News for hackers. Stuff that matters.
Re:StarCoder and StarCoderBase: 15.5B parameter models with ... by og_kalu on Tuesday May 16, 2023 @07:41AM (#35960414)
Any paper that's trained more than one model size on the same data affirms the same thing.

Llama 13b was better than 7b and Llama 66b was better than 33b.

Literally pick any one.

There are no comments on this comment yet.