trained on 25 images of one red orlinsky sculpture, the model DOES NOT WORK well
This model doesn't have a readme.