| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 2634, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 1.2870991230010986, | |
| "learning_rate": 1.6203492786636296e-05, | |
| "loss": 0.1607, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9892251037546519, | |
| "eval_f1": 0.932541011047874, | |
| "eval_loss": 0.03849802538752556, | |
| "eval_precision": 0.9266467065868264, | |
| "eval_recall": 0.938510781671159, | |
| "eval_runtime": 24.1011, | |
| "eval_samples_per_second": 134.849, | |
| "eval_steps_per_second": 8.464, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 1.1381593942642212, | |
| "learning_rate": 1.240698557327259e-05, | |
| "loss": 0.0444, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "grad_norm": 2.2290921211242676, | |
| "learning_rate": 8.610478359908885e-06, | |
| "loss": 0.0281, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9912320012470043, | |
| "eval_f1": 0.9497939618198639, | |
| "eval_loss": 0.03679182380437851, | |
| "eval_precision": 0.9482787573467675, | |
| "eval_recall": 0.9513140161725068, | |
| "eval_runtime": 24.0032, | |
| "eval_samples_per_second": 135.398, | |
| "eval_steps_per_second": 8.499, | |
| "step": 1756 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 2.8408632278442383, | |
| "learning_rate": 4.8139711465451785e-06, | |
| "loss": 0.019, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.847380410022779, | |
| "grad_norm": 0.38438597321510315, | |
| "learning_rate": 1.0174639331814731e-06, | |
| "loss": 0.0146, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2634, | |
| "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": 2751824963545344.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |