| { | |
| "best_metric": 0.9606727878443162, | |
| "best_model_checkpoint": "./output//roberta_0_100_0.000008_8_0.01_0.000008_01-26-22_23-25/checkpoint-8000", | |
| "epoch": 0.6081800212863008, | |
| "global_step": 8000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2.0271126314455844e-06, | |
| "loss": 0.3799, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "eval_accuracy": 0.9360740357434579, | |
| "eval_combined_score": 116916.96803701787, | |
| "eval_loss": 0.16771526634693146, | |
| "eval_number_of_examples": 233833, | |
| "eval_runtime": 565.9226, | |
| "eval_samples_per_second": 413.189, | |
| "eval_steps_per_second": 2.583, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.054225262891169e-06, | |
| "loss": 0.162, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "eval_accuracy": 0.9526029260198517, | |
| "eval_combined_score": 116916.97630146302, | |
| "eval_loss": 0.12519623339176178, | |
| "eval_number_of_examples": 233833, | |
| "eval_runtime": 566.1477, | |
| "eval_samples_per_second": 413.025, | |
| "eval_steps_per_second": 2.582, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 6.081337894336754e-06, | |
| "loss": 0.127, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_accuracy": 0.958962165306009, | |
| "eval_combined_score": 116916.97948108266, | |
| "eval_loss": 0.11369086056947708, | |
| "eval_number_of_examples": 233833, | |
| "eval_runtime": 566.6337, | |
| "eval_samples_per_second": 412.67, | |
| "eval_steps_per_second": 2.58, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 7.993077066164161e-06, | |
| "loss": 0.1168, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "eval_accuracy": 0.9606727878443162, | |
| "eval_combined_score": 116916.98033639393, | |
| "eval_loss": 0.10579771548509598, | |
| "eval_number_of_examples": 233833, | |
| "eval_runtime": 566.6673, | |
| "eval_samples_per_second": 412.646, | |
| "eval_steps_per_second": 2.58, | |
| "step": 8000 | |
| } | |
| ], | |
| "max_steps": 131540, | |
| "num_train_epochs": 10, | |
| "total_flos": 1.1928721221631345e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |