quaterion.eval.attached_metric module

class AttachedMetric(name: str, metric: BaseMetric, logger: bool = True, prog_bar: bool = False, on_step: Optional[bool] = None, on_epoch: Optional[bool] = None, **log_options)

Bases: object

Attach batch-wise metric to TrainableModel

Contain required parameters to compute and log batch-wise metric during training process.

Parameters:

• name – name of an attached metric to be used in log.

• metric – metric to be calculated.

• on_step – Logs the metric at the current step.

• on_epoch – Automatically accumulates and logs at the end of the epoch.

• prog_bar – Logs to the progress bar (Default: False).

• logger – Logs to the logger like Tensorboard, or any other custom logger passed to the Trainer (Default: True).

• **log_options – additional kwargs to be passed to model’s log.

• at (The remaining options can be found) –

• https – //pytorch-lightning.readthedocs.io/en/stable/extensions/logging.html