Microsoft drops Florence-2, a unified model to handle a variety of vision tasks https://t.co/Dej5BIsM19 https://t.co/5yx3wDhhgn on https://twitter.com/AcerboLivio/status/1803488068558627026