EvaluationLoopTrainingCallback

class EvaluationLoopTrainingCallback(factory, frequency=1, prefix=None, evaluator=None, evaluator_kwargs=None, additional_filter_triples=None, **kwargs)

Bases: TrainingCallback

A callback for regular evaluation using new-style evaluation loops.

Initialize the callback.

Parameters:

• factory (CoreTriplesFactory) – the triples factory comprising the evaluation triples

• frequency (int) – the evaluation frequency

• prefix (Optional[str]) – a prefix to use for logging (e.g., to distinguish between different splits)

• evaluator (Union[str, Evaluator, Type[Evaluator], None]) – the evaluator, or a hint thereof

• evaluator_kwargs (Optional[Mapping[str, Any]]) – additional keyword-based parameters used for the evaluation instantiation

• additional_filter_triples (Union[LongTensor, CoreTriplesFactory, Sequence[Union[LongTensor, CoreTriplesFactory]], None]) – additional filter triples to use for creating the filter

• kwargs – additional keyword-based parameters passed to EvaluationLoop.evaluate()

Attributes Summary

evaluation_loop

Return the evaluation loop instance (lazy-initialization).

Methods Summary

post_epoch(epoch, epoch_loss, **kwargs)

Call after epoch.

Attributes Documentation

evaluation_loop

Return the evaluation loop instance (lazy-initialization).

Methods Documentation

post_epoch(epoch, epoch_loss, **kwargs)

Call after epoch.

Return type:

None

Parameters:

• epoch (int) –

• epoch_loss (float) –

• kwargs (Any) –