nvflare.app_common.widgets.streaming module

class AnalyticsReceiver(events: Optional[List[str]] = None)

Bases: Widget, ABC

Receives analytic data.

Parameters

events (optional, List[str]) – A list of event that this receiver will handle.

abstract finalize(fl_ctx: FLContext)

Finalizes the receiver.

Parameters

fl_ctx (FLContext) – fl context.

handle_event(event_type: str, fl_ctx: FLContext)

Handles events.

Parameters

• event_type (str) – event type fired by workflow.

• fl_ctx (FLContext) – FLContext information.

abstract initialize(fl_ctx: FLContext)

Initializes the receiver.

Parameters

fl_ctx (FLContext) – fl context.

abstract save(fl_ctx: FLContext, shareable: Shareable, record_origin: str)

Saves the received data.

Parameters

• fl_ctx (FLContext) – fl context.

• shareable (Shareable) – the received message.

• record_origin (str) – the sender of this message / record.

class AnalyticsSender(event_type='analytix_log_stats')

Bases: Widget

Sends analytics data.

Note::

This class implements some common methods follows signatures from PyTorch SummaryWriter. It provides a convenient way for Learner to use.

Parameters

event_type (str) – event type to fire.

add_image(tag: str, image, global_step: Optional[int] = None, **kwargs)

Sends an image.

Parameters

• tag (str) – Data identifier.

• image – Image to send.

• global_step (optional, int) – Global step value.

• **kwargs – Additional arguments to pass to the receiver side.

add_scalar(tag: str, scalar: float, global_step: Optional[int] = None, **kwargs)

Sends a scalar.

Parameters

• tag (str) – Data identifier.

• scalar (float) – Value to send.

• global_step (optional, int) – Global step value.

• **kwargs – Additional arguments to pass to the receiver side.

add_scalars(tag: str, scalars: dict, global_step: Optional[int] = None, **kwargs)

Sends scalars.

Parameters

• tag (str) – The parent name for the tags.

• scalars (dict) – Key-value pair storing the tag and corresponding values.

• global_step (optional, int) – Global step value.

• **kwargs – Additional arguments to pass to the receiver side.

add_text(tag: str, text: str, global_step: Optional[int] = None, **kwargs)

Sends a text.

Parameters

• tag (str) – Data identifier.

• text (str) – String to send.

• global_step (optional, int) – Global step value.

• **kwargs – Additional arguments to pass to the receiver side.

close()

Close resources.

flush()

Flushes out the message.

This is doing nothing, it is defined for mimic the PyTorch SummaryWriter behavior.

handle_event(event_type: str, fl_ctx: FLContext)

Handles events.

Parameters

• event_type (str) – event type fired by workflow.

• fl_ctx (FLContext) – FLContext information.

create_analytic_dxo(tag: str, value, data_type: AnalyticsDataType, **kwargs) → DXO

Creates the analytic DXO.

Parameters

• tag (str) – the tag associated with this value.

• value – the analytic data.

• data_type (AnalyticsDataType) – analytic data type.

• kwargs – additional arguments to be passed into the receiver side’s function.

Returns

A DXO object that contains the analytic data.

send_analytic_dxo(comp: FLComponent, dxo: DXO, fl_ctx: FLContext, event_type: str = 'analytix_log_stats')

Sends analytic dxo.

Parameters

• comp (FLComponent) – An FLComponent.

• dxo (DXO) – analytic data in dxo.

• fl_ctx (FLContext) – fl context info.

• event_type (str) – Event type.