| { | |
| "stage": "2_SFT", | |
| "base_model": "/workspace/nishka-gkc-phi3-base", | |
| "training_start": "2025-12-29 04:16:19.019832", | |
| "dataset": { | |
| "path": "/workspace/pql_sft_training.jsonl", | |
| "examples": 10038, | |
| "estimated_tokens": 7000000 | |
| }, | |
| "lora_config": { | |
| "r": 32, | |
| "alpha": 64, | |
| "dropout": 0.05, | |
| "target_modules": [ | |
| "q_proj", | |
| "k_proj", | |
| "v_proj", | |
| "o_proj", | |
| "gate_proj", | |
| "up_proj", | |
| "down_proj" | |
| ] | |
| }, | |
| "training_args": { | |
| "epochs": 2, | |
| "batch_size": 2, | |
| "learning_rate": 0.0001, | |
| "max_length": 2048 | |
| } | |
| } |