spaCy/spacy/ml/models/multi_task.py

242 lines
8.2 KiB
Python
Raw Normal View History

from typing import Optional, Iterable, Tuple, List, Callable, TYPE_CHECKING
2020-06-03 15:45:00 +03:00
from thinc.api import chain, Maxout, LayerNorm, Softmax, Linear, zero_init, Model
from thinc.api import MultiSoftmax, list2array
from thinc.api import to_categorical, CosineDistance, L2Distance
from ...util import registry
from ...errors import Errors
from ...attrs import ID
import numpy
from functools import partial
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
2020-07-31 18:02:54 +03:00
if TYPE_CHECKING:
# This lets us add type hints for mypy etc. without causing circular imports
from ...vocab import Vocab # noqa: F401
from ...tokens import Doc # noqa: F401
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
2021-03-02 19:56:28 +03:00
@registry.architectures("spacy.PretrainVectors.v1")
def create_pretrain_vectors(
maxout_pieces: int, hidden_size: int, loss: str
) -> Callable[["Vocab", Model], Model]:
def create_vectors_objective(vocab: "Vocab", tok2vec: Model) -> Model:
if vocab.vectors.data.shape[1] == 0:
raise ValueError(Errors.E875)
model = build_cloze_multi_task_model(
vocab, tok2vec, hidden_size=hidden_size, maxout_pieces=maxout_pieces
)
model.attrs["loss"] = create_vectors_loss()
return model
def create_vectors_loss() -> Callable:
if loss == "cosine":
distance = CosineDistance(normalize=True, ignore_zeros=True)
return partial(get_vectors_loss, distance=distance)
elif loss == "L2":
distance = L2Distance(normalize=True)
return partial(get_vectors_loss, distance=distance)
else:
raise ValueError(Errors.E906.format(found=loss, supported="'cosine', 'L2'"))
return create_vectors_objective
2021-03-02 19:56:28 +03:00
@registry.architectures("spacy.PretrainCharacters.v1")
def create_pretrain_characters(
maxout_pieces: int, hidden_size: int, n_characters: int
) -> Callable[["Vocab", Model], Model]:
def create_characters_objective(vocab: "Vocab", tok2vec: Model) -> Model:
model = build_cloze_characters_multi_task_model(
vocab,
tok2vec,
hidden_size=hidden_size,
maxout_pieces=maxout_pieces,
nr_char=n_characters,
)
model.attrs["loss"] = partial(get_characters_loss, nr_char=n_characters)
return model
return create_characters_objective
def get_vectors_loss(ops, docs, prediction, distance):
"""Compute a loss based on a distance between the documents' vectors and
the prediction.
"""
# The simplest way to implement this would be to vstack the
# token.vector values, but that's a bit inefficient, especially on GPU.
# Instead we fetch the index into the vectors table for each of our tokens,
# and look them up all at once. This prevents data copying.
ids = ops.flatten([doc.to_array(ID).ravel() for doc in docs])
target = docs[0].vocab.vectors.data[ids]
d_target, loss = distance(prediction, target)
return loss, d_target
def get_characters_loss(ops, docs, prediction, nr_char):
"""Compute a loss based on a number of characters predicted from the docs."""
target_ids = numpy.vstack([doc.to_utf8_array(nr_char=nr_char) for doc in docs])
target_ids = target_ids.reshape((-1,))
target = ops.asarray(to_categorical(target_ids, n_classes=256), dtype="f")
target = target.reshape((-1, 256 * nr_char))
diff = prediction - target
loss = (diff ** 2).sum()
d_target = diff / float(prediction.shape[0])
return loss, d_target
2020-07-31 18:02:54 +03:00
def build_multi_task_model(
tok2vec: Model,
maxout_pieces: int,
token_vector_width: int,
nO: Optional[int] = None,
) -> Model:
softmax = Softmax(nO=nO, nI=token_vector_width * 2)
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
model = chain(
tok2vec,
2020-06-20 15:15:04 +03:00
Maxout(
nO=token_vector_width * 2,
nI=token_vector_width,
nP=maxout_pieces,
dropout=0.0,
),
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
LayerNorm(token_vector_width * 2),
softmax,
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
)
model.set_ref("tok2vec", tok2vec)
model.set_ref("output_layer", softmax)
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
return model
2020-07-31 18:02:54 +03:00
def build_cloze_multi_task_model(
vocab: "Vocab", tok2vec: Model, maxout_pieces: int, hidden_size: int
2020-07-31 18:02:54 +03:00
) -> Model:
nO = vocab.vectors.data.shape[1]
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
output_layer = chain(
list2array(),
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
Maxout(
nO=hidden_size,
2020-06-20 15:15:04 +03:00
nI=tok2vec.get_dim("nO"),
nP=maxout_pieces,
normalize=True,
dropout=0.0,
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
),
Linear(nO=nO, nI=hidden_size, init_W=zero_init),
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
)
model = chain(tok2vec, output_layer)
model = build_masked_language_model(vocab, model)
model.set_ref("tok2vec", tok2vec)
model.set_ref("output_layer", output_layer)
Default settings to configurations (#4995) * fix grad_clip naming * cleaning up pretrained_vectors out of cfg * further refactoring Model init's * move Model building out of pipes * further refactor to require a model config when creating a pipe * small fixes * making cfg in nn_parser more consistent * fixing nr_class for parser * fixing nn_parser's nO * fix printing of loss * architectures in own file per type, consistent naming * convenience methods default_tagger_config and default_tok2vec_config * let create_pipe access default config if available for that component * default_parser_config * move defaults to separate folder * allow reading nlp from package or dir with argument 'name' * architecture spacy.VocabVectors.v1 to read static vectors from file * cleanup * default configs for nel, textcat, morphologizer, tensorizer * fix imports * fixing unit tests * fixes and clean up * fixing defaults, nO, fix unit tests * restore parser IO * fix IO * 'fix' serialization test * add *.cfg to manifest * fix example configs with additional arguments * replace Morpohologizer with Tagger * add IO bit when testing overfitting of tagger (currently failing) * fix IO - don't initialize when reading from disk * expand overfitting tests to also check IO goes OK * remove dropout from HashEmbed to fix Tagger performance * add defaults for sentrec * update thinc * always pass a Model instance to a Pipe * fix piped_added statement * remove obsolete W029 * remove obsolete errors * restore byte checking tests (work again) * clean up test * further test cleanup * convert from config to Model in create_pipe * bring back error when component is not initialized * cleanup * remove calls for nlp2.begin_training * use thinc.api in imports * allow setting charembed's nM and nC * fix for hardcoded nM/nC + unit test * formatting fixes * trigger build
2020-02-27 20:42:27 +03:00
return model
def build_cloze_characters_multi_task_model(
2020-07-31 18:02:54 +03:00
vocab: "Vocab", tok2vec: Model, maxout_pieces: int, hidden_size: int, nr_char: int
) -> Model:
output_layer = chain(
list2array(),
Maxout(nO=hidden_size, nP=maxout_pieces),
LayerNorm(nI=hidden_size),
MultiSoftmax([256] * nr_char, nI=hidden_size),
)
model = build_masked_language_model(vocab, chain(tok2vec, output_layer))
model.set_ref("tok2vec", tok2vec)
model.set_ref("output_layer", output_layer)
return model
2020-07-31 18:02:54 +03:00
def build_masked_language_model(
vocab: "Vocab", wrapped_model: Model, mask_prob: float = 0.15
) -> Model:
2020-06-03 15:45:00 +03:00
"""Convert a model into a BERT-style masked language model"""
random_words = _RandomWords(vocab)
def mlm_forward(model, docs, is_train):
mask, docs = _apply_mask(docs, random_words, mask_prob=mask_prob)
mask = model.ops.asarray(mask).reshape((mask.shape[0], 1))
output, backprop = model.layers[0](docs, is_train)
2020-06-03 15:45:00 +03:00
def mlm_backward(d_output):
d_output *= 1 - mask
return backprop(d_output)
return output, mlm_backward
2020-07-31 18:02:54 +03:00
def mlm_initialize(model: Model, X=None, Y=None):
wrapped = model.layers[0]
wrapped.initialize(X=X, Y=Y)
for dim in wrapped.dim_names:
if wrapped.has_dim(dim):
model.set_dim(dim, wrapped.get_dim(dim))
mlm_model = Model(
"masked-language-model",
mlm_forward,
layers=[wrapped_model],
init=mlm_initialize,
refs={"wrapped": wrapped_model},
dims={dim: None for dim in wrapped_model.dim_names},
)
mlm_model.set_ref("wrapped", wrapped_model)
2020-06-03 15:45:00 +03:00
return mlm_model
2020-07-12 15:03:23 +03:00
class _RandomWords:
2020-07-31 18:02:54 +03:00
def __init__(self, vocab: "Vocab") -> None:
2020-06-03 15:45:00 +03:00
self.words = [lex.text for lex in vocab if lex.prob != 0.0]
self.probs = [lex.prob for lex in vocab if lex.prob != 0.0]
self.words = self.words[:10000]
self.probs = self.probs[:10000]
self.probs = numpy.exp(numpy.array(self.probs, dtype="f"))
self.probs /= self.probs.sum()
self._cache = []
2020-07-31 18:02:54 +03:00
def next(self) -> str:
2020-06-03 15:45:00 +03:00
if not self._cache:
self._cache.extend(
numpy.random.choice(len(self.words), 10000, p=self.probs)
)
index = self._cache.pop()
return self.words[index]
2020-07-31 18:02:54 +03:00
def _apply_mask(
docs: Iterable["Doc"], random_words: _RandomWords, mask_prob: float = 0.15
) -> Tuple[numpy.ndarray, List["Doc"]]:
2020-06-03 15:45:00 +03:00
# This needs to be here to avoid circular imports
2020-07-31 18:02:54 +03:00
from ...tokens import Doc # noqa: F811
2020-06-03 15:45:00 +03:00
N = sum(len(doc) for doc in docs)
mask = numpy.random.uniform(0.0, 1.0, (N,))
mask = mask >= mask_prob
i = 0
masked_docs = []
for doc in docs:
words = []
for token in doc:
if not mask[i]:
word = _replace_word(token.text, random_words)
else:
word = token.text
words.append(word)
i += 1
spaces = [bool(w.whitespace_) for w in doc]
# NB: If you change this implementation to instead modify
# the docs in place, take care that the IDs reflect the original
# words. Currently we use the original docs to make the vectors
# for the target, so we don't lose the original tokens. But if
# you modified the docs in place here, you would.
masked_docs.append(Doc(doc.vocab, words=words, spaces=spaces))
return mask, masked_docs
2020-07-31 18:02:54 +03:00
def _replace_word(word: str, random_words: _RandomWords, mask: str = "[MASK]") -> str:
2020-06-03 15:45:00 +03:00
roll = numpy.random.random()
if roll < 0.8:
return mask
elif roll < 0.9:
return random_words.next()
else:
return word