MMS-LID 126 (Core ML, INT8)

Core ML conversion of MMS-LID for 126 languages, INT8 linear quantized for on-device inference on iOS/macOS.

Contents

  • INT8 quantized Core ML model (.mlpackage)
  • Label mapping file(s) for index to language code

Input / Output

  • Input: 16 kHz mono float32 audio, 10 seconds (160,000 samples)
  • Output: Logits over 126 language classes; argmax then look up ISO 639-3 in labels file.

Related

Variant Repo
Float16 mms-lid-126-coreml
4-bit mms-lid-126-coreml-4bit
INT8 this repo

Citation

@article{pratap2023mms,
  title={Scaling Speech Technology to 1,000+ Languages},
  author={Vineel Pratap and Andros Tjandra and Bowen Shi and Paden Tomasello and Arun Babu and Sayani Kundu and Ali Elkahky and Zhaoheng Ni and Apoorv Vyas and Maryam Fazel-Zarandi and Alexei Baevski and Yossi Adi and Xiaohui Zhang and Wei-Ning Hsu and Alexis Conneau and Michael Auli},
  journal={arXiv},
  year={2023}
}

License

CC-BY-NC-4.0 (inherited from MMS-LID).

Downloads last month
9
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support