facebook

/

wav2vec2-xlsr-53-espeak-cv-ft

like 24

Follow

AI at Meta 3,280

Even though the model worked fine on a local machine with windows, when passing the path to the espeak-ng library according to this installation guide https://bootphon.github.io/phonemizer/install.html , I could not make it work in a VM under Ubuntu 22.04.4 LTS x86-64. When running my script to transcribe phonemes via wav2vec2phoneme, I got the following message

Traceback (most recent call last):
File "/dialrec/phoneme_transcription/phoneme_recognizers/transcribe.py", line 50, in
phoneme_recognizer = Wav2Vec2Phoneme()
File "/dialrec/phoneme_transcription/phoneme_recognizers/wav2vec2phoneme.py", line 24, in init
self.processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-xlsr-53-espeak-cv-ft")
File "/usr/local/lib/python3.10/site-packages/transformers/models/wav2vec2/processing_wav2vec2.py", line 52, in from_pretrained
return super().from_pretrained(pretrained_model_name_or_path, **kwargs)
File "/usr/local/lib/python3.10/site-packages/transformers/processing_utils.py", line 465, in from_pretrained
args = cls._get_arguments_from_pretrained(pretrained_model_name_or_path, **kwargs)
File "/usr/local/lib/python3.10/site-packages/transformers/processing_utils.py", line 511, in _get_arguments_from_pretrained
args.append(attribute_class.from_pretrained(pretrained_model_name_or_path, **kwargs))
File "/usr/local/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py", line 837, in from_pretrained
return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
File "/usr/local/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 2086, in from_pretrained
return cls._from_pretrained(
File "/usr/local/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 2325, in _from_pretrained
tokenizer = cls(*init_inputs, **init_kwargs)
File "/usr/local/lib/python3.10/site-packages/transformers/models/wav2vec2_phoneme/tokenization_wav2vec2_phoneme.py", line 153, in init
self.init_backend(self.phonemizer_lang)
File "/usr/local/lib/python3.10/site-packages/transformers/models/wav2vec2_phoneme/tokenization_wav2vec2_phoneme.py", line 202, in init_backend
self.backend = BACKENDS[self.phonemizer_backend](phonemizer_lang, language_switch="remove-flags")
File "/usr/local/lib/python3.10/site-packages/phonemizer/backend/espeak/espeak.py", line 45, in init
super().init(
File "/usr/local/lib/python3.10/site-packages/phonemizer/backend/espeak/base.py", line 39, in init
super().init(
File "/usr/local/lib/python3.10/site-packages/phonemizer/backend/base.py", line 77, in init
raise RuntimeError( # pragma: nocover
RuntimeError: espeak not installed on your system

For installing espeak, I followed these steps:

1. apt-get install espeak-ng
2. pip3 install phonemizer
3. pip3 install espeakng (also tried pip3 install py-espeak-ng)

Espeak is definitely installed under /usr/lib/x86_64-linux-gnu/libespeak-ng.so.1 and /usr/bin/espeak-ng.

I tried the following:

• without additional steps
• Setting the environmental variable PHONEMIZER_ESPEAK_LIBRARY='/usr/lib/x86_64-linux-gnu/libespeak-ng.so.1' and PHONEMIZER_ESPEAK_PATH='/usr/bin/espeak-ng'.
• Setting the environmental variable directly in the script with
  os.environ['PHONEMIZER_ESPEAK_LIBRARY'] = '/usr/lib/x86_64-linux-gnu/libespeak-ng.so.1'
  os.environ['PHONEMIZER_ESPEAK_PATH'] = '/usr/bin/espeak-ng'

I would appreciate any help. Thanks in advance.