My first LoRA, trained to depict tiny flattened vehicles on a character's soles. Training data made of 100% furries, so I'm not confident in it working with humans. The cars can look a bit too realistic at times, as a byproduct of the way the training images were made.Trigger Word: carstickerI've also seen that adding car-related things into the positive prompt helps. Examples include "cars stuck to sole".

Files with SHA256: 55af1e26c9df00fd3ce284c84ad8b94dca451201beee69d72a2a0a7744251030