Sebastian Raschka’s bestselling book Build a Large Language Model (From Scratch) is the best way to learn how Large Language Models function. With over 350 pages in seven chapters and five Appendixes, it guides you step-by-step in building an entire large language model similar to GPT-2. It uses Python and the PyTorch deep learning library. It’s a unique way to learn this subject, which some believe is the only way to truly learn: you build a model yourself.

Even with the clear explanations, diagrams, and code in the book, learning a complex subject is still hard. This Test Yourself guide intends to make it a little easier. The structure mirrors the structure of Build a Large Language Model (From Scratch), focusing on key concepts from each chapter. You can test yourself with multiple-choice quizzes, questions on code and key concepts, and questions with longer answers that push you to think critically. The answers to all questions are provided.

Depending on what you know at any point, this Test Yourself guide can help you in different ways. It will solidify your knowledge if used after reading a chapter. But it will also benefit you if you digest it before reading. By testing yourself on the main concepts and their relationships you are primed to navigate a chapter more easily and be ready for its messages.

We recommend using it before and after reading, as well as later when you have started forgetting. Repeated learning solidifies our knowledge and integrates it with related knowledge already in our long-term memory.