| { | |
| "cache_dir": null, | |
| "model_max_length": 2560, | |
| "max_steps": -1, | |
| "num_train_epoches": 2, | |
| "per_device_train_batch_size": 8, | |
| "gradient_accumulation_steps": 1, | |
| "gradient_checkpointing": true, | |
| "eval_steps": 500000000, | |
| "eval_epoches": 1, | |
| "per_device_eval_batch_size": 4, | |
| "learning_rate": 1e-05, | |
| "weight_decay": 0.0, | |
| "lr_scheduler_type": "linear", | |
| "warmup_steps": 0, | |
| "warmup_ratio": 0, | |
| "logging_steps": 8, | |
| "logging_epoches": 1, | |
| "save_steps": 50000000000, | |
| "save_epoches": 1, | |
| "save_total_limit": 0, | |
| "save_best": false, | |
| "seed": 42, | |
| "num_training_steps_per_epoch": 6160, | |
| "num_updating_steps_per_epoch": 6160, | |
| "num_eval_steps_per_epoch": null, | |
| "num_training_steps": 12320, | |
| "num_training_steps_aggr_devices": 98560, | |
| "num_updating_steps": 12320, | |
| "num_updating_steps_aggr_devices": 98560, | |
| "num_eval_steps": null, | |
| "per_eval_steps": 500000000, | |
| "num_updating_warmup_steps": 0, | |
| "num_updating_warmup_steps_aggr_devices": 0, | |
| "num_logging_steps": 8, | |
| "per_save_steps": 50000000000 | |
| } |