This article covers a deeper level understanding of Question Answering models in NLP, the datasets commonly used, and how to choose a pre-trained model by considering various factors like the document structure, runtime cost, etc.