| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 152, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013157894736842105, | |
| "grad_norm": 13.235902064816667, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.3431, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02631578947368421, | |
| "grad_norm": 13.138228566029483, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.3429, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.039473684210526314, | |
| "grad_norm": 11.866560617710881, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.2846, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 11.872638763731226, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.3455, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06578947368421052, | |
| "grad_norm": 7.262421698320824, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.2313, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07894736842105263, | |
| "grad_norm": 5.162108989306741, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.1935, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09210526315789473, | |
| "grad_norm": 4.989963293072479, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.1607, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 3.2196850103315575, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1746, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11842105263157894, | |
| "grad_norm": 9.267917781353754, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.2141, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13157894736842105, | |
| "grad_norm": 2.9550126149670968, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.1466, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14473684210526316, | |
| "grad_norm": 3.2305434151019665, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.1559, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 2.303728369668543, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.1459, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.17105263157894737, | |
| "grad_norm": 2.1650046636974225, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.1158, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.18421052631578946, | |
| "grad_norm": 2.43429839604823, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.1429, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19736842105263158, | |
| "grad_norm": 2.256199461941204, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.1569, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 2.540675164259285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1389, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2236842105263158, | |
| "grad_norm": 3.0603124517729934, | |
| "learning_rate": 9.998666040558187e-06, | |
| "loss": 0.1481, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23684210526315788, | |
| "grad_norm": 3.15679704599401, | |
| "learning_rate": 9.994664874011864e-06, | |
| "loss": 0.1543, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.2073031998301942, | |
| "learning_rate": 9.987998635318586e-06, | |
| "loss": 0.0912, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 1.5485358071033657, | |
| "learning_rate": 9.978670881475173e-06, | |
| "loss": 0.1454, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.27631578947368424, | |
| "grad_norm": 1.837098464284506, | |
| "learning_rate": 9.96668658961975e-06, | |
| "loss": 0.1263, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2894736842105263, | |
| "grad_norm": 2.195521777946691, | |
| "learning_rate": 9.952052154376027e-06, | |
| "loss": 0.1347, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3026315789473684, | |
| "grad_norm": 1.4759607642496113, | |
| "learning_rate": 9.93477538444123e-06, | |
| "loss": 0.1036, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 3.4617468292311653, | |
| "learning_rate": 9.91486549841951e-06, | |
| "loss": 0.1401, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32894736842105265, | |
| "grad_norm": 1.8290234660376485, | |
| "learning_rate": 9.892333119903045e-06, | |
| "loss": 0.1191, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.34210526315789475, | |
| "grad_norm": 1.9446124807007885, | |
| "learning_rate": 9.867190271803466e-06, | |
| "loss": 0.1156, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.35526315789473684, | |
| "grad_norm": 1.924481905745065, | |
| "learning_rate": 9.839450369936615e-06, | |
| "loss": 0.1203, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 0.9472794741039898, | |
| "learning_rate": 9.809128215864096e-06, | |
| "loss": 0.084, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3815789473684211, | |
| "grad_norm": 1.8614087289828243, | |
| "learning_rate": 9.776239988995401e-06, | |
| "loss": 0.1237, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.39473684210526316, | |
| "grad_norm": 1.8725601311672047, | |
| "learning_rate": 9.74080323795483e-06, | |
| "loss": 0.109, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.40789473684210525, | |
| "grad_norm": 1.6989866208387698, | |
| "learning_rate": 9.702836871217838e-06, | |
| "loss": 0.1078, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 1.4620493708990816, | |
