tf.test.is_built_with_xla  |  TensorFlow v2.16.1 (original) (raw)

tf.test.is_built_with_xla

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

Returns whether TensorFlow was built with XLA support.

View aliases

Compat aliases for migration

SeeMigration guide for more details.

tf.compat.v1.test.is_built_with_xla

tf.test.is_built_with_xla()

This method should only be used in tests written with tf.test.TestCase. A typical usage is to skip tests that should only run with XLA.

class MyTest(tf.test.TestCase): `` def test_add_on_xla(self): if not tf.test.is_built_with_xla(): self.skipTest("test is only applicable on XLA")

... @tf.function(jit_compile=True) ... def add(x, y): ... return tf.math.add(x, y) ... ... self.assertEqual(add(tf.ones(()), tf.ones(())), 2.0)

TensorFlow official binary is built with XLA.

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.