round03/adapter_config.json

{
    "adapter_path": "/Users/dellboy/Documents/Vibe_Coding/TopPDBLX/data/interim/slm/runs/r1-parse-residual-smollm2-360m-round03",
    "batch_size": 32,
    "clear_cache_threshold": 0,
    "config": "/Users/dellboy/Documents/Vibe_Coding/TopPDBLX/data/interim/slm/runs/r1-parse-residual-smollm2-360m-round03/training_config.yaml",
    "data": "/Users/dellboy/Documents/Vibe_Coding/TopPDBLX/data/interim/slm",
    "fine_tune_type": "lora",
    "grad_accumulation_steps": 1,
    "grad_checkpoint": true,
    "iters": 1000,
    "learning_rate": 0.0001,
    "lora_parameters": {
        "rank": 8,
        "dropout": 0.05,
        "scale": 20.0
    },
    "lr_schedule": {
        "name": "cosine_decay",
        "arguments": [
            0.0001,
            950,
            1e-05
        ],
        "warmup": 50,
        "warmup_init": 1e-06
    },
    "mask_prompt": true,
    "max_seq_length": 1024,
    "model": "mlx-community/SmolLM2-360M-Instruct",
    "num_layers": 32,
    "optimizer": "adam",
    "optimizer_config": {
        "adam": {},
        "adamw": {},
        "muon": {},
        "sgd": {},
        "adafactor": {}
    },
    "project_name": "toppdblx",
    "report_to": "wandb",
    "resume_adapter_file": null,
    "save_every": 100,
    "seed": 17,
    "steps_per_eval": 50,
    "steps_per_report": 10,
    "test": false,
    "test_batches": 500,
    "train": true,
    "val_batches": 25
}