| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5001406645619081, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001000281329123816, | |
| "grad_norm": 6.540109634399414, | |
| "learning_rate": 0.0001999999450590425, | |
| "loss": 2.6368, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002000562658247632, | |
| "grad_norm": 3.004662036895752, | |
| "learning_rate": 0.00019999978023623033, | |
| "loss": 2.0338, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.003000843987371448, | |
| "grad_norm": 2.942652463912964, | |
| "learning_rate": 0.0001999995055317446, | |
| "loss": 1.6784, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.004001125316495264, | |
| "grad_norm": 1.9385607242584229, | |
| "learning_rate": 0.00019999912094588717, | |
| "loss": 1.5838, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.005001406645619081, | |
| "grad_norm": 1.7167695760726929, | |
| "learning_rate": 0.00019999862647908064, | |
| "loss": 1.4409, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006001687974742896, | |
| "grad_norm": 1.8187777996063232, | |
| "learning_rate": 0.00019999802213186834, | |
| "loss": 1.4089, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.007001969303866712, | |
| "grad_norm": 1.9561201333999634, | |
| "learning_rate": 0.0001999973079049143, | |
| "loss": 1.4378, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.008002250632990529, | |
| "grad_norm": 1.5050302743911743, | |
| "learning_rate": 0.00019999648379900338, | |
| "loss": 1.1313, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.009002531962114344, | |
| "grad_norm": 1.2175908088684082, | |
| "learning_rate": 0.0001999955498150411, | |
| "loss": 1.161, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.010002813291238161, | |
| "grad_norm": 1.0915567874908447, | |
| "learning_rate": 0.00019999450595405374, | |
| "loss": 1.1367, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.011003094620361977, | |
| "grad_norm": 1.0935776233673096, | |
| "learning_rate": 0.0001999933522171883, | |
| "loss": 0.9341, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.012003375949485792, | |
| "grad_norm": 1.1541457176208496, | |
| "learning_rate": 0.00019999208860571255, | |
| "loss": 1.0971, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01300365727860961, | |
| "grad_norm": 1.0622767210006714, | |
| "learning_rate": 0.00019999071512101496, | |
| "loss": 1.0928, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.014003938607733425, | |
| "grad_norm": 1.0497690439224243, | |
| "learning_rate": 0.00019998923176460474, | |
| "loss": 1.0885, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.01500421993685724, | |
| "grad_norm": 0.9466670155525208, | |
| "learning_rate": 0.00019998763853811184, | |
| "loss": 0.9602, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.016004501265981057, | |
| "grad_norm": 1.1238515377044678, | |
| "learning_rate": 0.00019998593544328692, | |
| "loss": 0.9708, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.017004782595104875, | |
| "grad_norm": 1.2019788026809692, | |
| "learning_rate": 0.00019998412248200138, | |
| "loss": 0.9854, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.01800506392422869, | |
| "grad_norm": 0.8101840019226074, | |
| "learning_rate": 0.00019998219965624734, | |
| "loss": 0.9747, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.019005345253352506, | |
| "grad_norm": 0.7606379985809326, | |
| "learning_rate": 0.0001999801669681376, | |
| "loss": 0.9558, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.020005626582476323, | |
| "grad_norm": 0.7452303767204285, | |
| "learning_rate": 0.00019997802441990573, | |
| "loss": 0.8407, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.021005907911600136, | |
| "grad_norm": 0.810576856136322, | |
| "learning_rate": 0.00019997577201390606, | |
| "loss": 0.8321, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.022006189240723954, | |
| "grad_norm": 0.8486515879631042, | |
| "learning_rate": 0.00019997340975261353, | |
| "loss": 0.8239, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.02300647056984777, | |
| "grad_norm": 1.194822907447815, | |
| "learning_rate": 0.00019997093763862383, | |
| "loss": 0.85, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.024006751898971584, | |
| "grad_norm": 0.8571400046348572, | |
| "learning_rate": 0.0001999683556746534, | |
| "loss": 0.8917, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0250070332280954, | |
| "grad_norm": 0.7690589427947998, | |
| "learning_rate": 0.0001999656638635393, | |
| "loss": 0.9666, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02600731455721922, | |
| "grad_norm": 0.9334922432899475, | |
| "learning_rate": 0.0001999628622082394, | |
| "loss": 0.8808, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.027007595886343033, | |
| "grad_norm": 0.9193708896636963, | |
| "learning_rate": 0.0001999599507118322, | |
| "loss": 0.9316, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02800787721546685, | |
| "grad_norm": 0.9622653126716614, | |
| "learning_rate": 0.00019995692937751683, | |
| "loss": 0.9242, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.029008158544590667, | |
| "grad_norm": 0.76296067237854, | |
| "learning_rate": 0.0001999537982086133, | |
| "loss": 0.7989, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.03000843987371448, | |
| "grad_norm": 0.84053635597229, | |
| "learning_rate": 0.00019995055720856218, | |
| "loss": 0.8608, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.031008721202838298, | |
| "grad_norm": 0.7417949438095093, | |
| "learning_rate": 0.00019994720638092468, | |
| "loss": 0.822, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.032009002531962115, | |
| "grad_norm": 0.8762928247451782, | |
| "learning_rate": 0.00019994374572938277, | |
| "loss": 0.9001, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03300928386108593, | |
| "grad_norm": 0.6854381561279297, | |
| "learning_rate": 0.00019994017525773913, | |
| "loss": 0.7946, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03400956519020975, | |
| "grad_norm": 0.6967355012893677, | |
| "learning_rate": 0.00019993649496991705, | |
| "loss": 0.8277, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.03500984651933356, | |
| "grad_norm": 1.393047571182251, | |
| "learning_rate": 0.00019993270486996046, | |
| "loss": 0.7741, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03601012784845738, | |
| "grad_norm": 0.676814615726471, | |
| "learning_rate": 0.000199928804962034, | |
| "loss": 0.7317, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.037010409177581194, | |
| "grad_norm": 1.0221902132034302, | |
| "learning_rate": 0.00019992479525042303, | |
| "loss": 0.8664, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03801069050670501, | |
| "grad_norm": 0.5960223078727722, | |
| "learning_rate": 0.00019992067573953342, | |
| "loss": 0.8324, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03901097183582883, | |
| "grad_norm": 0.6729836463928223, | |
| "learning_rate": 0.0001999164464338918, | |
| "loss": 0.7634, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.040011253164952645, | |
| "grad_norm": 0.802059531211853, | |
| "learning_rate": 0.0001999121073381454, | |
| "loss": 0.8515, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.041011534494076456, | |
| "grad_norm": 1.074352741241455, | |
| "learning_rate": 0.0001999076584570621, | |
| "loss": 0.811, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04201181582320027, | |
| "grad_norm": 1.1671184301376343, | |
| "learning_rate": 0.00019990309979553045, | |
| "loss": 0.8336, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.04301209715232409, | |
| "grad_norm": 1.1403908729553223, | |
| "learning_rate": 0.00019989843135855958, | |
| "loss": 0.8674, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04401237848144791, | |
| "grad_norm": 0.9987945556640625, | |
| "learning_rate": 0.00019989365315127922, | |
| "loss": 0.8558, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.045012659810571724, | |
| "grad_norm": 0.827928900718689, | |
| "learning_rate": 0.0001998887651789398, | |
| "loss": 0.7749, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04601294113969554, | |
| "grad_norm": 0.694059431552887, | |
| "learning_rate": 0.0001998837674469123, | |
| "loss": 0.8992, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.04701322246881936, | |
| "grad_norm": 0.5862646102905273, | |
| "learning_rate": 0.00019987865996068833, | |
| "loss": 0.7361, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.04801350379794317, | |
| "grad_norm": 0.5515244007110596, | |
| "learning_rate": 0.00019987344272588006, | |
| "loss": 0.7065, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.049013785127066986, | |
| "grad_norm": 0.5773719549179077, | |
| "learning_rate": 0.00019986811574822033, | |
| "loss": 0.7041, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0500140664561908, | |
| "grad_norm": 0.4614185094833374, | |
| "learning_rate": 0.00019986267903356254, | |
| "loss": 0.6977, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05101434778531462, | |
| "grad_norm": 0.5293365716934204, | |
| "learning_rate": 0.0001998571325878806, | |
| "loss": 0.7435, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.05201462911443844, | |
| "grad_norm": 0.4691166281700134, | |
| "learning_rate": 0.0001998514764172691, | |
| "loss": 0.684, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.053014910443562255, | |
| "grad_norm": 0.6643468141555786, | |
| "learning_rate": 0.00019984571052794313, | |
| "loss": 0.7708, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.054015191772686065, | |
| "grad_norm": 0.6328372955322266, | |
| "learning_rate": 0.00019983983492623833, | |
| "loss": 0.8069, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.05501547310180988, | |
| "grad_norm": 0.7191805243492126, | |
| "learning_rate": 0.00019983384961861096, | |
| "loss": 0.757, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0560157544309337, | |
| "grad_norm": 0.6610512137413025, | |
| "learning_rate": 0.0001998277546116378, | |
| "loss": 0.6549, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.05701603576005752, | |
| "grad_norm": 0.6019318699836731, | |
| "learning_rate": 0.00019982154991201608, | |
| "loss": 0.7135, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.058016317089181334, | |
| "grad_norm": 0.6172312498092651, | |
| "learning_rate": 0.00019981523552656377, | |
| "loss": 0.7454, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.05901659841830515, | |
| "grad_norm": 0.6532041430473328, | |
| "learning_rate": 0.00019980881146221914, | |
| "loss": 0.6976, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.06001687974742896, | |
| "grad_norm": 0.7116113305091858, | |
| "learning_rate": 0.00019980227772604112, | |
| "loss": 0.7581, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06101716107655278, | |
| "grad_norm": 0.6532421708106995, | |
| "learning_rate": 0.0001997956343252091, | |
| "loss": 0.7042, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.062017442405676595, | |
| "grad_norm": 0.6296349167823792, | |
| "learning_rate": 0.00019978888126702296, | |
| "loss": 0.6981, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.06301772373480041, | |
