Artist Style for TGedNathan. Trained on 50 images.

Activation text: tg3dnathan

Negative text: open mouth

I'm trying out a new method for getting the dataset for this LoRa with text bubbles present throughout a lot of the base images used for training. I might make a new version using frames from their animations since it's a bit cleaner there if I think I need to. Characters tend to have open mouths so you can add 'open mouth' to the negative prompt if it continues to be an issue.