "epoch": 0.7002701641496276, + "grad_norm": 0.3568688631057739, + "learning_rate": 4.1169640526595487e-05, + "loss": 0.0175, + "step": 29873 + }, + { + "epoch": 0.7002936057244326, + "grad_norm": 0.7362297773361206, + "learning_rate": 4.116368411528823e-05, + "loss": 0.0818, + "step": 29874 + }, + { + "epoch": 0.7003170472992376, + "grad_norm": 2.4845733642578125, + "learning_rate": 4.115772802322847e-05, + "loss": 0.2028, + "step": 29875 + }, + { + "epoch": 0.7003404888740425, + "grad_norm": 2.7123615741729736, + "learning_rate": 4.11517722504485e-05, + "loss": 0.2196, + "step": 29876 + }, + { + "epoch": 0.7003639304488476, + "grad_norm": 1.5739370584487915, + "learning_rate": 4.1145816796980705e-05, + "loss": 0.1815, + "step": 29877 + }, + { + "epoch": 0.7003873720236525, + "grad_norm": 3.65161395072937, + "learning_rate": 4.113986166285736e-05, + "loss": 0.3723, + "step": 29878 + }, + { + "epoch": 0.7004108135984576, + "grad_norm": 1.3455370664596558, + "learning_rate": 4.1133906848110745e-05, + "loss": 0.1252, + "step": 29879 + }, + { + "epoch": 0.7004342551732625, + "grad_norm": 3.0739681720733643, + "learning_rate": 4.112795235277325e-05, + "loss": 0.4131, + "step": 29880 + }, + { + "epoch": 0.7004576967480676, + "grad_norm": 1.6181058883666992, + "learning_rate": 4.112199817687714e-05, + "loss": 0.2064, + "step": 29881 + }, + { + "epoch": 0.7004811383228725, + "grad_norm": 0.3465345799922943, + "learning_rate": 4.111604432045471e-05, + "loss": 0.0382, + "step": 29882 + }, + { + "epoch": 0.7005045798976776, + "grad_norm": 0.4460281431674957, + "learning_rate": 4.111009078353829e-05, + "loss": 0.0375, + "step": 29883 + }, + { + "epoch": 0.7005280214724825, + "grad_norm": 3.3585543632507324, + "learning_rate": 4.110413756616018e-05, + "loss": 0.4626, + "step": 29884 + }, + { + "epoch": 0.7005514630472875, + "grad_norm": 1.3649215698242188, + "learning_rate": 4.1098184668352636e-05, + "loss": 0.0793, + "step": 29885 + }, + { + "epoch": 0.7005749046220925, + "grad_norm": 5.239516735076904, + "learning_rate": 4.109223209014802e-05, + "loss": 0.3793, + "step": 29886 + }, + { + "epoch": 0.7005983461968975, + "grad_norm": 2.112509250640869, + "learning_rate": 4.1086279831578624e-05, + "loss": 0.1771, + "step": 29887 + }, + { + "epoch": 0.7006217877717025, + "grad_norm": 1.3682793378829956, + "learning_rate": 4.108032789267669e-05, + "loss": 0.1076, + "step": 29888 + }, + { + "epoch": 0.7006452293465075, + "grad_norm": 3.114089250564575, + "learning_rate": 4.107437627347459e-05, + "loss": 0.165, + "step": 29889 + }, + { + "epoch": 0.7006686709213125, + "grad_norm": 0.9312923550605774, + "learning_rate": 4.106842497400458e-05, + "loss": 0.1218, + "step": 29890 + }, + { + "epoch": 0.7006921124961175, + "grad_norm": 2.4355130195617676, + "learning_rate": 4.106247399429896e-05, + "loss": 1.4362, + "step": 29891 + }, + { + "epoch": 0.7007155540709225, + "grad_norm": 4.6254191398620605, + "learning_rate": 4.1056523334389965e-05, + "loss": 0.3918, + "step": 29892 + }, + { + "epoch": 0.7007389956457275, + "grad_norm": 3.318516254425049, + "learning_rate": 4.105057299430997e-05, + "loss": 0.2499, + "step": 29893 + }, + { + "epoch": 0.7007624372205324, + "grad_norm": 2.3331916332244873, + "learning_rate": 4.104462297409123e-05, + "loss": 0.1963, + "step": 29894 + }, + { + "epoch": 0.7007858787953375, + "grad_norm": 0.7288552522659302, + "learning_rate": 4.103867327376601e-05, + "loss": 0.0423, + "step": 29895 + }, + { + "epoch": 0.7008093203701424, + "grad_norm": 2.065326452255249, + "learning_rate": 4.103272389336662e-05, + "loss": 0.1407, + "step": 29896 + }, + { + "epoch": 0.7008327619449475, + "grad_norm": 2.975482940673828, + "learning_rate": 4.102677483292529e-05, + "loss": 0.2634, + "step": 29897 + }, + { + "epoch": 0.7008562035197524, + "grad_norm": 1.0541796684265137, + "learning_rate": 4.102082609247437e-05, + "loss": 0.0989, + "step": 29898 + }, + { + "epoch": 0.7008796450945575, + "grad_norm": 2.54253888130188, + "learning_rate": 4.1014877672046115e-05, + "loss": 0.2895, + "step": 29899 + }, + { + "epoch": 0.7009030866693624, + "grad_norm": 0.9049668908119202, + "learning_rate": 4.100892957167275e-05, + "loss": 0.0473, + "step": 29900 + }, + { + "epoch": 0.7009265282441675, + "grad_norm": 3.762812376022339, + "learning_rate": 4.100298179138663e-05, + "loss": 0.2368, + "step": 29901 + }, + { + "epoch": 0.7009499698189724, + "grad_norm": 0.5481036305427551, + "learning_rate": 4.099703433121998e-05, + "loss": 0.0566, + "step": 29902 + }, + { + "epoch": 0.7009734113937774, + "grad_norm": 3.4600861072540283, + "learning_rate": 4.099108719120509e-05, + "loss": 0.2048, + "step": 29903 + }, + { + "epoch": 0.7009968529685824, + "grad_norm": 2.3795042037963867, + "learning_rate": 4.098514037137419e-05, + "loss": 0.2563, + "step": 29904 + }, + { + "epoch": 0.7010202945433874, + "grad_norm": 1.2392723560333252, + "learning_rate": 4.0979193871759604e-05, + "loss": 0.1744, + "step": 29905 + }, + { + "epoch": 0.7010437361181924, + "grad_norm": 1.781969428062439, + "learning_rate": 4.097324769239357e-05, + "loss": 0.2642, + "step": 29906 + }, + { + "epoch": 0.7010671776929974, + "grad_norm": 0.5152778029441833, + "learning_rate": 4.0967301833308357e-05, + "loss": 0.0521, + "step": 29907 + }, + { + "epoch": 0.7010906192678024, + "grad_norm": 3.069936752319336, + "learning_rate": 4.096135629453622e-05, + "loss": 0.2439, + "step": 29908 + }, + { + "epoch": 0.7011140608426074, + "grad_norm": 2.7943320274353027, + "learning_rate": 4.095541107610943e-05, + "loss": 0.2726, + "step": 29909 + }, + { + "epoch": 0.7011375024174124, + "grad_norm": 4.1763458251953125, + "learning_rate": 4.09494661780602e-05, + "loss": 0.4699, + "step": 29910 + }, + { + "epoch": 0.7011609439922174, + "grad_norm": 3.641526460647583, + "learning_rate": 4.0943521600420866e-05, + "loss": 0.5063, + "step": 29911 + }, + { + "epoch": 0.7011843855670223, + "grad_norm": 1.6718318462371826, + "learning_rate": 4.093757734322361e-05, + "loss": 0.0623, + "step": 29912 + }, + { + "epoch": 0.7012078271418274, + "grad_norm": 3.6788599491119385, + "learning_rate": 4.0931633406500744e-05, + "loss": 0.3857, + "step": 29913 + }, + { + "epoch": 0.7012312687166324, + "grad_norm": 1.982812762260437, + "learning_rate": 4.0925689790284515e-05, + "loss": 0.1149, + "step": 29914 + }, + { + "epoch": 0.7012547102914374, + "grad_norm": 2.9758079051971436, + "learning_rate": 4.0919746494607135e-05, + "loss": 0.2537, + "step": 29915 + }, + { + "epoch": 0.7012781518662424, + "grad_norm": 3.6827869415283203, + "learning_rate": 4.091380351950088e-05, + "loss": 0.2323, + "step": 29916 + }, + { + "epoch": 0.7013015934410474, + "grad_norm": 2.8859922885894775, + "learning_rate": 4.090786086499796e-05, + "loss": 0.2265, + "step": 29917 + }, + { + "epoch": 0.7013250350158524, + "grad_norm": 1.5602034330368042, + "learning_rate": 4.090191853113068e-05, + "loss": 0.8433, + "step": 29918 + }, + { + "epoch": 0.7013484765906574, + "grad_norm": 0.5140725374221802, + "learning_rate": 4.089597651793124e-05, + "loss": 0.0418, + "step": 29919 + }, + { + "epoch": 0.7013719181654624, + "grad_norm": 2.707444429397583, + "learning_rate": 4.089003482543191e-05, + "loss": 0.3251, + "step": 29920 + }, + { + "epoch": 0.7013953597402673, + "grad_norm": 2.7672364711761475, + "learning_rate": 4.088409345366491e-05, + "loss": 0.3243, + "step": 29921 + }, + { + "epoch": 0.7014188013150724, + "grad_norm": 3.1431307792663574, + "learning_rate": 4.0878152402662475e-05, + "loss": 0.2391, + "step": 29922 + }, + { + "epoch": 0.7014422428898773, + "grad_norm": 0.7760409116744995, + "learning_rate": 4.087221167245681e-05, + "loss": 0.0983, + "step": 29923 + }, + { + "epoch": 0.7014656844646824, + "grad_norm": 2.5141022205352783, + "learning_rate": 4.086627126308019e-05, + "loss": 0.2605, + "step": 29924 + }, + { + "epoch": 0.7014891260394873, + "grad_norm": 2.3817050457000732, + "learning_rate": 4.086033117456488e-05, + "loss": 0.1945, + "step": 29925 + }, + { + "epoch": 0.7015125676142924, + "grad_norm": 2.9197072982788086, + "learning_rate": 4.085439140694308e-05, + "loss": 0.1857, + "step": 29926 + }, + { + "epoch": 0.7015360091890973, + "grad_norm": 3.980417013168335, + "learning_rate": 4.084845196024701e-05, + "loss": 1.8462, + "step": 29927 + }, + { + "epoch": 0.7015594507639024, + "grad_norm": 2.4994356632232666, + "learning_rate": 4.084251283450891e-05, + "loss": 0.2557, + "step": 29928 + }, + { + "epoch": 0.7015828923387073, + "grad_norm": 0.7843250632286072, + "learning_rate": 4.083657402976099e-05, + "loss": 0.0303, + "step": 29929 + }, + { + "epoch": 0.7016063339135123, + "grad_norm": 2.26676869392395, + "learning_rate": 4.083063554603546e-05, + "loss": 1.5182, + "step": 29930 + }, + { + "epoch": 0.7016297754883173, + "grad_norm": 2.9056739807128906, + "learning_rate": 4.0824697383364585e-05, + "loss": 0.3038, + "step": 29931 + }, + { + "epoch": 0.7016532170631223, + "grad_norm": 0.8882656693458557, + "learning_rate": 4.081875954178057e-05, + "loss": 0.1031, + "step": 29932 + }, + { + "epoch": 0.7016766586379273, + "grad_norm": 1.9382171630859375, + "learning_rate": 4.0812822021315635e-05, + "loss": 0.1364, + "step": 29933 + }, + { + "epoch": 0.7017001002127323, + "grad_norm": 1.1492950916290283, + "learning_rate": 4.080688482200199e-05, + "loss": 0.0988, + "step": 29934 + }, + { + "epoch": 0.7017235417875373, + "grad_norm": 2.77426815032959, + "learning_rate": 4.080094794387181e-05, + "loss": 0.2294, + "step": 29935 + }, + { + "epoch": 0.7017469833623423, + "grad_norm": 2.1456642150878906, + "learning_rate": 4.0795011386957395e-05, + "loss": 1.2512, + "step": 29936 + }, + { + "epoch": 0.7017704249371473, + "grad_norm": 3.1611480712890625, + "learning_rate": 4.0789075151290876e-05, + "loss": 0.2395, + "step": 29937 + }, + { + "epoch": 0.7017938665119523, + "grad_norm": 1.173182487487793, + "learning_rate": 4.078313923690453e-05, + "loss": 0.0482, + "step": 29938 + }, + { + "epoch": 0.7018173080867572, + "grad_norm": 4.362974166870117, + "learning_rate": 4.077720364383055e-05, + "loss": 0.2911, + "step": 29939 + }, + { + "epoch": 0.7018407496615623, + "grad_norm": 3.5298149585723877, + "learning_rate": 4.077126837210111e-05, + "loss": 0.3138, + "step": 29940 + }, + { + "epoch": 0.7018641912363672, + "grad_norm": 1.254557490348816, + "learning_rate": 4.0765333421748456e-05, + "loss": 0.1168, + "step": 29941 + }, + { + "epoch": 0.7018876328111723, + "grad_norm": 1.9664663076400757, + "learning_rate": 4.0759398792804715e-05, + "loss": 0.081, + "step": 29942 + }, + { + "epoch": 0.7019110743859772, + "grad_norm": 0.9673453569412231, + "learning_rate": 4.075346448530217e-05, + "loss": 0.1375, + "step": 29943 + }, + { + "epoch": 0.7019345159607823, + "grad_norm": 0.447597473859787, + "learning_rate": 4.074753049927301e-05, + "loss": 0.0355, + "step": 29944 + }, + { + "epoch": 0.7019579575355872, + "grad_norm": 1.0734895467758179, + "learning_rate": 4.074159683474941e-05, + "loss": 0.1171, + "step": 29945 + }, + { + "epoch": 0.7019813991103923, + "grad_norm": 2.503161668777466, + "learning_rate": 4.073566349176357e-05, + "loss": 0.1965, + "step": 29946 + }, + { + "epoch": 0.7020048406851972, + "grad_norm": 2.78753399848938, + "learning_rate": 4.0729730470347685e-05, + "loss": 0.4639, + "step": 29947 + }, + { + "epoch": 0.7020282822600022, + "grad_norm": 3.0882391929626465, + "learning_rate": 4.072379777053392e-05, + "loss": 0.3042, + "step": 29948 + }, + { + "epoch": 0.7020517238348072, + "grad_norm": 1.6363765001296997, + "learning_rate": 4.0717865392354494e-05, + "loss": 0.092, + "step": 29949 + }, + { + "epoch": 0.7020751654096122, + "grad_norm": 0.05825016275048256, + "learning_rate": 4.071193333584162e-05, + "loss": 0.0016, + "step": 29950 + }, + { + "epoch": 0.7020986069844172, + "grad_norm": 0.8653936386108398, + "learning_rate": 4.070600160102748e-05, + "loss": 0.0687, + "step": 29951 + }, + { + "epoch": 0.7021220485592222, + "grad_norm": 1.1667002439498901, + "learning_rate": 4.070007018794423e-05, + "loss": 0.0904, + "step": 29952 + }, + { + "epoch": 0.7021454901340272, + "grad_norm": 3.2090978622436523, + "learning_rate": 4.069413909662407e-05, + "loss": 0.1931, + "step": 29953 + }, + { + "epoch": 0.7021689317088322, + "grad_norm": 2.4071853160858154, + "learning_rate": 4.0688208327099184e-05, + "loss": 1.7491, + "step": 29954 + }, + { + "epoch": 0.7021923732836372, + "grad_norm": 2.46323299407959, + "learning_rate": 4.068227787940171e-05, + "loss": 0.15, + "step": 29955 + }, + { + "epoch": 0.7022158148584422, + "grad_norm": 2.5268452167510986, + "learning_rate": 4.0676347753563896e-05, + "loss": 0.2364, + "step": 29956 + }, + { + "epoch": 0.7022392564332471, + "grad_norm": 0.9267473220825195, + "learning_rate": 4.067041794961789e-05, + "loss": 0.1182, + "step": 29957 + }, + { + "epoch": 0.7022626980080522, + "grad_norm": 0.8122453093528748, + "learning_rate": 4.066448846759586e-05, + "loss": 0.0459, + "step": 29958 + }, + { + "epoch": 0.7022861395828571, + "grad_norm": 2.3124146461486816, + "learning_rate": 4.065855930752999e-05, + "loss": 0.0958, + "step": 29959 + }, + { + "epoch": 0.7023095811576622, + "grad_norm": 3.672159194946289, + "learning_rate": 4.065263046945243e-05, + "loss": 0.3275, + "step": 29960 + }, + { + "epoch": 0.7023330227324671, + "grad_norm": 3.290720224380493, + "learning_rate": 4.064670195339535e-05, + "loss": 0.2404, + "step": 29961 + }, + { + "epoch": 0.7023564643072722, + "grad_norm": 3.3719019889831543, + "learning_rate": 4.064077375939093e-05, + "loss": 0.3169, + "step": 29962 + }, + { + "epoch": 0.7023799058820771, + "grad_norm": 2.7375595569610596, + "learning_rate": 4.063484588747137e-05, + "loss": 0.2209, + "step": 29963 + }, + { + "epoch": 0.7024033474568822, + "grad_norm": 1.9210271835327148, + "learning_rate": 4.0628918337668806e-05, + "loss": 0.1281, + "step": 29964 + }, + { + "epoch": 0.7024267890316871, + "grad_norm": 3.198118209838867, + "learning_rate": 4.0622991110015407e-05, + "loss": 0.2951, + "step": 29965 + }, + { + "epoch": 0.7024502306064921, + "grad_norm": 4.620938777923584, + "learning_rate": 4.061706420454332e-05, + "loss": 0.3637, + "step": 29966 + }, + { + "epoch": 0.7024736721812972, + "grad_norm": 1.8690842390060425, + "learning_rate": 4.0611137621284724e-05, + "loss": 0.1251, + "step": 29967 + }, + { + "epoch": 0.7024971137561021, + "grad_norm": 0.693924069404602, + "learning_rate": 4.0605211360271714e-05, + "loss": 0.0629, + "step": 29968 + }, + { + "epoch": 0.7025205553309072, + "grad_norm": 3.3566040992736816, + "learning_rate": 4.059928542153654e-05, + "loss": 0.3696, + "step": 29969 + }, + { + "epoch": 0.7025439969057121, + "grad_norm": 0.3926146328449249, + "learning_rate": 4.059335980511132e-05, + "loss": 0.0265, + "step": 29970 + }, + { + "epoch": 0.7025674384805172, + "grad_norm": 3.9760704040527344, + "learning_rate": 4.058743451102819e-05, + "loss": 0.2557, + "step": 29971 + }, + { + "epoch": 0.7025908800553221, + "grad_norm": 2.6295902729034424, + "learning_rate": 4.058150953931932e-05, + "loss": 0.3353, + "step": 29972 + }, + { + "epoch": 0.7026143216301272, + "grad_norm": 4.626300811767578, + "learning_rate": 4.057558489001681e-05, + "loss": 0.4339, + "step": 29973 + }, + { + "epoch": 0.7026377632049321, + "grad_norm": 3.844083786010742, + "learning_rate": 4.056966056315288e-05, + "loss": 0.2479, + "step": 29974 + }, + { + "epoch": 0.7026612047797371, + "grad_norm": 1.332167387008667, + "learning_rate": 4.05637365587596e-05, + "loss": 0.0798, + "step": 29975 + }, + { + "epoch": 0.7026846463545421, + "grad_norm": 2.0366547107696533, + "learning_rate": 4.055781287686921e-05, + "loss": 0.1847, + "step": 29976 + }, + { + "epoch": 0.7027080879293471, + "grad_norm": 0.3917026221752167, + "learning_rate": 4.0551889517513784e-05, + "loss": 0.0178, + "step": 29977 + }, + { + "epoch": 0.7027315295041521, + "grad_norm": 2.3888099193573, + "learning_rate": 4.054596648072549e-05, + "loss": 0.1506, + "step": 29978 + }, + { + "epoch": 0.7027549710789571, + "grad_norm": 2.458887815475464, + "learning_rate": 4.054004376653644e-05, + "loss": 0.2187, + "step": 29979 + }, + { + "epoch": 0.7027784126537621, + "grad_norm": 0.8258459568023682, + "learning_rate": 4.053412137497876e-05, + "loss": 0.0624, + "step": 29980 + }, + { + "epoch": 0.7028018542285671, + "grad_norm": 3.0746312141418457, + "learning_rate": 4.052819930608463e-05, + "loss": 0.3541, + "step": 29981 + }, + { + "epoch": 0.702825295803372, + "grad_norm": 1.9739927053451538, + "learning_rate": 4.0522277559886176e-05, + "loss": 0.1308, + "step": 29982 + }, + { + "epoch": 0.7028487373781771, + "grad_norm": 2.663679599761963, + "learning_rate": 4.051635613641552e-05, + "loss": 1.5996, + "step": 29983 + }, + { + "epoch": 0.702872178952982, + "grad_norm": 4.272182464599609, + "learning_rate": 4.0510435035704777e-05, + "loss": 0.4477, + "step": 29984 + }, + { + "epoch": 0.7028956205277871, + "grad_norm": 1.0725817680358887, + "learning_rate": 4.050451425778605e-05, + "loss": 0.1363, + "step": 29985 + }, + { + "epoch": 0.702919062102592, + "grad_norm": 0.6296547651290894, + "learning_rate": 4.049859380269155e-05, + "loss": 0.0714, + "step": 29986 + }, + { + "epoch": 0.7029425036773971, + "grad_norm": 2.410653591156006, + "learning_rate": 4.049267367045331e-05, + "loss": 0.132, + "step": 29987 + }, + { + "epoch": 0.702965945252202, + "grad_norm": 3.6033432483673096, + "learning_rate": 4.048675386110353e-05, + "loss": 0.3108, + "step": 29988 + }, + { + "epoch": 0.7029893868270071, + "grad_norm": 2.4039735794067383, + "learning_rate": 4.0480834374674304e-05, + "loss": 0.1389, + "step": 29989 + }, + { + "epoch": 0.703012828401812, + "grad_norm": 2.10579252243042, + "learning_rate": 4.047491521119774e-05, + "loss": 0.1462, + "step": 29990 + }, + { + "epoch": 0.7030362699766171, + "grad_norm": 4.581515789031982, + "learning_rate": 4.046899637070596e-05, + "loss": 0.2817, + "step": 29991 + }, + { + "epoch": 0.703059711551422, + "grad_norm": 1.873464584350586, + "learning_rate": 4.0463077853231066e-05, + "loss": 0.9551, + "step": 29992 + }, + { + "epoch": 0.703083153126227, + "grad_norm": 1.9301952123641968, + "learning_rate": 4.045715965880517e-05, + "loss": 0.1519, + "step": 29993 + }, + { + "epoch": 0.703106594701032, + "grad_norm": 2.608351707458496, + "learning_rate": 4.0451241787460424e-05, + "loss": 0.1288, + "step": 29994 + }, + { + "epoch": 0.703130036275837, + "grad_norm": 1.1934376955032349, + "learning_rate": 4.044532423922892e-05, + "loss": 0.1133, + "step": 29995 + }, + { + "epoch": 0.703153477850642, + "grad_norm": 3.7584640979766846, + "learning_rate": 4.0439407014142764e-05, + "loss": 0.198, + "step": 29996 + }, + { + "epoch": 0.703176919425447, + "grad_norm": 2.618007183074951, + "learning_rate": 4.0433490112234004e-05, + "loss": 1.2665, + "step": 29997 + }, + { + "epoch": 0.703200361000252, + "grad_norm": 2.394200325012207, + "learning_rate": 4.042757353353486e-05, + "loss": 1.6877, + "step": 29998 + }, + { + "epoch": 0.703223802575057, + "grad_norm": 1.1279255151748657, + "learning_rate": 4.0421657278077375e-05, + "loss": 0.0983, + "step": 29999 + }, + { + "epoch": 0.703247244149862, + "grad_norm": 4.328824520111084, + "learning_rate": 4.041574134589361e-05, + "loss": 0.2758, + "step": 30000 + }, + { + "epoch": 0.703270685724667, + "grad_norm": 0.9228619337081909, + "learning_rate": 4.0409825737015736e-05, + "loss": 0.0876, + "step": 30001 + }, + { + "epoch": 0.7032941272994719, + "grad_norm": 1.8119057416915894, + "learning_rate": 4.0403910451475834e-05, + "loss": 0.0953, + "step": 30002 + }, + { + "epoch": 0.703317568874277, + "grad_norm": 3.042930841445923, + "learning_rate": 4.0397995489306e-05, + "loss": 0.1094, + "step": 30003 + }, + { + "epoch": 0.7033410104490819, + "grad_norm": 2.204943895339966, + "learning_rate": 4.0392080850538296e-05, + "loss": 0.151, + "step": 30004 + }, + { + "epoch": 0.703364452023887, + "grad_norm": 2.5484540462493896, + "learning_rate": 4.038616653520485e-05, + "loss": 0.2522, + "step": 30005 + }, + { + "epoch": 0.7033878935986919, + "grad_norm": 1.1387945413589478, + "learning_rate": 4.0380252543337706e-05, + "loss": 0.0965, + "step": 30006 + }, + { + "epoch": 0.703411335173497, + "grad_norm": 2.839564085006714, + "learning_rate": 4.037433887496902e-05, + "loss": 1.6391, + "step": 30007 + }, + { + "epoch": 0.7034347767483019, + "grad_norm": 0.5250778198242188, + "learning_rate": 4.036842553013085e-05, + "loss": 0.0732, + "step": 30008 + }, + { + "epoch": 0.703458218323107, + "grad_norm": 2.646604299545288, + "learning_rate": 4.0362512508855244e-05, + "loss": 0.2056, + "step": 30009 + }, + { + "epoch": 0.7034816598979119, + "grad_norm": 1.159328818321228, + "learning_rate": 4.035659981117435e-05, + "loss": 0.0987, + "step": 30010 + }, + { + "epoch": 0.7035051014727169, + "grad_norm": 1.5348210334777832, + "learning_rate": 4.035068743712023e-05, + "loss": 0.1372, + "step": 30011 + }, + { + "epoch": 0.7035285430475219, + "grad_norm": 2.0948245525360107, + "learning_rate": 4.034477538672495e-05, + "loss": 0.2029, + "step": 30012 + }, + { + "epoch": 0.7035519846223269, + "grad_norm": 2.5197620391845703, + "learning_rate": 4.033886366002057e-05, + "loss": 0.2225, + "step": 30013 + }, + { + "epoch": 0.7035754261971319, + "grad_norm": 0.991603672504425, + "learning_rate": 4.0332952257039226e-05, + "loss": 0.1006, + "step": 30014 + }, + { + "epoch": 0.7035988677719369, + "grad_norm": 2.892916679382324, + "learning_rate": 4.0327041177812965e-05, + "loss": 0.1136, + "step": 30015 + }, + { + "epoch": 0.7036223093467419, + "grad_norm": 3.1732680797576904, + "learning_rate": 4.0321130422373844e-05, + "loss": 0.2326, + "step": 30016 + }, + { + "epoch": 0.7036457509215469, + "grad_norm": 3.8477206230163574, + "learning_rate": 4.031521999075395e-05, + "loss": 0.3306, + "step": 30017 + }, + { + "epoch": 0.703669192496352, + "grad_norm": 1.6068633794784546, + "learning_rate": 4.0309309882985323e-05, + "loss": 0.0864, + "step": 30018 + }, + { + "epoch": 0.7036926340711569, + "grad_norm": 0.38729292154312134, + "learning_rate": 4.0303400099100095e-05, + "loss": 0.0171, + "step": 30019 + }, + { + "epoch": 0.703716075645962, + "grad_norm": 1.5656654834747314, + "learning_rate": 4.029749063913029e-05, + "loss": 0.0996, + "step": 30020 + }, + { + "epoch": 0.7037395172207669, + "grad_norm": 1.9773460626602173, + "learning_rate": 4.029158150310798e-05, + "loss": 0.1135, + "step": 30021 + }, + { + "epoch": 0.7037629587955719, + "grad_norm": 3.305314540863037, + "learning_rate": 4.028567269106519e-05, + "loss": 0.5534, + "step": 30022 + }, + { + "epoch": 0.7037864003703769, + "grad_norm": 0.5343354344367981, + "learning_rate": 4.0279764203034063e-05, + "loss": 0.0377, + "step": 30023 + }, + { + "epoch": 0.7038098419451819, + "grad_norm": 0.7644348740577698, + "learning_rate": 4.0273856039046607e-05, + "loss": 0.0441, + "step": 30024 + }, + { + "epoch": 0.7038332835199869, + "grad_norm": 5.347696304321289, + "learning_rate": 4.026794819913486e-05, + "loss": 0.4712, + "step": 30025 + }, + { + "epoch": 0.7038567250947919, + "grad_norm": 0.9912646412849426, + "learning_rate": 4.026204068333094e-05, + "loss": 0.0519, + "step": 30026 + }, + { + "epoch": 0.7038801666695969, + "grad_norm": 0.8214983940124512, + "learning_rate": 4.025613349166686e-05, + "loss": 0.0331, + "step": 30027 + }, + { + "epoch": 0.7039036082444019, + "grad_norm": 4.133302211761475, + "learning_rate": 4.025022662417468e-05, + "loss": 0.6101, + "step": 30028 + }, + { + "epoch": 0.7039270498192068, + "grad_norm": 3.1588337421417236, + "learning_rate": 4.0244320080886454e-05, + "loss": 1.2446, + "step": 30029 + }, + { + "epoch": 0.7039504913940119, + "grad_norm": 1.5822348594665527, + "learning_rate": 4.0238413861834224e-05, + "loss": 0.0931, + "step": 30030 + }, + { + "epoch": 0.7039739329688168, + "grad_norm": 1.0185368061065674, + "learning_rate": 4.023250796705001e-05, + "loss": 0.063, + "step": 30031 + }, + { + "epoch": 0.7039973745436219, + "grad_norm": 2.6223018169403076, + "learning_rate": 4.022660239656593e-05, + "loss": 0.249, + "step": 30032 + }, + { + "epoch": 0.7040208161184268, + "grad_norm": 0.9482018947601318, + "learning_rate": 4.022069715041397e-05, + "loss": 0.063, + "step": 30033 + }, + { + "epoch": 0.7040442576932319, + "grad_norm": 3.3673670291900635, + "learning_rate": 4.0214792228626164e-05, + "loss": 0.3782, + "step": 30034 + }, + { + "epoch": 0.7040676992680368, + "grad_norm": 2.2466392517089844, + "learning_rate": 4.0208887631234606e-05, + "loss": 1.5353, + "step": 30035 + }, + { + "epoch": 0.7040911408428419, + "grad_norm": 3.7663745880126953, + "learning_rate": 4.020298335827131e-05, + "loss": 0.3529, + "step": 30036 + }, + { + "epoch": 0.7041145824176468, + "grad_norm": 0.8972687125205994, + "learning_rate": 4.01970794097683e-05, + "loss": 0.0871, + "step": 30037 + }, + { + "epoch": 0.7041380239924518, + "grad_norm": 2.132155418395996, + "learning_rate": 4.019117578575758e-05, + "loss": 1.6296, + "step": 30038 + }, + { + "epoch": 0.7041614655672568, + "grad_norm": 1.9389784336090088, + "learning_rate": 4.0185272486271274e-05, + "loss": 0.1938, + "step": 30039 + }, + { + "epoch": 0.7041849071420618, + "grad_norm": 2.179208517074585, + "learning_rate": 4.017936951134135e-05, + "loss": 1.4276, + "step": 30040 + }, + { + "epoch": 0.7042083487168668, + "grad_norm": 1.355873465538025, + "learning_rate": 4.0173466860999854e-05, + "loss": 0.0791, + "step": 30041 + }, + { + "epoch": 0.7042317902916718, + "grad_norm": 2.582278251647949, + "learning_rate": 4.01675645352788e-05, + "loss": 0.2062, + "step": 30042 + }, + { + "epoch": 0.7042552318664768, + "grad_norm": 1.8821731805801392, + "learning_rate": 4.016166253421023e-05, + "loss": 0.1145, + "step": 30043 + }, + { + "epoch": 0.7042786734412818, + "grad_norm": 4.5279011726379395, + "learning_rate": 4.015576085782613e-05, + "loss": 0.249, + "step": 30044 + }, + { + "epoch": 0.7043021150160867, + "grad_norm": 1.9080109596252441, + "learning_rate": 4.014985950615858e-05, + "loss": 0.1389, + "step": 30045 + }, + { + "epoch": 0.7043255565908918, + "grad_norm": 1.2592833042144775, + "learning_rate": 4.014395847923955e-05, + "loss": 0.1568, + "step": 30046 + }, + { + "epoch": 0.7043489981656967, + "grad_norm": 1.3794405460357666, + "learning_rate": 4.013805777710111e-05, + "loss": 0.0936, + "step": 30047 + }, + { + "epoch": 0.7043724397405018, + "grad_norm": 2.481984853744507, + "learning_rate": 4.013215739977526e-05, + "loss": 0.2984, + "step": 30048 + }, + { + "epoch": 0.7043958813153067, + "grad_norm": 0.2971767485141754, + "learning_rate": 4.0126257347294e-05, + "loss": 0.0123, + "step": 30049 + }, + { + "epoch": 0.7044193228901118, + "grad_norm": 3.5368781089782715, + "learning_rate": 4.0120357619689364e-05, + "loss": 0.2225, + "step": 30050 + }, + { + "epoch": 0.7044427644649167, + "grad_norm": 1.5218453407287598, + "learning_rate": 4.01144582169933e-05, + "loss": 0.1384, + "step": 30051 + }, + { + "epoch": 0.7044662060397218, + "grad_norm": 4.560697078704834, + "learning_rate": 4.010855913923791e-05, + "loss": 0.2983, + "step": 30052 + }, + { + "epoch": 0.7044896476145267, + "grad_norm": 2.188154935836792, + "learning_rate": 4.010266038645517e-05, + "loss": 0.1597, + "step": 30053 + }, + { + "epoch": 0.7045130891893318, + "grad_norm": 1.7154533863067627, + "learning_rate": 4.0096761958677065e-05, + "loss": 0.0634, + "step": 30054 + }, + { + "epoch": 0.7045365307641367, + "grad_norm": 1.6404471397399902, + "learning_rate": 4.009086385593562e-05, + "loss": 1.0044, + "step": 30055 + }, + { + "epoch": 0.7045599723389417, + "grad_norm": 2.0175833702087402, + "learning_rate": 4.008496607826281e-05, + "loss": 0.1575, + "step": 30056 + }, + { + "epoch": 0.7045834139137467, + "grad_norm": 0.5779262185096741, + "learning_rate": 4.0079068625690675e-05, + "loss": 0.0511, + "step": 30057 + }, + { + "epoch": 0.7046068554885517, + "grad_norm": 3.248427152633667, + "learning_rate": 4.0073171498251175e-05, + "loss": 0.222, + "step": 30058 + }, + { + "epoch": 0.7046302970633567, + "grad_norm": 0.8926514387130737, + "learning_rate": 4.006727469597636e-05, + "loss": 0.0854, + "step": 30059 + }, + { + "epoch": 0.7046537386381617, + "grad_norm": 1.5283942222595215, + "learning_rate": 4.0061378218898205e-05, + "loss": 0.0856, + "step": 30060 + }, + { + "epoch": 0.7046771802129667, + "grad_norm": 3.4887208938598633, + "learning_rate": 4.005548206704869e-05, + "loss": 0.2924, + "step": 30061 + }, + { + "epoch": 0.7047006217877717, + "grad_norm": 1.4853618144989014, + "learning_rate": 4.004958624045981e-05, + "loss": 0.1701, + "step": 30062 + }, + { + "epoch": 0.7047240633625766, + "grad_norm": 3.7597312927246094, + "learning_rate": 4.004369073916354e-05, + "loss": 0.2197, + "step": 30063 + }, + { + "epoch": 0.7047475049373817, + "grad_norm": 0.8606910109519958, + "learning_rate": 4.003779556319191e-05, + "loss": 0.0675, + "step": 30064 + }, + { + "epoch": 0.7047709465121866, + "grad_norm": 5.542140007019043, + "learning_rate": 4.0031900712576894e-05, + "loss": 0.2954, + "step": 30065 + }, + { + "epoch": 0.7047943880869917, + "grad_norm": 0.7753850221633911, + "learning_rate": 4.002600618735047e-05, + "loss": 0.0895, + "step": 30066 + }, + { + "epoch": 0.7048178296617966, + "grad_norm": 0.6666287183761597, + "learning_rate": 4.002011198754463e-05, + "loss": 0.0772, + "step": 30067 + }, + { + "epoch": 0.7048412712366017, + "grad_norm": 1.4193397760391235, + "learning_rate": 4.001421811319135e-05, + "loss": 0.1449, + "step": 30068 + }, + { + "epoch": 0.7048647128114067, + "grad_norm": 0.6503329277038574, + "learning_rate": 4.000832456432256e-05, + "loss": 0.038, + "step": 30069 + }, + { + "epoch": 0.7048881543862117, + "grad_norm": 2.278364896774292, + "learning_rate": 4.00024313409703e-05, + "loss": 0.1207, + "step": 30070 + }, + { + "epoch": 0.7049115959610167, + "grad_norm": 0.8704830408096313, + "learning_rate": 3.999653844316657e-05, + "loss": 0.0959, + "step": 30071 + }, + { + "epoch": 0.7049350375358217, + "grad_norm": 1.0281723737716675, + "learning_rate": 3.99906458709433e-05, + "loss": 0.1334, + "step": 30072 + }, + { + "epoch": 0.7049584791106267, + "grad_norm": 2.7411396503448486, + "learning_rate": 3.998475362433247e-05, + "loss": 0.2113, + "step": 30073 + }, + { + "epoch": 0.7049819206854316, + "grad_norm": 0.46734005212783813, + "learning_rate": 3.997886170336607e-05, + "loss": 0.034, + "step": 30074 + }, + { + "epoch": 0.7050053622602367, + "grad_norm": 3.7412259578704834, + "learning_rate": 3.997297010807605e-05, + "loss": 0.2183, + "step": 30075 + }, + { + "epoch": 0.7050288038350416, + "grad_norm": 1.5792078971862793, + "learning_rate": 3.996707883849433e-05, + "loss": 0.0646, + "step": 30076 + }, + { + "epoch": 0.7050522454098467, + "grad_norm": 0.6849073171615601, + "learning_rate": 3.9961187894652974e-05, + "loss": 0.0591, + "step": 30077 + }, + { + "epoch": 0.7050756869846516, + "grad_norm": 3.568290948867798, + "learning_rate": 3.99552972765839e-05, + "loss": 0.2477, + "step": 30078 + }, + { + "epoch": 0.7050991285594567, + "grad_norm": 3.424943685531616, + "learning_rate": 3.994940698431908e-05, + "loss": 0.3419, + "step": 30079 + }, + { + "epoch": 0.7051225701342616, + "grad_norm": 4.082692623138428, + "learning_rate": 3.994351701789046e-05, + "loss": 0.2328, + "step": 30080 + }, + { + "epoch": 0.7051460117090667, + "grad_norm": 0.5916653275489807, + "learning_rate": 3.993762737733001e-05, + "loss": 0.052, + "step": 30081 + }, + { + "epoch": 0.7051694532838716, + "grad_norm": 2.33689546585083, + "learning_rate": 3.993173806266964e-05, + "loss": 0.2387, + "step": 30082 + }, + { + "epoch": 0.7051928948586766, + "grad_norm": 2.3374054431915283, + "learning_rate": 3.9925849073941345e-05, + "loss": 1.5222, + "step": 30083 + }, + { + "epoch": 0.7052163364334816, + "grad_norm": 2.2212674617767334, + "learning_rate": 3.991996041117713e-05, + "loss": 0.2326, + "step": 30084 + }, + { + "epoch": 0.7052397780082866, + "grad_norm": 2.027040958404541, + "learning_rate": 3.991407207440889e-05, + "loss": 0.1261, + "step": 30085 + }, + { + "epoch": 0.7052632195830916, + "grad_norm": 2.3447265625, + "learning_rate": 3.990818406366859e-05, + "loss": 0.2031, + "step": 30086 + }, + { + "epoch": 0.7052866611578966, + "grad_norm": 2.115086078643799, + "learning_rate": 3.990229637898817e-05, + "loss": 0.1802, + "step": 30087 + }, + { + "epoch": 0.7053101027327016, + "grad_norm": 4.198083400726318, + "learning_rate": 3.989640902039957e-05, + "loss": 0.4261, + "step": 30088 + }, + { + "epoch": 0.7053335443075066, + "grad_norm": 2.833752155303955, + "learning_rate": 3.989052198793473e-05, + "loss": 0.1719, + "step": 30089 + }, + { + "epoch": 0.7053569858823115, + "grad_norm": 1.1447803974151611, + "learning_rate": 3.9884635281625624e-05, + "loss": 0.1711, + "step": 30090 + }, + { + "epoch": 0.7053804274571166, + "grad_norm": 1.2226883172988892, + "learning_rate": 3.987874890150418e-05, + "loss": 0.1304, + "step": 30091 + }, + { + "epoch": 0.7054038690319215, + "grad_norm": 2.1447343826293945, + "learning_rate": 3.987286284760234e-05, + "loss": 1.1612, + "step": 30092 + }, + { + "epoch": 0.7054273106067266, + "grad_norm": 1.308294415473938, + "learning_rate": 3.986697711995204e-05, + "loss": 0.1631, + "step": 30093 + }, + { + "epoch": 0.7054507521815315, + "grad_norm": 0.8531020283699036, + "learning_rate": 3.986109171858518e-05, + "loss": 0.0605, + "step": 30094 + }, + { + "epoch": 0.7054741937563366, + "grad_norm": 4.239933490753174, + "learning_rate": 3.985520664353375e-05, + "loss": 0.4308, + "step": 30095 + }, + { + "epoch": 0.7054976353311415, + "grad_norm": 3.5596694946289062, + "learning_rate": 3.9849321894829636e-05, + "loss": 0.3461, + "step": 30096 + }, + { + "epoch": 0.7055210769059466, + "grad_norm": 3.872917890548706, + "learning_rate": 3.984343747250482e-05, + "loss": 0.2816, + "step": 30097 + }, + { + "epoch": 0.7055445184807515, + "grad_norm": 2.469895124435425, + "learning_rate": 3.9837553376591205e-05, + "loss": 0.1522, + "step": 30098 + }, + { + "epoch": 0.7055679600555566, + "grad_norm": 2.3416197299957275, + "learning_rate": 3.9831669607120726e-05, + "loss": 0.1181, + "step": 30099 + }, + { + "epoch": 0.7055914016303615, + "grad_norm": 0.8947708010673523, + "learning_rate": 3.982578616412529e-05, + "loss": 0.0869, + "step": 30100 + }, + { + "epoch": 0.7056148432051665, + "grad_norm": 5.903491973876953, + "learning_rate": 3.98199030476368e-05, + "loss": 0.208, + "step": 30101 + }, + { + "epoch": 0.7056382847799715, + "grad_norm": 2.679917335510254, + "learning_rate": 3.981402025768723e-05, + "loss": 1.3233, + "step": 30102 + }, + { + "epoch": 0.7056617263547765, + "grad_norm": 0.44438204169273376, + "learning_rate": 3.98081377943085e-05, + "loss": 0.0202, + "step": 30103 + }, + { + "epoch": 0.7056851679295815, + "grad_norm": 2.7193026542663574, + "learning_rate": 3.9802255657532494e-05, + "loss": 0.369, + "step": 30104 + }, + { + "epoch": 0.7057086095043865, + "grad_norm": 2.6233808994293213, + "learning_rate": 3.9796373847391144e-05, + "loss": 0.1623, + "step": 30105 + }, + { + "epoch": 0.7057320510791915, + "grad_norm": 2.413297414779663, + "learning_rate": 3.9790492363916365e-05, + "loss": 1.5569, + "step": 30106 + }, + { + "epoch": 0.7057554926539965, + "grad_norm": 1.1865432262420654, + "learning_rate": 3.978461120714002e-05, + "loss": 0.0836, + "step": 30107 + }, + { + "epoch": 0.7057789342288014, + "grad_norm": 2.4940240383148193, + "learning_rate": 3.9778730377094075e-05, + "loss": 0.2854, + "step": 30108 + }, + { + "epoch": 0.7058023758036065, + "grad_norm": 1.2503430843353271, + "learning_rate": 3.977284987381047e-05, + "loss": 0.1417, + "step": 30109 + }, + { + "epoch": 0.7058258173784114, + "grad_norm": 4.011404514312744, + "learning_rate": 3.9766969697321076e-05, + "loss": 0.3591, + "step": 30110 + }, + { + "epoch": 0.7058492589532165, + "grad_norm": 3.6901369094848633, + "learning_rate": 3.97610898476578e-05, + "loss": 0.4118, + "step": 30111 + }, + { + "epoch": 0.7058727005280214, + "grad_norm": 3.0069591999053955, + "learning_rate": 3.975521032485254e-05, + "loss": 0.1118, + "step": 30112 + }, + { + "epoch": 0.7058961421028265, + "grad_norm": 2.4665286540985107, + "learning_rate": 3.9749331128937206e-05, + "loss": 0.1205, + "step": 30113 + }, + { + "epoch": 0.7059195836776314, + "grad_norm": 3.1288466453552246, + "learning_rate": 3.9743452259943656e-05, + "loss": 0.1958, + "step": 30114 + }, + { + "epoch": 0.7059430252524365, + "grad_norm": 4.833426475524902, + "learning_rate": 3.973757371790386e-05, + "loss": 0.5397, + "step": 30115 + }, + { + "epoch": 0.7059664668272414, + "grad_norm": 5.356766700744629, + "learning_rate": 3.9731695502849685e-05, + "loss": 0.7302, + "step": 30116 + }, + { + "epoch": 0.7059899084020465, + "grad_norm": 3.255335807800293, + "learning_rate": 3.9725817614813035e-05, + "loss": 0.3126, + "step": 30117 + }, + { + "epoch": 0.7060133499768514, + "grad_norm": 1.1992796659469604, + "learning_rate": 3.971994005382579e-05, + "loss": 0.1068, + "step": 30118 + }, + { + "epoch": 0.7060367915516564, + "grad_norm": 2.5714359283447266, + "learning_rate": 3.9714062819919805e-05, + "loss": 0.1646, + "step": 30119 + }, + { + "epoch": 0.7060602331264615, + "grad_norm": 4.041020393371582, + "learning_rate": 3.9708185913127036e-05, + "loss": 0.4364, + "step": 30120 + }, + { + "epoch": 0.7060836747012664, + "grad_norm": 2.7191004753112793, + "learning_rate": 3.970230933347932e-05, + "loss": 0.1852, + "step": 30121 + }, + { + "epoch": 0.7061071162760715, + "grad_norm": 1.0015647411346436, + "learning_rate": 3.969643308100861e-05, + "loss": 0.1, + "step": 30122 + }, + { + "epoch": 0.7061305578508764, + "grad_norm": 5.107039928436279, + "learning_rate": 3.9690557155746745e-05, + "loss": 0.4364, + "step": 30123 + }, + { + "epoch": 0.7061539994256815, + "grad_norm": 1.3016172647476196, + "learning_rate": 3.968468155772561e-05, + "loss": 0.056, + "step": 30124 + }, + { + "epoch": 0.7061774410004864, + "grad_norm": 3.6646475791931152, + "learning_rate": 3.96788062869771e-05, + "loss": 0.4879, + "step": 30125 + }, + { + "epoch": 0.7062008825752915, + "grad_norm": 3.9278552532196045, + "learning_rate": 3.967293134353307e-05, + "loss": 0.3473, + "step": 30126 + }, + { + "epoch": 0.7062243241500964, + "grad_norm": 3.5463948249816895, + "learning_rate": 3.966705672742538e-05, + "loss": 0.2741, + "step": 30127 + }, + { + "epoch": 0.7062477657249014, + "grad_norm": 3.898272752761841, + "learning_rate": 3.9661182438685976e-05, + "loss": 0.2905, + "step": 30128 + }, + { + "epoch": 0.7062712072997064, + "grad_norm": 0.9029995799064636, + "learning_rate": 3.9655308477346686e-05, + "loss": 0.0336, + "step": 30129 + }, + { + "epoch": 0.7062946488745114, + "grad_norm": 1.740079402923584, + "learning_rate": 3.964943484343939e-05, + "loss": 1.0942, + "step": 30130 + }, + { + "epoch": 0.7063180904493164, + "grad_norm": 3.07133412361145, + "learning_rate": 3.964356153699593e-05, + "loss": 0.3514, + "step": 30131 + }, + { + "epoch": 0.7063415320241214, + "grad_norm": 2.481963634490967, + "learning_rate": 3.9637688558048225e-05, + "loss": 0.225, + "step": 30132 + }, + { + "epoch": 0.7063649735989264, + "grad_norm": 0.982602059841156, + "learning_rate": 3.963181590662809e-05, + "loss": 0.0842, + "step": 30133 + }, + { + "epoch": 0.7063884151737314, + "grad_norm": 3.9635231494903564, + "learning_rate": 3.962594358276746e-05, + "loss": 0.4328, + "step": 30134 + }, + { + "epoch": 0.7064118567485363, + "grad_norm": 0.9645938873291016, + "learning_rate": 3.962007158649815e-05, + "loss": 0.0444, + "step": 30135 + }, + { + "epoch": 0.7064352983233414, + "grad_norm": 3.0369997024536133, + "learning_rate": 3.961419991785203e-05, + "loss": 0.1894, + "step": 30136 + }, + { + "epoch": 0.7064587398981463, + "grad_norm": 0.31838154792785645, + "learning_rate": 3.960832857686096e-05, + "loss": 0.0263, + "step": 30137 + }, + { + "epoch": 0.7064821814729514, + "grad_norm": 3.136944055557251, + "learning_rate": 3.9602457563556805e-05, + "loss": 0.3101, + "step": 30138 + }, + { + "epoch": 0.7065056230477563, + "grad_norm": 2.0869529247283936, + "learning_rate": 3.959658687797136e-05, + "loss": 0.1279, + "step": 30139 + }, + { + "epoch": 0.7065290646225614, + "grad_norm": 4.04018497467041, + "learning_rate": 3.9590716520136584e-05, + "loss": 0.3182, + "step": 30140 + }, + { + "epoch": 0.7065525061973663, + "grad_norm": 2.932756185531616, + "learning_rate": 3.958484649008427e-05, + "loss": 0.1672, + "step": 30141 + }, + { + "epoch": 0.7065759477721714, + "grad_norm": 4.100789546966553, + "learning_rate": 3.957897678784629e-05, + "loss": 0.2922, + "step": 30142 + }, + { + "epoch": 0.7065993893469763, + "grad_norm": 1.6545875072479248, + "learning_rate": 3.9573107413454435e-05, + "loss": 1.0239, + "step": 30143 + }, + { + "epoch": 0.7066228309217814, + "grad_norm": 4.029358386993408, + "learning_rate": 3.956723836694064e-05, + "loss": 0.1726, + "step": 30144 + }, + { + "epoch": 0.7066462724965863, + "grad_norm": 2.7172763347625732, + "learning_rate": 3.9561369648336696e-05, + "loss": 0.2982, + "step": 30145 + }, + { + "epoch": 0.7066697140713913, + "grad_norm": 2.520902633666992, + "learning_rate": 3.955550125767444e-05, + "loss": 0.2073, + "step": 30146 + }, + { + "epoch": 0.7066931556461963, + "grad_norm": 2.004793643951416, + "learning_rate": 3.954963319498576e-05, + "loss": 0.211, + "step": 30147 + }, + { + "epoch": 0.7067165972210013, + "grad_norm": 1.6269558668136597, + "learning_rate": 3.9543765460302474e-05, + "loss": 0.1196, + "step": 30148 + }, + { + "epoch": 0.7067400387958063, + "grad_norm": 1.8222910165786743, + "learning_rate": 3.953789805365641e-05, + "loss": 0.1672, + "step": 30149 + }, + { + "epoch": 0.7067634803706113, + "grad_norm": 2.5637357234954834, + "learning_rate": 3.9532030975079415e-05, + "loss": 0.1628, + "step": 30150 + }, + { + "epoch": 0.7067869219454163, + "grad_norm": 2.37789249420166, + "learning_rate": 3.9526164224603325e-05, + "loss": 0.2018, + "step": 30151 + }, + { + "epoch": 0.7068103635202213, + "grad_norm": 1.9124276638031006, + "learning_rate": 3.952029780225993e-05, + "loss": 0.1243, + "step": 30152 + }, + { + "epoch": 0.7068338050950262, + "grad_norm": 2.2938520908355713, + "learning_rate": 3.9514431708081137e-05, + "loss": 0.1684, + "step": 30153 + }, + { + "epoch": 0.7068572466698313, + "grad_norm": 2.90486741065979, + "learning_rate": 3.950856594209873e-05, + "loss": 0.3054, + "step": 30154 + }, + { + "epoch": 0.7068806882446362, + "grad_norm": 3.7728872299194336, + "learning_rate": 3.9502700504344515e-05, + "loss": 0.3623, + "step": 30155 + }, + { + "epoch": 0.7069041298194413, + "grad_norm": 0.9032248854637146, + "learning_rate": 3.9496835394850385e-05, + "loss": 0.0791, + "step": 30156 + }, + { + "epoch": 0.7069275713942462, + "grad_norm": 2.04465913772583, + "learning_rate": 3.9490970613648126e-05, + "loss": 1.3949, + "step": 30157 + }, + { + "epoch": 0.7069510129690513, + "grad_norm": 4.277573108673096, + "learning_rate": 3.948510616076956e-05, + "loss": 0.3945, + "step": 30158 + }, + { + "epoch": 0.7069744545438562, + "grad_norm": 0.196975976228714, + "learning_rate": 3.947924203624648e-05, + "loss": 0.0215, + "step": 30159 + }, + { + "epoch": 0.7069978961186613, + "grad_norm": 6.078815460205078, + "learning_rate": 3.947337824011076e-05, + "loss": 0.3924, + "step": 30160 + }, + { + "epoch": 0.7070213376934662, + "grad_norm": 3.4936025142669678, + "learning_rate": 3.94675147723942e-05, + "loss": 0.4297, + "step": 30161 + }, + { + "epoch": 0.7070447792682713, + "grad_norm": 0.8031219840049744, + "learning_rate": 3.94616516331286e-05, + "loss": 0.0463, + "step": 30162 + }, + { + "epoch": 0.7070682208430762, + "grad_norm": 3.3574581146240234, + "learning_rate": 3.945578882234578e-05, + "loss": 0.204, + "step": 30163 + }, + { + "epoch": 0.7070916624178812, + "grad_norm": 2.563525676727295, + "learning_rate": 3.9449926340077556e-05, + "loss": 1.2741, + "step": 30164 + }, + { + "epoch": 0.7071151039926862, + "grad_norm": 2.5397701263427734, + "learning_rate": 3.944406418635571e-05, + "loss": 0.1533, + "step": 30165 + }, + { + "epoch": 0.7071385455674912, + "grad_norm": 1.7486960887908936, + "learning_rate": 3.9438202361212095e-05, + "loss": 0.1317, + "step": 30166 + }, + { + "epoch": 0.7071619871422962, + "grad_norm": 1.970596194267273, + "learning_rate": 3.9432340864678495e-05, + "loss": 0.0918, + "step": 30167 + }, + { + "epoch": 0.7071854287171012, + "grad_norm": 1.059228539466858, + "learning_rate": 3.942647969678669e-05, + "loss": 0.0834, + "step": 30168 + }, + { + "epoch": 0.7072088702919062, + "grad_norm": 2.6558315753936768, + "learning_rate": 3.942061885756855e-05, + "loss": 1.0986, + "step": 30169 + }, + { + "epoch": 0.7072323118667112, + "grad_norm": 2.8541102409362793, + "learning_rate": 3.9414758347055825e-05, + "loss": 0.1929, + "step": 30170 + }, + { + "epoch": 0.7072557534415163, + "grad_norm": 3.203242063522339, + "learning_rate": 3.9408898165280295e-05, + "loss": 0.3244, + "step": 30171 + }, + { + "epoch": 0.7072791950163212, + "grad_norm": 3.572497606277466, + "learning_rate": 3.940303831227381e-05, + "loss": 0.2583, + "step": 30172 + }, + { + "epoch": 0.7073026365911262, + "grad_norm": 0.194033682346344, + "learning_rate": 3.9397178788068154e-05, + "loss": 0.009, + "step": 30173 + }, + { + "epoch": 0.7073260781659312, + "grad_norm": 2.368842124938965, + "learning_rate": 3.939131959269511e-05, + "loss": 1.6329, + "step": 30174 + }, + { + "epoch": 0.7073495197407362, + "grad_norm": 0.9859929084777832, + "learning_rate": 3.938546072618646e-05, + "loss": 0.0369, + "step": 30175 + }, + { + "epoch": 0.7073729613155412, + "grad_norm": 0.3672371208667755, + "learning_rate": 3.9379602188574014e-05, + "loss": 0.0392, + "step": 30176 + }, + { + "epoch": 0.7073964028903462, + "grad_norm": 0.981590211391449, + "learning_rate": 3.937374397988952e-05, + "loss": 0.0525, + "step": 30177 + }, + { + "epoch": 0.7074198444651512, + "grad_norm": 0.9394968152046204, + "learning_rate": 3.936788610016483e-05, + "loss": 0.1187, + "step": 30178 + }, + { + "epoch": 0.7074432860399562, + "grad_norm": 1.4668540954589844, + "learning_rate": 3.9362028549431686e-05, + "loss": 0.1426, + "step": 30179 + }, + { + "epoch": 0.7074667276147611, + "grad_norm": 1.2057846784591675, + "learning_rate": 3.935617132772185e-05, + "loss": 0.1039, + "step": 30180 + }, + { + "epoch": 0.7074901691895662, + "grad_norm": 3.787670850753784, + "learning_rate": 3.935031443506717e-05, + "loss": 0.3838, + "step": 30181 + }, + { + "epoch": 0.7075136107643711, + "grad_norm": 1.51580011844635, + "learning_rate": 3.9344457871499384e-05, + "loss": 0.1402, + "step": 30182 + }, + { + "epoch": 0.7075370523391762, + "grad_norm": 1.5459303855895996, + "learning_rate": 3.9338601637050274e-05, + "loss": 0.0848, + "step": 30183 + }, + { + "epoch": 0.7075604939139811, + "grad_norm": 2.7057530879974365, + "learning_rate": 3.933274573175159e-05, + "loss": 0.1016, + "step": 30184 + }, + { + "epoch": 0.7075839354887862, + "grad_norm": 1.7836737632751465, + "learning_rate": 3.9326890155635166e-05, + "loss": 0.1494, + "step": 30185 + }, + { + "epoch": 0.7076073770635911, + "grad_norm": 3.3367252349853516, + "learning_rate": 3.9321034908732736e-05, + "loss": 0.3694, + "step": 30186 + }, + { + "epoch": 0.7076308186383962, + "grad_norm": 3.791208505630493, + "learning_rate": 3.931517999107608e-05, + "loss": 0.1837, + "step": 30187 + }, + { + "epoch": 0.7076542602132011, + "grad_norm": 1.618718147277832, + "learning_rate": 3.9309325402696975e-05, + "loss": 0.1077, + "step": 30188 + }, + { + "epoch": 0.7076777017880062, + "grad_norm": 2.179884195327759, + "learning_rate": 3.930347114362716e-05, + "loss": 0.2619, + "step": 30189 + }, + { + "epoch": 0.7077011433628111, + "grad_norm": 2.8519275188446045, + "learning_rate": 3.9297617213898385e-05, + "loss": 1.9096, + "step": 30190 + }, + { + "epoch": 0.7077245849376161, + "grad_norm": 0.6990102529525757, + "learning_rate": 3.929176361354249e-05, + "loss": 0.0576, + "step": 30191 + }, + { + "epoch": 0.7077480265124211, + "grad_norm": 3.787044048309326, + "learning_rate": 3.9285910342591146e-05, + "loss": 0.2511, + "step": 30192 + }, + { + "epoch": 0.7077714680872261, + "grad_norm": 4.527109146118164, + "learning_rate": 3.928005740107621e-05, + "loss": 0.4721, + "step": 30193 + }, + { + "epoch": 0.7077949096620311, + "grad_norm": 2.3458609580993652, + "learning_rate": 3.927420478902938e-05, + "loss": 0.2251, + "step": 30194 + }, + { + "epoch": 0.7078183512368361, + "grad_norm": 1.835128903388977, + "learning_rate": 3.926835250648242e-05, + "loss": 0.2104, + "step": 30195 + }, + { + "epoch": 0.7078417928116411, + "grad_norm": 1.3735662698745728, + "learning_rate": 3.926250055346709e-05, + "loss": 0.1749, + "step": 30196 + }, + { + "epoch": 0.7078652343864461, + "grad_norm": 1.7818944454193115, + "learning_rate": 3.92566489300151e-05, + "loss": 0.0988, + "step": 30197 + }, + { + "epoch": 0.707888675961251, + "grad_norm": 2.113976240158081, + "learning_rate": 3.9250797636158287e-05, + "loss": 0.1097, + "step": 30198 + }, + { + "epoch": 0.7079121175360561, + "grad_norm": 1.0811525583267212, + "learning_rate": 3.9244946671928354e-05, + "loss": 0.087, + "step": 30199 + }, + { + "epoch": 0.707935559110861, + "grad_norm": 2.160658597946167, + "learning_rate": 3.923909603735704e-05, + "loss": 0.188, + "step": 30200 + }, + { + "epoch": 0.7079590006856661, + "grad_norm": 2.4119131565093994, + "learning_rate": 3.92332457324761e-05, + "loss": 0.1448, + "step": 30201 + }, + { + "epoch": 0.707982442260471, + "grad_norm": 1.4228460788726807, + "learning_rate": 3.922739575731724e-05, + "loss": 0.0958, + "step": 30202 + }, + { + "epoch": 0.7080058838352761, + "grad_norm": 2.728549003601074, + "learning_rate": 3.922154611191228e-05, + "loss": 0.1857, + "step": 30203 + }, + { + "epoch": 0.708029325410081, + "grad_norm": 2.850955009460449, + "learning_rate": 3.9215696796292875e-05, + "loss": 0.3179, + "step": 30204 + }, + { + "epoch": 0.7080527669848861, + "grad_norm": 2.0117909908294678, + "learning_rate": 3.920984781049086e-05, + "loss": 0.176, + "step": 30205 + }, + { + "epoch": 0.708076208559691, + "grad_norm": 3.9620158672332764, + "learning_rate": 3.9203999154537905e-05, + "loss": 0.5026, + "step": 30206 + }, + { + "epoch": 0.708099650134496, + "grad_norm": 2.478266716003418, + "learning_rate": 3.919815082846577e-05, + "loss": 0.2279, + "step": 30207 + }, + { + "epoch": 0.708123091709301, + "grad_norm": 1.12270188331604, + "learning_rate": 3.919230283230618e-05, + "loss": 0.0861, + "step": 30208 + }, + { + "epoch": 0.708146533284106, + "grad_norm": 0.77884840965271, + "learning_rate": 3.918645516609082e-05, + "loss": 0.0856, + "step": 30209 + }, + { + "epoch": 0.708169974858911, + "grad_norm": 0.4640565812587738, + "learning_rate": 3.91806078298515e-05, + "loss": 0.0332, + "step": 30210 + }, + { + "epoch": 0.708193416433716, + "grad_norm": 3.944866180419922, + "learning_rate": 3.917476082361991e-05, + "loss": 0.5438, + "step": 30211 + }, + { + "epoch": 0.708216858008521, + "grad_norm": 4.27445650100708, + "learning_rate": 3.9168914147427794e-05, + "loss": 0.4743, + "step": 30212 + }, + { + "epoch": 0.708240299583326, + "grad_norm": 2.548635244369507, + "learning_rate": 3.916306780130685e-05, + "loss": 0.2235, + "step": 30213 + }, + { + "epoch": 0.708263741158131, + "grad_norm": 2.446268081665039, + "learning_rate": 3.915722178528881e-05, + "loss": 0.1781, + "step": 30214 + }, + { + "epoch": 0.708287182732936, + "grad_norm": 1.3814482688903809, + "learning_rate": 3.9151376099405354e-05, + "loss": 0.1506, + "step": 30215 + }, + { + "epoch": 0.7083106243077409, + "grad_norm": 2.2643649578094482, + "learning_rate": 3.9145530743688255e-05, + "loss": 0.2661, + "step": 30216 + }, + { + "epoch": 0.708334065882546, + "grad_norm": 3.2360992431640625, + "learning_rate": 3.913968571816925e-05, + "loss": 0.3161, + "step": 30217 + }, + { + "epoch": 0.7083575074573509, + "grad_norm": 2.4885988235473633, + "learning_rate": 3.9133841022880013e-05, + "loss": 0.2577, + "step": 30218 + }, + { + "epoch": 0.708380949032156, + "grad_norm": 2.4387104511260986, + "learning_rate": 3.912799665785227e-05, + "loss": 0.2095, + "step": 30219 + }, + { + "epoch": 0.7084043906069609, + "grad_norm": 2.595890760421753, + "learning_rate": 3.912215262311773e-05, + "loss": 0.3068, + "step": 30220 + }, + { + "epoch": 0.708427832181766, + "grad_norm": 1.8832143545150757, + "learning_rate": 3.91163089187081e-05, + "loss": 0.0917, + "step": 30221 + }, + { + "epoch": 0.708451273756571, + "grad_norm": 4.29249382019043, + "learning_rate": 3.911046554465506e-05, + "loss": 0.4518, + "step": 30222 + }, + { + "epoch": 0.708474715331376, + "grad_norm": 2.5787603855133057, + "learning_rate": 3.9104622500990374e-05, + "loss": 0.1135, + "step": 30223 + }, + { + "epoch": 0.708498156906181, + "grad_norm": 1.1844086647033691, + "learning_rate": 3.909877978774572e-05, + "loss": 0.0746, + "step": 30224 + }, + { + "epoch": 0.708521598480986, + "grad_norm": 4.010961532592773, + "learning_rate": 3.90929374049528e-05, + "loss": 0.389, + "step": 30225 + }, + { + "epoch": 0.708545040055791, + "grad_norm": 2.9224441051483154, + "learning_rate": 3.908709535264332e-05, + "loss": 0.3711, + "step": 30226 + }, + { + "epoch": 0.7085684816305959, + "grad_norm": 2.7822985649108887, + "learning_rate": 3.908125363084897e-05, + "loss": 0.1289, + "step": 30227 + }, + { + "epoch": 0.708591923205401, + "grad_norm": 3.507694721221924, + "learning_rate": 3.907541223960142e-05, + "loss": 0.1662, + "step": 30228 + }, + { + "epoch": 0.7086153647802059, + "grad_norm": 3.9196410179138184, + "learning_rate": 3.9069571178932386e-05, + "loss": 0.3, + "step": 30229 + }, + { + "epoch": 0.708638806355011, + "grad_norm": 0.752027690410614, + "learning_rate": 3.90637304488736e-05, + "loss": 0.0621, + "step": 30230 + }, + { + "epoch": 0.7086622479298159, + "grad_norm": 1.9468971490859985, + "learning_rate": 3.9057890049456746e-05, + "loss": 0.2043, + "step": 30231 + }, + { + "epoch": 0.708685689504621, + "grad_norm": 2.997152805328369, + "learning_rate": 3.905204998071348e-05, + "loss": 0.3338, + "step": 30232 + }, + { + "epoch": 0.7087091310794259, + "grad_norm": 3.2554593086242676, + "learning_rate": 3.9046210242675515e-05, + "loss": 0.2115, + "step": 30233 + }, + { + "epoch": 0.708732572654231, + "grad_norm": 2.052133798599243, + "learning_rate": 3.904037083537452e-05, + "loss": 0.1719, + "step": 30234 + }, + { + "epoch": 0.7087560142290359, + "grad_norm": 1.6642282009124756, + "learning_rate": 3.903453175884215e-05, + "loss": 0.0917, + "step": 30235 + }, + { + "epoch": 0.7087794558038409, + "grad_norm": 1.980574131011963, + "learning_rate": 3.902869301311016e-05, + "loss": 0.085, + "step": 30236 + }, + { + "epoch": 0.7088028973786459, + "grad_norm": 2.686393976211548, + "learning_rate": 3.902285459821021e-05, + "loss": 0.1901, + "step": 30237 + }, + { + "epoch": 0.7088263389534509, + "grad_norm": 0.7054540514945984, + "learning_rate": 3.9017016514173956e-05, + "loss": 0.0843, + "step": 30238 + }, + { + "epoch": 0.7088497805282559, + "grad_norm": 1.7757488489151, + "learning_rate": 3.9011178761033086e-05, + "loss": 0.1222, + "step": 30239 + }, + { + "epoch": 0.7088732221030609, + "grad_norm": 2.705841064453125, + "learning_rate": 3.900534133881923e-05, + "loss": 0.252, + "step": 30240 + }, + { + "epoch": 0.7088966636778659, + "grad_norm": 0.8968443870544434, + "learning_rate": 3.899950424756416e-05, + "loss": 0.0568, + "step": 30241 + }, + { + "epoch": 0.7089201052526709, + "grad_norm": 1.053113341331482, + "learning_rate": 3.8993667487299446e-05, + "loss": 0.1, + "step": 30242 + }, + { + "epoch": 0.7089435468274758, + "grad_norm": 2.8670661449432373, + "learning_rate": 3.898783105805684e-05, + "loss": 0.4339, + "step": 30243 + }, + { + "epoch": 0.7089669884022809, + "grad_norm": 3.978065013885498, + "learning_rate": 3.8981994959867986e-05, + "loss": 0.3709, + "step": 30244 + }, + { + "epoch": 0.7089904299770858, + "grad_norm": 2.8151071071624756, + "learning_rate": 3.897615919276455e-05, + "loss": 0.2213, + "step": 30245 + }, + { + "epoch": 0.7090138715518909, + "grad_norm": 2.4086732864379883, + "learning_rate": 3.897032375677818e-05, + "loss": 0.9422, + "step": 30246 + }, + { + "epoch": 0.7090373131266958, + "grad_norm": 0.6633596420288086, + "learning_rate": 3.8964488651940514e-05, + "loss": 0.0454, + "step": 30247 + }, + { + "epoch": 0.7090607547015009, + "grad_norm": 0.6506020426750183, + "learning_rate": 3.895865387828328e-05, + "loss": 0.0321, + "step": 30248 + }, + { + "epoch": 0.7090841962763058, + "grad_norm": 1.1553819179534912, + "learning_rate": 3.8952819435838116e-05, + "loss": 1.0911, + "step": 30249 + }, + { + "epoch": 0.7091076378511109, + "grad_norm": 2.8686838150024414, + "learning_rate": 3.8946985324636664e-05, + "loss": 0.2209, + "step": 30250 + }, + { + "epoch": 0.7091310794259158, + "grad_norm": 5.889098644256592, + "learning_rate": 3.89411515447106e-05, + "loss": 0.4917, + "step": 30251 + }, + { + "epoch": 0.7091545210007208, + "grad_norm": 3.628444194793701, + "learning_rate": 3.893531809609155e-05, + "loss": 0.3015, + "step": 30252 + }, + { + "epoch": 0.7091779625755258, + "grad_norm": 2.4875965118408203, + "learning_rate": 3.8929484978811145e-05, + "loss": 0.2, + "step": 30253 + }, + { + "epoch": 0.7092014041503308, + "grad_norm": 1.676805853843689, + "learning_rate": 3.892365219290107e-05, + "loss": 0.1043, + "step": 30254 + }, + { + "epoch": 0.7092248457251358, + "grad_norm": 1.737181544303894, + "learning_rate": 3.8917819738393024e-05, + "loss": 0.0824, + "step": 30255 + }, + { + "epoch": 0.7092482872999408, + "grad_norm": 2.613939046859741, + "learning_rate": 3.89119876153186e-05, + "loss": 0.3967, + "step": 30256 + }, + { + "epoch": 0.7092717288747458, + "grad_norm": 3.7917587757110596, + "learning_rate": 3.890615582370944e-05, + "loss": 0.4503, + "step": 30257 + }, + { + "epoch": 0.7092951704495508, + "grad_norm": 2.6630256175994873, + "learning_rate": 3.890032436359721e-05, + "loss": 0.2337, + "step": 30258 + }, + { + "epoch": 0.7093186120243558, + "grad_norm": 2.556185245513916, + "learning_rate": 3.889449323501353e-05, + "loss": 1.8011, + "step": 30259 + }, + { + "epoch": 0.7093420535991608, + "grad_norm": 1.5316675901412964, + "learning_rate": 3.888866243799002e-05, + "loss": 0.0667, + "step": 30260 + }, + { + "epoch": 0.7093654951739657, + "grad_norm": 2.1077518463134766, + "learning_rate": 3.888283197255836e-05, + "loss": 0.224, + "step": 30261 + }, + { + "epoch": 0.7093889367487708, + "grad_norm": 2.7815327644348145, + "learning_rate": 3.8877001838750195e-05, + "loss": 0.1838, + "step": 30262 + }, + { + "epoch": 0.7094123783235757, + "grad_norm": 3.0728538036346436, + "learning_rate": 3.8871172036597134e-05, + "loss": 0.2458, + "step": 30263 + }, + { + "epoch": 0.7094358198983808, + "grad_norm": 2.891897439956665, + "learning_rate": 3.8865342566130805e-05, + "loss": 0.1582, + "step": 30264 + }, + { + "epoch": 0.7094592614731857, + "grad_norm": 1.1153258085250854, + "learning_rate": 3.885951342738281e-05, + "loss": 0.0768, + "step": 30265 + }, + { + "epoch": 0.7094827030479908, + "grad_norm": 3.389094352722168, + "learning_rate": 3.8853684620384854e-05, + "loss": 0.2181, + "step": 30266 + }, + { + "epoch": 0.7095061446227957, + "grad_norm": 1.665693998336792, + "learning_rate": 3.8847856145168495e-05, + "loss": 0.0722, + "step": 30267 + }, + { + "epoch": 0.7095295861976008, + "grad_norm": 1.0228657722473145, + "learning_rate": 3.884202800176541e-05, + "loss": 0.0665, + "step": 30268 + }, + { + "epoch": 0.7095530277724057, + "grad_norm": 1.966391921043396, + "learning_rate": 3.8836200190207204e-05, + "loss": 0.1485, + "step": 30269 + }, + { + "epoch": 0.7095764693472107, + "grad_norm": 1.1149653196334839, + "learning_rate": 3.88303727105255e-05, + "loss": 0.0451, + "step": 30270 + }, + { + "epoch": 0.7095999109220157, + "grad_norm": 2.3934264183044434, + "learning_rate": 3.8824545562751916e-05, + "loss": 0.316, + "step": 30271 + }, + { + "epoch": 0.7096233524968207, + "grad_norm": 1.1749013662338257, + "learning_rate": 3.881871874691806e-05, + "loss": 0.036, + "step": 30272 + }, + { + "epoch": 0.7096467940716258, + "grad_norm": 3.368483066558838, + "learning_rate": 3.881289226305552e-05, + "loss": 0.1773, + "step": 30273 + }, + { + "epoch": 0.7096702356464307, + "grad_norm": 1.8895418643951416, + "learning_rate": 3.8807066111196e-05, + "loss": 0.1296, + "step": 30274 + }, + { + "epoch": 0.7096936772212358, + "grad_norm": 4.256265163421631, + "learning_rate": 3.8801240291371046e-05, + "loss": 0.4362, + "step": 30275 + }, + { + "epoch": 0.7097171187960407, + "grad_norm": 0.659794807434082, + "learning_rate": 3.8795414803612286e-05, + "loss": 0.0439, + "step": 30276 + }, + { + "epoch": 0.7097405603708458, + "grad_norm": 2.5868051052093506, + "learning_rate": 3.8789589647951286e-05, + "loss": 0.1917, + "step": 30277 + }, + { + "epoch": 0.7097640019456507, + "grad_norm": 1.2598894834518433, + "learning_rate": 3.878376482441973e-05, + "loss": 0.1427, + "step": 30278 + }, + { + "epoch": 0.7097874435204558, + "grad_norm": 2.354525089263916, + "learning_rate": 3.877794033304919e-05, + "loss": 1.6714, + "step": 30279 + }, + { + "epoch": 0.7098108850952607, + "grad_norm": 2.873674154281616, + "learning_rate": 3.877211617387123e-05, + "loss": 0.2438, + "step": 30280 + }, + { + "epoch": 0.7098343266700657, + "grad_norm": 2.1237478256225586, + "learning_rate": 3.876629234691752e-05, + "loss": 0.1101, + "step": 30281 + }, + { + "epoch": 0.7098577682448707, + "grad_norm": 1.0321722030639648, + "learning_rate": 3.8760468852219644e-05, + "loss": 0.0576, + "step": 30282 + }, + { + "epoch": 0.7098812098196757, + "grad_norm": 2.5158021450042725, + "learning_rate": 3.875464568980918e-05, + "loss": 0.185, + "step": 30283 + }, + { + "epoch": 0.7099046513944807, + "grad_norm": 3.9652276039123535, + "learning_rate": 3.8748822859717724e-05, + "loss": 0.3321, + "step": 30284 + }, + { + "epoch": 0.7099280929692857, + "grad_norm": 2.352034091949463, + "learning_rate": 3.8743000361976845e-05, + "loss": 0.1489, + "step": 30285 + }, + { + "epoch": 0.7099515345440907, + "grad_norm": 0.7443190813064575, + "learning_rate": 3.8737178196618206e-05, + "loss": 0.0445, + "step": 30286 + }, + { + "epoch": 0.7099749761188957, + "grad_norm": 1.180288314819336, + "learning_rate": 3.873135636367335e-05, + "loss": 0.0786, + "step": 30287 + }, + { + "epoch": 0.7099984176937006, + "grad_norm": 2.1065385341644287, + "learning_rate": 3.87255348631739e-05, + "loss": 0.2021, + "step": 30288 + }, + { + "epoch": 0.7100218592685057, + "grad_norm": 3.2032246589660645, + "learning_rate": 3.871971369515136e-05, + "loss": 0.2596, + "step": 30289 + }, + { + "epoch": 0.7100453008433106, + "grad_norm": 3.108415126800537, + "learning_rate": 3.871389285963742e-05, + "loss": 0.2319, + "step": 30290 + }, + { + "epoch": 0.7100687424181157, + "grad_norm": 4.503974437713623, + "learning_rate": 3.8708072356663625e-05, + "loss": 0.4804, + "step": 30291 + }, + { + "epoch": 0.7100921839929206, + "grad_norm": 0.6654266715049744, + "learning_rate": 3.870225218626152e-05, + "loss": 0.0728, + "step": 30292 + }, + { + "epoch": 0.7101156255677257, + "grad_norm": 3.656170606613159, + "learning_rate": 3.869643234846275e-05, + "loss": 0.3033, + "step": 30293 + }, + { + "epoch": 0.7101390671425306, + "grad_norm": 2.3936033248901367, + "learning_rate": 3.8690612843298865e-05, + "loss": 0.1112, + "step": 30294 + }, + { + "epoch": 0.7101625087173357, + "grad_norm": 1.5305596590042114, + "learning_rate": 3.8684793670801434e-05, + "loss": 0.2111, + "step": 30295 + }, + { + "epoch": 0.7101859502921406, + "grad_norm": 1.8619177341461182, + "learning_rate": 3.867897483100204e-05, + "loss": 0.1151, + "step": 30296 + }, + { + "epoch": 0.7102093918669456, + "grad_norm": 3.8253731727600098, + "learning_rate": 3.867315632393225e-05, + "loss": 0.3642, + "step": 30297 + }, + { + "epoch": 0.7102328334417506, + "grad_norm": 2.244399309158325, + "learning_rate": 3.86673381496236e-05, + "loss": 1.364, + "step": 30298 + }, + { + "epoch": 0.7102562750165556, + "grad_norm": 4.039371013641357, + "learning_rate": 3.8661520308107744e-05, + "loss": 0.2431, + "step": 30299 + }, + { + "epoch": 0.7102797165913606, + "grad_norm": 1.1000398397445679, + "learning_rate": 3.865570279941619e-05, + "loss": 0.0801, + "step": 30300 + }, + { + "epoch": 0.7103031581661656, + "grad_norm": 2.8137314319610596, + "learning_rate": 3.8649885623580495e-05, + "loss": 0.2256, + "step": 30301 + }, + { + "epoch": 0.7103265997409706, + "grad_norm": 2.8366520404815674, + "learning_rate": 3.864406878063227e-05, + "loss": 0.1712, + "step": 30302 + }, + { + "epoch": 0.7103500413157756, + "grad_norm": 1.1556864976882935, + "learning_rate": 3.863825227060306e-05, + "loss": 0.0735, + "step": 30303 + }, + { + "epoch": 0.7103734828905806, + "grad_norm": 2.506427526473999, + "learning_rate": 3.8632436093524416e-05, + "loss": 0.2123, + "step": 30304 + }, + { + "epoch": 0.7103969244653856, + "grad_norm": 3.346367835998535, + "learning_rate": 3.862662024942786e-05, + "loss": 0.2877, + "step": 30305 + }, + { + "epoch": 0.7104203660401905, + "grad_norm": 2.4248299598693848, + "learning_rate": 3.8620804738345026e-05, + "loss": 0.1939, + "step": 30306 + }, + { + "epoch": 0.7104438076149956, + "grad_norm": 1.5925661325454712, + "learning_rate": 3.8614989560307435e-05, + "loss": 0.1883, + "step": 30307 + }, + { + "epoch": 0.7104672491898005, + "grad_norm": 2.780580759048462, + "learning_rate": 3.860917471534663e-05, + "loss": 0.1044, + "step": 30308 + }, + { + "epoch": 0.7104906907646056, + "grad_norm": 1.2879878282546997, + "learning_rate": 3.860336020349418e-05, + "loss": 0.1253, + "step": 30309 + }, + { + "epoch": 0.7105141323394105, + "grad_norm": 2.9074208736419678, + "learning_rate": 3.859754602478162e-05, + "loss": 0.123, + "step": 30310 + }, + { + "epoch": 0.7105375739142156, + "grad_norm": 1.4824227094650269, + "learning_rate": 3.859173217924046e-05, + "loss": 0.1419, + "step": 30311 + }, + { + "epoch": 0.7105610154890205, + "grad_norm": 2.894044876098633, + "learning_rate": 3.858591866690232e-05, + "loss": 0.1323, + "step": 30312 + }, + { + "epoch": 0.7105844570638256, + "grad_norm": 1.6446563005447388, + "learning_rate": 3.858010548779869e-05, + "loss": 0.1643, + "step": 30313 + }, + { + "epoch": 0.7106078986386305, + "grad_norm": 5.101386547088623, + "learning_rate": 3.857429264196116e-05, + "loss": 0.2234, + "step": 30314 + }, + { + "epoch": 0.7106313402134355, + "grad_norm": 0.7415658235549927, + "learning_rate": 3.856848012942125e-05, + "loss": 0.0451, + "step": 30315 + }, + { + "epoch": 0.7106547817882405, + "grad_norm": 3.76662015914917, + "learning_rate": 3.85626679502105e-05, + "loss": 0.4256, + "step": 30316 + }, + { + "epoch": 0.7106782233630455, + "grad_norm": 3.354266881942749, + "learning_rate": 3.8556856104360425e-05, + "loss": 0.2598, + "step": 30317 + }, + { + "epoch": 0.7107016649378505, + "grad_norm": 0.8313061594963074, + "learning_rate": 3.855104459190255e-05, + "loss": 0.045, + "step": 30318 + }, + { + "epoch": 0.7107251065126555, + "grad_norm": 3.2650952339172363, + "learning_rate": 3.854523341286848e-05, + "loss": 0.2622, + "step": 30319 + }, + { + "epoch": 0.7107485480874605, + "grad_norm": 3.8434739112854004, + "learning_rate": 3.8539422567289684e-05, + "loss": 0.2338, + "step": 30320 + }, + { + "epoch": 0.7107719896622655, + "grad_norm": 1.3076711893081665, + "learning_rate": 3.853361205519772e-05, + "loss": 0.5582, + "step": 30321 + }, + { + "epoch": 0.7107954312370705, + "grad_norm": 2.309553384780884, + "learning_rate": 3.85278018766241e-05, + "loss": 0.1208, + "step": 30322 + }, + { + "epoch": 0.7108188728118755, + "grad_norm": 3.2864913940429688, + "learning_rate": 3.852199203160032e-05, + "loss": 0.2047, + "step": 30323 + }, + { + "epoch": 0.7108423143866806, + "grad_norm": 0.8359276056289673, + "learning_rate": 3.8516182520157986e-05, + "loss": 0.0422, + "step": 30324 + }, + { + "epoch": 0.7108657559614855, + "grad_norm": 2.027839183807373, + "learning_rate": 3.851037334232855e-05, + "loss": 0.2278, + "step": 30325 + }, + { + "epoch": 0.7108891975362905, + "grad_norm": 2.8920061588287354, + "learning_rate": 3.8504564498143545e-05, + "loss": 0.2101, + "step": 30326 + }, + { + "epoch": 0.7109126391110955, + "grad_norm": 2.002964973449707, + "learning_rate": 3.849875598763453e-05, + "loss": 0.084, + "step": 30327 + }, + { + "epoch": 0.7109360806859005, + "grad_norm": 2.024052143096924, + "learning_rate": 3.849294781083299e-05, + "loss": 0.0926, + "step": 30328 + }, + { + "epoch": 0.7109595222607055, + "grad_norm": 3.216193199157715, + "learning_rate": 3.848713996777046e-05, + "loss": 0.2166, + "step": 30329 + }, + { + "epoch": 0.7109829638355105, + "grad_norm": 2.7290306091308594, + "learning_rate": 3.848133245847839e-05, + "loss": 0.2351, + "step": 30330 + }, + { + "epoch": 0.7110064054103155, + "grad_norm": 1.1521962881088257, + "learning_rate": 3.847552528298838e-05, + "loss": 0.0767, + "step": 30331 + }, + { + "epoch": 0.7110298469851205, + "grad_norm": 3.2593023777008057, + "learning_rate": 3.846971844133189e-05, + "loss": 0.3333, + "step": 30332 + }, + { + "epoch": 0.7110532885599254, + "grad_norm": 3.096588134765625, + "learning_rate": 3.846391193354044e-05, + "loss": 0.2231, + "step": 30333 + }, + { + "epoch": 0.7110767301347305, + "grad_norm": 0.7019892930984497, + "learning_rate": 3.8458105759645545e-05, + "loss": 0.0833, + "step": 30334 + }, + { + "epoch": 0.7111001717095354, + "grad_norm": 1.3251359462738037, + "learning_rate": 3.845229991967869e-05, + "loss": 0.1347, + "step": 30335 + }, + { + "epoch": 0.7111236132843405, + "grad_norm": 2.624544143676758, + "learning_rate": 3.844649441367135e-05, + "loss": 0.1822, + "step": 30336 + }, + { + "epoch": 0.7111470548591454, + "grad_norm": 1.1472086906433105, + "learning_rate": 3.84406892416551e-05, + "loss": 0.1349, + "step": 30337 + }, + { + "epoch": 0.7111704964339505, + "grad_norm": 2.8653249740600586, + "learning_rate": 3.8434884403661374e-05, + "loss": 0.2074, + "step": 30338 + }, + { + "epoch": 0.7111939380087554, + "grad_norm": 3.3356637954711914, + "learning_rate": 3.842907989972172e-05, + "loss": 0.1907, + "step": 30339 + }, + { + "epoch": 0.7112173795835605, + "grad_norm": 0.5181065797805786, + "learning_rate": 3.8423275729867616e-05, + "loss": 0.0294, + "step": 30340 + }, + { + "epoch": 0.7112408211583654, + "grad_norm": 4.602202415466309, + "learning_rate": 3.8417471894130544e-05, + "loss": 0.1827, + "step": 30341 + }, + { + "epoch": 0.7112642627331704, + "grad_norm": 2.548081398010254, + "learning_rate": 3.841166839254201e-05, + "loss": 0.9925, + "step": 30342 + }, + { + "epoch": 0.7112877043079754, + "grad_norm": 3.211618661880493, + "learning_rate": 3.840586522513345e-05, + "loss": 0.1175, + "step": 30343 + }, + { + "epoch": 0.7113111458827804, + "grad_norm": 3.792280912399292, + "learning_rate": 3.840006239193643e-05, + "loss": 0.4381, + "step": 30344 + }, + { + "epoch": 0.7113345874575854, + "grad_norm": 2.3900177478790283, + "learning_rate": 3.839425989298242e-05, + "loss": 1.3246, + "step": 30345 + }, + { + "epoch": 0.7113580290323904, + "grad_norm": 2.174116849899292, + "learning_rate": 3.838845772830287e-05, + "loss": 0.2052, + "step": 30346 + }, + { + "epoch": 0.7113814706071954, + "grad_norm": 3.104041576385498, + "learning_rate": 3.8382655897929285e-05, + "loss": 0.2568, + "step": 30347 + }, + { + "epoch": 0.7114049121820004, + "grad_norm": 0.7542067766189575, + "learning_rate": 3.837685440189315e-05, + "loss": 0.0444, + "step": 30348 + }, + { + "epoch": 0.7114283537568054, + "grad_norm": 3.9472458362579346, + "learning_rate": 3.83710532402259e-05, + "loss": 0.3572, + "step": 30349 + }, + { + "epoch": 0.7114517953316104, + "grad_norm": 1.0449109077453613, + "learning_rate": 3.836525241295904e-05, + "loss": 0.0744, + "step": 30350 + }, + { + "epoch": 0.7114752369064153, + "grad_norm": 2.7095448970794678, + "learning_rate": 3.835945192012409e-05, + "loss": 1.6073, + "step": 30351 + }, + { + "epoch": 0.7114986784812204, + "grad_norm": 2.6194169521331787, + "learning_rate": 3.8353651761752495e-05, + "loss": 0.1336, + "step": 30352 + }, + { + "epoch": 0.7115221200560253, + "grad_norm": 3.950976848602295, + "learning_rate": 3.834785193787571e-05, + "loss": 0.316, + "step": 30353 + }, + { + "epoch": 0.7115455616308304, + "grad_norm": 1.944688081741333, + "learning_rate": 3.834205244852521e-05, + "loss": 0.149, + "step": 30354 + }, + { + "epoch": 0.7115690032056353, + "grad_norm": 1.8857895135879517, + "learning_rate": 3.8336253293732485e-05, + "loss": 0.2559, + "step": 30355 + }, + { + "epoch": 0.7115924447804404, + "grad_norm": 2.279219150543213, + "learning_rate": 3.8330454473528933e-05, + "loss": 0.2673, + "step": 30356 + }, + { + "epoch": 0.7116158863552453, + "grad_norm": 1.094398856163025, + "learning_rate": 3.8324655987946113e-05, + "loss": 0.0892, + "step": 30357 + }, + { + "epoch": 0.7116393279300504, + "grad_norm": 3.5801453590393066, + "learning_rate": 3.831885783701543e-05, + "loss": 0.2251, + "step": 30358 + }, + { + "epoch": 0.7116627695048553, + "grad_norm": 0.9900601506233215, + "learning_rate": 3.8313060020768375e-05, + "loss": 0.092, + "step": 30359 + }, + { + "epoch": 0.7116862110796603, + "grad_norm": 1.8602879047393799, + "learning_rate": 3.8307262539236374e-05, + "loss": 0.0661, + "step": 30360 + }, + { + "epoch": 0.7117096526544653, + "grad_norm": 4.095069885253906, + "learning_rate": 3.8301465392450886e-05, + "loss": 0.4581, + "step": 30361 + }, + { + "epoch": 0.7117330942292703, + "grad_norm": 3.7616991996765137, + "learning_rate": 3.82956685804434e-05, + "loss": 0.3218, + "step": 30362 + }, + { + "epoch": 0.7117565358040753, + "grad_norm": 0.9114193916320801, + "learning_rate": 3.8289872103245325e-05, + "loss": 0.0652, + "step": 30363 + }, + { + "epoch": 0.7117799773788803, + "grad_norm": 4.24197244644165, + "learning_rate": 3.828407596088817e-05, + "loss": 0.4648, + "step": 30364 + }, + { + "epoch": 0.7118034189536853, + "grad_norm": 3.7065751552581787, + "learning_rate": 3.827828015340335e-05, + "loss": 0.3209, + "step": 30365 + }, + { + "epoch": 0.7118268605284903, + "grad_norm": 3.430041551589966, + "learning_rate": 3.827248468082233e-05, + "loss": 0.2796, + "step": 30366 + }, + { + "epoch": 0.7118503021032953, + "grad_norm": 1.0055783987045288, + "learning_rate": 3.8266689543176534e-05, + "loss": 0.0312, + "step": 30367 + }, + { + "epoch": 0.7118737436781003, + "grad_norm": 2.112088918685913, + "learning_rate": 3.826089474049739e-05, + "loss": 0.082, + "step": 30368 + }, + { + "epoch": 0.7118971852529052, + "grad_norm": 3.098689079284668, + "learning_rate": 3.825510027281639e-05, + "loss": 0.2308, + "step": 30369 + }, + { + "epoch": 0.7119206268277103, + "grad_norm": 2.0422308444976807, + "learning_rate": 3.824930614016496e-05, + "loss": 0.1725, + "step": 30370 + }, + { + "epoch": 0.7119440684025152, + "grad_norm": 0.5172281861305237, + "learning_rate": 3.824351234257453e-05, + "loss": 0.0238, + "step": 30371 + }, + { + "epoch": 0.7119675099773203, + "grad_norm": 0.7487195730209351, + "learning_rate": 3.823771888007653e-05, + "loss": 0.0361, + "step": 30372 + }, + { + "epoch": 0.7119909515521252, + "grad_norm": 3.392258882522583, + "learning_rate": 3.823192575270242e-05, + "loss": 0.1632, + "step": 30373 + }, + { + "epoch": 0.7120143931269303, + "grad_norm": 2.4171864986419678, + "learning_rate": 3.8226132960483575e-05, + "loss": 0.0991, + "step": 30374 + }, + { + "epoch": 0.7120378347017353, + "grad_norm": 2.2779576778411865, + "learning_rate": 3.822034050345147e-05, + "loss": 0.1535, + "step": 30375 + }, + { + "epoch": 0.7120612762765403, + "grad_norm": 1.4962517023086548, + "learning_rate": 3.821454838163757e-05, + "loss": 0.21, + "step": 30376 + }, + { + "epoch": 0.7120847178513453, + "grad_norm": 1.0272496938705444, + "learning_rate": 3.820875659507327e-05, + "loss": 0.0496, + "step": 30377 + }, + { + "epoch": 0.7121081594261502, + "grad_norm": 4.266015529632568, + "learning_rate": 3.820296514378998e-05, + "loss": 0.4408, + "step": 30378 + }, + { + "epoch": 0.7121316010009553, + "grad_norm": 0.9856177568435669, + "learning_rate": 3.819717402781915e-05, + "loss": 0.0988, + "step": 30379 + }, + { + "epoch": 0.7121550425757602, + "grad_norm": 5.069182872772217, + "learning_rate": 3.8191383247192195e-05, + "loss": 0.442, + "step": 30380 + }, + { + "epoch": 0.7121784841505653, + "grad_norm": 3.125300884246826, + "learning_rate": 3.81855928019405e-05, + "loss": 0.1794, + "step": 30381 + }, + { + "epoch": 0.7122019257253702, + "grad_norm": 2.5762343406677246, + "learning_rate": 3.817980269209553e-05, + "loss": 0.2217, + "step": 30382 + }, + { + "epoch": 0.7122253673001753, + "grad_norm": 1.9427611827850342, + "learning_rate": 3.8174012917688706e-05, + "loss": 1.3852, + "step": 30383 + }, + { + "epoch": 0.7122488088749802, + "grad_norm": 2.5888078212738037, + "learning_rate": 3.816822347875143e-05, + "loss": 0.2248, + "step": 30384 + }, + { + "epoch": 0.7122722504497853, + "grad_norm": 3.3701937198638916, + "learning_rate": 3.816243437531511e-05, + "loss": 0.2502, + "step": 30385 + }, + { + "epoch": 0.7122956920245902, + "grad_norm": 2.20222806930542, + "learning_rate": 3.815664560741111e-05, + "loss": 1.8012, + "step": 30386 + }, + { + "epoch": 0.7123191335993952, + "grad_norm": 3.6151645183563232, + "learning_rate": 3.815085717507094e-05, + "loss": 0.2778, + "step": 30387 + }, + { + "epoch": 0.7123425751742002, + "grad_norm": 1.684730052947998, + "learning_rate": 3.8145069078325926e-05, + "loss": 0.9285, + "step": 30388 + }, + { + "epoch": 0.7123660167490052, + "grad_norm": 1.9090816974639893, + "learning_rate": 3.8139281317207534e-05, + "loss": 0.1637, + "step": 30389 + }, + { + "epoch": 0.7123894583238102, + "grad_norm": 0.25802871584892273, + "learning_rate": 3.813349389174713e-05, + "loss": 0.0098, + "step": 30390 + }, + { + "epoch": 0.7124128998986152, + "grad_norm": 2.030951738357544, + "learning_rate": 3.8127706801976135e-05, + "loss": 1.4897, + "step": 30391 + }, + { + "epoch": 0.7124363414734202, + "grad_norm": 3.455883264541626, + "learning_rate": 3.812192004792595e-05, + "loss": 0.3939, + "step": 30392 + }, + { + "epoch": 0.7124597830482252, + "grad_norm": 5.155013084411621, + "learning_rate": 3.811613362962796e-05, + "loss": 0.6157, + "step": 30393 + }, + { + "epoch": 0.7124832246230302, + "grad_norm": 0.6083592772483826, + "learning_rate": 3.8110347547113545e-05, + "loss": 0.0527, + "step": 30394 + }, + { + "epoch": 0.7125066661978352, + "grad_norm": 2.699218988418579, + "learning_rate": 3.810456180041415e-05, + "loss": 0.2425, + "step": 30395 + }, + { + "epoch": 0.7125301077726401, + "grad_norm": 1.3305864334106445, + "learning_rate": 3.809877638956114e-05, + "loss": 1.1182, + "step": 30396 + }, + { + "epoch": 0.7125535493474452, + "grad_norm": 1.8240641355514526, + "learning_rate": 3.8092991314585914e-05, + "loss": 0.1151, + "step": 30397 + }, + { + "epoch": 0.7125769909222501, + "grad_norm": 3.5770130157470703, + "learning_rate": 3.808720657551986e-05, + "loss": 0.3144, + "step": 30398 + }, + { + "epoch": 0.7126004324970552, + "grad_norm": 4.078422546386719, + "learning_rate": 3.808142217239432e-05, + "loss": 0.4563, + "step": 30399 + }, + { + "epoch": 0.7126238740718601, + "grad_norm": 1.7985516786575317, + "learning_rate": 3.8075638105240777e-05, + "loss": 0.1777, + "step": 30400 + }, + { + "epoch": 0.7126473156466652, + "grad_norm": 3.278442144393921, + "learning_rate": 3.806985437409051e-05, + "loss": 0.2776, + "step": 30401 + }, + { + "epoch": 0.7126707572214701, + "grad_norm": 7.385042190551758, + "learning_rate": 3.8064070978975e-05, + "loss": 0.5603, + "step": 30402 + }, + { + "epoch": 0.7126941987962752, + "grad_norm": 1.98713219165802, + "learning_rate": 3.805828791992557e-05, + "loss": 0.0735, + "step": 30403 + }, + { + "epoch": 0.7127176403710801, + "grad_norm": 1.9868961572647095, + "learning_rate": 3.805250519697362e-05, + "loss": 0.1354, + "step": 30404 + }, + { + "epoch": 0.7127410819458851, + "grad_norm": 1.6482915878295898, + "learning_rate": 3.804672281015053e-05, + "loss": 0.08, + "step": 30405 + }, + { + "epoch": 0.7127645235206901, + "grad_norm": 2.9007177352905273, + "learning_rate": 3.8040940759487606e-05, + "loss": 0.2575, + "step": 30406 + }, + { + "epoch": 0.7127879650954951, + "grad_norm": 3.112107038497925, + "learning_rate": 3.803515904501633e-05, + "loss": 0.2456, + "step": 30407 + }, + { + "epoch": 0.7128114066703001, + "grad_norm": 3.045057535171509, + "learning_rate": 3.8029377666768015e-05, + "loss": 0.1793, + "step": 30408 + }, + { + "epoch": 0.7128348482451051, + "grad_norm": 3.3880772590637207, + "learning_rate": 3.802359662477405e-05, + "loss": 0.3352, + "step": 30409 + }, + { + "epoch": 0.7128582898199101, + "grad_norm": 1.6670076847076416, + "learning_rate": 3.801781591906578e-05, + "loss": 0.0776, + "step": 30410 + }, + { + "epoch": 0.7128817313947151, + "grad_norm": 4.763289451599121, + "learning_rate": 3.801203554967456e-05, + "loss": 0.2826, + "step": 30411 + }, + { + "epoch": 0.71290517296952, + "grad_norm": 0.740927517414093, + "learning_rate": 3.800625551663181e-05, + "loss": 0.0507, + "step": 30412 + }, + { + "epoch": 0.7129286145443251, + "grad_norm": 1.6831780672073364, + "learning_rate": 3.8000475819968816e-05, + "loss": 0.0777, + "step": 30413 + }, + { + "epoch": 0.71295205611913, + "grad_norm": 0.8218187689781189, + "learning_rate": 3.799469645971703e-05, + "loss": 0.084, + "step": 30414 + }, + { + "epoch": 0.7129754976939351, + "grad_norm": 2.245008707046509, + "learning_rate": 3.798891743590776e-05, + "loss": 0.1229, + "step": 30415 + }, + { + "epoch": 0.71299893926874, + "grad_norm": 1.7617928981781006, + "learning_rate": 3.798313874857237e-05, + "loss": 0.1591, + "step": 30416 + }, + { + "epoch": 0.7130223808435451, + "grad_norm": 1.890709638595581, + "learning_rate": 3.7977360397742214e-05, + "loss": 0.1968, + "step": 30417 + }, + { + "epoch": 0.71304582241835, + "grad_norm": 4.523632526397705, + "learning_rate": 3.7971582383448635e-05, + "loss": 0.3603, + "step": 30418 + }, + { + "epoch": 0.7130692639931551, + "grad_norm": 1.0563546419143677, + "learning_rate": 3.796580470572296e-05, + "loss": 0.0837, + "step": 30419 + }, + { + "epoch": 0.71309270556796, + "grad_norm": 1.3118915557861328, + "learning_rate": 3.796002736459662e-05, + "loss": 0.0594, + "step": 30420 + }, + { + "epoch": 0.7131161471427651, + "grad_norm": 1.4141261577606201, + "learning_rate": 3.7954250360100906e-05, + "loss": 0.2262, + "step": 30421 + }, + { + "epoch": 0.71313958871757, + "grad_norm": 1.5976054668426514, + "learning_rate": 3.794847369226717e-05, + "loss": 1.2141, + "step": 30422 + }, + { + "epoch": 0.713163030292375, + "grad_norm": 2.540724039077759, + "learning_rate": 3.794269736112675e-05, + "loss": 0.1918, + "step": 30423 + }, + { + "epoch": 0.71318647186718, + "grad_norm": 0.691222071647644, + "learning_rate": 3.793692136671101e-05, + "loss": 0.034, + "step": 30424 + }, + { + "epoch": 0.713209913441985, + "grad_norm": 2.2289721965789795, + "learning_rate": 3.793114570905128e-05, + "loss": 0.1372, + "step": 30425 + }, + { + "epoch": 0.7132333550167901, + "grad_norm": 4.747527599334717, + "learning_rate": 3.792537038817887e-05, + "loss": 0.3353, + "step": 30426 + }, + { + "epoch": 0.713256796591595, + "grad_norm": 3.9620141983032227, + "learning_rate": 3.791959540412519e-05, + "loss": 0.264, + "step": 30427 + }, + { + "epoch": 0.7132802381664001, + "grad_norm": 1.179604172706604, + "learning_rate": 3.7913820756921514e-05, + "loss": 0.1312, + "step": 30428 + }, + { + "epoch": 0.713303679741205, + "grad_norm": 2.8618733882904053, + "learning_rate": 3.79080464465992e-05, + "loss": 0.3114, + "step": 30429 + }, + { + "epoch": 0.7133271213160101, + "grad_norm": 4.164484977722168, + "learning_rate": 3.790227247318957e-05, + "loss": 0.1695, + "step": 30430 + }, + { + "epoch": 0.713350562890815, + "grad_norm": 3.272195816040039, + "learning_rate": 3.789649883672396e-05, + "loss": 0.4353, + "step": 30431 + }, + { + "epoch": 0.71337400446562, + "grad_norm": 4.382615089416504, + "learning_rate": 3.789072553723365e-05, + "loss": 0.2655, + "step": 30432 + }, + { + "epoch": 0.713397446040425, + "grad_norm": 2.433899402618408, + "learning_rate": 3.7884952574750055e-05, + "loss": 0.153, + "step": 30433 + }, + { + "epoch": 0.71342088761523, + "grad_norm": 0.8500926494598389, + "learning_rate": 3.787917994930444e-05, + "loss": 0.0469, + "step": 30434 + }, + { + "epoch": 0.713444329190035, + "grad_norm": 0.5369430184364319, + "learning_rate": 3.787340766092813e-05, + "loss": 0.0453, + "step": 30435 + }, + { + "epoch": 0.71346777076484, + "grad_norm": 2.562934637069702, + "learning_rate": 3.786763570965247e-05, + "loss": 0.1785, + "step": 30436 + }, + { + "epoch": 0.713491212339645, + "grad_norm": 2.7138679027557373, + "learning_rate": 3.786186409550877e-05, + "loss": 0.1152, + "step": 30437 + }, + { + "epoch": 0.71351465391445, + "grad_norm": 2.3672635555267334, + "learning_rate": 3.785609281852831e-05, + "loss": 1.379, + "step": 30438 + }, + { + "epoch": 0.713538095489255, + "grad_norm": 2.826932191848755, + "learning_rate": 3.7850321878742476e-05, + "loss": 0.181, + "step": 30439 + }, + { + "epoch": 0.71356153706406, + "grad_norm": 2.964733600616455, + "learning_rate": 3.784455127618254e-05, + "loss": 0.154, + "step": 30440 + }, + { + "epoch": 0.7135849786388649, + "grad_norm": 1.9886276721954346, + "learning_rate": 3.783878101087982e-05, + "loss": 0.1324, + "step": 30441 + }, + { + "epoch": 0.71360842021367, + "grad_norm": 1.3004010915756226, + "learning_rate": 3.7833011082865625e-05, + "loss": 0.0346, + "step": 30442 + }, + { + "epoch": 0.7136318617884749, + "grad_norm": 6.184518337249756, + "learning_rate": 3.7827241492171253e-05, + "loss": 0.4054, + "step": 30443 + }, + { + "epoch": 0.71365530336328, + "grad_norm": 3.498878240585327, + "learning_rate": 3.7821472238827974e-05, + "loss": 0.3046, + "step": 30444 + }, + { + "epoch": 0.7136787449380849, + "grad_norm": 2.256990909576416, + "learning_rate": 3.781570332286718e-05, + "loss": 0.1049, + "step": 30445 + }, + { + "epoch": 0.71370218651289, + "grad_norm": 1.9834470748901367, + "learning_rate": 3.780993474432013e-05, + "loss": 0.2134, + "step": 30446 + }, + { + "epoch": 0.7137256280876949, + "grad_norm": 0.658376157283783, + "learning_rate": 3.7804166503218086e-05, + "loss": 0.0471, + "step": 30447 + }, + { + "epoch": 0.7137490696625, + "grad_norm": 2.389082431793213, + "learning_rate": 3.779839859959241e-05, + "loss": 1.5736, + "step": 30448 + }, + { + "epoch": 0.7137725112373049, + "grad_norm": 5.154268741607666, + "learning_rate": 3.779263103347438e-05, + "loss": 0.7335, + "step": 30449 + }, + { + "epoch": 0.71379595281211, + "grad_norm": 3.6895666122436523, + "learning_rate": 3.7786863804895275e-05, + "loss": 0.2694, + "step": 30450 + }, + { + "epoch": 0.7138193943869149, + "grad_norm": 4.196305274963379, + "learning_rate": 3.778109691388637e-05, + "loss": 0.2353, + "step": 30451 + }, + { + "epoch": 0.7138428359617199, + "grad_norm": 3.320651054382324, + "learning_rate": 3.777533036047902e-05, + "loss": 0.3545, + "step": 30452 + }, + { + "epoch": 0.7138662775365249, + "grad_norm": 1.7258905172348022, + "learning_rate": 3.776956414470447e-05, + "loss": 0.1779, + "step": 30453 + }, + { + "epoch": 0.7138897191113299, + "grad_norm": 3.5965094566345215, + "learning_rate": 3.776379826659402e-05, + "loss": 0.3914, + "step": 30454 + }, + { + "epoch": 0.7139131606861349, + "grad_norm": 3.1559934616088867, + "learning_rate": 3.775803272617895e-05, + "loss": 0.2271, + "step": 30455 + }, + { + "epoch": 0.7139366022609399, + "grad_norm": 1.274346113204956, + "learning_rate": 3.7752267523490537e-05, + "loss": 1.0476, + "step": 30456 + }, + { + "epoch": 0.7139600438357449, + "grad_norm": 3.4311625957489014, + "learning_rate": 3.774650265856005e-05, + "loss": 0.2449, + "step": 30457 + }, + { + "epoch": 0.7139834854105499, + "grad_norm": 2.194105863571167, + "learning_rate": 3.7740738131418815e-05, + "loss": 0.1561, + "step": 30458 + }, + { + "epoch": 0.7140069269853548, + "grad_norm": 2.9862911701202393, + "learning_rate": 3.7734973942098053e-05, + "loss": 0.3581, + "step": 30459 + }, + { + "epoch": 0.7140303685601599, + "grad_norm": 0.5201568603515625, + "learning_rate": 3.772921009062911e-05, + "loss": 0.0364, + "step": 30460 + }, + { + "epoch": 0.7140538101349648, + "grad_norm": 0.7688965201377869, + "learning_rate": 3.7723446577043206e-05, + "loss": 0.0713, + "step": 30461 + }, + { + "epoch": 0.7140772517097699, + "grad_norm": 2.320736885070801, + "learning_rate": 3.771768340137165e-05, + "loss": 0.1904, + "step": 30462 + }, + { + "epoch": 0.7141006932845748, + "grad_norm": 4.094718933105469, + "learning_rate": 3.7711920563645696e-05, + "loss": 0.4003, + "step": 30463 + }, + { + "epoch": 0.7141241348593799, + "grad_norm": 1.0030757188796997, + "learning_rate": 3.770615806389657e-05, + "loss": 0.0562, + "step": 30464 + }, + { + "epoch": 0.7141475764341848, + "grad_norm": 3.228437900543213, + "learning_rate": 3.7700395902155624e-05, + "loss": 0.4313, + "step": 30465 + }, + { + "epoch": 0.7141710180089899, + "grad_norm": 2.5301547050476074, + "learning_rate": 3.769463407845407e-05, + "loss": 0.1137, + "step": 30466 + }, + { + "epoch": 0.7141944595837948, + "grad_norm": 1.8697447776794434, + "learning_rate": 3.76888725928232e-05, + "loss": 0.1609, + "step": 30467 + }, + { + "epoch": 0.7142179011585998, + "grad_norm": 2.056730031967163, + "learning_rate": 3.768311144529425e-05, + "loss": 0.2354, + "step": 30468 + }, + { + "epoch": 0.7142413427334048, + "grad_norm": 2.990234613418579, + "learning_rate": 3.767735063589849e-05, + "loss": 0.1426, + "step": 30469 + }, + { + "epoch": 0.7142647843082098, + "grad_norm": 0.22359073162078857, + "learning_rate": 3.7671590164667135e-05, + "loss": 0.0192, + "step": 30470 + }, + { + "epoch": 0.7142882258830148, + "grad_norm": 1.2044726610183716, + "learning_rate": 3.76658300316315e-05, + "loss": 0.4667, + "step": 30471 + }, + { + "epoch": 0.7143116674578198, + "grad_norm": 2.6262547969818115, + "learning_rate": 3.766007023682285e-05, + "loss": 0.1793, + "step": 30472 + }, + { + "epoch": 0.7143351090326248, + "grad_norm": 3.729388952255249, + "learning_rate": 3.765431078027241e-05, + "loss": 0.4214, + "step": 30473 + }, + { + "epoch": 0.7143585506074298, + "grad_norm": 2.3766467571258545, + "learning_rate": 3.764855166201143e-05, + "loss": 0.3402, + "step": 30474 + }, + { + "epoch": 0.7143819921822347, + "grad_norm": 2.5212082862854004, + "learning_rate": 3.764279288207117e-05, + "loss": 1.1417, + "step": 30475 + }, + { + "epoch": 0.7144054337570398, + "grad_norm": 1.5673242807388306, + "learning_rate": 3.7637034440482834e-05, + "loss": 0.1362, + "step": 30476 + }, + { + "epoch": 0.7144288753318448, + "grad_norm": 1.2678419351577759, + "learning_rate": 3.763127633727773e-05, + "loss": 0.0944, + "step": 30477 + }, + { + "epoch": 0.7144523169066498, + "grad_norm": 6.1700897216796875, + "learning_rate": 3.762551857248708e-05, + "loss": 0.5702, + "step": 30478 + }, + { + "epoch": 0.7144757584814548, + "grad_norm": 0.7070662379264832, + "learning_rate": 3.761976114614213e-05, + "loss": 0.0161, + "step": 30479 + }, + { + "epoch": 0.7144992000562598, + "grad_norm": 1.354854702949524, + "learning_rate": 3.7614004058274085e-05, + "loss": 0.0739, + "step": 30480 + }, + { + "epoch": 0.7145226416310648, + "grad_norm": 0.7212334871292114, + "learning_rate": 3.7608247308914236e-05, + "loss": 0.0711, + "step": 30481 + }, + { + "epoch": 0.7145460832058698, + "grad_norm": 2.7271149158477783, + "learning_rate": 3.760249089809374e-05, + "loss": 0.2681, + "step": 30482 + }, + { + "epoch": 0.7145695247806748, + "grad_norm": 3.3552680015563965, + "learning_rate": 3.7596734825843915e-05, + "loss": 0.4871, + "step": 30483 + }, + { + "epoch": 0.7145929663554798, + "grad_norm": 2.236422300338745, + "learning_rate": 3.7590979092195934e-05, + "loss": 1.2208, + "step": 30484 + }, + { + "epoch": 0.7146164079302848, + "grad_norm": 3.1165363788604736, + "learning_rate": 3.758522369718109e-05, + "loss": 0.3024, + "step": 30485 + }, + { + "epoch": 0.7146398495050897, + "grad_norm": 2.178149700164795, + "learning_rate": 3.757946864083057e-05, + "loss": 0.2043, + "step": 30486 + }, + { + "epoch": 0.7146632910798948, + "grad_norm": 1.9214611053466797, + "learning_rate": 3.7573713923175624e-05, + "loss": 0.1371, + "step": 30487 + }, + { + "epoch": 0.7146867326546997, + "grad_norm": 3.5387392044067383, + "learning_rate": 3.7567959544247445e-05, + "loss": 0.2051, + "step": 30488 + }, + { + "epoch": 0.7147101742295048, + "grad_norm": 2.5395984649658203, + "learning_rate": 3.756220550407724e-05, + "loss": 1.4812, + "step": 30489 + }, + { + "epoch": 0.7147336158043097, + "grad_norm": 3.8105854988098145, + "learning_rate": 3.755645180269631e-05, + "loss": 0.268, + "step": 30490 + }, + { + "epoch": 0.7147570573791148, + "grad_norm": 0.7792101502418518, + "learning_rate": 3.755069844013582e-05, + "loss": 0.0768, + "step": 30491 + }, + { + "epoch": 0.7147804989539197, + "grad_norm": 2.519927501678467, + "learning_rate": 3.7544945416426994e-05, + "loss": 0.192, + "step": 30492 + }, + { + "epoch": 0.7148039405287248, + "grad_norm": 2.759667158126831, + "learning_rate": 3.753919273160106e-05, + "loss": 1.6466, + "step": 30493 + }, + { + "epoch": 0.7148273821035297, + "grad_norm": 3.161937952041626, + "learning_rate": 3.7533440385689214e-05, + "loss": 1.1883, + "step": 30494 + }, + { + "epoch": 0.7148508236783347, + "grad_norm": 0.7911866307258606, + "learning_rate": 3.752768837872264e-05, + "loss": 0.0551, + "step": 30495 + }, + { + "epoch": 0.7148742652531397, + "grad_norm": 2.9794845581054688, + "learning_rate": 3.752193671073259e-05, + "loss": 1.5611, + "step": 30496 + }, + { + "epoch": 0.7148977068279447, + "grad_norm": 0.7626209855079651, + "learning_rate": 3.75161853817503e-05, + "loss": 0.0628, + "step": 30497 + }, + { + "epoch": 0.7149211484027497, + "grad_norm": 2.7213149070739746, + "learning_rate": 3.751043439180695e-05, + "loss": 0.1234, + "step": 30498 + }, + { + "epoch": 0.7149445899775547, + "grad_norm": 2.7182745933532715, + "learning_rate": 3.750468374093373e-05, + "loss": 0.2132, + "step": 30499 + }, + { + "epoch": 0.7149680315523597, + "grad_norm": 2.4049315452575684, + "learning_rate": 3.7498933429161865e-05, + "loss": 0.2138, + "step": 30500 + }, + { + "epoch": 0.7149914731271647, + "grad_norm": 3.5542032718658447, + "learning_rate": 3.749318345652254e-05, + "loss": 0.37, + "step": 30501 + }, + { + "epoch": 0.7150149147019697, + "grad_norm": 4.209552764892578, + "learning_rate": 3.7487433823046926e-05, + "loss": 0.1906, + "step": 30502 + }, + { + "epoch": 0.7150383562767747, + "grad_norm": 0.7395839095115662, + "learning_rate": 3.7481684528766294e-05, + "loss": 0.0266, + "step": 30503 + }, + { + "epoch": 0.7150617978515796, + "grad_norm": 2.2518444061279297, + "learning_rate": 3.747593557371179e-05, + "loss": 0.2057, + "step": 30504 + }, + { + "epoch": 0.7150852394263847, + "grad_norm": 1.5848875045776367, + "learning_rate": 3.747018695791462e-05, + "loss": 0.1398, + "step": 30505 + }, + { + "epoch": 0.7151086810011896, + "grad_norm": 3.49143385887146, + "learning_rate": 3.7464438681405965e-05, + "loss": 0.4631, + "step": 30506 + }, + { + "epoch": 0.7151321225759947, + "grad_norm": 2.9548356533050537, + "learning_rate": 3.7458690744217005e-05, + "loss": 0.2035, + "step": 30507 + }, + { + "epoch": 0.7151555641507996, + "grad_norm": 2.6055877208709717, + "learning_rate": 3.745294314637897e-05, + "loss": 0.2448, + "step": 30508 + }, + { + "epoch": 0.7151790057256047, + "grad_norm": 2.218851327896118, + "learning_rate": 3.7447195887923005e-05, + "loss": 0.1245, + "step": 30509 + }, + { + "epoch": 0.7152024473004096, + "grad_norm": 2.518315553665161, + "learning_rate": 3.744144896888033e-05, + "loss": 1.1852, + "step": 30510 + }, + { + "epoch": 0.7152258888752147, + "grad_norm": 2.2005980014801025, + "learning_rate": 3.743570238928212e-05, + "loss": 0.122, + "step": 30511 + }, + { + "epoch": 0.7152493304500196, + "grad_norm": 1.7932648658752441, + "learning_rate": 3.742995614915955e-05, + "loss": 0.1108, + "step": 30512 + }, + { + "epoch": 0.7152727720248246, + "grad_norm": 2.1099977493286133, + "learning_rate": 3.74242102485438e-05, + "loss": 0.1386, + "step": 30513 + }, + { + "epoch": 0.7152962135996296, + "grad_norm": 1.3872673511505127, + "learning_rate": 3.7418464687466e-05, + "loss": 0.0797, + "step": 30514 + }, + { + "epoch": 0.7153196551744346, + "grad_norm": 1.97589910030365, + "learning_rate": 3.741271946595741e-05, + "loss": 0.1236, + "step": 30515 + }, + { + "epoch": 0.7153430967492396, + "grad_norm": 0.7138494253158569, + "learning_rate": 3.740697458404917e-05, + "loss": 0.0494, + "step": 30516 + }, + { + "epoch": 0.7153665383240446, + "grad_norm": 2.1440460681915283, + "learning_rate": 3.740123004177244e-05, + "loss": 0.1822, + "step": 30517 + }, + { + "epoch": 0.7153899798988496, + "grad_norm": 1.8155975341796875, + "learning_rate": 3.73954858391584e-05, + "loss": 0.1969, + "step": 30518 + }, + { + "epoch": 0.7154134214736546, + "grad_norm": 3.5131115913391113, + "learning_rate": 3.738974197623821e-05, + "loss": 0.3632, + "step": 30519 + }, + { + "epoch": 0.7154368630484595, + "grad_norm": 0.70120769739151, + "learning_rate": 3.7383998453043e-05, + "loss": 0.0317, + "step": 30520 + }, + { + "epoch": 0.7154603046232646, + "grad_norm": 3.136627435684204, + "learning_rate": 3.737825526960399e-05, + "loss": 0.213, + "step": 30521 + }, + { + "epoch": 0.7154837461980695, + "grad_norm": 0.7567683458328247, + "learning_rate": 3.7372512425952345e-05, + "loss": 0.113, + "step": 30522 + }, + { + "epoch": 0.7155071877728746, + "grad_norm": 1.4185348749160767, + "learning_rate": 3.7366769922119224e-05, + "loss": 0.0661, + "step": 30523 + }, + { + "epoch": 0.7155306293476795, + "grad_norm": 4.768754959106445, + "learning_rate": 3.736102775813576e-05, + "loss": 0.421, + "step": 30524 + }, + { + "epoch": 0.7155540709224846, + "grad_norm": 2.991163969039917, + "learning_rate": 3.735528593403312e-05, + "loss": 0.288, + "step": 30525 + }, + { + "epoch": 0.7155775124972895, + "grad_norm": 3.237499952316284, + "learning_rate": 3.734954444984247e-05, + "loss": 0.3041, + "step": 30526 + }, + { + "epoch": 0.7156009540720946, + "grad_norm": 1.888382911682129, + "learning_rate": 3.734380330559491e-05, + "loss": 0.1021, + "step": 30527 + }, + { + "epoch": 0.7156243956468996, + "grad_norm": 2.4113352298736572, + "learning_rate": 3.733806250132168e-05, + "loss": 0.1518, + "step": 30528 + }, + { + "epoch": 0.7156478372217046, + "grad_norm": 1.98886239528656, + "learning_rate": 3.733232203705387e-05, + "loss": 0.1449, + "step": 30529 + }, + { + "epoch": 0.7156712787965096, + "grad_norm": 2.8501718044281006, + "learning_rate": 3.732658191282266e-05, + "loss": 0.2037, + "step": 30530 + }, + { + "epoch": 0.7156947203713145, + "grad_norm": 1.862202525138855, + "learning_rate": 3.732084212865916e-05, + "loss": 0.1029, + "step": 30531 + }, + { + "epoch": 0.7157181619461196, + "grad_norm": 2.0358614921569824, + "learning_rate": 3.731510268459451e-05, + "loss": 0.0785, + "step": 30532 + }, + { + "epoch": 0.7157416035209245, + "grad_norm": 2.1506803035736084, + "learning_rate": 3.730936358065991e-05, + "loss": 0.1955, + "step": 30533 + }, + { + "epoch": 0.7157650450957296, + "grad_norm": 2.9738471508026123, + "learning_rate": 3.730362481688643e-05, + "loss": 1.4013, + "step": 30534 + }, + { + "epoch": 0.7157884866705345, + "grad_norm": 1.6528332233428955, + "learning_rate": 3.729788639330527e-05, + "loss": 0.1506, + "step": 30535 + }, + { + "epoch": 0.7158119282453396, + "grad_norm": 2.500304698944092, + "learning_rate": 3.729214830994755e-05, + "loss": 1.7451, + "step": 30536 + }, + { + "epoch": 0.7158353698201445, + "grad_norm": 0.9878610968589783, + "learning_rate": 3.72864105668444e-05, + "loss": 0.1189, + "step": 30537 + }, + { + "epoch": 0.7158588113949496, + "grad_norm": 1.9810054302215576, + "learning_rate": 3.7280673164026944e-05, + "loss": 0.1891, + "step": 30538 + }, + { + "epoch": 0.7158822529697545, + "grad_norm": 2.8272345066070557, + "learning_rate": 3.7274936101526314e-05, + "loss": 0.3853, + "step": 30539 + }, + { + "epoch": 0.7159056945445595, + "grad_norm": 2.0494396686553955, + "learning_rate": 3.726919937937362e-05, + "loss": 0.1617, + "step": 30540 + }, + { + "epoch": 0.7159291361193645, + "grad_norm": 2.502802848815918, + "learning_rate": 3.726346299760005e-05, + "loss": 0.1456, + "step": 30541 + }, + { + "epoch": 0.7159525776941695, + "grad_norm": 1.5968583822250366, + "learning_rate": 3.725772695623668e-05, + "loss": 0.1035, + "step": 30542 + }, + { + "epoch": 0.7159760192689745, + "grad_norm": 3.3015522956848145, + "learning_rate": 3.7251991255314654e-05, + "loss": 0.2627, + "step": 30543 + }, + { + "epoch": 0.7159994608437795, + "grad_norm": 1.1544747352600098, + "learning_rate": 3.724625589486505e-05, + "loss": 0.1005, + "step": 30544 + }, + { + "epoch": 0.7160229024185845, + "grad_norm": 3.0657565593719482, + "learning_rate": 3.724052087491908e-05, + "loss": 0.2235, + "step": 30545 + }, + { + "epoch": 0.7160463439933895, + "grad_norm": 0.9794889092445374, + "learning_rate": 3.723478619550779e-05, + "loss": 0.0509, + "step": 30546 + }, + { + "epoch": 0.7160697855681945, + "grad_norm": 2.652243137359619, + "learning_rate": 3.722905185666229e-05, + "loss": 0.2145, + "step": 30547 + }, + { + "epoch": 0.7160932271429995, + "grad_norm": 1.0889900922775269, + "learning_rate": 3.722331785841374e-05, + "loss": 0.0489, + "step": 30548 + }, + { + "epoch": 0.7161166687178044, + "grad_norm": 0.9493957161903381, + "learning_rate": 3.721758420079324e-05, + "loss": 0.0978, + "step": 30549 + }, + { + "epoch": 0.7161401102926095, + "grad_norm": 1.8900834321975708, + "learning_rate": 3.72118508838319e-05, + "loss": 0.1087, + "step": 30550 + }, + { + "epoch": 0.7161635518674144, + "grad_norm": 3.021550416946411, + "learning_rate": 3.720611790756082e-05, + "loss": 0.2875, + "step": 30551 + }, + { + "epoch": 0.7161869934422195, + "grad_norm": 0.9615432024002075, + "learning_rate": 3.7200385272011075e-05, + "loss": 0.1096, + "step": 30552 + }, + { + "epoch": 0.7162104350170244, + "grad_norm": 0.9762307405471802, + "learning_rate": 3.719465297721383e-05, + "loss": 0.0506, + "step": 30553 + }, + { + "epoch": 0.7162338765918295, + "grad_norm": 0.9588679075241089, + "learning_rate": 3.7188921023200175e-05, + "loss": 0.0759, + "step": 30554 + }, + { + "epoch": 0.7162573181666344, + "grad_norm": 2.263807535171509, + "learning_rate": 3.71831894100012e-05, + "loss": 0.1321, + "step": 30555 + }, + { + "epoch": 0.7162807597414395, + "grad_norm": 0.4816121459007263, + "learning_rate": 3.7177458137648e-05, + "loss": 0.0218, + "step": 30556 + }, + { + "epoch": 0.7163042013162444, + "grad_norm": 1.1042572259902954, + "learning_rate": 3.717172720617165e-05, + "loss": 0.0845, + "step": 30557 + }, + { + "epoch": 0.7163276428910494, + "grad_norm": 0.43147000670433044, + "learning_rate": 3.7165996615603304e-05, + "loss": 0.016, + "step": 30558 + }, + { + "epoch": 0.7163510844658544, + "grad_norm": 1.2254921197891235, + "learning_rate": 3.716026636597399e-05, + "loss": 0.1096, + "step": 30559 + }, + { + "epoch": 0.7163745260406594, + "grad_norm": 2.9095211029052734, + "learning_rate": 3.715453645731488e-05, + "loss": 0.1385, + "step": 30560 + }, + { + "epoch": 0.7163979676154644, + "grad_norm": 1.69048273563385, + "learning_rate": 3.7148806889657016e-05, + "loss": 0.1645, + "step": 30561 + }, + { + "epoch": 0.7164214091902694, + "grad_norm": 0.34731730818748474, + "learning_rate": 3.714307766303149e-05, + "loss": 0.0261, + "step": 30562 + }, + { + "epoch": 0.7164448507650744, + "grad_norm": 9.309684753417969, + "learning_rate": 3.71373487774694e-05, + "loss": 0.3271, + "step": 30563 + }, + { + "epoch": 0.7164682923398794, + "grad_norm": 2.01981782913208, + "learning_rate": 3.713162023300182e-05, + "loss": 0.2213, + "step": 30564 + }, + { + "epoch": 0.7164917339146843, + "grad_norm": 4.628416061401367, + "learning_rate": 3.71258920296598e-05, + "loss": 0.1641, + "step": 30565 + }, + { + "epoch": 0.7165151754894894, + "grad_norm": 2.982710361480713, + "learning_rate": 3.71201641674745e-05, + "loss": 0.2325, + "step": 30566 + }, + { + "epoch": 0.7165386170642943, + "grad_norm": 1.9446684122085571, + "learning_rate": 3.711443664647695e-05, + "loss": 1.2983, + "step": 30567 + }, + { + "epoch": 0.7165620586390994, + "grad_norm": 1.7281625270843506, + "learning_rate": 3.710870946669823e-05, + "loss": 0.1765, + "step": 30568 + }, + { + "epoch": 0.7165855002139043, + "grad_norm": 2.151972770690918, + "learning_rate": 3.71029826281694e-05, + "loss": 0.2401, + "step": 30569 + }, + { + "epoch": 0.7166089417887094, + "grad_norm": 2.300534725189209, + "learning_rate": 3.709725613092159e-05, + "loss": 0.0891, + "step": 30570 + }, + { + "epoch": 0.7166323833635143, + "grad_norm": 0.9148743152618408, + "learning_rate": 3.7091529974985814e-05, + "loss": 0.1485, + "step": 30571 + }, + { + "epoch": 0.7166558249383194, + "grad_norm": 2.688138484954834, + "learning_rate": 3.708580416039316e-05, + "loss": 0.2027, + "step": 30572 + }, + { + "epoch": 0.7166792665131243, + "grad_norm": 1.4205082654953003, + "learning_rate": 3.7080078687174714e-05, + "loss": 0.0923, + "step": 30573 + }, + { + "epoch": 0.7167027080879294, + "grad_norm": 2.228912830352783, + "learning_rate": 3.707435355536154e-05, + "loss": 0.324, + "step": 30574 + }, + { + "epoch": 0.7167261496627343, + "grad_norm": 0.9316880702972412, + "learning_rate": 3.706862876498469e-05, + "loss": 0.0964, + "step": 30575 + }, + { + "epoch": 0.7167495912375393, + "grad_norm": 2.758652687072754, + "learning_rate": 3.706290431607523e-05, + "loss": 0.1511, + "step": 30576 + }, + { + "epoch": 0.7167730328123443, + "grad_norm": 0.857100248336792, + "learning_rate": 3.705718020866422e-05, + "loss": 0.1201, + "step": 30577 + }, + { + "epoch": 0.7167964743871493, + "grad_norm": 2.8771984577178955, + "learning_rate": 3.7051456442782694e-05, + "loss": 0.2041, + "step": 30578 + }, + { + "epoch": 0.7168199159619544, + "grad_norm": 2.580134868621826, + "learning_rate": 3.704573301846176e-05, + "loss": 0.1872, + "step": 30579 + }, + { + "epoch": 0.7168433575367593, + "grad_norm": 1.0419107675552368, + "learning_rate": 3.704000993573246e-05, + "loss": 0.1216, + "step": 30580 + }, + { + "epoch": 0.7168667991115644, + "grad_norm": 1.8949161767959595, + "learning_rate": 3.7034287194625786e-05, + "loss": 0.1011, + "step": 30581 + }, + { + "epoch": 0.7168902406863693, + "grad_norm": 2.296693801879883, + "learning_rate": 3.702856479517288e-05, + "loss": 0.2441, + "step": 30582 + }, + { + "epoch": 0.7169136822611744, + "grad_norm": 0.5197176933288574, + "learning_rate": 3.702284273740476e-05, + "loss": 0.0514, + "step": 30583 + }, + { + "epoch": 0.7169371238359793, + "grad_norm": 0.6228872537612915, + "learning_rate": 3.701712102135245e-05, + "loss": 0.0555, + "step": 30584 + }, + { + "epoch": 0.7169605654107843, + "grad_norm": 3.6430630683898926, + "learning_rate": 3.7011399647047e-05, + "loss": 0.1724, + "step": 30585 + }, + { + "epoch": 0.7169840069855893, + "grad_norm": 3.2687506675720215, + "learning_rate": 3.700567861451948e-05, + "loss": 0.2198, + "step": 30586 + }, + { + "epoch": 0.7170074485603943, + "grad_norm": 5.316707611083984, + "learning_rate": 3.699995792380093e-05, + "loss": 0.3837, + "step": 30587 + }, + { + "epoch": 0.7170308901351993, + "grad_norm": 1.635890245437622, + "learning_rate": 3.699423757492239e-05, + "loss": 0.1036, + "step": 30588 + }, + { + "epoch": 0.7170543317100043, + "grad_norm": 2.412959098815918, + "learning_rate": 3.698851756791487e-05, + "loss": 0.1649, + "step": 30589 + }, + { + "epoch": 0.7170777732848093, + "grad_norm": 2.195862054824829, + "learning_rate": 3.6982797902809395e-05, + "loss": 0.2197, + "step": 30590 + }, + { + "epoch": 0.7171012148596143, + "grad_norm": 3.0495083332061768, + "learning_rate": 3.697707857963707e-05, + "loss": 1.3582, + "step": 30591 + }, + { + "epoch": 0.7171246564344192, + "grad_norm": 3.152538537979126, + "learning_rate": 3.6971359598428876e-05, + "loss": 0.2062, + "step": 30592 + }, + { + "epoch": 0.7171480980092243, + "grad_norm": 2.8748016357421875, + "learning_rate": 3.696564095921584e-05, + "loss": 0.2357, + "step": 30593 + }, + { + "epoch": 0.7171715395840292, + "grad_norm": 2.729552984237671, + "learning_rate": 3.695992266202903e-05, + "loss": 0.1204, + "step": 30594 + }, + { + "epoch": 0.7171949811588343, + "grad_norm": 3.392101287841797, + "learning_rate": 3.695420470689947e-05, + "loss": 0.14, + "step": 30595 + }, + { + "epoch": 0.7172184227336392, + "grad_norm": 1.4548479318618774, + "learning_rate": 3.6948487093858154e-05, + "loss": 0.1493, + "step": 30596 + }, + { + "epoch": 0.7172418643084443, + "grad_norm": 0.9434329867362976, + "learning_rate": 3.694276982293608e-05, + "loss": 0.1191, + "step": 30597 + }, + { + "epoch": 0.7172653058832492, + "grad_norm": 4.166364669799805, + "learning_rate": 3.693705289416436e-05, + "loss": 0.2667, + "step": 30598 + }, + { + "epoch": 0.7172887474580543, + "grad_norm": 2.0406768321990967, + "learning_rate": 3.693133630757395e-05, + "loss": 0.1677, + "step": 30599 + }, + { + "epoch": 0.7173121890328592, + "grad_norm": 1.512343406677246, + "learning_rate": 3.692562006319589e-05, + "loss": 0.057, + "step": 30600 + }, + { + "epoch": 0.7173356306076643, + "grad_norm": 1.739197850227356, + "learning_rate": 3.6919904161061183e-05, + "loss": 0.1094, + "step": 30601 + }, + { + "epoch": 0.7173590721824692, + "grad_norm": 4.744856834411621, + "learning_rate": 3.691418860120086e-05, + "loss": 0.1642, + "step": 30602 + }, + { + "epoch": 0.7173825137572742, + "grad_norm": 0.6339352130889893, + "learning_rate": 3.690847338364588e-05, + "loss": 0.023, + "step": 30603 + }, + { + "epoch": 0.7174059553320792, + "grad_norm": 3.786489725112915, + "learning_rate": 3.690275850842734e-05, + "loss": 0.2215, + "step": 30604 + }, + { + "epoch": 0.7174293969068842, + "grad_norm": 3.465545177459717, + "learning_rate": 3.689704397557616e-05, + "loss": 0.3326, + "step": 30605 + }, + { + "epoch": 0.7174528384816892, + "grad_norm": 4.362977981567383, + "learning_rate": 3.689132978512344e-05, + "loss": 0.306, + "step": 30606 + }, + { + "epoch": 0.7174762800564942, + "grad_norm": 4.359230041503906, + "learning_rate": 3.688561593710013e-05, + "loss": 0.2899, + "step": 30607 + }, + { + "epoch": 0.7174997216312992, + "grad_norm": 3.7334697246551514, + "learning_rate": 3.687990243153725e-05, + "loss": 0.3168, + "step": 30608 + }, + { + "epoch": 0.7175231632061042, + "grad_norm": 2.5075125694274902, + "learning_rate": 3.687418926846579e-05, + "loss": 0.2713, + "step": 30609 + }, + { + "epoch": 0.7175466047809091, + "grad_norm": 2.101243019104004, + "learning_rate": 3.6868476447916724e-05, + "loss": 1.7676, + "step": 30610 + }, + { + "epoch": 0.7175700463557142, + "grad_norm": 2.896939277648926, + "learning_rate": 3.68627639699211e-05, + "loss": 0.3203, + "step": 30611 + }, + { + "epoch": 0.7175934879305191, + "grad_norm": 3.778642177581787, + "learning_rate": 3.6857051834509906e-05, + "loss": 0.4758, + "step": 30612 + }, + { + "epoch": 0.7176169295053242, + "grad_norm": 3.7485549449920654, + "learning_rate": 3.685134004171412e-05, + "loss": 0.4387, + "step": 30613 + }, + { + "epoch": 0.7176403710801291, + "grad_norm": 2.4929041862487793, + "learning_rate": 3.684562859156475e-05, + "loss": 1.5514, + "step": 30614 + }, + { + "epoch": 0.7176638126549342, + "grad_norm": 2.9682846069335938, + "learning_rate": 3.6839917484092766e-05, + "loss": 0.1671, + "step": 30615 + }, + { + "epoch": 0.7176872542297391, + "grad_norm": 1.7808643579483032, + "learning_rate": 3.683420671932912e-05, + "loss": 0.0719, + "step": 30616 + }, + { + "epoch": 0.7177106958045442, + "grad_norm": 2.967393398284912, + "learning_rate": 3.682849629730486e-05, + "loss": 0.2147, + "step": 30617 + }, + { + "epoch": 0.7177341373793491, + "grad_norm": 0.8388933539390564, + "learning_rate": 3.6822786218050984e-05, + "loss": 0.0319, + "step": 30618 + }, + { + "epoch": 0.7177575789541542, + "grad_norm": 2.439758062362671, + "learning_rate": 3.6817076481598456e-05, + "loss": 0.3323, + "step": 30619 + }, + { + "epoch": 0.7177810205289591, + "grad_norm": 2.8594939708709717, + "learning_rate": 3.681136708797823e-05, + "loss": 0.2194, + "step": 30620 + }, + { + "epoch": 0.7178044621037641, + "grad_norm": 2.3012096881866455, + "learning_rate": 3.680565803722132e-05, + "loss": 0.165, + "step": 30621 + }, + { + "epoch": 0.7178279036785691, + "grad_norm": 10.382295608520508, + "learning_rate": 3.679994932935867e-05, + "loss": 0.2868, + "step": 30622 + }, + { + "epoch": 0.7178513452533741, + "grad_norm": 2.3082404136657715, + "learning_rate": 3.6794240964421255e-05, + "loss": 0.2203, + "step": 30623 + }, + { + "epoch": 0.7178747868281791, + "grad_norm": 1.2852578163146973, + "learning_rate": 3.678853294244009e-05, + "loss": 0.0769, + "step": 30624 + }, + { + "epoch": 0.7178982284029841, + "grad_norm": 1.260995864868164, + "learning_rate": 3.678282526344613e-05, + "loss": 0.097, + "step": 30625 + }, + { + "epoch": 0.7179216699777891, + "grad_norm": 3.0689940452575684, + "learning_rate": 3.677711792747033e-05, + "loss": 0.2475, + "step": 30626 + }, + { + "epoch": 0.7179451115525941, + "grad_norm": 2.9554615020751953, + "learning_rate": 3.677141093454366e-05, + "loss": 0.27, + "step": 30627 + }, + { + "epoch": 0.717968553127399, + "grad_norm": 1.7597341537475586, + "learning_rate": 3.676570428469707e-05, + "loss": 0.1362, + "step": 30628 + }, + { + "epoch": 0.7179919947022041, + "grad_norm": 2.8431942462921143, + "learning_rate": 3.675999797796158e-05, + "loss": 1.9235, + "step": 30629 + }, + { + "epoch": 0.7180154362770091, + "grad_norm": 4.9073357582092285, + "learning_rate": 3.675429201436808e-05, + "loss": 0.4937, + "step": 30630 + }, + { + "epoch": 0.7180388778518141, + "grad_norm": 0.67401522397995, + "learning_rate": 3.67485863939476e-05, + "loss": 0.0547, + "step": 30631 + }, + { + "epoch": 0.7180623194266191, + "grad_norm": 3.174072742462158, + "learning_rate": 3.674288111673108e-05, + "loss": 1.0945, + "step": 30632 + }, + { + "epoch": 0.7180857610014241, + "grad_norm": 3.6998140811920166, + "learning_rate": 3.673717618274946e-05, + "loss": 0.2681, + "step": 30633 + }, + { + "epoch": 0.7181092025762291, + "grad_norm": 1.45991051197052, + "learning_rate": 3.673147159203369e-05, + "loss": 0.1551, + "step": 30634 + }, + { + "epoch": 0.7181326441510341, + "grad_norm": 1.7122774124145508, + "learning_rate": 3.6725767344614714e-05, + "loss": 0.1491, + "step": 30635 + }, + { + "epoch": 0.7181560857258391, + "grad_norm": 3.1003360748291016, + "learning_rate": 3.672006344052353e-05, + "loss": 0.3853, + "step": 30636 + }, + { + "epoch": 0.718179527300644, + "grad_norm": 3.5236318111419678, + "learning_rate": 3.671435987979107e-05, + "loss": 0.4873, + "step": 30637 + }, + { + "epoch": 0.7182029688754491, + "grad_norm": 2.215644121170044, + "learning_rate": 3.670865666244827e-05, + "loss": 0.219, + "step": 30638 + }, + { + "epoch": 0.718226410450254, + "grad_norm": 2.3458597660064697, + "learning_rate": 3.670295378852607e-05, + "loss": 0.0483, + "step": 30639 + }, + { + "epoch": 0.7182498520250591, + "grad_norm": 1.1616508960723877, + "learning_rate": 3.669725125805543e-05, + "loss": 0.1531, + "step": 30640 + }, + { + "epoch": 0.718273293599864, + "grad_norm": 1.845895528793335, + "learning_rate": 3.669154907106726e-05, + "loss": 0.0885, + "step": 30641 + }, + { + "epoch": 0.7182967351746691, + "grad_norm": 5.252330303192139, + "learning_rate": 3.6685847227592504e-05, + "loss": 0.4888, + "step": 30642 + }, + { + "epoch": 0.718320176749474, + "grad_norm": 3.4002232551574707, + "learning_rate": 3.6680145727662164e-05, + "loss": 0.1999, + "step": 30643 + }, + { + "epoch": 0.7183436183242791, + "grad_norm": 4.830691337585449, + "learning_rate": 3.6674444571307144e-05, + "loss": 0.334, + "step": 30644 + }, + { + "epoch": 0.718367059899084, + "grad_norm": 1.8496769666671753, + "learning_rate": 3.666874375855836e-05, + "loss": 0.0814, + "step": 30645 + }, + { + "epoch": 0.718390501473889, + "grad_norm": 2.7375543117523193, + "learning_rate": 3.666304328944674e-05, + "loss": 0.2438, + "step": 30646 + }, + { + "epoch": 0.718413943048694, + "grad_norm": 1.4093196392059326, + "learning_rate": 3.665734316400324e-05, + "loss": 0.0685, + "step": 30647 + }, + { + "epoch": 0.718437384623499, + "grad_norm": 3.6823196411132812, + "learning_rate": 3.665164338225875e-05, + "loss": 0.2855, + "step": 30648 + }, + { + "epoch": 0.718460826198304, + "grad_norm": 1.3298368453979492, + "learning_rate": 3.6645943944244254e-05, + "loss": 0.088, + "step": 30649 + }, + { + "epoch": 0.718484267773109, + "grad_norm": 0.9594412446022034, + "learning_rate": 3.6640244849990646e-05, + "loss": 0.0717, + "step": 30650 + }, + { + "epoch": 0.718507709347914, + "grad_norm": 0.9825144410133362, + "learning_rate": 3.663454609952885e-05, + "loss": 0.5071, + "step": 30651 + }, + { + "epoch": 0.718531150922719, + "grad_norm": 0.6887317299842834, + "learning_rate": 3.662884769288979e-05, + "loss": 0.0463, + "step": 30652 + }, + { + "epoch": 0.718554592497524, + "grad_norm": 2.614636182785034, + "learning_rate": 3.662314963010438e-05, + "loss": 0.1723, + "step": 30653 + }, + { + "epoch": 0.718578034072329, + "grad_norm": 3.4830470085144043, + "learning_rate": 3.6617451911203505e-05, + "loss": 0.2658, + "step": 30654 + }, + { + "epoch": 0.718601475647134, + "grad_norm": 3.129753351211548, + "learning_rate": 3.6611754536218126e-05, + "loss": 0.3727, + "step": 30655 + }, + { + "epoch": 0.718624917221939, + "grad_norm": 0.7962478995323181, + "learning_rate": 3.660605750517918e-05, + "loss": 0.082, + "step": 30656 + }, + { + "epoch": 0.7186483587967439, + "grad_norm": 1.1899000406265259, + "learning_rate": 3.660036081811755e-05, + "loss": 0.1019, + "step": 30657 + }, + { + "epoch": 0.718671800371549, + "grad_norm": 4.876618385314941, + "learning_rate": 3.6594664475064146e-05, + "loss": 0.4228, + "step": 30658 + }, + { + "epoch": 0.7186952419463539, + "grad_norm": 4.402978897094727, + "learning_rate": 3.658896847604988e-05, + "loss": 0.2271, + "step": 30659 + }, + { + "epoch": 0.718718683521159, + "grad_norm": 1.614412546157837, + "learning_rate": 3.6583272821105644e-05, + "loss": 0.6804, + "step": 30660 + }, + { + "epoch": 0.7187421250959639, + "grad_norm": 2.0906786918640137, + "learning_rate": 3.6577577510262314e-05, + "loss": 0.1741, + "step": 30661 + }, + { + "epoch": 0.718765566670769, + "grad_norm": 3.687255620956421, + "learning_rate": 3.657188254355089e-05, + "loss": 0.3529, + "step": 30662 + }, + { + "epoch": 0.7187890082455739, + "grad_norm": 1.4662678241729736, + "learning_rate": 3.656618792100219e-05, + "loss": 0.0567, + "step": 30663 + }, + { + "epoch": 0.718812449820379, + "grad_norm": 3.087592601776123, + "learning_rate": 3.6560493642647154e-05, + "loss": 0.2787, + "step": 30664 + }, + { + "epoch": 0.7188358913951839, + "grad_norm": 2.1390726566314697, + "learning_rate": 3.6554799708516655e-05, + "loss": 0.1626, + "step": 30665 + }, + { + "epoch": 0.7188593329699889, + "grad_norm": 2.56789493560791, + "learning_rate": 3.654910611864156e-05, + "loss": 0.2672, + "step": 30666 + }, + { + "epoch": 0.7188827745447939, + "grad_norm": 2.22670578956604, + "learning_rate": 3.654341287305284e-05, + "loss": 0.1801, + "step": 30667 + }, + { + "epoch": 0.7189062161195989, + "grad_norm": 2.2949838638305664, + "learning_rate": 3.65377199717813e-05, + "loss": 0.1318, + "step": 30668 + }, + { + "epoch": 0.7189296576944039, + "grad_norm": 2.069131851196289, + "learning_rate": 3.653202741485793e-05, + "loss": 0.026, + "step": 30669 + }, + { + "epoch": 0.7189530992692089, + "grad_norm": 2.4151182174682617, + "learning_rate": 3.652633520231355e-05, + "loss": 1.8594, + "step": 30670 + }, + { + "epoch": 0.7189765408440139, + "grad_norm": 0.9033542275428772, + "learning_rate": 3.652064333417906e-05, + "loss": 0.0841, + "step": 30671 + }, + { + "epoch": 0.7189999824188189, + "grad_norm": 2.5833899974823, + "learning_rate": 3.651495181048536e-05, + "loss": 0.2445, + "step": 30672 + }, + { + "epoch": 0.7190234239936238, + "grad_norm": 1.8332384824752808, + "learning_rate": 3.6509260631263254e-05, + "loss": 0.1443, + "step": 30673 + }, + { + "epoch": 0.7190468655684289, + "grad_norm": 1.7989038228988647, + "learning_rate": 3.6503569796543735e-05, + "loss": 1.4537, + "step": 30674 + }, + { + "epoch": 0.7190703071432338, + "grad_norm": 3.2687010765075684, + "learning_rate": 3.649787930635764e-05, + "loss": 0.2822, + "step": 30675 + }, + { + "epoch": 0.7190937487180389, + "grad_norm": 2.50227689743042, + "learning_rate": 3.6492189160735824e-05, + "loss": 0.1266, + "step": 30676 + }, + { + "epoch": 0.7191171902928438, + "grad_norm": 2.918654680252075, + "learning_rate": 3.648649935970918e-05, + "loss": 1.5412, + "step": 30677 + }, + { + "epoch": 0.7191406318676489, + "grad_norm": 2.373079299926758, + "learning_rate": 3.6480809903308545e-05, + "loss": 0.1718, + "step": 30678 + }, + { + "epoch": 0.7191640734424538, + "grad_norm": 2.962355852127075, + "learning_rate": 3.647512079156485e-05, + "loss": 0.2923, + "step": 30679 + }, + { + "epoch": 0.7191875150172589, + "grad_norm": 1.6513689756393433, + "learning_rate": 3.6469432024508906e-05, + "loss": 0.9841, + "step": 30680 + }, + { + "epoch": 0.7192109565920639, + "grad_norm": 2.3398239612579346, + "learning_rate": 3.646374360217164e-05, + "loss": 1.1822, + "step": 30681 + }, + { + "epoch": 0.7192343981668688, + "grad_norm": 1.1769245862960815, + "learning_rate": 3.645805552458389e-05, + "loss": 0.0911, + "step": 30682 + }, + { + "epoch": 0.7192578397416739, + "grad_norm": 2.1405885219573975, + "learning_rate": 3.645236779177652e-05, + "loss": 0.1316, + "step": 30683 + }, + { + "epoch": 0.7192812813164788, + "grad_norm": 3.6991381645202637, + "learning_rate": 3.6446680403780386e-05, + "loss": 0.3926, + "step": 30684 + }, + { + "epoch": 0.7193047228912839, + "grad_norm": 1.6607781648635864, + "learning_rate": 3.644099336062635e-05, + "loss": 0.1002, + "step": 30685 + }, + { + "epoch": 0.7193281644660888, + "grad_norm": 0.7553552389144897, + "learning_rate": 3.6435306662345236e-05, + "loss": 0.0711, + "step": 30686 + }, + { + "epoch": 0.7193516060408939, + "grad_norm": 1.4825457334518433, + "learning_rate": 3.642962030896798e-05, + "loss": 1.163, + "step": 30687 + }, + { + "epoch": 0.7193750476156988, + "grad_norm": 4.484459400177002, + "learning_rate": 3.642393430052539e-05, + "loss": 0.3366, + "step": 30688 + }, + { + "epoch": 0.7193984891905039, + "grad_norm": 3.731761932373047, + "learning_rate": 3.6418248637048304e-05, + "loss": 0.1668, + "step": 30689 + }, + { + "epoch": 0.7194219307653088, + "grad_norm": 1.8184385299682617, + "learning_rate": 3.641256331856757e-05, + "loss": 0.1043, + "step": 30690 + }, + { + "epoch": 0.7194453723401139, + "grad_norm": 3.3731002807617188, + "learning_rate": 3.640687834511408e-05, + "loss": 0.2506, + "step": 30691 + }, + { + "epoch": 0.7194688139149188, + "grad_norm": 3.5399816036224365, + "learning_rate": 3.6401193716718664e-05, + "loss": 0.2392, + "step": 30692 + }, + { + "epoch": 0.7194922554897238, + "grad_norm": 4.385914325714111, + "learning_rate": 3.639550943341212e-05, + "loss": 0.2472, + "step": 30693 + }, + { + "epoch": 0.7195156970645288, + "grad_norm": 2.8202648162841797, + "learning_rate": 3.638982549522536e-05, + "loss": 1.1296, + "step": 30694 + }, + { + "epoch": 0.7195391386393338, + "grad_norm": 3.0151309967041016, + "learning_rate": 3.638414190218921e-05, + "loss": 0.1704, + "step": 30695 + }, + { + "epoch": 0.7195625802141388, + "grad_norm": 1.1519361734390259, + "learning_rate": 3.637845865433448e-05, + "loss": 0.1086, + "step": 30696 + }, + { + "epoch": 0.7195860217889438, + "grad_norm": 2.307586669921875, + "learning_rate": 3.6372775751692033e-05, + "loss": 0.0903, + "step": 30697 + }, + { + "epoch": 0.7196094633637488, + "grad_norm": 1.7392312288284302, + "learning_rate": 3.636709319429269e-05, + "loss": 1.0538, + "step": 30698 + }, + { + "epoch": 0.7196329049385538, + "grad_norm": 0.48448798060417175, + "learning_rate": 3.636141098216726e-05, + "loss": 0.038, + "step": 30699 + }, + { + "epoch": 0.7196563465133587, + "grad_norm": 3.515716075897217, + "learning_rate": 3.635572911534662e-05, + "loss": 0.2752, + "step": 30700 + }, + { + "epoch": 0.7196797880881638, + "grad_norm": 2.5055253505706787, + "learning_rate": 3.6350047593861594e-05, + "loss": 0.2077, + "step": 30701 + }, + { + "epoch": 0.7197032296629687, + "grad_norm": 1.8126178979873657, + "learning_rate": 3.634436641774297e-05, + "loss": 0.1254, + "step": 30702 + }, + { + "epoch": 0.7197266712377738, + "grad_norm": 2.2659850120544434, + "learning_rate": 3.6338685587021626e-05, + "loss": 0.3253, + "step": 30703 + }, + { + "epoch": 0.7197501128125787, + "grad_norm": 3.1357381343841553, + "learning_rate": 3.633300510172838e-05, + "loss": 1.3018, + "step": 30704 + }, + { + "epoch": 0.7197735543873838, + "grad_norm": 1.5871275663375854, + "learning_rate": 3.632732496189403e-05, + "loss": 0.775, + "step": 30705 + }, + { + "epoch": 0.7197969959621887, + "grad_norm": 0.8763269782066345, + "learning_rate": 3.632164516754937e-05, + "loss": 0.0859, + "step": 30706 + }, + { + "epoch": 0.7198204375369938, + "grad_norm": 0.9508457183837891, + "learning_rate": 3.6315965718725284e-05, + "loss": 0.5847, + "step": 30707 + }, + { + "epoch": 0.7198438791117987, + "grad_norm": 3.303016185760498, + "learning_rate": 3.631028661545256e-05, + "loss": 0.2965, + "step": 30708 + }, + { + "epoch": 0.7198673206866038, + "grad_norm": 1.815531611442566, + "learning_rate": 3.630460785776202e-05, + "loss": 1.1421, + "step": 30709 + }, + { + "epoch": 0.7198907622614087, + "grad_norm": 1.7925587892532349, + "learning_rate": 3.629892944568446e-05, + "loss": 0.075, + "step": 30710 + }, + { + "epoch": 0.7199142038362137, + "grad_norm": 2.6632442474365234, + "learning_rate": 3.629325137925067e-05, + "loss": 0.1953, + "step": 30711 + }, + { + "epoch": 0.7199376454110187, + "grad_norm": 0.9098001718521118, + "learning_rate": 3.6287573658491525e-05, + "loss": 0.0711, + "step": 30712 + }, + { + "epoch": 0.7199610869858237, + "grad_norm": 2.327511787414551, + "learning_rate": 3.628189628343779e-05, + "loss": 0.2378, + "step": 30713 + }, + { + "epoch": 0.7199845285606287, + "grad_norm": 0.8091129660606384, + "learning_rate": 3.627621925412028e-05, + "loss": 0.0505, + "step": 30714 + }, + { + "epoch": 0.7200079701354337, + "grad_norm": 5.006060600280762, + "learning_rate": 3.627054257056977e-05, + "loss": 0.3898, + "step": 30715 + }, + { + "epoch": 0.7200314117102387, + "grad_norm": 2.9913928508758545, + "learning_rate": 3.626486623281712e-05, + "loss": 0.2691, + "step": 30716 + }, + { + "epoch": 0.7200548532850437, + "grad_norm": 1.6394354104995728, + "learning_rate": 3.625919024089309e-05, + "loss": 0.0873, + "step": 30717 + }, + { + "epoch": 0.7200782948598486, + "grad_norm": 1.9034779071807861, + "learning_rate": 3.625351459482845e-05, + "loss": 0.1436, + "step": 30718 + }, + { + "epoch": 0.7201017364346537, + "grad_norm": 1.375828742980957, + "learning_rate": 3.6247839294654074e-05, + "loss": 0.1368, + "step": 30719 + }, + { + "epoch": 0.7201251780094586, + "grad_norm": 0.6067838072776794, + "learning_rate": 3.624216434040072e-05, + "loss": 0.0457, + "step": 30720 + }, + { + "epoch": 0.7201486195842637, + "grad_norm": 1.2909791469573975, + "learning_rate": 3.623648973209916e-05, + "loss": 0.0896, + "step": 30721 + }, + { + "epoch": 0.7201720611590686, + "grad_norm": 3.211099863052368, + "learning_rate": 3.6230815469780214e-05, + "loss": 0.3059, + "step": 30722 + }, + { + "epoch": 0.7201955027338737, + "grad_norm": 1.174046277999878, + "learning_rate": 3.622514155347464e-05, + "loss": 0.1114, + "step": 30723 + }, + { + "epoch": 0.7202189443086786, + "grad_norm": 3.6742138862609863, + "learning_rate": 3.621946798321322e-05, + "loss": 0.4177, + "step": 30724 + }, + { + "epoch": 0.7202423858834837, + "grad_norm": 2.7513160705566406, + "learning_rate": 3.6213794759026785e-05, + "loss": 1.0795, + "step": 30725 + }, + { + "epoch": 0.7202658274582886, + "grad_norm": 6.3061628341674805, + "learning_rate": 3.6208121880946114e-05, + "loss": 0.4339, + "step": 30726 + }, + { + "epoch": 0.7202892690330936, + "grad_norm": 3.19156551361084, + "learning_rate": 3.620244934900191e-05, + "loss": 0.2797, + "step": 30727 + }, + { + "epoch": 0.7203127106078986, + "grad_norm": 0.3115478456020355, + "learning_rate": 3.619677716322505e-05, + "loss": 0.03, + "step": 30728 + }, + { + "epoch": 0.7203361521827036, + "grad_norm": 1.7217824459075928, + "learning_rate": 3.619110532364628e-05, + "loss": 0.256, + "step": 30729 + }, + { + "epoch": 0.7203595937575086, + "grad_norm": 3.6659724712371826, + "learning_rate": 3.618543383029635e-05, + "loss": 0.3237, + "step": 30730 + }, + { + "epoch": 0.7203830353323136, + "grad_norm": 5.657516956329346, + "learning_rate": 3.617976268320603e-05, + "loss": 0.5796, + "step": 30731 + }, + { + "epoch": 0.7204064769071186, + "grad_norm": 3.558535099029541, + "learning_rate": 3.6174091882406146e-05, + "loss": 0.3246, + "step": 30732 + }, + { + "epoch": 0.7204299184819236, + "grad_norm": 3.274505853652954, + "learning_rate": 3.616842142792742e-05, + "loss": 1.2267, + "step": 30733 + }, + { + "epoch": 0.7204533600567287, + "grad_norm": 2.243415117263794, + "learning_rate": 3.616275131980065e-05, + "loss": 0.1551, + "step": 30734 + }, + { + "epoch": 0.7204768016315336, + "grad_norm": 3.384434938430786, + "learning_rate": 3.615708155805658e-05, + "loss": 0.2598, + "step": 30735 + }, + { + "epoch": 0.7205002432063387, + "grad_norm": 0.6658962965011597, + "learning_rate": 3.615141214272598e-05, + "loss": 0.0458, + "step": 30736 + }, + { + "epoch": 0.7205236847811436, + "grad_norm": 1.6167731285095215, + "learning_rate": 3.6145743073839575e-05, + "loss": 0.0933, + "step": 30737 + }, + { + "epoch": 0.7205471263559486, + "grad_norm": 4.1939287185668945, + "learning_rate": 3.614007435142821e-05, + "loss": 0.2804, + "step": 30738 + }, + { + "epoch": 0.7205705679307536, + "grad_norm": 0.6258390545845032, + "learning_rate": 3.613440597552256e-05, + "loss": 0.0627, + "step": 30739 + }, + { + "epoch": 0.7205940095055586, + "grad_norm": 1.2332247495651245, + "learning_rate": 3.612873794615345e-05, + "loss": 0.092, + "step": 30740 + }, + { + "epoch": 0.7206174510803636, + "grad_norm": 2.5929534435272217, + "learning_rate": 3.61230702633516e-05, + "loss": 0.3242, + "step": 30741 + }, + { + "epoch": 0.7206408926551686, + "grad_norm": 0.3436654508113861, + "learning_rate": 3.611740292714777e-05, + "loss": 0.0189, + "step": 30742 + }, + { + "epoch": 0.7206643342299736, + "grad_norm": 3.7985949516296387, + "learning_rate": 3.6111735937572696e-05, + "loss": 0.5001, + "step": 30743 + }, + { + "epoch": 0.7206877758047786, + "grad_norm": 2.5512044429779053, + "learning_rate": 3.610606929465712e-05, + "loss": 0.1826, + "step": 30744 + }, + { + "epoch": 0.7207112173795835, + "grad_norm": 3.2457611560821533, + "learning_rate": 3.6100402998431836e-05, + "loss": 0.2197, + "step": 30745 + }, + { + "epoch": 0.7207346589543886, + "grad_norm": 0.9124017357826233, + "learning_rate": 3.609473704892757e-05, + "loss": 0.0775, + "step": 30746 + }, + { + "epoch": 0.7207581005291935, + "grad_norm": 2.1832802295684814, + "learning_rate": 3.6089071446175046e-05, + "loss": 0.2219, + "step": 30747 + }, + { + "epoch": 0.7207815421039986, + "grad_norm": 3.7409157752990723, + "learning_rate": 3.608340619020503e-05, + "loss": 0.1576, + "step": 30748 + }, + { + "epoch": 0.7208049836788035, + "grad_norm": 2.5700125694274902, + "learning_rate": 3.6077741281048216e-05, + "loss": 0.2868, + "step": 30749 + }, + { + "epoch": 0.7208284252536086, + "grad_norm": 4.828099250793457, + "learning_rate": 3.607207671873539e-05, + "loss": 0.2605, + "step": 30750 + }, + { + "epoch": 0.7208518668284135, + "grad_norm": 1.4018337726593018, + "learning_rate": 3.606641250329725e-05, + "loss": 0.074, + "step": 30751 + }, + { + "epoch": 0.7208753084032186, + "grad_norm": 0.8139071464538574, + "learning_rate": 3.6060748634764597e-05, + "loss": 0.0738, + "step": 30752 + }, + { + "epoch": 0.7208987499780235, + "grad_norm": 3.687437057495117, + "learning_rate": 3.605508511316811e-05, + "loss": 0.2, + "step": 30753 + }, + { + "epoch": 0.7209221915528286, + "grad_norm": 1.1543564796447754, + "learning_rate": 3.604942193853853e-05, + "loss": 0.0837, + "step": 30754 + }, + { + "epoch": 0.7209456331276335, + "grad_norm": 1.9763306379318237, + "learning_rate": 3.6043759110906594e-05, + "loss": 1.2366, + "step": 30755 + }, + { + "epoch": 0.7209690747024385, + "grad_norm": 1.5713872909545898, + "learning_rate": 3.603809663030298e-05, + "loss": 0.1, + "step": 30756 + }, + { + "epoch": 0.7209925162772435, + "grad_norm": 1.4286473989486694, + "learning_rate": 3.6032434496758475e-05, + "loss": 0.8078, + "step": 30757 + }, + { + "epoch": 0.7210159578520485, + "grad_norm": 2.972177267074585, + "learning_rate": 3.602677271030379e-05, + "loss": 0.2257, + "step": 30758 + }, + { + "epoch": 0.7210393994268535, + "grad_norm": 3.9773788452148438, + "learning_rate": 3.602111127096963e-05, + "loss": 0.4613, + "step": 30759 + }, + { + "epoch": 0.7210628410016585, + "grad_norm": 1.3603609800338745, + "learning_rate": 3.6015450178786725e-05, + "loss": 0.1553, + "step": 30760 + }, + { + "epoch": 0.7210862825764635, + "grad_norm": 0.9702263474464417, + "learning_rate": 3.600978943378578e-05, + "loss": 0.139, + "step": 30761 + }, + { + "epoch": 0.7211097241512685, + "grad_norm": 4.195504188537598, + "learning_rate": 3.600412903599748e-05, + "loss": 0.3375, + "step": 30762 + }, + { + "epoch": 0.7211331657260734, + "grad_norm": 2.5085511207580566, + "learning_rate": 3.599846898545258e-05, + "loss": 0.1533, + "step": 30763 + }, + { + "epoch": 0.7211566073008785, + "grad_norm": 2.793914794921875, + "learning_rate": 3.5992809282181827e-05, + "loss": 0.1881, + "step": 30764 + }, + { + "epoch": 0.7211800488756834, + "grad_norm": 3.815800428390503, + "learning_rate": 3.598714992621588e-05, + "loss": 0.3394, + "step": 30765 + }, + { + "epoch": 0.7212034904504885, + "grad_norm": 1.5904771089553833, + "learning_rate": 3.598149091758547e-05, + "loss": 0.215, + "step": 30766 + }, + { + "epoch": 0.7212269320252934, + "grad_norm": 2.5632519721984863, + "learning_rate": 3.597583225632127e-05, + "loss": 0.176, + "step": 30767 + }, + { + "epoch": 0.7212503736000985, + "grad_norm": 3.387014627456665, + "learning_rate": 3.5970173942454024e-05, + "loss": 0.251, + "step": 30768 + }, + { + "epoch": 0.7212738151749034, + "grad_norm": 2.2057573795318604, + "learning_rate": 3.596451597601438e-05, + "loss": 0.1942, + "step": 30769 + }, + { + "epoch": 0.7212972567497085, + "grad_norm": 2.396852731704712, + "learning_rate": 3.59588583570331e-05, + "loss": 0.3021, + "step": 30770 + }, + { + "epoch": 0.7213206983245134, + "grad_norm": 2.7859208583831787, + "learning_rate": 3.595320108554086e-05, + "loss": 0.2163, + "step": 30771 + }, + { + "epoch": 0.7213441398993184, + "grad_norm": 3.7869362831115723, + "learning_rate": 3.594754416156834e-05, + "loss": 0.276, + "step": 30772 + }, + { + "epoch": 0.7213675814741234, + "grad_norm": 3.779416084289551, + "learning_rate": 3.594188758514625e-05, + "loss": 0.2565, + "step": 30773 + }, + { + "epoch": 0.7213910230489284, + "grad_norm": 3.0339226722717285, + "learning_rate": 3.5936231356305285e-05, + "loss": 0.1902, + "step": 30774 + }, + { + "epoch": 0.7214144646237334, + "grad_norm": 1.9594327211380005, + "learning_rate": 3.593057547507609e-05, + "loss": 0.1967, + "step": 30775 + }, + { + "epoch": 0.7214379061985384, + "grad_norm": 3.4616053104400635, + "learning_rate": 3.59249199414894e-05, + "loss": 0.2463, + "step": 30776 + }, + { + "epoch": 0.7214613477733434, + "grad_norm": 3.5548808574676514, + "learning_rate": 3.5919264755575935e-05, + "loss": 0.2017, + "step": 30777 + }, + { + "epoch": 0.7214847893481484, + "grad_norm": 3.0763847827911377, + "learning_rate": 3.5913609917366334e-05, + "loss": 0.3624, + "step": 30778 + }, + { + "epoch": 0.7215082309229534, + "grad_norm": 3.062244176864624, + "learning_rate": 3.590795542689128e-05, + "loss": 0.2946, + "step": 30779 + }, + { + "epoch": 0.7215316724977584, + "grad_norm": 3.434734582901001, + "learning_rate": 3.590230128418147e-05, + "loss": 0.227, + "step": 30780 + }, + { + "epoch": 0.7215551140725633, + "grad_norm": 3.065143346786499, + "learning_rate": 3.589664748926754e-05, + "loss": 0.2745, + "step": 30781 + }, + { + "epoch": 0.7215785556473684, + "grad_norm": 0.490225225687027, + "learning_rate": 3.589099404218024e-05, + "loss": 0.057, + "step": 30782 + }, + { + "epoch": 0.7216019972221733, + "grad_norm": 3.379955768585205, + "learning_rate": 3.588534094295021e-05, + "loss": 0.2621, + "step": 30783 + }, + { + "epoch": 0.7216254387969784, + "grad_norm": 3.6239559650421143, + "learning_rate": 3.587968819160812e-05, + "loss": 0.3217, + "step": 30784 + }, + { + "epoch": 0.7216488803717834, + "grad_norm": 1.0710527896881104, + "learning_rate": 3.587403578818465e-05, + "loss": 0.0762, + "step": 30785 + }, + { + "epoch": 0.7216723219465884, + "grad_norm": 4.172421455383301, + "learning_rate": 3.586838373271045e-05, + "loss": 0.3523, + "step": 30786 + }, + { + "epoch": 0.7216957635213934, + "grad_norm": 3.800855875015259, + "learning_rate": 3.586273202521618e-05, + "loss": 0.3526, + "step": 30787 + }, + { + "epoch": 0.7217192050961984, + "grad_norm": 3.58901047706604, + "learning_rate": 3.5857080665732536e-05, + "loss": 0.5092, + "step": 30788 + }, + { + "epoch": 0.7217426466710034, + "grad_norm": 2.854944944381714, + "learning_rate": 3.585142965429019e-05, + "loss": 0.2332, + "step": 30789 + }, + { + "epoch": 0.7217660882458083, + "grad_norm": 0.9553126096725464, + "learning_rate": 3.584577899091981e-05, + "loss": 0.0995, + "step": 30790 + }, + { + "epoch": 0.7217895298206134, + "grad_norm": 2.614875078201294, + "learning_rate": 3.584012867565203e-05, + "loss": 1.5597, + "step": 30791 + }, + { + "epoch": 0.7218129713954183, + "grad_norm": 2.6568076610565186, + "learning_rate": 3.583447870851752e-05, + "loss": 0.1597, + "step": 30792 + }, + { + "epoch": 0.7218364129702234, + "grad_norm": 2.838484287261963, + "learning_rate": 3.5828829089546925e-05, + "loss": 0.2037, + "step": 30793 + }, + { + "epoch": 0.7218598545450283, + "grad_norm": 1.0103464126586914, + "learning_rate": 3.582317981877087e-05, + "loss": 0.0879, + "step": 30794 + }, + { + "epoch": 0.7218832961198334, + "grad_norm": 0.6091668605804443, + "learning_rate": 3.5817530896220085e-05, + "loss": 0.03, + "step": 30795 + }, + { + "epoch": 0.7219067376946383, + "grad_norm": 4.267800331115723, + "learning_rate": 3.581188232192519e-05, + "loss": 0.3592, + "step": 30796 + }, + { + "epoch": 0.7219301792694434, + "grad_norm": 4.435940265655518, + "learning_rate": 3.580623409591682e-05, + "loss": 0.4138, + "step": 30797 + }, + { + "epoch": 0.7219536208442483, + "grad_norm": 2.48502254486084, + "learning_rate": 3.580058621822563e-05, + "loss": 1.7699, + "step": 30798 + }, + { + "epoch": 0.7219770624190534, + "grad_norm": 2.783456802368164, + "learning_rate": 3.579493868888226e-05, + "loss": 0.1827, + "step": 30799 + }, + { + "epoch": 0.7220005039938583, + "grad_norm": 2.9946110248565674, + "learning_rate": 3.5789291507917325e-05, + "loss": 0.2228, + "step": 30800 + }, + { + "epoch": 0.7220239455686633, + "grad_norm": 1.651222825050354, + "learning_rate": 3.578364467536151e-05, + "loss": 0.129, + "step": 30801 + }, + { + "epoch": 0.7220473871434683, + "grad_norm": 1.0126873254776, + "learning_rate": 3.577799819124547e-05, + "loss": 0.1173, + "step": 30802 + }, + { + "epoch": 0.7220708287182733, + "grad_norm": 0.43971678614616394, + "learning_rate": 3.577235205559981e-05, + "loss": 0.0396, + "step": 30803 + }, + { + "epoch": 0.7220942702930783, + "grad_norm": 2.0588366985321045, + "learning_rate": 3.576670626845519e-05, + "loss": 0.1542, + "step": 30804 + }, + { + "epoch": 0.7221177118678833, + "grad_norm": 2.123167037963867, + "learning_rate": 3.5761060829842206e-05, + "loss": 0.1476, + "step": 30805 + }, + { + "epoch": 0.7221411534426883, + "grad_norm": 2.8437979221343994, + "learning_rate": 3.575541573979153e-05, + "loss": 0.206, + "step": 30806 + }, + { + "epoch": 0.7221645950174933, + "grad_norm": 3.848422050476074, + "learning_rate": 3.5749770998333746e-05, + "loss": 0.3143, + "step": 30807 + }, + { + "epoch": 0.7221880365922982, + "grad_norm": 2.6055808067321777, + "learning_rate": 3.574412660549953e-05, + "loss": 0.2708, + "step": 30808 + }, + { + "epoch": 0.7222114781671033, + "grad_norm": 3.0208182334899902, + "learning_rate": 3.57384825613195e-05, + "loss": 0.3219, + "step": 30809 + }, + { + "epoch": 0.7222349197419082, + "grad_norm": 1.2637317180633545, + "learning_rate": 3.573283886582426e-05, + "loss": 0.0941, + "step": 30810 + }, + { + "epoch": 0.7222583613167133, + "grad_norm": 2.3429007530212402, + "learning_rate": 3.572719551904444e-05, + "loss": 1.2675, + "step": 30811 + }, + { + "epoch": 0.7222818028915182, + "grad_norm": 1.4227409362792969, + "learning_rate": 3.5721552521010646e-05, + "loss": 0.1597, + "step": 30812 + }, + { + "epoch": 0.7223052444663233, + "grad_norm": 2.189453601837158, + "learning_rate": 3.571590987175354e-05, + "loss": 0.1902, + "step": 30813 + }, + { + "epoch": 0.7223286860411282, + "grad_norm": 3.7486572265625, + "learning_rate": 3.571026757130368e-05, + "loss": 0.1653, + "step": 30814 + }, + { + "epoch": 0.7223521276159333, + "grad_norm": 2.155027389526367, + "learning_rate": 3.5704625619691756e-05, + "loss": 0.1754, + "step": 30815 + }, + { + "epoch": 0.7223755691907382, + "grad_norm": 0.4589904248714447, + "learning_rate": 3.5698984016948344e-05, + "loss": 0.0219, + "step": 30816 + }, + { + "epoch": 0.7223990107655432, + "grad_norm": 0.5331829786300659, + "learning_rate": 3.5693342763104045e-05, + "loss": 0.0305, + "step": 30817 + }, + { + "epoch": 0.7224224523403482, + "grad_norm": 1.6207480430603027, + "learning_rate": 3.568770185818947e-05, + "loss": 0.139, + "step": 30818 + }, + { + "epoch": 0.7224458939151532, + "grad_norm": 1.9764221906661987, + "learning_rate": 3.5682061302235215e-05, + "loss": 0.1322, + "step": 30819 + }, + { + "epoch": 0.7224693354899582, + "grad_norm": 1.8085793256759644, + "learning_rate": 3.567642109527193e-05, + "loss": 0.1009, + "step": 30820 + }, + { + "epoch": 0.7224927770647632, + "grad_norm": 2.5839107036590576, + "learning_rate": 3.567078123733019e-05, + "loss": 0.189, + "step": 30821 + }, + { + "epoch": 0.7225162186395682, + "grad_norm": 2.2972397804260254, + "learning_rate": 3.56651417284406e-05, + "loss": 0.1913, + "step": 30822 + }, + { + "epoch": 0.7225396602143732, + "grad_norm": 2.6269516944885254, + "learning_rate": 3.565950256863376e-05, + "loss": 0.2158, + "step": 30823 + }, + { + "epoch": 0.7225631017891782, + "grad_norm": 4.466193675994873, + "learning_rate": 3.565386375794024e-05, + "loss": 0.4996, + "step": 30824 + }, + { + "epoch": 0.7225865433639832, + "grad_norm": 1.4112383127212524, + "learning_rate": 3.564822529639069e-05, + "loss": 0.1081, + "step": 30825 + }, + { + "epoch": 0.7226099849387881, + "grad_norm": 2.810349941253662, + "learning_rate": 3.564258718401565e-05, + "loss": 0.1994, + "step": 30826 + }, + { + "epoch": 0.7226334265135932, + "grad_norm": 2.183006763458252, + "learning_rate": 3.563694942084578e-05, + "loss": 0.2046, + "step": 30827 + }, + { + "epoch": 0.7226568680883981, + "grad_norm": 2.90561842918396, + "learning_rate": 3.563131200691163e-05, + "loss": 0.319, + "step": 30828 + }, + { + "epoch": 0.7226803096632032, + "grad_norm": 2.3159756660461426, + "learning_rate": 3.562567494224379e-05, + "loss": 0.2664, + "step": 30829 + }, + { + "epoch": 0.7227037512380081, + "grad_norm": 0.8225297331809998, + "learning_rate": 3.562003822687284e-05, + "loss": 0.0773, + "step": 30830 + }, + { + "epoch": 0.7227271928128132, + "grad_norm": 0.19480934739112854, + "learning_rate": 3.5614401860829384e-05, + "loss": 0.0105, + "step": 30831 + }, + { + "epoch": 0.7227506343876181, + "grad_norm": 2.857064723968506, + "learning_rate": 3.5608765844143956e-05, + "loss": 0.2152, + "step": 30832 + }, + { + "epoch": 0.7227740759624232, + "grad_norm": 2.870584726333618, + "learning_rate": 3.560313017684721e-05, + "loss": 0.2362, + "step": 30833 + }, + { + "epoch": 0.7227975175372281, + "grad_norm": 2.7631611824035645, + "learning_rate": 3.5597494858969684e-05, + "loss": 0.193, + "step": 30834 + }, + { + "epoch": 0.7228209591120331, + "grad_norm": 0.9351234436035156, + "learning_rate": 3.559185989054197e-05, + "loss": 0.0873, + "step": 30835 + }, + { + "epoch": 0.7228444006868382, + "grad_norm": 2.0458364486694336, + "learning_rate": 3.55862252715946e-05, + "loss": 0.2079, + "step": 30836 + }, + { + "epoch": 0.7228678422616431, + "grad_norm": 1.8487361669540405, + "learning_rate": 3.558059100215822e-05, + "loss": 0.1398, + "step": 30837 + }, + { + "epoch": 0.7228912838364482, + "grad_norm": 2.0635411739349365, + "learning_rate": 3.557495708226335e-05, + "loss": 0.1351, + "step": 30838 + }, + { + "epoch": 0.7229147254112531, + "grad_norm": 4.056140422821045, + "learning_rate": 3.556932351194054e-05, + "loss": 0.2783, + "step": 30839 + }, + { + "epoch": 0.7229381669860582, + "grad_norm": 3.564976215362549, + "learning_rate": 3.556369029122044e-05, + "loss": 0.2944, + "step": 30840 + }, + { + "epoch": 0.7229616085608631, + "grad_norm": 2.9070024490356445, + "learning_rate": 3.555805742013356e-05, + "loss": 0.2339, + "step": 30841 + }, + { + "epoch": 0.7229850501356682, + "grad_norm": 3.5743112564086914, + "learning_rate": 3.5552424898710476e-05, + "loss": 0.3879, + "step": 30842 + }, + { + "epoch": 0.7230084917104731, + "grad_norm": 2.274226427078247, + "learning_rate": 3.554679272698175e-05, + "loss": 0.6722, + "step": 30843 + }, + { + "epoch": 0.7230319332852781, + "grad_norm": 2.375936269760132, + "learning_rate": 3.5541160904977944e-05, + "loss": 0.209, + "step": 30844 + }, + { + "epoch": 0.7230553748600831, + "grad_norm": 1.9649330377578735, + "learning_rate": 3.553552943272956e-05, + "loss": 0.1243, + "step": 30845 + }, + { + "epoch": 0.7230788164348881, + "grad_norm": 1.7855799198150635, + "learning_rate": 3.552989831026725e-05, + "loss": 0.1113, + "step": 30846 + }, + { + "epoch": 0.7231022580096931, + "grad_norm": 0.6538904309272766, + "learning_rate": 3.5524267537621527e-05, + "loss": 0.1463, + "step": 30847 + }, + { + "epoch": 0.7231256995844981, + "grad_norm": 2.41702938079834, + "learning_rate": 3.551863711482291e-05, + "loss": 0.1965, + "step": 30848 + }, + { + "epoch": 0.7231491411593031, + "grad_norm": 0.8702973127365112, + "learning_rate": 3.5513007041902004e-05, + "loss": 0.0456, + "step": 30849 + }, + { + "epoch": 0.7231725827341081, + "grad_norm": 4.75075626373291, + "learning_rate": 3.550737731888933e-05, + "loss": 0.374, + "step": 30850 + }, + { + "epoch": 0.723196024308913, + "grad_norm": 2.788001537322998, + "learning_rate": 3.5501747945815455e-05, + "loss": 0.1557, + "step": 30851 + }, + { + "epoch": 0.7232194658837181, + "grad_norm": 1.751125693321228, + "learning_rate": 3.549611892271087e-05, + "loss": 0.9731, + "step": 30852 + }, + { + "epoch": 0.723242907458523, + "grad_norm": 3.3040413856506348, + "learning_rate": 3.549049024960619e-05, + "loss": 0.278, + "step": 30853 + }, + { + "epoch": 0.7232663490333281, + "grad_norm": 1.0493168830871582, + "learning_rate": 3.548486192653193e-05, + "loss": 0.1109, + "step": 30854 + }, + { + "epoch": 0.723289790608133, + "grad_norm": 4.116488456726074, + "learning_rate": 3.5479233953518624e-05, + "loss": 0.2874, + "step": 30855 + }, + { + "epoch": 0.7233132321829381, + "grad_norm": 2.5181941986083984, + "learning_rate": 3.54736063305968e-05, + "loss": 0.1564, + "step": 30856 + }, + { + "epoch": 0.723336673757743, + "grad_norm": 1.9499027729034424, + "learning_rate": 3.5467979057796964e-05, + "loss": 0.8304, + "step": 30857 + }, + { + "epoch": 0.7233601153325481, + "grad_norm": 1.5415666103363037, + "learning_rate": 3.5462352135149734e-05, + "loss": 0.0792, + "step": 30858 + }, + { + "epoch": 0.723383556907353, + "grad_norm": 2.2804324626922607, + "learning_rate": 3.545672556268559e-05, + "loss": 0.0817, + "step": 30859 + }, + { + "epoch": 0.7234069984821581, + "grad_norm": 5.368160724639893, + "learning_rate": 3.5451099340435065e-05, + "loss": 0.3881, + "step": 30860 + }, + { + "epoch": 0.723430440056963, + "grad_norm": 2.5744993686676025, + "learning_rate": 3.544547346842866e-05, + "loss": 0.2666, + "step": 30861 + }, + { + "epoch": 0.723453881631768, + "grad_norm": 3.488548517227173, + "learning_rate": 3.5439847946696966e-05, + "loss": 0.3275, + "step": 30862 + }, + { + "epoch": 0.723477323206573, + "grad_norm": 3.970309019088745, + "learning_rate": 3.543422277527047e-05, + "loss": 0.4221, + "step": 30863 + }, + { + "epoch": 0.723500764781378, + "grad_norm": 1.8554129600524902, + "learning_rate": 3.542859795417967e-05, + "loss": 0.9876, + "step": 30864 + }, + { + "epoch": 0.723524206356183, + "grad_norm": 1.8407429456710815, + "learning_rate": 3.5422973483455136e-05, + "loss": 1.2977, + "step": 30865 + }, + { + "epoch": 0.723547647930988, + "grad_norm": 2.0083279609680176, + "learning_rate": 3.541734936312737e-05, + "loss": 0.0975, + "step": 30866 + }, + { + "epoch": 0.723571089505793, + "grad_norm": 2.470717191696167, + "learning_rate": 3.5411725593226884e-05, + "loss": 0.1983, + "step": 30867 + }, + { + "epoch": 0.723594531080598, + "grad_norm": 1.3083301782608032, + "learning_rate": 3.540610217378418e-05, + "loss": 0.9256, + "step": 30868 + }, + { + "epoch": 0.723617972655403, + "grad_norm": 0.5690341591835022, + "learning_rate": 3.54004791048298e-05, + "loss": 0.0633, + "step": 30869 + }, + { + "epoch": 0.723641414230208, + "grad_norm": 2.309436559677124, + "learning_rate": 3.539485638639419e-05, + "loss": 0.1616, + "step": 30870 + }, + { + "epoch": 0.7236648558050129, + "grad_norm": 4.233821868896484, + "learning_rate": 3.5389234018507944e-05, + "loss": 0.3034, + "step": 30871 + }, + { + "epoch": 0.723688297379818, + "grad_norm": 2.8458681106567383, + "learning_rate": 3.538361200120153e-05, + "loss": 0.1825, + "step": 30872 + }, + { + "epoch": 0.7237117389546229, + "grad_norm": 2.269188642501831, + "learning_rate": 3.537799033450542e-05, + "loss": 1.401, + "step": 30873 + }, + { + "epoch": 0.723735180529428, + "grad_norm": 0.14109306037425995, + "learning_rate": 3.537236901845018e-05, + "loss": 0.0044, + "step": 30874 + }, + { + "epoch": 0.7237586221042329, + "grad_norm": 0.6590839624404907, + "learning_rate": 3.536674805306629e-05, + "loss": 0.0937, + "step": 30875 + }, + { + "epoch": 0.723782063679038, + "grad_norm": 1.1099859476089478, + "learning_rate": 3.536112743838424e-05, + "loss": 0.1532, + "step": 30876 + }, + { + "epoch": 0.7238055052538429, + "grad_norm": 2.4380977153778076, + "learning_rate": 3.53555071744345e-05, + "loss": 0.17, + "step": 30877 + }, + { + "epoch": 0.723828946828648, + "grad_norm": 3.751004219055176, + "learning_rate": 3.534988726124763e-05, + "loss": 0.4101, + "step": 30878 + }, + { + "epoch": 0.7238523884034529, + "grad_norm": 3.1351287364959717, + "learning_rate": 3.5344267698854084e-05, + "loss": 0.2772, + "step": 30879 + }, + { + "epoch": 0.723875829978258, + "grad_norm": 2.407412052154541, + "learning_rate": 3.533864848728435e-05, + "loss": 0.961, + "step": 30880 + }, + { + "epoch": 0.7238992715530629, + "grad_norm": 2.7278709411621094, + "learning_rate": 3.533302962656895e-05, + "loss": 0.1989, + "step": 30881 + }, + { + "epoch": 0.7239227131278679, + "grad_norm": 2.5294339656829834, + "learning_rate": 3.532741111673833e-05, + "loss": 0.2659, + "step": 30882 + }, + { + "epoch": 0.7239461547026729, + "grad_norm": 1.825420618057251, + "learning_rate": 3.5321792957822976e-05, + "loss": 0.3111, + "step": 30883 + }, + { + "epoch": 0.7239695962774779, + "grad_norm": 5.318667411804199, + "learning_rate": 3.5316175149853414e-05, + "loss": 0.5157, + "step": 30884 + }, + { + "epoch": 0.7239930378522829, + "grad_norm": 0.913087785243988, + "learning_rate": 3.531055769286008e-05, + "loss": 0.0469, + "step": 30885 + }, + { + "epoch": 0.7240164794270879, + "grad_norm": 3.2373905181884766, + "learning_rate": 3.5304940586873504e-05, + "loss": 0.424, + "step": 30886 + }, + { + "epoch": 0.724039921001893, + "grad_norm": 3.1217305660247803, + "learning_rate": 3.529932383192415e-05, + "loss": 0.3117, + "step": 30887 + }, + { + "epoch": 0.7240633625766979, + "grad_norm": 0.575123131275177, + "learning_rate": 3.529370742804248e-05, + "loss": 0.0168, + "step": 30888 + }, + { + "epoch": 0.724086804151503, + "grad_norm": 2.553616762161255, + "learning_rate": 3.528809137525897e-05, + "loss": 0.3838, + "step": 30889 + }, + { + "epoch": 0.7241102457263079, + "grad_norm": 3.1997592449188232, + "learning_rate": 3.528247567360407e-05, + "loss": 0.4569, + "step": 30890 + }, + { + "epoch": 0.7241336873011129, + "grad_norm": 0.5002959370613098, + "learning_rate": 3.52768603231083e-05, + "loss": 0.0687, + "step": 30891 + }, + { + "epoch": 0.7241571288759179, + "grad_norm": 1.5583977699279785, + "learning_rate": 3.527124532380211e-05, + "loss": 0.093, + "step": 30892 + }, + { + "epoch": 0.7241805704507229, + "grad_norm": 0.36808454990386963, + "learning_rate": 3.526563067571595e-05, + "loss": 0.0266, + "step": 30893 + }, + { + "epoch": 0.7242040120255279, + "grad_norm": 2.1125335693359375, + "learning_rate": 3.526001637888031e-05, + "loss": 0.1339, + "step": 30894 + }, + { + "epoch": 0.7242274536003329, + "grad_norm": 1.9915874004364014, + "learning_rate": 3.5254402433325614e-05, + "loss": 1.4772, + "step": 30895 + }, + { + "epoch": 0.7242508951751379, + "grad_norm": 3.1102843284606934, + "learning_rate": 3.524878883908238e-05, + "loss": 0.2596, + "step": 30896 + }, + { + "epoch": 0.7242743367499429, + "grad_norm": 0.7406324744224548, + "learning_rate": 3.5243175596181e-05, + "loss": 0.0622, + "step": 30897 + }, + { + "epoch": 0.7242977783247478, + "grad_norm": 3.338055372238159, + "learning_rate": 3.5237562704652014e-05, + "loss": 0.3985, + "step": 30898 + }, + { + "epoch": 0.7243212198995529, + "grad_norm": 0.5445425510406494, + "learning_rate": 3.523195016452583e-05, + "loss": 0.0501, + "step": 30899 + }, + { + "epoch": 0.7243446614743578, + "grad_norm": 2.4852209091186523, + "learning_rate": 3.52263379758329e-05, + "loss": 0.2109, + "step": 30900 + }, + { + "epoch": 0.7243681030491629, + "grad_norm": 3.0072646141052246, + "learning_rate": 3.522072613860368e-05, + "loss": 0.2238, + "step": 30901 + }, + { + "epoch": 0.7243915446239678, + "grad_norm": 2.612961769104004, + "learning_rate": 3.52151146528686e-05, + "loss": 0.1864, + "step": 30902 + }, + { + "epoch": 0.7244149861987729, + "grad_norm": 1.8753527402877808, + "learning_rate": 3.520950351865815e-05, + "loss": 0.1083, + "step": 30903 + }, + { + "epoch": 0.7244384277735778, + "grad_norm": 3.9121406078338623, + "learning_rate": 3.520389273600275e-05, + "loss": 0.2334, + "step": 30904 + }, + { + "epoch": 0.7244618693483829, + "grad_norm": 0.6037682294845581, + "learning_rate": 3.519828230493287e-05, + "loss": 0.0871, + "step": 30905 + }, + { + "epoch": 0.7244853109231878, + "grad_norm": 3.0256683826446533, + "learning_rate": 3.519267222547892e-05, + "loss": 0.2194, + "step": 30906 + }, + { + "epoch": 0.7245087524979928, + "grad_norm": 1.7610015869140625, + "learning_rate": 3.518706249767135e-05, + "loss": 0.1292, + "step": 30907 + }, + { + "epoch": 0.7245321940727978, + "grad_norm": 1.1336020231246948, + "learning_rate": 3.5181453121540575e-05, + "loss": 0.716, + "step": 30908 + }, + { + "epoch": 0.7245556356476028, + "grad_norm": 3.525424003601074, + "learning_rate": 3.517584409711706e-05, + "loss": 0.3208, + "step": 30909 + }, + { + "epoch": 0.7245790772224078, + "grad_norm": 2.5271668434143066, + "learning_rate": 3.517023542443126e-05, + "loss": 0.0895, + "step": 30910 + }, + { + "epoch": 0.7246025187972128, + "grad_norm": 1.2379542589187622, + "learning_rate": 3.5164627103513605e-05, + "loss": 0.0711, + "step": 30911 + }, + { + "epoch": 0.7246259603720178, + "grad_norm": 1.9848417043685913, + "learning_rate": 3.515901913439449e-05, + "loss": 0.1467, + "step": 30912 + }, + { + "epoch": 0.7246494019468228, + "grad_norm": 2.258896827697754, + "learning_rate": 3.515341151710436e-05, + "loss": 0.3502, + "step": 30913 + }, + { + "epoch": 0.7246728435216278, + "grad_norm": 2.6579713821411133, + "learning_rate": 3.514780425167364e-05, + "loss": 0.2666, + "step": 30914 + }, + { + "epoch": 0.7246962850964328, + "grad_norm": 2.253831624984741, + "learning_rate": 3.514219733813274e-05, + "loss": 0.1957, + "step": 30915 + }, + { + "epoch": 0.7247197266712377, + "grad_norm": 3.1974496841430664, + "learning_rate": 3.5136590776512126e-05, + "loss": 0.2146, + "step": 30916 + }, + { + "epoch": 0.7247431682460428, + "grad_norm": 3.038220167160034, + "learning_rate": 3.5130984566842196e-05, + "loss": 0.3912, + "step": 30917 + }, + { + "epoch": 0.7247666098208477, + "grad_norm": 2.772273302078247, + "learning_rate": 3.512537870915336e-05, + "loss": 0.2639, + "step": 30918 + }, + { + "epoch": 0.7247900513956528, + "grad_norm": 3.035087823867798, + "learning_rate": 3.5119773203476045e-05, + "loss": 0.3137, + "step": 30919 + }, + { + "epoch": 0.7248134929704577, + "grad_norm": 10.942256927490234, + "learning_rate": 3.511416804984067e-05, + "loss": 0.4264, + "step": 30920 + }, + { + "epoch": 0.7248369345452628, + "grad_norm": 0.9206984043121338, + "learning_rate": 3.510856324827761e-05, + "loss": 0.0638, + "step": 30921 + }, + { + "epoch": 0.7248603761200677, + "grad_norm": 3.5597636699676514, + "learning_rate": 3.510295879881731e-05, + "loss": 0.3996, + "step": 30922 + }, + { + "epoch": 0.7248838176948728, + "grad_norm": 3.917941093444824, + "learning_rate": 3.509735470149022e-05, + "loss": 0.3085, + "step": 30923 + }, + { + "epoch": 0.7249072592696777, + "grad_norm": 3.0939717292785645, + "learning_rate": 3.50917509563267e-05, + "loss": 0.2901, + "step": 30924 + }, + { + "epoch": 0.7249307008444827, + "grad_norm": 3.6514060497283936, + "learning_rate": 3.508614756335715e-05, + "loss": 0.2383, + "step": 30925 + }, + { + "epoch": 0.7249541424192877, + "grad_norm": 1.8097261190414429, + "learning_rate": 3.5080544522612014e-05, + "loss": 0.0977, + "step": 30926 + }, + { + "epoch": 0.7249775839940927, + "grad_norm": 2.6140947341918945, + "learning_rate": 3.5074941834121655e-05, + "loss": 0.1267, + "step": 30927 + }, + { + "epoch": 0.7250010255688977, + "grad_norm": 2.3868045806884766, + "learning_rate": 3.506933949791645e-05, + "loss": 0.1174, + "step": 30928 + }, + { + "epoch": 0.7250244671437027, + "grad_norm": 0.530529797077179, + "learning_rate": 3.5063737514026886e-05, + "loss": 0.057, + "step": 30929 + }, + { + "epoch": 0.7250479087185077, + "grad_norm": 3.3769917488098145, + "learning_rate": 3.505813588248329e-05, + "loss": 0.3702, + "step": 30930 + }, + { + "epoch": 0.7250713502933127, + "grad_norm": 4.319328308105469, + "learning_rate": 3.505253460331609e-05, + "loss": 0.2658, + "step": 30931 + }, + { + "epoch": 0.7250947918681176, + "grad_norm": 2.4549050331115723, + "learning_rate": 3.504693367655565e-05, + "loss": 0.1891, + "step": 30932 + }, + { + "epoch": 0.7251182334429227, + "grad_norm": 0.5551085472106934, + "learning_rate": 3.5041333102232346e-05, + "loss": 0.0354, + "step": 30933 + }, + { + "epoch": 0.7251416750177276, + "grad_norm": 1.527352213859558, + "learning_rate": 3.5035732880376624e-05, + "loss": 0.1218, + "step": 30934 + }, + { + "epoch": 0.7251651165925327, + "grad_norm": 1.7019679546356201, + "learning_rate": 3.503013301101882e-05, + "loss": 1.0497, + "step": 30935 + }, + { + "epoch": 0.7251885581673376, + "grad_norm": 0.8115684390068054, + "learning_rate": 3.502453349418937e-05, + "loss": 0.0424, + "step": 30936 + }, + { + "epoch": 0.7252119997421427, + "grad_norm": 2.011150598526001, + "learning_rate": 3.501893432991862e-05, + "loss": 0.07, + "step": 30937 + }, + { + "epoch": 0.7252354413169477, + "grad_norm": 2.718663454055786, + "learning_rate": 3.5013335518236964e-05, + "loss": 0.3824, + "step": 30938 + }, + { + "epoch": 0.7252588828917527, + "grad_norm": 1.7726105451583862, + "learning_rate": 3.500773705917478e-05, + "loss": 0.1084, + "step": 30939 + }, + { + "epoch": 0.7252823244665577, + "grad_norm": 3.7611193656921387, + "learning_rate": 3.500213895276241e-05, + "loss": 0.4, + "step": 30940 + }, + { + "epoch": 0.7253057660413627, + "grad_norm": 2.602403163909912, + "learning_rate": 3.499654119903029e-05, + "loss": 0.2253, + "step": 30941 + }, + { + "epoch": 0.7253292076161677, + "grad_norm": 1.2150113582611084, + "learning_rate": 3.499094379800877e-05, + "loss": 0.1125, + "step": 30942 + }, + { + "epoch": 0.7253526491909726, + "grad_norm": 2.235071897506714, + "learning_rate": 3.498534674972821e-05, + "loss": 0.2082, + "step": 30943 + }, + { + "epoch": 0.7253760907657777, + "grad_norm": 2.6564531326293945, + "learning_rate": 3.4979750054218996e-05, + "loss": 0.1173, + "step": 30944 + }, + { + "epoch": 0.7253995323405826, + "grad_norm": 2.6815831661224365, + "learning_rate": 3.497415371151148e-05, + "loss": 0.2114, + "step": 30945 + }, + { + "epoch": 0.7254229739153877, + "grad_norm": 2.1808879375457764, + "learning_rate": 3.4968557721636e-05, + "loss": 0.2062, + "step": 30946 + }, + { + "epoch": 0.7254464154901926, + "grad_norm": 0.28289929032325745, + "learning_rate": 3.4962962084622955e-05, + "loss": 0.0358, + "step": 30947 + }, + { + "epoch": 0.7254698570649977, + "grad_norm": 1.7005239725112915, + "learning_rate": 3.495736680050274e-05, + "loss": 0.1121, + "step": 30948 + }, + { + "epoch": 0.7254932986398026, + "grad_norm": 0.6074746251106262, + "learning_rate": 3.4951771869305686e-05, + "loss": 0.0297, + "step": 30949 + }, + { + "epoch": 0.7255167402146077, + "grad_norm": 1.6606086492538452, + "learning_rate": 3.494617729106214e-05, + "loss": 0.1282, + "step": 30950 + }, + { + "epoch": 0.7255401817894126, + "grad_norm": 2.704221725463867, + "learning_rate": 3.494058306580247e-05, + "loss": 0.2048, + "step": 30951 + }, + { + "epoch": 0.7255636233642176, + "grad_norm": 0.9134582877159119, + "learning_rate": 3.4934989193557023e-05, + "loss": 0.0919, + "step": 30952 + }, + { + "epoch": 0.7255870649390226, + "grad_norm": 1.8253339529037476, + "learning_rate": 3.4929395674356124e-05, + "loss": 0.0633, + "step": 30953 + }, + { + "epoch": 0.7256105065138276, + "grad_norm": 2.560389518737793, + "learning_rate": 3.492380250823019e-05, + "loss": 0.223, + "step": 30954 + }, + { + "epoch": 0.7256339480886326, + "grad_norm": 0.3353309631347656, + "learning_rate": 3.491820969520953e-05, + "loss": 0.0164, + "step": 30955 + }, + { + "epoch": 0.7256573896634376, + "grad_norm": 4.996998310089111, + "learning_rate": 3.4912617235324484e-05, + "loss": 0.5465, + "step": 30956 + }, + { + "epoch": 0.7256808312382426, + "grad_norm": 2.6283118724823, + "learning_rate": 3.490702512860542e-05, + "loss": 1.2387, + "step": 30957 + }, + { + "epoch": 0.7257042728130476, + "grad_norm": 3.119417428970337, + "learning_rate": 3.490143337508263e-05, + "loss": 1.3073, + "step": 30958 + }, + { + "epoch": 0.7257277143878526, + "grad_norm": 4.159555435180664, + "learning_rate": 3.4895841974786527e-05, + "loss": 0.6357, + "step": 30959 + }, + { + "epoch": 0.7257511559626576, + "grad_norm": 3.455195903778076, + "learning_rate": 3.489025092774738e-05, + "loss": 0.4343, + "step": 30960 + }, + { + "epoch": 0.7257745975374625, + "grad_norm": 0.9861881136894226, + "learning_rate": 3.4884660233995595e-05, + "loss": 0.0405, + "step": 30961 + }, + { + "epoch": 0.7257980391122676, + "grad_norm": 5.90411901473999, + "learning_rate": 3.4879069893561475e-05, + "loss": 0.4105, + "step": 30962 + }, + { + "epoch": 0.7258214806870725, + "grad_norm": 3.6149351596832275, + "learning_rate": 3.487347990647536e-05, + "loss": 0.3342, + "step": 30963 + }, + { + "epoch": 0.7258449222618776, + "grad_norm": 1.1993963718414307, + "learning_rate": 3.4867890272767565e-05, + "loss": 0.0834, + "step": 30964 + }, + { + "epoch": 0.7258683638366825, + "grad_norm": 1.8423444032669067, + "learning_rate": 3.486230099246843e-05, + "loss": 0.0907, + "step": 30965 + }, + { + "epoch": 0.7258918054114876, + "grad_norm": 1.215970754623413, + "learning_rate": 3.485671206560825e-05, + "loss": 0.1296, + "step": 30966 + }, + { + "epoch": 0.7259152469862925, + "grad_norm": 3.2541863918304443, + "learning_rate": 3.485112349221742e-05, + "loss": 1.777, + "step": 30967 + }, + { + "epoch": 0.7259386885610976, + "grad_norm": 3.6643130779266357, + "learning_rate": 3.484553527232622e-05, + "loss": 0.2438, + "step": 30968 + }, + { + "epoch": 0.7259621301359025, + "grad_norm": 3.6704185009002686, + "learning_rate": 3.483994740596498e-05, + "loss": 0.3411, + "step": 30969 + }, + { + "epoch": 0.7259855717107075, + "grad_norm": 0.9920206665992737, + "learning_rate": 3.483435989316398e-05, + "loss": 0.0748, + "step": 30970 + }, + { + "epoch": 0.7260090132855125, + "grad_norm": 0.4300832450389862, + "learning_rate": 3.482877273395362e-05, + "loss": 0.0239, + "step": 30971 + }, + { + "epoch": 0.7260324548603175, + "grad_norm": 3.2946548461914062, + "learning_rate": 3.482318592836416e-05, + "loss": 0.2703, + "step": 30972 + }, + { + "epoch": 0.7260558964351225, + "grad_norm": 2.519592761993408, + "learning_rate": 3.48175994764259e-05, + "loss": 0.1446, + "step": 30973 + }, + { + "epoch": 0.7260793380099275, + "grad_norm": 2.5656802654266357, + "learning_rate": 3.48120133781692e-05, + "loss": 0.2275, + "step": 30974 + }, + { + "epoch": 0.7261027795847325, + "grad_norm": 1.792170524597168, + "learning_rate": 3.480642763362436e-05, + "loss": 0.7858, + "step": 30975 + }, + { + "epoch": 0.7261262211595375, + "grad_norm": 1.3578487634658813, + "learning_rate": 3.4800842242821675e-05, + "loss": 0.0925, + "step": 30976 + }, + { + "epoch": 0.7261496627343424, + "grad_norm": 3.9779584407806396, + "learning_rate": 3.479525720579146e-05, + "loss": 0.3174, + "step": 30977 + }, + { + "epoch": 0.7261731043091475, + "grad_norm": 2.72745418548584, + "learning_rate": 3.478967252256395e-05, + "loss": 0.1873, + "step": 30978 + }, + { + "epoch": 0.7261965458839524, + "grad_norm": 2.7754628658294678, + "learning_rate": 3.4784088193169575e-05, + "loss": 0.3446, + "step": 30979 + }, + { + "epoch": 0.7262199874587575, + "grad_norm": 2.0818753242492676, + "learning_rate": 3.477850421763857e-05, + "loss": 0.8467, + "step": 30980 + }, + { + "epoch": 0.7262434290335624, + "grad_norm": 2.388070583343506, + "learning_rate": 3.477292059600121e-05, + "loss": 0.1796, + "step": 30981 + }, + { + "epoch": 0.7262668706083675, + "grad_norm": 1.651479959487915, + "learning_rate": 3.476733732828781e-05, + "loss": 0.105, + "step": 30982 + }, + { + "epoch": 0.7262903121831724, + "grad_norm": 4.6660919189453125, + "learning_rate": 3.476175441452869e-05, + "loss": 0.3989, + "step": 30983 + }, + { + "epoch": 0.7263137537579775, + "grad_norm": 2.506436586380005, + "learning_rate": 3.4756171854754125e-05, + "loss": 1.6013, + "step": 30984 + }, + { + "epoch": 0.7263371953327824, + "grad_norm": 2.4197864532470703, + "learning_rate": 3.475058964899438e-05, + "loss": 0.1713, + "step": 30985 + }, + { + "epoch": 0.7263606369075875, + "grad_norm": 1.5313446521759033, + "learning_rate": 3.4745007797279797e-05, + "loss": 0.1055, + "step": 30986 + }, + { + "epoch": 0.7263840784823924, + "grad_norm": 1.6057929992675781, + "learning_rate": 3.473942629964064e-05, + "loss": 0.0417, + "step": 30987 + }, + { + "epoch": 0.7264075200571974, + "grad_norm": 0.7096280455589294, + "learning_rate": 3.473384515610719e-05, + "loss": 0.0417, + "step": 30988 + }, + { + "epoch": 0.7264309616320025, + "grad_norm": 3.0195159912109375, + "learning_rate": 3.472826436670973e-05, + "loss": 0.2031, + "step": 30989 + }, + { + "epoch": 0.7264544032068074, + "grad_norm": 1.2063419818878174, + "learning_rate": 3.472268393147855e-05, + "loss": 0.1083, + "step": 30990 + }, + { + "epoch": 0.7264778447816125, + "grad_norm": 2.0399885177612305, + "learning_rate": 3.4717103850443876e-05, + "loss": 0.1096, + "step": 30991 + }, + { + "epoch": 0.7265012863564174, + "grad_norm": 1.4050812721252441, + "learning_rate": 3.4711524123636073e-05, + "loss": 0.0982, + "step": 30992 + }, + { + "epoch": 0.7265247279312225, + "grad_norm": 1.1770344972610474, + "learning_rate": 3.470594475108538e-05, + "loss": 0.0877, + "step": 30993 + }, + { + "epoch": 0.7265481695060274, + "grad_norm": 2.119051218032837, + "learning_rate": 3.470036573282203e-05, + "loss": 0.1992, + "step": 30994 + }, + { + "epoch": 0.7265716110808325, + "grad_norm": 4.992456912994385, + "learning_rate": 3.4694787068876365e-05, + "loss": 0.2563, + "step": 30995 + }, + { + "epoch": 0.7265950526556374, + "grad_norm": 1.3474682569503784, + "learning_rate": 3.468920875927862e-05, + "loss": 0.0471, + "step": 30996 + }, + { + "epoch": 0.7266184942304424, + "grad_norm": 2.713958740234375, + "learning_rate": 3.4683630804059066e-05, + "loss": 0.2249, + "step": 30997 + }, + { + "epoch": 0.7266419358052474, + "grad_norm": 2.6451780796051025, + "learning_rate": 3.467805320324794e-05, + "loss": 0.1893, + "step": 30998 + }, + { + "epoch": 0.7266653773800524, + "grad_norm": 2.650902509689331, + "learning_rate": 3.467247595687556e-05, + "loss": 1.4835, + "step": 30999 + }, + { + "epoch": 0.7266888189548574, + "grad_norm": 2.8302910327911377, + "learning_rate": 3.466689906497216e-05, + "loss": 0.3049, + "step": 31000 + }, + { + "epoch": 0.7267122605296624, + "grad_norm": 1.6664297580718994, + "learning_rate": 3.4661322527568006e-05, + "loss": 0.0768, + "step": 31001 + }, + { + "epoch": 0.7267357021044674, + "grad_norm": 4.200141906738281, + "learning_rate": 3.465574634469335e-05, + "loss": 0.2853, + "step": 31002 + }, + { + "epoch": 0.7267591436792724, + "grad_norm": 2.945972204208374, + "learning_rate": 3.465017051637846e-05, + "loss": 0.2523, + "step": 31003 + }, + { + "epoch": 0.7267825852540774, + "grad_norm": 1.3877129554748535, + "learning_rate": 3.464459504265355e-05, + "loss": 0.1648, + "step": 31004 + }, + { + "epoch": 0.7268060268288824, + "grad_norm": 2.3807754516601562, + "learning_rate": 3.4639019923548934e-05, + "loss": 1.6265, + "step": 31005 + }, + { + "epoch": 0.7268294684036873, + "grad_norm": 1.3988378047943115, + "learning_rate": 3.463344515909479e-05, + "loss": 0.111, + "step": 31006 + }, + { + "epoch": 0.7268529099784924, + "grad_norm": 0.9568138718605042, + "learning_rate": 3.4627870749321455e-05, + "loss": 0.0607, + "step": 31007 + }, + { + "epoch": 0.7268763515532973, + "grad_norm": 1.9598274230957031, + "learning_rate": 3.462229669425913e-05, + "loss": 0.1147, + "step": 31008 + }, + { + "epoch": 0.7268997931281024, + "grad_norm": 0.6367562413215637, + "learning_rate": 3.4616722993938056e-05, + "loss": 0.0342, + "step": 31009 + }, + { + "epoch": 0.7269232347029073, + "grad_norm": 0.7152503132820129, + "learning_rate": 3.461114964838849e-05, + "loss": 0.07, + "step": 31010 + }, + { + "epoch": 0.7269466762777124, + "grad_norm": 0.8342567086219788, + "learning_rate": 3.460557665764062e-05, + "loss": 0.0498, + "step": 31011 + }, + { + "epoch": 0.7269701178525173, + "grad_norm": 2.0080723762512207, + "learning_rate": 3.4600004021724774e-05, + "loss": 0.1538, + "step": 31012 + }, + { + "epoch": 0.7269935594273224, + "grad_norm": 2.2542881965637207, + "learning_rate": 3.459443174067114e-05, + "loss": 0.2511, + "step": 31013 + }, + { + "epoch": 0.7270170010021273, + "grad_norm": 3.5436315536499023, + "learning_rate": 3.458885981450996e-05, + "loss": 0.3876, + "step": 31014 + }, + { + "epoch": 0.7270404425769323, + "grad_norm": 0.686659038066864, + "learning_rate": 3.458328824327147e-05, + "loss": 0.0824, + "step": 31015 + }, + { + "epoch": 0.7270638841517373, + "grad_norm": 0.9653209447860718, + "learning_rate": 3.457771702698586e-05, + "loss": 0.0476, + "step": 31016 + }, + { + "epoch": 0.7270873257265423, + "grad_norm": 1.681147575378418, + "learning_rate": 3.4572146165683425e-05, + "loss": 0.1334, + "step": 31017 + }, + { + "epoch": 0.7271107673013473, + "grad_norm": 3.57470965385437, + "learning_rate": 3.4566575659394374e-05, + "loss": 0.2043, + "step": 31018 + }, + { + "epoch": 0.7271342088761523, + "grad_norm": 0.7373645305633545, + "learning_rate": 3.456100550814888e-05, + "loss": 0.0492, + "step": 31019 + }, + { + "epoch": 0.7271576504509573, + "grad_norm": 1.6306511163711548, + "learning_rate": 3.4555435711977244e-05, + "loss": 0.1193, + "step": 31020 + }, + { + "epoch": 0.7271810920257623, + "grad_norm": 2.2184364795684814, + "learning_rate": 3.454986627090967e-05, + "loss": 0.1006, + "step": 31021 + }, + { + "epoch": 0.7272045336005672, + "grad_norm": 3.0963077545166016, + "learning_rate": 3.454429718497635e-05, + "loss": 0.2205, + "step": 31022 + }, + { + "epoch": 0.7272279751753723, + "grad_norm": 2.1992313861846924, + "learning_rate": 3.453872845420748e-05, + "loss": 0.087, + "step": 31023 + }, + { + "epoch": 0.7272514167501772, + "grad_norm": 4.348994731903076, + "learning_rate": 3.453316007863334e-05, + "loss": 0.2429, + "step": 31024 + }, + { + "epoch": 0.7272748583249823, + "grad_norm": 3.2226712703704834, + "learning_rate": 3.4527592058284116e-05, + "loss": 0.2854, + "step": 31025 + }, + { + "epoch": 0.7272982998997872, + "grad_norm": 4.024064064025879, + "learning_rate": 3.452202439319002e-05, + "loss": 0.4167, + "step": 31026 + }, + { + "epoch": 0.7273217414745923, + "grad_norm": 2.5691521167755127, + "learning_rate": 3.451645708338127e-05, + "loss": 0.1196, + "step": 31027 + }, + { + "epoch": 0.7273451830493972, + "grad_norm": 3.750936985015869, + "learning_rate": 3.451089012888805e-05, + "loss": 0.22, + "step": 31028 + }, + { + "epoch": 0.7273686246242023, + "grad_norm": 1.2130948305130005, + "learning_rate": 3.4505323529740564e-05, + "loss": 0.0882, + "step": 31029 + }, + { + "epoch": 0.7273920661990072, + "grad_norm": 0.8541309237480164, + "learning_rate": 3.449975728596906e-05, + "loss": 0.0782, + "step": 31030 + }, + { + "epoch": 0.7274155077738123, + "grad_norm": 2.3296279907226562, + "learning_rate": 3.449419139760367e-05, + "loss": 0.1866, + "step": 31031 + }, + { + "epoch": 0.7274389493486172, + "grad_norm": 3.9422054290771484, + "learning_rate": 3.448862586467469e-05, + "loss": 0.2609, + "step": 31032 + }, + { + "epoch": 0.7274623909234222, + "grad_norm": 2.4683034420013428, + "learning_rate": 3.4483060687212256e-05, + "loss": 0.14, + "step": 31033 + }, + { + "epoch": 0.7274858324982272, + "grad_norm": 3.240044355392456, + "learning_rate": 3.447749586524659e-05, + "loss": 0.2566, + "step": 31034 + }, + { + "epoch": 0.7275092740730322, + "grad_norm": 2.3293240070343018, + "learning_rate": 3.4471931398807854e-05, + "loss": 0.2328, + "step": 31035 + }, + { + "epoch": 0.7275327156478372, + "grad_norm": 4.237416744232178, + "learning_rate": 3.446636728792625e-05, + "loss": 0.5623, + "step": 31036 + }, + { + "epoch": 0.7275561572226422, + "grad_norm": 1.0388131141662598, + "learning_rate": 3.446080353263199e-05, + "loss": 0.066, + "step": 31037 + }, + { + "epoch": 0.7275795987974472, + "grad_norm": 2.665280342102051, + "learning_rate": 3.445524013295527e-05, + "loss": 0.2154, + "step": 31038 + }, + { + "epoch": 0.7276030403722522, + "grad_norm": 0.5087025165557861, + "learning_rate": 3.4449677088926256e-05, + "loss": 0.0656, + "step": 31039 + }, + { + "epoch": 0.7276264819470573, + "grad_norm": 1.3117064237594604, + "learning_rate": 3.444411440057513e-05, + "loss": 0.1134, + "step": 31040 + }, + { + "epoch": 0.7276499235218622, + "grad_norm": 3.3714704513549805, + "learning_rate": 3.44385520679321e-05, + "loss": 0.23, + "step": 31041 + }, + { + "epoch": 0.7276733650966672, + "grad_norm": 3.5751283168792725, + "learning_rate": 3.443299009102728e-05, + "loss": 0.4456, + "step": 31042 + }, + { + "epoch": 0.7276968066714722, + "grad_norm": 3.524523973464966, + "learning_rate": 3.4427428469890896e-05, + "loss": 0.4126, + "step": 31043 + }, + { + "epoch": 0.7277202482462772, + "grad_norm": 3.0021393299102783, + "learning_rate": 3.4421867204553184e-05, + "loss": 0.1084, + "step": 31044 + }, + { + "epoch": 0.7277436898210822, + "grad_norm": 0.5897262096405029, + "learning_rate": 3.441630629504424e-05, + "loss": 0.0167, + "step": 31045 + }, + { + "epoch": 0.7277671313958872, + "grad_norm": 1.232127070426941, + "learning_rate": 3.441074574139428e-05, + "loss": 0.0933, + "step": 31046 + }, + { + "epoch": 0.7277905729706922, + "grad_norm": 2.3899693489074707, + "learning_rate": 3.440518554363344e-05, + "loss": 0.1872, + "step": 31047 + }, + { + "epoch": 0.7278140145454972, + "grad_norm": 5.1184868812561035, + "learning_rate": 3.439962570179193e-05, + "loss": 0.6326, + "step": 31048 + }, + { + "epoch": 0.7278374561203022, + "grad_norm": 3.1889376640319824, + "learning_rate": 3.4394066215899845e-05, + "loss": 0.2868, + "step": 31049 + }, + { + "epoch": 0.7278608976951072, + "grad_norm": 2.1308770179748535, + "learning_rate": 3.438850708598744e-05, + "loss": 0.112, + "step": 31050 + }, + { + "epoch": 0.7278843392699121, + "grad_norm": 2.9778616428375244, + "learning_rate": 3.438294831208484e-05, + "loss": 0.0779, + "step": 31051 + }, + { + "epoch": 0.7279077808447172, + "grad_norm": 2.8592092990875244, + "learning_rate": 3.4377389894222214e-05, + "loss": 0.2324, + "step": 31052 + }, + { + "epoch": 0.7279312224195221, + "grad_norm": 3.3342130184173584, + "learning_rate": 3.437183183242971e-05, + "loss": 0.1693, + "step": 31053 + }, + { + "epoch": 0.7279546639943272, + "grad_norm": 2.653224229812622, + "learning_rate": 3.436627412673745e-05, + "loss": 0.314, + "step": 31054 + }, + { + "epoch": 0.7279781055691321, + "grad_norm": 1.0699574947357178, + "learning_rate": 3.436071677717565e-05, + "loss": 0.0821, + "step": 31055 + }, + { + "epoch": 0.7280015471439372, + "grad_norm": 3.6786484718322754, + "learning_rate": 3.435515978377446e-05, + "loss": 0.1443, + "step": 31056 + }, + { + "epoch": 0.7280249887187421, + "grad_norm": 2.3845479488372803, + "learning_rate": 3.434960314656404e-05, + "loss": 0.1398, + "step": 31057 + }, + { + "epoch": 0.7280484302935472, + "grad_norm": 2.401827096939087, + "learning_rate": 3.43440468655745e-05, + "loss": 0.1932, + "step": 31058 + }, + { + "epoch": 0.7280718718683521, + "grad_norm": 6.474218845367432, + "learning_rate": 3.433849094083602e-05, + "loss": 0.1677, + "step": 31059 + }, + { + "epoch": 0.7280953134431571, + "grad_norm": 2.7052133083343506, + "learning_rate": 3.433293537237873e-05, + "loss": 0.2588, + "step": 31060 + }, + { + "epoch": 0.7281187550179621, + "grad_norm": 2.9025683403015137, + "learning_rate": 3.4327380160232746e-05, + "loss": 0.1854, + "step": 31061 + }, + { + "epoch": 0.7281421965927671, + "grad_norm": 1.258427381515503, + "learning_rate": 3.432182530442828e-05, + "loss": 0.1584, + "step": 31062 + }, + { + "epoch": 0.7281656381675721, + "grad_norm": 3.8151276111602783, + "learning_rate": 3.4316270804995434e-05, + "loss": 0.3867, + "step": 31063 + }, + { + "epoch": 0.7281890797423771, + "grad_norm": 0.2763816714286804, + "learning_rate": 3.4310716661964346e-05, + "loss": 0.0257, + "step": 31064 + }, + { + "epoch": 0.7282125213171821, + "grad_norm": 6.027409076690674, + "learning_rate": 3.430516287536516e-05, + "loss": 0.357, + "step": 31065 + }, + { + "epoch": 0.7282359628919871, + "grad_norm": 4.054040431976318, + "learning_rate": 3.429960944522801e-05, + "loss": 0.2262, + "step": 31066 + }, + { + "epoch": 0.728259404466792, + "grad_norm": 2.7466437816619873, + "learning_rate": 3.429405637158298e-05, + "loss": 0.2106, + "step": 31067 + }, + { + "epoch": 0.7282828460415971, + "grad_norm": 1.2499597072601318, + "learning_rate": 3.428850365446026e-05, + "loss": 0.1004, + "step": 31068 + }, + { + "epoch": 0.728306287616402, + "grad_norm": 1.349583625793457, + "learning_rate": 3.428295129389e-05, + "loss": 0.078, + "step": 31069 + }, + { + "epoch": 0.7283297291912071, + "grad_norm": 2.8804428577423096, + "learning_rate": 3.427739928990229e-05, + "loss": 0.1825, + "step": 31070 + }, + { + "epoch": 0.728353170766012, + "grad_norm": 6.3947014808654785, + "learning_rate": 3.4271847642527265e-05, + "loss": 0.5288, + "step": 31071 + }, + { + "epoch": 0.7283766123408171, + "grad_norm": 3.46429181098938, + "learning_rate": 3.426629635179503e-05, + "loss": 0.4511, + "step": 31072 + }, + { + "epoch": 0.728400053915622, + "grad_norm": 0.9098070859909058, + "learning_rate": 3.426074541773574e-05, + "loss": 0.0554, + "step": 31073 + }, + { + "epoch": 0.7284234954904271, + "grad_norm": 1.05787193775177, + "learning_rate": 3.425519484037944e-05, + "loss": 0.073, + "step": 31074 + }, + { + "epoch": 0.728446937065232, + "grad_norm": 2.076122522354126, + "learning_rate": 3.424964461975635e-05, + "loss": 1.083, + "step": 31075 + }, + { + "epoch": 0.728470378640037, + "grad_norm": 2.7573094367980957, + "learning_rate": 3.424409475589653e-05, + "loss": 0.2597, + "step": 31076 + }, + { + "epoch": 0.728493820214842, + "grad_norm": 2.4880945682525635, + "learning_rate": 3.42385452488301e-05, + "loss": 0.2388, + "step": 31077 + }, + { + "epoch": 0.728517261789647, + "grad_norm": 1.6688567399978638, + "learning_rate": 3.423299609858718e-05, + "loss": 0.0758, + "step": 31078 + }, + { + "epoch": 0.728540703364452, + "grad_norm": 0.4510307013988495, + "learning_rate": 3.422744730519782e-05, + "loss": 0.0275, + "step": 31079 + }, + { + "epoch": 0.728564144939257, + "grad_norm": 1.969365119934082, + "learning_rate": 3.422189886869224e-05, + "loss": 0.138, + "step": 31080 + }, + { + "epoch": 0.728587586514062, + "grad_norm": 0.6244909167289734, + "learning_rate": 3.421635078910044e-05, + "loss": 0.0284, + "step": 31081 + }, + { + "epoch": 0.728611028088867, + "grad_norm": 1.753655195236206, + "learning_rate": 3.421080306645259e-05, + "loss": 0.181, + "step": 31082 + }, + { + "epoch": 0.728634469663672, + "grad_norm": 2.118314743041992, + "learning_rate": 3.420525570077879e-05, + "loss": 0.2206, + "step": 31083 + }, + { + "epoch": 0.728657911238477, + "grad_norm": 3.4772861003875732, + "learning_rate": 3.419970869210912e-05, + "loss": 0.2166, + "step": 31084 + }, + { + "epoch": 0.728681352813282, + "grad_norm": 0.9215489029884338, + "learning_rate": 3.4194162040473675e-05, + "loss": 0.0868, + "step": 31085 + }, + { + "epoch": 0.728704794388087, + "grad_norm": 0.443981796503067, + "learning_rate": 3.418861574590253e-05, + "loss": 0.0181, + "step": 31086 + }, + { + "epoch": 0.7287282359628919, + "grad_norm": 2.3698794841766357, + "learning_rate": 3.418306980842584e-05, + "loss": 0.2009, + "step": 31087 + }, + { + "epoch": 0.728751677537697, + "grad_norm": 1.0290932655334473, + "learning_rate": 3.417752422807365e-05, + "loss": 0.0577, + "step": 31088 + }, + { + "epoch": 0.7287751191125019, + "grad_norm": 1.4906617403030396, + "learning_rate": 3.4171979004876084e-05, + "loss": 0.0866, + "step": 31089 + }, + { + "epoch": 0.728798560687307, + "grad_norm": 7.247554302215576, + "learning_rate": 3.41664341388632e-05, + "loss": 0.1403, + "step": 31090 + }, + { + "epoch": 0.728822002262112, + "grad_norm": 0.5606358051300049, + "learning_rate": 3.4160889630065095e-05, + "loss": 0.0412, + "step": 31091 + }, + { + "epoch": 0.728845443836917, + "grad_norm": 2.0055997371673584, + "learning_rate": 3.415534547851183e-05, + "loss": 0.1741, + "step": 31092 + }, + { + "epoch": 0.728868885411722, + "grad_norm": 3.286430597305298, + "learning_rate": 3.4149801684233505e-05, + "loss": 0.3148, + "step": 31093 + }, + { + "epoch": 0.728892326986527, + "grad_norm": 1.0186158418655396, + "learning_rate": 3.414425824726024e-05, + "loss": 0.1095, + "step": 31094 + }, + { + "epoch": 0.728915768561332, + "grad_norm": 0.7488203048706055, + "learning_rate": 3.413871516762208e-05, + "loss": 0.0822, + "step": 31095 + }, + { + "epoch": 0.7289392101361369, + "grad_norm": 2.222327947616577, + "learning_rate": 3.4133172445349104e-05, + "loss": 1.5951, + "step": 31096 + }, + { + "epoch": 0.728962651710942, + "grad_norm": 0.3080708384513855, + "learning_rate": 3.412763008047139e-05, + "loss": 0.0265, + "step": 31097 + }, + { + "epoch": 0.7289860932857469, + "grad_norm": 1.6373611688613892, + "learning_rate": 3.412208807301901e-05, + "loss": 0.1419, + "step": 31098 + }, + { + "epoch": 0.729009534860552, + "grad_norm": 2.2475996017456055, + "learning_rate": 3.4116546423021986e-05, + "loss": 0.0951, + "step": 31099 + }, + { + "epoch": 0.7290329764353569, + "grad_norm": 5.280879974365234, + "learning_rate": 3.411100513051049e-05, + "loss": 0.211, + "step": 31100 + }, + { + "epoch": 0.729056418010162, + "grad_norm": 2.584581136703491, + "learning_rate": 3.410546419551451e-05, + "loss": 0.2539, + "step": 31101 + }, + { + "epoch": 0.7290798595849669, + "grad_norm": 2.4871411323547363, + "learning_rate": 3.4099923618064144e-05, + "loss": 0.2132, + "step": 31102 + }, + { + "epoch": 0.729103301159772, + "grad_norm": 1.749198079109192, + "learning_rate": 3.409438339818945e-05, + "loss": 0.1294, + "step": 31103 + }, + { + "epoch": 0.7291267427345769, + "grad_norm": 4.105876922607422, + "learning_rate": 3.408884353592043e-05, + "loss": 0.4055, + "step": 31104 + }, + { + "epoch": 0.7291501843093819, + "grad_norm": 0.4368017315864563, + "learning_rate": 3.408330403128725e-05, + "loss": 0.0283, + "step": 31105 + }, + { + "epoch": 0.7291736258841869, + "grad_norm": 5.159033298492432, + "learning_rate": 3.407776488431988e-05, + "loss": 0.5205, + "step": 31106 + }, + { + "epoch": 0.7291970674589919, + "grad_norm": 0.9226130843162537, + "learning_rate": 3.4072226095048454e-05, + "loss": 0.0759, + "step": 31107 + }, + { + "epoch": 0.7292205090337969, + "grad_norm": 0.918602466583252, + "learning_rate": 3.406668766350297e-05, + "loss": 0.0796, + "step": 31108 + }, + { + "epoch": 0.7292439506086019, + "grad_norm": 2.536013603210449, + "learning_rate": 3.406114958971349e-05, + "loss": 0.2295, + "step": 31109 + }, + { + "epoch": 0.7292673921834069, + "grad_norm": 0.6720831394195557, + "learning_rate": 3.4055611873710084e-05, + "loss": 0.0583, + "step": 31110 + }, + { + "epoch": 0.7292908337582119, + "grad_norm": 1.7241830825805664, + "learning_rate": 3.405007451552277e-05, + "loss": 0.1045, + "step": 31111 + }, + { + "epoch": 0.7293142753330168, + "grad_norm": 1.066328525543213, + "learning_rate": 3.404453751518156e-05, + "loss": 0.1206, + "step": 31112 + }, + { + "epoch": 0.7293377169078219, + "grad_norm": 1.059282660484314, + "learning_rate": 3.4039000872716596e-05, + "loss": 0.0852, + "step": 31113 + }, + { + "epoch": 0.7293611584826268, + "grad_norm": 2.8114967346191406, + "learning_rate": 3.403346458815786e-05, + "loss": 0.1726, + "step": 31114 + }, + { + "epoch": 0.7293846000574319, + "grad_norm": 3.839326858520508, + "learning_rate": 3.4027928661535405e-05, + "loss": 0.2628, + "step": 31115 + }, + { + "epoch": 0.7294080416322368, + "grad_norm": 3.527336597442627, + "learning_rate": 3.4022393092879225e-05, + "loss": 0.304, + "step": 31116 + }, + { + "epoch": 0.7294314832070419, + "grad_norm": 2.2010622024536133, + "learning_rate": 3.401685788221943e-05, + "loss": 0.1113, + "step": 31117 + }, + { + "epoch": 0.7294549247818468, + "grad_norm": 2.2289583683013916, + "learning_rate": 3.4011323029586014e-05, + "loss": 0.1719, + "step": 31118 + }, + { + "epoch": 0.7294783663566519, + "grad_norm": 2.9669482707977295, + "learning_rate": 3.4005788535009e-05, + "loss": 2.1462, + "step": 31119 + }, + { + "epoch": 0.7295018079314568, + "grad_norm": 1.105296015739441, + "learning_rate": 3.400025439851845e-05, + "loss": 0.0829, + "step": 31120 + }, + { + "epoch": 0.7295252495062619, + "grad_norm": 2.2988057136535645, + "learning_rate": 3.399472062014438e-05, + "loss": 0.178, + "step": 31121 + }, + { + "epoch": 0.7295486910810668, + "grad_norm": 5.0758209228515625, + "learning_rate": 3.398918719991682e-05, + "loss": 0.4528, + "step": 31122 + }, + { + "epoch": 0.7295721326558718, + "grad_norm": 2.121511459350586, + "learning_rate": 3.398365413786577e-05, + "loss": 1.2786, + "step": 31123 + }, + { + "epoch": 0.7295955742306768, + "grad_norm": 2.482978343963623, + "learning_rate": 3.397812143402125e-05, + "loss": 1.0116, + "step": 31124 + }, + { + "epoch": 0.7296190158054818, + "grad_norm": 3.2312357425689697, + "learning_rate": 3.397258908841332e-05, + "loss": 0.1486, + "step": 31125 + }, + { + "epoch": 0.7296424573802868, + "grad_norm": 0.7618697285652161, + "learning_rate": 3.396705710107199e-05, + "loss": 0.0946, + "step": 31126 + }, + { + "epoch": 0.7296658989550918, + "grad_norm": 2.8408148288726807, + "learning_rate": 3.3961525472027255e-05, + "loss": 0.1055, + "step": 31127 + }, + { + "epoch": 0.7296893405298968, + "grad_norm": 3.006286859512329, + "learning_rate": 3.3955994201309105e-05, + "loss": 0.1836, + "step": 31128 + }, + { + "epoch": 0.7297127821047018, + "grad_norm": 3.858962297439575, + "learning_rate": 3.395046328894763e-05, + "loss": 0.2612, + "step": 31129 + }, + { + "epoch": 0.7297362236795067, + "grad_norm": 2.877732992172241, + "learning_rate": 3.394493273497279e-05, + "loss": 1.6167, + "step": 31130 + }, + { + "epoch": 0.7297596652543118, + "grad_norm": 1.7271426916122437, + "learning_rate": 3.393940253941458e-05, + "loss": 0.0502, + "step": 31131 + }, + { + "epoch": 0.7297831068291167, + "grad_norm": 0.8413387537002563, + "learning_rate": 3.393387270230305e-05, + "loss": 0.0883, + "step": 31132 + }, + { + "epoch": 0.7298065484039218, + "grad_norm": 3.2796545028686523, + "learning_rate": 3.3928343223668194e-05, + "loss": 0.1781, + "step": 31133 + }, + { + "epoch": 0.7298299899787267, + "grad_norm": 4.197279930114746, + "learning_rate": 3.3922814103540005e-05, + "loss": 0.1711, + "step": 31134 + }, + { + "epoch": 0.7298534315535318, + "grad_norm": 3.7041618824005127, + "learning_rate": 3.391728534194848e-05, + "loss": 1.2429, + "step": 31135 + }, + { + "epoch": 0.7298768731283367, + "grad_norm": 1.9547747373580933, + "learning_rate": 3.3911756938923624e-05, + "loss": 0.1389, + "step": 31136 + }, + { + "epoch": 0.7299003147031418, + "grad_norm": 3.1737518310546875, + "learning_rate": 3.3906228894495396e-05, + "loss": 0.2535, + "step": 31137 + }, + { + "epoch": 0.7299237562779467, + "grad_norm": 2.06649112701416, + "learning_rate": 3.390070120869386e-05, + "loss": 0.218, + "step": 31138 + }, + { + "epoch": 0.7299471978527518, + "grad_norm": 3.818403720855713, + "learning_rate": 3.389517388154899e-05, + "loss": 0.335, + "step": 31139 + }, + { + "epoch": 0.7299706394275567, + "grad_norm": 2.35821533203125, + "learning_rate": 3.388964691309074e-05, + "loss": 0.2079, + "step": 31140 + }, + { + "epoch": 0.7299940810023617, + "grad_norm": 2.760632038116455, + "learning_rate": 3.3884120303349134e-05, + "loss": 0.2983, + "step": 31141 + }, + { + "epoch": 0.7300175225771668, + "grad_norm": 2.1372315883636475, + "learning_rate": 3.387859405235416e-05, + "loss": 0.2488, + "step": 31142 + }, + { + "epoch": 0.7300409641519717, + "grad_norm": 2.806502342224121, + "learning_rate": 3.387306816013579e-05, + "loss": 0.158, + "step": 31143 + }, + { + "epoch": 0.7300644057267768, + "grad_norm": 0.4138263165950775, + "learning_rate": 3.386754262672398e-05, + "loss": 0.038, + "step": 31144 + }, + { + "epoch": 0.7300878473015817, + "grad_norm": 4.597236633300781, + "learning_rate": 3.3862017452148784e-05, + "loss": 0.3301, + "step": 31145 + }, + { + "epoch": 0.7301112888763868, + "grad_norm": 3.9572534561157227, + "learning_rate": 3.385649263644013e-05, + "loss": 0.2272, + "step": 31146 + }, + { + "epoch": 0.7301347304511917, + "grad_norm": 1.991317868232727, + "learning_rate": 3.385096817962802e-05, + "loss": 0.2429, + "step": 31147 + }, + { + "epoch": 0.7301581720259968, + "grad_norm": 3.5751659870147705, + "learning_rate": 3.3845444081742406e-05, + "loss": 0.3048, + "step": 31148 + }, + { + "epoch": 0.7301816136008017, + "grad_norm": 2.262821912765503, + "learning_rate": 3.383992034281327e-05, + "loss": 0.0913, + "step": 31149 + }, + { + "epoch": 0.7302050551756067, + "grad_norm": 1.0078167915344238, + "learning_rate": 3.383439696287056e-05, + "loss": 0.0511, + "step": 31150 + }, + { + "epoch": 0.7302284967504117, + "grad_norm": 2.554480791091919, + "learning_rate": 3.382887394194431e-05, + "loss": 1.4086, + "step": 31151 + }, + { + "epoch": 0.7302519383252167, + "grad_norm": 2.5942351818084717, + "learning_rate": 3.382335128006441e-05, + "loss": 0.1715, + "step": 31152 + }, + { + "epoch": 0.7302753799000217, + "grad_norm": 2.1333489418029785, + "learning_rate": 3.38178289772609e-05, + "loss": 0.189, + "step": 31153 + }, + { + "epoch": 0.7302988214748267, + "grad_norm": 1.7337901592254639, + "learning_rate": 3.381230703356371e-05, + "loss": 0.1347, + "step": 31154 + }, + { + "epoch": 0.7303222630496317, + "grad_norm": 2.001526117324829, + "learning_rate": 3.380678544900281e-05, + "loss": 0.1595, + "step": 31155 + }, + { + "epoch": 0.7303457046244367, + "grad_norm": 0.8012782335281372, + "learning_rate": 3.380126422360814e-05, + "loss": 0.0325, + "step": 31156 + }, + { + "epoch": 0.7303691461992416, + "grad_norm": 0.9822121858596802, + "learning_rate": 3.379574335740965e-05, + "loss": 0.0431, + "step": 31157 + }, + { + "epoch": 0.7303925877740467, + "grad_norm": 0.9593426585197449, + "learning_rate": 3.379022285043735e-05, + "loss": 0.0702, + "step": 31158 + }, + { + "epoch": 0.7304160293488516, + "grad_norm": 1.6086395978927612, + "learning_rate": 3.378470270272116e-05, + "loss": 0.1268, + "step": 31159 + }, + { + "epoch": 0.7304394709236567, + "grad_norm": 3.179628849029541, + "learning_rate": 3.377918291429103e-05, + "loss": 0.2842, + "step": 31160 + }, + { + "epoch": 0.7304629124984616, + "grad_norm": 0.5824756026268005, + "learning_rate": 3.377366348517692e-05, + "loss": 0.0433, + "step": 31161 + }, + { + "epoch": 0.7304863540732667, + "grad_norm": 3.5918421745300293, + "learning_rate": 3.3768144415408755e-05, + "loss": 0.1747, + "step": 31162 + }, + { + "epoch": 0.7305097956480716, + "grad_norm": 1.00015127658844, + "learning_rate": 3.376262570501651e-05, + "loss": 0.0408, + "step": 31163 + }, + { + "epoch": 0.7305332372228767, + "grad_norm": 1.4105230569839478, + "learning_rate": 3.375710735403011e-05, + "loss": 0.1058, + "step": 31164 + }, + { + "epoch": 0.7305566787976816, + "grad_norm": 2.798989772796631, + "learning_rate": 3.375158936247952e-05, + "loss": 0.1382, + "step": 31165 + }, + { + "epoch": 0.7305801203724867, + "grad_norm": 4.581045150756836, + "learning_rate": 3.374607173039468e-05, + "loss": 0.3229, + "step": 31166 + }, + { + "epoch": 0.7306035619472916, + "grad_norm": 1.2560311555862427, + "learning_rate": 3.374055445780552e-05, + "loss": 0.1197, + "step": 31167 + }, + { + "epoch": 0.7306270035220966, + "grad_norm": 0.3330230414867401, + "learning_rate": 3.373503754474197e-05, + "loss": 0.0119, + "step": 31168 + }, + { + "epoch": 0.7306504450969016, + "grad_norm": 1.1566221714019775, + "learning_rate": 3.3729520991233934e-05, + "loss": 0.1362, + "step": 31169 + }, + { + "epoch": 0.7306738866717066, + "grad_norm": 3.7534728050231934, + "learning_rate": 3.372400479731142e-05, + "loss": 0.338, + "step": 31170 + }, + { + "epoch": 0.7306973282465116, + "grad_norm": 2.553741455078125, + "learning_rate": 3.3718488963004314e-05, + "loss": 0.1194, + "step": 31171 + }, + { + "epoch": 0.7307207698213166, + "grad_norm": 1.6844749450683594, + "learning_rate": 3.371297348834255e-05, + "loss": 0.173, + "step": 31172 + }, + { + "epoch": 0.7307442113961216, + "grad_norm": 1.5153682231903076, + "learning_rate": 3.370745837335606e-05, + "loss": 0.1257, + "step": 31173 + }, + { + "epoch": 0.7307676529709266, + "grad_norm": 3.7703921794891357, + "learning_rate": 3.370194361807476e-05, + "loss": 0.4097, + "step": 31174 + }, + { + "epoch": 0.7307910945457315, + "grad_norm": 0.6196513772010803, + "learning_rate": 3.369642922252855e-05, + "loss": 0.0459, + "step": 31175 + }, + { + "epoch": 0.7308145361205366, + "grad_norm": 2.9072346687316895, + "learning_rate": 3.369091518674741e-05, + "loss": 0.2122, + "step": 31176 + }, + { + "epoch": 0.7308379776953415, + "grad_norm": 3.4174394607543945, + "learning_rate": 3.3685401510761196e-05, + "loss": 0.213, + "step": 31177 + }, + { + "epoch": 0.7308614192701466, + "grad_norm": 0.8863962888717651, + "learning_rate": 3.367988819459989e-05, + "loss": 0.0615, + "step": 31178 + }, + { + "epoch": 0.7308848608449515, + "grad_norm": 1.6365286111831665, + "learning_rate": 3.3674375238293376e-05, + "loss": 0.1226, + "step": 31179 + }, + { + "epoch": 0.7309083024197566, + "grad_norm": 2.445431709289551, + "learning_rate": 3.366886264187157e-05, + "loss": 0.1474, + "step": 31180 + }, + { + "epoch": 0.7309317439945615, + "grad_norm": 2.454824209213257, + "learning_rate": 3.366335040536438e-05, + "loss": 0.3507, + "step": 31181 + }, + { + "epoch": 0.7309551855693666, + "grad_norm": 2.156641721725464, + "learning_rate": 3.365783852880168e-05, + "loss": 0.115, + "step": 31182 + }, + { + "epoch": 0.7309786271441715, + "grad_norm": 0.791804850101471, + "learning_rate": 3.3652327012213446e-05, + "loss": 0.0487, + "step": 31183 + }, + { + "epoch": 0.7310020687189765, + "grad_norm": 0.9282804727554321, + "learning_rate": 3.3646815855629544e-05, + "loss": 0.1105, + "step": 31184 + }, + { + "epoch": 0.7310255102937815, + "grad_norm": 0.6089640259742737, + "learning_rate": 3.36413050590799e-05, + "loss": 0.0334, + "step": 31185 + }, + { + "epoch": 0.7310489518685865, + "grad_norm": 2.9905009269714355, + "learning_rate": 3.3635794622594394e-05, + "loss": 0.2067, + "step": 31186 + }, + { + "epoch": 0.7310723934433915, + "grad_norm": 2.497826099395752, + "learning_rate": 3.3630284546202926e-05, + "loss": 0.129, + "step": 31187 + }, + { + "epoch": 0.7310958350181965, + "grad_norm": 2.6160032749176025, + "learning_rate": 3.3624774829935376e-05, + "loss": 1.3564, + "step": 31188 + }, + { + "epoch": 0.7311192765930015, + "grad_norm": 2.4126741886138916, + "learning_rate": 3.361926547382166e-05, + "loss": 1.6444, + "step": 31189 + }, + { + "epoch": 0.7311427181678065, + "grad_norm": 1.3064956665039062, + "learning_rate": 3.361375647789171e-05, + "loss": 0.0812, + "step": 31190 + }, + { + "epoch": 0.7311661597426115, + "grad_norm": 3.737957715988159, + "learning_rate": 3.3608247842175386e-05, + "loss": 0.2265, + "step": 31191 + }, + { + "epoch": 0.7311896013174165, + "grad_norm": 1.6092665195465088, + "learning_rate": 3.360273956670258e-05, + "loss": 0.0882, + "step": 31192 + }, + { + "epoch": 0.7312130428922216, + "grad_norm": 3.6992218494415283, + "learning_rate": 3.3597231651503166e-05, + "loss": 0.3815, + "step": 31193 + }, + { + "epoch": 0.7312364844670265, + "grad_norm": 0.8187867999076843, + "learning_rate": 3.3591724096607046e-05, + "loss": 0.0826, + "step": 31194 + }, + { + "epoch": 0.7312599260418315, + "grad_norm": 2.4767239093780518, + "learning_rate": 3.358621690204407e-05, + "loss": 1.3542, + "step": 31195 + }, + { + "epoch": 0.7312833676166365, + "grad_norm": 0.8528246879577637, + "learning_rate": 3.358071006784417e-05, + "loss": 0.0919, + "step": 31196 + }, + { + "epoch": 0.7313068091914415, + "grad_norm": 1.105150818824768, + "learning_rate": 3.357520359403721e-05, + "loss": 0.0887, + "step": 31197 + }, + { + "epoch": 0.7313302507662465, + "grad_norm": 2.0977962017059326, + "learning_rate": 3.3569697480653064e-05, + "loss": 0.173, + "step": 31198 + }, + { + "epoch": 0.7313536923410515, + "grad_norm": 1.230865478515625, + "learning_rate": 3.35641917277216e-05, + "loss": 0.111, + "step": 31199 + }, + { + "epoch": 0.7313771339158565, + "grad_norm": 3.320889949798584, + "learning_rate": 3.355868633527267e-05, + "loss": 0.1907, + "step": 31200 + }, + { + "epoch": 0.7314005754906615, + "grad_norm": 2.18145489692688, + "learning_rate": 3.355318130333621e-05, + "loss": 1.539, + "step": 31201 + }, + { + "epoch": 0.7314240170654664, + "grad_norm": 3.0852444171905518, + "learning_rate": 3.354767663194203e-05, + "loss": 0.2728, + "step": 31202 + }, + { + "epoch": 0.7314474586402715, + "grad_norm": 2.2707455158233643, + "learning_rate": 3.354217232112006e-05, + "loss": 0.2559, + "step": 31203 + }, + { + "epoch": 0.7314709002150764, + "grad_norm": 3.909355640411377, + "learning_rate": 3.353666837090013e-05, + "loss": 0.4046, + "step": 31204 + }, + { + "epoch": 0.7314943417898815, + "grad_norm": 2.1794769763946533, + "learning_rate": 3.35311647813121e-05, + "loss": 0.2111, + "step": 31205 + }, + { + "epoch": 0.7315177833646864, + "grad_norm": 2.2046561241149902, + "learning_rate": 3.352566155238583e-05, + "loss": 0.1911, + "step": 31206 + }, + { + "epoch": 0.7315412249394915, + "grad_norm": 2.7841289043426514, + "learning_rate": 3.352015868415117e-05, + "loss": 0.204, + "step": 31207 + }, + { + "epoch": 0.7315646665142964, + "grad_norm": 1.6779954433441162, + "learning_rate": 3.351465617663803e-05, + "loss": 0.1587, + "step": 31208 + }, + { + "epoch": 0.7315881080891015, + "grad_norm": 1.3867181539535522, + "learning_rate": 3.3509154029876236e-05, + "loss": 0.1201, + "step": 31209 + }, + { + "epoch": 0.7316115496639064, + "grad_norm": 3.481473684310913, + "learning_rate": 3.3503652243895634e-05, + "loss": 0.2381, + "step": 31210 + }, + { + "epoch": 0.7316349912387115, + "grad_norm": 2.421895742416382, + "learning_rate": 3.3498150818726096e-05, + "loss": 0.1245, + "step": 31211 + }, + { + "epoch": 0.7316584328135164, + "grad_norm": 0.7861517071723938, + "learning_rate": 3.3492649754397446e-05, + "loss": 0.0814, + "step": 31212 + }, + { + "epoch": 0.7316818743883214, + "grad_norm": 0.49163877964019775, + "learning_rate": 3.3487149050939525e-05, + "loss": 0.0192, + "step": 31213 + }, + { + "epoch": 0.7317053159631264, + "grad_norm": 1.9979125261306763, + "learning_rate": 3.3481648708382205e-05, + "loss": 1.3837, + "step": 31214 + }, + { + "epoch": 0.7317287575379314, + "grad_norm": 3.7725963592529297, + "learning_rate": 3.347614872675536e-05, + "loss": 0.2917, + "step": 31215 + }, + { + "epoch": 0.7317521991127364, + "grad_norm": 4.3692522048950195, + "learning_rate": 3.34706491060888e-05, + "loss": 0.3851, + "step": 31216 + }, + { + "epoch": 0.7317756406875414, + "grad_norm": 1.4005237817764282, + "learning_rate": 3.346514984641237e-05, + "loss": 0.0764, + "step": 31217 + }, + { + "epoch": 0.7317990822623464, + "grad_norm": 1.1339111328125, + "learning_rate": 3.3459650947755916e-05, + "loss": 0.0589, + "step": 31218 + }, + { + "epoch": 0.7318225238371514, + "grad_norm": 3.1321115493774414, + "learning_rate": 3.3454152410149255e-05, + "loss": 0.2691, + "step": 31219 + }, + { + "epoch": 0.7318459654119563, + "grad_norm": 1.5477713346481323, + "learning_rate": 3.3448654233622215e-05, + "loss": 0.0934, + "step": 31220 + }, + { + "epoch": 0.7318694069867614, + "grad_norm": 3.9619815349578857, + "learning_rate": 3.344315641820467e-05, + "loss": 0.3193, + "step": 31221 + }, + { + "epoch": 0.7318928485615663, + "grad_norm": 1.7621747255325317, + "learning_rate": 3.343765896392643e-05, + "loss": 0.1385, + "step": 31222 + }, + { + "epoch": 0.7319162901363714, + "grad_norm": 3.818057060241699, + "learning_rate": 3.343216187081732e-05, + "loss": 0.3896, + "step": 31223 + }, + { + "epoch": 0.7319397317111763, + "grad_norm": 2.0522985458374023, + "learning_rate": 3.342666513890719e-05, + "loss": 0.1147, + "step": 31224 + }, + { + "epoch": 0.7319631732859814, + "grad_norm": 7.707479476928711, + "learning_rate": 3.34211687682258e-05, + "loss": 0.5482, + "step": 31225 + }, + { + "epoch": 0.7319866148607863, + "grad_norm": 2.504328489303589, + "learning_rate": 3.341567275880305e-05, + "loss": 0.1428, + "step": 31226 + }, + { + "epoch": 0.7320100564355914, + "grad_norm": 0.6826940774917603, + "learning_rate": 3.341017711066871e-05, + "loss": 0.0696, + "step": 31227 + }, + { + "epoch": 0.7320334980103963, + "grad_norm": 3.8335628509521484, + "learning_rate": 3.3404681823852634e-05, + "loss": 0.3271, + "step": 31228 + }, + { + "epoch": 0.7320569395852013, + "grad_norm": 4.464621543884277, + "learning_rate": 3.339918689838465e-05, + "loss": 0.4843, + "step": 31229 + }, + { + "epoch": 0.7320803811600063, + "grad_norm": 1.8765134811401367, + "learning_rate": 3.339369233429453e-05, + "loss": 0.8772, + "step": 31230 + }, + { + "epoch": 0.7321038227348113, + "grad_norm": 0.5102606415748596, + "learning_rate": 3.338819813161212e-05, + "loss": 0.061, + "step": 31231 + }, + { + "epoch": 0.7321272643096163, + "grad_norm": 0.739957869052887, + "learning_rate": 3.338270429036723e-05, + "loss": 0.0643, + "step": 31232 + }, + { + "epoch": 0.7321507058844213, + "grad_norm": 4.358633995056152, + "learning_rate": 3.337721081058961e-05, + "loss": 0.4414, + "step": 31233 + }, + { + "epoch": 0.7321741474592263, + "grad_norm": 4.00924015045166, + "learning_rate": 3.3371717692309156e-05, + "loss": 0.3301, + "step": 31234 + }, + { + "epoch": 0.7321975890340313, + "grad_norm": 2.708921194076538, + "learning_rate": 3.336622493555565e-05, + "loss": 1.5785, + "step": 31235 + }, + { + "epoch": 0.7322210306088363, + "grad_norm": 2.369138240814209, + "learning_rate": 3.336073254035885e-05, + "loss": 0.1702, + "step": 31236 + }, + { + "epoch": 0.7322444721836413, + "grad_norm": 3.1314916610717773, + "learning_rate": 3.335524050674857e-05, + "loss": 0.2985, + "step": 31237 + }, + { + "epoch": 0.7322679137584462, + "grad_norm": 5.220622539520264, + "learning_rate": 3.334974883475468e-05, + "loss": 0.4741, + "step": 31238 + }, + { + "epoch": 0.7322913553332513, + "grad_norm": 2.7910854816436768, + "learning_rate": 3.334425752440691e-05, + "loss": 0.4227, + "step": 31239 + }, + { + "epoch": 0.7323147969080562, + "grad_norm": 1.5263584852218628, + "learning_rate": 3.333876657573504e-05, + "loss": 0.0531, + "step": 31240 + }, + { + "epoch": 0.7323382384828613, + "grad_norm": 1.922126293182373, + "learning_rate": 3.333327598876893e-05, + "loss": 0.08, + "step": 31241 + }, + { + "epoch": 0.7323616800576662, + "grad_norm": 0.9970744252204895, + "learning_rate": 3.3327785763538343e-05, + "loss": 0.0672, + "step": 31242 + }, + { + "epoch": 0.7323851216324713, + "grad_norm": 4.245748996734619, + "learning_rate": 3.3322295900073064e-05, + "loss": 0.3632, + "step": 31243 + }, + { + "epoch": 0.7324085632072763, + "grad_norm": 0.6071755290031433, + "learning_rate": 3.331680639840288e-05, + "loss": 0.084, + "step": 31244 + }, + { + "epoch": 0.7324320047820813, + "grad_norm": 2.674001693725586, + "learning_rate": 3.331131725855754e-05, + "loss": 0.1703, + "step": 31245 + }, + { + "epoch": 0.7324554463568863, + "grad_norm": 2.4949593544006348, + "learning_rate": 3.330582848056691e-05, + "loss": 1.536, + "step": 31246 + }, + { + "epoch": 0.7324788879316912, + "grad_norm": 2.5082883834838867, + "learning_rate": 3.330034006446073e-05, + "loss": 0.2793, + "step": 31247 + }, + { + "epoch": 0.7325023295064963, + "grad_norm": 1.2880150079727173, + "learning_rate": 3.329485201026877e-05, + "loss": 0.092, + "step": 31248 + }, + { + "epoch": 0.7325257710813012, + "grad_norm": 3.6038477420806885, + "learning_rate": 3.328936431802083e-05, + "loss": 0.2138, + "step": 31249 + }, + { + "epoch": 0.7325492126561063, + "grad_norm": 2.482128381729126, + "learning_rate": 3.3283876987746644e-05, + "loss": 0.2302, + "step": 31250 + }, + { + "epoch": 0.7325726542309112, + "grad_norm": 0.8096610307693481, + "learning_rate": 3.327839001947605e-05, + "loss": 0.0691, + "step": 31251 + }, + { + "epoch": 0.7325960958057163, + "grad_norm": 0.8960228562355042, + "learning_rate": 3.3272903413238745e-05, + "loss": 0.3467, + "step": 31252 + }, + { + "epoch": 0.7326195373805212, + "grad_norm": 0.7120388746261597, + "learning_rate": 3.326741716906458e-05, + "loss": 0.0535, + "step": 31253 + }, + { + "epoch": 0.7326429789553263, + "grad_norm": 2.379997968673706, + "learning_rate": 3.3261931286983286e-05, + "loss": 0.2293, + "step": 31254 + }, + { + "epoch": 0.7326664205301312, + "grad_norm": 2.527939796447754, + "learning_rate": 3.3256445767024626e-05, + "loss": 0.2423, + "step": 31255 + }, + { + "epoch": 0.7326898621049363, + "grad_norm": 3.7665624618530273, + "learning_rate": 3.325096060921837e-05, + "loss": 0.2481, + "step": 31256 + }, + { + "epoch": 0.7327133036797412, + "grad_norm": 1.077601671218872, + "learning_rate": 3.3245475813594286e-05, + "loss": 0.1386, + "step": 31257 + }, + { + "epoch": 0.7327367452545462, + "grad_norm": 2.0203917026519775, + "learning_rate": 3.323999138018208e-05, + "loss": 0.1953, + "step": 31258 + }, + { + "epoch": 0.7327601868293512, + "grad_norm": 1.2142157554626465, + "learning_rate": 3.32345073090116e-05, + "loss": 0.0719, + "step": 31259 + }, + { + "epoch": 0.7327836284041562, + "grad_norm": 1.4182409048080444, + "learning_rate": 3.322902360011256e-05, + "loss": 0.0719, + "step": 31260 + }, + { + "epoch": 0.7328070699789612, + "grad_norm": 0.9154912233352661, + "learning_rate": 3.322354025351472e-05, + "loss": 0.0537, + "step": 31261 + }, + { + "epoch": 0.7328305115537662, + "grad_norm": 2.9653091430664062, + "learning_rate": 3.321805726924778e-05, + "loss": 0.2949, + "step": 31262 + }, + { + "epoch": 0.7328539531285712, + "grad_norm": 2.3378140926361084, + "learning_rate": 3.3212574647341586e-05, + "loss": 1.8111, + "step": 31263 + }, + { + "epoch": 0.7328773947033762, + "grad_norm": 1.5441093444824219, + "learning_rate": 3.320709238782583e-05, + "loss": 0.0659, + "step": 31264 + }, + { + "epoch": 0.7329008362781811, + "grad_norm": 2.567612409591675, + "learning_rate": 3.320161049073025e-05, + "loss": 0.305, + "step": 31265 + }, + { + "epoch": 0.7329242778529862, + "grad_norm": 2.403853416442871, + "learning_rate": 3.3196128956084626e-05, + "loss": 0.1936, + "step": 31266 + }, + { + "epoch": 0.7329477194277911, + "grad_norm": 1.2640198469161987, + "learning_rate": 3.3190647783918695e-05, + "loss": 0.0653, + "step": 31267 + }, + { + "epoch": 0.7329711610025962, + "grad_norm": 1.1200467348098755, + "learning_rate": 3.318516697426219e-05, + "loss": 0.0727, + "step": 31268 + }, + { + "epoch": 0.7329946025774011, + "grad_norm": 2.6915512084960938, + "learning_rate": 3.3179686527144834e-05, + "loss": 0.2314, + "step": 31269 + }, + { + "epoch": 0.7330180441522062, + "grad_norm": 0.5210897326469421, + "learning_rate": 3.31742064425964e-05, + "loss": 0.0293, + "step": 31270 + }, + { + "epoch": 0.7330414857270111, + "grad_norm": 4.995628356933594, + "learning_rate": 3.316872672064655e-05, + "loss": 0.7611, + "step": 31271 + }, + { + "epoch": 0.7330649273018162, + "grad_norm": 0.503016471862793, + "learning_rate": 3.31632473613251e-05, + "loss": 0.0202, + "step": 31272 + }, + { + "epoch": 0.7330883688766211, + "grad_norm": 0.5027032494544983, + "learning_rate": 3.315776836466175e-05, + "loss": 0.0497, + "step": 31273 + }, + { + "epoch": 0.7331118104514261, + "grad_norm": 3.2079010009765625, + "learning_rate": 3.315228973068622e-05, + "loss": 0.266, + "step": 31274 + }, + { + "epoch": 0.7331352520262311, + "grad_norm": 0.9749830961227417, + "learning_rate": 3.3146811459428254e-05, + "loss": 0.0456, + "step": 31275 + }, + { + "epoch": 0.7331586936010361, + "grad_norm": 3.0921709537506104, + "learning_rate": 3.3141333550917565e-05, + "loss": 0.2139, + "step": 31276 + }, + { + "epoch": 0.7331821351758411, + "grad_norm": 2.462723731994629, + "learning_rate": 3.31358560051839e-05, + "loss": 0.2181, + "step": 31277 + }, + { + "epoch": 0.7332055767506461, + "grad_norm": 4.035024166107178, + "learning_rate": 3.3130378822256905e-05, + "loss": 0.3467, + "step": 31278 + }, + { + "epoch": 0.7332290183254511, + "grad_norm": 1.6213313341140747, + "learning_rate": 3.31249020021664e-05, + "loss": 0.1347, + "step": 31279 + }, + { + "epoch": 0.7332524599002561, + "grad_norm": 3.7247655391693115, + "learning_rate": 3.311942554494205e-05, + "loss": 0.2661, + "step": 31280 + }, + { + "epoch": 0.733275901475061, + "grad_norm": 2.3829169273376465, + "learning_rate": 3.311394945061359e-05, + "loss": 0.166, + "step": 31281 + }, + { + "epoch": 0.7332993430498661, + "grad_norm": 4.396070957183838, + "learning_rate": 3.310847371921071e-05, + "loss": 0.5096, + "step": 31282 + }, + { + "epoch": 0.733322784624671, + "grad_norm": 2.0121729373931885, + "learning_rate": 3.3102998350763105e-05, + "loss": 0.1613, + "step": 31283 + }, + { + "epoch": 0.7333462261994761, + "grad_norm": 1.8980849981307983, + "learning_rate": 3.309752334530054e-05, + "loss": 0.086, + "step": 31284 + }, + { + "epoch": 0.733369667774281, + "grad_norm": 1.0502656698226929, + "learning_rate": 3.3092048702852705e-05, + "loss": 0.0683, + "step": 31285 + }, + { + "epoch": 0.7333931093490861, + "grad_norm": 2.2008254528045654, + "learning_rate": 3.308657442344926e-05, + "loss": 0.1397, + "step": 31286 + }, + { + "epoch": 0.733416550923891, + "grad_norm": 1.8238368034362793, + "learning_rate": 3.3081100507119965e-05, + "loss": 0.0898, + "step": 31287 + }, + { + "epoch": 0.7334399924986961, + "grad_norm": 0.9089747071266174, + "learning_rate": 3.30756269538945e-05, + "loss": 0.5278, + "step": 31288 + }, + { + "epoch": 0.733463434073501, + "grad_norm": 1.478155493736267, + "learning_rate": 3.307015376380258e-05, + "loss": 0.0722, + "step": 31289 + }, + { + "epoch": 0.7334868756483061, + "grad_norm": 0.9240373373031616, + "learning_rate": 3.306468093687385e-05, + "loss": 0.0834, + "step": 31290 + }, + { + "epoch": 0.733510317223111, + "grad_norm": 2.4003124237060547, + "learning_rate": 3.305920847313807e-05, + "loss": 0.1768, + "step": 31291 + }, + { + "epoch": 0.733533758797916, + "grad_norm": 3.8423264026641846, + "learning_rate": 3.305373637262491e-05, + "loss": 0.3967, + "step": 31292 + }, + { + "epoch": 0.733557200372721, + "grad_norm": 2.430797815322876, + "learning_rate": 3.304826463536406e-05, + "loss": 0.2301, + "step": 31293 + }, + { + "epoch": 0.733580641947526, + "grad_norm": 2.304302930831909, + "learning_rate": 3.3042793261385195e-05, + "loss": 0.1335, + "step": 31294 + }, + { + "epoch": 0.7336040835223311, + "grad_norm": 2.6579136848449707, + "learning_rate": 3.303732225071804e-05, + "loss": 0.1773, + "step": 31295 + }, + { + "epoch": 0.733627525097136, + "grad_norm": 4.1053876876831055, + "learning_rate": 3.3031851603392205e-05, + "loss": 0.2507, + "step": 31296 + }, + { + "epoch": 0.7336509666719411, + "grad_norm": 0.4967813491821289, + "learning_rate": 3.302638131943747e-05, + "loss": 0.0313, + "step": 31297 + }, + { + "epoch": 0.733674408246746, + "grad_norm": 2.698448657989502, + "learning_rate": 3.302091139888344e-05, + "loss": 0.2796, + "step": 31298 + }, + { + "epoch": 0.7336978498215511, + "grad_norm": 1.0433588027954102, + "learning_rate": 3.301544184175985e-05, + "loss": 0.0783, + "step": 31299 + }, + { + "epoch": 0.733721291396356, + "grad_norm": 0.6618617177009583, + "learning_rate": 3.3009972648096364e-05, + "loss": 0.0217, + "step": 31300 + }, + { + "epoch": 0.733744732971161, + "grad_norm": 1.3506296873092651, + "learning_rate": 3.300450381792265e-05, + "loss": 0.0905, + "step": 31301 + }, + { + "epoch": 0.733768174545966, + "grad_norm": 0.8561906218528748, + "learning_rate": 3.299903535126838e-05, + "loss": 0.1178, + "step": 31302 + }, + { + "epoch": 0.733791616120771, + "grad_norm": 2.809116840362549, + "learning_rate": 3.29935672481632e-05, + "loss": 0.1849, + "step": 31303 + }, + { + "epoch": 0.733815057695576, + "grad_norm": 1.0452462434768677, + "learning_rate": 3.298809950863684e-05, + "loss": 0.0884, + "step": 31304 + }, + { + "epoch": 0.733838499270381, + "grad_norm": 2.562352180480957, + "learning_rate": 3.2982632132718935e-05, + "loss": 0.1621, + "step": 31305 + }, + { + "epoch": 0.733861940845186, + "grad_norm": 1.846177101135254, + "learning_rate": 3.2977165120439145e-05, + "loss": 0.1084, + "step": 31306 + }, + { + "epoch": 0.733885382419991, + "grad_norm": 2.4367871284484863, + "learning_rate": 3.297169847182715e-05, + "loss": 0.219, + "step": 31307 + }, + { + "epoch": 0.733908823994796, + "grad_norm": 2.7751383781433105, + "learning_rate": 3.29662321869126e-05, + "loss": 1.3153, + "step": 31308 + }, + { + "epoch": 0.733932265569601, + "grad_norm": 3.2819950580596924, + "learning_rate": 3.2960766265725144e-05, + "loss": 0.3196, + "step": 31309 + }, + { + "epoch": 0.7339557071444059, + "grad_norm": 2.5190680027008057, + "learning_rate": 3.295530070829443e-05, + "loss": 0.2084, + "step": 31310 + }, + { + "epoch": 0.733979148719211, + "grad_norm": 1.3410581350326538, + "learning_rate": 3.294983551465018e-05, + "loss": 0.1395, + "step": 31311 + }, + { + "epoch": 0.7340025902940159, + "grad_norm": 2.035545825958252, + "learning_rate": 3.294437068482201e-05, + "loss": 1.211, + "step": 31312 + }, + { + "epoch": 0.734026031868821, + "grad_norm": 3.5840892791748047, + "learning_rate": 3.293890621883957e-05, + "loss": 0.2244, + "step": 31313 + }, + { + "epoch": 0.7340494734436259, + "grad_norm": 3.113645553588867, + "learning_rate": 3.293344211673252e-05, + "loss": 0.2909, + "step": 31314 + }, + { + "epoch": 0.734072915018431, + "grad_norm": 3.3911514282226562, + "learning_rate": 3.292797837853048e-05, + "loss": 0.2744, + "step": 31315 + }, + { + "epoch": 0.7340963565932359, + "grad_norm": 4.9345622062683105, + "learning_rate": 3.2922515004263086e-05, + "loss": 0.7281, + "step": 31316 + }, + { + "epoch": 0.734119798168041, + "grad_norm": 0.561894953250885, + "learning_rate": 3.291705199396006e-05, + "loss": 0.0474, + "step": 31317 + }, + { + "epoch": 0.7341432397428459, + "grad_norm": 2.64599347114563, + "learning_rate": 3.291158934765099e-05, + "loss": 0.2457, + "step": 31318 + }, + { + "epoch": 0.734166681317651, + "grad_norm": 3.116663932800293, + "learning_rate": 3.290612706536551e-05, + "loss": 0.3032, + "step": 31319 + }, + { + "epoch": 0.7341901228924559, + "grad_norm": 2.5819013118743896, + "learning_rate": 3.290066514713328e-05, + "loss": 0.3844, + "step": 31320 + }, + { + "epoch": 0.7342135644672609, + "grad_norm": 0.8735735416412354, + "learning_rate": 3.289520359298389e-05, + "loss": 0.032, + "step": 31321 + }, + { + "epoch": 0.7342370060420659, + "grad_norm": 3.0593760013580322, + "learning_rate": 3.288974240294706e-05, + "loss": 0.255, + "step": 31322 + }, + { + "epoch": 0.7342604476168709, + "grad_norm": 4.144814968109131, + "learning_rate": 3.288428157705232e-05, + "loss": 0.4182, + "step": 31323 + }, + { + "epoch": 0.7342838891916759, + "grad_norm": 3.3259756565093994, + "learning_rate": 3.28788211153294e-05, + "loss": 0.2895, + "step": 31324 + }, + { + "epoch": 0.7343073307664809, + "grad_norm": 3.6568403244018555, + "learning_rate": 3.287336101780789e-05, + "loss": 0.3156, + "step": 31325 + }, + { + "epoch": 0.7343307723412859, + "grad_norm": 3.078446626663208, + "learning_rate": 3.28679012845174e-05, + "loss": 0.1709, + "step": 31326 + }, + { + "epoch": 0.7343542139160909, + "grad_norm": 1.0169602632522583, + "learning_rate": 3.286244191548757e-05, + "loss": 0.0724, + "step": 31327 + }, + { + "epoch": 0.7343776554908958, + "grad_norm": 2.2550899982452393, + "learning_rate": 3.2856982910747977e-05, + "loss": 0.1096, + "step": 31328 + }, + { + "epoch": 0.7344010970657009, + "grad_norm": 1.8225888013839722, + "learning_rate": 3.285152427032832e-05, + "loss": 0.1096, + "step": 31329 + }, + { + "epoch": 0.7344245386405058, + "grad_norm": 1.9430642127990723, + "learning_rate": 3.284606599425819e-05, + "loss": 0.1517, + "step": 31330 + }, + { + "epoch": 0.7344479802153109, + "grad_norm": 2.2834980487823486, + "learning_rate": 3.284060808256719e-05, + "loss": 0.1904, + "step": 31331 + }, + { + "epoch": 0.7344714217901158, + "grad_norm": 3.529849052429199, + "learning_rate": 3.283515053528493e-05, + "loss": 0.3664, + "step": 31332 + }, + { + "epoch": 0.7344948633649209, + "grad_norm": 1.100932002067566, + "learning_rate": 3.282969335244104e-05, + "loss": 0.0622, + "step": 31333 + }, + { + "epoch": 0.7345183049397258, + "grad_norm": 2.8214786052703857, + "learning_rate": 3.282423653406509e-05, + "loss": 0.2626, + "step": 31334 + }, + { + "epoch": 0.7345417465145309, + "grad_norm": 3.1793901920318604, + "learning_rate": 3.2818780080186715e-05, + "loss": 0.2697, + "step": 31335 + }, + { + "epoch": 0.7345651880893358, + "grad_norm": 2.476951837539673, + "learning_rate": 3.281332399083556e-05, + "loss": 1.4597, + "step": 31336 + }, + { + "epoch": 0.7345886296641408, + "grad_norm": 2.9275269508361816, + "learning_rate": 3.280786826604121e-05, + "loss": 0.2294, + "step": 31337 + }, + { + "epoch": 0.7346120712389458, + "grad_norm": 4.688579559326172, + "learning_rate": 3.2802412905833245e-05, + "loss": 0.2767, + "step": 31338 + }, + { + "epoch": 0.7346355128137508, + "grad_norm": 1.0658841133117676, + "learning_rate": 3.279695791024129e-05, + "loss": 0.1118, + "step": 31339 + }, + { + "epoch": 0.7346589543885558, + "grad_norm": 2.2463059425354004, + "learning_rate": 3.279150327929492e-05, + "loss": 0.2124, + "step": 31340 + }, + { + "epoch": 0.7346823959633608, + "grad_norm": 1.8192095756530762, + "learning_rate": 3.278604901302371e-05, + "loss": 0.1132, + "step": 31341 + }, + { + "epoch": 0.7347058375381658, + "grad_norm": 2.725788116455078, + "learning_rate": 3.278059511145733e-05, + "loss": 0.1798, + "step": 31342 + }, + { + "epoch": 0.7347292791129708, + "grad_norm": 2.07696270942688, + "learning_rate": 3.277514157462531e-05, + "loss": 0.1673, + "step": 31343 + }, + { + "epoch": 0.7347527206877758, + "grad_norm": 3.2944798469543457, + "learning_rate": 3.2769688402557287e-05, + "loss": 0.3577, + "step": 31344 + }, + { + "epoch": 0.7347761622625808, + "grad_norm": 1.821855902671814, + "learning_rate": 3.276423559528281e-05, + "loss": 0.1071, + "step": 31345 + }, + { + "epoch": 0.7347996038373859, + "grad_norm": 2.017292022705078, + "learning_rate": 3.2758783152831485e-05, + "loss": 1.5036, + "step": 31346 + }, + { + "epoch": 0.7348230454121908, + "grad_norm": 3.246880292892456, + "learning_rate": 3.2753331075232865e-05, + "loss": 0.1922, + "step": 31347 + }, + { + "epoch": 0.7348464869869958, + "grad_norm": 3.2645933628082275, + "learning_rate": 3.274787936251656e-05, + "loss": 0.4021, + "step": 31348 + }, + { + "epoch": 0.7348699285618008, + "grad_norm": 4.9343461990356445, + "learning_rate": 3.274242801471218e-05, + "loss": 0.3536, + "step": 31349 + }, + { + "epoch": 0.7348933701366058, + "grad_norm": 3.811424493789673, + "learning_rate": 3.273697703184927e-05, + "loss": 0.2818, + "step": 31350 + }, + { + "epoch": 0.7349168117114108, + "grad_norm": 2.098243236541748, + "learning_rate": 3.273152641395742e-05, + "loss": 0.1165, + "step": 31351 + }, + { + "epoch": 0.7349402532862158, + "grad_norm": 4.17331075668335, + "learning_rate": 3.272607616106621e-05, + "loss": 0.3566, + "step": 31352 + }, + { + "epoch": 0.7349636948610208, + "grad_norm": 3.581875801086426, + "learning_rate": 3.272062627320519e-05, + "loss": 0.2855, + "step": 31353 + }, + { + "epoch": 0.7349871364358258, + "grad_norm": 1.838921308517456, + "learning_rate": 3.271517675040392e-05, + "loss": 0.8724, + "step": 31354 + }, + { + "epoch": 0.7350105780106307, + "grad_norm": 1.1544768810272217, + "learning_rate": 3.2709727592692016e-05, + "loss": 0.1493, + "step": 31355 + }, + { + "epoch": 0.7350340195854358, + "grad_norm": 2.1368963718414307, + "learning_rate": 3.270427880009903e-05, + "loss": 0.2322, + "step": 31356 + }, + { + "epoch": 0.7350574611602407, + "grad_norm": 3.865429162979126, + "learning_rate": 3.269883037265451e-05, + "loss": 0.1766, + "step": 31357 + }, + { + "epoch": 0.7350809027350458, + "grad_norm": 4.359877109527588, + "learning_rate": 3.2693382310388046e-05, + "loss": 0.1864, + "step": 31358 + }, + { + "epoch": 0.7351043443098507, + "grad_norm": 3.3270645141601562, + "learning_rate": 3.268793461332913e-05, + "loss": 1.6665, + "step": 31359 + }, + { + "epoch": 0.7351277858846558, + "grad_norm": 3.3401706218719482, + "learning_rate": 3.268248728150738e-05, + "loss": 1.8757, + "step": 31360 + }, + { + "epoch": 0.7351512274594607, + "grad_norm": 1.2224726676940918, + "learning_rate": 3.267704031495239e-05, + "loss": 0.7374, + "step": 31361 + }, + { + "epoch": 0.7351746690342658, + "grad_norm": 5.635310649871826, + "learning_rate": 3.267159371369366e-05, + "loss": 0.3017, + "step": 31362 + }, + { + "epoch": 0.7351981106090707, + "grad_norm": 0.6860349774360657, + "learning_rate": 3.266614747776075e-05, + "loss": 0.0608, + "step": 31363 + }, + { + "epoch": 0.7352215521838757, + "grad_norm": 2.0536861419677734, + "learning_rate": 3.266070160718323e-05, + "loss": 0.1686, + "step": 31364 + }, + { + "epoch": 0.7352449937586807, + "grad_norm": 2.7838656902313232, + "learning_rate": 3.265525610199064e-05, + "loss": 1.5032, + "step": 31365 + }, + { + "epoch": 0.7352684353334857, + "grad_norm": 5.492789268493652, + "learning_rate": 3.264981096221248e-05, + "loss": 0.3939, + "step": 31366 + }, + { + "epoch": 0.7352918769082907, + "grad_norm": 2.3211774826049805, + "learning_rate": 3.2644366187878375e-05, + "loss": 0.1621, + "step": 31367 + }, + { + "epoch": 0.7353153184830957, + "grad_norm": 2.2148549556732178, + "learning_rate": 3.263892177901784e-05, + "loss": 0.1501, + "step": 31368 + }, + { + "epoch": 0.7353387600579007, + "grad_norm": 2.1767334938049316, + "learning_rate": 3.26334777356604e-05, + "loss": 0.1587, + "step": 31369 + }, + { + "epoch": 0.7353622016327057, + "grad_norm": 2.544597864151001, + "learning_rate": 3.262803405783561e-05, + "loss": 1.2432, + "step": 31370 + }, + { + "epoch": 0.7353856432075107, + "grad_norm": 4.8168840408325195, + "learning_rate": 3.2622590745572957e-05, + "loss": 0.5597, + "step": 31371 + }, + { + "epoch": 0.7354090847823157, + "grad_norm": 2.065239191055298, + "learning_rate": 3.261714779890207e-05, + "loss": 0.1175, + "step": 31372 + }, + { + "epoch": 0.7354325263571206, + "grad_norm": 4.117275238037109, + "learning_rate": 3.2611705217852393e-05, + "loss": 0.6088, + "step": 31373 + }, + { + "epoch": 0.7354559679319257, + "grad_norm": 2.939568042755127, + "learning_rate": 3.260626300245353e-05, + "loss": 0.3097, + "step": 31374 + }, + { + "epoch": 0.7354794095067306, + "grad_norm": 2.2790493965148926, + "learning_rate": 3.260082115273498e-05, + "loss": 0.1347, + "step": 31375 + }, + { + "epoch": 0.7355028510815357, + "grad_norm": 4.912323474884033, + "learning_rate": 3.259537966872627e-05, + "loss": 0.3948, + "step": 31376 + }, + { + "epoch": 0.7355262926563406, + "grad_norm": 0.6486014127731323, + "learning_rate": 3.258993855045692e-05, + "loss": 0.067, + "step": 31377 + }, + { + "epoch": 0.7355497342311457, + "grad_norm": 3.6519198417663574, + "learning_rate": 3.258449779795647e-05, + "loss": 0.2573, + "step": 31378 + }, + { + "epoch": 0.7355731758059506, + "grad_norm": 1.5770834684371948, + "learning_rate": 3.257905741125439e-05, + "loss": 1.6613, + "step": 31379 + }, + { + "epoch": 0.7355966173807557, + "grad_norm": 2.8130314350128174, + "learning_rate": 3.2573617390380276e-05, + "loss": 0.2979, + "step": 31380 + }, + { + "epoch": 0.7356200589555606, + "grad_norm": 2.1255693435668945, + "learning_rate": 3.2568177735363615e-05, + "loss": 0.1493, + "step": 31381 + }, + { + "epoch": 0.7356435005303656, + "grad_norm": 3.3907742500305176, + "learning_rate": 3.256273844623391e-05, + "loss": 0.1257, + "step": 31382 + }, + { + "epoch": 0.7356669421051706, + "grad_norm": 1.279024362564087, + "learning_rate": 3.2557299523020644e-05, + "loss": 0.1044, + "step": 31383 + }, + { + "epoch": 0.7356903836799756, + "grad_norm": 0.5998579263687134, + "learning_rate": 3.255186096575341e-05, + "loss": 0.0518, + "step": 31384 + }, + { + "epoch": 0.7357138252547806, + "grad_norm": 3.197885751724243, + "learning_rate": 3.254642277446167e-05, + "loss": 0.1888, + "step": 31385 + }, + { + "epoch": 0.7357372668295856, + "grad_norm": 2.9258201122283936, + "learning_rate": 3.254098494917491e-05, + "loss": 0.16, + "step": 31386 + }, + { + "epoch": 0.7357607084043906, + "grad_norm": 0.8201632499694824, + "learning_rate": 3.2535547489922694e-05, + "loss": 0.0629, + "step": 31387 + }, + { + "epoch": 0.7357841499791956, + "grad_norm": 2.8785152435302734, + "learning_rate": 3.2530110396734494e-05, + "loss": 0.193, + "step": 31388 + }, + { + "epoch": 0.7358075915540006, + "grad_norm": 0.593030571937561, + "learning_rate": 3.25246736696398e-05, + "loss": 0.0335, + "step": 31389 + }, + { + "epoch": 0.7358310331288056, + "grad_norm": 2.651092767715454, + "learning_rate": 3.251923730866814e-05, + "loss": 0.1766, + "step": 31390 + }, + { + "epoch": 0.7358544747036105, + "grad_norm": 3.1486334800720215, + "learning_rate": 3.251380131384899e-05, + "loss": 0.3797, + "step": 31391 + }, + { + "epoch": 0.7358779162784156, + "grad_norm": 0.7951274514198303, + "learning_rate": 3.2508365685211815e-05, + "loss": 0.0778, + "step": 31392 + }, + { + "epoch": 0.7359013578532205, + "grad_norm": 0.9672961831092834, + "learning_rate": 3.250293042278618e-05, + "loss": 0.0955, + "step": 31393 + }, + { + "epoch": 0.7359247994280256, + "grad_norm": 1.0517112016677856, + "learning_rate": 3.249749552660154e-05, + "loss": 0.1304, + "step": 31394 + }, + { + "epoch": 0.7359482410028305, + "grad_norm": 2.907538652420044, + "learning_rate": 3.2492060996687356e-05, + "loss": 0.2309, + "step": 31395 + }, + { + "epoch": 0.7359716825776356, + "grad_norm": 3.232663154602051, + "learning_rate": 3.2486626833073186e-05, + "loss": 0.1716, + "step": 31396 + }, + { + "epoch": 0.7359951241524406, + "grad_norm": 1.177027940750122, + "learning_rate": 3.248119303578847e-05, + "loss": 0.1756, + "step": 31397 + }, + { + "epoch": 0.7360185657272456, + "grad_norm": 2.473189115524292, + "learning_rate": 3.2475759604862676e-05, + "loss": 0.2348, + "step": 31398 + }, + { + "epoch": 0.7360420073020506, + "grad_norm": 1.5456175804138184, + "learning_rate": 3.247032654032534e-05, + "loss": 0.1119, + "step": 31399 + }, + { + "epoch": 0.7360654488768555, + "grad_norm": 1.1792107820510864, + "learning_rate": 3.24648938422059e-05, + "loss": 0.1125, + "step": 31400 + }, + { + "epoch": 0.7360888904516606, + "grad_norm": 4.479422092437744, + "learning_rate": 3.245946151053385e-05, + "loss": 0.2912, + "step": 31401 + }, + { + "epoch": 0.7361123320264655, + "grad_norm": 2.3474879264831543, + "learning_rate": 3.245402954533866e-05, + "loss": 0.2325, + "step": 31402 + }, + { + "epoch": 0.7361357736012706, + "grad_norm": 0.9317773580551147, + "learning_rate": 3.244859794664981e-05, + "loss": 0.0605, + "step": 31403 + }, + { + "epoch": 0.7361592151760755, + "grad_norm": 2.866701602935791, + "learning_rate": 3.244316671449673e-05, + "loss": 0.3173, + "step": 31404 + }, + { + "epoch": 0.7361826567508806, + "grad_norm": 4.018904209136963, + "learning_rate": 3.2437735848908954e-05, + "loss": 0.2928, + "step": 31405 + }, + { + "epoch": 0.7362060983256855, + "grad_norm": 0.8894152641296387, + "learning_rate": 3.2432305349915936e-05, + "loss": 0.1251, + "step": 31406 + }, + { + "epoch": 0.7362295399004906, + "grad_norm": 2.7706615924835205, + "learning_rate": 3.242687521754713e-05, + "loss": 0.2908, + "step": 31407 + }, + { + "epoch": 0.7362529814752955, + "grad_norm": 2.351573944091797, + "learning_rate": 3.2421445451831956e-05, + "loss": 0.3045, + "step": 31408 + }, + { + "epoch": 0.7362764230501005, + "grad_norm": 1.5404926538467407, + "learning_rate": 3.241601605279995e-05, + "loss": 0.1445, + "step": 31409 + }, + { + "epoch": 0.7362998646249055, + "grad_norm": 3.1488375663757324, + "learning_rate": 3.241058702048054e-05, + "loss": 0.2878, + "step": 31410 + }, + { + "epoch": 0.7363233061997105, + "grad_norm": 2.5483789443969727, + "learning_rate": 3.2405158354903164e-05, + "loss": 1.6026, + "step": 31411 + }, + { + "epoch": 0.7363467477745155, + "grad_norm": 2.3375258445739746, + "learning_rate": 3.2399730056097335e-05, + "loss": 0.1709, + "step": 31412 + }, + { + "epoch": 0.7363701893493205, + "grad_norm": 3.2584643363952637, + "learning_rate": 3.2394302124092456e-05, + "loss": 1.3089, + "step": 31413 + }, + { + "epoch": 0.7363936309241255, + "grad_norm": 1.6152164936065674, + "learning_rate": 3.2388874558918004e-05, + "loss": 0.1046, + "step": 31414 + }, + { + "epoch": 0.7364170724989305, + "grad_norm": 1.538534164428711, + "learning_rate": 3.2383447360603414e-05, + "loss": 0.096, + "step": 31415 + }, + { + "epoch": 0.7364405140737355, + "grad_norm": 3.8831210136413574, + "learning_rate": 3.2378020529178144e-05, + "loss": 0.2526, + "step": 31416 + }, + { + "epoch": 0.7364639556485405, + "grad_norm": 2.802370548248291, + "learning_rate": 3.23725940646716e-05, + "loss": 0.269, + "step": 31417 + }, + { + "epoch": 0.7364873972233454, + "grad_norm": 3.219364643096924, + "learning_rate": 3.23671679671133e-05, + "loss": 0.2977, + "step": 31418 + }, + { + "epoch": 0.7365108387981505, + "grad_norm": 1.9990428686141968, + "learning_rate": 3.236174223653264e-05, + "loss": 0.205, + "step": 31419 + }, + { + "epoch": 0.7365342803729554, + "grad_norm": 1.056837558746338, + "learning_rate": 3.2356316872959044e-05, + "loss": 0.097, + "step": 31420 + }, + { + "epoch": 0.7365577219477605, + "grad_norm": 2.3280386924743652, + "learning_rate": 3.2350891876422007e-05, + "loss": 0.1303, + "step": 31421 + }, + { + "epoch": 0.7365811635225654, + "grad_norm": 0.4046774208545685, + "learning_rate": 3.234546724695093e-05, + "loss": 0.0198, + "step": 31422 + }, + { + "epoch": 0.7366046050973705, + "grad_norm": 2.6982052326202393, + "learning_rate": 3.234004298457525e-05, + "loss": 1.6045, + "step": 31423 + }, + { + "epoch": 0.7366280466721754, + "grad_norm": 3.0286076068878174, + "learning_rate": 3.233461908932437e-05, + "loss": 0.2234, + "step": 31424 + }, + { + "epoch": 0.7366514882469805, + "grad_norm": 1.395918846130371, + "learning_rate": 3.2329195561227776e-05, + "loss": 0.1796, + "step": 31425 + }, + { + "epoch": 0.7366749298217854, + "grad_norm": 1.5071662664413452, + "learning_rate": 3.232377240031488e-05, + "loss": 0.0808, + "step": 31426 + }, + { + "epoch": 0.7366983713965904, + "grad_norm": 3.6165847778320312, + "learning_rate": 3.2318349606615086e-05, + "loss": 0.4672, + "step": 31427 + }, + { + "epoch": 0.7367218129713954, + "grad_norm": 2.6941423416137695, + "learning_rate": 3.231292718015784e-05, + "loss": 0.238, + "step": 31428 + }, + { + "epoch": 0.7367452545462004, + "grad_norm": 2.9135706424713135, + "learning_rate": 3.2307505120972516e-05, + "loss": 0.1784, + "step": 31429 + }, + { + "epoch": 0.7367686961210054, + "grad_norm": 2.8882830142974854, + "learning_rate": 3.2302083429088606e-05, + "loss": 0.3393, + "step": 31430 + }, + { + "epoch": 0.7367921376958104, + "grad_norm": 0.823320746421814, + "learning_rate": 3.229666210453549e-05, + "loss": 0.0585, + "step": 31431 + }, + { + "epoch": 0.7368155792706154, + "grad_norm": 2.2492613792419434, + "learning_rate": 3.229124114734255e-05, + "loss": 0.1453, + "step": 31432 + }, + { + "epoch": 0.7368390208454204, + "grad_norm": 2.8912158012390137, + "learning_rate": 3.2285820557539295e-05, + "loss": 0.1986, + "step": 31433 + }, + { + "epoch": 0.7368624624202254, + "grad_norm": 1.819100022315979, + "learning_rate": 3.228040033515507e-05, + "loss": 0.1056, + "step": 31434 + }, + { + "epoch": 0.7368859039950304, + "grad_norm": 3.03574800491333, + "learning_rate": 3.2274980480219296e-05, + "loss": 0.2903, + "step": 31435 + }, + { + "epoch": 0.7369093455698353, + "grad_norm": 1.005369782447815, + "learning_rate": 3.226956099276135e-05, + "loss": 0.1229, + "step": 31436 + }, + { + "epoch": 0.7369327871446404, + "grad_norm": 3.6467373371124268, + "learning_rate": 3.22641418728107e-05, + "loss": 0.3261, + "step": 31437 + }, + { + "epoch": 0.7369562287194453, + "grad_norm": 2.871690511703491, + "learning_rate": 3.225872312039673e-05, + "loss": 0.1566, + "step": 31438 + }, + { + "epoch": 0.7369796702942504, + "grad_norm": 1.672269582748413, + "learning_rate": 3.225330473554882e-05, + "loss": 0.0713, + "step": 31439 + }, + { + "epoch": 0.7370031118690553, + "grad_norm": 2.982544183731079, + "learning_rate": 3.2247886718296396e-05, + "loss": 1.2406, + "step": 31440 + }, + { + "epoch": 0.7370265534438604, + "grad_norm": 3.3233282566070557, + "learning_rate": 3.2242469068668836e-05, + "loss": 1.5266, + "step": 31441 + }, + { + "epoch": 0.7370499950186653, + "grad_norm": 3.087468147277832, + "learning_rate": 3.223705178669551e-05, + "loss": 0.405, + "step": 31442 + }, + { + "epoch": 0.7370734365934704, + "grad_norm": 2.1160242557525635, + "learning_rate": 3.223163487240588e-05, + "loss": 0.1797, + "step": 31443 + }, + { + "epoch": 0.7370968781682753, + "grad_norm": 2.7153544425964355, + "learning_rate": 3.222621832582927e-05, + "loss": 0.2176, + "step": 31444 + }, + { + "epoch": 0.7371203197430803, + "grad_norm": 1.7487062215805054, + "learning_rate": 3.222080214699514e-05, + "loss": 0.1821, + "step": 31445 + }, + { + "epoch": 0.7371437613178853, + "grad_norm": 3.196796178817749, + "learning_rate": 3.221538633593285e-05, + "loss": 0.3132, + "step": 31446 + }, + { + "epoch": 0.7371672028926903, + "grad_norm": 0.625499427318573, + "learning_rate": 3.220997089267178e-05, + "loss": 0.0394, + "step": 31447 + }, + { + "epoch": 0.7371906444674954, + "grad_norm": 3.2828261852264404, + "learning_rate": 3.22045558172413e-05, + "loss": 0.2563, + "step": 31448 + }, + { + "epoch": 0.7372140860423003, + "grad_norm": 0.966103732585907, + "learning_rate": 3.2199141109670784e-05, + "loss": 0.0807, + "step": 31449 + }, + { + "epoch": 0.7372375276171054, + "grad_norm": 1.3863730430603027, + "learning_rate": 3.219372676998967e-05, + "loss": 0.091, + "step": 31450 + }, + { + "epoch": 0.7372609691919103, + "grad_norm": 1.6146632432937622, + "learning_rate": 3.218831279822729e-05, + "loss": 0.0752, + "step": 31451 + }, + { + "epoch": 0.7372844107667154, + "grad_norm": 1.5419951677322388, + "learning_rate": 3.218289919441304e-05, + "loss": 1.0739, + "step": 31452 + }, + { + "epoch": 0.7373078523415203, + "grad_norm": 2.721193552017212, + "learning_rate": 3.217748595857629e-05, + "loss": 0.2203, + "step": 31453 + }, + { + "epoch": 0.7373312939163253, + "grad_norm": 2.363907814025879, + "learning_rate": 3.21720730907464e-05, + "loss": 0.0942, + "step": 31454 + }, + { + "epoch": 0.7373547354911303, + "grad_norm": 3.9459643363952637, + "learning_rate": 3.2166660590952724e-05, + "loss": 0.2961, + "step": 31455 + }, + { + "epoch": 0.7373781770659353, + "grad_norm": 1.0578724145889282, + "learning_rate": 3.216124845922465e-05, + "loss": 0.0402, + "step": 31456 + }, + { + "epoch": 0.7374016186407403, + "grad_norm": 2.3596904277801514, + "learning_rate": 3.215583669559158e-05, + "loss": 0.8599, + "step": 31457 + }, + { + "epoch": 0.7374250602155453, + "grad_norm": 2.3908960819244385, + "learning_rate": 3.215042530008284e-05, + "loss": 0.1864, + "step": 31458 + }, + { + "epoch": 0.7374485017903503, + "grad_norm": 1.0059202909469604, + "learning_rate": 3.2145014272727814e-05, + "loss": 0.1134, + "step": 31459 + }, + { + "epoch": 0.7374719433651553, + "grad_norm": 3.6169629096984863, + "learning_rate": 3.2139603613555845e-05, + "loss": 0.2698, + "step": 31460 + }, + { + "epoch": 0.7374953849399603, + "grad_norm": 3.1968588829040527, + "learning_rate": 3.2134193322596295e-05, + "loss": 0.2261, + "step": 31461 + }, + { + "epoch": 0.7375188265147653, + "grad_norm": 0.6813744306564331, + "learning_rate": 3.2128783399878484e-05, + "loss": 0.0263, + "step": 31462 + }, + { + "epoch": 0.7375422680895702, + "grad_norm": 0.6851043701171875, + "learning_rate": 3.2123373845431836e-05, + "loss": 0.0477, + "step": 31463 + }, + { + "epoch": 0.7375657096643753, + "grad_norm": 2.5274112224578857, + "learning_rate": 3.211796465928567e-05, + "loss": 0.1817, + "step": 31464 + }, + { + "epoch": 0.7375891512391802, + "grad_norm": 2.583975076675415, + "learning_rate": 3.2112555841469335e-05, + "loss": 0.1667, + "step": 31465 + }, + { + "epoch": 0.7376125928139853, + "grad_norm": 1.8720347881317139, + "learning_rate": 3.210714739201218e-05, + "loss": 1.1828, + "step": 31466 + }, + { + "epoch": 0.7376360343887902, + "grad_norm": 3.190066337585449, + "learning_rate": 3.2101739310943536e-05, + "loss": 0.1446, + "step": 31467 + }, + { + "epoch": 0.7376594759635953, + "grad_norm": 1.0288461446762085, + "learning_rate": 3.209633159829278e-05, + "loss": 0.0746, + "step": 31468 + }, + { + "epoch": 0.7376829175384002, + "grad_norm": 0.747349739074707, + "learning_rate": 3.20909242540892e-05, + "loss": 0.1025, + "step": 31469 + }, + { + "epoch": 0.7377063591132053, + "grad_norm": 2.569281578063965, + "learning_rate": 3.208551727836222e-05, + "loss": 0.2981, + "step": 31470 + }, + { + "epoch": 0.7377298006880102, + "grad_norm": 0.669064462184906, + "learning_rate": 3.208011067114113e-05, + "loss": 0.0431, + "step": 31471 + }, + { + "epoch": 0.7377532422628152, + "grad_norm": 1.833563208580017, + "learning_rate": 3.207470443245527e-05, + "loss": 0.1332, + "step": 31472 + }, + { + "epoch": 0.7377766838376202, + "grad_norm": 1.5743091106414795, + "learning_rate": 3.206929856233397e-05, + "loss": 0.166, + "step": 31473 + }, + { + "epoch": 0.7378001254124252, + "grad_norm": 0.7892115116119385, + "learning_rate": 3.206389306080654e-05, + "loss": 0.073, + "step": 31474 + }, + { + "epoch": 0.7378235669872302, + "grad_norm": 0.4075133502483368, + "learning_rate": 3.2058487927902357e-05, + "loss": 0.0367, + "step": 31475 + }, + { + "epoch": 0.7378470085620352, + "grad_norm": 1.0757871866226196, + "learning_rate": 3.205308316365074e-05, + "loss": 0.1394, + "step": 31476 + }, + { + "epoch": 0.7378704501368402, + "grad_norm": 2.529203176498413, + "learning_rate": 3.2047678768081e-05, + "loss": 0.3599, + "step": 31477 + }, + { + "epoch": 0.7378938917116452, + "grad_norm": 2.793537139892578, + "learning_rate": 3.2042274741222464e-05, + "loss": 0.2957, + "step": 31478 + }, + { + "epoch": 0.7379173332864501, + "grad_norm": 0.3590867817401886, + "learning_rate": 3.203687108310446e-05, + "loss": 0.0169, + "step": 31479 + }, + { + "epoch": 0.7379407748612552, + "grad_norm": 3.2027218341827393, + "learning_rate": 3.2031467793756265e-05, + "loss": 0.9796, + "step": 31480 + }, + { + "epoch": 0.7379642164360601, + "grad_norm": 2.5883841514587402, + "learning_rate": 3.202606487320724e-05, + "loss": 0.08, + "step": 31481 + }, + { + "epoch": 0.7379876580108652, + "grad_norm": 0.9411638379096985, + "learning_rate": 3.202066232148674e-05, + "loss": 0.1188, + "step": 31482 + }, + { + "epoch": 0.7380110995856701, + "grad_norm": 0.5573824048042297, + "learning_rate": 3.2015260138624025e-05, + "loss": 0.0611, + "step": 31483 + }, + { + "epoch": 0.7380345411604752, + "grad_norm": 2.359830379486084, + "learning_rate": 3.200985832464842e-05, + "loss": 0.2083, + "step": 31484 + }, + { + "epoch": 0.7380579827352801, + "grad_norm": 2.8419277667999268, + "learning_rate": 3.200445687958925e-05, + "loss": 0.2926, + "step": 31485 + }, + { + "epoch": 0.7380814243100852, + "grad_norm": 2.277858257293701, + "learning_rate": 3.1999055803475796e-05, + "loss": 0.2072, + "step": 31486 + }, + { + "epoch": 0.7381048658848901, + "grad_norm": 2.6300904750823975, + "learning_rate": 3.199365509633735e-05, + "loss": 0.0918, + "step": 31487 + }, + { + "epoch": 0.7381283074596952, + "grad_norm": 1.6406660079956055, + "learning_rate": 3.1988254758203275e-05, + "loss": 0.6559, + "step": 31488 + }, + { + "epoch": 0.7381517490345001, + "grad_norm": 0.4334245026111603, + "learning_rate": 3.198285478910284e-05, + "loss": 0.0228, + "step": 31489 + }, + { + "epoch": 0.7381751906093051, + "grad_norm": 2.1693742275238037, + "learning_rate": 3.197745518906535e-05, + "loss": 0.1352, + "step": 31490 + }, + { + "epoch": 0.7381986321841101, + "grad_norm": 2.4189140796661377, + "learning_rate": 3.1972055958120094e-05, + "loss": 0.1738, + "step": 31491 + }, + { + "epoch": 0.7382220737589151, + "grad_norm": 2.0278639793395996, + "learning_rate": 3.196665709629638e-05, + "loss": 0.1927, + "step": 31492 + }, + { + "epoch": 0.7382455153337201, + "grad_norm": 3.3082334995269775, + "learning_rate": 3.196125860362347e-05, + "loss": 0.3322, + "step": 31493 + }, + { + "epoch": 0.7382689569085251, + "grad_norm": 5.783205509185791, + "learning_rate": 3.1955860480130686e-05, + "loss": 0.5867, + "step": 31494 + }, + { + "epoch": 0.7382923984833301, + "grad_norm": 1.0517878532409668, + "learning_rate": 3.195046272584734e-05, + "loss": 0.1016, + "step": 31495 + }, + { + "epoch": 0.7383158400581351, + "grad_norm": 2.9940507411956787, + "learning_rate": 3.1945065340802694e-05, + "loss": 0.3464, + "step": 31496 + }, + { + "epoch": 0.73833928163294, + "grad_norm": 1.411360263824463, + "learning_rate": 3.193966832502604e-05, + "loss": 0.0715, + "step": 31497 + }, + { + "epoch": 0.7383627232077451, + "grad_norm": 2.3500537872314453, + "learning_rate": 3.193427167854667e-05, + "loss": 0.1491, + "step": 31498 + }, + { + "epoch": 0.73838616478255, + "grad_norm": 2.3517205715179443, + "learning_rate": 3.192887540139384e-05, + "loss": 0.1317, + "step": 31499 + }, + { + "epoch": 0.7384096063573551, + "grad_norm": 3.1990432739257812, + "learning_rate": 3.192347949359682e-05, + "loss": 1.9731, + "step": 31500 + }, + { + "epoch": 0.7384330479321601, + "grad_norm": 0.8980423212051392, + "learning_rate": 3.1918083955184955e-05, + "loss": 0.073, + "step": 31501 + }, + { + "epoch": 0.7384564895069651, + "grad_norm": 2.7089929580688477, + "learning_rate": 3.191268878618747e-05, + "loss": 1.2041, + "step": 31502 + }, + { + "epoch": 0.7384799310817701, + "grad_norm": 2.8210108280181885, + "learning_rate": 3.1907293986633646e-05, + "loss": 0.2458, + "step": 31503 + }, + { + "epoch": 0.7385033726565751, + "grad_norm": 1.4809739589691162, + "learning_rate": 3.1901899556552774e-05, + "loss": 0.0614, + "step": 31504 + }, + { + "epoch": 0.7385268142313801, + "grad_norm": 1.0935057401657104, + "learning_rate": 3.189650549597407e-05, + "loss": 0.0926, + "step": 31505 + }, + { + "epoch": 0.738550255806185, + "grad_norm": 2.5284109115600586, + "learning_rate": 3.189111180492688e-05, + "loss": 0.2449, + "step": 31506 + }, + { + "epoch": 0.7385736973809901, + "grad_norm": 3.7541215419769287, + "learning_rate": 3.188571848344038e-05, + "loss": 0.2929, + "step": 31507 + }, + { + "epoch": 0.738597138955795, + "grad_norm": 3.0731937885284424, + "learning_rate": 3.188032553154394e-05, + "loss": 0.335, + "step": 31508 + }, + { + "epoch": 0.7386205805306001, + "grad_norm": 3.0219802856445312, + "learning_rate": 3.1874932949266764e-05, + "loss": 0.1523, + "step": 31509 + }, + { + "epoch": 0.738644022105405, + "grad_norm": 0.6021931171417236, + "learning_rate": 3.186954073663811e-05, + "loss": 0.0602, + "step": 31510 + }, + { + "epoch": 0.7386674636802101, + "grad_norm": 1.1694583892822266, + "learning_rate": 3.1864148893687254e-05, + "loss": 0.1552, + "step": 31511 + }, + { + "epoch": 0.738690905255015, + "grad_norm": 3.628464937210083, + "learning_rate": 3.185875742044341e-05, + "loss": 0.2758, + "step": 31512 + }, + { + "epoch": 0.7387143468298201, + "grad_norm": 1.7096316814422607, + "learning_rate": 3.1853366316935886e-05, + "loss": 0.176, + "step": 31513 + }, + { + "epoch": 0.738737788404625, + "grad_norm": 0.9638229608535767, + "learning_rate": 3.184797558319392e-05, + "loss": 0.0915, + "step": 31514 + }, + { + "epoch": 0.7387612299794301, + "grad_norm": 1.6511205434799194, + "learning_rate": 3.184258521924676e-05, + "loss": 0.1035, + "step": 31515 + }, + { + "epoch": 0.738784671554235, + "grad_norm": 2.8308708667755127, + "learning_rate": 3.183719522512364e-05, + "loss": 0.1554, + "step": 31516 + }, + { + "epoch": 0.73880811312904, + "grad_norm": 3.2470762729644775, + "learning_rate": 3.183180560085378e-05, + "loss": 0.3044, + "step": 31517 + }, + { + "epoch": 0.738831554703845, + "grad_norm": 3.361128568649292, + "learning_rate": 3.1826416346466495e-05, + "loss": 0.2907, + "step": 31518 + }, + { + "epoch": 0.73885499627865, + "grad_norm": 0.3871641457080841, + "learning_rate": 3.182102746199096e-05, + "loss": 0.02, + "step": 31519 + }, + { + "epoch": 0.738878437853455, + "grad_norm": 5.467631816864014, + "learning_rate": 3.181563894745648e-05, + "loss": 0.2658, + "step": 31520 + }, + { + "epoch": 0.73890187942826, + "grad_norm": 2.5199687480926514, + "learning_rate": 3.1810250802892264e-05, + "loss": 0.2725, + "step": 31521 + }, + { + "epoch": 0.738925321003065, + "grad_norm": 2.962158679962158, + "learning_rate": 3.1804863028327536e-05, + "loss": 0.2434, + "step": 31522 + }, + { + "epoch": 0.73894876257787, + "grad_norm": 0.22192271053791046, + "learning_rate": 3.179947562379155e-05, + "loss": 0.0078, + "step": 31523 + }, + { + "epoch": 0.738972204152675, + "grad_norm": 2.190762519836426, + "learning_rate": 3.179408858931352e-05, + "loss": 0.1909, + "step": 31524 + }, + { + "epoch": 0.73899564572748, + "grad_norm": 2.29556941986084, + "learning_rate": 3.178870192492265e-05, + "loss": 0.2161, + "step": 31525 + }, + { + "epoch": 0.7390190873022849, + "grad_norm": 2.8818016052246094, + "learning_rate": 3.1783315630648236e-05, + "loss": 0.2624, + "step": 31526 + }, + { + "epoch": 0.73904252887709, + "grad_norm": 1.2859481573104858, + "learning_rate": 3.177792970651947e-05, + "loss": 0.0946, + "step": 31527 + }, + { + "epoch": 0.7390659704518949, + "grad_norm": 2.5368168354034424, + "learning_rate": 3.177254415256558e-05, + "loss": 0.192, + "step": 31528 + }, + { + "epoch": 0.7390894120267, + "grad_norm": 2.312734842300415, + "learning_rate": 3.1767158968815734e-05, + "loss": 0.153, + "step": 31529 + }, + { + "epoch": 0.7391128536015049, + "grad_norm": 2.057814359664917, + "learning_rate": 3.176177415529925e-05, + "loss": 0.0686, + "step": 31530 + }, + { + "epoch": 0.73913629517631, + "grad_norm": 1.551249623298645, + "learning_rate": 3.17563897120453e-05, + "loss": 0.707, + "step": 31531 + }, + { + "epoch": 0.7391597367511149, + "grad_norm": 0.6547471284866333, + "learning_rate": 3.1751005639083054e-05, + "loss": 0.0345, + "step": 31532 + }, + { + "epoch": 0.73918317832592, + "grad_norm": 3.1002860069274902, + "learning_rate": 3.17456219364418e-05, + "loss": 1.7761, + "step": 31533 + }, + { + "epoch": 0.7392066199007249, + "grad_norm": 2.392677068710327, + "learning_rate": 3.174023860415074e-05, + "loss": 0.19, + "step": 31534 + }, + { + "epoch": 0.7392300614755299, + "grad_norm": 3.025893211364746, + "learning_rate": 3.173485564223906e-05, + "loss": 0.2039, + "step": 31535 + }, + { + "epoch": 0.7392535030503349, + "grad_norm": 2.012052297592163, + "learning_rate": 3.172947305073596e-05, + "loss": 1.56, + "step": 31536 + }, + { + "epoch": 0.7392769446251399, + "grad_norm": 5.1790876388549805, + "learning_rate": 3.172409082967066e-05, + "loss": 0.4663, + "step": 31537 + }, + { + "epoch": 0.7393003861999449, + "grad_norm": 3.6053996086120605, + "learning_rate": 3.171870897907234e-05, + "loss": 0.2038, + "step": 31538 + }, + { + "epoch": 0.7393238277747499, + "grad_norm": 3.363790988922119, + "learning_rate": 3.171332749897026e-05, + "loss": 0.4055, + "step": 31539 + }, + { + "epoch": 0.7393472693495549, + "grad_norm": 2.380164861679077, + "learning_rate": 3.170794638939357e-05, + "loss": 0.19, + "step": 31540 + }, + { + "epoch": 0.7393707109243599, + "grad_norm": 3.0742907524108887, + "learning_rate": 3.170256565037146e-05, + "loss": 0.1353, + "step": 31541 + }, + { + "epoch": 0.7393941524991648, + "grad_norm": 5.375200271606445, + "learning_rate": 3.1697185281933186e-05, + "loss": 0.7896, + "step": 31542 + }, + { + "epoch": 0.7394175940739699, + "grad_norm": 0.9581630825996399, + "learning_rate": 3.16918052841079e-05, + "loss": 0.0572, + "step": 31543 + }, + { + "epoch": 0.7394410356487748, + "grad_norm": 0.8788305521011353, + "learning_rate": 3.16864256569248e-05, + "loss": 0.0342, + "step": 31544 + }, + { + "epoch": 0.7394644772235799, + "grad_norm": 2.0647430419921875, + "learning_rate": 3.1681046400413036e-05, + "loss": 0.2181, + "step": 31545 + }, + { + "epoch": 0.7394879187983848, + "grad_norm": 0.7627224922180176, + "learning_rate": 3.167566751460188e-05, + "loss": 0.0539, + "step": 31546 + }, + { + "epoch": 0.7395113603731899, + "grad_norm": 1.073431372642517, + "learning_rate": 3.167028899952046e-05, + "loss": 0.071, + "step": 31547 + }, + { + "epoch": 0.7395348019479948, + "grad_norm": 2.697968006134033, + "learning_rate": 3.166491085519798e-05, + "loss": 0.1431, + "step": 31548 + }, + { + "epoch": 0.7395582435227999, + "grad_norm": 1.4365575313568115, + "learning_rate": 3.165953308166362e-05, + "loss": 0.1432, + "step": 31549 + }, + { + "epoch": 0.7395816850976048, + "grad_norm": 2.9242045879364014, + "learning_rate": 3.165415567894652e-05, + "loss": 0.2081, + "step": 31550 + }, + { + "epoch": 0.7396051266724099, + "grad_norm": 2.9721646308898926, + "learning_rate": 3.164877864707591e-05, + "loss": 0.2504, + "step": 31551 + }, + { + "epoch": 0.7396285682472149, + "grad_norm": 0.9609702229499817, + "learning_rate": 3.1643401986080967e-05, + "loss": 0.1011, + "step": 31552 + }, + { + "epoch": 0.7396520098220198, + "grad_norm": 1.0986641645431519, + "learning_rate": 3.163802569599081e-05, + "loss": 0.1142, + "step": 31553 + }, + { + "epoch": 0.7396754513968249, + "grad_norm": 1.2133464813232422, + "learning_rate": 3.163264977683468e-05, + "loss": 0.1051, + "step": 31554 + }, + { + "epoch": 0.7396988929716298, + "grad_norm": 2.533059597015381, + "learning_rate": 3.1627274228641714e-05, + "loss": 0.2158, + "step": 31555 + }, + { + "epoch": 0.7397223345464349, + "grad_norm": 2.0830748081207275, + "learning_rate": 3.1621899051441084e-05, + "loss": 0.1417, + "step": 31556 + }, + { + "epoch": 0.7397457761212398, + "grad_norm": 0.880563497543335, + "learning_rate": 3.161652424526191e-05, + "loss": 0.0625, + "step": 31557 + }, + { + "epoch": 0.7397692176960449, + "grad_norm": 3.607672691345215, + "learning_rate": 3.161114981013345e-05, + "loss": 0.3968, + "step": 31558 + }, + { + "epoch": 0.7397926592708498, + "grad_norm": 4.9498291015625, + "learning_rate": 3.16057757460848e-05, + "loss": 0.3111, + "step": 31559 + }, + { + "epoch": 0.7398161008456549, + "grad_norm": 0.47770532965660095, + "learning_rate": 3.1600402053145126e-05, + "loss": 0.0397, + "step": 31560 + }, + { + "epoch": 0.7398395424204598, + "grad_norm": 2.0228846073150635, + "learning_rate": 3.159502873134361e-05, + "loss": 0.9257, + "step": 31561 + }, + { + "epoch": 0.7398629839952648, + "grad_norm": 3.574249505996704, + "learning_rate": 3.1589655780709384e-05, + "loss": 0.3871, + "step": 31562 + }, + { + "epoch": 0.7398864255700698, + "grad_norm": 3.3710601329803467, + "learning_rate": 3.158428320127157e-05, + "loss": 0.124, + "step": 31563 + }, + { + "epoch": 0.7399098671448748, + "grad_norm": 2.3270084857940674, + "learning_rate": 3.1578910993059406e-05, + "loss": 0.2002, + "step": 31564 + }, + { + "epoch": 0.7399333087196798, + "grad_norm": 2.5492727756500244, + "learning_rate": 3.1573539156101984e-05, + "loss": 0.2209, + "step": 31565 + }, + { + "epoch": 0.7399567502944848, + "grad_norm": 4.121927738189697, + "learning_rate": 3.156816769042843e-05, + "loss": 0.4271, + "step": 31566 + }, + { + "epoch": 0.7399801918692898, + "grad_norm": 1.2652727365493774, + "learning_rate": 3.1562796596067965e-05, + "loss": 0.1245, + "step": 31567 + }, + { + "epoch": 0.7400036334440948, + "grad_norm": 0.7966936826705933, + "learning_rate": 3.1557425873049685e-05, + "loss": 0.0889, + "step": 31568 + }, + { + "epoch": 0.7400270750188997, + "grad_norm": 3.9028780460357666, + "learning_rate": 3.1552055521402734e-05, + "loss": 0.4353, + "step": 31569 + }, + { + "epoch": 0.7400505165937048, + "grad_norm": 3.2641217708587646, + "learning_rate": 3.154668554115622e-05, + "loss": 0.3295, + "step": 31570 + }, + { + "epoch": 0.7400739581685097, + "grad_norm": 1.110826849937439, + "learning_rate": 3.1541315932339346e-05, + "loss": 0.0521, + "step": 31571 + }, + { + "epoch": 0.7400973997433148, + "grad_norm": 0.7103314399719238, + "learning_rate": 3.153594669498122e-05, + "loss": 0.0664, + "step": 31572 + }, + { + "epoch": 0.7401208413181197, + "grad_norm": 2.424504041671753, + "learning_rate": 3.153057782911097e-05, + "loss": 0.1033, + "step": 31573 + }, + { + "epoch": 0.7401442828929248, + "grad_norm": 2.916764259338379, + "learning_rate": 3.152520933475775e-05, + "loss": 0.2313, + "step": 31574 + }, + { + "epoch": 0.7401677244677297, + "grad_norm": 1.9077823162078857, + "learning_rate": 3.1519841211950644e-05, + "loss": 0.1685, + "step": 31575 + }, + { + "epoch": 0.7401911660425348, + "grad_norm": 2.375519037246704, + "learning_rate": 3.1514473460718775e-05, + "loss": 0.1142, + "step": 31576 + }, + { + "epoch": 0.7402146076173397, + "grad_norm": 2.7034378051757812, + "learning_rate": 3.1509106081091346e-05, + "loss": 0.2491, + "step": 31577 + }, + { + "epoch": 0.7402380491921448, + "grad_norm": 4.089900493621826, + "learning_rate": 3.150373907309738e-05, + "loss": 0.5267, + "step": 31578 + }, + { + "epoch": 0.7402614907669497, + "grad_norm": 0.6856374144554138, + "learning_rate": 3.14983724367661e-05, + "loss": 0.0382, + "step": 31579 + }, + { + "epoch": 0.7402849323417547, + "grad_norm": 3.8451592922210693, + "learning_rate": 3.149300617212657e-05, + "loss": 0.3738, + "step": 31580 + }, + { + "epoch": 0.7403083739165597, + "grad_norm": 2.911463737487793, + "learning_rate": 3.148764027920792e-05, + "loss": 0.1714, + "step": 31581 + }, + { + "epoch": 0.7403318154913647, + "grad_norm": 1.5436252355575562, + "learning_rate": 3.1482274758039265e-05, + "loss": 0.0945, + "step": 31582 + }, + { + "epoch": 0.7403552570661697, + "grad_norm": 2.1416144371032715, + "learning_rate": 3.1476909608649675e-05, + "loss": 0.1702, + "step": 31583 + }, + { + "epoch": 0.7403786986409747, + "grad_norm": 2.9284005165100098, + "learning_rate": 3.147154483106833e-05, + "loss": 0.2281, + "step": 31584 + }, + { + "epoch": 0.7404021402157797, + "grad_norm": 1.3719273805618286, + "learning_rate": 3.146618042532431e-05, + "loss": 0.119, + "step": 31585 + }, + { + "epoch": 0.7404255817905847, + "grad_norm": 2.4904539585113525, + "learning_rate": 3.146081639144672e-05, + "loss": 0.2401, + "step": 31586 + }, + { + "epoch": 0.7404490233653896, + "grad_norm": 3.5580039024353027, + "learning_rate": 3.145545272946467e-05, + "loss": 0.2823, + "step": 31587 + }, + { + "epoch": 0.7404724649401947, + "grad_norm": 3.0313663482666016, + "learning_rate": 3.145008943940724e-05, + "loss": 0.197, + "step": 31588 + }, + { + "epoch": 0.7404959065149996, + "grad_norm": 1.8535488843917847, + "learning_rate": 3.1444726521303566e-05, + "loss": 0.2209, + "step": 31589 + }, + { + "epoch": 0.7405193480898047, + "grad_norm": 2.5035347938537598, + "learning_rate": 3.143936397518271e-05, + "loss": 0.2972, + "step": 31590 + }, + { + "epoch": 0.7405427896646096, + "grad_norm": 2.661292552947998, + "learning_rate": 3.1434001801073834e-05, + "loss": 0.1248, + "step": 31591 + }, + { + "epoch": 0.7405662312394147, + "grad_norm": 4.630566596984863, + "learning_rate": 3.1428639999005985e-05, + "loss": 0.0949, + "step": 31592 + }, + { + "epoch": 0.7405896728142196, + "grad_norm": 3.7434587478637695, + "learning_rate": 3.1423278569008254e-05, + "loss": 0.4118, + "step": 31593 + }, + { + "epoch": 0.7406131143890247, + "grad_norm": 0.9681973457336426, + "learning_rate": 3.141791751110975e-05, + "loss": 0.0496, + "step": 31594 + }, + { + "epoch": 0.7406365559638296, + "grad_norm": 3.2311809062957764, + "learning_rate": 3.141255682533952e-05, + "loss": 0.3233, + "step": 31595 + }, + { + "epoch": 0.7406599975386347, + "grad_norm": 2.93192458152771, + "learning_rate": 3.140719651172671e-05, + "loss": 0.215, + "step": 31596 + }, + { + "epoch": 0.7406834391134396, + "grad_norm": 1.8333818912506104, + "learning_rate": 3.140183657030039e-05, + "loss": 0.146, + "step": 31597 + }, + { + "epoch": 0.7407068806882446, + "grad_norm": 2.6297292709350586, + "learning_rate": 3.139647700108963e-05, + "loss": 0.1043, + "step": 31598 + }, + { + "epoch": 0.7407303222630496, + "grad_norm": 1.9446618556976318, + "learning_rate": 3.1391117804123523e-05, + "loss": 0.1073, + "step": 31599 + }, + { + "epoch": 0.7407537638378546, + "grad_norm": 2.889561414718628, + "learning_rate": 3.1385758979431134e-05, + "loss": 1.0943, + "step": 31600 + }, + { + "epoch": 0.7407772054126596, + "grad_norm": 5.422562122344971, + "learning_rate": 3.1380400527041506e-05, + "loss": 0.4044, + "step": 31601 + }, + { + "epoch": 0.7408006469874646, + "grad_norm": 2.281919240951538, + "learning_rate": 3.137504244698377e-05, + "loss": 0.1135, + "step": 31602 + }, + { + "epoch": 0.7408240885622697, + "grad_norm": 2.3010382652282715, + "learning_rate": 3.136968473928701e-05, + "loss": 0.1422, + "step": 31603 + }, + { + "epoch": 0.7408475301370746, + "grad_norm": 1.4679242372512817, + "learning_rate": 3.136432740398026e-05, + "loss": 0.0346, + "step": 31604 + }, + { + "epoch": 0.7408709717118797, + "grad_norm": 3.0968778133392334, + "learning_rate": 3.1358970441092604e-05, + "loss": 0.3498, + "step": 31605 + }, + { + "epoch": 0.7408944132866846, + "grad_norm": 1.8468502759933472, + "learning_rate": 3.1353613850653105e-05, + "loss": 0.1686, + "step": 31606 + }, + { + "epoch": 0.7409178548614896, + "grad_norm": 2.0662803649902344, + "learning_rate": 3.1348257632690835e-05, + "loss": 0.2349, + "step": 31607 + }, + { + "epoch": 0.7409412964362946, + "grad_norm": 2.560673713684082, + "learning_rate": 3.134290178723481e-05, + "loss": 0.0806, + "step": 31608 + }, + { + "epoch": 0.7409647380110996, + "grad_norm": 2.174546241760254, + "learning_rate": 3.133754631431417e-05, + "loss": 0.2682, + "step": 31609 + }, + { + "epoch": 0.7409881795859046, + "grad_norm": 4.045844554901123, + "learning_rate": 3.133219121395793e-05, + "loss": 0.3735, + "step": 31610 + }, + { + "epoch": 0.7410116211607096, + "grad_norm": 1.4979784488677979, + "learning_rate": 3.132683648619514e-05, + "loss": 0.1067, + "step": 31611 + }, + { + "epoch": 0.7410350627355146, + "grad_norm": 3.594691514968872, + "learning_rate": 3.132148213105488e-05, + "loss": 0.4388, + "step": 31612 + }, + { + "epoch": 0.7410585043103196, + "grad_norm": 0.6272702217102051, + "learning_rate": 3.1316128148566184e-05, + "loss": 0.0445, + "step": 31613 + }, + { + "epoch": 0.7410819458851245, + "grad_norm": 4.187933444976807, + "learning_rate": 3.131077453875808e-05, + "loss": 0.3667, + "step": 31614 + }, + { + "epoch": 0.7411053874599296, + "grad_norm": 3.224961042404175, + "learning_rate": 3.130542130165964e-05, + "loss": 0.3243, + "step": 31615 + }, + { + "epoch": 0.7411288290347345, + "grad_norm": 0.9783452749252319, + "learning_rate": 3.1300068437299944e-05, + "loss": 0.0986, + "step": 31616 + }, + { + "epoch": 0.7411522706095396, + "grad_norm": 1.177824854850769, + "learning_rate": 3.129471594570802e-05, + "loss": 0.5298, + "step": 31617 + }, + { + "epoch": 0.7411757121843445, + "grad_norm": 1.128755807876587, + "learning_rate": 3.128936382691289e-05, + "loss": 0.1062, + "step": 31618 + }, + { + "epoch": 0.7411991537591496, + "grad_norm": 2.824897289276123, + "learning_rate": 3.12840120809436e-05, + "loss": 1.4588, + "step": 31619 + }, + { + "epoch": 0.7412225953339545, + "grad_norm": 0.6812125444412231, + "learning_rate": 3.12786607078292e-05, + "loss": 0.0336, + "step": 31620 + }, + { + "epoch": 0.7412460369087596, + "grad_norm": 2.8856523036956787, + "learning_rate": 3.127330970759869e-05, + "loss": 0.362, + "step": 31621 + }, + { + "epoch": 0.7412694784835645, + "grad_norm": 0.6347734928131104, + "learning_rate": 3.126795908028116e-05, + "loss": 0.047, + "step": 31622 + }, + { + "epoch": 0.7412929200583696, + "grad_norm": 2.8804404735565186, + "learning_rate": 3.1262608825905624e-05, + "loss": 0.2259, + "step": 31623 + }, + { + "epoch": 0.7413163616331745, + "grad_norm": 1.3900213241577148, + "learning_rate": 3.12572589445011e-05, + "loss": 0.0375, + "step": 31624 + }, + { + "epoch": 0.7413398032079795, + "grad_norm": 2.71317982673645, + "learning_rate": 3.1251909436096616e-05, + "loss": 0.2406, + "step": 31625 + }, + { + "epoch": 0.7413632447827845, + "grad_norm": 1.9804277420043945, + "learning_rate": 3.124656030072118e-05, + "loss": 0.1682, + "step": 31626 + }, + { + "epoch": 0.7413866863575895, + "grad_norm": 2.124380588531494, + "learning_rate": 3.124121153840387e-05, + "loss": 0.0964, + "step": 31627 + }, + { + "epoch": 0.7414101279323945, + "grad_norm": 2.2892720699310303, + "learning_rate": 3.123586314917365e-05, + "loss": 1.4477, + "step": 31628 + }, + { + "epoch": 0.7414335695071995, + "grad_norm": 1.0469013452529907, + "learning_rate": 3.12305151330596e-05, + "loss": 0.0513, + "step": 31629 + }, + { + "epoch": 0.7414570110820045, + "grad_norm": 0.930504322052002, + "learning_rate": 3.122516749009071e-05, + "loss": 0.0687, + "step": 31630 + }, + { + "epoch": 0.7414804526568095, + "grad_norm": 0.5195337533950806, + "learning_rate": 3.1219820220295994e-05, + "loss": 0.0421, + "step": 31631 + }, + { + "epoch": 0.7415038942316144, + "grad_norm": 3.0553362369537354, + "learning_rate": 3.121447332370447e-05, + "loss": 0.2826, + "step": 31632 + }, + { + "epoch": 0.7415273358064195, + "grad_norm": 1.9695881605148315, + "learning_rate": 3.120912680034511e-05, + "loss": 0.0586, + "step": 31633 + }, + { + "epoch": 0.7415507773812244, + "grad_norm": 3.0197153091430664, + "learning_rate": 3.120378065024699e-05, + "loss": 0.354, + "step": 31634 + }, + { + "epoch": 0.7415742189560295, + "grad_norm": 2.347123622894287, + "learning_rate": 3.119843487343911e-05, + "loss": 0.289, + "step": 31635 + }, + { + "epoch": 0.7415976605308344, + "grad_norm": 1.339377522468567, + "learning_rate": 3.119308946995044e-05, + "loss": 0.1391, + "step": 31636 + }, + { + "epoch": 0.7416211021056395, + "grad_norm": 2.4657514095306396, + "learning_rate": 3.1187744439810006e-05, + "loss": 0.1105, + "step": 31637 + }, + { + "epoch": 0.7416445436804444, + "grad_norm": 2.002107620239258, + "learning_rate": 3.118239978304681e-05, + "loss": 0.1479, + "step": 31638 + }, + { + "epoch": 0.7416679852552495, + "grad_norm": 5.363429069519043, + "learning_rate": 3.117705549968981e-05, + "loss": 0.5305, + "step": 31639 + }, + { + "epoch": 0.7416914268300544, + "grad_norm": 2.155658483505249, + "learning_rate": 3.117171158976805e-05, + "loss": 1.4422, + "step": 31640 + }, + { + "epoch": 0.7417148684048595, + "grad_norm": 0.7376279234886169, + "learning_rate": 3.116636805331053e-05, + "loss": 0.0473, + "step": 31641 + }, + { + "epoch": 0.7417383099796644, + "grad_norm": 3.546379327774048, + "learning_rate": 3.1161024890346256e-05, + "loss": 0.2436, + "step": 31642 + }, + { + "epoch": 0.7417617515544694, + "grad_norm": 3.0086958408355713, + "learning_rate": 3.1155682100904184e-05, + "loss": 0.1823, + "step": 31643 + }, + { + "epoch": 0.7417851931292744, + "grad_norm": 2.8800551891326904, + "learning_rate": 3.115033968501332e-05, + "loss": 0.2531, + "step": 31644 + }, + { + "epoch": 0.7418086347040794, + "grad_norm": 1.3918713331222534, + "learning_rate": 3.1144997642702647e-05, + "loss": 0.1247, + "step": 31645 + }, + { + "epoch": 0.7418320762788844, + "grad_norm": 2.4635705947875977, + "learning_rate": 3.1139655974001126e-05, + "loss": 0.1278, + "step": 31646 + }, + { + "epoch": 0.7418555178536894, + "grad_norm": 2.6152188777923584, + "learning_rate": 3.113431467893779e-05, + "loss": 0.3505, + "step": 31647 + }, + { + "epoch": 0.7418789594284944, + "grad_norm": 1.2915198802947998, + "learning_rate": 3.112897375754161e-05, + "loss": 0.0955, + "step": 31648 + }, + { + "epoch": 0.7419024010032994, + "grad_norm": 3.341172933578491, + "learning_rate": 3.112363320984155e-05, + "loss": 0.417, + "step": 31649 + }, + { + "epoch": 0.7419258425781043, + "grad_norm": 1.7400014400482178, + "learning_rate": 3.1118293035866584e-05, + "loss": 0.1149, + "step": 31650 + }, + { + "epoch": 0.7419492841529094, + "grad_norm": 3.3321235179901123, + "learning_rate": 3.111295323564568e-05, + "loss": 0.2826, + "step": 31651 + }, + { + "epoch": 0.7419727257277143, + "grad_norm": 3.396013021469116, + "learning_rate": 3.1107613809207846e-05, + "loss": 0.3356, + "step": 31652 + }, + { + "epoch": 0.7419961673025194, + "grad_norm": 2.452500581741333, + "learning_rate": 3.1102274756582017e-05, + "loss": 0.2391, + "step": 31653 + }, + { + "epoch": 0.7420196088773244, + "grad_norm": 0.8823307752609253, + "learning_rate": 3.1096936077797214e-05, + "loss": 0.0661, + "step": 31654 + }, + { + "epoch": 0.7420430504521294, + "grad_norm": 2.7538859844207764, + "learning_rate": 3.1091597772882366e-05, + "loss": 0.2159, + "step": 31655 + }, + { + "epoch": 0.7420664920269344, + "grad_norm": 1.7140038013458252, + "learning_rate": 3.108625984186645e-05, + "loss": 0.1723, + "step": 31656 + }, + { + "epoch": 0.7420899336017394, + "grad_norm": 3.2458934783935547, + "learning_rate": 3.108092228477842e-05, + "loss": 0.3329, + "step": 31657 + }, + { + "epoch": 0.7421133751765444, + "grad_norm": 1.4519697427749634, + "learning_rate": 3.107558510164725e-05, + "loss": 0.0997, + "step": 31658 + }, + { + "epoch": 0.7421368167513493, + "grad_norm": 3.3812127113342285, + "learning_rate": 3.107024829250186e-05, + "loss": 0.3445, + "step": 31659 + }, + { + "epoch": 0.7421602583261544, + "grad_norm": 1.212131142616272, + "learning_rate": 3.1064911857371274e-05, + "loss": 0.1222, + "step": 31660 + }, + { + "epoch": 0.7421836999009593, + "grad_norm": 2.579477548599243, + "learning_rate": 3.105957579628441e-05, + "loss": 0.3188, + "step": 31661 + }, + { + "epoch": 0.7422071414757644, + "grad_norm": 0.32306161522865295, + "learning_rate": 3.1054240109270226e-05, + "loss": 0.0206, + "step": 31662 + }, + { + "epoch": 0.7422305830505693, + "grad_norm": 1.6275410652160645, + "learning_rate": 3.104890479635765e-05, + "loss": 0.0916, + "step": 31663 + }, + { + "epoch": 0.7422540246253744, + "grad_norm": 1.790439486503601, + "learning_rate": 3.1043569857575686e-05, + "loss": 0.0825, + "step": 31664 + }, + { + "epoch": 0.7422774662001793, + "grad_norm": 0.668363630771637, + "learning_rate": 3.103823529295321e-05, + "loss": 0.0623, + "step": 31665 + }, + { + "epoch": 0.7423009077749844, + "grad_norm": 1.3621468544006348, + "learning_rate": 3.103290110251925e-05, + "loss": 0.0623, + "step": 31666 + }, + { + "epoch": 0.7423243493497893, + "grad_norm": 2.2239327430725098, + "learning_rate": 3.1027567286302703e-05, + "loss": 0.1593, + "step": 31667 + }, + { + "epoch": 0.7423477909245944, + "grad_norm": 2.5498454570770264, + "learning_rate": 3.102223384433252e-05, + "loss": 0.1241, + "step": 31668 + }, + { + "epoch": 0.7423712324993993, + "grad_norm": 1.8758600950241089, + "learning_rate": 3.101690077663762e-05, + "loss": 0.0866, + "step": 31669 + }, + { + "epoch": 0.7423946740742043, + "grad_norm": 3.145529270172119, + "learning_rate": 3.1011568083246966e-05, + "loss": 0.2157, + "step": 31670 + }, + { + "epoch": 0.7424181156490093, + "grad_norm": 2.2299273014068604, + "learning_rate": 3.100623576418946e-05, + "loss": 0.1995, + "step": 31671 + }, + { + "epoch": 0.7424415572238143, + "grad_norm": 1.247952938079834, + "learning_rate": 3.1000903819494075e-05, + "loss": 0.1412, + "step": 31672 + }, + { + "epoch": 0.7424649987986193, + "grad_norm": 4.103375434875488, + "learning_rate": 3.099557224918974e-05, + "loss": 0.2906, + "step": 31673 + }, + { + "epoch": 0.7424884403734243, + "grad_norm": 0.5646727085113525, + "learning_rate": 3.099024105330537e-05, + "loss": 0.0409, + "step": 31674 + }, + { + "epoch": 0.7425118819482293, + "grad_norm": 4.225828647613525, + "learning_rate": 3.0984910231869844e-05, + "loss": 0.35, + "step": 31675 + }, + { + "epoch": 0.7425353235230343, + "grad_norm": 4.69609260559082, + "learning_rate": 3.097957978491218e-05, + "loss": 0.3279, + "step": 31676 + }, + { + "epoch": 0.7425587650978392, + "grad_norm": 1.3779319524765015, + "learning_rate": 3.097424971246126e-05, + "loss": 0.0978, + "step": 31677 + }, + { + "epoch": 0.7425822066726443, + "grad_norm": 0.8363553881645203, + "learning_rate": 3.0968920014545964e-05, + "loss": 0.0682, + "step": 31678 + }, + { + "epoch": 0.7426056482474492, + "grad_norm": 2.813054084777832, + "learning_rate": 3.096359069119528e-05, + "loss": 0.1226, + "step": 31679 + }, + { + "epoch": 0.7426290898222543, + "grad_norm": 0.38840875029563904, + "learning_rate": 3.09582617424381e-05, + "loss": 0.0261, + "step": 31680 + }, + { + "epoch": 0.7426525313970592, + "grad_norm": 1.0997496843338013, + "learning_rate": 3.095293316830332e-05, + "loss": 0.0991, + "step": 31681 + }, + { + "epoch": 0.7426759729718643, + "grad_norm": 0.5315475463867188, + "learning_rate": 3.094760496881988e-05, + "loss": 0.0299, + "step": 31682 + }, + { + "epoch": 0.7426994145466692, + "grad_norm": 0.18850859999656677, + "learning_rate": 3.0942277144016676e-05, + "loss": 0.012, + "step": 31683 + }, + { + "epoch": 0.7427228561214743, + "grad_norm": 1.5220489501953125, + "learning_rate": 3.093694969392258e-05, + "loss": 0.07, + "step": 31684 + }, + { + "epoch": 0.7427462976962792, + "grad_norm": 4.449049949645996, + "learning_rate": 3.0931622618566566e-05, + "loss": 0.4195, + "step": 31685 + }, + { + "epoch": 0.7427697392710843, + "grad_norm": 0.5952180624008179, + "learning_rate": 3.092629591797752e-05, + "loss": 0.0522, + "step": 31686 + }, + { + "epoch": 0.7427931808458892, + "grad_norm": 2.0746445655822754, + "learning_rate": 3.09209695921843e-05, + "loss": 0.1078, + "step": 31687 + }, + { + "epoch": 0.7428166224206942, + "grad_norm": 2.052501916885376, + "learning_rate": 3.0915643641215864e-05, + "loss": 0.1251, + "step": 31688 + }, + { + "epoch": 0.7428400639954992, + "grad_norm": 1.9025815725326538, + "learning_rate": 3.091031806510109e-05, + "loss": 0.1101, + "step": 31689 + }, + { + "epoch": 0.7428635055703042, + "grad_norm": 1.8278639316558838, + "learning_rate": 3.090499286386889e-05, + "loss": 0.0494, + "step": 31690 + }, + { + "epoch": 0.7428869471451092, + "grad_norm": 6.140609264373779, + "learning_rate": 3.0899668037548094e-05, + "loss": 0.4177, + "step": 31691 + }, + { + "epoch": 0.7429103887199142, + "grad_norm": 2.9619853496551514, + "learning_rate": 3.0894343586167675e-05, + "loss": 0.2612, + "step": 31692 + }, + { + "epoch": 0.7429338302947192, + "grad_norm": 3.669396162033081, + "learning_rate": 3.088901950975649e-05, + "loss": 0.3027, + "step": 31693 + }, + { + "epoch": 0.7429572718695242, + "grad_norm": 3.7604851722717285, + "learning_rate": 3.088369580834344e-05, + "loss": 0.2494, + "step": 31694 + }, + { + "epoch": 0.7429807134443291, + "grad_norm": 1.3757524490356445, + "learning_rate": 3.0878372481957386e-05, + "loss": 0.0635, + "step": 31695 + }, + { + "epoch": 0.7430041550191342, + "grad_norm": 1.8060718774795532, + "learning_rate": 3.087304953062723e-05, + "loss": 1.129, + "step": 31696 + }, + { + "epoch": 0.7430275965939391, + "grad_norm": 3.511953830718994, + "learning_rate": 3.0867726954381824e-05, + "loss": 0.2447, + "step": 31697 + }, + { + "epoch": 0.7430510381687442, + "grad_norm": 1.6854252815246582, + "learning_rate": 3.08624047532501e-05, + "loss": 0.0808, + "step": 31698 + }, + { + "epoch": 0.7430744797435491, + "grad_norm": 1.8420367240905762, + "learning_rate": 3.085708292726088e-05, + "loss": 0.0972, + "step": 31699 + }, + { + "epoch": 0.7430979213183542, + "grad_norm": 4.203614711761475, + "learning_rate": 3.0851761476443105e-05, + "loss": 0.5903, + "step": 31700 + }, + { + "epoch": 0.7431213628931591, + "grad_norm": 0.5420281291007996, + "learning_rate": 3.0846440400825626e-05, + "loss": 0.0416, + "step": 31701 + }, + { + "epoch": 0.7431448044679642, + "grad_norm": 3.2127225399017334, + "learning_rate": 3.0841119700437305e-05, + "loss": 0.3151, + "step": 31702 + }, + { + "epoch": 0.7431682460427691, + "grad_norm": 2.511920690536499, + "learning_rate": 3.0835799375306974e-05, + "loss": 0.1271, + "step": 31703 + }, + { + "epoch": 0.7431916876175741, + "grad_norm": 3.352893829345703, + "learning_rate": 3.0830479425463563e-05, + "loss": 0.4753, + "step": 31704 + }, + { + "epoch": 0.7432151291923792, + "grad_norm": 4.269184589385986, + "learning_rate": 3.082515985093594e-05, + "loss": 0.3313, + "step": 31705 + }, + { + "epoch": 0.7432385707671841, + "grad_norm": 2.8682782649993896, + "learning_rate": 3.081984065175292e-05, + "loss": 0.3537, + "step": 31706 + }, + { + "epoch": 0.7432620123419892, + "grad_norm": 4.285766124725342, + "learning_rate": 3.08145218279434e-05, + "loss": 0.3652, + "step": 31707 + }, + { + "epoch": 0.7432854539167941, + "grad_norm": 1.599065899848938, + "learning_rate": 3.080920337953623e-05, + "loss": 0.7553, + "step": 31708 + }, + { + "epoch": 0.7433088954915992, + "grad_norm": 2.5340023040771484, + "learning_rate": 3.080388530656024e-05, + "loss": 0.1151, + "step": 31709 + }, + { + "epoch": 0.7433323370664041, + "grad_norm": 2.2736642360687256, + "learning_rate": 3.079856760904434e-05, + "loss": 0.0566, + "step": 31710 + }, + { + "epoch": 0.7433557786412092, + "grad_norm": 1.8281632661819458, + "learning_rate": 3.0793250287017364e-05, + "loss": 0.1123, + "step": 31711 + }, + { + "epoch": 0.7433792202160141, + "grad_norm": 0.46467655897140503, + "learning_rate": 3.0787933340508126e-05, + "loss": 0.0303, + "step": 31712 + }, + { + "epoch": 0.7434026617908192, + "grad_norm": 4.519301891326904, + "learning_rate": 3.0782616769545536e-05, + "loss": 0.3334, + "step": 31713 + }, + { + "epoch": 0.7434261033656241, + "grad_norm": 4.824827194213867, + "learning_rate": 3.0777300574158404e-05, + "loss": 0.5508, + "step": 31714 + }, + { + "epoch": 0.7434495449404291, + "grad_norm": 1.2890411615371704, + "learning_rate": 3.077198475437559e-05, + "loss": 0.1309, + "step": 31715 + }, + { + "epoch": 0.7434729865152341, + "grad_norm": 2.4092187881469727, + "learning_rate": 3.076666931022592e-05, + "loss": 0.996, + "step": 31716 + }, + { + "epoch": 0.7434964280900391, + "grad_norm": 0.902754545211792, + "learning_rate": 3.076135424173826e-05, + "loss": 0.0956, + "step": 31717 + }, + { + "epoch": 0.7435198696648441, + "grad_norm": 2.9532454013824463, + "learning_rate": 3.075603954894145e-05, + "loss": 1.5852, + "step": 31718 + }, + { + "epoch": 0.7435433112396491, + "grad_norm": 2.376180410385132, + "learning_rate": 3.075072523186431e-05, + "loss": 0.2331, + "step": 31719 + }, + { + "epoch": 0.7435667528144541, + "grad_norm": 1.85052490234375, + "learning_rate": 3.0745411290535684e-05, + "loss": 0.741, + "step": 31720 + }, + { + "epoch": 0.7435901943892591, + "grad_norm": 0.9858327507972717, + "learning_rate": 3.074009772498441e-05, + "loss": 0.0751, + "step": 31721 + }, + { + "epoch": 0.743613635964064, + "grad_norm": 3.009413003921509, + "learning_rate": 3.073478453523927e-05, + "loss": 0.1929, + "step": 31722 + }, + { + "epoch": 0.7436370775388691, + "grad_norm": 2.7176687717437744, + "learning_rate": 3.072947172132917e-05, + "loss": 0.2274, + "step": 31723 + }, + { + "epoch": 0.743660519113674, + "grad_norm": 0.9549853801727295, + "learning_rate": 3.0724159283282886e-05, + "loss": 0.1086, + "step": 31724 + }, + { + "epoch": 0.7436839606884791, + "grad_norm": 2.412743091583252, + "learning_rate": 3.071884722112928e-05, + "loss": 1.2834, + "step": 31725 + }, + { + "epoch": 0.743707402263284, + "grad_norm": 3.4907352924346924, + "learning_rate": 3.071353553489715e-05, + "loss": 0.3415, + "step": 31726 + }, + { + "epoch": 0.7437308438380891, + "grad_norm": 4.432168483734131, + "learning_rate": 3.070822422461535e-05, + "loss": 0.209, + "step": 31727 + }, + { + "epoch": 0.743754285412894, + "grad_norm": 2.778151273727417, + "learning_rate": 3.070291329031265e-05, + "loss": 0.2688, + "step": 31728 + }, + { + "epoch": 0.7437777269876991, + "grad_norm": 1.4486008882522583, + "learning_rate": 3.0697602732017875e-05, + "loss": 0.1038, + "step": 31729 + }, + { + "epoch": 0.743801168562504, + "grad_norm": 2.3705708980560303, + "learning_rate": 3.069229254975987e-05, + "loss": 0.0875, + "step": 31730 + }, + { + "epoch": 0.743824610137309, + "grad_norm": 2.1556074619293213, + "learning_rate": 3.068698274356745e-05, + "loss": 1.6494, + "step": 31731 + }, + { + "epoch": 0.743848051712114, + "grad_norm": 0.8743637800216675, + "learning_rate": 3.0681673313469405e-05, + "loss": 0.1361, + "step": 31732 + }, + { + "epoch": 0.743871493286919, + "grad_norm": 2.1821508407592773, + "learning_rate": 3.067636425949455e-05, + "loss": 1.5203, + "step": 31733 + }, + { + "epoch": 0.743894934861724, + "grad_norm": 1.2215359210968018, + "learning_rate": 3.067105558167166e-05, + "loss": 0.2024, + "step": 31734 + }, + { + "epoch": 0.743918376436529, + "grad_norm": 2.4367125034332275, + "learning_rate": 3.06657472800296e-05, + "loss": 0.3167, + "step": 31735 + }, + { + "epoch": 0.743941818011334, + "grad_norm": 0.5204566717147827, + "learning_rate": 3.066043935459711e-05, + "loss": 0.051, + "step": 31736 + }, + { + "epoch": 0.743965259586139, + "grad_norm": 4.196536064147949, + "learning_rate": 3.065513180540307e-05, + "loss": 0.2302, + "step": 31737 + }, + { + "epoch": 0.743988701160944, + "grad_norm": 2.9041507244110107, + "learning_rate": 3.0649824632476234e-05, + "loss": 1.3341, + "step": 31738 + }, + { + "epoch": 0.744012142735749, + "grad_norm": 0.35587278008461, + "learning_rate": 3.0644517835845384e-05, + "loss": 0.0333, + "step": 31739 + }, + { + "epoch": 0.7440355843105539, + "grad_norm": 0.6406548023223877, + "learning_rate": 3.063921141553935e-05, + "loss": 0.0437, + "step": 31740 + }, + { + "epoch": 0.744059025885359, + "grad_norm": 2.943286418914795, + "learning_rate": 3.0633905371586854e-05, + "loss": 0.2059, + "step": 31741 + }, + { + "epoch": 0.7440824674601639, + "grad_norm": 1.7152754068374634, + "learning_rate": 3.062859970401679e-05, + "loss": 0.1462, + "step": 31742 + }, + { + "epoch": 0.744105909034969, + "grad_norm": 2.006408929824829, + "learning_rate": 3.0623294412857884e-05, + "loss": 0.0943, + "step": 31743 + }, + { + "epoch": 0.7441293506097739, + "grad_norm": 3.0465712547302246, + "learning_rate": 3.061798949813893e-05, + "loss": 0.2321, + "step": 31744 + }, + { + "epoch": 0.744152792184579, + "grad_norm": 4.482028484344482, + "learning_rate": 3.061268495988871e-05, + "loss": 0.3574, + "step": 31745 + }, + { + "epoch": 0.7441762337593839, + "grad_norm": 1.7772213220596313, + "learning_rate": 3.060738079813603e-05, + "loss": 0.1314, + "step": 31746 + }, + { + "epoch": 0.744199675334189, + "grad_norm": 2.909560203552246, + "learning_rate": 3.0602077012909616e-05, + "loss": 0.2486, + "step": 31747 + }, + { + "epoch": 0.7442231169089939, + "grad_norm": 1.2326689958572388, + "learning_rate": 3.059677360423828e-05, + "loss": 0.0982, + "step": 31748 + }, + { + "epoch": 0.744246558483799, + "grad_norm": 3.0626516342163086, + "learning_rate": 3.059147057215084e-05, + "loss": 0.1903, + "step": 31749 + }, + { + "epoch": 0.7442700000586039, + "grad_norm": 2.4344322681427, + "learning_rate": 3.058616791667602e-05, + "loss": 1.687, + "step": 31750 + }, + { + "epoch": 0.7442934416334089, + "grad_norm": 3.7066218852996826, + "learning_rate": 3.0580865637842604e-05, + "loss": 0.2547, + "step": 31751 + }, + { + "epoch": 0.7443168832082139, + "grad_norm": 3.370015859603882, + "learning_rate": 3.057556373567938e-05, + "loss": 1.7808, + "step": 31752 + }, + { + "epoch": 0.7443403247830189, + "grad_norm": 3.302213668823242, + "learning_rate": 3.057026221021508e-05, + "loss": 0.2462, + "step": 31753 + }, + { + "epoch": 0.7443637663578239, + "grad_norm": 0.5300624966621399, + "learning_rate": 3.056496106147847e-05, + "loss": 0.0434, + "step": 31754 + }, + { + "epoch": 0.7443872079326289, + "grad_norm": 3.5707571506500244, + "learning_rate": 3.055966028949836e-05, + "loss": 0.2732, + "step": 31755 + }, + { + "epoch": 0.744410649507434, + "grad_norm": 0.9639098048210144, + "learning_rate": 3.0554359894303486e-05, + "loss": 0.0272, + "step": 31756 + }, + { + "epoch": 0.7444340910822389, + "grad_norm": 2.333599805831909, + "learning_rate": 3.0549059875922615e-05, + "loss": 0.1556, + "step": 31757 + }, + { + "epoch": 0.744457532657044, + "grad_norm": 0.9509584903717041, + "learning_rate": 3.054376023438449e-05, + "loss": 0.0423, + "step": 31758 + }, + { + "epoch": 0.7444809742318489, + "grad_norm": 4.080393314361572, + "learning_rate": 3.053846096971787e-05, + "loss": 0.3176, + "step": 31759 + }, + { + "epoch": 0.7445044158066539, + "grad_norm": 4.022922039031982, + "learning_rate": 3.053316208195149e-05, + "loss": 0.3797, + "step": 31760 + }, + { + "epoch": 0.7445278573814589, + "grad_norm": 3.554861545562744, + "learning_rate": 3.0527863571114135e-05, + "loss": 0.3171, + "step": 31761 + }, + { + "epoch": 0.7445512989562639, + "grad_norm": 2.5834929943084717, + "learning_rate": 3.052256543723456e-05, + "loss": 1.0728, + "step": 31762 + }, + { + "epoch": 0.7445747405310689, + "grad_norm": 2.9887094497680664, + "learning_rate": 3.051726768034151e-05, + "loss": 0.158, + "step": 31763 + }, + { + "epoch": 0.7445981821058739, + "grad_norm": 2.0948691368103027, + "learning_rate": 3.051197030046372e-05, + "loss": 0.2035, + "step": 31764 + }, + { + "epoch": 0.7446216236806789, + "grad_norm": 2.041996717453003, + "learning_rate": 3.050667329762993e-05, + "loss": 0.1959, + "step": 31765 + }, + { + "epoch": 0.7446450652554839, + "grad_norm": 4.265064239501953, + "learning_rate": 3.0501376671868875e-05, + "loss": 0.4866, + "step": 31766 + }, + { + "epoch": 0.7446685068302888, + "grad_norm": 2.383136749267578, + "learning_rate": 3.0496080423209285e-05, + "loss": 0.1316, + "step": 31767 + }, + { + "epoch": 0.7446919484050939, + "grad_norm": 3.394702434539795, + "learning_rate": 3.049078455167994e-05, + "loss": 0.3006, + "step": 31768 + }, + { + "epoch": 0.7447153899798988, + "grad_norm": 4.8277363777160645, + "learning_rate": 3.0485489057309556e-05, + "loss": 0.277, + "step": 31769 + }, + { + "epoch": 0.7447388315547039, + "grad_norm": 2.8170628547668457, + "learning_rate": 3.0480193940126867e-05, + "loss": 0.1483, + "step": 31770 + }, + { + "epoch": 0.7447622731295088, + "grad_norm": 3.475243330001831, + "learning_rate": 3.04748992001606e-05, + "loss": 0.254, + "step": 31771 + }, + { + "epoch": 0.7447857147043139, + "grad_norm": 3.133598566055298, + "learning_rate": 3.0469604837439447e-05, + "loss": 0.2188, + "step": 31772 + }, + { + "epoch": 0.7448091562791188, + "grad_norm": 1.3418467044830322, + "learning_rate": 3.0464310851992206e-05, + "loss": 0.136, + "step": 31773 + }, + { + "epoch": 0.7448325978539239, + "grad_norm": 1.3486500978469849, + "learning_rate": 3.0459017243847542e-05, + "loss": 0.1202, + "step": 31774 + }, + { + "epoch": 0.7448560394287288, + "grad_norm": 2.687281370162964, + "learning_rate": 3.0453724013034236e-05, + "loss": 0.1917, + "step": 31775 + }, + { + "epoch": 0.7448794810035338, + "grad_norm": 2.3701577186584473, + "learning_rate": 3.044843115958098e-05, + "loss": 0.1977, + "step": 31776 + }, + { + "epoch": 0.7449029225783388, + "grad_norm": 1.1409348249435425, + "learning_rate": 3.0443138683516493e-05, + "loss": 0.1539, + "step": 31777 + }, + { + "epoch": 0.7449263641531438, + "grad_norm": 3.9826395511627197, + "learning_rate": 3.0437846584869478e-05, + "loss": 0.3841, + "step": 31778 + }, + { + "epoch": 0.7449498057279488, + "grad_norm": 0.3490578830242157, + "learning_rate": 3.043255486366865e-05, + "loss": 0.021, + "step": 31779 + }, + { + "epoch": 0.7449732473027538, + "grad_norm": 2.183593988418579, + "learning_rate": 3.042726351994276e-05, + "loss": 0.2022, + "step": 31780 + }, + { + "epoch": 0.7449966888775588, + "grad_norm": 2.5943000316619873, + "learning_rate": 3.042197255372049e-05, + "loss": 0.2023, + "step": 31781 + }, + { + "epoch": 0.7450201304523638, + "grad_norm": 2.020245313644409, + "learning_rate": 3.0416681965030557e-05, + "loss": 0.1058, + "step": 31782 + }, + { + "epoch": 0.7450435720271688, + "grad_norm": 0.8228121399879456, + "learning_rate": 3.0411391753901665e-05, + "loss": 0.0353, + "step": 31783 + }, + { + "epoch": 0.7450670136019738, + "grad_norm": 3.6025309562683105, + "learning_rate": 3.0406101920362483e-05, + "loss": 0.3558, + "step": 31784 + }, + { + "epoch": 0.7450904551767787, + "grad_norm": 1.583024501800537, + "learning_rate": 3.0400812464441787e-05, + "loss": 0.0541, + "step": 31785 + }, + { + "epoch": 0.7451138967515838, + "grad_norm": 1.9368391036987305, + "learning_rate": 3.03955233861682e-05, + "loss": 0.935, + "step": 31786 + }, + { + "epoch": 0.7451373383263887, + "grad_norm": 1.8397926092147827, + "learning_rate": 3.0390234685570496e-05, + "loss": 0.0679, + "step": 31787 + }, + { + "epoch": 0.7451607799011938, + "grad_norm": 0.6520081758499146, + "learning_rate": 3.0384946362677336e-05, + "loss": 0.0471, + "step": 31788 + }, + { + "epoch": 0.7451842214759987, + "grad_norm": 2.918226718902588, + "learning_rate": 3.037965841751741e-05, + "loss": 0.1977, + "step": 31789 + }, + { + "epoch": 0.7452076630508038, + "grad_norm": 3.049903154373169, + "learning_rate": 3.0374370850119427e-05, + "loss": 0.1259, + "step": 31790 + }, + { + "epoch": 0.7452311046256087, + "grad_norm": 1.3147400617599487, + "learning_rate": 3.0369083660512054e-05, + "loss": 0.8214, + "step": 31791 + }, + { + "epoch": 0.7452545462004138, + "grad_norm": 2.264070510864258, + "learning_rate": 3.036379684872396e-05, + "loss": 0.1823, + "step": 31792 + }, + { + "epoch": 0.7452779877752187, + "grad_norm": 3.2986836433410645, + "learning_rate": 3.0358510414783904e-05, + "loss": 0.2453, + "step": 31793 + }, + { + "epoch": 0.7453014293500237, + "grad_norm": 2.114142417907715, + "learning_rate": 3.035322435872051e-05, + "loss": 0.159, + "step": 31794 + }, + { + "epoch": 0.7453248709248287, + "grad_norm": 1.1454702615737915, + "learning_rate": 3.0347938680562494e-05, + "loss": 0.1192, + "step": 31795 + }, + { + "epoch": 0.7453483124996337, + "grad_norm": 1.9220061302185059, + "learning_rate": 3.0342653380338513e-05, + "loss": 1.3446, + "step": 31796 + }, + { + "epoch": 0.7453717540744387, + "grad_norm": 4.604255199432373, + "learning_rate": 3.033736845807723e-05, + "loss": 0.2285, + "step": 31797 + }, + { + "epoch": 0.7453951956492437, + "grad_norm": 6.878903865814209, + "learning_rate": 3.0332083913807374e-05, + "loss": 0.4848, + "step": 31798 + }, + { + "epoch": 0.7454186372240487, + "grad_norm": 2.673532009124756, + "learning_rate": 3.0326799747557553e-05, + "loss": 0.2783, + "step": 31799 + }, + { + "epoch": 0.7454420787988537, + "grad_norm": 2.624492883682251, + "learning_rate": 3.0321515959356518e-05, + "loss": 0.1859, + "step": 31800 + }, + { + "epoch": 0.7454655203736587, + "grad_norm": 4.077646255493164, + "learning_rate": 3.031623254923288e-05, + "loss": 0.3551, + "step": 31801 + }, + { + "epoch": 0.7454889619484637, + "grad_norm": 3.365251302719116, + "learning_rate": 3.031094951721534e-05, + "loss": 0.3191, + "step": 31802 + }, + { + "epoch": 0.7455124035232686, + "grad_norm": 0.6603769659996033, + "learning_rate": 3.030566686333254e-05, + "loss": 0.0912, + "step": 31803 + }, + { + "epoch": 0.7455358450980737, + "grad_norm": 3.2883522510528564, + "learning_rate": 3.030038458761315e-05, + "loss": 0.2815, + "step": 31804 + }, + { + "epoch": 0.7455592866728786, + "grad_norm": 0.695713460445404, + "learning_rate": 3.02951026900858e-05, + "loss": 0.0293, + "step": 31805 + }, + { + "epoch": 0.7455827282476837, + "grad_norm": 1.8984565734863281, + "learning_rate": 3.028982117077922e-05, + "loss": 0.1166, + "step": 31806 + }, + { + "epoch": 0.7456061698224887, + "grad_norm": 3.1762359142303467, + "learning_rate": 3.028454002972202e-05, + "loss": 0.148, + "step": 31807 + }, + { + "epoch": 0.7456296113972937, + "grad_norm": 2.186915874481201, + "learning_rate": 3.027925926694288e-05, + "loss": 0.9824, + "step": 31808 + }, + { + "epoch": 0.7456530529720987, + "grad_norm": 3.2267799377441406, + "learning_rate": 3.02739788824704e-05, + "loss": 0.2734, + "step": 31809 + }, + { + "epoch": 0.7456764945469037, + "grad_norm": 2.9486472606658936, + "learning_rate": 3.0268698876333303e-05, + "loss": 0.2171, + "step": 31810 + }, + { + "epoch": 0.7456999361217087, + "grad_norm": 1.6837782859802246, + "learning_rate": 3.026341924856021e-05, + "loss": 0.1069, + "step": 31811 + }, + { + "epoch": 0.7457233776965136, + "grad_norm": 2.9050180912017822, + "learning_rate": 3.0258139999179723e-05, + "loss": 0.2819, + "step": 31812 + }, + { + "epoch": 0.7457468192713187, + "grad_norm": 2.368922710418701, + "learning_rate": 3.0252861128220578e-05, + "loss": 0.158, + "step": 31813 + }, + { + "epoch": 0.7457702608461236, + "grad_norm": 2.8698596954345703, + "learning_rate": 3.0247582635711358e-05, + "loss": 0.2446, + "step": 31814 + }, + { + "epoch": 0.7457937024209287, + "grad_norm": 1.4126033782958984, + "learning_rate": 3.024230452168072e-05, + "loss": 0.1643, + "step": 31815 + }, + { + "epoch": 0.7458171439957336, + "grad_norm": 2.946488380432129, + "learning_rate": 3.0237026786157296e-05, + "loss": 0.3299, + "step": 31816 + }, + { + "epoch": 0.7458405855705387, + "grad_norm": 3.4530088901519775, + "learning_rate": 3.0231749429169698e-05, + "loss": 0.1819, + "step": 31817 + }, + { + "epoch": 0.7458640271453436, + "grad_norm": 1.7571240663528442, + "learning_rate": 3.0226472450746614e-05, + "loss": 0.1598, + "step": 31818 + }, + { + "epoch": 0.7458874687201487, + "grad_norm": 2.361093282699585, + "learning_rate": 3.0221195850916662e-05, + "loss": 0.228, + "step": 31819 + }, + { + "epoch": 0.7459109102949536, + "grad_norm": 1.862143635749817, + "learning_rate": 3.0215919629708456e-05, + "loss": 0.1114, + "step": 31820 + }, + { + "epoch": 0.7459343518697586, + "grad_norm": 2.4710335731506348, + "learning_rate": 3.0210643787150604e-05, + "loss": 0.1866, + "step": 31821 + }, + { + "epoch": 0.7459577934445636, + "grad_norm": 1.0974950790405273, + "learning_rate": 3.0205368323271788e-05, + "loss": 0.1224, + "step": 31822 + }, + { + "epoch": 0.7459812350193686, + "grad_norm": 0.7397799491882324, + "learning_rate": 3.0200093238100612e-05, + "loss": 0.054, + "step": 31823 + }, + { + "epoch": 0.7460046765941736, + "grad_norm": 0.17658069729804993, + "learning_rate": 3.0194818531665657e-05, + "loss": 0.0079, + "step": 31824 + }, + { + "epoch": 0.7460281181689786, + "grad_norm": 0.7454248666763306, + "learning_rate": 3.0189544203995613e-05, + "loss": 0.0772, + "step": 31825 + }, + { + "epoch": 0.7460515597437836, + "grad_norm": 2.67211651802063, + "learning_rate": 3.0184270255119063e-05, + "loss": 0.2698, + "step": 31826 + }, + { + "epoch": 0.7460750013185886, + "grad_norm": 0.40775036811828613, + "learning_rate": 3.017899668506463e-05, + "loss": 0.0447, + "step": 31827 + }, + { + "epoch": 0.7460984428933936, + "grad_norm": 2.9266483783721924, + "learning_rate": 3.017372349386093e-05, + "loss": 0.3676, + "step": 31828 + }, + { + "epoch": 0.7461218844681986, + "grad_norm": 3.8661603927612305, + "learning_rate": 3.0168450681536554e-05, + "loss": 0.3199, + "step": 31829 + }, + { + "epoch": 0.7461453260430035, + "grad_norm": 1.8284879922866821, + "learning_rate": 3.0163178248120107e-05, + "loss": 0.1621, + "step": 31830 + }, + { + "epoch": 0.7461687676178086, + "grad_norm": 3.2297303676605225, + "learning_rate": 3.015790619364024e-05, + "loss": 0.2784, + "step": 31831 + }, + { + "epoch": 0.7461922091926135, + "grad_norm": 1.5530807971954346, + "learning_rate": 3.0152634518125544e-05, + "loss": 0.1966, + "step": 31832 + }, + { + "epoch": 0.7462156507674186, + "grad_norm": 2.0991058349609375, + "learning_rate": 3.0147363221604596e-05, + "loss": 0.2261, + "step": 31833 + }, + { + "epoch": 0.7462390923422235, + "grad_norm": 2.0429606437683105, + "learning_rate": 3.014209230410604e-05, + "loss": 0.1131, + "step": 31834 + }, + { + "epoch": 0.7462625339170286, + "grad_norm": 1.923768162727356, + "learning_rate": 3.013682176565845e-05, + "loss": 0.1188, + "step": 31835 + }, + { + "epoch": 0.7462859754918335, + "grad_norm": 3.380014181137085, + "learning_rate": 3.0131551606290443e-05, + "loss": 1.7403, + "step": 31836 + }, + { + "epoch": 0.7463094170666386, + "grad_norm": 1.1932262182235718, + "learning_rate": 3.0126281826030557e-05, + "loss": 0.0672, + "step": 31837 + }, + { + "epoch": 0.7463328586414435, + "grad_norm": 0.35137319564819336, + "learning_rate": 3.012101242490747e-05, + "loss": 0.0147, + "step": 31838 + }, + { + "epoch": 0.7463563002162485, + "grad_norm": 2.689112424850464, + "learning_rate": 3.0115743402949736e-05, + "loss": 0.3339, + "step": 31839 + }, + { + "epoch": 0.7463797417910535, + "grad_norm": 1.7525328397750854, + "learning_rate": 3.0110474760185935e-05, + "loss": 0.0799, + "step": 31840 + }, + { + "epoch": 0.7464031833658585, + "grad_norm": 0.8353830575942993, + "learning_rate": 3.0105206496644667e-05, + "loss": 0.0837, + "step": 31841 + }, + { + "epoch": 0.7464266249406635, + "grad_norm": 2.0047271251678467, + "learning_rate": 3.0099938612354517e-05, + "loss": 0.1288, + "step": 31842 + }, + { + "epoch": 0.7464500665154685, + "grad_norm": 1.4199806451797485, + "learning_rate": 3.009467110734403e-05, + "loss": 0.1064, + "step": 31843 + }, + { + "epoch": 0.7464735080902735, + "grad_norm": 1.807195782661438, + "learning_rate": 3.0089403981641863e-05, + "loss": 0.1691, + "step": 31844 + }, + { + "epoch": 0.7464969496650785, + "grad_norm": 1.779316782951355, + "learning_rate": 3.008413723527651e-05, + "loss": 0.0566, + "step": 31845 + }, + { + "epoch": 0.7465203912398835, + "grad_norm": 3.1583073139190674, + "learning_rate": 3.007887086827663e-05, + "loss": 0.209, + "step": 31846 + }, + { + "epoch": 0.7465438328146885, + "grad_norm": 3.7469587326049805, + "learning_rate": 3.007360488067077e-05, + "loss": 0.2142, + "step": 31847 + }, + { + "epoch": 0.7465672743894934, + "grad_norm": 2.590240955352783, + "learning_rate": 3.006833927248749e-05, + "loss": 0.2144, + "step": 31848 + }, + { + "epoch": 0.7465907159642985, + "grad_norm": 3.3717997074127197, + "learning_rate": 3.006307404375537e-05, + "loss": 0.1994, + "step": 31849 + }, + { + "epoch": 0.7466141575391034, + "grad_norm": 3.2469565868377686, + "learning_rate": 3.005780919450295e-05, + "loss": 0.1472, + "step": 31850 + }, + { + "epoch": 0.7466375991139085, + "grad_norm": 1.143771767616272, + "learning_rate": 3.0052544724758847e-05, + "loss": 0.064, + "step": 31851 + }, + { + "epoch": 0.7466610406887134, + "grad_norm": 2.984895944595337, + "learning_rate": 3.00472806345516e-05, + "loss": 1.4216, + "step": 31852 + }, + { + "epoch": 0.7466844822635185, + "grad_norm": 1.9429064989089966, + "learning_rate": 3.0042016923909788e-05, + "loss": 0.7327, + "step": 31853 + }, + { + "epoch": 0.7467079238383234, + "grad_norm": 2.659616708755493, + "learning_rate": 3.0036753592861956e-05, + "loss": 0.2439, + "step": 31854 + }, + { + "epoch": 0.7467313654131285, + "grad_norm": 4.526101589202881, + "learning_rate": 3.0031490641436622e-05, + "loss": 0.3427, + "step": 31855 + }, + { + "epoch": 0.7467548069879334, + "grad_norm": 3.299800395965576, + "learning_rate": 3.0026228069662433e-05, + "loss": 0.2636, + "step": 31856 + }, + { + "epoch": 0.7467782485627384, + "grad_norm": 1.5739805698394775, + "learning_rate": 3.0020965877567852e-05, + "loss": 0.0987, + "step": 31857 + }, + { + "epoch": 0.7468016901375435, + "grad_norm": 3.3737776279449463, + "learning_rate": 3.0015704065181525e-05, + "loss": 0.2435, + "step": 31858 + }, + { + "epoch": 0.7468251317123484, + "grad_norm": 3.241607189178467, + "learning_rate": 3.0010442632531945e-05, + "loss": 0.3557, + "step": 31859 + }, + { + "epoch": 0.7468485732871535, + "grad_norm": 1.2116822004318237, + "learning_rate": 3.000518157964768e-05, + "loss": 0.1254, + "step": 31860 + }, + { + "epoch": 0.7468720148619584, + "grad_norm": 4.382615566253662, + "learning_rate": 2.9999920906557265e-05, + "loss": 0.4869, + "step": 31861 + }, + { + "epoch": 0.7468954564367635, + "grad_norm": 10.20581340789795, + "learning_rate": 2.9994660613289204e-05, + "loss": 0.3721, + "step": 31862 + }, + { + "epoch": 0.7469188980115684, + "grad_norm": 2.767930269241333, + "learning_rate": 2.9989400699872128e-05, + "loss": 0.1924, + "step": 31863 + }, + { + "epoch": 0.7469423395863735, + "grad_norm": 3.700923442840576, + "learning_rate": 2.9984141166334535e-05, + "loss": 0.4362, + "step": 31864 + }, + { + "epoch": 0.7469657811611784, + "grad_norm": 1.9434863328933716, + "learning_rate": 2.9978882012704945e-05, + "loss": 0.0853, + "step": 31865 + }, + { + "epoch": 0.7469892227359834, + "grad_norm": 2.3828847408294678, + "learning_rate": 2.997362323901193e-05, + "loss": 0.1863, + "step": 31866 + }, + { + "epoch": 0.7470126643107884, + "grad_norm": 2.8113269805908203, + "learning_rate": 2.996836484528399e-05, + "loss": 0.1753, + "step": 31867 + }, + { + "epoch": 0.7470361058855934, + "grad_norm": 3.725710868835449, + "learning_rate": 2.9963106831549637e-05, + "loss": 0.341, + "step": 31868 + }, + { + "epoch": 0.7470595474603984, + "grad_norm": 1.919737458229065, + "learning_rate": 2.9957849197837474e-05, + "loss": 0.0921, + "step": 31869 + }, + { + "epoch": 0.7470829890352034, + "grad_norm": 2.7550485134124756, + "learning_rate": 2.9952591944175968e-05, + "loss": 0.3663, + "step": 31870 + }, + { + "epoch": 0.7471064306100084, + "grad_norm": 0.7934921979904175, + "learning_rate": 2.9947335070593684e-05, + "loss": 0.0283, + "step": 31871 + }, + { + "epoch": 0.7471298721848134, + "grad_norm": 0.3038308620452881, + "learning_rate": 2.9942078577119138e-05, + "loss": 0.0182, + "step": 31872 + }, + { + "epoch": 0.7471533137596184, + "grad_norm": 2.5195960998535156, + "learning_rate": 2.9936822463780843e-05, + "loss": 0.3137, + "step": 31873 + }, + { + "epoch": 0.7471767553344234, + "grad_norm": 0.9839394688606262, + "learning_rate": 2.993156673060732e-05, + "loss": 0.0997, + "step": 31874 + }, + { + "epoch": 0.7472001969092283, + "grad_norm": 0.865824282169342, + "learning_rate": 2.9926311377627046e-05, + "loss": 0.0602, + "step": 31875 + }, + { + "epoch": 0.7472236384840334, + "grad_norm": 2.00986385345459, + "learning_rate": 2.992105640486862e-05, + "loss": 0.1201, + "step": 31876 + }, + { + "epoch": 0.7472470800588383, + "grad_norm": 3.2425992488861084, + "learning_rate": 2.99158018123605e-05, + "loss": 1.6218, + "step": 31877 + }, + { + "epoch": 0.7472705216336434, + "grad_norm": 2.310826301574707, + "learning_rate": 2.991054760013122e-05, + "loss": 0.9156, + "step": 31878 + }, + { + "epoch": 0.7472939632084483, + "grad_norm": 1.1420834064483643, + "learning_rate": 2.9905293768209275e-05, + "loss": 0.1474, + "step": 31879 + }, + { + "epoch": 0.7473174047832534, + "grad_norm": 2.0134170055389404, + "learning_rate": 2.990004031662318e-05, + "loss": 0.1604, + "step": 31880 + }, + { + "epoch": 0.7473408463580583, + "grad_norm": 4.778905868530273, + "learning_rate": 2.98947872454014e-05, + "loss": 0.1622, + "step": 31881 + }, + { + "epoch": 0.7473642879328634, + "grad_norm": 0.8667196035385132, + "learning_rate": 2.9889534554572473e-05, + "loss": 0.0474, + "step": 31882 + }, + { + "epoch": 0.7473877295076683, + "grad_norm": 3.0230162143707275, + "learning_rate": 2.988428224416495e-05, + "loss": 0.294, + "step": 31883 + }, + { + "epoch": 0.7474111710824733, + "grad_norm": 0.4470337927341461, + "learning_rate": 2.9879030314207267e-05, + "loss": 0.0398, + "step": 31884 + }, + { + "epoch": 0.7474346126572783, + "grad_norm": 2.786621332168579, + "learning_rate": 2.9873778764727943e-05, + "loss": 0.1336, + "step": 31885 + }, + { + "epoch": 0.7474580542320833, + "grad_norm": 2.9649507999420166, + "learning_rate": 2.9868527595755468e-05, + "loss": 0.3975, + "step": 31886 + }, + { + "epoch": 0.7474814958068883, + "grad_norm": 2.24090313911438, + "learning_rate": 2.986327680731834e-05, + "loss": 0.0874, + "step": 31887 + }, + { + "epoch": 0.7475049373816933, + "grad_norm": 2.8295345306396484, + "learning_rate": 2.9858026399445006e-05, + "loss": 0.3159, + "step": 31888 + }, + { + "epoch": 0.7475283789564983, + "grad_norm": 0.6104263663291931, + "learning_rate": 2.9852776372164025e-05, + "loss": 0.0537, + "step": 31889 + }, + { + "epoch": 0.7475518205313033, + "grad_norm": 1.3240827322006226, + "learning_rate": 2.9847526725503848e-05, + "loss": 0.0565, + "step": 31890 + }, + { + "epoch": 0.7475752621061083, + "grad_norm": 2.0333409309387207, + "learning_rate": 2.9842277459492973e-05, + "loss": 0.2012, + "step": 31891 + }, + { + "epoch": 0.7475987036809133, + "grad_norm": 2.7792649269104004, + "learning_rate": 2.9837028574159863e-05, + "loss": 0.2944, + "step": 31892 + }, + { + "epoch": 0.7476221452557182, + "grad_norm": 2.468371868133545, + "learning_rate": 2.983178006953299e-05, + "loss": 0.098, + "step": 31893 + }, + { + "epoch": 0.7476455868305233, + "grad_norm": 1.0614044666290283, + "learning_rate": 2.982653194564087e-05, + "loss": 0.1028, + "step": 31894 + }, + { + "epoch": 0.7476690284053282, + "grad_norm": 3.581357002258301, + "learning_rate": 2.9821284202511935e-05, + "loss": 0.3316, + "step": 31895 + }, + { + "epoch": 0.7476924699801333, + "grad_norm": 1.1726493835449219, + "learning_rate": 2.981603684017471e-05, + "loss": 0.1033, + "step": 31896 + }, + { + "epoch": 0.7477159115549382, + "grad_norm": 3.2491183280944824, + "learning_rate": 2.981078985865765e-05, + "loss": 0.2748, + "step": 31897 + }, + { + "epoch": 0.7477393531297433, + "grad_norm": 0.7382663488388062, + "learning_rate": 2.9805543257989223e-05, + "loss": 0.0516, + "step": 31898 + }, + { + "epoch": 0.7477627947045482, + "grad_norm": 0.8774160146713257, + "learning_rate": 2.9800297038197877e-05, + "loss": 0.0843, + "step": 31899 + }, + { + "epoch": 0.7477862362793533, + "grad_norm": 2.896461248397827, + "learning_rate": 2.9795051199312075e-05, + "loss": 0.3938, + "step": 31900 + }, + { + "epoch": 0.7478096778541582, + "grad_norm": 1.3253685235977173, + "learning_rate": 2.9789805741360323e-05, + "loss": 0.1225, + "step": 31901 + }, + { + "epoch": 0.7478331194289632, + "grad_norm": 1.5565773248672485, + "learning_rate": 2.9784560664371065e-05, + "loss": 1.2621, + "step": 31902 + }, + { + "epoch": 0.7478565610037682, + "grad_norm": 4.192167282104492, + "learning_rate": 2.9779315968372744e-05, + "loss": 0.2028, + "step": 31903 + }, + { + "epoch": 0.7478800025785732, + "grad_norm": 2.964630365371704, + "learning_rate": 2.9774071653393844e-05, + "loss": 0.1904, + "step": 31904 + }, + { + "epoch": 0.7479034441533782, + "grad_norm": 4.8144402503967285, + "learning_rate": 2.9768827719462798e-05, + "loss": 0.2797, + "step": 31905 + }, + { + "epoch": 0.7479268857281832, + "grad_norm": 10.20104694366455, + "learning_rate": 2.9763584166608037e-05, + "loss": 0.544, + "step": 31906 + }, + { + "epoch": 0.7479503273029882, + "grad_norm": 2.288264751434326, + "learning_rate": 2.9758340994858037e-05, + "loss": 1.1621, + "step": 31907 + }, + { + "epoch": 0.7479737688777932, + "grad_norm": 0.833680272102356, + "learning_rate": 2.9753098204241293e-05, + "loss": 0.1541, + "step": 31908 + }, + { + "epoch": 0.7479972104525983, + "grad_norm": 0.577433168888092, + "learning_rate": 2.9747855794786207e-05, + "loss": 0.0222, + "step": 31909 + }, + { + "epoch": 0.7480206520274032, + "grad_norm": 3.1180355548858643, + "learning_rate": 2.9742613766521244e-05, + "loss": 0.3051, + "step": 31910 + }, + { + "epoch": 0.7480440936022082, + "grad_norm": 1.5779130458831787, + "learning_rate": 2.9737372119474815e-05, + "loss": 0.0411, + "step": 31911 + }, + { + "epoch": 0.7480675351770132, + "grad_norm": 2.341588020324707, + "learning_rate": 2.9732130853675398e-05, + "loss": 1.0625, + "step": 31912 + }, + { + "epoch": 0.7480909767518182, + "grad_norm": 2.6846020221710205, + "learning_rate": 2.9726889969151373e-05, + "loss": 0.1192, + "step": 31913 + }, + { + "epoch": 0.7481144183266232, + "grad_norm": 4.726840019226074, + "learning_rate": 2.9721649465931256e-05, + "loss": 0.1311, + "step": 31914 + }, + { + "epoch": 0.7481378599014282, + "grad_norm": 2.8761026859283447, + "learning_rate": 2.9716409344043438e-05, + "loss": 1.7247, + "step": 31915 + }, + { + "epoch": 0.7481613014762332, + "grad_norm": 3.9551119804382324, + "learning_rate": 2.9711169603516365e-05, + "loss": 0.2338, + "step": 31916 + }, + { + "epoch": 0.7481847430510382, + "grad_norm": 4.493601322174072, + "learning_rate": 2.9705930244378467e-05, + "loss": 0.2353, + "step": 31917 + }, + { + "epoch": 0.7482081846258432, + "grad_norm": 2.3329286575317383, + "learning_rate": 2.9700691266658122e-05, + "loss": 0.1616, + "step": 31918 + }, + { + "epoch": 0.7482316262006482, + "grad_norm": 3.6648647785186768, + "learning_rate": 2.969545267038384e-05, + "loss": 0.4113, + "step": 31919 + }, + { + "epoch": 0.7482550677754531, + "grad_norm": 1.9680253267288208, + "learning_rate": 2.9690214455583986e-05, + "loss": 0.2028, + "step": 31920 + }, + { + "epoch": 0.7482785093502582, + "grad_norm": 2.092531681060791, + "learning_rate": 2.968497662228703e-05, + "loss": 0.1246, + "step": 31921 + }, + { + "epoch": 0.7483019509250631, + "grad_norm": 5.058372974395752, + "learning_rate": 2.967973917052137e-05, + "loss": 0.3708, + "step": 31922 + }, + { + "epoch": 0.7483253924998682, + "grad_norm": 3.318295478820801, + "learning_rate": 2.9674502100315417e-05, + "loss": 1.4307, + "step": 31923 + }, + { + "epoch": 0.7483488340746731, + "grad_norm": 0.8073258996009827, + "learning_rate": 2.96692654116976e-05, + "loss": 0.0822, + "step": 31924 + }, + { + "epoch": 0.7483722756494782, + "grad_norm": 3.880502462387085, + "learning_rate": 2.9664029104696324e-05, + "loss": 0.3503, + "step": 31925 + }, + { + "epoch": 0.7483957172242831, + "grad_norm": 0.5934364199638367, + "learning_rate": 2.965879317933996e-05, + "loss": 0.0238, + "step": 31926 + }, + { + "epoch": 0.7484191587990882, + "grad_norm": 2.3309109210968018, + "learning_rate": 2.9653557635657013e-05, + "loss": 0.1757, + "step": 31927 + }, + { + "epoch": 0.7484426003738931, + "grad_norm": 1.5394748449325562, + "learning_rate": 2.9648322473675826e-05, + "loss": 0.125, + "step": 31928 + }, + { + "epoch": 0.7484660419486981, + "grad_norm": 2.783585786819458, + "learning_rate": 2.9643087693424832e-05, + "loss": 0.2151, + "step": 31929 + }, + { + "epoch": 0.7484894835235031, + "grad_norm": 0.19340291619300842, + "learning_rate": 2.9637853294932384e-05, + "loss": 0.0099, + "step": 31930 + }, + { + "epoch": 0.7485129250983081, + "grad_norm": 2.015310525894165, + "learning_rate": 2.963261927822695e-05, + "loss": 1.4778, + "step": 31931 + }, + { + "epoch": 0.7485363666731131, + "grad_norm": 3.2475717067718506, + "learning_rate": 2.9627385643336915e-05, + "loss": 0.2561, + "step": 31932 + }, + { + "epoch": 0.7485598082479181, + "grad_norm": 5.329776287078857, + "learning_rate": 2.9622152390290615e-05, + "loss": 0.356, + "step": 31933 + }, + { + "epoch": 0.7485832498227231, + "grad_norm": 1.071384310722351, + "learning_rate": 2.961691951911654e-05, + "loss": 0.0513, + "step": 31934 + }, + { + "epoch": 0.7486066913975281, + "grad_norm": 2.959625244140625, + "learning_rate": 2.961168702984304e-05, + "loss": 0.2325, + "step": 31935 + }, + { + "epoch": 0.748630132972333, + "grad_norm": 2.193253993988037, + "learning_rate": 2.9606454922498493e-05, + "loss": 0.3427, + "step": 31936 + }, + { + "epoch": 0.7486535745471381, + "grad_norm": 3.4183096885681152, + "learning_rate": 2.9601223197111306e-05, + "loss": 0.444, + "step": 31937 + }, + { + "epoch": 0.748677016121943, + "grad_norm": 1.2104909420013428, + "learning_rate": 2.9595991853709838e-05, + "loss": 0.0902, + "step": 31938 + }, + { + "epoch": 0.7487004576967481, + "grad_norm": 3.074406385421753, + "learning_rate": 2.9590760892322524e-05, + "loss": 0.241, + "step": 31939 + }, + { + "epoch": 0.748723899271553, + "grad_norm": 1.7768070697784424, + "learning_rate": 2.9585530312977717e-05, + "loss": 0.1323, + "step": 31940 + }, + { + "epoch": 0.7487473408463581, + "grad_norm": 0.5503659844398499, + "learning_rate": 2.9580300115703808e-05, + "loss": 0.0437, + "step": 31941 + }, + { + "epoch": 0.748770782421163, + "grad_norm": 2.789499044418335, + "learning_rate": 2.9575070300529162e-05, + "loss": 1.3878, + "step": 31942 + }, + { + "epoch": 0.7487942239959681, + "grad_norm": 1.070564866065979, + "learning_rate": 2.9569840867482134e-05, + "loss": 0.1016, + "step": 31943 + }, + { + "epoch": 0.748817665570773, + "grad_norm": 1.2678909301757812, + "learning_rate": 2.956461181659117e-05, + "loss": 0.0418, + "step": 31944 + }, + { + "epoch": 0.748841107145578, + "grad_norm": 2.8880913257598877, + "learning_rate": 2.9559383147884567e-05, + "loss": 0.1409, + "step": 31945 + }, + { + "epoch": 0.748864548720383, + "grad_norm": 2.308668375015259, + "learning_rate": 2.955415486139077e-05, + "loss": 0.139, + "step": 31946 + }, + { + "epoch": 0.748887990295188, + "grad_norm": 0.7999188303947449, + "learning_rate": 2.95489269571381e-05, + "loss": 0.0749, + "step": 31947 + }, + { + "epoch": 0.748911431869993, + "grad_norm": 0.5522040128707886, + "learning_rate": 2.9543699435154937e-05, + "loss": 0.0715, + "step": 31948 + }, + { + "epoch": 0.748934873444798, + "grad_norm": 3.169583320617676, + "learning_rate": 2.9538472295469644e-05, + "loss": 0.2448, + "step": 31949 + }, + { + "epoch": 0.748958315019603, + "grad_norm": 2.8419973850250244, + "learning_rate": 2.9533245538110575e-05, + "loss": 0.2598, + "step": 31950 + }, + { + "epoch": 0.748981756594408, + "grad_norm": 2.373363494873047, + "learning_rate": 2.9528019163106068e-05, + "loss": 0.1346, + "step": 31951 + }, + { + "epoch": 0.749005198169213, + "grad_norm": 2.482759952545166, + "learning_rate": 2.9522793170484543e-05, + "loss": 0.1549, + "step": 31952 + }, + { + "epoch": 0.749028639744018, + "grad_norm": 0.7814955711364746, + "learning_rate": 2.9517567560274327e-05, + "loss": 0.0911, + "step": 31953 + }, + { + "epoch": 0.749052081318823, + "grad_norm": 4.281362533569336, + "learning_rate": 2.9512342332503773e-05, + "loss": 0.3135, + "step": 31954 + }, + { + "epoch": 0.749075522893628, + "grad_norm": 3.583059787750244, + "learning_rate": 2.9507117487201195e-05, + "loss": 0.3437, + "step": 31955 + }, + { + "epoch": 0.7490989644684329, + "grad_norm": 0.16349482536315918, + "learning_rate": 2.950189302439501e-05, + "loss": 0.0096, + "step": 31956 + }, + { + "epoch": 0.749122406043238, + "grad_norm": 0.7900330424308777, + "learning_rate": 2.9496668944113535e-05, + "loss": 0.0503, + "step": 31957 + }, + { + "epoch": 0.7491458476180429, + "grad_norm": 0.6256396174430847, + "learning_rate": 2.9491445246385086e-05, + "loss": 0.0442, + "step": 31958 + }, + { + "epoch": 0.749169289192848, + "grad_norm": 3.335831642150879, + "learning_rate": 2.9486221931238078e-05, + "loss": 0.5822, + "step": 31959 + }, + { + "epoch": 0.749192730767653, + "grad_norm": 3.95359468460083, + "learning_rate": 2.9480998998700802e-05, + "loss": 0.3932, + "step": 31960 + }, + { + "epoch": 0.749216172342458, + "grad_norm": 1.6646634340286255, + "learning_rate": 2.9475776448801607e-05, + "loss": 0.0723, + "step": 31961 + }, + { + "epoch": 0.749239613917263, + "grad_norm": 3.4262993335723877, + "learning_rate": 2.9470554281568842e-05, + "loss": 0.4098, + "step": 31962 + }, + { + "epoch": 0.749263055492068, + "grad_norm": 2.7629480361938477, + "learning_rate": 2.9465332497030827e-05, + "loss": 0.3341, + "step": 31963 + }, + { + "epoch": 0.749286497066873, + "grad_norm": 2.573899269104004, + "learning_rate": 2.946011109521587e-05, + "loss": 0.1557, + "step": 31964 + }, + { + "epoch": 0.7493099386416779, + "grad_norm": 1.6813514232635498, + "learning_rate": 2.9454890076152362e-05, + "loss": 0.0824, + "step": 31965 + }, + { + "epoch": 0.749333380216483, + "grad_norm": 2.7210469245910645, + "learning_rate": 2.944966943986861e-05, + "loss": 1.5535, + "step": 31966 + }, + { + "epoch": 0.7493568217912879, + "grad_norm": 3.5512208938598633, + "learning_rate": 2.9444449186392896e-05, + "loss": 0.2071, + "step": 31967 + }, + { + "epoch": 0.749380263366093, + "grad_norm": 2.9330222606658936, + "learning_rate": 2.9439229315753625e-05, + "loss": 0.3789, + "step": 31968 + }, + { + "epoch": 0.7494037049408979, + "grad_norm": 0.8414952754974365, + "learning_rate": 2.943400982797907e-05, + "loss": 0.0733, + "step": 31969 + }, + { + "epoch": 0.749427146515703, + "grad_norm": 1.3225337266921997, + "learning_rate": 2.9428790723097564e-05, + "loss": 0.1191, + "step": 31970 + }, + { + "epoch": 0.7494505880905079, + "grad_norm": 1.1046546697616577, + "learning_rate": 2.94235720011374e-05, + "loss": 0.1407, + "step": 31971 + }, + { + "epoch": 0.749474029665313, + "grad_norm": 2.582693099975586, + "learning_rate": 2.9418353662126953e-05, + "loss": 0.1236, + "step": 31972 + }, + { + "epoch": 0.7494974712401179, + "grad_norm": 4.36942720413208, + "learning_rate": 2.9413135706094485e-05, + "loss": 0.4202, + "step": 31973 + }, + { + "epoch": 0.749520912814923, + "grad_norm": 0.9066629409790039, + "learning_rate": 2.9407918133068347e-05, + "loss": 0.0995, + "step": 31974 + }, + { + "epoch": 0.7495443543897279, + "grad_norm": 3.9814722537994385, + "learning_rate": 2.9402700943076823e-05, + "loss": 0.2166, + "step": 31975 + }, + { + "epoch": 0.7495677959645329, + "grad_norm": 1.3266117572784424, + "learning_rate": 2.939748413614819e-05, + "loss": 0.1448, + "step": 31976 + }, + { + "epoch": 0.7495912375393379, + "grad_norm": 2.5458242893218994, + "learning_rate": 2.939226771231083e-05, + "loss": 0.1927, + "step": 31977 + }, + { + "epoch": 0.7496146791141429, + "grad_norm": 4.422389030456543, + "learning_rate": 2.9387051671593014e-05, + "loss": 0.2184, + "step": 31978 + }, + { + "epoch": 0.7496381206889479, + "grad_norm": 1.985916256904602, + "learning_rate": 2.9381836014023012e-05, + "loss": 0.0941, + "step": 31979 + }, + { + "epoch": 0.7496615622637529, + "grad_norm": 2.9302682876586914, + "learning_rate": 2.937662073962918e-05, + "loss": 0.3075, + "step": 31980 + }, + { + "epoch": 0.7496850038385579, + "grad_norm": 3.451193332672119, + "learning_rate": 2.9371405848439783e-05, + "loss": 0.2981, + "step": 31981 + }, + { + "epoch": 0.7497084454133629, + "grad_norm": 3.888108968734741, + "learning_rate": 2.936619134048313e-05, + "loss": 0.2619, + "step": 31982 + }, + { + "epoch": 0.7497318869881678, + "grad_norm": 4.119363307952881, + "learning_rate": 2.9360977215787477e-05, + "loss": 0.331, + "step": 31983 + }, + { + "epoch": 0.7497553285629729, + "grad_norm": 2.3573384284973145, + "learning_rate": 2.935576347438118e-05, + "loss": 1.5972, + "step": 31984 + }, + { + "epoch": 0.7497787701377778, + "grad_norm": 4.02830696105957, + "learning_rate": 2.9350550116292497e-05, + "loss": 0.6254, + "step": 31985 + }, + { + "epoch": 0.7498022117125829, + "grad_norm": 3.1667230129241943, + "learning_rate": 2.9345337141549722e-05, + "loss": 0.3951, + "step": 31986 + }, + { + "epoch": 0.7498256532873878, + "grad_norm": 2.1831743717193604, + "learning_rate": 2.9340124550181126e-05, + "loss": 0.1469, + "step": 31987 + }, + { + "epoch": 0.7498490948621929, + "grad_norm": 1.6042100191116333, + "learning_rate": 2.9334912342215004e-05, + "loss": 0.0951, + "step": 31988 + }, + { + "epoch": 0.7498725364369978, + "grad_norm": 3.174039840698242, + "learning_rate": 2.9329700517679603e-05, + "loss": 0.3105, + "step": 31989 + }, + { + "epoch": 0.7498959780118029, + "grad_norm": 1.4733061790466309, + "learning_rate": 2.9324489076603258e-05, + "loss": 0.0717, + "step": 31990 + }, + { + "epoch": 0.7499194195866078, + "grad_norm": 0.870048999786377, + "learning_rate": 2.9319278019014194e-05, + "loss": 0.0383, + "step": 31991 + }, + { + "epoch": 0.7499428611614128, + "grad_norm": 1.7548404932022095, + "learning_rate": 2.9314067344940755e-05, + "loss": 1.0354, + "step": 31992 + }, + { + "epoch": 0.7499663027362178, + "grad_norm": 3.7103474140167236, + "learning_rate": 2.9308857054411166e-05, + "loss": 0.2408, + "step": 31993 + }, + { + "epoch": 0.7499897443110228, + "grad_norm": 0.36637023091316223, + "learning_rate": 2.9303647147453715e-05, + "loss": 0.0122, + "step": 31994 + }, + { + "epoch": 0.7500131858858278, + "grad_norm": 1.0223037004470825, + "learning_rate": 2.9298437624096652e-05, + "loss": 0.1117, + "step": 31995 + }, + { + "epoch": 0.7500366274606328, + "grad_norm": 1.5788681507110596, + "learning_rate": 2.9293228484368228e-05, + "loss": 0.1717, + "step": 31996 + }, + { + "epoch": 0.7500600690354378, + "grad_norm": 4.718828201293945, + "learning_rate": 2.9288019728296767e-05, + "loss": 0.4116, + "step": 31997 + }, + { + "epoch": 0.7500835106102428, + "grad_norm": 0.851808488368988, + "learning_rate": 2.92828113559105e-05, + "loss": 0.0711, + "step": 31998 + }, + { + "epoch": 0.7501069521850477, + "grad_norm": 0.5504887700080872, + "learning_rate": 2.9277603367237684e-05, + "loss": 0.0286, + "step": 31999 + }, + { + "epoch": 0.7501303937598528, + "grad_norm": 3.4499146938323975, + "learning_rate": 2.927239576230658e-05, + "loss": 0.2113, + "step": 32000 + }, + { + "epoch": 0.7501538353346577, + "grad_norm": 1.758622407913208, + "learning_rate": 2.926718854114544e-05, + "loss": 1.164, + "step": 32001 + }, + { + "epoch": 0.7501772769094628, + "grad_norm": 1.680633783340454, + "learning_rate": 2.9261981703782505e-05, + "loss": 0.1356, + "step": 32002 + }, + { + "epoch": 0.7502007184842677, + "grad_norm": 3.980862855911255, + "learning_rate": 2.9256775250246037e-05, + "loss": 0.27, + "step": 32003 + }, + { + "epoch": 0.7502241600590728, + "grad_norm": 2.056077718734741, + "learning_rate": 2.9251569180564332e-05, + "loss": 0.2155, + "step": 32004 + }, + { + "epoch": 0.7502476016338777, + "grad_norm": 2.8488004207611084, + "learning_rate": 2.9246363494765592e-05, + "loss": 0.3772, + "step": 32005 + }, + { + "epoch": 0.7502710432086828, + "grad_norm": 3.6339263916015625, + "learning_rate": 2.924115819287808e-05, + "loss": 0.3748, + "step": 32006 + }, + { + "epoch": 0.7502944847834877, + "grad_norm": 1.7479695081710815, + "learning_rate": 2.923595327493004e-05, + "loss": 0.128, + "step": 32007 + }, + { + "epoch": 0.7503179263582928, + "grad_norm": 2.4828124046325684, + "learning_rate": 2.9230748740949666e-05, + "loss": 0.2213, + "step": 32008 + }, + { + "epoch": 0.7503413679330977, + "grad_norm": 3.248272180557251, + "learning_rate": 2.9225544590965282e-05, + "loss": 0.2797, + "step": 32009 + }, + { + "epoch": 0.7503648095079027, + "grad_norm": 4.968419075012207, + "learning_rate": 2.9220340825005068e-05, + "loss": 0.5155, + "step": 32010 + }, + { + "epoch": 0.7503882510827078, + "grad_norm": 1.804101586341858, + "learning_rate": 2.9215137443097286e-05, + "loss": 0.0733, + "step": 32011 + }, + { + "epoch": 0.7504116926575127, + "grad_norm": 4.460813522338867, + "learning_rate": 2.920993444527015e-05, + "loss": 0.4453, + "step": 32012 + }, + { + "epoch": 0.7504351342323178, + "grad_norm": 3.633810520172119, + "learning_rate": 2.9204731831551912e-05, + "loss": 0.221, + "step": 32013 + }, + { + "epoch": 0.7504585758071227, + "grad_norm": 1.250644326210022, + "learning_rate": 2.9199529601970744e-05, + "loss": 0.0489, + "step": 32014 + }, + { + "epoch": 0.7504820173819278, + "grad_norm": 0.29631295800209045, + "learning_rate": 2.9194327756554927e-05, + "loss": 0.0135, + "step": 32015 + }, + { + "epoch": 0.7505054589567327, + "grad_norm": 2.8590900897979736, + "learning_rate": 2.9189126295332714e-05, + "loss": 0.1691, + "step": 32016 + }, + { + "epoch": 0.7505289005315378, + "grad_norm": 6.579126358032227, + "learning_rate": 2.9183925218332285e-05, + "loss": 0.3896, + "step": 32017 + }, + { + "epoch": 0.7505523421063427, + "grad_norm": 3.5123677253723145, + "learning_rate": 2.9178724525581868e-05, + "loss": 0.3344, + "step": 32018 + }, + { + "epoch": 0.7505757836811477, + "grad_norm": 2.4537360668182373, + "learning_rate": 2.9173524217109693e-05, + "loss": 0.2255, + "step": 32019 + }, + { + "epoch": 0.7505992252559527, + "grad_norm": 4.10603666305542, + "learning_rate": 2.9168324292943962e-05, + "loss": 0.1724, + "step": 32020 + }, + { + "epoch": 0.7506226668307577, + "grad_norm": 0.6331937313079834, + "learning_rate": 2.9163124753112858e-05, + "loss": 0.0606, + "step": 32021 + }, + { + "epoch": 0.7506461084055627, + "grad_norm": 3.76998233795166, + "learning_rate": 2.9157925597644675e-05, + "loss": 0.2157, + "step": 32022 + }, + { + "epoch": 0.7506695499803677, + "grad_norm": 3.0502371788024902, + "learning_rate": 2.9152726826567568e-05, + "loss": 0.3307, + "step": 32023 + }, + { + "epoch": 0.7506929915551727, + "grad_norm": 9.439197540283203, + "learning_rate": 2.9147528439909754e-05, + "loss": 0.2979, + "step": 32024 + }, + { + "epoch": 0.7507164331299777, + "grad_norm": 2.2785398960113525, + "learning_rate": 2.914233043769945e-05, + "loss": 1.4723, + "step": 32025 + }, + { + "epoch": 0.7507398747047827, + "grad_norm": 3.7069568634033203, + "learning_rate": 2.9137132819964864e-05, + "loss": 0.1524, + "step": 32026 + }, + { + "epoch": 0.7507633162795877, + "grad_norm": 4.714395523071289, + "learning_rate": 2.913193558673414e-05, + "loss": 0.2453, + "step": 32027 + }, + { + "epoch": 0.7507867578543926, + "grad_norm": 1.7175012826919556, + "learning_rate": 2.9126738738035532e-05, + "loss": 1.3762, + "step": 32028 + }, + { + "epoch": 0.7508101994291977, + "grad_norm": 2.5688469409942627, + "learning_rate": 2.9121542273897262e-05, + "loss": 0.1654, + "step": 32029 + }, + { + "epoch": 0.7508336410040026, + "grad_norm": 3.8673481941223145, + "learning_rate": 2.91163461943475e-05, + "loss": 0.2267, + "step": 32030 + }, + { + "epoch": 0.7508570825788077, + "grad_norm": 1.0803595781326294, + "learning_rate": 2.911115049941443e-05, + "loss": 0.0488, + "step": 32031 + }, + { + "epoch": 0.7508805241536126, + "grad_norm": 2.7515981197357178, + "learning_rate": 2.9105955189126254e-05, + "loss": 2.0312, + "step": 32032 + }, + { + "epoch": 0.7509039657284177, + "grad_norm": 1.2658711671829224, + "learning_rate": 2.9100760263511162e-05, + "loss": 0.0911, + "step": 32033 + }, + { + "epoch": 0.7509274073032226, + "grad_norm": 4.08188533782959, + "learning_rate": 2.9095565722597306e-05, + "loss": 0.3708, + "step": 32034 + }, + { + "epoch": 0.7509508488780277, + "grad_norm": 7.463450908660889, + "learning_rate": 2.9090371566412933e-05, + "loss": 0.1858, + "step": 32035 + }, + { + "epoch": 0.7509742904528326, + "grad_norm": 4.578990459442139, + "learning_rate": 2.9085177794986196e-05, + "loss": 0.1437, + "step": 32036 + }, + { + "epoch": 0.7509977320276376, + "grad_norm": 2.8146142959594727, + "learning_rate": 2.9079984408345286e-05, + "loss": 0.1835, + "step": 32037 + }, + { + "epoch": 0.7510211736024426, + "grad_norm": 1.715957760810852, + "learning_rate": 2.9074791406518366e-05, + "loss": 0.1732, + "step": 32038 + }, + { + "epoch": 0.7510446151772476, + "grad_norm": 1.9219205379486084, + "learning_rate": 2.9069598789533625e-05, + "loss": 0.2294, + "step": 32039 + }, + { + "epoch": 0.7510680567520526, + "grad_norm": 3.18454647064209, + "learning_rate": 2.9064406557419212e-05, + "loss": 0.278, + "step": 32040 + }, + { + "epoch": 0.7510914983268576, + "grad_norm": 4.211578369140625, + "learning_rate": 2.905921471020331e-05, + "loss": 0.2975, + "step": 32041 + }, + { + "epoch": 0.7511149399016626, + "grad_norm": 2.4843862056732178, + "learning_rate": 2.9054023247914143e-05, + "loss": 0.1607, + "step": 32042 + }, + { + "epoch": 0.7511383814764676, + "grad_norm": 3.144300699234009, + "learning_rate": 2.904883217057984e-05, + "loss": 0.2443, + "step": 32043 + }, + { + "epoch": 0.7511618230512725, + "grad_norm": 3.1682212352752686, + "learning_rate": 2.9043641478228568e-05, + "loss": 0.3871, + "step": 32044 + }, + { + "epoch": 0.7511852646260776, + "grad_norm": 3.7984797954559326, + "learning_rate": 2.9038451170888493e-05, + "loss": 0.3363, + "step": 32045 + }, + { + "epoch": 0.7512087062008825, + "grad_norm": 3.1676697731018066, + "learning_rate": 2.903326124858774e-05, + "loss": 0.2683, + "step": 32046 + }, + { + "epoch": 0.7512321477756876, + "grad_norm": 2.219604969024658, + "learning_rate": 2.9028071711354544e-05, + "loss": 0.1655, + "step": 32047 + }, + { + "epoch": 0.7512555893504925, + "grad_norm": 3.5540590286254883, + "learning_rate": 2.9022882559217024e-05, + "loss": 0.2286, + "step": 32048 + }, + { + "epoch": 0.7512790309252976, + "grad_norm": 1.629711627960205, + "learning_rate": 2.901769379220335e-05, + "loss": 0.1497, + "step": 32049 + }, + { + "epoch": 0.7513024725001025, + "grad_norm": 2.0969083309173584, + "learning_rate": 2.901250541034165e-05, + "loss": 0.1657, + "step": 32050 + }, + { + "epoch": 0.7513259140749076, + "grad_norm": 3.080519199371338, + "learning_rate": 2.9007317413660096e-05, + "loss": 0.2257, + "step": 32051 + }, + { + "epoch": 0.7513493556497125, + "grad_norm": 5.19071102142334, + "learning_rate": 2.9002129802186806e-05, + "loss": 0.4297, + "step": 32052 + }, + { + "epoch": 0.7513727972245176, + "grad_norm": 3.9574081897735596, + "learning_rate": 2.8996942575949958e-05, + "loss": 0.2674, + "step": 32053 + }, + { + "epoch": 0.7513962387993225, + "grad_norm": 2.5029687881469727, + "learning_rate": 2.899175573497772e-05, + "loss": 0.2189, + "step": 32054 + }, + { + "epoch": 0.7514196803741275, + "grad_norm": 2.4196598529815674, + "learning_rate": 2.8986569279298215e-05, + "loss": 0.1184, + "step": 32055 + }, + { + "epoch": 0.7514431219489325, + "grad_norm": 3.0734312534332275, + "learning_rate": 2.8981383208939582e-05, + "loss": 0.1592, + "step": 32056 + }, + { + "epoch": 0.7514665635237375, + "grad_norm": 3.818732738494873, + "learning_rate": 2.897619752392996e-05, + "loss": 0.3148, + "step": 32057 + }, + { + "epoch": 0.7514900050985425, + "grad_norm": 1.9190043210983276, + "learning_rate": 2.897101222429749e-05, + "loss": 0.1487, + "step": 32058 + }, + { + "epoch": 0.7515134466733475, + "grad_norm": 1.6935819387435913, + "learning_rate": 2.896582731007027e-05, + "loss": 0.1077, + "step": 32059 + }, + { + "epoch": 0.7515368882481525, + "grad_norm": 4.7535481452941895, + "learning_rate": 2.8960642781276502e-05, + "loss": 0.2484, + "step": 32060 + }, + { + "epoch": 0.7515603298229575, + "grad_norm": 2.8407342433929443, + "learning_rate": 2.8955458637944276e-05, + "loss": 0.2189, + "step": 32061 + }, + { + "epoch": 0.7515837713977626, + "grad_norm": 1.2699663639068604, + "learning_rate": 2.8950274880101746e-05, + "loss": 0.0386, + "step": 32062 + }, + { + "epoch": 0.7516072129725675, + "grad_norm": 3.8128437995910645, + "learning_rate": 2.894509150777701e-05, + "loss": 0.3145, + "step": 32063 + }, + { + "epoch": 0.7516306545473725, + "grad_norm": 9.361820220947266, + "learning_rate": 2.893990852099817e-05, + "loss": 0.3695, + "step": 32064 + }, + { + "epoch": 0.7516540961221775, + "grad_norm": 2.990100860595703, + "learning_rate": 2.8934725919793427e-05, + "loss": 0.313, + "step": 32065 + }, + { + "epoch": 0.7516775376969825, + "grad_norm": 3.2366750240325928, + "learning_rate": 2.892954370419082e-05, + "loss": 0.26, + "step": 32066 + }, + { + "epoch": 0.7517009792717875, + "grad_norm": 2.4600367546081543, + "learning_rate": 2.892436187421854e-05, + "loss": 0.3505, + "step": 32067 + }, + { + "epoch": 0.7517244208465925, + "grad_norm": 3.075326442718506, + "learning_rate": 2.891918042990468e-05, + "loss": 0.3073, + "step": 32068 + }, + { + "epoch": 0.7517478624213975, + "grad_norm": 4.257193088531494, + "learning_rate": 2.8913999371277335e-05, + "loss": 0.2585, + "step": 32069 + }, + { + "epoch": 0.7517713039962025, + "grad_norm": 1.8968793153762817, + "learning_rate": 2.8908818698364627e-05, + "loss": 0.1574, + "step": 32070 + }, + { + "epoch": 0.7517947455710075, + "grad_norm": 3.614515542984009, + "learning_rate": 2.8903638411194678e-05, + "loss": 0.1912, + "step": 32071 + }, + { + "epoch": 0.7518181871458125, + "grad_norm": 1.8567066192626953, + "learning_rate": 2.8898458509795544e-05, + "loss": 0.1217, + "step": 32072 + }, + { + "epoch": 0.7518416287206174, + "grad_norm": 0.416597843170166, + "learning_rate": 2.8893278994195404e-05, + "loss": 0.0368, + "step": 32073 + }, + { + "epoch": 0.7518650702954225, + "grad_norm": 0.5558127164840698, + "learning_rate": 2.8888099864422347e-05, + "loss": 0.038, + "step": 32074 + }, + { + "epoch": 0.7518885118702274, + "grad_norm": 0.5250779986381531, + "learning_rate": 2.888292112050445e-05, + "loss": 0.0426, + "step": 32075 + }, + { + "epoch": 0.7519119534450325, + "grad_norm": 3.9971823692321777, + "learning_rate": 2.8877742762469784e-05, + "loss": 0.3205, + "step": 32076 + }, + { + "epoch": 0.7519353950198374, + "grad_norm": 2.7104434967041016, + "learning_rate": 2.8872564790346524e-05, + "loss": 1.221, + "step": 32077 + }, + { + "epoch": 0.7519588365946425, + "grad_norm": 5.862802505493164, + "learning_rate": 2.886738720416272e-05, + "loss": 0.6277, + "step": 32078 + }, + { + "epoch": 0.7519822781694474, + "grad_norm": 3.698607921600342, + "learning_rate": 2.8862210003946456e-05, + "loss": 0.4302, + "step": 32079 + }, + { + "epoch": 0.7520057197442525, + "grad_norm": 2.2073442935943604, + "learning_rate": 2.8857033189725868e-05, + "loss": 0.0606, + "step": 32080 + }, + { + "epoch": 0.7520291613190574, + "grad_norm": 2.820661783218384, + "learning_rate": 2.8851856761529007e-05, + "loss": 0.183, + "step": 32081 + }, + { + "epoch": 0.7520526028938624, + "grad_norm": 2.048624277114868, + "learning_rate": 2.884668071938398e-05, + "loss": 1.3127, + "step": 32082 + }, + { + "epoch": 0.7520760444686674, + "grad_norm": 3.067366600036621, + "learning_rate": 2.8841505063318862e-05, + "loss": 0.2077, + "step": 32083 + }, + { + "epoch": 0.7520994860434724, + "grad_norm": 1.6878970861434937, + "learning_rate": 2.8836329793361717e-05, + "loss": 0.1131, + "step": 32084 + }, + { + "epoch": 0.7521229276182774, + "grad_norm": 4.899374485015869, + "learning_rate": 2.883115490954067e-05, + "loss": 0.5556, + "step": 32085 + }, + { + "epoch": 0.7521463691930824, + "grad_norm": 1.8582733869552612, + "learning_rate": 2.882598041188378e-05, + "loss": 0.0736, + "step": 32086 + }, + { + "epoch": 0.7521698107678874, + "grad_norm": 2.5449273586273193, + "learning_rate": 2.8820806300419113e-05, + "loss": 0.1673, + "step": 32087 + }, + { + "epoch": 0.7521932523426924, + "grad_norm": 2.405331611633301, + "learning_rate": 2.8815632575174734e-05, + "loss": 0.1651, + "step": 32088 + }, + { + "epoch": 0.7522166939174973, + "grad_norm": 3.229973554611206, + "learning_rate": 2.8810459236178755e-05, + "loss": 0.1778, + "step": 32089 + }, + { + "epoch": 0.7522401354923024, + "grad_norm": 2.575808525085449, + "learning_rate": 2.8805286283459232e-05, + "loss": 0.0977, + "step": 32090 + }, + { + "epoch": 0.7522635770671073, + "grad_norm": 2.2291085720062256, + "learning_rate": 2.8800113717044186e-05, + "loss": 0.1186, + "step": 32091 + }, + { + "epoch": 0.7522870186419124, + "grad_norm": 3.1139163970947266, + "learning_rate": 2.879494153696177e-05, + "loss": 0.2335, + "step": 32092 + }, + { + "epoch": 0.7523104602167173, + "grad_norm": 2.10308837890625, + "learning_rate": 2.8789769743240002e-05, + "loss": 0.1839, + "step": 32093 + }, + { + "epoch": 0.7523339017915224, + "grad_norm": 3.4241626262664795, + "learning_rate": 2.878459833590693e-05, + "loss": 0.3102, + "step": 32094 + }, + { + "epoch": 0.7523573433663273, + "grad_norm": 2.2330920696258545, + "learning_rate": 2.8779427314990646e-05, + "loss": 1.5506, + "step": 32095 + }, + { + "epoch": 0.7523807849411324, + "grad_norm": 2.1324522495269775, + "learning_rate": 2.8774256680519185e-05, + "loss": 0.088, + "step": 32096 + }, + { + "epoch": 0.7524042265159373, + "grad_norm": 2.673992395401001, + "learning_rate": 2.8769086432520587e-05, + "loss": 0.3046, + "step": 32097 + }, + { + "epoch": 0.7524276680907424, + "grad_norm": 3.5526087284088135, + "learning_rate": 2.8763916571022953e-05, + "loss": 0.3439, + "step": 32098 + }, + { + "epoch": 0.7524511096655473, + "grad_norm": 4.53336238861084, + "learning_rate": 2.8758747096054307e-05, + "loss": 0.3728, + "step": 32099 + }, + { + "epoch": 0.7524745512403523, + "grad_norm": 1.91270911693573, + "learning_rate": 2.8753578007642712e-05, + "loss": 0.1049, + "step": 32100 + }, + { + "epoch": 0.7524979928151573, + "grad_norm": 1.0717276334762573, + "learning_rate": 2.8748409305816172e-05, + "loss": 0.1289, + "step": 32101 + }, + { + "epoch": 0.7525214343899623, + "grad_norm": 0.9502055048942566, + "learning_rate": 2.8743240990602795e-05, + "loss": 0.105, + "step": 32102 + }, + { + "epoch": 0.7525448759647673, + "grad_norm": 2.912163257598877, + "learning_rate": 2.873807306203059e-05, + "loss": 0.1154, + "step": 32103 + }, + { + "epoch": 0.7525683175395723, + "grad_norm": 1.1933118104934692, + "learning_rate": 2.8732905520127586e-05, + "loss": 0.1215, + "step": 32104 + }, + { + "epoch": 0.7525917591143773, + "grad_norm": 3.3558895587921143, + "learning_rate": 2.8727738364921865e-05, + "loss": 0.1647, + "step": 32105 + }, + { + "epoch": 0.7526152006891823, + "grad_norm": 3.202319860458374, + "learning_rate": 2.8722571596441437e-05, + "loss": 0.1951, + "step": 32106 + }, + { + "epoch": 0.7526386422639872, + "grad_norm": 3.0372724533081055, + "learning_rate": 2.8717405214714342e-05, + "loss": 0.2457, + "step": 32107 + }, + { + "epoch": 0.7526620838387923, + "grad_norm": 1.3347078561782837, + "learning_rate": 2.871223921976861e-05, + "loss": 0.07, + "step": 32108 + }, + { + "epoch": 0.7526855254135972, + "grad_norm": 3.128910779953003, + "learning_rate": 2.8707073611632274e-05, + "loss": 0.1155, + "step": 32109 + }, + { + "epoch": 0.7527089669884023, + "grad_norm": 3.277963876724243, + "learning_rate": 2.8701908390333333e-05, + "loss": 0.2564, + "step": 32110 + }, + { + "epoch": 0.7527324085632072, + "grad_norm": 0.8278598189353943, + "learning_rate": 2.8696743555899863e-05, + "loss": 0.0569, + "step": 32111 + }, + { + "epoch": 0.7527558501380123, + "grad_norm": 1.9594274759292603, + "learning_rate": 2.869157910835988e-05, + "loss": 0.1642, + "step": 32112 + }, + { + "epoch": 0.7527792917128173, + "grad_norm": 1.0032391548156738, + "learning_rate": 2.8686415047741357e-05, + "loss": 0.0883, + "step": 32113 + }, + { + "epoch": 0.7528027332876223, + "grad_norm": 2.9641427993774414, + "learning_rate": 2.868125137407238e-05, + "loss": 0.3517, + "step": 32114 + }, + { + "epoch": 0.7528261748624273, + "grad_norm": 2.2517619132995605, + "learning_rate": 2.8676088087380937e-05, + "loss": 0.2041, + "step": 32115 + }, + { + "epoch": 0.7528496164372322, + "grad_norm": 0.7482832670211792, + "learning_rate": 2.8670925187695053e-05, + "loss": 0.0412, + "step": 32116 + }, + { + "epoch": 0.7528730580120373, + "grad_norm": 3.253476858139038, + "learning_rate": 2.8665762675042696e-05, + "loss": 0.3353, + "step": 32117 + }, + { + "epoch": 0.7528964995868422, + "grad_norm": 3.0477893352508545, + "learning_rate": 2.8660600549451955e-05, + "loss": 1.262, + "step": 32118 + }, + { + "epoch": 0.7529199411616473, + "grad_norm": 2.4887888431549072, + "learning_rate": 2.8655438810950798e-05, + "loss": 0.181, + "step": 32119 + }, + { + "epoch": 0.7529433827364522, + "grad_norm": 3.0632691383361816, + "learning_rate": 2.8650277459567232e-05, + "loss": 0.1264, + "step": 32120 + }, + { + "epoch": 0.7529668243112573, + "grad_norm": 2.4094579219818115, + "learning_rate": 2.864511649532926e-05, + "loss": 0.1797, + "step": 32121 + }, + { + "epoch": 0.7529902658860622, + "grad_norm": 4.245368480682373, + "learning_rate": 2.863995591826487e-05, + "loss": 0.3734, + "step": 32122 + }, + { + "epoch": 0.7530137074608673, + "grad_norm": 3.807685613632202, + "learning_rate": 2.863479572840212e-05, + "loss": 0.2729, + "step": 32123 + }, + { + "epoch": 0.7530371490356722, + "grad_norm": 0.44956865906715393, + "learning_rate": 2.8629635925768973e-05, + "loss": 0.0488, + "step": 32124 + }, + { + "epoch": 0.7530605906104773, + "grad_norm": 0.7682528495788574, + "learning_rate": 2.8624476510393395e-05, + "loss": 0.0199, + "step": 32125 + }, + { + "epoch": 0.7530840321852822, + "grad_norm": 3.2550065517425537, + "learning_rate": 2.8619317482303452e-05, + "loss": 0.395, + "step": 32126 + }, + { + "epoch": 0.7531074737600872, + "grad_norm": 0.4418160915374756, + "learning_rate": 2.8614158841527094e-05, + "loss": 0.0473, + "step": 32127 + }, + { + "epoch": 0.7531309153348922, + "grad_norm": 0.9838016033172607, + "learning_rate": 2.8609000588092316e-05, + "loss": 0.0663, + "step": 32128 + }, + { + "epoch": 0.7531543569096972, + "grad_norm": 0.866142988204956, + "learning_rate": 2.8603842722027087e-05, + "loss": 0.0712, + "step": 32129 + }, + { + "epoch": 0.7531777984845022, + "grad_norm": 0.6192330718040466, + "learning_rate": 2.8598685243359437e-05, + "loss": 0.0367, + "step": 32130 + }, + { + "epoch": 0.7532012400593072, + "grad_norm": 2.9519710540771484, + "learning_rate": 2.8593528152117322e-05, + "loss": 0.1707, + "step": 32131 + }, + { + "epoch": 0.7532246816341122, + "grad_norm": 2.3263750076293945, + "learning_rate": 2.8588371448328745e-05, + "loss": 0.2212, + "step": 32132 + }, + { + "epoch": 0.7532481232089172, + "grad_norm": 3.664963483810425, + "learning_rate": 2.8583215132021657e-05, + "loss": 0.2773, + "step": 32133 + }, + { + "epoch": 0.7532715647837221, + "grad_norm": 1.8915952444076538, + "learning_rate": 2.8578059203224063e-05, + "loss": 0.1549, + "step": 32134 + }, + { + "epoch": 0.7532950063585272, + "grad_norm": 2.270402669906616, + "learning_rate": 2.857290366196389e-05, + "loss": 0.1487, + "step": 32135 + }, + { + "epoch": 0.7533184479333321, + "grad_norm": 0.5684643387794495, + "learning_rate": 2.8567748508269187e-05, + "loss": 0.0382, + "step": 32136 + }, + { + "epoch": 0.7533418895081372, + "grad_norm": 1.6660739183425903, + "learning_rate": 2.8562593742167864e-05, + "loss": 1.31, + "step": 32137 + }, + { + "epoch": 0.7533653310829421, + "grad_norm": 3.5292537212371826, + "learning_rate": 2.8557439363687933e-05, + "loss": 0.3082, + "step": 32138 + }, + { + "epoch": 0.7533887726577472, + "grad_norm": 2.6211001873016357, + "learning_rate": 2.8552285372857358e-05, + "loss": 0.2949, + "step": 32139 + }, + { + "epoch": 0.7534122142325521, + "grad_norm": 2.985546112060547, + "learning_rate": 2.8547131769704082e-05, + "loss": 0.2104, + "step": 32140 + }, + { + "epoch": 0.7534356558073572, + "grad_norm": 0.6034526228904724, + "learning_rate": 2.8541978554256075e-05, + "loss": 0.0412, + "step": 32141 + }, + { + "epoch": 0.7534590973821621, + "grad_norm": 2.8113460540771484, + "learning_rate": 2.8536825726541283e-05, + "loss": 0.2692, + "step": 32142 + }, + { + "epoch": 0.7534825389569672, + "grad_norm": 2.847066879272461, + "learning_rate": 2.8531673286587702e-05, + "loss": 0.2526, + "step": 32143 + }, + { + "epoch": 0.7535059805317721, + "grad_norm": 3.1301374435424805, + "learning_rate": 2.8526521234423286e-05, + "loss": 0.1851, + "step": 32144 + }, + { + "epoch": 0.7535294221065771, + "grad_norm": 2.182196617126465, + "learning_rate": 2.852136957007596e-05, + "loss": 0.1142, + "step": 32145 + }, + { + "epoch": 0.7535528636813821, + "grad_norm": 0.21705828607082367, + "learning_rate": 2.8516218293573704e-05, + "loss": 0.0175, + "step": 32146 + }, + { + "epoch": 0.7535763052561871, + "grad_norm": 1.8746730089187622, + "learning_rate": 2.8511067404944448e-05, + "loss": 0.1332, + "step": 32147 + }, + { + "epoch": 0.7535997468309921, + "grad_norm": 0.940307080745697, + "learning_rate": 2.8505916904216123e-05, + "loss": 0.1149, + "step": 32148 + }, + { + "epoch": 0.7536231884057971, + "grad_norm": 2.574488878250122, + "learning_rate": 2.850076679141671e-05, + "loss": 0.1835, + "step": 32149 + }, + { + "epoch": 0.7536466299806021, + "grad_norm": 3.5245602130889893, + "learning_rate": 2.8495617066574176e-05, + "loss": 0.356, + "step": 32150 + }, + { + "epoch": 0.7536700715554071, + "grad_norm": 3.25075364112854, + "learning_rate": 2.8490467729716432e-05, + "loss": 0.1875, + "step": 32151 + }, + { + "epoch": 0.753693513130212, + "grad_norm": 2.8005404472351074, + "learning_rate": 2.8485318780871416e-05, + "loss": 0.2716, + "step": 32152 + }, + { + "epoch": 0.7537169547050171, + "grad_norm": 2.016975164413452, + "learning_rate": 2.848017022006707e-05, + "loss": 0.2403, + "step": 32153 + }, + { + "epoch": 0.753740396279822, + "grad_norm": 2.6189966201782227, + "learning_rate": 2.8475022047331335e-05, + "loss": 0.3745, + "step": 32154 + }, + { + "epoch": 0.7537638378546271, + "grad_norm": 2.826019287109375, + "learning_rate": 2.846987426269212e-05, + "loss": 0.1801, + "step": 32155 + }, + { + "epoch": 0.753787279429432, + "grad_norm": 1.2916033267974854, + "learning_rate": 2.8464726866177393e-05, + "loss": 0.036, + "step": 32156 + }, + { + "epoch": 0.7538107210042371, + "grad_norm": 2.6362671852111816, + "learning_rate": 2.8459579857815077e-05, + "loss": 0.2568, + "step": 32157 + }, + { + "epoch": 0.753834162579042, + "grad_norm": 3.4484403133392334, + "learning_rate": 2.8454433237633094e-05, + "loss": 1.5222, + "step": 32158 + }, + { + "epoch": 0.7538576041538471, + "grad_norm": 3.288461923599243, + "learning_rate": 2.8449287005659362e-05, + "loss": 0.306, + "step": 32159 + }, + { + "epoch": 0.753881045728652, + "grad_norm": 3.516554594039917, + "learning_rate": 2.844414116192179e-05, + "loss": 0.3364, + "step": 32160 + }, + { + "epoch": 0.753904487303457, + "grad_norm": 2.0754311084747314, + "learning_rate": 2.8438995706448348e-05, + "loss": 0.0944, + "step": 32161 + }, + { + "epoch": 0.753927928878262, + "grad_norm": 1.2605524063110352, + "learning_rate": 2.843385063926689e-05, + "loss": 0.1762, + "step": 32162 + }, + { + "epoch": 0.753951370453067, + "grad_norm": 2.649386405944824, + "learning_rate": 2.8428705960405412e-05, + "loss": 0.2514, + "step": 32163 + }, + { + "epoch": 0.7539748120278721, + "grad_norm": 0.4438009262084961, + "learning_rate": 2.8423561669891775e-05, + "loss": 0.0321, + "step": 32164 + }, + { + "epoch": 0.753998253602677, + "grad_norm": 2.4447779655456543, + "learning_rate": 2.8418417767753915e-05, + "loss": 1.5907, + "step": 32165 + }, + { + "epoch": 0.7540216951774821, + "grad_norm": 4.012757301330566, + "learning_rate": 2.841327425401973e-05, + "loss": 0.3444, + "step": 32166 + }, + { + "epoch": 0.754045136752287, + "grad_norm": 2.106241226196289, + "learning_rate": 2.8408131128717097e-05, + "loss": 0.1481, + "step": 32167 + }, + { + "epoch": 0.7540685783270921, + "grad_norm": 0.1681956648826599, + "learning_rate": 2.8402988391873985e-05, + "loss": 0.007, + "step": 32168 + }, + { + "epoch": 0.754092019901897, + "grad_norm": 2.801118850708008, + "learning_rate": 2.8397846043518283e-05, + "loss": 0.2223, + "step": 32169 + }, + { + "epoch": 0.754115461476702, + "grad_norm": 2.422046184539795, + "learning_rate": 2.8392704083677878e-05, + "loss": 1.1208, + "step": 32170 + }, + { + "epoch": 0.754138903051507, + "grad_norm": 1.0158907175064087, + "learning_rate": 2.838756251238067e-05, + "loss": 0.0692, + "step": 32171 + }, + { + "epoch": 0.754162344626312, + "grad_norm": 0.9750000238418579, + "learning_rate": 2.8382421329654563e-05, + "loss": 0.091, + "step": 32172 + }, + { + "epoch": 0.754185786201117, + "grad_norm": 3.1274702548980713, + "learning_rate": 2.8377280535527418e-05, + "loss": 0.225, + "step": 32173 + }, + { + "epoch": 0.754209227775922, + "grad_norm": 0.6333909630775452, + "learning_rate": 2.837214013002717e-05, + "loss": 0.0469, + "step": 32174 + }, + { + "epoch": 0.754232669350727, + "grad_norm": 3.004581928253174, + "learning_rate": 2.836700011318173e-05, + "loss": 0.2713, + "step": 32175 + }, + { + "epoch": 0.754256110925532, + "grad_norm": 3.0802738666534424, + "learning_rate": 2.8361860485018954e-05, + "loss": 0.2876, + "step": 32176 + }, + { + "epoch": 0.754279552500337, + "grad_norm": 1.0063352584838867, + "learning_rate": 2.8356721245566754e-05, + "loss": 0.116, + "step": 32177 + }, + { + "epoch": 0.754302994075142, + "grad_norm": 1.7209821939468384, + "learning_rate": 2.8351582394852983e-05, + "loss": 0.7551, + "step": 32178 + }, + { + "epoch": 0.754326435649947, + "grad_norm": 4.236483097076416, + "learning_rate": 2.8346443932905552e-05, + "loss": 0.2041, + "step": 32179 + }, + { + "epoch": 0.754349877224752, + "grad_norm": 1.6529018878936768, + "learning_rate": 2.83413058597523e-05, + "loss": 0.2142, + "step": 32180 + }, + { + "epoch": 0.7543733187995569, + "grad_norm": 4.457986831665039, + "learning_rate": 2.833616817542115e-05, + "loss": 0.3472, + "step": 32181 + }, + { + "epoch": 0.754396760374362, + "grad_norm": 4.586315155029297, + "learning_rate": 2.833103087993999e-05, + "loss": 0.2623, + "step": 32182 + }, + { + "epoch": 0.7544202019491669, + "grad_norm": 1.653559923171997, + "learning_rate": 2.832589397333666e-05, + "loss": 0.0999, + "step": 32183 + }, + { + "epoch": 0.754443643523972, + "grad_norm": 4.949633598327637, + "learning_rate": 2.8320757455639047e-05, + "loss": 0.5168, + "step": 32184 + }, + { + "epoch": 0.7544670850987769, + "grad_norm": 4.798006534576416, + "learning_rate": 2.831562132687502e-05, + "loss": 0.4018, + "step": 32185 + }, + { + "epoch": 0.754490526673582, + "grad_norm": 2.058858871459961, + "learning_rate": 2.8310485587072412e-05, + "loss": 0.1017, + "step": 32186 + }, + { + "epoch": 0.7545139682483869, + "grad_norm": 2.5564284324645996, + "learning_rate": 2.830535023625913e-05, + "loss": 0.1398, + "step": 32187 + }, + { + "epoch": 0.754537409823192, + "grad_norm": 5.818802356719971, + "learning_rate": 2.830021527446306e-05, + "loss": 0.2683, + "step": 32188 + }, + { + "epoch": 0.7545608513979969, + "grad_norm": 1.3159791231155396, + "learning_rate": 2.829508070171204e-05, + "loss": 0.0552, + "step": 32189 + }, + { + "epoch": 0.7545842929728019, + "grad_norm": 2.404865026473999, + "learning_rate": 2.8289946518033927e-05, + "loss": 0.2177, + "step": 32190 + }, + { + "epoch": 0.7546077345476069, + "grad_norm": 2.6681530475616455, + "learning_rate": 2.828481272345659e-05, + "loss": 0.1322, + "step": 32191 + }, + { + "epoch": 0.7546311761224119, + "grad_norm": 1.6124067306518555, + "learning_rate": 2.8279679318007867e-05, + "loss": 0.0844, + "step": 32192 + }, + { + "epoch": 0.7546546176972169, + "grad_norm": 2.3500959873199463, + "learning_rate": 2.8274546301715586e-05, + "loss": 0.1731, + "step": 32193 + }, + { + "epoch": 0.7546780592720219, + "grad_norm": 2.0092005729675293, + "learning_rate": 2.826941367460767e-05, + "loss": 0.1552, + "step": 32194 + }, + { + "epoch": 0.7547015008468269, + "grad_norm": 2.64444899559021, + "learning_rate": 2.8264281436711927e-05, + "loss": 0.2629, + "step": 32195 + }, + { + "epoch": 0.7547249424216319, + "grad_norm": 1.017052173614502, + "learning_rate": 2.825914958805622e-05, + "loss": 0.1064, + "step": 32196 + }, + { + "epoch": 0.7547483839964368, + "grad_norm": 5.841007232666016, + "learning_rate": 2.825401812866838e-05, + "loss": 0.4254, + "step": 32197 + }, + { + "epoch": 0.7547718255712419, + "grad_norm": 0.8732542991638184, + "learning_rate": 2.8248887058576222e-05, + "loss": 0.0773, + "step": 32198 + }, + { + "epoch": 0.7547952671460468, + "grad_norm": 0.8885060548782349, + "learning_rate": 2.824375637780765e-05, + "loss": 0.0835, + "step": 32199 + }, + { + "epoch": 0.7548187087208519, + "grad_norm": 0.4957595765590668, + "learning_rate": 2.823862608639044e-05, + "loss": 0.0291, + "step": 32200 + }, + { + "epoch": 0.7548421502956568, + "grad_norm": 3.205345630645752, + "learning_rate": 2.8233496184352505e-05, + "loss": 0.2132, + "step": 32201 + }, + { + "epoch": 0.7548655918704619, + "grad_norm": 0.765892744064331, + "learning_rate": 2.822836667172163e-05, + "loss": 0.0576, + "step": 32202 + }, + { + "epoch": 0.7548890334452668, + "grad_norm": 0.9903472661972046, + "learning_rate": 2.822323754852565e-05, + "loss": 0.0523, + "step": 32203 + }, + { + "epoch": 0.7549124750200719, + "grad_norm": 0.7914475202560425, + "learning_rate": 2.821810881479241e-05, + "loss": 0.0504, + "step": 32204 + }, + { + "epoch": 0.7549359165948768, + "grad_norm": 2.208160161972046, + "learning_rate": 2.82129804705497e-05, + "loss": 1.5636, + "step": 32205 + }, + { + "epoch": 0.7549593581696818, + "grad_norm": 3.2200980186462402, + "learning_rate": 2.8207852515825406e-05, + "loss": 0.2457, + "step": 32206 + }, + { + "epoch": 0.7549827997444868, + "grad_norm": 2.9488236904144287, + "learning_rate": 2.820272495064732e-05, + "loss": 0.1204, + "step": 32207 + }, + { + "epoch": 0.7550062413192918, + "grad_norm": 1.1595300436019897, + "learning_rate": 2.8197597775043262e-05, + "loss": 0.1297, + "step": 32208 + }, + { + "epoch": 0.7550296828940968, + "grad_norm": 1.9376506805419922, + "learning_rate": 2.8192470989041063e-05, + "loss": 0.8905, + "step": 32209 + }, + { + "epoch": 0.7550531244689018, + "grad_norm": 2.74285888671875, + "learning_rate": 2.8187344592668507e-05, + "loss": 0.1582, + "step": 32210 + }, + { + "epoch": 0.7550765660437068, + "grad_norm": 2.740591287612915, + "learning_rate": 2.8182218585953468e-05, + "loss": 0.309, + "step": 32211 + }, + { + "epoch": 0.7551000076185118, + "grad_norm": 2.0302469730377197, + "learning_rate": 2.8177092968923702e-05, + "loss": 0.1103, + "step": 32212 + }, + { + "epoch": 0.7551234491933168, + "grad_norm": 1.3766006231307983, + "learning_rate": 2.8171967741607077e-05, + "loss": 0.1051, + "step": 32213 + }, + { + "epoch": 0.7551468907681218, + "grad_norm": 2.8189988136291504, + "learning_rate": 2.8166842904031364e-05, + "loss": 0.1669, + "step": 32214 + }, + { + "epoch": 0.7551703323429269, + "grad_norm": 3.66795015335083, + "learning_rate": 2.8161718456224394e-05, + "loss": 0.3414, + "step": 32215 + }, + { + "epoch": 0.7551937739177318, + "grad_norm": 3.317758798599243, + "learning_rate": 2.8156594398213954e-05, + "loss": 0.2008, + "step": 32216 + }, + { + "epoch": 0.7552172154925368, + "grad_norm": 1.4305262565612793, + "learning_rate": 2.8151470730027862e-05, + "loss": 0.125, + "step": 32217 + }, + { + "epoch": 0.7552406570673418, + "grad_norm": 0.578458845615387, + "learning_rate": 2.8146347451693865e-05, + "loss": 0.038, + "step": 32218 + }, + { + "epoch": 0.7552640986421468, + "grad_norm": 0.46147459745407104, + "learning_rate": 2.8141224563239854e-05, + "loss": 0.0269, + "step": 32219 + }, + { + "epoch": 0.7552875402169518, + "grad_norm": 1.9278982877731323, + "learning_rate": 2.813610206469357e-05, + "loss": 1.7061, + "step": 32220 + }, + { + "epoch": 0.7553109817917568, + "grad_norm": 2.0876941680908203, + "learning_rate": 2.8130979956082816e-05, + "loss": 0.1047, + "step": 32221 + }, + { + "epoch": 0.7553344233665618, + "grad_norm": 2.069931745529175, + "learning_rate": 2.8125858237435352e-05, + "loss": 0.9957, + "step": 32222 + }, + { + "epoch": 0.7553578649413668, + "grad_norm": 3.1741445064544678, + "learning_rate": 2.8120736908779045e-05, + "loss": 1.4621, + "step": 32223 + }, + { + "epoch": 0.7553813065161717, + "grad_norm": 1.847754955291748, + "learning_rate": 2.8115615970141627e-05, + "loss": 0.2242, + "step": 32224 + }, + { + "epoch": 0.7554047480909768, + "grad_norm": 0.5909542441368103, + "learning_rate": 2.8110495421550874e-05, + "loss": 0.0653, + "step": 32225 + }, + { + "epoch": 0.7554281896657817, + "grad_norm": 0.9889354109764099, + "learning_rate": 2.810537526303463e-05, + "loss": 0.0578, + "step": 32226 + }, + { + "epoch": 0.7554516312405868, + "grad_norm": 3.370110034942627, + "learning_rate": 2.810025549462063e-05, + "loss": 0.1486, + "step": 32227 + }, + { + "epoch": 0.7554750728153917, + "grad_norm": 4.130216121673584, + "learning_rate": 2.8095136116336672e-05, + "loss": 0.3427, + "step": 32228 + }, + { + "epoch": 0.7554985143901968, + "grad_norm": 2.3281168937683105, + "learning_rate": 2.8090017128210522e-05, + "loss": 0.108, + "step": 32229 + }, + { + "epoch": 0.7555219559650017, + "grad_norm": 1.2344021797180176, + "learning_rate": 2.8084898530269964e-05, + "loss": 0.102, + "step": 32230 + }, + { + "epoch": 0.7555453975398068, + "grad_norm": 2.7066633701324463, + "learning_rate": 2.8079780322542747e-05, + "loss": 0.1372, + "step": 32231 + }, + { + "epoch": 0.7555688391146117, + "grad_norm": 1.4360061883926392, + "learning_rate": 2.807466250505668e-05, + "loss": 0.044, + "step": 32232 + }, + { + "epoch": 0.7555922806894168, + "grad_norm": 1.6380236148834229, + "learning_rate": 2.8069545077839522e-05, + "loss": 0.1599, + "step": 32233 + }, + { + "epoch": 0.7556157222642217, + "grad_norm": 1.8353123664855957, + "learning_rate": 2.806442804091901e-05, + "loss": 0.1388, + "step": 32234 + }, + { + "epoch": 0.7556391638390267, + "grad_norm": 3.6068108081817627, + "learning_rate": 2.8059311394322962e-05, + "loss": 0.2982, + "step": 32235 + }, + { + "epoch": 0.7556626054138317, + "grad_norm": 3.158970832824707, + "learning_rate": 2.805419513807911e-05, + "loss": 0.1874, + "step": 32236 + }, + { + "epoch": 0.7556860469886367, + "grad_norm": 4.278970241546631, + "learning_rate": 2.8049079272215218e-05, + "loss": 0.5408, + "step": 32237 + }, + { + "epoch": 0.7557094885634417, + "grad_norm": 4.449819087982178, + "learning_rate": 2.804396379675902e-05, + "loss": 0.4149, + "step": 32238 + }, + { + "epoch": 0.7557329301382467, + "grad_norm": 3.658407211303711, + "learning_rate": 2.803884871173832e-05, + "loss": 0.1987, + "step": 32239 + }, + { + "epoch": 0.7557563717130517, + "grad_norm": 3.095306873321533, + "learning_rate": 2.8033734017180856e-05, + "loss": 1.6082, + "step": 32240 + }, + { + "epoch": 0.7557798132878567, + "grad_norm": 1.977666974067688, + "learning_rate": 2.802861971311437e-05, + "loss": 0.0841, + "step": 32241 + }, + { + "epoch": 0.7558032548626616, + "grad_norm": 2.8638980388641357, + "learning_rate": 2.8023505799566617e-05, + "loss": 0.2279, + "step": 32242 + }, + { + "epoch": 0.7558266964374667, + "grad_norm": 2.532418966293335, + "learning_rate": 2.8018392276565318e-05, + "loss": 0.1693, + "step": 32243 + }, + { + "epoch": 0.7558501380122716, + "grad_norm": 2.9783735275268555, + "learning_rate": 2.8013279144138283e-05, + "loss": 0.1166, + "step": 32244 + }, + { + "epoch": 0.7558735795870767, + "grad_norm": 0.7166436910629272, + "learning_rate": 2.8008166402313207e-05, + "loss": 0.0465, + "step": 32245 + }, + { + "epoch": 0.7558970211618816, + "grad_norm": 1.2652826309204102, + "learning_rate": 2.800305405111783e-05, + "loss": 0.1283, + "step": 32246 + }, + { + "epoch": 0.7559204627366867, + "grad_norm": 2.2784035205841064, + "learning_rate": 2.799794209057992e-05, + "loss": 0.1142, + "step": 32247 + }, + { + "epoch": 0.7559439043114916, + "grad_norm": 5.275961399078369, + "learning_rate": 2.799283052072721e-05, + "loss": 0.4126, + "step": 32248 + }, + { + "epoch": 0.7559673458862967, + "grad_norm": 1.6052151918411255, + "learning_rate": 2.7987719341587427e-05, + "loss": 0.1898, + "step": 32249 + }, + { + "epoch": 0.7559907874611016, + "grad_norm": 1.528937578201294, + "learning_rate": 2.798260855318827e-05, + "loss": 0.1081, + "step": 32250 + }, + { + "epoch": 0.7560142290359066, + "grad_norm": 2.9674830436706543, + "learning_rate": 2.7977498155557535e-05, + "loss": 0.266, + "step": 32251 + }, + { + "epoch": 0.7560376706107116, + "grad_norm": 3.34702467918396, + "learning_rate": 2.7972388148722916e-05, + "loss": 1.4841, + "step": 32252 + }, + { + "epoch": 0.7560611121855166, + "grad_norm": 3.2780380249023438, + "learning_rate": 2.796727853271215e-05, + "loss": 0.4461, + "step": 32253 + }, + { + "epoch": 0.7560845537603216, + "grad_norm": 2.6366822719573975, + "learning_rate": 2.7962169307552954e-05, + "loss": 1.7581, + "step": 32254 + }, + { + "epoch": 0.7561079953351266, + "grad_norm": 1.3116445541381836, + "learning_rate": 2.7957060473273057e-05, + "loss": 0.0505, + "step": 32255 + }, + { + "epoch": 0.7561314369099316, + "grad_norm": 0.5958566665649414, + "learning_rate": 2.795195202990013e-05, + "loss": 0.011, + "step": 32256 + }, + { + "epoch": 0.7561548784847366, + "grad_norm": 2.3255412578582764, + "learning_rate": 2.794684397746198e-05, + "loss": 0.1282, + "step": 32257 + }, + { + "epoch": 0.7561783200595416, + "grad_norm": 2.7488346099853516, + "learning_rate": 2.7941736315986277e-05, + "loss": 1.5315, + "step": 32258 + }, + { + "epoch": 0.7562017616343466, + "grad_norm": 0.8557714223861694, + "learning_rate": 2.7936629045500718e-05, + "loss": 0.1149, + "step": 32259 + }, + { + "epoch": 0.7562252032091515, + "grad_norm": 0.7989992499351501, + "learning_rate": 2.7931522166033052e-05, + "loss": 0.0175, + "step": 32260 + }, + { + "epoch": 0.7562486447839566, + "grad_norm": 2.7029199600219727, + "learning_rate": 2.7926415677610984e-05, + "loss": 0.1907, + "step": 32261 + }, + { + "epoch": 0.7562720863587615, + "grad_norm": 0.20351628959178925, + "learning_rate": 2.7921309580262202e-05, + "loss": 0.0098, + "step": 32262 + }, + { + "epoch": 0.7562955279335666, + "grad_norm": 3.5184412002563477, + "learning_rate": 2.7916203874014403e-05, + "loss": 0.3435, + "step": 32263 + }, + { + "epoch": 0.7563189695083715, + "grad_norm": 2.1587648391723633, + "learning_rate": 2.7911098558895333e-05, + "loss": 0.1667, + "step": 32264 + }, + { + "epoch": 0.7563424110831766, + "grad_norm": 2.0457210540771484, + "learning_rate": 2.790599363493266e-05, + "loss": 0.1721, + "step": 32265 + }, + { + "epoch": 0.7563658526579815, + "grad_norm": 3.5671346187591553, + "learning_rate": 2.7900889102154104e-05, + "loss": 1.5778, + "step": 32266 + }, + { + "epoch": 0.7563892942327866, + "grad_norm": 1.466219425201416, + "learning_rate": 2.7895784960587344e-05, + "loss": 0.0581, + "step": 32267 + }, + { + "epoch": 0.7564127358075916, + "grad_norm": 2.235572338104248, + "learning_rate": 2.7890681210260085e-05, + "loss": 0.1839, + "step": 32268 + }, + { + "epoch": 0.7564361773823965, + "grad_norm": 2.167774200439453, + "learning_rate": 2.7885577851199983e-05, + "loss": 0.1443, + "step": 32269 + }, + { + "epoch": 0.7564596189572016, + "grad_norm": 4.574032783508301, + "learning_rate": 2.7880474883434805e-05, + "loss": 0.4428, + "step": 32270 + }, + { + "epoch": 0.7564830605320065, + "grad_norm": 0.9292036890983582, + "learning_rate": 2.7875372306992154e-05, + "loss": 0.0667, + "step": 32271 + }, + { + "epoch": 0.7565065021068116, + "grad_norm": 2.875291347503662, + "learning_rate": 2.78702701218998e-05, + "loss": 0.3065, + "step": 32272 + }, + { + "epoch": 0.7565299436816165, + "grad_norm": 2.4761857986450195, + "learning_rate": 2.7865168328185387e-05, + "loss": 0.2071, + "step": 32273 + }, + { + "epoch": 0.7565533852564216, + "grad_norm": 2.3700473308563232, + "learning_rate": 2.7860066925876605e-05, + "loss": 0.1498, + "step": 32274 + }, + { + "epoch": 0.7565768268312265, + "grad_norm": 4.097436904907227, + "learning_rate": 2.785496591500112e-05, + "loss": 0.1903, + "step": 32275 + }, + { + "epoch": 0.7566002684060316, + "grad_norm": 2.462385416030884, + "learning_rate": 2.7849865295586587e-05, + "loss": 0.1606, + "step": 32276 + }, + { + "epoch": 0.7566237099808365, + "grad_norm": 0.7984828948974609, + "learning_rate": 2.784476506766075e-05, + "loss": 0.1025, + "step": 32277 + }, + { + "epoch": 0.7566471515556416, + "grad_norm": 2.747274160385132, + "learning_rate": 2.7839665231251234e-05, + "loss": 0.2664, + "step": 32278 + }, + { + "epoch": 0.7566705931304465, + "grad_norm": 0.7300719618797302, + "learning_rate": 2.7834565786385734e-05, + "loss": 0.0873, + "step": 32279 + }, + { + "epoch": 0.7566940347052515, + "grad_norm": 1.547783613204956, + "learning_rate": 2.7829466733091914e-05, + "loss": 0.0761, + "step": 32280 + }, + { + "epoch": 0.7567174762800565, + "grad_norm": 4.062858581542969, + "learning_rate": 2.78243680713974e-05, + "loss": 0.2023, + "step": 32281 + }, + { + "epoch": 0.7567409178548615, + "grad_norm": 2.846986770629883, + "learning_rate": 2.7819269801329917e-05, + "loss": 0.1937, + "step": 32282 + }, + { + "epoch": 0.7567643594296665, + "grad_norm": 3.4303128719329834, + "learning_rate": 2.7814171922917087e-05, + "loss": 1.3316, + "step": 32283 + }, + { + "epoch": 0.7567878010044715, + "grad_norm": 0.4324769973754883, + "learning_rate": 2.780907443618661e-05, + "loss": 0.0188, + "step": 32284 + }, + { + "epoch": 0.7568112425792765, + "grad_norm": 1.3271859884262085, + "learning_rate": 2.7803977341166132e-05, + "loss": 0.0857, + "step": 32285 + }, + { + "epoch": 0.7568346841540815, + "grad_norm": 0.858884334564209, + "learning_rate": 2.7798880637883305e-05, + "loss": 0.0555, + "step": 32286 + }, + { + "epoch": 0.7568581257288864, + "grad_norm": 3.794264316558838, + "learning_rate": 2.779378432636578e-05, + "loss": 0.4166, + "step": 32287 + }, + { + "epoch": 0.7568815673036915, + "grad_norm": 0.50169438123703, + "learning_rate": 2.7788688406641183e-05, + "loss": 0.0336, + "step": 32288 + }, + { + "epoch": 0.7569050088784964, + "grad_norm": 1.8922632932662964, + "learning_rate": 2.778359287873722e-05, + "loss": 0.8885, + "step": 32289 + }, + { + "epoch": 0.7569284504533015, + "grad_norm": 3.299694299697876, + "learning_rate": 2.7778497742681507e-05, + "loss": 0.3216, + "step": 32290 + }, + { + "epoch": 0.7569518920281064, + "grad_norm": 1.1411241292953491, + "learning_rate": 2.7773402998501706e-05, + "loss": 0.0435, + "step": 32291 + }, + { + "epoch": 0.7569753336029115, + "grad_norm": 1.052157998085022, + "learning_rate": 2.776830864622545e-05, + "loss": 0.0841, + "step": 32292 + }, + { + "epoch": 0.7569987751777164, + "grad_norm": 1.1217373609542847, + "learning_rate": 2.776321468588038e-05, + "loss": 0.1121, + "step": 32293 + }, + { + "epoch": 0.7570222167525215, + "grad_norm": 2.2046189308166504, + "learning_rate": 2.7758121117494107e-05, + "loss": 0.1124, + "step": 32294 + }, + { + "epoch": 0.7570456583273264, + "grad_norm": 0.7225340008735657, + "learning_rate": 2.77530279410943e-05, + "loss": 0.0682, + "step": 32295 + }, + { + "epoch": 0.7570690999021314, + "grad_norm": 3.0349440574645996, + "learning_rate": 2.7747935156708625e-05, + "loss": 0.2381, + "step": 32296 + }, + { + "epoch": 0.7570925414769364, + "grad_norm": 2.438133478164673, + "learning_rate": 2.7742842764364686e-05, + "loss": 0.1682, + "step": 32297 + }, + { + "epoch": 0.7571159830517414, + "grad_norm": 0.522054135799408, + "learning_rate": 2.773775076409011e-05, + "loss": 0.0202, + "step": 32298 + }, + { + "epoch": 0.7571394246265464, + "grad_norm": 0.4110153615474701, + "learning_rate": 2.7732659155912532e-05, + "loss": 0.0396, + "step": 32299 + }, + { + "epoch": 0.7571628662013514, + "grad_norm": 1.6833969354629517, + "learning_rate": 2.7727567939859578e-05, + "loss": 0.065, + "step": 32300 + }, + { + "epoch": 0.7571863077761564, + "grad_norm": 2.375300884246826, + "learning_rate": 2.7722477115958846e-05, + "loss": 0.1014, + "step": 32301 + }, + { + "epoch": 0.7572097493509614, + "grad_norm": 1.9155091047286987, + "learning_rate": 2.771738668423801e-05, + "loss": 0.111, + "step": 32302 + }, + { + "epoch": 0.7572331909257664, + "grad_norm": 2.078486442565918, + "learning_rate": 2.7712296644724665e-05, + "loss": 0.1683, + "step": 32303 + }, + { + "epoch": 0.7572566325005714, + "grad_norm": 1.904770851135254, + "learning_rate": 2.7707206997446443e-05, + "loss": 0.0702, + "step": 32304 + }, + { + "epoch": 0.7572800740753763, + "grad_norm": 1.9999301433563232, + "learning_rate": 2.7702117742430945e-05, + "loss": 0.0817, + "step": 32305 + }, + { + "epoch": 0.7573035156501814, + "grad_norm": 0.818901002407074, + "learning_rate": 2.769702887970578e-05, + "loss": 0.065, + "step": 32306 + }, + { + "epoch": 0.7573269572249863, + "grad_norm": 2.780923843383789, + "learning_rate": 2.7691940409298546e-05, + "loss": 0.1699, + "step": 32307 + }, + { + "epoch": 0.7573503987997914, + "grad_norm": 1.4368000030517578, + "learning_rate": 2.768685233123687e-05, + "loss": 0.1586, + "step": 32308 + }, + { + "epoch": 0.7573738403745963, + "grad_norm": 3.4482967853546143, + "learning_rate": 2.76817646455484e-05, + "loss": 0.2984, + "step": 32309 + }, + { + "epoch": 0.7573972819494014, + "grad_norm": 1.1190310716629028, + "learning_rate": 2.7676677352260717e-05, + "loss": 0.0674, + "step": 32310 + }, + { + "epoch": 0.7574207235242063, + "grad_norm": 1.866999864578247, + "learning_rate": 2.767159045140141e-05, + "loss": 0.2141, + "step": 32311 + }, + { + "epoch": 0.7574441650990114, + "grad_norm": 0.692392110824585, + "learning_rate": 2.766650394299808e-05, + "loss": 0.0707, + "step": 32312 + }, + { + "epoch": 0.7574676066738163, + "grad_norm": 3.1428518295288086, + "learning_rate": 2.7661417827078307e-05, + "loss": 0.1531, + "step": 32313 + }, + { + "epoch": 0.7574910482486213, + "grad_norm": 3.5108633041381836, + "learning_rate": 2.7656332103669747e-05, + "loss": 0.2963, + "step": 32314 + }, + { + "epoch": 0.7575144898234263, + "grad_norm": 1.1097174882888794, + "learning_rate": 2.7651246772799965e-05, + "loss": 0.0868, + "step": 32315 + }, + { + "epoch": 0.7575379313982313, + "grad_norm": 2.5855252742767334, + "learning_rate": 2.7646161834496554e-05, + "loss": 1.5228, + "step": 32316 + }, + { + "epoch": 0.7575613729730363, + "grad_norm": 1.1039342880249023, + "learning_rate": 2.7641077288787097e-05, + "loss": 0.0954, + "step": 32317 + }, + { + "epoch": 0.7575848145478413, + "grad_norm": 1.5952250957489014, + "learning_rate": 2.763599313569919e-05, + "loss": 0.1111, + "step": 32318 + }, + { + "epoch": 0.7576082561226464, + "grad_norm": 2.3026039600372314, + "learning_rate": 2.7630909375260395e-05, + "loss": 0.1516, + "step": 32319 + }, + { + "epoch": 0.7576316976974513, + "grad_norm": 0.969617486000061, + "learning_rate": 2.7625826007498312e-05, + "loss": 0.0893, + "step": 32320 + }, + { + "epoch": 0.7576551392722564, + "grad_norm": 1.8062529563903809, + "learning_rate": 2.7620743032440565e-05, + "loss": 1.3337, + "step": 32321 + }, + { + "epoch": 0.7576785808470613, + "grad_norm": 2.2970430850982666, + "learning_rate": 2.76156604501147e-05, + "loss": 0.7935, + "step": 32322 + }, + { + "epoch": 0.7577020224218664, + "grad_norm": 0.2579580247402191, + "learning_rate": 2.7610578260548303e-05, + "loss": 0.0233, + "step": 32323 + }, + { + "epoch": 0.7577254639966713, + "grad_norm": 2.9897658824920654, + "learning_rate": 2.7605496463768944e-05, + "loss": 0.2042, + "step": 32324 + }, + { + "epoch": 0.7577489055714763, + "grad_norm": 4.765656471252441, + "learning_rate": 2.7600415059804187e-05, + "loss": 0.4834, + "step": 32325 + }, + { + "epoch": 0.7577723471462813, + "grad_norm": 0.23951733112335205, + "learning_rate": 2.759533404868159e-05, + "loss": 0.018, + "step": 32326 + }, + { + "epoch": 0.7577957887210863, + "grad_norm": 1.4042983055114746, + "learning_rate": 2.759025343042878e-05, + "loss": 0.1071, + "step": 32327 + }, + { + "epoch": 0.7578192302958913, + "grad_norm": 3.451211452484131, + "learning_rate": 2.7585173205073288e-05, + "loss": 1.657, + "step": 32328 + }, + { + "epoch": 0.7578426718706963, + "grad_norm": 2.6452088356018066, + "learning_rate": 2.758009337264268e-05, + "loss": 0.2924, + "step": 32329 + }, + { + "epoch": 0.7578661134455013, + "grad_norm": 2.299907922744751, + "learning_rate": 2.7575013933164517e-05, + "loss": 0.19, + "step": 32330 + }, + { + "epoch": 0.7578895550203063, + "grad_norm": 0.8721822500228882, + "learning_rate": 2.7569934886666372e-05, + "loss": 0.0681, + "step": 32331 + }, + { + "epoch": 0.7579129965951112, + "grad_norm": 2.4650566577911377, + "learning_rate": 2.7564856233175764e-05, + "loss": 0.123, + "step": 32332 + }, + { + "epoch": 0.7579364381699163, + "grad_norm": 2.8062703609466553, + "learning_rate": 2.755977797272029e-05, + "loss": 0.2175, + "step": 32333 + }, + { + "epoch": 0.7579598797447212, + "grad_norm": 1.4742639064788818, + "learning_rate": 2.7554700105327513e-05, + "loss": 0.0646, + "step": 32334 + }, + { + "epoch": 0.7579833213195263, + "grad_norm": 1.7998056411743164, + "learning_rate": 2.7549622631024985e-05, + "loss": 0.1049, + "step": 32335 + }, + { + "epoch": 0.7580067628943312, + "grad_norm": 2.860603094100952, + "learning_rate": 2.754454554984024e-05, + "loss": 0.2778, + "step": 32336 + }, + { + "epoch": 0.7580302044691363, + "grad_norm": 4.180389404296875, + "learning_rate": 2.753946886180082e-05, + "loss": 0.4596, + "step": 32337 + }, + { + "epoch": 0.7580536460439412, + "grad_norm": 2.588806390762329, + "learning_rate": 2.7534392566934286e-05, + "loss": 0.3292, + "step": 32338 + }, + { + "epoch": 0.7580770876187463, + "grad_norm": 2.769655466079712, + "learning_rate": 2.7529316665268135e-05, + "loss": 0.189, + "step": 32339 + }, + { + "epoch": 0.7581005291935512, + "grad_norm": 3.4395453929901123, + "learning_rate": 2.752424115682999e-05, + "loss": 0.1925, + "step": 32340 + }, + { + "epoch": 0.7581239707683562, + "grad_norm": 1.7349863052368164, + "learning_rate": 2.751916604164735e-05, + "loss": 0.1343, + "step": 32341 + }, + { + "epoch": 0.7581474123431612, + "grad_norm": 2.9177961349487305, + "learning_rate": 2.7514091319747748e-05, + "loss": 0.2196, + "step": 32342 + }, + { + "epoch": 0.7581708539179662, + "grad_norm": 3.0716702938079834, + "learning_rate": 2.750901699115873e-05, + "loss": 0.1461, + "step": 32343 + }, + { + "epoch": 0.7581942954927712, + "grad_norm": 2.9527995586395264, + "learning_rate": 2.7503943055907787e-05, + "loss": 0.187, + "step": 32344 + }, + { + "epoch": 0.7582177370675762, + "grad_norm": 2.0739848613739014, + "learning_rate": 2.7498869514022517e-05, + "loss": 0.1858, + "step": 32345 + }, + { + "epoch": 0.7582411786423812, + "grad_norm": 2.3693296909332275, + "learning_rate": 2.74937963655304e-05, + "loss": 0.1772, + "step": 32346 + }, + { + "epoch": 0.7582646202171862, + "grad_norm": 4.06795072555542, + "learning_rate": 2.748872361045901e-05, + "loss": 0.268, + "step": 32347 + }, + { + "epoch": 0.7582880617919912, + "grad_norm": 2.573169231414795, + "learning_rate": 2.748365124883584e-05, + "loss": 1.84, + "step": 32348 + }, + { + "epoch": 0.7583115033667962, + "grad_norm": 3.2115707397460938, + "learning_rate": 2.747857928068842e-05, + "loss": 0.3501, + "step": 32349 + }, + { + "epoch": 0.7583349449416011, + "grad_norm": 1.9646661281585693, + "learning_rate": 2.7473507706044265e-05, + "loss": 0.1151, + "step": 32350 + }, + { + "epoch": 0.7583583865164062, + "grad_norm": 0.7961140275001526, + "learning_rate": 2.7468436524930862e-05, + "loss": 0.0374, + "step": 32351 + }, + { + "epoch": 0.7583818280912111, + "grad_norm": 0.5748711824417114, + "learning_rate": 2.74633657373758e-05, + "loss": 0.0314, + "step": 32352 + }, + { + "epoch": 0.7584052696660162, + "grad_norm": 2.732130289077759, + "learning_rate": 2.7458295343406548e-05, + "loss": 0.2962, + "step": 32353 + }, + { + "epoch": 0.7584287112408211, + "grad_norm": 3.138777256011963, + "learning_rate": 2.7453225343050636e-05, + "loss": 0.2838, + "step": 32354 + }, + { + "epoch": 0.7584521528156262, + "grad_norm": 1.5251802206039429, + "learning_rate": 2.7448155736335556e-05, + "loss": 1.3623, + "step": 32355 + }, + { + "epoch": 0.7584755943904311, + "grad_norm": 3.5209078788757324, + "learning_rate": 2.7443086523288796e-05, + "loss": 0.1509, + "step": 32356 + }, + { + "epoch": 0.7584990359652362, + "grad_norm": 0.8031435012817383, + "learning_rate": 2.7438017703937912e-05, + "loss": 0.0487, + "step": 32357 + }, + { + "epoch": 0.7585224775400411, + "grad_norm": 0.9515165686607361, + "learning_rate": 2.743294927831036e-05, + "loss": 0.1226, + "step": 32358 + }, + { + "epoch": 0.7585459191148461, + "grad_norm": 1.6263567209243774, + "learning_rate": 2.7427881246433696e-05, + "loss": 0.0703, + "step": 32359 + }, + { + "epoch": 0.7585693606896511, + "grad_norm": 2.2300620079040527, + "learning_rate": 2.7422813608335383e-05, + "loss": 1.4854, + "step": 32360 + }, + { + "epoch": 0.7585928022644561, + "grad_norm": 2.79011869430542, + "learning_rate": 2.741774636404293e-05, + "loss": 0.2382, + "step": 32361 + }, + { + "epoch": 0.7586162438392611, + "grad_norm": 1.0175607204437256, + "learning_rate": 2.7412679513583827e-05, + "loss": 0.1243, + "step": 32362 + }, + { + "epoch": 0.7586396854140661, + "grad_norm": 1.4209030866622925, + "learning_rate": 2.7407613056985558e-05, + "loss": 0.0614, + "step": 32363 + }, + { + "epoch": 0.7586631269888711, + "grad_norm": 3.5248630046844482, + "learning_rate": 2.74025469942756e-05, + "loss": 0.1812, + "step": 32364 + }, + { + "epoch": 0.7586865685636761, + "grad_norm": 3.53629207611084, + "learning_rate": 2.7397481325481488e-05, + "loss": 0.1696, + "step": 32365 + }, + { + "epoch": 0.758710010138481, + "grad_norm": 2.0951077938079834, + "learning_rate": 2.7392416050630686e-05, + "loss": 0.1132, + "step": 32366 + }, + { + "epoch": 0.7587334517132861, + "grad_norm": 1.403381586074829, + "learning_rate": 2.7387351169750685e-05, + "loss": 0.0313, + "step": 32367 + }, + { + "epoch": 0.758756893288091, + "grad_norm": 2.7035982608795166, + "learning_rate": 2.7382286682868918e-05, + "loss": 0.2158, + "step": 32368 + }, + { + "epoch": 0.7587803348628961, + "grad_norm": 3.0350358486175537, + "learning_rate": 2.737722259001293e-05, + "loss": 0.3489, + "step": 32369 + }, + { + "epoch": 0.7588037764377011, + "grad_norm": 2.627880573272705, + "learning_rate": 2.7372158891210188e-05, + "loss": 0.2988, + "step": 32370 + }, + { + "epoch": 0.7588272180125061, + "grad_norm": 3.612783193588257, + "learning_rate": 2.736709558648811e-05, + "loss": 0.2997, + "step": 32371 + }, + { + "epoch": 0.7588506595873111, + "grad_norm": 2.1311638355255127, + "learning_rate": 2.736203267587425e-05, + "loss": 0.1245, + "step": 32372 + }, + { + "epoch": 0.7588741011621161, + "grad_norm": 0.8178113102912903, + "learning_rate": 2.7356970159396046e-05, + "loss": 0.059, + "step": 32373 + }, + { + "epoch": 0.7588975427369211, + "grad_norm": 5.488483428955078, + "learning_rate": 2.7351908037080965e-05, + "loss": 0.3761, + "step": 32374 + }, + { + "epoch": 0.758920984311726, + "grad_norm": 0.301570862531662, + "learning_rate": 2.7346846308956477e-05, + "loss": 0.0165, + "step": 32375 + }, + { + "epoch": 0.7589444258865311, + "grad_norm": 2.9239249229431152, + "learning_rate": 2.7341784975050033e-05, + "loss": 0.1848, + "step": 32376 + }, + { + "epoch": 0.758967867461336, + "grad_norm": 2.9679386615753174, + "learning_rate": 2.733672403538908e-05, + "loss": 0.2575, + "step": 32377 + }, + { + "epoch": 0.7589913090361411, + "grad_norm": 0.8866702318191528, + "learning_rate": 2.7331663490001137e-05, + "loss": 0.0888, + "step": 32378 + }, + { + "epoch": 0.759014750610946, + "grad_norm": 2.750623941421509, + "learning_rate": 2.7326603338913625e-05, + "loss": 0.152, + "step": 32379 + }, + { + "epoch": 0.7590381921857511, + "grad_norm": 2.4258902072906494, + "learning_rate": 2.7321543582153985e-05, + "loss": 0.2797, + "step": 32380 + }, + { + "epoch": 0.759061633760556, + "grad_norm": 3.086348533630371, + "learning_rate": 2.7316484219749716e-05, + "loss": 0.2093, + "step": 32381 + }, + { + "epoch": 0.7590850753353611, + "grad_norm": 2.719395160675049, + "learning_rate": 2.731142525172826e-05, + "loss": 0.2037, + "step": 32382 + }, + { + "epoch": 0.759108516910166, + "grad_norm": 2.008364200592041, + "learning_rate": 2.7306366678117044e-05, + "loss": 0.1819, + "step": 32383 + }, + { + "epoch": 0.7591319584849711, + "grad_norm": 3.818944215774536, + "learning_rate": 2.73013084989435e-05, + "loss": 0.3811, + "step": 32384 + }, + { + "epoch": 0.759155400059776, + "grad_norm": 2.018275022506714, + "learning_rate": 2.7296250714235127e-05, + "loss": 1.5356, + "step": 32385 + }, + { + "epoch": 0.759178841634581, + "grad_norm": 1.445910930633545, + "learning_rate": 2.729119332401935e-05, + "loss": 0.0788, + "step": 32386 + }, + { + "epoch": 0.759202283209386, + "grad_norm": 2.696453809738159, + "learning_rate": 2.7286136328323585e-05, + "loss": 0.1331, + "step": 32387 + }, + { + "epoch": 0.759225724784191, + "grad_norm": 3.1225290298461914, + "learning_rate": 2.7281079727175306e-05, + "loss": 0.2655, + "step": 32388 + }, + { + "epoch": 0.759249166358996, + "grad_norm": 2.5529332160949707, + "learning_rate": 2.727602352060189e-05, + "loss": 0.2104, + "step": 32389 + }, + { + "epoch": 0.759272607933801, + "grad_norm": 2.5937981605529785, + "learning_rate": 2.7270967708630857e-05, + "loss": 0.0796, + "step": 32390 + }, + { + "epoch": 0.759296049508606, + "grad_norm": 1.8799595832824707, + "learning_rate": 2.7265912291289598e-05, + "loss": 0.9956, + "step": 32391 + }, + { + "epoch": 0.759319491083411, + "grad_norm": 3.6014556884765625, + "learning_rate": 2.72608572686055e-05, + "loss": 0.2414, + "step": 32392 + }, + { + "epoch": 0.759342932658216, + "grad_norm": 0.8585211038589478, + "learning_rate": 2.7255802640606078e-05, + "loss": 0.0731, + "step": 32393 + }, + { + "epoch": 0.759366374233021, + "grad_norm": 0.6909350752830505, + "learning_rate": 2.7250748407318715e-05, + "loss": 0.0526, + "step": 32394 + }, + { + "epoch": 0.7593898158078259, + "grad_norm": 1.1044418811798096, + "learning_rate": 2.724569456877084e-05, + "loss": 0.0737, + "step": 32395 + }, + { + "epoch": 0.759413257382631, + "grad_norm": 2.6644020080566406, + "learning_rate": 2.724064112498983e-05, + "loss": 1.3117, + "step": 32396 + }, + { + "epoch": 0.7594366989574359, + "grad_norm": 3.3851218223571777, + "learning_rate": 2.7235588076003183e-05, + "loss": 1.5619, + "step": 32397 + }, + { + "epoch": 0.759460140532241, + "grad_norm": 2.738887310028076, + "learning_rate": 2.7230535421838287e-05, + "loss": 0.2155, + "step": 32398 + }, + { + "epoch": 0.7594835821070459, + "grad_norm": 2.279615640640259, + "learning_rate": 2.722548316252255e-05, + "loss": 0.2627, + "step": 32399 + }, + { + "epoch": 0.759507023681851, + "grad_norm": 4.498619556427002, + "learning_rate": 2.7220431298083382e-05, + "loss": 0.3623, + "step": 32400 + }, + { + "epoch": 0.7595304652566559, + "grad_norm": 1.8414767980575562, + "learning_rate": 2.7215379828548204e-05, + "loss": 0.1639, + "step": 32401 + }, + { + "epoch": 0.759553906831461, + "grad_norm": 2.831667423248291, + "learning_rate": 2.721032875394439e-05, + "loss": 0.3114, + "step": 32402 + }, + { + "epoch": 0.7595773484062659, + "grad_norm": 2.7830352783203125, + "learning_rate": 2.72052780742994e-05, + "loss": 0.1953, + "step": 32403 + }, + { + "epoch": 0.759600789981071, + "grad_norm": 4.290798187255859, + "learning_rate": 2.7200227789640632e-05, + "loss": 0.4056, + "step": 32404 + }, + { + "epoch": 0.7596242315558759, + "grad_norm": 2.448869466781616, + "learning_rate": 2.7195177899995438e-05, + "loss": 0.2121, + "step": 32405 + }, + { + "epoch": 0.7596476731306809, + "grad_norm": 0.24168036878108978, + "learning_rate": 2.7190128405391292e-05, + "loss": 0.024, + "step": 32406 + }, + { + "epoch": 0.7596711147054859, + "grad_norm": 2.9131269454956055, + "learning_rate": 2.7185079305855555e-05, + "loss": 1.327, + "step": 32407 + }, + { + "epoch": 0.7596945562802909, + "grad_norm": 4.040773868560791, + "learning_rate": 2.7180030601415618e-05, + "loss": 0.2687, + "step": 32408 + }, + { + "epoch": 0.7597179978550959, + "grad_norm": 2.135587215423584, + "learning_rate": 2.7174982292098862e-05, + "loss": 0.1588, + "step": 32409 + }, + { + "epoch": 0.7597414394299009, + "grad_norm": 2.5650808811187744, + "learning_rate": 2.7169934377932716e-05, + "loss": 0.1606, + "step": 32410 + }, + { + "epoch": 0.7597648810047058, + "grad_norm": 2.77420973777771, + "learning_rate": 2.716488685894457e-05, + "loss": 0.2001, + "step": 32411 + }, + { + "epoch": 0.7597883225795109, + "grad_norm": 1.5576117038726807, + "learning_rate": 2.7159839735161786e-05, + "loss": 0.217, + "step": 32412 + }, + { + "epoch": 0.7598117641543158, + "grad_norm": 0.5781150460243225, + "learning_rate": 2.7154793006611757e-05, + "loss": 0.0404, + "step": 32413 + }, + { + "epoch": 0.7598352057291209, + "grad_norm": 3.6039528846740723, + "learning_rate": 2.7149746673321873e-05, + "loss": 0.2866, + "step": 32414 + }, + { + "epoch": 0.7598586473039258, + "grad_norm": 0.9410005807876587, + "learning_rate": 2.7144700735319484e-05, + "loss": 0.057, + "step": 32415 + }, + { + "epoch": 0.7598820888787309, + "grad_norm": 2.6806013584136963, + "learning_rate": 2.713965519263203e-05, + "loss": 0.3335, + "step": 32416 + }, + { + "epoch": 0.7599055304535358, + "grad_norm": 1.8907935619354248, + "learning_rate": 2.7134610045286823e-05, + "loss": 1.2259, + "step": 32417 + }, + { + "epoch": 0.7599289720283409, + "grad_norm": 2.6552109718322754, + "learning_rate": 2.71295652933113e-05, + "loss": 0.2956, + "step": 32418 + }, + { + "epoch": 0.7599524136031458, + "grad_norm": 1.0968910455703735, + "learning_rate": 2.71245209367328e-05, + "loss": 0.1156, + "step": 32419 + }, + { + "epoch": 0.7599758551779509, + "grad_norm": 5.18863582611084, + "learning_rate": 2.7119476975578706e-05, + "loss": 0.5357, + "step": 32420 + }, + { + "epoch": 0.7599992967527559, + "grad_norm": 1.0669912099838257, + "learning_rate": 2.7114433409876383e-05, + "loss": 0.0954, + "step": 32421 + }, + { + "epoch": 0.7600227383275608, + "grad_norm": 2.088162899017334, + "learning_rate": 2.7109390239653154e-05, + "loss": 0.1234, + "step": 32422 + }, + { + "epoch": 0.7600461799023659, + "grad_norm": 0.9799680113792419, + "learning_rate": 2.7104347464936452e-05, + "loss": 0.0797, + "step": 32423 + }, + { + "epoch": 0.7600696214771708, + "grad_norm": 2.0459158420562744, + "learning_rate": 2.7099305085753612e-05, + "loss": 0.2851, + "step": 32424 + }, + { + "epoch": 0.7600930630519759, + "grad_norm": 1.8405343294143677, + "learning_rate": 2.7094263102132002e-05, + "loss": 0.1009, + "step": 32425 + }, + { + "epoch": 0.7601165046267808, + "grad_norm": 0.7798348665237427, + "learning_rate": 2.7089221514098962e-05, + "loss": 0.035, + "step": 32426 + }, + { + "epoch": 0.7601399462015859, + "grad_norm": 2.7788257598876953, + "learning_rate": 2.708418032168183e-05, + "loss": 0.3015, + "step": 32427 + }, + { + "epoch": 0.7601633877763908, + "grad_norm": 0.9447530508041382, + "learning_rate": 2.7079139524908004e-05, + "loss": 0.1135, + "step": 32428 + }, + { + "epoch": 0.7601868293511959, + "grad_norm": 3.1474685668945312, + "learning_rate": 2.7074099123804798e-05, + "loss": 0.194, + "step": 32429 + }, + { + "epoch": 0.7602102709260008, + "grad_norm": 4.058638572692871, + "learning_rate": 2.706905911839961e-05, + "loss": 0.283, + "step": 32430 + }, + { + "epoch": 0.7602337125008058, + "grad_norm": 0.764118492603302, + "learning_rate": 2.706401950871975e-05, + "loss": 0.0447, + "step": 32431 + }, + { + "epoch": 0.7602571540756108, + "grad_norm": 2.8046481609344482, + "learning_rate": 2.705898029479258e-05, + "loss": 0.3501, + "step": 32432 + }, + { + "epoch": 0.7602805956504158, + "grad_norm": 1.4092954397201538, + "learning_rate": 2.705394147664544e-05, + "loss": 0.1048, + "step": 32433 + }, + { + "epoch": 0.7603040372252208, + "grad_norm": 0.5167406797409058, + "learning_rate": 2.704890305430562e-05, + "loss": 0.0501, + "step": 32434 + }, + { + "epoch": 0.7603274788000258, + "grad_norm": 2.0236592292785645, + "learning_rate": 2.704386502780054e-05, + "loss": 0.1097, + "step": 32435 + }, + { + "epoch": 0.7603509203748308, + "grad_norm": 1.7318229675292969, + "learning_rate": 2.7038827397157496e-05, + "loss": 0.164, + "step": 32436 + }, + { + "epoch": 0.7603743619496358, + "grad_norm": 2.883629083633423, + "learning_rate": 2.7033790162403837e-05, + "loss": 0.2336, + "step": 32437 + }, + { + "epoch": 0.7603978035244408, + "grad_norm": 2.586796998977661, + "learning_rate": 2.702875332356687e-05, + "loss": 0.124, + "step": 32438 + }, + { + "epoch": 0.7604212450992458, + "grad_norm": 2.3945071697235107, + "learning_rate": 2.7023716880673942e-05, + "loss": 0.2486, + "step": 32439 + }, + { + "epoch": 0.7604446866740507, + "grad_norm": 1.2215847969055176, + "learning_rate": 2.7018680833752353e-05, + "loss": 0.1, + "step": 32440 + }, + { + "epoch": 0.7604681282488558, + "grad_norm": 2.398629665374756, + "learning_rate": 2.701364518282946e-05, + "loss": 0.1338, + "step": 32441 + }, + { + "epoch": 0.7604915698236607, + "grad_norm": 1.1561228036880493, + "learning_rate": 2.7008609927932605e-05, + "loss": 0.0941, + "step": 32442 + }, + { + "epoch": 0.7605150113984658, + "grad_norm": 1.1305509805679321, + "learning_rate": 2.700357506908908e-05, + "loss": 0.6904, + "step": 32443 + }, + { + "epoch": 0.7605384529732707, + "grad_norm": 2.6918368339538574, + "learning_rate": 2.6998540606326206e-05, + "loss": 0.123, + "step": 32444 + }, + { + "epoch": 0.7605618945480758, + "grad_norm": 1.589538812637329, + "learning_rate": 2.699350653967131e-05, + "loss": 1.2323, + "step": 32445 + }, + { + "epoch": 0.7605853361228807, + "grad_norm": 5.346999645233154, + "learning_rate": 2.6988472869151703e-05, + "loss": 0.4065, + "step": 32446 + }, + { + "epoch": 0.7606087776976858, + "grad_norm": 2.1366896629333496, + "learning_rate": 2.6983439594794657e-05, + "loss": 0.0775, + "step": 32447 + }, + { + "epoch": 0.7606322192724907, + "grad_norm": 1.7684683799743652, + "learning_rate": 2.697840671662756e-05, + "loss": 1.0517, + "step": 32448 + }, + { + "epoch": 0.7606556608472957, + "grad_norm": 0.4884110689163208, + "learning_rate": 2.697337423467767e-05, + "loss": 0.0206, + "step": 32449 + }, + { + "epoch": 0.7606791024221007, + "grad_norm": 2.5996575355529785, + "learning_rate": 2.6968342148972315e-05, + "loss": 0.1705, + "step": 32450 + }, + { + "epoch": 0.7607025439969057, + "grad_norm": 2.875117063522339, + "learning_rate": 2.6963310459538783e-05, + "loss": 0.2755, + "step": 32451 + }, + { + "epoch": 0.7607259855717107, + "grad_norm": 2.316528797149658, + "learning_rate": 2.6958279166404388e-05, + "loss": 0.1246, + "step": 32452 + }, + { + "epoch": 0.7607494271465157, + "grad_norm": 2.2200355529785156, + "learning_rate": 2.6953248269596387e-05, + "loss": 0.2637, + "step": 32453 + }, + { + "epoch": 0.7607728687213207, + "grad_norm": 1.8283264636993408, + "learning_rate": 2.6948217769142126e-05, + "loss": 0.1505, + "step": 32454 + }, + { + "epoch": 0.7607963102961257, + "grad_norm": 1.7643214464187622, + "learning_rate": 2.6943187665068902e-05, + "loss": 0.0906, + "step": 32455 + }, + { + "epoch": 0.7608197518709306, + "grad_norm": 5.023301601409912, + "learning_rate": 2.6938157957404008e-05, + "loss": 0.4091, + "step": 32456 + }, + { + "epoch": 0.7608431934457357, + "grad_norm": 2.240105390548706, + "learning_rate": 2.693312864617472e-05, + "loss": 1.9956, + "step": 32457 + }, + { + "epoch": 0.7608666350205406, + "grad_norm": 3.0434823036193848, + "learning_rate": 2.692809973140834e-05, + "loss": 0.2052, + "step": 32458 + }, + { + "epoch": 0.7608900765953457, + "grad_norm": 1.9997042417526245, + "learning_rate": 2.6923071213132135e-05, + "loss": 0.2059, + "step": 32459 + }, + { + "epoch": 0.7609135181701506, + "grad_norm": 1.299408197402954, + "learning_rate": 2.6918043091373367e-05, + "loss": 0.1037, + "step": 32460 + }, + { + "epoch": 0.7609369597449557, + "grad_norm": 0.6849254369735718, + "learning_rate": 2.691301536615939e-05, + "loss": 0.0821, + "step": 32461 + }, + { + "epoch": 0.7609604013197606, + "grad_norm": 3.5832881927490234, + "learning_rate": 2.6907988037517438e-05, + "loss": 0.4043, + "step": 32462 + }, + { + "epoch": 0.7609838428945657, + "grad_norm": 2.364449977874756, + "learning_rate": 2.6902961105474812e-05, + "loss": 0.1452, + "step": 32463 + }, + { + "epoch": 0.7610072844693706, + "grad_norm": 0.5834431052207947, + "learning_rate": 2.689793457005876e-05, + "loss": 0.0366, + "step": 32464 + }, + { + "epoch": 0.7610307260441757, + "grad_norm": 2.6904208660125732, + "learning_rate": 2.689290843129655e-05, + "loss": 0.2589, + "step": 32465 + }, + { + "epoch": 0.7610541676189806, + "grad_norm": 2.0304317474365234, + "learning_rate": 2.6887882689215505e-05, + "loss": 0.1123, + "step": 32466 + }, + { + "epoch": 0.7610776091937856, + "grad_norm": 1.3044695854187012, + "learning_rate": 2.6882857343842825e-05, + "loss": 0.1276, + "step": 32467 + }, + { + "epoch": 0.7611010507685906, + "grad_norm": 2.651782512664795, + "learning_rate": 2.6877832395205847e-05, + "loss": 0.2352, + "step": 32468 + }, + { + "epoch": 0.7611244923433956, + "grad_norm": 1.892966628074646, + "learning_rate": 2.6872807843331817e-05, + "loss": 0.2052, + "step": 32469 + }, + { + "epoch": 0.7611479339182006, + "grad_norm": 3.381682872772217, + "learning_rate": 2.6867783688247982e-05, + "loss": 0.2977, + "step": 32470 + }, + { + "epoch": 0.7611713754930056, + "grad_norm": 1.0071262121200562, + "learning_rate": 2.6862759929981607e-05, + "loss": 0.0694, + "step": 32471 + }, + { + "epoch": 0.7611948170678107, + "grad_norm": 4.9020233154296875, + "learning_rate": 2.6857736568559933e-05, + "loss": 0.4781, + "step": 32472 + }, + { + "epoch": 0.7612182586426156, + "grad_norm": 3.3561694622039795, + "learning_rate": 2.685271360401026e-05, + "loss": 0.2142, + "step": 32473 + }, + { + "epoch": 0.7612417002174207, + "grad_norm": 0.762371838092804, + "learning_rate": 2.6847691036359813e-05, + "loss": 0.0797, + "step": 32474 + }, + { + "epoch": 0.7612651417922256, + "grad_norm": 3.664763927459717, + "learning_rate": 2.6842668865635856e-05, + "loss": 0.3784, + "step": 32475 + }, + { + "epoch": 0.7612885833670306, + "grad_norm": 5.0559468269348145, + "learning_rate": 2.6837647091865637e-05, + "loss": 0.3332, + "step": 32476 + }, + { + "epoch": 0.7613120249418356, + "grad_norm": 3.9086692333221436, + "learning_rate": 2.6832625715076378e-05, + "loss": 0.1605, + "step": 32477 + }, + { + "epoch": 0.7613354665166406, + "grad_norm": 1.3513846397399902, + "learning_rate": 2.6827604735295364e-05, + "loss": 0.0591, + "step": 32478 + }, + { + "epoch": 0.7613589080914456, + "grad_norm": 0.7241690158843994, + "learning_rate": 2.6822584152549813e-05, + "loss": 0.0591, + "step": 32479 + }, + { + "epoch": 0.7613823496662506, + "grad_norm": 4.105838298797607, + "learning_rate": 2.681756396686699e-05, + "loss": 0.3764, + "step": 32480 + }, + { + "epoch": 0.7614057912410556, + "grad_norm": 1.0918971300125122, + "learning_rate": 2.6812544178274135e-05, + "loss": 0.1162, + "step": 32481 + }, + { + "epoch": 0.7614292328158606, + "grad_norm": 2.0799286365509033, + "learning_rate": 2.6807524786798465e-05, + "loss": 1.8486, + "step": 32482 + }, + { + "epoch": 0.7614526743906656, + "grad_norm": 2.077346086502075, + "learning_rate": 2.6802505792467224e-05, + "loss": 0.176, + "step": 32483 + }, + { + "epoch": 0.7614761159654706, + "grad_norm": 3.9726459980010986, + "learning_rate": 2.6797487195307647e-05, + "loss": 0.3514, + "step": 32484 + }, + { + "epoch": 0.7614995575402755, + "grad_norm": 0.719325602054596, + "learning_rate": 2.679246899534693e-05, + "loss": 0.0585, + "step": 32485 + }, + { + "epoch": 0.7615229991150806, + "grad_norm": 3.4084153175354004, + "learning_rate": 2.6787451192612378e-05, + "loss": 0.2218, + "step": 32486 + }, + { + "epoch": 0.7615464406898855, + "grad_norm": 1.9814722537994385, + "learning_rate": 2.6782433787131157e-05, + "loss": 0.2178, + "step": 32487 + }, + { + "epoch": 0.7615698822646906, + "grad_norm": 1.8585293292999268, + "learning_rate": 2.677741677893052e-05, + "loss": 0.0641, + "step": 32488 + }, + { + "epoch": 0.7615933238394955, + "grad_norm": 2.804417371749878, + "learning_rate": 2.677240016803767e-05, + "loss": 0.0859, + "step": 32489 + }, + { + "epoch": 0.7616167654143006, + "grad_norm": 3.7062718868255615, + "learning_rate": 2.676738395447982e-05, + "loss": 0.3619, + "step": 32490 + }, + { + "epoch": 0.7616402069891055, + "grad_norm": 2.1825616359710693, + "learning_rate": 2.6762368138284223e-05, + "loss": 0.1908, + "step": 32491 + }, + { + "epoch": 0.7616636485639106, + "grad_norm": 0.3068094253540039, + "learning_rate": 2.6757352719478047e-05, + "loss": 0.0268, + "step": 32492 + }, + { + "epoch": 0.7616870901387155, + "grad_norm": 0.8632696866989136, + "learning_rate": 2.675233769808856e-05, + "loss": 0.0802, + "step": 32493 + }, + { + "epoch": 0.7617105317135205, + "grad_norm": 2.6216068267822266, + "learning_rate": 2.674732307414295e-05, + "loss": 0.2046, + "step": 32494 + }, + { + "epoch": 0.7617339732883255, + "grad_norm": 2.370847463607788, + "learning_rate": 2.674230884766843e-05, + "loss": 0.181, + "step": 32495 + }, + { + "epoch": 0.7617574148631305, + "grad_norm": 3.2088515758514404, + "learning_rate": 2.67372950186922e-05, + "loss": 0.297, + "step": 32496 + }, + { + "epoch": 0.7617808564379355, + "grad_norm": 3.6637542247772217, + "learning_rate": 2.673228158724147e-05, + "loss": 0.3679, + "step": 32497 + }, + { + "epoch": 0.7618042980127405, + "grad_norm": 2.495309591293335, + "learning_rate": 2.67272685533434e-05, + "loss": 0.2486, + "step": 32498 + }, + { + "epoch": 0.7618277395875455, + "grad_norm": 1.4351288080215454, + "learning_rate": 2.6722255917025275e-05, + "loss": 0.0556, + "step": 32499 + }, + { + "epoch": 0.7618511811623505, + "grad_norm": 4.066262722015381, + "learning_rate": 2.671724367831424e-05, + "loss": 0.4872, + "step": 32500 + }, + { + "epoch": 0.7618746227371554, + "grad_norm": 2.762397050857544, + "learning_rate": 2.67122318372375e-05, + "loss": 0.2861, + "step": 32501 + }, + { + "epoch": 0.7618980643119605, + "grad_norm": 3.122744560241699, + "learning_rate": 2.670722039382223e-05, + "loss": 0.2056, + "step": 32502 + }, + { + "epoch": 0.7619215058867654, + "grad_norm": 1.8754677772521973, + "learning_rate": 2.6702209348095665e-05, + "loss": 0.0783, + "step": 32503 + }, + { + "epoch": 0.7619449474615705, + "grad_norm": 3.0350887775421143, + "learning_rate": 2.669719870008497e-05, + "loss": 0.1331, + "step": 32504 + }, + { + "epoch": 0.7619683890363754, + "grad_norm": 4.185822010040283, + "learning_rate": 2.6692188449817313e-05, + "loss": 0.4576, + "step": 32505 + }, + { + "epoch": 0.7619918306111805, + "grad_norm": 2.6168129444122314, + "learning_rate": 2.6687178597319927e-05, + "loss": 0.1709, + "step": 32506 + }, + { + "epoch": 0.7620152721859854, + "grad_norm": 2.2394351959228516, + "learning_rate": 2.6682169142619984e-05, + "loss": 0.1318, + "step": 32507 + }, + { + "epoch": 0.7620387137607905, + "grad_norm": 2.4357078075408936, + "learning_rate": 2.6677160085744647e-05, + "loss": 1.5911, + "step": 32508 + }, + { + "epoch": 0.7620621553355954, + "grad_norm": 3.4299983978271484, + "learning_rate": 2.6672151426721092e-05, + "loss": 0.1288, + "step": 32509 + }, + { + "epoch": 0.7620855969104005, + "grad_norm": 2.9262609481811523, + "learning_rate": 2.666714316557648e-05, + "loss": 0.1161, + "step": 32510 + }, + { + "epoch": 0.7621090384852054, + "grad_norm": 2.4638378620147705, + "learning_rate": 2.6662135302338042e-05, + "loss": 0.3466, + "step": 32511 + }, + { + "epoch": 0.7621324800600104, + "grad_norm": 1.8667988777160645, + "learning_rate": 2.6657127837032924e-05, + "loss": 0.164, + "step": 32512 + }, + { + "epoch": 0.7621559216348154, + "grad_norm": 3.2570948600769043, + "learning_rate": 2.6652120769688295e-05, + "loss": 0.2547, + "step": 32513 + }, + { + "epoch": 0.7621793632096204, + "grad_norm": 1.0747005939483643, + "learning_rate": 2.6647114100331282e-05, + "loss": 0.1176, + "step": 32514 + }, + { + "epoch": 0.7622028047844254, + "grad_norm": 3.6407523155212402, + "learning_rate": 2.6642107828989115e-05, + "loss": 0.4104, + "step": 32515 + }, + { + "epoch": 0.7622262463592304, + "grad_norm": 2.0962677001953125, + "learning_rate": 2.6637101955688947e-05, + "loss": 0.1356, + "step": 32516 + }, + { + "epoch": 0.7622496879340354, + "grad_norm": 1.019498586654663, + "learning_rate": 2.6632096480457892e-05, + "loss": 0.0654, + "step": 32517 + }, + { + "epoch": 0.7622731295088404, + "grad_norm": 4.203105926513672, + "learning_rate": 2.6627091403323178e-05, + "loss": 0.3956, + "step": 32518 + }, + { + "epoch": 0.7622965710836453, + "grad_norm": 2.157564163208008, + "learning_rate": 2.6622086724311923e-05, + "loss": 0.2313, + "step": 32519 + }, + { + "epoch": 0.7623200126584504, + "grad_norm": 0.6980197429656982, + "learning_rate": 2.661708244345129e-05, + "loss": 0.0571, + "step": 32520 + }, + { + "epoch": 0.7623434542332553, + "grad_norm": 3.9604034423828125, + "learning_rate": 2.6612078560768427e-05, + "loss": 0.2729, + "step": 32521 + }, + { + "epoch": 0.7623668958080604, + "grad_norm": 1.5675761699676514, + "learning_rate": 2.66070750762905e-05, + "loss": 0.9625, + "step": 32522 + }, + { + "epoch": 0.7623903373828654, + "grad_norm": 1.7488987445831299, + "learning_rate": 2.660207199004461e-05, + "loss": 0.1576, + "step": 32523 + }, + { + "epoch": 0.7624137789576704, + "grad_norm": 1.8733171224594116, + "learning_rate": 2.6597069302057975e-05, + "loss": 1.2724, + "step": 32524 + }, + { + "epoch": 0.7624372205324754, + "grad_norm": 5.77050256729126, + "learning_rate": 2.65920670123577e-05, + "loss": 0.7656, + "step": 32525 + }, + { + "epoch": 0.7624606621072804, + "grad_norm": 2.335193395614624, + "learning_rate": 2.658706512097091e-05, + "loss": 0.2849, + "step": 32526 + }, + { + "epoch": 0.7624841036820854, + "grad_norm": 3.2124695777893066, + "learning_rate": 2.6582063627924792e-05, + "loss": 0.141, + "step": 32527 + }, + { + "epoch": 0.7625075452568904, + "grad_norm": 2.9452593326568604, + "learning_rate": 2.6577062533246468e-05, + "loss": 0.2253, + "step": 32528 + }, + { + "epoch": 0.7625309868316954, + "grad_norm": 2.3282976150512695, + "learning_rate": 2.657206183696306e-05, + "loss": 0.1331, + "step": 32529 + }, + { + "epoch": 0.7625544284065003, + "grad_norm": 0.37649810314178467, + "learning_rate": 2.656706153910169e-05, + "loss": 0.0209, + "step": 32530 + }, + { + "epoch": 0.7625778699813054, + "grad_norm": 3.816599130630493, + "learning_rate": 2.656206163968953e-05, + "loss": 0.154, + "step": 32531 + }, + { + "epoch": 0.7626013115561103, + "grad_norm": 1.7543842792510986, + "learning_rate": 2.6557062138753686e-05, + "loss": 0.1147, + "step": 32532 + }, + { + "epoch": 0.7626247531309154, + "grad_norm": 1.6571896076202393, + "learning_rate": 2.6552063036321296e-05, + "loss": 0.097, + "step": 32533 + }, + { + "epoch": 0.7626481947057203, + "grad_norm": 1.9582442045211792, + "learning_rate": 2.654706433241947e-05, + "loss": 0.2014, + "step": 32534 + }, + { + "epoch": 0.7626716362805254, + "grad_norm": 0.4027699828147888, + "learning_rate": 2.6542066027075342e-05, + "loss": 0.0399, + "step": 32535 + }, + { + "epoch": 0.7626950778553303, + "grad_norm": 1.4185532331466675, + "learning_rate": 2.6537068120316e-05, + "loss": 0.0809, + "step": 32536 + }, + { + "epoch": 0.7627185194301354, + "grad_norm": 0.19673311710357666, + "learning_rate": 2.6532070612168615e-05, + "loss": 0.0087, + "step": 32537 + }, + { + "epoch": 0.7627419610049403, + "grad_norm": 0.5843323469161987, + "learning_rate": 2.652707350266026e-05, + "loss": 0.0703, + "step": 32538 + }, + { + "epoch": 0.7627654025797453, + "grad_norm": 1.1174695491790771, + "learning_rate": 2.6522076791818097e-05, + "loss": 0.0556, + "step": 32539 + }, + { + "epoch": 0.7627888441545503, + "grad_norm": 2.5512523651123047, + "learning_rate": 2.651708047966921e-05, + "loss": 0.1506, + "step": 32540 + }, + { + "epoch": 0.7628122857293553, + "grad_norm": 5.014354705810547, + "learning_rate": 2.6512084566240715e-05, + "loss": 0.3051, + "step": 32541 + }, + { + "epoch": 0.7628357273041603, + "grad_norm": 1.9466692209243774, + "learning_rate": 2.6507089051559708e-05, + "loss": 0.0999, + "step": 32542 + }, + { + "epoch": 0.7628591688789653, + "grad_norm": 1.877242922782898, + "learning_rate": 2.650209393565327e-05, + "loss": 0.1247, + "step": 32543 + }, + { + "epoch": 0.7628826104537703, + "grad_norm": 2.004901647567749, + "learning_rate": 2.6497099218548583e-05, + "loss": 0.2362, + "step": 32544 + }, + { + "epoch": 0.7629060520285753, + "grad_norm": 3.832490921020508, + "learning_rate": 2.6492104900272686e-05, + "loss": 0.2657, + "step": 32545 + }, + { + "epoch": 0.7629294936033802, + "grad_norm": 0.7649604678153992, + "learning_rate": 2.6487110980852704e-05, + "loss": 0.026, + "step": 32546 + }, + { + "epoch": 0.7629529351781853, + "grad_norm": 3.3220083713531494, + "learning_rate": 2.6482117460315713e-05, + "loss": 0.3055, + "step": 32547 + }, + { + "epoch": 0.7629763767529902, + "grad_norm": 3.7898361682891846, + "learning_rate": 2.6477124338688807e-05, + "loss": 0.3828, + "step": 32548 + }, + { + "epoch": 0.7629998183277953, + "grad_norm": 2.5981838703155518, + "learning_rate": 2.647213161599911e-05, + "loss": 0.2464, + "step": 32549 + }, + { + "epoch": 0.7630232599026002, + "grad_norm": 0.7315416932106018, + "learning_rate": 2.6467139292273668e-05, + "loss": 0.1011, + "step": 32550 + }, + { + "epoch": 0.7630467014774053, + "grad_norm": 2.1944384574890137, + "learning_rate": 2.6462147367539624e-05, + "loss": 0.0834, + "step": 32551 + }, + { + "epoch": 0.7630701430522102, + "grad_norm": 1.0982431173324585, + "learning_rate": 2.6457155841824032e-05, + "loss": 0.0998, + "step": 32552 + }, + { + "epoch": 0.7630935846270153, + "grad_norm": 2.3464760780334473, + "learning_rate": 2.6452164715153992e-05, + "loss": 0.2069, + "step": 32553 + }, + { + "epoch": 0.7631170262018202, + "grad_norm": 2.8646514415740967, + "learning_rate": 2.6447173987556563e-05, + "loss": 0.1426, + "step": 32554 + }, + { + "epoch": 0.7631404677766253, + "grad_norm": 2.489297866821289, + "learning_rate": 2.6442183659058815e-05, + "loss": 0.1583, + "step": 32555 + }, + { + "epoch": 0.7631639093514302, + "grad_norm": 1.477311611175537, + "learning_rate": 2.643719372968787e-05, + "loss": 0.0504, + "step": 32556 + }, + { + "epoch": 0.7631873509262352, + "grad_norm": 2.6569457054138184, + "learning_rate": 2.643220419947079e-05, + "loss": 0.184, + "step": 32557 + }, + { + "epoch": 0.7632107925010402, + "grad_norm": 2.346386671066284, + "learning_rate": 2.642721506843463e-05, + "loss": 0.2396, + "step": 32558 + }, + { + "epoch": 0.7632342340758452, + "grad_norm": 0.30183857679367065, + "learning_rate": 2.6422226336606472e-05, + "loss": 0.0127, + "step": 32559 + }, + { + "epoch": 0.7632576756506502, + "grad_norm": 2.1531753540039062, + "learning_rate": 2.6417238004013377e-05, + "loss": 0.2636, + "step": 32560 + }, + { + "epoch": 0.7632811172254552, + "grad_norm": 2.3299198150634766, + "learning_rate": 2.6412250070682387e-05, + "loss": 0.1694, + "step": 32561 + }, + { + "epoch": 0.7633045588002602, + "grad_norm": 3.0836455821990967, + "learning_rate": 2.640726253664063e-05, + "loss": 0.134, + "step": 32562 + }, + { + "epoch": 0.7633280003750652, + "grad_norm": 3.9843409061431885, + "learning_rate": 2.6402275401915112e-05, + "loss": 0.2441, + "step": 32563 + }, + { + "epoch": 0.7633514419498701, + "grad_norm": 2.7770752906799316, + "learning_rate": 2.639728866653295e-05, + "loss": 0.2803, + "step": 32564 + }, + { + "epoch": 0.7633748835246752, + "grad_norm": 0.8170757293701172, + "learning_rate": 2.6392302330521157e-05, + "loss": 0.0562, + "step": 32565 + }, + { + "epoch": 0.7633983250994801, + "grad_norm": 1.5094918012619019, + "learning_rate": 2.63873163939068e-05, + "loss": 0.1717, + "step": 32566 + }, + { + "epoch": 0.7634217666742852, + "grad_norm": 3.956864595413208, + "learning_rate": 2.638233085671695e-05, + "loss": 0.235, + "step": 32567 + }, + { + "epoch": 0.7634452082490901, + "grad_norm": 4.403086185455322, + "learning_rate": 2.6377345718978597e-05, + "loss": 0.2366, + "step": 32568 + }, + { + "epoch": 0.7634686498238952, + "grad_norm": 2.864708423614502, + "learning_rate": 2.637236098071888e-05, + "loss": 0.1709, + "step": 32569 + }, + { + "epoch": 0.7634920913987001, + "grad_norm": 1.6149656772613525, + "learning_rate": 2.6367376641964792e-05, + "loss": 0.0693, + "step": 32570 + }, + { + "epoch": 0.7635155329735052, + "grad_norm": 1.8528774976730347, + "learning_rate": 2.6362392702743387e-05, + "loss": 0.1626, + "step": 32571 + }, + { + "epoch": 0.7635389745483101, + "grad_norm": 2.938760757446289, + "learning_rate": 2.6357409163081714e-05, + "loss": 0.2873, + "step": 32572 + }, + { + "epoch": 0.7635624161231152, + "grad_norm": 3.8340237140655518, + "learning_rate": 2.6352426023006814e-05, + "loss": 0.2924, + "step": 32573 + }, + { + "epoch": 0.7635858576979202, + "grad_norm": 2.951742172241211, + "learning_rate": 2.634744328254569e-05, + "loss": 0.2968, + "step": 32574 + }, + { + "epoch": 0.7636092992727251, + "grad_norm": 0.7598261833190918, + "learning_rate": 2.634246094172541e-05, + "loss": 0.0583, + "step": 32575 + }, + { + "epoch": 0.7636327408475302, + "grad_norm": 2.560589075088501, + "learning_rate": 2.633747900057303e-05, + "loss": 0.2851, + "step": 32576 + }, + { + "epoch": 0.7636561824223351, + "grad_norm": 0.8015984296798706, + "learning_rate": 2.633249745911557e-05, + "loss": 0.0701, + "step": 32577 + }, + { + "epoch": 0.7636796239971402, + "grad_norm": 2.797783851623535, + "learning_rate": 2.6327516317380042e-05, + "loss": 0.1552, + "step": 32578 + }, + { + "epoch": 0.7637030655719451, + "grad_norm": 1.889338493347168, + "learning_rate": 2.6322535575393482e-05, + "loss": 0.0992, + "step": 32579 + }, + { + "epoch": 0.7637265071467502, + "grad_norm": 1.8301290273666382, + "learning_rate": 2.6317555233182923e-05, + "loss": 0.1062, + "step": 32580 + }, + { + "epoch": 0.7637499487215551, + "grad_norm": 5.206506252288818, + "learning_rate": 2.6312575290775343e-05, + "loss": 0.4467, + "step": 32581 + }, + { + "epoch": 0.7637733902963602, + "grad_norm": 2.0562381744384766, + "learning_rate": 2.6307595748197834e-05, + "loss": 0.146, + "step": 32582 + }, + { + "epoch": 0.7637968318711651, + "grad_norm": 2.8587372303009033, + "learning_rate": 2.630261660547738e-05, + "loss": 0.2782, + "step": 32583 + }, + { + "epoch": 0.7638202734459701, + "grad_norm": 1.3260585069656372, + "learning_rate": 2.629763786264101e-05, + "loss": 0.0523, + "step": 32584 + }, + { + "epoch": 0.7638437150207751, + "grad_norm": 3.0793890953063965, + "learning_rate": 2.6292659519715724e-05, + "loss": 0.2102, + "step": 32585 + }, + { + "epoch": 0.7638671565955801, + "grad_norm": 2.4160008430480957, + "learning_rate": 2.6287681576728505e-05, + "loss": 1.3214, + "step": 32586 + }, + { + "epoch": 0.7638905981703851, + "grad_norm": 0.5410684943199158, + "learning_rate": 2.6282704033706408e-05, + "loss": 0.0522, + "step": 32587 + }, + { + "epoch": 0.7639140397451901, + "grad_norm": 1.6813240051269531, + "learning_rate": 2.6277726890676457e-05, + "loss": 0.102, + "step": 32588 + }, + { + "epoch": 0.7639374813199951, + "grad_norm": 1.7180137634277344, + "learning_rate": 2.6272750147665626e-05, + "loss": 0.1098, + "step": 32589 + }, + { + "epoch": 0.7639609228948001, + "grad_norm": 1.7067914009094238, + "learning_rate": 2.6267773804700936e-05, + "loss": 0.0989, + "step": 32590 + }, + { + "epoch": 0.763984364469605, + "grad_norm": 1.6177257299423218, + "learning_rate": 2.6262797861809373e-05, + "loss": 0.6908, + "step": 32591 + }, + { + "epoch": 0.7640078060444101, + "grad_norm": 3.1223630905151367, + "learning_rate": 2.6257822319017942e-05, + "loss": 0.3227, + "step": 32592 + }, + { + "epoch": 0.764031247619215, + "grad_norm": 3.045473575592041, + "learning_rate": 2.625284717635361e-05, + "loss": 1.0906, + "step": 32593 + }, + { + "epoch": 0.7640546891940201, + "grad_norm": 0.7022808790206909, + "learning_rate": 2.6247872433843445e-05, + "loss": 0.0369, + "step": 32594 + }, + { + "epoch": 0.764078130768825, + "grad_norm": 3.142071485519409, + "learning_rate": 2.624289809151438e-05, + "loss": 0.2515, + "step": 32595 + }, + { + "epoch": 0.7641015723436301, + "grad_norm": 4.787646770477295, + "learning_rate": 2.6237924149393434e-05, + "loss": 0.4025, + "step": 32596 + }, + { + "epoch": 0.764125013918435, + "grad_norm": 3.0441365242004395, + "learning_rate": 2.623295060750759e-05, + "loss": 0.1613, + "step": 32597 + }, + { + "epoch": 0.7641484554932401, + "grad_norm": 3.3768324851989746, + "learning_rate": 2.6227977465883812e-05, + "loss": 0.2201, + "step": 32598 + }, + { + "epoch": 0.764171897068045, + "grad_norm": 0.7862772345542908, + "learning_rate": 2.6223004724549093e-05, + "loss": 0.0486, + "step": 32599 + }, + { + "epoch": 0.76419533864285, + "grad_norm": 0.9877697229385376, + "learning_rate": 2.621803238353041e-05, + "loss": 0.094, + "step": 32600 + }, + { + "epoch": 0.764218780217655, + "grad_norm": 1.6819970607757568, + "learning_rate": 2.6213060442854797e-05, + "loss": 0.1383, + "step": 32601 + }, + { + "epoch": 0.76424222179246, + "grad_norm": 3.3338623046875, + "learning_rate": 2.620808890254919e-05, + "loss": 0.3143, + "step": 32602 + }, + { + "epoch": 0.764265663367265, + "grad_norm": 1.8802632093429565, + "learning_rate": 2.620311776264056e-05, + "loss": 1.1606, + "step": 32603 + }, + { + "epoch": 0.76428910494207, + "grad_norm": 1.041670322418213, + "learning_rate": 2.6198147023155893e-05, + "loss": 0.0978, + "step": 32604 + }, + { + "epoch": 0.764312546516875, + "grad_norm": 2.925096035003662, + "learning_rate": 2.619317668412217e-05, + "loss": 0.166, + "step": 32605 + }, + { + "epoch": 0.76433598809168, + "grad_norm": 1.2530661821365356, + "learning_rate": 2.6188206745566302e-05, + "loss": 0.0984, + "step": 32606 + }, + { + "epoch": 0.764359429666485, + "grad_norm": 0.4330393671989441, + "learning_rate": 2.618323720751532e-05, + "loss": 0.0348, + "step": 32607 + }, + { + "epoch": 0.76438287124129, + "grad_norm": 3.9808526039123535, + "learning_rate": 2.6178268069996183e-05, + "loss": 0.3909, + "step": 32608 + }, + { + "epoch": 0.764406312816095, + "grad_norm": 2.136929512023926, + "learning_rate": 2.6173299333035838e-05, + "loss": 1.5305, + "step": 32609 + }, + { + "epoch": 0.7644297543909, + "grad_norm": 2.9018983840942383, + "learning_rate": 2.616833099666124e-05, + "loss": 0.2232, + "step": 32610 + }, + { + "epoch": 0.7644531959657049, + "grad_norm": 0.7830193042755127, + "learning_rate": 2.6163363060899327e-05, + "loss": 0.0277, + "step": 32611 + }, + { + "epoch": 0.76447663754051, + "grad_norm": 2.271955728530884, + "learning_rate": 2.6158395525777115e-05, + "loss": 0.1962, + "step": 32612 + }, + { + "epoch": 0.7645000791153149, + "grad_norm": 2.8190996646881104, + "learning_rate": 2.615342839132149e-05, + "loss": 0.2775, + "step": 32613 + }, + { + "epoch": 0.76452352069012, + "grad_norm": 1.857096791267395, + "learning_rate": 2.6148461657559475e-05, + "loss": 0.1413, + "step": 32614 + }, + { + "epoch": 0.7645469622649249, + "grad_norm": 3.298548698425293, + "learning_rate": 2.6143495324517974e-05, + "loss": 0.2908, + "step": 32615 + }, + { + "epoch": 0.76457040383973, + "grad_norm": 2.9230148792266846, + "learning_rate": 2.6138529392223953e-05, + "loss": 0.2519, + "step": 32616 + }, + { + "epoch": 0.7645938454145349, + "grad_norm": 3.1390841007232666, + "learning_rate": 2.6133563860704346e-05, + "loss": 0.2139, + "step": 32617 + }, + { + "epoch": 0.76461728698934, + "grad_norm": 4.5983381271362305, + "learning_rate": 2.6128598729986096e-05, + "loss": 0.2872, + "step": 32618 + }, + { + "epoch": 0.7646407285641449, + "grad_norm": 1.6049506664276123, + "learning_rate": 2.612363400009612e-05, + "loss": 0.0877, + "step": 32619 + }, + { + "epoch": 0.7646641701389499, + "grad_norm": 4.429106712341309, + "learning_rate": 2.611866967106141e-05, + "loss": 0.4007, + "step": 32620 + }, + { + "epoch": 0.7646876117137549, + "grad_norm": 3.412985324859619, + "learning_rate": 2.611370574290888e-05, + "loss": 0.2528, + "step": 32621 + }, + { + "epoch": 0.7647110532885599, + "grad_norm": 1.6256574392318726, + "learning_rate": 2.6108742215665462e-05, + "loss": 0.0843, + "step": 32622 + }, + { + "epoch": 0.7647344948633649, + "grad_norm": 0.5951262712478638, + "learning_rate": 2.610377908935805e-05, + "loss": 0.0311, + "step": 32623 + }, + { + "epoch": 0.7647579364381699, + "grad_norm": 1.5915584564208984, + "learning_rate": 2.6098816364013644e-05, + "loss": 0.0773, + "step": 32624 + }, + { + "epoch": 0.764781378012975, + "grad_norm": 1.7595257759094238, + "learning_rate": 2.6093854039659104e-05, + "loss": 0.1489, + "step": 32625 + }, + { + "epoch": 0.7648048195877799, + "grad_norm": 2.307816505432129, + "learning_rate": 2.6088892116321417e-05, + "loss": 0.1791, + "step": 32626 + }, + { + "epoch": 0.764828261162585, + "grad_norm": 2.4391772747039795, + "learning_rate": 2.6083930594027494e-05, + "loss": 0.2712, + "step": 32627 + }, + { + "epoch": 0.7648517027373899, + "grad_norm": 1.5296202898025513, + "learning_rate": 2.6078969472804227e-05, + "loss": 0.0613, + "step": 32628 + }, + { + "epoch": 0.7648751443121949, + "grad_norm": 1.1785030364990234, + "learning_rate": 2.607400875267856e-05, + "loss": 0.1707, + "step": 32629 + }, + { + "epoch": 0.7648985858869999, + "grad_norm": 3.1346805095672607, + "learning_rate": 2.606904843367739e-05, + "loss": 0.2476, + "step": 32630 + }, + { + "epoch": 0.7649220274618049, + "grad_norm": 6.508656978607178, + "learning_rate": 2.6064088515827622e-05, + "loss": 0.5502, + "step": 32631 + }, + { + "epoch": 0.7649454690366099, + "grad_norm": 3.1971092224121094, + "learning_rate": 2.605912899915621e-05, + "loss": 0.1982, + "step": 32632 + }, + { + "epoch": 0.7649689106114149, + "grad_norm": 1.3223750591278076, + "learning_rate": 2.605416988369005e-05, + "loss": 0.0949, + "step": 32633 + }, + { + "epoch": 0.7649923521862199, + "grad_norm": 1.6803178787231445, + "learning_rate": 2.6049211169456044e-05, + "loss": 0.1197, + "step": 32634 + }, + { + "epoch": 0.7650157937610249, + "grad_norm": 2.441774368286133, + "learning_rate": 2.60442528564811e-05, + "loss": 0.2947, + "step": 32635 + }, + { + "epoch": 0.7650392353358298, + "grad_norm": 0.33652806282043457, + "learning_rate": 2.603929494479208e-05, + "loss": 0.0182, + "step": 32636 + }, + { + "epoch": 0.7650626769106349, + "grad_norm": 3.1920742988586426, + "learning_rate": 2.6034337434415955e-05, + "loss": 0.268, + "step": 32637 + }, + { + "epoch": 0.7650861184854398, + "grad_norm": 1.4242151975631714, + "learning_rate": 2.6029380325379572e-05, + "loss": 0.0659, + "step": 32638 + }, + { + "epoch": 0.7651095600602449, + "grad_norm": 3.5221049785614014, + "learning_rate": 2.602442361770988e-05, + "loss": 0.2816, + "step": 32639 + }, + { + "epoch": 0.7651330016350498, + "grad_norm": 4.111822605133057, + "learning_rate": 2.6019467311433744e-05, + "loss": 0.1972, + "step": 32640 + }, + { + "epoch": 0.7651564432098549, + "grad_norm": 2.4444875717163086, + "learning_rate": 2.6014511406578057e-05, + "loss": 0.1525, + "step": 32641 + }, + { + "epoch": 0.7651798847846598, + "grad_norm": 2.2134835720062256, + "learning_rate": 2.6009555903169713e-05, + "loss": 0.123, + "step": 32642 + }, + { + "epoch": 0.7652033263594649, + "grad_norm": 1.1354830265045166, + "learning_rate": 2.6004600801235602e-05, + "loss": 0.1191, + "step": 32643 + }, + { + "epoch": 0.7652267679342698, + "grad_norm": 2.394138813018799, + "learning_rate": 2.5999646100802576e-05, + "loss": 0.1748, + "step": 32644 + }, + { + "epoch": 0.7652502095090749, + "grad_norm": 0.5951058268547058, + "learning_rate": 2.5994691801897576e-05, + "loss": 0.0751, + "step": 32645 + }, + { + "epoch": 0.7652736510838798, + "grad_norm": 2.4095215797424316, + "learning_rate": 2.598973790454746e-05, + "loss": 0.2294, + "step": 32646 + }, + { + "epoch": 0.7652970926586848, + "grad_norm": 0.4766218066215515, + "learning_rate": 2.598478440877912e-05, + "loss": 0.0232, + "step": 32647 + }, + { + "epoch": 0.7653205342334898, + "grad_norm": 2.348952531814575, + "learning_rate": 2.5979831314619373e-05, + "loss": 0.2407, + "step": 32648 + }, + { + "epoch": 0.7653439758082948, + "grad_norm": 4.506111145019531, + "learning_rate": 2.5974878622095177e-05, + "loss": 0.3528, + "step": 32649 + }, + { + "epoch": 0.7653674173830998, + "grad_norm": 1.0709339380264282, + "learning_rate": 2.5969926331233374e-05, + "loss": 0.0774, + "step": 32650 + }, + { + "epoch": 0.7653908589579048, + "grad_norm": 7.1895527839660645, + "learning_rate": 2.596497444206081e-05, + "loss": 0.4123, + "step": 32651 + }, + { + "epoch": 0.7654143005327098, + "grad_norm": 1.539189338684082, + "learning_rate": 2.5960022954604403e-05, + "loss": 0.6147, + "step": 32652 + }, + { + "epoch": 0.7654377421075148, + "grad_norm": 2.345353126525879, + "learning_rate": 2.595507186889098e-05, + "loss": 0.2297, + "step": 32653 + }, + { + "epoch": 0.7654611836823197, + "grad_norm": 2.0323903560638428, + "learning_rate": 2.595012118494743e-05, + "loss": 1.5236, + "step": 32654 + }, + { + "epoch": 0.7654846252571248, + "grad_norm": 1.1256496906280518, + "learning_rate": 2.594517090280061e-05, + "loss": 0.0731, + "step": 32655 + }, + { + "epoch": 0.7655080668319297, + "grad_norm": 2.014087438583374, + "learning_rate": 2.594022102247733e-05, + "loss": 0.1112, + "step": 32656 + }, + { + "epoch": 0.7655315084067348, + "grad_norm": 2.992539882659912, + "learning_rate": 2.5935271544004535e-05, + "loss": 0.1942, + "step": 32657 + }, + { + "epoch": 0.7655549499815397, + "grad_norm": 3.893369197845459, + "learning_rate": 2.5930322467409028e-05, + "loss": 0.2939, + "step": 32658 + }, + { + "epoch": 0.7655783915563448, + "grad_norm": 3.883913040161133, + "learning_rate": 2.5925373792717688e-05, + "loss": 1.7668, + "step": 32659 + }, + { + "epoch": 0.7656018331311497, + "grad_norm": 0.7571802735328674, + "learning_rate": 2.5920425519957316e-05, + "loss": 0.0779, + "step": 32660 + }, + { + "epoch": 0.7656252747059548, + "grad_norm": 3.9955971240997314, + "learning_rate": 2.591547764915483e-05, + "loss": 0.2978, + "step": 32661 + }, + { + "epoch": 0.7656487162807597, + "grad_norm": 2.3411316871643066, + "learning_rate": 2.5910530180337035e-05, + "loss": 0.1865, + "step": 32662 + }, + { + "epoch": 0.7656721578555648, + "grad_norm": 2.644576072692871, + "learning_rate": 2.590558311353076e-05, + "loss": 0.1699, + "step": 32663 + }, + { + "epoch": 0.7656955994303697, + "grad_norm": 7.005458831787109, + "learning_rate": 2.5900636448762917e-05, + "loss": 0.4837, + "step": 32664 + }, + { + "epoch": 0.7657190410051747, + "grad_norm": 1.7910916805267334, + "learning_rate": 2.5895690186060297e-05, + "loss": 0.135, + "step": 32665 + }, + { + "epoch": 0.7657424825799797, + "grad_norm": 3.029555320739746, + "learning_rate": 2.5890744325449746e-05, + "loss": 0.236, + "step": 32666 + }, + { + "epoch": 0.7657659241547847, + "grad_norm": 5.016878128051758, + "learning_rate": 2.58857988669581e-05, + "loss": 0.7205, + "step": 32667 + }, + { + "epoch": 0.7657893657295897, + "grad_norm": 3.2788448333740234, + "learning_rate": 2.5880853810612184e-05, + "loss": 0.2976, + "step": 32668 + }, + { + "epoch": 0.7658128073043947, + "grad_norm": 4.869080066680908, + "learning_rate": 2.5875909156438815e-05, + "loss": 0.3504, + "step": 32669 + }, + { + "epoch": 0.7658362488791997, + "grad_norm": 2.73171329498291, + "learning_rate": 2.587096490446489e-05, + "loss": 1.4427, + "step": 32670 + }, + { + "epoch": 0.7658596904540047, + "grad_norm": 1.5454176664352417, + "learning_rate": 2.5866021054717183e-05, + "loss": 0.0956, + "step": 32671 + }, + { + "epoch": 0.7658831320288096, + "grad_norm": 2.675595998764038, + "learning_rate": 2.5861077607222496e-05, + "loss": 0.157, + "step": 32672 + }, + { + "epoch": 0.7659065736036147, + "grad_norm": 1.602251410484314, + "learning_rate": 2.5856134562007717e-05, + "loss": 0.1267, + "step": 32673 + }, + { + "epoch": 0.7659300151784196, + "grad_norm": 1.6145299673080444, + "learning_rate": 2.5851191919099638e-05, + "loss": 0.1281, + "step": 32674 + }, + { + "epoch": 0.7659534567532247, + "grad_norm": 0.7161829471588135, + "learning_rate": 2.584624967852508e-05, + "loss": 0.0301, + "step": 32675 + }, + { + "epoch": 0.7659768983280297, + "grad_norm": 1.3691504001617432, + "learning_rate": 2.5841307840310825e-05, + "loss": 0.0922, + "step": 32676 + }, + { + "epoch": 0.7660003399028347, + "grad_norm": 1.0839776992797852, + "learning_rate": 2.5836366404483747e-05, + "loss": 0.0843, + "step": 32677 + }, + { + "epoch": 0.7660237814776397, + "grad_norm": 0.8821547031402588, + "learning_rate": 2.583142537107064e-05, + "loss": 0.0476, + "step": 32678 + }, + { + "epoch": 0.7660472230524447, + "grad_norm": 1.282576560974121, + "learning_rate": 2.582648474009829e-05, + "loss": 0.1659, + "step": 32679 + }, + { + "epoch": 0.7660706646272497, + "grad_norm": 0.8338629603385925, + "learning_rate": 2.582154451159353e-05, + "loss": 0.089, + "step": 32680 + }, + { + "epoch": 0.7660941062020546, + "grad_norm": 1.9582366943359375, + "learning_rate": 2.5816604685583145e-05, + "loss": 0.135, + "step": 32681 + }, + { + "epoch": 0.7661175477768597, + "grad_norm": 1.8312106132507324, + "learning_rate": 2.5811665262093932e-05, + "loss": 0.1744, + "step": 32682 + }, + { + "epoch": 0.7661409893516646, + "grad_norm": 2.1144256591796875, + "learning_rate": 2.5806726241152736e-05, + "loss": 0.1314, + "step": 32683 + }, + { + "epoch": 0.7661644309264697, + "grad_norm": 3.9711127281188965, + "learning_rate": 2.580178762278629e-05, + "loss": 0.2741, + "step": 32684 + }, + { + "epoch": 0.7661878725012746, + "grad_norm": 0.5488907694816589, + "learning_rate": 2.5796849407021463e-05, + "loss": 0.0387, + "step": 32685 + }, + { + "epoch": 0.7662113140760797, + "grad_norm": 0.775188148021698, + "learning_rate": 2.579191159388502e-05, + "loss": 0.0501, + "step": 32686 + }, + { + "epoch": 0.7662347556508846, + "grad_norm": 1.6571542024612427, + "learning_rate": 2.5786974183403756e-05, + "loss": 0.1611, + "step": 32687 + }, + { + "epoch": 0.7662581972256897, + "grad_norm": 2.026125431060791, + "learning_rate": 2.5782037175604458e-05, + "loss": 0.1103, + "step": 32688 + }, + { + "epoch": 0.7662816388004946, + "grad_norm": 3.0132968425750732, + "learning_rate": 2.577710057051387e-05, + "loss": 0.1829, + "step": 32689 + }, + { + "epoch": 0.7663050803752997, + "grad_norm": 2.1972403526306152, + "learning_rate": 2.5772164368158857e-05, + "loss": 0.2647, + "step": 32690 + }, + { + "epoch": 0.7663285219501046, + "grad_norm": 1.883607029914856, + "learning_rate": 2.576722856856617e-05, + "loss": 0.096, + "step": 32691 + }, + { + "epoch": 0.7663519635249096, + "grad_norm": 3.0045461654663086, + "learning_rate": 2.5762293171762598e-05, + "loss": 1.469, + "step": 32692 + }, + { + "epoch": 0.7663754050997146, + "grad_norm": 2.066307783126831, + "learning_rate": 2.5757358177774892e-05, + "loss": 0.0749, + "step": 32693 + }, + { + "epoch": 0.7663988466745196, + "grad_norm": 1.6728246212005615, + "learning_rate": 2.5752423586629826e-05, + "loss": 0.1565, + "step": 32694 + }, + { + "epoch": 0.7664222882493246, + "grad_norm": 2.6068124771118164, + "learning_rate": 2.574748939835423e-05, + "loss": 0.2439, + "step": 32695 + }, + { + "epoch": 0.7664457298241296, + "grad_norm": 4.077351093292236, + "learning_rate": 2.5742555612974816e-05, + "loss": 0.2669, + "step": 32696 + }, + { + "epoch": 0.7664691713989346, + "grad_norm": 3.080333948135376, + "learning_rate": 2.573762223051841e-05, + "loss": 1.197, + "step": 32697 + }, + { + "epoch": 0.7664926129737396, + "grad_norm": 4.308412551879883, + "learning_rate": 2.5732689251011744e-05, + "loss": 0.4583, + "step": 32698 + }, + { + "epoch": 0.7665160545485445, + "grad_norm": 2.119971990585327, + "learning_rate": 2.5727756674481607e-05, + "loss": 0.2723, + "step": 32699 + }, + { + "epoch": 0.7665394961233496, + "grad_norm": 2.2858526706695557, + "learning_rate": 2.5722824500954746e-05, + "loss": 1.0745, + "step": 32700 + }, + { + "epoch": 0.7665629376981545, + "grad_norm": 4.042911052703857, + "learning_rate": 2.5717892730457905e-05, + "loss": 0.3383, + "step": 32701 + }, + { + "epoch": 0.7665863792729596, + "grad_norm": 1.3902778625488281, + "learning_rate": 2.5712961363017883e-05, + "loss": 0.065, + "step": 32702 + }, + { + "epoch": 0.7666098208477645, + "grad_norm": 2.3027355670928955, + "learning_rate": 2.570803039866143e-05, + "loss": 0.2155, + "step": 32703 + }, + { + "epoch": 0.7666332624225696, + "grad_norm": 3.728410005569458, + "learning_rate": 2.570309983741529e-05, + "loss": 0.317, + "step": 32704 + }, + { + "epoch": 0.7666567039973745, + "grad_norm": 1.5091962814331055, + "learning_rate": 2.5698169679306227e-05, + "loss": 0.046, + "step": 32705 + }, + { + "epoch": 0.7666801455721796, + "grad_norm": 2.322465419769287, + "learning_rate": 2.569323992436098e-05, + "loss": 0.1952, + "step": 32706 + }, + { + "epoch": 0.7667035871469845, + "grad_norm": 2.882596015930176, + "learning_rate": 2.5688310572606266e-05, + "loss": 0.2969, + "step": 32707 + }, + { + "epoch": 0.7667270287217895, + "grad_norm": 2.5228660106658936, + "learning_rate": 2.568338162406887e-05, + "loss": 0.1096, + "step": 32708 + }, + { + "epoch": 0.7667504702965945, + "grad_norm": 2.084843397140503, + "learning_rate": 2.5678453078775565e-05, + "loss": 0.1098, + "step": 32709 + }, + { + "epoch": 0.7667739118713995, + "grad_norm": 1.9574923515319824, + "learning_rate": 2.567352493675307e-05, + "loss": 0.1532, + "step": 32710 + }, + { + "epoch": 0.7667973534462045, + "grad_norm": 2.8359344005584717, + "learning_rate": 2.5668597198028123e-05, + "loss": 0.215, + "step": 32711 + }, + { + "epoch": 0.7668207950210095, + "grad_norm": 2.547375202178955, + "learning_rate": 2.566366986262745e-05, + "loss": 0.1806, + "step": 32712 + }, + { + "epoch": 0.7668442365958145, + "grad_norm": 2.0825142860412598, + "learning_rate": 2.5658742930577783e-05, + "loss": 0.098, + "step": 32713 + }, + { + "epoch": 0.7668676781706195, + "grad_norm": 2.6018283367156982, + "learning_rate": 2.565381640190585e-05, + "loss": 0.1897, + "step": 32714 + }, + { + "epoch": 0.7668911197454245, + "grad_norm": 2.3913793563842773, + "learning_rate": 2.5648890276638427e-05, + "loss": 0.2351, + "step": 32715 + }, + { + "epoch": 0.7669145613202295, + "grad_norm": 3.2388312816619873, + "learning_rate": 2.5643964554802225e-05, + "loss": 0.1999, + "step": 32716 + }, + { + "epoch": 0.7669380028950344, + "grad_norm": 5.120748996734619, + "learning_rate": 2.5639039236423944e-05, + "loss": 0.4772, + "step": 32717 + }, + { + "epoch": 0.7669614444698395, + "grad_norm": 0.531260073184967, + "learning_rate": 2.5634114321530343e-05, + "loss": 0.0416, + "step": 32718 + }, + { + "epoch": 0.7669848860446444, + "grad_norm": 1.8114460706710815, + "learning_rate": 2.5629189810148123e-05, + "loss": 0.0973, + "step": 32719 + }, + { + "epoch": 0.7670083276194495, + "grad_norm": 3.1532936096191406, + "learning_rate": 2.5624265702303974e-05, + "loss": 0.3389, + "step": 32720 + }, + { + "epoch": 0.7670317691942544, + "grad_norm": 2.8826119899749756, + "learning_rate": 2.561934199802466e-05, + "loss": 0.2085, + "step": 32721 + }, + { + "epoch": 0.7670552107690595, + "grad_norm": 3.6772279739379883, + "learning_rate": 2.561441869733692e-05, + "loss": 0.3509, + "step": 32722 + }, + { + "epoch": 0.7670786523438644, + "grad_norm": 1.5801438093185425, + "learning_rate": 2.5609495800267424e-05, + "loss": 0.075, + "step": 32723 + }, + { + "epoch": 0.7671020939186695, + "grad_norm": 3.6384544372558594, + "learning_rate": 2.5604573306842904e-05, + "loss": 0.375, + "step": 32724 + }, + { + "epoch": 0.7671255354934744, + "grad_norm": 2.6819512844085693, + "learning_rate": 2.559965121709006e-05, + "loss": 0.2594, + "step": 32725 + }, + { + "epoch": 0.7671489770682794, + "grad_norm": 2.0675158500671387, + "learning_rate": 2.5594729531035587e-05, + "loss": 0.8288, + "step": 32726 + }, + { + "epoch": 0.7671724186430845, + "grad_norm": 1.9049936532974243, + "learning_rate": 2.5589808248706192e-05, + "loss": 0.1181, + "step": 32727 + }, + { + "epoch": 0.7671958602178894, + "grad_norm": 0.807555079460144, + "learning_rate": 2.5584887370128608e-05, + "loss": 0.0973, + "step": 32728 + }, + { + "epoch": 0.7672193017926945, + "grad_norm": 1.975966215133667, + "learning_rate": 2.5579966895329532e-05, + "loss": 0.113, + "step": 32729 + }, + { + "epoch": 0.7672427433674994, + "grad_norm": 1.3107579946517944, + "learning_rate": 2.5575046824335636e-05, + "loss": 0.0683, + "step": 32730 + }, + { + "epoch": 0.7672661849423045, + "grad_norm": 2.587090015411377, + "learning_rate": 2.557012715717363e-05, + "loss": 0.2116, + "step": 32731 + }, + { + "epoch": 0.7672896265171094, + "grad_norm": 4.166045188903809, + "learning_rate": 2.556520789387018e-05, + "loss": 0.2562, + "step": 32732 + }, + { + "epoch": 0.7673130680919145, + "grad_norm": 1.2962443828582764, + "learning_rate": 2.5560289034452035e-05, + "loss": 0.0463, + "step": 32733 + }, + { + "epoch": 0.7673365096667194, + "grad_norm": 2.6607818603515625, + "learning_rate": 2.5555370578945814e-05, + "loss": 0.1438, + "step": 32734 + }, + { + "epoch": 0.7673599512415245, + "grad_norm": 1.5872777700424194, + "learning_rate": 2.555045252737829e-05, + "loss": 0.0912, + "step": 32735 + }, + { + "epoch": 0.7673833928163294, + "grad_norm": 2.879530191421509, + "learning_rate": 2.55455348797761e-05, + "loss": 0.1203, + "step": 32736 + }, + { + "epoch": 0.7674068343911344, + "grad_norm": 0.9326315522193909, + "learning_rate": 2.554061763616592e-05, + "loss": 0.0488, + "step": 32737 + }, + { + "epoch": 0.7674302759659394, + "grad_norm": 2.1373953819274902, + "learning_rate": 2.5535700796574456e-05, + "loss": 1.3267, + "step": 32738 + }, + { + "epoch": 0.7674537175407444, + "grad_norm": 3.065941095352173, + "learning_rate": 2.5530784361028338e-05, + "loss": 1.2007, + "step": 32739 + }, + { + "epoch": 0.7674771591155494, + "grad_norm": 2.225264310836792, + "learning_rate": 2.5525868329554314e-05, + "loss": 0.894, + "step": 32740 + }, + { + "epoch": 0.7675006006903544, + "grad_norm": 4.389184474945068, + "learning_rate": 2.5520952702179014e-05, + "loss": 0.3555, + "step": 32741 + }, + { + "epoch": 0.7675240422651594, + "grad_norm": 3.317568302154541, + "learning_rate": 2.5516037478929124e-05, + "loss": 0.3529, + "step": 32742 + }, + { + "epoch": 0.7675474838399644, + "grad_norm": 0.6583110094070435, + "learning_rate": 2.55111226598313e-05, + "loss": 0.0497, + "step": 32743 + }, + { + "epoch": 0.7675709254147693, + "grad_norm": 2.4055063724517822, + "learning_rate": 2.5506208244912233e-05, + "loss": 0.2359, + "step": 32744 + }, + { + "epoch": 0.7675943669895744, + "grad_norm": 0.9780978560447693, + "learning_rate": 2.5501294234198535e-05, + "loss": 0.07, + "step": 32745 + }, + { + "epoch": 0.7676178085643793, + "grad_norm": 3.1887118816375732, + "learning_rate": 2.549638062771691e-05, + "loss": 0.2737, + "step": 32746 + }, + { + "epoch": 0.7676412501391844, + "grad_norm": 2.6780192852020264, + "learning_rate": 2.5491467425494053e-05, + "loss": 0.2344, + "step": 32747 + }, + { + "epoch": 0.7676646917139893, + "grad_norm": 3.0426931381225586, + "learning_rate": 2.5486554627556582e-05, + "loss": 0.2117, + "step": 32748 + }, + { + "epoch": 0.7676881332887944, + "grad_norm": 6.081753253936768, + "learning_rate": 2.548164223393117e-05, + "loss": 0.4654, + "step": 32749 + }, + { + "epoch": 0.7677115748635993, + "grad_norm": 0.8763041496276855, + "learning_rate": 2.5476730244644452e-05, + "loss": 0.0809, + "step": 32750 + }, + { + "epoch": 0.7677350164384044, + "grad_norm": 2.2574384212493896, + "learning_rate": 2.54718186597231e-05, + "loss": 1.3824, + "step": 32751 + }, + { + "epoch": 0.7677584580132093, + "grad_norm": 2.6107640266418457, + "learning_rate": 2.5466907479193726e-05, + "loss": 0.2588, + "step": 32752 + }, + { + "epoch": 0.7677818995880143, + "grad_norm": 2.099043369293213, + "learning_rate": 2.5461996703083025e-05, + "loss": 0.2409, + "step": 32753 + }, + { + "epoch": 0.7678053411628193, + "grad_norm": 2.7620792388916016, + "learning_rate": 2.5457086331417635e-05, + "loss": 0.2554, + "step": 32754 + }, + { + "epoch": 0.7678287827376243, + "grad_norm": 2.5489120483398438, + "learning_rate": 2.545217636422419e-05, + "loss": 0.1409, + "step": 32755 + }, + { + "epoch": 0.7678522243124293, + "grad_norm": 3.345533847808838, + "learning_rate": 2.5447266801529334e-05, + "loss": 0.2602, + "step": 32756 + }, + { + "epoch": 0.7678756658872343, + "grad_norm": 0.7961232662200928, + "learning_rate": 2.544235764335967e-05, + "loss": 0.0523, + "step": 32757 + }, + { + "epoch": 0.7678991074620393, + "grad_norm": 3.508708953857422, + "learning_rate": 2.54374488897419e-05, + "loss": 1.8162, + "step": 32758 + }, + { + "epoch": 0.7679225490368443, + "grad_norm": 3.1941022872924805, + "learning_rate": 2.5432540540702608e-05, + "loss": 0.4322, + "step": 32759 + }, + { + "epoch": 0.7679459906116493, + "grad_norm": 3.7454190254211426, + "learning_rate": 2.542763259626847e-05, + "loss": 0.2843, + "step": 32760 + }, + { + "epoch": 0.7679694321864543, + "grad_norm": 2.7394185066223145, + "learning_rate": 2.542272505646609e-05, + "loss": 1.8779, + "step": 32761 + }, + { + "epoch": 0.7679928737612592, + "grad_norm": 2.853790760040283, + "learning_rate": 2.5417817921322106e-05, + "loss": 0.1688, + "step": 32762 + }, + { + "epoch": 0.7680163153360643, + "grad_norm": 4.0396904945373535, + "learning_rate": 2.5412911190863142e-05, + "loss": 0.41, + "step": 32763 + }, + { + "epoch": 0.7680397569108692, + "grad_norm": 0.7151923775672913, + "learning_rate": 2.540800486511582e-05, + "loss": 0.0316, + "step": 32764 + }, + { + "epoch": 0.7680631984856743, + "grad_norm": 1.9533108472824097, + "learning_rate": 2.5403098944106716e-05, + "loss": 0.0443, + "step": 32765 + }, + { + "epoch": 0.7680866400604792, + "grad_norm": 3.487092971801758, + "learning_rate": 2.5398193427862536e-05, + "loss": 0.2336, + "step": 32766 + }, + { + "epoch": 0.7681100816352843, + "grad_norm": 0.7756311297416687, + "learning_rate": 2.539328831640986e-05, + "loss": 0.0501, + "step": 32767 + }, + { + "epoch": 0.7681335232100892, + "grad_norm": 2.1484291553497314, + "learning_rate": 2.53883836097753e-05, + "loss": 0.2447, + "step": 32768 + }, + { + "epoch": 0.7681569647848943, + "grad_norm": 2.332052707672119, + "learning_rate": 2.5383479307985426e-05, + "loss": 1.1475, + "step": 32769 + }, + { + "epoch": 0.7681804063596992, + "grad_norm": 1.9532166719436646, + "learning_rate": 2.537857541106693e-05, + "loss": 0.1336, + "step": 32770 + }, + { + "epoch": 0.7682038479345042, + "grad_norm": 2.466264486312866, + "learning_rate": 2.5373671919046382e-05, + "loss": 0.1207, + "step": 32771 + }, + { + "epoch": 0.7682272895093092, + "grad_norm": 1.7476894855499268, + "learning_rate": 2.5368768831950363e-05, + "loss": 0.1044, + "step": 32772 + }, + { + "epoch": 0.7682507310841142, + "grad_norm": 0.45939287543296814, + "learning_rate": 2.5363866149805525e-05, + "loss": 0.0323, + "step": 32773 + }, + { + "epoch": 0.7682741726589192, + "grad_norm": 3.319809913635254, + "learning_rate": 2.5358963872638453e-05, + "loss": 0.2473, + "step": 32774 + }, + { + "epoch": 0.7682976142337242, + "grad_norm": 2.7450664043426514, + "learning_rate": 2.5354062000475732e-05, + "loss": 0.2122, + "step": 32775 + }, + { + "epoch": 0.7683210558085292, + "grad_norm": 1.164842963218689, + "learning_rate": 2.5349160533343974e-05, + "loss": 0.0843, + "step": 32776 + }, + { + "epoch": 0.7683444973833342, + "grad_norm": 0.20423515141010284, + "learning_rate": 2.534425947126975e-05, + "loss": 0.013, + "step": 32777 + }, + { + "epoch": 0.7683679389581393, + "grad_norm": 4.404086112976074, + "learning_rate": 2.533935881427969e-05, + "loss": 0.3122, + "step": 32778 + }, + { + "epoch": 0.7683913805329442, + "grad_norm": 0.8289499878883362, + "learning_rate": 2.5334458562400378e-05, + "loss": 0.0706, + "step": 32779 + }, + { + "epoch": 0.7684148221077493, + "grad_norm": 3.5613138675689697, + "learning_rate": 2.5329558715658385e-05, + "loss": 0.2485, + "step": 32780 + }, + { + "epoch": 0.7684382636825542, + "grad_norm": 1.024850845336914, + "learning_rate": 2.5324659274080274e-05, + "loss": 0.1273, + "step": 32781 + }, + { + "epoch": 0.7684617052573592, + "grad_norm": 0.4062451124191284, + "learning_rate": 2.5319760237692703e-05, + "loss": 0.0255, + "step": 32782 + }, + { + "epoch": 0.7684851468321642, + "grad_norm": 2.842600107192993, + "learning_rate": 2.531486160652221e-05, + "loss": 0.1856, + "step": 32783 + }, + { + "epoch": 0.7685085884069692, + "grad_norm": 2.527773141860962, + "learning_rate": 2.5309963380595338e-05, + "loss": 0.1946, + "step": 32784 + }, + { + "epoch": 0.7685320299817742, + "grad_norm": 2.533604860305786, + "learning_rate": 2.5305065559938747e-05, + "loss": 0.4161, + "step": 32785 + }, + { + "epoch": 0.7685554715565792, + "grad_norm": 2.6090314388275146, + "learning_rate": 2.530016814457896e-05, + "loss": 1.4045, + "step": 32786 + }, + { + "epoch": 0.7685789131313842, + "grad_norm": 1.1238255500793457, + "learning_rate": 2.5295271134542563e-05, + "loss": 0.1487, + "step": 32787 + }, + { + "epoch": 0.7686023547061892, + "grad_norm": 1.1459838151931763, + "learning_rate": 2.529037452985613e-05, + "loss": 0.0502, + "step": 32788 + }, + { + "epoch": 0.7686257962809941, + "grad_norm": 2.1677134037017822, + "learning_rate": 2.528547833054623e-05, + "loss": 0.1495, + "step": 32789 + }, + { + "epoch": 0.7686492378557992, + "grad_norm": 1.7863829135894775, + "learning_rate": 2.5280582536639385e-05, + "loss": 0.1339, + "step": 32790 + }, + { + "epoch": 0.7686726794306041, + "grad_norm": 4.551220893859863, + "learning_rate": 2.527568714816224e-05, + "loss": 0.4908, + "step": 32791 + }, + { + "epoch": 0.7686961210054092, + "grad_norm": 1.7188773155212402, + "learning_rate": 2.527079216514131e-05, + "loss": 0.1796, + "step": 32792 + }, + { + "epoch": 0.7687195625802141, + "grad_norm": 2.3515591621398926, + "learning_rate": 2.526589758760316e-05, + "loss": 0.1069, + "step": 32793 + }, + { + "epoch": 0.7687430041550192, + "grad_norm": 0.6899904608726501, + "learning_rate": 2.5261003415574326e-05, + "loss": 0.0382, + "step": 32794 + }, + { + "epoch": 0.7687664457298241, + "grad_norm": 3.1861860752105713, + "learning_rate": 2.5256109649081427e-05, + "loss": 0.2493, + "step": 32795 + }, + { + "epoch": 0.7687898873046292, + "grad_norm": 1.8473652601242065, + "learning_rate": 2.525121628815097e-05, + "loss": 0.1028, + "step": 32796 + }, + { + "epoch": 0.7688133288794341, + "grad_norm": 2.085867404937744, + "learning_rate": 2.5246323332809486e-05, + "loss": 0.2684, + "step": 32797 + }, + { + "epoch": 0.7688367704542391, + "grad_norm": 2.6143593788146973, + "learning_rate": 2.5241430783083577e-05, + "loss": 0.1995, + "step": 32798 + }, + { + "epoch": 0.7688602120290441, + "grad_norm": 2.075193166732788, + "learning_rate": 2.5236538638999774e-05, + "loss": 0.1375, + "step": 32799 + }, + { + "epoch": 0.7688836536038491, + "grad_norm": 9.05774211883545, + "learning_rate": 2.5231646900584616e-05, + "loss": 0.2155, + "step": 32800 + }, + { + "epoch": 0.7689070951786541, + "grad_norm": 2.1402103900909424, + "learning_rate": 2.5226755567864634e-05, + "loss": 0.2177, + "step": 32801 + }, + { + "epoch": 0.7689305367534591, + "grad_norm": 2.4110655784606934, + "learning_rate": 2.522186464086639e-05, + "loss": 0.2701, + "step": 32802 + }, + { + "epoch": 0.7689539783282641, + "grad_norm": 2.6419153213500977, + "learning_rate": 2.5216974119616378e-05, + "loss": 0.2213, + "step": 32803 + }, + { + "epoch": 0.7689774199030691, + "grad_norm": 2.150420904159546, + "learning_rate": 2.52120840041412e-05, + "loss": 1.4367, + "step": 32804 + }, + { + "epoch": 0.769000861477874, + "grad_norm": 2.0920825004577637, + "learning_rate": 2.520719429446735e-05, + "loss": 0.0729, + "step": 32805 + }, + { + "epoch": 0.7690243030526791, + "grad_norm": 2.536041736602783, + "learning_rate": 2.520230499062134e-05, + "loss": 1.3071, + "step": 32806 + }, + { + "epoch": 0.769047744627484, + "grad_norm": 3.76285457611084, + "learning_rate": 2.5197416092629755e-05, + "loss": 0.3153, + "step": 32807 + }, + { + "epoch": 0.7690711862022891, + "grad_norm": 4.575661659240723, + "learning_rate": 2.51925276005191e-05, + "loss": 0.3399, + "step": 32808 + }, + { + "epoch": 0.769094627777094, + "grad_norm": 1.8752272129058838, + "learning_rate": 2.5187639514315887e-05, + "loss": 0.1088, + "step": 32809 + }, + { + "epoch": 0.7691180693518991, + "grad_norm": 2.1351959705352783, + "learning_rate": 2.5182751834046612e-05, + "loss": 0.1688, + "step": 32810 + }, + { + "epoch": 0.769141510926704, + "grad_norm": 0.8945106267929077, + "learning_rate": 2.5177864559737874e-05, + "loss": 0.0676, + "step": 32811 + }, + { + "epoch": 0.7691649525015091, + "grad_norm": 2.7909748554229736, + "learning_rate": 2.5172977691416134e-05, + "loss": 0.2401, + "step": 32812 + }, + { + "epoch": 0.769188394076314, + "grad_norm": 3.3015635013580322, + "learning_rate": 2.5168091229107927e-05, + "loss": 0.2628, + "step": 32813 + }, + { + "epoch": 0.7692118356511191, + "grad_norm": 3.639713764190674, + "learning_rate": 2.5163205172839753e-05, + "loss": 0.2649, + "step": 32814 + }, + { + "epoch": 0.769235277225924, + "grad_norm": 2.665130853652954, + "learning_rate": 2.5158319522638107e-05, + "loss": 0.1609, + "step": 32815 + }, + { + "epoch": 0.769258718800729, + "grad_norm": 2.751255512237549, + "learning_rate": 2.5153434278529554e-05, + "loss": 0.1446, + "step": 32816 + }, + { + "epoch": 0.769282160375534, + "grad_norm": 1.1220217943191528, + "learning_rate": 2.514854944054057e-05, + "loss": 0.118, + "step": 32817 + }, + { + "epoch": 0.769305601950339, + "grad_norm": 1.9977312088012695, + "learning_rate": 2.5143665008697638e-05, + "loss": 0.1631, + "step": 32818 + }, + { + "epoch": 0.769329043525144, + "grad_norm": 2.755972385406494, + "learning_rate": 2.5138780983027298e-05, + "loss": 0.0696, + "step": 32819 + }, + { + "epoch": 0.769352485099949, + "grad_norm": 2.4466733932495117, + "learning_rate": 2.5133897363556057e-05, + "loss": 0.1803, + "step": 32820 + }, + { + "epoch": 0.769375926674754, + "grad_norm": 2.3638269901275635, + "learning_rate": 2.5129014150310382e-05, + "loss": 1.0486, + "step": 32821 + }, + { + "epoch": 0.769399368249559, + "grad_norm": 12.791092872619629, + "learning_rate": 2.512413134331676e-05, + "loss": 0.1999, + "step": 32822 + }, + { + "epoch": 0.769422809824364, + "grad_norm": 0.5218335390090942, + "learning_rate": 2.5119248942601726e-05, + "loss": 0.0394, + "step": 32823 + }, + { + "epoch": 0.769446251399169, + "grad_norm": 2.9099011421203613, + "learning_rate": 2.511436694819176e-05, + "loss": 0.2143, + "step": 32824 + }, + { + "epoch": 0.7694696929739739, + "grad_norm": 2.5080249309539795, + "learning_rate": 2.510948536011334e-05, + "loss": 0.0563, + "step": 32825 + }, + { + "epoch": 0.769493134548779, + "grad_norm": 2.905632972717285, + "learning_rate": 2.5104604178392953e-05, + "loss": 0.1962, + "step": 32826 + }, + { + "epoch": 0.7695165761235839, + "grad_norm": 2.7227046489715576, + "learning_rate": 2.5099723403057107e-05, + "loss": 0.1717, + "step": 32827 + }, + { + "epoch": 0.769540017698389, + "grad_norm": 2.629218816757202, + "learning_rate": 2.509484303413222e-05, + "loss": 0.0724, + "step": 32828 + }, + { + "epoch": 0.769563459273194, + "grad_norm": 1.2467529773712158, + "learning_rate": 2.5089963071644863e-05, + "loss": 0.1215, + "step": 32829 + }, + { + "epoch": 0.769586900847999, + "grad_norm": 1.1876575946807861, + "learning_rate": 2.5085083515621442e-05, + "loss": 0.0671, + "step": 32830 + }, + { + "epoch": 0.769610342422804, + "grad_norm": 1.7448407411575317, + "learning_rate": 2.508020436608849e-05, + "loss": 0.1894, + "step": 32831 + }, + { + "epoch": 0.769633783997609, + "grad_norm": 4.038880348205566, + "learning_rate": 2.5075325623072454e-05, + "loss": 0.1916, + "step": 32832 + }, + { + "epoch": 0.769657225572414, + "grad_norm": 2.7553188800811768, + "learning_rate": 2.5070447286599806e-05, + "loss": 1.079, + "step": 32833 + }, + { + "epoch": 0.7696806671472189, + "grad_norm": 1.3308066129684448, + "learning_rate": 2.5065569356697027e-05, + "loss": 0.0743, + "step": 32834 + }, + { + "epoch": 0.769704108722024, + "grad_norm": 0.41753149032592773, + "learning_rate": 2.5060691833390537e-05, + "loss": 0.0305, + "step": 32835 + }, + { + "epoch": 0.7697275502968289, + "grad_norm": 2.395078659057617, + "learning_rate": 2.5055814716706872e-05, + "loss": 1.7144, + "step": 32836 + }, + { + "epoch": 0.769750991871634, + "grad_norm": 3.15350341796875, + "learning_rate": 2.5050938006672463e-05, + "loss": 0.2922, + "step": 32837 + }, + { + "epoch": 0.7697744334464389, + "grad_norm": 4.29312801361084, + "learning_rate": 2.5046061703313785e-05, + "loss": 0.2861, + "step": 32838 + }, + { + "epoch": 0.769797875021244, + "grad_norm": 0.6934155225753784, + "learning_rate": 2.5041185806657264e-05, + "loss": 0.0414, + "step": 32839 + }, + { + "epoch": 0.7698213165960489, + "grad_norm": 2.8638298511505127, + "learning_rate": 2.5036310316729384e-05, + "loss": 0.3236, + "step": 32840 + }, + { + "epoch": 0.769844758170854, + "grad_norm": 4.097775936126709, + "learning_rate": 2.5031435233556567e-05, + "loss": 0.3043, + "step": 32841 + }, + { + "epoch": 0.7698681997456589, + "grad_norm": 2.4246888160705566, + "learning_rate": 2.5026560557165292e-05, + "loss": 0.121, + "step": 32842 + }, + { + "epoch": 0.769891641320464, + "grad_norm": 3.042937755584717, + "learning_rate": 2.5021686287582024e-05, + "loss": 1.3323, + "step": 32843 + }, + { + "epoch": 0.7699150828952689, + "grad_norm": 4.286488056182861, + "learning_rate": 2.5016812424833204e-05, + "loss": 0.2152, + "step": 32844 + }, + { + "epoch": 0.7699385244700739, + "grad_norm": 3.115849256515503, + "learning_rate": 2.501193896894527e-05, + "loss": 1.7283, + "step": 32845 + }, + { + "epoch": 0.7699619660448789, + "grad_norm": 1.225548267364502, + "learning_rate": 2.500706591994466e-05, + "loss": 0.0931, + "step": 32846 + }, + { + "epoch": 0.7699854076196839, + "grad_norm": 2.4221222400665283, + "learning_rate": 2.5002193277857822e-05, + "loss": 0.1691, + "step": 32847 + }, + { + "epoch": 0.7700088491944889, + "grad_norm": 0.3274119198322296, + "learning_rate": 2.499732104271115e-05, + "loss": 0.0153, + "step": 32848 + }, + { + "epoch": 0.7700322907692939, + "grad_norm": 2.7371468544006348, + "learning_rate": 2.4992449214531167e-05, + "loss": 0.1575, + "step": 32849 + }, + { + "epoch": 0.7700557323440989, + "grad_norm": 0.4062751531600952, + "learning_rate": 2.4987577793344263e-05, + "loss": 0.0155, + "step": 32850 + }, + { + "epoch": 0.7700791739189039, + "grad_norm": 3.518779754638672, + "learning_rate": 2.4982706779176867e-05, + "loss": 0.3785, + "step": 32851 + }, + { + "epoch": 0.7701026154937088, + "grad_norm": 8.69038200378418, + "learning_rate": 2.4977836172055412e-05, + "loss": 0.3229, + "step": 32852 + }, + { + "epoch": 0.7701260570685139, + "grad_norm": 2.649653673171997, + "learning_rate": 2.4972965972006303e-05, + "loss": 0.2051, + "step": 32853 + }, + { + "epoch": 0.7701494986433188, + "grad_norm": 2.4351210594177246, + "learning_rate": 2.496809617905601e-05, + "loss": 0.1873, + "step": 32854 + }, + { + "epoch": 0.7701729402181239, + "grad_norm": 0.9351732134819031, + "learning_rate": 2.4963226793230908e-05, + "loss": 0.0687, + "step": 32855 + }, + { + "epoch": 0.7701963817929288, + "grad_norm": 2.5021770000457764, + "learning_rate": 2.4958357814557488e-05, + "loss": 0.22, + "step": 32856 + }, + { + "epoch": 0.7702198233677339, + "grad_norm": 3.3080129623413086, + "learning_rate": 2.4953489243062122e-05, + "loss": 0.3655, + "step": 32857 + }, + { + "epoch": 0.7702432649425388, + "grad_norm": 0.8300984501838684, + "learning_rate": 2.494862107877123e-05, + "loss": 0.0596, + "step": 32858 + }, + { + "epoch": 0.7702667065173439, + "grad_norm": 2.530972480773926, + "learning_rate": 2.494375332171124e-05, + "loss": 0.1552, + "step": 32859 + }, + { + "epoch": 0.7702901480921488, + "grad_norm": 0.8558405637741089, + "learning_rate": 2.4938885971908523e-05, + "loss": 0.0725, + "step": 32860 + }, + { + "epoch": 0.7703135896669538, + "grad_norm": 4.581014156341553, + "learning_rate": 2.4934019029389544e-05, + "loss": 0.5635, + "step": 32861 + }, + { + "epoch": 0.7703370312417588, + "grad_norm": 1.0116944313049316, + "learning_rate": 2.4929152494180686e-05, + "loss": 0.0985, + "step": 32862 + }, + { + "epoch": 0.7703604728165638, + "grad_norm": 2.6409876346588135, + "learning_rate": 2.492428636630837e-05, + "loss": 0.1685, + "step": 32863 + }, + { + "epoch": 0.7703839143913688, + "grad_norm": 4.073922634124756, + "learning_rate": 2.4919420645798975e-05, + "loss": 0.3233, + "step": 32864 + }, + { + "epoch": 0.7704073559661738, + "grad_norm": 2.175618886947632, + "learning_rate": 2.4914555332678923e-05, + "loss": 0.2193, + "step": 32865 + }, + { + "epoch": 0.7704307975409788, + "grad_norm": 0.5481880903244019, + "learning_rate": 2.4909690426974576e-05, + "loss": 0.0625, + "step": 32866 + }, + { + "epoch": 0.7704542391157838, + "grad_norm": 1.818938136100769, + "learning_rate": 2.490482592871235e-05, + "loss": 0.1468, + "step": 32867 + }, + { + "epoch": 0.7704776806905888, + "grad_norm": 4.533562660217285, + "learning_rate": 2.4899961837918685e-05, + "loss": 0.3503, + "step": 32868 + }, + { + "epoch": 0.7705011222653938, + "grad_norm": 2.088286876678467, + "learning_rate": 2.4895098154619945e-05, + "loss": 0.1213, + "step": 32869 + }, + { + "epoch": 0.7705245638401987, + "grad_norm": 1.1434036493301392, + "learning_rate": 2.48902348788425e-05, + "loss": 0.1147, + "step": 32870 + }, + { + "epoch": 0.7705480054150038, + "grad_norm": 1.824623942375183, + "learning_rate": 2.488537201061276e-05, + "loss": 0.1155, + "step": 32871 + }, + { + "epoch": 0.7705714469898087, + "grad_norm": 3.1347734928131104, + "learning_rate": 2.4880509549957108e-05, + "loss": 0.2104, + "step": 32872 + }, + { + "epoch": 0.7705948885646138, + "grad_norm": 2.0610995292663574, + "learning_rate": 2.487564749690189e-05, + "loss": 0.1155, + "step": 32873 + }, + { + "epoch": 0.7706183301394187, + "grad_norm": 0.9010323286056519, + "learning_rate": 2.4870785851473556e-05, + "loss": 0.0854, + "step": 32874 + }, + { + "epoch": 0.7706417717142238, + "grad_norm": 1.5391933917999268, + "learning_rate": 2.4865924613698444e-05, + "loss": 0.1059, + "step": 32875 + }, + { + "epoch": 0.7706652132890287, + "grad_norm": 3.0758211612701416, + "learning_rate": 2.4861063783602933e-05, + "loss": 0.2258, + "step": 32876 + }, + { + "epoch": 0.7706886548638338, + "grad_norm": 1.9757405519485474, + "learning_rate": 2.4856203361213415e-05, + "loss": 0.1519, + "step": 32877 + }, + { + "epoch": 0.7707120964386387, + "grad_norm": 2.298382043838501, + "learning_rate": 2.4851343346556244e-05, + "loss": 0.1526, + "step": 32878 + }, + { + "epoch": 0.7707355380134437, + "grad_norm": 4.307348728179932, + "learning_rate": 2.4846483739657766e-05, + "loss": 0.4114, + "step": 32879 + }, + { + "epoch": 0.7707589795882488, + "grad_norm": 0.7361313700675964, + "learning_rate": 2.4841624540544384e-05, + "loss": 0.0482, + "step": 32880 + }, + { + "epoch": 0.7707824211630537, + "grad_norm": 2.096614360809326, + "learning_rate": 2.4836765749242485e-05, + "loss": 0.2025, + "step": 32881 + }, + { + "epoch": 0.7708058627378588, + "grad_norm": 3.7324371337890625, + "learning_rate": 2.4831907365778416e-05, + "loss": 0.414, + "step": 32882 + }, + { + "epoch": 0.7708293043126637, + "grad_norm": 0.9127450585365295, + "learning_rate": 2.4827049390178526e-05, + "loss": 0.0537, + "step": 32883 + }, + { + "epoch": 0.7708527458874688, + "grad_norm": 0.9194483757019043, + "learning_rate": 2.4822191822469178e-05, + "loss": 0.0565, + "step": 32884 + }, + { + "epoch": 0.7708761874622737, + "grad_norm": 0.7861972451210022, + "learning_rate": 2.4817334662676728e-05, + "loss": 0.0461, + "step": 32885 + }, + { + "epoch": 0.7708996290370788, + "grad_norm": 2.8811702728271484, + "learning_rate": 2.4812477910827515e-05, + "loss": 0.2223, + "step": 32886 + }, + { + "epoch": 0.7709230706118837, + "grad_norm": 1.9102506637573242, + "learning_rate": 2.4807621566947937e-05, + "loss": 0.1511, + "step": 32887 + }, + { + "epoch": 0.7709465121866887, + "grad_norm": 3.6011459827423096, + "learning_rate": 2.4802765631064316e-05, + "loss": 0.3401, + "step": 32888 + }, + { + "epoch": 0.7709699537614937, + "grad_norm": 1.9036482572555542, + "learning_rate": 2.4797910103203005e-05, + "loss": 0.1646, + "step": 32889 + }, + { + "epoch": 0.7709933953362987, + "grad_norm": 5.186609268188477, + "learning_rate": 2.479305498339034e-05, + "loss": 0.2725, + "step": 32890 + }, + { + "epoch": 0.7710168369111037, + "grad_norm": 1.8834152221679688, + "learning_rate": 2.4788200271652652e-05, + "loss": 0.206, + "step": 32891 + }, + { + "epoch": 0.7710402784859087, + "grad_norm": 3.3874692916870117, + "learning_rate": 2.4783345968016303e-05, + "loss": 0.1661, + "step": 32892 + }, + { + "epoch": 0.7710637200607137, + "grad_norm": 3.2185001373291016, + "learning_rate": 2.4778492072507665e-05, + "loss": 0.317, + "step": 32893 + }, + { + "epoch": 0.7710871616355187, + "grad_norm": 2.0263147354125977, + "learning_rate": 2.4773638585153047e-05, + "loss": 0.1555, + "step": 32894 + }, + { + "epoch": 0.7711106032103237, + "grad_norm": 3.194985866546631, + "learning_rate": 2.4768785505978775e-05, + "loss": 1.5785, + "step": 32895 + }, + { + "epoch": 0.7711340447851287, + "grad_norm": 1.5634934902191162, + "learning_rate": 2.4763932835011184e-05, + "loss": 0.1316, + "step": 32896 + }, + { + "epoch": 0.7711574863599336, + "grad_norm": 0.44024163484573364, + "learning_rate": 2.4759080572276615e-05, + "loss": 0.0366, + "step": 32897 + }, + { + "epoch": 0.7711809279347387, + "grad_norm": 2.6348347663879395, + "learning_rate": 2.4754228717801365e-05, + "loss": 0.291, + "step": 32898 + }, + { + "epoch": 0.7712043695095436, + "grad_norm": 1.6693729162216187, + "learning_rate": 2.4749377271611805e-05, + "loss": 0.1122, + "step": 32899 + }, + { + "epoch": 0.7712278110843487, + "grad_norm": 2.3892016410827637, + "learning_rate": 2.4744526233734243e-05, + "loss": 0.2031, + "step": 32900 + }, + { + "epoch": 0.7712512526591536, + "grad_norm": 1.2178691625595093, + "learning_rate": 2.4739675604195013e-05, + "loss": 0.0503, + "step": 32901 + }, + { + "epoch": 0.7712746942339587, + "grad_norm": 2.522657871246338, + "learning_rate": 2.47348253830204e-05, + "loss": 0.2878, + "step": 32902 + }, + { + "epoch": 0.7712981358087636, + "grad_norm": 2.6315624713897705, + "learning_rate": 2.472997557023672e-05, + "loss": 0.1617, + "step": 32903 + }, + { + "epoch": 0.7713215773835687, + "grad_norm": 1.7395484447479248, + "learning_rate": 2.4725126165870338e-05, + "loss": 0.2224, + "step": 32904 + }, + { + "epoch": 0.7713450189583736, + "grad_norm": 0.8370128273963928, + "learning_rate": 2.4720277169947503e-05, + "loss": 0.0528, + "step": 32905 + }, + { + "epoch": 0.7713684605331786, + "grad_norm": 0.6511896252632141, + "learning_rate": 2.4715428582494594e-05, + "loss": 0.0396, + "step": 32906 + }, + { + "epoch": 0.7713919021079836, + "grad_norm": 1.483017921447754, + "learning_rate": 2.4710580403537886e-05, + "loss": 0.0747, + "step": 32907 + }, + { + "epoch": 0.7714153436827886, + "grad_norm": 3.695108652114868, + "learning_rate": 2.4705732633103684e-05, + "loss": 0.4686, + "step": 32908 + }, + { + "epoch": 0.7714387852575936, + "grad_norm": 3.7590060234069824, + "learning_rate": 2.4700885271218298e-05, + "loss": 0.3951, + "step": 32909 + }, + { + "epoch": 0.7714622268323986, + "grad_norm": 3.0764079093933105, + "learning_rate": 2.4696038317908022e-05, + "loss": 0.3441, + "step": 32910 + }, + { + "epoch": 0.7714856684072036, + "grad_norm": 3.474428415298462, + "learning_rate": 2.469119177319913e-05, + "loss": 0.1463, + "step": 32911 + }, + { + "epoch": 0.7715091099820086, + "grad_norm": 2.858069896697998, + "learning_rate": 2.468634563711798e-05, + "loss": 0.2321, + "step": 32912 + }, + { + "epoch": 0.7715325515568136, + "grad_norm": 2.986985206604004, + "learning_rate": 2.4681499909690842e-05, + "loss": 0.2926, + "step": 32913 + }, + { + "epoch": 0.7715559931316186, + "grad_norm": 7.036481857299805, + "learning_rate": 2.4676654590943992e-05, + "loss": 0.4794, + "step": 32914 + }, + { + "epoch": 0.7715794347064235, + "grad_norm": 3.6305599212646484, + "learning_rate": 2.4671809680903712e-05, + "loss": 0.2765, + "step": 32915 + }, + { + "epoch": 0.7716028762812286, + "grad_norm": 0.5010088682174683, + "learning_rate": 2.4666965179596335e-05, + "loss": 0.0575, + "step": 32916 + }, + { + "epoch": 0.7716263178560335, + "grad_norm": 2.4973807334899902, + "learning_rate": 2.466212108704812e-05, + "loss": 0.1837, + "step": 32917 + }, + { + "epoch": 0.7716497594308386, + "grad_norm": 3.3544516563415527, + "learning_rate": 2.4657277403285328e-05, + "loss": 0.3683, + "step": 32918 + }, + { + "epoch": 0.7716732010056435, + "grad_norm": 2.0673367977142334, + "learning_rate": 2.46524341283343e-05, + "loss": 0.1261, + "step": 32919 + }, + { + "epoch": 0.7716966425804486, + "grad_norm": 4.1027984619140625, + "learning_rate": 2.464759126222128e-05, + "loss": 0.0828, + "step": 32920 + }, + { + "epoch": 0.7717200841552535, + "grad_norm": 2.384937047958374, + "learning_rate": 2.464274880497255e-05, + "loss": 0.1584, + "step": 32921 + }, + { + "epoch": 0.7717435257300586, + "grad_norm": 2.6648383140563965, + "learning_rate": 2.4637906756614382e-05, + "loss": 0.6272, + "step": 32922 + }, + { + "epoch": 0.7717669673048635, + "grad_norm": 1.795088291168213, + "learning_rate": 2.463306511717305e-05, + "loss": 0.0576, + "step": 32923 + }, + { + "epoch": 0.7717904088796685, + "grad_norm": 1.8363200426101685, + "learning_rate": 2.4628223886674806e-05, + "loss": 0.334, + "step": 32924 + }, + { + "epoch": 0.7718138504544735, + "grad_norm": 2.7961976528167725, + "learning_rate": 2.4623383065145957e-05, + "loss": 0.2389, + "step": 32925 + }, + { + "epoch": 0.7718372920292785, + "grad_norm": 0.6690612435340881, + "learning_rate": 2.4618542652612753e-05, + "loss": 0.0717, + "step": 32926 + }, + { + "epoch": 0.7718607336040835, + "grad_norm": 0.7090986371040344, + "learning_rate": 2.4613702649101432e-05, + "loss": 0.0608, + "step": 32927 + }, + { + "epoch": 0.7718841751788885, + "grad_norm": 1.4071400165557861, + "learning_rate": 2.4608863054638297e-05, + "loss": 0.1596, + "step": 32928 + }, + { + "epoch": 0.7719076167536935, + "grad_norm": 2.212710380554199, + "learning_rate": 2.4604023869249605e-05, + "loss": 0.1994, + "step": 32929 + }, + { + "epoch": 0.7719310583284985, + "grad_norm": 0.614509105682373, + "learning_rate": 2.4599185092961563e-05, + "loss": 0.0455, + "step": 32930 + }, + { + "epoch": 0.7719544999033036, + "grad_norm": 3.450087070465088, + "learning_rate": 2.4594346725800498e-05, + "loss": 0.3216, + "step": 32931 + }, + { + "epoch": 0.7719779414781085, + "grad_norm": 2.921668767929077, + "learning_rate": 2.4589508767792623e-05, + "loss": 0.2373, + "step": 32932 + }, + { + "epoch": 0.7720013830529135, + "grad_norm": 0.6952066421508789, + "learning_rate": 2.4584671218964196e-05, + "loss": 0.0459, + "step": 32933 + }, + { + "epoch": 0.7720248246277185, + "grad_norm": 1.5073214769363403, + "learning_rate": 2.457983407934146e-05, + "loss": 0.1024, + "step": 32934 + }, + { + "epoch": 0.7720482662025235, + "grad_norm": 1.5113424062728882, + "learning_rate": 2.457499734895066e-05, + "loss": 0.0746, + "step": 32935 + }, + { + "epoch": 0.7720717077773285, + "grad_norm": 3.3703694343566895, + "learning_rate": 2.457016102781803e-05, + "loss": 0.3275, + "step": 32936 + }, + { + "epoch": 0.7720951493521335, + "grad_norm": 0.4978719651699066, + "learning_rate": 2.456532511596984e-05, + "loss": 0.0227, + "step": 32937 + }, + { + "epoch": 0.7721185909269385, + "grad_norm": 2.726776599884033, + "learning_rate": 2.4560489613432326e-05, + "loss": 0.3417, + "step": 32938 + }, + { + "epoch": 0.7721420325017435, + "grad_norm": 3.3694827556610107, + "learning_rate": 2.4555654520231685e-05, + "loss": 0.3609, + "step": 32939 + }, + { + "epoch": 0.7721654740765485, + "grad_norm": 2.0874698162078857, + "learning_rate": 2.4550819836394224e-05, + "loss": 1.5249, + "step": 32940 + }, + { + "epoch": 0.7721889156513535, + "grad_norm": 0.9596160650253296, + "learning_rate": 2.4545985561946127e-05, + "loss": 0.0798, + "step": 32941 + }, + { + "epoch": 0.7722123572261584, + "grad_norm": 2.7428832054138184, + "learning_rate": 2.4541151696913634e-05, + "loss": 0.1084, + "step": 32942 + }, + { + "epoch": 0.7722357988009635, + "grad_norm": 4.011343479156494, + "learning_rate": 2.4536318241322943e-05, + "loss": 0.4793, + "step": 32943 + }, + { + "epoch": 0.7722592403757684, + "grad_norm": 2.7878551483154297, + "learning_rate": 2.4531485195200345e-05, + "loss": 0.1398, + "step": 32944 + }, + { + "epoch": 0.7722826819505735, + "grad_norm": 1.6358758211135864, + "learning_rate": 2.4526652558572017e-05, + "loss": 0.1128, + "step": 32945 + }, + { + "epoch": 0.7723061235253784, + "grad_norm": 10.690817832946777, + "learning_rate": 2.4521820331464207e-05, + "loss": 0.4041, + "step": 32946 + }, + { + "epoch": 0.7723295651001835, + "grad_norm": 2.3251047134399414, + "learning_rate": 2.4516988513903115e-05, + "loss": 0.1842, + "step": 32947 + }, + { + "epoch": 0.7723530066749884, + "grad_norm": 2.3547897338867188, + "learning_rate": 2.451215710591498e-05, + "loss": 0.1607, + "step": 32948 + }, + { + "epoch": 0.7723764482497935, + "grad_norm": 1.3600883483886719, + "learning_rate": 2.4507326107525963e-05, + "loss": 0.121, + "step": 32949 + }, + { + "epoch": 0.7723998898245984, + "grad_norm": 2.4209322929382324, + "learning_rate": 2.4502495518762348e-05, + "loss": 0.1854, + "step": 32950 + }, + { + "epoch": 0.7724233313994034, + "grad_norm": 2.0815329551696777, + "learning_rate": 2.449766533965031e-05, + "loss": 0.1664, + "step": 32951 + }, + { + "epoch": 0.7724467729742084, + "grad_norm": 2.7251670360565186, + "learning_rate": 2.4492835570216044e-05, + "loss": 0.2256, + "step": 32952 + }, + { + "epoch": 0.7724702145490134, + "grad_norm": 1.9874476194381714, + "learning_rate": 2.4488006210485793e-05, + "loss": 0.159, + "step": 32953 + }, + { + "epoch": 0.7724936561238184, + "grad_norm": 3.849680185317993, + "learning_rate": 2.4483177260485755e-05, + "loss": 0.3724, + "step": 32954 + }, + { + "epoch": 0.7725170976986234, + "grad_norm": 0.8510468602180481, + "learning_rate": 2.4478348720242117e-05, + "loss": 0.0657, + "step": 32955 + }, + { + "epoch": 0.7725405392734284, + "grad_norm": 1.8443291187286377, + "learning_rate": 2.4473520589781042e-05, + "loss": 0.2542, + "step": 32956 + }, + { + "epoch": 0.7725639808482334, + "grad_norm": 3.005929470062256, + "learning_rate": 2.4468692869128808e-05, + "loss": 0.2081, + "step": 32957 + }, + { + "epoch": 0.7725874224230384, + "grad_norm": 1.437221884727478, + "learning_rate": 2.446386555831157e-05, + "loss": 0.1934, + "step": 32958 + }, + { + "epoch": 0.7726108639978434, + "grad_norm": 2.5545265674591064, + "learning_rate": 2.4459038657355516e-05, + "loss": 0.2215, + "step": 32959 + }, + { + "epoch": 0.7726343055726483, + "grad_norm": 5.045050144195557, + "learning_rate": 2.445421216628685e-05, + "loss": 0.5308, + "step": 32960 + }, + { + "epoch": 0.7726577471474534, + "grad_norm": 0.7165827751159668, + "learning_rate": 2.444938608513172e-05, + "loss": 0.0559, + "step": 32961 + }, + { + "epoch": 0.7726811887222583, + "grad_norm": 2.7623071670532227, + "learning_rate": 2.4444560413916372e-05, + "loss": 0.1367, + "step": 32962 + }, + { + "epoch": 0.7727046302970634, + "grad_norm": 1.2569427490234375, + "learning_rate": 2.4439735152666963e-05, + "loss": 0.1322, + "step": 32963 + }, + { + "epoch": 0.7727280718718683, + "grad_norm": 2.595550775527954, + "learning_rate": 2.4434910301409654e-05, + "loss": 0.1833, + "step": 32964 + }, + { + "epoch": 0.7727515134466734, + "grad_norm": 4.228603839874268, + "learning_rate": 2.443008586017067e-05, + "loss": 0.4036, + "step": 32965 + }, + { + "epoch": 0.7727749550214783, + "grad_norm": 5.120110988616943, + "learning_rate": 2.4425261828976165e-05, + "loss": 0.3732, + "step": 32966 + }, + { + "epoch": 0.7727983965962834, + "grad_norm": 3.4098119735717773, + "learning_rate": 2.4420438207852315e-05, + "loss": 0.3057, + "step": 32967 + }, + { + "epoch": 0.7728218381710883, + "grad_norm": 0.7544739842414856, + "learning_rate": 2.4415614996825264e-05, + "loss": 0.1235, + "step": 32968 + }, + { + "epoch": 0.7728452797458933, + "grad_norm": 1.774844765663147, + "learning_rate": 2.441079219592124e-05, + "loss": 0.0711, + "step": 32969 + }, + { + "epoch": 0.7728687213206983, + "grad_norm": 2.851059675216675, + "learning_rate": 2.4405969805166384e-05, + "loss": 0.1573, + "step": 32970 + }, + { + "epoch": 0.7728921628955033, + "grad_norm": 0.8489516377449036, + "learning_rate": 2.4401147824586856e-05, + "loss": 0.0834, + "step": 32971 + }, + { + "epoch": 0.7729156044703083, + "grad_norm": 2.6087405681610107, + "learning_rate": 2.4396326254208835e-05, + "loss": 0.1399, + "step": 32972 + }, + { + "epoch": 0.7729390460451133, + "grad_norm": 0.5754113793373108, + "learning_rate": 2.4391505094058464e-05, + "loss": 0.0265, + "step": 32973 + }, + { + "epoch": 0.7729624876199183, + "grad_norm": 3.3871657848358154, + "learning_rate": 2.438668434416189e-05, + "loss": 0.2747, + "step": 32974 + }, + { + "epoch": 0.7729859291947233, + "grad_norm": 3.37192964553833, + "learning_rate": 2.4381864004545317e-05, + "loss": 0.2599, + "step": 32975 + }, + { + "epoch": 0.7730093707695282, + "grad_norm": 3.163090705871582, + "learning_rate": 2.4377044075234856e-05, + "loss": 1.2497, + "step": 32976 + }, + { + "epoch": 0.7730328123443333, + "grad_norm": 1.4997423887252808, + "learning_rate": 2.4372224556256697e-05, + "loss": 0.0938, + "step": 32977 + }, + { + "epoch": 0.7730562539191382, + "grad_norm": 2.851145029067993, + "learning_rate": 2.436740544763698e-05, + "loss": 0.3424, + "step": 32978 + }, + { + "epoch": 0.7730796954939433, + "grad_norm": 4.346548557281494, + "learning_rate": 2.436258674940185e-05, + "loss": 0.3854, + "step": 32979 + }, + { + "epoch": 0.7731031370687482, + "grad_norm": 3.198549270629883, + "learning_rate": 2.4357768461577447e-05, + "loss": 0.1949, + "step": 32980 + }, + { + "epoch": 0.7731265786435533, + "grad_norm": 3.307849884033203, + "learning_rate": 2.4352950584189895e-05, + "loss": 0.3819, + "step": 32981 + }, + { + "epoch": 0.7731500202183583, + "grad_norm": 3.2672765254974365, + "learning_rate": 2.434813311726538e-05, + "loss": 0.2554, + "step": 32982 + }, + { + "epoch": 0.7731734617931633, + "grad_norm": 2.767380475997925, + "learning_rate": 2.4343316060830025e-05, + "loss": 0.126, + "step": 32983 + }, + { + "epoch": 0.7731969033679683, + "grad_norm": 0.6463024616241455, + "learning_rate": 2.4338499414909966e-05, + "loss": 0.0621, + "step": 32984 + }, + { + "epoch": 0.7732203449427733, + "grad_norm": 3.793889045715332, + "learning_rate": 2.433368317953134e-05, + "loss": 0.4005, + "step": 32985 + }, + { + "epoch": 0.7732437865175783, + "grad_norm": 3.4446768760681152, + "learning_rate": 2.4328867354720275e-05, + "loss": 0.3459, + "step": 32986 + }, + { + "epoch": 0.7732672280923832, + "grad_norm": 8.78866958618164, + "learning_rate": 2.432405194050287e-05, + "loss": 0.3468, + "step": 32987 + }, + { + "epoch": 0.7732906696671883, + "grad_norm": 4.783642768859863, + "learning_rate": 2.43192369369053e-05, + "loss": 0.5565, + "step": 32988 + }, + { + "epoch": 0.7733141112419932, + "grad_norm": 1.9910056591033936, + "learning_rate": 2.4314422343953692e-05, + "loss": 0.1511, + "step": 32989 + }, + { + "epoch": 0.7733375528167983, + "grad_norm": 2.0768344402313232, + "learning_rate": 2.4309608161674158e-05, + "loss": 0.1467, + "step": 32990 + }, + { + "epoch": 0.7733609943916032, + "grad_norm": 2.276195526123047, + "learning_rate": 2.4304794390092823e-05, + "loss": 1.321, + "step": 32991 + }, + { + "epoch": 0.7733844359664083, + "grad_norm": 3.3285858631134033, + "learning_rate": 2.429998102923581e-05, + "loss": 0.4088, + "step": 32992 + }, + { + "epoch": 0.7734078775412132, + "grad_norm": 2.0221211910247803, + "learning_rate": 2.4295168079129215e-05, + "loss": 0.1431, + "step": 32993 + }, + { + "epoch": 0.7734313191160183, + "grad_norm": 0.2813664972782135, + "learning_rate": 2.4290355539799136e-05, + "loss": 0.014, + "step": 32994 + }, + { + "epoch": 0.7734547606908232, + "grad_norm": 2.881570339202881, + "learning_rate": 2.428554341127176e-05, + "loss": 0.2679, + "step": 32995 + }, + { + "epoch": 0.7734782022656282, + "grad_norm": 4.321974277496338, + "learning_rate": 2.4280731693573143e-05, + "loss": 0.3199, + "step": 32996 + }, + { + "epoch": 0.7735016438404332, + "grad_norm": 1.3750452995300293, + "learning_rate": 2.4275920386729413e-05, + "loss": 0.0915, + "step": 32997 + }, + { + "epoch": 0.7735250854152382, + "grad_norm": 2.351496934890747, + "learning_rate": 2.4271109490766665e-05, + "loss": 0.1899, + "step": 32998 + }, + { + "epoch": 0.7735485269900432, + "grad_norm": 1.1163437366485596, + "learning_rate": 2.426629900571097e-05, + "loss": 0.1029, + "step": 32999 + }, + { + "epoch": 0.7735719685648482, + "grad_norm": 2.0904464721679688, + "learning_rate": 2.4261488931588504e-05, + "loss": 0.0675, + "step": 33000 + }, + { + "epoch": 0.7735954101396532, + "grad_norm": 5.568660736083984, + "learning_rate": 2.4256679268425297e-05, + "loss": 0.7064, + "step": 33001 + }, + { + "epoch": 0.7736188517144582, + "grad_norm": 2.6532092094421387, + "learning_rate": 2.4251870016247503e-05, + "loss": 1.5491, + "step": 33002 + }, + { + "epoch": 0.7736422932892631, + "grad_norm": 2.373955488204956, + "learning_rate": 2.42470611750812e-05, + "loss": 0.1604, + "step": 33003 + }, + { + "epoch": 0.7736657348640682, + "grad_norm": 0.46037471294403076, + "learning_rate": 2.4242252744952465e-05, + "loss": 0.0289, + "step": 33004 + }, + { + "epoch": 0.7736891764388731, + "grad_norm": 2.2397220134735107, + "learning_rate": 2.4237444725887405e-05, + "loss": 0.1893, + "step": 33005 + }, + { + "epoch": 0.7737126180136782, + "grad_norm": 3.6467676162719727, + "learning_rate": 2.4232637117912062e-05, + "loss": 0.4077, + "step": 33006 + }, + { + "epoch": 0.7737360595884831, + "grad_norm": 1.7272783517837524, + "learning_rate": 2.4227829921052593e-05, + "loss": 0.1038, + "step": 33007 + }, + { + "epoch": 0.7737595011632882, + "grad_norm": 2.173926830291748, + "learning_rate": 2.4223023135335053e-05, + "loss": 0.0987, + "step": 33008 + }, + { + "epoch": 0.7737829427380931, + "grad_norm": 3.293325424194336, + "learning_rate": 2.4218216760785517e-05, + "loss": 0.1238, + "step": 33009 + }, + { + "epoch": 0.7738063843128982, + "grad_norm": 4.641502380371094, + "learning_rate": 2.421341079743007e-05, + "loss": 0.4881, + "step": 33010 + }, + { + "epoch": 0.7738298258877031, + "grad_norm": 1.3704264163970947, + "learning_rate": 2.420860524529478e-05, + "loss": 0.0942, + "step": 33011 + }, + { + "epoch": 0.7738532674625082, + "grad_norm": 2.5716965198516846, + "learning_rate": 2.4203800104405705e-05, + "loss": 0.2064, + "step": 33012 + }, + { + "epoch": 0.7738767090373131, + "grad_norm": 2.8353641033172607, + "learning_rate": 2.4198995374788948e-05, + "loss": 0.2017, + "step": 33013 + }, + { + "epoch": 0.7739001506121181, + "grad_norm": 2.118422269821167, + "learning_rate": 2.4194191056470594e-05, + "loss": 0.2125, + "step": 33014 + }, + { + "epoch": 0.7739235921869231, + "grad_norm": 0.6268172860145569, + "learning_rate": 2.4189387149476694e-05, + "loss": 0.0499, + "step": 33015 + }, + { + "epoch": 0.7739470337617281, + "grad_norm": 2.985553026199341, + "learning_rate": 2.4184583653833303e-05, + "loss": 0.2197, + "step": 33016 + }, + { + "epoch": 0.7739704753365331, + "grad_norm": 1.0115232467651367, + "learning_rate": 2.4179780569566503e-05, + "loss": 0.062, + "step": 33017 + }, + { + "epoch": 0.7739939169113381, + "grad_norm": 0.7430335879325867, + "learning_rate": 2.4174977896702345e-05, + "loss": 0.0531, + "step": 33018 + }, + { + "epoch": 0.7740173584861431, + "grad_norm": 0.2929348945617676, + "learning_rate": 2.4170175635266856e-05, + "loss": 0.0103, + "step": 33019 + }, + { + "epoch": 0.7740408000609481, + "grad_norm": 1.4288591146469116, + "learning_rate": 2.416537378528615e-05, + "loss": 0.056, + "step": 33020 + }, + { + "epoch": 0.774064241635753, + "grad_norm": 3.4916937351226807, + "learning_rate": 2.416057234678627e-05, + "loss": 0.1844, + "step": 33021 + }, + { + "epoch": 0.7740876832105581, + "grad_norm": 2.3484208583831787, + "learning_rate": 2.415577131979325e-05, + "loss": 0.2161, + "step": 33022 + }, + { + "epoch": 0.774111124785363, + "grad_norm": 0.5712521076202393, + "learning_rate": 2.4150970704333155e-05, + "loss": 0.0346, + "step": 33023 + }, + { + "epoch": 0.7741345663601681, + "grad_norm": 2.005810260772705, + "learning_rate": 2.414617050043201e-05, + "loss": 0.0937, + "step": 33024 + }, + { + "epoch": 0.774158007934973, + "grad_norm": 2.442749261856079, + "learning_rate": 2.414137070811586e-05, + "loss": 0.1692, + "step": 33025 + }, + { + "epoch": 0.7741814495097781, + "grad_norm": 1.6385600566864014, + "learning_rate": 2.4136571327410763e-05, + "loss": 0.0778, + "step": 33026 + }, + { + "epoch": 0.774204891084583, + "grad_norm": 3.7507710456848145, + "learning_rate": 2.4131772358342796e-05, + "loss": 0.2525, + "step": 33027 + }, + { + "epoch": 0.7742283326593881, + "grad_norm": 0.18176989257335663, + "learning_rate": 2.4126973800937957e-05, + "loss": 0.0116, + "step": 33028 + }, + { + "epoch": 0.774251774234193, + "grad_norm": 0.27429887652397156, + "learning_rate": 2.41221756552223e-05, + "loss": 0.0178, + "step": 33029 + }, + { + "epoch": 0.774275215808998, + "grad_norm": 3.664544105529785, + "learning_rate": 2.411737792122184e-05, + "loss": 0.2883, + "step": 33030 + }, + { + "epoch": 0.774298657383803, + "grad_norm": 2.5476388931274414, + "learning_rate": 2.411258059896263e-05, + "loss": 0.1719, + "step": 33031 + }, + { + "epoch": 0.774322098958608, + "grad_norm": 1.351728916168213, + "learning_rate": 2.4107783688470663e-05, + "loss": 0.6386, + "step": 33032 + }, + { + "epoch": 0.774345540533413, + "grad_norm": 3.0718536376953125, + "learning_rate": 2.410298718977202e-05, + "loss": 0.2443, + "step": 33033 + }, + { + "epoch": 0.774368982108218, + "grad_norm": 2.646523952484131, + "learning_rate": 2.409819110289271e-05, + "loss": 0.1331, + "step": 33034 + }, + { + "epoch": 0.7743924236830231, + "grad_norm": 0.3727813959121704, + "learning_rate": 2.4093395427858732e-05, + "loss": 0.0416, + "step": 33035 + }, + { + "epoch": 0.774415865257828, + "grad_norm": 1.77193021774292, + "learning_rate": 2.408860016469614e-05, + "loss": 0.1494, + "step": 33036 + }, + { + "epoch": 0.7744393068326331, + "grad_norm": 3.5372142791748047, + "learning_rate": 2.4083805313430907e-05, + "loss": 1.7033, + "step": 33037 + }, + { + "epoch": 0.774462748407438, + "grad_norm": 0.3574815094470978, + "learning_rate": 2.4079010874089113e-05, + "loss": 0.022, + "step": 33038 + }, + { + "epoch": 0.7744861899822431, + "grad_norm": 3.1999380588531494, + "learning_rate": 2.4074216846696707e-05, + "loss": 0.3932, + "step": 33039 + }, + { + "epoch": 0.774509631557048, + "grad_norm": 2.014498233795166, + "learning_rate": 2.4069423231279764e-05, + "loss": 0.1094, + "step": 33040 + }, + { + "epoch": 0.774533073131853, + "grad_norm": 2.395146608352661, + "learning_rate": 2.4064630027864266e-05, + "loss": 0.0979, + "step": 33041 + }, + { + "epoch": 0.774556514706658, + "grad_norm": 1.2681312561035156, + "learning_rate": 2.4059837236476224e-05, + "loss": 0.1428, + "step": 33042 + }, + { + "epoch": 0.774579956281463, + "grad_norm": 0.8541616201400757, + "learning_rate": 2.4055044857141628e-05, + "loss": 0.0488, + "step": 33043 + }, + { + "epoch": 0.774603397856268, + "grad_norm": 2.93902850151062, + "learning_rate": 2.4050252889886483e-05, + "loss": 1.1577, + "step": 33044 + }, + { + "epoch": 0.774626839431073, + "grad_norm": 1.8075121641159058, + "learning_rate": 2.4045461334736818e-05, + "loss": 0.1323, + "step": 33045 + }, + { + "epoch": 0.774650281005878, + "grad_norm": 11.008333206176758, + "learning_rate": 2.4040670191718616e-05, + "loss": 0.3648, + "step": 33046 + }, + { + "epoch": 0.774673722580683, + "grad_norm": 3.356576442718506, + "learning_rate": 2.403587946085788e-05, + "loss": 0.2243, + "step": 33047 + }, + { + "epoch": 0.774697164155488, + "grad_norm": 1.6678898334503174, + "learning_rate": 2.4031089142180587e-05, + "loss": 0.1479, + "step": 33048 + }, + { + "epoch": 0.774720605730293, + "grad_norm": 2.2444660663604736, + "learning_rate": 2.4026299235712723e-05, + "loss": 0.1972, + "step": 33049 + }, + { + "epoch": 0.7747440473050979, + "grad_norm": 3.6446328163146973, + "learning_rate": 2.4021509741480318e-05, + "loss": 0.3795, + "step": 33050 + }, + { + "epoch": 0.774767488879903, + "grad_norm": 3.5561752319335938, + "learning_rate": 2.4016720659509307e-05, + "loss": 1.4246, + "step": 33051 + }, + { + "epoch": 0.7747909304547079, + "grad_norm": 0.5443956851959229, + "learning_rate": 2.4011931989825742e-05, + "loss": 0.0513, + "step": 33052 + }, + { + "epoch": 0.774814372029513, + "grad_norm": 4.488882541656494, + "learning_rate": 2.4007143732455573e-05, + "loss": 0.3368, + "step": 33053 + }, + { + "epoch": 0.7748378136043179, + "grad_norm": 2.2392380237579346, + "learning_rate": 2.4002355887424766e-05, + "loss": 0.2003, + "step": 33054 + }, + { + "epoch": 0.774861255179123, + "grad_norm": 3.7598037719726562, + "learning_rate": 2.3997568454759322e-05, + "loss": 0.3356, + "step": 33055 + }, + { + "epoch": 0.7748846967539279, + "grad_norm": 1.8306019306182861, + "learning_rate": 2.39927814344852e-05, + "loss": 0.0796, + "step": 33056 + }, + { + "epoch": 0.774908138328733, + "grad_norm": 2.232048511505127, + "learning_rate": 2.398799482662837e-05, + "loss": 0.1419, + "step": 33057 + }, + { + "epoch": 0.7749315799035379, + "grad_norm": 3.5261807441711426, + "learning_rate": 2.3983208631214837e-05, + "loss": 1.6393, + "step": 33058 + }, + { + "epoch": 0.7749550214783429, + "grad_norm": 0.3718276023864746, + "learning_rate": 2.3978422848270545e-05, + "loss": 0.0145, + "step": 33059 + }, + { + "epoch": 0.7749784630531479, + "grad_norm": 1.7535490989685059, + "learning_rate": 2.3973637477821475e-05, + "loss": 0.7674, + "step": 33060 + }, + { + "epoch": 0.7750019046279529, + "grad_norm": 4.727066993713379, + "learning_rate": 2.396885251989356e-05, + "loss": 0.3354, + "step": 33061 + }, + { + "epoch": 0.7750253462027579, + "grad_norm": 2.587751626968384, + "learning_rate": 2.3964067974512815e-05, + "loss": 0.2509, + "step": 33062 + }, + { + "epoch": 0.7750487877775629, + "grad_norm": 3.0400843620300293, + "learning_rate": 2.395928384170517e-05, + "loss": 0.1742, + "step": 33063 + }, + { + "epoch": 0.7750722293523679, + "grad_norm": 1.5197880268096924, + "learning_rate": 2.395450012149656e-05, + "loss": 0.048, + "step": 33064 + }, + { + "epoch": 0.7750956709271729, + "grad_norm": 3.65024995803833, + "learning_rate": 2.3949716813913003e-05, + "loss": 0.2781, + "step": 33065 + }, + { + "epoch": 0.7751191125019778, + "grad_norm": 1.0096007585525513, + "learning_rate": 2.394493391898043e-05, + "loss": 0.1231, + "step": 33066 + }, + { + "epoch": 0.7751425540767829, + "grad_norm": 2.550489902496338, + "learning_rate": 2.3940151436724767e-05, + "loss": 0.1736, + "step": 33067 + }, + { + "epoch": 0.7751659956515878, + "grad_norm": 1.5640028715133667, + "learning_rate": 2.3935369367172e-05, + "loss": 0.1632, + "step": 33068 + }, + { + "epoch": 0.7751894372263929, + "grad_norm": 0.7404163479804993, + "learning_rate": 2.393058771034804e-05, + "loss": 0.0424, + "step": 33069 + }, + { + "epoch": 0.7752128788011978, + "grad_norm": 0.7231813073158264, + "learning_rate": 2.3925806466278834e-05, + "loss": 0.0644, + "step": 33070 + }, + { + "epoch": 0.7752363203760029, + "grad_norm": 0.159030482172966, + "learning_rate": 2.3921025634990357e-05, + "loss": 0.0083, + "step": 33071 + }, + { + "epoch": 0.7752597619508078, + "grad_norm": 3.1402597427368164, + "learning_rate": 2.391624521650855e-05, + "loss": 0.2414, + "step": 33072 + }, + { + "epoch": 0.7752832035256129, + "grad_norm": 1.1286404132843018, + "learning_rate": 2.3911465210859295e-05, + "loss": 0.0622, + "step": 33073 + }, + { + "epoch": 0.7753066451004178, + "grad_norm": 0.6805478930473328, + "learning_rate": 2.3906685618068603e-05, + "loss": 0.0474, + "step": 33074 + }, + { + "epoch": 0.7753300866752229, + "grad_norm": 2.5238490104675293, + "learning_rate": 2.390190643816237e-05, + "loss": 0.2004, + "step": 33075 + }, + { + "epoch": 0.7753535282500278, + "grad_norm": 1.4723294973373413, + "learning_rate": 2.389712767116653e-05, + "loss": 0.0955, + "step": 33076 + }, + { + "epoch": 0.7753769698248328, + "grad_norm": 4.952635288238525, + "learning_rate": 2.3892349317107e-05, + "loss": 0.4402, + "step": 33077 + }, + { + "epoch": 0.7754004113996378, + "grad_norm": 1.196662187576294, + "learning_rate": 2.3887571376009732e-05, + "loss": 0.0862, + "step": 33078 + }, + { + "epoch": 0.7754238529744428, + "grad_norm": 1.4657139778137207, + "learning_rate": 2.3882793847900653e-05, + "loss": 0.6021, + "step": 33079 + }, + { + "epoch": 0.7754472945492478, + "grad_norm": 3.2306225299835205, + "learning_rate": 2.3878016732805673e-05, + "loss": 0.2607, + "step": 33080 + }, + { + "epoch": 0.7754707361240528, + "grad_norm": 3.2094786167144775, + "learning_rate": 2.387324003075072e-05, + "loss": 0.316, + "step": 33081 + }, + { + "epoch": 0.7754941776988578, + "grad_norm": 0.892493486404419, + "learning_rate": 2.3868463741761672e-05, + "loss": 0.0869, + "step": 33082 + }, + { + "epoch": 0.7755176192736628, + "grad_norm": 3.823329448699951, + "learning_rate": 2.3863687865864513e-05, + "loss": 0.1619, + "step": 33083 + }, + { + "epoch": 0.7755410608484677, + "grad_norm": 1.3541381359100342, + "learning_rate": 2.385891240308512e-05, + "loss": 0.1526, + "step": 33084 + }, + { + "epoch": 0.7755645024232728, + "grad_norm": 1.4751150608062744, + "learning_rate": 2.385413735344939e-05, + "loss": 0.1062, + "step": 33085 + }, + { + "epoch": 0.7755879439980778, + "grad_norm": 3.1159842014312744, + "learning_rate": 2.384936271698327e-05, + "loss": 0.3987, + "step": 33086 + }, + { + "epoch": 0.7756113855728828, + "grad_norm": 3.6321892738342285, + "learning_rate": 2.384458849371265e-05, + "loss": 0.2047, + "step": 33087 + }, + { + "epoch": 0.7756348271476878, + "grad_norm": 1.987131118774414, + "learning_rate": 2.3839814683663443e-05, + "loss": 0.1255, + "step": 33088 + }, + { + "epoch": 0.7756582687224928, + "grad_norm": 2.4841363430023193, + "learning_rate": 2.3835041286861504e-05, + "loss": 1.5126, + "step": 33089 + }, + { + "epoch": 0.7756817102972978, + "grad_norm": 3.282991886138916, + "learning_rate": 2.3830268303332802e-05, + "loss": 0.254, + "step": 33090 + }, + { + "epoch": 0.7757051518721028, + "grad_norm": 2.547503709793091, + "learning_rate": 2.3825495733103208e-05, + "loss": 1.3974, + "step": 33091 + }, + { + "epoch": 0.7757285934469078, + "grad_norm": 0.9779024124145508, + "learning_rate": 2.382072357619862e-05, + "loss": 0.1042, + "step": 33092 + }, + { + "epoch": 0.7757520350217127, + "grad_norm": 0.9095590114593506, + "learning_rate": 2.3815951832644923e-05, + "loss": 0.1032, + "step": 33093 + }, + { + "epoch": 0.7757754765965178, + "grad_norm": 4.836070537567139, + "learning_rate": 2.3811180502468024e-05, + "loss": 0.252, + "step": 33094 + }, + { + "epoch": 0.7757989181713227, + "grad_norm": 0.840823769569397, + "learning_rate": 2.3806409585693756e-05, + "loss": 0.3111, + "step": 33095 + }, + { + "epoch": 0.7758223597461278, + "grad_norm": 0.782072901725769, + "learning_rate": 2.3801639082348092e-05, + "loss": 0.0615, + "step": 33096 + }, + { + "epoch": 0.7758458013209327, + "grad_norm": 2.986827850341797, + "learning_rate": 2.3796868992456878e-05, + "loss": 1.0588, + "step": 33097 + }, + { + "epoch": 0.7758692428957378, + "grad_norm": 0.6626651883125305, + "learning_rate": 2.3792099316045956e-05, + "loss": 0.0532, + "step": 33098 + }, + { + "epoch": 0.7758926844705427, + "grad_norm": 2.0769741535186768, + "learning_rate": 2.378733005314129e-05, + "loss": 0.0746, + "step": 33099 + }, + { + "epoch": 0.7759161260453478, + "grad_norm": 0.7817726135253906, + "learning_rate": 2.3782561203768705e-05, + "loss": 0.4736, + "step": 33100 + }, + { + "epoch": 0.7759395676201527, + "grad_norm": 4.157502174377441, + "learning_rate": 2.377779276795409e-05, + "loss": 0.2861, + "step": 33101 + }, + { + "epoch": 0.7759630091949578, + "grad_norm": 2.687617540359497, + "learning_rate": 2.377302474572327e-05, + "loss": 0.1871, + "step": 33102 + }, + { + "epoch": 0.7759864507697627, + "grad_norm": 2.509833812713623, + "learning_rate": 2.3768257137102213e-05, + "loss": 0.1266, + "step": 33103 + }, + { + "epoch": 0.7760098923445677, + "grad_norm": 0.9011265635490417, + "learning_rate": 2.3763489942116723e-05, + "loss": 0.0943, + "step": 33104 + }, + { + "epoch": 0.7760333339193727, + "grad_norm": 0.22052650153636932, + "learning_rate": 2.3758723160792684e-05, + "loss": 0.0126, + "step": 33105 + }, + { + "epoch": 0.7760567754941777, + "grad_norm": 0.5442859530448914, + "learning_rate": 2.3753956793155953e-05, + "loss": 0.0428, + "step": 33106 + }, + { + "epoch": 0.7760802170689827, + "grad_norm": 1.80485999584198, + "learning_rate": 2.3749190839232393e-05, + "loss": 0.0503, + "step": 33107 + }, + { + "epoch": 0.7761036586437877, + "grad_norm": 0.8017258644104004, + "learning_rate": 2.3744425299047846e-05, + "loss": 0.0787, + "step": 33108 + }, + { + "epoch": 0.7761271002185927, + "grad_norm": 2.528818368911743, + "learning_rate": 2.373966017262822e-05, + "loss": 0.1874, + "step": 33109 + }, + { + "epoch": 0.7761505417933977, + "grad_norm": 2.831505298614502, + "learning_rate": 2.3734895459999306e-05, + "loss": 0.3252, + "step": 33110 + }, + { + "epoch": 0.7761739833682026, + "grad_norm": 0.8549838662147522, + "learning_rate": 2.373013116118703e-05, + "loss": 0.053, + "step": 33111 + }, + { + "epoch": 0.7761974249430077, + "grad_norm": 2.707484722137451, + "learning_rate": 2.3725367276217215e-05, + "loss": 0.274, + "step": 33112 + }, + { + "epoch": 0.7762208665178126, + "grad_norm": 0.6812296509742737, + "learning_rate": 2.372060380511568e-05, + "loss": 0.0628, + "step": 33113 + }, + { + "epoch": 0.7762443080926177, + "grad_norm": 0.9769071936607361, + "learning_rate": 2.3715840747908314e-05, + "loss": 0.0696, + "step": 33114 + }, + { + "epoch": 0.7762677496674226, + "grad_norm": 0.8576635718345642, + "learning_rate": 2.37110781046209e-05, + "loss": 0.0934, + "step": 33115 + }, + { + "epoch": 0.7762911912422277, + "grad_norm": 2.681382179260254, + "learning_rate": 2.370631587527935e-05, + "loss": 1.6362, + "step": 33116 + }, + { + "epoch": 0.7763146328170326, + "grad_norm": 3.2331807613372803, + "learning_rate": 2.370155405990948e-05, + "loss": 0.283, + "step": 33117 + }, + { + "epoch": 0.7763380743918377, + "grad_norm": 2.196647882461548, + "learning_rate": 2.3696792658537127e-05, + "loss": 0.1338, + "step": 33118 + }, + { + "epoch": 0.7763615159666426, + "grad_norm": 2.0173840522766113, + "learning_rate": 2.369203167118812e-05, + "loss": 1.094, + "step": 33119 + }, + { + "epoch": 0.7763849575414477, + "grad_norm": 2.8854570388793945, + "learning_rate": 2.3687271097888264e-05, + "loss": 0.2371, + "step": 33120 + }, + { + "epoch": 0.7764083991162526, + "grad_norm": 2.241955041885376, + "learning_rate": 2.368251093866345e-05, + "loss": 0.1661, + "step": 33121 + }, + { + "epoch": 0.7764318406910576, + "grad_norm": 4.623466968536377, + "learning_rate": 2.3677751193539444e-05, + "loss": 0.2399, + "step": 33122 + }, + { + "epoch": 0.7764552822658626, + "grad_norm": 4.135062217712402, + "learning_rate": 2.3672991862542138e-05, + "loss": 0.324, + "step": 33123 + }, + { + "epoch": 0.7764787238406676, + "grad_norm": 3.6454482078552246, + "learning_rate": 2.3668232945697323e-05, + "loss": 0.4731, + "step": 33124 + }, + { + "epoch": 0.7765021654154726, + "grad_norm": 2.400489330291748, + "learning_rate": 2.3663474443030832e-05, + "loss": 0.338, + "step": 33125 + }, + { + "epoch": 0.7765256069902776, + "grad_norm": 2.256025791168213, + "learning_rate": 2.3658716354568468e-05, + "loss": 1.2265, + "step": 33126 + }, + { + "epoch": 0.7765490485650826, + "grad_norm": 2.4469034671783447, + "learning_rate": 2.365395868033602e-05, + "loss": 0.1023, + "step": 33127 + }, + { + "epoch": 0.7765724901398876, + "grad_norm": 2.4585869312286377, + "learning_rate": 2.3649201420359368e-05, + "loss": 0.0722, + "step": 33128 + }, + { + "epoch": 0.7765959317146925, + "grad_norm": 2.768646478652954, + "learning_rate": 2.3644444574664303e-05, + "loss": 1.3315, + "step": 33129 + }, + { + "epoch": 0.7766193732894976, + "grad_norm": 3.819321632385254, + "learning_rate": 2.363968814327663e-05, + "loss": 0.3698, + "step": 33130 + }, + { + "epoch": 0.7766428148643025, + "grad_norm": 3.02933931350708, + "learning_rate": 2.3634932126222153e-05, + "loss": 0.1636, + "step": 33131 + }, + { + "epoch": 0.7766662564391076, + "grad_norm": 1.8912428617477417, + "learning_rate": 2.363017652352668e-05, + "loss": 0.1074, + "step": 33132 + }, + { + "epoch": 0.7766896980139125, + "grad_norm": 3.91607666015625, + "learning_rate": 2.3625421335215993e-05, + "loss": 0.3812, + "step": 33133 + }, + { + "epoch": 0.7767131395887176, + "grad_norm": 0.8267709016799927, + "learning_rate": 2.362066656131592e-05, + "loss": 0.0377, + "step": 33134 + }, + { + "epoch": 0.7767365811635225, + "grad_norm": 0.9678270816802979, + "learning_rate": 2.3615912201852276e-05, + "loss": 0.0882, + "step": 33135 + }, + { + "epoch": 0.7767600227383276, + "grad_norm": 1.8642827272415161, + "learning_rate": 2.361115825685084e-05, + "loss": 0.2891, + "step": 33136 + }, + { + "epoch": 0.7767834643131326, + "grad_norm": 2.539303779602051, + "learning_rate": 2.3606404726337416e-05, + "loss": 1.2389, + "step": 33137 + }, + { + "epoch": 0.7768069058879375, + "grad_norm": 1.1538845300674438, + "learning_rate": 2.3601651610337783e-05, + "loss": 0.0922, + "step": 33138 + }, + { + "epoch": 0.7768303474627426, + "grad_norm": 3.5274672508239746, + "learning_rate": 2.359689890887774e-05, + "loss": 0.2696, + "step": 33139 + }, + { + "epoch": 0.7768537890375475, + "grad_norm": 3.975048780441284, + "learning_rate": 2.3592146621983035e-05, + "loss": 0.4837, + "step": 33140 + }, + { + "epoch": 0.7768772306123526, + "grad_norm": 1.2587482929229736, + "learning_rate": 2.3587394749679525e-05, + "loss": 0.185, + "step": 33141 + }, + { + "epoch": 0.7769006721871575, + "grad_norm": 3.489076852798462, + "learning_rate": 2.3582643291992956e-05, + "loss": 0.2769, + "step": 33142 + }, + { + "epoch": 0.7769241137619626, + "grad_norm": 1.3118329048156738, + "learning_rate": 2.3577892248949117e-05, + "loss": 0.0552, + "step": 33143 + }, + { + "epoch": 0.7769475553367675, + "grad_norm": 2.7037410736083984, + "learning_rate": 2.357314162057379e-05, + "loss": 0.2599, + "step": 33144 + }, + { + "epoch": 0.7769709969115726, + "grad_norm": 2.969698190689087, + "learning_rate": 2.3568391406892732e-05, + "loss": 0.1462, + "step": 33145 + }, + { + "epoch": 0.7769944384863775, + "grad_norm": 3.572467803955078, + "learning_rate": 2.356364160793171e-05, + "loss": 0.3158, + "step": 33146 + }, + { + "epoch": 0.7770178800611826, + "grad_norm": 0.656267523765564, + "learning_rate": 2.3558892223716523e-05, + "loss": 0.0291, + "step": 33147 + }, + { + "epoch": 0.7770413216359875, + "grad_norm": 4.230108261108398, + "learning_rate": 2.3554143254272955e-05, + "loss": 0.3315, + "step": 33148 + }, + { + "epoch": 0.7770647632107925, + "grad_norm": 0.6593889594078064, + "learning_rate": 2.354939469962676e-05, + "loss": 0.0356, + "step": 33149 + }, + { + "epoch": 0.7770882047855975, + "grad_norm": 4.065143585205078, + "learning_rate": 2.3544646559803696e-05, + "loss": 0.4193, + "step": 33150 + }, + { + "epoch": 0.7771116463604025, + "grad_norm": 2.4905025959014893, + "learning_rate": 2.3539898834829533e-05, + "loss": 0.1274, + "step": 33151 + }, + { + "epoch": 0.7771350879352075, + "grad_norm": 0.5312016606330872, + "learning_rate": 2.3535151524730027e-05, + "loss": 0.0165, + "step": 33152 + }, + { + "epoch": 0.7771585295100125, + "grad_norm": 2.3608157634735107, + "learning_rate": 2.353040462953091e-05, + "loss": 0.1818, + "step": 33153 + }, + { + "epoch": 0.7771819710848175, + "grad_norm": 5.26185941696167, + "learning_rate": 2.3525658149258002e-05, + "loss": 0.4772, + "step": 33154 + }, + { + "epoch": 0.7772054126596225, + "grad_norm": 3.029170513153076, + "learning_rate": 2.3520912083937018e-05, + "loss": 0.2795, + "step": 33155 + }, + { + "epoch": 0.7772288542344274, + "grad_norm": 2.3450424671173096, + "learning_rate": 2.351616643359371e-05, + "loss": 0.3494, + "step": 33156 + }, + { + "epoch": 0.7772522958092325, + "grad_norm": 1.150067687034607, + "learning_rate": 2.3511421198253835e-05, + "loss": 0.0637, + "step": 33157 + }, + { + "epoch": 0.7772757373840374, + "grad_norm": 3.7377192974090576, + "learning_rate": 2.3506676377943117e-05, + "loss": 0.1832, + "step": 33158 + }, + { + "epoch": 0.7772991789588425, + "grad_norm": 2.3862218856811523, + "learning_rate": 2.3501931972687352e-05, + "loss": 0.2027, + "step": 33159 + }, + { + "epoch": 0.7773226205336474, + "grad_norm": 2.8608760833740234, + "learning_rate": 2.3497187982512225e-05, + "loss": 0.265, + "step": 33160 + }, + { + "epoch": 0.7773460621084525, + "grad_norm": 2.2224395275115967, + "learning_rate": 2.3492444407443538e-05, + "loss": 0.1275, + "step": 33161 + }, + { + "epoch": 0.7773695036832574, + "grad_norm": 1.7006548643112183, + "learning_rate": 2.348770124750699e-05, + "loss": 0.168, + "step": 33162 + }, + { + "epoch": 0.7773929452580625, + "grad_norm": 3.5235626697540283, + "learning_rate": 2.3482958502728336e-05, + "loss": 0.2661, + "step": 33163 + }, + { + "epoch": 0.7774163868328674, + "grad_norm": 3.8410139083862305, + "learning_rate": 2.3478216173133293e-05, + "loss": 0.4329, + "step": 33164 + }, + { + "epoch": 0.7774398284076725, + "grad_norm": 2.1663522720336914, + "learning_rate": 2.3473474258747573e-05, + "loss": 0.2355, + "step": 33165 + }, + { + "epoch": 0.7774632699824774, + "grad_norm": 1.8038201332092285, + "learning_rate": 2.3468732759596967e-05, + "loss": 0.1713, + "step": 33166 + }, + { + "epoch": 0.7774867115572824, + "grad_norm": 1.4139845371246338, + "learning_rate": 2.3463991675707164e-05, + "loss": 0.105, + "step": 33167 + }, + { + "epoch": 0.7775101531320874, + "grad_norm": 2.0052998065948486, + "learning_rate": 2.34592510071039e-05, + "loss": 0.083, + "step": 33168 + }, + { + "epoch": 0.7775335947068924, + "grad_norm": 2.015105724334717, + "learning_rate": 2.3454510753812887e-05, + "loss": 0.131, + "step": 33169 + }, + { + "epoch": 0.7775570362816974, + "grad_norm": 3.790494918823242, + "learning_rate": 2.3449770915859824e-05, + "loss": 0.4923, + "step": 33170 + }, + { + "epoch": 0.7775804778565024, + "grad_norm": 3.1201694011688232, + "learning_rate": 2.3445031493270485e-05, + "loss": 0.0652, + "step": 33171 + }, + { + "epoch": 0.7776039194313074, + "grad_norm": 0.934461772441864, + "learning_rate": 2.3440292486070536e-05, + "loss": 0.1051, + "step": 33172 + }, + { + "epoch": 0.7776273610061124, + "grad_norm": 2.6695737838745117, + "learning_rate": 2.3435553894285732e-05, + "loss": 0.2994, + "step": 33173 + }, + { + "epoch": 0.7776508025809173, + "grad_norm": 4.400851726531982, + "learning_rate": 2.343081571794177e-05, + "loss": 0.4568, + "step": 33174 + }, + { + "epoch": 0.7776742441557224, + "grad_norm": 3.00006365776062, + "learning_rate": 2.3426077957064353e-05, + "loss": 0.2445, + "step": 33175 + }, + { + "epoch": 0.7776976857305273, + "grad_norm": 1.9925053119659424, + "learning_rate": 2.3421340611679186e-05, + "loss": 0.085, + "step": 33176 + }, + { + "epoch": 0.7777211273053324, + "grad_norm": 8.462993621826172, + "learning_rate": 2.341660368181199e-05, + "loss": 0.4753, + "step": 33177 + }, + { + "epoch": 0.7777445688801373, + "grad_norm": 2.2802860736846924, + "learning_rate": 2.341186716748841e-05, + "loss": 0.1355, + "step": 33178 + }, + { + "epoch": 0.7777680104549424, + "grad_norm": 0.846436619758606, + "learning_rate": 2.3407131068734233e-05, + "loss": 0.0256, + "step": 33179 + }, + { + "epoch": 0.7777914520297473, + "grad_norm": 2.4762535095214844, + "learning_rate": 2.3402395385575114e-05, + "loss": 0.1574, + "step": 33180 + }, + { + "epoch": 0.7778148936045524, + "grad_norm": 2.3746330738067627, + "learning_rate": 2.3397660118036747e-05, + "loss": 0.1879, + "step": 33181 + }, + { + "epoch": 0.7778383351793573, + "grad_norm": 3.4161553382873535, + "learning_rate": 2.339292526614484e-05, + "loss": 0.1684, + "step": 33182 + }, + { + "epoch": 0.7778617767541623, + "grad_norm": 2.822777032852173, + "learning_rate": 2.338819082992504e-05, + "loss": 0.2356, + "step": 33183 + }, + { + "epoch": 0.7778852183289673, + "grad_norm": 2.5277411937713623, + "learning_rate": 2.3383456809403092e-05, + "loss": 0.2507, + "step": 33184 + }, + { + "epoch": 0.7779086599037723, + "grad_norm": 1.1847691535949707, + "learning_rate": 2.3378723204604647e-05, + "loss": 0.6727, + "step": 33185 + }, + { + "epoch": 0.7779321014785773, + "grad_norm": 1.8214795589447021, + "learning_rate": 2.3373990015555424e-05, + "loss": 0.1133, + "step": 33186 + }, + { + "epoch": 0.7779555430533823, + "grad_norm": 1.8739553689956665, + "learning_rate": 2.3369257242281085e-05, + "loss": 0.0757, + "step": 33187 + }, + { + "epoch": 0.7779789846281874, + "grad_norm": 1.9092737436294556, + "learning_rate": 2.3364524884807315e-05, + "loss": 0.1123, + "step": 33188 + }, + { + "epoch": 0.7780024262029923, + "grad_norm": 1.8745540380477905, + "learning_rate": 2.3359792943159787e-05, + "loss": 0.1591, + "step": 33189 + }, + { + "epoch": 0.7780258677777974, + "grad_norm": 3.3713889122009277, + "learning_rate": 2.3355061417364178e-05, + "loss": 0.3072, + "step": 33190 + }, + { + "epoch": 0.7780493093526023, + "grad_norm": 0.9583431482315063, + "learning_rate": 2.3350330307446134e-05, + "loss": 0.0901, + "step": 33191 + }, + { + "epoch": 0.7780727509274074, + "grad_norm": 4.043739318847656, + "learning_rate": 2.334559961343138e-05, + "loss": 0.3676, + "step": 33192 + }, + { + "epoch": 0.7780961925022123, + "grad_norm": 0.9607272744178772, + "learning_rate": 2.334086933534555e-05, + "loss": 0.0176, + "step": 33193 + }, + { + "epoch": 0.7781196340770173, + "grad_norm": 1.7177362442016602, + "learning_rate": 2.3336139473214324e-05, + "loss": 0.1291, + "step": 33194 + }, + { + "epoch": 0.7781430756518223, + "grad_norm": 1.893201470375061, + "learning_rate": 2.333141002706334e-05, + "loss": 0.188, + "step": 33195 + }, + { + "epoch": 0.7781665172266273, + "grad_norm": 1.8520134687423706, + "learning_rate": 2.3326680996918294e-05, + "loss": 1.3202, + "step": 33196 + }, + { + "epoch": 0.7781899588014323, + "grad_norm": 1.820846438407898, + "learning_rate": 2.3321952382804845e-05, + "loss": 1.0104, + "step": 33197 + }, + { + "epoch": 0.7782134003762373, + "grad_norm": 3.998128890991211, + "learning_rate": 2.331722418474861e-05, + "loss": 0.375, + "step": 33198 + }, + { + "epoch": 0.7782368419510423, + "grad_norm": 5.0496625900268555, + "learning_rate": 2.3312496402775296e-05, + "loss": 0.2149, + "step": 33199 + }, + { + "epoch": 0.7782602835258473, + "grad_norm": 3.084313154220581, + "learning_rate": 2.330776903691053e-05, + "loss": 0.2738, + "step": 33200 + }, + { + "epoch": 0.7782837251006522, + "grad_norm": 1.017297625541687, + "learning_rate": 2.330304208717997e-05, + "loss": 0.0559, + "step": 33201 + }, + { + "epoch": 0.7783071666754573, + "grad_norm": 1.4488539695739746, + "learning_rate": 2.3298315553609263e-05, + "loss": 0.1434, + "step": 33202 + }, + { + "epoch": 0.7783306082502622, + "grad_norm": 1.858436942100525, + "learning_rate": 2.3293589436224018e-05, + "loss": 0.2461, + "step": 33203 + }, + { + "epoch": 0.7783540498250673, + "grad_norm": 1.1257394552230835, + "learning_rate": 2.328886373504995e-05, + "loss": 0.0637, + "step": 33204 + }, + { + "epoch": 0.7783774913998722, + "grad_norm": 2.9879398345947266, + "learning_rate": 2.3284138450112658e-05, + "loss": 0.2715, + "step": 33205 + }, + { + "epoch": 0.7784009329746773, + "grad_norm": 13.536602973937988, + "learning_rate": 2.3279413581437782e-05, + "loss": 0.259, + "step": 33206 + }, + { + "epoch": 0.7784243745494822, + "grad_norm": 1.2539151906967163, + "learning_rate": 2.3274689129050954e-05, + "loss": 0.1604, + "step": 33207 + }, + { + "epoch": 0.7784478161242873, + "grad_norm": 1.880170226097107, + "learning_rate": 2.326996509297783e-05, + "loss": 0.1528, + "step": 33208 + }, + { + "epoch": 0.7784712576990922, + "grad_norm": 1.2866957187652588, + "learning_rate": 2.3265241473244048e-05, + "loss": 0.0967, + "step": 33209 + }, + { + "epoch": 0.7784946992738973, + "grad_norm": 4.1204400062561035, + "learning_rate": 2.3260518269875188e-05, + "loss": 0.2443, + "step": 33210 + }, + { + "epoch": 0.7785181408487022, + "grad_norm": 4.13543701171875, + "learning_rate": 2.3255795482896937e-05, + "loss": 0.4313, + "step": 33211 + }, + { + "epoch": 0.7785415824235072, + "grad_norm": 7.011837959289551, + "learning_rate": 2.3251073112334896e-05, + "loss": 0.6863, + "step": 33212 + }, + { + "epoch": 0.7785650239983122, + "grad_norm": 3.1526429653167725, + "learning_rate": 2.3246351158214707e-05, + "loss": 1.7774, + "step": 33213 + }, + { + "epoch": 0.7785884655731172, + "grad_norm": 0.7689312696456909, + "learning_rate": 2.3241629620561966e-05, + "loss": 0.0461, + "step": 33214 + }, + { + "epoch": 0.7786119071479222, + "grad_norm": 3.007638931274414, + "learning_rate": 2.3236908499402297e-05, + "loss": 0.2847, + "step": 33215 + }, + { + "epoch": 0.7786353487227272, + "grad_norm": 0.29616406559944153, + "learning_rate": 2.3232187794761296e-05, + "loss": 0.038, + "step": 33216 + }, + { + "epoch": 0.7786587902975322, + "grad_norm": 1.904422640800476, + "learning_rate": 2.3227467506664623e-05, + "loss": 0.1084, + "step": 33217 + }, + { + "epoch": 0.7786822318723372, + "grad_norm": 1.005788803100586, + "learning_rate": 2.3222747635137876e-05, + "loss": 0.0515, + "step": 33218 + }, + { + "epoch": 0.7787056734471421, + "grad_norm": 4.432047367095947, + "learning_rate": 2.3218028180206643e-05, + "loss": 0.2504, + "step": 33219 + }, + { + "epoch": 0.7787291150219472, + "grad_norm": 2.540809392929077, + "learning_rate": 2.3213309141896557e-05, + "loss": 0.1836, + "step": 33220 + }, + { + "epoch": 0.7787525565967521, + "grad_norm": 2.859743356704712, + "learning_rate": 2.320859052023322e-05, + "loss": 0.2354, + "step": 33221 + }, + { + "epoch": 0.7787759981715572, + "grad_norm": 1.3493199348449707, + "learning_rate": 2.3203872315242237e-05, + "loss": 0.0801, + "step": 33222 + }, + { + "epoch": 0.7787994397463621, + "grad_norm": 3.295457363128662, + "learning_rate": 2.3199154526949175e-05, + "loss": 0.1966, + "step": 33223 + }, + { + "epoch": 0.7788228813211672, + "grad_norm": 3.0669407844543457, + "learning_rate": 2.3194437155379677e-05, + "loss": 0.1977, + "step": 33224 + }, + { + "epoch": 0.7788463228959721, + "grad_norm": 3.045637369155884, + "learning_rate": 2.3189720200559327e-05, + "loss": 0.2956, + "step": 33225 + }, + { + "epoch": 0.7788697644707772, + "grad_norm": 2.989176034927368, + "learning_rate": 2.3185003662513715e-05, + "loss": 0.2128, + "step": 33226 + }, + { + "epoch": 0.7788932060455821, + "grad_norm": 0.7716618180274963, + "learning_rate": 2.3180287541268443e-05, + "loss": 0.0693, + "step": 33227 + }, + { + "epoch": 0.7789166476203871, + "grad_norm": 2.2798242568969727, + "learning_rate": 2.3175571836849076e-05, + "loss": 0.3319, + "step": 33228 + }, + { + "epoch": 0.7789400891951921, + "grad_norm": 4.252710819244385, + "learning_rate": 2.3170856549281193e-05, + "loss": 0.4549, + "step": 33229 + }, + { + "epoch": 0.7789635307699971, + "grad_norm": 2.333176851272583, + "learning_rate": 2.3166141678590438e-05, + "loss": 0.2132, + "step": 33230 + }, + { + "epoch": 0.7789869723448021, + "grad_norm": 2.9110524654388428, + "learning_rate": 2.3161427224802323e-05, + "loss": 0.2015, + "step": 33231 + }, + { + "epoch": 0.7790104139196071, + "grad_norm": 2.566758394241333, + "learning_rate": 2.3156713187942493e-05, + "loss": 0.2885, + "step": 33232 + }, + { + "epoch": 0.7790338554944121, + "grad_norm": 2.8666372299194336, + "learning_rate": 2.31519995680365e-05, + "loss": 0.1698, + "step": 33233 + }, + { + "epoch": 0.7790572970692171, + "grad_norm": 1.0919770002365112, + "learning_rate": 2.314728636510991e-05, + "loss": 0.0958, + "step": 33234 + }, + { + "epoch": 0.779080738644022, + "grad_norm": 0.8412837982177734, + "learning_rate": 2.3142573579188285e-05, + "loss": 0.0441, + "step": 33235 + }, + { + "epoch": 0.7791041802188271, + "grad_norm": 3.389296770095825, + "learning_rate": 2.313786121029724e-05, + "loss": 0.3901, + "step": 33236 + }, + { + "epoch": 0.779127621793632, + "grad_norm": 2.554499387741089, + "learning_rate": 2.3133149258462327e-05, + "loss": 0.1207, + "step": 33237 + }, + { + "epoch": 0.7791510633684371, + "grad_norm": 1.1036514043807983, + "learning_rate": 2.3128437723709106e-05, + "loss": 0.0921, + "step": 33238 + }, + { + "epoch": 0.7791745049432421, + "grad_norm": 3.518144130706787, + "learning_rate": 2.3123726606063144e-05, + "loss": 0.4103, + "step": 33239 + }, + { + "epoch": 0.7791979465180471, + "grad_norm": 2.219190835952759, + "learning_rate": 2.3119015905550003e-05, + "loss": 0.1629, + "step": 33240 + }, + { + "epoch": 0.7792213880928521, + "grad_norm": 2.387615442276001, + "learning_rate": 2.311430562219522e-05, + "loss": 0.2078, + "step": 33241 + }, + { + "epoch": 0.7792448296676571, + "grad_norm": 2.2571237087249756, + "learning_rate": 2.3109595756024393e-05, + "loss": 0.1868, + "step": 33242 + }, + { + "epoch": 0.7792682712424621, + "grad_norm": 2.945594549179077, + "learning_rate": 2.3104886307063045e-05, + "loss": 0.1204, + "step": 33243 + }, + { + "epoch": 0.7792917128172671, + "grad_norm": 3.8344287872314453, + "learning_rate": 2.3100177275336764e-05, + "loss": 0.3501, + "step": 33244 + }, + { + "epoch": 0.7793151543920721, + "grad_norm": 2.4831979274749756, + "learning_rate": 2.309546866087109e-05, + "loss": 0.1782, + "step": 33245 + }, + { + "epoch": 0.779338595966877, + "grad_norm": 4.357810020446777, + "learning_rate": 2.3090760463691565e-05, + "loss": 0.3507, + "step": 33246 + }, + { + "epoch": 0.7793620375416821, + "grad_norm": 2.4507791996002197, + "learning_rate": 2.3086052683823733e-05, + "loss": 0.1679, + "step": 33247 + }, + { + "epoch": 0.779385479116487, + "grad_norm": 2.5324366092681885, + "learning_rate": 2.3081345321293114e-05, + "loss": 0.1782, + "step": 33248 + }, + { + "epoch": 0.7794089206912921, + "grad_norm": 3.3163928985595703, + "learning_rate": 2.3076638376125305e-05, + "loss": 0.3242, + "step": 33249 + }, + { + "epoch": 0.779432362266097, + "grad_norm": 0.4788587987422943, + "learning_rate": 2.3071931848345818e-05, + "loss": 0.0461, + "step": 33250 + }, + { + "epoch": 0.7794558038409021, + "grad_norm": 0.9180229306221008, + "learning_rate": 2.3067225737980193e-05, + "loss": 0.0949, + "step": 33251 + }, + { + "epoch": 0.779479245415707, + "grad_norm": 3.463927984237671, + "learning_rate": 2.306252004505397e-05, + "loss": 0.3367, + "step": 33252 + }, + { + "epoch": 0.7795026869905121, + "grad_norm": 0.8759869933128357, + "learning_rate": 2.305781476959267e-05, + "loss": 0.055, + "step": 33253 + }, + { + "epoch": 0.779526128565317, + "grad_norm": 0.5790903568267822, + "learning_rate": 2.3053109911621807e-05, + "loss": 0.0255, + "step": 33254 + }, + { + "epoch": 0.779549570140122, + "grad_norm": 1.0388880968093872, + "learning_rate": 2.3048405471166957e-05, + "loss": 0.0826, + "step": 33255 + }, + { + "epoch": 0.779573011714927, + "grad_norm": 3.708754539489746, + "learning_rate": 2.3043701448253597e-05, + "loss": 0.2324, + "step": 33256 + }, + { + "epoch": 0.779596453289732, + "grad_norm": 4.86841344833374, + "learning_rate": 2.3038997842907307e-05, + "loss": 0.3912, + "step": 33257 + }, + { + "epoch": 0.779619894864537, + "grad_norm": 2.6403095722198486, + "learning_rate": 2.3034294655153575e-05, + "loss": 0.1645, + "step": 33258 + }, + { + "epoch": 0.779643336439342, + "grad_norm": 2.6886608600616455, + "learning_rate": 2.3029591885017932e-05, + "loss": 0.3173, + "step": 33259 + }, + { + "epoch": 0.779666778014147, + "grad_norm": 2.0695230960845947, + "learning_rate": 2.3024889532525874e-05, + "loss": 0.1315, + "step": 33260 + }, + { + "epoch": 0.779690219588952, + "grad_norm": 3.3605101108551025, + "learning_rate": 2.3020187597702912e-05, + "loss": 0.4822, + "step": 33261 + }, + { + "epoch": 0.779713661163757, + "grad_norm": 3.2094874382019043, + "learning_rate": 2.30154860805746e-05, + "loss": 0.1955, + "step": 33262 + }, + { + "epoch": 0.779737102738562, + "grad_norm": 2.5820813179016113, + "learning_rate": 2.301078498116642e-05, + "loss": 0.2231, + "step": 33263 + }, + { + "epoch": 0.7797605443133669, + "grad_norm": 2.2002289295196533, + "learning_rate": 2.3006084299503904e-05, + "loss": 0.1999, + "step": 33264 + }, + { + "epoch": 0.779783985888172, + "grad_norm": 2.3152897357940674, + "learning_rate": 2.3001384035612527e-05, + "loss": 0.1343, + "step": 33265 + }, + { + "epoch": 0.7798074274629769, + "grad_norm": 0.35959360003471375, + "learning_rate": 2.2996684189517804e-05, + "loss": 0.0207, + "step": 33266 + }, + { + "epoch": 0.779830869037782, + "grad_norm": 1.1568154096603394, + "learning_rate": 2.2991984761245212e-05, + "loss": 0.08, + "step": 33267 + }, + { + "epoch": 0.7798543106125869, + "grad_norm": 2.9647269248962402, + "learning_rate": 2.2987285750820275e-05, + "loss": 0.1518, + "step": 33268 + }, + { + "epoch": 0.779877752187392, + "grad_norm": 5.093358039855957, + "learning_rate": 2.298258715826852e-05, + "loss": 0.402, + "step": 33269 + }, + { + "epoch": 0.7799011937621969, + "grad_norm": 2.060664176940918, + "learning_rate": 2.2977888983615414e-05, + "loss": 0.1825, + "step": 33270 + }, + { + "epoch": 0.779924635337002, + "grad_norm": 2.3827857971191406, + "learning_rate": 2.2973191226886436e-05, + "loss": 0.1427, + "step": 33271 + }, + { + "epoch": 0.7799480769118069, + "grad_norm": 2.122798442840576, + "learning_rate": 2.2968493888107102e-05, + "loss": 0.1812, + "step": 33272 + }, + { + "epoch": 0.779971518486612, + "grad_norm": 2.1614115238189697, + "learning_rate": 2.296379696730285e-05, + "loss": 0.0722, + "step": 33273 + }, + { + "epoch": 0.7799949600614169, + "grad_norm": 2.703277587890625, + "learning_rate": 2.295910046449924e-05, + "loss": 0.2042, + "step": 33274 + }, + { + "epoch": 0.7800184016362219, + "grad_norm": 0.8744557499885559, + "learning_rate": 2.2954404379721715e-05, + "loss": 0.06, + "step": 33275 + }, + { + "epoch": 0.7800418432110269, + "grad_norm": 3.019277572631836, + "learning_rate": 2.2949708712995753e-05, + "loss": 0.232, + "step": 33276 + }, + { + "epoch": 0.7800652847858319, + "grad_norm": 4.385270118713379, + "learning_rate": 2.2945013464346843e-05, + "loss": 0.2683, + "step": 33277 + }, + { + "epoch": 0.7800887263606369, + "grad_norm": 3.6652920246124268, + "learning_rate": 2.294031863380045e-05, + "loss": 0.2022, + "step": 33278 + }, + { + "epoch": 0.7801121679354419, + "grad_norm": 3.093480348587036, + "learning_rate": 2.2935624221382025e-05, + "loss": 0.2424, + "step": 33279 + }, + { + "epoch": 0.7801356095102469, + "grad_norm": 0.6587817668914795, + "learning_rate": 2.2930930227117075e-05, + "loss": 0.03, + "step": 33280 + }, + { + "epoch": 0.7801590510850519, + "grad_norm": 3.109569549560547, + "learning_rate": 2.2926236651031098e-05, + "loss": 0.3476, + "step": 33281 + }, + { + "epoch": 0.7801824926598568, + "grad_norm": 0.8379619121551514, + "learning_rate": 2.2921543493149522e-05, + "loss": 0.0478, + "step": 33282 + }, + { + "epoch": 0.7802059342346619, + "grad_norm": 3.5773565769195557, + "learning_rate": 2.291685075349782e-05, + "loss": 0.3813, + "step": 33283 + }, + { + "epoch": 0.7802293758094668, + "grad_norm": 1.5741057395935059, + "learning_rate": 2.291215843210145e-05, + "loss": 0.1961, + "step": 33284 + }, + { + "epoch": 0.7802528173842719, + "grad_norm": 2.5604965686798096, + "learning_rate": 2.2907466528985876e-05, + "loss": 0.2127, + "step": 33285 + }, + { + "epoch": 0.7802762589590768, + "grad_norm": 3.9285929203033447, + "learning_rate": 2.2902775044176528e-05, + "loss": 0.3548, + "step": 33286 + }, + { + "epoch": 0.7802997005338819, + "grad_norm": 2.098179817199707, + "learning_rate": 2.2898083977698927e-05, + "loss": 0.1615, + "step": 33287 + }, + { + "epoch": 0.7803231421086868, + "grad_norm": 2.277574300765991, + "learning_rate": 2.289339332957847e-05, + "loss": 0.1326, + "step": 33288 + }, + { + "epoch": 0.7803465836834919, + "grad_norm": 2.49812912940979, + "learning_rate": 2.2888703099840646e-05, + "loss": 1.5915, + "step": 33289 + }, + { + "epoch": 0.7803700252582969, + "grad_norm": 3.13633131980896, + "learning_rate": 2.2884013288510886e-05, + "loss": 0.2301, + "step": 33290 + }, + { + "epoch": 0.7803934668331018, + "grad_norm": 2.801114320755005, + "learning_rate": 2.2879323895614636e-05, + "loss": 0.231, + "step": 33291 + }, + { + "epoch": 0.7804169084079069, + "grad_norm": 2.6836440563201904, + "learning_rate": 2.2874634921177318e-05, + "loss": 0.9612, + "step": 33292 + }, + { + "epoch": 0.7804403499827118, + "grad_norm": 1.1876636743545532, + "learning_rate": 2.2869946365224393e-05, + "loss": 0.0315, + "step": 33293 + }, + { + "epoch": 0.7804637915575169, + "grad_norm": 2.368804454803467, + "learning_rate": 2.286525822778134e-05, + "loss": 0.1764, + "step": 33294 + }, + { + "epoch": 0.7804872331323218, + "grad_norm": 3.9690089225769043, + "learning_rate": 2.2860570508873556e-05, + "loss": 0.5233, + "step": 33295 + }, + { + "epoch": 0.7805106747071269, + "grad_norm": 5.157387733459473, + "learning_rate": 2.2855883208526497e-05, + "loss": 0.4239, + "step": 33296 + }, + { + "epoch": 0.7805341162819318, + "grad_norm": 2.1805450916290283, + "learning_rate": 2.2851196326765577e-05, + "loss": 0.0982, + "step": 33297 + }, + { + "epoch": 0.7805575578567369, + "grad_norm": 2.0747809410095215, + "learning_rate": 2.284650986361625e-05, + "loss": 0.1619, + "step": 33298 + }, + { + "epoch": 0.7805809994315418, + "grad_norm": 1.6879220008850098, + "learning_rate": 2.284182381910389e-05, + "loss": 0.0512, + "step": 33299 + }, + { + "epoch": 0.7806044410063468, + "grad_norm": 0.6890121102333069, + "learning_rate": 2.283713819325399e-05, + "loss": 0.0669, + "step": 33300 + }, + { + "epoch": 0.7806278825811518, + "grad_norm": 1.7084766626358032, + "learning_rate": 2.2832452986091946e-05, + "loss": 0.1386, + "step": 33301 + }, + { + "epoch": 0.7806513241559568, + "grad_norm": 2.8946661949157715, + "learning_rate": 2.2827768197643186e-05, + "loss": 1.4858, + "step": 33302 + }, + { + "epoch": 0.7806747657307618, + "grad_norm": 5.076989650726318, + "learning_rate": 2.2823083827933122e-05, + "loss": 0.3095, + "step": 33303 + }, + { + "epoch": 0.7806982073055668, + "grad_norm": 1.640716791152954, + "learning_rate": 2.281839987698715e-05, + "loss": 0.1563, + "step": 33304 + }, + { + "epoch": 0.7807216488803718, + "grad_norm": 2.5768814086914062, + "learning_rate": 2.281371634483075e-05, + "loss": 0.1238, + "step": 33305 + }, + { + "epoch": 0.7807450904551768, + "grad_norm": 2.096911907196045, + "learning_rate": 2.280903323148925e-05, + "loss": 0.1047, + "step": 33306 + }, + { + "epoch": 0.7807685320299818, + "grad_norm": 4.128399848937988, + "learning_rate": 2.280435053698814e-05, + "loss": 0.4122, + "step": 33307 + }, + { + "epoch": 0.7807919736047868, + "grad_norm": 0.7611221671104431, + "learning_rate": 2.2799668261352792e-05, + "loss": 0.0725, + "step": 33308 + }, + { + "epoch": 0.7808154151795917, + "grad_norm": 2.043224811553955, + "learning_rate": 2.279498640460862e-05, + "loss": 1.0854, + "step": 33309 + }, + { + "epoch": 0.7808388567543968, + "grad_norm": 1.9347363710403442, + "learning_rate": 2.279030496678102e-05, + "loss": 0.1131, + "step": 33310 + }, + { + "epoch": 0.7808622983292017, + "grad_norm": 4.138265609741211, + "learning_rate": 2.2785623947895375e-05, + "loss": 0.2985, + "step": 33311 + }, + { + "epoch": 0.7808857399040068, + "grad_norm": 2.9760987758636475, + "learning_rate": 2.2780943347977125e-05, + "loss": 0.2286, + "step": 33312 + }, + { + "epoch": 0.7809091814788117, + "grad_norm": 0.7097312211990356, + "learning_rate": 2.2776263167051648e-05, + "loss": 0.0445, + "step": 33313 + }, + { + "epoch": 0.7809326230536168, + "grad_norm": 2.9920458793640137, + "learning_rate": 2.2771583405144337e-05, + "loss": 0.1754, + "step": 33314 + }, + { + "epoch": 0.7809560646284217, + "grad_norm": 2.2587668895721436, + "learning_rate": 2.2766904062280593e-05, + "loss": 0.2232, + "step": 33315 + }, + { + "epoch": 0.7809795062032268, + "grad_norm": 3.716559648513794, + "learning_rate": 2.276222513848576e-05, + "loss": 0.3014, + "step": 33316 + }, + { + "epoch": 0.7810029477780317, + "grad_norm": 4.108968257904053, + "learning_rate": 2.275754663378531e-05, + "loss": 0.3279, + "step": 33317 + }, + { + "epoch": 0.7810263893528367, + "grad_norm": 1.058425784111023, + "learning_rate": 2.275286854820454e-05, + "loss": 0.0811, + "step": 33318 + }, + { + "epoch": 0.7810498309276417, + "grad_norm": 1.8080792427062988, + "learning_rate": 2.2748190881768917e-05, + "loss": 0.0712, + "step": 33319 + }, + { + "epoch": 0.7810732725024467, + "grad_norm": 0.9148707985877991, + "learning_rate": 2.2743513634503777e-05, + "loss": 0.0347, + "step": 33320 + }, + { + "epoch": 0.7810967140772517, + "grad_norm": 1.490112543106079, + "learning_rate": 2.2738836806434494e-05, + "loss": 0.1117, + "step": 33321 + }, + { + "epoch": 0.7811201556520567, + "grad_norm": 3.208110809326172, + "learning_rate": 2.2734160397586468e-05, + "loss": 1.6527, + "step": 33322 + }, + { + "epoch": 0.7811435972268617, + "grad_norm": 2.8891050815582275, + "learning_rate": 2.2729484407985057e-05, + "loss": 0.2145, + "step": 33323 + }, + { + "epoch": 0.7811670388016667, + "grad_norm": 2.762528419494629, + "learning_rate": 2.272480883765561e-05, + "loss": 1.4708, + "step": 33324 + }, + { + "epoch": 0.7811904803764717, + "grad_norm": 1.7525124549865723, + "learning_rate": 2.2720133686623545e-05, + "loss": 0.084, + "step": 33325 + }, + { + "epoch": 0.7812139219512767, + "grad_norm": 0.7148317694664001, + "learning_rate": 2.2715458954914216e-05, + "loss": 0.045, + "step": 33326 + }, + { + "epoch": 0.7812373635260816, + "grad_norm": 1.3334227800369263, + "learning_rate": 2.2710784642552972e-05, + "loss": 0.7933, + "step": 33327 + }, + { + "epoch": 0.7812608051008867, + "grad_norm": 2.49151611328125, + "learning_rate": 2.270611074956519e-05, + "loss": 0.2573, + "step": 33328 + }, + { + "epoch": 0.7812842466756916, + "grad_norm": 0.9478671550750732, + "learning_rate": 2.270143727597619e-05, + "loss": 0.1008, + "step": 33329 + }, + { + "epoch": 0.7813076882504967, + "grad_norm": 4.181211471557617, + "learning_rate": 2.2696764221811407e-05, + "loss": 0.4364, + "step": 33330 + }, + { + "epoch": 0.7813311298253016, + "grad_norm": 3.207756757736206, + "learning_rate": 2.2692091587096108e-05, + "loss": 0.2128, + "step": 33331 + }, + { + "epoch": 0.7813545714001067, + "grad_norm": 3.2938544750213623, + "learning_rate": 2.2687419371855735e-05, + "loss": 0.2747, + "step": 33332 + }, + { + "epoch": 0.7813780129749116, + "grad_norm": 2.4113593101501465, + "learning_rate": 2.2682747576115592e-05, + "loss": 0.1673, + "step": 33333 + }, + { + "epoch": 0.7814014545497167, + "grad_norm": 3.707991123199463, + "learning_rate": 2.2678076199901032e-05, + "loss": 0.3002, + "step": 33334 + }, + { + "epoch": 0.7814248961245216, + "grad_norm": 0.622240424156189, + "learning_rate": 2.2673405243237412e-05, + "loss": 0.0356, + "step": 33335 + }, + { + "epoch": 0.7814483376993266, + "grad_norm": 1.423643708229065, + "learning_rate": 2.2668734706150062e-05, + "loss": 0.2061, + "step": 33336 + }, + { + "epoch": 0.7814717792741316, + "grad_norm": 1.0248312950134277, + "learning_rate": 2.2664064588664314e-05, + "loss": 0.0563, + "step": 33337 + }, + { + "epoch": 0.7814952208489366, + "grad_norm": 1.6806684732437134, + "learning_rate": 2.2659394890805552e-05, + "loss": 0.1557, + "step": 33338 + }, + { + "epoch": 0.7815186624237416, + "grad_norm": 1.3129228353500366, + "learning_rate": 2.2654725612599083e-05, + "loss": 0.1389, + "step": 33339 + }, + { + "epoch": 0.7815421039985466, + "grad_norm": 2.7353932857513428, + "learning_rate": 2.265005675407025e-05, + "loss": 0.1967, + "step": 33340 + }, + { + "epoch": 0.7815655455733517, + "grad_norm": 4.313100814819336, + "learning_rate": 2.2645388315244354e-05, + "loss": 0.3463, + "step": 33341 + }, + { + "epoch": 0.7815889871481566, + "grad_norm": 3.519435167312622, + "learning_rate": 2.2640720296146777e-05, + "loss": 0.2044, + "step": 33342 + }, + { + "epoch": 0.7816124287229617, + "grad_norm": 2.270789623260498, + "learning_rate": 2.2636052696802835e-05, + "loss": 0.1989, + "step": 33343 + }, + { + "epoch": 0.7816358702977666, + "grad_norm": 1.872855544090271, + "learning_rate": 2.2631385517237824e-05, + "loss": 1.0478, + "step": 33344 + }, + { + "epoch": 0.7816593118725716, + "grad_norm": 2.4506494998931885, + "learning_rate": 2.2626718757477106e-05, + "loss": 1.2416, + "step": 33345 + }, + { + "epoch": 0.7816827534473766, + "grad_norm": 5.087891578674316, + "learning_rate": 2.2622052417545993e-05, + "loss": 0.5814, + "step": 33346 + }, + { + "epoch": 0.7817061950221816, + "grad_norm": 2.9049532413482666, + "learning_rate": 2.2617386497469805e-05, + "loss": 0.087, + "step": 33347 + }, + { + "epoch": 0.7817296365969866, + "grad_norm": 2.186704158782959, + "learning_rate": 2.261272099727385e-05, + "loss": 1.2009, + "step": 33348 + }, + { + "epoch": 0.7817530781717916, + "grad_norm": 1.9230031967163086, + "learning_rate": 2.2608055916983416e-05, + "loss": 0.1426, + "step": 33349 + }, + { + "epoch": 0.7817765197465966, + "grad_norm": 0.2779344916343689, + "learning_rate": 2.2603391256623874e-05, + "loss": 0.0195, + "step": 33350 + }, + { + "epoch": 0.7817999613214016, + "grad_norm": 3.465806722640991, + "learning_rate": 2.259872701622051e-05, + "loss": 0.1612, + "step": 33351 + }, + { + "epoch": 0.7818234028962066, + "grad_norm": 0.3206181228160858, + "learning_rate": 2.259406319579863e-05, + "loss": 0.0345, + "step": 33352 + }, + { + "epoch": 0.7818468444710116, + "grad_norm": 1.0423507690429688, + "learning_rate": 2.2589399795383502e-05, + "loss": 0.0767, + "step": 33353 + }, + { + "epoch": 0.7818702860458165, + "grad_norm": 3.0666074752807617, + "learning_rate": 2.258473681500051e-05, + "loss": 1.6736, + "step": 33354 + }, + { + "epoch": 0.7818937276206216, + "grad_norm": 1.699487566947937, + "learning_rate": 2.258007425467491e-05, + "loss": 0.1577, + "step": 33355 + }, + { + "epoch": 0.7819171691954265, + "grad_norm": 3.6839046478271484, + "learning_rate": 2.257541211443197e-05, + "loss": 0.2935, + "step": 33356 + }, + { + "epoch": 0.7819406107702316, + "grad_norm": 2.128833293914795, + "learning_rate": 2.2570750394297046e-05, + "loss": 0.0846, + "step": 33357 + }, + { + "epoch": 0.7819640523450365, + "grad_norm": 3.6007704734802246, + "learning_rate": 2.2566089094295418e-05, + "loss": 0.2172, + "step": 33358 + }, + { + "epoch": 0.7819874939198416, + "grad_norm": 3.5931479930877686, + "learning_rate": 2.256142821445236e-05, + "loss": 0.2054, + "step": 33359 + }, + { + "epoch": 0.7820109354946465, + "grad_norm": 0.8825416564941406, + "learning_rate": 2.2556767754793174e-05, + "loss": 0.0637, + "step": 33360 + }, + { + "epoch": 0.7820343770694516, + "grad_norm": 2.053929567337036, + "learning_rate": 2.255210771534314e-05, + "loss": 0.1068, + "step": 33361 + }, + { + "epoch": 0.7820578186442565, + "grad_norm": 1.5867594480514526, + "learning_rate": 2.254744809612752e-05, + "loss": 0.1181, + "step": 33362 + }, + { + "epoch": 0.7820812602190615, + "grad_norm": 0.5599542260169983, + "learning_rate": 2.2542788897171664e-05, + "loss": 0.0358, + "step": 33363 + }, + { + "epoch": 0.7821047017938665, + "grad_norm": 4.785211086273193, + "learning_rate": 2.25381301185008e-05, + "loss": 0.2327, + "step": 33364 + }, + { + "epoch": 0.7821281433686715, + "grad_norm": 1.880979061126709, + "learning_rate": 2.253347176014019e-05, + "loss": 0.1346, + "step": 33365 + }, + { + "epoch": 0.7821515849434765, + "grad_norm": 3.8938584327697754, + "learning_rate": 2.252881382211518e-05, + "loss": 0.2288, + "step": 33366 + }, + { + "epoch": 0.7821750265182815, + "grad_norm": 0.48129141330718994, + "learning_rate": 2.2524156304450994e-05, + "loss": 0.0218, + "step": 33367 + }, + { + "epoch": 0.7821984680930865, + "grad_norm": 0.7252079844474792, + "learning_rate": 2.2519499207172923e-05, + "loss": 0.0605, + "step": 33368 + }, + { + "epoch": 0.7822219096678915, + "grad_norm": 0.507931113243103, + "learning_rate": 2.251484253030619e-05, + "loss": 0.0455, + "step": 33369 + }, + { + "epoch": 0.7822453512426965, + "grad_norm": 2.3006703853607178, + "learning_rate": 2.251018627387613e-05, + "loss": 0.1913, + "step": 33370 + }, + { + "epoch": 0.7822687928175015, + "grad_norm": 6.342188835144043, + "learning_rate": 2.2505530437907986e-05, + "loss": 0.5341, + "step": 33371 + }, + { + "epoch": 0.7822922343923064, + "grad_norm": 1.1793540716171265, + "learning_rate": 2.2500875022427015e-05, + "loss": 0.1331, + "step": 33372 + }, + { + "epoch": 0.7823156759671115, + "grad_norm": 3.6368300914764404, + "learning_rate": 2.2496220027458458e-05, + "loss": 0.343, + "step": 33373 + }, + { + "epoch": 0.7823391175419164, + "grad_norm": 3.528120756149292, + "learning_rate": 2.2491565453027607e-05, + "loss": 0.2445, + "step": 33374 + }, + { + "epoch": 0.7823625591167215, + "grad_norm": 3.455840587615967, + "learning_rate": 2.2486911299159663e-05, + "loss": 0.2583, + "step": 33375 + }, + { + "epoch": 0.7823860006915264, + "grad_norm": 1.5342034101486206, + "learning_rate": 2.248225756587995e-05, + "loss": 0.1269, + "step": 33376 + }, + { + "epoch": 0.7824094422663315, + "grad_norm": 3.6449670791625977, + "learning_rate": 2.2477604253213656e-05, + "loss": 0.2366, + "step": 33377 + }, + { + "epoch": 0.7824328838411364, + "grad_norm": 2.2971489429473877, + "learning_rate": 2.24729513611861e-05, + "loss": 0.1609, + "step": 33378 + }, + { + "epoch": 0.7824563254159415, + "grad_norm": 1.551161289215088, + "learning_rate": 2.246829888982248e-05, + "loss": 0.148, + "step": 33379 + }, + { + "epoch": 0.7824797669907464, + "grad_norm": 4.037929058074951, + "learning_rate": 2.2463646839148055e-05, + "loss": 0.4262, + "step": 33380 + }, + { + "epoch": 0.7825032085655514, + "grad_norm": 3.870246648788452, + "learning_rate": 2.245899520918806e-05, + "loss": 0.5045, + "step": 33381 + }, + { + "epoch": 0.7825266501403564, + "grad_norm": 2.093141794204712, + "learning_rate": 2.2454343999967708e-05, + "loss": 0.1713, + "step": 33382 + }, + { + "epoch": 0.7825500917151614, + "grad_norm": 2.832038164138794, + "learning_rate": 2.244969321151229e-05, + "loss": 0.1623, + "step": 33383 + }, + { + "epoch": 0.7825735332899664, + "grad_norm": 1.5974376201629639, + "learning_rate": 2.244504284384702e-05, + "loss": 0.0961, + "step": 33384 + }, + { + "epoch": 0.7825969748647714, + "grad_norm": 3.418822765350342, + "learning_rate": 2.244039289699713e-05, + "loss": 0.2894, + "step": 33385 + }, + { + "epoch": 0.7826204164395764, + "grad_norm": 2.829998016357422, + "learning_rate": 2.243574337098785e-05, + "loss": 0.2643, + "step": 33386 + }, + { + "epoch": 0.7826438580143814, + "grad_norm": 2.194274425506592, + "learning_rate": 2.2431094265844377e-05, + "loss": 1.6496, + "step": 33387 + }, + { + "epoch": 0.7826672995891863, + "grad_norm": 2.7506771087646484, + "learning_rate": 2.2426445581592004e-05, + "loss": 0.2132, + "step": 33388 + }, + { + "epoch": 0.7826907411639914, + "grad_norm": 3.811018466949463, + "learning_rate": 2.2421797318255877e-05, + "loss": 0.3723, + "step": 33389 + }, + { + "epoch": 0.7827141827387963, + "grad_norm": 0.8170151114463806, + "learning_rate": 2.24171494758613e-05, + "loss": 0.0911, + "step": 33390 + }, + { + "epoch": 0.7827376243136014, + "grad_norm": 1.334930419921875, + "learning_rate": 2.241250205443344e-05, + "loss": 0.0843, + "step": 33391 + }, + { + "epoch": 0.7827610658884064, + "grad_norm": 4.1747355461120605, + "learning_rate": 2.240785505399753e-05, + "loss": 0.3987, + "step": 33392 + }, + { + "epoch": 0.7827845074632114, + "grad_norm": 2.3904080390930176, + "learning_rate": 2.2403208474578785e-05, + "loss": 0.184, + "step": 33393 + }, + { + "epoch": 0.7828079490380164, + "grad_norm": 2.8660051822662354, + "learning_rate": 2.2398562316202375e-05, + "loss": 0.1852, + "step": 33394 + }, + { + "epoch": 0.7828313906128214, + "grad_norm": 4.25522518157959, + "learning_rate": 2.239391657889358e-05, + "loss": 0.2735, + "step": 33395 + }, + { + "epoch": 0.7828548321876264, + "grad_norm": 3.3567914962768555, + "learning_rate": 2.2389271262677582e-05, + "loss": 0.2099, + "step": 33396 + }, + { + "epoch": 0.7828782737624314, + "grad_norm": 3.070460557937622, + "learning_rate": 2.2384626367579577e-05, + "loss": 0.232, + "step": 33397 + }, + { + "epoch": 0.7829017153372364, + "grad_norm": 1.6447621583938599, + "learning_rate": 2.2379981893624757e-05, + "loss": 0.0835, + "step": 33398 + }, + { + "epoch": 0.7829251569120413, + "grad_norm": 2.5045530796051025, + "learning_rate": 2.237533784083835e-05, + "loss": 0.2351, + "step": 33399 + }, + { + "epoch": 0.7829485984868464, + "grad_norm": 2.4277660846710205, + "learning_rate": 2.2370694209245523e-05, + "loss": 1.3502, + "step": 33400 + }, + { + "epoch": 0.7829720400616513, + "grad_norm": 1.3382878303527832, + "learning_rate": 2.236605099887148e-05, + "loss": 0.0619, + "step": 33401 + }, + { + "epoch": 0.7829954816364564, + "grad_norm": 0.4221794903278351, + "learning_rate": 2.2361408209741462e-05, + "loss": 0.0326, + "step": 33402 + }, + { + "epoch": 0.7830189232112613, + "grad_norm": 1.3997873067855835, + "learning_rate": 2.2356765841880622e-05, + "loss": 0.0424, + "step": 33403 + }, + { + "epoch": 0.7830423647860664, + "grad_norm": 2.331305980682373, + "learning_rate": 2.235212389531416e-05, + "loss": 0.1353, + "step": 33404 + }, + { + "epoch": 0.7830658063608713, + "grad_norm": 3.946302890777588, + "learning_rate": 2.234748237006724e-05, + "loss": 0.6717, + "step": 33405 + }, + { + "epoch": 0.7830892479356764, + "grad_norm": 1.8199379444122314, + "learning_rate": 2.2342841266165082e-05, + "loss": 0.2286, + "step": 33406 + }, + { + "epoch": 0.7831126895104813, + "grad_norm": 2.9658772945404053, + "learning_rate": 2.2338200583632818e-05, + "loss": 0.173, + "step": 33407 + }, + { + "epoch": 0.7831361310852863, + "grad_norm": 3.431410551071167, + "learning_rate": 2.2333560322495685e-05, + "loss": 0.2886, + "step": 33408 + }, + { + "epoch": 0.7831595726600913, + "grad_norm": 4.186598777770996, + "learning_rate": 2.2328920482778836e-05, + "loss": 0.3352, + "step": 33409 + }, + { + "epoch": 0.7831830142348963, + "grad_norm": 1.1835824251174927, + "learning_rate": 2.2324281064507458e-05, + "loss": 0.1694, + "step": 33410 + }, + { + "epoch": 0.7832064558097013, + "grad_norm": 1.2570356130599976, + "learning_rate": 2.23196420677067e-05, + "loss": 0.1217, + "step": 33411 + }, + { + "epoch": 0.7832298973845063, + "grad_norm": 2.5119893550872803, + "learning_rate": 2.2315003492401764e-05, + "loss": 0.1589, + "step": 33412 + }, + { + "epoch": 0.7832533389593113, + "grad_norm": 4.014684200286865, + "learning_rate": 2.231036533861777e-05, + "loss": 0.1749, + "step": 33413 + }, + { + "epoch": 0.7832767805341163, + "grad_norm": 2.4132189750671387, + "learning_rate": 2.2305727606379922e-05, + "loss": 0.1494, + "step": 33414 + }, + { + "epoch": 0.7833002221089213, + "grad_norm": 0.491199791431427, + "learning_rate": 2.23010902957134e-05, + "loss": 0.0281, + "step": 33415 + }, + { + "epoch": 0.7833236636837263, + "grad_norm": 2.2381339073181152, + "learning_rate": 2.2296453406643347e-05, + "loss": 0.21, + "step": 33416 + }, + { + "epoch": 0.7833471052585312, + "grad_norm": 1.4810758829116821, + "learning_rate": 2.2291816939194932e-05, + "loss": 0.1119, + "step": 33417 + }, + { + "epoch": 0.7833705468333363, + "grad_norm": 0.6783087849617004, + "learning_rate": 2.22871808933933e-05, + "loss": 0.0448, + "step": 33418 + }, + { + "epoch": 0.7833939884081412, + "grad_norm": 1.1173428297042847, + "learning_rate": 2.22825452692636e-05, + "loss": 0.0538, + "step": 33419 + }, + { + "epoch": 0.7834174299829463, + "grad_norm": 0.6557444334030151, + "learning_rate": 2.2277910066830977e-05, + "loss": 0.0608, + "step": 33420 + }, + { + "epoch": 0.7834408715577512, + "grad_norm": 3.369417905807495, + "learning_rate": 2.2273275286120622e-05, + "loss": 0.2877, + "step": 33421 + }, + { + "epoch": 0.7834643131325563, + "grad_norm": 3.7212514877319336, + "learning_rate": 2.2268640927157668e-05, + "loss": 0.2181, + "step": 33422 + }, + { + "epoch": 0.7834877547073612, + "grad_norm": 1.1042081117630005, + "learning_rate": 2.2264006989967245e-05, + "loss": 0.0509, + "step": 33423 + }, + { + "epoch": 0.7835111962821663, + "grad_norm": 3.0445966720581055, + "learning_rate": 2.2259373474574507e-05, + "loss": 0.2684, + "step": 33424 + }, + { + "epoch": 0.7835346378569712, + "grad_norm": 0.9244415760040283, + "learning_rate": 2.225474038100457e-05, + "loss": 0.0926, + "step": 33425 + }, + { + "epoch": 0.7835580794317762, + "grad_norm": 4.000858783721924, + "learning_rate": 2.225010770928263e-05, + "loss": 0.2064, + "step": 33426 + }, + { + "epoch": 0.7835815210065812, + "grad_norm": 4.509444236755371, + "learning_rate": 2.224547545943375e-05, + "loss": 0.4113, + "step": 33427 + }, + { + "epoch": 0.7836049625813862, + "grad_norm": 2.4211294651031494, + "learning_rate": 2.2240843631483143e-05, + "loss": 0.1104, + "step": 33428 + }, + { + "epoch": 0.7836284041561912, + "grad_norm": 0.4251936674118042, + "learning_rate": 2.223621222545591e-05, + "loss": 0.023, + "step": 33429 + }, + { + "epoch": 0.7836518457309962, + "grad_norm": 3.876556396484375, + "learning_rate": 2.2231581241377163e-05, + "loss": 0.4525, + "step": 33430 + }, + { + "epoch": 0.7836752873058012, + "grad_norm": 2.858186721801758, + "learning_rate": 2.222695067927205e-05, + "loss": 0.1728, + "step": 33431 + }, + { + "epoch": 0.7836987288806062, + "grad_norm": 5.316250324249268, + "learning_rate": 2.2222320539165663e-05, + "loss": 0.3447, + "step": 33432 + }, + { + "epoch": 0.7837221704554111, + "grad_norm": 5.95754337310791, + "learning_rate": 2.221769082108318e-05, + "loss": 0.553, + "step": 33433 + }, + { + "epoch": 0.7837456120302162, + "grad_norm": 5.0168914794921875, + "learning_rate": 2.2213061525049695e-05, + "loss": 0.475, + "step": 33434 + }, + { + "epoch": 0.7837690536050211, + "grad_norm": 3.35975980758667, + "learning_rate": 2.2208432651090327e-05, + "loss": 0.1756, + "step": 33435 + }, + { + "epoch": 0.7837924951798262, + "grad_norm": 1.7507610321044922, + "learning_rate": 2.2203804199230182e-05, + "loss": 0.1109, + "step": 33436 + }, + { + "epoch": 0.7838159367546311, + "grad_norm": 4.975030899047852, + "learning_rate": 2.2199176169494383e-05, + "loss": 0.6551, + "step": 33437 + }, + { + "epoch": 0.7838393783294362, + "grad_norm": 0.9569250345230103, + "learning_rate": 2.2194548561908024e-05, + "loss": 0.0962, + "step": 33438 + }, + { + "epoch": 0.7838628199042411, + "grad_norm": 2.4288198947906494, + "learning_rate": 2.2189921376496224e-05, + "loss": 0.1112, + "step": 33439 + }, + { + "epoch": 0.7838862614790462, + "grad_norm": 2.39866042137146, + "learning_rate": 2.2185294613284125e-05, + "loss": 1.8536, + "step": 33440 + }, + { + "epoch": 0.7839097030538511, + "grad_norm": 2.8048274517059326, + "learning_rate": 2.218066827229681e-05, + "loss": 0.2797, + "step": 33441 + }, + { + "epoch": 0.7839331446286562, + "grad_norm": 1.785355806350708, + "learning_rate": 2.217604235355938e-05, + "loss": 0.1374, + "step": 33442 + }, + { + "epoch": 0.7839565862034612, + "grad_norm": 5.080653667449951, + "learning_rate": 2.2171416857096928e-05, + "loss": 0.4727, + "step": 33443 + }, + { + "epoch": 0.7839800277782661, + "grad_norm": 1.9320751428604126, + "learning_rate": 2.216679178293456e-05, + "loss": 0.1124, + "step": 33444 + }, + { + "epoch": 0.7840034693530712, + "grad_norm": 3.0009067058563232, + "learning_rate": 2.2162167131097335e-05, + "loss": 0.2279, + "step": 33445 + }, + { + "epoch": 0.7840269109278761, + "grad_norm": 1.9233120679855347, + "learning_rate": 2.2157542901610405e-05, + "loss": 0.1375, + "step": 33446 + }, + { + "epoch": 0.7840503525026812, + "grad_norm": 2.4077110290527344, + "learning_rate": 2.2152919094498846e-05, + "loss": 0.1238, + "step": 33447 + }, + { + "epoch": 0.7840737940774861, + "grad_norm": 3.088015079498291, + "learning_rate": 2.2148295709787727e-05, + "loss": 0.3204, + "step": 33448 + }, + { + "epoch": 0.7840972356522912, + "grad_norm": 2.296441078186035, + "learning_rate": 2.2143672747502154e-05, + "loss": 0.161, + "step": 33449 + }, + { + "epoch": 0.7841206772270961, + "grad_norm": 4.048628807067871, + "learning_rate": 2.2139050207667165e-05, + "loss": 0.3695, + "step": 33450 + }, + { + "epoch": 0.7841441188019012, + "grad_norm": 2.8810324668884277, + "learning_rate": 2.213442809030791e-05, + "loss": 0.3252, + "step": 33451 + }, + { + "epoch": 0.7841675603767061, + "grad_norm": 1.3643219470977783, + "learning_rate": 2.2129806395449416e-05, + "loss": 0.0847, + "step": 33452 + }, + { + "epoch": 0.7841910019515111, + "grad_norm": 2.8323683738708496, + "learning_rate": 2.21251851231168e-05, + "loss": 0.2472, + "step": 33453 + }, + { + "epoch": 0.7842144435263161, + "grad_norm": 2.3178436756134033, + "learning_rate": 2.212056427333512e-05, + "loss": 1.3621, + "step": 33454 + }, + { + "epoch": 0.7842378851011211, + "grad_norm": 3.3733766078948975, + "learning_rate": 2.211594384612945e-05, + "loss": 0.2218, + "step": 33455 + }, + { + "epoch": 0.7842613266759261, + "grad_norm": 2.3651583194732666, + "learning_rate": 2.2111323841524867e-05, + "loss": 0.1804, + "step": 33456 + }, + { + "epoch": 0.7842847682507311, + "grad_norm": 3.066250801086426, + "learning_rate": 2.2106704259546417e-05, + "loss": 0.1946, + "step": 33457 + }, + { + "epoch": 0.7843082098255361, + "grad_norm": 2.315324306488037, + "learning_rate": 2.210208510021916e-05, + "loss": 1.0212, + "step": 33458 + }, + { + "epoch": 0.7843316514003411, + "grad_norm": 5.333724498748779, + "learning_rate": 2.2097466363568208e-05, + "loss": 0.4619, + "step": 33459 + }, + { + "epoch": 0.784355092975146, + "grad_norm": 2.4254794120788574, + "learning_rate": 2.2092848049618598e-05, + "loss": 0.1315, + "step": 33460 + }, + { + "epoch": 0.7843785345499511, + "grad_norm": 3.065934181213379, + "learning_rate": 2.2088230158395373e-05, + "loss": 0.1523, + "step": 33461 + }, + { + "epoch": 0.784401976124756, + "grad_norm": 2.230515480041504, + "learning_rate": 2.208361268992358e-05, + "loss": 0.1028, + "step": 33462 + }, + { + "epoch": 0.7844254176995611, + "grad_norm": 1.3996340036392212, + "learning_rate": 2.207899564422833e-05, + "loss": 0.8956, + "step": 33463 + }, + { + "epoch": 0.784448859274366, + "grad_norm": 2.203336000442505, + "learning_rate": 2.2074379021334635e-05, + "loss": 0.1632, + "step": 33464 + }, + { + "epoch": 0.7844723008491711, + "grad_norm": 8.271069526672363, + "learning_rate": 2.2069762821267516e-05, + "loss": 0.2246, + "step": 33465 + }, + { + "epoch": 0.784495742423976, + "grad_norm": 3.7631583213806152, + "learning_rate": 2.206514704405209e-05, + "loss": 0.2749, + "step": 33466 + }, + { + "epoch": 0.7845191839987811, + "grad_norm": 2.5367038249969482, + "learning_rate": 2.206053168971337e-05, + "loss": 1.5838, + "step": 33467 + }, + { + "epoch": 0.784542625573586, + "grad_norm": 4.092071056365967, + "learning_rate": 2.2055916758276384e-05, + "loss": 0.3705, + "step": 33468 + }, + { + "epoch": 0.784566067148391, + "grad_norm": 0.5774593949317932, + "learning_rate": 2.2051302249766203e-05, + "loss": 0.0304, + "step": 33469 + }, + { + "epoch": 0.784589508723196, + "grad_norm": 4.088608264923096, + "learning_rate": 2.2046688164207807e-05, + "loss": 0.2516, + "step": 33470 + }, + { + "epoch": 0.784612950298001, + "grad_norm": 1.047658085823059, + "learning_rate": 2.2042074501626307e-05, + "loss": 0.0395, + "step": 33471 + }, + { + "epoch": 0.784636391872806, + "grad_norm": 3.3471648693084717, + "learning_rate": 2.20374612620467e-05, + "loss": 0.5, + "step": 33472 + }, + { + "epoch": 0.784659833447611, + "grad_norm": 1.8301070928573608, + "learning_rate": 2.2032848445494016e-05, + "loss": 0.1751, + "step": 33473 + }, + { + "epoch": 0.784683275022416, + "grad_norm": 2.4149365425109863, + "learning_rate": 2.2028236051993266e-05, + "loss": 0.1716, + "step": 33474 + }, + { + "epoch": 0.784706716597221, + "grad_norm": 0.898744523525238, + "learning_rate": 2.2023624081569526e-05, + "loss": 0.0782, + "step": 33475 + }, + { + "epoch": 0.784730158172026, + "grad_norm": 3.916429042816162, + "learning_rate": 2.2019012534247796e-05, + "loss": 0.4317, + "step": 33476 + }, + { + "epoch": 0.784753599746831, + "grad_norm": 1.3679163455963135, + "learning_rate": 2.2014401410053064e-05, + "loss": 0.0693, + "step": 33477 + }, + { + "epoch": 0.784777041321636, + "grad_norm": 1.7064539194107056, + "learning_rate": 2.2009790709010412e-05, + "loss": 0.0508, + "step": 33478 + }, + { + "epoch": 0.784800482896441, + "grad_norm": 3.991717576980591, + "learning_rate": 2.2005180431144835e-05, + "loss": 0.4148, + "step": 33479 + }, + { + "epoch": 0.7848239244712459, + "grad_norm": 0.7737177610397339, + "learning_rate": 2.200057057648134e-05, + "loss": 0.0359, + "step": 33480 + }, + { + "epoch": 0.784847366046051, + "grad_norm": 1.7701917886734009, + "learning_rate": 2.199596114504494e-05, + "loss": 0.0916, + "step": 33481 + }, + { + "epoch": 0.7848708076208559, + "grad_norm": 0.9570810794830322, + "learning_rate": 2.1991352136860642e-05, + "loss": 0.1117, + "step": 33482 + }, + { + "epoch": 0.784894249195661, + "grad_norm": 2.014216423034668, + "learning_rate": 2.1986743551953447e-05, + "loss": 0.0872, + "step": 33483 + }, + { + "epoch": 0.7849176907704659, + "grad_norm": 3.1655893325805664, + "learning_rate": 2.198213539034839e-05, + "loss": 0.2872, + "step": 33484 + }, + { + "epoch": 0.784941132345271, + "grad_norm": 2.626584053039551, + "learning_rate": 2.1977527652070474e-05, + "loss": 0.2004, + "step": 33485 + }, + { + "epoch": 0.7849645739200759, + "grad_norm": 1.895869255065918, + "learning_rate": 2.1972920337144675e-05, + "loss": 0.1526, + "step": 33486 + }, + { + "epoch": 0.784988015494881, + "grad_norm": 2.9606711864471436, + "learning_rate": 2.196831344559598e-05, + "loss": 0.1377, + "step": 33487 + }, + { + "epoch": 0.7850114570696859, + "grad_norm": 5.334167957305908, + "learning_rate": 2.196370697744943e-05, + "loss": 0.6234, + "step": 33488 + }, + { + "epoch": 0.7850348986444909, + "grad_norm": 0.8773413300514221, + "learning_rate": 2.1959100932730016e-05, + "loss": 0.0611, + "step": 33489 + }, + { + "epoch": 0.7850583402192959, + "grad_norm": 0.9209595322608948, + "learning_rate": 2.1954495311462676e-05, + "loss": 0.0802, + "step": 33490 + }, + { + "epoch": 0.7850817817941009, + "grad_norm": 3.129054069519043, + "learning_rate": 2.1949890113672466e-05, + "loss": 0.2505, + "step": 33491 + }, + { + "epoch": 0.7851052233689059, + "grad_norm": 2.4937236309051514, + "learning_rate": 2.1945285339384347e-05, + "loss": 0.233, + "step": 33492 + }, + { + "epoch": 0.7851286649437109, + "grad_norm": 1.845194935798645, + "learning_rate": 2.1940680988623307e-05, + "loss": 1.184, + "step": 33493 + }, + { + "epoch": 0.785152106518516, + "grad_norm": 1.138001561164856, + "learning_rate": 2.193607706141433e-05, + "loss": 0.0896, + "step": 33494 + }, + { + "epoch": 0.7851755480933209, + "grad_norm": 4.3987016677856445, + "learning_rate": 2.1931473557782388e-05, + "loss": 0.5789, + "step": 33495 + }, + { + "epoch": 0.785198989668126, + "grad_norm": 3.090524196624756, + "learning_rate": 2.1926870477752437e-05, + "loss": 0.196, + "step": 33496 + }, + { + "epoch": 0.7852224312429309, + "grad_norm": 1.72396719455719, + "learning_rate": 2.1922267821349517e-05, + "loss": 0.1801, + "step": 33497 + }, + { + "epoch": 0.785245872817736, + "grad_norm": 1.4362894296646118, + "learning_rate": 2.191766558859857e-05, + "loss": 1.4133, + "step": 33498 + }, + { + "epoch": 0.7852693143925409, + "grad_norm": 4.385042667388916, + "learning_rate": 2.1913063779524524e-05, + "loss": 0.2375, + "step": 33499 + }, + { + "epoch": 0.7852927559673459, + "grad_norm": 3.62532901763916, + "learning_rate": 2.1908462394152428e-05, + "loss": 0.4587, + "step": 33500 + }, + { + "epoch": 0.7853161975421509, + "grad_norm": 2.8398430347442627, + "learning_rate": 2.19038614325072e-05, + "loss": 0.3737, + "step": 33501 + }, + { + "epoch": 0.7853396391169559, + "grad_norm": 1.7196851968765259, + "learning_rate": 2.1899260894613826e-05, + "loss": 0.0958, + "step": 33502 + }, + { + "epoch": 0.7853630806917609, + "grad_norm": 2.9406886100769043, + "learning_rate": 2.189466078049722e-05, + "loss": 0.0868, + "step": 33503 + }, + { + "epoch": 0.7853865222665659, + "grad_norm": 0.9409865140914917, + "learning_rate": 2.1890061090182425e-05, + "loss": 0.0731, + "step": 33504 + }, + { + "epoch": 0.7854099638413709, + "grad_norm": 2.201141834259033, + "learning_rate": 2.1885461823694343e-05, + "loss": 0.211, + "step": 33505 + }, + { + "epoch": 0.7854334054161759, + "grad_norm": 2.507385492324829, + "learning_rate": 2.188086298105795e-05, + "loss": 0.1437, + "step": 33506 + }, + { + "epoch": 0.7854568469909808, + "grad_norm": 2.0235812664031982, + "learning_rate": 2.1876264562298178e-05, + "loss": 1.1252, + "step": 33507 + }, + { + "epoch": 0.7854802885657859, + "grad_norm": 1.9482676982879639, + "learning_rate": 2.1871666567439975e-05, + "loss": 1.1751, + "step": 33508 + }, + { + "epoch": 0.7855037301405908, + "grad_norm": 3.520902395248413, + "learning_rate": 2.186706899650832e-05, + "loss": 0.3122, + "step": 33509 + }, + { + "epoch": 0.7855271717153959, + "grad_norm": 1.2629504203796387, + "learning_rate": 2.1862471849528154e-05, + "loss": 0.1358, + "step": 33510 + }, + { + "epoch": 0.7855506132902008, + "grad_norm": 3.66188383102417, + "learning_rate": 2.1857875126524386e-05, + "loss": 0.4322, + "step": 33511 + }, + { + "epoch": 0.7855740548650059, + "grad_norm": 1.919978141784668, + "learning_rate": 2.1853278827522017e-05, + "loss": 0.1748, + "step": 33512 + }, + { + "epoch": 0.7855974964398108, + "grad_norm": 1.8284848928451538, + "learning_rate": 2.184868295254595e-05, + "loss": 0.1233, + "step": 33513 + }, + { + "epoch": 0.7856209380146159, + "grad_norm": 3.0537798404693604, + "learning_rate": 2.1844087501621125e-05, + "loss": 0.162, + "step": 33514 + }, + { + "epoch": 0.7856443795894208, + "grad_norm": 1.6774542331695557, + "learning_rate": 2.1839492474772448e-05, + "loss": 0.0933, + "step": 33515 + }, + { + "epoch": 0.7856678211642258, + "grad_norm": 3.6883325576782227, + "learning_rate": 2.1834897872024917e-05, + "loss": 0.356, + "step": 33516 + }, + { + "epoch": 0.7856912627390308, + "grad_norm": 4.691127777099609, + "learning_rate": 2.1830303693403432e-05, + "loss": 0.3895, + "step": 33517 + }, + { + "epoch": 0.7857147043138358, + "grad_norm": 4.402252197265625, + "learning_rate": 2.1825709938932903e-05, + "loss": 0.3961, + "step": 33518 + }, + { + "epoch": 0.7857381458886408, + "grad_norm": 3.0475573539733887, + "learning_rate": 2.1821116608638283e-05, + "loss": 0.2519, + "step": 33519 + }, + { + "epoch": 0.7857615874634458, + "grad_norm": 0.860304057598114, + "learning_rate": 2.1816523702544478e-05, + "loss": 0.0731, + "step": 33520 + }, + { + "epoch": 0.7857850290382508, + "grad_norm": 2.8043324947357178, + "learning_rate": 2.1811931220676384e-05, + "loss": 0.2257, + "step": 33521 + }, + { + "epoch": 0.7858084706130558, + "grad_norm": 0.4703480005264282, + "learning_rate": 2.1807339163058972e-05, + "loss": 0.0301, + "step": 33522 + }, + { + "epoch": 0.7858319121878607, + "grad_norm": 1.4776086807250977, + "learning_rate": 2.1802747529717115e-05, + "loss": 0.0879, + "step": 33523 + }, + { + "epoch": 0.7858553537626658, + "grad_norm": 2.058544158935547, + "learning_rate": 2.1798156320675778e-05, + "loss": 0.0569, + "step": 33524 + }, + { + "epoch": 0.7858787953374707, + "grad_norm": 2.8757166862487793, + "learning_rate": 2.1793565535959838e-05, + "loss": 1.9781, + "step": 33525 + }, + { + "epoch": 0.7859022369122758, + "grad_norm": 2.0013577938079834, + "learning_rate": 2.178897517559422e-05, + "loss": 0.1347, + "step": 33526 + }, + { + "epoch": 0.7859256784870807, + "grad_norm": 2.6865997314453125, + "learning_rate": 2.178438523960381e-05, + "loss": 0.3265, + "step": 33527 + }, + { + "epoch": 0.7859491200618858, + "grad_norm": 2.3473737239837646, + "learning_rate": 2.177979572801351e-05, + "loss": 0.0623, + "step": 33528 + }, + { + "epoch": 0.7859725616366907, + "grad_norm": 3.8945653438568115, + "learning_rate": 2.1775206640848256e-05, + "loss": 0.2063, + "step": 33529 + }, + { + "epoch": 0.7859960032114958, + "grad_norm": 3.176610231399536, + "learning_rate": 2.1770617978132934e-05, + "loss": 0.3002, + "step": 33530 + }, + { + "epoch": 0.7860194447863007, + "grad_norm": 1.894875407218933, + "learning_rate": 2.1766029739892435e-05, + "loss": 0.1426, + "step": 33531 + }, + { + "epoch": 0.7860428863611058, + "grad_norm": 1.699929118156433, + "learning_rate": 2.176144192615166e-05, + "loss": 1.1842, + "step": 33532 + }, + { + "epoch": 0.7860663279359107, + "grad_norm": 3.3005878925323486, + "learning_rate": 2.17568545369355e-05, + "loss": 0.2371, + "step": 33533 + }, + { + "epoch": 0.7860897695107157, + "grad_norm": 1.848171353340149, + "learning_rate": 2.1752267572268827e-05, + "loss": 1.3697, + "step": 33534 + }, + { + "epoch": 0.7861132110855207, + "grad_norm": 0.6780487895011902, + "learning_rate": 2.1747681032176535e-05, + "loss": 0.0686, + "step": 33535 + }, + { + "epoch": 0.7861366526603257, + "grad_norm": 3.446805238723755, + "learning_rate": 2.174309491668357e-05, + "loss": 1.7228, + "step": 33536 + }, + { + "epoch": 0.7861600942351307, + "grad_norm": 1.241605281829834, + "learning_rate": 2.1738509225814762e-05, + "loss": 0.108, + "step": 33537 + }, + { + "epoch": 0.7861835358099357, + "grad_norm": 0.5799511671066284, + "learning_rate": 2.1733923959595015e-05, + "loss": 0.0471, + "step": 33538 + }, + { + "epoch": 0.7862069773847407, + "grad_norm": 2.5676472187042236, + "learning_rate": 2.1729339118049185e-05, + "loss": 0.3427, + "step": 33539 + }, + { + "epoch": 0.7862304189595457, + "grad_norm": 1.9526058435440063, + "learning_rate": 2.1724754701202144e-05, + "loss": 0.891, + "step": 33540 + }, + { + "epoch": 0.7862538605343506, + "grad_norm": 0.32410505414009094, + "learning_rate": 2.1720170709078802e-05, + "loss": 0.0126, + "step": 33541 + }, + { + "epoch": 0.7862773021091557, + "grad_norm": 2.417262315750122, + "learning_rate": 2.1715587141704018e-05, + "loss": 0.1686, + "step": 33542 + }, + { + "epoch": 0.7863007436839606, + "grad_norm": 2.043966054916382, + "learning_rate": 2.1711003999102664e-05, + "loss": 0.1481, + "step": 33543 + }, + { + "epoch": 0.7863241852587657, + "grad_norm": 0.3704449534416199, + "learning_rate": 2.1706421281299595e-05, + "loss": 0.0125, + "step": 33544 + }, + { + "epoch": 0.7863476268335707, + "grad_norm": 1.9477286338806152, + "learning_rate": 2.170183898831969e-05, + "loss": 0.159, + "step": 33545 + }, + { + "epoch": 0.7863710684083757, + "grad_norm": 2.876892566680908, + "learning_rate": 2.1697257120187787e-05, + "loss": 0.1791, + "step": 33546 + }, + { + "epoch": 0.7863945099831807, + "grad_norm": 2.9600372314453125, + "learning_rate": 2.1692675676928763e-05, + "loss": 0.2938, + "step": 33547 + }, + { + "epoch": 0.7864179515579857, + "grad_norm": 2.6633682250976562, + "learning_rate": 2.1688094658567514e-05, + "loss": 0.1317, + "step": 33548 + }, + { + "epoch": 0.7864413931327907, + "grad_norm": 0.9917047023773193, + "learning_rate": 2.1683514065128863e-05, + "loss": 0.0738, + "step": 33549 + }, + { + "epoch": 0.7864648347075957, + "grad_norm": 0.7511430382728577, + "learning_rate": 2.1678933896637667e-05, + "loss": 0.0751, + "step": 33550 + }, + { + "epoch": 0.7864882762824007, + "grad_norm": 2.290957450866699, + "learning_rate": 2.167435415311877e-05, + "loss": 0.1701, + "step": 33551 + }, + { + "epoch": 0.7865117178572056, + "grad_norm": 2.862689733505249, + "learning_rate": 2.1669774834597036e-05, + "loss": 0.1426, + "step": 33552 + }, + { + "epoch": 0.7865351594320107, + "grad_norm": 2.3308358192443848, + "learning_rate": 2.166519594109728e-05, + "loss": 0.2315, + "step": 33553 + }, + { + "epoch": 0.7865586010068156, + "grad_norm": 5.763546466827393, + "learning_rate": 2.1660617472644395e-05, + "loss": 0.4353, + "step": 33554 + }, + { + "epoch": 0.7865820425816207, + "grad_norm": 2.960228443145752, + "learning_rate": 2.165603942926321e-05, + "loss": 0.245, + "step": 33555 + }, + { + "epoch": 0.7866054841564256, + "grad_norm": 2.353247880935669, + "learning_rate": 2.1651461810978557e-05, + "loss": 0.1599, + "step": 33556 + }, + { + "epoch": 0.7866289257312307, + "grad_norm": 2.1657772064208984, + "learning_rate": 2.1646884617815265e-05, + "loss": 0.2087, + "step": 33557 + }, + { + "epoch": 0.7866523673060356, + "grad_norm": 5.445622444152832, + "learning_rate": 2.1642307849798195e-05, + "loss": 0.3871, + "step": 33558 + }, + { + "epoch": 0.7866758088808407, + "grad_norm": 2.7025859355926514, + "learning_rate": 2.163773150695213e-05, + "loss": 0.2143, + "step": 33559 + }, + { + "epoch": 0.7866992504556456, + "grad_norm": 1.4590331315994263, + "learning_rate": 2.1633155589301935e-05, + "loss": 0.0704, + "step": 33560 + }, + { + "epoch": 0.7867226920304506, + "grad_norm": 2.5263659954071045, + "learning_rate": 2.1628580096872475e-05, + "loss": 0.0967, + "step": 33561 + }, + { + "epoch": 0.7867461336052556, + "grad_norm": 2.438979148864746, + "learning_rate": 2.1624005029688544e-05, + "loss": 0.1091, + "step": 33562 + }, + { + "epoch": 0.7867695751800606, + "grad_norm": 3.397369623184204, + "learning_rate": 2.1619430387774963e-05, + "loss": 0.2966, + "step": 33563 + }, + { + "epoch": 0.7867930167548656, + "grad_norm": 1.0367934703826904, + "learning_rate": 2.161485617115655e-05, + "loss": 0.0834, + "step": 33564 + }, + { + "epoch": 0.7868164583296706, + "grad_norm": 1.716569423675537, + "learning_rate": 2.161028237985814e-05, + "loss": 0.0581, + "step": 33565 + }, + { + "epoch": 0.7868398999044756, + "grad_norm": 5.428438663482666, + "learning_rate": 2.1605709013904507e-05, + "loss": 0.7283, + "step": 33566 + }, + { + "epoch": 0.7868633414792806, + "grad_norm": 1.1158778667449951, + "learning_rate": 2.1601136073320538e-05, + "loss": 0.0996, + "step": 33567 + }, + { + "epoch": 0.7868867830540855, + "grad_norm": 1.240929365158081, + "learning_rate": 2.1596563558131e-05, + "loss": 0.0546, + "step": 33568 + }, + { + "epoch": 0.7869102246288906, + "grad_norm": 2.467357873916626, + "learning_rate": 2.1591991468360717e-05, + "loss": 0.1905, + "step": 33569 + }, + { + "epoch": 0.7869336662036955, + "grad_norm": 1.0347603559494019, + "learning_rate": 2.1587419804034493e-05, + "loss": 0.092, + "step": 33570 + }, + { + "epoch": 0.7869571077785006, + "grad_norm": 3.5663552284240723, + "learning_rate": 2.1582848565177138e-05, + "loss": 0.3916, + "step": 33571 + }, + { + "epoch": 0.7869805493533055, + "grad_norm": 0.5442981123924255, + "learning_rate": 2.1578277751813413e-05, + "loss": 0.0428, + "step": 33572 + }, + { + "epoch": 0.7870039909281106, + "grad_norm": 2.369919538497925, + "learning_rate": 2.1573707363968166e-05, + "loss": 0.2369, + "step": 33573 + }, + { + "epoch": 0.7870274325029155, + "grad_norm": 2.404942035675049, + "learning_rate": 2.1569137401666216e-05, + "loss": 0.1261, + "step": 33574 + }, + { + "epoch": 0.7870508740777206, + "grad_norm": 0.8727399706840515, + "learning_rate": 2.156456786493233e-05, + "loss": 0.049, + "step": 33575 + }, + { + "epoch": 0.7870743156525255, + "grad_norm": 1.3549033403396606, + "learning_rate": 2.15599987537913e-05, + "loss": 0.1555, + "step": 33576 + }, + { + "epoch": 0.7870977572273306, + "grad_norm": 2.4738850593566895, + "learning_rate": 2.1555430068267934e-05, + "loss": 0.2122, + "step": 33577 + }, + { + "epoch": 0.7871211988021355, + "grad_norm": 2.2101078033447266, + "learning_rate": 2.1550861808386968e-05, + "loss": 1.3266, + "step": 33578 + }, + { + "epoch": 0.7871446403769405, + "grad_norm": 3.9412972927093506, + "learning_rate": 2.1546293974173272e-05, + "loss": 0.2854, + "step": 33579 + }, + { + "epoch": 0.7871680819517455, + "grad_norm": 0.7721675634384155, + "learning_rate": 2.154172656565159e-05, + "loss": 0.0492, + "step": 33580 + }, + { + "epoch": 0.7871915235265505, + "grad_norm": 1.6692628860473633, + "learning_rate": 2.153715958284669e-05, + "loss": 0.0952, + "step": 33581 + }, + { + "epoch": 0.7872149651013555, + "grad_norm": 2.7833242416381836, + "learning_rate": 2.1532593025783377e-05, + "loss": 0.2067, + "step": 33582 + }, + { + "epoch": 0.7872384066761605, + "grad_norm": 2.2898285388946533, + "learning_rate": 2.152802689448643e-05, + "loss": 0.1524, + "step": 33583 + }, + { + "epoch": 0.7872618482509655, + "grad_norm": 0.6560136675834656, + "learning_rate": 2.1523461188980577e-05, + "loss": 0.0536, + "step": 33584 + }, + { + "epoch": 0.7872852898257705, + "grad_norm": 3.0320630073547363, + "learning_rate": 2.1518895909290636e-05, + "loss": 0.3184, + "step": 33585 + }, + { + "epoch": 0.7873087314005754, + "grad_norm": 2.369389533996582, + "learning_rate": 2.1514331055441396e-05, + "loss": 0.0936, + "step": 33586 + }, + { + "epoch": 0.7873321729753805, + "grad_norm": 1.3271995782852173, + "learning_rate": 2.1509766627457605e-05, + "loss": 0.1011, + "step": 33587 + }, + { + "epoch": 0.7873556145501854, + "grad_norm": 0.9348220229148865, + "learning_rate": 2.150520262536403e-05, + "loss": 0.5758, + "step": 33588 + }, + { + "epoch": 0.7873790561249905, + "grad_norm": 2.9462666511535645, + "learning_rate": 2.1500639049185423e-05, + "loss": 0.21, + "step": 33589 + }, + { + "epoch": 0.7874024976997954, + "grad_norm": 1.2859944105148315, + "learning_rate": 2.1496075898946567e-05, + "loss": 0.1171, + "step": 33590 + }, + { + "epoch": 0.7874259392746005, + "grad_norm": 2.4671812057495117, + "learning_rate": 2.1491513174672175e-05, + "loss": 0.1194, + "step": 33591 + }, + { + "epoch": 0.7874493808494054, + "grad_norm": 2.983431577682495, + "learning_rate": 2.148695087638707e-05, + "loss": 0.1798, + "step": 33592 + }, + { + "epoch": 0.7874728224242105, + "grad_norm": 1.0067362785339355, + "learning_rate": 2.1482389004115976e-05, + "loss": 0.0793, + "step": 33593 + }, + { + "epoch": 0.7874962639990154, + "grad_norm": 2.2468459606170654, + "learning_rate": 2.1477827557883645e-05, + "loss": 1.2414, + "step": 33594 + }, + { + "epoch": 0.7875197055738204, + "grad_norm": 3.0313103199005127, + "learning_rate": 2.147326653771483e-05, + "loss": 0.1908, + "step": 33595 + }, + { + "epoch": 0.7875431471486255, + "grad_norm": 0.5742824673652649, + "learning_rate": 2.146870594363426e-05, + "loss": 0.0362, + "step": 33596 + }, + { + "epoch": 0.7875665887234304, + "grad_norm": 1.8643723726272583, + "learning_rate": 2.146414577566672e-05, + "loss": 0.1451, + "step": 33597 + }, + { + "epoch": 0.7875900302982355, + "grad_norm": 2.301365375518799, + "learning_rate": 2.145958603383691e-05, + "loss": 0.2705, + "step": 33598 + }, + { + "epoch": 0.7876134718730404, + "grad_norm": 2.3047287464141846, + "learning_rate": 2.1455026718169612e-05, + "loss": 0.131, + "step": 33599 + }, + { + "epoch": 0.7876369134478455, + "grad_norm": 1.3603829145431519, + "learning_rate": 2.1450467828689558e-05, + "loss": 0.1596, + "step": 33600 + }, + { + "epoch": 0.7876603550226504, + "grad_norm": 1.8868788480758667, + "learning_rate": 2.1445909365421467e-05, + "loss": 1.0127, + "step": 33601 + }, + { + "epoch": 0.7876837965974555, + "grad_norm": 3.429724931716919, + "learning_rate": 2.144135132839009e-05, + "loss": 0.3311, + "step": 33602 + }, + { + "epoch": 0.7877072381722604, + "grad_norm": 3.8494460582733154, + "learning_rate": 2.1436793717620142e-05, + "loss": 0.2985, + "step": 33603 + }, + { + "epoch": 0.7877306797470655, + "grad_norm": 1.6667954921722412, + "learning_rate": 2.143223653313633e-05, + "loss": 0.1396, + "step": 33604 + }, + { + "epoch": 0.7877541213218704, + "grad_norm": 1.2904245853424072, + "learning_rate": 2.1427679774963448e-05, + "loss": 0.0806, + "step": 33605 + }, + { + "epoch": 0.7877775628966754, + "grad_norm": 0.5602394342422485, + "learning_rate": 2.1423123443126182e-05, + "loss": 0.0484, + "step": 33606 + }, + { + "epoch": 0.7878010044714804, + "grad_norm": 4.397265434265137, + "learning_rate": 2.141856753764926e-05, + "loss": 0.162, + "step": 33607 + }, + { + "epoch": 0.7878244460462854, + "grad_norm": 0.39010027050971985, + "learning_rate": 2.1414012058557364e-05, + "loss": 0.0156, + "step": 33608 + }, + { + "epoch": 0.7878478876210904, + "grad_norm": 1.6317682266235352, + "learning_rate": 2.1409457005875276e-05, + "loss": 0.1304, + "step": 33609 + }, + { + "epoch": 0.7878713291958954, + "grad_norm": 3.225147008895874, + "learning_rate": 2.140490237962769e-05, + "loss": 0.2501, + "step": 33610 + }, + { + "epoch": 0.7878947707707004, + "grad_norm": 1.5652810335159302, + "learning_rate": 2.140034817983929e-05, + "loss": 0.0803, + "step": 33611 + }, + { + "epoch": 0.7879182123455054, + "grad_norm": 2.2593390941619873, + "learning_rate": 2.139579440653483e-05, + "loss": 0.1428, + "step": 33612 + }, + { + "epoch": 0.7879416539203103, + "grad_norm": 2.1006529331207275, + "learning_rate": 2.1391241059739e-05, + "loss": 0.24, + "step": 33613 + }, + { + "epoch": 0.7879650954951154, + "grad_norm": 1.1486154794692993, + "learning_rate": 2.1386688139476508e-05, + "loss": 0.0735, + "step": 33614 + }, + { + "epoch": 0.7879885370699203, + "grad_norm": 1.6649174690246582, + "learning_rate": 2.138213564577206e-05, + "loss": 0.1204, + "step": 33615 + }, + { + "epoch": 0.7880119786447254, + "grad_norm": 0.7784475088119507, + "learning_rate": 2.1377583578650328e-05, + "loss": 0.0529, + "step": 33616 + }, + { + "epoch": 0.7880354202195303, + "grad_norm": 1.6529085636138916, + "learning_rate": 2.1373031938136055e-05, + "loss": 0.0854, + "step": 33617 + }, + { + "epoch": 0.7880588617943354, + "grad_norm": 4.197923183441162, + "learning_rate": 2.1368480724253935e-05, + "loss": 0.2117, + "step": 33618 + }, + { + "epoch": 0.7880823033691403, + "grad_norm": 3.844926595687866, + "learning_rate": 2.1363929937028636e-05, + "loss": 0.156, + "step": 33619 + }, + { + "epoch": 0.7881057449439454, + "grad_norm": 2.029557704925537, + "learning_rate": 2.135937957648485e-05, + "loss": 0.2047, + "step": 33620 + }, + { + "epoch": 0.7881291865187503, + "grad_norm": 3.796889305114746, + "learning_rate": 2.13548296426473e-05, + "loss": 0.3351, + "step": 33621 + }, + { + "epoch": 0.7881526280935554, + "grad_norm": 2.2721433639526367, + "learning_rate": 2.1350280135540657e-05, + "loss": 0.095, + "step": 33622 + }, + { + "epoch": 0.7881760696683603, + "grad_norm": 2.1095426082611084, + "learning_rate": 2.1345731055189576e-05, + "loss": 0.082, + "step": 33623 + }, + { + "epoch": 0.7881995112431653, + "grad_norm": 0.5310831665992737, + "learning_rate": 2.1341182401618807e-05, + "loss": 0.062, + "step": 33624 + }, + { + "epoch": 0.7882229528179703, + "grad_norm": 3.0854432582855225, + "learning_rate": 2.133663417485299e-05, + "loss": 0.338, + "step": 33625 + }, + { + "epoch": 0.7882463943927753, + "grad_norm": 2.2670652866363525, + "learning_rate": 2.1332086374916803e-05, + "loss": 0.9729, + "step": 33626 + }, + { + "epoch": 0.7882698359675803, + "grad_norm": 1.7031468152999878, + "learning_rate": 2.132753900183493e-05, + "loss": 0.1252, + "step": 33627 + }, + { + "epoch": 0.7882932775423853, + "grad_norm": 2.573676347732544, + "learning_rate": 2.132299205563204e-05, + "loss": 0.2767, + "step": 33628 + }, + { + "epoch": 0.7883167191171903, + "grad_norm": 1.5441279411315918, + "learning_rate": 2.1318445536332787e-05, + "loss": 0.0946, + "step": 33629 + }, + { + "epoch": 0.7883401606919953, + "grad_norm": 2.008157968521118, + "learning_rate": 2.1313899443961884e-05, + "loss": 0.1492, + "step": 33630 + }, + { + "epoch": 0.7883636022668002, + "grad_norm": 2.2416303157806396, + "learning_rate": 2.130935377854397e-05, + "loss": 0.1703, + "step": 33631 + }, + { + "epoch": 0.7883870438416053, + "grad_norm": 0.2809949517250061, + "learning_rate": 2.1304808540103693e-05, + "loss": 0.0111, + "step": 33632 + }, + { + "epoch": 0.7884104854164102, + "grad_norm": 2.452960729598999, + "learning_rate": 2.130026372866577e-05, + "loss": 0.2258, + "step": 33633 + }, + { + "epoch": 0.7884339269912153, + "grad_norm": 1.7859073877334595, + "learning_rate": 2.1295719344254815e-05, + "loss": 0.0997, + "step": 33634 + }, + { + "epoch": 0.7884573685660202, + "grad_norm": 1.4691256284713745, + "learning_rate": 2.129117538689551e-05, + "loss": 0.1055, + "step": 33635 + }, + { + "epoch": 0.7884808101408253, + "grad_norm": 0.5918492674827576, + "learning_rate": 2.1286631856612472e-05, + "loss": 0.0444, + "step": 33636 + }, + { + "epoch": 0.7885042517156302, + "grad_norm": 1.5462766885757446, + "learning_rate": 2.1282088753430407e-05, + "loss": 0.0708, + "step": 33637 + }, + { + "epoch": 0.7885276932904353, + "grad_norm": 2.98380184173584, + "learning_rate": 2.1277546077373946e-05, + "loss": 0.2115, + "step": 33638 + }, + { + "epoch": 0.7885511348652402, + "grad_norm": 2.487456798553467, + "learning_rate": 2.127300382846773e-05, + "loss": 0.1643, + "step": 33639 + }, + { + "epoch": 0.7885745764400452, + "grad_norm": 0.771451473236084, + "learning_rate": 2.1268462006736413e-05, + "loss": 0.0538, + "step": 33640 + }, + { + "epoch": 0.7885980180148502, + "grad_norm": 0.6151390075683594, + "learning_rate": 2.1263920612204626e-05, + "loss": 0.0408, + "step": 33641 + }, + { + "epoch": 0.7886214595896552, + "grad_norm": 4.77860689163208, + "learning_rate": 2.1259379644897002e-05, + "loss": 0.157, + "step": 33642 + }, + { + "epoch": 0.7886449011644602, + "grad_norm": 2.508730411529541, + "learning_rate": 2.125483910483822e-05, + "loss": 0.1438, + "step": 33643 + }, + { + "epoch": 0.7886683427392652, + "grad_norm": 0.626545250415802, + "learning_rate": 2.1250298992052896e-05, + "loss": 0.0306, + "step": 33644 + }, + { + "epoch": 0.7886917843140702, + "grad_norm": 2.9583818912506104, + "learning_rate": 2.124575930656564e-05, + "loss": 0.1364, + "step": 33645 + }, + { + "epoch": 0.7887152258888752, + "grad_norm": 0.764798104763031, + "learning_rate": 2.1241220048401134e-05, + "loss": 0.0907, + "step": 33646 + }, + { + "epoch": 0.7887386674636803, + "grad_norm": 3.929380178451538, + "learning_rate": 2.1236681217583974e-05, + "loss": 0.4152, + "step": 33647 + }, + { + "epoch": 0.7887621090384852, + "grad_norm": 2.4916017055511475, + "learning_rate": 2.123214281413881e-05, + "loss": 0.1536, + "step": 33648 + }, + { + "epoch": 0.7887855506132903, + "grad_norm": 2.382112741470337, + "learning_rate": 2.122760483809022e-05, + "loss": 0.122, + "step": 33649 + }, + { + "epoch": 0.7888089921880952, + "grad_norm": 2.983894109725952, + "learning_rate": 2.1223067289462884e-05, + "loss": 0.3289, + "step": 33650 + }, + { + "epoch": 0.7888324337629002, + "grad_norm": 4.215008735656738, + "learning_rate": 2.1218530168281414e-05, + "loss": 0.2842, + "step": 33651 + }, + { + "epoch": 0.7888558753377052, + "grad_norm": 0.9310508966445923, + "learning_rate": 2.1213993474570403e-05, + "loss": 0.0607, + "step": 33652 + }, + { + "epoch": 0.7888793169125102, + "grad_norm": 2.4241838455200195, + "learning_rate": 2.120945720835448e-05, + "loss": 0.1019, + "step": 33653 + }, + { + "epoch": 0.7889027584873152, + "grad_norm": 1.9548858404159546, + "learning_rate": 2.1204921369658228e-05, + "loss": 0.0911, + "step": 33654 + }, + { + "epoch": 0.7889262000621202, + "grad_norm": 2.6285574436187744, + "learning_rate": 2.120038595850632e-05, + "loss": 1.6112, + "step": 33655 + }, + { + "epoch": 0.7889496416369252, + "grad_norm": 2.642373561859131, + "learning_rate": 2.1195850974923336e-05, + "loss": 0.1321, + "step": 33656 + }, + { + "epoch": 0.7889730832117302, + "grad_norm": 2.8332974910736084, + "learning_rate": 2.1191316418933848e-05, + "loss": 0.2, + "step": 33657 + }, + { + "epoch": 0.7889965247865351, + "grad_norm": 1.329107403755188, + "learning_rate": 2.1186782290562522e-05, + "loss": 0.5211, + "step": 33658 + }, + { + "epoch": 0.7890199663613402, + "grad_norm": 0.6167054772377014, + "learning_rate": 2.118224858983393e-05, + "loss": 0.0602, + "step": 33659 + }, + { + "epoch": 0.7890434079361451, + "grad_norm": 3.5488555431365967, + "learning_rate": 2.1177715316772673e-05, + "loss": 0.1702, + "step": 33660 + }, + { + "epoch": 0.7890668495109502, + "grad_norm": 1.6023849248886108, + "learning_rate": 2.117318247140332e-05, + "loss": 0.1637, + "step": 33661 + }, + { + "epoch": 0.7890902910857551, + "grad_norm": 2.5336477756500244, + "learning_rate": 2.1168650053750528e-05, + "loss": 0.1532, + "step": 33662 + }, + { + "epoch": 0.7891137326605602, + "grad_norm": 5.008472442626953, + "learning_rate": 2.1164118063838856e-05, + "loss": 0.417, + "step": 33663 + }, + { + "epoch": 0.7891371742353651, + "grad_norm": 4.02490758895874, + "learning_rate": 2.1159586501692897e-05, + "loss": 0.3993, + "step": 33664 + }, + { + "epoch": 0.7891606158101702, + "grad_norm": 2.2249345779418945, + "learning_rate": 2.1155055367337227e-05, + "loss": 0.2039, + "step": 33665 + }, + { + "epoch": 0.7891840573849751, + "grad_norm": 2.9566946029663086, + "learning_rate": 2.1150524660796456e-05, + "loss": 0.3322, + "step": 33666 + }, + { + "epoch": 0.7892074989597802, + "grad_norm": 1.5255078077316284, + "learning_rate": 2.1145994382095123e-05, + "loss": 0.143, + "step": 33667 + }, + { + "epoch": 0.7892309405345851, + "grad_norm": 0.8055552244186401, + "learning_rate": 2.114146453125787e-05, + "loss": 0.0738, + "step": 33668 + }, + { + "epoch": 0.7892543821093901, + "grad_norm": 6.159266948699951, + "learning_rate": 2.113693510830922e-05, + "loss": 0.4549, + "step": 33669 + }, + { + "epoch": 0.7892778236841951, + "grad_norm": 3.1406102180480957, + "learning_rate": 2.1132406113273805e-05, + "loss": 0.2272, + "step": 33670 + }, + { + "epoch": 0.7893012652590001, + "grad_norm": 3.0979161262512207, + "learning_rate": 2.1127877546176177e-05, + "loss": 0.1602, + "step": 33671 + }, + { + "epoch": 0.7893247068338051, + "grad_norm": 2.5867507457733154, + "learning_rate": 2.1123349407040892e-05, + "loss": 0.1838, + "step": 33672 + }, + { + "epoch": 0.7893481484086101, + "grad_norm": 3.3391735553741455, + "learning_rate": 2.1118821695892545e-05, + "loss": 0.3018, + "step": 33673 + }, + { + "epoch": 0.7893715899834151, + "grad_norm": 1.760242223739624, + "learning_rate": 2.1114294412755654e-05, + "loss": 1.085, + "step": 33674 + }, + { + "epoch": 0.7893950315582201, + "grad_norm": 0.7828405499458313, + "learning_rate": 2.1109767557654846e-05, + "loss": 0.0633, + "step": 33675 + }, + { + "epoch": 0.789418473133025, + "grad_norm": 0.9122900366783142, + "learning_rate": 2.110524113061466e-05, + "loss": 0.0778, + "step": 33676 + }, + { + "epoch": 0.7894419147078301, + "grad_norm": 1.292775273323059, + "learning_rate": 2.1100715131659654e-05, + "loss": 0.0619, + "step": 33677 + }, + { + "epoch": 0.789465356282635, + "grad_norm": 4.071953296661377, + "learning_rate": 2.1096189560814385e-05, + "loss": 0.4904, + "step": 33678 + }, + { + "epoch": 0.7894887978574401, + "grad_norm": 3.6147711277008057, + "learning_rate": 2.1091664418103406e-05, + "loss": 0.3165, + "step": 33679 + }, + { + "epoch": 0.789512239432245, + "grad_norm": 0.7070180773735046, + "learning_rate": 2.108713970355126e-05, + "loss": 0.0554, + "step": 33680 + }, + { + "epoch": 0.7895356810070501, + "grad_norm": 1.7715644836425781, + "learning_rate": 2.1082615417182505e-05, + "loss": 0.1111, + "step": 33681 + }, + { + "epoch": 0.789559122581855, + "grad_norm": 1.025364637374878, + "learning_rate": 2.1078091559021717e-05, + "loss": 0.0713, + "step": 33682 + }, + { + "epoch": 0.7895825641566601, + "grad_norm": 4.050548553466797, + "learning_rate": 2.1073568129093435e-05, + "loss": 0.2205, + "step": 33683 + }, + { + "epoch": 0.789606005731465, + "grad_norm": 1.019856333732605, + "learning_rate": 2.1069045127422193e-05, + "loss": 0.0877, + "step": 33684 + }, + { + "epoch": 0.78962944730627, + "grad_norm": 2.065908670425415, + "learning_rate": 2.1064522554032516e-05, + "loss": 0.1273, + "step": 33685 + }, + { + "epoch": 0.789652888881075, + "grad_norm": 1.4513112306594849, + "learning_rate": 2.1060000408948976e-05, + "loss": 0.1345, + "step": 33686 + }, + { + "epoch": 0.78967633045588, + "grad_norm": 3.3075551986694336, + "learning_rate": 2.105547869219605e-05, + "loss": 0.3394, + "step": 33687 + }, + { + "epoch": 0.789699772030685, + "grad_norm": 2.489579439163208, + "learning_rate": 2.1050957403798354e-05, + "loss": 1.0909, + "step": 33688 + }, + { + "epoch": 0.78972321360549, + "grad_norm": 4.4511847496032715, + "learning_rate": 2.104643654378039e-05, + "loss": 0.4526, + "step": 33689 + }, + { + "epoch": 0.789746655180295, + "grad_norm": 2.8900816440582275, + "learning_rate": 2.104191611216667e-05, + "loss": 0.1566, + "step": 33690 + }, + { + "epoch": 0.7897700967551, + "grad_norm": 3.047307014465332, + "learning_rate": 2.1037396108981734e-05, + "loss": 0.1691, + "step": 33691 + }, + { + "epoch": 0.789793538329905, + "grad_norm": 0.9716901779174805, + "learning_rate": 2.1032876534250078e-05, + "loss": 0.069, + "step": 33692 + }, + { + "epoch": 0.78981697990471, + "grad_norm": 0.6692085862159729, + "learning_rate": 2.102835738799629e-05, + "loss": 0.0293, + "step": 33693 + }, + { + "epoch": 0.7898404214795149, + "grad_norm": 1.1392920017242432, + "learning_rate": 2.1023838670244822e-05, + "loss": 0.0936, + "step": 33694 + }, + { + "epoch": 0.78986386305432, + "grad_norm": 0.5593820214271545, + "learning_rate": 2.1019320381020257e-05, + "loss": 0.0464, + "step": 33695 + }, + { + "epoch": 0.7898873046291249, + "grad_norm": 2.6424577236175537, + "learning_rate": 2.1014802520347077e-05, + "loss": 0.194, + "step": 33696 + }, + { + "epoch": 0.78991074620393, + "grad_norm": 1.6776336431503296, + "learning_rate": 2.1010285088249805e-05, + "loss": 0.0807, + "step": 33697 + }, + { + "epoch": 0.789934187778735, + "grad_norm": 2.5008161067962646, + "learning_rate": 2.100576808475294e-05, + "loss": 0.131, + "step": 33698 + }, + { + "epoch": 0.78995762935354, + "grad_norm": 2.0458943843841553, + "learning_rate": 2.100125150988097e-05, + "loss": 0.1993, + "step": 33699 + }, + { + "epoch": 0.789981070928345, + "grad_norm": 1.4473787546157837, + "learning_rate": 2.099673536365846e-05, + "loss": 0.0799, + "step": 33700 + }, + { + "epoch": 0.79000451250315, + "grad_norm": 4.285305500030518, + "learning_rate": 2.099221964610989e-05, + "loss": 0.4255, + "step": 33701 + }, + { + "epoch": 0.790027954077955, + "grad_norm": 2.7343568801879883, + "learning_rate": 2.098770435725975e-05, + "loss": 0.3367, + "step": 33702 + }, + { + "epoch": 0.79005139565276, + "grad_norm": 1.007234811782837, + "learning_rate": 2.098318949713256e-05, + "loss": 0.1225, + "step": 33703 + }, + { + "epoch": 0.790074837227565, + "grad_norm": 2.9186205863952637, + "learning_rate": 2.0978675065752794e-05, + "loss": 0.1971, + "step": 33704 + }, + { + "epoch": 0.7900982788023699, + "grad_norm": 2.2345759868621826, + "learning_rate": 2.0974161063144937e-05, + "loss": 0.2257, + "step": 33705 + }, + { + "epoch": 0.790121720377175, + "grad_norm": 2.296231985092163, + "learning_rate": 2.09696474893335e-05, + "loss": 0.1891, + "step": 33706 + }, + { + "epoch": 0.7901451619519799, + "grad_norm": 1.4332692623138428, + "learning_rate": 2.0965134344343006e-05, + "loss": 0.0836, + "step": 33707 + }, + { + "epoch": 0.790168603526785, + "grad_norm": 3.033864974975586, + "learning_rate": 2.0960621628197918e-05, + "loss": 0.2368, + "step": 33708 + }, + { + "epoch": 0.7901920451015899, + "grad_norm": 3.2810616493225098, + "learning_rate": 2.0956109340922723e-05, + "loss": 0.3198, + "step": 33709 + }, + { + "epoch": 0.790215486676395, + "grad_norm": 2.212963104248047, + "learning_rate": 2.0951597482541896e-05, + "loss": 1.778, + "step": 33710 + }, + { + "epoch": 0.7902389282511999, + "grad_norm": 0.23676107823848724, + "learning_rate": 2.0947086053079922e-05, + "loss": 0.0136, + "step": 33711 + }, + { + "epoch": 0.790262369826005, + "grad_norm": 0.755964457988739, + "learning_rate": 2.0942575052561264e-05, + "loss": 0.0493, + "step": 33712 + }, + { + "epoch": 0.7902858114008099, + "grad_norm": 2.632741689682007, + "learning_rate": 2.0938064481010434e-05, + "loss": 0.2247, + "step": 33713 + }, + { + "epoch": 0.7903092529756149, + "grad_norm": 0.4947936534881592, + "learning_rate": 2.093355433845189e-05, + "loss": 0.0176, + "step": 33714 + }, + { + "epoch": 0.7903326945504199, + "grad_norm": 1.72298002243042, + "learning_rate": 2.092904462491011e-05, + "loss": 0.0747, + "step": 33715 + }, + { + "epoch": 0.7903561361252249, + "grad_norm": 0.7639625668525696, + "learning_rate": 2.0924535340409546e-05, + "loss": 0.0576, + "step": 33716 + }, + { + "epoch": 0.7903795777000299, + "grad_norm": 1.373630404472351, + "learning_rate": 2.0920026484974688e-05, + "loss": 0.095, + "step": 33717 + }, + { + "epoch": 0.7904030192748349, + "grad_norm": 2.9354400634765625, + "learning_rate": 2.0915518058629956e-05, + "loss": 1.3387, + "step": 33718 + }, + { + "epoch": 0.7904264608496399, + "grad_norm": 2.370513916015625, + "learning_rate": 2.091101006139985e-05, + "loss": 0.1307, + "step": 33719 + }, + { + "epoch": 0.7904499024244449, + "grad_norm": 1.3047562837600708, + "learning_rate": 2.0906502493308854e-05, + "loss": 0.1052, + "step": 33720 + }, + { + "epoch": 0.7904733439992498, + "grad_norm": 3.1705050468444824, + "learning_rate": 2.0901995354381398e-05, + "loss": 0.2056, + "step": 33721 + }, + { + "epoch": 0.7904967855740549, + "grad_norm": 1.8023813962936401, + "learning_rate": 2.0897488644641938e-05, + "loss": 0.1678, + "step": 33722 + }, + { + "epoch": 0.7905202271488598, + "grad_norm": 2.0828659534454346, + "learning_rate": 2.0892982364114933e-05, + "loss": 0.177, + "step": 33723 + }, + { + "epoch": 0.7905436687236649, + "grad_norm": 2.327436923980713, + "learning_rate": 2.088847651282483e-05, + "loss": 0.1823, + "step": 33724 + }, + { + "epoch": 0.7905671102984698, + "grad_norm": 0.9071512818336487, + "learning_rate": 2.088397109079604e-05, + "loss": 0.0562, + "step": 33725 + }, + { + "epoch": 0.7905905518732749, + "grad_norm": 0.8490937352180481, + "learning_rate": 2.0879466098053092e-05, + "loss": 0.1398, + "step": 33726 + }, + { + "epoch": 0.7906139934480798, + "grad_norm": 2.4598376750946045, + "learning_rate": 2.0874961534620375e-05, + "loss": 0.1088, + "step": 33727 + }, + { + "epoch": 0.7906374350228849, + "grad_norm": 1.6213055849075317, + "learning_rate": 2.0870457400522336e-05, + "loss": 0.0565, + "step": 33728 + }, + { + "epoch": 0.7906608765976898, + "grad_norm": 1.8406634330749512, + "learning_rate": 2.0865953695783425e-05, + "loss": 0.1748, + "step": 33729 + }, + { + "epoch": 0.7906843181724948, + "grad_norm": 5.041447162628174, + "learning_rate": 2.0861450420428053e-05, + "loss": 0.2126, + "step": 33730 + }, + { + "epoch": 0.7907077597472998, + "grad_norm": 1.9883553981781006, + "learning_rate": 2.08569475744807e-05, + "loss": 0.0998, + "step": 33731 + }, + { + "epoch": 0.7907312013221048, + "grad_norm": 0.5684309601783752, + "learning_rate": 2.085244515796575e-05, + "loss": 0.0331, + "step": 33732 + }, + { + "epoch": 0.7907546428969098, + "grad_norm": 3.908780097961426, + "learning_rate": 2.084794317090768e-05, + "loss": 0.202, + "step": 33733 + }, + { + "epoch": 0.7907780844717148, + "grad_norm": 4.459428787231445, + "learning_rate": 2.0843441613330895e-05, + "loss": 0.317, + "step": 33734 + }, + { + "epoch": 0.7908015260465198, + "grad_norm": 3.1379549503326416, + "learning_rate": 2.0838940485259826e-05, + "loss": 0.1966, + "step": 33735 + }, + { + "epoch": 0.7908249676213248, + "grad_norm": 3.048192024230957, + "learning_rate": 2.0834439786718897e-05, + "loss": 0.2098, + "step": 33736 + }, + { + "epoch": 0.7908484091961298, + "grad_norm": 2.8684213161468506, + "learning_rate": 2.082993951773249e-05, + "loss": 0.1494, + "step": 33737 + }, + { + "epoch": 0.7908718507709348, + "grad_norm": 1.5683321952819824, + "learning_rate": 2.0825439678325078e-05, + "loss": 0.1258, + "step": 33738 + }, + { + "epoch": 0.7908952923457397, + "grad_norm": 0.5966500043869019, + "learning_rate": 2.082094026852106e-05, + "loss": 0.0533, + "step": 33739 + }, + { + "epoch": 0.7909187339205448, + "grad_norm": 3.2523930072784424, + "learning_rate": 2.0816441288344857e-05, + "loss": 0.2442, + "step": 33740 + }, + { + "epoch": 0.7909421754953497, + "grad_norm": 3.2835094928741455, + "learning_rate": 2.0811942737820854e-05, + "loss": 0.28, + "step": 33741 + }, + { + "epoch": 0.7909656170701548, + "grad_norm": 3.666999101638794, + "learning_rate": 2.080744461697346e-05, + "loss": 0.311, + "step": 33742 + }, + { + "epoch": 0.7909890586449597, + "grad_norm": 2.9952304363250732, + "learning_rate": 2.0802946925827126e-05, + "loss": 0.21, + "step": 33743 + }, + { + "epoch": 0.7910125002197648, + "grad_norm": 1.502799153327942, + "learning_rate": 2.0798449664406194e-05, + "loss": 0.043, + "step": 33744 + }, + { + "epoch": 0.7910359417945697, + "grad_norm": 2.8739211559295654, + "learning_rate": 2.079395283273513e-05, + "loss": 1.3606, + "step": 33745 + }, + { + "epoch": 0.7910593833693748, + "grad_norm": 1.5094252824783325, + "learning_rate": 2.0789456430838305e-05, + "loss": 0.0806, + "step": 33746 + }, + { + "epoch": 0.7910828249441797, + "grad_norm": 0.8165386915206909, + "learning_rate": 2.078496045874012e-05, + "loss": 0.0716, + "step": 33747 + }, + { + "epoch": 0.7911062665189847, + "grad_norm": 0.5027596950531006, + "learning_rate": 2.078046491646497e-05, + "loss": 0.0446, + "step": 33748 + }, + { + "epoch": 0.7911297080937897, + "grad_norm": 1.4391632080078125, + "learning_rate": 2.0775969804037244e-05, + "loss": 0.0781, + "step": 33749 + }, + { + "epoch": 0.7911531496685947, + "grad_norm": 1.127136468887329, + "learning_rate": 2.07714751214813e-05, + "loss": 0.1158, + "step": 33750 + }, + { + "epoch": 0.7911765912433998, + "grad_norm": 1.7798665761947632, + "learning_rate": 2.0766980868821594e-05, + "loss": 0.1596, + "step": 33751 + }, + { + "epoch": 0.7912000328182047, + "grad_norm": 2.427738904953003, + "learning_rate": 2.076248704608247e-05, + "loss": 0.1757, + "step": 33752 + }, + { + "epoch": 0.7912234743930098, + "grad_norm": 2.3332290649414062, + "learning_rate": 2.075799365328832e-05, + "loss": 0.137, + "step": 33753 + }, + { + "epoch": 0.7912469159678147, + "grad_norm": 1.8073432445526123, + "learning_rate": 2.0753500690463502e-05, + "loss": 0.1521, + "step": 33754 + }, + { + "epoch": 0.7912703575426198, + "grad_norm": 4.685239315032959, + "learning_rate": 2.0749008157632444e-05, + "loss": 0.3876, + "step": 33755 + }, + { + "epoch": 0.7912937991174247, + "grad_norm": 1.0217777490615845, + "learning_rate": 2.0744516054819496e-05, + "loss": 0.0701, + "step": 33756 + }, + { + "epoch": 0.7913172406922298, + "grad_norm": 3.780304193496704, + "learning_rate": 2.0740024382049007e-05, + "loss": 0.3547, + "step": 33757 + }, + { + "epoch": 0.7913406822670347, + "grad_norm": 2.588007926940918, + "learning_rate": 2.07355331393454e-05, + "loss": 0.2059, + "step": 33758 + }, + { + "epoch": 0.7913641238418397, + "grad_norm": 2.287688970565796, + "learning_rate": 2.073104232673301e-05, + "loss": 0.1471, + "step": 33759 + }, + { + "epoch": 0.7913875654166447, + "grad_norm": 3.3960587978363037, + "learning_rate": 2.0726551944236227e-05, + "loss": 0.1818, + "step": 33760 + }, + { + "epoch": 0.7914110069914497, + "grad_norm": 3.3299522399902344, + "learning_rate": 2.0722061991879392e-05, + "loss": 0.0437, + "step": 33761 + }, + { + "epoch": 0.7914344485662547, + "grad_norm": 1.4346272945404053, + "learning_rate": 2.0717572469686884e-05, + "loss": 0.0505, + "step": 33762 + }, + { + "epoch": 0.7914578901410597, + "grad_norm": 3.6317989826202393, + "learning_rate": 2.071308337768303e-05, + "loss": 0.2735, + "step": 33763 + }, + { + "epoch": 0.7914813317158647, + "grad_norm": 3.29956316947937, + "learning_rate": 2.0708594715892237e-05, + "loss": 0.1392, + "step": 33764 + }, + { + "epoch": 0.7915047732906697, + "grad_norm": 4.235085487365723, + "learning_rate": 2.0704106484338837e-05, + "loss": 0.3288, + "step": 33765 + }, + { + "epoch": 0.7915282148654746, + "grad_norm": 1.711836338043213, + "learning_rate": 2.0699618683047162e-05, + "loss": 0.1875, + "step": 33766 + }, + { + "epoch": 0.7915516564402797, + "grad_norm": 3.1994986534118652, + "learning_rate": 2.0695131312041614e-05, + "loss": 0.2836, + "step": 33767 + }, + { + "epoch": 0.7915750980150846, + "grad_norm": 2.9300670623779297, + "learning_rate": 2.0690644371346513e-05, + "loss": 0.3821, + "step": 33768 + }, + { + "epoch": 0.7915985395898897, + "grad_norm": 1.2453484535217285, + "learning_rate": 2.06861578609862e-05, + "loss": 0.0597, + "step": 33769 + }, + { + "epoch": 0.7916219811646946, + "grad_norm": 1.5928829908370972, + "learning_rate": 2.0681671780985e-05, + "loss": 0.0953, + "step": 33770 + }, + { + "epoch": 0.7916454227394997, + "grad_norm": 1.7500470876693726, + "learning_rate": 2.067718613136731e-05, + "loss": 1.1014, + "step": 33771 + }, + { + "epoch": 0.7916688643143046, + "grad_norm": 2.60605788230896, + "learning_rate": 2.0672700912157437e-05, + "loss": 1.4724, + "step": 33772 + }, + { + "epoch": 0.7916923058891097, + "grad_norm": 2.240380048751831, + "learning_rate": 2.0668216123379724e-05, + "loss": 0.9622, + "step": 33773 + }, + { + "epoch": 0.7917157474639146, + "grad_norm": 0.19932827353477478, + "learning_rate": 2.0663731765058493e-05, + "loss": 0.009, + "step": 33774 + }, + { + "epoch": 0.7917391890387196, + "grad_norm": 4.756407260894775, + "learning_rate": 2.0659247837218054e-05, + "loss": 0.335, + "step": 33775 + }, + { + "epoch": 0.7917626306135246, + "grad_norm": 1.8354195356369019, + "learning_rate": 2.0654764339882804e-05, + "loss": 0.1385, + "step": 33776 + }, + { + "epoch": 0.7917860721883296, + "grad_norm": 2.340927839279175, + "learning_rate": 2.065028127307703e-05, + "loss": 0.208, + "step": 33777 + }, + { + "epoch": 0.7918095137631346, + "grad_norm": 1.2177107334136963, + "learning_rate": 2.064579863682503e-05, + "loss": 0.0958, + "step": 33778 + }, + { + "epoch": 0.7918329553379396, + "grad_norm": 10.979233741760254, + "learning_rate": 2.064131643115119e-05, + "loss": 0.4948, + "step": 33779 + }, + { + "epoch": 0.7918563969127446, + "grad_norm": 1.7334903478622437, + "learning_rate": 2.0636834656079806e-05, + "loss": 0.0816, + "step": 33780 + }, + { + "epoch": 0.7918798384875496, + "grad_norm": 2.701636791229248, + "learning_rate": 2.063235331163518e-05, + "loss": 0.1668, + "step": 33781 + }, + { + "epoch": 0.7919032800623546, + "grad_norm": 1.9210394620895386, + "learning_rate": 2.0627872397841607e-05, + "loss": 0.0798, + "step": 33782 + }, + { + "epoch": 0.7919267216371596, + "grad_norm": 1.4933323860168457, + "learning_rate": 2.0623391914723455e-05, + "loss": 0.1506, + "step": 33783 + }, + { + "epoch": 0.7919501632119645, + "grad_norm": 2.5102975368499756, + "learning_rate": 2.0618911862305012e-05, + "loss": 0.1078, + "step": 33784 + }, + { + "epoch": 0.7919736047867696, + "grad_norm": 0.6811136603355408, + "learning_rate": 2.0614432240610593e-05, + "loss": 0.0283, + "step": 33785 + }, + { + "epoch": 0.7919970463615745, + "grad_norm": 2.259124994277954, + "learning_rate": 2.060995304966449e-05, + "loss": 0.1489, + "step": 33786 + }, + { + "epoch": 0.7920204879363796, + "grad_norm": 5.33499813079834, + "learning_rate": 2.0605474289491013e-05, + "loss": 0.4338, + "step": 33787 + }, + { + "epoch": 0.7920439295111845, + "grad_norm": 0.30918532609939575, + "learning_rate": 2.0600995960114446e-05, + "loss": 0.0178, + "step": 33788 + }, + { + "epoch": 0.7920673710859896, + "grad_norm": 1.417362928390503, + "learning_rate": 2.0596518061559123e-05, + "loss": 0.1065, + "step": 33789 + }, + { + "epoch": 0.7920908126607945, + "grad_norm": 0.8199198246002197, + "learning_rate": 2.059204059384934e-05, + "loss": 0.0851, + "step": 33790 + }, + { + "epoch": 0.7921142542355996, + "grad_norm": 0.8858949542045593, + "learning_rate": 2.058756355700934e-05, + "loss": 0.0707, + "step": 33791 + }, + { + "epoch": 0.7921376958104045, + "grad_norm": 3.877866506576538, + "learning_rate": 2.058308695106348e-05, + "loss": 0.1887, + "step": 33792 + }, + { + "epoch": 0.7921611373852095, + "grad_norm": 2.831362009048462, + "learning_rate": 2.057861077603602e-05, + "loss": 0.2365, + "step": 33793 + }, + { + "epoch": 0.7921845789600145, + "grad_norm": 2.692678213119507, + "learning_rate": 2.0574135031951258e-05, + "loss": 0.1576, + "step": 33794 + }, + { + "epoch": 0.7922080205348195, + "grad_norm": 2.161017656326294, + "learning_rate": 2.0569659718833445e-05, + "loss": 0.181, + "step": 33795 + }, + { + "epoch": 0.7922314621096245, + "grad_norm": 3.152963161468506, + "learning_rate": 2.0565184836706918e-05, + "loss": 0.2128, + "step": 33796 + }, + { + "epoch": 0.7922549036844295, + "grad_norm": 2.177492141723633, + "learning_rate": 2.056071038559594e-05, + "loss": 0.1103, + "step": 33797 + }, + { + "epoch": 0.7922783452592345, + "grad_norm": 2.815927505493164, + "learning_rate": 2.0556236365524763e-05, + "loss": 0.1952, + "step": 33798 + }, + { + "epoch": 0.7923017868340395, + "grad_norm": 2.205972671508789, + "learning_rate": 2.055176277651769e-05, + "loss": 0.177, + "step": 33799 + }, + { + "epoch": 0.7923252284088445, + "grad_norm": 2.4174270629882812, + "learning_rate": 2.0547289618598985e-05, + "loss": 0.1238, + "step": 33800 + }, + { + "epoch": 0.7923486699836495, + "grad_norm": 1.6061112880706787, + "learning_rate": 2.0542816891792905e-05, + "loss": 0.9639, + "step": 33801 + }, + { + "epoch": 0.7923721115584546, + "grad_norm": 3.17024564743042, + "learning_rate": 2.053834459612375e-05, + "loss": 0.1308, + "step": 33802 + }, + { + "epoch": 0.7923955531332595, + "grad_norm": 2.687866687774658, + "learning_rate": 2.0533872731615743e-05, + "loss": 0.1326, + "step": 33803 + }, + { + "epoch": 0.7924189947080645, + "grad_norm": 3.51869797706604, + "learning_rate": 2.0529401298293206e-05, + "loss": 0.2372, + "step": 33804 + }, + { + "epoch": 0.7924424362828695, + "grad_norm": 1.5487115383148193, + "learning_rate": 2.0524930296180377e-05, + "loss": 0.0642, + "step": 33805 + }, + { + "epoch": 0.7924658778576745, + "grad_norm": 2.0702643394470215, + "learning_rate": 2.052045972530151e-05, + "loss": 0.1767, + "step": 33806 + }, + { + "epoch": 0.7924893194324795, + "grad_norm": 2.5590531826019287, + "learning_rate": 2.051598958568086e-05, + "loss": 0.1688, + "step": 33807 + }, + { + "epoch": 0.7925127610072845, + "grad_norm": 0.7613240480422974, + "learning_rate": 2.0511519877342656e-05, + "loss": 0.0491, + "step": 33808 + }, + { + "epoch": 0.7925362025820895, + "grad_norm": 1.811443567276001, + "learning_rate": 2.0507050600311205e-05, + "loss": 0.1106, + "step": 33809 + }, + { + "epoch": 0.7925596441568945, + "grad_norm": 2.864562511444092, + "learning_rate": 2.050258175461074e-05, + "loss": 1.5809, + "step": 33810 + }, + { + "epoch": 0.7925830857316994, + "grad_norm": 3.8417046070098877, + "learning_rate": 2.04981133402655e-05, + "loss": 0.315, + "step": 33811 + }, + { + "epoch": 0.7926065273065045, + "grad_norm": 0.8680474162101746, + "learning_rate": 2.049364535729973e-05, + "loss": 0.0608, + "step": 33812 + }, + { + "epoch": 0.7926299688813094, + "grad_norm": 0.9357786774635315, + "learning_rate": 2.048917780573766e-05, + "loss": 0.0844, + "step": 33813 + }, + { + "epoch": 0.7926534104561145, + "grad_norm": 2.629239797592163, + "learning_rate": 2.048471068560356e-05, + "loss": 0.1622, + "step": 33814 + }, + { + "epoch": 0.7926768520309194, + "grad_norm": 2.068453550338745, + "learning_rate": 2.0480243996921633e-05, + "loss": 0.1461, + "step": 33815 + }, + { + "epoch": 0.7927002936057245, + "grad_norm": 2.3554635047912598, + "learning_rate": 2.047577773971616e-05, + "loss": 0.1788, + "step": 33816 + }, + { + "epoch": 0.7927237351805294, + "grad_norm": 3.954766035079956, + "learning_rate": 2.047131191401136e-05, + "loss": 0.3511, + "step": 33817 + }, + { + "epoch": 0.7927471767553345, + "grad_norm": 0.5323067903518677, + "learning_rate": 2.0466846519831463e-05, + "loss": 0.0484, + "step": 33818 + }, + { + "epoch": 0.7927706183301394, + "grad_norm": 0.4775485098361969, + "learning_rate": 2.0462381557200682e-05, + "loss": 0.0526, + "step": 33819 + }, + { + "epoch": 0.7927940599049444, + "grad_norm": 1.7262011766433716, + "learning_rate": 2.0457917026143237e-05, + "loss": 0.8952, + "step": 33820 + }, + { + "epoch": 0.7928175014797494, + "grad_norm": 3.3687491416931152, + "learning_rate": 2.045345292668339e-05, + "loss": 0.1428, + "step": 33821 + }, + { + "epoch": 0.7928409430545544, + "grad_norm": 5.826610088348389, + "learning_rate": 2.044898925884534e-05, + "loss": 0.5219, + "step": 33822 + }, + { + "epoch": 0.7928643846293594, + "grad_norm": 4.847960472106934, + "learning_rate": 2.0444526022653322e-05, + "loss": 0.3246, + "step": 33823 + }, + { + "epoch": 0.7928878262041644, + "grad_norm": 0.39917853474617004, + "learning_rate": 2.0440063218131533e-05, + "loss": 0.0272, + "step": 33824 + }, + { + "epoch": 0.7929112677789694, + "grad_norm": 4.017943382263184, + "learning_rate": 2.0435600845304205e-05, + "loss": 0.1665, + "step": 33825 + }, + { + "epoch": 0.7929347093537744, + "grad_norm": 2.616417169570923, + "learning_rate": 2.0431138904195524e-05, + "loss": 0.211, + "step": 33826 + }, + { + "epoch": 0.7929581509285794, + "grad_norm": 3.4410600662231445, + "learning_rate": 2.0426677394829706e-05, + "loss": 0.2651, + "step": 33827 + }, + { + "epoch": 0.7929815925033844, + "grad_norm": 3.3322982788085938, + "learning_rate": 2.0422216317231013e-05, + "loss": 0.3119, + "step": 33828 + }, + { + "epoch": 0.7930050340781893, + "grad_norm": 3.0551564693450928, + "learning_rate": 2.04177556714236e-05, + "loss": 0.1369, + "step": 33829 + }, + { + "epoch": 0.7930284756529944, + "grad_norm": 1.2457062005996704, + "learning_rate": 2.0413295457431693e-05, + "loss": 0.6489, + "step": 33830 + }, + { + "epoch": 0.7930519172277993, + "grad_norm": 2.8013229370117188, + "learning_rate": 2.040883567527947e-05, + "loss": 0.3732, + "step": 33831 + }, + { + "epoch": 0.7930753588026044, + "grad_norm": 1.858317494392395, + "learning_rate": 2.0404376324991148e-05, + "loss": 0.1318, + "step": 33832 + }, + { + "epoch": 0.7930988003774093, + "grad_norm": 1.9588533639907837, + "learning_rate": 2.0399917406590896e-05, + "loss": 0.0942, + "step": 33833 + }, + { + "epoch": 0.7931222419522144, + "grad_norm": 3.1787426471710205, + "learning_rate": 2.0395458920102963e-05, + "loss": 0.1905, + "step": 33834 + }, + { + "epoch": 0.7931456835270193, + "grad_norm": 2.700216293334961, + "learning_rate": 2.039100086555149e-05, + "loss": 0.2414, + "step": 33835 + }, + { + "epoch": 0.7931691251018244, + "grad_norm": 1.5354478359222412, + "learning_rate": 2.0386543242960697e-05, + "loss": 0.1778, + "step": 33836 + }, + { + "epoch": 0.7931925666766293, + "grad_norm": 3.7556965351104736, + "learning_rate": 2.0382086052354755e-05, + "loss": 0.5329, + "step": 33837 + }, + { + "epoch": 0.7932160082514343, + "grad_norm": 2.6274526119232178, + "learning_rate": 2.0377629293757862e-05, + "loss": 0.2319, + "step": 33838 + }, + { + "epoch": 0.7932394498262393, + "grad_norm": 5.477041244506836, + "learning_rate": 2.0373172967194153e-05, + "loss": 0.3649, + "step": 33839 + }, + { + "epoch": 0.7932628914010443, + "grad_norm": 1.4035733938217163, + "learning_rate": 2.0368717072687848e-05, + "loss": 0.1307, + "step": 33840 + }, + { + "epoch": 0.7932863329758493, + "grad_norm": 1.8719407320022583, + "learning_rate": 2.0364261610263156e-05, + "loss": 1.3756, + "step": 33841 + }, + { + "epoch": 0.7933097745506543, + "grad_norm": 0.6171284914016724, + "learning_rate": 2.035980657994421e-05, + "loss": 0.0602, + "step": 33842 + }, + { + "epoch": 0.7933332161254593, + "grad_norm": 2.5724456310272217, + "learning_rate": 2.03553519817552e-05, + "loss": 0.216, + "step": 33843 + }, + { + "epoch": 0.7933566577002643, + "grad_norm": 3.5681560039520264, + "learning_rate": 2.0350897815720294e-05, + "loss": 0.3936, + "step": 33844 + }, + { + "epoch": 0.7933800992750693, + "grad_norm": 0.8610503673553467, + "learning_rate": 2.0346444081863646e-05, + "loss": 0.0759, + "step": 33845 + }, + { + "epoch": 0.7934035408498743, + "grad_norm": 2.114189624786377, + "learning_rate": 2.034199078020941e-05, + "loss": 0.1178, + "step": 33846 + }, + { + "epoch": 0.7934269824246792, + "grad_norm": 1.660941243171692, + "learning_rate": 2.03375379107818e-05, + "loss": 0.1992, + "step": 33847 + }, + { + "epoch": 0.7934504239994843, + "grad_norm": 0.9879715442657471, + "learning_rate": 2.033308547360494e-05, + "loss": 0.0963, + "step": 33848 + }, + { + "epoch": 0.7934738655742892, + "grad_norm": 2.907132148742676, + "learning_rate": 2.0328633468703006e-05, + "loss": 0.2005, + "step": 33849 + }, + { + "epoch": 0.7934973071490943, + "grad_norm": 4.508358001708984, + "learning_rate": 2.032418189610015e-05, + "loss": 0.4552, + "step": 33850 + }, + { + "epoch": 0.7935207487238992, + "grad_norm": 0.2763440012931824, + "learning_rate": 2.031973075582049e-05, + "loss": 0.0107, + "step": 33851 + }, + { + "epoch": 0.7935441902987043, + "grad_norm": 2.3513286113739014, + "learning_rate": 2.0315280047888218e-05, + "loss": 0.1963, + "step": 33852 + }, + { + "epoch": 0.7935676318735093, + "grad_norm": 5.083581924438477, + "learning_rate": 2.03108297723275e-05, + "loss": 0.3473, + "step": 33853 + }, + { + "epoch": 0.7935910734483143, + "grad_norm": 1.559072732925415, + "learning_rate": 2.0306379929162456e-05, + "loss": 0.1562, + "step": 33854 + }, + { + "epoch": 0.7936145150231193, + "grad_norm": 1.82040274143219, + "learning_rate": 2.0301930518417244e-05, + "loss": 0.1054, + "step": 33855 + }, + { + "epoch": 0.7936379565979242, + "grad_norm": 3.4903323650360107, + "learning_rate": 2.029748154011599e-05, + "loss": 0.4242, + "step": 33856 + }, + { + "epoch": 0.7936613981727293, + "grad_norm": 3.9519338607788086, + "learning_rate": 2.0293032994282847e-05, + "loss": 0.3474, + "step": 33857 + }, + { + "epoch": 0.7936848397475342, + "grad_norm": 0.6696199178695679, + "learning_rate": 2.028858488094192e-05, + "loss": 0.0348, + "step": 33858 + }, + { + "epoch": 0.7937082813223393, + "grad_norm": 1.9051893949508667, + "learning_rate": 2.0284137200117404e-05, + "loss": 0.1609, + "step": 33859 + }, + { + "epoch": 0.7937317228971442, + "grad_norm": 1.7346975803375244, + "learning_rate": 2.027968995183339e-05, + "loss": 0.1332, + "step": 33860 + }, + { + "epoch": 0.7937551644719493, + "grad_norm": 3.177567720413208, + "learning_rate": 2.027524313611403e-05, + "loss": 0.1488, + "step": 33861 + }, + { + "epoch": 0.7937786060467542, + "grad_norm": 0.918285071849823, + "learning_rate": 2.0270796752983435e-05, + "loss": 0.0781, + "step": 33862 + }, + { + "epoch": 0.7938020476215593, + "grad_norm": 1.334466814994812, + "learning_rate": 2.0266350802465715e-05, + "loss": 0.0494, + "step": 33863 + }, + { + "epoch": 0.7938254891963642, + "grad_norm": 0.5199759006500244, + "learning_rate": 2.026190528458505e-05, + "loss": 0.0397, + "step": 33864 + }, + { + "epoch": 0.7938489307711692, + "grad_norm": 3.694868564605713, + "learning_rate": 2.0257460199365495e-05, + "loss": 0.3485, + "step": 33865 + }, + { + "epoch": 0.7938723723459742, + "grad_norm": 3.137295722961426, + "learning_rate": 2.0253015546831232e-05, + "loss": 0.1662, + "step": 33866 + }, + { + "epoch": 0.7938958139207792, + "grad_norm": 1.9566971063613892, + "learning_rate": 2.0248571327006348e-05, + "loss": 0.144, + "step": 33867 + }, + { + "epoch": 0.7939192554955842, + "grad_norm": 1.8222031593322754, + "learning_rate": 2.0244127539914957e-05, + "loss": 0.1317, + "step": 33868 + }, + { + "epoch": 0.7939426970703892, + "grad_norm": 0.896073579788208, + "learning_rate": 2.0239684185581176e-05, + "loss": 0.0392, + "step": 33869 + }, + { + "epoch": 0.7939661386451942, + "grad_norm": 2.7260584831237793, + "learning_rate": 2.0235241264029115e-05, + "loss": 0.482, + "step": 33870 + }, + { + "epoch": 0.7939895802199992, + "grad_norm": 0.8285788297653198, + "learning_rate": 2.0230798775282844e-05, + "loss": 0.0566, + "step": 33871 + }, + { + "epoch": 0.7940130217948042, + "grad_norm": 2.804314136505127, + "learning_rate": 2.0226356719366525e-05, + "loss": 0.1914, + "step": 33872 + }, + { + "epoch": 0.7940364633696092, + "grad_norm": 0.9443777799606323, + "learning_rate": 2.0221915096304246e-05, + "loss": 0.0557, + "step": 33873 + }, + { + "epoch": 0.7940599049444141, + "grad_norm": 0.8724778890609741, + "learning_rate": 2.02174739061201e-05, + "loss": 0.0607, + "step": 33874 + }, + { + "epoch": 0.7940833465192192, + "grad_norm": 0.622830867767334, + "learning_rate": 2.0213033148838178e-05, + "loss": 0.0265, + "step": 33875 + }, + { + "epoch": 0.7941067880940241, + "grad_norm": 0.6647590398788452, + "learning_rate": 2.020859282448255e-05, + "loss": 0.0419, + "step": 33876 + }, + { + "epoch": 0.7941302296688292, + "grad_norm": 0.6696595549583435, + "learning_rate": 2.020415293307738e-05, + "loss": 0.0704, + "step": 33877 + }, + { + "epoch": 0.7941536712436341, + "grad_norm": 3.218026876449585, + "learning_rate": 2.0199713474646687e-05, + "loss": 0.1737, + "step": 33878 + }, + { + "epoch": 0.7941771128184392, + "grad_norm": 0.8671960234642029, + "learning_rate": 2.0195274449214618e-05, + "loss": 0.0482, + "step": 33879 + }, + { + "epoch": 0.7942005543932441, + "grad_norm": 0.419550359249115, + "learning_rate": 2.0190835856805233e-05, + "loss": 0.0242, + "step": 33880 + }, + { + "epoch": 0.7942239959680492, + "grad_norm": 0.4159657955169678, + "learning_rate": 2.0186397697442615e-05, + "loss": 0.034, + "step": 33881 + }, + { + "epoch": 0.7942474375428541, + "grad_norm": 2.3402388095855713, + "learning_rate": 2.0181959971150856e-05, + "loss": 1.7377, + "step": 33882 + }, + { + "epoch": 0.7942708791176591, + "grad_norm": 1.0461064577102661, + "learning_rate": 2.0177522677954e-05, + "loss": 0.1185, + "step": 33883 + }, + { + "epoch": 0.7942943206924641, + "grad_norm": 3.1004579067230225, + "learning_rate": 2.017308581787617e-05, + "loss": 0.1476, + "step": 33884 + }, + { + "epoch": 0.7943177622672691, + "grad_norm": 3.8013253211975098, + "learning_rate": 2.016864939094142e-05, + "loss": 0.2177, + "step": 33885 + }, + { + "epoch": 0.7943412038420741, + "grad_norm": 3.002826690673828, + "learning_rate": 2.0164213397173825e-05, + "loss": 0.329, + "step": 33886 + }, + { + "epoch": 0.7943646454168791, + "grad_norm": 1.616731882095337, + "learning_rate": 2.0159777836597447e-05, + "loss": 0.1199, + "step": 33887 + }, + { + "epoch": 0.7943880869916841, + "grad_norm": 0.9852834939956665, + "learning_rate": 2.0155342709236348e-05, + "loss": 0.1009, + "step": 33888 + }, + { + "epoch": 0.7944115285664891, + "grad_norm": 1.9169843196868896, + "learning_rate": 2.015090801511462e-05, + "loss": 0.1258, + "step": 33889 + }, + { + "epoch": 0.794434970141294, + "grad_norm": 4.794885635375977, + "learning_rate": 2.014647375425629e-05, + "loss": 0.4967, + "step": 33890 + }, + { + "epoch": 0.7944584117160991, + "grad_norm": 0.7412974834442139, + "learning_rate": 2.0142039926685454e-05, + "loss": 0.0436, + "step": 33891 + }, + { + "epoch": 0.794481853290904, + "grad_norm": 3.0839385986328125, + "learning_rate": 2.013760653242617e-05, + "loss": 0.2607, + "step": 33892 + }, + { + "epoch": 0.7945052948657091, + "grad_norm": 3.168517589569092, + "learning_rate": 2.0133173571502462e-05, + "loss": 0.1825, + "step": 33893 + }, + { + "epoch": 0.794528736440514, + "grad_norm": 1.9956836700439453, + "learning_rate": 2.0128741043938416e-05, + "loss": 0.1166, + "step": 33894 + }, + { + "epoch": 0.7945521780153191, + "grad_norm": 1.3508809804916382, + "learning_rate": 2.0124308949758052e-05, + "loss": 0.1431, + "step": 33895 + }, + { + "epoch": 0.794575619590124, + "grad_norm": 3.0402252674102783, + "learning_rate": 2.0119877288985423e-05, + "loss": 0.2884, + "step": 33896 + }, + { + "epoch": 0.7945990611649291, + "grad_norm": 1.7522000074386597, + "learning_rate": 2.0115446061644594e-05, + "loss": 0.1018, + "step": 33897 + }, + { + "epoch": 0.794622502739734, + "grad_norm": 3.7966673374176025, + "learning_rate": 2.0111015267759616e-05, + "loss": 0.3651, + "step": 33898 + }, + { + "epoch": 0.794645944314539, + "grad_norm": 2.0345723628997803, + "learning_rate": 2.0106584907354508e-05, + "loss": 0.0716, + "step": 33899 + }, + { + "epoch": 0.794669385889344, + "grad_norm": 3.724701404571533, + "learning_rate": 2.010215498045329e-05, + "loss": 0.4517, + "step": 33900 + }, + { + "epoch": 0.794692827464149, + "grad_norm": 2.6936542987823486, + "learning_rate": 2.0097725487080056e-05, + "loss": 0.1859, + "step": 33901 + }, + { + "epoch": 0.794716269038954, + "grad_norm": 2.396933078765869, + "learning_rate": 2.0093296427258813e-05, + "loss": 0.1882, + "step": 33902 + }, + { + "epoch": 0.794739710613759, + "grad_norm": 0.6506970524787903, + "learning_rate": 2.0088867801013556e-05, + "loss": 0.0428, + "step": 33903 + }, + { + "epoch": 0.7947631521885641, + "grad_norm": 1.3455730676651, + "learning_rate": 2.0084439608368377e-05, + "loss": 0.0913, + "step": 33904 + }, + { + "epoch": 0.794786593763369, + "grad_norm": 2.0190725326538086, + "learning_rate": 2.008001184934728e-05, + "loss": 1.14, + "step": 33905 + }, + { + "epoch": 0.7948100353381741, + "grad_norm": 2.464442253112793, + "learning_rate": 2.0075584523974287e-05, + "loss": 0.2342, + "step": 33906 + }, + { + "epoch": 0.794833476912979, + "grad_norm": 4.018328666687012, + "learning_rate": 2.007115763227342e-05, + "loss": 0.2489, + "step": 33907 + }, + { + "epoch": 0.7948569184877841, + "grad_norm": 3.452650308609009, + "learning_rate": 2.0066731174268694e-05, + "loss": 0.217, + "step": 33908 + }, + { + "epoch": 0.794880360062589, + "grad_norm": 4.204708099365234, + "learning_rate": 2.0062305149984107e-05, + "loss": 0.3039, + "step": 33909 + }, + { + "epoch": 0.794903801637394, + "grad_norm": 1.1996345520019531, + "learning_rate": 2.0057879559443726e-05, + "loss": 0.1181, + "step": 33910 + }, + { + "epoch": 0.794927243212199, + "grad_norm": 1.0269701480865479, + "learning_rate": 2.0053454402671533e-05, + "loss": 0.063, + "step": 33911 + }, + { + "epoch": 0.794950684787004, + "grad_norm": 1.9151829481124878, + "learning_rate": 2.0049029679691522e-05, + "loss": 0.2467, + "step": 33912 + }, + { + "epoch": 0.794974126361809, + "grad_norm": 3.0281598567962646, + "learning_rate": 2.004460539052776e-05, + "loss": 0.1876, + "step": 33913 + }, + { + "epoch": 0.794997567936614, + "grad_norm": 2.1516637802124023, + "learning_rate": 2.0040181535204205e-05, + "loss": 0.1014, + "step": 33914 + }, + { + "epoch": 0.795021009511419, + "grad_norm": 3.4944705963134766, + "learning_rate": 2.0035758113744874e-05, + "loss": 0.1116, + "step": 33915 + }, + { + "epoch": 0.795044451086224, + "grad_norm": 3.2363927364349365, + "learning_rate": 2.0031335126173745e-05, + "loss": 0.3651, + "step": 33916 + }, + { + "epoch": 0.795067892661029, + "grad_norm": 2.4939632415771484, + "learning_rate": 2.0026912572514867e-05, + "loss": 0.1686, + "step": 33917 + }, + { + "epoch": 0.795091334235834, + "grad_norm": 4.437617301940918, + "learning_rate": 2.002249045279221e-05, + "loss": 0.4526, + "step": 33918 + }, + { + "epoch": 0.7951147758106389, + "grad_norm": 1.7814074754714966, + "learning_rate": 2.0018068767029774e-05, + "loss": 0.2104, + "step": 33919 + }, + { + "epoch": 0.795138217385444, + "grad_norm": 3.4254894256591797, + "learning_rate": 2.0013647515251544e-05, + "loss": 0.2396, + "step": 33920 + }, + { + "epoch": 0.7951616589602489, + "grad_norm": 1.976853370666504, + "learning_rate": 2.0009226697481477e-05, + "loss": 0.1447, + "step": 33921 + }, + { + "epoch": 0.795185100535054, + "grad_norm": 2.7496891021728516, + "learning_rate": 2.0004806313743628e-05, + "loss": 0.3095, + "step": 33922 + }, + { + "epoch": 0.7952085421098589, + "grad_norm": 1.1053904294967651, + "learning_rate": 2.000038636406194e-05, + "loss": 0.0666, + "step": 33923 + }, + { + "epoch": 0.795231983684664, + "grad_norm": 2.370772361755371, + "learning_rate": 1.999596684846039e-05, + "loss": 0.2045, + "step": 33924 + }, + { + "epoch": 0.7952554252594689, + "grad_norm": 0.47186943888664246, + "learning_rate": 1.9991547766962992e-05, + "loss": 0.0371, + "step": 33925 + }, + { + "epoch": 0.795278866834274, + "grad_norm": 3.443418025970459, + "learning_rate": 1.9987129119593706e-05, + "loss": 0.3093, + "step": 33926 + }, + { + "epoch": 0.7953023084090789, + "grad_norm": 1.6604112386703491, + "learning_rate": 1.998271090637651e-05, + "loss": 0.087, + "step": 33927 + }, + { + "epoch": 0.795325749983884, + "grad_norm": 2.137096643447876, + "learning_rate": 1.9978293127335356e-05, + "loss": 1.3882, + "step": 33928 + }, + { + "epoch": 0.7953491915586889, + "grad_norm": 0.6472088098526001, + "learning_rate": 1.9973875782494244e-05, + "loss": 0.054, + "step": 33929 + }, + { + "epoch": 0.7953726331334939, + "grad_norm": 1.9626941680908203, + "learning_rate": 1.9969458871877146e-05, + "loss": 0.2582, + "step": 33930 + }, + { + "epoch": 0.7953960747082989, + "grad_norm": 1.9433985948562622, + "learning_rate": 1.9965042395508005e-05, + "loss": 0.1687, + "step": 33931 + }, + { + "epoch": 0.7954195162831039, + "grad_norm": 2.561973810195923, + "learning_rate": 1.9960626353410804e-05, + "loss": 0.1943, + "step": 33932 + }, + { + "epoch": 0.7954429578579089, + "grad_norm": 2.8137218952178955, + "learning_rate": 1.9956210745609484e-05, + "loss": 0.224, + "step": 33933 + }, + { + "epoch": 0.7954663994327139, + "grad_norm": 0.8310822248458862, + "learning_rate": 1.9951795572127995e-05, + "loss": 0.1242, + "step": 33934 + }, + { + "epoch": 0.7954898410075188, + "grad_norm": 2.974868059158325, + "learning_rate": 1.9947380832990337e-05, + "loss": 0.1784, + "step": 33935 + }, + { + "epoch": 0.7955132825823239, + "grad_norm": 4.053675651550293, + "learning_rate": 1.994296652822042e-05, + "loss": 0.2466, + "step": 33936 + }, + { + "epoch": 0.7955367241571288, + "grad_norm": 2.6065142154693604, + "learning_rate": 1.9938552657842234e-05, + "loss": 0.2017, + "step": 33937 + }, + { + "epoch": 0.7955601657319339, + "grad_norm": 4.777054786682129, + "learning_rate": 1.993413922187972e-05, + "loss": 0.286, + "step": 33938 + }, + { + "epoch": 0.7955836073067388, + "grad_norm": 2.2626149654388428, + "learning_rate": 1.9929726220356825e-05, + "loss": 0.2095, + "step": 33939 + }, + { + "epoch": 0.7956070488815439, + "grad_norm": 3.2913570404052734, + "learning_rate": 1.9925313653297474e-05, + "loss": 0.3026, + "step": 33940 + }, + { + "epoch": 0.7956304904563488, + "grad_norm": 6.523021221160889, + "learning_rate": 1.99209015207256e-05, + "loss": 0.6596, + "step": 33941 + }, + { + "epoch": 0.7956539320311539, + "grad_norm": 1.0795856714248657, + "learning_rate": 1.9916489822665197e-05, + "loss": 0.0916, + "step": 33942 + }, + { + "epoch": 0.7956773736059588, + "grad_norm": 0.5888175964355469, + "learning_rate": 1.991207855914017e-05, + "loss": 0.0504, + "step": 33943 + }, + { + "epoch": 0.7957008151807639, + "grad_norm": 3.98610782623291, + "learning_rate": 1.990766773017445e-05, + "loss": 0.3878, + "step": 33944 + }, + { + "epoch": 0.7957242567555688, + "grad_norm": 1.5262227058410645, + "learning_rate": 1.990325733579198e-05, + "loss": 1.376, + "step": 33945 + }, + { + "epoch": 0.7957476983303738, + "grad_norm": 1.8974539041519165, + "learning_rate": 1.9898847376016694e-05, + "loss": 1.438, + "step": 33946 + }, + { + "epoch": 0.7957711399051788, + "grad_norm": 13.066818237304688, + "learning_rate": 1.989443785087248e-05, + "loss": 0.2867, + "step": 33947 + }, + { + "epoch": 0.7957945814799838, + "grad_norm": 1.7850663661956787, + "learning_rate": 1.9890028760383338e-05, + "loss": 0.1636, + "step": 33948 + }, + { + "epoch": 0.7958180230547888, + "grad_norm": 0.6867619156837463, + "learning_rate": 1.9885620104573112e-05, + "loss": 0.0585, + "step": 33949 + }, + { + "epoch": 0.7958414646295938, + "grad_norm": 1.5152485370635986, + "learning_rate": 1.98812118834658e-05, + "loss": 0.1615, + "step": 33950 + }, + { + "epoch": 0.7958649062043988, + "grad_norm": 0.48366960883140564, + "learning_rate": 1.9876804097085277e-05, + "loss": 0.0299, + "step": 33951 + }, + { + "epoch": 0.7958883477792038, + "grad_norm": 2.4152987003326416, + "learning_rate": 1.9872396745455467e-05, + "loss": 0.1974, + "step": 33952 + }, + { + "epoch": 0.7959117893540087, + "grad_norm": 2.564042568206787, + "learning_rate": 1.9867989828600297e-05, + "loss": 1.423, + "step": 33953 + }, + { + "epoch": 0.7959352309288138, + "grad_norm": 2.082939624786377, + "learning_rate": 1.9863583346543636e-05, + "loss": 0.1986, + "step": 33954 + }, + { + "epoch": 0.7959586725036188, + "grad_norm": 2.2164201736450195, + "learning_rate": 1.985917729930944e-05, + "loss": 0.0502, + "step": 33955 + }, + { + "epoch": 0.7959821140784238, + "grad_norm": 1.8984601497650146, + "learning_rate": 1.9854771686921615e-05, + "loss": 0.1222, + "step": 33956 + }, + { + "epoch": 0.7960055556532288, + "grad_norm": 2.828437328338623, + "learning_rate": 1.9850366509404052e-05, + "loss": 0.2345, + "step": 33957 + }, + { + "epoch": 0.7960289972280338, + "grad_norm": 2.054619312286377, + "learning_rate": 1.984596176678064e-05, + "loss": 0.0869, + "step": 33958 + }, + { + "epoch": 0.7960524388028388, + "grad_norm": 0.7943907976150513, + "learning_rate": 1.9841557459075278e-05, + "loss": 0.0577, + "step": 33959 + }, + { + "epoch": 0.7960758803776438, + "grad_norm": 1.31416654586792, + "learning_rate": 1.983715358631191e-05, + "loss": 0.0972, + "step": 33960 + }, + { + "epoch": 0.7960993219524488, + "grad_norm": 3.147050619125366, + "learning_rate": 1.9832750148514367e-05, + "loss": 0.2859, + "step": 33961 + }, + { + "epoch": 0.7961227635272538, + "grad_norm": 2.0421218872070312, + "learning_rate": 1.9828347145706594e-05, + "loss": 0.1218, + "step": 33962 + }, + { + "epoch": 0.7961462051020588, + "grad_norm": 2.6211817264556885, + "learning_rate": 1.982394457791248e-05, + "loss": 0.2062, + "step": 33963 + }, + { + "epoch": 0.7961696466768637, + "grad_norm": 4.97735595703125, + "learning_rate": 1.9819542445155882e-05, + "loss": 0.2436, + "step": 33964 + }, + { + "epoch": 0.7961930882516688, + "grad_norm": 2.1807632446289062, + "learning_rate": 1.9815140747460713e-05, + "loss": 0.1809, + "step": 33965 + }, + { + "epoch": 0.7962165298264737, + "grad_norm": 1.2124850749969482, + "learning_rate": 1.9810739484850817e-05, + "loss": 0.0782, + "step": 33966 + }, + { + "epoch": 0.7962399714012788, + "grad_norm": 2.6199653148651123, + "learning_rate": 1.980633865735012e-05, + "loss": 0.1122, + "step": 33967 + }, + { + "epoch": 0.7962634129760837, + "grad_norm": 1.4711267948150635, + "learning_rate": 1.98019382649825e-05, + "loss": 0.0991, + "step": 33968 + }, + { + "epoch": 0.7962868545508888, + "grad_norm": 1.8600503206253052, + "learning_rate": 1.9797538307771802e-05, + "loss": 0.1578, + "step": 33969 + }, + { + "epoch": 0.7963102961256937, + "grad_norm": 0.41391345858573914, + "learning_rate": 1.9793138785741927e-05, + "loss": 0.0233, + "step": 33970 + }, + { + "epoch": 0.7963337377004988, + "grad_norm": 1.6098142862319946, + "learning_rate": 1.9788739698916737e-05, + "loss": 0.1771, + "step": 33971 + }, + { + "epoch": 0.7963571792753037, + "grad_norm": 3.8139126300811768, + "learning_rate": 1.9784341047320067e-05, + "loss": 0.2486, + "step": 33972 + }, + { + "epoch": 0.7963806208501087, + "grad_norm": 7.070291042327881, + "learning_rate": 1.9779942830975817e-05, + "loss": 0.3213, + "step": 33973 + }, + { + "epoch": 0.7964040624249137, + "grad_norm": 3.161134719848633, + "learning_rate": 1.9775545049907883e-05, + "loss": 0.7936, + "step": 33974 + }, + { + "epoch": 0.7964275039997187, + "grad_norm": 3.01906418800354, + "learning_rate": 1.97711477041401e-05, + "loss": 1.5338, + "step": 33975 + }, + { + "epoch": 0.7964509455745237, + "grad_norm": 2.898836851119995, + "learning_rate": 1.976675079369632e-05, + "loss": 0.2608, + "step": 33976 + }, + { + "epoch": 0.7964743871493287, + "grad_norm": 2.2709381580352783, + "learning_rate": 1.9762354318600408e-05, + "loss": 0.1858, + "step": 33977 + }, + { + "epoch": 0.7964978287241337, + "grad_norm": 1.8091756105422974, + "learning_rate": 1.975795827887622e-05, + "loss": 0.1336, + "step": 33978 + }, + { + "epoch": 0.7965212702989387, + "grad_norm": 2.139610528945923, + "learning_rate": 1.9753562674547575e-05, + "loss": 1.0971, + "step": 33979 + }, + { + "epoch": 0.7965447118737436, + "grad_norm": 0.7597137093544006, + "learning_rate": 1.974916750563839e-05, + "loss": 0.0683, + "step": 33980 + }, + { + "epoch": 0.7965681534485487, + "grad_norm": 2.1878325939178467, + "learning_rate": 1.974477277217247e-05, + "loss": 0.1672, + "step": 33981 + }, + { + "epoch": 0.7965915950233536, + "grad_norm": 1.2385975122451782, + "learning_rate": 1.974037847417368e-05, + "loss": 0.0467, + "step": 33982 + }, + { + "epoch": 0.7966150365981587, + "grad_norm": 1.6130651235580444, + "learning_rate": 1.9735984611665847e-05, + "loss": 0.1073, + "step": 33983 + }, + { + "epoch": 0.7966384781729636, + "grad_norm": 1.268908143043518, + "learning_rate": 1.9731591184672815e-05, + "loss": 0.1011, + "step": 33984 + }, + { + "epoch": 0.7966619197477687, + "grad_norm": 0.46028196811676025, + "learning_rate": 1.9727198193218398e-05, + "loss": 0.0257, + "step": 33985 + }, + { + "epoch": 0.7966853613225736, + "grad_norm": 2.482576370239258, + "learning_rate": 1.972280563732647e-05, + "loss": 0.2416, + "step": 33986 + }, + { + "epoch": 0.7967088028973787, + "grad_norm": 2.656425714492798, + "learning_rate": 1.9718413517020884e-05, + "loss": 1.7579, + "step": 33987 + }, + { + "epoch": 0.7967322444721836, + "grad_norm": 4.373317718505859, + "learning_rate": 1.9714021832325435e-05, + "loss": 0.2678, + "step": 33988 + }, + { + "epoch": 0.7967556860469887, + "grad_norm": 3.967738628387451, + "learning_rate": 1.970963058326396e-05, + "loss": 0.2307, + "step": 33989 + }, + { + "epoch": 0.7967791276217936, + "grad_norm": 2.2490217685699463, + "learning_rate": 1.9705239769860296e-05, + "loss": 0.1361, + "step": 33990 + }, + { + "epoch": 0.7968025691965986, + "grad_norm": 1.8744325637817383, + "learning_rate": 1.9700849392138243e-05, + "loss": 0.145, + "step": 33991 + }, + { + "epoch": 0.7968260107714036, + "grad_norm": 3.268864393234253, + "learning_rate": 1.969645945012163e-05, + "loss": 0.2712, + "step": 33992 + }, + { + "epoch": 0.7968494523462086, + "grad_norm": 2.0684993267059326, + "learning_rate": 1.96920699438343e-05, + "loss": 0.1245, + "step": 33993 + }, + { + "epoch": 0.7968728939210136, + "grad_norm": 3.469433069229126, + "learning_rate": 1.9687680873300062e-05, + "loss": 0.239, + "step": 33994 + }, + { + "epoch": 0.7968963354958186, + "grad_norm": 4.1779584884643555, + "learning_rate": 1.9683292238542726e-05, + "loss": 0.5113, + "step": 33995 + }, + { + "epoch": 0.7969197770706236, + "grad_norm": 2.179654359817505, + "learning_rate": 1.9678904039586098e-05, + "loss": 0.1745, + "step": 33996 + }, + { + "epoch": 0.7969432186454286, + "grad_norm": 3.7057480812072754, + "learning_rate": 1.967451627645398e-05, + "loss": 0.3501, + "step": 33997 + }, + { + "epoch": 0.7969666602202335, + "grad_norm": 3.572848081588745, + "learning_rate": 1.967012894917022e-05, + "loss": 0.2696, + "step": 33998 + }, + { + "epoch": 0.7969901017950386, + "grad_norm": 2.3189895153045654, + "learning_rate": 1.9665742057758562e-05, + "loss": 0.1907, + "step": 33999 + }, + { + "epoch": 0.7970135433698435, + "grad_norm": 0.13440150022506714, + "learning_rate": 1.966135560224288e-05, + "loss": 0.0045, + "step": 34000 + }, + { + "epoch": 0.7970369849446486, + "grad_norm": 0.9170103669166565, + "learning_rate": 1.9656969582646944e-05, + "loss": 0.0463, + "step": 34001 + }, + { + "epoch": 0.7970604265194535, + "grad_norm": 0.855383038520813, + "learning_rate": 1.9652583998994543e-05, + "loss": 0.0462, + "step": 34002 + }, + { + "epoch": 0.7970838680942586, + "grad_norm": 3.506340265274048, + "learning_rate": 1.9648198851309485e-05, + "loss": 0.145, + "step": 34003 + }, + { + "epoch": 0.7971073096690635, + "grad_norm": 0.35781773924827576, + "learning_rate": 1.9643814139615534e-05, + "loss": 0.0228, + "step": 34004 + }, + { + "epoch": 0.7971307512438686, + "grad_norm": 2.136641263961792, + "learning_rate": 1.9639429863936533e-05, + "loss": 0.0879, + "step": 34005 + }, + { + "epoch": 0.7971541928186736, + "grad_norm": 4.852766513824463, + "learning_rate": 1.9635046024296246e-05, + "loss": 0.3488, + "step": 34006 + }, + { + "epoch": 0.7971776343934786, + "grad_norm": 3.528773307800293, + "learning_rate": 1.9630662620718454e-05, + "loss": 0.3319, + "step": 34007 + }, + { + "epoch": 0.7972010759682836, + "grad_norm": 2.084921360015869, + "learning_rate": 1.9626279653226952e-05, + "loss": 1.7475, + "step": 34008 + }, + { + "epoch": 0.7972245175430885, + "grad_norm": 2.810396909713745, + "learning_rate": 1.962189712184549e-05, + "loss": 0.1352, + "step": 34009 + }, + { + "epoch": 0.7972479591178936, + "grad_norm": 1.5837594270706177, + "learning_rate": 1.961751502659789e-05, + "loss": 0.0934, + "step": 34010 + }, + { + "epoch": 0.7972714006926985, + "grad_norm": 3.6959269046783447, + "learning_rate": 1.9613133367507898e-05, + "loss": 0.2559, + "step": 34011 + }, + { + "epoch": 0.7972948422675036, + "grad_norm": 2.745530843734741, + "learning_rate": 1.960875214459933e-05, + "loss": 0.1264, + "step": 34012 + }, + { + "epoch": 0.7973182838423085, + "grad_norm": 4.914734840393066, + "learning_rate": 1.9604371357895922e-05, + "loss": 0.361, + "step": 34013 + }, + { + "epoch": 0.7973417254171136, + "grad_norm": 0.9695359468460083, + "learning_rate": 1.959999100742147e-05, + "loss": 0.1104, + "step": 34014 + }, + { + "epoch": 0.7973651669919185, + "grad_norm": 2.357553243637085, + "learning_rate": 1.959561109319973e-05, + "loss": 0.1731, + "step": 34015 + }, + { + "epoch": 0.7973886085667236, + "grad_norm": 3.1940455436706543, + "learning_rate": 1.9591231615254448e-05, + "loss": 0.3288, + "step": 34016 + }, + { + "epoch": 0.7974120501415285, + "grad_norm": 0.8007090091705322, + "learning_rate": 1.9586852573609393e-05, + "loss": 0.0366, + "step": 34017 + }, + { + "epoch": 0.7974354917163335, + "grad_norm": 1.7630609273910522, + "learning_rate": 1.958247396828835e-05, + "loss": 0.175, + "step": 34018 + }, + { + "epoch": 0.7974589332911385, + "grad_norm": 2.1776342391967773, + "learning_rate": 1.9578095799315076e-05, + "loss": 0.1663, + "step": 34019 + }, + { + "epoch": 0.7974823748659435, + "grad_norm": 2.5723698139190674, + "learning_rate": 1.957371806671331e-05, + "loss": 0.2516, + "step": 34020 + }, + { + "epoch": 0.7975058164407485, + "grad_norm": 1.7141613960266113, + "learning_rate": 1.9569340770506782e-05, + "loss": 1.5061, + "step": 34021 + }, + { + "epoch": 0.7975292580155535, + "grad_norm": 2.6284356117248535, + "learning_rate": 1.9564963910719303e-05, + "loss": 0.1945, + "step": 34022 + }, + { + "epoch": 0.7975526995903585, + "grad_norm": 0.8188660740852356, + "learning_rate": 1.9560587487374594e-05, + "loss": 0.0983, + "step": 34023 + }, + { + "epoch": 0.7975761411651635, + "grad_norm": 3.597055435180664, + "learning_rate": 1.9556211500496367e-05, + "loss": 0.197, + "step": 34024 + }, + { + "epoch": 0.7975995827399684, + "grad_norm": 2.2793288230895996, + "learning_rate": 1.955183595010843e-05, + "loss": 0.0837, + "step": 34025 + }, + { + "epoch": 0.7976230243147735, + "grad_norm": 0.29623937606811523, + "learning_rate": 1.95474608362345e-05, + "loss": 0.016, + "step": 34026 + }, + { + "epoch": 0.7976464658895784, + "grad_norm": 4.262115478515625, + "learning_rate": 1.9543086158898293e-05, + "loss": 0.2682, + "step": 34027 + }, + { + "epoch": 0.7976699074643835, + "grad_norm": 3.516418695449829, + "learning_rate": 1.9538711918123575e-05, + "loss": 0.2648, + "step": 34028 + }, + { + "epoch": 0.7976933490391884, + "grad_norm": 2.6070053577423096, + "learning_rate": 1.9534338113934057e-05, + "loss": 1.4464, + "step": 34029 + }, + { + "epoch": 0.7977167906139935, + "grad_norm": 2.653033971786499, + "learning_rate": 1.952996474635347e-05, + "loss": 0.118, + "step": 34030 + }, + { + "epoch": 0.7977402321887984, + "grad_norm": 4.404234409332275, + "learning_rate": 1.9525591815405574e-05, + "loss": 0.6524, + "step": 34031 + }, + { + "epoch": 0.7977636737636035, + "grad_norm": 3.5041675567626953, + "learning_rate": 1.952121932111408e-05, + "loss": 0.1634, + "step": 34032 + }, + { + "epoch": 0.7977871153384084, + "grad_norm": 0.38974806666374207, + "learning_rate": 1.9516847263502715e-05, + "loss": 0.0152, + "step": 34033 + }, + { + "epoch": 0.7978105569132135, + "grad_norm": 3.6110730171203613, + "learning_rate": 1.9512475642595174e-05, + "loss": 0.2349, + "step": 34034 + }, + { + "epoch": 0.7978339984880184, + "grad_norm": 2.8578293323516846, + "learning_rate": 1.9508104458415232e-05, + "loss": 0.1513, + "step": 34035 + }, + { + "epoch": 0.7978574400628234, + "grad_norm": 2.4719176292419434, + "learning_rate": 1.9503733710986572e-05, + "loss": 0.2749, + "step": 34036 + }, + { + "epoch": 0.7978808816376284, + "grad_norm": 3.6058921813964844, + "learning_rate": 1.9499363400332894e-05, + "loss": 0.1814, + "step": 34037 + }, + { + "epoch": 0.7979043232124334, + "grad_norm": 0.7135379910469055, + "learning_rate": 1.9494993526477956e-05, + "loss": 0.0242, + "step": 34038 + }, + { + "epoch": 0.7979277647872384, + "grad_norm": 1.939075231552124, + "learning_rate": 1.9490624089445454e-05, + "loss": 0.1093, + "step": 34039 + }, + { + "epoch": 0.7979512063620434, + "grad_norm": 2.8024635314941406, + "learning_rate": 1.9486255089259085e-05, + "loss": 0.2279, + "step": 34040 + }, + { + "epoch": 0.7979746479368484, + "grad_norm": 1.7858244180679321, + "learning_rate": 1.948188652594256e-05, + "loss": 0.0792, + "step": 34041 + }, + { + "epoch": 0.7979980895116534, + "grad_norm": 2.4638161659240723, + "learning_rate": 1.9477518399519556e-05, + "loss": 0.2012, + "step": 34042 + }, + { + "epoch": 0.7980215310864583, + "grad_norm": 2.3869481086730957, + "learning_rate": 1.9473150710013833e-05, + "loss": 0.1133, + "step": 34043 + }, + { + "epoch": 0.7980449726612634, + "grad_norm": 4.3552327156066895, + "learning_rate": 1.9468783457449046e-05, + "loss": 0.2781, + "step": 34044 + }, + { + "epoch": 0.7980684142360683, + "grad_norm": 1.9898576736450195, + "learning_rate": 1.9464416641848916e-05, + "loss": 1.3859, + "step": 34045 + }, + { + "epoch": 0.7980918558108734, + "grad_norm": 4.862150192260742, + "learning_rate": 1.9460050263237095e-05, + "loss": 0.4743, + "step": 34046 + }, + { + "epoch": 0.7981152973856783, + "grad_norm": 1.3503165245056152, + "learning_rate": 1.9455684321637337e-05, + "loss": 0.0921, + "step": 34047 + }, + { + "epoch": 0.7981387389604834, + "grad_norm": 4.4425225257873535, + "learning_rate": 1.9451318817073294e-05, + "loss": 0.3985, + "step": 34048 + }, + { + "epoch": 0.7981621805352883, + "grad_norm": 1.9752192497253418, + "learning_rate": 1.944695374956863e-05, + "loss": 0.9631, + "step": 34049 + }, + { + "epoch": 0.7981856221100934, + "grad_norm": 6.4378180503845215, + "learning_rate": 1.9442589119147092e-05, + "loss": 0.7406, + "step": 34050 + }, + { + "epoch": 0.7982090636848983, + "grad_norm": 1.0177971124649048, + "learning_rate": 1.9438224925832326e-05, + "loss": 0.0691, + "step": 34051 + }, + { + "epoch": 0.7982325052597034, + "grad_norm": 0.41502806544303894, + "learning_rate": 1.9433861169648028e-05, + "loss": 0.0252, + "step": 34052 + }, + { + "epoch": 0.7982559468345083, + "grad_norm": 2.5137760639190674, + "learning_rate": 1.942949785061785e-05, + "loss": 0.0799, + "step": 34053 + }, + { + "epoch": 0.7982793884093133, + "grad_norm": 2.6822187900543213, + "learning_rate": 1.942513496876549e-05, + "loss": 0.2014, + "step": 34054 + }, + { + "epoch": 0.7983028299841183, + "grad_norm": 2.4961345195770264, + "learning_rate": 1.9420772524114595e-05, + "loss": 0.1478, + "step": 34055 + }, + { + "epoch": 0.7983262715589233, + "grad_norm": 3.7681291103363037, + "learning_rate": 1.9416410516688865e-05, + "loss": 0.3303, + "step": 34056 + }, + { + "epoch": 0.7983497131337284, + "grad_norm": 1.7388209104537964, + "learning_rate": 1.9412048946511964e-05, + "loss": 0.142, + "step": 34057 + }, + { + "epoch": 0.7983731547085333, + "grad_norm": 4.408529758453369, + "learning_rate": 1.9407687813607535e-05, + "loss": 0.2137, + "step": 34058 + }, + { + "epoch": 0.7983965962833384, + "grad_norm": 1.3285326957702637, + "learning_rate": 1.940332711799927e-05, + "loss": 0.0916, + "step": 34059 + }, + { + "epoch": 0.7984200378581433, + "grad_norm": 1.7721383571624756, + "learning_rate": 1.939896685971082e-05, + "loss": 0.0874, + "step": 34060 + }, + { + "epoch": 0.7984434794329484, + "grad_norm": 0.5460383296012878, + "learning_rate": 1.9394607038765843e-05, + "loss": 0.0449, + "step": 34061 + }, + { + "epoch": 0.7984669210077533, + "grad_norm": 3.6593592166900635, + "learning_rate": 1.9390247655187975e-05, + "loss": 0.3495, + "step": 34062 + }, + { + "epoch": 0.7984903625825583, + "grad_norm": 1.8623361587524414, + "learning_rate": 1.938588870900091e-05, + "loss": 0.1429, + "step": 34063 + }, + { + "epoch": 0.7985138041573633, + "grad_norm": 2.6190521717071533, + "learning_rate": 1.938153020022828e-05, + "loss": 0.2683, + "step": 34064 + }, + { + "epoch": 0.7985372457321683, + "grad_norm": 3.673074960708618, + "learning_rate": 1.937717212889374e-05, + "loss": 0.4624, + "step": 34065 + }, + { + "epoch": 0.7985606873069733, + "grad_norm": 4.094229221343994, + "learning_rate": 1.937281449502092e-05, + "loss": 0.3372, + "step": 34066 + }, + { + "epoch": 0.7985841288817783, + "grad_norm": 0.5034448504447937, + "learning_rate": 1.936845729863348e-05, + "loss": 0.0475, + "step": 34067 + }, + { + "epoch": 0.7986075704565833, + "grad_norm": 1.2420858144760132, + "learning_rate": 1.9364100539755035e-05, + "loss": 0.0934, + "step": 34068 + }, + { + "epoch": 0.7986310120313883, + "grad_norm": 2.0115668773651123, + "learning_rate": 1.935974421840927e-05, + "loss": 0.1327, + "step": 34069 + }, + { + "epoch": 0.7986544536061932, + "grad_norm": 0.8602313995361328, + "learning_rate": 1.9355388334619785e-05, + "loss": 0.0546, + "step": 34070 + }, + { + "epoch": 0.7986778951809983, + "grad_norm": 3.5323646068573, + "learning_rate": 1.9351032888410238e-05, + "loss": 0.2944, + "step": 34071 + }, + { + "epoch": 0.7987013367558032, + "grad_norm": 1.7011504173278809, + "learning_rate": 1.9346677879804265e-05, + "loss": 0.1612, + "step": 34072 + }, + { + "epoch": 0.7987247783306083, + "grad_norm": 2.2357048988342285, + "learning_rate": 1.9342323308825484e-05, + "loss": 0.9745, + "step": 34073 + }, + { + "epoch": 0.7987482199054132, + "grad_norm": 1.9105781316757202, + "learning_rate": 1.9337969175497528e-05, + "loss": 0.0797, + "step": 34074 + }, + { + "epoch": 0.7987716614802183, + "grad_norm": 1.5024986267089844, + "learning_rate": 1.933361547984399e-05, + "loss": 1.1894, + "step": 34075 + }, + { + "epoch": 0.7987951030550232, + "grad_norm": 1.9489641189575195, + "learning_rate": 1.9329262221888545e-05, + "loss": 0.0901, + "step": 34076 + }, + { + "epoch": 0.7988185446298283, + "grad_norm": 2.001110553741455, + "learning_rate": 1.93249094016548e-05, + "loss": 0.1434, + "step": 34077 + }, + { + "epoch": 0.7988419862046332, + "grad_norm": 2.4750471115112305, + "learning_rate": 1.9320557019166362e-05, + "loss": 0.2563, + "step": 34078 + }, + { + "epoch": 0.7988654277794383, + "grad_norm": 2.998136043548584, + "learning_rate": 1.9316205074446847e-05, + "loss": 0.334, + "step": 34079 + }, + { + "epoch": 0.7988888693542432, + "grad_norm": 1.3592935800552368, + "learning_rate": 1.9311853567519846e-05, + "loss": 0.1247, + "step": 34080 + }, + { + "epoch": 0.7989123109290482, + "grad_norm": 3.0385842323303223, + "learning_rate": 1.9307502498409026e-05, + "loss": 0.2272, + "step": 34081 + }, + { + "epoch": 0.7989357525038532, + "grad_norm": 2.47050142288208, + "learning_rate": 1.9303151867137936e-05, + "loss": 0.0897, + "step": 34082 + }, + { + "epoch": 0.7989591940786582, + "grad_norm": 1.3832345008850098, + "learning_rate": 1.929880167373024e-05, + "loss": 0.1142, + "step": 34083 + }, + { + "epoch": 0.7989826356534632, + "grad_norm": 2.6377432346343994, + "learning_rate": 1.929445191820952e-05, + "loss": 0.0943, + "step": 34084 + }, + { + "epoch": 0.7990060772282682, + "grad_norm": 1.4516253471374512, + "learning_rate": 1.929010260059937e-05, + "loss": 0.0648, + "step": 34085 + }, + { + "epoch": 0.7990295188030732, + "grad_norm": 0.9524024724960327, + "learning_rate": 1.9285753720923384e-05, + "loss": 0.1337, + "step": 34086 + }, + { + "epoch": 0.7990529603778782, + "grad_norm": 2.3032710552215576, + "learning_rate": 1.9281405279205155e-05, + "loss": 0.2038, + "step": 34087 + }, + { + "epoch": 0.7990764019526831, + "grad_norm": 1.709739327430725, + "learning_rate": 1.9277057275468312e-05, + "loss": 0.0811, + "step": 34088 + }, + { + "epoch": 0.7990998435274882, + "grad_norm": 2.2848057746887207, + "learning_rate": 1.9272709709736425e-05, + "loss": 0.207, + "step": 34089 + }, + { + "epoch": 0.7991232851022931, + "grad_norm": 4.292600154876709, + "learning_rate": 1.9268362582033085e-05, + "loss": 0.4173, + "step": 34090 + }, + { + "epoch": 0.7991467266770982, + "grad_norm": 2.7890355587005615, + "learning_rate": 1.926401589238188e-05, + "loss": 0.2325, + "step": 34091 + }, + { + "epoch": 0.7991701682519031, + "grad_norm": 1.8830279111862183, + "learning_rate": 1.9259669640806388e-05, + "loss": 1.3489, + "step": 34092 + }, + { + "epoch": 0.7991936098267082, + "grad_norm": 2.379612445831299, + "learning_rate": 1.9255323827330186e-05, + "loss": 0.2548, + "step": 34093 + }, + { + "epoch": 0.7992170514015131, + "grad_norm": 1.1068553924560547, + "learning_rate": 1.9250978451976854e-05, + "loss": 0.0768, + "step": 34094 + }, + { + "epoch": 0.7992404929763182, + "grad_norm": 0.2892569303512573, + "learning_rate": 1.924663351477002e-05, + "loss": 0.0112, + "step": 34095 + }, + { + "epoch": 0.7992639345511231, + "grad_norm": 1.3848624229431152, + "learning_rate": 1.9242289015733217e-05, + "loss": 0.0935, + "step": 34096 + }, + { + "epoch": 0.7992873761259282, + "grad_norm": 7.328225612640381, + "learning_rate": 1.923794495489003e-05, + "loss": 0.3101, + "step": 34097 + }, + { + "epoch": 0.7993108177007331, + "grad_norm": 3.5910234451293945, + "learning_rate": 1.9233601332264016e-05, + "loss": 0.1674, + "step": 34098 + }, + { + "epoch": 0.7993342592755381, + "grad_norm": 2.2653467655181885, + "learning_rate": 1.922925814787876e-05, + "loss": 0.2202, + "step": 34099 + }, + { + "epoch": 0.7993577008503431, + "grad_norm": 3.7967605590820312, + "learning_rate": 1.9224915401757794e-05, + "loss": 0.4486, + "step": 34100 + }, + { + "epoch": 0.7993811424251481, + "grad_norm": 1.796019196510315, + "learning_rate": 1.9220573093924732e-05, + "loss": 0.1043, + "step": 34101 + }, + { + "epoch": 0.7994045839999531, + "grad_norm": 2.4374427795410156, + "learning_rate": 1.9216231224403113e-05, + "loss": 0.2474, + "step": 34102 + }, + { + "epoch": 0.7994280255747581, + "grad_norm": 4.446595191955566, + "learning_rate": 1.9211889793216487e-05, + "loss": 0.3972, + "step": 34103 + }, + { + "epoch": 0.799451467149563, + "grad_norm": 5.3534345626831055, + "learning_rate": 1.9207548800388432e-05, + "loss": 0.5215, + "step": 34104 + }, + { + "epoch": 0.7994749087243681, + "grad_norm": 2.063235282897949, + "learning_rate": 1.9203208245942482e-05, + "loss": 0.111, + "step": 34105 + }, + { + "epoch": 0.799498350299173, + "grad_norm": 2.1564948558807373, + "learning_rate": 1.919886812990217e-05, + "loss": 0.1442, + "step": 34106 + }, + { + "epoch": 0.7995217918739781, + "grad_norm": 1.1719763278961182, + "learning_rate": 1.919452845229107e-05, + "loss": 0.0593, + "step": 34107 + }, + { + "epoch": 0.7995452334487831, + "grad_norm": 4.1998419761657715, + "learning_rate": 1.919018921313276e-05, + "loss": 0.2728, + "step": 34108 + }, + { + "epoch": 0.7995686750235881, + "grad_norm": 3.504624128341675, + "learning_rate": 1.9185850412450746e-05, + "loss": 0.3137, + "step": 34109 + }, + { + "epoch": 0.7995921165983931, + "grad_norm": 4.7414069175720215, + "learning_rate": 1.9181512050268592e-05, + "loss": 0.4055, + "step": 34110 + }, + { + "epoch": 0.7996155581731981, + "grad_norm": 2.4254517555236816, + "learning_rate": 1.9177174126609822e-05, + "loss": 0.1104, + "step": 34111 + }, + { + "epoch": 0.7996389997480031, + "grad_norm": 4.762918472290039, + "learning_rate": 1.9172836641497972e-05, + "loss": 0.3, + "step": 34112 + }, + { + "epoch": 0.7996624413228081, + "grad_norm": 2.0102341175079346, + "learning_rate": 1.9168499594956567e-05, + "loss": 0.149, + "step": 34113 + }, + { + "epoch": 0.7996858828976131, + "grad_norm": 1.743287444114685, + "learning_rate": 1.916416298700917e-05, + "loss": 0.1099, + "step": 34114 + }, + { + "epoch": 0.799709324472418, + "grad_norm": 2.4631052017211914, + "learning_rate": 1.9159826817679305e-05, + "loss": 0.2182, + "step": 34115 + }, + { + "epoch": 0.7997327660472231, + "grad_norm": 2.1784842014312744, + "learning_rate": 1.9155491086990496e-05, + "loss": 1.8347, + "step": 34116 + }, + { + "epoch": 0.799756207622028, + "grad_norm": 4.4692792892456055, + "learning_rate": 1.9151155794966256e-05, + "loss": 0.3773, + "step": 34117 + }, + { + "epoch": 0.7997796491968331, + "grad_norm": 2.209888219833374, + "learning_rate": 1.9146820941630104e-05, + "loss": 0.1503, + "step": 34118 + }, + { + "epoch": 0.799803090771638, + "grad_norm": 3.448875904083252, + "learning_rate": 1.9142486527005567e-05, + "loss": 0.272, + "step": 34119 + }, + { + "epoch": 0.7998265323464431, + "grad_norm": 2.129844903945923, + "learning_rate": 1.9138152551116195e-05, + "loss": 0.1046, + "step": 34120 + }, + { + "epoch": 0.799849973921248, + "grad_norm": 2.2028231620788574, + "learning_rate": 1.9133819013985488e-05, + "loss": 0.0895, + "step": 34121 + }, + { + "epoch": 0.7998734154960531, + "grad_norm": 0.9332128763198853, + "learning_rate": 1.9129485915636957e-05, + "loss": 0.1127, + "step": 34122 + }, + { + "epoch": 0.799896857070858, + "grad_norm": 2.24991774559021, + "learning_rate": 1.91251532560941e-05, + "loss": 0.1365, + "step": 34123 + }, + { + "epoch": 0.799920298645663, + "grad_norm": 0.41129621863365173, + "learning_rate": 1.9120821035380433e-05, + "loss": 0.0354, + "step": 34124 + }, + { + "epoch": 0.799943740220468, + "grad_norm": 0.4738618731498718, + "learning_rate": 1.9116489253519444e-05, + "loss": 0.0469, + "step": 34125 + }, + { + "epoch": 0.799967181795273, + "grad_norm": 0.5460931658744812, + "learning_rate": 1.911215791053469e-05, + "loss": 0.0562, + "step": 34126 + }, + { + "epoch": 0.799990623370078, + "grad_norm": 1.7018250226974487, + "learning_rate": 1.910782700644963e-05, + "loss": 0.1197, + "step": 34127 + }, + { + "epoch": 0.800014064944883, + "grad_norm": 3.60795521736145, + "learning_rate": 1.910349654128778e-05, + "loss": 0.2522, + "step": 34128 + }, + { + "epoch": 0.800037506519688, + "grad_norm": 3.1422910690307617, + "learning_rate": 1.9099166515072643e-05, + "loss": 0.2831, + "step": 34129 + }, + { + "epoch": 0.800060948094493, + "grad_norm": 1.090989112854004, + "learning_rate": 1.9094836927827687e-05, + "loss": 0.105, + "step": 34130 + }, + { + "epoch": 0.800084389669298, + "grad_norm": 2.2553017139434814, + "learning_rate": 1.9090507779576404e-05, + "loss": 0.1359, + "step": 34131 + }, + { + "epoch": 0.800107831244103, + "grad_norm": 4.319013595581055, + "learning_rate": 1.908617907034229e-05, + "loss": 0.466, + "step": 34132 + }, + { + "epoch": 0.800131272818908, + "grad_norm": 1.3981893062591553, + "learning_rate": 1.9081850800148882e-05, + "loss": 0.0903, + "step": 34133 + }, + { + "epoch": 0.800154714393713, + "grad_norm": 0.7807343602180481, + "learning_rate": 1.9077522969019622e-05, + "loss": 0.071, + "step": 34134 + }, + { + "epoch": 0.8001781559685179, + "grad_norm": 2.942040205001831, + "learning_rate": 1.9073195576978e-05, + "loss": 1.7831, + "step": 34135 + }, + { + "epoch": 0.800201597543323, + "grad_norm": 2.2611560821533203, + "learning_rate": 1.9068868624047498e-05, + "loss": 0.2026, + "step": 34136 + }, + { + "epoch": 0.8002250391181279, + "grad_norm": 1.9727036952972412, + "learning_rate": 1.9064542110251582e-05, + "loss": 0.0853, + "step": 34137 + }, + { + "epoch": 0.800248480692933, + "grad_norm": 3.8180360794067383, + "learning_rate": 1.9060216035613708e-05, + "loss": 0.3029, + "step": 34138 + }, + { + "epoch": 0.8002719222677379, + "grad_norm": 1.9764307737350464, + "learning_rate": 1.9055890400157417e-05, + "loss": 0.1127, + "step": 34139 + }, + { + "epoch": 0.800295363842543, + "grad_norm": 3.2777273654937744, + "learning_rate": 1.905156520390613e-05, + "loss": 0.2174, + "step": 34140 + }, + { + "epoch": 0.8003188054173479, + "grad_norm": 1.3742214441299438, + "learning_rate": 1.904724044688333e-05, + "loss": 0.4795, + "step": 34141 + }, + { + "epoch": 0.800342246992153, + "grad_norm": 0.6532623767852783, + "learning_rate": 1.9042916129112477e-05, + "loss": 0.0336, + "step": 34142 + }, + { + "epoch": 0.8003656885669579, + "grad_norm": 2.3134677410125732, + "learning_rate": 1.9038592250617004e-05, + "loss": 0.1291, + "step": 34143 + }, + { + "epoch": 0.8003891301417629, + "grad_norm": 0.3940143883228302, + "learning_rate": 1.903426881142044e-05, + "loss": 0.0159, + "step": 34144 + }, + { + "epoch": 0.8004125717165679, + "grad_norm": 3.568523406982422, + "learning_rate": 1.902994581154619e-05, + "loss": 0.3367, + "step": 34145 + }, + { + "epoch": 0.8004360132913729, + "grad_norm": 2.920818328857422, + "learning_rate": 1.9025623251017756e-05, + "loss": 0.1997, + "step": 34146 + }, + { + "epoch": 0.8004594548661779, + "grad_norm": 0.710341215133667, + "learning_rate": 1.902130112985856e-05, + "loss": 0.0538, + "step": 34147 + }, + { + "epoch": 0.8004828964409829, + "grad_norm": 1.346016764640808, + "learning_rate": 1.9016979448092064e-05, + "loss": 0.0658, + "step": 34148 + }, + { + "epoch": 0.8005063380157879, + "grad_norm": 2.8586483001708984, + "learning_rate": 1.9012658205741708e-05, + "loss": 0.1042, + "step": 34149 + }, + { + "epoch": 0.8005297795905929, + "grad_norm": 0.8684011697769165, + "learning_rate": 1.9008337402830957e-05, + "loss": 0.0589, + "step": 34150 + }, + { + "epoch": 0.8005532211653978, + "grad_norm": 2.2287187576293945, + "learning_rate": 1.9004017039383203e-05, + "loss": 0.0861, + "step": 34151 + }, + { + "epoch": 0.8005766627402029, + "grad_norm": 4.153589248657227, + "learning_rate": 1.8999697115421967e-05, + "loss": 0.2611, + "step": 34152 + }, + { + "epoch": 0.8006001043150078, + "grad_norm": 0.3081001043319702, + "learning_rate": 1.8995377630970656e-05, + "loss": 0.0315, + "step": 34153 + }, + { + "epoch": 0.8006235458898129, + "grad_norm": 2.678802013397217, + "learning_rate": 1.89910585860527e-05, + "loss": 0.1765, + "step": 34154 + }, + { + "epoch": 0.8006469874646178, + "grad_norm": 0.2846355140209198, + "learning_rate": 1.8986739980691503e-05, + "loss": 0.0181, + "step": 34155 + }, + { + "epoch": 0.8006704290394229, + "grad_norm": 2.761263608932495, + "learning_rate": 1.898242181491057e-05, + "loss": 0.1643, + "step": 34156 + }, + { + "epoch": 0.8006938706142278, + "grad_norm": 2.4834039211273193, + "learning_rate": 1.8978104088733262e-05, + "loss": 0.1484, + "step": 34157 + }, + { + "epoch": 0.8007173121890329, + "grad_norm": 2.5629518032073975, + "learning_rate": 1.8973786802183067e-05, + "loss": 1.3358, + "step": 34158 + }, + { + "epoch": 0.8007407537638379, + "grad_norm": 2.9300570487976074, + "learning_rate": 1.8969469955283392e-05, + "loss": 0.1589, + "step": 34159 + }, + { + "epoch": 0.8007641953386428, + "grad_norm": 3.3207504749298096, + "learning_rate": 1.8965153548057636e-05, + "loss": 0.26, + "step": 34160 + }, + { + "epoch": 0.8007876369134479, + "grad_norm": 5.302896022796631, + "learning_rate": 1.8960837580529244e-05, + "loss": 0.5268, + "step": 34161 + }, + { + "epoch": 0.8008110784882528, + "grad_norm": 2.183945655822754, + "learning_rate": 1.8956522052721625e-05, + "loss": 0.1534, + "step": 34162 + }, + { + "epoch": 0.8008345200630579, + "grad_norm": 0.3733849823474884, + "learning_rate": 1.895220696465817e-05, + "loss": 0.0305, + "step": 34163 + }, + { + "epoch": 0.8008579616378628, + "grad_norm": 4.87369966506958, + "learning_rate": 1.8947892316362347e-05, + "loss": 0.4047, + "step": 34164 + }, + { + "epoch": 0.8008814032126679, + "grad_norm": 2.946462392807007, + "learning_rate": 1.894357810785753e-05, + "loss": 0.1777, + "step": 34165 + }, + { + "epoch": 0.8009048447874728, + "grad_norm": 5.076092720031738, + "learning_rate": 1.8939264339167152e-05, + "loss": 0.433, + "step": 34166 + }, + { + "epoch": 0.8009282863622779, + "grad_norm": 1.9829951524734497, + "learning_rate": 1.893495101031457e-05, + "loss": 0.1844, + "step": 34167 + }, + { + "epoch": 0.8009517279370828, + "grad_norm": 1.9963135719299316, + "learning_rate": 1.893063812132325e-05, + "loss": 1.4904, + "step": 34168 + }, + { + "epoch": 0.8009751695118879, + "grad_norm": 0.6974406838417053, + "learning_rate": 1.8926325672216562e-05, + "loss": 0.0446, + "step": 34169 + }, + { + "epoch": 0.8009986110866928, + "grad_norm": 2.310971736907959, + "learning_rate": 1.892201366301789e-05, + "loss": 0.0677, + "step": 34170 + }, + { + "epoch": 0.8010220526614978, + "grad_norm": 1.7614177465438843, + "learning_rate": 1.8917702093750676e-05, + "loss": 0.12, + "step": 34171 + }, + { + "epoch": 0.8010454942363028, + "grad_norm": 2.0145912170410156, + "learning_rate": 1.8913390964438294e-05, + "loss": 0.2066, + "step": 34172 + }, + { + "epoch": 0.8010689358111078, + "grad_norm": 1.005293607711792, + "learning_rate": 1.8909080275104128e-05, + "loss": 0.062, + "step": 34173 + }, + { + "epoch": 0.8010923773859128, + "grad_norm": 1.4329286813735962, + "learning_rate": 1.8904770025771577e-05, + "loss": 0.1005, + "step": 34174 + }, + { + "epoch": 0.8011158189607178, + "grad_norm": 0.575209379196167, + "learning_rate": 1.8900460216464024e-05, + "loss": 0.0436, + "step": 34175 + }, + { + "epoch": 0.8011392605355228, + "grad_norm": 0.7211806774139404, + "learning_rate": 1.8896150847204817e-05, + "loss": 0.0355, + "step": 34176 + }, + { + "epoch": 0.8011627021103278, + "grad_norm": 3.5803513526916504, + "learning_rate": 1.8891841918017418e-05, + "loss": 0.2438, + "step": 34177 + }, + { + "epoch": 0.8011861436851327, + "grad_norm": 4.828100681304932, + "learning_rate": 1.8887533428925154e-05, + "loss": 1.8859, + "step": 34178 + }, + { + "epoch": 0.8012095852599378, + "grad_norm": 0.5214434266090393, + "learning_rate": 1.888322537995142e-05, + "loss": 0.0382, + "step": 34179 + }, + { + "epoch": 0.8012330268347427, + "grad_norm": 4.253694534301758, + "learning_rate": 1.8878917771119554e-05, + "loss": 0.3669, + "step": 34180 + }, + { + "epoch": 0.8012564684095478, + "grad_norm": 4.425541400909424, + "learning_rate": 1.887461060245299e-05, + "loss": 0.471, + "step": 34181 + }, + { + "epoch": 0.8012799099843527, + "grad_norm": 0.9770222902297974, + "learning_rate": 1.8870303873975083e-05, + "loss": 0.0753, + "step": 34182 + }, + { + "epoch": 0.8013033515591578, + "grad_norm": 1.8997787237167358, + "learning_rate": 1.8865997585709138e-05, + "loss": 0.091, + "step": 34183 + }, + { + "epoch": 0.8013267931339627, + "grad_norm": 2.0571677684783936, + "learning_rate": 1.886169173767861e-05, + "loss": 0.0791, + "step": 34184 + }, + { + "epoch": 0.8013502347087678, + "grad_norm": 4.455572128295898, + "learning_rate": 1.8857386329906836e-05, + "loss": 0.392, + "step": 34185 + }, + { + "epoch": 0.8013736762835727, + "grad_norm": 0.43322473764419556, + "learning_rate": 1.8853081362417147e-05, + "loss": 0.0328, + "step": 34186 + }, + { + "epoch": 0.8013971178583777, + "grad_norm": 1.1543346643447876, + "learning_rate": 1.8848776835232927e-05, + "loss": 0.0832, + "step": 34187 + }, + { + "epoch": 0.8014205594331827, + "grad_norm": 0.6999908089637756, + "learning_rate": 1.8844472748377494e-05, + "loss": 0.0411, + "step": 34188 + }, + { + "epoch": 0.8014440010079877, + "grad_norm": 2.2415854930877686, + "learning_rate": 1.8840169101874274e-05, + "loss": 0.1065, + "step": 34189 + }, + { + "epoch": 0.8014674425827927, + "grad_norm": 2.9867100715637207, + "learning_rate": 1.8835865895746562e-05, + "loss": 0.2181, + "step": 34190 + }, + { + "epoch": 0.8014908841575977, + "grad_norm": 3.593449831008911, + "learning_rate": 1.8831563130017738e-05, + "loss": 0.4075, + "step": 34191 + }, + { + "epoch": 0.8015143257324027, + "grad_norm": 1.7410578727722168, + "learning_rate": 1.88272608047111e-05, + "loss": 0.126, + "step": 34192 + }, + { + "epoch": 0.8015377673072077, + "grad_norm": 0.5618290305137634, + "learning_rate": 1.8822958919850053e-05, + "loss": 0.0194, + "step": 34193 + }, + { + "epoch": 0.8015612088820127, + "grad_norm": 2.135896921157837, + "learning_rate": 1.8818657475457912e-05, + "loss": 0.1537, + "step": 34194 + }, + { + "epoch": 0.8015846504568177, + "grad_norm": 4.435616493225098, + "learning_rate": 1.8814356471558004e-05, + "loss": 0.376, + "step": 34195 + }, + { + "epoch": 0.8016080920316226, + "grad_norm": 9.451090812683105, + "learning_rate": 1.8810055908173695e-05, + "loss": 0.4615, + "step": 34196 + }, + { + "epoch": 0.8016315336064277, + "grad_norm": 2.5334982872009277, + "learning_rate": 1.8805755785328304e-05, + "loss": 0.101, + "step": 34197 + }, + { + "epoch": 0.8016549751812326, + "grad_norm": 0.6242969036102295, + "learning_rate": 1.880145610304517e-05, + "loss": 0.0376, + "step": 34198 + }, + { + "epoch": 0.8016784167560377, + "grad_norm": 2.7070329189300537, + "learning_rate": 1.879715686134762e-05, + "loss": 0.2405, + "step": 34199 + }, + { + "epoch": 0.8017018583308426, + "grad_norm": 3.798989772796631, + "learning_rate": 1.8792858060258977e-05, + "loss": 0.4204, + "step": 34200 + }, + { + "epoch": 0.8017252999056477, + "grad_norm": 2.2546379566192627, + "learning_rate": 1.8788559699802544e-05, + "loss": 0.1644, + "step": 34201 + }, + { + "epoch": 0.8017487414804526, + "grad_norm": 0.879182755947113, + "learning_rate": 1.87842617800017e-05, + "loss": 0.0339, + "step": 34202 + }, + { + "epoch": 0.8017721830552577, + "grad_norm": 1.0303716659545898, + "learning_rate": 1.8779964300879728e-05, + "loss": 0.0548, + "step": 34203 + }, + { + "epoch": 0.8017956246300626, + "grad_norm": 2.9521172046661377, + "learning_rate": 1.877566726245993e-05, + "loss": 0.1792, + "step": 34204 + }, + { + "epoch": 0.8018190662048676, + "grad_norm": 2.0573348999023438, + "learning_rate": 1.8771370664765676e-05, + "loss": 0.0686, + "step": 34205 + }, + { + "epoch": 0.8018425077796726, + "grad_norm": 5.094662189483643, + "learning_rate": 1.8767074507820237e-05, + "loss": 0.5392, + "step": 34206 + }, + { + "epoch": 0.8018659493544776, + "grad_norm": 4.555140018463135, + "learning_rate": 1.8762778791646953e-05, + "loss": 0.2266, + "step": 34207 + }, + { + "epoch": 0.8018893909292826, + "grad_norm": 4.179629325866699, + "learning_rate": 1.8758483516269075e-05, + "loss": 0.294, + "step": 34208 + }, + { + "epoch": 0.8019128325040876, + "grad_norm": 3.270416021347046, + "learning_rate": 1.8754188681709983e-05, + "loss": 0.3551, + "step": 34209 + }, + { + "epoch": 0.8019362740788927, + "grad_norm": 3.569949150085449, + "learning_rate": 1.8749894287992942e-05, + "loss": 0.2252, + "step": 34210 + }, + { + "epoch": 0.8019597156536976, + "grad_norm": 3.154093027114868, + "learning_rate": 1.8745600335141266e-05, + "loss": 0.1448, + "step": 34211 + }, + { + "epoch": 0.8019831572285027, + "grad_norm": 1.315564513206482, + "learning_rate": 1.874130682317824e-05, + "loss": 0.0958, + "step": 34212 + }, + { + "epoch": 0.8020065988033076, + "grad_norm": 3.130692720413208, + "learning_rate": 1.8737013752127174e-05, + "loss": 0.2154, + "step": 34213 + }, + { + "epoch": 0.8020300403781127, + "grad_norm": 2.3901920318603516, + "learning_rate": 1.873272112201133e-05, + "loss": 0.1539, + "step": 34214 + }, + { + "epoch": 0.8020534819529176, + "grad_norm": 3.277639150619507, + "learning_rate": 1.8728428932854046e-05, + "loss": 0.3445, + "step": 34215 + }, + { + "epoch": 0.8020769235277226, + "grad_norm": 3.7842679023742676, + "learning_rate": 1.872413718467857e-05, + "loss": 0.4796, + "step": 34216 + }, + { + "epoch": 0.8021003651025276, + "grad_norm": 2.2715508937835693, + "learning_rate": 1.8719845877508237e-05, + "loss": 0.0739, + "step": 34217 + }, + { + "epoch": 0.8021238066773326, + "grad_norm": 1.6873644590377808, + "learning_rate": 1.8715555011366302e-05, + "loss": 0.0824, + "step": 34218 + }, + { + "epoch": 0.8021472482521376, + "grad_norm": 1.953597068786621, + "learning_rate": 1.8711264586276056e-05, + "loss": 0.1494, + "step": 34219 + }, + { + "epoch": 0.8021706898269426, + "grad_norm": 2.7849032878875732, + "learning_rate": 1.8706974602260772e-05, + "loss": 0.3242, + "step": 34220 + }, + { + "epoch": 0.8021941314017476, + "grad_norm": 3.2042977809906006, + "learning_rate": 1.8702685059343706e-05, + "loss": 0.4046, + "step": 34221 + }, + { + "epoch": 0.8022175729765526, + "grad_norm": 2.641279697418213, + "learning_rate": 1.869839595754819e-05, + "loss": 0.1029, + "step": 34222 + }, + { + "epoch": 0.8022410145513575, + "grad_norm": 3.339810371398926, + "learning_rate": 1.8694107296897445e-05, + "loss": 0.2657, + "step": 34223 + }, + { + "epoch": 0.8022644561261626, + "grad_norm": 2.293599843978882, + "learning_rate": 1.8689819077414773e-05, + "loss": 0.2131, + "step": 34224 + }, + { + "epoch": 0.8022878977009675, + "grad_norm": 2.9320404529571533, + "learning_rate": 1.8685531299123428e-05, + "loss": 0.2073, + "step": 34225 + }, + { + "epoch": 0.8023113392757726, + "grad_norm": 1.155109167098999, + "learning_rate": 1.8681243962046646e-05, + "loss": 0.076, + "step": 34226 + }, + { + "epoch": 0.8023347808505775, + "grad_norm": 3.5092482566833496, + "learning_rate": 1.8676957066207744e-05, + "loss": 0.3372, + "step": 34227 + }, + { + "epoch": 0.8023582224253826, + "grad_norm": 0.47216299176216125, + "learning_rate": 1.8672670611629938e-05, + "loss": 0.0284, + "step": 34228 + }, + { + "epoch": 0.8023816640001875, + "grad_norm": 3.9743096828460693, + "learning_rate": 1.866838459833654e-05, + "loss": 0.3306, + "step": 34229 + }, + { + "epoch": 0.8024051055749926, + "grad_norm": 2.781799554824829, + "learning_rate": 1.866409902635077e-05, + "loss": 0.291, + "step": 34230 + }, + { + "epoch": 0.8024285471497975, + "grad_norm": 4.329686164855957, + "learning_rate": 1.8659813895695877e-05, + "loss": 0.4015, + "step": 34231 + }, + { + "epoch": 0.8024519887246025, + "grad_norm": 1.0655146837234497, + "learning_rate": 1.865552920639513e-05, + "loss": 0.0629, + "step": 34232 + }, + { + "epoch": 0.8024754302994075, + "grad_norm": 2.976057291030884, + "learning_rate": 1.8651244958471735e-05, + "loss": 0.1957, + "step": 34233 + }, + { + "epoch": 0.8024988718742125, + "grad_norm": 2.1713483333587646, + "learning_rate": 1.8646961151949006e-05, + "loss": 0.1448, + "step": 34234 + }, + { + "epoch": 0.8025223134490175, + "grad_norm": 2.494203805923462, + "learning_rate": 1.864267778685015e-05, + "loss": 0.1626, + "step": 34235 + }, + { + "epoch": 0.8025457550238225, + "grad_norm": 1.5327684879302979, + "learning_rate": 1.8638394863198416e-05, + "loss": 0.0509, + "step": 34236 + }, + { + "epoch": 0.8025691965986275, + "grad_norm": 2.407900094985962, + "learning_rate": 1.8634112381017034e-05, + "loss": 0.2061, + "step": 34237 + }, + { + "epoch": 0.8025926381734325, + "grad_norm": 1.5343610048294067, + "learning_rate": 1.8629830340329247e-05, + "loss": 0.1184, + "step": 34238 + }, + { + "epoch": 0.8026160797482375, + "grad_norm": 0.6083729863166809, + "learning_rate": 1.862554874115826e-05, + "loss": 0.0526, + "step": 34239 + }, + { + "epoch": 0.8026395213230425, + "grad_norm": 1.2042049169540405, + "learning_rate": 1.8621267583527347e-05, + "loss": 0.1534, + "step": 34240 + }, + { + "epoch": 0.8026629628978474, + "grad_norm": 4.196978569030762, + "learning_rate": 1.861698686745974e-05, + "loss": 0.2894, + "step": 34241 + }, + { + "epoch": 0.8026864044726525, + "grad_norm": 5.263103485107422, + "learning_rate": 1.8612706592978647e-05, + "loss": 0.2036, + "step": 34242 + }, + { + "epoch": 0.8027098460474574, + "grad_norm": 3.2570838928222656, + "learning_rate": 1.86084267601073e-05, + "loss": 1.5177, + "step": 34243 + }, + { + "epoch": 0.8027332876222625, + "grad_norm": 2.5420868396759033, + "learning_rate": 1.860414736886892e-05, + "loss": 0.2431, + "step": 34244 + }, + { + "epoch": 0.8027567291970674, + "grad_norm": 3.0777463912963867, + "learning_rate": 1.8599868419286727e-05, + "loss": 0.1542, + "step": 34245 + }, + { + "epoch": 0.8027801707718725, + "grad_norm": 0.9189311265945435, + "learning_rate": 1.8595589911383902e-05, + "loss": 0.0895, + "step": 34246 + }, + { + "epoch": 0.8028036123466774, + "grad_norm": 3.8488411903381348, + "learning_rate": 1.8591311845183725e-05, + "loss": 0.426, + "step": 34247 + }, + { + "epoch": 0.8028270539214825, + "grad_norm": 0.6193170547485352, + "learning_rate": 1.8587034220709376e-05, + "loss": 0.062, + "step": 34248 + }, + { + "epoch": 0.8028504954962874, + "grad_norm": 2.122870683670044, + "learning_rate": 1.8582757037984078e-05, + "loss": 0.1083, + "step": 34249 + }, + { + "epoch": 0.8028739370710924, + "grad_norm": 1.8821433782577515, + "learning_rate": 1.857848029703102e-05, + "loss": 0.1625, + "step": 34250 + }, + { + "epoch": 0.8028973786458974, + "grad_norm": 0.8905324339866638, + "learning_rate": 1.857420399787342e-05, + "loss": 0.0945, + "step": 34251 + }, + { + "epoch": 0.8029208202207024, + "grad_norm": 1.255116581916809, + "learning_rate": 1.8569928140534455e-05, + "loss": 0.0639, + "step": 34252 + }, + { + "epoch": 0.8029442617955074, + "grad_norm": 3.0958712100982666, + "learning_rate": 1.8565652725037353e-05, + "loss": 0.1665, + "step": 34253 + }, + { + "epoch": 0.8029677033703124, + "grad_norm": 2.1673169136047363, + "learning_rate": 1.8561377751405328e-05, + "loss": 0.1788, + "step": 34254 + }, + { + "epoch": 0.8029911449451174, + "grad_norm": 1.0195705890655518, + "learning_rate": 1.8557103219661553e-05, + "loss": 0.0536, + "step": 34255 + }, + { + "epoch": 0.8030145865199224, + "grad_norm": 4.189062118530273, + "learning_rate": 1.855282912982924e-05, + "loss": 0.4086, + "step": 34256 + }, + { + "epoch": 0.8030380280947274, + "grad_norm": 0.4305962026119232, + "learning_rate": 1.854855548193155e-05, + "loss": 0.034, + "step": 34257 + }, + { + "epoch": 0.8030614696695324, + "grad_norm": 2.0567173957824707, + "learning_rate": 1.85442822759917e-05, + "loss": 0.1226, + "step": 34258 + }, + { + "epoch": 0.8030849112443373, + "grad_norm": 3.3027477264404297, + "learning_rate": 1.854000951203284e-05, + "loss": 0.2378, + "step": 34259 + }, + { + "epoch": 0.8031083528191424, + "grad_norm": 1.1555129289627075, + "learning_rate": 1.8535737190078195e-05, + "loss": 0.1111, + "step": 34260 + }, + { + "epoch": 0.8031317943939474, + "grad_norm": 1.8582597970962524, + "learning_rate": 1.853146531015094e-05, + "loss": 0.1423, + "step": 34261 + }, + { + "epoch": 0.8031552359687524, + "grad_norm": 2.4728689193725586, + "learning_rate": 1.852719387227424e-05, + "loss": 0.2142, + "step": 34262 + }, + { + "epoch": 0.8031786775435574, + "grad_norm": 2.234269142150879, + "learning_rate": 1.852292287647128e-05, + "loss": 0.1896, + "step": 34263 + }, + { + "epoch": 0.8032021191183624, + "grad_norm": 2.5668675899505615, + "learning_rate": 1.8518652322765206e-05, + "loss": 0.1327, + "step": 34264 + }, + { + "epoch": 0.8032255606931674, + "grad_norm": 4.249267101287842, + "learning_rate": 1.8514382211179236e-05, + "loss": 0.3634, + "step": 34265 + }, + { + "epoch": 0.8032490022679724, + "grad_norm": 0.32164305448532104, + "learning_rate": 1.851011254173649e-05, + "loss": 0.0288, + "step": 34266 + }, + { + "epoch": 0.8032724438427774, + "grad_norm": 2.036700487136841, + "learning_rate": 1.85058433144602e-05, + "loss": 0.157, + "step": 34267 + }, + { + "epoch": 0.8032958854175823, + "grad_norm": 1.280909538269043, + "learning_rate": 1.850157452937349e-05, + "loss": 0.1413, + "step": 34268 + }, + { + "epoch": 0.8033193269923874, + "grad_norm": 1.5796184539794922, + "learning_rate": 1.849730618649953e-05, + "loss": 0.0769, + "step": 34269 + }, + { + "epoch": 0.8033427685671923, + "grad_norm": 1.5321762561798096, + "learning_rate": 1.8493038285861485e-05, + "loss": 0.0749, + "step": 34270 + }, + { + "epoch": 0.8033662101419974, + "grad_norm": 2.497044801712036, + "learning_rate": 1.848877082748247e-05, + "loss": 0.1068, + "step": 34271 + }, + { + "epoch": 0.8033896517168023, + "grad_norm": 3.9256350994110107, + "learning_rate": 1.8484503811385702e-05, + "loss": 0.2308, + "step": 34272 + }, + { + "epoch": 0.8034130932916074, + "grad_norm": 3.82936429977417, + "learning_rate": 1.848023723759431e-05, + "loss": 0.2963, + "step": 34273 + }, + { + "epoch": 0.8034365348664123, + "grad_norm": 3.040557384490967, + "learning_rate": 1.847597110613145e-05, + "loss": 0.2262, + "step": 34274 + }, + { + "epoch": 0.8034599764412174, + "grad_norm": 3.8182425498962402, + "learning_rate": 1.8471705417020258e-05, + "loss": 0.3917, + "step": 34275 + }, + { + "epoch": 0.8034834180160223, + "grad_norm": 0.22501029074192047, + "learning_rate": 1.846744017028389e-05, + "loss": 0.0126, + "step": 34276 + }, + { + "epoch": 0.8035068595908273, + "grad_norm": 2.9112017154693604, + "learning_rate": 1.846317536594545e-05, + "loss": 0.2052, + "step": 34277 + }, + { + "epoch": 0.8035303011656323, + "grad_norm": 2.2196385860443115, + "learning_rate": 1.8458911004028113e-05, + "loss": 0.122, + "step": 34278 + }, + { + "epoch": 0.8035537427404373, + "grad_norm": 1.7625927925109863, + "learning_rate": 1.8454647084555053e-05, + "loss": 1.4765, + "step": 34279 + }, + { + "epoch": 0.8035771843152423, + "grad_norm": 3.2667007446289062, + "learning_rate": 1.845038360754936e-05, + "loss": 0.2082, + "step": 34280 + }, + { + "epoch": 0.8036006258900473, + "grad_norm": 1.9406235218048096, + "learning_rate": 1.8446120573034188e-05, + "loss": 0.1355, + "step": 34281 + }, + { + "epoch": 0.8036240674648523, + "grad_norm": 2.018465518951416, + "learning_rate": 1.844185798103265e-05, + "loss": 0.1262, + "step": 34282 + }, + { + "epoch": 0.8036475090396573, + "grad_norm": 1.2645666599273682, + "learning_rate": 1.8437595831567888e-05, + "loss": 0.0373, + "step": 34283 + }, + { + "epoch": 0.8036709506144623, + "grad_norm": 4.485666751861572, + "learning_rate": 1.8433334124662994e-05, + "loss": 0.2407, + "step": 34284 + }, + { + "epoch": 0.8036943921892673, + "grad_norm": 2.3037681579589844, + "learning_rate": 1.8429072860341156e-05, + "loss": 0.2327, + "step": 34285 + }, + { + "epoch": 0.8037178337640722, + "grad_norm": 3.182379961013794, + "learning_rate": 1.842481203862546e-05, + "loss": 0.2889, + "step": 34286 + }, + { + "epoch": 0.8037412753388773, + "grad_norm": 0.9775959253311157, + "learning_rate": 1.842055165953903e-05, + "loss": 0.0683, + "step": 34287 + }, + { + "epoch": 0.8037647169136822, + "grad_norm": 2.048628091812134, + "learning_rate": 1.8416291723104974e-05, + "loss": 0.1506, + "step": 34288 + }, + { + "epoch": 0.8037881584884873, + "grad_norm": 3.6424694061279297, + "learning_rate": 1.8412032229346388e-05, + "loss": 0.4092, + "step": 34289 + }, + { + "epoch": 0.8038116000632922, + "grad_norm": 3.0114731788635254, + "learning_rate": 1.8407773178286435e-05, + "loss": 0.3057, + "step": 34290 + }, + { + "epoch": 0.8038350416380973, + "grad_norm": 2.4992380142211914, + "learning_rate": 1.8403514569948178e-05, + "loss": 0.2017, + "step": 34291 + }, + { + "epoch": 0.8038584832129022, + "grad_norm": 4.586641788482666, + "learning_rate": 1.839925640435477e-05, + "loss": 0.2586, + "step": 34292 + }, + { + "epoch": 0.8038819247877073, + "grad_norm": 2.092075824737549, + "learning_rate": 1.8394998681529286e-05, + "loss": 0.0855, + "step": 34293 + }, + { + "epoch": 0.8039053663625122, + "grad_norm": 1.8256937265396118, + "learning_rate": 1.839074140149484e-05, + "loss": 1.1893, + "step": 34294 + }, + { + "epoch": 0.8039288079373172, + "grad_norm": 2.4686217308044434, + "learning_rate": 1.838648456427452e-05, + "loss": 0.1429, + "step": 34295 + }, + { + "epoch": 0.8039522495121222, + "grad_norm": 0.9590938687324524, + "learning_rate": 1.8382228169891436e-05, + "loss": 0.0574, + "step": 34296 + }, + { + "epoch": 0.8039756910869272, + "grad_norm": 2.713392496109009, + "learning_rate": 1.8377972218368644e-05, + "loss": 0.2309, + "step": 34297 + }, + { + "epoch": 0.8039991326617322, + "grad_norm": 1.0623447895050049, + "learning_rate": 1.8373716709729307e-05, + "loss": 0.0775, + "step": 34298 + }, + { + "epoch": 0.8040225742365372, + "grad_norm": 2.437086343765259, + "learning_rate": 1.8369461643996467e-05, + "loss": 1.0288, + "step": 34299 + }, + { + "epoch": 0.8040460158113422, + "grad_norm": 2.826284646987915, + "learning_rate": 1.836520702119322e-05, + "loss": 0.1598, + "step": 34300 + }, + { + "epoch": 0.8040694573861472, + "grad_norm": 0.4545815885066986, + "learning_rate": 1.836095284134264e-05, + "loss": 0.0358, + "step": 34301 + }, + { + "epoch": 0.8040928989609522, + "grad_norm": 2.311760663986206, + "learning_rate": 1.8356699104467843e-05, + "loss": 0.0952, + "step": 34302 + }, + { + "epoch": 0.8041163405357572, + "grad_norm": 4.088245391845703, + "learning_rate": 1.83524458105919e-05, + "loss": 0.5738, + "step": 34303 + }, + { + "epoch": 0.8041397821105621, + "grad_norm": 1.2956287860870361, + "learning_rate": 1.8348192959737853e-05, + "loss": 0.1218, + "step": 34304 + }, + { + "epoch": 0.8041632236853672, + "grad_norm": 3.8081562519073486, + "learning_rate": 1.834394055192883e-05, + "loss": 0.3985, + "step": 34305 + }, + { + "epoch": 0.8041866652601721, + "grad_norm": 0.9316178560256958, + "learning_rate": 1.8339688587187888e-05, + "loss": 0.0796, + "step": 34306 + }, + { + "epoch": 0.8042101068349772, + "grad_norm": 2.680569887161255, + "learning_rate": 1.8335437065538085e-05, + "loss": 0.1499, + "step": 34307 + }, + { + "epoch": 0.8042335484097821, + "grad_norm": 3.670966863632202, + "learning_rate": 1.8331185987002496e-05, + "loss": 0.2675, + "step": 34308 + }, + { + "epoch": 0.8042569899845872, + "grad_norm": 4.317390441894531, + "learning_rate": 1.8326935351604167e-05, + "loss": 0.3322, + "step": 34309 + }, + { + "epoch": 0.8042804315593921, + "grad_norm": 2.6722569465637207, + "learning_rate": 1.8322685159366205e-05, + "loss": 0.2986, + "step": 34310 + }, + { + "epoch": 0.8043038731341972, + "grad_norm": 1.2519373893737793, + "learning_rate": 1.8318435410311653e-05, + "loss": 0.851, + "step": 34311 + }, + { + "epoch": 0.8043273147090022, + "grad_norm": 3.8980987071990967, + "learning_rate": 1.831418610446357e-05, + "loss": 0.2451, + "step": 34312 + }, + { + "epoch": 0.8043507562838071, + "grad_norm": 2.347121000289917, + "learning_rate": 1.8309937241844976e-05, + "loss": 1.0845, + "step": 34313 + }, + { + "epoch": 0.8043741978586122, + "grad_norm": 3.1634116172790527, + "learning_rate": 1.8305688822478994e-05, + "loss": 0.3209, + "step": 34314 + }, + { + "epoch": 0.8043976394334171, + "grad_norm": 2.320953845977783, + "learning_rate": 1.830144084638864e-05, + "loss": 0.1119, + "step": 34315 + }, + { + "epoch": 0.8044210810082222, + "grad_norm": 3.857449531555176, + "learning_rate": 1.829719331359694e-05, + "loss": 0.4328, + "step": 34316 + }, + { + "epoch": 0.8044445225830271, + "grad_norm": 4.20146369934082, + "learning_rate": 1.8292946224126983e-05, + "loss": 0.5429, + "step": 34317 + }, + { + "epoch": 0.8044679641578322, + "grad_norm": 0.7351998686790466, + "learning_rate": 1.82886995780018e-05, + "loss": 0.0578, + "step": 34318 + }, + { + "epoch": 0.8044914057326371, + "grad_norm": 1.722200632095337, + "learning_rate": 1.8284453375244438e-05, + "loss": 0.0888, + "step": 34319 + }, + { + "epoch": 0.8045148473074422, + "grad_norm": 2.380281448364258, + "learning_rate": 1.828020761587792e-05, + "loss": 1.4767, + "step": 34320 + }, + { + "epoch": 0.8045382888822471, + "grad_norm": 3.198934555053711, + "learning_rate": 1.82759622999253e-05, + "loss": 0.2494, + "step": 34321 + }, + { + "epoch": 0.8045617304570521, + "grad_norm": 0.3420599699020386, + "learning_rate": 1.8271717427409586e-05, + "loss": 0.0233, + "step": 34322 + }, + { + "epoch": 0.8045851720318571, + "grad_norm": 2.733523368835449, + "learning_rate": 1.826747299835384e-05, + "loss": 0.2458, + "step": 34323 + }, + { + "epoch": 0.8046086136066621, + "grad_norm": 2.478851795196533, + "learning_rate": 1.82632290127811e-05, + "loss": 1.4048, + "step": 34324 + }, + { + "epoch": 0.8046320551814671, + "grad_norm": 1.1498183012008667, + "learning_rate": 1.825898547071434e-05, + "loss": 0.0504, + "step": 34325 + }, + { + "epoch": 0.8046554967562721, + "grad_norm": 0.8389394879341125, + "learning_rate": 1.825474237217666e-05, + "loss": 0.0845, + "step": 34326 + }, + { + "epoch": 0.8046789383310771, + "grad_norm": 3.3437745571136475, + "learning_rate": 1.8250499717191038e-05, + "loss": 0.1927, + "step": 34327 + }, + { + "epoch": 0.8047023799058821, + "grad_norm": 3.0040462017059326, + "learning_rate": 1.8246257505780507e-05, + "loss": 0.3027, + "step": 34328 + }, + { + "epoch": 0.804725821480687, + "grad_norm": 1.8389290571212769, + "learning_rate": 1.8242015737968053e-05, + "loss": 0.1885, + "step": 34329 + }, + { + "epoch": 0.8047492630554921, + "grad_norm": 3.4557530879974365, + "learning_rate": 1.8237774413776752e-05, + "loss": 0.2837, + "step": 34330 + }, + { + "epoch": 0.804772704630297, + "grad_norm": 4.340514659881592, + "learning_rate": 1.8233533533229586e-05, + "loss": 0.2745, + "step": 34331 + }, + { + "epoch": 0.8047961462051021, + "grad_norm": 3.003734588623047, + "learning_rate": 1.8229293096349553e-05, + "loss": 0.1973, + "step": 34332 + }, + { + "epoch": 0.804819587779907, + "grad_norm": 2.216484785079956, + "learning_rate": 1.8225053103159683e-05, + "loss": 1.6812, + "step": 34333 + }, + { + "epoch": 0.8048430293547121, + "grad_norm": 2.6820430755615234, + "learning_rate": 1.822081355368298e-05, + "loss": 0.1897, + "step": 34334 + }, + { + "epoch": 0.804866470929517, + "grad_norm": 2.3302571773529053, + "learning_rate": 1.8216574447942413e-05, + "loss": 0.1196, + "step": 34335 + }, + { + "epoch": 0.8048899125043221, + "grad_norm": 2.22969913482666, + "learning_rate": 1.8212335785961032e-05, + "loss": 0.17, + "step": 34336 + }, + { + "epoch": 0.804913354079127, + "grad_norm": 3.136664628982544, + "learning_rate": 1.820809756776183e-05, + "loss": 0.3104, + "step": 34337 + }, + { + "epoch": 0.8049367956539321, + "grad_norm": 0.9537093043327332, + "learning_rate": 1.820385979336775e-05, + "loss": 0.063, + "step": 34338 + }, + { + "epoch": 0.804960237228737, + "grad_norm": 3.141735792160034, + "learning_rate": 1.8199622462801856e-05, + "loss": 0.1944, + "step": 34339 + }, + { + "epoch": 0.804983678803542, + "grad_norm": 0.3901561498641968, + "learning_rate": 1.8195385576087122e-05, + "loss": 0.0273, + "step": 34340 + }, + { + "epoch": 0.805007120378347, + "grad_norm": 7.18742561340332, + "learning_rate": 1.819114913324651e-05, + "loss": 0.4811, + "step": 34341 + }, + { + "epoch": 0.805030561953152, + "grad_norm": 1.0463861227035522, + "learning_rate": 1.818691313430301e-05, + "loss": 0.1209, + "step": 34342 + }, + { + "epoch": 0.805054003527957, + "grad_norm": 2.991237163543701, + "learning_rate": 1.8182677579279637e-05, + "loss": 0.2361, + "step": 34343 + }, + { + "epoch": 0.805077445102762, + "grad_norm": 1.1526429653167725, + "learning_rate": 1.8178442468199363e-05, + "loss": 0.0489, + "step": 34344 + }, + { + "epoch": 0.805100886677567, + "grad_norm": 2.350661516189575, + "learning_rate": 1.8174207801085152e-05, + "loss": 0.2389, + "step": 34345 + }, + { + "epoch": 0.805124328252372, + "grad_norm": 0.8486694693565369, + "learning_rate": 1.816997357796e-05, + "loss": 0.0411, + "step": 34346 + }, + { + "epoch": 0.805147769827177, + "grad_norm": 1.1548254489898682, + "learning_rate": 1.8165739798846837e-05, + "loss": 0.1317, + "step": 34347 + }, + { + "epoch": 0.805171211401982, + "grad_norm": 2.860755681991577, + "learning_rate": 1.8161506463768698e-05, + "loss": 0.1455, + "step": 34348 + }, + { + "epoch": 0.8051946529767869, + "grad_norm": 3.1342220306396484, + "learning_rate": 1.8157273572748533e-05, + "loss": 0.261, + "step": 34349 + }, + { + "epoch": 0.805218094551592, + "grad_norm": 2.9950637817382812, + "learning_rate": 1.8153041125809268e-05, + "loss": 0.3088, + "step": 34350 + }, + { + "epoch": 0.8052415361263969, + "grad_norm": 3.5944488048553467, + "learning_rate": 1.8148809122973944e-05, + "loss": 0.3809, + "step": 34351 + }, + { + "epoch": 0.805264977701202, + "grad_norm": 1.9126683473587036, + "learning_rate": 1.8144577564265465e-05, + "loss": 1.5432, + "step": 34352 + }, + { + "epoch": 0.8052884192760069, + "grad_norm": 2.837491750717163, + "learning_rate": 1.8140346449706813e-05, + "loss": 0.3339, + "step": 34353 + }, + { + "epoch": 0.805311860850812, + "grad_norm": 3.968583583831787, + "learning_rate": 1.813611577932093e-05, + "loss": 0.3387, + "step": 34354 + }, + { + "epoch": 0.8053353024256169, + "grad_norm": 1.1113736629486084, + "learning_rate": 1.81318855531308e-05, + "loss": 0.1293, + "step": 34355 + }, + { + "epoch": 0.805358744000422, + "grad_norm": 2.751612901687622, + "learning_rate": 1.8127655771159357e-05, + "loss": 0.2453, + "step": 34356 + }, + { + "epoch": 0.8053821855752269, + "grad_norm": 1.2586679458618164, + "learning_rate": 1.812342643342957e-05, + "loss": 0.118, + "step": 34357 + }, + { + "epoch": 0.8054056271500319, + "grad_norm": 0.9816782474517822, + "learning_rate": 1.811919753996436e-05, + "loss": 0.0879, + "step": 34358 + }, + { + "epoch": 0.8054290687248369, + "grad_norm": 1.0496209859848022, + "learning_rate": 1.8114969090786693e-05, + "loss": 0.098, + "step": 34359 + }, + { + "epoch": 0.8054525102996419, + "grad_norm": 0.8338420391082764, + "learning_rate": 1.811074108591948e-05, + "loss": 0.0282, + "step": 34360 + }, + { + "epoch": 0.8054759518744469, + "grad_norm": 3.5416312217712402, + "learning_rate": 1.8106513525385717e-05, + "loss": 0.2898, + "step": 34361 + }, + { + "epoch": 0.8054993934492519, + "grad_norm": 2.323397159576416, + "learning_rate": 1.810228640920828e-05, + "loss": 1.6058, + "step": 34362 + }, + { + "epoch": 0.805522835024057, + "grad_norm": 2.5326428413391113, + "learning_rate": 1.8098059737410167e-05, + "loss": 0.144, + "step": 34363 + }, + { + "epoch": 0.8055462765988619, + "grad_norm": 1.8661208152770996, + "learning_rate": 1.8093833510014292e-05, + "loss": 0.1805, + "step": 34364 + }, + { + "epoch": 0.805569718173667, + "grad_norm": 0.8230056166648865, + "learning_rate": 1.8089607727043577e-05, + "loss": 0.4843, + "step": 34365 + }, + { + "epoch": 0.8055931597484719, + "grad_norm": 1.3192577362060547, + "learning_rate": 1.8085382388520954e-05, + "loss": 0.0737, + "step": 34366 + }, + { + "epoch": 0.805616601323277, + "grad_norm": 2.8458285331726074, + "learning_rate": 1.8081157494469315e-05, + "loss": 0.111, + "step": 34367 + }, + { + "epoch": 0.8056400428980819, + "grad_norm": 4.167299270629883, + "learning_rate": 1.807693304491166e-05, + "loss": 0.1467, + "step": 34368 + }, + { + "epoch": 0.8056634844728869, + "grad_norm": 3.5343427658081055, + "learning_rate": 1.807270903987085e-05, + "loss": 0.2432, + "step": 34369 + }, + { + "epoch": 0.8056869260476919, + "grad_norm": 3.3180716037750244, + "learning_rate": 1.8068485479369847e-05, + "loss": 0.2018, + "step": 34370 + }, + { + "epoch": 0.8057103676224969, + "grad_norm": 6.330245018005371, + "learning_rate": 1.806426236343153e-05, + "loss": 0.8947, + "step": 34371 + }, + { + "epoch": 0.8057338091973019, + "grad_norm": 1.0721709728240967, + "learning_rate": 1.8060039692078834e-05, + "loss": 0.0739, + "step": 34372 + }, + { + "epoch": 0.8057572507721069, + "grad_norm": 1.0555791854858398, + "learning_rate": 1.8055817465334633e-05, + "loss": 0.1021, + "step": 34373 + }, + { + "epoch": 0.8057806923469119, + "grad_norm": 2.042741298675537, + "learning_rate": 1.805159568322188e-05, + "loss": 0.2325, + "step": 34374 + }, + { + "epoch": 0.8058041339217169, + "grad_norm": 2.642871379852295, + "learning_rate": 1.8047374345763503e-05, + "loss": 0.2178, + "step": 34375 + }, + { + "epoch": 0.8058275754965218, + "grad_norm": 4.775494575500488, + "learning_rate": 1.8043153452982363e-05, + "loss": 0.3379, + "step": 34376 + }, + { + "epoch": 0.8058510170713269, + "grad_norm": 1.6463929414749146, + "learning_rate": 1.803893300490138e-05, + "loss": 0.1184, + "step": 34377 + }, + { + "epoch": 0.8058744586461318, + "grad_norm": 2.7887983322143555, + "learning_rate": 1.8034713001543458e-05, + "loss": 0.2748, + "step": 34378 + }, + { + "epoch": 0.8058979002209369, + "grad_norm": 1.9404312372207642, + "learning_rate": 1.8030493442931486e-05, + "loss": 0.1433, + "step": 34379 + }, + { + "epoch": 0.8059213417957418, + "grad_norm": 2.667266845703125, + "learning_rate": 1.8026274329088332e-05, + "loss": 0.2683, + "step": 34380 + }, + { + "epoch": 0.8059447833705469, + "grad_norm": 2.9134323596954346, + "learning_rate": 1.802205566003695e-05, + "loss": 0.0713, + "step": 34381 + }, + { + "epoch": 0.8059682249453518, + "grad_norm": 4.011430263519287, + "learning_rate": 1.80178374358002e-05, + "loss": 0.4142, + "step": 34382 + }, + { + "epoch": 0.8059916665201569, + "grad_norm": 3.496009349822998, + "learning_rate": 1.801361965640096e-05, + "loss": 0.2642, + "step": 34383 + }, + { + "epoch": 0.8060151080949618, + "grad_norm": 0.5302899479866028, + "learning_rate": 1.8009402321862135e-05, + "loss": 0.0377, + "step": 34384 + }, + { + "epoch": 0.8060385496697668, + "grad_norm": 0.9695305824279785, + "learning_rate": 1.800518543220656e-05, + "loss": 0.0913, + "step": 34385 + }, + { + "epoch": 0.8060619912445718, + "grad_norm": 2.7989001274108887, + "learning_rate": 1.8000968987457188e-05, + "loss": 0.193, + "step": 34386 + }, + { + "epoch": 0.8060854328193768, + "grad_norm": 3.7614645957946777, + "learning_rate": 1.7996752987636844e-05, + "loss": 0.4703, + "step": 34387 + }, + { + "epoch": 0.8061088743941818, + "grad_norm": 1.5649493932724, + "learning_rate": 1.799253743276844e-05, + "loss": 0.1407, + "step": 34388 + }, + { + "epoch": 0.8061323159689868, + "grad_norm": 1.4669607877731323, + "learning_rate": 1.798832232287484e-05, + "loss": 0.1087, + "step": 34389 + }, + { + "epoch": 0.8061557575437918, + "grad_norm": 2.942272424697876, + "learning_rate": 1.79841076579789e-05, + "loss": 0.2352, + "step": 34390 + }, + { + "epoch": 0.8061791991185968, + "grad_norm": 3.226193904876709, + "learning_rate": 1.79798934381035e-05, + "loss": 0.2673, + "step": 34391 + }, + { + "epoch": 0.8062026406934018, + "grad_norm": 4.106404781341553, + "learning_rate": 1.7975679663271495e-05, + "loss": 0.1945, + "step": 34392 + }, + { + "epoch": 0.8062260822682068, + "grad_norm": 2.6696937084198, + "learning_rate": 1.797146633350577e-05, + "loss": 0.1966, + "step": 34393 + }, + { + "epoch": 0.8062495238430117, + "grad_norm": 3.192652463912964, + "learning_rate": 1.796725344882917e-05, + "loss": 0.2346, + "step": 34394 + }, + { + "epoch": 0.8062729654178168, + "grad_norm": 1.2179515361785889, + "learning_rate": 1.7963041009264568e-05, + "loss": 0.0744, + "step": 34395 + }, + { + "epoch": 0.8062964069926217, + "grad_norm": 1.621469259262085, + "learning_rate": 1.795882901483481e-05, + "loss": 0.1007, + "step": 34396 + }, + { + "epoch": 0.8063198485674268, + "grad_norm": 2.5090842247009277, + "learning_rate": 1.795461746556275e-05, + "loss": 0.1789, + "step": 34397 + }, + { + "epoch": 0.8063432901422317, + "grad_norm": 3.5745112895965576, + "learning_rate": 1.795040636147123e-05, + "loss": 0.1963, + "step": 34398 + }, + { + "epoch": 0.8063667317170368, + "grad_norm": 1.645925521850586, + "learning_rate": 1.79461957025831e-05, + "loss": 0.104, + "step": 34399 + }, + { + "epoch": 0.8063901732918417, + "grad_norm": 3.2633745670318604, + "learning_rate": 1.7941985488921242e-05, + "loss": 0.2135, + "step": 34400 + }, + { + "epoch": 0.8064136148666468, + "grad_norm": 3.9817521572113037, + "learning_rate": 1.7937775720508486e-05, + "loss": 0.1917, + "step": 34401 + }, + { + "epoch": 0.8064370564414517, + "grad_norm": 1.799182415008545, + "learning_rate": 1.7933566397367663e-05, + "loss": 0.0992, + "step": 34402 + }, + { + "epoch": 0.8064604980162567, + "grad_norm": 1.3555361032485962, + "learning_rate": 1.7929357519521618e-05, + "loss": 0.1626, + "step": 34403 + }, + { + "epoch": 0.8064839395910617, + "grad_norm": 0.6178498268127441, + "learning_rate": 1.7925149086993187e-05, + "loss": 0.0599, + "step": 34404 + }, + { + "epoch": 0.8065073811658667, + "grad_norm": 2.743109941482544, + "learning_rate": 1.7920941099805176e-05, + "loss": 0.1844, + "step": 34405 + }, + { + "epoch": 0.8065308227406717, + "grad_norm": 3.2932989597320557, + "learning_rate": 1.7916733557980468e-05, + "loss": 0.1754, + "step": 34406 + }, + { + "epoch": 0.8065542643154767, + "grad_norm": 3.114497423171997, + "learning_rate": 1.791252646154188e-05, + "loss": 0.2637, + "step": 34407 + }, + { + "epoch": 0.8065777058902817, + "grad_norm": 2.3512144088745117, + "learning_rate": 1.7908319810512232e-05, + "loss": 1.055, + "step": 34408 + }, + { + "epoch": 0.8066011474650867, + "grad_norm": 1.9473706483840942, + "learning_rate": 1.790411360491434e-05, + "loss": 0.1061, + "step": 34409 + }, + { + "epoch": 0.8066245890398916, + "grad_norm": 3.4464187622070312, + "learning_rate": 1.7899907844771048e-05, + "loss": 0.1659, + "step": 34410 + }, + { + "epoch": 0.8066480306146967, + "grad_norm": 2.3804190158843994, + "learning_rate": 1.789570253010513e-05, + "loss": 0.2344, + "step": 34411 + }, + { + "epoch": 0.8066714721895016, + "grad_norm": 2.4118576049804688, + "learning_rate": 1.7891497660939438e-05, + "loss": 0.1433, + "step": 34412 + }, + { + "epoch": 0.8066949137643067, + "grad_norm": 3.1814098358154297, + "learning_rate": 1.788729323729682e-05, + "loss": 0.2272, + "step": 34413 + }, + { + "epoch": 0.8067183553391117, + "grad_norm": 3.635568618774414, + "learning_rate": 1.788308925920006e-05, + "loss": 0.2815, + "step": 34414 + }, + { + "epoch": 0.8067417969139167, + "grad_norm": 2.9153034687042236, + "learning_rate": 1.7878885726671956e-05, + "loss": 0.2246, + "step": 34415 + }, + { + "epoch": 0.8067652384887217, + "grad_norm": 1.4099106788635254, + "learning_rate": 1.787468263973533e-05, + "loss": 0.4956, + "step": 34416 + }, + { + "epoch": 0.8067886800635267, + "grad_norm": 0.7454961538314819, + "learning_rate": 1.7870479998412992e-05, + "loss": 0.042, + "step": 34417 + }, + { + "epoch": 0.8068121216383317, + "grad_norm": 5.066442489624023, + "learning_rate": 1.7866277802727706e-05, + "loss": 0.4369, + "step": 34418 + }, + { + "epoch": 0.8068355632131367, + "grad_norm": 0.6063239574432373, + "learning_rate": 1.7862076052702336e-05, + "loss": 0.0364, + "step": 34419 + }, + { + "epoch": 0.8068590047879417, + "grad_norm": 2.3502349853515625, + "learning_rate": 1.785787474835965e-05, + "loss": 0.2158, + "step": 34420 + }, + { + "epoch": 0.8068824463627466, + "grad_norm": 2.0105392932891846, + "learning_rate": 1.7853673889722443e-05, + "loss": 0.0983, + "step": 34421 + }, + { + "epoch": 0.8069058879375517, + "grad_norm": 0.7744452953338623, + "learning_rate": 1.7849473476813517e-05, + "loss": 0.0755, + "step": 34422 + }, + { + "epoch": 0.8069293295123566, + "grad_norm": 3.8216371536254883, + "learning_rate": 1.7845273509655624e-05, + "loss": 0.2144, + "step": 34423 + }, + { + "epoch": 0.8069527710871617, + "grad_norm": 1.3959487676620483, + "learning_rate": 1.7841073988271616e-05, + "loss": 0.0363, + "step": 34424 + }, + { + "epoch": 0.8069762126619666, + "grad_norm": 2.724226713180542, + "learning_rate": 1.7836874912684232e-05, + "loss": 1.5154, + "step": 34425 + }, + { + "epoch": 0.8069996542367717, + "grad_norm": 3.2173686027526855, + "learning_rate": 1.7832676282916293e-05, + "loss": 0.1754, + "step": 34426 + }, + { + "epoch": 0.8070230958115766, + "grad_norm": 0.9405546188354492, + "learning_rate": 1.7828478098990566e-05, + "loss": 0.0571, + "step": 34427 + }, + { + "epoch": 0.8070465373863817, + "grad_norm": 2.996021270751953, + "learning_rate": 1.7824280360929822e-05, + "loss": 0.1993, + "step": 34428 + }, + { + "epoch": 0.8070699789611866, + "grad_norm": 4.165895462036133, + "learning_rate": 1.782008306875684e-05, + "loss": 0.5386, + "step": 34429 + }, + { + "epoch": 0.8070934205359916, + "grad_norm": 0.5136016607284546, + "learning_rate": 1.781588622249438e-05, + "loss": 0.0339, + "step": 34430 + }, + { + "epoch": 0.8071168621107966, + "grad_norm": 2.3005623817443848, + "learning_rate": 1.7811689822165257e-05, + "loss": 0.2768, + "step": 34431 + }, + { + "epoch": 0.8071403036856016, + "grad_norm": 1.114876627922058, + "learning_rate": 1.7807493867792223e-05, + "loss": 0.0817, + "step": 34432 + }, + { + "epoch": 0.8071637452604066, + "grad_norm": 2.4879565238952637, + "learning_rate": 1.7803298359398025e-05, + "loss": 0.2194, + "step": 34433 + }, + { + "epoch": 0.8071871868352116, + "grad_norm": 2.1223206520080566, + "learning_rate": 1.7799103297005447e-05, + "loss": 0.1903, + "step": 34434 + }, + { + "epoch": 0.8072106284100166, + "grad_norm": 0.3905327022075653, + "learning_rate": 1.7794908680637224e-05, + "loss": 0.0362, + "step": 34435 + }, + { + "epoch": 0.8072340699848216, + "grad_norm": 2.2818191051483154, + "learning_rate": 1.7790714510316165e-05, + "loss": 0.1197, + "step": 34436 + }, + { + "epoch": 0.8072575115596266, + "grad_norm": 0.7403072714805603, + "learning_rate": 1.7786520786064964e-05, + "loss": 0.06, + "step": 34437 + }, + { + "epoch": 0.8072809531344316, + "grad_norm": 2.6576356887817383, + "learning_rate": 1.778232750790645e-05, + "loss": 0.2121, + "step": 34438 + }, + { + "epoch": 0.8073043947092365, + "grad_norm": 2.7759671211242676, + "learning_rate": 1.7778134675863335e-05, + "loss": 0.1999, + "step": 34439 + }, + { + "epoch": 0.8073278362840416, + "grad_norm": 3.746539831161499, + "learning_rate": 1.7773942289958366e-05, + "loss": 0.6428, + "step": 34440 + }, + { + "epoch": 0.8073512778588465, + "grad_norm": 3.2927675247192383, + "learning_rate": 1.7769750350214297e-05, + "loss": 0.2768, + "step": 34441 + }, + { + "epoch": 0.8073747194336516, + "grad_norm": 3.6709580421447754, + "learning_rate": 1.7765558856653886e-05, + "loss": 0.4187, + "step": 34442 + }, + { + "epoch": 0.8073981610084565, + "grad_norm": 2.367940664291382, + "learning_rate": 1.7761367809299834e-05, + "loss": 1.3757, + "step": 34443 + }, + { + "epoch": 0.8074216025832616, + "grad_norm": 1.919431209564209, + "learning_rate": 1.7757177208174936e-05, + "loss": 1.366, + "step": 34444 + }, + { + "epoch": 0.8074450441580665, + "grad_norm": 2.7097527980804443, + "learning_rate": 1.775298705330192e-05, + "loss": 0.3052, + "step": 34445 + }, + { + "epoch": 0.8074684857328716, + "grad_norm": 0.4230891764163971, + "learning_rate": 1.7748797344703494e-05, + "loss": 0.0387, + "step": 34446 + }, + { + "epoch": 0.8074919273076765, + "grad_norm": 2.464277505874634, + "learning_rate": 1.7744608082402382e-05, + "loss": 0.1636, + "step": 34447 + }, + { + "epoch": 0.8075153688824815, + "grad_norm": 8.297163963317871, + "learning_rate": 1.774041926642137e-05, + "loss": 0.3007, + "step": 34448 + }, + { + "epoch": 0.8075388104572865, + "grad_norm": 1.1527551412582397, + "learning_rate": 1.7736230896783156e-05, + "loss": 0.7189, + "step": 34449 + }, + { + "epoch": 0.8075622520320915, + "grad_norm": 2.7115023136138916, + "learning_rate": 1.7732042973510434e-05, + "loss": 0.1916, + "step": 34450 + }, + { + "epoch": 0.8075856936068965, + "grad_norm": 3.3644566535949707, + "learning_rate": 1.772785549662599e-05, + "loss": 0.2641, + "step": 34451 + }, + { + "epoch": 0.8076091351817015, + "grad_norm": 5.697545051574707, + "learning_rate": 1.772366846615252e-05, + "loss": 0.45, + "step": 34452 + }, + { + "epoch": 0.8076325767565065, + "grad_norm": 3.49635648727417, + "learning_rate": 1.7719481882112732e-05, + "loss": 0.1795, + "step": 34453 + }, + { + "epoch": 0.8076560183313115, + "grad_norm": 3.6194584369659424, + "learning_rate": 1.771529574452935e-05, + "loss": 0.2124, + "step": 34454 + }, + { + "epoch": 0.8076794599061164, + "grad_norm": 2.611011028289795, + "learning_rate": 1.7711110053425083e-05, + "loss": 0.1933, + "step": 34455 + }, + { + "epoch": 0.8077029014809215, + "grad_norm": 2.0677502155303955, + "learning_rate": 1.7706924808822634e-05, + "loss": 0.1035, + "step": 34456 + }, + { + "epoch": 0.8077263430557264, + "grad_norm": 0.87849360704422, + "learning_rate": 1.7702740010744733e-05, + "loss": 0.0264, + "step": 34457 + }, + { + "epoch": 0.8077497846305315, + "grad_norm": 4.145693778991699, + "learning_rate": 1.7698555659214088e-05, + "loss": 0.2853, + "step": 34458 + }, + { + "epoch": 0.8077732262053364, + "grad_norm": 2.591273546218872, + "learning_rate": 1.7694371754253368e-05, + "loss": 0.1886, + "step": 34459 + }, + { + "epoch": 0.8077966677801415, + "grad_norm": 3.95898175239563, + "learning_rate": 1.7690188295885335e-05, + "loss": 0.3811, + "step": 34460 + }, + { + "epoch": 0.8078201093549464, + "grad_norm": 1.4859815835952759, + "learning_rate": 1.7686005284132645e-05, + "loss": 0.1049, + "step": 34461 + }, + { + "epoch": 0.8078435509297515, + "grad_norm": 4.082418918609619, + "learning_rate": 1.7681822719017982e-05, + "loss": 0.5111, + "step": 34462 + }, + { + "epoch": 0.8078669925045564, + "grad_norm": 3.1591837406158447, + "learning_rate": 1.7677640600564084e-05, + "loss": 0.1913, + "step": 34463 + }, + { + "epoch": 0.8078904340793615, + "grad_norm": 3.865490674972534, + "learning_rate": 1.767345892879363e-05, + "loss": 0.2778, + "step": 34464 + }, + { + "epoch": 0.8079138756541665, + "grad_norm": 5.053914546966553, + "learning_rate": 1.7669277703729302e-05, + "loss": 0.288, + "step": 34465 + }, + { + "epoch": 0.8079373172289714, + "grad_norm": 2.4514262676239014, + "learning_rate": 1.766509692539379e-05, + "loss": 0.1898, + "step": 34466 + }, + { + "epoch": 0.8079607588037765, + "grad_norm": 3.2163894176483154, + "learning_rate": 1.7660916593809764e-05, + "loss": 0.3325, + "step": 34467 + }, + { + "epoch": 0.8079842003785814, + "grad_norm": 0.9039844870567322, + "learning_rate": 1.76567367089999e-05, + "loss": 0.0715, + "step": 34468 + }, + { + "epoch": 0.8080076419533865, + "grad_norm": 0.6606438755989075, + "learning_rate": 1.7652557270986934e-05, + "loss": 0.0413, + "step": 34469 + }, + { + "epoch": 0.8080310835281914, + "grad_norm": 0.8051503896713257, + "learning_rate": 1.7648378279793497e-05, + "loss": 0.0485, + "step": 34470 + }, + { + "epoch": 0.8080545251029965, + "grad_norm": 2.000697374343872, + "learning_rate": 1.764419973544226e-05, + "loss": 0.1263, + "step": 34471 + }, + { + "epoch": 0.8080779666778014, + "grad_norm": 3.119084119796753, + "learning_rate": 1.7640021637955917e-05, + "loss": 0.4242, + "step": 34472 + }, + { + "epoch": 0.8081014082526065, + "grad_norm": 0.9854485988616943, + "learning_rate": 1.7635843987357147e-05, + "loss": 0.0309, + "step": 34473 + }, + { + "epoch": 0.8081248498274114, + "grad_norm": 2.823572874069214, + "learning_rate": 1.7631666783668597e-05, + "loss": 0.144, + "step": 34474 + }, + { + "epoch": 0.8081482914022164, + "grad_norm": 2.726482629776001, + "learning_rate": 1.762749002691292e-05, + "loss": 0.2145, + "step": 34475 + }, + { + "epoch": 0.8081717329770214, + "grad_norm": 9.256189346313477, + "learning_rate": 1.762331371711282e-05, + "loss": 0.2405, + "step": 34476 + }, + { + "epoch": 0.8081951745518264, + "grad_norm": 2.2805230617523193, + "learning_rate": 1.7619137854290935e-05, + "loss": 0.1663, + "step": 34477 + }, + { + "epoch": 0.8082186161266314, + "grad_norm": 0.5619716048240662, + "learning_rate": 1.7614962438469928e-05, + "loss": 0.0636, + "step": 34478 + }, + { + "epoch": 0.8082420577014364, + "grad_norm": 2.230055570602417, + "learning_rate": 1.7610787469672453e-05, + "loss": 0.1033, + "step": 34479 + }, + { + "epoch": 0.8082654992762414, + "grad_norm": 3.1656980514526367, + "learning_rate": 1.7606612947921152e-05, + "loss": 0.1707, + "step": 34480 + }, + { + "epoch": 0.8082889408510464, + "grad_norm": 1.9479849338531494, + "learning_rate": 1.760243887323867e-05, + "loss": 0.1245, + "step": 34481 + }, + { + "epoch": 0.8083123824258514, + "grad_norm": 2.655210494995117, + "learning_rate": 1.7598265245647694e-05, + "loss": 0.92, + "step": 34482 + }, + { + "epoch": 0.8083358240006564, + "grad_norm": 1.020185947418213, + "learning_rate": 1.759409206517082e-05, + "loss": 0.0748, + "step": 34483 + }, + { + "epoch": 0.8083592655754613, + "grad_norm": 1.8487637042999268, + "learning_rate": 1.7589919331830752e-05, + "loss": 0.1108, + "step": 34484 + }, + { + "epoch": 0.8083827071502664, + "grad_norm": 3.248054027557373, + "learning_rate": 1.758574704565009e-05, + "loss": 0.2821, + "step": 34485 + }, + { + "epoch": 0.8084061487250713, + "grad_norm": 3.0345237255096436, + "learning_rate": 1.758157520665149e-05, + "loss": 0.2977, + "step": 34486 + }, + { + "epoch": 0.8084295902998764, + "grad_norm": 2.584644079208374, + "learning_rate": 1.7577403814857574e-05, + "loss": 0.2027, + "step": 34487 + }, + { + "epoch": 0.8084530318746813, + "grad_norm": 4.273730754852295, + "learning_rate": 1.7573232870290958e-05, + "loss": 0.3617, + "step": 34488 + }, + { + "epoch": 0.8084764734494864, + "grad_norm": 3.115324020385742, + "learning_rate": 1.756906237297432e-05, + "loss": 0.1149, + "step": 34489 + }, + { + "epoch": 0.8084999150242913, + "grad_norm": 1.6275453567504883, + "learning_rate": 1.756489232293027e-05, + "loss": 1.0229, + "step": 34490 + }, + { + "epoch": 0.8085233565990964, + "grad_norm": 2.882014751434326, + "learning_rate": 1.7560722720181445e-05, + "loss": 1.5529, + "step": 34491 + }, + { + "epoch": 0.8085467981739013, + "grad_norm": 3.1867880821228027, + "learning_rate": 1.7556553564750444e-05, + "loss": 0.3349, + "step": 34492 + }, + { + "epoch": 0.8085702397487063, + "grad_norm": 3.0176963806152344, + "learning_rate": 1.7552384856659898e-05, + "loss": 0.1792, + "step": 34493 + }, + { + "epoch": 0.8085936813235113, + "grad_norm": 6.1538286209106445, + "learning_rate": 1.7548216595932408e-05, + "loss": 0.4672, + "step": 34494 + }, + { + "epoch": 0.8086171228983163, + "grad_norm": 1.6759886741638184, + "learning_rate": 1.7544048782590638e-05, + "loss": 0.0866, + "step": 34495 + }, + { + "epoch": 0.8086405644731213, + "grad_norm": 2.6074435710906982, + "learning_rate": 1.7539881416657156e-05, + "loss": 0.1518, + "step": 34496 + }, + { + "epoch": 0.8086640060479263, + "grad_norm": 1.6577318906784058, + "learning_rate": 1.7535714498154622e-05, + "loss": 0.1765, + "step": 34497 + }, + { + "epoch": 0.8086874476227313, + "grad_norm": 0.9972710609436035, + "learning_rate": 1.7531548027105614e-05, + "loss": 0.0794, + "step": 34498 + }, + { + "epoch": 0.8087108891975363, + "grad_norm": 2.4955625534057617, + "learning_rate": 1.7527382003532756e-05, + "loss": 0.0928, + "step": 34499 + }, + { + "epoch": 0.8087343307723412, + "grad_norm": 1.2155104875564575, + "learning_rate": 1.752321642745861e-05, + "loss": 0.0833, + "step": 34500 + }, + { + "epoch": 0.8087577723471463, + "grad_norm": 1.880984902381897, + "learning_rate": 1.751905129890583e-05, + "loss": 0.1016, + "step": 34501 + }, + { + "epoch": 0.8087812139219512, + "grad_norm": 2.758096694946289, + "learning_rate": 1.751488661789701e-05, + "loss": 0.1982, + "step": 34502 + }, + { + "epoch": 0.8088046554967563, + "grad_norm": 1.8462848663330078, + "learning_rate": 1.7510722384454724e-05, + "loss": 0.152, + "step": 34503 + }, + { + "epoch": 0.8088280970715612, + "grad_norm": 2.2201807498931885, + "learning_rate": 1.7506558598601573e-05, + "loss": 0.1117, + "step": 34504 + }, + { + "epoch": 0.8088515386463663, + "grad_norm": 0.4384317696094513, + "learning_rate": 1.750239526036016e-05, + "loss": 0.0342, + "step": 34505 + }, + { + "epoch": 0.8088749802211712, + "grad_norm": 1.2886512279510498, + "learning_rate": 1.7498232369753055e-05, + "loss": 0.0843, + "step": 34506 + }, + { + "epoch": 0.8088984217959763, + "grad_norm": 3.550023317337036, + "learning_rate": 1.749406992680287e-05, + "loss": 0.4889, + "step": 34507 + }, + { + "epoch": 0.8089218633707812, + "grad_norm": 3.012341022491455, + "learning_rate": 1.7489907931532167e-05, + "loss": 0.1946, + "step": 34508 + }, + { + "epoch": 0.8089453049455863, + "grad_norm": 2.8899171352386475, + "learning_rate": 1.7485746383963562e-05, + "loss": 0.252, + "step": 34509 + }, + { + "epoch": 0.8089687465203912, + "grad_norm": 3.355445146560669, + "learning_rate": 1.748158528411963e-05, + "loss": 0.279, + "step": 34510 + }, + { + "epoch": 0.8089921880951962, + "grad_norm": 3.228945016860962, + "learning_rate": 1.747742463202292e-05, + "loss": 0.5828, + "step": 34511 + }, + { + "epoch": 0.8090156296700012, + "grad_norm": 0.6376218199729919, + "learning_rate": 1.7473264427696035e-05, + "loss": 0.1413, + "step": 34512 + }, + { + "epoch": 0.8090390712448062, + "grad_norm": 0.8536211252212524, + "learning_rate": 1.7469104671161507e-05, + "loss": 0.0549, + "step": 34513 + }, + { + "epoch": 0.8090625128196112, + "grad_norm": 0.8563997149467468, + "learning_rate": 1.7464945362441965e-05, + "loss": 0.0502, + "step": 34514 + }, + { + "epoch": 0.8090859543944162, + "grad_norm": 2.9956157207489014, + "learning_rate": 1.7460786501559957e-05, + "loss": 0.2356, + "step": 34515 + }, + { + "epoch": 0.8091093959692212, + "grad_norm": 2.0892527103424072, + "learning_rate": 1.745662808853803e-05, + "loss": 0.138, + "step": 34516 + }, + { + "epoch": 0.8091328375440262, + "grad_norm": 3.0180554389953613, + "learning_rate": 1.745247012339877e-05, + "loss": 0.3428, + "step": 34517 + }, + { + "epoch": 0.8091562791188313, + "grad_norm": 1.7655658721923828, + "learning_rate": 1.744831260616472e-05, + "loss": 0.1472, + "step": 34518 + }, + { + "epoch": 0.8091797206936362, + "grad_norm": 3.0266692638397217, + "learning_rate": 1.7444155536858432e-05, + "loss": 0.2684, + "step": 34519 + }, + { + "epoch": 0.8092031622684412, + "grad_norm": 2.939866065979004, + "learning_rate": 1.7439998915502466e-05, + "loss": 0.3736, + "step": 34520 + }, + { + "epoch": 0.8092266038432462, + "grad_norm": 0.45471063256263733, + "learning_rate": 1.7435842742119414e-05, + "loss": 0.05, + "step": 34521 + }, + { + "epoch": 0.8092500454180512, + "grad_norm": 1.3673522472381592, + "learning_rate": 1.74316870167318e-05, + "loss": 0.0636, + "step": 34522 + }, + { + "epoch": 0.8092734869928562, + "grad_norm": 1.2169421911239624, + "learning_rate": 1.7427531739362178e-05, + "loss": 0.0726, + "step": 34523 + }, + { + "epoch": 0.8092969285676612, + "grad_norm": 0.6692149639129639, + "learning_rate": 1.7423376910033097e-05, + "loss": 0.0423, + "step": 34524 + }, + { + "epoch": 0.8093203701424662, + "grad_norm": 2.3120758533477783, + "learning_rate": 1.7419222528767088e-05, + "loss": 0.2275, + "step": 34525 + }, + { + "epoch": 0.8093438117172712, + "grad_norm": 2.8982057571411133, + "learning_rate": 1.741506859558667e-05, + "loss": 0.1079, + "step": 34526 + }, + { + "epoch": 0.8093672532920761, + "grad_norm": 4.508543968200684, + "learning_rate": 1.7410915110514446e-05, + "loss": 0.242, + "step": 34527 + }, + { + "epoch": 0.8093906948668812, + "grad_norm": 1.4774762392044067, + "learning_rate": 1.7406762073572913e-05, + "loss": 0.1063, + "step": 34528 + }, + { + "epoch": 0.8094141364416861, + "grad_norm": 2.649664878845215, + "learning_rate": 1.7402609484784616e-05, + "loss": 0.2364, + "step": 34529 + }, + { + "epoch": 0.8094375780164912, + "grad_norm": 2.9027445316314697, + "learning_rate": 1.739845734417207e-05, + "loss": 1.4722, + "step": 34530 + }, + { + "epoch": 0.8094610195912961, + "grad_norm": 0.774142861366272, + "learning_rate": 1.7394305651757803e-05, + "loss": 0.0279, + "step": 34531 + }, + { + "epoch": 0.8094844611661012, + "grad_norm": 0.6753091216087341, + "learning_rate": 1.7390154407564386e-05, + "loss": 0.031, + "step": 34532 + }, + { + "epoch": 0.8095079027409061, + "grad_norm": 2.8743951320648193, + "learning_rate": 1.7386003611614276e-05, + "loss": 0.1335, + "step": 34533 + }, + { + "epoch": 0.8095313443157112, + "grad_norm": 2.026810884475708, + "learning_rate": 1.738185326393007e-05, + "loss": 0.1479, + "step": 34534 + }, + { + "epoch": 0.8095547858905161, + "grad_norm": 2.3720877170562744, + "learning_rate": 1.737770336453424e-05, + "loss": 0.3004, + "step": 34535 + }, + { + "epoch": 0.8095782274653212, + "grad_norm": 5.75503396987915, + "learning_rate": 1.7373553913449326e-05, + "loss": 0.4118, + "step": 34536 + }, + { + "epoch": 0.8096016690401261, + "grad_norm": 3.4455089569091797, + "learning_rate": 1.7369404910697818e-05, + "loss": 0.2283, + "step": 34537 + }, + { + "epoch": 0.8096251106149311, + "grad_norm": 2.4635801315307617, + "learning_rate": 1.7365256356302228e-05, + "loss": 0.2216, + "step": 34538 + }, + { + "epoch": 0.8096485521897361, + "grad_norm": 2.3728816509246826, + "learning_rate": 1.7361108250285108e-05, + "loss": 0.1894, + "step": 34539 + }, + { + "epoch": 0.8096719937645411, + "grad_norm": 0.5247290730476379, + "learning_rate": 1.7356960592668923e-05, + "loss": 0.0253, + "step": 34540 + }, + { + "epoch": 0.8096954353393461, + "grad_norm": 3.538465738296509, + "learning_rate": 1.7352813383476206e-05, + "loss": 0.2053, + "step": 34541 + }, + { + "epoch": 0.8097188769141511, + "grad_norm": 3.4667885303497314, + "learning_rate": 1.7348666622729447e-05, + "loss": 0.2276, + "step": 34542 + }, + { + "epoch": 0.8097423184889561, + "grad_norm": 1.2304214239120483, + "learning_rate": 1.734452031045114e-05, + "loss": 0.1003, + "step": 34543 + }, + { + "epoch": 0.8097657600637611, + "grad_norm": 2.107741117477417, + "learning_rate": 1.7340374446663776e-05, + "loss": 0.2088, + "step": 34544 + }, + { + "epoch": 0.809789201638566, + "grad_norm": 1.603477954864502, + "learning_rate": 1.733622903138986e-05, + "loss": 0.079, + "step": 34545 + }, + { + "epoch": 0.8098126432133711, + "grad_norm": 3.118093252182007, + "learning_rate": 1.733208406465191e-05, + "loss": 0.1322, + "step": 34546 + }, + { + "epoch": 0.809836084788176, + "grad_norm": 2.0664048194885254, + "learning_rate": 1.7327939546472393e-05, + "loss": 0.1188, + "step": 34547 + }, + { + "epoch": 0.8098595263629811, + "grad_norm": 1.2470123767852783, + "learning_rate": 1.7323795476873818e-05, + "loss": 0.0435, + "step": 34548 + }, + { + "epoch": 0.809882967937786, + "grad_norm": 0.7943102717399597, + "learning_rate": 1.7319651855878636e-05, + "loss": 0.0709, + "step": 34549 + }, + { + "epoch": 0.8099064095125911, + "grad_norm": 2.027867078781128, + "learning_rate": 1.7315508683509353e-05, + "loss": 0.0669, + "step": 34550 + }, + { + "epoch": 0.809929851087396, + "grad_norm": 1.1416316032409668, + "learning_rate": 1.7311365959788417e-05, + "loss": 0.108, + "step": 34551 + }, + { + "epoch": 0.8099532926622011, + "grad_norm": 2.5283219814300537, + "learning_rate": 1.730722368473837e-05, + "loss": 0.1539, + "step": 34552 + }, + { + "epoch": 0.809976734237006, + "grad_norm": 3.0591788291931152, + "learning_rate": 1.7303081858381642e-05, + "loss": 0.3983, + "step": 34553 + }, + { + "epoch": 0.810000175811811, + "grad_norm": 0.7114490270614624, + "learning_rate": 1.7298940480740723e-05, + "loss": 0.0749, + "step": 34554 + }, + { + "epoch": 0.810023617386616, + "grad_norm": 2.242640256881714, + "learning_rate": 1.729479955183807e-05, + "loss": 1.3396, + "step": 34555 + }, + { + "epoch": 0.810047058961421, + "grad_norm": 0.35576915740966797, + "learning_rate": 1.7290659071696146e-05, + "loss": 0.0216, + "step": 34556 + }, + { + "epoch": 0.810070500536226, + "grad_norm": 2.820751190185547, + "learning_rate": 1.7286519040337447e-05, + "loss": 0.2421, + "step": 34557 + }, + { + "epoch": 0.810093942111031, + "grad_norm": 1.4636112451553345, + "learning_rate": 1.7282379457784403e-05, + "loss": 0.0809, + "step": 34558 + }, + { + "epoch": 0.810117383685836, + "grad_norm": 2.1221795082092285, + "learning_rate": 1.7278240324059514e-05, + "loss": 0.0696, + "step": 34559 + }, + { + "epoch": 0.810140825260641, + "grad_norm": 2.5532162189483643, + "learning_rate": 1.7274101639185225e-05, + "loss": 0.1784, + "step": 34560 + }, + { + "epoch": 0.810164266835446, + "grad_norm": 3.015007257461548, + "learning_rate": 1.7269963403183974e-05, + "loss": 0.2437, + "step": 34561 + }, + { + "epoch": 0.810187708410251, + "grad_norm": 0.863984227180481, + "learning_rate": 1.7265825616078245e-05, + "loss": 0.0949, + "step": 34562 + }, + { + "epoch": 0.810211149985056, + "grad_norm": 1.984840750694275, + "learning_rate": 1.7261688277890465e-05, + "loss": 0.0529, + "step": 34563 + }, + { + "epoch": 0.810234591559861, + "grad_norm": 0.6440538167953491, + "learning_rate": 1.7257551388643055e-05, + "loss": 0.0404, + "step": 34564 + }, + { + "epoch": 0.8102580331346659, + "grad_norm": 3.206638813018799, + "learning_rate": 1.7253414948358537e-05, + "loss": 0.2432, + "step": 34565 + }, + { + "epoch": 0.810281474709471, + "grad_norm": 3.65364670753479, + "learning_rate": 1.7249278957059313e-05, + "loss": 0.1986, + "step": 34566 + }, + { + "epoch": 0.8103049162842759, + "grad_norm": 0.6818262338638306, + "learning_rate": 1.724514341476783e-05, + "loss": 0.067, + "step": 34567 + }, + { + "epoch": 0.810328357859081, + "grad_norm": 2.534550666809082, + "learning_rate": 1.7241008321506525e-05, + "loss": 1.1947, + "step": 34568 + }, + { + "epoch": 0.810351799433886, + "grad_norm": 0.8914284110069275, + "learning_rate": 1.7236873677297804e-05, + "loss": 0.0931, + "step": 34569 + }, + { + "epoch": 0.810375241008691, + "grad_norm": 1.8872673511505127, + "learning_rate": 1.7232739482164163e-05, + "loss": 0.2617, + "step": 34570 + }, + { + "epoch": 0.810398682583496, + "grad_norm": 0.5515681505203247, + "learning_rate": 1.722860573612798e-05, + "loss": 0.0623, + "step": 34571 + }, + { + "epoch": 0.810422124158301, + "grad_norm": 3.2934036254882812, + "learning_rate": 1.7224472439211738e-05, + "loss": 0.1786, + "step": 34572 + }, + { + "epoch": 0.810445565733106, + "grad_norm": 2.3209831714630127, + "learning_rate": 1.722033959143783e-05, + "loss": 0.1599, + "step": 34573 + }, + { + "epoch": 0.8104690073079109, + "grad_norm": 0.584356427192688, + "learning_rate": 1.7216207192828682e-05, + "loss": 0.0244, + "step": 34574 + }, + { + "epoch": 0.810492448882716, + "grad_norm": 1.105037808418274, + "learning_rate": 1.7212075243406734e-05, + "loss": 0.1141, + "step": 34575 + }, + { + "epoch": 0.8105158904575209, + "grad_norm": 0.5668594837188721, + "learning_rate": 1.7207943743194364e-05, + "loss": 0.0383, + "step": 34576 + }, + { + "epoch": 0.810539332032326, + "grad_norm": 1.965382695198059, + "learning_rate": 1.720381269221404e-05, + "loss": 0.0845, + "step": 34577 + }, + { + "epoch": 0.8105627736071309, + "grad_norm": 3.1109132766723633, + "learning_rate": 1.7199682090488155e-05, + "loss": 1.5133, + "step": 34578 + }, + { + "epoch": 0.810586215181936, + "grad_norm": 0.20913094282150269, + "learning_rate": 1.7195551938039134e-05, + "loss": 0.0236, + "step": 34579 + }, + { + "epoch": 0.8106096567567409, + "grad_norm": 2.3933730125427246, + "learning_rate": 1.7191422234889365e-05, + "loss": 0.2441, + "step": 34580 + }, + { + "epoch": 0.810633098331546, + "grad_norm": 0.3789898753166199, + "learning_rate": 1.7187292981061253e-05, + "loss": 0.0337, + "step": 34581 + }, + { + "epoch": 0.8106565399063509, + "grad_norm": 3.7179291248321533, + "learning_rate": 1.7183164176577238e-05, + "loss": 0.3877, + "step": 34582 + }, + { + "epoch": 0.8106799814811559, + "grad_norm": 2.4124722480773926, + "learning_rate": 1.7179035821459676e-05, + "loss": 0.1724, + "step": 34583 + }, + { + "epoch": 0.8107034230559609, + "grad_norm": 1.75765061378479, + "learning_rate": 1.717490791573102e-05, + "loss": 0.0814, + "step": 34584 + }, + { + "epoch": 0.8107268646307659, + "grad_norm": 4.061883449554443, + "learning_rate": 1.717078045941364e-05, + "loss": 0.4548, + "step": 34585 + }, + { + "epoch": 0.8107503062055709, + "grad_norm": 3.0394887924194336, + "learning_rate": 1.7166653452529935e-05, + "loss": 0.2992, + "step": 34586 + }, + { + "epoch": 0.8107737477803759, + "grad_norm": 0.968662440776825, + "learning_rate": 1.7162526895102306e-05, + "loss": 0.115, + "step": 34587 + }, + { + "epoch": 0.8107971893551809, + "grad_norm": 2.5017080307006836, + "learning_rate": 1.7158400787153127e-05, + "loss": 0.2211, + "step": 34588 + }, + { + "epoch": 0.8108206309299859, + "grad_norm": 3.510833501815796, + "learning_rate": 1.7154275128704767e-05, + "loss": 0.3241, + "step": 34589 + }, + { + "epoch": 0.8108440725047908, + "grad_norm": 3.932678461074829, + "learning_rate": 1.7150149919779668e-05, + "loss": 0.4393, + "step": 34590 + }, + { + "epoch": 0.8108675140795959, + "grad_norm": 3.3180882930755615, + "learning_rate": 1.7146025160400193e-05, + "loss": 0.2815, + "step": 34591 + }, + { + "epoch": 0.8108909556544008, + "grad_norm": 4.734318256378174, + "learning_rate": 1.71419008505887e-05, + "loss": 0.4301, + "step": 34592 + }, + { + "epoch": 0.8109143972292059, + "grad_norm": 2.534115791320801, + "learning_rate": 1.7137776990367573e-05, + "loss": 0.2097, + "step": 34593 + }, + { + "epoch": 0.8109378388040108, + "grad_norm": 2.3198435306549072, + "learning_rate": 1.7133653579759212e-05, + "loss": 0.1169, + "step": 34594 + }, + { + "epoch": 0.8109612803788159, + "grad_norm": 2.247786283493042, + "learning_rate": 1.7129530618785982e-05, + "loss": 0.1975, + "step": 34595 + }, + { + "epoch": 0.8109847219536208, + "grad_norm": 3.234372854232788, + "learning_rate": 1.712540810747022e-05, + "loss": 0.3435, + "step": 34596 + }, + { + "epoch": 0.8110081635284259, + "grad_norm": 0.6028929948806763, + "learning_rate": 1.7121286045834352e-05, + "loss": 0.0325, + "step": 34597 + }, + { + "epoch": 0.8110316051032308, + "grad_norm": 3.7806148529052734, + "learning_rate": 1.7117164433900712e-05, + "loss": 0.2219, + "step": 34598 + }, + { + "epoch": 0.8110550466780359, + "grad_norm": 1.480395793914795, + "learning_rate": 1.7113043271691677e-05, + "loss": 0.0936, + "step": 34599 + }, + { + "epoch": 0.8110784882528408, + "grad_norm": 0.4441123306751251, + "learning_rate": 1.7108922559229588e-05, + "loss": 0.0275, + "step": 34600 + }, + { + "epoch": 0.8111019298276458, + "grad_norm": 2.788027048110962, + "learning_rate": 1.7104802296536827e-05, + "loss": 0.1741, + "step": 34601 + }, + { + "epoch": 0.8111253714024508, + "grad_norm": 1.4651176929473877, + "learning_rate": 1.710068248363571e-05, + "loss": 0.0755, + "step": 34602 + }, + { + "epoch": 0.8111488129772558, + "grad_norm": 2.5555849075317383, + "learning_rate": 1.709656312054865e-05, + "loss": 0.0974, + "step": 34603 + }, + { + "epoch": 0.8111722545520608, + "grad_norm": 2.585561752319336, + "learning_rate": 1.7092444207297964e-05, + "loss": 0.2267, + "step": 34604 + }, + { + "epoch": 0.8111956961268658, + "grad_norm": 0.5554454326629639, + "learning_rate": 1.7088325743905987e-05, + "loss": 0.0467, + "step": 34605 + }, + { + "epoch": 0.8112191377016708, + "grad_norm": 0.8741777539253235, + "learning_rate": 1.7084207730395095e-05, + "loss": 0.1086, + "step": 34606 + }, + { + "epoch": 0.8112425792764758, + "grad_norm": 4.22415018081665, + "learning_rate": 1.7080090166787644e-05, + "loss": 0.3435, + "step": 34607 + }, + { + "epoch": 0.8112660208512807, + "grad_norm": 1.533645510673523, + "learning_rate": 1.7075973053105944e-05, + "loss": 0.1581, + "step": 34608 + }, + { + "epoch": 0.8112894624260858, + "grad_norm": 3.1600089073181152, + "learning_rate": 1.707185638937232e-05, + "loss": 0.1916, + "step": 34609 + }, + { + "epoch": 0.8113129040008907, + "grad_norm": 2.5797863006591797, + "learning_rate": 1.706774017560917e-05, + "loss": 0.1566, + "step": 34610 + }, + { + "epoch": 0.8113363455756958, + "grad_norm": 2.768720865249634, + "learning_rate": 1.7063624411838786e-05, + "loss": 0.3189, + "step": 34611 + }, + { + "epoch": 0.8113597871505007, + "grad_norm": 2.9487810134887695, + "learning_rate": 1.7059509098083504e-05, + "loss": 0.2559, + "step": 34612 + }, + { + "epoch": 0.8113832287253058, + "grad_norm": 1.5327491760253906, + "learning_rate": 1.705539423436566e-05, + "loss": 0.0651, + "step": 34613 + }, + { + "epoch": 0.8114066703001107, + "grad_norm": 2.950949192047119, + "learning_rate": 1.7051279820707566e-05, + "loss": 0.266, + "step": 34614 + }, + { + "epoch": 0.8114301118749158, + "grad_norm": 0.9659404158592224, + "learning_rate": 1.7047165857131575e-05, + "loss": 0.072, + "step": 34615 + }, + { + "epoch": 0.8114535534497207, + "grad_norm": 1.1849371194839478, + "learning_rate": 1.7043052343659994e-05, + "loss": 0.1381, + "step": 34616 + }, + { + "epoch": 0.8114769950245257, + "grad_norm": 2.8743722438812256, + "learning_rate": 1.7038939280315112e-05, + "loss": 0.1331, + "step": 34617 + }, + { + "epoch": 0.8115004365993307, + "grad_norm": 0.9854826331138611, + "learning_rate": 1.703482666711932e-05, + "loss": 0.1063, + "step": 34618 + }, + { + "epoch": 0.8115238781741357, + "grad_norm": 3.0598084926605225, + "learning_rate": 1.703071450409488e-05, + "loss": 0.1518, + "step": 34619 + }, + { + "epoch": 0.8115473197489408, + "grad_norm": 0.8303956389427185, + "learning_rate": 1.702660279126411e-05, + "loss": 0.0611, + "step": 34620 + }, + { + "epoch": 0.8115707613237457, + "grad_norm": 2.0235178470611572, + "learning_rate": 1.702249152864931e-05, + "loss": 0.2019, + "step": 34621 + }, + { + "epoch": 0.8115942028985508, + "grad_norm": 2.7111988067626953, + "learning_rate": 1.701838071627283e-05, + "loss": 0.1744, + "step": 34622 + }, + { + "epoch": 0.8116176444733557, + "grad_norm": 1.8410921096801758, + "learning_rate": 1.7014270354156946e-05, + "loss": 0.0535, + "step": 34623 + }, + { + "epoch": 0.8116410860481608, + "grad_norm": 0.8948186635971069, + "learning_rate": 1.7010160442323975e-05, + "loss": 0.0677, + "step": 34624 + }, + { + "epoch": 0.8116645276229657, + "grad_norm": 2.1526377201080322, + "learning_rate": 1.7006050980796197e-05, + "loss": 0.1536, + "step": 34625 + }, + { + "epoch": 0.8116879691977708, + "grad_norm": 3.027015209197998, + "learning_rate": 1.7001941969595914e-05, + "loss": 0.2744, + "step": 34626 + }, + { + "epoch": 0.8117114107725757, + "grad_norm": 2.197097063064575, + "learning_rate": 1.699783340874541e-05, + "loss": 0.1363, + "step": 34627 + }, + { + "epoch": 0.8117348523473807, + "grad_norm": 2.0864343643188477, + "learning_rate": 1.6993725298267028e-05, + "loss": 1.407, + "step": 34628 + }, + { + "epoch": 0.8117582939221857, + "grad_norm": 1.8057806491851807, + "learning_rate": 1.6989617638182985e-05, + "loss": 0.142, + "step": 34629 + }, + { + "epoch": 0.8117817354969907, + "grad_norm": 4.884799003601074, + "learning_rate": 1.698551042851565e-05, + "loss": 0.386, + "step": 34630 + }, + { + "epoch": 0.8118051770717957, + "grad_norm": 2.6713221073150635, + "learning_rate": 1.6981403669287266e-05, + "loss": 0.2072, + "step": 34631 + }, + { + "epoch": 0.8118286186466007, + "grad_norm": 3.4234938621520996, + "learning_rate": 1.6977297360520115e-05, + "loss": 0.2433, + "step": 34632 + }, + { + "epoch": 0.8118520602214057, + "grad_norm": 0.8949642181396484, + "learning_rate": 1.6973191502236484e-05, + "loss": 0.0769, + "step": 34633 + }, + { + "epoch": 0.8118755017962107, + "grad_norm": 4.988978862762451, + "learning_rate": 1.696908609445863e-05, + "loss": 0.4675, + "step": 34634 + }, + { + "epoch": 0.8118989433710156, + "grad_norm": 2.4314498901367188, + "learning_rate": 1.6964981137208867e-05, + "loss": 1.3899, + "step": 34635 + }, + { + "epoch": 0.8119223849458207, + "grad_norm": 0.727978527545929, + "learning_rate": 1.696087663050946e-05, + "loss": 0.0689, + "step": 34636 + }, + { + "epoch": 0.8119458265206256, + "grad_norm": 7.87601375579834, + "learning_rate": 1.6956772574382674e-05, + "loss": 0.4092, + "step": 34637 + }, + { + "epoch": 0.8119692680954307, + "grad_norm": 3.3753254413604736, + "learning_rate": 1.695266896885076e-05, + "loss": 0.2556, + "step": 34638 + }, + { + "epoch": 0.8119927096702356, + "grad_norm": 1.6718720197677612, + "learning_rate": 1.6948565813936012e-05, + "loss": 0.136, + "step": 34639 + }, + { + "epoch": 0.8120161512450407, + "grad_norm": 1.9871562719345093, + "learning_rate": 1.6944463109660645e-05, + "loss": 1.2744, + "step": 34640 + }, + { + "epoch": 0.8120395928198456, + "grad_norm": 1.2102614641189575, + "learning_rate": 1.694036085604699e-05, + "loss": 0.1378, + "step": 34641 + }, + { + "epoch": 0.8120630343946507, + "grad_norm": 0.7185034155845642, + "learning_rate": 1.6936259053117253e-05, + "loss": 0.0349, + "step": 34642 + }, + { + "epoch": 0.8120864759694556, + "grad_norm": 0.9386242032051086, + "learning_rate": 1.6932157700893725e-05, + "loss": 0.049, + "step": 34643 + }, + { + "epoch": 0.8121099175442607, + "grad_norm": 2.4492135047912598, + "learning_rate": 1.692805679939865e-05, + "loss": 0.2628, + "step": 34644 + }, + { + "epoch": 0.8121333591190656, + "grad_norm": 2.83607816696167, + "learning_rate": 1.6923956348654267e-05, + "loss": 0.1581, + "step": 34645 + }, + { + "epoch": 0.8121568006938706, + "grad_norm": 2.8042242527008057, + "learning_rate": 1.6919856348682837e-05, + "loss": 0.2096, + "step": 34646 + }, + { + "epoch": 0.8121802422686756, + "grad_norm": 5.0270280838012695, + "learning_rate": 1.6915756799506577e-05, + "loss": 0.319, + "step": 34647 + }, + { + "epoch": 0.8122036838434806, + "grad_norm": 2.195796489715576, + "learning_rate": 1.6911657701147787e-05, + "loss": 1.6713, + "step": 34648 + }, + { + "epoch": 0.8122271254182856, + "grad_norm": 1.8645880222320557, + "learning_rate": 1.6907559053628673e-05, + "loss": 0.1943, + "step": 34649 + }, + { + "epoch": 0.8122505669930906, + "grad_norm": 0.6547631025314331, + "learning_rate": 1.6903460856971485e-05, + "loss": 0.0284, + "step": 34650 + }, + { + "epoch": 0.8122740085678956, + "grad_norm": 2.2305030822753906, + "learning_rate": 1.6899363111198453e-05, + "loss": 0.2422, + "step": 34651 + }, + { + "epoch": 0.8122974501427006, + "grad_norm": 2.7272896766662598, + "learning_rate": 1.6895265816331786e-05, + "loss": 0.2182, + "step": 34652 + }, + { + "epoch": 0.8123208917175055, + "grad_norm": 3.449474811553955, + "learning_rate": 1.6891168972393767e-05, + "loss": 0.2628, + "step": 34653 + }, + { + "epoch": 0.8123443332923106, + "grad_norm": 2.1495018005371094, + "learning_rate": 1.6887072579406582e-05, + "loss": 0.1821, + "step": 34654 + }, + { + "epoch": 0.8123677748671155, + "grad_norm": 0.8908755779266357, + "learning_rate": 1.68829766373925e-05, + "loss": 0.0524, + "step": 34655 + }, + { + "epoch": 0.8123912164419206, + "grad_norm": 0.4217323362827301, + "learning_rate": 1.6878881146373714e-05, + "loss": 0.0357, + "step": 34656 + }, + { + "epoch": 0.8124146580167255, + "grad_norm": 1.501766324043274, + "learning_rate": 1.6874786106372476e-05, + "loss": 0.0917, + "step": 34657 + }, + { + "epoch": 0.8124380995915306, + "grad_norm": 1.1951515674591064, + "learning_rate": 1.6870691517410973e-05, + "loss": 0.0607, + "step": 34658 + }, + { + "epoch": 0.8124615411663355, + "grad_norm": 0.8355004191398621, + "learning_rate": 1.6866597379511405e-05, + "loss": 0.0874, + "step": 34659 + }, + { + "epoch": 0.8124849827411406, + "grad_norm": 3.1708984375, + "learning_rate": 1.686250369269605e-05, + "loss": 0.2379, + "step": 34660 + }, + { + "epoch": 0.8125084243159455, + "grad_norm": 3.3711090087890625, + "learning_rate": 1.6858410456987094e-05, + "loss": 0.2874, + "step": 34661 + }, + { + "epoch": 0.8125318658907505, + "grad_norm": 4.382408142089844, + "learning_rate": 1.685431767240674e-05, + "loss": 0.391, + "step": 34662 + }, + { + "epoch": 0.8125553074655555, + "grad_norm": 3.938197374343872, + "learning_rate": 1.6850225338977187e-05, + "loss": 0.4588, + "step": 34663 + }, + { + "epoch": 0.8125787490403605, + "grad_norm": 1.418285608291626, + "learning_rate": 1.6846133456720648e-05, + "loss": 0.1289, + "step": 34664 + }, + { + "epoch": 0.8126021906151655, + "grad_norm": 3.1246118545532227, + "learning_rate": 1.6842042025659312e-05, + "loss": 0.1558, + "step": 34665 + }, + { + "epoch": 0.8126256321899705, + "grad_norm": 4.135168552398682, + "learning_rate": 1.683795104581538e-05, + "loss": 0.2565, + "step": 34666 + }, + { + "epoch": 0.8126490737647755, + "grad_norm": 2.8716440200805664, + "learning_rate": 1.683386051721111e-05, + "loss": 0.2678, + "step": 34667 + }, + { + "epoch": 0.8126725153395805, + "grad_norm": 2.652631998062134, + "learning_rate": 1.682977043986863e-05, + "loss": 0.1358, + "step": 34668 + }, + { + "epoch": 0.8126959569143855, + "grad_norm": 3.549220561981201, + "learning_rate": 1.6825680813810163e-05, + "loss": 0.2504, + "step": 34669 + }, + { + "epoch": 0.8127193984891905, + "grad_norm": 1.9787200689315796, + "learning_rate": 1.68215916390579e-05, + "loss": 1.2803, + "step": 34670 + }, + { + "epoch": 0.8127428400639956, + "grad_norm": 2.9280402660369873, + "learning_rate": 1.681750291563401e-05, + "loss": 0.2799, + "step": 34671 + }, + { + "epoch": 0.8127662816388005, + "grad_norm": 0.6975424289703369, + "learning_rate": 1.6813414643560653e-05, + "loss": 0.0675, + "step": 34672 + }, + { + "epoch": 0.8127897232136055, + "grad_norm": 5.401224613189697, + "learning_rate": 1.6809326822860083e-05, + "loss": 0.2285, + "step": 34673 + }, + { + "epoch": 0.8128131647884105, + "grad_norm": 4.441967964172363, + "learning_rate": 1.6805239453554444e-05, + "loss": 0.369, + "step": 34674 + }, + { + "epoch": 0.8128366063632155, + "grad_norm": 0.6131439805030823, + "learning_rate": 1.680115253566592e-05, + "loss": 0.0219, + "step": 34675 + }, + { + "epoch": 0.8128600479380205, + "grad_norm": 0.620658814907074, + "learning_rate": 1.6797066069216672e-05, + "loss": 0.0628, + "step": 34676 + }, + { + "epoch": 0.8128834895128255, + "grad_norm": 2.158149003982544, + "learning_rate": 1.679298005422889e-05, + "loss": 0.1848, + "step": 34677 + }, + { + "epoch": 0.8129069310876305, + "grad_norm": 2.8473453521728516, + "learning_rate": 1.6788894490724714e-05, + "loss": 0.1546, + "step": 34678 + }, + { + "epoch": 0.8129303726624355, + "grad_norm": 2.8163576126098633, + "learning_rate": 1.678480937872633e-05, + "loss": 0.2692, + "step": 34679 + }, + { + "epoch": 0.8129538142372404, + "grad_norm": 1.8236734867095947, + "learning_rate": 1.6780724718255935e-05, + "loss": 0.0949, + "step": 34680 + }, + { + "epoch": 0.8129772558120455, + "grad_norm": 1.874855637550354, + "learning_rate": 1.6776640509335674e-05, + "loss": 0.208, + "step": 34681 + }, + { + "epoch": 0.8130006973868504, + "grad_norm": 3.286776304244995, + "learning_rate": 1.67725567519877e-05, + "loss": 0.2829, + "step": 34682 + }, + { + "epoch": 0.8130241389616555, + "grad_norm": 2.296274423599243, + "learning_rate": 1.6768473446234168e-05, + "loss": 1.6539, + "step": 34683 + }, + { + "epoch": 0.8130475805364604, + "grad_norm": 0.44774219393730164, + "learning_rate": 1.6764390592097244e-05, + "loss": 0.034, + "step": 34684 + }, + { + "epoch": 0.8130710221112655, + "grad_norm": 1.1474041938781738, + "learning_rate": 1.676030818959905e-05, + "loss": 0.1121, + "step": 34685 + }, + { + "epoch": 0.8130944636860704, + "grad_norm": 1.0608575344085693, + "learning_rate": 1.6756226238761798e-05, + "loss": 0.0603, + "step": 34686 + }, + { + "epoch": 0.8131179052608755, + "grad_norm": 4.147610664367676, + "learning_rate": 1.675214473960759e-05, + "loss": 0.1795, + "step": 34687 + }, + { + "epoch": 0.8131413468356804, + "grad_norm": 0.8851705193519592, + "learning_rate": 1.674806369215859e-05, + "loss": 0.0197, + "step": 34688 + }, + { + "epoch": 0.8131647884104855, + "grad_norm": 2.387847900390625, + "learning_rate": 1.6743983096436945e-05, + "loss": 0.1894, + "step": 34689 + }, + { + "epoch": 0.8131882299852904, + "grad_norm": 0.4975351393222809, + "learning_rate": 1.673990295246475e-05, + "loss": 0.0448, + "step": 34690 + }, + { + "epoch": 0.8132116715600954, + "grad_norm": 1.9966626167297363, + "learning_rate": 1.6735823260264215e-05, + "loss": 0.1773, + "step": 34691 + }, + { + "epoch": 0.8132351131349004, + "grad_norm": 3.92427921295166, + "learning_rate": 1.6731744019857422e-05, + "loss": 0.3502, + "step": 34692 + }, + { + "epoch": 0.8132585547097054, + "grad_norm": 2.7704408168792725, + "learning_rate": 1.6727665231266544e-05, + "loss": 0.1531, + "step": 34693 + }, + { + "epoch": 0.8132819962845104, + "grad_norm": 1.626943588256836, + "learning_rate": 1.67235868945137e-05, + "loss": 0.1474, + "step": 34694 + }, + { + "epoch": 0.8133054378593154, + "grad_norm": 2.3979928493499756, + "learning_rate": 1.6719509009621016e-05, + "loss": 0.1584, + "step": 34695 + }, + { + "epoch": 0.8133288794341204, + "grad_norm": 4.854002475738525, + "learning_rate": 1.671543157661062e-05, + "loss": 0.3978, + "step": 34696 + }, + { + "epoch": 0.8133523210089254, + "grad_norm": 4.610321521759033, + "learning_rate": 1.6711354595504604e-05, + "loss": 0.3843, + "step": 34697 + }, + { + "epoch": 0.8133757625837303, + "grad_norm": 3.122431755065918, + "learning_rate": 1.6707278066325148e-05, + "loss": 0.284, + "step": 34698 + }, + { + "epoch": 0.8133992041585354, + "grad_norm": 0.7405994534492493, + "learning_rate": 1.670320198909434e-05, + "loss": 0.0512, + "step": 34699 + }, + { + "epoch": 0.8134226457333403, + "grad_norm": 3.9318528175354004, + "learning_rate": 1.6699126363834307e-05, + "loss": 0.4653, + "step": 34700 + }, + { + "epoch": 0.8134460873081454, + "grad_norm": 0.7377188801765442, + "learning_rate": 1.6695051190567144e-05, + "loss": 0.0531, + "step": 34701 + }, + { + "epoch": 0.8134695288829503, + "grad_norm": 0.21377786993980408, + "learning_rate": 1.6690976469314967e-05, + "loss": 0.0128, + "step": 34702 + }, + { + "epoch": 0.8134929704577554, + "grad_norm": 2.4148263931274414, + "learning_rate": 1.668690220009991e-05, + "loss": 0.2148, + "step": 34703 + }, + { + "epoch": 0.8135164120325603, + "grad_norm": 0.5659083724021912, + "learning_rate": 1.6682828382944038e-05, + "loss": 0.0335, + "step": 34704 + }, + { + "epoch": 0.8135398536073654, + "grad_norm": 2.427064895629883, + "learning_rate": 1.6678755017869506e-05, + "loss": 1.5368, + "step": 34705 + }, + { + "epoch": 0.8135632951821703, + "grad_norm": 4.336215972900391, + "learning_rate": 1.66746821048984e-05, + "loss": 0.2592, + "step": 34706 + }, + { + "epoch": 0.8135867367569753, + "grad_norm": 3.1688623428344727, + "learning_rate": 1.6670609644052805e-05, + "loss": 0.1881, + "step": 34707 + }, + { + "epoch": 0.8136101783317803, + "grad_norm": 1.1411188840866089, + "learning_rate": 1.6666537635354818e-05, + "loss": 0.1129, + "step": 34708 + }, + { + "epoch": 0.8136336199065853, + "grad_norm": 3.194761037826538, + "learning_rate": 1.6662466078826555e-05, + "loss": 0.3347, + "step": 34709 + }, + { + "epoch": 0.8136570614813903, + "grad_norm": 1.6709237098693848, + "learning_rate": 1.6658394974490067e-05, + "loss": 0.0704, + "step": 34710 + }, + { + "epoch": 0.8136805030561953, + "grad_norm": 2.4691967964172363, + "learning_rate": 1.665432432236749e-05, + "loss": 0.1185, + "step": 34711 + }, + { + "epoch": 0.8137039446310003, + "grad_norm": 2.4719223976135254, + "learning_rate": 1.665025412248089e-05, + "loss": 0.1672, + "step": 34712 + }, + { + "epoch": 0.8137273862058053, + "grad_norm": 2.8920300006866455, + "learning_rate": 1.664618437485237e-05, + "loss": 0.258, + "step": 34713 + }, + { + "epoch": 0.8137508277806103, + "grad_norm": 2.9798405170440674, + "learning_rate": 1.664211507950396e-05, + "loss": 0.1495, + "step": 34714 + }, + { + "epoch": 0.8137742693554153, + "grad_norm": 4.084819793701172, + "learning_rate": 1.6638046236457804e-05, + "loss": 0.4003, + "step": 34715 + }, + { + "epoch": 0.8137977109302202, + "grad_norm": 2.369187355041504, + "learning_rate": 1.6633977845735947e-05, + "loss": 0.1478, + "step": 34716 + }, + { + "epoch": 0.8138211525050253, + "grad_norm": 3.3344666957855225, + "learning_rate": 1.6629909907360463e-05, + "loss": 0.2565, + "step": 34717 + }, + { + "epoch": 0.8138445940798302, + "grad_norm": 4.517098903656006, + "learning_rate": 1.6625842421353444e-05, + "loss": 0.3479, + "step": 34718 + }, + { + "epoch": 0.8138680356546353, + "grad_norm": 4.455656051635742, + "learning_rate": 1.6621775387736947e-05, + "loss": 0.3936, + "step": 34719 + }, + { + "epoch": 0.8138914772294402, + "grad_norm": 2.29917573928833, + "learning_rate": 1.661770880653305e-05, + "loss": 0.2132, + "step": 34720 + }, + { + "epoch": 0.8139149188042453, + "grad_norm": 3.515993118286133, + "learning_rate": 1.6613642677763795e-05, + "loss": 0.1975, + "step": 34721 + }, + { + "epoch": 0.8139383603790503, + "grad_norm": 0.7612792253494263, + "learning_rate": 1.660957700145127e-05, + "loss": 0.0365, + "step": 34722 + }, + { + "epoch": 0.8139618019538553, + "grad_norm": 3.166088581085205, + "learning_rate": 1.66055117776175e-05, + "loss": 0.2923, + "step": 34723 + }, + { + "epoch": 0.8139852435286603, + "grad_norm": 3.0043416023254395, + "learning_rate": 1.6601447006284597e-05, + "loss": 0.1298, + "step": 34724 + }, + { + "epoch": 0.8140086851034652, + "grad_norm": 4.0778398513793945, + "learning_rate": 1.659738268747458e-05, + "loss": 0.2403, + "step": 34725 + }, + { + "epoch": 0.8140321266782703, + "grad_norm": 0.51999831199646, + "learning_rate": 1.6593318821209513e-05, + "loss": 0.0531, + "step": 34726 + }, + { + "epoch": 0.8140555682530752, + "grad_norm": 0.6478433609008789, + "learning_rate": 1.6589255407511418e-05, + "loss": 0.0549, + "step": 34727 + }, + { + "epoch": 0.8140790098278803, + "grad_norm": 2.938750982284546, + "learning_rate": 1.6585192446402386e-05, + "loss": 0.3278, + "step": 34728 + }, + { + "epoch": 0.8141024514026852, + "grad_norm": 0.7740217447280884, + "learning_rate": 1.658112993790446e-05, + "loss": 0.0873, + "step": 34729 + }, + { + "epoch": 0.8141258929774903, + "grad_norm": 4.287428855895996, + "learning_rate": 1.657706788203963e-05, + "loss": 0.5136, + "step": 34730 + }, + { + "epoch": 0.8141493345522952, + "grad_norm": 1.5022518634796143, + "learning_rate": 1.6573006278830007e-05, + "loss": 0.0705, + "step": 34731 + }, + { + "epoch": 0.8141727761271003, + "grad_norm": 4.612793922424316, + "learning_rate": 1.6568945128297597e-05, + "loss": 0.4926, + "step": 34732 + }, + { + "epoch": 0.8141962177019052, + "grad_norm": 3.1215627193450928, + "learning_rate": 1.6564884430464435e-05, + "loss": 0.2846, + "step": 34733 + }, + { + "epoch": 0.8142196592767103, + "grad_norm": 2.6533820629119873, + "learning_rate": 1.6560824185352552e-05, + "loss": 1.1972, + "step": 34734 + }, + { + "epoch": 0.8142431008515152, + "grad_norm": 3.3914825916290283, + "learning_rate": 1.655676439298397e-05, + "loss": 0.2426, + "step": 34735 + }, + { + "epoch": 0.8142665424263202, + "grad_norm": 1.3584431409835815, + "learning_rate": 1.6552705053380747e-05, + "loss": 1.6417, + "step": 34736 + }, + { + "epoch": 0.8142899840011252, + "grad_norm": 2.4263696670532227, + "learning_rate": 1.6548646166564898e-05, + "loss": 0.1498, + "step": 34737 + }, + { + "epoch": 0.8143134255759302, + "grad_norm": 2.1600584983825684, + "learning_rate": 1.6544587732558435e-05, + "loss": 0.0923, + "step": 34738 + }, + { + "epoch": 0.8143368671507352, + "grad_norm": 0.587587296962738, + "learning_rate": 1.654052975138336e-05, + "loss": 0.0285, + "step": 34739 + }, + { + "epoch": 0.8143603087255402, + "grad_norm": 1.6895992755889893, + "learning_rate": 1.653647222306175e-05, + "loss": 0.1257, + "step": 34740 + }, + { + "epoch": 0.8143837503003452, + "grad_norm": 1.179383397102356, + "learning_rate": 1.6532415147615588e-05, + "loss": 0.0833, + "step": 34741 + }, + { + "epoch": 0.8144071918751502, + "grad_norm": 4.066610813140869, + "learning_rate": 1.6528358525066855e-05, + "loss": 0.4505, + "step": 34742 + }, + { + "epoch": 0.8144306334499551, + "grad_norm": 3.833696126937866, + "learning_rate": 1.6524302355437626e-05, + "loss": 0.335, + "step": 34743 + }, + { + "epoch": 0.8144540750247602, + "grad_norm": 1.803468942642212, + "learning_rate": 1.652024663874988e-05, + "loss": 0.0841, + "step": 34744 + }, + { + "epoch": 0.8144775165995651, + "grad_norm": 2.913012742996216, + "learning_rate": 1.6516191375025614e-05, + "loss": 0.2207, + "step": 34745 + }, + { + "epoch": 0.8145009581743702, + "grad_norm": 1.0742243528366089, + "learning_rate": 1.651213656428685e-05, + "loss": 0.092, + "step": 34746 + }, + { + "epoch": 0.8145243997491751, + "grad_norm": 0.645357072353363, + "learning_rate": 1.6508082206555574e-05, + "loss": 0.0466, + "step": 34747 + }, + { + "epoch": 0.8145478413239802, + "grad_norm": 3.654024600982666, + "learning_rate": 1.650402830185378e-05, + "loss": 0.1999, + "step": 34748 + }, + { + "epoch": 0.8145712828987851, + "grad_norm": 1.2762175798416138, + "learning_rate": 1.6499974850203493e-05, + "loss": 0.1111, + "step": 34749 + }, + { + "epoch": 0.8145947244735902, + "grad_norm": 1.8434014320373535, + "learning_rate": 1.6495921851626685e-05, + "loss": 1.0567, + "step": 34750 + }, + { + "epoch": 0.8146181660483951, + "grad_norm": 0.32933616638183594, + "learning_rate": 1.6491869306145336e-05, + "loss": 0.0387, + "step": 34751 + }, + { + "epoch": 0.8146416076232001, + "grad_norm": 1.8659453392028809, + "learning_rate": 1.6487817213781475e-05, + "loss": 0.6339, + "step": 34752 + }, + { + "epoch": 0.8146650491980051, + "grad_norm": 0.6426877379417419, + "learning_rate": 1.6483765574557065e-05, + "loss": 0.0426, + "step": 34753 + }, + { + "epoch": 0.8146884907728101, + "grad_norm": 1.7667770385742188, + "learning_rate": 1.6479714388494093e-05, + "loss": 0.0719, + "step": 34754 + }, + { + "epoch": 0.8147119323476151, + "grad_norm": 1.6880061626434326, + "learning_rate": 1.6475663655614503e-05, + "loss": 0.0742, + "step": 34755 + }, + { + "epoch": 0.8147353739224201, + "grad_norm": 1.4342912435531616, + "learning_rate": 1.6471613375940353e-05, + "loss": 0.0541, + "step": 34756 + }, + { + "epoch": 0.8147588154972251, + "grad_norm": 1.1030746698379517, + "learning_rate": 1.646756354949357e-05, + "loss": 0.1292, + "step": 34757 + }, + { + "epoch": 0.8147822570720301, + "grad_norm": 5.260685920715332, + "learning_rate": 1.6463514176296136e-05, + "loss": 0.3032, + "step": 34758 + }, + { + "epoch": 0.814805698646835, + "grad_norm": 6.604206085205078, + "learning_rate": 1.645946525637002e-05, + "loss": 0.2776, + "step": 34759 + }, + { + "epoch": 0.8148291402216401, + "grad_norm": 0.9509475231170654, + "learning_rate": 1.6455416789737198e-05, + "loss": 0.082, + "step": 34760 + }, + { + "epoch": 0.814852581796445, + "grad_norm": 2.175604820251465, + "learning_rate": 1.6451368776419607e-05, + "loss": 0.0865, + "step": 34761 + }, + { + "epoch": 0.8148760233712501, + "grad_norm": 3.7928738594055176, + "learning_rate": 1.644732121643926e-05, + "loss": 0.3953, + "step": 34762 + }, + { + "epoch": 0.814899464946055, + "grad_norm": 0.48136162757873535, + "learning_rate": 1.6443274109818073e-05, + "loss": 0.0283, + "step": 34763 + }, + { + "epoch": 0.8149229065208601, + "grad_norm": 2.9909560680389404, + "learning_rate": 1.6439227456578054e-05, + "loss": 0.1428, + "step": 34764 + }, + { + "epoch": 0.814946348095665, + "grad_norm": 2.8272998332977295, + "learning_rate": 1.643518125674115e-05, + "loss": 0.3116, + "step": 34765 + }, + { + "epoch": 0.8149697896704701, + "grad_norm": 2.5804593563079834, + "learning_rate": 1.643113551032929e-05, + "loss": 0.2248, + "step": 34766 + }, + { + "epoch": 0.814993231245275, + "grad_norm": 1.0677188634872437, + "learning_rate": 1.6427090217364416e-05, + "loss": 0.0536, + "step": 34767 + }, + { + "epoch": 0.8150166728200801, + "grad_norm": 4.430739402770996, + "learning_rate": 1.6423045377868518e-05, + "loss": 0.4715, + "step": 34768 + }, + { + "epoch": 0.815040114394885, + "grad_norm": 3.586574077606201, + "learning_rate": 1.6419000991863543e-05, + "loss": 0.1919, + "step": 34769 + }, + { + "epoch": 0.81506355596969, + "grad_norm": 0.5244281888008118, + "learning_rate": 1.6414957059371406e-05, + "loss": 0.0163, + "step": 34770 + }, + { + "epoch": 0.815086997544495, + "grad_norm": 3.423614501953125, + "learning_rate": 1.6410913580414068e-05, + "loss": 0.4537, + "step": 34771 + }, + { + "epoch": 0.8151104391193, + "grad_norm": 3.8761379718780518, + "learning_rate": 1.6406870555013455e-05, + "loss": 0.3319, + "step": 34772 + }, + { + "epoch": 0.8151338806941051, + "grad_norm": 2.397221088409424, + "learning_rate": 1.6402827983191493e-05, + "loss": 0.2851, + "step": 34773 + }, + { + "epoch": 0.81515732226891, + "grad_norm": 3.5682215690612793, + "learning_rate": 1.6398785864970156e-05, + "loss": 0.2541, + "step": 34774 + }, + { + "epoch": 0.8151807638437151, + "grad_norm": 1.238114833831787, + "learning_rate": 1.6394744200371347e-05, + "loss": 0.7594, + "step": 34775 + }, + { + "epoch": 0.81520420541852, + "grad_norm": 0.9664465188980103, + "learning_rate": 1.6390702989417018e-05, + "loss": 0.0647, + "step": 34776 + }, + { + "epoch": 0.8152276469933251, + "grad_norm": 0.32120874524116516, + "learning_rate": 1.638666223212909e-05, + "loss": 0.0097, + "step": 34777 + }, + { + "epoch": 0.81525108856813, + "grad_norm": 1.0725367069244385, + "learning_rate": 1.6382621928529486e-05, + "loss": 0.0926, + "step": 34778 + }, + { + "epoch": 0.815274530142935, + "grad_norm": 0.7342886328697205, + "learning_rate": 1.6378582078640125e-05, + "loss": 0.0389, + "step": 34779 + }, + { + "epoch": 0.81529797171774, + "grad_norm": 2.736032724380493, + "learning_rate": 1.6374542682482908e-05, + "loss": 0.1757, + "step": 34780 + }, + { + "epoch": 0.815321413292545, + "grad_norm": 3.4742515087127686, + "learning_rate": 1.637050374007979e-05, + "loss": 0.1939, + "step": 34781 + }, + { + "epoch": 0.81534485486735, + "grad_norm": 0.6967513561248779, + "learning_rate": 1.6366465251452678e-05, + "loss": 0.0564, + "step": 34782 + }, + { + "epoch": 0.815368296442155, + "grad_norm": 3.9877023696899414, + "learning_rate": 1.6362427216623467e-05, + "loss": 0.256, + "step": 34783 + }, + { + "epoch": 0.81539173801696, + "grad_norm": 2.120265007019043, + "learning_rate": 1.6358389635614078e-05, + "loss": 0.1975, + "step": 34784 + }, + { + "epoch": 0.815415179591765, + "grad_norm": 1.532757043838501, + "learning_rate": 1.6354352508446437e-05, + "loss": 0.0966, + "step": 34785 + }, + { + "epoch": 0.81543862116657, + "grad_norm": 3.7512288093566895, + "learning_rate": 1.6350315835142393e-05, + "loss": 0.2002, + "step": 34786 + }, + { + "epoch": 0.815462062741375, + "grad_norm": 2.7643091678619385, + "learning_rate": 1.634627961572388e-05, + "loss": 0.1731, + "step": 34787 + }, + { + "epoch": 0.8154855043161799, + "grad_norm": 3.58186674118042, + "learning_rate": 1.6342243850212846e-05, + "loss": 0.3812, + "step": 34788 + }, + { + "epoch": 0.815508945890985, + "grad_norm": 3.6191766262054443, + "learning_rate": 1.633820853863114e-05, + "loss": 0.3127, + "step": 34789 + }, + { + "epoch": 0.8155323874657899, + "grad_norm": 4.204402923583984, + "learning_rate": 1.6334173681000662e-05, + "loss": 0.2121, + "step": 34790 + }, + { + "epoch": 0.815555829040595, + "grad_norm": 3.1165738105773926, + "learning_rate": 1.6330139277343325e-05, + "loss": 0.489, + "step": 34791 + }, + { + "epoch": 0.8155792706153999, + "grad_norm": 2.397552490234375, + "learning_rate": 1.6326105327680997e-05, + "loss": 0.0721, + "step": 34792 + }, + { + "epoch": 0.815602712190205, + "grad_norm": 1.1686936616897583, + "learning_rate": 1.6322071832035546e-05, + "loss": 0.1162, + "step": 34793 + }, + { + "epoch": 0.8156261537650099, + "grad_norm": 3.5556976795196533, + "learning_rate": 1.6318038790428913e-05, + "loss": 0.3189, + "step": 34794 + }, + { + "epoch": 0.815649595339815, + "grad_norm": 2.5365054607391357, + "learning_rate": 1.6314006202882958e-05, + "loss": 0.2533, + "step": 34795 + }, + { + "epoch": 0.8156730369146199, + "grad_norm": 0.7965208292007446, + "learning_rate": 1.6309974069419554e-05, + "loss": 0.0529, + "step": 34796 + }, + { + "epoch": 0.815696478489425, + "grad_norm": 3.130488395690918, + "learning_rate": 1.630594239006059e-05, + "loss": 0.1251, + "step": 34797 + }, + { + "epoch": 0.8157199200642299, + "grad_norm": 1.677581548690796, + "learning_rate": 1.6301911164827933e-05, + "loss": 0.0822, + "step": 34798 + }, + { + "epoch": 0.8157433616390349, + "grad_norm": 2.6628522872924805, + "learning_rate": 1.6297880393743436e-05, + "loss": 0.1209, + "step": 34799 + }, + { + "epoch": 0.8157668032138399, + "grad_norm": 2.243375539779663, + "learning_rate": 1.629385007682899e-05, + "loss": 0.1946, + "step": 34800 + }, + { + "epoch": 0.8157902447886449, + "grad_norm": 1.9721373319625854, + "learning_rate": 1.628982021410649e-05, + "loss": 0.1314, + "step": 34801 + }, + { + "epoch": 0.8158136863634499, + "grad_norm": 2.8873448371887207, + "learning_rate": 1.6285790805597778e-05, + "loss": 0.1457, + "step": 34802 + }, + { + "epoch": 0.8158371279382549, + "grad_norm": 3.786580801010132, + "learning_rate": 1.6281761851324728e-05, + "loss": 0.3519, + "step": 34803 + }, + { + "epoch": 0.8158605695130599, + "grad_norm": 2.562870740890503, + "learning_rate": 1.6277733351309187e-05, + "loss": 1.1044, + "step": 34804 + }, + { + "epoch": 0.8158840110878649, + "grad_norm": 2.6273763179779053, + "learning_rate": 1.6273705305572994e-05, + "loss": 0.3369, + "step": 34805 + }, + { + "epoch": 0.8159074526626698, + "grad_norm": 2.9009759426116943, + "learning_rate": 1.6269677714138053e-05, + "loss": 0.249, + "step": 34806 + }, + { + "epoch": 0.8159308942374749, + "grad_norm": 3.6005451679229736, + "learning_rate": 1.6265650577026205e-05, + "loss": 0.2775, + "step": 34807 + }, + { + "epoch": 0.8159543358122798, + "grad_norm": 1.2858392000198364, + "learning_rate": 1.6261623894259283e-05, + "loss": 0.1173, + "step": 34808 + }, + { + "epoch": 0.8159777773870849, + "grad_norm": 2.5006649494171143, + "learning_rate": 1.6257597665859136e-05, + "loss": 0.2387, + "step": 34809 + }, + { + "epoch": 0.8160012189618898, + "grad_norm": 0.21869926154613495, + "learning_rate": 1.625357189184763e-05, + "loss": 0.0065, + "step": 34810 + }, + { + "epoch": 0.8160246605366949, + "grad_norm": 0.9377149939537048, + "learning_rate": 1.6249546572246576e-05, + "loss": 0.1096, + "step": 34811 + }, + { + "epoch": 0.8160481021114998, + "grad_norm": 1.6930471658706665, + "learning_rate": 1.6245521707077825e-05, + "loss": 0.1301, + "step": 34812 + }, + { + "epoch": 0.8160715436863049, + "grad_norm": 3.6339361667633057, + "learning_rate": 1.624149729636326e-05, + "loss": 0.2779, + "step": 34813 + }, + { + "epoch": 0.8160949852611098, + "grad_norm": 2.5227386951446533, + "learning_rate": 1.623747334012469e-05, + "loss": 1.7839, + "step": 34814 + }, + { + "epoch": 0.8161184268359148, + "grad_norm": 2.2055130004882812, + "learning_rate": 1.623344983838394e-05, + "loss": 0.1888, + "step": 34815 + }, + { + "epoch": 0.8161418684107198, + "grad_norm": 2.2785558700561523, + "learning_rate": 1.622942679116285e-05, + "loss": 0.2541, + "step": 34816 + }, + { + "epoch": 0.8161653099855248, + "grad_norm": 0.950619101524353, + "learning_rate": 1.622540419848324e-05, + "loss": 0.0684, + "step": 34817 + }, + { + "epoch": 0.8161887515603298, + "grad_norm": 0.41098541021347046, + "learning_rate": 1.6221382060366918e-05, + "loss": 0.0237, + "step": 34818 + }, + { + "epoch": 0.8162121931351348, + "grad_norm": 3.1454265117645264, + "learning_rate": 1.6217360376835755e-05, + "loss": 0.1907, + "step": 34819 + }, + { + "epoch": 0.8162356347099398, + "grad_norm": 0.6715973615646362, + "learning_rate": 1.6213339147911556e-05, + "loss": 0.053, + "step": 34820 + }, + { + "epoch": 0.8162590762847448, + "grad_norm": 1.4368181228637695, + "learning_rate": 1.6209318373616133e-05, + "loss": 0.1015, + "step": 34821 + }, + { + "epoch": 0.8162825178595497, + "grad_norm": 1.874693512916565, + "learning_rate": 1.6205298053971306e-05, + "loss": 1.6231, + "step": 34822 + }, + { + "epoch": 0.8163059594343548, + "grad_norm": 3.6004981994628906, + "learning_rate": 1.6201278188998892e-05, + "loss": 0.3084, + "step": 34823 + }, + { + "epoch": 0.8163294010091598, + "grad_norm": 4.278038501739502, + "learning_rate": 1.6197258778720658e-05, + "loss": 0.4396, + "step": 34824 + }, + { + "epoch": 0.8163528425839648, + "grad_norm": 0.8882210850715637, + "learning_rate": 1.619323982315847e-05, + "loss": 0.0684, + "step": 34825 + }, + { + "epoch": 0.8163762841587698, + "grad_norm": 2.6440086364746094, + "learning_rate": 1.6189221322334138e-05, + "loss": 0.1352, + "step": 34826 + }, + { + "epoch": 0.8163997257335748, + "grad_norm": 3.0567257404327393, + "learning_rate": 1.6185203276269446e-05, + "loss": 0.1768, + "step": 34827 + }, + { + "epoch": 0.8164231673083798, + "grad_norm": 2.4356043338775635, + "learning_rate": 1.61811856849862e-05, + "loss": 0.1325, + "step": 34828 + }, + { + "epoch": 0.8164466088831848, + "grad_norm": 0.5040231943130493, + "learning_rate": 1.6177168548506204e-05, + "loss": 0.0344, + "step": 34829 + }, + { + "epoch": 0.8164700504579898, + "grad_norm": 2.7054100036621094, + "learning_rate": 1.6173151866851244e-05, + "loss": 0.2506, + "step": 34830 + }, + { + "epoch": 0.8164934920327948, + "grad_norm": 3.4099478721618652, + "learning_rate": 1.6169135640043086e-05, + "loss": 0.1385, + "step": 34831 + }, + { + "epoch": 0.8165169336075998, + "grad_norm": 2.155930757522583, + "learning_rate": 1.6165119868103584e-05, + "loss": 0.1077, + "step": 34832 + }, + { + "epoch": 0.8165403751824047, + "grad_norm": 1.1811203956604004, + "learning_rate": 1.6161104551054507e-05, + "loss": 0.114, + "step": 34833 + }, + { + "epoch": 0.8165638167572098, + "grad_norm": 1.9873746633529663, + "learning_rate": 1.6157089688917627e-05, + "loss": 0.2257, + "step": 34834 + }, + { + "epoch": 0.8165872583320147, + "grad_norm": 3.1409835815429688, + "learning_rate": 1.6153075281714737e-05, + "loss": 1.4962, + "step": 34835 + }, + { + "epoch": 0.8166106999068198, + "grad_norm": 2.5130136013031006, + "learning_rate": 1.6149061329467584e-05, + "loss": 0.2053, + "step": 34836 + }, + { + "epoch": 0.8166341414816247, + "grad_norm": 3.602099657058716, + "learning_rate": 1.6145047832198023e-05, + "loss": 0.31, + "step": 34837 + }, + { + "epoch": 0.8166575830564298, + "grad_norm": 2.9583535194396973, + "learning_rate": 1.6141034789927754e-05, + "loss": 0.2428, + "step": 34838 + }, + { + "epoch": 0.8166810246312347, + "grad_norm": 2.387524127960205, + "learning_rate": 1.613702220267862e-05, + "loss": 0.195, + "step": 34839 + }, + { + "epoch": 0.8167044662060398, + "grad_norm": 2.505666732788086, + "learning_rate": 1.6133010070472355e-05, + "loss": 0.1985, + "step": 34840 + }, + { + "epoch": 0.8167279077808447, + "grad_norm": 2.104027032852173, + "learning_rate": 1.6128998393330742e-05, + "loss": 0.1247, + "step": 34841 + }, + { + "epoch": 0.8167513493556497, + "grad_norm": 3.1176555156707764, + "learning_rate": 1.612498717127554e-05, + "loss": 0.2334, + "step": 34842 + }, + { + "epoch": 0.8167747909304547, + "grad_norm": 2.675164222717285, + "learning_rate": 1.6120976404328492e-05, + "loss": 0.139, + "step": 34843 + }, + { + "epoch": 0.8167982325052597, + "grad_norm": 2.3289740085601807, + "learning_rate": 1.61169660925114e-05, + "loss": 0.213, + "step": 34844 + }, + { + "epoch": 0.8168216740800647, + "grad_norm": 1.4637205600738525, + "learning_rate": 1.611295623584602e-05, + "loss": 0.1543, + "step": 34845 + }, + { + "epoch": 0.8168451156548697, + "grad_norm": 4.1243510246276855, + "learning_rate": 1.6108946834354087e-05, + "loss": 0.3216, + "step": 34846 + }, + { + "epoch": 0.8168685572296747, + "grad_norm": 2.903890609741211, + "learning_rate": 1.6104937888057382e-05, + "loss": 0.3663, + "step": 34847 + }, + { + "epoch": 0.8168919988044797, + "grad_norm": 1.8524987697601318, + "learning_rate": 1.6100929396977603e-05, + "loss": 1.1544, + "step": 34848 + }, + { + "epoch": 0.8169154403792847, + "grad_norm": 1.911027431488037, + "learning_rate": 1.6096921361136575e-05, + "loss": 0.0755, + "step": 34849 + }, + { + "epoch": 0.8169388819540897, + "grad_norm": 1.0490679740905762, + "learning_rate": 1.6092913780555984e-05, + "loss": 0.0874, + "step": 34850 + }, + { + "epoch": 0.8169623235288946, + "grad_norm": 3.1291961669921875, + "learning_rate": 1.6088906655257628e-05, + "loss": 0.2465, + "step": 34851 + }, + { + "epoch": 0.8169857651036997, + "grad_norm": 0.7999172210693359, + "learning_rate": 1.6084899985263225e-05, + "loss": 0.0746, + "step": 34852 + }, + { + "epoch": 0.8170092066785046, + "grad_norm": 2.396497964859009, + "learning_rate": 1.6080893770594507e-05, + "loss": 0.1576, + "step": 34853 + }, + { + "epoch": 0.8170326482533097, + "grad_norm": 3.8992507457733154, + "learning_rate": 1.6076888011273228e-05, + "loss": 0.2128, + "step": 34854 + }, + { + "epoch": 0.8170560898281146, + "grad_norm": 2.071319818496704, + "learning_rate": 1.60728827073211e-05, + "loss": 0.172, + "step": 34855 + }, + { + "epoch": 0.8170795314029197, + "grad_norm": 3.290191173553467, + "learning_rate": 1.6068877858759856e-05, + "loss": 0.2249, + "step": 34856 + }, + { + "epoch": 0.8171029729777246, + "grad_norm": 1.533762812614441, + "learning_rate": 1.6064873465611276e-05, + "loss": 0.0895, + "step": 34857 + }, + { + "epoch": 0.8171264145525297, + "grad_norm": 2.570829153060913, + "learning_rate": 1.6060869527897037e-05, + "loss": 0.1583, + "step": 34858 + }, + { + "epoch": 0.8171498561273346, + "grad_norm": 2.234928846359253, + "learning_rate": 1.6056866045638886e-05, + "loss": 0.1043, + "step": 34859 + }, + { + "epoch": 0.8171732977021396, + "grad_norm": 1.7387491464614868, + "learning_rate": 1.605286301885851e-05, + "loss": 0.2261, + "step": 34860 + }, + { + "epoch": 0.8171967392769446, + "grad_norm": 3.0567283630371094, + "learning_rate": 1.6048860447577687e-05, + "loss": 0.2359, + "step": 34861 + }, + { + "epoch": 0.8172201808517496, + "grad_norm": 1.9164482355117798, + "learning_rate": 1.6044858331818114e-05, + "loss": 0.1028, + "step": 34862 + }, + { + "epoch": 0.8172436224265546, + "grad_norm": 1.53068208694458, + "learning_rate": 1.6040856671601467e-05, + "loss": 0.1003, + "step": 34863 + }, + { + "epoch": 0.8172670640013596, + "grad_norm": 1.3873239755630493, + "learning_rate": 1.6036855466949517e-05, + "loss": 0.114, + "step": 34864 + }, + { + "epoch": 0.8172905055761646, + "grad_norm": 3.093759298324585, + "learning_rate": 1.603285471788396e-05, + "loss": 0.2246, + "step": 34865 + }, + { + "epoch": 0.8173139471509696, + "grad_norm": 4.630692005157471, + "learning_rate": 1.6028854424426488e-05, + "loss": 0.4329, + "step": 34866 + }, + { + "epoch": 0.8173373887257745, + "grad_norm": 1.5262752771377563, + "learning_rate": 1.6024854586598814e-05, + "loss": 0.0915, + "step": 34867 + }, + { + "epoch": 0.8173608303005796, + "grad_norm": 2.561030149459839, + "learning_rate": 1.6020855204422635e-05, + "loss": 0.1553, + "step": 34868 + }, + { + "epoch": 0.8173842718753845, + "grad_norm": 0.961212158203125, + "learning_rate": 1.6016856277919633e-05, + "loss": 0.0881, + "step": 34869 + }, + { + "epoch": 0.8174077134501896, + "grad_norm": 2.7588951587677, + "learning_rate": 1.601285780711156e-05, + "loss": 0.0534, + "step": 34870 + }, + { + "epoch": 0.8174311550249945, + "grad_norm": 6.148392677307129, + "learning_rate": 1.600885979202008e-05, + "loss": 0.6117, + "step": 34871 + }, + { + "epoch": 0.8174545965997996, + "grad_norm": 4.913458824157715, + "learning_rate": 1.6004862232666885e-05, + "loss": 0.422, + "step": 34872 + }, + { + "epoch": 0.8174780381746045, + "grad_norm": 3.081502914428711, + "learning_rate": 1.600086512907365e-05, + "loss": 1.5481, + "step": 34873 + }, + { + "epoch": 0.8175014797494096, + "grad_norm": 3.3439371585845947, + "learning_rate": 1.5996868481262106e-05, + "loss": 0.3449, + "step": 34874 + }, + { + "epoch": 0.8175249213242146, + "grad_norm": 1.7092243432998657, + "learning_rate": 1.5992872289253912e-05, + "loss": 0.1671, + "step": 34875 + }, + { + "epoch": 0.8175483628990196, + "grad_norm": 1.4664736986160278, + "learning_rate": 1.5988876553070742e-05, + "loss": 0.0805, + "step": 34876 + }, + { + "epoch": 0.8175718044738246, + "grad_norm": 3.1241021156311035, + "learning_rate": 1.5984881272734297e-05, + "loss": 0.2067, + "step": 34877 + }, + { + "epoch": 0.8175952460486295, + "grad_norm": 1.467206597328186, + "learning_rate": 1.5980886448266265e-05, + "loss": 0.9716, + "step": 34878 + }, + { + "epoch": 0.8176186876234346, + "grad_norm": 2.9628210067749023, + "learning_rate": 1.5976892079688298e-05, + "loss": 0.1203, + "step": 34879 + }, + { + "epoch": 0.8176421291982395, + "grad_norm": 4.1054463386535645, + "learning_rate": 1.5972898167022075e-05, + "loss": 0.1499, + "step": 34880 + }, + { + "epoch": 0.8176655707730446, + "grad_norm": 2.2905375957489014, + "learning_rate": 1.5968904710289246e-05, + "loss": 0.2951, + "step": 34881 + }, + { + "epoch": 0.8176890123478495, + "grad_norm": 1.4476478099822998, + "learning_rate": 1.596491170951153e-05, + "loss": 0.0424, + "step": 34882 + }, + { + "epoch": 0.8177124539226546, + "grad_norm": 1.9865247011184692, + "learning_rate": 1.5960919164710565e-05, + "loss": 0.9536, + "step": 34883 + }, + { + "epoch": 0.8177358954974595, + "grad_norm": 6.517726898193359, + "learning_rate": 1.5956927075908013e-05, + "loss": 0.3421, + "step": 34884 + }, + { + "epoch": 0.8177593370722646, + "grad_norm": 1.2942376136779785, + "learning_rate": 1.5952935443125517e-05, + "loss": 0.0789, + "step": 34885 + }, + { + "epoch": 0.8177827786470695, + "grad_norm": 4.95007848739624, + "learning_rate": 1.594894426638478e-05, + "loss": 0.1557, + "step": 34886 + }, + { + "epoch": 0.8178062202218745, + "grad_norm": 2.216827154159546, + "learning_rate": 1.594495354570743e-05, + "loss": 0.2175, + "step": 34887 + }, + { + "epoch": 0.8178296617966795, + "grad_norm": 1.1544853448867798, + "learning_rate": 1.5940963281115108e-05, + "loss": 0.0772, + "step": 34888 + }, + { + "epoch": 0.8178531033714845, + "grad_norm": 3.0239217281341553, + "learning_rate": 1.59369734726295e-05, + "loss": 0.2911, + "step": 34889 + }, + { + "epoch": 0.8178765449462895, + "grad_norm": 2.392700672149658, + "learning_rate": 1.5932984120272242e-05, + "loss": 1.1562, + "step": 34890 + }, + { + "epoch": 0.8178999865210945, + "grad_norm": 0.9799839854240417, + "learning_rate": 1.592899522406498e-05, + "loss": 0.0867, + "step": 34891 + }, + { + "epoch": 0.8179234280958995, + "grad_norm": 0.723118782043457, + "learning_rate": 1.592500678402935e-05, + "loss": 0.0322, + "step": 34892 + }, + { + "epoch": 0.8179468696707045, + "grad_norm": 2.151193141937256, + "learning_rate": 1.5921018800186992e-05, + "loss": 1.6687, + "step": 34893 + }, + { + "epoch": 0.8179703112455095, + "grad_norm": 3.354515552520752, + "learning_rate": 1.5917031272559534e-05, + "loss": 0.3031, + "step": 34894 + }, + { + "epoch": 0.8179937528203145, + "grad_norm": 2.132272481918335, + "learning_rate": 1.5913044201168646e-05, + "loss": 0.1895, + "step": 34895 + }, + { + "epoch": 0.8180171943951194, + "grad_norm": 2.179828643798828, + "learning_rate": 1.5909057586035945e-05, + "loss": 0.0457, + "step": 34896 + }, + { + "epoch": 0.8180406359699245, + "grad_norm": 3.418680191040039, + "learning_rate": 1.5905071427183037e-05, + "loss": 0.224, + "step": 34897 + }, + { + "epoch": 0.8180640775447294, + "grad_norm": 1.3882520198822021, + "learning_rate": 1.5901085724631602e-05, + "loss": 0.1685, + "step": 34898 + }, + { + "epoch": 0.8180875191195345, + "grad_norm": 5.082427978515625, + "learning_rate": 1.5897100478403247e-05, + "loss": 0.6182, + "step": 34899 + }, + { + "epoch": 0.8181109606943394, + "grad_norm": 3.965430736541748, + "learning_rate": 1.5893115688519587e-05, + "loss": 0.3047, + "step": 34900 + }, + { + "epoch": 0.8181344022691445, + "grad_norm": 5.3480963706970215, + "learning_rate": 1.588913135500222e-05, + "loss": 0.35, + "step": 34901 + }, + { + "epoch": 0.8181578438439494, + "grad_norm": 0.6641038060188293, + "learning_rate": 1.5885147477872807e-05, + "loss": 0.061, + "step": 34902 + }, + { + "epoch": 0.8181812854187545, + "grad_norm": 1.9893513917922974, + "learning_rate": 1.5881164057152963e-05, + "loss": 0.1284, + "step": 34903 + }, + { + "epoch": 0.8182047269935594, + "grad_norm": 2.6342461109161377, + "learning_rate": 1.587718109286428e-05, + "loss": 0.2933, + "step": 34904 + }, + { + "epoch": 0.8182281685683644, + "grad_norm": 1.0802793502807617, + "learning_rate": 1.587319858502838e-05, + "loss": 0.1254, + "step": 34905 + }, + { + "epoch": 0.8182516101431694, + "grad_norm": 1.2909873723983765, + "learning_rate": 1.5869216533666865e-05, + "loss": 1.3647, + "step": 34906 + }, + { + "epoch": 0.8182750517179744, + "grad_norm": 1.636175513267517, + "learning_rate": 1.5865234938801332e-05, + "loss": 0.0789, + "step": 34907 + }, + { + "epoch": 0.8182984932927794, + "grad_norm": 1.5963997840881348, + "learning_rate": 1.5861253800453414e-05, + "loss": 0.0822, + "step": 34908 + }, + { + "epoch": 0.8183219348675844, + "grad_norm": 1.897929072380066, + "learning_rate": 1.585727311864469e-05, + "loss": 0.1487, + "step": 34909 + }, + { + "epoch": 0.8183453764423894, + "grad_norm": 3.9888970851898193, + "learning_rate": 1.5853292893396777e-05, + "loss": 0.2895, + "step": 34910 + }, + { + "epoch": 0.8183688180171944, + "grad_norm": 3.219853401184082, + "learning_rate": 1.584931312473127e-05, + "loss": 0.1855, + "step": 34911 + }, + { + "epoch": 0.8183922595919993, + "grad_norm": 2.117760181427002, + "learning_rate": 1.5845333812669748e-05, + "loss": 0.0887, + "step": 34912 + }, + { + "epoch": 0.8184157011668044, + "grad_norm": 2.022287607192993, + "learning_rate": 1.5841354957233812e-05, + "loss": 0.1864, + "step": 34913 + }, + { + "epoch": 0.8184391427416093, + "grad_norm": 2.5115792751312256, + "learning_rate": 1.5837376558445037e-05, + "loss": 0.14, + "step": 34914 + }, + { + "epoch": 0.8184625843164144, + "grad_norm": 2.2692081928253174, + "learning_rate": 1.5833398616325036e-05, + "loss": 0.1268, + "step": 34915 + }, + { + "epoch": 0.8184860258912193, + "grad_norm": 3.065816640853882, + "learning_rate": 1.5829421130895385e-05, + "loss": 0.1645, + "step": 34916 + }, + { + "epoch": 0.8185094674660244, + "grad_norm": 1.0068132877349854, + "learning_rate": 1.5825444102177665e-05, + "loss": 0.1106, + "step": 34917 + }, + { + "epoch": 0.8185329090408293, + "grad_norm": 3.0075109004974365, + "learning_rate": 1.5821467530193447e-05, + "loss": 0.1774, + "step": 34918 + }, + { + "epoch": 0.8185563506156344, + "grad_norm": 2.7731916904449463, + "learning_rate": 1.581749141496429e-05, + "loss": 0.2493, + "step": 34919 + }, + { + "epoch": 0.8185797921904393, + "grad_norm": 2.470649242401123, + "learning_rate": 1.581351575651182e-05, + "loss": 0.1976, + "step": 34920 + }, + { + "epoch": 0.8186032337652444, + "grad_norm": 3.5908191204071045, + "learning_rate": 1.580954055485755e-05, + "loss": 0.2794, + "step": 34921 + }, + { + "epoch": 0.8186266753400493, + "grad_norm": 2.472879648208618, + "learning_rate": 1.5805565810023114e-05, + "loss": 0.1884, + "step": 34922 + }, + { + "epoch": 0.8186501169148543, + "grad_norm": 1.1316734552383423, + "learning_rate": 1.5801591522030034e-05, + "loss": 0.0629, + "step": 34923 + }, + { + "epoch": 0.8186735584896593, + "grad_norm": 4.4839630126953125, + "learning_rate": 1.5797617690899892e-05, + "loss": 0.2823, + "step": 34924 + }, + { + "epoch": 0.8186970000644643, + "grad_norm": 6.20732307434082, + "learning_rate": 1.5793644316654242e-05, + "loss": 0.4909, + "step": 34925 + }, + { + "epoch": 0.8187204416392694, + "grad_norm": 3.8376171588897705, + "learning_rate": 1.578967139931462e-05, + "loss": 0.2744, + "step": 34926 + }, + { + "epoch": 0.8187438832140743, + "grad_norm": 1.865358591079712, + "learning_rate": 1.5785698938902637e-05, + "loss": 0.1123, + "step": 34927 + }, + { + "epoch": 0.8187673247888794, + "grad_norm": 3.0874078273773193, + "learning_rate": 1.5781726935439822e-05, + "loss": 0.4086, + "step": 34928 + }, + { + "epoch": 0.8187907663636843, + "grad_norm": 3.210479259490967, + "learning_rate": 1.5777755388947724e-05, + "loss": 0.219, + "step": 34929 + }, + { + "epoch": 0.8188142079384894, + "grad_norm": 2.1017961502075195, + "learning_rate": 1.577378429944788e-05, + "loss": 0.0924, + "step": 34930 + }, + { + "epoch": 0.8188376495132943, + "grad_norm": 2.277959108352661, + "learning_rate": 1.5769813666961862e-05, + "loss": 0.108, + "step": 34931 + }, + { + "epoch": 0.8188610910880993, + "grad_norm": 2.0919783115386963, + "learning_rate": 1.5765843491511178e-05, + "loss": 0.1505, + "step": 34932 + }, + { + "epoch": 0.8188845326629043, + "grad_norm": 0.5172796845436096, + "learning_rate": 1.5761873773117398e-05, + "loss": 0.0528, + "step": 34933 + }, + { + "epoch": 0.8189079742377093, + "grad_norm": 0.8500393629074097, + "learning_rate": 1.575790451180208e-05, + "loss": 0.0517, + "step": 34934 + }, + { + "epoch": 0.8189314158125143, + "grad_norm": 5.660991191864014, + "learning_rate": 1.575393570758674e-05, + "loss": 0.5425, + "step": 34935 + }, + { + "epoch": 0.8189548573873193, + "grad_norm": 2.75400447845459, + "learning_rate": 1.5749967360492912e-05, + "loss": 0.261, + "step": 34936 + }, + { + "epoch": 0.8189782989621243, + "grad_norm": 2.363588809967041, + "learning_rate": 1.574599947054214e-05, + "loss": 0.0289, + "step": 34937 + }, + { + "epoch": 0.8190017405369293, + "grad_norm": 3.0922017097473145, + "learning_rate": 1.5742032037755938e-05, + "loss": 0.2162, + "step": 34938 + }, + { + "epoch": 0.8190251821117343, + "grad_norm": 1.4486333131790161, + "learning_rate": 1.5738065062155815e-05, + "loss": 0.1194, + "step": 34939 + }, + { + "epoch": 0.8190486236865393, + "grad_norm": 4.767146587371826, + "learning_rate": 1.573409854376334e-05, + "loss": 0.3241, + "step": 34940 + }, + { + "epoch": 0.8190720652613442, + "grad_norm": 2.0110182762145996, + "learning_rate": 1.573013248260001e-05, + "loss": 0.0643, + "step": 34941 + }, + { + "epoch": 0.8190955068361493, + "grad_norm": 1.8494443893432617, + "learning_rate": 1.5726166878687364e-05, + "loss": 0.1085, + "step": 34942 + }, + { + "epoch": 0.8191189484109542, + "grad_norm": 0.3483686149120331, + "learning_rate": 1.5722201732046903e-05, + "loss": 0.0361, + "step": 34943 + }, + { + "epoch": 0.8191423899857593, + "grad_norm": 1.1062047481536865, + "learning_rate": 1.5718237042700146e-05, + "loss": 0.0889, + "step": 34944 + }, + { + "epoch": 0.8191658315605642, + "grad_norm": 3.5482139587402344, + "learning_rate": 1.5714272810668586e-05, + "loss": 0.2607, + "step": 34945 + }, + { + "epoch": 0.8191892731353693, + "grad_norm": 1.9944837093353271, + "learning_rate": 1.5710309035973736e-05, + "loss": 0.1395, + "step": 34946 + }, + { + "epoch": 0.8192127147101742, + "grad_norm": 2.124209403991699, + "learning_rate": 1.570634571863715e-05, + "loss": 0.1377, + "step": 34947 + }, + { + "epoch": 0.8192361562849793, + "grad_norm": 2.228746175765991, + "learning_rate": 1.570238285868031e-05, + "loss": 0.1558, + "step": 34948 + }, + { + "epoch": 0.8192595978597842, + "grad_norm": 3.8759758472442627, + "learning_rate": 1.5698420456124698e-05, + "loss": 0.2773, + "step": 34949 + }, + { + "epoch": 0.8192830394345892, + "grad_norm": 0.218766450881958, + "learning_rate": 1.5694458510991838e-05, + "loss": 0.011, + "step": 34950 + }, + { + "epoch": 0.8193064810093942, + "grad_norm": 0.6874687671661377, + "learning_rate": 1.5690497023303208e-05, + "loss": 0.0331, + "step": 34951 + }, + { + "epoch": 0.8193299225841992, + "grad_norm": 2.1833765506744385, + "learning_rate": 1.5686535993080298e-05, + "loss": 0.1597, + "step": 34952 + }, + { + "epoch": 0.8193533641590042, + "grad_norm": 0.9814507961273193, + "learning_rate": 1.5682575420344626e-05, + "loss": 0.1402, + "step": 34953 + }, + { + "epoch": 0.8193768057338092, + "grad_norm": 5.067755222320557, + "learning_rate": 1.5678615305117683e-05, + "loss": 0.247, + "step": 34954 + }, + { + "epoch": 0.8194002473086142, + "grad_norm": 2.4980592727661133, + "learning_rate": 1.5674655647420934e-05, + "loss": 0.1747, + "step": 34955 + }, + { + "epoch": 0.8194236888834192, + "grad_norm": 2.4931650161743164, + "learning_rate": 1.5670696447275878e-05, + "loss": 1.8336, + "step": 34956 + }, + { + "epoch": 0.8194471304582241, + "grad_norm": 3.1611392498016357, + "learning_rate": 1.566673770470397e-05, + "loss": 0.1688, + "step": 34957 + }, + { + "epoch": 0.8194705720330292, + "grad_norm": 6.334897518157959, + "learning_rate": 1.5662779419726738e-05, + "loss": 0.5324, + "step": 34958 + }, + { + "epoch": 0.8194940136078341, + "grad_norm": 1.8437620401382446, + "learning_rate": 1.5658821592365614e-05, + "loss": 0.2259, + "step": 34959 + }, + { + "epoch": 0.8195174551826392, + "grad_norm": 3.199519395828247, + "learning_rate": 1.565486422264212e-05, + "loss": 0.3162, + "step": 34960 + }, + { + "epoch": 0.8195408967574441, + "grad_norm": 2.5347652435302734, + "learning_rate": 1.56509073105777e-05, + "loss": 0.1748, + "step": 34961 + }, + { + "epoch": 0.8195643383322492, + "grad_norm": 1.4537765979766846, + "learning_rate": 1.5646950856193833e-05, + "loss": 0.1193, + "step": 34962 + }, + { + "epoch": 0.8195877799070541, + "grad_norm": 2.6123263835906982, + "learning_rate": 1.5642994859511985e-05, + "loss": 0.1439, + "step": 34963 + }, + { + "epoch": 0.8196112214818592, + "grad_norm": 5.461097240447998, + "learning_rate": 1.5639039320553595e-05, + "loss": 0.3248, + "step": 34964 + }, + { + "epoch": 0.8196346630566641, + "grad_norm": 3.067807912826538, + "learning_rate": 1.563508423934017e-05, + "loss": 0.1099, + "step": 34965 + }, + { + "epoch": 0.8196581046314692, + "grad_norm": 1.9932981729507446, + "learning_rate": 1.5631129615893146e-05, + "loss": 0.1656, + "step": 34966 + }, + { + "epoch": 0.8196815462062741, + "grad_norm": 3.1377227306365967, + "learning_rate": 1.562717545023399e-05, + "loss": 0.4907, + "step": 34967 + }, + { + "epoch": 0.8197049877810791, + "grad_norm": 3.149585008621216, + "learning_rate": 1.562322174238415e-05, + "loss": 0.2209, + "step": 34968 + }, + { + "epoch": 0.8197284293558841, + "grad_norm": 4.004940509796143, + "learning_rate": 1.561926849236508e-05, + "loss": 0.3053, + "step": 34969 + }, + { + "epoch": 0.8197518709306891, + "grad_norm": 0.31855469942092896, + "learning_rate": 1.5615315700198207e-05, + "loss": 0.0235, + "step": 34970 + }, + { + "epoch": 0.8197753125054941, + "grad_norm": 4.605819225311279, + "learning_rate": 1.5611363365905006e-05, + "loss": 0.4916, + "step": 34971 + }, + { + "epoch": 0.8197987540802991, + "grad_norm": 1.5516546964645386, + "learning_rate": 1.560741148950695e-05, + "loss": 0.2025, + "step": 34972 + }, + { + "epoch": 0.8198221956551041, + "grad_norm": 2.7835428714752197, + "learning_rate": 1.5603460071025445e-05, + "loss": 0.2589, + "step": 34973 + }, + { + "epoch": 0.8198456372299091, + "grad_norm": 1.6700503826141357, + "learning_rate": 1.5599509110481946e-05, + "loss": 0.17, + "step": 34974 + }, + { + "epoch": 0.819869078804714, + "grad_norm": 4.275335788726807, + "learning_rate": 1.5595558607897876e-05, + "loss": 0.3444, + "step": 34975 + }, + { + "epoch": 0.8198925203795191, + "grad_norm": 2.7073111534118652, + "learning_rate": 1.559160856329468e-05, + "loss": 0.1792, + "step": 34976 + }, + { + "epoch": 0.8199159619543241, + "grad_norm": 2.809342384338379, + "learning_rate": 1.5587658976693764e-05, + "loss": 1.739, + "step": 34977 + }, + { + "epoch": 0.8199394035291291, + "grad_norm": 2.463425397872925, + "learning_rate": 1.5583709848116613e-05, + "loss": 0.195, + "step": 34978 + }, + { + "epoch": 0.8199628451039341, + "grad_norm": 3.450634002685547, + "learning_rate": 1.5579761177584617e-05, + "loss": 0.1983, + "step": 34979 + }, + { + "epoch": 0.8199862866787391, + "grad_norm": 2.843846321105957, + "learning_rate": 1.557581296511922e-05, + "loss": 0.938, + "step": 34980 + }, + { + "epoch": 0.8200097282535441, + "grad_norm": 0.9603383541107178, + "learning_rate": 1.557186521074183e-05, + "loss": 0.093, + "step": 34981 + }, + { + "epoch": 0.8200331698283491, + "grad_norm": 1.361000657081604, + "learning_rate": 1.556791791447385e-05, + "loss": 0.1227, + "step": 34982 + }, + { + "epoch": 0.8200566114031541, + "grad_norm": 3.2383673191070557, + "learning_rate": 1.5563971076336747e-05, + "loss": 0.1586, + "step": 34983 + }, + { + "epoch": 0.820080052977959, + "grad_norm": 3.245142936706543, + "learning_rate": 1.5560024696351894e-05, + "loss": 1.4241, + "step": 34984 + }, + { + "epoch": 0.8201034945527641, + "grad_norm": 1.3162331581115723, + "learning_rate": 1.555607877454074e-05, + "loss": 0.1453, + "step": 34985 + }, + { + "epoch": 0.820126936127569, + "grad_norm": 2.8391926288604736, + "learning_rate": 1.5552133310924667e-05, + "loss": 0.193, + "step": 34986 + }, + { + "epoch": 0.8201503777023741, + "grad_norm": 2.1130735874176025, + "learning_rate": 1.55481883055251e-05, + "loss": 0.1847, + "step": 34987 + }, + { + "epoch": 0.820173819277179, + "grad_norm": 4.036220550537109, + "learning_rate": 1.554424375836344e-05, + "loss": 0.3943, + "step": 34988 + }, + { + "epoch": 0.8201972608519841, + "grad_norm": 2.4187674522399902, + "learning_rate": 1.554029966946109e-05, + "loss": 0.2499, + "step": 34989 + }, + { + "epoch": 0.820220702426789, + "grad_norm": 1.2017539739608765, + "learning_rate": 1.5536356038839417e-05, + "loss": 0.085, + "step": 34990 + }, + { + "epoch": 0.8202441440015941, + "grad_norm": 0.46117714047431946, + "learning_rate": 1.5532412866519886e-05, + "loss": 0.0166, + "step": 34991 + }, + { + "epoch": 0.820267585576399, + "grad_norm": 1.8842954635620117, + "learning_rate": 1.5528470152523844e-05, + "loss": 0.1495, + "step": 34992 + }, + { + "epoch": 0.820291027151204, + "grad_norm": 2.69035267829895, + "learning_rate": 1.552452789687271e-05, + "loss": 1.4684, + "step": 34993 + }, + { + "epoch": 0.820314468726009, + "grad_norm": 0.31357496976852417, + "learning_rate": 1.5520586099587832e-05, + "loss": 0.0094, + "step": 34994 + }, + { + "epoch": 0.820337910300814, + "grad_norm": 2.6108057498931885, + "learning_rate": 1.5516644760690658e-05, + "loss": 0.1689, + "step": 34995 + }, + { + "epoch": 0.820361351875619, + "grad_norm": 3.041749954223633, + "learning_rate": 1.551270388020254e-05, + "loss": 0.1136, + "step": 34996 + }, + { + "epoch": 0.820384793450424, + "grad_norm": 2.9868226051330566, + "learning_rate": 1.550876345814485e-05, + "loss": 0.2162, + "step": 34997 + }, + { + "epoch": 0.820408235025229, + "grad_norm": 3.1037228107452393, + "learning_rate": 1.5504823494539e-05, + "loss": 0.1951, + "step": 34998 + }, + { + "epoch": 0.820431676600034, + "grad_norm": 3.533958673477173, + "learning_rate": 1.5500883989406358e-05, + "loss": 0.3256, + "step": 34999 + }, + { + "epoch": 0.820455118174839, + "grad_norm": 1.4837145805358887, + "learning_rate": 1.5496944942768298e-05, + "loss": 0.1211, + "step": 35000 + }, + { + "epoch": 0.820478559749644, + "grad_norm": 4.271894454956055, + "learning_rate": 1.549300635464619e-05, + "loss": 0.4209, + "step": 35001 + }, + { + "epoch": 0.820502001324449, + "grad_norm": 4.09445858001709, + "learning_rate": 1.5489068225061377e-05, + "loss": 0.3578, + "step": 35002 + }, + { + "epoch": 0.820525442899254, + "grad_norm": 7.9670729637146, + "learning_rate": 1.5485130554035286e-05, + "loss": 0.295, + "step": 35003 + }, + { + "epoch": 0.8205488844740589, + "grad_norm": 2.7819981575012207, + "learning_rate": 1.548119334158925e-05, + "loss": 0.1715, + "step": 35004 + }, + { + "epoch": 0.820572326048864, + "grad_norm": 2.4529402256011963, + "learning_rate": 1.5477256587744627e-05, + "loss": 0.1399, + "step": 35005 + }, + { + "epoch": 0.8205957676236689, + "grad_norm": 1.1639125347137451, + "learning_rate": 1.5473320292522775e-05, + "loss": 0.0726, + "step": 35006 + }, + { + "epoch": 0.820619209198474, + "grad_norm": 3.754229784011841, + "learning_rate": 1.546938445594508e-05, + "loss": 0.3785, + "step": 35007 + }, + { + "epoch": 0.8206426507732789, + "grad_norm": 2.282352924346924, + "learning_rate": 1.546544907803289e-05, + "loss": 0.1553, + "step": 35008 + }, + { + "epoch": 0.820666092348084, + "grad_norm": 2.229229211807251, + "learning_rate": 1.5461514158807523e-05, + "loss": 0.219, + "step": 35009 + }, + { + "epoch": 0.8206895339228889, + "grad_norm": 2.615510940551758, + "learning_rate": 1.545757969829037e-05, + "loss": 0.3263, + "step": 35010 + }, + { + "epoch": 0.820712975497694, + "grad_norm": 2.5606119632720947, + "learning_rate": 1.5453645696502782e-05, + "loss": 0.1939, + "step": 35011 + }, + { + "epoch": 0.8207364170724989, + "grad_norm": 0.8036302328109741, + "learning_rate": 1.5449712153466078e-05, + "loss": 0.0256, + "step": 35012 + }, + { + "epoch": 0.8207598586473039, + "grad_norm": 2.7091057300567627, + "learning_rate": 1.5445779069201626e-05, + "loss": 0.1866, + "step": 35013 + }, + { + "epoch": 0.8207833002221089, + "grad_norm": 3.1661155223846436, + "learning_rate": 1.5441846443730746e-05, + "loss": 0.1475, + "step": 35014 + }, + { + "epoch": 0.8208067417969139, + "grad_norm": 3.7852327823638916, + "learning_rate": 1.5437914277074762e-05, + "loss": 0.2315, + "step": 35015 + }, + { + "epoch": 0.8208301833717189, + "grad_norm": 0.5834289193153381, + "learning_rate": 1.5433982569255057e-05, + "loss": 0.0526, + "step": 35016 + }, + { + "epoch": 0.8208536249465239, + "grad_norm": 1.7933566570281982, + "learning_rate": 1.5430051320292938e-05, + "loss": 0.1499, + "step": 35017 + }, + { + "epoch": 0.8208770665213289, + "grad_norm": 6.546329975128174, + "learning_rate": 1.542612053020971e-05, + "loss": 0.3403, + "step": 35018 + }, + { + "epoch": 0.8209005080961339, + "grad_norm": 1.9580183029174805, + "learning_rate": 1.5422190199026765e-05, + "loss": 0.1335, + "step": 35019 + }, + { + "epoch": 0.8209239496709388, + "grad_norm": 0.32472044229507446, + "learning_rate": 1.5418260326765387e-05, + "loss": 0.0324, + "step": 35020 + }, + { + "epoch": 0.8209473912457439, + "grad_norm": 2.797999143600464, + "learning_rate": 1.5414330913446905e-05, + "loss": 0.2143, + "step": 35021 + }, + { + "epoch": 0.8209708328205488, + "grad_norm": 3.0446276664733887, + "learning_rate": 1.5410401959092625e-05, + "loss": 0.2688, + "step": 35022 + }, + { + "epoch": 0.8209942743953539, + "grad_norm": 2.9250669479370117, + "learning_rate": 1.5406473463723903e-05, + "loss": 0.1663, + "step": 35023 + }, + { + "epoch": 0.8210177159701588, + "grad_norm": 0.3078010082244873, + "learning_rate": 1.5402545427362037e-05, + "loss": 0.0218, + "step": 35024 + }, + { + "epoch": 0.8210411575449639, + "grad_norm": 0.9513726830482483, + "learning_rate": 1.5398617850028342e-05, + "loss": 0.0907, + "step": 35025 + }, + { + "epoch": 0.8210645991197688, + "grad_norm": 1.0832566022872925, + "learning_rate": 1.539469073174411e-05, + "loss": 0.0551, + "step": 35026 + }, + { + "epoch": 0.8210880406945739, + "grad_norm": 3.9168660640716553, + "learning_rate": 1.5390764072530682e-05, + "loss": 0.3308, + "step": 35027 + }, + { + "epoch": 0.8211114822693789, + "grad_norm": 4.268289566040039, + "learning_rate": 1.538683787240931e-05, + "loss": 0.4409, + "step": 35028 + }, + { + "epoch": 0.8211349238441839, + "grad_norm": 1.108397126197815, + "learning_rate": 1.538291213140136e-05, + "loss": 0.9334, + "step": 35029 + }, + { + "epoch": 0.8211583654189889, + "grad_norm": 0.839305579662323, + "learning_rate": 1.537898684952812e-05, + "loss": 0.0286, + "step": 35030 + }, + { + "epoch": 0.8211818069937938, + "grad_norm": 4.664501667022705, + "learning_rate": 1.537506202681084e-05, + "loss": 0.4368, + "step": 35031 + }, + { + "epoch": 0.8212052485685989, + "grad_norm": 1.8478341102600098, + "learning_rate": 1.5371137663270885e-05, + "loss": 0.2026, + "step": 35032 + }, + { + "epoch": 0.8212286901434038, + "grad_norm": 1.4486349821090698, + "learning_rate": 1.536721375892951e-05, + "loss": 0.1323, + "step": 35033 + }, + { + "epoch": 0.8212521317182089, + "grad_norm": 2.226142644882202, + "learning_rate": 1.536329031380802e-05, + "loss": 0.118, + "step": 35034 + }, + { + "epoch": 0.8212755732930138, + "grad_norm": 2.4988880157470703, + "learning_rate": 1.5359367327927664e-05, + "loss": 0.2212, + "step": 35035 + }, + { + "epoch": 0.8212990148678189, + "grad_norm": 0.8134183287620544, + "learning_rate": 1.5355444801309793e-05, + "loss": 0.0489, + "step": 35036 + }, + { + "epoch": 0.8213224564426238, + "grad_norm": 4.278689384460449, + "learning_rate": 1.5351522733975655e-05, + "loss": 0.3837, + "step": 35037 + }, + { + "epoch": 0.8213458980174289, + "grad_norm": 1.6336597204208374, + "learning_rate": 1.534760112594653e-05, + "loss": 0.1363, + "step": 35038 + }, + { + "epoch": 0.8213693395922338, + "grad_norm": 4.163861274719238, + "learning_rate": 1.5343679977243706e-05, + "loss": 0.2628, + "step": 35039 + }, + { + "epoch": 0.8213927811670388, + "grad_norm": 3.388359308242798, + "learning_rate": 1.5339759287888433e-05, + "loss": 0.3745, + "step": 35040 + }, + { + "epoch": 0.8214162227418438, + "grad_norm": 1.0194337368011475, + "learning_rate": 1.5335839057902034e-05, + "loss": 0.0723, + "step": 35041 + }, + { + "epoch": 0.8214396643166488, + "grad_norm": 4.812357425689697, + "learning_rate": 1.5331919287305753e-05, + "loss": 0.4176, + "step": 35042 + }, + { + "epoch": 0.8214631058914538, + "grad_norm": 1.3413472175598145, + "learning_rate": 1.5327999976120823e-05, + "loss": 0.1064, + "step": 35043 + }, + { + "epoch": 0.8214865474662588, + "grad_norm": 4.715148448944092, + "learning_rate": 1.5324081124368573e-05, + "loss": 0.2639, + "step": 35044 + }, + { + "epoch": 0.8215099890410638, + "grad_norm": 1.2133437395095825, + "learning_rate": 1.5320162732070243e-05, + "loss": 0.1068, + "step": 35045 + }, + { + "epoch": 0.8215334306158688, + "grad_norm": 5.294346332550049, + "learning_rate": 1.5316244799247093e-05, + "loss": 0.3478, + "step": 35046 + }, + { + "epoch": 0.8215568721906737, + "grad_norm": 1.032638669013977, + "learning_rate": 1.531232732592035e-05, + "loss": 0.0462, + "step": 35047 + }, + { + "epoch": 0.8215803137654788, + "grad_norm": 3.4088635444641113, + "learning_rate": 1.530841031211133e-05, + "loss": 0.3147, + "step": 35048 + }, + { + "epoch": 0.8216037553402837, + "grad_norm": 2.18078875541687, + "learning_rate": 1.5304493757841255e-05, + "loss": 0.162, + "step": 35049 + }, + { + "epoch": 0.8216271969150888, + "grad_norm": 0.6006251573562622, + "learning_rate": 1.5300577663131375e-05, + "loss": 0.0483, + "step": 35050 + }, + { + "epoch": 0.8216506384898937, + "grad_norm": 3.8161723613739014, + "learning_rate": 1.529666202800295e-05, + "loss": 0.3232, + "step": 35051 + }, + { + "epoch": 0.8216740800646988, + "grad_norm": 0.6998589634895325, + "learning_rate": 1.5292746852477214e-05, + "loss": 0.0448, + "step": 35052 + }, + { + "epoch": 0.8216975216395037, + "grad_norm": 3.6642446517944336, + "learning_rate": 1.5288832136575392e-05, + "loss": 0.4059, + "step": 35053 + }, + { + "epoch": 0.8217209632143088, + "grad_norm": 3.4653289318084717, + "learning_rate": 1.5284917880318774e-05, + "loss": 0.3794, + "step": 35054 + }, + { + "epoch": 0.8217444047891137, + "grad_norm": 2.2155964374542236, + "learning_rate": 1.5281004083728545e-05, + "loss": 0.1144, + "step": 35055 + }, + { + "epoch": 0.8217678463639188, + "grad_norm": 2.2210733890533447, + "learning_rate": 1.5277090746825994e-05, + "loss": 0.1358, + "step": 35056 + }, + { + "epoch": 0.8217912879387237, + "grad_norm": 0.9904513955116272, + "learning_rate": 1.5273177869632327e-05, + "loss": 0.0796, + "step": 35057 + }, + { + "epoch": 0.8218147295135287, + "grad_norm": 0.3547556400299072, + "learning_rate": 1.526926545216879e-05, + "loss": 0.0145, + "step": 35058 + }, + { + "epoch": 0.8218381710883337, + "grad_norm": 3.9150846004486084, + "learning_rate": 1.526535349445659e-05, + "loss": 0.374, + "step": 35059 + }, + { + "epoch": 0.8218616126631387, + "grad_norm": 2.130415916442871, + "learning_rate": 1.5261441996516957e-05, + "loss": 0.1463, + "step": 35060 + }, + { + "epoch": 0.8218850542379437, + "grad_norm": 1.9554815292358398, + "learning_rate": 1.5257530958371125e-05, + "loss": 1.2934, + "step": 35061 + }, + { + "epoch": 0.8219084958127487, + "grad_norm": 1.2766553163528442, + "learning_rate": 1.5253620380040323e-05, + "loss": 0.0688, + "step": 35062 + }, + { + "epoch": 0.8219319373875537, + "grad_norm": 1.0692691802978516, + "learning_rate": 1.524971026154577e-05, + "loss": 0.0849, + "step": 35063 + }, + { + "epoch": 0.8219553789623587, + "grad_norm": 2.6250224113464355, + "learning_rate": 1.5245800602908656e-05, + "loss": 0.3041, + "step": 35064 + }, + { + "epoch": 0.8219788205371636, + "grad_norm": 2.4453189373016357, + "learning_rate": 1.5241891404150222e-05, + "loss": 0.149, + "step": 35065 + }, + { + "epoch": 0.8220022621119687, + "grad_norm": 3.195390224456787, + "learning_rate": 1.5237982665291638e-05, + "loss": 0.3766, + "step": 35066 + }, + { + "epoch": 0.8220257036867736, + "grad_norm": 1.8922748565673828, + "learning_rate": 1.5234074386354147e-05, + "loss": 0.2316, + "step": 35067 + }, + { + "epoch": 0.8220491452615787, + "grad_norm": 0.5958972573280334, + "learning_rate": 1.5230166567358972e-05, + "loss": 0.0388, + "step": 35068 + }, + { + "epoch": 0.8220725868363836, + "grad_norm": 1.0191340446472168, + "learning_rate": 1.5226259208327297e-05, + "loss": 0.1199, + "step": 35069 + }, + { + "epoch": 0.8220960284111887, + "grad_norm": 0.3685397207736969, + "learning_rate": 1.522235230928033e-05, + "loss": 0.0274, + "step": 35070 + }, + { + "epoch": 0.8221194699859936, + "grad_norm": 2.6016805171966553, + "learning_rate": 1.521844587023925e-05, + "loss": 0.2013, + "step": 35071 + }, + { + "epoch": 0.8221429115607987, + "grad_norm": 1.9756220579147339, + "learning_rate": 1.5214539891225277e-05, + "loss": 0.1077, + "step": 35072 + }, + { + "epoch": 0.8221663531356036, + "grad_norm": 3.9104557037353516, + "learning_rate": 1.5210634372259568e-05, + "loss": 0.2072, + "step": 35073 + }, + { + "epoch": 0.8221897947104087, + "grad_norm": 2.863030433654785, + "learning_rate": 1.5206729313363366e-05, + "loss": 0.2713, + "step": 35074 + }, + { + "epoch": 0.8222132362852136, + "grad_norm": 1.8407936096191406, + "learning_rate": 1.520282471455784e-05, + "loss": 0.1427, + "step": 35075 + }, + { + "epoch": 0.8222366778600186, + "grad_norm": 2.183068037033081, + "learning_rate": 1.5198920575864163e-05, + "loss": 0.1756, + "step": 35076 + }, + { + "epoch": 0.8222601194348236, + "grad_norm": 3.673879861831665, + "learning_rate": 1.5195016897303527e-05, + "loss": 0.1803, + "step": 35077 + }, + { + "epoch": 0.8222835610096286, + "grad_norm": 3.9638516902923584, + "learning_rate": 1.5191113678897095e-05, + "loss": 0.2171, + "step": 35078 + }, + { + "epoch": 0.8223070025844337, + "grad_norm": 2.618077278137207, + "learning_rate": 1.5187210920666062e-05, + "loss": 0.1656, + "step": 35079 + }, + { + "epoch": 0.8223304441592386, + "grad_norm": 2.9091122150421143, + "learning_rate": 1.5183308622631632e-05, + "loss": 0.1089, + "step": 35080 + }, + { + "epoch": 0.8223538857340437, + "grad_norm": 2.4718446731567383, + "learning_rate": 1.5179406784814954e-05, + "loss": 0.2818, + "step": 35081 + }, + { + "epoch": 0.8223773273088486, + "grad_norm": 0.2957579791545868, + "learning_rate": 1.5175505407237201e-05, + "loss": 0.0288, + "step": 35082 + }, + { + "epoch": 0.8224007688836537, + "grad_norm": 2.3357479572296143, + "learning_rate": 1.5171604489919533e-05, + "loss": 0.1633, + "step": 35083 + }, + { + "epoch": 0.8224242104584586, + "grad_norm": 3.6799569129943848, + "learning_rate": 1.5167704032883135e-05, + "loss": 0.1757, + "step": 35084 + }, + { + "epoch": 0.8224476520332636, + "grad_norm": 1.0764801502227783, + "learning_rate": 1.5163804036149131e-05, + "loss": 0.0739, + "step": 35085 + }, + { + "epoch": 0.8224710936080686, + "grad_norm": 3.362349033355713, + "learning_rate": 1.515990449973873e-05, + "loss": 0.291, + "step": 35086 + }, + { + "epoch": 0.8224945351828736, + "grad_norm": 2.724259853363037, + "learning_rate": 1.515600542367308e-05, + "loss": 0.1426, + "step": 35087 + }, + { + "epoch": 0.8225179767576786, + "grad_norm": 2.8278234004974365, + "learning_rate": 1.5152106807973332e-05, + "loss": 0.3637, + "step": 35088 + }, + { + "epoch": 0.8225414183324836, + "grad_norm": 5.17609167098999, + "learning_rate": 1.5148208652660633e-05, + "loss": 0.3386, + "step": 35089 + }, + { + "epoch": 0.8225648599072886, + "grad_norm": 1.9926347732543945, + "learning_rate": 1.5144310957756148e-05, + "loss": 1.4865, + "step": 35090 + }, + { + "epoch": 0.8225883014820936, + "grad_norm": 0.7009405493736267, + "learning_rate": 1.5140413723280989e-05, + "loss": 0.0614, + "step": 35091 + }, + { + "epoch": 0.8226117430568985, + "grad_norm": 2.638777256011963, + "learning_rate": 1.513651694925633e-05, + "loss": 0.1943, + "step": 35092 + }, + { + "epoch": 0.8226351846317036, + "grad_norm": 5.839062213897705, + "learning_rate": 1.5132620635703332e-05, + "loss": 0.5044, + "step": 35093 + }, + { + "epoch": 0.8226586262065085, + "grad_norm": 3.5982627868652344, + "learning_rate": 1.5128724782643133e-05, + "loss": 1.4607, + "step": 35094 + }, + { + "epoch": 0.8226820677813136, + "grad_norm": 1.2217808961868286, + "learning_rate": 1.512482939009685e-05, + "loss": 0.0425, + "step": 35095 + }, + { + "epoch": 0.8227055093561185, + "grad_norm": 2.828932523727417, + "learning_rate": 1.5120934458085634e-05, + "loss": 0.2238, + "step": 35096 + }, + { + "epoch": 0.8227289509309236, + "grad_norm": 2.4832372665405273, + "learning_rate": 1.511703998663062e-05, + "loss": 0.1785, + "step": 35097 + }, + { + "epoch": 0.8227523925057285, + "grad_norm": 4.54820442199707, + "learning_rate": 1.5113145975752906e-05, + "loss": 0.4478, + "step": 35098 + }, + { + "epoch": 0.8227758340805336, + "grad_norm": 2.6888210773468018, + "learning_rate": 1.5109252425473675e-05, + "loss": 0.2999, + "step": 35099 + }, + { + "epoch": 0.8227992756553385, + "grad_norm": 4.599812984466553, + "learning_rate": 1.5105359335814017e-05, + "loss": 0.443, + "step": 35100 + }, + { + "epoch": 0.8228227172301436, + "grad_norm": 1.0042402744293213, + "learning_rate": 1.5101466706795075e-05, + "loss": 0.0697, + "step": 35101 + }, + { + "epoch": 0.8228461588049485, + "grad_norm": 3.01011061668396, + "learning_rate": 1.5097574538437963e-05, + "loss": 0.1814, + "step": 35102 + }, + { + "epoch": 0.8228696003797535, + "grad_norm": 0.9286180734634399, + "learning_rate": 1.5093682830763789e-05, + "loss": 0.0954, + "step": 35103 + }, + { + "epoch": 0.8228930419545585, + "grad_norm": 0.5236092209815979, + "learning_rate": 1.5089791583793667e-05, + "loss": 0.041, + "step": 35104 + }, + { + "epoch": 0.8229164835293635, + "grad_norm": 7.287298679351807, + "learning_rate": 1.5085900797548714e-05, + "loss": 0.4104, + "step": 35105 + }, + { + "epoch": 0.8229399251041685, + "grad_norm": 2.3415627479553223, + "learning_rate": 1.5082010472050079e-05, + "loss": 1.2793, + "step": 35106 + }, + { + "epoch": 0.8229633666789735, + "grad_norm": 2.746403932571411, + "learning_rate": 1.5078120607318846e-05, + "loss": 1.4839, + "step": 35107 + }, + { + "epoch": 0.8229868082537785, + "grad_norm": 0.588618814945221, + "learning_rate": 1.507423120337611e-05, + "loss": 0.0592, + "step": 35108 + }, + { + "epoch": 0.8230102498285835, + "grad_norm": 3.0430831909179688, + "learning_rate": 1.5070342260242987e-05, + "loss": 0.1854, + "step": 35109 + }, + { + "epoch": 0.8230336914033884, + "grad_norm": 3.7359652519226074, + "learning_rate": 1.506645377794056e-05, + "loss": 0.3224, + "step": 35110 + }, + { + "epoch": 0.8230571329781935, + "grad_norm": 4.350727081298828, + "learning_rate": 1.5062565756489955e-05, + "loss": 0.206, + "step": 35111 + }, + { + "epoch": 0.8230805745529984, + "grad_norm": 2.1156272888183594, + "learning_rate": 1.5058678195912268e-05, + "loss": 0.1063, + "step": 35112 + }, + { + "epoch": 0.8231040161278035, + "grad_norm": 0.716391384601593, + "learning_rate": 1.5054791096228571e-05, + "loss": 0.0405, + "step": 35113 + }, + { + "epoch": 0.8231274577026084, + "grad_norm": 0.40901869535446167, + "learning_rate": 1.5050904457459968e-05, + "loss": 0.0372, + "step": 35114 + }, + { + "epoch": 0.8231508992774135, + "grad_norm": 4.15488862991333, + "learning_rate": 1.5047018279627546e-05, + "loss": 0.2278, + "step": 35115 + }, + { + "epoch": 0.8231743408522184, + "grad_norm": 3.8936655521392822, + "learning_rate": 1.504313256275237e-05, + "loss": 0.2637, + "step": 35116 + }, + { + "epoch": 0.8231977824270235, + "grad_norm": 3.6697731018066406, + "learning_rate": 1.5039247306855542e-05, + "loss": 0.3909, + "step": 35117 + }, + { + "epoch": 0.8232212240018284, + "grad_norm": 3.1452784538269043, + "learning_rate": 1.5035362511958173e-05, + "loss": 0.2014, + "step": 35118 + }, + { + "epoch": 0.8232446655766334, + "grad_norm": 1.9740756750106812, + "learning_rate": 1.5031478178081315e-05, + "loss": 0.0804, + "step": 35119 + }, + { + "epoch": 0.8232681071514384, + "grad_norm": 1.9310773611068726, + "learning_rate": 1.5027594305246051e-05, + "loss": 0.0628, + "step": 35120 + }, + { + "epoch": 0.8232915487262434, + "grad_norm": 0.2444426417350769, + "learning_rate": 1.5023710893473442e-05, + "loss": 0.024, + "step": 35121 + }, + { + "epoch": 0.8233149903010484, + "grad_norm": 1.1052441596984863, + "learning_rate": 1.501982794278457e-05, + "loss": 0.0973, + "step": 35122 + }, + { + "epoch": 0.8233384318758534, + "grad_norm": 2.4035112857818604, + "learning_rate": 1.5015945453200475e-05, + "loss": 0.1116, + "step": 35123 + }, + { + "epoch": 0.8233618734506584, + "grad_norm": 0.5051367282867432, + "learning_rate": 1.5012063424742285e-05, + "loss": 0.0367, + "step": 35124 + }, + { + "epoch": 0.8233853150254634, + "grad_norm": 2.970018148422241, + "learning_rate": 1.5008181857431015e-05, + "loss": 0.2265, + "step": 35125 + }, + { + "epoch": 0.8234087566002684, + "grad_norm": 0.8862134218215942, + "learning_rate": 1.5004300751287748e-05, + "loss": 0.2276, + "step": 35126 + }, + { + "epoch": 0.8234321981750734, + "grad_norm": 2.7029776573181152, + "learning_rate": 1.5000420106333524e-05, + "loss": 1.2634, + "step": 35127 + }, + { + "epoch": 0.8234556397498783, + "grad_norm": 1.7653441429138184, + "learning_rate": 1.4996539922589392e-05, + "loss": 0.043, + "step": 35128 + }, + { + "epoch": 0.8234790813246834, + "grad_norm": 3.800243377685547, + "learning_rate": 1.4992660200076458e-05, + "loss": 0.3887, + "step": 35129 + }, + { + "epoch": 0.8235025228994884, + "grad_norm": 1.9860780239105225, + "learning_rate": 1.4988780938815706e-05, + "loss": 0.1364, + "step": 35130 + }, + { + "epoch": 0.8235259644742934, + "grad_norm": 1.2227916717529297, + "learning_rate": 1.498490213882825e-05, + "loss": 0.1352, + "step": 35131 + }, + { + "epoch": 0.8235494060490984, + "grad_norm": 1.412521481513977, + "learning_rate": 1.4981023800135097e-05, + "loss": 0.072, + "step": 35132 + }, + { + "epoch": 0.8235728476239034, + "grad_norm": 1.8737317323684692, + "learning_rate": 1.4977145922757297e-05, + "loss": 0.0713, + "step": 35133 + }, + { + "epoch": 0.8235962891987084, + "grad_norm": 1.373022437095642, + "learning_rate": 1.4973268506715899e-05, + "loss": 0.0946, + "step": 35134 + }, + { + "epoch": 0.8236197307735134, + "grad_norm": 2.905059337615967, + "learning_rate": 1.496939155203194e-05, + "loss": 0.2137, + "step": 35135 + }, + { + "epoch": 0.8236431723483184, + "grad_norm": 7.9861578941345215, + "learning_rate": 1.4965515058726431e-05, + "loss": 0.2235, + "step": 35136 + }, + { + "epoch": 0.8236666139231233, + "grad_norm": 2.2195160388946533, + "learning_rate": 1.496163902682045e-05, + "loss": 0.1978, + "step": 35137 + }, + { + "epoch": 0.8236900554979284, + "grad_norm": 1.9921729564666748, + "learning_rate": 1.4957763456335006e-05, + "loss": 0.121, + "step": 35138 + }, + { + "epoch": 0.8237134970727333, + "grad_norm": 1.706533432006836, + "learning_rate": 1.4953888347291123e-05, + "loss": 0.1213, + "step": 35139 + }, + { + "epoch": 0.8237369386475384, + "grad_norm": 3.5354537963867188, + "learning_rate": 1.4950013699709808e-05, + "loss": 0.31, + "step": 35140 + }, + { + "epoch": 0.8237603802223433, + "grad_norm": 2.912487030029297, + "learning_rate": 1.4946139513612145e-05, + "loss": 1.8208, + "step": 35141 + }, + { + "epoch": 0.8237838217971484, + "grad_norm": 2.037384510040283, + "learning_rate": 1.4942265789019117e-05, + "loss": 0.0859, + "step": 35142 + }, + { + "epoch": 0.8238072633719533, + "grad_norm": 1.9439537525177002, + "learning_rate": 1.4938392525951716e-05, + "loss": 0.1042, + "step": 35143 + }, + { + "epoch": 0.8238307049467584, + "grad_norm": 2.333972215652466, + "learning_rate": 1.4934519724431018e-05, + "loss": 0.2129, + "step": 35144 + }, + { + "epoch": 0.8238541465215633, + "grad_norm": 3.6971278190612793, + "learning_rate": 1.4930647384478013e-05, + "loss": 0.3825, + "step": 35145 + }, + { + "epoch": 0.8238775880963684, + "grad_norm": 2.7917983531951904, + "learning_rate": 1.49267755061137e-05, + "loss": 0.2632, + "step": 35146 + }, + { + "epoch": 0.8239010296711733, + "grad_norm": 2.4315876960754395, + "learning_rate": 1.4922904089359091e-05, + "loss": 0.2168, + "step": 35147 + }, + { + "epoch": 0.8239244712459783, + "grad_norm": 2.1282174587249756, + "learning_rate": 1.4919033134235184e-05, + "loss": 0.1232, + "step": 35148 + }, + { + "epoch": 0.8239479128207833, + "grad_norm": 2.5178565979003906, + "learning_rate": 1.4915162640763015e-05, + "loss": 0.0934, + "step": 35149 + }, + { + "epoch": 0.8239713543955883, + "grad_norm": 0.9587988257408142, + "learning_rate": 1.4911292608963557e-05, + "loss": 0.1077, + "step": 35150 + }, + { + "epoch": 0.8239947959703933, + "grad_norm": 3.2572669982910156, + "learning_rate": 1.4907423038857826e-05, + "loss": 0.2596, + "step": 35151 + }, + { + "epoch": 0.8240182375451983, + "grad_norm": 3.125073194503784, + "learning_rate": 1.4903553930466784e-05, + "loss": 0.3794, + "step": 35152 + }, + { + "epoch": 0.8240416791200033, + "grad_norm": 1.6085602045059204, + "learning_rate": 1.4899685283811482e-05, + "loss": 0.1051, + "step": 35153 + }, + { + "epoch": 0.8240651206948083, + "grad_norm": 1.1251587867736816, + "learning_rate": 1.489581709891288e-05, + "loss": 0.08, + "step": 35154 + }, + { + "epoch": 0.8240885622696132, + "grad_norm": 2.004284620285034, + "learning_rate": 1.4891949375791937e-05, + "loss": 0.1235, + "step": 35155 + }, + { + "epoch": 0.8241120038444183, + "grad_norm": 2.438964366912842, + "learning_rate": 1.4888082114469692e-05, + "loss": 0.0725, + "step": 35156 + }, + { + "epoch": 0.8241354454192232, + "grad_norm": 3.681196689605713, + "learning_rate": 1.488421531496711e-05, + "loss": 0.2344, + "step": 35157 + }, + { + "epoch": 0.8241588869940283, + "grad_norm": 0.7389715313911438, + "learning_rate": 1.488034897730517e-05, + "loss": 0.039, + "step": 35158 + }, + { + "epoch": 0.8241823285688332, + "grad_norm": 3.723691940307617, + "learning_rate": 1.4876483101504846e-05, + "loss": 0.2331, + "step": 35159 + }, + { + "epoch": 0.8242057701436383, + "grad_norm": 1.9277687072753906, + "learning_rate": 1.487261768758712e-05, + "loss": 0.1069, + "step": 35160 + }, + { + "epoch": 0.8242292117184432, + "grad_norm": 1.3210923671722412, + "learning_rate": 1.4868752735572943e-05, + "loss": 0.0954, + "step": 35161 + }, + { + "epoch": 0.8242526532932483, + "grad_norm": 0.39862850308418274, + "learning_rate": 1.4864888245483322e-05, + "loss": 0.043, + "step": 35162 + }, + { + "epoch": 0.8242760948680532, + "grad_norm": 2.2588813304901123, + "learning_rate": 1.4861024217339226e-05, + "loss": 0.0738, + "step": 35163 + }, + { + "epoch": 0.8242995364428582, + "grad_norm": 1.7580212354660034, + "learning_rate": 1.4857160651161572e-05, + "loss": 0.1671, + "step": 35164 + }, + { + "epoch": 0.8243229780176632, + "grad_norm": 2.5318822860717773, + "learning_rate": 1.485329754697139e-05, + "loss": 1.7903, + "step": 35165 + }, + { + "epoch": 0.8243464195924682, + "grad_norm": 2.8997230529785156, + "learning_rate": 1.4849434904789595e-05, + "loss": 0.1416, + "step": 35166 + }, + { + "epoch": 0.8243698611672732, + "grad_norm": 1.9753910303115845, + "learning_rate": 1.4845572724637168e-05, + "loss": 0.0833, + "step": 35167 + }, + { + "epoch": 0.8243933027420782, + "grad_norm": 4.039829730987549, + "learning_rate": 1.4841711006535041e-05, + "loss": 0.322, + "step": 35168 + }, + { + "epoch": 0.8244167443168832, + "grad_norm": 4.163285255432129, + "learning_rate": 1.4837849750504206e-05, + "loss": 0.4866, + "step": 35169 + }, + { + "epoch": 0.8244401858916882, + "grad_norm": 1.7001936435699463, + "learning_rate": 1.4833988956565593e-05, + "loss": 0.1104, + "step": 35170 + }, + { + "epoch": 0.8244636274664932, + "grad_norm": 1.0269267559051514, + "learning_rate": 1.4830128624740147e-05, + "loss": 0.0904, + "step": 35171 + }, + { + "epoch": 0.8244870690412982, + "grad_norm": 3.591884136199951, + "learning_rate": 1.4826268755048811e-05, + "loss": 0.3367, + "step": 35172 + }, + { + "epoch": 0.8245105106161031, + "grad_norm": 1.4939968585968018, + "learning_rate": 1.4822409347512545e-05, + "loss": 0.081, + "step": 35173 + }, + { + "epoch": 0.8245339521909082, + "grad_norm": 2.3339591026306152, + "learning_rate": 1.4818550402152253e-05, + "loss": 0.2299, + "step": 35174 + }, + { + "epoch": 0.8245573937657131, + "grad_norm": 3.212752342224121, + "learning_rate": 1.4814691918988932e-05, + "loss": 0.1712, + "step": 35175 + }, + { + "epoch": 0.8245808353405182, + "grad_norm": 4.481584072113037, + "learning_rate": 1.4810833898043464e-05, + "loss": 0.3558, + "step": 35176 + }, + { + "epoch": 0.8246042769153231, + "grad_norm": 3.1518681049346924, + "learning_rate": 1.480697633933682e-05, + "loss": 0.3918, + "step": 35177 + }, + { + "epoch": 0.8246277184901282, + "grad_norm": 2.0122554302215576, + "learning_rate": 1.4803119242889929e-05, + "loss": 0.2821, + "step": 35178 + }, + { + "epoch": 0.8246511600649331, + "grad_norm": 0.8418650031089783, + "learning_rate": 1.4799262608723697e-05, + "loss": 0.0891, + "step": 35179 + }, + { + "epoch": 0.8246746016397382, + "grad_norm": 0.7107909321784973, + "learning_rate": 1.4795406436859071e-05, + "loss": 0.0266, + "step": 35180 + }, + { + "epoch": 0.8246980432145432, + "grad_norm": 3.937922239303589, + "learning_rate": 1.4791550727316938e-05, + "loss": 0.2806, + "step": 35181 + }, + { + "epoch": 0.8247214847893481, + "grad_norm": 2.563772201538086, + "learning_rate": 1.4787695480118268e-05, + "loss": 0.1828, + "step": 35182 + }, + { + "epoch": 0.8247449263641532, + "grad_norm": 0.9724085927009583, + "learning_rate": 1.4783840695283968e-05, + "loss": 0.0355, + "step": 35183 + }, + { + "epoch": 0.8247683679389581, + "grad_norm": 1.996343970298767, + "learning_rate": 1.477998637283493e-05, + "loss": 0.0918, + "step": 35184 + }, + { + "epoch": 0.8247918095137632, + "grad_norm": 1.196927547454834, + "learning_rate": 1.4776132512792096e-05, + "loss": 0.1106, + "step": 35185 + }, + { + "epoch": 0.8248152510885681, + "grad_norm": 2.678025245666504, + "learning_rate": 1.4772279115176335e-05, + "loss": 0.1805, + "step": 35186 + }, + { + "epoch": 0.8248386926633732, + "grad_norm": 3.3223538398742676, + "learning_rate": 1.47684261800086e-05, + "loss": 1.4203, + "step": 35187 + }, + { + "epoch": 0.8248621342381781, + "grad_norm": 1.480829119682312, + "learning_rate": 1.4764573707309793e-05, + "loss": 0.0825, + "step": 35188 + }, + { + "epoch": 0.8248855758129832, + "grad_norm": 3.650818347930908, + "learning_rate": 1.4760721697100765e-05, + "loss": 0.4468, + "step": 35189 + }, + { + "epoch": 0.8249090173877881, + "grad_norm": 3.084721803665161, + "learning_rate": 1.4756870149402503e-05, + "loss": 1.8133, + "step": 35190 + }, + { + "epoch": 0.8249324589625932, + "grad_norm": 2.1868982315063477, + "learning_rate": 1.475301906423584e-05, + "loss": 0.0843, + "step": 35191 + }, + { + "epoch": 0.8249559005373981, + "grad_norm": 2.653520345687866, + "learning_rate": 1.4749168441621708e-05, + "loss": 0.1821, + "step": 35192 + }, + { + "epoch": 0.8249793421122031, + "grad_norm": 2.2541422843933105, + "learning_rate": 1.4745318281580955e-05, + "loss": 0.1502, + "step": 35193 + }, + { + "epoch": 0.8250027836870081, + "grad_norm": 1.3933497667312622, + "learning_rate": 1.4741468584134532e-05, + "loss": 0.0767, + "step": 35194 + }, + { + "epoch": 0.8250262252618131, + "grad_norm": 2.7995803356170654, + "learning_rate": 1.4737619349303289e-05, + "loss": 0.1808, + "step": 35195 + }, + { + "epoch": 0.8250496668366181, + "grad_norm": 2.1091785430908203, + "learning_rate": 1.4733770577108118e-05, + "loss": 1.6676, + "step": 35196 + }, + { + "epoch": 0.8250731084114231, + "grad_norm": 2.91015625, + "learning_rate": 1.4729922267569918e-05, + "loss": 0.1592, + "step": 35197 + }, + { + "epoch": 0.8250965499862281, + "grad_norm": 2.8037734031677246, + "learning_rate": 1.4726074420709545e-05, + "loss": 0.1785, + "step": 35198 + }, + { + "epoch": 0.8251199915610331, + "grad_norm": 0.8585876822471619, + "learning_rate": 1.4722227036547876e-05, + "loss": 0.0698, + "step": 35199 + }, + { + "epoch": 0.825143433135838, + "grad_norm": 2.231579065322876, + "learning_rate": 1.4718380115105823e-05, + "loss": 0.0963, + "step": 35200 + }, + { + "epoch": 0.8251668747106431, + "grad_norm": 0.6773978471755981, + "learning_rate": 1.4714533656404206e-05, + "loss": 0.0584, + "step": 35201 + }, + { + "epoch": 0.825190316285448, + "grad_norm": 1.3664766550064087, + "learning_rate": 1.4710687660463962e-05, + "loss": 0.1124, + "step": 35202 + }, + { + "epoch": 0.8252137578602531, + "grad_norm": 2.6861746311187744, + "learning_rate": 1.4706842127305919e-05, + "loss": 1.4149, + "step": 35203 + }, + { + "epoch": 0.825237199435058, + "grad_norm": 2.992042064666748, + "learning_rate": 1.4702997056950951e-05, + "loss": 0.2358, + "step": 35204 + }, + { + "epoch": 0.8252606410098631, + "grad_norm": 2.3513100147247314, + "learning_rate": 1.4699152449419917e-05, + "loss": 0.0737, + "step": 35205 + }, + { + "epoch": 0.825284082584668, + "grad_norm": 0.23221756517887115, + "learning_rate": 1.4695308304733657e-05, + "loss": 0.0089, + "step": 35206 + }, + { + "epoch": 0.8253075241594731, + "grad_norm": 2.575937032699585, + "learning_rate": 1.4691464622913076e-05, + "loss": 0.2409, + "step": 35207 + }, + { + "epoch": 0.825330965734278, + "grad_norm": 0.6339043974876404, + "learning_rate": 1.4687621403979002e-05, + "loss": 0.0508, + "step": 35208 + }, + { + "epoch": 0.825354407309083, + "grad_norm": 2.98423433303833, + "learning_rate": 1.4683778647952295e-05, + "loss": 0.2878, + "step": 35209 + }, + { + "epoch": 0.825377848883888, + "grad_norm": 1.2958539724349976, + "learning_rate": 1.4679936354853808e-05, + "loss": 0.4723, + "step": 35210 + }, + { + "epoch": 0.825401290458693, + "grad_norm": 3.110572099685669, + "learning_rate": 1.4676094524704387e-05, + "loss": 0.1215, + "step": 35211 + }, + { + "epoch": 0.825424732033498, + "grad_norm": 3.4021413326263428, + "learning_rate": 1.4672253157524852e-05, + "loss": 0.3101, + "step": 35212 + }, + { + "epoch": 0.825448173608303, + "grad_norm": 1.0194586515426636, + "learning_rate": 1.4668412253336061e-05, + "loss": 0.1048, + "step": 35213 + }, + { + "epoch": 0.825471615183108, + "grad_norm": 4.448056697845459, + "learning_rate": 1.46645718121589e-05, + "loss": 0.2301, + "step": 35214 + }, + { + "epoch": 0.825495056757913, + "grad_norm": 1.8636873960494995, + "learning_rate": 1.4660731834014163e-05, + "loss": 0.2019, + "step": 35215 + }, + { + "epoch": 0.825518498332718, + "grad_norm": 2.8602511882781982, + "learning_rate": 1.4656892318922699e-05, + "loss": 0.2307, + "step": 35216 + }, + { + "epoch": 0.825541939907523, + "grad_norm": 3.1412365436553955, + "learning_rate": 1.4653053266905337e-05, + "loss": 0.3445, + "step": 35217 + }, + { + "epoch": 0.8255653814823279, + "grad_norm": 3.3642475605010986, + "learning_rate": 1.4649214677982915e-05, + "loss": 0.2498, + "step": 35218 + }, + { + "epoch": 0.825588823057133, + "grad_norm": 2.7233426570892334, + "learning_rate": 1.4645376552176216e-05, + "loss": 0.3556, + "step": 35219 + }, + { + "epoch": 0.8256122646319379, + "grad_norm": 4.478315353393555, + "learning_rate": 1.4641538889506135e-05, + "loss": 0.5211, + "step": 35220 + }, + { + "epoch": 0.825635706206743, + "grad_norm": 2.925935745239258, + "learning_rate": 1.4637701689993466e-05, + "loss": 0.2784, + "step": 35221 + }, + { + "epoch": 0.8256591477815479, + "grad_norm": 3.23587965965271, + "learning_rate": 1.4633864953659027e-05, + "loss": 0.2627, + "step": 35222 + }, + { + "epoch": 0.825682589356353, + "grad_norm": 2.3587453365325928, + "learning_rate": 1.4630028680523633e-05, + "loss": 1.6207, + "step": 35223 + }, + { + "epoch": 0.8257060309311579, + "grad_norm": 1.656336784362793, + "learning_rate": 1.462619287060808e-05, + "loss": 0.6081, + "step": 35224 + }, + { + "epoch": 0.825729472505963, + "grad_norm": 1.801405906677246, + "learning_rate": 1.4622357523933239e-05, + "loss": 1.3109, + "step": 35225 + }, + { + "epoch": 0.8257529140807679, + "grad_norm": 2.2351081371307373, + "learning_rate": 1.461852264051985e-05, + "loss": 0.132, + "step": 35226 + }, + { + "epoch": 0.825776355655573, + "grad_norm": 2.480820894241333, + "learning_rate": 1.4614688220388784e-05, + "loss": 0.2455, + "step": 35227 + }, + { + "epoch": 0.8257997972303779, + "grad_norm": 3.5630247592926025, + "learning_rate": 1.4610854263560826e-05, + "loss": 0.4231, + "step": 35228 + }, + { + "epoch": 0.8258232388051829, + "grad_norm": 0.5387688279151917, + "learning_rate": 1.4607020770056767e-05, + "loss": 0.0439, + "step": 35229 + }, + { + "epoch": 0.8258466803799879, + "grad_norm": 1.7158693075180054, + "learning_rate": 1.4603187739897417e-05, + "loss": 0.0917, + "step": 35230 + }, + { + "epoch": 0.8258701219547929, + "grad_norm": 3.0242838859558105, + "learning_rate": 1.4599355173103558e-05, + "loss": 0.1621, + "step": 35231 + }, + { + "epoch": 0.825893563529598, + "grad_norm": 3.581076145172119, + "learning_rate": 1.4595523069696004e-05, + "loss": 0.1613, + "step": 35232 + }, + { + "epoch": 0.8259170051044029, + "grad_norm": 3.212357997894287, + "learning_rate": 1.4591691429695554e-05, + "loss": 0.3213, + "step": 35233 + }, + { + "epoch": 0.825940446679208, + "grad_norm": 13.108565330505371, + "learning_rate": 1.4587860253122986e-05, + "loss": 0.153, + "step": 35234 + }, + { + "epoch": 0.8259638882540129, + "grad_norm": 1.1079226732254028, + "learning_rate": 1.4584029539999088e-05, + "loss": 0.1349, + "step": 35235 + }, + { + "epoch": 0.825987329828818, + "grad_norm": 0.48997023701667786, + "learning_rate": 1.4580199290344653e-05, + "loss": 0.0394, + "step": 35236 + }, + { + "epoch": 0.8260107714036229, + "grad_norm": 3.5422496795654297, + "learning_rate": 1.4576369504180431e-05, + "loss": 0.3778, + "step": 35237 + }, + { + "epoch": 0.8260342129784279, + "grad_norm": 2.71583890914917, + "learning_rate": 1.4572540181527239e-05, + "loss": 0.3081, + "step": 35238 + }, + { + "epoch": 0.8260576545532329, + "grad_norm": 1.5126092433929443, + "learning_rate": 1.456871132240586e-05, + "loss": 0.104, + "step": 35239 + }, + { + "epoch": 0.8260810961280379, + "grad_norm": 2.6451148986816406, + "learning_rate": 1.4564882926837053e-05, + "loss": 0.1054, + "step": 35240 + }, + { + "epoch": 0.8261045377028429, + "grad_norm": 1.0089504718780518, + "learning_rate": 1.4561054994841594e-05, + "loss": 0.0477, + "step": 35241 + }, + { + "epoch": 0.8261279792776479, + "grad_norm": 1.2100096940994263, + "learning_rate": 1.4557227526440264e-05, + "loss": 0.1026, + "step": 35242 + }, + { + "epoch": 0.8261514208524529, + "grad_norm": 2.49904727935791, + "learning_rate": 1.4553400521653804e-05, + "loss": 0.3323, + "step": 35243 + }, + { + "epoch": 0.8261748624272579, + "grad_norm": 1.8622716665267944, + "learning_rate": 1.4549573980502985e-05, + "loss": 0.1604, + "step": 35244 + }, + { + "epoch": 0.8261983040020628, + "grad_norm": 2.5127696990966797, + "learning_rate": 1.45457479030086e-05, + "loss": 1.2648, + "step": 35245 + }, + { + "epoch": 0.8262217455768679, + "grad_norm": 1.9150447845458984, + "learning_rate": 1.4541922289191389e-05, + "loss": 0.1709, + "step": 35246 + }, + { + "epoch": 0.8262451871516728, + "grad_norm": 1.439179539680481, + "learning_rate": 1.4538097139072104e-05, + "loss": 0.112, + "step": 35247 + }, + { + "epoch": 0.8262686287264779, + "grad_norm": 0.5163447856903076, + "learning_rate": 1.4534272452671516e-05, + "loss": 0.0278, + "step": 35248 + }, + { + "epoch": 0.8262920703012828, + "grad_norm": 2.483450174331665, + "learning_rate": 1.453044823001034e-05, + "loss": 0.0891, + "step": 35249 + }, + { + "epoch": 0.8263155118760879, + "grad_norm": 3.732074022293091, + "learning_rate": 1.4526624471109384e-05, + "loss": 0.3239, + "step": 35250 + }, + { + "epoch": 0.8263389534508928, + "grad_norm": 0.8756126165390015, + "learning_rate": 1.4522801175989332e-05, + "loss": 0.0663, + "step": 35251 + }, + { + "epoch": 0.8263623950256979, + "grad_norm": 1.926385760307312, + "learning_rate": 1.4518978344670997e-05, + "loss": 0.2461, + "step": 35252 + }, + { + "epoch": 0.8263858366005028, + "grad_norm": 4.118419170379639, + "learning_rate": 1.4515155977175087e-05, + "loss": 0.2539, + "step": 35253 + }, + { + "epoch": 0.8264092781753078, + "grad_norm": 1.7174086570739746, + "learning_rate": 1.4511334073522342e-05, + "loss": 0.154, + "step": 35254 + }, + { + "epoch": 0.8264327197501128, + "grad_norm": 0.5009793043136597, + "learning_rate": 1.4507512633733511e-05, + "loss": 0.0357, + "step": 35255 + }, + { + "epoch": 0.8264561613249178, + "grad_norm": 0.4228331744670868, + "learning_rate": 1.450369165782931e-05, + "loss": 0.0404, + "step": 35256 + }, + { + "epoch": 0.8264796028997228, + "grad_norm": 3.0727086067199707, + "learning_rate": 1.4499871145830468e-05, + "loss": 0.2039, + "step": 35257 + }, + { + "epoch": 0.8265030444745278, + "grad_norm": 0.8186599016189575, + "learning_rate": 1.4496051097757745e-05, + "loss": 0.0501, + "step": 35258 + }, + { + "epoch": 0.8265264860493328, + "grad_norm": 5.653944492340088, + "learning_rate": 1.449223151363186e-05, + "loss": 0.2517, + "step": 35259 + }, + { + "epoch": 0.8265499276241378, + "grad_norm": 1.0032939910888672, + "learning_rate": 1.4488412393473527e-05, + "loss": 0.0969, + "step": 35260 + }, + { + "epoch": 0.8265733691989428, + "grad_norm": 4.059329509735107, + "learning_rate": 1.4484593737303464e-05, + "loss": 0.3989, + "step": 35261 + }, + { + "epoch": 0.8265968107737478, + "grad_norm": 2.2305617332458496, + "learning_rate": 1.4480775545142389e-05, + "loss": 1.1255, + "step": 35262 + }, + { + "epoch": 0.8266202523485527, + "grad_norm": 3.5976572036743164, + "learning_rate": 1.447695781701105e-05, + "loss": 0.41, + "step": 35263 + }, + { + "epoch": 0.8266436939233578, + "grad_norm": 4.134981632232666, + "learning_rate": 1.447314055293012e-05, + "loss": 0.315, + "step": 35264 + }, + { + "epoch": 0.8266671354981627, + "grad_norm": 2.0915591716766357, + "learning_rate": 1.446932375292035e-05, + "loss": 0.1532, + "step": 35265 + }, + { + "epoch": 0.8266905770729678, + "grad_norm": 2.2556393146514893, + "learning_rate": 1.4465507417002444e-05, + "loss": 0.2291, + "step": 35266 + }, + { + "epoch": 0.8267140186477727, + "grad_norm": 7.164727687835693, + "learning_rate": 1.4461691545197098e-05, + "loss": 0.5584, + "step": 35267 + }, + { + "epoch": 0.8267374602225778, + "grad_norm": 1.745322346687317, + "learning_rate": 1.4457876137525017e-05, + "loss": 0.1807, + "step": 35268 + }, + { + "epoch": 0.8267609017973827, + "grad_norm": 2.6309943199157715, + "learning_rate": 1.4454061194006885e-05, + "loss": 0.166, + "step": 35269 + }, + { + "epoch": 0.8267843433721878, + "grad_norm": 2.944885730743408, + "learning_rate": 1.445024671466344e-05, + "loss": 0.1565, + "step": 35270 + }, + { + "epoch": 0.8268077849469927, + "grad_norm": 0.2872968018054962, + "learning_rate": 1.4446432699515367e-05, + "loss": 0.0262, + "step": 35271 + }, + { + "epoch": 0.8268312265217977, + "grad_norm": 1.662754774093628, + "learning_rate": 1.4442619148583347e-05, + "loss": 0.0705, + "step": 35272 + }, + { + "epoch": 0.8268546680966027, + "grad_norm": 1.0363796949386597, + "learning_rate": 1.4438806061888089e-05, + "loss": 0.0916, + "step": 35273 + }, + { + "epoch": 0.8268781096714077, + "grad_norm": 1.8559094667434692, + "learning_rate": 1.443499343945025e-05, + "loss": 0.0683, + "step": 35274 + }, + { + "epoch": 0.8269015512462127, + "grad_norm": 2.2636971473693848, + "learning_rate": 1.4431181281290563e-05, + "loss": 0.1432, + "step": 35275 + }, + { + "epoch": 0.8269249928210177, + "grad_norm": 0.8601071834564209, + "learning_rate": 1.4427369587429674e-05, + "loss": 0.0677, + "step": 35276 + }, + { + "epoch": 0.8269484343958227, + "grad_norm": 2.154052972793579, + "learning_rate": 1.44235583578883e-05, + "loss": 0.1498, + "step": 35277 + }, + { + "epoch": 0.8269718759706277, + "grad_norm": 5.083555698394775, + "learning_rate": 1.4419747592687116e-05, + "loss": 0.3578, + "step": 35278 + }, + { + "epoch": 0.8269953175454327, + "grad_norm": 2.023615598678589, + "learning_rate": 1.4415937291846782e-05, + "loss": 0.1318, + "step": 35279 + }, + { + "epoch": 0.8270187591202377, + "grad_norm": 3.221748113632202, + "learning_rate": 1.441212745538797e-05, + "loss": 0.1741, + "step": 35280 + }, + { + "epoch": 0.8270422006950426, + "grad_norm": 2.502784013748169, + "learning_rate": 1.4408318083331374e-05, + "loss": 0.1848, + "step": 35281 + }, + { + "epoch": 0.8270656422698477, + "grad_norm": 0.5327817797660828, + "learning_rate": 1.4404509175697622e-05, + "loss": 0.0309, + "step": 35282 + }, + { + "epoch": 0.8270890838446526, + "grad_norm": 2.8530492782592773, + "learning_rate": 1.4400700732507433e-05, + "loss": 0.1242, + "step": 35283 + }, + { + "epoch": 0.8271125254194577, + "grad_norm": 0.4233769178390503, + "learning_rate": 1.4396892753781454e-05, + "loss": 0.0294, + "step": 35284 + }, + { + "epoch": 0.8271359669942627, + "grad_norm": 2.1326913833618164, + "learning_rate": 1.4393085239540326e-05, + "loss": 0.1238, + "step": 35285 + }, + { + "epoch": 0.8271594085690677, + "grad_norm": 2.5309994220733643, + "learning_rate": 1.4389278189804723e-05, + "loss": 0.2038, + "step": 35286 + }, + { + "epoch": 0.8271828501438727, + "grad_norm": 3.143038511276245, + "learning_rate": 1.4385471604595313e-05, + "loss": 0.1415, + "step": 35287 + }, + { + "epoch": 0.8272062917186777, + "grad_norm": 4.568894386291504, + "learning_rate": 1.4381665483932738e-05, + "loss": 0.4353, + "step": 35288 + }, + { + "epoch": 0.8272297332934827, + "grad_norm": 3.1283514499664307, + "learning_rate": 1.4377859827837647e-05, + "loss": 0.2888, + "step": 35289 + }, + { + "epoch": 0.8272531748682876, + "grad_norm": 3.889888048171997, + "learning_rate": 1.4374054636330702e-05, + "loss": 0.203, + "step": 35290 + }, + { + "epoch": 0.8272766164430927, + "grad_norm": 5.557178974151611, + "learning_rate": 1.4370249909432554e-05, + "loss": 0.3613, + "step": 35291 + }, + { + "epoch": 0.8273000580178976, + "grad_norm": 0.49800193309783936, + "learning_rate": 1.436644564716384e-05, + "loss": 0.0308, + "step": 35292 + }, + { + "epoch": 0.8273234995927027, + "grad_norm": 3.9181737899780273, + "learning_rate": 1.4362641849545189e-05, + "loss": 0.5162, + "step": 35293 + }, + { + "epoch": 0.8273469411675076, + "grad_norm": 1.8500703573226929, + "learning_rate": 1.4358838516597262e-05, + "loss": 0.1548, + "step": 35294 + }, + { + "epoch": 0.8273703827423127, + "grad_norm": 1.2801096439361572, + "learning_rate": 1.4355035648340654e-05, + "loss": 0.1561, + "step": 35295 + }, + { + "epoch": 0.8273938243171176, + "grad_norm": 0.8623959422111511, + "learning_rate": 1.4351233244796058e-05, + "loss": 0.0452, + "step": 35296 + }, + { + "epoch": 0.8274172658919227, + "grad_norm": 1.9257926940917969, + "learning_rate": 1.4347431305984071e-05, + "loss": 0.0729, + "step": 35297 + }, + { + "epoch": 0.8274407074667276, + "grad_norm": 1.4694290161132812, + "learning_rate": 1.434362983192532e-05, + "loss": 0.0817, + "step": 35298 + }, + { + "epoch": 0.8274641490415326, + "grad_norm": 1.0326509475708008, + "learning_rate": 1.4339828822640456e-05, + "loss": 0.1029, + "step": 35299 + }, + { + "epoch": 0.8274875906163376, + "grad_norm": 0.7402521967887878, + "learning_rate": 1.4336028278150093e-05, + "loss": 0.071, + "step": 35300 + }, + { + "epoch": 0.8275110321911426, + "grad_norm": 1.4487794637680054, + "learning_rate": 1.433222819847485e-05, + "loss": 0.7499, + "step": 35301 + }, + { + "epoch": 0.8275344737659476, + "grad_norm": 1.2727309465408325, + "learning_rate": 1.4328428583635311e-05, + "loss": 0.0826, + "step": 35302 + }, + { + "epoch": 0.8275579153407526, + "grad_norm": 2.673367977142334, + "learning_rate": 1.4324629433652171e-05, + "loss": 0.2916, + "step": 35303 + }, + { + "epoch": 0.8275813569155576, + "grad_norm": 2.9529666900634766, + "learning_rate": 1.4320830748545988e-05, + "loss": 0.2777, + "step": 35304 + }, + { + "epoch": 0.8276047984903626, + "grad_norm": 0.7190320491790771, + "learning_rate": 1.4317032528337381e-05, + "loss": 0.0771, + "step": 35305 + }, + { + "epoch": 0.8276282400651676, + "grad_norm": 3.8060550689697266, + "learning_rate": 1.431323477304698e-05, + "loss": 0.3775, + "step": 35306 + }, + { + "epoch": 0.8276516816399726, + "grad_norm": 2.811028003692627, + "learning_rate": 1.4309437482695342e-05, + "loss": 0.1102, + "step": 35307 + }, + { + "epoch": 0.8276751232147775, + "grad_norm": 0.8739011287689209, + "learning_rate": 1.4305640657303132e-05, + "loss": 0.0757, + "step": 35308 + }, + { + "epoch": 0.8276985647895826, + "grad_norm": 3.392411947250366, + "learning_rate": 1.4301844296890921e-05, + "loss": 0.2436, + "step": 35309 + }, + { + "epoch": 0.8277220063643875, + "grad_norm": 2.4830336570739746, + "learning_rate": 1.4298048401479291e-05, + "loss": 0.1543, + "step": 35310 + }, + { + "epoch": 0.8277454479391926, + "grad_norm": 1.4632502794265747, + "learning_rate": 1.4294252971088895e-05, + "loss": 0.1365, + "step": 35311 + }, + { + "epoch": 0.8277688895139975, + "grad_norm": 2.9212212562561035, + "learning_rate": 1.4290458005740281e-05, + "loss": 0.2391, + "step": 35312 + }, + { + "epoch": 0.8277923310888026, + "grad_norm": 2.1790072917938232, + "learning_rate": 1.4286663505454057e-05, + "loss": 0.1767, + "step": 35313 + }, + { + "epoch": 0.8278157726636075, + "grad_norm": 3.1030197143554688, + "learning_rate": 1.428286947025077e-05, + "loss": 0.4131, + "step": 35314 + }, + { + "epoch": 0.8278392142384126, + "grad_norm": 0.7613886594772339, + "learning_rate": 1.4279075900151085e-05, + "loss": 0.061, + "step": 35315 + }, + { + "epoch": 0.8278626558132175, + "grad_norm": 3.8991687297821045, + "learning_rate": 1.4275282795175538e-05, + "loss": 0.2349, + "step": 35316 + }, + { + "epoch": 0.8278860973880225, + "grad_norm": 0.8411162495613098, + "learning_rate": 1.4271490155344714e-05, + "loss": 0.064, + "step": 35317 + }, + { + "epoch": 0.8279095389628275, + "grad_norm": 3.247476816177368, + "learning_rate": 1.4267697980679196e-05, + "loss": 0.1727, + "step": 35318 + }, + { + "epoch": 0.8279329805376325, + "grad_norm": 2.660597085952759, + "learning_rate": 1.4263906271199557e-05, + "loss": 0.1533, + "step": 35319 + }, + { + "epoch": 0.8279564221124375, + "grad_norm": 2.367234230041504, + "learning_rate": 1.4260115026926358e-05, + "loss": 0.1205, + "step": 35320 + }, + { + "epoch": 0.8279798636872425, + "grad_norm": 2.398420810699463, + "learning_rate": 1.4256324247880192e-05, + "loss": 0.2375, + "step": 35321 + }, + { + "epoch": 0.8280033052620475, + "grad_norm": 3.786797046661377, + "learning_rate": 1.4252533934081613e-05, + "loss": 0.3743, + "step": 35322 + }, + { + "epoch": 0.8280267468368525, + "grad_norm": 2.9415535926818848, + "learning_rate": 1.4248744085551214e-05, + "loss": 0.1318, + "step": 35323 + }, + { + "epoch": 0.8280501884116575, + "grad_norm": 5.621253967285156, + "learning_rate": 1.4244954702309542e-05, + "loss": 0.4912, + "step": 35324 + }, + { + "epoch": 0.8280736299864625, + "grad_norm": 2.702608823776245, + "learning_rate": 1.4241165784377164e-05, + "loss": 0.0932, + "step": 35325 + }, + { + "epoch": 0.8280970715612674, + "grad_norm": 0.605430006980896, + "learning_rate": 1.4237377331774625e-05, + "loss": 0.0385, + "step": 35326 + }, + { + "epoch": 0.8281205131360725, + "grad_norm": 0.5084793567657471, + "learning_rate": 1.423358934452247e-05, + "loss": 0.0417, + "step": 35327 + }, + { + "epoch": 0.8281439547108774, + "grad_norm": 0.5731228590011597, + "learning_rate": 1.4229801822641287e-05, + "loss": 0.0437, + "step": 35328 + }, + { + "epoch": 0.8281673962856825, + "grad_norm": 4.253736972808838, + "learning_rate": 1.4226014766151608e-05, + "loss": 0.7933, + "step": 35329 + }, + { + "epoch": 0.8281908378604874, + "grad_norm": 2.5285403728485107, + "learning_rate": 1.4222228175073993e-05, + "loss": 0.2042, + "step": 35330 + }, + { + "epoch": 0.8282142794352925, + "grad_norm": 2.0557641983032227, + "learning_rate": 1.4218442049428971e-05, + "loss": 0.2335, + "step": 35331 + }, + { + "epoch": 0.8282377210100974, + "grad_norm": 3.0304527282714844, + "learning_rate": 1.4214656389237102e-05, + "loss": 0.3055, + "step": 35332 + }, + { + "epoch": 0.8282611625849025, + "grad_norm": 0.7235838174819946, + "learning_rate": 1.4210871194518892e-05, + "loss": 0.0346, + "step": 35333 + }, + { + "epoch": 0.8282846041597074, + "grad_norm": 3.1027121543884277, + "learning_rate": 1.4207086465294939e-05, + "loss": 0.2998, + "step": 35334 + }, + { + "epoch": 0.8283080457345124, + "grad_norm": 3.1028499603271484, + "learning_rate": 1.420330220158571e-05, + "loss": 0.2744, + "step": 35335 + }, + { + "epoch": 0.8283314873093175, + "grad_norm": 2.82242751121521, + "learning_rate": 1.4199518403411816e-05, + "loss": 0.2736, + "step": 35336 + }, + { + "epoch": 0.8283549288841224, + "grad_norm": 1.374498963356018, + "learning_rate": 1.4195735070793736e-05, + "loss": 0.1007, + "step": 35337 + }, + { + "epoch": 0.8283783704589275, + "grad_norm": 2.939079999923706, + "learning_rate": 1.4191952203752013e-05, + "loss": 0.1309, + "step": 35338 + }, + { + "epoch": 0.8284018120337324, + "grad_norm": 2.3382177352905273, + "learning_rate": 1.418816980230716e-05, + "loss": 1.5379, + "step": 35339 + }, + { + "epoch": 0.8284252536085375, + "grad_norm": 1.8690316677093506, + "learning_rate": 1.4184387866479698e-05, + "loss": 0.1002, + "step": 35340 + }, + { + "epoch": 0.8284486951833424, + "grad_norm": 2.6913864612579346, + "learning_rate": 1.4180606396290186e-05, + "loss": 0.1972, + "step": 35341 + }, + { + "epoch": 0.8284721367581475, + "grad_norm": 0.9017285108566284, + "learning_rate": 1.4176825391759107e-05, + "loss": 0.0536, + "step": 35342 + }, + { + "epoch": 0.8284955783329524, + "grad_norm": 1.0741790533065796, + "learning_rate": 1.4173044852907002e-05, + "loss": 0.1004, + "step": 35343 + }, + { + "epoch": 0.8285190199077574, + "grad_norm": 3.0093469619750977, + "learning_rate": 1.4169264779754355e-05, + "loss": 0.2739, + "step": 35344 + }, + { + "epoch": 0.8285424614825624, + "grad_norm": 3.375791549682617, + "learning_rate": 1.416548517232167e-05, + "loss": 0.126, + "step": 35345 + }, + { + "epoch": 0.8285659030573674, + "grad_norm": 2.00998854637146, + "learning_rate": 1.4161706030629508e-05, + "loss": 0.1735, + "step": 35346 + }, + { + "epoch": 0.8285893446321724, + "grad_norm": 2.245988368988037, + "learning_rate": 1.4157927354698309e-05, + "loss": 0.2321, + "step": 35347 + }, + { + "epoch": 0.8286127862069774, + "grad_norm": 0.7389955520629883, + "learning_rate": 1.4154149144548634e-05, + "loss": 0.0327, + "step": 35348 + }, + { + "epoch": 0.8286362277817824, + "grad_norm": 2.5136208534240723, + "learning_rate": 1.4150371400200957e-05, + "loss": 0.1714, + "step": 35349 + }, + { + "epoch": 0.8286596693565874, + "grad_norm": 1.716691493988037, + "learning_rate": 1.4146594121675783e-05, + "loss": 0.1724, + "step": 35350 + }, + { + "epoch": 0.8286831109313924, + "grad_norm": 2.2590954303741455, + "learning_rate": 1.4142817308993605e-05, + "loss": 0.1276, + "step": 35351 + }, + { + "epoch": 0.8287065525061974, + "grad_norm": 3.2940714359283447, + "learning_rate": 1.4139040962174888e-05, + "loss": 0.2574, + "step": 35352 + }, + { + "epoch": 0.8287299940810023, + "grad_norm": 0.4921230971813202, + "learning_rate": 1.413526508124018e-05, + "loss": 0.0177, + "step": 35353 + }, + { + "epoch": 0.8287534356558074, + "grad_norm": 2.4967761039733887, + "learning_rate": 1.413148966620993e-05, + "loss": 0.1241, + "step": 35354 + }, + { + "epoch": 0.8287768772306123, + "grad_norm": 3.809880495071411, + "learning_rate": 1.4127714717104634e-05, + "loss": 0.212, + "step": 35355 + }, + { + "epoch": 0.8288003188054174, + "grad_norm": 2.2622532844543457, + "learning_rate": 1.4123940233944777e-05, + "loss": 0.1103, + "step": 35356 + }, + { + "epoch": 0.8288237603802223, + "grad_norm": 1.8687094449996948, + "learning_rate": 1.4120166216750818e-05, + "loss": 0.1279, + "step": 35357 + }, + { + "epoch": 0.8288472019550274, + "grad_norm": 4.061148166656494, + "learning_rate": 1.4116392665543244e-05, + "loss": 0.2604, + "step": 35358 + }, + { + "epoch": 0.8288706435298323, + "grad_norm": 2.763827085494995, + "learning_rate": 1.4112619580342535e-05, + "loss": 0.244, + "step": 35359 + }, + { + "epoch": 0.8288940851046374, + "grad_norm": 2.063455820083618, + "learning_rate": 1.4108846961169197e-05, + "loss": 0.1052, + "step": 35360 + }, + { + "epoch": 0.8289175266794423, + "grad_norm": 1.3997125625610352, + "learning_rate": 1.4105074808043662e-05, + "loss": 0.0569, + "step": 35361 + }, + { + "epoch": 0.8289409682542473, + "grad_norm": 1.0501452684402466, + "learning_rate": 1.4101303120986408e-05, + "loss": 0.0866, + "step": 35362 + }, + { + "epoch": 0.8289644098290523, + "grad_norm": 1.6008542776107788, + "learning_rate": 1.409753190001789e-05, + "loss": 0.126, + "step": 35363 + }, + { + "epoch": 0.8289878514038573, + "grad_norm": 2.8855412006378174, + "learning_rate": 1.4093761145158591e-05, + "loss": 0.3111, + "step": 35364 + }, + { + "epoch": 0.8290112929786623, + "grad_norm": 1.4053288698196411, + "learning_rate": 1.4089990856428936e-05, + "loss": 0.0674, + "step": 35365 + }, + { + "epoch": 0.8290347345534673, + "grad_norm": 2.786175012588501, + "learning_rate": 1.4086221033849422e-05, + "loss": 0.266, + "step": 35366 + }, + { + "epoch": 0.8290581761282723, + "grad_norm": 2.092816114425659, + "learning_rate": 1.4082451677440489e-05, + "loss": 0.1386, + "step": 35367 + }, + { + "epoch": 0.8290816177030773, + "grad_norm": 0.9901809096336365, + "learning_rate": 1.4078682787222598e-05, + "loss": 0.1157, + "step": 35368 + }, + { + "epoch": 0.8291050592778823, + "grad_norm": 0.46947941184043884, + "learning_rate": 1.4074914363216184e-05, + "loss": 0.027, + "step": 35369 + }, + { + "epoch": 0.8291285008526873, + "grad_norm": 4.046176433563232, + "learning_rate": 1.4071146405441704e-05, + "loss": 0.3644, + "step": 35370 + }, + { + "epoch": 0.8291519424274922, + "grad_norm": 2.8512368202209473, + "learning_rate": 1.4067378913919571e-05, + "loss": 0.1809, + "step": 35371 + }, + { + "epoch": 0.8291753840022973, + "grad_norm": 3.2629542350769043, + "learning_rate": 1.4063611888670258e-05, + "loss": 0.4693, + "step": 35372 + }, + { + "epoch": 0.8291988255771022, + "grad_norm": 2.498198986053467, + "learning_rate": 1.4059845329714228e-05, + "loss": 1.2091, + "step": 35373 + }, + { + "epoch": 0.8292222671519073, + "grad_norm": 0.5880749225616455, + "learning_rate": 1.4056079237071894e-05, + "loss": 0.0341, + "step": 35374 + }, + { + "epoch": 0.8292457087267122, + "grad_norm": 3.7564539909362793, + "learning_rate": 1.405231361076369e-05, + "loss": 0.3862, + "step": 35375 + }, + { + "epoch": 0.8292691503015173, + "grad_norm": 3.307525157928467, + "learning_rate": 1.4048548450810051e-05, + "loss": 0.3461, + "step": 35376 + }, + { + "epoch": 0.8292925918763222, + "grad_norm": 3.196925163269043, + "learning_rate": 1.4044783757231395e-05, + "loss": 0.1999, + "step": 35377 + }, + { + "epoch": 0.8293160334511273, + "grad_norm": 2.3361012935638428, + "learning_rate": 1.404101953004815e-05, + "loss": 0.1651, + "step": 35378 + }, + { + "epoch": 0.8293394750259322, + "grad_norm": 2.4213039875030518, + "learning_rate": 1.4037255769280766e-05, + "loss": 0.2854, + "step": 35379 + }, + { + "epoch": 0.8293629166007372, + "grad_norm": 3.3613760471343994, + "learning_rate": 1.403349247494965e-05, + "loss": 0.3418, + "step": 35380 + }, + { + "epoch": 0.8293863581755422, + "grad_norm": 2.271751880645752, + "learning_rate": 1.4029729647075218e-05, + "loss": 0.2177, + "step": 35381 + }, + { + "epoch": 0.8294097997503472, + "grad_norm": 2.220987558364868, + "learning_rate": 1.4025967285677888e-05, + "loss": 0.2214, + "step": 35382 + }, + { + "epoch": 0.8294332413251522, + "grad_norm": 3.8295140266418457, + "learning_rate": 1.4022205390778065e-05, + "loss": 0.2881, + "step": 35383 + }, + { + "epoch": 0.8294566828999572, + "grad_norm": 2.608372688293457, + "learning_rate": 1.4018443962396165e-05, + "loss": 0.189, + "step": 35384 + }, + { + "epoch": 0.8294801244747622, + "grad_norm": 3.2714650630950928, + "learning_rate": 1.4014683000552631e-05, + "loss": 0.1132, + "step": 35385 + }, + { + "epoch": 0.8295035660495672, + "grad_norm": 0.7267898917198181, + "learning_rate": 1.4010922505267843e-05, + "loss": 0.0309, + "step": 35386 + }, + { + "epoch": 0.8295270076243723, + "grad_norm": 0.7518488764762878, + "learning_rate": 1.4007162476562208e-05, + "loss": 0.0385, + "step": 35387 + }, + { + "epoch": 0.8295504491991772, + "grad_norm": 3.0709750652313232, + "learning_rate": 1.4003402914456132e-05, + "loss": 0.1325, + "step": 35388 + }, + { + "epoch": 0.8295738907739822, + "grad_norm": 3.5944111347198486, + "learning_rate": 1.3999643818969999e-05, + "loss": 0.2249, + "step": 35389 + }, + { + "epoch": 0.8295973323487872, + "grad_norm": 1.257583737373352, + "learning_rate": 1.3995885190124202e-05, + "loss": 1.0257, + "step": 35390 + }, + { + "epoch": 0.8296207739235922, + "grad_norm": 2.5827138423919678, + "learning_rate": 1.3992127027939173e-05, + "loss": 0.1205, + "step": 35391 + }, + { + "epoch": 0.8296442154983972, + "grad_norm": 3.312553644180298, + "learning_rate": 1.398836933243528e-05, + "loss": 0.2505, + "step": 35392 + }, + { + "epoch": 0.8296676570732022, + "grad_norm": 2.6106255054473877, + "learning_rate": 1.398461210363291e-05, + "loss": 0.1884, + "step": 35393 + }, + { + "epoch": 0.8296910986480072, + "grad_norm": 2.3703689575195312, + "learning_rate": 1.3980855341552457e-05, + "loss": 0.1618, + "step": 35394 + }, + { + "epoch": 0.8297145402228122, + "grad_norm": 1.8419389724731445, + "learning_rate": 1.3977099046214281e-05, + "loss": 0.1368, + "step": 35395 + }, + { + "epoch": 0.8297379817976172, + "grad_norm": 1.5120465755462646, + "learning_rate": 1.3973343217638802e-05, + "loss": 0.0949, + "step": 35396 + }, + { + "epoch": 0.8297614233724222, + "grad_norm": 3.966189384460449, + "learning_rate": 1.3969587855846366e-05, + "loss": 0.3565, + "step": 35397 + }, + { + "epoch": 0.8297848649472271, + "grad_norm": 3.2261195182800293, + "learning_rate": 1.3965832960857384e-05, + "loss": 0.1497, + "step": 35398 + }, + { + "epoch": 0.8298083065220322, + "grad_norm": 3.7199606895446777, + "learning_rate": 1.3962078532692201e-05, + "loss": 0.229, + "step": 35399 + }, + { + "epoch": 0.8298317480968371, + "grad_norm": 1.0173580646514893, + "learning_rate": 1.3958324571371218e-05, + "loss": 0.1133, + "step": 35400 + }, + { + "epoch": 0.8298551896716422, + "grad_norm": 1.9833842515945435, + "learning_rate": 1.3954571076914768e-05, + "loss": 0.0693, + "step": 35401 + }, + { + "epoch": 0.8298786312464471, + "grad_norm": 3.333691358566284, + "learning_rate": 1.395081804934325e-05, + "loss": 0.2072, + "step": 35402 + }, + { + "epoch": 0.8299020728212522, + "grad_norm": 2.0389561653137207, + "learning_rate": 1.394706548867698e-05, + "loss": 0.1329, + "step": 35403 + }, + { + "epoch": 0.8299255143960571, + "grad_norm": 2.3430707454681396, + "learning_rate": 1.3943313394936376e-05, + "loss": 0.1151, + "step": 35404 + }, + { + "epoch": 0.8299489559708622, + "grad_norm": 2.0120410919189453, + "learning_rate": 1.3939561768141774e-05, + "loss": 0.0921, + "step": 35405 + }, + { + "epoch": 0.8299723975456671, + "grad_norm": 3.5937459468841553, + "learning_rate": 1.3935810608313527e-05, + "loss": 0.1768, + "step": 35406 + }, + { + "epoch": 0.8299958391204721, + "grad_norm": 0.4128401279449463, + "learning_rate": 1.3932059915471974e-05, + "loss": 0.0547, + "step": 35407 + }, + { + "epoch": 0.8300192806952771, + "grad_norm": 3.9702513217926025, + "learning_rate": 1.3928309689637508e-05, + "loss": 0.3893, + "step": 35408 + }, + { + "epoch": 0.8300427222700821, + "grad_norm": 1.5010918378829956, + "learning_rate": 1.392455993083045e-05, + "loss": 0.0576, + "step": 35409 + }, + { + "epoch": 0.8300661638448871, + "grad_norm": 0.431298166513443, + "learning_rate": 1.3920810639071125e-05, + "loss": 0.0212, + "step": 35410 + }, + { + "epoch": 0.8300896054196921, + "grad_norm": 0.45876544713974, + "learning_rate": 1.3917061814379928e-05, + "loss": 0.0371, + "step": 35411 + }, + { + "epoch": 0.8301130469944971, + "grad_norm": 3.2251031398773193, + "learning_rate": 1.3913313456777177e-05, + "loss": 0.3098, + "step": 35412 + }, + { + "epoch": 0.8301364885693021, + "grad_norm": 2.257293224334717, + "learning_rate": 1.39095655662832e-05, + "loss": 0.1043, + "step": 35413 + }, + { + "epoch": 0.830159930144107, + "grad_norm": 2.9532456398010254, + "learning_rate": 1.3905818142918347e-05, + "loss": 0.184, + "step": 35414 + }, + { + "epoch": 0.8301833717189121, + "grad_norm": 3.5179665088653564, + "learning_rate": 1.3902071186702913e-05, + "loss": 0.2686, + "step": 35415 + }, + { + "epoch": 0.830206813293717, + "grad_norm": 0.9668119549751282, + "learning_rate": 1.3898324697657294e-05, + "loss": 0.0742, + "step": 35416 + }, + { + "epoch": 0.8302302548685221, + "grad_norm": 0.36043107509613037, + "learning_rate": 1.3894578675801783e-05, + "loss": 0.0183, + "step": 35417 + }, + { + "epoch": 0.830253696443327, + "grad_norm": 1.44444739818573, + "learning_rate": 1.389083312115672e-05, + "loss": 0.0509, + "step": 35418 + }, + { + "epoch": 0.8302771380181321, + "grad_norm": 0.6958783864974976, + "learning_rate": 1.3887088033742401e-05, + "loss": 0.0693, + "step": 35419 + }, + { + "epoch": 0.830300579592937, + "grad_norm": 0.7038781642913818, + "learning_rate": 1.3883343413579153e-05, + "loss": 0.0468, + "step": 35420 + }, + { + "epoch": 0.8303240211677421, + "grad_norm": 2.71169376373291, + "learning_rate": 1.3879599260687314e-05, + "loss": 1.8168, + "step": 35421 + }, + { + "epoch": 0.830347462742547, + "grad_norm": 2.782477617263794, + "learning_rate": 1.3875855575087182e-05, + "loss": 1.1613, + "step": 35422 + }, + { + "epoch": 0.830370904317352, + "grad_norm": 1.996529459953308, + "learning_rate": 1.3872112356799093e-05, + "loss": 0.1789, + "step": 35423 + }, + { + "epoch": 0.830394345892157, + "grad_norm": 3.257237672805786, + "learning_rate": 1.3868369605843356e-05, + "loss": 0.2403, + "step": 35424 + }, + { + "epoch": 0.830417787466962, + "grad_norm": 0.7686460614204407, + "learning_rate": 1.3864627322240254e-05, + "loss": 0.0716, + "step": 35425 + }, + { + "epoch": 0.830441229041767, + "grad_norm": 3.1808481216430664, + "learning_rate": 1.3860885506010112e-05, + "loss": 0.1329, + "step": 35426 + }, + { + "epoch": 0.830464670616572, + "grad_norm": 2.307326316833496, + "learning_rate": 1.385714415717323e-05, + "loss": 0.1645, + "step": 35427 + }, + { + "epoch": 0.830488112191377, + "grad_norm": 0.9698432683944702, + "learning_rate": 1.3853403275749877e-05, + "loss": 0.0814, + "step": 35428 + }, + { + "epoch": 0.830511553766182, + "grad_norm": 1.3422871828079224, + "learning_rate": 1.3849662861760404e-05, + "loss": 0.1561, + "step": 35429 + }, + { + "epoch": 0.830534995340987, + "grad_norm": 1.2776645421981812, + "learning_rate": 1.3845922915225084e-05, + "loss": 0.1108, + "step": 35430 + }, + { + "epoch": 0.830558436915792, + "grad_norm": 3.455451726913452, + "learning_rate": 1.3842183436164213e-05, + "loss": 0.2372, + "step": 35431 + }, + { + "epoch": 0.830581878490597, + "grad_norm": 0.6706648468971252, + "learning_rate": 1.3838444424598052e-05, + "loss": 0.0428, + "step": 35432 + }, + { + "epoch": 0.830605320065402, + "grad_norm": 2.7937755584716797, + "learning_rate": 1.3834705880546927e-05, + "loss": 0.1992, + "step": 35433 + }, + { + "epoch": 0.8306287616402069, + "grad_norm": 1.4226957559585571, + "learning_rate": 1.3830967804031126e-05, + "loss": 0.0651, + "step": 35434 + }, + { + "epoch": 0.830652203215012, + "grad_norm": 3.648250102996826, + "learning_rate": 1.3827230195070884e-05, + "loss": 0.2939, + "step": 35435 + }, + { + "epoch": 0.8306756447898169, + "grad_norm": 3.805068254470825, + "learning_rate": 1.3823493053686542e-05, + "loss": 0.2706, + "step": 35436 + }, + { + "epoch": 0.830699086364622, + "grad_norm": 2.021407127380371, + "learning_rate": 1.3819756379898342e-05, + "loss": 0.1294, + "step": 35437 + }, + { + "epoch": 0.830722527939427, + "grad_norm": 1.9267069101333618, + "learning_rate": 1.3816020173726574e-05, + "loss": 0.114, + "step": 35438 + }, + { + "epoch": 0.830745969514232, + "grad_norm": 2.129595994949341, + "learning_rate": 1.3812284435191503e-05, + "loss": 0.1951, + "step": 35439 + }, + { + "epoch": 0.830769411089037, + "grad_norm": 0.2532135248184204, + "learning_rate": 1.38085491643134e-05, + "loss": 0.0076, + "step": 35440 + }, + { + "epoch": 0.830792852663842, + "grad_norm": 3.496333599090576, + "learning_rate": 1.3804814361112506e-05, + "loss": 0.318, + "step": 35441 + }, + { + "epoch": 0.830816294238647, + "grad_norm": 1.5888302326202393, + "learning_rate": 1.3801080025609136e-05, + "loss": 0.1628, + "step": 35442 + }, + { + "epoch": 0.8308397358134519, + "grad_norm": 0.7233164310455322, + "learning_rate": 1.3797346157823533e-05, + "loss": 0.1061, + "step": 35443 + }, + { + "epoch": 0.830863177388257, + "grad_norm": 3.2694904804229736, + "learning_rate": 1.3793612757775931e-05, + "loss": 0.3928, + "step": 35444 + }, + { + "epoch": 0.8308866189630619, + "grad_norm": 0.978293240070343, + "learning_rate": 1.378987982548664e-05, + "loss": 0.0513, + "step": 35445 + }, + { + "epoch": 0.830910060537867, + "grad_norm": 2.5758063793182373, + "learning_rate": 1.3786147360975876e-05, + "loss": 0.1902, + "step": 35446 + }, + { + "epoch": 0.8309335021126719, + "grad_norm": 2.0670299530029297, + "learning_rate": 1.3782415364263901e-05, + "loss": 0.1422, + "step": 35447 + }, + { + "epoch": 0.830956943687477, + "grad_norm": 2.2531356811523438, + "learning_rate": 1.3778683835370943e-05, + "loss": 1.5472, + "step": 35448 + }, + { + "epoch": 0.8309803852622819, + "grad_norm": 3.0850236415863037, + "learning_rate": 1.3774952774317284e-05, + "loss": 0.1885, + "step": 35449 + }, + { + "epoch": 0.831003826837087, + "grad_norm": 2.2564144134521484, + "learning_rate": 1.3771222181123167e-05, + "loss": 0.1238, + "step": 35450 + }, + { + "epoch": 0.8310272684118919, + "grad_norm": 3.444884777069092, + "learning_rate": 1.3767492055808817e-05, + "loss": 0.1974, + "step": 35451 + }, + { + "epoch": 0.831050709986697, + "grad_norm": 1.0288605690002441, + "learning_rate": 1.3763762398394486e-05, + "loss": 0.1141, + "step": 35452 + }, + { + "epoch": 0.8310741515615019, + "grad_norm": 2.7102530002593994, + "learning_rate": 1.376003320890038e-05, + "loss": 0.2737, + "step": 35453 + }, + { + "epoch": 0.8310975931363069, + "grad_norm": 2.552183151245117, + "learning_rate": 1.3756304487346783e-05, + "loss": 0.2341, + "step": 35454 + }, + { + "epoch": 0.8311210347111119, + "grad_norm": 1.9799039363861084, + "learning_rate": 1.37525762337539e-05, + "loss": 0.1093, + "step": 35455 + }, + { + "epoch": 0.8311444762859169, + "grad_norm": 2.7418665885925293, + "learning_rate": 1.3748848448141938e-05, + "loss": 0.1141, + "step": 35456 + }, + { + "epoch": 0.8311679178607219, + "grad_norm": 1.3195744752883911, + "learning_rate": 1.374512113053118e-05, + "loss": 0.1204, + "step": 35457 + }, + { + "epoch": 0.8311913594355269, + "grad_norm": 2.3807661533355713, + "learning_rate": 1.374139428094181e-05, + "loss": 0.2169, + "step": 35458 + }, + { + "epoch": 0.8312148010103318, + "grad_norm": 3.8018991947174072, + "learning_rate": 1.3737667899394068e-05, + "loss": 0.3181, + "step": 35459 + }, + { + "epoch": 0.8312382425851369, + "grad_norm": 2.861539363861084, + "learning_rate": 1.3733941985908139e-05, + "loss": 1.6807, + "step": 35460 + }, + { + "epoch": 0.8312616841599418, + "grad_norm": 0.31848788261413574, + "learning_rate": 1.3730216540504292e-05, + "loss": 0.0203, + "step": 35461 + }, + { + "epoch": 0.8312851257347469, + "grad_norm": 0.38313695788383484, + "learning_rate": 1.3726491563202715e-05, + "loss": 0.0307, + "step": 35462 + }, + { + "epoch": 0.8313085673095518, + "grad_norm": 3.6047539710998535, + "learning_rate": 1.3722767054023622e-05, + "loss": 1.4383, + "step": 35463 + }, + { + "epoch": 0.8313320088843569, + "grad_norm": 3.4547505378723145, + "learning_rate": 1.371904301298722e-05, + "loss": 0.2177, + "step": 35464 + }, + { + "epoch": 0.8313554504591618, + "grad_norm": 1.847681999206543, + "learning_rate": 1.3715319440113706e-05, + "loss": 0.0634, + "step": 35465 + }, + { + "epoch": 0.8313788920339669, + "grad_norm": 1.7292232513427734, + "learning_rate": 1.3711596335423282e-05, + "loss": 0.1974, + "step": 35466 + }, + { + "epoch": 0.8314023336087718, + "grad_norm": 2.487293004989624, + "learning_rate": 1.3707873698936191e-05, + "loss": 1.7476, + "step": 35467 + }, + { + "epoch": 0.8314257751835769, + "grad_norm": 0.7285254597663879, + "learning_rate": 1.3704151530672571e-05, + "loss": 0.0256, + "step": 35468 + }, + { + "epoch": 0.8314492167583818, + "grad_norm": 3.3158321380615234, + "learning_rate": 1.3700429830652683e-05, + "loss": 0.1822, + "step": 35469 + }, + { + "epoch": 0.8314726583331868, + "grad_norm": 2.7983577251434326, + "learning_rate": 1.3696708598896679e-05, + "loss": 0.2159, + "step": 35470 + }, + { + "epoch": 0.8314960999079918, + "grad_norm": 3.729576587677002, + "learning_rate": 1.3692987835424764e-05, + "loss": 0.2021, + "step": 35471 + }, + { + "epoch": 0.8315195414827968, + "grad_norm": 3.353409767150879, + "learning_rate": 1.3689267540257122e-05, + "loss": 0.2277, + "step": 35472 + }, + { + "epoch": 0.8315429830576018, + "grad_norm": 0.8845450282096863, + "learning_rate": 1.3685547713413927e-05, + "loss": 0.0755, + "step": 35473 + }, + { + "epoch": 0.8315664246324068, + "grad_norm": 1.1568970680236816, + "learning_rate": 1.3681828354915394e-05, + "loss": 0.0884, + "step": 35474 + }, + { + "epoch": 0.8315898662072118, + "grad_norm": 1.8111437559127808, + "learning_rate": 1.3678109464781686e-05, + "loss": 0.0668, + "step": 35475 + }, + { + "epoch": 0.8316133077820168, + "grad_norm": 1.3187127113342285, + "learning_rate": 1.3674391043032987e-05, + "loss": 0.1359, + "step": 35476 + }, + { + "epoch": 0.8316367493568217, + "grad_norm": 1.3817464113235474, + "learning_rate": 1.3670673089689468e-05, + "loss": 0.1124, + "step": 35477 + }, + { + "epoch": 0.8316601909316268, + "grad_norm": 0.7746293544769287, + "learning_rate": 1.3666955604771304e-05, + "loss": 0.0654, + "step": 35478 + }, + { + "epoch": 0.8316836325064317, + "grad_norm": 2.403472423553467, + "learning_rate": 1.3663238588298643e-05, + "loss": 0.0845, + "step": 35479 + }, + { + "epoch": 0.8317070740812368, + "grad_norm": 0.19447152316570282, + "learning_rate": 1.3659522040291672e-05, + "loss": 0.008, + "step": 35480 + }, + { + "epoch": 0.8317305156560417, + "grad_norm": 2.2665603160858154, + "learning_rate": 1.3655805960770596e-05, + "loss": 0.2042, + "step": 35481 + }, + { + "epoch": 0.8317539572308468, + "grad_norm": 0.7619906067848206, + "learning_rate": 1.365209034975553e-05, + "loss": 0.0394, + "step": 35482 + }, + { + "epoch": 0.8317773988056517, + "grad_norm": 1.1359593868255615, + "learning_rate": 1.3648375207266661e-05, + "loss": 0.0897, + "step": 35483 + }, + { + "epoch": 0.8318008403804568, + "grad_norm": 3.368978261947632, + "learning_rate": 1.3644660533324138e-05, + "loss": 0.2932, + "step": 35484 + }, + { + "epoch": 0.8318242819552617, + "grad_norm": 4.198289394378662, + "learning_rate": 1.3640946327948112e-05, + "loss": 0.438, + "step": 35485 + }, + { + "epoch": 0.8318477235300668, + "grad_norm": 1.833067774772644, + "learning_rate": 1.363723259115871e-05, + "loss": 0.0858, + "step": 35486 + }, + { + "epoch": 0.8318711651048717, + "grad_norm": 2.3964695930480957, + "learning_rate": 1.363351932297614e-05, + "loss": 1.4652, + "step": 35487 + }, + { + "epoch": 0.8318946066796767, + "grad_norm": 0.2956489622592926, + "learning_rate": 1.362980652342053e-05, + "loss": 0.0285, + "step": 35488 + }, + { + "epoch": 0.8319180482544818, + "grad_norm": 1.9597349166870117, + "learning_rate": 1.3626094192512007e-05, + "loss": 1.1269, + "step": 35489 + }, + { + "epoch": 0.8319414898292867, + "grad_norm": 1.7714673280715942, + "learning_rate": 1.3622382330270732e-05, + "loss": 0.8433, + "step": 35490 + }, + { + "epoch": 0.8319649314040918, + "grad_norm": 1.37998628616333, + "learning_rate": 1.3618670936716815e-05, + "loss": 0.1033, + "step": 35491 + }, + { + "epoch": 0.8319883729788967, + "grad_norm": 2.33005690574646, + "learning_rate": 1.3614960011870436e-05, + "loss": 0.126, + "step": 35492 + }, + { + "epoch": 0.8320118145537018, + "grad_norm": 1.2115898132324219, + "learning_rate": 1.3611249555751693e-05, + "loss": 0.1179, + "step": 35493 + }, + { + "epoch": 0.8320352561285067, + "grad_norm": 2.7306580543518066, + "learning_rate": 1.3607539568380768e-05, + "loss": 0.1681, + "step": 35494 + }, + { + "epoch": 0.8320586977033118, + "grad_norm": 1.5849300622940063, + "learning_rate": 1.3603830049777755e-05, + "loss": 0.1375, + "step": 35495 + }, + { + "epoch": 0.8320821392781167, + "grad_norm": 2.707937479019165, + "learning_rate": 1.3600120999962796e-05, + "loss": 0.2608, + "step": 35496 + }, + { + "epoch": 0.8321055808529217, + "grad_norm": 0.9797545671463013, + "learning_rate": 1.3596412418955994e-05, + "loss": 0.1063, + "step": 35497 + }, + { + "epoch": 0.8321290224277267, + "grad_norm": 4.238323211669922, + "learning_rate": 1.3592704306777482e-05, + "loss": 0.4555, + "step": 35498 + }, + { + "epoch": 0.8321524640025317, + "grad_norm": 3.849287271499634, + "learning_rate": 1.3588996663447406e-05, + "loss": 0.1921, + "step": 35499 + }, + { + "epoch": 0.8321759055773367, + "grad_norm": 3.9316580295562744, + "learning_rate": 1.3585289488985864e-05, + "loss": 0.5618, + "step": 35500 + }, + { + "epoch": 0.8321993471521417, + "grad_norm": 4.461272716522217, + "learning_rate": 1.3581582783412961e-05, + "loss": 0.2056, + "step": 35501 + }, + { + "epoch": 0.8322227887269467, + "grad_norm": 2.8941032886505127, + "learning_rate": 1.3577876546748824e-05, + "loss": 0.1562, + "step": 35502 + }, + { + "epoch": 0.8322462303017517, + "grad_norm": 1.2733914852142334, + "learning_rate": 1.3574170779013561e-05, + "loss": 0.1728, + "step": 35503 + }, + { + "epoch": 0.8322696718765566, + "grad_norm": 1.1445894241333008, + "learning_rate": 1.3570465480227256e-05, + "loss": 0.0169, + "step": 35504 + }, + { + "epoch": 0.8322931134513617, + "grad_norm": 0.5447431802749634, + "learning_rate": 1.3566760650410038e-05, + "loss": 0.019, + "step": 35505 + }, + { + "epoch": 0.8323165550261666, + "grad_norm": 4.210526466369629, + "learning_rate": 1.3563056289582032e-05, + "loss": 0.4877, + "step": 35506 + }, + { + "epoch": 0.8323399966009717, + "grad_norm": 1.1297844648361206, + "learning_rate": 1.3559352397763302e-05, + "loss": 0.1206, + "step": 35507 + }, + { + "epoch": 0.8323634381757766, + "grad_norm": 2.2494242191314697, + "learning_rate": 1.3555648974973966e-05, + "loss": 0.074, + "step": 35508 + }, + { + "epoch": 0.8323868797505817, + "grad_norm": 0.6841112375259399, + "learning_rate": 1.3551946021234097e-05, + "loss": 0.0737, + "step": 35509 + }, + { + "epoch": 0.8324103213253866, + "grad_norm": 2.6877963542938232, + "learning_rate": 1.354824353656381e-05, + "loss": 0.1862, + "step": 35510 + }, + { + "epoch": 0.8324337629001917, + "grad_norm": 0.8467460870742798, + "learning_rate": 1.3544541520983167e-05, + "loss": 0.0421, + "step": 35511 + }, + { + "epoch": 0.8324572044749966, + "grad_norm": 3.1913375854492188, + "learning_rate": 1.3540839974512288e-05, + "loss": 0.2701, + "step": 35512 + }, + { + "epoch": 0.8324806460498017, + "grad_norm": 2.0467331409454346, + "learning_rate": 1.3537138897171231e-05, + "loss": 0.2391, + "step": 35513 + }, + { + "epoch": 0.8325040876246066, + "grad_norm": 2.4978983402252197, + "learning_rate": 1.3533438288980105e-05, + "loss": 0.2047, + "step": 35514 + }, + { + "epoch": 0.8325275291994116, + "grad_norm": 0.43560102581977844, + "learning_rate": 1.352973814995897e-05, + "loss": 0.0215, + "step": 35515 + }, + { + "epoch": 0.8325509707742166, + "grad_norm": 4.993399620056152, + "learning_rate": 1.3526038480127912e-05, + "loss": 0.227, + "step": 35516 + }, + { + "epoch": 0.8325744123490216, + "grad_norm": 1.0303155183792114, + "learning_rate": 1.3522339279506969e-05, + "loss": 0.0827, + "step": 35517 + }, + { + "epoch": 0.8325978539238266, + "grad_norm": 2.099595785140991, + "learning_rate": 1.3518640548116246e-05, + "loss": 0.27, + "step": 35518 + }, + { + "epoch": 0.8326212954986316, + "grad_norm": 1.8409888744354248, + "learning_rate": 1.351494228597584e-05, + "loss": 1.276, + "step": 35519 + }, + { + "epoch": 0.8326447370734366, + "grad_norm": 1.3824269771575928, + "learning_rate": 1.3511244493105779e-05, + "loss": 0.1185, + "step": 35520 + }, + { + "epoch": 0.8326681786482416, + "grad_norm": 2.5179967880249023, + "learning_rate": 1.3507547169526147e-05, + "loss": 0.3671, + "step": 35521 + }, + { + "epoch": 0.8326916202230465, + "grad_norm": 1.4428157806396484, + "learning_rate": 1.3503850315256994e-05, + "loss": 0.0797, + "step": 35522 + }, + { + "epoch": 0.8327150617978516, + "grad_norm": 0.8671565055847168, + "learning_rate": 1.3500153930318371e-05, + "loss": 0.0816, + "step": 35523 + }, + { + "epoch": 0.8327385033726565, + "grad_norm": 2.5734052658081055, + "learning_rate": 1.349645801473034e-05, + "loss": 0.1576, + "step": 35524 + }, + { + "epoch": 0.8327619449474616, + "grad_norm": 0.9578935503959656, + "learning_rate": 1.3492762568512973e-05, + "loss": 0.0471, + "step": 35525 + }, + { + "epoch": 0.8327853865222665, + "grad_norm": 1.286781668663025, + "learning_rate": 1.3489067591686299e-05, + "loss": 0.1215, + "step": 35526 + }, + { + "epoch": 0.8328088280970716, + "grad_norm": 2.7383944988250732, + "learning_rate": 1.3485373084270392e-05, + "loss": 0.3761, + "step": 35527 + }, + { + "epoch": 0.8328322696718765, + "grad_norm": 2.1463711261749268, + "learning_rate": 1.3481679046285278e-05, + "loss": 0.1146, + "step": 35528 + }, + { + "epoch": 0.8328557112466816, + "grad_norm": 4.505195617675781, + "learning_rate": 1.3477985477750988e-05, + "loss": 0.4638, + "step": 35529 + }, + { + "epoch": 0.8328791528214865, + "grad_norm": 2.25824236869812, + "learning_rate": 1.3474292378687593e-05, + "loss": 0.1138, + "step": 35530 + }, + { + "epoch": 0.8329025943962916, + "grad_norm": 2.475457191467285, + "learning_rate": 1.3470599749115108e-05, + "loss": 0.0839, + "step": 35531 + }, + { + "epoch": 0.8329260359710965, + "grad_norm": 1.1577818393707275, + "learning_rate": 1.34669075890536e-05, + "loss": 0.6132, + "step": 35532 + }, + { + "epoch": 0.8329494775459015, + "grad_norm": 2.088585615158081, + "learning_rate": 1.3463215898523085e-05, + "loss": 0.1458, + "step": 35533 + }, + { + "epoch": 0.8329729191207065, + "grad_norm": 2.352416515350342, + "learning_rate": 1.34595246775436e-05, + "loss": 0.1446, + "step": 35534 + }, + { + "epoch": 0.8329963606955115, + "grad_norm": 2.7330546379089355, + "learning_rate": 1.3455833926135153e-05, + "loss": 0.173, + "step": 35535 + }, + { + "epoch": 0.8330198022703165, + "grad_norm": 2.710118293762207, + "learning_rate": 1.3452143644317772e-05, + "loss": 0.1825, + "step": 35536 + }, + { + "epoch": 0.8330432438451215, + "grad_norm": 2.0109846591949463, + "learning_rate": 1.3448453832111518e-05, + "loss": 0.1454, + "step": 35537 + }, + { + "epoch": 0.8330666854199265, + "grad_norm": 2.528256893157959, + "learning_rate": 1.3444764489536377e-05, + "loss": 0.1897, + "step": 35538 + }, + { + "epoch": 0.8330901269947315, + "grad_norm": 5.655437469482422, + "learning_rate": 1.3441075616612387e-05, + "loss": 0.1746, + "step": 35539 + }, + { + "epoch": 0.8331135685695366, + "grad_norm": 0.9905204772949219, + "learning_rate": 1.3437387213359553e-05, + "loss": 0.045, + "step": 35540 + }, + { + "epoch": 0.8331370101443415, + "grad_norm": 2.5377330780029297, + "learning_rate": 1.3433699279797863e-05, + "loss": 0.1836, + "step": 35541 + }, + { + "epoch": 0.8331604517191465, + "grad_norm": 0.448703795671463, + "learning_rate": 1.3430011815947386e-05, + "loss": 0.0193, + "step": 35542 + }, + { + "epoch": 0.8331838932939515, + "grad_norm": 0.625171422958374, + "learning_rate": 1.3426324821828062e-05, + "loss": 0.0604, + "step": 35543 + }, + { + "epoch": 0.8332073348687565, + "grad_norm": 1.367631196975708, + "learning_rate": 1.3422638297459966e-05, + "loss": 0.1119, + "step": 35544 + }, + { + "epoch": 0.8332307764435615, + "grad_norm": 3.8904383182525635, + "learning_rate": 1.341895224286307e-05, + "loss": 0.1469, + "step": 35545 + }, + { + "epoch": 0.8332542180183665, + "grad_norm": 2.2950847148895264, + "learning_rate": 1.3415266658057368e-05, + "loss": 0.2215, + "step": 35546 + }, + { + "epoch": 0.8332776595931715, + "grad_norm": 3.17478346824646, + "learning_rate": 1.3411581543062867e-05, + "loss": 0.2497, + "step": 35547 + }, + { + "epoch": 0.8333011011679765, + "grad_norm": 1.4358034133911133, + "learning_rate": 1.3407896897899563e-05, + "loss": 0.0759, + "step": 35548 + }, + { + "epoch": 0.8333245427427814, + "grad_norm": 1.0593252182006836, + "learning_rate": 1.3404212722587416e-05, + "loss": 0.075, + "step": 35549 + }, + { + "epoch": 0.8333479843175865, + "grad_norm": 2.61344838142395, + "learning_rate": 1.3400529017146479e-05, + "loss": 0.0509, + "step": 35550 + }, + { + "epoch": 0.8333714258923914, + "grad_norm": 2.6620073318481445, + "learning_rate": 1.3396845781596689e-05, + "loss": 0.3289, + "step": 35551 + }, + { + "epoch": 0.8333948674671965, + "grad_norm": 3.6477243900299072, + "learning_rate": 1.3393163015958066e-05, + "loss": 0.1441, + "step": 35552 + }, + { + "epoch": 0.8334183090420014, + "grad_norm": 0.6893517374992371, + "learning_rate": 1.3389480720250546e-05, + "loss": 0.0547, + "step": 35553 + }, + { + "epoch": 0.8334417506168065, + "grad_norm": 2.465343952178955, + "learning_rate": 1.3385798894494161e-05, + "loss": 1.5711, + "step": 35554 + }, + { + "epoch": 0.8334651921916114, + "grad_norm": 1.1946345567703247, + "learning_rate": 1.338211753870886e-05, + "loss": 0.1041, + "step": 35555 + }, + { + "epoch": 0.8334886337664165, + "grad_norm": 1.5016690492630005, + "learning_rate": 1.3378436652914616e-05, + "loss": 0.0698, + "step": 35556 + }, + { + "epoch": 0.8335120753412214, + "grad_norm": 1.9745572805404663, + "learning_rate": 1.3374756237131425e-05, + "loss": 0.1363, + "step": 35557 + }, + { + "epoch": 0.8335355169160265, + "grad_norm": 2.3127810955047607, + "learning_rate": 1.3371076291379236e-05, + "loss": 1.4118, + "step": 35558 + }, + { + "epoch": 0.8335589584908314, + "grad_norm": 3.493004322052002, + "learning_rate": 1.3367396815678035e-05, + "loss": 0.3221, + "step": 35559 + }, + { + "epoch": 0.8335824000656364, + "grad_norm": 1.835978627204895, + "learning_rate": 1.336371781004777e-05, + "loss": 0.2335, + "step": 35560 + }, + { + "epoch": 0.8336058416404414, + "grad_norm": 3.647951126098633, + "learning_rate": 1.3360039274508407e-05, + "loss": 0.105, + "step": 35561 + }, + { + "epoch": 0.8336292832152464, + "grad_norm": 2.7802951335906982, + "learning_rate": 1.3356361209079881e-05, + "loss": 0.1894, + "step": 35562 + }, + { + "epoch": 0.8336527247900514, + "grad_norm": 1.1752471923828125, + "learning_rate": 1.3352683613782202e-05, + "loss": 0.1156, + "step": 35563 + }, + { + "epoch": 0.8336761663648564, + "grad_norm": 2.925657033920288, + "learning_rate": 1.3349006488635296e-05, + "loss": 0.231, + "step": 35564 + }, + { + "epoch": 0.8336996079396614, + "grad_norm": 2.8177478313446045, + "learning_rate": 1.3345329833659126e-05, + "loss": 0.2746, + "step": 35565 + }, + { + "epoch": 0.8337230495144664, + "grad_norm": 2.558688163757324, + "learning_rate": 1.3341653648873599e-05, + "loss": 1.2247, + "step": 35566 + }, + { + "epoch": 0.8337464910892713, + "grad_norm": 1.750231146812439, + "learning_rate": 1.3337977934298729e-05, + "loss": 0.1223, + "step": 35567 + }, + { + "epoch": 0.8337699326640764, + "grad_norm": 1.4549286365509033, + "learning_rate": 1.3334302689954415e-05, + "loss": 0.0961, + "step": 35568 + }, + { + "epoch": 0.8337933742388813, + "grad_norm": 1.8472479581832886, + "learning_rate": 1.3330627915860594e-05, + "loss": 0.1622, + "step": 35569 + }, + { + "epoch": 0.8338168158136864, + "grad_norm": 2.5678818225860596, + "learning_rate": 1.3326953612037252e-05, + "loss": 0.0996, + "step": 35570 + }, + { + "epoch": 0.8338402573884913, + "grad_norm": 4.436872482299805, + "learning_rate": 1.3323279778504283e-05, + "loss": 0.3136, + "step": 35571 + }, + { + "epoch": 0.8338636989632964, + "grad_norm": 6.076173782348633, + "learning_rate": 1.331960641528165e-05, + "loss": 0.8921, + "step": 35572 + }, + { + "epoch": 0.8338871405381013, + "grad_norm": 0.6389304995536804, + "learning_rate": 1.3315933522389257e-05, + "loss": 0.054, + "step": 35573 + }, + { + "epoch": 0.8339105821129064, + "grad_norm": 4.911074638366699, + "learning_rate": 1.3312261099847034e-05, + "loss": 0.4092, + "step": 35574 + }, + { + "epoch": 0.8339340236877113, + "grad_norm": 2.1758620738983154, + "learning_rate": 1.3308589147674944e-05, + "loss": 0.2862, + "step": 35575 + }, + { + "epoch": 0.8339574652625164, + "grad_norm": 3.2488598823547363, + "learning_rate": 1.330491766589288e-05, + "loss": 0.2093, + "step": 35576 + }, + { + "epoch": 0.8339809068373213, + "grad_norm": 3.1652843952178955, + "learning_rate": 1.3301246654520771e-05, + "loss": 0.2111, + "step": 35577 + }, + { + "epoch": 0.8340043484121263, + "grad_norm": 2.911787509918213, + "learning_rate": 1.3297576113578513e-05, + "loss": 0.2933, + "step": 35578 + }, + { + "epoch": 0.8340277899869313, + "grad_norm": 1.8085706233978271, + "learning_rate": 1.3293906043086057e-05, + "loss": 0.144, + "step": 35579 + }, + { + "epoch": 0.8340512315617363, + "grad_norm": 0.04913252219557762, + "learning_rate": 1.329023644306332e-05, + "loss": 0.0013, + "step": 35580 + }, + { + "epoch": 0.8340746731365413, + "grad_norm": 2.0299251079559326, + "learning_rate": 1.3286567313530174e-05, + "loss": 0.082, + "step": 35581 + }, + { + "epoch": 0.8340981147113463, + "grad_norm": 5.1518659591674805, + "learning_rate": 1.328289865450657e-05, + "loss": 0.4781, + "step": 35582 + }, + { + "epoch": 0.8341215562861513, + "grad_norm": 2.8192901611328125, + "learning_rate": 1.3279230466012393e-05, + "loss": 0.2012, + "step": 35583 + }, + { + "epoch": 0.8341449978609563, + "grad_norm": 0.2665548324584961, + "learning_rate": 1.327556274806756e-05, + "loss": 0.0087, + "step": 35584 + }, + { + "epoch": 0.8341684394357612, + "grad_norm": 2.270535945892334, + "learning_rate": 1.3271895500691955e-05, + "loss": 0.1142, + "step": 35585 + }, + { + "epoch": 0.8341918810105663, + "grad_norm": 0.6314849257469177, + "learning_rate": 1.3268228723905485e-05, + "loss": 0.0433, + "step": 35586 + }, + { + "epoch": 0.8342153225853712, + "grad_norm": 0.41691556572914124, + "learning_rate": 1.3264562417728021e-05, + "loss": 0.0435, + "step": 35587 + }, + { + "epoch": 0.8342387641601763, + "grad_norm": 1.6334333419799805, + "learning_rate": 1.3260896582179505e-05, + "loss": 0.0667, + "step": 35588 + }, + { + "epoch": 0.8342622057349812, + "grad_norm": 2.328418254852295, + "learning_rate": 1.3257231217279798e-05, + "loss": 1.5897, + "step": 35589 + }, + { + "epoch": 0.8342856473097863, + "grad_norm": 2.920438766479492, + "learning_rate": 1.325356632304877e-05, + "loss": 0.1357, + "step": 35590 + }, + { + "epoch": 0.8343090888845913, + "grad_norm": 2.912729263305664, + "learning_rate": 1.3249901899506367e-05, + "loss": 0.318, + "step": 35591 + }, + { + "epoch": 0.8343325304593963, + "grad_norm": 0.6110005974769592, + "learning_rate": 1.3246237946672424e-05, + "loss": 0.0683, + "step": 35592 + }, + { + "epoch": 0.8343559720342013, + "grad_norm": 2.132582426071167, + "learning_rate": 1.3242574464566838e-05, + "loss": 0.1413, + "step": 35593 + }, + { + "epoch": 0.8343794136090062, + "grad_norm": 2.214231491088867, + "learning_rate": 1.3238911453209457e-05, + "loss": 0.1346, + "step": 35594 + }, + { + "epoch": 0.8344028551838113, + "grad_norm": 3.6010329723358154, + "learning_rate": 1.3235248912620202e-05, + "loss": 0.2748, + "step": 35595 + }, + { + "epoch": 0.8344262967586162, + "grad_norm": 1.423984169960022, + "learning_rate": 1.3231586842818922e-05, + "loss": 0.1016, + "step": 35596 + }, + { + "epoch": 0.8344497383334213, + "grad_norm": 1.2750614881515503, + "learning_rate": 1.32279252438255e-05, + "loss": 0.3134, + "step": 35597 + }, + { + "epoch": 0.8344731799082262, + "grad_norm": 2.1850993633270264, + "learning_rate": 1.322426411565979e-05, + "loss": 1.1747, + "step": 35598 + }, + { + "epoch": 0.8344966214830313, + "grad_norm": 1.540189266204834, + "learning_rate": 1.3220603458341674e-05, + "loss": 0.1497, + "step": 35599 + }, + { + "epoch": 0.8345200630578362, + "grad_norm": 1.6023203134536743, + "learning_rate": 1.3216943271890969e-05, + "loss": 0.1302, + "step": 35600 + }, + { + "epoch": 0.8345435046326413, + "grad_norm": 2.0287539958953857, + "learning_rate": 1.3213283556327594e-05, + "loss": 1.5217, + "step": 35601 + }, + { + "epoch": 0.8345669462074462, + "grad_norm": 1.7550547122955322, + "learning_rate": 1.3209624311671365e-05, + "loss": 0.0602, + "step": 35602 + }, + { + "epoch": 0.8345903877822513, + "grad_norm": 2.170621633529663, + "learning_rate": 1.3205965537942178e-05, + "loss": 0.1834, + "step": 35603 + }, + { + "epoch": 0.8346138293570562, + "grad_norm": 2.5982933044433594, + "learning_rate": 1.3202307235159861e-05, + "loss": 0.1428, + "step": 35604 + }, + { + "epoch": 0.8346372709318612, + "grad_norm": 1.2977901697158813, + "learning_rate": 1.3198649403344265e-05, + "loss": 0.1108, + "step": 35605 + }, + { + "epoch": 0.8346607125066662, + "grad_norm": 3.8132128715515137, + "learning_rate": 1.3194992042515231e-05, + "loss": 0.206, + "step": 35606 + }, + { + "epoch": 0.8346841540814712, + "grad_norm": 1.8124908208847046, + "learning_rate": 1.3191335152692596e-05, + "loss": 0.1528, + "step": 35607 + }, + { + "epoch": 0.8347075956562762, + "grad_norm": 0.7607443928718567, + "learning_rate": 1.3187678733896236e-05, + "loss": 0.0548, + "step": 35608 + }, + { + "epoch": 0.8347310372310812, + "grad_norm": 3.5075812339782715, + "learning_rate": 1.3184022786145977e-05, + "loss": 0.4113, + "step": 35609 + }, + { + "epoch": 0.8347544788058862, + "grad_norm": 3.653395414352417, + "learning_rate": 1.3180367309461649e-05, + "loss": 0.2061, + "step": 35610 + }, + { + "epoch": 0.8347779203806912, + "grad_norm": 0.7443704605102539, + "learning_rate": 1.3176712303863092e-05, + "loss": 0.0395, + "step": 35611 + }, + { + "epoch": 0.8348013619554961, + "grad_norm": 2.2534091472625732, + "learning_rate": 1.3173057769370112e-05, + "loss": 0.1606, + "step": 35612 + }, + { + "epoch": 0.8348248035303012, + "grad_norm": 2.982401132583618, + "learning_rate": 1.3169403706002581e-05, + "loss": 0.3536, + "step": 35613 + }, + { + "epoch": 0.8348482451051061, + "grad_norm": 3.5368149280548096, + "learning_rate": 1.3165750113780284e-05, + "loss": 0.2878, + "step": 35614 + }, + { + "epoch": 0.8348716866799112, + "grad_norm": 3.394280195236206, + "learning_rate": 1.3162096992723095e-05, + "loss": 0.1416, + "step": 35615 + }, + { + "epoch": 0.8348951282547161, + "grad_norm": 0.6485012769699097, + "learning_rate": 1.3158444342850807e-05, + "loss": 0.0681, + "step": 35616 + }, + { + "epoch": 0.8349185698295212, + "grad_norm": 0.6323642730712891, + "learning_rate": 1.3154792164183238e-05, + "loss": 0.055, + "step": 35617 + }, + { + "epoch": 0.8349420114043261, + "grad_norm": 1.8637655973434448, + "learning_rate": 1.3151140456740207e-05, + "loss": 0.0858, + "step": 35618 + }, + { + "epoch": 0.8349654529791312, + "grad_norm": 1.2736799716949463, + "learning_rate": 1.314748922054152e-05, + "loss": 0.9587, + "step": 35619 + }, + { + "epoch": 0.8349888945539361, + "grad_norm": 1.4119791984558105, + "learning_rate": 1.3143838455607017e-05, + "loss": 0.1401, + "step": 35620 + }, + { + "epoch": 0.8350123361287412, + "grad_norm": 1.045811414718628, + "learning_rate": 1.314018816195648e-05, + "loss": 0.1231, + "step": 35621 + }, + { + "epoch": 0.8350357777035461, + "grad_norm": 2.263375997543335, + "learning_rate": 1.3136538339609728e-05, + "loss": 0.1395, + "step": 35622 + }, + { + "epoch": 0.8350592192783511, + "grad_norm": 1.5519845485687256, + "learning_rate": 1.3132888988586567e-05, + "loss": 0.1038, + "step": 35623 + }, + { + "epoch": 0.8350826608531561, + "grad_norm": 2.0175044536590576, + "learning_rate": 1.3129240108906782e-05, + "loss": 0.2429, + "step": 35624 + }, + { + "epoch": 0.8351061024279611, + "grad_norm": 5.133491039276123, + "learning_rate": 1.3125591700590167e-05, + "loss": 0.2621, + "step": 35625 + }, + { + "epoch": 0.8351295440027661, + "grad_norm": 1.6570515632629395, + "learning_rate": 1.3121943763656541e-05, + "loss": 0.1319, + "step": 35626 + }, + { + "epoch": 0.8351529855775711, + "grad_norm": 1.8107773065567017, + "learning_rate": 1.311829629812571e-05, + "loss": 0.0751, + "step": 35627 + }, + { + "epoch": 0.835176427152376, + "grad_norm": 2.168052911758423, + "learning_rate": 1.3114649304017435e-05, + "loss": 0.2123, + "step": 35628 + }, + { + "epoch": 0.8351998687271811, + "grad_norm": 1.8533717393875122, + "learning_rate": 1.3111002781351523e-05, + "loss": 0.1477, + "step": 35629 + }, + { + "epoch": 0.835223310301986, + "grad_norm": 2.7307517528533936, + "learning_rate": 1.3107356730147758e-05, + "loss": 0.2361, + "step": 35630 + }, + { + "epoch": 0.8352467518767911, + "grad_norm": 0.3295128345489502, + "learning_rate": 1.3103711150425912e-05, + "loss": 0.0106, + "step": 35631 + }, + { + "epoch": 0.835270193451596, + "grad_norm": 3.251739740371704, + "learning_rate": 1.310006604220575e-05, + "loss": 0.35, + "step": 35632 + }, + { + "epoch": 0.8352936350264011, + "grad_norm": 3.431173801422119, + "learning_rate": 1.3096421405507098e-05, + "loss": 0.2371, + "step": 35633 + }, + { + "epoch": 0.835317076601206, + "grad_norm": 3.1534900665283203, + "learning_rate": 1.3092777240349707e-05, + "loss": 0.2782, + "step": 35634 + }, + { + "epoch": 0.8353405181760111, + "grad_norm": 1.8968133926391602, + "learning_rate": 1.3089133546753352e-05, + "loss": 0.1795, + "step": 35635 + }, + { + "epoch": 0.835363959750816, + "grad_norm": 1.8419523239135742, + "learning_rate": 1.3085490324737803e-05, + "loss": 0.1067, + "step": 35636 + }, + { + "epoch": 0.8353874013256211, + "grad_norm": 0.5752428770065308, + "learning_rate": 1.3081847574322825e-05, + "loss": 0.0234, + "step": 35637 + }, + { + "epoch": 0.835410842900426, + "grad_norm": 1.2385549545288086, + "learning_rate": 1.3078205295528167e-05, + "loss": 0.0471, + "step": 35638 + }, + { + "epoch": 0.835434284475231, + "grad_norm": 3.580388069152832, + "learning_rate": 1.3074563488373604e-05, + "loss": 0.3574, + "step": 35639 + }, + { + "epoch": 0.835457726050036, + "grad_norm": 1.4317835569381714, + "learning_rate": 1.3070922152878928e-05, + "loss": 0.0639, + "step": 35640 + }, + { + "epoch": 0.835481167624841, + "grad_norm": 3.811631917953491, + "learning_rate": 1.306728128906387e-05, + "loss": 0.196, + "step": 35641 + }, + { + "epoch": 0.8355046091996461, + "grad_norm": 2.5989627838134766, + "learning_rate": 1.3063640896948193e-05, + "loss": 1.6171, + "step": 35642 + }, + { + "epoch": 0.835528050774451, + "grad_norm": 2.5974926948547363, + "learning_rate": 1.3060000976551646e-05, + "loss": 0.1822, + "step": 35643 + }, + { + "epoch": 0.8355514923492561, + "grad_norm": 1.986283540725708, + "learning_rate": 1.3056361527893978e-05, + "loss": 0.133, + "step": 35644 + }, + { + "epoch": 0.835574933924061, + "grad_norm": 3.826230525970459, + "learning_rate": 1.3052722550994912e-05, + "loss": 0.3992, + "step": 35645 + }, + { + "epoch": 0.8355983754988661, + "grad_norm": 3.051194429397583, + "learning_rate": 1.3049084045874239e-05, + "loss": 1.8534, + "step": 35646 + }, + { + "epoch": 0.835621817073671, + "grad_norm": 1.1384629011154175, + "learning_rate": 1.3045446012551688e-05, + "loss": 0.0536, + "step": 35647 + }, + { + "epoch": 0.835645258648476, + "grad_norm": 0.8001468777656555, + "learning_rate": 1.3041808451046989e-05, + "loss": 0.0613, + "step": 35648 + }, + { + "epoch": 0.835668700223281, + "grad_norm": 0.756709098815918, + "learning_rate": 1.303817136137988e-05, + "loss": 0.0668, + "step": 35649 + }, + { + "epoch": 0.835692141798086, + "grad_norm": 1.9067717790603638, + "learning_rate": 1.303453474357007e-05, + "loss": 0.2815, + "step": 35650 + }, + { + "epoch": 0.835715583372891, + "grad_norm": 3.3328094482421875, + "learning_rate": 1.3030898597637352e-05, + "loss": 0.4017, + "step": 35651 + }, + { + "epoch": 0.835739024947696, + "grad_norm": 1.8469960689544678, + "learning_rate": 1.3027262923601403e-05, + "loss": 0.0675, + "step": 35652 + }, + { + "epoch": 0.835762466522501, + "grad_norm": 1.694730281829834, + "learning_rate": 1.3023627721481979e-05, + "loss": 0.0504, + "step": 35653 + }, + { + "epoch": 0.835785908097306, + "grad_norm": 3.830458641052246, + "learning_rate": 1.3019992991298813e-05, + "loss": 0.3646, + "step": 35654 + }, + { + "epoch": 0.835809349672111, + "grad_norm": 0.9620602130889893, + "learning_rate": 1.30163587330716e-05, + "loss": 0.1096, + "step": 35655 + }, + { + "epoch": 0.835832791246916, + "grad_norm": 2.8840575218200684, + "learning_rate": 1.3012724946820077e-05, + "loss": 0.0814, + "step": 35656 + }, + { + "epoch": 0.835856232821721, + "grad_norm": 3.0724775791168213, + "learning_rate": 1.300909163256393e-05, + "loss": 0.2636, + "step": 35657 + }, + { + "epoch": 0.835879674396526, + "grad_norm": 1.8083913326263428, + "learning_rate": 1.300545879032291e-05, + "loss": 0.1573, + "step": 35658 + }, + { + "epoch": 0.8359031159713309, + "grad_norm": 1.0510566234588623, + "learning_rate": 1.3001826420116725e-05, + "loss": 0.0592, + "step": 35659 + }, + { + "epoch": 0.835926557546136, + "grad_norm": 3.511510133743286, + "learning_rate": 1.2998194521965079e-05, + "loss": 0.3925, + "step": 35660 + }, + { + "epoch": 0.8359499991209409, + "grad_norm": 0.23729386925697327, + "learning_rate": 1.2994563095887668e-05, + "loss": 0.0094, + "step": 35661 + }, + { + "epoch": 0.835973440695746, + "grad_norm": 3.361851930618286, + "learning_rate": 1.2990932141904211e-05, + "loss": 0.2752, + "step": 35662 + }, + { + "epoch": 0.8359968822705509, + "grad_norm": 3.0908615589141846, + "learning_rate": 1.298730166003438e-05, + "loss": 0.2494, + "step": 35663 + }, + { + "epoch": 0.836020323845356, + "grad_norm": 2.294645309448242, + "learning_rate": 1.2983671650297902e-05, + "loss": 0.2164, + "step": 35664 + }, + { + "epoch": 0.8360437654201609, + "grad_norm": 3.006917715072632, + "learning_rate": 1.2980042112714474e-05, + "loss": 0.1777, + "step": 35665 + }, + { + "epoch": 0.836067206994966, + "grad_norm": 1.7321689128875732, + "learning_rate": 1.2976413047303804e-05, + "loss": 0.1408, + "step": 35666 + }, + { + "epoch": 0.8360906485697709, + "grad_norm": 2.8789119720458984, + "learning_rate": 1.297278445408555e-05, + "loss": 0.1881, + "step": 35667 + }, + { + "epoch": 0.8361140901445759, + "grad_norm": 2.145293951034546, + "learning_rate": 1.2969156333079414e-05, + "loss": 0.194, + "step": 35668 + }, + { + "epoch": 0.8361375317193809, + "grad_norm": 0.33453893661499023, + "learning_rate": 1.2965528684305095e-05, + "loss": 0.0187, + "step": 35669 + }, + { + "epoch": 0.8361609732941859, + "grad_norm": 2.198667287826538, + "learning_rate": 1.2961901507782226e-05, + "loss": 0.9421, + "step": 35670 + }, + { + "epoch": 0.8361844148689909, + "grad_norm": 6.9462690353393555, + "learning_rate": 1.2958274803530556e-05, + "loss": 0.3526, + "step": 35671 + }, + { + "epoch": 0.8362078564437959, + "grad_norm": 2.688319683074951, + "learning_rate": 1.2954648571569738e-05, + "loss": 0.2637, + "step": 35672 + }, + { + "epoch": 0.8362312980186009, + "grad_norm": 2.4556827545166016, + "learning_rate": 1.2951022811919444e-05, + "loss": 0.3086, + "step": 35673 + }, + { + "epoch": 0.8362547395934059, + "grad_norm": 2.9141316413879395, + "learning_rate": 1.2947397524599347e-05, + "loss": 0.1262, + "step": 35674 + }, + { + "epoch": 0.8362781811682108, + "grad_norm": 1.7661582231521606, + "learning_rate": 1.2943772709629099e-05, + "loss": 0.0807, + "step": 35675 + }, + { + "epoch": 0.8363016227430159, + "grad_norm": 1.9388848543167114, + "learning_rate": 1.2940148367028403e-05, + "loss": 0.197, + "step": 35676 + }, + { + "epoch": 0.8363250643178208, + "grad_norm": 3.8489041328430176, + "learning_rate": 1.2936524496816893e-05, + "loss": 0.4119, + "step": 35677 + }, + { + "epoch": 0.8363485058926259, + "grad_norm": 4.120088577270508, + "learning_rate": 1.2932901099014273e-05, + "loss": 0.3386, + "step": 35678 + }, + { + "epoch": 0.8363719474674308, + "grad_norm": 1.9919110536575317, + "learning_rate": 1.2929278173640168e-05, + "loss": 0.252, + "step": 35679 + }, + { + "epoch": 0.8363953890422359, + "grad_norm": 1.0252376794815063, + "learning_rate": 1.2925655720714259e-05, + "loss": 0.0963, + "step": 35680 + }, + { + "epoch": 0.8364188306170408, + "grad_norm": 2.0161216259002686, + "learning_rate": 1.292203374025619e-05, + "loss": 0.1353, + "step": 35681 + }, + { + "epoch": 0.8364422721918459, + "grad_norm": 0.91402268409729, + "learning_rate": 1.2918412232285615e-05, + "loss": 0.084, + "step": 35682 + }, + { + "epoch": 0.8364657137666508, + "grad_norm": 3.187880516052246, + "learning_rate": 1.2914791196822152e-05, + "loss": 0.4699, + "step": 35683 + }, + { + "epoch": 0.8364891553414558, + "grad_norm": 0.6086177229881287, + "learning_rate": 1.2911170633885516e-05, + "loss": 0.0423, + "step": 35684 + }, + { + "epoch": 0.8365125969162608, + "grad_norm": 2.599138021469116, + "learning_rate": 1.2907550543495306e-05, + "loss": 1.0706, + "step": 35685 + }, + { + "epoch": 0.8365360384910658, + "grad_norm": 0.8531581163406372, + "learning_rate": 1.2903930925671182e-05, + "loss": 0.0265, + "step": 35686 + }, + { + "epoch": 0.8365594800658708, + "grad_norm": 1.2599563598632812, + "learning_rate": 1.290031178043275e-05, + "loss": 0.1084, + "step": 35687 + }, + { + "epoch": 0.8365829216406758, + "grad_norm": 3.69757342338562, + "learning_rate": 1.2896693107799707e-05, + "loss": 0.3066, + "step": 35688 + }, + { + "epoch": 0.8366063632154808, + "grad_norm": 1.6530017852783203, + "learning_rate": 1.2893074907791624e-05, + "loss": 0.1131, + "step": 35689 + }, + { + "epoch": 0.8366298047902858, + "grad_norm": 2.819110155105591, + "learning_rate": 1.2889457180428189e-05, + "loss": 0.3776, + "step": 35690 + }, + { + "epoch": 0.8366532463650908, + "grad_norm": 1.8325307369232178, + "learning_rate": 1.2885839925729014e-05, + "loss": 1.2862, + "step": 35691 + }, + { + "epoch": 0.8366766879398958, + "grad_norm": 3.6740849018096924, + "learning_rate": 1.2882223143713723e-05, + "loss": 0.1925, + "step": 35692 + }, + { + "epoch": 0.8367001295147009, + "grad_norm": 2.161914825439453, + "learning_rate": 1.287860683440193e-05, + "loss": 0.1621, + "step": 35693 + }, + { + "epoch": 0.8367235710895058, + "grad_norm": 1.8490186929702759, + "learning_rate": 1.2874990997813274e-05, + "loss": 0.0576, + "step": 35694 + }, + { + "epoch": 0.8367470126643108, + "grad_norm": 1.139219880104065, + "learning_rate": 1.2871375633967342e-05, + "loss": 0.0849, + "step": 35695 + }, + { + "epoch": 0.8367704542391158, + "grad_norm": 1.689177393913269, + "learning_rate": 1.2867760742883795e-05, + "loss": 0.1027, + "step": 35696 + }, + { + "epoch": 0.8367938958139208, + "grad_norm": 2.9369375705718994, + "learning_rate": 1.2864146324582228e-05, + "loss": 0.2071, + "step": 35697 + }, + { + "epoch": 0.8368173373887258, + "grad_norm": 0.6631166338920593, + "learning_rate": 1.2860532379082258e-05, + "loss": 0.0906, + "step": 35698 + }, + { + "epoch": 0.8368407789635308, + "grad_norm": 0.9578538537025452, + "learning_rate": 1.2856918906403458e-05, + "loss": 0.0732, + "step": 35699 + }, + { + "epoch": 0.8368642205383358, + "grad_norm": 0.4026896357536316, + "learning_rate": 1.2853305906565494e-05, + "loss": 0.0331, + "step": 35700 + }, + { + "epoch": 0.8368876621131408, + "grad_norm": 3.492306709289551, + "learning_rate": 1.2849693379587947e-05, + "loss": 0.2826, + "step": 35701 + }, + { + "epoch": 0.8369111036879457, + "grad_norm": 1.907521367073059, + "learning_rate": 1.284608132549039e-05, + "loss": 0.1431, + "step": 35702 + }, + { + "epoch": 0.8369345452627508, + "grad_norm": 0.6134662628173828, + "learning_rate": 1.2842469744292462e-05, + "loss": 0.0512, + "step": 35703 + }, + { + "epoch": 0.8369579868375557, + "grad_norm": 3.9222121238708496, + "learning_rate": 1.283885863601375e-05, + "loss": 0.2961, + "step": 35704 + }, + { + "epoch": 0.8369814284123608, + "grad_norm": 1.8304405212402344, + "learning_rate": 1.2835248000673838e-05, + "loss": 0.199, + "step": 35705 + }, + { + "epoch": 0.8370048699871657, + "grad_norm": 6.033811569213867, + "learning_rate": 1.2831637838292322e-05, + "loss": 0.2071, + "step": 35706 + }, + { + "epoch": 0.8370283115619708, + "grad_norm": 2.985217332839966, + "learning_rate": 1.2828028148888793e-05, + "loss": 0.2697, + "step": 35707 + }, + { + "epoch": 0.8370517531367757, + "grad_norm": 2.564054012298584, + "learning_rate": 1.2824418932482806e-05, + "loss": 1.0496, + "step": 35708 + }, + { + "epoch": 0.8370751947115808, + "grad_norm": 2.8138089179992676, + "learning_rate": 1.2820810189094002e-05, + "loss": 1.2317, + "step": 35709 + }, + { + "epoch": 0.8370986362863857, + "grad_norm": 4.039480686187744, + "learning_rate": 1.2817201918741927e-05, + "loss": 0.23, + "step": 35710 + }, + { + "epoch": 0.8371220778611907, + "grad_norm": 3.080620765686035, + "learning_rate": 1.2813594121446148e-05, + "loss": 0.178, + "step": 35711 + }, + { + "epoch": 0.8371455194359957, + "grad_norm": 2.3448262214660645, + "learning_rate": 1.2809986797226281e-05, + "loss": 0.1038, + "step": 35712 + }, + { + "epoch": 0.8371689610108007, + "grad_norm": 1.6507161855697632, + "learning_rate": 1.2806379946101876e-05, + "loss": 0.7214, + "step": 35713 + }, + { + "epoch": 0.8371924025856057, + "grad_norm": 3.049799919128418, + "learning_rate": 1.2802773568092507e-05, + "loss": 0.2253, + "step": 35714 + }, + { + "epoch": 0.8372158441604107, + "grad_norm": 2.392764091491699, + "learning_rate": 1.2799167663217715e-05, + "loss": 1.0936, + "step": 35715 + }, + { + "epoch": 0.8372392857352157, + "grad_norm": 2.3641817569732666, + "learning_rate": 1.2795562231497116e-05, + "loss": 1.2516, + "step": 35716 + }, + { + "epoch": 0.8372627273100207, + "grad_norm": 1.763633370399475, + "learning_rate": 1.279195727295025e-05, + "loss": 0.1187, + "step": 35717 + }, + { + "epoch": 0.8372861688848257, + "grad_norm": 2.9482414722442627, + "learning_rate": 1.278835278759668e-05, + "loss": 0.2593, + "step": 35718 + }, + { + "epoch": 0.8373096104596307, + "grad_norm": 0.6361598968505859, + "learning_rate": 1.2784748775455945e-05, + "loss": 0.0779, + "step": 35719 + }, + { + "epoch": 0.8373330520344356, + "grad_norm": 2.7068026065826416, + "learning_rate": 1.2781145236547632e-05, + "loss": 0.1819, + "step": 35720 + }, + { + "epoch": 0.8373564936092407, + "grad_norm": 6.159414768218994, + "learning_rate": 1.2777542170891244e-05, + "loss": 0.4466, + "step": 35721 + }, + { + "epoch": 0.8373799351840456, + "grad_norm": 2.8120405673980713, + "learning_rate": 1.277393957850639e-05, + "loss": 0.191, + "step": 35722 + }, + { + "epoch": 0.8374033767588507, + "grad_norm": 3.6316308975219727, + "learning_rate": 1.2770337459412585e-05, + "loss": 0.1462, + "step": 35723 + }, + { + "epoch": 0.8374268183336556, + "grad_norm": 2.5414557456970215, + "learning_rate": 1.2766735813629371e-05, + "loss": 0.1038, + "step": 35724 + }, + { + "epoch": 0.8374502599084607, + "grad_norm": 1.7339890003204346, + "learning_rate": 1.2763134641176322e-05, + "loss": 1.5696, + "step": 35725 + }, + { + "epoch": 0.8374737014832656, + "grad_norm": 0.6590974926948547, + "learning_rate": 1.2759533942072955e-05, + "loss": 0.0349, + "step": 35726 + }, + { + "epoch": 0.8374971430580707, + "grad_norm": 0.7485032081604004, + "learning_rate": 1.2755933716338785e-05, + "loss": 0.0719, + "step": 35727 + }, + { + "epoch": 0.8375205846328756, + "grad_norm": 4.137410640716553, + "learning_rate": 1.275233396399339e-05, + "loss": 0.3794, + "step": 35728 + }, + { + "epoch": 0.8375440262076806, + "grad_norm": 2.366049289703369, + "learning_rate": 1.2748734685056295e-05, + "loss": 0.2298, + "step": 35729 + }, + { + "epoch": 0.8375674677824856, + "grad_norm": 0.9788901209831238, + "learning_rate": 1.274513587954701e-05, + "loss": 0.0281, + "step": 35730 + }, + { + "epoch": 0.8375909093572906, + "grad_norm": 3.224719285964966, + "learning_rate": 1.2741537547485083e-05, + "loss": 0.332, + "step": 35731 + }, + { + "epoch": 0.8376143509320956, + "grad_norm": 2.4748857021331787, + "learning_rate": 1.2737939688890011e-05, + "loss": 0.2121, + "step": 35732 + }, + { + "epoch": 0.8376377925069006, + "grad_norm": 1.7562777996063232, + "learning_rate": 1.2734342303781322e-05, + "loss": 0.089, + "step": 35733 + }, + { + "epoch": 0.8376612340817056, + "grad_norm": 1.8578382730484009, + "learning_rate": 1.2730745392178567e-05, + "loss": 0.0931, + "step": 35734 + }, + { + "epoch": 0.8376846756565106, + "grad_norm": 7.783594608306885, + "learning_rate": 1.272714895410123e-05, + "loss": 0.3865, + "step": 35735 + }, + { + "epoch": 0.8377081172313156, + "grad_norm": 2.694356918334961, + "learning_rate": 1.272355298956882e-05, + "loss": 0.3439, + "step": 35736 + }, + { + "epoch": 0.8377315588061206, + "grad_norm": 1.668993353843689, + "learning_rate": 1.2719957498600899e-05, + "loss": 0.0413, + "step": 35737 + }, + { + "epoch": 0.8377550003809255, + "grad_norm": 3.3110604286193848, + "learning_rate": 1.2716362481216926e-05, + "loss": 0.3916, + "step": 35738 + }, + { + "epoch": 0.8377784419557306, + "grad_norm": 0.4219292998313904, + "learning_rate": 1.2712767937436431e-05, + "loss": 0.0197, + "step": 35739 + }, + { + "epoch": 0.8378018835305355, + "grad_norm": 3.2770416736602783, + "learning_rate": 1.270917386727889e-05, + "loss": 0.2159, + "step": 35740 + }, + { + "epoch": 0.8378253251053406, + "grad_norm": 2.4193615913391113, + "learning_rate": 1.270558027076385e-05, + "loss": 0.1592, + "step": 35741 + }, + { + "epoch": 0.8378487666801455, + "grad_norm": 4.99550199508667, + "learning_rate": 1.2701987147910788e-05, + "loss": 0.2896, + "step": 35742 + }, + { + "epoch": 0.8378722082549506, + "grad_norm": 2.644542694091797, + "learning_rate": 1.2698394498739185e-05, + "loss": 0.1941, + "step": 35743 + }, + { + "epoch": 0.8378956498297556, + "grad_norm": 1.9628134965896606, + "learning_rate": 1.2694802323268562e-05, + "loss": 0.1454, + "step": 35744 + }, + { + "epoch": 0.8379190914045606, + "grad_norm": 3.2405354976654053, + "learning_rate": 1.2691210621518391e-05, + "loss": 0.2555, + "step": 35745 + }, + { + "epoch": 0.8379425329793656, + "grad_norm": 2.9886574745178223, + "learning_rate": 1.2687619393508143e-05, + "loss": 0.1886, + "step": 35746 + }, + { + "epoch": 0.8379659745541705, + "grad_norm": 1.3948782682418823, + "learning_rate": 1.268402863925735e-05, + "loss": 0.0832, + "step": 35747 + }, + { + "epoch": 0.8379894161289756, + "grad_norm": 1.1741104125976562, + "learning_rate": 1.2680438358785452e-05, + "loss": 0.1393, + "step": 35748 + }, + { + "epoch": 0.8380128577037805, + "grad_norm": 1.3948503732681274, + "learning_rate": 1.2676848552111975e-05, + "loss": 0.16, + "step": 35749 + }, + { + "epoch": 0.8380362992785856, + "grad_norm": 1.9987555742263794, + "learning_rate": 1.2673259219256373e-05, + "loss": 0.1487, + "step": 35750 + }, + { + "epoch": 0.8380597408533905, + "grad_norm": 3.7000153064727783, + "learning_rate": 1.2669670360238118e-05, + "loss": 0.3762, + "step": 35751 + }, + { + "epoch": 0.8380831824281956, + "grad_norm": 5.062994480133057, + "learning_rate": 1.2666081975076694e-05, + "loss": 0.3414, + "step": 35752 + }, + { + "epoch": 0.8381066240030005, + "grad_norm": 6.06378173828125, + "learning_rate": 1.266249406379153e-05, + "loss": 0.5204, + "step": 35753 + }, + { + "epoch": 0.8381300655778056, + "grad_norm": 8.43982219696045, + "learning_rate": 1.2658906626402155e-05, + "loss": 0.996, + "step": 35754 + }, + { + "epoch": 0.8381535071526105, + "grad_norm": 3.162416458129883, + "learning_rate": 1.2655319662928011e-05, + "loss": 0.2088, + "step": 35755 + }, + { + "epoch": 0.8381769487274155, + "grad_norm": 2.933305501937866, + "learning_rate": 1.2651733173388569e-05, + "loss": 0.14, + "step": 35756 + }, + { + "epoch": 0.8382003903022205, + "grad_norm": 2.411836624145508, + "learning_rate": 1.264814715780327e-05, + "loss": 0.1676, + "step": 35757 + }, + { + "epoch": 0.8382238318770255, + "grad_norm": 2.660573959350586, + "learning_rate": 1.2644561616191552e-05, + "loss": 0.1799, + "step": 35758 + }, + { + "epoch": 0.8382472734518305, + "grad_norm": 3.1960999965667725, + "learning_rate": 1.2640976548572925e-05, + "loss": 0.2133, + "step": 35759 + }, + { + "epoch": 0.8382707150266355, + "grad_norm": 2.4151721000671387, + "learning_rate": 1.2637391954966792e-05, + "loss": 0.0783, + "step": 35760 + }, + { + "epoch": 0.8382941566014405, + "grad_norm": 3.3416800498962402, + "learning_rate": 1.2633807835392653e-05, + "loss": 0.274, + "step": 35761 + }, + { + "epoch": 0.8383175981762455, + "grad_norm": 3.9039664268493652, + "learning_rate": 1.2630224189869932e-05, + "loss": 0.2697, + "step": 35762 + }, + { + "epoch": 0.8383410397510505, + "grad_norm": 1.779964804649353, + "learning_rate": 1.2626641018418061e-05, + "loss": 0.923, + "step": 35763 + }, + { + "epoch": 0.8383644813258555, + "grad_norm": 2.365989923477173, + "learning_rate": 1.2623058321056502e-05, + "loss": 1.5852, + "step": 35764 + }, + { + "epoch": 0.8383879229006604, + "grad_norm": 0.8787763118743896, + "learning_rate": 1.2619476097804661e-05, + "loss": 0.0499, + "step": 35765 + }, + { + "epoch": 0.8384113644754655, + "grad_norm": 4.290493011474609, + "learning_rate": 1.2615894348682022e-05, + "loss": 0.4417, + "step": 35766 + }, + { + "epoch": 0.8384348060502704, + "grad_norm": 3.8663690090179443, + "learning_rate": 1.2612313073707992e-05, + "loss": 0.4024, + "step": 35767 + }, + { + "epoch": 0.8384582476250755, + "grad_norm": 1.625540852546692, + "learning_rate": 1.2608732272902012e-05, + "loss": 0.1223, + "step": 35768 + }, + { + "epoch": 0.8384816891998804, + "grad_norm": 3.2154369354248047, + "learning_rate": 1.2605151946283511e-05, + "loss": 0.3547, + "step": 35769 + }, + { + "epoch": 0.8385051307746855, + "grad_norm": 2.6023645401000977, + "learning_rate": 1.2601572093871904e-05, + "loss": 0.0998, + "step": 35770 + }, + { + "epoch": 0.8385285723494904, + "grad_norm": 1.0161150693893433, + "learning_rate": 1.2597992715686613e-05, + "loss": 0.0729, + "step": 35771 + }, + { + "epoch": 0.8385520139242955, + "grad_norm": 2.2328975200653076, + "learning_rate": 1.2594413811747074e-05, + "loss": 1.5424, + "step": 35772 + }, + { + "epoch": 0.8385754554991004, + "grad_norm": 2.556195020675659, + "learning_rate": 1.2590835382072707e-05, + "loss": 0.1283, + "step": 35773 + }, + { + "epoch": 0.8385988970739054, + "grad_norm": 1.3262732028961182, + "learning_rate": 1.2587257426682942e-05, + "loss": 0.1223, + "step": 35774 + }, + { + "epoch": 0.8386223386487104, + "grad_norm": 3.679426670074463, + "learning_rate": 1.258367994559717e-05, + "loss": 0.2212, + "step": 35775 + }, + { + "epoch": 0.8386457802235154, + "grad_norm": 4.134055137634277, + "learning_rate": 1.2580102938834804e-05, + "loss": 0.4216, + "step": 35776 + }, + { + "epoch": 0.8386692217983204, + "grad_norm": 2.6831820011138916, + "learning_rate": 1.2576526406415268e-05, + "loss": 0.1402, + "step": 35777 + }, + { + "epoch": 0.8386926633731254, + "grad_norm": 3.154825210571289, + "learning_rate": 1.2572950348357925e-05, + "loss": 0.3325, + "step": 35778 + }, + { + "epoch": 0.8387161049479304, + "grad_norm": 2.429753541946411, + "learning_rate": 1.256937476468224e-05, + "loss": 0.1809, + "step": 35779 + }, + { + "epoch": 0.8387395465227354, + "grad_norm": 2.8188321590423584, + "learning_rate": 1.2565799655407584e-05, + "loss": 0.2989, + "step": 35780 + }, + { + "epoch": 0.8387629880975404, + "grad_norm": 3.3233470916748047, + "learning_rate": 1.2562225020553353e-05, + "loss": 0.3282, + "step": 35781 + }, + { + "epoch": 0.8387864296723454, + "grad_norm": 0.7983399033546448, + "learning_rate": 1.2558650860138955e-05, + "loss": 0.0794, + "step": 35782 + }, + { + "epoch": 0.8388098712471503, + "grad_norm": 1.9824588298797607, + "learning_rate": 1.2555077174183772e-05, + "loss": 0.088, + "step": 35783 + }, + { + "epoch": 0.8388333128219554, + "grad_norm": 2.0546834468841553, + "learning_rate": 1.2551503962707178e-05, + "loss": 0.1283, + "step": 35784 + }, + { + "epoch": 0.8388567543967603, + "grad_norm": 3.964632749557495, + "learning_rate": 1.2547931225728582e-05, + "loss": 0.3226, + "step": 35785 + }, + { + "epoch": 0.8388801959715654, + "grad_norm": 7.926181793212891, + "learning_rate": 1.2544358963267399e-05, + "loss": 0.3085, + "step": 35786 + }, + { + "epoch": 0.8389036375463703, + "grad_norm": 2.50663685798645, + "learning_rate": 1.2540787175342972e-05, + "loss": 0.3795, + "step": 35787 + }, + { + "epoch": 0.8389270791211754, + "grad_norm": 1.8751928806304932, + "learning_rate": 1.2537215861974693e-05, + "loss": 0.1105, + "step": 35788 + }, + { + "epoch": 0.8389505206959803, + "grad_norm": 1.111946702003479, + "learning_rate": 1.253364502318195e-05, + "loss": 0.0737, + "step": 35789 + }, + { + "epoch": 0.8389739622707854, + "grad_norm": 0.8668358325958252, + "learning_rate": 1.2530074658984104e-05, + "loss": 0.0685, + "step": 35790 + }, + { + "epoch": 0.8389974038455903, + "grad_norm": 3.899001359939575, + "learning_rate": 1.2526504769400504e-05, + "loss": 0.3415, + "step": 35791 + }, + { + "epoch": 0.8390208454203953, + "grad_norm": 2.514498710632324, + "learning_rate": 1.2522935354450582e-05, + "loss": 0.1718, + "step": 35792 + }, + { + "epoch": 0.8390442869952003, + "grad_norm": 3.093801736831665, + "learning_rate": 1.2519366414153666e-05, + "loss": 1.7947, + "step": 35793 + }, + { + "epoch": 0.8390677285700053, + "grad_norm": 1.4226117134094238, + "learning_rate": 1.2515797948529129e-05, + "loss": 0.0358, + "step": 35794 + }, + { + "epoch": 0.8390911701448104, + "grad_norm": 0.7954151630401611, + "learning_rate": 1.2512229957596334e-05, + "loss": 0.0563, + "step": 35795 + }, + { + "epoch": 0.8391146117196153, + "grad_norm": 1.9474331140518188, + "learning_rate": 1.2508662441374609e-05, + "loss": 0.1921, + "step": 35796 + }, + { + "epoch": 0.8391380532944204, + "grad_norm": 3.0958163738250732, + "learning_rate": 1.2505095399883371e-05, + "loss": 1.7746, + "step": 35797 + }, + { + "epoch": 0.8391614948692253, + "grad_norm": 1.1668323278427124, + "learning_rate": 1.2501528833141917e-05, + "loss": 0.0399, + "step": 35798 + }, + { + "epoch": 0.8391849364440304, + "grad_norm": 3.832399606704712, + "learning_rate": 1.2497962741169655e-05, + "loss": 0.4625, + "step": 35799 + }, + { + "epoch": 0.8392083780188353, + "grad_norm": 0.4127638339996338, + "learning_rate": 1.2494397123985913e-05, + "loss": 0.0446, + "step": 35800 + }, + { + "epoch": 0.8392318195936403, + "grad_norm": 3.918113946914673, + "learning_rate": 1.2490831981610029e-05, + "loss": 0.1941, + "step": 35801 + }, + { + "epoch": 0.8392552611684453, + "grad_norm": 3.167175054550171, + "learning_rate": 1.2487267314061346e-05, + "loss": 0.2135, + "step": 35802 + }, + { + "epoch": 0.8392787027432503, + "grad_norm": 0.9117016196250916, + "learning_rate": 1.2483703121359191e-05, + "loss": 0.0681, + "step": 35803 + }, + { + "epoch": 0.8393021443180553, + "grad_norm": 2.755823850631714, + "learning_rate": 1.248013940352295e-05, + "loss": 0.215, + "step": 35804 + }, + { + "epoch": 0.8393255858928603, + "grad_norm": 0.47246962785720825, + "learning_rate": 1.2476576160571929e-05, + "loss": 0.0157, + "step": 35805 + }, + { + "epoch": 0.8393490274676653, + "grad_norm": 1.650619387626648, + "learning_rate": 1.2473013392525467e-05, + "loss": 0.1048, + "step": 35806 + }, + { + "epoch": 0.8393724690424703, + "grad_norm": 1.259211778640747, + "learning_rate": 1.2469451099402895e-05, + "loss": 0.0474, + "step": 35807 + }, + { + "epoch": 0.8393959106172753, + "grad_norm": 2.2824740409851074, + "learning_rate": 1.2465889281223541e-05, + "loss": 0.1345, + "step": 35808 + }, + { + "epoch": 0.8394193521920803, + "grad_norm": 2.9031317234039307, + "learning_rate": 1.2462327938006713e-05, + "loss": 0.3711, + "step": 35809 + }, + { + "epoch": 0.8394427937668852, + "grad_norm": 3.9834723472595215, + "learning_rate": 1.2458767069771759e-05, + "loss": 0.2512, + "step": 35810 + }, + { + "epoch": 0.8394662353416903, + "grad_norm": 4.359045028686523, + "learning_rate": 1.2455206676538012e-05, + "loss": 0.1611, + "step": 35811 + }, + { + "epoch": 0.8394896769164952, + "grad_norm": 1.872315764427185, + "learning_rate": 1.2451646758324776e-05, + "loss": 0.101, + "step": 35812 + }, + { + "epoch": 0.8395131184913003, + "grad_norm": 3.377413034439087, + "learning_rate": 1.244808731515137e-05, + "loss": 0.2065, + "step": 35813 + }, + { + "epoch": 0.8395365600661052, + "grad_norm": 1.9156270027160645, + "learning_rate": 1.244452834703711e-05, + "loss": 0.0907, + "step": 35814 + }, + { + "epoch": 0.8395600016409103, + "grad_norm": 2.5324349403381348, + "learning_rate": 1.2440969854001283e-05, + "loss": 0.1299, + "step": 35815 + }, + { + "epoch": 0.8395834432157152, + "grad_norm": 1.8419703245162964, + "learning_rate": 1.2437411836063207e-05, + "loss": 0.8793, + "step": 35816 + }, + { + "epoch": 0.8396068847905203, + "grad_norm": 1.9863123893737793, + "learning_rate": 1.2433854293242219e-05, + "loss": 0.137, + "step": 35817 + }, + { + "epoch": 0.8396303263653252, + "grad_norm": 1.7803659439086914, + "learning_rate": 1.2430297225557597e-05, + "loss": 0.141, + "step": 35818 + }, + { + "epoch": 0.8396537679401302, + "grad_norm": 0.9929376840591431, + "learning_rate": 1.2426740633028644e-05, + "loss": 0.0629, + "step": 35819 + }, + { + "epoch": 0.8396772095149352, + "grad_norm": 0.6784601211547852, + "learning_rate": 1.2423184515674657e-05, + "loss": 0.0492, + "step": 35820 + }, + { + "epoch": 0.8397006510897402, + "grad_norm": 1.446205496788025, + "learning_rate": 1.2419628873514911e-05, + "loss": 0.0837, + "step": 35821 + }, + { + "epoch": 0.8397240926645452, + "grad_norm": 1.8971056938171387, + "learning_rate": 1.2416073706568743e-05, + "loss": 0.0737, + "step": 35822 + }, + { + "epoch": 0.8397475342393502, + "grad_norm": 1.6756960153579712, + "learning_rate": 1.2412519014855407e-05, + "loss": 0.132, + "step": 35823 + }, + { + "epoch": 0.8397709758141552, + "grad_norm": 2.978405714035034, + "learning_rate": 1.2408964798394217e-05, + "loss": 0.191, + "step": 35824 + }, + { + "epoch": 0.8397944173889602, + "grad_norm": 3.1899874210357666, + "learning_rate": 1.2405411057204453e-05, + "loss": 0.4242, + "step": 35825 + }, + { + "epoch": 0.8398178589637652, + "grad_norm": 3.4496943950653076, + "learning_rate": 1.2401857791305394e-05, + "loss": 0.2614, + "step": 35826 + }, + { + "epoch": 0.8398413005385702, + "grad_norm": 3.388619899749756, + "learning_rate": 1.2398305000716315e-05, + "loss": 0.3144, + "step": 35827 + }, + { + "epoch": 0.8398647421133751, + "grad_norm": 4.81517219543457, + "learning_rate": 1.239475268545649e-05, + "loss": 0.4619, + "step": 35828 + }, + { + "epoch": 0.8398881836881802, + "grad_norm": 3.5892531871795654, + "learning_rate": 1.239120084554518e-05, + "loss": 0.2918, + "step": 35829 + }, + { + "epoch": 0.8399116252629851, + "grad_norm": 8.019682884216309, + "learning_rate": 1.2387649481001707e-05, + "loss": 0.3574, + "step": 35830 + }, + { + "epoch": 0.8399350668377902, + "grad_norm": 1.1128016710281372, + "learning_rate": 1.2384098591845305e-05, + "loss": 0.0655, + "step": 35831 + }, + { + "epoch": 0.8399585084125951, + "grad_norm": 2.0682802200317383, + "learning_rate": 1.2380548178095253e-05, + "loss": 1.1163, + "step": 35832 + }, + { + "epoch": 0.8399819499874002, + "grad_norm": 1.4843239784240723, + "learning_rate": 1.2376998239770787e-05, + "loss": 0.0631, + "step": 35833 + }, + { + "epoch": 0.8400053915622051, + "grad_norm": 8.56342887878418, + "learning_rate": 1.2373448776891205e-05, + "loss": 0.2555, + "step": 35834 + }, + { + "epoch": 0.8400288331370102, + "grad_norm": 0.4138874113559723, + "learning_rate": 1.236989978947577e-05, + "loss": 0.0321, + "step": 35835 + }, + { + "epoch": 0.8400522747118151, + "grad_norm": 0.35733410716056824, + "learning_rate": 1.2366351277543686e-05, + "loss": 0.0311, + "step": 35836 + }, + { + "epoch": 0.8400757162866201, + "grad_norm": 3.0771331787109375, + "learning_rate": 1.2362803241114274e-05, + "loss": 2.0077, + "step": 35837 + }, + { + "epoch": 0.8400991578614251, + "grad_norm": 1.9059444665908813, + "learning_rate": 1.235925568020675e-05, + "loss": 0.1966, + "step": 35838 + }, + { + "epoch": 0.8401225994362301, + "grad_norm": 1.6650081872940063, + "learning_rate": 1.2355708594840376e-05, + "loss": 0.2479, + "step": 35839 + }, + { + "epoch": 0.8401460410110351, + "grad_norm": 0.6982275247573853, + "learning_rate": 1.2352161985034382e-05, + "loss": 0.029, + "step": 35840 + }, + { + "epoch": 0.8401694825858401, + "grad_norm": 2.1500802040100098, + "learning_rate": 1.234861585080801e-05, + "loss": 0.0402, + "step": 35841 + }, + { + "epoch": 0.8401929241606451, + "grad_norm": 1.3963969945907593, + "learning_rate": 1.234507019218053e-05, + "loss": 0.0714, + "step": 35842 + }, + { + "epoch": 0.8402163657354501, + "grad_norm": 3.9706125259399414, + "learning_rate": 1.2341525009171162e-05, + "loss": 0.3694, + "step": 35843 + }, + { + "epoch": 0.840239807310255, + "grad_norm": 2.362229347229004, + "learning_rate": 1.2337980301799145e-05, + "loss": 0.1573, + "step": 35844 + }, + { + "epoch": 0.8402632488850601, + "grad_norm": 1.9873123168945312, + "learning_rate": 1.23344360700837e-05, + "loss": 0.0691, + "step": 35845 + }, + { + "epoch": 0.8402866904598651, + "grad_norm": 0.9507126808166504, + "learning_rate": 1.2330892314044085e-05, + "loss": 0.1263, + "step": 35846 + }, + { + "epoch": 0.8403101320346701, + "grad_norm": 2.8146743774414062, + "learning_rate": 1.232734903369951e-05, + "loss": 0.3368, + "step": 35847 + }, + { + "epoch": 0.8403335736094751, + "grad_norm": 0.48325690627098083, + "learning_rate": 1.232380622906918e-05, + "loss": 0.0276, + "step": 35848 + }, + { + "epoch": 0.8403570151842801, + "grad_norm": 2.596522569656372, + "learning_rate": 1.2320263900172369e-05, + "loss": 0.2006, + "step": 35849 + }, + { + "epoch": 0.8403804567590851, + "grad_norm": 1.5920263528823853, + "learning_rate": 1.2316722047028272e-05, + "loss": 0.129, + "step": 35850 + }, + { + "epoch": 0.8404038983338901, + "grad_norm": 2.3758349418640137, + "learning_rate": 1.23131806696561e-05, + "loss": 0.1884, + "step": 35851 + }, + { + "epoch": 0.8404273399086951, + "grad_norm": 2.27829647064209, + "learning_rate": 1.2309639768075087e-05, + "loss": 0.2455, + "step": 35852 + }, + { + "epoch": 0.8404507814835, + "grad_norm": 1.0190324783325195, + "learning_rate": 1.2306099342304423e-05, + "loss": 0.1258, + "step": 35853 + }, + { + "epoch": 0.8404742230583051, + "grad_norm": 6.128387928009033, + "learning_rate": 1.2302559392363311e-05, + "loss": 0.6235, + "step": 35854 + }, + { + "epoch": 0.84049766463311, + "grad_norm": 4.217854022979736, + "learning_rate": 1.2299019918271004e-05, + "loss": 0.1336, + "step": 35855 + }, + { + "epoch": 0.8405211062079151, + "grad_norm": 4.431812763214111, + "learning_rate": 1.2295480920046675e-05, + "loss": 0.2689, + "step": 35856 + }, + { + "epoch": 0.84054454778272, + "grad_norm": 1.2275547981262207, + "learning_rate": 1.229194239770951e-05, + "loss": 0.0959, + "step": 35857 + }, + { + "epoch": 0.8405679893575251, + "grad_norm": 4.118928909301758, + "learning_rate": 1.228840435127876e-05, + "loss": 0.1461, + "step": 35858 + }, + { + "epoch": 0.84059143093233, + "grad_norm": 5.123678207397461, + "learning_rate": 1.2284866780773597e-05, + "loss": 0.3378, + "step": 35859 + }, + { + "epoch": 0.8406148725071351, + "grad_norm": 0.36539968848228455, + "learning_rate": 1.2281329686213205e-05, + "loss": 0.0329, + "step": 35860 + }, + { + "epoch": 0.84063831408194, + "grad_norm": 3.551701307296753, + "learning_rate": 1.227779306761676e-05, + "loss": 0.4477, + "step": 35861 + }, + { + "epoch": 0.8406617556567451, + "grad_norm": 1.7241193056106567, + "learning_rate": 1.2274256925003502e-05, + "loss": 0.1278, + "step": 35862 + }, + { + "epoch": 0.84068519723155, + "grad_norm": 1.934635043144226, + "learning_rate": 1.2270721258392603e-05, + "loss": 0.2085, + "step": 35863 + }, + { + "epoch": 0.840708638806355, + "grad_norm": 0.18558669090270996, + "learning_rate": 1.2267186067803227e-05, + "loss": 0.0148, + "step": 35864 + }, + { + "epoch": 0.84073208038116, + "grad_norm": 2.9241976737976074, + "learning_rate": 1.2263651353254569e-05, + "loss": 0.2361, + "step": 35865 + }, + { + "epoch": 0.840755521955965, + "grad_norm": 0.4314035177230835, + "learning_rate": 1.2260117114765801e-05, + "loss": 0.0339, + "step": 35866 + }, + { + "epoch": 0.84077896353077, + "grad_norm": 2.1534385681152344, + "learning_rate": 1.2256583352356088e-05, + "loss": 0.0713, + "step": 35867 + }, + { + "epoch": 0.840802405105575, + "grad_norm": 1.8038395643234253, + "learning_rate": 1.2253050066044636e-05, + "loss": 0.0983, + "step": 35868 + }, + { + "epoch": 0.84082584668038, + "grad_norm": 2.611640453338623, + "learning_rate": 1.2249517255850584e-05, + "loss": 0.126, + "step": 35869 + }, + { + "epoch": 0.840849288255185, + "grad_norm": 3.5957822799682617, + "learning_rate": 1.2245984921793141e-05, + "loss": 0.5847, + "step": 35870 + }, + { + "epoch": 0.84087272982999, + "grad_norm": 4.039188861846924, + "learning_rate": 1.2242453063891446e-05, + "loss": 0.3441, + "step": 35871 + }, + { + "epoch": 0.840896171404795, + "grad_norm": 4.195590972900391, + "learning_rate": 1.2238921682164684e-05, + "loss": 0.3533, + "step": 35872 + }, + { + "epoch": 0.8409196129795999, + "grad_norm": 2.942101240158081, + "learning_rate": 1.2235390776631982e-05, + "loss": 0.1694, + "step": 35873 + }, + { + "epoch": 0.840943054554405, + "grad_norm": 2.787036895751953, + "learning_rate": 1.2231860347312507e-05, + "loss": 0.1824, + "step": 35874 + }, + { + "epoch": 0.8409664961292099, + "grad_norm": 0.493502140045166, + "learning_rate": 1.222833039422544e-05, + "loss": 0.0526, + "step": 35875 + }, + { + "epoch": 0.840989937704015, + "grad_norm": 0.6257320642471313, + "learning_rate": 1.2224800917389923e-05, + "loss": 0.0475, + "step": 35876 + }, + { + "epoch": 0.8410133792788199, + "grad_norm": 0.917156457901001, + "learning_rate": 1.2221271916825106e-05, + "loss": 0.0614, + "step": 35877 + }, + { + "epoch": 0.841036820853625, + "grad_norm": 2.9725677967071533, + "learning_rate": 1.2217743392550141e-05, + "loss": 1.2059, + "step": 35878 + }, + { + "epoch": 0.8410602624284299, + "grad_norm": 0.7993043661117554, + "learning_rate": 1.2214215344584135e-05, + "loss": 0.0584, + "step": 35879 + }, + { + "epoch": 0.841083704003235, + "grad_norm": 2.8400235176086426, + "learning_rate": 1.2210687772946295e-05, + "loss": 0.3513, + "step": 35880 + }, + { + "epoch": 0.8411071455780399, + "grad_norm": 0.5524574518203735, + "learning_rate": 1.2207160677655727e-05, + "loss": 0.0358, + "step": 35881 + }, + { + "epoch": 0.8411305871528449, + "grad_norm": 4.26461935043335, + "learning_rate": 1.220363405873155e-05, + "loss": 0.2708, + "step": 35882 + }, + { + "epoch": 0.8411540287276499, + "grad_norm": 1.6737494468688965, + "learning_rate": 1.2200107916192948e-05, + "loss": 1.0868, + "step": 35883 + }, + { + "epoch": 0.8411774703024549, + "grad_norm": 2.2992875576019287, + "learning_rate": 1.219658225005903e-05, + "loss": 0.2563, + "step": 35884 + }, + { + "epoch": 0.8412009118772599, + "grad_norm": 2.2117748260498047, + "learning_rate": 1.2193057060348933e-05, + "loss": 0.1356, + "step": 35885 + }, + { + "epoch": 0.8412243534520649, + "grad_norm": 1.8303014039993286, + "learning_rate": 1.2189532347081745e-05, + "loss": 0.9124, + "step": 35886 + }, + { + "epoch": 0.8412477950268699, + "grad_norm": 1.387186050415039, + "learning_rate": 1.2186008110276648e-05, + "loss": 0.0407, + "step": 35887 + }, + { + "epoch": 0.8412712366016749, + "grad_norm": 0.945052981376648, + "learning_rate": 1.2182484349952738e-05, + "loss": 0.0946, + "step": 35888 + }, + { + "epoch": 0.8412946781764798, + "grad_norm": 1.5244472026824951, + "learning_rate": 1.2178961066129136e-05, + "loss": 0.0489, + "step": 35889 + }, + { + "epoch": 0.8413181197512849, + "grad_norm": 1.2632238864898682, + "learning_rate": 1.2175438258824967e-05, + "loss": 0.0917, + "step": 35890 + }, + { + "epoch": 0.8413415613260898, + "grad_norm": 0.5134739279747009, + "learning_rate": 1.217191592805934e-05, + "loss": 0.0406, + "step": 35891 + }, + { + "epoch": 0.8413650029008949, + "grad_norm": 4.454177379608154, + "learning_rate": 1.2168394073851341e-05, + "loss": 0.5508, + "step": 35892 + }, + { + "epoch": 0.8413884444756998, + "grad_norm": 3.406845808029175, + "learning_rate": 1.216487269622012e-05, + "loss": 0.4387, + "step": 35893 + }, + { + "epoch": 0.8414118860505049, + "grad_norm": 0.7924333810806274, + "learning_rate": 1.216135179518476e-05, + "loss": 0.0491, + "step": 35894 + }, + { + "epoch": 0.8414353276253098, + "grad_norm": 1.7899880409240723, + "learning_rate": 1.2157831370764394e-05, + "loss": 1.5768, + "step": 35895 + }, + { + "epoch": 0.8414587692001149, + "grad_norm": 1.9516690969467163, + "learning_rate": 1.2154311422978104e-05, + "loss": 0.094, + "step": 35896 + }, + { + "epoch": 0.8414822107749199, + "grad_norm": 1.2161778211593628, + "learning_rate": 1.2150791951844986e-05, + "loss": 0.075, + "step": 35897 + }, + { + "epoch": 0.8415056523497249, + "grad_norm": 0.7680575847625732, + "learning_rate": 1.2147272957384149e-05, + "loss": 0.0605, + "step": 35898 + }, + { + "epoch": 0.8415290939245299, + "grad_norm": 0.7884153127670288, + "learning_rate": 1.2143754439614651e-05, + "loss": 0.0707, + "step": 35899 + }, + { + "epoch": 0.8415525354993348, + "grad_norm": 1.5982706546783447, + "learning_rate": 1.2140236398555627e-05, + "loss": 0.1047, + "step": 35900 + }, + { + "epoch": 0.8415759770741399, + "grad_norm": 2.345745801925659, + "learning_rate": 1.2136718834226157e-05, + "loss": 0.1506, + "step": 35901 + }, + { + "epoch": 0.8415994186489448, + "grad_norm": 2.5176033973693848, + "learning_rate": 1.2133201746645328e-05, + "loss": 0.2456, + "step": 35902 + }, + { + "epoch": 0.8416228602237499, + "grad_norm": 2.545501947402954, + "learning_rate": 1.2129685135832213e-05, + "loss": 0.2602, + "step": 35903 + }, + { + "epoch": 0.8416463017985548, + "grad_norm": 2.707932949066162, + "learning_rate": 1.2126169001805899e-05, + "loss": 0.2996, + "step": 35904 + }, + { + "epoch": 0.8416697433733599, + "grad_norm": 2.639166831970215, + "learning_rate": 1.2122653344585432e-05, + "loss": 1.6981, + "step": 35905 + }, + { + "epoch": 0.8416931849481648, + "grad_norm": 4.86952018737793, + "learning_rate": 1.2119138164189925e-05, + "loss": 0.2838, + "step": 35906 + }, + { + "epoch": 0.8417166265229699, + "grad_norm": 2.0578441619873047, + "learning_rate": 1.211562346063847e-05, + "loss": 0.2164, + "step": 35907 + }, + { + "epoch": 0.8417400680977748, + "grad_norm": 2.973019599914551, + "learning_rate": 1.2112109233950109e-05, + "loss": 0.202, + "step": 35908 + }, + { + "epoch": 0.8417635096725798, + "grad_norm": 0.3541240692138672, + "learning_rate": 1.2108595484143915e-05, + "loss": 0.0291, + "step": 35909 + }, + { + "epoch": 0.8417869512473848, + "grad_norm": 0.856185257434845, + "learning_rate": 1.210508221123895e-05, + "loss": 0.0253, + "step": 35910 + }, + { + "epoch": 0.8418103928221898, + "grad_norm": 1.622029185295105, + "learning_rate": 1.2101569415254289e-05, + "loss": 0.0587, + "step": 35911 + }, + { + "epoch": 0.8418338343969948, + "grad_norm": 2.6295716762542725, + "learning_rate": 1.2098057096208947e-05, + "loss": 0.1295, + "step": 35912 + }, + { + "epoch": 0.8418572759717998, + "grad_norm": 3.8796582221984863, + "learning_rate": 1.2094545254122058e-05, + "loss": 0.2609, + "step": 35913 + }, + { + "epoch": 0.8418807175466048, + "grad_norm": 2.9367904663085938, + "learning_rate": 1.2091033889012625e-05, + "loss": 1.5872, + "step": 35914 + }, + { + "epoch": 0.8419041591214098, + "grad_norm": 0.8059682846069336, + "learning_rate": 1.2087523000899725e-05, + "loss": 0.0719, + "step": 35915 + }, + { + "epoch": 0.8419276006962148, + "grad_norm": 1.4324671030044556, + "learning_rate": 1.2084012589802385e-05, + "loss": 0.0772, + "step": 35916 + }, + { + "epoch": 0.8419510422710198, + "grad_norm": 1.1439380645751953, + "learning_rate": 1.2080502655739656e-05, + "loss": 0.6746, + "step": 35917 + }, + { + "epoch": 0.8419744838458247, + "grad_norm": 1.8959721326828003, + "learning_rate": 1.2076993198730601e-05, + "loss": 0.1962, + "step": 35918 + }, + { + "epoch": 0.8419979254206298, + "grad_norm": 1.6068743467330933, + "learning_rate": 1.207348421879424e-05, + "loss": 0.0673, + "step": 35919 + }, + { + "epoch": 0.8420213669954347, + "grad_norm": 3.154232978820801, + "learning_rate": 1.2069975715949645e-05, + "loss": 0.1728, + "step": 35920 + }, + { + "epoch": 0.8420448085702398, + "grad_norm": 2.0050597190856934, + "learning_rate": 1.2066467690215832e-05, + "loss": 0.0858, + "step": 35921 + }, + { + "epoch": 0.8420682501450447, + "grad_norm": 3.320467710494995, + "learning_rate": 1.2062960141611845e-05, + "loss": 0.2934, + "step": 35922 + }, + { + "epoch": 0.8420916917198498, + "grad_norm": 2.843580484390259, + "learning_rate": 1.2059453070156702e-05, + "loss": 0.1635, + "step": 35923 + }, + { + "epoch": 0.8421151332946547, + "grad_norm": 1.1447056531906128, + "learning_rate": 1.2055946475869417e-05, + "loss": 0.0678, + "step": 35924 + }, + { + "epoch": 0.8421385748694598, + "grad_norm": 1.956363320350647, + "learning_rate": 1.2052440358769058e-05, + "loss": 0.1179, + "step": 35925 + }, + { + "epoch": 0.8421620164442647, + "grad_norm": 0.9018441438674927, + "learning_rate": 1.2048934718874638e-05, + "loss": 0.0701, + "step": 35926 + }, + { + "epoch": 0.8421854580190697, + "grad_norm": 2.0718586444854736, + "learning_rate": 1.2045429556205168e-05, + "loss": 0.1593, + "step": 35927 + }, + { + "epoch": 0.8422088995938747, + "grad_norm": 3.169797658920288, + "learning_rate": 1.2041924870779675e-05, + "loss": 0.2776, + "step": 35928 + }, + { + "epoch": 0.8422323411686797, + "grad_norm": 0.9769853353500366, + "learning_rate": 1.2038420662617167e-05, + "loss": 0.0353, + "step": 35929 + }, + { + "epoch": 0.8422557827434847, + "grad_norm": 1.1432104110717773, + "learning_rate": 1.203491693173664e-05, + "loss": 0.1357, + "step": 35930 + }, + { + "epoch": 0.8422792243182897, + "grad_norm": 0.9224042892456055, + "learning_rate": 1.2031413678157122e-05, + "loss": 0.0551, + "step": 35931 + }, + { + "epoch": 0.8423026658930947, + "grad_norm": 3.2544357776641846, + "learning_rate": 1.2027910901897654e-05, + "loss": 0.1938, + "step": 35932 + }, + { + "epoch": 0.8423261074678997, + "grad_norm": 2.5416557788848877, + "learning_rate": 1.2024408602977222e-05, + "loss": 0.1151, + "step": 35933 + }, + { + "epoch": 0.8423495490427046, + "grad_norm": 3.5459702014923096, + "learning_rate": 1.202090678141481e-05, + "loss": 0.3419, + "step": 35934 + }, + { + "epoch": 0.8423729906175097, + "grad_norm": 3.2331604957580566, + "learning_rate": 1.2017405437229434e-05, + "loss": 0.2533, + "step": 35935 + }, + { + "epoch": 0.8423964321923146, + "grad_norm": 1.457999587059021, + "learning_rate": 1.2013904570440093e-05, + "loss": 0.0626, + "step": 35936 + }, + { + "epoch": 0.8424198737671197, + "grad_norm": 1.7488006353378296, + "learning_rate": 1.201040418106576e-05, + "loss": 0.0751, + "step": 35937 + }, + { + "epoch": 0.8424433153419246, + "grad_norm": 2.187708616256714, + "learning_rate": 1.2006904269125464e-05, + "loss": 1.6299, + "step": 35938 + }, + { + "epoch": 0.8424667569167297, + "grad_norm": 2.0701205730438232, + "learning_rate": 1.2003404834638187e-05, + "loss": 0.1464, + "step": 35939 + }, + { + "epoch": 0.8424901984915346, + "grad_norm": 2.2104554176330566, + "learning_rate": 1.1999905877622907e-05, + "loss": 1.1856, + "step": 35940 + }, + { + "epoch": 0.8425136400663397, + "grad_norm": 1.33004891872406, + "learning_rate": 1.1996407398098607e-05, + "loss": 0.1006, + "step": 35941 + }, + { + "epoch": 0.8425370816411446, + "grad_norm": 2.372859477996826, + "learning_rate": 1.1992909396084263e-05, + "loss": 1.6134, + "step": 35942 + }, + { + "epoch": 0.8425605232159497, + "grad_norm": 2.064868211746216, + "learning_rate": 1.1989411871598877e-05, + "loss": 0.2443, + "step": 35943 + }, + { + "epoch": 0.8425839647907546, + "grad_norm": 2.399012804031372, + "learning_rate": 1.1985914824661404e-05, + "loss": 0.1407, + "step": 35944 + }, + { + "epoch": 0.8426074063655596, + "grad_norm": 0.3216117322444916, + "learning_rate": 1.198241825529085e-05, + "loss": 0.0249, + "step": 35945 + }, + { + "epoch": 0.8426308479403646, + "grad_norm": 4.891662120819092, + "learning_rate": 1.1978922163506168e-05, + "loss": 0.3056, + "step": 35946 + }, + { + "epoch": 0.8426542895151696, + "grad_norm": 0.26193967461586, + "learning_rate": 1.197542654932633e-05, + "loss": 0.0103, + "step": 35947 + }, + { + "epoch": 0.8426777310899747, + "grad_norm": 1.009522795677185, + "learning_rate": 1.1971931412770298e-05, + "loss": 0.1183, + "step": 35948 + }, + { + "epoch": 0.8427011726647796, + "grad_norm": 3.6174209117889404, + "learning_rate": 1.1968436753857049e-05, + "loss": 0.1838, + "step": 35949 + }, + { + "epoch": 0.8427246142395847, + "grad_norm": 2.270955801010132, + "learning_rate": 1.196494257260552e-05, + "loss": 1.4448, + "step": 35950 + }, + { + "epoch": 0.8427480558143896, + "grad_norm": 0.8415481448173523, + "learning_rate": 1.1961448869034698e-05, + "loss": 0.056, + "step": 35951 + }, + { + "epoch": 0.8427714973891947, + "grad_norm": 1.2848756313323975, + "learning_rate": 1.1957955643163531e-05, + "loss": 0.1455, + "step": 35952 + }, + { + "epoch": 0.8427949389639996, + "grad_norm": 0.9122716784477234, + "learning_rate": 1.1954462895010975e-05, + "loss": 0.074, + "step": 35953 + }, + { + "epoch": 0.8428183805388046, + "grad_norm": 0.6950517892837524, + "learning_rate": 1.195097062459598e-05, + "loss": 0.0872, + "step": 35954 + }, + { + "epoch": 0.8428418221136096, + "grad_norm": 1.5073260068893433, + "learning_rate": 1.1947478831937475e-05, + "loss": 0.2082, + "step": 35955 + }, + { + "epoch": 0.8428652636884146, + "grad_norm": 2.5163915157318115, + "learning_rate": 1.1943987517054445e-05, + "loss": 0.0851, + "step": 35956 + }, + { + "epoch": 0.8428887052632196, + "grad_norm": 2.374718189239502, + "learning_rate": 1.1940496679965796e-05, + "loss": 0.0977, + "step": 35957 + }, + { + "epoch": 0.8429121468380246, + "grad_norm": 5.7703728675842285, + "learning_rate": 1.1937006320690503e-05, + "loss": 0.1912, + "step": 35958 + }, + { + "epoch": 0.8429355884128296, + "grad_norm": 3.0484728813171387, + "learning_rate": 1.1933516439247494e-05, + "loss": 0.2152, + "step": 35959 + }, + { + "epoch": 0.8429590299876346, + "grad_norm": 1.3699233531951904, + "learning_rate": 1.19300270356557e-05, + "loss": 0.1076, + "step": 35960 + }, + { + "epoch": 0.8429824715624396, + "grad_norm": 3.064157485961914, + "learning_rate": 1.192653810993406e-05, + "loss": 0.2039, + "step": 35961 + }, + { + "epoch": 0.8430059131372446, + "grad_norm": 1.7092009782791138, + "learning_rate": 1.192304966210147e-05, + "loss": 0.1227, + "step": 35962 + }, + { + "epoch": 0.8430293547120495, + "grad_norm": 1.1749013662338257, + "learning_rate": 1.1919561692176917e-05, + "loss": 0.0514, + "step": 35963 + }, + { + "epoch": 0.8430527962868546, + "grad_norm": 2.5584285259246826, + "learning_rate": 1.1916074200179306e-05, + "loss": 0.1245, + "step": 35964 + }, + { + "epoch": 0.8430762378616595, + "grad_norm": 3.137758255004883, + "learning_rate": 1.1912587186127545e-05, + "loss": 0.2278, + "step": 35965 + }, + { + "epoch": 0.8430996794364646, + "grad_norm": 4.000250339508057, + "learning_rate": 1.190910065004056e-05, + "loss": 0.302, + "step": 35966 + }, + { + "epoch": 0.8431231210112695, + "grad_norm": 2.779111862182617, + "learning_rate": 1.1905614591937252e-05, + "loss": 0.2266, + "step": 35967 + }, + { + "epoch": 0.8431465625860746, + "grad_norm": 2.951554536819458, + "learning_rate": 1.190212901183657e-05, + "loss": 0.1467, + "step": 35968 + }, + { + "epoch": 0.8431700041608795, + "grad_norm": 2.7718048095703125, + "learning_rate": 1.1898643909757412e-05, + "loss": 1.3729, + "step": 35969 + }, + { + "epoch": 0.8431934457356846, + "grad_norm": 2.8642280101776123, + "learning_rate": 1.1895159285718693e-05, + "loss": 0.2577, + "step": 35970 + }, + { + "epoch": 0.8432168873104895, + "grad_norm": 1.177870750427246, + "learning_rate": 1.1891675139739321e-05, + "loss": 0.1095, + "step": 35971 + }, + { + "epoch": 0.8432403288852945, + "grad_norm": 0.642498791217804, + "learning_rate": 1.1888191471838205e-05, + "loss": 0.0389, + "step": 35972 + }, + { + "epoch": 0.8432637704600995, + "grad_norm": 1.8349906206130981, + "learning_rate": 1.188470828203423e-05, + "loss": 0.1222, + "step": 35973 + }, + { + "epoch": 0.8432872120349045, + "grad_norm": 3.286208152770996, + "learning_rate": 1.188122557034631e-05, + "loss": 0.1826, + "step": 35974 + }, + { + "epoch": 0.8433106536097095, + "grad_norm": 3.6307692527770996, + "learning_rate": 1.1877743336793313e-05, + "loss": 0.3926, + "step": 35975 + }, + { + "epoch": 0.8433340951845145, + "grad_norm": 1.689772367477417, + "learning_rate": 1.1874261581394187e-05, + "loss": 1.7884, + "step": 35976 + }, + { + "epoch": 0.8433575367593195, + "grad_norm": 2.222111463546753, + "learning_rate": 1.1870780304167783e-05, + "loss": 0.0767, + "step": 35977 + }, + { + "epoch": 0.8433809783341245, + "grad_norm": 3.361598491668701, + "learning_rate": 1.1867299505133012e-05, + "loss": 0.2995, + "step": 35978 + }, + { + "epoch": 0.8434044199089294, + "grad_norm": 2.4814605712890625, + "learning_rate": 1.1863819184308733e-05, + "loss": 0.1643, + "step": 35979 + }, + { + "epoch": 0.8434278614837345, + "grad_norm": 0.29888415336608887, + "learning_rate": 1.1860339341713866e-05, + "loss": 0.0132, + "step": 35980 + }, + { + "epoch": 0.8434513030585394, + "grad_norm": 2.1617507934570312, + "learning_rate": 1.1856859977367273e-05, + "loss": 0.2228, + "step": 35981 + }, + { + "epoch": 0.8434747446333445, + "grad_norm": 2.1200735569000244, + "learning_rate": 1.185338109128782e-05, + "loss": 0.2126, + "step": 35982 + }, + { + "epoch": 0.8434981862081494, + "grad_norm": 4.632262706756592, + "learning_rate": 1.1849902683494418e-05, + "loss": 0.2786, + "step": 35983 + }, + { + "epoch": 0.8435216277829545, + "grad_norm": 2.158616065979004, + "learning_rate": 1.1846424754005925e-05, + "loss": 0.1195, + "step": 35984 + }, + { + "epoch": 0.8435450693577594, + "grad_norm": 2.1350739002227783, + "learning_rate": 1.1842947302841213e-05, + "loss": 0.1159, + "step": 35985 + }, + { + "epoch": 0.8435685109325645, + "grad_norm": 1.2733345031738281, + "learning_rate": 1.1839470330019143e-05, + "loss": 0.1268, + "step": 35986 + }, + { + "epoch": 0.8435919525073694, + "grad_norm": 5.126357555389404, + "learning_rate": 1.1835993835558589e-05, + "loss": 0.4627, + "step": 35987 + }, + { + "epoch": 0.8436153940821745, + "grad_norm": 2.186534881591797, + "learning_rate": 1.1832517819478383e-05, + "loss": 1.4596, + "step": 35988 + }, + { + "epoch": 0.8436388356569794, + "grad_norm": 0.9439871311187744, + "learning_rate": 1.1829042281797443e-05, + "loss": 0.0821, + "step": 35989 + }, + { + "epoch": 0.8436622772317844, + "grad_norm": 0.22300799190998077, + "learning_rate": 1.1825567222534595e-05, + "loss": 0.0076, + "step": 35990 + }, + { + "epoch": 0.8436857188065894, + "grad_norm": 3.125619649887085, + "learning_rate": 1.1822092641708683e-05, + "loss": 0.2216, + "step": 35991 + }, + { + "epoch": 0.8437091603813944, + "grad_norm": 2.856182813644409, + "learning_rate": 1.1818618539338588e-05, + "loss": 0.1788, + "step": 35992 + }, + { + "epoch": 0.8437326019561994, + "grad_norm": 2.967231512069702, + "learning_rate": 1.1815144915443155e-05, + "loss": 0.204, + "step": 35993 + }, + { + "epoch": 0.8437560435310044, + "grad_norm": 3.4721767902374268, + "learning_rate": 1.18116717700412e-05, + "loss": 0.2658, + "step": 35994 + }, + { + "epoch": 0.8437794851058094, + "grad_norm": 3.4427173137664795, + "learning_rate": 1.1808199103151619e-05, + "loss": 0.1652, + "step": 35995 + }, + { + "epoch": 0.8438029266806144, + "grad_norm": 2.6945574283599854, + "learning_rate": 1.1804726914793218e-05, + "loss": 0.2455, + "step": 35996 + }, + { + "epoch": 0.8438263682554193, + "grad_norm": 5.016499042510986, + "learning_rate": 1.1801255204984862e-05, + "loss": 0.2817, + "step": 35997 + }, + { + "epoch": 0.8438498098302244, + "grad_norm": 2.144686460494995, + "learning_rate": 1.1797783973745369e-05, + "loss": 0.2512, + "step": 35998 + }, + { + "epoch": 0.8438732514050294, + "grad_norm": 1.5846375226974487, + "learning_rate": 1.1794313221093579e-05, + "loss": 0.1002, + "step": 35999 + }, + { + "epoch": 0.8438966929798344, + "grad_norm": 2.3562428951263428, + "learning_rate": 1.1790842947048308e-05, + "loss": 0.1588, + "step": 36000 + }, + { + "epoch": 0.8439201345546394, + "grad_norm": 1.0298197269439697, + "learning_rate": 1.1787373151628423e-05, + "loss": 0.0832, + "step": 36001 + }, + { + "epoch": 0.8439435761294444, + "grad_norm": 2.4366838932037354, + "learning_rate": 1.1783903834852738e-05, + "loss": 0.2213, + "step": 36002 + }, + { + "epoch": 0.8439670177042494, + "grad_norm": 4.090799808502197, + "learning_rate": 1.1780434996740042e-05, + "loss": 0.1473, + "step": 36003 + }, + { + "epoch": 0.8439904592790544, + "grad_norm": 2.0824084281921387, + "learning_rate": 1.1776966637309205e-05, + "loss": 0.1235, + "step": 36004 + }, + { + "epoch": 0.8440139008538594, + "grad_norm": 3.0702502727508545, + "learning_rate": 1.1773498756579037e-05, + "loss": 0.1892, + "step": 36005 + }, + { + "epoch": 0.8440373424286644, + "grad_norm": 3.3027634620666504, + "learning_rate": 1.1770031354568345e-05, + "loss": 0.3222, + "step": 36006 + }, + { + "epoch": 0.8440607840034694, + "grad_norm": 1.0185514688491821, + "learning_rate": 1.1766564431295912e-05, + "loss": 0.0921, + "step": 36007 + }, + { + "epoch": 0.8440842255782743, + "grad_norm": 2.0605647563934326, + "learning_rate": 1.1763097986780613e-05, + "loss": 0.1189, + "step": 36008 + }, + { + "epoch": 0.8441076671530794, + "grad_norm": 1.4844671487808228, + "learning_rate": 1.1759632021041223e-05, + "loss": 0.0826, + "step": 36009 + }, + { + "epoch": 0.8441311087278843, + "grad_norm": 0.5170246362686157, + "learning_rate": 1.1756166534096547e-05, + "loss": 0.0336, + "step": 36010 + }, + { + "epoch": 0.8441545503026894, + "grad_norm": 1.4706791639328003, + "learning_rate": 1.1752701525965393e-05, + "loss": 0.1588, + "step": 36011 + }, + { + "epoch": 0.8441779918774943, + "grad_norm": 1.423570156097412, + "learning_rate": 1.1749236996666568e-05, + "loss": 0.8976, + "step": 36012 + }, + { + "epoch": 0.8442014334522994, + "grad_norm": 3.510988235473633, + "learning_rate": 1.1745772946218835e-05, + "loss": 0.3142, + "step": 36013 + }, + { + "epoch": 0.8442248750271043, + "grad_norm": 1.3896682262420654, + "learning_rate": 1.1742309374641047e-05, + "loss": 0.143, + "step": 36014 + }, + { + "epoch": 0.8442483166019094, + "grad_norm": 6.769565105438232, + "learning_rate": 1.1738846281951943e-05, + "loss": 0.2702, + "step": 36015 + }, + { + "epoch": 0.8442717581767143, + "grad_norm": 3.3797073364257812, + "learning_rate": 1.1735383668170363e-05, + "loss": 0.1881, + "step": 36016 + }, + { + "epoch": 0.8442951997515193, + "grad_norm": 2.914292812347412, + "learning_rate": 1.173192153331507e-05, + "loss": 1.5385, + "step": 36017 + }, + { + "epoch": 0.8443186413263243, + "grad_norm": 2.0693016052246094, + "learning_rate": 1.1728459877404863e-05, + "loss": 0.1301, + "step": 36018 + }, + { + "epoch": 0.8443420829011293, + "grad_norm": 2.169481039047241, + "learning_rate": 1.1724998700458512e-05, + "loss": 0.181, + "step": 36019 + }, + { + "epoch": 0.8443655244759343, + "grad_norm": 2.55147123336792, + "learning_rate": 1.172153800249477e-05, + "loss": 1.4105, + "step": 36020 + }, + { + "epoch": 0.8443889660507393, + "grad_norm": 1.5163744688034058, + "learning_rate": 1.1718077783532477e-05, + "loss": 0.2499, + "step": 36021 + }, + { + "epoch": 0.8444124076255443, + "grad_norm": 2.7797176837921143, + "learning_rate": 1.1714618043590376e-05, + "loss": 0.2223, + "step": 36022 + }, + { + "epoch": 0.8444358492003493, + "grad_norm": 2.440319776535034, + "learning_rate": 1.1711158782687226e-05, + "loss": 0.1616, + "step": 36023 + }, + { + "epoch": 0.8444592907751542, + "grad_norm": 0.5872730016708374, + "learning_rate": 1.1707700000841825e-05, + "loss": 0.0328, + "step": 36024 + }, + { + "epoch": 0.8444827323499593, + "grad_norm": 2.8289079666137695, + "learning_rate": 1.1704241698072916e-05, + "loss": 0.1895, + "step": 36025 + }, + { + "epoch": 0.8445061739247642, + "grad_norm": 4.018872261047363, + "learning_rate": 1.1700783874399269e-05, + "loss": 0.1766, + "step": 36026 + }, + { + "epoch": 0.8445296154995693, + "grad_norm": 1.2203236818313599, + "learning_rate": 1.1697326529839658e-05, + "loss": 0.0953, + "step": 36027 + }, + { + "epoch": 0.8445530570743742, + "grad_norm": 3.1292216777801514, + "learning_rate": 1.1693869664412815e-05, + "loss": 0.2086, + "step": 36028 + }, + { + "epoch": 0.8445764986491793, + "grad_norm": 1.5616544485092163, + "learning_rate": 1.1690413278137547e-05, + "loss": 0.08, + "step": 36029 + }, + { + "epoch": 0.8445999402239842, + "grad_norm": 1.1079479455947876, + "learning_rate": 1.168695737103258e-05, + "loss": 0.0661, + "step": 36030 + }, + { + "epoch": 0.8446233817987893, + "grad_norm": 0.5381124019622803, + "learning_rate": 1.1683501943116658e-05, + "loss": 0.0339, + "step": 36031 + }, + { + "epoch": 0.8446468233735942, + "grad_norm": 2.7251133918762207, + "learning_rate": 1.1680046994408522e-05, + "loss": 2.0346, + "step": 36032 + }, + { + "epoch": 0.8446702649483993, + "grad_norm": 1.0013726949691772, + "learning_rate": 1.1676592524926955e-05, + "loss": 0.0966, + "step": 36033 + }, + { + "epoch": 0.8446937065232042, + "grad_norm": 2.1867778301239014, + "learning_rate": 1.1673138534690675e-05, + "loss": 0.2012, + "step": 36034 + }, + { + "epoch": 0.8447171480980092, + "grad_norm": 1.6146423816680908, + "learning_rate": 1.1669685023718435e-05, + "loss": 0.0673, + "step": 36035 + }, + { + "epoch": 0.8447405896728142, + "grad_norm": 1.483388066291809, + "learning_rate": 1.1666231992028964e-05, + "loss": 0.123, + "step": 36036 + }, + { + "epoch": 0.8447640312476192, + "grad_norm": 2.5867998600006104, + "learning_rate": 1.1662779439641003e-05, + "loss": 0.2606, + "step": 36037 + }, + { + "epoch": 0.8447874728224242, + "grad_norm": 4.0451812744140625, + "learning_rate": 1.165932736657327e-05, + "loss": 0.2286, + "step": 36038 + }, + { + "epoch": 0.8448109143972292, + "grad_norm": 3.076920747756958, + "learning_rate": 1.1655875772844516e-05, + "loss": 0.2163, + "step": 36039 + }, + { + "epoch": 0.8448343559720342, + "grad_norm": 2.4833874702453613, + "learning_rate": 1.165242465847346e-05, + "loss": 0.1881, + "step": 36040 + }, + { + "epoch": 0.8448577975468392, + "grad_norm": 1.8528109788894653, + "learning_rate": 1.1648974023478842e-05, + "loss": 0.1057, + "step": 36041 + }, + { + "epoch": 0.8448812391216441, + "grad_norm": 2.741861343383789, + "learning_rate": 1.1645523867879382e-05, + "loss": 0.1426, + "step": 36042 + }, + { + "epoch": 0.8449046806964492, + "grad_norm": 1.6594964265823364, + "learning_rate": 1.1642074191693785e-05, + "loss": 0.0684, + "step": 36043 + }, + { + "epoch": 0.8449281222712541, + "grad_norm": 2.375408887863159, + "learning_rate": 1.1638624994940783e-05, + "loss": 0.213, + "step": 36044 + }, + { + "epoch": 0.8449515638460592, + "grad_norm": 0.4083400368690491, + "learning_rate": 1.163517627763907e-05, + "loss": 0.0428, + "step": 36045 + }, + { + "epoch": 0.8449750054208641, + "grad_norm": 2.6132214069366455, + "learning_rate": 1.1631728039807388e-05, + "loss": 0.1752, + "step": 36046 + }, + { + "epoch": 0.8449984469956692, + "grad_norm": 1.254619836807251, + "learning_rate": 1.1628280281464443e-05, + "loss": 0.0785, + "step": 36047 + }, + { + "epoch": 0.8450218885704741, + "grad_norm": 0.9714934229850769, + "learning_rate": 1.1624833002628932e-05, + "loss": 0.0746, + "step": 36048 + }, + { + "epoch": 0.8450453301452792, + "grad_norm": 0.26366907358169556, + "learning_rate": 1.1621386203319551e-05, + "loss": 0.0068, + "step": 36049 + }, + { + "epoch": 0.8450687717200841, + "grad_norm": 3.5777790546417236, + "learning_rate": 1.1617939883555029e-05, + "loss": 0.244, + "step": 36050 + }, + { + "epoch": 0.8450922132948891, + "grad_norm": 2.712634325027466, + "learning_rate": 1.1614494043354018e-05, + "loss": 0.1778, + "step": 36051 + }, + { + "epoch": 0.8451156548696942, + "grad_norm": 1.331211805343628, + "learning_rate": 1.1611048682735248e-05, + "loss": 0.0852, + "step": 36052 + }, + { + "epoch": 0.8451390964444991, + "grad_norm": 2.6384849548339844, + "learning_rate": 1.1607603801717437e-05, + "loss": 0.1516, + "step": 36053 + }, + { + "epoch": 0.8451625380193042, + "grad_norm": 1.961334466934204, + "learning_rate": 1.160415940031926e-05, + "loss": 0.0924, + "step": 36054 + }, + { + "epoch": 0.8451859795941091, + "grad_norm": 2.2173588275909424, + "learning_rate": 1.1600715478559387e-05, + "loss": 0.1731, + "step": 36055 + }, + { + "epoch": 0.8452094211689142, + "grad_norm": 1.0654244422912598, + "learning_rate": 1.159727203645653e-05, + "loss": 0.0687, + "step": 36056 + }, + { + "epoch": 0.8452328627437191, + "grad_norm": 1.8520315885543823, + "learning_rate": 1.159382907402935e-05, + "loss": 0.1213, + "step": 36057 + }, + { + "epoch": 0.8452563043185242, + "grad_norm": 1.3232923746109009, + "learning_rate": 1.1590386591296521e-05, + "loss": 0.0734, + "step": 36058 + }, + { + "epoch": 0.8452797458933291, + "grad_norm": 2.1052677631378174, + "learning_rate": 1.1586944588276772e-05, + "loss": 0.1352, + "step": 36059 + }, + { + "epoch": 0.8453031874681342, + "grad_norm": 0.6080660223960876, + "learning_rate": 1.1583503064988732e-05, + "loss": 0.0359, + "step": 36060 + }, + { + "epoch": 0.8453266290429391, + "grad_norm": 1.8027141094207764, + "learning_rate": 1.15800620214511e-05, + "loss": 0.1406, + "step": 36061 + }, + { + "epoch": 0.8453500706177441, + "grad_norm": 1.9201072454452515, + "learning_rate": 1.1576621457682535e-05, + "loss": 0.1385, + "step": 36062 + }, + { + "epoch": 0.8453735121925491, + "grad_norm": 1.8217155933380127, + "learning_rate": 1.1573181373701703e-05, + "loss": 0.1636, + "step": 36063 + }, + { + "epoch": 0.8453969537673541, + "grad_norm": 2.7560830116271973, + "learning_rate": 1.1569741769527265e-05, + "loss": 0.0926, + "step": 36064 + }, + { + "epoch": 0.8454203953421591, + "grad_norm": 2.735694408416748, + "learning_rate": 1.1566302645177884e-05, + "loss": 0.1464, + "step": 36065 + }, + { + "epoch": 0.8454438369169641, + "grad_norm": 2.6140403747558594, + "learning_rate": 1.1562864000672258e-05, + "loss": 0.1006, + "step": 36066 + }, + { + "epoch": 0.8454672784917691, + "grad_norm": 2.3948357105255127, + "learning_rate": 1.1559425836029013e-05, + "loss": 1.4407, + "step": 36067 + }, + { + "epoch": 0.8454907200665741, + "grad_norm": 2.572802782058716, + "learning_rate": 1.1555988151266816e-05, + "loss": 0.229, + "step": 36068 + }, + { + "epoch": 0.845514161641379, + "grad_norm": 1.8522396087646484, + "learning_rate": 1.1552550946404317e-05, + "loss": 1.7326, + "step": 36069 + }, + { + "epoch": 0.8455376032161841, + "grad_norm": 2.0290091037750244, + "learning_rate": 1.1549114221460133e-05, + "loss": 0.1352, + "step": 36070 + }, + { + "epoch": 0.845561044790989, + "grad_norm": 0.7559751868247986, + "learning_rate": 1.1545677976452974e-05, + "loss": 0.031, + "step": 36071 + }, + { + "epoch": 0.8455844863657941, + "grad_norm": 3.4257001876831055, + "learning_rate": 1.1542242211401444e-05, + "loss": 0.3245, + "step": 36072 + }, + { + "epoch": 0.845607927940599, + "grad_norm": 5.123607635498047, + "learning_rate": 1.1538806926324198e-05, + "loss": 1.3067, + "step": 36073 + }, + { + "epoch": 0.8456313695154041, + "grad_norm": 3.456705331802368, + "learning_rate": 1.1535372121239874e-05, + "loss": 0.2632, + "step": 36074 + }, + { + "epoch": 0.845654811090209, + "grad_norm": 0.6341959238052368, + "learning_rate": 1.1531937796167114e-05, + "loss": 0.0434, + "step": 36075 + }, + { + "epoch": 0.8456782526650141, + "grad_norm": 4.415319442749023, + "learning_rate": 1.1528503951124514e-05, + "loss": 0.4583, + "step": 36076 + }, + { + "epoch": 0.845701694239819, + "grad_norm": 1.5557457208633423, + "learning_rate": 1.1525070586130749e-05, + "loss": 0.1461, + "step": 36077 + }, + { + "epoch": 0.845725135814624, + "grad_norm": 2.5110087394714355, + "learning_rate": 1.1521637701204458e-05, + "loss": 0.0578, + "step": 36078 + }, + { + "epoch": 0.845748577389429, + "grad_norm": 3.9917478561401367, + "learning_rate": 1.1518205296364238e-05, + "loss": 0.31, + "step": 36079 + }, + { + "epoch": 0.845772018964234, + "grad_norm": 1.8731763362884521, + "learning_rate": 1.151477337162874e-05, + "loss": 0.1486, + "step": 36080 + }, + { + "epoch": 0.845795460539039, + "grad_norm": 4.175073623657227, + "learning_rate": 1.151134192701656e-05, + "loss": 0.2152, + "step": 36081 + }, + { + "epoch": 0.845818902113844, + "grad_norm": 2.5184566974639893, + "learning_rate": 1.1507910962546331e-05, + "loss": 0.1119, + "step": 36082 + }, + { + "epoch": 0.845842343688649, + "grad_norm": 2.4446916580200195, + "learning_rate": 1.1504480478236635e-05, + "loss": 0.1681, + "step": 36083 + }, + { + "epoch": 0.845865785263454, + "grad_norm": 1.7236428260803223, + "learning_rate": 1.1501050474106146e-05, + "loss": 0.1223, + "step": 36084 + }, + { + "epoch": 0.845889226838259, + "grad_norm": 0.501498818397522, + "learning_rate": 1.1497620950173448e-05, + "loss": 0.0471, + "step": 36085 + }, + { + "epoch": 0.845912668413064, + "grad_norm": 1.0576014518737793, + "learning_rate": 1.1494191906457141e-05, + "loss": 0.089, + "step": 36086 + }, + { + "epoch": 0.845936109987869, + "grad_norm": 1.0448236465454102, + "learning_rate": 1.1490763342975842e-05, + "loss": 0.0833, + "step": 36087 + }, + { + "epoch": 0.845959551562674, + "grad_norm": 0.9704782962799072, + "learning_rate": 1.1487335259748122e-05, + "loss": 0.0754, + "step": 36088 + }, + { + "epoch": 0.8459829931374789, + "grad_norm": 2.941929817199707, + "learning_rate": 1.1483907656792637e-05, + "loss": 0.2149, + "step": 36089 + }, + { + "epoch": 0.846006434712284, + "grad_norm": 2.501932144165039, + "learning_rate": 1.1480480534127935e-05, + "loss": 1.5438, + "step": 36090 + }, + { + "epoch": 0.8460298762870889, + "grad_norm": 2.516146659851074, + "learning_rate": 1.1477053891772649e-05, + "loss": 0.0915, + "step": 36091 + }, + { + "epoch": 0.846053317861894, + "grad_norm": 0.6116661429405212, + "learning_rate": 1.1473627729745373e-05, + "loss": 0.0221, + "step": 36092 + }, + { + "epoch": 0.8460767594366989, + "grad_norm": 2.828110694885254, + "learning_rate": 1.147020204806467e-05, + "loss": 0.3263, + "step": 36093 + }, + { + "epoch": 0.846100201011504, + "grad_norm": 2.726045608520508, + "learning_rate": 1.1466776846749138e-05, + "loss": 0.3561, + "step": 36094 + }, + { + "epoch": 0.8461236425863089, + "grad_norm": 2.754817485809326, + "learning_rate": 1.1463352125817372e-05, + "loss": 0.0579, + "step": 36095 + }, + { + "epoch": 0.846147084161114, + "grad_norm": 4.449141025543213, + "learning_rate": 1.1459927885287924e-05, + "loss": 0.2188, + "step": 36096 + }, + { + "epoch": 0.8461705257359189, + "grad_norm": 2.837801933288574, + "learning_rate": 1.1456504125179412e-05, + "loss": 1.3478, + "step": 36097 + }, + { + "epoch": 0.8461939673107239, + "grad_norm": 0.9138662219047546, + "learning_rate": 1.1453080845510411e-05, + "loss": 0.0829, + "step": 36098 + }, + { + "epoch": 0.8462174088855289, + "grad_norm": 1.2276047468185425, + "learning_rate": 1.1449658046299471e-05, + "loss": 0.0847, + "step": 36099 + }, + { + "epoch": 0.8462408504603339, + "grad_norm": 2.6977319717407227, + "learning_rate": 1.1446235727565158e-05, + "loss": 0.166, + "step": 36100 + }, + { + "epoch": 0.8462642920351389, + "grad_norm": 3.548095464706421, + "learning_rate": 1.1442813889326087e-05, + "loss": 0.2407, + "step": 36101 + }, + { + "epoch": 0.8462877336099439, + "grad_norm": 3.731489419937134, + "learning_rate": 1.143939253160079e-05, + "loss": 0.2592, + "step": 36102 + }, + { + "epoch": 0.846311175184749, + "grad_norm": 1.738823652267456, + "learning_rate": 1.1435971654407817e-05, + "loss": 0.0878, + "step": 36103 + }, + { + "epoch": 0.8463346167595539, + "grad_norm": 1.7330073118209839, + "learning_rate": 1.1432551257765777e-05, + "loss": 1.1711, + "step": 36104 + }, + { + "epoch": 0.846358058334359, + "grad_norm": 2.5842998027801514, + "learning_rate": 1.1429131341693211e-05, + "loss": 0.2031, + "step": 36105 + }, + { + "epoch": 0.8463814999091639, + "grad_norm": 3.86816143989563, + "learning_rate": 1.1425711906208658e-05, + "loss": 0.5806, + "step": 36106 + }, + { + "epoch": 0.8464049414839689, + "grad_norm": 2.7169275283813477, + "learning_rate": 1.1422292951330683e-05, + "loss": 0.2122, + "step": 36107 + }, + { + "epoch": 0.8464283830587739, + "grad_norm": 3.018000602722168, + "learning_rate": 1.1418874477077823e-05, + "loss": 0.1813, + "step": 36108 + }, + { + "epoch": 0.8464518246335789, + "grad_norm": 6.395005702972412, + "learning_rate": 1.1415456483468657e-05, + "loss": 0.2883, + "step": 36109 + }, + { + "epoch": 0.8464752662083839, + "grad_norm": 1.209041714668274, + "learning_rate": 1.141203897052172e-05, + "loss": 0.1256, + "step": 36110 + }, + { + "epoch": 0.8464987077831889, + "grad_norm": 1.5565096139907837, + "learning_rate": 1.1408621938255548e-05, + "loss": 0.2496, + "step": 36111 + }, + { + "epoch": 0.8465221493579939, + "grad_norm": 4.4468913078308105, + "learning_rate": 1.1405205386688678e-05, + "loss": 0.4928, + "step": 36112 + }, + { + "epoch": 0.8465455909327989, + "grad_norm": 4.915998458862305, + "learning_rate": 1.140178931583964e-05, + "loss": 0.3199, + "step": 36113 + }, + { + "epoch": 0.8465690325076038, + "grad_norm": 1.0774368047714233, + "learning_rate": 1.1398373725727008e-05, + "loss": 0.0963, + "step": 36114 + }, + { + "epoch": 0.8465924740824089, + "grad_norm": 1.8686480522155762, + "learning_rate": 1.1394958616369267e-05, + "loss": 0.9957, + "step": 36115 + }, + { + "epoch": 0.8466159156572138, + "grad_norm": 0.6457666754722595, + "learning_rate": 1.1391543987784992e-05, + "loss": 0.042, + "step": 36116 + }, + { + "epoch": 0.8466393572320189, + "grad_norm": 3.41670298576355, + "learning_rate": 1.138812983999269e-05, + "loss": 0.2513, + "step": 36117 + }, + { + "epoch": 0.8466627988068238, + "grad_norm": 0.5710232853889465, + "learning_rate": 1.138471617301089e-05, + "loss": 0.0366, + "step": 36118 + }, + { + "epoch": 0.8466862403816289, + "grad_norm": 1.9071685075759888, + "learning_rate": 1.1381302986858112e-05, + "loss": 0.0981, + "step": 36119 + }, + { + "epoch": 0.8467096819564338, + "grad_norm": 0.63547682762146, + "learning_rate": 1.1377890281552872e-05, + "loss": 0.0462, + "step": 36120 + }, + { + "epoch": 0.8467331235312389, + "grad_norm": 1.1151357889175415, + "learning_rate": 1.1374478057113669e-05, + "loss": 0.1515, + "step": 36121 + }, + { + "epoch": 0.8467565651060438, + "grad_norm": 2.253324270248413, + "learning_rate": 1.1371066313559065e-05, + "loss": 0.1658, + "step": 36122 + }, + { + "epoch": 0.8467800066808489, + "grad_norm": 1.5034456253051758, + "learning_rate": 1.1367655050907544e-05, + "loss": 0.0627, + "step": 36123 + }, + { + "epoch": 0.8468034482556538, + "grad_norm": 2.5172438621520996, + "learning_rate": 1.1364244269177615e-05, + "loss": 0.1902, + "step": 36124 + }, + { + "epoch": 0.8468268898304588, + "grad_norm": 5.363185405731201, + "learning_rate": 1.1360833968387774e-05, + "loss": 0.4968, + "step": 36125 + }, + { + "epoch": 0.8468503314052638, + "grad_norm": 0.6106475591659546, + "learning_rate": 1.1357424148556561e-05, + "loss": 0.0139, + "step": 36126 + }, + { + "epoch": 0.8468737729800688, + "grad_norm": 1.623458743095398, + "learning_rate": 1.1354014809702462e-05, + "loss": 0.1274, + "step": 36127 + }, + { + "epoch": 0.8468972145548738, + "grad_norm": 3.8214423656463623, + "learning_rate": 1.135060595184394e-05, + "loss": 0.2454, + "step": 36128 + }, + { + "epoch": 0.8469206561296788, + "grad_norm": 0.48085516691207886, + "learning_rate": 1.1347197574999557e-05, + "loss": 0.0286, + "step": 36129 + }, + { + "epoch": 0.8469440977044838, + "grad_norm": 3.1791608333587646, + "learning_rate": 1.1343789679187756e-05, + "loss": 0.4121, + "step": 36130 + }, + { + "epoch": 0.8469675392792888, + "grad_norm": 0.5761058330535889, + "learning_rate": 1.1340382264427064e-05, + "loss": 0.0297, + "step": 36131 + }, + { + "epoch": 0.8469909808540937, + "grad_norm": 3.712629556655884, + "learning_rate": 1.1336975330735932e-05, + "loss": 0.3356, + "step": 36132 + }, + { + "epoch": 0.8470144224288988, + "grad_norm": 3.1122193336486816, + "learning_rate": 1.1333568878132872e-05, + "loss": 0.2198, + "step": 36133 + }, + { + "epoch": 0.8470378640037037, + "grad_norm": 2.4091391563415527, + "learning_rate": 1.1330162906636343e-05, + "loss": 0.1993, + "step": 36134 + }, + { + "epoch": 0.8470613055785088, + "grad_norm": 1.4452908039093018, + "learning_rate": 1.1326757416264866e-05, + "loss": 0.1922, + "step": 36135 + }, + { + "epoch": 0.8470847471533137, + "grad_norm": 0.8612813949584961, + "learning_rate": 1.1323352407036891e-05, + "loss": 0.0579, + "step": 36136 + }, + { + "epoch": 0.8471081887281188, + "grad_norm": 1.9075236320495605, + "learning_rate": 1.1319947878970872e-05, + "loss": 0.1388, + "step": 36137 + }, + { + "epoch": 0.8471316303029237, + "grad_norm": 1.2985849380493164, + "learning_rate": 1.1316543832085335e-05, + "loss": 0.1569, + "step": 36138 + }, + { + "epoch": 0.8471550718777288, + "grad_norm": 0.8524725437164307, + "learning_rate": 1.1313140266398736e-05, + "loss": 0.0877, + "step": 36139 + }, + { + "epoch": 0.8471785134525337, + "grad_norm": 1.3029247522354126, + "learning_rate": 1.1309737181929525e-05, + "loss": 0.0547, + "step": 36140 + }, + { + "epoch": 0.8472019550273387, + "grad_norm": 4.042747974395752, + "learning_rate": 1.1306334578696142e-05, + "loss": 0.4463, + "step": 36141 + }, + { + "epoch": 0.8472253966021437, + "grad_norm": 1.2755972146987915, + "learning_rate": 1.130293245671712e-05, + "loss": 0.0392, + "step": 36142 + }, + { + "epoch": 0.8472488381769487, + "grad_norm": 3.918200969696045, + "learning_rate": 1.1299530816010861e-05, + "loss": 0.2297, + "step": 36143 + }, + { + "epoch": 0.8472722797517537, + "grad_norm": 5.0648980140686035, + "learning_rate": 1.1296129656595856e-05, + "loss": 0.1962, + "step": 36144 + }, + { + "epoch": 0.8472957213265587, + "grad_norm": 2.416930913925171, + "learning_rate": 1.129272897849054e-05, + "loss": 0.2228, + "step": 36145 + }, + { + "epoch": 0.8473191629013637, + "grad_norm": 0.959492564201355, + "learning_rate": 1.128932878171336e-05, + "loss": 0.053, + "step": 36146 + }, + { + "epoch": 0.8473426044761687, + "grad_norm": 3.4592418670654297, + "learning_rate": 1.1285929066282785e-05, + "loss": 0.1764, + "step": 36147 + }, + { + "epoch": 0.8473660460509737, + "grad_norm": 0.8541285991668701, + "learning_rate": 1.1282529832217259e-05, + "loss": 0.0833, + "step": 36148 + }, + { + "epoch": 0.8473894876257787, + "grad_norm": 0.7709214687347412, + "learning_rate": 1.12791310795352e-05, + "loss": 0.0356, + "step": 36149 + }, + { + "epoch": 0.8474129292005836, + "grad_norm": 2.228813409805298, + "learning_rate": 1.127573280825508e-05, + "loss": 0.1467, + "step": 36150 + }, + { + "epoch": 0.8474363707753887, + "grad_norm": 1.7593690156936646, + "learning_rate": 1.1272335018395341e-05, + "loss": 0.0636, + "step": 36151 + }, + { + "epoch": 0.8474598123501936, + "grad_norm": 1.29677414894104, + "learning_rate": 1.1268937709974403e-05, + "loss": 0.1161, + "step": 36152 + }, + { + "epoch": 0.8474832539249987, + "grad_norm": 3.1121392250061035, + "learning_rate": 1.1265540883010684e-05, + "loss": 0.1222, + "step": 36153 + }, + { + "epoch": 0.8475066954998037, + "grad_norm": 2.383166551589966, + "learning_rate": 1.1262144537522657e-05, + "loss": 0.1966, + "step": 36154 + }, + { + "epoch": 0.8475301370746087, + "grad_norm": 1.5972130298614502, + "learning_rate": 1.125874867352873e-05, + "loss": 0.1002, + "step": 36155 + }, + { + "epoch": 0.8475535786494137, + "grad_norm": 3.090324640274048, + "learning_rate": 1.1255353291047321e-05, + "loss": 0.2016, + "step": 36156 + }, + { + "epoch": 0.8475770202242187, + "grad_norm": 2.399888038635254, + "learning_rate": 1.1251958390096874e-05, + "loss": 0.2166, + "step": 36157 + }, + { + "epoch": 0.8476004617990237, + "grad_norm": 1.0146117210388184, + "learning_rate": 1.124856397069578e-05, + "loss": 0.0317, + "step": 36158 + }, + { + "epoch": 0.8476239033738286, + "grad_norm": 1.1189751625061035, + "learning_rate": 1.1245170032862463e-05, + "loss": 0.1388, + "step": 36159 + }, + { + "epoch": 0.8476473449486337, + "grad_norm": 2.059664487838745, + "learning_rate": 1.1241776576615359e-05, + "loss": 0.0824, + "step": 36160 + }, + { + "epoch": 0.8476707865234386, + "grad_norm": 2.617129325866699, + "learning_rate": 1.1238383601972857e-05, + "loss": 0.2528, + "step": 36161 + }, + { + "epoch": 0.8476942280982437, + "grad_norm": 2.9378502368927, + "learning_rate": 1.1234991108953408e-05, + "loss": 0.2091, + "step": 36162 + }, + { + "epoch": 0.8477176696730486, + "grad_norm": 2.3132734298706055, + "learning_rate": 1.1231599097575385e-05, + "loss": 0.165, + "step": 36163 + }, + { + "epoch": 0.8477411112478537, + "grad_norm": 2.148484230041504, + "learning_rate": 1.1228207567857208e-05, + "loss": 0.0848, + "step": 36164 + }, + { + "epoch": 0.8477645528226586, + "grad_norm": 1.748273491859436, + "learning_rate": 1.1224816519817272e-05, + "loss": 0.1365, + "step": 36165 + }, + { + "epoch": 0.8477879943974637, + "grad_norm": 2.7642462253570557, + "learning_rate": 1.1221425953473952e-05, + "loss": 0.1379, + "step": 36166 + }, + { + "epoch": 0.8478114359722686, + "grad_norm": 3.221064329147339, + "learning_rate": 1.121803586884569e-05, + "loss": 0.3095, + "step": 36167 + }, + { + "epoch": 0.8478348775470737, + "grad_norm": 3.5495879650115967, + "learning_rate": 1.1214646265950868e-05, + "loss": 0.1861, + "step": 36168 + }, + { + "epoch": 0.8478583191218786, + "grad_norm": 5.9134931564331055, + "learning_rate": 1.1211257144807863e-05, + "loss": 0.7149, + "step": 36169 + }, + { + "epoch": 0.8478817606966836, + "grad_norm": 1.9030060768127441, + "learning_rate": 1.1207868505435082e-05, + "loss": 0.2038, + "step": 36170 + }, + { + "epoch": 0.8479052022714886, + "grad_norm": 3.5792572498321533, + "learning_rate": 1.1204480347850898e-05, + "loss": 0.2895, + "step": 36171 + }, + { + "epoch": 0.8479286438462936, + "grad_norm": 2.1772255897521973, + "learning_rate": 1.1201092672073688e-05, + "loss": 0.1689, + "step": 36172 + }, + { + "epoch": 0.8479520854210986, + "grad_norm": 2.5584819316864014, + "learning_rate": 1.1197705478121833e-05, + "loss": 0.1373, + "step": 36173 + }, + { + "epoch": 0.8479755269959036, + "grad_norm": 1.0415661334991455, + "learning_rate": 1.1194318766013756e-05, + "loss": 0.0941, + "step": 36174 + }, + { + "epoch": 0.8479989685707086, + "grad_norm": 3.140957832336426, + "learning_rate": 1.1190932535767795e-05, + "loss": 0.2543, + "step": 36175 + }, + { + "epoch": 0.8480224101455136, + "grad_norm": 2.3149027824401855, + "learning_rate": 1.1187546787402336e-05, + "loss": 0.2015, + "step": 36176 + }, + { + "epoch": 0.8480458517203185, + "grad_norm": 2.6290411949157715, + "learning_rate": 1.1184161520935742e-05, + "loss": 0.1853, + "step": 36177 + }, + { + "epoch": 0.8480692932951236, + "grad_norm": 1.3176709413528442, + "learning_rate": 1.118077673638639e-05, + "loss": 0.1192, + "step": 36178 + }, + { + "epoch": 0.8480927348699285, + "grad_norm": 4.060552597045898, + "learning_rate": 1.1177392433772615e-05, + "loss": 0.3299, + "step": 36179 + }, + { + "epoch": 0.8481161764447336, + "grad_norm": 0.6558215022087097, + "learning_rate": 1.1174008613112829e-05, + "loss": 0.0552, + "step": 36180 + }, + { + "epoch": 0.8481396180195385, + "grad_norm": 3.1350045204162598, + "learning_rate": 1.1170625274425373e-05, + "loss": 0.1971, + "step": 36181 + }, + { + "epoch": 0.8481630595943436, + "grad_norm": 2.081794261932373, + "learning_rate": 1.1167242417728608e-05, + "loss": 0.1426, + "step": 36182 + }, + { + "epoch": 0.8481865011691485, + "grad_norm": 2.0400495529174805, + "learning_rate": 1.1163860043040863e-05, + "loss": 1.3018, + "step": 36183 + }, + { + "epoch": 0.8482099427439536, + "grad_norm": 2.9186341762542725, + "learning_rate": 1.1160478150380504e-05, + "loss": 0.1273, + "step": 36184 + }, + { + "epoch": 0.8482333843187585, + "grad_norm": 1.0420894622802734, + "learning_rate": 1.1157096739765916e-05, + "loss": 0.0873, + "step": 36185 + }, + { + "epoch": 0.8482568258935635, + "grad_norm": 3.615163564682007, + "learning_rate": 1.1153715811215382e-05, + "loss": 1.2792, + "step": 36186 + }, + { + "epoch": 0.8482802674683685, + "grad_norm": 0.9515607953071594, + "learning_rate": 1.1150335364747311e-05, + "loss": 0.0657, + "step": 36187 + }, + { + "epoch": 0.8483037090431735, + "grad_norm": 2.4063498973846436, + "learning_rate": 1.1146955400380021e-05, + "loss": 0.1595, + "step": 36188 + }, + { + "epoch": 0.8483271506179785, + "grad_norm": 3.648190975189209, + "learning_rate": 1.1143575918131854e-05, + "loss": 0.1691, + "step": 36189 + }, + { + "epoch": 0.8483505921927835, + "grad_norm": 2.980754852294922, + "learning_rate": 1.1140196918021128e-05, + "loss": 0.2552, + "step": 36190 + }, + { + "epoch": 0.8483740337675885, + "grad_norm": 2.2212696075439453, + "learning_rate": 1.1136818400066184e-05, + "loss": 0.1989, + "step": 36191 + }, + { + "epoch": 0.8483974753423935, + "grad_norm": 2.6930179595947266, + "learning_rate": 1.1133440364285374e-05, + "loss": 0.2599, + "step": 36192 + }, + { + "epoch": 0.8484209169171985, + "grad_norm": 3.4066474437713623, + "learning_rate": 1.1130062810697028e-05, + "loss": 0.1254, + "step": 36193 + }, + { + "epoch": 0.8484443584920035, + "grad_norm": 2.5548462867736816, + "learning_rate": 1.1126685739319454e-05, + "loss": 0.1059, + "step": 36194 + }, + { + "epoch": 0.8484678000668084, + "grad_norm": 6.050387859344482, + "learning_rate": 1.112330915017098e-05, + "loss": 0.4871, + "step": 36195 + }, + { + "epoch": 0.8484912416416135, + "grad_norm": 2.532334327697754, + "learning_rate": 1.1119933043269926e-05, + "loss": 0.338, + "step": 36196 + }, + { + "epoch": 0.8485146832164184, + "grad_norm": 1.2207114696502686, + "learning_rate": 1.11165574186346e-05, + "loss": 0.0684, + "step": 36197 + }, + { + "epoch": 0.8485381247912235, + "grad_norm": 3.3153445720672607, + "learning_rate": 1.111318227628333e-05, + "loss": 0.2071, + "step": 36198 + }, + { + "epoch": 0.8485615663660284, + "grad_norm": 3.6362271308898926, + "learning_rate": 1.1109807616234447e-05, + "loss": 0.2234, + "step": 36199 + }, + { + "epoch": 0.8485850079408335, + "grad_norm": 7.240307331085205, + "learning_rate": 1.1106433438506258e-05, + "loss": 0.2974, + "step": 36200 + }, + { + "epoch": 0.8486084495156384, + "grad_norm": 1.1414474248886108, + "learning_rate": 1.110305974311705e-05, + "loss": 0.0562, + "step": 36201 + }, + { + "epoch": 0.8486318910904435, + "grad_norm": 0.954872727394104, + "learning_rate": 1.1099686530085151e-05, + "loss": 0.0811, + "step": 36202 + }, + { + "epoch": 0.8486553326652484, + "grad_norm": 0.23389825224876404, + "learning_rate": 1.1096313799428848e-05, + "loss": 0.009, + "step": 36203 + }, + { + "epoch": 0.8486787742400534, + "grad_norm": 3.725968599319458, + "learning_rate": 1.1092941551166425e-05, + "loss": 0.408, + "step": 36204 + }, + { + "epoch": 0.8487022158148585, + "grad_norm": 2.5979106426239014, + "learning_rate": 1.1089569785316212e-05, + "loss": 1.0984, + "step": 36205 + }, + { + "epoch": 0.8487256573896634, + "grad_norm": 3.943326711654663, + "learning_rate": 1.1086198501896505e-05, + "loss": 0.5007, + "step": 36206 + }, + { + "epoch": 0.8487490989644685, + "grad_norm": 2.0621109008789062, + "learning_rate": 1.108282770092558e-05, + "loss": 0.2078, + "step": 36207 + }, + { + "epoch": 0.8487725405392734, + "grad_norm": 1.955975890159607, + "learning_rate": 1.1079457382421742e-05, + "loss": 0.2429, + "step": 36208 + }, + { + "epoch": 0.8487959821140785, + "grad_norm": 0.8392907977104187, + "learning_rate": 1.1076087546403258e-05, + "loss": 0.0942, + "step": 36209 + }, + { + "epoch": 0.8488194236888834, + "grad_norm": 0.3702365458011627, + "learning_rate": 1.1072718192888409e-05, + "loss": 0.0233, + "step": 36210 + }, + { + "epoch": 0.8488428652636885, + "grad_norm": 5.313909530639648, + "learning_rate": 1.1069349321895483e-05, + "loss": 0.3212, + "step": 36211 + }, + { + "epoch": 0.8488663068384934, + "grad_norm": 0.2219880074262619, + "learning_rate": 1.1065980933442799e-05, + "loss": 0.016, + "step": 36212 + }, + { + "epoch": 0.8488897484132985, + "grad_norm": 2.8337244987487793, + "learning_rate": 1.1062613027548597e-05, + "loss": 0.2202, + "step": 36213 + }, + { + "epoch": 0.8489131899881034, + "grad_norm": 2.4997611045837402, + "learning_rate": 1.105924560423115e-05, + "loss": 0.1898, + "step": 36214 + }, + { + "epoch": 0.8489366315629084, + "grad_norm": 2.697080612182617, + "learning_rate": 1.1055878663508745e-05, + "loss": 0.1645, + "step": 36215 + }, + { + "epoch": 0.8489600731377134, + "grad_norm": 2.657111644744873, + "learning_rate": 1.1052512205399646e-05, + "loss": 0.1485, + "step": 36216 + }, + { + "epoch": 0.8489835147125184, + "grad_norm": 1.3801840543746948, + "learning_rate": 1.104914622992209e-05, + "loss": 0.0823, + "step": 36217 + }, + { + "epoch": 0.8490069562873234, + "grad_norm": 1.7400518655776978, + "learning_rate": 1.1045780737094392e-05, + "loss": 1.1795, + "step": 36218 + }, + { + "epoch": 0.8490303978621284, + "grad_norm": 2.5960655212402344, + "learning_rate": 1.1042415726934796e-05, + "loss": 0.2438, + "step": 36219 + }, + { + "epoch": 0.8490538394369334, + "grad_norm": 2.2997422218322754, + "learning_rate": 1.1039051199461536e-05, + "loss": 0.106, + "step": 36220 + }, + { + "epoch": 0.8490772810117384, + "grad_norm": 4.4639668464660645, + "learning_rate": 1.1035687154692898e-05, + "loss": 0.536, + "step": 36221 + }, + { + "epoch": 0.8491007225865433, + "grad_norm": 1.4250454902648926, + "learning_rate": 1.1032323592647099e-05, + "loss": 1.194, + "step": 36222 + }, + { + "epoch": 0.8491241641613484, + "grad_norm": 1.7447154521942139, + "learning_rate": 1.1028960513342435e-05, + "loss": 0.1538, + "step": 36223 + }, + { + "epoch": 0.8491476057361533, + "grad_norm": 1.7102041244506836, + "learning_rate": 1.1025597916797115e-05, + "loss": 1.1784, + "step": 36224 + }, + { + "epoch": 0.8491710473109584, + "grad_norm": 1.3627828359603882, + "learning_rate": 1.1022235803029423e-05, + "loss": 0.0668, + "step": 36225 + }, + { + "epoch": 0.8491944888857633, + "grad_norm": 1.5629600286483765, + "learning_rate": 1.101887417205758e-05, + "loss": 0.7778, + "step": 36226 + }, + { + "epoch": 0.8492179304605684, + "grad_norm": 3.5619990825653076, + "learning_rate": 1.1015513023899826e-05, + "loss": 0.423, + "step": 36227 + }, + { + "epoch": 0.8492413720353733, + "grad_norm": 3.516369581222534, + "learning_rate": 1.1012152358574401e-05, + "loss": 0.3215, + "step": 36228 + }, + { + "epoch": 0.8492648136101784, + "grad_norm": 0.9231324195861816, + "learning_rate": 1.1008792176099514e-05, + "loss": 0.0781, + "step": 36229 + }, + { + "epoch": 0.8492882551849833, + "grad_norm": 1.1547480821609497, + "learning_rate": 1.1005432476493449e-05, + "loss": 0.068, + "step": 36230 + }, + { + "epoch": 0.8493116967597883, + "grad_norm": 1.7739691734313965, + "learning_rate": 1.1002073259774414e-05, + "loss": 0.0833, + "step": 36231 + }, + { + "epoch": 0.8493351383345933, + "grad_norm": 3.2039968967437744, + "learning_rate": 1.099871452596063e-05, + "loss": 0.2206, + "step": 36232 + }, + { + "epoch": 0.8493585799093983, + "grad_norm": 2.6671547889709473, + "learning_rate": 1.0995356275070312e-05, + "loss": 0.1384, + "step": 36233 + }, + { + "epoch": 0.8493820214842033, + "grad_norm": 1.0266425609588623, + "learning_rate": 1.0991998507121682e-05, + "loss": 0.0904, + "step": 36234 + }, + { + "epoch": 0.8494054630590083, + "grad_norm": 4.318839073181152, + "learning_rate": 1.098864122213299e-05, + "loss": 0.3538, + "step": 36235 + }, + { + "epoch": 0.8494289046338133, + "grad_norm": 1.1298047304153442, + "learning_rate": 1.0985284420122422e-05, + "loss": 0.1064, + "step": 36236 + }, + { + "epoch": 0.8494523462086183, + "grad_norm": 1.0364676713943481, + "learning_rate": 1.0981928101108208e-05, + "loss": 0.1196, + "step": 36237 + }, + { + "epoch": 0.8494757877834233, + "grad_norm": 4.626591682434082, + "learning_rate": 1.0978572265108567e-05, + "loss": 0.2661, + "step": 36238 + }, + { + "epoch": 0.8494992293582283, + "grad_norm": 3.102911949157715, + "learning_rate": 1.0975216912141695e-05, + "loss": 0.3915, + "step": 36239 + }, + { + "epoch": 0.8495226709330332, + "grad_norm": 0.6095808148384094, + "learning_rate": 1.0971862042225789e-05, + "loss": 0.0348, + "step": 36240 + }, + { + "epoch": 0.8495461125078383, + "grad_norm": 1.3649773597717285, + "learning_rate": 1.0968507655379067e-05, + "loss": 0.1263, + "step": 36241 + }, + { + "epoch": 0.8495695540826432, + "grad_norm": 2.1819190979003906, + "learning_rate": 1.0965153751619717e-05, + "loss": 0.1745, + "step": 36242 + }, + { + "epoch": 0.8495929956574483, + "grad_norm": 2.7716310024261475, + "learning_rate": 1.0961800330965955e-05, + "loss": 0.1535, + "step": 36243 + }, + { + "epoch": 0.8496164372322532, + "grad_norm": 3.2871739864349365, + "learning_rate": 1.0958447393435967e-05, + "loss": 0.1787, + "step": 36244 + }, + { + "epoch": 0.8496398788070583, + "grad_norm": 1.9300687313079834, + "learning_rate": 1.095509493904795e-05, + "loss": 0.1202, + "step": 36245 + }, + { + "epoch": 0.8496633203818632, + "grad_norm": 0.3595808446407318, + "learning_rate": 1.0951742967820078e-05, + "loss": 0.0235, + "step": 36246 + }, + { + "epoch": 0.8496867619566683, + "grad_norm": 3.370509147644043, + "learning_rate": 1.094839147977057e-05, + "loss": 0.1567, + "step": 36247 + }, + { + "epoch": 0.8497102035314732, + "grad_norm": 1.3298004865646362, + "learning_rate": 1.094504047491759e-05, + "loss": 0.0998, + "step": 36248 + }, + { + "epoch": 0.8497336451062782, + "grad_norm": 2.4210638999938965, + "learning_rate": 1.0941689953279299e-05, + "loss": 1.5069, + "step": 36249 + }, + { + "epoch": 0.8497570866810832, + "grad_norm": 3.4372105598449707, + "learning_rate": 1.093833991487393e-05, + "loss": 0.2287, + "step": 36250 + }, + { + "epoch": 0.8497805282558882, + "grad_norm": 2.8484816551208496, + "learning_rate": 1.0934990359719633e-05, + "loss": 0.196, + "step": 36251 + }, + { + "epoch": 0.8498039698306932, + "grad_norm": 1.9709954261779785, + "learning_rate": 1.0931641287834582e-05, + "loss": 0.0865, + "step": 36252 + }, + { + "epoch": 0.8498274114054982, + "grad_norm": 0.7964591383934021, + "learning_rate": 1.0928292699236941e-05, + "loss": 0.037, + "step": 36253 + }, + { + "epoch": 0.8498508529803032, + "grad_norm": 2.3554089069366455, + "learning_rate": 1.0924944593944896e-05, + "loss": 1.6146, + "step": 36254 + }, + { + "epoch": 0.8498742945551082, + "grad_norm": 1.8672457933425903, + "learning_rate": 1.0921596971976588e-05, + "loss": 0.1656, + "step": 36255 + }, + { + "epoch": 0.8498977361299133, + "grad_norm": 1.2148507833480835, + "learning_rate": 1.0918249833350225e-05, + "loss": 0.1826, + "step": 36256 + }, + { + "epoch": 0.8499211777047182, + "grad_norm": 2.589920997619629, + "learning_rate": 1.0914903178083935e-05, + "loss": 0.1759, + "step": 36257 + }, + { + "epoch": 0.8499446192795233, + "grad_norm": 2.7490766048431396, + "learning_rate": 1.0911557006195883e-05, + "loss": 0.1965, + "step": 36258 + }, + { + "epoch": 0.8499680608543282, + "grad_norm": 3.080413341522217, + "learning_rate": 1.0908211317704208e-05, + "loss": 0.2861, + "step": 36259 + }, + { + "epoch": 0.8499915024291332, + "grad_norm": 1.7829477787017822, + "learning_rate": 1.090486611262711e-05, + "loss": 0.1373, + "step": 36260 + }, + { + "epoch": 0.8500149440039382, + "grad_norm": 3.6044657230377197, + "learning_rate": 1.0901521390982705e-05, + "loss": 0.2941, + "step": 36261 + }, + { + "epoch": 