arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2404.12299cs.CLcs.AIcs.LGcs.SDeess.AS
收藏

大语言模型在远距离语言对中的同声传译语料库构建

Simultaneous Interpretation Corpus Construction by Large Language Models in Distant Language Pair

Nara Institute of Science and Technology(奈良先端科学技术大学院大学)

Yusuke Sakai, Mana Makinae, Hidetaka Kamigaito, Taro Watanabe

AI总结:

本研究提出利用大语言模型将现有语音翻译语料库转换为同声传译风格语料,以微调同声机器翻译模型,在保持质量的同时降低延迟。

摘要:

在同声机器翻译(SiMT)系统中,使用同声传译(SI)语料库进行训练是获得高质量且低延迟系统的有效方法。然而,由于标注者能力的限制,构建这样的语料库非常具有挑战性,因此现有的SI语料库十分有限。为此,我们提出了一种方法,利用大语言模型(LLM-SI-Corpus)将现有的语音翻译语料库转换为口译风格的数据,同时保持原始词序并保留完整的源语言内容。我们证明,在文本到文本和语音到文本的设置下,使用LLM-SI-Corpus对SiMT模型进行微调,可以在保持与离线数据集训练的模型相同质量水平的同时降低延迟。LLM-SI-Corpus可在\url{https://github.com/yusuke1997/LLM-SI-Corpus}获取。

英文摘要:

In Simultaneous Machine Translation (SiMT) systems, training with a simultaneous interpretation (SI) corpus is an effective method for achieving high-quality yet low-latency systems. However, it is very challenging to curate such a corpus due to limitations in the abilities of annotators, and hence, existing SI corpora are limited. Therefore, we propose a method to convert existing speech translation corpora into interpretation-style data, maintaining the original word order and preserving the entire source content using Large Language Models (LLM-SI-Corpus). We demonstrate that fine-tuning SiMT models in text-to-text and speech-to-text settings with the LLM-SI-Corpus reduces latencies while maintaining the same level of quality as the models trained with offline datasets. The LLM-SI-Corpus is available at \url{https://github.com/yusuke1997/LLM-SI-Corpus}.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