|
FlexNN 1
Fully connected neural network built from scratch with flexible n-layer design and multiple activations.
|
Shared PODs for FlexNN::Layers — LayerType, DType, and per-type Params. More...
#include <cstdint>#include <type_traits>

Go to the source code of this file.
Classes | |
| struct | FlexNN::Layers::DenseParams |
| Parameters for a Dense layer. More... | |
| struct | FlexNN::Layers::Conv1DParams |
| Parameters for a Conv1D layer. More... | |
| struct | FlexNN::Layers::Pool1DParams |
| Parameters for pooling layers (Max and Avg share this POD). More... | |
| struct | FlexNN::Layers::BatchNormParams |
| Parameters for BatchNorm1D. More... | |
Namespaces | |
| namespace | FlexNN |
| Namespace for the FlexNN neural network library. | |
| namespace | FlexNN::Layers |
Enumerations | |
| enum class | FlexNN::Layers::LayerType : uint8_t { FlexNN::Layers::Dense = 0 , FlexNN::Layers::Conv1D = 1 , FlexNN::Layers::BatchNorm1D = 2 , FlexNN::Layers::MaxPool1D = 3 , FlexNN::Layers::AvgPool1D = 4 } |
| Layer kind, stable on-wire (uint8_t). More... | |
| enum class | FlexNN::Layers::DType : uint16_t { FlexNN::Layers::F32 = 0 } |
Data type for blobs in model.bin. More... | |
Shared PODs for FlexNN::Layers — LayerType, DType, and per-type Params.
This header is standalone (no Eigen) so both ModelIO and per-layer headers can include it without pulling heavy dependencies. Values are stable on-wire for model.bin (see LLD_FLEXNN §5, §7, Appendix A).