|
{ |
|
"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": 5.162445545196533, |
|
"learning_rate": 4.75e-05, |
|
"loss": 0.5622, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.7192982456140351, |
|
"eval_f1": 0.6505255255255256, |
|
"eval_loss": 0.5075693130493164, |
|
"eval_precision": 0.6564625850340136, |
|
"eval_recall": 0.6463902527732315, |
|
"eval_runtime": 5.145, |
|
"eval_samples_per_second": 77.551, |
|
"eval_steps_per_second": 9.718, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 3.832602024078369, |
|
"learning_rate": 4.5e-05, |
|
"loss": 0.5003, |
|
"step": 244 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.7468671679197995, |
|
"eval_f1": 0.7178277095863939, |
|
"eval_loss": 0.48389074206352234, |
|
"eval_precision": 0.7106679035250464, |
|
"eval_recall": 0.7409074377159484, |
|
"eval_runtime": 5.0628, |
|
"eval_samples_per_second": 78.81, |
|
"eval_steps_per_second": 9.876, |
|
"step": 244 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 4.173563480377197, |
|
"learning_rate": 4.25e-05, |
|
"loss": 0.4614, |
|
"step": 366 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.7719298245614035, |
|
"eval_f1": 0.7343911163780806, |
|
"eval_loss": 0.425637811422348, |
|
"eval_precision": 0.7280813755532857, |
|
"eval_recall": 0.7436352064011639, |
|
"eval_runtime": 5.0694, |
|
"eval_samples_per_second": 78.708, |
|
"eval_steps_per_second": 9.863, |
|
"step": 366 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 2.580731153488159, |
|
"learning_rate": 4e-05, |
|
"loss": 0.4022, |
|
"step": 488 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.8170426065162907, |
|
"eval_f1": 0.777617444284111, |
|
"eval_loss": 0.387967973947525, |
|
"eval_precision": 0.7798245614035089, |
|
"eval_recall": 0.7755501000181851, |
|
"eval_runtime": 5.0609, |
|
"eval_samples_per_second": 78.84, |
|
"eval_steps_per_second": 9.88, |
|
"step": 488 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"grad_norm": 1.3641459941864014, |
|
"learning_rate": 3.7500000000000003e-05, |
|
"loss": 0.3678, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.8020050125313283, |
|
"eval_f1": 0.7759781102306244, |
|
"eval_loss": 0.4130999743938446, |
|
"eval_precision": 0.765661378526179, |
|
"eval_recall": 0.7974177123113293, |
|
"eval_runtime": 5.1051, |
|
"eval_samples_per_second": 78.158, |
|
"eval_steps_per_second": 9.794, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"grad_norm": 2.4856526851654053, |
|
"learning_rate": 3.5e-05, |
|
"loss": 0.3376, |
|
"step": 732 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.8320802005012531, |
|
"eval_f1": 0.7999041923338897, |
|
"eval_loss": 0.36453139781951904, |
|
"eval_precision": 0.7965023376930815, |
|
"eval_recall": 0.803691580287325, |
|
"eval_runtime": 5.0825, |
|
"eval_samples_per_second": 78.504, |
|
"eval_steps_per_second": 9.838, |
|
"step": 732 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"grad_norm": 0.8218249678611755, |
|
"learning_rate": 3.2500000000000004e-05, |
|
"loss": 0.3268, |
|
"step": 854 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.8345864661654135, |
|
"eval_f1": 0.8069354838709677, |
|
"eval_loss": 0.3639658987522125, |
|
"eval_precision": 0.798820781588242, |
|
"eval_recall": 0.8179669030732861, |
|
"eval_runtime": 5.0648, |
|
"eval_samples_per_second": 78.779, |
|
"eval_steps_per_second": 9.872, |
|
"step": 854 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"grad_norm": 3.0957868099212646, |
|
"learning_rate": 3e-05, |
|
"loss": 0.3044, |
|
"step": 976 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.8345864661654135, |
|
"eval_f1": 0.8024159663865547, |
|
"eval_loss": 0.3551309108734131, |
|
"eval_precision": 0.7996165051429931, |
|
"eval_recall": 0.8054646299327151, |
|
"eval_runtime": 5.1012, |
|
"eval_samples_per_second": 78.216, |
|
"eval_steps_per_second": 9.802, |
|
"step": 976 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"grad_norm": 5.1089277267456055, |
|
"learning_rate": 2.7500000000000004e-05, |
|
"loss": 0.2984, |
|
"step": 1098 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.849624060150376, |
|
"eval_f1": 0.821236559139785, |
|
"eval_loss": 0.3508576452732086, |
|
"eval_precision": 0.8169406150583245, |
|
"eval_recall": 0.8261047463175123, |
|
"eval_runtime": 5.0536, |
|
"eval_samples_per_second": 78.953, |
|
"eval_steps_per_second": 9.894, |
|
"step": 1098 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 5.958853244781494, |
|
"learning_rate": 2.5e-05, |
|
"loss": 0.2922, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.8521303258145363, |
|
"eval_f1": 0.8220702165451041, |
|
"eval_loss": 0.341298907995224, |
|
"eval_precision": 0.8212799324446589, |
|
"eval_recall": 0.8228768867066739, |
|
"eval_runtime": 5.0649, |
|
"eval_samples_per_second": 78.778, |
|
"eval_steps_per_second": 9.872, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"grad_norm": 3.4268646240234375, |
|
"learning_rate": 2.25e-05, |
|
"loss": 0.2666, |
|
"step": 1342 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.8521303258145363, |
|
"eval_f1": 0.8254476698921143, |
|
"eval_loss": 0.3493994474411011, |
|
"eval_precision": 0.8192918192918193, |
