PiperOrigin-RevId: 505787358

1. Both `value_serialization` and `executor_factory` were defining the symbol `CardinalitiesType` and they were defining it differently. Removed the `value_serialization.CardinalitiesType` symbol.

2. A `grpc.Call`, not a `grpc.RpcError`, has a `code` method. Improve the typing when handling `grpc.RpcError`s by using `TypeGuard`.

3. Added a `typing_extentions` dependency to enable `TypeGuard` in older versions of Python.

PiperOrigin-RevId: 505725528

This change includes:

* Moving metrics typings into their own file. In particular, model libraries now access the typing through `tff.learning.metrics`, rather than the other way around.
* Renaming `finalizer.py` to `keras_finalizer.py`.
* Sorting the BUILD file in alphabetical order.

PiperOrigin-RevId: 505228356

PiperOrigin-RevId: 505201820

Remove the single quote in the type, which breaks the rendering on https://www.tensorflow.org/federated/api_docs/python/tff/learning/templates/DistributionProcess.

PiperOrigin-RevId: 504907703

PiperOrigin-RevId: 504692476

PiperOrigin-RevId: 504677663

PiperOrigin-RevId: 504599751

PiperOrigin-RevId: 504569014

PiperOrigin-RevId: 504560730

PiperOrigin-RevId: 504548024

PiperOrigin-RevId: 504384140

This is one of many steps to deleting the Python execution context.

PiperOrigin-RevId: 504367165

PiperOrigin-RevId: 504358730

`ValueImpl` was merged into `Value` in https://github.com/tensorflow/federated/commit/42c469e5fb99838bd5ba18bf6be738260ed06662

PiperOrigin-RevId: 504320469

Currently the C++ execution contexts do not have a way to return sizing information to the user, so this capability will be removed until we add the functionality to the C++ execution context.

This is one of many steps to deleting the Python execution context.

PiperOrigin-RevId: 504310574

Rename the `cpp_execution_context.(create|set)_local_cpp_execution_context` function to match the name of `execution_context.(create|set)_(sync|async)_local_cpp_execution_context`.

* Renamed the `cpp_execution_context.(create|set)_local_cpp_execution_context` function.
* Fixed internal usage of this API.
* Added test to ensure that these function signatures match.

This is a follow up to cl/500021834.

PiperOrigin-RevId: 504287557

PiperOrigin-RevId: 504281487

This revealed two Python typing-related issues:
1. Every subclass of `executor_base.Executor` was using a property on the corresponding subclass of `executor_value_base.ExecutorValue` without having access to that property. This was fixed by:
2. The methods on the subclasses of `federating_executor.FederatingStrategy` (`FederatedComposingStrategy` and `FederatedResolvingStrategy`) were annotated to return types specific to that subclass. This return type conflicted with the return types of the shared implementations defined in `executor_utils`.
3. The type annotation of `target_executors` parameter of `FederatedResolvingStrategy` was incorrect.

* Enabled pytype in the following modules:
    * `federated_composing_strategy`
    * `federated_resolving_strategy`
    * `federating_executor`
    * `reference_resolving_executor`
    * `remote_executor`
    * `thread_delegating_executor`
    * `value_serialization`

To fix issue #1:

* Promoted the `reference` property from the subclasses of `executor_value_base.ExecutorValue` to `executor_value_base.ExecutorValue`.
* Renamed usage of this API to be consistent.

To fix issue #2:

* Updated the methods on the subclasses of `federating_executor.FederatingStrategy` (`FederatedComposingStrategy` and `FederatedResolvingStrategy`) to accept and return `executor_value_base.ExecutorValue` types.

Note: That this does make the return types more generic, but I believe this is the intended interface. An alternative fix could be to update the `executor_utils` function to take and return `TypeVar`'s (i.e. generic functions). However, I don't think this is the intended interface.

To fix issue #3:

* Updated the type annotation of the `target_executors` parameter to `dict[placements.PlacementLiteral, Union[list[executor_base.Executor], executor_base.Executor]]`.

PiperOrigin-RevId: 504114859

PiperOrigin-RevId: 504091659

PiperOrigin-RevId: 504081415

PiperOrigin-RevId: 504069777

The C++ XLA executor does yet handle `federated_sum` and `federated_mean` intrinsics or unplaced literal values, so this also refactors tests to use work arounds:

- Implement all aggregations in terms of `federated_aggregate` with XLA computations.
- Wrap all literals in no-arg XLA computations that are immediately called.

PiperOrigin-RevId: 504046551

PiperOrigin-RevId: 504015530

These functions are supposed to be high level convenience functions. If you need to change the executor or compiler of this contexts, you are probably a "power user" and should manually construct an execution context for your purpose.

PiperOrigin-RevId: 504015371

Extend JAX/XLA serialization to preserve the Python containers as additional annotations on structure types.

PiperOrigin-RevId: 504012748

Even though the baseclass `Struct` is registered, this does not appear to be
inherited by subclasses. This is required to flatten structured arguments during jax serialization,
otherwise flatten stops when it encounters a `_XlaSerializerStructArg` typed value, when it should flatten to lists of `_XlaSerializerTensorArg`s.

PiperOrigin-RevId: 504003136

Align the context types allowed by the `tff.program.NativeFederatedContext` and the `tff.program.PrefetchingDataSource`.

PiperOrigin-RevId: 504002913

Adds a ReferenceResolvingExecutor after the SequenceExecutor to resolve lambdas
inside sequence intrinsics. This uncovers a bug in the FederatingExecutor where
non-federated intrinsics raise an error and fail computation when they should
be forwarded to the unplaced child executor.

PiperOrigin-RevId: 503993984

PiperOrigin-RevId: 503545574

This map will be specified by user while declaring tf_computation and will be utilized only for DTensor executor is used. The layout map attributes will be ignored for other types of executors.

PiperOrigin-RevId: 503224849

Add a bool parameter `stream_structs` to set_borg_* APIs. Remote executors stream the values in a structure across the gRPC interface when `stream_structs` is set to true.

PiperOrigin-RevId: 503184338

This works enables using TFF's C++ runtime with TPUs, which uses the MLIR bridge and XLA to place the loop body on TPU cores.

PiperOrigin-RevId: 503023318

The two values here have different dimensions and shapes, and the implicit operation numpy was previously performing while computing these norms seems to be failing with the numpy upgrade.

PiperOrigin-RevId: 503018459

PiperOrigin-RevId: 503013253

Fix a bug in XlaExecutor to always expect results as tuples (possibly of length 1), to match MLIR based JAX tracing.

PiperOrigin-RevId: 502965936

PiperOrigin-RevId: 502932624

PiperOrigin-RevId: 502894191

PiperOrigin-RevId: 502723099

PiperOrigin-RevId: 502713076