explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-07-10 08:12:24 +03:00
Code Issues Packages Projects Releases Wiki Activity

[ci skip] Add breakpoint for debugging

Browse Source
This commit is contained in:
Lj Miranda 2022-08-25 13:23:15 +08:00
parent 5452e71b05
commit d6e56b62b9
1 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
spacy/pipeline/spancat_exclusive.py
View File

@ -79,9 +79,9 @@ def make_spancat(
spans_key: str,
scorer: Optional[Callable],
negative_weight: float = 1.0,
allow_overlap: Optional[bool] = True,
allow_overlap: bool = True,
) -> "SpanCategorizerExclusive":
"""Create a SpanCategorizer component. The span categorizer consists of two
"""Create a SpanCategorizerExclusive component. The span categorizer consists of two
parts: a suggester function that proposes candidate spans, and a labeller
model that predicts one or more labels for each span.
@ -97,7 +97,7 @@ def make_spancat(
reference document under the same key.
negative_weight (float): Multiplier for the loss terms.
Can be used to down weigh the negative samples if there are too many.
allow_overlap (Optional[bool]): If True the data is assumed to
allow_overlap (bool): If True the data is assumed to
contain overlapping spans.
"""
return SpanCategorizerExclusive(
@ -148,8 +148,8 @@ class SpanCategorizerExclusive(TrainablePipe):
*,
spans_key: str = "spans",
negative_weight: float = 1.0,
allow_overlap: bool = True,
scorer: Optional[Callable] = spancat_score,
allow_overlap: Optional[bool] = True,
) -> None:
"""Initialize the span categorizer.
vocab (Vocab): The shared vocabulary.
@ -162,11 +162,11 @@ class SpanCategorizerExclusive(TrainablePipe):
`"spans"`.
negative_weight (float): Multiplier for the loss terms.
Can be used to down weigh the negative samples if there are too many.
allow_overlap (bool): If True the data is assumed to
contain overlapping spans.
scorer (Optional[Callable]): The scoring method. Defaults to
Scorer.score_spans for the Doc.spans[spans_key] with overlapping
spans allowed.
allow_overlap (Optional[bool]): If True the data is assumed to
contain overlapping spans.
DOCS: https://spacy.io/api/spancategorizerexclusive#init
"""
@ -376,6 +376,7 @@ class SpanCategorizerExclusive(TrainablePipe):
offset += spans.lengths[i]
target = self.model.ops.asarray(target, dtype="f") # type: ignore
negative_samples = numpy.nonzero(negative_spans)[0]
breakpoint()
target[negative_samples, self._negative_label] = 1.0
d_scores = scores - target
neg_weight = self.cfg["negative_weight"]