| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.5590297095805052, |
| "eval_steps": 100, |
| "global_step": 2000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 0.744562653824687, |
| "epoch": 0.017803493935684877, |
| "grad_norm": 8.4375, |
| "learning_rate": 1.8e-07, |
| "loss": 1.1008, |
| "mean_token_accuracy": 0.7502190172672272, |
| "num_tokens": 259048.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 0.723362203873694, |
| "epoch": 0.035606987871369754, |
| "grad_norm": 8.6875, |
| "learning_rate": 3.8e-07, |
| "loss": 1.0517, |
| "mean_token_accuracy": 0.7539289794862271, |
| "num_tokens": 532019.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 0.7459586564451456, |
| "epoch": 0.053410481807054634, |
| "grad_norm": 8.375, |
| "learning_rate": 5.800000000000001e-07, |
| "loss": 1.0774, |
| "mean_token_accuracy": 0.7505976643413306, |
| "num_tokens": 807220.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 0.7810037085786462, |
| "epoch": 0.07121397574273951, |
| "grad_norm": 7.71875, |
| "learning_rate": 7.8e-07, |
| "loss": 1.0764, |
| "mean_token_accuracy": 0.7482377212494612, |
| "num_tokens": 1068429.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 0.7945824306458235, |
| "epoch": 0.08901746967842439, |
| "grad_norm": 5.3125, |
| "learning_rate": 9.800000000000001e-07, |
| "loss": 1.0429, |
| "mean_token_accuracy": 0.7540664587169885, |
| "num_tokens": 1336698.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 0.8338877715170383, |
| "epoch": 0.10682096361410927, |
| "grad_norm": 4.90625, |
| "learning_rate": 1.1800000000000001e-06, |
| "loss": 1.0373, |
| "mean_token_accuracy": 0.7545978050678969, |
| "num_tokens": 1608574.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 0.9112611550837755, |
| "epoch": 0.12462445754979415, |
| "grad_norm": 3.359375, |
| "learning_rate": 1.3800000000000001e-06, |
| "loss": 1.0542, |
| "mean_token_accuracy": 0.7494721282273531, |
| "num_tokens": 1872747.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 0.9293210502713919, |
| "epoch": 0.14242795148547902, |
| "grad_norm": 3.234375, |
| "learning_rate": 1.5800000000000001e-06, |
| "loss": 1.0293, |
| "mean_token_accuracy": 0.7546770930290222, |
| "num_tokens": 2133173.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 0.9038813605904579, |
| "epoch": 0.1602314454211639, |
| "grad_norm": 2.953125, |
| "learning_rate": 1.7800000000000001e-06, |
| "loss": 0.9937, |
| "mean_token_accuracy": 0.7605839747935533, |
| "num_tokens": 2406145.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 0.9136581432074309, |
| "epoch": 0.17803493935684878, |
| "grad_norm": 3.15625, |
| "learning_rate": 1.98e-06, |
| "loss": 1.0088, |
| "mean_token_accuracy": 0.757991149649024, |
| "num_tokens": 2663709.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.17803493935684878, |
| "eval_biology_entropy": 1.1452245082855224, |
| "eval_biology_loss": 1.2803255319595337, |
| "eval_biology_mean_token_accuracy": 0.7091106538772582, |
| "eval_biology_num_tokens": 2663709.0, |
| "eval_biology_runtime": 27.5039, |
| "eval_biology_samples_per_second": 18.179, |
| "eval_biology_steps_per_second": 4.545, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.17803493935684878, |
| "eval_chemistry_entropy": 0.8995756044387817, |
| "eval_chemistry_loss": 1.014352798461914, |
| "eval_chemistry_mean_token_accuracy": 0.7613508405685425, |
| "eval_chemistry_num_tokens": 2663709.0, |
| "eval_chemistry_runtime": 34.6715, |
| "eval_chemistry_samples_per_second": 14.421, |
| "eval_chemistry_steps_per_second": 3.605, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.17803493935684878, |
| "eval_math_entropy": 0.9318298659324646, |
| "eval_math_loss": 1.267881989479065, |
| "eval_math_mean_token_accuracy": 0.7267651476860046, |
| "eval_math_num_tokens": 2663709.0, |
| "eval_math_runtime": 35.7096, |
| "eval_math_samples_per_second": 14.002, |
| "eval_math_steps_per_second": 3.5, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.17803493935684878, |
| "eval_physics_entropy": 0.9398559331893921, |
| "eval_physics_loss": 1.0559574365615845, |
| "eval_physics_mean_token_accuracy": 0.75667662525177, |
| "eval_physics_num_tokens": 2663709.0, |
| "eval_physics_runtime": 41.3576, |
| "eval_physics_samples_per_second": 12.09, |
| "eval_physics_steps_per_second": 3.022, |
| "step": 100 |
| }, |
| { |
| "entropy": 0.8964114494621753, |
| "epoch": 0.19583843329253367, |
| "grad_norm": 2.921875, |
| "learning_rate": 2.1800000000000003e-06, |
| "loss": 0.9784, |
| "mean_token_accuracy": 0.763974778726697, |
| "num_tokens": 2934980.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 0.8946480920538307, |
| "epoch": 0.21364192722821854, |
| "grad_norm": 2.6875, |
| "learning_rate": 2.38e-06, |
| "loss": 0.9663, |
| "mean_token_accuracy": 0.7627720884978771, |
| "num_tokens": 3208702.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 0.8829301230609417, |
| "epoch": 0.2314454211639034, |
| "grad_norm": 2.859375, |
| "learning_rate": 2.5800000000000003e-06, |
| "loss": 0.9684, |
| "mean_token_accuracy": 0.7646027702838183, |
| "num_tokens": 3480119.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 0.891739396750927, |
| "epoch": 0.2492489150995883, |
| "grad_norm": 3.265625, |
| "learning_rate": 2.7800000000000005e-06, |
| "loss": 0.9634, |
| "mean_token_accuracy": 0.7636573657393455, |
| "num_tokens": 3746392.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 0.8886399254202842, |
| "epoch": 0.26705240903527316, |
| "grad_norm": 3.171875, |
| "learning_rate": 2.9800000000000003e-06, |
| "loss": 0.9551, |
| "mean_token_accuracy": 0.7625629913061858, |
| "num_tokens": 4019487.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 0.8688891407102346, |
| "epoch": 0.28485590297095803, |
| "grad_norm": 2.921875, |
| "learning_rate": 3.1800000000000005e-06, |
| "loss": 0.9428, |
| "mean_token_accuracy": 0.7663193177431822, |
| "num_tokens": 4298100.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 0.8776854418218136, |
| "epoch": 0.30265939690664295, |
| "grad_norm": 2.671875, |
| "learning_rate": 3.3800000000000007e-06, |
| "loss": 0.926, |
| "mean_token_accuracy": 0.7682191282510757, |
| "num_tokens": 4568116.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 0.8811742646619678, |
| "epoch": 0.3204628908423278, |
| "grad_norm": 2.90625, |
| "learning_rate": 3.58e-06, |
| "loss": 0.9262, |
| "mean_token_accuracy": 0.7651723474264145, |
| "num_tokens": 4831334.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 0.8657637692987918, |
| "epoch": 0.3382663847780127, |
| "grad_norm": 2.28125, |
| "learning_rate": 3.7800000000000002e-06, |
| "loss": 0.902, |
| "mean_token_accuracy": 0.7699528813362122, |
| "num_tokens": 5103301.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 0.8889979477971792, |
| "epoch": 0.35606987871369755, |
| "grad_norm": 2.78125, |
| "learning_rate": 3.980000000000001e-06, |
| "loss": 0.921, |
| "mean_token_accuracy": 0.7651176899671555, |
| "num_tokens": 5362915.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.35606987871369755, |
| "eval_biology_entropy": 1.1483394474983215, |
| "eval_biology_loss": 1.2182775735855103, |
| "eval_biology_mean_token_accuracy": 0.7090820508003235, |
| "eval_biology_num_tokens": 5362915.0, |
| "eval_biology_runtime": 27.4751, |
| "eval_biology_samples_per_second": 18.198, |
| "eval_biology_steps_per_second": 4.55, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.35606987871369755, |
| "eval_chemistry_entropy": 0.8670397129058838, |
| "eval_chemistry_loss": 0.9158776998519897, |
| "eval_chemistry_mean_token_accuracy": 0.7710415840148925, |
| "eval_chemistry_num_tokens": 5362915.0, |
| "eval_chemistry_runtime": 34.6851, |
| "eval_chemistry_samples_per_second": 14.415, |
| "eval_chemistry_steps_per_second": 3.604, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.35606987871369755, |
| "eval_math_entropy": 0.8976822423934937, |
| "eval_math_loss": 1.1883426904678345, |
| "eval_math_mean_token_accuracy": 0.7329299259185791, |
| "eval_math_num_tokens": 5362915.0, |
| "eval_math_runtime": 35.713, |
| "eval_math_samples_per_second": 14.001, |
| "eval_math_steps_per_second": 3.5, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.35606987871369755, |
| "eval_physics_entropy": 0.9075017051696778, |
| "eval_physics_loss": 0.9745987057685852, |
| "eval_physics_mean_token_accuracy": 0.7644435825347901, |
| "eval_physics_num_tokens": 5362915.0, |
| "eval_physics_runtime": 41.3483, |
| "eval_physics_samples_per_second": 12.092, |
| "eval_physics_steps_per_second": 3.023, |
| "step": 200 |
| }, |
| { |
| "entropy": 0.8561842858791351, |
| "epoch": 0.3738733726493824, |
| "grad_norm": 2.53125, |
| "learning_rate": 4.18e-06, |
| "loss": 0.8773, |
| "mean_token_accuracy": 0.7733380310237408, |
| "num_tokens": 5625455.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 0.8662637937813997, |
| "epoch": 0.39167686658506734, |
| "grad_norm": 2.484375, |
| "learning_rate": 4.38e-06, |
| "loss": 0.8794, |
| "mean_token_accuracy": 0.7717520732432603, |
| "num_tokens": 5890806.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 0.8469413101673127, |
| "epoch": 0.4094803605207522, |
| "grad_norm": 2.828125, |
| "learning_rate": 4.58e-06, |
| "loss": 0.8482, |
| "mean_token_accuracy": 0.7760931551456451, |
| "num_tokens": 6158751.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 0.8607879716902971, |
| "epoch": 0.4272838544564371, |
| "grad_norm": 2.75, |
| "learning_rate": 4.78e-06, |
| "loss": 0.8507, |
| "mean_token_accuracy": 0.7743628785014153, |
| "num_tokens": 6423747.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 0.8727764803916216, |
| "epoch": 0.44508734839212194, |
| "grad_norm": 2.546875, |
| "learning_rate": 4.980000000000001e-06, |
| "loss": 0.8631, |
| "mean_token_accuracy": 0.7715923175215721, |
| "num_tokens": 6684517.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 0.8464060451835393, |
| "epoch": 0.4628908423278068, |
| "grad_norm": 2.71875, |
| "learning_rate": 5.18e-06, |
| "loss": 0.8341, |
| "mean_token_accuracy": 0.7778113219887018, |
| "num_tokens": 6945659.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 0.8537958716973663, |
| "epoch": 0.48069433626349173, |
| "grad_norm": 2.5625, |
| "learning_rate": 5.380000000000001e-06, |
| "loss": 0.8447, |
| "mean_token_accuracy": 0.776023967564106, |
| "num_tokens": 7217797.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 0.8301472842693329, |
| "epoch": 0.4984978301991766, |
| "grad_norm": 2.59375, |
| "learning_rate": 5.580000000000001e-06, |
| "loss": 0.8049, |
| "mean_token_accuracy": 0.7804847627878189, |
| "num_tokens": 7498182.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 0.8598408252000809, |
| "epoch": 0.5163013241348615, |
| "grad_norm": 2.421875, |
| "learning_rate": 5.78e-06, |
| "loss": 0.8373, |
| "mean_token_accuracy": 0.7750447608530522, |
| "num_tokens": 7761686.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 0.8562261991202831, |
| "epoch": 0.5341048180705463, |
| "grad_norm": 2.484375, |
| "learning_rate": 5.98e-06, |
| "loss": 0.8303, |
| "mean_token_accuracy": 0.776263014599681, |
| "num_tokens": 8026923.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5341048180705463, |
| "eval_biology_entropy": 1.1636152486801148, |
| "eval_biology_loss": 1.1685189008712769, |
| "eval_biology_mean_token_accuracy": 0.7092800517082214, |
| "eval_biology_num_tokens": 8026923.0, |
| "eval_biology_runtime": 27.4927, |
| "eval_biology_samples_per_second": 18.187, |
| "eval_biology_steps_per_second": 4.547, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5341048180705463, |
| "eval_chemistry_entropy": 0.8379480366706848, |
| "eval_chemistry_loss": 0.8368871212005615, |
| "eval_chemistry_mean_token_accuracy": 0.7790313405990601, |
| "eval_chemistry_num_tokens": 8026923.0, |
| "eval_chemistry_runtime": 34.7081, |
| "eval_chemistry_samples_per_second": 14.406, |
| "eval_chemistry_steps_per_second": 3.601, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5341048180705463, |
| "eval_math_entropy": 0.8852076215744018, |
| "eval_math_loss": 1.1242812871932983, |
| "eval_math_mean_token_accuracy": 0.7380555834770203, |
| "eval_math_num_tokens": 8026923.0, |
| "eval_math_runtime": 35.7116, |
| "eval_math_samples_per_second": 14.001, |
| "eval_math_steps_per_second": 3.5, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5341048180705463, |
| "eval_physics_entropy": 0.8859062085151672, |
| "eval_physics_loss": 0.9132099747657776, |
| "eval_physics_mean_token_accuracy": 0.769776885509491, |
| "eval_physics_num_tokens": 8026923.0, |
| "eval_physics_runtime": 41.3565, |
| "eval_physics_samples_per_second": 12.09, |
| "eval_physics_steps_per_second": 3.022, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.8374487660825253, |
| "epoch": 0.5519083120062312, |
| "grad_norm": 2.640625, |
| "learning_rate": 6.18e-06, |
| "loss": 0.816, |
| "mean_token_accuracy": 0.7786551136523485, |
| "num_tokens": 8301506.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 0.8338102400302887, |
| "epoch": 0.5697118059419161, |
| "grad_norm": 2.234375, |
| "learning_rate": 6.380000000000001e-06, |
| "loss": 0.814, |
| "mean_token_accuracy": 0.7796321768313647, |
| "num_tokens": 8559126.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 0.8385110892355442, |
| "epoch": 0.587515299877601, |
| "grad_norm": 2.515625, |
| "learning_rate": 6.5800000000000005e-06, |
| "loss": 0.8112, |
| "mean_token_accuracy": 0.78204450532794, |
| "num_tokens": 8821088.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 0.829721986502409, |
| "epoch": 0.6053187938132859, |
| "grad_norm": 2.453125, |
| "learning_rate": 6.780000000000001e-06, |
| "loss": 0.8, |
| "mean_token_accuracy": 0.7853520669043064, |
| "num_tokens": 9089866.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 0.8165910333395004, |
| "epoch": 0.6231222877489707, |
| "grad_norm": 2.375, |
| "learning_rate": 6.98e-06, |
| "loss": 0.8006, |
| "mean_token_accuracy": 0.7839661207050085, |
| "num_tokens": 9358866.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 0.8157743275165558, |
| "epoch": 0.6409257816846556, |
| "grad_norm": 2.03125, |
| "learning_rate": 7.180000000000001e-06, |
| "loss": 0.7816, |
| "mean_token_accuracy": 0.7866707690060138, |
| "num_tokens": 9629977.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 0.8082325670868158, |
| "epoch": 0.6587292756203404, |
| "grad_norm": 2.46875, |
| "learning_rate": 7.3800000000000005e-06, |
| "loss": 0.7762, |
| "mean_token_accuracy": 0.7889597825706005, |
| "num_tokens": 9897757.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 0.7897130275145173, |
| "epoch": 0.6765327695560254, |
| "grad_norm": 2.28125, |
| "learning_rate": 7.58e-06, |
| "loss": 0.7648, |
| "mean_token_accuracy": 0.7922043181955815, |
| "num_tokens": 10168367.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 0.7785528032109141, |
| "epoch": 0.6943362634917103, |
| "grad_norm": 2.203125, |
| "learning_rate": 7.78e-06, |
| "loss": 0.754, |
| "mean_token_accuracy": 0.7948459524661302, |
| "num_tokens": 10439039.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 0.7947212610393762, |
| "epoch": 0.7121397574273951, |
| "grad_norm": 2.390625, |
| "learning_rate": 7.980000000000002e-06, |
| "loss": 0.7651, |
| "mean_token_accuracy": 0.7918860759586096, |
| "num_tokens": 10709549.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7121397574273951, |
| "eval_biology_entropy": 1.170575973033905, |
| "eval_biology_loss": 1.1608192920684814, |
| "eval_biology_mean_token_accuracy": 0.7121581416130066, |
| "eval_biology_num_tokens": 10709549.0, |
| "eval_biology_runtime": 27.464, |
| "eval_biology_samples_per_second": 18.206, |
| "eval_biology_steps_per_second": 4.551, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7121397574273951, |
| "eval_chemistry_entropy": 0.8082829179763794, |
| "eval_chemistry_loss": 0.7949892282485962, |
| "eval_chemistry_mean_token_accuracy": 0.7886457958221436, |
| "eval_chemistry_num_tokens": 10709549.0, |
| "eval_chemistry_runtime": 34.6794, |
| "eval_chemistry_samples_per_second": 14.418, |
| "eval_chemistry_steps_per_second": 3.604, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7121397574273951, |
| "eval_math_entropy": 0.8507072486877442, |
| "eval_math_loss": 1.0929126739501953, |
| "eval_math_mean_token_accuracy": 0.7454545521736144, |
| "eval_math_num_tokens": 10709549.0, |
| "eval_math_runtime": 35.7089, |
| "eval_math_samples_per_second": 14.002, |
| "eval_math_steps_per_second": 3.501, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7121397574273951, |
| "eval_physics_entropy": 0.853618821144104, |
| "eval_physics_loss": 0.8825188279151917, |
| "eval_physics_mean_token_accuracy": 0.77662775182724, |
| "eval_physics_num_tokens": 10709549.0, |
| "eval_physics_runtime": 41.4089, |
| "eval_physics_samples_per_second": 12.075, |
| "eval_physics_steps_per_second": 3.019, |
| "step": 400 |
| }, |
| { |
| "entropy": 0.7872084211558104, |
| "epoch": 0.72994325136308, |
| "grad_norm": 2.28125, |
| "learning_rate": 8.18e-06, |
| "loss": 0.7584, |
| "mean_token_accuracy": 0.7928647272288799, |
| "num_tokens": 10976011.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 0.7894711047410965, |
| "epoch": 0.7477467452987648, |
| "grad_norm": 2.59375, |
| "learning_rate": 8.380000000000001e-06, |
| "loss": 0.7551, |
| "mean_token_accuracy": 0.7918795708566904, |
| "num_tokens": 11245251.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 0.800145055539906, |
| "epoch": 0.7655502392344498, |
| "grad_norm": 2.203125, |
| "learning_rate": 8.580000000000001e-06, |
| "loss": 0.7678, |
| "mean_token_accuracy": 0.788029194995761, |
| "num_tokens": 11518514.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 0.8004930332303047, |
| "epoch": 0.7833537331701347, |
| "grad_norm": 2.40625, |
| "learning_rate": 8.78e-06, |
| "loss": 0.7698, |
| "mean_token_accuracy": 0.7889820158481597, |
| "num_tokens": 11787117.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 0.7886414948850871, |
| "epoch": 0.8011572271058195, |
| "grad_norm": 2.578125, |
| "learning_rate": 8.98e-06, |
| "loss": 0.7561, |
| "mean_token_accuracy": 0.7918269295245409, |
| "num_tokens": 12050993.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 0.7945233354344964, |
| "epoch": 0.8189607210415044, |
| "grad_norm": 2.3125, |
| "learning_rate": 9.180000000000002e-06, |
| "loss": 0.7586, |
| "mean_token_accuracy": 0.7906661573797464, |
| "num_tokens": 12311688.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 0.7684864999726415, |
| "epoch": 0.8367642149771892, |
| "grad_norm": 2.109375, |
| "learning_rate": 9.38e-06, |
| "loss": 0.7375, |
| "mean_token_accuracy": 0.7950909677892923, |
| "num_tokens": 12598408.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 0.7694930030032993, |
| "epoch": 0.8545677089128741, |
| "grad_norm": 2.171875, |
| "learning_rate": 9.58e-06, |
| "loss": 0.7446, |
| "mean_token_accuracy": 0.7951429858803749, |
| "num_tokens": 12869642.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 0.7591588348150253, |
| "epoch": 0.8723712028485591, |
| "grad_norm": 2.234375, |
| "learning_rate": 9.780000000000001e-06, |
| "loss": 0.7266, |
| "mean_token_accuracy": 0.7976758178323508, |
| "num_tokens": 13139555.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 0.7894993636757135, |
| "epoch": 0.8901746967842439, |
| "grad_norm": 2.265625, |
| "learning_rate": 9.980000000000001e-06, |
| "loss": 0.7593, |
| "mean_token_accuracy": 0.7908468171954155, |
| "num_tokens": 13396545.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8901746967842439, |
| "eval_biology_entropy": 1.1589476070404052, |
| "eval_biology_loss": 1.1534974575042725, |
| "eval_biology_mean_token_accuracy": 0.7124825038909912, |
| "eval_biology_num_tokens": 13396545.0, |
| "eval_biology_runtime": 27.4926, |
| "eval_biology_samples_per_second": 18.187, |
| "eval_biology_steps_per_second": 4.547, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8901746967842439, |
| "eval_chemistry_entropy": 0.781173488855362, |
| "eval_chemistry_loss": 0.7634457945823669, |
| "eval_chemistry_mean_token_accuracy": 0.7938974785804749, |
| "eval_chemistry_num_tokens": 13396545.0, |
| "eval_chemistry_runtime": 34.3839, |
| "eval_chemistry_samples_per_second": 14.542, |
| "eval_chemistry_steps_per_second": 3.635, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8901746967842439, |
| "eval_math_entropy": 0.8440827627182007, |
| "eval_math_loss": 1.07049560546875, |
| "eval_math_mean_token_accuracy": 0.7487871046066285, |
| "eval_math_num_tokens": 13396545.0, |
| "eval_math_runtime": 35.8298, |
| "eval_math_samples_per_second": 13.955, |
| "eval_math_steps_per_second": 3.489, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8901746967842439, |
| "eval_physics_entropy": 0.8325447106361389, |
| "eval_physics_loss": 0.8577290773391724, |
| "eval_physics_mean_token_accuracy": 0.7805949940681457, |
| "eval_physics_num_tokens": 13396545.0, |
| "eval_physics_runtime": 41.2595, |
