Skip to main content

Command Palette

Search for a command to run...

Introduction

Updated
1 min read
B

With 2+ years of experience in web backend development, I now specialize in AI engineering, building intelligent systems and scalable solutions. Passionate about crafting innovative software, I love exploring new technologies, experimenting with AI models, and bringing ideas to life. Always learning, always building.

The best way to understand how things work is to build them by yourself. I believe implementing deep learning in Rust is a great way to develop a strong understanding in deep learning while enjoying the process of learning Rust. That is exactly what this series is about.

Instead of relying on high-level machine learning frameworks, we’ll build major components from scratch: forward and backward propagation, gradient descent, normalization, and more. While we'll use the Burn crate as our computational backbone, we’ll limit its role to essential tensor operations.

By the end of this series, I hope you will have a solid understanding of how deep learning works under the hood.

All source code is available in the repository, with each article’s implementation tagged by chapter.

Understanding Deep Learning by Building It in Rust

Part 1 of 8

Learn deep learning by building it from scratch in Rust using Burn only for tensors. We’ll implement activations, losses, backprop, and optimizers step by step to understand how neural networks truly work.

Up next

Part 1: Tensors

As you begin your implementation, the first thing you will encounter is the tensor. In order to know how deep learning works, you must understand this fundamental structure. A tensor is simply a multi