Zippy's Disco Diffusion Cheatsheet v0.3

- Billions of images are scraped from the internet. These images, along with their text descriptions, are saved in a database. - The AI model uses this database to train through reverse diffusion. - Diffusion adds noise to an image (from the dog to random pixels). - Reverse diffusion turns noise back into an image.

Whether you’re new to AI, or quite experienced with Disco, I think this guide will be of good use to you. This guide was designed as a response to the most frequent questions I’ve been asked by the AI Art community: “What’s the prompt?” “What are the settings?”

Key features of the 2.0 release

• New text-to-image diffusion models  using a brand new text encoder (OpenCLIP). Result: improved image quality, although not everyone likes the style.


• Super-resolution upscaler diffusion models  which enhance the resolution of images by a factor of 4.


• Depth-to-image diffusion model  which infers the de

... See more