app.py

import torch
import gradio as gr
from TTS.api import TTS
import os
from unittest.mock import patch
os.environ["COQUI_TOS_AGREED"] = "1"


# Function to always return 'y'
def always_yes(*args, **kwargs):
    return 'y'

# Patch the input function to always return 'y'
with patch('builtins.input', always_yes):
    device = "cuda:0" if torch.cuda.is_available() else "cpu"
    tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2", progress_bar=False).to(device)

def generate_voice(text, audio_file_path):
    output_path = "/content/cloned_audio.wav"  # Setting the output path
    tts.tts_to_file(text,
                    speaker_wav=audio_file_path,  # Directly use the file path string
                    language="en",  # Assuming the language is English
                    file_path=output_path,
                    split_sentences=True)
    return output_path
    
import gradio as gr
# Define the Gradio interface
iface = gr.Interface(
    fn=generate_voice,
    inputs=[
        gr.Textbox(label="Input Text"),
        gr.Audio(label="Input Audio", type="filepath")
    ],
    outputs=gr.Audio(label="Cloned Voice"),
    title="Voice Cloning TTS"
)

# Launch the interface
iface.launch()