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

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

Parameters for the forward and backward batch normalization layers. More...

Public Member Functions

def __init__
 
- Public Member Functions inherited from Parameter
def __init__
 
- Public Member Functions inherited from Parameter
def __init__
 
def check
 

Static Public Attributes

 alpha = ...
 
 epsilon = ...
 
 dimension = ...
 
- Static Public Attributes inherited from Parameter
 weightsInitializer = ...
 
 biasesInitializer = ...
 
 predictionStage = ...
 
 propagateGradient = ...
 
 weightsAndBiasesInitialized = ...
 
 allowInplaceComputation = ...
 

Detailed Description

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

Constructor & Destructor Documentation

def __init__ (   self,
  alpha = 0.01,
  epsilon = 0.00001,
  dimension = 1 
)

Constructs the parameters of the batch normalization layer

Parameters
alphaSmoothing factor that is used in population mean and population variance computations
epsilonA constant added to the mini-batch variance for numerical stability
dimensionIndex of the dimension for which the normalization is performed
Deprecated:
This item will be removed in a future release.

Member Data Documentation

alpha = ...
static

Smoothing factor that is used in population mean and population variance computations

dimension = ...
static

Index of the dimension for which the normalization is performed

epsilon = ...
static

A constant added to the mini-batch variance for numerical stability


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

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