princeton-nlp / SimPO

SimPO: Simple Preference Optimization with a Reference-Free Reward

Geek Repo

Github PK Tool