| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.8666985188724317, | |
| "eval_steps": 500, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 24.413671493530273, | |
| "learning_rate": 4.601847427934385e-05, | |
| "loss": 1.7935, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 27.336660385131836, | |
| "learning_rate": 4.203694855868769e-05, | |
| "loss": 1.2168, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 20.986051559448242, | |
| "learning_rate": 3.805542283803154e-05, | |
| "loss": 0.9801, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 21.99080467224121, | |
| "learning_rate": 3.407389711737538e-05, | |
| "loss": 0.83, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 16.28312873840332, | |
| "learning_rate": 3.0092371396719225e-05, | |
| "loss": 0.5681, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 19.05654525756836, | |
| "learning_rate": 2.6110845676063068e-05, | |
| "loss": 0.4901, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 11.311978340148926, | |
| "learning_rate": 2.2129319955406913e-05, | |
| "loss": 0.4685, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 7.999639987945557, | |
| "learning_rate": 1.8147794234750756e-05, | |
| "loss": 0.4105, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 7.035294532775879, | |
| "learning_rate": 1.4166268514094602e-05, | |
| "loss": 0.2915, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 9.238717079162598, | |
| "learning_rate": 1.0184742793438446e-05, | |
| "loss": 0.2247, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 5.481063365936279, | |
| "learning_rate": 6.20321707278229e-06, | |
| "loss": 0.2151, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 5.940337181091309, | |
| "learning_rate": 2.221691352126135e-06, | |
| "loss": 0.1981, | |
| "step": 6000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6279, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 5.126075524128e+16, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |