Inside Deep Learning (IDL): Jupyter notebooks exploring machine learning concepts. This repo provides clear PyTorch implementations and explanations from scratch for concepts that are often hard to find.

Note

This repository uses workshop-template with MystMD.

View this repository in your browser

Tip

All notebooks are supported for Colab and Jupyter NBViewer.

Table of Contents

Tip

🤖 Programming ML models. ➗ Focus on a specific concept, such as mathematics proof. 🔵 Minor variations on the main topics.

1. Linear regression 📈
   1. 🤖 Simple linear regression
   2. 🤖 Multivariate linear regression
   3. 🤖 Multivariate linear regression
   4. 🔵 L2 regularization
2. Classification 📊
   1. 🤖 Multiclass classfication
   • ➗ Softmax function and its derivative
3. Multilayer Perceptron 🧠
   1. 🤖 Multilayer perceptron (MLP)
   • ➗ Gradients and activation functions
   • 🔵 MLP for classification
   • 🔵 MLP like PyTorch

How to Use

Important

Supported on Linux and Windows. For macOS, check the PyTorch install guide.

1. Clone the repository:

   git clone https://github.com/PilotLeoYan/inside-deep-learning.git
   cd inside-deep-learning

2. Install uv (if not already installed):

   curl -LsSf https://astral.sh/uv/install.sh | sh

3. Create environment:
   Inside Deep Learning is written in python=3.14. We recommend using uv to manage dependencies.

   A. Install dependencies with cuda:

   uv sync --extra cuda

   B. Install dependencies without cuda:

   uv sync --extra cpu

4. Build myst-cli in local:

   uv run jupyter book start

   or also launch JupyterLab with:

   uv run jupyter lab

Used Hardware

• CPU: AMD Ryzen 7
• GPU: Nvidia Geforce RTX 2070-SUPER (8GB VRAM)
• RAM: 16GB DDR4

Support

If you find this repo useful, please consider starring ★ it on Github:

If you use this work for something, please cite it using the following BibTeX:

@software{ortegarivera2025insidedeeplearning,
  author={Ortega Rivera, Leonardo F.},
  orcid={0009-0004-0497-2808},
  title={Inside Deep learning},
  url={https://github.com/PilotLeoYan/inside-deep-learning},
  year={2025}
}

Find the Draw.io figures used in this repo in this Drive.

Star History

---

If you would like to contact me you can send me an email.