AI4BharatIndicBART
IndicBART XLSum
IndicBART XLSum fine-tunes the separate-script IndicBART model on the Indian-language portion of XLSum.
Its language coverage is narrower than the base model
The card documents seven languages for this summarization checkpoint. The base IndicBART language list should not be substituted for the fine-tune's published scope. Model-specific license terms are not stated in this card.IndicBART XLSum official model card
Use a summary reference from the same language
Practical guidance. A summary can preserve the topic while changing a specific claim. Review factual retention separately from how well the generated wording matches a reference.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | IndicBART-XLSum |
| Languages | Bengali, Gujarati, Hindi, Marathi, Punjabi, Tamil and Telugu.IndicBART XLSum official model card |
| Scripts | Native scripts are retained.IndicBART XLSum official model card |
| Tokenizer | SentencePiece through the AlbertTokenizer class, rather than MBartTokenizer's expected setup.IndicBART XLSum official model card |
Related models
- IndicBART Separate Scripts
AI4Bharat · IndicBART
IndicBART generation variant that retains each Indian language's own script.
Launch date not confirmedDownloadable Separate-script IndicBART model for sentence summaries in 11 Indian languages.
Launch date not confirmedDownloadable