This is a Bertweet model: It maps sentences & paragraphs to a 768 dimensional dense vector space and classifies them into 5 multi labels.
label2id={ "misogynous": 0, "shaming": 1, "stereotype": 2, "objectification": 3, "violence": 4, },