TRL Transformers

HuggingFace Transformer Reinforcement Learning — SFT, DPO, PPO training for aligning LLMs locally.