{"max_new_tokens":16,"llm_model":"llm.mnn","llm_weight":"llm.mnn.weight","backend_type":"cpu","thread_num":2,"precision":"low","memory":"low","sampler_type":"greedy"}