explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-05-29 18:23:06 +03:00
Code Issues Packages Projects Releases Wiki Activity

Use Softmax v2 directly from thinc

Browse Source
This commit is contained in:
Lj Miranda 2022-09-05 11:28:30 +08:00
parent 43bf05275f
commit 2bbab641e9
1 changed files with 4 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
spacy/pipeline/spancat_exclusive.py
View File

@ -2,33 +2,24 @@ from dataclasses import dataclass
from typing import Any, Callable, Dict, Iterable, List, Optional, Tuple, cast
import numpy
from thinc.api import Config, Model, Ops, Optimizer, Softmax_v2
from thinc.api import get_current_ops, set_dropout_rate
from thinc.types import Floats2d, Ints1d, Ints2d, Ragged
from thinc.api import Config, Model, Ops, Optimizer
from thinc.api import set_dropout_rate
from thinc.types import Floats2d, Ints2d, Ragged
from ..compat import Protocol, runtime_checkable
from ..errors import Errors
from ..language import Language
from ..tokens import Doc, Span, SpanGroup
from ..training import Example, validate_examples
from ..util import registry
from ..vocab import Vocab
from .spancat import spancat_score, build_ngram_suggester
from .trainable_pipe import TrainablePipe
@registry.layers("spacy.Softmax.v1")
def build_softmax(nO=None, nI=None) -> Model[Floats2d, Floats2d]:
"""
An output layer for softmax classification.
"""
return Softmax_v2(nI=nI, nO=nO)
spancat_exclusive_default_config = """
[model]
@architectures = "spacy.SpanCategorizer.v1"
scorer = {"@layers": "spacy.Softmax.v1"}
scorer = {"@layers": "Softmax.v2"}
[model.reducer]
@layers = spacy.mean_max_reducer.v1