While I'm between working on some other models, this one is for fun and for a learning experience. What happens when training a FLUX model on so few images, so little variation, and there's a non-human feature on the character?

The images of this model are from screenshots of the music video for the song Anoana by Heilung. The video is on YouTube and a description about the lyrics on Genius.

Overall, I think it turned out well. Usually, it will generate the character without the antlers unless they are prompted.