hf_quant_config.json

{
    "producer": {
        "name": "modelopt",
        "version": "0.17.0"
    },
    "quantization": {
        "quant_algo": "FP8",
        "kv_cache_quant_algo": null
    }
}