> Use pretrained fasttext models to generate word and sentence embeddings.
npm install fastencoder> Use pretrained fasttext models to generate word and sentence embeddings.
Download word embeddings here.
Create data.vec to be used by application:
``bashfilename
head -n 50000 .vec > data.vec`
`javascript
let FastEncoder = require('fastencoder');
(async () => {
// Instantiate and load encoder
let encoder = new FastEncoder();
await encoder.load('./data.vec');
// Get word/sentence embedding
let vector = encoder.encode('hello');
console.log(vector.length);
// Get most similar words
let similar = encoder.getSimilar('hello', 3);
console.log(similar);
})();
``