Trained with 81 image and 10 epoch

Cosine with 0.05 offset

Contains all outfit

I am using dim of 48 so it captures more detail and should work better then most other loras

The Civit AI backupsite is almost finished, you can upload and share your models there:

https://miyukiai.com/

If you like my work please consider supporting me: ko-fi.com/suzushi2024