p2pfl.learning.aggregators.aggregator module

Abstract aggregator.

class p2pfl.learning.aggregators.aggregator.Aggregator(disable_partial_aggregation=False)[source]

Bases: NodeComponent

Class to manage the aggregation of models.

Parameters:

node_addr – Address of the node.

SUPPORTS_PARTIAL_AGGREGATION: bool = False
add_model(*args: Any, **kwargs: Any) → Any
Return type:

Any

aggregate(*args: Any, **kwargs: Any) → Any
Return type:

Any

clear(*args: Any, **kwargs: Any) → Any
Return type:

Any

get_aggregated_models(*args: Any, **kwargs: Any) → Any
Return type:

Any

get_missing_models(*args: Any, **kwargs: Any) → Any
Return type:

Any

get_model(*args: Any, **kwargs: Any) → Any
Return type:

Any

get_required_callbacks(*args: Any, **kwargs: Any) → Any
Return type:

Any

set_nodes_to_aggregate(*args: Any, **kwargs: Any) → Any
Return type:

Any

wait_and_get_aggregation(*args: Any, **kwargs: Any) → Any
Return type:

Any

exception p2pfl.learning.aggregators.aggregator.NoModelsToAggregateError[source]

Bases: Exception

Exception raised when there are no models to aggregate.