| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 200, | |
| "global_step": 1505, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013289036544850499, | |
| "grad_norm": 0.10525072365999222, | |
| "learning_rate": 4.9978216198586135e-05, | |
| "loss": 0.6155, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.026578073089700997, | |
| "grad_norm": 0.08554615080356598, | |
| "learning_rate": 4.991290275706486e-05, | |
| "loss": 0.5694, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03986710963455149, | |
| "grad_norm": 0.08361516892910004, | |
| "learning_rate": 4.980417349743936e-05, | |
| "loss": 0.557, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.053156146179401995, | |
| "grad_norm": 0.08680060505867004, | |
| "learning_rate": 4.9652217902637596e-05, | |
| "loss": 0.548, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0664451827242525, | |
| "grad_norm": 0.08960291743278503, | |
| "learning_rate": 4.945730078629964e-05, | |
| "loss": 0.5427, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07973421926910298, | |
| "grad_norm": 0.09262242168188095, | |
| "learning_rate": 4.921976183128585e-05, | |
| "loss": 0.5384, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09302325581395349, | |
| "grad_norm": 0.08780515193939209, | |
| "learning_rate": 4.894001499771015e-05, | |
| "loss": 0.5362, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.10631229235880399, | |
| "grad_norm": 0.09249912202358246, | |
| "learning_rate": 4.861854780153004e-05, | |
| "loss": 0.5324, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11960132890365449, | |
| "grad_norm": 0.09562400728464127, | |
| "learning_rate": 4.825592046495054e-05, | |
| "loss": 0.5311, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.132890365448505, | |
| "grad_norm": 0.09372778236865997, | |
| "learning_rate": 4.785276494012263e-05, | |
| "loss": 0.5278, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.132890365448505, | |
| "eval_accuracy": 0.19452303794312395, | |
| "eval_loss": 0.5592088103294373, | |
| "eval_runtime": 19.5284, | |
| "eval_samples_per_second": 93.914, | |
| "eval_steps_per_second": 0.41, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1461794019933555, | |
| "grad_norm": 0.08762918412685394, | |
| "learning_rate": 4.740978380783765e-05, | |
| "loss": 0.5253, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.15946843853820597, | |
| "grad_norm": 0.08518578112125397, | |
| "learning_rate": 4.6927749053136866e-05, | |
| "loss": 0.5192, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.17275747508305647, | |
| "grad_norm": 0.09664598107337952, | |
| "learning_rate": 4.640750071996995e-05, | |
| "loss": 0.5217, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.18604651162790697, | |
| "grad_norm": 0.08245342969894409, | |
| "learning_rate": 4.584994544724695e-05, | |
| "loss": 0.5172, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.19933554817275748, | |
| "grad_norm": 0.08551981300115585, | |
| "learning_rate": 4.5256054888834934e-05, | |
| "loss": 0.5152, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.21262458471760798, | |
| "grad_norm": 0.09647104889154434, | |
| "learning_rate": 4.4626864020252774e-05, | |
| "loss": 0.5139, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.22591362126245848, | |
| "grad_norm": 0.09810427576303482, | |
| "learning_rate": 4.3963469335015085e-05, | |
| "loss": 0.5129, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.23920265780730898, | |
| "grad_norm": 0.08342389762401581, | |
| "learning_rate": 4.326702693376844e-05, | |
| "loss": 0.5119, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.25249169435215946, | |
| "grad_norm": 0.08738644421100616, | |
| "learning_rate": 4.2538750509550054e-05, | |
| "loss": 0.511, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.26578073089701, | |
| "grad_norm": 0.08475251495838165, | |
| "learning_rate": 4.177990923267986e-05, | |
| "loss": 0.5117, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.26578073089701, | |
| "eval_accuracy": 0.1953402564276045, | |
| "eval_loss": 0.5438870787620544, | |
| "eval_runtime": 15.5302, | |
| "eval_samples_per_second": 118.093, | |
| "eval_steps_per_second": 0.515, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27906976744186046, | |
| "grad_norm": 0.07873477786779404, | |
| "learning_rate": 4.099182553897229e-05, | |
| "loss": 0.5084, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.292358803986711, | |
| "grad_norm": 0.09158772230148315, | |
| "learning_rate": 4.017587282512181e-05, | |
| "loss": 0.5065, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.30564784053156147, | |
| "grad_norm": 0.07729614526033401, | |
| "learning_rate": 3.933347305527898e-05, | |
| "loss": 0.5047, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.31893687707641194, | |
| "grad_norm": 0.08530613034963608, | |
| "learning_rate": 3.846609428298757e-05, | |
| "loss": 0.5049, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.33222591362126247, | |
| "grad_norm": 0.07760792225599289, | |
| "learning_rate": 3.7575248092801686e-05, | |
| "loss": 0.5035, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34551495016611294, | |
| "grad_norm": 0.08521712571382523, | |
| "learning_rate": 3.66624869660411e-05, | |
| "loss": 0.5042, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.3588039867109635, | |
| "grad_norm": 0.08439727872610092, | |
| "learning_rate": 3.572940157527572e-05, | |
| "loss": 0.5021, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.37209302325581395, | |
| "grad_norm": 0.09042590111494064, | |
| "learning_rate": 3.47776180122539e-05, | |
| "loss": 0.5019, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3853820598006645, | |
| "grad_norm": 0.08219762146472931, | |
| "learning_rate": 3.3808794954105716e-05, | |
| "loss": 0.501, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.39867109634551495, | |
| "grad_norm": 0.08426713198423386, | |
| "learning_rate": 3.282462077275947e-05, | |
| "loss": 0.5013, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.39867109634551495, | |
| "eval_accuracy": 0.19588631180347973, | |
| "eval_loss": 0.5341373682022095, | |
| "eval_runtime": 16.1072, | |
| "eval_samples_per_second": 113.862, | |
| "eval_steps_per_second": 0.497, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4119601328903654, | |
| "grad_norm": 0.08020314574241638, | |
| "learning_rate": 3.1826810592609036e-05, | |
| "loss": 0.4968, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.42524916943521596, | |
| "grad_norm": 0.07975760847330093, | |
| "learning_rate": 3.081710330155942e-05, | |
| "loss": 0.4997, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.43853820598006643, | |
| "grad_norm": 0.08056964725255966, | |
| "learning_rate": 2.979725852065981e-05, | |
| "loss": 0.4968, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.45182724252491696, | |
| "grad_norm": 0.08022565394639969, | |
| "learning_rate": 2.876905353760459e-05, | |
| "loss": 0.4976, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 0.08131925761699677, | |
| "learning_rate": 2.7734280209446865e-05, | |
| "loss": 0.4973, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.47840531561461797, | |
| "grad_norm": 0.07562076300382614, | |
| "learning_rate": 2.6694741839921732e-05, | |
| "loss": 0.4956, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.49169435215946844, | |
| "grad_norm": 0.07877329736948013, | |
| "learning_rate": 2.5652250036821523e-05, | |
| "loss": 0.4966, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5049833887043189, | |
| "grad_norm": 0.08014395087957382, | |
| "learning_rate": 2.4608621554899362e-05, | |
| "loss": 0.4934, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.5182724252491694, | |
| "grad_norm": 0.07770328223705292, | |
| "learning_rate": 2.356567512980326e-05, | |
| "loss": 0.4934, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.53156146179402, | |
| "grad_norm": 0.07732851803302765, | |
| "learning_rate": 2.252522830855798e-05, | |
| "loss": 0.4951, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.53156146179402, | |
| "eval_accuracy": 0.19623978277118043, | |
| "eval_loss": 0.5274041295051575, | |
| "eval_runtime": 16.4552, | |
| "eval_samples_per_second": 111.454, | |
| "eval_steps_per_second": 0.486, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5448504983388704, | |
| "grad_norm": 0.07608461380004883, | |
| "learning_rate": 2.1489094282118395e-05, | |
| "loss": 0.4896, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.5581395348837209, | |
| "grad_norm": 0.07657533138990402, | |
| "learning_rate": 2.0459078725514092e-05, | |
| "loss": 0.4918, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.07983728498220444, | |
| "learning_rate": 1.9436976651092144e-05, | |
| "loss": 0.4927, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.584717607973422, | |
| "grad_norm": 0.07355430722236633, | |
| "learning_rate": 1.8424569280341653e-05, | |
| "loss": 0.493, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5980066445182725, | |
| "grad_norm": 0.08014149218797684, | |
| "learning_rate": 1.7423620939751788e-05, | |
| "loss": 0.4922, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6112956810631229, | |
| "grad_norm": 0.07500924915075302, | |
| "learning_rate": 1.6435875986112685e-05, | |
| "loss": 0.491, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.6245847176079734, | |
| "grad_norm": 0.07356715947389603, | |
| "learning_rate": 1.546305576661776e-05, | |
| "loss": 0.4909, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.6378737541528239, | |
| "grad_norm": 0.07140863686800003, | |
| "learning_rate": 1.4506855619064846e-05, | |
| "loss": 0.489, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.6511627906976745, | |
| "grad_norm": 0.07692987471818924, | |
| "learning_rate": 1.3568941917384036e-05, | |
| "loss": 0.4902, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.6644518272425249, | |
| "grad_norm": 0.07356040179729462, | |
| "learning_rate": 1.2650949167640997e-05, | |
| "loss": 0.4894, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6644518272425249, | |
| "eval_accuracy": 0.19652373156663552, | |
| "eval_loss": 0.5229406952857971, | |
| "eval_runtime": 15.6791, | |
| "eval_samples_per_second": 116.971, | |
| "eval_steps_per_second": 0.51, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6777408637873754, | |
| "grad_norm": 0.0691773071885109, | |
| "learning_rate": 1.1754477159576499e-05, | |
