Nematus: a Toolkit for Neural Machine Translation

By Rico Sennrich, Orhan Firat, Kyunghyun Cho, Alexandra Birch, Barry Haddow, Julian Hitschler, Marcin Junczys-Dowmunt, Samuel Läubli, Antonio Valerio Miceli Barone, Jozef Mokry and 1 more. 2017

407 citations recorded by the source corpus.

We present Nematus, a toolkit for Neural Machine Translation. The toolkit prioritizes high translation accuracy, usability, and extensibility. Nematus has been used to build top-performing submissions to shared translation tasks at WMT and IWSLT, and has been used to train systems for production environments.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

A Call for Clarity in Reporting BLEU Scores (2018, background): “However, this script in-conveniently requires the data to be wrapped into XML. Nematus (Sennrich et al., 2017) contains a version ( multi-bleu-detok.perl ) that re-moves the XML requirement.”
Root Mean Square Layer Normalization (2019, methodology): “We perform experiments with RNNSearch using the WeightNorm implementation provided by Theano-version Nematus.”
Improving a Neural Semantic Parser by Counterfactual Learning from Human Bandit Feedback (2018, methodology): “In our experiments we use the sequence-to-sequence neural network package N EMATUS (Sennrich et al., 2017).”
The WebNLG Challenge: Generating Text from RDF Data (2017, methodology): “ADAPTCENTRE makes use of the Nematus (Sennrich et al., 2017) system.”
From Recognition to Cognition: Visual Commonsense Reasoning (2018, methodology): “Additionally, we found that even after training on these corpora, the model would struggle with paraphrases, so we also translated SNLI sentences from English to German and back using the Nematus machine translation system [81, 73].”
Six Challenges for Neural Machine Translation (2017, methodology): “We use the toolkit Nematus 2 (Sennrich et al., 2017) which has been shown to give state-of-the-art results (Sennrich et al., 2016a) at the WMT 2016 evaluation campaign (Bojar et al., 2016).”
Adversarial Example Generation with Syntactically Controlled Paraphrase Networks (2018, methodology): “The pretrained Czech-English model used for translation came from the Nematus NMT system (Sennrich et al., 2017).”
Findings of the 2018 Conference on Machine Translation (WMT18) (2018, methodology): “…size is 10K. 2.4.13 LMU-NMT (Huck et al., 2018) For the WMT18 news translation shared task, LMU Munich (Huck et al., 2018) has trained basic shallow attentional encoder-decoder systems (Bahdanau et al., 2014) with the Nematus toolkit (Sennrich et al., 2017), like last year (Huck et al., 2017a).”
The IIT Bombay English-Hindi Parallel Corpus (2017, methodology): “We used Nematus 5 (Sennrich et al., 2017) for training our NMT systems.”
Marian: Fast Neural Machine Translation in C++ (2018, methodology): “Marian grew out of a C++ re-implementation of Nematus (Sennrich et al., 2017b), and still maintains binary-compatibility for common models.”
Regularization techniques for fine-tuning in neural machine translation (2017, methodology): “(2016a), using Nematus (Sennrich et al., 2017) as the neural machine translation toolkit.”
Probing the Need for Visual Context in Multimodal Machine Translation (2019, methodology): “Our baseline NMT is an attentive model (Bahdanau et al., 2014) with a 2-layer bidirectional GRU encoder (Cho et al., 2014) and a 2-layer conditional GRU decoder (Sennrich et al., 2017).”
Evaluating the morphological competence of Machine Translation Systems (2017, methodology): “• BPE-based NMT: LIMSI NMT (Burlot et al., 2017) is based on NMTPY and UEDIN NMT (Sennrich et al., 2017a) on Nematus.”
Evaluating Discourse Phenomena in Neural Machine Translation (2017, methodology): “All models are sequence-to-sequence models with attention (Bahdanau et al., 2015), implemented in Nematus (Sennrich et al., 2017).”
Stronger Baselines for Trustable Results in Neural Machine Translation (2017, background): “However, Adam’s speed and reputation of generally being “good enough” have made it a popular choice for researchers and NMT toolkit authors 6 (Arthur et al., 2016; Lee et al., 2016; Britz et al., 2017; Sennrich et al., 2017).”
Multilingual Neural Machine Translation with Task-Specific Attention (2018, background): “The decoder states are updated according to where q implements the conditional GRU with attention of Sennrich et al. (2017) and H t is an attention-weighted representation of the encoded source sequence h .”