Stability-AI
`ComfyUI` plugin
Initial ComfyUI plugin implementation
Uses basic workflows for text-to-image, image-to-image, inpaint/outpaint
- Image count in UI set to value from plugin, image count != batch size, images are queued, seeds are incremental
- Image strength default value configurable by plugin, also changed the default value because 100% image strength just returns the same image
- Image strength is backwards to denoise in ComfyUI, e.g. denoise 1.0 is equivalent to Image strength 0%, 75% image strength is equivalent to 0.25 denoise
Note: Plugin UI extension support would be useful for additional features, something like this, just a quick implementation, not pushed for now because it could be done better, maybe the types could be the UI elements in Theme, definitions below would change "select" -> "Popout" and "boolean" -> "Checkbox"
TODO: lots
Let me know if there are any changes or anything to add for this initial version
@hlky We're considering moving out most plugins from this monorepo into separate repositories so you don't need our approval.
Beyond WebGPU + SDWebUI (which we have the capacity to support first-party), I'm afraid of @KAJdev and myself becoming a bottleneck.
We'd maintain a list of recommended plugins in this project linking out to efforts like this for maximum visibility.
How does that sound?
Open to suggestions/ideation here!
@hlky We're moving to ComfyUI for lots of things internally now, so it makes sense for this to be a first-party plugin.
Sorry for the whiplash!
Would you mind re-opening this PR? We are interested in having official ComfyUI support by June 22nd and would love to help get this done. 🙏