Sebastian Raschka · 2026-10-24 · notable
Sebastian Raschka — an 11-hour video course on building reasoning models
Sebastian Raschka's 'Building Reasoning Models from the Ground Up' is a from-scratch video course: eight YouTube videos and over 11 hours of PyTorch coding on evaluation, inference-time scaling, reinforcement learning and distillation.
Eight videos that code a reasoning model's key parts from scratch in PyTorch.
What is it?
'Building Reasoning Models from the Ground Up' is Sebastian Raschka's new video course, published on October 24, 2026 as the companion to his book Build a Reasoning Model (From Scratch). It has eight videos with over 11 hours of coding.
How does it work?
The course works through the main methods behind reasoning LLMs in Python and PyTorch: how to evaluate them, how to spend more compute at inference time, how to train with reinforcement learning, and how to distill reasoning into smaller models. The code lives in the rasbt/reasoning-from-scratch repository on GitHub.
Why does it matter?
Most reasoning-model material is papers or high-level talks. Raschka's course lets engineers write the code themselves and see each step work, which is a fast way to understand what labs mean by RL training and test-time scaling.
Who is it for?
ML engineers and students who learn by coding
Try it
git clone https://github.com/rasbt/reasoning-from-scratch