2 ms·Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks1 points by zerojames 3y ago