Where to Find AI-For-Beginners Documentation: Complete Guide to the Microsoft Curriculum

You can find all AI-For-Beginners documentation in the repository's README.md, index.html (Docsify configuration), and the lessons/ directory structure, with the interactive web version hosted on GitHub Pages.

The microsoft/AI-For-Beginners repository provides a comprehensive, self-contained curriculum for learning artificial intelligence. All AI-For-Beginners documentation is organized as a Docsify-powered static site that renders Markdown files directly from the repository, making it accessible both on GitHub and via an interactive web interface. The documentation spans from high-level curriculum overviews to runnable Jupyter notebooks and setup guides.

Primary Documentation Entry Points

The curriculum uses three interconnected entry points to serve different learning preferences.

Root README and GitHub Pages

The README.md at the repository root serves as the primary documentation hub. It contains a high-level curriculum summary, badge links, and a comprehensive "Content" table that maps lesson numbers to their folder locations. When you visit the GitHub Pages URL (https://microsoft.github.io/AI-For-Beginners/), you are viewing the same README.md content rendered through the Docsify engine.

Docsify Configuration in index.html

The index.html file configures the Docsify engine with window.$docsify settings including routerMode: 'history' for hash-less routing. This transforms the flat Markdown files into a navigable single-page application. The configuration treats README.md as the homepage and automatically generates navigation from the repository's file structure.

Lesson Structure and Navigation

Deep learning concepts live in the lessons/ directory, which follows a strict organizational pattern to maintain consistency across topics.

Standard Lesson Layout

Each lesson directory (lessons/<topic>/) contains:

• A README.md explaining the theoretical foundations
• Jupyter notebooks (*.ipynb) with runnable implementations
• An optional lab/ sub-folder with hands-on exercises

For example, the Perceptron lesson resides at lessons/3-NeuralNetworks/03-Perceptron/ and includes both the theory README and implementation notebooks.

Framework-Specific Notebooks

Every lesson provides dual implementations in both PyTorch and TensorFlow. The notebooks are clearly labeled and contain identical concepts implemented with their respective frameworks. The README.md in each lesson folder links to both versions, allowing you to choose based on your preferred deep learning library.

Supporting Documentation Resources

Beyond the core curriculum, the repository includes specialized documentation for environment setup and supplementary tools.

Setup and Environment Configuration

The environment.yml file pins exact dependency versions including TensorFlow 2.17, Keras 3.5, NumPy 1.26, PyTorch, and OpenCV. Detailed setup instructions live in lessons/0-course-setup/setup.md, which explains how to create a Conda environment, run Jupyter locally, or launch the repository in a VS Code dev-container.

Standalone Examples

The examples/ folder contains four minimal scripts that demonstrate core AI concepts without requiring the full lesson hierarchy:

• 01-hello-ai-world.py – Basic environment verification
• 02-simple-neural-network.py – Minimal neural network implementation
• 03-image-classifier.ipynb – Computer vision quickstart
• 04-text-sentiment.py – Natural language processing sample

Reference the examples/README.md for execution instructions and dependencies.

Quiz Application

The Vue.js quiz application resides in etc/quiz-app/ with its own dedicated documentation. The etc/quiz-app/README.md explains how to build, run, and deploy the interactive assessment tool, including ESLint configuration and development server setup.

How to Access Documentation Locally

You can serve the full documentation site on your machine to browse offline or modify content.

Clone and Serve the Docsify Site

# Clone with sparse checkout to skip large translation files

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'

# Install the pinned environment

conda env create -f environment.yml
conda activate ai4beg

# Serve documentation locally

npm install -g docsify-cli
docsify serve .

This launches the interactive documentation site at http://localhost:3000, rendering the same content available on GitHub Pages.

Open Specific Lesson Notebooks

import nbformat as nbf

# Load the Perceptron notebook (PyTorch version)

path = "lessons/3-NeuralNetworks/03-Perceptron/Perceptron.ipynb"
with open(path, "r", encoding="utf-8") as f:
    notebook = nbf.read(f, as_version=4)

print(f"Cells: {len(notebook['cells'])}")
print(f"Metadata: {notebook['metadata']}")

Launch the Quiz Application

cd etc/quiz-app
npm install   # Installs Vue 2 and ESLint dependencies

npm run serve # Starts dev server on http://localhost:8080

Community and Contribution Guidelines

The CONTRIBUTING.md file establishes guidelines for pull requests, issue reporting, and community etiquette. It links to the Discord server and Microsoft forum communities where you can ask questions about the curriculum content. Translation efforts and sketchnote illustrations are organized under translations/ and lessons/sketchnotes/ respectively, with contribution workflows documented in the main README.

Summary

• Primary entry point: The root README.md serves as both GitHub documentation and the Docsify homepage
• Interactive site: The index.html file powers the GitHub Pages version with hash-less routing
• Lesson structure: Each topic in lessons/<topic>/ contains a README, dual-framework notebooks (PyTorch/TensorFlow), and optional labs
• Environment setup: environment.yml and lessons/0-course-setup/setup.md provide Conda and container configuration
• Quick examples: The examples/ folder offers standalone scripts for immediate experimentation
• Assessment tools: etc/quiz-app/README.md documents the Vue.js quiz application

Frequently Asked Questions

Do I need internet access to view the AI-For-Beginners documentation?

No, you can view all documentation offline by cloning the repository and opening the Markdown files directly in any text editor. To view the interactive Docsify version offline, run docsify serve . after cloning, or simply open index.html in a browser (though search functionality requires a local server).

What is the difference between the README and the Docsify site?

The README.md and the Docsify site display identical content, but the Docsify site (accessed via index.html or GitHub Pages) adds a navigation sidebar, search functionality, and hash-less routing. The README is the source file; the Docsify site is the rendered interface that makes the documentation browsable like a traditional website.

How do I switch between PyTorch and TensorFlow examples?

Each lesson folder contains separate Jupyter notebooks with clear filename conventions. Look for notebooks ending in -PyTorch.ipynb or -TensorFlow.ipynb within the lesson directory. The lesson's README.md contains links to both versions, and the table of contents in the main README indicates which framework-specific resources are available.

Where can I find troubleshooting help for local setup?

The lessons/0-course-setup/setup.md file contains detailed troubleshooting for Conda environment creation, Jupyter configuration, and dev-container setup. For platform-specific issues or dependency conflicts, refer to the environment.yml file to verify exact package versions (TensorFlow 2.17, Keras 3.5, NumPy 1.26) and the repository's Issues tab for community solutions.

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