arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1502.00512cs.CLcs.LG
收藏

缩放循环神经网络语言模型

Scaling Recurrent Neural Network Language Models

Cantab Research(坎塔布研究公司)

Will Williams, Niranjani Prasad, David Mrva, Tom Ash, Tony Robinson

摘要:

研究了循环神经网络语言模型的标度特性。我们讨论如何在GPU上训练非常大的RNN,并解决RNNLM如何在模型大小,训练集大小,计算成本和内存方面扩展的问题。我们的分析表明,尽管训练的成本更高,RNNLM在标准基准测试中获得的困惑比n-gram模型低得多。我们训练了已知最大的RNN,并在ASR任务上实现了18%的相对字出错率增益。我们还在最近发布的十亿字语言建模基准上提出了新的最低困惑,机器翻译的BLEU点增益和单词预测的17%相对命中率增益。

英文摘要:

This paper investigates the scaling properties of Recurrent Neural Network Language Models (RNNLMs). We discuss how to train very large RNNs on GPUs and address the questions of how RNNLMs scale with respect to model size, training-set size, computational costs and memory. Our analysis shows that despite being more costly to train, RNNLMs obtain much lower perplexities on standard benchmarks than n-gram models. We train the largest known RNNs and present relative word error rates gains of 18% on an ASR task. We also present the new lowest perplexities on the recently released billion word language modelling benchmark, 1 BLEU point gain on machine translation and a 17% relative hit rate gain in word prediction.

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。

↑