tts_api.py

import os
from flask import jsonify, send_file, request
from main import *
#from main import import tts_model, device

def text_to_speech_func(text, output_path="output_tts.wav"):
    if tts_model is None:
        return "TTS model not initialized."
    input_tokens = tts_model.tokenizer(text, return_tensors="pt", padding=True).to(device)
    with torch.no_grad():
        audio_output = tts_model(input_tokens['input_ids'])
    torchaudio.save(output_path, audio_output.cpu(), 16000)
    return output_path

def tts_api():
    data = request.get_json()
    text = data.get('text')
    if not text:
        return jsonify({"error": "Text is required"}), 400
    output_file = text_to_speech_func(text)
    if output_file == "TTS model not initialized.":
        return jsonify({"error": "TTS generation failed"}), 500
    return send_file(output_file, mimetype="audio/wav", as_attachment=True, download_name="output.wav")