paper-with-me

홈 › Papers

On the Sparsity of Neural Machine Translation Models

2020-10-06 · EMNLP 2020 11 · Yong Wang, Longyue Wang, Victor O. K. Li, Zhaopeng Tu

Modern neural machine translation (NMT) models employ a large number of parameters, which leads to serious over-parameterization and typically causes the underutilization of computational resources. In response to this problem, we empirically investigate whether the redundant parameters can be reused to achieve better performance. Experiments and analyses are systematically conducted on different datasets and NMT architectures. We show that: 1) the pruned parameters can be rejuvenated to improve the baseline model by up to +0.8 BLEU points; 2) the rejuvenated parameters are reallocated to enhance the ability of modeling low-level lexical information.

📄 PDF Abstract BibTeX arXiv:2010.02646

Code (0)

등록된 구현이 없습니다.

Tasks

Machine TranslationNMTTranslation

Similar Papers 제목 키워드 기반

Reducing the Impact of Data Sparsity in Statistical Machine Translation

2014-10-01 · WS 2014 10 · Karan Singla, Kunal Sachdeva, Srinivas Bangalore, Dipti Misra Sharma 외
Machine TranslationTranslation

Three-phase training to address data sparsity in Neural Machine Translation

2017-12-01 · WS 2017 12 · Ruchit Agrawal, Mihir Shekhar, Dipti Sharma
Machine TranslationTranslation

Target-Side Augmentation for Document-Level Machine Translation

2023-05-08 · Guangsheng Bao, Zhiyang Teng, Yue Zhang

Document-level machine translation faces the challenge of data sparsity due to its long input length and a small amount of training data, increasing the risk of learning spurious patterns. To address this challenge, we p…

Data AugmentationDocument Level Machine TranslationMachine TranslationTranslation

Semantic Neural Machine Translation using AMR

2019-02-19 · TACL 2019 3 · Linfeng Song, Daniel Gildea, Yue Zhang, Zhiguo Wang 외

It is intuitive that semantic representations can be useful for machine translation, mainly because they can help in enforcing meaning preservation and handling data sparsity (many sentences correspond to one meaning) of…

Abstract Meaning RepresentationMachine TranslationNMTTranslation

Simulated Multiple Reference Training Improves Low-Resource Machine Translation

2020-04-30 · EMNLP 2020 11 · Huda Khayrallah, Brian Thompson, Matt Post, Philipp Koehn

Many valid translations exist for a given sentence, yet machine translation (MT) is trained with a single reference translation, exacerbating data sparsity in low-resource settings. We introduce Simulated Multiple Refere…

Machine TranslationSentenceTranslationvalid