Tags: #llm-alignment

An efficient and user-friendly library for scaling Reinforcement Learning with Large Language Models on large-scale GPU resources.

Provides robust recipes and training code to align language models with human and AI preferences, enhancing helpfulness and safety.

A modular framework for aligning any-modality large models with human intentions and values using diverse fine-tuning and reinforcement learning methods.