| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1175, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.312751594516966, | |
| "las": 67.21955477957225, | |
| "step": 235, | |
| "uas": 76.90964644260148 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.2455898308091693, | |
| "las": 70.40593627237014, | |
| "step": 470, | |
| "uas": 79.92143168921868 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 3.060515880584717, | |
| "learning_rate": 0.00011489361702127661, | |
| "loss": 0.8516, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.2293221056461334, | |
| "las": 73.41772151898735, | |
| "step": 705, | |
| "uas": 81.92928852029682 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.4363478405608072, | |
| "las": 73.46137058053252, | |
| "step": 940, | |
| "uas": 81.71104321257093 | |
| }, | |
| { | |
| "epoch": 4.25531914893617, | |
| "grad_norm": 3.269766092300415, | |
| "learning_rate": 2.9787234042553192e-05, | |
| "loss": 0.2066, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1175, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.288547626897152e+16, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |