Spaces:
Sleeping
Sleeping
Pradeep Kumar
commited on
Delete classifier_data_lib_test.py
Browse files- classifier_data_lib_test.py +0 -95
classifier_data_lib_test.py
DELETED
@@ -1,95 +0,0 @@
|
|
1 |
-
# Copyright 2024 The TensorFlow Authors. All Rights Reserved.
|
2 |
-
#
|
3 |
-
# Licensed under the Apache License, Version 2.0 (the "License");
|
4 |
-
# you may not use this file except in compliance with the License.
|
5 |
-
# You may obtain a copy of the License at
|
6 |
-
#
|
7 |
-
# http://www.apache.org/licenses/LICENSE-2.0
|
8 |
-
#
|
9 |
-
# Unless required by applicable law or agreed to in writing, software
|
10 |
-
# distributed under the License is distributed on an "AS IS" BASIS,
|
11 |
-
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
12 |
-
# See the License for the specific language governing permissions and
|
13 |
-
# limitations under the License.
|
14 |
-
|
15 |
-
"""Tests for third_party.tensorflow_models.official.nlp.data.classifier_data_lib."""
|
16 |
-
|
17 |
-
import os
|
18 |
-
import tempfile
|
19 |
-
|
20 |
-
from absl.testing import parameterized
|
21 |
-
import tensorflow as tf, tf_keras
|
22 |
-
import tensorflow_datasets as tfds
|
23 |
-
|
24 |
-
from official.nlp.data import classifier_data_lib
|
25 |
-
from official.nlp.tools import tokenization
|
26 |
-
|
27 |
-
|
28 |
-
def decode_record(record, name_to_features):
|
29 |
-
"""Decodes a record to a TensorFlow example."""
|
30 |
-
return tf.io.parse_single_example(record, name_to_features)
|
31 |
-
|
32 |
-
|
33 |
-
class BertClassifierLibTest(tf.test.TestCase, parameterized.TestCase):
|
34 |
-
|
35 |
-
def setUp(self):
|
36 |
-
super(BertClassifierLibTest, self).setUp()
|
37 |
-
self.model_dir = self.get_temp_dir()
|
38 |
-
self.processors = {
|
39 |
-
"CB": classifier_data_lib.CBProcessor,
|
40 |
-
"SUPERGLUE-RTE": classifier_data_lib.SuperGLUERTEProcessor,
|
41 |
-
"BOOLQ": classifier_data_lib.BoolQProcessor,
|
42 |
-
"WIC": classifier_data_lib.WiCProcessor,
|
43 |
-
}
|
44 |
-
|
45 |
-
vocab_tokens = [
|
46 |
-
"[UNK]", "[CLS]", "[SEP]", "want", "##want", "##ed", "wa", "un", "runn",
|
47 |
-
"##ing", ","
|
48 |
-
]
|
49 |
-
with tempfile.NamedTemporaryFile(delete=False) as vocab_writer:
|
50 |
-
vocab_writer.write("".join([x + "\n" for x in vocab_tokens
|
51 |
-
]).encode("utf-8"))
|
52 |
-
vocab_file = vocab_writer.name
|
53 |
-
self.tokenizer = tokenization.FullTokenizer(vocab_file)
|
54 |
-
|
55 |
-
@parameterized.parameters(
|
56 |
-
{"task_type": "CB"},
|
57 |
-
{"task_type": "BOOLQ"},
|
58 |
-
{"task_type": "SUPERGLUE-RTE"},
|
59 |
-
{"task_type": "WIC"},
|
60 |
-
)
|
61 |
-
def test_generate_dataset_from_tfds_processor(self, task_type):
|
62 |
-
with tfds.testing.mock_data(num_examples=5):
|
63 |
-
output_path = os.path.join(self.model_dir, task_type)
|
64 |
-
|
65 |
-
processor = self.processors[task_type]()
|
66 |
-
|
67 |
-
classifier_data_lib.generate_tf_record_from_data_file(
|
68 |
-
processor,
|
69 |
-
None,
|
70 |
-
self.tokenizer,
|
71 |
-
train_data_output_path=output_path,
|
72 |
-
eval_data_output_path=output_path,
|
73 |
-
test_data_output_path=output_path)
|
74 |
-
files = tf.io.gfile.glob(output_path)
|
75 |
-
self.assertNotEmpty(files)
|
76 |
-
|
77 |
-
train_dataset = tf.data.TFRecordDataset(output_path)
|
78 |
-
seq_length = 128
|
79 |
-
label_type = tf.int64
|
80 |
-
name_to_features = {
|
81 |
-
"input_ids": tf.io.FixedLenFeature([seq_length], tf.int64),
|
82 |
-
"input_mask": tf.io.FixedLenFeature([seq_length], tf.int64),
|
83 |
-
"segment_ids": tf.io.FixedLenFeature([seq_length], tf.int64),
|
84 |
-
"label_ids": tf.io.FixedLenFeature([], label_type),
|
85 |
-
}
|
86 |
-
train_dataset = train_dataset.map(
|
87 |
-
lambda record: decode_record(record, name_to_features))
|
88 |
-
|
89 |
-
# If data is retrieved without error, then all requirements
|
90 |
-
# including data type/shapes are met.
|
91 |
-
_ = next(iter(train_dataset))
|
92 |
-
|
93 |
-
|
94 |
-
if __name__ == "__main__":
|
95 |
-
tf.test.main()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|