| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 3180, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.4123988151550293, | |
| "learning_rate": 0.00012162031188058806, | |
| "loss": 0.2161, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8945161290322581, | |
| "eval_loss": 0.04954709857702255, | |
| "eval_runtime": 5.4811, | |
| "eval_samples_per_second": 565.584, | |
| "eval_steps_per_second": 11.859, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.2431638091802597, | |
| "learning_rate": 0.0001128208605910619, | |
| "loss": 0.0477, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9296774193548387, | |
| "eval_loss": 0.030576715245842934, | |
| "eval_runtime": 5.8135, | |
| "eval_samples_per_second": 533.243, | |
| "eval_steps_per_second": 11.181, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.1598445326089859, | |
| "learning_rate": 9.910939257015842e-05, | |
| "loss": 0.0317, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9406451612903226, | |
| "eval_loss": 0.024058301001787186, | |
| "eval_runtime": 5.4176, | |
| "eval_samples_per_second": 572.208, | |
| "eval_steps_per_second": 11.998, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.14414510130882263, | |
| "learning_rate": 8.181972276552748e-05, | |
| "loss": 0.0252, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9435483870967742, | |
| "eval_loss": 0.0203920416533947, | |
| "eval_runtime": 5.5058, | |
| "eval_samples_per_second": 563.045, | |
| "eval_steps_per_second": 11.806, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.10477118194103241, | |
| "learning_rate": 6.263374402702216e-05, | |
| "loss": 0.0221, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9454838709677419, | |
| "eval_loss": 0.018697841092944145, | |
| "eval_runtime": 5.4543, | |
| "eval_samples_per_second": 568.359, | |
| "eval_steps_per_second": 11.917, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.11606243997812271, | |
| "learning_rate": 4.3417817067273426e-05, | |
| "loss": 0.02, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9448387096774193, | |
| "eval_loss": 0.018037663772702217, | |
| "eval_runtime": 5.4037, | |
| "eval_samples_per_second": 573.685, | |
| "eval_steps_per_second": 12.029, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.07421741634607315, | |
| "learning_rate": 2.604121588179992e-05, | |
| "loss": 0.0186, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9448387096774193, | |
| "eval_loss": 0.01729719527065754, | |
| "eval_runtime": 5.4857, | |
| "eval_samples_per_second": 565.101, | |
| "eval_steps_per_second": 11.849, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.07452843338251114, | |
| "learning_rate": 1.2194289772747888e-05, | |
| "loss": 0.0176, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9470967741935484, | |
| "eval_loss": 0.01650388352572918, | |
| "eval_runtime": 5.5176, | |
| "eval_samples_per_second": 561.836, | |
| "eval_steps_per_second": 11.78, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.0748431384563446, | |
| "learning_rate": 3.2240306883973477e-06, | |
| "loss": 0.017, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9464516129032258, | |
| "eval_loss": 0.016214776784181595, | |
| "eval_runtime": 5.4173, | |
| "eval_samples_per_second": 572.241, | |
| "eval_steps_per_second": 11.999, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.06282064318656921, | |
| "learning_rate": 3.041447991138999e-09, | |
| "loss": 0.0168, | |
| "step": 3170 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 3180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 1000000000.0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 825254092458012.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.17478103003354462, | |
| "learning_rate": 0.00012465175934507248, | |
| "lr_scheduler_type": "cosine", | |
| "num_train_epochs": 10, | |
| "temperature": 6.90073840111308, | |
| "weight_decay": 0.25246423960377606 | |
| } | |
| } | |