首页> 外文会议>China National Conference on Computational Linguistics >Multi-reward Based Reinforcement Learning for Neural Machine Translation
【24h】

Multi-reward Based Reinforcement Learning for Neural Machine Translation

机译:基于多奖励的神经电脑翻译加固学习

获取原文

摘要

Reinforcement learning (RL) has made remarkable progress in neural machine translation (NMT). However, it exists the problems with uneven sampling distribution, sparse rewards and high variance in training phase. Therefore, we propose a multi-reward reinforcement learning training strategy to decouple action selection and value estimation. Meanwhile, our method combines with language model rewards to jointly optimize model parameters. In addition, we add Gumbel noise in sampling to obtain more effective semantic information. To verify the robustness of our method, we not only conducted experiments on large corpora, but also performed on low-resource languages. Experimental results show that our work is superior to the baselines in WMT14 English-German, LDC2014 Chinese-English and CWMT2018 Mongolian-Chinese tasks, which fully certificates the effectiveness of our method.
机译:强化学习(RL)在神经机翻译(NMT)中取得了显着进展。 然而,它存在不均匀的采样分布,稀疏奖励和训练阶段的高方差存在问题。 因此,我们提出了一种多奖励加强学习培训策略来解耦行动选择和价值估计。 同时,我们的方法与语言模型奖励相结合,共同优化了模型参数。 此外,我们在采样中添加Gumbel噪声以获得更有效的语义信息。 为了验证我们方法的稳健性,我们不仅对大型语料库进行了实验,还对低资源语言进行了实验。 实验结果表明,我们的工作优于WMT14英语 - 德语,LDC2014中英和CWMT2018蒙古 - 中文任务的基线,这完全证明了我们方法的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号