Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks Paper โข 2311.06242 โข Published Nov 10, 2023 โข 84