Fetching the paper…

Text-to-Image Diffusion Models Cannot Count, and Prompt Refinement Cannot Help · Around