What it does

This repo contains exercises for the Deep Learning textbook by Goodfellow, Bengio, and Courville. The book itself has been the standard reference since 2016; these are the problem sets that were meant to go with it. The exercises are written in LaTeX and cover the same territory: linear algebra, probability, machine learning basics, and deep architectures.

The interesting bit

The textbook shipped without exercises for years. This repo is essentially the author retroactively filling in a gap that frustrated self-learners and instructors who wanted to assign homework. It’s a rare case of a core reference getting its missing piece from the source.

Key highlights

• Exercises map directly to chapters of the printed book
• LaTeX source included, so you can reformat or extract subsets
• 1,400+ stars suggest pent-up demand from the ML education crowd
• Maintained by Goodfellow himself, not a third-party study guide

Caveats

• No solutions provided; you’re on your own for verification
• Sparse README gives no guidance on difficulty progression or prerequisites
• TeX-only format means you’ll need to compile or read raw source

Verdict

Grab this if you’re working through the book solo and need structure, or if you’re teaching and want vetted problems. Skip it if you need hand-holding, solutions, or interactive notebooks — this is old-school pencil-and-paper stuff.