| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1590, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6690322580645162, | |
| "eval_loss": 0.42948082089424133, | |
| "eval_runtime": 5.393, | |
| "eval_samples_per_second": 574.82, | |
| "eval_steps_per_second": 12.053, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.5723270440251573, | |
| "grad_norm": 0.8646999001502991, | |
| "learning_rate": 1.371069182389937e-05, | |
| "loss": 0.6591, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8229032258064516, | |
| "eval_loss": 0.16900987923145294, | |
| "eval_runtime": 5.2956, | |
| "eval_samples_per_second": 585.388, | |
| "eval_steps_per_second": 12.274, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8729032258064516, | |
| "eval_loss": 0.0963619202375412, | |
| "eval_runtime": 5.3142, | |
| "eval_samples_per_second": 583.338, | |
| "eval_steps_per_second": 12.231, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.1446540880503147, | |
| "grad_norm": 0.7748064398765564, | |
| "learning_rate": 7.421383647798742e-06, | |
| "loss": 0.1958, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8977419354838709, | |
| "eval_loss": 0.07352616637945175, | |
| "eval_runtime": 5.3982, | |
| "eval_samples_per_second": 574.264, | |
| "eval_steps_per_second": 12.041, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.716981132075472, | |
| "grad_norm": 0.542826771736145, | |
| "learning_rate": 1.1320754716981133e-06, | |
| "loss": 0.113, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1590, | |
| "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": 391368939443328.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.4868927644085208, | |
| "num_train_epochs": 5, | |
| "temperature": 2 | |
| } | |
| } | |