| "eval_physics_samples_per_second": 12.118, |
| "eval_physics_steps_per_second": 3.03, |
| "step": 500 |
| }, |
| { |
| "entropy": 0.757487927749753, |
| "epoch": 0.9079781907199288, |
| "grad_norm": 2.703125, |
| "learning_rate": 1.018e-05, |
| "loss": 0.73, |
| "mean_token_accuracy": 0.7988032333552837, |
| "num_tokens": 13667857.0, |
| "step": 510 |
| }, |
| { |
| "entropy": 0.7804714627563953, |
| "epoch": 0.9257816846556136, |
| "grad_norm": 2.25, |
| "learning_rate": 1.038e-05, |
| "loss": 0.7412, |
| "mean_token_accuracy": 0.7939899254590272, |
| "num_tokens": 13932771.0, |
| "step": 520 |
| }, |
| { |
| "entropy": 0.7541740108281374, |
| "epoch": 0.9435851785912985, |
| "grad_norm": 2.125, |
| "learning_rate": 1.0580000000000002e-05, |
| "loss": 0.7236, |
| "mean_token_accuracy": 0.7992219775915146, |
| "num_tokens": 14194632.0, |
| "step": 530 |
| }, |
| { |
| "entropy": 0.7748582396656275, |
| "epoch": 0.9613886725269835, |
| "grad_norm": 2.078125, |
| "learning_rate": 1.0780000000000002e-05, |
| "loss": 0.7462, |
| "mean_token_accuracy": 0.7930019188672304, |
| "num_tokens": 14457650.0, |
| "step": 540 |
| }, |
| { |
| "entropy": 0.7507146148011088, |
| "epoch": 0.9791921664626683, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.0980000000000002e-05, |
| "loss": 0.7263, |
| "mean_token_accuracy": 0.798946413770318, |
| "num_tokens": 14724967.0, |
| "step": 550 |
| }, |
| { |
| "entropy": 0.7736786268651485, |
| "epoch": 0.9969956603983532, |
| "grad_norm": 2.265625, |
| "learning_rate": 1.1180000000000001e-05, |
| "loss": 0.745, |
| "mean_token_accuracy": 0.7954828392714262, |
| "num_tokens": 14993272.0, |
| "step": 560 |
| }, |
| { |
| "entropy": 0.712507297915797, |
| "epoch": 1.014242795148548, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.138e-05, |
| "loss": 0.6882, |
| "mean_token_accuracy": 0.806304710142074, |
| "num_tokens": 15254936.0, |
| "step": 570 |
| }, |
| { |
| "entropy": 0.7331789376214146, |
| "epoch": 1.0320462890842328, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.1580000000000001e-05, |
| "loss": 0.7013, |
| "mean_token_accuracy": 0.8014222860336304, |
| "num_tokens": 15515007.0, |
| "step": 580 |
| }, |
| { |
| "entropy": 0.709935688227415, |
| "epoch": 1.0498497830199176, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.178e-05, |
| "loss": 0.6839, |
| "mean_token_accuracy": 0.8082031298428773, |
| "num_tokens": 15794203.0, |
| "step": 590 |
| }, |
| { |
| "entropy": 0.7278657285496593, |
| "epoch": 1.0676532769556026, |
| "grad_norm": 2.109375, |
| "learning_rate": 1.198e-05, |
| "loss": 0.6985, |
| "mean_token_accuracy": 0.8037012912333011, |
| "num_tokens": 16063664.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.0676532769556026, |
| "eval_biology_entropy": 1.1254952058792114, |
| "eval_biology_loss": 1.1531407833099365, |
| "eval_biology_mean_token_accuracy": 0.7121509776115418, |
| "eval_biology_num_tokens": 16063664.0, |
| "eval_biology_runtime": 27.4617, |
| "eval_biology_samples_per_second": 18.207, |
| "eval_biology_steps_per_second": 4.552, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.0676532769556026, |
| "eval_chemistry_entropy": 0.7266788492202759, |
| "eval_chemistry_loss": 0.741141140460968, |
| "eval_chemistry_mean_token_accuracy": 0.7978090291023254, |
| "eval_chemistry_num_tokens": 16063664.0, |
| "eval_chemistry_runtime": 34.637, |
| "eval_chemistry_samples_per_second": 14.435, |
| "eval_chemistry_steps_per_second": 3.609, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.0676532769556026, |
| "eval_math_entropy": 0.7815078992843628, |
| "eval_math_loss": 1.0565334558486938, |
| "eval_math_mean_token_accuracy": 0.7529526872634887, |
| "eval_math_num_tokens": 16063664.0, |
| "eval_math_runtime": 35.7135, |
| "eval_math_samples_per_second": 14.0, |
| "eval_math_steps_per_second": 3.5, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.0676532769556026, |
| "eval_physics_entropy": 0.7784343235492707, |
| "eval_physics_loss": 0.8401147723197937, |
| "eval_physics_mean_token_accuracy": 0.7849010486602783, |
| "eval_physics_num_tokens": 16063664.0, |
| "eval_physics_runtime": 41.3341, |
| "eval_physics_samples_per_second": 12.097, |
| "eval_physics_steps_per_second": 3.024, |
| "step": 600 |
| }, |
| { |
| "entropy": 0.7066504143178463, |
| "epoch": 1.0854567708912874, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.218e-05, |
| "loss": 0.6839, |
| "mean_token_accuracy": 0.8087195225059987, |
| "num_tokens": 16338303.0, |
| "step": 610 |
| }, |
| { |
| "entropy": 0.7359958736225962, |
| "epoch": 1.1032602648269723, |
| "grad_norm": 2.125, |
| "learning_rate": 1.2380000000000002e-05, |
| "loss": 0.7001, |
| "mean_token_accuracy": 0.8021640948951244, |
| "num_tokens": 16606302.0, |
| "step": 620 |
| }, |
| { |
| "entropy": 0.7180680487304926, |
| "epoch": 1.121063758762657, |
| "grad_norm": 2.171875, |
| "learning_rate": 1.2580000000000002e-05, |
| "loss": 0.6851, |
| "mean_token_accuracy": 0.8039267532527447, |
| "num_tokens": 16877071.0, |
| "step": 630 |
| }, |
| { |
| "entropy": 0.7086811441928148, |
| "epoch": 1.138867252698342, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.2780000000000001e-05, |
| "loss": 0.6871, |
| "mean_token_accuracy": 0.8053031664341688, |
| "num_tokens": 17145775.0, |
| "step": 640 |
| }, |
| { |
| "entropy": 0.721453401632607, |
| "epoch": 1.156670746634027, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.2980000000000001e-05, |
| "loss": 0.6849, |
| "mean_token_accuracy": 0.8055552441626788, |
| "num_tokens": 17410775.0, |
| "step": 650 |
| }, |
| { |
| "entropy": 0.7386436846107245, |
| "epoch": 1.1744742405697117, |
| "grad_norm": 2.203125, |
| "learning_rate": 1.3180000000000001e-05, |
| "loss": 0.7053, |
| "mean_token_accuracy": 0.7988783929497003, |
| "num_tokens": 17677791.0, |
| "step": 660 |
| }, |
| { |
| "entropy": 0.7144078232347966, |
| "epoch": 1.1922777345053968, |
| "grad_norm": 2.359375, |
| "learning_rate": 1.3380000000000002e-05, |
| "loss": 0.6799, |
| "mean_token_accuracy": 0.805137374997139, |
| "num_tokens": 17949441.0, |
| "step": 670 |
| }, |
| { |
| "entropy": 0.7001266788691283, |
| "epoch": 1.2100812284410816, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.3580000000000002e-05, |
| "loss": 0.6704, |
| "mean_token_accuracy": 0.8073600005358458, |
| "num_tokens": 18217742.0, |
| "step": 680 |
| }, |
| { |
| "entropy": 0.7174897050485015, |
| "epoch": 1.2278847223767664, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.378e-05, |
| "loss": 0.6865, |
| "mean_token_accuracy": 0.8053013660013676, |
| "num_tokens": 18487769.0, |
| "step": 690 |
| }, |
| { |
| "entropy": 0.7023122169077396, |
| "epoch": 1.2456882163124514, |
| "grad_norm": 2.046875, |
| "learning_rate": 1.398e-05, |
| "loss": 0.6765, |
| "mean_token_accuracy": 0.8065393235534429, |
| "num_tokens": 18760296.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.2456882163124514, |
| "eval_biology_entropy": 1.1227134470939637, |
| "eval_biology_loss": 1.153878927230835, |
| "eval_biology_mean_token_accuracy": 0.7108333306312561, |
| "eval_biology_num_tokens": 18760296.0, |
| "eval_biology_runtime": 27.4519, |
| "eval_biology_samples_per_second": 18.214, |
| "eval_biology_steps_per_second": 4.553, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.2456882163124514, |
| "eval_chemistry_entropy": 0.7248331642150879, |
| "eval_chemistry_loss": 0.7243748307228088, |
| "eval_chemistry_mean_token_accuracy": 0.8007906107902527, |
| "eval_chemistry_num_tokens": 18760296.0, |
| "eval_chemistry_runtime": 34.6367, |
| "eval_chemistry_samples_per_second": 14.436, |
| "eval_chemistry_steps_per_second": 3.609, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.2456882163124514, |
| "eval_math_entropy": 0.7919260501861572, |
| "eval_math_loss": 1.0444875955581665, |
| "eval_math_mean_token_accuracy": 0.7544966869354248, |
| "eval_math_num_tokens": 18760296.0, |
| "eval_math_runtime": 35.6953, |
| "eval_math_samples_per_second": 14.007, |
| "eval_math_steps_per_second": 3.502, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.2456882163124514, |
| "eval_physics_entropy": 0.7797312350273132, |
| "eval_physics_loss": 0.8263967633247375, |
| "eval_physics_mean_token_accuracy": 0.7863735747337341, |
| "eval_physics_num_tokens": 18760296.0, |
| "eval_physics_runtime": 41.3137, |
| "eval_physics_samples_per_second": 12.103, |
| "eval_physics_steps_per_second": 3.026, |
| "step": 700 |
| }, |
| { |
| "entropy": 0.7069545309990645, |
| "epoch": 1.2634917102481362, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.418e-05, |
| "loss": 0.6753, |
| "mean_token_accuracy": 0.8080522440373897, |
| "num_tokens": 19035733.0, |
| "step": 710 |
| }, |
| { |
| "entropy": 0.7217417221516371, |
| "epoch": 1.281295204183821, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.4380000000000001e-05, |
| "loss": 0.6938, |
| "mean_token_accuracy": 0.8029382523149252, |
| "num_tokens": 19311432.0, |
| "step": 720 |
| }, |
| { |
| "entropy": 0.6935903675854206, |
| "epoch": 1.299098698119506, |
| "grad_norm": 1.9375, |
| "learning_rate": 1.4580000000000001e-05, |
| "loss": 0.6676, |
| "mean_token_accuracy": 0.8101335410028696, |
| "num_tokens": 19593016.0, |
| "step": 730 |
| }, |
| { |
| "entropy": 0.700637718103826, |
| "epoch": 1.3169021920551909, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.478e-05, |
| "loss": 0.6669, |
| "mean_token_accuracy": 0.8078547552227974, |
| "num_tokens": 19860476.0, |
| "step": 740 |
| }, |
| { |
| "entropy": 0.7031819498166442, |
| "epoch": 1.3347056859908757, |
| "grad_norm": 2.125, |
| "learning_rate": 1.498e-05, |
| "loss": 0.6689, |
| "mean_token_accuracy": 0.8074837666004896, |
| "num_tokens": 20124436.0, |
| "step": 750 |
| }, |
| { |
| "entropy": 0.7036608405411243, |
| "epoch": 1.3525091799265607, |
| "grad_norm": 2.015625, |
| "learning_rate": 1.5180000000000002e-05, |
| "loss": 0.6786, |
| "mean_token_accuracy": 0.8082354873418808, |
| "num_tokens": 20388027.0, |
| "step": 760 |
| }, |
| { |
| "entropy": 0.7016012903302908, |
| "epoch": 1.3703126738622455, |
| "grad_norm": 1.9921875, |
| "learning_rate": 1.5380000000000002e-05, |
| "loss": 0.6705, |
| "mean_token_accuracy": 0.8066376801580191, |
| "num_tokens": 20656076.0, |
| "step": 770 |
| }, |
| { |
| "entropy": 0.7241682725027203, |
| "epoch": 1.3881161677979303, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.5580000000000003e-05, |
| "loss": 0.6972, |
| "mean_token_accuracy": 0.8026757381856442, |
| "num_tokens": 20911633.0, |
