@inproceedings{singh-etal-2025-adapting,
title = "Adapting {I}ndic{T}rans2 for Legal Domain {MT} via {QL}o{RA} Fine-Tuning at {JUST}-{NLP} 2025",
author = "Singh, Akoijam Jenil and
Meetei, Loitongbam Sanayai and
Surajkanta, Yumnam",
editor = "Modi, Ashutosh and
Ghosh, Saptarshi and
Ekbal, Asif and
Goyal, Pawan and
Jain, Sarika and
Joshi, Abhinav and
Mishra, Shivani and
Datta, Debtanu and
Paul, Shounak and
Singh, Kshetrimayum Boynao and
Kumar, Sandeep",
booktitle = "Proceedings of the 1st Workshop on NLP for Empowering Justice (JUST-NLP 2025)",
month = dec,
year = "2025",
address = "Mumbai, India",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.justnlp-main.15/",
pages = "142--147",
ISBN = "979-8-89176-312-8",
abstract = "Machine Translation (MT) in the legal domain presents substantial challenges due to its complex terminology, lengthy statutes, and rigid syntactic structures. The JUST-NLP 2025 Shared Task on Legal Machine Translation was organized to advance research on domain-specific MT systems for legal texts. In this work, we propose a fine-tuned version of the pretrained large language model (LLM) ai4bharat/indictrans2-en-indic-1B, a transformer-based English-to-Indic translation model. Fine-tuning was performed using the parallel corpus provided by the JUST-NLP 2025 Shared Task organizers.Our adapted model demonstrates notable improvements over the baseline system, particularly in handling domain-specific legal terminology and complex syntactic constructions. In automatic evaluation, our system obtained BLEU = 46.67 and chrF = 70.03.In human evaluation, it achieved adequacy = 4.085 and fluency = 4.006. Our approach achieved an AutoRank score of 58.79, highlighting the effectiveness of domain adaptation through fine-tuning for legal machine translation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="singh-etal-2025-adapting">
<titleInfo>
<title>Adapting IndicTrans2 for Legal Domain MT via QLoRA Fine-Tuning at JUST-NLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Akoijam</namePart>
<namePart type="given">Jenil</namePart>
<namePart type="family">Singh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Loitongbam</namePart>
<namePart type="given">Sanayai</namePart>
<namePart type="family">Meetei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yumnam</namePart>
<namePart type="family">Surajkanta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 1st Workshop on NLP for Empowering Justice (JUST-NLP 2025)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ashutosh</namePart>
<namePart type="family">Modi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Saptarshi</namePart>
<namePart type="family">Ghosh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Asif</namePart>
<namePart type="family">Ekbal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pawan</namePart>
<namePart type="family">Goyal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sarika</namePart>
<namePart type="family">Jain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abhinav</namePart>
<namePart type="family">Joshi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shivani</namePart>
<namePart type="family">Mishra</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Debtanu</namePart>
<namePart type="family">Datta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shounak</namePart>
<namePart type="family">Paul</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kshetrimayum</namePart>
<namePart type="given">Boynao</namePart>
<namePart type="family">Singh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sandeep</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Mumbai, India</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-312-8</identifier>
</relatedItem>
<abstract>Machine Translation (MT) in the legal domain presents substantial challenges due to its complex terminology, lengthy statutes, and rigid syntactic structures. The JUST-NLP 2025 Shared Task on Legal Machine Translation was organized to advance research on domain-specific MT systems for legal texts. In this work, we propose a fine-tuned version of the pretrained large language model (LLM) ai4bharat/indictrans2-en-indic-1B, a transformer-based English-to-Indic translation model. Fine-tuning was performed using the parallel corpus provided by the JUST-NLP 2025 Shared Task organizers.Our adapted model demonstrates notable improvements over the baseline system, particularly in handling domain-specific legal terminology and complex syntactic constructions. In automatic evaluation, our system obtained BLEU = 46.67 and chrF = 70.03.In human evaluation, it achieved adequacy = 4.085 and fluency = 4.006. Our approach achieved an AutoRank score of 58.79, highlighting the effectiveness of domain adaptation through fine-tuning for legal machine translation.</abstract>
<identifier type="citekey">singh-etal-2025-adapting</identifier>
<location>
<url>https://aclanthology.org/2025.justnlp-main.15/</url>
</location>
<part>
<date>2025-12</date>
<extent unit="page">
<start>142</start>
<end>147</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Adapting IndicTrans2 for Legal Domain MT via QLoRA Fine-Tuning at JUST-NLP 2025
%A Singh, Akoijam Jenil
%A Meetei, Loitongbam Sanayai
%A Surajkanta, Yumnam
%Y Modi, Ashutosh
%Y Ghosh, Saptarshi
%Y Ekbal, Asif
%Y Goyal, Pawan
%Y Jain, Sarika
%Y Joshi, Abhinav
%Y Mishra, Shivani
%Y Datta, Debtanu
%Y Paul, Shounak
%Y Singh, Kshetrimayum Boynao
%Y Kumar, Sandeep
%S Proceedings of the 1st Workshop on NLP for Empowering Justice (JUST-NLP 2025)
%D 2025
%8 December
%I Association for Computational Linguistics
%C Mumbai, India
%@ 979-8-89176-312-8
%F singh-etal-2025-adapting
%X Machine Translation (MT) in the legal domain presents substantial challenges due to its complex terminology, lengthy statutes, and rigid syntactic structures. The JUST-NLP 2025 Shared Task on Legal Machine Translation was organized to advance research on domain-specific MT systems for legal texts. In this work, we propose a fine-tuned version of the pretrained large language model (LLM) ai4bharat/indictrans2-en-indic-1B, a transformer-based English-to-Indic translation model. Fine-tuning was performed using the parallel corpus provided by the JUST-NLP 2025 Shared Task organizers.Our adapted model demonstrates notable improvements over the baseline system, particularly in handling domain-specific legal terminology and complex syntactic constructions. In automatic evaluation, our system obtained BLEU = 46.67 and chrF = 70.03.In human evaluation, it achieved adequacy = 4.085 and fluency = 4.006. Our approach achieved an AutoRank score of 58.79, highlighting the effectiveness of domain adaptation through fine-tuning for legal machine translation.
%U https://aclanthology.org/2025.justnlp-main.15/
%P 142-147
Markdown (Informal)
[Adapting IndicTrans2 for Legal Domain MT via QLoRA Fine-Tuning at JUST-NLP 2025](https://aclanthology.org/2025.justnlp-main.15/) (Singh et al., JUSTNLP 2025)
ACL