The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits
The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits
arxiv.org /abs/2402.17764
HN discussion here: https://news.ycombinator.com/item?id=39535800
The paper is coming out of Microsoft Research and the University of Chinese Academy of Sciences, so highly likely to not be a fake.
This is huge. I wonder when the NVIDIA stock is going to react.
0 comments