common

Common components shared between server and client.

Functions

array_from_numpy(ndarray)	Create Array from NumPy ndarray.
bytes_to_ndarray(tensor)	Deserialize NumPy ndarray from bytes.
configure(identifier[, filename, host])	Configure logging to file and/or remote log server.
event(event_type[, event_details])	Submit create_event to ThreadPoolExecutor to avoid blocking.
log(level, msg, *args, **kwargs)	Log 'msg % args' with the integer severity 'level'.
ndarray_to_bytes(ndarray)	Serialize NumPy ndarray to bytes.
now()	Construct a datetime from time.time() with time zone set to UTC.
ndarrays_to_parameters(ndarrays)	Convert NumPy ndarrays to parameters object.
parameters_to_ndarrays(parameters)	Convert parameters object to NumPy ndarrays.

Classes

Array(dtype, shape, stype, data)	Array type.
ClientMessage([get_properties_res, ...])	ClientMessage is a container used to hold one result message.
Code(value)	Client status codes.
ConfigsRecord([configs_dict, keep_input])	Configs record.
Context(state)	State of your run.
DisconnectRes(reason)	DisconnectRes message from client to server.
EvaluateIns(parameters, config)	Evaluate instructions for a client.
EvaluateRes(status, loss, num_examples, metrics)	Evaluate response from a client.
EventType(value)	Types of telemetry events.
FitIns(parameters, config)	Fit instructions for a client.
FitRes(status, parameters, num_examples, metrics)	Fit response from a client.
Error(code[, reason])	A dataclass that stores information about an error that occurred.
GetParametersIns(config)	Parameters request for a client.
GetParametersRes(status, parameters)	Response when asked to return parameters.
GetPropertiesIns(config)	Properties request for a client.
GetPropertiesRes(status, properties)	Properties response from a client.
Message(metadata[, content, error])	State of your application from the viewpoint of the entity using it.
MessageType()	Message type.
MessageTypeLegacy()	Legacy message type.
Metadata(run_id, message_id, src_node_id, ...)	A dataclass holding metadata associated with the current message.
MetricsRecord([metrics_dict, keep_input])	Metrics record.
Parameters(tensors, tensor_type)	Model parameters.
ParametersRecord([array_dict, keep_input])	Parameters record.
ReconnectIns(seconds)	ReconnectIns message from server to client.
RecordSet([parameters_records, ...])	RecordSet stores groups of parameters, metrics and configs.
ServerMessage([get_properties_ins, ...])	ServerMessage is a container used to hold one instruction message.
Status(code, message)	Client status.