node-llama-cpp:本地运行大语言模型，支持多硬件加速与函数调用，提供预编译二进制

Run AI models locally on your machine with node.js bindings for llama.cpp. Enforce a JSON schema on the model output on the generation level