This is just a very fast model created for educational purposes. Its goal is to classify images of castles, where the training set contains categories "spain", "france", and "japan". Images come from Duck Duck Go image search, using "castle interior"/"castle exterior" as suffix.
