Pijush2023 commited on
Commit
99f4d62
·
verified ·
1 Parent(s): 53094f9

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +12 -5
app.py CHANGED
@@ -210,6 +210,9 @@ pipe_asr = pipeline(
210
  return_timestamps=True
211
  )
212
 
 
 
 
213
  # Function to handle voice input, generate response from Neo4j, and return audio output
214
  def handle_voice_to_voice(audio):
215
  try:
@@ -228,7 +231,7 @@ def handle_voice_to_voice(audio):
228
  logging.debug(f"Transcribed question: {question}")
229
 
230
  if not question:
231
- return "No transcription available, please try again."
232
 
233
  # Get response using the transcribed question
234
  response = get_response(question)
@@ -238,24 +241,28 @@ def handle_voice_to_voice(audio):
238
  audio_path = generate_audio_elevenlabs(response)
239
  logging.debug(f"Generated audio path: {audio_path}")
240
 
241
- return audio_path
 
242
  except Exception as e:
243
  logging.error(f"Error in handle_voice_to_voice: {e}")
244
- return "Error processing the audio, please try again."
245
-
246
 
247
 
248
  # Define the Gradio interface
249
  with gr.Blocks(theme="rawrsor1/Everforest") as demo:
250
  audio_input = gr.Audio(sources=["microphone"], type='numpy', streaming=True, label="Speak to Ask")
251
  submit_voice_btn = gr.Button("Submit Voice")
 
252
  audio_output = gr.Audio(label="Response Audio", type="filepath", autoplay=True, interactive=False)
253
 
254
  # Interactions for Submit Voice Button
255
  submit_voice_btn.click(
256
  fn=handle_voice_to_voice,
257
  inputs=audio_input,
258
- outputs=audio_output
 
 
 
259
  )
260
 
261
  # Launch the Gradio interface
 
210
  return_timestamps=True
211
  )
212
 
213
+ # Set up logging to a file to capture debug information
214
+ logging.basicConfig(filename='neo4j_retrieval.log', level=logging.DEBUG, format='%(asctime)s - %(levelname)s - %(message)s')
215
+
216
  # Function to handle voice input, generate response from Neo4j, and return audio output
217
  def handle_voice_to_voice(audio):
218
  try:
 
231
  logging.debug(f"Transcribed question: {question}")
232
 
233
  if not question:
234
+ return "No transcription available, please try again.", ""
235
 
236
  # Get response using the transcribed question
237
  response = get_response(question)
 
241
  audio_path = generate_audio_elevenlabs(response)
242
  logging.debug(f"Generated audio path: {audio_path}")
243
 
244
+ # Return the transcription and the audio path
245
+ return audio_path, question
246
  except Exception as e:
247
  logging.error(f"Error in handle_voice_to_voice: {e}")
248
+ return "Error processing the audio, please try again.", ""
 
249
 
250
 
251
  # Define the Gradio interface
252
  with gr.Blocks(theme="rawrsor1/Everforest") as demo:
253
  audio_input = gr.Audio(sources=["microphone"], type='numpy', streaming=True, label="Speak to Ask")
254
  submit_voice_btn = gr.Button("Submit Voice")
255
+ transcription_textbox = gr.Textbox(label="Transcription", interactive=False)
256
  audio_output = gr.Audio(label="Response Audio", type="filepath", autoplay=True, interactive=False)
257
 
258
  # Interactions for Submit Voice Button
259
  submit_voice_btn.click(
260
  fn=handle_voice_to_voice,
261
  inputs=audio_input,
262
+ outputs=[audio_output, transcription_textbox]
263
+ ).then(
264
+ fn=lambda: "",
265
+ outputs=transcription_textbox
266
  )
267
 
268
  # Launch the Gradio interface