| "learning_rate": 9.66236114702178e-06, | |
| "loss": 0.0993, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4342105263157895, | |
| "grad_norm": 1.4349318310245505, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.1168, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4473684210526316, | |
| "grad_norm": 1.2971799391770196, | |
| "learning_rate": 9.573969342440107e-06, | |
| "loss": 0.0989, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.4605263157894737, | |
| "grad_norm": 1.3284875209748295, | |
| "learning_rate": 9.52610042648741e-06, | |
| "loss": 0.1052, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 1.517294107060111, | |
| "learning_rate": 9.475816456775313e-06, | |
| "loss": 0.1095, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4868421052631579, | |
| "grad_norm": 1.8550537994393244, | |
| "learning_rate": 9.423144264014278e-06, | |
| "loss": 0.0975, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.2490574894196733, | |
| "learning_rate": 9.368111953231849e-06, | |
| "loss": 0.1171, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5131578947368421, | |
| "grad_norm": 1.839240097339726, | |
| "learning_rate": 9.310748888776254e-06, | |
| "loss": 0.0916, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.0701426736883897, | |
| "learning_rate": 9.251085678648072e-06, | |
| "loss": 0.1004, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5394736842105263, | |
| "grad_norm": 1.3170647678764655, | |
| "learning_rate": 9.189154158168293e-06, | |
| "loss": 0.0873, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5526315789473685, | |
| "grad_norm": 1.2646530838963848, | |
| "learning_rate": 9.124987372991512e-06, | |
| "loss": 0.1062, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5657894736842105, | |
| "grad_norm": 2.0376323047774947, | |
| "learning_rate": 9.058619561473308e-06, | |
| "loss": 0.1446, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 1.4525149631147896, | |
| "learning_rate": 8.990086136401199e-06, | |
| "loss": 0.1063, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5921052631578947, | |
| "grad_norm": 1.7227367203376753, | |
| "learning_rate": 8.91942366609897e-06, | |
| "loss": 0.1288, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6052631578947368, | |
| "grad_norm": 1.6374368378542756, | |
| "learning_rate": 8.846669854914395e-06, | |
| "loss": 0.1101, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.618421052631579, | |
| "grad_norm": 1.776661892531926, | |
| "learning_rate": 8.771863523100821e-06, | |
| "loss": 0.0863, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 2.101189002973872, | |
| "learning_rate": 8.695044586103297e-06, | |
| "loss": 0.1155, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6447368421052632, | |
| "grad_norm": 1.4990143401140024, | |
| "learning_rate": 8.616254033260351e-06, | |
| "loss": 0.1233, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6578947368421053, | |
| "grad_norm": 1.300205397968391, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.0722, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6710526315789473, | |
| "grad_norm": 1.2976123144962795, | |
| "learning_rate": 8.452927275070858e-06, | |
| "loss": 0.0955, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 3.1993999412027807, | |
| "learning_rate": 8.368478218232787e-06, | |
| "loss": 0.0992, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6973684210526315, | |
| "grad_norm": 1.511040367087983, | |
| "learning_rate": 8.282231796065215e-06, | |
| "loss": 0.1239, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.7105263157894737, | |
| "grad_norm": 1.735863282573467, | |
| "learning_rate": 8.194234028259806e-06, | |
| "loss": 0.1114, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.7236842105263158, | |
| "grad_norm": 1.062538506538223, | |
| "learning_rate": 8.104531868997858e-06, | |
| "loss": 0.0833, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 1.4057184361489892, | |
| "learning_rate": 8.013173181896283e-06, | |
| "loss": 0.1348, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.0217071440957173, | |
| "learning_rate": 7.920206714468383e-06, | |
| "loss": 0.0869, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7631578947368421, | |
| "grad_norm": 0.9647935123156808, | |
| "learning_rate": 7.82568207211296e-06, | |
| "loss": 0.0774, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7763157894736842, | |
| "grad_norm": 1.2119765768642377, | |
| "learning_rate": 7.729649691645673e-06, | |
| "loss": 0.0867, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.684404887958961, | |
| "learning_rate": 7.63216081438678e-06, | |
| "loss": 0.077, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8026315789473685, | |
| "grad_norm": 1.202460612915356, | |
| "learning_rate": 7.533267458819597e-06, | |
| "loss": 0.0823, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.8157894736842105, | |
| "grad_norm": 1.307993207575058, | |
| "learning_rate": 7.4330223928342814e-06, | |
| "loss": 0.0902, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8289473684210527, | |
| "grad_norm": 1.6640855566750032, | |
| "learning_rate": 7.33147910557174e-06, | |
