RAFT: Reward rAnked Fine-Tuning for Generative Foundation Model Alignment 논문리뷰

https://arxiv.org/abs/2304.06767