| "loss": 0.4869, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.6910299003322259, | |
| "grad_norm": 0.07505939155817032, | |
| "learning_rate": 1.088108817864629e-05, | |
| "loss": 0.4865, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.7043189368770764, | |
| "grad_norm": 0.06973451375961304, | |
| "learning_rate": 1.003230428341979e-05, | |
| "loss": 0.4888, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.717607973421927, | |
| "grad_norm": 0.07225219160318375, | |
| "learning_rate": 9.209604653082326e-06, | |
| "loss": 0.4858, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.7308970099667774, | |
| "grad_norm": 0.07558443397283554, | |
| "learning_rate": 8.414423009663563e-06, | |
| "loss": 0.4891, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7441860465116279, | |
| "grad_norm": 0.0698658898472786, | |
| "learning_rate": 7.648145119484152e-06, | |
| "loss": 0.4871, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.7574750830564784, | |
| "grad_norm": 0.06963298469781876, | |
| "learning_rate": 6.912106378175098e-06, | |
| "loss": 0.4884, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.770764119601329, | |
| "grad_norm": 0.0692787617444992, | |
| "learning_rate": 6.207589483478266e-06, | |
| "loss": 0.4877, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.7840531561461794, | |
| "grad_norm": 0.07016126066446304, | |
| "learning_rate": 5.53582219988382e-06, | |
| "loss": 0.4856, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.7973421926910299, | |
| "grad_norm": 0.06945677101612091, | |
| "learning_rate": 4.897975218999926e-06, | |
| "loss": 0.4868, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7973421926910299, | |
| "eval_accuracy": 0.19665158843513314, | |
| "eval_loss": 0.5205041170120239, | |
| "eval_runtime": 14.8321, | |
| "eval_samples_per_second": 123.651, | |
| "eval_steps_per_second": 0.539, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8106312292358804, | |
| "grad_norm": 0.07045505195856094, | |
| "learning_rate": 4.295160119383712e-06, | |
| "loss": 0.4859, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.8239202657807309, | |
| "grad_norm": 0.06839559227228165, | |
| "learning_rate": 3.728427429388709e-06, | |
| "loss": 0.4863, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.8372093023255814, | |
| "grad_norm": 0.06684821844100952, | |
| "learning_rate": 3.198764796404807e-06, | |
| "loss": 0.4856, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.8504983388704319, | |
| "grad_norm": 0.06731660664081573, | |
| "learning_rate": 2.707095265681081e-06, | |
| "loss": 0.4854, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.8637873754152824, | |
| "grad_norm": 0.06780705600976944, | |
| "learning_rate": 2.254275671731007e-06, | |
| "loss": 0.4868, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8770764119601329, | |
| "grad_norm": 0.06815515458583832, | |
| "learning_rate": 1.8410951451234533e-06, | |
| "loss": 0.4854, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.8903654485049833, | |
| "grad_norm": 0.0670180469751358, | |
| "learning_rate": 1.4682737372615967e-06, | |
| "loss": 0.485, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.9036544850498339, | |
| "grad_norm": 0.06649608910083771, | |
| "learning_rate": 1.1364611655463736e-06, | |
| "loss": 0.4867, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.9169435215946844, | |
| "grad_norm": 0.0674930214881897, | |
| "learning_rate": 8.462356811112987e-07, | |
| "loss": 0.4865, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 0.06808231770992279, | |
| "learning_rate": 5.981030611018234e-07, | |
| "loss": 0.4864, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "eval_accuracy": 0.19667556159797644, | |
| "eval_loss": 0.519675612449646, | |
| "eval_runtime": 14.9507, | |
| "eval_samples_per_second": 122.67, | |
| "eval_steps_per_second": 0.535, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9435215946843853, | |
| "grad_norm": 0.06696037203073502, | |
| "learning_rate": 3.9249572725543196e-07, | |
| "loss": 0.4852, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.9568106312292359, | |
| "grad_norm": 0.06675516068935394, | |
| "learning_rate": 2.297719923185032e-07, | |
| "loss": 0.4875, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.9700996677740864, | |
| "grad_norm": 0.06678403913974762, | |
| "learning_rate": 1.1021543561322012e-07, | |
| "loss": 0.4852, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.9833887043189369, | |
| "grad_norm": 0.0660882443189621, | |
| "learning_rate": 3.403440884269526e-08, | |
| "loss": 0.4848, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.9966777408637874, | |
| "grad_norm": 0.06698651611804962, | |
| "learning_rate": 1.3616729956228425e-09, | |
| "loss": 0.4847, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1505, | |
| "total_flos": 2.786803439690685e+19, | |
| "train_loss": 0.0, | |
| "train_runtime": 4.5361, | |
| "train_samples_per_second": 339673.082, | |
| "train_steps_per_second": 331.781 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 1505, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.786803439690685e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |