tff.backends.native.set_local_execution_context

Sets an execution context that executes computations locally.

tff.backends.native.set_local_execution_context(
    default_num_clients: int = 0,
    max_fanout=100,
    clients_per_thread=1,
    server_tf_device=None,
    client_tf_devices=tuple(),
    reference_resolving_clients=False
)