wait, does anyone have a "digital clone benchmark" ? Like, if I trained it on 15k of my tweets, how well does it answer/auto complete the last 1k? Can I take out random gaps and see if it fills it in?

Posters: DefenderOfBasic. DefenderOfBasic

3 likes recorded by the source corpus.

Topics: AI & LLMs

View Post: https://x.com/DefenderOfBasic/status/1856542898457604253