π Add multimodal AI capabilities with image-text-to-text pipeline 4e10023 ndc8 commited on about 1 month ago