LLaMA-Pro-8B-AWQ/quant_config.json

{
    "zero_point": true,
    "q_group_size": 128,
    "w_bit": 4,
    "version": "GEMM",
    "modules_to_not_convert": null
}