AI4BharatIndicBART
IndicBART XX–English
IndicBART-XXEN is a translation fine-tune of IndicBART trained on PMI and PIB data for Indian-language-to-English translation.
A translation fine-tune is separate from the base generator
The card uses MBartForConditionalGeneration but explains why its tokenizer setup differs from standard mBART. It does not independently state a full language list or model-specific license.IndicBART XX–English official model card
Verify the language pair before deployment
Practical guidance. Use a held-out sample from the required source language. Do not infer that every language in a base model has equal coverage in one task fine-tune.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | IndicBART-XXEN |
| Direction | XX → English, as specified in the card.IndicBART XX–English official model card |
| Tokenizer | The example uses AlbertTokenizer with SentencePiece.IndicBART XX–English official model card |
Related models
Multilingual generation model that maps Indian-language text into shared Devanagari script.
Launch date not confirmedDownloadable- IndicTrans2 Indic to English
AI4Bharat · IndicTrans2
Billion-parameter-class IndicTrans2 model for translation from 22 Indian languages into English.
Launch date not confirmedRestricted accessDownloadable