| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9365853658536586, | |
| "eval_steps": 500, | |
| "global_step": 3, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3121951219512195, | |
| "grad_norm": 3.576138715967291, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7047, | |
| "mean_token_accuracy": 0.794938562437892, | |
| "num_tokens": 524288.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.624390243902439, | |
| "grad_norm": 3.416216144709776, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6693, | |
| "mean_token_accuracy": 0.7996451482176781, | |
| "num_tokens": 1048576.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.9365853658536586, | |
| "grad_norm": 2.6744297925211464, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.6693, | |
| "mean_token_accuracy": 0.7972947089001536, | |
| "num_tokens": 1570323.0, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.9365853658536586, | |
| "step": 3, | |
| "total_flos": 3124062060544.0, | |
| "train_loss": 0.6811053355534872, | |
| "train_runtime": 330.2876, | |
| "train_samples_per_second": 1.238, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 3, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3124062060544.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |