- Neural machine translation (NMT) is an approach to machine translation that uses an artificial neural network to predict the likelihood of a sequence of words1. It models entire sentences in a single integrated model1. NMT utilizes a large neural network and has rapidly become the leading technology behind automated translation services, offering significant improvements over traditional rule-based and statistical machine translation methods2.了解详细信息:✕以上信息汇总借助人工智能技术,若访问原文请查看了解详细信息。
Neural machine translation (NMT) is an approach to machine translation that uses an artificial neural network to predict the likelihood of a sequence of words, typically modeling entire sentences in a single...
en.wikipedia.org/wiki/Neural_machine_translationNeural Machine Translation (NMT) is an approach to machine translation that utilizes a large neural network. It has rapidly become the leading technology behind automated translation services, offering significant...
deepai.org/machine-learning-glossary-and-terms/n… - 其他用户还问了以下问题
- 查看更多前往 Wikipedia 查看全部内容
Neural machine translation - Wikipedia
Neural machine translation (NMT) is an approach to machine translation that uses an artificial neural network to predict the likelihood of a sequence of words, typically modeling entire sentences in a single integrated model. It is the dominant approach today and can produce translations that rival … 展开
In the translation task, a sentence $${\displaystyle \mathbf {x} =x_{1,I}}$$ (consisting of $${\displaystyle I}$$ tokens $${\displaystyle x_{i}}$$) in the source language is to be translated into a sentence 展开
Early approaches
In 1987, Robert B. Allen demonstrated the use of feed-forward neural networks for translating auto-generated English sentences with a limited … 展开Cross-entropy loss
NMT models are usually trained to maximize the likelihood of observing the training data. I.e., for a … 展开• Koehn, Philipp (2020). Neural Machine Translation. Cambridge University Press.
• Stahlberg, Felix (2020). Neural Machine Translation: A Review and Survey. 展开1987Robert B. Allen demonstrated the use of feed-forward neural networks for translating auto-generated English sentences with a limited vocabulary of 31 words into Spanish.1991Lonnie Chrisman built upon Allen's work by training separate recursive auto-associative memory (RAAM) networks for the source and the target language.1997Castaño and Casacuberta employed an Elman's recurrent neural network in another machine translation task with very limited vocabulary and complexity.2013Kalchbrenner & Blunsom using a convolutional neural network (CNN) for encoding the source.2014Cho et al. and Sutskever et al. using a recurrent neural network (RNN) instead.2015Baidu launched the "first large-scale NMT system".2016Google launched its NMT system.2017Gehring et al. combined a CNN encoder with an attention mechanism in 2017, which handled long-range dependencies in the source better than previous approaches and also increased translation speed because a CNN encoder is parallelizable.2017Vaswani et al. introduced the transformer.2020GPT-3 released in 2020 can function as a neural machine translation system.NMT has overcome several challenges that were present in statistical machine translation (SMT):
• NMT's full reliance on continuous representation of tokens overcame sparsity issues caused by rare words or phrases. Models were … 展开As outlined in the history section above, instead of using an NMT system that is trained on parallel text, one can also prompt a generative LLM to translate a text. These models differ from an encoder-decoder NMT system in a number of ways: 展开
CC-BY-SA 许可证中的维基百科文本 Google Neural Machine Translation - Wikipedia
Google Translate - Wikipedia
网页Google Translate is a multilingual neural machine translation service developed by Google to translate text, documents and websites from one language into another. It offers a website interface , a mobile app for …
神经机器翻译 - 维基百科,自由的百科全书
A Gentle Introduction to Neural Machine Translation
网页2019年8月7日 · Learn the challenge and history of machine translation, and how neural networks can learn to translate text end-to-end. This post covers the basics of neural machine translation, its benefits and …
Scaling neural machine translation to 200 languages - Nature
Neural machine translation: A review of methods, resources, and …
Neural machine translation: past, present, and future
A Neural Network for Machine Translation, at …
网页2016年9月27日 · Learn how Google developed and deployed a neural network for machine translation, achieving state-of-the-art quality and speed. The paper describes the system architecture, training techniques, …
Zero-Shot Translation with Google’s Multilingual …
网页2016年11月22日 · Learn how Google Translate uses a single system to translate between multiple languages, enabling zero-shot translation and improving quality. See how the system learns a common representation …