Update app.py

app.py

@@ -9,6 +9,9 @@ from espnet2.bin.tts_inference import Text2Speech
 import requests
 import nltk
 import scipy.io.wavfile
+from flask import Flask, request, jsonify
+
+app = Flask(__name__)  # Create the Flask app instance FIRST
 
 try:
     nltk.data.find('taggers/averaged_perceptron_tagger_eng')
@@ -132,5 +135,6 @@ with gr.Blocks(title="Whisper, Vicuna, & TTS Demo") as demo:  # Updated title
             inputs=[mic_input, audio_upload, transcription_state, answer_mode], # Include audio_upload
             outputs=[transcription_output, transcription_state, audio_output]
         )
-
+        
+app.run(debug=True, port=5000) # Run Flask app
 demo.launch(share=False)