File size: 1,976 Bytes
74e8f2f |
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 |
# Copyright 2024 Big Vision Authors.
#
# 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.
"""Tests for bert_ops."""
import tempfile
from big_vision import input_pipeline
import big_vision.pp.builder as pp_builder
import big_vision.pp.ops_general # pylint: disable=unused-import
from big_vision.pp.proj.flaxformer import bert_ops # pylint: disable=unused-import
import tensorflow as tf
# BERT vocabulary for testing.
_BERT_VOCAB = [
"[PAD]",
"[UNK]",
"more",
"than",
"one",
"[CLS]",
"[SEP]",
]
def _create_ds(pp_str, tensor_slices, num_examples):
return input_pipeline.make_for_inference(
tf.data.Dataset.from_tensor_slices(tensor_slices),
num_ex_per_process=[num_examples],
preprocess_fn=pp_builder.get_preprocess_fn(pp_str),
batch_size=num_examples,
)[0]
class BertOpsTest(tf.test.TestCase):
def test_tokenize(self):
inkey = "texts"
vocab_path = f"{tempfile.mkdtemp()}/vocab.txt"
with open(vocab_path, "w") as f:
f.write("\n".join(_BERT_VOCAB))
pp_str = (
f"bert_tokenize(inkey='{inkey}', vocab_path='{vocab_path}', max_len=5)"
f"|keep('labels')"
)
tensor_slices = {
inkey: tf.ragged.constant([["one more"], ["more than one"], [""]])
}
ds = _create_ds(pp_str, tensor_slices, 3)
self.assertAllEqual(
next(iter(ds))["labels"],
[[5, 4, 2, 0, 0], [5, 2, 3, 4, 0], [5, 0, 0, 0, 0]],
)
if __name__ == "__main__":
tf.test.main()
|