Georgi Gerganov and the GGML team behind llama.cpp are joining Hugging Face, ensuring long-term open-source support for local AI inference while keeping the project fully community-driven.