| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.779627806351935, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 0.8321, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 7.733700169087985, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.876, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 7.238197182487587, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 0.9343, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 7.603617636922086, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.9237, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.818599974024641, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.8329, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.828681062053993, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.7508, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.946872126320063, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.7864, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.0446358678116603, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.724, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 3.739861312910538, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.7341, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.4227376657752946, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.6941, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.391771433690599, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.6809, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.2009407499112648, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.6218, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.3765840923942827, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5491, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.3999756241646164, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.6052, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.8401582850030842, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.5614, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.0515653437541763, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.5627, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.4292668893585168, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.5473, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.948375472471034, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.5612, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.1500236041505767, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.5184, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.56511087486709, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.5239, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.4550230706327607, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.4476, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.4683221764519665, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.4343, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.9598764618893638, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.4513, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.8748226437296447, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.4223, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.9915570304867101, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.3872, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.0232047465010423, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.4575, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.2242842361295692, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.4301, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.0130665777865844, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.4214, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.9414805063646674, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.4099, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0749237269903371, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.3338, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 1.060179945113561, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.3138, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.8224397239963451, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.3326, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 1.4173542158443497, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.3049, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 1.4399741661681684, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.2553, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.9425436420467089, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.2908, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 1.4458019448596084, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.2717, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.968676737156165, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.262, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.8827324918142179, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.2566, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.811311364516425, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.2688, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.898907543496273, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.2442, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.828754132086278, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.2016, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.7758159042595095, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.1948, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.7488933544212325, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.1538, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 1.0658916255129685, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.2353, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.903164330113684, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.2287, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.7447429897898407, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.1544, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.8077667346749938, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.2015, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.7466964986415479, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.1635, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.8252912787487685, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.1652, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.7489178905315844, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.1806, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.5821897945848419, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.1139, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.7241098658980318, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.1273, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.6255431736720212, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.1295, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.691002790072626, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.1265, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.8846004437927963, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.1054, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.7954288609280562, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.1334, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.7354679769808076, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.1613, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.6454613491986969, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.1378, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.5480281731422116, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.0912, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6713079133136983, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.1419, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.5735993174644404, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.0855, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.5260788939778733, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.0776, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.5086340807856387, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.0567, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.6863184640140559, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.1049, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.7053508342490321, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.1187, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.6907230517007713, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.0778, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.5662192151886665, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.1261, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.5500097783951419, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.0729, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.5199610486134152, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.0639, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.5703471993073134, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.0868, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.48459618363091345, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.0603, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.4221018661488874, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.0564, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.4479624595473982, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.0676, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.6159422075741188, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.0969, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.4778532002901978, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.0515, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.519918127935992, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.0989, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.5518238984180506, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.0518, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.6162483926314095, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.0357, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.5410786019743518, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.0759, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.34043415172692976, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.0213, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.45491123612455514, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.0563, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.405793593930519, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.0323, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.4128595719166626, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.0497, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.3943040299802371, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.0698, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.40453966325471846, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.0485, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.3527227038646973, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.0474, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.3113189499234012, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.0223, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.47969508626162477, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.052, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.45232502093737514, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.0431, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.42026240894441896, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.0395, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.3105540765776395, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0422, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.33705487084981545, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.0131, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.2651102122483395, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.0389, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.23502236871851126, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.0203, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.2852899739737784, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.0455, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.5135177571505845, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.0374, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.3109771044901478, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.0273, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.2794741292253182, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.0498, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.46775927936415107, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.0501, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.2544518030008082, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.0248, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.21067652452529076, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.0311, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.16891889075081276, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.0164, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.19557809712868815, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.039, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.21154537490243636, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.027, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.19588171489260067, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.0119, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.4830359892264546, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.0232, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.2483409232485105, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.0233, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.2978512584627456, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.0419, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.24272181112135727, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.0517, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.26268861351063316, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.0285, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.1660762052563307, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.0228, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.28125815338610055, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.0436, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.1085402493742939, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.0137, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.19439728399973843, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.0318, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.1411492540257263, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.0132, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.2552476524426269, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.0472, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.18961590990356175, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.0225, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.16134959351312292, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.01, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.20865845319758783, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.0491, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.15530459393079682, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.0155, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.1555202802598813, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.0133, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.2052351226381838, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.0254, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.17749939068887433, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.0222, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.21338816615021317, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.0349, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.16918304347541172, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.0292, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.18651719491638674, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.0331, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.18264072089250763, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.029, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.18867489235805426, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.0284, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.19979335321317201, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.0269, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.15388280708793917, | |
| "learning_rate": 0.0, | |
| "loss": 0.0146, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 2.50173978574848e+16, | |
| "train_loss": 0.20831933366134764, | |
| "train_runtime": 2041.6117, | |
| "train_samples_per_second": 2.012, | |
| "train_steps_per_second": 0.064 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "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": 2.50173978574848e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |