# How Lessons Are Organized in AI-For-Beginners: The Complete 12-Week Curriculum Structure

> Explore the AI-For-Beginners curriculum structure. Discover how lessons are organized week-by-week with READMEs, notebooks, and labs for a complete 12-week AI journey.

- Repository: [Microsoft/AI-For-Beginners](https://github.com/microsoft/AI-For-Beginners)
- Tags: architecture
- Published: 2026-08-24

---

**The AI-For-Beginners curriculum uses a hierarchical folder structure under `lessons/` with numeric prefixes to indicate learning order, containing consistent README.md files, executable Jupyter notebooks, and hands-on labs across 12 weeks of AI instruction.**

The microsoft/AI-For-Beginners repository structures its educational content as a fully-featured, 12-week curriculum designed for progressive skill building. Understanding how the lessons are organized in AI-For-Beginners allows learners to navigate efficiently from environment setup through advanced neural network architectures. The repository implements a standardized organizational pattern that supports both self-paced study and classroom instruction through predictable file layouts and comprehensive documentation.

## The Hierarchical Curriculum Structure

The repository’s content lives in the top-level **`lessons/`** directory, where each major topic receives a numeric prefix (e.g., `1-Intro`, `2-Symbolic`) to enforce natural learning progression. This design makes the curriculum order immediately apparent from folder names alone and corresponds directly to the master table documented in the root **[`README.md`](https://github.com/microsoft/AI-For-Beginners/blob/main/README.md)** (lines 81–99).

The curriculum flows through eight major thematic sections:

- **0 – Course Setup**: Environment configuration and instructor resources
- **I – Introduction**: Historical context and fundamental AI concepts  
- **II – Symbolic AI**: Knowledge representation and expert systems
- **III – Neural Networks**: Perceptrons to deep learning frameworks
- **IV – Computer Vision**: CNNs, transfer learning, and generative models
- **V – Natural Language Processing**: Embeddings, transformers, and LLMs
- **VI – Other AI Techniques**: Genetic algorithms, reinforcement learning, and multi-agent systems
- **VII – AI Ethics**: Responsible AI principles and fairness

### Weeks 0–1: Environment Setup and Foundations

The journey begins in **`lessons/0-course-setup/`**, which contains [`setup.md`](https://github.com/microsoft/AI-For-Beginners/blob/main/setup.md) for environment installation, [`how-to-run.md`](https://github.com/microsoft/AI-For-Beginners/blob/main/how-to-run.md) for execution guidance, and [`for-teachers.md`](https://github.com/microsoft/AI-For-Beginners/blob/main/for-teachers.md) with instructor notes. **`lessons/1-Intro/`** follows with historical context, basic AI concepts, overview materials, and introductory assignments.

### Weeks 2–3: Symbolic AI and Neural Networks

**`lessons/2-Symbolic/`** houses knowledge representation materials including `Animals.ipynb`, `FamilyOntology.ipynb`, and `MSConceptGraph.ipynb` for ontology exploration. **`lessons/3-NeuralNetworks/`** progresses through `Perceptron.ipynb`, `OwnFramework.ipynb`, and framework introductions (`IntroPyTorch.ipynb`, `IntroKerasTF.ipynb`), allowing learners to build neural networks from scratch before using established libraries.

### Weeks 4–5: Computer Vision and Natural Language Processing

**`lessons/4-ComputerVision/`** covers image processing through `ConvNetsPyTorch.ipynb` and `ConvNetsTF.ipynb`, alongside auto-encoders, GANs, object detection, and segmentation notebooks. **`lessons/5-NLP/`** explores text representation, word embeddings, RNNs, generative networks, BERT implementations, and large language model demonstrations.

### Weeks 6–7: Specialized Techniques and Ethics

**`lessons/6-Other/`** contains `Genetic.ipynb` for genetic algorithms, `CartPole-RL-PyTorch.ipynb` for reinforcement learning, and multi-agent system examples. **`lessons/7-Ethics/`** provides responsible AI reading materials and links to Microsoft Learn modules on fairness and accountability.

