The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits Paper • 2402.17764 • Published Feb 27 • 603