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

training

Adjusting a model's parameters until its outputs match labelled examples.

Training fits a model's parameters to labelled examples by repeatedly reducing a loss on the training set; for TinyML it happens on a computer, never on the device.

Examples

  • The gesture network is trained on a computer or in Colab from labelled 50 × 3 windows; the board only runs the finished model.

  • Improving a deployed model means collecting more data, training again, converting again and shipping a new model file to the fleet.

See also

inference TinyML LiteRT
Read the full section →