My stuff

  • My Workflows

  • Liked Workflows

  • Following Workflows

Go to OpenArt main site
Upload workflow

Controlnet++Inpainting

5.0

0 reviews
40
26.6K
6.7K
5
Description

A usage method summerized from Github.

1. The Promax model on huggingface is required to use the Inpaint function.

https://huggingface.co/xinsir/controlnet-union-sdxl-1.0/blob/main/diffusion_pytorch_model_promax.safetensors

2. The images generated by Inpaint preprocessing are RGBA and need to be converted to RGB through other nodes.

3. If a broadcast error (output with shape [1, 1280] doesn't match the broadcast shape [2, 1280]) occurs, ComfyUI needs to be restarted.

2 and 3 may be bugs, and they will be able to be used normally after being corrected in the future.

Discussion

(No comments yet)

Loading...

Author

67
126.7K
2.0K
518.8K

No reviews yet

  • - latest (a year ago)

  • - v20240720-062640

Primitive Nodes (2)

ACN_ControlNet++InputNode (1)

ACN_ControlNet++LoaderAdvanced (1)

Custom Nodes (16)

ComfyUI

  • - CheckpointLoaderSimple (1)

  • - CLIPTextEncode (2)

  • - VAEDecode (1)

  • - EmptyLatentImage (1)

  • - PreviewImage (2)

  • - LoadImage (1)

ComfyUI Inspire Pack

  • - KSampler //Inspire (1)

  • - LayerUtility: ImageRemoveAlpha (1)

  • - LayerMask: MaskEdgeUltraDetail V2 (1)

  • - InpaintPreprocessor (1)

  • - ACN_AdvancedControlNetApply (1)

  • - GetImageSizeAndCount (1)

  • - ResizeMask (1)

  • - JWImageResizeByLongerSide (1)

Checkpoints (1)

realvisxlV40_v40LightningBakedvae.safetensors

LoRAs (0)