AI Book Club: Build a Reasoning Model (From Scratch)
August's book is "Build a Reasoning Model (From Scratch)"!
This is a casual-style event. Not a structured presentation on topics. Sometimes, the discussion even drifts away from the chapters, but feel free to grab the mic to help steer it back.
Feel free to join the discussion even if you have not read the book chapters! :)
Want to discuss the contents during the reading week? Join the Flyte MLOps Slack group.
-------------------------------------------------
About the book:
Title: Build a Reasoning Model (From Scratch)
Authors: Sebastian Raschka
Published: July 2026
Manning (Promo code: AIBookClub should give you 45% off: https://www.manning.com/books/build-a-reasoning-model-from-scratch
O'rielly platform: https://learning.oreilly.com/library/view/build-a-reasoning/9781633434677/
Chapters:
1 Understanding reasoning models
2 Generating text with a pretrained 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
Book Description
Build a Reasoning Model (From Scratch) 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.