Dowon

`stable-diffusion-webui\models\adetailer`. If folder not exists, make one.

I've only tested with torch 2.0, and the output will vary depending on your settings.

Increase the detection model confidence threshold to around 80.

You sent `ad_use_cfg_scale` as `None`. Options will be added as I update. I apologise for this.

I've made some big changes to API usage. Please see the wiki.

Well, it could be an issue with the adetailer, or it could be an issue with your prompts and denoising strength.

The ddetailer does nothing in img2img and goes straight to face detection. In contrast, adetailer goes through the whole process. This error occurs when you don't specify any masks in...

I'll try, but I don't think I'll be able to control it as intended.

How about this: https://github.com/Bing-su/adetailer/wiki/Advanced

Sorry about that. It's hard to write instructions...