ThreadPoolHandle

public final class ThreadPoolHandle

Creates a dataset that uses a custom thread pool to compute `input_dataset`.

Nested Classes

class ThreadPoolHandle.Options Optional attributes for ThreadPoolHandle  

Public Methods

Output<Object>
asOutput()
Returns the symbolic handle of a tensor.
static ThreadPoolHandle.Options
container(String container)
static ThreadPoolHandle
create(Scope scope, Long numThreads, String displayName, Options... options)
Factory method to create a class wrapping a new ThreadPoolHandle operation.
Output<?>
handle()
A resource that can be consumed by one or more ExperimentalThreadPoolDataset ops.
static ThreadPoolHandle.Options
maxIntraOpParallelism(Long maxIntraOpParallelism)
static ThreadPoolHandle.Options
sharedName(String sharedName)

Inherited Methods

Public Methods

public Output<Object> 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 ThreadPoolHandle.Options container (String container)

public static ThreadPoolHandle create (Scope scope, Long numThreads, String displayName, Options... options)

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

Parameters
scope current scope
numThreads The number of threads in the thread pool.
displayName A human-readable name for the threads that may be visible in some visualizations. threadpool.
options carries optional attributes values
Returns
  • a new instance of ThreadPoolHandle

public Output<?> handle ()

A resource that can be consumed by one or more ExperimentalThreadPoolDataset ops.

public static ThreadPoolHandle.Options maxIntraOpParallelism (Long maxIntraOpParallelism)

Parameters
maxIntraOpParallelism The maximum degree of parallelism to use within operations that execute on this threadpool.

public static ThreadPoolHandle.Options sharedName (String sharedName)