Fine grained annotations extending hateful memes dataset with additional labels for identifying protected categories and attack types. - View it on GitHub