[N] Stable Diffusion Image Variations released, allows you to do variations like DALLE-2

Justin Pinkey has been experimenting with fine-tuning stable diffusion to use CLIP image embedding as the conditioning instead of a text prompt. This allows you do do the dalle2 like "image variations"

tweet: [https://twitter.com/Buntworthy/status/1566744186153484288](https://twitter.com/Buntworthy/status/1566744186153484288)

github: [https://github.com/justinpinkney/stable-diffusion](https://github.com/justinpinkney/stable-diffusion)

demo made with gradio: [https://github.com/gradio-app/gradio](https://github.com/gradio-app/gradio)

https://preview.redd.it/9jbdu4n1g2m91.jpg?width=3294&format=pjpg&auto=webp&s=716de69799bf14dcb2b84461f21832b679406e22

dalle dalle-2 diffusion image machinelearning stable diffusion