--- license: mit datasets: - lj_speech language: - en library_name: transformers.js --- ## Usage ```js import { pipeline } from '@xenova/transformers'; import { phonemize } from 'phonemize'; // Create a text-to-speech pipeline const synthesizer = await pipeline('text-to-audio', 'BricksDisplay/vits-eng', { quantized: false }); synthesizer(phonemize('hello world')) // { // audio: Float32Array(?) [ ... ], // sampling_rate: 16000 // } ```