BatchNorm Properties

The BatchNorm type exposes the following members.

Properties


  Name Description
Public property BetaInitializer
Initializer for the beta weight.
Public property BlendTimeConst
The blend time constant in samples.
Public property Epsilon
Small float added to variance to avoid dividing by zero.
Public property GammaInitializer
Initializer for the gamma weight.
Public property Name
Gets or sets the name of the network layer
(Inherited from LayerConfig.)
Public property NormalizationTimeConstant
The time constant in samples of the first-order low-pass filter that is used to compute mean/variance statistics for use in inference
Public property Params
Gets or sets the parameters.
(Inherited from LayerConfig.)
Public property RunningMeanInitializer
Initializer for the running mean weight.
Public property RunningStdInvInitializer
Initializer for the running standard inv weight.
Public property Shape
The input shape for batch norm layer
Public property Spatial
Boolean, if yes the input data is spatial (2D). If not, then sets to 1D