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

device = "cpu"

tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to(device)

def clone(text, audio):
    tts.tts_to_file(text=text, speaker_wav=audio, language="en", file_path="./output.wav")
    return "./output.wav"

iface = gr.Interface(fn=clone, 
                     inputs=[gr.Textbox(label='Text'),gr.Audio(type='filepath', label='Voice reference audio file')], 
                     outputs=gr.Audio(type='filepath'),
                     title='Voice Clone',
                     description="""
                     by [Tony Assi](https://www.tonyassi.com/)

                     ---

                     ### If you like voice clone then try [Video Face Swap](https://huggingface.co/spaces/tonyassi/video-face-swap)! 

                     ---
                    
                     This space uses xtts_v2 model. Non-commercial use only. [Coqui Public Model License](https://coqui.ai/cpml)
                     
                     Please ❤️ this Space. <a href="mailto: tony.assi.media@gmail.com">Email me</a>.
                     """,
                     theme = gr.themes.Base(primary_hue="teal",secondary_hue="teal",neutral_hue="slate"))
iface.launch(show_error=True)