Transformer-based autoregressive (AR) methods have achieved appealing performance for varied sequence-to-sequence generation tasks, e.g., neural machine translation, summarization, and code generation, but suffer from low inference efficiency. To speed up the inference stage, many non-autoregressive (NAR) strategies have been proposed in the past few years. Among them, the conditional masked language model (CMLM) is one of the most versatile frameworks, as it can support many different sequence generation scenarios and achieve very competitive performance on these tasks. In this paper, we further introduce a simple yet effective adaptive masking over masking strategy to enhance the refinement capability of the decoder and make the encoder optimization easier. Experiments on \textbf{3} different tasks (neural machine translation, summarization, and code generation) with \textbf{15} datasets in total confirm that our proposed simple method achieves significant performance improvement over the strong CMLM model. Surprisingly, our proposed model yields state-of-the-art performance on neural machine translation (\textbf{34.62} BLEU on WMT16 EN$\to$RO, \textbf{34.82} BLEU on WMT16 RO$\to$EN, and \textbf{34.84} BLEU on IWSLT De$\to$En) and even better performance than the \textbf{AR} Transformer on \textbf{7} benchmark datasets with at least \textbf{2.2$\times$} speedup. Our code is available at GitHub.
Conditional masked language model (CMLM) training has proven successful for non-autoregressive and semi-autoregressive sequence generation tasks, such as machine translation. Given a trained CMLM, however, it is not clear what the best inference strategy is. We formulate masked inference as a factorization of conditional probabilities of partial sequences, show that this does not harm performance, and investigate a number of simple heuristics motivated by this perspective. We identify a thresholding strategy that has advantages over the standard "mask-predict" algorithm, and provide analyses o…
Masked Latent Transformer with the Random Masking Ratio to Advance the Diagnosis of Dental Fluorosis
State Key Laboratory of Public Big Data, Guizhou University（贵州大学公共大数据国家重点实验室）；College of Computer Science and Technology, Guizhou University（贵州大学计算机科学与技术学院）
Microsoft Dynamics 365 AI（微软Dynamics 365人工智能）；Microsoft Research（微软研究院）
Masked Language Model (MLM) framework has been widely adopted for self-supervised language pre-training. In this paper, we argue that randomly sampled masks in MLM would lead to undesirably large gradient variance. Thus, we theoretically quantify the gradient variance via correlating the gradient covariance with the Hamming distance between two different masks (given a certain text sequence). To reduce the variance due to the sampling of masks, we propose a fully-explored masking strategy, where a text sequence is divided into a certain number of non-overlapping segments. Thereafter, the token…
Korea University（高丽大学）；The Catholic University of Korea（韩国天主教大学）
Masked language modeling (MLM) has been widely used for pre-training effective bidirectional representations, but incurs substantial training costs. In this paper, we propose a novel concept-based curriculum masking (CCM) method to efficiently pre-train a language model. CCM has two key differences from existing curriculum learning approaches to effectively reflect the nature of MLM. First, we introduce a carefully-designed linguistic difficulty criterion that evaluates the MLM difficulty of each token. Second, we construct a curriculum that gradually masks words related to the previously mask…
Masked language models (MLMs) conventionally mask 15% of tokens due to the belief that more masking would leave insufficient context to learn good representations; this masking rate has been widely used, regardless of model sizes or masking strategies. In this work, we revisit this important choice of MLM pre-training. We first establish that 15% is not universally optimal, and larger models should adopt a higher masking rate. Specifically, we find that masking 40% outperforms 15% for BERT-large size models on GLUE and SQuAD. Interestingly, an extremely high masking rate of 80% can still prese…
We show that a simple unsupervised masking objective can approach near supervised performance on abstractive multi-document news summarization. Our method trains a state-of-the-art neural summarization model to predict the masked out source document with highest lexical centrality relative to the multi-document group. In experiments on the Multi-News dataset, our masked training objective yields a system that outperforms past unsupervised methods and, in human evaluation, surpasses the best supervised method without requiring access to any ground-truth summaries. Further, we evaluate how diffe…
Masked language model and autoregressive language model are two types of language models. While pretrained masked language models such as BERT overwhelm the line of natural language understanding (NLU) tasks, autoregressive language models such as GPT are especially capable in natural language generation (NLG). In this paper, we propose a probabilistic masking scheme for the masked language model, which we call probabilistically masked language model (PMLM). We implement a specific PMLM with a uniform prior distribution on the masking ratio named u-PMLM. We prove that u-PMLM is equivalent to a…
Beijing Foreign Studies University（北京外国语大学）；Moutai Institute（茅台学院）；Jinan University（暨南大学）；Microsoft Research Asia（微软亚洲研究院）
Sentiment transfer aims at revising the input text to satisfy a given sentiment polarity while retaining the original semantic content. The nucleus of sentiment transfer lies in precisely separating the sentiment information from the content information. Existing explicit approaches generally identify and mask sentiment tokens simply based on prior linguistic knowledge and manually-defined rules, leading to low generality and undesirable transfer performance. In this paper, we view the positions to be masked as the learnable parameters, and further propose a novel AM-ST model to learn adaptive…
Medical multiple-choice question answering (MCQA) is particularly difficult. Questions may describe patient symptoms and ask for the correct diagnosis, which requires domain knowledge and complex reasoning. Standard language modeling pretraining alone is not sufficient to achieve the best results. \citet{jin2020disease} showed that focusing masked language modeling on disease name prediction when using medical encyclopedic paragraphs as input leads to considerable MCQA accuracy improvement. In this work, we show that (1) fine-tuning on generated MCQA dataset outperforms the masked language mod…
The Hebrew University of Jerusalem（耶路撒冷希伯来大学）；Ben Gurion University（本·古里安大学）
Masked language modeling (MLM) is one of the key sub-tasks in vision-language pretraining. In the cross-modal setting, tokens in the sentence are masked at random, and the model predicts the masked tokens given the image and the text. In this paper, we observe several key disadvantages of MLM in this setting. First, as captions tend to be short, in a third of the sentences no token is sampled. Second, the majority of masked tokens are stop-words and punctuation, leading to under-utilization of the image. We investigate a range of alternative masking strategies specific to the cross-modal setti…
Masked language modeling (MLM) pre-training models such as BERT corrupt the input by replacing some tokens with [MASK] and then train a model to reconstruct the original tokens. This is an effective technique which has led to good results on all NLP benchmarks. We propose to expand upon this idea by masking the positions of some tokens along with the masked input token ids. We follow the same standard approach as BERT masking a percentage of the tokens positions and then predicting their original values using an additional fully connected classifier stage. This approach has shown good performa…
École Normale Supérieure-PSL（巴黎高等师范学院-巴黎文理研究大学）；CNRS（法国国家科学研究中心）；U. Sorbonne Nouvelle（新索邦大学）；University of Pisa（比萨大学）
Word order, an essential property of natural languages, is injected in Transformer-based neural language models using position encoding. However, recent experiments have shown that explicit position encoding is not always useful, since some models without such feature managed to achieve state-of-the art performance on some tasks. To understand better this phenomenon, we examine the effect of removing position encodings on the pre-training objective itself (i.e., masked language modelling), to test whether models can reconstruct position information from co-occurrences alone. We do so by contro…
We study how masking and predicting tokens in an unsupervised fashion can give rise to linguistic structures and downstream performance gains. Recent theories have suggested that pretrained language models acquire useful inductive biases through masks that implicitly act as cloze reductions for downstream tasks. While appealing, we show that the success of the random masking strategy used in practice cannot be explained by such cloze-like masks alone. We construct cloze-like masks using task-specific lexicons for three different classification datasets and show that the majority of pretrained …
Masked Thought: Simply Masking Partial Reasoning Steps Can Improve Mathematical Reasoning Learning of Language Models
Gaoling School of Artificial Intelligence, Renmin University of China（中国人民大学高瓴人工智能学院）；Alibaba Group（阿里巴巴集团）；Computational Bioscience Research Center, KAUST（阿卜杜拉国王科技大学计算生物科学研究中心）
Pre-trained language model (PTM) has been shown to yield powerful text representations for dense passage retrieval task. The Masked Language Modeling (MLM) is a major sub-task of the pre-training process. However, we found that the conventional random masking strategy tend to select a large number of tokens that have limited effect on the passage retrieval task (e,g. stop-words and punctuation). By noticing the term importance weight can provide valuable information for passage retrieval, we hereby propose alternative retrieval oriented masking (dubbed as ROM) strategy where more important tok…
Paraphrase generation is a long-standing problem and serves an essential role in many natural language processing problems. Despite some encouraging results, recent methods either confront the problem of favoring generic utterance or need to retrain the model from scratch for each new dataset. This paper presents a novel approach to paraphrasing sentences, extended from the GPT-2 model. We develop a template masking technique, named first-order masking, to masked out irrelevant words in exemplars utilizing POS taggers. So that, the paraphrasing task is changed to predicting spans in masked tem…
Coarse-grained linguistic information, such as named entities or phrases, facilitates adequately representation learning in pre-training. Previous works mainly focus on extending the objective of BERT's Masked Language Modeling (MLM) from masking individual tokens to contiguous sequences of n tokens. We argue that such contiguously masking method neglects to model the intra-dependencies and inter-relation of coarse-grained linguistic information. As an alternative, we propose ERNIE-Gram, an explicitly n-gram masking method to enhance the integration of coarse-grained information into pre-train…
Indian Institute of Technology Hyderabad（印度理工学院海得拉巴分校）
Text Style Transfer (TST) is performable through approaches such as latent space disentanglement, cycle-consistency losses, prototype editing etc. The prototype editing approach, which is known to be quite successful in TST, involves two key phases a) Masking of source style-associated tokens and b) Reconstruction of this source-style masked sentence conditioned with the target style. We follow a similar transduction method, in which we transpose the more difficult direct source to target TST task to a simpler Style-Masked Language Model (SMLM) Task, wherein, similar to BERT \cite{bert}, the g…
College of Computer Science and Technology（计算机科学与技术学院）；Nanjing University of Aeronautics and Astronautics（南京航空航天大学）；MIIT Key Laboratory of Pattern Analysis and Machine Intelligence（工业和信息化部模式分析与机器智能重点实验室）
The task of Chinese Spelling Check (CSC) is aiming to detect and correct spelling errors that can be found in the text. While manually annotating a high-quality dataset is expensive and time-consuming, thus the scale of the training dataset is usually very small (e.g., SIGHAN15 only contains 2339 samples for training), therefore supervised-learning based models usually suffer the data sparsity limitation and over-fitting issue, especially in the era of big language models. In this paper, we are dedicated to investigating the \textbf{unsupervised} paradigm to address the CSC problem and we prop…
We present an efficient method of utilizing pretrained language models, where we learn selective binary masks for pretrained weights in lieu of modifying them through finetuning. Extensive evaluations of masking BERT and RoBERTa on a series of NLP tasks show that our masking scheme yields performance comparable to finetuning, yet has a much smaller memory footprint when several tasks need to be inferred simultaneously. Through intrinsic evaluations, we show that representations computed by masked language models encode information necessary for solving downstream tasks. Analyzing the loss land…
Recent work has shown the importance of adaptation of broad-coverage contextualised embedding models on the domain of the target task of interest. Current self-supervised adaptation methods are simplistic, as the training signal comes from a small percentage of \emph{randomly} masked-out tokens. In this paper, we show that careful masking strategies can bridge the knowledge gap of masked language models (MLMs) about the domains more effectively by allocating self-supervision where it is needed. Furthermore, we propose an effective training strategy by adversarially masking out those tokens whi…
MedFLIP: Medical Vision-and-Language Self-supervised Fast Pre-Training with Masked Autoencoder
University of Copenhagen（哥本哈根大学）；Tsinghua University（清华大学）；Shanghai Medical Image Insights（上海医学影像洞察）；Harvard University（哈佛大学）；University of Waterloo（滑铁卢大学）