0.8500383855787432, + "grad_norm": 0.38991060853004456, + "learning_rate": 1.0898177152789135e-05, + "loss": 0.0137, + "step": 36262 + }, + { + "epoch": 0.8500618271535482, + "grad_norm": 0.5970289707183838, + "learning_rate": 1.0894833398064585e-05, + "loss": 0.0539, + "step": 36263 + }, + { + "epoch": 0.8500852687283532, + "grad_norm": 0.5906796455383301, + "learning_rate": 1.0891490126827165e-05, + "loss": 0.0603, + "step": 36264 + }, + { + "epoch": 0.8501087103031582, + "grad_norm": 3.4367918968200684, + "learning_rate": 1.0888147339095034e-05, + "loss": 0.3216, + "step": 36265 + }, + { + "epoch": 0.8501321518779632, + "grad_norm": 0.7776431441307068, + "learning_rate": 1.0884805034886314e-05, + "loss": 0.0581, + "step": 36266 + }, + { + "epoch": 0.8501555934527681, + "grad_norm": 1.5387697219848633, + "learning_rate": 1.0881463214219135e-05, + "loss": 0.1134, + "step": 36267 + }, + { + "epoch": 0.8501790350275732, + "grad_norm": 2.0934414863586426, + "learning_rate": 1.0878121877111657e-05, + "loss": 0.1586, + "step": 36268 + }, + { + "epoch": 0.8502024766023781, + "grad_norm": 3.5464296340942383, + "learning_rate": 1.0874781023581993e-05, + "loss": 0.2957, + "step": 36269 + }, + { + "epoch": 0.8502259181771832, + "grad_norm": 3.045438051223755, + "learning_rate": 1.087144065364828e-05, + "loss": 0.1562, + "step": 36270 + }, + { + "epoch": 0.8502493597519881, + "grad_norm": 3.2207019329071045, + "learning_rate": 1.0868100767328616e-05, + "loss": 0.457, + "step": 36271 + }, + { + "epoch": 0.8502728013267932, + "grad_norm": 1.7889384031295776, + "learning_rate": 1.0864761364641151e-05, + "loss": 1.2267, + "step": 36272 + }, + { + "epoch": 0.8502962429015981, + "grad_norm": 3.5165481567382812, + "learning_rate": 1.0861422445604009e-05, + "loss": 0.2168, + "step": 36273 + }, + { + "epoch": 0.8503196844764032, + "grad_norm": 1.1239739656448364, + "learning_rate": 1.085808401023527e-05, + "loss": 0.1146, + "step": 36274 + }, + { + "epoch": 0.8503431260512081, + "grad_norm": 3.9513866901397705, + "learning_rate": 1.08547460585531e-05, + "loss": 0.2534, + "step": 36275 + }, + { + "epoch": 0.8503665676260131, + "grad_norm": 2.0077149868011475, + "learning_rate": 1.0851408590575585e-05, + "loss": 0.1356, + "step": 36276 + }, + { + "epoch": 0.8503900092008181, + "grad_norm": 2.1312451362609863, + "learning_rate": 1.0848071606320832e-05, + "loss": 0.1032, + "step": 36277 + }, + { + "epoch": 0.8504134507756231, + "grad_norm": 1.7269983291625977, + "learning_rate": 1.0844735105806947e-05, + "loss": 0.1923, + "step": 36278 + }, + { + "epoch": 0.8504368923504281, + "grad_norm": 2.2777132987976074, + "learning_rate": 1.084139908905203e-05, + "loss": 1.4976, + "step": 36279 + }, + { + "epoch": 0.8504603339252331, + "grad_norm": 1.602271318435669, + "learning_rate": 1.0838063556074175e-05, + "loss": 0.0577, + "step": 36280 + }, + { + "epoch": 0.8504837755000381, + "grad_norm": 2.8256027698516846, + "learning_rate": 1.0834728506891512e-05, + "loss": 0.163, + "step": 36281 + }, + { + "epoch": 0.8505072170748431, + "grad_norm": 2.0406911373138428, + "learning_rate": 1.0831393941522117e-05, + "loss": 0.1055, + "step": 36282 + }, + { + "epoch": 0.850530658649648, + "grad_norm": 2.066783905029297, + "learning_rate": 1.0828059859984074e-05, + "loss": 0.0768, + "step": 36283 + }, + { + "epoch": 0.8505541002244531, + "grad_norm": 3.630335569381714, + "learning_rate": 1.0824726262295503e-05, + "loss": 0.3027, + "step": 36284 + }, + { + "epoch": 0.850577541799258, + "grad_norm": 3.3278253078460693, + "learning_rate": 1.0821393148474468e-05, + "loss": 0.2442, + "step": 36285 + }, + { + "epoch": 0.8506009833740631, + "grad_norm": 1.4374009370803833, + "learning_rate": 1.0818060518539064e-05, + "loss": 0.0645, + "step": 36286 + }, + { + "epoch": 0.850624424948868, + "grad_norm": 0.9129589796066284, + "learning_rate": 1.0814728372507355e-05, + "loss": 0.0925, + "step": 36287 + }, + { + "epoch": 0.8506478665236731, + "grad_norm": 2.0846943855285645, + "learning_rate": 1.081139671039746e-05, + "loss": 0.1562, + "step": 36288 + }, + { + "epoch": 0.850671308098478, + "grad_norm": 4.310152530670166, + "learning_rate": 1.080806553222743e-05, + "loss": 0.3623, + "step": 36289 + }, + { + "epoch": 0.8506947496732831, + "grad_norm": 2.3875391483306885, + "learning_rate": 1.080473483801534e-05, + "loss": 1.0018, + "step": 36290 + }, + { + "epoch": 0.850718191248088, + "grad_norm": 1.106520414352417, + "learning_rate": 1.0801404627779278e-05, + "loss": 0.1145, + "step": 36291 + }, + { + "epoch": 0.8507416328228931, + "grad_norm": 3.3354101181030273, + "learning_rate": 1.0798074901537291e-05, + "loss": 0.2626, + "step": 36292 + }, + { + "epoch": 0.850765074397698, + "grad_norm": 1.1231833696365356, + "learning_rate": 1.0794745659307437e-05, + "loss": 0.0475, + "step": 36293 + }, + { + "epoch": 0.850788515972503, + "grad_norm": 3.5716233253479004, + "learning_rate": 1.079141690110783e-05, + "loss": 0.1916, + "step": 36294 + }, + { + "epoch": 0.850811957547308, + "grad_norm": 4.50718879699707, + "learning_rate": 1.0788088626956482e-05, + "loss": 0.2584, + "step": 36295 + }, + { + "epoch": 0.850835399122113, + "grad_norm": 2.0931739807128906, + "learning_rate": 1.0784760836871489e-05, + "loss": 1.4642, + "step": 36296 + }, + { + "epoch": 0.850858840696918, + "grad_norm": 1.2334179878234863, + "learning_rate": 1.0781433530870888e-05, + "loss": 0.0829, + "step": 36297 + }, + { + "epoch": 0.850882282271723, + "grad_norm": 2.998974323272705, + "learning_rate": 1.0778106708972746e-05, + "loss": 0.2308, + "step": 36298 + }, + { + "epoch": 0.850905723846528, + "grad_norm": 0.9778029918670654, + "learning_rate": 1.0774780371195104e-05, + "loss": 0.0719, + "step": 36299 + }, + { + "epoch": 0.850929165421333, + "grad_norm": 3.5862061977386475, + "learning_rate": 1.077145451755599e-05, + "loss": 0.3197, + "step": 36300 + }, + { + "epoch": 0.850952606996138, + "grad_norm": 0.3990434408187866, + "learning_rate": 1.076812914807349e-05, + "loss": 0.0173, + "step": 36301 + }, + { + "epoch": 0.850976048570943, + "grad_norm": 1.328631043434143, + "learning_rate": 1.0764804262765638e-05, + "loss": 0.1327, + "step": 36302 + }, + { + "epoch": 0.8509994901457479, + "grad_norm": 1.4203417301177979, + "learning_rate": 1.076147986165047e-05, + "loss": 0.0542, + "step": 36303 + }, + { + "epoch": 0.851022931720553, + "grad_norm": 0.47461870312690735, + "learning_rate": 1.0758155944746007e-05, + "loss": 0.0341, + "step": 36304 + }, + { + "epoch": 0.8510463732953579, + "grad_norm": 2.5844473838806152, + "learning_rate": 1.075483251207029e-05, + "loss": 0.2453, + "step": 36305 + }, + { + "epoch": 0.851069814870163, + "grad_norm": 2.4906389713287354, + "learning_rate": 1.0751509563641381e-05, + "loss": 0.2724, + "step": 36306 + }, + { + "epoch": 0.851093256444968, + "grad_norm": 0.9413013458251953, + "learning_rate": 1.074818709947727e-05, + "loss": 0.1113, + "step": 36307 + }, + { + "epoch": 0.851116698019773, + "grad_norm": 1.8282897472381592, + "learning_rate": 1.0744865119596026e-05, + "loss": 0.0778, + "step": 36308 + }, + { + "epoch": 0.851140139594578, + "grad_norm": 0.9110792279243469, + "learning_rate": 1.0741543624015648e-05, + "loss": 0.1123, + "step": 36309 + }, + { + "epoch": 0.851163581169383, + "grad_norm": 3.1694650650024414, + "learning_rate": 1.0738222612754168e-05, + "loss": 0.1596, + "step": 36310 + }, + { + "epoch": 0.851187022744188, + "grad_norm": 3.150693416595459, + "learning_rate": 1.0734902085829601e-05, + "loss": 0.2275, + "step": 36311 + }, + { + "epoch": 0.8512104643189929, + "grad_norm": 3.3233933448791504, + "learning_rate": 1.0731582043259936e-05, + "loss": 0.236, + "step": 36312 + }, + { + "epoch": 0.851233905893798, + "grad_norm": 1.898604393005371, + "learning_rate": 1.0728262485063246e-05, + "loss": 0.1309, + "step": 36313 + }, + { + "epoch": 0.8512573474686029, + "grad_norm": 4.205657958984375, + "learning_rate": 1.0724943411257515e-05, + "loss": 0.2176, + "step": 36314 + }, + { + "epoch": 0.851280789043408, + "grad_norm": 2.3008787631988525, + "learning_rate": 1.0721624821860754e-05, + "loss": 0.2357, + "step": 36315 + }, + { + "epoch": 0.8513042306182129, + "grad_norm": 1.8722175359725952, + "learning_rate": 1.0718306716890958e-05, + "loss": 0.1422, + "step": 36316 + }, + { + "epoch": 0.851327672193018, + "grad_norm": 3.393432855606079, + "learning_rate": 1.0714989096366145e-05, + "loss": 0.2011, + "step": 36317 + }, + { + "epoch": 0.8513511137678229, + "grad_norm": 4.477087497711182, + "learning_rate": 1.0711671960304292e-05, + "loss": 0.262, + "step": 36318 + }, + { + "epoch": 0.851374555342628, + "grad_norm": 0.9407199621200562, + "learning_rate": 1.0708355308723417e-05, + "loss": 0.0387, + "step": 36319 + }, + { + "epoch": 0.8513979969174329, + "grad_norm": 3.2392334938049316, + "learning_rate": 1.0705039141641537e-05, + "loss": 0.2639, + "step": 36320 + }, + { + "epoch": 0.851421438492238, + "grad_norm": 1.6170356273651123, + "learning_rate": 1.0701723459076629e-05, + "loss": 0.0651, + "step": 36321 + }, + { + "epoch": 0.8514448800670429, + "grad_norm": 0.7768978476524353, + "learning_rate": 1.0698408261046677e-05, + "loss": 0.0447, + "step": 36322 + }, + { + "epoch": 0.8514683216418479, + "grad_norm": 1.8419111967086792, + "learning_rate": 1.0695093547569668e-05, + "loss": 0.1515, + "step": 36323 + }, + { + "epoch": 0.8514917632166529, + "grad_norm": 2.9256608486175537, + "learning_rate": 1.0691779318663597e-05, + "loss": 0.0802, + "step": 36324 + }, + { + "epoch": 0.8515152047914579, + "grad_norm": 3.7119810581207275, + "learning_rate": 1.068846557434643e-05, + "loss": 0.2336, + "step": 36325 + }, + { + "epoch": 0.8515386463662629, + "grad_norm": 2.4274299144744873, + "learning_rate": 1.0685152314636171e-05, + "loss": 0.401, + "step": 36326 + }, + { + "epoch": 0.8515620879410679, + "grad_norm": 1.6727185249328613, + "learning_rate": 1.0681839539550787e-05, + "loss": 0.7186, + "step": 36327 + }, + { + "epoch": 0.8515855295158729, + "grad_norm": 2.2751870155334473, + "learning_rate": 1.067852724910825e-05, + "loss": 0.1344, + "step": 36328 + }, + { + "epoch": 0.8516089710906779, + "grad_norm": 1.1001595258712769, + "learning_rate": 1.0675215443326536e-05, + "loss": 0.9682, + "step": 36329 + }, + { + "epoch": 0.8516324126654828, + "grad_norm": 1.783442735671997, + "learning_rate": 1.067190412222362e-05, + "loss": 1.0497, + "step": 36330 + }, + { + "epoch": 0.8516558542402879, + "grad_norm": 1.2367526292800903, + "learning_rate": 1.066859328581744e-05, + "loss": 0.1523, + "step": 36331 + }, + { + "epoch": 0.8516792958150928, + "grad_norm": 2.527438163757324, + "learning_rate": 1.0665282934125998e-05, + "loss": 1.441, + "step": 36332 + }, + { + "epoch": 0.8517027373898979, + "grad_norm": 0.9715538024902344, + "learning_rate": 1.0661973067167241e-05, + "loss": 0.0578, + "step": 36333 + }, + { + "epoch": 0.8517261789647028, + "grad_norm": 2.9329941272735596, + "learning_rate": 1.0658663684959146e-05, + "loss": 0.1582, + "step": 36334 + }, + { + "epoch": 0.8517496205395079, + "grad_norm": 2.3615832328796387, + "learning_rate": 1.0655354787519645e-05, + "loss": 0.1279, + "step": 36335 + }, + { + "epoch": 0.8517730621143128, + "grad_norm": 2.216639757156372, + "learning_rate": 1.065204637486671e-05, + "loss": 0.1374, + "step": 36336 + }, + { + "epoch": 0.8517965036891179, + "grad_norm": 2.412929058074951, + "learning_rate": 1.0648738447018259e-05, + "loss": 0.1096, + "step": 36337 + }, + { + "epoch": 0.8518199452639228, + "grad_norm": 2.8559036254882812, + "learning_rate": 1.064543100399228e-05, + "loss": 0.1115, + "step": 36338 + }, + { + "epoch": 0.8518433868387278, + "grad_norm": 2.9711220264434814, + "learning_rate": 1.0642124045806712e-05, + "loss": 1.7492, + "step": 36339 + }, + { + "epoch": 0.8518668284135328, + "grad_norm": 2.8484952449798584, + "learning_rate": 1.0638817572479486e-05, + "loss": 0.1611, + "step": 36340 + }, + { + "epoch": 0.8518902699883378, + "grad_norm": 1.7861922979354858, + "learning_rate": 1.0635511584028557e-05, + "loss": 0.1736, + "step": 36341 + }, + { + "epoch": 0.8519137115631428, + "grad_norm": 2.2709414958953857, + "learning_rate": 1.0632206080471851e-05, + "loss": 0.12, + "step": 36342 + }, + { + "epoch": 0.8519371531379478, + "grad_norm": 0.34430912137031555, + "learning_rate": 1.062890106182728e-05, + "loss": 0.0147, + "step": 36343 + }, + { + "epoch": 0.8519605947127528, + "grad_norm": 1.891726016998291, + "learning_rate": 1.0625596528112813e-05, + "loss": 1.4839, + "step": 36344 + }, + { + "epoch": 0.8519840362875578, + "grad_norm": 1.3304388523101807, + "learning_rate": 1.0622292479346385e-05, + "loss": 0.1165, + "step": 36345 + }, + { + "epoch": 0.8520074778623627, + "grad_norm": 1.4880790710449219, + "learning_rate": 1.0618988915545924e-05, + "loss": 0.0825, + "step": 36346 + }, + { + "epoch": 0.8520309194371678, + "grad_norm": 2.0258798599243164, + "learning_rate": 1.0615685836729328e-05, + "loss": 1.1619, + "step": 36347 + }, + { + "epoch": 0.8520543610119727, + "grad_norm": 2.8344409465789795, + "learning_rate": 1.0612383242914547e-05, + "loss": 0.2703, + "step": 36348 + }, + { + "epoch": 0.8520778025867778, + "grad_norm": 2.8559088706970215, + "learning_rate": 1.0609081134119481e-05, + "loss": 0.295, + "step": 36349 + }, + { + "epoch": 0.8521012441615827, + "grad_norm": 0.7282575368881226, + "learning_rate": 1.0605779510362035e-05, + "loss": 0.0349, + "step": 36350 + }, + { + "epoch": 0.8521246857363878, + "grad_norm": 3.9625132083892822, + "learning_rate": 1.0602478371660163e-05, + "loss": 0.2855, + "step": 36351 + }, + { + "epoch": 0.8521481273111927, + "grad_norm": 0.7427232265472412, + "learning_rate": 1.059917771803175e-05, + "loss": 0.0631, + "step": 36352 + }, + { + "epoch": 0.8521715688859978, + "grad_norm": 2.5937411785125732, + "learning_rate": 1.0595877549494727e-05, + "loss": 0.1732, + "step": 36353 + }, + { + "epoch": 0.8521950104608027, + "grad_norm": 2.073568820953369, + "learning_rate": 1.0592577866066977e-05, + "loss": 0.2103, + "step": 36354 + }, + { + "epoch": 0.8522184520356078, + "grad_norm": 2.3997881412506104, + "learning_rate": 1.0589278667766412e-05, + "loss": 0.2093, + "step": 36355 + }, + { + "epoch": 0.8522418936104127, + "grad_norm": 0.9310762286186218, + "learning_rate": 1.0585979954610915e-05, + "loss": 0.0723, + "step": 36356 + }, + { + "epoch": 0.8522653351852177, + "grad_norm": 2.470912218093872, + "learning_rate": 1.0582681726618404e-05, + "loss": 0.1689, + "step": 36357 + }, + { + "epoch": 0.8522887767600228, + "grad_norm": 2.337399482727051, + "learning_rate": 1.05793839838068e-05, + "loss": 0.156, + "step": 36358 + }, + { + "epoch": 0.8523122183348277, + "grad_norm": 2.497636556625366, + "learning_rate": 1.0576086726193979e-05, + "loss": 0.0982, + "step": 36359 + }, + { + "epoch": 0.8523356599096328, + "grad_norm": 1.9214080572128296, + "learning_rate": 1.0572789953797813e-05, + "loss": 0.0939, + "step": 36360 + }, + { + "epoch": 0.8523591014844377, + "grad_norm": 2.326099157333374, + "learning_rate": 1.056949366663621e-05, + "loss": 0.1813, + "step": 36361 + }, + { + "epoch": 0.8523825430592428, + "grad_norm": 3.1383450031280518, + "learning_rate": 1.0566197864727046e-05, + "loss": 0.3246, + "step": 36362 + }, + { + "epoch": 0.8524059846340477, + "grad_norm": 3.5428545475006104, + "learning_rate": 1.0562902548088194e-05, + "loss": 0.1556, + "step": 36363 + }, + { + "epoch": 0.8524294262088528, + "grad_norm": 4.101941108703613, + "learning_rate": 1.0559607716737563e-05, + "loss": 0.4136, + "step": 36364 + }, + { + "epoch": 0.8524528677836577, + "grad_norm": 3.2177698612213135, + "learning_rate": 1.0556313370693017e-05, + "loss": 0.3021, + "step": 36365 + }, + { + "epoch": 0.8524763093584627, + "grad_norm": 0.562679648399353, + "learning_rate": 1.0553019509972428e-05, + "loss": 0.0461, + "step": 36366 + }, + { + "epoch": 0.8524997509332677, + "grad_norm": 2.022397518157959, + "learning_rate": 1.0549726134593663e-05, + "loss": 0.1252, + "step": 36367 + }, + { + "epoch": 0.8525231925080727, + "grad_norm": 2.412306547164917, + "learning_rate": 1.0546433244574583e-05, + "loss": 0.1698, + "step": 36368 + }, + { + "epoch": 0.8525466340828777, + "grad_norm": 2.2796244621276855, + "learning_rate": 1.0543140839933097e-05, + "loss": 0.1181, + "step": 36369 + }, + { + "epoch": 0.8525700756576827, + "grad_norm": 3.1593730449676514, + "learning_rate": 1.0539848920687024e-05, + "loss": 0.1215, + "step": 36370 + }, + { + "epoch": 0.8525935172324877, + "grad_norm": 1.4304606914520264, + "learning_rate": 1.053655748685426e-05, + "loss": 0.918, + "step": 36371 + }, + { + "epoch": 0.8526169588072927, + "grad_norm": 0.999811589717865, + "learning_rate": 1.0533266538452658e-05, + "loss": 0.1008, + "step": 36372 + }, + { + "epoch": 0.8526404003820977, + "grad_norm": 1.5285592079162598, + "learning_rate": 1.0529976075500058e-05, + "loss": 0.1412, + "step": 36373 + }, + { + "epoch": 0.8526638419569027, + "grad_norm": 3.4606895446777344, + "learning_rate": 1.0526686098014338e-05, + "loss": 0.2745, + "step": 36374 + }, + { + "epoch": 0.8526872835317076, + "grad_norm": 2.4982776641845703, + "learning_rate": 1.0523396606013302e-05, + "loss": 0.1565, + "step": 36375 + }, + { + "epoch": 0.8527107251065127, + "grad_norm": 2.65415358543396, + "learning_rate": 1.052010759951485e-05, + "loss": 0.1226, + "step": 36376 + }, + { + "epoch": 0.8527341666813176, + "grad_norm": 2.8238260746002197, + "learning_rate": 1.0516819078536821e-05, + "loss": 0.1004, + "step": 36377 + }, + { + "epoch": 0.8527576082561227, + "grad_norm": 0.819612443447113, + "learning_rate": 1.0513531043097047e-05, + "loss": 0.0738, + "step": 36378 + }, + { + "epoch": 0.8527810498309276, + "grad_norm": 2.579885244369507, + "learning_rate": 1.0510243493213356e-05, + "loss": 0.183, + "step": 36379 + }, + { + "epoch": 0.8528044914057327, + "grad_norm": 1.7344781160354614, + "learning_rate": 1.0506956428903592e-05, + "loss": 0.0936, + "step": 36380 + }, + { + "epoch": 0.8528279329805376, + "grad_norm": 0.1597987413406372, + "learning_rate": 1.0503669850185615e-05, + "loss": 0.0131, + "step": 36381 + }, + { + "epoch": 0.8528513745553427, + "grad_norm": 2.6824123859405518, + "learning_rate": 1.0500383757077225e-05, + "loss": 0.2228, + "step": 36382 + }, + { + "epoch": 0.8528748161301476, + "grad_norm": 1.6736186742782593, + "learning_rate": 1.0497098149596286e-05, + "loss": 0.1102, + "step": 36383 + }, + { + "epoch": 0.8528982577049526, + "grad_norm": 4.950158596038818, + "learning_rate": 1.0493813027760612e-05, + "loss": 0.3327, + "step": 36384 + }, + { + "epoch": 0.8529216992797576, + "grad_norm": 0.9521492719650269, + "learning_rate": 1.0490528391588028e-05, + "loss": 0.0829, + "step": 36385 + }, + { + "epoch": 0.8529451408545626, + "grad_norm": 2.5068764686584473, + "learning_rate": 1.048724424109635e-05, + "loss": 0.2337, + "step": 36386 + }, + { + "epoch": 0.8529685824293676, + "grad_norm": 1.2964988946914673, + "learning_rate": 1.0483960576303398e-05, + "loss": 0.9434, + "step": 36387 + }, + { + "epoch": 0.8529920240041726, + "grad_norm": 2.3016536235809326, + "learning_rate": 1.048067739722698e-05, + "loss": 1.5558, + "step": 36388 + }, + { + "epoch": 0.8530154655789776, + "grad_norm": 2.2511043548583984, + "learning_rate": 1.0477394703884936e-05, + "loss": 0.1165, + "step": 36389 + }, + { + "epoch": 0.8530389071537826, + "grad_norm": 5.504744052886963, + "learning_rate": 1.0474112496295064e-05, + "loss": 0.5128, + "step": 36390 + }, + { + "epoch": 0.8530623487285875, + "grad_norm": 3.0494544506073, + "learning_rate": 1.0470830774475182e-05, + "loss": 0.1757, + "step": 36391 + }, + { + "epoch": 0.8530857903033926, + "grad_norm": 4.23091983795166, + "learning_rate": 1.0467549538443066e-05, + "loss": 0.2658, + "step": 36392 + }, + { + "epoch": 0.8531092318781975, + "grad_norm": 3.5995993614196777, + "learning_rate": 1.0464268788216568e-05, + "loss": 0.3019, + "step": 36393 + }, + { + "epoch": 0.8531326734530026, + "grad_norm": 0.954186737537384, + "learning_rate": 1.0460988523813464e-05, + "loss": 0.059, + "step": 36394 + }, + { + "epoch": 0.8531561150278075, + "grad_norm": 1.451997995376587, + "learning_rate": 1.0457708745251527e-05, + "loss": 0.0407, + "step": 36395 + }, + { + "epoch": 0.8531795566026126, + "grad_norm": 0.9369937777519226, + "learning_rate": 1.0454429452548608e-05, + "loss": 0.0307, + "step": 36396 + }, + { + "epoch": 0.8532029981774175, + "grad_norm": 1.672222375869751, + "learning_rate": 1.0451150645722463e-05, + "loss": 0.1001, + "step": 36397 + }, + { + "epoch": 0.8532264397522226, + "grad_norm": 1.6901488304138184, + "learning_rate": 1.0447872324790898e-05, + "loss": 0.1519, + "step": 36398 + }, + { + "epoch": 0.8532498813270275, + "grad_norm": 0.4873538613319397, + "learning_rate": 1.0444594489771698e-05, + "loss": 0.0312, + "step": 36399 + }, + { + "epoch": 0.8532733229018326, + "grad_norm": 0.6522741913795471, + "learning_rate": 1.044131714068265e-05, + "loss": 0.0364, + "step": 36400 + }, + { + "epoch": 0.8532967644766375, + "grad_norm": 2.021674633026123, + "learning_rate": 1.0438040277541506e-05, + "loss": 0.0874, + "step": 36401 + }, + { + "epoch": 0.8533202060514425, + "grad_norm": 3.937389373779297, + "learning_rate": 1.0434763900366096e-05, + "loss": 0.3033, + "step": 36402 + }, + { + "epoch": 0.8533436476262475, + "grad_norm": 2.213498830795288, + "learning_rate": 1.0431488009174174e-05, + "loss": 0.2038, + "step": 36403 + }, + { + "epoch": 0.8533670892010525, + "grad_norm": 1.4596649408340454, + "learning_rate": 1.04282126039835e-05, + "loss": 0.0973, + "step": 36404 + }, + { + "epoch": 0.8533905307758575, + "grad_norm": 2.536381721496582, + "learning_rate": 1.0424937684811876e-05, + "loss": 0.2618, + "step": 36405 + }, + { + "epoch": 0.8534139723506625, + "grad_norm": 4.319916248321533, + "learning_rate": 1.0421663251677071e-05, + "loss": 0.4003, + "step": 36406 + }, + { + "epoch": 0.8534374139254675, + "grad_norm": 3.1108558177948, + "learning_rate": 1.0418389304596832e-05, + "loss": 0.4349, + "step": 36407 + }, + { + "epoch": 0.8534608555002725, + "grad_norm": 2.338656425476074, + "learning_rate": 1.0415115843588907e-05, + "loss": 0.2511, + "step": 36408 + }, + { + "epoch": 0.8534842970750776, + "grad_norm": 2.0960803031921387, + "learning_rate": 1.0411842868671106e-05, + "loss": 1.3081, + "step": 36409 + }, + { + "epoch": 0.8535077386498825, + "grad_norm": 2.243065118789673, + "learning_rate": 1.0408570379861172e-05, + "loss": 1.0728, + "step": 36410 + }, + { + "epoch": 0.8535311802246875, + "grad_norm": 0.5164622664451599, + "learning_rate": 1.0405298377176842e-05, + "loss": 0.0607, + "step": 36411 + }, + { + "epoch": 0.8535546217994925, + "grad_norm": 1.630899429321289, + "learning_rate": 1.0402026860635894e-05, + "loss": 0.9962, + "step": 36412 + }, + { + "epoch": 0.8535780633742975, + "grad_norm": 2.2206037044525146, + "learning_rate": 1.0398755830256035e-05, + "loss": 0.1168, + "step": 36413 + }, + { + "epoch": 0.8536015049491025, + "grad_norm": 1.2876986265182495, + "learning_rate": 1.0395485286055074e-05, + "loss": 1.0194, + "step": 36414 + }, + { + "epoch": 0.8536249465239075, + "grad_norm": 1.4322198629379272, + "learning_rate": 1.0392215228050727e-05, + "loss": 0.1057, + "step": 36415 + }, + { + "epoch": 0.8536483880987125, + "grad_norm": 1.0339314937591553, + "learning_rate": 1.038894565626074e-05, + "loss": 0.0769, + "step": 36416 + }, + { + "epoch": 0.8536718296735175, + "grad_norm": 1.0963053703308105, + "learning_rate": 1.038567657070284e-05, + "loss": 0.0728, + "step": 36417 + }, + { + "epoch": 0.8536952712483225, + "grad_norm": 1.1328973770141602, + "learning_rate": 1.0382407971394792e-05, + "loss": 0.0857, + "step": 36418 + }, + { + "epoch": 0.8537187128231275, + "grad_norm": 1.4935041666030884, + "learning_rate": 1.0379139858354326e-05, + "loss": 0.0802, + "step": 36419 + }, + { + "epoch": 0.8537421543979324, + "grad_norm": 2.593726396560669, + "learning_rate": 1.0375872231599137e-05, + "loss": 0.1766, + "step": 36420 + }, + { + "epoch": 0.8537655959727375, + "grad_norm": 2.8264524936676025, + "learning_rate": 1.0372605091147003e-05, + "loss": 0.1377, + "step": 36421 + }, + { + "epoch": 0.8537890375475424, + "grad_norm": 3.0176146030426025, + "learning_rate": 1.0369338437015641e-05, + "loss": 1.697, + "step": 36422 + }, + { + "epoch": 0.8538124791223475, + "grad_norm": 1.9671307802200317, + "learning_rate": 1.036607226922277e-05, + "loss": 0.0773, + "step": 36423 + }, + { + "epoch": 0.8538359206971524, + "grad_norm": 2.5367729663848877, + "learning_rate": 1.0362806587786111e-05, + "loss": 0.127, + "step": 36424 + }, + { + "epoch": 0.8538593622719575, + "grad_norm": 1.1972423791885376, + "learning_rate": 1.0359541392723382e-05, + "loss": 0.0761, + "step": 36425 + }, + { + "epoch": 0.8538828038467624, + "grad_norm": 2.752793073654175, + "learning_rate": 1.0356276684052279e-05, + "loss": 1.453, + "step": 36426 + }, + { + "epoch": 0.8539062454215675, + "grad_norm": 4.8074564933776855, + "learning_rate": 1.0353012461790557e-05, + "loss": 0.5639, + "step": 36427 + }, + { + "epoch": 0.8539296869963724, + "grad_norm": 2.409990072250366, + "learning_rate": 1.0349748725955921e-05, + "loss": 1.0139, + "step": 36428 + }, + { + "epoch": 0.8539531285711774, + "grad_norm": 4.191486835479736, + "learning_rate": 1.0346485476566038e-05, + "loss": 0.2789, + "step": 36429 + }, + { + "epoch": 0.8539765701459824, + "grad_norm": 5.146047592163086, + "learning_rate": 1.0343222713638668e-05, + "loss": 0.1859, + "step": 36430 + }, + { + "epoch": 0.8540000117207874, + "grad_norm": 0.6955975294113159, + "learning_rate": 1.0339960437191499e-05, + "loss": 0.0533, + "step": 36431 + }, + { + "epoch": 0.8540234532955924, + "grad_norm": 1.6077731847763062, + "learning_rate": 1.0336698647242227e-05, + "loss": 0.0905, + "step": 36432 + }, + { + "epoch": 0.8540468948703974, + "grad_norm": 3.7337453365325928, + "learning_rate": 1.0333437343808516e-05, + "loss": 0.2867, + "step": 36433 + }, + { + "epoch": 0.8540703364452024, + "grad_norm": 2.753847360610962, + "learning_rate": 1.0330176526908132e-05, + "loss": 0.2752, + "step": 36434 + }, + { + "epoch": 0.8540937780200074, + "grad_norm": 2.5481767654418945, + "learning_rate": 1.0326916196558711e-05, + "loss": 0.2214, + "step": 36435 + }, + { + "epoch": 0.8541172195948123, + "grad_norm": 0.5827184319496155, + "learning_rate": 1.0323656352777978e-05, + "loss": 0.0203, + "step": 36436 + }, + { + "epoch": 0.8541406611696174, + "grad_norm": 1.9836151599884033, + "learning_rate": 1.0320396995583603e-05, + "loss": 1.6471, + "step": 36437 + }, + { + "epoch": 0.8541641027444223, + "grad_norm": 2.4986579418182373, + "learning_rate": 1.0317138124993286e-05, + "loss": 1.4557, + "step": 36438 + }, + { + "epoch": 0.8541875443192274, + "grad_norm": 2.791468620300293, + "learning_rate": 1.0313879741024669e-05, + "loss": 0.1587, + "step": 36439 + }, + { + "epoch": 0.8542109858940323, + "grad_norm": 0.24154610931873322, + "learning_rate": 1.0310621843695478e-05, + "loss": 0.0182, + "step": 36440 + }, + { + "epoch": 0.8542344274688374, + "grad_norm": 2.6350340843200684, + "learning_rate": 1.0307364433023358e-05, + "loss": 0.1212, + "step": 36441 + }, + { + "epoch": 0.8542578690436423, + "grad_norm": 3.7894880771636963, + "learning_rate": 1.0304107509026028e-05, + "loss": 0.2883, + "step": 36442 + }, + { + "epoch": 0.8542813106184474, + "grad_norm": 0.8703776597976685, + "learning_rate": 1.030085107172113e-05, + "loss": 0.0825, + "step": 36443 + }, + { + "epoch": 0.8543047521932523, + "grad_norm": 2.4688146114349365, + "learning_rate": 1.0297595121126336e-05, + "loss": 0.1754, + "step": 36444 + }, + { + "epoch": 0.8543281937680574, + "grad_norm": 0.4054601490497589, + "learning_rate": 1.029433965725931e-05, + "loss": 0.0505, + "step": 36445 + }, + { + "epoch": 0.8543516353428623, + "grad_norm": 2.3242850303649902, + "learning_rate": 1.0291084680137697e-05, + "loss": 0.2001, + "step": 36446 + }, + { + "epoch": 0.8543750769176673, + "grad_norm": 1.7132999897003174, + "learning_rate": 1.0287830189779201e-05, + "loss": 0.0914, + "step": 36447 + }, + { + "epoch": 0.8543985184924723, + "grad_norm": 2.8526997566223145, + "learning_rate": 1.0284576186201466e-05, + "loss": 1.5079, + "step": 36448 + }, + { + "epoch": 0.8544219600672773, + "grad_norm": 2.2891194820404053, + "learning_rate": 1.0281322669422144e-05, + "loss": 0.1592, + "step": 36449 + }, + { + "epoch": 0.8544454016420823, + "grad_norm": 3.795407772064209, + "learning_rate": 1.0278069639458876e-05, + "loss": 0.3023, + "step": 36450 + }, + { + "epoch": 0.8544688432168873, + "grad_norm": 1.8845007419586182, + "learning_rate": 1.0274817096329315e-05, + "loss": 0.1826, + "step": 36451 + }, + { + "epoch": 0.8544922847916923, + "grad_norm": 1.3761495351791382, + "learning_rate": 1.0271565040051146e-05, + "loss": 0.172, + "step": 36452 + }, + { + "epoch": 0.8545157263664973, + "grad_norm": 3.742136001586914, + "learning_rate": 1.0268313470641955e-05, + "loss": 0.2727, + "step": 36453 + }, + { + "epoch": 0.8545391679413022, + "grad_norm": 1.2596088647842407, + "learning_rate": 1.026506238811945e-05, + "loss": 0.1585, + "step": 36454 + }, + { + "epoch": 0.8545626095161073, + "grad_norm": 3.004356861114502, + "learning_rate": 1.0261811792501231e-05, + "loss": 0.1297, + "step": 36455 + }, + { + "epoch": 0.8545860510909122, + "grad_norm": 2.409592628479004, + "learning_rate": 1.0258561683804945e-05, + "loss": 0.2317, + "step": 36456 + }, + { + "epoch": 0.8546094926657173, + "grad_norm": 3.2154266834259033, + "learning_rate": 1.0255312062048227e-05, + "loss": 1.5962, + "step": 36457 + }, + { + "epoch": 0.8546329342405222, + "grad_norm": 0.8673989772796631, + "learning_rate": 1.0252062927248684e-05, + "loss": 0.083, + "step": 36458 + }, + { + "epoch": 0.8546563758153273, + "grad_norm": 3.036057472229004, + "learning_rate": 1.0248814279424002e-05, + "loss": 0.2364, + "step": 36459 + }, + { + "epoch": 0.8546798173901323, + "grad_norm": 2.9361467361450195, + "learning_rate": 1.0245566118591776e-05, + "loss": 0.1655, + "step": 36460 + }, + { + "epoch": 0.8547032589649373, + "grad_norm": 1.5455936193466187, + "learning_rate": 1.0242318444769627e-05, + "loss": 0.1006, + "step": 36461 + }, + { + "epoch": 0.8547267005397423, + "grad_norm": 0.7861774563789368, + "learning_rate": 1.0239071257975186e-05, + "loss": 0.0846, + "step": 36462 + }, + { + "epoch": 0.8547501421145473, + "grad_norm": 2.808697462081909, + "learning_rate": 1.023582455822606e-05, + "loss": 0.3002, + "step": 36463 + }, + { + "epoch": 0.8547735836893523, + "grad_norm": 2.9910202026367188, + "learning_rate": 1.0232578345539856e-05, + "loss": 0.2933, + "step": 36464 + }, + { + "epoch": 0.8547970252641572, + "grad_norm": 1.4498268365859985, + "learning_rate": 1.0229332619934218e-05, + "loss": 0.1223, + "step": 36465 + }, + { + "epoch": 0.8548204668389623, + "grad_norm": 0.8257948160171509, + "learning_rate": 1.0226087381426752e-05, + "loss": 0.0673, + "step": 36466 + }, + { + "epoch": 0.8548439084137672, + "grad_norm": 2.5199472904205322, + "learning_rate": 1.0222842630035068e-05, + "loss": 0.3148, + "step": 36467 + }, + { + "epoch": 0.8548673499885723, + "grad_norm": 2.0553674697875977, + "learning_rate": 1.021959836577676e-05, + "loss": 0.159, + "step": 36468 + }, + { + "epoch": 0.8548907915633772, + "grad_norm": 1.3847168684005737, + "learning_rate": 1.021635458866943e-05, + "loss": 0.0539, + "step": 36469 + }, + { + "epoch": 0.8549142331381823, + "grad_norm": 1.796660304069519, + "learning_rate": 1.0213111298730693e-05, + "loss": 0.1122, + "step": 36470 + }, + { + "epoch": 0.8549376747129872, + "grad_norm": 1.4998555183410645, + "learning_rate": 1.0209868495978114e-05, + "loss": 0.1155, + "step": 36471 + }, + { + "epoch": 0.8549611162877923, + "grad_norm": 2.4464526176452637, + "learning_rate": 1.0206626180429347e-05, + "loss": 0.1454, + "step": 36472 + }, + { + "epoch": 0.8549845578625972, + "grad_norm": 1.2787375450134277, + "learning_rate": 1.020338435210193e-05, + "loss": 0.0865, + "step": 36473 + }, + { + "epoch": 0.8550079994374022, + "grad_norm": 3.9935667514801025, + "learning_rate": 1.0200143011013485e-05, + "loss": 0.3508, + "step": 36474 + }, + { + "epoch": 0.8550314410122072, + "grad_norm": 3.287508487701416, + "learning_rate": 1.0196902157181588e-05, + "loss": 0.1397, + "step": 36475 + }, + { + "epoch": 0.8550548825870122, + "grad_norm": 3.0755715370178223, + "learning_rate": 1.0193661790623831e-05, + "loss": 1.2786, + "step": 36476 + }, + { + "epoch": 0.8550783241618172, + "grad_norm": 3.3009843826293945, + "learning_rate": 1.0190421911357762e-05, + "loss": 0.2433, + "step": 36477 + }, + { + "epoch": 0.8551017657366222, + "grad_norm": 3.9271938800811768, + "learning_rate": 1.0187182519400995e-05, + "loss": 0.2749, + "step": 36478 + }, + { + "epoch": 0.8551252073114272, + "grad_norm": 3.180917978286743, + "learning_rate": 1.018394361477112e-05, + "loss": 0.2012, + "step": 36479 + }, + { + "epoch": 0.8551486488862322, + "grad_norm": 0.925900936126709, + "learning_rate": 1.0180705197485685e-05, + "loss": 0.0712, + "step": 36480 + }, + { + "epoch": 0.8551720904610371, + "grad_norm": 0.43656212091445923, + "learning_rate": 1.017746726756228e-05, + "loss": 0.0225, + "step": 36481 + }, + { + "epoch": 0.8551955320358422, + "grad_norm": 1.388657808303833, + "learning_rate": 1.0174229825018455e-05, + "loss": 0.052, + "step": 36482 + }, + { + "epoch": 0.8552189736106471, + "grad_norm": 1.0382404327392578, + "learning_rate": 1.0170992869871787e-05, + "loss": 0.0896, + "step": 36483 + }, + { + "epoch": 0.8552424151854522, + "grad_norm": 3.782804250717163, + "learning_rate": 1.0167756402139816e-05, + "loss": 0.3825, + "step": 36484 + }, + { + "epoch": 0.8552658567602571, + "grad_norm": 1.392983078956604, + "learning_rate": 1.016452042184014e-05, + "loss": 0.1352, + "step": 36485 + }, + { + "epoch": 0.8552892983350622, + "grad_norm": 0.5550116300582886, + "learning_rate": 1.0161284928990312e-05, + "loss": 0.0624, + "step": 36486 + }, + { + "epoch": 0.8553127399098671, + "grad_norm": 2.499997615814209, + "learning_rate": 1.015804992360787e-05, + "loss": 0.2633, + "step": 36487 + }, + { + "epoch": 0.8553361814846722, + "grad_norm": 0.7469810247421265, + "learning_rate": 1.0154815405710383e-05, + "loss": 0.059, + "step": 36488 + }, + { + "epoch": 0.8553596230594771, + "grad_norm": 2.49249267578125, + "learning_rate": 1.0151581375315367e-05, + "loss": 0.2552, + "step": 36489 + }, + { + "epoch": 0.8553830646342822, + "grad_norm": 0.4108889698982239, + "learning_rate": 1.0148347832440419e-05, + "loss": 0.0125, + "step": 36490 + }, + { + "epoch": 0.8554065062090871, + "grad_norm": 1.1914721727371216, + "learning_rate": 1.014511477710305e-05, + "loss": 0.1454, + "step": 36491 + }, + { + "epoch": 0.8554299477838921, + "grad_norm": 0.7149349451065063, + "learning_rate": 1.014188220932083e-05, + "loss": 0.0571, + "step": 36492 + }, + { + "epoch": 0.8554533893586971, + "grad_norm": 4.672049522399902, + "learning_rate": 1.0138650129111272e-05, + "loss": 0.2832, + "step": 36493 + }, + { + "epoch": 0.8554768309335021, + "grad_norm": 2.347829818725586, + "learning_rate": 1.0135418536491936e-05, + "loss": 0.1096, + "step": 36494 + }, + { + "epoch": 0.8555002725083071, + "grad_norm": 3.6387240886688232, + "learning_rate": 1.0132187431480344e-05, + "loss": 0.3188, + "step": 36495 + }, + { + "epoch": 0.8555237140831121, + "grad_norm": 2.365370750427246, + "learning_rate": 1.0128956814094015e-05, + "loss": 1.635, + "step": 36496 + }, + { + "epoch": 0.8555471556579171, + "grad_norm": 0.872336208820343, + "learning_rate": 1.01257266843505e-05, + "loss": 0.1183, + "step": 36497 + }, + { + "epoch": 0.8555705972327221, + "grad_norm": 3.076160192489624, + "learning_rate": 1.0122497042267332e-05, + "loss": 0.2805, + "step": 36498 + }, + { + "epoch": 0.855594038807527, + "grad_norm": 0.6520993709564209, + "learning_rate": 1.0119267887862017e-05, + "loss": 0.0352, + "step": 36499 + }, + { + "epoch": 0.8556174803823321, + "grad_norm": 1.1025785207748413, + "learning_rate": 1.0116039221152086e-05, + "loss": 0.054, + "step": 36500 + }, + { + "epoch": 0.855640921957137, + "grad_norm": 0.8395069241523743, + "learning_rate": 1.0112811042155045e-05, + "loss": 0.1054, + "step": 36501 + }, + { + "epoch": 0.8556643635319421, + "grad_norm": 0.7786880731582642, + "learning_rate": 1.0109583350888407e-05, + "loss": 0.0395, + "step": 36502 + }, + { + "epoch": 0.855687805106747, + "grad_norm": 3.440589427947998, + "learning_rate": 1.0106356147369689e-05, + "loss": 0.2815, + "step": 36503 + }, + { + "epoch": 0.8557112466815521, + "grad_norm": 1.1601141691207886, + "learning_rate": 1.0103129431616442e-05, + "loss": 0.0673, + "step": 36504 + }, + { + "epoch": 0.855734688256357, + "grad_norm": 1.3376089334487915, + "learning_rate": 1.0099903203646132e-05, + "loss": 0.0695, + "step": 36505 + }, + { + "epoch": 0.8557581298311621, + "grad_norm": 0.1943497359752655, + "learning_rate": 1.0096677463476278e-05, + "loss": 0.0116, + "step": 36506 + }, + { + "epoch": 0.855781571405967, + "grad_norm": 3.3489062786102295, + "learning_rate": 1.0093452211124376e-05, + "loss": 0.1933, + "step": 36507 + }, + { + "epoch": 0.855805012980772, + "grad_norm": 2.257930040359497, + "learning_rate": 1.0090227446607936e-05, + "loss": 0.1495, + "step": 36508 + }, + { + "epoch": 0.855828454555577, + "grad_norm": 2.2191498279571533, + "learning_rate": 1.0087003169944432e-05, + "loss": 0.1112, + "step": 36509 + }, + { + "epoch": 0.855851896130382, + "grad_norm": 3.0571064949035645, + "learning_rate": 1.0083779381151393e-05, + "loss": 0.2137, + "step": 36510 + }, + { + "epoch": 0.8558753377051871, + "grad_norm": 2.920488119125366, + "learning_rate": 1.0080556080246295e-05, + "loss": 0.159, + "step": 36511 + }, + { + "epoch": 0.855898779279992, + "grad_norm": 2.1316518783569336, + "learning_rate": 1.0077333267246624e-05, + "loss": 0.07, + "step": 36512 + }, + { + "epoch": 0.8559222208547971, + "grad_norm": 2.7854678630828857, + "learning_rate": 1.0074110942169878e-05, + "loss": 0.2774, + "step": 36513 + }, + { + "epoch": 0.855945662429602, + "grad_norm": 0.2608129680156708, + "learning_rate": 1.007088910503351e-05, + "loss": 0.0229, + "step": 36514 + }, + { + "epoch": 0.8559691040044071, + "grad_norm": 0.41251009702682495, + "learning_rate": 1.0067667755855048e-05, + "loss": 0.0138, + "step": 36515 + }, + { + "epoch": 0.855992545579212, + "grad_norm": 0.7527337074279785, + "learning_rate": 1.0064446894651936e-05, + "loss": 0.0603, + "step": 36516 + }, + { + "epoch": 0.856015987154017, + "grad_norm": 3.5139923095703125, + "learning_rate": 1.006122652144168e-05, + "loss": 0.1884, + "step": 36517 + }, + { + "epoch": 0.856039428728822, + "grad_norm": 1.9858442544937134, + "learning_rate": 1.0058006636241735e-05, + "loss": 0.1944, + "step": 36518 + }, + { + "epoch": 0.856062870303627, + "grad_norm": 2.8762834072113037, + "learning_rate": 1.0054787239069574e-05, + "loss": 0.2237, + "step": 36519 + }, + { + "epoch": 0.856086311878432, + "grad_norm": 1.4910180568695068, + "learning_rate": 1.0051568329942662e-05, + "loss": 0.1287, + "step": 36520 + }, + { + "epoch": 0.856109753453237, + "grad_norm": 3.287205219268799, + "learning_rate": 1.0048349908878484e-05, + "loss": 0.29, + "step": 36521 + }, + { + "epoch": 0.856133195028042, + "grad_norm": 3.7037932872772217, + "learning_rate": 1.0045131975894451e-05, + "loss": 0.5806, + "step": 36522 + }, + { + "epoch": 0.856156636602847, + "grad_norm": 3.168921709060669, + "learning_rate": 1.004191453100809e-05, + "loss": 0.1712, + "step": 36523 + }, + { + "epoch": 0.856180078177652, + "grad_norm": 0.5294380187988281, + "learning_rate": 1.0038697574236833e-05, + "loss": 0.0661, + "step": 36524 + }, + { + "epoch": 0.856203519752457, + "grad_norm": 0.7282454967498779, + "learning_rate": 1.0035481105598133e-05, + "loss": 0.0257, + "step": 36525 + }, + { + "epoch": 0.856226961327262, + "grad_norm": 3.066591262817383, + "learning_rate": 1.0032265125109419e-05, + "loss": 0.3476, + "step": 36526 + }, + { + "epoch": 0.856250402902067, + "grad_norm": 1.091138482093811, + "learning_rate": 1.0029049632788178e-05, + "loss": 0.1175, + "step": 36527 + }, + { + "epoch": 0.8562738444768719, + "grad_norm": 1.8505479097366333, + "learning_rate": 1.0025834628651854e-05, + "loss": 0.961, + "step": 36528 + }, + { + "epoch": 0.856297286051677, + "grad_norm": 3.562150716781616, + "learning_rate": 1.0022620112717862e-05, + "loss": 0.3915, + "step": 36529 + }, + { + "epoch": 0.8563207276264819, + "grad_norm": 0.8660587072372437, + "learning_rate": 1.0019406085003669e-05, + "loss": 0.0597, + "step": 36530 + }, + { + "epoch": 0.856344169201287, + "grad_norm": 1.4053962230682373, + "learning_rate": 1.0016192545526725e-05, + "loss": 0.0839, + "step": 36531 + }, + { + "epoch": 0.8563676107760919, + "grad_norm": 2.4639971256256104, + "learning_rate": 1.0012979494304441e-05, + "loss": 0.2923, + "step": 36532 + }, + { + "epoch": 0.856391052350897, + "grad_norm": 3.5772786140441895, + "learning_rate": 1.0009766931354258e-05, + "loss": 0.1626, + "step": 36533 + }, + { + "epoch": 0.8564144939257019, + "grad_norm": 1.744754433631897, + "learning_rate": 1.0006554856693594e-05, + "loss": 0.113, + "step": 36534 + }, + { + "epoch": 0.856437935500507, + "grad_norm": 2.318877935409546, + "learning_rate": 1.0003343270339915e-05, + "loss": 1.3675, + "step": 36535 + }, + { + "epoch": 0.8564613770753119, + "grad_norm": 5.129202842712402, + "learning_rate": 1.0000132172310629e-05, + "loss": 0.3006, + "step": 36536 + }, + { + "epoch": 0.8564848186501169, + "grad_norm": 2.5543220043182373, + "learning_rate": 9.996921562623151e-06, + "loss": 0.0804, + "step": 36537 + }, + { + "epoch": 0.8565082602249219, + "grad_norm": 4.904815673828125, + "learning_rate": 9.993711441294884e-06, + "loss": 0.6293, + "step": 36538 + }, + { + "epoch": 0.8565317017997269, + "grad_norm": 4.620468616485596, + "learning_rate": 9.9905018083433e-06, + "loss": 0.4735, + "step": 36539 + }, + { + "epoch": 0.8565551433745319, + "grad_norm": 2.3342669010162354, + "learning_rate": 9.987292663785775e-06, + "loss": 0.1183, + "step": 36540 + }, + { + "epoch": 0.8565785849493369, + "grad_norm": 2.9923267364501953, + "learning_rate": 9.984084007639716e-06, + "loss": 0.2356, + "step": 36541 + }, + { + "epoch": 0.8566020265241419, + "grad_norm": 1.4113284349441528, + "learning_rate": 9.980875839922566e-06, + "loss": 0.1549, + "step": 36542 + }, + { + "epoch": 0.8566254680989469, + "grad_norm": 3.2626802921295166, + "learning_rate": 9.97766816065171e-06, + "loss": 0.2986, + "step": 36543 + }, + { + "epoch": 0.8566489096737518, + "grad_norm": 1.0300811529159546, + "learning_rate": 9.974460969844567e-06, + "loss": 0.0863, + "step": 36544 + }, + { + "epoch": 0.8566723512485569, + "grad_norm": 2.081082582473755, + "learning_rate": 9.971254267518526e-06, + "loss": 0.1233, + "step": 36545 + }, + { + "epoch": 0.8566957928233618, + "grad_norm": 2.1223058700561523, + "learning_rate": 9.968048053690992e-06, + "loss": 0.0769, + "step": 36546 + }, + { + "epoch": 0.8567192343981669, + "grad_norm": 2.8000833988189697, + "learning_rate": 9.964842328379332e-06, + "loss": 0.2845, + "step": 36547 + }, + { + "epoch": 0.8567426759729718, + "grad_norm": 1.8907139301300049, + "learning_rate": 9.961637091600995e-06, + "loss": 0.1839, + "step": 36548 + }, + { + "epoch": 0.8567661175477769, + "grad_norm": 2.1191351413726807, + "learning_rate": 9.958432343373359e-06, + "loss": 0.8326, + "step": 36549 + }, + { + "epoch": 0.8567895591225818, + "grad_norm": 1.7957673072814941, + "learning_rate": 9.955228083713763e-06, + "loss": 0.1692, + "step": 36550 + }, + { + "epoch": 0.8568130006973869, + "grad_norm": 2.5183138847351074, + "learning_rate": 9.952024312639662e-06, + "loss": 0.297, + "step": 36551 + }, + { + "epoch": 0.8568364422721918, + "grad_norm": 2.95158052444458, + "learning_rate": 9.948821030168398e-06, + "loss": 0.2609, + "step": 36552 + }, + { + "epoch": 0.8568598838469969, + "grad_norm": 1.3487942218780518, + "learning_rate": 9.945618236317377e-06, + "loss": 0.0802, + "step": 36553 + }, + { + "epoch": 0.8568833254218018, + "grad_norm": 3.131556749343872, + "learning_rate": 9.942415931103933e-06, + "loss": 0.2191, + "step": 36554 + }, + { + "epoch": 0.8569067669966068, + "grad_norm": 3.347092628479004, + "learning_rate": 9.939214114545503e-06, + "loss": 1.212, + "step": 36555 + }, + { + "epoch": 0.8569302085714118, + "grad_norm": 3.3992185592651367, + "learning_rate": 9.936012786659421e-06, + "loss": 0.2708, + "step": 36556 + }, + { + "epoch": 0.8569536501462168, + "grad_norm": 2.823035955429077, + "learning_rate": 9.932811947463061e-06, + "loss": 0.2231, + "step": 36557 + }, + { + "epoch": 0.8569770917210218, + "grad_norm": 2.0760014057159424, + "learning_rate": 9.929611596973809e-06, + "loss": 0.1134, + "step": 36558 + }, + { + "epoch": 0.8570005332958268, + "grad_norm": 1.5723896026611328, + "learning_rate": 9.926411735209006e-06, + "loss": 0.0573, + "step": 36559 + }, + { + "epoch": 0.8570239748706318, + "grad_norm": 2.8375043869018555, + "learning_rate": 9.923212362186007e-06, + "loss": 0.1945, + "step": 36560 + }, + { + "epoch": 0.8570474164454368, + "grad_norm": 2.3985273838043213, + "learning_rate": 9.920013477922219e-06, + "loss": 0.2497, + "step": 36561 + }, + { + "epoch": 0.8570708580202419, + "grad_norm": 0.3456117510795593, + "learning_rate": 9.91681508243495e-06, + "loss": 0.0228, + "step": 36562 + }, + { + "epoch": 0.8570942995950468, + "grad_norm": 1.5335171222686768, + "learning_rate": 9.913617175741585e-06, + "loss": 0.9786, + "step": 36563 + }, + { + "epoch": 0.8571177411698518, + "grad_norm": 2.0688226222991943, + "learning_rate": 9.910419757859468e-06, + "loss": 0.1433, + "step": 36564 + }, + { + "epoch": 0.8571411827446568, + "grad_norm": 0.9291883707046509, + "learning_rate": 9.907222828805951e-06, + "loss": 0.0829, + "step": 36565 + }, + { + "epoch": 0.8571646243194618, + "grad_norm": 3.298189878463745, + "learning_rate": 9.904026388598375e-06, + "loss": 0.4837, + "step": 36566 + }, + { + "epoch": 0.8571880658942668, + "grad_norm": 3.589831829071045, + "learning_rate": 9.900830437254071e-06, + "loss": 0.2828, + "step": 36567 + }, + { + "epoch": 0.8572115074690718, + "grad_norm": 0.8864837884902954, + "learning_rate": 9.897634974790404e-06, + "loss": 0.4532, + "step": 36568 + }, + { + "epoch": 0.8572349490438768, + "grad_norm": 1.8273905515670776, + "learning_rate": 9.894440001224714e-06, + "loss": 0.08, + "step": 36569 + }, + { + "epoch": 0.8572583906186818, + "grad_norm": 1.9867063760757446, + "learning_rate": 9.891245516574321e-06, + "loss": 0.1383, + "step": 36570 + }, + { + "epoch": 0.8572818321934867, + "grad_norm": 2.237147092819214, + "learning_rate": 9.888051520856557e-06, + "loss": 1.2708, + "step": 36571 + }, + { + "epoch": 0.8573052737682918, + "grad_norm": 1.2120281457901, + "learning_rate": 9.884858014088749e-06, + "loss": 0.0929, + "step": 36572 + }, + { + "epoch": 0.8573287153430967, + "grad_norm": 2.404672622680664, + "learning_rate": 9.881664996288254e-06, + "loss": 0.1821, + "step": 36573 + }, + { + "epoch": 0.8573521569179018, + "grad_norm": 1.054373025894165, + "learning_rate": 9.878472467472366e-06, + "loss": 0.0937, + "step": 36574 + }, + { + "epoch": 0.8573755984927067, + "grad_norm": 0.8314735293388367, + "learning_rate": 9.875280427658418e-06, + "loss": 0.0419, + "step": 36575 + }, + { + "epoch": 0.8573990400675118, + "grad_norm": 1.7539067268371582, + "learning_rate": 9.872088876863739e-06, + "loss": 0.1954, + "step": 36576 + }, + { + "epoch": 0.8574224816423167, + "grad_norm": 2.178853988647461, + "learning_rate": 9.86889781510565e-06, + "loss": 0.232, + "step": 36577 + }, + { + "epoch": 0.8574459232171218, + "grad_norm": 3.7537503242492676, + "learning_rate": 9.865707242401445e-06, + "loss": 0.232, + "step": 36578 + }, + { + "epoch": 0.8574693647919267, + "grad_norm": 2.2600653171539307, + "learning_rate": 9.862517158768436e-06, + "loss": 0.1059, + "step": 36579 + }, + { + "epoch": 0.8574928063667318, + "grad_norm": 1.9629501104354858, + "learning_rate": 9.859327564223963e-06, + "loss": 0.07, + "step": 36580 + }, + { + "epoch": 0.8575162479415367, + "grad_norm": 3.404867649078369, + "learning_rate": 9.856138458785314e-06, + "loss": 0.1213, + "step": 36581 + }, + { + "epoch": 0.8575396895163417, + "grad_norm": 2.008401870727539, + "learning_rate": 9.852949842469783e-06, + "loss": 0.1522, + "step": 36582 + }, + { + "epoch": 0.8575631310911467, + "grad_norm": 2.6020379066467285, + "learning_rate": 9.849761715294692e-06, + "loss": 0.106, + "step": 36583 + }, + { + "epoch": 0.8575865726659517, + "grad_norm": 2.2583625316619873, + "learning_rate": 9.846574077277315e-06, + "loss": 1.594, + "step": 36584 + }, + { + "epoch": 0.8576100142407567, + "grad_norm": 3.2470624446868896, + "learning_rate": 9.84338692843495e-06, + "loss": 0.2824, + "step": 36585 + }, + { + "epoch": 0.8576334558155617, + "grad_norm": 2.116353750228882, + "learning_rate": 9.840200268784928e-06, + "loss": 0.1148, + "step": 36586 + }, + { + "epoch": 0.8576568973903667, + "grad_norm": 0.320353239774704, + "learning_rate": 9.837014098344488e-06, + "loss": 0.0176, + "step": 36587 + }, + { + "epoch": 0.8576803389651717, + "grad_norm": 3.288318157196045, + "learning_rate": 9.833828417130964e-06, + "loss": 0.2846, + "step": 36588 + }, + { + "epoch": 0.8577037805399766, + "grad_norm": 1.434748649597168, + "learning_rate": 9.830643225161628e-06, + "loss": 0.0857, + "step": 36589 + }, + { + "epoch": 0.8577272221147817, + "grad_norm": 6.991396427154541, + "learning_rate": 9.827458522453748e-06, + "loss": 0.328, + "step": 36590 + }, + { + "epoch": 0.8577506636895866, + "grad_norm": 1.9250456094741821, + "learning_rate": 9.824274309024628e-06, + "loss": 0.1239, + "step": 36591 + }, + { + "epoch": 0.8577741052643917, + "grad_norm": 1.6497224569320679, + "learning_rate": 9.821090584891501e-06, + "loss": 1.2538, + "step": 36592 + }, + { + "epoch": 0.8577975468391966, + "grad_norm": 4.167444705963135, + "learning_rate": 9.817907350071687e-06, + "loss": 0.3938, + "step": 36593 + }, + { + "epoch": 0.8578209884140017, + "grad_norm": 2.036762237548828, + "learning_rate": 9.814724604582448e-06, + "loss": 1.3573, + "step": 36594 + }, + { + "epoch": 0.8578444299888066, + "grad_norm": 4.058417320251465, + "learning_rate": 9.811542348441049e-06, + "loss": 0.1679, + "step": 36595 + }, + { + "epoch": 0.8578678715636117, + "grad_norm": 1.6692750453948975, + "learning_rate": 9.808360581664767e-06, + "loss": 0.1164, + "step": 36596 + }, + { + "epoch": 0.8578913131384166, + "grad_norm": 3.1969478130340576, + "learning_rate": 9.80517930427084e-06, + "loss": 0.1827, + "step": 36597 + }, + { + "epoch": 0.8579147547132217, + "grad_norm": 0.6575228571891785, + "learning_rate": 9.801998516276534e-06, + "loss": 0.0803, + "step": 36598 + }, + { + "epoch": 0.8579381962880266, + "grad_norm": 5.333802223205566, + "learning_rate": 9.798818217699124e-06, + "loss": 0.5196, + "step": 36599 + }, + { + "epoch": 0.8579616378628316, + "grad_norm": 3.140848159790039, + "learning_rate": 9.795638408555885e-06, + "loss": 0.2063, + "step": 36600 + }, + { + "epoch": 0.8579850794376366, + "grad_norm": 2.343323230743408, + "learning_rate": 9.792459088864037e-06, + "loss": 0.149, + "step": 36601 + }, + { + "epoch": 0.8580085210124416, + "grad_norm": 2.1982264518737793, + "learning_rate": 9.789280258640853e-06, + "loss": 1.3393, + "step": 36602 + }, + { + "epoch": 0.8580319625872466, + "grad_norm": 1.779826045036316, + "learning_rate": 9.786101917903556e-06, + "loss": 0.1631, + "step": 36603 + }, + { + "epoch": 0.8580554041620516, + "grad_norm": 0.6297589540481567, + "learning_rate": 9.78292406666942e-06, + "loss": 0.0388, + "step": 36604 + }, + { + "epoch": 0.8580788457368566, + "grad_norm": 2.064850330352783, + "learning_rate": 9.77974670495565e-06, + "loss": 0.1792, + "step": 36605 + }, + { + "epoch": 0.8581022873116616, + "grad_norm": 1.080102801322937, + "learning_rate": 9.776569832779537e-06, + "loss": 0.9397, + "step": 36606 + }, + { + "epoch": 0.8581257288864665, + "grad_norm": 3.1797101497650146, + "learning_rate": 9.773393450158285e-06, + "loss": 0.213, + "step": 36607 + }, + { + "epoch": 0.8581491704612716, + "grad_norm": 3.2301254272460938, + "learning_rate": 9.770217557109141e-06, + "loss": 0.456, + "step": 36608 + }, + { + "epoch": 0.8581726120360765, + "grad_norm": 3.6687960624694824, + "learning_rate": 9.76704215364933e-06, + "loss": 0.1876, + "step": 36609 + }, + { + "epoch": 0.8581960536108816, + "grad_norm": 1.2252033948898315, + "learning_rate": 9.763867239796065e-06, + "loss": 0.1393, + "step": 36610 + }, + { + "epoch": 0.8582194951856865, + "grad_norm": 1.9162147045135498, + "learning_rate": 9.760692815566608e-06, + "loss": 0.103, + "step": 36611 + }, + { + "epoch": 0.8582429367604916, + "grad_norm": 2.753418207168579, + "learning_rate": 9.757518880978167e-06, + "loss": 0.1176, + "step": 36612 + }, + { + "epoch": 0.8582663783352966, + "grad_norm": 2.1384739875793457, + "learning_rate": 9.754345436047984e-06, + "loss": 0.1343, + "step": 36613 + }, + { + "epoch": 0.8582898199101016, + "grad_norm": 3.057284116744995, + "learning_rate": 9.751172480793247e-06, + "loss": 0.3547, + "step": 36614 + }, + { + "epoch": 0.8583132614849066, + "grad_norm": 0.9605786204338074, + "learning_rate": 9.748000015231196e-06, + "loss": 0.0688, + "step": 36615 + }, + { + "epoch": 0.8583367030597115, + "grad_norm": 1.7342498302459717, + "learning_rate": 9.744828039379028e-06, + "loss": 0.0689, + "step": 36616 + }, + { + "epoch": 0.8583601446345166, + "grad_norm": 1.3688178062438965, + "learning_rate": 9.741656553253942e-06, + "loss": 0.1448, + "step": 36617 + }, + { + "epoch": 0.8583835862093215, + "grad_norm": 5.263716220855713, + "learning_rate": 9.738485556873189e-06, + "loss": 0.5081, + "step": 36618 + }, + { + "epoch": 0.8584070277841266, + "grad_norm": 0.7919503450393677, + "learning_rate": 9.735315050253946e-06, + "loss": 0.0342, + "step": 36619 + }, + { + "epoch": 0.8584304693589315, + "grad_norm": 4.3699469566345215, + "learning_rate": 9.73214503341342e-06, + "loss": 0.4373, + "step": 36620 + }, + { + "epoch": 0.8584539109337366, + "grad_norm": 3.686035394668579, + "learning_rate": 9.72897550636882e-06, + "loss": 0.2059, + "step": 36621 + }, + { + "epoch": 0.8584773525085415, + "grad_norm": 2.2563798427581787, + "learning_rate": 9.725806469137333e-06, + "loss": 0.1573, + "step": 36622 + }, + { + "epoch": 0.8585007940833466, + "grad_norm": 3.789231538772583, + "learning_rate": 9.722637921736144e-06, + "loss": 0.345, + "step": 36623 + }, + { + "epoch": 0.8585242356581515, + "grad_norm": 2.249354600906372, + "learning_rate": 9.719469864182452e-06, + "loss": 1.1453, + "step": 36624 + }, + { + "epoch": 0.8585476772329566, + "grad_norm": 2.3447935581207275, + "learning_rate": 9.716302296493484e-06, + "loss": 0.1453, + "step": 36625 + }, + { + "epoch": 0.8585711188077615, + "grad_norm": 1.329430103302002, + "learning_rate": 9.713135218686386e-06, + "loss": 0.1368, + "step": 36626 + }, + { + "epoch": 0.8585945603825665, + "grad_norm": 4.060398578643799, + "learning_rate": 9.709968630778366e-06, + "loss": 0.2119, + "step": 36627 + }, + { + "epoch": 0.8586180019573715, + "grad_norm": 0.7553934454917908, + "learning_rate": 9.706802532786596e-06, + "loss": 0.0633, + "step": 36628 + }, + { + "epoch": 0.8586414435321765, + "grad_norm": 2.805856943130493, + "learning_rate": 9.703636924728244e-06, + "loss": 0.306, + "step": 36629 + }, + { + "epoch": 0.8586648851069815, + "grad_norm": 0.7741292119026184, + "learning_rate": 9.70047180662048e-06, + "loss": 0.0676, + "step": 36630 + }, + { + "epoch": 0.8586883266817865, + "grad_norm": 6.32374906539917, + "learning_rate": 9.69730717848052e-06, + "loss": 0.5609, + "step": 36631 + }, + { + "epoch": 0.8587117682565915, + "grad_norm": 3.1495072841644287, + "learning_rate": 9.69414304032551e-06, + "loss": 0.2213, + "step": 36632 + }, + { + "epoch": 0.8587352098313965, + "grad_norm": 3.5969860553741455, + "learning_rate": 9.690979392172616e-06, + "loss": 0.2593, + "step": 36633 + }, + { + "epoch": 0.8587586514062014, + "grad_norm": 3.503777503967285, + "learning_rate": 9.687816234039004e-06, + "loss": 0.2057, + "step": 36634 + }, + { + "epoch": 0.8587820929810065, + "grad_norm": 2.620743751525879, + "learning_rate": 9.684653565941825e-06, + "loss": 0.2802, + "step": 36635 + }, + { + "epoch": 0.8588055345558114, + "grad_norm": 3.1297554969787598, + "learning_rate": 9.681491387898278e-06, + "loss": 0.2495, + "step": 36636 + }, + { + "epoch": 0.8588289761306165, + "grad_norm": 2.8911948204040527, + "learning_rate": 9.67832969992547e-06, + "loss": 0.2166, + "step": 36637 + }, + { + "epoch": 0.8588524177054214, + "grad_norm": 2.5297582149505615, + "learning_rate": 9.67516850204061e-06, + "loss": 1.183, + "step": 36638 + }, + { + "epoch": 0.8588758592802265, + "grad_norm": 0.9988753199577332, + "learning_rate": 9.672007794260829e-06, + "loss": 0.0553, + "step": 36639 + }, + { + "epoch": 0.8588993008550314, + "grad_norm": 1.5144230127334595, + "learning_rate": 9.668847576603269e-06, + "loss": 0.1386, + "step": 36640 + }, + { + "epoch": 0.8589227424298365, + "grad_norm": 2.953822135925293, + "learning_rate": 9.665687849085082e-06, + "loss": 0.3032, + "step": 36641 + }, + { + "epoch": 0.8589461840046414, + "grad_norm": 2.255192279815674, + "learning_rate": 9.662528611723387e-06, + "loss": 0.1615, + "step": 36642 + }, + { + "epoch": 0.8589696255794464, + "grad_norm": 3.1984119415283203, + "learning_rate": 9.659369864535383e-06, + "loss": 0.1865, + "step": 36643 + }, + { + "epoch": 0.8589930671542514, + "grad_norm": 3.298760414123535, + "learning_rate": 9.656211607538169e-06, + "loss": 0.2853, + "step": 36644 + }, + { + "epoch": 0.8590165087290564, + "grad_norm": 2.9017956256866455, + "learning_rate": 9.653053840748893e-06, + "loss": 0.2643, + "step": 36645 + }, + { + "epoch": 0.8590399503038614, + "grad_norm": 3.3188648223876953, + "learning_rate": 9.64989656418469e-06, + "loss": 0.2731, + "step": 36646 + }, + { + "epoch": 0.8590633918786664, + "grad_norm": 3.535158157348633, + "learning_rate": 9.64673977786269e-06, + "loss": 0.3426, + "step": 36647 + }, + { + "epoch": 0.8590868334534714, + "grad_norm": 3.8479647636413574, + "learning_rate": 9.6435834818e-06, + "loss": 0.3317, + "step": 36648 + }, + { + "epoch": 0.8591102750282764, + "grad_norm": 1.0608365535736084, + "learning_rate": 9.640427676013764e-06, + "loss": 0.0708, + "step": 36649 + }, + { + "epoch": 0.8591337166030814, + "grad_norm": 2.982196092605591, + "learning_rate": 9.637272360521133e-06, + "loss": 0.162, + "step": 36650 + }, + { + "epoch": 0.8591571581778864, + "grad_norm": 0.7495642304420471, + "learning_rate": 9.634117535339193e-06, + "loss": 0.0327, + "step": 36651 + }, + { + "epoch": 0.8591805997526913, + "grad_norm": 2.3420605659484863, + "learning_rate": 9.630963200485077e-06, + "loss": 0.1412, + "step": 36652 + }, + { + "epoch": 0.8592040413274964, + "grad_norm": 0.4711824953556061, + "learning_rate": 9.627809355975904e-06, + "loss": 0.0382, + "step": 36653 + }, + { + "epoch": 0.8592274829023013, + "grad_norm": 2.671457052230835, + "learning_rate": 9.624656001828769e-06, + "loss": 0.2334, + "step": 36654 + }, + { + "epoch": 0.8592509244771064, + "grad_norm": 1.4420417547225952, + "learning_rate": 9.621503138060772e-06, + "loss": 0.1124, + "step": 36655 + }, + { + "epoch": 0.8592743660519113, + "grad_norm": 2.428903341293335, + "learning_rate": 9.618350764689065e-06, + "loss": 0.1856, + "step": 36656 + }, + { + "epoch": 0.8592978076267164, + "grad_norm": 2.199045419692993, + "learning_rate": 9.615198881730726e-06, + "loss": 0.0603, + "step": 36657 + }, + { + "epoch": 0.8593212492015213, + "grad_norm": 3.1552538871765137, + "learning_rate": 9.612047489202858e-06, + "loss": 0.3173, + "step": 36658 + }, + { + "epoch": 0.8593446907763264, + "grad_norm": 0.24751821160316467, + "learning_rate": 9.608896587122562e-06, + "loss": 0.0149, + "step": 36659 + }, + { + "epoch": 0.8593681323511313, + "grad_norm": 2.0245118141174316, + "learning_rate": 9.605746175506913e-06, + "loss": 1.4119, + "step": 36660 + }, + { + "epoch": 0.8593915739259363, + "grad_norm": 2.516385555267334, + "learning_rate": 9.602596254373041e-06, + "loss": 1.4658, + "step": 36661 + }, + { + "epoch": 0.8594150155007413, + "grad_norm": 3.4787700176239014, + "learning_rate": 9.599446823738022e-06, + "loss": 0.1636, + "step": 36662 + }, + { + "epoch": 0.8594384570755463, + "grad_norm": 0.6644935607910156, + "learning_rate": 9.596297883618955e-06, + "loss": 0.0191, + "step": 36663 + }, + { + "epoch": 0.8594618986503514, + "grad_norm": 1.8402721881866455, + "learning_rate": 9.593149434032912e-06, + "loss": 0.0685, + "step": 36664 + }, + { + "epoch": 0.8594853402251563, + "grad_norm": 0.37103983759880066, + "learning_rate": 9.59000147499699e-06, + "loss": 0.0201, + "step": 36665 + }, + { + "epoch": 0.8595087817999614, + "grad_norm": 1.0938197374343872, + "learning_rate": 9.586854006528256e-06, + "loss": 0.0781, + "step": 36666 + }, + { + "epoch": 0.8595322233747663, + "grad_norm": 2.024320363998413, + "learning_rate": 9.583707028643807e-06, + "loss": 0.1306, + "step": 36667 + }, + { + "epoch": 0.8595556649495714, + "grad_norm": 0.4345240890979767, + "learning_rate": 9.580560541360673e-06, + "loss": 0.0292, + "step": 36668 + }, + { + "epoch": 0.8595791065243763, + "grad_norm": 2.7012827396392822, + "learning_rate": 9.577414544695984e-06, + "loss": 1.7657, + "step": 36669 + }, + { + "epoch": 0.8596025480991814, + "grad_norm": 0.25579747557640076, + "learning_rate": 9.574269038666784e-06, + "loss": 0.0124, + "step": 36670 + }, + { + "epoch": 0.8596259896739863, + "grad_norm": 4.156151294708252, + "learning_rate": 9.571124023290146e-06, + "loss": 0.2716, + "step": 36671 + }, + { + "epoch": 0.8596494312487913, + "grad_norm": 0.8099514245986938, + "learning_rate": 9.567979498583113e-06, + "loss": 0.084, + "step": 36672 + }, + { + "epoch": 0.8596728728235963, + "grad_norm": 2.2164032459259033, + "learning_rate": 9.564835464562783e-06, + "loss": 1.0449, + "step": 36673 + }, + { + "epoch": 0.8596963143984013, + "grad_norm": 2.650049924850464, + "learning_rate": 9.561691921246207e-06, + "loss": 0.2191, + "step": 36674 + }, + { + "epoch": 0.8597197559732063, + "grad_norm": 3.844109296798706, + "learning_rate": 9.558548868650407e-06, + "loss": 0.2232, + "step": 36675 + }, + { + "epoch": 0.8597431975480113, + "grad_norm": 0.3158988654613495, + "learning_rate": 9.55540630679248e-06, + "loss": 0.0297, + "step": 36676 + }, + { + "epoch": 0.8597666391228163, + "grad_norm": 0.7383468151092529, + "learning_rate": 9.552264235689467e-06, + "loss": 0.089, + "step": 36677 + }, + { + "epoch": 0.8597900806976213, + "grad_norm": 1.52143394947052, + "learning_rate": 9.5491226553584e-06, + "loss": 0.0835, + "step": 36678 + }, + { + "epoch": 0.8598135222724262, + "grad_norm": 2.9365832805633545, + "learning_rate": 9.545981565816353e-06, + "loss": 0.2329, + "step": 36679 + }, + { + "epoch": 0.8598369638472313, + "grad_norm": 0.44625502824783325, + "learning_rate": 9.542840967080324e-06, + "loss": 0.035, + "step": 36680 + }, + { + "epoch": 0.8598604054220362, + "grad_norm": 1.2423248291015625, + "learning_rate": 9.53970085916741e-06, + "loss": 0.1179, + "step": 36681 + }, + { + "epoch": 0.8598838469968413, + "grad_norm": 3.8364899158477783, + "learning_rate": 9.536561242094611e-06, + "loss": 0.4667, + "step": 36682 + }, + { + "epoch": 0.8599072885716462, + "grad_norm": 2.431318521499634, + "learning_rate": 9.533422115878987e-06, + "loss": 1.9899, + "step": 36683 + }, + { + "epoch": 0.8599307301464513, + "grad_norm": 1.7826930284500122, + "learning_rate": 9.530283480537528e-06, + "loss": 0.0734, + "step": 36684 + }, + { + "epoch": 0.8599541717212562, + "grad_norm": 0.8178507685661316, + "learning_rate": 9.527145336087317e-06, + "loss": 0.04, + "step": 36685 + }, + { + "epoch": 0.8599776132960613, + "grad_norm": 2.7352747917175293, + "learning_rate": 9.524007682545366e-06, + "loss": 0.3201, + "step": 36686 + }, + { + "epoch": 0.8600010548708662, + "grad_norm": 3.232564687728882, + "learning_rate": 9.520870519928671e-06, + "loss": 0.395, + "step": 36687 + }, + { + "epoch": 0.8600244964456712, + "grad_norm": 0.3961043059825897, + "learning_rate": 9.517733848254295e-06, + "loss": 0.047, + "step": 36688 + }, + { + "epoch": 0.8600479380204762, + "grad_norm": 0.6123956441879272, + "learning_rate": 9.514597667539238e-06, + "loss": 0.0225, + "step": 36689 + }, + { + "epoch": 0.8600713795952812, + "grad_norm": 0.577536404132843, + "learning_rate": 9.511461977800518e-06, + "loss": 0.0777, + "step": 36690 + }, + { + "epoch": 0.8600948211700862, + "grad_norm": 3.3083651065826416, + "learning_rate": 9.508326779055154e-06, + "loss": 0.3233, + "step": 36691 + }, + { + "epoch": 0.8601182627448912, + "grad_norm": 1.8709512948989868, + "learning_rate": 9.505192071320145e-06, + "loss": 0.2375, + "step": 36692 + }, + { + "epoch": 0.8601417043196962, + "grad_norm": 1.3020284175872803, + "learning_rate": 9.50205785461249e-06, + "loss": 0.6007, + "step": 36693 + }, + { + "epoch": 0.8601651458945012, + "grad_norm": 0.8010807037353516, + "learning_rate": 9.498924128949239e-06, + "loss": 0.032, + "step": 36694 + }, + { + "epoch": 0.8601885874693062, + "grad_norm": 2.926337480545044, + "learning_rate": 9.495790894347367e-06, + "loss": 0.4829, + "step": 36695 + }, + { + "epoch": 0.8602120290441112, + "grad_norm": 3.191375970840454, + "learning_rate": 9.492658150823864e-06, + "loss": 0.2097, + "step": 36696 + }, + { + "epoch": 0.8602354706189161, + "grad_norm": 3.9374706745147705, + "learning_rate": 9.489525898395758e-06, + "loss": 1.7461, + "step": 36697 + }, + { + "epoch": 0.8602589121937212, + "grad_norm": 2.796407461166382, + "learning_rate": 9.486394137080034e-06, + "loss": 1.699, + "step": 36698 + }, + { + "epoch": 0.8602823537685261, + "grad_norm": 0.8931389451026917, + "learning_rate": 9.483262866893672e-06, + "loss": 0.0254, + "step": 36699 + }, + { + "epoch": 0.8603057953433312, + "grad_norm": 2.941277503967285, + "learning_rate": 9.480132087853654e-06, + "loss": 1.7884, + "step": 36700 + }, + { + "epoch": 0.8603292369181361, + "grad_norm": 2.6046669483184814, + "learning_rate": 9.477001799977003e-06, + "loss": 0.2453, + "step": 36701 + }, + { + "epoch": 0.8603526784929412, + "grad_norm": 3.5788798332214355, + "learning_rate": 9.473872003280693e-06, + "loss": 0.3043, + "step": 36702 + }, + { + "epoch": 0.8603761200677461, + "grad_norm": 0.9386955499649048, + "learning_rate": 9.470742697781687e-06, + "loss": 0.0636, + "step": 36703 + }, + { + "epoch": 0.8603995616425512, + "grad_norm": 2.7122042179107666, + "learning_rate": 9.467613883496983e-06, + "loss": 0.2682, + "step": 36704 + }, + { + "epoch": 0.8604230032173561, + "grad_norm": 2.3782217502593994, + "learning_rate": 9.464485560443559e-06, + "loss": 0.2401, + "step": 36705 + }, + { + "epoch": 0.8604464447921611, + "grad_norm": 3.6400203704833984, + "learning_rate": 9.461357728638354e-06, + "loss": 0.2224, + "step": 36706 + }, + { + "epoch": 0.8604698863669661, + "grad_norm": 3.7548956871032715, + "learning_rate": 9.458230388098388e-06, + "loss": 0.3961, + "step": 36707 + }, + { + "epoch": 0.8604933279417711, + "grad_norm": 4.063116550445557, + "learning_rate": 9.455103538840592e-06, + "loss": 0.2949, + "step": 36708 + }, + { + "epoch": 0.8605167695165761, + "grad_norm": 0.5232821702957153, + "learning_rate": 9.451977180881966e-06, + "loss": 0.0612, + "step": 36709 + }, + { + "epoch": 0.8605402110913811, + "grad_norm": 0.9575291275978088, + "learning_rate": 9.44885131423946e-06, + "loss": 0.0846, + "step": 36710 + }, + { + "epoch": 0.8605636526661861, + "grad_norm": 2.3685050010681152, + "learning_rate": 9.445725938930039e-06, + "loss": 0.1326, + "step": 36711 + }, + { + "epoch": 0.8605870942409911, + "grad_norm": 0.9683002829551697, + "learning_rate": 9.442601054970645e-06, + "loss": 0.0581, + "step": 36712 + }, + { + "epoch": 0.860610535815796, + "grad_norm": 1.1762211322784424, + "learning_rate": 9.439476662378233e-06, + "loss": 0.1011, + "step": 36713 + }, + { + "epoch": 0.8606339773906011, + "grad_norm": 1.746681571006775, + "learning_rate": 9.436352761169775e-06, + "loss": 0.0894, + "step": 36714 + }, + { + "epoch": 0.8606574189654062, + "grad_norm": 1.4244532585144043, + "learning_rate": 9.433229351362227e-06, + "loss": 0.765, + "step": 36715 + }, + { + "epoch": 0.8606808605402111, + "grad_norm": 0.20653477311134338, + "learning_rate": 9.430106432972519e-06, + "loss": 0.0176, + "step": 36716 + }, + { + "epoch": 0.8607043021150161, + "grad_norm": 1.2239325046539307, + "learning_rate": 9.426984006017592e-06, + "loss": 0.1794, + "step": 36717 + }, + { + "epoch": 0.8607277436898211, + "grad_norm": 3.598623514175415, + "learning_rate": 9.423862070514377e-06, + "loss": 0.1425, + "step": 36718 + }, + { + "epoch": 0.8607511852646261, + "grad_norm": 1.4276303052902222, + "learning_rate": 9.420740626479863e-06, + "loss": 0.0783, + "step": 36719 + }, + { + "epoch": 0.8607746268394311, + "grad_norm": 2.319331407546997, + "learning_rate": 9.417619673930932e-06, + "loss": 0.1721, + "step": 36720 + }, + { + "epoch": 0.8607980684142361, + "grad_norm": 3.2420575618743896, + "learning_rate": 9.414499212884564e-06, + "loss": 1.1702, + "step": 36721 + }, + { + "epoch": 0.8608215099890411, + "grad_norm": 2.8533108234405518, + "learning_rate": 9.411379243357677e-06, + "loss": 0.2302, + "step": 36722 + }, + { + "epoch": 0.8608449515638461, + "grad_norm": 2.074843406677246, + "learning_rate": 9.408259765367178e-06, + "loss": 0.1548, + "step": 36723 + }, + { + "epoch": 0.860868393138651, + "grad_norm": 2.459465742111206, + "learning_rate": 9.405140778930021e-06, + "loss": 0.1618, + "step": 36724 + }, + { + "epoch": 0.8608918347134561, + "grad_norm": 1.4053400754928589, + "learning_rate": 9.402022284063095e-06, + "loss": 0.1723, + "step": 36725 + }, + { + "epoch": 0.860915276288261, + "grad_norm": 3.1433651447296143, + "learning_rate": 9.39890428078336e-06, + "loss": 0.2684, + "step": 36726 + }, + { + "epoch": 0.8609387178630661, + "grad_norm": 1.1220206022262573, + "learning_rate": 9.395786769107728e-06, + "loss": 0.1505, + "step": 36727 + }, + { + "epoch": 0.860962159437871, + "grad_norm": 2.255148410797119, + "learning_rate": 9.392669749053096e-06, + "loss": 0.2702, + "step": 36728 + }, + { + "epoch": 0.8609856010126761, + "grad_norm": 3.7101190090179443, + "learning_rate": 9.389553220636383e-06, + "loss": 0.3253, + "step": 36729 + }, + { + "epoch": 0.861009042587481, + "grad_norm": 1.8407341241836548, + "learning_rate": 9.386437183874508e-06, + "loss": 0.1691, + "step": 36730 + }, + { + "epoch": 0.8610324841622861, + "grad_norm": 1.1054017543792725, + "learning_rate": 9.383321638784349e-06, + "loss": 0.0951, + "step": 36731 + }, + { + "epoch": 0.861055925737091, + "grad_norm": 0.09435002505779266, + "learning_rate": 9.380206585382856e-06, + "loss": 0.0025, + "step": 36732 + }, + { + "epoch": 0.861079367311896, + "grad_norm": 1.8342794179916382, + "learning_rate": 9.377092023686884e-06, + "loss": 0.1364, + "step": 36733 + }, + { + "epoch": 0.861102808886701, + "grad_norm": 1.9464994668960571, + "learning_rate": 9.373977953713387e-06, + "loss": 0.172, + "step": 36734 + }, + { + "epoch": 0.861126250461506, + "grad_norm": 1.88998281955719, + "learning_rate": 9.370864375479238e-06, + "loss": 1.2057, + "step": 36735 + }, + { + "epoch": 0.861149692036311, + "grad_norm": 2.23551869392395, + "learning_rate": 9.367751289001314e-06, + "loss": 0.2504, + "step": 36736 + }, + { + "epoch": 0.861173133611116, + "grad_norm": 2.724600315093994, + "learning_rate": 9.364638694296534e-06, + "loss": 0.1717, + "step": 36737 + }, + { + "epoch": 0.861196575185921, + "grad_norm": 1.5498634576797485, + "learning_rate": 9.361526591381742e-06, + "loss": 0.0894, + "step": 36738 + }, + { + "epoch": 0.861220016760726, + "grad_norm": 0.7059367299079895, + "learning_rate": 9.358414980273878e-06, + "loss": 0.0411, + "step": 36739 + }, + { + "epoch": 0.861243458335531, + "grad_norm": 1.809088110923767, + "learning_rate": 9.355303860989805e-06, + "loss": 0.1231, + "step": 36740 + }, + { + "epoch": 0.861266899910336, + "grad_norm": 4.003778457641602, + "learning_rate": 9.352193233546402e-06, + "loss": 0.0643, + "step": 36741 + }, + { + "epoch": 0.8612903414851409, + "grad_norm": 3.594256639480591, + "learning_rate": 9.349083097960543e-06, + "loss": 0.35, + "step": 36742 + }, + { + "epoch": 0.861313783059946, + "grad_norm": 2.8929495811462402, + "learning_rate": 9.345973454249114e-06, + "loss": 0.153, + "step": 36743 + }, + { + "epoch": 0.8613372246347509, + "grad_norm": 1.3681762218475342, + "learning_rate": 9.342864302428955e-06, + "loss": 0.7742, + "step": 36744 + }, + { + "epoch": 0.861360666209556, + "grad_norm": 1.0025399923324585, + "learning_rate": 9.339755642516968e-06, + "loss": 0.0904, + "step": 36745 + }, + { + "epoch": 0.8613841077843609, + "grad_norm": 2.885277271270752, + "learning_rate": 9.336647474530036e-06, + "loss": 0.3646, + "step": 36746 + }, + { + "epoch": 0.861407549359166, + "grad_norm": 1.7523443698883057, + "learning_rate": 9.333539798485003e-06, + "loss": 0.1511, + "step": 36747 + }, + { + "epoch": 0.8614309909339709, + "grad_norm": 3.983086585998535, + "learning_rate": 9.330432614398732e-06, + "loss": 0.4963, + "step": 36748 + }, + { + "epoch": 0.861454432508776, + "grad_norm": 2.4404659271240234, + "learning_rate": 9.327325922288089e-06, + "loss": 0.2238, + "step": 36749 + }, + { + "epoch": 0.8614778740835809, + "grad_norm": 3.0749595165252686, + "learning_rate": 9.324219722169913e-06, + "loss": 0.2902, + "step": 36750 + }, + { + "epoch": 0.861501315658386, + "grad_norm": 0.719353199005127, + "learning_rate": 9.321114014061062e-06, + "loss": 0.0756, + "step": 36751 + }, + { + "epoch": 0.8615247572331909, + "grad_norm": 1.183656930923462, + "learning_rate": 9.318008797978406e-06, + "loss": 0.185, + "step": 36752 + }, + { + "epoch": 0.8615481988079959, + "grad_norm": 2.3391597270965576, + "learning_rate": 9.31490407393879e-06, + "loss": 1.3058, + "step": 36753 + }, + { + "epoch": 0.8615716403828009, + "grad_norm": 4.28443717956543, + "learning_rate": 9.311799841959057e-06, + "loss": 0.3017, + "step": 36754 + }, + { + "epoch": 0.8615950819576059, + "grad_norm": 1.0118423700332642, + "learning_rate": 9.308696102056047e-06, + "loss": 0.0576, + "step": 36755 + }, + { + "epoch": 0.8616185235324109, + "grad_norm": 2.0286872386932373, + "learning_rate": 9.30559285424657e-06, + "loss": 0.1482, + "step": 36756 + }, + { + "epoch": 0.8616419651072159, + "grad_norm": 3.174853801727295, + "learning_rate": 9.302490098547534e-06, + "loss": 0.2846, + "step": 36757 + }, + { + "epoch": 0.8616654066820209, + "grad_norm": 3.091907262802124, + "learning_rate": 9.299387834975703e-06, + "loss": 0.3143, + "step": 36758 + }, + { + "epoch": 0.8616888482568259, + "grad_norm": 0.12272539734840393, + "learning_rate": 9.296286063547977e-06, + "loss": 0.0035, + "step": 36759 + }, + { + "epoch": 0.8617122898316308, + "grad_norm": 1.3208422660827637, + "learning_rate": 9.293184784281139e-06, + "loss": 0.122, + "step": 36760 + }, + { + "epoch": 0.8617357314064359, + "grad_norm": 1.3313456773757935, + "learning_rate": 9.290083997192034e-06, + "loss": 0.1262, + "step": 36761 + }, + { + "epoch": 0.8617591729812408, + "grad_norm": 2.29760479927063, + "learning_rate": 9.286983702297491e-06, + "loss": 0.1253, + "step": 36762 + }, + { + "epoch": 0.8617826145560459, + "grad_norm": 7.23256254196167, + "learning_rate": 9.283883899614299e-06, + "loss": 0.2531, + "step": 36763 + }, + { + "epoch": 0.8618060561308508, + "grad_norm": 3.3101794719696045, + "learning_rate": 9.28078458915933e-06, + "loss": 0.2596, + "step": 36764 + }, + { + "epoch": 0.8618294977056559, + "grad_norm": 2.0810835361480713, + "learning_rate": 9.277685770949373e-06, + "loss": 0.118, + "step": 36765 + }, + { + "epoch": 0.8618529392804609, + "grad_norm": 3.128993511199951, + "learning_rate": 9.274587445001238e-06, + "loss": 0.1045, + "step": 36766 + }, + { + "epoch": 0.8618763808552659, + "grad_norm": 0.61441969871521, + "learning_rate": 9.271489611331741e-06, + "loss": 0.0322, + "step": 36767 + }, + { + "epoch": 0.8618998224300709, + "grad_norm": 3.2186803817749023, + "learning_rate": 9.268392269957704e-06, + "loss": 0.1941, + "step": 36768 + }, + { + "epoch": 0.8619232640048758, + "grad_norm": 0.6766477823257446, + "learning_rate": 9.265295420895891e-06, + "loss": 0.0472, + "step": 36769 + }, + { + "epoch": 0.8619467055796809, + "grad_norm": 0.567472517490387, + "learning_rate": 9.262199064163147e-06, + "loss": 0.0395, + "step": 36770 + }, + { + "epoch": 0.8619701471544858, + "grad_norm": 2.6066932678222656, + "learning_rate": 9.259103199776275e-06, + "loss": 0.1425, + "step": 36771 + }, + { + "epoch": 0.8619935887292909, + "grad_norm": 2.3004419803619385, + "learning_rate": 9.256007827752055e-06, + "loss": 0.215, + "step": 36772 + }, + { + "epoch": 0.8620170303040958, + "grad_norm": 1.8471256494522095, + "learning_rate": 9.252912948107295e-06, + "loss": 0.1446, + "step": 36773 + }, + { + "epoch": 0.8620404718789009, + "grad_norm": 0.9554805159568787, + "learning_rate": 9.24981856085878e-06, + "loss": 0.1063, + "step": 36774 + }, + { + "epoch": 0.8620639134537058, + "grad_norm": 1.0110468864440918, + "learning_rate": 9.246724666023299e-06, + "loss": 0.0327, + "step": 36775 + }, + { + "epoch": 0.8620873550285109, + "grad_norm": 0.4364291727542877, + "learning_rate": 9.243631263617625e-06, + "loss": 0.0232, + "step": 36776 + }, + { + "epoch": 0.8621107966033158, + "grad_norm": 2.902207612991333, + "learning_rate": 9.24053835365858e-06, + "loss": 0.2427, + "step": 36777 + }, + { + "epoch": 0.8621342381781208, + "grad_norm": 2.1916279792785645, + "learning_rate": 9.237445936162925e-06, + "loss": 0.2067, + "step": 36778 + }, + { + "epoch": 0.8621576797529258, + "grad_norm": 1.8996554613113403, + "learning_rate": 9.23435401114744e-06, + "loss": 0.1614, + "step": 36779 + }, + { + "epoch": 0.8621811213277308, + "grad_norm": 1.9075862169265747, + "learning_rate": 9.231262578628908e-06, + "loss": 0.1174, + "step": 36780 + }, + { + "epoch": 0.8622045629025358, + "grad_norm": 2.056405782699585, + "learning_rate": 9.228171638624062e-06, + "loss": 0.1268, + "step": 36781 + }, + { + "epoch": 0.8622280044773408, + "grad_norm": 4.43155574798584, + "learning_rate": 9.225081191149743e-06, + "loss": 0.4643, + "step": 36782 + }, + { + "epoch": 0.8622514460521458, + "grad_norm": 0.8717533946037292, + "learning_rate": 9.22199123622266e-06, + "loss": 0.0716, + "step": 36783 + }, + { + "epoch": 0.8622748876269508, + "grad_norm": 2.679433822631836, + "learning_rate": 9.218901773859623e-06, + "loss": 0.175, + "step": 36784 + }, + { + "epoch": 0.8622983292017558, + "grad_norm": 2.189069986343384, + "learning_rate": 9.215812804077384e-06, + "loss": 0.1174, + "step": 36785 + }, + { + "epoch": 0.8623217707765608, + "grad_norm": 2.397671937942505, + "learning_rate": 9.212724326892696e-06, + "loss": 0.1654, + "step": 36786 + }, + { + "epoch": 0.8623452123513657, + "grad_norm": 0.9153658151626587, + "learning_rate": 9.209636342322314e-06, + "loss": 0.0856, + "step": 36787 + }, + { + "epoch": 0.8623686539261708, + "grad_norm": 1.8791007995605469, + "learning_rate": 9.206548850383002e-06, + "loss": 0.1535, + "step": 36788 + }, + { + "epoch": 0.8623920955009757, + "grad_norm": 3.494457483291626, + "learning_rate": 9.203461851091488e-06, + "loss": 0.2639, + "step": 36789 + }, + { + "epoch": 0.8624155370757808, + "grad_norm": 2.4119796752929688, + "learning_rate": 9.200375344464563e-06, + "loss": 0.1023, + "step": 36790 + }, + { + "epoch": 0.8624389786505857, + "grad_norm": 2.7232306003570557, + "learning_rate": 9.197289330518955e-06, + "loss": 0.1589, + "step": 36791 + }, + { + "epoch": 0.8624624202253908, + "grad_norm": 0.6374641060829163, + "learning_rate": 9.194203809271406e-06, + "loss": 0.0295, + "step": 36792 + }, + { + "epoch": 0.8624858618001957, + "grad_norm": 1.9432891607284546, + "learning_rate": 9.19111878073864e-06, + "loss": 0.1172, + "step": 36793 + }, + { + "epoch": 0.8625093033750008, + "grad_norm": 2.626041889190674, + "learning_rate": 9.188034244937438e-06, + "loss": 0.2242, + "step": 36794 + }, + { + "epoch": 0.8625327449498057, + "grad_norm": 3.794426918029785, + "learning_rate": 9.184950201884512e-06, + "loss": 0.2004, + "step": 36795 + }, + { + "epoch": 0.8625561865246107, + "grad_norm": 0.989607572555542, + "learning_rate": 9.181866651596593e-06, + "loss": 0.0499, + "step": 36796 + }, + { + "epoch": 0.8625796280994157, + "grad_norm": 1.9027210474014282, + "learning_rate": 9.178783594090423e-06, + "loss": 0.1202, + "step": 36797 + }, + { + "epoch": 0.8626030696742207, + "grad_norm": 2.4532253742218018, + "learning_rate": 9.175701029382733e-06, + "loss": 0.1499, + "step": 36798 + }, + { + "epoch": 0.8626265112490257, + "grad_norm": 5.9761762619018555, + "learning_rate": 9.172618957490242e-06, + "loss": 0.4653, + "step": 36799 + }, + { + "epoch": 0.8626499528238307, + "grad_norm": 2.761758804321289, + "learning_rate": 9.169537378429683e-06, + "loss": 0.1498, + "step": 36800 + }, + { + "epoch": 0.8626733943986357, + "grad_norm": 1.6853828430175781, + "learning_rate": 9.16645629221774e-06, + "loss": 0.0858, + "step": 36801 + }, + { + "epoch": 0.8626968359734407, + "grad_norm": 2.6346192359924316, + "learning_rate": 9.16337569887118e-06, + "loss": 0.2308, + "step": 36802 + }, + { + "epoch": 0.8627202775482457, + "grad_norm": 3.0227267742156982, + "learning_rate": 9.1602955984067e-06, + "loss": 0.1796, + "step": 36803 + }, + { + "epoch": 0.8627437191230507, + "grad_norm": 0.22288845479488373, + "learning_rate": 9.157215990841005e-06, + "loss": 0.0115, + "step": 36804 + }, + { + "epoch": 0.8627671606978556, + "grad_norm": 0.5584088563919067, + "learning_rate": 9.15413687619081e-06, + "loss": 0.0353, + "step": 36805 + }, + { + "epoch": 0.8627906022726607, + "grad_norm": 1.3020027875900269, + "learning_rate": 9.15105825447281e-06, + "loss": 0.0935, + "step": 36806 + }, + { + "epoch": 0.8628140438474656, + "grad_norm": 3.7196261882781982, + "learning_rate": 9.147980125703736e-06, + "loss": 0.1887, + "step": 36807 + }, + { + "epoch": 0.8628374854222707, + "grad_norm": 0.9201107025146484, + "learning_rate": 9.144902489900265e-06, + "loss": 0.0491, + "step": 36808 + }, + { + "epoch": 0.8628609269970756, + "grad_norm": 1.961959958076477, + "learning_rate": 9.141825347079114e-06, + "loss": 0.1327, + "step": 36809 + }, + { + "epoch": 0.8628843685718807, + "grad_norm": 2.7039566040039062, + "learning_rate": 9.138748697256982e-06, + "loss": 0.1846, + "step": 36810 + }, + { + "epoch": 0.8629078101466856, + "grad_norm": 2.3870632648468018, + "learning_rate": 9.135672540450558e-06, + "loss": 0.172, + "step": 36811 + }, + { + "epoch": 0.8629312517214907, + "grad_norm": 2.1434245109558105, + "learning_rate": 9.132596876676524e-06, + "loss": 0.1401, + "step": 36812 + }, + { + "epoch": 0.8629546932962956, + "grad_norm": 2.797077178955078, + "learning_rate": 9.129521705951572e-06, + "loss": 0.2427, + "step": 36813 + }, + { + "epoch": 0.8629781348711006, + "grad_norm": 2.197777509689331, + "learning_rate": 9.126447028292372e-06, + "loss": 0.1998, + "step": 36814 + }, + { + "epoch": 0.8630015764459056, + "grad_norm": 2.091543674468994, + "learning_rate": 9.12337284371565e-06, + "loss": 0.1245, + "step": 36815 + }, + { + "epoch": 0.8630250180207106, + "grad_norm": 1.9876967668533325, + "learning_rate": 9.120299152238054e-06, + "loss": 1.3152, + "step": 36816 + }, + { + "epoch": 0.8630484595955156, + "grad_norm": 2.6871426105499268, + "learning_rate": 9.117225953876274e-06, + "loss": 0.3194, + "step": 36817 + }, + { + "epoch": 0.8630719011703206, + "grad_norm": 2.7157986164093018, + "learning_rate": 9.114153248646962e-06, + "loss": 0.2162, + "step": 36818 + }, + { + "epoch": 0.8630953427451257, + "grad_norm": 4.7490973472595215, + "learning_rate": 9.111081036566837e-06, + "loss": 0.3898, + "step": 36819 + }, + { + "epoch": 0.8631187843199306, + "grad_norm": 3.53414249420166, + "learning_rate": 9.108009317652534e-06, + "loss": 0.2571, + "step": 36820 + }, + { + "epoch": 0.8631422258947357, + "grad_norm": 2.173804759979248, + "learning_rate": 9.104938091920712e-06, + "loss": 0.1569, + "step": 36821 + }, + { + "epoch": 0.8631656674695406, + "grad_norm": 2.1440138816833496, + "learning_rate": 9.101867359388061e-06, + "loss": 0.0869, + "step": 36822 + }, + { + "epoch": 0.8631891090443456, + "grad_norm": 2.493726968765259, + "learning_rate": 9.098797120071245e-06, + "loss": 1.8769, + "step": 36823 + }, + { + "epoch": 0.8632125506191506, + "grad_norm": 0.31698814034461975, + "learning_rate": 9.095727373986907e-06, + "loss": 0.0284, + "step": 36824 + }, + { + "epoch": 0.8632359921939556, + "grad_norm": 4.201055526733398, + "learning_rate": 9.092658121151709e-06, + "loss": 0.2855, + "step": 36825 + }, + { + "epoch": 0.8632594337687606, + "grad_norm": 2.4703056812286377, + "learning_rate": 9.089589361582307e-06, + "loss": 0.1008, + "step": 36826 + }, + { + "epoch": 0.8632828753435656, + "grad_norm": 1.3630599975585938, + "learning_rate": 9.086521095295331e-06, + "loss": 0.1015, + "step": 36827 + }, + { + "epoch": 0.8633063169183706, + "grad_norm": 4.595876693725586, + "learning_rate": 9.083453322307466e-06, + "loss": 0.5507, + "step": 36828 + }, + { + "epoch": 0.8633297584931756, + "grad_norm": 3.791832447052002, + "learning_rate": 9.080386042635336e-06, + "loss": 0.2369, + "step": 36829 + }, + { + "epoch": 0.8633532000679806, + "grad_norm": 1.5627386569976807, + "learning_rate": 9.077319256295569e-06, + "loss": 0.0895, + "step": 36830 + }, + { + "epoch": 0.8633766416427856, + "grad_norm": 1.8952581882476807, + "learning_rate": 9.07425296330484e-06, + "loss": 0.1658, + "step": 36831 + }, + { + "epoch": 0.8634000832175905, + "grad_norm": 3.7522876262664795, + "learning_rate": 9.071187163679773e-06, + "loss": 0.2116, + "step": 36832 + }, + { + "epoch": 0.8634235247923956, + "grad_norm": 3.156179189682007, + "learning_rate": 9.068121857437007e-06, + "loss": 0.2463, + "step": 36833 + }, + { + "epoch": 0.8634469663672005, + "grad_norm": 3.3762032985687256, + "learning_rate": 9.06505704459315e-06, + "loss": 1.7329, + "step": 36834 + }, + { + "epoch": 0.8634704079420056, + "grad_norm": 3.220205068588257, + "learning_rate": 9.061992725164858e-06, + "loss": 0.1212, + "step": 36835 + }, + { + "epoch": 0.8634938495168105, + "grad_norm": 1.6427972316741943, + "learning_rate": 9.05892889916876e-06, + "loss": 0.0535, + "step": 36836 + }, + { + "epoch": 0.8635172910916156, + "grad_norm": 2.8655812740325928, + "learning_rate": 9.055865566621469e-06, + "loss": 0.1999, + "step": 36837 + }, + { + "epoch": 0.8635407326664205, + "grad_norm": 4.362773895263672, + "learning_rate": 9.05280272753961e-06, + "loss": 0.1437, + "step": 36838 + }, + { + "epoch": 0.8635641742412256, + "grad_norm": 1.3434432744979858, + "learning_rate": 9.049740381939787e-06, + "loss": 0.0558, + "step": 36839 + }, + { + "epoch": 0.8635876158160305, + "grad_norm": 1.1772658824920654, + "learning_rate": 9.046678529838637e-06, + "loss": 0.5396, + "step": 36840 + }, + { + "epoch": 0.8636110573908355, + "grad_norm": 2.7652909755706787, + "learning_rate": 9.043617171252771e-06, + "loss": 0.2771, + "step": 36841 + }, + { + "epoch": 0.8636344989656405, + "grad_norm": 2.594717264175415, + "learning_rate": 9.040556306198777e-06, + "loss": 1.4022, + "step": 36842 + }, + { + "epoch": 0.8636579405404455, + "grad_norm": 0.7018815279006958, + "learning_rate": 9.037495934693296e-06, + "loss": 0.0508, + "step": 36843 + }, + { + "epoch": 0.8636813821152505, + "grad_norm": 1.6664677858352661, + "learning_rate": 9.034436056752927e-06, + "loss": 0.127, + "step": 36844 + }, + { + "epoch": 0.8637048236900555, + "grad_norm": 1.864143967628479, + "learning_rate": 9.031376672394265e-06, + "loss": 0.2753, + "step": 36845 + }, + { + "epoch": 0.8637282652648605, + "grad_norm": 3.4336421489715576, + "learning_rate": 9.028317781633889e-06, + "loss": 0.2804, + "step": 36846 + }, + { + "epoch": 0.8637517068396655, + "grad_norm": 1.1882835626602173, + "learning_rate": 9.025259384488439e-06, + "loss": 0.0955, + "step": 36847 + }, + { + "epoch": 0.8637751484144705, + "grad_norm": 2.073601007461548, + "learning_rate": 9.02220148097449e-06, + "loss": 0.1706, + "step": 36848 + }, + { + "epoch": 0.8637985899892755, + "grad_norm": 3.109504461288452, + "learning_rate": 9.019144071108631e-06, + "loss": 0.3086, + "step": 36849 + }, + { + "epoch": 0.8638220315640804, + "grad_norm": 4.9728169441223145, + "learning_rate": 9.01608715490746e-06, + "loss": 0.3498, + "step": 36850 + }, + { + "epoch": 0.8638454731388855, + "grad_norm": 0.4342491924762726, + "learning_rate": 9.013030732387562e-06, + "loss": 0.0408, + "step": 36851 + }, + { + "epoch": 0.8638689147136904, + "grad_norm": 4.245776176452637, + "learning_rate": 9.00997480356549e-06, + "loss": 0.3701, + "step": 36852 + }, + { + "epoch": 0.8638923562884955, + "grad_norm": 2.0538406372070312, + "learning_rate": 9.006919368457867e-06, + "loss": 0.1838, + "step": 36853 + }, + { + "epoch": 0.8639157978633004, + "grad_norm": 2.362327814102173, + "learning_rate": 9.003864427081254e-06, + "loss": 1.7935, + "step": 36854 + }, + { + "epoch": 0.8639392394381055, + "grad_norm": 2.7987611293792725, + "learning_rate": 9.00080997945224e-06, + "loss": 0.1722, + "step": 36855 + }, + { + "epoch": 0.8639626810129104, + "grad_norm": 0.9192147254943848, + "learning_rate": 8.99775602558739e-06, + "loss": 0.0762, + "step": 36856 + }, + { + "epoch": 0.8639861225877155, + "grad_norm": 2.896959066390991, + "learning_rate": 8.994702565503278e-06, + "loss": 0.159, + "step": 36857 + }, + { + "epoch": 0.8640095641625204, + "grad_norm": 1.9148141145706177, + "learning_rate": 8.991649599216467e-06, + "loss": 0.1956, + "step": 36858 + }, + { + "epoch": 0.8640330057373254, + "grad_norm": 2.397688150405884, + "learning_rate": 8.988597126743502e-06, + "loss": 0.0761, + "step": 36859 + }, + { + "epoch": 0.8640564473121304, + "grad_norm": 2.202589511871338, + "learning_rate": 8.985545148100994e-06, + "loss": 1.5558, + "step": 36860 + }, + { + "epoch": 0.8640798888869354, + "grad_norm": 4.087432384490967, + "learning_rate": 8.982493663305468e-06, + "loss": 0.5411, + "step": 36861 + }, + { + "epoch": 0.8641033304617404, + "grad_norm": 4.669478893280029, + "learning_rate": 8.979442672373506e-06, + "loss": 0.3343, + "step": 36862 + }, + { + "epoch": 0.8641267720365454, + "grad_norm": 3.9785544872283936, + "learning_rate": 8.976392175321635e-06, + "loss": 0.4179, + "step": 36863 + }, + { + "epoch": 0.8641502136113504, + "grad_norm": 2.7246506214141846, + "learning_rate": 8.973342172166422e-06, + "loss": 0.1868, + "step": 36864 + }, + { + "epoch": 0.8641736551861554, + "grad_norm": 2.723022937774658, + "learning_rate": 8.970292662924396e-06, + "loss": 0.2265, + "step": 36865 + }, + { + "epoch": 0.8641970967609603, + "grad_norm": 0.8079411387443542, + "learning_rate": 8.967243647612122e-06, + "loss": 0.0472, + "step": 36866 + }, + { + "epoch": 0.8642205383357654, + "grad_norm": 3.3423001766204834, + "learning_rate": 8.964195126246155e-06, + "loss": 0.3186, + "step": 36867 + }, + { + "epoch": 0.8642439799105703, + "grad_norm": 1.4454964399337769, + "learning_rate": 8.961147098843038e-06, + "loss": 0.1871, + "step": 36868 + }, + { + "epoch": 0.8642674214853754, + "grad_norm": 1.1561750173568726, + "learning_rate": 8.958099565419287e-06, + "loss": 0.0723, + "step": 36869 + }, + { + "epoch": 0.8642908630601804, + "grad_norm": 3.742483139038086, + "learning_rate": 8.955052525991448e-06, + "loss": 0.1971, + "step": 36870 + }, + { + "epoch": 0.8643143046349854, + "grad_norm": 0.40481746196746826, + "learning_rate": 8.95200598057605e-06, + "loss": 0.0342, + "step": 36871 + }, + { + "epoch": 0.8643377462097904, + "grad_norm": 2.34887957572937, + "learning_rate": 8.948959929189615e-06, + "loss": 0.136, + "step": 36872 + }, + { + "epoch": 0.8643611877845954, + "grad_norm": 1.715543270111084, + "learning_rate": 8.945914371848707e-06, + "loss": 0.2301, + "step": 36873 + }, + { + "epoch": 0.8643846293594004, + "grad_norm": 2.458470344543457, + "learning_rate": 8.942869308569824e-06, + "loss": 0.1776, + "step": 36874 + }, + { + "epoch": 0.8644080709342054, + "grad_norm": 1.874023675918579, + "learning_rate": 8.939824739369496e-06, + "loss": 1.2945, + "step": 36875 + }, + { + "epoch": 0.8644315125090104, + "grad_norm": 2.6106886863708496, + "learning_rate": 8.936780664264233e-06, + "loss": 0.2714, + "step": 36876 + }, + { + "epoch": 0.8644549540838153, + "grad_norm": 3.3801114559173584, + "learning_rate": 8.933737083270544e-06, + "loss": 0.323, + "step": 36877 + }, + { + "epoch": 0.8644783956586204, + "grad_norm": 1.9156546592712402, + "learning_rate": 8.930693996404971e-06, + "loss": 1.3596, + "step": 36878 + }, + { + "epoch": 0.8645018372334253, + "grad_norm": 0.7801571488380432, + "learning_rate": 8.927651403684e-06, + "loss": 0.0865, + "step": 36879 + }, + { + "epoch": 0.8645252788082304, + "grad_norm": 3.027287721633911, + "learning_rate": 8.924609305124165e-06, + "loss": 2.2138, + "step": 36880 + }, + { + "epoch": 0.8645487203830353, + "grad_norm": 0.6499428153038025, + "learning_rate": 8.921567700741962e-06, + "loss": 0.0706, + "step": 36881 + }, + { + "epoch": 0.8645721619578404, + "grad_norm": 1.099534034729004, + "learning_rate": 8.918526590553899e-06, + "loss": 0.057, + "step": 36882 + }, + { + "epoch": 0.8645956035326453, + "grad_norm": 2.4959890842437744, + "learning_rate": 8.915485974576477e-06, + "loss": 0.1729, + "step": 36883 + }, + { + "epoch": 0.8646190451074504, + "grad_norm": 1.8405343294143677, + "learning_rate": 8.912445852826157e-06, + "loss": 0.0516, + "step": 36884 + }, + { + "epoch": 0.8646424866822553, + "grad_norm": 0.7724652290344238, + "learning_rate": 8.909406225319495e-06, + "loss": 0.065, + "step": 36885 + }, + { + "epoch": 0.8646659282570603, + "grad_norm": 2.1714749336242676, + "learning_rate": 8.906367092072954e-06, + "loss": 0.2367, + "step": 36886 + }, + { + "epoch": 0.8646893698318653, + "grad_norm": 1.2115263938903809, + "learning_rate": 8.903328453103021e-06, + "loss": 0.0469, + "step": 36887 + }, + { + "epoch": 0.8647128114066703, + "grad_norm": 2.225754737854004, + "learning_rate": 8.900290308426195e-06, + "loss": 0.1165, + "step": 36888 + }, + { + "epoch": 0.8647362529814753, + "grad_norm": 0.7822840809822083, + "learning_rate": 8.897252658058952e-06, + "loss": 0.0622, + "step": 36889 + }, + { + "epoch": 0.8647596945562803, + "grad_norm": 1.097314476966858, + "learning_rate": 8.894215502017766e-06, + "loss": 0.0931, + "step": 36890 + }, + { + "epoch": 0.8647831361310853, + "grad_norm": 3.334643840789795, + "learning_rate": 8.891178840319126e-06, + "loss": 0.212, + "step": 36891 + }, + { + "epoch": 0.8648065777058903, + "grad_norm": 2.578878879547119, + "learning_rate": 8.888142672979527e-06, + "loss": 0.1645, + "step": 36892 + }, + { + "epoch": 0.8648300192806953, + "grad_norm": 2.5826356410980225, + "learning_rate": 8.885107000015424e-06, + "loss": 0.1483, + "step": 36893 + }, + { + "epoch": 0.8648534608555003, + "grad_norm": 2.314207077026367, + "learning_rate": 8.882071821443294e-06, + "loss": 0.0818, + "step": 36894 + }, + { + "epoch": 0.8648769024303052, + "grad_norm": 0.71971195936203, + "learning_rate": 8.879037137279611e-06, + "loss": 0.1012, + "step": 36895 + }, + { + "epoch": 0.8649003440051103, + "grad_norm": 2.920327663421631, + "learning_rate": 8.87600294754083e-06, + "loss": 0.2798, + "step": 36896 + }, + { + "epoch": 0.8649237855799152, + "grad_norm": 4.229564666748047, + "learning_rate": 8.872969252243401e-06, + "loss": 0.4816, + "step": 36897 + }, + { + "epoch": 0.8649472271547203, + "grad_norm": 2.158902406692505, + "learning_rate": 8.869936051403815e-06, + "loss": 0.2357, + "step": 36898 + }, + { + "epoch": 0.8649706687295252, + "grad_norm": 0.9332676529884338, + "learning_rate": 8.866903345038523e-06, + "loss": 0.1536, + "step": 36899 + }, + { + "epoch": 0.8649941103043303, + "grad_norm": 0.8448096513748169, + "learning_rate": 8.863871133163982e-06, + "loss": 0.0657, + "step": 36900 + }, + { + "epoch": 0.8650175518791352, + "grad_norm": 1.0304865837097168, + "learning_rate": 8.86083941579664e-06, + "loss": 0.0749, + "step": 36901 + }, + { + "epoch": 0.8650409934539403, + "grad_norm": 2.249392509460449, + "learning_rate": 8.857808192952932e-06, + "loss": 0.1766, + "step": 36902 + }, + { + "epoch": 0.8650644350287452, + "grad_norm": 0.44884130358695984, + "learning_rate": 8.85477746464931e-06, + "loss": 0.0369, + "step": 36903 + }, + { + "epoch": 0.8650878766035502, + "grad_norm": 2.892824411392212, + "learning_rate": 8.85174723090223e-06, + "loss": 0.2104, + "step": 36904 + }, + { + "epoch": 0.8651113181783552, + "grad_norm": 2.1605424880981445, + "learning_rate": 8.848717491728154e-06, + "loss": 1.5441, + "step": 36905 + }, + { + "epoch": 0.8651347597531602, + "grad_norm": 2.251500368118286, + "learning_rate": 8.845688247143492e-06, + "loss": 0.5534, + "step": 36906 + }, + { + "epoch": 0.8651582013279652, + "grad_norm": 1.9964923858642578, + "learning_rate": 8.842659497164684e-06, + "loss": 0.2094, + "step": 36907 + }, + { + "epoch": 0.8651816429027702, + "grad_norm": 2.823885679244995, + "learning_rate": 8.839631241808178e-06, + "loss": 0.17, + "step": 36908 + }, + { + "epoch": 0.8652050844775752, + "grad_norm": 1.4934660196304321, + "learning_rate": 8.8366034810904e-06, + "loss": 0.1559, + "step": 36909 + }, + { + "epoch": 0.8652285260523802, + "grad_norm": 3.1512272357940674, + "learning_rate": 8.83357621502775e-06, + "loss": 0.188, + "step": 36910 + }, + { + "epoch": 0.8652519676271851, + "grad_norm": 3.126028060913086, + "learning_rate": 8.830549443636693e-06, + "loss": 0.3984, + "step": 36911 + }, + { + "epoch": 0.8652754092019902, + "grad_norm": 2.465610980987549, + "learning_rate": 8.827523166933637e-06, + "loss": 0.1005, + "step": 36912 + }, + { + "epoch": 0.8652988507767951, + "grad_norm": 2.2888870239257812, + "learning_rate": 8.824497384935016e-06, + "loss": 0.1468, + "step": 36913 + }, + { + "epoch": 0.8653222923516002, + "grad_norm": 2.0778164863586426, + "learning_rate": 8.821472097657224e-06, + "loss": 0.1557, + "step": 36914 + }, + { + "epoch": 0.8653457339264051, + "grad_norm": 3.9159200191497803, + "learning_rate": 8.818447305116672e-06, + "loss": 0.2274, + "step": 36915 + }, + { + "epoch": 0.8653691755012102, + "grad_norm": 5.102921009063721, + "learning_rate": 8.815423007329793e-06, + "loss": 0.1881, + "step": 36916 + }, + { + "epoch": 0.8653926170760151, + "grad_norm": 4.9423017501831055, + "learning_rate": 8.812399204313015e-06, + "loss": 0.3832, + "step": 36917 + }, + { + "epoch": 0.8654160586508202, + "grad_norm": 2.199124574661255, + "learning_rate": 8.809375896082717e-06, + "loss": 0.1739, + "step": 36918 + }, + { + "epoch": 0.8654395002256251, + "grad_norm": 2.857461929321289, + "learning_rate": 8.806353082655305e-06, + "loss": 0.2083, + "step": 36919 + }, + { + "epoch": 0.8654629418004302, + "grad_norm": 1.299351692199707, + "learning_rate": 8.803330764047191e-06, + "loss": 0.0578, + "step": 36920 + }, + { + "epoch": 0.8654863833752352, + "grad_norm": 1.958390474319458, + "learning_rate": 8.80030894027477e-06, + "loss": 0.0941, + "step": 36921 + }, + { + "epoch": 0.8655098249500401, + "grad_norm": 9.274782180786133, + "learning_rate": 8.797287611354421e-06, + "loss": 0.3084, + "step": 36922 + }, + { + "epoch": 0.8655332665248452, + "grad_norm": 4.057296276092529, + "learning_rate": 8.794266777302563e-06, + "loss": 0.4526, + "step": 36923 + }, + { + "epoch": 0.8655567080996501, + "grad_norm": 2.8827924728393555, + "learning_rate": 8.79124643813558e-06, + "loss": 0.1884, + "step": 36924 + }, + { + "epoch": 0.8655801496744552, + "grad_norm": 1.5681942701339722, + "learning_rate": 8.788226593869864e-06, + "loss": 0.0533, + "step": 36925 + }, + { + "epoch": 0.8656035912492601, + "grad_norm": 1.318924903869629, + "learning_rate": 8.785207244521797e-06, + "loss": 0.0637, + "step": 36926 + }, + { + "epoch": 0.8656270328240652, + "grad_norm": 2.9737062454223633, + "learning_rate": 8.782188390107737e-06, + "loss": 0.3664, + "step": 36927 + }, + { + "epoch": 0.8656504743988701, + "grad_norm": 3.188904285430908, + "learning_rate": 8.779170030644102e-06, + "loss": 0.2979, + "step": 36928 + }, + { + "epoch": 0.8656739159736752, + "grad_norm": 1.7712130546569824, + "learning_rate": 8.776152166147245e-06, + "loss": 1.1686, + "step": 36929 + }, + { + "epoch": 0.8656973575484801, + "grad_norm": 1.1671983003616333, + "learning_rate": 8.773134796633564e-06, + "loss": 0.0691, + "step": 36930 + }, + { + "epoch": 0.8657207991232851, + "grad_norm": 2.2429287433624268, + "learning_rate": 8.770117922119414e-06, + "loss": 0.193, + "step": 36931 + }, + { + "epoch": 0.8657442406980901, + "grad_norm": 3.3586442470550537, + "learning_rate": 8.76710154262118e-06, + "loss": 0.2598, + "step": 36932 + }, + { + "epoch": 0.8657676822728951, + "grad_norm": 1.4770855903625488, + "learning_rate": 8.764085658155209e-06, + "loss": 0.0468, + "step": 36933 + }, + { + "epoch": 0.8657911238477001, + "grad_norm": 4.545464038848877, + "learning_rate": 8.761070268737882e-06, + "loss": 0.5325, + "step": 36934 + }, + { + "epoch": 0.8658145654225051, + "grad_norm": 1.0717204809188843, + "learning_rate": 8.758055374385532e-06, + "loss": 0.09, + "step": 36935 + }, + { + "epoch": 0.8658380069973101, + "grad_norm": 0.7386196255683899, + "learning_rate": 8.755040975114559e-06, + "loss": 0.0525, + "step": 36936 + }, + { + "epoch": 0.8658614485721151, + "grad_norm": 0.9215503334999084, + "learning_rate": 8.752027070941294e-06, + "loss": 0.1158, + "step": 36937 + }, + { + "epoch": 0.86588489014692, + "grad_norm": 2.1048150062561035, + "learning_rate": 8.7490136618821e-06, + "loss": 0.0947, + "step": 36938 + }, + { + "epoch": 0.8659083317217251, + "grad_norm": 3.895376205444336, + "learning_rate": 8.746000747953309e-06, + "loss": 0.2242, + "step": 36939 + }, + { + "epoch": 0.86593177329653, + "grad_norm": 0.6271817088127136, + "learning_rate": 8.742988329171297e-06, + "loss": 0.0295, + "step": 36940 + }, + { + "epoch": 0.8659552148713351, + "grad_norm": 4.817882537841797, + "learning_rate": 8.739976405552398e-06, + "loss": 0.2891, + "step": 36941 + }, + { + "epoch": 0.86597865644614, + "grad_norm": 2.783439874649048, + "learning_rate": 8.736964977112926e-06, + "loss": 0.3473, + "step": 36942 + }, + { + "epoch": 0.8660020980209451, + "grad_norm": 2.4499900341033936, + "learning_rate": 8.733954043869285e-06, + "loss": 0.2093, + "step": 36943 + }, + { + "epoch": 0.86602553959575, + "grad_norm": 5.048238754272461, + "learning_rate": 8.73094360583776e-06, + "loss": 0.5037, + "step": 36944 + }, + { + "epoch": 0.8660489811705551, + "grad_norm": 3.464174747467041, + "learning_rate": 8.727933663034704e-06, + "loss": 0.2879, + "step": 36945 + }, + { + "epoch": 0.86607242274536, + "grad_norm": 9.326526641845703, + "learning_rate": 8.724924215476448e-06, + "loss": 0.467, + "step": 36946 + }, + { + "epoch": 0.866095864320165, + "grad_norm": 2.737269163131714, + "learning_rate": 8.721915263179325e-06, + "loss": 0.3358, + "step": 36947 + }, + { + "epoch": 0.86611930589497, + "grad_norm": 3.273153305053711, + "learning_rate": 8.718906806159632e-06, + "loss": 0.2807, + "step": 36948 + }, + { + "epoch": 0.866142747469775, + "grad_norm": 4.135514736175537, + "learning_rate": 8.715898844433734e-06, + "loss": 0.4073, + "step": 36949 + }, + { + "epoch": 0.86616618904458, + "grad_norm": 0.7477502822875977, + "learning_rate": 8.71289137801794e-06, + "loss": 0.0451, + "step": 36950 + }, + { + "epoch": 0.866189630619385, + "grad_norm": 3.87675404548645, + "learning_rate": 8.709884406928549e-06, + "loss": 0.2066, + "step": 36951 + }, + { + "epoch": 0.86621307219419, + "grad_norm": 1.9271129369735718, + "learning_rate": 8.7068779311819e-06, + "loss": 0.1123, + "step": 36952 + }, + { + "epoch": 0.866236513768995, + "grad_norm": 1.8502192497253418, + "learning_rate": 8.703871950794306e-06, + "loss": 1.2972, + "step": 36953 + }, + { + "epoch": 0.8662599553438, + "grad_norm": 3.556774854660034, + "learning_rate": 8.700866465782042e-06, + "loss": 0.3609, + "step": 36954 + }, + { + "epoch": 0.866283396918605, + "grad_norm": 1.1769534349441528, + "learning_rate": 8.697861476161472e-06, + "loss": 0.1865, + "step": 36955 + }, + { + "epoch": 0.86630683849341, + "grad_norm": 1.5029324293136597, + "learning_rate": 8.694856981948873e-06, + "loss": 1.0876, + "step": 36956 + }, + { + "epoch": 0.866330280068215, + "grad_norm": 2.2192845344543457, + "learning_rate": 8.691852983160542e-06, + "loss": 0.0948, + "step": 36957 + }, + { + "epoch": 0.8663537216430199, + "grad_norm": 0.8857112526893616, + "learning_rate": 8.688849479812788e-06, + "loss": 0.0758, + "step": 36958 + }, + { + "epoch": 0.866377163217825, + "grad_norm": 1.6022751331329346, + "learning_rate": 8.6858464719219e-06, + "loss": 0.092, + "step": 36959 + }, + { + "epoch": 0.8664006047926299, + "grad_norm": 2.322669744491577, + "learning_rate": 8.68284395950416e-06, + "loss": 0.2637, + "step": 36960 + }, + { + "epoch": 0.866424046367435, + "grad_norm": 1.4016053676605225, + "learning_rate": 8.679841942575895e-06, + "loss": 0.1034, + "step": 36961 + }, + { + "epoch": 0.8664474879422399, + "grad_norm": 1.085272192955017, + "learning_rate": 8.676840421153376e-06, + "loss": 0.1102, + "step": 36962 + }, + { + "epoch": 0.866470929517045, + "grad_norm": 1.522424340248108, + "learning_rate": 8.673839395252892e-06, + "loss": 0.1158, + "step": 36963 + }, + { + "epoch": 0.8664943710918499, + "grad_norm": 3.538698196411133, + "learning_rate": 8.670838864890696e-06, + "loss": 0.2891, + "step": 36964 + }, + { + "epoch": 0.866517812666655, + "grad_norm": 2.526815176010132, + "learning_rate": 8.66783883008312e-06, + "loss": 0.316, + "step": 36965 + }, + { + "epoch": 0.8665412542414599, + "grad_norm": 0.8164492249488831, + "learning_rate": 8.664839290846427e-06, + "loss": 0.0724, + "step": 36966 + }, + { + "epoch": 0.8665646958162649, + "grad_norm": 1.0457087755203247, + "learning_rate": 8.66184024719685e-06, + "loss": 0.1078, + "step": 36967 + }, + { + "epoch": 0.8665881373910699, + "grad_norm": 2.7458369731903076, + "learning_rate": 8.658841699150722e-06, + "loss": 0.1904, + "step": 36968 + }, + { + "epoch": 0.8666115789658749, + "grad_norm": 3.745817184448242, + "learning_rate": 8.655843646724294e-06, + "loss": 0.2446, + "step": 36969 + }, + { + "epoch": 0.8666350205406799, + "grad_norm": 3.4132277965545654, + "learning_rate": 8.652846089933818e-06, + "loss": 0.1996, + "step": 36970 + }, + { + "epoch": 0.8666584621154849, + "grad_norm": 2.519746780395508, + "learning_rate": 8.649849028795565e-06, + "loss": 0.1391, + "step": 36971 + }, + { + "epoch": 0.86668190369029, + "grad_norm": 1.5362848043441772, + "learning_rate": 8.646852463325805e-06, + "loss": 0.1403, + "step": 36972 + }, + { + "epoch": 0.8667053452650949, + "grad_norm": 0.46911656856536865, + "learning_rate": 8.643856393540772e-06, + "loss": 0.0405, + "step": 36973 + }, + { + "epoch": 0.8667287868399, + "grad_norm": 0.3885030746459961, + "learning_rate": 8.640860819456764e-06, + "loss": 0.0222, + "step": 36974 + }, + { + "epoch": 0.8667522284147049, + "grad_norm": 2.91908860206604, + "learning_rate": 8.637865741090012e-06, + "loss": 0.1348, + "step": 36975 + }, + { + "epoch": 0.86677566998951, + "grad_norm": 3.1526129245758057, + "learning_rate": 8.634871158456747e-06, + "loss": 0.3793, + "step": 36976 + }, + { + "epoch": 0.8667991115643149, + "grad_norm": 0.8265628218650818, + "learning_rate": 8.63187707157327e-06, + "loss": 0.0494, + "step": 36977 + }, + { + "epoch": 0.8668225531391199, + "grad_norm": 4.229286193847656, + "learning_rate": 8.628883480455785e-06, + "loss": 0.2927, + "step": 36978 + }, + { + "epoch": 0.8668459947139249, + "grad_norm": 2.1340458393096924, + "learning_rate": 8.625890385120549e-06, + "loss": 1.0668, + "step": 36979 + }, + { + "epoch": 0.8668694362887299, + "grad_norm": 3.136773109436035, + "learning_rate": 8.622897785583784e-06, + "loss": 0.2469, + "step": 36980 + }, + { + "epoch": 0.8668928778635349, + "grad_norm": 3.0868053436279297, + "learning_rate": 8.619905681861762e-06, + "loss": 0.2021, + "step": 36981 + }, + { + "epoch": 0.8669163194383399, + "grad_norm": 4.307802200317383, + "learning_rate": 8.616914073970705e-06, + "loss": 0.3372, + "step": 36982 + }, + { + "epoch": 0.8669397610131448, + "grad_norm": 5.232659339904785, + "learning_rate": 8.613922961926845e-06, + "loss": 0.5442, + "step": 36983 + }, + { + "epoch": 0.8669632025879499, + "grad_norm": 0.36935222148895264, + "learning_rate": 8.610932345746403e-06, + "loss": 0.0343, + "step": 36984 + }, + { + "epoch": 0.8669866441627548, + "grad_norm": 4.219695568084717, + "learning_rate": 8.607942225445597e-06, + "loss": 0.3191, + "step": 36985 + }, + { + "epoch": 0.8670100857375599, + "grad_norm": 2.3116636276245117, + "learning_rate": 8.604952601040683e-06, + "loss": 0.1293, + "step": 36986 + }, + { + "epoch": 0.8670335273123648, + "grad_norm": 3.5231282711029053, + "learning_rate": 8.601963472547859e-06, + "loss": 0.3586, + "step": 36987 + }, + { + "epoch": 0.8670569688871699, + "grad_norm": 1.8913153409957886, + "learning_rate": 8.598974839983342e-06, + "loss": 1.0374, + "step": 36988 + }, + { + "epoch": 0.8670804104619748, + "grad_norm": 1.708818793296814, + "learning_rate": 8.595986703363368e-06, + "loss": 0.0982, + "step": 36989 + }, + { + "epoch": 0.8671038520367799, + "grad_norm": 1.3253669738769531, + "learning_rate": 8.592999062704143e-06, + "loss": 0.1259, + "step": 36990 + }, + { + "epoch": 0.8671272936115848, + "grad_norm": 2.303297758102417, + "learning_rate": 8.590011918021878e-06, + "loss": 0.1373, + "step": 36991 + }, + { + "epoch": 0.8671507351863899, + "grad_norm": 3.1436784267425537, + "learning_rate": 8.58702526933276e-06, + "loss": 0.2146, + "step": 36992 + }, + { + "epoch": 0.8671741767611948, + "grad_norm": 0.5542626976966858, + "learning_rate": 8.584039116653031e-06, + "loss": 0.0397, + "step": 36993 + }, + { + "epoch": 0.8671976183359998, + "grad_norm": 1.130555272102356, + "learning_rate": 8.581053459998879e-06, + "loss": 0.0406, + "step": 36994 + }, + { + "epoch": 0.8672210599108048, + "grad_norm": 1.6118296384811401, + "learning_rate": 8.5780682993865e-06, + "loss": 0.1499, + "step": 36995 + }, + { + "epoch": 0.8672445014856098, + "grad_norm": 2.817913055419922, + "learning_rate": 8.575083634832092e-06, + "loss": 0.2902, + "step": 36996 + }, + { + "epoch": 0.8672679430604148, + "grad_norm": 0.6132343411445618, + "learning_rate": 8.572099466351846e-06, + "loss": 0.0538, + "step": 36997 + }, + { + "epoch": 0.8672913846352198, + "grad_norm": 0.9657545685768127, + "learning_rate": 8.569115793961946e-06, + "loss": 0.0404, + "step": 36998 + }, + { + "epoch": 0.8673148262100248, + "grad_norm": 1.6381866931915283, + "learning_rate": 8.566132617678612e-06, + "loss": 0.1677, + "step": 36999 + }, + { + "epoch": 0.8673382677848298, + "grad_norm": 2.535308599472046, + "learning_rate": 8.563149937517989e-06, + "loss": 0.1198, + "step": 37000 + }, + { + "epoch": 0.8673617093596347, + "grad_norm": 0.3419870436191559, + "learning_rate": 8.560167753496318e-06, + "loss": 0.0193, + "step": 37001 + }, + { + "epoch": 0.8673851509344398, + "grad_norm": 1.2530368566513062, + "learning_rate": 8.55718606562973e-06, + "loss": 0.1049, + "step": 37002 + }, + { + "epoch": 0.8674085925092447, + "grad_norm": 2.930187463760376, + "learning_rate": 8.554204873934435e-06, + "loss": 0.1166, + "step": 37003 + }, + { + "epoch": 0.8674320340840498, + "grad_norm": 3.1491973400115967, + "learning_rate": 8.551224178426587e-06, + "loss": 0.2264, + "step": 37004 + }, + { + "epoch": 0.8674554756588547, + "grad_norm": 1.3166743516921997, + "learning_rate": 8.548243979122362e-06, + "loss": 0.0575, + "step": 37005 + }, + { + "epoch": 0.8674789172336598, + "grad_norm": 2.1748738288879395, + "learning_rate": 8.545264276037945e-06, + "loss": 0.1912, + "step": 37006 + }, + { + "epoch": 0.8675023588084647, + "grad_norm": 3.1672823429107666, + "learning_rate": 8.542285069189493e-06, + "loss": 0.2953, + "step": 37007 + }, + { + "epoch": 0.8675258003832698, + "grad_norm": 2.211132049560547, + "learning_rate": 8.53930635859318e-06, + "loss": 0.1825, + "step": 37008 + }, + { + "epoch": 0.8675492419580747, + "grad_norm": 0.7824918627738953, + "learning_rate": 8.53632814426516e-06, + "loss": 0.0594, + "step": 37009 + }, + { + "epoch": 0.8675726835328798, + "grad_norm": 0.7890198826789856, + "learning_rate": 8.533350426221597e-06, + "loss": 0.0526, + "step": 37010 + }, + { + "epoch": 0.8675961251076847, + "grad_norm": 1.4528836011886597, + "learning_rate": 8.530373204478626e-06, + "loss": 0.0771, + "step": 37011 + }, + { + "epoch": 0.8676195666824897, + "grad_norm": 3.8317058086395264, + "learning_rate": 8.527396479052419e-06, + "loss": 0.2599, + "step": 37012 + }, + { + "epoch": 0.8676430082572947, + "grad_norm": 3.348202705383301, + "learning_rate": 8.524420249959152e-06, + "loss": 0.2547, + "step": 37013 + }, + { + "epoch": 0.8676664498320997, + "grad_norm": 1.7688255310058594, + "learning_rate": 8.52144451721496e-06, + "loss": 0.1624, + "step": 37014 + }, + { + "epoch": 0.8676898914069047, + "grad_norm": 2.1980531215667725, + "learning_rate": 8.518469280835973e-06, + "loss": 0.0933, + "step": 37015 + }, + { + "epoch": 0.8677133329817097, + "grad_norm": 1.9005144834518433, + "learning_rate": 8.515494540838342e-06, + "loss": 0.1259, + "step": 37016 + }, + { + "epoch": 0.8677367745565147, + "grad_norm": 3.309873104095459, + "learning_rate": 8.512520297238213e-06, + "loss": 0.2858, + "step": 37017 + }, + { + "epoch": 0.8677602161313197, + "grad_norm": 2.3697891235351562, + "learning_rate": 8.509546550051706e-06, + "loss": 0.143, + "step": 37018 + }, + { + "epoch": 0.8677836577061246, + "grad_norm": 2.8847265243530273, + "learning_rate": 8.506573299294996e-06, + "loss": 0.29, + "step": 37019 + }, + { + "epoch": 0.8678070992809297, + "grad_norm": 1.2558671236038208, + "learning_rate": 8.50360054498418e-06, + "loss": 0.0706, + "step": 37020 + }, + { + "epoch": 0.8678305408557346, + "grad_norm": 1.7133458852767944, + "learning_rate": 8.500628287135404e-06, + "loss": 0.1802, + "step": 37021 + }, + { + "epoch": 0.8678539824305397, + "grad_norm": 1.5671030282974243, + "learning_rate": 8.497656525764797e-06, + "loss": 0.0965, + "step": 37022 + }, + { + "epoch": 0.8678774240053447, + "grad_norm": 0.5960574746131897, + "learning_rate": 8.494685260888458e-06, + "loss": 0.0369, + "step": 37023 + }, + { + "epoch": 0.8679008655801497, + "grad_norm": 3.934386968612671, + "learning_rate": 8.49171449252254e-06, + "loss": 0.3036, + "step": 37024 + }, + { + "epoch": 0.8679243071549547, + "grad_norm": 0.9029322862625122, + "learning_rate": 8.488744220683143e-06, + "loss": 0.076, + "step": 37025 + }, + { + "epoch": 0.8679477487297597, + "grad_norm": 1.438672423362732, + "learning_rate": 8.485774445386407e-06, + "loss": 0.1042, + "step": 37026 + }, + { + "epoch": 0.8679711903045647, + "grad_norm": 2.612576723098755, + "learning_rate": 8.482805166648433e-06, + "loss": 0.1545, + "step": 37027 + }, + { + "epoch": 0.8679946318793696, + "grad_norm": 2.190727472305298, + "learning_rate": 8.479836384485329e-06, + "loss": 0.1431, + "step": 37028 + }, + { + "epoch": 0.8680180734541747, + "grad_norm": 1.8941779136657715, + "learning_rate": 8.476868098913205e-06, + "loss": 0.0895, + "step": 37029 + }, + { + "epoch": 0.8680415150289796, + "grad_norm": 2.751763105392456, + "learning_rate": 8.473900309948158e-06, + "loss": 0.2768, + "step": 37030 + }, + { + "epoch": 0.8680649566037847, + "grad_norm": 3.6771581172943115, + "learning_rate": 8.470933017606309e-06, + "loss": 0.349, + "step": 37031 + }, + { + "epoch": 0.8680883981785896, + "grad_norm": 3.1657700538635254, + "learning_rate": 8.467966221903756e-06, + "loss": 0.3817, + "step": 37032 + }, + { + "epoch": 0.8681118397533947, + "grad_norm": 1.1360697746276855, + "learning_rate": 8.464999922856597e-06, + "loss": 0.0599, + "step": 37033 + }, + { + "epoch": 0.8681352813281996, + "grad_norm": 1.4924520254135132, + "learning_rate": 8.46203412048091e-06, + "loss": 0.0949, + "step": 37034 + }, + { + "epoch": 0.8681587229030047, + "grad_norm": 2.739072322845459, + "learning_rate": 8.459068814792814e-06, + "loss": 1.2423, + "step": 37035 + }, + { + "epoch": 0.8681821644778096, + "grad_norm": 1.4041446447372437, + "learning_rate": 8.456104005808352e-06, + "loss": 0.0583, + "step": 37036 + }, + { + "epoch": 0.8682056060526147, + "grad_norm": 3.190232038497925, + "learning_rate": 8.453139693543654e-06, + "loss": 0.1408, + "step": 37037 + }, + { + "epoch": 0.8682290476274196, + "grad_norm": 3.6455183029174805, + "learning_rate": 8.45017587801481e-06, + "loss": 0.3717, + "step": 37038 + }, + { + "epoch": 0.8682524892022246, + "grad_norm": 0.6575945615768433, + "learning_rate": 8.447212559237883e-06, + "loss": 0.0305, + "step": 37039 + }, + { + "epoch": 0.8682759307770296, + "grad_norm": 0.9440435767173767, + "learning_rate": 8.44424973722896e-06, + "loss": 0.0697, + "step": 37040 + }, + { + "epoch": 0.8682993723518346, + "grad_norm": 3.866265296936035, + "learning_rate": 8.441287412004106e-06, + "loss": 0.257, + "step": 37041 + }, + { + "epoch": 0.8683228139266396, + "grad_norm": 1.2292757034301758, + "learning_rate": 8.438325583579409e-06, + "loss": 0.0602, + "step": 37042 + }, + { + "epoch": 0.8683462555014446, + "grad_norm": 3.3898403644561768, + "learning_rate": 8.435364251970911e-06, + "loss": 0.1737, + "step": 37043 + }, + { + "epoch": 0.8683696970762496, + "grad_norm": 4.579004764556885, + "learning_rate": 8.432403417194712e-06, + "loss": 0.3235, + "step": 37044 + }, + { + "epoch": 0.8683931386510546, + "grad_norm": 0.7326071262359619, + "learning_rate": 8.429443079266875e-06, + "loss": 0.0189, + "step": 37045 + }, + { + "epoch": 0.8684165802258595, + "grad_norm": 0.8482064008712769, + "learning_rate": 8.426483238203452e-06, + "loss": 0.0774, + "step": 37046 + }, + { + "epoch": 0.8684400218006646, + "grad_norm": 0.6501697301864624, + "learning_rate": 8.423523894020502e-06, + "loss": 0.0397, + "step": 37047 + }, + { + "epoch": 0.8684634633754695, + "grad_norm": 2.799426794052124, + "learning_rate": 8.420565046734097e-06, + "loss": 0.1774, + "step": 37048 + }, + { + "epoch": 0.8684869049502746, + "grad_norm": 3.026585102081299, + "learning_rate": 8.417606696360247e-06, + "loss": 0.2116, + "step": 37049 + }, + { + "epoch": 0.8685103465250795, + "grad_norm": 3.7483108043670654, + "learning_rate": 8.414648842915051e-06, + "loss": 0.3578, + "step": 37050 + }, + { + "epoch": 0.8685337880998846, + "grad_norm": 2.552934169769287, + "learning_rate": 8.411691486414563e-06, + "loss": 0.2371, + "step": 37051 + }, + { + "epoch": 0.8685572296746895, + "grad_norm": 4.04066276550293, + "learning_rate": 8.408734626874804e-06, + "loss": 0.2807, + "step": 37052 + }, + { + "epoch": 0.8685806712494946, + "grad_norm": 0.3388349413871765, + "learning_rate": 8.405778264311837e-06, + "loss": 0.0192, + "step": 37053 + }, + { + "epoch": 0.8686041128242995, + "grad_norm": 2.483673095703125, + "learning_rate": 8.402822398741684e-06, + "loss": 0.1646, + "step": 37054 + }, + { + "epoch": 0.8686275543991046, + "grad_norm": 2.4931223392486572, + "learning_rate": 8.399867030180398e-06, + "loss": 0.213, + "step": 37055 + }, + { + "epoch": 0.8686509959739095, + "grad_norm": 2.6982476711273193, + "learning_rate": 8.396912158643977e-06, + "loss": 0.3605, + "step": 37056 + }, + { + "epoch": 0.8686744375487145, + "grad_norm": 1.3811436891555786, + "learning_rate": 8.39395778414852e-06, + "loss": 0.1057, + "step": 37057 + }, + { + "epoch": 0.8686978791235195, + "grad_norm": 1.1433244943618774, + "learning_rate": 8.391003906710016e-06, + "loss": 0.1026, + "step": 37058 + }, + { + "epoch": 0.8687213206983245, + "grad_norm": 3.1293063163757324, + "learning_rate": 8.388050526344505e-06, + "loss": 0.1935, + "step": 37059 + }, + { + "epoch": 0.8687447622731295, + "grad_norm": 1.0187791585922241, + "learning_rate": 8.385097643068007e-06, + "loss": 0.0738, + "step": 37060 + }, + { + "epoch": 0.8687682038479345, + "grad_norm": 2.3288826942443848, + "learning_rate": 8.382145256896523e-06, + "loss": 0.2124, + "step": 37061 + }, + { + "epoch": 0.8687916454227395, + "grad_norm": 1.1293727159500122, + "learning_rate": 8.379193367846116e-06, + "loss": 0.1724, + "step": 37062 + }, + { + "epoch": 0.8688150869975445, + "grad_norm": 3.1143970489501953, + "learning_rate": 8.376241975932764e-06, + "loss": 0.3012, + "step": 37063 + }, + { + "epoch": 0.8688385285723494, + "grad_norm": 2.806760311126709, + "learning_rate": 8.373291081172518e-06, + "loss": 0.121, + "step": 37064 + }, + { + "epoch": 0.8688619701471545, + "grad_norm": 3.8342177867889404, + "learning_rate": 8.370340683581368e-06, + "loss": 0.3047, + "step": 37065 + }, + { + "epoch": 0.8688854117219594, + "grad_norm": 2.069524049758911, + "learning_rate": 8.367390783175333e-06, + "loss": 0.1423, + "step": 37066 + }, + { + "epoch": 0.8689088532967645, + "grad_norm": 3.017185926437378, + "learning_rate": 8.364441379970411e-06, + "loss": 0.1453, + "step": 37067 + }, + { + "epoch": 0.8689322948715694, + "grad_norm": 0.32183384895324707, + "learning_rate": 8.361492473982579e-06, + "loss": 0.0268, + "step": 37068 + }, + { + "epoch": 0.8689557364463745, + "grad_norm": 3.357635498046875, + "learning_rate": 8.358544065227891e-06, + "loss": 0.1536, + "step": 37069 + }, + { + "epoch": 0.8689791780211794, + "grad_norm": 0.9217638969421387, + "learning_rate": 8.355596153722323e-06, + "loss": 0.0664, + "step": 37070 + }, + { + "epoch": 0.8690026195959845, + "grad_norm": 0.6524062752723694, + "learning_rate": 8.35264873948186e-06, + "loss": 0.0189, + "step": 37071 + }, + { + "epoch": 0.8690260611707894, + "grad_norm": 3.7415826320648193, + "learning_rate": 8.349701822522504e-06, + "loss": 0.3805, + "step": 37072 + }, + { + "epoch": 0.8690495027455944, + "grad_norm": 3.1576366424560547, + "learning_rate": 8.346755402860218e-06, + "loss": 0.3706, + "step": 37073 + }, + { + "epoch": 0.8690729443203995, + "grad_norm": 1.6350761651992798, + "learning_rate": 8.343809480511045e-06, + "loss": 0.087, + "step": 37074 + }, + { + "epoch": 0.8690963858952044, + "grad_norm": 1.4194810390472412, + "learning_rate": 8.340864055490916e-06, + "loss": 0.1373, + "step": 37075 + }, + { + "epoch": 0.8691198274700095, + "grad_norm": 5.158720016479492, + "learning_rate": 8.337919127815853e-06, + "loss": 0.276, + "step": 37076 + }, + { + "epoch": 0.8691432690448144, + "grad_norm": 1.480072021484375, + "learning_rate": 8.334974697501818e-06, + "loss": 0.1295, + "step": 37077 + }, + { + "epoch": 0.8691667106196195, + "grad_norm": 2.584962844848633, + "learning_rate": 8.33203076456478e-06, + "loss": 0.1525, + "step": 37078 + }, + { + "epoch": 0.8691901521944244, + "grad_norm": 0.8764552474021912, + "learning_rate": 8.329087329020735e-06, + "loss": 0.0802, + "step": 37079 + }, + { + "epoch": 0.8692135937692295, + "grad_norm": 1.0673785209655762, + "learning_rate": 8.326144390885626e-06, + "loss": 0.0906, + "step": 37080 + }, + { + "epoch": 0.8692370353440344, + "grad_norm": 2.8958051204681396, + "learning_rate": 8.32320195017543e-06, + "loss": 0.4154, + "step": 37081 + }, + { + "epoch": 0.8692604769188395, + "grad_norm": 2.6925504207611084, + "learning_rate": 8.320260006906122e-06, + "loss": 0.1094, + "step": 37082 + }, + { + "epoch": 0.8692839184936444, + "grad_norm": 1.0145052671432495, + "learning_rate": 8.31731856109368e-06, + "loss": 0.1021, + "step": 37083 + }, + { + "epoch": 0.8693073600684494, + "grad_norm": 0.6363232135772705, + "learning_rate": 8.314377612754032e-06, + "loss": 0.0788, + "step": 37084 + }, + { + "epoch": 0.8693308016432544, + "grad_norm": 2.317082166671753, + "learning_rate": 8.311437161903124e-06, + "loss": 1.213, + "step": 37085 + }, + { + "epoch": 0.8693542432180594, + "grad_norm": 0.2614709138870239, + "learning_rate": 8.308497208556964e-06, + "loss": 0.0194, + "step": 37086 + }, + { + "epoch": 0.8693776847928644, + "grad_norm": 1.9178460836410522, + "learning_rate": 8.305557752731485e-06, + "loss": 0.0967, + "step": 37087 + }, + { + "epoch": 0.8694011263676694, + "grad_norm": 0.5406562089920044, + "learning_rate": 8.302618794442596e-06, + "loss": 0.0271, + "step": 37088 + }, + { + "epoch": 0.8694245679424744, + "grad_norm": 3.96734881401062, + "learning_rate": 8.299680333706295e-06, + "loss": 0.2571, + "step": 37089 + }, + { + "epoch": 0.8694480095172794, + "grad_norm": 3.7089998722076416, + "learning_rate": 8.296742370538513e-06, + "loss": 0.3031, + "step": 37090 + }, + { + "epoch": 0.8694714510920843, + "grad_norm": 2.73665714263916, + "learning_rate": 8.293804904955182e-06, + "loss": 0.2044, + "step": 37091 + }, + { + "epoch": 0.8694948926668894, + "grad_norm": 0.5607686638832092, + "learning_rate": 8.290867936972246e-06, + "loss": 0.0401, + "step": 37092 + }, + { + "epoch": 0.8695183342416943, + "grad_norm": 2.7045395374298096, + "learning_rate": 8.287931466605636e-06, + "loss": 0.2483, + "step": 37093 + }, + { + "epoch": 0.8695417758164994, + "grad_norm": 0.3371833264827728, + "learning_rate": 8.284995493871273e-06, + "loss": 0.0263, + "step": 37094 + }, + { + "epoch": 0.8695652173913043, + "grad_norm": 0.6648495197296143, + "learning_rate": 8.282060018785121e-06, + "loss": 0.045, + "step": 37095 + }, + { + "epoch": 0.8695886589661094, + "grad_norm": 0.33632221817970276, + "learning_rate": 8.279125041363089e-06, + "loss": 0.0197, + "step": 37096 + }, + { + "epoch": 0.8696121005409143, + "grad_norm": 0.4052259027957916, + "learning_rate": 8.276190561621089e-06, + "loss": 0.0622, + "step": 37097 + }, + { + "epoch": 0.8696355421157194, + "grad_norm": 1.9752652645111084, + "learning_rate": 8.273256579575073e-06, + "loss": 1.2915, + "step": 37098 + }, + { + "epoch": 0.8696589836905243, + "grad_norm": 2.0901200771331787, + "learning_rate": 8.27032309524094e-06, + "loss": 0.1105, + "step": 37099 + }, + { + "epoch": 0.8696824252653294, + "grad_norm": 2.999213695526123, + "learning_rate": 8.26739010863462e-06, + "loss": 0.3132, + "step": 37100 + }, + { + "epoch": 0.8697058668401343, + "grad_norm": 2.172553300857544, + "learning_rate": 8.264457619772004e-06, + "loss": 0.1135, + "step": 37101 + }, + { + "epoch": 0.8697293084149393, + "grad_norm": 2.3158907890319824, + "learning_rate": 8.261525628669043e-06, + "loss": 0.1358, + "step": 37102 + }, + { + "epoch": 0.8697527499897443, + "grad_norm": 1.6508269309997559, + "learning_rate": 8.258594135341612e-06, + "loss": 0.054, + "step": 37103 + }, + { + "epoch": 0.8697761915645493, + "grad_norm": 0.6665193438529968, + "learning_rate": 8.255663139805636e-06, + "loss": 0.0297, + "step": 37104 + }, + { + "epoch": 0.8697996331393543, + "grad_norm": 1.549036979675293, + "learning_rate": 8.25273264207701e-06, + "loss": 0.1213, + "step": 37105 + }, + { + "epoch": 0.8698230747141593, + "grad_norm": 2.72570538520813, + "learning_rate": 8.24980264217161e-06, + "loss": 0.2779, + "step": 37106 + }, + { + "epoch": 0.8698465162889643, + "grad_norm": 2.4103431701660156, + "learning_rate": 8.246873140105394e-06, + "loss": 0.2411, + "step": 37107 + }, + { + "epoch": 0.8698699578637693, + "grad_norm": 3.316869020462036, + "learning_rate": 8.24394413589421e-06, + "loss": 0.29, + "step": 37108 + }, + { + "epoch": 0.8698933994385742, + "grad_norm": 2.224168300628662, + "learning_rate": 8.241015629553972e-06, + "loss": 0.1219, + "step": 37109 + }, + { + "epoch": 0.8699168410133793, + "grad_norm": 0.5074768662452698, + "learning_rate": 8.238087621100543e-06, + "loss": 0.0513, + "step": 37110 + }, + { + "epoch": 0.8699402825881842, + "grad_norm": 2.794816017150879, + "learning_rate": 8.235160110549856e-06, + "loss": 0.2464, + "step": 37111 + }, + { + "epoch": 0.8699637241629893, + "grad_norm": 4.158167362213135, + "learning_rate": 8.232233097917763e-06, + "loss": 0.4187, + "step": 37112 + }, + { + "epoch": 0.8699871657377942, + "grad_norm": 2.4772629737854004, + "learning_rate": 8.22930658322013e-06, + "loss": 1.5019, + "step": 37113 + }, + { + "epoch": 0.8700106073125993, + "grad_norm": 2.015510320663452, + "learning_rate": 8.22638056647289e-06, + "loss": 0.1551, + "step": 37114 + }, + { + "epoch": 0.8700340488874042, + "grad_norm": 1.6930499076843262, + "learning_rate": 8.223455047691886e-06, + "loss": 0.0849, + "step": 37115 + }, + { + "epoch": 0.8700574904622093, + "grad_norm": 1.1306685209274292, + "learning_rate": 8.220530026892992e-06, + "loss": 0.0888, + "step": 37116 + }, + { + "epoch": 0.8700809320370142, + "grad_norm": 3.5758347511291504, + "learning_rate": 8.217605504092085e-06, + "loss": 0.2269, + "step": 37117 + }, + { + "epoch": 0.8701043736118192, + "grad_norm": 2.7489101886749268, + "learning_rate": 8.21468147930503e-06, + "loss": 0.1989, + "step": 37118 + }, + { + "epoch": 0.8701278151866242, + "grad_norm": 2.7248048782348633, + "learning_rate": 8.211757952547683e-06, + "loss": 1.4484, + "step": 37119 + }, + { + "epoch": 0.8701512567614292, + "grad_norm": 3.9405527114868164, + "learning_rate": 8.208834923835928e-06, + "loss": 0.4896, + "step": 37120 + }, + { + "epoch": 0.8701746983362342, + "grad_norm": 0.8746351003646851, + "learning_rate": 8.205912393185611e-06, + "loss": 0.0985, + "step": 37121 + }, + { + "epoch": 0.8701981399110392, + "grad_norm": 3.121476173400879, + "learning_rate": 8.202990360612583e-06, + "loss": 0.2358, + "step": 37122 + }, + { + "epoch": 0.8702215814858442, + "grad_norm": 1.893315315246582, + "learning_rate": 8.200068826132733e-06, + "loss": 0.8348, + "step": 37123 + }, + { + "epoch": 0.8702450230606492, + "grad_norm": 1.4297096729278564, + "learning_rate": 8.197147789761884e-06, + "loss": 0.0812, + "step": 37124 + }, + { + "epoch": 0.8702684646354543, + "grad_norm": 3.2367584705352783, + "learning_rate": 8.194227251515896e-06, + "loss": 0.2833, + "step": 37125 + }, + { + "epoch": 0.8702919062102592, + "grad_norm": 3.125415325164795, + "learning_rate": 8.191307211410592e-06, + "loss": 0.2776, + "step": 37126 + }, + { + "epoch": 0.8703153477850643, + "grad_norm": 0.6469566226005554, + "learning_rate": 8.18838766946186e-06, + "loss": 0.0416, + "step": 37127 + }, + { + "epoch": 0.8703387893598692, + "grad_norm": 0.6265138983726501, + "learning_rate": 8.18546862568551e-06, + "loss": 0.0447, + "step": 37128 + }, + { + "epoch": 0.8703622309346742, + "grad_norm": 1.3737657070159912, + "learning_rate": 8.182550080097396e-06, + "loss": 0.1058, + "step": 37129 + }, + { + "epoch": 0.8703856725094792, + "grad_norm": 1.8566172122955322, + "learning_rate": 8.179632032713335e-06, + "loss": 0.2127, + "step": 37130 + }, + { + "epoch": 0.8704091140842842, + "grad_norm": 2.2725791931152344, + "learning_rate": 8.176714483549186e-06, + "loss": 0.1611, + "step": 37131 + }, + { + "epoch": 0.8704325556590892, + "grad_norm": 0.8132784366607666, + "learning_rate": 8.173797432620744e-06, + "loss": 0.0804, + "step": 37132 + }, + { + "epoch": 0.8704559972338942, + "grad_norm": 2.1849205493927, + "learning_rate": 8.170880879943876e-06, + "loss": 0.1452, + "step": 37133 + }, + { + "epoch": 0.8704794388086992, + "grad_norm": 0.823988676071167, + "learning_rate": 8.167964825534369e-06, + "loss": 0.1205, + "step": 37134 + }, + { + "epoch": 0.8705028803835042, + "grad_norm": 3.7068614959716797, + "learning_rate": 8.165049269408087e-06, + "loss": 0.3362, + "step": 37135 + }, + { + "epoch": 0.8705263219583091, + "grad_norm": 2.8836495876312256, + "learning_rate": 8.16213421158083e-06, + "loss": 0.0874, + "step": 37136 + }, + { + "epoch": 0.8705497635331142, + "grad_norm": 2.7664847373962402, + "learning_rate": 8.159219652068418e-06, + "loss": 0.1782, + "step": 37137 + }, + { + "epoch": 0.8705732051079191, + "grad_norm": 4.426172256469727, + "learning_rate": 8.15630559088666e-06, + "loss": 0.2576, + "step": 37138 + }, + { + "epoch": 0.8705966466827242, + "grad_norm": 4.110998630523682, + "learning_rate": 8.153392028051355e-06, + "loss": 0.518, + "step": 37139 + }, + { + "epoch": 0.8706200882575291, + "grad_norm": 0.613105833530426, + "learning_rate": 8.150478963578346e-06, + "loss": 0.0523, + "step": 37140 + }, + { + "epoch": 0.8706435298323342, + "grad_norm": 2.432980537414551, + "learning_rate": 8.147566397483409e-06, + "loss": 0.2729, + "step": 37141 + }, + { + "epoch": 0.8706669714071391, + "grad_norm": 3.0899300575256348, + "learning_rate": 8.144654329782375e-06, + "loss": 0.3273, + "step": 37142 + }, + { + "epoch": 0.8706904129819442, + "grad_norm": 2.3391060829162598, + "learning_rate": 8.141742760491022e-06, + "loss": 1.4496, + "step": 37143 + }, + { + "epoch": 0.8707138545567491, + "grad_norm": 2.831493854522705, + "learning_rate": 8.138831689625137e-06, + "loss": 0.2388, + "step": 37144 + }, + { + "epoch": 0.8707372961315542, + "grad_norm": 2.660407304763794, + "learning_rate": 8.135921117200551e-06, + "loss": 0.2094, + "step": 37145 + }, + { + "epoch": 0.8707607377063591, + "grad_norm": 2.2212765216827393, + "learning_rate": 8.133011043233019e-06, + "loss": 0.2107, + "step": 37146 + }, + { + "epoch": 0.8707841792811641, + "grad_norm": 3.3579766750335693, + "learning_rate": 8.130101467738383e-06, + "loss": 0.1409, + "step": 37147 + }, + { + "epoch": 0.8708076208559691, + "grad_norm": 3.005739688873291, + "learning_rate": 8.127192390732386e-06, + "loss": 0.1156, + "step": 37148 + }, + { + "epoch": 0.8708310624307741, + "grad_norm": 2.5103297233581543, + "learning_rate": 8.124283812230838e-06, + "loss": 0.2011, + "step": 37149 + }, + { + "epoch": 0.8708545040055791, + "grad_norm": 0.7848950028419495, + "learning_rate": 8.121375732249503e-06, + "loss": 0.0613, + "step": 37150 + }, + { + "epoch": 0.8708779455803841, + "grad_norm": 3.464203119277954, + "learning_rate": 8.118468150804148e-06, + "loss": 0.2692, + "step": 37151 + }, + { + "epoch": 0.870901387155189, + "grad_norm": 3.3491814136505127, + "learning_rate": 8.115561067910583e-06, + "loss": 0.2846, + "step": 37152 + }, + { + "epoch": 0.8709248287299941, + "grad_norm": 2.3231093883514404, + "learning_rate": 8.11265448358457e-06, + "loss": 1.1122, + "step": 37153 + }, + { + "epoch": 0.870948270304799, + "grad_norm": 1.712003469467163, + "learning_rate": 8.109748397841888e-06, + "loss": 0.1386, + "step": 37154 + }, + { + "epoch": 0.8709717118796041, + "grad_norm": 2.1093380451202393, + "learning_rate": 8.106842810698278e-06, + "loss": 0.1587, + "step": 37155 + }, + { + "epoch": 0.870995153454409, + "grad_norm": 1.1836192607879639, + "learning_rate": 8.10393772216953e-06, + "loss": 0.086, + "step": 37156 + }, + { + "epoch": 0.8710185950292141, + "grad_norm": 1.4286556243896484, + "learning_rate": 8.101033132271385e-06, + "loss": 0.0841, + "step": 37157 + }, + { + "epoch": 0.871042036604019, + "grad_norm": 4.5702900886535645, + "learning_rate": 8.09812904101962e-06, + "loss": 0.2271, + "step": 37158 + }, + { + "epoch": 0.8710654781788241, + "grad_norm": 2.33858060836792, + "learning_rate": 8.09522544843e-06, + "loss": 1.7713, + "step": 37159 + }, + { + "epoch": 0.871088919753629, + "grad_norm": 2.40921950340271, + "learning_rate": 8.092322354518278e-06, + "loss": 0.2041, + "step": 37160 + }, + { + "epoch": 0.8711123613284341, + "grad_norm": 0.7308445572853088, + "learning_rate": 8.089419759300198e-06, + "loss": 0.0805, + "step": 37161 + }, + { + "epoch": 0.871135802903239, + "grad_norm": 1.1641972064971924, + "learning_rate": 8.086517662791503e-06, + "loss": 0.1327, + "step": 37162 + }, + { + "epoch": 0.871159244478044, + "grad_norm": 2.5754971504211426, + "learning_rate": 8.08361606500796e-06, + "loss": 0.29, + "step": 37163 + }, + { + "epoch": 0.871182686052849, + "grad_norm": 0.7147566676139832, + "learning_rate": 8.080714965965274e-06, + "loss": 0.0386, + "step": 37164 + }, + { + "epoch": 0.871206127627654, + "grad_norm": 2.7286148071289062, + "learning_rate": 8.077814365679237e-06, + "loss": 1.8419, + "step": 37165 + }, + { + "epoch": 0.871229569202459, + "grad_norm": 3.0579044818878174, + "learning_rate": 8.074914264165556e-06, + "loss": 0.2284, + "step": 37166 + }, + { + "epoch": 0.871253010777264, + "grad_norm": 2.6250298023223877, + "learning_rate": 8.072014661439987e-06, + "loss": 0.116, + "step": 37167 + }, + { + "epoch": 0.871276452352069, + "grad_norm": 1.3404744863510132, + "learning_rate": 8.069115557518236e-06, + "loss": 0.0949, + "step": 37168 + }, + { + "epoch": 0.871299893926874, + "grad_norm": 2.915177583694458, + "learning_rate": 8.066216952416061e-06, + "loss": 1.2823, + "step": 37169 + }, + { + "epoch": 0.871323335501679, + "grad_norm": 0.7334538698196411, + "learning_rate": 8.063318846149159e-06, + "loss": 0.036, + "step": 37170 + }, + { + "epoch": 0.871346777076484, + "grad_norm": 3.1321969032287598, + "learning_rate": 8.060421238733262e-06, + "loss": 0.2249, + "step": 37171 + }, + { + "epoch": 0.8713702186512889, + "grad_norm": 2.452690362930298, + "learning_rate": 8.057524130184135e-06, + "loss": 0.1107, + "step": 37172 + }, + { + "epoch": 0.871393660226094, + "grad_norm": 2.181321382522583, + "learning_rate": 8.054627520517467e-06, + "loss": 1.3427, + "step": 37173 + }, + { + "epoch": 0.8714171018008989, + "grad_norm": 2.3447611331939697, + "learning_rate": 8.051731409748964e-06, + "loss": 0.1328, + "step": 37174 + }, + { + "epoch": 0.871440543375704, + "grad_norm": 2.349602699279785, + "learning_rate": 8.048835797894361e-06, + "loss": 0.113, + "step": 37175 + }, + { + "epoch": 0.871463984950509, + "grad_norm": 2.209038734436035, + "learning_rate": 8.045940684969356e-06, + "loss": 0.1436, + "step": 37176 + }, + { + "epoch": 0.871487426525314, + "grad_norm": 2.222672939300537, + "learning_rate": 8.043046070989647e-06, + "loss": 0.2489, + "step": 37177 + }, + { + "epoch": 0.871510868100119, + "grad_norm": 0.5414958596229553, + "learning_rate": 8.040151955970977e-06, + "loss": 0.0363, + "step": 37178 + }, + { + "epoch": 0.871534309674924, + "grad_norm": 2.828458786010742, + "learning_rate": 8.037258339929021e-06, + "loss": 1.3753, + "step": 37179 + }, + { + "epoch": 0.871557751249729, + "grad_norm": 2.8131232261657715, + "learning_rate": 8.034365222879491e-06, + "loss": 0.1795, + "step": 37180 + }, + { + "epoch": 0.871581192824534, + "grad_norm": 2.5692532062530518, + "learning_rate": 8.031472604838087e-06, + "loss": 0.2908, + "step": 37181 + }, + { + "epoch": 0.871604634399339, + "grad_norm": 3.8801679611206055, + "learning_rate": 8.02858048582047e-06, + "loss": 0.3315, + "step": 37182 + }, + { + "epoch": 0.8716280759741439, + "grad_norm": 2.0900845527648926, + "learning_rate": 8.025688865842385e-06, + "loss": 0.1904, + "step": 37183 + }, + { + "epoch": 0.871651517548949, + "grad_norm": 4.190168857574463, + "learning_rate": 8.022797744919475e-06, + "loss": 0.2532, + "step": 37184 + }, + { + "epoch": 0.8716749591237539, + "grad_norm": 2.2252793312072754, + "learning_rate": 8.019907123067482e-06, + "loss": 0.2303, + "step": 37185 + }, + { + "epoch": 0.871698400698559, + "grad_norm": 2.6742653846740723, + "learning_rate": 8.017017000302041e-06, + "loss": 0.2681, + "step": 37186 + }, + { + "epoch": 0.8717218422733639, + "grad_norm": 2.5564098358154297, + "learning_rate": 8.014127376638869e-06, + "loss": 0.2615, + "step": 37187 + }, + { + "epoch": 0.871745283848169, + "grad_norm": 0.6384747624397278, + "learning_rate": 8.011238252093611e-06, + "loss": 0.0358, + "step": 37188 + }, + { + "epoch": 0.8717687254229739, + "grad_norm": 0.5371701717376709, + "learning_rate": 8.008349626681954e-06, + "loss": 0.0541, + "step": 37189 + }, + { + "epoch": 0.871792166997779, + "grad_norm": 1.014891505241394, + "learning_rate": 8.005461500419598e-06, + "loss": 0.0656, + "step": 37190 + }, + { + "epoch": 0.8718156085725839, + "grad_norm": 1.9426087141036987, + "learning_rate": 8.002573873322195e-06, + "loss": 0.1391, + "step": 37191 + }, + { + "epoch": 0.8718390501473889, + "grad_norm": 1.0101816654205322, + "learning_rate": 7.9996867454054e-06, + "loss": 0.0512, + "step": 37192 + }, + { + "epoch": 0.8718624917221939, + "grad_norm": 2.8679237365722656, + "learning_rate": 7.9968001166849e-06, + "loss": 1.558, + "step": 37193 + }, + { + "epoch": 0.8718859332969989, + "grad_norm": 2.519415855407715, + "learning_rate": 7.99391398717635e-06, + "loss": 0.1386, + "step": 37194 + }, + { + "epoch": 0.8719093748718039, + "grad_norm": 2.4962666034698486, + "learning_rate": 7.991028356895391e-06, + "loss": 1.6597, + "step": 37195 + }, + { + "epoch": 0.8719328164466089, + "grad_norm": 3.4120914936065674, + "learning_rate": 7.988143225857703e-06, + "loss": 0.1235, + "step": 37196 + }, + { + "epoch": 0.8719562580214139, + "grad_norm": 3.928868532180786, + "learning_rate": 7.985258594078948e-06, + "loss": 0.2715, + "step": 37197 + }, + { + "epoch": 0.8719796995962189, + "grad_norm": 1.1283060312271118, + "learning_rate": 7.98237446157477e-06, + "loss": 0.0454, + "step": 37198 + }, + { + "epoch": 0.8720031411710238, + "grad_norm": 0.6572888493537903, + "learning_rate": 7.979490828360826e-06, + "loss": 0.0246, + "step": 37199 + }, + { + "epoch": 0.8720265827458289, + "grad_norm": 4.636784076690674, + "learning_rate": 7.976607694452742e-06, + "loss": 0.2596, + "step": 37200 + }, + { + "epoch": 0.8720500243206338, + "grad_norm": 0.8932363390922546, + "learning_rate": 7.973725059866166e-06, + "loss": 0.0755, + "step": 37201 + }, + { + "epoch": 0.8720734658954389, + "grad_norm": 6.264013767242432, + "learning_rate": 7.970842924616739e-06, + "loss": 0.1937, + "step": 37202 + }, + { + "epoch": 0.8720969074702438, + "grad_norm": 2.2230818271636963, + "learning_rate": 7.967961288720117e-06, + "loss": 0.2107, + "step": 37203 + }, + { + "epoch": 0.8721203490450489, + "grad_norm": 1.4060975313186646, + "learning_rate": 7.965080152191929e-06, + "loss": 0.0551, + "step": 37204 + }, + { + "epoch": 0.8721437906198538, + "grad_norm": 2.1673874855041504, + "learning_rate": 7.962199515047796e-06, + "loss": 1.0292, + "step": 37205 + }, + { + "epoch": 0.8721672321946589, + "grad_norm": 1.4389283657073975, + "learning_rate": 7.959319377303364e-06, + "loss": 0.0526, + "step": 37206 + }, + { + "epoch": 0.8721906737694638, + "grad_norm": 2.5585858821868896, + "learning_rate": 7.95643973897423e-06, + "loss": 1.7577, + "step": 37207 + }, + { + "epoch": 0.8722141153442688, + "grad_norm": 4.278708457946777, + "learning_rate": 7.953560600076072e-06, + "loss": 0.361, + "step": 37208 + }, + { + "epoch": 0.8722375569190738, + "grad_norm": 1.621334195137024, + "learning_rate": 7.950681960624451e-06, + "loss": 0.1525, + "step": 37209 + }, + { + "epoch": 0.8722609984938788, + "grad_norm": 2.3605566024780273, + "learning_rate": 7.947803820635035e-06, + "loss": 0.1554, + "step": 37210 + }, + { + "epoch": 0.8722844400686838, + "grad_norm": 2.727867603302002, + "learning_rate": 7.944926180123435e-06, + "loss": 0.344, + "step": 37211 + }, + { + "epoch": 0.8723078816434888, + "grad_norm": 0.8034011125564575, + "learning_rate": 7.942049039105248e-06, + "loss": 0.0277, + "step": 37212 + }, + { + "epoch": 0.8723313232182938, + "grad_norm": 2.335103988647461, + "learning_rate": 7.939172397596085e-06, + "loss": 0.1851, + "step": 37213 + }, + { + "epoch": 0.8723547647930988, + "grad_norm": 0.7624934315681458, + "learning_rate": 7.936296255611575e-06, + "loss": 0.0354, + "step": 37214 + }, + { + "epoch": 0.8723782063679038, + "grad_norm": 4.4930219650268555, + "learning_rate": 7.933420613167287e-06, + "loss": 0.4168, + "step": 37215 + }, + { + "epoch": 0.8724016479427088, + "grad_norm": 0.566010594367981, + "learning_rate": 7.93054547027886e-06, + "loss": 0.0374, + "step": 37216 + }, + { + "epoch": 0.8724250895175137, + "grad_norm": 0.8391523361206055, + "learning_rate": 7.927670826961887e-06, + "loss": 0.0227, + "step": 37217 + }, + { + "epoch": 0.8724485310923188, + "grad_norm": 3.5822365283966064, + "learning_rate": 7.92479668323195e-06, + "loss": 1.5544, + "step": 37218 + }, + { + "epoch": 0.8724719726671237, + "grad_norm": 1.2194545269012451, + "learning_rate": 7.92192303910465e-06, + "loss": 0.0832, + "step": 37219 + }, + { + "epoch": 0.8724954142419288, + "grad_norm": 2.374333381652832, + "learning_rate": 7.919049894595598e-06, + "loss": 0.1383, + "step": 37220 + }, + { + "epoch": 0.8725188558167337, + "grad_norm": 3.1589207649230957, + "learning_rate": 7.916177249720357e-06, + "loss": 0.26, + "step": 37221 + }, + { + "epoch": 0.8725422973915388, + "grad_norm": 2.6218297481536865, + "learning_rate": 7.91330510449454e-06, + "loss": 0.2264, + "step": 37222 + }, + { + "epoch": 0.8725657389663437, + "grad_norm": 2.8358304500579834, + "learning_rate": 7.91043345893372e-06, + "loss": 0.1815, + "step": 37223 + }, + { + "epoch": 0.8725891805411488, + "grad_norm": 2.4962005615234375, + "learning_rate": 7.907562313053473e-06, + "loss": 1.5327, + "step": 37224 + }, + { + "epoch": 0.8726126221159537, + "grad_norm": 1.3669933080673218, + "learning_rate": 7.904691666869391e-06, + "loss": 0.0893, + "step": 37225 + }, + { + "epoch": 0.8726360636907587, + "grad_norm": 1.975052833557129, + "learning_rate": 7.901821520397034e-06, + "loss": 0.1174, + "step": 37226 + }, + { + "epoch": 0.8726595052655638, + "grad_norm": 1.1910911798477173, + "learning_rate": 7.898951873651972e-06, + "loss": 0.0569, + "step": 37227 + }, + { + "epoch": 0.8726829468403687, + "grad_norm": 3.1836562156677246, + "learning_rate": 7.8960827266498e-06, + "loss": 1.4831, + "step": 37228 + }, + { + "epoch": 0.8727063884151738, + "grad_norm": 2.1917574405670166, + "learning_rate": 7.893214079406075e-06, + "loss": 0.1109, + "step": 37229 + }, + { + "epoch": 0.8727298299899787, + "grad_norm": 0.6468079090118408, + "learning_rate": 7.89034593193635e-06, + "loss": 0.0357, + "step": 37230 + }, + { + "epoch": 0.8727532715647838, + "grad_norm": 2.8595523834228516, + "learning_rate": 7.887478284256188e-06, + "loss": 0.1846, + "step": 37231 + }, + { + "epoch": 0.8727767131395887, + "grad_norm": 1.3467274904251099, + "learning_rate": 7.884611136381182e-06, + "loss": 0.156, + "step": 37232 + }, + { + "epoch": 0.8728001547143938, + "grad_norm": 2.0480308532714844, + "learning_rate": 7.881744488326859e-06, + "loss": 0.0969, + "step": 37233 + }, + { + "epoch": 0.8728235962891987, + "grad_norm": 3.170449733734131, + "learning_rate": 7.878878340108765e-06, + "loss": 0.1563, + "step": 37234 + }, + { + "epoch": 0.8728470378640037, + "grad_norm": 0.5674164295196533, + "learning_rate": 7.876012691742485e-06, + "loss": 0.0215, + "step": 37235 + }, + { + "epoch": 0.8728704794388087, + "grad_norm": 1.526834487915039, + "learning_rate": 7.873147543243554e-06, + "loss": 0.1385, + "step": 37236 + }, + { + "epoch": 0.8728939210136137, + "grad_norm": 4.228145122528076, + "learning_rate": 7.870282894627512e-06, + "loss": 0.1994, + "step": 37237 + }, + { + "epoch": 0.8729173625884187, + "grad_norm": 4.1629958152771, + "learning_rate": 7.867418745909905e-06, + "loss": 0.415, + "step": 37238 + }, + { + "epoch": 0.8729408041632237, + "grad_norm": 1.5040836334228516, + "learning_rate": 7.864555097106274e-06, + "loss": 0.1242, + "step": 37239 + }, + { + "epoch": 0.8729642457380287, + "grad_norm": 2.226154088973999, + "learning_rate": 7.861691948232141e-06, + "loss": 0.159, + "step": 37240 + }, + { + "epoch": 0.8729876873128337, + "grad_norm": 2.2188570499420166, + "learning_rate": 7.858829299303072e-06, + "loss": 0.2311, + "step": 37241 + }, + { + "epoch": 0.8730111288876387, + "grad_norm": 3.1926658153533936, + "learning_rate": 7.855967150334598e-06, + "loss": 1.8493, + "step": 37242 + }, + { + "epoch": 0.8730345704624437, + "grad_norm": 1.9053456783294678, + "learning_rate": 7.853105501342206e-06, + "loss": 0.0826, + "step": 37243 + }, + { + "epoch": 0.8730580120372486, + "grad_norm": 2.2134006023406982, + "learning_rate": 7.850244352341484e-06, + "loss": 0.1988, + "step": 37244 + }, + { + "epoch": 0.8730814536120537, + "grad_norm": 2.107490301132202, + "learning_rate": 7.84738370334791e-06, + "loss": 0.1596, + "step": 37245 + }, + { + "epoch": 0.8731048951868586, + "grad_norm": 0.8977077603340149, + "learning_rate": 7.844523554377037e-06, + "loss": 0.1237, + "step": 37246 + }, + { + "epoch": 0.8731283367616637, + "grad_norm": 5.011837005615234, + "learning_rate": 7.841663905444351e-06, + "loss": 0.5413, + "step": 37247 + }, + { + "epoch": 0.8731517783364686, + "grad_norm": 2.5956101417541504, + "learning_rate": 7.838804756565398e-06, + "loss": 0.2001, + "step": 37248 + }, + { + "epoch": 0.8731752199112737, + "grad_norm": 3.8242528438568115, + "learning_rate": 7.835946107755698e-06, + "loss": 0.1394, + "step": 37249 + }, + { + "epoch": 0.8731986614860786, + "grad_norm": 0.6523830890655518, + "learning_rate": 7.833087959030727e-06, + "loss": 0.043, + "step": 37250 + }, + { + "epoch": 0.8732221030608837, + "grad_norm": 1.4111697673797607, + "learning_rate": 7.830230310406029e-06, + "loss": 0.0609, + "step": 37251 + }, + { + "epoch": 0.8732455446356886, + "grad_norm": 1.0354849100112915, + "learning_rate": 7.827373161897078e-06, + "loss": 0.0831, + "step": 37252 + }, + { + "epoch": 0.8732689862104936, + "grad_norm": 3.219104290008545, + "learning_rate": 7.824516513519387e-06, + "loss": 1.5979, + "step": 37253 + }, + { + "epoch": 0.8732924277852986, + "grad_norm": 2.4785399436950684, + "learning_rate": 7.821660365288463e-06, + "loss": 0.1451, + "step": 37254 + }, + { + "epoch": 0.8733158693601036, + "grad_norm": 2.890699863433838, + "learning_rate": 7.818804717219796e-06, + "loss": 0.221, + "step": 37255 + }, + { + "epoch": 0.8733393109349086, + "grad_norm": 2.552107334136963, + "learning_rate": 7.815949569328906e-06, + "loss": 0.1368, + "step": 37256 + }, + { + "epoch": 0.8733627525097136, + "grad_norm": 1.4025505781173706, + "learning_rate": 7.81309492163126e-06, + "loss": 0.147, + "step": 37257 + }, + { + "epoch": 0.8733861940845186, + "grad_norm": 1.991646647453308, + "learning_rate": 7.810240774142352e-06, + "loss": 0.1698, + "step": 37258 + }, + { + "epoch": 0.8734096356593236, + "grad_norm": 1.3719902038574219, + "learning_rate": 7.807387126877652e-06, + "loss": 0.0551, + "step": 37259 + }, + { + "epoch": 0.8734330772341286, + "grad_norm": 1.613844394683838, + "learning_rate": 7.80453397985269e-06, + "loss": 0.0905, + "step": 37260 + }, + { + "epoch": 0.8734565188089336, + "grad_norm": 2.6811258792877197, + "learning_rate": 7.801681333082911e-06, + "loss": 0.2933, + "step": 37261 + }, + { + "epoch": 0.8734799603837385, + "grad_norm": 2.220071315765381, + "learning_rate": 7.798829186583801e-06, + "loss": 0.1981, + "step": 37262 + }, + { + "epoch": 0.8735034019585436, + "grad_norm": 2.279909610748291, + "learning_rate": 7.795977540370835e-06, + "loss": 0.1312, + "step": 37263 + }, + { + "epoch": 0.8735268435333485, + "grad_norm": 0.8925409317016602, + "learning_rate": 7.793126394459482e-06, + "loss": 0.0704, + "step": 37264 + }, + { + "epoch": 0.8735502851081536, + "grad_norm": 2.7119057178497314, + "learning_rate": 7.790275748865205e-06, + "loss": 0.1342, + "step": 37265 + }, + { + "epoch": 0.8735737266829585, + "grad_norm": 2.5894360542297363, + "learning_rate": 7.787425603603504e-06, + "loss": 0.1263, + "step": 37266 + }, + { + "epoch": 0.8735971682577636, + "grad_norm": 2.849987506866455, + "learning_rate": 7.784575958689833e-06, + "loss": 0.2739, + "step": 37267 + }, + { + "epoch": 0.8736206098325685, + "grad_norm": 2.626737356185913, + "learning_rate": 7.781726814139612e-06, + "loss": 0.2335, + "step": 37268 + }, + { + "epoch": 0.8736440514073736, + "grad_norm": 3.247518539428711, + "learning_rate": 7.778878169968363e-06, + "loss": 0.2751, + "step": 37269 + }, + { + "epoch": 0.8736674929821785, + "grad_norm": 3.2978618144989014, + "learning_rate": 7.776030026191516e-06, + "loss": 0.454, + "step": 37270 + }, + { + "epoch": 0.8736909345569835, + "grad_norm": 0.4690181016921997, + "learning_rate": 7.773182382824517e-06, + "loss": 0.0309, + "step": 37271 + }, + { + "epoch": 0.8737143761317885, + "grad_norm": 3.597982883453369, + "learning_rate": 7.770335239882809e-06, + "loss": 0.3167, + "step": 37272 + }, + { + "epoch": 0.8737378177065935, + "grad_norm": 0.6596841216087341, + "learning_rate": 7.767488597381867e-06, + "loss": 0.0922, + "step": 37273 + }, + { + "epoch": 0.8737612592813985, + "grad_norm": 3.5144753456115723, + "learning_rate": 7.764642455337123e-06, + "loss": 0.5629, + "step": 37274 + }, + { + "epoch": 0.8737847008562035, + "grad_norm": 3.0530667304992676, + "learning_rate": 7.761796813764032e-06, + "loss": 0.2106, + "step": 37275 + }, + { + "epoch": 0.8738081424310085, + "grad_norm": 3.785724401473999, + "learning_rate": 7.758951672678016e-06, + "loss": 0.3092, + "step": 37276 + }, + { + "epoch": 0.8738315840058135, + "grad_norm": 6.219729423522949, + "learning_rate": 7.756107032094528e-06, + "loss": 0.6263, + "step": 37277 + }, + { + "epoch": 0.8738550255806186, + "grad_norm": 2.870760679244995, + "learning_rate": 7.753262892028979e-06, + "loss": 0.2401, + "step": 37278 + }, + { + "epoch": 0.8738784671554235, + "grad_norm": 3.859347343444824, + "learning_rate": 7.750419252496844e-06, + "loss": 0.4183, + "step": 37279 + }, + { + "epoch": 0.8739019087302285, + "grad_norm": 1.9915060997009277, + "learning_rate": 7.747576113513499e-06, + "loss": 0.1794, + "step": 37280 + }, + { + "epoch": 0.8739253503050335, + "grad_norm": 0.9169178605079651, + "learning_rate": 7.744733475094423e-06, + "loss": 0.0723, + "step": 37281 + }, + { + "epoch": 0.8739487918798385, + "grad_norm": 3.68277907371521, + "learning_rate": 7.741891337255014e-06, + "loss": 0.2881, + "step": 37282 + }, + { + "epoch": 0.8739722334546435, + "grad_norm": 3.8812413215637207, + "learning_rate": 7.739049700010703e-06, + "loss": 0.3459, + "step": 37283 + }, + { + "epoch": 0.8739956750294485, + "grad_norm": 4.909111976623535, + "learning_rate": 7.73620856337689e-06, + "loss": 0.4568, + "step": 37284 + }, + { + "epoch": 0.8740191166042535, + "grad_norm": 2.50767183303833, + "learning_rate": 7.733367927368995e-06, + "loss": 1.4162, + "step": 37285 + }, + { + "epoch": 0.8740425581790585, + "grad_norm": 2.9038748741149902, + "learning_rate": 7.730527792002462e-06, + "loss": 1.8614, + "step": 37286 + }, + { + "epoch": 0.8740659997538635, + "grad_norm": 1.4447991847991943, + "learning_rate": 7.727688157292668e-06, + "loss": 0.1012, + "step": 37287 + }, + { + "epoch": 0.8740894413286685, + "grad_norm": 1.6515978574752808, + "learning_rate": 7.724849023255042e-06, + "loss": 0.1771, + "step": 37288 + }, + { + "epoch": 0.8741128829034734, + "grad_norm": 1.0022531747817993, + "learning_rate": 7.722010389904977e-06, + "loss": 0.0354, + "step": 37289 + }, + { + "epoch": 0.8741363244782785, + "grad_norm": 0.7032926678657532, + "learning_rate": 7.719172257257878e-06, + "loss": 0.0346, + "step": 37290 + }, + { + "epoch": 0.8741597660530834, + "grad_norm": 2.4815673828125, + "learning_rate": 7.716334625329125e-06, + "loss": 1.6763, + "step": 37291 + }, + { + "epoch": 0.8741832076278885, + "grad_norm": 4.448746681213379, + "learning_rate": 7.713497494134136e-06, + "loss": 0.452, + "step": 37292 + }, + { + "epoch": 0.8742066492026934, + "grad_norm": 1.437998652458191, + "learning_rate": 7.710660863688323e-06, + "loss": 0.0464, + "step": 37293 + }, + { + "epoch": 0.8742300907774985, + "grad_norm": 3.065195083618164, + "learning_rate": 7.707824734007053e-06, + "loss": 0.0772, + "step": 37294 + }, + { + "epoch": 0.8742535323523034, + "grad_norm": 2.7622673511505127, + "learning_rate": 7.704989105105731e-06, + "loss": 0.1558, + "step": 37295 + }, + { + "epoch": 0.8742769739271085, + "grad_norm": 0.31046566367149353, + "learning_rate": 7.702153976999726e-06, + "loss": 0.0154, + "step": 37296 + }, + { + "epoch": 0.8743004155019134, + "grad_norm": 2.0243217945098877, + "learning_rate": 7.699319349704415e-06, + "loss": 0.0734, + "step": 37297 + }, + { + "epoch": 0.8743238570767184, + "grad_norm": 4.002205848693848, + "learning_rate": 7.696485223235205e-06, + "loss": 0.4244, + "step": 37298 + }, + { + "epoch": 0.8743472986515234, + "grad_norm": 1.773809552192688, + "learning_rate": 7.693651597607477e-06, + "loss": 0.0602, + "step": 37299 + }, + { + "epoch": 0.8743707402263284, + "grad_norm": 2.5344913005828857, + "learning_rate": 7.69081847283657e-06, + "loss": 0.1067, + "step": 37300 + }, + { + "epoch": 0.8743941818011334, + "grad_norm": 1.4727540016174316, + "learning_rate": 7.687985848937896e-06, + "loss": 0.1518, + "step": 37301 + }, + { + "epoch": 0.8744176233759384, + "grad_norm": 3.2771995067596436, + "learning_rate": 7.685153725926797e-06, + "loss": 0.2714, + "step": 37302 + }, + { + "epoch": 0.8744410649507434, + "grad_norm": 2.4536783695220947, + "learning_rate": 7.682322103818629e-06, + "loss": 0.2029, + "step": 37303 + }, + { + "epoch": 0.8744645065255484, + "grad_norm": 1.4948070049285889, + "learning_rate": 7.67949098262878e-06, + "loss": 0.077, + "step": 37304 + }, + { + "epoch": 0.8744879481003534, + "grad_norm": 1.6545346975326538, + "learning_rate": 7.676660362372622e-06, + "loss": 1.5717, + "step": 37305 + }, + { + "epoch": 0.8745113896751584, + "grad_norm": 3.238765239715576, + "learning_rate": 7.673830243065506e-06, + "loss": 0.2079, + "step": 37306 + }, + { + "epoch": 0.8745348312499633, + "grad_norm": 3.6884074211120605, + "learning_rate": 7.67100062472279e-06, + "loss": 0.3062, + "step": 37307 + }, + { + "epoch": 0.8745582728247684, + "grad_norm": 4.530335426330566, + "learning_rate": 7.668171507359811e-06, + "loss": 0.2729, + "step": 37308 + }, + { + "epoch": 0.8745817143995733, + "grad_norm": 4.016420841217041, + "learning_rate": 7.665342890991934e-06, + "loss": 0.4233, + "step": 37309 + }, + { + "epoch": 0.8746051559743784, + "grad_norm": 0.9082427024841309, + "learning_rate": 7.662514775634477e-06, + "loss": 0.0834, + "step": 37310 + }, + { + "epoch": 0.8746285975491833, + "grad_norm": 2.613290309906006, + "learning_rate": 7.659687161302842e-06, + "loss": 0.2029, + "step": 37311 + }, + { + "epoch": 0.8746520391239884, + "grad_norm": 0.8368127346038818, + "learning_rate": 7.656860048012326e-06, + "loss": 0.0953, + "step": 37312 + }, + { + "epoch": 0.8746754806987933, + "grad_norm": 2.953303098678589, + "learning_rate": 7.654033435778296e-06, + "loss": 0.2688, + "step": 37313 + }, + { + "epoch": 0.8746989222735984, + "grad_norm": 2.3294904232025146, + "learning_rate": 7.651207324616071e-06, + "loss": 0.2063, + "step": 37314 + }, + { + "epoch": 0.8747223638484033, + "grad_norm": 2.401245355606079, + "learning_rate": 7.648381714540998e-06, + "loss": 0.0785, + "step": 37315 + }, + { + "epoch": 0.8747458054232083, + "grad_norm": 4.592925071716309, + "learning_rate": 7.645556605568372e-06, + "loss": 0.2964, + "step": 37316 + }, + { + "epoch": 0.8747692469980133, + "grad_norm": 1.1891252994537354, + "learning_rate": 7.64273199771357e-06, + "loss": 0.0698, + "step": 37317 + }, + { + "epoch": 0.8747926885728183, + "grad_norm": 2.9520959854125977, + "learning_rate": 7.639907890991904e-06, + "loss": 0.283, + "step": 37318 + }, + { + "epoch": 0.8748161301476233, + "grad_norm": 1.8227452039718628, + "learning_rate": 7.637084285418705e-06, + "loss": 0.85, + "step": 37319 + }, + { + "epoch": 0.8748395717224283, + "grad_norm": 0.8068342208862305, + "learning_rate": 7.634261181009273e-06, + "loss": 0.1089, + "step": 37320 + }, + { + "epoch": 0.8748630132972333, + "grad_norm": 3.2234115600585938, + "learning_rate": 7.63143857777895e-06, + "loss": 1.2603, + "step": 37321 + }, + { + "epoch": 0.8748864548720383, + "grad_norm": 1.1466786861419678, + "learning_rate": 7.628616475743022e-06, + "loss": 0.0979, + "step": 37322 + }, + { + "epoch": 0.8749098964468432, + "grad_norm": 0.851193904876709, + "learning_rate": 7.625794874916814e-06, + "loss": 0.0636, + "step": 37323 + }, + { + "epoch": 0.8749333380216483, + "grad_norm": 5.3125810623168945, + "learning_rate": 7.622973775315656e-06, + "loss": 0.4089, + "step": 37324 + }, + { + "epoch": 0.8749567795964532, + "grad_norm": 2.395648241043091, + "learning_rate": 7.620153176954836e-06, + "loss": 0.1732, + "step": 37325 + }, + { + "epoch": 0.8749802211712583, + "grad_norm": 3.039738655090332, + "learning_rate": 7.617333079849665e-06, + "loss": 0.315, + "step": 37326 + }, + { + "epoch": 0.8750036627460632, + "grad_norm": 2.3103837966918945, + "learning_rate": 7.61451348401544e-06, + "loss": 0.1699, + "step": 37327 + }, + { + "epoch": 0.8750271043208683, + "grad_norm": 1.7704437971115112, + "learning_rate": 7.611694389467439e-06, + "loss": 0.1377, + "step": 37328 + }, + { + "epoch": 0.8750505458956733, + "grad_norm": 2.881990671157837, + "learning_rate": 7.6088757962210154e-06, + "loss": 0.158, + "step": 37329 + }, + { + "epoch": 0.8750739874704783, + "grad_norm": 2.9341278076171875, + "learning_rate": 7.606057704291403e-06, + "loss": 0.1442, + "step": 37330 + }, + { + "epoch": 0.8750974290452833, + "grad_norm": 0.4207315444946289, + "learning_rate": 7.603240113693932e-06, + "loss": 0.0491, + "step": 37331 + }, + { + "epoch": 0.8751208706200883, + "grad_norm": 0.4401416480541229, + "learning_rate": 7.6004230244438815e-06, + "loss": 0.0389, + "step": 37332 + }, + { + "epoch": 0.8751443121948933, + "grad_norm": 3.489199161529541, + "learning_rate": 7.597606436556537e-06, + "loss": 0.3046, + "step": 37333 + }, + { + "epoch": 0.8751677537696982, + "grad_norm": 0.6611656546592712, + "learning_rate": 7.594790350047176e-06, + "loss": 0.0575, + "step": 37334 + }, + { + "epoch": 0.8751911953445033, + "grad_norm": 2.1099963188171387, + "learning_rate": 7.591974764931064e-06, + "loss": 0.1762, + "step": 37335 + }, + { + "epoch": 0.8752146369193082, + "grad_norm": 1.9307886362075806, + "learning_rate": 7.589159681223512e-06, + "loss": 1.0422, + "step": 37336 + }, + { + "epoch": 0.8752380784941133, + "grad_norm": 2.127462863922119, + "learning_rate": 7.586345098939784e-06, + "loss": 0.1505, + "step": 37337 + }, + { + "epoch": 0.8752615200689182, + "grad_norm": 3.849154233932495, + "learning_rate": 7.583531018095136e-06, + "loss": 0.2044, + "step": 37338 + }, + { + "epoch": 0.8752849616437233, + "grad_norm": 0.3138620853424072, + "learning_rate": 7.580717438704854e-06, + "loss": 0.0256, + "step": 37339 + }, + { + "epoch": 0.8753084032185282, + "grad_norm": 2.3081884384155273, + "learning_rate": 7.577904360784194e-06, + "loss": 1.3211, + "step": 37340 + }, + { + "epoch": 0.8753318447933333, + "grad_norm": 2.8020122051239014, + "learning_rate": 7.57509178434841e-06, + "loss": 0.2065, + "step": 37341 + }, + { + "epoch": 0.8753552863681382, + "grad_norm": 2.17154598236084, + "learning_rate": 7.572279709412777e-06, + "loss": 0.0994, + "step": 37342 + }, + { + "epoch": 0.8753787279429432, + "grad_norm": 2.282505750656128, + "learning_rate": 7.569468135992564e-06, + "loss": 0.1637, + "step": 37343 + }, + { + "epoch": 0.8754021695177482, + "grad_norm": 0.7238446474075317, + "learning_rate": 7.566657064103022e-06, + "loss": 0.022, + "step": 37344 + }, + { + "epoch": 0.8754256110925532, + "grad_norm": 1.23201584815979, + "learning_rate": 7.563846493759397e-06, + "loss": 0.0871, + "step": 37345 + }, + { + "epoch": 0.8754490526673582, + "grad_norm": 0.8535782098770142, + "learning_rate": 7.561036424976931e-06, + "loss": 0.0603, + "step": 37346 + }, + { + "epoch": 0.8754724942421632, + "grad_norm": 3.4477391242980957, + "learning_rate": 7.558226857770889e-06, + "loss": 0.1482, + "step": 37347 + }, + { + "epoch": 0.8754959358169682, + "grad_norm": 2.4692413806915283, + "learning_rate": 7.555417792156494e-06, + "loss": 0.1456, + "step": 37348 + }, + { + "epoch": 0.8755193773917732, + "grad_norm": 2.1967971324920654, + "learning_rate": 7.55260922814901e-06, + "loss": 0.1657, + "step": 37349 + }, + { + "epoch": 0.8755428189665782, + "grad_norm": 0.6393860578536987, + "learning_rate": 7.5498011657636705e-06, + "loss": 0.0926, + "step": 37350 + }, + { + "epoch": 0.8755662605413832, + "grad_norm": 0.3472619652748108, + "learning_rate": 7.546993605015706e-06, + "loss": 0.0106, + "step": 37351 + }, + { + "epoch": 0.8755897021161881, + "grad_norm": 0.20586109161376953, + "learning_rate": 7.544186545920362e-06, + "loss": 0.0119, + "step": 37352 + }, + { + "epoch": 0.8756131436909932, + "grad_norm": 0.37282490730285645, + "learning_rate": 7.541379988492836e-06, + "loss": 0.0327, + "step": 37353 + }, + { + "epoch": 0.8756365852657981, + "grad_norm": 2.4099063873291016, + "learning_rate": 7.538573932748405e-06, + "loss": 0.1147, + "step": 37354 + }, + { + "epoch": 0.8756600268406032, + "grad_norm": 2.642059326171875, + "learning_rate": 7.535768378702257e-06, + "loss": 0.1223, + "step": 37355 + }, + { + "epoch": 0.8756834684154081, + "grad_norm": 1.5606913566589355, + "learning_rate": 7.532963326369636e-06, + "loss": 0.0562, + "step": 37356 + }, + { + "epoch": 0.8757069099902132, + "grad_norm": 3.80000638961792, + "learning_rate": 7.530158775765772e-06, + "loss": 0.4286, + "step": 37357 + }, + { + "epoch": 0.8757303515650181, + "grad_norm": 1.1948193311691284, + "learning_rate": 7.527354726905856e-06, + "loss": 0.079, + "step": 37358 + }, + { + "epoch": 0.8757537931398232, + "grad_norm": 2.9537549018859863, + "learning_rate": 7.524551179805106e-06, + "loss": 0.1913, + "step": 37359 + }, + { + "epoch": 0.8757772347146281, + "grad_norm": 2.7977702617645264, + "learning_rate": 7.521748134478756e-06, + "loss": 0.2655, + "step": 37360 + }, + { + "epoch": 0.8758006762894331, + "grad_norm": 2.8184092044830322, + "learning_rate": 7.518945590941972e-06, + "loss": 0.2098, + "step": 37361 + }, + { + "epoch": 0.8758241178642381, + "grad_norm": 1.040304183959961, + "learning_rate": 7.5161435492100195e-06, + "loss": 0.0555, + "step": 37362 + }, + { + "epoch": 0.8758475594390431, + "grad_norm": 2.38725209236145, + "learning_rate": 7.513342009298063e-06, + "loss": 0.1624, + "step": 37363 + }, + { + "epoch": 0.8758710010138481, + "grad_norm": 4.782529354095459, + "learning_rate": 7.510540971221314e-06, + "loss": 0.2828, + "step": 37364 + }, + { + "epoch": 0.8758944425886531, + "grad_norm": 4.83221435546875, + "learning_rate": 7.507740434994959e-06, + "loss": 0.154, + "step": 37365 + }, + { + "epoch": 0.8759178841634581, + "grad_norm": 1.7837129831314087, + "learning_rate": 7.504940400634208e-06, + "loss": 1.0111, + "step": 37366 + }, + { + "epoch": 0.8759413257382631, + "grad_norm": 2.8891401290893555, + "learning_rate": 7.502140868154262e-06, + "loss": 0.1686, + "step": 37367 + }, + { + "epoch": 0.875964767313068, + "grad_norm": 2.5206496715545654, + "learning_rate": 7.499341837570284e-06, + "loss": 0.1677, + "step": 37368 + }, + { + "epoch": 0.8759882088878731, + "grad_norm": 0.8585243821144104, + "learning_rate": 7.496543308897497e-06, + "loss": 0.0872, + "step": 37369 + }, + { + "epoch": 0.876011650462678, + "grad_norm": 2.730421781539917, + "learning_rate": 7.4937452821510654e-06, + "loss": 1.5925, + "step": 37370 + }, + { + "epoch": 0.8760350920374831, + "grad_norm": 2.140120506286621, + "learning_rate": 7.490947757346167e-06, + "loss": 0.2237, + "step": 37371 + }, + { + "epoch": 0.876058533612288, + "grad_norm": 3.886655330657959, + "learning_rate": 7.488150734497989e-06, + "loss": 0.1997, + "step": 37372 + }, + { + "epoch": 0.8760819751870931, + "grad_norm": 2.520390272140503, + "learning_rate": 7.485354213621698e-06, + "loss": 0.24, + "step": 37373 + }, + { + "epoch": 0.876105416761898, + "grad_norm": 2.2171742916107178, + "learning_rate": 7.482558194732481e-06, + "loss": 0.1634, + "step": 37374 + }, + { + "epoch": 0.8761288583367031, + "grad_norm": 0.5174683928489685, + "learning_rate": 7.479762677845514e-06, + "loss": 0.032, + "step": 37375 + }, + { + "epoch": 0.876152299911508, + "grad_norm": 4.7277607917785645, + "learning_rate": 7.4769676629759534e-06, + "loss": 0.4991, + "step": 37376 + }, + { + "epoch": 0.876175741486313, + "grad_norm": 2.945732831954956, + "learning_rate": 7.474173150138941e-06, + "loss": 0.2746, + "step": 37377 + }, + { + "epoch": 0.876199183061118, + "grad_norm": 1.1179314851760864, + "learning_rate": 7.471379139349699e-06, + "loss": 0.0635, + "step": 37378 + }, + { + "epoch": 0.876222624635923, + "grad_norm": 1.1693115234375, + "learning_rate": 7.468585630623348e-06, + "loss": 0.0784, + "step": 37379 + }, + { + "epoch": 0.8762460662107281, + "grad_norm": 3.754628896713257, + "learning_rate": 7.465792623975032e-06, + "loss": 0.3568, + "step": 37380 + }, + { + "epoch": 0.876269507785533, + "grad_norm": 2.076134204864502, + "learning_rate": 7.463000119419949e-06, + "loss": 1.3631, + "step": 37381 + }, + { + "epoch": 0.8762929493603381, + "grad_norm": 0.36687979102134705, + "learning_rate": 7.460208116973233e-06, + "loss": 0.0058, + "step": 37382 + }, + { + "epoch": 0.876316390935143, + "grad_norm": 3.7059168815612793, + "learning_rate": 7.457416616650026e-06, + "loss": 0.4367, + "step": 37383 + }, + { + "epoch": 0.8763398325099481, + "grad_norm": 0.468333899974823, + "learning_rate": 7.454625618465483e-06, + "loss": 0.0114, + "step": 37384 + }, + { + "epoch": 0.876363274084753, + "grad_norm": 3.086216449737549, + "learning_rate": 7.451835122434736e-06, + "loss": 0.3049, + "step": 37385 + }, + { + "epoch": 0.8763867156595581, + "grad_norm": 1.9847179651260376, + "learning_rate": 7.449045128572918e-06, + "loss": 0.9305, + "step": 37386 + }, + { + "epoch": 0.876410157234363, + "grad_norm": 4.153652191162109, + "learning_rate": 7.446255636895205e-06, + "loss": 0.3871, + "step": 37387 + }, + { + "epoch": 0.876433598809168, + "grad_norm": 4.279313087463379, + "learning_rate": 7.443466647416708e-06, + "loss": 0.2717, + "step": 37388 + }, + { + "epoch": 0.876457040383973, + "grad_norm": 1.939799189567566, + "learning_rate": 7.440678160152547e-06, + "loss": 1.1856, + "step": 37389 + }, + { + "epoch": 0.876480481958778, + "grad_norm": 0.4860302805900574, + "learning_rate": 7.43789017511789e-06, + "loss": 0.0367, + "step": 37390 + }, + { + "epoch": 0.876503923533583, + "grad_norm": 2.5802247524261475, + "learning_rate": 7.435102692327833e-06, + "loss": 1.8121, + "step": 37391 + }, + { + "epoch": 0.876527365108388, + "grad_norm": 0.6339633464813232, + "learning_rate": 7.432315711797522e-06, + "loss": 0.0379, + "step": 37392 + }, + { + "epoch": 0.876550806683193, + "grad_norm": 3.5021708011627197, + "learning_rate": 7.429529233542043e-06, + "loss": 0.3357, + "step": 37393 + }, + { + "epoch": 0.876574248257998, + "grad_norm": 0.8629961609840393, + "learning_rate": 7.426743257576563e-06, + "loss": 0.0428, + "step": 37394 + }, + { + "epoch": 0.876597689832803, + "grad_norm": 2.6032333374023438, + "learning_rate": 7.42395778391618e-06, + "loss": 0.1353, + "step": 37395 + }, + { + "epoch": 0.876621131407608, + "grad_norm": 1.1954845190048218, + "learning_rate": 7.421172812576005e-06, + "loss": 0.066, + "step": 37396 + }, + { + "epoch": 0.8766445729824129, + "grad_norm": 1.947220802307129, + "learning_rate": 7.418388343571147e-06, + "loss": 0.0421, + "step": 37397 + }, + { + "epoch": 0.876668014557218, + "grad_norm": 1.1141653060913086, + "learning_rate": 7.415604376916718e-06, + "loss": 0.1118, + "step": 37398 + }, + { + "epoch": 0.8766914561320229, + "grad_norm": 3.0435149669647217, + "learning_rate": 7.412820912627805e-06, + "loss": 0.131, + "step": 37399 + }, + { + "epoch": 0.876714897706828, + "grad_norm": 0.8901415467262268, + "learning_rate": 7.410037950719539e-06, + "loss": 0.0746, + "step": 37400 + }, + { + "epoch": 0.8767383392816329, + "grad_norm": 0.4596942961215973, + "learning_rate": 7.407255491206999e-06, + "loss": 0.038, + "step": 37401 + }, + { + "epoch": 0.876761780856438, + "grad_norm": 1.1794495582580566, + "learning_rate": 7.404473534105317e-06, + "loss": 0.0979, + "step": 37402 + }, + { + "epoch": 0.8767852224312429, + "grad_norm": 1.7464278936386108, + "learning_rate": 7.401692079429556e-06, + "loss": 0.0876, + "step": 37403 + }, + { + "epoch": 0.876808664006048, + "grad_norm": 1.604369044303894, + "learning_rate": 7.39891112719483e-06, + "loss": 0.1012, + "step": 37404 + }, + { + "epoch": 0.8768321055808529, + "grad_norm": 4.319499969482422, + "learning_rate": 7.396130677416202e-06, + "loss": 0.2428, + "step": 37405 + }, + { + "epoch": 0.8768555471556579, + "grad_norm": 1.0095865726470947, + "learning_rate": 7.393350730108762e-06, + "loss": 0.0772, + "step": 37406 + }, + { + "epoch": 0.8768789887304629, + "grad_norm": 0.450739324092865, + "learning_rate": 7.390571285287618e-06, + "loss": 0.0389, + "step": 37407 + }, + { + "epoch": 0.8769024303052679, + "grad_norm": 3.947000503540039, + "learning_rate": 7.387792342967847e-06, + "loss": 0.1951, + "step": 37408 + }, + { + "epoch": 0.8769258718800729, + "grad_norm": 1.6688148975372314, + "learning_rate": 7.385013903164506e-06, + "loss": 0.0404, + "step": 37409 + }, + { + "epoch": 0.8769493134548779, + "grad_norm": 3.0817863941192627, + "learning_rate": 7.382235965892692e-06, + "loss": 0.35, + "step": 37410 + }, + { + "epoch": 0.8769727550296829, + "grad_norm": 1.984774112701416, + "learning_rate": 7.379458531167449e-06, + "loss": 0.1734, + "step": 37411 + }, + { + "epoch": 0.8769961966044879, + "grad_norm": 0.31769901514053345, + "learning_rate": 7.376681599003899e-06, + "loss": 0.0373, + "step": 37412 + }, + { + "epoch": 0.8770196381792928, + "grad_norm": 2.326406478881836, + "learning_rate": 7.3739051694170505e-06, + "loss": 0.1768, + "step": 37413 + }, + { + "epoch": 0.8770430797540979, + "grad_norm": 0.9473302960395813, + "learning_rate": 7.371129242422003e-06, + "loss": 0.0541, + "step": 37414 + }, + { + "epoch": 0.8770665213289028, + "grad_norm": 0.8123084902763367, + "learning_rate": 7.368353818033835e-06, + "loss": 0.1097, + "step": 37415 + }, + { + "epoch": 0.8770899629037079, + "grad_norm": 2.4218311309814453, + "learning_rate": 7.365578896267567e-06, + "loss": 0.1687, + "step": 37416 + }, + { + "epoch": 0.8771134044785128, + "grad_norm": 0.571479082107544, + "learning_rate": 7.362804477138274e-06, + "loss": 0.032, + "step": 37417 + }, + { + "epoch": 0.8771368460533179, + "grad_norm": 1.2838226556777954, + "learning_rate": 7.360030560661002e-06, + "loss": 0.0874, + "step": 37418 + }, + { + "epoch": 0.8771602876281228, + "grad_norm": 1.8094027042388916, + "learning_rate": 7.357257146850816e-06, + "loss": 0.0777, + "step": 37419 + }, + { + "epoch": 0.8771837292029279, + "grad_norm": 0.9306883811950684, + "learning_rate": 7.354484235722769e-06, + "loss": 0.0745, + "step": 37420 + }, + { + "epoch": 0.8772071707777328, + "grad_norm": 2.596447467803955, + "learning_rate": 7.351711827291885e-06, + "loss": 1.6089, + "step": 37421 + }, + { + "epoch": 0.8772306123525379, + "grad_norm": 2.671060562133789, + "learning_rate": 7.348939921573217e-06, + "loss": 0.2407, + "step": 37422 + }, + { + "epoch": 0.8772540539273428, + "grad_norm": 4.129461288452148, + "learning_rate": 7.346168518581809e-06, + "loss": 0.316, + "step": 37423 + }, + { + "epoch": 0.8772774955021478, + "grad_norm": 1.6353999376296997, + "learning_rate": 7.343397618332681e-06, + "loss": 0.1897, + "step": 37424 + }, + { + "epoch": 0.8773009370769528, + "grad_norm": 3.2017838954925537, + "learning_rate": 7.340627220840901e-06, + "loss": 0.4357, + "step": 37425 + }, + { + "epoch": 0.8773243786517578, + "grad_norm": 2.4195685386657715, + "learning_rate": 7.337857326121466e-06, + "loss": 0.1853, + "step": 37426 + }, + { + "epoch": 0.8773478202265628, + "grad_norm": 1.5618191957473755, + "learning_rate": 7.335087934189433e-06, + "loss": 0.086, + "step": 37427 + }, + { + "epoch": 0.8773712618013678, + "grad_norm": 2.8864822387695312, + "learning_rate": 7.332319045059821e-06, + "loss": 0.1896, + "step": 37428 + }, + { + "epoch": 0.8773947033761728, + "grad_norm": 1.1865029335021973, + "learning_rate": 7.329550658747653e-06, + "loss": 0.1281, + "step": 37429 + }, + { + "epoch": 0.8774181449509778, + "grad_norm": 0.9106546640396118, + "learning_rate": 7.3267827752679485e-06, + "loss": 0.0861, + "step": 37430 + }, + { + "epoch": 0.8774415865257829, + "grad_norm": 1.5208855867385864, + "learning_rate": 7.324015394635708e-06, + "loss": 0.0833, + "step": 37431 + }, + { + "epoch": 0.8774650281005878, + "grad_norm": 4.072812080383301, + "learning_rate": 7.321248516865986e-06, + "loss": 0.3557, + "step": 37432 + }, + { + "epoch": 0.8774884696753928, + "grad_norm": 2.762092113494873, + "learning_rate": 7.318482141973781e-06, + "loss": 0.1451, + "step": 37433 + }, + { + "epoch": 0.8775119112501978, + "grad_norm": 1.5313022136688232, + "learning_rate": 7.3157162699740825e-06, + "loss": 0.0844, + "step": 37434 + }, + { + "epoch": 0.8775353528250028, + "grad_norm": 4.716525554656982, + "learning_rate": 7.312950900881921e-06, + "loss": 0.1696, + "step": 37435 + }, + { + "epoch": 0.8775587943998078, + "grad_norm": 3.2509822845458984, + "learning_rate": 7.310186034712296e-06, + "loss": 0.2825, + "step": 37436 + }, + { + "epoch": 0.8775822359746128, + "grad_norm": 0.6556521058082581, + "learning_rate": 7.307421671480197e-06, + "loss": 0.0372, + "step": 37437 + }, + { + "epoch": 0.8776056775494178, + "grad_norm": 1.6604461669921875, + "learning_rate": 7.3046578112006215e-06, + "loss": 1.304, + "step": 37438 + }, + { + "epoch": 0.8776291191242228, + "grad_norm": 3.064023733139038, + "learning_rate": 7.301894453888602e-06, + "loss": 0.2034, + "step": 37439 + }, + { + "epoch": 0.8776525606990278, + "grad_norm": 2.110569953918457, + "learning_rate": 7.299131599559106e-06, + "loss": 0.0783, + "step": 37440 + }, + { + "epoch": 0.8776760022738328, + "grad_norm": 2.0419864654541016, + "learning_rate": 7.296369248227131e-06, + "loss": 0.1083, + "step": 37441 + }, + { + "epoch": 0.8776994438486377, + "grad_norm": 2.0251152515411377, + "learning_rate": 7.293607399907665e-06, + "loss": 0.2123, + "step": 37442 + }, + { + "epoch": 0.8777228854234428, + "grad_norm": 2.525275468826294, + "learning_rate": 7.2908460546156965e-06, + "loss": 0.1694, + "step": 37443 + }, + { + "epoch": 0.8777463269982477, + "grad_norm": 2.308913230895996, + "learning_rate": 7.28808521236618e-06, + "loss": 0.1643, + "step": 37444 + }, + { + "epoch": 0.8777697685730528, + "grad_norm": 2.6315557956695557, + "learning_rate": 7.2853248731741375e-06, + "loss": 0.0965, + "step": 37445 + }, + { + "epoch": 0.8777932101478577, + "grad_norm": 2.9153647422790527, + "learning_rate": 7.282565037054534e-06, + "loss": 0.115, + "step": 37446 + }, + { + "epoch": 0.8778166517226628, + "grad_norm": 4.4340362548828125, + "learning_rate": 7.279805704022324e-06, + "loss": 0.5611, + "step": 37447 + }, + { + "epoch": 0.8778400932974677, + "grad_norm": 0.4130610227584839, + "learning_rate": 7.277046874092508e-06, + "loss": 0.0309, + "step": 37448 + }, + { + "epoch": 0.8778635348722728, + "grad_norm": 0.22475078701972961, + "learning_rate": 7.274288547280028e-06, + "loss": 0.0175, + "step": 37449 + }, + { + "epoch": 0.8778869764470777, + "grad_norm": 2.9689877033233643, + "learning_rate": 7.271530723599884e-06, + "loss": 0.1721, + "step": 37450 + }, + { + "epoch": 0.8779104180218827, + "grad_norm": 5.066195487976074, + "learning_rate": 7.268773403066998e-06, + "loss": 0.2589, + "step": 37451 + }, + { + "epoch": 0.8779338595966877, + "grad_norm": 2.2610039710998535, + "learning_rate": 7.266016585696367e-06, + "loss": 0.1111, + "step": 37452 + }, + { + "epoch": 0.8779573011714927, + "grad_norm": 3.705193519592285, + "learning_rate": 7.263260271502948e-06, + "loss": 0.1796, + "step": 37453 + }, + { + "epoch": 0.8779807427462977, + "grad_norm": 0.6620181798934937, + "learning_rate": 7.260504460501682e-06, + "loss": 0.0346, + "step": 37454 + }, + { + "epoch": 0.8780041843211027, + "grad_norm": 6.266302108764648, + "learning_rate": 7.257749152707527e-06, + "loss": 0.4723, + "step": 37455 + }, + { + "epoch": 0.8780276258959077, + "grad_norm": 2.591066360473633, + "learning_rate": 7.254994348135413e-06, + "loss": 0.1644, + "step": 37456 + }, + { + "epoch": 0.8780510674707127, + "grad_norm": 3.5720174312591553, + "learning_rate": 7.25224004680033e-06, + "loss": 0.6388, + "step": 37457 + }, + { + "epoch": 0.8780745090455176, + "grad_norm": 2.71016001701355, + "learning_rate": 7.249486248717196e-06, + "loss": 0.2009, + "step": 37458 + }, + { + "epoch": 0.8780979506203227, + "grad_norm": 2.903132200241089, + "learning_rate": 7.2467329539009585e-06, + "loss": 0.19, + "step": 37459 + }, + { + "epoch": 0.8781213921951276, + "grad_norm": 3.4970293045043945, + "learning_rate": 7.24398016236656e-06, + "loss": 0.1628, + "step": 37460 + }, + { + "epoch": 0.8781448337699327, + "grad_norm": 5.801788806915283, + "learning_rate": 7.241227874128931e-06, + "loss": 0.4625, + "step": 37461 + }, + { + "epoch": 0.8781682753447376, + "grad_norm": 0.7730886936187744, + "learning_rate": 7.238476089202983e-06, + "loss": 0.0949, + "step": 37462 + }, + { + "epoch": 0.8781917169195427, + "grad_norm": 3.8559656143188477, + "learning_rate": 7.235724807603683e-06, + "loss": 0.3609, + "step": 37463 + }, + { + "epoch": 0.8782151584943476, + "grad_norm": 2.7853074073791504, + "learning_rate": 7.232974029345963e-06, + "loss": 1.1434, + "step": 37464 + }, + { + "epoch": 0.8782386000691527, + "grad_norm": 2.6374893188476562, + "learning_rate": 7.2302237544447205e-06, + "loss": 0.1687, + "step": 37465 + }, + { + "epoch": 0.8782620416439576, + "grad_norm": 1.093117117881775, + "learning_rate": 7.227473982914912e-06, + "loss": 0.0837, + "step": 37466 + }, + { + "epoch": 0.8782854832187627, + "grad_norm": 3.0214364528656006, + "learning_rate": 7.224724714771424e-06, + "loss": 0.2972, + "step": 37467 + }, + { + "epoch": 0.8783089247935676, + "grad_norm": 0.6949450969696045, + "learning_rate": 7.221975950029192e-06, + "loss": 0.0181, + "step": 37468 + }, + { + "epoch": 0.8783323663683726, + "grad_norm": 1.5497729778289795, + "learning_rate": 7.219227688703123e-06, + "loss": 1.2522, + "step": 37469 + }, + { + "epoch": 0.8783558079431776, + "grad_norm": 0.6861217021942139, + "learning_rate": 7.216479930808129e-06, + "loss": 0.0697, + "step": 37470 + }, + { + "epoch": 0.8783792495179826, + "grad_norm": 3.8272361755371094, + "learning_rate": 7.2137326763591416e-06, + "loss": 0.2892, + "step": 37471 + }, + { + "epoch": 0.8784026910927876, + "grad_norm": 2.418097496032715, + "learning_rate": 7.210985925371039e-06, + "loss": 0.2673, + "step": 37472 + }, + { + "epoch": 0.8784261326675926, + "grad_norm": 1.2165583372116089, + "learning_rate": 7.208239677858741e-06, + "loss": 0.1321, + "step": 37473 + }, + { + "epoch": 0.8784495742423976, + "grad_norm": 2.244758367538452, + "learning_rate": 7.205493933837127e-06, + "loss": 0.1483, + "step": 37474 + }, + { + "epoch": 0.8784730158172026, + "grad_norm": 1.330613136291504, + "learning_rate": 7.202748693321126e-06, + "loss": 0.1115, + "step": 37475 + }, + { + "epoch": 0.8784964573920075, + "grad_norm": 1.0357112884521484, + "learning_rate": 7.200003956325607e-06, + "loss": 0.1069, + "step": 37476 + }, + { + "epoch": 0.8785198989668126, + "grad_norm": 2.3226535320281982, + "learning_rate": 7.1972597228654795e-06, + "loss": 0.0882, + "step": 37477 + }, + { + "epoch": 0.8785433405416175, + "grad_norm": 1.2506539821624756, + "learning_rate": 7.194515992955642e-06, + "loss": 0.0749, + "step": 37478 + }, + { + "epoch": 0.8785667821164226, + "grad_norm": 1.878199577331543, + "learning_rate": 7.191772766610972e-06, + "loss": 0.1485, + "step": 37479 + }, + { + "epoch": 0.8785902236912275, + "grad_norm": 4.566648483276367, + "learning_rate": 7.189030043846345e-06, + "loss": 0.4245, + "step": 37480 + }, + { + "epoch": 0.8786136652660326, + "grad_norm": 2.4610233306884766, + "learning_rate": 7.186287824676663e-06, + "loss": 0.1549, + "step": 37481 + }, + { + "epoch": 0.8786371068408376, + "grad_norm": 2.423743724822998, + "learning_rate": 7.183546109116768e-06, + "loss": 0.176, + "step": 37482 + }, + { + "epoch": 0.8786605484156426, + "grad_norm": 1.8501137495040894, + "learning_rate": 7.18080489718157e-06, + "loss": 0.105, + "step": 37483 + }, + { + "epoch": 0.8786839899904476, + "grad_norm": 0.5923429131507874, + "learning_rate": 7.1780641888859465e-06, + "loss": 0.0329, + "step": 37484 + }, + { + "epoch": 0.8787074315652526, + "grad_norm": 2.7810182571411133, + "learning_rate": 7.175323984244753e-06, + "loss": 1.2877, + "step": 37485 + }, + { + "epoch": 0.8787308731400576, + "grad_norm": 2.951303005218506, + "learning_rate": 7.172584283272843e-06, + "loss": 0.2952, + "step": 37486 + }, + { + "epoch": 0.8787543147148625, + "grad_norm": 1.3888728618621826, + "learning_rate": 7.169845085985117e-06, + "loss": 0.0554, + "step": 37487 + }, + { + "epoch": 0.8787777562896676, + "grad_norm": 2.0501461029052734, + "learning_rate": 7.167106392396427e-06, + "loss": 0.1505, + "step": 37488 + }, + { + "epoch": 0.8788011978644725, + "grad_norm": 1.4751229286193848, + "learning_rate": 7.16436820252161e-06, + "loss": 0.1284, + "step": 37489 + }, + { + "epoch": 0.8788246394392776, + "grad_norm": 0.4153165817260742, + "learning_rate": 7.16163051637555e-06, + "loss": 0.0182, + "step": 37490 + }, + { + "epoch": 0.8788480810140825, + "grad_norm": 3.4055709838867188, + "learning_rate": 7.158893333973105e-06, + "loss": 0.1784, + "step": 37491 + }, + { + "epoch": 0.8788715225888876, + "grad_norm": 2.2996206283569336, + "learning_rate": 7.156156655329105e-06, + "loss": 0.1542, + "step": 37492 + }, + { + "epoch": 0.8788949641636925, + "grad_norm": 3.96047043800354, + "learning_rate": 7.153420480458417e-06, + "loss": 0.4294, + "step": 37493 + }, + { + "epoch": 0.8789184057384976, + "grad_norm": 4.0906982421875, + "learning_rate": 7.150684809375863e-06, + "loss": 0.1341, + "step": 37494 + }, + { + "epoch": 0.8789418473133025, + "grad_norm": 1.8534526824951172, + "learning_rate": 7.147949642096308e-06, + "loss": 0.065, + "step": 37495 + }, + { + "epoch": 0.8789652888881075, + "grad_norm": 4.228043556213379, + "learning_rate": 7.145214978634607e-06, + "loss": 0.3917, + "step": 37496 + }, + { + "epoch": 0.8789887304629125, + "grad_norm": 1.4488729238510132, + "learning_rate": 7.142480819005571e-06, + "loss": 0.1463, + "step": 37497 + }, + { + "epoch": 0.8790121720377175, + "grad_norm": 2.5664079189300537, + "learning_rate": 7.1397471632240424e-06, + "loss": 0.2194, + "step": 37498 + }, + { + "epoch": 0.8790356136125225, + "grad_norm": 1.777754306793213, + "learning_rate": 7.1370140113048435e-06, + "loss": 0.2445, + "step": 37499 + }, + { + "epoch": 0.8790590551873275, + "grad_norm": 2.863370895385742, + "learning_rate": 7.13428136326284e-06, + "loss": 0.1457, + "step": 37500 + }, + { + "epoch": 0.8790824967621325, + "grad_norm": 2.120115041732788, + "learning_rate": 7.13154921911281e-06, + "loss": 0.1567, + "step": 37501 + }, + { + "epoch": 0.8791059383369375, + "grad_norm": 1.6262379884719849, + "learning_rate": 7.128817578869629e-06, + "loss": 0.136, + "step": 37502 + }, + { + "epoch": 0.8791293799117424, + "grad_norm": 1.007185697555542, + "learning_rate": 7.126086442548096e-06, + "loss": 0.6557, + "step": 37503 + }, + { + "epoch": 0.8791528214865475, + "grad_norm": 3.2018444538116455, + "learning_rate": 7.123355810163024e-06, + "loss": 0.3955, + "step": 37504 + }, + { + "epoch": 0.8791762630613524, + "grad_norm": 4.781523704528809, + "learning_rate": 7.120625681729231e-06, + "loss": 0.257, + "step": 37505 + }, + { + "epoch": 0.8791997046361575, + "grad_norm": 0.8006435632705688, + "learning_rate": 7.117896057261541e-06, + "loss": 0.0877, + "step": 37506 + }, + { + "epoch": 0.8792231462109624, + "grad_norm": 2.1761114597320557, + "learning_rate": 7.115166936774742e-06, + "loss": 0.1591, + "step": 37507 + }, + { + "epoch": 0.8792465877857675, + "grad_norm": 5.355931758880615, + "learning_rate": 7.112438320283676e-06, + "loss": 0.6591, + "step": 37508 + }, + { + "epoch": 0.8792700293605724, + "grad_norm": 4.300551891326904, + "learning_rate": 7.109710207803133e-06, + "loss": 0.2178, + "step": 37509 + }, + { + "epoch": 0.8792934709353775, + "grad_norm": 2.9082415103912354, + "learning_rate": 7.106982599347911e-06, + "loss": 0.0818, + "step": 37510 + }, + { + "epoch": 0.8793169125101824, + "grad_norm": 0.41569194197654724, + "learning_rate": 7.104255494932799e-06, + "loss": 0.0324, + "step": 37511 + }, + { + "epoch": 0.8793403540849875, + "grad_norm": 1.985754370689392, + "learning_rate": 7.101528894572629e-06, + "loss": 0.1776, + "step": 37512 + }, + { + "epoch": 0.8793637956597924, + "grad_norm": 2.6894140243530273, + "learning_rate": 7.098802798282167e-06, + "loss": 1.5907, + "step": 37513 + }, + { + "epoch": 0.8793872372345974, + "grad_norm": 2.412382125854492, + "learning_rate": 7.096077206076202e-06, + "loss": 0.1264, + "step": 37514 + }, + { + "epoch": 0.8794106788094024, + "grad_norm": 2.738299608230591, + "learning_rate": 7.093352117969543e-06, + "loss": 0.1265, + "step": 37515 + }, + { + "epoch": 0.8794341203842074, + "grad_norm": 2.461430311203003, + "learning_rate": 7.090627533976979e-06, + "loss": 0.2061, + "step": 37516 + }, + { + "epoch": 0.8794575619590124, + "grad_norm": 3.087336778640747, + "learning_rate": 7.087903454113276e-06, + "loss": 0.1986, + "step": 37517 + }, + { + "epoch": 0.8794810035338174, + "grad_norm": 0.6631993055343628, + "learning_rate": 7.085179878393222e-06, + "loss": 0.0551, + "step": 37518 + }, + { + "epoch": 0.8795044451086224, + "grad_norm": 2.2293784618377686, + "learning_rate": 7.082456806831595e-06, + "loss": 0.1191, + "step": 37519 + }, + { + "epoch": 0.8795278866834274, + "grad_norm": 2.6520261764526367, + "learning_rate": 7.079734239443148e-06, + "loss": 0.2067, + "step": 37520 + }, + { + "epoch": 0.8795513282582323, + "grad_norm": 0.666343092918396, + "learning_rate": 7.077012176242703e-06, + "loss": 0.0548, + "step": 37521 + }, + { + "epoch": 0.8795747698330374, + "grad_norm": 4.149672985076904, + "learning_rate": 7.0742906172449945e-06, + "loss": 0.1659, + "step": 37522 + }, + { + "epoch": 0.8795982114078423, + "grad_norm": 1.9981496334075928, + "learning_rate": 7.071569562464786e-06, + "loss": 1.4075, + "step": 37523 + }, + { + "epoch": 0.8796216529826474, + "grad_norm": 1.0147684812545776, + "learning_rate": 7.068849011916867e-06, + "loss": 0.081, + "step": 37524 + }, + { + "epoch": 0.8796450945574523, + "grad_norm": 3.754981279373169, + "learning_rate": 7.066128965615993e-06, + "loss": 0.3493, + "step": 37525 + }, + { + "epoch": 0.8796685361322574, + "grad_norm": 1.83586585521698, + "learning_rate": 7.063409423576906e-06, + "loss": 0.136, + "step": 37526 + }, + { + "epoch": 0.8796919777070623, + "grad_norm": 2.1061508655548096, + "learning_rate": 7.060690385814361e-06, + "loss": 0.1672, + "step": 37527 + }, + { + "epoch": 0.8797154192818674, + "grad_norm": 1.0888279676437378, + "learning_rate": 7.057971852343137e-06, + "loss": 0.0475, + "step": 37528 + }, + { + "epoch": 0.8797388608566723, + "grad_norm": 0.44335275888442993, + "learning_rate": 7.055253823177976e-06, + "loss": 0.0357, + "step": 37529 + }, + { + "epoch": 0.8797623024314773, + "grad_norm": 1.357795238494873, + "learning_rate": 7.0525362983336225e-06, + "loss": 0.0831, + "step": 37530 + }, + { + "epoch": 0.8797857440062823, + "grad_norm": 2.7345187664031982, + "learning_rate": 7.0498192778248095e-06, + "loss": 0.168, + "step": 37531 + }, + { + "epoch": 0.8798091855810873, + "grad_norm": 2.6543872356414795, + "learning_rate": 7.0471027616662795e-06, + "loss": 0.1428, + "step": 37532 + }, + { + "epoch": 0.8798326271558924, + "grad_norm": 2.4472427368164062, + "learning_rate": 7.0443867498728115e-06, + "loss": 0.1807, + "step": 37533 + }, + { + "epoch": 0.8798560687306973, + "grad_norm": 3.273955821990967, + "learning_rate": 7.041671242459102e-06, + "loss": 1.5169, + "step": 37534 + }, + { + "epoch": 0.8798795103055024, + "grad_norm": 2.923152446746826, + "learning_rate": 7.038956239439886e-06, + "loss": 0.3177, + "step": 37535 + }, + { + "epoch": 0.8799029518803073, + "grad_norm": 3.171159505844116, + "learning_rate": 7.0362417408299185e-06, + "loss": 0.2318, + "step": 37536 + }, + { + "epoch": 0.8799263934551124, + "grad_norm": 4.238326072692871, + "learning_rate": 7.03352774664392e-06, + "loss": 0.3065, + "step": 37537 + }, + { + "epoch": 0.8799498350299173, + "grad_norm": 6.394739627838135, + "learning_rate": 7.030814256896623e-06, + "loss": 0.4259, + "step": 37538 + }, + { + "epoch": 0.8799732766047224, + "grad_norm": 3.5137221813201904, + "learning_rate": 7.028101271602716e-06, + "loss": 0.1854, + "step": 37539 + }, + { + "epoch": 0.8799967181795273, + "grad_norm": 4.09668493270874, + "learning_rate": 7.0253887907769654e-06, + "loss": 0.3009, + "step": 37540 + }, + { + "epoch": 0.8800201597543323, + "grad_norm": 2.7461304664611816, + "learning_rate": 7.022676814434059e-06, + "loss": 1.9815, + "step": 37541 + }, + { + "epoch": 0.8800436013291373, + "grad_norm": 1.8026061058044434, + "learning_rate": 7.019965342588742e-06, + "loss": 0.0889, + "step": 37542 + }, + { + "epoch": 0.8800670429039423, + "grad_norm": 3.348966121673584, + "learning_rate": 7.017254375255689e-06, + "loss": 0.1511, + "step": 37543 + }, + { + "epoch": 0.8800904844787473, + "grad_norm": 1.8106036186218262, + "learning_rate": 7.0145439124496334e-06, + "loss": 0.1587, + "step": 37544 + }, + { + "epoch": 0.8801139260535523, + "grad_norm": 3.025844097137451, + "learning_rate": 7.011833954185265e-06, + "loss": 0.3183, + "step": 37545 + }, + { + "epoch": 0.8801373676283573, + "grad_norm": 1.9120945930480957, + "learning_rate": 7.009124500477316e-06, + "loss": 0.0766, + "step": 37546 + }, + { + "epoch": 0.8801608092031623, + "grad_norm": 0.6317944526672363, + "learning_rate": 7.006415551340462e-06, + "loss": 0.0544, + "step": 37547 + }, + { + "epoch": 0.8801842507779672, + "grad_norm": 2.7477216720581055, + "learning_rate": 7.003707106789414e-06, + "loss": 0.1595, + "step": 37548 + }, + { + "epoch": 0.8802076923527723, + "grad_norm": 0.5986048579216003, + "learning_rate": 7.000999166838873e-06, + "loss": 0.0364, + "step": 37549 + }, + { + "epoch": 0.8802311339275772, + "grad_norm": 6.826652526855469, + "learning_rate": 6.998291731503526e-06, + "loss": 0.1514, + "step": 37550 + }, + { + "epoch": 0.8802545755023823, + "grad_norm": 0.7580511569976807, + "learning_rate": 6.99558480079806e-06, + "loss": 0.0262, + "step": 37551 + }, + { + "epoch": 0.8802780170771872, + "grad_norm": 0.7598084807395935, + "learning_rate": 6.992878374737155e-06, + "loss": 0.0497, + "step": 37552 + }, + { + "epoch": 0.8803014586519923, + "grad_norm": 1.2664172649383545, + "learning_rate": 6.990172453335519e-06, + "loss": 0.9583, + "step": 37553 + }, + { + "epoch": 0.8803249002267972, + "grad_norm": 4.018610954284668, + "learning_rate": 6.987467036607831e-06, + "loss": 0.399, + "step": 37554 + }, + { + "epoch": 0.8803483418016023, + "grad_norm": 3.596766948699951, + "learning_rate": 6.984762124568744e-06, + "loss": 0.2632, + "step": 37555 + }, + { + "epoch": 0.8803717833764072, + "grad_norm": 0.44126468896865845, + "learning_rate": 6.982057717232971e-06, + "loss": 0.0361, + "step": 37556 + }, + { + "epoch": 0.8803952249512123, + "grad_norm": 1.5482964515686035, + "learning_rate": 6.979353814615164e-06, + "loss": 0.0689, + "step": 37557 + }, + { + "epoch": 0.8804186665260172, + "grad_norm": 2.0956830978393555, + "learning_rate": 6.97665041672998e-06, + "loss": 0.1163, + "step": 37558 + }, + { + "epoch": 0.8804421081008222, + "grad_norm": 1.5376709699630737, + "learning_rate": 6.973947523592106e-06, + "loss": 0.1559, + "step": 37559 + }, + { + "epoch": 0.8804655496756272, + "grad_norm": 1.7370823621749878, + "learning_rate": 6.97124513521622e-06, + "loss": 1.3992, + "step": 37560 + }, + { + "epoch": 0.8804889912504322, + "grad_norm": 2.0346627235412598, + "learning_rate": 6.968543251616977e-06, + "loss": 0.1286, + "step": 37561 + }, + { + "epoch": 0.8805124328252372, + "grad_norm": 0.9988531470298767, + "learning_rate": 6.965841872809042e-06, + "loss": 0.0804, + "step": 37562 + }, + { + "epoch": 0.8805358744000422, + "grad_norm": 2.124462604522705, + "learning_rate": 6.96314099880705e-06, + "loss": 0.2137, + "step": 37563 + }, + { + "epoch": 0.8805593159748472, + "grad_norm": 1.361069917678833, + "learning_rate": 6.9604406296256645e-06, + "loss": 0.0789, + "step": 37564 + }, + { + "epoch": 0.8805827575496522, + "grad_norm": 0.8891620635986328, + "learning_rate": 6.957740765279563e-06, + "loss": 0.0676, + "step": 37565 + }, + { + "epoch": 0.8806061991244571, + "grad_norm": 2.5277655124664307, + "learning_rate": 6.955041405783369e-06, + "loss": 0.228, + "step": 37566 + }, + { + "epoch": 0.8806296406992622, + "grad_norm": 1.397530198097229, + "learning_rate": 6.952342551151747e-06, + "loss": 0.0959, + "step": 37567 + }, + { + "epoch": 0.8806530822740671, + "grad_norm": 3.3452329635620117, + "learning_rate": 6.949644201399319e-06, + "loss": 0.3309, + "step": 37568 + }, + { + "epoch": 0.8806765238488722, + "grad_norm": 2.622704029083252, + "learning_rate": 6.94694635654074e-06, + "loss": 0.1582, + "step": 37569 + }, + { + "epoch": 0.8806999654236771, + "grad_norm": 1.2605278491973877, + "learning_rate": 6.944249016590631e-06, + "loss": 0.1045, + "step": 37570 + }, + { + "epoch": 0.8807234069984822, + "grad_norm": 0.9182525873184204, + "learning_rate": 6.94155218156366e-06, + "loss": 0.0985, + "step": 37571 + }, + { + "epoch": 0.8807468485732871, + "grad_norm": 4.212023735046387, + "learning_rate": 6.938855851474435e-06, + "loss": 0.241, + "step": 37572 + }, + { + "epoch": 0.8807702901480922, + "grad_norm": 2.461662530899048, + "learning_rate": 6.936160026337601e-06, + "loss": 0.1246, + "step": 37573 + }, + { + "epoch": 0.8807937317228971, + "grad_norm": 1.7850593328475952, + "learning_rate": 6.9334647061677806e-06, + "loss": 0.1696, + "step": 37574 + }, + { + "epoch": 0.8808171732977021, + "grad_norm": 1.3301565647125244, + "learning_rate": 6.930769890979605e-06, + "loss": 0.0703, + "step": 37575 + }, + { + "epoch": 0.8808406148725071, + "grad_norm": 1.3656299114227295, + "learning_rate": 6.928075580787685e-06, + "loss": 0.091, + "step": 37576 + }, + { + "epoch": 0.8808640564473121, + "grad_norm": 1.0193437337875366, + "learning_rate": 6.925381775606621e-06, + "loss": 0.056, + "step": 37577 + }, + { + "epoch": 0.8808874980221171, + "grad_norm": 3.8066844940185547, + "learning_rate": 6.922688475451078e-06, + "loss": 0.3102, + "step": 37578 + }, + { + "epoch": 0.8809109395969221, + "grad_norm": 2.17447566986084, + "learning_rate": 6.919995680335634e-06, + "loss": 0.2005, + "step": 37579 + }, + { + "epoch": 0.8809343811717271, + "grad_norm": 1.091374158859253, + "learning_rate": 6.917303390274921e-06, + "loss": 0.0745, + "step": 37580 + }, + { + "epoch": 0.8809578227465321, + "grad_norm": 3.955169677734375, + "learning_rate": 6.914611605283539e-06, + "loss": 0.4287, + "step": 37581 + }, + { + "epoch": 0.880981264321337, + "grad_norm": 1.6126368045806885, + "learning_rate": 6.911920325376076e-06, + "loss": 0.0756, + "step": 37582 + }, + { + "epoch": 0.8810047058961421, + "grad_norm": 1.1933623552322388, + "learning_rate": 6.9092295505671535e-06, + "loss": 0.165, + "step": 37583 + }, + { + "epoch": 0.881028147470947, + "grad_norm": 1.147104263305664, + "learning_rate": 6.906539280871361e-06, + "loss": 0.1635, + "step": 37584 + }, + { + "epoch": 0.8810515890457521, + "grad_norm": 2.8171958923339844, + "learning_rate": 6.903849516303318e-06, + "loss": 0.1413, + "step": 37585 + }, + { + "epoch": 0.8810750306205571, + "grad_norm": 0.5898934602737427, + "learning_rate": 6.901160256877603e-06, + "loss": 0.0494, + "step": 37586 + }, + { + "epoch": 0.8810984721953621, + "grad_norm": 3.1248812675476074, + "learning_rate": 6.898471502608816e-06, + "loss": 0.3589, + "step": 37587 + }, + { + "epoch": 0.8811219137701671, + "grad_norm": 1.446602702140808, + "learning_rate": 6.8957832535115345e-06, + "loss": 0.0867, + "step": 37588 + }, + { + "epoch": 0.8811453553449721, + "grad_norm": 2.1374058723449707, + "learning_rate": 6.893095509600356e-06, + "loss": 0.275, + "step": 37589 + }, + { + "epoch": 0.8811687969197771, + "grad_norm": 0.18879812955856323, + "learning_rate": 6.890408270889837e-06, + "loss": 0.0071, + "step": 37590 + }, + { + "epoch": 0.8811922384945821, + "grad_norm": 0.9399098753929138, + "learning_rate": 6.887721537394598e-06, + "loss": 0.1166, + "step": 37591 + }, + { + "epoch": 0.8812156800693871, + "grad_norm": 1.1216729879379272, + "learning_rate": 6.885035309129207e-06, + "loss": 0.0944, + "step": 37592 + }, + { + "epoch": 0.881239121644192, + "grad_norm": 3.0038933753967285, + "learning_rate": 6.882349586108227e-06, + "loss": 0.3808, + "step": 37593 + }, + { + "epoch": 0.8812625632189971, + "grad_norm": 2.870683431625366, + "learning_rate": 6.879664368346239e-06, + "loss": 1.6369, + "step": 37594 + }, + { + "epoch": 0.881286004793802, + "grad_norm": 0.8085561990737915, + "learning_rate": 6.876979655857807e-06, + "loss": 0.0674, + "step": 37595 + }, + { + "epoch": 0.8813094463686071, + "grad_norm": 2.1659092903137207, + "learning_rate": 6.8742954486574864e-06, + "loss": 0.1908, + "step": 37596 + }, + { + "epoch": 0.881332887943412, + "grad_norm": 3.7852234840393066, + "learning_rate": 6.871611746759865e-06, + "loss": 0.2839, + "step": 37597 + }, + { + "epoch": 0.8813563295182171, + "grad_norm": 0.5037530660629272, + "learning_rate": 6.86892855017951e-06, + "loss": 0.0433, + "step": 37598 + }, + { + "epoch": 0.881379771093022, + "grad_norm": 0.9291506409645081, + "learning_rate": 6.866245858930964e-06, + "loss": 0.0934, + "step": 37599 + }, + { + "epoch": 0.8814032126678271, + "grad_norm": 3.408921957015991, + "learning_rate": 6.8635636730287835e-06, + "loss": 0.281, + "step": 37600 + }, + { + "epoch": 0.881426654242632, + "grad_norm": 2.315216064453125, + "learning_rate": 6.860881992487522e-06, + "loss": 0.1461, + "step": 37601 + }, + { + "epoch": 0.881450095817437, + "grad_norm": 2.5215187072753906, + "learning_rate": 6.858200817321725e-06, + "loss": 0.1428, + "step": 37602 + }, + { + "epoch": 0.881473537392242, + "grad_norm": 4.3172712326049805, + "learning_rate": 6.855520147545969e-06, + "loss": 0.5055, + "step": 37603 + }, + { + "epoch": 0.881496978967047, + "grad_norm": 3.032693386077881, + "learning_rate": 6.852839983174775e-06, + "loss": 0.2472, + "step": 37604 + }, + { + "epoch": 0.881520420541852, + "grad_norm": 1.5499179363250732, + "learning_rate": 6.850160324222687e-06, + "loss": 0.1173, + "step": 37605 + }, + { + "epoch": 0.881543862116657, + "grad_norm": 2.371443510055542, + "learning_rate": 6.8474811707042505e-06, + "loss": 0.0948, + "step": 37606 + }, + { + "epoch": 0.881567303691462, + "grad_norm": 2.6094019412994385, + "learning_rate": 6.844802522634008e-06, + "loss": 0.1296, + "step": 37607 + }, + { + "epoch": 0.881590745266267, + "grad_norm": 0.9777491092681885, + "learning_rate": 6.84212438002646e-06, + "loss": 0.0691, + "step": 37608 + }, + { + "epoch": 0.881614186841072, + "grad_norm": 2.925830841064453, + "learning_rate": 6.8394467428961715e-06, + "loss": 0.1642, + "step": 37609 + }, + { + "epoch": 0.881637628415877, + "grad_norm": 3.2190353870391846, + "learning_rate": 6.836769611257676e-06, + "loss": 0.4058, + "step": 37610 + }, + { + "epoch": 0.881661069990682, + "grad_norm": 1.9094423055648804, + "learning_rate": 6.834092985125484e-06, + "loss": 1.1575, + "step": 37611 + }, + { + "epoch": 0.881684511565487, + "grad_norm": 1.0386384725570679, + "learning_rate": 6.831416864514128e-06, + "loss": 0.09, + "step": 37612 + }, + { + "epoch": 0.8817079531402919, + "grad_norm": 2.9376461505889893, + "learning_rate": 6.828741249438131e-06, + "loss": 0.2348, + "step": 37613 + }, + { + "epoch": 0.881731394715097, + "grad_norm": 2.013458728790283, + "learning_rate": 6.826066139911991e-06, + "loss": 0.1574, + "step": 37614 + }, + { + "epoch": 0.8817548362899019, + "grad_norm": 1.80964195728302, + "learning_rate": 6.82339153595023e-06, + "loss": 0.0414, + "step": 37615 + }, + { + "epoch": 0.881778277864707, + "grad_norm": 2.8193483352661133, + "learning_rate": 6.820717437567381e-06, + "loss": 0.3545, + "step": 37616 + }, + { + "epoch": 0.8818017194395119, + "grad_norm": 2.8711860179901123, + "learning_rate": 6.818043844777933e-06, + "loss": 0.2171, + "step": 37617 + }, + { + "epoch": 0.881825161014317, + "grad_norm": 3.1062748432159424, + "learning_rate": 6.815370757596407e-06, + "loss": 0.2108, + "step": 37618 + }, + { + "epoch": 0.8818486025891219, + "grad_norm": 2.0317654609680176, + "learning_rate": 6.812698176037291e-06, + "loss": 0.2283, + "step": 37619 + }, + { + "epoch": 0.881872044163927, + "grad_norm": 2.5498576164245605, + "learning_rate": 6.810026100115086e-06, + "loss": 0.1667, + "step": 37620 + }, + { + "epoch": 0.8818954857387319, + "grad_norm": 1.342537522315979, + "learning_rate": 6.807354529844312e-06, + "loss": 0.0287, + "step": 37621 + }, + { + "epoch": 0.8819189273135369, + "grad_norm": 2.7521984577178955, + "learning_rate": 6.804683465239437e-06, + "loss": 0.1065, + "step": 37622 + }, + { + "epoch": 0.8819423688883419, + "grad_norm": 2.5589280128479004, + "learning_rate": 6.8020129063149805e-06, + "loss": 0.1304, + "step": 37623 + }, + { + "epoch": 0.8819658104631469, + "grad_norm": 2.049260377883911, + "learning_rate": 6.799342853085433e-06, + "loss": 0.1724, + "step": 37624 + }, + { + "epoch": 0.8819892520379519, + "grad_norm": 2.1091315746307373, + "learning_rate": 6.7966733055652595e-06, + "loss": 0.172, + "step": 37625 + }, + { + "epoch": 0.8820126936127569, + "grad_norm": 1.225294589996338, + "learning_rate": 6.794004263768972e-06, + "loss": 0.0896, + "step": 37626 + }, + { + "epoch": 0.8820361351875619, + "grad_norm": 3.5563101768493652, + "learning_rate": 6.791335727711023e-06, + "loss": 0.3718, + "step": 37627 + }, + { + "epoch": 0.8820595767623669, + "grad_norm": 3.3588173389434814, + "learning_rate": 6.788667697405904e-06, + "loss": 0.2629, + "step": 37628 + }, + { + "epoch": 0.8820830183371718, + "grad_norm": 1.8686572313308716, + "learning_rate": 6.786000172868101e-06, + "loss": 0.1969, + "step": 37629 + }, + { + "epoch": 0.8821064599119769, + "grad_norm": 2.48193359375, + "learning_rate": 6.783333154112093e-06, + "loss": 0.1989, + "step": 37630 + }, + { + "epoch": 0.8821299014867818, + "grad_norm": 2.818699359893799, + "learning_rate": 6.7806666411523335e-06, + "loss": 0.3433, + "step": 37631 + }, + { + "epoch": 0.8821533430615869, + "grad_norm": 3.7723724842071533, + "learning_rate": 6.7780006340032786e-06, + "loss": 0.2383, + "step": 37632 + }, + { + "epoch": 0.8821767846363918, + "grad_norm": 1.1913663148880005, + "learning_rate": 6.775335132679439e-06, + "loss": 0.1449, + "step": 37633 + }, + { + "epoch": 0.8822002262111969, + "grad_norm": 2.2011115550994873, + "learning_rate": 6.772670137195247e-06, + "loss": 0.1878, + "step": 37634 + }, + { + "epoch": 0.8822236677860018, + "grad_norm": 3.1858439445495605, + "learning_rate": 6.770005647565147e-06, + "loss": 0.2535, + "step": 37635 + }, + { + "epoch": 0.8822471093608069, + "grad_norm": 4.499666690826416, + "learning_rate": 6.767341663803639e-06, + "loss": 0.4117, + "step": 37636 + }, + { + "epoch": 0.8822705509356119, + "grad_norm": 1.4987609386444092, + "learning_rate": 6.764678185925155e-06, + "loss": 0.0349, + "step": 37637 + }, + { + "epoch": 0.8822939925104168, + "grad_norm": 3.1912877559661865, + "learning_rate": 6.762015213944151e-06, + "loss": 0.2631, + "step": 37638 + }, + { + "epoch": 0.8823174340852219, + "grad_norm": 4.169366359710693, + "learning_rate": 6.7593527478750696e-06, + "loss": 0.2179, + "step": 37639 + }, + { + "epoch": 0.8823408756600268, + "grad_norm": 2.756723165512085, + "learning_rate": 6.756690787732356e-06, + "loss": 0.1732, + "step": 37640 + }, + { + "epoch": 0.8823643172348319, + "grad_norm": 1.6716724634170532, + "learning_rate": 6.754029333530465e-06, + "loss": 0.0902, + "step": 37641 + }, + { + "epoch": 0.8823877588096368, + "grad_norm": 0.7002871036529541, + "learning_rate": 6.75136838528384e-06, + "loss": 0.0386, + "step": 37642 + }, + { + "epoch": 0.8824112003844419, + "grad_norm": 5.098428249359131, + "learning_rate": 6.748707943006904e-06, + "loss": 0.616, + "step": 37643 + }, + { + "epoch": 0.8824346419592468, + "grad_norm": 5.371604919433594, + "learning_rate": 6.74604800671409e-06, + "loss": 0.474, + "step": 37644 + }, + { + "epoch": 0.8824580835340519, + "grad_norm": 1.5390466451644897, + "learning_rate": 6.743388576419851e-06, + "loss": 1.0889, + "step": 37645 + }, + { + "epoch": 0.8824815251088568, + "grad_norm": 0.9557821750640869, + "learning_rate": 6.740729652138622e-06, + "loss": 0.1096, + "step": 37646 + }, + { + "epoch": 0.8825049666836619, + "grad_norm": 0.9246702194213867, + "learning_rate": 6.73807123388479e-06, + "loss": 0.0465, + "step": 37647 + }, + { + "epoch": 0.8825284082584668, + "grad_norm": 1.7290946245193481, + "learning_rate": 6.735413321672823e-06, + "loss": 0.1156, + "step": 37648 + }, + { + "epoch": 0.8825518498332718, + "grad_norm": 2.7721166610717773, + "learning_rate": 6.73275591551713e-06, + "loss": 0.2284, + "step": 37649 + }, + { + "epoch": 0.8825752914080768, + "grad_norm": 1.8828022480010986, + "learning_rate": 6.730099015432123e-06, + "loss": 1.6608, + "step": 37650 + }, + { + "epoch": 0.8825987329828818, + "grad_norm": 1.5682271718978882, + "learning_rate": 6.727442621432212e-06, + "loss": 0.1378, + "step": 37651 + }, + { + "epoch": 0.8826221745576868, + "grad_norm": 0.6520892977714539, + "learning_rate": 6.7247867335318296e-06, + "loss": 0.0379, + "step": 37652 + }, + { + "epoch": 0.8826456161324918, + "grad_norm": 0.9448367953300476, + "learning_rate": 6.722131351745365e-06, + "loss": 0.0211, + "step": 37653 + }, + { + "epoch": 0.8826690577072968, + "grad_norm": 0.9823487997055054, + "learning_rate": 6.71947647608725e-06, + "loss": 0.1137, + "step": 37654 + }, + { + "epoch": 0.8826924992821018, + "grad_norm": 1.1627225875854492, + "learning_rate": 6.716822106571874e-06, + "loss": 0.1031, + "step": 37655 + }, + { + "epoch": 0.8827159408569067, + "grad_norm": 2.191897392272949, + "learning_rate": 6.714168243213648e-06, + "loss": 0.1703, + "step": 37656 + }, + { + "epoch": 0.8827393824317118, + "grad_norm": 2.982412815093994, + "learning_rate": 6.711514886026948e-06, + "loss": 0.3786, + "step": 37657 + }, + { + "epoch": 0.8827628240065167, + "grad_norm": 1.9556208848953247, + "learning_rate": 6.70886203502622e-06, + "loss": 0.0612, + "step": 37658 + }, + { + "epoch": 0.8827862655813218, + "grad_norm": 1.375540852546692, + "learning_rate": 6.706209690225818e-06, + "loss": 0.0675, + "step": 37659 + }, + { + "epoch": 0.8828097071561267, + "grad_norm": 4.049326419830322, + "learning_rate": 6.70355785164013e-06, + "loss": 0.3776, + "step": 37660 + }, + { + "epoch": 0.8828331487309318, + "grad_norm": 2.6434149742126465, + "learning_rate": 6.700906519283578e-06, + "loss": 0.0814, + "step": 37661 + }, + { + "epoch": 0.8828565903057367, + "grad_norm": 2.96744441986084, + "learning_rate": 6.69825569317053e-06, + "loss": 0.1931, + "step": 37662 + }, + { + "epoch": 0.8828800318805418, + "grad_norm": 0.6280543804168701, + "learning_rate": 6.695605373315372e-06, + "loss": 0.0678, + "step": 37663 + }, + { + "epoch": 0.8829034734553467, + "grad_norm": 1.0171126127243042, + "learning_rate": 6.692955559732483e-06, + "loss": 0.0549, + "step": 37664 + }, + { + "epoch": 0.8829269150301517, + "grad_norm": 1.70745050907135, + "learning_rate": 6.69030625243624e-06, + "loss": 0.1207, + "step": 37665 + }, + { + "epoch": 0.8829503566049567, + "grad_norm": 3.4730565547943115, + "learning_rate": 6.687657451440999e-06, + "loss": 1.5253, + "step": 37666 + }, + { + "epoch": 0.8829737981797617, + "grad_norm": 0.8713648915290833, + "learning_rate": 6.6850091567611795e-06, + "loss": 0.0554, + "step": 37667 + }, + { + "epoch": 0.8829972397545667, + "grad_norm": 1.8257312774658203, + "learning_rate": 6.682361368411116e-06, + "loss": 0.1714, + "step": 37668 + }, + { + "epoch": 0.8830206813293717, + "grad_norm": 2.024132251739502, + "learning_rate": 6.679714086405164e-06, + "loss": 0.109, + "step": 37669 + }, + { + "epoch": 0.8830441229041767, + "grad_norm": 2.035778045654297, + "learning_rate": 6.677067310757734e-06, + "loss": 0.1664, + "step": 37670 + }, + { + "epoch": 0.8830675644789817, + "grad_norm": 2.289684772491455, + "learning_rate": 6.674421041483148e-06, + "loss": 0.2331, + "step": 37671 + }, + { + "epoch": 0.8830910060537867, + "grad_norm": 2.156919240951538, + "learning_rate": 6.671775278595782e-06, + "loss": 0.1498, + "step": 37672 + }, + { + "epoch": 0.8831144476285917, + "grad_norm": 2.714475393295288, + "learning_rate": 6.669130022109981e-06, + "loss": 0.3676, + "step": 37673 + }, + { + "epoch": 0.8831378892033966, + "grad_norm": 2.898583173751831, + "learning_rate": 6.6664852720401125e-06, + "loss": 0.1734, + "step": 37674 + }, + { + "epoch": 0.8831613307782017, + "grad_norm": 2.902639865875244, + "learning_rate": 6.663841028400519e-06, + "loss": 0.1531, + "step": 37675 + }, + { + "epoch": 0.8831847723530066, + "grad_norm": 2.293539524078369, + "learning_rate": 6.6611972912055455e-06, + "loss": 0.1385, + "step": 37676 + }, + { + "epoch": 0.8832082139278117, + "grad_norm": 1.8777414560317993, + "learning_rate": 6.658554060469546e-06, + "loss": 0.2065, + "step": 37677 + }, + { + "epoch": 0.8832316555026166, + "grad_norm": 3.2890732288360596, + "learning_rate": 6.6559113362068324e-06, + "loss": 0.1427, + "step": 37678 + }, + { + "epoch": 0.8832550970774217, + "grad_norm": 3.405041456222534, + "learning_rate": 6.653269118431793e-06, + "loss": 0.3397, + "step": 37679 + }, + { + "epoch": 0.8832785386522266, + "grad_norm": 2.593125581741333, + "learning_rate": 6.6506274071587385e-06, + "loss": 1.3592, + "step": 37680 + }, + { + "epoch": 0.8833019802270317, + "grad_norm": 0.5133357644081116, + "learning_rate": 6.64798620240199e-06, + "loss": 0.0351, + "step": 37681 + }, + { + "epoch": 0.8833254218018366, + "grad_norm": 2.1476664543151855, + "learning_rate": 6.645345504175904e-06, + "loss": 0.2275, + "step": 37682 + }, + { + "epoch": 0.8833488633766416, + "grad_norm": 3.32724666595459, + "learning_rate": 6.642705312494802e-06, + "loss": 0.2266, + "step": 37683 + }, + { + "epoch": 0.8833723049514466, + "grad_norm": 0.856539785861969, + "learning_rate": 6.640065627373005e-06, + "loss": 0.0705, + "step": 37684 + }, + { + "epoch": 0.8833957465262516, + "grad_norm": 0.22936619818210602, + "learning_rate": 6.637426448824824e-06, + "loss": 0.0332, + "step": 37685 + }, + { + "epoch": 0.8834191881010566, + "grad_norm": 3.142927646636963, + "learning_rate": 6.634787776864604e-06, + "loss": 0.3093, + "step": 37686 + }, + { + "epoch": 0.8834426296758616, + "grad_norm": 5.025412559509277, + "learning_rate": 6.632149611506644e-06, + "loss": 0.1447, + "step": 37687 + }, + { + "epoch": 0.8834660712506667, + "grad_norm": 2.507082462310791, + "learning_rate": 6.629511952765277e-06, + "loss": 0.2724, + "step": 37688 + }, + { + "epoch": 0.8834895128254716, + "grad_norm": 4.5883073806762695, + "learning_rate": 6.626874800654803e-06, + "loss": 0.341, + "step": 37689 + }, + { + "epoch": 0.8835129544002767, + "grad_norm": 1.5999565124511719, + "learning_rate": 6.6242381551895216e-06, + "loss": 0.1291, + "step": 37690 + }, + { + "epoch": 0.8835363959750816, + "grad_norm": 0.9096196293830872, + "learning_rate": 6.621602016383743e-06, + "loss": 0.0471, + "step": 37691 + }, + { + "epoch": 0.8835598375498867, + "grad_norm": 2.7612247467041016, + "learning_rate": 6.61896638425179e-06, + "loss": 0.1421, + "step": 37692 + }, + { + "epoch": 0.8835832791246916, + "grad_norm": 2.829083204269409, + "learning_rate": 6.616331258807939e-06, + "loss": 0.1466, + "step": 37693 + }, + { + "epoch": 0.8836067206994966, + "grad_norm": 1.401741862297058, + "learning_rate": 6.613696640066514e-06, + "loss": 0.1591, + "step": 37694 + }, + { + "epoch": 0.8836301622743016, + "grad_norm": 1.1283106803894043, + "learning_rate": 6.611062528041789e-06, + "loss": 0.0896, + "step": 37695 + }, + { + "epoch": 0.8836536038491066, + "grad_norm": 1.4958285093307495, + "learning_rate": 6.608428922748078e-06, + "loss": 0.0714, + "step": 37696 + }, + { + "epoch": 0.8836770454239116, + "grad_norm": 2.6727652549743652, + "learning_rate": 6.605795824199645e-06, + "loss": 0.0808, + "step": 37697 + }, + { + "epoch": 0.8837004869987166, + "grad_norm": 3.3849875926971436, + "learning_rate": 6.603163232410781e-06, + "loss": 0.3223, + "step": 37698 + }, + { + "epoch": 0.8837239285735216, + "grad_norm": 2.7737250328063965, + "learning_rate": 6.600531147395794e-06, + "loss": 1.7059, + "step": 37699 + }, + { + "epoch": 0.8837473701483266, + "grad_norm": 3.0585336685180664, + "learning_rate": 6.597899569168953e-06, + "loss": 0.1542, + "step": 37700 + }, + { + "epoch": 0.8837708117231315, + "grad_norm": 1.1859265565872192, + "learning_rate": 6.595268497744533e-06, + "loss": 0.0901, + "step": 37701 + }, + { + "epoch": 0.8837942532979366, + "grad_norm": 3.4301438331604004, + "learning_rate": 6.592637933136814e-06, + "loss": 0.1367, + "step": 37702 + }, + { + "epoch": 0.8838176948727415, + "grad_norm": 3.0344860553741455, + "learning_rate": 6.5900078753600715e-06, + "loss": 0.2384, + "step": 37703 + }, + { + "epoch": 0.8838411364475466, + "grad_norm": 2.166527271270752, + "learning_rate": 6.58737832442855e-06, + "loss": 0.0861, + "step": 37704 + }, + { + "epoch": 0.8838645780223515, + "grad_norm": 6.227917194366455, + "learning_rate": 6.584749280356539e-06, + "loss": 0.385, + "step": 37705 + }, + { + "epoch": 0.8838880195971566, + "grad_norm": 2.737635374069214, + "learning_rate": 6.582120743158326e-06, + "loss": 0.1296, + "step": 37706 + }, + { + "epoch": 0.8839114611719615, + "grad_norm": 0.6641346216201782, + "learning_rate": 6.579492712848156e-06, + "loss": 0.0533, + "step": 37707 + }, + { + "epoch": 0.8839349027467666, + "grad_norm": 3.5846445560455322, + "learning_rate": 6.576865189440274e-06, + "loss": 0.2918, + "step": 37708 + }, + { + "epoch": 0.8839583443215715, + "grad_norm": 3.626237630844116, + "learning_rate": 6.574238172948954e-06, + "loss": 0.1279, + "step": 37709 + }, + { + "epoch": 0.8839817858963765, + "grad_norm": 3.284236431121826, + "learning_rate": 6.571611663388444e-06, + "loss": 0.2878, + "step": 37710 + }, + { + "epoch": 0.8840052274711815, + "grad_norm": 2.712106466293335, + "learning_rate": 6.568985660772986e-06, + "loss": 0.3133, + "step": 37711 + }, + { + "epoch": 0.8840286690459865, + "grad_norm": 3.5824074745178223, + "learning_rate": 6.566360165116847e-06, + "loss": 0.1882, + "step": 37712 + }, + { + "epoch": 0.8840521106207915, + "grad_norm": 2.634533643722534, + "learning_rate": 6.563735176434271e-06, + "loss": 0.1868, + "step": 37713 + }, + { + "epoch": 0.8840755521955965, + "grad_norm": 0.9288517236709595, + "learning_rate": 6.561110694739492e-06, + "loss": 0.0706, + "step": 37714 + }, + { + "epoch": 0.8840989937704015, + "grad_norm": 2.801273822784424, + "learning_rate": 6.558486720046753e-06, + "loss": 0.1677, + "step": 37715 + }, + { + "epoch": 0.8841224353452065, + "grad_norm": 3.485368490219116, + "learning_rate": 6.555863252370276e-06, + "loss": 1.3104, + "step": 37716 + }, + { + "epoch": 0.8841458769200115, + "grad_norm": 3.6011767387390137, + "learning_rate": 6.5532402917243275e-06, + "loss": 0.3406, + "step": 37717 + }, + { + "epoch": 0.8841693184948165, + "grad_norm": 2.0313456058502197, + "learning_rate": 6.550617838123119e-06, + "loss": 0.096, + "step": 37718 + }, + { + "epoch": 0.8841927600696214, + "grad_norm": 2.136974334716797, + "learning_rate": 6.547995891580894e-06, + "loss": 1.4709, + "step": 37719 + }, + { + "epoch": 0.8842162016444265, + "grad_norm": 0.8269014358520508, + "learning_rate": 6.5453744521118635e-06, + "loss": 0.0624, + "step": 37720 + }, + { + "epoch": 0.8842396432192314, + "grad_norm": 3.022805690765381, + "learning_rate": 6.542753519730271e-06, + "loss": 0.1078, + "step": 37721 + }, + { + "epoch": 0.8842630847940365, + "grad_norm": 0.9182561635971069, + "learning_rate": 6.540133094450318e-06, + "loss": 0.0413, + "step": 37722 + }, + { + "epoch": 0.8842865263688414, + "grad_norm": 0.7833830714225769, + "learning_rate": 6.537513176286214e-06, + "loss": 0.061, + "step": 37723 + }, + { + "epoch": 0.8843099679436465, + "grad_norm": 3.4105114936828613, + "learning_rate": 6.534893765252214e-06, + "loss": 0.1421, + "step": 37724 + }, + { + "epoch": 0.8843334095184514, + "grad_norm": 1.5133103132247925, + "learning_rate": 6.532274861362508e-06, + "loss": 0.1847, + "step": 37725 + }, + { + "epoch": 0.8843568510932565, + "grad_norm": 2.666924238204956, + "learning_rate": 6.529656464631295e-06, + "loss": 0.1589, + "step": 37726 + }, + { + "epoch": 0.8843802926680614, + "grad_norm": 1.9556212425231934, + "learning_rate": 6.527038575072808e-06, + "loss": 1.4406, + "step": 37727 + }, + { + "epoch": 0.8844037342428664, + "grad_norm": 1.3577972650527954, + "learning_rate": 6.524421192701235e-06, + "loss": 0.0937, + "step": 37728 + }, + { + "epoch": 0.8844271758176714, + "grad_norm": 3.6033174991607666, + "learning_rate": 6.521804317530755e-06, + "loss": 0.2665, + "step": 37729 + }, + { + "epoch": 0.8844506173924764, + "grad_norm": 1.321400761604309, + "learning_rate": 6.519187949575612e-06, + "loss": 0.1507, + "step": 37730 + }, + { + "epoch": 0.8844740589672814, + "grad_norm": 2.903770685195923, + "learning_rate": 6.5165720888499814e-06, + "loss": 0.1964, + "step": 37731 + }, + { + "epoch": 0.8844975005420864, + "grad_norm": 2.6963977813720703, + "learning_rate": 6.5139567353680656e-06, + "loss": 0.2048, + "step": 37732 + }, + { + "epoch": 0.8845209421168914, + "grad_norm": 0.6496042013168335, + "learning_rate": 6.511341889144051e-06, + "loss": 0.0388, + "step": 37733 + }, + { + "epoch": 0.8845443836916964, + "grad_norm": 0.4552983343601227, + "learning_rate": 6.5087275501921285e-06, + "loss": 0.0258, + "step": 37734 + }, + { + "epoch": 0.8845678252665014, + "grad_norm": 3.4178097248077393, + "learning_rate": 6.5061137185264634e-06, + "loss": 0.1818, + "step": 37735 + }, + { + "epoch": 0.8845912668413064, + "grad_norm": 2.1269466876983643, + "learning_rate": 6.503500394161255e-06, + "loss": 0.0862, + "step": 37736 + }, + { + "epoch": 0.8846147084161113, + "grad_norm": 2.21445894241333, + "learning_rate": 6.500887577110693e-06, + "loss": 0.1283, + "step": 37737 + }, + { + "epoch": 0.8846381499909164, + "grad_norm": 1.9803014993667603, + "learning_rate": 6.4982752673889426e-06, + "loss": 0.0815, + "step": 37738 + }, + { + "epoch": 0.8846615915657214, + "grad_norm": 1.2233330011367798, + "learning_rate": 6.4956634650101714e-06, + "loss": 0.1138, + "step": 37739 + }, + { + "epoch": 0.8846850331405264, + "grad_norm": 2.6893701553344727, + "learning_rate": 6.493052169988567e-06, + "loss": 0.1933, + "step": 37740 + }, + { + "epoch": 0.8847084747153314, + "grad_norm": 4.2271952629089355, + "learning_rate": 6.490441382338297e-06, + "loss": 0.2569, + "step": 37741 + }, + { + "epoch": 0.8847319162901364, + "grad_norm": 2.5574636459350586, + "learning_rate": 6.487831102073494e-06, + "loss": 1.3593, + "step": 37742 + }, + { + "epoch": 0.8847553578649414, + "grad_norm": 2.4623289108276367, + "learning_rate": 6.485221329208357e-06, + "loss": 0.2128, + "step": 37743 + }, + { + "epoch": 0.8847787994397464, + "grad_norm": 1.199683666229248, + "learning_rate": 6.482612063757054e-06, + "loss": 0.1238, + "step": 37744 + }, + { + "epoch": 0.8848022410145514, + "grad_norm": 1.8483389616012573, + "learning_rate": 6.480003305733728e-06, + "loss": 0.1832, + "step": 37745 + }, + { + "epoch": 0.8848256825893563, + "grad_norm": 2.299267530441284, + "learning_rate": 6.477395055152524e-06, + "loss": 0.2233, + "step": 37746 + }, + { + "epoch": 0.8848491241641614, + "grad_norm": 0.5447219610214233, + "learning_rate": 6.474787312027608e-06, + "loss": 0.0196, + "step": 37747 + }, + { + "epoch": 0.8848725657389663, + "grad_norm": 0.6007869243621826, + "learning_rate": 6.472180076373136e-06, + "loss": 0.0405, + "step": 37748 + }, + { + "epoch": 0.8848960073137714, + "grad_norm": 1.9322675466537476, + "learning_rate": 6.469573348203217e-06, + "loss": 0.1165, + "step": 37749 + }, + { + "epoch": 0.8849194488885763, + "grad_norm": 3.1669387817382812, + "learning_rate": 6.466967127532042e-06, + "loss": 0.1475, + "step": 37750 + }, + { + "epoch": 0.8849428904633814, + "grad_norm": 1.364654302597046, + "learning_rate": 6.464361414373732e-06, + "loss": 0.0525, + "step": 37751 + }, + { + "epoch": 0.8849663320381863, + "grad_norm": 2.5165693759918213, + "learning_rate": 6.461756208742431e-06, + "loss": 0.2544, + "step": 37752 + }, + { + "epoch": 0.8849897736129914, + "grad_norm": 2.8966071605682373, + "learning_rate": 6.4591515106522725e-06, + "loss": 0.2961, + "step": 37753 + }, + { + "epoch": 0.8850132151877963, + "grad_norm": 1.750998616218567, + "learning_rate": 6.456547320117368e-06, + "loss": 1.3474, + "step": 37754 + }, + { + "epoch": 0.8850366567626013, + "grad_norm": 2.011854887008667, + "learning_rate": 6.453943637151882e-06, + "loss": 0.1315, + "step": 37755 + }, + { + "epoch": 0.8850600983374063, + "grad_norm": 0.6500109434127808, + "learning_rate": 6.451340461769917e-06, + "loss": 0.0286, + "step": 37756 + }, + { + "epoch": 0.8850835399122113, + "grad_norm": 2.840923309326172, + "learning_rate": 6.448737793985626e-06, + "loss": 0.3287, + "step": 37757 + }, + { + "epoch": 0.8851069814870163, + "grad_norm": 0.7071940898895264, + "learning_rate": 6.446135633813111e-06, + "loss": 0.0755, + "step": 37758 + }, + { + "epoch": 0.8851304230618213, + "grad_norm": 2.5272765159606934, + "learning_rate": 6.4435339812664915e-06, + "loss": 1.2463, + "step": 37759 + }, + { + "epoch": 0.8851538646366263, + "grad_norm": 0.6992717981338501, + "learning_rate": 6.440932836359892e-06, + "loss": 0.0362, + "step": 37760 + }, + { + "epoch": 0.8851773062114313, + "grad_norm": 3.348771572113037, + "learning_rate": 6.4383321991074e-06, + "loss": 0.1324, + "step": 37761 + }, + { + "epoch": 0.8852007477862363, + "grad_norm": 2.976560354232788, + "learning_rate": 6.43573206952317e-06, + "loss": 0.1937, + "step": 37762 + }, + { + "epoch": 0.8852241893610413, + "grad_norm": 2.4176690578460693, + "learning_rate": 6.433132447621282e-06, + "loss": 0.1787, + "step": 37763 + }, + { + "epoch": 0.8852476309358462, + "grad_norm": 1.2762861251831055, + "learning_rate": 6.430533333415856e-06, + "loss": 0.0906, + "step": 37764 + }, + { + "epoch": 0.8852710725106513, + "grad_norm": 2.427328109741211, + "learning_rate": 6.427934726920981e-06, + "loss": 0.1395, + "step": 37765 + }, + { + "epoch": 0.8852945140854562, + "grad_norm": 3.5938892364501953, + "learning_rate": 6.425336628150758e-06, + "loss": 0.3157, + "step": 37766 + }, + { + "epoch": 0.8853179556602613, + "grad_norm": 2.8766353130340576, + "learning_rate": 6.4227390371192965e-06, + "loss": 0.3353, + "step": 37767 + }, + { + "epoch": 0.8853413972350662, + "grad_norm": 2.3414700031280518, + "learning_rate": 6.420141953840664e-06, + "loss": 0.084, + "step": 37768 + }, + { + "epoch": 0.8853648388098713, + "grad_norm": 0.7461988925933838, + "learning_rate": 6.417545378328993e-06, + "loss": 0.0312, + "step": 37769 + }, + { + "epoch": 0.8853882803846762, + "grad_norm": 1.0040910243988037, + "learning_rate": 6.4149493105983506e-06, + "loss": 0.0842, + "step": 37770 + }, + { + "epoch": 0.8854117219594813, + "grad_norm": 3.1133790016174316, + "learning_rate": 6.412353750662836e-06, + "loss": 0.1959, + "step": 37771 + }, + { + "epoch": 0.8854351635342862, + "grad_norm": 2.1670687198638916, + "learning_rate": 6.4097586985365055e-06, + "loss": 0.1742, + "step": 37772 + }, + { + "epoch": 0.8854586051090912, + "grad_norm": 2.051591157913208, + "learning_rate": 6.407164154233469e-06, + "loss": 0.2116, + "step": 37773 + }, + { + "epoch": 0.8854820466838962, + "grad_norm": 1.5842777490615845, + "learning_rate": 6.404570117767761e-06, + "loss": 0.0593, + "step": 37774 + }, + { + "epoch": 0.8855054882587012, + "grad_norm": 3.2510061264038086, + "learning_rate": 6.401976589153513e-06, + "loss": 0.2227, + "step": 37775 + }, + { + "epoch": 0.8855289298335062, + "grad_norm": 3.2311646938323975, + "learning_rate": 6.3993835684047706e-06, + "loss": 0.1805, + "step": 37776 + }, + { + "epoch": 0.8855523714083112, + "grad_norm": 0.5675491690635681, + "learning_rate": 6.396791055535601e-06, + "loss": 0.0434, + "step": 37777 + }, + { + "epoch": 0.8855758129831162, + "grad_norm": 2.4385123252868652, + "learning_rate": 6.394199050560068e-06, + "loss": 0.129, + "step": 37778 + }, + { + "epoch": 0.8855992545579212, + "grad_norm": 1.2187496423721313, + "learning_rate": 6.391607553492251e-06, + "loss": 0.1257, + "step": 37779 + }, + { + "epoch": 0.8856226961327262, + "grad_norm": 3.039910078048706, + "learning_rate": 6.389016564346207e-06, + "loss": 0.1561, + "step": 37780 + }, + { + "epoch": 0.8856461377075312, + "grad_norm": 2.380941152572632, + "learning_rate": 6.386426083135977e-06, + "loss": 0.1027, + "step": 37781 + }, + { + "epoch": 0.8856695792823361, + "grad_norm": 2.4914638996124268, + "learning_rate": 6.383836109875641e-06, + "loss": 0.2362, + "step": 37782 + }, + { + "epoch": 0.8856930208571412, + "grad_norm": 2.2266101837158203, + "learning_rate": 6.381246644579242e-06, + "loss": 0.1583, + "step": 37783 + }, + { + "epoch": 0.8857164624319461, + "grad_norm": 2.7872273921966553, + "learning_rate": 6.378657687260836e-06, + "loss": 0.1807, + "step": 37784 + }, + { + "epoch": 0.8857399040067512, + "grad_norm": 3.188511371612549, + "learning_rate": 6.376069237934468e-06, + "loss": 0.3165, + "step": 37785 + }, + { + "epoch": 0.8857633455815561, + "grad_norm": 1.6470612287521362, + "learning_rate": 6.37348129661417e-06, + "loss": 0.1216, + "step": 37786 + }, + { + "epoch": 0.8857867871563612, + "grad_norm": 1.1929775476455688, + "learning_rate": 6.370893863313987e-06, + "loss": 0.0633, + "step": 37787 + }, + { + "epoch": 0.8858102287311661, + "grad_norm": 2.4141860008239746, + "learning_rate": 6.368306938047985e-06, + "loss": 0.0957, + "step": 37788 + }, + { + "epoch": 0.8858336703059712, + "grad_norm": 2.548264741897583, + "learning_rate": 6.365720520830176e-06, + "loss": 0.2005, + "step": 37789 + }, + { + "epoch": 0.8858571118807762, + "grad_norm": 3.6758079528808594, + "learning_rate": 6.363134611674593e-06, + "loss": 0.2302, + "step": 37790 + }, + { + "epoch": 0.8858805534555811, + "grad_norm": 1.6527235507965088, + "learning_rate": 6.3605492105952794e-06, + "loss": 1.1039, + "step": 37791 + }, + { + "epoch": 0.8859039950303862, + "grad_norm": 4.178040027618408, + "learning_rate": 6.357964317606269e-06, + "loss": 0.3491, + "step": 37792 + }, + { + "epoch": 0.8859274366051911, + "grad_norm": 3.645934820175171, + "learning_rate": 6.355379932721561e-06, + "loss": 0.2347, + "step": 37793 + }, + { + "epoch": 0.8859508781799962, + "grad_norm": 0.7994982600212097, + "learning_rate": 6.35279605595519e-06, + "loss": 0.0188, + "step": 37794 + }, + { + "epoch": 0.8859743197548011, + "grad_norm": 3.1029603481292725, + "learning_rate": 6.3502126873212e-06, + "loss": 0.1466, + "step": 37795 + }, + { + "epoch": 0.8859977613296062, + "grad_norm": 1.217498779296875, + "learning_rate": 6.34762982683359e-06, + "loss": 1.2291, + "step": 37796 + }, + { + "epoch": 0.8860212029044111, + "grad_norm": 3.4257559776306152, + "learning_rate": 6.34504747450636e-06, + "loss": 0.3478, + "step": 37797 + }, + { + "epoch": 0.8860446444792162, + "grad_norm": 2.4263837337493896, + "learning_rate": 6.342465630353556e-06, + "loss": 1.0868, + "step": 37798 + }, + { + "epoch": 0.8860680860540211, + "grad_norm": 1.417846441268921, + "learning_rate": 6.339884294389131e-06, + "loss": 0.0852, + "step": 37799 + }, + { + "epoch": 0.8860915276288261, + "grad_norm": 2.662513017654419, + "learning_rate": 6.337303466627153e-06, + "loss": 0.1922, + "step": 37800 + }, + { + "epoch": 0.8861149692036311, + "grad_norm": 3.4262542724609375, + "learning_rate": 6.3347231470816005e-06, + "loss": 0.2956, + "step": 37801 + }, + { + "epoch": 0.8861384107784361, + "grad_norm": 0.7703806161880493, + "learning_rate": 6.332143335766472e-06, + "loss": 0.3029, + "step": 37802 + }, + { + "epoch": 0.8861618523532411, + "grad_norm": 3.4182322025299072, + "learning_rate": 6.3295640326957566e-06, + "loss": 0.4158, + "step": 37803 + }, + { + "epoch": 0.8861852939280461, + "grad_norm": 3.268110752105713, + "learning_rate": 6.326985237883465e-06, + "loss": 0.2664, + "step": 37804 + }, + { + "epoch": 0.8862087355028511, + "grad_norm": 4.745784759521484, + "learning_rate": 6.3244069513435975e-06, + "loss": 0.2872, + "step": 37805 + }, + { + "epoch": 0.8862321770776561, + "grad_norm": 0.5521789789199829, + "learning_rate": 6.32182917309011e-06, + "loss": 0.0171, + "step": 37806 + }, + { + "epoch": 0.886255618652461, + "grad_norm": 3.577758312225342, + "learning_rate": 6.319251903137025e-06, + "loss": 0.215, + "step": 37807 + }, + { + "epoch": 0.8862790602272661, + "grad_norm": 3.666712760925293, + "learning_rate": 6.316675141498318e-06, + "loss": 0.3083, + "step": 37808 + }, + { + "epoch": 0.886302501802071, + "grad_norm": 1.8254996538162231, + "learning_rate": 6.314098888187969e-06, + "loss": 1.4038, + "step": 37809 + }, + { + "epoch": 0.8863259433768761, + "grad_norm": 2.036569118499756, + "learning_rate": 6.311523143219944e-06, + "loss": 0.1732, + "step": 37810 + }, + { + "epoch": 0.886349384951681, + "grad_norm": 0.6565904021263123, + "learning_rate": 6.308947906608231e-06, + "loss": 0.0894, + "step": 37811 + }, + { + "epoch": 0.8863728265264861, + "grad_norm": 1.0713032484054565, + "learning_rate": 6.306373178366787e-06, + "loss": 0.7858, + "step": 37812 + }, + { + "epoch": 0.886396268101291, + "grad_norm": 3.287571430206299, + "learning_rate": 6.30379895850961e-06, + "loss": 0.2054, + "step": 37813 + }, + { + "epoch": 0.8864197096760961, + "grad_norm": 1.9360706806182861, + "learning_rate": 6.301225247050668e-06, + "loss": 1.5858, + "step": 37814 + }, + { + "epoch": 0.886443151250901, + "grad_norm": 2.541550636291504, + "learning_rate": 6.298652044003883e-06, + "loss": 0.154, + "step": 37815 + }, + { + "epoch": 0.8864665928257061, + "grad_norm": 2.6809628009796143, + "learning_rate": 6.2960793493832656e-06, + "loss": 1.3826, + "step": 37816 + }, + { + "epoch": 0.886490034400511, + "grad_norm": 2.8744237422943115, + "learning_rate": 6.29350716320275e-06, + "loss": 0.1774, + "step": 37817 + }, + { + "epoch": 0.886513475975316, + "grad_norm": 2.1219050884246826, + "learning_rate": 6.290935485476312e-06, + "loss": 0.1273, + "step": 37818 + }, + { + "epoch": 0.886536917550121, + "grad_norm": 0.4357074797153473, + "learning_rate": 6.288364316217865e-06, + "loss": 0.0333, + "step": 37819 + }, + { + "epoch": 0.886560359124926, + "grad_norm": 0.6031304001808167, + "learning_rate": 6.285793655441408e-06, + "loss": 0.0594, + "step": 37820 + }, + { + "epoch": 0.886583800699731, + "grad_norm": 1.7793561220169067, + "learning_rate": 6.283223503160873e-06, + "loss": 0.0813, + "step": 37821 + }, + { + "epoch": 0.886607242274536, + "grad_norm": 0.7602694630622864, + "learning_rate": 6.280653859390206e-06, + "loss": 0.073, + "step": 37822 + }, + { + "epoch": 0.886630683849341, + "grad_norm": 2.548365354537964, + "learning_rate": 6.27808472414334e-06, + "loss": 0.1756, + "step": 37823 + }, + { + "epoch": 0.886654125424146, + "grad_norm": 3.551210880279541, + "learning_rate": 6.27551609743422e-06, + "loss": 0.3158, + "step": 37824 + }, + { + "epoch": 0.886677566998951, + "grad_norm": 5.156259059906006, + "learning_rate": 6.2729479792767665e-06, + "loss": 0.226, + "step": 37825 + }, + { + "epoch": 0.886701008573756, + "grad_norm": 1.7223483324050903, + "learning_rate": 6.270380369684959e-06, + "loss": 0.1195, + "step": 37826 + }, + { + "epoch": 0.8867244501485609, + "grad_norm": 1.8772917985916138, + "learning_rate": 6.2678132686726846e-06, + "loss": 0.1431, + "step": 37827 + }, + { + "epoch": 0.886747891723366, + "grad_norm": 1.6225502490997314, + "learning_rate": 6.2652466762539e-06, + "loss": 0.8844, + "step": 37828 + }, + { + "epoch": 0.8867713332981709, + "grad_norm": 1.244866967201233, + "learning_rate": 6.2626805924425265e-06, + "loss": 0.05, + "step": 37829 + }, + { + "epoch": 0.886794774872976, + "grad_norm": 1.472852349281311, + "learning_rate": 6.260115017252488e-06, + "loss": 0.1298, + "step": 37830 + }, + { + "epoch": 0.8868182164477809, + "grad_norm": 2.325310230255127, + "learning_rate": 6.257549950697694e-06, + "loss": 0.1848, + "step": 37831 + }, + { + "epoch": 0.886841658022586, + "grad_norm": 1.9544317722320557, + "learning_rate": 6.254985392792057e-06, + "loss": 0.2013, + "step": 37832 + }, + { + "epoch": 0.8868650995973909, + "grad_norm": 2.2210710048675537, + "learning_rate": 6.252421343549519e-06, + "loss": 0.1276, + "step": 37833 + }, + { + "epoch": 0.886888541172196, + "grad_norm": 4.355260372161865, + "learning_rate": 6.249857802983983e-06, + "loss": 0.1912, + "step": 37834 + }, + { + "epoch": 0.8869119827470009, + "grad_norm": 3.177229166030884, + "learning_rate": 6.247294771109347e-06, + "loss": 0.2505, + "step": 37835 + }, + { + "epoch": 0.8869354243218059, + "grad_norm": 0.3749920129776001, + "learning_rate": 6.2447322479395335e-06, + "loss": 0.0425, + "step": 37836 + }, + { + "epoch": 0.8869588658966109, + "grad_norm": 1.9274239540100098, + "learning_rate": 6.24217023348842e-06, + "loss": 0.1163, + "step": 37837 + }, + { + "epoch": 0.8869823074714159, + "grad_norm": 0.9092330932617188, + "learning_rate": 6.2396087277699415e-06, + "loss": 0.1017, + "step": 37838 + }, + { + "epoch": 0.8870057490462209, + "grad_norm": 1.455329179763794, + "learning_rate": 6.2370477307979735e-06, + "loss": 0.08, + "step": 37839 + }, + { + "epoch": 0.8870291906210259, + "grad_norm": 0.5746575593948364, + "learning_rate": 6.234487242586428e-06, + "loss": 0.0429, + "step": 37840 + }, + { + "epoch": 0.887052632195831, + "grad_norm": 5.3537278175354, + "learning_rate": 6.2319272631491954e-06, + "loss": 0.2365, + "step": 37841 + }, + { + "epoch": 0.8870760737706359, + "grad_norm": 2.4109842777252197, + "learning_rate": 6.229367792500163e-06, + "loss": 0.1892, + "step": 37842 + }, + { + "epoch": 0.887099515345441, + "grad_norm": 1.3084220886230469, + "learning_rate": 6.2268088306532076e-06, + "loss": 0.1092, + "step": 37843 + }, + { + "epoch": 0.8871229569202459, + "grad_norm": 1.553346037864685, + "learning_rate": 6.224250377622221e-06, + "loss": 0.5729, + "step": 37844 + }, + { + "epoch": 0.887146398495051, + "grad_norm": 3.8306915760040283, + "learning_rate": 6.2216924334211e-06, + "loss": 0.3382, + "step": 37845 + }, + { + "epoch": 0.8871698400698559, + "grad_norm": 2.0369229316711426, + "learning_rate": 6.219134998063713e-06, + "loss": 0.1526, + "step": 37846 + }, + { + "epoch": 0.8871932816446609, + "grad_norm": 2.060209035873413, + "learning_rate": 6.216578071563939e-06, + "loss": 0.2738, + "step": 37847 + }, + { + "epoch": 0.8872167232194659, + "grad_norm": 2.6557741165161133, + "learning_rate": 6.2140216539356426e-06, + "loss": 0.3566, + "step": 37848 + }, + { + "epoch": 0.8872401647942709, + "grad_norm": 4.47089147567749, + "learning_rate": 6.211465745192702e-06, + "loss": 0.3938, + "step": 37849 + }, + { + "epoch": 0.8872636063690759, + "grad_norm": 0.5316997766494751, + "learning_rate": 6.208910345348973e-06, + "loss": 0.042, + "step": 37850 + }, + { + "epoch": 0.8872870479438809, + "grad_norm": 3.1067445278167725, + "learning_rate": 6.2063554544183334e-06, + "loss": 0.2971, + "step": 37851 + }, + { + "epoch": 0.8873104895186859, + "grad_norm": 3.8162741661071777, + "learning_rate": 6.203801072414661e-06, + "loss": 0.3547, + "step": 37852 + }, + { + "epoch": 0.8873339310934909, + "grad_norm": 2.2208762168884277, + "learning_rate": 6.2012471993518116e-06, + "loss": 0.161, + "step": 37853 + }, + { + "epoch": 0.8873573726682958, + "grad_norm": 4.378176689147949, + "learning_rate": 6.198693835243619e-06, + "loss": 1.9236, + "step": 37854 + }, + { + "epoch": 0.8873808142431009, + "grad_norm": 3.9728057384490967, + "learning_rate": 6.19614098010396e-06, + "loss": 1.9747, + "step": 37855 + }, + { + "epoch": 0.8874042558179058, + "grad_norm": 2.488337516784668, + "learning_rate": 6.193588633946668e-06, + "loss": 0.2464, + "step": 37856 + }, + { + "epoch": 0.8874276973927109, + "grad_norm": 2.6166136264801025, + "learning_rate": 6.191036796785587e-06, + "loss": 0.3079, + "step": 37857 + }, + { + "epoch": 0.8874511389675158, + "grad_norm": 1.793553352355957, + "learning_rate": 6.1884854686345974e-06, + "loss": 0.0948, + "step": 37858 + }, + { + "epoch": 0.8874745805423209, + "grad_norm": 2.228435754776001, + "learning_rate": 6.18593464950753e-06, + "loss": 0.1505, + "step": 37859 + }, + { + "epoch": 0.8874980221171258, + "grad_norm": 2.850569248199463, + "learning_rate": 6.183384339418208e-06, + "loss": 0.3086, + "step": 37860 + }, + { + "epoch": 0.8875214636919309, + "grad_norm": 2.9987099170684814, + "learning_rate": 6.180834538380486e-06, + "loss": 0.1569, + "step": 37861 + }, + { + "epoch": 0.8875449052667358, + "grad_norm": 1.6241673231124878, + "learning_rate": 6.178285246408199e-06, + "loss": 0.0664, + "step": 37862 + }, + { + "epoch": 0.8875683468415408, + "grad_norm": 5.470936298370361, + "learning_rate": 6.1757364635151475e-06, + "loss": 0.469, + "step": 37863 + }, + { + "epoch": 0.8875917884163458, + "grad_norm": 