mirror of https://github.com/vladmandic/automatic
2.3 KiB
2.3 KiB
TODO
Issues
Stuff to be fixed, in no particular order...
- SD-XL Sketch/Inpaint
- Kandinsky 2.2 (2.1 is working)
- Misterious Extensions auto-enabling
- Misterious Extra network corruptions
Features
Stuff to be added, in no particular order...
- Diffusers: save before refiner
- Diffusers: prompt parser
- Diffusers: SD-XL Lora
- Train: use
interrogator - Train: use
rembg - Train: new Lora train UI
- Port
p.all_hr_prompts - Image watermark using
image-watermark - Image phash and hdash using
imagehash - Update
Wiki - Create new
GitHubhooks/actions for CI/CD - Import core repos
- Docker PR
- Model merge using
git-rebasin - Additional upscalers
- XYZ grid upscalers
- New image browser
- Update
gradio - Rename repo: automatic -> sdnext
- New icons
- Enable refiner workflow for
ldmbackend - Add
sgmbackend - Improve
lycologging - Cache models when switching backends
- Style editor
- Built-in motd-style notifications
- Localization
Investigate
Stuff to be investigated...
Merge PRs
Pick & merge PRs from main repo...
- up-to-date with: df004be
- current todo list: https://github.com/AUTOMATIC1111/stable-diffusion-webui/compare/df004be...394ffa7
Integration
Tech that can be integrated as part of the core workflow...
- Merge without distortion
- Weighted merges
- Null-text inversion
- Custom diffusion, Custom diffusion
- Dream artist
- QuickEmbedding
- DataComp CLiP
- ClipSeg
- DragGAN
- LamaCleaner
TensorRT