Class GradientDescent

java.lang.Object

org.tensorflow.framework.optimizers.Optimizer

org.tensorflow.framework.optimizers.GradientDescent

public class GradientDescent
extends Optimizer

Basic Stochastic gradient descent optimizer. GradientDescent updates the current weight using the current gradient ?L/?w multiplied by the learning rate.

Nested Class Summary

Nested classes/interfaces inherited from class Optimizer

Optimizer.GradAndVar<T>, Optimizer.Options

Modifier and Type	Class	Description
static class	Optimizer.GradAndVar<T extends TType>	A class that holds a paired gradient and variable.
static class	Optimizer.Options	Optional attributes for Optimizer

Field Summary

Fields

Modifier and Type	Field	Description
static final float	LEARNING_RATE_DEFAULT

Fields inherited from class Optimizer

globals, graph, tf, VARIABLE_V2

Modifier and Type	Field	Description
protected final List<Variable<?>>	globals	Global state variables
protected final Graph	graph	The Graph this optimizer is operating on.
protected final Ops	tf	The ops builder for the graph.
static final String	VARIABLE_V2

Constructor Summary

Constructors

Constructor	Description
GradientDescent(Graph graph)	Creates a GradientDescent Optimizer
GradientDescent(Graph graph, float learningRate)	Creates a GradientDescent Optimizer
GradientDescent(Graph graph, String name, float learningRate)	Creates a GradientDescent Optimizer

Method Summary

Modifier and Type	Method	Description
protected <T extends TType> Op	applyDense(Ops deps, Output<T> gradient, Output<T> variable)	Generates the gradient update operations for the specific variable and gradient.
String	getOptimizerName()	Get the Name of the optimizer.
String	toString()

Methods inherited from class Optimizer

applyGradients, computeGradients, createName, createSlot, createSlots, finish, getSlot, getTF, minimize, minimize, prepare

Modifier and Type	Method	Description
Op	applyGradients(List<Optimizer.GradAndVar<? extends TType>> gradsAndVars, String name)	Applies gradients to variables
<T extends TType> List<Optimizer.GradAndVar<?>>	computeGradients(Operand<?> loss)	Computes the gradients based on a loss operand.
static String	createName(Output<? extends TType> variable, String slotName)	Creates a name by combining a variable name and a slot name
protected <T extends TType> void	createSlot(Output<T> variable, String slotName, Operand<T> initializer)	Creates a slot in the graph for the specified variable with the specified name.
protected void	createSlots(List<Output<? extends TType>> variables)	Performs a No-op slot creation method.
protected Op	finish(List<Op> updateOperations, String name)	Gathers up the update operations into a single op that can be used as a run target.
<T extends TType> Optional<Variable<T>>	getSlot(Output<T> var, String slotName)	Gets the slot associated with the specified variable and slot name.
final Ops	getTF()	Gets the Optimizer's Ops instance
Op	minimize(Operand<?> loss)	Minimizes the loss by updating the variables
Op	minimize(Operand<?> loss, String name)	Minimizes the loss by updating the variables
protected Optional<Op>	prepare(String scopeName)	Returns a No-op prepare.

Methods inherited from class Object

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

Field Details

LEARNING_RATE_DEFAULT

public static final float LEARNING_RATE_DEFAULT

See Also:

• Constant Field Values

Constructor Details

GradientDescent

public GradientDescent(Graph graph)

Creates a GradientDescent Optimizer

Parameters:

graph - the TensorFlow graph

GradientDescent

public GradientDescent(Graph graph,
 float learningRate)

Creates a GradientDescent Optimizer

Parameters:

graph - the TensorFlow graph

learningRate - the learning rate, defaults to 0.01

GradientDescent

public GradientDescent(Graph graph,
 String name,
 float learningRate)

Creates a GradientDescent Optimizer

Parameters:

graph - the TensorFlow graph

name - the name for this Optimizer, default is "GradientDescent"

learningRate - the learning rate, defaults to 0.01

Method Details

applyDense

protected <T extends TType> Op applyDense(Ops deps,
 Output<T> gradient,
 Output<T> variable)

Generates the gradient update operations for the specific variable and gradient.

Specified by:

applyDense in class Optimizer

Type Parameters:

T - The type of the variable.

Parameters:

gradient - The gradient to use.

variable - The variable to update.

Returns:

An operand which applies the desired optimizer update to the variable.

toString

public String toString()

Overrides:

toString in class Object

getOptimizerName

public String getOptimizerName()

Get the Name of the optimizer.

Specified by:

getOptimizerName in class Optimizer

Returns:

The optimizer name.