| { | |
| "best_metric": 0.6319243311882019, | |
| "best_model_checkpoint": "BERT-WMM/run-3/checkpoint-533", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1599, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 6.5503740310668945, | |
| "learning_rate": 1.7399056118476298e-05, | |
| "loss": 0.713, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7384976525821596, | |
| "eval_loss": 0.6319243311882019, | |
| "eval_runtime": 2.0773, | |
| "eval_samples_per_second": 1025.361, | |
| "eval_steps_per_second": 64.506, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 8.555328369140625, | |
| "learning_rate": 9.483198011799183e-06, | |
| "loss": 0.4521, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7619718309859155, | |
| "eval_loss": 0.6519739627838135, | |
| "eval_runtime": 2.0756, | |
| "eval_samples_per_second": 1026.207, | |
| "eval_steps_per_second": 64.559, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 5.210901260375977, | |
| "learning_rate": 1.5673399051220685e-06, | |
| "loss": 0.2844, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7741784037558685, | |
| "eval_loss": 0.7667396068572998, | |
| "eval_runtime": 2.0571, | |
| "eval_samples_per_second": 1035.437, | |
| "eval_steps_per_second": 65.14, | |
| "step": 1599 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1599, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 507646505902536.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 2.531491422515341e-05 | |
| } | |
| } | |