RT @EMostaque: State of the art CLIP model for image to text released by the DataComp team supported by a Stability AI grant - better data =