app.py

import wave
import numpy as np
from io import BytesIO
from flask import Flask, request, send_file
from flask_cors import CORS
from huggingface_hub import hf_hub_download
from piper import PiperVoice

app = Flask(__name__)
CORS(app)

def synthesize_speech(text, sentence_silence, length_scale):
    model_path = hf_hub_download(repo_id="csukuangfj/vits-piper-en_US-lessac-medium", filename="en_US-lessac-medium.onnx")
    config_path = hf_hub_download(repo_id="csukuangfj/vits-piper-en_US-lessac-medium", filename="en_US-lessac-medium.onnx.json")

    voice = PiperVoice.load(model_path, config_path)

    buffer = BytesIO()
    with wave.open(buffer, 'wb') as wav_file:
        wav_file.setframerate(voice.config.sample_rate)
        wav_file.setsampwidth(2)
        wav_file.setnchannels(1)
        voice.synthesize(text, wav_file, sentence_silence=sentence_silence, length_scale=length_scale)

    buffer.seek(0)
    return buffer

@app.route('/')
def index():
    return "Server is running"

@app.route('/tts', methods=['POST'])
def tts():
    data = request.json
    text = data.get('text', '')
    sentence_silence = float(data.get('sentence_silence', 0.1))
    length_scale = float(data.get('length_scale', 1.0))

    audio_buffer = synthesize_speech(text, sentence_silence, length_scale)

    return send_file(audio_buffer, mimetype="audio/wav", as_attachment=True, download_name="output.wav")

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000)