CS/NLP at SemEval-2022 Task 4: Effective Data Augmentation Methods for Patronizing Language Detection and Multi-label Classification with RoBERTa and GPT3

25105

Knowledge Triples