I wanted to see if I could train a Jessica Rabbit LoRA that would keep her original cel shaded styling from the movie Who Framed Roger Rabbit.

Trained with AI-Toolkit, 3500 steps, 50 images.

You may need to tweak the strength between 0.8 to 1.1, depending on your checkpoint.

Trained with Hyphoria [Illu & NAI] - you may see varied results on other checkpoints (eg SmoothMix will heavily adapt with its style)

Please post images against the model - will tip those I particularly enjoy!