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 2202.12142cs.CL
收藏

没有词块的预训练:学习数百万个单词的词汇

Pretraining without Wordpieces: Learning Over a Vocabulary of Millions of Words

Zhangyin Feng, Duyu Tang, Cong Zhou, Junwei Liao, Shuangzhi Wu, Xiaocheng Feng, Bing Qin, Yunbo Cao, Shuming Shi

摘要:

标准BERT采用基于子词的标记化,其可以将词分成两个或更多个词段(例如,将“无损”转换为“损失”和“较少”)。这将在以下情况下带来不便:(1)如何获得一个词被分成多个词块的上下文向量?(2)在不知道单词个数的情况下,如何通过完形填空预测单词?在这项工作中,我们探索了在单词词汇表而不是单词片段上开发BERT风格的预训练模型的可能性。我们把这种词级的BERT模型称为词级BERT。我们训练具有不同词汇大小、初始化配置和语言的模型。实验结果表明,与标准的基于词块的BERT相比,WordBERT在完形填空测试和机器阅读理解测试中有显著的提高。在许多其他自然语言理解任务中,包括POS标记,组块和NER,WordBERT始终比BERT表现得更好。模型分析表明,WordBERT相对于BERT的主要优势在于对低频词和生僻词的理解。此外,由于管道是独立于语言的,我们训练WordBERT中文语言,并获得五个自然语言理解数据集的显着收益。最后,通过对推理速度的分析,说明在自然语言理解任务中,WordBERT与BERT具有相当的时间开销。

英文摘要:

The standard BERT adopts subword-based tokenization, which may break a word into two or more wordpieces (e.g., converting "lossless" to "loss" and "less"). This will bring inconvenience in following situations: (1) what is the best way to obtain the contextual vector of a word that is divided into multiple wordpieces? (2) how to predict a word via cloze test without knowing the number of wordpieces in advance? In this work, we explore the possibility of developing BERT-style pretrained model over a vocabulary of words instead of wordpieces. We call such word-level BERT model as WordBERT. We train models with different vocabulary sizes, initialization configurations and languages. Results show that, compared to standard wordpiece-based BERT, WordBERT makes significant improvements on cloze test and machine reading comprehension. On many other natural language understanding tasks, including POS tagging, chunking and NER, WordBERT consistently performs better than BERT. Model analysis indicates that the major advantage of WordBERT over BERT lies in the understanding for low-frequency words and rare words. Furthermore, since the pipeline is language-independent, we train WordBERT for Chinese language and obtain significant gains on five natural language understanding datasets. Lastly, the analyse on inference speed illustrates WordBERT has comparable time cost to BERT in natural language understanding tasks.

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。

↑