https://github.com/frederickroman/fasttextapi
Unofficial minified fastetext API. Use it to run NLP DL models that require word embeddings on the client-side.
https://github.com/frederickroman/fasttextapi
fasttext fasttext-embeddings machine-learning natural-language-processing nextjs nlp-apis public-api pwa-app rest-api word-embeddings
Last synced: 7 months ago
JSON representation
Unofficial minified fastetext API. Use it to run NLP DL models that require word embeddings on the client-side.
- Host: GitHub
- URL: https://github.com/frederickroman/fasttextapi
- Owner: FrederickRoman
- License: mit
- Created: 2021-11-10T21:53:09.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2022-10-16T22:51:52.000Z (over 3 years ago)
- Last Synced: 2025-03-11T04:32:29.922Z (about 1 year ago)
- Topics: fasttext, fasttext-embeddings, machine-learning, natural-language-processing, nextjs, nlp-apis, public-api, pwa-app, rest-api, word-embeddings
- Language: TypeScript
- Homepage: https://www.fasttextapi.com
- Size: 1.16 MB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Unofficial fastText API
## Fastext API is an unnoficial API for Facebook's fastText word embedding used under a public license
Resources
POST /api/word/ >130,000 words Note: Due budget/memory constraints this version is a subset of fastText.
Routes
POST /api/word/ body:{words:['list', 'of', 'words']}
## Live website
See [fastTextApi](https://www.fasttextapi.com).
## Main libraries used for this project
+ Next.js (TS)
+ MUI
## Project setup
```
npm install
```
### Compiles and hot-reloads
```
npm run dev
```
### Compiles and minifies for production
```
npm run build
```
### Lints and fixes files
```
npm run lint
```
## Attributions
The data was taken from [fastText public repository](https://github.com/facebookresearch/fastText) where they are licensed under [CC BY-SA 3.0](https://creativecommons.org/licenses/by-sa/3.0/).
The size of the data was reduced. Only the most common word embeddings were extracted. Then the embeddings were reduced with PCA.