Lightweight fully connected neural network library (also for Arduino and AVR boards).
Implements a memory-efficient neural network using variadic C++ templates. Supports Sigmoid, Fast Sigmoid, ReLU, Tanh, and Fast Tanh, ... activation functions. Training can be done on PC and models imported to Arduino. No heap memory usage, suitable for AVR microcontrollers.
| Filename | Release Date | File Size |
|---|---|---|
| LightweightNeuralNetwork-1.0.5.zip | 2026-07-21 | 22.30 KiB |
| LightweightNeuralNetwork-1.0.4.zip | 2026-07-20 | 18.13 KiB |
| LightweightNeuralNetwork-1.0.3.zip | 2026-05-30 | 17.26 KiB |
| LightweightNeuralNetwork-1.0.2.zip | 2026-02-12 | 12.51 KiB |
| LightweightNeuralNetwork-1.0.1.zip | 2025-12-13 | 12.52 KiB |