the thing about expert fine tunes is you need good data lmao, and to get good data you need real expertise in whatever expert discourse, oooor you'll end up fine-tuning it to the mid l0l

Posters: LericDax. LericDax

2 likes recorded by the source corpus.