|
FlexNN 1
Fully connected neural network built from scratch with flexible n-layer design and multiple activations.
|
Pool1D layers — MaxPool1D and AvgPool1D (weightless). More...
#include <utility>#include <vector>#include <Eigen/Dense>#include "LayerTypes.hpp"#include "../activations/Activation.hpp"

Go to the source code of this file.
Classes | |
| class | FlexNN::Layers::MaxPool1D |
| class | FlexNN::Layers::AvgPool1D |
Namespaces | |
| namespace | FlexNN |
| Namespace for the FlexNN neural network library. | |
| namespace | FlexNN::Layers |
Pool1D layers — MaxPool1D and AvgPool1D (weightless).
Both share Pool1DParams (channels, kernelSize, stride, padding) and are Activation::None only in v0.1. MaxPool routes gradient to argmax; AvgPool distributes 1/K. No weights, so model.bin has w_cnt=b_cnt=0.