For producing images of Wander.
(NEW) V2! 100+ more images have been trained to (hopefully) improve the model.
