| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 5.0, |
| "global_step": 1230, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 2.03, |
| "learning_rate": 1.7804878048780488e-05, |
| "loss": 0.4863, |
| "step": 500 |
| }, |
| { |
| "epoch": 4.07, |
| "learning_rate": 5.609756097560976e-06, |
| "loss": 0.1397, |
| "step": 1000 |
| }, |
| { |
| "epoch": 5.0, |
| "step": 1230, |
| "total_flos": 2588026118284800.0, |
| "train_loss": 0.26801115826862615, |
| "train_runtime": 238.2502, |
| "train_samples_per_second": 165.141, |
| "train_steps_per_second": 5.163 |
| } |
| ], |
| "max_steps": 1230, |
| "num_train_epochs": 5, |
| "total_flos": 2588026118284800.0, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
