llama.cpp and llama.c are both libraries to perform fast inference for Llama-based models.