Recent research has revealed that neural language models at scale suffer from poor temporal generalization capability, i.e., language model pre-trained on static data from past years performs worse over time on emerging data. Existing methods mainly perform continual training to mitigate such a misalignment. While effective to some extent but is far from being addressed on both the language modeling and downstream tasks. In this paper, we empirically observe that temporal generalization is closely affiliated with lexical semantic change, which is one of the essential phenomena of natural languages. Based on this observation, we propose a simple yet effective lexical-level masking strategy to post-train a converged language model. Experiments on two pre-trained language models, two different classification tasks, and four benchmark datasets demonstrate the effectiveness of our proposed method over existing temporal adaptation methods, i.e., continual training with new data. Our code is available at https://github.com/zhaochen0110/LMLM.

Transformer-based autoregressive and non-autoregressive models have played an essential role in sequence generation tasks. The autoregressive model can obtain excellent performance, while the non-autoregressive model brings fast decoding speed for inference. In this paper, we propose JANUS, a Joint Autoregressive and Non-autoregressive training method using aUxiliary losS to enhance the model performance in both AR and NAR manner simultaneously and effectively alleviate the problem of distribution discrepancy.Further, we pre-train BART with JANUS on a large corpus with minimal cost (16 GPU days) and make the BART-JANUS capable of non-autoregressive generation, demonstrating that our approach can transfer the AR knowledge to NAR. Empirically, we show our approach and BART-JANUS can achieve significant improvement on multiple generation tasks, including machine translation and GLGE benchmarks. Our code is available at Github.

Healthcare is becoming a more and more important research topic recently. With the growing data in the healthcare domain, it offers a great opportunity for deep learning to improve the quality of service and reduce costs. However, the complexity of electronic health records (EHR) data is a challenge for the application of deep learning. Specifically, the data produced in the hospital admissions are monitored by the EHR system, which includes structured data like daily body temperature and unstructured data like free text and laboratory measurements. Although there are some preprocessing frameworks proposed for specific EHR data, the clinical notes that contain significant clinical value are beyond the realm of their consideration. Besides, whether these different data from various views are all beneficial to the medical tasks and how to best utilize these data remain unclear. Therefore, in this paper, we first extract the accompanying clinical notes from EHR and propose a method to integrate these data, we also comprehensively study the different models and the data leverage methods for better medical task prediction performance. The results on two prediction tasks show that our fused model with different data outperforms the state-of-the-art method without clinical notes, which illustrates the importance of our fusion method and the clinical note features.

Data augmentation, which refers to manipulating the inputs (e.g., adding random noise,masking specific parts) to enlarge the dataset,has been widely adopted in machine learning. Most data augmentation techniques operate on a single input, which limits the diversity of the training corpus. In this paper, we propose a simple yet effective data augmentation technique for neural machine translation, mixSeq, which operates on multiple inputs and their corresponding targets. Specifically, we randomly select two input sequences,concatenate them together as a longer input aswell as their corresponding target sequencesas an enlarged target, and train models on theaugmented dataset. Experiments on nine machine translation tasks demonstrate that such asimple method boosts the baselines by a non-trivial margin. Our method can be further combined with single input based data augmentation methods to obtain further improvements.

Transformer architecture achieves great success in abundant natural language processing tasks. The over-parameterization of the Transformer model has motivated plenty of works to alleviate its overfitting for superior performances. With some explorations, we find simple techniques such as dropout, can greatly boost model performance with a careful design. Therefore, in this paper, we integrate different dropout techniques into the training of Transformer models. Specifically, we propose an approach named UniDrop to unites three different dropout techniques from fine-grain to coarse-grain, i.e., feature dropout, structure dropout, and data dropout. Theoretically, we demonstrate that these three dropouts play different roles from regularization perspectives. Empirically, we conduct experiments on both neural machine translation and text classification benchmark datasets. Extensive results indicate that Transformer with UniDrop can achieve around 1.5 BLEU improvement on IWSLT14 translation tasks, and better accuracy for the classification even using strong pre-trained RoBERTa as backbone.

We Microsoft Research Asia made submissions to 11 language directions in the WMT19 news translation tasks. We won the first place for 8 of the 11 directions and the second place for the other three. Our basic systems are built on Transformer, back translation and knowledge distillation. We integrate several of our rececent techniques to enhance the baseline systems: multi-agent dual learning (MADL), masked sequence-to-sequence pre-training (MASS), neural architecture optimization (NAO), and soft contextual data augmentation (SCA).

While data augmentation is an important trick to boost the accuracy of deep learning methods in computer vision tasks, its study in natural language tasks is still very limited. In this paper, we present a novel data augmentation method for neural machine translation.Different from previous augmentation methods that randomly drop, swap or replace words with other words in a sentence, we softly augment a randomly chosen word in a sentence by its contextual mixture of multiple related words. More accurately, we replace the one-hot representation of a word by a distribution (provided by a language model) over the vocabulary, i.e., replacing the embedding of this word by a weighted combination of multiple semantically similar words. Since the weights of those words depend on the contextual information of the word to be replaced,the newly generated sentences capture much richer information than previous augmentation methods. Experimental results on both small scale and large scale machine translation data sets demonstrate the superiority of our method over strong baselines.

