Transformer is a deep machine learning model based on the encoder–decoder network architecture and is used primarily in the field of NLP.