Quantize model (llama.cpp)

./build/bin/llama-quantize model.gguf model-q4.gguf Q4_K_M