{
    "batcher": null,
    "cacher": null,
    "compiler": null,
    "factorizer": null,
    "pruner": null,
    "quantizer": "hqq",
    "hqq_backend": "torchao_int4",
    "hqq_compute_dtype": "torch.float16",
    "hqq_group_size": 64,
    "hqq_weight_bits": 4,
    "batch_size": 1,
    "device": "cuda",
    "save_fns": [
        "hqq"
    ],
    "load_fns": [
        "hqq"
    ],
    "reapply_after_load": {
        "factorizer": null,
        "pruner": null,
        "quantizer": null,
        "cacher": null,
        "compiler": null,
        "batcher": null
    },
    "api_key": null
}