Instructions to use trysem/mms-tts-ml-ONNX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers.js
How to use trysem/mms-tts-ml-ONNX with Transformers.js:
// npm i @huggingface/transformers import { pipeline } from '@huggingface/transformers'; // Allocate pipeline const pipe = await pipeline('text-to-audio', 'trysem/mms-tts-ml-ONNX');
| language: | |
| - ml | |
| license: cc-by-nc-4.0 | |
| library_name: transformers.js | |
| pipeline_tag: text-to-audio | |
| tags: | |
| - onnx | |
| - text-to-speech | |
| - tts | |
| - mms | |
| - ml | |
| - onnxruntime-web | |
| base_model: facebook/mms-tts-mal | |
| # MMS-TTS Malayalam — ONNX | |
| Verbatim ONNX export of [`facebook/mms-tts-mal`](https://huggingface.co/facebook/mms-tts-mal) | |
| for in-browser inference via [transformers.js](https://huggingface.co/docs/transformers.js) | |
| or [onnxruntime-web](https://onnxruntime.ai/). | |
| This bundle was produced by the [onnx-conversion](https://github.com/NakliTechie/onnx-conversion) pipeline at | |
| `scripts/seq2seq/03_mms_batch_export.py` so the [Ghoshana](https://github.com/NakliTechie/Ghoshana) | |
| browser app can ship MMS-TTS coverage for Malayalam without depending on | |
| a system voice install. | |
| The MMS family from [Massively Multilingual Speech](https://research.facebook.com/publications/scaling-speech-technology-to-1000-languages/) | |
| covers 1107+ languages but only ~12 of those have community ONNX exports | |
| (Xenova mostly), with `kn` (Kannada) at onnx-community. This repo fills | |
| in the rest of the Malayalam gap. | |
| ## License | |
| CC-BY-NC-4.0 (inherited from the upstream `facebook/mms-tts-mal`). | |