| { | |
| "best_metric": 0.8624554253693326, | |
| "best_model_checkpoint": "./outputs/electra-base-uncased/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": 8.69429684732554e-05, | |
| "loss": 0.5068, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8221090168110036, | |
| "eval_loss": 0.47897759079933167, | |
| "eval_runtime": 24.4089, | |
| "eval_samples_per_second": 402.107, | |
| "eval_steps_per_second": 50.268, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 6.520722635494156e-05, | |
| "loss": 0.3939, | |
| "step": 24544 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8475802343352012, | |
| "eval_loss": 0.4636411666870117, | |
| "eval_runtime": 24.3906, | |
| "eval_samples_per_second": 402.409, | |
| "eval_steps_per_second": 50.306, | |
| "step": 24544 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.34714842366277e-05, | |
| "loss": 0.2818, | |
| "step": 36816 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8509424350483953, | |
| "eval_loss": 0.43753835558891296, | |
| "eval_runtime": 24.3809, | |
| "eval_samples_per_second": 402.569, | |
| "eval_steps_per_second": 50.326, | |
| "step": 36816 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 2.173574211831385e-05, | |
| "loss": 0.1747, | |
| "step": 49088 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8599083036169128, | |
| "eval_loss": 0.5222671031951904, | |
| "eval_runtime": 24.4161, | |
| "eval_samples_per_second": 401.989, | |
| "eval_steps_per_second": 50.254, | |
| "step": 49088 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0934, | |
| "step": 61360 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8624554253693326, | |
| "eval_loss": 0.6794368624687195, | |
| "eval_runtime": 24.4311, | |
| "eval_samples_per_second": 401.742, | |
| "eval_steps_per_second": 50.223, | |
| "step": 61360 | |
| } | |
| ], | |
| "max_steps": 61360, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.6510003421181696e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |