LostTech.TensorFlow : API Documentation

Type ClusterSpec

Namespace tensorflow.train

Parent PythonObjectContainer

Interfaces IClusterSpec

Represents a cluster as a set of "tasks", organized into "jobs".

A tf.train.ClusterSpec represents the set of processes that participate in a distributed TensorFlow computation. Every tf.distribute.Server is constructed in a particular cluster.

To create a cluster with two jobs and five tasks, you specify the mapping from job names to lists of network addresses (typically hostname-port pairs). Each job may also be specified as a sparse mapping from task indices to network addresses. This enables a server to be configured without needing to know the identity of (for example) all other worker tasks:
Show Example 
cluster = tf.train.ClusterSpec({"worker": ["worker0.example.com:2222",
                                                       "worker1.example.com:2222",
                                                       "worker2.example.com:2222"],
                                            "ps": ["ps0.example.com:2222",
                                                   "ps1.example.com:2222"]})

Methods

Properties

Public instance methods

object as_cluster_def()

Returns a tf.train.ClusterDef protocol buffer based on this cluster.

object as_cluster_def_dyn()

Returns a tf.train.ClusterDef protocol buffer based on this cluster.

IList<object> job_tasks(string job_name)

Returns a mapping from task ID to address in the given job.

NOTE: For backwards compatibility, this method returns a list. If the given job was defined with a sparse set of task indices, the length of this list may not reflect the number of tasks defined in this job. Use the tf.train.ClusterSpec.num_tasks method to find the number of tasks defined in a particular job.
Parameters
string job_name
The string name of a job in this cluster.
Returns
IList<object>
A list of task addresses, where the index in the list corresponds to the task index of each task. The list may contain `None` if the job was defined with a sparse set of task indices.

object job_tasks_dyn(object job_name)

Returns a mapping from task ID to address in the given job.

NOTE: For backwards compatibility, this method returns a list. If the given job was defined with a sparse set of task indices, the length of this list may not reflect the number of tasks defined in this job. Use the tf.train.ClusterSpec.num_tasks method to find the number of tasks defined in a particular job.
Parameters
object job_name
The string name of a job in this cluster.
Returns
object
A list of task addresses, where the index in the list corresponds to the task index of each task. The list may contain `None` if the job was defined with a sparse set of task indices.

int num_tasks(string job_name)

Returns the number of tasks defined in the given job.
Parameters
string job_name
The string name of a job in this cluster.
Returns
int
The number of tasks defined in the given job.

object num_tasks_dyn(object job_name)

Returns the number of tasks defined in the given job.
Parameters
object job_name
The string name of a job in this cluster.
Returns
object
The number of tasks defined in the given job.

object task_address(string job_name, Nullable<int> task_index)

Returns the address of the given task in the given job.
Parameters
string job_name
The string name of a job in this cluster.
Nullable<int> task_index
A non-negative integer.
Returns
object
The address of the given task in the given job.

object task_address_dyn(object job_name, object task_index)

Returns the address of the given task in the given job.
Parameters
object job_name
The string name of a job in this cluster.
object task_index
A non-negative integer.
Returns
object
The address of the given task in the given job.

IList<object> task_indices(string job_name)

Returns a list of valid task indices in the given job.
Parameters
string job_name
The string name of a job in this cluster.
Returns
IList<object>
A list of valid task indices in the given job.

object task_indices_dyn(object job_name)

Returns a list of valid task indices in the given job.
Parameters
object job_name
The string name of a job in this cluster.
Returns
object
A list of valid task indices in the given job.

Public properties

IList<object> jobs get;

Returns a list of job names in this cluster.

object jobs_dyn get;

Returns a list of job names in this cluster.

object PythonObject get;