Offensive Language Detection with BERT-based models, By Customizing Attention Probabilities

Open in new window