| { | |
| "best_metric": 0.5995926260948181, | |
| "best_model_checkpoint": "hBERTv2_new_pretrain_48_mrpc/checkpoint-87", | |
| "epoch": 8.0, | |
| "global_step": 232, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 3.9200000000000004e-05, | |
| "loss": 0.6634, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6862745098039216, | |
| "eval_combined_score": 0.7371769900012985, | |
| "eval_f1": 0.7880794701986754, | |
| "eval_loss": 0.6017037034034729, | |
| "eval_runtime": 0.668, | |
| "eval_samples_per_second": 610.794, | |
| "eval_steps_per_second": 5.988, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.8400000000000005e-05, | |
| "loss": 0.6054, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6691176470588235, | |
| "eval_combined_score": 0.7003639925154952, | |
| "eval_f1": 0.731610337972167, | |
| "eval_loss": 0.6601347327232361, | |
| "eval_runtime": 0.6859, | |
| "eval_samples_per_second": 594.842, | |
| "eval_steps_per_second": 5.832, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.76e-05, | |
| "loss": 0.5623, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6936274509803921, | |
| "eval_combined_score": 0.7513938781619518, | |
| "eval_f1": 0.8091603053435115, | |
| "eval_loss": 0.5995926260948181, | |
| "eval_runtime": 0.6727, | |
| "eval_samples_per_second": 606.513, | |
| "eval_steps_per_second": 5.946, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.680000000000001e-05, | |
| "loss": 0.4773, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7009803921568627, | |
| "eval_combined_score": 0.753356438116648, | |
| "eval_f1": 0.8057324840764333, | |
| "eval_loss": 0.6380115747451782, | |
| "eval_runtime": 0.6759, | |
| "eval_samples_per_second": 603.635, | |
| "eval_steps_per_second": 5.918, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.3781, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6470588235294118, | |
| "eval_combined_score": 0.6930946291560103, | |
| "eval_f1": 0.7391304347826086, | |
| "eval_loss": 0.8475584387779236, | |
| "eval_runtime": 0.6849, | |
| "eval_samples_per_second": 595.667, | |
| "eval_steps_per_second": 5.84, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.52e-05, | |
| "loss": 0.258, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6642156862745098, | |
| "eval_combined_score": 0.7077884238995054, | |
| "eval_f1": 0.7513611615245008, | |
| "eval_loss": 0.8256893754005432, | |
| "eval_runtime": 0.6757, | |
| "eval_samples_per_second": 603.855, | |
| "eval_steps_per_second": 5.92, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3.44e-05, | |
| "loss": 0.2236, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6495098039215687, | |
| "eval_combined_score": 0.6973039215686274, | |
| "eval_f1": 0.7450980392156862, | |
| "eval_loss": 1.1872520446777344, | |
| "eval_runtime": 0.6794, | |
| "eval_samples_per_second": 600.55, | |
| "eval_steps_per_second": 5.888, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 3.3600000000000004e-05, | |
| "loss": 0.1818, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6029411764705882, | |
| "eval_combined_score": 0.6468904355635383, | |
| "eval_f1": 0.6908396946564885, | |
| "eval_loss": 1.2389172315597534, | |
| "eval_runtime": 0.6785, | |
| "eval_samples_per_second": 601.284, | |
| "eval_steps_per_second": 5.895, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 232, | |
| "total_flos": 4294176753582080.0, | |
| "train_loss": 0.4187329986999775, | |
| "train_runtime": 172.7626, | |
| "train_samples_per_second": 1061.572, | |
| "train_steps_per_second": 8.393 | |
| } | |
| ], | |
| "max_steps": 1450, | |
| "num_train_epochs": 50, | |
| "total_flos": 4294176753582080.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |