Skip to content
Free public beta until December 4, 2026: every course is open, no card needed. After that, the first chapter of each course stays free and full access is a one-time $39 ($19 for up to 100 founding members chosen for helpful feedback). Your progress is kept. How the beta works

TinyML

Running trained machine-learning models on microcontrollers, with kilobytes of RAM and milliwatts of power.

TinyML is on-device inference on microcontroller-class hardware; the model is trained and converted elsewhere and runs on the chip as a fixed function in statically allocated memory.

Examples

  • The gesture classifier is trained in Colab, converted to an int8 .tflite file, and only the last step, running it on accelerometer windows, happens on the ESP32.

  • A vibration RMS threshold on the Chapter 23 fan needs no TinyML; a model earns its place only when one number can't separate the cases.

See also

inference training TensorFlow Lite Micro edge computing
Read the full section →