PyTorchKR Papers

최신 논문

총 5,206편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2006.04050

Growing Together: Modeling Human Language Learning With n-Best Multi-Checkpoint Machine Translation

We describe our submission to the 2020 Duolingo Shared Task on Simultaneous Translation And Paraphrase for Language Education (STAPLE) (Mayhew et al., 2020). We view MT models at various training stages (i.e., checkpoints) as human learners at different levels. Hence, we employ an ensemble of multi-checkpoints from the same model to generate translation sequences with various levels of fluency. From each checkpoint, for our best model, we sample n-Best sequences (n=10) with a beam width =100. We achieve 37.57 macro F1 with a 6 checkpoint model ensemble on the official English to Portuguese shared task test data, outperforming a baseline Amazon translation system of 21.30 macro F1 and ultimately demonstrating the utility of our intuitive method.

arXiv:2006.04229공식 코드 · Papers with Code

Pre-training Polish Transformer-based Language Models at Scale

Transformer-based language models are now widely used in Natural Language Processing (NLP). This statement is especially true for English language, in which many pre-trained models utilizing transformer-based architecture have been published in recent years. This has driven forward the state of the art for a variety of standard NLP tasks such as classification, regression, and sequence labeling, as well as text-to-text tasks, such as machine translation, question answering, or summarization. The situation have been different for low-resource languages, such as Polish, however. Although some transformer-based language models for Polish are available, none of them have come close to the scale, in terms of corpus size and the number of parameters, of the largest English-language models. In this study, we present two language models for Polish based on the popular BERT architecture. The larger model was trained on a dataset consisting of over 1 billion polish sentences, or 135GB of raw text. We describe our methodology for collecting the data, preparing the corpus, and pre-training the model. We then evaluate our models on thirteen Polish linguistic tasks, and demonstrate improvements over previous approaches in eleven of them.

arXiv:2006.04781

What's the Difference Between Professional Human and Machine Translation? A Blind Multi-language Study on Domain-specific MT

Machine translation (MT) has been shown to produce a number of errors that require human post-editing, but the extent to which professional human translation (HT) contains such errors has not yet been compared to MT. We compile pre-translated documents in which MT and HT are interleaved, and ask professional translators to flag errors and post-edit these documents in a blind evaluation. We find that the post-editing effort for MT segments is only higher in two out of three language pairs, and that the number of segments with wrong terminology, omissions, and typographical problems is similar in HT.

arXiv:2006.04721

Modeling Discourse Structure for Document-level Neural Machine Translation

Recently, document-level neural machine translation (NMT) has become a hot topic in the community of machine translation. Despite its success, most of existing studies ignored the discourse structure information of the input document to be translated, which has shown effective in other tasks. In this paper, we propose to improve document-level NMT with the aid of discourse structure information. Our encoder is based on a hierarchical attention network (HAN). Specifically, we first parse the input document to obtain its discourse structure. Then, we introduce a Transformer-based path encoder to embed the discourse structure information of each word. Finally, we combine the discourse structure information with the word embedding before it is fed into the encoder. Experimental results on the English-to-German dataset show that our model can significantly outperform both Transformer and Transformer+HAN.

arXiv:2006.04702공식 코드 · Papers with Code

CycleGT: Unsupervised Graph-to-Text and Text-to-Graph Generation via Cycle Training

Two important tasks at the intersection of knowledge graphs and natural language processing are graph-to-text (G2T) and text-to-graph (T2G) conversion. Due to the difficulty and high cost of data collection, the supervised data available in the two fields are usually on the magnitude of tens of thousands, for example, 18K in the WebNLG~2017 dataset after preprocessing, which is far fewer than the millions of data for other tasks such as machine translation. Consequently, deep learning models for G2T and T2G suffer largely from scarce training data. We present CycleGT, an unsupervised training method that can bootstrap from fully non-parallel graph and text data, and iteratively back translate between the two forms. Experiments on WebNLG datasets show that our unsupervised model trained on the same number of data achieves performance on par with several fully supervised models. Further experiments on the non-parallel GenWiki dataset verify that our method performs the best among unsupervised baselines. This validates our framework as an effective approach to overcome the data scarcity problem in the fields of G2T and T2G. Our code is available at https://github.com/QipengGuo/CycleGT.

arXiv:2006.03158공식 코드 · Papers with Code

MLE-guided parameter search for task loss minimization in neural sequence modeling

Neural autoregressive sequence models are used to generate sequences in a variety of natural language processing (NLP) tasks, where they are evaluated according to sequence-level task losses. These models are typically trained with maximum likelihood estimation, which ignores the task loss, yet empirically performs well as a surrogate objective. Typical approaches to directly optimizing the task loss such as policy gradient and minimum risk training are based around sampling in the sequence space to obtain candidate update directions that are scored based on the loss of a single sequence. In this paper, we develop an alternative method based on random search in the parameter space that leverages access to the maximum likelihood gradient. We propose maximum likelihood guided parameter search (MGS), which samples from a distribution over update directions that is a mixture of random search around the current parameters and around the maximum likelihood gradient, with each direction weighted by its improvement in the task loss. MGS shifts sampling to the parameter space, and scores candidates using losses that are pooled from multiple sequences. Our experiments show that MGS is capable of optimizing sequence-level losses, with substantial reductions in repetition and non-termination in sequence completion, and similar improvements to those of minimum risk training in machine translation.

arXiv:2006.03511공식 코드 · Papers with Code

Unsupervised Translation of Programming Languages

A transcompiler, also known as source-to-source translator, is a system that converts source code from a high-level programming language (such as C++ or Python) to another. Transcompilers are primarily used for interoperability, and to port codebases written in an obsolete or deprecated language (e.g. COBOL, Python 2) to a modern one. They typically rely on handcrafted rewrite rules, applied to the source code abstract syntax tree. Unfortunately, the resulting translations often lack readability, fail to respect the target language conventions, and require manual modifications in order to work properly. The overall translation process is timeconsuming and requires expertise in both the source and target languages, making code-translation projects expensive. Although neural models significantly outperform their rule-based counterparts in the context of natural language translation, their applications to transcompilation have been limited due to the scarcity of parallel data in this domain. In this paper, we propose to leverage recent approaches in unsupervised machine translation to train a fully unsupervised neural transcompiler. We train our model on source code from open source GitHub projects, and show that it can translate functions between C++, Java, and Python with high accuracy. Our method relies exclusively on monolingual source code, requires no expertise in the source or target languages, and can easily be generalized to other programming languages. We also build and release a test set composed of 852 parallel functions, along with unit tests to check the correctness of translations. We show that our model outperforms rule-based commercial baselines by a significant margin.

arXiv:2006.02635

M3P: Learning Universal Representations via Multitask Multilingual Multimodal Pre-training

We present M3P, a Multitask Multilingual Multimodal Pre-trained model that combines multilingual pre-training and multimodal pre-training into a unified framework via multitask pre-training. Our goal is to learn universal representations that can map objects occurred in different modalities or texts expressed in different languages into a common semantic space. In addition, to explicitly encourage fine-grained alignment between images and non-English languages, we also propose Multimodal Code-switched Training (MCT) to combine monolingual pre-training and multimodal pre-training via a code-switch strategy. Experiments are performed on the multilingual image retrieval task across two benchmark datasets, including MSCOCO and Multi30K. M3P can achieve comparable results for English and new state-of-the-art results for non-English languages.

arXiv:2006.02876

Enhanced back-translation for low resource neural machine translation using self-training

Improving neural machine translation (NMT) models using the back-translations of the monolingual target data (synthetic parallel data) is currently the state-of-the-art approach for training improved translation systems. The quality of the backward system - which is trained on the available parallel data and used for the back-translation - has been shown in many studies to affect the performance of the final NMT model. In low resource conditions, the available parallel data is usually not enough to train a backward model that can produce the qualitative synthetic data needed to train a standard translation model. This work proposes a self-training strategy where the output of the backward model is used to improve the model itself through the forward translation technique. The technique was shown to improve baseline low resource IWSLT'14 English-German and IWSLT'15 English-Vietnamese backward translation models by 11.06 and 1.5 BLEUs respectively. The synthetic data generated by the improved English-German backward model was used to train a forward model which out-performed another forward model trained using standard back-translation by 2.7 BLEU.

arXiv:2006.02163공식 코드 · Papers with Code

Cross-model Back-translated Distillation for Unsupervised Machine Translation

Recent unsupervised machine translation (UMT) systems usually employ three main principles: initialization, language modeling and iterative back-translation, though they may apply them differently. Crucially, iterative back-translation and denoising auto-encoding for language modeling provide data diversity to train the UMT systems. However, the gains from these diversification processes has seemed to plateau. We introduce a novel component to the standard UMT framework called Cross-model Back-translated Distillation (CBD), that is aimed to induce another level of data diversification that existing principles lack. CBD is applicable to all previous UMT approaches. In our experiments, CBD achieves the state of the art in the WMT'14 English-French, WMT'16 English-German and English-Romanian bilingual unsupervised translation tasks, with 38.2, 30.1, and 36.3 BLEU respectively. It also yields 1.5-3.3 BLEU improvements in IWSLT English-French and English-German tasks. Through extensive experimental analyses, we show that CBD is effective because it embraces data diversity while other similar variants do not.

