:mod:`fedlab_core.server.handler`
=================================

.. py:module:: fedlab_core.server.handler


Module Contents
---------------

Classes
~~~~~~~

.. autoapisummary::

   fedlab_core.server.handler.ParameterServerHandler
   fedlab_core.server.handler.SyncParameterServerHandler
   fedlab_core.server.handler.AsyncParameterServerHandler



.. class:: ParameterServerHandler(model, cuda=False)


   Bases: :class:`object`

   An abstract class representing handler for parameter server.

   Please make sure that you self-defined server handler class subclasses this class

   .. rubric:: Example

   read sourcecode of :class:`SyncSGDParameterServerHandler` below

   .. method:: on_receive(self)
      :abstractmethod:

      Override this function to define what the server to do when receiving message from client


   .. method:: update(self, model_list)
      :abstractmethod:

      Override this function to update global model

      :param model_list: a list of model parameters serialized by :func:`ravel_model_params`
      :type model_list: list


   .. method:: buffer(self)
      :property:


   .. method:: model(self)
      :property:



.. class:: SyncParameterServerHandler(model, client_num_in_total, cuda=False, select_ratio=1.0, logger_path='server_handler.txt', logger_name='server handler')


   Bases: :class:`fedlab_core.server.handler.ParameterServerHandler`

   Synchronous Parameter Server Handler

   Backend of synchronous parameter server: this class is responsible for backend computing.

   Synchronous parameter server will wait for every client to finish local training process before the
   next FL round.

   :param model: Model used in this federation
   :type model: torch.nn.Module
   :param client_num_in_total: Total number of clients in this federation
   :type client_num_in_total: int
   :param cuda: Use GPUs or not
   :type cuda: bool
   :param select_ratio: ``select_ratio * client_num`` is the number of clients to join every FL round
   :type select_ratio: float

   .. method:: on_receive(self, sender, message_code, payload) -> None

      Define what parameter server does when receiving a single client's message

      :param sender: Index of client in distributed
      :type sender: int
      :param message_code: Agreements code defined in :class:`MessageCode` class
      :type message_code: MessageCode
      :param payload: Serialized model parameters
      :type payload: torch.Tensor


   .. method:: select_clients(self)

      Return a list of client rank indices selected randomly


   .. method:: is_updated(self) -> bool


   .. method:: start_round(self)



.. class:: AsyncParameterServerHandler(model, cuda)


   Bases: :class:`fedlab_core.server.handler.ParameterServerHandler`

   Asynchronous ParameterServer Handler

   Update global model immediately after receiving a ParameterUpdate message
   paper: https://arxiv.org/abs/1903.03934

   :param model: Global model in server
   :type model: torch.nn.Module
   :param cuda: Use GPUs or not
   :type cuda: bool

   .. method:: update(self, model_list)

      Override this function to update global model

      :param model_list: a list of model parameters serialized by :func:`ravel_model_params`
      :type model_list: list


   .. method:: on_receive(self, sender, message_code, parameter)

      Override this function to define what the server to do when receiving message from client