| { | |
| "best_metric": 0.7299432626496387, | |
| "best_model_checkpoint": "robertuito-finetuned-class/run-0/checkpoint-2088", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 2088, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8374, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7242765273311897, | |
| "eval_f1": 0.72858215075172, | |
| "eval_loss": 0.75840163230896, | |
| "eval_recall": 0.7666197221538275, | |
| "eval_runtime": 2.775, | |
| "eval_samples_per_second": 448.282, | |
| "eval_steps_per_second": 14.054, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4823, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7564308681672026, | |
| "eval_f1": 0.7049375903197748, | |
| "eval_loss": 0.9376571774482727, | |
| "eval_recall": 0.6838350828130171, | |
| "eval_runtime": 2.7021, | |
| "eval_samples_per_second": 460.38, | |
| "eval_steps_per_second": 14.433, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3293, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1491, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7564308681672026, | |
| "eval_f1": 0.7299432626496387, | |
| "eval_loss": 1.3851027488708496, | |
| "eval_recall": 0.7322288045696094, | |
| "eval_runtime": 2.6853, | |
| "eval_samples_per_second": 463.259, | |
| "eval_steps_per_second": 14.523, | |
| "step": 2088 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2088, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 621962468524032.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 8.740829729953022e-05, | |
| "num_train_epochs": 3, | |
| "per_device_train_batch_size": 8, | |
| "seed": 26 | |
| } | |
| } | |