| { | |
| "best_metric": 0.598778486251831, | |
| "best_model_checkpoint": "BERT-WMM/run-6/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": 0.6478816270828247, | |
| "eval_runtime": 2.0455, | |
| "eval_samples_per_second": 1041.294, | |
| "eval_steps_per_second": 65.509, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7568075117370892, | |
| "eval_loss": 0.598778486251831, | |
| "eval_runtime": 2.5526, | |
| "eval_samples_per_second": 834.457, | |
| "eval_steps_per_second": 52.496, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7582159624413145, | |
| "eval_loss": 0.6211540102958679, | |
| "eval_runtime": 2.1567, | |
| "eval_samples_per_second": 987.64, | |
| "eval_steps_per_second": 62.133, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 9.620325088500977, | |
| "learning_rate": 1.3117656765629316e-06, | |
| "loss": 0.5199, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7633802816901408, | |
| "eval_loss": 0.6602327227592468, | |
| "eval_runtime": 2.0894, | |
| "eval_samples_per_second": 1019.414, | |
| "eval_steps_per_second": 64.132, | |
| "step": 536 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 536, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 732330975627216.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.9530733406603648e-05, | |
| "num_train_epochs": 4, | |
| "per_device_train_batch_size": 64, | |
| "weight_decay": 0.01257257791736148 | |
| } | |
| } | |