|TensorFlow 1 version||View source on GitHub|
An in-process TensorFlow server, for use in distributed training.
Compat aliases for migration
See Migration guide for more details.
tf.distribute.Server( server_or_cluster_def, job_name=None, task_index=None, protocol=None, config=None, start=True )
Used in the notebooks
|Used in the guide||Used in the tutorials|
tf.distribute.Server instance encapsulates a set of devices and a
tf.compat.v1.Session target that
can participate in distributed training. A server belongs to a
cluster (specified by a
corresponds to a particular task in a named job. The server can
communicate with any other server in the same cluster.
(Optional.) Specifies the name of the job of which the server is
a member. Defaults to the value in
(Optional.) Specifies the task index of the server in its job.
Defaults to the value in
(Optional.) Specifies the protocol to be used by the server.
Acceptable values include |