/trlx

A repo for distributed training of language models with Reinforcement Learning via Human Feedback (RLHF)

Primary LanguagePythonMIT LicenseMIT

Stargazers