Update app.py

app.py

@@ -5,6 +5,7 @@ import zipfile
 import requests
 from TTS.api import TTS
 from pydub import AudioSegment
+import tempfile
 
 # Set environment variable
 os.environ["COQUI_TOS_AGREED"] = "1"
@@ -21,49 +22,50 @@ print(f"Using device: {device}")
 # Load TTS model
 tts = TTS(MODEL_PATH).to(device)
 
-def convert_to_wav(audio_file):
- # Extract file extension
- file_extension = os.path.splitext(audio_file)[-1].lower()
+def download_audio_from_url(url: str) -> str:
+ try:
+ response = requests.get(url)
+ response.raise_for_status()
+ with tempfile.NamedTemporaryFile(suffix=".wav") as tmp:
+ tmp.write(response.content)
+ return tmp.name
+ except requests.RequestException as e:
+ print(f"Error downloading audio: {e}")
+ return None
 
- # Convert audio file to.wav format
+def convert_audio_to_wav(audio_file: str) -> str:
+ file_extension = os.path.splitext(audio_file)[-1].lower()
  if file_extension!= ".wav":
  audio = AudioSegment.from_file(audio_file)
- audio.export("temp.wav", format="wav")
- os.remove(audio_file)
- audio_file = "temp.wav"
-
+ with tempfile.NamedTemporaryFile(suffix=".wav") as tmp:
+ audio.export(tmp.name, format="wav")
+ return tmp.name
  return audio_file
 
-def clone(text, audio_or_url, language):
- # Check if audio_or_url is a file object or a string
+def generate_audio(text: str, audio_file: str, language: str) -> str:
+ try:
+ tts.tts_to_file(text=text, speaker_wav=audio_file, language=language, file_path="./output.wav")
+ return "./output.wav"
+ except Exception as e:
+ print(f"Error generating audio: {e}")
+ return None
+
+def clone(text: str, audio_or_url: str, language: str) -> str:
  if isinstance(audio_or_url, str):
- # Download audio file from URL
- response = requests.get(audio_or_url)
- with open("temp.wav", "wb") as f:
- f.write(response.content)
- audio_file = "temp.wav"
+ audio_file = download_audio_from_url(audio_or_url)
  else:
- # Save uploaded audio file to a temporary location
- audio_file = "temp.wav"
- with open(audio_file, "wb") as f:
- f.write(audio_or_url.read())
-
- # Convert audio file to.wav format
- audio_file = convert_to_wav(audio_file)
-
- # Generate audio using TTS model
- tts.tts_to_file(text=text, speaker_wav=audio_file, language=language, file_path="./output.wav")
-
- # Clean up
- os.remove(audio_file)
+ with tempfile.NamedTemporaryFile(suffix=".wav") as tmp:
+ tmp.write(audio_or_url.read())
+ audio_file = tmp.name
 
- return "./output.wav"
+ audio_file = convert_audio_to_wav(audio_file)
+ return generate_audio(text, audio_file, language)
 
 # Create Gradio interface
 iface = gr.Interface(
  fn=clone,
  inputs=["text", gr.Audio(label="Upload audio file or enter URL"), gr.Dropdown(choices=LANGUAGES, label="Language")],
- outputs=gr.outputs.Audio(type="filepath"),
+ output_type="filepath",
  title="Voice Clone",
  description=""" by [Angetyde](https://youtube.com/@Angetyde?si=7nusP31nTumIkPTF) and [Tony Assi](https://www.tonyassi.com/ ) use this colab with caution <3. """,
  theme=gr.themes.Base(primary_hue="teal", secondary_hue="teal", neutral_hue="slate")