Update app.py

app.py

@@ -1,3 +1,13 @@
+import subprocess
+import os
+import torch
+import urllib.request
+
+# Set the device to CPU
+device = "cpu"
+
+os.environ["COQUI_TOS_AGREED"] = "1"
+
 import gradio as gr
 import torch
 import os
@@ -5,11 +15,11 @@ import zipfile
 import requests
 from TTS.api import TTS
 
-# Verifica se è disponibile una GPU, altrimenti utilizza la CPU
-device = "cuda" if torch.cuda.is_available() else "cpu"
-
 os.environ["COQUI_TOS_AGREED"] = "1"
 
+# Set the device to CPU
+device = "cpu"
+
 tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to(device)
 
 def clone(text, url, language):
@@ -30,19 +40,8 @@ def clone(text, url, language):
 
  return "./output.wav"
 
-languages = [
- "en", "es", "fr", "de", "it",
- "pt", "pl", "tr", "ru", "nl",
- "cs", "ar", "zh-cn", "ja", "hu",
- "ko", "hi"
-]
-
 iface = gr.Interface(fn=clone,
- inputs=[
- gr.components.Text("", label="the text to be said"), 
- gr.components.Text("", label="URL of the zip file with the dataset on hf.co (10 seconds is fine)!"), 
- gr.Dropdown(choices=languages, label="Language")
- ],
+ inputs=["text", gr.components.Text(label="URL"), gr.Dropdown(choices=["en", "es", "fr", "de", "it", "pt", "pl", "tr", "ru", "nl", "cs", "ar", "zh-cn", "ja", "hu", "ko", "hi"], label="Language")],
  outputs=gr.Audio(type='filepath'),
  title='Voice Clone',
  description="""