Abstract
This paper describes NLIP Lab’s multilingual machine translation system for the WAT24 shared task on multilingual Indic MT task for 22 scheduled languages belonging to 4 language families. We explore pre-training for Indic languages using alignment agreement objectives. We utilize bi-lingual dictionaries to substitute words from source sentences. Furthermore, we fine-tuned language direction-specific multilingual translation models using small and high-quality seed data. Our primary submission is a 243M parameters multilingual translation model covering 22 Indic languages. In the IN22-Gen benchmark, we achieved an average chrF++ score of 46.80 and 18.19 BLEU score for the En-Indic direction. In the Indic-En direction, we achieved an average chrF++ score of 56.34 and 30.82 BLEU score. In the In22-Conv benchmark, we achieved an average chrF++ score of 43.43 and BLEU score of 16.58 in the En-Indic direction, and in the Indic-En direction, we achieved an average of 52.44 and 29.77 for chrF++ and BLEU respectively. Our model is competitive with IndicTransv1 (474M parameter model).- Anthology ID:
- 2024.wmt-1.74
- Volume:
- Proceedings of the Ninth Conference on Machine Translation
- Month:
- November
- Year:
- 2024
- Address:
- Miami, Florida, USA
- Editors:
- Barry Haddow, Tom Kocmi, Philipp Koehn, Christof Monz
- Venue:
- WMT
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 804–809
- Language:
- URL:
- https://preview.aclanthology.org/add_missing_videos/2024.wmt-1.74/
- DOI:
- 10.18653/v1/2024.wmt-1.74
- Cite (ACL):
- Maharaj Brahma, Pramit Sahoo, and Maunendra Sankar Desarkar. 2024. NLIP-Lab-IITH Multilingual MT System for WAT24 MT Shared Task. In Proceedings of the Ninth Conference on Machine Translation, pages 804–809, Miami, Florida, USA. Association for Computational Linguistics.
- Cite (Informal):
- NLIP-Lab-IITH Multilingual MT System for WAT24 MT Shared Task (Brahma et al., WMT 2024)
- PDF:
- https://preview.aclanthology.org/add_missing_videos/2024.wmt-1.74.pdf