We recommend using inference engine from
IndicTrans2 for inference of these models. Please execute the following block of code from the root directory:
1from inference.engine import Model
2
3model = Model(ckpt_dir, model_type="ctranslate2")
4
5sents = [sent1, sent2,...]
6
7# for a batch of sentences
8model.batch_translate(sents, src_lang, tgt_lang)
9
10# for a paragraph
11model.translate_paragraph(text, src_lang, tgt_lang)
1@inproceedings{gumma-etal-2025-towards,
2 title = "Towards Inducing Long-Context Abilities in Multilingual Neural Machine Translation Models",
3 author = "Gumma, Varun and
4 Chitale, Pranjal A and
5 Bali, Kalika",
6 editor = "Chiruzzo, Luis and
7 Ritter, Alan and
8 Wang, Lu",
9 booktitle = "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
10 month = apr,
11 year = "2025",
12 address = "Albuquerque, New Mexico",
13 publisher = "Association for Computational Linguistics",
14 url = "https://aclanthology.org/2025.naacl-long.366/",
15 pages = "7158--7170",
16 ISBN = "979-8-89176-189-6"
17}
