Help protect the Great Barrier Reef with TensorFlow on Kaggle Join Challenge


Cluster Resolver for Google Cloud TPUs.

Inherits From: ClusterResolver

Used in the notebooks

Used in the guide Used in the tutorials

This is an implementation of cluster resolvers for the Google Cloud TPU service.

TPUClusterResolver supports the following distinct environments: Google Compute Engine Google Kubernetes Engine Google internal

It can be passed into tf.distribute.TPUStrategy to support TF2 training on Cloud TPUs.

tpu A string corresponding to the TPU to use. It can be the TPU name or TPU worker gRPC address. If not set, it will try automatically resolve the TPU address on Cloud TPUs. If set to "local", it will assume that the TPU is directly connected to the VM instead of over the network.
zone Zone where the TPUs are located. If omitted or empty, we will assume that the zone of the TPU is the same as the zone of the GCE VM, which we will try to discover from the GCE metadata service.
project Name of the GCP project containing Cloud TPUs. If omitted or empty, we will try to discover the project name of the GCE VM from the GCE metadata service.
job_name Name of the TensorFlow job the TPUs belong to.
coordinator_name The name to use for the coordinator. Set to None if the coordinator should not be included in the computed ClusterSpec.
coordinator_address The address of the coordinator (typically an ip:port pair). If set to None, a TF server will be started. If coordinator_name is None, a TF server will not be started even if coordinator_address is None.
credentials GCE Credentials. If None, then we use default credentials from the oauth2client
service The GCE API object returned by the googleapiclient.discovery function. If you specify a custom service object, then the credentials parameter will be ignored.
discovery_url A URL template that points to the location of the discovery service. It should have two parameters {api} and {apiVersion} that when filled in produce an absolute URL to the discovery document for that service. The environment variable 'TPU_API_DISCOVERY_URL' will override this.

ImportError If the googleapiclient is not installed.
ValueError If no TPUs are specified.
RuntimeError If an empty TPU name is specified and this is running in a Google Cloud environment.

environment Returns the current environment which TensorFlow is running in.
task_id Returns the task id this ClusterResolver indicates.

In TensorFlow distributed environment, each job may have an applicable task id, which is the index of the instance within its task type. This is useful when user needs to run specific code according to task index. For example,

cluster_spec = tf.train.ClusterSpec({
    "ps": ["localhost:2222", "localhost:2223"],
    "worker": ["localhost:2224", "localhost:2225", "localhost:2226"]

# SimpleClusterResolver is used here for illustration; other cluster
# resolvers may be used for other source of task type/id.
simple_resolver = SimpleClusterResolver(cluster_spec, task_type="worker",


if cluster_resolver.task_type == 'worker' and cluster_resolver.task_id == 0:
  # Perform something that's only applicable on 'worker' type, id 0. This
  # block will run on this particular instance since we've specified this
  # task to be a 'worker', id 0 in above cluster resolver.
  # Perform something that's only applicable on other ids. This block will
  # not run on this particular instance.

Returns None if such information is not available or is not applicable in the current distributed environment, such as training with tf.distribute.cluster_resolver.TPUClusterResolver.

For more information, please see tf.distribute.cluster_resolver.ClusterResolver's class docstring.

task_type Returns the task type this ClusterResolver indicates.

In TensorFlow distributed environment, each job may have an applicable task type. Valid task types in TensorFlow include 'chief': a worker that is designated with more responsibility, 'worker': a regular worker for training/evaluation, 'ps': a parameter server, or 'evaluator': an evaluator that evaluates the checkpoints for metrics.

See Multi-worker configuration for more information about 'chief' and &#