p2pfl.learning.frameworks.pytorch.lightning_logger module

Lightning Logger for P2PFL.

class p2pfl.learning.frameworks.pytorch.lightning_logger.FederatedLogger(addr)

Bases: Logger

Pytorch Lightning Logger for Federated Learning. Handles local training loggin.

Parameters:

• node_name – Name of the node.

• addr (str)

finalize(status)

Finalize the logger.

Parameters:

status (str)

Return type:

None

log_hyperparams(*args, **kwargs)

Log hyperparameters.

Parameters:

• args (Any)

• kwargs (Any)

Return type:

None

log_metrics(metrics, step=None)

Log metrics (in a pytorch format).

Parameters:

• metrics (dict[str, float])

• step (int | None)

Return type:

None

property name: None

Name of the logger.

save()

Save the logger.

Return type:

None

property version: None

Version of the logger.