LASER is a library to calculate and use multilingual sentence embeddings.
NEWS
- 2022/07/06 Updated LASER models with support for over 200 languages are now available
- 2022/07/06 Multilingual similarity search (xsim) evaluation pipeline released
- 2022/05/03 Librivox S2S is available: Speech-to-Speech translations automatically mined in Librivox [9]
- 2019/11/08 CCMatrix is available: Mining billions of high-quality parallel sentences on the WEB [8]
- 2019/07/31 Gilles Bodard and Jérémy Rapin provided a Docker environment to use LASER
- 2019/07/11 WikiMatrix is available: bitext extraction for 1620 language pairs in WikiPedia [7]
- 2019/03/18 switch to BSD license
- 2019/02/13 The code to perform bitext mining is now available
CURRENT VERSION:
- We now provide updated LASER models which support over 200 languages. Please see here for more details including how to download the models and perform inference.
According to our experience, the sentence encoder also supports code-switching, i.e. the same sentences can contain words in several different languages.
We have also some evidence that the encoder can generalize to other languages which have not been seen during training, but which are in a language family which is covered by other languages.
A detailed description of how the multilingual sentence embeddings are trained can be found here, together with an experimental evaluation.
- Python >= 3.7
- PyTorch 1.0
- NumPy, tested with 1.15.4
- Cython, needed by Python wrapper of FastBPE, tested with 0.29.6
- Faiss, for fast similarity search and bitext mining
- transliterate 1.10.2 (
pip install transliterate
) - jieba 0.39, Chinese segmenter (
pip install jieba
) - mecab 0.996, Japanese segmenter
- tokenization from the Moses encoder (installed automatically)
- FastBPE, fast C++ implementation of byte-pair encoding (installed automatically)
- Fairseq, sequence modeling toolkit (
pip install fairseq==0.12.1
) - tabulate, pretty-print tabular data (
pip install tabulate
) - pandas, data analysis toolkit (
pip install pandas
) - Sentencepiece, subword tokenization (installed automatically)
- set the environment variable 'LASER' to the root of the installation, e.g.
export LASER="${HOME}/projects/laser"
- download encoders from Amazon s3 by e.g.
bash ./nllb/download_models.sh
- download third party software by
bash ./install_external_tools.sh
- download the data used in the example tasks (see description for each task)
We showcase several applications of multilingual sentence embeddings with code to reproduce our results (in the directory "tasks").
- Cross-lingual document classification using the MLDoc corpus [2,6]
- WikiMatrix Mining 135M Parallel Sentences in 1620 Language Pairs from Wikipedia [7]
- Bitext mining using the BUCC corpus [3,5]
- Cross-lingual NLI using the XNLI corpus [4,5,6]
- Multilingual similarity search [1,6]
- Sentence embedding of text files example how to calculate sentence embeddings for arbitrary text files in any of the supported language.
For all tasks, we use exactly the same multilingual encoder, without any task specific optimization or fine-tuning.
LASER is BSD-licensed, as found in the LICENSE
file in the root directory of this source tree.
The original LASER model was trained on the following languages:
Afrikaans, Albanian, Amharic, Arabic, Armenian, Aymara, Azerbaijani, Basque, Belarusian, Bengali, Berber languages, Bosnian, Breton, Bulgarian, Burmese, Catalan, Central/Kadazan Dusun, Central Khmer, Chavacano, Chinese, Coastal Kadazan, Cornish, Croatian, Czech, Danish, Dutch, Eastern Mari, English, Esperanto, Estonian, Finnish, French, Galician, Georgian, German, Greek, Hausa, Hebrew, Hindi, Hungarian, Icelandic, Ido, Indonesian, Interlingua, Interlingue, Irish, Italian, Japanese, Kabyle, Kazakh, Korean, Kurdish, Latvian, Latin, Lingua Franca Nova, Lithuanian, Low German/Saxon, Macedonian, Malagasy, Malay, Malayalam, Maldivian (Divehi), Marathi, Norwegian (Bokmål), Occitan, Persian (Farsi), Polish, Portuguese, Romanian, Russian, Serbian, Sindhi, Sinhala, Slovak, Slovenian, Somali, Spanish, Swahili, Swedish, Tagalog, Tajik, Tamil, Tatar, Telugu, Thai, Turkish, Uighur, Ukrainian, Urdu, Uzbek, Vietnamese, Wu Chinese and Yue Chinese.
We have also observed that the model seems to generalize well to other (minority) languages or dialects, e.g.
Asturian, Egyptian Arabic, Faroese, Kashubian, North Moluccan Malay, Nynorsk Norwegian, Piedmontese, Sorbian, Swabian, Swiss German or Western Frisian.
Updated LASER models referred to as LASER3 supplement the above list with support for 147 languages. The full list of supported languages can be seen here.
[1] Holger Schwenk and Matthijs Douze, Learning Joint Multilingual Sentence Representations with Neural Machine Translation, ACL workshop on Representation Learning for NLP, 2017
[2] Holger Schwenk and Xian Li, A Corpus for Multilingual Document Classification in Eight Languages, LREC, pages 3548-3551, 2018.
[3] Holger Schwenk, Filtering and Mining Parallel Data in a Joint Multilingual Space ACL, July 2018
[4] Alexis Conneau, Guillaume Lample, Ruty Rinott, Adina Williams, Samuel R. Bowman, Holger Schwenk and Veselin Stoyanov, XNLI: Cross-lingual Sentence Understanding through Inference, EMNLP, 2018.
[5] Mikel Artetxe and Holger Schwenk, Margin-based Parallel Corpus Mining with Multilingual Sentence Embeddings arXiv, Nov 3 2018.
[6] Mikel Artetxe and Holger Schwenk, Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond arXiv, Dec 26 2018.
[7] Holger Schwenk, Vishrav Chaudhary, Shuo Sun, Hongyu Gong and Paco Guzman, WikiMatrix: Mining 135M Parallel Sentences in 1620 Language Pairs from Wikipedia arXiv, July 11 2019.
[8] Holger Schwenk, Guillaume Wenzek, Sergey Edunov, Edouard Grave and Armand Joulin CCMatrix: Mining Billions of High-Quality Parallel Sentences on the WEB
[9] Paul-Ambroise Duquenne, Hongyu Gong, Holger Schwenk, Multimodal and Multilingual Embeddings for Large-Scale Speech Mining,, NeurIPS 2021, pages 15748-15761.
[10] Kevin Heffernan, Onur Celebi, and Holger Schwenk, Bitext Mining Using Distilled Sentence Representations for Low-Resource Languages