View source on GitHub
|
Returns the preferred device type for the accelerators.
tf.experimental.dtensor.preferred_device_type() -> str
The returned device type is determined by checking the first present device type from all supported device types in the order of 'TPU', 'GPU', 'CPU'.
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.