Welcome aMUSEd: Efficient Text-to-Image Generation

In three linesHugging Face introduces aMUSEd, an efficient text-to-image generation model based on MUSE architecture. The model uses image tokenization and a decoder to generate quality images with reduced latency compared to standard diffusion models.

Summary generated by Claude — human-verified