Transformers for Natural Language Processing: Build innovative deep neural network architectures for NLP with Python, PyTorch, TensorFlow, BERT, RoBERTa, and more

El autor, muy reputado, explica de forma muy sencilla las bases y aplicaciones de Transformers

A well intended hands-on book for NLP enthusiasts; however, when it comes to explaining the core concepts the language barrier is quite evident and the explanations are brief. For example, the author mentions about paying extra focus on the chapter about transformers, which ironically is brief and difficult to digest without referring to the nice free-of-cost references listed at the end of the chapter and in the github notebooks. This holds for the following chapters as well. Additionally, there are quite a few errors in the text. A bit more proofreading is what one could expect for the price you pay. All in all a decent attempt with an overview of transformers, transformer-based architectures and its applications. However, if you are looking for step-by-step conceptual explanations about transformers and their variations this is NOT your go to reference. Personally, I would recommend Getting started with Google BERT by Sudharsan Ravichandiran. On the other hand, if you are is looking for clear steps to set up your first transformers notebook project, this could be a resource to refer to. On a totally different note, if still interested in the book, the publisher offers, both, print + ebook (immediate access) for a lower price of about 29 euros.

This book is a comprehensive reference on Transformers, the new technologies used in natural language processing. The book covers all the mathematics and architectures. It goes in details over HuggingFaces, Bert, Roberta, GPT2 , GPT3, T5, and many more. It is recommended if you want to have an handson understanding on the technologies that took natural language processing by Storm and made things like CNN and ever Menon absolute obsolete. Very much liked chapter 4 with the challenges observed in modern nlp. definitely recommend the book. 5 Stars, also because updated to 2021.

This is the book for which I was finding from the last six months. It expertly introduces transformers and mentors the reader for building innovative deep neural network architectures for NLP. The book covers almost all game-changing applications for natural language processing (NLP), natural language understanting (NLU), and natural laguage generation (NLG). The book is very useful even for beginners in the domain as the questions of each chapter are answered in the Appendix.

may require some googling if new to NLP, but it's the best resource I've found so far, and provides a great framework to direct study material and then apply it.