same idea, different artworks in the training datav1.0
trained on random art images in different styles that i collected/stole from xwitter
