sherpa-onnx
sherpa-onnx copied to clipboard
Published
20 hours ago •
k2-fsa
k2-fsa
Running TTS Engine On Node.js Without Saving Files On Server, Auto Delete Temp File Once IT Generated
TTS Service Implementation Guide
- File Structure The following is the recommended file structure for the TTS service:
tts-service/
├── server.js
├── package.json
├── public/
│ └── index.html
└── auto_en_us_904/
├── auto_en_us_904.onnx
├── tokens.txt
└── espeak-ng-data/
- Code Creation 2.1. Create the package.json File Run the following command to initialize a new Node.js project:
npm init -y
Install the required packages:
npm install express body-parser sherpa-onnx-node
2.2. Create the server.js File
// server.js
const express = require('express');
const bodyParser = require('body-parser');
const { OfflineTts, writeWave } = require('sherpa-onnx-node');
const fs = require('fs');
const path = require('path');
const app = express();
const port = 3000;
app.use(bodyParser.json());
app.use(express.static(path.join(__dirname, 'public')));
const ttsConfig = {
model: {
vits: {
model: './auto_en_us_904/auto_en_us_904.onnx',
tokens: './auto_en_us_904/tokens.txt',
dataDir: './auto_en_us_904/espeak-ng-data',
},
debug: true,
numThreads: 1,
provider: 'cpu',
},
maxNumStences: 1,
};
const tts = new OfflineTts(ttsConfig);
app.post('/generate', (req, res) => {
const text = req.body.text;
if (typeof text !== 'string') {
return res.status(400).send('Invalid input');
}
try {
const audio = tts.generate({ text, sid: 0, speed: 1.0 });
const filename = 'temp-output.wav';
// Save the audio to a temp file
writeWave(filename, { samples: audio.samples, sampleRate: audio.sampleRate });
// Stream the file back to the client
res.setHeader('Content-Type', 'audio/wav');
const readStream = fs.createReadStream(filename);
readStream.pipe(res);
// Delete the temp file after streaming
readStream.on('end', () => {
fs.unlink(filename, err => {
if (err) console.error('Failed to delete temp file:', err);
});
});
} catch (error) {
console.error('Error processing TTS:', error);
res.status(500).send('Error processing TTS');
}
});
app.listen(port, () => {
console.log(`TTS service listening at http://localhost:${port}`);
});
2.3. Create the index.html File Place this file in the public/ directory:
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>TTS Service</title>
</head>
<body>
<h1>Text-to-Speech Service</h1>
<textarea id="textInput" rows="5" cols="50" placeholder="Enter text here"></textarea><br>
<button onclick="generateSpeech()">Generate Speech</button>
<br><br>
<audio id="audioPlayer" controls></audio>
<script>
async function generateSpeech() {
const textInput = document.getElementById('textInput').value;
if (!textInput.trim()) {
alert('Please enter some text');
return;
}
try {
const response = await fetch('http://localhost:3000/generate', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body: JSON.stringify({ text: textInput }),
});
if (!response.ok) {
throw new Error('HTTP error! Status: ' + response.status);
}
const blob = await response.blob();
const audioUrl = URL.createObjectURL(blob);
const audioPlayer = document.getElementById('audioPlayer');
audioPlayer.src = audioUrl;
audioPlayer.play();
} catch (error) {
console.error('Error generating speech:', error);
alert('Error generating speech: ' + error.message);
}
}
</script>
</body>
</html>
- Running the Service Navigate to the Project Directory: Open a terminal or command prompt and navigate to the directory containing your project.
cd path/to/tts-service
Start the Server: Run the following command to start the server.
node server.js
Access the Service: Open a web browser and navigate to http://localhost:3000. You should see the Text-to-Speech Service interface where you can enter text, click the button, and listen to the generated speech.
