Upload an image of a pet, and compare results from a trained ViT model, a zero-shot CLIP model, and an OpenAI vision model.