Würstchen: An Efficient Architecture for Large-Scale Text-to-Image Diffusion Models
Want to make some of these yourself?