Generative AI with Python and TensorFlow 2: Create images, text, and music with VAEs, GANs, LSTMs, Transformer models Paperback – 30 April 2021

The book is targeted at programmers and machine learning practitioners who want to learn about generative models. It is not for complete beginners - as the authors themselves state, one should have a basic understanding of probability, linear algebra, and deep learning.The book starts by introducing core concepts such as Docker, Kubernetes and Tensorflow, and then moves onto more advanced (and interesting) topics such as generating digits, images, music and deep fakes, and even copying artist’s styles to generate new art. Consequently, as the reader progresses through the book there are ample opportunities to use his/her coding skills to experiment.For this reader there were two chapters of special interest.- Chapter 8 on using transformers to generate text, in which BERT and GPT are introduced and explained, and then a program created to generate fake headlines. Having recently read Rothman’s book on transformers this chapter was very useful.- Chapter 10 in which various aspects of reinforcement learning are discussed.By the end of the book, this reader definitely felt that he had acquired a much firmer understanding of GANs and what they are capable of. Recommended and a good book to gain knowledge to have in one’s arsenal.

This book provides thorough insight not only with deep neural networks, GANs and emerging applications in generative AI, but also covers some very important theory with mathematical derivations. What really stood out to me, was the containerisation using Docker and Kubernetes which is vital for developing and deploying your AI projects. Example configurations are provided within the book itself, with detailed explanations and an overview of what those mean. I have only scratched the surface with this text, but I am looking forward to thoroughly studying the techniques for composing music with generative models. I am also very excited about the emerging applications such as discovering pharmaceutical drugs using generative models. Generative AI for medical applications is going to be huge and this book is a great way to become familiar with this exciting new field.