Hugging Face's logo

utter-project
/
mHuBERT-147

Feature Extraction
Transformers
PyTorch
Safetensors
hubert
Model card Files
xet
 Community
16

Error with loading tokenizer.

#8
by lilias - opened
Discussion
lilias

Hello,
I am using: AutoProcessor.from_pretrained("utter-project/mHuBERT-147")

I receive this error:
OSError: Can't load tokenizer for 'utter-project/mHuBERT-147'. If you were trying to load it from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, make sure 'utter-project/mHuBERT-147' is the correct path to a directory containing all relevant files for a Wav2Vec2CTCTokenizer tokenizer.
mzboito
 UTTER - Unified Transcription and Translation for Extended Reality org

Hello,

Thanks for the interest in using our model.
There is no tokenizer associated to this release, as mHuBERT-147 is not an ASR model. It is a speech representation model.

You can, however, use it to train an ASR system, if you want. :)

All the best,
mzboito changed discussion status to closed
lilias

Hello,
Thank you for your answer.
I don't use it as an ASR model.

So my question is:
self.feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained("utter-project/mHuBERT-147")
self.hubert = HubertModel.from_pretrained("utter-project/mHuBERT-147")

Do I have to use it in this way?
mzboito
 UTTER - Unified Transcription and Translation for Extended Reality org

I'm not sure what you want to do, but the code you sent is correct. It loads the pretrained model correctly!
clementruhm

I think they try to follow generic Hubert class documentation: https://huggingface.co/docs/transformers/en/model_doc/hubert#transformers.HubertModel.forward.example

Sign up or log in to comment