Run a private OpenAI-compatible LLM endpoint on Apple Silicon with one Rust binary

Disclosure: I maintain Ferrum, an MIT-licensed local LLM inference server written in Rust. If you...