spaCy/spacy/_morphologizer.pyx
2018-09-25 10:58:13 +02:00

139 lines
4.9 KiB
Cython

from __future__ import unicode_literals
from collections import OrderedDict, defaultdict
import cytoolz
import ujson
import numpy
cimport numpy as np
from .util import msgpack
from .util import msgpack_numpy
from thinc.api import chain
from thinc.neural.util import to_categorical, copy_array
from . import util
from .pipe import Pipe
from ._ml import Tok2Vec, build_morphologizer_model
from ._ml import link_vectors_to_models, zero_init, flatten
from ._ml import create_default_optimizer
from .errors import Errors, TempErrors
from .compat import json_dumps, basestring_
from .tokens.doc cimport Doc
from .vocab cimport Vocab
from .morphology cimport Morphology
from .pipeline import Pipe
class Morphologizer(Pipe):
name = 'morphologizer'
@classmethod
def Model(cls, attr_nums, **cfg):
if cfg.get('pretrained_dims') and not cfg.get('pretrained_vectors'):
raise ValueError(TempErrors.T008)
return build_morphologizer_model(attr_nums, **cfg)
def __init__(self, vocab, model=True, **cfg):
self.vocab = vocab
self.model = model
self.cfg = OrderedDict(sorted(cfg.items()))
self.cfg.setdefault('cnn_maxout_pieces', 2)
@property
def labels(self):
return self.vocab.morphology.tag_names
@property
def tok2vec(self):
if self.model in (None, True, False):
return None
else:
return chain(self.model.tok2vec, flatten)
def __call__(self, doc):
features, tokvecs = self.predict([doc])
self.set_annotations([doc], features, tensors=tokvecs)
return doc
def pipe(self, stream, batch_size=128, n_threads=-1):
for docs in cytoolz.partition_all(batch_size, stream):
docs = list(docs)
features, tokvecs = self.predict(docs)
self.set_annotations(docs, features, tensors=tokvecs)
yield from docs
def predict(self, docs):
if not any(len(doc) for doc in docs):
# Handle case where there are no tokens in any docs.
n_labels = self.model.nO
guesses = [self.model.ops.allocate((0, n_labels)) for doc in docs]
tokvecs = self.model.ops.allocate((0, self.model.tok2vec.nO))
return guesses, tokvecs
tokvecs = self.model.tok2vec(docs)
scores = self.model.softmax(tokvecs)
guesses = []
# Resolve multisoftmax into guesses
for doc_scores in scores:
guesses.append(scores_to_guesses(doc_scores, self.model.softmax.out_sizes))
return guesses, tokvecs
def set_annotations(self, docs, batch_feature_ids, tensors=None):
if isinstance(docs, Doc):
docs = [docs]
cdef Doc doc
cdef Vocab vocab = self.vocab
for i, doc in enumerate(docs):
doc_feat_ids = batch_feature_ids[i]
if hasattr(doc_feat_ids, 'get'):
doc_feat_ids = doc_feat_ids.get()
# Convert the neuron indices into feature IDs.
offset = self.vocab.morphology.first_feature
for j, nr_feat in enumerate(self.model.softmax.out_sizes):
doc_feat_ids[:, j] += offset
offset += nr_feat
# Now add the analysis, and set the hash.
for j in range(doc_feat_ids.shape[0]):
doc.c[j].morph = self.vocab.morphology.add(doc_feat_ids[j])
def update(self, docs, golds, drop=0., sgd=None, losses=None):
if losses is not None and self.name not in losses:
losses[self.name] = 0.
tag_scores, bp_tag_scores = self.model.begin_update(docs, drop=drop)
loss, d_tag_scores = self.get_loss(docs, golds, tag_scores)
bp_tag_scores(d_tag_scores, sgd=sgd)
if losses is not None:
losses[self.name] += loss
def get_loss(self, docs, golds, scores):
guesses = []
for doc_scores in scores:
guesses.append(scores_to_guesses(doc_scores, self.model.softmax.out_sizes))
guesses = self.model.ops.flatten(guesses)
cdef int idx = 0
target = numpy.zeros(scores.shape, dtype='f')
for gold in golds:
for features in gold.morphology:
if features is None:
target[idx] = guesses[idx]
else:
for feature in features:
column = feature_to_column(feature) # TODO
target[idx, column] = 1
idx += 1
target = self.model.ops.xp.array(target, dtype='f')
d_scores = scores - target
loss = (d_scores**2).sum()
d_scores = self.model.ops.unflatten(d_scores, [len(d) for d in docs])
return float(loss), d_scores
def use_params(self, params):
with self.model.use_params(params):
yield
def scores_to_guesses(scores, out_sizes):
raise NotImplementedError
def feature_to_column(feature):
raise NotImplementedError