explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-04-30 13:53:45 +03:00
Code Issues Packages Projects Releases Wiki Activity

[WIP] Update

Browse Source
This commit is contained in:
Lj Miranda 2022-08-25 13:08:37 +08:00
parent 3d07c05cba
commit 5452e71b05
1 changed files with 7 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
spacy/pipeline/spancat_exclusive.py
View File

@ -272,7 +272,7 @@ class SpanCategorizerExclusive(TrainablePipe):
"""Modify a batch of Doc objects, using pre-computed scores. """Modify a batch of Doc objects, using pre-computed scores.
docs (Iterable[Doc]): The documents to modify. docs (Iterable[Doc]): The documents to modify.
scores: The scores to set, produced by SpanCategorizer.predict. scores: The scores to set, produced by SpanCategorizerExclusive.predict.
DOCS: https://spacy.io/api/spancategorizerexclusive#set_annotations DOCS: https://spacy.io/api/spancategorizerexclusive#set_annotations
""" """
@ -446,10 +446,12 @@ class SpanCategorizerExclusive(TrainablePipe):
scores = self.model.ops.to_numpy(scores) scores = self.model.ops.to_numpy(scores)
indices = self.model.ops.to_numpy(indices) indices = self.model.ops.to_numpy(indices)
predicted = scores.argmax(axis=1) predicted = scores.argmax(axis=1)
# Remove samples where the negative label is the argmax # Remove samples where the negative label is the argmax
positive = numpy.where(predicted != self._negative_label) positive = numpy.where(predicted != self._negative_label)
predicted = predicted[positive[0]] predicted = predicted[positive[0]]
indices = indices[positive[0]] indices = indices[positive[0]]
# Sort spans according to argmax probability # Sort spans according to argmax probability
if not allow_overlap: if not allow_overlap:
argmax_probs = numpy.take_along_axis( argmax_probs = numpy.take_along_axis(
@ -459,16 +461,20 @@ class SpanCategorizerExclusive(TrainablePipe):
sort_idx = (argmax_probs * -1).argsort() sort_idx = (argmax_probs * -1).argsort()
predicted = predicted[sort_idx] predicted = predicted[sort_idx]
indices = indices[sort_idx] indices = indices[sort_idx]
seen = Ranges() seen = Ranges()
spans = SpanGroup(doc, name=self.key) spans = SpanGroup(doc, name=self.key)
for i in range(len(predicted)): for i in range(len(predicted)):
label = predicted[i] label = predicted[i]
start = indices[i, 0] start = indices[i, 0]
end = indices[i, 1] end = indices[i, 1]
if not allow_overlap: if not allow_overlap:
if (start, end) in seen: if (start, end) in seen:
continue continue
else: else:
seen.add(start, end) seen.add(start, end)
spans.append(Span(doc, start, end, label=labels[label])) spans.append(Span(doc, start, end, label=labels[label]))
return spans return spans