|---| | [ View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v2.21.0/tensorflow/python/ops/stateful_random_ops.py#L1000-L1031) | Replaces the global generator with another `Generator` object. #### View aliases **Main aliases** [`tf.random.experimental.set_global_generator`](https://www.tensorflow.org/api_docs/python/tf/random/set_global_generator) **Compat aliases for migration** See [Migration guide](https://www.tensorflow.org/guide/migrate) for more details. [`tf.compat.v1.random.experimental.set_global_generator`](https://www.tensorflow.org/api_docs/python/tf/random/set_global_generator), [`tf.compat.v1.random.set_global_generator`](https://www.tensorflow.org/api_docs/python/tf/random/set_global_generator)
tf.random.set_global_generator( generator ) This function replaces the global generator with the provided `generator` object. A random number generator utilizes a [`tf.Variable`](https://www.tensorflow.org/api_docs/python/tf/Variable) object to store its state. The user shall be aware of caveats how `set_global_generator` interacts with [`tf.function`](https://www.tensorflow.org/api_docs/python/tf/function): - tf.function puts restrictions on Variable creation thus one cannot freely create a new random generator instance inside [`tf.function`](https://www.tensorflow.org/api_docs/python/tf/function). To call `set_global_generator` inside [`tf.function`](https://www.tensorflow.org/api_docs/python/tf/function), the generator instance must have already been created eagerly. - tf.function captures the Variable during trace-compilation, thus a compiled f.function will not be affected `set_global_generator` as demonstrated by random_test.py/RandomTest.testResetGlobalGeneratorBadWithDefun . For most use cases, avoid calling `set_global_generator` after program initialization, and prefer to reset the state of the existing global generator instead, such as, rng = tf.random.get_global_generator() rng.reset_from_seed(30)


| ## Args || |---|---| | `generator` | the new `Generator` object. |