UNBNLP at SemEval-2019 Task 5 and 6: Using Language Models to Detect Hate Speech and Offensive Language