arXiv:2006.02014공식 코드 · Papers with Code

Norm-Based Curriculum Learning for Neural Machine Translation

A neural machine translation (NMT) system is expensive to train, especially with high-resource settings. As the NMT architectures become deeper and wider, this issue gets worse and worse. In this paper, we aim to improve the efficiency of training an NMT by introducing a novel norm-based curriculum learning method. We use the norm (aka length or module) of a word embedding as a measure of 1) the difficulty of the sentence, 2) the competence of the model, and 3) the weight of the sentence. The norm-based sentence difficulty takes the advantages of both linguistically motivated and model-based sentence difficulties. It is easy to determine and contains learning-dependent features. The norm-based model competence makes NMT learn the curriculum in a fully automated way, while the norm-based sentence weight further enhances the learning of the vector representation of the NMT. Experimental results for the WMT'14 English-German and WMT'17 Chinese-English translation tasks demonstrate that the proposed method outperforms strong baselines in terms of BLEU score (+1.17/+1.56) and training speedup (2.22x/3.33x).

arXiv:2006.01594

Training Multilingual Machine Translation by Alternately Freezing Language-Specific Encoders-Decoders

We propose a modular architecture of language-specific encoder-decoders that constitutes a multilingual machine translation system that can be incrementally extended to new languages without the need for retraining the existing system when adding new languages. Differently from previous works, we simultaneously train $N$ languages in all translation directions by alternately freezing encoder or decoder modules, which indirectly forces the system to train in a common intermediate representation for all languages. Experimental results from multilingual machine translation show that we can successfully train this modular architecture improving on the initial languages while falling slightly behind when adding new languages or doing zero-shot translation. Additional comparison of the quality of sentence representation in the task of natural language inference shows that the alternately freezing training is also beneficial in this direction.

arXiv:2006.01112공식 코드 · Papers with Code

Cascaded Text Generation with Markov Transformers

The two dominant approaches to neural text generation are fully autoregressive models, using serial beam search decoding, and non-autoregressive models, using parallel decoding with no output dependencies. This work proposes an autoregressive model with sub-linear parallel time generation. Noting that conditional random fields with bounded context can be decoded in parallel, we propose an efficient cascaded decoding approach for generating high-quality output. To parameterize this cascade, we introduce a Markov transformer, a variant of the popular fully autoregressive model that allows us to simultaneously decode with specific autoregressive context cutoffs. This approach requires only a small modification from standard autoregressive training, while showing competitive accuracy/speed tradeoff compared to existing methods on five machine translation datasets.

arXiv:2006.01080

Is 42 the Answer to Everything in Subtitling-oriented Speech Translation?

Subtitling is becoming increasingly important for disseminating information, given the enormous amounts of audiovisual content becoming available daily. Although Neural Machine Translation (NMT) can speed up the process of translating audiovisual content, large manual effort is still required for transcribing the source language, and for spotting and segmenting the text into proper subtitles. Creating proper subtitles in terms of timing and segmentation highly depends on information present in the audio (utterance duration, natural pauses). In this work, we explore two methods for applying Speech Translation (ST) to subtitling: a) a direct end-to-end and b) a classical cascade approach. We discuss the benefit of having access to the source language speech for improving the conformity of the generated subtitles to the spatial and temporal subtitling constraints and show that length is not the answer to everything in the case of subtitling-oriented ST.

arXiv:2006.00814공식 코드 · Papers with Code

Online Versus Offline NMT Quality: An In-depth Analysis on English-German and German-English

We conduct in this work an evaluation study comparing offline and online neural machine translation architectures. Two sequence-to-sequence models: convolutional Pervasive Attention (Elbayad et al. 2018) and attention-based Transformer (Vaswani et al. 2017) are considered. We investigate, for both architectures, the impact of online decoding constraints on the translation quality through a carefully designed human evaluation on English-German and German-English language pairs, the latter being particularly sensitive to latency constraints. The evaluation results allow us to identify the strengths and shortcomings of each model when we shift to the online setup.

arXiv:2006.00262

Data Augmentation with Unsupervised Machine Translation Improves the Structural Similarity of Cross-lingual Word Embeddings

Unsupervised cross-lingual word embedding (CLWE) methods learn a linear transformation matrix that maps two monolingual embedding spaces that are separately trained with monolingual corpora. This method relies on the assumption that the two embedding spaces are structurally similar, which does not necessarily hold true in general. In this paper, we argue that using a pseudo-parallel corpus generated by an unsupervised machine translation model facilitates the structural similarity of the two embedding spaces and improves the quality of CLWEs in the unsupervised mapping method. We show that our approach outperforms other alternative approaches given the same amount of data, and, through detailed analysis, we show that data augmentation with the pseudo data from unsupervised machine translation is especially effective for mapping-based CLWEs because (1) the pseudo data makes the source and target corpora (partially) parallel; (2) the pseudo data contains information on the original language that helps to learn similar embedding spaces between the source and target languages.

arXiv:2006.00249

Dynamic Masking for Improved Stability in Spoken Language Translation

For spoken language translation (SLT) in live scenarios such as conferences, lectures and meetings, it is desirable to show the translation to the user as quickly as possible, avoiding an annoying lag between speaker and translated captions. In other words, we would like low-latency, online SLT. If we assume a pipeline of automatic speech recognition (ASR) and machine translation (MT) then a viable approach to online SLT is to pair an online ASR system, with a a retranslation strategy, where the MT system re-translates every update received from ASR. However this can result in annoying "flicker" as the MT system updates its translation. A possible solution is to add a fixed delay, or "mask" to the the output of the MT system, but a fixed global mask introduces undesirable latency to the output. We show how this mask can be set dynamically, improving the latency-flicker trade-off without sacrificing translation quality.

arXiv:2006.00114

Design and Implementation of a Virtual 3D Educational Environment to improve Deaf Education

Advances in NLP, knowledge representation and computer graphic technologies can provide us insights into the development of educational tool for Deaf people. Actual education materials and tools for deaf pupils present several problems, since textbooks are designed to support normal students in the classroom and most of them are not suitable for people with hearing disabilities. Virtual Reality (VR) technologies appear to be a good tool and a promising framework in the education of pupils with hearing disabilities. In this paper, we present a current research tasks surrounding the design and implementation of a virtual 3D educational environment based on X3D and H-Anim standards. The system generates and animates automatically Sign language sentence from a semantic representation that encode the whole meaning of the Arabic input text. Some aspects and issues in Sign language generation will be discussed, including the model of Sign representation that facilitate reuse and reduces the time of Sign generation, conversion of semantic components to sign features representation with regard to Sign language linguistics characteristics and how to generate realistic smooth gestural sequences using X3D content to performs transition between signs for natural-looking of animated avatar. Sign language sentences were evaluated by Algerian native Deaf people. The goal of the project is the development of a machine translation system from Arabic to Algerian Sign Language that can be used as educational tool for Deaf children in algerian primary schools.

arXiv:2005.14489

Neural Simultaneous Speech Translation Using Alignment-Based Chunking

In simultaneous machine translation, the objective is to determine when to produce a partial translation given a continuous stream of source words, with a trade-off between latency and quality. We propose a neural machine translation (NMT) model that makes dynamic decisions when to continue feeding on input or generate output words. The model is composed of two main components: one to dynamically decide on ending a source chunk, and another that translates the consumed chunk. We train the components jointly and in a manner consistent with the inference conditions. To generate chunked training data, we propose a method that utilizes word alignment while also preserving enough context. We compare models with bidirectional and unidirectional encoders of different depths, both on real speech and text input. Our results on the IWSLT 2020 English-to-German task outperform a wait-k baseline by 2.6 to 3.7% BLEU absolute.

arXiv:2005.13681공식 코드 · Papers with Code

Phone Features Improve Speech Translation

End-to-end models for speech translation (ST) more tightly couple speech recognition (ASR) and machine translation (MT) than a traditional cascade of separate ASR and MT models, with simpler model architectures and the potential for reduced error propagation. Their performance is often assumed to be superior, though in many conditions this is not yet the case. We compare cascaded and end-to-end models across high, medium, and low-resource conditions, and show that cascades remain stronger baselines. Further, we introduce two methods to incorporate phone features into ST models. We show that these features improve both architectures, closing the gap between end-to-end models and cascades, and outperforming previous academic work -- by up to 9 BLEU on our low-resource setting.