Comment by giancarlostoro

10 hours ago

It didnt, just small enough, which is good, its way more useful to train small purposeful models than to use an entire frontier LLM.

0 comments