Self-Play Fine-Tuning of Diffusion Models for Text-to-Image Generation Paper β’ 2402.10210 β’ Published Feb 15 β’ 32