Labs

Learn it by building it.

Our books are free, and so is running any chapter's notebook on your own computer. Labs is for when you'd rather we run it: ready in about 30 seconds, with nothing to install.

Labs: 40 hours a month in the Lab. Running a chapter yourself stays free.

Why it matters

Reading about code isn't the same as running it.

Here's what one chapter of A Primer on Search Relevance asks of you before you can run it, and what it asks in the Lab.

On your own machine

Set it up first

  • Install Python 3.12.15
  • Install bm25s 0.3.16, faiss-cpu 1.16.0, lightgbm 4.7.0, numpy 2.5.5, pandas 3.1.1, pystemmer 3.1.0, scikit-learn 1.9.1 and scipy 1.18.1, at exactly those versions
  • Get the book's data
  • Fix whatever doesn't match, until your output matches the book's

Then you can start the chapter.

Free: the book's notebooks, with its code and data, from GitHub

In the Lab

Open the chapter

  • Python 3.12.15
  • bm25s 0.3.16, faiss-cpu 1.16.0, lightgbm 4.7.0, numpy 2.5.5, pandas 3.1.1, pystemmer 3.1.0, scikit-learn 1.9.1, scipy 1.18.1
  • The book's data
  • The chapter's notebook, every example a cell

Ready in about 30 seconds. You start on the chapter.

The versions are the ones A Primer on Search Relevance's examples ran on in its current edition, 2026.11.10. In the Lab they're installed already, with the data, on a machine of your own.

In the Lab

Run it. Change it. Keep it.

Each chapter opens in JupyterLab, every example a cell, on the book's exact Python and libraries.

A chapter of A Primer on Search Relevance in the Lab: each example under the words that lead to it, as the book ran it, and a cell for your own.

Run it as the book did

Every example runs on the versions the book pins, so you see the output the book shows.

Make it your own

Change the code, add cells, and try your own ideas on the book's code and data. Reset to the book whenever you like.

Keep your work

Your notebooks are saved for each chapter and edition, and kept for 12 months after you last open them.

Do your class's work

Courses set assignments you complete in the Lab, and they're graded there.

How it works

From any chapter, in three steps.

  1. 1

    Open a chapter

    On any chapter of a book, choose Open in Lab. There's nothing to install.

  2. 2

    Give it 30 seconds

    A machine of your own starts with the book's environment and notebook.

  3. 3

    Start building

    Hours count only while it's open, and it stops after 15 minutes idle.

JupyterLab works best on a laptop or a desktop.

With Labs

Build it here, then use it at work.

Labs gives you 40 hours a month in the Lab. Running a chapter yourself stays free: download its book's notebooks from any chapter.