view article Article Unbelievable! Run 70B LLM Inference on a Single 4GB GPU with This NEW Technique lyogavin β’ Nov 30, 2023 β’ 47