riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 02:02:21 +06:00
Code Issues Actions 33 Packages Projects Releases Wiki Activity

Exclude the madeup words from M2M100Tokenizer.vocab_size (#20976)

Browse Source
This commit is contained in:
Guillaume Klein 2023-02-08 15:19:06 +01:00 committed by GitHub
parent cc1d0685b3
commit ca905ba28e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 8 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/transformers/models/m2m_100/tokenization_m2m_100.py
View File

@ -193,7 +193,7 @@ class M2M100Tokenizer(PreTrainedTokenizer):
@property
def vocab_size(self) -> int:
return len(self.encoder) + len(self.lang_token_to_id) + self.num_madeup_words
return len(self.encoder) + len(self.lang_token_to_id)
@property
def src_lang(self) -> str:

13
tests/models/m2m_100/test_tokenization_m2m_100.py
View File

@ -84,15 +84,13 @@ class M2M100TokenizationTest(TokenizerTesterMixin, unittest.TestCase):
self.assertEqual(self.get_tokenizer()._convert_id_to_token(token_id), token)
def test_get_vocab(self):
vocab_keys = list(self.get_tokenizer().get_vocab().keys())
tokenizer = self.get_tokenizer()
vocab_keys = list(tokenizer.get_vocab().keys())
self.assertEqual(vocab_keys[0], "</s>")
self.assertEqual(vocab_keys[1], "<unk>")
self.assertEqual(vocab_keys[-1], "<s>")
self.assertEqual(len(vocab_keys), 110)
def test_vocab_size(self):
self.assertEqual(self.get_tokenizer().vocab_size, 117)
self.assertEqual(len(vocab_keys), tokenizer.vocab_size + len(tokenizer.get_added_vocab()))
@unittest.skip("Skip this test while all models are still to be uploaded.")
def test_pretrained_model_lists(self):
@ -161,7 +159,10 @@ class M2M100TokenizerIntegrationTest(unittest.TestCase):
self.assertEqual(self.tokenizer.get_lang_id("mr"), 128063)
def test_get_vocab(self):
self.assertIn(self.tokenizer.get_lang_token("en"), self.tokenizer.get_vocab())
vocab = self.tokenizer.get_vocab()
self.assertEqual(len(vocab), self.tokenizer.vocab_size)
self.assertEqual(vocab["<unk>"], 3)
self.assertIn(self.tokenizer.get_lang_token("en"), vocab)
def test_tokenizer_batch_encode_plus(self):
self.tokenizer.src_lang = "en"