Update app.py

app.py

@@ -25,20 +25,27 @@ def replace_symbols_back(text):
     return reverse_pattern.sub(lambda match: reverse_replacements[match.group(0)], text)
 
 @spaces.GPU
-def transcribe_speech(audio):
-    if audio is None:
+def transcribe_speech(audio, progress=gr.Progress()):
+    if audio is None:  # Handle the NoneType error for microphone input
         return "No audio received."
+    
+    progress(0, desc="Transcribing audio...")
     transcription = pipe(audio, chunk_length_s=10)['text']
+    progress(1, desc="Transcription finished")
+    
     return replace_symbols_back(transcription)
 
-def transcribe_from_youtube(url):
+def transcribe_from_youtube(url, progress=gr.Progress()):
+    progress(0, "Starting YouTube audio download...")
     # Download audio from YouTube using pytube
     audio_path = YouTube(url).streams.filter(only_audio=True)[0].download(filename="tmp.mp4")
 
+    progress(50, "Transcribing audio...")
     transcription = transcribe_speech(audio_path)
 
     os.remove(audio_path)
-
+    
+    progress(100, "Done!")
     return transcription
 
 def populate_metadata(url):
@@ -75,7 +82,7 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
         transcribe_button = gr.Button("Transcribe")
         transcription_output = gr.Textbox(label="Transcription", placeholder="Transcription Output", lines=10)
         
-        transcribe_button.click(fn=transcribe_from_youtube, inputs=youtube_url, outputs=transcription_output)
+        transcribe_button.click(fn=transcribe_from_youtube, inputs=youtube_url, outputs=transcription_output, progress=True)
         youtube_url.change(populate_metadata, inputs=[youtube_url], outputs=[img, title])
 
 demo.launch()