{
    "adapter_path": "adapters",
    "batch_size": 1,
    "clear_cache_threshold": 0,
    "config": "training/lora.yaml",
    "data": "data/out",
    "fine_tune_type": "lora",
    "grad_accumulation_steps": 4,
    "grad_checkpoint": true,
    "iters": 550,
    "learning_rate": 5e-05,
    "lora_parameters": {
        "rank": 16,
        "scale": 20.0,
        "dropout": 0.05
    },
    "lr_schedule": {
        "name": "cosine_decay",
        "warmup": 40,
        "arguments": [
            5e-05,
            650,
            5e-06
        ]
    },
    "mask_prompt": true,
    "max_seq_length": 2048,
    "model": "mlx-community/Qwen3-4B-Instruct-2507-4bit",
    "num_layers": 16,
    "optimizer": "adamw",
    "optimizer_config": {
        "adam": {},
        "adamw": {},
        "muon": {},
        "sgd": {},
        "adafactor": {}
    },
    "project_name": null,
    "report_to": null,
    "resume_adapter_file": "adapters/0000100_adapters.safetensors",
    "save_every": 100,
    "seed": 7,
    "steps_per_eval": 100,
    "steps_per_report": 20,
    "test": false,
    "test_batches": 500,
    "train": true,
    "val_batches": 25
}