| "grad_norm": 0.6616079211235046, | |
| "learning_rate": 0.00019978201855890308, | |
| "loss": 0.6885, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.06401800506392423, | |
| "grad_norm": 0.6802906394004822, | |
| "learning_rate": 0.00019977504620839035, | |
| "loss": 0.6685, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.06501828639304805, | |
| "grad_norm": 0.5653461217880249, | |
| "learning_rate": 0.00019976796422314615, | |
| "loss": 0.7135, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06601856772217186, | |
| "grad_norm": 0.5990989804267883, | |
| "learning_rate": 0.00019976077261095226, | |
| "loss": 0.8312, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.06701884905129568, | |
| "grad_norm": 0.6948749423027039, | |
| "learning_rate": 0.00019975347137971098, | |
| "loss": 0.742, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.0680191303804195, | |
| "grad_norm": 0.6895416378974915, | |
| "learning_rate": 0.00019974606053744503, | |
| "loss": 0.6902, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.06901941170954332, | |
| "grad_norm": 0.6278745532035828, | |
| "learning_rate": 0.00019973854009229763, | |
| "loss": 0.65, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.07001969303866712, | |
| "grad_norm": 0.6024957299232483, | |
| "learning_rate": 0.00019973091005253232, | |
| "loss": 0.6611, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07101997436779094, | |
| "grad_norm": 1.4569993019104004, | |
| "learning_rate": 0.0001997231704265332, | |
| "loss": 0.6809, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.07202025569691475, | |
| "grad_norm": 0.42510971426963806, | |
| "learning_rate": 0.00019971532122280464, | |
| "loss": 0.6415, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.07302053702603857, | |
| "grad_norm": 0.5024048089981079, | |
| "learning_rate": 0.0001997073624499716, | |
| "loss": 0.6515, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.07402081835516239, | |
| "grad_norm": 0.4638510048389435, | |
| "learning_rate": 0.0001996992941167792, | |
| "loss": 0.726, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.0750210996842862, | |
| "grad_norm": 0.5222762823104858, | |
| "learning_rate": 0.00019969111623209323, | |
| "loss": 0.5953, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07602138101341002, | |
| "grad_norm": 0.5868512392044067, | |
| "learning_rate": 0.00019968282880489957, | |
| "loss": 0.7317, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.07702166234253384, | |
| "grad_norm": 0.6825968027114868, | |
| "learning_rate": 0.00019967443184430467, | |
| "loss": 0.6028, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.07802194367165766, | |
| "grad_norm": 0.7539929747581482, | |
| "learning_rate": 0.0001996659253595353, | |
| "loss": 0.747, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.07902222500078147, | |
| "grad_norm": 0.7608583569526672, | |
| "learning_rate": 0.0001996573093599385, | |
| "loss": 0.6783, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.08002250632990529, | |
| "grad_norm": 0.7046730518341064, | |
| "learning_rate": 0.00019964858385498172, | |
| "loss": 0.6443, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08102278765902911, | |
| "grad_norm": 0.7782636284828186, | |
| "learning_rate": 0.00019963974885425266, | |
| "loss": 0.6666, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.08202306898815291, | |
| "grad_norm": 0.7668784856796265, | |
| "learning_rate": 0.00019963080436745945, | |
| "loss": 0.6789, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.08302335031727673, | |
| "grad_norm": 0.759543240070343, | |
| "learning_rate": 0.00019962175040443044, | |
| "loss": 0.7122, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.08402363164640055, | |
| "grad_norm": 0.4773256778717041, | |
| "learning_rate": 0.0001996125869751143, | |
| "loss": 0.61, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.08502391297552436, | |
| "grad_norm": 0.4939149022102356, | |
| "learning_rate": 0.00019960331408957997, | |
| "loss": 0.6302, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08602419430464818, | |
| "grad_norm": 0.5174906849861145, | |
| "learning_rate": 0.00019959393175801671, | |
| "loss": 0.682, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.087024475633772, | |
| "grad_norm": 0.4596863090991974, | |
| "learning_rate": 0.00019958443999073397, | |
| "loss": 0.6479, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.08802475696289581, | |
| "grad_norm": 0.5851575136184692, | |
| "learning_rate": 0.00019957483879816151, | |
| "loss": 0.6814, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.08902503829201963, | |
| "grad_norm": 0.5562273859977722, | |
| "learning_rate": 0.00019956512819084928, | |
| "loss": 0.6806, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.09002531962114345, | |
| "grad_norm": 0.5329018235206604, | |
| "learning_rate": 0.00019955530817946748, | |
| "loss": 0.6147, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09102560095026727, | |
| "grad_norm": 0.5893882513046265, | |
| "learning_rate": 0.00019954537877480655, | |
| "loss": 0.5858, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.09202588227939108, | |
| "grad_norm": 0.5248862504959106, | |
| "learning_rate": 0.00019953533998777706, | |
| "loss": 0.602, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0930261636085149, | |
| "grad_norm": 0.5627758502960205, | |
| "learning_rate": 0.00019952519182940993, | |
| "loss": 0.6584, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.09402644493763872, | |
| "grad_norm": 0.6081700325012207, | |
| "learning_rate": 0.00019951493431085603, | |
| "loss": 0.6227, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.09502672626676252, | |
| "grad_norm": 0.6908998489379883, | |
| "learning_rate": 0.00019950456744338658, | |
| "loss": 0.6615, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.09602700759588634, | |
| "grad_norm": 0.49264904856681824, | |
| "learning_rate": 0.00019949409123839288, | |
| "loss": 0.6784, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.09702728892501016, | |
| "grad_norm": 0.6333140730857849, | |
| "learning_rate": 0.00019948350570738642, | |
| "loss": 0.661, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.09802757025413397, | |
| "grad_norm": 0.6504009962081909, | |
| "learning_rate": 0.0001994728108619987, | |
| "loss": 0.649, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.09902785158325779, | |
| "grad_norm": 0.6558539867401123, | |
| "learning_rate": 0.0001994620067139815, | |
| "loss": 0.5927, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.1000281329123816, | |
| "grad_norm": 0.750813901424408, | |
| "learning_rate": 0.00019945109327520658, | |
| "loss": 0.6479, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10102841424150542, | |
| "grad_norm": 0.6053513884544373, | |
| "learning_rate": 0.00019944007055766586, | |
| "loss": 0.6851, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.10202869557062924, | |
| "grad_norm": 0.5293607711791992, | |
| "learning_rate": 0.00019942893857347128, | |
| "loss": 0.6544, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.10302897689975306, | |
| "grad_norm": 0.5119344592094421, | |
| "learning_rate": 0.00019941769733485494, | |
| "loss": 0.6704, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.10402925822887688, | |
| "grad_norm": 0.49486231803894043, | |
| "learning_rate": 0.00019940634685416888, | |
| "loss": 0.6309, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.10502953955800069, | |
| "grad_norm": 0.408372700214386, | |
| "learning_rate": 0.00019939488714388524, | |
| "loss": 0.5601, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.10602982088712451, | |
| "grad_norm": 0.45692771673202515, | |
| "learning_rate": 0.00019938331821659614, | |
| "loss": 0.583, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.10703010221624833, | |
| "grad_norm": 0.5024964809417725, | |
| "learning_rate": 0.0001993716400850138, | |
| "loss": 0.5944, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.10803038354537213, | |
| "grad_norm": 0.621824324131012, | |
| "learning_rate": 0.0001993598527619703, | |
| "loss": 0.6284, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.10903066487449595, | |
| "grad_norm": 0.5267654657363892, | |
| "learning_rate": 0.00019934795626041783, | |
| "loss": 0.6178, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.11003094620361976, | |
| "grad_norm": 0.497410386800766, | |
| "learning_rate": 0.0001993359505934285, | |
| "loss": 0.557, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11103122753274358, | |
| "grad_norm": 0.5206043720245361, | |
| "learning_rate": 0.00019932383577419432, | |
| "loss": 0.577, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.1120315088618674, | |
| "grad_norm": 1.2140614986419678, | |
| "learning_rate": 0.0001993116118160273, | |
| "loss": 0.6339, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.11303179019099122, | |
| "grad_norm": 0.7585737109184265, | |
| "learning_rate": 0.00019929927873235938, | |
| "loss": 0.6587, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.11403207152011503, | |
| "grad_norm": 0.5759165287017822, | |
| "learning_rate": 0.00019928683653674237, | |
| "loss": 0.6275, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.11503235284923885, | |
| "grad_norm": 0.5694018006324768, | |
| "learning_rate": 0.00019927428524284805, | |
| "loss": 0.6535, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.11603263417836267, | |
| "grad_norm": 0.4953472912311554, | |
| "learning_rate": 0.00019926162486446792, | |
| "loss": 0.6077, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.11703291550748648, | |
| "grad_norm": 0.6509867310523987, | |
| "learning_rate": 0.0001992488554155135, | |
| "loss": 0.6451, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.1180331968366103, | |
| "grad_norm": 0.6137562990188599, | |
| "learning_rate": 0.00019923597691001615, | |
| "loss": 0.6224, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.11903347816573412, | |
| "grad_norm": 0.595355212688446, | |
| "learning_rate": 0.0001992229893621269, | |
| "loss": 0.6059, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.12003375949485792, | |
| "grad_norm": 0.5127987265586853, | |
| "learning_rate": 0.00019920989278611687, | |
| "loss": 0.6359, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12103404082398174, | |
| "grad_norm": 0.5433658361434937, | |
| "learning_rate": 0.0001991966871963767, | |
| "loss": 0.6817, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.12203432215310556, | |
| "grad_norm": 0.634130597114563, | |
| "learning_rate": 0.000199183372607417, | |
| "loss": 0.6174, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.12303460348222937, | |
| "grad_norm": 0.5337157845497131, | |
| "learning_rate": 0.0001991699490338681, | |
| "loss": 0.671, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.12403488481135319, | |
| "grad_norm": 0.5023785829544067, | |
| "learning_rate": 0.00019915641649048005, | |
| "loss": 0.5828, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.12503516614047702, | |
| "grad_norm": 0.711187481880188, | |
| "learning_rate": 0.0001991427749921227, | |
