Does a good job at drawing her face and I managed to get some really beautiful images from it but it took some work. My problem is that too many of the images were trained on her wearing some type of animal ears. I can't just get hair or hair with flowers in it.

Also, the model seems to have somewhat of a Chibi effect, giving the character a big head and small body or making the image look cute and child like. Maybe this was the intention but I would rather be able to make a more realistic person and then add ears or whatever else I want in the prompt.

I give you mad respect for doing an amazing job training the file perfectly without overtraining. It really does work great and better than most lora files, many of which are overtrained. Good job.

Would love to see a version of this retrained without the head ears and childification.