blazefl.core.BaseClientTrainer

class blazefl.core.BaseClientTrainer(*args, **kwargs)

Bases: Protocol[UplinkPackage, DownlinkPackage]

Abstract base class for serial client training in federated learning.

This class defines the interface for training clients in a serial manner, where each client is processed one after the other.

Raises:

NotImplementedError – If the methods are not implemented in a subclass.

__init__(*args, **kwargs)

Methods

__init__(*args, **kwargs)
local_process(payload, cid_list)	Process the downlink payload from the server for a list of client IDs.
uplink_package()	Prepare the data package to be sent from the client to the server.

local_process(payload: DownlinkPackage, cid_list: list[int]) → None

Process the downlink payload from the server for a list of client IDs.

Parameters:

• payload (DownlinkPackage) – The data package received from the server.

• cid_list (list[int]) – A list of client IDs to process.

Returns:

None

uplink_package() → list[UplinkPackage]

Prepare the data package to be sent from the client to the server.

Returns:

A list of data packages prepared for uplink transmission.

Return type:

list[UplinkPackage]