TensorFlow 2 version | View source on GitHub |
Represents options for dataset threading.
tf.data.experimental.ThreadingOptions()
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)
Attributes | |
---|---|
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__
__eq__(
other
)
Return self==value.
__ne__
__ne__(
other
)
Return self!=value.