org.deeplearning4j.nn.layers.mkldnn

Class MKLDNNConvHelper

java.lang.Object

org.deeplearning4j.nn.layers.mkldnn.MKLDNNConvHelper

All Implemented Interfaces:

ConvolutionHelper, LayerHelper

public class MKLDNNConvHelper
extends Object
implements ConvolutionHelper

MKL-DNN Convolution (2d) helper

Author:

Alex Black

Field Summary

Fields

Modifier and Type	Field and Description
protected OpContext	context
protected OpContext	contextBwd

Constructor Summary

Constructors

Constructor and Description
MKLDNNConvHelper(org.nd4j.linalg.api.buffer.DataType dataType)

Method Summary

Modifier and Type	Method and Description
INDArray	activate(INDArray z, IActivation afn, boolean training)
Pair<Gradient,INDArray>	backpropGradient(INDArray input, INDArray weights, INDArray bias, INDArray delta, int[] kernel, int[] strides, int[] pad, INDArray biasGradView, INDArray weightGradView, IActivation afn, ConvolutionLayer.AlgoMode mode, ConvolutionLayer.BwdFilterAlgo bwdFilterAlgo, ConvolutionLayer.BwdDataAlgo bwdDataAlgo, ConvolutionMode convolutionMode, int[] dilation, LayerWorkspaceMgr workspaceMgr)
boolean	checkSupported()
Map<String,Long>	helperMemoryUse() Return the currently allocated memory for the helper. (a) Excludes: any shared memory used by multiple helpers/layers (b) Excludes any temporary memory (c) Includes all memory that persists for longer than the helper method This is mainly used for debugging and reporting purposes.
INDArray	preOutput(INDArray input, INDArray weights, INDArray bias, int[] kernel, int[] strides, int[] pad, ConvolutionLayer.AlgoMode mode, ConvolutionLayer.FwdAlgo fwdAlgo, ConvolutionMode convolutionMode, int[] dilation, LayerWorkspaceMgr workspaceMgr)

Methods inherited from class java.lang.Object

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

Field Detail

context

protected OpContext context

contextBwd

protected OpContext contextBwd

Constructor Detail

MKLDNNConvHelper

public MKLDNNConvHelper(org.nd4j.linalg.api.buffer.DataType dataType)

Method Detail

checkSupported

public boolean checkSupported()

Specified by:

checkSupported in interface ConvolutionHelper

backpropGradient

public Pair<Gradient,INDArray> backpropGradient(INDArray input,
                                                INDArray weights,
                                                INDArray bias,
                                                INDArray delta,
                                                int[] kernel,
                                                int[] strides,
                                                int[] pad,
                                                INDArray biasGradView,
                                                INDArray weightGradView,
                                                IActivation afn,
                                                ConvolutionLayer.AlgoMode mode,
                                                ConvolutionLayer.BwdFilterAlgo bwdFilterAlgo,
                                                ConvolutionLayer.BwdDataAlgo bwdDataAlgo,
                                                ConvolutionMode convolutionMode,
                                                int[] dilation,
                                                LayerWorkspaceMgr workspaceMgr)

Specified by:

backpropGradient in interface ConvolutionHelper

preOutput

public INDArray preOutput(INDArray input,
                          INDArray weights,
                          INDArray bias,
                          int[] kernel,
                          int[] strides,
                          int[] pad,
                          ConvolutionLayer.AlgoMode mode,
                          ConvolutionLayer.FwdAlgo fwdAlgo,
                          ConvolutionMode convolutionMode,
                          int[] dilation,
                          LayerWorkspaceMgr workspaceMgr)

Specified by:

preOutput in interface ConvolutionHelper

activate

public INDArray activate(INDArray z,
                         IActivation afn,
                         boolean training)

Specified by:

activate in interface ConvolutionHelper

helperMemoryUse

public Map<String,Long> helperMemoryUse()

Description copied from interface: LayerHelper

Return the currently allocated memory for the helper.
(a) Excludes: any shared memory used by multiple helpers/layers
(b) Excludes any temporary memory (c) Includes all memory that persists for longer than the helper method
This is mainly used for debugging and reporting purposes. Returns a map:
Key: The name of the type of memory
Value: The amount of memory

Specified by:

helperMemoryUse in interface LayerHelper

Returns:

Map of memory, may be null if none is used.