Make OpenAIGPTTokenizer work with SpaCy 2.x and 3.x (#15019)

* Make OpenAIGPTTokenizer work with SpaCy 3.x

SpaCy 3.x introduced an API change to creating the tokenizer that
breaks OpenAIGPTTokenizer. The old API for creating the tokenizer in
SpaCy 2.x no longer works under SpaCy 3.x, but the new API for creating
the tokenizer in SpaCy 3.x DOES work under SpaCy 2.x. Switching to the
new API should allow OpenAIGPTTokenizer to work under both SpaCy 2.x and
SpaCy 3.x versions.

* Add is_spacy_available and is_ftfy_available methods to file utils

* Add spacy and ftfy unittest decorator to testing utils

* Add tests for OpenAIGPTTokenizer that require spacy and ftfy

* Modify CircleCI config to run tests that require spacy and ftfy

* Remove unneeded unittest decorators are reuse test code

* Run make fixup

.circleci/config.yml

@@ -99,7 +99,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_torch_and_tf_all:
         working_directory: ~/transformers
         docker:
@@ -169,7 +169,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_torch_and_flax_all:
         working_directory: ~/transformers
         docker:
@@ -237,7 +237,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_torch_all:
         working_directory: ~/transformers
         docker:
@@ -304,7 +304,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_tf_all:
         working_directory: ~/transformers
         docker:
@@ -370,7 +370,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_flax_all:
         working_directory: ~/transformers
         docker:
@@ -437,7 +437,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_pipelines_torch_all:
         working_directory: ~/transformers
         docker:
@@ -549,7 +549,7 @@ jobs:
                       - v0.4-custom_tokenizers-{{ checksum "setup.py" }}
                       - v0.4-{{ checksum "setup.py" }}
             - run: pip install --upgrade pip
-            - run: pip install .[ja,testing,sentencepiece,jieba]
+            - run: pip install .[ja,testing,sentencepiece,jieba,spacy,ftfy]
             - run: python -m unidic download
             - save_cache:
                   key: v0.4-custom_tokenizers-{{ checksum "setup.py" }}
@@ -557,7 +557,7 @@ jobs:
                       - '~/.cache/pip'
             - run: |
                   if [ -f test_list.txt ]; then
-                    python -m pytest -s --make-reports=tests_custom_tokenizers ./tests/test_tokenization_bert_japanese.py | tee tests_output.txt
+                    python -m pytest -s --make-reports=tests_custom_tokenizers ./tests/test_tokenization_bert_japanese.py ./tests/test_tokenization_openai.py | tee tests_output.txt
                   fi
             - store_artifacts:
                   path: ~/transformers/tests_output.txt
@@ -662,7 +662,7 @@ jobs:
                   path: ~/transformers/flax_examples_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_examples_flax_all:
         working_directory: ~/transformers
         docker:
@@ -729,7 +729,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_hub_all:
         working_directory: ~/transformers
         docker:
@@ -795,7 +795,7 @@ jobs:
                   path: ~/transformers/tests_output.txt
             - store_artifacts:
                   path: ~/transformers/reports
-    
+
     run_tests_onnxruntime_all:
         working_directory: ~/transformers
         docker:

src/transformers/file_utils.py

@@ -512,6 +512,14 @@ def is_pytesseract_available():
     return importlib.util.find_spec("pytesseract") is not None
 
 
+def is_spacy_available():
+    return importlib.util.find_spec("spacy") is not None
+
+
+def is_ftfy_available():
+    return importlib.util.find_spec("ftfy") is not None
+
+
 def is_in_notebook():
     try:
         # Test adapted from tqdm.autonotebook: https://github.com/tqdm/tqdm/blob/master/tqdm/autonotebook.py

src/transformers/models/openai/tokenization_openai.py

@@ -104,7 +104,7 @@ class OpenAIGPTTokenizer(PreTrainedTokenizer):
             from spacy.lang.en import English
 
             _nlp = English()
-            self.nlp = _nlp.Defaults.create_tokenizer(_nlp)
+            self.nlp = _nlp.tokenizer
             self.fix_text = ftfy.fix_text
         except ImportError:
             logger.warning("ftfy or spacy is not installed using BERT BasicTokenizer instead of SpaCy & ftfy.")

src/transformers/testing_utils.py

@@ -34,6 +34,7 @@ from .file_utils import (
     is_detectron2_available,
     is_faiss_available,
     is_flax_available,
+    is_ftfy_available,
     is_keras2onnx_available,
     is_librosa_available,
     is_onnx_available,
@@ -46,6 +47,7 @@ from .file_utils import (
     is_scatter_available,
     is_sentencepiece_available,
     is_soundfile_availble,
+    is_spacy_available,
     is_tensorflow_probability_available,
     is_tf_available,
     is_timm_available,
@@ -412,6 +414,26 @@ def require_vision(test_case):
         return test_case
 
 
+def require_ftfy(test_case):
+    """
+    Decorator marking a test that requires ftfy. These tests are skipped when ftfy isn't installed.
+    """
+    if not is_ftfy_available():
+        return unittest.skip("test requires ftfy")(test_case)
+    else:
+        return test_case
+
+
+def require_spacy(test_case):
+    """
+    Decorator marking a test that requires SpaCy. These tests are skipped when SpaCy isn't installed.
+    """
+    if not is_spacy_available():
+        return unittest.skip("test requires spacy")(test_case)
+    else:
+        return test_case
+
+
 def require_torch_multi_gpu(test_case):
     """
     Decorator marking a test that requires a multi-GPU setup (in PyTorch). These tests are skipped on a machine without

tests/test_tokenization_openai.py

@@ -20,13 +20,14 @@ import unittest
 
 from transformers import OpenAIGPTTokenizer, OpenAIGPTTokenizerFast
 from transformers.models.openai.tokenization_openai import VOCAB_FILES_NAMES
-from transformers.testing_utils import require_tokenizers
+from transformers.testing_utils import require_ftfy, require_spacy, require_tokenizers
 
 from .test_tokenization_common import TokenizerTesterMixin
 
 
 @require_tokenizers
 class OpenAIGPTTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
+    """Tests OpenAIGPTTokenizer that uses BERT BasicTokenizer."""
 
     tokenizer_class = OpenAIGPTTokenizer
     rust_tokenizer_class = OpenAIGPTTokenizerFast
@@ -132,3 +133,12 @@ class OpenAIGPTTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
     # tokenizer has no padding token
     def test_padding_different_model_input_name(self):
         pass
+
+
+@require_ftfy
+@require_spacy
+@require_tokenizers
+class OpenAIGPTTokenizationTestWithSpacy(OpenAIGPTTokenizationTest):
+    """Tests OpenAIGPTTokenizer that uses SpaCy and ftfy."""
+
+    pass