Align Anything: Training All-modality Model with Feedback
JMLR: OmniSafe is an infrastructural framework for accelerating SafeRL research.
Open on GitHub