2.248655319213867, + "learning_rate": 6.1731881897151955e-06, + "loss": 0.1839, + "step": 37864 + }, + { + "epoch": 0.8876152299911508, + "grad_norm": 2.1555655002593994, + "learning_rate": 6.170640425022178e-06, + "loss": 0.3237, + "step": 37865 + }, + { + "epoch": 0.8876386715659558, + "grad_norm": 2.605468273162842, + "learning_rate": 6.168093169449895e-06, + "loss": 1.4345, + "step": 37866 + }, + { + "epoch": 0.8876621131407608, + "grad_norm": 2.1950817108154297, + "learning_rate": 6.16554642301218e-06, + "loss": 0.0853, + "step": 37867 + }, + { + "epoch": 0.8876855547155658, + "grad_norm": 3.5144522190093994, + "learning_rate": 6.163000185722856e-06, + "loss": 0.3954, + "step": 37868 + }, + { + "epoch": 0.8877089962903708, + "grad_norm": 3.04266357421875, + "learning_rate": 6.160454457595733e-06, + "loss": 0.1714, + "step": 37869 + }, + { + "epoch": 0.8877324378651757, + "grad_norm": 3.0256383419036865, + "learning_rate": 6.1579092386445995e-06, + "loss": 0.1977, + "step": 37870 + }, + { + "epoch": 0.8877558794399808, + "grad_norm": 2.5293350219726562, + "learning_rate": 6.155364528883301e-06, + "loss": 1.6914, + "step": 37871 + }, + { + "epoch": 0.8877793210147857, + "grad_norm": 2.8358492851257324, + "learning_rate": 6.1528203283256366e-06, + "loss": 0.2045, + "step": 37872 + }, + { + "epoch": 0.8878027625895908, + "grad_norm": 3.148106336593628, + "learning_rate": 6.150276636985419e-06, + "loss": 0.3259, + "step": 37873 + }, + { + "epoch": 0.8878262041643957, + "grad_norm": 3.4466195106506348, + "learning_rate": 6.147733454876425e-06, + "loss": 0.1644, + "step": 37874 + }, + { + "epoch": 0.8878496457392008, + "grad_norm": 1.0277196168899536, + "learning_rate": 6.1451907820124555e-06, + "loss": 0.0719, + "step": 37875 + }, + { + "epoch": 0.8878730873140057, + "grad_norm": 1.220471978187561, + "learning_rate": 6.142648618407332e-06, + "loss": 0.0538, + "step": 37876 + }, + { + "epoch": 0.8878965288888108, + "grad_norm": 2.3159396648406982, + "learning_rate": 6.140106964074832e-06, + "loss": 0.211, + "step": 37877 + }, + { + "epoch": 0.8879199704636157, + "grad_norm": 2.677086353302002, + "learning_rate": 6.137565819028768e-06, + "loss": 0.2078, + "step": 37878 + }, + { + "epoch": 0.8879434120384208, + "grad_norm": 6.037191390991211, + "learning_rate": 6.135025183282906e-06, + "loss": 0.5589, + "step": 37879 + }, + { + "epoch": 0.8879668536132257, + "grad_norm": 2.7608816623687744, + "learning_rate": 6.132485056851034e-06, + "loss": 0.3969, + "step": 37880 + }, + { + "epoch": 0.8879902951880307, + "grad_norm": 3.697197198867798, + "learning_rate": 6.129945439746931e-06, + "loss": 0.2533, + "step": 37881 + }, + { + "epoch": 0.8880137367628357, + "grad_norm": 3.3482248783111572, + "learning_rate": 6.127406331984375e-06, + "loss": 0.2897, + "step": 37882 + }, + { + "epoch": 0.8880371783376407, + "grad_norm": 0.9447148442268372, + "learning_rate": 6.124867733577167e-06, + "loss": 0.0432, + "step": 37883 + }, + { + "epoch": 0.8880606199124457, + "grad_norm": 1.4451706409454346, + "learning_rate": 6.122329644539071e-06, + "loss": 0.5239, + "step": 37884 + }, + { + "epoch": 0.8880840614872507, + "grad_norm": 2.629702091217041, + "learning_rate": 6.119792064883844e-06, + "loss": 0.2231, + "step": 37885 + }, + { + "epoch": 0.8881075030620557, + "grad_norm": 1.7213797569274902, + "learning_rate": 6.117254994625265e-06, + "loss": 0.1758, + "step": 37886 + }, + { + "epoch": 0.8881309446368607, + "grad_norm": 0.9165969491004944, + "learning_rate": 6.114718433777112e-06, + "loss": 0.0961, + "step": 37887 + }, + { + "epoch": 0.8881543862116656, + "grad_norm": 3.9503722190856934, + "learning_rate": 6.112182382353116e-06, + "loss": 0.5336, + "step": 37888 + }, + { + "epoch": 0.8881778277864707, + "grad_norm": 2.8655502796173096, + "learning_rate": 6.109646840367056e-06, + "loss": 0.1896, + "step": 37889 + }, + { + "epoch": 0.8882012693612756, + "grad_norm": 1.206079363822937, + "learning_rate": 6.107111807832699e-06, + "loss": 0.0816, + "step": 37890 + }, + { + "epoch": 0.8882247109360807, + "grad_norm": 0.6194387674331665, + "learning_rate": 6.104577284763813e-06, + "loss": 0.0692, + "step": 37891 + }, + { + "epoch": 0.8882481525108857, + "grad_norm": 1.3063551187515259, + "learning_rate": 6.102043271174118e-06, + "loss": 0.0767, + "step": 37892 + }, + { + "epoch": 0.8882715940856907, + "grad_norm": 1.951191782951355, + "learning_rate": 6.0995097670773716e-06, + "loss": 1.1342, + "step": 37893 + }, + { + "epoch": 0.8882950356604957, + "grad_norm": 1.9015088081359863, + "learning_rate": 6.09697677248734e-06, + "loss": 0.657, + "step": 37894 + }, + { + "epoch": 0.8883184772353007, + "grad_norm": 2.506800413131714, + "learning_rate": 6.094444287417722e-06, + "loss": 0.0997, + "step": 37895 + }, + { + "epoch": 0.8883419188101057, + "grad_norm": 1.4912201166152954, + "learning_rate": 6.09191231188232e-06, + "loss": 0.1544, + "step": 37896 + }, + { + "epoch": 0.8883653603849107, + "grad_norm": 5.412325382232666, + "learning_rate": 6.089380845894832e-06, + "loss": 0.3364, + "step": 37897 + }, + { + "epoch": 0.8883888019597157, + "grad_norm": 1.1572145223617554, + "learning_rate": 6.086849889469004e-06, + "loss": 0.0409, + "step": 37898 + }, + { + "epoch": 0.8884122435345206, + "grad_norm": 3.0549869537353516, + "learning_rate": 6.08431944261858e-06, + "loss": 0.2963, + "step": 37899 + }, + { + "epoch": 0.8884356851093257, + "grad_norm": 1.3152835369110107, + "learning_rate": 6.081789505357249e-06, + "loss": 0.1352, + "step": 37900 + }, + { + "epoch": 0.8884591266841306, + "grad_norm": 1.5059618949890137, + "learning_rate": 6.07926007769879e-06, + "loss": 0.731, + "step": 37901 + }, + { + "epoch": 0.8884825682589357, + "grad_norm": 3.7441940307617188, + "learning_rate": 6.076731159656901e-06, + "loss": 0.2167, + "step": 37902 + }, + { + "epoch": 0.8885060098337406, + "grad_norm": 3.6855125427246094, + "learning_rate": 6.074202751245317e-06, + "loss": 0.4867, + "step": 37903 + }, + { + "epoch": 0.8885294514085457, + "grad_norm": 2.14017391204834, + "learning_rate": 6.071674852477749e-06, + "loss": 0.1202, + "step": 37904 + }, + { + "epoch": 0.8885528929833506, + "grad_norm": 0.6839768290519714, + "learning_rate": 6.069147463367908e-06, + "loss": 0.0654, + "step": 37905 + }, + { + "epoch": 0.8885763345581557, + "grad_norm": 3.2520620822906494, + "learning_rate": 6.0666205839295275e-06, + "loss": 0.2168, + "step": 37906 + }, + { + "epoch": 0.8885997761329606, + "grad_norm": 1.924399733543396, + "learning_rate": 6.064094214176275e-06, + "loss": 0.1396, + "step": 37907 + }, + { + "epoch": 0.8886232177077656, + "grad_norm": 1.004211187362671, + "learning_rate": 6.061568354121916e-06, + "loss": 0.1046, + "step": 37908 + }, + { + "epoch": 0.8886466592825706, + "grad_norm": 2.7845537662506104, + "learning_rate": 6.059043003780129e-06, + "loss": 0.1392, + "step": 37909 + }, + { + "epoch": 0.8886701008573756, + "grad_norm": 2.4372220039367676, + "learning_rate": 6.056518163164604e-06, + "loss": 0.1078, + "step": 37910 + }, + { + "epoch": 0.8886935424321806, + "grad_norm": 2.378810167312622, + "learning_rate": 6.053993832289062e-06, + "loss": 0.2203, + "step": 37911 + }, + { + "epoch": 0.8887169840069856, + "grad_norm": 2.341878890991211, + "learning_rate": 6.051470011167171e-06, + "loss": 0.0862, + "step": 37912 + }, + { + "epoch": 0.8887404255817906, + "grad_norm": 3.430299997329712, + "learning_rate": 6.048946699812663e-06, + "loss": 0.302, + "step": 37913 + }, + { + "epoch": 0.8887638671565956, + "grad_norm": 1.3448156118392944, + "learning_rate": 6.046423898239207e-06, + "loss": 0.1267, + "step": 37914 + }, + { + "epoch": 0.8887873087314005, + "grad_norm": 3.7206315994262695, + "learning_rate": 6.0439016064605004e-06, + "loss": 0.4262, + "step": 37915 + }, + { + "epoch": 0.8888107503062056, + "grad_norm": 2.969090700149536, + "learning_rate": 6.041379824490223e-06, + "loss": 0.1689, + "step": 37916 + }, + { + "epoch": 0.8888341918810105, + "grad_norm": 1.7940226793289185, + "learning_rate": 6.0388585523420636e-06, + "loss": 0.1659, + "step": 37917 + }, + { + "epoch": 0.8888576334558156, + "grad_norm": 2.254298210144043, + "learning_rate": 6.0363377900297006e-06, + "loss": 0.1087, + "step": 37918 + }, + { + "epoch": 0.8888810750306205, + "grad_norm": 2.5216310024261475, + "learning_rate": 6.033817537566811e-06, + "loss": 0.0866, + "step": 37919 + }, + { + "epoch": 0.8889045166054256, + "grad_norm": 1.3564765453338623, + "learning_rate": 6.03129779496705e-06, + "loss": 0.0664, + "step": 37920 + }, + { + "epoch": 0.8889279581802305, + "grad_norm": 3.49261474609375, + "learning_rate": 6.028778562244131e-06, + "loss": 0.2044, + "step": 37921 + }, + { + "epoch": 0.8889513997550356, + "grad_norm": 1.0040345191955566, + "learning_rate": 6.026259839411708e-06, + "loss": 0.072, + "step": 37922 + }, + { + "epoch": 0.8889748413298405, + "grad_norm": 2.2100610733032227, + "learning_rate": 6.023741626483426e-06, + "loss": 0.1198, + "step": 37923 + }, + { + "epoch": 0.8889982829046456, + "grad_norm": 2.4950618743896484, + "learning_rate": 6.0212239234729516e-06, + "loss": 0.78, + "step": 37924 + }, + { + "epoch": 0.8890217244794505, + "grad_norm": 0.5558326840400696, + "learning_rate": 6.018706730393986e-06, + "loss": 0.0547, + "step": 37925 + }, + { + "epoch": 0.8890451660542555, + "grad_norm": 0.20467139780521393, + "learning_rate": 6.016190047260151e-06, + "loss": 0.0189, + "step": 37926 + }, + { + "epoch": 0.8890686076290605, + "grad_norm": 1.4295111894607544, + "learning_rate": 6.013673874085102e-06, + "loss": 0.0542, + "step": 37927 + }, + { + "epoch": 0.8890920492038655, + "grad_norm": 4.167678356170654, + "learning_rate": 6.011158210882506e-06, + "loss": 0.29, + "step": 37928 + }, + { + "epoch": 0.8891154907786705, + "grad_norm": 3.0605506896972656, + "learning_rate": 6.008643057666019e-06, + "loss": 0.2324, + "step": 37929 + }, + { + "epoch": 0.8891389323534755, + "grad_norm": 3.96071457862854, + "learning_rate": 6.006128414449275e-06, + "loss": 0.2669, + "step": 37930 + }, + { + "epoch": 0.8891623739282805, + "grad_norm": 2.7754743099212646, + "learning_rate": 6.003614281245917e-06, + "loss": 0.1445, + "step": 37931 + }, + { + "epoch": 0.8891858155030855, + "grad_norm": 1.095199704170227, + "learning_rate": 6.001100658069592e-06, + "loss": 0.1441, + "step": 37932 + }, + { + "epoch": 0.8892092570778904, + "grad_norm": 0.9674882888793945, + "learning_rate": 5.99858754493392e-06, + "loss": 0.041, + "step": 37933 + }, + { + "epoch": 0.8892326986526955, + "grad_norm": 3.3999133110046387, + "learning_rate": 5.99607494185257e-06, + "loss": 0.3721, + "step": 37934 + }, + { + "epoch": 0.8892561402275004, + "grad_norm": 2.8544511795043945, + "learning_rate": 5.993562848839174e-06, + "loss": 0.2475, + "step": 37935 + }, + { + "epoch": 0.8892795818023055, + "grad_norm": 2.1961910724639893, + "learning_rate": 5.991051265907321e-06, + "loss": 1.4179, + "step": 37936 + }, + { + "epoch": 0.8893030233771104, + "grad_norm": 1.6501027345657349, + "learning_rate": 5.98854019307068e-06, + "loss": 0.1305, + "step": 37937 + }, + { + "epoch": 0.8893264649519155, + "grad_norm": 2.3952267169952393, + "learning_rate": 5.9860296303428705e-06, + "loss": 0.1805, + "step": 37938 + }, + { + "epoch": 0.8893499065267204, + "grad_norm": 0.3847734034061432, + "learning_rate": 5.983519577737506e-06, + "loss": 0.0407, + "step": 37939 + }, + { + "epoch": 0.8893733481015255, + "grad_norm": 1.1125566959381104, + "learning_rate": 5.981010035268198e-06, + "loss": 0.0696, + "step": 37940 + }, + { + "epoch": 0.8893967896763304, + "grad_norm": 1.2319340705871582, + "learning_rate": 5.9785010029485774e-06, + "loss": 0.0864, + "step": 37941 + }, + { + "epoch": 0.8894202312511355, + "grad_norm": 3.1365420818328857, + "learning_rate": 5.975992480792259e-06, + "loss": 0.2387, + "step": 37942 + }, + { + "epoch": 0.8894436728259405, + "grad_norm": 2.808347463607788, + "learning_rate": 5.973484468812851e-06, + "loss": 0.2069, + "step": 37943 + }, + { + "epoch": 0.8894671144007454, + "grad_norm": 1.901079773902893, + "learning_rate": 5.970976967023956e-06, + "loss": 0.1034, + "step": 37944 + }, + { + "epoch": 0.8894905559755505, + "grad_norm": 0.8116573095321655, + "learning_rate": 5.968469975439173e-06, + "loss": 0.0991, + "step": 37945 + }, + { + "epoch": 0.8895139975503554, + "grad_norm": 5.111205577850342, + "learning_rate": 5.965963494072135e-06, + "loss": 0.3078, + "step": 37946 + }, + { + "epoch": 0.8895374391251605, + "grad_norm": 1.5163402557373047, + "learning_rate": 5.963457522936411e-06, + "loss": 0.1001, + "step": 37947 + }, + { + "epoch": 0.8895608806999654, + "grad_norm": 3.486880302429199, + "learning_rate": 5.96095206204561e-06, + "loss": 0.2589, + "step": 37948 + }, + { + "epoch": 0.8895843222747705, + "grad_norm": 1.6767250299453735, + "learning_rate": 5.958447111413334e-06, + "loss": 0.0557, + "step": 37949 + }, + { + "epoch": 0.8896077638495754, + "grad_norm": 0.6979978680610657, + "learning_rate": 5.955942671053161e-06, + "loss": 0.0568, + "step": 37950 + }, + { + "epoch": 0.8896312054243805, + "grad_norm": 3.639822483062744, + "learning_rate": 5.953438740978701e-06, + "loss": 0.3802, + "step": 37951 + }, + { + "epoch": 0.8896546469991854, + "grad_norm": 2.4734768867492676, + "learning_rate": 5.9509353212034994e-06, + "loss": 0.1883, + "step": 37952 + }, + { + "epoch": 0.8896780885739904, + "grad_norm": 1.8952269554138184, + "learning_rate": 5.948432411741189e-06, + "loss": 1.2287, + "step": 37953 + }, + { + "epoch": 0.8897015301487954, + "grad_norm": 0.624580979347229, + "learning_rate": 5.945930012605327e-06, + "loss": 0.0269, + "step": 37954 + }, + { + "epoch": 0.8897249717236004, + "grad_norm": 2.611699342727661, + "learning_rate": 5.943428123809503e-06, + "loss": 0.2704, + "step": 37955 + }, + { + "epoch": 0.8897484132984054, + "grad_norm": 0.23795968294143677, + "learning_rate": 5.94092674536727e-06, + "loss": 0.0309, + "step": 37956 + }, + { + "epoch": 0.8897718548732104, + "grad_norm": 1.85348641872406, + "learning_rate": 5.93842587729222e-06, + "loss": 0.0893, + "step": 37957 + }, + { + "epoch": 0.8897952964480154, + "grad_norm": 2.172618865966797, + "learning_rate": 5.935925519597896e-06, + "loss": 0.1222, + "step": 37958 + }, + { + "epoch": 0.8898187380228204, + "grad_norm": 2.590139627456665, + "learning_rate": 5.933425672297899e-06, + "loss": 0.1739, + "step": 37959 + }, + { + "epoch": 0.8898421795976253, + "grad_norm": 0.3243628740310669, + "learning_rate": 5.9309263354057844e-06, + "loss": 0.0287, + "step": 37960 + }, + { + "epoch": 0.8898656211724304, + "grad_norm": 1.5276196002960205, + "learning_rate": 5.928427508935097e-06, + "loss": 1.417, + "step": 37961 + }, + { + "epoch": 0.8898890627472353, + "grad_norm": 4.088330268859863, + "learning_rate": 5.9259291928994155e-06, + "loss": 0.1883, + "step": 37962 + }, + { + "epoch": 0.8899125043220404, + "grad_norm": 0.9648884534835815, + "learning_rate": 5.923431387312284e-06, + "loss": 0.0599, + "step": 37963 + }, + { + "epoch": 0.8899359458968453, + "grad_norm": 2.0553414821624756, + "learning_rate": 5.920934092187269e-06, + "loss": 0.1265, + "step": 37964 + }, + { + "epoch": 0.8899593874716504, + "grad_norm": 2.293055295944214, + "learning_rate": 5.918437307537883e-06, + "loss": 0.0863, + "step": 37965 + }, + { + "epoch": 0.8899828290464553, + "grad_norm": 3.0720486640930176, + "learning_rate": 5.9159410333777255e-06, + "loss": 0.1425, + "step": 37966 + }, + { + "epoch": 0.8900062706212604, + "grad_norm": 3.051187038421631, + "learning_rate": 5.913445269720319e-06, + "loss": 0.24, + "step": 37967 + }, + { + "epoch": 0.8900297121960653, + "grad_norm": 1.6543687582015991, + "learning_rate": 5.9109500165791975e-06, + "loss": 0.0564, + "step": 37968 + }, + { + "epoch": 0.8900531537708704, + "grad_norm": 0.6814701557159424, + "learning_rate": 5.9084552739679165e-06, + "loss": 0.0447, + "step": 37969 + }, + { + "epoch": 0.8900765953456753, + "grad_norm": 1.8687011003494263, + "learning_rate": 5.905961041899988e-06, + "loss": 0.1482, + "step": 37970 + }, + { + "epoch": 0.8901000369204803, + "grad_norm": 2.8530874252319336, + "learning_rate": 5.903467320388955e-06, + "loss": 0.3134, + "step": 37971 + }, + { + "epoch": 0.8901234784952853, + "grad_norm": 2.2173566818237305, + "learning_rate": 5.900974109448376e-06, + "loss": 1.7716, + "step": 37972 + }, + { + "epoch": 0.8901469200700903, + "grad_norm": 3.685495615005493, + "learning_rate": 5.898481409091727e-06, + "loss": 0.2705, + "step": 37973 + }, + { + "epoch": 0.8901703616448953, + "grad_norm": 2.7444636821746826, + "learning_rate": 5.895989219332587e-06, + "loss": 0.2407, + "step": 37974 + }, + { + "epoch": 0.8901938032197003, + "grad_norm": 2.770169258117676, + "learning_rate": 5.893497540184456e-06, + "loss": 0.2021, + "step": 37975 + }, + { + "epoch": 0.8902172447945053, + "grad_norm": 3.4484007358551025, + "learning_rate": 5.891006371660856e-06, + "loss": 0.1509, + "step": 37976 + }, + { + "epoch": 0.8902406863693103, + "grad_norm": 0.44109296798706055, + "learning_rate": 5.888515713775289e-06, + "loss": 0.0395, + "step": 37977 + }, + { + "epoch": 0.8902641279441152, + "grad_norm": 1.6099821329116821, + "learning_rate": 5.886025566541276e-06, + "loss": 0.106, + "step": 37978 + }, + { + "epoch": 0.8902875695189203, + "grad_norm": 1.5328962802886963, + "learning_rate": 5.8835359299723505e-06, + "loss": 0.8847, + "step": 37979 + }, + { + "epoch": 0.8903110110937252, + "grad_norm": 0.6544080972671509, + "learning_rate": 5.881046804082002e-06, + "loss": 0.0399, + "step": 37980 + }, + { + "epoch": 0.8903344526685303, + "grad_norm": 1.2538846731185913, + "learning_rate": 5.878558188883743e-06, + "loss": 0.0912, + "step": 37981 + }, + { + "epoch": 0.8903578942433352, + "grad_norm": 3.109004020690918, + "learning_rate": 5.876070084391072e-06, + "loss": 0.1363, + "step": 37982 + }, + { + "epoch": 0.8903813358181403, + "grad_norm": 0.616441547870636, + "learning_rate": 5.87358249061748e-06, + "loss": 0.0363, + "step": 37983 + }, + { + "epoch": 0.8904047773929452, + "grad_norm": 1.9309887886047363, + "learning_rate": 5.871095407576488e-06, + "loss": 0.1435, + "step": 37984 + }, + { + "epoch": 0.8904282189677503, + "grad_norm": 2.3732240200042725, + "learning_rate": 5.868608835281564e-06, + "loss": 1.5862, + "step": 37985 + }, + { + "epoch": 0.8904516605425552, + "grad_norm": 0.8707643151283264, + "learning_rate": 5.86612277374623e-06, + "loss": 0.061, + "step": 37986 + }, + { + "epoch": 0.8904751021173603, + "grad_norm": 2.969153881072998, + "learning_rate": 5.863637222983953e-06, + "loss": 0.3475, + "step": 37987 + }, + { + "epoch": 0.8904985436921652, + "grad_norm": 1.8849358558654785, + "learning_rate": 5.861152183008245e-06, + "loss": 0.1034, + "step": 37988 + }, + { + "epoch": 0.8905219852669702, + "grad_norm": 0.7086827754974365, + "learning_rate": 5.858667653832561e-06, + "loss": 0.0541, + "step": 37989 + }, + { + "epoch": 0.8905454268417752, + "grad_norm": 3.337068557739258, + "learning_rate": 5.856183635470369e-06, + "loss": 0.2818, + "step": 37990 + }, + { + "epoch": 0.8905688684165802, + "grad_norm": 2.9722254276275635, + "learning_rate": 5.853700127935202e-06, + "loss": 0.3009, + "step": 37991 + }, + { + "epoch": 0.8905923099913852, + "grad_norm": 3.7160162925720215, + "learning_rate": 5.851217131240505e-06, + "loss": 0.2946, + "step": 37992 + }, + { + "epoch": 0.8906157515661902, + "grad_norm": 2.155200481414795, + "learning_rate": 5.848734645399745e-06, + "loss": 0.1354, + "step": 37993 + }, + { + "epoch": 0.8906391931409953, + "grad_norm": 2.57912015914917, + "learning_rate": 5.8462526704264e-06, + "loss": 1.6691, + "step": 37994 + }, + { + "epoch": 0.8906626347158002, + "grad_norm": 1.7983399629592896, + "learning_rate": 5.843771206333937e-06, + "loss": 0.1069, + "step": 37995 + }, + { + "epoch": 0.8906860762906053, + "grad_norm": 1.8514783382415771, + "learning_rate": 5.8412902531358e-06, + "loss": 0.0983, + "step": 37996 + }, + { + "epoch": 0.8907095178654102, + "grad_norm": 0.7136591076850891, + "learning_rate": 5.838809810845469e-06, + "loss": 0.0309, + "step": 37997 + }, + { + "epoch": 0.8907329594402152, + "grad_norm": 1.6763489246368408, + "learning_rate": 5.836329879476421e-06, + "loss": 0.0973, + "step": 37998 + }, + { + "epoch": 0.8907564010150202, + "grad_norm": 2.5498735904693604, + "learning_rate": 5.833850459042089e-06, + "loss": 0.2283, + "step": 37999 + }, + { + "epoch": 0.8907798425898252, + "grad_norm": 1.9847261905670166, + "learning_rate": 5.831371549555941e-06, + "loss": 0.0645, + "step": 38000 + }, + { + "epoch": 0.8908032841646302, + "grad_norm": 1.7948079109191895, + "learning_rate": 5.828893151031412e-06, + "loss": 0.0782, + "step": 38001 + }, + { + "epoch": 0.8908267257394352, + "grad_norm": 3.056973934173584, + "learning_rate": 5.826415263481966e-06, + "loss": 0.2368, + "step": 38002 + }, + { + "epoch": 0.8908501673142402, + "grad_norm": 4.962286949157715, + "learning_rate": 5.823937886921016e-06, + "loss": 0.2667, + "step": 38003 + }, + { + "epoch": 0.8908736088890452, + "grad_norm": 0.7791330814361572, + "learning_rate": 5.8214610213620406e-06, + "loss": 0.051, + "step": 38004 + }, + { + "epoch": 0.8908970504638501, + "grad_norm": 1.4953714609146118, + "learning_rate": 5.8189846668184724e-06, + "loss": 0.0611, + "step": 38005 + }, + { + "epoch": 0.8909204920386552, + "grad_norm": 2.7883684635162354, + "learning_rate": 5.816508823303735e-06, + "loss": 0.2679, + "step": 38006 + }, + { + "epoch": 0.8909439336134601, + "grad_norm": 2.0040760040283203, + "learning_rate": 5.814033490831272e-06, + "loss": 0.1187, + "step": 38007 + }, + { + "epoch": 0.8909673751882652, + "grad_norm": 1.556980013847351, + "learning_rate": 5.811558669414519e-06, + "loss": 0.129, + "step": 38008 + }, + { + "epoch": 0.8909908167630701, + "grad_norm": 3.214520215988159, + "learning_rate": 5.809084359066874e-06, + "loss": 0.2699, + "step": 38009 + }, + { + "epoch": 0.8910142583378752, + "grad_norm": 3.258002758026123, + "learning_rate": 5.806610559801784e-06, + "loss": 0.2294, + "step": 38010 + }, + { + "epoch": 0.8910376999126801, + "grad_norm": 1.5266268253326416, + "learning_rate": 5.804137271632692e-06, + "loss": 0.0599, + "step": 38011 + }, + { + "epoch": 0.8910611414874852, + "grad_norm": 2.7535009384155273, + "learning_rate": 5.801664494573e-06, + "loss": 0.2727, + "step": 38012 + }, + { + "epoch": 0.8910845830622901, + "grad_norm": 1.1246862411499023, + "learning_rate": 5.799192228636119e-06, + "loss": 0.0619, + "step": 38013 + }, + { + "epoch": 0.8911080246370952, + "grad_norm": 0.594123363494873, + "learning_rate": 5.796720473835471e-06, + "loss": 0.043, + "step": 38014 + }, + { + "epoch": 0.8911314662119001, + "grad_norm": 2.881334066390991, + "learning_rate": 5.794249230184467e-06, + "loss": 1.1936, + "step": 38015 + }, + { + "epoch": 0.8911549077867051, + "grad_norm": 1.3968347311019897, + "learning_rate": 5.79177849769651e-06, + "loss": 0.0377, + "step": 38016 + }, + { + "epoch": 0.8911783493615101, + "grad_norm": 4.318878650665283, + "learning_rate": 5.789308276385019e-06, + "loss": 0.3098, + "step": 38017 + }, + { + "epoch": 0.8912017909363151, + "grad_norm": 2.2128818035125732, + "learning_rate": 5.786838566263386e-06, + "loss": 0.087, + "step": 38018 + }, + { + "epoch": 0.8912252325111201, + "grad_norm": 2.800618886947632, + "learning_rate": 5.7843693673450215e-06, + "loss": 1.6196, + "step": 38019 + }, + { + "epoch": 0.8912486740859251, + "grad_norm": 1.4217498302459717, + "learning_rate": 5.781900679643326e-06, + "loss": 0.9547, + "step": 38020 + }, + { + "epoch": 0.8912721156607301, + "grad_norm": 1.643539547920227, + "learning_rate": 5.779432503171655e-06, + "loss": 0.1618, + "step": 38021 + }, + { + "epoch": 0.8912955572355351, + "grad_norm": 3.910804271697998, + "learning_rate": 5.776964837943466e-06, + "loss": 0.3118, + "step": 38022 + }, + { + "epoch": 0.89131899881034, + "grad_norm": 1.0518691539764404, + "learning_rate": 5.77449768397208e-06, + "loss": 0.0823, + "step": 38023 + }, + { + "epoch": 0.8913424403851451, + "grad_norm": 2.7739417552948, + "learning_rate": 5.772031041270953e-06, + "loss": 0.095, + "step": 38024 + }, + { + "epoch": 0.89136588195995, + "grad_norm": 2.2532100677490234, + "learning_rate": 5.76956490985342e-06, + "loss": 0.1354, + "step": 38025 + }, + { + "epoch": 0.8913893235347551, + "grad_norm": 0.669614315032959, + "learning_rate": 5.7670992897328805e-06, + "loss": 0.0666, + "step": 38026 + }, + { + "epoch": 0.89141276510956, + "grad_norm": 4.1493425369262695, + "learning_rate": 5.7646341809227125e-06, + "loss": 0.241, + "step": 38027 + }, + { + "epoch": 0.8914362066843651, + "grad_norm": 1.9242528676986694, + "learning_rate": 5.762169583436283e-06, + "loss": 0.1417, + "step": 38028 + }, + { + "epoch": 0.89145964825917, + "grad_norm": 2.844621419906616, + "learning_rate": 5.759705497286971e-06, + "loss": 0.2438, + "step": 38029 + }, + { + "epoch": 0.8914830898339751, + "grad_norm": 1.8040335178375244, + "learning_rate": 5.757241922488154e-06, + "loss": 1.0387, + "step": 38030 + }, + { + "epoch": 0.89150653140878, + "grad_norm": 3.1870410442352295, + "learning_rate": 5.7547788590531894e-06, + "loss": 0.1806, + "step": 38031 + }, + { + "epoch": 0.891529972983585, + "grad_norm": 2.5170013904571533, + "learning_rate": 5.752316306995453e-06, + "loss": 0.1089, + "step": 38032 + }, + { + "epoch": 0.89155341455839, + "grad_norm": 2.4973931312561035, + "learning_rate": 5.749854266328292e-06, + "loss": 0.5179, + "step": 38033 + }, + { + "epoch": 0.891576856133195, + "grad_norm": 2.1367361545562744, + "learning_rate": 5.747392737065049e-06, + "loss": 0.1866, + "step": 38034 + }, + { + "epoch": 0.891600297708, + "grad_norm": 3.2945070266723633, + "learning_rate": 5.744931719219116e-06, + "loss": 0.2404, + "step": 38035 + }, + { + "epoch": 0.891623739282805, + "grad_norm": 2.1831271648406982, + "learning_rate": 5.742471212803846e-06, + "loss": 0.1782, + "step": 38036 + }, + { + "epoch": 0.89164718085761, + "grad_norm": 2.1164536476135254, + "learning_rate": 5.7400112178325745e-06, + "loss": 0.203, + "step": 38037 + }, + { + "epoch": 0.891670622432415, + "grad_norm": 0.5398465394973755, + "learning_rate": 5.7375517343186466e-06, + "loss": 0.058, + "step": 38038 + }, + { + "epoch": 0.89169406400722, + "grad_norm": 1.9325692653656006, + "learning_rate": 5.735092762275418e-06, + "loss": 0.1663, + "step": 38039 + }, + { + "epoch": 0.891717505582025, + "grad_norm": 2.684476375579834, + "learning_rate": 5.732634301716233e-06, + "loss": 0.2192, + "step": 38040 + }, + { + "epoch": 0.8917409471568299, + "grad_norm": 3.0874664783477783, + "learning_rate": 5.730176352654404e-06, + "loss": 0.2755, + "step": 38041 + }, + { + "epoch": 0.891764388731635, + "grad_norm": 4.0223236083984375, + "learning_rate": 5.727718915103309e-06, + "loss": 0.2031, + "step": 38042 + }, + { + "epoch": 0.8917878303064399, + "grad_norm": 0.761376678943634, + "learning_rate": 5.7252619890762585e-06, + "loss": 0.0672, + "step": 38043 + }, + { + "epoch": 0.891811271881245, + "grad_norm": 1.040822982788086, + "learning_rate": 5.722805574586587e-06, + "loss": 0.6917, + "step": 38044 + }, + { + "epoch": 0.89183471345605, + "grad_norm": 0.8986183404922485, + "learning_rate": 5.720349671647618e-06, + "loss": 0.0433, + "step": 38045 + }, + { + "epoch": 0.891858155030855, + "grad_norm": 0.4147665500640869, + "learning_rate": 5.717894280272673e-06, + "loss": 0.0386, + "step": 38046 + }, + { + "epoch": 0.89188159660566, + "grad_norm": 2.6142003536224365, + "learning_rate": 5.7154394004750975e-06, + "loss": 0.1678, + "step": 38047 + }, + { + "epoch": 0.891905038180465, + "grad_norm": 0.9365708231925964, + "learning_rate": 5.712985032268192e-06, + "loss": 0.0473, + "step": 38048 + }, + { + "epoch": 0.89192847975527, + "grad_norm": 1.8477486371994019, + "learning_rate": 5.710531175665279e-06, + "loss": 0.2244, + "step": 38049 + }, + { + "epoch": 0.891951921330075, + "grad_norm": 3.857059955596924, + "learning_rate": 5.708077830679692e-06, + "loss": 0.2612, + "step": 38050 + }, + { + "epoch": 0.89197536290488, + "grad_norm": 1.784461259841919, + "learning_rate": 5.705624997324721e-06, + "loss": 0.0749, + "step": 38051 + }, + { + "epoch": 0.8919988044796849, + "grad_norm": 2.874882221221924, + "learning_rate": 5.703172675613677e-06, + "loss": 0.1481, + "step": 38052 + }, + { + "epoch": 0.89202224605449, + "grad_norm": 0.7364901900291443, + "learning_rate": 5.700720865559872e-06, + "loss": 0.0141, + "step": 38053 + }, + { + "epoch": 0.8920456876292949, + "grad_norm": 2.5259053707122803, + "learning_rate": 5.698269567176584e-06, + "loss": 0.2209, + "step": 38054 + }, + { + "epoch": 0.8920691292041, + "grad_norm": 1.9267488718032837, + "learning_rate": 5.695818780477158e-06, + "loss": 0.1496, + "step": 38055 + }, + { + "epoch": 0.8920925707789049, + "grad_norm": 2.738565444946289, + "learning_rate": 5.693368505474872e-06, + "loss": 0.2132, + "step": 38056 + }, + { + "epoch": 0.89211601235371, + "grad_norm": 2.22029447555542, + "learning_rate": 5.6909187421830265e-06, + "loss": 0.0964, + "step": 38057 + }, + { + "epoch": 0.8921394539285149, + "grad_norm": 3.0404810905456543, + "learning_rate": 5.688469490614878e-06, + "loss": 0.2552, + "step": 38058 + }, + { + "epoch": 0.89216289550332, + "grad_norm": 0.9089040160179138, + "learning_rate": 5.686020750783772e-06, + "loss": 0.038, + "step": 38059 + }, + { + "epoch": 0.8921863370781249, + "grad_norm": 2.3030407428741455, + "learning_rate": 5.683572522702962e-06, + "loss": 0.1511, + "step": 38060 + }, + { + "epoch": 0.8922097786529299, + "grad_norm": 2.9581100940704346, + "learning_rate": 5.68112480638573e-06, + "loss": 0.2042, + "step": 38061 + }, + { + "epoch": 0.8922332202277349, + "grad_norm": 3.012892246246338, + "learning_rate": 5.678677601845384e-06, + "loss": 0.2405, + "step": 38062 + }, + { + "epoch": 0.8922566618025399, + "grad_norm": 2.2197558879852295, + "learning_rate": 5.676230909095181e-06, + "loss": 0.1149, + "step": 38063 + }, + { + "epoch": 0.8922801033773449, + "grad_norm": 2.0073909759521484, + "learning_rate": 5.673784728148401e-06, + "loss": 0.1668, + "step": 38064 + }, + { + "epoch": 0.8923035449521499, + "grad_norm": 2.767799139022827, + "learning_rate": 5.671339059018321e-06, + "loss": 1.6257, + "step": 38065 + }, + { + "epoch": 0.8923269865269549, + "grad_norm": 2.402414083480835, + "learning_rate": 5.668893901718186e-06, + "loss": 0.1284, + "step": 38066 + }, + { + "epoch": 0.8923504281017599, + "grad_norm": 3.9853150844573975, + "learning_rate": 5.666449256261308e-06, + "loss": 0.0981, + "step": 38067 + }, + { + "epoch": 0.8923738696765648, + "grad_norm": 3.1622936725616455, + "learning_rate": 5.66400512266092e-06, + "loss": 0.2992, + "step": 38068 + }, + { + "epoch": 0.8923973112513699, + "grad_norm": 0.9930292367935181, + "learning_rate": 5.661561500930301e-06, + "loss": 0.0887, + "step": 38069 + }, + { + "epoch": 0.8924207528261748, + "grad_norm": 6.017672061920166, + "learning_rate": 5.659118391082685e-06, + "loss": 0.1757, + "step": 38070 + }, + { + "epoch": 0.8924441944009799, + "grad_norm": 0.88703453540802, + "learning_rate": 5.6566757931313496e-06, + "loss": 0.0671, + "step": 38071 + }, + { + "epoch": 0.8924676359757848, + "grad_norm": 0.9692531228065491, + "learning_rate": 5.654233707089562e-06, + "loss": 0.0919, + "step": 38072 + }, + { + "epoch": 0.8924910775505899, + "grad_norm": 1.6908204555511475, + "learning_rate": 5.651792132970524e-06, + "loss": 0.1384, + "step": 38073 + }, + { + "epoch": 0.8925145191253948, + "grad_norm": 2.2799887657165527, + "learning_rate": 5.649351070787534e-06, + "loss": 1.4447, + "step": 38074 + }, + { + "epoch": 0.8925379607001999, + "grad_norm": 0.820689857006073, + "learning_rate": 5.6469105205538275e-06, + "loss": 0.0475, + "step": 38075 + }, + { + "epoch": 0.8925614022750048, + "grad_norm": 3.821800470352173, + "learning_rate": 5.644470482282626e-06, + "loss": 0.2445, + "step": 38076 + }, + { + "epoch": 0.8925848438498099, + "grad_norm": 1.4820011854171753, + "learning_rate": 5.642030955987187e-06, + "loss": 0.0993, + "step": 38077 + }, + { + "epoch": 0.8926082854246148, + "grad_norm": 1.6876722574234009, + "learning_rate": 5.639591941680744e-06, + "loss": 0.0891, + "step": 38078 + }, + { + "epoch": 0.8926317269994198, + "grad_norm": 3.578248977661133, + "learning_rate": 5.637153439376508e-06, + "loss": 0.2216, + "step": 38079 + }, + { + "epoch": 0.8926551685742248, + "grad_norm": 0.7090355157852173, + "learning_rate": 5.634715449087746e-06, + "loss": 0.0759, + "step": 38080 + }, + { + "epoch": 0.8926786101490298, + "grad_norm": 3.0086829662323, + "learning_rate": 5.63227797082767e-06, + "loss": 0.1645, + "step": 38081 + }, + { + "epoch": 0.8927020517238348, + "grad_norm": 2.1729819774627686, + "learning_rate": 5.629841004609493e-06, + "loss": 0.1561, + "step": 38082 + }, + { + "epoch": 0.8927254932986398, + "grad_norm": 2.156287670135498, + "learning_rate": 5.627404550446469e-06, + "loss": 0.0739, + "step": 38083 + }, + { + "epoch": 0.8927489348734448, + "grad_norm": 2.301532030105591, + "learning_rate": 5.624968608351799e-06, + "loss": 1.2776, + "step": 38084 + }, + { + "epoch": 0.8927723764482498, + "grad_norm": 2.205709218978882, + "learning_rate": 5.622533178338707e-06, + "loss": 1.3068, + "step": 38085 + }, + { + "epoch": 0.8927958180230547, + "grad_norm": 2.1855385303497314, + "learning_rate": 5.620098260420381e-06, + "loss": 0.2163, + "step": 38086 + }, + { + "epoch": 0.8928192595978598, + "grad_norm": 2.0161619186401367, + "learning_rate": 5.617663854610078e-06, + "loss": 0.2059, + "step": 38087 + }, + { + "epoch": 0.8928427011726647, + "grad_norm": 7.154105186462402, + "learning_rate": 5.615229960920987e-06, + "loss": 0.6428, + "step": 38088 + }, + { + "epoch": 0.8928661427474698, + "grad_norm": 1.0827009677886963, + "learning_rate": 5.612796579366297e-06, + "loss": 0.0358, + "step": 38089 + }, + { + "epoch": 0.8928895843222747, + "grad_norm": 0.6878411173820496, + "learning_rate": 5.610363709959243e-06, + "loss": 0.0247, + "step": 38090 + }, + { + "epoch": 0.8929130258970798, + "grad_norm": 0.7803559899330139, + "learning_rate": 5.607931352713003e-06, + "loss": 0.1006, + "step": 38091 + }, + { + "epoch": 0.8929364674718847, + "grad_norm": 3.2012436389923096, + "learning_rate": 5.605499507640766e-06, + "loss": 0.26, + "step": 38092 + }, + { + "epoch": 0.8929599090466898, + "grad_norm": 2.282226085662842, + "learning_rate": 5.603068174755755e-06, + "loss": 1.5965, + "step": 38093 + }, + { + "epoch": 0.8929833506214947, + "grad_norm": 0.9284749627113342, + "learning_rate": 5.600637354071147e-06, + "loss": 0.075, + "step": 38094 + }, + { + "epoch": 0.8930067921962997, + "grad_norm": 2.9068350791931152, + "learning_rate": 5.598207045600135e-06, + "loss": 0.2991, + "step": 38095 + }, + { + "epoch": 0.8930302337711048, + "grad_norm": 0.4996291995048523, + "learning_rate": 5.5957772493559154e-06, + "loss": 0.0336, + "step": 38096 + }, + { + "epoch": 0.8930536753459097, + "grad_norm": 1.2676266431808472, + "learning_rate": 5.593347965351658e-06, + "loss": 1.0302, + "step": 38097 + }, + { + "epoch": 0.8930771169207148, + "grad_norm": 2.942014217376709, + "learning_rate": 5.590919193600552e-06, + "loss": 0.2192, + "step": 38098 + }, + { + "epoch": 0.8931005584955197, + "grad_norm": 2.0146119594573975, + "learning_rate": 5.588490934115764e-06, + "loss": 0.1074, + "step": 38099 + }, + { + "epoch": 0.8931240000703248, + "grad_norm": 0.6519384384155273, + "learning_rate": 5.586063186910484e-06, + "loss": 0.0664, + "step": 38100 + }, + { + "epoch": 0.8931474416451297, + "grad_norm": 2.077822208404541, + "learning_rate": 5.583635951997878e-06, + "loss": 0.2212, + "step": 38101 + }, + { + "epoch": 0.8931708832199348, + "grad_norm": 2.738947629928589, + "learning_rate": 5.581209229391126e-06, + "loss": 0.1587, + "step": 38102 + }, + { + "epoch": 0.8931943247947397, + "grad_norm": 2.357089042663574, + "learning_rate": 5.578783019103384e-06, + "loss": 0.1411, + "step": 38103 + }, + { + "epoch": 0.8932177663695448, + "grad_norm": 2.6754462718963623, + "learning_rate": 5.5763573211478074e-06, + "loss": 0.1954, + "step": 38104 + }, + { + "epoch": 0.8932412079443497, + "grad_norm": 3.5362696647644043, + "learning_rate": 5.573932135537585e-06, + "loss": 1.2447, + "step": 38105 + }, + { + "epoch": 0.8932646495191547, + "grad_norm": 2.54231858253479, + "learning_rate": 5.571507462285852e-06, + "loss": 0.1664, + "step": 38106 + }, + { + "epoch": 0.8932880910939597, + "grad_norm": 2.162896156311035, + "learning_rate": 5.5690833014057864e-06, + "loss": 0.102, + "step": 38107 + }, + { + "epoch": 0.8933115326687647, + "grad_norm": 3.331355094909668, + "learning_rate": 5.566659652910522e-06, + "loss": 0.1851, + "step": 38108 + }, + { + "epoch": 0.8933349742435697, + "grad_norm": 4.045475959777832, + "learning_rate": 5.564236516813226e-06, + "loss": 0.3802, + "step": 38109 + }, + { + "epoch": 0.8933584158183747, + "grad_norm": 4.1854705810546875, + "learning_rate": 5.561813893127032e-06, + "loss": 0.5892, + "step": 38110 + }, + { + "epoch": 0.8933818573931797, + "grad_norm": 1.0565099716186523, + "learning_rate": 5.559391781865075e-06, + "loss": 0.1095, + "step": 38111 + }, + { + "epoch": 0.8934052989679847, + "grad_norm": 1.7682197093963623, + "learning_rate": 5.556970183040533e-06, + "loss": 0.0839, + "step": 38112 + }, + { + "epoch": 0.8934287405427896, + "grad_norm": 2.253112316131592, + "learning_rate": 5.554549096666528e-06, + "loss": 0.1979, + "step": 38113 + }, + { + "epoch": 0.8934521821175947, + "grad_norm": 2.9425158500671387, + "learning_rate": 5.552128522756195e-06, + "loss": 0.3089, + "step": 38114 + }, + { + "epoch": 0.8934756236923996, + "grad_norm": 3.7145254611968994, + "learning_rate": 5.549708461322667e-06, + "loss": 0.3384, + "step": 38115 + }, + { + "epoch": 0.8934990652672047, + "grad_norm": 0.3903592824935913, + "learning_rate": 5.547288912379078e-06, + "loss": 0.0098, + "step": 38116 + }, + { + "epoch": 0.8935225068420096, + "grad_norm": 1.3928769826889038, + "learning_rate": 5.544869875938541e-06, + "loss": 0.1831, + "step": 38117 + }, + { + "epoch": 0.8935459484168147, + "grad_norm": 1.3667162656784058, + "learning_rate": 5.542451352014199e-06, + "loss": 0.5294, + "step": 38118 + }, + { + "epoch": 0.8935693899916196, + "grad_norm": 0.931392252445221, + "learning_rate": 5.540033340619177e-06, + "loss": 0.1083, + "step": 38119 + }, + { + "epoch": 0.8935928315664247, + "grad_norm": 3.1177287101745605, + "learning_rate": 5.5376158417665855e-06, + "loss": 0.1998, + "step": 38120 + }, + { + "epoch": 0.8936162731412296, + "grad_norm": 4.247686862945557, + "learning_rate": 5.535198855469559e-06, + "loss": 0.1612, + "step": 38121 + }, + { + "epoch": 0.8936397147160346, + "grad_norm": 0.741349458694458, + "learning_rate": 5.532782381741197e-06, + "loss": 0.0407, + "step": 38122 + }, + { + "epoch": 0.8936631562908396, + "grad_norm": 2.2756800651550293, + "learning_rate": 5.530366420594612e-06, + "loss": 0.146, + "step": 38123 + }, + { + "epoch": 0.8936865978656446, + "grad_norm": 4.2863922119140625, + "learning_rate": 5.5279509720429055e-06, + "loss": 0.4813, + "step": 38124 + }, + { + "epoch": 0.8937100394404496, + "grad_norm": 1.5129148960113525, + "learning_rate": 5.525536036099199e-06, + "loss": 0.078, + "step": 38125 + }, + { + "epoch": 0.8937334810152546, + "grad_norm": 3.294801712036133, + "learning_rate": 5.523121612776594e-06, + "loss": 0.1986, + "step": 38126 + }, + { + "epoch": 0.8937569225900596, + "grad_norm": 1.7649340629577637, + "learning_rate": 5.52070770208819e-06, + "loss": 0.0945, + "step": 38127 + }, + { + "epoch": 0.8937803641648646, + "grad_norm": 1.9128615856170654, + "learning_rate": 5.518294304047078e-06, + "loss": 0.1019, + "step": 38128 + }, + { + "epoch": 0.8938038057396696, + "grad_norm": 5.4767608642578125, + "learning_rate": 5.515881418666369e-06, + "loss": 0.2957, + "step": 38129 + }, + { + "epoch": 0.8938272473144746, + "grad_norm": 0.9236876368522644, + "learning_rate": 5.513469045959118e-06, + "loss": 0.0502, + "step": 38130 + }, + { + "epoch": 0.8938506888892795, + "grad_norm": 1.0501121282577515, + "learning_rate": 5.511057185938439e-06, + "loss": 0.0621, + "step": 38131 + }, + { + "epoch": 0.8938741304640846, + "grad_norm": 3.6600022315979004, + "learning_rate": 5.5086458386174435e-06, + "loss": 0.3142, + "step": 38132 + }, + { + "epoch": 0.8938975720388895, + "grad_norm": 0.7276793718338013, + "learning_rate": 5.506235004009186e-06, + "loss": 0.0537, + "step": 38133 + }, + { + "epoch": 0.8939210136136946, + "grad_norm": 2.859818935394287, + "learning_rate": 5.503824682126757e-06, + "loss": 1.488, + "step": 38134 + }, + { + "epoch": 0.8939444551884995, + "grad_norm": 1.0763624906539917, + "learning_rate": 5.501414872983235e-06, + "loss": 0.0676, + "step": 38135 + }, + { + "epoch": 0.8939678967633046, + "grad_norm": 2.5982909202575684, + "learning_rate": 5.499005576591687e-06, + "loss": 0.2362, + "step": 38136 + }, + { + "epoch": 0.8939913383381095, + "grad_norm": 1.5762425661087036, + "learning_rate": 5.496596792965181e-06, + "loss": 0.05, + "step": 38137 + }, + { + "epoch": 0.8940147799129146, + "grad_norm": 3.048828125, + "learning_rate": 5.494188522116816e-06, + "loss": 0.1633, + "step": 38138 + }, + { + "epoch": 0.8940382214877195, + "grad_norm": 4.403196334838867, + "learning_rate": 5.491780764059639e-06, + "loss": 0.3723, + "step": 38139 + }, + { + "epoch": 0.8940616630625245, + "grad_norm": 2.567718982696533, + "learning_rate": 5.489373518806718e-06, + "loss": 0.3001, + "step": 38140 + }, + { + "epoch": 0.8940851046373295, + "grad_norm": 0.3216460645198822, + "learning_rate": 5.486966786371106e-06, + "loss": 0.017, + "step": 38141 + }, + { + "epoch": 0.8941085462121345, + "grad_norm": 0.48927217721939087, + "learning_rate": 5.484560566765862e-06, + "loss": 0.0595, + "step": 38142 + }, + { + "epoch": 0.8941319877869395, + "grad_norm": 3.239043951034546, + "learning_rate": 5.482154860004052e-06, + "loss": 0.3401, + "step": 38143 + }, + { + "epoch": 0.8941554293617445, + "grad_norm": 5.3171281814575195, + "learning_rate": 5.4797496660987325e-06, + "loss": 0.215, + "step": 38144 + }, + { + "epoch": 0.8941788709365495, + "grad_norm": 1.4191818237304688, + "learning_rate": 5.47734498506296e-06, + "loss": 0.1048, + "step": 38145 + }, + { + "epoch": 0.8942023125113545, + "grad_norm": 0.8515853881835938, + "learning_rate": 5.474940816909768e-06, + "loss": 0.0724, + "step": 38146 + }, + { + "epoch": 0.8942257540861596, + "grad_norm": 2.604499101638794, + "learning_rate": 5.472537161652203e-06, + "loss": 1.5109, + "step": 38147 + }, + { + "epoch": 0.8942491956609645, + "grad_norm": 2.53045916557312, + "learning_rate": 5.470134019303308e-06, + "loss": 0.2646, + "step": 38148 + }, + { + "epoch": 0.8942726372357696, + "grad_norm": 3.120543956756592, + "learning_rate": 5.467731389876107e-06, + "loss": 0.2282, + "step": 38149 + }, + { + "epoch": 0.8942960788105745, + "grad_norm": 0.8663738369941711, + "learning_rate": 5.4653292733836794e-06, + "loss": 0.0675, + "step": 38150 + }, + { + "epoch": 0.8943195203853795, + "grad_norm": 2.571887254714966, + "learning_rate": 5.462927669839024e-06, + "loss": 0.1554, + "step": 38151 + }, + { + "epoch": 0.8943429619601845, + "grad_norm": 3.503558874130249, + "learning_rate": 5.460526579255176e-06, + "loss": 0.2802, + "step": 38152 + }, + { + "epoch": 0.8943664035349895, + "grad_norm": 2.544923782348633, + "learning_rate": 5.458126001645181e-06, + "loss": 0.3221, + "step": 38153 + }, + { + "epoch": 0.8943898451097945, + "grad_norm": 2.6803269386291504, + "learning_rate": 5.45572593702205e-06, + "loss": 0.1007, + "step": 38154 + }, + { + "epoch": 0.8944132866845995, + "grad_norm": 0.6291919946670532, + "learning_rate": 5.4533263853987825e-06, + "loss": 0.0536, + "step": 38155 + }, + { + "epoch": 0.8944367282594045, + "grad_norm": 1.7889076471328735, + "learning_rate": 5.4509273467884366e-06, + "loss": 0.1247, + "step": 38156 + }, + { + "epoch": 0.8944601698342095, + "grad_norm": 0.33040672540664673, + "learning_rate": 5.448528821204036e-06, + "loss": 0.0429, + "step": 38157 + }, + { + "epoch": 0.8944836114090144, + "grad_norm": 3.2072396278381348, + "learning_rate": 5.4461308086585675e-06, + "loss": 1.8171, + "step": 38158 + }, + { + "epoch": 0.8945070529838195, + "grad_norm": 2.4644815921783447, + "learning_rate": 5.443733309165044e-06, + "loss": 0.2169, + "step": 38159 + }, + { + "epoch": 0.8945304945586244, + "grad_norm": 2.558715343475342, + "learning_rate": 5.441336322736501e-06, + "loss": 0.2089, + "step": 38160 + }, + { + "epoch": 0.8945539361334295, + "grad_norm": 2.8520798683166504, + "learning_rate": 5.4389398493859156e-06, + "loss": 0.1262, + "step": 38161 + }, + { + "epoch": 0.8945773777082344, + "grad_norm": 3.236013174057007, + "learning_rate": 5.436543889126289e-06, + "loss": 0.1997, + "step": 38162 + }, + { + "epoch": 0.8946008192830395, + "grad_norm": 2.7261691093444824, + "learning_rate": 5.434148441970643e-06, + "loss": 0.1423, + "step": 38163 + }, + { + "epoch": 0.8946242608578444, + "grad_norm": 2.3135595321655273, + "learning_rate": 5.4317535079319694e-06, + "loss": 0.1491, + "step": 38164 + }, + { + "epoch": 0.8946477024326495, + "grad_norm": 3.1745352745056152, + "learning_rate": 5.429359087023267e-06, + "loss": 0.306, + "step": 38165 + }, + { + "epoch": 0.8946711440074544, + "grad_norm": 2.1088788509368896, + "learning_rate": 5.426965179257515e-06, + "loss": 1.3081, + "step": 38166 + }, + { + "epoch": 0.8946945855822594, + "grad_norm": 2.2277283668518066, + "learning_rate": 5.424571784647692e-06, + "loss": 0.1603, + "step": 38167 + }, + { + "epoch": 0.8947180271570644, + "grad_norm": 3.2255001068115234, + "learning_rate": 5.422178903206821e-06, + "loss": 1.24, + "step": 38168 + }, + { + "epoch": 0.8947414687318694, + "grad_norm": 2.4219794273376465, + "learning_rate": 5.419786534947846e-06, + "loss": 0.1565, + "step": 38169 + }, + { + "epoch": 0.8947649103066744, + "grad_norm": 2.06547212600708, + "learning_rate": 5.4173946798837915e-06, + "loss": 0.1637, + "step": 38170 + }, + { + "epoch": 0.8947883518814794, + "grad_norm": 0.7322423458099365, + "learning_rate": 5.415003338027613e-06, + "loss": 0.0625, + "step": 38171 + }, + { + "epoch": 0.8948117934562844, + "grad_norm": 0.5375235080718994, + "learning_rate": 5.412612509392279e-06, + "loss": 0.0474, + "step": 38172 + }, + { + "epoch": 0.8948352350310894, + "grad_norm": 0.33145883679389954, + "learning_rate": 5.410222193990777e-06, + "loss": 0.0295, + "step": 38173 + }, + { + "epoch": 0.8948586766058944, + "grad_norm": 1.414617896080017, + "learning_rate": 5.407832391836065e-06, + "loss": 0.0833, + "step": 38174 + }, + { + "epoch": 0.8948821181806994, + "grad_norm": 2.6368215084075928, + "learning_rate": 5.405443102941099e-06, + "loss": 1.2496, + "step": 38175 + }, + { + "epoch": 0.8949055597555043, + "grad_norm": 3.9951608180999756, + "learning_rate": 5.403054327318868e-06, + "loss": 0.3813, + "step": 38176 + }, + { + "epoch": 0.8949290013303094, + "grad_norm": 1.8928272724151611, + "learning_rate": 5.400666064982329e-06, + "loss": 0.1048, + "step": 38177 + }, + { + "epoch": 0.8949524429051143, + "grad_norm": 0.5698238611221313, + "learning_rate": 5.398278315944438e-06, + "loss": 0.0561, + "step": 38178 + }, + { + "epoch": 0.8949758844799194, + "grad_norm": 7.340559959411621, + "learning_rate": 5.39589108021813e-06, + "loss": 0.555, + "step": 38179 + }, + { + "epoch": 0.8949993260547243, + "grad_norm": 2.610942840576172, + "learning_rate": 5.393504357816392e-06, + "loss": 0.1324, + "step": 38180 + }, + { + "epoch": 0.8950227676295294, + "grad_norm": 2.866913080215454, + "learning_rate": 5.39111814875215e-06, + "loss": 0.2153, + "step": 38181 + }, + { + "epoch": 0.8950462092043343, + "grad_norm": 2.4846866130828857, + "learning_rate": 5.38873245303837e-06, + "loss": 0.2329, + "step": 38182 + }, + { + "epoch": 0.8950696507791394, + "grad_norm": 1.097302794456482, + "learning_rate": 5.386347270687997e-06, + "loss": 0.0809, + "step": 38183 + }, + { + "epoch": 0.8950930923539443, + "grad_norm": 1.6497225761413574, + "learning_rate": 5.3839626017139545e-06, + "loss": 0.1025, + "step": 38184 + }, + { + "epoch": 0.8951165339287493, + "grad_norm": 3.4162724018096924, + "learning_rate": 5.381578446129198e-06, + "loss": 0.3808, + "step": 38185 + }, + { + "epoch": 0.8951399755035543, + "grad_norm": 3.037493944168091, + "learning_rate": 5.3791948039466525e-06, + "loss": 0.1706, + "step": 38186 + }, + { + "epoch": 0.8951634170783593, + "grad_norm": 3.038750648498535, + "learning_rate": 5.376811675179239e-06, + "loss": 0.0555, + "step": 38187 + }, + { + "epoch": 0.8951868586531643, + "grad_norm": 2.4281649589538574, + "learning_rate": 5.374429059839925e-06, + "loss": 0.1798, + "step": 38188 + }, + { + "epoch": 0.8952103002279693, + "grad_norm": 3.1569457054138184, + "learning_rate": 5.372046957941623e-06, + "loss": 0.1303, + "step": 38189 + }, + { + "epoch": 0.8952337418027743, + "grad_norm": 1.7961766719818115, + "learning_rate": 5.369665369497245e-06, + "loss": 0.0824, + "step": 38190 + }, + { + "epoch": 0.8952571833775793, + "grad_norm": 0.9429186582565308, + "learning_rate": 5.367284294519736e-06, + "loss": 0.0747, + "step": 38191 + }, + { + "epoch": 0.8952806249523843, + "grad_norm": 3.4360833168029785, + "learning_rate": 5.364903733021975e-06, + "loss": 0.2982, + "step": 38192 + }, + { + "epoch": 0.8953040665271893, + "grad_norm": 1.697879433631897, + "learning_rate": 5.362523685016929e-06, + "loss": 0.9785, + "step": 38193 + }, + { + "epoch": 0.8953275081019942, + "grad_norm": 1.0570805072784424, + "learning_rate": 5.360144150517476e-06, + "loss": 0.084, + "step": 38194 + }, + { + "epoch": 0.8953509496767993, + "grad_norm": 2.4143753051757812, + "learning_rate": 5.357765129536563e-06, + "loss": 0.1803, + "step": 38195 + }, + { + "epoch": 0.8953743912516042, + "grad_norm": 1.0679551362991333, + "learning_rate": 5.355386622087066e-06, + "loss": 0.1477, + "step": 38196 + }, + { + "epoch": 0.8953978328264093, + "grad_norm": 1.2911251783370972, + "learning_rate": 5.35300862818191e-06, + "loss": 0.0548, + "step": 38197 + }, + { + "epoch": 0.8954212744012143, + "grad_norm": 1.7477226257324219, + "learning_rate": 5.350631147833984e-06, + "loss": 1.2768, + "step": 38198 + }, + { + "epoch": 0.8954447159760193, + "grad_norm": 3.3799004554748535, + "learning_rate": 5.3482541810562004e-06, + "loss": 0.1722, + "step": 38199 + }, + { + "epoch": 0.8954681575508243, + "grad_norm": 0.44542405009269714, + "learning_rate": 5.345877727861437e-06, + "loss": 0.0254, + "step": 38200 + }, + { + "epoch": 0.8954915991256293, + "grad_norm": 0.2810465693473816, + "learning_rate": 5.343501788262606e-06, + "loss": 0.0255, + "step": 38201 + }, + { + "epoch": 0.8955150407004343, + "grad_norm": 4.620611667633057, + "learning_rate": 5.341126362272608e-06, + "loss": 0.4458, + "step": 38202 + }, + { + "epoch": 0.8955384822752392, + "grad_norm": 1.8469250202178955, + "learning_rate": 5.338751449904322e-06, + "loss": 1.5012, + "step": 38203 + }, + { + "epoch": 0.8955619238500443, + "grad_norm": 0.784155011177063, + "learning_rate": 5.3363770511706044e-06, + "loss": 0.0565, + "step": 38204 + }, + { + "epoch": 0.8955853654248492, + "grad_norm": 0.7168461680412292, + "learning_rate": 5.334003166084389e-06, + "loss": 0.0671, + "step": 38205 + }, + { + "epoch": 0.8956088069996543, + "grad_norm": 1.8817932605743408, + "learning_rate": 5.331629794658533e-06, + "loss": 0.1075, + "step": 38206 + }, + { + "epoch": 0.8956322485744592, + "grad_norm": 1.9666390419006348, + "learning_rate": 5.329256936905902e-06, + "loss": 0.172, + "step": 38207 + }, + { + "epoch": 0.8956556901492643, + "grad_norm": 0.3614542484283447, + "learning_rate": 5.326884592839398e-06, + "loss": 0.0162, + "step": 38208 + }, + { + "epoch": 0.8956791317240692, + "grad_norm": 0.8564059734344482, + "learning_rate": 5.32451276247189e-06, + "loss": 0.0823, + "step": 38209 + }, + { + "epoch": 0.8957025732988743, + "grad_norm": 0.8566598296165466, + "learning_rate": 5.32214144581622e-06, + "loss": 0.081, + "step": 38210 + }, + { + "epoch": 0.8957260148736792, + "grad_norm": 2.211033821105957, + "learning_rate": 5.31977064288528e-06, + "loss": 0.1216, + "step": 38211 + }, + { + "epoch": 0.8957494564484842, + "grad_norm": 3.4175360202789307, + "learning_rate": 5.317400353691915e-06, + "loss": 0.1909, + "step": 38212 + }, + { + "epoch": 0.8957728980232892, + "grad_norm": 2.8236401081085205, + "learning_rate": 5.315030578249014e-06, + "loss": 0.3438, + "step": 38213 + }, + { + "epoch": 0.8957963395980942, + "grad_norm": 1.1391266584396362, + "learning_rate": 5.312661316569423e-06, + "loss": 0.7231, + "step": 38214 + }, + { + "epoch": 0.8958197811728992, + "grad_norm": 3.1115810871124268, + "learning_rate": 5.310292568665986e-06, + "loss": 0.3101, + "step": 38215 + }, + { + "epoch": 0.8958432227477042, + "grad_norm": 3.9269604682922363, + "learning_rate": 5.307924334551551e-06, + "loss": 0.3284, + "step": 38216 + }, + { + "epoch": 0.8958666643225092, + "grad_norm": 3.7820322513580322, + "learning_rate": 5.305556614238994e-06, + "loss": 0.5268, + "step": 38217 + }, + { + "epoch": 0.8958901058973142, + "grad_norm": 1.9081188440322876, + "learning_rate": 5.303189407741149e-06, + "loss": 0.1307, + "step": 38218 + }, + { + "epoch": 0.8959135474721192, + "grad_norm": 6.283193111419678, + "learning_rate": 5.300822715070852e-06, + "loss": 0.6337, + "step": 38219 + }, + { + "epoch": 0.8959369890469242, + "grad_norm": 4.05999755859375, + "learning_rate": 5.298456536240959e-06, + "loss": 0.4005, + "step": 38220 + }, + { + "epoch": 0.8959604306217291, + "grad_norm": 3.658377170562744, + "learning_rate": 5.2960908712643145e-06, + "loss": 0.3101, + "step": 38221 + }, + { + "epoch": 0.8959838721965342, + "grad_norm": 1.7040088176727295, + "learning_rate": 5.2937257201537305e-06, + "loss": 0.1901, + "step": 38222 + }, + { + "epoch": 0.8960073137713391, + "grad_norm": 2.4815053939819336, + "learning_rate": 5.291361082922064e-06, + "loss": 1.4175, + "step": 38223 + }, + { + "epoch": 0.8960307553461442, + "grad_norm": 1.7669658660888672, + "learning_rate": 5.288996959582138e-06, + "loss": 0.6174, + "step": 38224 + }, + { + "epoch": 0.8960541969209491, + "grad_norm": 2.9991276264190674, + "learning_rate": 5.286633350146752e-06, + "loss": 0.2863, + "step": 38225 + }, + { + "epoch": 0.8960776384957542, + "grad_norm": 2.4051620960235596, + "learning_rate": 5.284270254628776e-06, + "loss": 1.7153, + "step": 38226 + }, + { + "epoch": 0.8961010800705591, + "grad_norm": 2.7150444984436035, + "learning_rate": 5.28190767304102e-06, + "loss": 0.164, + "step": 38227 + }, + { + "epoch": 0.8961245216453642, + "grad_norm": 1.7082862854003906, + "learning_rate": 5.279545605396274e-06, + "loss": 0.111, + "step": 38228 + }, + { + "epoch": 0.8961479632201691, + "grad_norm": 2.2483456134796143, + "learning_rate": 5.277184051707395e-06, + "loss": 0.1048, + "step": 38229 + }, + { + "epoch": 0.8961714047949741, + "grad_norm": 1.147660493850708, + "learning_rate": 5.274823011987173e-06, + "loss": 0.0994, + "step": 38230 + }, + { + "epoch": 0.8961948463697791, + "grad_norm": 3.111104726791382, + "learning_rate": 5.272462486248431e-06, + "loss": 0.2419, + "step": 38231 + }, + { + "epoch": 0.8962182879445841, + "grad_norm": 1.2976418733596802, + "learning_rate": 5.2701024745039575e-06, + "loss": 0.1159, + "step": 38232 + }, + { + "epoch": 0.8962417295193891, + "grad_norm": 1.2360388040542603, + "learning_rate": 5.267742976766576e-06, + "loss": 0.0686, + "step": 38233 + }, + { + "epoch": 0.8962651710941941, + "grad_norm": 1.9747276306152344, + "learning_rate": 5.265383993049089e-06, + "loss": 1.2836, + "step": 38234 + }, + { + "epoch": 0.8962886126689991, + "grad_norm": 2.8011882305145264, + "learning_rate": 5.263025523364296e-06, + "loss": 0.3193, + "step": 38235 + }, + { + "epoch": 0.8963120542438041, + "grad_norm": 0.7321584820747375, + "learning_rate": 5.260667567724986e-06, + "loss": 0.0448, + "step": 38236 + }, + { + "epoch": 0.896335495818609, + "grad_norm": 3.4183413982391357, + "learning_rate": 5.2583101261439615e-06, + "loss": 0.2853, + "step": 38237 + }, + { + "epoch": 0.8963589373934141, + "grad_norm": 0.3945792019367218, + "learning_rate": 5.255953198634001e-06, + "loss": 0.0281, + "step": 38238 + }, + { + "epoch": 0.896382378968219, + "grad_norm": 2.7475552558898926, + "learning_rate": 5.253596785207904e-06, + "loss": 0.1576, + "step": 38239 + }, + { + "epoch": 0.8964058205430241, + "grad_norm": 2.6943776607513428, + "learning_rate": 5.25124088587845e-06, + "loss": 0.0996, + "step": 38240 + }, + { + "epoch": 0.896429262117829, + "grad_norm": 0.646511435508728, + "learning_rate": 5.24888550065844e-06, + "loss": 0.0718, + "step": 38241 + }, + { + "epoch": 0.8964527036926341, + "grad_norm": 1.6766523122787476, + "learning_rate": 5.246530629560642e-06, + "loss": 0.0933, + "step": 38242 + }, + { + "epoch": 0.896476145267439, + "grad_norm": 2.4475855827331543, + "learning_rate": 5.244176272597822e-06, + "loss": 0.1902, + "step": 38243 + }, + { + "epoch": 0.8964995868422441, + "grad_norm": 0.9522435069084167, + "learning_rate": 5.241822429782772e-06, + "loss": 0.0554, + "step": 38244 + }, + { + "epoch": 0.896523028417049, + "grad_norm": 1.5765074491500854, + "learning_rate": 5.239469101128247e-06, + "loss": 0.1151, + "step": 38245 + }, + { + "epoch": 0.8965464699918541, + "grad_norm": 6.0198869705200195, + "learning_rate": 5.237116286647048e-06, + "loss": 0.2697, + "step": 38246 + }, + { + "epoch": 0.896569911566659, + "grad_norm": 2.069638967514038, + "learning_rate": 5.234763986351909e-06, + "loss": 0.0885, + "step": 38247 + }, + { + "epoch": 0.896593353141464, + "grad_norm": 2.709179401397705, + "learning_rate": 5.232412200255609e-06, + "loss": 0.1587, + "step": 38248 + }, + { + "epoch": 0.8966167947162691, + "grad_norm": 2.0405068397521973, + "learning_rate": 5.230060928370905e-06, + "loss": 1.3511, + "step": 38249 + }, + { + "epoch": 0.896640236291074, + "grad_norm": 3.996070384979248, + "learning_rate": 5.227710170710542e-06, + "loss": 0.3074, + "step": 38250 + }, + { + "epoch": 0.8966636778658791, + "grad_norm": 2.7455356121063232, + "learning_rate": 5.225359927287299e-06, + "loss": 0.2348, + "step": 38251 + }, + { + "epoch": 0.896687119440684, + "grad_norm": 2.8880488872528076, + "learning_rate": 5.223010198113909e-06, + "loss": 0.2084, + "step": 38252 + }, + { + "epoch": 0.8967105610154891, + "grad_norm": 4.41391658782959, + "learning_rate": 5.220660983203141e-06, + "loss": 0.3274, + "step": 38253 + }, + { + "epoch": 0.896734002590294, + "grad_norm": 1.8993146419525146, + "learning_rate": 5.218312282567739e-06, + "loss": 0.1002, + "step": 38254 + }, + { + "epoch": 0.8967574441650991, + "grad_norm": 1.9597394466400146, + "learning_rate": 5.215964096220427e-06, + "loss": 0.1239, + "step": 38255 + }, + { + "epoch": 0.896780885739904, + "grad_norm": 3.581376075744629, + "learning_rate": 5.213616424173962e-06, + "loss": 1.4394, + "step": 38256 + }, + { + "epoch": 0.896804327314709, + "grad_norm": 1.3924643993377686, + "learning_rate": 5.211269266441077e-06, + "loss": 0.023, + "step": 38257 + }, + { + "epoch": 0.896827768889514, + "grad_norm": 3.347834587097168, + "learning_rate": 5.208922623034507e-06, + "loss": 0.3357, + "step": 38258 + }, + { + "epoch": 0.896851210464319, + "grad_norm": 2.7499773502349854, + "learning_rate": 5.206576493966997e-06, + "loss": 0.1289, + "step": 38259 + }, + { + "epoch": 0.896874652039124, + "grad_norm": 0.961847186088562, + "learning_rate": 5.204230879251271e-06, + "loss": 0.0505, + "step": 38260 + }, + { + "epoch": 0.896898093613929, + "grad_norm": 2.256464958190918, + "learning_rate": 5.20188577890005e-06, + "loss": 0.1708, + "step": 38261 + }, + { + "epoch": 0.896921535188734, + "grad_norm": 3.3656938076019287, + "learning_rate": 5.1995411929260606e-06, + "loss": 0.1634, + "step": 38262 + }, + { + "epoch": 0.896944976763539, + "grad_norm": 2.6998140811920166, + "learning_rate": 5.1971971213420124e-06, + "loss": 0.2954, + "step": 38263 + }, + { + "epoch": 0.896968418338344, + "grad_norm": 0.7794342041015625, + "learning_rate": 5.194853564160662e-06, + "loss": 0.0686, + "step": 38264 + }, + { + "epoch": 0.896991859913149, + "grad_norm": 1.4901539087295532, + "learning_rate": 5.192510521394678e-06, + "loss": 0.082, + "step": 38265 + }, + { + "epoch": 0.8970153014879539, + "grad_norm": 3.794743776321411, + "learning_rate": 5.190167993056816e-06, + "loss": 0.4772, + "step": 38266 + }, + { + "epoch": 0.897038743062759, + "grad_norm": 2.0664594173431396, + "learning_rate": 5.187825979159766e-06, + "loss": 1.5586, + "step": 38267 + }, + { + "epoch": 0.8970621846375639, + "grad_norm": 1.7643282413482666, + "learning_rate": 5.185484479716241e-06, + "loss": 0.1039, + "step": 38268 + }, + { + "epoch": 0.897085626212369, + "grad_norm": 3.4264278411865234, + "learning_rate": 5.1831434947389405e-06, + "loss": 0.2758, + "step": 38269 + }, + { + "epoch": 0.8971090677871739, + "grad_norm": 4.517719268798828, + "learning_rate": 5.180803024240554e-06, + "loss": 0.3264, + "step": 38270 + }, + { + "epoch": 0.897132509361979, + "grad_norm": 3.720615863800049, + "learning_rate": 5.178463068233807e-06, + "loss": 0.1488, + "step": 38271 + }, + { + "epoch": 0.8971559509367839, + "grad_norm": 4.9303765296936035, + "learning_rate": 5.176123626731389e-06, + "loss": 0.4481, + "step": 38272 + }, + { + "epoch": 0.897179392511589, + "grad_norm": 0.6495131850242615, + "learning_rate": 5.173784699745998e-06, + "loss": 0.031, + "step": 38273 + }, + { + "epoch": 0.8972028340863939, + "grad_norm": 2.9685981273651123, + "learning_rate": 5.1714462872903045e-06, + "loss": 0.2852, + "step": 38274 + }, + { + "epoch": 0.897226275661199, + "grad_norm": 2.489340305328369, + "learning_rate": 5.1691083893770195e-06, + "loss": 0.2016, + "step": 38275 + }, + { + "epoch": 0.8972497172360039, + "grad_norm": 3.1207756996154785, + "learning_rate": 5.1667710060188e-06, + "loss": 0.2989, + "step": 38276 + }, + { + "epoch": 0.8972731588108089, + "grad_norm": 3.5398433208465576, + "learning_rate": 5.164434137228347e-06, + "loss": 0.2424, + "step": 38277 + }, + { + "epoch": 0.8972966003856139, + "grad_norm": 2.4085822105407715, + "learning_rate": 5.1620977830183605e-06, + "loss": 0.1727, + "step": 38278 + }, + { + "epoch": 0.8973200419604189, + "grad_norm": 3.365999698638916, + "learning_rate": 5.159761943401498e-06, + "loss": 0.3836, + "step": 38279 + }, + { + "epoch": 0.8973434835352239, + "grad_norm": 1.9918705224990845, + "learning_rate": 5.157426618390426e-06, + "loss": 0.1005, + "step": 38280 + }, + { + "epoch": 0.8973669251100289, + "grad_norm": 0.9336322546005249, + "learning_rate": 5.155091807997825e-06, + "loss": 0.0862, + "step": 38281 + }, + { + "epoch": 0.8973903666848339, + "grad_norm": 2.0537447929382324, + "learning_rate": 5.152757512236372e-06, + "loss": 0.1162, + "step": 38282 + }, + { + "epoch": 0.8974138082596389, + "grad_norm": 1.5829895734786987, + "learning_rate": 5.150423731118703e-06, + "loss": 0.1363, + "step": 38283 + }, + { + "epoch": 0.8974372498344438, + "grad_norm": 1.215894103050232, + "learning_rate": 5.148090464657518e-06, + "loss": 0.0889, + "step": 38284 + }, + { + "epoch": 0.8974606914092489, + "grad_norm": 3.9288790225982666, + "learning_rate": 5.1457577128654625e-06, + "loss": 0.4411, + "step": 38285 + }, + { + "epoch": 0.8974841329840538, + "grad_norm": 2.7347567081451416, + "learning_rate": 5.1434254757551815e-06, + "loss": 0.229, + "step": 38286 + }, + { + "epoch": 0.8975075745588589, + "grad_norm": 3.064972162246704, + "learning_rate": 5.141093753339355e-06, + "loss": 0.1329, + "step": 38287 + }, + { + "epoch": 0.8975310161336638, + "grad_norm": 2.862133502960205, + "learning_rate": 5.138762545630604e-06, + "loss": 0.2978, + "step": 38288 + }, + { + "epoch": 0.8975544577084689, + "grad_norm": 4.141231060028076, + "learning_rate": 5.136431852641599e-06, + "loss": 0.4731, + "step": 38289 + }, + { + "epoch": 0.8975778992832738, + "grad_norm": 2.7982306480407715, + "learning_rate": 5.134101674384973e-06, + "loss": 0.2626, + "step": 38290 + }, + { + "epoch": 0.8976013408580789, + "grad_norm": 2.9758718013763428, + "learning_rate": 5.131772010873403e-06, + "loss": 0.203, + "step": 38291 + }, + { + "epoch": 0.8976247824328838, + "grad_norm": 2.769756317138672, + "learning_rate": 5.129442862119494e-06, + "loss": 0.1119, + "step": 38292 + }, + { + "epoch": 0.8976482240076888, + "grad_norm": 1.4860374927520752, + "learning_rate": 5.127114228135899e-06, + "loss": 0.0568, + "step": 38293 + }, + { + "epoch": 0.8976716655824938, + "grad_norm": 3.0151569843292236, + "learning_rate": 5.124786108935254e-06, + "loss": 0.2328, + "step": 38294 + }, + { + "epoch": 0.8976951071572988, + "grad_norm": 0.34807276725769043, + "learning_rate": 5.122458504530159e-06, + "loss": 0.0255, + "step": 38295 + }, + { + "epoch": 0.8977185487321038, + "grad_norm": 3.600668430328369, + "learning_rate": 5.120131414933304e-06, + "loss": 0.3116, + "step": 38296 + }, + { + "epoch": 0.8977419903069088, + "grad_norm": 2.764587163925171, + "learning_rate": 5.11780484015727e-06, + "loss": 0.1079, + "step": 38297 + }, + { + "epoch": 0.8977654318817138, + "grad_norm": 1.0556743144989014, + "learning_rate": 5.1154787802147e-06, + "loss": 0.0531, + "step": 38298 + }, + { + "epoch": 0.8977888734565188, + "grad_norm": 3.0388524532318115, + "learning_rate": 5.113153235118217e-06, + "loss": 0.3052, + "step": 38299 + }, + { + "epoch": 0.8978123150313237, + "grad_norm": 2.4413902759552, + "learning_rate": 5.110828204880424e-06, + "loss": 0.128, + "step": 38300 + }, + { + "epoch": 0.8978357566061288, + "grad_norm": 2.367713689804077, + "learning_rate": 5.108503689513933e-06, + "loss": 0.1899, + "step": 38301 + }, + { + "epoch": 0.8978591981809338, + "grad_norm": 0.6585769653320312, + "learning_rate": 5.1061796890313655e-06, + "loss": 0.0442, + "step": 38302 + }, + { + "epoch": 0.8978826397557388, + "grad_norm": 3.583219289779663, + "learning_rate": 5.103856203445356e-06, + "loss": 1.6718, + "step": 38303 + }, + { + "epoch": 0.8979060813305438, + "grad_norm": 1.4673750400543213, + "learning_rate": 5.101533232768485e-06, + "loss": 0.0608, + "step": 38304 + }, + { + "epoch": 0.8979295229053488, + "grad_norm": 2.295344352722168, + "learning_rate": 5.0992107770133635e-06, + "loss": 1.8195, + "step": 38305 + }, + { + "epoch": 0.8979529644801538, + "grad_norm": 2.0066936016082764, + "learning_rate": 5.0968888361926034e-06, + "loss": 1.2567, + "step": 38306 + }, + { + "epoch": 0.8979764060549588, + "grad_norm": 2.5353760719299316, + "learning_rate": 5.094567410318785e-06, + "loss": 0.1486, + "step": 38307 + }, + { + "epoch": 0.8979998476297638, + "grad_norm": 2.5635032653808594, + "learning_rate": 5.092246499404496e-06, + "loss": 0.107, + "step": 38308 + }, + { + "epoch": 0.8980232892045688, + "grad_norm": 3.613208532333374, + "learning_rate": 5.089926103462361e-06, + "loss": 0.2482, + "step": 38309 + }, + { + "epoch": 0.8980467307793738, + "grad_norm": 0.7422364354133606, + "learning_rate": 5.087606222504959e-06, + "loss": 0.0209, + "step": 38310 + }, + { + "epoch": 0.8980701723541787, + "grad_norm": 2.7887680530548096, + "learning_rate": 5.085286856544868e-06, + "loss": 0.2926, + "step": 38311 + }, + { + "epoch": 0.8980936139289838, + "grad_norm": 3.0498299598693848, + "learning_rate": 5.082968005594679e-06, + "loss": 0.3846, + "step": 38312 + }, + { + "epoch": 0.8981170555037887, + "grad_norm": 2.7233693599700928, + "learning_rate": 5.0806496696669704e-06, + "loss": 0.1431, + "step": 38313 + }, + { + "epoch": 0.8981404970785938, + "grad_norm": 3.3998701572418213, + "learning_rate": 5.078331848774332e-06, + "loss": 0.2037, + "step": 38314 + }, + { + "epoch": 0.8981639386533987, + "grad_norm": 2.768770217895508, + "learning_rate": 5.0760145429293194e-06, + "loss": 0.1416, + "step": 38315 + }, + { + "epoch": 0.8981873802282038, + "grad_norm": 2.3083622455596924, + "learning_rate": 5.073697752144535e-06, + "loss": 0.1483, + "step": 38316 + }, + { + "epoch": 0.8982108218030087, + "grad_norm": 0.6262428164482117, + "learning_rate": 5.071381476432546e-06, + "loss": 0.0572, + "step": 38317 + }, + { + "epoch": 0.8982342633778138, + "grad_norm": 1.730238676071167, + "learning_rate": 5.069065715805898e-06, + "loss": 0.9407, + "step": 38318 + }, + { + "epoch": 0.8982577049526187, + "grad_norm": 3.634486675262451, + "learning_rate": 5.066750470277171e-06, + "loss": 0.2861, + "step": 38319 + }, + { + "epoch": 0.8982811465274237, + "grad_norm": 7.294936656951904, + "learning_rate": 5.064435739858931e-06, + "loss": 0.2487, + "step": 38320 + }, + { + "epoch": 0.8983045881022287, + "grad_norm": 0.6714000105857849, + "learning_rate": 5.062121524563712e-06, + "loss": 0.0272, + "step": 38321 + }, + { + "epoch": 0.8983280296770337, + "grad_norm": 3.2365646362304688, + "learning_rate": 5.059807824404106e-06, + "loss": 0.3052, + "step": 38322 + }, + { + "epoch": 0.8983514712518387, + "grad_norm": 3.8838999271392822, + "learning_rate": 5.057494639392657e-06, + "loss": 0.3224, + "step": 38323 + }, + { + "epoch": 0.8983749128266437, + "grad_norm": 3.1587724685668945, + "learning_rate": 5.055181969541911e-06, + "loss": 0.2021, + "step": 38324 + }, + { + "epoch": 0.8983983544014487, + "grad_norm": 3.151149034500122, + "learning_rate": 5.0528698148643915e-06, + "loss": 0.1391, + "step": 38325 + }, + { + "epoch": 0.8984217959762537, + "grad_norm": 2.085193634033203, + "learning_rate": 5.0505581753726996e-06, + "loss": 0.1164, + "step": 38326 + }, + { + "epoch": 0.8984452375510587, + "grad_norm": 1.1384141445159912, + "learning_rate": 5.0482470510793355e-06, + "loss": 0.0595, + "step": 38327 + }, + { + "epoch": 0.8984686791258637, + "grad_norm": 2.7141220569610596, + "learning_rate": 5.045936441996846e-06, + "loss": 0.1794, + "step": 38328 + }, + { + "epoch": 0.8984921207006686, + "grad_norm": 2.773021697998047, + "learning_rate": 5.043626348137787e-06, + "loss": 0.2258, + "step": 38329 + }, + { + "epoch": 0.8985155622754737, + "grad_norm": 2.654359817504883, + "learning_rate": 5.0413167695146924e-06, + "loss": 0.1261, + "step": 38330 + }, + { + "epoch": 0.8985390038502786, + "grad_norm": 1.0613752603530884, + "learning_rate": 5.039007706140064e-06, + "loss": 0.1248, + "step": 38331 + }, + { + "epoch": 0.8985624454250837, + "grad_norm": 0.44882723689079285, + "learning_rate": 5.036699158026459e-06, + "loss": 0.0332, + "step": 38332 + }, + { + "epoch": 0.8985858869998886, + "grad_norm": 2.1525185108184814, + "learning_rate": 5.034391125186388e-06, + "loss": 1.4295, + "step": 38333 + }, + { + "epoch": 0.8986093285746937, + "grad_norm": 0.4558921158313751, + "learning_rate": 5.032083607632387e-06, + "loss": 0.0434, + "step": 38334 + }, + { + "epoch": 0.8986327701494986, + "grad_norm": 0.8246422410011292, + "learning_rate": 5.029776605376968e-06, + "loss": 0.0838, + "step": 38335 + }, + { + "epoch": 0.8986562117243037, + "grad_norm": 3.5993919372558594, + "learning_rate": 5.027470118432653e-06, + "loss": 2.5213, + "step": 38336 + }, + { + "epoch": 0.8986796532991086, + "grad_norm": 1.4539899826049805, + "learning_rate": 5.0251641468119445e-06, + "loss": 0.0815, + "step": 38337 + }, + { + "epoch": 0.8987030948739136, + "grad_norm": 2.0306456089019775, + "learning_rate": 5.022858690527377e-06, + "loss": 0.1384, + "step": 38338 + }, + { + "epoch": 0.8987265364487186, + "grad_norm": 0.9128825664520264, + "learning_rate": 5.02055374959145e-06, + "loss": 0.0848, + "step": 38339 + }, + { + "epoch": 0.8987499780235236, + "grad_norm": 1.6858371496200562, + "learning_rate": 5.018249324016655e-06, + "loss": 0.082, + "step": 38340 + }, + { + "epoch": 0.8987734195983286, + "grad_norm": 0.9272050261497498, + "learning_rate": 5.015945413815526e-06, + "loss": 0.0427, + "step": 38341 + }, + { + "epoch": 0.8987968611731336, + "grad_norm": 0.8604831695556641, + "learning_rate": 5.013642019000553e-06, + "loss": 0.0708, + "step": 38342 + }, + { + "epoch": 0.8988203027479386, + "grad_norm": 2.6449685096740723, + "learning_rate": 5.011339139584214e-06, + "loss": 0.1899, + "step": 38343 + }, + { + "epoch": 0.8988437443227436, + "grad_norm": 2.2905590534210205, + "learning_rate": 5.009036775579035e-06, + "loss": 0.1578, + "step": 38344 + }, + { + "epoch": 0.8988671858975485, + "grad_norm": 3.7038445472717285, + "learning_rate": 5.006734926997481e-06, + "loss": 0.2156, + "step": 38345 + }, + { + "epoch": 0.8988906274723536, + "grad_norm": 4.303318977355957, + "learning_rate": 5.004433593852053e-06, + "loss": 0.3526, + "step": 38346 + }, + { + "epoch": 0.8989140690471585, + "grad_norm": 3.9395394325256348, + "learning_rate": 5.002132776155244e-06, + "loss": 0.3895, + "step": 38347 + }, + { + "epoch": 0.8989375106219636, + "grad_norm": 4.21840238571167, + "learning_rate": 4.99983247391953e-06, + "loss": 0.1924, + "step": 38348 + }, + { + "epoch": 0.8989609521967685, + "grad_norm": 1.8883060216903687, + "learning_rate": 4.997532687157402e-06, + "loss": 0.1115, + "step": 38349 + }, + { + "epoch": 0.8989843937715736, + "grad_norm": 0.30701592564582825, + "learning_rate": 4.995233415881317e-06, + "loss": 0.0112, + "step": 38350 + }, + { + "epoch": 0.8990078353463785, + "grad_norm": 2.6523752212524414, + "learning_rate": 4.992934660103788e-06, + "loss": 0.221, + "step": 38351 + }, + { + "epoch": 0.8990312769211836, + "grad_norm": 2.265394687652588, + "learning_rate": 4.9906364198372575e-06, + "loss": 0.1703, + "step": 38352 + }, + { + "epoch": 0.8990547184959886, + "grad_norm": 2.3718619346618652, + "learning_rate": 4.9883386950941964e-06, + "loss": 1.9095, + "step": 38353 + }, + { + "epoch": 0.8990781600707936, + "grad_norm": 2.7055983543395996, + "learning_rate": 4.986041485887094e-06, + "loss": 0.1805, + "step": 38354 + }, + { + "epoch": 0.8991016016455986, + "grad_norm": 2.374389410018921, + "learning_rate": 4.983744792228407e-06, + "loss": 0.1416, + "step": 38355 + }, + { + "epoch": 0.8991250432204035, + "grad_norm": 0.945324718952179, + "learning_rate": 4.981448614130579e-06, + "loss": 0.0332, + "step": 38356 + }, + { + "epoch": 0.8991484847952086, + "grad_norm": 0.23512274026870728, + "learning_rate": 4.979152951606092e-06, + "loss": 0.0177, + "step": 38357 + }, + { + "epoch": 0.8991719263700135, + "grad_norm": 1.1971819400787354, + "learning_rate": 4.9768578046673895e-06, + "loss": 0.0785, + "step": 38358 + }, + { + "epoch": 0.8991953679448186, + "grad_norm": 4.022367000579834, + "learning_rate": 4.974563173326918e-06, + "loss": 0.2938, + "step": 38359 + }, + { + "epoch": 0.8992188095196235, + "grad_norm": 2.2532880306243896, + "learning_rate": 4.972269057597156e-06, + "loss": 0.2729, + "step": 38360 + }, + { + "epoch": 0.8992422510944286, + "grad_norm": 2.3961150646209717, + "learning_rate": 4.969975457490528e-06, + "loss": 1.559, + "step": 38361 + }, + { + "epoch": 0.8992656926692335, + "grad_norm": 1.4453332424163818, + "learning_rate": 4.967682373019467e-06, + "loss": 0.0993, + "step": 38362 + }, + { + "epoch": 0.8992891342440386, + "grad_norm": 4.3051533699035645, + "learning_rate": 4.965389804196452e-06, + "loss": 0.1787, + "step": 38363 + }, + { + "epoch": 0.8993125758188435, + "grad_norm": 0.7494373321533203, + "learning_rate": 4.9630977510339075e-06, + "loss": 0.0916, + "step": 38364 + }, + { + "epoch": 0.8993360173936485, + "grad_norm": 5.378782272338867, + "learning_rate": 4.9608062135442665e-06, + "loss": 0.4782, + "step": 38365 + }, + { + "epoch": 0.8993594589684535, + "grad_norm": 2.4574012756347656, + "learning_rate": 4.958515191739943e-06, + "loss": 0.142, + "step": 38366 + }, + { + "epoch": 0.8993829005432585, + "grad_norm": 0.9870123863220215, + "learning_rate": 4.956224685633415e-06, + "loss": 0.0805, + "step": 38367 + }, + { + "epoch": 0.8994063421180635, + "grad_norm": 0.5856090784072876, + "learning_rate": 4.953934695237072e-06, + "loss": 0.024, + "step": 38368 + }, + { + "epoch": 0.8994297836928685, + "grad_norm": 0.2754824459552765, + "learning_rate": 4.95164522056335e-06, + "loss": 0.0136, + "step": 38369 + }, + { + "epoch": 0.8994532252676735, + "grad_norm": 5.064234733581543, + "learning_rate": 4.949356261624683e-06, + "loss": 0.4268, + "step": 38370 + }, + { + "epoch": 0.8994766668424785, + "grad_norm": 2.485060930252075, + "learning_rate": 4.947067818433459e-06, + "loss": 0.1628, + "step": 38371 + }, + { + "epoch": 0.8995001084172835, + "grad_norm": 0.9938589334487915, + "learning_rate": 4.944779891002138e-06, + "loss": 0.0606, + "step": 38372 + }, + { + "epoch": 0.8995235499920885, + "grad_norm": 2.8432443141937256, + "learning_rate": 4.9424924793431085e-06, + "loss": 0.1855, + "step": 38373 + }, + { + "epoch": 0.8995469915668934, + "grad_norm": 2.260899305343628, + "learning_rate": 4.940205583468782e-06, + "loss": 0.2312, + "step": 38374 + }, + { + "epoch": 0.8995704331416985, + "grad_norm": 0.9086979627609253, + "learning_rate": 4.937919203391583e-06, + "loss": 0.0714, + "step": 38375 + }, + { + "epoch": 0.8995938747165034, + "grad_norm": 1.0826960802078247, + "learning_rate": 4.935633339123902e-06, + "loss": 0.0849, + "step": 38376 + }, + { + "epoch": 0.8996173162913085, + "grad_norm": 0.9801568984985352, + "learning_rate": 4.93334799067815e-06, + "loss": 0.0573, + "step": 38377 + }, + { + "epoch": 0.8996407578661134, + "grad_norm": 2.186343193054199, + "learning_rate": 4.931063158066718e-06, + "loss": 0.1795, + "step": 38378 + }, + { + "epoch": 0.8996641994409185, + "grad_norm": 1.215401530265808, + "learning_rate": 4.928778841302017e-06, + "loss": 0.0889, + "step": 38379 + }, + { + "epoch": 0.8996876410157234, + "grad_norm": 4.116802215576172, + "learning_rate": 4.926495040396439e-06, + "loss": 0.1867, + "step": 38380 + }, + { + "epoch": 0.8997110825905285, + "grad_norm": 3.1410608291625977, + "learning_rate": 4.924211755362373e-06, + "loss": 0.2324, + "step": 38381 + }, + { + "epoch": 0.8997345241653334, + "grad_norm": 3.483708143234253, + "learning_rate": 4.921928986212199e-06, + "loss": 0.2908, + "step": 38382 + }, + { + "epoch": 0.8997579657401384, + "grad_norm": 0.9505228400230408, + "learning_rate": 4.919646732958316e-06, + "loss": 0.0627, + "step": 38383 + }, + { + "epoch": 0.8997814073149434, + "grad_norm": 1.4568474292755127, + "learning_rate": 4.917364995613094e-06, + "loss": 0.0726, + "step": 38384 + }, + { + "epoch": 0.8998048488897484, + "grad_norm": 1.347920536994934, + "learning_rate": 4.915083774188934e-06, + "loss": 0.0968, + "step": 38385 + }, + { + "epoch": 0.8998282904645534, + "grad_norm": 1.2967948913574219, + "learning_rate": 4.912803068698191e-06, + "loss": 0.1448, + "step": 38386 + }, + { + "epoch": 0.8998517320393584, + "grad_norm": 4.330373764038086, + "learning_rate": 4.9105228791532566e-06, + "loss": 0.4329, + "step": 38387 + }, + { + "epoch": 0.8998751736141634, + "grad_norm": 2.1850626468658447, + "learning_rate": 4.90824320556651e-06, + "loss": 0.1772, + "step": 38388 + }, + { + "epoch": 0.8998986151889684, + "grad_norm": 0.39619943499565125, + "learning_rate": 4.905964047950306e-06, + "loss": 0.0149, + "step": 38389 + }, + { + "epoch": 0.8999220567637733, + "grad_norm": 1.270337700843811, + "learning_rate": 4.903685406317016e-06, + "loss": 0.0791, + "step": 38390 + }, + { + "epoch": 0.8999454983385784, + "grad_norm": 3.775153875350952, + "learning_rate": 4.901407280678982e-06, + "loss": 0.311, + "step": 38391 + }, + { + "epoch": 0.8999689399133833, + "grad_norm": 3.4245011806488037, + "learning_rate": 4.899129671048608e-06, + "loss": 0.2621, + "step": 38392 + }, + { + "epoch": 0.8999923814881884, + "grad_norm": 2.3358380794525146, + "learning_rate": 4.896852577438227e-06, + "loss": 0.103, + "step": 38393 + }, + { + "epoch": 0.9000158230629933, + "grad_norm": 1.723556399345398, + "learning_rate": 4.894575999860196e-06, + "loss": 0.1731, + "step": 38394 + }, + { + "epoch": 0.9000392646377984, + "grad_norm": 2.9897243976593018, + "learning_rate": 4.892299938326872e-06, + "loss": 0.2319, + "step": 38395 + }, + { + "epoch": 0.9000627062126033, + "grad_norm": 1.2452507019042969, + "learning_rate": 4.890024392850601e-06, + "loss": 0.1628, + "step": 38396 + }, + { + "epoch": 0.9000861477874084, + "grad_norm": 3.9368343353271484, + "learning_rate": 4.887749363443728e-06, + "loss": 0.3431, + "step": 38397 + }, + { + "epoch": 0.9001095893622133, + "grad_norm": 3.4396910667419434, + "learning_rate": 4.885474850118587e-06, + "loss": 0.1362, + "step": 38398 + }, + { + "epoch": 0.9001330309370184, + "grad_norm": 0.30418428778648376, + "learning_rate": 4.8832008528875595e-06, + "loss": 0.017, + "step": 38399 + }, + { + "epoch": 0.9001564725118233, + "grad_norm": 1.6032367944717407, + "learning_rate": 4.880927371762955e-06, + "loss": 0.8071, + "step": 38400 + }, + { + "epoch": 0.9001799140866283, + "grad_norm": 2.7603747844696045, + "learning_rate": 4.878654406757111e-06, + "loss": 0.3318, + "step": 38401 + }, + { + "epoch": 0.9002033556614333, + "grad_norm": 1.9017928838729858, + "learning_rate": 4.8763819578823586e-06, + "loss": 0.1191, + "step": 38402 + }, + { + "epoch": 0.9002267972362383, + "grad_norm": 3.934235095977783, + "learning_rate": 4.874110025151035e-06, + "loss": 0.3873, + "step": 38403 + }, + { + "epoch": 0.9002502388110434, + "grad_norm": 3.363257646560669, + "learning_rate": 4.871838608575452e-06, + "loss": 0.2958, + "step": 38404 + }, + { + "epoch": 0.9002736803858483, + "grad_norm": 0.5248829126358032, + "learning_rate": 4.869567708167966e-06, + "loss": 0.0412, + "step": 38405 + }, + { + "epoch": 0.9002971219606534, + "grad_norm": 1.7977757453918457, + "learning_rate": 4.867297323940878e-06, + "loss": 0.0707, + "step": 38406 + }, + { + "epoch": 0.9003205635354583, + "grad_norm": 0.7325446605682373, + "learning_rate": 4.865027455906502e-06, + "loss": 0.0454, + "step": 38407 + }, + { + "epoch": 0.9003440051102634, + "grad_norm": 3.1711816787719727, + "learning_rate": 4.86275810407717e-06, + "loss": 0.2295, + "step": 38408 + }, + { + "epoch": 0.9003674466850683, + "grad_norm": 0.3889509439468384, + "learning_rate": 4.8604892684651735e-06, + "loss": 0.0257, + "step": 38409 + }, + { + "epoch": 0.9003908882598733, + "grad_norm": 2.899138927459717, + "learning_rate": 4.8582209490828474e-06, + "loss": 0.2089, + "step": 38410 + }, + { + "epoch": 0.9004143298346783, + "grad_norm": 4.34229040145874, + "learning_rate": 4.855953145942482e-06, + "loss": 0.4685, + "step": 38411 + }, + { + "epoch": 0.9004377714094833, + "grad_norm": 1.517685890197754, + "learning_rate": 4.853685859056401e-06, + "loss": 0.0695, + "step": 38412 + }, + { + "epoch": 0.9004612129842883, + "grad_norm": 2.2426064014434814, + "learning_rate": 4.8514190884368924e-06, + "loss": 0.1294, + "step": 38413 + }, + { + "epoch": 0.9004846545590933, + "grad_norm": 0.7579936385154724, + "learning_rate": 4.84915283409626e-06, + "loss": 0.0408, + "step": 38414 + }, + { + "epoch": 0.9005080961338983, + "grad_norm": 2.021207809448242, + "learning_rate": 4.8468870960468036e-06, + "loss": 0.1309, + "step": 38415 + }, + { + "epoch": 0.9005315377087033, + "grad_norm": 1.8634049892425537, + "learning_rate": 4.844621874300792e-06, + "loss": 1.2078, + "step": 38416 + }, + { + "epoch": 0.9005549792835083, + "grad_norm": 1.6979016065597534, + "learning_rate": 4.842357168870559e-06, + "loss": 0.0821, + "step": 38417 + }, + { + "epoch": 0.9005784208583133, + "grad_norm": 0.3897829055786133, + "learning_rate": 4.840092979768374e-06, + "loss": 0.0332, + "step": 38418 + }, + { + "epoch": 0.9006018624331182, + "grad_norm": 2.800110101699829, + "learning_rate": 4.837829307006514e-06, + "loss": 0.2592, + "step": 38419 + }, + { + "epoch": 0.9006253040079233, + "grad_norm": 1.572885513305664, + "learning_rate": 4.83556615059727e-06, + "loss": 0.1401, + "step": 38420 + }, + { + "epoch": 0.9006487455827282, + "grad_norm": 2.6413767337799072, + "learning_rate": 4.833303510552911e-06, + "loss": 0.1986, + "step": 38421 + }, + { + "epoch": 0.9006721871575333, + "grad_norm": 1.7533756494522095, + "learning_rate": 4.831041386885726e-06, + "loss": 0.0939, + "step": 38422 + }, + { + "epoch": 0.9006956287323382, + "grad_norm": 6.789989471435547, + "learning_rate": 4.828779779607972e-06, + "loss": 0.1866, + "step": 38423 + }, + { + "epoch": 0.9007190703071433, + "grad_norm": 2.182126760482788, + "learning_rate": 4.826518688731951e-06, + "loss": 0.1636, + "step": 38424 + }, + { + "epoch": 0.9007425118819482, + "grad_norm": 3.0872862339019775, + "learning_rate": 4.82425811426992e-06, + "loss": 0.2266, + "step": 38425 + }, + { + "epoch": 0.9007659534567533, + "grad_norm": 1.4217040538787842, + "learning_rate": 4.821998056234145e-06, + "loss": 0.1191, + "step": 38426 + }, + { + "epoch": 0.9007893950315582, + "grad_norm": 2.9509329795837402, + "learning_rate": 4.8197385146368734e-06, + "loss": 0.2696, + "step": 38427 + }, + { + "epoch": 0.9008128366063632, + "grad_norm": 3.8312835693359375, + "learning_rate": 4.817479489490384e-06, + "loss": 0.4495, + "step": 38428 + }, + { + "epoch": 0.9008362781811682, + "grad_norm": 2.058126449584961, + "learning_rate": 4.815220980806912e-06, + "loss": 0.1414, + "step": 38429 + }, + { + "epoch": 0.9008597197559732, + "grad_norm": 4.621946811676025, + "learning_rate": 4.812962988598746e-06, + "loss": 0.4521, + "step": 38430 + }, + { + "epoch": 0.9008831613307782, + "grad_norm": 0.4191737174987793, + "learning_rate": 4.810705512878111e-06, + "loss": 0.0341, + "step": 38431 + }, + { + "epoch": 0.9009066029055832, + "grad_norm": 3.449948310852051, + "learning_rate": 4.8084485536572745e-06, + "loss": 0.4562, + "step": 38432 + }, + { + "epoch": 0.9009300444803882, + "grad_norm": 2.2923226356506348, + "learning_rate": 4.8061921109484594e-06, + "loss": 0.1365, + "step": 38433 + }, + { + "epoch": 0.9009534860551932, + "grad_norm": 2.2540767192840576, + "learning_rate": 4.8039361847639355e-06, + "loss": 0.1969, + "step": 38434 + }, + { + "epoch": 0.9009769276299981, + "grad_norm": 2.688361167907715, + "learning_rate": 4.8016807751159024e-06, + "loss": 0.2548, + "step": 38435 + }, + { + "epoch": 0.9010003692048032, + "grad_norm": 3.2105772495269775, + "learning_rate": 4.799425882016639e-06, + "loss": 0.2024, + "step": 38436 + }, + { + "epoch": 0.9010238107796081, + "grad_norm": 1.6393781900405884, + "learning_rate": 4.79717150547837e-06, + "loss": 1.3215, + "step": 38437 + }, + { + "epoch": 0.9010472523544132, + "grad_norm": 3.0593082904815674, + "learning_rate": 4.794917645513342e-06, + "loss": 0.2456, + "step": 38438 + }, + { + "epoch": 0.9010706939292181, + "grad_norm": 3.2857372760772705, + "learning_rate": 4.7926643021337535e-06, + "loss": 0.1675, + "step": 38439 + }, + { + "epoch": 0.9010941355040232, + "grad_norm": 1.7038246393203735, + "learning_rate": 4.7904114753518525e-06, + "loss": 0.1342, + "step": 38440 + }, + { + "epoch": 0.9011175770788281, + "grad_norm": 1.2536219358444214, + "learning_rate": 4.788159165179851e-06, + "loss": 0.0934, + "step": 38441 + }, + { + "epoch": 0.9011410186536332, + "grad_norm": 0.9887672662734985, + "learning_rate": 4.78590737162996e-06, + "loss": 0.065, + "step": 38442 + }, + { + "epoch": 0.9011644602284381, + "grad_norm": 2.3075103759765625, + "learning_rate": 4.783656094714428e-06, + "loss": 0.1805, + "step": 38443 + }, + { + "epoch": 0.9011879018032432, + "grad_norm": 0.7104661464691162, + "learning_rate": 4.781405334445454e-06, + "loss": 0.0419, + "step": 38444 + }, + { + "epoch": 0.9012113433780481, + "grad_norm": 3.5206167697906494, + "learning_rate": 4.779155090835264e-06, + "loss": 0.1993, + "step": 38445 + }, + { + "epoch": 0.9012347849528531, + "grad_norm": 0.46076443791389465, + "learning_rate": 4.776905363896045e-06, + "loss": 0.0297, + "step": 38446 + }, + { + "epoch": 0.9012582265276581, + "grad_norm": 1.6628611087799072, + "learning_rate": 4.77465615364e-06, + "loss": 0.1003, + "step": 38447 + }, + { + "epoch": 0.9012816681024631, + "grad_norm": 5.034276008605957, + "learning_rate": 4.772407460079353e-06, + "loss": 0.3295, + "step": 38448 + }, + { + "epoch": 0.9013051096772681, + "grad_norm": 3.0527591705322266, + "learning_rate": 4.7701592832263165e-06, + "loss": 0.4863, + "step": 38449 + }, + { + "epoch": 0.9013285512520731, + "grad_norm": 0.965759813785553, + "learning_rate": 4.767911623093069e-06, + "loss": 0.036, + "step": 38450 + }, + { + "epoch": 0.9013519928268781, + "grad_norm": 1.4023874998092651, + "learning_rate": 4.765664479691811e-06, + "loss": 1.3333, + "step": 38451 + }, + { + "epoch": 0.9013754344016831, + "grad_norm": 2.3582091331481934, + "learning_rate": 4.763417853034735e-06, + "loss": 0.2415, + "step": 38452 + }, + { + "epoch": 0.901398875976488, + "grad_norm": 1.4759814739227295, + "learning_rate": 4.761171743134029e-06, + "loss": 0.0717, + "step": 38453 + }, + { + "epoch": 0.9014223175512931, + "grad_norm": 3.577427387237549, + "learning_rate": 4.758926150001874e-06, + "loss": 0.2682, + "step": 38454 + }, + { + "epoch": 0.9014457591260981, + "grad_norm": 3.1262245178222656, + "learning_rate": 4.7566810736504705e-06, + "loss": 0.2011, + "step": 38455 + }, + { + "epoch": 0.9014692007009031, + "grad_norm": 3.48362135887146, + "learning_rate": 4.754436514091998e-06, + "loss": 0.2594, + "step": 38456 + }, + { + "epoch": 0.9014926422757081, + "grad_norm": 1.1558340787887573, + "learning_rate": 4.752192471338635e-06, + "loss": 0.0432, + "step": 38457 + }, + { + "epoch": 0.9015160838505131, + "grad_norm": 1.2753056287765503, + "learning_rate": 4.749948945402549e-06, + "loss": 0.0787, + "step": 38458 + }, + { + "epoch": 0.9015395254253181, + "grad_norm": 3.3923792839050293, + "learning_rate": 4.74770593629591e-06, + "loss": 0.2191, + "step": 38459 + }, + { + "epoch": 0.9015629670001231, + "grad_norm": 2.591278076171875, + "learning_rate": 4.745463444030906e-06, + "loss": 0.2523, + "step": 38460 + }, + { + "epoch": 0.9015864085749281, + "grad_norm": 1.7647150754928589, + "learning_rate": 4.743221468619685e-06, + "loss": 0.141, + "step": 38461 + }, + { + "epoch": 0.901609850149733, + "grad_norm": 2.9949429035186768, + "learning_rate": 4.740980010074425e-06, + "loss": 0.2727, + "step": 38462 + }, + { + "epoch": 0.9016332917245381, + "grad_norm": 2.0102806091308594, + "learning_rate": 4.738739068407305e-06, + "loss": 0.4329, + "step": 38463 + }, + { + "epoch": 0.901656733299343, + "grad_norm": 3.070863723754883, + "learning_rate": 4.736498643630449e-06, + "loss": 0.1118, + "step": 38464 + }, + { + "epoch": 0.9016801748741481, + "grad_norm": 4.136011123657227, + "learning_rate": 4.734258735756037e-06, + "loss": 0.3289, + "step": 38465 + }, + { + "epoch": 0.901703616448953, + "grad_norm": 1.3344204425811768, + "learning_rate": 4.732019344796212e-06, + "loss": 1.6436, + "step": 38466 + }, + { + "epoch": 0.9017270580237581, + "grad_norm": 2.5734846591949463, + "learning_rate": 4.729780470763123e-06, + "loss": 0.2785, + "step": 38467 + }, + { + "epoch": 0.901750499598563, + "grad_norm": 2.866816282272339, + "learning_rate": 4.7275421136689255e-06, + "loss": 0.2219, + "step": 38468 + }, + { + "epoch": 0.9017739411733681, + "grad_norm": 0.9271827340126038, + "learning_rate": 4.725304273525777e-06, + "loss": 0.0824, + "step": 38469 + }, + { + "epoch": 0.901797382748173, + "grad_norm": 2.837744951248169, + "learning_rate": 4.723066950345801e-06, + "loss": 0.1315, + "step": 38470 + }, + { + "epoch": 0.901820824322978, + "grad_norm": 2.1915111541748047, + "learning_rate": 4.72083014414112e-06, + "loss": 0.1156, + "step": 38471 + }, + { + "epoch": 0.901844265897783, + "grad_norm": 0.4376201629638672, + "learning_rate": 4.718593854923925e-06, + "loss": 0.041, + "step": 38472 + }, + { + "epoch": 0.901867707472588, + "grad_norm": 2.6387577056884766, + "learning_rate": 4.716358082706307e-06, + "loss": 0.1709, + "step": 38473 + }, + { + "epoch": 0.901891149047393, + "grad_norm": 1.032145380973816, + "learning_rate": 4.714122827500389e-06, + "loss": 0.1353, + "step": 38474 + }, + { + "epoch": 0.901914590622198, + "grad_norm": 2.6461181640625, + "learning_rate": 4.711888089318351e-06, + "loss": 0.1657, + "step": 38475 + }, + { + "epoch": 0.901938032197003, + "grad_norm": 1.4473565816879272, + "learning_rate": 4.70965386817227e-06, + "loss": 0.0716, + "step": 38476 + }, + { + "epoch": 0.901961473771808, + "grad_norm": 1.145905613899231, + "learning_rate": 4.707420164074294e-06, + "loss": 0.1358, + "step": 38477 + }, + { + "epoch": 0.901984915346613, + "grad_norm": 3.408883810043335, + "learning_rate": 4.705186977036546e-06, + "loss": 0.1286, + "step": 38478 + }, + { + "epoch": 0.902008356921418, + "grad_norm": 3.260166883468628, + "learning_rate": 4.702954307071117e-06, + "loss": 0.158, + "step": 38479 + }, + { + "epoch": 0.902031798496223, + "grad_norm": 0.6382853388786316, + "learning_rate": 4.7007221541901295e-06, + "loss": 0.0512, + "step": 38480 + }, + { + "epoch": 0.902055240071028, + "grad_norm": 1.633447527885437, + "learning_rate": 4.698490518405718e-06, + "loss": 0.12, + "step": 38481 + }, + { + "epoch": 0.9020786816458329, + "grad_norm": 3.008981227874756, + "learning_rate": 4.696259399729974e-06, + "loss": 0.3528, + "step": 38482 + }, + { + "epoch": 0.902102123220638, + "grad_norm": 2.398341417312622, + "learning_rate": 4.694028798174999e-06, + "loss": 0.2867, + "step": 38483 + }, + { + "epoch": 0.9021255647954429, + "grad_norm": 4.301868915557861, + "learning_rate": 4.691798713752915e-06, + "loss": 0.3581, + "step": 38484 + }, + { + "epoch": 0.902149006370248, + "grad_norm": 2.545542001724243, + "learning_rate": 4.689569146475803e-06, + "loss": 0.107, + "step": 38485 + }, + { + "epoch": 0.9021724479450529, + "grad_norm": 8.234943389892578, + "learning_rate": 4.687340096355764e-06, + "loss": 0.5139, + "step": 38486 + }, + { + "epoch": 0.902195889519858, + "grad_norm": 2.071000099182129, + "learning_rate": 4.685111563404909e-06, + "loss": 0.7915, + "step": 38487 + }, + { + "epoch": 0.9022193310946629, + "grad_norm": 4.814145088195801, + "learning_rate": 4.68288354763532e-06, + "loss": 0.4586, + "step": 38488 + }, + { + "epoch": 0.902242772669468, + "grad_norm": 1.3356245756149292, + "learning_rate": 4.680656049059073e-06, + "loss": 0.1053, + "step": 38489 + }, + { + "epoch": 0.9022662142442729, + "grad_norm": 3.451918363571167, + "learning_rate": 4.678429067688283e-06, + "loss": 0.1303, + "step": 38490 + }, + { + "epoch": 0.9022896558190779, + "grad_norm": 3.437654733657837, + "learning_rate": 4.676202603535007e-06, + "loss": 0.2223, + "step": 38491 + }, + { + "epoch": 0.9023130973938829, + "grad_norm": 0.6788774728775024, + "learning_rate": 4.673976656611323e-06, + "loss": 0.0309, + "step": 38492 + }, + { + "epoch": 0.9023365389686879, + "grad_norm": 1.1317414045333862, + "learning_rate": 4.671751226929344e-06, + "loss": 0.0734, + "step": 38493 + }, + { + "epoch": 0.9023599805434929, + "grad_norm": 4.617288112640381, + "learning_rate": 4.669526314501116e-06, + "loss": 0.4555, + "step": 38494 + }, + { + "epoch": 0.9023834221182979, + "grad_norm": 2.113985776901245, + "learning_rate": 4.667301919338718e-06, + "loss": 0.1018, + "step": 38495 + }, + { + "epoch": 0.9024068636931029, + "grad_norm": 3.772313356399536, + "learning_rate": 4.665078041454207e-06, + "loss": 0.2549, + "step": 38496 + }, + { + "epoch": 0.9024303052679079, + "grad_norm": 3.145486831665039, + "learning_rate": 4.662854680859685e-06, + "loss": 0.3171, + "step": 38497 + }, + { + "epoch": 0.9024537468427128, + "grad_norm": 3.1404476165771484, + "learning_rate": 4.660631837567186e-06, + "loss": 0.3626, + "step": 38498 + }, + { + "epoch": 0.9024771884175179, + "grad_norm": 1.7431682348251343, + "learning_rate": 4.658409511588768e-06, + "loss": 0.1616, + "step": 38499 + }, + { + "epoch": 0.9025006299923228, + "grad_norm": 2.2901601791381836, + "learning_rate": 4.656187702936509e-06, + "loss": 0.0756, + "step": 38500 + }, + { + "epoch": 0.9025240715671279, + "grad_norm": 2.225559711456299, + "learning_rate": 4.653966411622457e-06, + "loss": 1.2624, + "step": 38501 + }, + { + "epoch": 0.9025475131419328, + "grad_norm": 0.9111387133598328, + "learning_rate": 4.6517456376586775e-06, + "loss": 0.0644, + "step": 38502 + }, + { + "epoch": 0.9025709547167379, + "grad_norm": 0.6843660473823547, + "learning_rate": 4.649525381057196e-06, + "loss": 0.0798, + "step": 38503 + }, + { + "epoch": 0.9025943962915428, + "grad_norm": 4.03378963470459, + "learning_rate": 4.6473056418300686e-06, + "loss": 0.1877, + "step": 38504 + }, + { + "epoch": 0.9026178378663479, + "grad_norm": 0.7409359812736511, + "learning_rate": 4.645086419989331e-06, + "loss": 0.0302, + "step": 38505 + }, + { + "epoch": 0.9026412794411529, + "grad_norm": 1.0680311918258667, + "learning_rate": 4.64286771554705e-06, + "loss": 0.0918, + "step": 38506 + }, + { + "epoch": 0.9026647210159578, + "grad_norm": 1.5131908655166626, + "learning_rate": 4.64064952851525e-06, + "loss": 1.2132, + "step": 38507 + }, + { + "epoch": 0.9026881625907629, + "grad_norm": 1.0340607166290283, + "learning_rate": 4.6384318589059565e-06, + "loss": 0.0695, + "step": 38508 + }, + { + "epoch": 0.9027116041655678, + "grad_norm": 2.8367600440979004, + "learning_rate": 4.636214706731224e-06, + "loss": 0.18, + "step": 38509 + }, + { + "epoch": 0.9027350457403729, + "grad_norm": 3.5551159381866455, + "learning_rate": 4.633998072003076e-06, + "loss": 0.2074, + "step": 38510 + }, + { + "epoch": 0.9027584873151778, + "grad_norm": 2.6366817951202393, + "learning_rate": 4.631781954733527e-06, + "loss": 0.2916, + "step": 38511 + }, + { + "epoch": 0.9027819288899829, + "grad_norm": 2.6063506603240967, + "learning_rate": 4.6295663549346e-06, + "loss": 0.2692, + "step": 38512 + }, + { + "epoch": 0.9028053704647878, + "grad_norm": 2.462543249130249, + "learning_rate": 4.627351272618341e-06, + "loss": 0.8025, + "step": 38513 + }, + { + "epoch": 0.9028288120395929, + "grad_norm": 2.3235981464385986, + "learning_rate": 4.625136707796763e-06, + "loss": 0.3127, + "step": 38514 + }, + { + "epoch": 0.9028522536143978, + "grad_norm": 1.397073745727539, + "learning_rate": 4.622922660481865e-06, + "loss": 0.0752, + "step": 38515 + }, + { + "epoch": 0.9028756951892029, + "grad_norm": 2.990363359451294, + "learning_rate": 4.620709130685674e-06, + "loss": 0.2928, + "step": 38516 + }, + { + "epoch": 0.9028991367640078, + "grad_norm": 0.7555739283561707, + "learning_rate": 4.618496118420202e-06, + "loss": 0.0395, + "step": 38517 + }, + { + "epoch": 0.9029225783388128, + "grad_norm": 0.15201842784881592, + "learning_rate": 4.616283623697437e-06, + "loss": 0.0055, + "step": 38518 + }, + { + "epoch": 0.9029460199136178, + "grad_norm": 3.0999667644500732, + "learning_rate": 4.614071646529405e-06, + "loss": 0.2341, + "step": 38519 + }, + { + "epoch": 0.9029694614884228, + "grad_norm": 0.7550830841064453, + "learning_rate": 4.611860186928097e-06, + "loss": 0.109, + "step": 38520 + }, + { + "epoch": 0.9029929030632278, + "grad_norm": 2.5036354064941406, + "learning_rate": 4.6096492449055225e-06, + "loss": 0.0966, + "step": 38521 + }, + { + "epoch": 0.9030163446380328, + "grad_norm": 3.047726631164551, + "learning_rate": 4.607438820473686e-06, + "loss": 0.2828, + "step": 38522 + }, + { + "epoch": 0.9030397862128378, + "grad_norm": 2.009270429611206, + "learning_rate": 4.6052289136445545e-06, + "loss": 1.4224, + "step": 38523 + }, + { + "epoch": 0.9030632277876428, + "grad_norm": 2.438790798187256, + "learning_rate": 4.60301952443013e-06, + "loss": 0.2426, + "step": 38524 + }, + { + "epoch": 0.9030866693624477, + "grad_norm": 1.9900038242340088, + "learning_rate": 4.600810652842413e-06, + "loss": 0.1931, + "step": 38525 + }, + { + "epoch": 0.9031101109372528, + "grad_norm": 2.5107638835906982, + "learning_rate": 4.598602298893384e-06, + "loss": 0.2458, + "step": 38526 + }, + { + "epoch": 0.9031335525120577, + "grad_norm": 5.345643520355225, + "learning_rate": 4.59639446259501e-06, + "loss": 0.3883, + "step": 38527 + }, + { + "epoch": 0.9031569940868628, + "grad_norm": 4.114660739898682, + "learning_rate": 4.594187143959294e-06, + "loss": 0.292, + "step": 38528 + }, + { + "epoch": 0.9031804356616677, + "grad_norm": 2.287869930267334, + "learning_rate": 4.591980342998203e-06, + "loss": 0.0992, + "step": 38529 + }, + { + "epoch": 0.9032038772364728, + "grad_norm": 0.621562659740448, + "learning_rate": 4.5897740597236835e-06, + "loss": 0.0752, + "step": 38530 + }, + { + "epoch": 0.9032273188112777, + "grad_norm": 0.20788677036762238, + "learning_rate": 4.587568294147759e-06, + "loss": 0.0066, + "step": 38531 + }, + { + "epoch": 0.9032507603860828, + "grad_norm": 1.594599962234497, + "learning_rate": 4.5853630462823425e-06, + "loss": 0.0912, + "step": 38532 + }, + { + "epoch": 0.9032742019608877, + "grad_norm": 0.7025209665298462, + "learning_rate": 4.583158316139458e-06, + "loss": 0.0546, + "step": 38533 + }, + { + "epoch": 0.9032976435356928, + "grad_norm": 3.288912773132324, + "learning_rate": 4.5809541037310275e-06, + "loss": 0.2451, + "step": 38534 + }, + { + "epoch": 0.9033210851104977, + "grad_norm": 1.3359096050262451, + "learning_rate": 4.5787504090690324e-06, + "loss": 0.0783, + "step": 38535 + }, + { + "epoch": 0.9033445266853027, + "grad_norm": 2.3760440349578857, + "learning_rate": 4.576547232165418e-06, + "loss": 0.1327, + "step": 38536 + }, + { + "epoch": 0.9033679682601077, + "grad_norm": 0.42403292655944824, + "learning_rate": 4.57434457303213e-06, + "loss": 0.0272, + "step": 38537 + }, + { + "epoch": 0.9033914098349127, + "grad_norm": 4.819157600402832, + "learning_rate": 4.5721424316811366e-06, + "loss": 0.1776, + "step": 38538 + }, + { + "epoch": 0.9034148514097177, + "grad_norm": 0.8858511447906494, + "learning_rate": 4.569940808124384e-06, + "loss": 0.1185, + "step": 38539 + }, + { + "epoch": 0.9034382929845227, + "grad_norm": 2.6284468173980713, + "learning_rate": 4.567739702373819e-06, + "loss": 0.2113, + "step": 38540 + }, + { + "epoch": 0.9034617345593277, + "grad_norm": 0.40998274087905884, + "learning_rate": 4.565539114441386e-06, + "loss": 0.0224, + "step": 38541 + }, + { + "epoch": 0.9034851761341327, + "grad_norm": 1.0190434455871582, + "learning_rate": 4.563339044339021e-06, + "loss": 0.0758, + "step": 38542 + }, + { + "epoch": 0.9035086177089376, + "grad_norm": 5.841423034667969, + "learning_rate": 4.561139492078637e-06, + "loss": 0.4853, + "step": 38543 + }, + { + "epoch": 0.9035320592837427, + "grad_norm": 3.3436849117279053, + "learning_rate": 4.5589404576722005e-06, + "loss": 0.2269, + "step": 38544 + }, + { + "epoch": 0.9035555008585476, + "grad_norm": 0.6623375415802002, + "learning_rate": 4.556741941131648e-06, + "loss": 0.0211, + "step": 38545 + }, + { + "epoch": 0.9035789424333527, + "grad_norm": 2.730464220046997, + "learning_rate": 4.554543942468903e-06, + "loss": 0.1172, + "step": 38546 + }, + { + "epoch": 0.9036023840081576, + "grad_norm": 1.3573577404022217, + "learning_rate": 4.552346461695878e-06, + "loss": 0.051, + "step": 38547 + }, + { + "epoch": 0.9036258255829627, + "grad_norm": 2.554730176925659, + "learning_rate": 4.5501494988245184e-06, + "loss": 0.0672, + "step": 38548 + }, + { + "epoch": 0.9036492671577676, + "grad_norm": 2.8136324882507324, + "learning_rate": 4.547953053866716e-06, + "loss": 0.1758, + "step": 38549 + }, + { + "epoch": 0.9036727087325727, + "grad_norm": 3.078810214996338, + "learning_rate": 4.545757126834405e-06, + "loss": 0.2414, + "step": 38550 + }, + { + "epoch": 0.9036961503073776, + "grad_norm": 0.8973540663719177, + "learning_rate": 4.543561717739497e-06, + "loss": 0.0976, + "step": 38551 + }, + { + "epoch": 0.9037195918821826, + "grad_norm": 3.1175429821014404, + "learning_rate": 4.541366826593918e-06, + "loss": 0.2858, + "step": 38552 + }, + { + "epoch": 0.9037430334569876, + "grad_norm": 0.30652710795402527, + "learning_rate": 4.539172453409568e-06, + "loss": 0.0309, + "step": 38553 + }, + { + "epoch": 0.9037664750317926, + "grad_norm": 1.2063331604003906, + "learning_rate": 4.536978598198349e-06, + "loss": 0.1036, + "step": 38554 + }, + { + "epoch": 0.9037899166065976, + "grad_norm": 0.3861486315727234, + "learning_rate": 4.534785260972152e-06, + "loss": 0.0343, + "step": 38555 + }, + { + "epoch": 0.9038133581814026, + "grad_norm": 1.804731845855713, + "learning_rate": 4.532592441742911e-06, + "loss": 0.0699, + "step": 38556 + }, + { + "epoch": 0.9038367997562077, + "grad_norm": 3.13913893699646, + "learning_rate": 4.530400140522495e-06, + "loss": 0.1253, + "step": 38557 + }, + { + "epoch": 0.9038602413310126, + "grad_norm": 3.9211106300354004, + "learning_rate": 4.528208357322816e-06, + "loss": 0.3147, + "step": 38558 + }, + { + "epoch": 0.9038836829058177, + "grad_norm": 0.8745445013046265, + "learning_rate": 4.526017092155776e-06, + "loss": 0.0833, + "step": 38559 + }, + { + "epoch": 0.9039071244806226, + "grad_norm": 2.713179111480713, + "learning_rate": 4.523826345033244e-06, + "loss": 0.2613, + "step": 38560 + }, + { + "epoch": 0.9039305660554277, + "grad_norm": 3.5408661365509033, + "learning_rate": 4.521636115967121e-06, + "loss": 0.3323, + "step": 38561 + }, + { + "epoch": 0.9039540076302326, + "grad_norm": 1.5660003423690796, + "learning_rate": 4.519446404969263e-06, + "loss": 0.0695, + "step": 38562 + }, + { + "epoch": 0.9039774492050376, + "grad_norm": 3.2445051670074463, + "learning_rate": 4.5172572120515955e-06, + "loss": 0.2409, + "step": 38563 + }, + { + "epoch": 0.9040008907798426, + "grad_norm": 2.267155408859253, + "learning_rate": 4.515068537225975e-06, + "loss": 0.1066, + "step": 38564 + }, + { + "epoch": 0.9040243323546476, + "grad_norm": 1.3075957298278809, + "learning_rate": 4.51288038050427e-06, + "loss": 0.1439, + "step": 38565 + }, + { + "epoch": 0.9040477739294526, + "grad_norm": 3.8126778602600098, + "learning_rate": 4.5106927418983704e-06, + "loss": 0.2754, + "step": 38566 + }, + { + "epoch": 0.9040712155042576, + "grad_norm": 3.266317129135132, + "learning_rate": 4.5085056214201335e-06, + "loss": 0.2833, + "step": 38567 + }, + { + "epoch": 0.9040946570790626, + "grad_norm": 2.258584499359131, + "learning_rate": 4.506319019081407e-06, + "loss": 1.4727, + "step": 38568 + }, + { + "epoch": 0.9041180986538676, + "grad_norm": 3.0577552318573, + "learning_rate": 4.50413293489409e-06, + "loss": 0.1763, + "step": 38569 + }, + { + "epoch": 0.9041415402286725, + "grad_norm": 2.952765464782715, + "learning_rate": 4.501947368870041e-06, + "loss": 0.2853, + "step": 38570 + }, + { + "epoch": 0.9041649818034776, + "grad_norm": 0.7483669519424438, + "learning_rate": 4.499762321021117e-06, + "loss": 0.0762, + "step": 38571 + }, + { + "epoch": 0.9041884233782825, + "grad_norm": 0.45732060074806213, + "learning_rate": 4.497577791359164e-06, + "loss": 0.0322, + "step": 38572 + }, + { + "epoch": 0.9042118649530876, + "grad_norm": 3.6372294425964355, + "learning_rate": 4.49539377989604e-06, + "loss": 0.2551, + "step": 38573 + }, + { + "epoch": 0.9042353065278925, + "grad_norm": 0.5013499855995178, + "learning_rate": 4.49321028664359e-06, + "loss": 0.0525, + "step": 38574 + }, + { + "epoch": 0.9042587481026976, + "grad_norm": 1.2034847736358643, + "learning_rate": 4.4910273116136495e-06, + "loss": 0.1197, + "step": 38575 + }, + { + "epoch": 0.9042821896775025, + "grad_norm": 0.29154330492019653, + "learning_rate": 4.488844854818108e-06, + "loss": 0.0242, + "step": 38576 + }, + { + "epoch": 0.9043056312523076, + "grad_norm": 2.01082706451416, + "learning_rate": 4.486662916268769e-06, + "loss": 0.7697, + "step": 38577 + }, + { + "epoch": 0.9043290728271125, + "grad_norm": 2.744478702545166, + "learning_rate": 4.484481495977488e-06, + "loss": 0.1967, + "step": 38578 + }, + { + "epoch": 0.9043525144019176, + "grad_norm": 2.7185442447662354, + "learning_rate": 4.48230059395609e-06, + "loss": 0.1773, + "step": 38579 + }, + { + "epoch": 0.9043759559767225, + "grad_norm": 4.737734317779541, + "learning_rate": 4.480120210216421e-06, + "loss": 0.486, + "step": 38580 + }, + { + "epoch": 0.9043993975515275, + "grad_norm": 4.453315734863281, + "learning_rate": 4.477940344770293e-06, + "loss": 0.3341, + "step": 38581 + }, + { + "epoch": 0.9044228391263325, + "grad_norm": 1.0691050291061401, + "learning_rate": 4.475760997629541e-06, + "loss": 0.0602, + "step": 38582 + }, + { + "epoch": 0.9044462807011375, + "grad_norm": 2.472033977508545, + "learning_rate": 4.473582168806012e-06, + "loss": 0.172, + "step": 38583 + }, + { + "epoch": 0.9044697222759425, + "grad_norm": 1.13400137424469, + "learning_rate": 4.471403858311507e-06, + "loss": 0.8023, + "step": 38584 + }, + { + "epoch": 0.9044931638507475, + "grad_norm": 1.1246614456176758, + "learning_rate": 4.469226066157861e-06, + "loss": 0.0417, + "step": 38585 + }, + { + "epoch": 0.9045166054255525, + "grad_norm": 2.3457143306732178, + "learning_rate": 4.467048792356876e-06, + "loss": 0.0711, + "step": 38586 + }, + { + "epoch": 0.9045400470003575, + "grad_norm": 1.9885846376419067, + "learning_rate": 4.464872036920365e-06, + "loss": 0.2288, + "step": 38587 + }, + { + "epoch": 0.9045634885751624, + "grad_norm": 2.832407236099243, + "learning_rate": 4.462695799860139e-06, + "loss": 0.2596, + "step": 38588 + }, + { + "epoch": 0.9045869301499675, + "grad_norm": 2.209726572036743, + "learning_rate": 4.460520081188014e-06, + "loss": 0.1867, + "step": 38589 + }, + { + "epoch": 0.9046103717247724, + "grad_norm": 4.665188789367676, + "learning_rate": 4.458344880915799e-06, + "loss": 0.5765, + "step": 38590 + }, + { + "epoch": 0.9046338132995775, + "grad_norm": 1.8957725763320923, + "learning_rate": 4.456170199055298e-06, + "loss": 0.0812, + "step": 38591 + }, + { + "epoch": 0.9046572548743824, + "grad_norm": 1.6142737865447998, + "learning_rate": 4.453996035618302e-06, + "loss": 0.1281, + "step": 38592 + }, + { + "epoch": 0.9046806964491875, + "grad_norm": 0.4798720180988312, + "learning_rate": 4.451822390616589e-06, + "loss": 0.0359, + "step": 38593 + }, + { + "epoch": 0.9047041380239924, + "grad_norm": 1.3492830991744995, + "learning_rate": 4.449649264061995e-06, + "loss": 0.083, + "step": 38594 + }, + { + "epoch": 0.9047275795987975, + "grad_norm": 1.4710805416107178, + "learning_rate": 4.4474766559662764e-06, + "loss": 0.0761, + "step": 38595 + }, + { + "epoch": 0.9047510211736024, + "grad_norm": 1.900087594985962, + "learning_rate": 4.445304566341246e-06, + "loss": 0.1689, + "step": 38596 + }, + { + "epoch": 0.9047744627484074, + "grad_norm": 2.8328773975372314, + "learning_rate": 4.4431329951986735e-06, + "loss": 0.3585, + "step": 38597 + }, + { + "epoch": 0.9047979043232124, + "grad_norm": 2.1526694297790527, + "learning_rate": 4.440961942550359e-06, + "loss": 0.1699, + "step": 38598 + }, + { + "epoch": 0.9048213458980174, + "grad_norm": 2.2399234771728516, + "learning_rate": 4.438791408408072e-06, + "loss": 0.1443, + "step": 38599 + }, + { + "epoch": 0.9048447874728224, + "grad_norm": 3.008939027786255, + "learning_rate": 4.436621392783569e-06, + "loss": 0.2862, + "step": 38600 + }, + { + "epoch": 0.9048682290476274, + "grad_norm": 4.49144172668457, + "learning_rate": 4.4344518956886625e-06, + "loss": 0.4964, + "step": 38601 + }, + { + "epoch": 0.9048916706224324, + "grad_norm": 1.843912124633789, + "learning_rate": 4.432282917135111e-06, + "loss": 0.0846, + "step": 38602 + }, + { + "epoch": 0.9049151121972374, + "grad_norm": 1.5845496654510498, + "learning_rate": 4.430114457134671e-06, + "loss": 0.9025, + "step": 38603 + }, + { + "epoch": 0.9049385537720424, + "grad_norm": 0.6609578132629395, + "learning_rate": 4.427946515699121e-06, + "loss": 0.0526, + "step": 38604 + }, + { + "epoch": 0.9049619953468474, + "grad_norm": 2.111151933670044, + "learning_rate": 4.42577909284021e-06, + "loss": 0.0627, + "step": 38605 + }, + { + "epoch": 0.9049854369216523, + "grad_norm": 1.7783883810043335, + "learning_rate": 4.423612188569725e-06, + "loss": 0.0642, + "step": 38606 + }, + { + "epoch": 0.9050088784964574, + "grad_norm": 6.199720859527588, + "learning_rate": 4.421445802899394e-06, + "loss": 0.3368, + "step": 38607 + }, + { + "epoch": 0.9050323200712624, + "grad_norm": 0.9752307534217834, + "learning_rate": 4.419279935840992e-06, + "loss": 0.0801, + "step": 38608 + }, + { + "epoch": 0.9050557616460674, + "grad_norm": 3.297865629196167, + "learning_rate": 4.4171145874062684e-06, + "loss": 0.0868, + "step": 38609 + }, + { + "epoch": 0.9050792032208724, + "grad_norm": 2.7711880207061768, + "learning_rate": 4.4149497576069675e-06, + "loss": 0.1945, + "step": 38610 + }, + { + "epoch": 0.9051026447956774, + "grad_norm": 2.5326085090637207, + "learning_rate": 4.412785446454837e-06, + "loss": 0.1586, + "step": 38611 + }, + { + "epoch": 0.9051260863704824, + "grad_norm": 2.1683738231658936, + "learning_rate": 4.4106216539616215e-06, + "loss": 0.0901, + "step": 38612 + }, + { + "epoch": 0.9051495279452874, + "grad_norm": 2.7725393772125244, + "learning_rate": 4.408458380139047e-06, + "loss": 0.0886, + "step": 38613 + }, + { + "epoch": 0.9051729695200924, + "grad_norm": 1.462687373161316, + "learning_rate": 4.4062956249988685e-06, + "loss": 0.0944, + "step": 38614 + }, + { + "epoch": 0.9051964110948973, + "grad_norm": 1.5908875465393066, + "learning_rate": 4.404133388552833e-06, + "loss": 0.0755, + "step": 38615 + }, + { + "epoch": 0.9052198526697024, + "grad_norm": 2.278799533843994, + "learning_rate": 4.401971670812655e-06, + "loss": 0.0941, + "step": 38616 + }, + { + "epoch": 0.9052432942445073, + "grad_norm": 2.194554090499878, + "learning_rate": 4.399810471790045e-06, + "loss": 0.1947, + "step": 38617 + }, + { + "epoch": 0.9052667358193124, + "grad_norm": 2.7209107875823975, + "learning_rate": 4.397649791496772e-06, + "loss": 0.3328, + "step": 38618 + }, + { + "epoch": 0.9052901773941173, + "grad_norm": 0.48469629883766174, + "learning_rate": 4.395489629944538e-06, + "loss": 0.0344, + "step": 38619 + }, + { + "epoch": 0.9053136189689224, + "grad_norm": 4.448728561401367, + "learning_rate": 4.3933299871450455e-06, + "loss": 0.3859, + "step": 38620 + }, + { + "epoch": 0.9053370605437273, + "grad_norm": 0.142959326505661, + "learning_rate": 4.39117086311005e-06, + "loss": 0.009, + "step": 38621 + }, + { + "epoch": 0.9053605021185324, + "grad_norm": 2.1082265377044678, + "learning_rate": 4.389012257851255e-06, + "loss": 0.1165, + "step": 38622 + }, + { + "epoch": 0.9053839436933373, + "grad_norm": 3.38826847076416, + "learning_rate": 4.386854171380361e-06, + "loss": 0.2846, + "step": 38623 + }, + { + "epoch": 0.9054073852681424, + "grad_norm": 2.0319294929504395, + "learning_rate": 4.384696603709093e-06, + "loss": 0.1911, + "step": 38624 + }, + { + "epoch": 0.9054308268429473, + "grad_norm": 1.3012174367904663, + "learning_rate": 4.382539554849141e-06, + "loss": 0.0783, + "step": 38625 + }, + { + "epoch": 0.9054542684177523, + "grad_norm": 3.9710638523101807, + "learning_rate": 4.380383024812207e-06, + "loss": 0.4126, + "step": 38626 + }, + { + "epoch": 0.9054777099925573, + "grad_norm": 0.6190412044525146, + "learning_rate": 4.378227013610015e-06, + "loss": 0.0541, + "step": 38627 + }, + { + "epoch": 0.9055011515673623, + "grad_norm": 2.204909324645996, + "learning_rate": 4.3760715212542546e-06, + "loss": 0.1896, + "step": 38628 + }, + { + "epoch": 0.9055245931421673, + "grad_norm": 3.2280023097991943, + "learning_rate": 4.373916547756607e-06, + "loss": 0.3096, + "step": 38629 + }, + { + "epoch": 0.9055480347169723, + "grad_norm": 3.3123652935028076, + "learning_rate": 4.3717620931287954e-06, + "loss": 0.212, + "step": 38630 + }, + { + "epoch": 0.9055714762917773, + "grad_norm": 1.3967561721801758, + "learning_rate": 4.369608157382476e-06, + "loss": 0.0728, + "step": 38631 + }, + { + "epoch": 0.9055949178665823, + "grad_norm": 1.7623714208602905, + "learning_rate": 4.367454740529364e-06, + "loss": 1.0939, + "step": 38632 + }, + { + "epoch": 0.9056183594413872, + "grad_norm": 3.235807180404663, + "learning_rate": 4.3653018425811155e-06, + "loss": 0.2898, + "step": 38633 + }, + { + "epoch": 0.9056418010161923, + "grad_norm": 1.6568607091903687, + "learning_rate": 4.363149463549443e-06, + "loss": 0.0511, + "step": 38634 + }, + { + "epoch": 0.9056652425909972, + "grad_norm": 2.9643867015838623, + "learning_rate": 4.3609976034460045e-06, + "loss": 0.2122, + "step": 38635 + }, + { + "epoch": 0.9056886841658023, + "grad_norm": 3.7110798358917236, + "learning_rate": 4.35884626228249e-06, + "loss": 0.2001, + "step": 38636 + }, + { + "epoch": 0.9057121257406072, + "grad_norm": 1.5215405225753784, + "learning_rate": 4.3566954400705575e-06, + "loss": 0.0661, + "step": 38637 + }, + { + "epoch": 0.9057355673154123, + "grad_norm": 2.425058126449585, + "learning_rate": 4.354545136821875e-06, + "loss": 0.131, + "step": 38638 + }, + { + "epoch": 0.9057590088902172, + "grad_norm": 2.331968069076538, + "learning_rate": 4.352395352548133e-06, + "loss": 1.451, + "step": 38639 + }, + { + "epoch": 0.9057824504650223, + "grad_norm": 4.289159774780273, + "learning_rate": 4.350246087260979e-06, + "loss": 0.3227, + "step": 38640 + }, + { + "epoch": 0.9058058920398272, + "grad_norm": 5.081238746643066, + "learning_rate": 4.34809734097208e-06, + "loss": 0.4383, + "step": 38641 + }, + { + "epoch": 0.9058293336146322, + "grad_norm": 0.48286205530166626, + "learning_rate": 4.345949113693093e-06, + "loss": 0.0248, + "step": 38642 + }, + { + "epoch": 0.9058527751894372, + "grad_norm": 1.3147706985473633, + "learning_rate": 4.343801405435677e-06, + "loss": 0.1412, + "step": 38643 + }, + { + "epoch": 0.9058762167642422, + "grad_norm": 2.0020644664764404, + "learning_rate": 4.341654216211488e-06, + "loss": 0.1592, + "step": 38644 + }, + { + "epoch": 0.9058996583390472, + "grad_norm": 1.3289504051208496, + "learning_rate": 4.3395075460321624e-06, + "loss": 0.0629, + "step": 38645 + }, + { + "epoch": 0.9059230999138522, + "grad_norm": 1.65150785446167, + "learning_rate": 4.3373613949093675e-06, + "loss": 0.2066, + "step": 38646 + }, + { + "epoch": 0.9059465414886572, + "grad_norm": 0.21947740018367767, + "learning_rate": 4.335215762854738e-06, + "loss": 0.0232, + "step": 38647 + }, + { + "epoch": 0.9059699830634622, + "grad_norm": 2.337695837020874, + "learning_rate": 4.333070649879922e-06, + "loss": 0.2338, + "step": 38648 + }, + { + "epoch": 0.9059934246382672, + "grad_norm": 3.594219446182251, + "learning_rate": 4.330926055996553e-06, + "loss": 0.4605, + "step": 38649 + }, + { + "epoch": 0.9060168662130722, + "grad_norm": 2.457685708999634, + "learning_rate": 4.3287819812162675e-06, + "loss": 0.2029, + "step": 38650 + }, + { + "epoch": 0.9060403077878771, + "grad_norm": 1.9149501323699951, + "learning_rate": 4.3266384255506885e-06, + "loss": 0.1518, + "step": 38651 + }, + { + "epoch": 0.9060637493626822, + "grad_norm": 1.0076923370361328, + "learning_rate": 4.324495389011474e-06, + "loss": 0.0832, + "step": 38652 + }, + { + "epoch": 0.9060871909374871, + "grad_norm": 0.32817506790161133, + "learning_rate": 4.322352871610236e-06, + "loss": 0.0257, + "step": 38653 + }, + { + "epoch": 0.9061106325122922, + "grad_norm": 2.642340898513794, + "learning_rate": 4.320210873358588e-06, + "loss": 0.1562, + "step": 38654 + }, + { + "epoch": 0.9061340740870971, + "grad_norm": 3.4439737796783447, + "learning_rate": 4.3180693942681875e-06, + "loss": 0.1211, + "step": 38655 + }, + { + "epoch": 0.9061575156619022, + "grad_norm": 0.9083487391471863, + "learning_rate": 4.315928434350624e-06, + "loss": 0.0807, + "step": 38656 + }, + { + "epoch": 0.9061809572367071, + "grad_norm": 0.5684742331504822, + "learning_rate": 4.313787993617524e-06, + "loss": 0.0334, + "step": 38657 + }, + { + "epoch": 0.9062043988115122, + "grad_norm": 2.268458843231201, + "learning_rate": 4.311648072080488e-06, + "loss": 1.2146, + "step": 38658 + }, + { + "epoch": 0.9062278403863172, + "grad_norm": 5.197117328643799, + "learning_rate": 4.309508669751161e-06, + "loss": 0.5352, + "step": 38659 + }, + { + "epoch": 0.9062512819611221, + "grad_norm": 3.2048449516296387, + "learning_rate": 4.307369786641124e-06, + "loss": 0.1888, + "step": 38660 + }, + { + "epoch": 0.9062747235359272, + "grad_norm": 2.2171578407287598, + "learning_rate": 4.3052314227619796e-06, + "loss": 0.1805, + "step": 38661 + }, + { + "epoch": 0.9062981651107321, + "grad_norm": 1.8342885971069336, + "learning_rate": 4.3030935781253504e-06, + "loss": 1.0222, + "step": 38662 + }, + { + "epoch": 0.9063216066855372, + "grad_norm": 1.7234563827514648, + "learning_rate": 4.300956252742827e-06, + "loss": 0.0946, + "step": 38663 + }, + { + "epoch": 0.9063450482603421, + "grad_norm": 0.9563323259353638, + "learning_rate": 4.29881944662599e-06, + "loss": 0.0748, + "step": 38664 + }, + { + "epoch": 0.9063684898351472, + "grad_norm": 2.1598734855651855, + "learning_rate": 4.296683159786463e-06, + "loss": 0.1541, + "step": 38665 + }, + { + "epoch": 0.9063919314099521, + "grad_norm": 0.3792617917060852, + "learning_rate": 4.294547392235815e-06, + "loss": 0.0316, + "step": 38666 + }, + { + "epoch": 0.9064153729847572, + "grad_norm": 1.2838951349258423, + "learning_rate": 4.292412143985658e-06, + "loss": 0.0486, + "step": 38667 + }, + { + "epoch": 0.9064388145595621, + "grad_norm": 1.2198083400726318, + "learning_rate": 4.29027741504755e-06, + "loss": 0.1826, + "step": 38668 + }, + { + "epoch": 0.9064622561343672, + "grad_norm": 2.3862884044647217, + "learning_rate": 4.288143205433104e-06, + "loss": 0.2877, + "step": 38669 + }, + { + "epoch": 0.9064856977091721, + "grad_norm": 1.0166130065917969, + "learning_rate": 4.286009515153877e-06, + "loss": 0.0996, + "step": 38670 + }, + { + "epoch": 0.9065091392839771, + "grad_norm": 3.075491189956665, + "learning_rate": 4.283876344221438e-06, + "loss": 0.2805, + "step": 38671 + }, + { + "epoch": 0.9065325808587821, + "grad_norm": 5.541085720062256, + "learning_rate": 4.281743692647389e-06, + "loss": 0.4453, + "step": 38672 + }, + { + "epoch": 0.9065560224335871, + "grad_norm": 1.5187933444976807, + "learning_rate": 4.279611560443297e-06, + "loss": 0.1277, + "step": 38673 + }, + { + "epoch": 0.9065794640083921, + "grad_norm": 5.853801250457764, + "learning_rate": 4.277479947620721e-06, + "loss": 0.4217, + "step": 38674 + }, + { + "epoch": 0.9066029055831971, + "grad_norm": 0.7239189147949219, + "learning_rate": 4.27534885419123e-06, + "loss": 0.0374, + "step": 38675 + }, + { + "epoch": 0.906626347158002, + "grad_norm": 1.899804711341858, + "learning_rate": 4.27321828016638e-06, + "loss": 0.1342, + "step": 38676 + }, + { + "epoch": 0.9066497887328071, + "grad_norm": 2.455648899078369, + "learning_rate": 4.271088225557751e-06, + "loss": 0.1259, + "step": 38677 + }, + { + "epoch": 0.906673230307612, + "grad_norm": 0.794631838798523, + "learning_rate": 4.268958690376868e-06, + "loss": 0.0571, + "step": 38678 + }, + { + "epoch": 0.9066966718824171, + "grad_norm": 1.8456553220748901, + "learning_rate": 4.26682967463532e-06, + "loss": 0.1418, + "step": 38679 + }, + { + "epoch": 0.906720113457222, + "grad_norm": 2.3874130249023438, + "learning_rate": 4.264701178344654e-06, + "loss": 0.2205, + "step": 38680 + }, + { + "epoch": 0.9067435550320271, + "grad_norm": 2.2675418853759766, + "learning_rate": 4.2625732015164065e-06, + "loss": 0.2288, + "step": 38681 + }, + { + "epoch": 0.906766996606832, + "grad_norm": 2.4612321853637695, + "learning_rate": 4.260445744162134e-06, + "loss": 0.1977, + "step": 38682 + }, + { + "epoch": 0.9067904381816371, + "grad_norm": 3.760587692260742, + "learning_rate": 4.258318806293349e-06, + "loss": 0.0815, + "step": 38683 + }, + { + "epoch": 0.906813879756442, + "grad_norm": 2.444791316986084, + "learning_rate": 4.256192387921643e-06, + "loss": 1.7777, + "step": 38684 + }, + { + "epoch": 0.9068373213312471, + "grad_norm": 2.771182060241699, + "learning_rate": 4.254066489058528e-06, + "loss": 0.2433, + "step": 38685 + }, + { + "epoch": 0.906860762906052, + "grad_norm": 1.4591784477233887, + "learning_rate": 4.25194110971554e-06, + "loss": 0.0607, + "step": 38686 + }, + { + "epoch": 0.906884204480857, + "grad_norm": 3.2041893005371094, + "learning_rate": 4.249816249904204e-06, + "loss": 0.2103, + "step": 38687 + }, + { + "epoch": 0.906907646055662, + "grad_norm": 3.5922434329986572, + "learning_rate": 4.247691909636064e-06, + "loss": 0.2074, + "step": 38688 + }, + { + "epoch": 0.906931087630467, + "grad_norm": 1.945990800857544, + "learning_rate": 4.245568088922625e-06, + "loss": 0.1507, + "step": 38689 + }, + { + "epoch": 0.906954529205272, + "grad_norm": 2.3261067867279053, + "learning_rate": 4.2434447877754305e-06, + "loss": 0.2621, + "step": 38690 + }, + { + "epoch": 0.906977970780077, + "grad_norm": 2.975015640258789, + "learning_rate": 4.241322006206005e-06, + "loss": 0.1262, + "step": 38691 + }, + { + "epoch": 0.907001412354882, + "grad_norm": 1.669382095336914, + "learning_rate": 4.239199744225864e-06, + "loss": 1.263, + "step": 38692 + }, + { + "epoch": 0.907024853929687, + "grad_norm": 7.942165374755859, + "learning_rate": 4.237078001846517e-06, + "loss": 0.5191, + "step": 38693 + }, + { + "epoch": 0.907048295504492, + "grad_norm": 3.8664333820343018, + "learning_rate": 4.234956779079469e-06, + "loss": 0.3438, + "step": 38694 + }, + { + "epoch": 0.907071737079297, + "grad_norm": 1.7856361865997314, + "learning_rate": 4.232836075936242e-06, + "loss": 0.0411, + "step": 38695 + }, + { + "epoch": 0.9070951786541019, + "grad_norm": 3.999861717224121, + "learning_rate": 4.230715892428327e-06, + "loss": 0.1269, + "step": 38696 + }, + { + "epoch": 0.907118620228907, + "grad_norm": 1.6157140731811523, + "learning_rate": 4.228596228567261e-06, + "loss": 0.1632, + "step": 38697 + }, + { + "epoch": 0.9071420618037119, + "grad_norm": 2.3841183185577393, + "learning_rate": 4.226477084364522e-06, + "loss": 0.2466, + "step": 38698 + }, + { + "epoch": 0.907165503378517, + "grad_norm": 3.357485771179199, + "learning_rate": 4.224358459831601e-06, + "loss": 1.4989, + "step": 38699 + }, + { + "epoch": 0.9071889449533219, + "grad_norm": 3.7291457653045654, + "learning_rate": 4.222240354980012e-06, + "loss": 0.329, + "step": 38700 + }, + { + "epoch": 0.907212386528127, + "grad_norm": 2.537541151046753, + "learning_rate": 4.220122769821233e-06, + "loss": 0.1592, + "step": 38701 + }, + { + "epoch": 0.9072358281029319, + "grad_norm": 4.487227916717529, + "learning_rate": 4.218005704366756e-06, + "loss": 0.4081, + "step": 38702 + }, + { + "epoch": 0.907259269677737, + "grad_norm": 2.422947406768799, + "learning_rate": 4.215889158628073e-06, + "loss": 1.316, + "step": 38703 + }, + { + "epoch": 0.9072827112525419, + "grad_norm": 0.8578968644142151, + "learning_rate": 4.213773132616683e-06, + "loss": 0.0444, + "step": 38704 + }, + { + "epoch": 0.907306152827347, + "grad_norm": 0.3989279568195343, + "learning_rate": 4.211657626344045e-06, + "loss": 0.0238, + "step": 38705 + }, + { + "epoch": 0.9073295944021519, + "grad_norm": 1.5813109874725342, + "learning_rate": 4.209542639821651e-06, + "loss": 0.0986, + "step": 38706 + }, + { + "epoch": 0.9073530359769569, + "grad_norm": 2.94707989692688, + "learning_rate": 4.207428173060968e-06, + "loss": 1.1681, + "step": 38707 + }, + { + "epoch": 0.9073764775517619, + "grad_norm": 2.1408660411834717, + "learning_rate": 4.205314226073476e-06, + "loss": 0.1366, + "step": 38708 + }, + { + "epoch": 0.9073999191265669, + "grad_norm": 5.638894557952881, + "learning_rate": 4.2032007988706325e-06, + "loss": 0.2667, + "step": 38709 + }, + { + "epoch": 0.907423360701372, + "grad_norm": 1.2958377599716187, + "learning_rate": 4.201087891463928e-06, + "loss": 0.1815, + "step": 38710 + }, + { + "epoch": 0.9074468022761769, + "grad_norm": 1.409195899963379, + "learning_rate": 4.19897550386481e-06, + "loss": 0.8514, + "step": 38711 + }, + { + "epoch": 0.907470243850982, + "grad_norm": 4.548293113708496, + "learning_rate": 4.196863636084747e-06, + "loss": 0.2348, + "step": 38712 + }, + { + "epoch": 0.9074936854257869, + "grad_norm": 2.1264617443084717, + "learning_rate": 4.1947522881351956e-06, + "loss": 0.1204, + "step": 38713 + }, + { + "epoch": 0.907517127000592, + "grad_norm": 1.7798324823379517, + "learning_rate": 4.192641460027602e-06, + "loss": 0.1469, + "step": 38714 + }, + { + "epoch": 0.9075405685753969, + "grad_norm": 0.5798603296279907, + "learning_rate": 4.190531151773436e-06, + "loss": 0.069, + "step": 38715 + }, + { + "epoch": 0.9075640101502019, + "grad_norm": 1.1281721591949463, + "learning_rate": 4.188421363384132e-06, + "loss": 0.0617, + "step": 38716 + }, + { + "epoch": 0.9075874517250069, + "grad_norm": 1.9537897109985352, + "learning_rate": 4.186312094871159e-06, + "loss": 0.1309, + "step": 38717 + }, + { + "epoch": 0.9076108932998119, + "grad_norm": 2.3327643871307373, + "learning_rate": 4.184203346245952e-06, + "loss": 1.7042, + "step": 38718 + }, + { + "epoch": 0.9076343348746169, + "grad_norm": 0.6359997987747192, + "learning_rate": 4.182095117519957e-06, + "loss": 0.0702, + "step": 38719 + }, + { + "epoch": 0.9076577764494219, + "grad_norm": 0.6349444389343262, + "learning_rate": 4.1799874087046e-06, + "loss": 0.0535, + "step": 38720 + }, + { + "epoch": 0.9076812180242269, + "grad_norm": 3.344089984893799, + "learning_rate": 4.177880219811314e-06, + "loss": 0.2634, + "step": 38721 + }, + { + "epoch": 0.9077046595990319, + "grad_norm": 2.2516114711761475, + "learning_rate": 4.1757735508515585e-06, + "loss": 0.0843, + "step": 38722 + }, + { + "epoch": 0.9077281011738368, + "grad_norm": 1.7233235836029053, + "learning_rate": 4.173667401836756e-06, + "loss": 0.1856, + "step": 38723 + }, + { + "epoch": 0.9077515427486419, + "grad_norm": 3.269439697265625, + "learning_rate": 4.171561772778321e-06, + "loss": 0.2975, + "step": 38724 + }, + { + "epoch": 0.9077749843234468, + "grad_norm": 0.7573312520980835, + "learning_rate": 4.169456663687688e-06, + "loss": 0.0661, + "step": 38725 + }, + { + "epoch": 0.9077984258982519, + "grad_norm": 1.9799875020980835, + "learning_rate": 4.16735207457628e-06, + "loss": 0.1247, + "step": 38726 + }, + { + "epoch": 0.9078218674730568, + "grad_norm": 1.151182770729065, + "learning_rate": 4.165248005455502e-06, + "loss": 0.0592, + "step": 38727 + }, + { + "epoch": 0.9078453090478619, + "grad_norm": 0.46840283274650574, + "learning_rate": 4.163144456336776e-06, + "loss": 0.0345, + "step": 38728 + }, + { + "epoch": 0.9078687506226668, + "grad_norm": 1.3283641338348389, + "learning_rate": 4.161041427231538e-06, + "loss": 0.0355, + "step": 38729 + }, + { + "epoch": 0.9078921921974719, + "grad_norm": 2.8143808841705322, + "learning_rate": 4.1589389181511895e-06, + "loss": 0.1505, + "step": 38730 + }, + { + "epoch": 0.9079156337722768, + "grad_norm": 1.6455328464508057, + "learning_rate": 4.156836929107133e-06, + "loss": 0.7927, + "step": 38731 + }, + { + "epoch": 0.9079390753470818, + "grad_norm": 1.9656299352645874, + "learning_rate": 4.15473546011077e-06, + "loss": 0.1416, + "step": 38732 + }, + { + "epoch": 0.9079625169218868, + "grad_norm": 7.313246250152588, + "learning_rate": 4.152634511173503e-06, + "loss": 0.3052, + "step": 38733 + }, + { + "epoch": 0.9079859584966918, + "grad_norm": 3.2344398498535156, + "learning_rate": 4.1505340823067344e-06, + "loss": 0.2703, + "step": 38734 + }, + { + "epoch": 0.9080094000714968, + "grad_norm": 4.889636039733887, + "learning_rate": 4.148434173521865e-06, + "loss": 0.3711, + "step": 38735 + }, + { + "epoch": 0.9080328416463018, + "grad_norm": 1.5250552892684937, + "learning_rate": 4.146334784830297e-06, + "loss": 0.0944, + "step": 38736 + }, + { + "epoch": 0.9080562832211068, + "grad_norm": 4.199056625366211, + "learning_rate": 4.1442359162433995e-06, + "loss": 0.3219, + "step": 38737 + }, + { + "epoch": 0.9080797247959118, + "grad_norm": 1.6657260656356812, + "learning_rate": 4.142137567772586e-06, + "loss": 0.1582, + "step": 38738 + }, + { + "epoch": 0.9081031663707168, + "grad_norm": 5.093998432159424, + "learning_rate": 4.140039739429202e-06, + "loss": 0.3092, + "step": 38739 + }, + { + "epoch": 0.9081266079455218, + "grad_norm": 0.467349648475647, + "learning_rate": 4.137942431224684e-06, + "loss": 0.0272, + "step": 38740 + }, + { + "epoch": 0.9081500495203267, + "grad_norm": 3.7148847579956055, + "learning_rate": 4.135845643170366e-06, + "loss": 0.2401, + "step": 38741 + }, + { + "epoch": 0.9081734910951318, + "grad_norm": 0.6252340078353882, + "learning_rate": 4.1337493752776625e-06, + "loss": 0.0405, + "step": 38742 + }, + { + "epoch": 0.9081969326699367, + "grad_norm": 0.5828227400779724, + "learning_rate": 4.13165362755793e-06, + "loss": 0.0265, + "step": 38743 + }, + { + "epoch": 0.9082203742447418, + "grad_norm": 2.3350982666015625, + "learning_rate": 4.129558400022537e-06, + "loss": 1.205, + "step": 38744 + }, + { + "epoch": 0.9082438158195467, + "grad_norm": 1.8520747423171997, + "learning_rate": 4.127463692682853e-06, + "loss": 0.0506, + "step": 38745 + }, + { + "epoch": 0.9082672573943518, + "grad_norm": 3.436520576477051, + "learning_rate": 4.125369505550258e-06, + "loss": 1.7108, + "step": 38746 + }, + { + "epoch": 0.9082906989691567, + "grad_norm": 1.6682125329971313, + "learning_rate": 4.123275838636076e-06, + "loss": 0.0654, + "step": 38747 + }, + { + "epoch": 0.9083141405439618, + "grad_norm": 2.07327938079834, + "learning_rate": 4.121182691951719e-06, + "loss": 0.1599, + "step": 38748 + }, + { + "epoch": 0.9083375821187667, + "grad_norm": 0.43029677867889404, + "learning_rate": 4.119090065508513e-06, + "loss": 0.0147, + "step": 38749 + }, + { + "epoch": 0.9083610236935717, + "grad_norm": 2.535942316055298, + "learning_rate": 4.116997959317825e-06, + "loss": 1.5558, + "step": 38750 + }, + { + "epoch": 0.9083844652683767, + "grad_norm": 2.3882131576538086, + "learning_rate": 4.1149063733909924e-06, + "loss": 0.4153, + "step": 38751 + }, + { + "epoch": 0.9084079068431817, + "grad_norm": 1.1170172691345215, + "learning_rate": 4.112815307739381e-06, + "loss": 0.1249, + "step": 38752 + }, + { + "epoch": 0.9084313484179867, + "grad_norm": 2.441235303878784, + "learning_rate": 4.110724762374329e-06, + "loss": 0.2275, + "step": 38753 + }, + { + "epoch": 0.9084547899927917, + "grad_norm": 2.3276631832122803, + "learning_rate": 4.108634737307171e-06, + "loss": 0.0928, + "step": 38754 + }, + { + "epoch": 0.9084782315675967, + "grad_norm": 2.4919850826263428, + "learning_rate": 4.106545232549275e-06, + "loss": 0.1093, + "step": 38755 + }, + { + "epoch": 0.9085016731424017, + "grad_norm": 1.7844685316085815, + "learning_rate": 4.104456248111954e-06, + "loss": 0.1224, + "step": 38756 + }, + { + "epoch": 0.9085251147172066, + "grad_norm": 1.7688615322113037, + "learning_rate": 4.102367784006555e-06, + "loss": 0.0722, + "step": 38757 + }, + { + "epoch": 0.9085485562920117, + "grad_norm": 4.348111152648926, + "learning_rate": 4.100279840244403e-06, + "loss": 0.2768, + "step": 38758 + }, + { + "epoch": 0.9085719978668166, + "grad_norm": 2.3053293228149414, + "learning_rate": 4.0981924168368105e-06, + "loss": 0.1173, + "step": 38759 + }, + { + "epoch": 0.9085954394416217, + "grad_norm": 1.7613741159439087, + "learning_rate": 4.096105513795134e-06, + "loss": 0.087, + "step": 38760 + }, + { + "epoch": 0.9086188810164267, + "grad_norm": 1.39346182346344, + "learning_rate": 4.0940191311307e-06, + "loss": 0.075, + "step": 38761 + }, + { + "epoch": 0.9086423225912317, + "grad_norm": 0.9372336864471436, + "learning_rate": 4.091933268854797e-06, + "loss": 0.0433, + "step": 38762 + }, + { + "epoch": 0.9086657641660367, + "grad_norm": 3.1366870403289795, + "learning_rate": 4.0898479269787625e-06, + "loss": 0.1712, + "step": 38763 + }, + { + "epoch": 0.9086892057408417, + "grad_norm": 0.5717966556549072, + "learning_rate": 4.08776310551392e-06, + "loss": 0.0549, + "step": 38764 + }, + { + "epoch": 0.9087126473156467, + "grad_norm": 2.2771189212799072, + "learning_rate": 4.08567880447156e-06, + "loss": 0.1625, + "step": 38765 + }, + { + "epoch": 0.9087360888904517, + "grad_norm": 3.875218391418457, + "learning_rate": 4.0835950238630076e-06, + "loss": 0.3766, + "step": 38766 + }, + { + "epoch": 0.9087595304652567, + "grad_norm": 1.6653294563293457, + "learning_rate": 4.081511763699564e-06, + "loss": 0.0926, + "step": 38767 + }, + { + "epoch": 0.9087829720400616, + "grad_norm": 3.107923746109009, + "learning_rate": 4.079429023992542e-06, + "loss": 0.3292, + "step": 38768 + }, + { + "epoch": 0.9088064136148667, + "grad_norm": 4.144940376281738, + "learning_rate": 4.077346804753235e-06, + "loss": 0.3756, + "step": 38769 + }, + { + "epoch": 0.9088298551896716, + "grad_norm": 1.3374580144882202, + "learning_rate": 4.0752651059929425e-06, + "loss": 0.0748, + "step": 38770 + }, + { + "epoch": 0.9088532967644767, + "grad_norm": 2.066340208053589, + "learning_rate": 4.073183927722956e-06, + "loss": 0.1204, + "step": 38771 + }, + { + "epoch": 0.9088767383392816, + "grad_norm": 0.29391831159591675, + "learning_rate": 4.071103269954557e-06, + "loss": 0.0268, + "step": 38772 + }, + { + "epoch": 0.9089001799140867, + "grad_norm": 2.7321317195892334, + "learning_rate": 4.069023132699057e-06, + "loss": 0.2315, + "step": 38773 + }, + { + "epoch": 0.9089236214888916, + "grad_norm": 3.129185199737549, + "learning_rate": 4.066943515967747e-06, + "loss": 0.1588, + "step": 38774 + }, + { + "epoch": 0.9089470630636967, + "grad_norm": 2.539797306060791, + "learning_rate": 4.0648644197718745e-06, + "loss": 0.2613, + "step": 38775 + }, + { + "epoch": 0.9089705046385016, + "grad_norm": 3.3507754802703857, + "learning_rate": 4.062785844122763e-06, + "loss": 0.3116, + "step": 38776 + }, + { + "epoch": 0.9089939462133066, + "grad_norm": 1.688094973564148, + "learning_rate": 4.06070778903167e-06, + "loss": 1.3098, + "step": 38777 + }, + { + "epoch": 0.9090173877881116, + "grad_norm": 4.019057750701904, + "learning_rate": 4.058630254509888e-06, + "loss": 0.2332, + "step": 38778 + }, + { + "epoch": 0.9090408293629166, + "grad_norm": 3.7175097465515137, + "learning_rate": 4.0565532405686505e-06, + "loss": 0.181, + "step": 38779 + }, + { + "epoch": 0.9090642709377216, + "grad_norm": 2.429786443710327, + "learning_rate": 4.05447674721926e-06, + "loss": 0.2307, + "step": 38780 + }, + { + "epoch": 0.9090877125125266, + "grad_norm": 4.307648181915283, + "learning_rate": 4.052400774472987e-06, + "loss": 0.3818, + "step": 38781 + }, + { + "epoch": 0.9091111540873316, + "grad_norm": 1.8657516241073608, + "learning_rate": 4.050325322341087e-06, + "loss": 1.0497, + "step": 38782 + }, + { + "epoch": 0.9091345956621366, + "grad_norm": 3.159730911254883, + "learning_rate": 4.0482503908348065e-06, + "loss": 0.2406, + "step": 38783 + }, + { + "epoch": 0.9091580372369416, + "grad_norm": 2.937208890914917, + "learning_rate": 4.0461759799654275e-06, + "loss": 0.1691, + "step": 38784 + }, + { + "epoch": 0.9091814788117466, + "grad_norm": 3.8886783123016357, + "learning_rate": 4.044102089744184e-06, + "loss": 0.3215, + "step": 38785 + }, + { + "epoch": 0.9092049203865515, + "grad_norm": 3.4448306560516357, + "learning_rate": 4.042028720182356e-06, + "loss": 0.1847, + "step": 38786 + }, + { + "epoch": 0.9092283619613566, + "grad_norm": 2.5104966163635254, + "learning_rate": 4.0399558712911565e-06, + "loss": 0.2065, + "step": 38787 + }, + { + "epoch": 0.9092518035361615, + "grad_norm": 4.924994945526123, + "learning_rate": 4.037883543081867e-06, + "loss": 0.2112, + "step": 38788 + }, + { + "epoch": 0.9092752451109666, + "grad_norm": 2.9797136783599854, + "learning_rate": 4.03581173556572e-06, + "loss": 0.275, + "step": 38789 + }, + { + "epoch": 0.9092986866857715, + "grad_norm": 1.9949429035186768, + "learning_rate": 4.0337404487539646e-06, + "loss": 0.1647, + "step": 38790 + }, + { + "epoch": 0.9093221282605766, + "grad_norm": 1.2313576936721802, + "learning_rate": 4.031669682657812e-06, + "loss": 0.0433, + "step": 38791 + }, + { + "epoch": 0.9093455698353815, + "grad_norm": 2.056786060333252, + "learning_rate": 4.029599437288534e-06, + "loss": 0.1428, + "step": 38792 + }, + { + "epoch": 0.9093690114101866, + "grad_norm": 0.7498824000358582, + "learning_rate": 4.027529712657352e-06, + "loss": 0.0544, + "step": 38793 + }, + { + "epoch": 0.9093924529849915, + "grad_norm": 3.1305606365203857, + "learning_rate": 4.025460508775492e-06, + "loss": 0.1621, + "step": 38794 + }, + { + "epoch": 0.9094158945597965, + "grad_norm": 2.0884366035461426, + "learning_rate": 4.023391825654177e-06, + "loss": 0.0955, + "step": 38795 + }, + { + "epoch": 0.9094393361346015, + "grad_norm": 1.002248764038086, + "learning_rate": 4.021323663304633e-06, + "loss": 0.0418, + "step": 38796 + }, + { + "epoch": 0.9094627777094065, + "grad_norm": 1.4828011989593506, + "learning_rate": 4.019256021738072e-06, + "loss": 0.1466, + "step": 38797 + }, + { + "epoch": 0.9094862192842115, + "grad_norm": 2.625284194946289, + "learning_rate": 4.017188900965752e-06, + "loss": 0.1164, + "step": 38798 + }, + { + "epoch": 0.9095096608590165, + "grad_norm": 2.9944751262664795, + "learning_rate": 4.0151223009988415e-06, + "loss": 0.2166, + "step": 38799 + }, + { + "epoch": 0.9095331024338215, + "grad_norm": 0.9689773917198181, + "learning_rate": 4.013056221848588e-06, + "loss": 0.0874, + "step": 38800 + }, + { + "epoch": 0.9095565440086265, + "grad_norm": 2.0227763652801514, + "learning_rate": 4.010990663526193e-06, + "loss": 0.1381, + "step": 38801 + }, + { + "epoch": 0.9095799855834314, + "grad_norm": 2.6186506748199463, + "learning_rate": 4.0089256260428584e-06, + "loss": 0.2402, + "step": 38802 + }, + { + "epoch": 0.9096034271582365, + "grad_norm": 1.5692256689071655, + "learning_rate": 4.006861109409799e-06, + "loss": 0.1071, + "step": 38803 + }, + { + "epoch": 0.9096268687330414, + "grad_norm": 1.4881831407546997, + "learning_rate": 4.004797113638192e-06, + "loss": 0.0907, + "step": 38804 + }, + { + "epoch": 0.9096503103078465, + "grad_norm": 3.0252413749694824, + "learning_rate": 4.002733638739275e-06, + "loss": 0.1659, + "step": 38805 + }, + { + "epoch": 0.9096737518826514, + "grad_norm": 2.8838558197021484, + "learning_rate": 4.000670684724217e-06, + "loss": 0.2884, + "step": 38806 + }, + { + "epoch": 0.9096971934574565, + "grad_norm": 1.6269254684448242, + "learning_rate": 3.998608251604219e-06, + "loss": 0.1145, + "step": 38807 + }, + { + "epoch": 0.9097206350322614, + "grad_norm": 3.171717643737793, + "learning_rate": 3.996546339390472e-06, + "loss": 0.297, + "step": 38808 + }, + { + "epoch": 0.9097440766070665, + "grad_norm": 2.1453428268432617, + "learning_rate": 3.9944849480941685e-06, + "loss": 1.4668, + "step": 38809 + }, + { + "epoch": 0.9097675181818714, + "grad_norm": 2.9974918365478516, + "learning_rate": 3.992424077726476e-06, + "loss": 0.2866, + "step": 38810 + }, + { + "epoch": 0.9097909597566765, + "grad_norm": 0.3330913782119751, + "learning_rate": 3.9903637282986076e-06, + "loss": 0.0228, + "step": 38811 + }, + { + "epoch": 0.9098144013314815, + "grad_norm": 2.020658254623413, + "learning_rate": 3.988303899821699e-06, + "loss": 0.1184, + "step": 38812 + }, + { + "epoch": 0.9098378429062864, + "grad_norm": 1.4830834865570068, + "learning_rate": 3.986244592306976e-06, + "loss": 0.2048, + "step": 38813 + }, + { + "epoch": 0.9098612844810915, + "grad_norm": 5.706773281097412, + "learning_rate": 3.984185805765583e-06, + "loss": 0.4275, + "step": 38814 + }, + { + "epoch": 0.9098847260558964, + "grad_norm": 3.4121596813201904, + "learning_rate": 3.982127540208702e-06, + "loss": 0.2088, + "step": 38815 + }, + { + "epoch": 0.9099081676307015, + "grad_norm": 2.0799784660339355, + "learning_rate": 3.980069795647501e-06, + "loss": 0.0894, + "step": 38816 + }, + { + "epoch": 0.9099316092055064, + "grad_norm": 3.594376802444458, + "learning_rate": 3.978012572093115e-06, + "loss": 0.4264, + "step": 38817 + }, + { + "epoch": 0.9099550507803115, + "grad_norm": 2.2149922847747803, + "learning_rate": 3.975955869556758e-06, + "loss": 0.1764, + "step": 38818 + }, + { + "epoch": 0.9099784923551164, + "grad_norm": 1.1880398988723755, + "learning_rate": 3.973899688049554e-06, + "loss": 0.1432, + "step": 38819 + }, + { + "epoch": 0.9100019339299215, + "grad_norm": 0.9959955215454102, + "learning_rate": 3.971844027582683e-06, + "loss": 0.1616, + "step": 38820 + }, + { + "epoch": 0.9100253755047264, + "grad_norm": 0.8672559261322021, + "learning_rate": 3.9697888881672806e-06, + "loss": 0.0695, + "step": 38821 + }, + { + "epoch": 0.9100488170795314, + "grad_norm": 4.461551666259766, + "learning_rate": 3.967734269814494e-06, + "loss": 0.4682, + "step": 38822 + }, + { + "epoch": 0.9100722586543364, + "grad_norm": 3.6225569248199463, + "learning_rate": 3.9656801725354795e-06, + "loss": 0.3922, + "step": 38823 + }, + { + "epoch": 0.9100957002291414, + "grad_norm": 1.555381178855896, + "learning_rate": 3.963626596341386e-06, + "loss": 1.3612, + "step": 38824 + }, + { + "epoch": 0.9101191418039464, + "grad_norm": 1.70304274559021, + "learning_rate": 3.961573541243369e-06, + "loss": 0.965, + "step": 38825 + }, + { + "epoch": 0.9101425833787514, + "grad_norm": 3.9101016521453857, + "learning_rate": 3.9595210072525425e-06, + "loss": 0.4081, + "step": 38826 + }, + { + "epoch": 0.9101660249535564, + "grad_norm": 0.8289559483528137, + "learning_rate": 3.957468994380064e-06, + "loss": 0.0497, + "step": 38827 + }, + { + "epoch": 0.9101894665283614, + "grad_norm": 1.0957874059677124, + "learning_rate": 3.95541750263706e-06, + "loss": 0.0921, + "step": 38828 + }, + { + "epoch": 0.9102129081031664, + "grad_norm": 2.231015682220459, + "learning_rate": 3.953366532034641e-06, + "loss": 0.1319, + "step": 38829 + }, + { + "epoch": 0.9102363496779714, + "grad_norm": 2.4654784202575684, + "learning_rate": 3.951316082583978e-06, + "loss": 0.3151, + "step": 38830 + }, + { + "epoch": 0.9102597912527763, + "grad_norm": 1.013578176498413, + "learning_rate": 3.949266154296172e-06, + "loss": 0.0646, + "step": 38831 + }, + { + "epoch": 0.9102832328275814, + "grad_norm": 2.949568748474121, + "learning_rate": 3.9472167471823475e-06, + "loss": 0.2714, + "step": 38832 + }, + { + "epoch": 0.9103066744023863, + "grad_norm": 2.9024617671966553, + "learning_rate": 3.945167861253618e-06, + "loss": 0.2969, + "step": 38833 + }, + { + "epoch": 0.9103301159771914, + "grad_norm": 1.4183019399642944, + "learning_rate": 3.943119496521119e-06, + "loss": 0.0716, + "step": 38834 + }, + { + "epoch": 0.9103535575519963, + "grad_norm": 1.33344304561615, + "learning_rate": 3.9410716529959425e-06, + "loss": 0.0956, + "step": 38835 + }, + { + "epoch": 0.9103769991268014, + "grad_norm": 2.7418723106384277, + "learning_rate": 3.939024330689201e-06, + "loss": 0.2967, + "step": 38836 + }, + { + "epoch": 0.9104004407016063, + "grad_norm": 1.820829153060913, + "learning_rate": 3.9369775296120405e-06, + "loss": 0.9402, + "step": 38837 + }, + { + "epoch": 0.9104238822764114, + "grad_norm": 2.1865885257720947, + "learning_rate": 3.934931249775531e-06, + "loss": 1.5917, + "step": 38838 + }, + { + "epoch": 0.9104473238512163, + "grad_norm": 3.4165267944335938, + "learning_rate": 3.932885491190785e-06, + "loss": 0.3693, + "step": 38839 + }, + { + "epoch": 0.9104707654260213, + "grad_norm": 3.5892674922943115, + "learning_rate": 3.930840253868906e-06, + "loss": 0.2029, + "step": 38840 + }, + { + "epoch": 0.9104942070008263, + "grad_norm": 3.0196194648742676, + "learning_rate": 3.928795537820995e-06, + "loss": 0.1732, + "step": 38841 + }, + { + "epoch": 0.9105176485756313, + "grad_norm": 1.725048542022705, + "learning_rate": 3.926751343058121e-06, + "loss": 0.0914, + "step": 38842 + }, + { + "epoch": 0.9105410901504363, + "grad_norm": 1.0327293872833252, + "learning_rate": 3.92470766959141e-06, + "loss": 0.0428, + "step": 38843 + }, + { + "epoch": 0.9105645317252413, + "grad_norm": 1.2757198810577393, + "learning_rate": 3.92266451743194e-06, + "loss": 0.0844, + "step": 38844 + }, + { + "epoch": 0.9105879733000463, + "grad_norm": 3.0393729209899902, + "learning_rate": 3.920621886590792e-06, + "loss": 0.1537, + "step": 38845 + }, + { + "epoch": 0.9106114148748513, + "grad_norm": 0.7294538021087646, + "learning_rate": 3.918579777079046e-06, + "loss": 0.0582, + "step": 38846 + }, + { + "epoch": 0.9106348564496562, + "grad_norm": 0.5193808674812317, + "learning_rate": 3.916538188907792e-06, + "loss": 0.0629, + "step": 38847 + }, + { + "epoch": 0.9106582980244613, + "grad_norm": 1.722123622894287, + "learning_rate": 3.914497122088101e-06, + "loss": 0.1059, + "step": 38848 + }, + { + "epoch": 0.9106817395992662, + "grad_norm": 5.553028106689453, + "learning_rate": 3.91245657663103e-06, + "loss": 0.2408, + "step": 38849 + }, + { + "epoch": 0.9107051811740713, + "grad_norm": 1.7745801210403442, + "learning_rate": 3.910416552547702e-06, + "loss": 0.1569, + "step": 38850 + }, + { + "epoch": 0.9107286227488762, + "grad_norm": 4.485363006591797, + "learning_rate": 3.9083770498491435e-06, + "loss": 0.0952, + "step": 38851 + }, + { + "epoch": 0.9107520643236813, + "grad_norm": 2.700664520263672, + "learning_rate": 3.9063380685464335e-06, + "loss": 0.1604, + "step": 38852 + }, + { + "epoch": 0.9107755058984862, + "grad_norm": 3.225761651992798, + "learning_rate": 3.904299608650641e-06, + "loss": 0.1698, + "step": 38853 + }, + { + "epoch": 0.9107989474732913, + "grad_norm": 0.9118355512619019, + "learning_rate": 3.902261670172824e-06, + "loss": 0.0974, + "step": 38854 + }, + { + "epoch": 0.9108223890480962, + "grad_norm": 4.8982954025268555, + "learning_rate": 3.9002242531240185e-06, + "loss": 0.4925, + "step": 38855 + }, + { + "epoch": 0.9108458306229013, + "grad_norm": 2.318096399307251, + "learning_rate": 3.898187357515315e-06, + "loss": 0.2834, + "step": 38856 + }, + { + "epoch": 0.9108692721977062, + "grad_norm": 1.4641553163528442, + "learning_rate": 3.896150983357738e-06, + "loss": 0.1439, + "step": 38857 + }, + { + "epoch": 0.9108927137725112, + "grad_norm": 4.068728923797607, + "learning_rate": 3.894115130662357e-06, + "loss": 0.2108, + "step": 38858 + }, + { + "epoch": 0.9109161553473162, + "grad_norm": 2.6949806213378906, + "learning_rate": 3.892079799440207e-06, + "loss": 0.2516, + "step": 38859 + }, + { + "epoch": 0.9109395969221212, + "grad_norm": 2.373330593109131, + "learning_rate": 3.890044989702324e-06, + "loss": 0.146, + "step": 38860 + }, + { + "epoch": 0.9109630384969262, + "grad_norm": 3.3914217948913574, + "learning_rate": 3.888010701459777e-06, + "loss": 0.2229, + "step": 38861 + }, + { + "epoch": 0.9109864800717312, + "grad_norm": 2.8487753868103027, + "learning_rate": 3.885976934723568e-06, + "loss": 0.2403, + "step": 38862 + }, + { + "epoch": 0.9110099216465363, + "grad_norm": 3.09271502494812, + "learning_rate": 3.883943689504754e-06, + "loss": 0.2908, + "step": 38863 + }, + { + "epoch": 0.9110333632213412, + "grad_norm": 0.8090235590934753, + "learning_rate": 3.881910965814373e-06, + "loss": 0.051, + "step": 38864 + }, + { + "epoch": 0.9110568047961463, + "grad_norm": 3.174806833267212, + "learning_rate": 3.879878763663447e-06, + "loss": 0.9352, + "step": 38865 + }, + { + "epoch": 0.9110802463709512, + "grad_norm": 2.3212461471557617, + "learning_rate": 3.877847083063002e-06, + "loss": 1.3934, + "step": 38866 + }, + { + "epoch": 0.9111036879457562, + "grad_norm": 4.351711273193359, + "learning_rate": 3.875815924024051e-06, + "loss": 0.4516, + "step": 38867 + }, + { + "epoch": 0.9111271295205612, + "grad_norm": 0.2828041613101959, + "learning_rate": 3.87378528655763e-06, + "loss": 0.0209, + "step": 38868 + }, + { + "epoch": 0.9111505710953662, + "grad_norm": 2.0324454307556152, + "learning_rate": 3.871755170674751e-06, + "loss": 0.2861, + "step": 38869 + }, + { + "epoch": 0.9111740126701712, + "grad_norm": 0.38481611013412476, + "learning_rate": 3.86972557638644e-06, + "loss": 0.0123, + "step": 38870 + }, + { + "epoch": 0.9111974542449762, + "grad_norm": 2.166719675064087, + "learning_rate": 3.8676965037037e-06, + "loss": 1.6621, + "step": 38871 + }, + { + "epoch": 0.9112208958197812, + "grad_norm": 1.5578776597976685, + "learning_rate": 3.86566795263753e-06, + "loss": 0.0904, + "step": 38872 + }, + { + "epoch": 0.9112443373945862, + "grad_norm": 2.3968849182128906, + "learning_rate": 3.8636399231989585e-06, + "loss": 0.1625, + "step": 38873 + }, + { + "epoch": 0.9112677789693912, + "grad_norm": 1.444420576095581, + "learning_rate": 3.861612415398963e-06, + "loss": 0.7946, + "step": 38874 + }, + { + "epoch": 0.9112912205441962, + "grad_norm": 2.0857889652252197, + "learning_rate": 3.85958542924858e-06, + "loss": 0.1604, + "step": 38875 + }, + { + "epoch": 0.9113146621190011, + "grad_norm": 3.2951693534851074, + "learning_rate": 3.85755896475879e-06, + "loss": 0.2641, + "step": 38876 + }, + { + "epoch": 0.9113381036938062, + "grad_norm": 1.9458608627319336, + "learning_rate": 3.855533021940583e-06, + "loss": 0.1683, + "step": 38877 + }, + { + "epoch": 0.9113615452686111, + "grad_norm": 2.044469118118286, + "learning_rate": 3.853507600804962e-06, + "loss": 1.6897, + "step": 38878 + }, + { + "epoch": 0.9113849868434162, + "grad_norm": 4.068728923797607, + "learning_rate": 3.851482701362918e-06, + "loss": 0.4077, + "step": 38879 + }, + { + "epoch": 0.9114084284182211, + "grad_norm": 0.7166047096252441, + "learning_rate": 3.84945832362541e-06, + "loss": 0.0767, + "step": 38880 + }, + { + "epoch": 0.9114318699930262, + "grad_norm": 2.36188006401062, + "learning_rate": 3.847434467603461e-06, + "loss": 0.1514, + "step": 38881 + }, + { + "epoch": 0.9114553115678311, + "grad_norm": 1.8748846054077148, + "learning_rate": 3.84541113330803e-06, + "loss": 0.1197, + "step": 38882 + }, + { + "epoch": 0.9114787531426362, + "grad_norm": 2.146759033203125, + "learning_rate": 3.843388320750107e-06, + "loss": 0.1703, + "step": 38883 + }, + { + "epoch": 0.9115021947174411, + "grad_norm": 0.7725376486778259, + "learning_rate": 3.841366029940663e-06, + "loss": 0.0563, + "step": 38884 + }, + { + "epoch": 0.9115256362922461, + "grad_norm": 0.9805461764335632, + "learning_rate": 3.839344260890665e-06, + "loss": 0.0408, + "step": 38885 + }, + { + "epoch": 0.9115490778670511, + "grad_norm": 2.112285614013672, + "learning_rate": 3.837323013611094e-06, + "loss": 0.19, + "step": 38886 + }, + { + "epoch": 0.9115725194418561, + "grad_norm": 1.999504566192627, + "learning_rate": 3.835302288112896e-06, + "loss": 0.1417, + "step": 38887 + }, + { + "epoch": 0.9115959610166611, + "grad_norm": 1.9324296712875366, + "learning_rate": 3.833282084407064e-06, + "loss": 1.3955, + "step": 38888 + }, + { + "epoch": 0.9116194025914661, + "grad_norm": 1.7541435956954956, + "learning_rate": 3.831262402504554e-06, + "loss": 0.2285, + "step": 38889 + }, + { + "epoch": 0.9116428441662711, + "grad_norm": 2.78092360496521, + "learning_rate": 3.829243242416314e-06, + "loss": 0.2237, + "step": 38890 + }, + { + "epoch": 0.9116662857410761, + "grad_norm": 2.3696839809417725, + "learning_rate": 3.827224604153312e-06, + "loss": 0.1747, + "step": 38891 + }, + { + "epoch": 0.911689727315881, + "grad_norm": 2.503993034362793, + "learning_rate": 3.825206487726485e-06, + "loss": 0.1133, + "step": 38892 + }, + { + "epoch": 0.9117131688906861, + "grad_norm": 3.2002363204956055, + "learning_rate": 3.8231888931467785e-06, + "loss": 0.1381, + "step": 38893 + }, + { + "epoch": 0.911736610465491, + "grad_norm": 0.681108832359314, + "learning_rate": 3.821171820425173e-06, + "loss": 0.0587, + "step": 38894 + }, + { + "epoch": 0.9117600520402961, + "grad_norm": 1.9478931427001953, + "learning_rate": 3.819155269572594e-06, + "loss": 0.1094, + "step": 38895 + }, + { + "epoch": 0.911783493615101, + "grad_norm": 1.4846669435501099, + "learning_rate": 3.817139240599976e-06, + "loss": 0.1288, + "step": 38896 + }, + { + "epoch": 0.9118069351899061, + "grad_norm": 2.942274332046509, + "learning_rate": 3.815123733518255e-06, + "loss": 0.2203, + "step": 38897 + }, + { + "epoch": 0.911830376764711, + "grad_norm": 0.7830906510353088, + "learning_rate": 3.81310874833839e-06, + "loss": 0.0481, + "step": 38898 + }, + { + "epoch": 0.9118538183395161, + "grad_norm": 2.1604490280151367, + "learning_rate": 3.811094285071293e-06, + "loss": 0.1038, + "step": 38899 + }, + { + "epoch": 0.911877259914321, + "grad_norm": 1.7949248552322388, + "learning_rate": 3.8090803437279e-06, + "loss": 0.0818, + "step": 38900 + }, + { + "epoch": 0.911900701489126, + "grad_norm": 0.22909799218177795, + "learning_rate": 3.807066924319158e-06, + "loss": 0.0103, + "step": 38901 + }, + { + "epoch": 0.911924143063931, + "grad_norm": 3.7690625190734863, + "learning_rate": 3.8050540268559697e-06, + "loss": 0.3023, + "step": 38902 + }, + { + "epoch": 0.911947584638736, + "grad_norm": 1.9277955293655396, + "learning_rate": 3.8030416513492595e-06, + "loss": 0.1513, + "step": 38903 + }, + { + "epoch": 0.911971026213541, + "grad_norm": 0.964534342288971, + "learning_rate": 3.8010297978099516e-06, + "loss": 0.0867, + "step": 38904 + }, + { + "epoch": 0.911994467788346, + "grad_norm": 1.506429672241211, + "learning_rate": 3.79901846624896e-06, + "loss": 0.0791, + "step": 38905 + }, + { + "epoch": 0.912017909363151, + "grad_norm": 1.5716297626495361, + "learning_rate": 3.7970076566771982e-06, + "loss": 0.1058, + "step": 38906 + }, + { + "epoch": 0.912041350937956, + "grad_norm": 3.0731616020202637, + "learning_rate": 3.7949973691055794e-06, + "loss": 0.2248, + "step": 38907 + }, + { + "epoch": 0.912064792512761, + "grad_norm": 0.4598565101623535, + "learning_rate": 3.7929876035450174e-06, + "loss": 0.0273, + "step": 38908 + }, + { + "epoch": 0.912088234087566, + "grad_norm": 4.1285600662231445, + "learning_rate": 3.7909783600063918e-06, + "loss": 0.2339, + "step": 38909 + }, + { + "epoch": 0.912111675662371, + "grad_norm": 0.7467500567436218, + "learning_rate": 3.788969638500639e-06, + "loss": 0.06, + "step": 38910 + }, + { + "epoch": 0.912135117237176, + "grad_norm": 2.635383367538452, + "learning_rate": 3.7869614390386386e-06, + "loss": 0.1854, + "step": 38911 + }, + { + "epoch": 0.9121585588119809, + "grad_norm": 4.575096607208252, + "learning_rate": 3.7849537616312713e-06, + "loss": 0.3655, + "step": 38912 + }, + { + "epoch": 0.912182000386786, + "grad_norm": 1.0864417552947998, + "learning_rate": 3.7829466062894723e-06, + "loss": 0.0988, + "step": 38913 + }, + { + "epoch": 0.912205441961591, + "grad_norm": 1.200173258781433, + "learning_rate": 3.7809399730241114e-06, + "loss": 0.1854, + "step": 38914 + }, + { + "epoch": 0.912228883536396, + "grad_norm": 2.360200881958008, + "learning_rate": 3.778933861846068e-06, + "loss": 0.2093, + "step": 38915 + }, + { + "epoch": 0.912252325111201, + "grad_norm": 2.5497515201568604, + "learning_rate": 3.776928272766245e-06, + "loss": 0.1305, + "step": 38916 + }, + { + "epoch": 0.912275766686006, + "grad_norm": 3.820966958999634, + "learning_rate": 3.7749232057955107e-06, + "loss": 0.1705, + "step": 38917 + }, + { + "epoch": 0.912299208260811, + "grad_norm": 4.617698669433594, + "learning_rate": 3.7729186609447355e-06, + "loss": 0.54, + "step": 38918 + }, + { + "epoch": 0.912322649835616, + "grad_norm": 2.791147470474243, + "learning_rate": 3.7709146382248205e-06, + "loss": 0.26, + "step": 38919 + }, + { + "epoch": 0.912346091410421, + "grad_norm": 4.065909385681152, + "learning_rate": 3.7689111376466246e-06, + "loss": 0.3678, + "step": 38920 + }, + { + "epoch": 0.9123695329852259, + "grad_norm": 1.7495561838150024, + "learning_rate": 3.7669081592210167e-06, + "loss": 0.174, + "step": 38921 + }, + { + "epoch": 0.912392974560031, + "grad_norm": 3.2308573722839355, + "learning_rate": 3.7649057029588763e-06, + "loss": 0.3246, + "step": 38922 + }, + { + "epoch": 0.9124164161348359, + "grad_norm": 0.9677290320396423, + "learning_rate": 3.762903768871073e-06, + "loss": 0.1245, + "step": 38923 + }, + { + "epoch": 0.912439857709641, + "grad_norm": 0.42663344740867615, + "learning_rate": 3.760902356968454e-06, + "loss": 0.0304, + "step": 38924 + }, + { + "epoch": 0.9124632992844459, + "grad_norm": 0.8699777722358704, + "learning_rate": 3.7589014672618772e-06, + "loss": 0.0751, + "step": 38925 + }, + { + "epoch": 0.912486740859251, + "grad_norm": 0.5998640060424805, + "learning_rate": 3.7569010997622113e-06, + "loss": 0.0596, + "step": 38926 + }, + { + "epoch": 0.9125101824340559, + "grad_norm": 3.5576553344726562, + "learning_rate": 3.754901254480314e-06, + "loss": 0.1464, + "step": 38927 + }, + { + "epoch": 0.912533624008861, + "grad_norm": 4.029393672943115, + "learning_rate": 3.752901931427022e-06, + "loss": 0.3147, + "step": 38928 + }, + { + "epoch": 0.9125570655836659, + "grad_norm": 2.430079698562622, + "learning_rate": 3.7509031306131926e-06, + "loss": 1.886, + "step": 38929 + }, + { + "epoch": 0.9125805071584709, + "grad_norm": 2.2919411659240723, + "learning_rate": 3.7489048520496726e-06, + "loss": 0.2067, + "step": 38930 + }, + { + "epoch": 0.9126039487332759, + "grad_norm": 1.3777481317520142, + "learning_rate": 3.7469070957472873e-06, + "loss": 0.117, + "step": 38931 + }, + { + "epoch": 0.9126273903080809, + "grad_norm": 2.7372872829437256, + "learning_rate": 3.7449098617169054e-06, + "loss": 0.1635, + "step": 38932 + }, + { + "epoch": 0.9126508318828859, + "grad_norm": 1.0927255153656006, + "learning_rate": 3.742913149969329e-06, + "loss": 0.0896, + "step": 38933 + }, + { + "epoch": 0.9126742734576909, + "grad_norm": 1.286799669265747, + "learning_rate": 3.740916960515428e-06, + "loss": 0.8105, + "step": 38934 + }, + { + "epoch": 0.9126977150324959, + "grad_norm": 3.734524965286255, + "learning_rate": 3.7389212933660266e-06, + "loss": 0.2417, + "step": 38935 + }, + { + "epoch": 0.9127211566073009, + "grad_norm": 1.362236738204956, + "learning_rate": 3.7369261485319275e-06, + "loss": 0.1302, + "step": 38936 + }, + { + "epoch": 0.9127445981821058, + "grad_norm": 0.4562554955482483, + "learning_rate": 3.7349315260239882e-06, + "loss": 0.0192, + "step": 38937 + }, + { + "epoch": 0.9127680397569109, + "grad_norm": 3.213355779647827, + "learning_rate": 3.7329374258530003e-06, + "loss": 0.1909, + "step": 38938 + }, + { + "epoch": 0.9127914813317158, + "grad_norm": 2.913252115249634, + "learning_rate": 3.730943848029811e-06, + "loss": 1.3699, + "step": 38939 + }, + { + "epoch": 0.9128149229065209, + "grad_norm": 3.5423407554626465, + "learning_rate": 3.7289507925652336e-06, + "loss": 0.3075, + "step": 38940 + }, + { + "epoch": 0.9128383644813258, + "grad_norm": 2.0445103645324707, + "learning_rate": 3.72695825947007e-06, + "loss": 0.1507, + "step": 38941 + }, + { + "epoch": 0.9128618060561309, + "grad_norm": 1.9415274858474731, + "learning_rate": 3.7249662487551353e-06, + "loss": 0.0897, + "step": 38942 + }, + { + "epoch": 0.9128852476309358, + "grad_norm": 4.585609436035156, + "learning_rate": 3.7229747604312415e-06, + "loss": 0.152, + "step": 38943 + }, + { + "epoch": 0.9129086892057409, + "grad_norm": 2.1342527866363525, + "learning_rate": 3.720983794509192e-06, + "loss": 0.1911, + "step": 38944 + }, + { + "epoch": 0.9129321307805458, + "grad_norm": 3.931771993637085, + "learning_rate": 3.7189933509997887e-06, + "loss": 0.4241, + "step": 38945 + }, + { + "epoch": 0.9129555723553509, + "grad_norm": 2.9989163875579834, + "learning_rate": 3.717003429913846e-06, + "loss": 0.1414, + "step": 38946 + }, + { + "epoch": 0.9129790139301558, + "grad_norm": 3.5074007511138916, + "learning_rate": 3.715014031262143e-06, + "loss": 0.1927, + "step": 38947 + }, + { + "epoch": 0.9130024555049608, + "grad_norm": 2.0729143619537354, + "learning_rate": 3.713025155055483e-06, + "loss": 1.0296, + "step": 38948 + }, + { + "epoch": 0.9130258970797658, + "grad_norm": 2.4418513774871826, + "learning_rate": 3.711036801304657e-06, + "loss": 1.1803, + "step": 38949 + }, + { + "epoch": 0.9130493386545708, + "grad_norm": 1.7973464727401733, + "learning_rate": 3.7090489700204456e-06, + "loss": 0.0786, + "step": 38950 + }, + { + "epoch": 0.9130727802293758, + "grad_norm": 2.0544373989105225, + "learning_rate": 3.707061661213651e-06, + "loss": 0.1083, + "step": 38951 + }, + { + "epoch": 0.9130962218041808, + "grad_norm": 2.647986650466919, + "learning_rate": 3.7050748748950424e-06, + "loss": 0.0662, + "step": 38952 + }, + { + "epoch": 0.9131196633789858, + "grad_norm": 3.2496440410614014, + "learning_rate": 3.7030886110754114e-06, + "loss": 0.1896, + "step": 38953 + }, + { + "epoch": 0.9131431049537908, + "grad_norm": 2.5288538932800293, + "learning_rate": 3.701102869765527e-06, + "loss": 1.8547, + "step": 38954 + }, + { + "epoch": 0.9131665465285957, + "grad_norm": 2.770615816116333, + "learning_rate": 3.699117650976169e-06, + "loss": 0.3496, + "step": 38955 + }, + { + "epoch": 0.9131899881034008, + "grad_norm": 3.6736648082733154, + "learning_rate": 3.6971329547180854e-06, + "loss": 0.3685, + "step": 38956 + }, + { + "epoch": 0.9132134296782057, + "grad_norm": 3.192553997039795, + "learning_rate": 3.695148781002078e-06, + "loss": 0.1944, + "step": 38957 + }, + { + "epoch": 0.9132368712530108, + "grad_norm": 3.4044344425201416, + "learning_rate": 3.693165129838905e-06, + "loss": 0.387, + "step": 38958 + }, + { + "epoch": 0.9132603128278157, + "grad_norm": 0.4842410087585449, + "learning_rate": 3.691182001239324e-06, + "loss": 0.047, + "step": 38959 + }, + { + "epoch": 0.9132837544026208, + "grad_norm": 1.0316470861434937, + "learning_rate": 3.6891993952141047e-06, + "loss": 0.0778, + "step": 38960 + }, + { + "epoch": 0.9133071959774257, + "grad_norm": 2.374398708343506, + "learning_rate": 3.687217311773994e-06, + "loss": 0.8853, + "step": 38961 + }, + { + "epoch": 0.9133306375522308, + "grad_norm": 0.6295188069343567, + "learning_rate": 3.6852357509297497e-06, + "loss": 0.0429, + "step": 38962 + }, + { + "epoch": 0.9133540791270357, + "grad_norm": 1.436113715171814, + "learning_rate": 3.683254712692108e-06, + "loss": 0.8087, + "step": 38963 + }, + { + "epoch": 0.9133775207018408, + "grad_norm": 2.740025758743286, + "learning_rate": 3.68127419707186e-06, + "loss": 1.4537, + "step": 38964 + }, + { + "epoch": 0.9134009622766458, + "grad_norm": 3.7559573650360107, + "learning_rate": 3.6792942040797196e-06, + "loss": 0.2651, + "step": 38965 + }, + { + "epoch": 0.9134244038514507, + "grad_norm": 3.28798508644104, + "learning_rate": 3.6773147337264336e-06, + "loss": 0.3041, + "step": 38966 + }, + { + "epoch": 0.9134478454262558, + "grad_norm": 2.681335926055908, + "learning_rate": 3.675335786022749e-06, + "loss": 1.4157, + "step": 38967 + }, + { + "epoch": 0.9134712870010607, + "grad_norm": 4.0444207191467285, + "learning_rate": 3.6733573609794015e-06, + "loss": 0.2658, + "step": 38968 + }, + { + "epoch": 0.9134947285758658, + "grad_norm": 5.096565246582031, + "learning_rate": 3.671379458607105e-06, + "loss": 0.5492, + "step": 38969 + }, + { + "epoch": 0.9135181701506707, + "grad_norm": 0.45125240087509155, + "learning_rate": 3.6694020789166172e-06, + "loss": 0.0195, + "step": 38970 + }, + { + "epoch": 0.9135416117254758, + "grad_norm": 1.448277473449707, + "learning_rate": 3.6674252219186857e-06, + "loss": 0.1441, + "step": 38971 + }, + { + "epoch": 0.9135650533002807, + "grad_norm": 2.783679962158203, + "learning_rate": 3.665448887624001e-06, + "loss": 0.106, + "step": 38972 + }, + { + "epoch": 0.9135884948750858, + "grad_norm": 0.6268703937530518, + "learning_rate": 3.6634730760433002e-06, + "loss": 0.025, + "step": 38973 + }, + { + "epoch": 0.9136119364498907, + "grad_norm": 3.534977436065674, + "learning_rate": 3.6614977871873066e-06, + "loss": 0.2136, + "step": 38974 + }, + { + "epoch": 0.9136353780246957, + "grad_norm": 2.9559566974639893, + "learning_rate": 3.6595230210667354e-06, + "loss": 0.2455, + "step": 38975 + }, + { + "epoch": 0.9136588195995007, + "grad_norm": 2.112151622772217, + "learning_rate": 3.657548777692288e-06, + "loss": 1.7238, + "step": 38976 + }, + { + "epoch": 0.9136822611743057, + "grad_norm": 3.3201773166656494, + "learning_rate": 3.6555750570747005e-06, + "loss": 0.2676, + "step": 38977 + }, + { + "epoch": 0.9137057027491107, + "grad_norm": 3.0561394691467285, + "learning_rate": 3.6536018592246646e-06, + "loss": 0.2775, + "step": 38978 + }, + { + "epoch": 0.9137291443239157, + "grad_norm": 4.0216498374938965, + "learning_rate": 3.651629184152905e-06, + "loss": 0.3309, + "step": 38979 + }, + { + "epoch": 0.9137525858987207, + "grad_norm": 2.270738363265991, + "learning_rate": 3.649657031870113e-06, + "loss": 0.2638, + "step": 38980 + }, + { + "epoch": 0.9137760274735257, + "grad_norm": 2.3130481243133545, + "learning_rate": 3.64768540238698e-06, + "loss": 0.0725, + "step": 38981 + }, + { + "epoch": 0.9137994690483306, + "grad_norm": 2.505180597305298, + "learning_rate": 3.64571429571422e-06, + "loss": 0.1681, + "step": 38982 + }, + { + "epoch": 0.9138229106231357, + "grad_norm": 2.6935031414031982, + "learning_rate": 3.643743711862524e-06, + "loss": 0.1868, + "step": 38983 + }, + { + "epoch": 0.9138463521979406, + "grad_norm": 2.446941614151001, + "learning_rate": 3.641773650842584e-06, + "loss": 1.1026, + "step": 38984 + }, + { + "epoch": 0.9138697937727457, + "grad_norm": 1.5328136682510376, + "learning_rate": 3.6398041126650906e-06, + "loss": 0.6749, + "step": 38985 + }, + { + "epoch": 0.9138932353475506, + "grad_norm": 0.32180583477020264, + "learning_rate": 3.637835097340736e-06, + "loss": 0.0168, + "step": 38986 + }, + { + "epoch": 0.9139166769223557, + "grad_norm": 6.693302154541016, + "learning_rate": 3.6358666048801894e-06, + "loss": 0.7041, + "step": 38987 + }, + { + "epoch": 0.9139401184971606, + "grad_norm": 2.092780351638794, + "learning_rate": 3.633898635294142e-06, + "loss": 0.118, + "step": 38988 + }, + { + "epoch": 0.9139635600719657, + "grad_norm": 1.5190623998641968, + "learning_rate": 3.6319311885932627e-06, + "loss": 0.0451, + "step": 38989 + }, + { + "epoch": 0.9139870016467706, + "grad_norm": 0.8297644257545471, + "learning_rate": 3.6299642647882547e-06, + "loss": 0.05, + "step": 38990 + }, + { + "epoch": 0.9140104432215757, + "grad_norm": 4.011929988861084, + "learning_rate": 3.6279978638897537e-06, + "loss": 0.4727, + "step": 38991 + }, + { + "epoch": 0.9140338847963806, + "grad_norm": 1.9139726161956787, + "learning_rate": 3.6260319859084513e-06, + "loss": 0.1435, + "step": 38992 + }, + { + "epoch": 0.9140573263711856, + "grad_norm": 2.110536575317383, + "learning_rate": 3.6240666308550163e-06, + "loss": 0.1017, + "step": 38993 + }, + { + "epoch": 0.9140807679459906, + "grad_norm": 5.212272644042969, + "learning_rate": 3.6221017987400853e-06, + "loss": 0.3293, + "step": 38994 + }, + { + "epoch": 0.9141042095207956, + "grad_norm": 1.65030038356781, + "learning_rate": 3.6201374895743377e-06, + "loss": 0.0573, + "step": 38995 + }, + { + "epoch": 0.9141276510956006, + "grad_norm": 0.6393111944198608, + "learning_rate": 3.618173703368455e-06, + "loss": 0.0392, + "step": 38996 + }, + { + "epoch": 0.9141510926704056, + "grad_norm": 1.1698743104934692, + "learning_rate": 3.6162104401330722e-06, + "loss": 0.081, + "step": 38997 + }, + { + "epoch": 0.9141745342452106, + "grad_norm": 1.226824164390564, + "learning_rate": 3.6142476998788365e-06, + "loss": 0.1112, + "step": 38998 + }, + { + "epoch": 0.9141979758200156, + "grad_norm": 1.4990390539169312, + "learning_rate": 3.6122854826163954e-06, + "loss": 0.0541, + "step": 38999 + }, + { + "epoch": 0.9142214173948205, + "grad_norm": 2.3839449882507324, + "learning_rate": 3.6103237883564178e-06, + "loss": 0.1558, + "step": 39000 + }, + { + "epoch": 0.9142448589696256, + "grad_norm": 1.0581437349319458, + "learning_rate": 3.6083626171095174e-06, + "loss": 0.109, + "step": 39001 + }, + { + "epoch": 0.9142683005444305, + "grad_norm": 2.1803314685821533, + "learning_rate": 3.6064019688863525e-06, + "loss": 0.1267, + "step": 39002 + }, + { + "epoch": 0.9142917421192356, + "grad_norm": 3.838440418243408, + "learning_rate": 3.60444184369757e-06, + "loss": 0.337, + "step": 39003 + }, + { + "epoch": 0.9143151836940405, + "grad_norm": 2.035083293914795, + "learning_rate": 3.6024822415537953e-06, + "loss": 0.1096, + "step": 39004 + }, + { + "epoch": 0.9143386252688456, + "grad_norm": 3.53592586517334, + "learning_rate": 3.600523162465663e-06, + "loss": 0.2643, + "step": 39005 + }, + { + "epoch": 0.9143620668436505, + "grad_norm": 4.587161540985107, + "learning_rate": 3.5985646064437994e-06, + "loss": 0.4091, + "step": 39006 + }, + { + "epoch": 0.9143855084184556, + "grad_norm": 0.7691444158554077, + "learning_rate": 3.5966065734988398e-06, + "loss": 0.0399, + "step": 39007 + }, + { + "epoch": 0.9144089499932605, + "grad_norm": 2.223080635070801, + "learning_rate": 3.5946490636413865e-06, + "loss": 0.2291, + "step": 39008 + }, + { + "epoch": 0.9144323915680656, + "grad_norm": 3.770843505859375, + "learning_rate": 3.5926920768820984e-06, + "loss": 0.2688, + "step": 39009 + }, + { + "epoch": 0.9144558331428705, + "grad_norm": 2.3870904445648193, + "learning_rate": 3.590735613231566e-06, + "loss": 0.1022, + "step": 39010 + }, + { + "epoch": 0.9144792747176755, + "grad_norm": 0.8885020613670349, + "learning_rate": 3.588779672700415e-06, + "loss": 0.0439, + "step": 39011 + }, + { + "epoch": 0.9145027162924805, + "grad_norm": 3.1009609699249268, + "learning_rate": 3.5868242552992594e-06, + "loss": 0.2673, + "step": 39012 + }, + { + "epoch": 0.9145261578672855, + "grad_norm": 2.022430658340454, + "learning_rate": 3.584869361038701e-06, + "loss": 0.1614, + "step": 39013 + }, + { + "epoch": 0.9145495994420905, + "grad_norm": 2.9832122325897217, + "learning_rate": 3.582914989929342e-06, + "loss": 1.6382, + "step": 39014 + }, + { + "epoch": 0.9145730410168955, + "grad_norm": 3.1909902095794678, + "learning_rate": 3.580961141981809e-06, + "loss": 0.1574, + "step": 39015 + }, + { + "epoch": 0.9145964825917006, + "grad_norm": 3.2353219985961914, + "learning_rate": 3.579007817206692e-06, + "loss": 0.2686, + "step": 39016 + }, + { + "epoch": 0.9146199241665055, + "grad_norm": 2.901080369949341, + "learning_rate": 3.5770550156145944e-06, + "loss": 0.2819, + "step": 39017 + }, + { + "epoch": 0.9146433657413106, + "grad_norm": 0.3389972150325775, + "learning_rate": 3.5751027372160963e-06, + "loss": 0.0146, + "step": 39018 + }, + { + "epoch": 0.9146668073161155, + "grad_norm": 3.4256319999694824, + "learning_rate": 3.5731509820218113e-06, + "loss": 0.5284, + "step": 39019 + }, + { + "epoch": 0.9146902488909205, + "grad_norm": 4.669649600982666, + "learning_rate": 3.57119975004232e-06, + "loss": 0.4662, + "step": 39020 + }, + { + "epoch": 0.9147136904657255, + "grad_norm": 0.5740401148796082, + "learning_rate": 3.569249041288192e-06, + "loss": 0.0183, + "step": 39021 + }, + { + "epoch": 0.9147371320405305, + "grad_norm": 1.5392265319824219, + "learning_rate": 3.5672988557700515e-06, + "loss": 0.0908, + "step": 39022 + }, + { + "epoch": 0.9147605736153355, + "grad_norm": 2.3536198139190674, + "learning_rate": 3.5653491934984573e-06, + "loss": 0.2772, + "step": 39023 + }, + { + "epoch": 0.9147840151901405, + "grad_norm": 2.966477870941162, + "learning_rate": 3.5634000544839896e-06, + "loss": 0.2125, + "step": 39024 + }, + { + "epoch": 0.9148074567649455, + "grad_norm": 2.2717294692993164, + "learning_rate": 3.5614514387372176e-06, + "loss": 1.4476, + "step": 39025 + }, + { + "epoch": 0.9148308983397505, + "grad_norm": 3.0194027423858643, + "learning_rate": 3.559503346268711e-06, + "loss": 0.2902, + "step": 39026 + }, + { + "epoch": 0.9148543399145554, + "grad_norm": 1.5105681419372559, + "learning_rate": 3.5575557770890723e-06, + "loss": 0.1492, + "step": 39027 + }, + { + "epoch": 0.9148777814893605, + "grad_norm": 9.696088790893555, + "learning_rate": 3.555608731208837e-06, + "loss": 0.7592, + "step": 39028 + }, + { + "epoch": 0.9149012230641654, + "grad_norm": 2.199704647064209, + "learning_rate": 3.5536622086385863e-06, + "loss": 1.5998, + "step": 39029 + }, + { + "epoch": 0.9149246646389705, + "grad_norm": 1.1548980474472046, + "learning_rate": 3.5517162093888665e-06, + "loss": 0.1015, + "step": 39030 + }, + { + "epoch": 0.9149481062137754, + "grad_norm": 3.0594422817230225, + "learning_rate": 3.549770733470259e-06, + "loss": 0.1721, + "step": 39031 + }, + { + "epoch": 0.9149715477885805, + "grad_norm": 2.9577908515930176, + "learning_rate": 3.5478257808932992e-06, + "loss": 0.2379, + "step": 39032 + }, + { + "epoch": 0.9149949893633854, + "grad_norm": 2.335177183151245, + "learning_rate": 3.5458813516685453e-06, + "loss": 0.2092, + "step": 39033 + }, + { + "epoch": 0.9150184309381905, + "grad_norm": 0.7548976540565491, + "learning_rate": 3.5439374458065556e-06, + "loss": 0.0755, + "step": 39034 + }, + { + "epoch": 0.9150418725129954, + "grad_norm": 0.47502464056015015, + "learning_rate": 3.5419940633178774e-06, + "loss": 0.0354, + "step": 39035 + }, + { + "epoch": 0.9150653140878005, + "grad_norm": 3.008258581161499, + "learning_rate": 3.5400512042130574e-06, + "loss": 0.2424, + "step": 39036 + }, + { + "epoch": 0.9150887556626054, + "grad_norm": 1.9314144849777222, + "learning_rate": 3.538108868502621e-06, + "loss": 0.0844, + "step": 39037 + }, + { + "epoch": 0.9151121972374104, + "grad_norm": 1.2718867063522339, + "learning_rate": 3.5361670561971262e-06, + "loss": 0.0581, + "step": 39038 + }, + { + "epoch": 0.9151356388122154, + "grad_norm": 1.9837062358856201, + "learning_rate": 3.5342257673070867e-06, + "loss": 0.106, + "step": 39039 + }, + { + "epoch": 0.9151590803870204, + "grad_norm": 0.841808557510376, + "learning_rate": 3.532285001843061e-06, + "loss": 0.1018, + "step": 39040 + }, + { + "epoch": 0.9151825219618254, + "grad_norm": 3.3326921463012695, + "learning_rate": 3.5303447598155624e-06, + "loss": 0.2958, + "step": 39041 + }, + { + "epoch": 0.9152059635366304, + "grad_norm": 0.9860419034957886, + "learning_rate": 3.528405041235139e-06, + "loss": 0.0629, + "step": 39042 + }, + { + "epoch": 0.9152294051114354, + "grad_norm": 2.936734676361084, + "learning_rate": 3.5264658461122814e-06, + "loss": 2.2527, + "step": 39043 + }, + { + "epoch": 0.9152528466862404, + "grad_norm": 2.499662160873413, + "learning_rate": 3.524527174457548e-06, + "loss": 0.2057, + "step": 39044 + }, + { + "epoch": 0.9152762882610453, + "grad_norm": 1.1637159585952759, + "learning_rate": 3.522589026281442e-06, + "loss": 0.4348, + "step": 39045 + }, + { + "epoch": 0.9152997298358504, + "grad_norm": 1.3432137966156006, + "learning_rate": 3.5206514015944767e-06, + "loss": 0.1317, + "step": 39046 + }, + { + "epoch": 0.9153231714106553, + "grad_norm": 0.8818445801734924, + "learning_rate": 3.5187143004071777e-06, + "loss": 0.0494, + "step": 39047 + }, + { + "epoch": 0.9153466129854604, + "grad_norm": 1.3446398973464966, + "learning_rate": 3.516777722730036e-06, + "loss": 0.063, + "step": 39048 + }, + { + "epoch": 0.9153700545602653, + "grad_norm": 2.302008867263794, + "learning_rate": 3.5148416685735876e-06, + "loss": 0.2609, + "step": 39049 + }, + { + "epoch": 0.9153934961350704, + "grad_norm": 1.589446783065796, + "learning_rate": 3.512906137948313e-06, + "loss": 0.0301, + "step": 39050 + }, + { + "epoch": 0.9154169377098753, + "grad_norm": 3.720402956008911, + "learning_rate": 3.510971130864715e-06, + "loss": 0.3146, + "step": 39051 + }, + { + "epoch": 0.9154403792846804, + "grad_norm": 1.0696293115615845, + "learning_rate": 3.509036647333297e-06, + "loss": 0.039, + "step": 39052 + }, + { + "epoch": 0.9154638208594853, + "grad_norm": 1.0847855806350708, + "learning_rate": 3.507102687364572e-06, + "loss": 0.085, + "step": 39053 + }, + { + "epoch": 0.9154872624342903, + "grad_norm": 1.7346986532211304, + "learning_rate": 3.5051692509690205e-06, + "loss": 0.1124, + "step": 39054 + }, + { + "epoch": 0.9155107040090953, + "grad_norm": 3.6342644691467285, + "learning_rate": 3.503236338157123e-06, + "loss": 0.2979, + "step": 39055 + }, + { + "epoch": 0.9155341455839003, + "grad_norm": 2.3531577587127686, + "learning_rate": 3.501303948939394e-06, + "loss": 0.2201, + "step": 39056 + }, + { + "epoch": 0.9155575871587053, + "grad_norm": 3.0400960445404053, + "learning_rate": 3.499372083326291e-06, + "loss": 0.3603, + "step": 39057 + }, + { + "epoch": 0.9155810287335103, + "grad_norm": 0.8605417013168335, + "learning_rate": 3.4974407413283172e-06, + "loss": 0.0462, + "step": 39058 + }, + { + "epoch": 0.9156044703083153, + "grad_norm": 2.9921419620513916, + "learning_rate": 3.495509922955931e-06, + "loss": 0.1832, + "step": 39059 + }, + { + "epoch": 0.9156279118831203, + "grad_norm": 2.92256760597229, + "learning_rate": 3.493579628219634e-06, + "loss": 0.3299, + "step": 39060 + }, + { + "epoch": 0.9156513534579253, + "grad_norm": 1.8718773126602173, + "learning_rate": 3.4916498571298862e-06, + "loss": 0.1075, + "step": 39061 + }, + { + "epoch": 0.9156747950327303, + "grad_norm": 2.4209327697753906, + "learning_rate": 3.489720609697156e-06, + "loss": 0.1022, + "step": 39062 + }, + { + "epoch": 0.9156982366075352, + "grad_norm": 3.1875741481781006, + "learning_rate": 3.4877918859319124e-06, + "loss": 0.2648, + "step": 39063 + }, + { + "epoch": 0.9157216781823403, + "grad_norm": 0.4246201515197754, + "learning_rate": 3.485863685844626e-06, + "loss": 0.0229, + "step": 39064 + }, + { + "epoch": 0.9157451197571452, + "grad_norm": 3.2269320487976074, + "learning_rate": 3.483936009445754e-06, + "loss": 0.1477, + "step": 39065 + }, + { + "epoch": 0.9157685613319503, + "grad_norm": 2.002068519592285, + "learning_rate": 3.4820088567457666e-06, + "loss": 0.0914, + "step": 39066 + }, + { + "epoch": 0.9157920029067552, + "grad_norm": 2.72355318069458, + "learning_rate": 3.4800822277550993e-06, + "loss": 0.3399, + "step": 39067 + }, + { + "epoch": 0.9158154444815603, + "grad_norm": 2.15535831451416, + "learning_rate": 3.478156122484233e-06, + "loss": 0.1124, + "step": 39068 + }, + { + "epoch": 0.9158388860563653, + "grad_norm": 3.402324914932251, + "learning_rate": 3.4762305409436037e-06, + "loss": 0.3232, + "step": 39069 + }, + { + "epoch": 0.9158623276311703, + "grad_norm": 1.92747163772583, + "learning_rate": 3.4743054831436695e-06, + "loss": 0.0865, + "step": 39070 + }, + { + "epoch": 0.9158857692059753, + "grad_norm": 0.3874748647212982, + "learning_rate": 3.4723809490948446e-06, + "loss": 0.0228, + "step": 39071 + }, + { + "epoch": 0.9159092107807802, + "grad_norm": 0.49955037236213684, + "learning_rate": 3.470456938807609e-06, + "loss": 0.0422, + "step": 39072 + }, + { + "epoch": 0.9159326523555853, + "grad_norm": 2.5371577739715576, + "learning_rate": 3.4685334522923997e-06, + "loss": 0.2911, + "step": 39073 + }, + { + "epoch": 0.9159560939303902, + "grad_norm": 4.055713176727295, + "learning_rate": 3.46661048955963e-06, + "loss": 0.2536, + "step": 39074 + }, + { + "epoch": 0.9159795355051953, + "grad_norm": 3.4698967933654785, + "learning_rate": 3.4646880506197466e-06, + "loss": 0.1511, + "step": 39075 + }, + { + "epoch": 0.9160029770800002, + "grad_norm": 3.0430076122283936, + "learning_rate": 3.4627661354831863e-06, + "loss": 0.102, + "step": 39076 + }, + { + "epoch": 0.9160264186548053, + "grad_norm": 4.531479358673096, + "learning_rate": 3.460844744160352e-06, + "loss": 0.1992, + "step": 39077 + }, + { + "epoch": 0.9160498602296102, + "grad_norm": 5.8338212966918945, + "learning_rate": 3.4589238766617015e-06, + "loss": 0.386, + "step": 39078 + }, + { + "epoch": 0.9160733018044153, + "grad_norm": 1.109158992767334, + "learning_rate": 3.457003532997638e-06, + "loss": 0.0669, + "step": 39079 + }, + { + "epoch": 0.9160967433792202, + "grad_norm": 1.8212261199951172, + "learning_rate": 3.455083713178597e-06, + "loss": 0.1043, + "step": 39080 + }, + { + "epoch": 0.9161201849540253, + "grad_norm": 3.2590224742889404, + "learning_rate": 3.4531644172149823e-06, + "loss": 1.8597, + "step": 39081 + }, + { + "epoch": 0.9161436265288302, + "grad_norm": 1.7523221969604492, + "learning_rate": 3.4512456451172183e-06, + "loss": 0.1547, + "step": 39082 + }, + { + "epoch": 0.9161670681036352, + "grad_norm": 3.108881950378418, + "learning_rate": 3.4493273968957073e-06, + "loss": 0.1457, + "step": 39083 + }, + { + "epoch": 0.9161905096784402, + "grad_norm": 0.908389687538147, + "learning_rate": 3.447409672560853e-06, + "loss": 0.0587, + "step": 39084 + }, + { + "epoch": 0.9162139512532452, + "grad_norm": 3.3821866512298584, + "learning_rate": 3.4454924721230795e-06, + "loss": 0.1882, + "step": 39085 + }, + { + "epoch": 0.9162373928280502, + "grad_norm": 3.7014448642730713, + "learning_rate": 3.4435757955927684e-06, + "loss": 0.272, + "step": 39086 + }, + { + "epoch": 0.9162608344028552, + "grad_norm": 1.3558162450790405, + "learning_rate": 3.441659642980344e-06, + "loss": 0.059, + "step": 39087 + }, + { + "epoch": 0.9162842759776602, + "grad_norm": 1.7514690160751343, + "learning_rate": 3.4397440142961757e-06, + "loss": 0.8019, + "step": 39088 + }, + { + "epoch": 0.9163077175524652, + "grad_norm": 2.351735830307007, + "learning_rate": 3.437828909550689e-06, + "loss": 0.206, + "step": 39089 + }, + { + "epoch": 0.9163311591272701, + "grad_norm": 0.8016148209571838, + "learning_rate": 3.4359143287542306e-06, + "loss": 0.0207, + "step": 39090 + }, + { + "epoch": 0.9163546007020752, + "grad_norm": 3.3555500507354736, + "learning_rate": 3.434000271917226e-06, + "loss": 0.1918, + "step": 39091 + }, + { + "epoch": 0.9163780422768801, + "grad_norm": 4.086506366729736, + "learning_rate": 3.4320867390500554e-06, + "loss": 0.5173, + "step": 39092 + }, + { + "epoch": 0.9164014838516852, + "grad_norm": 1.597104787826538, + "learning_rate": 3.4301737301631e-06, + "loss": 0.0968, + "step": 39093 + }, + { + "epoch": 0.9164249254264901, + "grad_norm": 1.0500611066818237, + "learning_rate": 3.4282612452667393e-06, + "loss": 0.1195, + "step": 39094 + }, + { + "epoch": 0.9164483670012952, + "grad_norm": 2.7006356716156006, + "learning_rate": 3.426349284371344e-06, + "loss": 0.2298, + "step": 39095 + }, + { + "epoch": 0.9164718085761001, + "grad_norm": 2.8751816749572754, + "learning_rate": 3.424437847487294e-06, + "loss": 0.2292, + "step": 39096 + }, + { + "epoch": 0.9164952501509052, + "grad_norm": 1.034539818763733, + "learning_rate": 3.4225269346249476e-06, + "loss": 0.1175, + "step": 39097 + }, + { + "epoch": 0.9165186917257101, + "grad_norm": 2.5849294662475586, + "learning_rate": 3.420616545794708e-06, + "loss": 0.1391, + "step": 39098 + }, + { + "epoch": 0.9165421333005151, + "grad_norm": 1.0528911352157593, + "learning_rate": 3.4187066810069003e-06, + "loss": 0.1193, + "step": 39099 + }, + { + "epoch": 0.9165655748753201, + "grad_norm": 0.9566137194633484, + "learning_rate": 3.4167973402719155e-06, + "loss": 0.0794, + "step": 39100 + }, + { + "epoch": 0.9165890164501251, + "grad_norm": 2.3948543071746826, + "learning_rate": 3.4148885236001128e-06, + "loss": 0.2201, + "step": 39101 + }, + { + "epoch": 0.9166124580249301, + "grad_norm": 2.7779481410980225, + "learning_rate": 3.4129802310018166e-06, + "loss": 0.3235, + "step": 39102 + }, + { + "epoch": 0.9166358995997351, + "grad_norm": 4.765302658081055, + "learning_rate": 3.411072462487408e-06, + "loss": 0.4809, + "step": 39103 + }, + { + "epoch": 0.9166593411745401, + "grad_norm": 1.9049123525619507, + "learning_rate": 3.4091652180672452e-06, + "loss": 0.092, + "step": 39104 + }, + { + "epoch": 0.9166827827493451, + "grad_norm": 1.6615568399429321, + "learning_rate": 3.4072584977516753e-06, + "loss": 0.1091, + "step": 39105 + }, + { + "epoch": 0.91670622432415, + "grad_norm": 1.9706134796142578, + "learning_rate": 3.405352301551035e-06, + "loss": 0.1017, + "step": 39106 + }, + { + "epoch": 0.9167296658989551, + "grad_norm": 0.5130146741867065, + "learning_rate": 3.4034466294756593e-06, + "loss": 0.0479, + "step": 39107 + }, + { + "epoch": 0.91675310747376, + "grad_norm": 2.183621406555176, + "learning_rate": 3.401541481535908e-06, + "loss": 0.1417, + "step": 39108 + }, + { + "epoch": 0.9167765490485651, + "grad_norm": 3.139997959136963, + "learning_rate": 3.3996368577420947e-06, + "loss": 0.2069, + "step": 39109 + }, + { + "epoch": 0.91679999062337, + "grad_norm": 3.0531718730926514, + "learning_rate": 3.397732758104577e-06, + "loss": 0.3442, + "step": 39110 + }, + { + "epoch": 0.9168234321981751, + "grad_norm": 1.3749092817306519, + "learning_rate": 3.395829182633681e-06, + "loss": 0.1487, + "step": 39111 + }, + { + "epoch": 0.91684687377298, + "grad_norm": 2.2213611602783203, + "learning_rate": 3.3939261313397308e-06, + "loss": 0.1031, + "step": 39112 + }, + { + "epoch": 0.9168703153477851, + "grad_norm": 3.2735204696655273, + "learning_rate": 3.392023604233063e-06, + "loss": 0.284, + "step": 39113 + }, + { + "epoch": 0.91689375692259, + "grad_norm": 2.0448405742645264, + "learning_rate": 3.390121601323981e-06, + "loss": 0.1144, + "step": 39114 + }, + { + "epoch": 0.9169171984973951, + "grad_norm": 0.8358117341995239, + "learning_rate": 3.3882201226228093e-06, + "loss": 0.0366, + "step": 39115 + }, + { + "epoch": 0.9169406400722, + "grad_norm": 0.5888127088546753, + "learning_rate": 3.3863191681398622e-06, + "loss": 0.0803, + "step": 39116 + }, + { + "epoch": 0.916964081647005, + "grad_norm": 0.8661376237869263, + "learning_rate": 3.3844187378854864e-06, + "loss": 0.0761, + "step": 39117 + }, + { + "epoch": 0.91698752322181, + "grad_norm": 1.423478603363037, + "learning_rate": 3.382518831869963e-06, + "loss": 0.1437, + "step": 39118 + }, + { + "epoch": 0.917010964796615, + "grad_norm": 0.2340449094772339, + "learning_rate": 3.380619450103606e-06, + "loss": 0.0188, + "step": 39119 + }, + { + "epoch": 0.9170344063714201, + "grad_norm": 3.722419500350952, + "learning_rate": 3.3787205925967403e-06, + "loss": 0.142, + "step": 39120 + }, + { + "epoch": 0.917057847946225, + "grad_norm": 2.713404655456543, + "learning_rate": 3.3768222593596353e-06, + "loss": 0.9661, + "step": 39121 + }, + { + "epoch": 0.9170812895210301, + "grad_norm": 2.7431020736694336, + "learning_rate": 3.374924450402617e-06, + "loss": 0.1454, + "step": 39122 + }, + { + "epoch": 0.917104731095835, + "grad_norm": 2.7192139625549316, + "learning_rate": 3.373027165735976e-06, + "loss": 0.2184, + "step": 39123 + }, + { + "epoch": 0.9171281726706401, + "grad_norm": 0.7959463596343994, + "learning_rate": 3.371130405370004e-06, + "loss": 0.0462, + "step": 39124 + }, + { + "epoch": 0.917151614245445, + "grad_norm": 3.8819079399108887, + "learning_rate": 3.3692341693149942e-06, + "loss": 1.9585, + "step": 39125 + }, + { + "epoch": 0.91717505582025, + "grad_norm": 1.5590205192565918, + "learning_rate": 3.3673384575812484e-06, + "loss": 0.1317, + "step": 39126 + }, + { + "epoch": 0.917198497395055, + "grad_norm": 1.9123800992965698, + "learning_rate": 3.365443270179025e-06, + "loss": 0.1291, + "step": 39127 + }, + { + "epoch": 0.91722193896986, + "grad_norm": 1.7391752004623413, + "learning_rate": 3.3635486071186273e-06, + "loss": 0.0627, + "step": 39128 + }, + { + "epoch": 0.917245380544665, + "grad_norm": 2.449763774871826, + "learning_rate": 3.3616544684103247e-06, + "loss": 1.6338, + "step": 39129 + }, + { + "epoch": 0.91726882211947, + "grad_norm": 0.8054003715515137, + "learning_rate": 3.359760854064409e-06, + "loss": 0.0459, + "step": 39130 + }, + { + "epoch": 0.917292263694275, + "grad_norm": 2.0264971256256104, + "learning_rate": 3.3578677640911603e-06, + "loss": 0.1511, + "step": 39131 + }, + { + "epoch": 0.91731570526908, + "grad_norm": 1.655760645866394, + "learning_rate": 3.355975198500827e-06, + "loss": 1.1468, + "step": 39132 + }, + { + "epoch": 0.917339146843885, + "grad_norm": 3.781343936920166, + "learning_rate": 3.3540831573037003e-06, + "loss": 0.2188, + "step": 39133 + }, + { + "epoch": 0.91736258841869, + "grad_norm": 1.641782283782959, + "learning_rate": 3.3521916405100163e-06, + "loss": 0.962, + "step": 39134 + }, + { + "epoch": 0.917386029993495, + "grad_norm": 0.9915540218353271, + "learning_rate": 3.350300648130067e-06, + "loss": 0.0422, + "step": 39135 + }, + { + "epoch": 0.9174094715683, + "grad_norm": 1.4388905763626099, + "learning_rate": 3.3484101801741107e-06, + "loss": 0.0454, + "step": 39136 + }, + { + "epoch": 0.9174329131431049, + "grad_norm": 3.954835891723633, + "learning_rate": 3.346520236652395e-06, + "loss": 0.1904, + "step": 39137 + }, + { + "epoch": 0.91745635471791, + "grad_norm": 0.48423051834106445, + "learning_rate": 3.3446308175751896e-06, + "loss": 0.0271, + "step": 39138 + }, + { + "epoch": 0.9174797962927149, + "grad_norm": 5.388535022735596, + "learning_rate": 3.342741922952719e-06, + "loss": 0.3098, + "step": 39139 + }, + { + "epoch": 0.91750323786752, + "grad_norm": 3.262633800506592, + "learning_rate": 3.340853552795242e-06, + "loss": 0.3582, + "step": 39140 + }, + { + "epoch": 0.9175266794423249, + "grad_norm": 2.6422369480133057, + "learning_rate": 3.3389657071130174e-06, + "loss": 0.1703, + "step": 39141 + }, + { + "epoch": 0.91755012101713, + "grad_norm": 0.7688231468200684, + "learning_rate": 3.3370783859162923e-06, + "loss": 0.0484, + "step": 39142 + }, + { + "epoch": 0.9175735625919349, + "grad_norm": 4.9484710693359375, + "learning_rate": 3.3351915892152918e-06, + "loss": 0.7246, + "step": 39143 + }, + { + "epoch": 0.91759700416674, + "grad_norm": 1.926772952079773, + "learning_rate": 3.333305317020263e-06, + "loss": 0.0998, + "step": 39144 + }, + { + "epoch": 0.9176204457415449, + "grad_norm": 0.44513678550720215, + "learning_rate": 3.331419569341432e-06, + "loss": 0.0198, + "step": 39145 + }, + { + "epoch": 0.9176438873163499, + "grad_norm": 2.5401430130004883, + "learning_rate": 3.329534346189045e-06, + "loss": 0.2186, + "step": 39146 + }, + { + "epoch": 0.9176673288911549, + "grad_norm": 1.1030123233795166, + "learning_rate": 3.3276496475733053e-06, + "loss": 0.0619, + "step": 39147 + }, + { + "epoch": 0.9176907704659599, + "grad_norm": 4.6037983894348145, + "learning_rate": 3.325765473504472e-06, + "loss": 0.4559, + "step": 39148 + }, + { + "epoch": 0.9177142120407649, + "grad_norm": 2.018442392349243, + "learning_rate": 3.323881823992758e-06, + "loss": 1.5511, + "step": 39149 + }, + { + "epoch": 0.9177376536155699, + "grad_norm": 1.7516549825668335, + "learning_rate": 3.3219986990483674e-06, + "loss": 0.1207, + "step": 39150 + }, + { + "epoch": 0.9177610951903749, + "grad_norm": 3.7691402435302734, + "learning_rate": 3.320116098681536e-06, + "loss": 0.3945, + "step": 39151 + }, + { + "epoch": 0.9177845367651799, + "grad_norm": 4.152720928192139, + "learning_rate": 3.318234022902467e-06, + "loss": 0.3312, + "step": 39152 + }, + { + "epoch": 0.9178079783399848, + "grad_norm": 0.7948034405708313, + "learning_rate": 3.3163524717213847e-06, + "loss": 0.0848, + "step": 39153 + }, + { + "epoch": 0.9178314199147899, + "grad_norm": 3.0527162551879883, + "learning_rate": 3.3144714451484813e-06, + "loss": 0.1552, + "step": 39154 + }, + { + "epoch": 0.9178548614895948, + "grad_norm": 3.2858214378356934, + "learning_rate": 3.312590943193983e-06, + "loss": 0.1337, + "step": 39155 + }, + { + "epoch": 0.9178783030643999, + "grad_norm": 0.7319098711013794, + "learning_rate": 3.3107109658680912e-06, + "loss": 0.0743, + "step": 39156 + }, + { + "epoch": 0.9179017446392048, + "grad_norm": 3.1955137252807617, + "learning_rate": 3.308831513180999e-06, + "loss": 0.391, + "step": 39157 + }, + { + "epoch": 0.9179251862140099, + "grad_norm": 1.902571439743042, + "learning_rate": 3.3069525851429085e-06, + "loss": 0.1413, + "step": 39158 + }, + { + "epoch": 0.9179486277888148, + "grad_norm": 0.5777831077575684, + "learning_rate": 3.3050741817640006e-06, + "loss": 0.0604, + "step": 39159 + }, + { + "epoch": 0.9179720693636199, + "grad_norm": 4.218545913696289, + "learning_rate": 3.303196303054479e-06, + "loss": 0.2503, + "step": 39160 + }, + { + "epoch": 0.9179955109384248, + "grad_norm": 2.442934989929199, + "learning_rate": 3.3013189490245344e-06, + "loss": 0.2745, + "step": 39161 + }, + { + "epoch": 0.9180189525132298, + "grad_norm": 1.2197355031967163, + "learning_rate": 3.2994421196843595e-06, + "loss": 0.1315, + "step": 39162 + }, + { + "epoch": 0.9180423940880348, + "grad_norm": 0.5279836058616638, + "learning_rate": 3.297565815044135e-06, + "loss": 0.0239, + "step": 39163 + }, + { + "epoch": 0.9180658356628398, + "grad_norm": 2.0603079795837402, + "learning_rate": 3.2956900351140187e-06, + "loss": 1.7387, + "step": 39164 + }, + { + "epoch": 0.9180892772376448, + "grad_norm": 2.374865770339966, + "learning_rate": 3.293814779904214e-06, + "loss": 0.0641, + "step": 39165 + }, + { + "epoch": 0.9181127188124498, + "grad_norm": 2.4931278228759766, + "learning_rate": 3.291940049424902e-06, + "loss": 0.1408, + "step": 39166 + }, + { + "epoch": 0.9181361603872548, + "grad_norm": 2.3559868335723877, + "learning_rate": 3.2900658436862186e-06, + "loss": 0.1593, + "step": 39167 + }, + { + "epoch": 0.9181596019620598, + "grad_norm": 3.764066457748413, + "learning_rate": 3.288192162698378e-06, + "loss": 0.3063, + "step": 39168 + }, + { + "epoch": 0.9181830435368648, + "grad_norm": 2.846731662750244, + "learning_rate": 3.286319006471528e-06, + "loss": 1.4726, + "step": 39169 + }, + { + "epoch": 0.9182064851116698, + "grad_norm": 3.74708890914917, + "learning_rate": 3.284446375015826e-06, + "loss": 0.196, + "step": 39170 + }, + { + "epoch": 0.9182299266864749, + "grad_norm": 1.574734091758728, + "learning_rate": 3.2825742683414317e-06, + "loss": 1.0612, + "step": 39171 + }, + { + "epoch": 0.9182533682612798, + "grad_norm": 0.777042806148529, + "learning_rate": 3.280702686458503e-06, + "loss": 0.0704, + "step": 39172 + }, + { + "epoch": 0.9182768098360848, + "grad_norm": 2.568335771560669, + "learning_rate": 3.27883162937721e-06, + "loss": 0.1716, + "step": 39173 + }, + { + "epoch": 0.9183002514108898, + "grad_norm": 2.2336225509643555, + "learning_rate": 3.2769610971077e-06, + "loss": 0.172, + "step": 39174 + }, + { + "epoch": 0.9183236929856948, + "grad_norm": 1.6940404176712036, + "learning_rate": 3.2750910896601207e-06, + "loss": 0.1249, + "step": 39175 + }, + { + "epoch": 0.9183471345604998, + "grad_norm": 0.44576507806777954, + "learning_rate": 3.2732216070445963e-06, + "loss": 0.0278, + "step": 39176 + }, + { + "epoch": 0.9183705761353048, + "grad_norm": 3.268662452697754, + "learning_rate": 3.2713526492713087e-06, + "loss": 0.1674, + "step": 39177 + }, + { + "epoch": 0.9183940177101098, + "grad_norm": 2.4000039100646973, + "learning_rate": 3.2694842163503825e-06, + "loss": 1.7864, + "step": 39178 + }, + { + "epoch": 0.9184174592849148, + "grad_norm": 1.6832071542739868, + "learning_rate": 3.267616308291943e-06, + "loss": 0.074, + "step": 39179 + }, + { + "epoch": 0.9184409008597197, + "grad_norm": 2.9133169651031494, + "learning_rate": 3.265748925106138e-06, + "loss": 0.2745, + "step": 39180 + }, + { + "epoch": 0.9184643424345248, + "grad_norm": 2.7930502891540527, + "learning_rate": 3.2638820668031145e-06, + "loss": 0.1549, + "step": 39181 + }, + { + "epoch": 0.9184877840093297, + "grad_norm": 4.278416156768799, + "learning_rate": 3.262015733392976e-06, + "loss": 0.4607, + "step": 39182 + }, + { + "epoch": 0.9185112255841348, + "grad_norm": 2.4508721828460693, + "learning_rate": 3.260149924885869e-06, + "loss": 0.1178, + "step": 39183 + }, + { + "epoch": 0.9185346671589397, + "grad_norm": 1.9008713960647583, + "learning_rate": 3.2582846412918976e-06, + "loss": 0.1468, + "step": 39184 + }, + { + "epoch": 0.9185581087337448, + "grad_norm": 2.816131353378296, + "learning_rate": 3.2564198826211976e-06, + "loss": 0.3127, + "step": 39185 + }, + { + "epoch": 0.9185815503085497, + "grad_norm": 4.04551362991333, + "learning_rate": 3.2545556488838837e-06, + "loss": 0.4575, + "step": 39186 + }, + { + "epoch": 0.9186049918833548, + "grad_norm": 4.448280334472656, + "learning_rate": 3.252691940090069e-06, + "loss": 0.37, + "step": 39187 + }, + { + "epoch": 0.9186284334581597, + "grad_norm": 5.917234420776367, + "learning_rate": 3.2508287562498684e-06, + "loss": 0.8041, + "step": 39188 + }, + { + "epoch": 0.9186518750329647, + "grad_norm": 1.5756292343139648, + "learning_rate": 3.248966097373396e-06, + "loss": 0.1209, + "step": 39189 + }, + { + "epoch": 0.9186753166077697, + "grad_norm": 3.4766297340393066, + "learning_rate": 3.2471039634707545e-06, + "loss": 0.1787, + "step": 39190 + }, + { + "epoch": 0.9186987581825747, + "grad_norm": 1.5102849006652832, + "learning_rate": 3.245242354552036e-06, + "loss": 0.1956, + "step": 39191 + }, + { + "epoch": 0.9187221997573797, + "grad_norm": 2.687173843383789, + "learning_rate": 3.2433812706273547e-06, + "loss": 0.2123, + "step": 39192 + }, + { + "epoch": 0.9187456413321847, + "grad_norm": 3.1628811359405518, + "learning_rate": 3.2415207117068135e-06, + "loss": 0.1892, + "step": 39193 + }, + { + "epoch": 0.9187690829069897, + "grad_norm": 0.6540158987045288, + "learning_rate": 3.2396606778005046e-06, + "loss": 0.0646, + "step": 39194 + }, + { + "epoch": 0.9187925244817947, + "grad_norm": 4.115820407867432, + "learning_rate": 3.2378011689185193e-06, + "loss": 0.4513, + "step": 39195 + }, + { + "epoch": 0.9188159660565997, + "grad_norm": 2.7472193241119385, + "learning_rate": 3.2359421850709503e-06, + "loss": 0.1017, + "step": 39196 + }, + { + "epoch": 0.9188394076314047, + "grad_norm": 3.2374515533447266, + "learning_rate": 3.234083726267867e-06, + "loss": 0.3319, + "step": 39197 + }, + { + "epoch": 0.9188628492062096, + "grad_norm": 1.6982687711715698, + "learning_rate": 3.232225792519361e-06, + "loss": 0.0883, + "step": 39198 + }, + { + "epoch": 0.9188862907810147, + "grad_norm": 0.9259372353553772, + "learning_rate": 3.230368383835536e-06, + "loss": 0.1086, + "step": 39199 + }, + { + "epoch": 0.9189097323558196, + "grad_norm": 1.6528953313827515, + "learning_rate": 3.228511500226439e-06, + "loss": 0.0938, + "step": 39200 + }, + { + "epoch": 0.9189331739306247, + "grad_norm": 2.276782751083374, + "learning_rate": 3.2266551417021617e-06, + "loss": 0.195, + "step": 39201 + }, + { + "epoch": 0.9189566155054296, + "grad_norm": 1.1427780389785767, + "learning_rate": 3.2247993082727747e-06, + "loss": 0.1055, + "step": 39202 + }, + { + "epoch": 0.9189800570802347, + "grad_norm": 2.3270962238311768, + "learning_rate": 3.222943999948358e-06, + "loss": 0.1275, + "step": 39203 + }, + { + "epoch": 0.9190034986550396, + "grad_norm": 2.3557114601135254, + "learning_rate": 3.2210892167389594e-06, + "loss": 0.1925, + "step": 39204 + }, + { + "epoch": 0.9190269402298447, + "grad_norm": 3.1724696159362793, + "learning_rate": 3.219234958654649e-06, + "loss": 0.3532, + "step": 39205 + }, + { + "epoch": 0.9190503818046496, + "grad_norm": 3.436769962310791, + "learning_rate": 3.2173812257054957e-06, + "loss": 0.2864, + "step": 39206 + }, + { + "epoch": 0.9190738233794546, + "grad_norm": 4.783199310302734, + "learning_rate": 3.215528017901559e-06, + "loss": 0.5244, + "step": 39207 + }, + { + "epoch": 0.9190972649542596, + "grad_norm": 0.8901227116584778, + "learning_rate": 3.2136753352528857e-06, + "loss": 0.1195, + "step": 39208 + }, + { + "epoch": 0.9191207065290646, + "grad_norm": 0.9177272319793701, + "learning_rate": 3.211823177769524e-06, + "loss": 0.0787, + "step": 39209 + }, + { + "epoch": 0.9191441481038696, + "grad_norm": 0.5512257218360901, + "learning_rate": 3.209971545461532e-06, + "loss": 0.0336, + "step": 39210 + }, + { + "epoch": 0.9191675896786746, + "grad_norm": 1.8840312957763672, + "learning_rate": 3.2081204383389572e-06, + "loss": 0.152, + "step": 39211 + }, + { + "epoch": 0.9191910312534796, + "grad_norm": 2.4487249851226807, + "learning_rate": 3.2062698564118478e-06, + "loss": 1.2502, + "step": 39212 + }, + { + "epoch": 0.9192144728282846, + "grad_norm": 3.058448553085327, + "learning_rate": 3.204419799690228e-06, + "loss": 0.3042, + "step": 39213 + }, + { + "epoch": 0.9192379144030896, + "grad_norm": 2.709827184677124, + "learning_rate": 3.2025702681841574e-06, + "loss": 1.2105, + "step": 39214 + }, + { + "epoch": 0.9192613559778946, + "grad_norm": 2.55216383934021, + "learning_rate": 3.200721261903661e-06, + "loss": 0.1985, + "step": 39215 + }, + { + "epoch": 0.9192847975526995, + "grad_norm": 1.7988921403884888, + "learning_rate": 3.198872780858786e-06, + "loss": 0.1017, + "step": 39216 + }, + { + "epoch": 0.9193082391275046, + "grad_norm": 2.9181623458862305, + "learning_rate": 3.1970248250595246e-06, + "loss": 0.1492, + "step": 39217 + }, + { + "epoch": 0.9193316807023095, + "grad_norm": 0.8924022316932678, + "learning_rate": 3.1951773945159357e-06, + "loss": 0.0844, + "step": 39218 + }, + { + "epoch": 0.9193551222771146, + "grad_norm": 1.4815313816070557, + "learning_rate": 3.1933304892380446e-06, + "loss": 0.0862, + "step": 39219 + }, + { + "epoch": 0.9193785638519195, + "grad_norm": 1.0733180046081543, + "learning_rate": 3.191484109235865e-06, + "loss": 0.0885, + "step": 39220 + }, + { + "epoch": 0.9194020054267246, + "grad_norm": 2.471686840057373, + "learning_rate": 3.1896382545194116e-06, + "loss": 0.2918, + "step": 39221 + }, + { + "epoch": 0.9194254470015296, + "grad_norm": 3.0250484943389893, + "learning_rate": 3.1877929250987093e-06, + "loss": 0.1437, + "step": 39222 + }, + { + "epoch": 0.9194488885763346, + "grad_norm": 1.7919203042984009, + "learning_rate": 3.1859481209837393e-06, + "loss": 0.1699, + "step": 39223 + }, + { + "epoch": 0.9194723301511396, + "grad_norm": 0.7597114443778992, + "learning_rate": 3.1841038421845603e-06, + "loss": 0.0635, + "step": 39224 + }, + { + "epoch": 0.9194957717259445, + "grad_norm": 2.784792184829712, + "learning_rate": 3.182260088711142e-06, + "loss": 0.1754, + "step": 39225 + }, + { + "epoch": 0.9195192133007496, + "grad_norm": 2.186768054962158, + "learning_rate": 3.1804168605735097e-06, + "loss": 0.1677, + "step": 39226 + }, + { + "epoch": 0.9195426548755545, + "grad_norm": 3.055298328399658, + "learning_rate": 3.1785741577816663e-06, + "loss": 0.3102, + "step": 39227 + }, + { + "epoch": 0.9195660964503596, + "grad_norm": 3.487501382827759, + "learning_rate": 3.1767319803455932e-06, + "loss": 0.2949, + "step": 39228 + }, + { + "epoch": 0.9195895380251645, + "grad_norm": 3.0724375247955322, + "learning_rate": 3.174890328275304e-06, + "loss": 0.2055, + "step": 39229 + }, + { + "epoch": 0.9196129795999696, + "grad_norm": 5.62855339050293, + "learning_rate": 3.173049201580769e-06, + "loss": 0.2399, + "step": 39230 + }, + { + "epoch": 0.9196364211747745, + "grad_norm": 0.6539332866668701, + "learning_rate": 3.171208600271991e-06, + "loss": 0.0522, + "step": 39231 + }, + { + "epoch": 0.9196598627495796, + "grad_norm": 5.464206218719482, + "learning_rate": 3.1693685243589733e-06, + "loss": 0.5159, + "step": 39232 + }, + { + "epoch": 0.9196833043243845, + "grad_norm": 2.6260101795196533, + "learning_rate": 3.1675289738516744e-06, + "loss": 0.1592, + "step": 39233 + }, + { + "epoch": 0.9197067458991895, + "grad_norm": 0.5251327157020569, + "learning_rate": 3.1656899487600867e-06, + "loss": 0.0622, + "step": 39234 + }, + { + "epoch": 0.9197301874739945, + "grad_norm": 2.478908061981201, + "learning_rate": 3.1638514490941908e-06, + "loss": 0.1667, + "step": 39235 + }, + { + "epoch": 0.9197536290487995, + "grad_norm": 4.04256534576416, + "learning_rate": 3.162013474863945e-06, + "loss": 0.365, + "step": 39236 + }, + { + "epoch": 0.9197770706236045, + "grad_norm": 2.681201696395874, + "learning_rate": 3.1601760260793423e-06, + "loss": 0.1826, + "step": 39237 + }, + { + "epoch": 0.9198005121984095, + "grad_norm": 0.4457859694957733, + "learning_rate": 3.158339102750363e-06, + "loss": 0.0286, + "step": 39238 + }, + { + "epoch": 0.9198239537732145, + "grad_norm": 2.0092227458953857, + "learning_rate": 3.1565027048869543e-06, + "loss": 0.2219, + "step": 39239 + }, + { + "epoch": 0.9198473953480195, + "grad_norm": 0.8404163122177124, + "learning_rate": 3.1546668324990756e-06, + "loss": 0.0732, + "step": 39240 + }, + { + "epoch": 0.9198708369228245, + "grad_norm": 1.9354944229125977, + "learning_rate": 3.152831485596708e-06, + "loss": 0.1375, + "step": 39241 + }, + { + "epoch": 0.9198942784976295, + "grad_norm": 3.2020983695983887, + "learning_rate": 3.1509966641897982e-06, + "loss": 0.2202, + "step": 39242 + }, + { + "epoch": 0.9199177200724344, + "grad_norm": 1.4292926788330078, + "learning_rate": 3.149162368288294e-06, + "loss": 0.0541, + "step": 39243 + }, + { + "epoch": 0.9199411616472395, + "grad_norm": 0.7576732635498047, + "learning_rate": 3.147328597902166e-06, + "loss": 0.0537, + "step": 39244 + }, + { + "epoch": 0.9199646032220444, + "grad_norm": 1.4307823181152344, + "learning_rate": 3.1454953530413502e-06, + "loss": 0.1376, + "step": 39245 + }, + { + "epoch": 0.9199880447968495, + "grad_norm": 2.4803128242492676, + "learning_rate": 3.1436626337158046e-06, + "loss": 0.1687, + "step": 39246 + }, + { + "epoch": 0.9200114863716544, + "grad_norm": 2.2638492584228516, + "learning_rate": 3.141830439935478e-06, + "loss": 0.1211, + "step": 39247 + }, + { + "epoch": 0.9200349279464595, + "grad_norm": 0.7866616249084473, + "learning_rate": 3.139998771710284e-06, + "loss": 0.0489, + "step": 39248 + }, + { + "epoch": 0.9200583695212644, + "grad_norm": 2.7468137741088867, + "learning_rate": 3.1381676290501927e-06, + "loss": 0.3194, + "step": 39249 + }, + { + "epoch": 0.9200818110960695, + "grad_norm": 2.310957431793213, + "learning_rate": 3.136337011965107e-06, + "loss": 1.3776, + "step": 39250 + }, + { + "epoch": 0.9201052526708744, + "grad_norm": 1.0347620248794556, + "learning_rate": 3.134506920465008e-06, + "loss": 0.0381, + "step": 39251 + }, + { + "epoch": 0.9201286942456794, + "grad_norm": 2.530116319656372, + "learning_rate": 3.1326773545597875e-06, + "loss": 0.1584, + "step": 39252 + }, + { + "epoch": 0.9201521358204844, + "grad_norm": 3.1009387969970703, + "learning_rate": 3.1308483142593825e-06, + "loss": 0.2531, + "step": 39253 + }, + { + "epoch": 0.9201755773952894, + "grad_norm": 1.019374966621399, + "learning_rate": 3.1290197995737182e-06, + "loss": 0.0546, + "step": 39254 + }, + { + "epoch": 0.9201990189700944, + "grad_norm": 0.536970853805542, + "learning_rate": 3.1271918105127087e-06, + "loss": 0.0414, + "step": 39255 + }, + { + "epoch": 0.9202224605448994, + "grad_norm": 0.2617732584476471, + "learning_rate": 3.1253643470862903e-06, + "loss": 0.0279, + "step": 39256 + }, + { + "epoch": 0.9202459021197044, + "grad_norm": 2.996068239212036, + "learning_rate": 3.1235374093043668e-06, + "loss": 0.2249, + "step": 39257 + }, + { + "epoch": 0.9202693436945094, + "grad_norm": 2.561835765838623, + "learning_rate": 3.121710997176863e-06, + "loss": 0.245, + "step": 39258 + }, + { + "epoch": 0.9202927852693144, + "grad_norm": 3.900677442550659, + "learning_rate": 3.119885110713672e-06, + "loss": 0.4169, + "step": 39259 + }, + { + "epoch": 0.9203162268441194, + "grad_norm": 0.7910061478614807, + "learning_rate": 3.1180597499247066e-06, + "loss": 0.0561, + "step": 39260 + }, + { + "epoch": 0.9203396684189243, + "grad_norm": 2.541633367538452, + "learning_rate": 3.11623491481986e-06, + "loss": 0.2591, + "step": 39261 + }, + { + "epoch": 0.9203631099937294, + "grad_norm": 1.8960002660751343, + "learning_rate": 3.1144106054090572e-06, + "loss": 0.1411, + "step": 39262 + }, + { + "epoch": 0.9203865515685343, + "grad_norm": 1.1101595163345337, + "learning_rate": 3.1125868217021903e-06, + "loss": 0.0974, + "step": 39263 + }, + { + "epoch": 0.9204099931433394, + "grad_norm": 0.3573494851589203, + "learning_rate": 3.1107635637091517e-06, + "loss": 0.0171, + "step": 39264 + }, + { + "epoch": 0.9204334347181443, + "grad_norm": 4.871189594268799, + "learning_rate": 3.108940831439833e-06, + "loss": 0.4936, + "step": 39265 + }, + { + "epoch": 0.9204568762929494, + "grad_norm": 2.8436570167541504, + "learning_rate": 3.107118624904126e-06, + "loss": 0.1422, + "step": 39266 + }, + { + "epoch": 0.9204803178677543, + "grad_norm": 1.089993953704834, + "learning_rate": 3.105296944111924e-06, + "loss": 0.1639, + "step": 39267 + }, + { + "epoch": 0.9205037594425594, + "grad_norm": 0.6261327266693115, + "learning_rate": 3.1034757890730847e-06, + "loss": 0.0336, + "step": 39268 + }, + { + "epoch": 0.9205272010173643, + "grad_norm": 2.568983554840088, + "learning_rate": 3.101655159797534e-06, + "loss": 0.2199, + "step": 39269 + }, + { + "epoch": 0.9205506425921693, + "grad_norm": 0.710443913936615, + "learning_rate": 3.0998350562951085e-06, + "loss": 0.0406, + "step": 39270 + }, + { + "epoch": 0.9205740841669743, + "grad_norm": 0.9347928166389465, + "learning_rate": 3.098015478575722e-06, + "loss": 0.0839, + "step": 39271 + }, + { + "epoch": 0.9205975257417793, + "grad_norm": 2.723867654800415, + "learning_rate": 3.0961964266492114e-06, + "loss": 0.2028, + "step": 39272 + }, + { + "epoch": 0.9206209673165844, + "grad_norm": 2.416998863220215, + "learning_rate": 3.0943779005254805e-06, + "loss": 0.2722, + "step": 39273 + }, + { + "epoch": 0.9206444088913893, + "grad_norm": 3.506319999694824, + "learning_rate": 3.0925599002143536e-06, + "loss": 0.3126, + "step": 39274 + }, + { + "epoch": 0.9206678504661944, + "grad_norm": 2.981764316558838, + "learning_rate": 3.0907424257257344e-06, + "loss": 0.1909, + "step": 39275 + }, + { + "epoch": 0.9206912920409993, + "grad_norm": 2.34696102142334, + "learning_rate": 3.088925477069471e-06, + "loss": 0.2702, + "step": 39276 + }, + { + "epoch": 0.9207147336158044, + "grad_norm": 4.99330472946167, + "learning_rate": 3.0871090542554326e-06, + "loss": 0.267, + "step": 39277 + }, + { + "epoch": 0.9207381751906093, + "grad_norm": 3.794267177581787, + "learning_rate": 3.0852931572934562e-06, + "loss": 0.2517, + "step": 39278 + }, + { + "epoch": 0.9207616167654143, + "grad_norm": 0.850975513458252, + "learning_rate": 3.083477786193412e-06, + "loss": 0.0533, + "step": 39279 + }, + { + "epoch": 0.9207850583402193, + "grad_norm": 3.4755117893218994, + "learning_rate": 3.0816629409651355e-06, + "loss": 0.2086, + "step": 39280 + }, + { + "epoch": 0.9208084999150243, + "grad_norm": 1.1721808910369873, + "learning_rate": 3.0798486216184753e-06, + "loss": 0.0543, + "step": 39281 + }, + { + "epoch": 0.9208319414898293, + "grad_norm": 2.1554760932922363, + "learning_rate": 3.07803482816329e-06, + "loss": 1.4915, + "step": 39282 + }, + { + "epoch": 0.9208553830646343, + "grad_norm": 4.690199851989746, + "learning_rate": 3.076221560609416e-06, + "loss": 0.3978, + "step": 39283 + }, + { + "epoch": 0.9208788246394393, + "grad_norm": 1.0731868743896484, + "learning_rate": 3.0744088189666897e-06, + "loss": 0.0322, + "step": 39284 + }, + { + "epoch": 0.9209022662142443, + "grad_norm": 1.450925588607788, + "learning_rate": 3.072596603244937e-06, + "loss": 0.0894, + "step": 39285 + }, + { + "epoch": 0.9209257077890493, + "grad_norm": 2.0637333393096924, + "learning_rate": 3.0707849134539945e-06, + "loss": 0.2562, + "step": 39286 + }, + { + "epoch": 0.9209491493638543, + "grad_norm": 1.0352991819381714, + "learning_rate": 3.0689737496037208e-06, + "loss": 0.0849, + "step": 39287 + }, + { + "epoch": 0.9209725909386592, + "grad_norm": 3.277815818786621, + "learning_rate": 3.067163111703897e-06, + "loss": 0.3141, + "step": 39288 + }, + { + "epoch": 0.9209960325134643, + "grad_norm": 1.025424599647522, + "learning_rate": 3.0653529997643814e-06, + "loss": 0.1158, + "step": 39289 + }, + { + "epoch": 0.9210194740882692, + "grad_norm": 3.187433958053589, + "learning_rate": 3.0635434137949894e-06, + "loss": 0.1575, + "step": 39290 + }, + { + "epoch": 0.9210429156630743, + "grad_norm": 1.0899677276611328, + "learning_rate": 3.0617343538055455e-06, + "loss": 0.0636, + "step": 39291 + }, + { + "epoch": 0.9210663572378792, + "grad_norm": 2.704421043395996, + "learning_rate": 3.0599258198058425e-06, + "loss": 1.3451, + "step": 39292 + }, + { + "epoch": 0.9210897988126843, + "grad_norm": 1.4312574863433838, + "learning_rate": 3.0581178118057053e-06, + "loss": 1.5765, + "step": 39293 + }, + { + "epoch": 0.9211132403874892, + "grad_norm": 1.1393743753433228, + "learning_rate": 3.05631032981496e-06, + "loss": 0.1084, + "step": 39294 + }, + { + "epoch": 0.9211366819622943, + "grad_norm": 4.355836391448975, + "learning_rate": 3.054503373843387e-06, + "loss": 0.3906, + "step": 39295 + }, + { + "epoch": 0.9211601235370992, + "grad_norm": 1.278630018234253, + "learning_rate": 3.0526969439008124e-06, + "loss": 0.1035, + "step": 39296 + }, + { + "epoch": 0.9211835651119042, + "grad_norm": 3.5590829849243164, + "learning_rate": 3.050891039997039e-06, + "loss": 0.2934, + "step": 39297 + }, + { + "epoch": 0.9212070066867092, + "grad_norm": 3.3113343715667725, + "learning_rate": 3.0490856621418374e-06, + "loss": 0.1782, + "step": 39298 + }, + { + "epoch": 0.9212304482615142, + "grad_norm": 2.889937400817871, + "learning_rate": 3.047280810345032e-06, + "loss": 0.2971, + "step": 39299 + }, + { + "epoch": 0.9212538898363192, + "grad_norm": 1.9336459636688232, + "learning_rate": 3.045476484616405e-06, + "loss": 0.2344, + "step": 39300 + }, + { + "epoch": 0.9212773314111242, + "grad_norm": 2.2118823528289795, + "learning_rate": 3.043672684965748e-06, + "loss": 0.1705, + "step": 39301 + }, + { + "epoch": 0.9213007729859292, + "grad_norm": 0.9700807929039001, + "learning_rate": 3.0418694114028533e-06, + "loss": 0.0675, + "step": 39302 + }, + { + "epoch": 0.9213242145607342, + "grad_norm": 3.0962820053100586, + "learning_rate": 3.0400666639375018e-06, + "loss": 0.2679, + "step": 39303 + }, + { + "epoch": 0.9213476561355392, + "grad_norm": 0.9640467762947083, + "learning_rate": 3.038264442579475e-06, + "loss": 0.0775, + "step": 39304 + }, + { + "epoch": 0.9213710977103442, + "grad_norm": 3.3405072689056396, + "learning_rate": 3.0364627473385533e-06, + "loss": 0.2012, + "step": 39305 + }, + { + "epoch": 0.9213945392851491, + "grad_norm": 2.2980353832244873, + "learning_rate": 3.034661578224507e-06, + "loss": 1.1288, + "step": 39306 + }, + { + "epoch": 0.9214179808599542, + "grad_norm": 2.8282852172851562, + "learning_rate": 3.032860935247106e-06, + "loss": 0.2676, + "step": 39307 + }, + { + "epoch": 0.9214414224347591, + "grad_norm": 1.1275126934051514, + "learning_rate": 3.0310608184161427e-06, + "loss": 0.1045, + "step": 39308 + }, + { + "epoch": 0.9214648640095642, + "grad_norm": 0.7847543358802795, + "learning_rate": 3.0292612277413648e-06, + "loss": 0.0214, + "step": 39309 + }, + { + "epoch": 0.9214883055843691, + "grad_norm": 2.3683857917785645, + "learning_rate": 3.0274621632325308e-06, + "loss": 0.0961, + "step": 39310 + }, + { + "epoch": 0.9215117471591742, + "grad_norm": 0.5263874530792236, + "learning_rate": 3.025663624899433e-06, + "loss": 0.0336, + "step": 39311 + }, + { + "epoch": 0.9215351887339791, + "grad_norm": 3.9931702613830566, + "learning_rate": 3.0238656127517974e-06, + "loss": 0.3143, + "step": 39312 + }, + { + "epoch": 0.9215586303087842, + "grad_norm": 2.050257682800293, + "learning_rate": 3.0220681267993933e-06, + "loss": 0.0994, + "step": 39313 + }, + { + "epoch": 0.9215820718835891, + "grad_norm": 4.53142786026001, + "learning_rate": 3.02027116705198e-06, + "loss": 0.2595, + "step": 39314 + }, + { + "epoch": 0.9216055134583941, + "grad_norm": 3.2885000705718994, + "learning_rate": 3.0184747335193054e-06, + "loss": 1.8477, + "step": 39315 + }, + { + "epoch": 0.9216289550331991, + "grad_norm": 2.5275514125823975, + "learning_rate": 3.0166788262111055e-06, + "loss": 0.1521, + "step": 39316 + }, + { + "epoch": 0.9216523966080041, + "grad_norm": 1.6385334730148315, + "learning_rate": 3.01488344513714e-06, + "loss": 1.5787, + "step": 39317 + }, + { + "epoch": 0.9216758381828091, + "grad_norm": 1.5441899299621582, + "learning_rate": 3.013088590307145e-06, + "loss": 0.078, + "step": 39318 + }, + { + "epoch": 0.9216992797576141, + "grad_norm": 0.3677608072757721, + "learning_rate": 3.011294261730846e-06, + "loss": 0.0341, + "step": 39319 + }, + { + "epoch": 0.9217227213324191, + "grad_norm": 2.3111743927001953, + "learning_rate": 3.009500459418002e-06, + "loss": 0.159, + "step": 39320 + }, + { + "epoch": 0.9217461629072241, + "grad_norm": 2.8227744102478027, + "learning_rate": 3.0077071833783386e-06, + "loss": 0.2495, + "step": 39321 + }, + { + "epoch": 0.921769604482029, + "grad_norm": 1.1276174783706665, + "learning_rate": 3.0059144336215706e-06, + "loss": 0.1155, + "step": 39322 + }, + { + "epoch": 0.9217930460568341, + "grad_norm": 1.5964604616165161, + "learning_rate": 3.0041222101574453e-06, + "loss": 0.1039, + "step": 39323 + }, + { + "epoch": 0.9218164876316391, + "grad_norm": 2.2983386516571045, + "learning_rate": 3.002330512995688e-06, + "loss": 0.1939, + "step": 39324 + }, + { + "epoch": 0.9218399292064441, + "grad_norm": 0.4228995144367218, + "learning_rate": 3.0005393421460027e-06, + "loss": 0.0384, + "step": 39325 + }, + { + "epoch": 0.9218633707812491, + "grad_norm": 2.98587965965271, + "learning_rate": 2.9987486976181144e-06, + "loss": 0.095, + "step": 39326 + }, + { + "epoch": 0.9218868123560541, + "grad_norm": 0.6918138265609741, + "learning_rate": 2.996958579421749e-06, + "loss": 0.0394, + "step": 39327 + }, + { + "epoch": 0.9219102539308591, + "grad_norm": 3.4519007205963135, + "learning_rate": 2.995168987566621e-06, + "loss": 0.3188, + "step": 39328 + }, + { + "epoch": 0.9219336955056641, + "grad_norm": 3.184199571609497, + "learning_rate": 2.9933799220624338e-06, + "loss": 0.2172, + "step": 39329 + }, + { + "epoch": 0.9219571370804691, + "grad_norm": 2.380411148071289, + "learning_rate": 2.9915913829188903e-06, + "loss": 0.1418, + "step": 39330 + }, + { + "epoch": 0.921980578655274, + "grad_norm": 1.8609111309051514, + "learning_rate": 2.9898033701456828e-06, + "loss": 0.0964, + "step": 39331 + }, + { + "epoch": 0.9220040202300791, + "grad_norm": 0.6308184862136841, + "learning_rate": 2.9880158837525487e-06, + "loss": 0.0461, + "step": 39332 + }, + { + "epoch": 0.922027461804884, + "grad_norm": 3.3402373790740967, + "learning_rate": 2.986228923749168e-06, + "loss": 0.2853, + "step": 39333 + }, + { + "epoch": 0.9220509033796891, + "grad_norm": 2.9713475704193115, + "learning_rate": 2.984442490145223e-06, + "loss": 0.3753, + "step": 39334 + }, + { + "epoch": 0.922074344954494, + "grad_norm": 0.7037737965583801, + "learning_rate": 2.982656582950427e-06, + "loss": 0.0419, + "step": 39335 + }, + { + "epoch": 0.9220977865292991, + "grad_norm": 1.7397884130477905, + "learning_rate": 2.980871202174473e-06, + "loss": 0.1689, + "step": 39336 + }, + { + "epoch": 0.922121228104104, + "grad_norm": 2.5232574939727783, + "learning_rate": 2.9790863478270427e-06, + "loss": 0.1911, + "step": 39337 + }, + { + "epoch": 0.9221446696789091, + "grad_norm": 3.2678258419036865, + "learning_rate": 2.9773020199177937e-06, + "loss": 0.2386, + "step": 39338 + }, + { + "epoch": 0.922168111253714, + "grad_norm": 2.1910722255706787, + "learning_rate": 2.9755182184564524e-06, + "loss": 0.1206, + "step": 39339 + }, + { + "epoch": 0.9221915528285191, + "grad_norm": 1.6723151206970215, + "learning_rate": 2.9737349434526775e-06, + "loss": 0.0972, + "step": 39340 + }, + { + "epoch": 0.922214994403324, + "grad_norm": 2.7748894691467285, + "learning_rate": 2.9719521949161387e-06, + "loss": 0.1024, + "step": 39341 + }, + { + "epoch": 0.922238435978129, + "grad_norm": 3.2124743461608887, + "learning_rate": 2.970169972856529e-06, + "loss": 0.1706, + "step": 39342 + }, + { + "epoch": 0.922261877552934, + "grad_norm": 1.8569328784942627, + "learning_rate": 2.968388277283496e-06, + "loss": 0.1826, + "step": 39343 + }, + { + "epoch": 0.922285319127739, + "grad_norm": 1.6550391912460327, + "learning_rate": 2.966607108206698e-06, + "loss": 0.112, + "step": 39344 + }, + { + "epoch": 0.922308760702544, + "grad_norm": 5.981138706207275, + "learning_rate": 2.9648264656358394e-06, + "loss": 0.3433, + "step": 39345 + }, + { + "epoch": 0.922332202277349, + "grad_norm": 0.6389743089675903, + "learning_rate": 2.963046349580556e-06, + "loss": 0.062, + "step": 39346 + }, + { + "epoch": 0.922355643852154, + "grad_norm": 1.0245486497879028, + "learning_rate": 2.9612667600505073e-06, + "loss": 0.1489, + "step": 39347 + }, + { + "epoch": 0.922379085426959, + "grad_norm": 2.3632588386535645, + "learning_rate": 2.959487697055352e-06, + "loss": 0.1557, + "step": 39348 + }, + { + "epoch": 0.922402527001764, + "grad_norm": 1.558372974395752, + "learning_rate": 2.95770916060476e-06, + "loss": 0.1947, + "step": 39349 + }, + { + "epoch": 0.922425968576569, + "grad_norm": 1.7319434881210327, + "learning_rate": 2.9559311507083576e-06, + "loss": 0.0903, + "step": 39350 + }, + { + "epoch": 0.9224494101513739, + "grad_norm": 0.3804222047328949, + "learning_rate": 2.954153667375792e-06, + "loss": 0.0376, + "step": 39351 + }, + { + "epoch": 0.922472851726179, + "grad_norm": 2.392606735229492, + "learning_rate": 2.952376710616733e-06, + "loss": 0.0995, + "step": 39352 + }, + { + "epoch": 0.9224962933009839, + "grad_norm": 1.333629846572876, + "learning_rate": 2.9506002804407963e-06, + "loss": 1.1563, + "step": 39353 + }, + { + "epoch": 0.922519734875789, + "grad_norm": 2.513888359069824, + "learning_rate": 2.9488243768576394e-06, + "loss": 0.1644, + "step": 39354 + }, + { + "epoch": 0.9225431764505939, + "grad_norm": 3.000990152359009, + "learning_rate": 2.947048999876889e-06, + "loss": 1.5318, + "step": 39355 + }, + { + "epoch": 0.922566618025399, + "grad_norm": 3.003213405609131, + "learning_rate": 2.9452741495081814e-06, + "loss": 0.3341, + "step": 39356 + }, + { + "epoch": 0.9225900596002039, + "grad_norm": 3.3086462020874023, + "learning_rate": 2.943499825761131e-06, + "loss": 0.232, + "step": 39357 + }, + { + "epoch": 0.922613501175009, + "grad_norm": 2.6969962120056152, + "learning_rate": 2.9417260286453863e-06, + "loss": 0.1622, + "step": 39358 + }, + { + "epoch": 0.9226369427498139, + "grad_norm": 1.1288470029830933, + "learning_rate": 2.939952758170561e-06, + "loss": 0.1145, + "step": 39359 + }, + { + "epoch": 0.9226603843246189, + "grad_norm": 0.7814863324165344, + "learning_rate": 2.9381800143462923e-06, + "loss": 0.0734, + "step": 39360 + }, + { + "epoch": 0.9226838258994239, + "grad_norm": 0.36269742250442505, + "learning_rate": 2.9364077971821834e-06, + "loss": 0.0356, + "step": 39361 + }, + { + "epoch": 0.9227072674742289, + "grad_norm": 1.500022530555725, + "learning_rate": 2.9346361066878604e-06, + "loss": 0.0936, + "step": 39362 + }, + { + "epoch": 0.9227307090490339, + "grad_norm": 2.356868267059326, + "learning_rate": 2.932864942872926e-06, + "loss": 0.1322, + "step": 39363 + }, + { + "epoch": 0.9227541506238389, + "grad_norm": 2.961160898208618, + "learning_rate": 2.931094305746984e-06, + "loss": 0.2171, + "step": 39364 + }, + { + "epoch": 0.9227775921986439, + "grad_norm": 3.5381388664245605, + "learning_rate": 2.9293241953196713e-06, + "loss": 0.177, + "step": 39365 + }, + { + "epoch": 0.9228010337734489, + "grad_norm": 3.1916682720184326, + "learning_rate": 2.927554611600569e-06, + "loss": 0.1689, + "step": 39366 + }, + { + "epoch": 0.9228244753482538, + "grad_norm": 2.905428647994995, + "learning_rate": 2.925785554599292e-06, + "loss": 0.1429, + "step": 39367 + }, + { + "epoch": 0.9228479169230589, + "grad_norm": 0.5786945819854736, + "learning_rate": 2.9240170243254205e-06, + "loss": 0.0293, + "step": 39368 + }, + { + "epoch": 0.9228713584978638, + "grad_norm": 2.2425193786621094, + "learning_rate": 2.922249020788559e-06, + "loss": 0.1683, + "step": 39369 + }, + { + "epoch": 0.9228948000726689, + "grad_norm": 5.370898723602295, + "learning_rate": 2.9204815439983104e-06, + "loss": 1.1874, + "step": 39370 + }, + { + "epoch": 0.9229182416474738, + "grad_norm": 2.806389093399048, + "learning_rate": 2.9187145939642555e-06, + "loss": 0.1694, + "step": 39371 + }, + { + "epoch": 0.9229416832222789, + "grad_norm": 0.35179632902145386, + "learning_rate": 2.916948170695988e-06, + "loss": 0.0414, + "step": 39372 + }, + { + "epoch": 0.9229651247970838, + "grad_norm": 1.690155029296875, + "learning_rate": 2.915182274203099e-06, + "loss": 0.1114, + "step": 39373 + }, + { + "epoch": 0.9229885663718889, + "grad_norm": 0.4940837323665619, + "learning_rate": 2.9134169044951478e-06, + "loss": 0.0368, + "step": 39374 + }, + { + "epoch": 0.9230120079466939, + "grad_norm": 2.6173980236053467, + "learning_rate": 2.9116520615817376e-06, + "loss": 0.3132, + "step": 39375 + }, + { + "epoch": 0.9230354495214989, + "grad_norm": 3.292769193649292, + "learning_rate": 2.909887745472417e-06, + "loss": 0.2335, + "step": 39376 + }, + { + "epoch": 0.9230588910963039, + "grad_norm": 0.9546391367912292, + "learning_rate": 2.908123956176789e-06, + "loss": 0.1093, + "step": 39377 + }, + { + "epoch": 0.9230823326711088, + "grad_norm": 1.8513681888580322, + "learning_rate": 2.9063606937044018e-06, + "loss": 0.1829, + "step": 39378 + }, + { + "epoch": 0.9231057742459139, + "grad_norm": 2.630463123321533, + "learning_rate": 2.9045979580648362e-06, + "loss": 0.2295, + "step": 39379 + }, + { + "epoch": 0.9231292158207188, + "grad_norm": 1.262466311454773, + "learning_rate": 2.9028357492676628e-06, + "loss": 0.1427, + "step": 39380 + }, + { + "epoch": 0.9231526573955239, + "grad_norm": 2.08418607711792, + "learning_rate": 2.901074067322418e-06, + "loss": 0.1559, + "step": 39381 + }, + { + "epoch": 0.9231760989703288, + "grad_norm": 2.913208484649658, + "learning_rate": 2.899312912238672e-06, + "loss": 0.1352, + "step": 39382 + }, + { + "epoch": 0.9231995405451339, + "grad_norm": 0.6515222787857056, + "learning_rate": 2.8975522840259837e-06, + "loss": 0.06, + "step": 39383 + }, + { + "epoch": 0.9232229821199388, + "grad_norm": 4.1781697273254395, + "learning_rate": 2.8957921826939128e-06, + "loss": 0.4417, + "step": 39384 + }, + { + "epoch": 0.9232464236947439, + "grad_norm": 1.4755017757415771, + "learning_rate": 2.8940326082520176e-06, + "loss": 0.1254, + "step": 39385 + }, + { + "epoch": 0.9232698652695488, + "grad_norm": 1.0011109113693237, + "learning_rate": 2.892273560709813e-06, + "loss": 0.4077, + "step": 39386 + }, + { + "epoch": 0.9232933068443538, + "grad_norm": 4.000321388244629, + "learning_rate": 2.8905150400768797e-06, + "loss": 0.1915, + "step": 39387 + }, + { + "epoch": 0.9233167484191588, + "grad_norm": 3.118929386138916, + "learning_rate": 2.8887570463627334e-06, + "loss": 0.2145, + "step": 39388 + }, + { + "epoch": 0.9233401899939638, + "grad_norm": 1.1574580669403076, + "learning_rate": 2.8869995795769102e-06, + "loss": 0.0386, + "step": 39389 + }, + { + "epoch": 0.9233636315687688, + "grad_norm": 2.2721400260925293, + "learning_rate": 2.885242639728969e-06, + "loss": 0.2404, + "step": 39390 + }, + { + "epoch": 0.9233870731435738, + "grad_norm": 2.11661958694458, + "learning_rate": 2.8834862268284357e-06, + "loss": 0.1398, + "step": 39391 + }, + { + "epoch": 0.9234105147183788, + "grad_norm": 3.7835278511047363, + "learning_rate": 2.8817303408848363e-06, + "loss": 0.2825, + "step": 39392 + }, + { + "epoch": 0.9234339562931838, + "grad_norm": 2.1233694553375244, + "learning_rate": 2.879974981907696e-06, + "loss": 0.1633, + "step": 39393 + }, + { + "epoch": 0.9234573978679887, + "grad_norm": 0.690980076789856, + "learning_rate": 2.8782201499065408e-06, + "loss": 0.0615, + "step": 39394 + }, + { + "epoch": 0.9234808394427938, + "grad_norm": 2.6102025508880615, + "learning_rate": 2.8764658448908856e-06, + "loss": 0.1643, + "step": 39395 + }, + { + "epoch": 0.9235042810175987, + "grad_norm": 1.8296996355056763, + "learning_rate": 2.8747120668702553e-06, + "loss": 0.1802, + "step": 39396 + }, + { + "epoch": 0.9235277225924038, + "grad_norm": 2.5675694942474365, + "learning_rate": 2.8729588158541875e-06, + "loss": 0.3204, + "step": 39397 + }, + { + "epoch": 0.9235511641672087, + "grad_norm": 2.281447649002075, + "learning_rate": 2.8712060918521633e-06, + "loss": 0.1557, + "step": 39398 + }, + { + "epoch": 0.9235746057420138, + "grad_norm": 2.499077796936035, + "learning_rate": 2.869453894873719e-06, + "loss": 0.1321, + "step": 39399 + }, + { + "epoch": 0.9235980473168187, + "grad_norm": 5.266241550445557, + "learning_rate": 2.8677022249283368e-06, + "loss": 0.23, + "step": 39400 + }, + { + "epoch": 0.9236214888916238, + "grad_norm": 0.5294190645217896, + "learning_rate": 2.865951082025542e-06, + "loss": 0.067, + "step": 39401 + }, + { + "epoch": 0.9236449304664287, + "grad_norm": 0.7309638261795044, + "learning_rate": 2.8642004661748156e-06, + "loss": 0.075, + "step": 39402 + }, + { + "epoch": 0.9236683720412338, + "grad_norm": 1.5153629779815674, + "learning_rate": 2.8624503773856727e-06, + "loss": 0.1575, + "step": 39403 + }, + { + "epoch": 0.9236918136160387, + "grad_norm": 1.2136095762252808, + "learning_rate": 2.860700815667616e-06, + "loss": 0.1158, + "step": 39404 + }, + { + "epoch": 0.9237152551908437, + "grad_norm": 2.0362162590026855, + "learning_rate": 2.858951781030117e-06, + "loss": 0.1458, + "step": 39405 + }, + { + "epoch": 0.9237386967656487, + "grad_norm": 1.9140944480895996, + "learning_rate": 2.857203273482689e-06, + "loss": 0.1518, + "step": 39406 + }, + { + "epoch": 0.9237621383404537, + "grad_norm": 3.201185941696167, + "learning_rate": 2.8554552930347812e-06, + "loss": 0.1765, + "step": 39407 + }, + { + "epoch": 0.9237855799152587, + "grad_norm": 1.6289118528366089, + "learning_rate": 2.853707839695918e-06, + "loss": 0.1251, + "step": 39408 + }, + { + "epoch": 0.9238090214900637, + "grad_norm": 2.424572467803955, + "learning_rate": 2.851960913475571e-06, + "loss": 0.296, + "step": 39409 + }, + { + "epoch": 0.9238324630648687, + "grad_norm": 0.45423296093940735, + "learning_rate": 2.8502145143832203e-06, + "loss": 0.0262, + "step": 39410 + }, + { + "epoch": 0.9238559046396737, + "grad_norm": 0.7045748233795166, + "learning_rate": 2.848468642428337e-06, + "loss": 0.0555, + "step": 39411 + }, + { + "epoch": 0.9238793462144786, + "grad_norm": 12.254233360290527, + "learning_rate": 2.8467232976203906e-06, + "loss": 0.1525, + "step": 39412 + }, + { + "epoch": 0.9239027877892837, + "grad_norm": 3.2355263233184814, + "learning_rate": 2.844978479968863e-06, + "loss": 0.2346, + "step": 39413 + }, + { + "epoch": 0.9239262293640886, + "grad_norm": 0.3651764690876007, + "learning_rate": 2.8432341894831903e-06, + "loss": 0.0444, + "step": 39414 + }, + { + "epoch": 0.9239496709388937, + "grad_norm": 1.5578137636184692, + "learning_rate": 2.841490426172877e-06, + "loss": 0.0521, + "step": 39415 + }, + { + "epoch": 0.9239731125136986, + "grad_norm": 2.0697460174560547, + "learning_rate": 2.8397471900473703e-06, + "loss": 1.2344, + "step": 39416 + }, + { + "epoch": 0.9239965540885037, + "grad_norm": 0.6622860431671143, + "learning_rate": 2.8380044811161297e-06, + "loss": 0.0508, + "step": 39417 + }, + { + "epoch": 0.9240199956633086, + "grad_norm": 3.129603862762451, + "learning_rate": 2.8362622993886145e-06, + "loss": 0.2023, + "step": 39418 + }, + { + "epoch": 0.9240434372381137, + "grad_norm": 0.23364728689193726, + "learning_rate": 2.83452064487425e-06, + "loss": 0.0178, + "step": 39419 + }, + { + "epoch": 0.9240668788129186, + "grad_norm": 2.249401092529297, + "learning_rate": 2.8327795175825176e-06, + "loss": 1.6887, + "step": 39420 + }, + { + "epoch": 0.9240903203877237, + "grad_norm": 2.8451602458953857, + "learning_rate": 2.8310389175228545e-06, + "loss": 1.512, + "step": 39421 + }, + { + "epoch": 0.9241137619625286, + "grad_norm": 3.0990946292877197, + "learning_rate": 2.8292988447047088e-06, + "loss": 0.3485, + "step": 39422 + }, + { + "epoch": 0.9241372035373336, + "grad_norm": 2.9976613521575928, + "learning_rate": 2.8275592991375278e-06, + "loss": 0.1745, + "step": 39423 + }, + { + "epoch": 0.9241606451121386, + "grad_norm": 1.28872811794281, + "learning_rate": 2.8258202808307377e-06, + "loss": 0.0638, + "step": 39424 + }, + { + "epoch": 0.9241840866869436, + "grad_norm": 4.295647621154785, + "learning_rate": 2.8240817897937754e-06, + "loss": 0.2547, + "step": 39425 + }, + { + "epoch": 0.9242075282617487, + "grad_norm": 2.385892629623413, + "learning_rate": 2.822343826036078e-06, + "loss": 0.1489, + "step": 39426 + }, + { + "epoch": 0.9242309698365536, + "grad_norm": 1.7772481441497803, + "learning_rate": 2.8206063895670707e-06, + "loss": 0.1584, + "step": 39427 + }, + { + "epoch": 0.9242544114113587, + "grad_norm": 2.9013657569885254, + "learning_rate": 2.8188694803961913e-06, + "loss": 0.3147, + "step": 39428 + }, + { + "epoch": 0.9242778529861636, + "grad_norm": 3.0704503059387207, + "learning_rate": 2.8171330985328647e-06, + "loss": 0.3486, + "step": 39429 + }, + { + "epoch": 0.9243012945609687, + "grad_norm": 2.0092735290527344, + "learning_rate": 2.8153972439865063e-06, + "loss": 0.1378, + "step": 39430 + }, + { + "epoch": 0.9243247361357736, + "grad_norm": 2.7358946800231934, + "learning_rate": 2.8136619167665303e-06, + "loss": 0.1895, + "step": 39431 + }, + { + "epoch": 0.9243481777105786, + "grad_norm": 2.64438533782959, + "learning_rate": 2.8119271168823515e-06, + "loss": 0.2634, + "step": 39432 + }, + { + "epoch": 0.9243716192853836, + "grad_norm": 3.31449556350708, + "learning_rate": 2.8101928443433957e-06, + "loss": 0.3567, + "step": 39433 + }, + { + "epoch": 0.9243950608601886, + "grad_norm": 1.4352264404296875, + "learning_rate": 2.8084590991590553e-06, + "loss": 0.8693, + "step": 39434 + }, + { + "epoch": 0.9244185024349936, + "grad_norm": 3.679011583328247, + "learning_rate": 2.8067258813387563e-06, + "loss": 0.2564, + "step": 39435 + }, + { + "epoch": 0.9244419440097986, + "grad_norm": 4.126211643218994, + "learning_rate": 2.8049931908919023e-06, + "loss": 0.2072, + "step": 39436 + }, + { + "epoch": 0.9244653855846036, + "grad_norm": 3.449122428894043, + "learning_rate": 2.803261027827886e-06, + "loss": 0.3183, + "step": 39437 + }, + { + "epoch": 0.9244888271594086, + "grad_norm": 2.260890483856201, + "learning_rate": 2.8015293921560994e-06, + "loss": 0.2492, + "step": 39438 + }, + { + "epoch": 0.9245122687342135, + "grad_norm": 3.0756397247314453, + "learning_rate": 2.799798283885946e-06, + "loss": 0.2217, + "step": 39439 + }, + { + "epoch": 0.9245357103090186, + "grad_norm": 2.5751941204071045, + "learning_rate": 2.79806770302683e-06, + "loss": 0.1193, + "step": 39440 + }, + { + "epoch": 0.9245591518838235, + "grad_norm": 0.9365156292915344, + "learning_rate": 2.796337649588132e-06, + "loss": 0.0449, + "step": 39441 + }, + { + "epoch": 0.9245825934586286, + "grad_norm": 4.944089889526367, + "learning_rate": 2.7946081235792344e-06, + "loss": 0.3831, + "step": 39442 + }, + { + "epoch": 0.9246060350334335, + "grad_norm": 2.3692667484283447, + "learning_rate": 2.7928791250095287e-06, + "loss": 1.4517, + "step": 39443 + }, + { + "epoch": 0.9246294766082386, + "grad_norm": 0.9129736423492432, + "learning_rate": 2.7911506538883856e-06, + "loss": 0.0741, + "step": 39444 + }, + { + "epoch": 0.9246529181830435, + "grad_norm": 1.0635417699813843, + "learning_rate": 2.789422710225209e-06, + "loss": 0.0274, + "step": 39445 + }, + { + "epoch": 0.9246763597578486, + "grad_norm": 1.5979750156402588, + "learning_rate": 2.7876952940293354e-06, + "loss": 0.0848, + "step": 39446 + }, + { + "epoch": 0.9246998013326535, + "grad_norm": 2.3266828060150146, + "learning_rate": 2.78596840531018e-06, + "loss": 0.2238, + "step": 39447 + }, + { + "epoch": 0.9247232429074586, + "grad_norm": 4.123460292816162, + "learning_rate": 2.78424204407709e-06, + "loss": 0.3823, + "step": 39448 + }, + { + "epoch": 0.9247466844822635, + "grad_norm": 0.4856018126010895, + "learning_rate": 2.7825162103394364e-06, + "loss": 0.0586, + "step": 39449 + }, + { + "epoch": 0.9247701260570685, + "grad_norm": 3.0402863025665283, + "learning_rate": 2.780790904106578e-06, + "loss": 0.1744, + "step": 39450 + }, + { + "epoch": 0.9247935676318735, + "grad_norm": 3.3743505477905273, + "learning_rate": 2.7790661253878857e-06, + "loss": 0.3068, + "step": 39451 + }, + { + "epoch": 0.9248170092066785, + "grad_norm": 3.698762893676758, + "learning_rate": 2.777341874192707e-06, + "loss": 0.5487, + "step": 39452 + }, + { + "epoch": 0.9248404507814835, + "grad_norm": 1.5782322883605957, + "learning_rate": 2.7756181505304126e-06, + "loss": 0.0609, + "step": 39453 + }, + { + "epoch": 0.9248638923562885, + "grad_norm": 2.8327419757843018, + "learning_rate": 2.7738949544103497e-06, + "loss": 0.1488, + "step": 39454 + }, + { + "epoch": 0.9248873339310935, + "grad_norm": 2.6868906021118164, + "learning_rate": 2.7721722858418674e-06, + "loss": 0.2111, + "step": 39455 + }, + { + "epoch": 0.9249107755058985, + "grad_norm": 4.164589881896973, + "learning_rate": 2.770450144834291e-06, + "loss": 0.3551, + "step": 39456 + }, + { + "epoch": 0.9249342170807034, + "grad_norm": 4.579200744628906, + "learning_rate": 2.7687285313970134e-06, + "loss": 0.6146, + "step": 39457 + }, + { + "epoch": 0.9249576586555085, + "grad_norm": 2.1552734375, + "learning_rate": 2.767007445539338e-06, + "loss": 0.1944, + "step": 39458 + }, + { + "epoch": 0.9249811002303134, + "grad_norm": 0.9597750902175903, + "learning_rate": 2.7652868872706016e-06, + "loss": 0.0447, + "step": 39459 + }, + { + "epoch": 0.9250045418051185, + "grad_norm": 1.640366792678833, + "learning_rate": 2.7635668566001637e-06, + "loss": 0.1125, + "step": 39460 + }, + { + "epoch": 0.9250279833799234, + "grad_norm": 0.8885765075683594, + "learning_rate": 2.76184735353735e-06, + "loss": 0.078, + "step": 39461 + }, + { + "epoch": 0.9250514249547285, + "grad_norm": 2.397149085998535, + "learning_rate": 2.7601283780914866e-06, + "loss": 1.425, + "step": 39462 + }, + { + "epoch": 0.9250748665295334, + "grad_norm": 2.3625361919403076, + "learning_rate": 2.7584099302718992e-06, + "loss": 0.8838, + "step": 39463 + }, + { + "epoch": 0.9250983081043385, + "grad_norm": 0.7730839848518372, + "learning_rate": 2.7566920100879134e-06, + "loss": 