Python* API Reference for Intel® Data Analytics Acceleration Library 2020 Update 1

Public Member Functions | Static Public Attributes | List of all members
Batch Class Reference

Computes the results of the backward softmax layer in the batch processing mode. More...

Public Member Functions

def __init__
 
def getMethod
 
def getLayerInput
 
def getLayerParameter
 
def getLayerResult
 
def getResult
 
def setResult
 
def clone
 
def allocateResult
 
def compute
 
- Public Member Functions inherited from LayerIface
def getLayerResult
 
def getLayerInput
 
def getLayerParameter
 
def clone
 
def allocateResult
 
def addInput
 
- Public Member Functions inherited from AlgorithmImpl
def computeNoThrow
 
def compute
 
def checkComputeParams
 
def checkResult
 
def setupCompute
 
def resetCompute
 
def enableResetOnCompute
 
def hostApp
 
def setHostApp
 
- Public Member Functions inherited from Algorithm
def checkComputeParams
 
def getBaseParameter
 
- Public Member Functions inherited from AlgorithmIfaceImpl
def enableChecks
 
def isChecksEnabled
 
- Public Member Functions inherited from AlgorithmIface
def checkComputeParams
 
def checkResult
 
def getMethod
 

Static Public Attributes

 parameter = ...
 
 input = ...
 

Detailed Description

Parameters
fptypeData type to use in intermediate computations of the backward softmax layer, double or float
methodBackward softmax layer method, Method
Enumerations
  • Method Computation methods for the backward softmax layer
  • backward.InputId Identifiers of input objects for the backward softmax layer
  • backward.InputLayerDataId Identifiers of extra results computed by the forward softmax layer
  • backward.ResultId Identifiers of result objects for the backward softmax layer
References
Deprecated:
This item will be removed in a future release.
Aliases
  • Batch_Float64DefaultDense is an alias of Batch(fptype=float64, method=daal.algorithms.neural_networks.layers.softmax.defaultDense)
  • Batch_Float32DefaultDense is an alias of Batch(fptype=float32, method=daal.algorithms.neural_networks.layers.softmax.defaultDense)

Constructor & Destructor Documentation

def __init__ (   self,
  args 
)

Variant 1

Default constructor

Deprecated:
This item will be removed in a future release.

Variant 2

Constructs a backward softmax layer in the batch processing mode and initializes its parameter with the provided parameter

Parameters
parameterParameter to initialize the parameter of the layer
Deprecated:
This item will be removed in a future release.

Variant 3

Constructs a backward softmax layer by copying input objects and parameters of another backward softmax layer in the batch processing mode

Parameters
otherAlgorithm to use as the source to initialize the input objects and parameters of the algorithm
Deprecated:
This item will be removed in a future release.

Member Function Documentation

def allocateResult (   self)

Allocates memory to store the result of the backward softmax layer

Returns
Status of computations
def clone (   self)

Returns a pointer to the newly allocated the backward softmax layer with a copy of input objects and parameters of this backward softmax layer in the batch processing mode

Returns
Pointer to the newly allocated layer
def compute (   self)

Invokes computations

def getLayerInput (   self)

Returns the structure that contains input objects of the softmax backward layer

Returns
Structure that contains input objects of the softmax backward layer
def getLayerParameter (   self)

Returns the structure that contains prameters of the backward softmax layer

Returns
Structure that contains parameters of the backward softmax layer
def getLayerResult (   self)

Returns the structure that contains result of the backward softmax layer

Returns
Structure that contains result of the backward softmax layer
def getMethod (   self)

Returns the method of the algorithm

Returns
Method of the algorithm
def getResult (   self)

Returns the structure that contains result of the backward softmax layer

Returns
Structure that contains result of the backward softmax layer
def setResult (   self,
  result 
)

Registers user-allocated memory to store results of the backward softmax layer

Parameters
resultStructure to store result of the backward softmax layer
Returns
Status of computations

Member Data Documentation

input = ...
static

Input objects of the layer

parameter = ...
static

softmax layer parameters structure


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

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