OFA

This is the official repo for the paper: UNIFYING ARCHITECTURES, TASKS, AND MODALITIES THROUGH A SIMPLE SEQUENCE-TO-SEQUENCE LEARNING FRAMEWORK

OFA is a unified multimodal pretrained model that unifies modalities (i.e., cross-modality, vision, language) and tasks (e.g., image generation, visual grounding, image captioning, image classification, text generation, etc.) to a simple sequence-tosequence learning framework.

Examples

Image Generation (normal query)

Image Generation (counterfactual query)

Open-Ended VQA

grounded QA

Viusal Grounding