AutoModelForImageTextToText

Load a vision-language text-generation model by name (a picture and a prompt in, text out): a :class:GenerativeModel.