Package io.github.kirstenali.deepj.nn

Class BatchNorm

java.lang.Object

io.github.kirstenali.deepj.nn.Module

io.github.kirstenali.deepj.nn.BatchNorm

All Implemented Interfaces:

Layer, Trainable

public final class BatchNorm
extends Module

Channel-axis (axis 1) normalization, using backend reductions and persistent FLOAT32 running statistics.

Constructor Summary

Constructors

Constructor	Description
BatchNorm(int channels)
BatchNorm(int channels, float epsilon, float momentum, ParameterFactory parameters)

Method Summary

Modifier and Type	Method	Description
List<Tensor>	buffers()	Persistent non-parameter tensors, in deterministic traversal order.
Tensor	forward(Tensor input)
List<Parameter>	parameters()
Tensor	runningMean()
Tensor	runningVariance()

Methods inherited from class io.github.kirstenali.deepj.nn.Module

children, eval, isTraining, setTraining, train

Methods inherited from class java.lang.Object

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

Methods inherited from interface io.github.kirstenali.deepj.layers.Layer

forward, supportsIncrementalInference

Methods inherited from interface io.github.kirstenali.deepj.training.Trainable

zeroGrad

Constructor Details

BatchNorm

public BatchNorm(int channels)

BatchNorm

public BatchNorm(int channels,
 float epsilon,
 float momentum,
 ParameterFactory parameters)

Method Details

parameters

public List<Parameter> parameters()

Specified by:

parameters in interface Trainable

Overrides:

parameters in class Module

buffers

public List<Tensor> buffers()

Description copied from class: Module

Persistent non-parameter tensors, in deterministic traversal order.

Overrides:

buffers in class Module

runningMean

public Tensor runningMean()

runningVariance

public Tensor runningVariance()

forward

public Tensor forward(Tensor input)