LoggerManager

class torch_ecg.components.LoggerManager(log_dir: str | bytes | PathLike | None = None, log_suffix: str | None = None)

Bases: ReprMixin

Manager of loggers.

This class manages multiple loggers and provides a unified interface to log metrics and messages.

Parameters:

• log_dir (path-like, optional) – The directory to save the log file.

• log_suffix (str, optional) – The suffix of the log file.

close() → None

Close the loggers.

epoch_end(epoch: int) → None

Actions to be performed at the end of each epoch.

Parameters:

epoch (int) – The number of the epoch.

epoch_start(epoch: int) → None

Message logged at the start of each epoch.

Parameters:

epoch (int) – The number of the epoch.

extra_repr_keys() → List[str]

Extra keys for __repr__() and __str__().

flush() → None

Flush the loggers.

classmethod from_config(config: Dict[str, Any]) → LoggerManager

Create a logger manager from a configuration.

Parameters:

config (dict) – The configuration of the logger manager.

Returns:

The instance of the created logger manager.

Return type:

LoggerManager

property log_dir: str

Directory to save the log file.

log_message(msg: str, level: int = 20) → None

Log a message.

Parameters:

• msg (str) – The message to be logged.

• level ({logging.DEBUG, logging.INFO, logging.WARNING, logging.ERROR, logging.CRITICAL}, optional) – The level of the message.

log_metrics(metrics: Dict[str, Real | Tensor], step: int | None = None, epoch: int | None = None, part: str = 'train') → None

Log metrics.

Parameters:

• metrics (dict) – The metrics to be logged.

• step (int, optional) – The number of (global) steps of training.

• epoch (int, optional,) – The epoch number of training.

• part (str, optional) – The part of the training data the metrics computed from, can be “train” or “val” or “test”, etc.

property log_suffix: str

Suffix of the log file.

property loggers: List[BaseLogger]

List of loggers.