| "loss": 0.5384, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.12603544746960083, | |
| "grad_norm": 0.7347482442855835, | |
| "learning_rate": 0.00019912902455378556, | |
| "loss": 0.6257, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.12703572879872463, | |
| "grad_norm": 0.583920955657959, | |
| "learning_rate": 0.00019911516519057788, | |
| "loss": 0.5547, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.12803601012784846, | |
| "grad_norm": 1.15117609500885, | |
| "learning_rate": 0.00019910119691772863, | |
| "loss": 0.6215, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.12903629145697226, | |
| "grad_norm": 0.5940455794334412, | |
| "learning_rate": 0.00019908711975058637, | |
| "loss": 0.6136, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.1300365727860961, | |
| "grad_norm": 0.5960786938667297, | |
| "learning_rate": 0.0001990729337046194, | |
| "loss": 0.5271, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1310368541152199, | |
| "grad_norm": 0.6204031705856323, | |
| "learning_rate": 0.0001990586387954156, | |
| "loss": 0.5584, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.13203713544434373, | |
| "grad_norm": 0.6092723608016968, | |
| "learning_rate": 0.00019904423503868247, | |
| "loss": 0.6402, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.13303741677346753, | |
| "grad_norm": 0.6005457639694214, | |
| "learning_rate": 0.00019902972245024715, | |
| "loss": 0.6422, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.13403769810259136, | |
| "grad_norm": 0.6619305610656738, | |
| "learning_rate": 0.00019901510104605637, | |
| "loss": 0.5395, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.13503797943171517, | |
| "grad_norm": 0.7267995476722717, | |
| "learning_rate": 0.00019900037084217637, | |
| "loss": 0.6358, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.136038260760839, | |
| "grad_norm": 0.3749965727329254, | |
| "learning_rate": 0.00019898553185479303, | |
| "loss": 0.5185, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.1370385420899628, | |
| "grad_norm": 0.6056541204452515, | |
| "learning_rate": 0.00019897058410021167, | |
| "loss": 0.5466, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.13803882341908663, | |
| "grad_norm": 0.4612765908241272, | |
| "learning_rate": 0.00019895552759485722, | |
| "loss": 0.6349, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.13903910474821043, | |
| "grad_norm": 0.47530749440193176, | |
| "learning_rate": 0.00019894036235527395, | |
| "loss": 0.5828, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.14003938607733424, | |
| "grad_norm": 0.5521253347396851, | |
| "learning_rate": 0.00019892508839812584, | |
| "loss": 0.6172, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14103966740645807, | |
| "grad_norm": 0.4842628538608551, | |
| "learning_rate": 0.00019890970574019617, | |
| "loss": 0.5327, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.14203994873558187, | |
| "grad_norm": 0.5947982668876648, | |
| "learning_rate": 0.00019889421439838763, | |
| "loss": 0.5432, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1430402300647057, | |
| "grad_norm": 0.6485494375228882, | |
| "learning_rate": 0.00019887861438972246, | |
| "loss": 0.5293, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.1440405113938295, | |
| "grad_norm": 0.4645004868507385, | |
| "learning_rate": 0.00019886290573134228, | |
| "loss": 0.6078, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.14504079272295334, | |
| "grad_norm": 0.49198296666145325, | |
| "learning_rate": 0.000198847088440508, | |
| "loss": 0.5283, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.14604107405207714, | |
| "grad_norm": 0.6396425366401672, | |
| "learning_rate": 0.0001988311625346, | |
| "loss": 0.6178, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.14704135538120097, | |
| "grad_norm": 0.49430540204048157, | |
| "learning_rate": 0.00019881512803111796, | |
| "loss": 0.5914, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.14804163671032478, | |
| "grad_norm": 0.5589768886566162, | |
| "learning_rate": 0.00019879898494768093, | |
| "loss": 0.633, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.1490419180394486, | |
| "grad_norm": 0.629703938961029, | |
| "learning_rate": 0.00019878273330202717, | |
| "loss": 0.5706, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.1500421993685724, | |
| "grad_norm": 0.4935409724712372, | |
| "learning_rate": 0.00019876637311201433, | |
| "loss": 0.5004, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15104248069769624, | |
| "grad_norm": 0.44915804266929626, | |
| "learning_rate": 0.00019874990439561934, | |
| "loss": 0.5326, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.15204276202682004, | |
| "grad_norm": 0.5451924204826355, | |
| "learning_rate": 0.0001987333271709383, | |
| "loss": 0.6474, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.15304304335594385, | |
| "grad_norm": 0.641677737236023, | |
| "learning_rate": 0.00019871664145618657, | |
| "loss": 0.5785, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.15404332468506768, | |
| "grad_norm": 0.40758514404296875, | |
| "learning_rate": 0.00019869984726969878, | |
| "loss": 0.5416, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.15504360601419148, | |
| "grad_norm": 0.8472537994384766, | |
| "learning_rate": 0.00019868294462992866, | |
| "loss": 0.5764, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.1560438873433153, | |
| "grad_norm": 0.49797093868255615, | |
| "learning_rate": 0.00019866593355544922, | |
| "loss": 0.5597, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.15704416867243912, | |
| "grad_norm": 0.5156570076942444, | |
| "learning_rate": 0.00019864881406495246, | |
| "loss": 0.6066, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.15804445000156295, | |
| "grad_norm": 0.5762325525283813, | |
| "learning_rate": 0.00019863158617724967, | |
| "loss": 0.5678, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.15904473133068675, | |
| "grad_norm": 0.4978669583797455, | |
| "learning_rate": 0.00019861424991127115, | |
| "loss": 0.5523, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.16004501265981058, | |
| "grad_norm": 1.5707205533981323, | |
| "learning_rate": 0.00019859680528606637, | |
| "loss": 0.6809, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16104529398893438, | |
| "grad_norm": 0.42276397347450256, | |
| "learning_rate": 0.00019857925232080373, | |
| "loss": 0.5389, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.16204557531805822, | |
| "grad_norm": 0.5103838443756104, | |
| "learning_rate": 0.00019856159103477086, | |
| "loss": 0.5572, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.16304585664718202, | |
| "grad_norm": 0.49496352672576904, | |
| "learning_rate": 0.00019854382144737426, | |
| "loss": 0.5608, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.16404613797630582, | |
| "grad_norm": 0.48654571175575256, | |
| "learning_rate": 0.00019852594357813952, | |
| "loss": 0.5678, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.16504641930542965, | |
| "grad_norm": 0.6116751432418823, | |
| "learning_rate": 0.00019850795744671116, | |
| "loss": 0.6197, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.16604670063455346, | |
| "grad_norm": 0.4954712688922882, | |
| "learning_rate": 0.0001984898630728527, | |
| "loss": 0.5498, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.1670469819636773, | |
| "grad_norm": 0.5408039689064026, | |
| "learning_rate": 0.0001984716604764466, | |
| "loss": 0.5202, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.1680472632928011, | |
| "grad_norm": 0.6083703637123108, | |
| "learning_rate": 0.0001984533496774942, | |
| "loss": 0.5584, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.16904754462192492, | |
| "grad_norm": 0.47014451026916504, | |
| "learning_rate": 0.0001984349306961158, | |
| "loss": 0.5508, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.17004782595104873, | |
| "grad_norm": 0.5064229369163513, | |
| "learning_rate": 0.00019841640355255043, | |
| "loss": 0.6152, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.17104810728017256, | |
| "grad_norm": 0.6776212453842163, | |
| "learning_rate": 0.00019839776826715614, | |
| "loss": 0.5392, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.17204838860929636, | |
| "grad_norm": 0.46285948157310486, | |
| "learning_rate": 0.00019837902486040978, | |
| "loss": 0.6151, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.1730486699384202, | |
| "grad_norm": 0.48528385162353516, | |
| "learning_rate": 0.0001983601733529069, | |
| "loss": 0.5609, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.174048951267544, | |
| "grad_norm": 0.5141404867172241, | |
| "learning_rate": 0.00019834121376536187, | |
| "loss": 0.6115, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.17504923259666783, | |
| "grad_norm": 0.42379871010780334, | |
| "learning_rate": 0.00019832214611860793, | |
| "loss": 0.6509, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.17604951392579163, | |
| "grad_norm": 0.4130716323852539, | |
| "learning_rate": 0.00019830297043359692, | |
| "loss": 0.5719, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.17704979525491543, | |
| "grad_norm": 0.41237375140190125, | |
| "learning_rate": 0.00019828368673139947, | |
| "loss": 0.5288, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.17805007658403926, | |
| "grad_norm": 0.5197924971580505, | |
| "learning_rate": 0.0001982642950332049, | |
| "loss": 0.5941, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.17905035791316307, | |
| "grad_norm": 0.4431571066379547, | |
| "learning_rate": 0.00019824479536032112, | |
| "loss": 0.5075, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.1800506392422869, | |
| "grad_norm": 0.5234508514404297, | |
| "learning_rate": 0.0001982251877341748, | |
| "loss": 0.5239, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1810509205714107, | |
| "grad_norm": 0.5809316635131836, | |
| "learning_rate": 0.00019820547217631117, | |
| "loss": 0.5646, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.18205120190053453, | |
| "grad_norm": 0.5900911092758179, | |
| "learning_rate": 0.00019818564870839405, | |
| "loss": 0.5413, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.18305148322965833, | |
| "grad_norm": 0.6365839242935181, | |
| "learning_rate": 0.00019816571735220583, | |
| "loss": 0.5463, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.18405176455878217, | |
| "grad_norm": 0.7070748805999756, | |
| "learning_rate": 0.00019814567812964748, | |
| "loss": 0.5335, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.18505204588790597, | |
| "grad_norm": 0.5305525064468384, | |
| "learning_rate": 0.00019812553106273847, | |
| "loss": 0.529, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1860523272170298, | |
| "grad_norm": 0.6620722413063049, | |
| "learning_rate": 0.00019810527617361681, | |
| "loss": 0.498, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1870526085461536, | |
| "grad_norm": 0.5616530776023865, | |
| "learning_rate": 0.00019808491348453894, | |
| "loss": 0.5551, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.18805288987527743, | |
