OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Fasttext Multilingual

Fasttext Multilingual is a public GitHub repository containing 78 transformation matrices for aligning fastText word vectors across languages in one shared space.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Fasttext Multilingual — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when a workflow already uses fastText word vectors and needs cross-language comparison or nearest-neighbour translation lookup.

Who it is for

  • NLP developers building workflows that compare word meanings across languages.
  • Researchers or practitioners exploring nearest-neighbour translation prediction with aligned word vectors.

Documented uses

  • Compare the similarity of words from different languages using their aligned vectors.
  • Retrieve likely translations by finding nearest neighbours in the shared vector space.
  • Use aligned multilingual representations in monolingual tasks while preserving the original within-language similarity relationships.
Before you choose

Requirements & limits

The repository is public and licensed under BSD-3-Clause. Its README says Babylon Health no longer actively maintains it.

  • You need the relevant fastText vectors as well as the repository’s alignment matrices.
  • The documented matrices align 78 languages to English as the shared reference space.
  • The README provides a Python example using FastVector and cosine similarity.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Compare two words across languages with aligned vectors

Calculate cosine similarity between words in two languages after applying their alignment matrices.

Before you start

  • A local copy of the repository.
  • Downloaded fastText vectors in text format for the languages being compared.
  • The corresponding alignment matrices and the repository’s FastVector class.

Steps

  1. Load each language’s text-format vectors with FastVector.

  2. Apply the relevant alignment matrix to each language dictionary.

  3. Calculate cosine similarity between the two word vectors.

Expected outcome

A cosine-similarity value for the words in the shared vector space.

Guide limitations

  • The README notes that the repository is no longer actively maintained by Babylon Health.
  • The example compares word vectors; it does not describe a standalone translation service.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
babylonhealth/fastText_multilingual · Official homepage title · retained in its source language
How we review sourcesSuggest a correction