| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 4074, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.36818851251840945, | |
| "grad_norm": 14.467572212219238, | |
| "learning_rate": 2.6318114874815905e-05, | |
| "loss": 0.198, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7363770250368189, | |
| "grad_norm": 10.934176445007324, | |
| "learning_rate": 2.263622974963181e-05, | |
| "loss": 0.1377, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1045655375552283, | |
| "grad_norm": 0.5147254467010498, | |
| "learning_rate": 1.8954344624447718e-05, | |
| "loss": 0.1138, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4727540500736378, | |
| "grad_norm": 33.273719787597656, | |
| "learning_rate": 1.5272459499263623e-05, | |
| "loss": 0.0751, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.8409425625920472, | |
| "grad_norm": 7.599516868591309, | |
| "learning_rate": 1.1590574374079529e-05, | |
| "loss": 0.0707, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.2091310751104567, | |
| "grad_norm": 13.896529197692871, | |
| "learning_rate": 7.908689248895435e-06, | |
| "loss": 0.0569, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.5773195876288657, | |
| "grad_norm": 12.114886283874512, | |
| "learning_rate": 4.2268041237113405e-06, | |
| "loss": 0.024, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.9455081001472756, | |
| "grad_norm": 0.2183772623538971, | |
| "learning_rate": 5.44918998527246e-07, | |
| "loss": 0.0249, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 4074, | |
| "total_flos": 6.855770591553946e+16, | |
| "train_loss": 0.08668158495607596, | |
| "train_runtime": 5343.4587, | |
| "train_samples_per_second": 12.191, | |
| "train_steps_per_second": 0.762 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4074, | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.855770591553946e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |