config.json

{
    "llm_model": "llm.mnn",
    "llm_weight": "llm.mnn.weight",
    "embedding_file": "embeddings_int4.bin",
    "backend_type": "cpu",
    "thread_num": 4,
    "precision": "low",
    "memory": "low",
    "mllm": {
        "backend_type": "cpu",
        "thread_num": 4,
        "precision": "low",
        "memory": "low"
    },
    "sampler_type": "mixed",
    "mixed_samplers": [
        "penalty",
        "topK",
        "topP",
        "min_p",
        "temperature"
    ],
    "penalty": 1.0,
    "temperature": 1.0,
    "topP": 0.95,
    "topK": 20,
    "min_p": 0,
    "max_new_tokens": 40960
}