llama.cppllama.cpp
llama.cpp
解决什么问题What problem it solves
在普通电脑上跑大模型的主流开源引擎，CPU 也能跑。The mainstream open-source engine for running large models on an ordinary computer, CPU included.
这是什么What it is
用 C/C++ 写的推理引擎，强调轻量与跨平台。支持 CPU + GPU 混合推理。
它是本地部署生态的底座——大量量化版本是为它准备的。
An inference engine written in C/C++ that emphasises lightness and cross-platform support, with mixed CPU and GPU inference.
It is the foundation of the local deployment ecosystem — many quantisations exist specifically for it.
为什么需要知道Why it matters
本地部署的默认选项。选它意味着你能用上最多的量化版本和社区支持。The default choice for local deployment. Picking it means access to the largest set of quantisations and community support.
现在能做什么What you can do now
回到名词地图看第5站的其他 29 个词 →Back to the map: the other 29 terms in station 5 →
出处Sources
- llama.cpp 官方仓库llama.cpp GitHub repository ↗