| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "global_step": 6980, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 4.641833810888253e-05, | |
| "loss": 2.5428, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 4.2836676217765046e-05, | |
| "loss": 2.4266, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "learning_rate": 3.9255014326647564e-05, | |
| "loss": 2.3389, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "learning_rate": 3.567335243553009e-05, | |
| "loss": 2.2819, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.16, | |
| "learning_rate": 3.2091690544412614e-05, | |
| "loss": 2.2307, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 2.851002865329513e-05, | |
| "loss": 2.1831, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "learning_rate": 2.492836676217765e-05, | |
| "loss": 2.1536, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 11.46, | |
| "learning_rate": 2.1346704871060173e-05, | |
| "loss": 2.112, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 12.89, | |
| "learning_rate": 1.7765042979942695e-05, | |
| "loss": 2.0913, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 14.33, | |
| "learning_rate": 1.4183381088825215e-05, | |
| "loss": 2.0638, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 15.76, | |
| "learning_rate": 1.0601719197707736e-05, | |
| "loss": 2.0455, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 17.19, | |
| "learning_rate": 7.020057306590258e-06, | |
| "loss": 2.0315, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 18.62, | |
| "learning_rate": 3.4383954154727795e-06, | |
| "loss": 2.0204, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 6980, | |
| "total_flos": 1.819637710848e+16, | |
| "train_loss": 2.1814393838702095, | |
| "train_runtime": 6610.5408, | |
| "train_samples_per_second": 5.267, | |
| "train_steps_per_second": 1.056 | |
| } | |
| ], | |
| "max_steps": 6980, | |
| "num_train_epochs": 20, | |
| "total_flos": 1.819637710848e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |