tff.framework.FederatedComposingStrategy

View source on GitHub

A strategy for composing federated types and intrinsics in disjoint scopes.

Inherits From: FederatingStrategy

tff.framework.FederatedComposingStrategy(
    executor: tff.framework.FederatingExecutor,
    server_executor: tff.framework.Executor,
    target_executors: list[executor_base.Executor],
    local_computation_factory: local_computation_factory_base.LocalComputationFactory = tensorflow_computation_factory.TensorFlowComputationFactory()
)

This strategy implements the FederatingStrategy interface and provides logic for handling federated types and federated intrinsics on an underlying collection of target executors. This strategy can be used in hierarchical aggregation structures with federating executors handling disjoint scopes at the leafs.

This strategy only supports the placements:

• tff.SERVER
• tff.CLIENTS

Args
executor	A federating_executor.FederatingExecutor to use to handle unplaced types, computations, and processing.
server_executor	The parent executor to use for all processing at the parent, such as combining values from child executors, unplaced and server-side processing, etc.
target_executors	The list of executors that manage disjoint scopes to combine in this executor, delegate to and collect or aggregate from.
local_computation_factory	An instance of LocalComputationFactory to use to construct local computations used as parameters in certain federated operators (such as tff.federated_sum, etc.). Defaults to a TensorFlow computation factory that generates TensorFlow code.

Raises
TypeError	If server_executor is not an executor_base.Executor or if target_executors is not a list of executor_base.Executors.

Methods

close

View source

close()

Release resources associated with this strategy, if any.

compute_federated_aggregate

View source

compute_federated_aggregate(
    arg
)

Returns an embedded call for a federated aggregate.

compute_federated_apply

View source

compute_federated_apply(
    arg
)

Returns an embedded call for a federated apply.

compute_federated_broadcast

View source

compute_federated_broadcast(
    arg
)

Returns an embedded call for a federated broadcast.

compute_federated_eval_at_clients

View source

compute_federated_eval_at_clients(
    arg
)

Returns an embedded call for a federated eval at clients.

compute_federated_eval_at_server

View source

compute_federated_eval_at_server(
    arg
)

Returns an embedded call for a federated eval at server.

compute_federated_intrinsic

View source

compute_federated_intrinsic(
    uri, arg=None
)

Returns an embedded call for a federated intrinsic.

Args
uri	The URI of an intrinsic to embed.
arg	An optional embedded argument of the call, or None if no argument is supplied.

compute_federated_map

View source

compute_federated_map(
    arg
)

Returns an embedded call for a federated map.

compute_federated_map_all_equal

View source

compute_federated_map_all_equal(
    arg
)

Returns an embedded call for a federated map all equal.

compute_federated_mean

View source

compute_federated_mean(
    arg
)

Returns an embedded call for a federated mean.

compute_federated_secure_select

View source

compute_federated_secure_select(
    arg
)

Returns an embedded call for a federated secure select.

compute_federated_secure_sum_bitwidth

View source

compute_federated_secure_sum_bitwidth(
    arg
)

Returns an embedded call for a federated secure sum.

compute_federated_select

View source

compute_federated_select(
    arg
)

Returns an embedded call for a federated select.

compute_federated_sum

View source

compute_federated_sum(
    arg
)

Returns an embedded call for a federated sum.

compute_federated_value

View source

compute_federated_value(
    value, type_signature
)

Returns an embedded value for a federated type.

Args
value	An object to embed in the executor.
type_signature	A tff.Type, the type of value.

compute_federated_value_at_clients

View source

compute_federated_value_at_clients(
    arg
)

Returns an embedded call for a federated value at clients.

compute_federated_value_at_server

View source

compute_federated_value_at_server(
    arg
)

Returns an embedded call for a federated value at server.

compute_federated_weighted_mean

View source

compute_federated_weighted_mean(
    arg
)

Returns an embedded call for a federated weighted mean.

compute_federated_zip_at_clients

View source

compute_federated_zip_at_clients(
    arg
)

Returns an embedded call for a federated zip at clients.

compute_federated_zip_at_server

View source

compute_federated_zip_at_server(
    arg
)

Returns an embedded call for a federated zip at server.

factory

View source

@classmethod
factory(
    server_executor: tff.framework.Executor,
    target_executors: list[executor_base.Executor],
    local_computation_factory: local_computation_factory_base.LocalComputationFactory = tensorflow_computation_factory.TensorFlowComputationFactory()
)

ingest_value

View source

ingest_value(
    value: Any,
    type_signature: tff.types.Type
) -> tff.framework.ExecutorValue