espnet/mms_1b_mlsuperb

ESPnet2 ASR model

This is a simple baseline for the ML-SUPERB 2.0 Challenge. It is a self-supervised MMS 1B model fine-tuned on 142 languages of ML-SUPERB using CTC loss. The MMS model is frozen and used as a feature extractor for a small Transformer encoder during fine-tuning, which took approximately 1 day on a single GPU.

The model was trained using the ML-SUPERB recipe in ESPnet. Inference can be performed with the following script:

from espnet2.bin.asr_inference import Speech2Text

model = Speech2Text.from_pretrained(
  "espnet/mms_1b_mlsuperb"
)

speech, rate = soundfile.read("speech.wav")
text, *_ = model(speech)[0]

Demo: How to use in ESPnet2

Follow the ESPnet installation instructions if you haven't done that already.

Citing ESPnet

@inproceedings{watanabe2018espnet,
  author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
  title={{ESPnet}: End-to-End Speech Processing Toolkit},
  year={2018},
  booktitle={Proceedings of Interspeech},
  pages={2207--2211},
  doi={10.21437/Interspeech.2018-1456},
  url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
}

or arXiv:

@misc{watanabe2018espnet,
  title={ESPnet: End-to-End Speech Processing Toolkit},
  author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
  year={2018},
  eprint={1804.00015},
  archivePrefix={arXiv},
  primaryClass={cs.CL}
}