| "grad_norm": 0.6696614623069763, | |
| "learning_rate": 0.00019806444301787978, | |
| "loss": 0.5957, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.18905317120440124, | |
| "grad_norm": 0.7320078015327454, | |
| "learning_rate": 0.0001980438647961327, | |
| "loss": 0.6154, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.19005345253352504, | |
| "grad_norm": 0.630651593208313, | |
| "learning_rate": 0.00019802317884190935, | |
| "loss": 0.5962, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.19105373386264887, | |
| "grad_norm": 0.6430137753486633, | |
| "learning_rate": 0.00019800238517793996, | |
| "loss": 0.5237, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.19205401519177268, | |
| "grad_norm": 0.5368555188179016, | |
| "learning_rate": 0.00019798148382707296, | |
| "loss": 0.5346, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.1930542965208965, | |
| "grad_norm": 0.45558878779411316, | |
| "learning_rate": 0.00019796047481227515, | |
| "loss": 0.567, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.1940545778500203, | |
| "grad_norm": 0.49816155433654785, | |
| "learning_rate": 0.00019793935815663163, | |
| "loss": 0.5741, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.19505485917914414, | |
| "grad_norm": 0.5015010833740234, | |
| "learning_rate": 0.00019791813388334581, | |
| "loss": 0.5198, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.19605514050826794, | |
| "grad_norm": 0.41044294834136963, | |
| "learning_rate": 0.00019789680201573933, | |
| "loss": 0.5111, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.19705542183739178, | |
| "grad_norm": 0.9961733222007751, | |
| "learning_rate": 0.00019787536257725202, | |
| "loss": 0.5427, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.19805570316651558, | |
| "grad_norm": 0.42929476499557495, | |
| "learning_rate": 0.00019785381559144196, | |
| "loss": 0.494, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.1990559844956394, | |
| "grad_norm": 0.4490405321121216, | |
| "learning_rate": 0.00019783216108198542, | |
| "loss": 0.5225, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.2000562658247632, | |
| "grad_norm": 0.513647198677063, | |
| "learning_rate": 0.00019781039907267677, | |
| "loss": 0.557, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.20105654715388704, | |
| "grad_norm": 0.4340246915817261, | |
| "learning_rate": 0.00019778852958742853, | |
| "loss": 0.572, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.20205682848301085, | |
| "grad_norm": 0.4050307273864746, | |
| "learning_rate": 0.00019776655265027127, | |
| "loss": 0.5278, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.20305710981213465, | |
| "grad_norm": 0.4073551595211029, | |
| "learning_rate": 0.00019774446828535371, | |
| "loss": 0.4888, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.20405739114125848, | |
| "grad_norm": 0.5057980418205261, | |
| "learning_rate": 0.00019772227651694256, | |
| "loss": 0.5399, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.20505767247038229, | |
| "grad_norm": 0.4432837963104248, | |
| "learning_rate": 0.00019769997736942258, | |
| "loss": 0.5073, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.20605795379950612, | |
| "grad_norm": 0.47015416622161865, | |
| "learning_rate": 0.00019767757086729647, | |
| "loss": 0.4948, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.20705823512862992, | |
| "grad_norm": 0.5023201107978821, | |
| "learning_rate": 0.00019765505703518496, | |
| "loss": 0.5286, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.20805851645775375, | |
| "grad_norm": 0.5784673690795898, | |
| "learning_rate": 0.00019763243589782662, | |
| "loss": 0.5247, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.20905879778687755, | |
| "grad_norm": 0.46152037382125854, | |
| "learning_rate": 0.00019760970748007803, | |
| "loss": 0.5048, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.21005907911600138, | |
| "grad_norm": 0.6008761525154114, | |
| "learning_rate": 0.0001975868718069136, | |
| "loss": 0.5575, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2110593604451252, | |
| "grad_norm": 0.4951910972595215, | |
| "learning_rate": 0.00019756392890342563, | |
| "loss": 0.4928, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.21205964177424902, | |
| "grad_norm": 0.5855419635772705, | |
| "learning_rate": 0.00019754087879482422, | |
| "loss": 0.542, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.21305992310337282, | |
| "grad_norm": 0.5546546578407288, | |
| "learning_rate": 0.00019751772150643722, | |
| "loss": 0.5762, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.21406020443249665, | |
| "grad_norm": 0.5548007488250732, | |
| "learning_rate": 0.00019749445706371038, | |
| "loss": 0.522, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.21506048576162046, | |
| "grad_norm": 0.5254936814308167, | |
| "learning_rate": 0.00019747108549220702, | |
| "loss": 0.5864, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.21606076709074426, | |
| "grad_norm": 0.5107640027999878, | |
| "learning_rate": 0.00019744760681760832, | |
| "loss": 0.5451, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.2170610484198681, | |
| "grad_norm": 0.5446146726608276, | |
| "learning_rate": 0.00019742402106571314, | |
| "loss": 0.5352, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2180613297489919, | |
| "grad_norm": 0.414689302444458, | |
| "learning_rate": 0.00019740032826243788, | |
| "loss": 0.5236, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.21906161107811573, | |
| "grad_norm": 0.40488916635513306, | |
| "learning_rate": 0.0001973765284338167, | |
| "loss": 0.4937, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.22006189240723953, | |
| "grad_norm": 0.42548874020576477, | |
| "learning_rate": 0.00019735262160600127, | |
| "loss": 0.5433, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22106217373636336, | |
| "grad_norm": 0.5061275362968445, | |
| "learning_rate": 0.00019732860780526088, | |
| "loss": 0.5747, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.22206245506548716, | |
| "grad_norm": 0.4437318742275238, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 0.5681, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.223062736394611, | |
| "grad_norm": 0.40250784158706665, | |
| "learning_rate": 0.00019728025939067008, | |
| "loss": 0.4735, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.2240630177237348, | |
| "grad_norm": 0.43195122480392456, | |
| "learning_rate": 0.00019725592482994583, | |
| "loss": 0.5166, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.22506329905285863, | |
| "grad_norm": 0.3928036689758301, | |
| "learning_rate": 0.00019723148340254892, | |
| "loss": 0.4718, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.22606358038198243, | |
| "grad_norm": 0.3949666917324066, | |
| "learning_rate": 0.00019720693513533598, | |
| "loss": 0.4851, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.22706386171110624, | |
| "grad_norm": 0.4328802227973938, | |
| "learning_rate": 0.00019718228005528122, | |
| "loss": 0.4482, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.22806414304023007, | |
| "grad_norm": 0.5324240326881409, | |
| "learning_rate": 0.00019715751818947603, | |
| "loss": 0.5488, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.22906442436935387, | |
| "grad_norm": 0.446676641702652, | |
| "learning_rate": 0.0001971326495651293, | |
| "loss": 0.5178, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.2300647056984777, | |
| "grad_norm": 0.5981528162956238, | |
| "learning_rate": 0.00019710767420956705, | |
| "loss": 0.4943, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2310649870276015, | |
| "grad_norm": 0.4626360237598419, | |
| "learning_rate": 0.0001970825921502328, | |
| "loss": 0.5424, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.23206526835672533, | |
| "grad_norm": 0.4125116467475891, | |
| "learning_rate": 0.0001970574034146871, | |
| "loss": 0.5126, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.23306554968584914, | |
| "grad_norm": 0.45425575971603394, | |
| "learning_rate": 0.00019703210803060782, | |
| "loss": 0.5083, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.23406583101497297, | |
| "grad_norm": 0.449907511472702, | |
| "learning_rate": 0.00019700670602579008, | |
| "loss": 0.5083, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.23506611234409677, | |
| "grad_norm": 0.42983028292655945, | |
| "learning_rate": 0.00019698119742814606, | |
| "loss": 0.5683, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2360663936732206, | |
| "grad_norm": 0.5434374809265137, | |
| "learning_rate": 0.00019695558226570507, | |
| "loss": 0.5554, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.2370666750023444, | |
| "grad_norm": 1.194429874420166, | |
| "learning_rate": 0.00019692986056661356, | |
| "loss": 0.5923, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.23806695633146824, | |
| "grad_norm": 0.40971899032592773, | |
| "learning_rate": 0.00019690403235913504, | |
| "loss": 0.5173, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.23906723766059204, | |
| "grad_norm": 0.40528661012649536, | |
| "learning_rate": 0.00019687809767165, | |
| "loss": 0.5331, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.24006751898971584, | |
| "grad_norm": 0.40911996364593506, | |
| "learning_rate": 0.000196852056532656, | |
| "loss": 0.4904, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.24106780031883968, | |
| "grad_norm": 0.5151987075805664, | |
| "learning_rate": 0.00019682590897076752, | |
| "loss": 0.5382, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.24206808164796348, | |
| "grad_norm": 0.3909003734588623, | |
| "learning_rate": 0.00019679965501471608, | |
| "loss": 0.447, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.2430683629770873, | |
| "grad_norm": 0.497062623500824, | |
| "learning_rate": 0.0001967732946933499, | |
| "loss": 0.596, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2440686443062111, | |
| "grad_norm": 0.4184584617614746, | |
| "learning_rate": 0.00019674682803563428, | |
| "loss": 0.5146, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.24506892563533494, | |
| "grad_norm": 0.4521709680557251, | |
| "learning_rate": 0.00019672025507065131, | |
| "loss": 0.5296, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.24606920696445875, | |
| "grad_norm": 0.44223159551620483, | |
| "learning_rate": 0.00019669357582759983, | |
| "loss": 0.5223, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.24706948829358258, | |
| "grad_norm": 0.5312020778656006, | |
| "learning_rate": 0.00019666679033579552, | |
| "loss": 0.5491, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.24806976962270638, | |
| "grad_norm": 0.4204310178756714, | |
| "learning_rate": 0.00019663989862467082, | |
| "loss": 0.4885, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.2490700509518302, | |
| "grad_norm": 0.4755087196826935, | |
| "learning_rate": 0.00019661290072377482, | |
| "loss": 0.5295, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.25007033228095404, | |
| "grad_norm": 0.8514156937599182, | |
| "learning_rate": 0.00019658579666277334, | |
