nvflare.edge.executors.ete module

class EdgeTaskExecutor(updater_id: str, update_timeout: float, learner_id='')[source]

Bases: HierarchicalUpdateGatherer

This is the base class for building executors to manage federated learning on edge devices. Subclasses must implement the required abstract methods defined here.

Note: This class is based on HUG (HierarchicalUpdateGatherer). All interactions with parent clients are already handled by HUG.

Constructor of EdgeTaskExecutor.

Parameters:

  • updater_id – ID of the updater

  • update_timeout – timeout for update messages sent to parent

  • learner_id – ID of the learner component. Optional.

abstract process_edge_result_report(request: ResultReport, current_task: TaskInfo, fl_ctx: FLContext) → ResultResponse | None[source]

This is called to process an edge result report sent from the edge device.

Parameters:

  • request – the request from edge device

  • current_task – the current pending task

  • fl_ctx – FLContext object

Returns: reply to the edge device

abstract process_edge_selection_request(request: SelectionRequest, current_task: TaskInfo, fl_ctx: FLContext) → SelectionResponse | None[source]

This is called to process an edge selection request sent from the edge device.

Parameters:

  • request – the request from edge device

  • current_task – the current pending task

  • fl_ctx – FLContext object

Returns: reply to the edge device

abstract process_edge_task_request(request: TaskRequest, current_task: TaskInfo, fl_ctx: FLContext) → TaskResponse | None[source]

This is called to process an edge task request sent from the edge device.

Parameters:

  • request – the request from edge device

  • current_task – the current pending task

  • fl_ctx – FLContext object

Returns: reply to the edge device