History

class History

기반 클래스: object

History class for training and/or evaluation metrics collection.

메소드

add_loss_centralized(server_round, loss)	Add one loss entry (from centralized evaluation).
add_loss_distributed(server_round, loss)	Add one loss entry (from distributed evaluation).
add_metrics_centralized(server_round, metrics)	Add metrics entries (from centralized evaluation).
add_metrics_distributed(server_round, metrics)	Add metrics entries (from distributed evaluation).
add_metrics_distributed_fit(server_round, ...)	Add metrics entries (from distributed fit).

add_loss_centralized(server_round: int, loss: float) → None

Add one loss entry (from centralized evaluation).

add_loss_distributed(server_round: int, loss: float) → None

Add one loss entry (from distributed evaluation).

add_metrics_centralized(server_round: int, metrics: dict[str, bool | bytes | float | int | str]) → None

Add metrics entries (from centralized evaluation).

add_metrics_distributed(server_round: int, metrics: dict[str, bool | bytes | float | int | str]) → None

Add metrics entries (from distributed evaluation).

add_metrics_distributed_fit(server_round: int, metrics: dict[str, bool | bytes | float | int | str]) → None

Add metrics entries (from distributed fit).