| { | |
| "best_metric": 2.3443617820739746, | |
| "best_model_checkpoint": "output_dir/checkpoint-500", | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 842, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.406175771971497e-05, | |
| "loss": 3.5848, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 3.812351543942993e-05, | |
| "loss": 3.0455, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.21852731591449e-05, | |
| "loss": 2.8043, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2.6247030878859858e-05, | |
| "loss": 2.6995, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.0308788598574824e-05, | |
| "loss": 2.611, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_loss": 2.3443617820739746, | |
| "eval_runtime": 34.7863, | |
| "eval_samples_per_second": 128.988, | |
| "eval_steps_per_second": 8.078, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.4370546318289787e-05, | |
| "loss": 2.5726, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.432304038004752e-06, | |
| "loss": 2.4666, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2.494061757719715e-06, | |
| "loss": 2.4894, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 842, | |
| "total_flos": 1784293008629760.0, | |
| "train_loss": 2.7663106034883693, | |
| "train_runtime": 344.8745, | |
| "train_samples_per_second": 39.029, | |
| "train_steps_per_second": 2.441 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.3465805053710938, | |
| "eval_runtime": 34.7916, | |
| "eval_samples_per_second": 128.968, | |
| "eval_steps_per_second": 8.077, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.332395076751709, | |
| "eval_runtime": 34.8727, | |
| "eval_samples_per_second": 128.668, | |
| "eval_steps_per_second": 8.058, | |
| "step": 842 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 842, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 1784293008629760.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |