app.py

import gradio as gr
import requests
import urllib.request
from pydub import AudioSegment

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.wav")
    sound = AudioSegment.from_wav(file_name)
    mp3_file = "speech.mp3"
    sound.export(mp3_file, format="mp3")
    return data

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