Daria from the final movie.

Issues:

1. Blurry data set leads to inherantly blurry lora. It looks more accurate to the show which is cool, but will look worse when using more realistic or detailed chekpoints.

2. Works well with the cartoony checkpoints, but, in testing, had a pretty bad (50/50) failure rate on some checkpoints.

3. Overzealous tagging in the data set has lead to some secret tags; i.e., 'drink' returns a blue fastfood drink container, complete with plastic top (but sometimes not straw?), so some generic words may lead to unexpected outcomes.