keras-team/keras · error · ValueError

Cannot create sharding when device mesh is not set for Tenso

Error message

Cannot create sharding when device mesh is not set for TensorLayout.

What it means

Error "Cannot create sharding when device mesh is not set for TensorLayout." thrown in keras-team/keras.

Source

Thrown at keras/src/backend/tensorflow/distribution_lib.py:78

        A `tf.dtensor.Mesh` instance.
    """
    mesh_dims = list(zip(device_mesh.axis_names, device_mesh.shape))
    return dtensor.create_distributed_mesh(
        mesh_dims=mesh_dims, local_devices=device_mesh.devices.flatten()
    )


def _to_backend_layout(tensor_layout):
    """Convert the TensorLayout to Tensorflow backend specific Sharding.

    Args:
        tensor_layout: TensorLayout instance to convert.

    Returns:
        A `tf.dtensor.Layout` instance.
    """
    if tensor_layout.device_mesh is None:
        raise ValueError(
            "Cannot create sharding when device mesh is not set for "
            "TensorLayout."
        )

    sharding_specs = [
        axis if axis else dtensor.UNSHARDED for axis in tensor_layout.axes
    ]
    dtensor_mesh = tensor_layout.device_mesh.backend_mesh
    return dtensor.Layout(sharding_specs=sharding_specs, mesh=dtensor_mesh)

View on GitHub (pinned to 7a34a03db6)

When it happens

Trigger: Thrown at keras/src/backend/tensorflow/distribution_lib.py:78 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/3fd41f6d8b494f31. Report an issue: GitHub.