logoalt Hacker News

htrp • today at 1:29 AM • 1 reply • view on HN

they always undertrain the model


Replies

cmrdporcupine • today at 1:32 AM

yeah it's kind of the point. they're pedagogical tools. they give you the full recipe for building it from scratch or fine-tuning it beyond what it is.

but i'm prepared to be surprised, maybe they'll conjure up something standalone useful this time

➕ show 1 reply