def test_new_processor_registration(self):
try:
AutoConfig.register("custom", CustomConfig)
AutoFeatureExtractor.register(CustomConfig, CustomFeatureExtractor)
AutoTokenizer.register(CustomConfig, slow_tokenizer_class=CustomTokenizer)
AutoProcessor.register(CustomConfig, CustomProcessor)
# Trying to register something existing in the Transformers library will raise an error
with self.assertRaises(ValueError):
AutoProcessor.register(Wav2Vec2Config, Wav2Vec2Processor)
# Now that the config is registered, it can be used as any other config with the auto-API
feature_extractor = CustomFeatureExtractor.from_pretrained(SAMPLE_PROCESSOR_CONFIG_DIR)
with tempfile.TemporaryDirectory() as tmp_dir:
vocab_file = os.path.join(tmp_dir, "vocab.txt")
with open(vocab_file, "w", encoding="utf-8") as vocab_writer:
vocab_writer.write("".join([x + "\n" for x in self.vocab_tokens]))
tokenizer = CustomTokenizer(vocab_file)
processor = CustomProcessor(feature_extractor, tokenizer)
with tempfile.TemporaryDirectory() as tmp_dir:
processor.save_pretrained(tmp_dir)
new_processor = AutoProcessor.from_pretrained(tmp_dir)
self.assertIsInstance(new_processor, CustomProcessor)
finally:
if "custom" in CONFIG_MAPPING._extra_content:
del CONFIG_MAPPING._extra_content["custom"]
if CustomConfig in FEATURE_EXTRACTOR_MAPPING._extra_content:
del FEATURE_EXTRACTOR_MAPPING._extra_content[CustomConfig]
if CustomConfig in TOKENIZER_MAPPING._extra_content:
del TOKENIZER_MAPPING._extra_content[CustomConfig]
if CustomConfig in PROCESSOR_MAPPING._extra_content:
del PROCESSOR_MAPPING._extra_content[CustomConfig]
def test_new_feature_extractor_registration(self):
try:
AutoConfig.register("custom", CustomConfig)
AutoFeatureExtractor.register(CustomConfig, CustomFeatureExtractor)
# Trying to register something existing in the Transformers library will raise an error
with self.assertRaises(ValueError):
AutoFeatureExtractor.register(Wav2Vec2Config, Wav2Vec2FeatureExtractor)
# Now that the config is registered, it can be used as any other config with the auto-API
feature_extractor = CustomFeatureExtractor.from_pretrained(SAMPLE_FEATURE_EXTRACTION_CONFIG_DIR)
with tempfile.TemporaryDirectory() as tmp_dir:
feature_extractor.save_pretrained(tmp_dir)
new_feature_extractor = AutoFeatureExtractor.from_pretrained(tmp_dir)
self.assertIsInstance(new_feature_extractor, CustomFeatureExtractor)
finally:
if "custom" in CONFIG_MAPPING._extra_content:
del CONFIG_MAPPING._extra_content["custom"]
if CustomConfig in FEATURE_EXTRACTOR_MAPPING._extra_content:
del FEATURE_EXTRACTOR_MAPPING._extra_content[CustomConfig]