arXiv cs.CL· SK Deeraj, Yash Ingle, Pruthwik Mishra·· 11 小时前AI 评分27
从 OPUS Books 构建德英平行语料并微调 NLLB-200
Dataset Collection and Fine-Tuning of NLLB for German--English Translation
AI 导读
研究者从 OPUS Books 整理出 2.5 万条高质量德英平行句对，并微调 Meta 的 NLLB-200 模型。微调后的 NLLB-600M 在文学与技术类德英翻译上较基座模型提升 +3.4 BLEU、+2.6 chrF++、降低 3.2 TER，且已集成进开源文档翻译系统 LinguaHaru，支持 PDF 与 Office 文档的格式保留翻译。
正文 · 原文
Abstract:This paper describes the curation of a high-quality parallel corpus from OPUS Books and the fine-tuning of Meta's NLLB-200 models for German--English translation. We detail a dataset extraction pipeline that filters 45k raw OPUS sentence pairs to 25k high-quality examples through encoding validation, length filtering, language identification, deduplication, and quality scoring. We present a reproducible fine-tuning methodology optimized for consumer GPUs using mixed precision, gradient accumulation, and gradient checkpointing. Our fine-tuned NLLB-600M model achieves statistically significant improvements (+3.4 BLEU, +2.6 chrF++, -3.2 TER) over the base model on literary and technical German-English translation. We demonstrate practical deployment by integrating the fine-tuned model into LinguaHaru, an existing open-source document translation system, achieving format-preserving translation across PDF and Office documents. All dataset preparation scripts, training code, and evaluation tools are provided for reproducibility. The fine-tuned model is available at HuggingFace.
| Comments: | 6 pages, 5 tables |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.07797 [cs.CL] |
| (or arXiv:2611.07797v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.07797 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Pruthwik Mishra [view email]
[v1]
Thu, 5 Nov 2026 10:35:22 UTC (31 KB)
来源：arXiv cs.CL · arxiv.org