| { | |
| "best_metric": 0.6025658845901489, | |
| "best_model_checkpoint": "BERT-WMM/run-4/checkpoint-268", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 536, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7352112676056338, | |
| "eval_loss": 0.632646918296814, | |
| "eval_runtime": 1.9849, | |
| "eval_samples_per_second": 1073.124, | |
| "eval_steps_per_second": 67.511, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7539906103286385, | |
| "eval_loss": 0.6025658845901489, | |
| "eval_runtime": 2.0385, | |
| "eval_samples_per_second": 1044.892, | |
| "eval_steps_per_second": 65.735, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.760093896713615, | |
| "eval_loss": 0.6695208549499512, | |
| "eval_runtime": 2.0168, | |
| "eval_samples_per_second": 1056.124, | |
| "eval_steps_per_second": 66.442, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 11.244783401489258, | |
| "learning_rate": 1.9838092768709567e-06, | |
| "loss": 0.4285, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.7637132406234741, | |
| "eval_runtime": 2.0114, | |
| "eval_samples_per_second": 1058.984, | |
| "eval_steps_per_second": 66.622, | |
| "step": 536 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 536, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 709335274032504.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 2.9536715900078686e-05, | |
| "num_train_epochs": 4, | |
| "per_device_train_batch_size": 64, | |
| "weight_decay": 0.0002947113945627026 | |
| } | |
| } | |