| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.5641188621521, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.5535, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7243107769423559, | |
| "eval_f1": 0.6401043033324587, | |
| "eval_loss": 0.5041041374206543, | |
| "eval_precision": 0.6583725987676694, | |
| "eval_recall": 0.6324331696672122, | |
| "eval_runtime": 1.8214, | |
| "eval_samples_per_second": 219.06, | |
| "eval_steps_per_second": 27.451, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.286947011947632, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4636, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7669172932330827, | |
| "eval_f1": 0.7331142070096449, | |
| "eval_loss": 0.4692240059375763, | |
| "eval_precision": 0.7252895752895754, | |
| "eval_recall": 0.7475904709947263, | |
| "eval_runtime": 1.8328, | |
| "eval_samples_per_second": 217.703, | |
| "eval_steps_per_second": 27.281, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.800495147705078, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.4023, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8370927318295739, | |
| "eval_f1": 0.7809488416091623, | |
| "eval_loss": 0.36048197746276855, | |
| "eval_precision": 0.832562695924765, | |
| "eval_recall": 0.7572285870158211, | |
| "eval_runtime": 1.8261, | |
| "eval_samples_per_second": 218.495, | |
| "eval_steps_per_second": 27.38, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.0396647453308105, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3202, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8546365914786967, | |
| "eval_f1": 0.8159125620465827, | |
| "eval_loss": 0.3256481885910034, | |
| "eval_precision": 0.8356565656565657, | |
| "eval_recall": 0.8021458446990362, | |
| "eval_runtime": 1.8317, | |
| "eval_samples_per_second": 217.825, | |
| "eval_steps_per_second": 27.296, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.8746237754821777, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.2919, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8771929824561403, | |
| "eval_f1": 0.8475258334958082, | |
| "eval_loss": 0.3067488968372345, | |
| "eval_precision": 0.8591828192414193, | |
| "eval_recall": 0.8381069285324605, | |
| "eval_runtime": 1.8315, | |
| "eval_samples_per_second": 217.857, | |
| "eval_steps_per_second": 27.3, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.942033290863037, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.2657, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8521303258145363, | |
| "eval_f1": 0.8320383569853806, | |
| "eval_loss": 0.3400041460990906, | |
| "eval_precision": 0.8193218954248366, | |
| "eval_recall": 0.8553827968721586, | |
| "eval_runtime": 1.8306, | |
| "eval_samples_per_second": 217.958, | |
| "eval_steps_per_second": 27.313, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.39800548553466797, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.2559, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.87468671679198, | |
| "eval_f1": 0.8524146298159436, | |
| "eval_loss": 0.2993007302284241, | |
| "eval_precision": 0.8451250578971746, | |
| "eval_recall": 0.8613384251682124, | |
| "eval_runtime": 1.8316, | |
| "eval_samples_per_second": 217.847, | |
| "eval_steps_per_second": 27.299, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 7.434815406799316, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2369, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.8584001703456596, | |
| "eval_loss": 0.30177775025367737, | |
| "eval_precision": 0.8759655377302435, | |
| "eval_recall": 0.8451991271140207, | |
| "eval_runtime": 1.8332, | |
| "eval_samples_per_second": 217.658, | |
| "eval_steps_per_second": 27.275, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 6.123136043548584, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.2178, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8847117794486216, | |
| "eval_f1": 0.8595070422535211, | |
| "eval_loss": 0.29259544610977173, | |
| "eval_precision": 0.8633733523114054, | |
| "eval_recall": 0.8559283506092017, | |
| "eval_runtime": 1.8408, | |
| "eval_samples_per_second": 216.759, | |
| "eval_steps_per_second": 27.163, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.8655648231506348, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2118, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.8622036668943447, | |
| "eval_loss": 0.29553094506263733, | |
| "eval_precision": 0.8671602787456446, | |
| "eval_recall": 0.8577014002545917, | |
| "eval_runtime": 1.8333, | |
| "eval_samples_per_second": 217.636, | |
| "eval_steps_per_second": 27.273, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 3.4139134883880615, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.2034, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8847117794486216, | |
| "eval_f1": 0.8572517421602788, | |
| "eval_loss": 0.2934134602546692, | |
| "eval_precision": 0.8679426449878376, | |
| "eval_recall": 0.8484269867248591, | |
| "eval_runtime": 1.8401, | |
| "eval_samples_per_second": 216.837, | |
| "eval_steps_per_second": 27.173, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 5.207653045654297, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1856, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8533986527862829, | |
| "eval_loss": 0.297758549451828, | |
| "eval_precision": 0.8572003218020917, | |
| "eval_recall": 0.8498817966903074, | |
| "eval_runtime": 1.8319, | |
| "eval_samples_per_second": 217.802, | |
| "eval_steps_per_second": 27.293, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.1223020553588867, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1775, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8493865995596099, | |
| "eval_loss": 0.3038978576660156, | |
| "eval_precision": 0.8651108632904749, | |
| "eval_recall": 0.8373795235497363, | |
| "eval_runtime": 1.8374, | |
| "eval_samples_per_second": 217.149, | |
| "eval_steps_per_second": 27.212, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 7.688318252563477, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1719, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.8622036668943447, | |
| "eval_loss": 0.3036334812641144, | |
| "eval_precision": 0.8671602787456446, | |
| "eval_recall": 0.8577014002545917, | |
| "eval_runtime": 1.8354, | |
| "eval_samples_per_second": 217.392, | |
| "eval_steps_per_second": 27.242, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.8339260816574097, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1621, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8822055137844611, | |
| "eval_f1": 0.8596342841745197, | |
| "eval_loss": 0.299029141664505, | |
| "eval_precision": 0.8555364857667042, | |
| "eval_recall": 0.8641571194762684, | |
| "eval_runtime": 1.8346, | |
| "eval_samples_per_second": 217.487, | |
| "eval_steps_per_second": 27.254, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.4192665815353394, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1535, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8847117794486216, | |
| "eval_f1": 0.8616171059774413, | |
| "eval_loss": 0.3039585053920746, | |
| "eval_precision": 0.859873949579832, | |
| "eval_recall": 0.8634297144935443, | |
| "eval_runtime": 1.8339, | |
| "eval_samples_per_second": 217.57, | |
| "eval_steps_per_second": 27.264, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.2236837148666382, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1504, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8510452961672474, | |
| "eval_loss": 0.31895172595977783, | |
| "eval_precision": 0.8615574190244527, | |
| "eval_recall": 0.8423804328059648, | |
| "eval_runtime": 1.8322, | |
| "eval_samples_per_second": 217.776, | |
| "eval_steps_per_second": 27.29, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.8861066102981567, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1459, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8771929824561403, | |
| "eval_f1": 0.8522278069611882, | |
| "eval_loss": 0.31010520458221436, | |
| "eval_precision": 0.8513631702756499, | |
| "eval_recall": 0.8531096563011457, | |
| "eval_runtime": 1.8303, | |
| "eval_samples_per_second": 217.993, | |
| "eval_steps_per_second": 27.317, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 2.291304349899292, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1444, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.8822055137844611, | |
| "eval_f1": 0.855319904024935, | |
| "eval_loss": 0.3119203448295593, | |
| "eval_precision": 0.862378106322743, | |
| "eval_recall": 0.8491543917075832, | |
| "eval_runtime": 1.8361, | |
| "eval_samples_per_second": 217.305, | |
| "eval_steps_per_second": 27.231, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.763808250427246, | |
| "learning_rate": 0.0, | |
| "loss": 0.1384, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8847117794486216, | |
| "eval_f1": 0.8609292598654301, | |
| "eval_loss": 0.3090469241142273, | |
| "eval_precision": 0.8609292598654301, | |
| "eval_recall": 0.8609292598654301, | |
| "eval_runtime": 1.8381, | |
| "eval_samples_per_second": 217.076, | |
| "eval_steps_per_second": 27.203, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 8444128359504000.0, | |
| "train_loss": 0.242632052937492, | |
| "train_runtime": 627.3825, | |
| "train_samples_per_second": 115.974, | |
| "train_steps_per_second": 3.889 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 8444128359504000.0, | |
| "train_batch_size": 30, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |