This repository contains the source code for Modeling Disagreement in Hateful Memes by a novel Contrastive Learning Paradigm