NanoQuant: Efficient Sub-1-Bit Quantization of Large Language Models

13 points| chrsw | 17 days ago |arxiv.org