tf.experimental.dtensor.copy_to_mesh

本页内容
Args
Returns

Copies a tf.Tensor onto the DTensor device with the given layout.

tf.experimental.dtensor.copy_to_mesh(
    tensor: Any,
    layout: tf.experimental.dtensor.Layout,
    source_layout: Optional[tf.experimental.dtensor.Layout] = None
) -> tf.Tensor

Copies a regular tf.Tensor onto the DTensor device. Use the mesh attached to layout as target mesh. This method currently only supports replicated layouts. To get a DTensor with a sharded layout, use the pack method.

Args
`tensor`	A regular tf.Tensor to be copied as a DTensor.
`layout`	Target layout (and mesh) for the result DTensor.
`source_layout`	Source layout of the tensor before copy, used for backward passes.

Returns
A DTensor on the DTensor device with the given layout.

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 2023-03-23 UTC.

tf.experimental.dtensor.copy_to_mesh

Args

Returns