| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "eval_steps": 500, | |
| "global_step": 7900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 6.33, | |
| "learning_rate": 4.683544303797468e-05, | |
| "loss": 1.1738, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "learning_rate": 4.367088607594937e-05, | |
| "loss": 0.1263, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "learning_rate": 4.050632911392405e-05, | |
| "loss": 0.0548, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "learning_rate": 3.7341772151898736e-05, | |
| "loss": 0.0291, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 31.65, | |
| "learning_rate": 3.4177215189873416e-05, | |
| "loss": 0.0192, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 37.97, | |
| "learning_rate": 3.10126582278481e-05, | |
| "loss": 0.0152, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 44.3, | |
| "learning_rate": 2.7848101265822786e-05, | |
| "loss": 0.0102, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 50.63, | |
| "learning_rate": 2.468354430379747e-05, | |
| "loss": 0.006, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 56.96, | |
| "learning_rate": 2.1518987341772153e-05, | |
| "loss": 0.0053, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 63.29, | |
| "learning_rate": 1.8354430379746836e-05, | |
| "loss": 0.0042, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 69.62, | |
| "learning_rate": 1.5189873417721521e-05, | |
| "loss": 0.0034, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 75.95, | |
| "learning_rate": 1.2025316455696203e-05, | |
| "loss": 0.003, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 82.28, | |
| "learning_rate": 8.860759493670886e-06, | |
| "loss": 0.0026, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 88.61, | |
| "learning_rate": 5.69620253164557e-06, | |
| "loss": 0.0025, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 94.94, | |
| "learning_rate": 2.531645569620253e-06, | |
| "loss": 0.0019, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 7900, | |
| "total_flos": 4.8994912512e+16, | |
| "train_loss": 0.09234336149843433, | |
| "train_runtime": 3031.3949, | |
| "train_samples_per_second": 164.941, | |
| "train_steps_per_second": 2.606 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 7900, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 4.8994912512e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |