Features:
- uses WebUi github.com/hlky/stable-di…
- so it is very easy to install and can adapt new features from there very easily
- two step generation: running multiple images like four with low step value (can be adjusted) and in preview window HQ version can be re rendered. So whole process is very fast on local computer
- after a "proof of concept" more features like img2img and inpainting will directly added
- I hope that I have enough time for this next week to finish it (if there will be no unexpected roadblocks)
So there would be fully end-to-end Open-Source AI image generation locally running with great User Interface (I prefer Krita over Gimp)
Possible future directions of this project: 1) stop it if somebody else it doing this already 2) rewrite as Photoshop Plugin
btw: it will have it's own dialogs/UI but uses existing WebUi for backend via WebUi Websocket interface
• • •
Missing some Tweet in this thread? You can try to
force a refresh