> For the complete documentation index, see [llms.txt](https://grigore-mihaela.gitbook.io/machine-learning/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://grigore-mihaela.gitbook.io/machine-learning/readings-and-other-media/nlp/mono-multilingual.md).

# Mono- / multilingual

* [Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation](https://arxiv.org/abs/2004.09813), by [Nils Reimers](https://arxiv.org/search/cs?searchtype=author\&query=Reimers%2C+N), [Iryna Gurevych](https://arxiv.org/search/cs?searchtype=author\&query=Gurevych%2C+I)\
  \[ paper ]\
  An easy and efficient method to extend existing sentence embedding models to new languages. This allows to create multilingual versions from previously monolingual models.
* [Multilingual Universal Sentence Encoder for Semantic Retrieval](https://arxiv.org/abs/1907.04307), by [Yinfei Yang](https://arxiv.org/search/cs?searchtype=author\&query=Yang%2C+Y), [Daniel Cer](https://arxiv.org/search/cs?searchtype=author\&query=Cer%2C+D), [Amin Ahmad](https://arxiv.org/search/cs?searchtype=author\&query=Ahmad%2C+A), [Mandy Guo](https://arxiv.org/search/cs?searchtype=author\&query=Guo%2C+M), [Jax Law](https://arxiv.org/search/cs?searchtype=author\&query=Law%2C+J), [Noah Constant](https://arxiv.org/search/cs?searchtype=author\&query=Constant%2C+N), [Gustavo Hernandez Abrego](https://arxiv.org/search/cs?searchtype=author\&query=Abrego%2C+G+H), [Steve Yuan](https://arxiv.org/search/cs?searchtype=author\&query=Yuan%2C+S), [Chris Tar](https://arxiv.org/search/cs?searchtype=author\&query=Tar%2C+C), [Yun-Hsuan Sung](https://arxiv.org/search/cs?searchtype=author\&query=Sung%2C+Y), [Brian Strope](https://arxiv.org/search/cs?searchtype=author\&query=Strope%2C+B), [Ray Kurzweil](https://arxiv.org/search/cs?searchtype=author\&query=Kurzweil%2C+R)\
  \[ paper ]\
  When the authors team is from Google and it comprises of 12 people, it's a sign of something worth reading. \
  \- They propose two pre-trained multilingual (16 languages) retrieval models based on Transformer and CNN model architectures.\
  \- They reach SOTA performance on the folowing tasks: semantic retrieval (SR), translation pair bitext retrieval (BR) and retrieval question answering (ReQA). \
  \- They make the models available on TensorFlow Hub.\
  \- They also wrote a[ short blog post](https://ai.googleblog.com/2019/07/multilingual-universal-sentence-encoder.html) about their findngs on Google Blog.