Build a Reasoning Model (From Scratch)
Get the eBook free when you register your print book at Manning.
“An exceptional deep dive into the next frontier of AI.”
—Aman Chadha, Google
This book is a practical guide to understanding how modern reasoning-oriented LLMs work by building their core methods step by step. The book tells a clear engineering story: start with a conventional pre-trained LLM, learn how text generation works, build reliable evaluation tools, improve reasoning through inference-time methods, then move into training-based approaches such as reinforcement learning and distillation.
The progression is deliberate. Early chapters establish the baseline model and explain text generation, KV caching, and evaluation with math verifiers. The middle chapters show how reasoning can be improved without changing model weights, using chain-of-thought prompting, sampling, self-consistency, response scoring, and self-refinement. Later chapters move to changing the model itself through reinforcement learning with verifiable rewards, GRPO improvements, format rewards, and finally distillation from stronger reasoning models into smaller ones.
The book is especially useful because it implements the core methods from scratch rather than treating them as black-box library calls. Readers see how self-consistency, self-refinement, Best-of-N, and training-based methods actually work, including their cost and latency trade-offs. It also discusses common failure modes, including cases where refinement can make answers worse. Difficult concepts such as softmax, temperature, and top-p sampling are clarified with code-linked explanations and diagrams, and visual workflows make pipelines and scoring methods easier to follow.
Reading the book feels like following a guided technical build rather than a loose survey of AI topics. Each concept is introduced because the project now needs it. Diagrams, roadmaps, code listings, exercises, and repeated workflow summaries help readers stay oriented through advanced material. This structure reflects Sebastian Raschka’s professional strength: explaining complex machine learning topics by making every detail concrete and showing exactly where each section fits in the larger story. He does not treat mechanisms like evaluation, log-probabilities, KL regularization, or distillation as isolated abstractions; he connects them to the goal of making reasoning models understandable and implementable.
Physically and organizationally, the book has eight chapters and seven substantial appendixes. That design keeps the main narrative focused while moving supporting material like references, exercise solutions, model source code, larger models, batching, evaluation alternatives, and chat interfaces into ordered appendixes. The result is a logically flowing book that remains hands-on, navigable, and technically deep without constantly interrupting the central build.
What’s insideFrom-scratch implementations of core LLM reasoning improvementsVerifier-based evaluation methodsRL with automatic verifiers for mathematics tasks
About the reader
For readers who know Python and have some knowledge of machine learning.
About the author
Sebastian Raschka is an LLM Research Engineer with over a decade of experience. He is the author of the bestselling book Build a Large Language Model (From Scratch).
From the Publisher
“Distills the profound ideas in the clearest, most accessible way.”
Byron Hsu, LMSYS
“Big. Dope. Great read, fun writing.”
Chris Alexiuk, NVIDIA
“The gold standard for developers wanting to build at the cutting edge of AI.”
Omar Sanseviero, Author of Hands-On Generative AI with Transformers and Diffusion Models
why this book?
Build a Reasoning Model (From Scratch) helps you understand modern reasoning-oriented LLMs by building their core techniques from scratch rather than treating them as black boxes.
The book gives practical experience with evaluation, inference-time scaling, self-consistency, self-refinement, reinforcement learning with verifiable rewards, GRPO, and distillation.
Readers who know Python and some machine learning will gain a clear, hands-on path from using a pretrained LLM to improving, training, and evaluating reasoning models.
about Manning
Manning helps developers and tech professionals stay ahead in a fast-moving industry with expert-led books, videos, and projects. Learning never stops, but it’s hard to keep up, so we focus on content that’s practical, clear, and trusted. As an independent publisher, we adapt quickly, from pioneering early-access books to offering DRM-free eBooks. Our series, like “In Action” and “In a Month of Lunches”, reflect a commitment to making complex topics accessible.
Customer Reviews
4.2 out of 5 stars 43
4.0 out of 5 stars 51
4.5 out of 5 stars 622
4.5 out of 5 stars 36
4.8 out of 5 stars 7
4.4 out of 5 stars 14
Level of proficiency
Intermediate Intermediate Intermediate Intermediate Intermediate Advanced
About the reader
For readers with intermediate Python skills. For intermediate Python programmers. Readers need intermediate Python skills and some knowledge of machine learning. For data scientists and ML engineers. For data scientists and data analysts. For data scientists and machine learning engineers.
Special features
Includes liveBook with out built-in AI assistant. Includes liveBook with out built-in AI assistant. Includes liveBook with out built-in AI assistant. Includes liveBook with out built-in AI assistant. Includes liveBook with out built-in AI assistant. Includes liveBook with out built-in AI assistant.
Pages
648 344 368 456 232 520