The ggml.ai team behind llama.cpp joins Hugging Face, uniting two pillars of open-source AI to accelerate local inference.