Tools for Deep Learning

This part is a practical reference for working with the book itself and the ecosystem around it. It documents the d2l library — the saved functions and classes the chapters accumulate — and how to use the notebooks across the four supported frameworks. It covers the everyday tooling a reader needs: running the code on your own hardware or in the cloud, choosing GPUs and reasoning about their memory and throughput, and the mechanics of contributing improvements back.

Where the rest of the book teaches models, this part teaches the workflow: environments, hardware, reproducibility, and the conventions that keep a multi-framework project maintainable. Treat it as a manual to consult rather than a chapter to read linearly.