CUNI at WMT24 General Translation Task: LLMs, (Q)LoRA, CPO and Model Merging

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F24%3A10492861" target="_blank" >RIV/00216208:11320/24:10492861 - isvavai.cz</a>

Výsledek na webu

<a href="https://aclanthology.org/2024.wmt-1.16/" target="_blank" >https://aclanthology.org/2024.wmt-1.16/</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/2024.wmt-1.16" target="_blank" >10.18653/v1/2024.wmt-1.16</a>

Jazyk výsledku

angličtina

Název v původním jazyce

CUNI at WMT24 General Translation Task: LLMs, (Q)LoRA, CPO and Model Merging

Popis výsledku v původním jazyce

This paper presents the contributions of Charles University teams to the WMT24 General Translation task (English to Czech, German and Russian, and Czech to Ukrainian) and the WMT24 Translation into Low-Resource Languages of Spain task. Our most elaborate submission, CUNI-MH for en2cs, is the result of fine-tuning Mistral 7B v0.1 for translation using a three-stage process: Supervised fine-tuning using QLoRA, Contrastive Preference Optimization, and merging of model checkpoints. We also describe the CUNI-GA, CUNI-Transformer and CUNI-DocTransformer submissions, which are based on our systems from the previous year. Our en2ru system CUNI-DS uses a similar first stage as CUNI-MH (QLoRA for en2cs) and follows with transfer learning for en2ru. For en2de (CUNI-NL), we experimented with an LLM-based speech translation system, to translate without the speech input. For the Translation into Low-Resource Languages of Spain task, we performed QLoRA fine-tuning of a large LLM on a small amount of synthetic (backt

Název v anglickém jazyce

CUNI at WMT24 General Translation Task: LLMs, (Q)LoRA, CPO and Model Merging

Popis výsledku anglicky

This paper presents the contributions of Charles University teams to the WMT24 General Translation task (English to Czech, German and Russian, and Czech to Ukrainian) and the WMT24 Translation into Low-Resource Languages of Spain task. Our most elaborate submission, CUNI-MH for en2cs, is the result of fine-tuning Mistral 7B v0.1 for translation using a three-stage process: Supervised fine-tuning using QLoRA, Contrastive Preference Optimization, and merging of model checkpoints. We also describe the CUNI-GA, CUNI-Transformer and CUNI-DocTransformer submissions, which are based on our systems from the previous year. Our en2ru system CUNI-DS uses a similar first stage as CUNI-MH (QLoRA for en2cs) and follows with transfer learning for en2ru. For en2de (CUNI-NL), we experimented with an LLM-based speech translation system, to translate without the speech input. For the Translation into Low-Resource Languages of Spain task, we performed QLoRA fine-tuning of a large LLM on a small amount of synthetic (backt

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the Ninth Conference on Machine Translation

ISBN

979-8-89176-179-7

ISSN

—

e-ISSN

—

Počet stran výsledku

15

Strana od-do

232-246

Název nakladatele

Association for Computational Linguistics

Místo vydání

Kerrville, TX, USA

Místo konání akce

Miami, FL, USA

Datum konání akce

15. 11. 2024

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—