How to Set Up the AI for Beginners Conda Environment Locally: Complete Installation Guide

Clone the Microsoft AI for Beginners repository and create the ai4beg Conda environment using the root environment.yml to run all curriculum notebooks with a pre-configured Jupyter kernel.

The AI for Beginners curriculum by Microsoft provides hands-on machine learning lessons through Jupyter notebooks. To run these notebooks locally, you need a dedicated Conda environment named ai4beg that bundles all required dependencies. This guide walks through the complete setup process using the official environment definitions in the repository.

Prerequisites

Before installing the AI for Beginners Conda environment, ensure you have:

• Conda installed (Anaconda or Miniconda)
• Git for cloning the repository
• Approximately 2-5 GB of free disk space (less with sparse checkout)

Step 1: Clone the Microsoft AI for Beginners Repository

Start by downloading the curriculum source code. Navigate to your preferred directory and run:

git clone https://github.com/microsoft/AI-For-Beginners.git
cd AI-For-Beginners

This creates a local copy of the entire repository, including notebooks, data files, and configuration files.

Optional: Skip Large Translation Assets

If you want to reduce download size, use a sparse checkout to exclude the translations/ and translated_images/ folders:

git clone --filter=blob:none --sparse https://github.com/microsoft/AI-For-Beginners.git
cd AI-For-Beginners
git sparse-checkout set --no-cone '/*' '!translations' '!translated_images'

This approach downloads only the core curriculum content, saving significant storage space.

Step 2: Create the AI for Beginners Conda Environment

The repository includes a root-level environment.yml file that defines the complete environment. This file is also duplicated at .devcontainer/environment.yml for container-based development.

Create the environment using Conda:

conda env create -f environment.yml

What the Environment Includes

According to the [environment.yml](https://github.com/microsoft/AI-For-Beginners/blob/main/environment.yml) source, the ai4beg environment:

• Pulls packages from defaults and conda-forge channels
• Installs core scientific Python packages
• References requirements.txt for additional Python dependencies
• Registers an IPython kernel (ipykernel) for Jupyter notebook execution

The environment is designed as a single Conda environment that supports the entire curriculum, eliminating the need to switch environments between lessons.

Step 3: Activate the Environment and Verify Jupyter Kernel

Once creation completes, activate the environment:

conda activate ai4beg

Verify the Jupyter kernel is available:

jupyter kernelspec list

You should see python3 or ai4beg in the list of available kernels. The ipykernel package installed by environment.yml automatically registers this kernel.

Step 4: Launch Jupyter and Start Learning

With the environment active, start Jupyter Notebook or JupyterLab:

jupyter notebook

Or for JupyterLab:

jupyter lab

Navigate to any lesson folder and open the .ipynb notebook files. All imports should resolve without additional configuration.

Alternative: Using VS Code Dev Container

For a containerized setup, the repository includes [.devcontainer/environment.yml](https://github.com/microsoft/AI-For-Beginners/blob/main/.devcontainer/environment.yml). If using VS Code with the Dev Containers extension:

1. Open the repository in VS Code
2. Click "Reopen in Container" when prompted
3. The environment builds automatically from the Dev Container configuration

This approach isolates the AI for Beginners setup from your host system entirely.

Troubleshooting Common Issues

Environment Creation Fails

If Conda cannot resolve dependencies, try updating Conda first:

conda update -n base conda
conda env create -f environment.yml --force

Kernel Not Showing in Jupyter

Manually install the kernel if needed:

ipython kernel install --user --name=ai4beg --display-name "Python (AI for Beginners)"

Package Conflicts

The environment.yml specifies channel priority. If conflicts persist, clear Conda's package cache:

conda clean --all
conda env create -f environment.yml

Summary

• Clone the microsoft/AI-For-Beginners repository using standard or sparse checkout
• Create the ai4beg Conda environment from root environment.yml
• Activate with conda activate ai4beg to access the pre-configured Jupyter kernel
• Launch Jupyter and run any notebook without additional setup
• The same environment definition exists at .devcontainer/environment.yml for container workflows

Frequently Asked Questions

What Python version does the AI for Beginners environment use?

The environment.yml specifies Python 3.x compatible with the curriculum's dependencies. Specific version pinning is handled through requirements.txt, which the environment references for comprehensive package management. Check the source file for the exact version constraint.

Can I use pip instead of Conda for this setup?

While pip can install individual packages, the official AI for Beginners workflow relies on Conda for environment isolation and dependency resolution across scientific computing packages. The environment.yml orchestrates both Conda and pip installations internally via requirements.txt.

How do I update the environment when requirements change?

Delete and recreate the environment:

conda activate base
conda remove --name ai4beg --all
conda env create -f environment.yml

Alternatively, update in place with conda env update -f environment.yml if the changes are additive.

Is GPU support included in the environment?

The base environment.yml provides CPU-based scientific computing. GPU-accelerated packages like PyTorch or TensorFlow with CUDA support may require additional installation steps or modified environment files depending on your hardware and the specific lesson requirements.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}