With the integration of the new version of IndicTrans2 model into MinT (T352690) it is possible to translate across 22 Indic languages. Currently MinT is provided as optional in Content Translation for those languages where Google Translate is the default service.
This ticket proposes to make MinT the default service for these languages. This will expose to users an open source service by default, while still allowing users to switch to other services if they prefer them. User feedback and analysis of future re-runs of the machine translation usage report will determine whether the change was useful and whether to revert it until future improvements of MinT or apply it to more languages in the future.
This is the list of languages:
- Assamese (as/asm_Beng)
- Bodo (brx/brx_Deva) No wiki yet
- Dogri (doi/doi_Deva) No wiki yet
- Goan (gom/gom_Deva)
- Gujarati (gu/guj_Gujr)
- Hindi (hi/hin_Deva)
- Kannada (kn/kan_Knda)
- Kashmiri (ks/kas_Arab & kas_Deva)
- Maithili (mai/mai_Deva)
- Malayalam (ml/mal_Mlym)
- Manipuri (mni/mni_Beng & mni_Mtei)
- Nepali (ne/npi_Deva)
- Oriya (or/ory_Orya)
- Panjabi (pa/pan_Guru)
- Sanskrit (sa/san_Deva)
- Tamil (ta/tam_Taml)
- Telugu (te/tel_Telu)
- Urdu (ur/urd_Arab)
Bangla (bn/ben_Beng)Marathi (mr/mar_Deva)Santali (sat/sat_Olck)Sindhi (sd/snd_Arab & snd_Deva)
- Communicate with the communities.
- Assamese
- Bangla
- Bangla Wikipedia community objected to having MinT as default.
- Goan
- Gujarati
- Kannada
- Kashmiri
- Maithili
- Malayalam
- Manipuri
- Marathi
- Community members objected to having MinT, please read this thread for reasons given.
- Hausa
- Greek
- Hindi
- Santali
- The community is of the opinion that NLLB-200 is better, they will prefer to have NLLB-200 model in MinT than IndicTrans2. See discussion for details
- Sindhi
- A community member objected to having MinT as default in this Wiki. Please read this thread for reasons given.
- Tamil
- Telugu
- Urdu
- Nepali
- Oriya
- Panjabi
- Sanskrit
- Yiddish
- Uzbek
- Enable MinT as default MT