@inproceedings{singh-etal-2025-instruction,
title = "Instruction-Tuned {E}nglish to {B}hojpuri Neural Machine Translation Using Contrastive Preference Optimization",
author = "Singh, Kshetrimayum Boynao and
Kumar, Deepak and
Ekbal, Asif",
editor = "Haddow, Barry and
Kocmi, Tom and
Koehn, Philipp and
Monz, Christof",
booktitle = "Proceedings of the Tenth Conference on Machine Translation",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.wmt-1.38/",
doi = "10.18653/v1/2025.wmt-1.38",
pages = "638--643",
ISBN = "979-8-89176-341-8",
abstract = "This paper presents an English to Bhojpuri machine translation (MT) system developed for the WMT25 General MT Shared Task. Given the low-resource nature of Bhojpuri, we adopt a two-stage training pipeline: unsupervised pretraining followed by supervised fine-tuning. During pretraining, we use a 300,000-sentence corpus comprising 70{\%} Bhojpuri monolingual data and 30{\%} English data to establish language grounding. The fine-tuning stage utilizes 29,749 bilingual English to Bhojpuri sentence pairs (including training, validation, and test sets). To adapt the system to instruction-following scenarios, we apply a novel optimization strategy: Contrastive Preference Optimization (CPO). This technique enables the model to capture fine-grained translation preferences and maintain semantic fidelity in instruction-tuned settings. We evaluate our system across multiple metrics, demonstrating moderate performance in low-resource MT tasks, particularly in diverse domains such as literary, news, social, and speech."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="singh-etal-2025-instruction">
<titleInfo>
<title>Instruction-Tuned English to Bhojpuri Neural Machine Translation Using Contrastive Preference Optimization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kshetrimayum</namePart>
<namePart type="given">Boynao</namePart>
<namePart type="family">Singh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Deepak</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Asif</namePart>
<namePart type="family">Ekbal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Tenth Conference on Machine Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Barry</namePart>
<namePart type="family">Haddow</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tom</namePart>
<namePart type="family">Kocmi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Philipp</namePart>
<namePart type="family">Koehn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Christof</namePart>
<namePart type="family">Monz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-341-8</identifier>
</relatedItem>
<abstract>This paper presents an English to Bhojpuri machine translation (MT) system developed for the WMT25 General MT Shared Task. Given the low-resource nature of Bhojpuri, we adopt a two-stage training pipeline: unsupervised pretraining followed by supervised fine-tuning. During pretraining, we use a 300,000-sentence corpus comprising 70% Bhojpuri monolingual data and 30% English data to establish language grounding. The fine-tuning stage utilizes 29,749 bilingual English to Bhojpuri sentence pairs (including training, validation, and test sets). To adapt the system to instruction-following scenarios, we apply a novel optimization strategy: Contrastive Preference Optimization (CPO). This technique enables the model to capture fine-grained translation preferences and maintain semantic fidelity in instruction-tuned settings. We evaluate our system across multiple metrics, demonstrating moderate performance in low-resource MT tasks, particularly in diverse domains such as literary, news, social, and speech.</abstract>
<identifier type="citekey">singh-etal-2025-instruction</identifier>
<identifier type="doi">10.18653/v1/2025.wmt-1.38</identifier>
<location>
<url>https://aclanthology.org/2025.wmt-1.38/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>638</start>
<end>643</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Instruction-Tuned English to Bhojpuri Neural Machine Translation Using Contrastive Preference Optimization
%A Singh, Kshetrimayum Boynao
%A Kumar, Deepak
%A Ekbal, Asif
%Y Haddow, Barry
%Y Kocmi, Tom
%Y Koehn, Philipp
%Y Monz, Christof
%S Proceedings of the Tenth Conference on Machine Translation
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-341-8
%F singh-etal-2025-instruction
%X This paper presents an English to Bhojpuri machine translation (MT) system developed for the WMT25 General MT Shared Task. Given the low-resource nature of Bhojpuri, we adopt a two-stage training pipeline: unsupervised pretraining followed by supervised fine-tuning. During pretraining, we use a 300,000-sentence corpus comprising 70% Bhojpuri monolingual data and 30% English data to establish language grounding. The fine-tuning stage utilizes 29,749 bilingual English to Bhojpuri sentence pairs (including training, validation, and test sets). To adapt the system to instruction-following scenarios, we apply a novel optimization strategy: Contrastive Preference Optimization (CPO). This technique enables the model to capture fine-grained translation preferences and maintain semantic fidelity in instruction-tuned settings. We evaluate our system across multiple metrics, demonstrating moderate performance in low-resource MT tasks, particularly in diverse domains such as literary, news, social, and speech.
%R 10.18653/v1/2025.wmt-1.38
%U https://aclanthology.org/2025.wmt-1.38/
%U https://doi.org/10.18653/v1/2025.wmt-1.38
%P 638-643
Markdown (Informal)
[Instruction-Tuned English to Bhojpuri Neural Machine Translation Using Contrastive Preference Optimization](https://aclanthology.org/2025.wmt-1.38/) (Singh et al., WMT 2025)
ACL