Running LLaMA models with int4 quantization in python with llama.cpp - View it on GitHub