| "loss": 0.5279, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2510706136100778, | |
| "grad_norm": 0.6153047680854797, | |
| "learning_rate": 0.0001965585864714488, | |
| "loss": 0.4662, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.25207089493920165, | |
| "grad_norm": 0.49649399518966675, | |
| "learning_rate": 0.00019653127017970034, | |
| "loss": 0.494, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2530711762683255, | |
| "grad_norm": 0.46047142148017883, | |
| "learning_rate": 0.0001965038478175436, | |
| "loss": 0.5341, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.25407145759744926, | |
| "grad_norm": 0.7829805016517639, | |
| "learning_rate": 0.00019647631941511082, | |
| "loss": 0.4661, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.2550717389265731, | |
| "grad_norm": 0.5202223062515259, | |
| "learning_rate": 0.0001964486850026507, | |
| "loss": 0.5413, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2560720202556969, | |
| "grad_norm": 0.5138536095619202, | |
| "learning_rate": 0.00019642094461052852, | |
| "loss": 0.4998, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.25707230158482075, | |
| "grad_norm": 0.44960975646972656, | |
| "learning_rate": 0.00019639309826922585, | |
| "loss": 0.5004, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.2580725829139445, | |
| "grad_norm": 0.37193313241004944, | |
| "learning_rate": 0.0001963651460093409, | |
| "loss": 0.4918, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.25907286424306836, | |
| "grad_norm": 0.41742751002311707, | |
| "learning_rate": 0.00019633708786158806, | |
| "loss": 0.5449, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.2600731455721922, | |
| "grad_norm": 0.4701450765132904, | |
| "learning_rate": 0.00019630892385679818, | |
| "loss": 0.5558, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.261073426901316, | |
| "grad_norm": 0.5549765229225159, | |
| "learning_rate": 0.00019628065402591845, | |
| "loss": 0.4763, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.2620737082304398, | |
| "grad_norm": 0.5767666697502136, | |
| "learning_rate": 0.00019625227840001225, | |
| "loss": 0.5085, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.2630739895595636, | |
| "grad_norm": 0.4518953561782837, | |
| "learning_rate": 0.0001962237970102593, | |
| "loss": 0.4957, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.26407427088868746, | |
| "grad_norm": 0.47013652324676514, | |
| "learning_rate": 0.0001961952098879555, | |
| "loss": 0.4965, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.26507455221781123, | |
| "grad_norm": 0.43617936968803406, | |
| "learning_rate": 0.00019616651706451287, | |
| "loss": 0.4466, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.26607483354693506, | |
| "grad_norm": 0.4727296531200409, | |
| "learning_rate": 0.0001961377185714597, | |
| "loss": 0.5034, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.2670751148760589, | |
| "grad_norm": 0.4447670876979828, | |
| "learning_rate": 0.0001961088144404403, | |
| "loss": 0.4418, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.2680753962051827, | |
| "grad_norm": 0.4598022699356079, | |
| "learning_rate": 0.00019607980470321505, | |
| "loss": 0.5056, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.2690756775343065, | |
| "grad_norm": 0.9431538581848145, | |
| "learning_rate": 0.00019605068939166045, | |
| "loss": 0.5425, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.27007595886343033, | |
| "grad_norm": 0.4852324426174164, | |
| "learning_rate": 0.00019602146853776894, | |
| "loss": 0.5254, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.27107624019255416, | |
| "grad_norm": 0.42556115984916687, | |
| "learning_rate": 0.000195992142173649, | |
| "loss": 0.4199, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.272076521521678, | |
| "grad_norm": 1.5440813302993774, | |
| "learning_rate": 0.0001959627103315249, | |
| "loss": 0.4965, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.27307680285080177, | |
| "grad_norm": 0.3926527202129364, | |
| "learning_rate": 0.00019593317304373705, | |
| "loss": 0.4596, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.2740770841799256, | |
| "grad_norm": 0.4427184462547302, | |
| "learning_rate": 0.00019590353034274144, | |
| "loss": 0.5012, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.27507736550904943, | |
| "grad_norm": 0.5410654544830322, | |
| "learning_rate": 0.00019587378226111014, | |
| "loss": 0.5594, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.27607764683817326, | |
| "grad_norm": 0.5077141523361206, | |
| "learning_rate": 0.00019584392883153088, | |
| "loss": 0.4821, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.27707792816729704, | |
| "grad_norm": 0.5456703901290894, | |
| "learning_rate": 0.00019581397008680717, | |
| "loss": 0.4654, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.27807820949642087, | |
| "grad_norm": 0.5426118969917297, | |
| "learning_rate": 0.00019578390605985826, | |
| "loss": 0.4844, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.2790784908255447, | |
| "grad_norm": 0.42402184009552, | |
| "learning_rate": 0.00019575373678371909, | |
| "loss": 0.4966, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2800787721546685, | |
| "grad_norm": 0.40676864981651306, | |
| "learning_rate": 0.00019572346229154025, | |
| "loss": 0.4758, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2810790534837923, | |
| "grad_norm": 0.4172145128250122, | |
| "learning_rate": 0.00019569308261658787, | |
| "loss": 0.4944, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.28207933481291614, | |
| "grad_norm": 0.39111289381980896, | |
| "learning_rate": 0.00019566259779224378, | |
| "loss": 0.5315, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.28307961614203997, | |
| "grad_norm": 0.36669549345970154, | |
| "learning_rate": 0.00019563200785200526, | |
| "loss": 0.4696, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.28407989747116374, | |
| "grad_norm": 0.48696157336235046, | |
| "learning_rate": 0.00019560131282948516, | |
| "loss": 0.4651, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.2850801788002876, | |
| "grad_norm": 0.4705601632595062, | |
| "learning_rate": 0.0001955705127584117, | |
| "loss": 0.5729, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2860804601294114, | |
| "grad_norm": 0.423532634973526, | |
| "learning_rate": 0.00019553960767262863, | |
| "loss": 0.5168, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.28708074145853524, | |
| "grad_norm": 0.7851701378822327, | |
| "learning_rate": 0.00019550859760609503, | |
| "loss": 0.5141, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.288081022787659, | |
| "grad_norm": 0.4529748857021332, | |
| "learning_rate": 0.00019547748259288536, | |
| "loss": 0.485, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.28908130411678284, | |
| "grad_norm": 0.7077706456184387, | |
| "learning_rate": 0.0001954462626671894, | |
| "loss": 0.557, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2900815854459067, | |
| "grad_norm": 0.5855894684791565, | |
| "learning_rate": 0.0001954149378633122, | |
| "loss": 0.5297, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.29108186677503045, | |
| "grad_norm": 0.7854194641113281, | |
| "learning_rate": 0.00019538350821567404, | |
| "loss": 0.4249, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.2920821481041543, | |
| "grad_norm": 0.5263127684593201, | |
| "learning_rate": 0.00019535197375881045, | |
| "loss": 0.5524, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.2930824294332781, | |
| "grad_norm": 0.38130998611450195, | |
| "learning_rate": 0.00019532033452737205, | |
| "loss": 0.5038, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.29408271076240194, | |
| "grad_norm": 0.4063386023044586, | |
| "learning_rate": 0.00019528859055612468, | |
| "loss": 0.5091, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.2950829920915257, | |
| "grad_norm": 0.4857774078845978, | |
| "learning_rate": 0.0001952567418799492, | |
| "loss": 0.5455, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.29608327342064955, | |
| "grad_norm": 0.45703741908073425, | |
| "learning_rate": 0.00019522478853384155, | |
| "loss": 0.4488, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.2970835547497734, | |
| "grad_norm": 0.3836863040924072, | |
| "learning_rate": 0.00019519273055291266, | |
| "loss": 0.4978, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.2980838360788972, | |
| "grad_norm": 0.35562822222709656, | |
| "learning_rate": 0.00019516056797238846, | |
| "loss": 0.4739, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.299084117408021, | |
| "grad_norm": 0.4030165672302246, | |
| "learning_rate": 0.00019512830082760987, | |
| "loss": 0.5039, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.3000843987371448, | |
| "grad_norm": 0.5485990047454834, | |
| "learning_rate": 0.00019509592915403255, | |
| "loss": 0.5515, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.30108468006626865, | |
| "grad_norm": 0.41883930563926697, | |
| "learning_rate": 0.00019506345298722717, | |
| "loss": 0.4664, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.3020849613953925, | |
| "grad_norm": 0.4625292420387268, | |
| "learning_rate": 0.00019503087236287913, | |
| "loss": 0.4959, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.30308524272451626, | |
| "grad_norm": 0.47342514991760254, | |
| "learning_rate": 0.00019499818731678873, | |
| "loss": 0.4354, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.3040855240536401, | |
| "grad_norm": 0.4889644384384155, | |
| "learning_rate": 0.00019496539788487082, | |
| "loss": 0.5214, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3050858053827639, | |
| "grad_norm": 0.5727283954620361, | |
| "learning_rate": 0.0001949325041031551, | |
| "loss": 0.5524, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3060860867118877, | |
| "grad_norm": 0.9029666781425476, | |
| "learning_rate": 0.0001948995060077859, | |
| "loss": 0.5522, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.3070863680410115, | |
| "grad_norm": 0.4093385934829712, | |
| "learning_rate": 0.0001948664036350221, | |
| "loss": 0.4834, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.30808664937013536, | |
| "grad_norm": 0.46531906723976135, | |
| "learning_rate": 0.00019483319702123732, | |
| "loss": 0.4903, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.3090869306992592, | |
| "grad_norm": 0.5990651249885559, | |
| "learning_rate": 0.00019479988620291956, | |
| "loss": 0.4129, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.31008721202838296, | |
| "grad_norm": 0.4223703444004059, | |
| "learning_rate": 0.00019476647121667137, | |
| "loss": 0.432, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3110874933575068, | |
| "grad_norm": 0.6224504709243774, | |
| "learning_rate": 0.00019473295209920983, | |
| "loss": 0.6105, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.3120877746866306, | |
| "grad_norm": 0.6995674967765808, | |
| "learning_rate": 0.00019469932888736632, | |
| "loss": 0.5236, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.31308805601575446, | |
| "grad_norm": 0.393481582403183, | |
| "learning_rate": 0.00019466560161808674, | |
