V2:
All you need to do is to prompt for the before picture and let the lora take care of the after picture.
I'm not sure v2 gives better looking results but the prompting works much better than the beta version.
If you enjoy my work please leave a like or a review. Pics of your work are also appreciated :)
Description
A quick project I did to try and do a concept lora. The tagging is far from perfect so if anyone has experience in these kinds of loras I would appreciate some advice on how to tag the dataset properly.
Anyway the lora works well enough so I am publishing it as a beta.