Graph

public final class Graph

A data flow graph representing a TensorFlow computation.

Instances of a Graph are thread-safe.

WARNING: Resources consumed by the Graph object must be explicitly freed by invoking the close() method then the Graph object is no longer needed.

Nested Classes

interface

Graph.WhileSubgraphBuilder

Used to instantiate an abstract class which overrides the buildSubgraph method to build a conditional or body subgraph for a while loop.

Public Constructors

Graph() Create an empty Graph.

Public Methods

Output[]<?>	addGradients(Output<?> y, Output[]<?> x) Adds operations to compute the partial derivatives of sum of ys w.r.t xs, i.e., dy/dx_1, dy/dx_2... This is a simplified version of addGradients(String, Output[], Output[], Output[]) where y is a single output, dx is null and prefix is null.
Output[]<?>	addGradients(String prefix, Output[]<?> y, Output[]<?> x, Output[]<?> dx) Adds operations to compute the partial derivatives of sum of ys w.r.t xs, i.e., d(y_1 + y_2 + ...)/dx_1, d(y_1 + y_2 + ...)/dx_2... dx are used as initial gradients (which represent the symbolic partial derivatives of some loss function L w.r.t.
synchronized void	addInitializer(Op initializer) Adds an initializer to the graph initializer list.
void	checkInput(Op input) Checks that input is valid to use as an input in this execution environment.
void	close() Release resources associated with the Graph.
ExecutionEnvironment.Types	environmentType() Get the type of this environment (from the `Environments` enumeration.
void	importGraphDef(GraphDef graphDef) Import a representation of a TensorFlow graph.
void	importGraphDef(GraphDef graphDef, String prefix) Import a representation of a TensorFlow graph.
List<Op>	initializers() Returns all initializers added to the graph via addInitializer(Op)
GraphOperationBuilder	opBuilder(String type, String name) Returns a builder to add Operations to the Graph.
GraphOperation	operation(String name) Returns the operation (node in the Graph) with the provided name.
Iterator<Operation>	operations() Iterator over all the Operations in the graph.
GraphDef	toGraphDef() Generate a representation of the Graph.
Output[]<?>	whileLoop(Output[]<?> inputs, Graph.WhileSubgraphBuilder cgBuilder, Graph.WhileSubgraphBuilder bgBuilder, String name) Builds a while loop.

Inherited Methods

From class java.lang.Object

boolean	equals(Object arg0)
final Class<?>	getClass()
int	hashCode()
final void	notify()
final void	notifyAll()
String	toString()
final void	wait(long arg0, int arg1)
final void	wait(long arg0)
final void	wait()

From interface org.tensorflow.ExecutionEnvironment

abstract void	checkInput(Op input) Checks that input is valid to use as an input in this execution environment.
abstract ExecutionEnvironment.Types	environmentType() Get the type of this environment (from the `Environments` enumeration.
abstract boolean	isEager()
abstract boolean	isGraph()
abstract boolean	isOpEnabled(String opType) Returns true if the given operation is valid in this execution environment.
abstract OperationBuilder	opBuilder(String type, String name) Returns a builder to create a new Operation.

From interface java.lang.AutoCloseable

abstract void

close()

Public Constructors

public Graph ()

Create an empty Graph.

Public Methods

public Output[]<?> addGradients (Output<?> y, Output[]<?> x)

Adds operations to compute the partial derivatives of sum of ys w.r.t xs, i.e., dy/dx_1, dy/dx_2...

This is a simplified version of addGradients(String, Output[], Output[], Output[]) where y is a single output, dx is null and prefix is null.

Parameters

y	output of the function to derive
x	inputs of the function for which partial derivatives are computed

Returns

• the partial derivatives dy with the size of x

public Output[]<?> addGradients (String prefix, Output[]<?> y, Output[]<?> x, Output[]<?> dx)

Adds operations to compute the partial derivatives of sum of ys w.r.t xs, i.e., d(y_1 + y_2 + ...)/dx_1, d(y_1 + y_2 + ...)/dx_2...

dx are used as initial gradients (which represent the symbolic partial derivatives of some loss function L w.r.t. y). dx must be null or have size of y.

If dx is null, the implementation will use dx of OnesLike for all shapes in y.

prefix is used as the name prefix applied to all nodes added to the graph to compute gradients. It must be unique within the provided graph or the operation will fail.

If prefix is null, then one will be chosen automatically.

Parameters

prefix	unique string prefix applied before the names of nodes added to the graph to compute gradients. If null, a default one will be chosen.
y	output of the function to derive
x	inputs of the function for which partial derivatives are computed
dx	if not null, the partial derivatives of some loss function L w.r.t. y

Returns

• the partial derivatives dy with the size of x

public synchronized void addInitializer (Op initializer)

Adds an initializer to the graph initializer list.

Parameters

initializer	An initializer to add to the list.

public void checkInput (Op input)

Checks that input is valid to use as an input in this execution environment. Throws IllegalArgumentException if not.

Parameters

input	The op to check

public void close ()

Release resources associated with the Graph.

Blocks until there are no active Session instances referring to this Graph. A Graph is not usable after close returns.

public ExecutionEnvironment.Types environmentType ()

Get the type of this environment (from the `Environments` enumeration.

Returns

• An `Environments` value indicating the type of execution environment.

public void importGraphDef (GraphDef graphDef)

Import a representation of a TensorFlow graph.

The representation of the graph, referred to as a GraphDef, can be generated by toGraphDef() and equivalents in other language APIs.

Parameters

graphDef	GraphDef proto to import

Throws

IllegalArgumentException	if graphDef is not a recognized serialization of a graph.

See Also

• importGraphDef(GraphDef, String)

public void importGraphDef (GraphDef graphDef, String prefix)

Import a representation of a TensorFlow graph.

Parameters

graphDef	the representation of a TensorFlow graph.
prefix	a prefix that will be prepended to names in graphDef

Throws

IllegalArgumentException	if graphDef is not a recognized serialization of a graph.

See Also

• importGraphDef(GraphDef)

public List<Op> initializers ()

Returns all initializers added to the graph via addInitializer(Op)

public GraphOperationBuilder opBuilder (String type, String name)

Returns a builder to add Operations to the Graph.

Parameters

type	of the Operation (i.e., identifies the computation to be performed)
name	to refer to the created Operation in the graph.

Returns

• an OperationBuilder, which will add the Operation to the graph when build() is invoked. If build() is not invoked, then some resources may leak.

public GraphOperation operation (String name)

Returns the operation (node in the Graph) with the provided name.

Or null if no such operation exists in the Graph.

Parameters

name	name of the operation to look for

Returns

• operation in the graph with this name or null if not found

public Iterator<Operation> operations ()

Iterator over all the Operations in the graph.

The order of iteration is unspecified. Consumers of the iterator will receive no notification should the underlying graph change during iteration.

public GraphDef toGraphDef ()

Generate a representation of the Graph.

See Also

• importGraphDef(GraphDef)
• importGraphDef(GraphDef, String)

public Output[]<?> whileLoop (Output[]<?> inputs, Graph.WhileSubgraphBuilder cgBuilder, Graph.WhileSubgraphBuilder bgBuilder, String name)

Builds a while loop.

Parameters

inputs	the loop inputs
cgBuilder	WhileSubgraphBuilder to build the conditional subgraph
bgBuilder	WhileSubgraphBuilder to build the body subgraph
name	name for the loop

Returns

• list of loop outputs, of the same length as inputs