|View source on GitHub|
Constructs an executor factory to execute computations locally.
tff.framework.local_executor_factory( num_clients=None, max_fanout=100, clients_per_thread=1, tf_devices=tuple() )
num_clients: The number of clients. If specified, the executor factory function returned by
local_executor_factorywill be configured to have exactly
num_clientsclients. If unspecified (
None), then the function returned will attempt to infer cardinalities of all placements for which it is passed values.
max_fanout: The maximum fanout at any point in the aggregation hierarchy. If
num_clients > max_fanout, the constructed executor stack will consist of multiple levels of aggregators. The height of the stack will be on the order of
log(num_clients) / log(max_fanout).
clients_per_thread: The number of clients which will share a single thread. Adjusting this parameter away from 1 can be useful if clients work is light.
tf_devices: The list of devices to run clients for simulation. The
tf.config.LogicalDevicelist can often be returned by
An instance of
executor_factory.ExecutorFactory encapsulating the
executor construction logic specified above.
ValueError: If the number of clients is specified and not one or larger.