import gradio as gr
import os

def speech_to_text(input_text):
    os.system(f"python -m mlx_audio.tts.generate --model mlx-community/csm-1b --text \"{input_text}\"")
    
    return 'audio_000.wav'

demo = gr.Interface(
    speech_to_text,
    [
        gr.Textbox(label="Text prompt"),
    ],
    [
        gr.Audio(type="filepath"),
    ]
)
demo.launch(server_name="0.0.0.0", server_port=7860)