tf.data.experimental.ThreadingOptions

TensorFlow 1 version View source on GitHub

Represents options for dataset threading.

You can set the threading options of a dataset through the experimental_threading property of tf.data.Options; the property is an instance of tf.data.experimental.ThreadingOptions.

options = tf.data.Options()
options.experimental_threading.private_threadpool_size = 10
dataset = dataset.with_options(options)

max_intra_op_parallelism If set, it overrides the maximum degree of intra-op parallelism.
private_threadpool_size If set, the dataset will use a private threadpool of the given size.

Methods

__eq__

View source

Return self==value.

__ne__

View source

Return self!=value.