from
Hacker News
Top
New
Speculative sampling: LLMs writing a lot faster using smaller LLMs
by
spolu
on 6/9/23, 3:33 PM with 0 comments