| { | |
| "model_name": "microsoft/Phi-3.5-mini-instruct", | |
| "training_data_path": "training_data/summary/training_dataset.jsonl", | |
| "training_config": { | |
| "per_device_train_batch_size": 2, | |
| "gradient_accumulation_steps": 4, | |
| "warmup_steps": 10, | |
| "max_steps": 200, | |
| "learning_rate": 0.0002, | |
| "fp16": false, | |
| "bf16": false, | |
| "logging_steps": 5, | |
| "optim": "adamw_torch", | |
| "weight_decay": 0.01, | |
| "lr_scheduler_type": "linear", | |
| "seed": 3407, | |
| "save_strategy": "steps", | |
| "save_steps": 50, | |
| "eval_steps": 50, | |
| "save_total_limit": 3 | |
| }, | |
| "lora_config": { | |
| "r": 16, | |
| "target_modules": [ | |
| "q_proj", | |
| "k_proj", | |
| "v_proj", | |
| "o_proj", | |
| "gate_proj", | |
| "up_proj", | |
| "down_proj" | |
| ], | |
| "lora_alpha": 16, | |
| "lora_dropout": 0.05, | |
| "bias": "none", | |
| "use_gradient_checkpointing": true, | |
| "random_state": 3407, | |
| "use_rslora": false, | |
| "loftq_config": null | |
| }, | |
| "training_date": "2025-08-22T14:26:32.116147", | |
| "max_seq_length": 1024, | |
| "base_model_dir": "model/cache", | |
| "output_dir": "model/fine_tuned" | |
| } |