| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "global_step": 6650, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 4.62406015037594e-05, | |
| "loss": 3.1912, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 4.24812030075188e-05, | |
| "loss": 2.4395, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 3.87218045112782e-05, | |
| "loss": 2.2379, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 3.49624060150376e-05, | |
| "loss": 2.0996, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 3.120300751879699e-05, | |
| "loss": 1.9879, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "learning_rate": 2.7443609022556393e-05, | |
| "loss": 1.924, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "learning_rate": 2.368421052631579e-05, | |
| "loss": 1.8705, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "learning_rate": 1.9924812030075188e-05, | |
| "loss": 1.822, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.77, | |
| "learning_rate": 1.6165413533834585e-05, | |
| "loss": 1.7756, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.52, | |
| "learning_rate": 1.2406015037593984e-05, | |
| "loss": 1.7552, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.27, | |
| "learning_rate": 8.646616541353383e-06, | |
| "loss": 1.7259, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "learning_rate": 4.887218045112782e-06, | |
| "loss": 1.7117, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 9.77, | |
| "learning_rate": 1.1278195488721805e-06, | |
| "loss": 1.699, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 6650, | |
| "total_flos": 3.398700317184e+16, | |
| "train_loss": 2.0113733139611725, | |
| "train_runtime": 2782.8401, | |
| "train_samples_per_second": 23.875, | |
| "train_steps_per_second": 2.39 | |
| } | |
| ], | |
| "max_steps": 6650, | |
| "num_train_epochs": 10, | |
| "total_flos": 3.398700317184e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |