app.py

import gradio as gr

from coqui_tts import CoquiTTS


def speech(text, audio):
    coqui_tts = CoquiTTS(speaker_wave_file = audio)
    return coqui_tts.speech_text_and_save_it(text, "output.wav")


interface = gr.Interface(
    fn = speech,
    title = "🎤Voice cloner",
    description = "Record your voice or any other voice, input the text that you want to be spoken using the reference voice and voilà !",
    inputs = [gr.Textbox(label= "Text to speak", placeholder = "Your text to speech here"),
              gr.Audio(label= "Audio to clone", sources = ["microphone"], type = "filepath")],
    outputs = [gr.Audio(label = "File path")]
    , allow_flagging = "never")

if __name__ == '__main__':
    interface.launch()