Machine Translation
The First-Ever Multilingual Model To Win WMT, Beating Out Bilingual Models
Building a universal translation system to help everyone access information and better connect with one another is the ultimate goal of the machine translation (MT) field. But the MT field needs to solve fundamental limitations in order to make that future a reality. Most MT systems today use groups of bilingual models, which typically require …
Tools For Language Access During COVID-19
Translation services make it easier to communicate with someone who doesn’t speak the same language, whether you’re traveling abroad or living in a new country. But in the context of a global pandemic, government and health officials urgently need to deliver vital information to their communities, and every member of the community needs access to …
CCMatrix: A Billion-scale Bitext Data Set For Training Translation Models
CCMatrix is the largest data set of high-quality, web-based bitexts for training translation models. With more than 4.5 billion parallel sentences in 576 language pairs pulled from snapshots of the CommonCrawl public data set, CCMatrix is more than 50 times larger than the WikiMatrix corpus that we shared last year. Gathering a data set of this size …
Shitloads And Zingers: On The Perils Of Machine Translation
Years ago, on a flight from Amsterdam to Boston, two American nuns seated to my right listened to a voluble young Dutchman who was out to discover the United States. He asked the nuns where they were from. Alas, Framingham, Massachusetts was not on his itinerary, but, he noted, he had ‘shitloads of time and …
PyCon 2019 | Applied Deep Learning For NLP Using PyTorch
PyCon 2019 | Applied Deep Learning For NLP Using PyTorch Speaker: Elvis Saravia Natural language processing (NLP) has experienced a rapid growth over the last few years and has become an important skill to build applications that range from social features to clinical and health solutions. In this tutorial, we will introduce PyTorch as …