08:56
The Hugging Face team is trying to speed things up with a new model called aMUSEd that can generate images in just a few seconds, faster than other competitors like Stable Diffusion, as reported by Webmaster's Home on January 5. This lightweight text-to-image model is based on Google's MUSE model and has a parameter size of about 800 million. aMUSEd can be deployed on devices such as mobile devices. Its speed comes from the way it's built. aMUSEd employs an architecture called the Masked Image Model (MIM) instead of the latent diffusion found in Stable Diffusion and other image generation models. According to the Hugging Face team, MIM reduces the inference steps, which increases the speed and explainability of the model. And its small size also makes it fast.
- 1

