Build a Reasoning Model (From Scratch)

AvSebastian Raschka

Inbunden, Engelska, 2026

663 kr

Beställningsvara. Skickas inom 11-20 vardagar. Fri frakt över 249 kr.

Fler format och utgåvor

E-bok

467 kr

Beskrivning

Large language models can perform strong reasoning tasks, but using them alone is not enough. To build powerful applications, you need to know how to create and customise them. This book guides you to build your own reasoning model. It explains complex ideas in a simple way and shows how to build a model that can reason. Pretrain a large language model for general text generation. Fine-tune the model to improve reasoning tasks. Evaluate model performance using standard benchmarks. Understand the complete, hands-on project of building a reasoning model. Learn to customise the model using your own data. Build a Reasoning Model (From Scratch) is a practical guide to creating your own large language models. The book takes you through a complete hands-on project. You will learn to prepare data, pretrain a model and fine-tune it for reasoning. The author explains each step clearly so that you can build a working model. After reading this book, you will have the skills to build and adapt your own reasoning models. You can move beyond using existing tools and start creating your own intelligent systems. This book is for readers who know Python and have some knowledge of machine learning.

Produktinformation

Utgivningsdatum:2026-11-02
Vikt:449 g
Format:Inbunden
Språk:Engelska
Antal sidor:440
Förlag:Manning Publications
ISBN:9781633434677

Utforska kategorier

Informationsteknik: allmänt inom Data och IT
Affärsapplikationer inom Data och IT
Hårdvara inom Data och IT

Mer om författaren

Sebastian Raschka is a machine learning and AI researcher. He is known for making complex topics easy to understand. With a focus on practical application, Sebastian brings a clear, step-by-step approach to building advanced models. He translates deep learning theory into hands-on projects that help readers build their own AI systems from scratch.

Recensioner i media

“Truly inspirational! It motivates you to put your new skills into action.” Benjamin Muskalla, Senior Engineer, GitHub “The most understandable and comprehensive explanation of language models yet! It's unique and practical teaching style achieves a level of understanding you can’t get any other way.” Cameron Wolfe, Senior Scientist, Netflix

Innehållsförteckning

1 UNDERSTANDING REASONING MODELS 2 GENERATING TEXT WITH A PRE-TRAINED LLM 3 EVALUATING REASONING MODELS 4 IMPROVING REASONING WITH INFERENCE-TIME SCALING 5 INFERENCE-TIME SCALING VIA SELF-REFINEMENT 6 TRAINING REASONING MODELS WITH REINFORCEMENT LEARNING 7 IMPROVING GRPO FOR REINFORCEMENT LEARNING 8 DISTILLING REASONING MODELS FOR EFFICIENT REASONING APPENDICES APPENDIX A: REFERENCES AND FURTHER READING APPENDIX B: EXERCISE SOLUTIONS APPENDIX C: QWEN3 LLM SOURCE CODE APPENDIX D: USING LARGER LLMS APPENDIX E: BATCHED INFERENCE APPENDIX F: COMMON APPROACHES TO MODEL EVALUATION