| { | |
| "best_metric": 0.6044057011604309, | |
| "best_model_checkpoint": "BERT-WMM/run-7/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": 10.467696189880371, | |
| "learning_rate": 1.2929689145457759e-05, | |
| "loss": 0.7152, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.747887323943662, | |
| "eval_loss": 0.6044057011604309, | |
| "eval_runtime": 2.0899, | |
| "eval_samples_per_second": 1019.204, | |
| "eval_steps_per_second": 64.119, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 12.011512756347656, | |
| "learning_rate": 7.047210007396903e-06, | |
| "loss": 0.456, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7633802816901408, | |
| "eval_loss": 0.6140081882476807, | |
| "eval_runtime": 2.0772, | |
| "eval_samples_per_second": 1025.443, | |
| "eval_steps_per_second": 64.511, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 8.719986915588379, | |
| "learning_rate": 1.1647308693360492e-06, | |
| "loss": 0.286, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7615023474178404, | |
| "eval_loss": 0.7435628771781921, | |
| "eval_runtime": 2.9291, | |
| "eval_samples_per_second": 727.196, | |
| "eval_steps_per_second": 45.748, | |
| "step": 1599 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1599, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 518314932761976.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.8812168283518612e-05 | |
| } | |
| } | |