| "loss": 0.141, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 1.3895288949997349, | |
| "learning_rate": 7.2286917788826926e-06, | |
| "loss": 0.0992, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.8552631578947368, | |
| "grad_norm": 0.86233230188121, | |
| "learning_rate": 7.124715258417111e-06, | |
| "loss": 0.0663, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.868421052631579, | |
| "grad_norm": 1.3734677592094817, | |
| "learning_rate": 7.019605024359475e-06, | |
| "loss": 0.0848, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.881578947368421, | |
| "grad_norm": 1.2016915391705267, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.1137, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 1.1962233631272159, | |
| "learning_rate": 6.806208330935766e-06, | |
| "loss": 0.0866, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.9078947368421053, | |
| "grad_norm": 1.3774507325980982, | |
| "learning_rate": 6.698035736583307e-06, | |
| "loss": 0.0975, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.9210526315789473, | |
| "grad_norm": 1.0910714190292217, | |
| "learning_rate": 6.588957097909509e-06, | |
| "loss": 0.0804, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9342105263157895, | |
| "grad_norm": 0.7559368152288192, | |
| "learning_rate": 6.4790306175063535e-06, | |
| "loss": 0.0748, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 1.320278904610911, | |
| "learning_rate": 6.368314950360416e-06, | |
| "loss": 0.1092, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9605263157894737, | |
| "grad_norm": 0.9978117274759357, | |
| "learning_rate": 6.2568691725555144e-06, | |
| "loss": 0.0669, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9736842105263158, | |
| "grad_norm": 1.184512537053634, | |
| "learning_rate": 6.144752749750671e-06, | |
| "loss": 0.0791, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.9868421052631579, | |
| "grad_norm": 1.3888364419849502, | |
| "learning_rate": 6.0320255054501985e-06, | |
| "loss": 0.0791, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.0696792385978997, | |
| "learning_rate": 5.918747589082853e-06, | |
| "loss": 0.1352, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.09702859073877335, | |
| "eval_runtime": 143.4285, | |
| "eval_samples_per_second": 35.593, | |
| "eval_steps_per_second": 1.116, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.013157894736842, | |
| "grad_norm": 1.130795788902584, | |
| "learning_rate": 5.804979443907065e-06, | |
| "loss": 0.0565, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.0263157894736843, | |
| "grad_norm": 0.9742675703939438, | |
| "learning_rate": 5.690781774759412e-06, | |
| "loss": 0.0826, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0394736842105263, | |
| "grad_norm": 1.4529332894640457, | |
| "learning_rate": 5.576215515663489e-06, | |
| "loss": 0.109, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.8745980190537407, | |
| "learning_rate": 5.46134179731651e-06, | |
| "loss": 0.0815, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0657894736842106, | |
| "grad_norm": 0.965352409023328, | |
| "learning_rate": 5.346221914470959e-06, | |
| "loss": 0.0753, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0789473684210527, | |
| "grad_norm": 0.9281912350813449, | |
| "learning_rate": 5.230917293228699e-06, | |
| "loss": 0.0604, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0921052631578947, | |
| "grad_norm": 0.7117517424840752, | |
| "learning_rate": 5.115489458265006e-06, | |
| "loss": 0.0548, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.8174226361772433, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0798, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.118421052631579, | |
| "grad_norm": 1.091499890906096, | |
| "learning_rate": 4.8845105417349955e-06, | |
| "loss": 0.0669, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.131578947368421, | |
| "grad_norm": 0.8489893677105097, | |
| "learning_rate": 4.7690827067713035e-06, | |
| "loss": 0.0799, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1447368421052633, | |
| "grad_norm": 1.0103580405572923, | |
| "learning_rate": 4.653778085529043e-06, | |
| "loss": 0.0697, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.8915773944210619, | |
| "learning_rate": 4.53865820268349e-06, | |
| "loss": 0.0786, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1710526315789473, | |
| "grad_norm": 0.9795278903306394, | |
| "learning_rate": 4.4237844843365126e-06, | |
| "loss": 0.0948, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.1842105263157894, | |
| "grad_norm": 0.6718237305799917, | |
| "learning_rate": 4.309218225240591e-06, | |
| "loss": 0.0751, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1973684210526316, | |
| "grad_norm": 0.9809953018950844, | |
| "learning_rate": 4.195020556092935e-06, | |
| "loss": 0.0717, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 1.3739884608587827, | |
| "learning_rate": 4.081252410917148e-06, | |
| "loss": 0.0834, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.2236842105263157, | |
| "grad_norm": 0.7354951934416454, | |
| "learning_rate": 3.967974494549803e-06, | |
