| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 48, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.041666666666666664, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 3.886, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 2e-05, | |
| "loss": 3.5195, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.9340161087325483e-05, | |
| "loss": 3.6848, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 1.744772182743782e-05, | |
| "loss": 3.5055, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 3.0, | |
| "learning_rate": 1.4572423233046386e-05, | |
| "loss": 3.4931, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 5.596227169036865, | |
| "eval_runtime": 0.0423, | |
| "eval_samples_per_second": 283.93, | |
| "eval_steps_per_second": 47.322, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.0416666666666667, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 1.1093712083778748e-05, | |
| "loss": 3.4722, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 7.470666176083193e-06, | |
| "loss": 3.4108, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.4583333333333333, | |
| "grad_norm": 2.5, | |
| "learning_rate": 4.181410844420473e-06, | |
| "loss": 3.4187, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 1.660021821101222e-06, | |
| "loss": 3.4852, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 2.392412244407294e-07, | |
| "loss": 3.4278, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 5.484647274017334, | |
| "eval_runtime": 0.0423, | |
| "eval_samples_per_second": 283.999, | |
| "eval_steps_per_second": 47.333, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 48, | |
| "total_flos": 394600362541056.0, | |
| "train_loss": 3.4869212259848914, | |
| "train_runtime": 14.4504, | |
| "train_samples_per_second": 102.281, | |
| "train_steps_per_second": 3.322 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 48, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 1000000000, | |
| "total_flos": 394600362541056.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |