A ai.googleblog.com MaMMUT: A simple vision-encoder text-decoder architecture for multimodal tasks 72 points mfiguiere 3 years ago 33 comments
