This model was trained on images of the city streets of Cairo from the film Raiders of the Lost Ark. It generates arabic, middle eastern, egyptian feeling streets, narrow alleys, roads, etc.