0.0575, + "step": 39464 + }, + { + "epoch": 0.9251217496791434, + "grad_norm": 2.87623929977417, + "learning_rate": 2.7549746175488443e-06, + "loss": 0.257, + "step": 39465 + }, + { + "epoch": 0.9251451912539485, + "grad_norm": 3.4117023944854736, + "learning_rate": 2.753257752664029e-06, + "loss": 0.3277, + "step": 39466 + }, + { + "epoch": 0.9251686328287534, + "grad_norm": 1.364829421043396, + "learning_rate": 2.751541415442771e-06, + "loss": 0.0777, + "step": 39467 + }, + { + "epoch": 0.9251920744035584, + "grad_norm": 1.0832877159118652, + "learning_rate": 2.749825605894363e-06, + "loss": 0.1099, + "step": 39468 + }, + { + "epoch": 0.9252155159783634, + "grad_norm": 4.451838493347168, + "learning_rate": 2.7481103240281526e-06, + "loss": 0.1685, + "step": 39469 + }, + { + "epoch": 0.9252389575531684, + "grad_norm": 3.3743183612823486, + "learning_rate": 2.7463955698534327e-06, + "loss": 0.2812, + "step": 39470 + }, + { + "epoch": 0.9252623991279734, + "grad_norm": 1.3842991590499878, + "learning_rate": 2.744681343379507e-06, + "loss": 0.0747, + "step": 39471 + }, + { + "epoch": 0.9252858407027784, + "grad_norm": 3.494314670562744, + "learning_rate": 2.742967644615657e-06, + "loss": 0.4196, + "step": 39472 + }, + { + "epoch": 0.9253092822775834, + "grad_norm": 2.1665775775909424, + "learning_rate": 2.7412544735712088e-06, + "loss": 1.6019, + "step": 39473 + }, + { + "epoch": 0.9253327238523884, + "grad_norm": 2.913145065307617, + "learning_rate": 2.7395418302554542e-06, + "loss": 0.2283, + "step": 39474 + }, + { + "epoch": 0.9253561654271933, + "grad_norm": 3.7498769760131836, + "learning_rate": 2.7378297146776753e-06, + "loss": 0.2208, + "step": 39475 + }, + { + "epoch": 0.9253796070019984, + "grad_norm": 2.3636794090270996, + "learning_rate": 2.736118126847176e-06, + "loss": 0.1676, + "step": 39476 + }, + { + "epoch": 0.9254030485768034, + "grad_norm": 3.4682421684265137, + "learning_rate": 2.7344070667732145e-06, + "loss": 0.3407, + "step": 39477 + }, + { + "epoch": 0.9254264901516084, + "grad_norm": 1.6362770795822144, + "learning_rate": 2.7326965344651177e-06, + "loss": 1.2789, + "step": 39478 + }, + { + "epoch": 0.9254499317264134, + "grad_norm": 3.853799343109131, + "learning_rate": 2.7309865299321334e-06, + "loss": 0.2707, + "step": 39479 + }, + { + "epoch": 0.9254733733012184, + "grad_norm": 2.089712142944336, + "learning_rate": 2.729277053183543e-06, + "loss": 0.1587, + "step": 39480 + }, + { + "epoch": 0.9254968148760234, + "grad_norm": 1.9575719833374023, + "learning_rate": 2.7275681042286395e-06, + "loss": 0.8767, + "step": 39481 + }, + { + "epoch": 0.9255202564508284, + "grad_norm": 4.167311191558838, + "learning_rate": 2.7258596830766925e-06, + "loss": 0.4729, + "step": 39482 + }, + { + "epoch": 0.9255436980256334, + "grad_norm": 2.675114393234253, + "learning_rate": 2.724151789736962e-06, + "loss": 0.2521, + "step": 39483 + }, + { + "epoch": 0.9255671396004383, + "grad_norm": 4.472937107086182, + "learning_rate": 2.722444424218706e-06, + "loss": 0.2765, + "step": 39484 + }, + { + "epoch": 0.9255905811752434, + "grad_norm": 2.429490089416504, + "learning_rate": 2.720737586531219e-06, + "loss": 1.3144, + "step": 39485 + }, + { + "epoch": 0.9256140227500483, + "grad_norm": 2.0064666271209717, + "learning_rate": 2.7190312766837365e-06, + "loss": 0.1109, + "step": 39486 + }, + { + "epoch": 0.9256374643248534, + "grad_norm": 0.9864164590835571, + "learning_rate": 2.7173254946855298e-06, + "loss": 0.0763, + "step": 39487 + }, + { + "epoch": 0.9256609058996583, + "grad_norm": 0.7313706874847412, + "learning_rate": 2.7156202405458463e-06, + "loss": 0.1093, + "step": 39488 + }, + { + "epoch": 0.9256843474744634, + "grad_norm": 2.043602466583252, + "learning_rate": 2.713915514273946e-06, + "loss": 0.0865, + "step": 39489 + }, + { + "epoch": 0.9257077890492683, + "grad_norm": 2.631880760192871, + "learning_rate": 2.7122113158790763e-06, + "loss": 0.0942, + "step": 39490 + }, + { + "epoch": 0.9257312306240734, + "grad_norm": 2.8157732486724854, + "learning_rate": 2.7105076453704747e-06, + "loss": 2.183, + "step": 39491 + }, + { + "epoch": 0.9257546721988783, + "grad_norm": 0.7431889772415161, + "learning_rate": 2.7088045027574006e-06, + "loss": 0.0576, + "step": 39492 + }, + { + "epoch": 0.9257781137736834, + "grad_norm": 1.8971117734909058, + "learning_rate": 2.7071018880490906e-06, + "loss": 0.0904, + "step": 39493 + }, + { + "epoch": 0.9258015553484883, + "grad_norm": 2.543179512023926, + "learning_rate": 2.705399801254782e-06, + "loss": 0.1873, + "step": 39494 + }, + { + "epoch": 0.9258249969232933, + "grad_norm": 3.488823413848877, + "learning_rate": 2.703698242383712e-06, + "loss": 0.2451, + "step": 39495 + }, + { + "epoch": 0.9258484384980983, + "grad_norm": 3.7277750968933105, + "learning_rate": 2.7019972114451064e-06, + "loss": 0.2802, + "step": 39496 + }, + { + "epoch": 0.9258718800729033, + "grad_norm": 2.872659921646118, + "learning_rate": 2.700296708448191e-06, + "loss": 0.2641, + "step": 39497 + }, + { + "epoch": 0.9258953216477083, + "grad_norm": 0.4152877926826477, + "learning_rate": 2.6985967334022142e-06, + "loss": 0.035, + "step": 39498 + }, + { + "epoch": 0.9259187632225133, + "grad_norm": 2.733039617538452, + "learning_rate": 2.69689728631638e-06, + "loss": 0.1474, + "step": 39499 + }, + { + "epoch": 0.9259422047973183, + "grad_norm": 1.4286342859268188, + "learning_rate": 2.695198367199925e-06, + "loss": 0.1336, + "step": 39500 + }, + { + "epoch": 0.9259656463721233, + "grad_norm": 1.8273060321807861, + "learning_rate": 2.693499976062064e-06, + "loss": 0.0992, + "step": 39501 + }, + { + "epoch": 0.9259890879469282, + "grad_norm": 2.438671827316284, + "learning_rate": 2.6918021129120006e-06, + "loss": 0.1795, + "step": 39502 + }, + { + "epoch": 0.9260125295217333, + "grad_norm": 0.9602465629577637, + "learning_rate": 2.6901047777589396e-06, + "loss": 0.0295, + "step": 39503 + }, + { + "epoch": 0.9260359710965382, + "grad_norm": 1.1988319158554077, + "learning_rate": 2.6884079706121277e-06, + "loss": 0.1006, + "step": 39504 + }, + { + "epoch": 0.9260594126713433, + "grad_norm": 1.4581880569458008, + "learning_rate": 2.6867116914807367e-06, + "loss": 0.0407, + "step": 39505 + }, + { + "epoch": 0.9260828542461482, + "grad_norm": 4.919469833374023, + "learning_rate": 2.6850159403739917e-06, + "loss": 0.6662, + "step": 39506 + }, + { + "epoch": 0.9261062958209533, + "grad_norm": 0.5179067254066467, + "learning_rate": 2.6833207173010854e-06, + "loss": 0.018, + "step": 39507 + }, + { + "epoch": 0.9261297373957582, + "grad_norm": 4.4907097816467285, + "learning_rate": 2.681626022271222e-06, + "loss": 0.3257, + "step": 39508 + }, + { + "epoch": 0.9261531789705633, + "grad_norm": 2.210005044937134, + "learning_rate": 2.6799318552935825e-06, + "loss": 0.197, + "step": 39509 + }, + { + "epoch": 0.9261766205453682, + "grad_norm": 0.5617991089820862, + "learning_rate": 2.678238216377371e-06, + "loss": 0.0647, + "step": 39510 + }, + { + "epoch": 0.9262000621201733, + "grad_norm": 2.653261661529541, + "learning_rate": 2.67654510553178e-06, + "loss": 0.1942, + "step": 39511 + }, + { + "epoch": 0.9262235036949782, + "grad_norm": 3.08212947845459, + "learning_rate": 2.6748525227659913e-06, + "loss": 0.3455, + "step": 39512 + }, + { + "epoch": 0.9262469452697832, + "grad_norm": 0.8317687511444092, + "learning_rate": 2.6731604680891865e-06, + "loss": 0.117, + "step": 39513 + }, + { + "epoch": 0.9262703868445882, + "grad_norm": 0.9798609614372253, + "learning_rate": 2.6714689415105576e-06, + "loss": 0.1202, + "step": 39514 + }, + { + "epoch": 0.9262938284193932, + "grad_norm": 0.7966490983963013, + "learning_rate": 2.6697779430392646e-06, + "loss": 0.0411, + "step": 39515 + }, + { + "epoch": 0.9263172699941982, + "grad_norm": 3.0164268016815186, + "learning_rate": 2.6680874726844994e-06, + "loss": 0.2046, + "step": 39516 + }, + { + "epoch": 0.9263407115690032, + "grad_norm": 2.514132022857666, + "learning_rate": 2.6663975304554112e-06, + "loss": 0.1915, + "step": 39517 + }, + { + "epoch": 0.9263641531438082, + "grad_norm": 2.970097064971924, + "learning_rate": 2.664708116361203e-06, + "loss": 0.1168, + "step": 39518 + }, + { + "epoch": 0.9263875947186132, + "grad_norm": 3.0827553272247314, + "learning_rate": 2.6630192304110344e-06, + "loss": 0.1774, + "step": 39519 + }, + { + "epoch": 0.9264110362934181, + "grad_norm": 2.4577927589416504, + "learning_rate": 2.661330872614054e-06, + "loss": 0.1596, + "step": 39520 + }, + { + "epoch": 0.9264344778682232, + "grad_norm": 0.8657861351966858, + "learning_rate": 2.6596430429794204e-06, + "loss": 0.1236, + "step": 39521 + }, + { + "epoch": 0.9264579194430281, + "grad_norm": 2.754765033721924, + "learning_rate": 2.6579557415163048e-06, + "loss": 0.2575, + "step": 39522 + }, + { + "epoch": 0.9264813610178332, + "grad_norm": 3.6143932342529297, + "learning_rate": 2.6562689682338547e-06, + "loss": 0.3088, + "step": 39523 + }, + { + "epoch": 0.9265048025926381, + "grad_norm": 0.5080747008323669, + "learning_rate": 2.65458272314123e-06, + "loss": 0.0327, + "step": 39524 + }, + { + "epoch": 0.9265282441674432, + "grad_norm": 1.7425260543823242, + "learning_rate": 2.652897006247579e-06, + "loss": 1.1898, + "step": 39525 + }, + { + "epoch": 0.9265516857422481, + "grad_norm": 2.380295991897583, + "learning_rate": 2.6512118175620492e-06, + "loss": 0.0832, + "step": 39526 + }, + { + "epoch": 0.9265751273170532, + "grad_norm": 2.4930179119110107, + "learning_rate": 2.6495271570937783e-06, + "loss": 0.2223, + "step": 39527 + }, + { + "epoch": 0.9265985688918582, + "grad_norm": 1.4842748641967773, + "learning_rate": 2.647843024851904e-06, + "loss": 0.1071, + "step": 39528 + }, + { + "epoch": 0.9266220104666631, + "grad_norm": 0.376513808965683, + "learning_rate": 2.646159420845562e-06, + "loss": 0.0418, + "step": 39529 + }, + { + "epoch": 0.9266454520414682, + "grad_norm": 1.0165300369262695, + "learning_rate": 2.644476345083913e-06, + "loss": 0.0906, + "step": 39530 + }, + { + "epoch": 0.9266688936162731, + "grad_norm": 0.8623316884040833, + "learning_rate": 2.6427937975760707e-06, + "loss": 0.0314, + "step": 39531 + }, + { + "epoch": 0.9266923351910782, + "grad_norm": 0.6987559795379639, + "learning_rate": 2.6411117783311735e-06, + "loss": 0.041, + "step": 39532 + }, + { + "epoch": 0.9267157767658831, + "grad_norm": 1.9847073554992676, + "learning_rate": 2.6394302873583355e-06, + "loss": 0.2126, + "step": 39533 + }, + { + "epoch": 0.9267392183406882, + "grad_norm": 0.6115304827690125, + "learning_rate": 2.637749324666683e-06, + "loss": 0.0714, + "step": 39534 + }, + { + "epoch": 0.9267626599154931, + "grad_norm": 0.5042852163314819, + "learning_rate": 2.6360688902653306e-06, + "loss": 0.0282, + "step": 39535 + }, + { + "epoch": 0.9267861014902982, + "grad_norm": 1.2488856315612793, + "learning_rate": 2.6343889841634164e-06, + "loss": 0.0863, + "step": 39536 + }, + { + "epoch": 0.9268095430651031, + "grad_norm": 3.167091131210327, + "learning_rate": 2.632709606370054e-06, + "loss": 0.4016, + "step": 39537 + }, + { + "epoch": 0.9268329846399082, + "grad_norm": 2.7068521976470947, + "learning_rate": 2.6310307568943373e-06, + "loss": 0.1505, + "step": 39538 + }, + { + "epoch": 0.9268564262147131, + "grad_norm": 1.689549446105957, + "learning_rate": 2.6293524357453912e-06, + "loss": 0.1896, + "step": 39539 + }, + { + "epoch": 0.9268798677895181, + "grad_norm": 2.9478747844696045, + "learning_rate": 2.6276746429323097e-06, + "loss": 0.2761, + "step": 39540 + }, + { + "epoch": 0.9269033093643231, + "grad_norm": 3.33786678314209, + "learning_rate": 2.6259973784641956e-06, + "loss": 0.3055, + "step": 39541 + }, + { + "epoch": 0.9269267509391281, + "grad_norm": 3.298969030380249, + "learning_rate": 2.6243206423501644e-06, + "loss": 0.4629, + "step": 39542 + }, + { + "epoch": 0.9269501925139331, + "grad_norm": 2.5389158725738525, + "learning_rate": 2.6226444345993086e-06, + "loss": 0.1469, + "step": 39543 + }, + { + "epoch": 0.9269736340887381, + "grad_norm": 2.8827803134918213, + "learning_rate": 2.620968755220732e-06, + "loss": 0.185, + "step": 39544 + }, + { + "epoch": 0.9269970756635431, + "grad_norm": 2.2563867568969727, + "learning_rate": 2.619293604223505e-06, + "loss": 0.2538, + "step": 39545 + }, + { + "epoch": 0.9270205172383481, + "grad_norm": 3.7388663291931152, + "learning_rate": 2.6176189816167317e-06, + "loss": 0.2473, + "step": 39546 + }, + { + "epoch": 0.927043958813153, + "grad_norm": 2.300208806991577, + "learning_rate": 2.615944887409505e-06, + "loss": 0.284, + "step": 39547 + }, + { + "epoch": 0.9270674003879581, + "grad_norm": 6.061248779296875, + "learning_rate": 2.6142713216108837e-06, + "loss": 0.4844, + "step": 39548 + }, + { + "epoch": 0.927090841962763, + "grad_norm": 1.3153188228607178, + "learning_rate": 2.6125982842299723e-06, + "loss": 0.1441, + "step": 39549 + }, + { + "epoch": 0.9271142835375681, + "grad_norm": 0.644544780254364, + "learning_rate": 2.610925775275841e-06, + "loss": 0.0448, + "step": 39550 + }, + { + "epoch": 0.927137725112373, + "grad_norm": 0.4956805408000946, + "learning_rate": 2.609253794757571e-06, + "loss": 0.0227, + "step": 39551 + }, + { + "epoch": 0.9271611666871781, + "grad_norm": 7.7863030433654785, + "learning_rate": 2.6075823426842226e-06, + "loss": 0.1735, + "step": 39552 + }, + { + "epoch": 0.927184608261983, + "grad_norm": 2.275571346282959, + "learning_rate": 2.6059114190648657e-06, + "loss": 0.1567, + "step": 39553 + }, + { + "epoch": 0.9272080498367881, + "grad_norm": 0.9022806882858276, + "learning_rate": 2.6042410239085824e-06, + "loss": 0.0582, + "step": 39554 + }, + { + "epoch": 0.927231491411593, + "grad_norm": 2.6108877658843994, + "learning_rate": 2.6025711572244093e-06, + "loss": 0.2907, + "step": 39555 + }, + { + "epoch": 0.927254932986398, + "grad_norm": 2.1678526401519775, + "learning_rate": 2.6009018190214396e-06, + "loss": 0.1766, + "step": 39556 + }, + { + "epoch": 0.927278374561203, + "grad_norm": 3.256103992462158, + "learning_rate": 2.5992330093087103e-06, + "loss": 0.2053, + "step": 39557 + }, + { + "epoch": 0.927301816136008, + "grad_norm": 4.345277786254883, + "learning_rate": 2.597564728095281e-06, + "loss": 0.7022, + "step": 39558 + }, + { + "epoch": 0.927325257710813, + "grad_norm": 3.2386083602905273, + "learning_rate": 2.595896975390211e-06, + "loss": 0.2412, + "step": 39559 + }, + { + "epoch": 0.927348699285618, + "grad_norm": 1.9545735120773315, + "learning_rate": 2.594229751202526e-06, + "loss": 0.1933, + "step": 39560 + }, + { + "epoch": 0.927372140860423, + "grad_norm": 9.063591957092285, + "learning_rate": 2.592563055541297e-06, + "loss": 0.4606, + "step": 39561 + }, + { + "epoch": 0.927395582435228, + "grad_norm": 2.13393497467041, + "learning_rate": 2.5908968884155728e-06, + "loss": 0.1174, + "step": 39562 + }, + { + "epoch": 0.927419024010033, + "grad_norm": 3.595240592956543, + "learning_rate": 2.5892312498343676e-06, + "loss": 0.239, + "step": 39563 + }, + { + "epoch": 0.927442465584838, + "grad_norm": 3.475768566131592, + "learning_rate": 2.587566139806741e-06, + "loss": 0.3711, + "step": 39564 + }, + { + "epoch": 0.9274659071596429, + "grad_norm": 3.5057084560394287, + "learning_rate": 2.585901558341697e-06, + "loss": 0.1539, + "step": 39565 + }, + { + "epoch": 0.927489348734448, + "grad_norm": 1.991071343421936, + "learning_rate": 2.584237505448317e-06, + "loss": 0.0917, + "step": 39566 + }, + { + "epoch": 0.9275127903092529, + "grad_norm": 2.563049793243408, + "learning_rate": 2.5825739811355833e-06, + "loss": 0.185, + "step": 39567 + }, + { + "epoch": 0.927536231884058, + "grad_norm": 1.2402164936065674, + "learning_rate": 2.5809109854125547e-06, + "loss": 0.0798, + "step": 39568 + }, + { + "epoch": 0.9275596734588629, + "grad_norm": 1.3972426652908325, + "learning_rate": 2.5792485182882354e-06, + "loss": 0.0586, + "step": 39569 + }, + { + "epoch": 0.927583115033668, + "grad_norm": 2.6155872344970703, + "learning_rate": 2.5775865797716625e-06, + "loss": 0.1829, + "step": 39570 + }, + { + "epoch": 0.9276065566084729, + "grad_norm": 1.6551975011825562, + "learning_rate": 2.57592516987184e-06, + "loss": 0.1443, + "step": 39571 + }, + { + "epoch": 0.927629998183278, + "grad_norm": 1.7827644348144531, + "learning_rate": 2.574264288597794e-06, + "loss": 0.1083, + "step": 39572 + }, + { + "epoch": 0.9276534397580829, + "grad_norm": 1.7240533828735352, + "learning_rate": 2.5726039359585064e-06, + "loss": 1.0681, + "step": 39573 + }, + { + "epoch": 0.927676881332888, + "grad_norm": 1.708324909210205, + "learning_rate": 2.570944111963036e-06, + "loss": 0.1921, + "step": 39574 + }, + { + "epoch": 0.9277003229076929, + "grad_norm": 0.8962551355361938, + "learning_rate": 2.5692848166203433e-06, + "loss": 0.0951, + "step": 39575 + }, + { + "epoch": 0.9277237644824979, + "grad_norm": 1.7824115753173828, + "learning_rate": 2.5676260499394643e-06, + "loss": 0.1683, + "step": 39576 + }, + { + "epoch": 0.9277472060573029, + "grad_norm": 1.9606884717941284, + "learning_rate": 2.565967811929382e-06, + "loss": 0.1615, + "step": 39577 + }, + { + "epoch": 0.9277706476321079, + "grad_norm": 3.035676956176758, + "learning_rate": 2.564310102599088e-06, + "loss": 0.2322, + "step": 39578 + }, + { + "epoch": 0.927794089206913, + "grad_norm": 1.9950063228607178, + "learning_rate": 2.5626529219575977e-06, + "loss": 0.1987, + "step": 39579 + }, + { + "epoch": 0.9278175307817179, + "grad_norm": 1.0249638557434082, + "learning_rate": 2.5609962700138824e-06, + "loss": 0.08, + "step": 39580 + }, + { + "epoch": 0.927840972356523, + "grad_norm": 3.9987592697143555, + "learning_rate": 2.559340146776945e-06, + "loss": 2.2615, + "step": 39581 + }, + { + "epoch": 0.9278644139313279, + "grad_norm": 2.4378175735473633, + "learning_rate": 2.557684552255768e-06, + "loss": 0.1712, + "step": 39582 + }, + { + "epoch": 0.927887855506133, + "grad_norm": 2.7393922805786133, + "learning_rate": 2.5560294864593436e-06, + "loss": 0.3067, + "step": 39583 + }, + { + "epoch": 0.9279112970809379, + "grad_norm": 1.8864891529083252, + "learning_rate": 2.554374949396643e-06, + "loss": 0.0612, + "step": 39584 + }, + { + "epoch": 0.9279347386557429, + "grad_norm": 3.0734503269195557, + "learning_rate": 2.5527209410766363e-06, + "loss": 0.4714, + "step": 39585 + }, + { + "epoch": 0.9279581802305479, + "grad_norm": 1.9492764472961426, + "learning_rate": 2.551067461508294e-06, + "loss": 0.1915, + "step": 39586 + }, + { + "epoch": 0.9279816218053529, + "grad_norm": 1.9132137298583984, + "learning_rate": 2.5494145107006208e-06, + "loss": 0.0781, + "step": 39587 + }, + { + "epoch": 0.9280050633801579, + "grad_norm": 4.303241729736328, + "learning_rate": 2.547762088662553e-06, + "loss": 0.3102, + "step": 39588 + }, + { + "epoch": 0.9280285049549629, + "grad_norm": 2.200914144515991, + "learning_rate": 2.5461101954030732e-06, + "loss": 0.1079, + "step": 39589 + }, + { + "epoch": 0.9280519465297679, + "grad_norm": 1.4998372793197632, + "learning_rate": 2.544458830931129e-06, + "loss": 0.8609, + "step": 39590 + }, + { + "epoch": 0.9280753881045729, + "grad_norm": 0.9280866384506226, + "learning_rate": 2.5428079952557026e-06, + "loss": 0.0926, + "step": 39591 + }, + { + "epoch": 0.9280988296793778, + "grad_norm": 1.4492723941802979, + "learning_rate": 2.5411576883857424e-06, + "loss": 0.0664, + "step": 39592 + }, + { + "epoch": 0.9281222712541829, + "grad_norm": 1.9803061485290527, + "learning_rate": 2.5395079103301854e-06, + "loss": 0.1033, + "step": 39593 + }, + { + "epoch": 0.9281457128289878, + "grad_norm": 0.7060295343399048, + "learning_rate": 2.5378586610980137e-06, + "loss": 0.0347, + "step": 39594 + }, + { + "epoch": 0.9281691544037929, + "grad_norm": 1.2744333744049072, + "learning_rate": 2.5362099406981644e-06, + "loss": 0.687, + "step": 39595 + }, + { + "epoch": 0.9281925959785978, + "grad_norm": 5.016531467437744, + "learning_rate": 2.5345617491395744e-06, + "loss": 0.3911, + "step": 39596 + }, + { + "epoch": 0.9282160375534029, + "grad_norm": 1.4055618047714233, + "learning_rate": 2.5329140864311927e-06, + "loss": 0.1075, + "step": 39597 + }, + { + "epoch": 0.9282394791282078, + "grad_norm": 0.6860155463218689, + "learning_rate": 2.5312669525819567e-06, + "loss": 0.0503, + "step": 39598 + }, + { + "epoch": 0.9282629207030129, + "grad_norm": 2.3849332332611084, + "learning_rate": 2.529620347600803e-06, + "loss": 1.6022, + "step": 39599 + }, + { + "epoch": 0.9282863622778178, + "grad_norm": 1.6137827634811401, + "learning_rate": 2.5279742714966804e-06, + "loss": 0.0774, + "step": 39600 + }, + { + "epoch": 0.9283098038526229, + "grad_norm": 2.194559335708618, + "learning_rate": 2.526328724278504e-06, + "loss": 0.137, + "step": 39601 + }, + { + "epoch": 0.9283332454274278, + "grad_norm": 1.07606840133667, + "learning_rate": 2.5246837059552107e-06, + "loss": 0.083, + "step": 39602 + }, + { + "epoch": 0.9283566870022328, + "grad_norm": 2.8649513721466064, + "learning_rate": 2.5230392165357163e-06, + "loss": 0.2097, + "step": 39603 + }, + { + "epoch": 0.9283801285770378, + "grad_norm": 1.337418794631958, + "learning_rate": 2.521395256028969e-06, + "loss": 0.1015, + "step": 39604 + }, + { + "epoch": 0.9284035701518428, + "grad_norm": 1.2076071500778198, + "learning_rate": 2.51975182444385e-06, + "loss": 0.1314, + "step": 39605 + }, + { + "epoch": 0.9284270117266478, + "grad_norm": 3.8913607597351074, + "learning_rate": 2.5181089217893197e-06, + "loss": 0.4433, + "step": 39606 + }, + { + "epoch": 0.9284504533014528, + "grad_norm": 1.0504273176193237, + "learning_rate": 2.5164665480742587e-06, + "loss": 0.1052, + "step": 39607 + }, + { + "epoch": 0.9284738948762578, + "grad_norm": 1.7868802547454834, + "learning_rate": 2.5148247033075945e-06, + "loss": 0.1619, + "step": 39608 + }, + { + "epoch": 0.9284973364510628, + "grad_norm": 1.8964298963546753, + "learning_rate": 2.513183387498241e-06, + "loss": 0.1075, + "step": 39609 + }, + { + "epoch": 0.9285207780258677, + "grad_norm": 1.4541239738464355, + "learning_rate": 2.5115426006550923e-06, + "loss": 0.0544, + "step": 39610 + }, + { + "epoch": 0.9285442196006728, + "grad_norm": 3.657726526260376, + "learning_rate": 2.5099023427870407e-06, + "loss": 0.3436, + "step": 39611 + }, + { + "epoch": 0.9285676611754777, + "grad_norm": 2.6142165660858154, + "learning_rate": 2.508262613903012e-06, + "loss": 0.1838, + "step": 39612 + }, + { + "epoch": 0.9285911027502828, + "grad_norm": 2.2083609104156494, + "learning_rate": 2.5066234140118884e-06, + "loss": 0.171, + "step": 39613 + }, + { + "epoch": 0.9286145443250877, + "grad_norm": 1.5794165134429932, + "learning_rate": 2.5049847431225627e-06, + "loss": 0.1391, + "step": 39614 + }, + { + "epoch": 0.9286379858998928, + "grad_norm": 1.1345651149749756, + "learning_rate": 2.503346601243939e-06, + "loss": 0.0409, + "step": 39615 + }, + { + "epoch": 0.9286614274746977, + "grad_norm": 4.416385173797607, + "learning_rate": 2.5017089883848876e-06, + "loss": 0.2917, + "step": 39616 + }, + { + "epoch": 0.9286848690495028, + "grad_norm": 3.86850643157959, + "learning_rate": 2.5000719045543243e-06, + "loss": 0.2319, + "step": 39617 + }, + { + "epoch": 0.9287083106243077, + "grad_norm": 2.459385633468628, + "learning_rate": 2.4984353497610857e-06, + "loss": 0.2282, + "step": 39618 + }, + { + "epoch": 0.9287317521991127, + "grad_norm": 2.591493844985962, + "learning_rate": 2.4967993240140985e-06, + "loss": 0.1856, + "step": 39619 + }, + { + "epoch": 0.9287551937739177, + "grad_norm": 4.776238441467285, + "learning_rate": 2.495163827322211e-06, + "loss": 0.4556, + "step": 39620 + }, + { + "epoch": 0.9287786353487227, + "grad_norm": 2.0666849613189697, + "learning_rate": 2.4935288596943164e-06, + "loss": 0.1191, + "step": 39621 + }, + { + "epoch": 0.9288020769235277, + "grad_norm": 2.869992971420288, + "learning_rate": 2.4918944211392736e-06, + "loss": 0.3399, + "step": 39622 + }, + { + "epoch": 0.9288255184983327, + "grad_norm": 3.1638379096984863, + "learning_rate": 2.490260511665943e-06, + "loss": 0.1545, + "step": 39623 + }, + { + "epoch": 0.9288489600731377, + "grad_norm": 1.7078402042388916, + "learning_rate": 2.4886271312832055e-06, + "loss": 0.155, + "step": 39624 + }, + { + "epoch": 0.9288724016479427, + "grad_norm": 5.3869757652282715, + "learning_rate": 2.4869942799999213e-06, + "loss": 0.536, + "step": 39625 + }, + { + "epoch": 0.9288958432227477, + "grad_norm": 3.068976879119873, + "learning_rate": 2.485361957824939e-06, + "loss": 0.1961, + "step": 39626 + }, + { + "epoch": 0.9289192847975527, + "grad_norm": 1.9973115921020508, + "learning_rate": 2.4837301647671287e-06, + "loss": 0.2216, + "step": 39627 + }, + { + "epoch": 0.9289427263723576, + "grad_norm": 3.7128775119781494, + "learning_rate": 2.48209890083535e-06, + "loss": 0.3049, + "step": 39628 + }, + { + "epoch": 0.9289661679471627, + "grad_norm": 2.5068023204803467, + "learning_rate": 2.480468166038441e-06, + "loss": 0.1151, + "step": 39629 + }, + { + "epoch": 0.9289896095219677, + "grad_norm": 3.078197717666626, + "learning_rate": 2.4788379603852496e-06, + "loss": 0.2675, + "step": 39630 + }, + { + "epoch": 0.9290130510967727, + "grad_norm": 3.39935040473938, + "learning_rate": 2.4772082838846245e-06, + "loss": 0.4056, + "step": 39631 + }, + { + "epoch": 0.9290364926715777, + "grad_norm": 3.052086353302002, + "learning_rate": 2.475579136545414e-06, + "loss": 0.2274, + "step": 39632 + }, + { + "epoch": 0.9290599342463827, + "grad_norm": 1.4747039079666138, + "learning_rate": 2.4739505183764445e-06, + "loss": 0.0847, + "step": 39633 + }, + { + "epoch": 0.9290833758211877, + "grad_norm": 0.7628024220466614, + "learning_rate": 2.4723224293865756e-06, + "loss": 0.077, + "step": 39634 + }, + { + "epoch": 0.9291068173959927, + "grad_norm": 1.838857650756836, + "learning_rate": 2.470694869584622e-06, + "loss": 0.989, + "step": 39635 + }, + { + "epoch": 0.9291302589707977, + "grad_norm": 1.1580580472946167, + "learning_rate": 2.469067838979411e-06, + "loss": 0.1301, + "step": 39636 + }, + { + "epoch": 0.9291537005456026, + "grad_norm": 0.7987567782402039, + "learning_rate": 2.46744133757979e-06, + "loss": 0.0481, + "step": 39637 + }, + { + "epoch": 0.9291771421204077, + "grad_norm": 2.2613909244537354, + "learning_rate": 2.4658153653945635e-06, + "loss": 0.176, + "step": 39638 + }, + { + "epoch": 0.9292005836952126, + "grad_norm": 1.5742782354354858, + "learning_rate": 2.4641899224325693e-06, + "loss": 0.1833, + "step": 39639 + }, + { + "epoch": 0.9292240252700177, + "grad_norm": 2.5737526416778564, + "learning_rate": 2.462565008702633e-06, + "loss": 0.2354, + "step": 39640 + }, + { + "epoch": 0.9292474668448226, + "grad_norm": 3.1414647102355957, + "learning_rate": 2.460940624213559e-06, + "loss": 0.2246, + "step": 39641 + }, + { + "epoch": 0.9292709084196277, + "grad_norm": 2.7874910831451416, + "learning_rate": 2.4593167689741626e-06, + "loss": 0.2556, + "step": 39642 + }, + { + "epoch": 0.9292943499944326, + "grad_norm": 1.3279128074645996, + "learning_rate": 2.4576934429932473e-06, + "loss": 0.3011, + "step": 39643 + }, + { + "epoch": 0.9293177915692377, + "grad_norm": 2.007099151611328, + "learning_rate": 2.45607064627964e-06, + "loss": 0.1434, + "step": 39644 + }, + { + "epoch": 0.9293412331440426, + "grad_norm": 2.767547845840454, + "learning_rate": 2.4544483788421334e-06, + "loss": 0.2227, + "step": 39645 + }, + { + "epoch": 0.9293646747188476, + "grad_norm": 2.105228900909424, + "learning_rate": 2.4528266406895315e-06, + "loss": 0.14, + "step": 39646 + }, + { + "epoch": 0.9293881162936526, + "grad_norm": 2.0576655864715576, + "learning_rate": 2.451205431830639e-06, + "loss": 0.1076, + "step": 39647 + }, + { + "epoch": 0.9294115578684576, + "grad_norm": 2.3145265579223633, + "learning_rate": 2.4495847522742476e-06, + "loss": 0.1403, + "step": 39648 + }, + { + "epoch": 0.9294349994432626, + "grad_norm": 1.3070200681686401, + "learning_rate": 2.4479646020291513e-06, + "loss": 0.0522, + "step": 39649 + }, + { + "epoch": 0.9294584410180676, + "grad_norm": 2.6686413288116455, + "learning_rate": 2.446344981104132e-06, + "loss": 0.1493, + "step": 39650 + }, + { + "epoch": 0.9294818825928726, + "grad_norm": 1.5012011528015137, + "learning_rate": 2.4447258895079926e-06, + "loss": 0.1102, + "step": 39651 + }, + { + "epoch": 0.9295053241676776, + "grad_norm": 1.4961013793945312, + "learning_rate": 2.4431073272495274e-06, + "loss": 0.0521, + "step": 39652 + }, + { + "epoch": 0.9295287657424826, + "grad_norm": 1.764428973197937, + "learning_rate": 2.441489294337496e-06, + "loss": 0.0708, + "step": 39653 + }, + { + "epoch": 0.9295522073172876, + "grad_norm": 1.6704390048980713, + "learning_rate": 2.439871790780679e-06, + "loss": 0.1222, + "step": 39654 + }, + { + "epoch": 0.9295756488920925, + "grad_norm": 2.430546998977661, + "learning_rate": 2.4382548165878705e-06, + "loss": 0.1817, + "step": 39655 + }, + { + "epoch": 0.9295990904668976, + "grad_norm": 1.9973573684692383, + "learning_rate": 2.43663837176783e-06, + "loss": 0.3905, + "step": 39656 + }, + { + "epoch": 0.9296225320417025, + "grad_norm": 1.7968000173568726, + "learning_rate": 2.435022456329339e-06, + "loss": 0.1559, + "step": 39657 + }, + { + "epoch": 0.9296459736165076, + "grad_norm": 3.195817708969116, + "learning_rate": 2.4334070702811572e-06, + "loss": 0.1674, + "step": 39658 + }, + { + "epoch": 0.9296694151913125, + "grad_norm": 3.149301528930664, + "learning_rate": 2.4317922136320447e-06, + "loss": 0.1845, + "step": 39659 + }, + { + "epoch": 0.9296928567661176, + "grad_norm": 4.782651901245117, + "learning_rate": 2.430177886390783e-06, + "loss": 0.2926, + "step": 39660 + }, + { + "epoch": 0.9297162983409225, + "grad_norm": 0.6884527802467346, + "learning_rate": 2.42856408856611e-06, + "loss": 0.0303, + "step": 39661 + }, + { + "epoch": 0.9297397399157276, + "grad_norm": 0.08923124521970749, + "learning_rate": 2.4269508201667957e-06, + "loss": 0.0023, + "step": 39662 + }, + { + "epoch": 0.9297631814905325, + "grad_norm": 2.531261682510376, + "learning_rate": 2.425338081201578e-06, + "loss": 1.1506, + "step": 39663 + }, + { + "epoch": 0.9297866230653375, + "grad_norm": 3.613473653793335, + "learning_rate": 2.423725871679228e-06, + "loss": 0.4355, + "step": 39664 + }, + { + "epoch": 0.9298100646401425, + "grad_norm": 3.179513454437256, + "learning_rate": 2.422114191608493e-06, + "loss": 0.2482, + "step": 39665 + }, + { + "epoch": 0.9298335062149475, + "grad_norm": 1.0545194149017334, + "learning_rate": 2.4205030409981012e-06, + "loss": 0.0935, + "step": 39666 + }, + { + "epoch": 0.9298569477897525, + "grad_norm": 2.56793212890625, + "learning_rate": 2.4188924198568106e-06, + "loss": 0.3308, + "step": 39667 + }, + { + "epoch": 0.9298803893645575, + "grad_norm": 3.3653624057769775, + "learning_rate": 2.417282328193349e-06, + "loss": 0.2154, + "step": 39668 + }, + { + "epoch": 0.9299038309393625, + "grad_norm": 1.2858027219772339, + "learning_rate": 2.4156727660164525e-06, + "loss": 0.1021, + "step": 39669 + }, + { + "epoch": 0.9299272725141675, + "grad_norm": 2.252758502960205, + "learning_rate": 2.41406373333487e-06, + "loss": 1.3765, + "step": 39670 + }, + { + "epoch": 0.9299507140889725, + "grad_norm": 3.0870301723480225, + "learning_rate": 2.412455230157318e-06, + "loss": 0.1557, + "step": 39671 + }, + { + "epoch": 0.9299741556637775, + "grad_norm": 2.356563091278076, + "learning_rate": 2.410847256492521e-06, + "loss": 0.1819, + "step": 39672 + }, + { + "epoch": 0.9299975972385824, + "grad_norm": 2.6190900802612305, + "learning_rate": 2.409239812349218e-06, + "loss": 0.1418, + "step": 39673 + }, + { + "epoch": 0.9300210388133875, + "grad_norm": 0.4945899248123169, + "learning_rate": 2.4076328977361117e-06, + "loss": 0.0296, + "step": 39674 + }, + { + "epoch": 0.9300444803881924, + "grad_norm": 0.6087806224822998, + "learning_rate": 2.406026512661941e-06, + "loss": 0.044, + "step": 39675 + }, + { + "epoch": 0.9300679219629975, + "grad_norm": 4.2865824699401855, + "learning_rate": 2.40442065713542e-06, + "loss": 0.4299, + "step": 39676 + }, + { + "epoch": 0.9300913635378024, + "grad_norm": 1.2737070322036743, + "learning_rate": 2.4028153311652535e-06, + "loss": 0.0918, + "step": 39677 + }, + { + "epoch": 0.9301148051126075, + "grad_norm": 2.288083553314209, + "learning_rate": 2.4012105347601565e-06, + "loss": 0.1421, + "step": 39678 + }, + { + "epoch": 0.9301382466874124, + "grad_norm": 1.1651989221572876, + "learning_rate": 2.3996062679288333e-06, + "loss": 0.4328, + "step": 39679 + }, + { + "epoch": 0.9301616882622175, + "grad_norm": 0.14204631745815277, + "learning_rate": 2.398002530679999e-06, + "loss": 0.0089, + "step": 39680 + }, + { + "epoch": 0.9301851298370225, + "grad_norm": 3.047598361968994, + "learning_rate": 2.3963993230223246e-06, + "loss": 0.217, + "step": 39681 + }, + { + "epoch": 0.9302085714118274, + "grad_norm": 2.172813892364502, + "learning_rate": 2.394796644964559e-06, + "loss": 0.0914, + "step": 39682 + }, + { + "epoch": 0.9302320129866325, + "grad_norm": 1.9128575325012207, + "learning_rate": 2.393194496515361e-06, + "loss": 0.1003, + "step": 39683 + }, + { + "epoch": 0.9302554545614374, + "grad_norm": 4.3916754722595215, + "learning_rate": 2.391592877683424e-06, + "loss": 0.341, + "step": 39684 + }, + { + "epoch": 0.9302788961362425, + "grad_norm": 1.2031090259552002, + "learning_rate": 2.3899917884774634e-06, + "loss": 0.1036, + "step": 39685 + }, + { + "epoch": 0.9303023377110474, + "grad_norm": 3.0567548274993896, + "learning_rate": 2.3883912289061506e-06, + "loss": 0.1992, + "step": 39686 + }, + { + "epoch": 0.9303257792858525, + "grad_norm": 2.831897497177124, + "learning_rate": 2.3867911989781556e-06, + "loss": 0.2723, + "step": 39687 + }, + { + "epoch": 0.9303492208606574, + "grad_norm": 2.0816597938537598, + "learning_rate": 2.385191698702183e-06, + "loss": 1.7374, + "step": 39688 + }, + { + "epoch": 0.9303726624354625, + "grad_norm": 2.5251688957214355, + "learning_rate": 2.383592728086903e-06, + "loss": 0.2135, + "step": 39689 + }, + { + "epoch": 0.9303961040102674, + "grad_norm": 3.480154514312744, + "learning_rate": 2.3819942871410093e-06, + "loss": 0.348, + "step": 39690 + }, + { + "epoch": 0.9304195455850724, + "grad_norm": 1.1389096975326538, + "learning_rate": 2.3803963758731506e-06, + "loss": 0.1352, + "step": 39691 + }, + { + "epoch": 0.9304429871598774, + "grad_norm": 2.975903272628784, + "learning_rate": 2.3787989942920086e-06, + "loss": 0.2346, + "step": 39692 + }, + { + "epoch": 0.9304664287346824, + "grad_norm": 4.092973232269287, + "learning_rate": 2.3772021424062427e-06, + "loss": 0.324, + "step": 39693 + }, + { + "epoch": 0.9304898703094874, + "grad_norm": 1.4488348960876465, + "learning_rate": 2.3756058202245134e-06, + "loss": 0.1289, + "step": 39694 + }, + { + "epoch": 0.9305133118842924, + "grad_norm": 3.9191012382507324, + "learning_rate": 2.3740100277555023e-06, + "loss": 0.279, + "step": 39695 + }, + { + "epoch": 0.9305367534590974, + "grad_norm": 0.7785283327102661, + "learning_rate": 2.3724147650078685e-06, + "loss": 0.0683, + "step": 39696 + }, + { + "epoch": 0.9305601950339024, + "grad_norm": 0.8935564756393433, + "learning_rate": 2.3708200319902507e-06, + "loss": 0.0621, + "step": 39697 + }, + { + "epoch": 0.9305836366087074, + "grad_norm": 3.39017391204834, + "learning_rate": 2.3692258287112966e-06, + "loss": 0.259, + "step": 39698 + }, + { + "epoch": 0.9306070781835124, + "grad_norm": 1.4273982048034668, + "learning_rate": 2.3676321551796776e-06, + "loss": 0.1094, + "step": 39699 + }, + { + "epoch": 0.9306305197583173, + "grad_norm": 3.9949584007263184, + "learning_rate": 2.366039011404031e-06, + "loss": 0.4336, + "step": 39700 + }, + { + "epoch": 0.9306539613331224, + "grad_norm": 0.890177309513092, + "learning_rate": 2.3644463973929943e-06, + "loss": 0.0841, + "step": 39701 + }, + { + "epoch": 0.9306774029079273, + "grad_norm": 1.2303028106689453, + "learning_rate": 2.3628543131552273e-06, + "loss": 0.095, + "step": 39702 + }, + { + "epoch": 0.9307008444827324, + "grad_norm": 0.41872650384902954, + "learning_rate": 2.3612627586993565e-06, + "loss": 0.0142, + "step": 39703 + }, + { + "epoch": 0.9307242860575373, + "grad_norm": 2.0450711250305176, + "learning_rate": 2.359671734034019e-06, + "loss": 0.1662, + "step": 39704 + }, + { + "epoch": 0.9307477276323424, + "grad_norm": 1.5994079113006592, + "learning_rate": 2.358081239167853e-06, + "loss": 0.1281, + "step": 39705 + }, + { + "epoch": 0.9307711692071473, + "grad_norm": 1.9277762174606323, + "learning_rate": 2.3564912741094736e-06, + "loss": 0.1753, + "step": 39706 + }, + { + "epoch": 0.9307946107819524, + "grad_norm": 2.2075676918029785, + "learning_rate": 2.3549018388675182e-06, + "loss": 0.131, + "step": 39707 + }, + { + "epoch": 0.9308180523567573, + "grad_norm": 0.9185525178909302, + "learning_rate": 2.3533129334506133e-06, + "loss": 0.0736, + "step": 39708 + }, + { + "epoch": 0.9308414939315623, + "grad_norm": 2.4884703159332275, + "learning_rate": 2.3517245578673854e-06, + "loss": 0.2373, + "step": 39709 + }, + { + "epoch": 0.9308649355063673, + "grad_norm": 2.2192838191986084, + "learning_rate": 2.3501367121264382e-06, + "loss": 0.1484, + "step": 39710 + }, + { + "epoch": 0.9308883770811723, + "grad_norm": 3.1297709941864014, + "learning_rate": 2.348549396236388e-06, + "loss": 0.2958, + "step": 39711 + }, + { + "epoch": 0.9309118186559773, + "grad_norm": 2.7333266735076904, + "learning_rate": 2.3469626102058606e-06, + "loss": 0.2837, + "step": 39712 + }, + { + "epoch": 0.9309352602307823, + "grad_norm": 1.3244613409042358, + "learning_rate": 2.3453763540434604e-06, + "loss": 0.0406, + "step": 39713 + }, + { + "epoch": 0.9309587018055873, + "grad_norm": 2.5922446250915527, + "learning_rate": 2.3437906277577916e-06, + "loss": 1.3989, + "step": 39714 + }, + { + "epoch": 0.9309821433803923, + "grad_norm": 1.5370360612869263, + "learning_rate": 2.342205431357469e-06, + "loss": 0.1045, + "step": 39715 + }, + { + "epoch": 0.9310055849551973, + "grad_norm": 1.8776991367340088, + "learning_rate": 2.340620764851087e-06, + "loss": 1.6326, + "step": 39716 + }, + { + "epoch": 0.9310290265300023, + "grad_norm": 2.9704928398132324, + "learning_rate": 2.3390366282472378e-06, + "loss": 1.2665, + "step": 39717 + }, + { + "epoch": 0.9310524681048072, + "grad_norm": 2.242215394973755, + "learning_rate": 2.3374530215545256e-06, + "loss": 0.1474, + "step": 39718 + }, + { + "epoch": 0.9310759096796123, + "grad_norm": 3.781635046005249, + "learning_rate": 2.3358699447815215e-06, + "loss": 0.4192, + "step": 39719 + }, + { + "epoch": 0.9310993512544172, + "grad_norm": 3.071038007736206, + "learning_rate": 2.3342873979368516e-06, + "loss": 0.2802, + "step": 39720 + }, + { + "epoch": 0.9311227928292223, + "grad_norm": 2.6589157581329346, + "learning_rate": 2.3327053810290766e-06, + "loss": 0.1965, + "step": 39721 + }, + { + "epoch": 0.9311462344040272, + "grad_norm": 2.8262553215026855, + "learning_rate": 2.3311238940668e-06, + "loss": 0.1214, + "step": 39722 + }, + { + "epoch": 0.9311696759788323, + "grad_norm": 3.083583354949951, + "learning_rate": 2.3295429370585707e-06, + "loss": 0.2389, + "step": 39723 + }, + { + "epoch": 0.9311931175536372, + "grad_norm": 1.7234903573989868, + "learning_rate": 2.3279625100130042e-06, + "loss": 0.0794, + "step": 39724 + }, + { + "epoch": 0.9312165591284423, + "grad_norm": 2.331059694290161, + "learning_rate": 2.32638261293866e-06, + "loss": 0.224, + "step": 39725 + }, + { + "epoch": 0.9312400007032472, + "grad_norm": 3.272614002227783, + "learning_rate": 2.3248032458440984e-06, + "loss": 0.0888, + "step": 39726 + }, + { + "epoch": 0.9312634422780522, + "grad_norm": 1.9838613271713257, + "learning_rate": 2.323224408737901e-06, + "loss": 0.0845, + "step": 39727 + }, + { + "epoch": 0.9312868838528572, + "grad_norm": 3.14973783493042, + "learning_rate": 2.3216461016286494e-06, + "loss": 0.1476, + "step": 39728 + }, + { + "epoch": 0.9313103254276622, + "grad_norm": 2.1730072498321533, + "learning_rate": 2.320068324524882e-06, + "loss": 0.1665, + "step": 39729 + }, + { + "epoch": 0.9313337670024672, + "grad_norm": 1.6370923519134521, + "learning_rate": 2.31849107743517e-06, + "loss": 0.1104, + "step": 39730 + }, + { + "epoch": 0.9313572085772722, + "grad_norm": 0.6239952445030212, + "learning_rate": 2.316914360368083e-06, + "loss": 0.021, + "step": 39731 + }, + { + "epoch": 0.9313806501520773, + "grad_norm": 3.3058180809020996, + "learning_rate": 2.3153381733321487e-06, + "loss": 1.6361, + "step": 39732 + }, + { + "epoch": 0.9314040917268822, + "grad_norm": 1.1454222202301025, + "learning_rate": 2.3137625163359377e-06, + "loss": 0.0962, + "step": 39733 + }, + { + "epoch": 0.9314275333016873, + "grad_norm": 0.5288586616516113, + "learning_rate": 2.3121873893880097e-06, + "loss": 0.0271, + "step": 39734 + }, + { + "epoch": 0.9314509748764922, + "grad_norm": 0.8283774256706238, + "learning_rate": 2.3106127924968913e-06, + "loss": 0.0859, + "step": 39735 + }, + { + "epoch": 0.9314744164512972, + "grad_norm": 3.2020153999328613, + "learning_rate": 2.309038725671131e-06, + "loss": 0.2039, + "step": 39736 + }, + { + "epoch": 0.9314978580261022, + "grad_norm": 4.143810272216797, + "learning_rate": 2.307465188919289e-06, + "loss": 0.1637, + "step": 39737 + }, + { + "epoch": 0.9315212996009072, + "grad_norm": 0.8670355081558228, + "learning_rate": 2.3058921822498803e-06, + "loss": 0.0458, + "step": 39738 + }, + { + "epoch": 0.9315447411757122, + "grad_norm": 2.243715763092041, + "learning_rate": 2.3043197056714315e-06, + "loss": 0.135, + "step": 39739 + }, + { + "epoch": 0.9315681827505172, + "grad_norm": 0.8216835260391235, + "learning_rate": 2.3027477591925138e-06, + "loss": 0.0555, + "step": 39740 + }, + { + "epoch": 0.9315916243253222, + "grad_norm": 0.6753237247467041, + "learning_rate": 2.3011763428216203e-06, + "loss": 0.0573, + "step": 39741 + }, + { + "epoch": 0.9316150659001272, + "grad_norm": 1.8147461414337158, + "learning_rate": 2.2996054565672998e-06, + "loss": 1.2688, + "step": 39742 + }, + { + "epoch": 0.9316385074749322, + "grad_norm": 0.5374711155891418, + "learning_rate": 2.2980351004380674e-06, + "loss": 0.0275, + "step": 39743 + }, + { + "epoch": 0.9316619490497372, + "grad_norm": 2.7806737422943115, + "learning_rate": 2.2964652744424387e-06, + "loss": 0.2199, + "step": 39744 + }, + { + "epoch": 0.9316853906245421, + "grad_norm": 1.3460545539855957, + "learning_rate": 2.294895978588929e-06, + "loss": 0.0737, + "step": 39745 + }, + { + "epoch": 0.9317088321993472, + "grad_norm": 2.367036819458008, + "learning_rate": 2.2933272128860762e-06, + "loss": 0.1407, + "step": 39746 + }, + { + "epoch": 0.9317322737741521, + "grad_norm": 2.239698886871338, + "learning_rate": 2.2917589773423732e-06, + "loss": 1.6312, + "step": 39747 + }, + { + "epoch": 0.9317557153489572, + "grad_norm": 0.5926405787467957, + "learning_rate": 2.2901912719663133e-06, + "loss": 0.0575, + "step": 39748 + }, + { + "epoch": 0.9317791569237621, + "grad_norm": 1.8532627820968628, + "learning_rate": 2.288624096766445e-06, + "loss": 0.0688, + "step": 39749 + }, + { + "epoch": 0.9318025984985672, + "grad_norm": 1.4301528930664062, + "learning_rate": 2.287057451751251e-06, + "loss": 0.1147, + "step": 39750 + }, + { + "epoch": 0.9318260400733721, + "grad_norm": 1.5716698169708252, + "learning_rate": 2.285491336929213e-06, + "loss": 0.1217, + "step": 39751 + }, + { + "epoch": 0.9318494816481772, + "grad_norm": 2.6039912700653076, + "learning_rate": 2.2839257523088575e-06, + "loss": 0.3132, + "step": 39752 + }, + { + "epoch": 0.9318729232229821, + "grad_norm": 2.507704734802246, + "learning_rate": 2.282360697898678e-06, + "loss": 0.155, + "step": 39753 + }, + { + "epoch": 0.9318963647977871, + "grad_norm": 3.7906980514526367, + "learning_rate": 2.280796173707145e-06, + "loss": 0.3324, + "step": 39754 + }, + { + "epoch": 0.9319198063725921, + "grad_norm": 1.2877708673477173, + "learning_rate": 2.279232179742763e-06, + "loss": 0.11, + "step": 39755 + }, + { + "epoch": 0.9319432479473971, + "grad_norm": 0.9790854454040527, + "learning_rate": 2.2776687160140144e-06, + "loss": 0.1254, + "step": 39756 + }, + { + "epoch": 0.9319666895222021, + "grad_norm": 3.6935973167419434, + "learning_rate": 2.276105782529381e-06, + "loss": 0.1343, + "step": 39757 + }, + { + "epoch": 0.9319901310970071, + "grad_norm": 2.5590593814849854, + "learning_rate": 2.274543379297345e-06, + "loss": 0.2027, + "step": 39758 + }, + { + "epoch": 0.9320135726718121, + "grad_norm": 2.1196722984313965, + "learning_rate": 2.272981506326388e-06, + "loss": 0.1076, + "step": 39759 + }, + { + "epoch": 0.9320370142466171, + "grad_norm": 2.61038875579834, + "learning_rate": 2.271420163624971e-06, + "loss": 0.1303, + "step": 39760 + }, + { + "epoch": 0.932060455821422, + "grad_norm": 2.3658084869384766, + "learning_rate": 2.2698593512015863e-06, + "loss": 0.1551, + "step": 39761 + }, + { + "epoch": 0.9320838973962271, + "grad_norm": 1.5385651588439941, + "learning_rate": 2.2682990690646945e-06, + "loss": 0.8464, + "step": 39762 + }, + { + "epoch": 0.932107338971032, + "grad_norm": 3.8551900386810303, + "learning_rate": 2.266739317222766e-06, + "loss": 0.3509, + "step": 39763 + }, + { + "epoch": 0.9321307805458371, + "grad_norm": 1.724408507347107, + "learning_rate": 2.2651800956842385e-06, + "loss": 1.0728, + "step": 39764 + }, + { + "epoch": 0.932154222120642, + "grad_norm": 1.7841157913208008, + "learning_rate": 2.2636214044576053e-06, + "loss": 0.0895, + "step": 39765 + }, + { + "epoch": 0.9321776636954471, + "grad_norm": 3.172983169555664, + "learning_rate": 2.2620632435513046e-06, + "loss": 0.2767, + "step": 39766 + }, + { + "epoch": 0.932201105270252, + "grad_norm": 1.505852460861206, + "learning_rate": 2.2605056129738065e-06, + "loss": 0.1081, + "step": 39767 + }, + { + "epoch": 0.9322245468450571, + "grad_norm": 3.0247867107391357, + "learning_rate": 2.2589485127335495e-06, + "loss": 0.1247, + "step": 39768 + }, + { + "epoch": 0.932247988419862, + "grad_norm": 3.5888259410858154, + "learning_rate": 2.2573919428389823e-06, + "loss": 0.2437, + "step": 39769 + }, + { + "epoch": 0.9322714299946671, + "grad_norm": 2.0160655975341797, + "learning_rate": 2.2558359032985422e-06, + "loss": 0.1422, + "step": 39770 + }, + { + "epoch": 0.932294871569472, + "grad_norm": 0.7914813756942749, + "learning_rate": 2.2542803941207e-06, + "loss": 0.0923, + "step": 39771 + }, + { + "epoch": 0.932318313144277, + "grad_norm": 2.7582976818084717, + "learning_rate": 2.2527254153138723e-06, + "loss": 0.1411, + "step": 39772 + }, + { + "epoch": 0.932341754719082, + "grad_norm": 2.544224977493286, + "learning_rate": 2.251170966886518e-06, + "loss": 1.3604, + "step": 39773 + }, + { + "epoch": 0.932365196293887, + "grad_norm": 2.4068989753723145, + "learning_rate": 2.2496170488470525e-06, + "loss": 0.1672, + "step": 39774 + }, + { + "epoch": 0.932388637868692, + "grad_norm": 1.8311867713928223, + "learning_rate": 2.248063661203914e-06, + "loss": 0.1309, + "step": 39775 + }, + { + "epoch": 0.932412079443497, + "grad_norm": 5.202707290649414, + "learning_rate": 2.246510803965529e-06, + "loss": 0.3488, + "step": 39776 + }, + { + "epoch": 0.932435521018302, + "grad_norm": 4.638556480407715, + "learning_rate": 2.2449584771403243e-06, + "loss": 0.5575, + "step": 39777 + }, + { + "epoch": 0.932458962593107, + "grad_norm": 3.042389392852783, + "learning_rate": 2.243406680736726e-06, + "loss": 0.2783, + "step": 39778 + }, + { + "epoch": 0.932482404167912, + "grad_norm": 1.1485390663146973, + "learning_rate": 2.24185541476315e-06, + "loss": 0.0904, + "step": 39779 + }, + { + "epoch": 0.932505845742717, + "grad_norm": 3.050358295440674, + "learning_rate": 2.240304679228011e-06, + "loss": 0.2108, + "step": 39780 + }, + { + "epoch": 0.9325292873175219, + "grad_norm": 2.0288610458374023, + "learning_rate": 2.238754474139737e-06, + "loss": 0.1574, + "step": 39781 + }, + { + "epoch": 0.932552728892327, + "grad_norm": 1.4754058122634888, + "learning_rate": 2.23720479950672e-06, + "loss": 0.1039, + "step": 39782 + }, + { + "epoch": 0.932576170467132, + "grad_norm": 1.0579673051834106, + "learning_rate": 2.2356556553373874e-06, + "loss": 0.0709, + "step": 39783 + }, + { + "epoch": 0.932599612041937, + "grad_norm": 1.2739171981811523, + "learning_rate": 2.234107041640121e-06, + "loss": 1.3249, + "step": 39784 + }, + { + "epoch": 0.932623053616742, + "grad_norm": 3.7031195163726807, + "learning_rate": 2.2325589584233475e-06, + "loss": 0.3992, + "step": 39785 + }, + { + "epoch": 0.932646495191547, + "grad_norm": 2.1695497035980225, + "learning_rate": 2.2310114056954713e-06, + "loss": 0.2019, + "step": 39786 + }, + { + "epoch": 0.932669936766352, + "grad_norm": 2.8142683506011963, + "learning_rate": 2.229464383464863e-06, + "loss": 0.1754, + "step": 39787 + }, + { + "epoch": 0.932693378341157, + "grad_norm": 3.692150115966797, + "learning_rate": 2.227917891739939e-06, + "loss": 0.2918, + "step": 39788 + }, + { + "epoch": 0.932716819915962, + "grad_norm": 2.7708427906036377, + "learning_rate": 2.2263719305290585e-06, + "loss": 0.1661, + "step": 39789 + }, + { + "epoch": 0.9327402614907669, + "grad_norm": 3.104672908782959, + "learning_rate": 2.22482649984066e-06, + "loss": 1.6576, + "step": 39790 + }, + { + "epoch": 0.932763703065572, + "grad_norm": 2.265047311782837, + "learning_rate": 2.2232815996830912e-06, + "loss": 0.1375, + "step": 39791 + }, + { + "epoch": 0.9327871446403769, + "grad_norm": 4.1009016036987305, + "learning_rate": 2.2217372300647463e-06, + "loss": 0.4257, + "step": 39792 + }, + { + "epoch": 0.932810586215182, + "grad_norm": 2.8396589756011963, + "learning_rate": 2.2201933909940074e-06, + "loss": 0.2635, + "step": 39793 + }, + { + "epoch": 0.9328340277899869, + "grad_norm": 1.7632681131362915, + "learning_rate": 2.2186500824792457e-06, + "loss": 0.0996, + "step": 39794 + }, + { + "epoch": 0.932857469364792, + "grad_norm": 1.9635441303253174, + "learning_rate": 2.2171073045288316e-06, + "loss": 0.1555, + "step": 39795 + }, + { + "epoch": 0.9328809109395969, + "grad_norm": 1.862685203552246, + "learning_rate": 2.2155650571511367e-06, + "loss": 0.2449, + "step": 39796 + }, + { + "epoch": 0.932904352514402, + "grad_norm": 0.8848963975906372, + "learning_rate": 2.214023340354554e-06, + "loss": 0.0425, + "step": 39797 + }, + { + "epoch": 0.9329277940892069, + "grad_norm": 1.9440470933914185, + "learning_rate": 2.2124821541474217e-06, + "loss": 1.0036, + "step": 39798 + }, + { + "epoch": 0.932951235664012, + "grad_norm": 2.3340015411376953, + "learning_rate": 2.2109414985381216e-06, + "loss": 0.2622, + "step": 39799 + }, + { + "epoch": 0.9329746772388169, + "grad_norm": 2.0276503562927246, + "learning_rate": 2.2094013735349917e-06, + "loss": 0.1893, + "step": 39800 + }, + { + "epoch": 0.9329981188136219, + "grad_norm": 2.194387197494507, + "learning_rate": 2.207861779146403e-06, + "loss": 0.1451, + "step": 39801 + }, + { + "epoch": 0.9330215603884269, + "grad_norm": 1.067723035812378, + "learning_rate": 2.2063227153806933e-06, + "loss": 0.0715, + "step": 39802 + }, + { + "epoch": 0.9330450019632319, + "grad_norm": 4.583799362182617, + "learning_rate": 2.2047841822462444e-06, + "loss": 0.3922, + "step": 39803 + }, + { + "epoch": 0.9330684435380369, + "grad_norm": 4.675017356872559, + "learning_rate": 2.2032461797513837e-06, + "loss": 0.3169, + "step": 39804 + }, + { + "epoch": 0.9330918851128419, + "grad_norm": 2.6392478942871094, + "learning_rate": 2.2017087079044486e-06, + "loss": 0.2188, + "step": 39805 + }, + { + "epoch": 0.9331153266876469, + "grad_norm": 1.226931095123291, + "learning_rate": 2.200171766713799e-06, + "loss": 0.0907, + "step": 39806 + }, + { + "epoch": 0.9331387682624519, + "grad_norm": 3.4103024005889893, + "learning_rate": 2.1986353561877727e-06, + "loss": 0.1713, + "step": 39807 + }, + { + "epoch": 0.9331622098372568, + "grad_norm": 3.1912131309509277, + "learning_rate": 2.197099476334685e-06, + "loss": 0.2349, + "step": 39808 + }, + { + "epoch": 0.9331856514120619, + "grad_norm": 2.3225302696228027, + "learning_rate": 2.195564127162886e-06, + "loss": 0.1294, + "step": 39809 + }, + { + "epoch": 0.9332090929868668, + "grad_norm": 2.0008983612060547, + "learning_rate": 2.1940293086807228e-06, + "loss": 0.1382, + "step": 39810 + }, + { + "epoch": 0.9332325345616719, + "grad_norm": 0.8991118669509888, + "learning_rate": 2.1924950208965013e-06, + "loss": 0.0718, + "step": 39811 + }, + { + "epoch": 0.9332559761364768, + "grad_norm": 2.568192720413208, + "learning_rate": 2.1909612638185473e-06, + "loss": 1.6003, + "step": 39812 + }, + { + "epoch": 0.9332794177112819, + "grad_norm": 2.558297634124756, + "learning_rate": 2.189428037455188e-06, + "loss": 0.1257, + "step": 39813 + }, + { + "epoch": 0.9333028592860868, + "grad_norm": 2.6145570278167725, + "learning_rate": 2.1878953418147498e-06, + "loss": 0.1837, + "step": 39814 + }, + { + "epoch": 0.9333263008608919, + "grad_norm": 1.6327106952667236, + "learning_rate": 2.186363176905526e-06, + "loss": 0.2155, + "step": 39815 + }, + { + "epoch": 0.9333497424356968, + "grad_norm": 2.651205062866211, + "learning_rate": 2.1848315427358655e-06, + "loss": 0.3619, + "step": 39816 + }, + { + "epoch": 0.9333731840105018, + "grad_norm": 0.9791213870048523, + "learning_rate": 2.183300439314051e-06, + "loss": 0.0737, + "step": 39817 + }, + { + "epoch": 0.9333966255853068, + "grad_norm": 2.6376800537109375, + "learning_rate": 2.1817698666483976e-06, + "loss": 0.209, + "step": 39818 + }, + { + "epoch": 0.9334200671601118, + "grad_norm": 1.3919427394866943, + "learning_rate": 2.18023982474721e-06, + "loss": 0.1433, + "step": 39819 + }, + { + "epoch": 0.9334435087349168, + "grad_norm": 3.9628684520721436, + "learning_rate": 2.1787103136187923e-06, + "loss": 0.3689, + "step": 39820 + }, + { + "epoch": 0.9334669503097218, + "grad_norm": 1.1123791933059692, + "learning_rate": 2.1771813332714385e-06, + "loss": 0.0939, + "step": 39821 + }, + { + "epoch": 0.9334903918845268, + "grad_norm": 1.1417310237884521, + "learning_rate": 2.1756528837134525e-06, + "loss": 0.03, + "step": 39822 + }, + { + "epoch": 0.9335138334593318, + "grad_norm": 1.054027795791626, + "learning_rate": 2.174124964953128e-06, + "loss": 0.1124, + "step": 39823 + }, + { + "epoch": 0.9335372750341367, + "grad_norm": 1.3329730033874512, + "learning_rate": 2.1725975769987474e-06, + "loss": 0.7408, + "step": 39824 + }, + { + "epoch": 0.9335607166089418, + "grad_norm": 3.546370267868042, + "learning_rate": 2.1710707198586035e-06, + "loss": 0.3669, + "step": 39825 + }, + { + "epoch": 0.9335841581837467, + "grad_norm": 3.0384318828582764, + "learning_rate": 2.16954439354099e-06, + "loss": 0.2319, + "step": 39826 + }, + { + "epoch": 0.9336075997585518, + "grad_norm": 0.5988396406173706, + "learning_rate": 2.1680185980541666e-06, + "loss": 0.0452, + "step": 39827 + }, + { + "epoch": 0.9336310413333567, + "grad_norm": 2.6396710872650146, + "learning_rate": 2.1664933334064274e-06, + "loss": 0.2663, + "step": 39828 + }, + { + "epoch": 0.9336544829081618, + "grad_norm": 3.330380439758301, + "learning_rate": 2.1649685996060543e-06, + "loss": 0.3645, + "step": 39829 + }, + { + "epoch": 0.9336779244829667, + "grad_norm": 1.740674376487732, + "learning_rate": 2.163444396661307e-06, + "loss": 0.0806, + "step": 39830 + }, + { + "epoch": 0.9337013660577718, + "grad_norm": 0.23405492305755615, + "learning_rate": 2.1619207245804685e-06, + "loss": 0.0109, + "step": 39831 + }, + { + "epoch": 0.9337248076325767, + "grad_norm": 2.6814053058624268, + "learning_rate": 2.1603975833717982e-06, + "loss": 0.2376, + "step": 39832 + }, + { + "epoch": 0.9337482492073818, + "grad_norm": 3.0830819606781006, + "learning_rate": 2.1588749730435455e-06, + "loss": 0.3183, + "step": 39833 + }, + { + "epoch": 0.9337716907821867, + "grad_norm": 0.7813961505889893, + "learning_rate": 2.157352893603992e-06, + "loss": 0.0234, + "step": 39834 + }, + { + "epoch": 0.9337951323569917, + "grad_norm": 2.1509647369384766, + "learning_rate": 2.1558313450613987e-06, + "loss": 0.2183, + "step": 39835 + }, + { + "epoch": 0.9338185739317968, + "grad_norm": 3.4521491527557373, + "learning_rate": 2.154310327424014e-06, + "loss": 0.2411, + "step": 39836 + }, + { + "epoch": 0.9338420155066017, + "grad_norm": 3.2736868858337402, + "learning_rate": 2.152789840700098e-06, + "loss": 0.3332, + "step": 39837 + }, + { + "epoch": 0.9338654570814068, + "grad_norm": 3.5779874324798584, + "learning_rate": 2.1512698848978995e-06, + "loss": 0.213, + "step": 39838 + }, + { + "epoch": 0.9338888986562117, + "grad_norm": 3.0198981761932373, + "learning_rate": 2.1497504600256566e-06, + "loss": 0.1597, + "step": 39839 + }, + { + "epoch": 0.9339123402310168, + "grad_norm": 3.362621545791626, + "learning_rate": 2.1482315660916074e-06, + "loss": 0.2184, + "step": 39840 + }, + { + "epoch": 0.9339357818058217, + "grad_norm": 0.7573781609535217, + "learning_rate": 2.1467132031040227e-06, + "loss": 0.05, + "step": 39841 + }, + { + "epoch": 0.9339592233806268, + "grad_norm": 0.7170029282569885, + "learning_rate": 2.1451953710711182e-06, + "loss": 0.0731, + "step": 39842 + }, + { + "epoch": 0.9339826649554317, + "grad_norm": 3.1235618591308594, + "learning_rate": 2.143678070001132e-06, + "loss": 0.2427, + "step": 39843 + }, + { + "epoch": 0.9340061065302367, + "grad_norm": 1.981977105140686, + "learning_rate": 2.142161299902301e-06, + "loss": 0.1437, + "step": 39844 + }, + { + "epoch": 0.9340295481050417, + "grad_norm": 3.109840154647827, + "learning_rate": 2.140645060782853e-06, + "loss": 1.1669, + "step": 39845 + }, + { + "epoch": 0.9340529896798467, + "grad_norm": 2.1301393508911133, + "learning_rate": 2.139129352651015e-06, + "loss": 0.1381, + "step": 39846 + }, + { + "epoch": 0.9340764312546517, + "grad_norm": 2.9917829036712646, + "learning_rate": 2.1376141755150125e-06, + "loss": 0.2066, + "step": 39847 + }, + { + "epoch": 0.9340998728294567, + "grad_norm": 2.066042423248291, + "learning_rate": 2.1360995293830733e-06, + "loss": 0.1228, + "step": 39848 + }, + { + "epoch": 0.9341233144042617, + "grad_norm": 3.0992419719696045, + "learning_rate": 2.1345854142634015e-06, + "loss": 0.2434, + "step": 39849 + }, + { + "epoch": 0.9341467559790667, + "grad_norm": 3.655170202255249, + "learning_rate": 2.133071830164235e-06, + "loss": 0.2808, + "step": 39850 + }, + { + "epoch": 0.9341701975538717, + "grad_norm": 2.94997501373291, + "learning_rate": 2.1315587770937563e-06, + "loss": 0.1881, + "step": 39851 + }, + { + "epoch": 0.9341936391286767, + "grad_norm": 2.42604660987854, + "learning_rate": 2.130046255060203e-06, + "loss": 1.4389, + "step": 39852 + }, + { + "epoch": 0.9342170807034816, + "grad_norm": 0.7710070610046387, + "learning_rate": 2.1285342640717575e-06, + "loss": 0.049, + "step": 39853 + }, + { + "epoch": 0.9342405222782867, + "grad_norm": 2.712317705154419, + "learning_rate": 2.127022804136647e-06, + "loss": 0.3584, + "step": 39854 + }, + { + "epoch": 0.9342639638530916, + "grad_norm": 2.4152443408966064, + "learning_rate": 2.125511875263064e-06, + "loss": 0.1441, + "step": 39855 + }, + { + "epoch": 0.9342874054278967, + "grad_norm": 0.38847365975379944, + "learning_rate": 2.1240014774592033e-06, + "loss": 0.0265, + "step": 39856 + }, + { + "epoch": 0.9343108470027016, + "grad_norm": 4.552140712738037, + "learning_rate": 2.1224916107332572e-06, + "loss": 0.1957, + "step": 39857 + }, + { + "epoch": 0.9343342885775067, + "grad_norm": 2.9643263816833496, + "learning_rate": 2.1209822750934306e-06, + "loss": 1.2906, + "step": 39858 + }, + { + "epoch": 0.9343577301523116, + "grad_norm": 2.1444993019104004, + "learning_rate": 2.1194734705479057e-06, + "loss": 0.1036, + "step": 39859 + }, + { + "epoch": 0.9343811717271167, + "grad_norm": 4.573276042938232, + "learning_rate": 2.1179651971048654e-06, + "loss": 0.4457, + "step": 39860 + }, + { + "epoch": 0.9344046133019216, + "grad_norm": 2.9273784160614014, + "learning_rate": 2.1164574547725025e-06, + "loss": 0.2436, + "step": 39861 + }, + { + "epoch": 0.9344280548767266, + "grad_norm": 0.784702479839325, + "learning_rate": 2.1149502435589995e-06, + "loss": 0.0596, + "step": 39862 + }, + { + "epoch": 0.9344514964515316, + "grad_norm": 2.6084847450256348, + "learning_rate": 2.113443563472517e-06, + "loss": 0.2711, + "step": 39863 + }, + { + "epoch": 0.9344749380263366, + "grad_norm": 1.682746171951294, + "learning_rate": 2.111937414521248e-06, + "loss": 1.0522, + "step": 39864 + }, + { + "epoch": 0.9344983796011416, + "grad_norm": 2.1663835048675537, + "learning_rate": 2.1104317967133524e-06, + "loss": 0.1792, + "step": 39865 + }, + { + "epoch": 0.9345218211759466, + "grad_norm": 0.4265560805797577, + "learning_rate": 2.1089267100570133e-06, + "loss": 0.0169, + "step": 39866 + }, + { + "epoch": 0.9345452627507516, + "grad_norm": 2.802133083343506, + "learning_rate": 2.1074221545603903e-06, + "loss": 0.3903, + "step": 39867 + }, + { + "epoch": 0.9345687043255566, + "grad_norm": 0.7561486959457397, + "learning_rate": 2.1059181302316546e-06, + "loss": 0.0647, + "step": 39868 + }, + { + "epoch": 0.9345921459003615, + "grad_norm": 3.3940205574035645, + "learning_rate": 2.104414637078944e-06, + "loss": 0.2257, + "step": 39869 + }, + { + "epoch": 0.9346155874751666, + "grad_norm": 0.7907858490943909, + "learning_rate": 2.1029116751104416e-06, + "loss": 0.0823, + "step": 39870 + }, + { + "epoch": 0.9346390290499715, + "grad_norm": 0.777735710144043, + "learning_rate": 2.101409244334285e-06, + "loss": 0.0599, + "step": 39871 + }, + { + "epoch": 0.9346624706247766, + "grad_norm": 1.1227798461914062, + "learning_rate": 2.0999073447586336e-06, + "loss": 0.1037, + "step": 39872 + }, + { + "epoch": 0.9346859121995815, + "grad_norm": 2.899550676345825, + "learning_rate": 2.0984059763916487e-06, + "loss": 0.3245, + "step": 39873 + }, + { + "epoch": 0.9347093537743866, + "grad_norm": 2.0306613445281982, + "learning_rate": 2.0969051392414674e-06, + "loss": 0.2553, + "step": 39874 + }, + { + "epoch": 0.9347327953491915, + "grad_norm": 2.5728015899658203, + "learning_rate": 2.095404833316217e-06, + "loss": 0.1893, + "step": 39875 + }, + { + "epoch": 0.9347562369239966, + "grad_norm": 3.3114640712738037, + "learning_rate": 2.0939050586240684e-06, + "loss": 0.3206, + "step": 39876 + }, + { + "epoch": 0.9347796784988015, + "grad_norm": 2.7647387981414795, + "learning_rate": 2.0924058151731374e-06, + "loss": 0.2387, + "step": 39877 + }, + { + "epoch": 0.9348031200736066, + "grad_norm": 2.6971302032470703, + "learning_rate": 2.090907102971562e-06, + "loss": 0.177, + "step": 39878 + }, + { + "epoch": 0.9348265616484115, + "grad_norm": 1.9788925647735596, + "learning_rate": 2.0894089220274805e-06, + "loss": 0.0886, + "step": 39879 + }, + { + "epoch": 0.9348500032232165, + "grad_norm": 2.7384278774261475, + "learning_rate": 2.0879112723490193e-06, + "loss": 0.2179, + "step": 39880 + }, + { + "epoch": 0.9348734447980215, + "grad_norm": 1.910575032234192, + "learning_rate": 2.086414153944294e-06, + "loss": 0.0927, + "step": 39881 + }, + { + "epoch": 0.9348968863728265, + "grad_norm": 2.164881944656372, + "learning_rate": 2.084917566821454e-06, + "loss": 0.1188, + "step": 39882 + }, + { + "epoch": 0.9349203279476315, + "grad_norm": 2.881963014602661, + "learning_rate": 2.0834215109886036e-06, + "loss": 0.2531, + "step": 39883 + }, + { + "epoch": 0.9349437695224365, + "grad_norm": 0.5923572182655334, + "learning_rate": 2.081925986453859e-06, + "loss": 0.0379, + "step": 39884 + }, + { + "epoch": 0.9349672110972415, + "grad_norm": 3.495506763458252, + "learning_rate": 2.080430993225324e-06, + "loss": 0.3028, + "step": 39885 + }, + { + "epoch": 0.9349906526720465, + "grad_norm": 2.915492534637451, + "learning_rate": 2.0789365313111374e-06, + "loss": 0.2075, + "step": 39886 + }, + { + "epoch": 0.9350140942468516, + "grad_norm": 2.2108519077301025, + "learning_rate": 2.0774426007193924e-06, + "loss": 0.2126, + "step": 39887 + }, + { + "epoch": 0.9350375358216565, + "grad_norm": 0.9881086349487305, + "learning_rate": 2.075949201458194e-06, + "loss": 0.0797, + "step": 39888 + }, + { + "epoch": 0.9350609773964615, + "grad_norm": 2.4744913578033447, + "learning_rate": 2.074456333535657e-06, + "loss": 0.0981, + "step": 39889 + }, + { + "epoch": 0.9350844189712665, + "grad_norm": 2.326519250869751, + "learning_rate": 2.0729639969598647e-06, + "loss": 1.5238, + "step": 39890 + }, + { + "epoch": 0.9351078605460715, + "grad_norm": 0.5015738010406494, + "learning_rate": 2.071472191738921e-06, + "loss": 0.0333, + "step": 39891 + }, + { + "epoch": 0.9351313021208765, + "grad_norm": 1.4278581142425537, + "learning_rate": 2.0699809178809314e-06, + "loss": 0.1913, + "step": 39892 + }, + { + "epoch": 0.9351547436956815, + "grad_norm": 3.0017316341400146, + "learning_rate": 2.0684901753939777e-06, + "loss": 0.3566, + "step": 39893 + }, + { + "epoch": 0.9351781852704865, + "grad_norm": 2.10717511177063, + "learning_rate": 2.0669999642861427e-06, + "loss": 0.1723, + "step": 39894 + }, + { + "epoch": 0.9352016268452915, + "grad_norm": 3.842796802520752, + "learning_rate": 2.0655102845655194e-06, + "loss": 0.2567, + "step": 39895 + }, + { + "epoch": 0.9352250684200965, + "grad_norm": 1.5843780040740967, + "learning_rate": 2.0640211362402017e-06, + "loss": 0.1369, + "step": 39896 + }, + { + "epoch": 0.9352485099949015, + "grad_norm": 3.247063398361206, + "learning_rate": 2.06253251931825e-06, + "loss": 0.2699, + "step": 39897 + }, + { + "epoch": 0.9352719515697064, + "grad_norm": 0.568550169467926, + "learning_rate": 2.0610444338077463e-06, + "loss": 0.0509, + "step": 39898 + }, + { + "epoch": 0.9352953931445115, + "grad_norm": 0.7672021389007568, + "learning_rate": 2.059556879716784e-06, + "loss": 0.0571, + "step": 39899 + }, + { + "epoch": 0.9353188347193164, + "grad_norm": 4.01094388961792, + "learning_rate": 2.058069857053413e-06, + "loss": 0.3284, + "step": 39900 + }, + { + "epoch": 0.9353422762941215, + "grad_norm": 1.6723248958587646, + "learning_rate": 2.056583365825715e-06, + "loss": 0.0984, + "step": 39901 + }, + { + "epoch": 0.9353657178689264, + "grad_norm": 2.581571102142334, + "learning_rate": 2.0550974060417394e-06, + "loss": 0.3139, + "step": 39902 + }, + { + "epoch": 0.9353891594437315, + "grad_norm": 0.4181279242038727, + "learning_rate": 2.0536119777095576e-06, + "loss": 0.0218, + "step": 39903 + }, + { + "epoch": 0.9354126010185364, + "grad_norm": 3.6066746711730957, + "learning_rate": 2.0521270808372406e-06, + "loss": 1.4792, + "step": 39904 + }, + { + "epoch": 0.9354360425933415, + "grad_norm": 0.37286925315856934, + "learning_rate": 2.0506427154328266e-06, + "loss": 0.0165, + "step": 39905 + }, + { + "epoch": 0.9354594841681464, + "grad_norm": 1.81426203250885, + "learning_rate": 2.049158881504387e-06, + "loss": 0.1333, + "step": 39906 + }, + { + "epoch": 0.9354829257429514, + "grad_norm": 4.13509464263916, + "learning_rate": 2.04767557905996e-06, + "loss": 0.1639, + "step": 39907 + }, + { + "epoch": 0.9355063673177564, + "grad_norm": 3.3043041229248047, + "learning_rate": 2.0461928081076055e-06, + "loss": 0.2112, + "step": 39908 + }, + { + "epoch": 0.9355298088925614, + "grad_norm": 3.93166446685791, + "learning_rate": 2.0447105686553616e-06, + "loss": 0.3979, + "step": 39909 + }, + { + "epoch": 0.9355532504673664, + "grad_norm": 1.0242621898651123, + "learning_rate": 2.043228860711266e-06, + "loss": 0.0481, + "step": 39910 + }, + { + "epoch": 0.9355766920421714, + "grad_norm": 0.6714812517166138, + "learning_rate": 2.041747684283368e-06, + "loss": 0.0394, + "step": 39911 + }, + { + "epoch": 0.9356001336169764, + "grad_norm": 2.8099050521850586, + "learning_rate": 2.0402670393797064e-06, + "loss": 0.1299, + "step": 39912 + }, + { + "epoch": 0.9356235751917814, + "grad_norm": 2.425891160964966, + "learning_rate": 2.038786926008307e-06, + "loss": 0.1937, + "step": 39913 + }, + { + "epoch": 0.9356470167665863, + "grad_norm": 1.722400426864624, + "learning_rate": 2.0373073441772085e-06, + "loss": 1.6273, + "step": 39914 + }, + { + "epoch": 0.9356704583413914, + "grad_norm": 2.746492624282837, + "learning_rate": 2.0358282938944372e-06, + "loss": 0.2027, + "step": 39915 + }, + { + "epoch": 0.9356938999161963, + "grad_norm": 1.7259832620620728, + "learning_rate": 2.0343497751679984e-06, + "loss": 0.1308, + "step": 39916 + }, + { + "epoch": 0.9357173414910014, + "grad_norm": 1.4146089553833008, + "learning_rate": 2.0328717880059413e-06, + "loss": 0.0696, + "step": 39917 + }, + { + "epoch": 0.9357407830658063, + "grad_norm": 2.1929311752319336, + "learning_rate": 2.03139433241627e-06, + "loss": 0.9541, + "step": 39918 + }, + { + "epoch": 0.9357642246406114, + "grad_norm": 2.755699396133423, + "learning_rate": 2.029917408407023e-06, + "loss": 0.1878, + "step": 39919 + }, + { + "epoch": 0.9357876662154163, + "grad_norm": 0.5014148354530334, + "learning_rate": 2.0284410159861934e-06, + "loss": 0.0363, + "step": 39920 + }, + { + "epoch": 0.9358111077902214, + "grad_norm": 2.247346878051758, + "learning_rate": 2.0269651551617974e-06, + "loss": 1.2039, + "step": 39921 + }, + { + "epoch": 0.9358345493650263, + "grad_norm": 2.861927032470703, + "learning_rate": 2.02548982594184e-06, + "loss": 0.1679, + "step": 39922 + }, + { + "epoch": 0.9358579909398314, + "grad_norm": 3.2326738834381104, + "learning_rate": 2.024015028334325e-06, + "loss": 0.2318, + "step": 39923 + }, + { + "epoch": 0.9358814325146363, + "grad_norm": 2.047293186187744, + "learning_rate": 2.022540762347269e-06, + "loss": 0.1044, + "step": 39924 + }, + { + "epoch": 0.9359048740894413, + "grad_norm": 1.238337516784668, + "learning_rate": 2.0210670279886657e-06, + "loss": 0.0539, + "step": 39925 + }, + { + "epoch": 0.9359283156642463, + "grad_norm": 2.315781354904175, + "learning_rate": 2.019593825266497e-06, + "loss": 0.2281, + "step": 39926 + }, + { + "epoch": 0.9359517572390513, + "grad_norm": 2.860447645187378, + "learning_rate": 2.018121154188779e-06, + "loss": 0.2505, + "step": 39927 + }, + { + "epoch": 0.9359751988138563, + "grad_norm": 0.8134136199951172, + "learning_rate": 2.016649014763483e-06, + "loss": 0.0289, + "step": 39928 + }, + { + "epoch": 0.9359986403886613, + "grad_norm": 2.4259870052337646, + "learning_rate": 2.015177406998592e-06, + "loss": 0.1538, + "step": 39929 + }, + { + "epoch": 0.9360220819634663, + "grad_norm": 0.8246551752090454, + "learning_rate": 2.0137063309021098e-06, + "loss": 0.0852, + "step": 39930 + }, + { + "epoch": 0.9360455235382713, + "grad_norm": 1.8055390119552612, + "learning_rate": 2.0122357864820198e-06, + "loss": 1.3526, + "step": 39931 + }, + { + "epoch": 0.9360689651130762, + "grad_norm": 1.29932701587677, + "learning_rate": 2.0107657737462925e-06, + "loss": 0.1306, + "step": 39932 + }, + { + "epoch": 0.9360924066878813, + "grad_norm": 3.401294469833374, + "learning_rate": 2.0092962927029112e-06, + "loss": 0.3758, + "step": 39933 + }, + { + "epoch": 0.9361158482626862, + "grad_norm": 1.068268895149231, + "learning_rate": 2.007827343359836e-06, + "loss": 0.1141, + "step": 39934 + }, + { + "epoch": 0.9361392898374913, + "grad_norm": 1.7489606142044067, + "learning_rate": 2.006358925725049e-06, + "loss": 0.0492, + "step": 39935 + }, + { + "epoch": 0.9361627314122962, + "grad_norm": 3.5611486434936523, + "learning_rate": 2.0048910398065e-06, + "loss": 0.2786, + "step": 39936 + }, + { + "epoch": 0.9361861729871013, + "grad_norm": 1.2376682758331299, + "learning_rate": 2.0034236856121825e-06, + "loss": 0.0351, + "step": 39937 + }, + { + "epoch": 0.9362096145619063, + "grad_norm": 2.9727585315704346, + "learning_rate": 2.001956863150034e-06, + "loss": 0.2406, + "step": 39938 + }, + { + "epoch": 0.9362330561367113, + "grad_norm": 1.7452137470245361, + "learning_rate": 2.0004905724280378e-06, + "loss": 0.0507, + "step": 39939 + }, + { + "epoch": 0.9362564977115163, + "grad_norm": 1.746225118637085, + "learning_rate": 1.99902481345412e-06, + "loss": 1.0284, + "step": 39940 + }, + { + "epoch": 0.9362799392863212, + "grad_norm": 0.5588154196739197, + "learning_rate": 1.997559586236253e-06, + "loss": 0.0582, + "step": 39941 + }, + { + "epoch": 0.9363033808611263, + "grad_norm": 2.469705820083618, + "learning_rate": 1.996094890782385e-06, + "loss": 1.873, + "step": 39942 + }, + { + "epoch": 0.9363268224359312, + "grad_norm": 3.4481899738311768, + "learning_rate": 1.994630727100455e-06, + "loss": 0.2822, + "step": 39943 + }, + { + "epoch": 0.9363502640107363, + "grad_norm": 1.8885831832885742, + "learning_rate": 1.9931670951984225e-06, + "loss": 0.1406, + "step": 39944 + }, + { + "epoch": 0.9363737055855412, + "grad_norm": 2.2462732791900635, + "learning_rate": 1.9917039950842265e-06, + "loss": 0.059, + "step": 39945 + }, + { + "epoch": 0.9363971471603463, + "grad_norm": 3.2455296516418457, + "learning_rate": 1.9902414267657933e-06, + "loss": 0.2188, + "step": 39946 + }, + { + "epoch": 0.9364205887351512, + "grad_norm": 2.174377202987671, + "learning_rate": 1.988779390251061e-06, + "loss": 0.1263, + "step": 39947 + }, + { + "epoch": 0.9364440303099563, + "grad_norm": 3.5223357677459717, + "learning_rate": 1.987317885547968e-06, + "loss": 0.2957, + "step": 39948 + }, + { + "epoch": 0.9364674718847612, + "grad_norm": 2.4048378467559814, + "learning_rate": 1.9858569126644523e-06, + "loss": 0.3426, + "step": 39949 + }, + { + "epoch": 0.9364909134595663, + "grad_norm": 1.172178864479065, + "learning_rate": 1.9843964716084297e-06, + "loss": 0.0874, + "step": 39950 + }, + { + "epoch": 0.9365143550343712, + "grad_norm": 2.488445520401001, + "learning_rate": 1.982936562387827e-06, + "loss": 1.4346, + "step": 39951 + }, + { + "epoch": 0.9365377966091762, + "grad_norm": 2.2899303436279297, + "learning_rate": 1.981477185010572e-06, + "loss": 0.0714, + "step": 39952 + }, + { + "epoch": 0.9365612381839812, + "grad_norm": 0.7484351992607117, + "learning_rate": 1.9800183394845683e-06, + "loss": 0.0753, + "step": 39953 + }, + { + "epoch": 0.9365846797587862, + "grad_norm": 2.004359722137451, + "learning_rate": 1.978560025817744e-06, + "loss": 1.3278, + "step": 39954 + }, + { + "epoch": 0.9366081213335912, + "grad_norm": 2.5438148975372314, + "learning_rate": 1.977102244018003e-06, + "loss": 0.1652, + "step": 39955 + }, + { + "epoch": 0.9366315629083962, + "grad_norm": 0.8234058022499084, + "learning_rate": 1.9756449940932733e-06, + "loss": 0.0362, + "step": 39956 + }, + { + "epoch": 0.9366550044832012, + "grad_norm": 1.8902641534805298, + "learning_rate": 1.9741882760514474e-06, + "loss": 0.1425, + "step": 39957 + }, + { + "epoch": 0.9366784460580062, + "grad_norm": 2.1532490253448486, + "learning_rate": 1.9727320899004308e-06, + "loss": 0.098, + "step": 39958 + }, + { + "epoch": 0.9367018876328111, + "grad_norm": 1.0037811994552612, + "learning_rate": 1.9712764356481283e-06, + "loss": 0.1086, + "step": 39959 + }, + { + "epoch": 0.9367253292076162, + "grad_norm": 1.7072595357894897, + "learning_rate": 1.9698213133024334e-06, + "loss": 0.1622, + "step": 39960 + }, + { + "epoch": 0.9367487707824211, + "grad_norm": 2.605059862136841, + "learning_rate": 1.9683667228712397e-06, + "loss": 0.1279, + "step": 39961 + }, + { + "epoch": 0.9367722123572262, + "grad_norm": 2.5385656356811523, + "learning_rate": 1.9669126643624525e-06, + "loss": 0.1204, + "step": 39962 + }, + { + "epoch": 0.9367956539320311, + "grad_norm": 1.1646074056625366, + "learning_rate": 1.965459137783954e-06, + "loss": 0.0426, + "step": 39963 + }, + { + "epoch": 0.9368190955068362, + "grad_norm": 2.2079951763153076, + "learning_rate": 1.9640061431436264e-06, + "loss": 0.0943, + "step": 39964 + }, + { + "epoch": 0.9368425370816411, + "grad_norm": 4.131021976470947, + "learning_rate": 1.962553680449364e-06, + "loss": 0.2653, + "step": 39965 + }, + { + "epoch": 0.9368659786564462, + "grad_norm": 0.7251577377319336, + "learning_rate": 1.9611017497090378e-06, + "loss": 0.0712, + "step": 39966 + }, + { + "epoch": 0.9368894202312511, + "grad_norm": 4.498300552368164, + "learning_rate": 1.9596503509305196e-06, + "loss": 0.3624, + "step": 39967 + }, + { + "epoch": 0.9369128618060562, + "grad_norm": 1.5828511714935303, + "learning_rate": 1.9581994841217034e-06, + "loss": 0.1435, + "step": 39968 + }, + { + "epoch": 0.9369363033808611, + "grad_norm": 1.7073184251785278, + "learning_rate": 1.956749149290449e-06, + "loss": 0.069, + "step": 39969 + }, + { + "epoch": 0.9369597449556661, + "grad_norm": 3.64929461479187, + "learning_rate": 1.955299346444639e-06, + "loss": 0.1657, + "step": 39970 + }, + { + "epoch": 0.9369831865304711, + "grad_norm": 1.721095323562622, + "learning_rate": 1.953850075592134e-06, + "loss": 0.3155, + "step": 39971 + }, + { + "epoch": 0.9370066281052761, + "grad_norm": 2.623375654220581, + "learning_rate": 1.9524013367407946e-06, + "loss": 0.3218, + "step": 39972 + }, + { + "epoch": 0.9370300696800811, + "grad_norm": 0.4750109314918518, + "learning_rate": 1.9509531298984806e-06, + "loss": 0.036, + "step": 39973 + }, + { + "epoch": 0.9370535112548861, + "grad_norm": 0.9852405190467834, + "learning_rate": 1.949505455073042e-06, + "loss": 0.0644, + "step": 39974 + }, + { + "epoch": 0.9370769528296911, + "grad_norm": 1.9692814350128174, + "learning_rate": 1.9480583122723495e-06, + "loss": 0.1778, + "step": 39975 + }, + { + "epoch": 0.9371003944044961, + "grad_norm": 1.045747995376587, + "learning_rate": 1.946611701504253e-06, + "loss": 0.0895, + "step": 39976 + }, + { + "epoch": 0.937123835979301, + "grad_norm": 2.939464569091797, + "learning_rate": 1.9451656227766014e-06, + "loss": 0.2204, + "step": 39977 + }, + { + "epoch": 0.9371472775541061, + "grad_norm": 4.187961578369141, + "learning_rate": 1.943720076097233e-06, + "loss": 0.2411, + "step": 39978 + }, + { + "epoch": 0.937170719128911, + "grad_norm": 1.6407283544540405, + "learning_rate": 1.942275061473986e-06, + "loss": 0.1354, + "step": 39979 + }, + { + "epoch": 0.9371941607037161, + "grad_norm": 0.6109353303909302, + "learning_rate": 1.940830578914732e-06, + "loss": 0.0289, + "step": 39980 + }, + { + "epoch": 0.937217602278521, + "grad_norm": 2.7839741706848145, + "learning_rate": 1.9393866284272644e-06, + "loss": 0.2697, + "step": 39981 + }, + { + "epoch": 0.9372410438533261, + "grad_norm": 2.389192819595337, + "learning_rate": 1.937943210019466e-06, + "loss": 0.2549, + "step": 39982 + }, + { + "epoch": 0.937264485428131, + "grad_norm": 1.6758770942687988, + "learning_rate": 1.936500323699131e-06, + "loss": 0.0811, + "step": 39983 + }, + { + "epoch": 0.9372879270029361, + "grad_norm": 0.44302213191986084, + "learning_rate": 1.935057969474108e-06, + "loss": 0.0299, + "step": 39984 + }, + { + "epoch": 0.937311368577741, + "grad_norm": 3.1161277294158936, + "learning_rate": 1.933616147352224e-06, + "loss": 0.3761, + "step": 39985 + }, + { + "epoch": 0.937334810152546, + "grad_norm": 0.7353126406669617, + "learning_rate": 1.9321748573412845e-06, + "loss": 0.0322, + "step": 39986 + }, + { + "epoch": 0.937358251727351, + "grad_norm": 1.7096632719039917, + "learning_rate": 1.9307340994491275e-06, + "loss": 1.2707, + "step": 39987 + }, + { + "epoch": 0.937381693302156, + "grad_norm": 0.29145342111587524, + "learning_rate": 1.9292938736835577e-06, + "loss": 0.0099, + "step": 39988 + }, + { + "epoch": 0.9374051348769611, + "grad_norm": 4.017167568206787, + "learning_rate": 1.927854180052413e-06, + "loss": 0.446, + "step": 39989 + }, + { + "epoch": 0.937428576451766, + "grad_norm": 1.8766306638717651, + "learning_rate": 1.9264150185634768e-06, + "loss": 0.0652, + "step": 39990 + }, + { + "epoch": 0.9374520180265711, + "grad_norm": 3.9814465045928955, + "learning_rate": 1.924976389224564e-06, + "loss": 0.2872, + "step": 39991 + }, + { + "epoch": 0.937475459601376, + "grad_norm": 2.234199047088623, + "learning_rate": 1.923538292043503e-06, + "loss": 0.1136, + "step": 39992 + }, + { + "epoch": 0.9374989011761811, + "grad_norm": 3.277667284011841, + "learning_rate": 1.9221007270280643e-06, + "loss": 0.299, + "step": 39993 + }, + { + "epoch": 0.937522342750986, + "grad_norm": 0.35174626111984253, + "learning_rate": 1.9206636941860757e-06, + "loss": 0.0331, + "step": 39994 + }, + { + "epoch": 0.937545784325791, + "grad_norm": 7.576033592224121, + "learning_rate": 1.9192271935253193e-06, + "loss": 0.1346, + "step": 39995 + }, + { + "epoch": 0.937569225900596, + "grad_norm": 4.04686975479126, + "learning_rate": 1.9177912250536e-06, + "loss": 0.257, + "step": 39996 + }, + { + "epoch": 0.937592667475401, + "grad_norm": 5.192465782165527, + "learning_rate": 1.916355788778701e-06, + "loss": 0.3718, + "step": 39997 + }, + { + "epoch": 0.937616109050206, + "grad_norm": 18.46865463256836, + "learning_rate": 1.9149208847084156e-06, + "loss": 0.3314, + "step": 39998 + }, + { + "epoch": 0.937639550625011, + "grad_norm": 1.776960015296936, + "learning_rate": 1.9134865128505153e-06, + "loss": 0.1538, + "step": 39999 + }, + { + "epoch": 0.937662992199816, + "grad_norm": 2.8224730491638184, + "learning_rate": 1.912052673212805e-06, + "loss": 0.2685, + "step": 40000 + }, + { + "epoch": 0.937686433774621, + "grad_norm": 0.400107741355896, + "learning_rate": 1.910619365803057e-06, + "loss": 0.0382, + "step": 40001 + }, + { + "epoch": 0.937709875349426, + "grad_norm": 3.8519017696380615, + "learning_rate": 1.909186590629042e-06, + "loss": 0.2134, + "step": 40002 + }, + { + "epoch": 0.937733316924231, + "grad_norm": 6.280003547668457, + "learning_rate": 1.9077543476985314e-06, + "loss": 0.2537, + "step": 40003 + }, + { + "epoch": 0.937756758499036, + "grad_norm": 1.781467318534851, + "learning_rate": 1.9063226370193199e-06, + "loss": 0.1465, + "step": 40004 + }, + { + "epoch": 0.937780200073841, + "grad_norm": 0.7659076452255249, + "learning_rate": 1.904891458599145e-06, + "loss": 0.0649, + "step": 40005 + }, + { + "epoch": 0.9378036416486459, + "grad_norm": 1.6805013418197632, + "learning_rate": 1.9034608124457898e-06, + "loss": 1.2467, + "step": 40006 + }, + { + "epoch": 0.937827083223451, + "grad_norm": 2.729794979095459, + "learning_rate": 1.9020306985670256e-06, + "loss": 0.2442, + "step": 40007 + }, + { + "epoch": 0.9378505247982559, + "grad_norm": 5.817837238311768, + "learning_rate": 1.9006011169705906e-06, + "loss": 0.1556, + "step": 40008 + }, + { + "epoch": 0.937873966373061, + "grad_norm": 2.092250347137451, + "learning_rate": 1.8991720676642565e-06, + "loss": 0.1421, + "step": 40009 + }, + { + "epoch": 0.9378974079478659, + "grad_norm": 2.387375831604004, + "learning_rate": 1.8977435506557728e-06, + "loss": 0.1775, + "step": 40010 + }, + { + "epoch": 0.937920849522671, + "grad_norm": 1.7808358669281006, + "learning_rate": 1.8963155659528998e-06, + "loss": 1.0304, + "step": 40011 + }, + { + "epoch": 0.9379442910974759, + "grad_norm": 3.6971354484558105, + "learning_rate": 1.8948881135633534e-06, + "loss": 0.1775, + "step": 40012 + }, + { + "epoch": 0.937967732672281, + "grad_norm": 3.6102874279022217, + "learning_rate": 1.8934611934949276e-06, + "loss": 0.2099, + "step": 40013 + }, + { + "epoch": 0.9379911742470859, + "grad_norm": 3.887075424194336, + "learning_rate": 1.8920348057553272e-06, + "loss": 0.1746, + "step": 40014 + }, + { + "epoch": 0.9380146158218909, + "grad_norm": 2.588465929031372, + "learning_rate": 1.8906089503523018e-06, + "loss": 0.1956, + "step": 40015 + }, + { + "epoch": 0.9380380573966959, + "grad_norm": 2.572960615158081, + "learning_rate": 1.8891836272936003e-06, + "loss": 1.2691, + "step": 40016 + }, + { + "epoch": 0.9380614989715009, + "grad_norm": 0.4411352276802063, + "learning_rate": 1.8877588365869391e-06, + "loss": 0.0337, + "step": 40017 + }, + { + "epoch": 0.9380849405463059, + "grad_norm": 1.0245364904403687, + "learning_rate": 1.8863345782400565e-06, + "loss": 0.1066, + "step": 40018 + }, + { + "epoch": 0.9381083821211109, + "grad_norm": 3.2180185317993164, + "learning_rate": 1.8849108522606796e-06, + "loss": 0.2976, + "step": 40019 + }, + { + "epoch": 0.9381318236959159, + "grad_norm": 2.4229817390441895, + "learning_rate": 1.8834876586565464e-06, + "loss": 0.1665, + "step": 40020 + }, + { + "epoch": 0.9381552652707209, + "grad_norm": 1.4414383172988892, + "learning_rate": 1.8820649974353622e-06, + "loss": 0.0595, + "step": 40021 + }, + { + "epoch": 0.9381787068455258, + "grad_norm": 1.9941860437393188, + "learning_rate": 1.8806428686048539e-06, + "loss": 0.0808, + "step": 40022 + }, + { + "epoch": 0.9382021484203309, + "grad_norm": 1.1421618461608887, + "learning_rate": 1.8792212721727376e-06, + "loss": 0.0978, + "step": 40023 + }, + { + "epoch": 0.9382255899951358, + "grad_norm": 2.412876605987549, + "learning_rate": 1.8778002081467183e-06, + "loss": 0.1507, + "step": 40024 + }, + { + "epoch": 0.9382490315699409, + "grad_norm": 1.9898430109024048, + "learning_rate": 1.8763796765345232e-06, + "loss": 0.2768, + "step": 40025 + }, + { + "epoch": 0.9382724731447458, + "grad_norm": 3.674837827682495, + "learning_rate": 1.874959677343846e-06, + "loss": 0.1917, + "step": 40026 + }, + { + "epoch": 0.9382959147195509, + "grad_norm": 1.8721166849136353, + "learning_rate": 1.8735402105823918e-06, + "loss": 0.0621, + "step": 40027 + }, + { + "epoch": 0.9383193562943558, + "grad_norm": 2.8061587810516357, + "learning_rate": 1.8721212762578656e-06, + "loss": 0.1934, + "step": 40028 + }, + { + "epoch": 0.9383427978691609, + "grad_norm": 0.7083305716514587, + "learning_rate": 1.8707028743779831e-06, + "loss": 0.0454, + "step": 40029 + }, + { + "epoch": 0.9383662394439658, + "grad_norm": 0.5925416350364685, + "learning_rate": 1.8692850049504163e-06, + "loss": 0.0756, + "step": 40030 + }, + { + "epoch": 0.9383896810187708, + "grad_norm": 1.1844524145126343, + "learning_rate": 1.86786766798287e-06, + "loss": 0.0623, + "step": 40031 + }, + { + "epoch": 0.9384131225935758, + "grad_norm": 2.5146374702453613, + "learning_rate": 1.8664508634830269e-06, + "loss": 0.2545, + "step": 40032 + }, + { + "epoch": 0.9384365641683808, + "grad_norm": 0.6610877513885498, + "learning_rate": 1.865034591458592e-06, + "loss": 0.0532, + "step": 40033 + }, + { + "epoch": 0.9384600057431858, + "grad_norm": 3.161092758178711, + "learning_rate": 1.863618851917237e-06, + "loss": 0.2425, + "step": 40034 + }, + { + "epoch": 0.9384834473179908, + "grad_norm": 0.6626964807510376, + "learning_rate": 1.8622036448666447e-06, + "loss": 0.0552, + "step": 40035 + }, + { + "epoch": 0.9385068888927958, + "grad_norm": 1.8528186082839966, + "learning_rate": 1.8607889703144864e-06, + "loss": 0.9905, + "step": 40036 + }, + { + "epoch": 0.9385303304676008, + "grad_norm": 2.976227283477783, + "learning_rate": 1.8593748282684454e-06, + "loss": 0.2568, + "step": 40037 + }, + { + "epoch": 0.9385537720424058, + "grad_norm": 1.9649766683578491, + "learning_rate": 1.857961218736215e-06, + "loss": 0.1751, + "step": 40038 + }, + { + "epoch": 0.9385772136172108, + "grad_norm": 0.436717689037323, + "learning_rate": 1.8565481417254337e-06, + "loss": 0.0164, + "step": 40039 + }, + { + "epoch": 0.9386006551920159, + "grad_norm": 0.741974949836731, + "learning_rate": 1.8551355972437734e-06, + "loss": 0.0511, + "step": 40040 + }, + { + "epoch": 0.9386240967668208, + "grad_norm": 2.3338000774383545, + "learning_rate": 1.8537235852989165e-06, + "loss": 0.1889, + "step": 40041 + }, + { + "epoch": 0.9386475383416258, + "grad_norm": 0.7958022952079773, + "learning_rate": 1.8523121058985127e-06, + "loss": 0.0595, + "step": 40042 + }, + { + "epoch": 0.9386709799164308, + "grad_norm": 3.584517240524292, + "learning_rate": 1.8509011590502335e-06, + "loss": 0.3175, + "step": 40043 + }, + { + "epoch": 0.9386944214912358, + "grad_norm": 4.388113021850586, + "learning_rate": 1.849490744761706e-06, + "loss": 0.3242, + "step": 40044 + }, + { + "epoch": 0.9387178630660408, + "grad_norm": 1.7181252241134644, + "learning_rate": 1.8480808630406132e-06, + "loss": 0.0451, + "step": 40045 + }, + { + "epoch": 0.9387413046408458, + "grad_norm": 2.737074375152588, + "learning_rate": 1.8466715138945934e-06, + "loss": 0.1756, + "step": 40046 + }, + { + "epoch": 0.9387647462156508, + "grad_norm": 2.758497953414917, + "learning_rate": 1.8452626973312958e-06, + "loss": 0.2823, + "step": 40047 + }, + { + "epoch": 0.9387881877904558, + "grad_norm": 3.6527864933013916, + "learning_rate": 1.843854413358359e-06, + "loss": 0.3148, + "step": 40048 + }, + { + "epoch": 0.9388116293652607, + "grad_norm": 2.5571835041046143, + "learning_rate": 1.8424466619834325e-06, + "loss": 0.2289, + "step": 40049 + }, + { + "epoch": 0.9388350709400658, + "grad_norm": 2.2476611137390137, + "learning_rate": 1.8410394432141321e-06, + "loss": 0.1515, + "step": 40050 + }, + { + "epoch": 0.9388585125148707, + "grad_norm": 2.4450488090515137, + "learning_rate": 1.8396327570581296e-06, + "loss": 0.1431, + "step": 40051 + }, + { + "epoch": 0.9388819540896758, + "grad_norm": 3.1692922115325928, + "learning_rate": 1.83822660352303e-06, + "loss": 0.3245, + "step": 40052 + }, + { + "epoch": 0.9389053956644807, + "grad_norm": 5.017033576965332, + "learning_rate": 1.8368209826164829e-06, + "loss": 0.1613, + "step": 40053 + }, + { + "epoch": 0.9389288372392858, + "grad_norm": 0.708585798740387, + "learning_rate": 1.8354158943461041e-06, + "loss": 0.05, + "step": 40054 + }, + { + "epoch": 0.9389522788140907, + "grad_norm": 1.6496707201004028, + "learning_rate": 1.8340113387195102e-06, + "loss": 0.0755, + "step": 40055 + }, + { + "epoch": 0.9389757203888958, + "grad_norm": 2.1473162174224854, + "learning_rate": 1.832607315744328e-06, + "loss": 0.099, + "step": 40056 + }, + { + "epoch": 0.9389991619637007, + "grad_norm": 2.342437982559204, + "learning_rate": 1.831203825428196e-06, + "loss": 0.1503, + "step": 40057 + }, + { + "epoch": 0.9390226035385058, + "grad_norm": 4.5127105712890625, + "learning_rate": 1.829800867778697e-06, + "loss": 0.4329, + "step": 40058 + }, + { + "epoch": 0.9390460451133107, + "grad_norm": 1.9885894060134888, + "learning_rate": 1.8283984428034694e-06, + "loss": 0.9807, + "step": 40059 + }, + { + "epoch": 0.9390694866881157, + "grad_norm": 3.4632930755615234, + "learning_rate": 1.8269965505101072e-06, + "loss": 0.2321, + "step": 40060 + }, + { + "epoch": 0.9390929282629207, + "grad_norm": 1.3570607900619507, + "learning_rate": 1.8255951909062264e-06, + "loss": 0.2009, + "step": 40061 + }, + { + "epoch": 0.9391163698377257, + "grad_norm": 1.5003983974456787, + "learning_rate": 1.8241943639994208e-06, + "loss": 0.0779, + "step": 40062 + }, + { + "epoch": 0.9391398114125307, + "grad_norm": 2.475058078765869, + "learning_rate": 1.8227940697972955e-06, + "loss": 0.1258, + "step": 40063 + }, + { + "epoch": 0.9391632529873357, + "grad_norm": 1.4977761507034302, + "learning_rate": 1.8213943083074447e-06, + "loss": 0.0432, + "step": 40064 + }, + { + "epoch": 0.9391866945621407, + "grad_norm": 2.0290277004241943, + "learning_rate": 1.8199950795374843e-06, + "loss": 1.386, + "step": 40065 + }, + { + "epoch": 0.9392101361369457, + "grad_norm": 3.055659532546997, + "learning_rate": 1.8185963834949859e-06, + "loss": 0.231, + "step": 40066 + }, + { + "epoch": 0.9392335777117506, + "grad_norm": 2.7154228687286377, + "learning_rate": 1.8171982201875437e-06, + "loss": 0.1824, + "step": 40067 + }, + { + "epoch": 0.9392570192865557, + "grad_norm": 4.787985324859619, + "learning_rate": 1.8158005896227515e-06, + "loss": 0.2059, + "step": 40068 + }, + { + "epoch": 0.9392804608613606, + "grad_norm": 3.0078580379486084, + "learning_rate": 1.8144034918081698e-06, + "loss": 1.284, + "step": 40069 + }, + { + "epoch": 0.9393039024361657, + "grad_norm": 1.632969856262207, + "learning_rate": 1.8130069267514037e-06, + "loss": 0.1138, + "step": 40070 + }, + { + "epoch": 0.9393273440109706, + "grad_norm": 3.1152002811431885, + "learning_rate": 1.8116108944600251e-06, + "loss": 0.3086, + "step": 40071 + }, + { + "epoch": 0.9393507855857757, + "grad_norm": 2.6834311485290527, + "learning_rate": 1.8102153949416056e-06, + "loss": 0.1851, + "step": 40072 + }, + { + "epoch": 0.9393742271605806, + "grad_norm": 1.3719338178634644, + "learning_rate": 1.8088204282037169e-06, + "loss": 0.0666, + "step": 40073 + }, + { + "epoch": 0.9393976687353857, + "grad_norm": 2.201540470123291, + "learning_rate": 1.8074259942539306e-06, + "loss": 0.1027, + "step": 40074 + }, + { + "epoch": 0.9394211103101906, + "grad_norm": 0.8981117606163025, + "learning_rate": 1.8060320930998076e-06, + "loss": 0.082, + "step": 40075 + }, + { + "epoch": 0.9394445518849956, + "grad_norm": 1.7854565382003784, + "learning_rate": 1.8046387247489083e-06, + "loss": 0.103, + "step": 40076 + }, + { + "epoch": 0.9394679934598006, + "grad_norm": 0.190641388297081, + "learning_rate": 1.8032458892088155e-06, + "loss": 0.0145, + "step": 40077 + }, + { + "epoch": 0.9394914350346056, + "grad_norm": 2.589547872543335, + "learning_rate": 1.801853586487068e-06, + "loss": 0.1763, + "step": 40078 + }, + { + "epoch": 0.9395148766094106, + "grad_norm": 3.8875861167907715, + "learning_rate": 1.8004618165912258e-06, + "loss": 0.1347, + "step": 40079 + }, + { + "epoch": 0.9395383181842156, + "grad_norm": 1.9921618700027466, + "learning_rate": 1.7990705795288386e-06, + "loss": 0.2081, + "step": 40080 + }, + { + "epoch": 0.9395617597590206, + "grad_norm": 2.4580843448638916, + "learning_rate": 1.7976798753074563e-06, + "loss": 0.1364, + "step": 40081 + }, + { + "epoch": 0.9395852013338256, + "grad_norm": 1.006004810333252, + "learning_rate": 1.796289703934617e-06, + "loss": 0.1193, + "step": 40082 + }, + { + "epoch": 0.9396086429086306, + "grad_norm": 1.5797284841537476, + "learning_rate": 1.7949000654178705e-06, + "loss": 0.1246, + "step": 40083 + }, + { + "epoch": 0.9396320844834356, + "grad_norm": 3.6759164333343506, + "learning_rate": 1.7935109597647658e-06, + "loss": 0.3484, + "step": 40084 + }, + { + "epoch": 0.9396555260582405, + "grad_norm": 1.766972303390503, + "learning_rate": 1.7921223869828308e-06, + "loss": 0.0909, + "step": 40085 + }, + { + "epoch": 0.9396789676330456, + "grad_norm": 0.8551183342933655, + "learning_rate": 1.7907343470796034e-06, + "loss": 0.0871, + "step": 40086 + }, + { + "epoch": 0.9397024092078505, + "grad_norm": 2.2581355571746826, + "learning_rate": 1.7893468400626e-06, + "loss": 0.1049, + "step": 40087 + }, + { + "epoch": 0.9397258507826556, + "grad_norm": 1.2411507368087769, + "learning_rate": 1.7879598659393704e-06, + "loss": 0.0639, + "step": 40088 + }, + { + "epoch": 0.9397492923574605, + "grad_norm": 1.2242909669876099, + "learning_rate": 1.7865734247174192e-06, + "loss": 0.1091, + "step": 40089 + }, + { + "epoch": 0.9397727339322656, + "grad_norm": 2.2830097675323486, + "learning_rate": 1.7851875164042963e-06, + "loss": 0.1531, + "step": 40090 + }, + { + "epoch": 0.9397961755070706, + "grad_norm": 4.123632907867432, + "learning_rate": 1.7838021410075068e-06, + "loss": 0.2242, + "step": 40091 + }, + { + "epoch": 0.9398196170818756, + "grad_norm": 0.8865787982940674, + "learning_rate": 1.7824172985345777e-06, + "loss": 0.0363, + "step": 40092 + }, + { + "epoch": 0.9398430586566806, + "grad_norm": 1.1643953323364258, + "learning_rate": 1.7810329889930034e-06, + "loss": 0.0555, + "step": 40093 + }, + { + "epoch": 0.9398665002314855, + "grad_norm": 2.4319913387298584, + "learning_rate": 1.7796492123902997e-06, + "loss": 0.2397, + "step": 40094 + }, + { + "epoch": 0.9398899418062906, + "grad_norm": 2.8673815727233887, + "learning_rate": 1.778265968733983e-06, + "loss": 0.2474, + "step": 40095 + }, + { + "epoch": 0.9399133833810955, + "grad_norm": 2.643481731414795, + "learning_rate": 1.7768832580315697e-06, + "loss": 0.2656, + "step": 40096 + }, + { + "epoch": 0.9399368249559006, + "grad_norm": 1.7494945526123047, + "learning_rate": 1.7755010802905426e-06, + "loss": 0.2221, + "step": 40097 + }, + { + "epoch": 0.9399602665307055, + "grad_norm": 2.510383129119873, + "learning_rate": 1.7741194355184066e-06, + "loss": 0.0995, + "step": 40098 + }, + { + "epoch": 0.9399837081055106, + "grad_norm": 1.564706563949585, + "learning_rate": 1.7727383237226669e-06, + "loss": 0.0705, + "step": 40099 + }, + { + "epoch": 0.9400071496803155, + "grad_norm": 3.8064918518066406, + "learning_rate": 1.7713577449107954e-06, + "loss": 0.2578, + "step": 40100 + }, + { + "epoch": 0.9400305912551206, + "grad_norm": 1.4515200853347778, + "learning_rate": 1.7699776990902972e-06, + "loss": 0.0504, + "step": 40101 + }, + { + "epoch": 0.9400540328299255, + "grad_norm": 3.3741838932037354, + "learning_rate": 1.7685981862686775e-06, + "loss": 0.2275, + "step": 40102 + }, + { + "epoch": 0.9400774744047306, + "grad_norm": 4.133012294769287, + "learning_rate": 1.7672192064534076e-06, + "loss": 0.3293, + "step": 40103 + }, + { + "epoch": 0.9401009159795355, + "grad_norm": 1.041383147239685, + "learning_rate": 1.76584075965196e-06, + "loss": 0.7938, + "step": 40104 + }, + { + "epoch": 0.9401243575543405, + "grad_norm": 1.2867932319641113, + "learning_rate": 1.7644628458718282e-06, + "loss": 0.134, + "step": 40105 + }, + { + "epoch": 0.9401477991291455, + "grad_norm": 0.5268332958221436, + "learning_rate": 1.7630854651204843e-06, + "loss": 0.0497, + "step": 40106 + }, + { + "epoch": 0.9401712407039505, + "grad_norm": 1.1394922733306885, + "learning_rate": 1.7617086174053887e-06, + "loss": 0.0906, + "step": 40107 + }, + { + "epoch": 0.9401946822787555, + "grad_norm": 1.009096622467041, + "learning_rate": 1.7603323027340358e-06, + "loss": 0.0625, + "step": 40108 + }, + { + "epoch": 0.9402181238535605, + "grad_norm": 0.7698789238929749, + "learning_rate": 1.7589565211138748e-06, + "loss": 0.0417, + "step": 40109 + }, + { + "epoch": 0.9402415654283655, + "grad_norm": 1.681542158126831, + "learning_rate": 1.7575812725523889e-06, + "loss": 0.0704, + "step": 40110 + }, + { + "epoch": 0.9402650070031705, + "grad_norm": 3.5073790550231934, + "learning_rate": 1.7562065570570163e-06, + "loss": 0.2053, + "step": 40111 + }, + { + "epoch": 0.9402884485779754, + "grad_norm": 0.5921160578727722, + "learning_rate": 1.754832374635229e-06, + "loss": 0.0263, + "step": 40112 + }, + { + "epoch": 0.9403118901527805, + "grad_norm": 2.2868144512176514, + "learning_rate": 1.7534587252944989e-06, + "loss": 1.5676, + "step": 40113 + }, + { + "epoch": 0.9403353317275854, + "grad_norm": 1.113369107246399, + "learning_rate": 1.752085609042242e-06, + "loss": 0.0988, + "step": 40114 + }, + { + "epoch": 0.9403587733023905, + "grad_norm": 0.6137747764587402, + "learning_rate": 1.7507130258859416e-06, + "loss": 0.0446, + "step": 40115 + }, + { + "epoch": 0.9403822148771954, + "grad_norm": 3.3953399658203125, + "learning_rate": 1.749340975833047e-06, + "loss": 0.2581, + "step": 40116 + }, + { + "epoch": 0.9404056564520005, + "grad_norm": 2.3084418773651123, + "learning_rate": 1.7479694588909857e-06, + "loss": 1.6735, + "step": 40117 + }, + { + "epoch": 0.9404290980268054, + "grad_norm": 2.4634695053100586, + "learning_rate": 1.746598475067196e-06, + "loss": 0.2928, + "step": 40118 + }, + { + "epoch": 0.9404525396016105, + "grad_norm": 3.400728464126587, + "learning_rate": 1.7452280243691277e-06, + "loss": 0.2755, + "step": 40119 + }, + { + "epoch": 0.9404759811764154, + "grad_norm": 1.133155345916748, + "learning_rate": 1.7438581068042193e-06, + "loss": 0.1799, + "step": 40120 + }, + { + "epoch": 0.9404994227512204, + "grad_norm": 1.8236092329025269, + "learning_rate": 1.742488722379898e-06, + "loss": 0.1224, + "step": 40121 + }, + { + "epoch": 0.9405228643260254, + "grad_norm": 1.331520438194275, + "learning_rate": 1.7411198711035914e-06, + "loss": 0.1278, + "step": 40122 + }, + { + "epoch": 0.9405463059008304, + "grad_norm": 1.500817894935608, + "learning_rate": 1.7397515529827379e-06, + "loss": 0.0943, + "step": 40123 + }, + { + "epoch": 0.9405697474756354, + "grad_norm": 1.0305976867675781, + "learning_rate": 1.7383837680247538e-06, + "loss": 0.0934, + "step": 40124 + }, + { + "epoch": 0.9405931890504404, + "grad_norm": 3.028759002685547, + "learning_rate": 1.7370165162370556e-06, + "loss": 0.0721, + "step": 40125 + }, + { + "epoch": 0.9406166306252454, + "grad_norm": 1.0368506908416748, + "learning_rate": 1.7356497976270702e-06, + "loss": 0.0393, + "step": 40126 + }, + { + "epoch": 0.9406400722000504, + "grad_norm": 5.020947456359863, + "learning_rate": 1.7342836122022144e-06, + "loss": 0.5498, + "step": 40127 + }, + { + "epoch": 0.9406635137748554, + "grad_norm": 2.176161050796509, + "learning_rate": 1.732917959969893e-06, + "loss": 0.3226, + "step": 40128 + }, + { + "epoch": 0.9406869553496604, + "grad_norm": 0.5114253759384155, + "learning_rate": 1.7315528409375336e-06, + "loss": 0.0715, + "step": 40129 + }, + { + "epoch": 0.9407103969244653, + "grad_norm": 2.055271625518799, + "learning_rate": 1.7301882551125303e-06, + "loss": 0.0782, + "step": 40130 + }, + { + "epoch": 0.9407338384992704, + "grad_norm": 2.4204585552215576, + "learning_rate": 1.7288242025022882e-06, + "loss": 0.1853, + "step": 40131 + }, + { + "epoch": 0.9407572800740753, + "grad_norm": 3.6840453147888184, + "learning_rate": 1.7274606831142014e-06, + "loss": 1.4615, + "step": 40132 + }, + { + "epoch": 0.9407807216488804, + "grad_norm": 1.2354626655578613, + "learning_rate": 1.7260976969556753e-06, + "loss": 0.1502, + "step": 40133 + }, + { + "epoch": 0.9408041632236853, + "grad_norm": 4.183233261108398, + "learning_rate": 1.7247352440341146e-06, + "loss": 0.4005, + "step": 40134 + }, + { + "epoch": 0.9408276047984904, + "grad_norm": 0.816865086555481, + "learning_rate": 1.7233733243569027e-06, + "loss": 0.066, + "step": 40135 + }, + { + "epoch": 0.9408510463732953, + "grad_norm": 1.6441437005996704, + "learning_rate": 1.7220119379314336e-06, + "loss": 0.164, + "step": 40136 + }, + { + "epoch": 0.9408744879481004, + "grad_norm": 2.255007266998291, + "learning_rate": 1.7206510847650792e-06, + "loss": 0.0991, + "step": 40137 + }, + { + "epoch": 0.9408979295229053, + "grad_norm": 0.5354436635971069, + "learning_rate": 1.7192907648652335e-06, + "loss": 0.0269, + "step": 40138 + }, + { + "epoch": 0.9409213710977103, + "grad_norm": 1.4836375713348389, + "learning_rate": 1.7179309782392904e-06, + "loss": 0.0561, + "step": 40139 + }, + { + "epoch": 0.9409448126725153, + "grad_norm": 0.6372954845428467, + "learning_rate": 1.7165717248946111e-06, + "loss": 0.0419, + "step": 40140 + }, + { + "epoch": 0.9409682542473203, + "grad_norm": 2.3515894412994385, + "learning_rate": 1.7152130048385783e-06, + "loss": 0.1198, + "step": 40141 + }, + { + "epoch": 0.9409916958221254, + "grad_norm": 1.7267415523529053, + "learning_rate": 1.7138548180785752e-06, + "loss": 0.1039, + "step": 40142 + }, + { + "epoch": 0.9410151373969303, + "grad_norm": 3.894143581390381, + "learning_rate": 1.7124971646219513e-06, + "loss": 0.2678, + "step": 40143 + }, + { + "epoch": 0.9410385789717354, + "grad_norm": 1.8274260759353638, + "learning_rate": 1.7111400444760783e-06, + "loss": 1.2141, + "step": 40144 + }, + { + "epoch": 0.9410620205465403, + "grad_norm": 4.586537837982178, + "learning_rate": 1.7097834576483175e-06, + "loss": 0.3572, + "step": 40145 + }, + { + "epoch": 0.9410854621213454, + "grad_norm": 1.9688451290130615, + "learning_rate": 1.7084274041460402e-06, + "loss": 0.1849, + "step": 40146 + }, + { + "epoch": 0.9411089036961503, + "grad_norm": 0.2942453920841217, + "learning_rate": 1.7070718839766075e-06, + "loss": 0.0178, + "step": 40147 + }, + { + "epoch": 0.9411323452709554, + "grad_norm": 2.0022974014282227, + "learning_rate": 1.7057168971473691e-06, + "loss": 0.1014, + "step": 40148 + }, + { + "epoch": 0.9411557868457603, + "grad_norm": 1.5015904903411865, + "learning_rate": 1.7043624436656525e-06, + "loss": 0.1054, + "step": 40149 + }, + { + "epoch": 0.9411792284205653, + "grad_norm": 1.769392728805542, + "learning_rate": 1.7030085235388516e-06, + "loss": 0.1167, + "step": 40150 + }, + { + "epoch": 0.9412026699953703, + "grad_norm": 0.9036487340927124, + "learning_rate": 1.7016551367742828e-06, + "loss": 0.0679, + "step": 40151 + }, + { + "epoch": 0.9412261115701753, + "grad_norm": 1.9869060516357422, + "learning_rate": 1.7003022833792848e-06, + "loss": 0.2171, + "step": 40152 + }, + { + "epoch": 0.9412495531449803, + "grad_norm": 2.0827431678771973, + "learning_rate": 1.6989499633612293e-06, + "loss": 1.4569, + "step": 40153 + }, + { + "epoch": 0.9412729947197853, + "grad_norm": 3.226318597793579, + "learning_rate": 1.6975981767274217e-06, + "loss": 1.4151, + "step": 40154 + }, + { + "epoch": 0.9412964362945903, + "grad_norm": 5.776237487792969, + "learning_rate": 1.696246923485223e-06, + "loss": 0.6482, + "step": 40155 + }, + { + "epoch": 0.9413198778693953, + "grad_norm": 1.7220863103866577, + "learning_rate": 1.694896203641949e-06, + "loss": 1.1673, + "step": 40156 + }, + { + "epoch": 0.9413433194442002, + "grad_norm": 2.4938933849334717, + "learning_rate": 1.6935460172049277e-06, + "loss": 0.1873, + "step": 40157 + }, + { + "epoch": 0.9413667610190053, + "grad_norm": 1.8809562921524048, + "learning_rate": 1.6921963641814864e-06, + "loss": 1.3025, + "step": 40158 + }, + { + "epoch": 0.9413902025938102, + "grad_norm": 1.959665298461914, + "learning_rate": 1.6908472445789526e-06, + "loss": 0.1706, + "step": 40159 + }, + { + "epoch": 0.9414136441686153, + "grad_norm": 2.6315290927886963, + "learning_rate": 1.6894986584046536e-06, + "loss": 0.2655, + "step": 40160 + }, + { + "epoch": 0.9414370857434202, + "grad_norm": 0.36963939666748047, + "learning_rate": 1.6881506056658946e-06, + "loss": 0.0263, + "step": 40161 + }, + { + "epoch": 0.9414605273182253, + "grad_norm": 0.7851833701133728, + "learning_rate": 1.6868030863699923e-06, + "loss": 0.0451, + "step": 40162 + }, + { + "epoch": 0.9414839688930302, + "grad_norm": 2.452584981918335, + "learning_rate": 1.685456100524263e-06, + "loss": 0.1808, + "step": 40163 + }, + { + "epoch": 0.9415074104678353, + "grad_norm": 2.118612051010132, + "learning_rate": 1.684109648136023e-06, + "loss": 0.1718, + "step": 40164 + }, + { + "epoch": 0.9415308520426402, + "grad_norm": 2.9165823459625244, + "learning_rate": 1.6827637292125553e-06, + "loss": 0.2094, + "step": 40165 + }, + { + "epoch": 0.9415542936174452, + "grad_norm": 1.7462443113327026, + "learning_rate": 1.6814183437611763e-06, + "loss": 0.0593, + "step": 40166 + }, + { + "epoch": 0.9415777351922502, + "grad_norm": 3.2727129459381104, + "learning_rate": 1.6800734917891913e-06, + "loss": 0.3625, + "step": 40167 + }, + { + "epoch": 0.9416011767670552, + "grad_norm": 0.4135969281196594, + "learning_rate": 1.6787291733038946e-06, + "loss": 0.0487, + "step": 40168 + }, + { + "epoch": 0.9416246183418602, + "grad_norm": 2.440784215927124, + "learning_rate": 1.67738538831258e-06, + "loss": 1.3501, + "step": 40169 + }, + { + "epoch": 0.9416480599166652, + "grad_norm": 3.03234601020813, + "learning_rate": 1.67604213682252e-06, + "loss": 0.1502, + "step": 40170 + }, + { + "epoch": 0.9416715014914702, + "grad_norm": 2.2172439098358154, + "learning_rate": 1.6746994188410303e-06, + "loss": 0.1333, + "step": 40171 + }, + { + "epoch": 0.9416949430662752, + "grad_norm": 1.9115290641784668, + "learning_rate": 1.6733572343753944e-06, + "loss": 0.1059, + "step": 40172 + }, + { + "epoch": 0.9417183846410802, + "grad_norm": 3.4754652976989746, + "learning_rate": 1.6720155834328731e-06, + "loss": 0.2139, + "step": 40173 + }, + { + "epoch": 0.9417418262158852, + "grad_norm": 2.165886163711548, + "learning_rate": 1.6706744660207606e-06, + "loss": 0.0948, + "step": 40174 + }, + { + "epoch": 0.9417652677906901, + "grad_norm": 3.610422134399414, + "learning_rate": 1.6693338821463511e-06, + "loss": 0.1613, + "step": 40175 + }, + { + "epoch": 0.9417887093654952, + "grad_norm": 1.9081493616104126, + "learning_rate": 1.6679938318168831e-06, + "loss": 0.0762, + "step": 40176 + }, + { + "epoch": 0.9418121509403001, + "grad_norm": 3.7480123043060303, + "learning_rate": 1.6666543150396507e-06, + "loss": 0.4595, + "step": 40177 + }, + { + "epoch": 0.9418355925151052, + "grad_norm": 2.5477027893066406, + "learning_rate": 1.665315331821915e-06, + "loss": 1.7314, + "step": 40178 + }, + { + "epoch": 0.9418590340899101, + "grad_norm": 1.6212430000305176, + "learning_rate": 1.6639768821709478e-06, + "loss": 0.0897, + "step": 40179 + }, + { + "epoch": 0.9418824756647152, + "grad_norm": 2.3179895877838135, + "learning_rate": 1.6626389660940101e-06, + "loss": 0.1966, + "step": 40180 + }, + { + "epoch": 0.9419059172395201, + "grad_norm": 0.826571524143219, + "learning_rate": 1.6613015835983515e-06, + "loss": 0.0791, + "step": 40181 + }, + { + "epoch": 0.9419293588143252, + "grad_norm": 1.663150429725647, + "learning_rate": 1.6599647346912328e-06, + "loss": 0.1158, + "step": 40182 + }, + { + "epoch": 0.9419528003891301, + "grad_norm": 3.021096706390381, + "learning_rate": 1.6586284193799152e-06, + "loss": 0.1693, + "step": 40183 + }, + { + "epoch": 0.9419762419639351, + "grad_norm": 2.4155218601226807, + "learning_rate": 1.657292637671648e-06, + "loss": 0.2485, + "step": 40184 + }, + { + "epoch": 0.9419996835387401, + "grad_norm": 2.19761323928833, + "learning_rate": 1.655957389573659e-06, + "loss": 0.1859, + "step": 40185 + }, + { + "epoch": 0.9420231251135451, + "grad_norm": 4.381444931030273, + "learning_rate": 1.6546226750932314e-06, + "loss": 0.187, + "step": 40186 + }, + { + "epoch": 0.9420465666883501, + "grad_norm": 3.8578805923461914, + "learning_rate": 1.6532884942375814e-06, + "loss": 0.3787, + "step": 40187 + }, + { + "epoch": 0.9420700082631551, + "grad_norm": 1.1817944049835205, + "learning_rate": 1.651954847013948e-06, + "loss": 0.1028, + "step": 40188 + }, + { + "epoch": 0.9420934498379601, + "grad_norm": 1.8700462579727173, + "learning_rate": 1.6506217334295805e-06, + "loss": 0.0939, + "step": 40189 + }, + { + "epoch": 0.9421168914127651, + "grad_norm": 1.1976598501205444, + "learning_rate": 1.6492891534916954e-06, + "loss": 0.1012, + "step": 40190 + }, + { + "epoch": 0.94214033298757, + "grad_norm": 2.838542938232422, + "learning_rate": 1.6479571072075316e-06, + "loss": 0.2775, + "step": 40191 + }, + { + "epoch": 0.9421637745623751, + "grad_norm": 2.103123426437378, + "learning_rate": 1.646625594584328e-06, + "loss": 0.1804, + "step": 40192 + }, + { + "epoch": 0.9421872161371802, + "grad_norm": 3.053098678588867, + "learning_rate": 1.6452946156293004e-06, + "loss": 0.201, + "step": 40193 + }, + { + "epoch": 0.9422106577119851, + "grad_norm": 2.4410321712493896, + "learning_rate": 1.6439641703496655e-06, + "loss": 0.2275, + "step": 40194 + }, + { + "epoch": 0.9422340992867901, + "grad_norm": 4.200092315673828, + "learning_rate": 1.64263425875264e-06, + "loss": 0.1892, + "step": 40195 + }, + { + "epoch": 0.9422575408615951, + "grad_norm": 0.8388473391532898, + "learning_rate": 1.641304880845451e-06, + "loss": 0.0789, + "step": 40196 + }, + { + "epoch": 0.9422809824364001, + "grad_norm": 2.36301851272583, + "learning_rate": 1.6399760366353044e-06, + "loss": 1.172, + "step": 40197 + }, + { + "epoch": 0.9423044240112051, + "grad_norm": 2.7625980377197266, + "learning_rate": 1.6386477261294165e-06, + "loss": 0.2063, + "step": 40198 + }, + { + "epoch": 0.9423278655860101, + "grad_norm": 3.271453857421875, + "learning_rate": 1.6373199493349922e-06, + "loss": 0.2078, + "step": 40199 + }, + { + "epoch": 0.942351307160815, + "grad_norm": 2.353315830230713, + "learning_rate": 1.6359927062592373e-06, + "loss": 0.2227, + "step": 40200 + }, + { + "epoch": 0.9423747487356201, + "grad_norm": 1.861570954322815, + "learning_rate": 1.6346659969093569e-06, + "loss": 1.4638, + "step": 40201 + }, + { + "epoch": 0.942398190310425, + "grad_norm": 0.6821273565292358, + "learning_rate": 1.6333398212925343e-06, + "loss": 0.0554, + "step": 40202 + }, + { + "epoch": 0.9424216318852301, + "grad_norm": 0.6774165630340576, + "learning_rate": 1.632014179415975e-06, + "loss": 0.0212, + "step": 40203 + }, + { + "epoch": 0.942445073460035, + "grad_norm": 3.128791332244873, + "learning_rate": 1.6306890712868727e-06, + "loss": 0.2643, + "step": 40204 + }, + { + "epoch": 0.9424685150348401, + "grad_norm": 0.7262317538261414, + "learning_rate": 1.6293644969124334e-06, + "loss": 0.056, + "step": 40205 + }, + { + "epoch": 0.942491956609645, + "grad_norm": 2.7666478157043457, + "learning_rate": 1.6280404562998174e-06, + "loss": 1.6597, + "step": 40206 + }, + { + "epoch": 0.9425153981844501, + "grad_norm": 3.674539089202881, + "learning_rate": 1.6267169494562196e-06, + "loss": 0.2549, + "step": 40207 + }, + { + "epoch": 0.942538839759255, + "grad_norm": 1.8206528425216675, + "learning_rate": 1.6253939763888115e-06, + "loss": 1.8464, + "step": 40208 + }, + { + "epoch": 0.9425622813340601, + "grad_norm": 1.3660266399383545, + "learning_rate": 1.6240715371047988e-06, + "loss": 0.1054, + "step": 40209 + }, + { + "epoch": 0.942585722908865, + "grad_norm": 0.6980900764465332, + "learning_rate": 1.62274963161132e-06, + "loss": 0.0433, + "step": 40210 + }, + { + "epoch": 0.94260916448367, + "grad_norm": 2.2951691150665283, + "learning_rate": 1.6214282599155805e-06, + "loss": 0.0748, + "step": 40211 + }, + { + "epoch": 0.942632606058475, + "grad_norm": 7.56807279586792, + "learning_rate": 1.6201074220247414e-06, + "loss": 0.7288, + "step": 40212 + }, + { + "epoch": 0.94265604763328, + "grad_norm": 0.5994062423706055, + "learning_rate": 1.6187871179459635e-06, + "loss": 0.041, + "step": 40213 + }, + { + "epoch": 0.942679489208085, + "grad_norm": 0.4347686767578125, + "learning_rate": 1.6174673476864078e-06, + "loss": 0.0318, + "step": 40214 + }, + { + "epoch": 0.94270293078289, + "grad_norm": 2.6121978759765625, + "learning_rate": 1.6161481112532461e-06, + "loss": 0.2499, + "step": 40215 + }, + { + "epoch": 0.942726372357695, + "grad_norm": 3.2546300888061523, + "learning_rate": 1.6148294086536286e-06, + "loss": 0.2965, + "step": 40216 + }, + { + "epoch": 0.9427498139325, + "grad_norm": 2.9464662075042725, + "learning_rate": 1.6135112398947161e-06, + "loss": 0.1799, + "step": 40217 + }, + { + "epoch": 0.942773255507305, + "grad_norm": 5.076132774353027, + "learning_rate": 1.6121936049836584e-06, + "loss": 0.2983, + "step": 40218 + }, + { + "epoch": 0.94279669708211, + "grad_norm": 0.700498640537262, + "learning_rate": 1.6108765039275941e-06, + "loss": 0.0612, + "step": 40219 + }, + { + "epoch": 0.9428201386569149, + "grad_norm": 2.958621025085449, + "learning_rate": 1.6095599367336954e-06, + "loss": 0.2503, + "step": 40220 + }, + { + "epoch": 0.94284358023172, + "grad_norm": 0.9103572964668274, + "learning_rate": 1.608243903409079e-06, + "loss": 0.6377, + "step": 40221 + }, + { + "epoch": 0.9428670218065249, + "grad_norm": 3.7567741870880127, + "learning_rate": 1.6069284039608945e-06, + "loss": 0.1922, + "step": 40222 + }, + { + "epoch": 0.94289046338133, + "grad_norm": 3.1868860721588135, + "learning_rate": 1.6056134383962917e-06, + "loss": 0.2191, + "step": 40223 + }, + { + "epoch": 0.9429139049561349, + "grad_norm": 1.2018629312515259, + "learning_rate": 1.6042990067223984e-06, + "loss": 0.6227, + "step": 40224 + }, + { + "epoch": 0.94293734653094, + "grad_norm": 3.09770131111145, + "learning_rate": 1.6029851089463422e-06, + "loss": 0.1722, + "step": 40225 + }, + { + "epoch": 0.9429607881057449, + "grad_norm": 0.30849677324295044, + "learning_rate": 1.6016717450752506e-06, + "loss": 0.0233, + "step": 40226 + }, + { + "epoch": 0.94298422968055, + "grad_norm": 4.344459533691406, + "learning_rate": 1.6003589151162624e-06, + "loss": 0.4398, + "step": 40227 + }, + { + "epoch": 0.9430076712553549, + "grad_norm": 0.7811950445175171, + "learning_rate": 1.599046619076483e-06, + "loss": 0.0871, + "step": 40228 + }, + { + "epoch": 0.94303111283016, + "grad_norm": 0.9128870368003845, + "learning_rate": 1.5977348569630623e-06, + "loss": 0.0713, + "step": 40229 + }, + { + "epoch": 0.9430545544049649, + "grad_norm": 2.8351354598999023, + "learning_rate": 1.5964236287830837e-06, + "loss": 0.1167, + "step": 40230 + }, + { + "epoch": 0.9430779959797699, + "grad_norm": 1.4459398984909058, + "learning_rate": 1.5951129345436856e-06, + "loss": 0.0902, + "step": 40231 + }, + { + "epoch": 0.9431014375545749, + "grad_norm": 2.8457798957824707, + "learning_rate": 1.5938027742519735e-06, + "loss": 0.1832, + "step": 40232 + }, + { + "epoch": 0.9431248791293799, + "grad_norm": 2.7925565242767334, + "learning_rate": 1.5924931479150529e-06, + "loss": 0.201, + "step": 40233 + }, + { + "epoch": 0.9431483207041849, + "grad_norm": 3.6326208114624023, + "learning_rate": 1.5911840555400182e-06, + "loss": 0.3296, + "step": 40234 + }, + { + "epoch": 0.9431717622789899, + "grad_norm": 2.092583656311035, + "learning_rate": 1.5898754971339858e-06, + "loss": 0.8393, + "step": 40235 + }, + { + "epoch": 0.9431952038537949, + "grad_norm": 2.8930253982543945, + "learning_rate": 1.5885674727040723e-06, + "loss": 0.1123, + "step": 40236 + }, + { + "epoch": 0.9432186454285999, + "grad_norm": 2.0117692947387695, + "learning_rate": 1.5872599822573497e-06, + "loss": 1.2155, + "step": 40237 + }, + { + "epoch": 0.9432420870034048, + "grad_norm": 3.244168996810913, + "learning_rate": 1.5859530258009235e-06, + "loss": 0.2409, + "step": 40238 + }, + { + "epoch": 0.9432655285782099, + "grad_norm": 2.6753804683685303, + "learning_rate": 1.5846466033418883e-06, + "loss": 0.1972, + "step": 40239 + }, + { + "epoch": 0.9432889701530148, + "grad_norm": 3.6571810245513916, + "learning_rate": 1.583340714887327e-06, + "loss": 0.2525, + "step": 40240 + }, + { + "epoch": 0.9433124117278199, + "grad_norm": 2.9193615913391113, + "learning_rate": 1.5820353604443118e-06, + "loss": 0.2891, + "step": 40241 + }, + { + "epoch": 0.9433358533026248, + "grad_norm": 2.3243393898010254, + "learning_rate": 1.5807305400199478e-06, + "loss": 0.1935, + "step": 40242 + }, + { + "epoch": 0.9433592948774299, + "grad_norm": 1.449568748474121, + "learning_rate": 1.579426253621319e-06, + "loss": 0.0486, + "step": 40243 + }, + { + "epoch": 0.9433827364522349, + "grad_norm": 2.4009294509887695, + "learning_rate": 1.5781225012554856e-06, + "loss": 0.1652, + "step": 40244 + }, + { + "epoch": 0.9434061780270399, + "grad_norm": 2.8851985931396484, + "learning_rate": 1.5768192829295203e-06, + "loss": 0.3156, + "step": 40245 + }, + { + "epoch": 0.9434296196018449, + "grad_norm": 3.753944158554077, + "learning_rate": 1.575516598650495e-06, + "loss": 0.2667, + "step": 40246 + }, + { + "epoch": 0.9434530611766498, + "grad_norm": 2.6743106842041016, + "learning_rate": 1.5742144484254928e-06, + "loss": 0.2174, + "step": 40247 + }, + { + "epoch": 0.9434765027514549, + "grad_norm": 3.818262815475464, + "learning_rate": 1.5729128322615637e-06, + "loss": 0.4172, + "step": 40248 + }, + { + "epoch": 0.9434999443262598, + "grad_norm": 3.264728307723999, + "learning_rate": 1.5716117501657912e-06, + "loss": 0.3242, + "step": 40249 + }, + { + "epoch": 0.9435233859010649, + "grad_norm": 1.0670161247253418, + "learning_rate": 1.5703112021452138e-06, + "loss": 0.0945, + "step": 40250 + }, + { + "epoch": 0.9435468274758698, + "grad_norm": 5.786457061767578, + "learning_rate": 1.5690111882068925e-06, + "loss": 0.4612, + "step": 40251 + }, + { + "epoch": 0.9435702690506749, + "grad_norm": 3.501297950744629, + "learning_rate": 1.5677117083578885e-06, + "loss": 0.3219, + "step": 40252 + }, + { + "epoch": 0.9435937106254798, + "grad_norm": 2.561204195022583, + "learning_rate": 1.5664127626052405e-06, + "loss": 0.1885, + "step": 40253 + }, + { + "epoch": 0.9436171522002849, + "grad_norm": 1.132492184638977, + "learning_rate": 1.5651143509560207e-06, + "loss": 0.0799, + "step": 40254 + }, + { + "epoch": 0.9436405937750898, + "grad_norm": 3.372887372970581, + "learning_rate": 1.5638164734172456e-06, + "loss": 0.3174, + "step": 40255 + }, + { + "epoch": 0.9436640353498948, + "grad_norm": 2.815850019454956, + "learning_rate": 1.5625191299959762e-06, + "loss": 0.4015, + "step": 40256 + }, + { + "epoch": 0.9436874769246998, + "grad_norm": 3.222991466522217, + "learning_rate": 1.5612223206992404e-06, + "loss": 0.2819, + "step": 40257 + }, + { + "epoch": 0.9437109184995048, + "grad_norm": 2.340285539627075, + "learning_rate": 1.559926045534077e-06, + "loss": 0.2034, + "step": 40258 + }, + { + "epoch": 0.9437343600743098, + "grad_norm": 1.671504259109497, + "learning_rate": 1.5586303045075246e-06, + "loss": 0.1812, + "step": 40259 + }, + { + "epoch": 0.9437578016491148, + "grad_norm": 2.6681766510009766, + "learning_rate": 1.557335097626611e-06, + "loss": 0.3701, + "step": 40260 + }, + { + "epoch": 0.9437812432239198, + "grad_norm": 1.6633249521255493, + "learning_rate": 1.5560404248983751e-06, + "loss": 0.0838, + "step": 40261 + }, + { + "epoch": 0.9438046847987248, + "grad_norm": 2.9469568729400635, + "learning_rate": 1.5547462863298222e-06, + "loss": 0.26, + "step": 40262 + }, + { + "epoch": 0.9438281263735298, + "grad_norm": 0.8709321022033691, + "learning_rate": 1.5534526819279915e-06, + "loss": 0.059, + "step": 40263 + }, + { + "epoch": 0.9438515679483348, + "grad_norm": 2.2790274620056152, + "learning_rate": 1.5521596116998994e-06, + "loss": 0.0823, + "step": 40264 + }, + { + "epoch": 0.9438750095231397, + "grad_norm": 3.0161728858947754, + "learning_rate": 1.5508670756525512e-06, + "loss": 1.4819, + "step": 40265 + }, + { + "epoch": 0.9438984510979448, + "grad_norm": 1.0000449419021606, + "learning_rate": 1.5495750737929527e-06, + "loss": 0.0802, + "step": 40266 + }, + { + "epoch": 0.9439218926727497, + "grad_norm": 2.2197425365448, + "learning_rate": 1.5482836061281424e-06, + "loss": 0.1218, + "step": 40267 + }, + { + "epoch": 0.9439453342475548, + "grad_norm": 1.7133698463439941, + "learning_rate": 1.5469926726651151e-06, + "loss": 0.0804, + "step": 40268 + }, + { + "epoch": 0.9439687758223597, + "grad_norm": 0.39967435598373413, + "learning_rate": 1.545702273410865e-06, + "loss": 0.0304, + "step": 40269 + }, + { + "epoch": 0.9439922173971648, + "grad_norm": 2.83943772315979, + "learning_rate": 1.5444124083724087e-06, + "loss": 0.3676, + "step": 40270 + }, + { + "epoch": 0.9440156589719697, + "grad_norm": 0.9687100648880005, + "learning_rate": 1.5431230775567296e-06, + "loss": 0.0885, + "step": 40271 + }, + { + "epoch": 0.9440391005467748, + "grad_norm": 0.782620370388031, + "learning_rate": 1.5418342809708442e-06, + "loss": 0.0712, + "step": 40272 + }, + { + "epoch": 0.9440625421215797, + "grad_norm": 2.834962844848633, + "learning_rate": 1.5405460186217247e-06, + "loss": 0.155, + "step": 40273 + }, + { + "epoch": 0.9440859836963847, + "grad_norm": 1.7387081384658813, + "learning_rate": 1.5392582905163766e-06, + "loss": 0.0539, + "step": 40274 + }, + { + "epoch": 0.9441094252711897, + "grad_norm": 2.8267977237701416, + "learning_rate": 1.5379710966617834e-06, + "loss": 0.2438, + "step": 40275 + }, + { + "epoch": 0.9441328668459947, + "grad_norm": 1.9750282764434814, + "learning_rate": 1.536684437064928e-06, + "loss": 0.2738, + "step": 40276 + }, + { + "epoch": 0.9441563084207997, + "grad_norm": 2.9813363552093506, + "learning_rate": 1.535398311732794e-06, + "loss": 0.379, + "step": 40277 + }, + { + "epoch": 0.9441797499956047, + "grad_norm": 2.27950382232666, + "learning_rate": 1.5341127206723538e-06, + "loss": 0.1307, + "step": 40278 + }, + { + "epoch": 0.9442031915704097, + "grad_norm": 2.325188636779785, + "learning_rate": 1.5328276638905792e-06, + "loss": 0.1566, + "step": 40279 + }, + { + "epoch": 0.9442266331452147, + "grad_norm": 1.5543655157089233, + "learning_rate": 1.5315431413944647e-06, + "loss": 0.1069, + "step": 40280 + }, + { + "epoch": 0.9442500747200196, + "grad_norm": 3.7271058559417725, + "learning_rate": 1.5302591531909606e-06, + "loss": 0.2508, + "step": 40281 + }, + { + "epoch": 0.9442735162948247, + "grad_norm": 2.3116612434387207, + "learning_rate": 1.5289756992870386e-06, + "loss": 0.1072, + "step": 40282 + }, + { + "epoch": 0.9442969578696296, + "grad_norm": 2.3930225372314453, + "learning_rate": 1.5276927796896489e-06, + "loss": 0.1918, + "step": 40283 + }, + { + "epoch": 0.9443203994444347, + "grad_norm": 0.908199667930603, + "learning_rate": 1.526410394405786e-06, + "loss": 0.0285, + "step": 40284 + }, + { + "epoch": 0.9443438410192396, + "grad_norm": 1.5007281303405762, + "learning_rate": 1.5251285434423779e-06, + "loss": 0.0794, + "step": 40285 + }, + { + "epoch": 0.9443672825940447, + "grad_norm": 1.0344736576080322, + "learning_rate": 1.5238472268063962e-06, + "loss": 0.1065, + "step": 40286 + }, + { + "epoch": 0.9443907241688496, + "grad_norm": 3.973299026489258, + "learning_rate": 1.5225664445047805e-06, + "loss": 0.5462, + "step": 40287 + }, + { + "epoch": 0.9444141657436547, + "grad_norm": 2.1770904064178467, + "learning_rate": 1.5212861965445025e-06, + "loss": 0.9394, + "step": 40288 + }, + { + "epoch": 0.9444376073184596, + "grad_norm": 1.567139983177185, + "learning_rate": 1.52000648293249e-06, + "loss": 0.1153, + "step": 40289 + }, + { + "epoch": 0.9444610488932647, + "grad_norm": 3.056450843811035, + "learning_rate": 1.5187273036756932e-06, + "loss": 1.5956, + "step": 40290 + }, + { + "epoch": 0.9444844904680696, + "grad_norm": 3.2550811767578125, + "learning_rate": 1.51744865878104e-06, + "loss": 0.3648, + "step": 40291 + }, + { + "epoch": 0.9445079320428746, + "grad_norm": 1.5296052694320679, + "learning_rate": 1.51617054825548e-06, + "loss": 0.0643, + "step": 40292 + }, + { + "epoch": 0.9445313736176796, + "grad_norm": 3.8808631896972656, + "learning_rate": 1.5148929721059634e-06, + "loss": 0.3313, + "step": 40293 + }, + { + "epoch": 0.9445548151924846, + "grad_norm": 1.676990270614624, + "learning_rate": 1.513615930339396e-06, + "loss": 0.1167, + "step": 40294 + }, + { + "epoch": 0.9445782567672897, + "grad_norm": 2.9122374057769775, + "learning_rate": 1.5123394229627164e-06, + "loss": 0.2066, + "step": 40295 + }, + { + "epoch": 0.9446016983420946, + "grad_norm": 1.658232569694519, + "learning_rate": 1.5110634499828525e-06, + "loss": 0.1203, + "step": 40296 + }, + { + "epoch": 0.9446251399168997, + "grad_norm": 4.377054214477539, + "learning_rate": 1.509788011406732e-06, + "loss": 0.5564, + "step": 40297 + }, + { + "epoch": 0.9446485814917046, + "grad_norm": 1.0083914995193481, + "learning_rate": 1.5085131072412607e-06, + "loss": 0.1337, + "step": 40298 + }, + { + "epoch": 0.9446720230665097, + "grad_norm": 2.6323623657226562, + "learning_rate": 1.5072387374933772e-06, + "loss": 0.2071, + "step": 40299 + }, + { + "epoch": 0.9446954646413146, + "grad_norm": 2.19212007522583, + "learning_rate": 1.5059649021699872e-06, + "loss": 1.7904, + "step": 40300 + }, + { + "epoch": 0.9447189062161196, + "grad_norm": 2.6648457050323486, + "learning_rate": 1.5046916012779965e-06, + "loss": 0.2712, + "step": 40301 + }, + { + "epoch": 0.9447423477909246, + "grad_norm": 3.534479856491089, + "learning_rate": 1.5034188348243217e-06, + "loss": 0.4292, + "step": 40302 + }, + { + "epoch": 0.9447657893657296, + "grad_norm": 1.9797269105911255, + "learning_rate": 1.5021466028158681e-06, + "loss": 0.0618, + "step": 40303 + }, + { + "epoch": 0.9447892309405346, + "grad_norm": 2.067164659500122, + "learning_rate": 1.5008749052595195e-06, + "loss": 1.4883, + "step": 40304 + }, + { + "epoch": 0.9448126725153396, + "grad_norm": 4.198977470397949, + "learning_rate": 1.4996037421622145e-06, + "loss": 0.2561, + "step": 40305 + }, + { + "epoch": 0.9448361140901446, + "grad_norm": 3.1105568408966064, + "learning_rate": 1.4983331135308142e-06, + "loss": 0.1829, + "step": 40306 + }, + { + "epoch": 0.9448595556649496, + "grad_norm": 2.8788833618164062, + "learning_rate": 1.4970630193722356e-06, + "loss": 0.2431, + "step": 40307 + }, + { + "epoch": 0.9448829972397546, + "grad_norm": 2.7719171047210693, + "learning_rate": 1.4957934596933621e-06, + "loss": 0.1205, + "step": 40308 + }, + { + "epoch": 0.9449064388145596, + "grad_norm": 2.5261240005493164, + "learning_rate": 1.4945244345010767e-06, + "loss": 0.1831, + "step": 40309 + }, + { + "epoch": 0.9449298803893645, + "grad_norm": 3.0585272312164307, + "learning_rate": 1.4932559438022854e-06, + "loss": 0.2351, + "step": 40310 + }, + { + "epoch": 0.9449533219641696, + "grad_norm": 1.207632064819336, + "learning_rate": 1.491987987603838e-06, + "loss": 0.1121, + "step": 40311 + }, + { + "epoch": 0.9449767635389745, + "grad_norm": 1.0816452503204346, + "learning_rate": 1.490720565912651e-06, + "loss": 0.0782, + "step": 40312 + }, + { + "epoch": 0.9450002051137796, + "grad_norm": 0.8337037563323975, + "learning_rate": 1.489453678735575e-06, + "loss": 0.067, + "step": 40313 + }, + { + "epoch": 0.9450236466885845, + "grad_norm": 0.9600521326065063, + "learning_rate": 1.4881873260794932e-06, + "loss": 0.0471, + "step": 40314 + }, + { + "epoch": 0.9450470882633896, + "grad_norm": 2.7774789333343506, + "learning_rate": 1.4869215079512888e-06, + "loss": 0.1931, + "step": 40315 + }, + { + "epoch": 0.9450705298381945, + "grad_norm": 1.444143295288086, + "learning_rate": 1.485656224357801e-06, + "loss": 0.0949, + "step": 40316 + }, + { + "epoch": 0.9450939714129996, + "grad_norm": 4.525035381317139, + "learning_rate": 1.484391475305924e-06, + "loss": 0.303, + "step": 40317 + }, + { + "epoch": 0.9451174129878045, + "grad_norm": 4.406754970550537, + "learning_rate": 1.483127260802497e-06, + "loss": 0.5878, + "step": 40318 + }, + { + "epoch": 0.9451408545626095, + "grad_norm": 1.949874758720398, + "learning_rate": 1.4818635808544034e-06, + "loss": 0.1753, + "step": 40319 + }, + { + "epoch": 0.9451642961374145, + "grad_norm": 2.4429147243499756, + "learning_rate": 1.4806004354684821e-06, + "loss": 1.5538, + "step": 40320 + }, + { + "epoch": 0.9451877377122195, + "grad_norm": 0.9159852862358093, + "learning_rate": 1.4793378246515942e-06, + "loss": 0.1033, + "step": 40321 + }, + { + "epoch": 0.9452111792870245, + "grad_norm": 2.7479186058044434, + "learning_rate": 1.4780757484105901e-06, + "loss": 0.134, + "step": 40322 + }, + { + "epoch": 0.9452346208618295, + "grad_norm": 0.3299195468425751, + "learning_rate": 1.4768142067523195e-06, + "loss": 0.0221, + "step": 40323 + }, + { + "epoch": 0.9452580624366345, + "grad_norm": 0.4691465497016907, + "learning_rate": 1.4755531996836213e-06, + "loss": 0.0295, + "step": 40324 + }, + { + "epoch": 0.9452815040114395, + "grad_norm": 1.0974842309951782, + "learning_rate": 1.474292727211346e-06, + "loss": 0.1456, + "step": 40325 + }, + { + "epoch": 0.9453049455862444, + "grad_norm": 1.1543267965316772, + "learning_rate": 1.4730327893423213e-06, + "loss": 0.0706, + "step": 40326 + }, + { + "epoch": 0.9453283871610495, + "grad_norm": 3.7435264587402344, + "learning_rate": 1.471773386083397e-06, + "loss": 0.2294, + "step": 40327 + }, + { + "epoch": 0.9453518287358544, + "grad_norm": 0.2303825169801712, + "learning_rate": 1.4705145174414016e-06, + "loss": 0.0146, + "step": 40328 + }, + { + "epoch": 0.9453752703106595, + "grad_norm": 1.7483676671981812, + "learning_rate": 1.4692561834231622e-06, + "loss": 0.0566, + "step": 40329 + }, + { + "epoch": 0.9453987118854644, + "grad_norm": 5.098404884338379, + "learning_rate": 1.4679983840355072e-06, + "loss": 0.3577, + "step": 40330 + }, + { + "epoch": 0.9454221534602695, + "grad_norm": 2.1705503463745117, + "learning_rate": 1.4667411192852642e-06, + "loss": 0.1465, + "step": 40331 + }, + { + "epoch": 0.9454455950350744, + "grad_norm": 2.7565293312072754, + "learning_rate": 1.4654843891792502e-06, + "loss": 0.1241, + "step": 40332 + }, + { + "epoch": 0.9454690366098795, + "grad_norm": 3.154773473739624, + "learning_rate": 1.464228193724304e-06, + "loss": 0.2273, + "step": 40333 + }, + { + "epoch": 0.9454924781846844, + "grad_norm": 1.5300835371017456, + "learning_rate": 1.4629725329272204e-06, + "loss": 0.129, + "step": 40334 + }, + { + "epoch": 0.9455159197594895, + "grad_norm": 0.9762030839920044, + "learning_rate": 1.4617174067948159e-06, + "loss": 0.0483, + "step": 40335 + }, + { + "epoch": 0.9455393613342944, + "grad_norm": 1.534105658531189, + "learning_rate": 1.460462815333896e-06, + "loss": 0.117, + "step": 40336 + }, + { + "epoch": 0.9455628029090994, + "grad_norm": 1.0978989601135254, + "learning_rate": 1.4592087585512782e-06, + "loss": 0.0753, + "step": 40337 + }, + { + "epoch": 0.9455862444839044, + "grad_norm": 0.9884806275367737, + "learning_rate": 1.4579552364537675e-06, + "loss": 0.0826, + "step": 40338 + }, + { + "epoch": 0.9456096860587094, + "grad_norm": 0.6193172931671143, + "learning_rate": 1.4567022490481697e-06, + "loss": 0.0365, + "step": 40339 + }, + { + "epoch": 0.9456331276335144, + "grad_norm": 1.8538938760757446, + "learning_rate": 1.4554497963412684e-06, + "loss": 1.2132, + "step": 40340 + }, + { + "epoch": 0.9456565692083194, + "grad_norm": 0.44996970891952515, + "learning_rate": 1.4541978783398691e-06, + "loss": 0.0401, + "step": 40341 + }, + { + "epoch": 0.9456800107831244, + "grad_norm": 1.685401439666748, + "learning_rate": 1.4529464950507554e-06, + "loss": 1.0046, + "step": 40342 + }, + { + "epoch": 0.9457034523579294, + "grad_norm": 0.8469193577766418, + "learning_rate": 1.451695646480733e-06, + "loss": 0.0333, + "step": 40343 + }, + { + "epoch": 0.9457268939327343, + "grad_norm": 2.4974095821380615, + "learning_rate": 1.450445332636574e-06, + "loss": 0.1661, + "step": 40344 + }, + { + "epoch": 0.9457503355075394, + "grad_norm": 2.7980093955993652, + "learning_rate": 1.4491955535250845e-06, + "loss": 0.2462, + "step": 40345 + }, + { + "epoch": 0.9457737770823444, + "grad_norm": 1.837066650390625, + "learning_rate": 1.4479463091530143e-06, + "loss": 0.0678, + "step": 40346 + }, + { + "epoch": 0.9457972186571494, + "grad_norm": 2.374086856842041, + "learning_rate": 1.4466975995271693e-06, + "loss": 0.1555, + "step": 40347 + }, + { + "epoch": 0.9458206602319544, + "grad_norm": 2.5198702812194824, + "learning_rate": 1.4454494246543104e-06, + "loss": 1.468, + "step": 40348 + }, + { + "epoch": 0.9458441018067594, + "grad_norm": 3.7670319080352783, + "learning_rate": 1.4442017845412104e-06, + "loss": 0.4722, + "step": 40349 + }, + { + "epoch": 0.9458675433815644, + "grad_norm": 3.8708434104919434, + "learning_rate": 1.4429546791946524e-06, + "loss": 0.2677, + "step": 40350 + }, + { + "epoch": 0.9458909849563694, + "grad_norm": 2.0407261848449707, + "learning_rate": 1.4417081086213868e-06, + "loss": 0.1313, + "step": 40351 + }, + { + "epoch": 0.9459144265311744, + "grad_norm": 2.9180848598480225, + "learning_rate": 1.440462072828186e-06, + "loss": 0.1063, + "step": 40352 + }, + { + "epoch": 0.9459378681059794, + "grad_norm": 2.175906181335449, + "learning_rate": 1.4392165718218108e-06, + "loss": 0.1882, + "step": 40353 + }, + { + "epoch": 0.9459613096807844, + "grad_norm": 1.1790359020233154, + "learning_rate": 1.437971605609023e-06, + "loss": 0.0955, + "step": 40354 + }, + { + "epoch": 0.9459847512555893, + "grad_norm": 3.1539008617401123, + "learning_rate": 1.4367271741965504e-06, + "loss": 0.2306, + "step": 40355 + }, + { + "epoch": 0.9460081928303944, + "grad_norm": 2.823375701904297, + "learning_rate": 1.4354832775911763e-06, + "loss": 1.9637, + "step": 40356 + }, + { + "epoch": 0.9460316344051993, + "grad_norm": 2.8581087589263916, + "learning_rate": 1.4342399157996512e-06, + "loss": 0.1957, + "step": 40357 + }, + { + "epoch": 0.9460550759800044, + "grad_norm": 3.939093589782715, + "learning_rate": 1.4329970888287136e-06, + "loss": 0.2126, + "step": 40358 + }, + { + "epoch": 0.9460785175548093, + "grad_norm": 0.9219315052032471, + "learning_rate": 1.431754796685103e-06, + "loss": 0.0608, + "step": 40359 + }, + { + "epoch": 0.9461019591296144, + "grad_norm": 2.9437594413757324, + "learning_rate": 1.4305130393755584e-06, + "loss": 0.1542, + "step": 40360 + }, + { + "epoch": 0.9461254007044193, + "grad_norm": 5.314517498016357, + "learning_rate": 1.4292718169068075e-06, + "loss": 0.4079, + "step": 40361 + }, + { + "epoch": 0.9461488422792244, + "grad_norm": 3.5997204780578613, + "learning_rate": 1.428031129285623e-06, + "loss": 0.3058, + "step": 40362 + }, + { + "epoch": 0.9461722838540293, + "grad_norm": 0.5593435168266296, + "learning_rate": 1.4267909765186994e-06, + "loss": 0.0386, + "step": 40363 + }, + { + "epoch": 0.9461957254288343, + "grad_norm": 2.1879634857177734, + "learning_rate": 1.4255513586127867e-06, + "loss": 0.1927, + "step": 40364 + }, + { + "epoch": 0.9462191670036393, + "grad_norm": 0.9920879006385803, + "learning_rate": 1.424312275574602e-06, + "loss": 0.0857, + "step": 40365 + }, + { + "epoch": 0.9462426085784443, + "grad_norm": 1.2978368997573853, + "learning_rate": 1.4230737274108729e-06, + "loss": 0.595, + "step": 40366 + }, + { + "epoch": 0.9462660501532493, + "grad_norm": 0.9726434350013733, + "learning_rate": 1.4218357141283166e-06, + "loss": 0.0636, + "step": 40367 + }, + { + "epoch": 0.9462894917280543, + "grad_norm": 0.9737340211868286, + "learning_rate": 1.420598235733639e-06, + "loss": 0.0744, + "step": 40368 + }, + { + "epoch": 0.9463129333028593, + "grad_norm": 1.6965739727020264, + "learning_rate": 1.4193612922335785e-06, + "loss": 0.124, + "step": 40369 + }, + { + "epoch": 0.9463363748776643, + "grad_norm": 3.738621711730957, + "learning_rate": 1.4181248836348414e-06, + "loss": 0.4034, + "step": 40370 + }, + { + "epoch": 0.9463598164524692, + "grad_norm": 4.528378963470459, + "learning_rate": 1.4168890099441224e-06, + "loss": 0.2157, + "step": 40371 + }, + { + "epoch": 0.9463832580272743, + "grad_norm": 1.9217547178268433, + "learning_rate": 1.4156536711681378e-06, + "loss": 0.0889, + "step": 40372 + }, + { + "epoch": 0.9464066996020792, + "grad_norm": 1.5518018007278442, + "learning_rate": 1.4144188673135827e-06, + "loss": 0.1113, + "step": 40373 + }, + { + "epoch": 0.9464301411768843, + "grad_norm": 2.002717971801758, + "learning_rate": 1.4131845983871628e-06, + "loss": 0.115, + "step": 40374 + }, + { + "epoch": 0.9464535827516892, + "grad_norm": 1.4861862659454346, + "learning_rate": 1.4119508643955837e-06, + "loss": 0.105, + "step": 40375 + }, + { + "epoch": 0.9464770243264943, + "grad_norm": 1.6161853075027466, + "learning_rate": 1.410717665345529e-06, + "loss": 0.0656, + "step": 40376 + }, + { + "epoch": 0.9465004659012992, + "grad_norm": 3.973717212677002, + "learning_rate": 1.4094850012436934e-06, + "loss": 0.4185, + "step": 40377 + }, + { + "epoch": 0.9465239074761043, + "grad_norm": 4.186716556549072, + "learning_rate": 1.4082528720967602e-06, + "loss": 0.3927, + "step": 40378 + }, + { + "epoch": 0.9465473490509092, + "grad_norm": 2.7385170459747314, + "learning_rate": 1.4070212779114244e-06, + "loss": 0.1782, + "step": 40379 + }, + { + "epoch": 0.9465707906257143, + "grad_norm": 0.23070178925991058, + "learning_rate": 1.4057902186943583e-06, + "loss": 0.0178, + "step": 40380 + }, + { + "epoch": 0.9465942322005192, + "grad_norm": 2.5945420265197754, + "learning_rate": 1.4045596944522343e-06, + "loss": 0.1906, + "step": 40381 + }, + { + "epoch": 0.9466176737753242, + "grad_norm": 2.830664873123169, + "learning_rate": 1.4033297051917581e-06, + "loss": 0.2724, + "step": 40382 + }, + { + "epoch": 0.9466411153501292, + "grad_norm": 1.84756600856781, + "learning_rate": 1.4021002509195912e-06, + "loss": 0.1088, + "step": 40383 + }, + { + "epoch": 0.9466645569249342, + "grad_norm": 2.9814627170562744, + "learning_rate": 1.400871331642395e-06, + "loss": 0.2114, + "step": 40384 + }, + { + "epoch": 0.9466879984997392, + "grad_norm": 2.1362287998199463, + "learning_rate": 1.3996429473668414e-06, + "loss": 0.1429, + "step": 40385 + }, + { + "epoch": 0.9467114400745442, + "grad_norm": 2.857816457748413, + "learning_rate": 1.3984150980996036e-06, + "loss": 0.2817, + "step": 40386 + }, + { + "epoch": 0.9467348816493492, + "grad_norm": 1.6620434522628784, + "learning_rate": 1.3971877838473203e-06, + "loss": 0.9209, + "step": 40387 + }, + { + "epoch": 0.9467583232241542, + "grad_norm": 4.31223201751709, + "learning_rate": 1.395961004616686e-06, + "loss": 0.2982, + "step": 40388 + }, + { + "epoch": 0.9467817647989591, + "grad_norm": 1.0321497917175293, + "learning_rate": 1.3947347604143402e-06, + "loss": 0.1183, + "step": 40389 + }, + { + "epoch": 0.9468052063737642, + "grad_norm": 3.0832293033599854, + "learning_rate": 1.393509051246933e-06, + "loss": 0.2925, + "step": 40390 + }, + { + "epoch": 0.9468286479485691, + "grad_norm": 0.8609453439712524, + "learning_rate": 1.3922838771211143e-06, + "loss": 0.0459, + "step": 40391 + }, + { + "epoch": 0.9468520895233742, + "grad_norm": 2.2184369564056396, + "learning_rate": 1.391059238043535e-06, + "loss": 0.1958, + "step": 40392 + }, + { + "epoch": 0.9468755310981791, + "grad_norm": 2.7330799102783203, + "learning_rate": 1.3898351340208448e-06, + "loss": 0.1564, + "step": 40393 + }, + { + "epoch": 0.9468989726729842, + "grad_norm": 2.0573716163635254, + "learning_rate": 1.388611565059672e-06, + "loss": 0.1551, + "step": 40394 + }, + { + "epoch": 0.9469224142477891, + "grad_norm": 0.3849600553512573, + "learning_rate": 1.387388531166678e-06, + "loss": 0.0304, + "step": 40395 + }, + { + "epoch": 0.9469458558225942, + "grad_norm": 1.571099877357483, + "learning_rate": 1.3861660323484793e-06, + "loss": 1.1577, + "step": 40396 + }, + { + "epoch": 0.9469692973973992, + "grad_norm": 2.4403269290924072, + "learning_rate": 1.3849440686117266e-06, + "loss": 0.2223, + "step": 40397 + }, + { + "epoch": 0.9469927389722042, + "grad_norm": 2.4335782527923584, + "learning_rate": 1.3837226399630365e-06, + "loss": 0.146, + "step": 40398 + }, + { + "epoch": 0.9470161805470092, + "grad_norm": 3.376538038253784, + "learning_rate": 1.3825017464090262e-06, + "loss": 0.2583, + "step": 40399 + }, + { + "epoch": 0.9470396221218141, + "grad_norm": 1.4322283267974854, + "learning_rate": 1.3812813879563457e-06, + "loss": 0.1083, + "step": 40400 + }, + { + "epoch": 0.9470630636966192, + "grad_norm": 1.9545162916183472, + "learning_rate": 1.3800615646116123e-06, + "loss": 0.2482, + "step": 40401 + }, + { + "epoch": 0.9470865052714241, + "grad_norm": 2.430004596710205, + "learning_rate": 1.3788422763814313e-06, + "loss": 1.7804, + "step": 40402 + }, + { + "epoch": 0.9471099468462292, + "grad_norm": 2.4939868450164795, + "learning_rate": 1.37762352327242e-06, + "loss": 0.2367, + "step": 40403 + }, + { + "epoch": 0.9471333884210341, + "grad_norm": 2.0981621742248535, + "learning_rate": 1.3764053052911953e-06, + "loss": 0.0985, + "step": 40404 + }, + { + "epoch": 0.9471568299958392, + "grad_norm": 1.4817793369293213, + "learning_rate": 1.3751876224443738e-06, + "loss": 0.1152, + "step": 40405 + }, + { + "epoch": 0.9471802715706441, + "grad_norm": 2.513881206512451, + "learning_rate": 1.3739704747385506e-06, + "loss": 0.1609, + "step": 40406 + }, + { + "epoch": 0.9472037131454492, + "grad_norm": 2.9476778507232666, + "learning_rate": 1.3727538621803538e-06, + "loss": 0.2458, + "step": 40407 + }, + { + "epoch": 0.9472271547202541, + "grad_norm": 2.547644853591919, + "learning_rate": 1.3715377847763554e-06, + "loss": 0.1869, + "step": 40408 + }, + { + "epoch": 0.9472505962950591, + "grad_norm": 3.807285785675049, + "learning_rate": 1.370322242533173e-06, + "loss": 0.2655, + "step": 40409 + }, + { + "epoch": 0.9472740378698641, + "grad_norm": 1.2070235013961792, + "learning_rate": 1.3691072354573898e-06, + "loss": 0.0715, + "step": 40410 + }, + { + "epoch": 0.9472974794446691, + "grad_norm": 1.490599513053894, + "learning_rate": 1.3678927635556004e-06, + "loss": 0.0475, + "step": 40411 + }, + { + "epoch": 0.9473209210194741, + "grad_norm": 1.574772834777832, + "learning_rate": 1.3666788268344e-06, + "loss": 0.0863, + "step": 40412 + }, + { + "epoch": 0.9473443625942791, + "grad_norm": 1.4921780824661255, + "learning_rate": 1.3654654253003828e-06, + "loss": 0.0656, + "step": 40413 + }, + { + "epoch": 0.9473678041690841, + "grad_norm": 1.2934894561767578, + "learning_rate": 1.3642525589601106e-06, + "loss": 0.1299, + "step": 40414 + }, + { + "epoch": 0.9473912457438891, + "grad_norm": 4.072547435760498, + "learning_rate": 1.3630402278201892e-06, + "loss": 0.3221, + "step": 40415 + }, + { + "epoch": 0.947414687318694, + "grad_norm": 1.344887614250183, + "learning_rate": 1.361828431887169e-06, + "loss": 0.1587, + "step": 40416 + }, + { + "epoch": 0.9474381288934991, + "grad_norm": 2.6073052883148193, + "learning_rate": 1.3606171711676552e-06, + "loss": 0.1944, + "step": 40417 + }, + { + "epoch": 0.947461570468304, + "grad_norm": 2.5080535411834717, + "learning_rate": 1.35940644566821e-06, + "loss": 0.1114, + "step": 40418 + }, + { + "epoch": 0.9474850120431091, + "grad_norm": 0.2029513269662857, + "learning_rate": 1.3581962553953832e-06, + "loss": 0.0028, + "step": 40419 + }, + { + "epoch": 0.947508453617914, + "grad_norm": 0.8447015285491943, + "learning_rate": 1.3569866003557697e-06, + "loss": 0.0337, + "step": 40420 + }, + { + "epoch": 0.9475318951927191, + "grad_norm": 2.5458712577819824, + "learning_rate": 1.3557774805559086e-06, + "loss": 0.2089, + "step": 40421 + }, + { + "epoch": 0.947555336767524, + "grad_norm": 2.9610302448272705, + "learning_rate": 1.354568896002384e-06, + "loss": 0.197, + "step": 40422 + }, + { + "epoch": 0.9475787783423291, + "grad_norm": 2.570859909057617, + "learning_rate": 1.3533608467017455e-06, + "loss": 0.1735, + "step": 40423 + }, + { + "epoch": 0.947602219917134, + "grad_norm": 0.6272452473640442, + "learning_rate": 1.3521533326605329e-06, + "loss": 0.0299, + "step": 40424 + }, + { + "epoch": 0.947625661491939, + "grad_norm": 2.7445969581604004, + "learning_rate": 1.3509463538853074e-06, + "loss": 0.1676, + "step": 40425 + }, + { + "epoch": 0.947649103066744, + "grad_norm": 0.9953302145004272, + "learning_rate": 1.3497399103826303e-06, + "loss": 0.0358, + "step": 40426 + }, + { + "epoch": 0.947672544641549, + "grad_norm": 0.9889453053474426, + "learning_rate": 1.348534002159041e-06, + "loss": 0.0698, + "step": 40427 + }, + { + "epoch": 0.947695986216354, + "grad_norm": 3.147307872772217, + "learning_rate": 1.3473286292210785e-06, + "loss": 0.153, + "step": 40428 + }, + { + "epoch": 0.947719427791159, + "grad_norm": 1.7706022262573242, + "learning_rate": 1.346123791575271e-06, + "loss": 0.1236, + "step": 40429 + }, + { + "epoch": 0.947742869365964, + "grad_norm": 0.4886449873447418, + "learning_rate": 1.3449194892281802e-06, + "loss": 0.0488, + "step": 40430 + }, + { + "epoch": 0.947766310940769, + "grad_norm": 1.3100817203521729, + "learning_rate": 1.343715722186334e-06, + "loss": 0.0458, + "step": 40431 + }, + { + "epoch": 0.947789752515574, + "grad_norm": 2.6694765090942383, + "learning_rate": 1.342512490456249e-06, + "loss": 1.5269, + "step": 40432 + }, + { + "epoch": 0.947813194090379, + "grad_norm": 0.5366988778114319, + "learning_rate": 1.3413097940444764e-06, + "loss": 0.0351, + "step": 40433 + }, + { + "epoch": 0.947836635665184, + "grad_norm": 0.5616278648376465, + "learning_rate": 1.3401076329575325e-06, + "loss": 0.0456, + "step": 40434 + }, + { + "epoch": 0.947860077239989, + "grad_norm": 1.9205814599990845, + "learning_rate": 1.3389060072019344e-06, + "loss": 0.1294, + "step": 40435 + }, + { + "epoch": 0.9478835188147939, + "grad_norm": 1.380842924118042, + "learning_rate": 1.3377049167841994e-06, + "loss": 0.024, + "step": 40436 + }, + { + "epoch": 0.947906960389599, + "grad_norm": 1.4507977962493896, + "learning_rate": 1.3365043617108553e-06, + "loss": 0.0794, + "step": 40437 + }, + { + "epoch": 0.9479304019644039, + "grad_norm": 1.1969338655471802, + "learning_rate": 1.3353043419884192e-06, + "loss": 0.0469, + "step": 40438 + }, + { + "epoch": 0.947953843539209, + "grad_norm": 2.803816556930542, + "learning_rate": 1.3341048576233862e-06, + "loss": 0.2338, + "step": 40439 + }, + { + "epoch": 0.9479772851140139, + "grad_norm": 3.388383388519287, + "learning_rate": 1.332905908622284e-06, + "loss": 0.2696, + "step": 40440 + }, + { + "epoch": 0.948000726688819, + "grad_norm": 2.3063037395477295, + "learning_rate": 1.3317074949915965e-06, + "loss": 0.1915, + "step": 40441 + }, + { + "epoch": 0.9480241682636239, + "grad_norm": 2.5914180278778076, + "learning_rate": 1.3305096167378516e-06, + "loss": 0.1496, + "step": 40442 + }, + { + "epoch": 0.948047609838429, + "grad_norm": 1.6939339637756348, + "learning_rate": 1.3293122738675333e-06, + "loss": 0.1322, + "step": 40443 + }, + { + "epoch": 0.9480710514132339, + "grad_norm": 0.86128169298172, + "learning_rate": 1.3281154663871254e-06, + "loss": 0.0507, + "step": 40444 + }, + { + "epoch": 0.9480944929880389, + "grad_norm": 4.967557430267334, + "learning_rate": 1.3269191943031555e-06, + "loss": 0.2641, + "step": 40445 + }, + { + "epoch": 0.9481179345628439, + "grad_norm": 3.1218957901000977, + "learning_rate": 1.3257234576220967e-06, + "loss": 0.1913, + "step": 40446 + }, + { + "epoch": 0.9481413761376489, + "grad_norm": 0.8760313987731934, + "learning_rate": 1.3245282563504323e-06, + "loss": 0.1016, + "step": 40447 + }, + { + "epoch": 0.948164817712454, + "grad_norm": 1.4842119216918945, + "learning_rate": 1.3233335904946465e-06, + "loss": 0.1236, + "step": 40448 + }, + { + "epoch": 0.9481882592872589, + "grad_norm": 0.4443213939666748, + "learning_rate": 1.3221394600612335e-06, + "loss": 0.0206, + "step": 40449 + }, + { + "epoch": 0.948211700862064, + "grad_norm": 1.8017195463180542, + "learning_rate": 1.3209458650566665e-06, + "loss": 0.1046, + "step": 40450 + }, + { + "epoch": 0.9482351424368689, + "grad_norm": 2.8236873149871826, + "learning_rate": 1.3197528054874175e-06, + "loss": 0.1565, + "step": 40451 + }, + { + "epoch": 0.948258584011674, + "grad_norm": 9.4290132522583, + "learning_rate": 1.3185602813599708e-06, + "loss": 0.1747, + "step": 40452 + }, + { + "epoch": 0.9482820255864789, + "grad_norm": 1.9241511821746826, + "learning_rate": 1.3173682926807763e-06, + "loss": 0.199, + "step": 40453 + }, + { + "epoch": 0.9483054671612839, + "grad_norm": 1.8593913316726685, + "learning_rate": 1.316176839456329e-06, + "loss": 1.4276, + "step": 40454 + }, + { + "epoch": 0.9483289087360889, + "grad_norm": 2.806929349899292, + "learning_rate": 1.3149859216930794e-06, + "loss": 1.1883, + "step": 40455 + }, + { + "epoch": 0.9483523503108939, + "grad_norm": 2.558865547180176, + "learning_rate": 1.3137955393974889e-06, + "loss": 0.2591, + "step": 40456 + }, + { + "epoch": 0.9483757918856989, + "grad_norm": 1.2989240884780884, + "learning_rate": 1.3126056925760189e-06, + "loss": 0.1525, + "step": 40457 + }, + { + "epoch": 0.9483992334605039, + "grad_norm": 3.5115506649017334, + "learning_rate": 1.311416381235131e-06, + "loss": 0.329, + "step": 40458 + }, + { + "epoch": 0.9484226750353089, + "grad_norm": 2.1446139812469482, + "learning_rate": 1.3102276053812646e-06, + "loss": 0.2284, + "step": 40459 + }, + { + "epoch": 0.9484461166101139, + "grad_norm": 3.2263998985290527, + "learning_rate": 1.309039365020881e-06, + "loss": 0.2711, + "step": 40460 + }, + { + "epoch": 0.9484695581849188, + "grad_norm": 2.339205026626587, + "learning_rate": 1.3078516601604308e-06, + "loss": 0.161, + "step": 40461 + }, + { + "epoch": 0.9484929997597239, + "grad_norm": 3.220343828201294, + "learning_rate": 1.306664490806353e-06, + "loss": 0.2491, + "step": 40462 + }, + { + "epoch": 0.9485164413345288, + "grad_norm": 2.4224703311920166, + "learning_rate": 1.305477856965076e-06, + "loss": 1.3556, + "step": 40463 + }, + { + "epoch": 0.9485398829093339, + "grad_norm": 2.6474530696868896, + "learning_rate": 1.3042917586430504e-06, + "loss": 0.1528, + "step": 40464 + }, + { + "epoch": 0.9485633244841388, + "grad_norm": 2.1114726066589355, + "learning_rate": 1.303106195846715e-06, + "loss": 0.2075, + "step": 40465 + }, + { + "epoch": 0.9485867660589439, + "grad_norm": 2.063443660736084, + "learning_rate": 1.3019211685825095e-06, + "loss": 0.1809, + "step": 40466 + }, + { + "epoch": 0.9486102076337488, + "grad_norm": 2.496173858642578, + "learning_rate": 1.3007366768568506e-06, + "loss": 0.1845, + "step": 40467 + }, + { + "epoch": 0.9486336492085539, + "grad_norm": 3.8411176204681396, + "learning_rate": 1.2995527206761782e-06, + "loss": 0.3109, + "step": 40468 + }, + { + "epoch": 0.9486570907833588, + "grad_norm": 2.3291666507720947, + "learning_rate": 1.2983693000468976e-06, + "loss": 0.1344, + "step": 40469 + }, + { + "epoch": 0.9486805323581639, + "grad_norm": 0.6386258006095886, + "learning_rate": 1.2971864149754375e-06, + "loss": 0.0563, + "step": 40470 + }, + { + "epoch": 0.9487039739329688, + "grad_norm": 1.2074558734893799, + "learning_rate": 1.296004065468226e-06, + "loss": 0.1386, + "step": 40471 + }, + { + "epoch": 0.9487274155077738, + "grad_norm": 4.590926170349121, + "learning_rate": 1.2948222515316688e-06, + "loss": 0.3667, + "step": 40472 + }, + { + "epoch": 0.9487508570825788, + "grad_norm": 2.327552556991577, + "learning_rate": 1.2936409731721832e-06, + "loss": 0.1448, + "step": 40473 + }, + { + "epoch": 0.9487742986573838, + "grad_norm": 2.6834793090820312, + "learning_rate": 1.2924602303961864e-06, + "loss": 0.1571, + "step": 40474 + }, + { + "epoch": 0.9487977402321888, + "grad_norm": 3.1776485443115234, + "learning_rate": 1.291280023210062e-06, + "loss": 0.1788, + "step": 40475 + }, + { + "epoch": 0.9488211818069938, + "grad_norm": 0.9504422545433044, + "learning_rate": 1.290100351620227e-06, + "loss": 0.0929, + "step": 40476 + }, + { + "epoch": 0.9488446233817988, + "grad_norm": 3.5456576347351074, + "learning_rate": 1.2889212156330877e-06, + "loss": 0.2402, + "step": 40477 + }, + { + "epoch": 0.9488680649566038, + "grad_norm": 2.9917023181915283, + "learning_rate": 1.2877426152550387e-06, + "loss": 0.2733, + "step": 40478 + }, + { + "epoch": 0.9488915065314087, + "grad_norm": 1.5249097347259521, + "learning_rate": 1.286564550492475e-06, + "loss": 0.064, + "step": 40479 + }, + { + "epoch": 0.9489149481062138, + "grad_norm": 1.1056650876998901, + "learning_rate": 1.2853870213517805e-06, + "loss": 0.0854, + "step": 40480 + }, + { + "epoch": 0.9489383896810187, + "grad_norm": 4.495150089263916, + "learning_rate": 1.284210027839361e-06, + "loss": 0.1074, + "step": 40481 + }, + { + "epoch": 0.9489618312558238, + "grad_norm": 3.1938316822052, + "learning_rate": 1.283033569961578e-06, + "loss": 0.2537, + "step": 40482 + }, + { + "epoch": 0.9489852728306287, + "grad_norm": 2.132412910461426, + "learning_rate": 1.281857647724849e-06, + "loss": 0.1765, + "step": 40483 + }, + { + "epoch": 0.9490087144054338, + "grad_norm": 1.7898730039596558, + "learning_rate": 1.280682261135524e-06, + "loss": 1.5528, + "step": 40484 + }, + { + "epoch": 0.9490321559802387, + "grad_norm": 1.6322627067565918, + "learning_rate": 1.2795074101999982e-06, + "loss": 0.0582, + "step": 40485 + }, + { + "epoch": 0.9490555975550438, + "grad_norm": 2.116863965988159, + "learning_rate": 1.2783330949246441e-06, + "loss": 0.1366, + "step": 40486 + }, + { + "epoch": 0.9490790391298487, + "grad_norm": 2.0031204223632812, + "learning_rate": 1.2771593153158234e-06, + "loss": 0.1213, + "step": 40487 + }, + { + "epoch": 0.9491024807046538, + "grad_norm": 4.796371936798096, + "learning_rate": 1.2759860713799088e-06, + "loss": 0.4254, + "step": 40488 + }, + { + "epoch": 0.9491259222794587, + "grad_norm": 3.6987802982330322, + "learning_rate": 1.2748133631232729e-06, + "loss": 0.2666, + "step": 40489 + }, + { + "epoch": 0.9491493638542637, + "grad_norm": 0.4964282214641571, + "learning_rate": 1.2736411905522882e-06, + "loss": 0.023, + "step": 40490 + }, + { + "epoch": 0.9491728054290687, + "grad_norm": 0.41562366485595703, + "learning_rate": 1.2724695536732833e-06, + "loss": 0.0289, + "step": 40491 + }, + { + "epoch": 0.9491962470038737, + "grad_norm": 3.4723103046417236, + "learning_rate": 1.2712984524926531e-06, + "loss": 0.3302, + "step": 40492 + }, + { + "epoch": 0.9492196885786787, + "grad_norm": 5.3096489906311035, + "learning_rate": 1.2701278870167255e-06, + "loss": 0.3528, + "step": 40493 + }, + { + "epoch": 0.9492431301534837, + "grad_norm": 1.2961243391036987, + "learning_rate": 1.2689578572518622e-06, + "loss": 0.0879, + "step": 40494 + }, + { + "epoch": 0.9492665717282887, + "grad_norm": 0.4512046277523041, + "learning_rate": 1.2677883632043918e-06, + "loss": 0.0283, + "step": 40495 + }, + { + "epoch": 0.9492900133030937, + "grad_norm": 2.68050217628479, + "learning_rate": 1.2666194048806978e-06, + "loss": 0.0736, + "step": 40496 + }, + { + "epoch": 0.9493134548778986, + "grad_norm": 0.9314528703689575, + "learning_rate": 1.2654509822870974e-06, + "loss": 0.079, + "step": 40497 + }, + { + "epoch": 0.9493368964527037, + "grad_norm": 1.2940958738327026, + "learning_rate": 1.264283095429941e-06, + "loss": 0.1181, + "step": 40498 + }, + { + "epoch": 0.9493603380275087, + "grad_norm": 2.9255547523498535, + "learning_rate": 1.2631157443155572e-06, + "loss": 0.2092, + "step": 40499 + }, + { + "epoch": 0.9493837796023137, + "grad_norm": 2.6247410774230957, + "learning_rate": 1.261948928950285e-06, + "loss": 0.1239, + "step": 40500 + }, + { + "epoch": 0.9494072211771187, + "grad_norm": 2.461416244506836, + "learning_rate": 1.2607826493404417e-06, + "loss": 0.3896, + "step": 40501 + }, + { + "epoch": 0.9494306627519237, + "grad_norm": 1.8949557542800903, + "learning_rate": 1.259616905492378e-06, + "loss": 0.2739, + "step": 40502 + }, + { + "epoch": 0.9494541043267287, + "grad_norm": 1.0205689668655396, + "learning_rate": 1.2584516974124105e-06, + "loss": 0.0961, + "step": 40503 + }, + { + "epoch": 0.9494775459015337, + "grad_norm": 2.640306234359741, + "learning_rate": 1.2572870251068569e-06, + "loss": 0.1341, + "step": 40504 + }, + { + "epoch": 0.9495009874763387, + "grad_norm": 0.5064034461975098, + "learning_rate": 1.2561228885820452e-06, + "loss": 0.0212, + "step": 40505 + }, + { + "epoch": 0.9495244290511436, + "grad_norm": 3.692653179168701, + "learning_rate": 1.2549592878442816e-06, + "loss": 0.1983, + "step": 40506 + }, + { + "epoch": 0.9495478706259487, + "grad_norm": 2.226691246032715, + "learning_rate": 1.2537962228998946e-06, + "loss": 0.1474, + "step": 40507 + }, + { + "epoch": 0.9495713122007536, + "grad_norm": 1.7757612466812134, + "learning_rate": 1.2526336937551786e-06, + "loss": 0.0705, + "step": 40508 + }, + { + "epoch": 0.9495947537755587, + "grad_norm": 3.4154610633850098, + "learning_rate": 1.2514717004164512e-06, + "loss": 0.2499, + "step": 40509 + }, + { + "epoch": 0.9496181953503636, + "grad_norm": 2.021235942840576, + "learning_rate": 1.2503102428900072e-06, + "loss": 1.4113, + "step": 40510 + }, + { + "epoch": 0.9496416369251687, + "grad_norm": 1.5028332471847534, + "learning_rate": 1.2491493211821637e-06, + "loss": 0.0965, + "step": 40511 + }, + { + "epoch": 0.9496650784999736, + "grad_norm": 0.5060668587684631, + "learning_rate": 1.2479889352992158e-06, + "loss": 0.0364, + "step": 40512 + }, + { + "epoch": 0.9496885200747787, + "grad_norm": 3.0142674446105957, + "learning_rate": 1.2468290852474474e-06, + "loss": 0.1249, + "step": 40513 + }, + { + "epoch": 0.9497119616495836, + "grad_norm": 2.341447114944458, + "learning_rate": 1.2456697710331644e-06, + "loss": 0.2086, + "step": 40514 + }, + { + "epoch": 0.9497354032243887, + "grad_norm": 2.531913995742798, + "learning_rate": 1.2445109926626508e-06, + "loss": 0.2663, + "step": 40515 + }, + { + "epoch": 0.9497588447991936, + "grad_norm": 2.9905529022216797, + "learning_rate": 1.2433527501422015e-06, + "loss": 0.2323, + "step": 40516 + }, + { + "epoch": 0.9497822863739986, + "grad_norm": 1.1035041809082031, + "learning_rate": 1.2421950434781004e-06, + "loss": 0.1037, + "step": 40517 + }, + { + "epoch": 0.9498057279488036, + "grad_norm": 2.6691102981567383, + "learning_rate": 1.2410378726766313e-06, + "loss": 0.2538, + "step": 40518 + }, + { + "epoch": 0.9498291695236086, + "grad_norm": 1.2060904502868652, + "learning_rate": 1.239881237744056e-06, + "loss": 0.0781, + "step": 40519 + }, + { + "epoch": 0.9498526110984136, + "grad_norm": 1.9755945205688477, + "learning_rate": 1.2387251386866583e-06, + "loss": 0.1906, + "step": 40520 + }, + { + "epoch": 0.9498760526732186, + "grad_norm": 0.6047903895378113, + "learning_rate": 1.237569575510722e-06, + "loss": 0.0302, + "step": 40521 + }, + { + "epoch": 0.9498994942480236, + "grad_norm": 2.99473237991333, + "learning_rate": 1.2364145482225198e-06, + "loss": 0.2627, + "step": 40522 + }, + { + "epoch": 0.9499229358228286, + "grad_norm": 0.45992714166641235, + "learning_rate": 1.2352600568282913e-06, + "loss": 0.0424, + "step": 40523 + }, + { + "epoch": 0.9499463773976335, + "grad_norm": 3.0753073692321777, + "learning_rate": 1.2341061013343313e-06, + "loss": 0.2566, + "step": 40524 + }, + { + "epoch": 0.9499698189724386, + "grad_norm": 3.056889057159424, + "learning_rate": 1.2329526817468906e-06, + "loss": 0.1561, + "step": 40525 + }, + { + "epoch": 0.9499932605472435, + "grad_norm": 6.167847633361816, + "learning_rate": 1.2317997980722084e-06, + "loss": 0.3303, + "step": 40526 + }, + { + "epoch": 0.9500167021220486, + "grad_norm": 1.3342472314834595, + "learning_rate": 1.2306474503165687e-06, + "loss": 0.0551, + "step": 40527 + }, + { + "epoch": 0.9500401436968535, + "grad_norm": 2.597480297088623, + "learning_rate": 1.229495638486211e-06, + "loss": 0.2547, + "step": 40528 + }, + { + "epoch": 0.9500635852716586, + "grad_norm": 2.5029048919677734, + "learning_rate": 1.2283443625873857e-06, + "loss": 0.1439, + "step": 40529 + }, + { + "epoch": 0.9500870268464635, + "grad_norm": 4.533350467681885, + "learning_rate": 1.2271936226263546e-06, + "loss": 0.5397, + "step": 40530 + }, + { + "epoch": 0.9501104684212686, + "grad_norm": 2.881369113922119, + "learning_rate": 1.2260434186093351e-06, + "loss": 0.2799, + "step": 40531 + }, + { + "epoch": 0.9501339099960735, + "grad_norm": 1.77545964717865, + "learning_rate": 1.2248937505425885e-06, + "loss": 0.1647, + "step": 40532 + }, + { + "epoch": 0.9501573515708786, + "grad_norm": 0.2703551948070526, + "learning_rate": 1.223744618432332e-06, + "loss": 0.016, + "step": 40533 + }, + { + "epoch": 0.9501807931456835, + "grad_norm": 1.1565769910812378, + "learning_rate": 1.2225960222848277e-06, + "loss": 0.0874, + "step": 40534 + }, + { + "epoch": 0.9502042347204885, + "grad_norm": 2.5935897827148438, + "learning_rate": 1.2214479621062925e-06, + "loss": 0.2084, + "step": 40535 + }, + { + "epoch": 0.9502276762952935, + "grad_norm": 0.3599233627319336, + "learning_rate": 1.2203004379029547e-06, + "loss": 0.063, + "step": 40536 + }, + { + "epoch": 0.9502511178700985, + "grad_norm": 3.5450119972229004, + "learning_rate": 1.2191534496810542e-06, + "loss": 0.1279, + "step": 40537 + }, + { + "epoch": 0.9502745594449035, + "grad_norm": 3.1418099403381348, + "learning_rate": 1.2180069974467856e-06, + "loss": 0.3934, + "step": 40538 + }, + { + "epoch": 0.9502980010197085, + "grad_norm": 2.975374221801758, + "learning_rate": 1.2168610812063996e-06, + "loss": 0.1897, + "step": 40539 + }, + { + "epoch": 0.9503214425945135, + "grad_norm": 4.398083686828613, + "learning_rate": 1.2157157009661023e-06, + "loss": 0.3341, + "step": 40540 + }, + { + "epoch": 0.9503448841693185, + "grad_norm": 5.680080890655518, + "learning_rate": 1.2145708567321113e-06, + "loss": 0.3831, + "step": 40541 + }, + { + "epoch": 0.9503683257441234, + "grad_norm": 2.8207180500030518, + "learning_rate": 1.2134265485106322e-06, + "loss": 0.1973, + "step": 40542 + }, + { + "epoch": 0.9503917673189285, + "grad_norm": 0.49438104033470154, + "learning_rate": 1.2122827763078825e-06, + "loss": 0.0332, + "step": 40543 + }, + { + "epoch": 0.9504152088937334, + "grad_norm": 2.4559526443481445, + "learning_rate": 1.2111395401300685e-06, + "loss": 0.1099, + "step": 40544 + }, + { + "epoch": 0.9504386504685385, + "grad_norm": 2.122392177581787, + "learning_rate": 1.209996839983385e-06, + "loss": 1.8092, + "step": 40545 + }, + { + "epoch": 0.9504620920433434, + "grad_norm": 0.7137613296508789, + "learning_rate": 1.208854675874027e-06, + "loss": 0.0563, + "step": 40546 + }, + { + "epoch": 0.9504855336181485, + "grad_norm": 2.194660186767578, + "learning_rate": 1.207713047808212e-06, + "loss": 0.1066, + "step": 40547 + }, + { + "epoch": 0.9505089751929534, + "grad_norm": 0.7114189863204956, + "learning_rate": 1.2065719557921128e-06, + "loss": 0.0592, + "step": 40548 + }, + { + "epoch": 0.9505324167677585, + "grad_norm": 3.117420196533203, + "learning_rate": 1.2054313998319467e-06, + "loss": 0.2602, + "step": 40549 + }, + { + "epoch": 0.9505558583425635, + "grad_norm": 0.8879574537277222, + "learning_rate": 1.204291379933875e-06, + "loss": 0.0783, + "step": 40550 + }, + { + "epoch": 0.9505792999173684, + "grad_norm": 4.005757808685303, + "learning_rate": 1.2031518961041043e-06, + "loss": 0.45, + "step": 40551 + }, + { + "epoch": 0.9506027414921735, + "grad_norm": 2.484940767288208, + "learning_rate": 1.2020129483488074e-06, + "loss": 0.1437, + "step": 40552 + }, + { + "epoch": 0.9506261830669784, + "grad_norm": 4.977088928222656, + "learning_rate": 1.2008745366741569e-06, + "loss": 0.2928, + "step": 40553 + }, + { + "epoch": 0.9506496246417835, + "grad_norm": 1.9342892169952393, + "learning_rate": 1.199736661086348e-06, + "loss": 0.0671, + "step": 40554 + }, + { + "epoch": 0.9506730662165884, + "grad_norm": 0.9207454919815063, + "learning_rate": 1.1985993215915536e-06, + "loss": 0.0274, + "step": 40555 + }, + { + "epoch": 0.9506965077913935, + "grad_norm": 2.570732593536377, + "learning_rate": 1.1974625181959242e-06, + "loss": 0.1582, + "step": 40556 + }, + { + "epoch": 0.9507199493661984, + "grad_norm": 3.726020097732544, + "learning_rate": 1.196326250905655e-06, + "loss": 0.1659, + "step": 40557 + }, + { + "epoch": 0.9507433909410035, + "grad_norm": 0.8301217555999756, + "learning_rate": 1.1951905197268854e-06, + "loss": 0.053, + "step": 40558 + }, + { + "epoch": 0.9507668325158084, + "grad_norm": 2.1767923831939697, + "learning_rate": 1.194055324665788e-06, + "loss": 0.1485, + "step": 40559 + }, + { + "epoch": 0.9507902740906135, + "grad_norm": 3.4652018547058105, + "learning_rate": 1.1929206657285364e-06, + "loss": 0.3047, + "step": 40560 + }, + { + "epoch": 0.9508137156654184, + "grad_norm": 2.5725131034851074, + "learning_rate": 1.1917865429212693e-06, + "loss": 0.488, + "step": 40561 + }, + { + "epoch": 0.9508371572402234, + "grad_norm": 3.0032896995544434, + "learning_rate": 1.1906529562501378e-06, + "loss": 0.3123, + "step": 40562 + }, + { + "epoch": 0.9508605988150284, + "grad_norm": 0.9061135649681091, + "learning_rate": 1.1895199057213147e-06, + "loss": 0.0571, + "step": 40563 + }, + { + "epoch": 0.9508840403898334, + "grad_norm": 0.27625617384910583, + "learning_rate": 1.1883873913409393e-06, + "loss": 0.0163, + "step": 40564 + }, + { + "epoch": 0.9509074819646384, + "grad_norm": 3.822050094604492, + "learning_rate": 1.1872554131151291e-06, + "loss": 0.2168, + "step": 40565 + }, + { + "epoch": 0.9509309235394434, + "grad_norm": 0.3286789059638977, + "learning_rate": 1.1861239710500683e-06, + "loss": 0.0172, + "step": 40566 + }, + { + "epoch": 0.9509543651142484, + "grad_norm": 2.5288503170013428, + "learning_rate": 1.1849930651518737e-06, + "loss": 0.205, + "step": 40567 + }, + { + "epoch": 0.9509778066890534, + "grad_norm": 3.5706465244293213, + "learning_rate": 1.1838626954266851e-06, + "loss": 0.3479, + "step": 40568 + }, + { + "epoch": 0.9510012482638583, + "grad_norm": 2.54284930229187, + "learning_rate": 1.1827328618806422e-06, + "loss": 0.192, + "step": 40569 + }, + { + "epoch": 0.9510246898386634, + "grad_norm": 0.8421108722686768, + "learning_rate": 1.181603564519862e-06, + "loss": 0.0543, + "step": 40570 + }, + { + "epoch": 0.9510481314134683, + "grad_norm": 0.6413118839263916, + "learning_rate": 1.1804748033504732e-06, + "loss": 0.0516, + "step": 40571 + }, + { + "epoch": 0.9510715729882734, + "grad_norm": 0.582316517829895, + "learning_rate": 1.1793465783786151e-06, + "loss": 0.0124, + "step": 40572 + }, + { + "epoch": 0.9510950145630783, + "grad_norm": 0.60512375831604, + "learning_rate": 1.1782188896104052e-06, + "loss": 0.0518, + "step": 40573 + }, + { + "epoch": 0.9511184561378834, + "grad_norm": 1.9477722644805908, + "learning_rate": 1.1770917370519497e-06, + "loss": 0.1839, + "step": 40574 + }, + { + "epoch": 0.9511418977126883, + "grad_norm": 0.9870845675468445, + "learning_rate": 1.175965120709377e-06, + "loss": 0.0622, + "step": 40575 + }, + { + "epoch": 0.9511653392874934, + "grad_norm": 2.5249011516571045, + "learning_rate": 1.1748390405887933e-06, + "loss": 0.0776, + "step": 40576 + }, + { + "epoch": 0.9511887808622983, + "grad_norm": 4.249883651733398, + "learning_rate": 1.1737134966963158e-06, + "loss": 0.5584, + "step": 40577 + }, + { + "epoch": 0.9512122224371033, + "grad_norm": 0.7153325080871582, + "learning_rate": 1.1725884890380401e-06, + "loss": 0.0382, + "step": 40578 + }, + { + "epoch": 0.9512356640119083, + "grad_norm": 2.318565607070923, + "learning_rate": 1.1714640176200831e-06, + "loss": 0.1547, + "step": 40579 + }, + { + "epoch": 0.9512591055867133, + "grad_norm": 2.5046749114990234, + "learning_rate": 1.17034008244854e-06, + "loss": 0.1606, + "step": 40580 + }, + { + "epoch": 0.9512825471615183, + "grad_norm": 1.0377333164215088, + "learning_rate": 1.1692166835295172e-06, + "loss": 0.0425, + "step": 40581 + }, + { + "epoch": 0.9513059887363233, + "grad_norm": 2.0109527111053467, + "learning_rate": 1.1680938208690983e-06, + "loss": 0.1725, + "step": 40582 + }, + { + "epoch": 0.9513294303111283, + "grad_norm": 0.3266054391860962, + "learning_rate": 1.1669714944733901e-06, + "loss": 0.0266, + "step": 40583 + }, + { + "epoch": 0.9513528718859333, + "grad_norm": 0.5424845814704895, + "learning_rate": 1.165849704348454e-06, + "loss": 0.0367, + "step": 40584 + }, + { + "epoch": 0.9513763134607383, + "grad_norm": 0.7874343991279602, + "learning_rate": 1.1647284505004075e-06, + "loss": 0.0601, + "step": 40585 + }, + { + "epoch": 0.9513997550355433, + "grad_norm": 3.4160513877868652, + "learning_rate": 1.1636077329353235e-06, + "loss": 0.1739, + "step": 40586 + }, + { + "epoch": 0.9514231966103482, + "grad_norm": 1.2036627531051636, + "learning_rate": 1.162487551659275e-06, + "loss": 0.067, + "step": 40587 + }, + { + "epoch": 0.9514466381851533, + "grad_norm": 3.394394636154175, + "learning_rate": 1.1613679066783568e-06, + "loss": 0.3834, + "step": 40588 + }, + { + "epoch": 0.9514700797599582, + "grad_norm": 2.663215398788452, + "learning_rate": 1.1602487979986421e-06, + "loss": 0.2385, + "step": 40589 + }, + { + "epoch": 0.9514935213347633, + "grad_norm": 2.662283182144165, + "learning_rate": 1.1591302256261815e-06, + "loss": 0.1892, + "step": 40590 + }, + { + "epoch": 0.9515169629095682, + "grad_norm": 4.449678897857666, + "learning_rate": 1.1580121895670704e-06, + "loss": 0.4733, + "step": 40591 + }, + { + "epoch": 0.9515404044843733, + "grad_norm": 0.42818695306777954, + "learning_rate": 1.1568946898273591e-06, + "loss": 0.0273, + "step": 40592 + }, + { + "epoch": 0.9515638460591782, + "grad_norm": 2.437810182571411, + "learning_rate": 1.155777726413121e-06, + "loss": 0.0847, + "step": 40593 + }, + { + "epoch": 0.9515872876339833, + "grad_norm": 0.914115846157074, + "learning_rate": 1.1546612993304062e-06, + "loss": 0.1162, + "step": 40594 + }, + { + "epoch": 0.9516107292087882, + "grad_norm": 1.1157058477401733, + "learning_rate": 1.1535454085852881e-06, + "loss": 0.1372, + "step": 40595 + }, + { + "epoch": 0.9516341707835932, + "grad_norm": 1.6956015825271606, + "learning_rate": 1.152430054183795e-06, + "loss": 0.0915, + "step": 40596 + }, + { + "epoch": 0.9516576123583982, + "grad_norm": 4.623432636260986, + "learning_rate": 1.1513152361320112e-06, + "loss": 0.3738, + "step": 40597 + }, + { + "epoch": 0.9516810539332032, + "grad_norm": 2.9962356090545654, + "learning_rate": 1.1502009544359648e-06, + "loss": 1.54, + "step": 40598 + }, + { + "epoch": 0.9517044955080082, + "grad_norm": 2.288236379623413, + "learning_rate": 1.1490872091016958e-06, + "loss": 0.124, + "step": 40599 + }, + { + "epoch": 0.9517279370828132, + "grad_norm": 3.178579568862915, + "learning_rate": 1.1479740001352769e-06, + "loss": 0.1455, + "step": 40600 + }, + { + "epoch": 0.9517513786576182, + "grad_norm": 3.5111382007598877, + "learning_rate": 1.1468613275427253e-06, + "loss": 1.2801, + "step": 40601 + }, + { + "epoch": 0.9517748202324232, + "grad_norm": 1.8214610815048218, + "learning_rate": 1.145749191330081e-06, + "loss": 0.0952, + "step": 40602 + }, + { + "epoch": 0.9517982618072283, + "grad_norm": 2.34000563621521, + "learning_rate": 1.1446375915033835e-06, + "loss": 0.1717, + "step": 40603 + }, + { + "epoch": 0.9518217033820332, + "grad_norm": 3.4993484020233154, + "learning_rate": 1.1435265280686614e-06, + "loss": 0.2349, + "step": 40604 + }, + { + "epoch": 0.9518451449568383, + "grad_norm": 0.6266669631004333, + "learning_rate": 1.1424160010319429e-06, + "loss": 0.0433, + "step": 40605 + }, + { + "epoch": 0.9518685865316432, + "grad_norm": 2.7846715450286865, + "learning_rate": 1.1413060103992568e-06, + "loss": 0.2891, + "step": 40606 + }, + { + "epoch": 0.9518920281064482, + "grad_norm": 3.426800489425659, + "learning_rate": 1.1401965561766315e-06, + "loss": 0.3219, + "step": 40607 + }, + { + "epoch": 0.9519154696812532, + "grad_norm": 2.2487452030181885, + "learning_rate": 1.1390876383700732e-06, + "loss": 0.1075, + "step": 40608 + }, + { + "epoch": 0.9519389112560582, + "grad_norm": 4.209052562713623, + "learning_rate": 1.1379792569855886e-06, + "loss": 0.145, + "step": 40609 + }, + { + "epoch": 0.9519623528308632, + "grad_norm": 2.581277370452881, + "learning_rate": 1.1368714120292279e-06, + "loss": 0.3378, + "step": 40610 + }, + { + "epoch": 0.9519857944056682, + "grad_norm": 2.0781219005584717, + "learning_rate": 1.1357641035069756e-06, + "loss": 0.2034, + "step": 40611 + }, + { + "epoch": 0.9520092359804732, + "grad_norm": 1.5470798015594482, + "learning_rate": 1.1346573314248487e-06, + "loss": 0.1506, + "step": 40612 + }, + { + "epoch": 0.9520326775552782, + "grad_norm": 2.7690017223358154, + "learning_rate": 1.133551095788854e-06, + "loss": 0.1613, + "step": 40613 + }, + { + "epoch": 0.9520561191300831, + "grad_norm": 1.2140721082687378, + "learning_rate": 1.1324453966049864e-06, + "loss": 0.0923, + "step": 40614 + }, + { + "epoch": 0.9520795607048882, + "grad_norm": 0.3553030788898468, + "learning_rate": 1.1313402338792522e-06, + "loss": 0.0223, + "step": 40615 + }, + { + "epoch": 0.9521030022796931, + "grad_norm": 0.7110946178436279, + "learning_rate": 1.1302356076176358e-06, + "loss": 0.0618, + "step": 40616 + }, + { + "epoch": 0.9521264438544982, + "grad_norm": 2.7470412254333496, + "learning_rate": 1.1291315178261542e-06, + "loss": 0.0809, + "step": 40617 + }, + { + "epoch": 0.9521498854293031, + "grad_norm": 3.6114041805267334, + "learning_rate": 1.1280279645107806e-06, + "loss": 0.2605, + "step": 40618 + }, + { + "epoch": 0.9521733270041082, + "grad_norm": 2.8739607334136963, + "learning_rate": 1.12692494767751e-06, + "loss": 1.4401, + "step": 40619 + }, + { + "epoch": 0.9521967685789131, + "grad_norm": 7.77672004699707, + "learning_rate": 1.125822467332327e-06, + "loss": 0.4642, + "step": 40620 + }, + { + "epoch": 0.9522202101537182, + "grad_norm": 1.8435813188552856, + "learning_rate": 1.124720523481204e-06, + "loss": 1.308, + "step": 40621 + }, + { + "epoch": 0.9522436517285231, + "grad_norm": 1.598244547843933, + "learning_rate": 1.1236191161301257e-06, + "loss": 0.1617, + "step": 40622 + }, + { + "epoch": 0.9522670933033281, + "grad_norm": 3.2376997470855713, + "learning_rate": 1.1225182452850758e-06, + "loss": 0.2337, + "step": 40623 + }, + { + "epoch": 0.9522905348781331, + "grad_norm": 1.1227892637252808, + "learning_rate": 1.1214179109520163e-06, + "loss": 0.0517, + "step": 40624 + }, + { + "epoch": 0.9523139764529381, + "grad_norm": 1.9740185737609863, + "learning_rate": 1.1203181131369312e-06, + "loss": 0.1013, + "step": 40625 + }, + { + "epoch": 0.9523374180277431, + "grad_norm": 1.7199124097824097, + "learning_rate": 1.1192188518457824e-06, + "loss": 1.0176, + "step": 40626 + }, + { + "epoch": 0.9523608596025481, + "grad_norm": 0.4713924527168274, + "learning_rate": 1.118120127084532e-06, + "loss": 0.025, + "step": 40627 + }, + { + "epoch": 0.9523843011773531, + "grad_norm": 1.9765024185180664, + "learning_rate": 1.117021938859142e-06, + "loss": 1.4181, + "step": 40628 + }, + { + "epoch": 0.9524077427521581, + "grad_norm": 1.5979448556900024, + "learning_rate": 1.1159242871755515e-06, + "loss": 0.0474, + "step": 40629 + }, + { + "epoch": 0.952431184326963, + "grad_norm": 3.8536272048950195, + "learning_rate": 1.1148271720397563e-06, + "loss": 0.2243, + "step": 40630 + }, + { + "epoch": 0.9524546259017681, + "grad_norm": 2.887742757797241, + "learning_rate": 1.1137305934576847e-06, + "loss": 0.3038, + "step": 40631 + }, + { + "epoch": 0.952478067476573, + "grad_norm": 1.1635887622833252, + "learning_rate": 1.1126345514352987e-06, + "loss": 0.0625, + "step": 40632 + }, + { + "epoch": 0.9525015090513781, + "grad_norm": 1.128660798072815, + "learning_rate": 1.111539045978527e-06, + "loss": 0.1298, + "step": 40633 + }, + { + "epoch": 0.952524950626183, + "grad_norm": 4.511726379394531, + "learning_rate": 1.110444077093331e-06, + "loss": 0.1673, + "step": 40634 + }, + { + "epoch": 0.9525483922009881, + "grad_norm": 3.867422580718994, + "learning_rate": 1.1093496447856288e-06, + "loss": 0.4376, + "step": 40635 + }, + { + "epoch": 0.952571833775793, + "grad_norm": 3.059807777404785, + "learning_rate": 1.108255749061393e-06, + "loss": 0.1992, + "step": 40636 + }, + { + "epoch": 0.9525952753505981, + "grad_norm": 3.422701835632324, + "learning_rate": 1.1071623899265416e-06, + "loss": 0.3984, + "step": 40637 + }, + { + "epoch": 0.952618716925403, + "grad_norm": 2.9847724437713623, + "learning_rate": 1.1060695673870137e-06, + "loss": 0.2696, + "step": 40638 + }, + { + "epoch": 0.9526421585002081, + "grad_norm": 3.7671103477478027, + "learning_rate": 1.104977281448727e-06, + "loss": 0.286, + "step": 40639 + }, + { + "epoch": 0.952665600075013, + "grad_norm": 3.4913976192474365, + "learning_rate": 1.10388553211761e-06, + "loss": 0.3875, + "step": 40640 + }, + { + "epoch": 0.952689041649818, + "grad_norm": 2.661522388458252, + "learning_rate": 1.1027943193995915e-06, + "loss": 0.233, + "step": 40641 + }, + { + "epoch": 0.952712483224623, + "grad_norm": 0.890395998954773, + "learning_rate": 1.1017036433005889e-06, + "loss": 0.0493, + "step": 40642 + }, + { + "epoch": 0.952735924799428, + "grad_norm": 3.068882942199707, + "learning_rate": 1.1006135038265308e-06, + "loss": 0.2244, + "step": 40643 + }, + { + "epoch": 0.952759366374233, + "grad_norm": 2.957674026489258, + "learning_rate": 1.0995239009833236e-06, + "loss": 0.1905, + "step": 40644 + }, + { + "epoch": 0.952782807949038, + "grad_norm": 2.1843087673187256, + "learning_rate": 1.0984348347768847e-06, + "loss": 0.1569, + "step": 40645 + }, + { + "epoch": 0.952806249523843, + "grad_norm": 3.6154000759124756, + "learning_rate": 1.0973463052131095e-06, + "loss": 0.1984, + "step": 40646 + }, + { + "epoch": 0.952829691098648, + "grad_norm": 3.5957796573638916, + "learning_rate": 1.0962583122979152e-06, + "loss": 0.225, + "step": 40647 + }, + { + "epoch": 0.952853132673453, + "grad_norm": 2.5791172981262207, + "learning_rate": 1.0951708560371975e-06, + "loss": 0.297, + "step": 40648 + }, + { + "epoch": 0.952876574248258, + "grad_norm": 1.7654904127120972, + "learning_rate": 1.0940839364368738e-06, + "loss": 0.1027, + "step": 40649 + }, + { + "epoch": 0.9529000158230629, + "grad_norm": 2.2257497310638428, + "learning_rate": 1.092997553502839e-06, + "loss": 0.1131, + "step": 40650 + }, + { + "epoch": 0.952923457397868, + "grad_norm": 3.670722723007202, + "learning_rate": 1.0919117072409668e-06, + "loss": 0.2787, + "step": 40651 + }, + { + "epoch": 0.9529468989726729, + "grad_norm": 1.8936488628387451, + "learning_rate": 1.0908263976571743e-06, + "loss": 0.1874, + "step": 40652 + }, + { + "epoch": 0.952970340547478, + "grad_norm": 2.622119188308716, + "learning_rate": 1.0897416247573344e-06, + "loss": 0.2095, + "step": 40653 + }, + { + "epoch": 0.952993782122283, + "grad_norm": 7.389554977416992, + "learning_rate": 1.0886573885473316e-06, + "loss": 0.1597, + "step": 40654 + }, + { + "epoch": 0.953017223697088, + "grad_norm": 4.525871276855469, + "learning_rate": 1.0875736890330613e-06, + "loss": 0.387, + "step": 40655 + }, + { + "epoch": 0.953040665271893, + "grad_norm": 0.24188601970672607, + "learning_rate": 1.086490526220396e-06, + "loss": 0.0238, + "step": 40656 + }, + { + "epoch": 0.953064106846698, + "grad_norm": 2.4039528369903564, + "learning_rate": 1.0854079001152095e-06, + "loss": 0.1644, + "step": 40657 + }, + { + "epoch": 0.953087548421503, + "grad_norm": 1.5319645404815674, + "learning_rate": 1.0843258107233856e-06, + "loss": 0.0876, + "step": 40658 + }, + { + "epoch": 0.953110989996308, + "grad_norm": 4.47539758682251, + "learning_rate": 1.0832442580507973e-06, + "loss": 0.4105, + "step": 40659 + }, + { + "epoch": 0.953134431571113, + "grad_norm": 1.8996691703796387, + "learning_rate": 1.0821632421032957e-06, + "loss": 0.2108, + "step": 40660 + }, + { + "epoch": 0.9531578731459179, + "grad_norm": 1.0445574522018433, + "learning_rate": 1.0810827628867538e-06, + "loss": 0.0433, + "step": 40661 + }, + { + "epoch": 0.953181314720723, + "grad_norm": 1.620009183883667, + "learning_rate": 1.0800028204070555e-06, + "loss": 0.0854, + "step": 40662 + }, + { + "epoch": 0.9532047562955279, + "grad_norm": 1.5401264429092407, + "learning_rate": 1.07892341467003e-06, + "loss": 0.1593, + "step": 40663 + }, + { + "epoch": 0.953228197870333, + "grad_norm": 3.34818696975708, + "learning_rate": 1.07784454568155e-06, + "loss": 0.1383, + "step": 40664 + }, + { + "epoch": 0.9532516394451379, + "grad_norm": 1.9570133686065674, + "learning_rate": 1.0767662134474777e-06, + "loss": 1.5594, + "step": 40665 + }, + { + "epoch": 0.953275081019943, + "grad_norm": 1.525823712348938, + "learning_rate": 1.0756884179736304e-06, + "loss": 0.124, + "step": 40666 + }, + { + "epoch": 0.9532985225947479, + "grad_norm": 0.6700523495674133, + "learning_rate": 1.0746111592659035e-06, + "loss": 0.0557, + "step": 40667 + }, + { + "epoch": 0.953321964169553, + "grad_norm": 2.435040235519409, + "learning_rate": 1.0735344373301036e-06, + "loss": 1.788, + "step": 40668 + }, + { + "epoch": 0.9533454057443579, + "grad_norm": 2.480262279510498, + "learning_rate": 1.0724582521720928e-06, + "loss": 0.1652, + "step": 40669 + }, + { + "epoch": 0.9533688473191629, + "grad_norm": 1.2861334085464478, + "learning_rate": 1.0713826037977103e-06, + "loss": 0.0695, + "step": 40670 + }, + { + "epoch": 0.9533922888939679, + "grad_norm": 2.4276013374328613, + "learning_rate": 1.0703074922127853e-06, + "loss": 0.2003, + "step": 40671 + }, + { + "epoch": 0.9534157304687729, + "grad_norm": 0.7655103802680969, + "learning_rate": 1.0692329174231352e-06, + "loss": 0.0773, + "step": 40672 + }, + { + "epoch": 0.9534391720435779, + "grad_norm": 2.154766082763672, + "learning_rate": 1.068158879434622e-06, + "loss": 0.1214, + "step": 40673 + }, + { + "epoch": 0.9534626136183829, + "grad_norm": 3.19012713432312, + "learning_rate": 1.0670853782530632e-06, + "loss": 0.2129, + "step": 40674 + }, + { + "epoch": 0.9534860551931879, + "grad_norm": 2.951777219772339, + "learning_rate": 1.0660124138842764e-06, + "loss": 1.6054, + "step": 40675 + }, + { + "epoch": 0.9535094967679929, + "grad_norm": 3.3077244758605957, + "learning_rate": 1.0649399863340903e-06, + "loss": 0.3016, + "step": 40676 + }, + { + "epoch": 0.9535329383427978, + "grad_norm": 3.072589874267578, + "learning_rate": 1.0638680956083336e-06, + "loss": 0.2951, + "step": 40677 + }, + { + "epoch": 0.9535563799176029, + "grad_norm": 0.982776939868927, + "learning_rate": 1.0627967417127905e-06, + "loss": 0.0873, + "step": 40678 + }, + { + "epoch": 0.9535798214924078, + "grad_norm": 2.3478615283966064, + "learning_rate": 1.0617259246533006e-06, + "loss": 0.1782, + "step": 40679 + }, + { + "epoch": 0.9536032630672129, + "grad_norm": 2.1975502967834473, + "learning_rate": 1.0606556444356709e-06, + "loss": 0.1443, + "step": 40680 + }, + { + "epoch": 0.9536267046420178, + "grad_norm": 0.5837947726249695, + "learning_rate": 1.0595859010657073e-06, + "loss": 0.0608, + "step": 40681 + }, + { + "epoch": 0.9536501462168229, + "grad_norm": 3.6567628383636475, + "learning_rate": 1.0585166945492165e-06, + "loss": 0.3666, + "step": 40682 + }, + { + "epoch": 0.9536735877916278, + "grad_norm": 0.268621027469635, + "learning_rate": 1.0574480248919826e-06, + "loss": 0.0111, + "step": 40683 + }, + { + "epoch": 0.9536970293664329, + "grad_norm": 0.31921032071113586, + "learning_rate": 1.0563798920998235e-06, + "loss": 0.0188, + "step": 40684 + }, + { + "epoch": 0.9537204709412378, + "grad_norm": 0.7513546347618103, + "learning_rate": 1.0553122961785344e-06, + "loss": 0.0569, + "step": 40685 + }, + { + "epoch": 0.9537439125160428, + "grad_norm": 2.437237024307251, + "learning_rate": 1.0542452371338885e-06, + "loss": 0.2185, + "step": 40686 + }, + { + "epoch": 0.9537673540908478, + "grad_norm": 2.8823459148406982, + "learning_rate": 1.053178714971703e-06, + "loss": 0.2263, + "step": 40687 + }, + { + "epoch": 0.9537907956656528, + "grad_norm": 1.6813957691192627, + "learning_rate": 1.0521127296977518e-06, + "loss": 0.1052, + "step": 40688 + }, + { + "epoch": 0.9538142372404578, + "grad_norm": 0.9645352959632874, + "learning_rate": 1.0510472813178185e-06, + "loss": 0.0607, + "step": 40689 + }, + { + "epoch": 0.9538376788152628, + "grad_norm": 1.910610556602478, + "learning_rate": 1.0499823698376876e-06, + "loss": 1.2287, + "step": 40690 + }, + { + "epoch": 0.9538611203900678, + "grad_norm": 1.870688796043396, + "learning_rate": 1.0489179952631212e-06, + "loss": 1.0248, + "step": 40691 + }, + { + "epoch": 0.9538845619648728, + "grad_norm": 2.734593391418457, + "learning_rate": 1.0478541575999145e-06, + "loss": 0.4826, + "step": 40692 + }, + { + "epoch": 0.9539080035396778, + "grad_norm": 1.4017716646194458, + "learning_rate": 1.046790856853841e-06, + "loss": 1.2535, + "step": 40693 + }, + { + "epoch": 0.9539314451144828, + "grad_norm": 4.604560852050781, + "learning_rate": 1.0457280930306513e-06, + "loss": 0.4676, + "step": 40694 + }, + { + "epoch": 0.9539548866892877, + "grad_norm": 4.007316589355469, + "learning_rate": 1.0446658661361408e-06, + "loss": 0.4503, + "step": 40695 + }, + { + "epoch": 0.9539783282640928, + "grad_norm": 3.2263906002044678, + "learning_rate": 1.0436041761760384e-06, + "loss": 0.3736, + "step": 40696 + }, + { + "epoch": 0.9540017698388977, + "grad_norm": 2.707651138305664, + "learning_rate": 1.0425430231561284e-06, + "loss": 0.2041, + "step": 40697 + }, + { + "epoch": 0.9540252114137028, + "grad_norm": 0.7963805198669434, + "learning_rate": 1.0414824070821617e-06, + "loss": 0.0789, + "step": 40698 + }, + { + "epoch": 0.9540486529885077, + "grad_norm": 1.2561076879501343, + "learning_rate": 1.040422327959889e-06, + "loss": 0.1399, + "step": 40699 + }, + { + "epoch": 0.9540720945633128, + "grad_norm": 2.064687728881836, + "learning_rate": 1.0393627857950727e-06, + "loss": 0.1429, + "step": 40700 + }, + { + "epoch": 0.9540955361381177, + "grad_norm": 4.4719390869140625, + "learning_rate": 1.0383037805934525e-06, + "loss": 0.4286, + "step": 40701 + }, + { + "epoch": 0.9541189777129228, + "grad_norm": 1.7064659595489502, + "learning_rate": 1.0372453123607905e-06, + "loss": 0.1576, + "step": 40702 + }, + { + "epoch": 0.9541424192877277, + "grad_norm": 2.0460901260375977, + "learning_rate": 1.0361873811028044e-06, + "loss": 0.1414, + "step": 40703 + }, + { + "epoch": 0.9541658608625327, + "grad_norm": 4.50952672958374, + "learning_rate": 1.0351299868252561e-06, + "loss": 0.3055, + "step": 40704 + }, + { + "epoch": 0.9541893024373378, + "grad_norm": 0.6261314153671265, + "learning_rate": 1.0340731295338635e-06, + "loss": 0.0667, + "step": 40705 + }, + { + "epoch": 0.9542127440121427, + "grad_norm": 2.916612148284912, + "learning_rate": 1.0330168092343884e-06, + "loss": 0.1462, + "step": 40706 + }, + { + "epoch": 0.9542361855869478, + "grad_norm": 2.3767378330230713, + "learning_rate": 1.0319610259325373e-06, + "loss": 0.1676, + "step": 40707 + }, + { + "epoch": 0.9542596271617527, + "grad_norm": 2.516310214996338, + "learning_rate": 1.0309057796340504e-06, + "loss": 0.0458, + "step": 40708 + }, + { + "epoch": 0.9542830687365578, + "grad_norm": 0.26239895820617676, + "learning_rate": 1.0298510703446563e-06, + "loss": 0.0355, + "step": 40709 + }, + { + "epoch": 0.9543065103113627, + "grad_norm": 3.086472749710083, + "learning_rate": 1.0287968980700725e-06, + "loss": 0.1126, + "step": 40710 + }, + { + "epoch": 0.9543299518861678, + "grad_norm": 3.0248332023620605, + "learning_rate": 1.0277432628160054e-06, + "loss": 0.1317, + "step": 40711 + }, + { + "epoch": 0.9543533934609727, + "grad_norm": 0.6138874292373657, + "learning_rate": 1.0266901645881955e-06, + "loss": 0.0393, + "step": 40712 + }, + { + "epoch": 0.9543768350357777, + "grad_norm": 0.9150004386901855, + "learning_rate": 1.0256376033923487e-06, + "loss": 0.0588, + "step": 40713 + }, + { + "epoch": 0.9544002766105827, + "grad_norm": 1.8222806453704834, + "learning_rate": 1.024585579234183e-06, + "loss": 0.0855, + "step": 40714 + }, + { + "epoch": 0.9544237181853877, + "grad_norm": 1.9388788938522339, + "learning_rate": 1.0235340921193936e-06, + "loss": 0.133, + "step": 40715 + }, + { + "epoch": 0.9544471597601927, + "grad_norm": 1.3330860137939453, + "learning_rate": 1.0224831420536873e-06, + "loss": 0.0513, + "step": 40716 + }, + { + "epoch": 0.9544706013349977, + "grad_norm": 1.819931149482727, + "learning_rate": 1.0214327290427706e-06, + "loss": 0.053, + "step": 40717 + }, + { + "epoch": 0.9544940429098027, + "grad_norm": 4.466858863830566, + "learning_rate": 1.0203828530923386e-06, + "loss": 0.2125, + "step": 40718 + }, + { + "epoch": 0.9545174844846077, + "grad_norm": 1.6725388765335083, + "learning_rate": 1.0193335142081096e-06, + "loss": 0.1014, + "step": 40719 + }, + { + "epoch": 0.9545409260594127, + "grad_norm": 1.051926612854004, + "learning_rate": 1.0182847123957339e-06, + "loss": 0.0595, + "step": 40720 + }, + { + "epoch": 0.9545643676342177, + "grad_norm": 4.033100128173828, + "learning_rate": 1.017236447660952e-06, + "loss": 0.4113, + "step": 40721 + }, + { + "epoch": 0.9545878092090226, + "grad_norm": 3.538645029067993, + "learning_rate": 1.0161887200094144e-06, + "loss": 0.2143, + "step": 40722 + }, + { + "epoch": 0.9546112507838277, + "grad_norm": 4.619296550750732, + "learning_rate": 1.015141529446828e-06, + "loss": 0.2933, + "step": 40723 + }, + { + "epoch": 0.9546346923586326, + "grad_norm": 2.1765713691711426, + "learning_rate": 1.014094875978866e-06, + "loss": 0.0938, + "step": 40724 + }, + { + "epoch": 0.9546581339334377, + "grad_norm": 0.7515145540237427, + "learning_rate": 1.0130487596112014e-06, + "loss": 0.0314, + "step": 40725 + }, + { + "epoch": 0.9546815755082426, + "grad_norm": 2.7383811473846436, + "learning_rate": 1.01200318034953e-06, + "loss": 0.2485, + "step": 40726 + }, + { + "epoch": 0.9547050170830477, + "grad_norm": 1.3725446462631226, + "learning_rate": 1.0109581381995026e-06, + "loss": 0.1524, + "step": 40727 + }, + { + "epoch": 0.9547284586578526, + "grad_norm": 3.2494893074035645, + "learning_rate": 1.0099136331668035e-06, + "loss": 0.3521, + "step": 40728 + }, + { + "epoch": 0.9547519002326577, + "grad_norm": 0.46966859698295593, + "learning_rate": 1.008869665257095e-06, + "loss": 0.0284, + "step": 40729 + }, + { + "epoch": 0.9547753418074626, + "grad_norm": 2.403759002685547, + "learning_rate": 1.0078262344760393e-06, + "loss": 0.238, + "step": 40730 + }, + { + "epoch": 0.9547987833822676, + "grad_norm": 2.928196668624878, + "learning_rate": 1.0067833408293092e-06, + "loss": 0.2664, + "step": 40731 + }, + { + "epoch": 0.9548222249570726, + "grad_norm": 0.5336623787879944, + "learning_rate": 1.0057409843225563e-06, + "loss": 0.0335, + "step": 40732 + }, + { + "epoch": 0.9548456665318776, + "grad_norm": 3.0510354042053223, + "learning_rate": 1.0046991649614201e-06, + "loss": 0.0899, + "step": 40733 + }, + { + "epoch": 0.9548691081066826, + "grad_norm": 1.075238585472107, + "learning_rate": 1.0036578827515852e-06, + "loss": 0.0298, + "step": 40734 + }, + { + "epoch": 0.9548925496814876, + "grad_norm": 2.1425392627716064, + "learning_rate": 1.0026171376986914e-06, + "loss": 0.091, + "step": 40735 + }, + { + "epoch": 0.9549159912562926, + "grad_norm": 0.6442045569419861, + "learning_rate": 1.0015769298083677e-06, + "loss": 0.0499, + "step": 40736 + }, + { + "epoch": 0.9549394328310976, + "grad_norm": 0.812361478805542, + "learning_rate": 1.000537259086276e-06, + "loss": 0.0295, + "step": 40737 + }, + { + "epoch": 0.9549628744059026, + "grad_norm": 1.7015302181243896, + "learning_rate": 9.994981255380565e-07, + "loss": 0.0897, + "step": 40738 + }, + { + "epoch": 0.9549863159807076, + "grad_norm": 3.068673610687256, + "learning_rate": 9.984595291693377e-07, + "loss": 0.2448, + "step": 40739 + }, + { + "epoch": 0.9550097575555125, + "grad_norm": 2.2123870849609375, + "learning_rate": 9.974214699857598e-07, + "loss": 0.115, + "step": 40740 + }, + { + "epoch": 0.9550331991303176, + "grad_norm": 3.9282286167144775, + "learning_rate": 9.963839479929625e-07, + "loss": 0.4006, + "step": 40741 + }, + { + "epoch": 0.9550566407051225, + "grad_norm": 0.7527332305908203, + "learning_rate": 9.95346963196564e-07, + "loss": 0.0961, + "step": 40742 + }, + { + "epoch": 0.9550800822799276, + "grad_norm": 2.748628854751587, + "learning_rate": 9.943105156022037e-07, + "loss": 0.1652, + "step": 40743 + }, + { + "epoch": 0.9551035238547325, + "grad_norm": 1.6397424936294556, + "learning_rate": 9.932746052154996e-07, + "loss": 0.137, + "step": 40744 + }, + { + "epoch": 0.9551269654295376, + "grad_norm": 2.4325149059295654, + "learning_rate": 9.922392320420582e-07, + "loss": 0.1847, + "step": 40745 + }, + { + "epoch": 0.9551504070043425, + "grad_norm": 2.5885472297668457, + "learning_rate": 9.912043960875195e-07, + "loss": 0.2412, + "step": 40746 + }, + { + "epoch": 0.9551738485791476, + "grad_norm": 1.1341663599014282, + "learning_rate": 9.901700973574902e-07, + "loss": 0.0793, + "step": 40747 + }, + { + "epoch": 0.9551972901539525, + "grad_norm": 2.2911269664764404, + "learning_rate": 9.891363358575768e-07, + "loss": 0.1169, + "step": 40748 + }, + { + "epoch": 0.9552207317287575, + "grad_norm": 0.42964404821395874, + "learning_rate": 9.88103111593397e-07, + "loss": 0.0141, + "step": 40749 + }, + { + "epoch": 0.9552441733035625, + "grad_norm": 0.9550474882125854, + "learning_rate": 9.870704245705575e-07, + "loss": 0.113, + "step": 40750 + }, + { + "epoch": 0.9552676148783675, + "grad_norm": 0.43140125274658203, + "learning_rate": 9.86038274794654e-07, + "loss": 0.0482, + "step": 40751 + }, + { + "epoch": 0.9552910564531725, + "grad_norm": 1.4131168127059937, + "learning_rate": 9.850066622712927e-07, + "loss": 0.1488, + "step": 40752 + }, + { + "epoch": 0.9553144980279775, + "grad_norm": 2.252178430557251, + "learning_rate": 9.839755870060697e-07, + "loss": 0.1717, + "step": 40753 + }, + { + "epoch": 0.9553379396027825, + "grad_norm": 1.9713300466537476, + "learning_rate": 9.829450490045799e-07, + "loss": 1.2639, + "step": 40754 + }, + { + "epoch": 0.9553613811775875, + "grad_norm": 2.565704345703125, + "learning_rate": 9.819150482724082e-07, + "loss": 0.1887, + "step": 40755 + }, + { + "epoch": 0.9553848227523926, + "grad_norm": 3.5983283519744873, + "learning_rate": 9.8088558481515e-07, + "loss": 0.339, + "step": 40756 + }, + { + "epoch": 0.9554082643271975, + "grad_norm": 2.6010076999664307, + "learning_rate": 9.798566586383895e-07, + "loss": 0.1773, + "step": 40757 + }, + { + "epoch": 0.9554317059020025, + "grad_norm": 1.608638882637024, + "learning_rate": 9.788282697477224e-07, + "loss": 0.1203, + "step": 40758 + }, + { + "epoch": 0.9554551474768075, + "grad_norm": 2.7938058376312256, + "learning_rate": 9.77800418148711e-07, + "loss": 0.1785, + "step": 40759 + }, + { + "epoch": 0.9554785890516125, + "grad_norm": 1.9869757890701294, + "learning_rate": 9.767731038469396e-07, + "loss": 0.1073, + "step": 40760 + }, + { + "epoch": 0.9555020306264175, + "grad_norm": 2.4417479038238525, + "learning_rate": 9.757463268479816e-07, + "loss": 1.093, + "step": 40761 + }, + { + "epoch": 0.9555254722012225, + "grad_norm": 2.96130633354187, + "learning_rate": 9.747200871573991e-07, + "loss": 0.3215, + "step": 40762 + }, + { + "epoch": 0.9555489137760275, + "grad_norm": 2.3239798545837402, + "learning_rate": 9.736943847807766e-07, + "loss": 0.1389, + "step": 40763 + }, + { + "epoch": 0.9555723553508325, + "grad_norm": 0.4291699528694153, + "learning_rate": 9.726692197236764e-07, + "loss": 0.0333, + "step": 40764 + }, + { + "epoch": 0.9555957969256375, + "grad_norm": 0.5153229832649231, + "learning_rate": 9.716445919916607e-07, + "loss": 0.0495, + "step": 40765 + }, + { + "epoch": 0.9556192385004425, + "grad_norm": 2.4747872352600098, + "learning_rate": 9.706205015902803e-07, + "loss": 0.2329, + "step": 40766 + }, + { + "epoch": 0.9556426800752474, + "grad_norm": 3.2612760066986084, + "learning_rate": 9.695969485250977e-07, + "loss": 1.4247, + "step": 40767 + }, + { + "epoch": 0.9556661216500525, + "grad_norm": 1.445029616355896, + "learning_rate": 9.685739328016642e-07, + "loss": 0.1256, + "step": 40768 + }, + { + "epoch": 0.9556895632248574, + "grad_norm": 1.711014986038208, + "learning_rate": 9.675514544255304e-07, + "loss": 0.1005, + "step": 40769 + }, + { + "epoch": 0.9557130047996625, + "grad_norm": 3.0191256999969482, + "learning_rate": 9.665295134022479e-07, + "loss": 0.1703, + "step": 40770 + }, + { + "epoch": 0.9557364463744674, + "grad_norm": 2.347168207168579, + "learning_rate": 9.655081097373676e-07, + "loss": 0.1012, + "step": 40771 + }, + { + "epoch": 0.9557598879492725, + "grad_norm": 3.4576056003570557, + "learning_rate": 9.644872434364293e-07, + "loss": 0.3246, + "step": 40772 + }, + { + "epoch": 0.9557833295240774, + "grad_norm": 2.296194314956665, + "learning_rate": 9.634669145049513e-07, + "loss": 0.0911, + "step": 40773 + }, + { + "epoch": 0.9558067710988825, + "grad_norm": 3.0611565113067627, + "learning_rate": 9.624471229485066e-07, + "loss": 0.2155, + "step": 40774 + }, + { + "epoch": 0.9558302126736874, + "grad_norm": 0.8598414659500122, + "learning_rate": 9.61427868772591e-07, + "loss": 0.0581, + "step": 40775 + }, + { + "epoch": 0.9558536542484924, + "grad_norm": 2.0750560760498047, + "learning_rate": 9.604091519827552e-07, + "loss": 0.1374, + "step": 40776 + }, + { + "epoch": 0.9558770958232974, + "grad_norm": 3.4794111251831055, + "learning_rate": 9.593909725845285e-07, + "loss": 0.1441, + "step": 40777 + }, + { + "epoch": 0.9559005373981024, + "grad_norm": 2.066304922103882, + "learning_rate": 9.583733305834397e-07, + "loss": 0.1092, + "step": 40778 + }, + { + "epoch": 0.9559239789729074, + "grad_norm": 0.2868356704711914, + "learning_rate": 9.573562259849844e-07, + "loss": 0.0145, + "step": 40779 + }, + { + "epoch": 0.9559474205477124, + "grad_norm": 0.7615112066268921, + "learning_rate": 9.563396587947028e-07, + "loss": 0.0693, + "step": 40780 + }, + { + "epoch": 0.9559708621225174, + "grad_norm": 2.1116368770599365, + "learning_rate": 9.553236290181234e-07, + "loss": 0.1933, + "step": 40781 + }, + { + "epoch": 0.9559943036973224, + "grad_norm": 3.7365963459014893, + "learning_rate": 9.543081366607198e-07, + "loss": 0.2638, + "step": 40782 + }, + { + "epoch": 0.9560177452721274, + "grad_norm": 1.0381587743759155, + "learning_rate": 9.532931817280432e-07, + "loss": 0.1007, + "step": 40783 + }, + { + "epoch": 0.9560411868469324, + "grad_norm": 2.492880344390869, + "learning_rate": 9.522787642255781e-07, + "loss": 1.5729, + "step": 40784 + }, + { + "epoch": 0.9560646284217373, + "grad_norm": 1.9629180431365967, + "learning_rate": 9.512648841588312e-07, + "loss": 0.1786, + "step": 40785 + }, + { + "epoch": 0.9560880699965424, + "grad_norm": 1.7823313474655151, + "learning_rate": 9.502515415333091e-07, + "loss": 0.1287, + "step": 40786 + }, + { + "epoch": 0.9561115115713473, + "grad_norm": 0.6845256686210632, + "learning_rate": 9.492387363544964e-07, + "loss": 0.0642, + "step": 40787 + }, + { + "epoch": 0.9561349531461524, + "grad_norm": 1.4234240055084229, + "learning_rate": 9.482264686279108e-07, + "loss": 0.0703, + "step": 40788 + }, + { + "epoch": 0.9561583947209573, + "grad_norm": 2.255124568939209, + "learning_rate": 9.472147383590257e-07, + "loss": 1.4902, + "step": 40789 + }, + { + "epoch": 0.9561818362957624, + "grad_norm": 2.0451111793518066, + "learning_rate": 9.46203545553348e-07, + "loss": 0.1133, + "step": 40790 + }, + { + "epoch": 0.9562052778705673, + "grad_norm": 3.3150014877319336, + "learning_rate": 9.45192890216351e-07, + "loss": 0.3168, + "step": 40791 + }, + { + "epoch": 0.9562287194453724, + "grad_norm": 1.171244502067566, + "learning_rate": 9.44182772353519e-07, + "loss": 0.0685, + "step": 40792 + }, + { + "epoch": 0.9562521610201773, + "grad_norm": 3.3906924724578857, + "learning_rate": 9.431731919703368e-07, + "loss": 0.3343, + "step": 40793 + }, + { + "epoch": 0.9562756025949823, + "grad_norm": 2.602879047393799, + "learning_rate": 9.421641490722777e-07, + "loss": 0.2224, + "step": 40794 + }, + { + "epoch": 0.9562990441697873, + "grad_norm": 1.1069344282150269, + "learning_rate": 9.411556436648261e-07, + "loss": 0.109, + "step": 40795 + }, + { + "epoch": 0.9563224857445923, + "grad_norm": 2.1841182708740234, + "learning_rate": 9.401476757534555e-07, + "loss": 0.2601, + "step": 40796 + }, + { + "epoch": 0.9563459273193973, + "grad_norm": 3.075985908508301, + "learning_rate": 9.391402453436283e-07, + "loss": 0.3363, + "step": 40797 + }, + { + "epoch": 0.9563693688942023, + "grad_norm": 4.618722438812256, + "learning_rate": 9.381333524408065e-07, + "loss": 0.2203, + "step": 40798 + }, + { + "epoch": 0.9563928104690073, + "grad_norm": 4.140332221984863, + "learning_rate": 9.371269970504637e-07, + "loss": 0.3543, + "step": 40799 + }, + { + "epoch": 0.9564162520438123, + "grad_norm": 2.7407655715942383, + "learning_rate": 9.3612117917804e-07, + "loss": 0.1988, + "step": 40800 + }, + { + "epoch": 0.9564396936186172, + "grad_norm": 4.738866329193115, + "learning_rate": 9.351158988290199e-07, + "loss": 0.2362, + "step": 40801 + }, + { + "epoch": 0.9564631351934223, + "grad_norm": 0.5432906746864319, + "learning_rate": 9.341111560088545e-07, + "loss": 0.0364, + "step": 40802 + }, + { + "epoch": 0.9564865767682272, + "grad_norm": 2.306364059448242, + "learning_rate": 9.331069507229728e-07, + "loss": 0.9725, + "step": 40803 + }, + { + "epoch": 0.9565100183430323, + "grad_norm": 3.2813825607299805, + "learning_rate": 9.321032829768484e-07, + "loss": 0.3015, + "step": 40804 + }, + { + "epoch": 0.9565334599178372, + "grad_norm": 3.061079740524292, + "learning_rate": 9.3110015277591e-07, + "loss": 0.2862, + "step": 40805 + }, + { + "epoch": 0.9565569014926423, + "grad_norm": 2.2210092544555664, + "learning_rate": 9.30097560125609e-07, + "loss": 0.1166, + "step": 40806 + }, + { + "epoch": 0.9565803430674473, + "grad_norm": 2.7263994216918945, + "learning_rate": 9.290955050313855e-07, + "loss": 0.1142, + "step": 40807 + }, + { + "epoch": 0.9566037846422523, + "grad_norm": 2.053014039993286, + "learning_rate": 9.280939874986794e-07, + "loss": 0.1266, + "step": 40808 + }, + { + "epoch": 0.9566272262170573, + "grad_norm": 2.418123245239258, + "learning_rate": 9.270930075329198e-07, + "loss": 0.1853, + "step": 40809 + }, + { + "epoch": 0.9566506677918623, + "grad_norm": 0.6632936596870422, + "learning_rate": 9.260925651395358e-07, + "loss": 0.0902, + "step": 40810 + }, + { + "epoch": 0.9566741093666673, + "grad_norm": 0.2683175802230835, + "learning_rate": 9.250926603239562e-07, + "loss": 0.0146, + "step": 40811 + }, + { + "epoch": 0.9566975509414722, + "grad_norm": 1.205773949623108, + "learning_rate": 9.240932930916213e-07, + "loss": 0.0991, + "step": 40812 + }, + { + "epoch": 0.9567209925162773, + "grad_norm": 2.921464681625366, + "learning_rate": 9.230944634479266e-07, + "loss": 0.1754, + "step": 40813 + }, + { + "epoch": 0.9567444340910822, + "grad_norm": 1.018187165260315, + "learning_rate": 9.220961713983123e-07, + "loss": 0.0771, + "step": 40814 + }, + { + "epoch": 0.9567678756658873, + "grad_norm": 3.390472888946533, + "learning_rate": 9.210984169481962e-07, + "loss": 0.2633, + "step": 40815 + }, + { + "epoch": 0.9567913172406922, + "grad_norm": 3.2479944229125977, + "learning_rate": 9.201012001029851e-07, + "loss": 0.3223, + "step": 40816 + }, + { + "epoch": 0.9568147588154973, + "grad_norm": 4.1944169998168945, + "learning_rate": 9.19104520868086e-07, + "loss": 0.3426, + "step": 40817 + }, + { + "epoch": 0.9568382003903022, + "grad_norm": 3.794424295425415, + "learning_rate": 9.181083792489053e-07, + "loss": 0.2499, + "step": 40818 + }, + { + "epoch": 0.9568616419651073, + "grad_norm": 2.188161849975586, + "learning_rate": 9.171127752508724e-07, + "loss": 0.1064, + "step": 40819 + }, + { + "epoch": 0.9568850835399122, + "grad_norm": 2.9819118976593018, + "learning_rate": 9.161177088793493e-07, + "loss": 0.2796, + "step": 40820 + }, + { + "epoch": 0.9569085251147172, + "grad_norm": 1.949804425239563, + "learning_rate": 9.151231801397764e-07, + "loss": 0.1202, + "step": 40821 + }, + { + "epoch": 0.9569319666895222, + "grad_norm": 0.9458889365196228, + "learning_rate": 9.14129189037527e-07, + "loss": 0.0443, + "step": 40822 + }, + { + "epoch": 0.9569554082643272, + "grad_norm": 0.2106626331806183, + "learning_rate": 9.131357355779968e-07, + "loss": 0.0153, + "step": 40823 + }, + { + "epoch": 0.9569788498391322, + "grad_norm": 2.383901596069336, + "learning_rate": 9.121428197665816e-07, + "loss": 0.1953, + "step": 40824 + }, + { + "epoch": 0.9570022914139372, + "grad_norm": 0.9583751559257507, + "learning_rate": 9.111504416086548e-07, + "loss": 0.115, + "step": 40825 + }, + { + "epoch": 0.9570257329887422, + "grad_norm": 1.2359895706176758, + "learning_rate": 9.101586011096231e-07, + "loss": 0.1233, + "step": 40826 + }, + { + "epoch": 0.9570491745635472, + "grad_norm": 0.605000376701355, + "learning_rate": 9.091672982748489e-07, + "loss": 0.0514, + "step": 40827 + }, + { + "epoch": 0.9570726161383522, + "grad_norm": 2.85263991355896, + "learning_rate": 9.081765331097281e-07, + "loss": 0.2865, + "step": 40828 + }, + { + "epoch": 0.9570960577131572, + "grad_norm": 2.4372074604034424, + "learning_rate": 9.071863056196228e-07, + "loss": 0.1503, + "step": 40829 + }, + { + "epoch": 0.9571194992879621, + "grad_norm": 2.1755666732788086, + "learning_rate": 9.061966158099067e-07, + "loss": 0.2179, + "step": 40830 + }, + { + "epoch": 0.9571429408627672, + "grad_norm": 2.9931108951568604, + "learning_rate": 9.052074636859531e-07, + "loss": 0.1691, + "step": 40831 + }, + { + "epoch": 0.9571663824375721, + "grad_norm": 2.6365129947662354, + "learning_rate": 9.042188492531356e-07, + "loss": 0.1493, + "step": 40832 + }, + { + "epoch": 0.9571898240123772, + "grad_norm": 0.9427235722541809, + "learning_rate": 9.032307725168165e-07, + "loss": 0.1126, + "step": 40833 + }, + { + "epoch": 0.9572132655871821, + "grad_norm": 2.874730348587036, + "learning_rate": 9.022432334823472e-07, + "loss": 0.3529, + "step": 40834 + }, + { + "epoch": 0.9572367071619872, + "grad_norm": 2.4338126182556152, + "learning_rate": 9.01256232155101e-07, + "loss": 1.2987, + "step": 40835 + }, + { + "epoch": 0.9572601487367921, + "grad_norm": 0.36551305651664734, + "learning_rate": 9.002697685404182e-07, + "loss": 0.0221, + "step": 40836 + }, + { + "epoch": 0.9572835903115972, + "grad_norm": 3.5416812896728516, + "learning_rate": 8.992838426436611e-07, + "loss": 0.2348, + "step": 40837 + }, + { + "epoch": 0.9573070318864021, + "grad_norm": 2.036055326461792, + "learning_rate": 8.9829845447017e-07, + "loss": 0.1139, + "step": 40838 + }, + { + "epoch": 0.9573304734612071, + "grad_norm": 2.132627487182617, + "learning_rate": 8.973136040252961e-07, + "loss": 0.217, + "step": 40839 + }, + { + "epoch": 0.9573539150360121, + "grad_norm": 0.2989051938056946, + "learning_rate": 8.963292913143906e-07, + "loss": 0.0178, + "step": 40840 + }, + { + "epoch": 0.9573773566108171, + "grad_norm": 4.171248912811279, + "learning_rate": 8.953455163427938e-07, + "loss": 0.4142, + "step": 40841 + }, + { + "epoch": 0.9574007981856221, + "grad_norm": 0.40186789631843567, + "learning_rate": 8.943622791158235e-07, + "loss": 0.03, + "step": 40842 + }, + { + "epoch": 0.9574242397604271, + "grad_norm": 0.6124466061592102, + "learning_rate": 8.933795796388422e-07, + "loss": 0.0391, + "step": 40843 + }, + { + "epoch": 0.9574476813352321, + "grad_norm": 3.9900166988372803, + "learning_rate": 8.92397417917179e-07, + "loss": 0.2669, + "step": 40844 + }, + { + "epoch": 0.9574711229100371, + "grad_norm": 3.2309679985046387, + "learning_rate": 8.914157939561296e-07, + "loss": 0.2544, + "step": 40845 + }, + { + "epoch": 0.957494564484842, + "grad_norm": 3.5767621994018555, + "learning_rate": 8.904347077610676e-07, + "loss": 0.4288, + "step": 40846 + }, + { + "epoch": 0.9575180060596471, + "grad_norm": 2.2554588317871094, + "learning_rate": 8.894541593372774e-07, + "loss": 0.1511, + "step": 40847 + }, + { + "epoch": 0.957541447634452, + "grad_norm": 2.3425872325897217, + "learning_rate": 8.884741486901104e-07, + "loss": 0.1938, + "step": 40848 + }, + { + "epoch": 0.9575648892092571, + "grad_norm": 0.3159635066986084, + "learning_rate": 8.874946758248625e-07, + "loss": 0.0108, + "step": 40849 + }, + { + "epoch": 0.957588330784062, + "grad_norm": 2.58889102935791, + "learning_rate": 8.865157407468516e-07, + "loss": 0.1866, + "step": 40850 + }, + { + "epoch": 0.9576117723588671, + "grad_norm": 0.5923829078674316, + "learning_rate": 8.855373434613957e-07, + "loss": 0.0332, + "step": 40851 + }, + { + "epoch": 0.957635213933672, + "grad_norm": 2.413048028945923, + "learning_rate": 8.845594839738014e-07, + "loss": 0.0906, + "step": 40852 + }, + { + "epoch": 0.9576586555084771, + "grad_norm": 0.5872024297714233, + "learning_rate": 8.835821622893759e-07, + "loss": 0.0392, + "step": 40853 + }, + { + "epoch": 0.957682097083282, + "grad_norm": 2.931368827819824, + "learning_rate": 8.826053784134147e-07, + "loss": 0.1893, + "step": 40854 + }, + { + "epoch": 0.957705538658087, + "grad_norm": 1.4859365224838257, + "learning_rate": 8.816291323512249e-07, + "loss": 0.1369, + "step": 40855 + }, + { + "epoch": 0.957728980232892, + "grad_norm": 0.32812267541885376, + "learning_rate": 8.806534241081133e-07, + "loss": 0.0342, + "step": 40856 + }, + { + "epoch": 0.957752421807697, + "grad_norm": 0.7080690860748291, + "learning_rate": 8.796782536893533e-07, + "loss": 0.0486, + "step": 40857 + }, + { + "epoch": 0.9577758633825021, + "grad_norm": 3.8514564037323, + "learning_rate": 8.787036211002408e-07, + "loss": 0.1419, + "step": 40858 + }, + { + "epoch": 0.957799304957307, + "grad_norm": 1.2806048393249512, + "learning_rate": 8.777295263460828e-07, + "loss": 0.1645, + "step": 40859 + }, + { + "epoch": 0.9578227465321121, + "grad_norm": 3.9472293853759766, + "learning_rate": 8.767559694321415e-07, + "loss": 0.2171, + "step": 40860 + }, + { + "epoch": 0.957846188106917, + "grad_norm": 2.345823287963867, + "learning_rate": 8.757829503637238e-07, + "loss": 0.1359, + "step": 40861 + }, + { + "epoch": 0.9578696296817221, + "grad_norm": 3.5536773204803467, + "learning_rate": 8.748104691460813e-07, + "loss": 0.2016, + "step": 40862 + }, + { + "epoch": 0.957893071256527, + "grad_norm": 2.221791982650757, + "learning_rate": 8.738385257844983e-07, + "loss": 0.1941, + "step": 40863 + }, + { + "epoch": 0.9579165128313321, + "grad_norm": 4.18239164352417, + "learning_rate": 8.728671202842709e-07, + "loss": 0.2314, + "step": 40864 + }, + { + "epoch": 0.957939954406137, + "grad_norm": 0.348423570394516, + "learning_rate": 8.718962526506502e-07, + "loss": 0.0127, + "step": 40865 + }, + { + "epoch": 0.957963395980942, + "grad_norm": 2.934722661972046, + "learning_rate": 8.709259228888988e-07, + "loss": 0.1813, + "step": 40866 + }, + { + "epoch": 0.957986837555747, + "grad_norm": 3.3717551231384277, + "learning_rate": 8.699561310042903e-07, + "loss": 0.1851, + "step": 40867 + }, + { + "epoch": 0.958010279130552, + "grad_norm": 2.3976402282714844, + "learning_rate": 8.689868770020981e-07, + "loss": 0.1436, + "step": 40868 + }, + { + "epoch": 0.958033720705357, + "grad_norm": 3.1001222133636475, + "learning_rate": 8.680181608875626e-07, + "loss": 0.1917, + "step": 40869 + }, + { + "epoch": 0.958057162280162, + "grad_norm": 2.93520188331604, + "learning_rate": 8.670499826659462e-07, + "loss": 0.1997, + "step": 40870 + }, + { + "epoch": 0.958080603854967, + "grad_norm": 3.6501033306121826, + "learning_rate": 8.660823423425113e-07, + "loss": 0.1348, + "step": 40871 + }, + { + "epoch": 0.958104045429772, + "grad_norm": 0.5702388882637024, + "learning_rate": 8.651152399224982e-07, + "loss": 0.029, + "step": 40872 + }, + { + "epoch": 0.958127487004577, + "grad_norm": 0.24462354183197021, + "learning_rate": 8.641486754111582e-07, + "loss": 0.0071, + "step": 40873 + }, + { + "epoch": 0.958150928579382, + "grad_norm": 0.3062233626842499, + "learning_rate": 8.631826488137429e-07, + "loss": 0.0171, + "step": 40874 + }, + { + "epoch": 0.9581743701541869, + "grad_norm": 2.1165120601654053, + "learning_rate": 8.622171601354701e-07, + "loss": 0.112, + "step": 40875 + }, + { + "epoch": 0.958197811728992, + "grad_norm": 3.0417659282684326, + "learning_rate": 8.612522093816022e-07, + "loss": 0.1853, + "step": 40876 + }, + { + "epoch": 0.9582212533037969, + "grad_norm": 1.7987031936645508, + "learning_rate": 8.602877965573686e-07, + "loss": 1.0041, + "step": 40877 + }, + { + "epoch": 0.958244694878602, + "grad_norm": 0.659274697303772, + "learning_rate": 8.593239216679872e-07, + "loss": 0.0564, + "step": 40878 + }, + { + "epoch": 0.9582681364534069, + "grad_norm": 3.914719343185425, + "learning_rate": 8.583605847187203e-07, + "loss": 0.4202, + "step": 40879 + }, + { + "epoch": 0.958291578028212, + "grad_norm": 2.0332963466644287, + "learning_rate": 8.573977857147641e-07, + "loss": 0.0354, + "step": 40880 + }, + { + "epoch": 0.9583150196030169, + "grad_norm": 2.4452145099639893, + "learning_rate": 8.564355246613698e-07, + "loss": 0.259, + "step": 40881 + }, + { + "epoch": 0.958338461177822, + "grad_norm": 2.6143319606781006, + "learning_rate": 8.554738015637331e-07, + "loss": 0.2123, + "step": 40882 + }, + { + "epoch": 0.9583619027526269, + "grad_norm": 0.6049901247024536, + "learning_rate": 8.545126164270834e-07, + "loss": 0.034, + "step": 40883 + }, + { + "epoch": 0.9583853443274319, + "grad_norm": 1.8730112314224243, + "learning_rate": 8.535519692566385e-07, + "loss": 0.1924, + "step": 40884 + }, + { + "epoch": 0.9584087859022369, + "grad_norm": 2.604250431060791, + "learning_rate": 8.525918600576055e-07, + "loss": 0.2543, + "step": 40885 + }, + { + "epoch": 0.9584322274770419, + "grad_norm": 3.136268377304077, + "learning_rate": 8.516322888352024e-07, + "loss": 0.2323, + "step": 40886 + }, + { + "epoch": 0.9584556690518469, + "grad_norm": 1.9010953903198242, + "learning_rate": 8.506732555946251e-07, + "loss": 0.157, + "step": 40887 + }, + { + "epoch": 0.9584791106266519, + "grad_norm": 2.9698598384857178, + "learning_rate": 8.497147603410804e-07, + "loss": 0.2915, + "step": 40888 + }, + { + "epoch": 0.9585025522014569, + "grad_norm": 2.283505916595459, + "learning_rate": 8.487568030797755e-07, + "loss": 0.1037, + "step": 40889 + }, + { + "epoch": 0.9585259937762619, + "grad_norm": 3.009997606277466, + "learning_rate": 8.477993838159059e-07, + "loss": 0.1736, + "step": 40890 + }, + { + "epoch": 0.9585494353510668, + "grad_norm": 3.1978447437286377, + "learning_rate": 8.468425025546567e-07, + "loss": 0.3056, + "step": 40891 + }, + { + "epoch": 0.9585728769258719, + "grad_norm": 2.1240036487579346, + "learning_rate": 8.458861593012346e-07, + "loss": 0.0815, + "step": 40892 + }, + { + "epoch": 0.9585963185006768, + "grad_norm": 1.5128048658370972, + "learning_rate": 8.449303540608244e-07, + "loss": 0.0784, + "step": 40893 + }, + { + "epoch": 0.9586197600754819, + "grad_norm": 2.2788004875183105, + "learning_rate": 8.439750868386109e-07, + "loss": 0.1427, + "step": 40894 + }, + { + "epoch": 0.9586432016502868, + "grad_norm": 1.0482298135757446, + "learning_rate": 8.430203576397678e-07, + "loss": 0.1039, + "step": 40895 + }, + { + "epoch": 0.9586666432250919, + "grad_norm": 0.6230494379997253, + "learning_rate": 8.420661664694907e-07, + "loss": 0.0268, + "step": 40896 + }, + { + "epoch": 0.9586900847998968, + "grad_norm": 2.055532693862915, + "learning_rate": 8.411125133329422e-07, + "loss": 0.1183, + "step": 40897 + }, + { + "epoch": 0.9587135263747019, + "grad_norm": 3.6422159671783447, + "learning_rate": 8.401593982353184e-07, + "loss": 0.3228, + "step": 40898 + }, + { + "epoch": 0.9587369679495068, + "grad_norm": 3.058698892593384, + "learning_rate": 8.392068211817705e-07, + "loss": 0.1548, + "step": 40899 + }, + { + "epoch": 0.9587604095243119, + "grad_norm": 2.2386622428894043, + "learning_rate": 8.382547821774722e-07, + "loss": 0.1158, + "step": 40900 + }, + { + "epoch": 0.9587838510991168, + "grad_norm": 3.707216739654541, + "learning_rate": 8.373032812275972e-07, + "loss": 0.1914, + "step": 40901 + }, + { + "epoch": 0.9588072926739218, + "grad_norm": 2.66841459274292, + "learning_rate": 8.363523183372968e-07, + "loss": 0.2275, + "step": 40902 + }, + { + "epoch": 0.9588307342487268, + "grad_norm": 2.2809786796569824, + "learning_rate": 8.354018935117447e-07, + "loss": 1.4039, + "step": 40903 + }, + { + "epoch": 0.9588541758235318, + "grad_norm": 4.390146255493164, + "learning_rate": 8.344520067560813e-07, + "loss": 0.1381, + "step": 40904 + }, + { + "epoch": 0.9588776173983368, + "grad_norm": 1.9341504573822021, + "learning_rate": 8.335026580754801e-07, + "loss": 0.1098, + "step": 40905 + }, + { + "epoch": 0.9589010589731418, + "grad_norm": 0.695221483707428, + "learning_rate": 8.325538474750705e-07, + "loss": 0.063, + "step": 40906 + }, + { + "epoch": 0.9589245005479468, + "grad_norm": 0.8951438069343567, + "learning_rate": 8.31605574960026e-07, + "loss": 0.0607, + "step": 40907 + }, + { + "epoch": 0.9589479421227518, + "grad_norm": 0.8330568075180054, + "learning_rate": 8.306578405354648e-07, + "loss": 0.0871, + "step": 40908 + }, + { + "epoch": 0.9589713836975569, + "grad_norm": 5.184148788452148, + "learning_rate": 8.297106442065495e-07, + "loss": 0.4027, + "step": 40909 + }, + { + "epoch": 0.9589948252723618, + "grad_norm": 3.2101731300354004, + "learning_rate": 8.28763985978409e-07, + "loss": 0.2864, + "step": 40910 + }, + { + "epoch": 0.9590182668471668, + "grad_norm": 6.094471454620361, + "learning_rate": 8.278178658561841e-07, + "loss": 0.4603, + "step": 40911 + }, + { + "epoch": 0.9590417084219718, + "grad_norm": 2.5811312198638916, + "learning_rate": 8.268722838450149e-07, + "loss": 0.1345, + "step": 40912 + }, + { + "epoch": 0.9590651499967768, + "grad_norm": 0.8438000082969666, + "learning_rate": 8.259272399500195e-07, + "loss": 0.0692, + "step": 40913 + }, + { + "epoch": 0.9590885915715818, + "grad_norm": 1.4122555255889893, + "learning_rate": 8.249827341763273e-07, + "loss": 0.0506, + "step": 40914 + }, + { + "epoch": 0.9591120331463868, + "grad_norm": 1.8067872524261475, + "learning_rate": 8.240387665290672e-07, + "loss": 0.1783, + "step": 40915 + }, + { + "epoch": 0.9591354747211918, + "grad_norm": 2.7970364093780518, + "learning_rate": 8.230953370133688e-07, + "loss": 0.2329, + "step": 40916 + }, + { + "epoch": 0.9591589162959968, + "grad_norm": 2.8352181911468506, + "learning_rate": 8.221524456343388e-07, + "loss": 1.7883, + "step": 40917 + }, + { + "epoch": 0.9591823578708017, + "grad_norm": 0.48657259345054626, + "learning_rate": 8.212100923971067e-07, + "loss": 0.0318, + "step": 40918 + }, + { + "epoch": 0.9592057994456068, + "grad_norm": 1.8322300910949707, + "learning_rate": 8.202682773067683e-07, + "loss": 0.091, + "step": 40919 + }, + { + "epoch": 0.9592292410204117, + "grad_norm": 2.7044475078582764, + "learning_rate": 8.193270003684417e-07, + "loss": 0.237, + "step": 40920 + }, + { + "epoch": 0.9592526825952168, + "grad_norm": 1.0471076965332031, + "learning_rate": 8.18386261587234e-07, + "loss": 0.5109, + "step": 40921 + }, + { + "epoch": 0.9592761241700217, + "grad_norm": 3.7508294582366943, + "learning_rate": 8.174460609682633e-07, + "loss": 0.2713, + "step": 40922 + }, + { + "epoch": 0.9592995657448268, + "grad_norm": 1.1855863332748413, + "learning_rate": 8.165063985166143e-07, + "loss": 0.0541, + "step": 40923 + }, + { + "epoch": 0.9593230073196317, + "grad_norm": 1.5127599239349365, + "learning_rate": 8.155672742373833e-07, + "loss": 0.1069, + "step": 40924 + }, + { + "epoch": 0.9593464488944368, + "grad_norm": 1.7208874225616455, + "learning_rate": 8.146286881356769e-07, + "loss": 0.1856, + "step": 40925 + }, + { + "epoch": 0.9593698904692417, + "grad_norm": 2.2569899559020996, + "learning_rate": 8.136906402165912e-07, + "loss": 0.2461, + "step": 40926 + }, + { + "epoch": 0.9593933320440468, + "grad_norm": 0.5562609434127808, + "learning_rate": 8.127531304851998e-07, + "loss": 0.0535, + "step": 40927 + }, + { + "epoch": 0.9594167736188517, + "grad_norm": 2.187124252319336, + "learning_rate": 8.118161589465989e-07, + "loss": 0.1672, + "step": 40928 + }, + { + "epoch": 0.9594402151936567, + "grad_norm": 3.335789442062378, + "learning_rate": 8.108797256058731e-07, + "loss": 0.1822, + "step": 40929 + }, + { + "epoch": 0.9594636567684617, + "grad_norm": 3.824897289276123, + "learning_rate": 8.099438304681073e-07, + "loss": 0.1424, + "step": 40930 + }, + { + "epoch": 0.9594870983432667, + "grad_norm": 1.1071596145629883, + "learning_rate": 8.090084735383751e-07, + "loss": 0.0908, + "step": 40931 + }, + { + "epoch": 0.9595105399180717, + "grad_norm": 1.9059503078460693, + "learning_rate": 8.080736548217505e-07, + "loss": 0.147, + "step": 40932 + }, + { + "epoch": 0.9595339814928767, + "grad_norm": 2.8576061725616455, + "learning_rate": 8.071393743233069e-07, + "loss": 0.2663, + "step": 40933 + }, + { + "epoch": 0.9595574230676817, + "grad_norm": 2.299013614654541, + "learning_rate": 8.06205632048107e-07, + "loss": 0.1346, + "step": 40934 + }, + { + "epoch": 0.9595808646424867, + "grad_norm": 3.336075782775879, + "learning_rate": 8.052724280012358e-07, + "loss": 0.2135, + "step": 40935 + }, + { + "epoch": 0.9596043062172916, + "grad_norm": 2.58613657951355, + "learning_rate": 8.043397621877446e-07, + "loss": 0.1993, + "step": 40936 + }, + { + "epoch": 0.9596277477920967, + "grad_norm": 3.6276910305023193, + "learning_rate": 8.03407634612685e-07, + "loss": 0.5477, + "step": 40937 + }, + { + "epoch": 0.9596511893669016, + "grad_norm": 2.4039478302001953, + "learning_rate": 8.02476045281142e-07, + "loss": 0.1149, + "step": 40938 + }, + { + "epoch": 0.9596746309417067, + "grad_norm": 1.4822670221328735, + "learning_rate": 8.015449941981334e-07, + "loss": 0.0855, + "step": 40939 + }, + { + "epoch": 0.9596980725165116, + "grad_norm": 0.7861407995223999, + "learning_rate": 8.006144813687333e-07, + "loss": 0.0847, + "step": 40940 + }, + { + "epoch": 0.9597215140913167, + "grad_norm": 3.2772228717803955, + "learning_rate": 7.99684506797993e-07, + "loss": 0.0836, + "step": 40941 + }, + { + "epoch": 0.9597449556661216, + "grad_norm": 1.2566235065460205, + "learning_rate": 7.987550704909641e-07, + "loss": 0.0867, + "step": 40942 + }, + { + "epoch": 0.9597683972409267, + "grad_norm": 1.140310525894165, + "learning_rate": 7.978261724526648e-07, + "loss": 0.1281, + "step": 40943 + }, + { + "epoch": 0.9597918388157316, + "grad_norm": 1.9799565076828003, + "learning_rate": 7.968978126881577e-07, + "loss": 0.1235, + "step": 40944 + }, + { + "epoch": 0.9598152803905367, + "grad_norm": 4.179902076721191, + "learning_rate": 7.959699912024721e-07, + "loss": 0.4094, + "step": 40945 + }, + { + "epoch": 0.9598387219653416, + "grad_norm": 1.9109740257263184, + "learning_rate": 7.950427080006373e-07, + "loss": 0.0905, + "step": 40946 + }, + { + "epoch": 0.9598621635401466, + "grad_norm": 3.8467154502868652, + "learning_rate": 7.941159630877049e-07, + "loss": 0.3364, + "step": 40947 + }, + { + "epoch": 0.9598856051149516, + "grad_norm": 2.1529998779296875, + "learning_rate": 7.931897564686707e-07, + "loss": 0.1627, + "step": 40948 + }, + { + "epoch": 0.9599090466897566, + "grad_norm": 2.2944436073303223, + "learning_rate": 7.922640881485865e-07, + "loss": 0.1089, + "step": 40949 + }, + { + "epoch": 0.9599324882645616, + "grad_norm": 8.312536239624023, + "learning_rate": 7.913389581324704e-07, + "loss": 0.1998, + "step": 40950 + }, + { + "epoch": 0.9599559298393666, + "grad_norm": 3.250589370727539, + "learning_rate": 7.904143664253294e-07, + "loss": 0.2207, + "step": 40951 + }, + { + "epoch": 0.9599793714141716, + "grad_norm": 1.4403659105300903, + "learning_rate": 7.894903130321929e-07, + "loss": 0.0394, + "step": 40952 + }, + { + "epoch": 0.9600028129889766, + "grad_norm": 2.2478387355804443, + "learning_rate": 7.88566797958068e-07, + "loss": 0.0765, + "step": 40953 + }, + { + "epoch": 0.9600262545637815, + "grad_norm": 2.637126922607422, + "learning_rate": 7.87643821207984e-07, + "loss": 0.1818, + "step": 40954 + }, + { + "epoch": 0.9600496961385866, + "grad_norm": 0.6105244159698486, + "learning_rate": 7.867213827869258e-07, + "loss": 0.0717, + "step": 40955 + }, + { + "epoch": 0.9600731377133915, + "grad_norm": 2.743544816970825, + "learning_rate": 7.857994826999004e-07, + "loss": 0.2358, + "step": 40956 + }, + { + "epoch": 0.9600965792881966, + "grad_norm": 0.8075942397117615, + "learning_rate": 7.848781209519263e-07, + "loss": 0.0526, + "step": 40957 + }, + { + "epoch": 0.9601200208630015, + "grad_norm": 2.0678603649139404, + "learning_rate": 7.839572975479881e-07, + "loss": 0.0643, + "step": 40958 + }, + { + "epoch": 0.9601434624378066, + "grad_norm": 2.1522092819213867, + "learning_rate": 7.830370124930819e-07, + "loss": 0.2214, + "step": 40959 + }, + { + "epoch": 0.9601669040126116, + "grad_norm": 0.7875931859016418, + "learning_rate": 7.821172657922149e-07, + "loss": 0.0384, + "step": 40960 + }, + { + "epoch": 0.9601903455874166, + "grad_norm": 1.3047608137130737, + "learning_rate": 7.811980574503718e-07, + "loss": 0.0934, + "step": 40961 + }, + { + "epoch": 0.9602137871622216, + "grad_norm": 0.8677595257759094, + "learning_rate": 7.802793874725378e-07, + "loss": 0.0956, + "step": 40962 + }, + { + "epoch": 0.9602372287370265, + "grad_norm": 1.4218742847442627, + "learning_rate": 7.793612558636864e-07, + "loss": 0.0491, + "step": 40963 + }, + { + "epoch": 0.9602606703118316, + "grad_norm": 2.15270733833313, + "learning_rate": 7.784436626288139e-07, + "loss": 0.2164, + "step": 40964 + }, + { + "epoch": 0.9602841118866365, + "grad_norm": 0.8496479392051697, + "learning_rate": 7.775266077729049e-07, + "loss": 0.1066, + "step": 40965 + }, + { + "epoch": 0.9603075534614416, + "grad_norm": 2.2645578384399414, + "learning_rate": 7.766100913009111e-07, + "loss": 0.1026, + "step": 40966 + }, + { + "epoch": 0.9603309950362465, + "grad_norm": 2.0025813579559326, + "learning_rate": 7.756941132178397e-07, + "loss": 1.085, + "step": 40967 + }, + { + "epoch": 0.9603544366110516, + "grad_norm": 3.1550416946411133, + "learning_rate": 7.74778673528631e-07, + "loss": 0.2586, + "step": 40968 + }, + { + "epoch": 0.9603778781858565, + "grad_norm": 1.488699197769165, + "learning_rate": 7.738637722382703e-07, + "loss": 0.1276, + "step": 40969 + }, + { + "epoch": 0.9604013197606616, + "grad_norm": 1.1652663946151733, + "learning_rate": 7.729494093517086e-07, + "loss": 0.0471, + "step": 40970 + }, + { + "epoch": 0.9604247613354665, + "grad_norm": 4.790688991546631, + "learning_rate": 7.720355848739202e-07, + "loss": 0.275, + "step": 40971 + }, + { + "epoch": 0.9604482029102716, + "grad_norm": 1.3547934293746948, + "learning_rate": 7.711222988098455e-07, + "loss": 0.1371, + "step": 40972 + }, + { + "epoch": 0.9604716444850765, + "grad_norm": 2.838059425354004, + "learning_rate": 7.70209551164458e-07, + "loss": 1.5126, + "step": 40973 + }, + { + "epoch": 0.9604950860598815, + "grad_norm": 0.973391592502594, + "learning_rate": 7.692973419426985e-07, + "loss": 0.0843, + "step": 40974 + }, + { + "epoch": 0.9605185276346865, + "grad_norm": 3.5144708156585693, + "learning_rate": 7.683856711495296e-07, + "loss": 0.2074, + "step": 40975 + }, + { + "epoch": 0.9605419692094915, + "grad_norm": 2.990955114364624, + "learning_rate": 7.674745387898807e-07, + "loss": 0.4282, + "step": 40976 + }, + { + "epoch": 0.9605654107842965, + "grad_norm": 2.3856053352355957, + "learning_rate": 7.665639448687145e-07, + "loss": 0.1211, + "step": 40977 + }, + { + "epoch": 0.9605888523591015, + "grad_norm": 1.1622051000595093, + "learning_rate": 7.656538893909493e-07, + "loss": 0.0522, + "step": 40978 + }, + { + "epoch": 0.9606122939339065, + "grad_norm": 3.4065139293670654, + "learning_rate": 7.647443723615477e-07, + "loss": 0.2327, + "step": 40979 + }, + { + "epoch": 0.9606357355087115, + "grad_norm": 2.0227761268615723, + "learning_rate": 7.638353937854392e-07, + "loss": 0.1417, + "step": 40980 + }, + { + "epoch": 0.9606591770835164, + "grad_norm": 0.7535014152526855, + "learning_rate": 7.629269536675421e-07, + "loss": 0.0299, + "step": 40981 + }, + { + "epoch": 0.9606826186583215, + "grad_norm": 4.120227813720703, + "learning_rate": 7.620190520127968e-07, + "loss": 0.3613, + "step": 40982 + }, + { + "epoch": 0.9607060602331264, + "grad_norm": 2.5091347694396973, + "learning_rate": 7.611116888261216e-07, + "loss": 0.1005, + "step": 40983 + }, + { + "epoch": 0.9607295018079315, + "grad_norm": 3.7832775115966797, + "learning_rate": 7.60204864112446e-07, + "loss": 0.3959, + "step": 40984 + }, + { + "epoch": 0.9607529433827364, + "grad_norm": 1.7006710767745972, + "learning_rate": 7.592985778766881e-07, + "loss": 0.1259, + "step": 40985 + }, + { + "epoch": 0.9607763849575415, + "grad_norm": 2.2658746242523193, + "learning_rate": 7.583928301237664e-07, + "loss": 0.0915, + "step": 40986 + }, + { + "epoch": 0.9607998265323464, + "grad_norm": 0.6074616312980652, + "learning_rate": 7.57487620858599e-07, + "loss": 0.0792, + "step": 40987 + }, + { + "epoch": 0.9608232681071515, + "grad_norm": 2.1548643112182617, + "learning_rate": 7.565829500860932e-07, + "loss": 0.1146, + "step": 40988 + }, + { + "epoch": 0.9608467096819564, + "grad_norm": 3.2301950454711914, + "learning_rate": 7.556788178111563e-07, + "loss": 0.2246, + "step": 40989 + }, + { + "epoch": 0.9608701512567615, + "grad_norm": 2.4198451042175293, + "learning_rate": 7.547752240386952e-07, + "loss": 0.3059, + "step": 40990 + }, + { + "epoch": 0.9608935928315664, + "grad_norm": 3.2873551845550537, + "learning_rate": 7.538721687736173e-07, + "loss": 0.1724, + "step": 40991 + }, + { + "epoch": 0.9609170344063714, + "grad_norm": 2.691742420196533, + "learning_rate": 7.529696520208184e-07, + "loss": 0.1941, + "step": 40992 + }, + { + "epoch": 0.9609404759811764, + "grad_norm": 3.565913200378418, + "learning_rate": 7.520676737852061e-07, + "loss": 0.3721, + "step": 40993 + }, + { + "epoch": 0.9609639175559814, + "grad_norm": 3.104186773300171, + "learning_rate": 7.511662340716541e-07, + "loss": 0.278, + "step": 40994 + }, + { + "epoch": 0.9609873591307864, + "grad_norm": 0.6200962066650391, + "learning_rate": 7.502653328850695e-07, + "loss": 0.0568, + "step": 40995 + }, + { + "epoch": 0.9610108007055914, + "grad_norm": 2.5977723598480225, + "learning_rate": 7.493649702303373e-07, + "loss": 0.2567, + "step": 40996 + }, + { + "epoch": 0.9610342422803964, + "grad_norm": 0.5832075476646423, + "learning_rate": 7.484651461123315e-07, + "loss": 0.0276, + "step": 40997 + }, + { + "epoch": 0.9610576838552014, + "grad_norm": 2.850867986679077, + "learning_rate": 7.475658605359592e-07, + "loss": 0.1754, + "step": 40998 + }, + { + "epoch": 0.9610811254300063, + "grad_norm": 2.0795397758483887, + "learning_rate": 7.466671135060721e-07, + "loss": 0.084, + "step": 40999 + }, + { + "epoch": 0.9611045670048114, + "grad_norm": 2.0997776985168457, + "learning_rate": 7.457689050275662e-07, + "loss": 0.1301, + "step": 41000 + }, + { + "epoch": 0.9611280085796163, + "grad_norm": 4.4699506759643555, + "learning_rate": 7.448712351053044e-07, + "loss": 0.5113, + "step": 41001 + }, + { + "epoch": 0.9611514501544214, + "grad_norm": 1.0486100912094116, + "learning_rate": 7.439741037441717e-07, + "loss": 0.0742, + "step": 41002 + }, + { + "epoch": 0.9611748917292263, + "grad_norm": 2.638413190841675, + "learning_rate": 7.430775109490196e-07, + "loss": 0.2662, + "step": 41003 + }, + { + "epoch": 0.9611983333040314, + "grad_norm": 3.4021520614624023, + "learning_rate": 7.421814567247221e-07, + "loss": 0.2321, + "step": 41004 + }, + { + "epoch": 0.9612217748788363, + "grad_norm": 3.1895627975463867, + "learning_rate": 7.412859410761308e-07, + "loss": 0.1773, + "step": 41005 + }, + { + "epoch": 0.9612452164536414, + "grad_norm": 3.876915454864502, + "learning_rate": 7.403909640081197e-07, + "loss": 0.3008, + "step": 41006 + }, + { + "epoch": 0.9612686580284463, + "grad_norm": 2.254938840866089, + "learning_rate": 7.394965255255404e-07, + "loss": 0.1118, + "step": 41007 + }, + { + "epoch": 0.9612920996032513, + "grad_norm": 1.3817793130874634, + "learning_rate": 7.386026256332446e-07, + "loss": 1.2214, + "step": 41008 + }, + { + "epoch": 0.9613155411780563, + "grad_norm": 2.375098466873169, + "learning_rate": 7.377092643360728e-07, + "loss": 0.1547, + "step": 41009 + }, + { + "epoch": 0.9613389827528613, + "grad_norm": 4.6697211265563965, + "learning_rate": 7.36816441638899e-07, + "loss": 0.4557, + "step": 41010 + }, + { + "epoch": 0.9613624243276664, + "grad_norm": 1.9507777690887451, + "learning_rate": 7.359241575465303e-07, + "loss": 0.159, + "step": 41011 + }, + { + "epoch": 0.9613858659024713, + "grad_norm": 3.183601140975952, + "learning_rate": 7.350324120638408e-07, + "loss": 0.2797, + "step": 41012 + }, + { + "epoch": 0.9614093074772764, + "grad_norm": 1.684314489364624, + "learning_rate": 7.341412051956487e-07, + "loss": 0.0952, + "step": 41013 + }, + { + "epoch": 0.9614327490520813, + "grad_norm": 2.9947509765625, + "learning_rate": 7.332505369468057e-07, + "loss": 0.1922, + "step": 41014 + }, + { + "epoch": 0.9614561906268864, + "grad_norm": 5.3087286949157715, + "learning_rate": 7.323604073221301e-07, + "loss": 0.2068, + "step": 41015 + }, + { + "epoch": 0.9614796322016913, + "grad_norm": 0.3084317743778229, + "learning_rate": 7.314708163264627e-07, + "loss": 0.0157, + "step": 41016 + }, + { + "epoch": 0.9615030737764964, + "grad_norm": 0.39893776178359985, + "learning_rate": 7.305817639646217e-07, + "loss": 0.0416, + "step": 41017 + }, + { + "epoch": 0.9615265153513013, + "grad_norm": 1.8695428371429443, + "learning_rate": 7.296932502414366e-07, + "loss": 0.0878, + "step": 41018 + }, + { + "epoch": 0.9615499569261063, + "grad_norm": 3.7283542156219482, + "learning_rate": 7.288052751617258e-07, + "loss": 0.2687, + "step": 41019 + }, + { + "epoch": 0.9615733985009113, + "grad_norm": 2.2203593254089355, + "learning_rate": 7.279178387303188e-07, + "loss": 0.1808, + "step": 41020 + }, + { + "epoch": 0.9615968400757163, + "grad_norm": 2.268216371536255, + "learning_rate": 7.270309409520116e-07, + "loss": 0.1658, + "step": 41021 + }, + { + "epoch": 0.9616202816505213, + "grad_norm": 2.8516881465911865, + "learning_rate": 7.261445818316226e-07, + "loss": 0.1284, + "step": 41022 + }, + { + "epoch": 0.9616437232253263, + "grad_norm": 1.9918603897094727, + "learning_rate": 7.252587613739703e-07, + "loss": 0.1173, + "step": 41023 + }, + { + "epoch": 0.9616671648001313, + "grad_norm": 1.9319857358932495, + "learning_rate": 7.243734795838509e-07, + "loss": 0.1179, + "step": 41024 + }, + { + "epoch": 0.9616906063749363, + "grad_norm": 4.167109966278076, + "learning_rate": 7.234887364660714e-07, + "loss": 0.3601, + "step": 41025 + }, + { + "epoch": 0.9617140479497412, + "grad_norm": 2.7627224922180176, + "learning_rate": 7.226045320254394e-07, + "loss": 0.1919, + "step": 41026 + }, + { + "epoch": 0.9617374895245463, + "grad_norm": 0.8404424786567688, + "learning_rate": 7.217208662667507e-07, + "loss": 0.1153, + "step": 41027 + }, + { + "epoch": 0.9617609310993512, + "grad_norm": 1.6829707622528076, + "learning_rate": 7.208377391947907e-07, + "loss": 0.1048, + "step": 41028 + }, + { + "epoch": 0.9617843726741563, + "grad_norm": 3.148470163345337, + "learning_rate": 7.199551508143443e-07, + "loss": 0.0689, + "step": 41029 + }, + { + "epoch": 0.9618078142489612, + "grad_norm": 1.4445726871490479, + "learning_rate": 7.190731011302299e-07, + "loss": 0.0732, + "step": 41030 + }, + { + "epoch": 0.9618312558237663, + "grad_norm": 0.4674065411090851, + "learning_rate": 7.181915901472102e-07, + "loss": 0.0408, + "step": 41031 + }, + { + "epoch": 0.9618546973985712, + "grad_norm": 1.7915452718734741, + "learning_rate": 7.173106178700706e-07, + "loss": 0.8576, + "step": 41032 + }, + { + "epoch": 0.9618781389733763, + "grad_norm": 4.7281951904296875, + "learning_rate": 7.164301843036069e-07, + "loss": 0.3613, + "step": 41033 + }, + { + "epoch": 0.9619015805481812, + "grad_norm": 3.0823187828063965, + "learning_rate": 7.155502894525823e-07, + "loss": 0.3528, + "step": 41034 + }, + { + "epoch": 0.9619250221229863, + "grad_norm": 1.7316218614578247, + "learning_rate": 7.146709333217594e-07, + "loss": 0.0991, + "step": 41035 + }, + { + "epoch": 0.9619484636977912, + "grad_norm": 2.8319501876831055, + "learning_rate": 7.137921159159233e-07, + "loss": 0.1344, + "step": 41036 + }, + { + "epoch": 0.9619719052725962, + "grad_norm": 3.7668585777282715, + "learning_rate": 7.129138372398592e-07, + "loss": 0.1988, + "step": 41037 + }, + { + "epoch": 0.9619953468474012, + "grad_norm": 1.5893523693084717, + "learning_rate": 7.120360972983075e-07, + "loss": 0.0793, + "step": 41038 + }, + { + "epoch": 0.9620187884222062, + "grad_norm": 0.8491044640541077, + "learning_rate": 7.111588960960425e-07, + "loss": 0.0604, + "step": 41039 + }, + { + "epoch": 0.9620422299970112, + "grad_norm": 0.6060433983802795, + "learning_rate": 7.102822336378267e-07, + "loss": 0.0483, + "step": 41040 + }, + { + "epoch": 0.9620656715718162, + "grad_norm": 1.3531782627105713, + "learning_rate": 7.09406109928401e-07, + "loss": 0.1261, + "step": 41041 + }, + { + "epoch": 0.9620891131466212, + "grad_norm": 3.092329740524292, + "learning_rate": 7.085305249725282e-07, + "loss": 0.1674, + "step": 41042 + }, + { + "epoch": 0.9621125547214262, + "grad_norm": 3.266401529312134, + "learning_rate": 7.076554787749712e-07, + "loss": 0.2502, + "step": 41043 + }, + { + "epoch": 0.9621359962962311, + "grad_norm": 1.9926958084106445, + "learning_rate": 7.067809713404594e-07, + "loss": 0.0952, + "step": 41044 + }, + { + "epoch": 0.9621594378710362, + "grad_norm": 0.815243124961853, + "learning_rate": 7.059070026737557e-07, + "loss": 0.0749, + "step": 41045 + }, + { + "epoch": 0.9621828794458411, + "grad_norm": 4.749818325042725, + "learning_rate": 7.050335727795787e-07, + "loss": 0.39, + "step": 41046 + }, + { + "epoch": 0.9622063210206462, + "grad_norm": 1.7423604726791382, + "learning_rate": 7.041606816626911e-07, + "loss": 0.0737, + "step": 41047 + }, + { + "epoch": 0.9622297625954511, + "grad_norm": 1.7186070680618286, + "learning_rate": 7.032883293278114e-07, + "loss": 1.1708, + "step": 41048 + }, + { + "epoch": 0.9622532041702562, + "grad_norm": 1.646560788154602, + "learning_rate": 7.024165157796803e-07, + "loss": 1.624, + "step": 41049 + }, + { + "epoch": 0.9622766457450611, + "grad_norm": 1.3110333681106567, + "learning_rate": 7.015452410230383e-07, + "loss": 0.104, + "step": 41050 + }, + { + "epoch": 0.9623000873198662, + "grad_norm": 1.637284755706787, + "learning_rate": 7.00674505062604e-07, + "loss": 0.1218, + "step": 41051 + }, + { + "epoch": 0.9623235288946711, + "grad_norm": 3.2679972648620605, + "learning_rate": 6.998043079030958e-07, + "loss": 0.1532, + "step": 41052 + }, + { + "epoch": 0.9623469704694761, + "grad_norm": 3.6246485710144043, + "learning_rate": 6.989346495492544e-07, + "loss": 0.3124, + "step": 41053 + }, + { + "epoch": 0.9623704120442811, + "grad_norm": 2.0894625186920166, + "learning_rate": 6.980655300057648e-07, + "loss": 0.1257, + "step": 41054 + }, + { + "epoch": 0.9623938536190861, + "grad_norm": 1.1016604900360107, + "learning_rate": 6.971969492773789e-07, + "loss": 0.03, + "step": 41055 + }, + { + "epoch": 0.9624172951938911, + "grad_norm": 0.43279197812080383, + "learning_rate": 6.963289073687929e-07, + "loss": 0.0284, + "step": 41056 + }, + { + "epoch": 0.9624407367686961, + "grad_norm": 0.4248036742210388, + "learning_rate": 6.95461404284714e-07, + "loss": 0.0146, + "step": 41057 + }, + { + "epoch": 0.9624641783435011, + "grad_norm": 2.720679998397827, + "learning_rate": 6.945944400298609e-07, + "loss": 0.1711, + "step": 41058 + }, + { + "epoch": 0.9624876199183061, + "grad_norm": 0.6773041486740112, + "learning_rate": 6.937280146089186e-07, + "loss": 0.0438, + "step": 41059 + }, + { + "epoch": 0.962511061493111, + "grad_norm": 3.1168384552001953, + "learning_rate": 6.928621280266056e-07, + "loss": 0.4351, + "step": 41060 + }, + { + "epoch": 0.9625345030679161, + "grad_norm": 2.84904146194458, + "learning_rate": 6.919967802876182e-07, + "loss": 0.1355, + "step": 41061 + }, + { + "epoch": 0.9625579446427212, + "grad_norm": 0.9302638173103333, + "learning_rate": 6.911319713966414e-07, + "loss": 0.1045, + "step": 41062 + }, + { + "epoch": 0.9625813862175261, + "grad_norm": 5.0449395179748535, + "learning_rate": 6.902677013583825e-07, + "loss": 0.4405, + "step": 41063 + }, + { + "epoch": 0.9626048277923311, + "grad_norm": 0.9232614636421204, + "learning_rate": 6.894039701775268e-07, + "loss": 0.0729, + "step": 41064 + }, + { + "epoch": 0.9626282693671361, + "grad_norm": 5.355392932891846, + "learning_rate": 6.885407778587482e-07, + "loss": 0.3289, + "step": 41065 + }, + { + "epoch": 0.9626517109419411, + "grad_norm": 3.332918167114258, + "learning_rate": 6.876781244067432e-07, + "loss": 0.2629, + "step": 41066 + }, + { + "epoch": 0.9626751525167461, + "grad_norm": 2.353607654571533, + "learning_rate": 6.868160098261855e-07, + "loss": 0.1913, + "step": 41067 + }, + { + "epoch": 0.9626985940915511, + "grad_norm": 2.4835617542266846, + "learning_rate": 6.859544341217605e-07, + "loss": 0.1116, + "step": 41068 + }, + { + "epoch": 0.9627220356663561, + "grad_norm": 1.615390658378601, + "learning_rate": 6.850933972981421e-07, + "loss": 0.1694, + "step": 41069 + }, + { + "epoch": 0.9627454772411611, + "grad_norm": 3.718806505203247, + "learning_rate": 6.842328993599933e-07, + "loss": 0.4167, + "step": 41070 + }, + { + "epoch": 0.962768918815966, + "grad_norm": 2.801445484161377, + "learning_rate": 6.833729403119993e-07, + "loss": 0.2166, + "step": 41071 + }, + { + "epoch": 0.9627923603907711, + "grad_norm": 2.570516586303711, + "learning_rate": 6.825135201588117e-07, + "loss": 0.2934, + "step": 41072 + }, + { + "epoch": 0.962815801965576, + "grad_norm": 2.4793128967285156, + "learning_rate": 6.816546389050826e-07, + "loss": 0.1445, + "step": 41073 + }, + { + "epoch": 0.9628392435403811, + "grad_norm": 3.307640314102173, + "learning_rate": 6.80796296555497e-07, + "loss": 0.2542, + "step": 41074 + }, + { + "epoch": 0.962862685115186, + "grad_norm": 4.382322788238525, + "learning_rate": 6.799384931147179e-07, + "loss": 0.2331, + "step": 41075 + }, + { + "epoch": 0.9628861266899911, + "grad_norm": 4.42119836807251, + "learning_rate": 6.790812285873749e-07, + "loss": 0.3706, + "step": 41076 + }, + { + "epoch": 0.962909568264796, + "grad_norm": 3.2957842350006104, + "learning_rate": 6.782245029781309e-07, + "loss": 0.1867, + "step": 41077 + }, + { + "epoch": 0.9629330098396011, + "grad_norm": 0.9320381283760071, + "learning_rate": 6.773683162916267e-07, + "loss": 0.0647, + "step": 41078 + }, + { + "epoch": 0.962956451414406, + "grad_norm": 4.462352752685547, + "learning_rate": 6.765126685325252e-07, + "loss": 0.2011, + "step": 41079 + }, + { + "epoch": 0.962979892989211, + "grad_norm": 0.6384648680686951, + "learning_rate": 6.756575597054449e-07, + "loss": 0.0373, + "step": 41080 + }, + { + "epoch": 0.963003334564016, + "grad_norm": 0.40897777676582336, + "learning_rate": 6.748029898150488e-07, + "loss": 0.0233, + "step": 41081 + }, + { + "epoch": 0.963026776138821, + "grad_norm": 2.7769529819488525, + "learning_rate": 6.739489588659664e-07, + "loss": 0.2619, + "step": 41082 + }, + { + "epoch": 0.963050217713626, + "grad_norm": 1.8528844118118286, + "learning_rate": 6.730954668628275e-07, + "loss": 0.1244, + "step": 41083 + }, + { + "epoch": 0.963073659288431, + "grad_norm": 3.356685161590576, + "learning_rate": 6.722425138102728e-07, + "loss": 0.2966, + "step": 41084 + }, + { + "epoch": 0.963097100863236, + "grad_norm": 2.864180564880371, + "learning_rate": 6.713900997129096e-07, + "loss": 0.2627, + "step": 41085 + }, + { + "epoch": 0.963120542438041, + "grad_norm": 1.5696197748184204, + "learning_rate": 6.705382245753788e-07, + "loss": 0.0783, + "step": 41086 + }, + { + "epoch": 0.963143984012846, + "grad_norm": 2.63252592086792, + "learning_rate": 6.696868884022988e-07, + "loss": 0.2016, + "step": 41087 + }, + { + "epoch": 0.963167425587651, + "grad_norm": 4.031850337982178, + "learning_rate": 6.688360911982994e-07, + "loss": 0.4358, + "step": 41088 + }, + { + "epoch": 0.9631908671624559, + "grad_norm": 4.087768077850342, + "learning_rate": 6.679858329679877e-07, + "loss": 0.173, + "step": 41089 + }, + { + "epoch": 0.963214308737261, + "grad_norm": 1.345868468284607, + "learning_rate": 6.671361137159715e-07, + "loss": 0.0634, + "step": 41090 + }, + { + "epoch": 0.9632377503120659, + "grad_norm": 2.603644609451294, + "learning_rate": 6.662869334468802e-07, + "loss": 0.2199, + "step": 41091 + }, + { + "epoch": 0.963261191886871, + "grad_norm": 0.7115285992622375, + "learning_rate": 6.65438292165288e-07, + "loss": 0.0608, + "step": 41092 + }, + { + "epoch": 0.9632846334616759, + "grad_norm": 2.6432645320892334, + "learning_rate": 6.645901898758355e-07, + "loss": 0.1281, + "step": 41093 + }, + { + "epoch": 0.963308075036481, + "grad_norm": 2.041302442550659, + "learning_rate": 6.637426265831081e-07, + "loss": 1.3286, + "step": 41094 + }, + { + "epoch": 0.9633315166112859, + "grad_norm": 2.9843437671661377, + "learning_rate": 6.628956022917132e-07, + "loss": 0.2103, + "step": 41095 + }, + { + "epoch": 0.963354958186091, + "grad_norm": 3.225292444229126, + "learning_rate": 6.620491170062249e-07, + "loss": 0.3338, + "step": 41096 + }, + { + "epoch": 0.9633783997608959, + "grad_norm": 2.339482307434082, + "learning_rate": 6.612031707312616e-07, + "loss": 0.1361, + "step": 41097 + }, + { + "epoch": 0.963401841335701, + "grad_norm": 1.3728920221328735, + "learning_rate": 6.603577634713975e-07, + "loss": 0.0742, + "step": 41098 + }, + { + "epoch": 0.9634252829105059, + "grad_norm": 1.822253704071045, + "learning_rate": 6.595128952312291e-07, + "loss": 0.1327, + "step": 41099 + }, + { + "epoch": 0.9634487244853109, + "grad_norm": 3.6377129554748535, + "learning_rate": 6.586685660153302e-07, + "loss": 0.2308, + "step": 41100 + }, + { + "epoch": 0.9634721660601159, + "grad_norm": 4.6547393798828125, + "learning_rate": 6.578247758282974e-07, + "loss": 1.5379, + "step": 41101 + }, + { + "epoch": 0.9634956076349209, + "grad_norm": 1.4246357679367065, + "learning_rate": 6.569815246746935e-07, + "loss": 0.0725, + "step": 41102 + }, + { + "epoch": 0.9635190492097259, + "grad_norm": 3.4401063919067383, + "learning_rate": 6.561388125591039e-07, + "loss": 0.3801, + "step": 41103 + }, + { + "epoch": 0.9635424907845309, + "grad_norm": 2.531231164932251, + "learning_rate": 6.552966394861027e-07, + "loss": 0.2019, + "step": 41104 + }, + { + "epoch": 0.9635659323593359, + "grad_norm": 2.2909488677978516, + "learning_rate": 6.544550054602527e-07, + "loss": 0.1221, + "step": 41105 + }, + { + "epoch": 0.9635893739341409, + "grad_norm": 2.444979190826416, + "learning_rate": 6.536139104861172e-07, + "loss": 0.152, + "step": 41106 + }, + { + "epoch": 0.9636128155089458, + "grad_norm": 2.6315810680389404, + "learning_rate": 6.527733545682812e-07, + "loss": 0.2105, + "step": 41107 + }, + { + "epoch": 0.9636362570837509, + "grad_norm": 1.824061393737793, + "learning_rate": 6.519333377112746e-07, + "loss": 0.107, + "step": 41108 + }, + { + "epoch": 0.9636596986585558, + "grad_norm": 1.520772933959961, + "learning_rate": 6.510938599196714e-07, + "loss": 1.0851, + "step": 41109 + }, + { + "epoch": 0.9636831402333609, + "grad_norm": 1.6202032566070557, + "learning_rate": 6.502549211980346e-07, + "loss": 0.1877, + "step": 41110 + }, + { + "epoch": 0.9637065818081658, + "grad_norm": 0.23316222429275513, + "learning_rate": 6.494165215509051e-07, + "loss": 0.0072, + "step": 41111 + }, + { + "epoch": 0.9637300233829709, + "grad_norm": 3.519365072250366, + "learning_rate": 6.485786609828348e-07, + "loss": 0.3219, + "step": 41112 + }, + { + "epoch": 0.9637534649577759, + "grad_norm": 0.8177173733711243, + "learning_rate": 6.477413394983755e-07, + "loss": 0.074, + "step": 41113 + }, + { + "epoch": 0.9637769065325809, + "grad_norm": 1.4814072847366333, + "learning_rate": 6.469045571020571e-07, + "loss": 0.0976, + "step": 41114 + }, + { + "epoch": 0.9638003481073859, + "grad_norm": 2.5114054679870605, + "learning_rate": 6.460683137984313e-07, + "loss": 0.1508, + "step": 41115 + }, + { + "epoch": 0.9638237896821908, + "grad_norm": 1.7764054536819458, + "learning_rate": 6.452326095920391e-07, + "loss": 0.097, + "step": 41116 + }, + { + "epoch": 0.9638472312569959, + "grad_norm": 0.4846791923046112, + "learning_rate": 6.44397444487399e-07, + "loss": 0.0347, + "step": 41117 + }, + { + "epoch": 0.9638706728318008, + "grad_norm": 3.616974353790283, + "learning_rate": 6.435628184890518e-07, + "loss": 0.2236, + "step": 41118 + }, + { + "epoch": 0.9638941144066059, + "grad_norm": 4.176868438720703, + "learning_rate": 6.427287316015274e-07, + "loss": 0.4123, + "step": 41119 + }, + { + "epoch": 0.9639175559814108, + "grad_norm": 1.6426982879638672, + "learning_rate": 6.418951838293552e-07, + "loss": 0.082, + "step": 41120 + }, + { + "epoch": 0.9639409975562159, + "grad_norm": 3.770052194595337, + "learning_rate": 6.41062175177054e-07, + "loss": 0.3031, + "step": 41121 + }, + { + "epoch": 0.9639644391310208, + "grad_norm": 0.6899734735488892, + "learning_rate": 6.402297056491313e-07, + "loss": 0.0656, + "step": 41122 + }, + { + "epoch": 0.9639878807058259, + "grad_norm": 1.0162580013275146, + "learning_rate": 6.393977752501279e-07, + "loss": 0.0434, + "step": 41123 + }, + { + "epoch": 0.9640113222806308, + "grad_norm": 3.1563994884490967, + "learning_rate": 6.385663839845402e-07, + "loss": 0.2029, + "step": 41124 + }, + { + "epoch": 0.9640347638554359, + "grad_norm": 0.4861656129360199, + "learning_rate": 6.377355318568867e-07, + "loss": 0.0346, + "step": 41125 + }, + { + "epoch": 0.9640582054302408, + "grad_norm": 1.5522570610046387, + "learning_rate": 6.369052188716751e-07, + "loss": 0.1328, + "step": 41126 + }, + { + "epoch": 0.9640816470050458, + "grad_norm": 3.0488736629486084, + "learning_rate": 6.360754450334128e-07, + "loss": 0.2017, + "step": 41127 + }, + { + "epoch": 0.9641050885798508, + "grad_norm": 2.9121294021606445, + "learning_rate": 6.352462103465961e-07, + "loss": 0.1777, + "step": 41128 + }, + { + "epoch": 0.9641285301546558, + "grad_norm": 0.43944674730300903, + "learning_rate": 6.344175148157328e-07, + "loss": 0.0291, + "step": 41129 + }, + { + "epoch": 0.9641519717294608, + "grad_norm": 3.204641580581665, + "learning_rate": 6.335893584452968e-07, + "loss": 0.3671, + "step": 41130 + }, + { + "epoch": 0.9641754133042658, + "grad_norm": 0.5438840985298157, + "learning_rate": 6.327617412398179e-07, + "loss": 0.0308, + "step": 41131 + }, + { + "epoch": 0.9641988548790708, + "grad_norm": 4.2266340255737305, + "learning_rate": 6.319346632037593e-07, + "loss": 0.3633, + "step": 41132 + }, + { + "epoch": 0.9642222964538758, + "grad_norm": 2.150686740875244, + "learning_rate": 6.311081243416173e-07, + "loss": 0.1706, + "step": 41133 + }, + { + "epoch": 0.9642457380286807, + "grad_norm": 2.8144280910491943, + "learning_rate": 6.302821246578771e-07, + "loss": 0.1723, + "step": 41134 + }, + { + "epoch": 0.9642691796034858, + "grad_norm": 3.354369878768921, + "learning_rate": 6.294566641570132e-07, + "loss": 0.2513, + "step": 41135 + }, + { + "epoch": 0.9642926211782907, + "grad_norm": 2.133124828338623, + "learning_rate": 6.286317428435217e-07, + "loss": 0.0718, + "step": 41136 + }, + { + "epoch": 0.9643160627530958, + "grad_norm": 0.8338159322738647, + "learning_rate": 6.27807360721866e-07, + "loss": 0.0663, + "step": 41137 + }, + { + "epoch": 0.9643395043279007, + "grad_norm": 3.540386915206909, + "learning_rate": 6.269835177965199e-07, + "loss": 0.3779, + "step": 41138 + }, + { + "epoch": 0.9643629459027058, + "grad_norm": 2.3731844425201416, + "learning_rate": 6.26160214071958e-07, + "loss": 0.2599, + "step": 41139 + }, + { + "epoch": 0.9643863874775107, + "grad_norm": 0.8410900235176086, + "learning_rate": 6.253374495526431e-07, + "loss": 0.0754, + "step": 41140 + }, + { + "epoch": 0.9644098290523158, + "grad_norm": 3.302842617034912, + "learning_rate": 6.245152242430385e-07, + "loss": 0.2914, + "step": 41141 + }, + { + "epoch": 0.9644332706271207, + "grad_norm": 2.0458340644836426, + "learning_rate": 6.236935381476184e-07, + "loss": 0.0908, + "step": 41142 + }, + { + "epoch": 0.9644567122019257, + "grad_norm": 0.4198777377605438, + "learning_rate": 6.228723912708234e-07, + "loss": 0.0265, + "step": 41143 + }, + { + "epoch": 0.9644801537767307, + "grad_norm": 3.0707428455352783, + "learning_rate": 6.220517836171169e-07, + "loss": 0.2375, + "step": 41144 + }, + { + "epoch": 0.9645035953515357, + "grad_norm": 1.7827054262161255, + "learning_rate": 6.212317151909508e-07, + "loss": 0.1606, + "step": 41145 + }, + { + "epoch": 0.9645270369263407, + "grad_norm": 2.6654701232910156, + "learning_rate": 6.204121859967771e-07, + "loss": 0.1515, + "step": 41146 + }, + { + "epoch": 0.9645504785011457, + "grad_norm": 3.5337953567504883, + "learning_rate": 6.195931960390477e-07, + "loss": 0.285, + "step": 41147 + }, + { + "epoch": 0.9645739200759507, + "grad_norm": 2.2053983211517334, + "learning_rate": 6.187747453221926e-07, + "loss": 0.1447, + "step": 41148 + }, + { + "epoch": 0.9645973616507557, + "grad_norm": 2.9791646003723145, + "learning_rate": 6.179568338506636e-07, + "loss": 0.2731, + "step": 41149 + }, + { + "epoch": 0.9646208032255607, + "grad_norm": 2.432630777359009, + "learning_rate": 6.171394616288906e-07, + "loss": 1.3272, + "step": 41150 + }, + { + "epoch": 0.9646442448003657, + "grad_norm": 2.0198967456817627, + "learning_rate": 6.163226286613144e-07, + "loss": 1.4608, + "step": 41151 + }, + { + "epoch": 0.9646676863751706, + "grad_norm": 4.393893241882324, + "learning_rate": 6.155063349523649e-07, + "loss": 0.4927, + "step": 41152 + }, + { + "epoch": 0.9646911279499757, + "grad_norm": 2.7973856925964355, + "learning_rate": 6.146905805064829e-07, + "loss": 0.356, + "step": 41153 + }, + { + "epoch": 0.9647145695247806, + "grad_norm": 1.782146692276001, + "learning_rate": 6.13875365328076e-07, + "loss": 0.1819, + "step": 41154 + }, + { + "epoch": 0.9647380110995857, + "grad_norm": 2.7370502948760986, + "learning_rate": 6.13060689421574e-07, + "loss": 0.2143, + "step": 41155 + }, + { + "epoch": 0.9647614526743906, + "grad_norm": 1.716163992881775, + "learning_rate": 6.122465527913956e-07, + "loss": 0.0953, + "step": 41156 + }, + { + "epoch": 0.9647848942491957, + "grad_norm": 1.9959214925765991, + "learning_rate": 6.114329554419707e-07, + "loss": 0.1667, + "step": 41157 + }, + { + "epoch": 0.9648083358240006, + "grad_norm": 1.0663666725158691, + "learning_rate": 6.106198973776956e-07, + "loss": 0.077, + "step": 41158 + }, + { + "epoch": 0.9648317773988057, + "grad_norm": 3.410820960998535, + "learning_rate": 6.09807378603e-07, + "loss": 0.2464, + "step": 41159 + }, + { + "epoch": 0.9648552189736106, + "grad_norm": 2.6507413387298584, + "learning_rate": 6.089953991222919e-07, + "loss": 0.2071, + "step": 41160 + }, + { + "epoch": 0.9648786605484156, + "grad_norm": 0.7078771591186523, + "learning_rate": 6.081839589399563e-07, + "loss": 0.0357, + "step": 41161 + }, + { + "epoch": 0.9649021021232206, + "grad_norm": 0.9868518114089966, + "learning_rate": 6.07373058060412e-07, + "loss": 0.0996, + "step": 41162 + }, + { + "epoch": 0.9649255436980256, + "grad_norm": 3.94102144241333, + "learning_rate": 6.065626964880556e-07, + "loss": 0.36, + "step": 41163 + }, + { + "epoch": 0.9649489852728307, + "grad_norm": 0.6140046715736389, + "learning_rate": 6.057528742272833e-07, + "loss": 0.0562, + "step": 41164 + }, + { + "epoch": 0.9649724268476356, + "grad_norm": 2.755307197570801, + "learning_rate": 6.04943591282503e-07, + "loss": 0.1916, + "step": 41165 + }, + { + "epoch": 0.9649958684224407, + "grad_norm": 6.624035835266113, + "learning_rate": 6.041348476580777e-07, + "loss": 0.3492, + "step": 41166 + }, + { + "epoch": 0.9650193099972456, + "grad_norm": 2.034050703048706, + "learning_rate": 6.033266433584261e-07, + "loss": 0.125, + "step": 41167 + }, + { + "epoch": 0.9650427515720507, + "grad_norm": 3.5161068439483643, + "learning_rate": 6.025189783879004e-07, + "loss": 0.1915, + "step": 41168 + }, + { + "epoch": 0.9650661931468556, + "grad_norm": 0.6126407384872437, + "learning_rate": 6.017118527509192e-07, + "loss": 0.0468, + "step": 41169 + }, + { + "epoch": 0.9650896347216606, + "grad_norm": 1.8438366651535034, + "learning_rate": 6.009052664518233e-07, + "loss": 0.0642, + "step": 41170 + }, + { + "epoch": 0.9651130762964656, + "grad_norm": 2.8031957149505615, + "learning_rate": 6.000992194950317e-07, + "loss": 0.1475, + "step": 41171 + }, + { + "epoch": 0.9651365178712706, + "grad_norm": 1.2149085998535156, + "learning_rate": 5.992937118848851e-07, + "loss": 0.1117, + "step": 41172 + }, + { + "epoch": 0.9651599594460756, + "grad_norm": 2.863542079925537, + "learning_rate": 5.984887436257691e-07, + "loss": 0.2402, + "step": 41173 + }, + { + "epoch": 0.9651834010208806, + "grad_norm": 0.3812270760536194, + "learning_rate": 5.976843147220579e-07, + "loss": 0.0212, + "step": 41174 + }, + { + "epoch": 0.9652068425956856, + "grad_norm": 0.7593246698379517, + "learning_rate": 5.968804251780924e-07, + "loss": 0.0635, + "step": 41175 + }, + { + "epoch": 0.9652302841704906, + "grad_norm": 2.262704372406006, + "learning_rate": 5.960770749982469e-07, + "loss": 0.1305, + "step": 41176 + }, + { + "epoch": 0.9652537257452956, + "grad_norm": 2.1209449768066406, + "learning_rate": 5.952742641868958e-07, + "loss": 0.1633, + "step": 41177 + }, + { + "epoch": 0.9652771673201006, + "grad_norm": 2.63674259185791, + "learning_rate": 5.944719927483689e-07, + "loss": 0.1139, + "step": 41178 + }, + { + "epoch": 0.9653006088949055, + "grad_norm": 2.0088858604431152, + "learning_rate": 5.936702606870403e-07, + "loss": 1.1871, + "step": 41179 + }, + { + "epoch": 0.9653240504697106, + "grad_norm": 3.2373478412628174, + "learning_rate": 5.928690680072512e-07, + "loss": 0.3248, + "step": 41180 + }, + { + "epoch": 0.9653474920445155, + "grad_norm": 2.301220178604126, + "learning_rate": 5.920684147133427e-07, + "loss": 0.1612, + "step": 41181 + }, + { + "epoch": 0.9653709336193206, + "grad_norm": 5.165053844451904, + "learning_rate": 5.912683008096665e-07, + "loss": 0.5191, + "step": 41182 + }, + { + "epoch": 0.9653943751941255, + "grad_norm": 1.3584860563278198, + "learning_rate": 5.904687263005748e-07, + "loss": 0.0446, + "step": 41183 + }, + { + "epoch": 0.9654178167689306, + "grad_norm": 3.1380531787872314, + "learning_rate": 5.896696911903865e-07, + "loss": 0.2308, + "step": 41184 + }, + { + "epoch": 0.9654412583437355, + "grad_norm": 2.256654739379883, + "learning_rate": 5.888711954834536e-07, + "loss": 0.1898, + "step": 41185 + }, + { + "epoch": 0.9654646999185406, + "grad_norm": 1.9367140531539917, + "learning_rate": 5.880732391840948e-07, + "loss": 0.1621, + "step": 41186 + }, + { + "epoch": 0.9654881414933455, + "grad_norm": 0.6196786761283875, + "learning_rate": 5.872758222966401e-07, + "loss": 0.0406, + "step": 41187 + }, + { + "epoch": 0.9655115830681505, + "grad_norm": 3.457540273666382, + "learning_rate": 5.864789448254304e-07, + "loss": 0.3263, + "step": 41188 + }, + { + "epoch": 0.9655350246429555, + "grad_norm": 1.0614690780639648, + "learning_rate": 5.856826067747734e-07, + "loss": 0.0549, + "step": 41189 + }, + { + "epoch": 0.9655584662177605, + "grad_norm": 0.9683782458305359, + "learning_rate": 5.848868081490099e-07, + "loss": 0.1153, + "step": 41190 + }, + { + "epoch": 0.9655819077925655, + "grad_norm": 2.500363826751709, + "learning_rate": 5.840915489524368e-07, + "loss": 0.2083, + "step": 41191 + }, + { + "epoch": 0.9656053493673705, + "grad_norm": 3.261495351791382, + "learning_rate": 5.832968291893837e-07, + "loss": 0.2082, + "step": 41192 + }, + { + "epoch": 0.9656287909421755, + "grad_norm": 3.1138267517089844, + "learning_rate": 5.825026488641471e-07, + "loss": 0.2147, + "step": 41193 + }, + { + "epoch": 0.9656522325169805, + "grad_norm": 1.2942941188812256, + "learning_rate": 5.817090079810572e-07, + "loss": 0.0988, + "step": 41194 + }, + { + "epoch": 0.9656756740917855, + "grad_norm": 0.46636298298835754, + "learning_rate": 5.809159065443992e-07, + "loss": 0.0527, + "step": 41195 + }, + { + "epoch": 0.9656991156665905, + "grad_norm": 1.9956125020980835, + "learning_rate": 5.80123344558492e-07, + "loss": 0.1564, + "step": 41196 + }, + { + "epoch": 0.9657225572413954, + "grad_norm": 1.9376697540283203, + "learning_rate": 5.793313220276319e-07, + "loss": 1.2829, + "step": 41197 + }, + { + "epoch": 0.9657459988162005, + "grad_norm": 3.5334420204162598, + "learning_rate": 5.785398389561159e-07, + "loss": 0.1389, + "step": 41198 + }, + { + "epoch": 0.9657694403910054, + "grad_norm": 2.0745747089385986, + "learning_rate": 5.777488953482401e-07, + "loss": 0.0975, + "step": 41199 + }, + { + "epoch": 0.9657928819658105, + "grad_norm": 2.8111751079559326, + "learning_rate": 5.769584912082904e-07, + "loss": 0.1621, + "step": 41200 + }, + { + "epoch": 0.9658163235406154, + "grad_norm": 1.4924311637878418, + "learning_rate": 5.76168626540563e-07, + "loss": 0.0967, + "step": 41201 + }, + { + "epoch": 0.9658397651154205, + "grad_norm": 1.0325653553009033, + "learning_rate": 5.753793013493436e-07, + "loss": 0.0712, + "step": 41202 + }, + { + "epoch": 0.9658632066902254, + "grad_norm": 3.673893451690674, + "learning_rate": 5.745905156389064e-07, + "loss": 0.246, + "step": 41203 + }, + { + "epoch": 0.9658866482650305, + "grad_norm": 1.8261609077453613, + "learning_rate": 5.73802269413537e-07, + "loss": 0.1679, + "step": 41204 + }, + { + "epoch": 0.9659100898398354, + "grad_norm": 1.4429724216461182, + "learning_rate": 5.730145626775096e-07, + "loss": 0.1629, + "step": 41205 + }, + { + "epoch": 0.9659335314146404, + "grad_norm": 2.3890278339385986, + "learning_rate": 5.722273954350988e-07, + "loss": 0.1792, + "step": 41206 + }, + { + "epoch": 0.9659569729894454, + "grad_norm": 2.0990402698516846, + "learning_rate": 5.714407676905898e-07, + "loss": 0.2029, + "step": 41207 + }, + { + "epoch": 0.9659804145642504, + "grad_norm": 4.209105014801025, + "learning_rate": 5.706546794482237e-07, + "loss": 0.4721, + "step": 41208 + }, + { + "epoch": 0.9660038561390554, + "grad_norm": 2.5188560485839844, + "learning_rate": 5.698691307122861e-07, + "loss": 0.1628, + "step": 41209 + }, + { + "epoch": 0.9660272977138604, + "grad_norm": 2.534498453140259, + "learning_rate": 5.690841214870401e-07, + "loss": 0.2299, + "step": 41210 + }, + { + "epoch": 0.9660507392886654, + "grad_norm": 1.6432139873504639, + "learning_rate": 5.68299651776738e-07, + "loss": 0.1135, + "step": 41211 + }, + { + "epoch": 0.9660741808634704, + "grad_norm": 2.052919387817383, + "learning_rate": 5.675157215856431e-07, + "loss": 0.1914, + "step": 41212 + }, + { + "epoch": 0.9660976224382753, + "grad_norm": 1.6965603828430176, + "learning_rate": 5.667323309179851e-07, + "loss": 0.1202, + "step": 41213 + }, + { + "epoch": 0.9661210640130804, + "grad_norm": 1.3489086627960205, + "learning_rate": 5.659494797780496e-07, + "loss": 0.0841, + "step": 41214 + }, + { + "epoch": 0.9661445055878854, + "grad_norm": 4.091005802154541, + "learning_rate": 5.651671681700666e-07, + "loss": 0.4657, + "step": 41215 + }, + { + "epoch": 0.9661679471626904, + "grad_norm": 1.5353862047195435, + "learning_rate": 5.64385396098277e-07, + "loss": 0.0869, + "step": 41216 + }, + { + "epoch": 0.9661913887374954, + "grad_norm": 2.993074655532837, + "learning_rate": 5.636041635669331e-07, + "loss": 0.255, + "step": 41217 + }, + { + "epoch": 0.9662148303123004, + "grad_norm": 2.484412431716919, + "learning_rate": 5.628234705802648e-07, + "loss": 0.193, + "step": 41218 + }, + { + "epoch": 0.9662382718871054, + "grad_norm": 2.2137813568115234, + "learning_rate": 5.620433171425133e-07, + "loss": 0.2285, + "step": 41219 + }, + { + "epoch": 0.9662617134619104, + "grad_norm": 1.206105351448059, + "learning_rate": 5.612637032579083e-07, + "loss": 0.1208, + "step": 41220 + }, + { + "epoch": 0.9662851550367154, + "grad_norm": 0.9866556525230408, + "learning_rate": 5.604846289306798e-07, + "loss": 0.0652, + "step": 41221 + }, + { + "epoch": 0.9663085966115204, + "grad_norm": 2.934237241744995, + "learning_rate": 5.597060941650689e-07, + "loss": 0.2857, + "step": 41222 + }, + { + "epoch": 0.9663320381863254, + "grad_norm": 1.9909226894378662, + "learning_rate": 5.589280989652834e-07, + "loss": 0.1329, + "step": 41223 + }, + { + "epoch": 0.9663554797611303, + "grad_norm": 1.3518043756484985, + "learning_rate": 5.581506433355533e-07, + "loss": 0.108, + "step": 41224 + }, + { + "epoch": 0.9663789213359354, + "grad_norm": 3.9598803520202637, + "learning_rate": 5.573737272800861e-07, + "loss": 0.2504, + "step": 41225 + }, + { + "epoch": 0.9664023629107403, + "grad_norm": 2.615203380584717, + "learning_rate": 5.565973508031009e-07, + "loss": 0.2654, + "step": 41226 + }, + { + "epoch": 0.9664258044855454, + "grad_norm": 2.546995162963867, + "learning_rate": 5.558215139088274e-07, + "loss": 0.12, + "step": 41227 + }, + { + "epoch": 0.9664492460603503, + "grad_norm": 3.689592123031616, + "learning_rate": 5.550462166014624e-07, + "loss": 0.1127, + "step": 41228 + }, + { + "epoch": 0.9664726876351554, + "grad_norm": 1.8230899572372437, + "learning_rate": 5.542714588852027e-07, + "loss": 0.0435, + "step": 41229 + }, + { + "epoch": 0.9664961292099603, + "grad_norm": 4.155491352081299, + "learning_rate": 5.534972407642669e-07, + "loss": 0.2561, + "step": 41230 + }, + { + "epoch": 0.9665195707847654, + "grad_norm": 1.3073121309280396, + "learning_rate": 5.527235622428517e-07, + "loss": 0.0334, + "step": 41231 + }, + { + "epoch": 0.9665430123595703, + "grad_norm": 1.8057562112808228, + "learning_rate": 5.519504233251538e-07, + "loss": 0.1349, + "step": 41232 + }, + { + "epoch": 0.9665664539343753, + "grad_norm": 2.2619736194610596, + "learning_rate": 5.511778240153698e-07, + "loss": 0.9598, + "step": 41233 + }, + { + "epoch": 0.9665898955091803, + "grad_norm": 1.0198051929473877, + "learning_rate": 5.504057643176963e-07, + "loss": 0.0981, + "step": 41234 + }, + { + "epoch": 0.9666133370839853, + "grad_norm": 2.5158770084381104, + "learning_rate": 5.496342442363078e-07, + "loss": 0.1879, + "step": 41235 + }, + { + "epoch": 0.9666367786587903, + "grad_norm": 2.484433174133301, + "learning_rate": 5.488632637754121e-07, + "loss": 0.178, + "step": 41236 + }, + { + "epoch": 0.9666602202335953, + "grad_norm": 3.7093605995178223, + "learning_rate": 5.480928229391724e-07, + "loss": 0.2956, + "step": 41237 + }, + { + "epoch": 0.9666836618084003, + "grad_norm": 1.3203985691070557, + "learning_rate": 5.473229217317855e-07, + "loss": 0.0785, + "step": 41238 + }, + { + "epoch": 0.9667071033832053, + "grad_norm": 2.5876622200012207, + "learning_rate": 5.465535601574035e-07, + "loss": 0.1985, + "step": 41239 + }, + { + "epoch": 0.9667305449580103, + "grad_norm": 1.2872754335403442, + "learning_rate": 5.457847382202341e-07, + "loss": 0.0908, + "step": 41240 + }, + { + "epoch": 0.9667539865328153, + "grad_norm": 1.4100263118743896, + "learning_rate": 5.450164559244298e-07, + "loss": 0.1648, + "step": 41241 + }, + { + "epoch": 0.9667774281076202, + "grad_norm": 1.3249566555023193, + "learning_rate": 5.442487132741537e-07, + "loss": 0.1579, + "step": 41242 + }, + { + "epoch": 0.9668008696824253, + "grad_norm": 0.820392370223999, + "learning_rate": 5.434815102735912e-07, + "loss": 0.0463, + "step": 41243 + }, + { + "epoch": 0.9668243112572302, + "grad_norm": 3.039642572402954, + "learning_rate": 5.427148469268839e-07, + "loss": 0.1855, + "step": 41244 + }, + { + "epoch": 0.9668477528320353, + "grad_norm": 1.4415615797042847, + "learning_rate": 5.419487232382059e-07, + "loss": 0.0752, + "step": 41245 + }, + { + "epoch": 0.9668711944068402, + "grad_norm": 2.9345614910125732, + "learning_rate": 5.411831392117206e-07, + "loss": 0.1143, + "step": 41246 + }, + { + "epoch": 0.9668946359816453, + "grad_norm": 0.8696581125259399, + "learning_rate": 5.404180948515692e-07, + "loss": 0.0455, + "step": 41247 + }, + { + "epoch": 0.9669180775564502, + "grad_norm": 7.8589630126953125, + "learning_rate": 5.396535901619038e-07, + "loss": 0.7577, + "step": 41248 + }, + { + "epoch": 0.9669415191312553, + "grad_norm": 3.432408094406128, + "learning_rate": 5.388896251468656e-07, + "loss": 0.2559, + "step": 41249 + }, + { + "epoch": 0.9669649607060602, + "grad_norm": 2.764463186264038, + "learning_rate": 5.38126199810618e-07, + "loss": 0.2473, + "step": 41250 + }, + { + "epoch": 0.9669884022808652, + "grad_norm": 0.9291497468948364, + "learning_rate": 5.373633141572909e-07, + "loss": 0.0429, + "step": 41251 + }, + { + "epoch": 0.9670118438556702, + "grad_norm": 2.4549548625946045, + "learning_rate": 5.366009681910256e-07, + "loss": 0.156, + "step": 41252 + }, + { + "epoch": 0.9670352854304752, + "grad_norm": 1.2790063619613647, + "learning_rate": 5.358391619159742e-07, + "loss": 0.0985, + "step": 41253 + }, + { + "epoch": 0.9670587270052802, + "grad_norm": 0.5205850005149841, + "learning_rate": 5.350778953362445e-07, + "loss": 0.0373, + "step": 41254 + }, + { + "epoch": 0.9670821685800852, + "grad_norm": 1.8692156076431274, + "learning_rate": 5.343171684559778e-07, + "loss": 0.11, + "step": 41255 + }, + { + "epoch": 0.9671056101548902, + "grad_norm": 1.8844635486602783, + "learning_rate": 5.33556981279304e-07, + "loss": 0.1611, + "step": 41256 + }, + { + "epoch": 0.9671290517296952, + "grad_norm": 0.6540302038192749, + "learning_rate": 5.32797333810342e-07, + "loss": 0.0719, + "step": 41257 + }, + { + "epoch": 0.9671524933045001, + "grad_norm": 1.5754836797714233, + "learning_rate": 5.32038226053222e-07, + "loss": 0.1507, + "step": 41258 + }, + { + "epoch": 0.9671759348793052, + "grad_norm": 1.1030843257904053, + "learning_rate": 5.312796580120627e-07, + "loss": 0.0827, + "step": 41259 + }, + { + "epoch": 0.9671993764541101, + "grad_norm": 4.731240272521973, + "learning_rate": 5.30521629690972e-07, + "loss": 0.0664, + "step": 41260 + }, + { + "epoch": 0.9672228180289152, + "grad_norm": 3.1576120853424072, + "learning_rate": 5.297641410940801e-07, + "loss": 0.2003, + "step": 41261 + }, + { + "epoch": 0.9672462596037201, + "grad_norm": 0.5696126222610474, + "learning_rate": 5.290071922254724e-07, + "loss": 0.0377, + "step": 41262 + }, + { + "epoch": 0.9672697011785252, + "grad_norm": 2.2300238609313965, + "learning_rate": 5.28250783089268e-07, + "loss": 0.1083, + "step": 41263 + }, + { + "epoch": 0.9672931427533301, + "grad_norm": 2.5048630237579346, + "learning_rate": 5.274949136895746e-07, + "loss": 1.5176, + "step": 41264 + }, + { + "epoch": 0.9673165843281352, + "grad_norm": 2.587568759918213, + "learning_rate": 5.267395840305001e-07, + "loss": 0.1039, + "step": 41265 + }, + { + "epoch": 0.9673400259029402, + "grad_norm": 2.445072889328003, + "learning_rate": 5.259847941161189e-07, + "loss": 1.8288, + "step": 41266 + }, + { + "epoch": 0.9673634674777452, + "grad_norm": 0.6007964611053467, + "learning_rate": 5.2523054395055e-07, + "loss": 0.0488, + "step": 41267 + }, + { + "epoch": 0.9673869090525502, + "grad_norm": 3.209864616394043, + "learning_rate": 5.24476833537868e-07, + "loss": 0.1966, + "step": 41268 + }, + { + "epoch": 0.9674103506273551, + "grad_norm": 1.3515950441360474, + "learning_rate": 5.237236628821807e-07, + "loss": 0.1488, + "step": 41269 + }, + { + "epoch": 0.9674337922021602, + "grad_norm": 2.7566165924072266, + "learning_rate": 5.229710319875624e-07, + "loss": 0.1468, + "step": 41270 + }, + { + "epoch": 0.9674572337769651, + "grad_norm": 1.3795398473739624, + "learning_rate": 5.222189408580991e-07, + "loss": 0.1331, + "step": 41271 + }, + { + "epoch": 0.9674806753517702, + "grad_norm": 2.736708879470825, + "learning_rate": 5.214673894978761e-07, + "loss": 0.313, + "step": 41272 + }, + { + "epoch": 0.9675041169265751, + "grad_norm": 1.5416518449783325, + "learning_rate": 5.20716377910968e-07, + "loss": 0.0976, + "step": 41273 + }, + { + "epoch": 0.9675275585013802, + "grad_norm": 2.2308130264282227, + "learning_rate": 5.199659061014494e-07, + "loss": 0.1839, + "step": 41274 + }, + { + "epoch": 0.9675510000761851, + "grad_norm": 3.80796480178833, + "learning_rate": 5.192159740733948e-07, + "loss": 0.2673, + "step": 41275 + }, + { + "epoch": 0.9675744416509902, + "grad_norm": 0.7158821821212769, + "learning_rate": 5.184665818308788e-07, + "loss": 0.0197, + "step": 41276 + }, + { + "epoch": 0.9675978832257951, + "grad_norm": 3.648961305618286, + "learning_rate": 5.177177293779423e-07, + "loss": 0.1008, + "step": 41277 + }, + { + "epoch": 0.9676213248006001, + "grad_norm": 5.1329240798950195, + "learning_rate": 5.169694167186711e-07, + "loss": 0.324, + "step": 41278 + }, + { + "epoch": 0.9676447663754051, + "grad_norm": 3.50463604927063, + "learning_rate": 5.162216438571288e-07, + "loss": 0.2163, + "step": 41279 + }, + { + "epoch": 0.9676682079502101, + "grad_norm": 2.428433656692505, + "learning_rate": 5.154744107973564e-07, + "loss": 0.1148, + "step": 41280 + }, + { + "epoch": 0.9676916495250151, + "grad_norm": 0.6321622729301453, + "learning_rate": 5.147277175434284e-07, + "loss": 0.0406, + "step": 41281 + }, + { + "epoch": 0.9677150910998201, + "grad_norm": 2.080615282058716, + "learning_rate": 5.139815640993751e-07, + "loss": 0.1099, + "step": 41282 + }, + { + "epoch": 0.9677385326746251, + "grad_norm": 1.0258761644363403, + "learning_rate": 5.132359504692596e-07, + "loss": 0.7774, + "step": 41283 + }, + { + "epoch": 0.9677619742494301, + "grad_norm": 1.3082325458526611, + "learning_rate": 5.124908766571235e-07, + "loss": 0.1646, + "step": 41284 + }, + { + "epoch": 0.967785415824235, + "grad_norm": 2.1405327320098877, + "learning_rate": 5.117463426670077e-07, + "loss": 0.1302, + "step": 41285 + }, + { + "epoch": 0.9678088573990401, + "grad_norm": 2.13983416557312, + "learning_rate": 5.110023485029536e-07, + "loss": 1.4193, + "step": 41286 + }, + { + "epoch": 0.967832298973845, + "grad_norm": 2.872292995452881, + "learning_rate": 5.102588941690023e-07, + "loss": 0.2553, + "step": 41287 + }, + { + "epoch": 0.9678557405486501, + "grad_norm": 3.393448829650879, + "learning_rate": 5.09515979669184e-07, + "loss": 0.2533, + "step": 41288 + }, + { + "epoch": 0.967879182123455, + "grad_norm": 0.44604361057281494, + "learning_rate": 5.087736050075176e-07, + "loss": 0.0302, + "step": 41289 + }, + { + "epoch": 0.9679026236982601, + "grad_norm": 1.5090607404708862, + "learning_rate": 5.080317701880555e-07, + "loss": 0.1303, + "step": 41290 + }, + { + "epoch": 0.967926065273065, + "grad_norm": 3.4190568923950195, + "learning_rate": 5.072904752148055e-07, + "loss": 0.1545, + "step": 41291 + }, + { + "epoch": 0.9679495068478701, + "grad_norm": 2.008329391479492, + "learning_rate": 5.065497200917868e-07, + "loss": 0.1004, + "step": 41292 + }, + { + "epoch": 0.967972948422675, + "grad_norm": 5.034021854400635, + "learning_rate": 5.058095048230405e-07, + "loss": 0.2978, + "step": 41293 + }, + { + "epoch": 0.9679963899974801, + "grad_norm": 2.405834197998047, + "learning_rate": 5.050698294125634e-07, + "loss": 0.1336, + "step": 41294 + }, + { + "epoch": 0.968019831572285, + "grad_norm": 2.5122082233428955, + "learning_rate": 5.043306938643744e-07, + "loss": 0.0877, + "step": 41295 + }, + { + "epoch": 0.96804327314709, + "grad_norm": 2.8408591747283936, + "learning_rate": 5.035920981824816e-07, + "loss": 0.2551, + "step": 41296 + }, + { + "epoch": 0.968066714721895, + "grad_norm": 3.209505796432495, + "learning_rate": 5.028540423709038e-07, + "loss": 0.1652, + "step": 41297 + }, + { + "epoch": 0.9680901562967, + "grad_norm": 1.9097555875778198, + "learning_rate": 5.02116526433638e-07, + "loss": 0.1672, + "step": 41298 + }, + { + "epoch": 0.968113597871505, + "grad_norm": 0.34188270568847656, + "learning_rate": 5.01379550374681e-07, + "loss": 0.0222, + "step": 41299 + }, + { + "epoch": 0.96813703944631, + "grad_norm": 3.8196845054626465, + "learning_rate": 5.006431141980406e-07, + "loss": 0.336, + "step": 41300 + }, + { + "epoch": 0.968160481021115, + "grad_norm": 1.7936362028121948, + "learning_rate": 4.999072179077024e-07, + "loss": 0.1055, + "step": 41301 + }, + { + "epoch": 0.96818392259592, + "grad_norm": 2.6343798637390137, + "learning_rate": 4.991718615076635e-07, + "loss": 1.598, + "step": 41302 + }, + { + "epoch": 0.968207364170725, + "grad_norm": 0.48462948203086853, + "learning_rate": 4.984370450019204e-07, + "loss": 0.051, + "step": 41303 + }, + { + "epoch": 0.96823080574553, + "grad_norm": 1.147558569908142, + "learning_rate": 4.97702768394448e-07, + "loss": 0.0742, + "step": 41304 + }, + { + "epoch": 0.9682542473203349, + "grad_norm": 0.8467426300048828, + "learning_rate": 4.969690316892428e-07, + "loss": 0.0619, + "step": 41305 + }, + { + "epoch": 0.96827768889514, + "grad_norm": 0.7775278091430664, + "learning_rate": 4.962358348902907e-07, + "loss": 0.068, + "step": 41306 + }, + { + "epoch": 0.9683011304699449, + "grad_norm": 1.6112099885940552, + "learning_rate": 4.955031780015551e-07, + "loss": 0.1256, + "step": 41307 + }, + { + "epoch": 0.96832457204475, + "grad_norm": 5.250941753387451, + "learning_rate": 4.947710610270107e-07, + "loss": 0.5127, + "step": 41308 + }, + { + "epoch": 0.9683480136195549, + "grad_norm": 2.1737964153289795, + "learning_rate": 4.94039483970643e-07, + "loss": 0.2297, + "step": 41309 + }, + { + "epoch": 0.96837145519436, + "grad_norm": 3.16483998298645, + "learning_rate": 4.933084468364157e-07, + "loss": 0.2196, + "step": 41310 + }, + { + "epoch": 0.9683948967691649, + "grad_norm": 1.7426881790161133, + "learning_rate": 4.925779496282923e-07, + "loss": 0.1435, + "step": 41311 + }, + { + "epoch": 0.96841833834397, + "grad_norm": 1.0341901779174805, + "learning_rate": 4.918479923502362e-07, + "loss": 0.0512, + "step": 41312 + }, + { + "epoch": 0.9684417799187749, + "grad_norm": 3.6651313304901123, + "learning_rate": 4.911185750062109e-07, + "loss": 0.2746, + "step": 41313 + }, + { + "epoch": 0.9684652214935799, + "grad_norm": 3.9349920749664307, + "learning_rate": 4.903896976001798e-07, + "loss": 0.2093, + "step": 41314 + }, + { + "epoch": 0.9684886630683849, + "grad_norm": 2.503528118133545, + "learning_rate": 4.896613601360845e-07, + "loss": 0.2392, + "step": 41315 + }, + { + "epoch": 0.9685121046431899, + "grad_norm": 2.3743393421173096, + "learning_rate": 4.889335626178882e-07, + "loss": 0.2037, + "step": 41316 + }, + { + "epoch": 0.968535546217995, + "grad_norm": 3.2828662395477295, + "learning_rate": 4.882063050495322e-07, + "loss": 0.0715, + "step": 41317 + }, + { + "epoch": 0.9685589877927999, + "grad_norm": 0.4001533091068268, + "learning_rate": 4.874795874349803e-07, + "loss": 0.0232, + "step": 41318 + }, + { + "epoch": 0.968582429367605, + "grad_norm": 1.5765811204910278, + "learning_rate": 4.867534097781512e-07, + "loss": 0.0788, + "step": 41319 + }, + { + "epoch": 0.9686058709424099, + "grad_norm": 4.55811882019043, + "learning_rate": 4.860277720829975e-07, + "loss": 0.4662, + "step": 41320 + }, + { + "epoch": 0.968629312517215, + "grad_norm": 3.196403980255127, + "learning_rate": 4.853026743534495e-07, + "loss": 0.2455, + "step": 41321 + }, + { + "epoch": 0.9686527540920199, + "grad_norm": 2.564103841781616, + "learning_rate": 4.845781165934594e-07, + "loss": 0.2153, + "step": 41322 + }, + { + "epoch": 0.968676195666825, + "grad_norm": 3.7454497814178467, + "learning_rate": 4.838540988069462e-07, + "loss": 0.11, + "step": 41323 + }, + { + "epoch": 0.9686996372416299, + "grad_norm": 1.3785573244094849, + "learning_rate": 4.831306209978293e-07, + "loss": 0.0922, + "step": 41324 + }, + { + "epoch": 0.9687230788164349, + "grad_norm": 0.8831751942634583, + "learning_rate": 4.824076831700496e-07, + "loss": 0.0491, + "step": 41325 + }, + { + "epoch": 0.9687465203912399, + "grad_norm": 1.3834283351898193, + "learning_rate": 4.816852853275266e-07, + "loss": 0.1396, + "step": 41326 + }, + { + "epoch": 0.9687699619660449, + "grad_norm": 3.4003193378448486, + "learning_rate": 4.809634274741681e-07, + "loss": 0.1559, + "step": 41327 + }, + { + "epoch": 0.9687934035408499, + "grad_norm": 0.7151206731796265, + "learning_rate": 4.802421096139042e-07, + "loss": 0.0717, + "step": 41328 + }, + { + "epoch": 0.9688168451156549, + "grad_norm": 5.168037414550781, + "learning_rate": 4.79521331750643e-07, + "loss": 0.3312, + "step": 41329 + }, + { + "epoch": 0.9688402866904599, + "grad_norm": 2.271056652069092, + "learning_rate": 4.788010938883036e-07, + "loss": 0.0553, + "step": 41330 + }, + { + "epoch": 0.9688637282652649, + "grad_norm": 0.7608160376548767, + "learning_rate": 4.780813960307829e-07, + "loss": 0.0558, + "step": 41331 + }, + { + "epoch": 0.9688871698400698, + "grad_norm": 1.6723105907440186, + "learning_rate": 4.773622381819997e-07, + "loss": 0.1617, + "step": 41332 + }, + { + "epoch": 0.9689106114148749, + "grad_norm": 0.9087970852851868, + "learning_rate": 4.766436203458402e-07, + "loss": 0.0585, + "step": 41333 + }, + { + "epoch": 0.9689340529896798, + "grad_norm": 0.4281790852546692, + "learning_rate": 4.7592554252621215e-07, + "loss": 0.0216, + "step": 41334 + }, + { + "epoch": 0.9689574945644849, + "grad_norm": 1.270397424697876, + "learning_rate": 4.7520800472700133e-07, + "loss": 0.1137, + "step": 41335 + }, + { + "epoch": 0.9689809361392898, + "grad_norm": 3.16155743598938, + "learning_rate": 4.7449100695212687e-07, + "loss": 0.1733, + "step": 41336 + }, + { + "epoch": 0.9690043777140949, + "grad_norm": 0.9418074488639832, + "learning_rate": 4.7377454920545237e-07, + "loss": 0.0945, + "step": 41337 + }, + { + "epoch": 0.9690278192888998, + "grad_norm": 1.2139792442321777, + "learning_rate": 4.730586314908747e-07, + "loss": 0.0518, + "step": 41338 + }, + { + "epoch": 0.9690512608637049, + "grad_norm": 2.7225117683410645, + "learning_rate": 4.7234325381229073e-07, + "loss": 0.173, + "step": 41339 + }, + { + "epoch": 0.9690747024385098, + "grad_norm": 3.2114171981811523, + "learning_rate": 4.716284161735529e-07, + "loss": 0.2149, + "step": 41340 + }, + { + "epoch": 0.9690981440133148, + "grad_norm": 2.6777045726776123, + "learning_rate": 4.709141185785693e-07, + "loss": 0.226, + "step": 41341 + }, + { + "epoch": 0.9691215855881198, + "grad_norm": 2.75191593170166, + "learning_rate": 4.7020036103119225e-07, + "loss": 0.3778, + "step": 41342 + }, + { + "epoch": 0.9691450271629248, + "grad_norm": 5.333038806915283, + "learning_rate": 4.6948714353531875e-07, + "loss": 0.3622, + "step": 41343 + }, + { + "epoch": 0.9691684687377298, + "grad_norm": 0.3157399296760559, + "learning_rate": 4.6877446609481235e-07, + "loss": 0.0205, + "step": 41344 + }, + { + "epoch": 0.9691919103125348, + "grad_norm": 2.379554033279419, + "learning_rate": 4.680623287135255e-07, + "loss": 0.1145, + "step": 41345 + }, + { + "epoch": 0.9692153518873398, + "grad_norm": 2.625612258911133, + "learning_rate": 4.673507313953329e-07, + "loss": 0.2557, + "step": 41346 + }, + { + "epoch": 0.9692387934621448, + "grad_norm": 1.2328693866729736, + "learning_rate": 4.6663967414408706e-07, + "loss": 0.0707, + "step": 41347 + }, + { + "epoch": 0.9692622350369497, + "grad_norm": 3.4205572605133057, + "learning_rate": 4.6592915696366255e-07, + "loss": 0.2806, + "step": 41348 + }, + { + "epoch": 0.9692856766117548, + "grad_norm": 2.5298898220062256, + "learning_rate": 4.652191798579009e-07, + "loss": 0.1248, + "step": 41349 + }, + { + "epoch": 0.9693091181865597, + "grad_norm": 0.4215523898601532, + "learning_rate": 4.645097428306544e-07, + "loss": 0.0343, + "step": 41350 + }, + { + "epoch": 0.9693325597613648, + "grad_norm": 2.623840093612671, + "learning_rate": 4.638008458857868e-07, + "loss": 0.2618, + "step": 41351 + }, + { + "epoch": 0.9693560013361697, + "grad_norm": 3.2013869285583496, + "learning_rate": 4.630924890271171e-07, + "loss": 0.1393, + "step": 41352 + }, + { + "epoch": 0.9693794429109748, + "grad_norm": 3.4512293338775635, + "learning_rate": 4.623846722585201e-07, + "loss": 0.3115, + "step": 41353 + }, + { + "epoch": 0.9694028844857797, + "grad_norm": 2.416463851928711, + "learning_rate": 4.6167739558381493e-07, + "loss": 0.3012, + "step": 41354 + }, + { + "epoch": 0.9694263260605848, + "grad_norm": 2.3847949504852295, + "learning_rate": 4.609706590068541e-07, + "loss": 0.1645, + "step": 41355 + }, + { + "epoch": 0.9694497676353897, + "grad_norm": 2.7917544841766357, + "learning_rate": 4.602644625314678e-07, + "loss": 2.2126, + "step": 41356 + }, + { + "epoch": 0.9694732092101948, + "grad_norm": 2.6190567016601562, + "learning_rate": 4.595588061614864e-07, + "loss": 0.1642, + "step": 41357 + }, + { + "epoch": 0.9694966507849997, + "grad_norm": 1.5586578845977783, + "learning_rate": 4.5885368990072896e-07, + "loss": 0.0526, + "step": 41358 + }, + { + "epoch": 0.9695200923598047, + "grad_norm": 2.771914005279541, + "learning_rate": 4.581491137530258e-07, + "loss": 0.2458, + "step": 41359 + }, + { + "epoch": 0.9695435339346097, + "grad_norm": 1.6598639488220215, + "learning_rate": 4.5744507772221835e-07, + "loss": 0.0552, + "step": 41360 + }, + { + "epoch": 0.9695669755094147, + "grad_norm": 1.9269739389419556, + "learning_rate": 4.567415818121035e-07, + "loss": 1.6375, + "step": 41361 + }, + { + "epoch": 0.9695904170842197, + "grad_norm": 0.5563507080078125, + "learning_rate": 4.560386260265115e-07, + "loss": 0.0316, + "step": 41362 + }, + { + "epoch": 0.9696138586590247, + "grad_norm": 2.344031810760498, + "learning_rate": 4.5533621036926154e-07, + "loss": 0.0875, + "step": 41363 + }, + { + "epoch": 0.9696373002338297, + "grad_norm": 0.9174139499664307, + "learning_rate": 4.546343348441395e-07, + "loss": 0.0781, + "step": 41364 + }, + { + "epoch": 0.9696607418086347, + "grad_norm": 1.5992152690887451, + "learning_rate": 4.539329994549757e-07, + "loss": 0.9303, + "step": 41365 + }, + { + "epoch": 0.9696841833834396, + "grad_norm": 0.8261155486106873, + "learning_rate": 4.53232204205567e-07, + "loss": 0.0631, + "step": 41366 + }, + { + "epoch": 0.9697076249582447, + "grad_norm": 2.4113235473632812, + "learning_rate": 4.5253194909973264e-07, + "loss": 0.1932, + "step": 41367 + }, + { + "epoch": 0.9697310665330496, + "grad_norm": 2.622852325439453, + "learning_rate": 4.518322341412473e-07, + "loss": 0.1362, + "step": 41368 + }, + { + "epoch": 0.9697545081078547, + "grad_norm": 1.3233615159988403, + "learning_rate": 4.511330593339302e-07, + "loss": 0.1271, + "step": 41369 + }, + { + "epoch": 0.9697779496826597, + "grad_norm": 2.5204474925994873, + "learning_rate": 4.5043442468155617e-07, + "loss": 0.1933, + "step": 41370 + }, + { + "epoch": 0.9698013912574647, + "grad_norm": 2.237136125564575, + "learning_rate": 4.4973633018792207e-07, + "loss": 0.1292, + "step": 41371 + }, + { + "epoch": 0.9698248328322697, + "grad_norm": 0.4182225167751312, + "learning_rate": 4.4903877585682487e-07, + "loss": 0.0265, + "step": 41372 + }, + { + "epoch": 0.9698482744070747, + "grad_norm": 2.438342809677124, + "learning_rate": 4.4834176169203935e-07, + "loss": 1.554, + "step": 41373 + }, + { + "epoch": 0.9698717159818797, + "grad_norm": 4.988926410675049, + "learning_rate": 4.476452876973514e-07, + "loss": 0.5691, + "step": 41374 + }, + { + "epoch": 0.9698951575566847, + "grad_norm": 0.5114757418632507, + "learning_rate": 4.469493538765357e-07, + "loss": 0.0199, + "step": 41375 + }, + { + "epoch": 0.9699185991314897, + "grad_norm": 1.1117802858352661, + "learning_rate": 4.4625396023337815e-07, + "loss": 0.0404, + "step": 41376 + }, + { + "epoch": 0.9699420407062946, + "grad_norm": 4.808274269104004, + "learning_rate": 4.4555910677163136e-07, + "loss": 0.1492, + "step": 41377 + }, + { + "epoch": 0.9699654822810997, + "grad_norm": 0.7918978929519653, + "learning_rate": 4.4486479349509224e-07, + "loss": 0.0731, + "step": 41378 + }, + { + "epoch": 0.9699889238559046, + "grad_norm": 1.8138127326965332, + "learning_rate": 4.4417102040751335e-07, + "loss": 1.34, + "step": 41379 + }, + { + "epoch": 0.9700123654307097, + "grad_norm": 0.753093421459198, + "learning_rate": 4.4347778751265834e-07, + "loss": 0.3309, + "step": 41380 + }, + { + "epoch": 0.9700358070055146, + "grad_norm": 2.3219563961029053, + "learning_rate": 4.4278509481430195e-07, + "loss": 0.1629, + "step": 41381 + }, + { + "epoch": 0.9700592485803197, + "grad_norm": 0.8846388459205627, + "learning_rate": 4.420929423161857e-07, + "loss": 0.0788, + "step": 41382 + }, + { + "epoch": 0.9700826901551246, + "grad_norm": 0.40498167276382446, + "learning_rate": 4.414013300220732e-07, + "loss": 0.0242, + "step": 41383 + }, + { + "epoch": 0.9701061317299297, + "grad_norm": 2.0118651390075684, + "learning_rate": 4.4071025793571694e-07, + "loss": 0.1563, + "step": 41384 + }, + { + "epoch": 0.9701295733047346, + "grad_norm": 3.0952794551849365, + "learning_rate": 4.400197260608696e-07, + "loss": 0.3694, + "step": 41385 + }, + { + "epoch": 0.9701530148795396, + "grad_norm": 2.1305861473083496, + "learning_rate": 4.393297344012726e-07, + "loss": 0.1984, + "step": 41386 + }, + { + "epoch": 0.9701764564543446, + "grad_norm": 3.289381504058838, + "learning_rate": 4.3864028296066727e-07, + "loss": 1.513, + "step": 41387 + }, + { + "epoch": 0.9701998980291496, + "grad_norm": 2.2013750076293945, + "learning_rate": 4.379513717428063e-07, + "loss": 0.18, + "step": 41388 + }, + { + "epoch": 0.9702233396039546, + "grad_norm": 4.3510942459106445, + "learning_rate": 4.372630007514089e-07, + "loss": 0.3541, + "step": 41389 + }, + { + "epoch": 0.9702467811787596, + "grad_norm": 0.959184467792511, + "learning_rate": 4.3657516999023875e-07, + "loss": 0.0795, + "step": 41390 + }, + { + "epoch": 0.9702702227535646, + "grad_norm": 4.429970741271973, + "learning_rate": 4.3588787946300394e-07, + "loss": 0.3407, + "step": 41391 + }, + { + "epoch": 0.9702936643283696, + "grad_norm": 1.8807854652404785, + "learning_rate": 4.352011291734348e-07, + "loss": 0.1562, + "step": 41392 + }, + { + "epoch": 0.9703171059031745, + "grad_norm": 1.110599160194397, + "learning_rate": 4.3451491912526177e-07, + "loss": 0.04, + "step": 41393 + }, + { + "epoch": 0.9703405474779796, + "grad_norm": 1.6498044729232788, + "learning_rate": 4.3382924932222625e-07, + "loss": 0.1632, + "step": 41394 + }, + { + "epoch": 0.9703639890527845, + "grad_norm": 3.255436897277832, + "learning_rate": 4.3314411976802525e-07, + "loss": 0.1482, + "step": 41395 + }, + { + "epoch": 0.9703874306275896, + "grad_norm": 3.664472818374634, + "learning_rate": 4.324595304663781e-07, + "loss": 0.303, + "step": 41396 + }, + { + "epoch": 0.9704108722023945, + "grad_norm": 1.561804175376892, + "learning_rate": 4.31775481421004e-07, + "loss": 1.1942, + "step": 41397 + }, + { + "epoch": 0.9704343137771996, + "grad_norm": 0.7189221978187561, + "learning_rate": 4.310919726356222e-07, + "loss": 0.0523, + "step": 41398 + }, + { + "epoch": 0.9704577553520045, + "grad_norm": 0.5798328518867493, + "learning_rate": 4.304090041139408e-07, + "loss": 0.0395, + "step": 41399 + }, + { + "epoch": 0.9704811969268096, + "grad_norm": 2.2038698196411133, + "learning_rate": 4.2972657585964584e-07, + "loss": 0.131, + "step": 41400 + }, + { + "epoch": 0.9705046385016145, + "grad_norm": 3.4903128147125244, + "learning_rate": 4.290446878764565e-07, + "loss": 0.2609, + "step": 41401 + }, + { + "epoch": 0.9705280800764196, + "grad_norm": 2.0768988132476807, + "learning_rate": 4.283633401680809e-07, + "loss": 0.1729, + "step": 41402 + }, + { + "epoch": 0.9705515216512245, + "grad_norm": 4.234241008758545, + "learning_rate": 4.2768253273820504e-07, + "loss": 0.3134, + "step": 41403 + }, + { + "epoch": 0.9705749632260295, + "grad_norm": 1.329690933227539, + "learning_rate": 4.270022655905148e-07, + "loss": 0.042, + "step": 41404 + }, + { + "epoch": 0.9705984048008345, + "grad_norm": 1.7559460401535034, + "learning_rate": 4.2632253872870733e-07, + "loss": 0.205, + "step": 41405 + }, + { + "epoch": 0.9706218463756395, + "grad_norm": 0.4637989103794098, + "learning_rate": 4.2564335215647957e-07, + "loss": 0.0249, + "step": 41406 + }, + { + "epoch": 0.9706452879504445, + "grad_norm": 3.5633702278137207, + "learning_rate": 4.249647058775175e-07, + "loss": 0.2629, + "step": 41407 + }, + { + "epoch": 0.9706687295252495, + "grad_norm": 2.6657888889312744, + "learning_rate": 4.242865998954848e-07, + "loss": 0.1922, + "step": 41408 + }, + { + "epoch": 0.9706921711000545, + "grad_norm": 2.618741035461426, + "learning_rate": 4.236090342140786e-07, + "loss": 0.2299, + "step": 41409 + }, + { + "epoch": 0.9707156126748595, + "grad_norm": 1.3743457794189453, + "learning_rate": 4.229320088369626e-07, + "loss": 0.0852, + "step": 41410 + }, + { + "epoch": 0.9707390542496644, + "grad_norm": 1.9559482336044312, + "learning_rate": 4.222555237678338e-07, + "loss": 1.369, + "step": 41411 + }, + { + "epoch": 0.9707624958244695, + "grad_norm": 2.42913818359375, + "learning_rate": 4.2157957901033386e-07, + "loss": 0.2648, + "step": 41412 + }, + { + "epoch": 0.9707859373992744, + "grad_norm": 0.5465770363807678, + "learning_rate": 4.2090417456813747e-07, + "loss": 0.0642, + "step": 41413 + }, + { + "epoch": 0.9708093789740795, + "grad_norm": 2.410834550857544, + "learning_rate": 4.202293104449306e-07, + "loss": 0.3064, + "step": 41414 + }, + { + "epoch": 0.9708328205488844, + "grad_norm": 4.514214992523193, + "learning_rate": 4.195549866443549e-07, + "loss": 0.2337, + "step": 41415 + }, + { + "epoch": 0.9708562621236895, + "grad_norm": 3.258571147918701, + "learning_rate": 4.1888120317006285e-07, + "loss": 0.0885, + "step": 41416 + }, + { + "epoch": 0.9708797036984944, + "grad_norm": 1.0630477666854858, + "learning_rate": 4.1820796002572936e-07, + "loss": 0.0402, + "step": 41417 + }, + { + "epoch": 0.9709031452732995, + "grad_norm": 2.055447816848755, + "learning_rate": 4.17535257214996e-07, + "loss": 0.1212, + "step": 41418 + }, + { + "epoch": 0.9709265868481044, + "grad_norm": 2.3183605670928955, + "learning_rate": 4.168630947415264e-07, + "loss": 0.2125, + "step": 41419 + }, + { + "epoch": 0.9709500284229095, + "grad_norm": 1.5767755508422852, + "learning_rate": 4.1619147260895106e-07, + "loss": 0.1269, + "step": 41420 + }, + { + "epoch": 0.9709734699977145, + "grad_norm": 2.300560474395752, + "learning_rate": 4.1552039082092267e-07, + "loss": 0.1038, + "step": 41421 + }, + { + "epoch": 0.9709969115725194, + "grad_norm": 1.5508898496627808, + "learning_rate": 4.1484984938108263e-07, + "loss": 0.0809, + "step": 41422 + }, + { + "epoch": 0.9710203531473245, + "grad_norm": 0.7231253981590271, + "learning_rate": 4.1417984829305037e-07, + "loss": 0.0146, + "step": 41423 + }, + { + "epoch": 0.9710437947221294, + "grad_norm": 2.717686176300049, + "learning_rate": 4.1351038756050063e-07, + "loss": 0.1763, + "step": 41424 + }, + { + "epoch": 0.9710672362969345, + "grad_norm": 1.763043999671936, + "learning_rate": 4.1284146718703066e-07, + "loss": 1.1408, + "step": 41425 + }, + { + "epoch": 0.9710906778717394, + "grad_norm": 0.9982231855392456, + "learning_rate": 4.121730871762819e-07, + "loss": 0.0566, + "step": 41426 + }, + { + "epoch": 0.9711141194465445, + "grad_norm": 0.41315922141075134, + "learning_rate": 4.115052475318959e-07, + "loss": 0.0297, + "step": 41427 + }, + { + "epoch": 0.9711375610213494, + "grad_norm": 0.514876663684845, + "learning_rate": 4.108379482574698e-07, + "loss": 0.0103, + "step": 41428 + }, + { + "epoch": 0.9711610025961545, + "grad_norm": 0.7338535785675049, + "learning_rate": 4.101711893566451e-07, + "loss": 0.0532, + "step": 41429 + }, + { + "epoch": 0.9711844441709594, + "grad_norm": 2.233853816986084, + "learning_rate": 4.09504970833019e-07, + "loss": 0.169, + "step": 41430 + }, + { + "epoch": 0.9712078857457644, + "grad_norm": 2.67279052734375, + "learning_rate": 4.0883929269023293e-07, + "loss": 0.148, + "step": 41431 + }, + { + "epoch": 0.9712313273205694, + "grad_norm": 3.825537919998169, + "learning_rate": 4.0817415493188406e-07, + "loss": 0.2231, + "step": 41432 + }, + { + "epoch": 0.9712547688953744, + "grad_norm": 2.3897645473480225, + "learning_rate": 4.075095575615695e-07, + "loss": 0.1852, + "step": 41433 + }, + { + "epoch": 0.9712782104701794, + "grad_norm": 1.0050727128982544, + "learning_rate": 4.068455005829197e-07, + "loss": 0.077, + "step": 41434 + }, + { + "epoch": 0.9713016520449844, + "grad_norm": 2.2939717769622803, + "learning_rate": 4.061819839995207e-07, + "loss": 0.9661, + "step": 41435 + }, + { + "epoch": 0.9713250936197894, + "grad_norm": 2.2439374923706055, + "learning_rate": 4.0551900781499175e-07, + "loss": 0.0999, + "step": 41436 + }, + { + "epoch": 0.9713485351945944, + "grad_norm": 0.7836329340934753, + "learning_rate": 4.0485657203290786e-07, + "loss": 0.0566, + "step": 41437 + }, + { + "epoch": 0.9713719767693993, + "grad_norm": 0.5457519888877869, + "learning_rate": 4.0419467665686605e-07, + "loss": 0.0323, + "step": 41438 + }, + { + "epoch": 0.9713954183442044, + "grad_norm": 1.7971538305282593, + "learning_rate": 4.035333216904746e-07, + "loss": 1.249, + "step": 41439 + }, + { + "epoch": 0.9714188599190093, + "grad_norm": 2.3865718841552734, + "learning_rate": 4.028725071373085e-07, + "loss": 0.1929, + "step": 41440 + }, + { + "epoch": 0.9714423014938144, + "grad_norm": 1.4455959796905518, + "learning_rate": 4.0221223300095365e-07, + "loss": 1.0115, + "step": 41441 + }, + { + "epoch": 0.9714657430686193, + "grad_norm": 2.8031911849975586, + "learning_rate": 4.0155249928499615e-07, + "loss": 0.1125, + "step": 41442 + }, + { + "epoch": 0.9714891846434244, + "grad_norm": 3.023054838180542, + "learning_rate": 4.008933059930109e-07, + "loss": 0.2901, + "step": 41443 + }, + { + "epoch": 0.9715126262182293, + "grad_norm": 3.718059539794922, + "learning_rate": 4.00234653128595e-07, + "loss": 0.3359, + "step": 41444 + }, + { + "epoch": 0.9715360677930344, + "grad_norm": 3.3281829357147217, + "learning_rate": 3.9957654069529004e-07, + "loss": 0.2866, + "step": 41445 + }, + { + "epoch": 0.9715595093678393, + "grad_norm": 2.0578935146331787, + "learning_rate": 3.989189686966932e-07, + "loss": 0.1279, + "step": 41446 + }, + { + "epoch": 0.9715829509426444, + "grad_norm": 2.6986441612243652, + "learning_rate": 3.9826193713635717e-07, + "loss": 0.1185, + "step": 41447 + }, + { + "epoch": 0.9716063925174493, + "grad_norm": 2.983888864517212, + "learning_rate": 3.9760544601785685e-07, + "loss": 0.1636, + "step": 41448 + }, + { + "epoch": 0.9716298340922543, + "grad_norm": 1.2508870363235474, + "learning_rate": 3.96949495344745e-07, + "loss": 0.1741, + "step": 41449 + }, + { + "epoch": 0.9716532756670593, + "grad_norm": 1.0681873559951782, + "learning_rate": 3.962940851205854e-07, + "loss": 0.0715, + "step": 41450 + }, + { + "epoch": 0.9716767172418643, + "grad_norm": 2.908473253250122, + "learning_rate": 3.9563921534894187e-07, + "loss": 1.4285, + "step": 41451 + }, + { + "epoch": 0.9717001588166693, + "grad_norm": 4.102509021759033, + "learning_rate": 3.9498488603335605e-07, + "loss": 0.3592, + "step": 41452 + }, + { + "epoch": 0.9717236003914743, + "grad_norm": 1.3129870891571045, + "learning_rate": 3.943310971773806e-07, + "loss": 0.0738, + "step": 41453 + }, + { + "epoch": 0.9717470419662793, + "grad_norm": 0.5607498288154602, + "learning_rate": 3.936778487845683e-07, + "loss": 0.0512, + "step": 41454 + }, + { + "epoch": 0.9717704835410843, + "grad_norm": 0.9058919548988342, + "learning_rate": 3.9302514085846066e-07, + "loss": 0.0687, + "step": 41455 + }, + { + "epoch": 0.9717939251158892, + "grad_norm": 1.5395910739898682, + "learning_rate": 3.9237297340259936e-07, + "loss": 0.0918, + "step": 41456 + }, + { + "epoch": 0.9718173666906943, + "grad_norm": 3.6120975017547607, + "learning_rate": 3.9172134642051493e-07, + "loss": 0.3376, + "step": 41457 + }, + { + "epoch": 0.9718408082654992, + "grad_norm": 1.3200565576553345, + "learning_rate": 3.910702599157601e-07, + "loss": 0.0893, + "step": 41458 + }, + { + "epoch": 0.9718642498403043, + "grad_norm": 2.5980985164642334, + "learning_rate": 3.904197138918653e-07, + "loss": 0.2187, + "step": 41459 + }, + { + "epoch": 0.9718876914151092, + "grad_norm": 0.7946816086769104, + "learning_rate": 3.8976970835233884e-07, + "loss": 0.0185, + "step": 41460 + }, + { + "epoch": 0.9719111329899143, + "grad_norm": 3.04093337059021, + "learning_rate": 3.891202433007335e-07, + "loss": 0.1826, + "step": 41461 + }, + { + "epoch": 0.9719345745647192, + "grad_norm": 5.2259674072265625, + "learning_rate": 3.884713187405575e-07, + "loss": 0.3913, + "step": 41462 + }, + { + "epoch": 0.9719580161395243, + "grad_norm": 0.6549270749092102, + "learning_rate": 3.8782293467533036e-07, + "loss": 0.0529, + "step": 41463 + }, + { + "epoch": 0.9719814577143292, + "grad_norm": 2.7915151119232178, + "learning_rate": 3.8717509110859363e-07, + "loss": 0.1856, + "step": 41464 + }, + { + "epoch": 0.9720048992891342, + "grad_norm": 1.4345567226409912, + "learning_rate": 3.865277880438334e-07, + "loss": 0.1107, + "step": 41465 + }, + { + "epoch": 0.9720283408639392, + "grad_norm": 0.7680564522743225, + "learning_rate": 3.8588102548458016e-07, + "loss": 0.1096, + "step": 41466 + }, + { + "epoch": 0.9720517824387442, + "grad_norm": 5.186408042907715, + "learning_rate": 3.8523480343433115e-07, + "loss": 0.6438, + "step": 41467 + }, + { + "epoch": 0.9720752240135492, + "grad_norm": 0.6532955169677734, + "learning_rate": 3.845891218966058e-07, + "loss": 0.0349, + "step": 41468 + }, + { + "epoch": 0.9720986655883542, + "grad_norm": 0.544491708278656, + "learning_rate": 3.8394398087490125e-07, + "loss": 0.0726, + "step": 41469 + }, + { + "epoch": 0.9721221071631592, + "grad_norm": 3.1932919025421143, + "learning_rate": 3.8329938037271474e-07, + "loss": 0.2614, + "step": 41470 + }, + { + "epoch": 0.9721455487379642, + "grad_norm": 2.3033292293548584, + "learning_rate": 3.8265532039354345e-07, + "loss": 0.1342, + "step": 41471 + }, + { + "epoch": 0.9721689903127693, + "grad_norm": 2.00852370262146, + "learning_rate": 3.820118009408846e-07, + "loss": 1.585, + "step": 41472 + }, + { + "epoch": 0.9721924318875742, + "grad_norm": 2.5515060424804688, + "learning_rate": 3.8136882201823545e-07, + "loss": 0.1236, + "step": 41473 + }, + { + "epoch": 0.9722158734623793, + "grad_norm": 1.0181273221969604, + "learning_rate": 3.8072638362908195e-07, + "loss": 0.0832, + "step": 41474 + }, + { + "epoch": 0.9722393150371842, + "grad_norm": 2.3031156063079834, + "learning_rate": 3.8008448577689927e-07, + "loss": 0.3053, + "step": 41475 + }, + { + "epoch": 0.9722627566119892, + "grad_norm": 1.07192862033844, + "learning_rate": 3.794431284651845e-07, + "loss": 0.099, + "step": 41476 + }, + { + "epoch": 0.9722861981867942, + "grad_norm": 1.803688645362854, + "learning_rate": 3.7880231169741267e-07, + "loss": 0.1629, + "step": 41477 + }, + { + "epoch": 0.9723096397615992, + "grad_norm": 1.5580910444259644, + "learning_rate": 3.7816203547705877e-07, + "loss": 0.0831, + "step": 41478 + }, + { + "epoch": 0.9723330813364042, + "grad_norm": 1.6560251712799072, + "learning_rate": 3.775222998076089e-07, + "loss": 0.162, + "step": 41479 + }, + { + "epoch": 0.9723565229112092, + "grad_norm": 2.6854560375213623, + "learning_rate": 3.768831046925159e-07, + "loss": 0.3334, + "step": 41480 + }, + { + "epoch": 0.9723799644860142, + "grad_norm": 2.7028019428253174, + "learning_rate": 3.762444501352547e-07, + "loss": 0.2845, + "step": 41481 + }, + { + "epoch": 0.9724034060608192, + "grad_norm": 2.8789470195770264, + "learning_rate": 3.756063361392892e-07, + "loss": 0.1702, + "step": 41482 + }, + { + "epoch": 0.9724268476356241, + "grad_norm": 2.933903217315674, + "learning_rate": 3.7496876270809444e-07, + "loss": 0.1036, + "step": 41483 + }, + { + "epoch": 0.9724502892104292, + "grad_norm": 1.5390007495880127, + "learning_rate": 3.743317298451121e-07, + "loss": 0.1617, + "step": 41484 + }, + { + "epoch": 0.9724737307852341, + "grad_norm": 2.8564534187316895, + "learning_rate": 3.736952375538172e-07, + "loss": 0.2675, + "step": 41485 + }, + { + "epoch": 0.9724971723600392, + "grad_norm": 0.7659358978271484, + "learning_rate": 3.730592858376403e-07, + "loss": 0.0515, + "step": 41486 + }, + { + "epoch": 0.9725206139348441, + "grad_norm": 3.6755826473236084, + "learning_rate": 3.7242387470005634e-07, + "loss": 0.4878, + "step": 41487 + }, + { + "epoch": 0.9725440555096492, + "grad_norm": 2.243330478668213, + "learning_rate": 3.7178900414449605e-07, + "loss": 1.1874, + "step": 41488 + }, + { + "epoch": 0.9725674970844541, + "grad_norm": 2.538228750228882, + "learning_rate": 3.711546741744121e-07, + "loss": 0.3216, + "step": 41489 + }, + { + "epoch": 0.9725909386592592, + "grad_norm": 1.848798394203186, + "learning_rate": 3.705208847932351e-07, + "loss": 0.2034, + "step": 41490 + }, + { + "epoch": 0.9726143802340641, + "grad_norm": 1.0292023420333862, + "learning_rate": 3.69887636004429e-07, + "loss": 0.0948, + "step": 41491 + }, + { + "epoch": 0.9726378218088692, + "grad_norm": 3.6602509021759033, + "learning_rate": 3.692549278114021e-07, + "loss": 0.3024, + "step": 41492 + }, + { + "epoch": 0.9726612633836741, + "grad_norm": 6.686319351196289, + "learning_rate": 3.6862276021759623e-07, + "loss": 0.2938, + "step": 41493 + }, + { + "epoch": 0.9726847049584791, + "grad_norm": 1.128522515296936, + "learning_rate": 3.6799113322645297e-07, + "loss": 0.0452, + "step": 41494 + }, + { + "epoch": 0.9727081465332841, + "grad_norm": 1.2600300312042236, + "learning_rate": 3.673600468413918e-07, + "loss": 0.0935, + "step": 41495 + }, + { + "epoch": 0.9727315881080891, + "grad_norm": 0.6867110133171082, + "learning_rate": 3.6672950106583224e-07, + "loss": 0.0349, + "step": 41496 + }, + { + "epoch": 0.9727550296828941, + "grad_norm": 2.9310760498046875, + "learning_rate": 3.6609949590320494e-07, + "loss": 0.3186, + "step": 41497 + }, + { + "epoch": 0.9727784712576991, + "grad_norm": 2.2063310146331787, + "learning_rate": 3.6547003135691815e-07, + "loss": 0.1025, + "step": 41498 + }, + { + "epoch": 0.9728019128325041, + "grad_norm": 0.9090109467506409, + "learning_rate": 3.6484110743040256e-07, + "loss": 0.0858, + "step": 41499 + }, + { + "epoch": 0.9728253544073091, + "grad_norm": 3.465698719024658, + "learning_rate": 3.6421272412704434e-07, + "loss": 0.1909, + "step": 41500 + }, + { + "epoch": 0.972848795982114, + "grad_norm": 0.7620819807052612, + "learning_rate": 3.6358488145027403e-07, + "loss": 0.0377, + "step": 41501 + }, + { + "epoch": 0.9728722375569191, + "grad_norm": 2.7467849254608154, + "learning_rate": 3.6295757940350005e-07, + "loss": 0.1926, + "step": 41502 + }, + { + "epoch": 0.972895679131724, + "grad_norm": 2.673809766769409, + "learning_rate": 3.623308179901197e-07, + "loss": 0.2543, + "step": 41503 + }, + { + "epoch": 0.9729191207065291, + "grad_norm": 1.2138415575027466, + "learning_rate": 3.6170459721353025e-07, + "loss": 0.05, + "step": 41504 + }, + { + "epoch": 0.972942562281334, + "grad_norm": 2.945638656616211, + "learning_rate": 3.6107891707714e-07, + "loss": 0.2511, + "step": 41505 + }, + { + "epoch": 0.9729660038561391, + "grad_norm": 0.5825858116149902, + "learning_rate": 3.604537775843242e-07, + "loss": 0.0491, + "step": 41506 + }, + { + "epoch": 0.972989445430944, + "grad_norm": 1.78069269657135, + "learning_rate": 3.598291787385022e-07, + "loss": 0.1347, + "step": 41507 + }, + { + "epoch": 0.9730128870057491, + "grad_norm": 2.795109272003174, + "learning_rate": 3.59205120543038e-07, + "loss": 0.2958, + "step": 41508 + }, + { + "epoch": 0.973036328580554, + "grad_norm": 1.9541325569152832, + "learning_rate": 3.585816030013289e-07, + "loss": 0.1816, + "step": 41509 + }, + { + "epoch": 0.973059770155359, + "grad_norm": 1.0334696769714355, + "learning_rate": 3.579586261167611e-07, + "loss": 0.12, + "step": 41510 + }, + { + "epoch": 0.973083211730164, + "grad_norm": 0.3677651286125183, + "learning_rate": 3.5733618989270966e-07, + "loss": 0.0114, + "step": 41511 + }, + { + "epoch": 0.973106653304969, + "grad_norm": 2.6392035484313965, + "learning_rate": 3.567142943325608e-07, + "loss": 0.2263, + "step": 41512 + }, + { + "epoch": 0.973130094879774, + "grad_norm": 2.8231239318847656, + "learning_rate": 3.5609293943966725e-07, + "loss": 1.5001, + "step": 41513 + }, + { + "epoch": 0.973153536454579, + "grad_norm": 2.482137680053711, + "learning_rate": 3.5547212521742645e-07, + "loss": 0.2243, + "step": 41514 + }, + { + "epoch": 0.973176978029384, + "grad_norm": 1.9816820621490479, + "learning_rate": 3.548518516691912e-07, + "loss": 0.0995, + "step": 41515 + }, + { + "epoch": 0.973200419604189, + "grad_norm": 2.9504358768463135, + "learning_rate": 3.542321187983366e-07, + "loss": 1.1255, + "step": 41516 + }, + { + "epoch": 0.973223861178994, + "grad_norm": 5.837112903594971, + "learning_rate": 3.5361292660821553e-07, + "loss": 0.3995, + "step": 41517 + }, + { + "epoch": 0.973247302753799, + "grad_norm": 0.6704616546630859, + "learning_rate": 3.5299427510219197e-07, + "loss": 0.0671, + "step": 41518 + }, + { + "epoch": 0.9732707443286039, + "grad_norm": 2.0271449089050293, + "learning_rate": 3.5237616428362987e-07, + "loss": 0.2076, + "step": 41519 + }, + { + "epoch": 0.973294185903409, + "grad_norm": 2.882824420928955, + "learning_rate": 3.51758594155871e-07, + "loss": 0.3035, + "step": 41520 + }, + { + "epoch": 0.9733176274782139, + "grad_norm": 0.5403229594230652, + "learning_rate": 3.511415647222793e-07, + "loss": 0.0542, + "step": 41521 + }, + { + "epoch": 0.973341069053019, + "grad_norm": 1.4706107378005981, + "learning_rate": 3.5052507598618556e-07, + "loss": 0.0892, + "step": 41522 + }, + { + "epoch": 0.973364510627824, + "grad_norm": 1.121272325515747, + "learning_rate": 3.4990912795095367e-07, + "loss": 0.1364, + "step": 41523 + }, + { + "epoch": 0.973387952202629, + "grad_norm": 0.7570439577102661, + "learning_rate": 3.492937206199032e-07, + "loss": 0.0808, + "step": 41524 + }, + { + "epoch": 0.973411393777434, + "grad_norm": 1.010210394859314, + "learning_rate": 3.486788539963981e-07, + "loss": 0.0926, + "step": 41525 + }, + { + "epoch": 0.973434835352239, + "grad_norm": 2.7883293628692627, + "learning_rate": 3.4806452808375803e-07, + "loss": 1.6817, + "step": 41526 + }, + { + "epoch": 0.973458276927044, + "grad_norm": 1.202067255973816, + "learning_rate": 3.4745074288532466e-07, + "loss": 0.0711, + "step": 41527 + }, + { + "epoch": 0.973481718501849, + "grad_norm": 3.497753858566284, + "learning_rate": 3.4683749840442867e-07, + "loss": 0.1744, + "step": 41528 + }, + { + "epoch": 0.973505160076654, + "grad_norm": 1.9619718790054321, + "learning_rate": 3.4622479464438975e-07, + "loss": 0.1228, + "step": 41529 + }, + { + "epoch": 0.9735286016514589, + "grad_norm": 0.9559443593025208, + "learning_rate": 3.4561263160853843e-07, + "loss": 0.0458, + "step": 41530 + }, + { + "epoch": 0.973552043226264, + "grad_norm": 3.517350196838379, + "learning_rate": 3.450010093001943e-07, + "loss": 0.3624, + "step": 41531 + }, + { + "epoch": 0.9735754848010689, + "grad_norm": 2.0324289798736572, + "learning_rate": 3.44389927722677e-07, + "loss": 0.9111, + "step": 41532 + }, + { + "epoch": 0.973598926375874, + "grad_norm": 3.8591811656951904, + "learning_rate": 3.43779386879306e-07, + "loss": 0.2837, + "step": 41533 + }, + { + "epoch": 0.9736223679506789, + "grad_norm": 1.6014825105667114, + "learning_rate": 3.4316938677338986e-07, + "loss": 0.1415, + "step": 41534 + }, + { + "epoch": 0.973645809525484, + "grad_norm": 2.7548367977142334, + "learning_rate": 3.4255992740822585e-07, + "loss": 1.6773, + "step": 41535 + }, + { + "epoch": 0.9736692511002889, + "grad_norm": 2.050098180770874, + "learning_rate": 3.419510087871447e-07, + "loss": 0.1575, + "step": 41536 + }, + { + "epoch": 0.973692692675094, + "grad_norm": 0.6541606187820435, + "learning_rate": 3.4134263091344374e-07, + "loss": 0.0291, + "step": 41537 + }, + { + "epoch": 0.9737161342498989, + "grad_norm": 0.621423065662384, + "learning_rate": 3.407347937904093e-07, + "loss": 0.0356, + "step": 41538 + }, + { + "epoch": 0.9737395758247039, + "grad_norm": 1.6613017320632935, + "learning_rate": 3.4012749742136087e-07, + "loss": 0.1378, + "step": 41539 + }, + { + "epoch": 0.9737630173995089, + "grad_norm": 1.6668962240219116, + "learning_rate": 3.395207418095736e-07, + "loss": 0.1061, + "step": 41540 + }, + { + "epoch": 0.9737864589743139, + "grad_norm": 3.119462490081787, + "learning_rate": 3.38914526958356e-07, + "loss": 0.2149, + "step": 41541 + }, + { + "epoch": 0.9738099005491189, + "grad_norm": 1.4751253128051758, + "learning_rate": 3.3830885287098326e-07, + "loss": 0.6907, + "step": 41542 + }, + { + "epoch": 0.9738333421239239, + "grad_norm": 1.1571767330169678, + "learning_rate": 3.3770371955075263e-07, + "loss": 0.1469, + "step": 41543 + }, + { + "epoch": 0.9738567836987289, + "grad_norm": 2.144813299179077, + "learning_rate": 3.370991270009394e-07, + "loss": 0.081, + "step": 41544 + }, + { + "epoch": 0.9738802252735339, + "grad_norm": 3.3278143405914307, + "learning_rate": 3.364950752248408e-07, + "loss": 0.2303, + "step": 41545 + }, + { + "epoch": 0.9739036668483388, + "grad_norm": 2.121471881866455, + "learning_rate": 3.358915642257099e-07, + "loss": 0.1352, + "step": 41546 + }, + { + "epoch": 0.9739271084231439, + "grad_norm": 0.5376843810081482, + "learning_rate": 3.3528859400683287e-07, + "loss": 0.0351, + "step": 41547 + }, + { + "epoch": 0.9739505499979488, + "grad_norm": 2.032829999923706, + "learning_rate": 3.3468616457149604e-07, + "loss": 1.2252, + "step": 41548 + }, + { + "epoch": 0.9739739915727539, + "grad_norm": 1.7861707210540771, + "learning_rate": 3.340842759229412e-07, + "loss": 0.2087, + "step": 41549 + }, + { + "epoch": 0.9739974331475588, + "grad_norm": 1.7662690877914429, + "learning_rate": 3.334829280644547e-07, + "loss": 1.3435, + "step": 41550 + }, + { + "epoch": 0.9740208747223639, + "grad_norm": 2.6908795833587646, + "learning_rate": 3.328821209992894e-07, + "loss": 0.2551, + "step": 41551 + }, + { + "epoch": 0.9740443162971688, + "grad_norm": 1.0089974403381348, + "learning_rate": 3.322818547307094e-07, + "loss": 0.0673, + "step": 41552 + }, + { + "epoch": 0.9740677578719739, + "grad_norm": 3.2891616821289062, + "learning_rate": 3.316821292619676e-07, + "loss": 0.2635, + "step": 41553 + }, + { + "epoch": 0.9740911994467788, + "grad_norm": 1.6603754758834839, + "learning_rate": 3.310829445963171e-07, + "loss": 0.1994, + "step": 41554 + }, + { + "epoch": 0.9741146410215838, + "grad_norm": 0.5307032465934753, + "learning_rate": 3.3048430073702175e-07, + "loss": 0.0335, + "step": 41555 + }, + { + "epoch": 0.9741380825963888, + "grad_norm": 2.5860543251037598, + "learning_rate": 3.2988619768732353e-07, + "loss": 0.2502, + "step": 41556 + }, + { + "epoch": 0.9741615241711938, + "grad_norm": 1.2199451923370361, + "learning_rate": 3.292886354504532e-07, + "loss": 0.0365, + "step": 41557 + }, + { + "epoch": 0.9741849657459988, + "grad_norm": 0.46813520789146423, + "learning_rate": 3.2869161402967473e-07, + "loss": 0.0206, + "step": 41558 + }, + { + "epoch": 0.9742084073208038, + "grad_norm": 2.5846426486968994, + "learning_rate": 3.280951334282079e-07, + "loss": 0.1372, + "step": 41559 + }, + { + "epoch": 0.9742318488956088, + "grad_norm": 2.3413593769073486, + "learning_rate": 3.274991936493166e-07, + "loss": 0.2392, + "step": 41560 + }, + { + "epoch": 0.9742552904704138, + "grad_norm": 2.1482253074645996, + "learning_rate": 3.2690379469619834e-07, + "loss": 0.1546, + "step": 41561 + }, + { + "epoch": 0.9742787320452188, + "grad_norm": 0.786949098110199, + "learning_rate": 3.2630893657211725e-07, + "loss": 0.0307, + "step": 41562 + }, + { + "epoch": 0.9743021736200238, + "grad_norm": 1.8344303369522095, + "learning_rate": 3.257146192802818e-07, + "loss": 0.09, + "step": 41563 + }, + { + "epoch": 0.9743256151948287, + "grad_norm": 3.537419557571411, + "learning_rate": 3.251208428239116e-07, + "loss": 0.2732, + "step": 41564 + }, + { + "epoch": 0.9743490567696338, + "grad_norm": 2.0897433757781982, + "learning_rate": 3.245276072062486e-07, + "loss": 1.5558, + "step": 41565 + }, + { + "epoch": 0.9743724983444387, + "grad_norm": 2.3447821140289307, + "learning_rate": 3.239349124305013e-07, + "loss": 0.148, + "step": 41566 + }, + { + "epoch": 0.9743959399192438, + "grad_norm": 3.005812168121338, + "learning_rate": 3.233427584998894e-07, + "loss": 0.3117, + "step": 41567 + }, + { + "epoch": 0.9744193814940487, + "grad_norm": 0.8233637809753418, + "learning_rate": 3.227511454176213e-07, + "loss": 0.0654, + "step": 41568 + }, + { + "epoch": 0.9744428230688538, + "grad_norm": 4.406890392303467, + "learning_rate": 3.2216007318689456e-07, + "loss": 0.2951, + "step": 41569 + }, + { + "epoch": 0.9744662646436587, + "grad_norm": 0.6285672187805176, + "learning_rate": 3.215695418109399e-07, + "loss": 0.0489, + "step": 41570 + }, + { + "epoch": 0.9744897062184638, + "grad_norm": 0.740148663520813, + "learning_rate": 3.209795512929548e-07, + "loss": 0.0493, + "step": 41571 + }, + { + "epoch": 0.9745131477932687, + "grad_norm": 3.246286630630493, + "learning_rate": 3.203901016361366e-07, + "loss": 0.2645, + "step": 41572 + }, + { + "epoch": 0.9745365893680737, + "grad_norm": 2.483567476272583, + "learning_rate": 3.1980119284367174e-07, + "loss": 0.1517, + "step": 41573 + }, + { + "epoch": 0.9745600309428788, + "grad_norm": 1.1322306394577026, + "learning_rate": 3.192128249187798e-07, + "loss": 0.0941, + "step": 41574 + }, + { + "epoch": 0.9745834725176837, + "grad_norm": 1.6890228986740112, + "learning_rate": 3.1862499786464715e-07, + "loss": 0.1011, + "step": 41575 + }, + { + "epoch": 0.9746069140924888, + "grad_norm": 2.6063499450683594, + "learning_rate": 3.1803771168444907e-07, + "loss": 0.1671, + "step": 41576 + }, + { + "epoch": 0.9746303556672937, + "grad_norm": 1.1394319534301758, + "learning_rate": 3.174509663813829e-07, + "loss": 0.0609, + "step": 41577 + }, + { + "epoch": 0.9746537972420988, + "grad_norm": 2.8515565395355225, + "learning_rate": 3.168647619586351e-07, + "loss": 1.164, + "step": 41578 + }, + { + "epoch": 0.9746772388169037, + "grad_norm": 3.3970048427581787, + "learning_rate": 3.1627909841936975e-07, + "loss": 0.3761, + "step": 41579 + }, + { + "epoch": 0.9747006803917088, + "grad_norm": 2.0208699703216553, + "learning_rate": 3.156939757667954e-07, + "loss": 0.0651, + "step": 41580 + }, + { + "epoch": 0.9747241219665137, + "grad_norm": 2.299164295196533, + "learning_rate": 3.1510939400405395e-07, + "loss": 0.157, + "step": 41581 + }, + { + "epoch": 0.9747475635413188, + "grad_norm": 0.9480617642402649, + "learning_rate": 3.145253531343317e-07, + "loss": 0.1277, + "step": 41582 + }, + { + "epoch": 0.9747710051161237, + "grad_norm": 6.279168605804443, + "learning_rate": 3.1394185316080403e-07, + "loss": 0.3747, + "step": 41583 + }, + { + "epoch": 0.9747944466909287, + "grad_norm": 2.324644088745117, + "learning_rate": 3.133588940866239e-07, + "loss": 0.2131, + "step": 41584 + }, + { + "epoch": 0.9748178882657337, + "grad_norm": 2.430341958999634, + "learning_rate": 3.127764759149665e-07, + "loss": 0.1375, + "step": 41585 + }, + { + "epoch": 0.9748413298405387, + "grad_norm": 3.0220022201538086, + "learning_rate": 3.12194598648996e-07, + "loss": 0.1634, + "step": 41586 + }, + { + "epoch": 0.9748647714153437, + "grad_norm": 0.5289327502250671, + "learning_rate": 3.116132622918433e-07, + "loss": 0.0667, + "step": 41587 + }, + { + "epoch": 0.9748882129901487, + "grad_norm": 1.782568335533142, + "learning_rate": 3.1103246684668354e-07, + "loss": 0.134, + "step": 41588 + }, + { + "epoch": 0.9749116545649537, + "grad_norm": 1.7663021087646484, + "learning_rate": 3.104522123166698e-07, + "loss": 0.1011, + "step": 41589 + }, + { + "epoch": 0.9749350961397587, + "grad_norm": 2.604452133178711, + "learning_rate": 3.0987249870495505e-07, + "loss": 1.3361, + "step": 41590 + }, + { + "epoch": 0.9749585377145636, + "grad_norm": 0.5655877590179443, + "learning_rate": 3.092933260146591e-07, + "loss": 0.0171, + "step": 41591 + }, + { + "epoch": 0.9749819792893687, + "grad_norm": 3.382983922958374, + "learning_rate": 3.0871469424894605e-07, + "loss": 0.2299, + "step": 41592 + }, + { + "epoch": 0.9750054208641736, + "grad_norm": 0.7269143462181091, + "learning_rate": 3.0813660341095785e-07, + "loss": 0.0357, + "step": 41593 + }, + { + "epoch": 0.9750288624389787, + "grad_norm": 0.7575817108154297, + "learning_rate": 3.075590535038142e-07, + "loss": 0.048, + "step": 41594 + }, + { + "epoch": 0.9750523040137836, + "grad_norm": 0.7449039220809937, + "learning_rate": 3.0698204453065705e-07, + "loss": 0.0457, + "step": 41595 + }, + { + "epoch": 0.9750757455885887, + "grad_norm": 1.6584582328796387, + "learning_rate": 3.064055764946172e-07, + "loss": 0.1171, + "step": 41596 + }, + { + "epoch": 0.9750991871633936, + "grad_norm": 2.8190903663635254, + "learning_rate": 3.0582964939882555e-07, + "loss": 0.1968, + "step": 41597 + }, + { + "epoch": 0.9751226287381987, + "grad_norm": 2.9492695331573486, + "learning_rate": 3.052542632464128e-07, + "loss": 0.1153, + "step": 41598 + }, + { + "epoch": 0.9751460703130036, + "grad_norm": 3.8637235164642334, + "learning_rate": 3.046794180404877e-07, + "loss": 0.2394, + "step": 41599 + }, + { + "epoch": 0.9751695118878086, + "grad_norm": 0.7294873595237732, + "learning_rate": 3.0410511378418103e-07, + "loss": 0.0445, + "step": 41600 + }, + { + "epoch": 0.9751929534626136, + "grad_norm": 1.1440438032150269, + "learning_rate": 3.0353135048059034e-07, + "loss": 0.1173, + "step": 41601 + }, + { + "epoch": 0.9752163950374186, + "grad_norm": 1.2224843502044678, + "learning_rate": 3.0295812813285753e-07, + "loss": 0.0533, + "step": 41602 + }, + { + "epoch": 0.9752398366122236, + "grad_norm": 4.856633186340332, + "learning_rate": 3.02385446744069e-07, + "loss": 0.3642, + "step": 41603 + }, + { + "epoch": 0.9752632781870286, + "grad_norm": 3.582785129547119, + "learning_rate": 3.018133063173334e-07, + "loss": 0.1056, + "step": 41604 + }, + { + "epoch": 0.9752867197618336, + "grad_norm": 1.9189332723617554, + "learning_rate": 3.0124170685577047e-07, + "loss": 0.1127, + "step": 41605 + }, + { + "epoch": 0.9753101613366386, + "grad_norm": 3.41740345954895, + "learning_rate": 3.006706483624777e-07, + "loss": 0.2598, + "step": 41606 + }, + { + "epoch": 0.9753336029114436, + "grad_norm": 3.6412606239318848, + "learning_rate": 3.0010013084054154e-07, + "loss": 0.2671, + "step": 41607 + }, + { + "epoch": 0.9753570444862486, + "grad_norm": 2.108039140701294, + "learning_rate": 2.9953015429305954e-07, + "loss": 0.0704, + "step": 41608 + }, + { + "epoch": 0.9753804860610535, + "grad_norm": 0.7187025547027588, + "learning_rate": 2.989607187231402e-07, + "loss": 0.0547, + "step": 41609 + }, + { + "epoch": 0.9754039276358586, + "grad_norm": 3.1403753757476807, + "learning_rate": 2.9839182413387013e-07, + "loss": 0.2249, + "step": 41610 + }, + { + "epoch": 0.9754273692106635, + "grad_norm": 1.8266407251358032, + "learning_rate": 2.978234705283134e-07, + "loss": 0.7037, + "step": 41611 + }, + { + "epoch": 0.9754508107854686, + "grad_norm": 1.6016379594802856, + "learning_rate": 2.972556579095787e-07, + "loss": 0.1038, + "step": 41612 + }, + { + "epoch": 0.9754742523602735, + "grad_norm": 3.4929370880126953, + "learning_rate": 2.9668838628074125e-07, + "loss": 0.2932, + "step": 41613 + }, + { + "epoch": 0.9754976939350786, + "grad_norm": 2.004025936126709, + "learning_rate": 2.961216556448654e-07, + "loss": 0.1272, + "step": 41614 + }, + { + "epoch": 0.9755211355098835, + "grad_norm": 2.7711915969848633, + "learning_rate": 2.955554660050486e-07, + "loss": 1.267, + "step": 41615 + }, + { + "epoch": 0.9755445770846886, + "grad_norm": 3.036986827850342, + "learning_rate": 2.9498981736434396e-07, + "loss": 0.3765, + "step": 41616 + }, + { + "epoch": 0.9755680186594935, + "grad_norm": 1.4630028009414673, + "learning_rate": 2.944247097258268e-07, + "loss": 0.1622, + "step": 41617 + }, + { + "epoch": 0.9755914602342985, + "grad_norm": 1.482581377029419, + "learning_rate": 2.938601430925725e-07, + "loss": 0.1253, + "step": 41618 + }, + { + "epoch": 0.9756149018091035, + "grad_norm": 5.558128833770752, + "learning_rate": 2.932961174676341e-07, + "loss": 0.2009, + "step": 41619 + }, + { + "epoch": 0.9756383433839085, + "grad_norm": 2.804629325866699, + "learning_rate": 2.927326328540647e-07, + "loss": 0.26, + "step": 41620 + }, + { + "epoch": 0.9756617849587135, + "grad_norm": 2.0521910190582275, + "learning_rate": 2.921696892549397e-07, + "loss": 0.1932, + "step": 41621 + }, + { + "epoch": 0.9756852265335185, + "grad_norm": 0.914719820022583, + "learning_rate": 2.9160728667330104e-07, + "loss": 0.0776, + "step": 41622 + }, + { + "epoch": 0.9757086681083235, + "grad_norm": 1.976592779159546, + "learning_rate": 2.91045425112213e-07, + "loss": 0.167, + "step": 41623 + }, + { + "epoch": 0.9757321096831285, + "grad_norm": 3.13399338722229, + "learning_rate": 2.904841045747175e-07, + "loss": 0.1234, + "step": 41624 + }, + { + "epoch": 0.9757555512579336, + "grad_norm": 3.1725921630859375, + "learning_rate": 2.8992332506384557e-07, + "loss": 0.2766, + "step": 41625 + }, + { + "epoch": 0.9757789928327385, + "grad_norm": 0.3224819302558899, + "learning_rate": 2.8936308658266133e-07, + "loss": 0.0298, + "step": 41626 + }, + { + "epoch": 0.9758024344075436, + "grad_norm": 1.7770940065383911, + "learning_rate": 2.888033891341957e-07, + "loss": 0.1397, + "step": 41627 + }, + { + "epoch": 0.9758258759823485, + "grad_norm": 0.44637101888656616, + "learning_rate": 2.882442327214907e-07, + "loss": 0.0436, + "step": 41628 + }, + { + "epoch": 0.9758493175571535, + "grad_norm": 1.6966298818588257, + "learning_rate": 2.8768561734756614e-07, + "loss": 1.0044, + "step": 41629 + }, + { + "epoch": 0.9758727591319585, + "grad_norm": 0.3358694016933441, + "learning_rate": 2.8712754301547517e-07, + "loss": 0.0093, + "step": 41630 + }, + { + "epoch": 0.9758962007067635, + "grad_norm": 1.6290720701217651, + "learning_rate": 2.8657000972822645e-07, + "loss": 1.6807, + "step": 41631 + }, + { + "epoch": 0.9759196422815685, + "grad_norm": 3.690683126449585, + "learning_rate": 2.860130174888509e-07, + "loss": 0.2948, + "step": 41632 + }, + { + "epoch": 0.9759430838563735, + "grad_norm": 2.5015876293182373, + "learning_rate": 2.8545656630037943e-07, + "loss": 0.1832, + "step": 41633 + }, + { + "epoch": 0.9759665254311785, + "grad_norm": 1.1527537107467651, + "learning_rate": 2.8490065616582074e-07, + "loss": 0.1199, + "step": 41634 + }, + { + "epoch": 0.9759899670059835, + "grad_norm": 1.3970427513122559, + "learning_rate": 2.8434528708819464e-07, + "loss": 0.0533, + "step": 41635 + }, + { + "epoch": 0.9760134085807884, + "grad_norm": 2.9363391399383545, + "learning_rate": 2.8379045907050983e-07, + "loss": 0.3356, + "step": 41636 + }, + { + "epoch": 0.9760368501555935, + "grad_norm": 0.7137044072151184, + "learning_rate": 2.8323617211578615e-07, + "loss": 0.0757, + "step": 41637 + }, + { + "epoch": 0.9760602917303984, + "grad_norm": 1.1293476819992065, + "learning_rate": 2.8268242622703225e-07, + "loss": 0.0315, + "step": 41638 + }, + { + "epoch": 0.9760837333052035, + "grad_norm": 1.4341285228729248, + "learning_rate": 2.8212922140723463e-07, + "loss": 0.073, + "step": 41639 + }, + { + "epoch": 0.9761071748800084, + "grad_norm": 1.384469985961914, + "learning_rate": 2.8157655765941315e-07, + "loss": 0.1281, + "step": 41640 + }, + { + "epoch": 0.9761306164548135, + "grad_norm": 0.8144757151603699, + "learning_rate": 2.810244349865654e-07, + "loss": 0.028, + "step": 41641 + }, + { + "epoch": 0.9761540580296184, + "grad_norm": 2.16477632522583, + "learning_rate": 2.8047285339168894e-07, + "loss": 0.1232, + "step": 41642 + }, + { + "epoch": 0.9761774996044235, + "grad_norm": 2.6148533821105957, + "learning_rate": 2.799218128777592e-07, + "loss": 0.2078, + "step": 41643 + }, + { + "epoch": 0.9762009411792284, + "grad_norm": 2.6885523796081543, + "learning_rate": 2.793713134477738e-07, + "loss": 1.5532, + "step": 41644 + }, + { + "epoch": 0.9762243827540334, + "grad_norm": 0.36205217242240906, + "learning_rate": 2.7882135510473027e-07, + "loss": 0.0323, + "step": 41645 + }, + { + "epoch": 0.9762478243288384, + "grad_norm": 4.102987289428711, + "learning_rate": 2.782719378516041e-07, + "loss": 0.2974, + "step": 41646 + }, + { + "epoch": 0.9762712659036434, + "grad_norm": 1.0991674661636353, + "learning_rate": 2.777230616913817e-07, + "loss": 0.06, + "step": 41647 + }, + { + "epoch": 0.9762947074784484, + "grad_norm": 0.2486228048801422, + "learning_rate": 2.7717472662703856e-07, + "loss": 0.0226, + "step": 41648 + }, + { + "epoch": 0.9763181490532534, + "grad_norm": 1.6102131605148315, + "learning_rate": 2.7662693266155004e-07, + "loss": 0.0696, + "step": 41649 + }, + { + "epoch": 0.9763415906280584, + "grad_norm": 10.02763557434082, + "learning_rate": 2.760796797978804e-07, + "loss": 0.4571, + "step": 41650 + }, + { + "epoch": 0.9763650322028634, + "grad_norm": 3.6016759872436523, + "learning_rate": 2.755329680390162e-07, + "loss": 0.122, + "step": 41651 + }, + { + "epoch": 0.9763884737776684, + "grad_norm": 1.7393906116485596, + "learning_rate": 2.749867973879106e-07, + "loss": 1.0538, + "step": 41652 + }, + { + "epoch": 0.9764119153524734, + "grad_norm": 2.340517520904541, + "learning_rate": 2.7444116784752784e-07, + "loss": 0.1376, + "step": 41653 + }, + { + "epoch": 0.9764353569272783, + "grad_norm": 2.2246906757354736, + "learning_rate": 2.738960794208323e-07, + "loss": 1.4317, + "step": 41654 + }, + { + "epoch": 0.9764587985020834, + "grad_norm": 2.6219756603240967, + "learning_rate": 2.7335153211077713e-07, + "loss": 0.1593, + "step": 41655 + }, + { + "epoch": 0.9764822400768883, + "grad_norm": 1.5037708282470703, + "learning_rate": 2.7280752592032664e-07, + "loss": 0.1225, + "step": 41656 + }, + { + "epoch": 0.9765056816516934, + "grad_norm": 2.646016836166382, + "learning_rate": 2.72264060852423e-07, + "loss": 0.2551, + "step": 41657 + }, + { + "epoch": 0.9765291232264983, + "grad_norm": 3.9177894592285156, + "learning_rate": 2.717211369100192e-07, + "loss": 0.4668, + "step": 41658 + }, + { + "epoch": 0.9765525648013034, + "grad_norm": 1.6117873191833496, + "learning_rate": 2.711787540960575e-07, + "loss": 0.1568, + "step": 41659 + }, + { + "epoch": 0.9765760063761083, + "grad_norm": 1.1813716888427734, + "learning_rate": 2.706369124134911e-07, + "loss": 0.1216, + "step": 41660 + }, + { + "epoch": 0.9765994479509134, + "grad_norm": 9.64221477508545, + "learning_rate": 2.700956118652509e-07, + "loss": 0.4436, + "step": 41661 + }, + { + "epoch": 0.9766228895257183, + "grad_norm": 1.7614802122116089, + "learning_rate": 2.69554852454279e-07, + "loss": 0.11, + "step": 41662 + }, + { + "epoch": 0.9766463311005233, + "grad_norm": 4.125524044036865, + "learning_rate": 2.690146341835065e-07, + "loss": 0.284, + "step": 41663 + }, + { + "epoch": 0.9766697726753283, + "grad_norm": 2.0029873847961426, + "learning_rate": 2.6847495705585314e-07, + "loss": 0.142, + "step": 41664 + }, + { + "epoch": 0.9766932142501333, + "grad_norm": 3.16624116897583, + "learning_rate": 2.679358210742722e-07, + "loss": 0.3841, + "step": 41665 + }, + { + "epoch": 0.9767166558249383, + "grad_norm": 1.1986559629440308, + "learning_rate": 2.6739722624166133e-07, + "loss": 0.0449, + "step": 41666 + }, + { + "epoch": 0.9767400973997433, + "grad_norm": 3.1080217361450195, + "learning_rate": 2.668591725609737e-07, + "loss": 0.1839, + "step": 41667 + }, + { + "epoch": 0.9767635389745483, + "grad_norm": 2.661306381225586, + "learning_rate": 2.6632166003510703e-07, + "loss": 0.2301, + "step": 41668 + }, + { + "epoch": 0.9767869805493533, + "grad_norm": 2.7831835746765137, + "learning_rate": 2.6578468866697015e-07, + "loss": 0.3143, + "step": 41669 + }, + { + "epoch": 0.9768104221241583, + "grad_norm": 2.3976666927337646, + "learning_rate": 2.6524825845950505e-07, + "loss": 0.2974, + "step": 41670 + }, + { + "epoch": 0.9768338636989633, + "grad_norm": 0.32682719826698303, + "learning_rate": 2.647123694155984e-07, + "loss": 0.0142, + "step": 41671 + }, + { + "epoch": 0.9768573052737682, + "grad_norm": 2.0935707092285156, + "learning_rate": 2.6417702153818114e-07, + "loss": 0.0555, + "step": 41672 + }, + { + "epoch": 0.9768807468485733, + "grad_norm": 3.247947931289673, + "learning_rate": 2.6364221483012875e-07, + "loss": 0.1393, + "step": 41673 + }, + { + "epoch": 0.9769041884233782, + "grad_norm": 0.4837140440940857, + "learning_rate": 2.631079492943611e-07, + "loss": 0.0404, + "step": 41674 + }, + { + "epoch": 0.9769276299981833, + "grad_norm": 2.5293359756469727, + "learning_rate": 2.625742249337759e-07, + "loss": 0.2072, + "step": 41675 + }, + { + "epoch": 0.9769510715729883, + "grad_norm": 1.459732174873352, + "learning_rate": 2.6204104175127084e-07, + "loss": 0.0493, + "step": 41676 + }, + { + "epoch": 0.9769745131477933, + "grad_norm": 0.9005033373832703, + "learning_rate": 2.615083997497325e-07, + "loss": 0.0724, + "step": 41677 + }, + { + "epoch": 0.9769979547225983, + "grad_norm": 2.0147132873535156, + "learning_rate": 2.6097629893204744e-07, + "loss": 0.1213, + "step": 41678 + }, + { + "epoch": 0.9770213962974033, + "grad_norm": 3.8912887573242188, + "learning_rate": 2.6044473930112445e-07, + "loss": 0.4299, + "step": 41679 + }, + { + "epoch": 0.9770448378722083, + "grad_norm": 2.6553573608398438, + "learning_rate": 2.599137208598168e-07, + "loss": 0.1224, + "step": 41680 + }, + { + "epoch": 0.9770682794470132, + "grad_norm": 2.4750261306762695, + "learning_rate": 2.5938324361103326e-07, + "loss": 1.4883, + "step": 41681 + }, + { + "epoch": 0.9770917210218183, + "grad_norm": 0.7731393575668335, + "learning_rate": 2.5885330755763826e-07, + "loss": 0.0505, + "step": 41682 + }, + { + "epoch": 0.9771151625966232, + "grad_norm": 0.9101647138595581, + "learning_rate": 2.583239127025072e-07, + "loss": 0.0566, + "step": 41683 + }, + { + "epoch": 0.9771386041714283, + "grad_norm": 2.1836445331573486, + "learning_rate": 2.5779505904851563e-07, + "loss": 0.2259, + "step": 41684 + }, + { + "epoch": 0.9771620457462332, + "grad_norm": 2.213477849960327, + "learning_rate": 2.572667465985279e-07, + "loss": 0.1444, + "step": 41685 + }, + { + "epoch": 0.9771854873210383, + "grad_norm": 1.9343196153640747, + "learning_rate": 2.5673897535541947e-07, + "loss": 0.125, + "step": 41686 + }, + { + "epoch": 0.9772089288958432, + "grad_norm": 0.2566395699977875, + "learning_rate": 2.562117453220547e-07, + "loss": 0.0328, + "step": 41687 + }, + { + "epoch": 0.9772323704706483, + "grad_norm": 1.6129887104034424, + "learning_rate": 2.5568505650128696e-07, + "loss": 0.0867, + "step": 41688 + }, + { + "epoch": 0.9772558120454532, + "grad_norm": 1.7567572593688965, + "learning_rate": 2.551589088959694e-07, + "loss": 0.1633, + "step": 41689 + }, + { + "epoch": 0.9772792536202582, + "grad_norm": 2.441448926925659, + "learning_rate": 2.5463330250896644e-07, + "loss": 0.2062, + "step": 41690 + }, + { + "epoch": 0.9773026951950632, + "grad_norm": 1.677810788154602, + "learning_rate": 2.541082373431314e-07, + "loss": 0.0591, + "step": 41691 + }, + { + "epoch": 0.9773261367698682, + "grad_norm": 2.3182690143585205, + "learning_rate": 2.535837134013175e-07, + "loss": 0.1125, + "step": 41692 + }, + { + "epoch": 0.9773495783446732, + "grad_norm": 0.9016867280006409, + "learning_rate": 2.530597306863447e-07, + "loss": 0.1005, + "step": 41693 + }, + { + "epoch": 0.9773730199194782, + "grad_norm": 4.241342544555664, + "learning_rate": 2.5253628920108853e-07, + "loss": 0.3096, + "step": 41694 + }, + { + "epoch": 0.9773964614942832, + "grad_norm": 3.1963837146759033, + "learning_rate": 2.520133889483689e-07, + "loss": 0.2913, + "step": 41695 + }, + { + "epoch": 0.9774199030690882, + "grad_norm": 2.402104377746582, + "learning_rate": 2.5149102993103914e-07, + "loss": 0.1345, + "step": 41696 + }, + { + "epoch": 0.9774433446438932, + "grad_norm": 5.172398090362549, + "learning_rate": 2.5096921215190803e-07, + "loss": 0.1712, + "step": 41697 + }, + { + "epoch": 0.9774667862186982, + "grad_norm": 0.8610276579856873, + "learning_rate": 2.5044793561384006e-07, + "loss": 0.0696, + "step": 41698 + }, + { + "epoch": 0.9774902277935031, + "grad_norm": 3.3005247116088867, + "learning_rate": 2.499272003196329e-07, + "loss": 0.32, + "step": 41699 + }, + { + "epoch": 0.9775136693683082, + "grad_norm": 1.9509230852127075, + "learning_rate": 2.4940700627212876e-07, + "loss": 0.8342, + "step": 41700 + }, + { + "epoch": 0.9775371109431131, + "grad_norm": 2.629450798034668, + "learning_rate": 2.488873534741476e-07, + "loss": 0.1488, + "step": 41701 + }, + { + "epoch": 0.9775605525179182, + "grad_norm": 4.120776653289795, + "learning_rate": 2.483682419285094e-07, + "loss": 0.3819, + "step": 41702 + }, + { + "epoch": 0.9775839940927231, + "grad_norm": 3.836758852005005, + "learning_rate": 2.4784967163802297e-07, + "loss": 0.2534, + "step": 41703 + }, + { + "epoch": 0.9776074356675282, + "grad_norm": 0.9988082647323608, + "learning_rate": 2.4733164260551945e-07, + "loss": 0.0286, + "step": 41704 + }, + { + "epoch": 0.9776308772423331, + "grad_norm": 0.9303717613220215, + "learning_rate": 2.4681415483379653e-07, + "loss": 0.0517, + "step": 41705 + }, + { + "epoch": 0.9776543188171382, + "grad_norm": 3.223979949951172, + "learning_rate": 2.4629720832566316e-07, + "loss": 0.2302, + "step": 41706 + }, + { + "epoch": 0.9776777603919431, + "grad_norm": 1.1142845153808594, + "learning_rate": 2.457808030839281e-07, + "loss": 0.0648, + "step": 41707 + }, + { + "epoch": 0.9777012019667481, + "grad_norm": 1.9417304992675781, + "learning_rate": 2.4526493911138926e-07, + "loss": 0.1092, + "step": 41708 + }, + { + "epoch": 0.9777246435415531, + "grad_norm": 2.8411872386932373, + "learning_rate": 2.447496164108443e-07, + "loss": 0.1659, + "step": 41709 + }, + { + "epoch": 0.9777480851163581, + "grad_norm": 2.892280101776123, + "learning_rate": 2.442348349850909e-07, + "loss": 0.2428, + "step": 41710 + }, + { + "epoch": 0.9777715266911631, + "grad_norm": 2.951260805130005, + "learning_rate": 2.437205948369381e-07, + "loss": 0.2876, + "step": 41711 + }, + { + "epoch": 0.9777949682659681, + "grad_norm": 4.8824262619018555, + "learning_rate": 2.4320689596915025e-07, + "loss": 0.4377, + "step": 41712 + }, + { + "epoch": 0.9778184098407731, + "grad_norm": 0.3754366934299469, + "learning_rate": 2.4269373838452513e-07, + "loss": 0.0401, + "step": 41713 + }, + { + "epoch": 0.9778418514155781, + "grad_norm": 3.1070146560668945, + "learning_rate": 2.421811220858605e-07, + "loss": 0.1429, + "step": 41714 + }, + { + "epoch": 0.977865292990383, + "grad_norm": 0.6188639998435974, + "learning_rate": 2.4166904707592084e-07, + "loss": 0.0202, + "step": 41715 + }, + { + "epoch": 0.9778887345651881, + "grad_norm": 1.9908121824264526, + "learning_rate": 2.4115751335748173e-07, + "loss": 1.2569, + "step": 41716 + }, + { + "epoch": 0.977912176139993, + "grad_norm": 0.5581676959991455, + "learning_rate": 2.406465209333297e-07, + "loss": 0.0548, + "step": 41717 + }, + { + "epoch": 0.9779356177147981, + "grad_norm": 2.6431593894958496, + "learning_rate": 2.401360698062405e-07, + "loss": 1.3127, + "step": 41718 + }, + { + "epoch": 0.977959059289603, + "grad_norm": 2.2073493003845215, + "learning_rate": 2.396261599789784e-07, + "loss": 0.1574, + "step": 41719 + }, + { + "epoch": 0.9779825008644081, + "grad_norm": 3.3272483348846436, + "learning_rate": 2.3911679145430797e-07, + "loss": 0.3019, + "step": 41720 + }, + { + "epoch": 0.978005942439213, + "grad_norm": 4.01917839050293, + "learning_rate": 2.3860796423499364e-07, + "loss": 0.237, + "step": 41721 + }, + { + "epoch": 0.9780293840140181, + "grad_norm": 0.6869193315505981, + "learning_rate": 2.3809967832379986e-07, + "loss": 0.0295, + "step": 41722 + }, + { + "epoch": 0.978052825588823, + "grad_norm": 4.5997819900512695, + "learning_rate": 2.3759193372347998e-07, + "loss": 0.3369, + "step": 41723 + }, + { + "epoch": 0.978076267163628, + "grad_norm": 2.084228992462158, + "learning_rate": 2.3708473043678737e-07, + "loss": 0.1307, + "step": 41724 + }, + { + "epoch": 0.978099708738433, + "grad_norm": 3.6763358116149902, + "learning_rate": 2.3657806846648645e-07, + "loss": 0.3315, + "step": 41725 + }, + { + "epoch": 0.978123150313238, + "grad_norm": 3.4285776615142822, + "learning_rate": 2.360719478153084e-07, + "loss": 0.2988, + "step": 41726 + }, + { + "epoch": 0.9781465918880431, + "grad_norm": 2.1149609088897705, + "learning_rate": 2.3556636848601765e-07, + "loss": 0.1241, + "step": 41727 + }, + { + "epoch": 0.978170033462848, + "grad_norm": 0.9256049394607544, + "learning_rate": 2.3506133048134537e-07, + "loss": 0.1077, + "step": 41728 + }, + { + "epoch": 0.9781934750376531, + "grad_norm": 3.033433198928833, + "learning_rate": 2.3455683380403383e-07, + "loss": 0.3975, + "step": 41729 + }, + { + "epoch": 0.978216916612458, + "grad_norm": 3.5767009258270264, + "learning_rate": 2.3405287845682522e-07, + "loss": 0.21, + "step": 41730 + }, + { + "epoch": 0.9782403581872631, + "grad_norm": 4.431881904602051, + "learning_rate": 2.3354946444245074e-07, + "loss": 0.2985, + "step": 41731 + }, + { + "epoch": 0.978263799762068, + "grad_norm": 2.705946207046509, + "learning_rate": 2.3304659176364153e-07, + "loss": 0.2486, + "step": 41732 + }, + { + "epoch": 0.9782872413368731, + "grad_norm": 2.6767807006835938, + "learning_rate": 2.325442604231287e-07, + "loss": 0.3078, + "step": 41733 + }, + { + "epoch": 0.978310682911678, + "grad_norm": 7.459668159484863, + "learning_rate": 2.3204247042363235e-07, + "loss": 0.1434, + "step": 41734 + }, + { + "epoch": 0.978334124486483, + "grad_norm": 0.5048977732658386, + "learning_rate": 2.315412217678725e-07, + "loss": 0.0378, + "step": 41735 + }, + { + "epoch": 0.978357566061288, + "grad_norm": 0.8890485763549805, + "learning_rate": 2.3104051445859144e-07, + "loss": 0.0502, + "step": 41736 + }, + { + "epoch": 0.978381007636093, + "grad_norm": 4.107095718383789, + "learning_rate": 2.3054034849847584e-07, + "loss": 0.3703, + "step": 41737 + }, + { + "epoch": 0.978404449210898, + "grad_norm": 2.8958582878112793, + "learning_rate": 2.3004072389025687e-07, + "loss": 1.3492, + "step": 41738 + }, + { + "epoch": 0.978427890785703, + "grad_norm": 2.406951427459717, + "learning_rate": 2.2954164063665463e-07, + "loss": 0.2744, + "step": 41739 + }, + { + "epoch": 0.978451332360508, + "grad_norm": 3.0109481811523438, + "learning_rate": 2.2904309874034468e-07, + "loss": 0.1918, + "step": 41740 + }, + { + "epoch": 0.978474773935313, + "grad_norm": 2.369281053543091, + "learning_rate": 2.2854509820406932e-07, + "loss": 0.1021, + "step": 41741 + }, + { + "epoch": 0.978498215510118, + "grad_norm": 1.656470537185669, + "learning_rate": 2.280476390305042e-07, + "loss": 0.1473, + "step": 41742 + }, + { + "epoch": 0.978521657084923, + "grad_norm": 3.11938738822937, + "learning_rate": 2.275507212223582e-07, + "loss": 0.2634, + "step": 41743 + }, + { + "epoch": 0.9785450986597279, + "grad_norm": 3.222724676132202, + "learning_rate": 2.2705434478232923e-07, + "loss": 0.1971, + "step": 41744 + }, + { + "epoch": 0.978568540234533, + "grad_norm": 3.2977588176727295, + "learning_rate": 2.26558509713104e-07, + "loss": 0.3239, + "step": 41745 + }, + { + "epoch": 0.9785919818093379, + "grad_norm": 0.6042739748954773, + "learning_rate": 2.2606321601738034e-07, + "loss": 0.0586, + "step": 41746 + }, + { + "epoch": 0.978615423384143, + "grad_norm": 2.5318336486816406, + "learning_rate": 2.2556846369784502e-07, + "loss": 0.1383, + "step": 41747 + }, + { + "epoch": 0.9786388649589479, + "grad_norm": 1.1241934299468994, + "learning_rate": 2.2507425275718476e-07, + "loss": 0.0321, + "step": 41748 + }, + { + "epoch": 0.978662306533753, + "grad_norm": 4.491013050079346, + "learning_rate": 2.2458058319807517e-07, + "loss": 0.3699, + "step": 41749 + }, + { + "epoch": 0.9786857481085579, + "grad_norm": 1.7542065382003784, + "learning_rate": 2.2408745502319194e-07, + "loss": 0.1918, + "step": 41750 + }, + { + "epoch": 0.978709189683363, + "grad_norm": 1.066967248916626, + "learning_rate": 2.235948682352218e-07, + "loss": 0.0736, + "step": 41751 + }, + { + "epoch": 0.9787326312581679, + "grad_norm": 1.7765158414840698, + "learning_rate": 2.2310282283682927e-07, + "loss": 0.1108, + "step": 41752 + }, + { + "epoch": 0.978756072832973, + "grad_norm": 2.2976601123809814, + "learning_rate": 2.2261131883067888e-07, + "loss": 0.1625, + "step": 41753 + }, + { + "epoch": 0.9787795144077779, + "grad_norm": 2.0555922985076904, + "learning_rate": 2.2212035621944628e-07, + "loss": 0.982, + "step": 41754 + }, + { + "epoch": 0.9788029559825829, + "grad_norm": 3.641842842102051, + "learning_rate": 2.21629935005796e-07, + "loss": 0.311, + "step": 41755 + }, + { + "epoch": 0.9788263975573879, + "grad_norm": 2.9673895835876465, + "learning_rate": 2.2114005519239257e-07, + "loss": 0.3098, + "step": 41756 + }, + { + "epoch": 0.9788498391321929, + "grad_norm": 2.774672269821167, + "learning_rate": 2.2065071678188943e-07, + "loss": 0.2261, + "step": 41757 + }, + { + "epoch": 0.9788732807069979, + "grad_norm": 0.8528953194618225, + "learning_rate": 2.2016191977694e-07, + "loss": 0.0627, + "step": 41758 + }, + { + "epoch": 0.9788967222818029, + "grad_norm": 2.035187244415283, + "learning_rate": 2.1967366418019775e-07, + "loss": 0.1411, + "step": 41759 + }, + { + "epoch": 0.9789201638566079, + "grad_norm": 3.5576131343841553, + "learning_rate": 2.1918594999430497e-07, + "loss": 0.2761, + "step": 41760 + }, + { + "epoch": 0.9789436054314129, + "grad_norm": 0.2894681692123413, + "learning_rate": 2.1869877722191512e-07, + "loss": 0.0098, + "step": 41761 + }, + { + "epoch": 0.9789670470062178, + "grad_norm": 2.6079165935516357, + "learning_rate": 2.1821214586568161e-07, + "loss": 0.2049, + "step": 41762 + }, + { + "epoch": 0.9789904885810229, + "grad_norm": 1.573696494102478, + "learning_rate": 2.177260559282246e-07, + "loss": 0.0521, + "step": 41763 + }, + { + "epoch": 0.9790139301558278, + "grad_norm": 2.011732816696167, + "learning_rate": 2.1724050741219748e-07, + "loss": 0.1338, + "step": 41764 + }, + { + "epoch": 0.9790373717306329, + "grad_norm": 3.704620122909546, + "learning_rate": 2.167555003202315e-07, + "loss": 0.1957, + "step": 41765 + }, + { + "epoch": 0.9790608133054378, + "grad_norm": 2.5711567401885986, + "learning_rate": 2.162710346549468e-07, + "loss": 0.1833, + "step": 41766 + }, + { + "epoch": 0.9790842548802429, + "grad_norm": 0.9183083176612854, + "learning_rate": 2.157871104189857e-07, + "loss": 0.0472, + "step": 41767 + }, + { + "epoch": 0.9791076964550478, + "grad_norm": 3.033536911010742, + "learning_rate": 2.153037276149683e-07, + "loss": 0.3124, + "step": 41768 + }, + { + "epoch": 0.9791311380298529, + "grad_norm": 2.7907373905181885, + "learning_rate": 2.1482088624551477e-07, + "loss": 1.7686, + "step": 41769 + }, + { + "epoch": 0.9791545796046578, + "grad_norm": 2.853839874267578, + "learning_rate": 2.1433858631325632e-07, + "loss": 0.2108, + "step": 41770 + }, + { + "epoch": 0.9791780211794628, + "grad_norm": 2.407217025756836, + "learning_rate": 2.1385682782080197e-07, + "loss": 1.5902, + "step": 41771 + }, + { + "epoch": 0.9792014627542678, + "grad_norm": 0.9686439037322998, + "learning_rate": 2.1337561077076073e-07, + "loss": 0.066, + "step": 41772 + }, + { + "epoch": 0.9792249043290728, + "grad_norm": 3.067108631134033, + "learning_rate": 2.1289493516575276e-07, + "loss": 0.1701, + "step": 41773 + }, + { + "epoch": 0.9792483459038778, + "grad_norm": 1.6540154218673706, + "learning_rate": 2.1241480100837597e-07, + "loss": 0.0887, + "step": 41774 + }, + { + "epoch": 0.9792717874786828, + "grad_norm": 1.3976496458053589, + "learning_rate": 2.1193520830125046e-07, + "loss": 0.0982, + "step": 41775 + }, + { + "epoch": 0.9792952290534878, + "grad_norm": 3.5985748767852783, + "learning_rate": 2.114561570469742e-07, + "loss": 1.3981, + "step": 41776 + }, + { + "epoch": 0.9793186706282928, + "grad_norm": 0.6222305297851562, + "learning_rate": 2.1097764724813395e-07, + "loss": 0.0433, + "step": 41777 + }, + { + "epoch": 0.9793421122030979, + "grad_norm": 1.3244611024856567, + "learning_rate": 2.104996789073388e-07, + "loss": 0.6022, + "step": 41778 + }, + { + "epoch": 0.9793655537779028, + "grad_norm": 2.666046619415283, + "learning_rate": 2.100222520271755e-07, + "loss": 0.2125, + "step": 41779 + }, + { + "epoch": 0.9793889953527078, + "grad_norm": 2.627610445022583, + "learning_rate": 2.0954536661024205e-07, + "loss": 0.1146, + "step": 41780 + }, + { + "epoch": 0.9794124369275128, + "grad_norm": 1.187923550605774, + "learning_rate": 2.0906902265911409e-07, + "loss": 0.1921, + "step": 41781 + }, + { + "epoch": 0.9794358785023178, + "grad_norm": 1.371445894241333, + "learning_rate": 2.085932201763896e-07, + "loss": 0.0723, + "step": 41782 + }, + { + "epoch": 0.9794593200771228, + "grad_norm": 1.139054775238037, + "learning_rate": 2.0811795916464427e-07, + "loss": 0.0723, + "step": 41783 + }, + { + "epoch": 0.9794827616519278, + "grad_norm": 1.1356947422027588, + "learning_rate": 2.076432396264538e-07, + "loss": 0.1043, + "step": 41784 + }, + { + "epoch": 0.9795062032267328, + "grad_norm": 1.3889847993850708, + "learning_rate": 2.0716906156439396e-07, + "loss": 0.0552, + "step": 41785 + }, + { + "epoch": 0.9795296448015378, + "grad_norm": 2.4267983436584473, + "learning_rate": 2.0669542498104043e-07, + "loss": 0.2087, + "step": 41786 + }, + { + "epoch": 0.9795530863763428, + "grad_norm": 1.581972360610962, + "learning_rate": 2.0622232987896896e-07, + "loss": 0.0589, + "step": 41787 + }, + { + "epoch": 0.9795765279511478, + "grad_norm": 2.981205463409424, + "learning_rate": 2.057497762607441e-07, + "loss": 0.0825, + "step": 41788 + }, + { + "epoch": 0.9795999695259527, + "grad_norm": 2.6257224082946777, + "learning_rate": 2.052777641289194e-07, + "loss": 0.0899, + "step": 41789 + }, + { + "epoch": 0.9796234111007578, + "grad_norm": 3.2052087783813477, + "learning_rate": 2.0480629348607062e-07, + "loss": 0.2121, + "step": 41790 + }, + { + "epoch": 0.9796468526755627, + "grad_norm": 2.3957064151763916, + "learning_rate": 2.043353643347401e-07, + "loss": 0.3285, + "step": 41791 + }, + { + "epoch": 0.9796702942503678, + "grad_norm": 2.8940157890319824, + "learning_rate": 2.038649766774925e-07, + "loss": 0.2008, + "step": 41792 + }, + { + "epoch": 0.9796937358251727, + "grad_norm": 2.572340726852417, + "learning_rate": 2.033951305168924e-07, + "loss": 0.1904, + "step": 41793 + }, + { + "epoch": 0.9797171773999778, + "grad_norm": 2.393360137939453, + "learning_rate": 2.0292582585546005e-07, + "loss": 0.8345, + "step": 41794 + }, + { + "epoch": 0.9797406189747827, + "grad_norm": 2.135303497314453, + "learning_rate": 2.0245706269577114e-07, + "loss": 0.0895, + "step": 41795 + }, + { + "epoch": 0.9797640605495878, + "grad_norm": 0.9915403723716736, + "learning_rate": 2.0198884104034588e-07, + "loss": 0.081, + "step": 41796 + }, + { + "epoch": 0.9797875021243927, + "grad_norm": 2.423626184463501, + "learning_rate": 2.015211608917489e-07, + "loss": 0.1712, + "step": 41797 + }, + { + "epoch": 0.9798109436991977, + "grad_norm": 1.5649906396865845, + "learning_rate": 2.0105402225248927e-07, + "loss": 0.1561, + "step": 41798 + }, + { + "epoch": 0.9798343852740027, + "grad_norm": 3.7502851486206055, + "learning_rate": 2.0058742512512053e-07, + "loss": 0.2742, + "step": 41799 + }, + { + "epoch": 0.9798578268488077, + "grad_norm": 2.9569170475006104, + "learning_rate": 2.0012136951217397e-07, + "loss": 0.1944, + "step": 41800 + }, + { + "epoch": 0.9798812684236127, + "grad_norm": 1.016549825668335, + "learning_rate": 1.9965585541618093e-07, + "loss": 0.095, + "step": 41801 + }, + { + "epoch": 0.9799047099984177, + "grad_norm": 3.410564422607422, + "learning_rate": 1.991908828396505e-07, + "loss": 0.1422, + "step": 41802 + }, + { + "epoch": 0.9799281515732227, + "grad_norm": 2.4796268939971924, + "learning_rate": 1.9872645178512505e-07, + "loss": 0.2816, + "step": 41803 + }, + { + "epoch": 0.9799515931480277, + "grad_norm": 2.2688117027282715, + "learning_rate": 1.9826256225510264e-07, + "loss": 0.1908, + "step": 41804 + }, + { + "epoch": 0.9799750347228326, + "grad_norm": 4.811615467071533, + "learning_rate": 1.9779921425212568e-07, + "loss": 0.2009, + "step": 41805 + }, + { + "epoch": 0.9799984762976377, + "grad_norm": 2.962040662765503, + "learning_rate": 1.9733640777869211e-07, + "loss": 0.2439, + "step": 41806 + }, + { + "epoch": 0.9800219178724426, + "grad_norm": 2.8515467643737793, + "learning_rate": 1.968741428373222e-07, + "loss": 0.2089, + "step": 41807 + }, + { + "epoch": 0.9800453594472477, + "grad_norm": 1.2171645164489746, + "learning_rate": 1.9641241943051393e-07, + "loss": 0.0755, + "step": 41808 + }, + { + "epoch": 0.9800688010220526, + "grad_norm": 1.3301584720611572, + "learning_rate": 1.9595123756077638e-07, + "loss": 0.0371, + "step": 41809 + }, + { + "epoch": 0.9800922425968577, + "grad_norm": 1.1542818546295166, + "learning_rate": 1.9549059723061868e-07, + "loss": 0.1185, + "step": 41810 + }, + { + "epoch": 0.9801156841716626, + "grad_norm": 0.8100607395172119, + "learning_rate": 1.9503049844253885e-07, + "loss": 0.0543, + "step": 41811 + }, + { + "epoch": 0.9801391257464677, + "grad_norm": 0.43455439805984497, + "learning_rate": 1.9457094119902376e-07, + "loss": 0.025, + "step": 41812 + }, + { + "epoch": 0.9801625673212726, + "grad_norm": 3.895029306411743, + "learning_rate": 1.941119255025714e-07, + "loss": 0.4307, + "step": 41813 + }, + { + "epoch": 0.9801860088960777, + "grad_norm": 1.8615782260894775, + "learning_rate": 1.9365345135567982e-07, + "loss": 0.2292, + "step": 41814 + }, + { + "epoch": 0.9802094504708826, + "grad_norm": 2.235929489135742, + "learning_rate": 1.9319551876082474e-07, + "loss": 0.25, + "step": 41815 + }, + { + "epoch": 0.9802328920456876, + "grad_norm": 0.6552281975746155, + "learning_rate": 1.9273812772050425e-07, + "loss": 0.0239, + "step": 41816 + }, + { + "epoch": 0.9802563336204926, + "grad_norm": 1.7421749830245972, + "learning_rate": 1.92281278237183e-07, + "loss": 0.0507, + "step": 41817 + }, + { + "epoch": 0.9802797751952976, + "grad_norm": 1.8670204877853394, + "learning_rate": 1.91824970313359e-07, + "loss": 0.093, + "step": 41818 + }, + { + "epoch": 0.9803032167701026, + "grad_norm": 2.176480531692505, + "learning_rate": 1.9136920395149693e-07, + "loss": 0.9223, + "step": 41819 + }, + { + "epoch": 0.9803266583449076, + "grad_norm": 0.7233095169067383, + "learning_rate": 1.909139791540726e-07, + "loss": 0.0785, + "step": 41820 + }, + { + "epoch": 0.9803500999197126, + "grad_norm": 0.691533625125885, + "learning_rate": 1.904592959235507e-07, + "loss": 0.0462, + "step": 41821 + }, + { + "epoch": 0.9803735414945176, + "grad_norm": 8.545406341552734, + "learning_rate": 1.9000515426240706e-07, + "loss": 0.5122, + "step": 41822 + }, + { + "epoch": 0.9803969830693225, + "grad_norm": 3.7459356784820557, + "learning_rate": 1.8955155417309523e-07, + "loss": 0.2376, + "step": 41823 + }, + { + "epoch": 0.9804204246441276, + "grad_norm": 2.9237866401672363, + "learning_rate": 1.89098495658091e-07, + "loss": 0.2069, + "step": 41824 + }, + { + "epoch": 0.9804438662189325, + "grad_norm": 5.0022501945495605, + "learning_rate": 1.886459787198369e-07, + "loss": 0.4053, + "step": 41825 + }, + { + "epoch": 0.9804673077937376, + "grad_norm": 1.903289794921875, + "learning_rate": 1.8819400336080873e-07, + "loss": 0.0384, + "step": 41826 + }, + { + "epoch": 0.9804907493685425, + "grad_norm": 2.7982349395751953, + "learning_rate": 1.8774256958343784e-07, + "loss": 0.2658, + "step": 41827 + }, + { + "epoch": 0.9805141909433476, + "grad_norm": 2.8566954135894775, + "learning_rate": 1.8729167739017784e-07, + "loss": 0.1393, + "step": 41828 + }, + { + "epoch": 0.9805376325181526, + "grad_norm": 3.783371925354004, + "learning_rate": 1.8684132678349342e-07, + "loss": 0.2765, + "step": 41829 + }, + { + "epoch": 0.9805610740929576, + "grad_norm": 0.5435754060745239, + "learning_rate": 1.8639151776580488e-07, + "loss": 0.0338, + "step": 41830 + }, + { + "epoch": 0.9805845156677626, + "grad_norm": 3.168984889984131, + "learning_rate": 1.859422503395658e-07, + "loss": 0.3351, + "step": 41831 + }, + { + "epoch": 0.9806079572425676, + "grad_norm": 1.9765619039535522, + "learning_rate": 1.854935245072076e-07, + "loss": 0.187, + "step": 41832 + }, + { + "epoch": 0.9806313988173726, + "grad_norm": 0.9937213063240051, + "learning_rate": 1.850453402711616e-07, + "loss": 0.049, + "step": 41833 + }, + { + "epoch": 0.9806548403921775, + "grad_norm": 0.8963032364845276, + "learning_rate": 1.8459769763388146e-07, + "loss": 0.0695, + "step": 41834 + }, + { + "epoch": 0.9806782819669826, + "grad_norm": 2.1256158351898193, + "learning_rate": 1.8415059659777633e-07, + "loss": 0.1218, + "step": 41835 + }, + { + "epoch": 0.9807017235417875, + "grad_norm": 1.560166597366333, + "learning_rate": 1.8370403716526652e-07, + "loss": 0.0961, + "step": 41836 + }, + { + "epoch": 0.9807251651165926, + "grad_norm": 1.7079081535339355, + "learning_rate": 1.832580193387945e-07, + "loss": 0.0893, + "step": 41837 + }, + { + "epoch": 0.9807486066913975, + "grad_norm": 2.2920289039611816, + "learning_rate": 1.8281254312076945e-07, + "loss": 0.2054, + "step": 41838 + }, + { + "epoch": 0.9807720482662026, + "grad_norm": 3.1583220958709717, + "learning_rate": 1.8236760851360056e-07, + "loss": 0.2854, + "step": 41839 + }, + { + "epoch": 0.9807954898410075, + "grad_norm": 0.3867858052253723, + "learning_rate": 1.8192321551973036e-07, + "loss": 0.0205, + "step": 41840 + }, + { + "epoch": 0.9808189314158126, + "grad_norm": 2.4374451637268066, + "learning_rate": 1.8147936414153466e-07, + "loss": 0.084, + "step": 41841 + }, + { + "epoch": 0.9808423729906175, + "grad_norm": 2.9324655532836914, + "learning_rate": 1.810360543814449e-07, + "loss": 0.2128, + "step": 41842 + }, + { + "epoch": 0.9808658145654225, + "grad_norm": 1.5532021522521973, + "learning_rate": 1.8059328624185913e-07, + "loss": 0.0896, + "step": 41843 + }, + { + "epoch": 0.9808892561402275, + "grad_norm": 1.7554172277450562, + "learning_rate": 1.8015105972518653e-07, + "loss": 0.1663, + "step": 41844 + }, + { + "epoch": 0.9809126977150325, + "grad_norm": 1.856431007385254, + "learning_rate": 1.7970937483381412e-07, + "loss": 1.2397, + "step": 41845 + }, + { + "epoch": 0.9809361392898375, + "grad_norm": 2.9174604415893555, + "learning_rate": 1.7926823157015104e-07, + "loss": 0.2926, + "step": 41846 + }, + { + "epoch": 0.9809595808646425, + "grad_norm": 3.348932981491089, + "learning_rate": 1.788276299365843e-07, + "loss": 0.2502, + "step": 41847 + }, + { + "epoch": 0.9809830224394475, + "grad_norm": 0.34654027223587036, + "learning_rate": 1.7838756993550087e-07, + "loss": 0.0335, + "step": 41848 + }, + { + "epoch": 0.9810064640142525, + "grad_norm": 0.8050699830055237, + "learning_rate": 1.7794805156929884e-07, + "loss": 0.0526, + "step": 41849 + }, + { + "epoch": 0.9810299055890574, + "grad_norm": 6.020442008972168, + "learning_rate": 1.7750907484035407e-07, + "loss": 0.3184, + "step": 41850 + }, + { + "epoch": 0.9810533471638625, + "grad_norm": 0.4298267364501953, + "learning_rate": 1.7707063975105353e-07, + "loss": 0.0182, + "step": 41851 + }, + { + "epoch": 0.9810767887386674, + "grad_norm": 4.006765365600586, + "learning_rate": 1.7663274630377314e-07, + "loss": 0.3697, + "step": 41852 + }, + { + "epoch": 0.9811002303134725, + "grad_norm": 0.8890538215637207, + "learning_rate": 1.7619539450088872e-07, + "loss": 0.0553, + "step": 41853 + }, + { + "epoch": 0.9811236718882774, + "grad_norm": 2.0033702850341797, + "learning_rate": 1.7575858434477622e-07, + "loss": 0.2624, + "step": 41854 + }, + { + "epoch": 0.9811471134630825, + "grad_norm": 1.694640874862671, + "learning_rate": 1.7532231583781145e-07, + "loss": 0.0743, + "step": 41855 + }, + { + "epoch": 0.9811705550378874, + "grad_norm": 1.6509954929351807, + "learning_rate": 1.74886588982337e-07, + "loss": 0.1222, + "step": 41856 + }, + { + "epoch": 0.9811939966126925, + "grad_norm": 3.7981927394866943, + "learning_rate": 1.74451403780751e-07, + "loss": 0.4051, + "step": 41857 + }, + { + "epoch": 0.9812174381874974, + "grad_norm": 1.7840687036514282, + "learning_rate": 1.7401676023539593e-07, + "loss": 0.1544, + "step": 41858 + }, + { + "epoch": 0.9812408797623025, + "grad_norm": 2.1616780757904053, + "learning_rate": 1.7358265834862553e-07, + "loss": 1.5956, + "step": 41859 + }, + { + "epoch": 0.9812643213371074, + "grad_norm": 1.814115047454834, + "learning_rate": 1.7314909812280455e-07, + "loss": 0.1176, + "step": 41860 + }, + { + "epoch": 0.9812877629119124, + "grad_norm": 2.9053456783294678, + "learning_rate": 1.727160795602867e-07, + "loss": 0.1896, + "step": 41861 + }, + { + "epoch": 0.9813112044867174, + "grad_norm": 2.933046817779541, + "learning_rate": 1.7228360266342558e-07, + "loss": 0.2781, + "step": 41862 + }, + { + "epoch": 0.9813346460615224, + "grad_norm": 2.3967740535736084, + "learning_rate": 1.7185166743454162e-07, + "loss": 0.1231, + "step": 41863 + }, + { + "epoch": 0.9813580876363274, + "grad_norm": 3.568186044692993, + "learning_rate": 1.7142027387601068e-07, + "loss": 0.2822, + "step": 41864 + }, + { + "epoch": 0.9813815292111324, + "grad_norm": 1.754014492034912, + "learning_rate": 1.7098942199016422e-07, + "loss": 0.0889, + "step": 41865 + }, + { + "epoch": 0.9814049707859374, + "grad_norm": 4.9563188552856445, + "learning_rate": 1.705591117793226e-07, + "loss": 0.4847, + "step": 41866 + }, + { + "epoch": 0.9814284123607424, + "grad_norm": 3.8121867179870605, + "learning_rate": 1.7012934324585063e-07, + "loss": 0.3363, + "step": 41867 + }, + { + "epoch": 0.9814518539355473, + "grad_norm": 1.9990559816360474, + "learning_rate": 1.697001163920464e-07, + "loss": 1.0709, + "step": 41868 + }, + { + "epoch": 0.9814752955103524, + "grad_norm": 1.7880903482437134, + "learning_rate": 1.6927143122026367e-07, + "loss": 0.6464, + "step": 41869 + }, + { + "epoch": 0.9814987370851573, + "grad_norm": 1.7450017929077148, + "learning_rate": 1.6884328773282276e-07, + "loss": 0.5591, + "step": 41870 + }, + { + "epoch": 0.9815221786599624, + "grad_norm": 3.238687753677368, + "learning_rate": 1.6841568593204405e-07, + "loss": 0.1269, + "step": 41871 + }, + { + "epoch": 0.9815456202347673, + "grad_norm": 0.4240644574165344, + "learning_rate": 1.6798862582025898e-07, + "loss": 0.0094, + "step": 41872 + }, + { + "epoch": 0.9815690618095724, + "grad_norm": 3.2814507484436035, + "learning_rate": 1.6756210739975465e-07, + "loss": 0.3388, + "step": 41873 + }, + { + "epoch": 0.9815925033843773, + "grad_norm": 2.3345165252685547, + "learning_rate": 1.671361306728736e-07, + "loss": 0.2374, + "step": 41874 + }, + { + "epoch": 0.9816159449591824, + "grad_norm": 0.7693871259689331, + "learning_rate": 1.667106956419251e-07, + "loss": 0.0707, + "step": 41875 + }, + { + "epoch": 0.9816393865339873, + "grad_norm": 3.614384889602661, + "learning_rate": 1.6628580230920732e-07, + "loss": 0.4776, + "step": 41876 + }, + { + "epoch": 0.9816628281087924, + "grad_norm": 3.313530445098877, + "learning_rate": 1.6586145067702951e-07, + "loss": 0.3004, + "step": 41877 + }, + { + "epoch": 0.9816862696835973, + "grad_norm": 1.8198378086090088, + "learning_rate": 1.6543764074770096e-07, + "loss": 0.2157, + "step": 41878 + }, + { + "epoch": 0.9817097112584023, + "grad_norm": 1.4643824100494385, + "learning_rate": 1.650143725235087e-07, + "loss": 0.1534, + "step": 41879 + }, + { + "epoch": 0.9817331528332074, + "grad_norm": 6.135099411010742, + "learning_rate": 1.6459164600675091e-07, + "loss": 0.3934, + "step": 41880 + }, + { + "epoch": 0.9817565944080123, + "grad_norm": 1.0430269241333008, + "learning_rate": 1.6416946119972576e-07, + "loss": 0.0634, + "step": 41881 + }, + { + "epoch": 0.9817800359828174, + "grad_norm": 1.5053486824035645, + "learning_rate": 1.6374781810473138e-07, + "loss": 0.0947, + "step": 41882 + }, + { + "epoch": 0.9818034775576223, + "grad_norm": 0.566524088382721, + "learning_rate": 1.6332671672404376e-07, + "loss": 0.0523, + "step": 41883 + }, + { + "epoch": 0.9818269191324274, + "grad_norm": 1.5790590047836304, + "learning_rate": 1.6290615705996104e-07, + "loss": 0.1249, + "step": 41884 + }, + { + "epoch": 0.9818503607072323, + "grad_norm": 1.9499506950378418, + "learning_rate": 1.6248613911474807e-07, + "loss": 0.1516, + "step": 41885 + }, + { + "epoch": 0.9818738022820374, + "grad_norm": 2.2588860988616943, + "learning_rate": 1.6206666289070304e-07, + "loss": 0.1294, + "step": 41886 + }, + { + "epoch": 0.9818972438568423, + "grad_norm": 2.8900017738342285, + "learning_rate": 1.616477283900797e-07, + "loss": 0.0866, + "step": 41887 + }, + { + "epoch": 0.9819206854316473, + "grad_norm": 2.9991447925567627, + "learning_rate": 1.6122933561516505e-07, + "loss": 0.2064, + "step": 41888 + }, + { + "epoch": 0.9819441270064523, + "grad_norm": 3.109586238861084, + "learning_rate": 1.6081148456822404e-07, + "loss": 0.2338, + "step": 41889 + }, + { + "epoch": 0.9819675685812573, + "grad_norm": 3.7365198135375977, + "learning_rate": 1.6039417525153255e-07, + "loss": 0.2843, + "step": 41890 + }, + { + "epoch": 0.9819910101560623, + "grad_norm": 1.4125049114227295, + "learning_rate": 1.5997740766735547e-07, + "loss": 0.121, + "step": 41891 + }, + { + "epoch": 0.9820144517308673, + "grad_norm": 2.989609479904175, + "learning_rate": 1.5956118181793544e-07, + "loss": 0.1956, + "step": 41892 + }, + { + "epoch": 0.9820378933056723, + "grad_norm": 1.3937351703643799, + "learning_rate": 1.5914549770554842e-07, + "loss": 0.8109, + "step": 41893 + }, + { + "epoch": 0.9820613348804773, + "grad_norm": 3.1104488372802734, + "learning_rate": 1.5873035533244819e-07, + "loss": 0.2634, + "step": 41894 + }, + { + "epoch": 0.9820847764552822, + "grad_norm": 1.9223395586013794, + "learning_rate": 1.5831575470088843e-07, + "loss": 0.1866, + "step": 41895 + }, + { + "epoch": 0.9821082180300873, + "grad_norm": 1.787859559059143, + "learning_rate": 1.5790169581311188e-07, + "loss": 0.1384, + "step": 41896 + }, + { + "epoch": 0.9821316596048922, + "grad_norm": 2.2709920406341553, + "learning_rate": 1.5748817867136112e-07, + "loss": 0.1177, + "step": 41897 + }, + { + "epoch": 0.9821551011796973, + "grad_norm": 0.37101760506629944, + "learning_rate": 1.5707520327788993e-07, + "loss": 0.0134, + "step": 41898 + }, + { + "epoch": 0.9821785427545022, + "grad_norm": 4.8247785568237305, + "learning_rate": 1.5666276963494097e-07, + "loss": 0.588, + "step": 41899 + }, + { + "epoch": 0.9822019843293073, + "grad_norm": 1.913043737411499, + "learning_rate": 1.5625087774473468e-07, + "loss": 0.1797, + "step": 41900 + }, + { + "epoch": 0.9822254259041122, + "grad_norm": 4.6542649269104, + "learning_rate": 1.558395276095359e-07, + "loss": 0.3001, + "step": 41901 + }, + { + "epoch": 0.9822488674789173, + "grad_norm": 3.483933448791504, + "learning_rate": 1.554287192315429e-07, + "loss": 0.3598, + "step": 41902 + }, + { + "epoch": 0.9822723090537222, + "grad_norm": 4.40418004989624, + "learning_rate": 1.550184526130094e-07, + "loss": 0.1691, + "step": 41903 + }, + { + "epoch": 0.9822957506285273, + "grad_norm": 1.0446425676345825, + "learning_rate": 1.5460872775615586e-07, + "loss": 0.1139, + "step": 41904 + }, + { + "epoch": 0.9823191922033322, + "grad_norm": 1.9356600046157837, + "learning_rate": 1.5419954466320274e-07, + "loss": 0.0876, + "step": 41905 + }, + { + "epoch": 0.9823426337781372, + "grad_norm": 3.379594087600708, + "learning_rate": 1.5379090333635937e-07, + "loss": 0.2794, + "step": 41906 + }, + { + "epoch": 0.9823660753529422, + "grad_norm": 2.802706003189087, + "learning_rate": 1.533828037778684e-07, + "loss": 0.1816, + "step": 41907 + }, + { + "epoch": 0.9823895169277472, + "grad_norm": 2.336930513381958, + "learning_rate": 1.5297524598991697e-07, + "loss": 0.133, + "step": 41908 + }, + { + "epoch": 0.9824129585025522, + "grad_norm": 0.5248034000396729, + "learning_rate": 1.5256822997473664e-07, + "loss": 0.0321, + "step": 41909 + }, + { + "epoch": 0.9824364000773572, + "grad_norm": 2.4527246952056885, + "learning_rate": 1.5216175573453672e-07, + "loss": 0.1624, + "step": 41910 + }, + { + "epoch": 0.9824598416521622, + "grad_norm": 2.0234382152557373, + "learning_rate": 1.5175582327150439e-07, + "loss": 0.1853, + "step": 41911 + }, + { + "epoch": 0.9824832832269672, + "grad_norm": 0.8230023384094238, + "learning_rate": 1.5135043258784899e-07, + "loss": 0.0928, + "step": 41912 + }, + { + "epoch": 0.9825067248017721, + "grad_norm": 3.0022084712982178, + "learning_rate": 1.5094558368577982e-07, + "loss": 0.2238, + "step": 41913 + }, + { + "epoch": 0.9825301663765772, + "grad_norm": 2.321112871170044, + "learning_rate": 1.5054127656748406e-07, + "loss": 0.1611, + "step": 41914 + }, + { + "epoch": 0.9825536079513821, + "grad_norm": 2.100057601928711, + "learning_rate": 1.5013751123515995e-07, + "loss": 0.2043, + "step": 41915 + }, + { + "epoch": 0.9825770495261872, + "grad_norm": 3.8781938552856445, + "learning_rate": 1.4973428769099462e-07, + "loss": 0.3984, + "step": 41916 + }, + { + "epoch": 0.9826004911009921, + "grad_norm": 4.583860397338867, + "learning_rate": 1.4933160593718632e-07, + "loss": 0.2633, + "step": 41917 + }, + { + "epoch": 0.9826239326757972, + "grad_norm": 1.0766944885253906, + "learning_rate": 1.4892946597589997e-07, + "loss": 0.0833, + "step": 41918 + }, + { + "epoch": 0.9826473742506021, + "grad_norm": 2.8434581756591797, + "learning_rate": 1.4852786780934492e-07, + "loss": 0.2117, + "step": 41919 + }, + { + "epoch": 0.9826708158254072, + "grad_norm": 1.2850902080535889, + "learning_rate": 1.48126811439675e-07, + "loss": 0.0716, + "step": 41920 + }, + { + "epoch": 0.9826942574002121, + "grad_norm": 1.2166980504989624, + "learning_rate": 1.4772629686907735e-07, + "loss": 0.0792, + "step": 41921 + }, + { + "epoch": 0.9827176989750172, + "grad_norm": 0.9714139103889465, + "learning_rate": 1.47326324099728e-07, + "loss": 0.0826, + "step": 41922 + }, + { + "epoch": 0.9827411405498221, + "grad_norm": 0.4533940553665161, + "learning_rate": 1.4692689313379194e-07, + "loss": 0.0239, + "step": 41923 + }, + { + "epoch": 0.9827645821246271, + "grad_norm": 1.2482106685638428, + "learning_rate": 1.4652800397343402e-07, + "loss": 0.0693, + "step": 41924 + }, + { + "epoch": 0.9827880236994321, + "grad_norm": 3.217890739440918, + "learning_rate": 1.4612965662083033e-07, + "loss": 0.1647, + "step": 41925 + }, + { + "epoch": 0.9828114652742371, + "grad_norm": 1.1039981842041016, + "learning_rate": 1.4573185107812358e-07, + "loss": 0.1118, + "step": 41926 + }, + { + "epoch": 0.9828349068490421, + "grad_norm": 0.4778010845184326, + "learning_rate": 1.4533458734750094e-07, + "loss": 0.0318, + "step": 41927 + }, + { + "epoch": 0.9828583484238471, + "grad_norm": 3.7728359699249268, + "learning_rate": 1.449378654310829e-07, + "loss": 1.3665, + "step": 41928 + }, + { + "epoch": 0.9828817899986521, + "grad_norm": 0.6423256993293762, + "learning_rate": 1.445416853310566e-07, + "loss": 0.0243, + "step": 41929 + }, + { + "epoch": 0.9829052315734571, + "grad_norm": 3.9977006912231445, + "learning_rate": 1.441460470495426e-07, + "loss": 0.165, + "step": 41930 + }, + { + "epoch": 0.9829286731482622, + "grad_norm": 2.831143379211426, + "learning_rate": 1.437509505887058e-07, + "loss": 0.3061, + "step": 41931 + }, + { + "epoch": 0.9829521147230671, + "grad_norm": 4.161294937133789, + "learning_rate": 1.4335639595067784e-07, + "loss": 0.3264, + "step": 41932 + }, + { + "epoch": 0.9829755562978721, + "grad_norm": 2.269569158554077, + "learning_rate": 1.4296238313761257e-07, + "loss": 0.0806, + "step": 41933 + }, + { + "epoch": 0.9829989978726771, + "grad_norm": 3.0042619705200195, + "learning_rate": 1.4256891215163047e-07, + "loss": 0.341, + "step": 41934 + }, + { + "epoch": 0.9830224394474821, + "grad_norm": 1.7032448053359985, + "learning_rate": 1.4217598299488544e-07, + "loss": 0.0976, + "step": 41935 + }, + { + "epoch": 0.9830458810222871, + "grad_norm": 1.1004093885421753, + "learning_rate": 1.4178359566949795e-07, + "loss": 0.076, + "step": 41936 + }, + { + "epoch": 0.9830693225970921, + "grad_norm": 2.20292592048645, + "learning_rate": 1.4139175017761076e-07, + "loss": 0.1496, + "step": 41937 + }, + { + "epoch": 0.9830927641718971, + "grad_norm": 0.3940734267234802, + "learning_rate": 1.4100044652132216e-07, + "loss": 0.0184, + "step": 41938 + }, + { + "epoch": 0.9831162057467021, + "grad_norm": 1.7438584566116333, + "learning_rate": 1.4060968470278602e-07, + "loss": 1.4045, + "step": 41939 + }, + { + "epoch": 0.983139647321507, + "grad_norm": 3.5994739532470703, + "learning_rate": 1.4021946472411175e-07, + "loss": 0.3536, + "step": 41940 + }, + { + "epoch": 0.9831630888963121, + "grad_norm": 3.828037977218628, + "learning_rate": 1.3982978658740874e-07, + "loss": 0.4249, + "step": 41941 + }, + { + "epoch": 0.983186530471117, + "grad_norm": 2.1413538455963135, + "learning_rate": 1.3944065029479757e-07, + "loss": 0.0931, + "step": 41942 + }, + { + "epoch": 0.9832099720459221, + "grad_norm": 1.8934054374694824, + "learning_rate": 1.3905205584838765e-07, + "loss": 0.0785, + "step": 41943 + }, + { + "epoch": 0.983233413620727, + "grad_norm": 2.453279733657837, + "learning_rate": 1.3866400325028838e-07, + "loss": 0.217, + "step": 41944 + }, + { + "epoch": 0.9832568551955321, + "grad_norm": 0.962029218673706, + "learning_rate": 1.382764925026092e-07, + "loss": 0.0742, + "step": 41945 + }, + { + "epoch": 0.983280296770337, + "grad_norm": 2.878955602645874, + "learning_rate": 1.3788952360744845e-07, + "loss": 0.2498, + "step": 41946 + }, + { + "epoch": 0.9833037383451421, + "grad_norm": 1.9759920835494995, + "learning_rate": 1.3750309656690442e-07, + "loss": 1.3266, + "step": 41947 + }, + { + "epoch": 0.983327179919947, + "grad_norm": 1.3454989194869995, + "learning_rate": 1.3711721138307544e-07, + "loss": 0.1078, + "step": 41948 + }, + { + "epoch": 0.983350621494752, + "grad_norm": 2.0678048133850098, + "learning_rate": 1.3673186805805983e-07, + "loss": 0.1939, + "step": 41949 + }, + { + "epoch": 0.983374063069557, + "grad_norm": 2.4263346195220947, + "learning_rate": 1.363470665939559e-07, + "loss": 0.2469, + "step": 41950 + }, + { + "epoch": 0.983397504644362, + "grad_norm": 6.426438808441162, + "learning_rate": 1.359628069928176e-07, + "loss": 0.319, + "step": 41951 + }, + { + "epoch": 0.983420946219167, + "grad_norm": 1.9669818878173828, + "learning_rate": 1.3557908925677653e-07, + "loss": 0.1084, + "step": 41952 + }, + { + "epoch": 0.983444387793972, + "grad_norm": 0.6850908398628235, + "learning_rate": 1.3519591338787551e-07, + "loss": 0.0447, + "step": 41953 + }, + { + "epoch": 0.983467829368777, + "grad_norm": 3.618176221847534, + "learning_rate": 1.3481327938821287e-07, + "loss": 0.2762, + "step": 41954 + }, + { + "epoch": 0.983491270943582, + "grad_norm": 2.6481854915618896, + "learning_rate": 1.344311872598647e-07, + "loss": 0.1532, + "step": 41955 + }, + { + "epoch": 0.983514712518387, + "grad_norm": 2.111395835876465, + "learning_rate": 1.3404963700489603e-07, + "loss": 0.1327, + "step": 41956 + }, + { + "epoch": 0.983538154093192, + "grad_norm": 3.6640260219573975, + "learning_rate": 1.3366862862538298e-07, + "loss": 0.1379, + "step": 41957 + }, + { + "epoch": 0.983561595667997, + "grad_norm": 6.481539249420166, + "learning_rate": 1.3328816212340166e-07, + "loss": 0.6265, + "step": 41958 + }, + { + "epoch": 0.983585037242802, + "grad_norm": 1.8222142457962036, + "learning_rate": 1.329082375009949e-07, + "loss": 0.1283, + "step": 41959 + }, + { + "epoch": 0.9836084788176069, + "grad_norm": 0.5179885625839233, + "learning_rate": 1.3252885476023879e-07, + "loss": 0.0581, + "step": 41960 + }, + { + "epoch": 0.983631920392412, + "grad_norm": 2.282433032989502, + "learning_rate": 1.3215001390318726e-07, + "loss": 0.1481, + "step": 41961 + }, + { + "epoch": 0.9836553619672169, + "grad_norm": 2.351588249206543, + "learning_rate": 1.3177171493190532e-07, + "loss": 0.1705, + "step": 41962 + }, + { + "epoch": 0.983678803542022, + "grad_norm": 0.38416779041290283, + "learning_rate": 1.313939578484358e-07, + "loss": 0.02, + "step": 41963 + }, + { + "epoch": 0.9837022451168269, + "grad_norm": 1.9266037940979004, + "learning_rate": 1.3101674265482146e-07, + "loss": 0.132, + "step": 41964 + }, + { + "epoch": 0.983725686691632, + "grad_norm": 3.813950777053833, + "learning_rate": 1.3064006935312734e-07, + "loss": 0.3001, + "step": 41965 + }, + { + "epoch": 0.9837491282664369, + "grad_norm": 2.093567371368408, + "learning_rate": 1.3026393794538516e-07, + "loss": 0.1498, + "step": 41966 + }, + { + "epoch": 0.983772569841242, + "grad_norm": 1.5473978519439697, + "learning_rate": 1.2988834843364884e-07, + "loss": 0.1113, + "step": 41967 + }, + { + "epoch": 0.9837960114160469, + "grad_norm": 3.2163686752319336, + "learning_rate": 1.2951330081993896e-07, + "loss": 0.1899, + "step": 41968 + }, + { + "epoch": 0.9838194529908519, + "grad_norm": 2.4355695247650146, + "learning_rate": 1.2913879510628723e-07, + "loss": 0.2215, + "step": 41969 + }, + { + "epoch": 0.9838428945656569, + "grad_norm": 4.059430122375488, + "learning_rate": 1.2876483129474758e-07, + "loss": 0.3148, + "step": 41970 + }, + { + "epoch": 0.9838663361404619, + "grad_norm": 0.3824087083339691, + "learning_rate": 1.2839140938734063e-07, + "loss": 0.0184, + "step": 41971 + }, + { + "epoch": 0.9838897777152669, + "grad_norm": 3.6191248893737793, + "learning_rate": 1.2801852938608695e-07, + "loss": 0.306, + "step": 41972 + }, + { + "epoch": 0.9839132192900719, + "grad_norm": 3.1832268238067627, + "learning_rate": 1.2764619129300715e-07, + "loss": 0.2484, + "step": 41973 + }, + { + "epoch": 0.9839366608648769, + "grad_norm": 3.4271466732025146, + "learning_rate": 1.2727439511013295e-07, + "loss": 0.2836, + "step": 41974 + }, + { + "epoch": 0.9839601024396819, + "grad_norm": 2.986720323562622, + "learning_rate": 1.2690314083947385e-07, + "loss": 0.2509, + "step": 41975 + }, + { + "epoch": 0.9839835440144868, + "grad_norm": 3.1092987060546875, + "learning_rate": 1.2653242848305048e-07, + "loss": 0.2649, + "step": 41976 + }, + { + "epoch": 0.9840069855892919, + "grad_norm": 1.736884593963623, + "learning_rate": 1.261622580428612e-07, + "loss": 0.9238, + "step": 41977 + }, + { + "epoch": 0.9840304271640968, + "grad_norm": 0.30303266644477844, + "learning_rate": 1.2579262952093773e-07, + "loss": 0.0279, + "step": 41978 + }, + { + "epoch": 0.9840538687389019, + "grad_norm": 1.6709635257720947, + "learning_rate": 1.254235429192674e-07, + "loss": 0.0381, + "step": 41979 + }, + { + "epoch": 0.9840773103137068, + "grad_norm": 2.552932024002075, + "learning_rate": 1.2505499823984857e-07, + "loss": 1.4249, + "step": 41980 + }, + { + "epoch": 0.9841007518885119, + "grad_norm": 0.7603167295455933, + "learning_rate": 1.2468699548469077e-07, + "loss": 0.0545, + "step": 41981 + }, + { + "epoch": 0.9841241934633169, + "grad_norm": 2.047649621963501, + "learning_rate": 1.2431953465579238e-07, + "loss": 0.0742, + "step": 41982 + }, + { + "epoch": 0.9841476350381219, + "grad_norm": 2.974341630935669, + "learning_rate": 1.2395261575515182e-07, + "loss": 0.2358, + "step": 41983 + }, + { + "epoch": 0.9841710766129269, + "grad_norm": 1.7846176624298096, + "learning_rate": 1.235862387847453e-07, + "loss": 0.1419, + "step": 41984 + }, + { + "epoch": 0.9841945181877318, + "grad_norm": 2.0201499462127686, + "learning_rate": 1.2322040374657118e-07, + "loss": 0.0754, + "step": 41985 + }, + { + "epoch": 0.9842179597625369, + "grad_norm": 3.406935930252075, + "learning_rate": 1.2285511064260567e-07, + "loss": 0.2136, + "step": 41986 + }, + { + "epoch": 0.9842414013373418, + "grad_norm": 4.437325477600098, + "learning_rate": 1.224903594748472e-07, + "loss": 0.3578, + "step": 41987 + }, + { + "epoch": 0.9842648429121469, + "grad_norm": 0.22038638591766357, + "learning_rate": 1.2212615024524977e-07, + "loss": 0.0215, + "step": 41988 + }, + { + "epoch": 0.9842882844869518, + "grad_norm": 0.9475652575492859, + "learning_rate": 1.2176248295581173e-07, + "loss": 0.091, + "step": 41989 + }, + { + "epoch": 0.9843117260617569, + "grad_norm": 3.363379716873169, + "learning_rate": 1.2139935760849818e-07, + "loss": 0.2595, + "step": 41990 + }, + { + "epoch": 0.9843351676365618, + "grad_norm": 3.5904362201690674, + "learning_rate": 1.2103677420528537e-07, + "loss": 0.3336, + "step": 41991 + }, + { + "epoch": 0.9843586092113669, + "grad_norm": 2.6677277088165283, + "learning_rate": 1.2067473274812723e-07, + "loss": 0.1427, + "step": 41992 + }, + { + "epoch": 0.9843820507861718, + "grad_norm": 0.31257516145706177, + "learning_rate": 1.2031323323901112e-07, + "loss": 0.0338, + "step": 41993 + }, + { + "epoch": 0.9844054923609769, + "grad_norm": 2.482217788696289, + "learning_rate": 1.1995227567986878e-07, + "loss": 0.2112, + "step": 41994 + }, + { + "epoch": 0.9844289339357818, + "grad_norm": 2.6496472358703613, + "learning_rate": 1.1959186007267643e-07, + "loss": 0.3046, + "step": 41995 + }, + { + "epoch": 0.9844523755105868, + "grad_norm": 1.1343066692352295, + "learning_rate": 1.1923198641938803e-07, + "loss": 0.1169, + "step": 41996 + }, + { + "epoch": 0.9844758170853918, + "grad_norm": 1.899754285812378, + "learning_rate": 1.1887265472195764e-07, + "loss": 0.123, + "step": 41997 + }, + { + "epoch": 0.9844992586601968, + "grad_norm": 2.284118890762329, + "learning_rate": 1.1851386498232809e-07, + "loss": 0.1206, + "step": 41998 + }, + { + "epoch": 0.9845227002350018, + "grad_norm": 5.145362377166748, + "learning_rate": 1.1815561720245338e-07, + "loss": 0.2044, + "step": 41999 + }, + { + "epoch": 0.9845461418098068, + "grad_norm": 3.092406988143921, + "learning_rate": 1.1779791138427643e-07, + "loss": 1.4318, + "step": 42000 + }, + { + "epoch": 0.9845695833846118, + "grad_norm": 2.752330780029297, + "learning_rate": 1.17440747529729e-07, + "loss": 0.2352, + "step": 42001 + }, + { + "epoch": 0.9845930249594168, + "grad_norm": 3.2449536323547363, + "learning_rate": 1.170841256407651e-07, + "loss": 0.3141, + "step": 42002 + }, + { + "epoch": 0.9846164665342217, + "grad_norm": 3.3876895904541016, + "learning_rate": 1.167280457193054e-07, + "loss": 0.2092, + "step": 42003 + }, + { + "epoch": 0.9846399081090268, + "grad_norm": 1.0291060209274292, + "learning_rate": 1.163725077672928e-07, + "loss": 0.2534, + "step": 42004 + }, + { + "epoch": 0.9846633496838317, + "grad_norm": 2.7607972621917725, + "learning_rate": 1.16017511786648e-07, + "loss": 1.2968, + "step": 42005 + }, + { + "epoch": 0.9846867912586368, + "grad_norm": 4.1424784660339355, + "learning_rate": 1.1566305777930275e-07, + "loss": 0.2297, + "step": 42006 + }, + { + "epoch": 0.9847102328334417, + "grad_norm": 1.6733089685440063, + "learning_rate": 1.1530914574717777e-07, + "loss": 0.203, + "step": 42007 + }, + { + "epoch": 0.9847336744082468, + "grad_norm": 3.910252332687378, + "learning_rate": 1.1495577569219373e-07, + "loss": 0.3804, + "step": 42008 + }, + { + "epoch": 0.9847571159830517, + "grad_norm": 2.1390790939331055, + "learning_rate": 1.1460294761627133e-07, + "loss": 0.0651, + "step": 42009 + }, + { + "epoch": 0.9847805575578568, + "grad_norm": 3.4450747966766357, + "learning_rate": 1.1425066152132014e-07, + "loss": 0.4129, + "step": 42010 + }, + { + "epoch": 0.9848039991326617, + "grad_norm": 0.7769533395767212, + "learning_rate": 1.1389891740924974e-07, + "loss": 0.0455, + "step": 42011 + }, + { + "epoch": 0.9848274407074668, + "grad_norm": 0.22861263155937195, + "learning_rate": 1.1354771528196973e-07, + "loss": 0.0177, + "step": 42012 + }, + { + "epoch": 0.9848508822822717, + "grad_norm": 2.9992361068725586, + "learning_rate": 1.1319705514140077e-07, + "loss": 0.2565, + "step": 42013 + }, + { + "epoch": 0.9848743238570767, + "grad_norm": 3.0374860763549805, + "learning_rate": 1.1284693698941917e-07, + "loss": 1.5277, + "step": 42014 + }, + { + "epoch": 0.9848977654318817, + "grad_norm": 2.1497840881347656, + "learning_rate": 1.124973608279456e-07, + "loss": 0.1159, + "step": 42015 + }, + { + "epoch": 0.9849212070066867, + "grad_norm": 0.587873637676239, + "learning_rate": 1.1214832665886743e-07, + "loss": 0.0291, + "step": 42016 + }, + { + "epoch": 0.9849446485814917, + "grad_norm": 5.839058876037598, + "learning_rate": 1.1179983448408316e-07, + "loss": 0.7601, + "step": 42017 + }, + { + "epoch": 0.9849680901562967, + "grad_norm": 1.8010510206222534, + "learning_rate": 1.1145188430549125e-07, + "loss": 0.3103, + "step": 42018 + }, + { + "epoch": 0.9849915317311017, + "grad_norm": 3.3699419498443604, + "learning_rate": 1.1110447612495689e-07, + "loss": 0.1172, + "step": 42019 + }, + { + "epoch": 0.9850149733059067, + "grad_norm": 4.591249465942383, + "learning_rate": 1.1075760994437856e-07, + "loss": 0.5532, + "step": 42020 + }, + { + "epoch": 0.9850384148807116, + "grad_norm": 2.1571338176727295, + "learning_rate": 1.1041128576564364e-07, + "loss": 1.5625, + "step": 42021 + }, + { + "epoch": 0.9850618564555167, + "grad_norm": 0.500667929649353, + "learning_rate": 1.1006550359061729e-07, + "loss": 0.0253, + "step": 42022 + }, + { + "epoch": 0.9850852980303216, + "grad_norm": 1.4008352756500244, + "learning_rate": 1.0972026342118691e-07, + "loss": 0.1053, + "step": 42023 + }, + { + "epoch": 0.9851087396051267, + "grad_norm": 0.5439323782920837, + "learning_rate": 1.0937556525922877e-07, + "loss": 0.0448, + "step": 42024 + }, + { + "epoch": 0.9851321811799316, + "grad_norm": 3.510481595993042, + "learning_rate": 1.0903140910660802e-07, + "loss": 0.2201, + "step": 42025 + }, + { + "epoch": 0.9851556227547367, + "grad_norm": 1.79990553855896, + "learning_rate": 1.0868779496518988e-07, + "loss": 0.6512, + "step": 42026 + }, + { + "epoch": 0.9851790643295416, + "grad_norm": 3.7158684730529785, + "learning_rate": 1.083447228368395e-07, + "loss": 0.2411, + "step": 42027 + }, + { + "epoch": 0.9852025059043467, + "grad_norm": 2.493105888366699, + "learning_rate": 1.0800219272342205e-07, + "loss": 0.2245, + "step": 42028 + }, + { + "epoch": 0.9852259474791516, + "grad_norm": 4.421070098876953, + "learning_rate": 1.076602046267805e-07, + "loss": 0.3185, + "step": 42029 + }, + { + "epoch": 0.9852493890539566, + "grad_norm": 1.2488833665847778, + "learning_rate": 1.0731875854880225e-07, + "loss": 0.0467, + "step": 42030 + }, + { + "epoch": 0.9852728306287616, + "grad_norm": 1.072107195854187, + "learning_rate": 1.0697785449130803e-07, + "loss": 0.092, + "step": 42031 + }, + { + "epoch": 0.9852962722035666, + "grad_norm": 2.9771347045898438, + "learning_rate": 1.0663749245616306e-07, + "loss": 0.1757, + "step": 42032 + }, + { + "epoch": 0.9853197137783717, + "grad_norm": 3.0620837211608887, + "learning_rate": 1.0629767244521028e-07, + "loss": 0.3756, + "step": 42033 + }, + { + "epoch": 0.9853431553531766, + "grad_norm": 1.994900107383728, + "learning_rate": 1.0595839446030375e-07, + "loss": 0.1427, + "step": 42034 + }, + { + "epoch": 0.9853665969279817, + "grad_norm": 1.2996412515640259, + "learning_rate": 1.0561965850326427e-07, + "loss": 0.0979, + "step": 42035 + }, + { + "epoch": 0.9853900385027866, + "grad_norm": 2.6171164512634277, + "learning_rate": 1.05281464575957e-07, + "loss": 0.0725, + "step": 42036 + }, + { + "epoch": 0.9854134800775917, + "grad_norm": 0.8994220495223999, + "learning_rate": 1.049438126801916e-07, + "loss": 0.0683, + "step": 42037 + }, + { + "epoch": 0.9854369216523966, + "grad_norm": 5.009870529174805, + "learning_rate": 1.0460670281781105e-07, + "loss": 0.324, + "step": 42038 + }, + { + "epoch": 0.9854603632272017, + "grad_norm": 1.3631559610366821, + "learning_rate": 1.042701349906361e-07, + "loss": 0.108, + "step": 42039 + }, + { + "epoch": 0.9854838048020066, + "grad_norm": 0.48085784912109375, + "learning_rate": 1.0393410920050973e-07, + "loss": 0.0466, + "step": 42040 + }, + { + "epoch": 0.9855072463768116, + "grad_norm": 3.3245465755462646, + "learning_rate": 1.035986254492416e-07, + "loss": 0.3481, + "step": 42041 + }, + { + "epoch": 0.9855306879516166, + "grad_norm": 3.2065768241882324, + "learning_rate": 1.0326368373865247e-07, + "loss": 0.2381, + "step": 42042 + }, + { + "epoch": 0.9855541295264216, + "grad_norm": 2.481092691421509, + "learning_rate": 1.0292928407057422e-07, + "loss": 0.1588, + "step": 42043 + }, + { + "epoch": 0.9855775711012266, + "grad_norm": 2.045658588409424, + "learning_rate": 1.0259542644680542e-07, + "loss": 0.099, + "step": 42044 + }, + { + "epoch": 0.9856010126760316, + "grad_norm": 1.900832176208496, + "learning_rate": 1.0226211086915571e-07, + "loss": 0.1716, + "step": 42045 + }, + { + "epoch": 0.9856244542508366, + "grad_norm": 2.768153190612793, + "learning_rate": 1.0192933733944588e-07, + "loss": 0.185, + "step": 42046 + }, + { + "epoch": 0.9856478958256416, + "grad_norm": 0.4685071110725403, + "learning_rate": 1.0159710585947446e-07, + "loss": 0.0416, + "step": 42047 + }, + { + "epoch": 0.9856713374004465, + "grad_norm": 0.19850477576255798, + "learning_rate": 1.0126541643105114e-07, + "loss": 0.0044, + "step": 42048 + }, + { + "epoch": 0.9856947789752516, + "grad_norm": 2.9937593936920166, + "learning_rate": 1.0093426905596337e-07, + "loss": 0.2287, + "step": 42049 + }, + { + "epoch": 0.9857182205500565, + "grad_norm": 4.074273586273193, + "learning_rate": 1.0060366373602081e-07, + "loss": 0.3779, + "step": 42050 + }, + { + "epoch": 0.9857416621248616, + "grad_norm": 1.586952805519104, + "learning_rate": 1.0027360047301093e-07, + "loss": 0.0921, + "step": 42051 + }, + { + "epoch": 0.9857651036996665, + "grad_norm": 1.0389206409454346, + "learning_rate": 9.994407926872118e-08, + "loss": 0.08, + "step": 42052 + }, + { + "epoch": 0.9857885452744716, + "grad_norm": 2.5811171531677246, + "learning_rate": 9.961510012495012e-08, + "loss": 0.2268, + "step": 42053 + }, + { + "epoch": 0.9858119868492765, + "grad_norm": 2.8049373626708984, + "learning_rate": 9.928666304347411e-08, + "loss": 0.2494, + "step": 42054 + }, + { + "epoch": 0.9858354284240816, + "grad_norm": 3.699392080307007, + "learning_rate": 9.895876802608062e-08, + "loss": 0.4635, + "step": 42055 + }, + { + "epoch": 0.9858588699988865, + "grad_norm": 5.6180009841918945, + "learning_rate": 9.863141507454598e-08, + "loss": 0.3491, + "step": 42056 + }, + { + "epoch": 0.9858823115736915, + "grad_norm": 3.5557122230529785, + "learning_rate": 9.830460419064657e-08, + "loss": 0.171, + "step": 42057 + }, + { + "epoch": 0.9859057531484965, + "grad_norm": 1.1744334697723389, + "learning_rate": 9.797833537614764e-08, + "loss": 0.0548, + "step": 42058 + }, + { + "epoch": 0.9859291947233015, + "grad_norm": 2.196171522140503, + "learning_rate": 9.765260863282554e-08, + "loss": 0.1173, + "step": 42059 + }, + { + "epoch": 0.9859526362981065, + "grad_norm": 1.8777239322662354, + "learning_rate": 9.732742396245664e-08, + "loss": 0.0596, + "step": 42060 + }, + { + "epoch": 0.9859760778729115, + "grad_norm": 2.6404552459716797, + "learning_rate": 9.700278136679508e-08, + "loss": 0.0653, + "step": 42061 + }, + { + "epoch": 0.9859995194477165, + "grad_norm": 1.9118621349334717, + "learning_rate": 9.667868084760612e-08, + "loss": 0.1448, + "step": 42062 + }, + { + "epoch": 0.9860229610225215, + "grad_norm": 1.0618417263031006, + "learning_rate": 9.635512240665501e-08, + "loss": 0.0661, + "step": 42063 + }, + { + "epoch": 0.9860464025973265, + "grad_norm": 3.2464723587036133, + "learning_rate": 9.60321060456848e-08, + "loss": 0.2375, + "step": 42064 + }, + { + "epoch": 0.9860698441721315, + "grad_norm": 0.7131055593490601, + "learning_rate": 9.570963176644965e-08, + "loss": 0.0831, + "step": 42065 + }, + { + "epoch": 0.9860932857469364, + "grad_norm": 1.1636903285980225, + "learning_rate": 9.538769957070371e-08, + "loss": 0.1513, + "step": 42066 + }, + { + "epoch": 0.9861167273217415, + "grad_norm": 3.1776180267333984, + "learning_rate": 9.506630946020112e-08, + "loss": 0.2874, + "step": 42067 + }, + { + "epoch": 0.9861401688965464, + "grad_norm": 1.0838271379470825, + "learning_rate": 9.474546143667384e-08, + "loss": 0.1096, + "step": 42068 + }, + { + "epoch": 0.9861636104713515, + "grad_norm": 1.6305047273635864, + "learning_rate": 9.442515550186492e-08, + "loss": 0.1087, + "step": 42069 + }, + { + "epoch": 0.9861870520461564, + "grad_norm": 2.447763681411743, + "learning_rate": 9.410539165750632e-08, + "loss": 0.1513, + "step": 42070 + }, + { + "epoch": 0.9862104936209615, + "grad_norm": 1.3761401176452637, + "learning_rate": 9.378616990535217e-08, + "loss": 0.1449, + "step": 42071 + }, + { + "epoch": 0.9862339351957664, + "grad_norm": 1.3795056343078613, + "learning_rate": 9.346749024712331e-08, + "loss": 0.1082, + "step": 42072 + }, + { + "epoch": 0.9862573767705715, + "grad_norm": 1.8403170108795166, + "learning_rate": 9.314935268454061e-08, + "loss": 0.1391, + "step": 42073 + }, + { + "epoch": 0.9862808183453764, + "grad_norm": 2.5151560306549072, + "learning_rate": 9.283175721934712e-08, + "loss": 0.2025, + "step": 42074 + }, + { + "epoch": 0.9863042599201814, + "grad_norm": 2.4383811950683594, + "learning_rate": 9.251470385325256e-08, + "loss": 0.1326, + "step": 42075 + }, + { + "epoch": 0.9863277014949864, + "grad_norm": 2.1497297286987305, + "learning_rate": 9.21981925879889e-08, + "loss": 1.2632, + "step": 42076 + }, + { + "epoch": 0.9863511430697914, + "grad_norm": 1.4392480850219727, + "learning_rate": 9.188222342525476e-08, + "loss": 0.0941, + "step": 42077 + }, + { + "epoch": 0.9863745846445964, + "grad_norm": 2.984994649887085, + "learning_rate": 9.156679636678212e-08, + "loss": 0.0945, + "step": 42078 + }, + { + "epoch": 0.9863980262194014, + "grad_norm": 4.307251930236816, + "learning_rate": 9.125191141428069e-08, + "loss": 0.2745, + "step": 42079 + }, + { + "epoch": 0.9864214677942064, + "grad_norm": 1.6549960374832153, + "learning_rate": 9.093756856946024e-08, + "loss": 0.0529, + "step": 42080 + }, + { + "epoch": 0.9864449093690114, + "grad_norm": 0.6169266104698181, + "learning_rate": 9.062376783400827e-08, + "loss": 0.0348, + "step": 42081 + }, + { + "epoch": 0.9864683509438164, + "grad_norm": 1.6158249378204346, + "learning_rate": 9.031050920965677e-08, + "loss": 0.0821, + "step": 42082 + }, + { + "epoch": 0.9864917925186214, + "grad_norm": 3.13289737701416, + "learning_rate": 8.999779269808217e-08, + "loss": 0.2102, + "step": 42083 + }, + { + "epoch": 0.9865152340934265, + "grad_norm": 1.254193902015686, + "learning_rate": 8.96856183009831e-08, + "loss": 1.4638, + "step": 42084 + }, + { + "epoch": 0.9865386756682314, + "grad_norm": 2.121067523956299, + "learning_rate": 8.93739860200582e-08, + "loss": 0.1124, + "step": 42085 + }, + { + "epoch": 0.9865621172430364, + "grad_norm": 0.782634437084198, + "learning_rate": 8.906289585700611e-08, + "loss": 0.0675, + "step": 42086 + }, + { + "epoch": 0.9865855588178414, + "grad_norm": 1.0411157608032227, + "learning_rate": 8.875234781351439e-08, + "loss": 0.0395, + "step": 42087 + }, + { + "epoch": 0.9866090003926464, + "grad_norm": 2.6826162338256836, + "learning_rate": 8.844234189124834e-08, + "loss": 0.1984, + "step": 42088 + }, + { + "epoch": 0.9866324419674514, + "grad_norm": 2.014930248260498, + "learning_rate": 8.813287809191772e-08, + "loss": 1.4428, + "step": 42089 + }, + { + "epoch": 0.9866558835422564, + "grad_norm": 1.565455675125122, + "learning_rate": 8.782395641717678e-08, + "loss": 0.0617, + "step": 42090 + }, + { + "epoch": 0.9866793251170614, + "grad_norm": 2.7461354732513428, + "learning_rate": 8.751557686871303e-08, + "loss": 0.2001, + "step": 42091 + }, + { + "epoch": 0.9867027666918664, + "grad_norm": 3.4702930450439453, + "learning_rate": 8.720773944820293e-08, + "loss": 0.1591, + "step": 42092 + }, + { + "epoch": 0.9867262082666713, + "grad_norm": 1.9577176570892334, + "learning_rate": 8.69004441573118e-08, + "loss": 0.057, + "step": 42093 + }, + { + "epoch": 0.9867496498414764, + "grad_norm": 3.4895288944244385, + "learning_rate": 8.659369099770498e-08, + "loss": 0.2129, + "step": 42094 + }, + { + "epoch": 0.9867730914162813, + "grad_norm": 3.3646833896636963, + "learning_rate": 8.62874799710478e-08, + "loss": 0.3066, + "step": 42095 + }, + { + "epoch": 0.9867965329910864, + "grad_norm": 0.8804746270179749, + "learning_rate": 8.59818110790056e-08, + "loss": 0.0421, + "step": 42096 + }, + { + "epoch": 0.9868199745658913, + "grad_norm": 2.023467540740967, + "learning_rate": 8.567668432324372e-08, + "loss": 0.2518, + "step": 42097 + }, + { + "epoch": 0.9868434161406964, + "grad_norm": 3.3537964820861816, + "learning_rate": 8.537209970540527e-08, + "loss": 0.1564, + "step": 42098 + }, + { + "epoch": 0.9868668577155013, + "grad_norm": 2.012871026992798, + "learning_rate": 8.506805722714451e-08, + "loss": 0.1001, + "step": 42099 + }, + { + "epoch": 0.9868902992903064, + "grad_norm": 3.520437002182007, + "learning_rate": 8.476455689011565e-08, + "loss": 0.1698, + "step": 42100 + }, + { + "epoch": 0.9869137408651113, + "grad_norm": 1.8306876420974731, + "learning_rate": 8.446159869596182e-08, + "loss": 0.1172, + "step": 42101 + }, + { + "epoch": 0.9869371824399163, + "grad_norm": 2.6495115756988525, + "learning_rate": 8.415918264632617e-08, + "loss": 0.202, + "step": 42102 + }, + { + "epoch": 0.9869606240147213, + "grad_norm": 1.122100591659546, + "learning_rate": 8.38573087428518e-08, + "loss": 0.0512, + "step": 42103 + }, + { + "epoch": 0.9869840655895263, + "grad_norm": 5.773512363433838, + "learning_rate": 8.355597698718187e-08, + "loss": 0.3387, + "step": 42104 + }, + { + "epoch": 0.9870075071643313, + "grad_norm": 3.0914318561553955, + "learning_rate": 8.32551873809484e-08, + "loss": 0.2014, + "step": 42105 + }, + { + "epoch": 0.9870309487391363, + "grad_norm": 2.141177177429199, + "learning_rate": 8.295493992578341e-08, + "loss": 0.1562, + "step": 42106 + }, + { + "epoch": 0.9870543903139413, + "grad_norm": 1.9942559003829956, + "learning_rate": 8.265523462330783e-08, + "loss": 0.0783, + "step": 42107 + }, + { + "epoch": 0.9870778318887463, + "grad_norm": 1.199556589126587, + "learning_rate": 8.235607147515367e-08, + "loss": 0.1363, + "step": 42108 + }, + { + "epoch": 0.9871012734635513, + "grad_norm": 3.389402151107788, + "learning_rate": 8.2057450482953e-08, + "loss": 0.4116, + "step": 42109 + }, + { + "epoch": 0.9871247150383563, + "grad_norm": 1.2305699586868286, + "learning_rate": 8.17593716483156e-08, + "loss": 0.126, + "step": 42110 + }, + { + "epoch": 0.9871481566131612, + "grad_norm": 0.782378613948822, + "learning_rate": 8.146183497285131e-08, + "loss": 0.0782, + "step": 42111 + }, + { + "epoch": 0.9871715981879663, + "grad_norm": 1.679179072380066, + "learning_rate": 8.116484045819218e-08, + "loss": 0.0959, + "step": 42112 + }, + { + "epoch": 0.9871950397627712, + "grad_norm": 1.0369161367416382, + "learning_rate": 8.086838810594799e-08, + "loss": 0.0773, + "step": 42113 + }, + { + "epoch": 0.9872184813375763, + "grad_norm": 3.935289144515991, + "learning_rate": 8.057247791771749e-08, + "loss": 0.1697, + "step": 42114 + }, + { + "epoch": 0.9872419229123812, + "grad_norm": 0.6469109058380127, + "learning_rate": 8.027710989511051e-08, + "loss": 0.0694, + "step": 42115 + }, + { + "epoch": 0.9872653644871863, + "grad_norm": 3.1341536045074463, + "learning_rate": 7.998228403972574e-08, + "loss": 0.1968, + "step": 42116 + }, + { + "epoch": 0.9872888060619912, + "grad_norm": 2.221557378768921, + "learning_rate": 7.968800035316194e-08, + "loss": 0.1108, + "step": 42117 + }, + { + "epoch": 0.9873122476367963, + "grad_norm": 3.987529754638672, + "learning_rate": 7.939425883702889e-08, + "loss": 0.1855, + "step": 42118 + }, + { + "epoch": 0.9873356892116012, + "grad_norm": 3.4277522563934326, + "learning_rate": 7.910105949290314e-08, + "loss": 0.3508, + "step": 42119 + }, + { + "epoch": 0.9873591307864062, + "grad_norm": 2.9818646907806396, + "learning_rate": 7.88084023223945e-08, + "loss": 0.2099, + "step": 42120 + }, + { + "epoch": 0.9873825723612112, + "grad_norm": 1.9510235786437988, + "learning_rate": 7.851628732707949e-08, + "loss": 0.1035, + "step": 42121 + }, + { + "epoch": 0.9874060139360162, + "grad_norm": 0.6220452785491943, + "learning_rate": 7.822471450854574e-08, + "loss": 0.0463, + "step": 42122 + }, + { + "epoch": 0.9874294555108212, + "grad_norm": 1.975846529006958, + "learning_rate": 7.793368386836975e-08, + "loss": 1.8771, + "step": 42123 + }, + { + "epoch": 0.9874528970856262, + "grad_norm": 2.1965456008911133, + "learning_rate": 7.764319540813914e-08, + "loss": 0.1106, + "step": 42124 + }, + { + "epoch": 0.9874763386604312, + "grad_norm": 4.863966941833496, + "learning_rate": 7.735324912941932e-08, + "loss": 0.3003, + "step": 42125 + }, + { + "epoch": 0.9874997802352362, + "grad_norm": 0.6386576294898987, + "learning_rate": 7.706384503379793e-08, + "loss": 0.0407, + "step": 42126 + }, + { + "epoch": 0.9875232218100412, + "grad_norm": 0.7375481724739075, + "learning_rate": 7.677498312282927e-08, + "loss": 0.0656, + "step": 42127 + }, + { + "epoch": 0.9875466633848462, + "grad_norm": 0.3888457715511322, + "learning_rate": 7.648666339810096e-08, + "loss": 0.0208, + "step": 42128 + }, + { + "epoch": 0.9875701049596511, + "grad_norm": 2.6821510791778564, + "learning_rate": 7.61988858611562e-08, + "loss": 0.2191, + "step": 42129 + }, + { + "epoch": 0.9875935465344562, + "grad_norm": 0.9253848791122437, + "learning_rate": 7.591165051357152e-08, + "loss": 0.0661, + "step": 42130 + }, + { + "epoch": 0.9876169881092611, + "grad_norm": 0.8384608626365662, + "learning_rate": 7.562495735689013e-08, + "loss": 0.0693, + "step": 42131 + }, + { + "epoch": 0.9876404296840662, + "grad_norm": 0.4991839528083801, + "learning_rate": 7.533880639268853e-08, + "loss": 0.0299, + "step": 42132 + }, + { + "epoch": 0.9876638712588711, + "grad_norm": 2.1896731853485107, + "learning_rate": 7.505319762249885e-08, + "loss": 0.0916, + "step": 42133 + }, + { + "epoch": 0.9876873128336762, + "grad_norm": 2.8150923252105713, + "learning_rate": 7.476813104788649e-08, + "loss": 0.1425, + "step": 42134 + }, + { + "epoch": 0.9877107544084811, + "grad_norm": 3.5388739109039307, + "learning_rate": 7.448360667039466e-08, + "loss": 0.2275, + "step": 42135 + }, + { + "epoch": 0.9877341959832862, + "grad_norm": 2.285255193710327, + "learning_rate": 7.419962449155548e-08, + "loss": 0.2433, + "step": 42136 + }, + { + "epoch": 0.9877576375580912, + "grad_norm": 2.103898286819458, + "learning_rate": 7.391618451292325e-08, + "loss": 0.1923, + "step": 42137 + }, + { + "epoch": 0.9877810791328961, + "grad_norm": 1.6181811094284058, + "learning_rate": 7.363328673604119e-08, + "loss": 0.1371, + "step": 42138 + }, + { + "epoch": 0.9878045207077012, + "grad_norm": 0.945162296295166, + "learning_rate": 7.33509311624192e-08, + "loss": 0.0599, + "step": 42139 + }, + { + "epoch": 0.9878279622825061, + "grad_norm": 3.3845596313476562, + "learning_rate": 7.306911779361159e-08, + "loss": 0.2087, + "step": 42140 + }, + { + "epoch": 0.9878514038573112, + "grad_norm": 0.3693538308143616, + "learning_rate": 7.278784663113935e-08, + "loss": 0.0272, + "step": 42141 + }, + { + "epoch": 0.9878748454321161, + "grad_norm": 2.9776511192321777, + "learning_rate": 7.250711767653461e-08, + "loss": 0.2482, + "step": 42142 + }, + { + "epoch": 0.9878982870069212, + "grad_norm": 2.7391104698181152, + "learning_rate": 7.222693093130729e-08, + "loss": 0.3401, + "step": 42143 + }, + { + "epoch": 0.9879217285817261, + "grad_norm": 2.247750759124756, + "learning_rate": 7.194728639698945e-08, + "loss": 0.1451, + "step": 42144 + }, + { + "epoch": 0.9879451701565312, + "grad_norm": 0.9701491594314575, + "learning_rate": 7.166818407510212e-08, + "loss": 0.0993, + "step": 42145 + }, + { + "epoch": 0.9879686117313361, + "grad_norm": 2.103339433670044, + "learning_rate": 7.13896239671441e-08, + "loss": 0.1298, + "step": 42146 + }, + { + "epoch": 0.9879920533061411, + "grad_norm": 0.18965192139148712, + "learning_rate": 7.111160607463641e-08, + "loss": 0.0087, + "step": 42147 + }, + { + "epoch": 0.9880154948809461, + "grad_norm": 0.407638818025589, + "learning_rate": 7.083413039908892e-08, + "loss": 0.0271, + "step": 42148 + }, + { + "epoch": 0.9880389364557511, + "grad_norm": 2.9652280807495117, + "learning_rate": 7.055719694201157e-08, + "loss": 0.1481, + "step": 42149 + }, + { + "epoch": 0.9880623780305561, + "grad_norm": 3.568369150161743, + "learning_rate": 7.028080570489204e-08, + "loss": 0.4282, + "step": 42150 + }, + { + "epoch": 0.9880858196053611, + "grad_norm": 4.47199010848999, + "learning_rate": 7.000495668924023e-08, + "loss": 0.7088, + "step": 42151 + }, + { + "epoch": 0.9881092611801661, + "grad_norm": 3.4765193462371826, + "learning_rate": 6.972964989655495e-08, + "loss": 0.1583, + "step": 42152 + }, + { + "epoch": 0.9881327027549711, + "grad_norm": 1.0963906049728394, + "learning_rate": 6.94548853283239e-08, + "loss": 0.1015, + "step": 42153 + }, + { + "epoch": 0.988156144329776, + "grad_norm": 0.38182562589645386, + "learning_rate": 6.918066298604586e-08, + "loss": 0.0288, + "step": 42154 + }, + { + "epoch": 0.9881795859045811, + "grad_norm": 1.661067247390747, + "learning_rate": 6.890698287119746e-08, + "loss": 0.1055, + "step": 42155 + }, + { + "epoch": 0.988203027479386, + "grad_norm": 3.813079595565796, + "learning_rate": 6.863384498526638e-08, + "loss": 0.3791, + "step": 42156 + }, + { + "epoch": 0.9882264690541911, + "grad_norm": 5.287487983703613, + "learning_rate": 6.836124932974031e-08, + "loss": 0.2779, + "step": 42157 + }, + { + "epoch": 0.988249910628996, + "grad_norm": 3.801525354385376, + "learning_rate": 6.808919590610696e-08, + "loss": 0.5254, + "step": 42158 + }, + { + "epoch": 0.9882733522038011, + "grad_norm": 2.149353265762329, + "learning_rate": 6.781768471582073e-08, + "loss": 0.1771, + "step": 42159 + }, + { + "epoch": 0.988296793778606, + "grad_norm": 2.879765510559082, + "learning_rate": 6.75467157603693e-08, + "loss": 0.2837, + "step": 42160 + }, + { + "epoch": 0.9883202353534111, + "grad_norm": 0.2562396228313446, + "learning_rate": 6.727628904120708e-08, + "loss": 0.0169, + "step": 42161 + }, + { + "epoch": 0.988343676928216, + "grad_norm": 4.176395893096924, + "learning_rate": 6.700640455982176e-08, + "loss": 0.2618, + "step": 42162 + }, + { + "epoch": 0.9883671185030211, + "grad_norm": 3.230591297149658, + "learning_rate": 6.673706231767884e-08, + "loss": 0.2333, + "step": 42163 + }, + { + "epoch": 0.988390560077826, + "grad_norm": 2.195133924484253, + "learning_rate": 6.646826231622161e-08, + "loss": 1.2883, + "step": 42164 + }, + { + "epoch": 0.988414001652631, + "grad_norm": 1.5143280029296875, + "learning_rate": 6.620000455691555e-08, + "loss": 0.8776, + "step": 42165 + }, + { + "epoch": 0.988437443227436, + "grad_norm": 3.5847349166870117, + "learning_rate": 6.593228904122617e-08, + "loss": 0.2462, + "step": 42166 + }, + { + "epoch": 0.988460884802241, + "grad_norm": 2.1267106533050537, + "learning_rate": 6.566511577058565e-08, + "loss": 0.0809, + "step": 42167 + }, + { + "epoch": 0.988484326377046, + "grad_norm": 2.4530465602874756, + "learning_rate": 6.539848474647058e-08, + "loss": 0.2716, + "step": 42168 + }, + { + "epoch": 0.988507767951851, + "grad_norm": 3.442610502243042, + "learning_rate": 6.513239597030207e-08, + "loss": 0.2497, + "step": 42169 + }, + { + "epoch": 0.988531209526656, + "grad_norm": 2.5225770473480225, + "learning_rate": 6.48668494435456e-08, + "loss": 1.2872, + "step": 42170 + }, + { + "epoch": 0.988554651101461, + "grad_norm": 4.276185989379883, + "learning_rate": 6.460184516762224e-08, + "loss": 0.4083, + "step": 42171 + }, + { + "epoch": 0.988578092676266, + "grad_norm": 3.1357381343841553, + "learning_rate": 6.43373831439864e-08, + "loss": 0.1759, + "step": 42172 + }, + { + "epoch": 0.988601534251071, + "grad_norm": 1.8294132947921753, + "learning_rate": 6.407346337405918e-08, + "loss": 0.2244, + "step": 42173 + }, + { + "epoch": 0.9886249758258759, + "grad_norm": 0.980790376663208, + "learning_rate": 6.381008585928383e-08, + "loss": 0.0981, + "step": 42174 + }, + { + "epoch": 0.988648417400681, + "grad_norm": 1.1197664737701416, + "learning_rate": 6.354725060108146e-08, + "loss": 0.0766, + "step": 42175 + }, + { + "epoch": 0.9886718589754859, + "grad_norm": 2.490387439727783, + "learning_rate": 6.328495760088426e-08, + "loss": 0.2579, + "step": 42176 + }, + { + "epoch": 0.988695300550291, + "grad_norm": 0.47627219557762146, + "learning_rate": 6.302320686011332e-08, + "loss": 0.0274, + "step": 42177 + }, + { + "epoch": 0.9887187421250959, + "grad_norm": 2.502973794937134, + "learning_rate": 6.27619983801786e-08, + "loss": 0.175, + "step": 42178 + }, + { + "epoch": 0.988742183699901, + "grad_norm": 0.5568432211875916, + "learning_rate": 6.250133216252341e-08, + "loss": 0.0597, + "step": 42179 + }, + { + "epoch": 0.9887656252747059, + "grad_norm": 1.7873483896255493, + "learning_rate": 6.224120820853552e-08, + "loss": 0.1299, + "step": 42180 + }, + { + "epoch": 0.988789066849511, + "grad_norm": 2.27949595451355, + "learning_rate": 6.198162651963601e-08, + "loss": 0.1186, + "step": 42181 + }, + { + "epoch": 0.9888125084243159, + "grad_norm": 0.4623124599456787, + "learning_rate": 6.172258709723488e-08, + "loss": 0.0328, + "step": 42182 + }, + { + "epoch": 0.9888359499991209, + "grad_norm": 4.354942321777344, + "learning_rate": 6.14640899427421e-08, + "loss": 0.1982, + "step": 42183 + }, + { + "epoch": 0.9888593915739259, + "grad_norm": 2.534221649169922, + "learning_rate": 6.120613505755656e-08, + "loss": 0.1543, + "step": 42184 + }, + { + "epoch": 0.9888828331487309, + "grad_norm": 0.5448474884033203, + "learning_rate": 6.094872244307715e-08, + "loss": 0.0347, + "step": 42185 + }, + { + "epoch": 0.9889062747235359, + "grad_norm": 1.7257325649261475, + "learning_rate": 6.069185210069161e-08, + "loss": 0.092, + "step": 42186 + }, + { + "epoch": 0.9889297162983409, + "grad_norm": 2.519953489303589, + "learning_rate": 6.043552403179887e-08, + "loss": 0.2116, + "step": 42187 + }, + { + "epoch": 0.988953157873146, + "grad_norm": 0.7189509272575378, + "learning_rate": 6.017973823780887e-08, + "loss": 0.0488, + "step": 42188 + }, + { + "epoch": 0.9889765994479509, + "grad_norm": 3.392146587371826, + "learning_rate": 5.992449472008721e-08, + "loss": 0.3379, + "step": 42189 + }, + { + "epoch": 0.989000041022756, + "grad_norm": 0.3959571421146393, + "learning_rate": 5.966979348002167e-08, + "loss": 0.0333, + "step": 42190 + }, + { + "epoch": 0.9890234825975609, + "grad_norm": 5.450596332550049, + "learning_rate": 5.9415634519000006e-08, + "loss": 0.3176, + "step": 42191 + }, + { + "epoch": 0.989046924172366, + "grad_norm": 3.9308197498321533, + "learning_rate": 5.916201783839892e-08, + "loss": 0.2444, + "step": 42192 + }, + { + "epoch": 0.9890703657471709, + "grad_norm": 3.0954604148864746, + "learning_rate": 5.8908943439583974e-08, + "loss": 0.1457, + "step": 42193 + }, + { + "epoch": 0.9890938073219759, + "grad_norm": 2.6365811824798584, + "learning_rate": 5.865641132395405e-08, + "loss": 0.1381, + "step": 42194 + }, + { + "epoch": 0.9891172488967809, + "grad_norm": 4.5718770027160645, + "learning_rate": 5.840442149286363e-08, + "loss": 0.2454, + "step": 42195 + }, + { + "epoch": 0.9891406904715859, + "grad_norm": 3.2466187477111816, + "learning_rate": 5.815297394767827e-08, + "loss": 0.2258, + "step": 42196 + }, + { + "epoch": 0.9891641320463909, + "grad_norm": 1.2043825387954712, + "learning_rate": 5.790206868976356e-08, + "loss": 0.0543, + "step": 42197 + }, + { + "epoch": 0.9891875736211959, + "grad_norm": 0.4793209731578827, + "learning_rate": 5.765170572048506e-08, + "loss": 0.0207, + "step": 42198 + }, + { + "epoch": 0.9892110151960009, + "grad_norm": 0.9393729567527771, + "learning_rate": 5.7401885041197254e-08, + "loss": 0.0733, + "step": 42199 + }, + { + "epoch": 0.9892344567708059, + "grad_norm": 3.5948903560638428, + "learning_rate": 5.7152606653254615e-08, + "loss": 0.2721, + "step": 42200 + }, + { + "epoch": 0.9892578983456108, + "grad_norm": 2.594827175140381, + "learning_rate": 5.690387055801161e-08, + "loss": 0.1807, + "step": 42201 + }, + { + "epoch": 0.9892813399204159, + "grad_norm": 1.0076888799667358, + "learning_rate": 5.6655676756822705e-08, + "loss": 0.0787, + "step": 42202 + }, + { + "epoch": 0.9893047814952208, + "grad_norm": 2.376591682434082, + "learning_rate": 5.640802525103128e-08, + "loss": 0.3342, + "step": 42203 + }, + { + "epoch": 0.9893282230700259, + "grad_norm": 0.8974959850311279, + "learning_rate": 5.6160916041969605e-08, + "loss": 0.503, + "step": 42204 + }, + { + "epoch": 0.9893516646448308, + "grad_norm": 2.938934326171875, + "learning_rate": 5.591434913099214e-08, + "loss": 0.2381, + "step": 42205 + }, + { + "epoch": 0.9893751062196359, + "grad_norm": 1.5410009622573853, + "learning_rate": 5.5668324519442263e-08, + "loss": 1.4206, + "step": 42206 + }, + { + "epoch": 0.9893985477944408, + "grad_norm": 2.125723361968994, + "learning_rate": 5.542284220864114e-08, + "loss": 0.1563, + "step": 42207 + }, + { + "epoch": 0.9894219893692459, + "grad_norm": 2.9632749557495117, + "learning_rate": 5.517790219993213e-08, + "loss": 0.2799, + "step": 42208 + }, + { + "epoch": 0.9894454309440508, + "grad_norm": 3.9212188720703125, + "learning_rate": 5.493350449462531e-08, + "loss": 0.5052, + "step": 42209 + }, + { + "epoch": 0.9894688725188558, + "grad_norm": 0.6305142045021057, + "learning_rate": 5.4689649094075145e-08, + "loss": 0.0247, + "step": 42210 + }, + { + "epoch": 0.9894923140936608, + "grad_norm": 4.076115608215332, + "learning_rate": 5.44463359995695e-08, + "loss": 0.3488, + "step": 42211 + }, + { + "epoch": 0.9895157556684658, + "grad_norm": 3.14839243888855, + "learning_rate": 5.420356521246284e-08, + "loss": 0.2874, + "step": 42212 + }, + { + "epoch": 0.9895391972432708, + "grad_norm": 0.5620961785316467, + "learning_rate": 5.396133673405412e-08, + "loss": 0.0496, + "step": 42213 + }, + { + "epoch": 0.9895626388180758, + "grad_norm": 3.5866880416870117, + "learning_rate": 5.3719650565664525e-08, + "loss": 0.1846, + "step": 42214 + }, + { + "epoch": 0.9895860803928808, + "grad_norm": 3.869328737258911, + "learning_rate": 5.3478506708604106e-08, + "loss": 0.1709, + "step": 42215 + }, + { + "epoch": 0.9896095219676858, + "grad_norm": 1.9215487241744995, + "learning_rate": 5.323790516417182e-08, + "loss": 0.1199, + "step": 42216 + }, + { + "epoch": 0.9896329635424908, + "grad_norm": 2.5671746730804443, + "learning_rate": 5.299784593367774e-08, + "loss": 0.3158, + "step": 42217 + }, + { + "epoch": 0.9896564051172958, + "grad_norm": 2.0231235027313232, + "learning_rate": 5.275832901843192e-08, + "loss": 0.1302, + "step": 42218 + }, + { + "epoch": 0.9896798466921007, + "grad_norm": 3.880422592163086, + "learning_rate": 5.251935441973332e-08, + "loss": 0.395, + "step": 42219 + }, + { + "epoch": 0.9897032882669058, + "grad_norm": 2.1300878524780273, + "learning_rate": 5.228092213888092e-08, + "loss": 0.0575, + "step": 42220 + }, + { + "epoch": 0.9897267298417107, + "grad_norm": 1.1138925552368164, + "learning_rate": 5.204303217715145e-08, + "loss": 0.158, + "step": 42221 + }, + { + "epoch": 0.9897501714165158, + "grad_norm": 0.305655300617218, + "learning_rate": 5.180568453585499e-08, + "loss": 0.0086, + "step": 42222 + }, + { + "epoch": 0.9897736129913207, + "grad_norm": 3.0326337814331055, + "learning_rate": 5.1568879216257194e-08, + "loss": 0.2647, + "step": 42223 + }, + { + "epoch": 0.9897970545661258, + "grad_norm": 2.3686108589172363, + "learning_rate": 5.133261621966812e-08, + "loss": 0.1379, + "step": 42224 + }, + { + "epoch": 0.9898204961409307, + "grad_norm": 1.3010599613189697, + "learning_rate": 5.109689554735342e-08, + "loss": 0.1312, + "step": 42225 + }, + { + "epoch": 0.9898439377157358, + "grad_norm": 3.697190284729004, + "learning_rate": 5.0861717200600955e-08, + "loss": 0.1972, + "step": 42226 + }, + { + "epoch": 0.9898673792905407, + "grad_norm": 1.2454272508621216, + "learning_rate": 5.0627081180687483e-08, + "loss": 0.109, + "step": 42227 + }, + { + "epoch": 0.9898908208653457, + "grad_norm": 3.6548893451690674, + "learning_rate": 5.0392987488878664e-08, + "loss": 0.2759, + "step": 42228 + }, + { + "epoch": 0.9899142624401507, + "grad_norm": 1.5825634002685547, + "learning_rate": 5.015943612644014e-08, + "loss": 0.043, + "step": 42229 + }, + { + "epoch": 0.9899377040149557, + "grad_norm": 3.21616268157959, + "learning_rate": 4.992642709465978e-08, + "loss": 0.2761, + "step": 42230 + }, + { + "epoch": 0.9899611455897607, + "grad_norm": 2.018847942352295, + "learning_rate": 4.969396039478102e-08, + "loss": 1.0622, + "step": 42231 + }, + { + "epoch": 0.9899845871645657, + "grad_norm": 2.9009718894958496, + "learning_rate": 4.9462036028069535e-08, + "loss": 0.13, + "step": 42232 + }, + { + "epoch": 0.9900080287393707, + "grad_norm": 2.037074327468872, + "learning_rate": 4.923065399579096e-08, + "loss": 1.3085, + "step": 42233 + }, + { + "epoch": 0.9900314703141757, + "grad_norm": 3.6506049633026123, + "learning_rate": 4.8999814299199866e-08, + "loss": 0.4281, + "step": 42234 + }, + { + "epoch": 0.9900549118889806, + "grad_norm": 1.6205850839614868, + "learning_rate": 4.8769516939550786e-08, + "loss": 0.1088, + "step": 42235 + }, + { + "epoch": 0.9900783534637857, + "grad_norm": 2.486304521560669, + "learning_rate": 4.853976191807608e-08, + "loss": 1.3256, + "step": 42236 + }, + { + "epoch": 0.9901017950385906, + "grad_norm": 1.455268144607544, + "learning_rate": 4.8310549236041394e-08, + "loss": 0.0882, + "step": 42237 + }, + { + "epoch": 0.9901252366133957, + "grad_norm": 3.8567380905151367, + "learning_rate": 4.808187889469018e-08, + "loss": 0.3008, + "step": 42238 + }, + { + "epoch": 0.9901486781882007, + "grad_norm": 0.7049946188926697, + "learning_rate": 4.7853750895243686e-08, + "loss": 0.0799, + "step": 42239 + }, + { + "epoch": 0.9901721197630057, + "grad_norm": 0.4401894509792328, + "learning_rate": 4.762616523896757e-08, + "loss": 0.0329, + "step": 42240 + }, + { + "epoch": 0.9901955613378107, + "grad_norm": 3.6914615631103516, + "learning_rate": 4.7399121927071964e-08, + "loss": 0.301, + "step": 42241 + }, + { + "epoch": 0.9902190029126157, + "grad_norm": 0.7914968132972717, + "learning_rate": 4.717262096080033e-08, + "loss": 0.0461, + "step": 42242 + }, + { + "epoch": 0.9902424444874207, + "grad_norm": 0.953057587146759, + "learning_rate": 4.6946662341385004e-08, + "loss": 0.0647, + "step": 42243 + }, + { + "epoch": 0.9902658860622257, + "grad_norm": 0.8069885969161987, + "learning_rate": 4.672124607004724e-08, + "loss": 0.1123, + "step": 42244 + }, + { + "epoch": 0.9902893276370307, + "grad_norm": 0.7725409865379333, + "learning_rate": 4.649637214800828e-08, + "loss": 0.0472, + "step": 42245 + }, + { + "epoch": 0.9903127692118356, + "grad_norm": 2.5454845428466797, + "learning_rate": 4.627204057648937e-08, + "loss": 0.1153, + "step": 42246 + }, + { + "epoch": 0.9903362107866407, + "grad_norm": 1.2547475099563599, + "learning_rate": 4.604825135671176e-08, + "loss": 0.1751, + "step": 42247 + }, + { + "epoch": 0.9903596523614456, + "grad_norm": 1.633609652519226, + "learning_rate": 4.5825004489885584e-08, + "loss": 0.104, + "step": 42248 + }, + { + "epoch": 0.9903830939362507, + "grad_norm": 1.2195900678634644, + "learning_rate": 4.5602299977220984e-08, + "loss": 0.0688, + "step": 42249 + }, + { + "epoch": 0.9904065355110556, + "grad_norm": 1.6085724830627441, + "learning_rate": 4.538013781992811e-08, + "loss": 0.0836, + "step": 42250 + }, + { + "epoch": 0.9904299770858607, + "grad_norm": 0.9883369207382202, + "learning_rate": 4.515851801921711e-08, + "loss": 0.0856, + "step": 42251 + }, + { + "epoch": 0.9904534186606656, + "grad_norm": 0.7092793583869934, + "learning_rate": 4.4937440576287014e-08, + "loss": 0.0364, + "step": 42252 + }, + { + "epoch": 0.9904768602354707, + "grad_norm": 1.3543678522109985, + "learning_rate": 4.4716905492325766e-08, + "loss": 0.1059, + "step": 42253 + }, + { + "epoch": 0.9905003018102756, + "grad_norm": 2.6939852237701416, + "learning_rate": 4.449691276855461e-08, + "loss": 0.1821, + "step": 42254 + }, + { + "epoch": 0.9905237433850806, + "grad_norm": 3.8036293983459473, + "learning_rate": 4.427746240615038e-08, + "loss": 0.2501, + "step": 42255 + }, + { + "epoch": 0.9905471849598856, + "grad_norm": 2.6240503787994385, + "learning_rate": 4.405855440630102e-08, + "loss": 0.2063, + "step": 42256 + }, + { + "epoch": 0.9905706265346906, + "grad_norm": 1.9043012857437134, + "learning_rate": 4.384018877020557e-08, + "loss": 0.0601, + "step": 42257 + }, + { + "epoch": 0.9905940681094956, + "grad_norm": 3.124020576477051, + "learning_rate": 4.3622365499051966e-08, + "loss": 0.1748, + "step": 42258 + }, + { + "epoch": 0.9906175096843006, + "grad_norm": 1.7150375843048096, + "learning_rate": 4.3405084593994835e-08, + "loss": 0.0934, + "step": 42259 + }, + { + "epoch": 0.9906409512591056, + "grad_norm": 6.968566417694092, + "learning_rate": 4.318834605624433e-08, + "loss": 0.4568, + "step": 42260 + }, + { + "epoch": 0.9906643928339106, + "grad_norm": 3.307011604309082, + "learning_rate": 4.2972149886966184e-08, + "loss": 0.3229, + "step": 42261 + }, + { + "epoch": 0.9906878344087156, + "grad_norm": 2.472623109817505, + "learning_rate": 4.275649608732613e-08, + "loss": 0.1925, + "step": 42262 + }, + { + "epoch": 0.9907112759835206, + "grad_norm": 2.3513519763946533, + "learning_rate": 4.2541384658500996e-08, + "loss": 0.1474, + "step": 42263 + }, + { + "epoch": 0.9907347175583255, + "grad_norm": 1.942715048789978, + "learning_rate": 4.2326815601656525e-08, + "loss": 0.1639, + "step": 42264 + }, + { + "epoch": 0.9907581591331306, + "grad_norm": 1.9776970148086548, + "learning_rate": 4.2112788917958444e-08, + "loss": 1.2397, + "step": 42265 + }, + { + "epoch": 0.9907816007079355, + "grad_norm": 2.692643404006958, + "learning_rate": 4.18993046085725e-08, + "loss": 1.3437, + "step": 42266 + }, + { + "epoch": 0.9908050422827406, + "grad_norm": 1.4056800603866577, + "learning_rate": 4.168636267464221e-08, + "loss": 0.0715, + "step": 42267 + }, + { + "epoch": 0.9908284838575455, + "grad_norm": 2.901934862136841, + "learning_rate": 4.1473963117344414e-08, + "loss": 0.2294, + "step": 42268 + }, + { + "epoch": 0.9908519254323506, + "grad_norm": 1.5413448810577393, + "learning_rate": 4.1262105937811546e-08, + "loss": 0.121, + "step": 42269 + }, + { + "epoch": 0.9908753670071555, + "grad_norm": 4.915070056915283, + "learning_rate": 4.105079113719823e-08, + "loss": 0.3853, + "step": 42270 + }, + { + "epoch": 0.9908988085819606, + "grad_norm": 3.214829921722412, + "learning_rate": 4.08400187166591e-08, + "loss": 0.2532, + "step": 42271 + }, + { + "epoch": 0.9909222501567655, + "grad_norm": 2.4730381965637207, + "learning_rate": 4.0629788677326587e-08, + "loss": 1.1869, + "step": 42272 + }, + { + "epoch": 0.9909456917315705, + "grad_norm": 1.7939461469650269, + "learning_rate": 4.0420101020355316e-08, + "loss": 0.0538, + "step": 42273 + }, + { + "epoch": 0.9909691333063755, + "grad_norm": 3.6395440101623535, + "learning_rate": 4.021095574686662e-08, + "loss": 0.276, + "step": 42274 + }, + { + "epoch": 0.9909925748811805, + "grad_norm": 12.665502548217773, + "learning_rate": 4.000235285800402e-08, + "loss": 0.3162, + "step": 42275 + }, + { + "epoch": 0.9910160164559855, + "grad_norm": 0.7110495567321777, + "learning_rate": 3.979429235491105e-08, + "loss": 0.0668, + "step": 42276 + }, + { + "epoch": 0.9910394580307905, + "grad_norm": 2.9412732124328613, + "learning_rate": 3.9586774238697944e-08, + "loss": 0.3045, + "step": 42277 + }, + { + "epoch": 0.9910628996055955, + "grad_norm": 2.1877284049987793, + "learning_rate": 3.937979851049711e-08, + "loss": 0.1117, + "step": 42278 + }, + { + "epoch": 0.9910863411804005, + "grad_norm": 3.915205240249634, + "learning_rate": 3.917336517142989e-08, + "loss": 0.2361, + "step": 42279 + }, + { + "epoch": 0.9911097827552054, + "grad_norm": 0.30668187141418457, + "learning_rate": 3.8967474222617596e-08, + "loss": 0.013, + "step": 42280 + }, + { + "epoch": 0.9911332243300105, + "grad_norm": 3.49735426902771, + "learning_rate": 3.876212566518156e-08, + "loss": 0.2228, + "step": 42281 + }, + { + "epoch": 0.9911566659048154, + "grad_norm": 1.7536725997924805, + "learning_rate": 3.8557319500232e-08, + "loss": 0.1153, + "step": 42282 + }, + { + "epoch": 0.9911801074796205, + "grad_norm": 2.001328945159912, + "learning_rate": 3.8353055728890255e-08, + "loss": 0.0915, + "step": 42283 + }, + { + "epoch": 0.9912035490544254, + "grad_norm": 2.2460989952087402, + "learning_rate": 3.814933435224432e-08, + "loss": 0.1017, + "step": 42284 + }, + { + "epoch": 0.9912269906292305, + "grad_norm": 2.304915428161621, + "learning_rate": 3.794615537141555e-08, + "loss": 0.1793, + "step": 42285 + }, + { + "epoch": 0.9912504322040354, + "grad_norm": 2.577078104019165, + "learning_rate": 3.774351878749194e-08, + "loss": 0.1629, + "step": 42286 + }, + { + "epoch": 0.9912738737788405, + "grad_norm": 0.45593681931495667, + "learning_rate": 3.754142460159482e-08, + "loss": 0.0174, + "step": 42287 + }, + { + "epoch": 0.9912973153536454, + "grad_norm": 2.602328062057495, + "learning_rate": 3.7339872814801115e-08, + "loss": 0.2012, + "step": 42288 + }, + { + "epoch": 0.9913207569284505, + "grad_norm": 1.532828450202942, + "learning_rate": 3.713886342822104e-08, + "loss": 1.062, + "step": 42289 + }, + { + "epoch": 0.9913441985032555, + "grad_norm": 2.3677797317504883, + "learning_rate": 3.693839644293151e-08, + "loss": 0.172, + "step": 42290 + }, + { + "epoch": 0.9913676400780604, + "grad_norm": 1.7076570987701416, + "learning_rate": 3.673847186002055e-08, + "loss": 0.2064, + "step": 42291 + }, + { + "epoch": 0.9913910816528655, + "grad_norm": 3.3859593868255615, + "learning_rate": 3.6539089680576176e-08, + "loss": 0.2727, + "step": 42292 + }, + { + "epoch": 0.9914145232276704, + "grad_norm": 0.9469587206840515, + "learning_rate": 3.634024990568641e-08, + "loss": 0.0515, + "step": 42293 + }, + { + "epoch": 0.9914379648024755, + "grad_norm": 2.515822410583496, + "learning_rate": 3.614195253642816e-08, + "loss": 0.1815, + "step": 42294 + }, + { + "epoch": 0.9914614063772804, + "grad_norm": 2.658931016921997, + "learning_rate": 3.5944197573867244e-08, + "loss": 0.3132, + "step": 42295 + }, + { + "epoch": 0.9914848479520855, + "grad_norm": 6.303678035736084, + "learning_rate": 3.574698501909168e-08, + "loss": 0.255, + "step": 42296 + }, + { + "epoch": 0.9915082895268904, + "grad_norm": 1.9745699167251587, + "learning_rate": 3.555031487315619e-08, + "loss": 1.3095, + "step": 42297 + }, + { + "epoch": 0.9915317311016955, + "grad_norm": 4.485254287719727, + "learning_rate": 3.535418713714877e-08, + "loss": 0.3656, + "step": 42298 + }, + { + "epoch": 0.9915551726765004, + "grad_norm": 3.8002989292144775, + "learning_rate": 3.5158601812101956e-08, + "loss": 0.1548, + "step": 42299 + }, + { + "epoch": 0.9915786142513054, + "grad_norm": 0.8629452586174011, + "learning_rate": 3.496355889910374e-08, + "loss": 0.0467, + "step": 42300 + }, + { + "epoch": 0.9916020558261104, + "grad_norm": 0.6145262122154236, + "learning_rate": 3.4769058399208855e-08, + "loss": 0.0624, + "step": 42301 + }, + { + "epoch": 0.9916254974009154, + "grad_norm": 0.7693286538124084, + "learning_rate": 3.4575100313460896e-08, + "loss": 0.0402, + "step": 42302 + }, + { + "epoch": 0.9916489389757204, + "grad_norm": 0.9370254874229431, + "learning_rate": 3.438168464292568e-08, + "loss": 0.0604, + "step": 42303 + }, + { + "epoch": 0.9916723805505254, + "grad_norm": 0.16318681836128235, + "learning_rate": 3.418881138863572e-08, + "loss": 0.0047, + "step": 42304 + }, + { + "epoch": 0.9916958221253304, + "grad_norm": 2.6107053756713867, + "learning_rate": 3.399648055165683e-08, + "loss": 0.1827, + "step": 42305 + }, + { + "epoch": 0.9917192637001354, + "grad_norm": 0.8046424388885498, + "learning_rate": 3.380469213302151e-08, + "loss": 0.0698, + "step": 42306 + }, + { + "epoch": 0.9917427052749404, + "grad_norm": 0.8337745070457458, + "learning_rate": 3.361344613377338e-08, + "loss": 0.0655, + "step": 42307 + }, + { + "epoch": 0.9917661468497454, + "grad_norm": 2.254826307296753, + "learning_rate": 3.342274255494493e-08, + "loss": 0.0792, + "step": 42308 + }, + { + "epoch": 0.9917895884245503, + "grad_norm": 0.8140287399291992, + "learning_rate": 3.323258139757979e-08, + "loss": 0.0634, + "step": 42309 + }, + { + "epoch": 0.9918130299993554, + "grad_norm": 4.754089832305908, + "learning_rate": 3.304296266271045e-08, + "loss": 0.363, + "step": 42310 + }, + { + "epoch": 0.9918364715741603, + "grad_norm": 2.493744134902954, + "learning_rate": 3.285388635134723e-08, + "loss": 0.1329, + "step": 42311 + }, + { + "epoch": 0.9918599131489654, + "grad_norm": 1.5748072862625122, + "learning_rate": 3.266535246454483e-08, + "loss": 0.0444, + "step": 42312 + }, + { + "epoch": 0.9918833547237703, + "grad_norm": 3.1401703357696533, + "learning_rate": 3.2477361003302455e-08, + "loss": 0.3044, + "step": 42313 + }, + { + "epoch": 0.9919067962985754, + "grad_norm": 0.5567004680633545, + "learning_rate": 3.228991196864151e-08, + "loss": 0.0459, + "step": 42314 + }, + { + "epoch": 0.9919302378733803, + "grad_norm": 1.2429310083389282, + "learning_rate": 3.2103005361605596e-08, + "loss": 0.0566, + "step": 42315 + }, + { + "epoch": 0.9919536794481854, + "grad_norm": 1.7214665412902832, + "learning_rate": 3.1916641183171724e-08, + "loss": 0.1377, + "step": 42316 + }, + { + "epoch": 0.9919771210229903, + "grad_norm": 2.3444595336914062, + "learning_rate": 3.1730819434383496e-08, + "loss": 0.8999, + "step": 42317 + }, + { + "epoch": 0.9920005625977953, + "grad_norm": 0.24976249039173126, + "learning_rate": 3.1545540116217906e-08, + "loss": 0.0169, + "step": 42318 + }, + { + "epoch": 0.9920240041726003, + "grad_norm": 2.331608295440674, + "learning_rate": 3.136080322970747e-08, + "loss": 0.1632, + "step": 42319 + }, + { + "epoch": 0.9920474457474053, + "grad_norm": 1.9919817447662354, + "learning_rate": 3.117660877585138e-08, + "loss": 1.6089, + "step": 42320 + }, + { + "epoch": 0.9920708873222103, + "grad_norm": 2.196627140045166, + "learning_rate": 3.0992956755626635e-08, + "loss": 0.1473, + "step": 42321 + }, + { + "epoch": 0.9920943288970153, + "grad_norm": 2.2461986541748047, + "learning_rate": 3.080984717005464e-08, + "loss": 0.1165, + "step": 42322 + }, + { + "epoch": 0.9921177704718203, + "grad_norm": 2.1891963481903076, + "learning_rate": 3.0627280020123495e-08, + "loss": 0.2081, + "step": 42323 + }, + { + "epoch": 0.9921412120466253, + "grad_norm": 6.906768321990967, + "learning_rate": 3.0445255306810194e-08, + "loss": 0.3953, + "step": 42324 + }, + { + "epoch": 0.9921646536214302, + "grad_norm": 0.49668771028518677, + "learning_rate": 3.026377303111394e-08, + "loss": 0.0289, + "step": 42325 + }, + { + "epoch": 0.9921880951962353, + "grad_norm": 0.295979380607605, + "learning_rate": 3.008283319401173e-08, + "loss": 0.0157, + "step": 42326 + }, + { + "epoch": 0.9922115367710402, + "grad_norm": 2.887963056564331, + "learning_rate": 2.990243579650276e-08, + "loss": 0.3777, + "step": 42327 + }, + { + "epoch": 0.9922349783458453, + "grad_norm": 3.7404963970184326, + "learning_rate": 2.9722580839552928e-08, + "loss": 0.3439, + "step": 42328 + }, + { + "epoch": 0.9922584199206502, + "grad_norm": 3.198385715484619, + "learning_rate": 2.9543268324128127e-08, + "loss": 0.1478, + "step": 42329 + }, + { + "epoch": 0.9922818614954553, + "grad_norm": 1.4646096229553223, + "learning_rate": 2.9364498251216453e-08, + "loss": 0.0715, + "step": 42330 + }, + { + "epoch": 0.9923053030702602, + "grad_norm": 1.2247508764266968, + "learning_rate": 2.918627062179491e-08, + "loss": 0.1126, + "step": 42331 + }, + { + "epoch": 0.9923287446450653, + "grad_norm": 2.381309747695923, + "learning_rate": 2.900858543680718e-08, + "loss": 0.1456, + "step": 42332 + }, + { + "epoch": 0.9923521862198702, + "grad_norm": 0.8658315539360046, + "learning_rate": 2.883144269723026e-08, + "loss": 0.0295, + "step": 42333 + }, + { + "epoch": 0.9923756277946753, + "grad_norm": 1.9735528230667114, + "learning_rate": 2.8654842404030047e-08, + "loss": 0.1769, + "step": 42334 + }, + { + "epoch": 0.9923990693694802, + "grad_norm": 2.670682907104492, + "learning_rate": 2.847878455816133e-08, + "loss": 1.4204, + "step": 42335 + }, + { + "epoch": 0.9924225109442852, + "grad_norm": 0.6299597024917603, + "learning_rate": 2.8303269160578903e-08, + "loss": 0.0317, + "step": 42336 + }, + { + "epoch": 0.9924459525190902, + "grad_norm": 2.3873651027679443, + "learning_rate": 2.8128296212226458e-08, + "loss": 0.1684, + "step": 42337 + }, + { + "epoch": 0.9924693940938952, + "grad_norm": 1.835183024406433, + "learning_rate": 2.795386571405878e-08, + "loss": 0.9776, + "step": 42338 + }, + { + "epoch": 0.9924928356687002, + "grad_norm": 1.795439600944519, + "learning_rate": 2.7779977667030664e-08, + "loss": 0.131, + "step": 42339 + }, + { + "epoch": 0.9925162772435052, + "grad_norm": 3.540198564529419, + "learning_rate": 2.76066320720858e-08, + "loss": 0.397, + "step": 42340 + }, + { + "epoch": 0.9925397188183103, + "grad_norm": 2.1527836322784424, + "learning_rate": 2.7433828930145677e-08, + "loss": 0.2257, + "step": 42341 + }, + { + "epoch": 0.9925631603931152, + "grad_norm": 4.276904582977295, + "learning_rate": 2.726156824216508e-08, + "loss": 0.3, + "step": 42342 + }, + { + "epoch": 0.9925866019679203, + "grad_norm": 2.565983295440674, + "learning_rate": 2.70898500090766e-08, + "loss": 0.2759, + "step": 42343 + }, + { + "epoch": 0.9926100435427252, + "grad_norm": 2.1360538005828857, + "learning_rate": 2.6918674231812823e-08, + "loss": 0.1138, + "step": 42344 + }, + { + "epoch": 0.9926334851175302, + "grad_norm": 0.9965245127677917, + "learning_rate": 2.6748040911295234e-08, + "loss": 0.0523, + "step": 42345 + }, + { + "epoch": 0.9926569266923352, + "grad_norm": 0.7683595418930054, + "learning_rate": 2.657795004844532e-08, + "loss": 0.055, + "step": 42346 + }, + { + "epoch": 0.9926803682671402, + "grad_norm": 4.151097297668457, + "learning_rate": 2.640840164420677e-08, + "loss": 0.3028, + "step": 42347 + }, + { + "epoch": 0.9927038098419452, + "grad_norm": 3.1664698123931885, + "learning_rate": 2.6239395699478863e-08, + "loss": 0.1853, + "step": 42348 + }, + { + "epoch": 0.9927272514167502, + "grad_norm": 2.0018279552459717, + "learning_rate": 2.6070932215194187e-08, + "loss": 0.1531, + "step": 42349 + }, + { + "epoch": 0.9927506929915552, + "grad_norm": 0.8984320759773254, + "learning_rate": 2.5903011192252025e-08, + "loss": 0.0722, + "step": 42350 + }, + { + "epoch": 0.9927741345663602, + "grad_norm": 2.777697801589966, + "learning_rate": 2.573563263157386e-08, + "loss": 0.2057, + "step": 42351 + }, + { + "epoch": 0.9927975761411652, + "grad_norm": 1.99907386302948, + "learning_rate": 2.5568796534070072e-08, + "loss": 0.239, + "step": 42352 + }, + { + "epoch": 0.9928210177159702, + "grad_norm": 1.1220003366470337, + "learning_rate": 2.5402502900639945e-08, + "loss": 0.096, + "step": 42353 + }, + { + "epoch": 0.9928444592907751, + "grad_norm": 2.683661937713623, + "learning_rate": 2.523675173218276e-08, + "loss": 1.6939, + "step": 42354 + }, + { + "epoch": 0.9928679008655802, + "grad_norm": 1.0581445693969727, + "learning_rate": 2.50715430295978e-08, + "loss": 0.0851, + "step": 42355 + }, + { + "epoch": 0.9928913424403851, + "grad_norm": 1.8371500968933105, + "learning_rate": 2.490687679379544e-08, + "loss": 0.2424, + "step": 42356 + }, + { + "epoch": 0.9929147840151902, + "grad_norm": 0.8102803826332092, + "learning_rate": 2.4742753025652766e-08, + "loss": 0.0444, + "step": 42357 + }, + { + "epoch": 0.9929382255899951, + "grad_norm": 3.8073678016662598, + "learning_rate": 2.4579171726069052e-08, + "loss": 0.2513, + "step": 42358 + }, + { + "epoch": 0.9929616671648002, + "grad_norm": 1.220821738243103, + "learning_rate": 2.4416132895932474e-08, + "loss": 0.0808, + "step": 42359 + }, + { + "epoch": 0.9929851087396051, + "grad_norm": 0.9042794108390808, + "learning_rate": 2.4253636536131218e-08, + "loss": 0.0935, + "step": 42360 + }, + { + "epoch": 0.9930085503144102, + "grad_norm": 3.4582157135009766, + "learning_rate": 2.4091682647531254e-08, + "loss": 0.2859, + "step": 42361 + }, + { + "epoch": 0.9930319918892151, + "grad_norm": 1.9825690984725952, + "learning_rate": 2.3930271231020762e-08, + "loss": 0.234, + "step": 42362 + }, + { + "epoch": 0.9930554334640201, + "grad_norm": 2.84464955329895, + "learning_rate": 2.3769402287476817e-08, + "loss": 0.2021, + "step": 42363 + }, + { + "epoch": 0.9930788750388251, + "grad_norm": 3.352778196334839, + "learning_rate": 2.3609075817765392e-08, + "loss": 0.173, + "step": 42364 + }, + { + "epoch": 0.9931023166136301, + "grad_norm": 1.1483486890792847, + "learning_rate": 2.3449291822774665e-08, + "loss": 0.0572, + "step": 42365 + }, + { + "epoch": 0.9931257581884351, + "grad_norm": 2.9513516426086426, + "learning_rate": 2.329005030334841e-08, + "loss": 0.1702, + "step": 42366 + }, + { + "epoch": 0.9931491997632401, + "grad_norm": 0.41795992851257324, + "learning_rate": 2.3131351260363697e-08, + "loss": 0.0309, + "step": 42367 + }, + { + "epoch": 0.9931726413380451, + "grad_norm": 1.4477394819259644, + "learning_rate": 2.2973194694686507e-08, + "loss": 0.1404, + "step": 42368 + }, + { + "epoch": 0.9931960829128501, + "grad_norm": 2.835533618927002, + "learning_rate": 2.2815580607160604e-08, + "loss": 0.244, + "step": 42369 + }, + { + "epoch": 0.993219524487655, + "grad_norm": 1.6993176937103271, + "learning_rate": 2.265850899865196e-08, + "loss": 0.8837, + "step": 42370 + }, + { + "epoch": 0.9932429660624601, + "grad_norm": 0.4204801917076111, + "learning_rate": 2.2501979870004354e-08, + "loss": 0.0251, + "step": 42371 + }, + { + "epoch": 0.993266407637265, + "grad_norm": 0.6003291606903076, + "learning_rate": 2.234599322208375e-08, + "loss": 0.0536, + "step": 42372 + }, + { + "epoch": 0.9932898492120701, + "grad_norm": 2.208984613418579, + "learning_rate": 2.2190549055711718e-08, + "loss": 0.1813, + "step": 42373 + }, + { + "epoch": 0.993313290786875, + "grad_norm": 1.3836504220962524, + "learning_rate": 2.203564737175423e-08, + "loss": 0.0976, + "step": 42374 + }, + { + "epoch": 0.9933367323616801, + "grad_norm": 2.9745543003082275, + "learning_rate": 2.1881288171032855e-08, + "loss": 0.2177, + "step": 42375 + }, + { + "epoch": 0.993360173936485, + "grad_norm": 11.210833549499512, + "learning_rate": 2.172747145440246e-08, + "loss": 0.3572, + "step": 42376 + }, + { + "epoch": 0.9933836155112901, + "grad_norm": 1.180356740951538, + "learning_rate": 2.1574197222695714e-08, + "loss": 0.1251, + "step": 42377 + }, + { + "epoch": 0.993407057086095, + "grad_norm": 3.1776156425476074, + "learning_rate": 2.142146547672308e-08, + "loss": 0.2936, + "step": 42378 + }, + { + "epoch": 0.9934304986609, + "grad_norm": 3.4161922931671143, + "learning_rate": 2.1269276217339428e-08, + "loss": 0.1918, + "step": 42379 + }, + { + "epoch": 0.993453940235705, + "grad_norm": 1.7368783950805664, + "learning_rate": 2.1117629445355225e-08, + "loss": 0.1698, + "step": 42380 + }, + { + "epoch": 0.99347738181051, + "grad_norm": 2.4543845653533936, + "learning_rate": 2.0966525161603136e-08, + "loss": 0.1372, + "step": 42381 + }, + { + "epoch": 0.993500823385315, + "grad_norm": 3.4325673580169678, + "learning_rate": 2.0815963366893622e-08, + "loss": 0.1835, + "step": 42382 + }, + { + "epoch": 0.99352426496012, + "grad_norm": 1.851020336151123, + "learning_rate": 2.0665944062048247e-08, + "loss": 1.5435, + "step": 42383 + }, + { + "epoch": 0.993547706534925, + "grad_norm": 0.3683747947216034, + "learning_rate": 2.051646724787748e-08, + "loss": 0.0287, + "step": 42384 + }, + { + "epoch": 0.99357114810973, + "grad_norm": 1.0731033086776733, + "learning_rate": 2.0367532925202882e-08, + "loss": 0.077, + "step": 42385 + }, + { + "epoch": 0.993594589684535, + "grad_norm": 3.083026170730591, + "learning_rate": 2.0219141094823812e-08, + "loss": 0.2229, + "step": 42386 + }, + { + "epoch": 0.99361803125934, + "grad_norm": 1.4794247150421143, + "learning_rate": 2.0071291757539633e-08, + "loss": 0.0526, + "step": 42387 + }, + { + "epoch": 0.993641472834145, + "grad_norm": 3.9293019771575928, + "learning_rate": 1.992398491416081e-08, + "loss": 0.2847, + "step": 42388 + }, + { + "epoch": 0.99366491440895, + "grad_norm": 2.2480108737945557, + "learning_rate": 1.97772205654978e-08, + "loss": 0.1695, + "step": 42389 + }, + { + "epoch": 0.9936883559837549, + "grad_norm": 2.8338446617126465, + "learning_rate": 1.9630998712316663e-08, + "loss": 1.506, + "step": 42390 + }, + { + "epoch": 0.99371179755856, + "grad_norm": 2.9949138164520264, + "learning_rate": 1.9485319355438958e-08, + "loss": 0.2938, + "step": 42391 + }, + { + "epoch": 0.993735239133365, + "grad_norm": 2.2701992988586426, + "learning_rate": 1.9340182495652948e-08, + "loss": 0.1086, + "step": 42392 + }, + { + "epoch": 0.99375868070817, + "grad_norm": 3.347991704940796, + "learning_rate": 1.9195588133724684e-08, + "loss": 0.3641, + "step": 42393 + }, + { + "epoch": 0.993782122282975, + "grad_norm": 2.0339314937591553, + "learning_rate": 1.9051536270464633e-08, + "loss": 0.1709, + "step": 42394 + }, + { + "epoch": 0.99380556385778, + "grad_norm": 3.0154500007629395, + "learning_rate": 1.890802690662774e-08, + "loss": 0.2858, + "step": 42395 + }, + { + "epoch": 0.993829005432585, + "grad_norm": 3.051215171813965, + "learning_rate": 1.876506004301337e-08, + "loss": 0.1387, + "step": 42396 + }, + { + "epoch": 0.99385244700739, + "grad_norm": 2.146303415298462, + "learning_rate": 1.862263568039868e-08, + "loss": 0.1294, + "step": 42397 + }, + { + "epoch": 0.993875888582195, + "grad_norm": 1.025585412979126, + "learning_rate": 1.848075381953862e-08, + "loss": 0.0412, + "step": 42398 + }, + { + "epoch": 0.9938993301569999, + "grad_norm": 0.45155537128448486, + "learning_rate": 1.833941446122145e-08, + "loss": 0.0549, + "step": 42399 + }, + { + "epoch": 0.993922771731805, + "grad_norm": 1.2174056768417358, + "learning_rate": 1.8198617606202118e-08, + "loss": 0.1083, + "step": 42400 + }, + { + "epoch": 0.9939462133066099, + "grad_norm": 2.518589973449707, + "learning_rate": 1.805836325524668e-08, + "loss": 0.1517, + "step": 42401 + }, + { + "epoch": 0.993969654881415, + "grad_norm": 0.9633538722991943, + "learning_rate": 1.791865140912119e-08, + "loss": 0.0313, + "step": 42402 + }, + { + "epoch": 0.9939930964562199, + "grad_norm": 0.25931209325790405, + "learning_rate": 1.77794820685695e-08, + "loss": 0.0139, + "step": 42403 + }, + { + "epoch": 0.994016538031025, + "grad_norm": 1.6809515953063965, + "learning_rate": 1.7640855234368758e-08, + "loss": 1.1226, + "step": 42404 + }, + { + "epoch": 0.9940399796058299, + "grad_norm": 4.474240779876709, + "learning_rate": 1.750277090726282e-08, + "loss": 0.6731, + "step": 42405 + }, + { + "epoch": 0.994063421180635, + "grad_norm": 2.0085299015045166, + "learning_rate": 1.7365229087984436e-08, + "loss": 0.1497, + "step": 42406 + }, + { + "epoch": 0.9940868627554399, + "grad_norm": 0.6582728028297424, + "learning_rate": 1.7228229777310754e-08, + "loss": 0.0386, + "step": 42407 + }, + { + "epoch": 0.9941103043302449, + "grad_norm": 2.2933897972106934, + "learning_rate": 1.7091772975952323e-08, + "loss": 1.326, + "step": 42408 + }, + { + "epoch": 0.9941337459050499, + "grad_norm": 1.1553335189819336, + "learning_rate": 1.6955858684675198e-08, + "loss": 0.0725, + "step": 42409 + }, + { + "epoch": 0.9941571874798549, + "grad_norm": 3.8907878398895264, + "learning_rate": 1.6820486904201015e-08, + "loss": 1.4135, + "step": 42410 + }, + { + "epoch": 0.9941806290546599, + "grad_norm": 5.552264213562012, + "learning_rate": 1.6685657635273633e-08, + "loss": 0.4886, + "step": 42411 + }, + { + "epoch": 0.9942040706294649, + "grad_norm": 3.1300759315490723, + "learning_rate": 1.6551370878614692e-08, + "loss": 0.3671, + "step": 42412 + }, + { + "epoch": 0.9942275122042699, + "grad_norm": 2.0604302883148193, + "learning_rate": 1.6417626634968043e-08, + "loss": 0.0973, + "step": 42413 + }, + { + "epoch": 0.9942509537790749, + "grad_norm": 4.44672966003418, + "learning_rate": 1.6284424905044227e-08, + "loss": 0.3008, + "step": 42414 + }, + { + "epoch": 0.9942743953538798, + "grad_norm": 1.294476866722107, + "learning_rate": 1.6151765689575993e-08, + "loss": 0.0697, + "step": 42415 + }, + { + "epoch": 0.9942978369286849, + "grad_norm": 3.582550525665283, + "learning_rate": 1.6019648989284985e-08, + "loss": 0.1915, + "step": 42416 + }, + { + "epoch": 0.9943212785034898, + "grad_norm": 1.269694447517395, + "learning_rate": 1.5888074804870645e-08, + "loss": 0.0493, + "step": 42417 + }, + { + "epoch": 0.9943447200782949, + "grad_norm": 1.2502914667129517, + "learning_rate": 1.575704313706572e-08, + "loss": 0.1011, + "step": 42418 + }, + { + "epoch": 0.9943681616530998, + "grad_norm": 3.6454381942749023, + "learning_rate": 1.562655398658075e-08, + "loss": 0.2837, + "step": 42419 + }, + { + "epoch": 0.9943916032279049, + "grad_norm": 2.8148257732391357, + "learning_rate": 1.5496607354104076e-08, + "loss": 0.1408, + "step": 42420 + }, + { + "epoch": 0.9944150448027098, + "grad_norm": 1.894032597541809, + "learning_rate": 1.5367203240368445e-08, + "loss": 0.1245, + "step": 42421 + }, + { + "epoch": 0.9944384863775149, + "grad_norm": 5.065504550933838, + "learning_rate": 1.523834164605109e-08, + "loss": 0.4694, + "step": 42422 + }, + { + "epoch": 0.9944619279523198, + "grad_norm": 3.749546766281128, + "learning_rate": 1.511002257186256e-08, + "loss": 0.2996, + "step": 42423 + }, + { + "epoch": 0.9944853695271249, + "grad_norm": 2.8591887950897217, + "learning_rate": 1.4982246018502287e-08, + "loss": 0.1656, + "step": 42424 + }, + { + "epoch": 0.9945088111019298, + "grad_norm": 1.6884433031082153, + "learning_rate": 1.4855011986658618e-08, + "loss": 0.0697, + "step": 42425 + }, + { + "epoch": 0.9945322526767348, + "grad_norm": 0.5228205323219299, + "learning_rate": 1.4728320477019886e-08, + "loss": 0.0178, + "step": 42426 + }, + { + "epoch": 0.9945556942515398, + "grad_norm": 1.7071858644485474, + "learning_rate": 1.460217149027443e-08, + "loss": 0.0957, + "step": 42427 + }, + { + "epoch": 0.9945791358263448, + "grad_norm": 0.9932582378387451, + "learning_rate": 1.447656502711059e-08, + "loss": 0.1107, + "step": 42428 + }, + { + "epoch": 0.9946025774011498, + "grad_norm": 3.65665340423584, + "learning_rate": 1.4351501088216701e-08, + "loss": 0.3972, + "step": 42429 + }, + { + "epoch": 0.9946260189759548, + "grad_norm": 2.1754190921783447, + "learning_rate": 1.42269796742589e-08, + "loss": 0.1936, + "step": 42430 + }, + { + "epoch": 0.9946494605507598, + "grad_norm": 0.38197052478790283, + "learning_rate": 1.4103000785914422e-08, + "loss": 0.0187, + "step": 42431 + }, + { + "epoch": 0.9946729021255648, + "grad_norm": 3.0465505123138428, + "learning_rate": 1.3979564423860503e-08, + "loss": 0.2026, + "step": 42432 + }, + { + "epoch": 0.9946963437003697, + "grad_norm": 2.4254119396209717, + "learning_rate": 1.3856670588763276e-08, + "loss": 1.7671, + "step": 42433 + }, + { + "epoch": 0.9947197852751748, + "grad_norm": 2.3700547218322754, + "learning_rate": 1.3734319281288877e-08, + "loss": 1.7524, + "step": 42434 + }, + { + "epoch": 0.9947432268499797, + "grad_norm": 4.311805248260498, + "learning_rate": 1.3612510502103438e-08, + "loss": 0.3308, + "step": 42435 + }, + { + "epoch": 0.9947666684247848, + "grad_norm": 3.2055106163024902, + "learning_rate": 1.3491244251873092e-08, + "loss": 0.1376, + "step": 42436 + }, + { + "epoch": 0.9947901099995897, + "grad_norm": 2.734506607055664, + "learning_rate": 1.337052053124177e-08, + "loss": 0.3019, + "step": 42437 + }, + { + "epoch": 0.9948135515743948, + "grad_norm": 0.5782621502876282, + "learning_rate": 1.3250339340886709e-08, + "loss": 0.0233, + "step": 42438 + }, + { + "epoch": 0.9948369931491997, + "grad_norm": 0.518308699131012, + "learning_rate": 1.3130700681440733e-08, + "loss": 0.0454, + "step": 42439 + }, + { + "epoch": 0.9948604347240048, + "grad_norm": 1.7771172523498535, + "learning_rate": 1.3011604553547774e-08, + "loss": 0.0665, + "step": 42440 + }, + { + "epoch": 0.9948838762988097, + "grad_norm": 0.47171133756637573, + "learning_rate": 1.2893050957885067e-08, + "loss": 0.0426, + "step": 42441 + }, + { + "epoch": 0.9949073178736147, + "grad_norm": 2.7210185527801514, + "learning_rate": 1.2775039895063234e-08, + "loss": 0.2341, + "step": 42442 + }, + { + "epoch": 0.9949307594484198, + "grad_norm": 4.459206581115723, + "learning_rate": 1.2657571365737308e-08, + "loss": 0.3074, + "step": 42443 + }, + { + "epoch": 0.9949542010232247, + "grad_norm": 3.604881763458252, + "learning_rate": 1.2540645370540115e-08, + "loss": 0.2295, + "step": 42444 + }, + { + "epoch": 0.9949776425980298, + "grad_norm": 1.1684577465057373, + "learning_rate": 1.2424261910115587e-08, + "loss": 0.0582, + "step": 42445 + }, + { + "epoch": 0.9950010841728347, + "grad_norm": 0.8224440813064575, + "learning_rate": 1.2308420985085444e-08, + "loss": 0.115, + "step": 42446 + }, + { + "epoch": 0.9950245257476398, + "grad_norm": 0.6195363998413086, + "learning_rate": 1.2193122596082519e-08, + "loss": 0.0393, + "step": 42447 + }, + { + "epoch": 0.9950479673224447, + "grad_norm": 2.2574262619018555, + "learning_rate": 1.2078366743728531e-08, + "loss": 0.1889, + "step": 42448 + }, + { + "epoch": 0.9950714088972498, + "grad_norm": 1.9510200023651123, + "learning_rate": 1.1964153428645208e-08, + "loss": 0.1622, + "step": 42449 + }, + { + "epoch": 0.9950948504720547, + "grad_norm": 3.693692922592163, + "learning_rate": 1.1850482651465378e-08, + "loss": 0.38, + "step": 42450 + }, + { + "epoch": 0.9951182920468598, + "grad_norm": 1.7783879041671753, + "learning_rate": 1.173735441278856e-08, + "loss": 0.1492, + "step": 42451 + }, + { + "epoch": 0.9951417336216647, + "grad_norm": 3.30122709274292, + "learning_rate": 1.1624768713236479e-08, + "loss": 0.2128, + "step": 42452 + }, + { + "epoch": 0.9951651751964697, + "grad_norm": 1.1122827529907227, + "learning_rate": 1.1512725553419756e-08, + "loss": 0.0518, + "step": 42453 + }, + { + "epoch": 0.9951886167712747, + "grad_norm": 3.3848278522491455, + "learning_rate": 1.1401224933949017e-08, + "loss": 0.2906, + "step": 42454 + }, + { + "epoch": 0.9952120583460797, + "grad_norm": 0.6250975728034973, + "learning_rate": 1.1290266855434883e-08, + "loss": 0.0562, + "step": 42455 + }, + { + "epoch": 0.9952354999208847, + "grad_norm": 0.782717227935791, + "learning_rate": 1.1179851318465772e-08, + "loss": 0.0584, + "step": 42456 + }, + { + "epoch": 0.9952589414956897, + "grad_norm": 2.4648149013519287, + "learning_rate": 1.1069978323641206e-08, + "loss": 0.1781, + "step": 42457 + }, + { + "epoch": 0.9952823830704947, + "grad_norm": 3.9961752891540527, + "learning_rate": 1.0960647871560703e-08, + "loss": 0.2203, + "step": 42458 + }, + { + "epoch": 0.9953058246452997, + "grad_norm": 3.7480084896087646, + "learning_rate": 1.0851859962823785e-08, + "loss": 0.4589, + "step": 42459 + }, + { + "epoch": 0.9953292662201046, + "grad_norm": 0.6540626883506775, + "learning_rate": 1.0743614598018869e-08, + "loss": 0.0798, + "step": 42460 + }, + { + "epoch": 0.9953527077949097, + "grad_norm": 1.7195205688476562, + "learning_rate": 1.0635911777723273e-08, + "loss": 0.1011, + "step": 42461 + }, + { + "epoch": 0.9953761493697146, + "grad_norm": 1.6139552593231201, + "learning_rate": 1.0528751502536516e-08, + "loss": 0.0862, + "step": 42462 + }, + { + "epoch": 0.9953995909445197, + "grad_norm": 5.178159236907959, + "learning_rate": 1.0422133773024812e-08, + "loss": 0.4462, + "step": 42463 + }, + { + "epoch": 0.9954230325193246, + "grad_norm": 2.576871871948242, + "learning_rate": 1.0316058589776579e-08, + "loss": 0.1181, + "step": 42464 + }, + { + "epoch": 0.9954464740941297, + "grad_norm": 3.4699342250823975, + "learning_rate": 1.0210525953369133e-08, + "loss": 0.3249, + "step": 42465 + }, + { + "epoch": 0.9954699156689346, + "grad_norm": 3.2183382511138916, + "learning_rate": 1.0105535864368688e-08, + "loss": 0.2381, + "step": 42466 + }, + { + "epoch": 0.9954933572437397, + "grad_norm": 2.3649845123291016, + "learning_rate": 1.0001088323341457e-08, + "loss": 0.1437, + "step": 42467 + }, + { + "epoch": 0.9955167988185446, + "grad_norm": 3.679969549179077, + "learning_rate": 9.897183330864757e-09, + "loss": 0.2571, + "step": 42468 + }, + { + "epoch": 0.9955402403933497, + "grad_norm": 1.8845468759536743, + "learning_rate": 9.793820887493698e-09, + "loss": 0.1218, + "step": 42469 + }, + { + "epoch": 0.9955636819681546, + "grad_norm": 0.6969287395477295, + "learning_rate": 9.691000993794497e-09, + "loss": 0.0358, + "step": 42470 + }, + { + "epoch": 0.9955871235429596, + "grad_norm": 3.079580783843994, + "learning_rate": 9.588723650322262e-09, + "loss": 0.1506, + "step": 42471 + }, + { + "epoch": 0.9956105651177646, + "grad_norm": 2.480194330215454, + "learning_rate": 9.486988857643208e-09, + "loss": 0.169, + "step": 42472 + }, + { + "epoch": 0.9956340066925696, + "grad_norm": 2.3826448917388916, + "learning_rate": 9.385796616290244e-09, + "loss": 0.1797, + "step": 42473 + }, + { + "epoch": 0.9956574482673746, + "grad_norm": 0.3312165439128876, + "learning_rate": 9.285146926818478e-09, + "loss": 0.0355, + "step": 42474 + }, + { + "epoch": 0.9956808898421796, + "grad_norm": 0.8476864099502563, + "learning_rate": 9.185039789783023e-09, + "loss": 0.085, + "step": 42475 + }, + { + "epoch": 0.9957043314169846, + "grad_norm": 2.616682529449463, + "learning_rate": 9.085475205716787e-09, + "loss": 0.2033, + "step": 42476 + }, + { + "epoch": 0.9957277729917896, + "grad_norm": 1.7701565027236938, + "learning_rate": 8.986453175174881e-09, + "loss": 0.0636, + "step": 42477 + }, + { + "epoch": 0.9957512145665945, + "grad_norm": 3.8396875858306885, + "learning_rate": 8.887973698668006e-09, + "loss": 0.2661, + "step": 42478 + }, + { + "epoch": 0.9957746561413996, + "grad_norm": 1.3488975763320923, + "learning_rate": 8.790036776762379e-09, + "loss": 0.1549, + "step": 42479 + }, + { + "epoch": 0.9957980977162045, + "grad_norm": 0.8331107497215271, + "learning_rate": 8.692642409957596e-09, + "loss": 0.0871, + "step": 42480 + }, + { + "epoch": 0.9958215392910096, + "grad_norm": 2.571810722351074, + "learning_rate": 8.595790598808773e-09, + "loss": 0.1444, + "step": 42481 + }, + { + "epoch": 0.9958449808658145, + "grad_norm": 1.862669587135315, + "learning_rate": 8.49948134382661e-09, + "loss": 0.0777, + "step": 42482 + }, + { + "epoch": 0.9958684224406196, + "grad_norm": 2.883774995803833, + "learning_rate": 8.403714645544014e-09, + "loss": 0.2645, + "step": 42483 + }, + { + "epoch": 0.9958918640154245, + "grad_norm": 1.1137480735778809, + "learning_rate": 8.308490504471689e-09, + "loss": 0.0595, + "step": 42484 + }, + { + "epoch": 0.9959153055902296, + "grad_norm": 1.6066069602966309, + "learning_rate": 8.213808921120337e-09, + "loss": 0.1245, + "step": 42485 + }, + { + "epoch": 0.9959387471650345, + "grad_norm": 1.485880732536316, + "learning_rate": 8.119669896022864e-09, + "loss": 1.2116, + "step": 42486 + }, + { + "epoch": 0.9959621887398395, + "grad_norm": 1.7652363777160645, + "learning_rate": 8.02607342967887e-09, + "loss": 1.1291, + "step": 42487 + }, + { + "epoch": 0.9959856303146445, + "grad_norm": 0.6811317205429077, + "learning_rate": 7.933019522587959e-09, + "loss": 0.0352, + "step": 42488 + }, + { + "epoch": 0.9960090718894495, + "grad_norm": 2.66744065284729, + "learning_rate": 7.840508175271933e-09, + "loss": 0.2173, + "step": 42489 + }, + { + "epoch": 0.9960325134642545, + "grad_norm": 2.1303062438964844, + "learning_rate": 7.74853938821929e-09, + "loss": 0.2065, + "step": 42490 + }, + { + "epoch": 0.9960559550390595, + "grad_norm": 2.335306406021118, + "learning_rate": 7.657113161940732e-09, + "loss": 0.1263, + "step": 42491 + }, + { + "epoch": 0.9960793966138645, + "grad_norm": 2.524503707885742, + "learning_rate": 7.56622949692476e-09, + "loss": 1.6576, + "step": 42492 + }, + { + "epoch": 0.9961028381886695, + "grad_norm": 3.362154245376587, + "learning_rate": 7.475888393659868e-09, + "loss": 0.3029, + "step": 42493 + }, + { + "epoch": 0.9961262797634746, + "grad_norm": 1.29243004322052, + "learning_rate": 7.3860898526456615e-09, + "loss": 0.0483, + "step": 42494 + }, + { + "epoch": 0.9961497213382795, + "grad_norm": 0.9566646814346313, + "learning_rate": 7.296833874370634e-09, + "loss": 0.0462, + "step": 42495 + }, + { + "epoch": 0.9961731629130846, + "grad_norm": 0.6901454925537109, + "learning_rate": 7.2081204593121845e-09, + "loss": 0.0571, + "step": 42496 + }, + { + "epoch": 0.9961966044878895, + "grad_norm": 2.652360200881958, + "learning_rate": 7.119949607947707e-09, + "loss": 0.1575, + "step": 42497 + }, + { + "epoch": 0.9962200460626945, + "grad_norm": 0.9212996363639832, + "learning_rate": 7.032321320776802e-09, + "loss": 0.0864, + "step": 42498 + }, + { + "epoch": 0.9962434876374995, + "grad_norm": 2.1367387771606445, + "learning_rate": 6.94523559824356e-09, + "loss": 0.1538, + "step": 42499 + }, + { + "epoch": 0.9962669292123045, + "grad_norm": 2.9582107067108154, + "learning_rate": 6.858692440847581e-09, + "loss": 0.2951, + "step": 42500 + }, + { + "epoch": 0.9962903707871095, + "grad_norm": 0.3226577043533325, + "learning_rate": 6.772691849044055e-09, + "loss": 0.0272, + "step": 42501 + }, + { + "epoch": 0.9963138123619145, + "grad_norm": 2.851408004760742, + "learning_rate": 6.687233823299277e-09, + "loss": 0.2559, + "step": 42502 + }, + { + "epoch": 0.9963372539367195, + "grad_norm": 2.272965431213379, + "learning_rate": 6.602318364090643e-09, + "loss": 0.1475, + "step": 42503 + }, + { + "epoch": 0.9963606955115245, + "grad_norm": 1.7904331684112549, + "learning_rate": 6.517945471862241e-09, + "loss": 0.1319, + "step": 42504 + }, + { + "epoch": 0.9963841370863294, + "grad_norm": 2.166717767715454, + "learning_rate": 6.434115147080366e-09, + "loss": 0.0968, + "step": 42505 + }, + { + "epoch": 0.9964075786611345, + "grad_norm": 1.5419402122497559, + "learning_rate": 6.350827390200209e-09, + "loss": 0.0923, + "step": 42506 + }, + { + "epoch": 0.9964310202359394, + "grad_norm": 2.6422393321990967, + "learning_rate": 6.2680822016769615e-09, + "loss": 0.1622, + "step": 42507 + }, + { + "epoch": 0.9964544618107445, + "grad_norm": 1.8855005502700806, + "learning_rate": 6.18587958194361e-09, + "loss": 0.1007, + "step": 42508 + }, + { + "epoch": 0.9964779033855494, + "grad_norm": 1.0523024797439575, + "learning_rate": 6.1042195314664486e-09, + "loss": 0.1014, + "step": 42509 + }, + { + "epoch": 0.9965013449603545, + "grad_norm": 2.459707498550415, + "learning_rate": 6.023102050678464e-09, + "loss": 0.2385, + "step": 42510 + }, + { + "epoch": 0.9965247865351594, + "grad_norm": 1.9259247779846191, + "learning_rate": 5.942527140012643e-09, + "loss": 0.1584, + "step": 42511 + }, + { + "epoch": 0.9965482281099645, + "grad_norm": 1.0857244729995728, + "learning_rate": 5.8624947999241785e-09, + "loss": 0.0581, + "step": 42512 + }, + { + "epoch": 0.9965716696847694, + "grad_norm": 1.6467639207839966, + "learning_rate": 5.783005030823852e-09, + "loss": 0.0947, + "step": 42513 + }, + { + "epoch": 0.9965951112595745, + "grad_norm": 0.9848331809043884, + "learning_rate": 5.7040578331668534e-09, + "loss": 0.0536, + "step": 42514 + }, + { + "epoch": 0.9966185528343794, + "grad_norm": 0.9338620901107788, + "learning_rate": 5.62565320737507e-09, + "loss": 0.0529, + "step": 42515 + }, + { + "epoch": 0.9966419944091844, + "grad_norm": 1.5353025197982788, + "learning_rate": 5.547791153859283e-09, + "loss": 0.1176, + "step": 42516 + }, + { + "epoch": 0.9966654359839894, + "grad_norm": 2.532470703125, + "learning_rate": 5.4704716730635815e-09, + "loss": 0.205, + "step": 42517 + }, + { + "epoch": 0.9966888775587944, + "grad_norm": 5.117066383361816, + "learning_rate": 5.393694765387647e-09, + "loss": 0.2145, + "step": 42518 + }, + { + "epoch": 0.9967123191335994, + "grad_norm": 1.7863490581512451, + "learning_rate": 5.317460431264465e-09, + "loss": 0.1132, + "step": 42519 + }, + { + "epoch": 0.9967357607084044, + "grad_norm": 1.9649713039398193, + "learning_rate": 5.241768671093717e-09, + "loss": 0.0889, + "step": 42520 + }, + { + "epoch": 0.9967592022832094, + "grad_norm": 4.010322093963623, + "learning_rate": 5.166619485297286e-09, + "loss": 0.3003, + "step": 42521 + }, + { + "epoch": 0.9967826438580144, + "grad_norm": 2.0591015815734863, + "learning_rate": 5.092012874274854e-09, + "loss": 0.074, + "step": 42522 + }, + { + "epoch": 0.9968060854328193, + "grad_norm": 2.321782112121582, + "learning_rate": 5.017948838437203e-09, + "loss": 0.1927, + "step": 42523 + }, + { + "epoch": 0.9968295270076244, + "grad_norm": 1.8566597700119019, + "learning_rate": 4.944427378184013e-09, + "loss": 0.1417, + "step": 42524 + }, + { + "epoch": 0.9968529685824293, + "grad_norm": 0.5370247960090637, + "learning_rate": 4.871448493914965e-09, + "loss": 0.0359, + "step": 42525 + }, + { + "epoch": 0.9968764101572344, + "grad_norm": 3.011568546295166, + "learning_rate": 4.799012186018636e-09, + "loss": 0.2608, + "step": 42526 + }, + { + "epoch": 0.9968998517320393, + "grad_norm": 2.1576263904571533, + "learning_rate": 4.72711845490581e-09, + "loss": 0.0928, + "step": 42527 + }, + { + "epoch": 0.9969232933068444, + "grad_norm": 1.0169057846069336, + "learning_rate": 4.655767300942859e-09, + "loss": 0.0954, + "step": 42528 + }, + { + "epoch": 0.9969467348816493, + "grad_norm": 3.8581061363220215, + "learning_rate": 4.584958724540566e-09, + "loss": 0.4725, + "step": 42529 + }, + { + "epoch": 0.9969701764564544, + "grad_norm": 6.860947132110596, + "learning_rate": 4.514692726065306e-09, + "loss": 0.2677, + "step": 42530 + }, + { + "epoch": 0.9969936180312593, + "grad_norm": 1.7323728799819946, + "learning_rate": 4.4449693059056556e-09, + "loss": 0.0737, + "step": 42531 + }, + { + "epoch": 0.9970170596060643, + "grad_norm": 1.6385031938552856, + "learning_rate": 4.37578846443909e-09, + "loss": 0.1327, + "step": 42532 + }, + { + "epoch": 0.9970405011808693, + "grad_norm": 0.42251500487327576, + "learning_rate": 4.3071502020430865e-09, + "loss": 0.0165, + "step": 42533 + }, + { + "epoch": 0.9970639427556743, + "grad_norm": 3.033186674118042, + "learning_rate": 4.239054519084018e-09, + "loss": 0.2137, + "step": 42534 + }, + { + "epoch": 0.9970873843304793, + "grad_norm": 4.647757053375244, + "learning_rate": 4.171501415939361e-09, + "loss": 0.4757, + "step": 42535 + }, + { + "epoch": 0.9971108259052843, + "grad_norm": 0.617689847946167, + "learning_rate": 4.1044908929643856e-09, + "loss": 0.0326, + "step": 42536 + }, + { + "epoch": 0.9971342674800893, + "grad_norm": 1.604709267616272, + "learning_rate": 4.038022950536568e-09, + "loss": 0.7352, + "step": 42537 + }, + { + "epoch": 0.9971577090548943, + "grad_norm": 1.9094854593276978, + "learning_rate": 3.9720975890111815e-09, + "loss": 0.0932, + "step": 42538 + }, + { + "epoch": 0.9971811506296993, + "grad_norm": 4.090206146240234, + "learning_rate": 3.906714808743494e-09, + "loss": 0.2625, + "step": 42539 + }, + { + "epoch": 0.9972045922045043, + "grad_norm": 1.5510300397872925, + "learning_rate": 3.841874610088781e-09, + "loss": 0.1046, + "step": 42540 + }, + { + "epoch": 0.9972280337793092, + "grad_norm": 1.9054867029190063, + "learning_rate": 3.77757699340231e-09, + "loss": 0.1424, + "step": 42541 + }, + { + "epoch": 0.9972514753541143, + "grad_norm": 0.619951605796814, + "learning_rate": 3.7138219590282517e-09, + "loss": 0.0437, + "step": 42542 + }, + { + "epoch": 0.9972749169289192, + "grad_norm": 1.0843721628189087, + "learning_rate": 3.650609507321878e-09, + "loss": 0.0802, + "step": 42543 + }, + { + "epoch": 0.9972983585037243, + "grad_norm": 1.5143513679504395, + "learning_rate": 3.5879396386051534e-09, + "loss": 0.1395, + "step": 42544 + }, + { + "epoch": 0.9973218000785293, + "grad_norm": 3.2383430004119873, + "learning_rate": 3.525812353244451e-09, + "loss": 0.1417, + "step": 42545 + }, + { + "epoch": 0.9973452416533343, + "grad_norm": 0.8099052906036377, + "learning_rate": 3.4642276515506334e-09, + "loss": 0.0731, + "step": 42546 + }, + { + "epoch": 0.9973686832281393, + "grad_norm": 3.3202664852142334, + "learning_rate": 3.4031855338900743e-09, + "loss": 0.1917, + "step": 42547 + }, + { + "epoch": 0.9973921248029443, + "grad_norm": 1.776336669921875, + "learning_rate": 3.342686000562534e-09, + "loss": 0.0583, + "step": 42548 + }, + { + "epoch": 0.9974155663777493, + "grad_norm": 3.540400266647339, + "learning_rate": 3.282729051912181e-09, + "loss": 0.3692, + "step": 42549 + }, + { + "epoch": 0.9974390079525542, + "grad_norm": 0.514330267906189, + "learning_rate": 3.223314688260981e-09, + "loss": 0.0257, + "step": 42550 + }, + { + "epoch": 0.9974624495273593, + "grad_norm": 2.4423887729644775, + "learning_rate": 3.164442909930898e-09, + "loss": 0.2798, + "step": 42551 + }, + { + "epoch": 0.9974858911021642, + "grad_norm": 3.5127785205841064, + "learning_rate": 3.106113717243897e-09, + "loss": 1.4333, + "step": 42552 + }, + { + "epoch": 0.9975093326769693, + "grad_norm": 2.3790011405944824, + "learning_rate": 3.0483271105108404e-09, + "loss": 0.1528, + "step": 42553 + }, + { + "epoch": 0.9975327742517742, + "grad_norm": 0.6446169018745422, + "learning_rate": 2.991083090053692e-09, + "loss": 0.0181, + "step": 42554 + }, + { + "epoch": 0.9975562158265793, + "grad_norm": 3.107334613800049, + "learning_rate": 2.9343816561722138e-09, + "loss": 0.3405, + "step": 42555 + }, + { + "epoch": 0.9975796574013842, + "grad_norm": 2.934650421142578, + "learning_rate": 2.878222809188369e-09, + "loss": 0.2772, + "step": 42556 + }, + { + "epoch": 0.9976030989761893, + "grad_norm": 3.1837046146392822, + "learning_rate": 2.8226065494019184e-09, + "loss": 0.3335, + "step": 42557 + }, + { + "epoch": 0.9976265405509942, + "grad_norm": 2.7842535972595215, + "learning_rate": 2.7675328771015197e-09, + "loss": 0.1819, + "step": 42558 + }, + { + "epoch": 0.9976499821257993, + "grad_norm": 2.2872447967529297, + "learning_rate": 2.7130017925980357e-09, + "loss": 0.1197, + "step": 42559 + }, + { + "epoch": 0.9976734237006042, + "grad_norm": 0.8993238806724548, + "learning_rate": 2.6590132961912263e-09, + "loss": 0.0746, + "step": 42560 + }, + { + "epoch": 0.9976968652754092, + "grad_norm": 1.7436739206314087, + "learning_rate": 2.6055673881586474e-09, + "loss": 1.4195, + "step": 42561 + }, + { + "epoch": 0.9977203068502142, + "grad_norm": 1.5647228956222534, + "learning_rate": 2.5526640688111614e-09, + "loss": 1.1256, + "step": 42562 + }, + { + "epoch": 0.9977437484250192, + "grad_norm": 3.5010974407196045, + "learning_rate": 2.5003033384152218e-09, + "loss": 0.2903, + "step": 42563 + }, + { + "epoch": 0.9977671899998242, + "grad_norm": 2.8839118480682373, + "learning_rate": 2.448485197270589e-09, + "loss": 0.2339, + "step": 42564 + }, + { + "epoch": 0.9977906315746292, + "grad_norm": 4.1483025550842285, + "learning_rate": 2.3972096456548187e-09, + "loss": 0.3399, + "step": 42565 + }, + { + "epoch": 0.9978140731494342, + "grad_norm": 1.1905837059020996, + "learning_rate": 2.3464766838343643e-09, + "loss": 0.1161, + "step": 42566 + }, + { + "epoch": 0.9978375147242392, + "grad_norm": 0.618249773979187, + "learning_rate": 2.296286312097884e-09, + "loss": 0.0507, + "step": 42567 + }, + { + "epoch": 0.9978609562990441, + "grad_norm": 2.7285983562469482, + "learning_rate": 2.246638530711831e-09, + "loss": 0.3239, + "step": 42568 + }, + { + "epoch": 0.9978843978738492, + "grad_norm": 3.3466145992279053, + "learning_rate": 2.1975333399537614e-09, + "loss": 1.1806, + "step": 42569 + }, + { + "epoch": 0.9979078394486541, + "grad_norm": 2.1250643730163574, + "learning_rate": 2.148970740079026e-09, + "loss": 0.0922, + "step": 42570 + }, + { + "epoch": 0.9979312810234592, + "grad_norm": 3.245269298553467, + "learning_rate": 2.100950731354079e-09, + "loss": 0.3779, + "step": 42571 + }, + { + "epoch": 0.9979547225982641, + "grad_norm": 2.3608546257019043, + "learning_rate": 2.053473314045373e-09, + "loss": 1.666, + "step": 42572 + }, + { + "epoch": 0.9979781641730692, + "grad_norm": 5.432130336761475, + "learning_rate": 2.006538488397158e-09, + "loss": 0.4091, + "step": 42573 + }, + { + "epoch": 0.9980016057478741, + "grad_norm": 3.887486219406128, + "learning_rate": 1.96014625468699e-09, + "loss": 0.2874, + "step": 42574 + }, + { + "epoch": 0.9980250473226792, + "grad_norm": 1.1106646060943604, + "learning_rate": 1.9142966131480144e-09, + "loss": 0.057, + "step": 42575 + }, + { + "epoch": 0.9980484888974841, + "grad_norm": 2.528735399246216, + "learning_rate": 1.8689895640244813e-09, + "loss": 0.1559, + "step": 42576 + }, + { + "epoch": 0.9980719304722891, + "grad_norm": 4.391139507293701, + "learning_rate": 1.8242251075828442e-09, + "loss": 0.2729, + "step": 42577 + }, + { + "epoch": 0.9980953720470941, + "grad_norm": 1.5248396396636963, + "learning_rate": 1.7800032440451475e-09, + "loss": 0.0564, + "step": 42578 + }, + { + "epoch": 0.9981188136218991, + "grad_norm": 0.9812974333763123, + "learning_rate": 1.7363239736667425e-09, + "loss": 0.0427, + "step": 42579 + }, + { + "epoch": 0.9981422551967041, + "grad_norm": 3.548988103866577, + "learning_rate": 1.693187296680776e-09, + "loss": 0.1327, + "step": 42580 + }, + { + "epoch": 0.9981656967715091, + "grad_norm": 2.979276657104492, + "learning_rate": 1.6505932133092928e-09, + "loss": 0.2245, + "step": 42581 + }, + { + "epoch": 0.9981891383463141, + "grad_norm": 1.8117095232009888, + "learning_rate": 1.6085417238076438e-09, + "loss": 0.098, + "step": 42582 + }, + { + "epoch": 0.9982125799211191, + "grad_norm": 1.5451363325119019, + "learning_rate": 1.5670328283756697e-09, + "loss": 0.8934, + "step": 42583 + }, + { + "epoch": 0.998236021495924, + "grad_norm": 3.832414150238037, + "learning_rate": 1.526066527257619e-09, + "loss": 0.2627, + "step": 42584 + }, + { + "epoch": 0.9982594630707291, + "grad_norm": 2.2625274658203125, + "learning_rate": 1.485642820675537e-09, + "loss": 0.2199, + "step": 42585 + }, + { + "epoch": 0.998282904645534, + "grad_norm": 0.5041182041168213, + "learning_rate": 1.4457617088403653e-09, + "loss": 0.0648, + "step": 42586 + }, + { + "epoch": 0.9983063462203391, + "grad_norm": 0.9075811505317688, + "learning_rate": 1.4064231919741488e-09, + "loss": 0.0818, + "step": 42587 + }, + { + "epoch": 0.998329787795144, + "grad_norm": 0.6439443826675415, + "learning_rate": 1.3676272702878301e-09, + "loss": 0.0414, + "step": 42588 + }, + { + "epoch": 0.9983532293699491, + "grad_norm": 3.350443124771118, + "learning_rate": 1.3293739439923514e-09, + "loss": 0.3069, + "step": 42589 + }, + { + "epoch": 0.998376670944754, + "grad_norm": 2.898890495300293, + "learning_rate": 1.291663213298655e-09, + "loss": 0.2167, + "step": 42590 + }, + { + "epoch": 0.9984001125195591, + "grad_norm": 2.6353578567504883, + "learning_rate": 1.254495078395479e-09, + "loss": 0.1336, + "step": 42591 + }, + { + "epoch": 0.998423554094364, + "grad_norm": 2.407470464706421, + "learning_rate": 1.2178695395048678e-09, + "loss": 0.3315, + "step": 42592 + }, + { + "epoch": 0.9984469956691691, + "grad_norm": 3.097933769226074, + "learning_rate": 1.1817865968266616e-09, + "loss": 0.3844, + "step": 42593 + }, + { + "epoch": 0.998470437243974, + "grad_norm": 2.466913938522339, + "learning_rate": 1.1462462505384963e-09, + "loss": 0.3829, + "step": 42594 + }, + { + "epoch": 0.998493878818779, + "grad_norm": 2.2140352725982666, + "learning_rate": 1.1112485008402118e-09, + "loss": 0.289, + "step": 42595 + }, + { + "epoch": 0.9985173203935841, + "grad_norm": 4.39260721206665, + "learning_rate": 1.0767933479316483e-09, + "loss": 0.5107, + "step": 42596 + }, + { + "epoch": 0.998540761968389, + "grad_norm": 2.1185226440429688, + "learning_rate": 1.0428807919904416e-09, + "loss": 1.1598, + "step": 42597 + }, + { + "epoch": 0.9985642035431941, + "grad_norm": 1.7981853485107422, + "learning_rate": 1.0095108331942271e-09, + "loss": 0.1751, + "step": 42598 + }, + { + "epoch": 0.998587645117999, + "grad_norm": 2.0052530765533447, + "learning_rate": 9.766834717428453e-10, + "loss": 0.0501, + "step": 42599 + }, + { + "epoch": 0.9986110866928041, + "grad_norm": 1.0125092267990112, + "learning_rate": 9.443987077917272e-10, + "loss": 0.0673, + "step": 42600 + }, + { + "epoch": 0.998634528267609, + "grad_norm": 4.0515594482421875, + "learning_rate": 9.126565415407129e-10, + "loss": 0.346, + "step": 42601 + }, + { + "epoch": 0.9986579698424141, + "grad_norm": 1.2149946689605713, + "learning_rate": 8.814569731452337e-10, + "loss": 0.0926, + "step": 42602 + }, + { + "epoch": 0.998681411417219, + "grad_norm": 1.53141188621521, + "learning_rate": 8.508000027718233e-10, + "loss": 0.0692, + "step": 42603 + }, + { + "epoch": 0.998704852992024, + "grad_norm": 2.388928174972534, + "learning_rate": 8.20685630598117e-10, + "loss": 1.5875, + "step": 42604 + }, + { + "epoch": 0.998728294566829, + "grad_norm": 0.8796454071998596, + "learning_rate": 7.911138567906484e-10, + "loss": 0.063, + "step": 42605 + }, + { + "epoch": 0.998751736141634, + "grad_norm": 1.1925865411758423, + "learning_rate": 7.620846814937466e-10, + "loss": 0.0376, + "step": 42606 + }, + { + "epoch": 0.998775177716439, + "grad_norm": 3.397768259048462, + "learning_rate": 7.335981048739449e-10, + "loss": 0.3712, + "step": 42607 + }, + { + "epoch": 0.998798619291244, + "grad_norm": 1.0405025482177734, + "learning_rate": 7.056541270866746e-10, + "loss": 0.0798, + "step": 42608 + }, + { + "epoch": 0.998822060866049, + "grad_norm": 4.439299583435059, + "learning_rate": 6.782527482873669e-10, + "loss": 0.1641, + "step": 42609 + }, + { + "epoch": 0.998845502440854, + "grad_norm": 1.4889435768127441, + "learning_rate": 6.513939686092485e-10, + "loss": 0.1477, + "step": 42610 + }, + { + "epoch": 0.998868944015659, + "grad_norm": 3.0820631980895996, + "learning_rate": 6.25077788218853e-10, + "loss": 0.2193, + "step": 42611 + }, + { + "epoch": 0.998892385590464, + "grad_norm": 3.4667301177978516, + "learning_rate": 5.993042072383048e-10, + "loss": 0.1899, + "step": 42612 + }, + { + "epoch": 0.9989158271652689, + "grad_norm": 2.363612413406372, + "learning_rate": 5.740732258230353e-10, + "loss": 0.2512, + "step": 42613 + }, + { + "epoch": 0.998939268740074, + "grad_norm": 2.6657638549804688, + "learning_rate": 5.49384844106271e-10, + "loss": 0.2198, + "step": 42614 + }, + { + "epoch": 0.9989627103148789, + "grad_norm": 1.8845077753067017, + "learning_rate": 5.252390622212388e-10, + "loss": 0.0811, + "step": 42615 + }, + { + "epoch": 0.998986151889684, + "grad_norm": 0.7506443858146667, + "learning_rate": 5.016358802900634e-10, + "loss": 0.0524, + "step": 42616 + }, + { + "epoch": 0.9990095934644889, + "grad_norm": 1.7403349876403809, + "learning_rate": 4.785752984570735e-10, + "loss": 0.0888, + "step": 42617 + }, + { + "epoch": 0.999033035039294, + "grad_norm": 0.37379178404808044, + "learning_rate": 4.5605731683329157e-10, + "loss": 0.051, + "step": 42618 + }, + { + "epoch": 0.9990564766140989, + "grad_norm": 0.7679322957992554, + "learning_rate": 4.3408193555194433e-10, + "loss": 0.0442, + "step": 42619 + }, + { + "epoch": 0.999079918188904, + "grad_norm": 1.8875765800476074, + "learning_rate": 4.1264915472405407e-10, + "loss": 0.8769, + "step": 42620 + }, + { + "epoch": 0.9991033597637089, + "grad_norm": 0.8985680937767029, + "learning_rate": 3.917589744717454e-10, + "loss": 0.0914, + "step": 42621 + }, + { + "epoch": 0.999126801338514, + "grad_norm": 2.837190628051758, + "learning_rate": 3.714113949060405e-10, + "loss": 0.197, + "step": 42622 + }, + { + "epoch": 0.9991502429133189, + "grad_norm": 3.6858887672424316, + "learning_rate": 3.516064161268595e-10, + "loss": 0.5731, + "step": 42623 + }, + { + "epoch": 0.9991736844881239, + "grad_norm": 3.8264896869659424, + "learning_rate": 3.3234403825632696e-10, + "loss": 0.2943, + "step": 42624 + }, + { + "epoch": 0.9991971260629289, + "grad_norm": 0.32057303190231323, + "learning_rate": 3.136242614054652e-10, + "loss": 0.007, + "step": 42625 + }, + { + "epoch": 0.9992205676377339, + "grad_norm": 1.8506560325622559, + "learning_rate": 2.9544708565198975e-10, + "loss": 0.7969, + "step": 42626 + }, + { + "epoch": 0.9992440092125389, + "grad_norm": 1.7679033279418945, + "learning_rate": 2.7781251110692297e-10, + "loss": 0.2139, + "step": 42627 + }, + { + "epoch": 0.9992674507873439, + "grad_norm": 1.9560060501098633, + "learning_rate": 2.607205378701849e-10, + "loss": 0.1277, + "step": 42628 + }, + { + "epoch": 0.9992908923621489, + "grad_norm": 0.6423216462135315, + "learning_rate": 2.4417116603059344e-10, + "loss": 0.0343, + "step": 42629 + }, + { + "epoch": 0.9993143339369539, + "grad_norm": 1.5601423978805542, + "learning_rate": 2.2816439567696635e-10, + "loss": 0.0354, + "step": 42630 + }, + { + "epoch": 0.9993377755117588, + "grad_norm": 2.215837001800537, + "learning_rate": 2.1270022689812152e-10, + "loss": 0.2325, + "step": 42631 + }, + { + "epoch": 0.9993612170865639, + "grad_norm": 3.0598695278167725, + "learning_rate": 1.977786597828768e-10, + "loss": 0.1446, + "step": 42632 + }, + { + "epoch": 0.9993846586613688, + "grad_norm": 1.151070475578308, + "learning_rate": 1.8339969439784554e-10, + "loss": 0.0962, + "step": 42633 + }, + { + "epoch": 0.9994081002361739, + "grad_norm": 1.896572470664978, + "learning_rate": 1.695633308318456e-10, + "loss": 0.1017, + "step": 42634 + }, + { + "epoch": 0.9994315418109788, + "grad_norm": 5.52318000793457, + "learning_rate": 1.5626956915149037e-10, + "loss": 0.6673, + "step": 42635 + }, + { + "epoch": 0.9994549833857839, + "grad_norm": 0.863865315914154, + "learning_rate": 1.435184094455977e-10, + "loss": 0.0819, + "step": 42636 + }, + { + "epoch": 0.9994784249605888, + "grad_norm": 0.8905113339424133, + "learning_rate": 1.313098517696787e-10, + "loss": 0.0576, + "step": 42637 + }, + { + "epoch": 0.9995018665353939, + "grad_norm": 1.619079351425171, + "learning_rate": 1.1964389619034678e-10, + "loss": 0.1466, + "step": 42638 + }, + { + "epoch": 0.9995253081101988, + "grad_norm": 3.4945409297943115, + "learning_rate": 1.085205427742153e-10, + "loss": 0.2452, + "step": 42639 + }, + { + "epoch": 0.9995487496850038, + "grad_norm": 3.3741540908813477, + "learning_rate": 9.793979157679545e-11, + "loss": 0.2654, + "step": 42640 + }, + { + "epoch": 0.9995721912598088, + "grad_norm": 2.8491196632385254, + "learning_rate": 8.790164266470058e-11, + "loss": 0.1823, + "step": 42641 + }, + { + "epoch": 0.9995956328346138, + "grad_norm": 2.8972082138061523, + "learning_rate": 7.840609608233963e-11, + "loss": 0.3261, + "step": 42642 + }, + { + "epoch": 0.9996190744094188, + "grad_norm": 3.2255804538726807, + "learning_rate": 6.945315189632595e-11, + "loss": 0.3785, + "step": 42643 + }, + { + "epoch": 0.9996425159842238, + "grad_norm": 4.029861927032471, + "learning_rate": 6.104281013996627e-11, + "loss": 0.2988, + "step": 42644 + }, + { + "epoch": 0.9996659575590288, + "grad_norm": 3.3224668502807617, + "learning_rate": 5.3175070857669483e-11, + "loss": 0.189, + "step": 42645 + }, + { + "epoch": 0.9996893991338338, + "grad_norm": 1.8188893795013428, + "learning_rate": 4.584993410494676e-11, + "loss": 0.169, + "step": 42646 + }, + { + "epoch": 0.9997128407086389, + "grad_norm": 2.3923044204711914, + "learning_rate": 3.906739991510477e-11, + "loss": 1.6304, + "step": 42647 + }, + { + "epoch": 0.9997362822834438, + "grad_norm": 1.9959824085235596, + "learning_rate": 3.282746833255246e-11, + "loss": 0.1076, + "step": 42648 + }, + { + "epoch": 0.9997597238582488, + "grad_norm": 1.413393259048462, + "learning_rate": 2.7130139368392037e-11, + "loss": 0.151, + "step": 42649 + }, + { + "epoch": 0.9997831654330538, + "grad_norm": 2.039055347442627, + "learning_rate": 2.1975413078134666e-11, + "loss": 0.1147, + "step": 42650 + }, + { + "epoch": 0.9998066070078588, + "grad_norm": 2.6629936695098877, + "learning_rate": 1.7363289483984802e-11, + "loss": 0.1691, + "step": 42651 + }, + { + "epoch": 0.9998300485826638, + "grad_norm": 1.8316164016723633, + "learning_rate": 1.3293768597044675e-11, + "loss": 0.073, + "step": 42652 + }, + { + "epoch": 0.9998534901574688, + "grad_norm": 3.086435556411743, + "learning_rate": 9.766850461723209e-12, + "loss": 0.2078, + "step": 42653 + }, + { + "epoch": 0.9998769317322738, + "grad_norm": 3.397658348083496, + "learning_rate": 6.782535078020402e-12, + "loss": 0.1328, + "step": 42654 + }, + { + "epoch": 0.9999003733070788, + "grad_norm": 1.5221973657608032, + "learning_rate": 4.340822468140715e-12, + "loss": 0.1212, + "step": 42655 + }, + { + "epoch": 0.9999238148818838, + "grad_norm": 2.2061495780944824, + "learning_rate": 2.441712643186378e-12, + "loss": 0.1983, + "step": 42656 + }, + { + "epoch": 0.9999472564566888, + "grad_norm": 3.143914222717285, + "learning_rate": 1.0852056253618514e-12, + "loss": 0.4873, + "step": 42657 + }, + { + "epoch": 0.9999706980314937, + "grad_norm": 1.677915096282959, + "learning_rate": 2.713014035649053e-13, + "loss": 0.1931, + "step": 42658 + }, + { + "epoch": 0.9999941396062988, + "grad_norm": 1.7505136728286743, + "learning_rate": 0.0, + "loss": 0.2172, + "step": 42659 + } + ], + "logging_steps": 1, + "max_steps": 42659, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.6963484591125955e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}