| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 4182, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3586800573888092, | |
| "grad_norm": 8.820939064025879, | |
| "learning_rate": 1.760879961740794e-05, | |
| "loss": 2.101, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7173601147776184, | |
| "grad_norm": 7.0845489501953125, | |
| "learning_rate": 1.5217599234815878e-05, | |
| "loss": 1.8819, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.590270757675171, | |
| "eval_rouge1": 0.3916416859053997, | |
| "eval_rouge2": 0.19880729723807156, | |
| "eval_rougeL": 0.382750399398817, | |
| "eval_rougeLsum": 0.38311641208619757, | |
| "eval_runtime": 1067.7291, | |
| "eval_samples_per_second": 2.611, | |
| "eval_steps_per_second": 0.653, | |
| "step": 1394 | |
| }, | |
| { | |
| "epoch": 1.0760401721664274, | |
| "grad_norm": 7.022219181060791, | |
| "learning_rate": 1.2826398852223818e-05, | |
| "loss": 1.8208, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4347202295552366, | |
| "grad_norm": 8.181060791015625, | |
| "learning_rate": 1.0435198469631755e-05, | |
| "loss": 1.718, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.793400286944046, | |
| "grad_norm": 8.304183959960938, | |
| "learning_rate": 8.043998087039695e-06, | |
| "loss": 1.7516, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.5554497241973877, | |
| "eval_rouge1": 0.38234006744658555, | |
| "eval_rouge2": 0.19842659831002687, | |
| "eval_rougeL": 0.3723334371691044, | |
| "eval_rougeLsum": 0.3724996919585559, | |
| "eval_runtime": 1048.7275, | |
| "eval_samples_per_second": 2.658, | |
| "eval_steps_per_second": 0.665, | |
| "step": 2788 | |
| }, | |
| { | |
| "epoch": 2.152080344332855, | |
| "grad_norm": 5.35423469543457, | |
| "learning_rate": 5.652797704447634e-06, | |
| "loss": 1.6662, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.5107604017216643, | |
| "grad_norm": 5.020022392272949, | |
| "learning_rate": 3.261597321855572e-06, | |
| "loss": 1.6696, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.869440459110473, | |
| "grad_norm": 6.37124490737915, | |
| "learning_rate": 8.703969392635103e-07, | |
| "loss": 1.6634, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4182, | |
| "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": 2776798057512960.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |