| { | |
| "best_metric": 0.839429444727458, | |
| "best_model_checkpoint": "./outputs/bert-base-cased/fine_tuned_models/checkpoint-61360", | |
| "epoch": 5.0, | |
| "global_step": 61360, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.7388593694651082e-05, | |
| "loss": 0.5897, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.822414671421294, | |
| "eval_loss": 0.45840176939964294, | |
| "eval_runtime": 73.2673, | |
| "eval_samples_per_second": 133.961, | |
| "eval_steps_per_second": 8.38, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.3041445270988312e-05, | |
| "loss": 0.3759, | |
| "step": 24544 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8327050433010698, | |
| "eval_loss": 0.45654305815696716, | |
| "eval_runtime": 73.2832, | |
| "eval_samples_per_second": 133.932, | |
| "eval_steps_per_second": 8.378, | |
| "step": 24544 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 8.694296847325541e-06, | |
| "loss": 0.2523, | |
| "step": 36816 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8353540499235863, | |
| "eval_loss": 0.4860617220401764, | |
| "eval_runtime": 73.3093, | |
| "eval_samples_per_second": 133.885, | |
| "eval_steps_per_second": 8.375, | |
| "step": 36816 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.3471484236627705e-06, | |
| "loss": 0.1672, | |
| "step": 49088 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8374936321956189, | |
| "eval_loss": 0.5898149609565735, | |
| "eval_runtime": 73.2627, | |
| "eval_samples_per_second": 133.97, | |
| "eval_steps_per_second": 8.381, | |
| "step": 49088 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.1151, | |
| "step": 61360 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.839429444727458, | |
| "eval_loss": 0.7346773743629456, | |
| "eval_runtime": 73.2837, | |
| "eval_samples_per_second": 133.932, | |
| "eval_steps_per_second": 8.378, | |
| "step": 61360 | |
| } | |
| ], | |
| "max_steps": 61360, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.6333273497007872e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |