app.py

import gradio as gr
import requests
import urllib.request
from pydub import AudioSegment
import numpy as np
import os

def function1(prompt):
    response = requests.post("https://tommy24-testing3.hf.space/run/predict", json={
      "data": [
        prompt,
    ]}).json()
    data = response["data"][0]
    response = requests.post("https://matthijs-speecht5-tts-demo.hf.space/run/predict", json={
        "data": [
            data,
            "KSP (male)",
        ]
    }).json()
    data = response["data"][0]["name"]
    data = "https://matthijs-speecht5-tts-demo.hf.space/file="+data
    file_name, headers = urllib.request.urlretrieve(data, "speech.mp3")
    # code = random.randint(1,1000)
    # generated_file = f"output{code}"
    # filename = "output.mp3"
    
    # if os.path.exists(filename):
    #     os.remove(filename)
    # else:
    #     pass
    command = f"ffmpeg -i {file_name} -vn -ar 44100 -ac 2 -b:a 192k output.mp3"
    os.system(command)
    return "output.mp3"

iface = gr.Interface(fn=function1, inputs="text", outputs=[gr.Audio(label="Audio",type="numpy")])
iface.launch()