| { | |
| "best_metric": 1.8468588590621948, | |
| "best_model_checkpoint": "output/max-korzh/checkpoint-66", | |
| "epoch": 1.0, | |
| "global_step": 66, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.00013526627358699495, | |
| "loss": 2.4173, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 0.00012957411177772776, | |
| "loss": 2.2555, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 0.00012044442080070213, | |
| "loss": 2.0536, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 0.0001083919039965842, | |
| "loss": 2.0403, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 9.409604445829847e-05, | |
| "loss": 1.9983, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 7.836279790554737e-05, | |
| "loss": 1.9702, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 6.207915542933309e-05, | |
| "loss": 1.9274, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 4.616313771642488e-05, | |
| "loss": 1.9929, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 3.151203992254603e-05, | |
| "loss": 1.927, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.8951844985992186e-05, | |
| "loss": 1.9614, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 9.190657300387505e-06, | |
| "loss": 1.8335, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2.778782010045488e-06, | |
| "loss": 1.8942, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 7.770053204565375e-08, | |
| "loss": 1.8146, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.8468588590621948, | |
| "eval_runtime": 4.2001, | |
| "eval_samples_per_second": 22.381, | |
| "eval_steps_per_second": 2.857, | |
| "step": 66 | |
| } | |
| ], | |
| "max_steps": 66, | |
| "num_train_epochs": 1, | |
| "total_flos": 68458512384000.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |