Merge branch 'add/exclusive-spancat' of github.com:ljvmiranda921/spaCy into exclusive-spancat

This commit is contained in:
kadarakos 2023-03-07 09:24:02 +00:00
commit 91e0c3d782

View File

@ -306,8 +306,6 @@ class SpanCategorizer(TrainablePipe):
vocab: Vocab, vocab: Vocab,
model: Model[Tuple[List[Doc], Ragged], Floats2d], model: Model[Tuple[List[Doc], Ragged], Floats2d],
suggester: Suggester, suggester: Suggester,
# TODO Not sure what's the best default name when it can both be spancat
# and spancat_singlelabel
name: str = "spancat", name: str = "spancat",
*, *,
add_negative_label: bool = False, add_negative_label: bool = False,
@ -324,6 +322,7 @@ class SpanCategorizer(TrainablePipe):
can produce multiple labels per span (multi-label). In the can produce multiple labels per span (multi-label). In the
multi-label case the classification layer is expected to be multi-label case the classification layer is expected to be
Logistic and Softmax in the multi-class case. Logistic and Softmax in the multi-class case.
vocab (Vocab): The shared vocabulary. vocab (Vocab): The shared vocabulary.
model (thinc.api.Model): The Thinc Model powering the pipeline component. model (thinc.api.Model): The Thinc Model powering the pipeline component.
suggester (Callable[[Iterable[Doc], Optional[Ops]], Ragged]): A function that suggests spans. suggester (Callable[[Iterable[Doc], Optional[Ops]], Ragged]): A function that suggests spans.