Using SD1.5 resources for SDXL based images #1227
ntdviet
started this conversation in
Show and tell
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I made a workflow to blend SD1.5 visual styles (determined by custom checkpoints and LORAs) and precision into SDXL1.0's fantastic compositions. It ends up also fixing bad faces and hands automatically, which is a common error for SDXL base images from midhot to fullbody. What surprised me was the ability to auto fix multiple faces and hands in group images.
Resource and doku: https://github.com/ntdviet/comfyui-ext/tree/main/custom_workflows/SDXL1.0_SD1.5_Mix_FixTune
Images using this workflow:
![0016_sdxl_refine2](https
Beta Was this translation helpful? Give feedback.
All reactions