Java* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 2

List of all members

Class that computes the results of the backward softmax cross-entropy layer in the batch processing mode. More...

Class Constructor

◆ SoftmaxCrossBackwardBatch() [1/2]

Constructs the backward softmax cross-entropy layer by copying input objects of backward softmax cross-entropy layer

Parameters
contextContext to manage the backward softmax cross-entropy layer
otherA backward softmax cross-entropy layer to be used as the source to initialize the input objects of the backward softmax cross-entropy layer

◆ SoftmaxCrossBackwardBatch() [2/2]

SoftmaxCrossBackwardBatch ( DaalContext  context,
Class<? extends Number >  cls,
SoftmaxCrossMethod  method 
)

Constructs the backward softmax cross-entropy layer

Parameters
contextContext to manage the backward softmax cross-entropy layer
clsData type to use in intermediate computations for the layer, Double.class or Float.class
methodThe layer computation method, SoftmaxCrossMethod

Detailed Description

References

Member Function Documentation

◆ clone()

Returns the newly allocated backward softmax cross-entropy layer with a copy of input objects of this backward softmax cross-entropy layer

Parameters
contextContext to manage the layer
Returns
The newly allocated backward softmax cross-entropy layer

◆ compute()

Computes the result of the backward softmax cross-entropy layer

Returns
Backward softmax cross-entropy layer result

◆ getLayerInput()

SoftmaxCrossBackwardInput getLayerInput ( )

Returns the structure that contains input object of the backward layer

Returns
Structure that contains input object of the backward layer

◆ getLayerParameter()

SoftmaxCrossParameter getLayerParameter ( )

Returns the structure that contains parameters of the backward layer

Returns
Structure that contains parameters of the backward layer

◆ getLayerResult()

SoftmaxCrossBackwardResult getLayerResult ( )

Returns the structure that contains result of the backward layer

Returns
Structure that contains result of the backward layer

◆ setResult()

void setResult ( SoftmaxCrossBackwardResult  result)

Registers user-allocated memory to store the result of the backward softmax cross-entropy layer

Parameters
resultStructure to store the result of the backward softmax cross-entropy layer

Member Data Documentation

◆ input

Input data

◆ method

Computation method for the layer

◆ parameter

SoftmaxCrossParameters of the layer


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

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