C++ API Reference for Intel® Data Analytics Acceleration Library 2018 Update 1

List of all members

Implements the abstract interface LayerIface. LayerIfaceImpl is, in turn, the base class for the classes interfacing the layers. More...

Class Declaration

Member Function Documentation

virtual services::Status addInput ( forward::ResultPtr  result,
size_t  resultIndex,
size_t  inputIndex 
)
inlinevirtual

Connects two layers in neural network by getting tensor with value from the result of the previous layer and adding it to the input object of this layer algorithm

Parameters
[in]resultStructure that contains results of the previous layer
[in]resultIndexIndex of the tensor with value in the structure that contains results of the previous layer
[in]inputIndexIndex in the input object of this layer algorithm where the tensor with value should be placed

Implements LayerIface.

virtual services::Status allocateInput ( )
inlinevirtual

Allocates memory buffers needed for the computations

Implements LayerIface.

virtual layers::forward::LayerIfacePtr getLayerForPrediction ( ) const
inlinevirtual

Returns the layer that corresponds to this layer on the prediction stage

Returns
The layer that corresponds to this layer on the prediction stage

Implements LayerIface.

virtual services::Status initializeInput ( )
inlinevirtual

Initializes values of weights and biases if needed

Implements LayerIface.


The documentation for this class was generated from the following file:

For more complete information about compiler optimizations, see our Optimization Notice.