G github.com vLLM-mlx – 65 tok/s LLM inference on Mac with tool calling and prompt caching 3 points raullen 8 months ago 1 comments
