fastText
fastText is a library where Facebook AI Research develops word representations for efficient learning and sentence/text classification. It is written in C++11; it is used with command line tool and Python bindings, offering language recognition models with ready word vectors for 157 languages trained on Wikipedia and Crawl. GitHub repository is archived (read-only) by the owner, with no active development.
fastText
10facebookresearch · Translation & NLP
Common commercial products in this category: Google Translate, DeepL, Microsoft Translator
- You train word vector learning models.
- You get vector representations for non-vocabulary words.
- You can perform text classification operations.
- • Repository is archived (no longer maintained)
- • Last update 2 years ago
- • High number of open issues (554)
- • NeutralHacker News
Facebook AI Research Open Sources fastText
“Jupyter notebook with some numbers on performance and accuracy of fastText: This compares fastText against the word2vec implementation in gensim, on the syntactic/semantic word analogy task. (raw pull request with comments: )”
Verified against README · Oct 26, 2026 · source
My note
Alternative pages
More in: Translation & NLP
- Transformers★ 168,376 · today
- spaCy★ 34,004 · 4 days ago
- rasa★ 21,364 · 3 mo ago
- NLTK★ 14,761 · today
- Stanza★ 7,905 · yesterday
- OpenNMT-py★ 7,020 · 1 yr ago