Chapters 7 and 8 cover arguably the most important subfield in this space, namely deep transfer learning techniques relying on transformer neural networks for key functions, such as BERT and GPT. This model architecture class is proving to have the most impact on recent applications, partly due to better scalability on parallel computing architecture than prior methods. Chapters 9 and 10 dig deeper into various adaptation strategies for making the transfer learning process more efficient. Chapter 11 concludes the book by reviewing important topics and briefly discussing emerging research topics and directions.