| "loss": 0.0497, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.236842105263158, | |
| "grad_norm": 0.9321611181418619, | |
| "learning_rate": 3.855247250249331e-06, | |
| "loss": 0.1091, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.8153028496714768, | |
| "learning_rate": 3.743130827444487e-06, | |
| "loss": 0.0719, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 1.7763781332034168, | |
| "learning_rate": 3.6316850496395863e-06, | |
| "loss": 0.0995, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2763157894736843, | |
| "grad_norm": 1.0403921146171855, | |
| "learning_rate": 3.5209693824936486e-06, | |
| "loss": 0.1035, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2894736842105263, | |
| "grad_norm": 0.9375966336686881, | |
| "learning_rate": 3.4110429020904924e-06, | |
| "loss": 0.0854, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.3026315789473684, | |
| "grad_norm": 0.8044967231814447, | |
| "learning_rate": 3.301964263416693e-06, | |
| "loss": 0.0545, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 1.057615646119616, | |
| "learning_rate": 3.1937916690642356e-06, | |
| "loss": 0.0721, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.3289473684210527, | |
| "grad_norm": 0.8460218568489392, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.0756, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3421052631578947, | |
| "grad_norm": 0.7973010176418315, | |
| "learning_rate": 2.980394975640526e-06, | |
| "loss": 0.0701, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3552631578947367, | |
| "grad_norm": 0.43625526231871226, | |
| "learning_rate": 2.8752847415828923e-06, | |
| "loss": 0.0439, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.7846073292481729, | |
| "learning_rate": 2.771308221117309e-06, | |
| "loss": 0.068, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.381578947368421, | |
| "grad_norm": 0.7894886291152677, | |
| "learning_rate": 2.668520894428259e-06, | |
| "loss": 0.0521, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3947368421052633, | |
| "grad_norm": 0.8898713056015292, | |
| "learning_rate": 2.5669776071657194e-06, | |
| "loss": 0.0832, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.4078947368421053, | |
| "grad_norm": 0.7239448528037634, | |
| "learning_rate": 2.466732541180404e-06, | |
| "loss": 0.0651, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 1.1924698653865555, | |
| "learning_rate": 2.3678391856132203e-06, | |
| "loss": 0.0871, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.4342105263157894, | |
| "grad_norm": 0.796957419192979, | |
| "learning_rate": 2.2703503083543288e-06, | |
| "loss": 0.0632, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.4473684210526316, | |
| "grad_norm": 0.8317972188965687, | |
| "learning_rate": 2.174317927887041e-06, | |
| "loss": 0.0627, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4605263157894737, | |
| "grad_norm": 0.6383766307808258, | |
| "learning_rate": 2.0797932855316183e-06, | |
| "loss": 0.0685, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.8167329309674938, | |
| "learning_rate": 1.9868268181037186e-06, | |
| "loss": 0.0674, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.486842105263158, | |
| "grad_norm": 0.7471966235227357, | |
| "learning_rate": 1.8954681310021434e-06, | |
| "loss": 0.0729, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.6587846660979715, | |
| "learning_rate": 1.8057659717401948e-06, | |
| "loss": 0.0481, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.513157894736842, | |
| "grad_norm": 0.8623712101197185, | |
| "learning_rate": 1.7177682039347875e-06, | |
| "loss": 0.0751, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.7600610777742871, | |
| "learning_rate": 1.6315217817672142e-06, | |
| "loss": 0.0754, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.5394736842105263, | |
| "grad_norm": 0.7028388956696653, | |
| "learning_rate": 1.5470727249291423e-06, | |
| "loss": 0.0591, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5526315789473686, | |
| "grad_norm": 0.9080662531327958, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.0665, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5657894736842106, | |
| "grad_norm": 0.8002081473664321, | |
| "learning_rate": 1.383745966739652e-06, | |
| "loss": 0.0567, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.8070376846944848, | |
| "learning_rate": 1.3049554138967052e-06, | |
| "loss": 0.074, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5921052631578947, | |
| "grad_norm": 0.7133857988769062, | |
| "learning_rate": 1.2281364768991804e-06, | |
| "loss": 0.082, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.6052631578947367, | |
| "grad_norm": 0.7063230781754961, | |
| "learning_rate": 1.1533301450856054e-06, | |
| "loss": 0.0793, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.618421052631579, | |
| "grad_norm": 0.5118303753159236, | |
| "learning_rate": 1.0805763339010329e-06, | |
| "loss": 0.0606, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.6255028097581671, | |
| "learning_rate": 1.0099138635988026e-06, | |
| "loss": 0.0624, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6447368421052633, | |
| "grad_norm": 1.2565545443199133, | |
| "learning_rate": 9.41380438526694e-07, | |
| "loss": 0.0834, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6578947368421053, | |
| "grad_norm": 0.6949398501709794, | |
| "learning_rate": 8.750126270084891e-07, | |
| "loss": 0.0554, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6710526315789473, | |
| "grad_norm": 0.6169370398639232, | |
| "learning_rate": 8.108458418317089e-07, | |
| "loss": 0.0756, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.5783403647508548, | |
| "learning_rate": 7.489143213519301e-07, | |
| "loss": 0.0484, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6973684210526314, | |
| "grad_norm": 0.613206478877195, | |
| "learning_rate": 6.892511112237472e-07, | |
| "loss": 0.0674, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.7105263157894737, | |
| "grad_norm": 0.7610207582376373, | |
| "learning_rate": 6.318880467681527e-07, | |
| "loss": 0.0802, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.723684210526316, | |
| "grad_norm": 0.6484932637313764, | |
| "learning_rate": 5.768557359857241e-07, | |
| "loss": 0.0799, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.801045963046033, | |
| "learning_rate": 5.241835432246888e-07, | |
| "loss": 0.0631, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.5006567755599417, | |
| "learning_rate": 4.738995735125895e-07, | |
| "loss": 0.0602, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.763157894736842, | |
| "grad_norm": 0.7321825355306667, | |
| "learning_rate": 4.2603065755989493e-07, | |
| "loss": 0.0574, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.776315789473684, | |
| "grad_norm": 0.5455281345073466, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.0585, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.6941675384816356, | |
| "learning_rate": 3.3763885297822153e-07, | |
| "loss": 0.0593, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.8026315789473686, | |
| "grad_norm": 0.607277109667664, | |
| "learning_rate": 2.9716312878216194e-07, | |
| "loss": 0.0596, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.8157894736842106, | |
| "grad_norm": 0.5526534850437881, | |
| "learning_rate": 2.5919676204517073e-07, | |
| "loss": 0.0574, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.8289473684210527, | |
| "grad_norm": 0.7027887496493381, | |
| "learning_rate": 2.237600110046001e-07, | |
| "loss": 0.0883, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.5658475866263035, | |
| "learning_rate": 1.908717841359048e-07, | |
| "loss": 0.0593, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8552631578947367, | |
| "grad_norm": 0.6440458312241749, | |
| "learning_rate": 1.6054963006338742e-07, | |
| "loss": 0.0645, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.868421052631579, | |
| "grad_norm": 0.8251777255090132, | |
| "learning_rate": 1.328097281965357e-07, | |
| "loss": 0.091, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.881578947368421, | |
| "grad_norm": 0.584859727969167, | |
| "learning_rate": 1.0766688009695548e-07, | |
| "loss": 0.0624, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.6478070234261503, | |
| "learning_rate": 8.513450158049109e-08, | |
| "loss": 0.0662, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.9078947368421053, | |
| "grad_norm": 0.5481010295457636, | |
| "learning_rate": 6.522461555877213e-08, | |
| "loss": 0.0717, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.9210526315789473, | |
| "grad_norm": 0.5866004853500736, | |
| "learning_rate": 4.794784562397459e-08, | |
| "loss": 0.0729, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.9342105263157894, | |
| "grad_norm": 0.7549384445552164, | |
| "learning_rate": 3.3313410380250157e-08, | |
| "loss": 0.0936, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.8887389895072535, | |
| "learning_rate": 2.1329118524827662e-08, | |
| "loss": 0.1033, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9605263157894737, | |
| "grad_norm": 0.6459019124164774, | |
| "learning_rate": 1.200136468141544e-08, | |
| "loss": 0.0951, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.973684210526316, | |
| "grad_norm": 0.5993524993653697, | |
| "learning_rate": 5.3351259881379016e-09, | |
| "loss": 0.063, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.986842105263158, | |
| "grad_norm": 0.6183094577100168, | |
| "learning_rate": 1.3339594418138036e-09, | |
| "loss": 0.0715, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.47166327201031033, | |
| "learning_rate": 0.0, | |
| "loss": 0.051, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.08257210999727249, | |
| "eval_runtime": 143.687, | |
| "eval_samples_per_second": 35.529, | |
| "eval_steps_per_second": 1.114, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 152, | |
| "total_flos": 4.516827336435302e+16, | |
| "train_loss": 0.09894711428665016, | |
| "train_runtime": 2051.5561, | |
| "train_samples_per_second": 9.454, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 152, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 4.516827336435302e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |