SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Add strip_accents to basic BertTokenizer. (#6280)

Browse Source 
* Add strip_accents to basic tokenizer

* Add tests for strip_accents.

* fix style with black

* Fix strip_accents test

* empty commit to trigger CI

* Improved strip_accents check

* Add code quality with is not False
This commit is contained in:
Philip May
2020-08-06 12:52:28 +02:00
committed by GitHub
parent 31da35cc89
commit d5bc32ce92
2 changed files with 58 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

38
tests/test_tokenization_bert.py
View File

@@ -130,6 +130,30 @@ class BertTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
)
self.assertListEqual(tokenizer.tokenize("H\u00E9llo"), ["hello"])
def test_basic_tokenizer_lower_strip_accents_false(self):
tokenizer = BasicTokenizer(do_lower_case=True, strip_accents=False)
self.assertListEqual(
tokenizer.tokenize(" \tHäLLo!how \n Are yoU? "), ["hällo", "!", "how", "are", "you", "?"]
)
self.assertListEqual(tokenizer.tokenize("H\u00E9llo"), ["h\u00E9llo"])
def test_basic_tokenizer_lower_strip_accents_true(self):
tokenizer = BasicTokenizer(do_lower_case=True, strip_accents=True)
self.assertListEqual(
tokenizer.tokenize(" \tHäLLo!how \n Are yoU? "), ["hallo", "!", "how", "are", "you", "?"]
)
self.assertListEqual(tokenizer.tokenize("H\u00E9llo"), ["hello"])
def test_basic_tokenizer_lower_strip_accents_default(self):
tokenizer = BasicTokenizer(do_lower_case=True)
self.assertListEqual(
tokenizer.tokenize(" \tHäLLo!how \n Are yoU? "), ["hallo", "!", "how", "are", "you", "?"]
)
self.assertListEqual(tokenizer.tokenize("H\u00E9llo"), ["hello"])
def test_basic_tokenizer_no_lower(self):
tokenizer = BasicTokenizer(do_lower_case=False)
@@ -137,6 +161,20 @@ class BertTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
tokenizer.tokenize(" \tHeLLo!how \n Are yoU? "), ["HeLLo", "!", "how", "Are", "yoU", "?"]
)
def test_basic_tokenizer_no_lower_strip_accents_false(self):
tokenizer = BasicTokenizer(do_lower_case=False, strip_accents=False)
self.assertListEqual(
tokenizer.tokenize(" \tHäLLo!how \n Are yoU? "), ["HäLLo", "!", "how", "Are", "yoU", "?"]
)
def test_basic_tokenizer_no_lower_strip_accents_true(self):
tokenizer = BasicTokenizer(do_lower_case=False, strip_accents=True)
self.assertListEqual(
tokenizer.tokenize(" \tHäLLo!how \n Are yoU? "), ["HaLLo", "!", "how", "Are", "yoU", "?"]
)
def test_basic_tokenizer_respects_never_split_tokens(self):
tokenizer = BasicTokenizer(do_lower_case=False, never_split=["[UNK]"])