|
"eval_recall": 0.8328787052191308, |
|
"eval_runtime": 5.0474, |
|
"eval_samples_per_second": 79.051, |
|
"eval_steps_per_second": 9.906, |
|
"step": 1342 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"grad_norm": 10.23725414276123, |
|
"learning_rate": 2e-05, |
|
"loss": 0.2641, |
|
"step": 1464 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.8546365914786967, |
|
"eval_f1": 0.8288009705864946, |
|
"eval_loss": 0.35203850269317627, |
|
"eval_precision": 0.822024085224641, |
|
"eval_recall": 0.837152209492635, |
|
"eval_runtime": 5.0872, |
|
"eval_samples_per_second": 78.432, |
|
"eval_steps_per_second": 9.829, |
|
"step": 1464 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"grad_norm": 1.6404447555541992, |
|
"learning_rate": 1.75e-05, |
|
"loss": 0.2694, |
|
"step": 1586 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.849624060150376, |
|
"eval_f1": 0.8167483159828537, |
|
"eval_loss": 0.33577361702919006, |
|
"eval_precision": 0.8201621387462095, |
|
"eval_recall": 0.8136024731769412, |
|
"eval_runtime": 5.0577, |
|
"eval_samples_per_second": 78.89, |
|
"eval_steps_per_second": 9.886, |
|
"step": 1586 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"grad_norm": 10.325481414794922, |
|
"learning_rate": 1.5e-05, |
|
"loss": 0.2678, |
|
"step": 1708 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.8646616541353384, |
|
"eval_f1": 0.8383403361344538, |
|
"eval_loss": 0.3354940712451935, |
|
"eval_precision": 0.8352012604792199, |
|
"eval_recall": 0.8417439534460811, |
|
"eval_runtime": 5.0671, |
|
"eval_samples_per_second": 78.744, |
|
"eval_steps_per_second": 9.868, |
|
"step": 1708 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"grad_norm": 0.9968625903129578, |
|
"learning_rate": 1.25e-05, |
|
"loss": 0.255, |
|
"step": 1830 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.8646616541353384, |
|
"eval_f1": 0.8391129032258065, |
|
"eval_loss": 0.3406221270561218, |
|
"eval_precision": 0.8345705196182396, |
|
"eval_recall": 0.8442444080741953, |
|
"eval_runtime": 5.0556, |
|
"eval_samples_per_second": 78.923, |
|
"eval_steps_per_second": 9.89, |
|
"step": 1830 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"grad_norm": 6.620565414428711, |
|
"learning_rate": 1e-05, |
|
"loss": 0.2482, |
|
"step": 1952 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.8621553884711779, |
|
"eval_f1": 0.8372817261706151, |
|
"eval_loss": 0.337022989988327, |
|
"eval_precision": 0.8308913308913308, |
|
"eval_recall": 0.8449718130569195, |
|
"eval_runtime": 5.0566, |
|
"eval_samples_per_second": 78.906, |
|
"eval_steps_per_second": 9.888, |
|
"step": 1952 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"grad_norm": 7.714181423187256, |
|
"learning_rate": 7.5e-06, |
|
"loss": 0.2444, |
|
"step": 2074 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_accuracy": 0.8696741854636592, |
|
"eval_f1": 0.844327731092437, |
|
"eval_loss": 0.3271777033805847, |
|
"eval_precision": 0.8411320530352577, |
|
"eval_recall": 0.8477905073649754, |
|
"eval_runtime": 5.0845, |
|
"eval_samples_per_second": 78.474, |
|
"eval_steps_per_second": 9.834, |
|
"step": 2074 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"grad_norm": 4.410248756408691, |
|
"learning_rate": 5e-06, |
|
"loss": 0.2521, |
|
"step": 2196 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.8621553884711779, |
|
"eval_f1": 0.8365204824303285, |
|
"eval_loss": 0.3318663537502289, |
|
"eval_precision": 0.8313636363636363, |
|
"eval_recall": 0.8424713584288053, |
|
"eval_runtime": 5.0547, |
|
"eval_samples_per_second": 78.936, |
|
"eval_steps_per_second": 9.892, |
|
"step": 2196 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"grad_norm": 3.304914951324463, |
|
"learning_rate": 2.5e-06, |
|
"loss": 0.2456, |
|
"step": 2318 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"eval_accuracy": 0.8696741854636592, |
|
"eval_f1": 0.844327731092437, |
|
"eval_loss": 0.3293265402317047, |
|
"eval_precision": 0.8411320530352577, |
|
"eval_recall": 0.8477905073649754, |
|
"eval_runtime": 5.0713, |
|
"eval_samples_per_second": 78.678, |
|
"eval_steps_per_second": 9.859, |
|
"step": 2318 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"grad_norm": 8.615702629089355, |
|
"learning_rate": 0.0, |
|
"loss": 0.2458, |
|
"step": 2440 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.8671679197994987, |
|
"eval_f1": 0.8417152566223307, |
|
"eval_loss": 0.33090195059776306, |
|
"eval_precision": 0.8378262413446174, |
|
"eval_recall": 0.8460174577195854, |
|
"eval_runtime": 5.0518, |
|
"eval_samples_per_second": 78.981, |
|
"eval_steps_per_second": 9.897, |
|
"step": 2440 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"step": 2440, |
|
"total_flos": 7609911792720000.0, |
|
"train_loss": 0.3206107397548488, |
|
"train_runtime": 1955.0616, |
|
"train_samples_per_second": 37.216, |
|
"train_steps_per_second": 1.248 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 2440, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 20, |
|
"save_steps": 500, |
|
"total_flos": 7609911792720000.0, |
|
"train_batch_size": 30, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|