-
I would like to use the controlnet(https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/controlnet_quantized/README.md) model pipeline with a 16:9 resolution input image, but currentl…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues and checked the recent builds/commits of both this extension and the webui
### What happened?
Cannot click o…
-
-
Firstly, thanks for this excellent work.
After reading the paper and experimenting with the code, I thought I'd drop a suggestion. Rather than altering a pretrained LDM model (Stable Diffusion) di…
-
I tested using the inpaint controlnet. It was an attempt to duplicate the input image, not for inpaining. The attempt was some success, but the color was changed. I tried various ways, but I couldn't …
-
如标题，谢谢
-
Hi,
it not always but there are some error happened when generating , error as below, any idea about this? thanks!
INFO len( region_list )=3 …
-
How is it now: in provisioning script now you can specify loras, controlnet models, etc.
I think it would be useful to have the same capability for text embeddings which go to separate folder. Tha…
-
I create and adjust my pose,
send to controlnet,
select and activate ControlNet
select openpose for model and preprocessor
enter a prompt and hit generate
the render looks nothing like the pose i…
-
Hi
I like to use controlnet to ensure poses are similar from image to image. However, each time Fast Stable Difussion re-create the poses. Is there a way to "lock" the pose in controlnet. This is es…