<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kyslyi-etal-2025-vuyko">
    <titleInfo>
        <title>Vuyko Mistral: Adapting LLMs for Low-Resource Dialectal Translation</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Roman</namePart>
        <namePart type="family">Kyslyi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yuliia</namePart>
        <namePart type="family">Maksymiuk</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ihor</namePart>
        <namePart type="family">Pysmennyi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the Fourth Ukrainian Natural Language Processing Workshop (UNLP 2025)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Mariana</namePart>
            <namePart type="family">Romanyshyn</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria (online)</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-269-5</identifier>
    </relatedItem>
    <abstract>In this paper we introduce the first effort to adapt large language models (LLMs) to the Ukrainian dialect (in our case Hutsul), a low-resource and morphologically complex dialect spoken in the Carpathian Highlands. We created a parallel corpus of 9852 dialect-to-standard Ukrainian sentence pairs and a dictionary of 7320 dialectal word mappings. We also addressed data shortage by proposing an advanced Retrieval-Augmented Generation (RAG) pipeline to generate synthetic parallel translation pairs, expanding the corpus with 52142 examples. We have fine-tuned multiple open-source LLMs using LoRA and evaluated them on a standard-to-dialect translation task, also comparing with few-shot GPT-4o translation. In the absence of human annotators, we adopt a multi-metric evaluation strategy combining BLEU, chrF++, TER, and LLM-based judgment (GPT-4o). The results show that even small(7B) finetuned models outperform zero-shot baselines such as GPT-4o across both automatic and LLM-evaluated metrics. All data, models, and code are publicly released at: https://github.com/woters/vuyko-hutsul.</abstract>
    <identifier type="citekey">kyslyi-etal-2025-vuyko</identifier>
    <identifier type="doi">10.18653/v1/2025.unlp-1.10</identifier>
    <location>
        <url>https://aclanthology.org/2025.unlp-1.10/</url>
    </location>
    <part>
        <date>2025-07</date>
        <extent unit="page">
            <start>86</start>
            <end>95</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Vuyko Mistral: Adapting LLMs for Low-Resource Dialectal Translation
%A Kyslyi, Roman
%A Maksymiuk, Yuliia
%A Pysmennyi, Ihor
%Y Romanyshyn, Mariana
%S Proceedings of the Fourth Ukrainian Natural Language Processing Workshop (UNLP 2025)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria (online)
%@ 979-8-89176-269-5
%F kyslyi-etal-2025-vuyko
%X In this paper we introduce the first effort to adapt large language models (LLMs) to the Ukrainian dialect (in our case Hutsul), a low-resource and morphologically complex dialect spoken in the Carpathian Highlands. We created a parallel corpus of 9852 dialect-to-standard Ukrainian sentence pairs and a dictionary of 7320 dialectal word mappings. We also addressed data shortage by proposing an advanced Retrieval-Augmented Generation (RAG) pipeline to generate synthetic parallel translation pairs, expanding the corpus with 52142 examples. We have fine-tuned multiple open-source LLMs using LoRA and evaluated them on a standard-to-dialect translation task, also comparing with few-shot GPT-4o translation. In the absence of human annotators, we adopt a multi-metric evaluation strategy combining BLEU, chrF++, TER, and LLM-based judgment (GPT-4o). The results show that even small(7B) finetuned models outperform zero-shot baselines such as GPT-4o across both automatic and LLM-evaluated metrics. All data, models, and code are publicly released at: https://github.com/woters/vuyko-hutsul.
%R 10.18653/v1/2025.unlp-1.10
%U https://aclanthology.org/2025.unlp-1.10/
%U https://doi.org/10.18653/v1/2025.unlp-1.10
%P 86-95

Markdown (Informal)

[Vuyko Mistral: Adapting LLMs for Low-Resource Dialectal Translation](https://aclanthology.org/2025.unlp-1.10/) (Kyslyi et al., UNLP 2025)

• Vuyko Mistral: Adapting LLMs for Low-Resource Dialectal Translation (Kyslyi et al., UNLP 2025)

ACL

• Roman Kyslyi, Yuliia Maksymiuk, and Ihor Pysmennyi. 2025. Vuyko Mistral: Adapting LLMs for Low-Resource Dialectal Translation. In Proceedings of the Fourth Ukrainian Natural Language Processing Workshop (UNLP 2025), pages 86–95, Vienna, Austria (online). Association for Computational Linguistics.