Annotate and describe images with text prompts
Answer questions about images by chatting
a tiny vision language model