| "step": 780 |
| }, |
| { |
| "entropy": 0.7188639907166362, |
| "epoch": 1.4059196617336152, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.578e-05, |
| "loss": 0.6947, |
| "mean_token_accuracy": 0.8022018726915121, |
| "num_tokens": 21176227.0, |
| "step": 790 |
| }, |
| { |
| "entropy": 0.6969450779259205, |
| "epoch": 1.4237231556693, |
| "grad_norm": 1.9921875, |
| "learning_rate": 1.5980000000000003e-05, |
| "loss": 0.6712, |
| "mean_token_accuracy": 0.8089527126401663, |
| "num_tokens": 21445242.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.4237231556693, |
| "eval_biology_entropy": 1.1155093412399293, |
| "eval_biology_loss": 1.1545594930648804, |
| "eval_biology_mean_token_accuracy": 0.7096608462333679, |
| "eval_biology_num_tokens": 21445242.0, |
| "eval_biology_runtime": 27.4718, |
| "eval_biology_samples_per_second": 18.2, |
| "eval_biology_steps_per_second": 4.55, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.4237231556693, |
| "eval_chemistry_entropy": 0.7045279297828674, |
| "eval_chemistry_loss": 0.7082335948944092, |
| "eval_chemistry_mean_token_accuracy": 0.8038255844116211, |
| "eval_chemistry_num_tokens": 21445242.0, |
| "eval_chemistry_runtime": 34.6567, |
| "eval_chemistry_samples_per_second": 14.427, |
| "eval_chemistry_steps_per_second": 3.607, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.4237231556693, |
| "eval_math_entropy": 0.7524438078403473, |
| "eval_math_loss": 1.0430735349655151, |
| "eval_math_mean_token_accuracy": 0.756148540019989, |
| "eval_math_num_tokens": 21445242.0, |
| "eval_math_runtime": 35.681, |
| "eval_math_samples_per_second": 14.013, |
| "eval_math_steps_per_second": 3.503, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.4237231556693, |
| "eval_physics_entropy": 0.7572180745601654, |
| "eval_physics_loss": 0.8156519532203674, |
| "eval_physics_mean_token_accuracy": 0.7885509886741638, |
| "eval_physics_num_tokens": 21445242.0, |
| "eval_physics_runtime": 41.3121, |
| "eval_physics_samples_per_second": 12.103, |
| "eval_physics_steps_per_second": 3.026, |
| "step": 800 |
| }, |
| { |
| "entropy": 0.702604778856039, |
| "epoch": 1.441526649604985, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.618e-05, |
| "loss": 0.6747, |
| "mean_token_accuracy": 0.8062219139188528, |
| "num_tokens": 21718455.0, |
| "step": 810 |
| }, |
| { |
| "entropy": 0.6979160573333502, |
| "epoch": 1.4593301435406698, |
| "grad_norm": 1.8984375, |
| "learning_rate": 1.638e-05, |
| "loss": 0.6672, |
| "mean_token_accuracy": 0.8101254928857088, |
| "num_tokens": 21974076.0, |
| "step": 820 |
| }, |
| { |
| "entropy": 0.7017049081623554, |
| "epoch": 1.4771336374763546, |
| "grad_norm": 1.9140625, |
| "learning_rate": 1.658e-05, |
| "loss": 0.6746, |
| "mean_token_accuracy": 0.8071672473102808, |
| "num_tokens": 22241599.0, |
| "step": 830 |
| }, |
| { |
| "entropy": 0.704755617491901, |
| "epoch": 1.4949371314120397, |
| "grad_norm": 1.921875, |
| "learning_rate": 1.6780000000000002e-05, |
| "loss": 0.679, |
| "mean_token_accuracy": 0.8053142219781876, |
| "num_tokens": 22510725.0, |
| "step": 840 |
| }, |
| { |
| "entropy": 0.682666652649641, |
| "epoch": 1.5127406253477245, |
| "grad_norm": 1.9453125, |
| "learning_rate": 1.698e-05, |
| "loss": 0.6637, |
| "mean_token_accuracy": 0.8104805506765842, |
| "num_tokens": 22778514.0, |
| "step": 850 |
| }, |
| { |
| "entropy": 0.686640658043325, |
| "epoch": 1.5305441192834093, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.718e-05, |
| "loss": 0.6563, |
| "mean_token_accuracy": 0.8099232520908117, |
| "num_tokens": 23051230.0, |
| "step": 860 |
| }, |
| { |
| "entropy": 0.7000547967851162, |
| "epoch": 1.5483476132190943, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.7380000000000003e-05, |
| "loss": 0.6674, |
| "mean_token_accuracy": 0.8080349907279014, |
| "num_tokens": 23306327.0, |
| "step": 870 |
| }, |
| { |
| "entropy": 0.7020668694749475, |
| "epoch": 1.5661511071547791, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.758e-05, |
| "loss": 0.6696, |
| "mean_token_accuracy": 0.8060734212398529, |
| "num_tokens": 23573127.0, |
| "step": 880 |
| }, |
| { |
| "entropy": 0.6865475872531533, |
| "epoch": 1.583954601090464, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 0.6575, |
| "mean_token_accuracy": 0.8091985180974006, |
| "num_tokens": 23841914.0, |
| "step": 890 |
| }, |
| { |
| "entropy": 0.6941418889909983, |
| "epoch": 1.601758095026149, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.798e-05, |
| "loss": 0.6705, |
| "mean_token_accuracy": 0.8074412882328034, |
| "num_tokens": 24105114.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.601758095026149, |
| "eval_biology_entropy": 1.1217118601799012, |
| "eval_biology_loss": 1.155590295791626, |
| "eval_biology_mean_token_accuracy": 0.7096499319076538, |
| "eval_biology_num_tokens": 24105114.0, |
| "eval_biology_runtime": 27.4449, |
| "eval_biology_samples_per_second": 18.218, |
| "eval_biology_steps_per_second": 4.555, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.601758095026149, |
| "eval_chemistry_entropy": 0.6972420279979705, |
| "eval_chemistry_loss": 0.6956310868263245, |
| "eval_chemistry_mean_token_accuracy": 0.806081386089325, |
| "eval_chemistry_num_tokens": 24105114.0, |
| "eval_chemistry_runtime": 34.6492, |
| "eval_chemistry_samples_per_second": 14.43, |
| "eval_chemistry_steps_per_second": 3.608, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.601758095026149, |
| "eval_math_entropy": 0.7655970025062561, |
| "eval_math_loss": 1.0178277492523193, |
| "eval_math_mean_token_accuracy": 0.7585388913154602, |
| "eval_math_num_tokens": 24105114.0, |
| "eval_math_runtime": 35.6809, |
| "eval_math_samples_per_second": 14.013, |
| "eval_math_steps_per_second": 3.503, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.601758095026149, |
| "eval_physics_entropy": 0.7582002460956574, |
| "eval_physics_loss": 0.8036030530929565, |
| "eval_physics_mean_token_accuracy": 0.7907866368293762, |
| "eval_physics_num_tokens": 24105114.0, |
| "eval_physics_runtime": 41.3271, |
| "eval_physics_samples_per_second": 12.099, |
| "eval_physics_steps_per_second": 3.025, |
| "step": 900 |
| }, |
| { |
| "entropy": 0.6704264059662819, |
| "epoch": 1.6195615889618338, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 0.6472, |
| "mean_token_accuracy": 0.8138602152466774, |
| "num_tokens": 24384008.0, |
| "step": 910 |
| }, |
| { |
| "entropy": 0.6972090976312757, |
| "epoch": 1.6373650828975186, |
| "grad_norm": 1.9453125, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 0.6582, |
| "mean_token_accuracy": 0.8084709718823433, |
| "num_tokens": 24638205.0, |
| "step": 920 |
| }, |
| { |
| "entropy": 0.6524842688813806, |
| "epoch": 1.6551685768332036, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.858e-05, |
| "loss": 0.6271, |
| "mean_token_accuracy": 0.8176452334970236, |
| "num_tokens": 24910104.0, |
| "step": 930 |
| }, |
| { |
| "entropy": 0.6726915189996362, |
| "epoch": 1.6729720707688884, |
| "grad_norm": 1.953125, |
| "learning_rate": 1.878e-05, |
| "loss": 0.6468, |
| "mean_token_accuracy": 0.8116984341293574, |
| "num_tokens": 25181306.0, |
| "step": 940 |
| }, |
| { |
| "entropy": 0.6833181194961071, |
| "epoch": 1.6907755647045732, |
| "grad_norm": 1.9140625, |
| "learning_rate": 1.898e-05, |
| "loss": 0.6543, |
| "mean_token_accuracy": 0.8111916728317737, |
| "num_tokens": 25449138.0, |
| "step": 950 |
| }, |
| { |
| "entropy": 0.6850541615858674, |
| "epoch": 1.7085790586402583, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.918e-05, |
| "loss": 0.6533, |
| "mean_token_accuracy": 0.8115210771560669, |
| "num_tokens": 25709030.0, |
| "step": 960 |
| }, |
| { |
| "entropy": 0.6901535639539361, |
| "epoch": 1.7263825525759429, |
| "grad_norm": 1.875, |
| "learning_rate": 1.938e-05, |
| "loss": 0.6626, |
| "mean_token_accuracy": 0.808983851224184, |
| "num_tokens": 25972807.0, |
| "step": 970 |
| }, |
| { |
| "entropy": 0.6824520044028759, |
| "epoch": 1.744186046511628, |
| "grad_norm": 1.8984375, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 0.6494, |
| "mean_token_accuracy": 0.8119773466140032, |
| "num_tokens": 26244955.0, |
| "step": 980 |
| }, |
| { |
| "entropy": 0.6511875489726663, |
| "epoch": 1.761989540447313, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.978e-05, |
| "loss": 0.6282, |
| "mean_token_accuracy": 0.8185035139322281, |
| "num_tokens": 26515548.0, |
| "step": 990 |
| }, |
| { |
| "entropy": 0.6718592615798116, |
| "epoch": 1.7797930343829975, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 0.6379, |
| "mean_token_accuracy": 0.8133674435317516, |
| "num_tokens": 26782376.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.7797930343829975, |
| "eval_biology_entropy": 1.1214026446342469, |
| "eval_biology_loss": 1.1586663722991943, |
| "eval_biology_mean_token_accuracy": 0.7074249668121337, |
| "eval_biology_num_tokens": 26782376.0, |
| "eval_biology_runtime": 27.4407, |
| "eval_biology_samples_per_second": 18.221, |
| "eval_biology_steps_per_second": 4.555, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.7797930343829975, |
| "eval_chemistry_entropy": 0.6865044567584991, |
| "eval_chemistry_loss": 0.6886394023895264, |
| "eval_chemistry_mean_token_accuracy": 0.8079376211166381, |
| "eval_chemistry_num_tokens": 26782376.0, |
| "eval_chemistry_runtime": 34.3725, |
| "eval_chemistry_samples_per_second": 14.547, |
| "eval_chemistry_steps_per_second": 3.637, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.7797930343829975, |
| "eval_math_entropy": 0.7426287050247192, |
| "eval_math_loss": 1.0259448289871216, |
| "eval_math_mean_token_accuracy": 0.759060863494873, |
| "eval_math_num_tokens": 26782376.0, |
| "eval_math_runtime": 35.5412, |
| "eval_math_samples_per_second": 14.068, |
| "eval_math_steps_per_second": 3.517, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.7797930343829975, |
| "eval_physics_entropy": 0.7432767500877381, |
| "eval_physics_loss": 0.8022939562797546, |
| "eval_physics_mean_token_accuracy": 0.7904350452423096, |
| "eval_physics_num_tokens": 26782376.0, |
| "eval_physics_runtime": 41.2462, |
| "eval_physics_samples_per_second": 12.122, |
| "eval_physics_steps_per_second": 3.031, |
| "step": 1000 |
| }, |
| { |
| "entropy": 0.681940340809524, |
| "epoch": 1.7975965283186826, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 0.6547, |
| "mean_token_accuracy": 0.8105040695518255, |
| "num_tokens": 27049648.0, |
| "step": 1010 |
| }, |
| { |
| "entropy": 0.6968798106536269, |
| "epoch": 1.8154000222543676, |
| "grad_norm": 1.75, |
| "learning_rate": 1.995777777777778e-05, |
| "loss": 0.6672, |
| "mean_token_accuracy": 0.8069467183202506, |
| "num_tokens": 27315061.0, |
| "step": 1020 |
| }, |
| { |
| "entropy": 0.6640733122825623, |
| "epoch": 1.8332035161900522, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.9935555555555557e-05, |
| "loss": 0.6298, |
| "mean_token_accuracy": 0.8162682700902224, |
| "num_tokens": 27587650.0, |
| "step": 1030 |
| }, |
| { |
| "entropy": 0.6761852929368615, |
| "epoch": 1.8510070101257372, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.9913333333333335e-05, |
| "loss": 0.6535, |
| "mean_token_accuracy": 0.8131658129394055, |
| "num_tokens": 27855458.0, |
| "step": 1040 |
| }, |
| { |
| "entropy": 0.6659740578383208, |
| "epoch": 1.868810504061422, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.9891111111111112e-05, |
| "loss": 0.6394, |
| "mean_token_accuracy": 0.8157707110047341, |
| "num_tokens": 28127703.0, |
| "step": 1050 |
| }, |
| { |
| "entropy": 0.6528725802898407, |
| "epoch": 1.8866139979971068, |
| "grad_norm": 1.6328125, |
| "learning_rate": 1.986888888888889e-05, |
| "loss": 0.6185, |
| "mean_token_accuracy": 0.818981533870101, |
| "num_tokens": 28395128.0, |
| "step": 1060 |
| }, |
| { |
| "entropy": 0.6687119539827109, |
| "epoch": 1.9044174919327919, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.9846666666666668e-05, |
| "loss": 0.6408, |
| "mean_token_accuracy": 0.8138482984155416, |
| "num_tokens": 28662344.0, |
| "step": 1070 |
| }, |
| { |
| "entropy": 0.6675814667716622, |
| "epoch": 1.9222209858684767, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.9824444444444445e-05, |
| "loss": 0.6419, |
| "mean_token_accuracy": 0.8137888047844172, |
| "num_tokens": 28928769.0, |
| "step": 1080 |
| }, |
| { |
| "entropy": 0.6718493867665529, |
| "epoch": 1.9400244798041615, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.9802222222222226e-05, |
| "loss": 0.6403, |
| "mean_token_accuracy": 0.8140697859227657, |
| "num_tokens": 29186214.0, |
| "step": 1090 |
| }, |
| { |
| "entropy": 0.6808245742693544, |
| "epoch": 1.9578279737398465, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.978e-05, |
| "loss": 0.6553, |
| "mean_token_accuracy": 0.8104663461446762, |
| "num_tokens": 29446445.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.9578279737398465, |
| "eval_biology_entropy": 1.1261132264137268, |
| "eval_biology_loss": 1.1582802534103394, |
| "eval_biology_mean_token_accuracy": 0.7076550316810608, |
| "eval_biology_num_tokens": 29446445.0, |
| "eval_biology_runtime": 27.4443, |
| "eval_biology_samples_per_second": 18.219, |
| "eval_biology_steps_per_second": 4.555, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.9578279737398465, |
| "eval_chemistry_entropy": 0.6661963372230529, |
| "eval_chemistry_loss": 0.6719214916229248, |
| "eval_chemistry_mean_token_accuracy": 0.8110591564178466, |
| "eval_chemistry_num_tokens": 29446445.0, |
| "eval_chemistry_runtime": 34.6282, |
| "eval_chemistry_samples_per_second": 14.439, |
| "eval_chemistry_steps_per_second": 3.61, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.9578279737398465, |
| "eval_math_entropy": 0.7202653360366821, |
| "eval_math_loss": 1.0047616958618164, |
| "eval_math_mean_token_accuracy": 0.7631195125579834, |
| "eval_math_num_tokens": 29446445.0, |
| "eval_math_runtime": 35.6935, |
| "eval_math_samples_per_second": 14.008, |
| "eval_math_steps_per_second": 3.502, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.9578279737398465, |
| "eval_physics_entropy": 0.7280703535079956, |
| "eval_physics_loss": 0.7826510071754456, |
| "eval_physics_mean_token_accuracy": 0.7939513020515442, |
| "eval_physics_num_tokens": 29446445.0, |
| "eval_physics_runtime": 41.3329, |
| "eval_physics_samples_per_second": 12.097, |
| "eval_physics_steps_per_second": 3.024, |
| "step": 1100 |
| }, |
| { |
| "entropy": 0.6665386298671365, |
| "epoch": 1.9756314676755313, |
| "grad_norm": 1.625, |
| "learning_rate": 1.975777777777778e-05, |
| "loss": 0.6318, |
| "mean_token_accuracy": 0.815965899080038, |
| "num_tokens": 29711771.0, |
| "step": 1110 |
| }, |
| { |
| "entropy": 0.6679358556866646, |
| "epoch": 1.9934349616112161, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.9735555555555556e-05, |
| "loss": 0.6412, |
| "mean_token_accuracy": 0.8136959586292505, |
| "num_tokens": 29979994.0, |
| "step": 1120 |
| }, |
| { |
| "entropy": 0.6079216349509454, |
| "epoch": 2.010682096361411, |
| "grad_norm": 1.953125, |
| "learning_rate": 1.9713333333333337e-05, |
| "loss": 0.5794, |
| "mean_token_accuracy": 0.8284662919659769, |
| "num_tokens": 30240793.0, |
| "step": 1130 |
| }, |
| { |
| "entropy": 0.550388932786882, |
| "epoch": 2.028485590297096, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.969111111111111e-05, |
| "loss": 0.523, |
| "mean_token_accuracy": 0.8395265836268664, |
| "num_tokens": 30503988.0, |
| "step": 1140 |
| }, |
| { |
| "entropy": 0.5524052530527115, |
| "epoch": 2.0462890842327806, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.9668888888888892e-05, |
| "loss": 0.5236, |
| "mean_token_accuracy": 0.838665135204792, |
| "num_tokens": 30771841.0, |
| "step": 1150 |
| }, |
| { |
| "entropy": 0.5385194117203355, |
| "epoch": 2.0640925781684656, |
| "grad_norm": 1.90625, |
| "learning_rate": 1.9646666666666666e-05, |
| "loss": 0.5127, |
| "mean_token_accuracy": 0.8424484401941299, |
| "num_tokens": 31045037.0, |
| "step": 1160 |
| }, |
| { |
| "entropy": 0.5770289728417992, |
| "epoch": 2.0818960721041506, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.9624444444444447e-05, |
| "loss": 0.5462, |
| "mean_token_accuracy": 0.8320943620055914, |
| "num_tokens": 31299575.0, |
| "step": 1170 |
| }, |
| { |
| "entropy": 0.5732878848910332, |
| "epoch": 2.099699566039835, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.9602222222222225e-05, |
| "loss": 0.5444, |
| "mean_token_accuracy": 0.8322335626929999, |
| "num_tokens": 31565963.0, |
| "step": 1180 |
| }, |
| { |
| "entropy": 0.5452515786513686, |
| "epoch": 2.1175030599755202, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 0.5234, |
| "mean_token_accuracy": 0.8394101049751044, |
| "num_tokens": 31834212.0, |
| "step": 1190 |
| }, |
| { |
| "entropy": 0.5641031274572015, |
| "epoch": 2.1353065539112053, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.955777777777778e-05, |
| "loss": 0.5358, |
| "mean_token_accuracy": 0.8341521114110947, |
| "num_tokens": 32103356.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.1353065539112053, |
| "eval_biology_entropy": 0.956549976348877, |
| "eval_biology_loss": 1.2057111263275146, |
| "eval_biology_mean_token_accuracy": 0.7040139017105103, |
| "eval_biology_num_tokens": 32103356.0, |
| "eval_biology_runtime": 35.5816, |
| "eval_biology_samples_per_second": 14.052, |
| "eval_biology_steps_per_second": 3.513, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.1353065539112053, |
| "eval_chemistry_entropy": 0.5792287123203278, |
| "eval_chemistry_loss": 0.684585452079773, |
| "eval_chemistry_mean_token_accuracy": 0.8109366421699524, |
| "eval_chemistry_num_tokens": 32103356.0, |
| "eval_chemistry_runtime": 34.5829, |
| "eval_chemistry_samples_per_second": 14.458, |
| "eval_chemistry_steps_per_second": 3.615, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.1353065539112053, |
| "eval_math_entropy": 0.6300130400657654, |
| "eval_math_loss": 1.078292727470398, |
| "eval_math_mean_token_accuracy": 0.758972237110138, |
| "eval_math_num_tokens": 32103356.0, |
| "eval_math_runtime": 35.6847, |
| "eval_math_samples_per_second": 14.012, |
| "eval_math_steps_per_second": 3.503, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.1353065539112053, |
| "eval_physics_entropy": 0.6210716564655304, |
| "eval_physics_loss": 0.8277297019958496, |
| "eval_physics_mean_token_accuracy": 0.7909838285446167, |
| "eval_physics_num_tokens": 32103356.0, |
| "eval_physics_runtime": 41.3252, |
| "eval_physics_samples_per_second": 12.099, |
| "eval_physics_steps_per_second": 3.025, |
| "step": 1200 |
| }, |
| { |
| "entropy": 0.5546079104766249, |
| "epoch": 2.15311004784689, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.9535555555555557e-05, |
| "loss": 0.5238, |
| "mean_token_accuracy": 0.8379202175885438, |
| "num_tokens": 32374830.0, |
| "step": 1210 |
| }, |
| { |
| "entropy": 0.5367290515452623, |
| "epoch": 2.170913541782575, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.9513333333333335e-05, |
| "loss": 0.5147, |
| "mean_token_accuracy": 0.8416944585740567, |
| "num_tokens": 32643479.0, |
| "step": 1220 |
| }, |
| { |
| "entropy": 0.5465414123609662, |
| "epoch": 2.18871703571826, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.9491111111111113e-05, |
| "loss": 0.5185, |
| "mean_token_accuracy": 0.8402627848088742, |
| "num_tokens": 32908761.0, |
| "step": 1230 |
| }, |
| { |
| "entropy": 0.5668866217136384, |
| "epoch": 2.2065205296539445, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.946888888888889e-05, |
| "loss": 0.5353, |
| "mean_token_accuracy": 0.8366922777146101, |
| "num_tokens": 33166413.0, |
| "step": 1240 |
| }, |
| { |
| "entropy": 0.5626859543845057, |
| "epoch": 2.2243240235896296, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.9446666666666668e-05, |
| "loss": 0.5369, |
| "mean_token_accuracy": 0.8356177676469088, |
| "num_tokens": 33432686.0, |
| "step": 1250 |
| }, |
| { |
| "entropy": 0.5568565413355827, |
| "epoch": 2.242127517525314, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.9424444444444446e-05, |
| "loss": 0.5338, |
| "mean_token_accuracy": 0.8363381318747998, |
| "num_tokens": 33693242.0, |
| "step": 1260 |
| }, |
| { |
| "entropy": 0.5357751946896314, |
| "epoch": 2.259931011460999, |
| "grad_norm": 1.75, |
| "learning_rate": 1.9402222222222223e-05, |
| "loss": 0.5046, |
| "mean_token_accuracy": 0.8440463934093714, |
| "num_tokens": 33970820.0, |
| "step": 1270 |
| }, |
| { |
| "entropy": 0.5442078202962876, |
| "epoch": 2.277734505396684, |
| "grad_norm": 1.8828125, |
| "learning_rate": 1.938e-05, |
| "loss": 0.5214, |
| "mean_token_accuracy": 0.8392076704651117, |
| "num_tokens": 34243042.0, |
| "step": 1280 |
| }, |
| { |
| "entropy": 0.5499420555308461, |
| "epoch": 2.295537999332369, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.935777777777778e-05, |
| "loss": 0.5281, |
| "mean_token_accuracy": 0.8386849842965602, |
| "num_tokens": 34515236.0, |
| "step": 1290 |
| }, |
| { |
| "entropy": 0.5299480393528938, |
| "epoch": 2.313341493268054, |
| "grad_norm": 1.75, |
| "learning_rate": 1.9335555555555556e-05, |
| "loss": 0.5046, |
| "mean_token_accuracy": 0.8450733765959739, |
| "num_tokens": 34785068.0, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.313341493268054, |
| "eval_biology_entropy": 0.9190451064109803, |
| "eval_biology_loss": 1.2270817756652832, |
| "eval_biology_mean_token_accuracy": 0.7022366142272949, |
| "eval_biology_num_tokens": 34785068.0, |
| "eval_biology_runtime": 46.5074, |
| "eval_biology_samples_per_second": 10.751, |
| "eval_biology_steps_per_second": 2.688, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.313341493268054, |
| "eval_chemistry_entropy": 0.5683628201484681, |
| "eval_chemistry_loss": 0.6834931969642639, |
| "eval_chemistry_mean_token_accuracy": 0.811858865737915, |
| "eval_chemistry_num_tokens": 34785068.0, |
| "eval_chemistry_runtime": 34.5417, |
| "eval_chemistry_samples_per_second": 14.475, |
| "eval_chemistry_steps_per_second": 3.619, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.313341493268054, |
| "eval_math_entropy": 0.6246570134162903, |
| "eval_math_loss": 1.0720049142837524, |
| "eval_math_mean_token_accuracy": 0.759788429737091, |
| "eval_math_num_tokens": 34785068.0, |
| "eval_math_runtime": 35.6064, |
| "eval_math_samples_per_second": 14.042, |
| "eval_math_steps_per_second": 3.511, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.313341493268054, |
| "eval_physics_entropy": 0.6201333129405975, |
| "eval_physics_loss": 0.8274801969528198, |
| "eval_physics_mean_token_accuracy": 0.7917842483520507, |
| "eval_physics_num_tokens": 34785068.0, |
| "eval_physics_runtime": 41.2855, |
| "eval_physics_samples_per_second": 12.111, |
| "eval_physics_steps_per_second": 3.028, |
| "step": 1300 |
| }, |
| { |
| "entropy": 0.5658086974173784, |
| "epoch": 2.331144987203739, |
| "grad_norm": 1.625, |
| "learning_rate": 1.9313333333333334e-05, |
| "loss": 0.5374, |
| "mean_token_accuracy": 0.8340179108083248, |
| "num_tokens": 35055845.0, |
| "step": 1310 |
| }, |
| { |
| "entropy": 0.5485894976183772, |
| "epoch": 2.3489484811394234, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.9291111111111115e-05, |
| "loss": 0.5214, |
| "mean_token_accuracy": 0.8392617363482714, |
| "num_tokens": 35327481.0, |
| "step": 1320 |
| }, |
| { |
| "entropy": 0.5623961247503757, |
| "epoch": 2.3667519750751085, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.926888888888889e-05, |
| "loss": 0.5378, |
| "mean_token_accuracy": 0.8353296346962452, |
| "num_tokens": 35596362.0, |
| "step": 1330 |
| }, |
| { |
| "entropy": 0.552398738451302, |
| "epoch": 2.3845554690107935, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.924666666666667e-05, |
| "loss": 0.5233, |
| "mean_token_accuracy": 0.839025941118598, |
| "num_tokens": 35869585.0, |
| "step": 1340 |
| }, |
| { |
| "entropy": 0.555871544405818, |
| "epoch": 2.402358962946478, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.9224444444444444e-05, |
| "loss": 0.5315, |
| "mean_token_accuracy": 0.8366593103855848, |
| "num_tokens": 36134072.0, |
| "step": 1350 |
| }, |
| { |
| "entropy": 0.549046789854765, |
| "epoch": 2.420162456882163, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.9202222222222225e-05, |
| "loss": 0.5258, |
| "mean_token_accuracy": 0.8397909853607416, |
| "num_tokens": 36406263.0, |
| "step": 1360 |
| }, |
| { |
| "entropy": 0.554560680128634, |
| "epoch": 2.437965950817848, |
| "grad_norm": 1.890625, |
| "learning_rate": 1.918e-05, |
| "loss": 0.5257, |
| "mean_token_accuracy": 0.8386047866195441, |
| "num_tokens": 36676822.0, |
| "step": 1370 |
| }, |
| { |
| "entropy": 0.5509166169911623, |
| "epoch": 2.4557694447535328, |
| "grad_norm": 1.9140625, |
| "learning_rate": 1.915777777777778e-05, |
| "loss": 0.5245, |
| "mean_token_accuracy": 0.8386286627501249, |
| "num_tokens": 36942837.0, |
| "step": 1380 |
| }, |
| { |
| "entropy": 0.553687858581543, |
| "epoch": 2.473572938689218, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.9135555555555555e-05, |
| "loss": 0.5288, |
| "mean_token_accuracy": 0.8373784240335226, |
| "num_tokens": 37210791.0, |
| "step": 1390 |
| }, |
| { |
| "entropy": 0.537860699184239, |
| "epoch": 2.491376432624903, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.9113333333333336e-05, |
| "loss": 0.5195, |
| "mean_token_accuracy": 0.8420177605003119, |
| "num_tokens": 37485822.0, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.491376432624903, |
| "eval_biology_entropy": 0.9255113205909729, |
| "eval_biology_loss": 1.2256031036376953, |
| "eval_biology_mean_token_accuracy": 0.703334276676178, |
| "eval_biology_num_tokens": 37485822.0, |
| "eval_biology_runtime": 27.455, |
| "eval_biology_samples_per_second": 18.212, |
| "eval_biology_steps_per_second": 4.553, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.491376432624903, |
| "eval_chemistry_entropy": 0.5651899397373199, |
| "eval_chemistry_loss": 0.6717930436134338, |
| "eval_chemistry_mean_token_accuracy": 0.8143071975708008, |
| "eval_chemistry_num_tokens": 37485822.0, |
| "eval_chemistry_runtime": 34.6518, |
| "eval_chemistry_samples_per_second": 14.429, |
| "eval_chemistry_steps_per_second": 3.607, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.491376432624903, |
| "eval_math_entropy": 0.6243210020065307, |
| "eval_math_loss": 1.0662870407104492, |
| "eval_math_mean_token_accuracy": 0.7615419855117798, |
| "eval_math_num_tokens": 37485822.0, |
| "eval_math_runtime": 35.6708, |
| "eval_math_samples_per_second": 14.017, |
| "eval_math_steps_per_second": 3.504, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.491376432624903, |
| "eval_physics_entropy": 0.6135793507099152, |
| "eval_physics_loss": 0.8177351355552673, |
| "eval_physics_mean_token_accuracy": 0.7939276299476623, |
| "eval_physics_num_tokens": 37485822.0, |
| "eval_physics_runtime": 41.3365, |
| "eval_physics_samples_per_second": 12.096, |
| "eval_physics_steps_per_second": 3.024, |
| "step": 1400 |
| }, |
| { |
| "entropy": 0.5481447698548436, |
| "epoch": 2.5091799265605874, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.9091111111111113e-05, |
| "loss": 0.5196, |
| "mean_token_accuracy": 0.8403281763195991, |
| "num_tokens": 37746150.0, |
| "step": 1410 |
| }, |
| { |
| "entropy": 0.5513471394777298, |
| "epoch": 2.5269834204962724, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.906888888888889e-05, |
| "loss": 0.5221, |
| "mean_token_accuracy": 0.8380666643381118, |
| "num_tokens": 38016863.0, |
| "step": 1420 |
| }, |
| { |
| "entropy": 0.5683571495115757, |
| "epoch": 2.544786914431957, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.904666666666667e-05, |
| "loss": 0.5372, |
| "mean_token_accuracy": 0.8328249994665384, |
| "num_tokens": 38277754.0, |
| "step": 1430 |
| }, |
| { |
| "entropy": 0.5602214207872749, |
| "epoch": 2.562590408367642, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.9024444444444446e-05, |
| "loss": 0.5359, |
| "mean_token_accuracy": 0.8351134214550257, |
| "num_tokens": 38542761.0, |
| "step": 1440 |
| }, |
| { |
| "entropy": 0.5457610888406634, |
| "epoch": 2.580393902303327, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.9002222222222224e-05, |
| "loss": 0.5188, |
| "mean_token_accuracy": 0.8395175360143184, |
| "num_tokens": 38820999.0, |
| "step": 1450 |
| }, |
| { |
| "entropy": 0.5467756260186434, |
| "epoch": 2.598197396239012, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.898e-05, |
| "loss": 0.5273, |
| "mean_token_accuracy": 0.8386720154434443, |
| "num_tokens": 39080016.0, |
| "step": 1460 |
| }, |
| { |
| "entropy": 0.5515998177230358, |
| "epoch": 2.6160008901746967, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.895777777777778e-05, |
| "loss": 0.5199, |
| "mean_token_accuracy": 0.8405467420816422, |
| "num_tokens": 39345434.0, |
| "step": 1470 |
| }, |
| { |
| "entropy": 0.5525520471855998, |
| "epoch": 2.6338043841103818, |
| "grad_norm": 1.9140625, |
| "learning_rate": 1.8935555555555556e-05, |
| "loss": 0.5275, |
| "mean_token_accuracy": 0.8375415705144406, |
| "num_tokens": 39610041.0, |
| "step": 1480 |
| }, |
| { |
| "entropy": 0.5331873701885342, |
| "epoch": 2.6516078780460663, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.8913333333333334e-05, |
| "loss": 0.5088, |
| "mean_token_accuracy": 0.8429557517170906, |
| "num_tokens": 39884185.0, |
| "step": 1490 |
| }, |
| { |
| "entropy": 0.5642509242519737, |
| "epoch": 2.6694113719817514, |
| "grad_norm": 1.96875, |
| "learning_rate": 1.8891111111111115e-05, |
| "loss": 0.5362, |
| "mean_token_accuracy": 0.8353289574384689, |
| "num_tokens": 40143543.0, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.6694113719817514, |
| "eval_biology_entropy": 0.9068852887153626, |
| "eval_biology_loss": 1.2326120138168335, |
| "eval_biology_mean_token_accuracy": 0.7021635494232178, |
| "eval_biology_num_tokens": 40143543.0, |
| "eval_biology_runtime": 27.4204, |
| "eval_biology_samples_per_second": 18.235, |
| "eval_biology_steps_per_second": 4.559, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.6694113719817514, |
| "eval_chemistry_entropy": 0.5678843791484833, |
| "eval_chemistry_loss": 0.6659799814224243, |
| "eval_chemistry_mean_token_accuracy": 0.8151456089019775, |
| "eval_chemistry_num_tokens": 40143543.0, |
| "eval_chemistry_runtime": 34.3579, |
| "eval_chemistry_samples_per_second": 14.553, |
| "eval_chemistry_steps_per_second": 3.638, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.6694113719817514, |
| "eval_math_entropy": 0.6364672091007233, |
| "eval_math_loss": 1.0642262697219849, |
| "eval_math_mean_token_accuracy": 0.7619977474212647, |
| "eval_math_num_tokens": 40143543.0, |
| "eval_math_runtime": 35.5283, |
| "eval_math_samples_per_second": 14.073, |
| "eval_math_steps_per_second": 3.518, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.6694113719817514, |
| "eval_physics_entropy": 0.6290303509235382, |
| "eval_physics_loss": 0.8116629719734192, |
| "eval_physics_mean_token_accuracy": 0.7934681911468506, |
| "eval_physics_num_tokens": 40143543.0, |
| "eval_physics_runtime": 41.2302, |
| "eval_physics_samples_per_second": 12.127, |
| "eval_physics_steps_per_second": 3.032, |
| "step": 1500 |
| }, |
| { |
| "entropy": 0.5444372620433569, |
| "epoch": 2.6872148659174364, |
| "grad_norm": 1.8828125, |
| "learning_rate": 1.886888888888889e-05, |
| "loss": 0.5189, |
| "mean_token_accuracy": 0.8400942515581846, |
| "num_tokens": 40406812.0, |
| "step": 1510 |
| }, |
| { |
| "entropy": 0.5620427438989282, |
| "epoch": 2.7050183598531214, |
| "grad_norm": 2.015625, |
| "learning_rate": 1.884666666666667e-05, |
| "loss": 0.5391, |
| "mean_token_accuracy": 0.8342275880277157, |
| "num_tokens": 40668543.0, |
| "step": 1520 |
| }, |
| { |
| "entropy": 0.556936101987958, |
| "epoch": 2.722821853788806, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.8824444444444445e-05, |
| "loss": 0.5235, |
| "mean_token_accuracy": 0.8389734990894795, |
| "num_tokens": 40942603.0, |
| "step": 1530 |
| }, |
| { |
| "entropy": 0.5577435137704014, |
| "epoch": 2.740625347724491, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.8802222222222226e-05, |
| "loss": 0.5244, |
| "mean_token_accuracy": 0.8378285989165306, |
| "num_tokens": 41208093.0, |
| "step": 1540 |
| }, |
| { |
| "entropy": 0.545657803863287, |
| "epoch": 2.7584288416601757, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.878e-05, |
| "loss": 0.5255, |
| "mean_token_accuracy": 0.8382109336555004, |
| "num_tokens": 41477562.0, |
| "step": 1550 |
| }, |
| { |
| "entropy": 0.539658653922379, |
| "epoch": 2.7762323355958607, |
| "grad_norm": 1.9453125, |
| "learning_rate": 1.875777777777778e-05, |
| "loss": 0.5171, |
| "mean_token_accuracy": 0.8404715970158577, |
| "num_tokens": 41746571.0, |
| "step": 1560 |
| }, |
| { |
| "entropy": 0.5481333693489432, |
| "epoch": 2.7940358295315457, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.873555555555556e-05, |
| "loss": 0.5196, |
| "mean_token_accuracy": 0.8405867479741573, |
| "num_tokens": 42011633.0, |
| "step": 1570 |
| }, |
| { |
| "entropy": 0.5563079403713346, |
| "epoch": 2.8118393234672303, |
| "grad_norm": 1.609375, |
| "learning_rate": 1.8713333333333336e-05, |
| "loss": 0.5298, |
| "mean_token_accuracy": 0.8373136855661869, |
| "num_tokens": 42283087.0, |
| "step": 1580 |
| }, |
| { |
| "entropy": 0.5321604607626795, |
| "epoch": 2.8296428174029153, |
| "grad_norm": 1.5625, |
| "learning_rate": 1.8691111111111114e-05, |
| "loss": 0.5089, |
| "mean_token_accuracy": 0.8434795420616865, |
| "num_tokens": 42561534.0, |
| "step": 1590 |
| }, |
| { |
| "entropy": 0.5442171167582274, |
| "epoch": 2.8474463113386, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.866888888888889e-05, |
| "loss": 0.5263, |
| "mean_token_accuracy": 0.840014697611332, |
| "num_tokens": 42834266.0, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.8474463113386, |
| "eval_biology_entropy": 0.9398055624961853, |
| "eval_biology_loss": 1.2283508777618408, |
| "eval_biology_mean_token_accuracy": 0.701851697921753, |
| "eval_biology_num_tokens": 42834266.0, |
| "eval_biology_runtime": 27.4556, |
| "eval_biology_samples_per_second": 18.211, |
| "eval_biology_steps_per_second": 4.553, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.8474463113386, |
| "eval_chemistry_entropy": 0.57317799949646, |
| "eval_chemistry_loss": 0.6582952737808228, |
| "eval_chemistry_mean_token_accuracy": 0.816745276927948, |
| "eval_chemistry_num_tokens": 42834266.0, |
| "eval_chemistry_runtime": 48.0292, |
| "eval_chemistry_samples_per_second": 10.41, |
| "eval_chemistry_steps_per_second": 2.603, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.8474463113386, |
| "eval_math_entropy": 0.6262275006771088, |
| "eval_math_loss": 1.0604408979415894, |
| "eval_math_mean_token_accuracy": 0.762318911075592, |
| "eval_math_num_tokens": 42834266.0, |
| "eval_math_runtime": 35.6417, |
| "eval_math_samples_per_second": 14.029, |
| "eval_math_steps_per_second": 3.507, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.8474463113386, |
| "eval_physics_entropy": 0.6248719379901886, |
| "eval_physics_loss": 0.8128731846809387, |
| "eval_physics_mean_token_accuracy": 0.7944864196777344, |
| "eval_physics_num_tokens": 42834266.0, |
| "eval_physics_runtime": 41.2757, |
| "eval_physics_samples_per_second": 12.114, |
| "eval_physics_steps_per_second": 3.028, |
| "step": 1600 |
| }, |
| { |
| "entropy": 0.570980884693563, |
| "epoch": 2.865249805274285, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.864666666666667e-05, |
| "loss": 0.5414, |
| "mean_token_accuracy": 0.8347698282450438, |
| "num_tokens": 43100513.0, |
| "step": 1610 |
| }, |
| { |
| "entropy": 0.5554070059210062, |
| "epoch": 2.88305329920997, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.8624444444444446e-05, |
| "loss": 0.5253, |
| "mean_token_accuracy": 0.8384413819760084, |
| "num_tokens": 43375632.0, |
| "step": 1620 |
| }, |
| { |
| "entropy": 0.5394276781007648, |
| "epoch": 2.900856793145655, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.8602222222222224e-05, |
| "loss": 0.52, |
| "mean_token_accuracy": 0.8410471968352795, |
| "num_tokens": 43644341.0, |
| "step": 1630 |
| }, |
| { |
| "entropy": 0.5360630100592971, |
| "epoch": 2.9186602870813396, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.858e-05, |
| "loss": 0.5094, |
| "mean_token_accuracy": 0.8415741622447968, |
| "num_tokens": 43917580.0, |
| "step": 1640 |
| }, |
| { |
| "entropy": 0.5425418116152286, |
| "epoch": 2.9364637810170247, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.855777777777778e-05, |
| "loss": 0.5179, |
| "mean_token_accuracy": 0.8425847593694925, |
| "num_tokens": 44177325.0, |
| "step": 1650 |
| }, |
| { |
| "entropy": 0.5552718479186296, |
| "epoch": 2.9542672749527092, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.8535555555555557e-05, |
| "loss": 0.5254, |
| "mean_token_accuracy": 0.8371932659298181, |
| "num_tokens": 44442947.0, |
| "step": 1660 |
| }, |
| { |
| "entropy": 0.5552935941144824, |
| "epoch": 2.9720707688883943, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.8513333333333335e-05, |
| "loss": 0.5323, |
| "mean_token_accuracy": 0.8363202281296254, |
| "num_tokens": 44700601.0, |
| "step": 1670 |
| }, |
| { |
| "entropy": 0.5467931361868977, |
| "epoch": 2.9898742628240793, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.8491111111111112e-05, |
| "loss": 0.5177, |
| "mean_token_accuracy": 0.8421835143119096, |
| "num_tokens": 44971479.0, |
| "step": 1680 |
| }, |
| { |
| "entropy": 0.5015917474223721, |
| "epoch": 3.007121397574274, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.846888888888889e-05, |
| "loss": 0.4417, |
| "mean_token_accuracy": 0.8615512405672381, |
| "num_tokens": 45231796.0, |
| "step": 1690 |
| }, |
| { |
| "entropy": 0.35349990567192435, |
| "epoch": 3.0249248915099587, |
| "grad_norm": 2.1875, |
| "learning_rate": 1.8446666666666667e-05, |
| "loss": 0.3434, |
| "mean_token_accuracy": 0.8858234889805316, |
| "num_tokens": 45500317.0, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.0249248915099587, |
| "eval_biology_entropy": 0.695198887348175, |
| "eval_biology_loss": 1.4266223907470703, |
| "eval_biology_mean_token_accuracy": 0.6922373852729797, |
| "eval_biology_num_tokens": 45500317.0, |
| "eval_biology_runtime": 27.4124, |
| "eval_biology_samples_per_second": 18.24, |
| "eval_biology_steps_per_second": 4.56, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.0249248915099587, |
| "eval_chemistry_entropy": 0.4068129107952118, |
| "eval_chemistry_loss": 0.7570622563362122, |
| "eval_chemistry_mean_token_accuracy": 0.8122591314315796, |
| "eval_chemistry_num_tokens": 45500317.0, |
| "eval_chemistry_runtime": 34.6241, |
| "eval_chemistry_samples_per_second": 14.441, |
| "eval_chemistry_steps_per_second": 3.61, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.0249248915099587, |
| "eval_math_entropy": 0.46907779622077944, |
| "eval_math_loss": 1.2473300695419312, |
| "eval_math_mean_token_accuracy": 0.7579210791587829, |
| "eval_math_num_tokens": 45500317.0, |
| "eval_math_runtime": 35.6827, |
| "eval_math_samples_per_second": 14.012, |
| "eval_math_steps_per_second": 3.503, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.0249248915099587, |
| "eval_physics_entropy": 0.4469644265174866, |
| "eval_physics_loss": 0.964731752872467, |
| "eval_physics_mean_token_accuracy": 0.7871066861152649, |
| "eval_physics_num_tokens": 45500317.0, |
| "eval_physics_runtime": 41.3134, |
| "eval_physics_samples_per_second": 12.103, |
| "eval_physics_steps_per_second": 3.026, |
| "step": 1700 |
| }, |
| { |
| "entropy": 0.358632389549166, |
| "epoch": 3.0427283854456437, |
| "grad_norm": 2.265625, |
| "learning_rate": 1.842444444444445e-05, |
| "loss": 0.3136, |
| "mean_token_accuracy": 0.8949032187461853, |
| "num_tokens": 45762680.0, |
| "step": 1710 |
| }, |
| { |
| "entropy": 0.33809250192716717, |
| "epoch": 3.0605318793813288, |
| "grad_norm": 2.328125, |
| "learning_rate": 1.8402222222222223e-05, |
| "loss": 0.3137, |
| "mean_token_accuracy": 0.8946652490645647, |
| "num_tokens": 46029923.0, |
| "step": 1720 |
| }, |
| { |
| "entropy": 0.34262730833142996, |
| "epoch": 3.0783353733170133, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 0.317, |
| "mean_token_accuracy": 0.8937140975147486, |
| "num_tokens": 46305156.0, |
| "step": 1730 |
| }, |
| { |
| "entropy": 0.33494829703122375, |
| "epoch": 3.0961388672526984, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.8357777777777778e-05, |
| "loss": 0.3106, |
| "mean_token_accuracy": 0.8961000189185142, |
| "num_tokens": 46582874.0, |
| "step": 1740 |
| }, |
| { |
| "entropy": 0.3524747174233198, |
| "epoch": 3.1139423611883834, |
| "grad_norm": 2.640625, |
| "learning_rate": 1.833555555555556e-05, |
| "loss": 0.3229, |
| "mean_token_accuracy": 0.8921468783169985, |
| "num_tokens": 46850175.0, |
| "step": 1750 |
| }, |
| { |
| "entropy": 0.33744557332247493, |
| "epoch": 3.131745855124068, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.8313333333333333e-05, |
| "loss": 0.3114, |
| "mean_token_accuracy": 0.8956554789096117, |
| "num_tokens": 47125899.0, |
| "step": 1760 |
| }, |
| { |
| "entropy": 0.3454610071144998, |
| "epoch": 3.149549349059753, |
| "grad_norm": 2.484375, |
| "learning_rate": 1.8291111111111114e-05, |
| "loss": 0.3187, |
| "mean_token_accuracy": 0.8930278174579144, |
| "num_tokens": 47390194.0, |
| "step": 1770 |
| }, |
| { |
| "entropy": 0.35187926813960074, |
| "epoch": 3.167352842995438, |
| "grad_norm": 2.265625, |
| "learning_rate": 1.8268888888888888e-05, |
| "loss": 0.3227, |
| "mean_token_accuracy": 0.891836655139923, |
| "num_tokens": 47654784.0, |
| "step": 1780 |
| }, |
| { |
| "entropy": 0.3448355795815587, |
| "epoch": 3.1851563369311227, |
| "grad_norm": 2.390625, |
| "learning_rate": 1.824666666666667e-05, |
| "loss": 0.3204, |
| "mean_token_accuracy": 0.8938332315534353, |
| "num_tokens": 47919400.0, |
| "step": 1790 |
| }, |
| { |
| "entropy": 0.3550509645603597, |
| "epoch": 3.2029598308668077, |
| "grad_norm": 2.40625, |
| "learning_rate": 1.8224444444444447e-05, |
| "loss": 0.3271, |
| "mean_token_accuracy": 0.8914212208241225, |
| "num_tokens": 48187796.0, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.2029598308668077, |
| "eval_biology_entropy": 0.6370464863777161, |
| "eval_biology_loss": 1.522788166999817, |
| "eval_biology_mean_token_accuracy": 0.6883723630905151, |
| "eval_biology_num_tokens": 48187796.0, |
| "eval_biology_runtime": 27.4271, |
| "eval_biology_samples_per_second": 18.23, |
| "eval_biology_steps_per_second": 4.558, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.2029598308668077, |
| "eval_chemistry_entropy": 0.39649688673019406, |
| "eval_chemistry_loss": 0.782845675945282, |
| "eval_chemistry_mean_token_accuracy": 0.8119130005836487, |
| "eval_chemistry_num_tokens": 48187796.0, |
| "eval_chemistry_runtime": 34.6311, |
| "eval_chemistry_samples_per_second": 14.438, |
| "eval_chemistry_steps_per_second": 3.609, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.2029598308668077, |
| "eval_math_entropy": 0.4450396730899811, |
| "eval_math_loss": 1.3091418743133545, |
| "eval_math_mean_token_accuracy": 0.7542828044891358, |
| "eval_math_num_tokens": 48187796.0, |
| "eval_math_runtime": 36.0092, |
| "eval_math_samples_per_second": 13.885, |
| "eval_math_steps_per_second": 3.471, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.2029598308668077, |
| "eval_physics_entropy": 0.4299520983695984, |
| "eval_physics_loss": 1.0064133405685425, |
| "eval_physics_mean_token_accuracy": 0.7850898914337158, |
| "eval_physics_num_tokens": 48187796.0, |
| "eval_physics_runtime": 41.2826, |
| "eval_physics_samples_per_second": 12.112, |
| "eval_physics_steps_per_second": 3.028, |
| "step": 1800 |
| }, |
| { |
| "entropy": 0.34619652703404424, |
| "epoch": 3.2207633248024923, |
| "grad_norm": 2.390625, |
| "learning_rate": 1.8202222222222225e-05, |
| "loss": 0.3217, |
| "mean_token_accuracy": 0.8925070457160473, |
| "num_tokens": 48457428.0, |
| "step": 1810 |
| }, |
| { |
| "entropy": 0.3736777358688414, |
| "epoch": 3.2385668187381773, |
| "grad_norm": 2.40625, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 0.3381, |
| "mean_token_accuracy": 0.8857701819390058, |
| "num_tokens": 48716417.0, |
| "step": 1820 |
| }, |
| { |
| "entropy": 0.34597971392795446, |
| "epoch": 3.2563703126738623, |
| "grad_norm": 2.359375, |
| "learning_rate": 1.815777777777778e-05, |
| "loss": 0.3178, |
| "mean_token_accuracy": 0.8930837467312813, |
| "num_tokens": 48977306.0, |
| "step": 1830 |
| }, |
| { |
| "entropy": 0.3459949854761362, |
| "epoch": 3.274173806609547, |
| "grad_norm": 2.1875, |
| "learning_rate": 1.8135555555555557e-05, |
| "loss": 0.319, |
| "mean_token_accuracy": 0.8932924944907427, |
| "num_tokens": 49249667.0, |
| "step": 1840 |
| }, |
| { |
| "entropy": 0.3554858529008925, |
| "epoch": 3.291977300545232, |
| "grad_norm": 2.234375, |
| "learning_rate": 1.8113333333333335e-05, |
| "loss": 0.3279, |
| "mean_token_accuracy": 0.8899349391460418, |
| "num_tokens": 49520900.0, |
| "step": 1850 |
| }, |
| { |
| "entropy": 0.3597189445048571, |
| "epoch": 3.309780794480917, |
| "grad_norm": 2.5, |
| "learning_rate": 1.8091111111111113e-05, |
| "loss": 0.3309, |
| "mean_token_accuracy": 0.8891872499138117, |
| "num_tokens": 49786411.0, |
| "step": 1860 |
| }, |
| { |
| "entropy": 0.34456501295790076, |
| "epoch": 3.3275842884166016, |
| "grad_norm": 2.3125, |
| "learning_rate": 1.806888888888889e-05, |
| "loss": 0.3169, |
| "mean_token_accuracy": 0.8927634313702584, |
| "num_tokens": 50056207.0, |
| "step": 1870 |
| }, |
| { |
| "entropy": 0.3534662628546357, |
| "epoch": 3.3453877823522866, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.8046666666666668e-05, |
| "loss": 0.3239, |
| "mean_token_accuracy": 0.8915927741676569, |
| "num_tokens": 50330876.0, |
| "step": 1880 |
| }, |
| { |
| "entropy": 0.35242587570101025, |
| "epoch": 3.3631912762879717, |
| "grad_norm": 2.53125, |
| "learning_rate": 1.8024444444444445e-05, |
| "loss": 0.3271, |
| "mean_token_accuracy": 0.8908862814307212, |
| "num_tokens": 50602004.0, |
| "step": 1890 |
| }, |
| { |
| "entropy": 0.3582916082814336, |
| "epoch": 3.3809947702236562, |
| "grad_norm": 2.328125, |
| "learning_rate": 1.8002222222222223e-05, |
| "loss": 0.3295, |
| "mean_token_accuracy": 0.8895761199295521, |
| "num_tokens": 50873444.0, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.3809947702236562, |
| "eval_biology_entropy": 0.6187065536975861, |
| "eval_biology_loss": 1.5503290891647339, |
| "eval_biology_mean_token_accuracy": 0.6876367254257202, |
| "eval_biology_num_tokens": 50873444.0, |
| "eval_biology_runtime": 27.4338, |
| "eval_biology_samples_per_second": 18.226, |
| "eval_biology_steps_per_second": 4.556, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.3809947702236562, |
| "eval_chemistry_entropy": 0.39340255749225617, |
| "eval_chemistry_loss": 0.790208101272583, |
| "eval_chemistry_mean_token_accuracy": 0.8109837841987609, |
| "eval_chemistry_num_tokens": 50873444.0, |
| "eval_chemistry_runtime": 34.6186, |
| "eval_chemistry_samples_per_second": 14.443, |
| "eval_chemistry_steps_per_second": 3.611, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.3809947702236562, |
| "eval_math_entropy": 0.43355243229866025, |
| "eval_math_loss": 1.338222622871399, |
| "eval_math_mean_token_accuracy": 0.7530519342422486, |
| "eval_math_num_tokens": 50873444.0, |
| "eval_math_runtime": 35.6604, |
| "eval_math_samples_per_second": 14.021, |
| "eval_math_steps_per_second": 3.505, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.3809947702236562, |
| "eval_physics_entropy": 0.427370148897171, |
| "eval_physics_loss": 1.0167144536972046, |
| "eval_physics_mean_token_accuracy": 0.7849561262130738, |
| "eval_physics_num_tokens": 50873444.0, |
| "eval_physics_runtime": 41.3101, |
| "eval_physics_samples_per_second": 12.104, |
| "eval_physics_steps_per_second": 3.026, |
| "step": 1900 |
| }, |
| { |
| "entropy": 0.3595650595612824, |
| "epoch": 3.3987982641593413, |
| "grad_norm": 2.34375, |
| "learning_rate": 1.798e-05, |
| "loss": 0.3329, |
| "mean_token_accuracy": 0.8875269912183285, |
| "num_tokens": 51141077.0, |
| "step": 1910 |
| }, |
| { |
| "entropy": 0.35742873027920724, |
| "epoch": 3.4166017580950263, |
| "grad_norm": 2.546875, |
| "learning_rate": 1.7957777777777778e-05, |
| "loss": 0.3303, |
| "mean_token_accuracy": 0.8895150814205408, |
| "num_tokens": 51406483.0, |
| "step": 1920 |
| }, |
| { |
| "entropy": 0.356181532330811, |
| "epoch": 3.434405252030711, |
| "grad_norm": 2.5, |
| "learning_rate": 1.7935555555555556e-05, |
| "loss": 0.3268, |
| "mean_token_accuracy": 0.8905583586543798, |
| "num_tokens": 51668458.0, |
| "step": 1930 |
| }, |
| { |
| "entropy": 0.36474447073414923, |
| "epoch": 3.452208745966396, |
| "grad_norm": 2.53125, |
| "learning_rate": 1.7913333333333337e-05, |
| "loss": 0.3353, |
| "mean_token_accuracy": 0.8880269512534141, |
| "num_tokens": 51928199.0, |
| "step": 1940 |
| }, |
| { |
| "entropy": 0.3577393182553351, |
| "epoch": 3.470012239902081, |
| "grad_norm": 2.34375, |
| "learning_rate": 1.789111111111111e-05, |
| "loss": 0.3306, |
| "mean_token_accuracy": 0.8887866042554379, |
| "num_tokens": 52199244.0, |
| "step": 1950 |
| }, |
| { |
| "entropy": 0.3538716332055628, |
| "epoch": 3.4878157338377656, |
| "grad_norm": 2.5625, |
| "learning_rate": 1.7868888888888892e-05, |
| "loss": 0.3253, |
| "mean_token_accuracy": 0.8902977962046862, |
| "num_tokens": 52473476.0, |
| "step": 1960 |
| }, |
| { |
| "entropy": 0.3571629270911217, |
| "epoch": 3.5056192277734506, |
| "grad_norm": 2.078125, |
| "learning_rate": 1.7846666666666666e-05, |
| "loss": 0.3299, |
| "mean_token_accuracy": 0.8891440749168396, |
| "num_tokens": 52743275.0, |
| "step": 1970 |
| }, |
| { |
| "entropy": 0.3572821389883757, |
| "epoch": 3.523422721709135, |
| "grad_norm": 2.375, |
| "learning_rate": 1.7824444444444447e-05, |
| "loss": 0.3279, |
| "mean_token_accuracy": 0.8893665768206119, |
| "num_tokens": 53014827.0, |
| "step": 1980 |
| }, |
| { |
| "entropy": 0.3552156362682581, |
| "epoch": 3.54122621564482, |
| "grad_norm": 2.640625, |
| "learning_rate": 1.780222222222222e-05, |
| "loss": 0.3302, |
| "mean_token_accuracy": 0.8896852925419807, |
| "num_tokens": 53285577.0, |
| "step": 1990 |
| }, |
| { |
| "entropy": 0.346796066313982, |
| "epoch": 3.5590297095805052, |
| "grad_norm": 2.34375, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 0.3239, |
| "mean_token_accuracy": 0.8911258917301893, |
| "num_tokens": 53553391.0, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.5590297095805052, |
| "eval_biology_entropy": 0.6294043908119201, |
| "eval_biology_loss": 1.5500138998031616, |
| "eval_biology_mean_token_accuracy": 0.687289656162262, |
| "eval_biology_num_tokens": 53553391.0, |
| "eval_biology_runtime": 27.4265, |
| "eval_biology_samples_per_second": 18.231, |
| "eval_biology_steps_per_second": 4.558, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.5590297095805052, |
| "eval_chemistry_entropy": 0.4015175952911377, |
| "eval_chemistry_loss": 0.7834813594818115, |
| "eval_chemistry_mean_token_accuracy": 0.8115957312583923, |
| "eval_chemistry_num_tokens": 53553391.0, |
| "eval_chemistry_runtime": 34.3329, |
| "eval_chemistry_samples_per_second": 14.563, |
| "eval_chemistry_steps_per_second": 3.641, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.5590297095805052, |
| "eval_math_entropy": 0.43964423608779907, |
| "eval_math_loss": 1.3351186513900757, |
| "eval_math_mean_token_accuracy": 0.7527407507896423, |
| "eval_math_num_tokens": 53553391.0, |
| "eval_math_runtime": 35.5354, |
| "eval_math_samples_per_second": 14.07, |
| "eval_math_steps_per_second": 3.518, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.5590297095805052, |
| "eval_physics_entropy": 0.43375833916664125, |
| "eval_physics_loss": 1.015561580657959, |
| "eval_physics_mean_token_accuracy": 0.7846694207191467, |
| "eval_physics_num_tokens": 53553391.0, |
| "eval_physics_runtime": 41.2328, |
| "eval_physics_samples_per_second": 12.126, |
| "eval_physics_steps_per_second": 3.032, |
| "step": 2000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 18, |
| "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": 4.16464083921255e+18, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
