Language model based pre-trained models such as BERT have provided significant gains across different NLP tasks. In this paper, we study different types of pre-trained transformer based models such as auto-regressive models (GPT-2), auto-encoder models (BERT), and seq2seq models (BART) for conditional data augmentation. We show that prepending the class labels to text sequences provides a simple yet effective way to condition the pre-trained models for data augmentation. On three classification benchmarks, pre-trained Seq2Seq model outperforms other models. Further, we explore how different pre-trained model based data augmentation differs in-terms of data diversity, and how well such methods preserve the class-label information.

8
下载
关闭预览

相关内容

ACM/IEEE第23届模型驱动工程语言和系统国际会议,是模型驱动软件和系统工程的首要会议系列,由ACM-SIGSOFT和IEEE-TCSE支持组织。自1998年以来,模型涵盖了建模的各个方面,从语言和方法到工具和应用程序。模特的参加者来自不同的背景,包括研究人员、学者、工程师和工业专业人士。MODELS 2019是一个论坛,参与者可以围绕建模和模型驱动的软件和系统交流前沿研究成果和创新实践经验。今年的版本将为建模社区提供进一步推进建模基础的机会,并在网络物理系统、嵌入式系统、社会技术系统、云计算、大数据、机器学习、安全、开源等新兴领域提出建模的创新应用以及可持续性。 官网链接:http://www.modelsconference.org/

Although pretrained Transformers such as BERT achieve high accuracy on in-distribution examples, do they generalize to new distributions? We systematically measure out-of-distribution (OOD) generalization for various NLP tasks by constructing a new robustness benchmark with realistic distribution shifts. We measure the generalization of previous models including bag-of-words models, ConvNets, and LSTMs, and we show that pretrained Transformers' performance declines are substantially smaller. Pretrained transformers are also more effective at detecting anomalous or OOD examples, while many previous models are frequently worse than chance. We examine which factors affect robustness, finding that larger models are not necessarily more robust, distillation can be harmful, and more diverse pretraining data can enhance robustness. Finally, we show where future work can improve OOD robustness.

0
5
下载
预览

The notion of "in-domain data" in NLP is often over-simplistic and vague, as textual data varies in many nuanced linguistic aspects such as topic, style or level of formality. In addition, domain labels are many times unavailable, making it challenging to build domain-specific systems. We show that massive pre-trained language models implicitly learn sentence representations that cluster by domains without supervision -- suggesting a simple data-driven definition of domains in textual data. We harness this property and propose domain data selection methods based on such models, which require only a small set of in-domain monolingual data. We evaluate our data selection methods for neural machine translation across five diverse domains, where they outperform an established approach as measured by both BLEU and by precision and recall of sentence selection with respect to an oracle.

0
11
下载
预览

题目: Data Augmentation using Pre-trained Transformer Models

简介:

基于语言模型的预训练模型,如BERT,在不同的NLP任务中提供了显著的收益。在本文中,我们研究了不同类型的基于自回归模型(GPT-2)、自编码器模型(BERT)和seq2seq模型(BART)等用于条件数据增强的预训练变压器模型。我们表明,将类标签前置到文本序列提供了一种简单而有效的方法来设置预训练模型的条件,以便进行数据扩充。在三个分类基准上,预先训练的Seq2Seq模型优于其他模型。此外,我们还探讨了不同的基于预训练模型的数据扩充在数据多样性方面是如何不同的,以及这些方法如何很好地保存类标签信息。

成为VIP会员查看完整内容
0
36

Modern neural network training relies heavily on data augmentation for improved generalization. After the initial success of label-preserving augmentations, there has been a recent surge of interest in label-perturbing approaches, which combine features and labels across training samples to smooth the learned decision surface. In this paper, we propose a new augmentation method that leverages the first and second moments extracted and re-injected by feature normalization. We replace the moments of the learned features of one training image by those of another, and also interpolate the target labels. As our approach is fast, operates entirely in feature space, and mixes different signals than prior methods, one can effectively combine it with existing augmentation methods. We demonstrate its efficacy across benchmark data sets in computer vision, speech, and natural language processing, where it consistently improves the generalization performance of highly competitive baseline networks.

0
10
下载
预览

Bidirectional Encoder Representations from Transformers (BERT) represents the latest incarnation of pretrained language models which have recently advanced a wide range of natural language processing tasks. In this paper, we showcase how BERT can be usefully applied in text summarization and propose a general framework for both extractive and abstractive models. We introduce a novel document-level encoder based on BERT which is able to express the semantics of a document and obtain representations for its sentences. Our extractive model is built on top of this encoder by stacking several inter-sentence Transformer layers. For abstractive summarization, we propose a new fine-tuning schedule which adopts different optimizers for the encoder and the decoder as a means of alleviating the mismatch between the two (the former is pretrained while the latter is not). We also demonstrate that a two-staged fine-tuning approach can further boost the quality of the generated summaries. Experiments on three datasets show that our model achieves state-of-the-art results across the board in both extractive and abstractive settings. Our code is available at https://github.com/nlpyang/PreSumm

0
4
下载
预览

We explore deep autoregressive Transformer models in language modeling for speech recognition. We focus on two aspects. First, we revisit Transformer model configurations specifically for language modeling. We show that well configured Transformer models outperform our baseline models based on the shallow stack of LSTM recurrent neural network layers. We carry out experiments on the open-source LibriSpeech 960hr task, for both 200K vocabulary word-level and 10K byte-pair encoding subword-level language modeling. We apply our word-level models to conventional hybrid speech recognition by lattice rescoring, and the subword-level models to attention based encoder-decoder models by shallow fusion. Second, we show that deep Transformer language models do not require positional encoding. The positional encoding is an essential augmentation for the self-attention mechanism which is invariant to sequence ordering. However, in autoregressive setup, as is the case for language modeling, the amount of information increases along the position dimension, which is a positional signal by its own. The analysis of attention weights shows that deep autoregressive self-attention models can automatically make use of such positional information. We find that removing the positional encoding even slightly improves the performance of these models.

0
5
下载
预览

Despite much success, deep learning generally does not perform well with small labeled training sets. In these scenarios, data augmentation has shown much promise in alleviating the need for more labeled data, but it so far has mostly been applied in supervised settings and achieved limited gains. In this work, we propose to apply data augmentation to unlabeled data in a semi-supervised learning setting. Our method, named Unsupervised Data Augmentation or UDA, encourages the model predictions to be consistent between an unlabeled example and an augmented unlabeled example. Unlike previous methods that use random noise such as Gaussian noise or dropout noise, UDA has a small twist in that it makes use of harder and more realistic noise generated by state-of-the-art data augmentation methods. This small twist leads to substantial improvements on six language tasks and three vision tasks even when the labeled set is extremely small. For example, on the IMDb text classification dataset, with only 20 labeled examples, UDA achieves an error rate of 4.20, outperforming the state-of-the-art model trained on 25,000 labeled examples. On standard semi-supervised learning benchmarks CIFAR-10 and SVHN, UDA outperforms all previous approaches and achieves an error rate of 2.7% on CIFAR-10 with only 4,000 examples and an error rate of 2.85% on SVHN with only 250 examples, nearly matching the performance of models trained on the full sets which are one or two orders of magnitude larger. UDA also works well on large-scale datasets such as ImageNet. When trained with 10% of the labeled set, UDA improves the top-1/top-5 accuracy from 55.1/77.3% to 68.7/88.5%. For the full ImageNet with 1.3M extra unlabeled data, UDA further pushes the performance from 78.3/94.4% to 79.0/94.5%.

0
3
下载
预览

Pre-trained language model representations have been successful in a wide range of language understanding tasks. In this paper, we examine different strategies to integrate pre-trained representations into sequence to sequence models and apply it to neural machine translation and abstractive summarization. We find that pre-trained representations are most effective when added to the encoder network which slows inference by only 14%. Our experiments in machine translation show gains of up to 5.3 BLEU in a simulated resource-poor setup. While returns diminish with more labeled data, we still observe improvements when millions of sentence-pairs are available. Finally, on abstractive summarization we achieve a new state of the art on the full text version of CNN/DailyMail.

