| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9995002498750625, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24987506246876562, | |
| "grad_norm": 1.422733187675476, | |
| "learning_rate": 4.5835415625520576e-05, | |
| "loss": 6.5696, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.49975012493753124, | |
| "grad_norm": 1.82273268699646, | |
| "learning_rate": 4.167083125104115e-05, | |
| "loss": 4.7727, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7496251874062968, | |
| "grad_norm": 2.662141799926758, | |
| "learning_rate": 3.7506246876561725e-05, | |
| "loss": 4.4216, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9995002498750625, | |
| "grad_norm": 2.524129629135132, | |
| "learning_rate": 3.334166250208229e-05, | |
| "loss": 4.178, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6003, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 270991884288000.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |