CivitAI removes the metadata of images that contains workflows so I decided I'd post my SD1.5 and Pony workflows.
This workflow has multiple nice features
Loading loras from the positive prompt
Correctly adding metadata for CivitAI resource detection
ComfyUI's replacement for Highres fix
ComfyUI's replacement for "Adetailer"
Theoretically you don't need to understand how this flow works you can just use it but let me know if you have questions!
This isn't a tutorial on how to setup ComfyUI (there are plenty of tutorials out there).
Note: The Pony Workflow has all the Highres and Adetailer stuff disabled because you dont really need it in Pony. It can be reenabled but you may need to rebuilt it based on the nodes in the SD1.5 work flow.
This workflow is my full workflow so it contains "Adetailer" nodes and a bunch of other random useful/useless nodes.
If you want a tutorial on how to use the "Adetailer" nodes please go here:
https://civarchive.com/models/364333/comfyui-face-eyes-and-lips-adetailer-flow-tutorial
Install Requirements
1. ComfyUI-Manager - https://github.com/ltdrdata/ComfyUI-Manager
2. todo
Description
FAQ
Comments (10)
Thank you so much my Friend 😍❤️
My pleasure! Let me know if you have any questions about it
Thank you so much! Technological illiterate here! If fate wills it i might get to use this workflow
It took me awhile to learn so if you do decide to pick it up don't feel frustrated if it doesn't make sense right away.
Feel free to ask any questions you have. I'm not qualified to give a full tutorial but I can give tips/tricks :D
You are awesome!!! 🙌🙌🙌
This works on sdxl models?
Lol thanks.
Yes it does. You wont have to change any nodes for SDXL/PonyXL just make sure you adjust the settings so its the right settings for XL.
For example, size for XL should be 832x1248.
Than, I will be sure to look at it, i struggled to find a good exemple for ADetailer.
I have a more detailed tutorial for my Adetailer flow here https://civitai.com/models/364333/comfyui-face-eyes-and-lips-adetailer-flow-tutorial
It can target Face, Eyes, and/or Lips with separate prompts for each. (Lips is a little bit too powerful so I rarely use it :P)
love the work flow the face detailer or eye is not or it doent work, my upscaler doesnt work is there wrong with work flow ?
Can you give more details about what isn't working?
For example are you seeing an error or is it just doing nothing?
Can you confirm that the UltralyticsDetectorProvider node is able to find your detection models (like "face_yolov8m.pt)