| { | |
| "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": 0.2636558711528778, | |
| "learning_rate": 0.0002540605571670213, | |
| "loss": 0.1873, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.892258064516129, | |
| "eval_loss": 0.055641546845436096, | |
| "eval_runtime": 5.4336, | |
| "eval_samples_per_second": 570.529, | |
| "eval_steps_per_second": 11.963, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.591749370098114, | |
| "learning_rate": 0.00023567881268032667, | |
| "loss": 0.045, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9219354838709677, | |
| "eval_loss": 0.036715276539325714, | |
| "eval_runtime": 5.4851, | |
| "eval_samples_per_second": 565.168, | |
| "eval_steps_per_second": 11.85, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.3078596889972687, | |
| "learning_rate": 0.0002070360378748417, | |
| "loss": 0.0315, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.03158052638173103, | |
| "eval_runtime": 5.4416, | |
| "eval_samples_per_second": 569.683, | |
| "eval_steps_per_second": 11.945, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.07862970232963562, | |
| "learning_rate": 0.00017091852529921845, | |
| "loss": 0.0241, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9364516129032258, | |
| "eval_loss": 0.028497053310275078, | |
| "eval_runtime": 5.4242, | |
| "eval_samples_per_second": 571.509, | |
| "eval_steps_per_second": 11.983, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.09211583435535431, | |
| "learning_rate": 0.00013083969000659743, | |
| "loss": 0.0203, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9438709677419355, | |
| "eval_loss": 0.023725289851427078, | |
| "eval_runtime": 5.4755, | |
| "eval_samples_per_second": 566.158, | |
| "eval_steps_per_second": 11.871, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.09916875511407852, | |
| "learning_rate": 9.069829393233059e-05, | |
| "loss": 0.0179, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9419354838709677, | |
| "eval_loss": 0.023092513903975487, | |
| "eval_runtime": 5.4547, | |
| "eval_samples_per_second": 568.314, | |
| "eval_steps_per_second": 11.916, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.056765951216220856, | |
| "learning_rate": 5.4399184757334654e-05, | |
| "loss": 0.0164, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.944516129032258, | |
| "eval_loss": 0.02133980579674244, | |
| "eval_runtime": 5.4134, | |
| "eval_samples_per_second": 572.656, | |
| "eval_steps_per_second": 12.007, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.05946100503206253, | |
| "learning_rate": 2.547344276638816e-05, | |
| "loss": 0.0152, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9448387096774193, | |
| "eval_loss": 0.0208445955067873, | |
| "eval_runtime": 5.5038, | |
| "eval_samples_per_second": 563.245, | |
| "eval_steps_per_second": 11.81, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.0664161741733551, | |
| "learning_rate": 6.734886799353313e-06, | |
| "loss": 0.0146, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9432258064516129, | |
| "eval_loss": 0.020406901836395264, | |
| "eval_runtime": 5.4482, | |
| "eval_samples_per_second": 568.996, | |
| "eval_steps_per_second": 11.931, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.05120823532342911, | |
| "learning_rate": 6.353477961658025e-09, | |
| "loss": 0.0143, | |
| "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": 826195242364884.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.1846334135887684, | |
| "learning_rate": 0.0002603931443799669, | |
| "lr_scheduler_type": "cosine", | |
| "num_train_epochs": 10, | |
| "temperature": 8.818478491241954, | |
| "weight_decay": 0.1682665124233767 | |
| } | |
| } | |