VPTQ: Extreme Low-bit Vector Post-Training Quantization for Large Language Models Paper • 2409.17066 • Published 9 days ago • 22
