mHuBERT-147: A Compact Multilingual HuBERT Model
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866909397965340672 |
|---|---|
| author | Boito, Marcely Zanon Iyer, Vivek Lagos, Nikolaos Besacier, Laurent Calapodescu, Ioan |
| author_facet | Boito, Marcely Zanon Iyer, Vivek Lagos, Nikolaos Besacier, Laurent Calapodescu, Ioan |
| contents | We present mHuBERT-147, the first general-purpose massively multilingual HuBERT speech representation model trained on 90K hours of clean, open-license data. To scale up the multi-iteration HuBERT approach, we use faiss-based clustering, achieving 5.2x faster label assignment than the original method. We also apply a new multilingual batching up-sampling strategy, leveraging both language and dataset diversity. After 3 training iterations, our compact 95M parameter mHuBERT-147 outperforms larger models trained on substantially more data. We rank second and first on the ML-SUPERB 10min and 1h leaderboards, with SOTA scores for 3 tasks. Across ASR/LID tasks, our model consistently surpasses XLS-R (300M params; 436K hours) and demonstrates strong competitiveness against the much larger MMS (1B params; 491K hours). Our findings indicate that mHuBERT-147 is a promising model for multilingual speech tasks, offering an unprecedented balance between high performance and parameter efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_06371 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | mHuBERT-147: A Compact Multilingual HuBERT Model Boito, Marcely Zanon Iyer, Vivek Lagos, Nikolaos Besacier, Laurent Calapodescu, Ioan Computation and Language Sound Audio and Speech Processing We present mHuBERT-147, the first general-purpose massively multilingual HuBERT speech representation model trained on 90K hours of clean, open-license data. To scale up the multi-iteration HuBERT approach, we use faiss-based clustering, achieving 5.2x faster label assignment than the original method. We also apply a new multilingual batching up-sampling strategy, leveraging both language and dataset diversity. After 3 training iterations, our compact 95M parameter mHuBERT-147 outperforms larger models trained on substantially more data. We rank second and first on the ML-SUPERB 10min and 1h leaderboards, with SOTA scores for 3 tasks. Across ASR/LID tasks, our model consistently surpasses XLS-R (300M params; 436K hours) and demonstrates strong competitiveness against the much larger MMS (1B params; 491K hours). Our findings indicate that mHuBERT-147 is a promising model for multilingual speech tasks, offering an unprecedented balance between high performance and parameter efficiency. |
| title | mHuBERT-147: A Compact Multilingual HuBERT Model |
| topic | Computation and Language Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2406.06371 |