Public

Get the Layer (if any). Returns null if #hasLayer() == false

getLayer

getVertexName

A representation of the vertices that are inputs to this vertex (inputs duing
forward pass)
Specific

getInputVertices

Get the array of inputs previously set for this GraphVertex

getInputs

This method clears inpjut for this vertex

clearVertex

doBackward

doForward

feedForwardMaskArrays

Get the epsilon/error (i.e., dL/dOutput) array previously set for this
GraphVertex

getEpsilon

A representation of the vertices that this vertex is connected to (outputs duing
forward pass) Speci

getOutputVertices

getVertexIndex

Whether the GraphVertex contains a Layer object or not

hasLayer

Whether the GraphVertex is an input vertex

isInputVertex

Whether the GraphVertex is an output vertex

isOutputVertex

See Layer#setBackpropGradientsViewArray(INDArray)

setBackpropGradientsViewArray

Set the errors (epsilon - aka dL/dActivation) for this GraphVertex

setEpsilon

setInput

setInputVertices

Only applies to layer vertices. Will throw exceptions on others. If applied to a
layer vertex it wil

setLayerAsFrozen

setOutputVertices

BufferedReader

Finding current android device location

 A Java virtual machine.

 A VirtualMachine represents a Java virtual machine to which this Java vir

com.sun.tools.attach.VirtualMachine

An immutable arbitrary-precision signed decimal.A value is represented by an
arbitrary-precision "un

java.math.BigDecimal

Formats and parses dates in a locale-sensitive manner. Formatting turns a Date
into a String, and pa

java.text.SimpleDateFormat

An iterator over a sequence of objects, such as a collection.If a collection has
been changed since 

java.util.Iterator

A parser that parses a text string of primitive types and strings with the help
of regular expressio

java.util.Scanner

Reflections one-stop-shop objectReflections scans your classpath, indexes the
metadata, allows you t

org.reflections.Reflections

getResourceAsStream

scheduleAtFixedRate

runOnUiThread

/**
     * Do backprop (gradient calculation)
     *
     * @param truncatedBPTT    false: normal backprop. true: calculate gradients using truncated BPTT for RNN layers
     * @param externalEpsilons null usually (for typical supervised learning). If not null (and length > 0) then assume that
     *                         the user has provided some errors externally, as they would do for example in reinforcement
     *                         learning situations.
     */

//: Nd4j.getWorkspaceManager().getWorkspaceForCurrentThread(wsConf, workspaceBackProp);

//Do backprop according to the reverse of the topological ordering of the network

//If true: already set epsilon for this vertex; later epsilons should be *added* to the existing one, not set

//FIXME: make the frozen vertex feature extraction more flexible

//Two reasons for a vertex to be an output vertex:

//(a) it's an output layer (i.e., instanceof IOutputLayer), or

//(b) it's a normal layer, but it has been marked as an output layer for use in external errors - for reinforcement learning, for example

"(but isn't an IOutputLayer). Only IOutputLayer layers can be fit via backprop with"

//Set epsilons for the vertices that provide inputs to this vertex:

//This vertex: must output to multiple vertices... we want to add the epsilons here

//TODO: in some circumstances, it may be safe  to do in-place add (but not always)

//try (MemoryWorkspace wsB = Nd4j.getMemoryManager().scopeOutOfWorkspaces()) {

//Now, add the gradients in the order we need them in for flattening (same as params order)

How to use doBackwardmethodin org.deeplearning4j.nn.graph.vertex.GraphVertex

Best Java code snippets using org.deeplearning4j.nn.graph.vertex.GraphVertex.doBackward (Showing top 1 results out of 315)

How to use
doBackward
method
in
org.deeplearning4j.nn.graph.vertex.GraphVertex