This repository provides a comprehensive, open-source textbook and learning stack for Machine Learning Systems Engineering, addressing the critical gap between AI model development and robust, real-world deployment. Aimed at engineers, researchers, and students, it establishes AI engineering as a foundational discipline by teaching the principles and practices of building efficient, reliable, and safe intelligent systems. The project offers a structured curriculum with hands-on components, enabling learners to master both theoretical concepts and practical implementation.

How It Works

The project is centered around a textbook being restructured into two focused volumes: Volume I covers foundational single-machine ML systems, while Volume II delves into distributed systems and production infrastructure at scale. Development occurs openly on the dev branch, allowing community observation and contribution. The learning stack complements the textbook with TinyTorch, a framework for understanding ML internals by implementing them from scratch, and hardware kits for hands-on deployment on edge devices like Arduino and Raspberry Pi.

Quick Start & Requirements

• Read Online: Access the textbook at https://mlsysbook.ai/book/.
• Download: PDFs and EPUBs are available at https://mlsysbook.ai/book/assets/downloads/Machine-Learning-Systems.pdf and https://mlsysbook.ai/epub.
• Volume I: Available now at https://mlsysbook.ai/vol1/.
• Volume II: Under active development, expected Summer 2026 at https://mlsysbook.ai/vol2/.
• TinyTorch: Get started with the framework at https://mlsysbook.ai/tinytorch/.
• Prerequisites: Specific hardware requirements for labs are detailed within the respective kit documentation. Python is implied for TinyTorch.

Highlighted Details

• Restructuring into two volumes: Volume I (Foundations) and Volume II (Scale), with Volume I content complete and undergoing polish.
• "Working in the Open" development model on the dev branch, providing transparency into the textbook's creation process.
• TinyTorch framework allows users to build ML frameworks from scratch, fostering a deep understanding of underlying mechanisms.
• Hardware kits and labs enable practical engineering experience under real-world constraints like memory, power, and timing.

Maintenance & Community

The project is actively developed on the dev branch, with a commitment to transparency. Community engagement is encouraged through GitHub Discussions. The broader ML Systems Ecosystem is accessible via https://mlsysbook.org.

Licensing & Compatibility

The book content is licensed under CC BY-NC-ND 4.0, permitting free sharing with attribution but prohibiting commercial use and derivative works. The TinyTorch code is licensed under Apache 2.0, allowing for free use, modification, and distribution, including in commercial projects.

Limitations & Caveats

Volume II is currently under active development, with content still being written and revised, meaning it may contain rough edges or placeholder figures. Components like Software Co-Labs and the AI Olympics are slated for release in 2026. The dev branch represents ongoing work and may not be as stable as the main branch, which hosts the last stable release.