ReplicateMetadata

public final class ReplicateMetadata

Metadata indicating how the TPU computation should be replicated.

This operation holds the metadata common to operations of a `tpu.replicate()` computation subgraph.

Nested Classes

class

ReplicateMetadata.Options

Optional attributes for ReplicateMetadata

Constants

String

OP_NAME

The name of this op, as known by TensorFlow core engine

Public Methods

static ReplicateMetadata.Options	allowSoftPlacement(Boolean allowSoftPlacement)
static ReplicateMetadata.Options	computationShape(List<Long> computationShape)
static ReplicateMetadata	create(Scope scope, Long numReplicas, Options... options) Factory method to create a class wrapping a new ReplicateMetadata operation.
static ReplicateMetadata.Options	deviceAssignment(List<Long> deviceAssignment)
static ReplicateMetadata.Options	hostComputeCore(List<String> hostComputeCore)
static ReplicateMetadata.Options	numCoresPerReplica(Long numCoresPerReplica)
static ReplicateMetadata.Options	paddingMap(List<String> paddingMap)
static ReplicateMetadata.Options	stepMarkerLocation(String stepMarkerLocation)
static ReplicateMetadata.Options	topology(String topology)
static ReplicateMetadata.Options	useSpmdForXlaPartitioning(Boolean useSpmdForXlaPartitioning)
static ReplicateMetadata.Options	useTpu(Boolean useTpu)

Inherited Methods

From class org.tensorflow.op.RawOp

final boolean	equals(Object obj)
final int	hashCode()
Operation	op() Return this unit of computation as a single Operation.
final String	toString()

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.op.Op

abstract ExecutionEnvironment	env() Return the execution environment this op was created in.
abstract Operation	op() Return this unit of computation as a single Operation.

Constants

public static final String OP_NAME

The name of this op, as known by TensorFlow core engine

Constant Value: "TPUReplicateMetadata"

Public Methods

public static ReplicateMetadata.Options allowSoftPlacement (Boolean allowSoftPlacement)

public static ReplicateMetadata.Options computationShape (List<Long> computationShape)

Parameters

computationShape	DEPRECATED. Use num_cores_per_replica instead.

public static ReplicateMetadata create (Scope scope, Long numReplicas, Options... options)

Factory method to create a class wrapping a new ReplicateMetadata operation.

Parameters

scope	current scope
numReplicas	Number of replicas of the computation
options	carries optional attributes values

Returns

• a new instance of ReplicateMetadata

public static ReplicateMetadata.Options deviceAssignment (List<Long> deviceAssignment)

Parameters

deviceAssignment	The assignment of devices for the computation.

public static ReplicateMetadata.Options hostComputeCore (List<String> hostComputeCore)

public static ReplicateMetadata.Options numCoresPerReplica (Long numCoresPerReplica)

Parameters

numCoresPerReplica	Number of cores per replica. Used for model parallelism.

public static ReplicateMetadata.Options paddingMap (List<String> paddingMap)

public static ReplicateMetadata.Options stepMarkerLocation (String stepMarkerLocation)

public static ReplicateMetadata.Options topology (String topology)

Parameters

topology	TopologyProto indicating the topology of the TPU pod slice.

public static ReplicateMetadata.Options useSpmdForXlaPartitioning (Boolean useSpmdForXlaPartitioning)

public static ReplicateMetadata.Options useTpu (Boolean useTpu)

Parameters

useTpu	Whether to place the computation on the TPU.