| { | |
| "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": 0.26659590005874634, | |
| "learning_rate": 0.00016775241359960757, | |
| "loss": 0.1876, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9, | |
| "eval_loss": 0.04595751687884331, | |
| "eval_runtime": 5.5039, | |
| "eval_samples_per_second": 563.242, | |
| "eval_steps_per_second": 11.81, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.1902252733707428, | |
| "learning_rate": 0.00015561522064766448, | |
| "loss": 0.0427, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9351612903225807, | |
| "eval_loss": 0.030528072267770767, | |
| "eval_runtime": 5.4415, | |
| "eval_samples_per_second": 569.694, | |
| "eval_steps_per_second": 11.945, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.21384550631046295, | |
| "learning_rate": 0.00013670282173226981, | |
| "loss": 0.0291, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9364516129032258, | |
| "eval_loss": 0.02449769712984562, | |
| "eval_runtime": 5.4302, | |
| "eval_samples_per_second": 570.878, | |
| "eval_steps_per_second": 11.97, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.10294859856367111, | |
| "learning_rate": 0.00011285496445235415, | |
| "loss": 0.023, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9435483870967742, | |
| "eval_loss": 0.021032005548477173, | |
| "eval_runtime": 5.5078, | |
| "eval_samples_per_second": 562.834, | |
| "eval_steps_per_second": 11.801, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.06512714922428131, | |
| "learning_rate": 8.639150460022785e-05, | |
| "loss": 0.0201, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9490322580645161, | |
| "eval_loss": 0.01880320906639099, | |
| "eval_runtime": 5.4592, | |
| "eval_samples_per_second": 567.844, | |
| "eval_steps_per_second": 11.906, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.09834893047809601, | |
| "learning_rate": 5.988673679288492e-05, | |
| "loss": 0.018, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9480645161290323, | |
| "eval_loss": 0.018181176856160164, | |
| "eval_runtime": 5.4267, | |
| "eval_samples_per_second": 571.246, | |
| "eval_steps_per_second": 11.978, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.09281986206769943, | |
| "learning_rate": 3.591897397475451e-05, | |
| "loss": 0.0167, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9474193548387096, | |
| "eval_loss": 0.017256448045372963, | |
| "eval_runtime": 5.5033, | |
| "eval_samples_per_second": 563.302, | |
| "eval_steps_per_second": 11.811, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.0637020543217659, | |
| "learning_rate": 1.681973602830377e-05, | |
| "loss": 0.0158, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9483870967741935, | |
| "eval_loss": 0.01661858893930912, | |
| "eval_runtime": 5.4493, | |
| "eval_samples_per_second": 568.879, | |
| "eval_steps_per_second": 11.928, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.0715932622551918, | |
| "learning_rate": 4.446945753838205e-06, | |
| "loss": 0.0152, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9487096774193549, | |
| "eval_loss": 0.016348928213119507, | |
| "eval_runtime": 5.423, | |
| "eval_samples_per_second": 571.636, | |
| "eval_steps_per_second": 11.986, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.057582587003707886, | |
| "learning_rate": 4.195107161476372e-09, | |
| "loss": 0.0149, | |
| "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": 0.2541194692185785, | |
| "learning_rate": 0.00017193372690989549, | |
| "lr_scheduler_type": "cosine", | |
| "num_train_epochs": 10, | |
| "temperature": 11.572807433105293, | |
| "weight_decay": 0.16868839944734676 | |
| } | |
| } | |