tf.experimental.dtensor.unpack
Stay organized with collections
Save and categorize content based on your preferences.
Unpacks a DTensor into tf.Tensor components.
tf.experimental.dtensor.unpack(
tensor: Any
) -> Sequence[Any]
Used in the notebooks
Packing and unpacking are inverse operations:
* unpack(pack(tensors)) == tensors
* pack(unpack(dtensor)) == dtensor
- For any DTensor on the mesh,
unpack returns the raw components placed on
each underlying device.
- Packing these raw components in the same order using
pack returns a
DTensor which should be identical to the original DTensor--both the content
value and the layout.
See the documentation for pack for more information about how packing and
unpacking works.
Args |
|---|
tensor
|
The DTensor to unpack.
|
Returns |
|---|
|
The individual component tensors of the DTensor. This will include only the
client-local components, i.e. the components placed on the local devices.
|
Raises |
|---|
RuntimeError
|
When unpack is not called eagerly.
|
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.
[null,null,["Last updated 2024-04-26 UTC."],[],[],null,["# tf.experimental.dtensor.unpack\n\n\u003cbr /\u003e\n\n|---------------------------------------------------------------------------------------------------------------------------|\n| [View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v2.16.1/tensorflow/dtensor/python/api.py#L342-L372) |\n\nUnpacks a DTensor into [`tf.Tensor`](../../../tf/Tensor) components. \n\n tf.experimental.dtensor.unpack(\n tensor: Any\n ) -\u003e Sequence[Any]\n\n### Used in the notebooks\n\n| Used in the guide |\n|-------------------------------------------------------------------------|\n| - [DTensor concepts](https://www.tensorflow.org/guide/dtensor_overview) |\n\nPacking and unpacking are inverse operations: \n\n * unpack(pack(tensors)) == tensors\n * pack(unpack(dtensor)) == dtensor\n\n1. For any DTensor on the mesh, `unpack` returns the raw components placed on each underlying device.\n2. Packing these raw components in the same order using `pack` returns a DTensor which should be identical to the original DTensor--both the content value and the layout.\n\nSee the documentation for `pack` for more information about how packing and\nunpacking works.\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Args ---- ||\n|----------|------------------------|\n| `tensor` | The DTensor to unpack. |\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Returns ------- ||\n|---|---|\n| The individual component tensors of the DTensor. This will include only the client-local components, i.e. the components placed on the local devices. ||\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Raises ------ ||\n|----------------|--------------------------------------|\n| `RuntimeError` | When `unpack` is not called eagerly. |\n\n\u003cbr /\u003e"]]
View source on GitHub