| { | |
| "best_metric": 0.7203035354614258, | |
| "best_model_checkpoint": "outputs/checkpoint-613", | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 613, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.147441434337073e-05, | |
| "loss": 2.727, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 3.705117131325856e-05, | |
| "loss": 2.0215, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.031336501627827e-05, | |
| "loss": 1.4483, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.262792828314637e-05, | |
| "loss": 1.2309, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 4.442324303011218e-05, | |
| "loss": 1.0732, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 4.589012198616609e-05, | |
| "loss": 1.0297, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.713035047190436e-05, | |
| "loss": 0.9676, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 4.82046852530342e-05, | |
| "loss": 0.9311, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 4.915231568918581e-05, | |
| "loss": 0.9144, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8811, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 4.971822886716504e-05, | |
| "loss": 0.8284, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 4.943070730304773e-05, | |
| "loss": 0.8349, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.7203035354614258, | |
| "eval_runtime": 2.0613, | |
| "eval_samples_per_second": 568.098, | |
| "eval_steps_per_second": 17.95, | |
| "step": 613 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 9195, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "total_flos": 1018911101288448.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |