| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.861635220125786, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6212903225806452, | |
| "eval_loss": 0.2233715057373047, | |
| "eval_runtime": 7.114, | |
| "eval_samples_per_second": 435.763, | |
| "eval_steps_per_second": 9.137, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 1.650593990216632e-05, | |
| "loss": 0.3549, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8422580645161291, | |
| "eval_loss": 0.10387030988931656, | |
| "eval_runtime": 5.636, | |
| "eval_samples_per_second": 550.034, | |
| "eval_steps_per_second": 11.533, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8970967741935484, | |
| "eval_loss": 0.06669718772172928, | |
| "eval_runtime": 6.0164, | |
| "eval_samples_per_second": 515.259, | |
| "eval_steps_per_second": 10.804, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 1.3011879804332637e-05, | |
| "loss": 0.1201, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9051612903225806, | |
| "eval_loss": 0.04942993074655533, | |
| "eval_runtime": 5.9806, | |
| "eval_samples_per_second": 518.343, | |
| "eval_steps_per_second": 10.868, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 9.517819706498952e-06, | |
| "loss": 0.0738, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9151612903225806, | |
| "eval_loss": 0.04068068042397499, | |
| "eval_runtime": 6.3498, | |
| "eval_samples_per_second": 488.208, | |
| "eval_steps_per_second": 10.237, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9180645161290323, | |
| "eval_loss": 0.035583361983299255, | |
| "eval_runtime": 6.1405, | |
| "eval_samples_per_second": 504.845, | |
| "eval_steps_per_second": 10.585, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 6.02375960866527e-06, | |
| "loss": 0.0576, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9203225806451613, | |
| "eval_loss": 0.03254067152738571, | |
| "eval_runtime": 5.5714, | |
| "eval_samples_per_second": 556.415, | |
| "eval_steps_per_second": 11.667, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 2.5296995108315863e-06, | |
| "loss": 0.0502, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2862, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 826126757366148.0, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.774428978900597, | |
| "num_train_epochs": 9, | |
| "temperature": 6 | |
| } | |
| } | |