This paper aims to explain our group SSN_NLP_MLRG submission to HASOC 2021: Hate speech and offensive language detection in English and Indo-Aryan language, wherein we discover one of a kind models to carry out the subtask1 includes subtask A: To discover the remarks is Hate speech and offensive (HOF) or now not and subtask B: to categorize the HOF remarks into profanity (PRFN), Hate speech (HATE), Offensive (OFFN) in English, Hindi language and subtask A in Marathi language. The experiments cowl unique gaining knowledge of strategies that consist of gadget getting to know, transfer studying, and Multilingual pre-educated models. Our exceptional fashions are Roberta for English subtask A, BERT for English subtask B, and MBERT for the Hindi subtask A, Hindi subtask B, and Marathi subtask A.
https://www.eurchembull.com/uploads/paper/3743641c46d49700b9557e15b8cb16e7.pdf