| { | |
| "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": 0.3132377564907074, | |
| "learning_rate": 0.0001724256975247876, | |
| "loss": 0.1837, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9041935483870968, | |
| "eval_loss": 0.04824049770832062, | |
| "eval_runtime": 5.4515, | |
| "eval_samples_per_second": 568.653, | |
| "eval_steps_per_second": 11.923, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.22080405056476593, | |
| "learning_rate": 0.0001599503839610332, | |
| "loss": 0.0431, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.927741935483871, | |
| "eval_loss": 0.03218682482838631, | |
| "eval_runtime": 5.4844, | |
| "eval_samples_per_second": 565.235, | |
| "eval_steps_per_second": 11.852, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.1997973769903183, | |
| "learning_rate": 0.00014051111924417916, | |
| "loss": 0.0295, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9345161290322581, | |
| "eval_loss": 0.02600180357694626, | |
| "eval_runtime": 5.461, | |
| "eval_samples_per_second": 567.657, | |
| "eval_steps_per_second": 11.902, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.11037289351224899, | |
| "learning_rate": 0.00011599890306959969, | |
| "loss": 0.0232, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9416129032258065, | |
| "eval_loss": 0.0218922421336174, | |
| "eval_runtime": 5.4181, | |
| "eval_samples_per_second": 572.161, | |
| "eval_steps_per_second": 11.997, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.09720052033662796, | |
| "learning_rate": 8.879821828652982e-05, | |
| "loss": 0.02, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9448387096774193, | |
| "eval_loss": 0.01982319913804531, | |
| "eval_runtime": 5.4788, | |
| "eval_samples_per_second": 565.817, | |
| "eval_steps_per_second": 11.864, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.07554798573255539, | |
| "learning_rate": 6.155507478206978e-05, | |
| "loss": 0.0179, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.944516129032258, | |
| "eval_loss": 0.019010677933692932, | |
| "eval_runtime": 5.4951, | |
| "eval_samples_per_second": 564.142, | |
| "eval_steps_per_second": 11.829, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.06709519028663635, | |
| "learning_rate": 3.691961271421151e-05, | |
| "loss": 0.0164, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9458064516129032, | |
| "eval_loss": 0.018042678013443947, | |
| "eval_runtime": 5.4272, | |
| "eval_samples_per_second": 571.201, | |
| "eval_steps_per_second": 11.977, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.06841150671243668, | |
| "learning_rate": 1.728830396315599e-05, | |
| "loss": 0.0154, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9461290322580646, | |
| "eval_loss": 0.01746629737317562, | |
| "eval_runtime": 5.4329, | |
| "eval_samples_per_second": 570.601, | |
| "eval_steps_per_second": 11.964, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.07178815454244614, | |
| "learning_rate": 4.570829754441392e-06, | |
| "loss": 0.0149, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9461290322580646, | |
| "eval_loss": 0.017176177352666855, | |
| "eval_runtime": 5.4823, | |
| "eval_samples_per_second": 565.452, | |
| "eval_steps_per_second": 11.856, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.05784892663359642, | |
| "learning_rate": 4.311975386746314e-09, | |
| "loss": 0.0147, | |
| "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": 825404033099184.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.8577901310898406, | |
| "learning_rate": 0.00017672349478818126, | |
| "lr_scheduler_type": "cosine", | |
| "num_train_epochs": 10, | |
| "temperature": 14.978266913116322, | |
| "weight_decay": 0.28559191608322254 | |
| } | |
| } | |