The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits Paper โข 2402.17764 โข Published Feb 27 โข 602