While very deep neural networks have shown effectiveness for computer vision and text classification applications, how to increase the network depth of the neural machine translation (NMT) models for better translation quality remains a challenging problem. Directly stacking more blocks to the NMT model results in no improvement and even drop in performance. In this work, we propose an effective two-stage approach with three specially designed components to construct deeper NMT models, which result in significant improvements over the strong Transformer baselines on WMT14 English→German and English→French translation tasks.

While target-side monolingual data has been proven to be very useful to improve neural machine translation (briefly, NMT) through back translation, source-side monolingual data is not well investigated. In this work, we study how to use both the source-side and target-side monolingual data for NMT, and propose an effective strategy leveraging both of them. First, we generate synthetic bitext by translating monolingual data from the two domains into the other domain using the models pretrained on genuine bitext. Next, a model is trained on a noised version of the concatenated synthetic bitext where each source sequence is randomly corrupted. Finally, the model is fine-tuned on the genuine bitext and a clean version of a subset of the synthetic bitext without adding any noise. Our approach achieves state-of-the-art results on WMT16, WMT17, WMT18 English↔German translations and WMT19 German→French translations, which demonstrate the effectiveness of our method. We also conduct a comprehensive study on how each part in the pipeline works.

Neural machine translation, which achieves near human-level performance in some languages, strongly relies on the large amounts of parallel sentences, which hinders its applicability to low-resource language pairs. Recent works explore the possibility of unsupervised machine translation with monolingual data only, leading to much lower accuracy compared with the supervised one. Observing that weakly paired bilingual documents are much easier to collect than bilingual sentences, e.g., from Wikipedia, news websites or books, in this paper, we investigate training translation models with weakly paired bilingual documents. Our approach contains two components. 1) We provide a simple approach to mine implicitly bilingual sentence pairs from document pairs which can then be used as supervised training signals. 2) We leverage the topic consistency of two weakly paired documents and learn the sentence translation model by constraining the word distribution-level alignments. We evaluate our method on weakly paired documents from Wikipedia on six tasks, the widely used WMT16 German↔English, WMT13 Spanish↔English and WMT16 Romanian↔English translation tasks. We obtain 24.1/30.3, 28.1/27.6 and 30.1/27.6 BLEU points separately, outperforming previous results by more than 5 BLEU points in each direction and reducing the gap between unsupervised translation and supervised translation up to 50%.

Neural machine translation usually adopts autoregressive models and suffers from exposure bias as well as the consequent error propagation problem. Many previous works have discussed the relationship between error propagation and the accuracy drop (i.e., the left part of the translated sentence is often better than its right part in left-to-right decoding models) problem. In this paper, we conduct a series of analyses to deeply understand this problem and get several interesting findings. (1) The role of error propagation on accuracy drop is overstated in the literature, although it indeed contributes to the accuracy drop problem. (2) Characteristics of a language play a more important role in causing the accuracy drop: the left part of the translation result in a right-branching language (e.g., English) is more likely to be more accurate than its right part, while the right part is more accurate for a left-branching language (e.g., Japanese). Our discoveries are confirmed on different model structures including Transformer and RNN, and in other sequence generation tasks such as text summarization.

Recent studies have shown that reinforcement learning (RL) is an effective approach for improving the performance of neural machine translation (NMT) system. However, due to its instability, successfully RL training is challenging, especially in real-world systems where deep models and large datasets are leveraged. In this paper, taking several large-scale translation tasks as testbeds, we conduct a systematic study on how to train better NMT models using reinforcement learning. We provide a comprehensive comparison of several important factors (e.g., baseline reward, reward shaping) in RL training. Furthermore, to fill in the gap that it remains unclear whether RL is still beneficial when monolingual data is used, we propose a new method to leverage RL to further boost the performance of NMT systems trained with source/target monolingual data. By integrating all our findings, we obtain competitive results on WMT14 English-German, WMT17 English-Chinese, and WMT17 Chinese-English translation tasks, especially setting a state-of-the-art performance on WMT17 Chinese-English translation task.

Recurrent neural networks have achieved state-of-the-art results in many artificial intelligence tasks, such as language modeling, neural machine translation, speech recognition and so on. One of the key factors to these successes is big models. However, training such big models usually takes days or even weeks of time even if using tens of GPU cards. In this paper, we propose an efficient architecture to improve the efficiency of such RNN model training, which adopts the group strategy for recurrent layers, while exploiting the representation rearrangement strategy between layers as well as time steps. To demonstrate the advantages of our models, we conduct experiments on several datasets and tasks. The results show that our architecture achieves comparable or better accuracy comparing with baselines, with a much smaller number of parameters and at a much lower computational cost.