org.deeplearning4j.optimize

Class GradientAdjustment

java.lang.Object

org.deeplearning4j.optimize.GradientAdjustment

public class GradientAdjustment
extends Object

Gradient adjustment

Author:

Adam Gibson

Method Summary

Modifier and Type	Method and Description
static void	updateGradientAccordingToParams(NeuralNetConfiguration conf, int iteration, org.nd4j.linalg.learning.AdaGrad adaGrad, org.nd4j.linalg.api.ndarray.INDArray gradient, org.nd4j.linalg.api.ndarray.INDArray params, int batchSize) Update the gradient according to the configuration such as adagrad, momentum, and sparsity
static void	updateGradientAccordingToParams(NeuralNetConfiguration conf, int iteration, Gradient gradient, int batchSize, Map<String,org.nd4j.linalg.learning.AdaGrad> adaGrad, Model model) Updates each variable wrt its gradient

Methods inherited from class java.lang.Object

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

Method Detail

updateGradientAccordingToParams

public static void updateGradientAccordingToParams(NeuralNetConfiguration conf,
                                                   int iteration,
                                                   Gradient gradient,
                                                   int batchSize,
                                                   Map<String,org.nd4j.linalg.learning.AdaGrad> adaGrad,
                                                   Model model)

Updates each variable wrt its gradient

Parameters:

conf - the configuration

iteration - the iteration

gradient - the gradients for the variables

batchSize - the batch size of the input

adaGrad - the adagrad map (per variable adagrad entries(

model - the model to use

updateGradientAccordingToParams

public static void updateGradientAccordingToParams(NeuralNetConfiguration conf,
                                                   int iteration,
                                                   org.nd4j.linalg.learning.AdaGrad adaGrad,
                                                   org.nd4j.linalg.api.ndarray.INDArray gradient,
                                                   org.nd4j.linalg.api.ndarray.INDArray params,
                                                   int batchSize)

Update the gradient according to the configuration such as adagrad, momentum, and sparsity

Parameters:

gradient - the gradient to modify