Class GroupNormLayer

java.lang.Object

smile.deep.layer.ModuleLayer

smile.deep.layer.GroupNormLayer

All Implemented Interfaces:

Function<Tensor,Tensor>, Layer

public class GroupNormLayer
extends ModuleLayer

Group normalization. The input channels are separated into groups. The mean and standard-deviation are calculated separately over each group.

Constructor Summary

Constructors

Constructor	Description
GroupNormLayer(int groups, int channels)	Constructor.
GroupNormLayer(int groups, int channels, double eps, boolean affine)	Constructor.

Method Summary

Modifier and Type	Method	Description
Tensor	forward(Tensor input)	Forward propagation (or forward pass) through the layer.

Methods inherited from class ModuleLayer

module, toString

Methods inherited from class Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Methods inherited from interface Function

andThen, compose

Methods inherited from interface Layer

apply, name, to, to

Constructor Details

GroupNormLayer

public GroupNormLayer(int groups,
 int channels)

Constructor.

Parameters:

groups - the number of groups to separate the channels into. The number of channels must be divisible by the number of groups.

channels - the number of input channels in (N,C,H,W).

GroupNormLayer

public GroupNormLayer(int groups,
 int channels,
 double eps,
 boolean affine)

Constructor.

Parameters:

groups - the number of groups to separate the channels into. The number of channels must be divisible by the number of groups.

channels - the number of input channels in (N,C,H,W).

eps - a value added to the denominator for numerical stability.

affine - when set to true, this layer has learnable affine parameters.

Method Details

forward

public Tensor forward(Tensor input)

Description copied from interface: Layer

Forward propagation (or forward pass) through the layer.

Parameters:

input - the input tensor.

Returns:

the output tensor.