|
FlexNN 1
Fully connected neural network built from scratch with flexible n-layer design and multiple activations.
|
Implements FlexNN::Layers::BatchNorm1D forward/backward. More...
#include "layers/BatchNorm1D.hpp"#include <cassert>#include <cmath>#include "activations/detail.hpp"
Namespaces | |
| namespace | FlexNN |
| Namespace for the FlexNN neural network library. | |
| namespace | FlexNN::Layers |
Implements FlexNN::Layers::BatchNorm1D forward/backward.
Forward uses batch stats for training (updates running) and running stats for eval. Backward implements the full BN chain with gamma scaling. The layer is Activation::None only in v0.1 — any other activation returns Status error at export.