| { |
| "adapter_path": "adapters", |
| "batch_size": 2, |
| "clear_cache_threshold": 0, |
| "config": "configs/lora.yaml", |
| "data": "data/sft", |
| "fine_tune_type": "lora", |
| "grad_accumulation_steps": 2, |
| "grad_checkpoint": true, |
| "iters": 1700, |
| "learning_rate": 3e-05, |
| "lora_parameters": { |
| "rank": 16, |
| "scale": 16.0, |
| "dropout": 0.05, |
| "keys": [ |
| "self_attn.q_proj", |
| "self_attn.k_proj", |
| "self_attn.v_proj", |
| "self_attn.o_proj", |
| "mlp.gate_proj", |
| "mlp.up_proj", |
| "mlp.down_proj" |
| ] |
| }, |
| "lr_schedule": { |
| "name": "cosine_decay", |
| "warmup": 50, |
| "arguments": [ |
| 3e-05, |
| 1700, |
| 1e-06 |
| ] |
| }, |
| "mask_prompt": true, |
| "max_seq_length": 1024, |
| "model": "Qwen/Qwen2.5-Coder-1.5B-Instruct", |
| "num_layers": 16, |
| "optimizer": "adam", |
| "optimizer_config": { |
| "adam": {}, |
| "adamw": {}, |
| "muon": {}, |
| "sgd": {}, |
| "adafactor": {} |
| }, |
| "project_name": null, |
| "report_to": null, |
| "resume_adapter_file": null, |
| "save_every": 200, |
| "seed": 7, |
| "steps_per_eval": 100, |
| "steps_per_report": 10, |
| "test": false, |
| "test_batches": 500, |
| "train": true, |
| "trust_remote_code": false, |
| "val_batches": 25 |
| } |
