mirror of https://github.com/vladmandic/automatic
2.7 KiB
2.7 KiB
Publish extensions
sd-extension-aesthetic-scorersd-extension-steps-animationsd-extension-system-info
Investigating
Things I'm actively investigating...
Embeddings
Need to study more to determine best out-of-the-box settings:
- Train object
- Train style
- Impact of Hires fix:
- e.g 25 steps and denoising strength 0.25-0.7
- Impact of non-square target resolution
Hypernetworks
LORA
Ideas
Things I'm looking into...
Automatic
Stuff to be fixed...
- Reconnect WebUI
- Settings params on updates install
Tech that can be integrated as part of the core workflow...
- Merge without distortion
- Weighted merges
- Use scripts from API
- Face swap
- Null-text inversion
- Custom diffusion
- Dream artist
- LoCon
Video Generation
- Deforum
- Latent blending
- VIDM: video implicit diffusion models
- Tune-a-Video
- Animator extension
- Prompt travel
- KLMC2 animation
- BOAAB-limit animation https://www.ajayjain.net/journey
- Disco diffusion
- Video killed the radio star
- Seed travel
- Prompt fusion
Experimental
Cool stuff that is not integrated anywhere...
- TensorRT
- GIT
- Bunch of stuff:https://pharmapsychotic.com/tools.html