| { | |
| "best_metric": 0.5784130692481995, | |
| "best_model_checkpoint": "BERT-WMM/run-3/checkpoint-534", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 1068, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7549295774647887, | |
| "eval_loss": 0.603171169757843, | |
| "eval_runtime": 2.3997, | |
| "eval_samples_per_second": 887.602, | |
| "eval_steps_per_second": 55.84, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 9.50742244720459, | |
| "learning_rate": 8.206819746059557e-06, | |
| "loss": 0.6057, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7676056338028169, | |
| "eval_loss": 0.5784130692481995, | |
| "eval_runtime": 2.4488, | |
| "eval_samples_per_second": 869.828, | |
| "eval_steps_per_second": 54.722, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7657276995305164, | |
| "eval_loss": 0.6564009189605713, | |
| "eval_runtime": 2.4626, | |
| "eval_samples_per_second": 864.929, | |
| "eval_steps_per_second": 54.413, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 23.906795501708984, | |
| "learning_rate": 9.825065893169892e-07, | |
| "loss": 0.2903, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7544600938967136, | |
| "eval_loss": 0.7257726788520813, | |
| "eval_runtime": 2.4681, | |
| "eval_samples_per_second": 863.014, | |
| "eval_steps_per_second": 54.293, | |
| "step": 1068 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1068, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 697511614823352.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "num_train_epochs": 4 | |
| } | |
| } | |