### Advanced Extras

The **`lessons/X-Extras/`** folder contains cutting-edge demonstrations such as `Clip.ipynb` for CLIP model exploration and VQ-GAN experiments, serving as extension material for advanced learners.

## Standard Internal Lesson Layout

Every lesson folder follows an identical internal structure to reduce cognitive load and support automated tooling:

- **[`README.md`](https://github.com/microsoft/AI-For-Beginners/blob/main/README.md)**: Topic introduction with conceptual explanations and navigation links to specific notebooks
- **Jupyter Notebooks (`*.ipynb`)**: Executable code cells with explanatory markdown, typically offered in both PyTorch and TensorFlow variants where applicable
- **`lab/` subfolder**: Optional hands-on assignments and practical exercises for skill reinforcement

This uniformity ensures that learners can open any lesson directory and immediately understand where to find theory, code examples, and practice problems.

## Programmatically Navigating the Curriculum

The consistent file structure enables automated exploration and processing of the educational content. You can interact with the curriculum programmatically using standard Python libraries:

```python

# List all lesson directories to verify curriculum completeness

import pathlib

lesson_root = pathlib.Path("lessons")
lesson_dirs = sorted([p.name for p in lesson_root.iterdir() if p.is_dir()])
print("Lesson folders:", lesson_dirs)

```

```python

# Load the Perceptron notebook as JSON for automated analysis

import json
import pathlib

perceptron_nb = pathlib.Path("lessons/3-NeuralNetworks/03-Perceptron/Perceptron.ipynb").read_text()
nb_data = json.loads(perceptron_nb)
print("Number of cells:", len(nb_data["cells"]))

```

```python

# Execute a TensorFlow notebook programmatically using nbclient

from nbclient import NotebookClient
import pathlib

nb_path = pathlib.Path("lessons/4-ComputerVision/07-ConvNets/ConvNetsTF.ipynb")
client = NotebookClient(nb_path.read_text())
client.execute()

```

These patterns demonstrate that every notebook in the curriculum can be parsed, analyzed, and executed through the same programmatic interface, regardless of subject matter.

## Summary

- The AI-For-Beginners curriculum organizes content into **12 sequential weeks** under the `lessons/` directory
- **Numeric prefixes** (0, 1, 2, etc.) enforce logical learning progression from setup to advanced topics
- Each lesson follows a **consistent internal layout** with README.md, Jupyter notebooks, and optional lab/ folders
- **Dual-framework support** provides both PyTorch and TensorFlow implementations for key neural network concepts
- The master curriculum table in **[`README.md`](https://github.com/microsoft/AI-For-Beginners/blob/main/README.md)** (lines 81–99) serves as the definitive navigation reference

## Frequently Asked Questions

### How are the lessons organized in AI-For-Beginners for self-paced learning?

The curriculum uses numbered folder prefixes (0 through 7, plus X-Extras) to indicate the recommended learning sequence. Each folder contains a README.md file that introduces the topic and links to executable Jupyter notebooks, allowing learners to progress from basic setup through advanced AI techniques without external guidance.

### What specific files are included in each lesson folder?

Every lesson contains a README.md for conceptual introductions, one or more Jupyter notebooks (*.ipynb) with runnable code examples, and optionally a `lab/` subfolder with hands-on assignments. According to the source code, notebooks like `Animals.ipynb` and `Perceptron.ipynb` include both explanatory markdown cells and executable Python code.

### Does the curriculum support both PyTorch and TensorFlow?

Yes, the repository provides dual implementations for many core concepts. For example, `lessons/4-ComputerVision/` contains both `ConvNetsPyTorch.ipynb` and `ConvNetsTF.ipynb`, while `lessons/3-NeuralNetworks/` includes `IntroPyTorch.ipynb` and `IntroKerasTF.ipynb` for framework comparison.

### Where is the master curriculum outline documented?

The definitive curriculum table appears in the root **[`README.md`](https://github.com/microsoft/AI-For-Beginners/blob/main/README.md)** file, specifically between lines 81 and 99, as implemented in the microsoft/AI-For-Beginners repository. This table maps each week number to its corresponding `lessons/` subfolder, core focus, and specific notebook filenames.