IndicTrans2
Translate between English and the 22 languages listed in India's Constitution, with an open model.
IndicTrans2 translates between English and the 22 languages listed in India's Constitution, and among those languages, including low-resource ones such as Kashmiri, Manipuri and Sindhi, across five scripts. The project describes it as the first open source translation model to cover them all.
And where does AI come in?
A multilingual transformer-based neural machine translation model. AI4Bharat releases everything: the training corpus (BPCC), the back-translation data, the models, the evaluation benchmarks (IN22) and the scripts, to move research on low-resource Indian languages forward.
The AI label
- Model or technique
- Transformer neural machine translation models, from English to Indian languages, the other way round, and between Indian languages.
- Model published
- Not filled in
- Where it runs
- Not filled in
- Data
- BPCC training corpus and back-translation data, released with the models.
- Role of people
- Not filled in
- Frugality
- Not filled in
“Not filled in”: MAUB could not check it with the project. Know more? Write to us.
Run this project? Show the “Listed on MAUB” badge
Paste this code on your site: an image and a link to this page. No script, no tracker: MAUB cannot tell who displays it.
<a href="https://maub.subalpine.it/en/projets/indictrans2"><img src="https://maub.subalpine.it/statique/img/badge-recense-en.svg" alt="Listed on MAUB" width="180" height="48"></a>Something wrong or out of date? Write to contact@subalpine.it.
The monthly newsletter
Once a month, what moved on MAUB, in your inbox.