Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/ageron/handson-ml3/llms.txt

Use this file to discover all available pages before exploring further.

The quickest way to run the Hands-On ML notebooks is directly in the cloud — no installation, no package management, just click a link and start coding. If you prefer to work locally, you can have a full environment running in about ten minutes using Anaconda. This page covers both paths.

​
Cloud options (no installation required)

You can open any notebook in several hosted environments. Google Colab is the recommended option because it offers free GPU access, deep integration with Google Drive, and the most consistent experience with the book’s code.
Google Colab is the recommended platform. It runs on Google’s infrastructure, provides optional free GPU/TPU access, and requires only a Google account.Open in Google Colab
Colab provides a temporary environment. Any files you create or data you download will be deleted when the session ends. Download anything you want to keep, or save notebooks to your Google Drive.

​
Local setup (3 steps)

Running locally gives you full control over your environment, persistent storage, and the ability to work offline. The setup uses Anaconda (or Miniconda) to create an isolated Python 3.10 environment with all required libraries.
1

Clone the repository and install Anaconda

Install Anaconda or Miniconda if you do not have it already. Then clone the repository:
git clone https://github.com/ageron/handson-ml3.git
cd handson-ml3
You will also need git. Alternatively, download main.zip and unzip it manually.
2

Create and activate the conda environment

From inside the handson-ml3 directory, create the homl3 environment from the included environment.yml file. This installs Python 3.10 along with Scikit-Learn, TensorFlow, Keras, and all other required libraries:
conda env create -f environment.yml
conda activate homl3
3

Register the kernel and launch Jupyter

Register the homl3 environment as a Jupyter kernel so the notebooks use the correct Python installation:
python -m ipykernel install --user --name=python3
jupyter notebook
Your browser will open Jupyter’s file browser. Click index.ipynb to see all chapters and start navigating.
If you have an NVIDIA GPU with Compute Capability 3.5 or higher, you can enable GPU acceleration for TensorFlow. See the full installation guide for GPU setup steps including CUDA and cuDNN.

​
Prerequisites checklist

Before you start, make sure you have the following:
  • Python — the conda environment installs Python 3.10 automatically
  • git — needed to clone the repository (or download the zip instead)
  • Anaconda or Miniconda — for local setup only
  • A Google account — for Colab only
  • Basic Python knowledge — familiarity with functions, loops, and importing libraries
No prior machine learning experience is required. The book and notebooks build up every concept from scratch.

​
What’s next

Full installation guide

Detailed walkthrough including GPU setup and update instructions

Docker setup

Run notebooks in an isolated container without touching your system Python

Cloud platforms compared

Detailed comparison of Colab, Kaggle, Binder, and Deepnote

Project overview

All 19 chapters at a glance with topic summaries

Build docs developers (and LLMs) love

Get started for freeTalk to us