TPUPartitionedInput

public final class TPUPartitionedInput

An op that groups a list of partitioned inputs together. This op

Nested Classes

class TPUPartitionedInput.Options Optional attributes for TPUPartitionedInput  

Public Methods

Output<T>
asOutput()
Returns the symbolic handle of a tensor.
static <T> TPUPartitionedInput<T>
create(Scope scope, Iterable<Operand<T>> inputs, Options... options)
Factory method to create a class wrapping a new TPUPartitionedInput operation.
Output<T>
output()
A handle which represents the full shape of partitioned tensors.
static TPUPartitionedInput.Options
partitionDim(Long partitionDim)

Inherited Methods

Public Methods

public Output<T> asOutput ()

Returns the symbolic handle of a 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 TPUPartitionedInput<T> create (Scope scope, Iterable<Operand<T>> inputs, Options... options)

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

Parameters
scope current scope
inputs A list of partitioned inputs which must have the same shape.
options carries optional attributes values
Returns
  • a new instance of TPUPartitionedInput

public Output<T> output ()

A handle which represents the full shape of partitioned tensors.

public static TPUPartitionedInput.Options partitionDim (Long partitionDim)

Parameters
partitionDim An integer describles which dimension is partitioned. -1 means those inputs are replicated.