| "loss": 0.4544, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.31408833734487823, | |
| "grad_norm": 0.9335904121398926, | |
| "learning_rate": 0.00019463177032843124, | |
| "loss": 0.4994, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.31508861867400206, | |
| "grad_norm": 0.4107264280319214, | |
| "learning_rate": 0.00019459783505557424, | |
| "loss": 0.5042, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3160889000031259, | |
| "grad_norm": 0.3685978651046753, | |
| "learning_rate": 0.00019456379583680452, | |
| "loss": 0.465, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.31708918133224967, | |
| "grad_norm": 0.36938750743865967, | |
| "learning_rate": 0.000194529652709525, | |
| "loss": 0.5036, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.3180894626613735, | |
| "grad_norm": 0.6016782522201538, | |
| "learning_rate": 0.00019449540571125286, | |
| "loss": 0.488, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.31908974399049733, | |
| "grad_norm": 0.5433719158172607, | |
| "learning_rate": 0.00019446105487961926, | |
| "loss": 0.4962, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.32009002531962116, | |
| "grad_norm": 0.387410044670105, | |
| "learning_rate": 0.0001944266002523696, | |
| "loss": 0.4358, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.32109030664874494, | |
| "grad_norm": 0.38905179500579834, | |
| "learning_rate": 0.0001943920418673633, | |
| "loss": 0.4784, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.32209058797786877, | |
| "grad_norm": 0.5600481033325195, | |
| "learning_rate": 0.00019435737976257377, | |
| "loss": 0.493, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.3230908693069926, | |
| "grad_norm": 0.4002164304256439, | |
| "learning_rate": 0.00019432261397608834, | |
| "loss": 0.4714, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.32409115063611643, | |
| "grad_norm": 0.3761899769306183, | |
| "learning_rate": 0.00019428774454610843, | |
| "loss": 0.4673, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.3250914319652402, | |
| "grad_norm": 0.387349396944046, | |
| "learning_rate": 0.00019425277151094913, | |
| "loss": 0.4321, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.32609171329436404, | |
| "grad_norm": 0.37781885266304016, | |
| "learning_rate": 0.00019421769490903957, | |
| "loss": 0.5345, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.32709199462348787, | |
| "grad_norm": 0.33928921818733215, | |
| "learning_rate": 0.0001941825147789225, | |
| "loss": 0.4169, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.32809227595261165, | |
| "grad_norm": 0.4935011565685272, | |
| "learning_rate": 0.00019414723115925456, | |
| "loss": 0.4755, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.3290925572817355, | |
| "grad_norm": 0.44166022539138794, | |
| "learning_rate": 0.0001941118440888061, | |
| "loss": 0.5851, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.3300928386108593, | |
| "grad_norm": 0.4288978576660156, | |
| "learning_rate": 0.0001940763536064611, | |
| "loss": 0.478, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.33109311993998314, | |
| "grad_norm": 0.45652419328689575, | |
| "learning_rate": 0.00019404075975121716, | |
| "loss": 0.5151, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.3320934012691069, | |
| "grad_norm": 0.4268530607223511, | |
| "learning_rate": 0.0001940050625621855, | |
| "loss": 0.4809, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.33309368259823074, | |
| "grad_norm": 0.40482282638549805, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.4439, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.3340939639273546, | |
| "grad_norm": 0.4686596691608429, | |
| "learning_rate": 0.0001939333583397715, | |
| "loss": 0.4791, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.3350942452564784, | |
| "grad_norm": 0.4924260377883911, | |
| "learning_rate": 0.00019389735138517915, | |
| "loss": 0.4592, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3360945265856022, | |
| "grad_norm": 0.5177869200706482, | |
| "learning_rate": 0.00019386124125437895, | |
| "loss": 0.4488, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.337094807914726, | |
| "grad_norm": 0.5712275505065918, | |
| "learning_rate": 0.00019382502798704935, | |
| "loss": 0.4791, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.33809508924384984, | |
| "grad_norm": 0.5464945435523987, | |
| "learning_rate": 0.00019378871162298227, | |
| "loss": 0.5554, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.3390953705729737, | |
| "grad_norm": 0.5113401412963867, | |
| "learning_rate": 0.00019375229220208276, | |
| "loss": 0.4479, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.34009565190209745, | |
| "grad_norm": 0.451701819896698, | |
| "learning_rate": 0.00019371576976436917, | |
| "loss": 0.5338, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3410959332312213, | |
| "grad_norm": 0.41110074520111084, | |
| "learning_rate": 0.00019367914434997312, | |
| "loss": 0.4551, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.3420962145603451, | |
| "grad_norm": 0.444566547870636, | |
| "learning_rate": 0.00019364241599913924, | |
| "loss": 0.5034, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.3430964958894689, | |
| "grad_norm": 1.1947789192199707, | |
| "learning_rate": 0.0001936055847522254, | |
| "loss": 0.4569, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.3440967772185927, | |
| "grad_norm": 0.5699276328086853, | |
| "learning_rate": 0.00019356865064970244, | |
| "loss": 0.4675, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.34509705854771655, | |
| "grad_norm": 0.3912697732448578, | |
| "learning_rate": 0.0001935316137321543, | |
| "loss": 0.5124, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3460973398768404, | |
| "grad_norm": 0.4574461579322815, | |
| "learning_rate": 0.00019349447404027782, | |
| "loss": 0.4276, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.34709762120596416, | |
| "grad_norm": 0.6651443839073181, | |
| "learning_rate": 0.00019345723161488283, | |
| "loss": 0.4731, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.348097902535088, | |
| "grad_norm": 0.7808742523193359, | |
| "learning_rate": 0.000193419886496892, | |
| "loss": 0.4953, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.3490981838642118, | |
| "grad_norm": 0.4347785711288452, | |
| "learning_rate": 0.00019338243872734086, | |
| "loss": 0.4121, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.35009846519333565, | |
| "grad_norm": 0.5245057344436646, | |
| "learning_rate": 0.00019334488834737775, | |
| "loss": 0.4786, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3510987465224594, | |
| "grad_norm": 0.3696887195110321, | |
| "learning_rate": 0.00019330723539826375, | |
| "loss": 0.4757, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.35209902785158326, | |
| "grad_norm": 0.4131191372871399, | |
| "learning_rate": 0.00019326947992137262, | |
| "loss": 0.4781, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.3530993091807071, | |
| "grad_norm": 0.419505774974823, | |
| "learning_rate": 0.00019323162195819082, | |
| "loss": 0.5067, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.35409959050983086, | |
| "grad_norm": 0.4195800721645355, | |
| "learning_rate": 0.0001931936615503174, | |
| "loss": 0.4244, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.3550998718389547, | |
| "grad_norm": 0.384988397359848, | |
| "learning_rate": 0.000193155598739464, | |
| "loss": 0.4633, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3561001531680785, | |
| "grad_norm": 0.432335764169693, | |
| "learning_rate": 0.0001931174335674547, | |
| "loss": 0.4616, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.35710043449720236, | |
| "grad_norm": 0.4696163237094879, | |
| "learning_rate": 0.0001930791660762262, | |
| "loss": 0.4931, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.35810071582632613, | |
| "grad_norm": 0.44555091857910156, | |
| "learning_rate": 0.00019304079630782752, | |
| "loss": 0.4916, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.35910099715544996, | |
| "grad_norm": 0.3690800368785858, | |
| "learning_rate": 0.0001930023243044201, | |
| "loss": 0.4411, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.3601012784845738, | |
| "grad_norm": 0.5480444431304932, | |
| "learning_rate": 0.00019296375010827773, | |
| "loss": 0.5076, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3611015598136976, | |
| "grad_norm": 0.45511335134506226, | |
| "learning_rate": 0.00019292507376178643, | |
| "loss": 0.5281, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.3621018411428214, | |
| "grad_norm": 0.42300713062286377, | |
| "learning_rate": 0.00019288629530744454, | |
| "loss": 0.532, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.36310212247194523, | |
| "grad_norm": 0.44190162420272827, | |
| "learning_rate": 0.0001928474147878626, | |
| "loss": 0.4408, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.36410240380106906, | |
| "grad_norm": 0.540590226650238, | |
| "learning_rate": 0.0001928084322457632, | |
| "loss": 0.4755, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.3651026851301929, | |
| "grad_norm": 0.559669554233551, | |
| "learning_rate": 0.00019276934772398114, | |
| "loss": 0.4183, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.36610296645931667, | |
| "grad_norm": 0.4601163864135742, | |
| "learning_rate": 0.00019273016126546323, | |
| "loss": 0.4451, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.3671032477884405, | |
| "grad_norm": 0.41059058904647827, | |
| "learning_rate": 0.00019269087291326833, | |
| "loss": 0.5118, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.36810352911756433, | |
| "grad_norm": 0.365362286567688, | |
| "learning_rate": 0.00019265148271056722, | |
| "loss": 0.4125, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.3691038104466881, | |
| "grad_norm": 0.4561755061149597, | |
| "learning_rate": 0.0001926119907006426, | |
| "loss": 0.5015, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.37010409177581194, | |
| "grad_norm": 0.37689074873924255, | |
| "learning_rate": 0.00019257239692688907, | |
| "loss": 0.4954, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.37110437310493577, | |
| "grad_norm": 0.3596641421318054, | |
| "learning_rate": 0.00019253270143281296, | |
| "loss": 0.452, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.3721046544340596, | |
| "grad_norm": 0.408954381942749, | |
| "learning_rate": 0.00019249290426203252, | |
| "loss": 0.4491, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.3731049357631834, | |
| "grad_norm": 0.46262991428375244, | |
| "learning_rate": 0.0001924530054582776, | |
| "loss": 0.452, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3741052170923072, | |
| "grad_norm": 0.3985409438610077, | |
| "learning_rate": 0.0001924130050653898, | |
| "loss": 0.4546, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.37510549842143104, | |
| "grad_norm": 0.3863021731376648, | |
| "learning_rate": 0.00019237290312732226, | |
| "loss": 0.4425, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.37610577975055487, | |
| "grad_norm": 0.36261245608329773, | |
| "learning_rate": 0.00019233269968813984, | |
| "loss": 0.4499, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.37710606107967864, | |
| "grad_norm": 0.34936317801475525, | |
| "learning_rate": 0.00019229239479201876, | |
| "loss": 0.4001, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.3781063424088025, | |
| "grad_norm": 0.36340323090553284, | |
| "learning_rate": 0.0001922519884832469, | |
| "loss": 0.4893, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.3791066237379263, | |
| "grad_norm": 0.4176476001739502, | |
| "learning_rate": 0.0001922114808062234, | |
| "loss": 0.4451, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.3801069050670501, | |
| "grad_norm": 0.7352374792098999, | |
| "learning_rate": 0.00019217087180545893, | |
| "loss": 0.4793, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3811071863961739, | |
| "grad_norm": 0.36440688371658325, | |
| "learning_rate": 0.0001921301615255754, | |
| "loss": 0.4881, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.38210746772529774, | |
| "grad_norm": 0.3963299095630646, | |
| "learning_rate": 0.0001920893500113061, | |
| "loss": 0.442, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.3831077490544216, | |
| "grad_norm": 0.3732217848300934, | |
| "learning_rate": 0.00019204843730749547, | |
| "loss": 0.4383, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.38410803038354535, | |
| "grad_norm": 0.3525499999523163, | |
| "learning_rate": 0.00019200742345909915, | |
| "loss": 0.4356, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.3851083117126692, | |
| "grad_norm": 0.35598984360694885, | |
| "learning_rate": 0.00019196630851118398, | |
| "loss": 0.4441, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.386108593041793, | |
| "grad_norm": 0.39109891653060913, | |
| "learning_rate": 0.0001919250925089278, | |
| "loss": 0.4396, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.38710887437091684, | |
| "grad_norm": 0.5131281018257141, | |
| "learning_rate": 0.00019188377549761963, | |
| "loss": 0.4569, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.3881091557000406, | |
| "grad_norm": 0.5972152948379517, | |
| "learning_rate": 0.00019184235752265928, | |
| "loss": 0.4571, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.38910943702916445, | |
| "grad_norm": 0.518309473991394, | |
| "learning_rate": 0.00019180083862955772, | |
| "loss": 0.4804, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3901097183582883, | |
| "grad_norm": 0.44072574377059937, | |
| "learning_rate": 0.00019175921886393666, | |
| "loss": 0.5042, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.39110999968741206, | |
| "grad_norm": 0.42578837275505066, | |
| "learning_rate": 0.00019171749827152869, | |
| "loss": 0.5209, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.3921102810165359, | |
| "grad_norm": 0.5175500512123108, | |
| "learning_rate": 0.0001916756768981772, | |
| "loss": 0.4601, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.3931105623456597, | |
| "grad_norm": 0.42719781398773193, | |
| "learning_rate": 0.00019163375478983632, | |
| "loss": 0.4508, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.39411084367478355, | |
| "grad_norm": 0.4961738586425781, | |
| "learning_rate": 0.00019159173199257085, | |
| "loss": 0.4627, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.3951111250039073, | |
| "grad_norm": 0.39553913474082947, | |
| "learning_rate": 0.00019154960855255628, | |
| "loss": 0.4836, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.39611140633303116, | |
| "grad_norm": 0.36840665340423584, | |
| "learning_rate": 0.0001915073845160786, | |
| "loss": 0.4378, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.397111687662155, | |
| "grad_norm": 0.5660284161567688, | |
| "learning_rate": 0.00019146505992953446, | |
| "loss": 0.3921, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.3981119689912788, | |
| "grad_norm": 0.3654325306415558, | |
| "learning_rate": 0.00019142263483943085, | |
| "loss": 0.4646, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.3991122503204026, | |
| "grad_norm": 0.6575559377670288, | |
| "learning_rate": 0.00019138010929238534, | |
| "loss": 0.4916, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.4001125316495264, | |
| "grad_norm": 0.37716901302337646, | |
| "learning_rate": 0.00019133748333512575, | |
| "loss": 0.3914, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.40111281297865026, | |
| "grad_norm": 0.40329572558403015, | |
| "learning_rate": 0.00019129475701449035, | |
| "loss": 0.4456, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.4021130943077741, | |
| "grad_norm": 0.3746117055416107, | |
| "learning_rate": 0.0001912519303774276, | |
| "loss": 0.4727, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.40311337563689786, | |
| "grad_norm": 0.580685019493103, | |
| "learning_rate": 0.0001912090034709963, | |
| "loss": 0.4281, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.4041136569660217, | |
| "grad_norm": 0.3667939007282257, | |
| "learning_rate": 0.00019116597634236525, | |
| "loss": 0.4145, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.4051139382951455, | |
| "grad_norm": 0.4624421000480652, | |
| "learning_rate": 0.0001911228490388136, | |
| "loss": 0.4459, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.4061142196242693, | |
| "grad_norm": 0.422579288482666, | |
| "learning_rate": 0.00019107962160773035, | |
| "loss": 0.3903, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.40711450095339313, | |
| "grad_norm": 0.40308403968811035, | |
| "learning_rate": 0.0001910362940966147, | |
| "loss": 0.4568, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.40811478228251696, | |
| "grad_norm": 0.3817251920700073, | |
| "learning_rate": 0.00019099286655307568, | |
| "loss": 0.4944, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.4091150636116408, | |
| "grad_norm": 0.6623880863189697, | |
| "learning_rate": 0.0001909493390248324, | |
| "loss": 0.3936, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.41011534494076457, | |
| "grad_norm": 0.3836446702480316, | |
| "learning_rate": 0.00019090571155971366, | |
| "loss": 0.4518, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4111156262698884, | |
| "grad_norm": 0.7342603206634521, | |
| "learning_rate": 0.00019086198420565823, | |
| "loss": 0.4008, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.41211590759901223, | |
| "grad_norm": 0.38605815172195435, | |
| "learning_rate": 0.00019081815701071445, | |
| "loss": 0.3825, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.41311618892813606, | |
| "grad_norm": 0.4390144646167755, | |
| "learning_rate": 0.0001907742300230406, | |
| "loss": 0.5158, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.41411647025725984, | |
| "grad_norm": 0.4054723381996155, | |
| "learning_rate": 0.00019073020329090444, | |
| "loss": 0.4426, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.41511675158638367, | |
| "grad_norm": 0.37295302748680115, | |
| "learning_rate": 0.0001906860768626834, | |
| "loss": 0.4147, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4161170329155075, | |
| "grad_norm": 0.41723310947418213, | |
| "learning_rate": 0.00019064185078686443, | |
| "loss": 0.4136, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.4171173142446313, | |
| "grad_norm": 0.4265085458755493, | |
| "learning_rate": 0.000190597525112044, | |
| "loss": 0.4546, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.4181175955737551, | |
| "grad_norm": 0.6407977342605591, | |
| "learning_rate": 0.000190553099886928, | |
| "loss": 0.4569, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.41911787690287894, | |
| "grad_norm": 0.41884028911590576, | |
| "learning_rate": 0.00019050857516033173, | |
| "loss": 0.4187, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.42011815823200277, | |
| "grad_norm": 0.4397021234035492, | |
| "learning_rate": 0.00019046395098117983, | |
| "loss": 0.4551, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.42111843956112655, | |
| "grad_norm": 0.39388948678970337, | |
| "learning_rate": 0.00019041922739850616, | |
| "loss": 0.4561, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.4221187208902504, | |
| "grad_norm": 0.4537939429283142, | |
| "learning_rate": 0.00019037440446145385, | |
| "loss": 0.5148, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.4231190022193742, | |
| "grad_norm": 0.41184279322624207, | |
| "learning_rate": 0.00019032948221927524, | |
| "loss": 0.4562, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.42411928354849804, | |
| "grad_norm": 0.4077985882759094, | |
| "learning_rate": 0.00019028446072133175, | |
| "loss": 0.4862, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.4251195648776218, | |
| "grad_norm": 0.3779621124267578, | |
| "learning_rate": 0.00019023934001709383, | |
| "loss": 0.4127, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.42611984620674564, | |
| "grad_norm": 0.7072260975837708, | |
| "learning_rate": 0.00019019412015614098, | |
| "loss": 0.4615, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.4271201275358695, | |
| "grad_norm": 0.5229703783988953, | |
| "learning_rate": 0.00019014880118816164, | |
| "loss": 0.4456, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.4281204088649933, | |
| "grad_norm": 0.4435012936592102, | |
| "learning_rate": 0.0001901033831629532, | |
| "loss": 0.455, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.4291206901941171, | |
| "grad_norm": 0.36504679918289185, | |
| "learning_rate": 0.00019005786613042185, | |
| "loss": 0.4972, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.4301209715232409, | |
| "grad_norm": 0.38636064529418945, | |
| "learning_rate": 0.00019001225014058255, | |
| "loss": 0.449, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.43112125285236474, | |
| "grad_norm": 0.4293528199195862, | |
| "learning_rate": 0.00018996653524355902, | |
| "loss": 0.4844, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.4321215341814885, | |
| "grad_norm": 0.37384432554244995, | |
| "learning_rate": 0.00018992072148958368, | |
| "loss": 0.479, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.43312181551061235, | |
| "grad_norm": 0.4167514145374298, | |
| "learning_rate": 0.00018987480892899758, | |
| "loss": 0.4173, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.4341220968397362, | |
| "grad_norm": 0.4135938882827759, | |
| "learning_rate": 0.00018982879761225027, | |
| "loss": 0.4681, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.43512237816886, | |
| "grad_norm": 0.3738775849342346, | |
| "learning_rate": 0.00018978268758989991, | |
| "loss": 0.4707, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4361226594979838, | |
| "grad_norm": 0.43221938610076904, | |
| "learning_rate": 0.00018973647891261307, | |
| "loss": 0.4261, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.4371229408271076, | |
| "grad_norm": 0.7024879455566406, | |
| "learning_rate": 0.00018969017163116472, | |
| "loss": 0.4826, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.43812322215623145, | |
| "grad_norm": 0.3586277961730957, | |
| "learning_rate": 0.0001896437657964382, | |
| "loss": 0.4867, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.4391235034853553, | |
| "grad_norm": 0.41892313957214355, | |
| "learning_rate": 0.00018959726145942508, | |
| "loss": 0.4156, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.44012378481447906, | |
| "grad_norm": 0.3496420681476593, | |
| "learning_rate": 0.00018955065867122528, | |
| "loss": 0.37, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4411240661436029, | |
| "grad_norm": 0.39177075028419495, | |
| "learning_rate": 0.00018950395748304678, | |
| "loss": 0.4554, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.4421243474727267, | |
| "grad_norm": 0.3440227508544922, | |
| "learning_rate": 0.0001894571579462058, | |
| "loss": 0.4359, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.4431246288018505, | |
| "grad_norm": 0.41202670335769653, | |
| "learning_rate": 0.00018941026011212654, | |
| "loss": 0.4068, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.4441249101309743, | |
| "grad_norm": 0.4585283696651459, | |
| "learning_rate": 0.00018936326403234125, | |
| "loss": 0.4962, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.44512519146009816, | |
| "grad_norm": 0.4074370861053467, | |
| "learning_rate": 0.00018931616975849006, | |
| "loss": 0.477, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.446125472789222, | |
| "grad_norm": 0.7314764857292175, | |
| "learning_rate": 0.00018926897734232115, | |
| "loss": 0.4972, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.44712575411834576, | |
| "grad_norm": 0.46429434418678284, | |
| "learning_rate": 0.0001892216868356904, | |
| "loss": 0.4283, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.4481260354474696, | |
| "grad_norm": 0.5269820094108582, | |
| "learning_rate": 0.0001891742982905615, | |
| "loss": 0.4563, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.4491263167765934, | |
| "grad_norm": 0.37672704458236694, | |
| "learning_rate": 0.00018912681175900598, | |
| "loss": 0.463, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.45012659810571726, | |
| "grad_norm": 1.1096312999725342, | |
| "learning_rate": 0.00018907922729320285, | |
| "loss": 0.4775, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.45112687943484103, | |
| "grad_norm": 0.4609079957008362, | |
| "learning_rate": 0.00018903154494543889, | |
| "loss": 0.5329, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.45212716076396486, | |
| "grad_norm": 0.42541447281837463, | |
| "learning_rate": 0.00018898376476810834, | |
| "loss": 0.4258, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.4531274420930887, | |
| "grad_norm": 0.42514634132385254, | |
| "learning_rate": 0.00018893588681371303, | |
| "loss": 0.4433, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.45412772342221247, | |
| "grad_norm": 0.6249622106552124, | |
| "learning_rate": 0.00018888791113486213, | |
| "loss": 0.5478, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.4551280047513363, | |
| "grad_norm": 0.46454930305480957, | |
| "learning_rate": 0.00018883983778427227, | |
| "loss": 0.4129, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.45612828608046013, | |
| "grad_norm": 0.3539360463619232, | |
| "learning_rate": 0.0001887916668147673, | |
| "loss": 0.4428, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.45712856740958396, | |
| "grad_norm": 0.3984419107437134, | |
| "learning_rate": 0.00018874339827927846, | |
| "loss": 0.4802, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.45812884873870774, | |
| "grad_norm": 0.4771546721458435, | |
| "learning_rate": 0.00018869503223084414, | |
| "loss": 0.4729, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.45912913006783157, | |
| "grad_norm": 0.40642663836479187, | |
| "learning_rate": 0.00018864656872260985, | |
| "loss": 0.378, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.4601294113969554, | |
| "grad_norm": 0.42019134759902954, | |
| "learning_rate": 0.00018859800780782828, | |
| "loss": 0.4128, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.46112969272607923, | |
| "grad_norm": 0.3868459463119507, | |
| "learning_rate": 0.000188549349539859, | |
| "loss": 0.4626, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.462129974055203, | |
| "grad_norm": 0.31789880990982056, | |
| "learning_rate": 0.00018850059397216876, | |
| "loss": 0.4054, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.46313025538432684, | |
| "grad_norm": 0.32979318499565125, | |
| "learning_rate": 0.00018845174115833099, | |
| "loss": 0.3991, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.46413053671345067, | |
| "grad_norm": 0.33265942335128784, | |
| "learning_rate": 0.0001884027911520262, | |
| "loss": 0.411, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.4651308180425745, | |
| "grad_norm": 0.3669120967388153, | |
| "learning_rate": 0.00018835374400704154, | |
| "loss": 0.3834, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4661310993716983, | |
| "grad_norm": 0.40603357553482056, | |
| "learning_rate": 0.00018830459977727096, | |
| "loss": 0.455, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.4671313807008221, | |
| "grad_norm": 0.43093356490135193, | |
| "learning_rate": 0.0001882553585167151, | |
| "loss": 0.3614, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.46813166202994594, | |
| "grad_norm": 0.5963067412376404, | |
| "learning_rate": 0.00018820602027948114, | |
| "loss": 0.4074, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.4691319433590697, | |
| "grad_norm": 0.4733903706073761, | |
| "learning_rate": 0.00018815658511978298, | |
| "loss": 0.4063, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.47013222468819355, | |
| "grad_norm": 1.338910698890686, | |
| "learning_rate": 0.00018810705309194083, | |
| "loss": 0.455, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4711325060173174, | |
| "grad_norm": 0.3329702317714691, | |
| "learning_rate": 0.00018805742425038145, | |
| "loss": 0.3776, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.4721327873464412, | |
| "grad_norm": 0.362064927816391, | |
| "learning_rate": 0.00018800769864963802, | |
| "loss": 0.4351, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.473133068675565, | |
| "grad_norm": 0.36146995425224304, | |
| "learning_rate": 0.00018795787634434994, | |
| "loss": 0.4226, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.4741333500046888, | |
| "grad_norm": 0.34001466631889343, | |
| "learning_rate": 0.0001879079573892629, | |
| "loss": 0.401, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.47513363133381264, | |
| "grad_norm": 0.3179371953010559, | |
| "learning_rate": 0.00018785794183922883, | |
| "loss": 0.4175, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4761339126629365, | |
| "grad_norm": 0.5295701622962952, | |
| "learning_rate": 0.00018780782974920572, | |
| "loss": 0.4233, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.47713419399206025, | |
| "grad_norm": 0.3790874779224396, | |
| "learning_rate": 0.00018775762117425777, | |
| "loss": 0.4353, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.4781344753211841, | |
| "grad_norm": 0.38352811336517334, | |
| "learning_rate": 0.0001877073161695551, | |
| "loss": 0.4574, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.4791347566503079, | |
| "grad_norm": 0.4280262291431427, | |
| "learning_rate": 0.00018765691479037376, | |
| "loss": 0.4525, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.4801350379794317, | |
| "grad_norm": 0.36147376894950867, | |
| "learning_rate": 0.00018760641709209583, | |
| "loss": 0.4475, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4811353193085555, | |
| "grad_norm": 0.6647056937217712, | |
| "learning_rate": 0.0001875558231302091, | |
| "loss": 0.4051, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.48213560063767935, | |
| "grad_norm": 0.4935802221298218, | |
| "learning_rate": 0.00018750513296030718, | |
| "loss": 0.4151, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.4831358819668032, | |
| "grad_norm": 0.5030160546302795, | |
| "learning_rate": 0.00018745434663808942, | |
| "loss": 0.5521, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.48413616329592696, | |
| "grad_norm": 0.3784274160861969, | |
| "learning_rate": 0.0001874034642193608, | |
| "loss": 0.417, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.4851364446250508, | |
| "grad_norm": 0.35883280634880066, | |
| "learning_rate": 0.0001873524857600319, | |
| "loss": 0.4672, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4861367259541746, | |
| "grad_norm": 0.3392554819583893, | |
| "learning_rate": 0.00018730141131611882, | |
| "loss": 0.4166, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.48713700728329845, | |
| "grad_norm": 0.5107550621032715, | |
| "learning_rate": 0.00018725024094374315, | |
| "loss": 0.409, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.4881372886124222, | |
| "grad_norm": 0.44673365354537964, | |
| "learning_rate": 0.00018719897469913184, | |
| "loss": 0.409, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.48913756994154606, | |
| "grad_norm": 0.4512506425380707, | |
| "learning_rate": 0.00018714761263861728, | |
| "loss": 0.3925, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.4901378512706699, | |
| "grad_norm": 0.42347124218940735, | |
| "learning_rate": 0.000187096154818637, | |
| "loss": 0.4185, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4911381325997937, | |
| "grad_norm": 0.5886827111244202, | |
| "learning_rate": 0.00018704460129573391, | |
| "loss": 0.3909, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.4921384139289175, | |
| "grad_norm": 0.40246817469596863, | |
| "learning_rate": 0.00018699295212655596, | |
| "loss": 0.4267, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.4931386952580413, | |
| "grad_norm": 0.5066978931427002, | |
| "learning_rate": 0.00018694120736785632, | |
| "loss": 0.4437, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.49413897658716516, | |
| "grad_norm": 0.36570316553115845, | |
| "learning_rate": 0.00018688936707649304, | |
| "loss": 0.4221, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.49513925791628893, | |
| "grad_norm": 0.4331252872943878, | |
| "learning_rate": 0.00018683743130942928, | |
| "loss": 0.4173, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.49613953924541276, | |
| "grad_norm": 0.9528374671936035, | |
| "learning_rate": 0.00018678540012373302, | |
| "loss": 0.4185, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.4971398205745366, | |
| "grad_norm": 0.3963034152984619, | |
| "learning_rate": 0.00018673327357657715, | |
| "loss": 0.4448, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.4981401019036604, | |
| "grad_norm": 0.7620049715042114, | |
| "learning_rate": 0.0001866810517252393, | |
| "loss": 0.4721, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.4991403832327842, | |
| "grad_norm": 0.5681329369544983, | |
| "learning_rate": 0.00018662873462710184, | |
| "loss": 0.4317, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.5001406645619081, | |
| "grad_norm": 0.4743039608001709, | |
| "learning_rate": 0.0001865763223396518, | |
| "loss": 0.4232, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 2997, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5343957202649088.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |