tf.VariableSynchronization  |  TensorFlow v2.16.1 (original) (raw)

tf.VariableSynchronization

Stay organized with collections Save and categorize content based on your preferences.

Indicates when a distributed variable will be synced.

View aliases

Compat aliases for migration

SeeMigration guide for more details.

tf.compat.v1.VariableSynchronization

• AUTO: Indicates that the synchronization will be determined by the currentDistributionStrategy (eg. With MirroredStrategy this would beON_WRITE).
• NONE: Indicates that there will only be one copy of the variable, so there is no need to sync.
• ON_WRITE: Indicates that the variable will be updated across devices every time it is written.
• ON_READ: Indicates that the variable will be aggregated across devices when it is read (eg. when checkpointing or when evaluating an op that uses the variable).
  Example:

>>> temp_grad=[tf.Variable([0.], trainable=False,
...                      synchronization=tf.VariableSynchronization.ON_READ,
...                      aggregation=tf.VariableAggregation.MEAN
...                      )]

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates. Some content is licensed under the numpy license.

Last updated 2024-04-26 UTC.