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 2210.09683cs.CL
收藏

阿里巴巴-翻译中国提交的WMT 2022中国国际汽车工业展览会共享任务

Alibaba-Translate China's Submission for WMT 2022 Metrics Shared Task

DAMO Academy, Alibaba Group(阿里巴巴集团达摩院);University of Macau(澳门大学);University of Science and Technology of China(中国科学技术大学);National University of Singapore(新加坡国立大学)

Yu Wan, Keqin Bao, Dayiheng Liu, Baosong Yang, Derek F. Wong, Lidia S. Chao, Wenqiang Lei, Jun Xie

摘要:

在本报告中,我们提交了我们对WMT 2022年可持续发展共同任务的提交。我们基于UNITE(统一翻译评估)的核心思想构建我们的系统,该思想将仅源代码,仅参考以及源代码-参考组合评估场景统一到一个模型中。具体来说,在模型预训练阶段,我们首先应用伪标记数据示例来连续预训练UNITE。值得注意的是,为了减少预训练和微调之间的差距,我们使用了数据裁剪和基于排名的分数归一化策略。在微调阶段,我们使用过去几年WMT比赛的直接评估(DA)和多维质量指标(MQM)数据。特别地,我们收集来自具有不同预训练的语言模型主干的模型的结果,并且针对所涉及的翻译方向使用不同的集成策略。

英文摘要:

In this report, we present our submission to the WMT 2022 Metrics Shared Task. We build our system based on the core idea of UNITE (Unified Translation Evaluation), which unifies source-only, reference-only, and source-reference-combined evaluation scenarios into one single model. Specifically, during the model pre-training phase, we first apply the pseudo-labeled data examples to continuously pre-train UNITE. Notably, to reduce the gap between pre-training and fine-tuning, we use data cropping and a ranking-based score normalization strategy. During the fine-tuning phase, we use both Direct Assessment (DA) and Multidimensional Quality Metrics (MQM) data from past years' WMT competitions. Specially, we collect the results from models with different pre-trained language model backbones, and use different ensembling strategies for involved translation directions.

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。

↑