I uploaded workflow with Facedetailer for people who need to improve face after txt2img / highres fix. Usually you need it if your character at distance (feet out of frame / full body poses).

How to use?

1) Upload your image in load node.

2) Choose Checkpoint Model (For example - Cyberfix V-pred);

3) Choose LoRA (or click "bypass" on LoRA node);

4) Write prompt for your picture (Positive and negative);

5) Run workflow.

If you need help you can ask.