TinyML: Edge & Embedded
HarvardX Professional Certificate in Tiny Machine Learning
The complete courseware from the HarvardX TinyML Professional Certificate on edX: four courses, in five chapters, that run from ML fundamentals through embedded deployment and MLOps. The materials include slide decks, readings, Colab notebooks, and supplementary material. Harvard SEAS and the Google TensorFlow team developed the original courseware.
Fundamentals of TinyML
ML basics, deep learning building blocks, CNNs, computer vision, and responsible AI design.
Course 2Applications of TinyML
TensorFlow Lite, quantization, keyword spotting, visual wake words, anomaly detection, and data engineering.
Course 3Deploying TinyML
Embedded hardware, TFLite Micro internals, and hands-on deployment of KWS, VWW, and gesture recognition on Arduino.
Course 4MLOps for Scaling TinyML
ML development lifecycle, continuous training, model conversion, deployment at scale, prediction serving, and monitoring.
| TinyML Syllabus | A 10–12 week semester plan with weekly assignments, learning objectives, and notes on adapting it. | Syllabus → |
| Hardware Kits | Deployment labs for TinyML projects on Arduino, Raspberry Pi, and Seeed boards. | Kits → |
| Textbook | Volume I covers the systems foundations: compression, hardware acceleration, benchmarking, and deployment. | Vol I → |
| edX Certificate | The self-paced online version of this curriculum, on HarvardX through edX. | edX → |
The TinyMLx Team
Instructors: Vijay Janapa Reddi, Laurence Moroney, Pete Warden, Lara Suzuki
Guest Instructor: Susan Kennedy
Staff Lead: Brian Plancher
Staff: Colby Banbury, Benjamin Brown, Dhilan Ramaprasad, J. Evan Smith, Matthew Stewart
Contributors: Sharad Chitlangia, Radhika Ghosal, Srivatsan Krishnan, Maximilian Lam, Mark Mazumder
These materials were originally developed for the HarvardX Professional Certificate in Tiny Machine Learning on edX. The original curriculum lists every item, including the forum prompts and quizzes not listed above.