spaCy/spacy/ml/models/textcat.py
Sofie Van Landeghem 311133e579
Train textcat with config (#5143)
* bring back default build_text_classifier method

* remove _set_dims_ hack in favor of proper dim inference

* add tok2vec initialize to unit test

* small fixes

* add unit test for various textcat config settings

* logistic output layer does not have nO

* fix window_size setting

* proper fix

* fix W initialization

* Update textcat training example

* Use ml_datasets
* Convert training data to `Example` format
* Use `n_texts` to set proportionate dev size

* fix _init renaming on latest thinc

* avoid setting a non-existing dim

* update to thinc==8.0.0a2

* add BOW and CNN defaults for easy testing

* various experiments with train_textcat script, fix softmax activation in textcat bow

* allow textcat train script to work on other datasets as well

* have dataset as a parameter

* train textcat from config, with example config

* add config for training textcat

* formatting

* fix exclusive_classes

* fixing BOW for GPU

* bump thinc to 8.0.0a3 (not published yet so CI will fail)

* add in link_vectors_to_models which got deleted

Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
2020-03-29 19:40:36 +02:00

136 lines
5.7 KiB
Python

from thinc.api import Model, reduce_mean, Linear, list2ragged, Logistic, ParametricAttention
from thinc.api import chain, concatenate, clone, Dropout
from thinc.api import SparseLinear, Softmax, softmax_activation, Maxout, reduce_sum, Relu, residual, expand_window
from thinc.api import HashEmbed, with_ragged, with_array, with_cpu, uniqued, FeatureExtractor
from ..spacy_vectors import SpacyVectors
from ... import util
from ...attrs import ID, ORTH, NORM, PREFIX, SUFFIX, SHAPE, LOWER
from ...util import registry
from ..extract_ngrams import extract_ngrams
@registry.architectures.register("spacy.TextCatCNN.v1")
def build_simple_cnn_text_classifier(tok2vec, exclusive_classes, nO=None):
"""
Build a simple CNN text classifier, given a token-to-vector model as inputs.
If exclusive_classes=True, a softmax non-linearity is applied, so that the
outputs sum to 1. If exclusive_classes=False, a logistic non-linearity
is applied instead, so that outputs are in the range [0, 1].
"""
with Model.define_operators({">>": chain}):
if exclusive_classes:
output_layer = Softmax(nO=nO, nI=tok2vec.get_dim("nO"))
model = tok2vec >> list2ragged() >> reduce_mean() >> output_layer
model.set_ref("output_layer", output_layer)
else:
linear_layer = Linear(nO=nO, nI=tok2vec.get_dim("nO"))
model = (
tok2vec >> list2ragged() >> reduce_mean() >> linear_layer >> Logistic()
)
model.set_ref("output_layer", linear_layer)
model.set_ref("tok2vec", tok2vec)
model.set_dim("nO", nO)
return model
@registry.architectures.register("spacy.TextCatBOW.v1")
def build_bow_text_classifier(exclusive_classes, ngram_size, no_output_layer, nO=None):
with Model.define_operators({">>": chain}):
sparse_linear = SparseLinear(nO)
model = extract_ngrams(ngram_size, attr=ORTH) >> sparse_linear
model = with_cpu(model, model.ops)
if not no_output_layer:
output_layer = softmax_activation() if exclusive_classes else Logistic()
model = model >> with_cpu(output_layer, output_layer.ops)
model.set_ref("output_layer", sparse_linear)
return model
@registry.architectures.register("spacy.TextCat.v1")
def build_text_classifier(width, embed_size, pretrained_vectors, exclusive_classes, ngram_size,
window_size, conv_depth, nO=None):
cols = [ORTH, LOWER, PREFIX, SUFFIX, SHAPE, ID]
with Model.define_operators({">>": chain, "|": concatenate, "**": clone}):
lower = HashEmbed(nO=width, nV=embed_size, column=cols.index(LOWER))
prefix = HashEmbed(nO=width // 2, nV=embed_size, column=cols.index(PREFIX))
suffix = HashEmbed(nO=width // 2, nV=embed_size, column=cols.index(SUFFIX))
shape = HashEmbed(nO=width // 2, nV=embed_size, column=cols.index(SHAPE))
width_nI = sum(layer.get_dim("nO") for layer in [lower, prefix, suffix, shape])
trained_vectors = FeatureExtractor(cols) >> with_array(
uniqued(
(lower | prefix | suffix | shape)
>> Maxout(nO=width, nI=width_nI, normalize=True),
column=cols.index(ORTH),
)
)
if pretrained_vectors:
nlp = util.load_model(pretrained_vectors)
vectors = nlp.vocab.vectors
vector_dim = vectors.data.shape[1]
static_vectors = SpacyVectors(vectors) >> with_array(
Linear(width, vector_dim)
)
vector_layer = trained_vectors | static_vectors
vectors_width = width * 2
else:
vector_layer = trained_vectors
vectors_width = width
tok2vec = vector_layer >> with_array(
Maxout(width, vectors_width, normalize=True)
>> residual((expand_window(window_size=window_size)
>> Maxout(nO=width, nI=width * ((window_size * 2) + 1), normalize=True))) ** conv_depth,
pad=conv_depth,
)
cnn_model = (
tok2vec
>> list2ragged()
>> ParametricAttention(width)
>> reduce_sum()
>> residual(Maxout(nO=width, nI=width))
>> Linear(nO=nO, nI=width)
>> Dropout(0.0)
)
linear_model = build_bow_text_classifier(
nO=nO, ngram_size=ngram_size, exclusive_classes=exclusive_classes, no_output_layer=False
)
nO_double = nO*2 if nO else None
if exclusive_classes:
output_layer = Softmax(nO=nO, nI=nO_double)
else:
output_layer = (
Linear(nO=nO, nI=nO_double) >> Dropout(0.0) >> Logistic()
)
model = (linear_model | cnn_model) >> output_layer
model.set_ref("tok2vec", tok2vec)
if model.has_dim("nO") is not False:
model.set_dim("nO", nO)
model.set_ref("output_layer", linear_model.get_ref("output_layer"))
return model
@registry.architectures.register("spacy.TextCatLowData.v1")
def build_text_classifier_lowdata(width, pretrained_vectors, nO=None):
nlp = util.load_model(pretrained_vectors)
vectors = nlp.vocab.vectors
vector_dim = vectors.data.shape[1]
# Note, before v.3, this was the default if setting "low_data" and "pretrained_dims"
with Model.define_operators({">>": chain, "**": clone}):
model = (
SpacyVectors(vectors)
>> list2ragged()
>> with_ragged(0, Linear(width, vector_dim))
>> ParametricAttention(width)
>> reduce_sum()
>> residual(Relu(width, width)) ** 2
>> Linear(nO, width)
>> Dropout(0.0)
>> Logistic()
)
return model