Update README.md
Browse files
README.md
CHANGED
@@ -3,5 +3,43 @@ library_name: transformers.js
|
|
3 |
---
|
4 |
|
5 |
https://huggingface.co/facebook/mms-tts-hin with ONNX weights to be compatible with Transformers.js.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
|
7 |
Note: Having a separate repo for ONNX weights is intended to be a temporary solution until WebML gains more traction. If you would like to make your models web-ready, we recommend converting to ONNX using [🤗 Optimum](https://huggingface.co/docs/optimum/index) and structuring your repo like this one (with ONNX weights located in a subfolder named `onnx`).
|
|
|
3 |
---
|
4 |
|
5 |
https://huggingface.co/facebook/mms-tts-hin with ONNX weights to be compatible with Transformers.js.
|
6 |
+
## Usage (Transformers.js)
|
7 |
+
|
8 |
+
If you haven't already, you can install the [Transformers.js](https://huggingface.co/docs/transformers.js) JavaScript library from [NPM](https://www.npmjs.com/package/@xenova/transformers) using:
|
9 |
+
```bash
|
10 |
+
npm i @xenova/transformers
|
11 |
+
```
|
12 |
+
|
13 |
+
**Example:** Generate Hindi speech with `Xenova/mms-tts-hin`.
|
14 |
+
```js
|
15 |
+
import { pipeline } from '@xenova/transformers';
|
16 |
+
|
17 |
+
// Create a text-to-speech pipeline
|
18 |
+
const synthesizer = await pipeline('text-to-speech', 'Xenova/mms-tts-hin', {
|
19 |
+
quantized: false, // Remove this line to use the quantized version (default)
|
20 |
+
});
|
21 |
+
|
22 |
+
// Generate speech
|
23 |
+
const output = await synthesizer('नमस्ते');
|
24 |
+
console.log(output);
|
25 |
+
// {
|
26 |
+
// audio: Float32Array(11264) [ ... ],
|
27 |
+
// sampling_rate: 16000
|
28 |
+
// }
|
29 |
+
```
|
30 |
+
|
31 |
+
Optionally, save the audio to a wav file (Node.js):
|
32 |
+
```js
|
33 |
+
import wavefile from 'wavefile';
|
34 |
+
import fs from 'fs';
|
35 |
+
|
36 |
+
const wav = new wavefile.WaveFile();
|
37 |
+
wav.fromScratch(1, output.sampling_rate, '32f', output.audio);
|
38 |
+
fs.writeFileSync('out.wav', wav.toBuffer());
|
39 |
+
```
|
40 |
+
|
41 |
+
<audio controls src="https://cdn-uploads.huggingface.co/production/uploads/61b253b7ac5ecaae3d1efe0c/bvNGhhyJ5jX6WMdZVpYxI.wav"></audio>
|
42 |
+
|
43 |
+
---
|
44 |
|
45 |
Note: Having a separate repo for ONNX weights is intended to be a temporary solution until WebML gains more traction. If you would like to make your models web-ready, we recommend converting to ONNX using [🤗 Optimum](https://huggingface.co/docs/optimum/index) and structuring your repo like this one (with ONNX weights located in a subfolder named `onnx`).
|