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

Deprecate prepare_seq2seq_batch (#10287)

Browse Source 
* Deprecate prepare_seq2seq_batch

* Fix last tests

* Apply suggestions from code review

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Suraj Patil <surajp815@gmail.com>

* More review comments

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Suraj Patil <surajp815@gmail.com>
This commit is contained in:
Sylvain Gugger
2021-02-22 12:36:16 -05:00
committed by GitHub
parent e73a3e1891
commit 9e147d31f6
31 changed files with 325 additions and 320 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
tests/test_tokenization_marian.py
View File

@@ -70,7 +70,7 @@ class MarianTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
def test_tokenizer_equivalence_en_de(self):
en_de_tokenizer = MarianTokenizer.from_pretrained(f"{ORG_NAME}opus-mt-en-de")
batch = en_de_tokenizer.prepare_seq2seq_batch(["I am a small frog"], return_tensors=None)
batch = en_de_tokenizer(["I am a small frog"], return_tensors=None)
self.assertIsInstance(batch, BatchEncoding)
expected = [38, 121, 14, 697, 38848, 0]
self.assertListEqual(expected, batch.input_ids[0])
@@ -84,12 +84,14 @@ class MarianTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
def test_outputs_not_longer_than_maxlen(self):
tok = self.get_tokenizer()
batch = tok.prepare_seq2seq_batch(["I am a small frog" * 1000, "I am a small frog"], return_tensors=FRAMEWORK)
batch = tok(
["I am a small frog" * 1000, "I am a small frog"], padding=True, truncation=True, return_tensors=FRAMEWORK
)
self.assertIsInstance(batch, BatchEncoding)
self.assertEqual(batch.input_ids.shape, (2, 512))
def test_outputs_can_be_shorter(self):
tok = self.get_tokenizer()
batch_smaller = tok.prepare_seq2seq_batch(["I am a tiny frog", "I am a small frog"], return_tensors=FRAMEWORK)
batch_smaller = tok(["I am a tiny frog", "I am a small frog"], padding=True, return_tensors=FRAMEWORK)
self.assertIsInstance(batch_smaller, BatchEncoding)
self.assertEqual(batch_smaller.input_ids.shape, (2, 10))