-
Notifications
You must be signed in to change notification settings - Fork 7.4k
/
question_answer.py
237 lines (205 loc) · 9.94 KB
/
question_answer.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
# Copyright 2020 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""APIs to train a model that can answer questions based on a predefined text."""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os
import tempfile
import tensorflow as tf
from tensorflow_examples.lite.model_maker.core import compat
from tensorflow_examples.lite.model_maker.core.api import mm_export
from tensorflow_examples.lite.model_maker.core.export_format import ExportFormat
from tensorflow_examples.lite.model_maker.core.task import custom_model
from tensorflow_examples.lite.model_maker.core.task import model_spec as ms
from tensorflow_examples.lite.model_maker.core.task import model_util
from tensorflow_examples.lite.model_maker.core.task.metadata_writers.bert.question_answerer import metadata_writer_for_bert_question_answerer as metadata_writer
def _get_model_info(model_spec, vocab_file):
"""Gets the specific info for the question answer model."""
return metadata_writer.QuestionAnswererInfo(
name=model_spec.name + ' Question and Answerer',
version='v1',
description=metadata_writer.DEFAULT_DESCRIPTION,
input_names=metadata_writer.bert_qa_inputs(
ids_name=model_spec.tflite_input_name['ids'],
mask_name=model_spec.tflite_input_name['mask'],
segment_ids_name=model_spec.tflite_input_name['segment_ids']),
output_names=metadata_writer.bert_qa_outputs(
start_logits_name=model_spec.tflite_output_name['start_logits'],
end_logits_name=model_spec.tflite_output_name['end_logits']),
tokenizer_type=metadata_writer.Tokenizer.BERT_TOKENIZER,
vocab_file=vocab_file)
@mm_export('question_answer.QuestionAnswer')
class QuestionAnswer(custom_model.CustomModel):
"""QuestionAnswer class for inference and exporting to tflite."""
DEFAULT_EXPORT_FORMAT = (ExportFormat.TFLITE, ExportFormat.VOCAB)
ALLOWED_EXPORT_FORMAT = (ExportFormat.TFLITE, ExportFormat.VOCAB,
ExportFormat.SAVED_MODEL)
def train(self,
train_data,
epochs=None,
batch_size=None,
steps_per_epoch=None):
"""Feeds the training data for training."""
if batch_size is None:
batch_size = self.model_spec.default_batch_size
if len(train_data) < batch_size:
raise ValueError('The size of the train_data (%d) couldn\'t be smaller '
'than batch_size (%d). To solve this problem, set '
'the batch_size smaller or increase the size of the '
'train_data.' % (len(train_data), batch_size))
train_ds = train_data.gen_dataset(batch_size, is_training=True)
steps_per_epoch = model_util.get_steps_per_epoch(steps_per_epoch,
batch_size, train_data)
if steps_per_epoch is not None:
train_ds = train_ds.take(steps_per_epoch)
self.model = self.model_spec.train(
train_ds=train_ds, epochs=epochs, steps_per_epoch=steps_per_epoch)
return self.model
def create_model(self):
self.model = self.model_spec.create_model()
def evaluate(self,
data,
max_answer_length=30,
null_score_diff_threshold=0.0,
verbose_logging=False,
output_dir=None):
"""Evaluate the model.
Args:
data: Data to be evaluated.
max_answer_length: The maximum length of an answer that can be generated.
This is needed because the start and end predictions are not conditioned
on one another.
null_score_diff_threshold: If null_score - best_non_null is greater than
the threshold, predict null. This is only used for SQuAD v2.
verbose_logging: If true, all of the warnings related to data processing
will be printed. A number of warnings are expected for a normal SQuAD
evaluation.
output_dir: The output directory to save output to json files:
predictions.json, nbest_predictions.json, null_odds.json. If None, skip
saving to json files.
Returns:
A dict contains two metrics: Exact match rate and F1 score.
"""
predict_batch_size = self.model_spec.predict_batch_size
ds = data.gen_dataset(predict_batch_size, is_training=False)
num_steps = int(len(data) / predict_batch_size)
return self.model_spec.evaluate(
self.model, None, ds, num_steps, data.examples, data.features,
data.squad_file, data.version_2_with_negative, max_answer_length,
null_score_diff_threshold, verbose_logging, output_dir)
def evaluate_tflite(self,
tflite_filepath,
data,
max_answer_length=30,
null_score_diff_threshold=0.0,
verbose_logging=False,
output_dir=None):
"""Evaluate the model.
Args:
tflite_filepath: File path to the TFLite model.
data: Data to be evaluated.
max_answer_length: The maximum length of an answer that can be generated.
This is needed because the start and end predictions are not conditioned
on one another.
null_score_diff_threshold: If null_score - best_non_null is greater than
the threshold, predict null. This is only used for SQuAD v2.
verbose_logging: If true, all of the warnings related to data processing
will be printed. A number of warnings are expected for a normal SQuAD
evaluation.
output_dir: The output directory to save output to json files:
predictions.json, nbest_predictions.json, null_odds.json. If None, skip
saving to json files.
Returns:
A dict contains two metrics: Exact match rate and F1 score.
"""
ds = data.gen_dataset(batch_size=1, is_training=False)
return self.model_spec.evaluate(
None, tflite_filepath, ds, len(data), data.examples, data.features,
data.squad_file, data.version_2_with_negative, max_answer_length,
null_score_diff_threshold, verbose_logging, output_dir)
def _export_tflite(self,
tflite_filepath,
quantization_config='default',
with_metadata=True,
export_metadata_json_file=False):
"""Converts the retrained model to tflite format and saves it.
Args:
tflite_filepath: File path to save tflite model.
quantization_config: Configuration for post-training quantization. If
'default', sets the `quantization_config` by default according to
`self.model_spec`. If None, exports the float tflite model without
quantization.
with_metadata: Whether the output tflite model contains metadata.
export_metadata_json_file: Whether to export metadata in json file. If
True, export the metadata in the same directory as tflite model.Used
only if `with_metadata` is True.
"""
if quantization_config == 'default':
quantization_config = self.model_spec.get_default_quantization_config()
# Sets batch size from None to 1 when converting to tflite.
model_util.set_batch_size(self.model, batch_size=1)
model_util.export_tflite(self.model, tflite_filepath, quantization_config,
self.model_spec.convert_from_saved_model_tf2)
# Sets batch size back to None to support retraining later.
model_util.set_batch_size(self.model, batch_size=None)
if with_metadata:
with tempfile.TemporaryDirectory() as temp_dir:
tf.compat.v1.logging.info(
'Vocab file is inside the TFLite model with metadata.')
vocab_filepath = os.path.join(temp_dir, 'vocab.txt')
self.model_spec.save_vocab(vocab_filepath)
model_info = _get_model_info(self.model_spec, vocab_filepath)
export_dir = os.path.dirname(tflite_filepath)
populator = metadata_writer.MetadataPopulatorForBertQuestionAndAnswer(
tflite_filepath, export_dir, model_info)
populator.populate(export_metadata_json_file)
@classmethod
def create(cls,
train_data,
model_spec,
batch_size=None,
epochs=2,
steps_per_epoch=None,
shuffle=False,
do_train=True):
"""Loads data and train the model for question answer.
Args:
train_data: Training data.
model_spec: Specification for the model.
batch_size: Batch size for training.
epochs: Number of epochs for training.
steps_per_epoch: Integer or None. Total number of steps (batches of
samples) before declaring one epoch finished and starting the next
epoch. If `steps_per_epoch` is None, the epoch will run until the input
dataset is exhausted.
shuffle: Whether the data should be shuffled.
do_train: Whether to run training.
Returns:
An instance based on QuestionAnswer.
"""
model_spec = ms.get(model_spec)
if compat.get_tf_behavior() not in model_spec.compat_tf_versions:
raise ValueError('Incompatible versions. Expect {}, but got {}.'.format(
model_spec.compat_tf_versions, compat.get_tf_behavior()))
model = cls(model_spec, shuffle=shuffle)
if do_train:
tf.compat.v1.logging.info('Retraining the models...')
model.train(train_data, epochs, batch_size, steps_per_epoch)
else:
model.create_model()
return model
# Shortcut function.
create = QuestionAnswer.create
mm_export('question_answer.create').export_constant(__name__, 'create')