explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2024-12-24 17:06:29 +03:00
Code Issues Packages Projects Releases Wiki Activity

Tidy up and auto-format

Browse Source
This commit is contained in:
Ines Montani 2020-10-05 21:55:18 +02:00
parent 181039bd17
commit 9614e53b02
1 changed files with 6 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
spacy/ml/models/tok2vec.py
View File

@ -1,4 +1,4 @@
from typing import Optional, List, Union, Dict from typing import Optional, List, Union
from thinc.types import Floats2d from thinc.types import Floats2d
from thinc.api import chain, clone, concatenate, with_array, with_padded from thinc.api import chain, clone, concatenate, with_array, with_padded
from thinc.api import Model, noop, list2ragged, ragged2list, HashEmbed from thinc.api import Model, noop, list2ragged, ragged2list, HashEmbed
@ -11,7 +11,7 @@ from ...ml import _character_embed
from ..staticvectors import StaticVectors from ..staticvectors import StaticVectors
from ..featureextractor import FeatureExtractor from ..featureextractor import FeatureExtractor
from ...pipeline.tok2vec import Tok2VecListener from ...pipeline.tok2vec import Tok2VecListener
from ...attrs import ORTH, NORM, LOWER, PREFIX, SUFFIX, SHAPE, intify_attr from ...attrs import intify_attr
@registry.architectures.register("spacy.Tok2VecListener.v1") @registry.architectures.register("spacy.Tok2VecListener.v1")
@ -29,7 +29,7 @@ def build_hash_embed_cnn_tok2vec(
window_size: int, window_size: int,
maxout_pieces: int, maxout_pieces: int,
subword_features: bool, subword_features: bool,
pretrained_vectors: Optional[bool] pretrained_vectors: Optional[bool],
) -> Model[List[Doc], List[Floats2d]]: ) -> Model[List[Doc], List[Floats2d]]:
"""Build spaCy's 'standard' tok2vec layer, which uses hash embedding """Build spaCy's 'standard' tok2vec layer, which uses hash embedding
with subword features and a CNN with layer-normalized maxout. with subword features and a CNN with layer-normalized maxout.
@ -56,7 +56,7 @@ def build_hash_embed_cnn_tok2vec(
""" """
if subword_features: if subword_features:
attrs = ["NORM", "PREFIX", "SUFFIX", "SHAPE"] attrs = ["NORM", "PREFIX", "SUFFIX", "SHAPE"]
row_sizes = [embed_size, embed_size//2, embed_size//2, embed_size//2] row_sizes = [embed_size, embed_size // 2, embed_size // 2, embed_size // 2]
else: else:
attrs = ["NORM"] attrs = ["NORM"]
row_sizes = [embed_size] row_sizes = [embed_size]
@ -120,7 +120,7 @@ def MultiHashEmbed(
layer is used to map the vectors to the specified width before concatenating layer is used to map the vectors to the specified width before concatenating
it with the other embedding outputs. A single Maxout layer is then used to it with the other embedding outputs. A single Maxout layer is then used to
reduce the concatenated vectors to the final width. reduce the concatenated vectors to the final width.
The `rows` parameter controls the number of rows used by the `HashEmbed` The `rows` parameter controls the number of rows used by the `HashEmbed`
tables. The HashEmbed layer needs surprisingly few rows, due to its use of tables. The HashEmbed layer needs surprisingly few rows, due to its use of
the hashing trick. Generally between 2000 and 10000 rows is sufficient, the hashing trick. Generally between 2000 and 10000 rows is sufficient,
@ -143,13 +143,7 @@ def MultiHashEmbed(
def make_hash_embed(index): def make_hash_embed(index):
nonlocal seed nonlocal seed
seed += 1 seed += 1
return HashEmbed( return HashEmbed(width, rows[index], column=index, seed=seed, dropout=0.0)
width,
rows[index],
column=index,
seed=seed,
dropout=0.0,
)
embeddings = [make_hash_embed(i) for i in range(len(attrs))] embeddings = [make_hash_embed(i) for i in range(len(attrs))]
concat_size = width * (len(embeddings) + include_static_vectors) concat_size = width * (len(embeddings) + include_static_vectors)