Saved in:
Bibliographic Details
Main Authors: de Mello, Felipe Ribeiro Fujita, Takada, Hideyuki
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.11388
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912759858331648
author de Mello, Felipe Ribeiro Fujita
Takada, Hideyuki
author_facet de Mello, Felipe Ribeiro Fujita
Takada, Hideyuki
contents We investigated the impact of data selection on machine translation fine-tuning for open LLMs. Using Japanese-English corpora, we compare five selectors: TF-IDF, COMET Kiwi, QuRate, FD-Score, and random selection, under controlled training conditions. We observed that semantic selectors consistently outperform lexical and geometry-based heuristics, and that even when the selected data differ by less than 3%, the impact on model performance is substantial, underscoring the sensitivity of fine-tuning to data quality.
format Preprint
id arxiv_https___arxiv_org_abs_2512_11388
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Improving Translation Quality by Selecting Better Data for LLM Fine-Tuning: A Comparative Analysis
de Mello, Felipe Ribeiro Fujita
Takada, Hideyuki
Computation and Language
We investigated the impact of data selection on machine translation fine-tuning for open LLMs. Using Japanese-English corpora, we compare five selectors: TF-IDF, COMET Kiwi, QuRate, FD-Score, and random selection, under controlled training conditions. We observed that semantic selectors consistently outperform lexical and geometry-based heuristics, and that even when the selected data differ by less than 3%, the impact on model performance is substantial, underscoring the sensitivity of fine-tuning to data quality.
title Improving Translation Quality by Selecting Better Data for LLM Fine-Tuning: A Comparative Analysis
topic Computation and Language
url https://arxiv.org/abs/2512.11388