| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "global_step": 4128, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 3.7832929782082326e-05, | |
| "loss": 1.8019, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5.920654149000606e-05, | |
| "loss": 1.3315, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 4.974258025439128e-05, | |
| "loss": 0.9419, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 4.0278619018776494e-05, | |
| "loss": 0.6627, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.49071434140205383, | |
| "eval_runtime": 1039.783, | |
| "eval_samples_per_second": 3.505, | |
| "eval_steps_per_second": 0.351, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 3.081465778316172e-05, | |
| "loss": 0.3682, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 2.135069654754694e-05, | |
| "loss": 0.2958, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 1.1886735311932163e-05, | |
| "loss": 0.2752, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 2.4227740763173836e-06, | |
| "loss": 0.246, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.32644888758659363, | |
| "eval_runtime": 1038.0256, | |
| "eval_samples_per_second": 3.511, | |
| "eval_steps_per_second": 0.352, | |
| "step": 4128 | |
| } | |
| ], | |
| "max_steps": 4128, | |
| "num_train_epochs": 2, | |
| "total_flos": 7.01816537088e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |