| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "global_step": 14688, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 2.897875816993464e-05, | |
| "loss": 2.0084, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 2.795751633986928e-05, | |
| "loss": 1.5875, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 2.6936274509803924e-05, | |
| "loss": 1.5189, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5915032679738562e-05, | |
| "loss": 1.447, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 2.48937908496732e-05, | |
| "loss": 1.4006, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2.3872549019607843e-05, | |
| "loss": 1.35, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2.2851307189542485e-05, | |
| "loss": 1.3541, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 2.1830065359477124e-05, | |
| "loss": 1.0592, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 2.0808823529411766e-05, | |
| "loss": 0.9374, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 1.9787581699346405e-05, | |
| "loss": 0.9286, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 1.8766339869281047e-05, | |
| "loss": 0.9295, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 1.7745098039215686e-05, | |
| "loss": 0.9148, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 1.6723856209150328e-05, | |
| "loss": 0.9295, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 1.5702614379084967e-05, | |
| "loss": 0.8957, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.4681372549019607e-05, | |
| "loss": 0.7651, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 1.366013071895425e-05, | |
| "loss": 0.4515, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 1.2638888888888888e-05, | |
| "loss": 0.4702, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 1.161764705882353e-05, | |
| "loss": 0.4578, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 1.059640522875817e-05, | |
| "loss": 0.4478, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 9.575163398692811e-06, | |
| "loss": 0.4501, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 8.553921568627451e-06, | |
| "loss": 0.4619, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 7.532679738562092e-06, | |
| "loss": 0.4468, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "learning_rate": 6.511437908496732e-06, | |
| "loss": 0.1485, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 5.490196078431373e-06, | |
| "loss": 0.1502, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 4.468954248366013e-06, | |
| "loss": 0.1412, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 3.447712418300654e-06, | |
| "loss": 0.1327, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 2.4264705882352943e-06, | |
| "loss": 0.1521, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "learning_rate": 1.4052287581699345e-06, | |
| "loss": 0.1475, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 3.839869281045752e-07, | |
| "loss": 0.1381, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 14688, | |
| "total_flos": 3.682861839160443e+17, | |
| "train_loss": 0.7583477986664034, | |
| "train_runtime": 15779.9103, | |
| "train_samples_per_second": 33.507, | |
| "train_steps_per_second": 0.931 | |
| } | |
| ], | |
| "max_steps": 14688, | |
| "num_train_epochs": 4, | |
| "total_flos": 3.682861839160443e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |