Currently, there are two pre-trained models to try - v1: example-based model which takes animation examples for conditioning on the style. - v2: label-based model which takes labels for conditioning on the style. This model was trained on a subset of styles