0
3
下载
预览

The classification of acoustic environments allows for machines to better understand the auditory world around them. The use of deep learning in order to teach machines to discriminate between different rooms is a new area of research. Similarly to other learning tasks, this task suffers from the high-dimensionality and the limited availability of training data. Data augmentation methods have proven useful in addressing this issue in the tasks of sound event detection and scene classification. This paper proposes a method for data augmentation for the task of room classification from reverberant speech. Generative Adversarial Networks (GANs) are trained that generate artificial data as if they were measured in real rooms. This provides additional training examples to the classifiers without the need for any additional data collection, which is time-consuming and often impractical. A representation of acoustic environments is proposed, which is used to train the GANs. The representation is based on a sparse model for the early reflections, a stochastic model for the reverberant tail and a mixing mechanism between the two. In the experiments shown, the proposed data augmentation method increases the test accuracy of a CNN-RNN room classifier from 89.4% to 95.5%.

0
4
下载
预览

We propose a novel data augmentation method for labeled sentences called conditional BERT contextual augmentation. Data augmentation methods are often applied to prevent overfitting and improve generalization of deep neural network models. Recently proposed contextual augmentation augments labeled sentences by randomly replacing words with more varied substitutions predicted by language model. BERT demonstrates that a deep bidirectional language model is more powerful than either an unidirectional language model or the shallow concatenation of a forward and backward model. We retrofit BERT to conditional BERT by introducing a new conditional masked language model\footnote{The term "conditional masked language model" appeared once in original BERT paper, which indicates context-conditional, is equivalent to term "masked language model". In our paper, "conditional masked language model" indicates we apply extra label-conditional constraint to the "masked language model".} task. The well trained conditional BERT can be applied to enhance contextual augmentation. Experiments on six various different text classification tasks show that our method can be easily applied to both convolutional or recurrent neural networks classifier to obtain obvious improvement.

0
8
下载
预览
小贴士
相关论文
Dan Hendrycks,Xiaoyuan Liu,Eric Wallace,Adam Dziedzic,Rishabh Krishnan,Dawn Song
5+阅读 · 2020年4月13日
Roee Aharoni,Yoav Goldberg
11+阅读 · 2020年4月5日
On Feature Normalization and Data Augmentation
Boyi Li,Felix Wu,Ser-Nam Lim,Serge Belongie,Kilian Q. Weinberger
10+阅读 · 2020年2月25日
Attention Forcing for Sequence-to-sequence Model Training
Qingyun Dou,Yiting Lu,Joshua Efiong,Mark J. F. Gales
6+阅读 · 2019年9月26日
Yang Liu,Mirella Lapata
4+阅读 · 2019年8月22日
Kazuki Irie,Albert Zeyer,Ralf Schlüter,Hermann Ney
5+阅读 · 2019年7月11日
Qizhe Xie,Zihang Dai,Eduard Hovy,Minh-Thang Luong,Quoc V. Le
3+阅读 · 2019年7月10日
Sergey Edunov,Alexei Baevski,Michael Auli
3+阅读 · 2019年4月1日
Data Augmentation of Room Classifiers using Generative Adversarial Networks
Constantinos Papayiannis,Christine Evers,Patrick A. Naylor
4+阅读 · 2019年1月10日
Conditional BERT Contextual Augmentation
Xing Wu,Shangwen Lv,Liangjun Zang,Jizhong Han,Songlin Hu
8+阅读 · 2018年12月17日
相关VIP内容
专知会员服务
41+阅读 · 2020年3月6日
专知会员服务
53+阅读 · 2020年2月3日
ExBert — 可视化分析Transformer学到的表示
专知会员服务
22+阅读 · 2019年10月16日
Top