Have a question? Connect with the community at the TensorFlow Forum Visit Forum

RebatchDataset

public final class RebatchDataset

Creates a dataset that changes the batch size.

Creates a dataset that changes the batch size of the dataset to current batch size // num_replicas.

Nested Classes

class RebatchDataset.Options Optional attributes for RebatchDataset

Constants

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

Public Methods

Output < TType >
asOutput ()
Returns the symbolic handle of the tensor.
static RebatchDataset
create ( Scope scope, Operand <?> inputDataset, Operand < TInt64 > numReplicas, List<Class<? extends TType >> outputTypes, List< Shape > outputShapes, Options... options)
Factory method to create a class wrapping a new RebatchDataset operation.
Output <?>
static RebatchDataset.Options
useFallback (Boolean useFallback)

Inherited Methods

Constants

public static final String OP_NAME

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

Constant Value: "ExperimentalRebatchDataset"

Public Methods

public Output < TType > asOutput ()

Returns the symbolic handle of the tensor.

Inputs to TensorFlow operations are outputs of another TensorFlow operation. This method is used to obtain a symbolic handle that represents the computation of the input.

public static RebatchDataset create ( Scope scope, Operand <?> inputDataset, Operand < TInt64 > numReplicas, List<Class<? extends TType >> outputTypes, List< Shape > outputShapes, Options... options)

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

Parameters
scope current scope
inputDataset A variant tensor representing the input dataset.
numReplicas A scalar representing the number of replicas to distribute this batch across. As a result of this transformation the current batch size would end up being divided by this parameter.
options carries optional attributes values
Returns
  • a new instance of RebatchDataset

public Output <?> handle ()

public static RebatchDataset.Options useFallback (Boolean useFallback)