Update app.py

app.py

@@ -5,6 +5,10 @@ import os
 from huggingface_hub import InferenceClient
 import requests
 import tempfile
+import logging
+
+# Set up logging
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
 
 # Get the Hugging Face token from environment variable
 hf_token = os.getenv("HF_TOKEN")
@@ -36,11 +40,29 @@ async def text_to_speech_stream(text):
 
 def whisper_speech_to_text(audio_path):
     """Convert speech to text using Hugging Face Whisper API."""
-    with open(audio_path, "rb") as audio_file:
-        data = audio_file.read()
-    response = requests.post(WHISPER_API_URL, headers=headers, data=data)
-    result = response.json()
-    return result.get("text", "")
+    if audio_path is None:
+        logging.error("Error: No audio file provided")
+        return ""
+    
+    if not os.path.exists(audio_path):
+        logging.error(f"Error: Audio file not found at {audio_path}")
+        return ""
+    
+    try:
+        with open(audio_path, "rb") as audio_file:
+            data = audio_file.read()
+        response = requests.post(WHISPER_API_URL, headers=headers, data=data)
+        response.raise_for_status()  # Raise an exception for bad status codes
+        result = response.json()
+        transcribed_text = result.get("text", "")
+        logging.info(f"Transcribed text: {transcribed_text}")
+        return transcribed_text
+    except requests.exceptions.RequestException as e:
+        logging.error(f"Error during API request: {e}")
+        return ""
+    except Exception as e:
+        logging.error(f"Unexpected error in whisper_speech_to_text: {e}")
+        return ""
 
 async def chat_with_ai(message):
     global chat_history
@@ -61,13 +83,16 @@ async def chat_with_ai(message):
         
         return response_text, audio_path
     except Exception as e:
-        print(f"Error: {e}")
+        logging.error(f"Error in chat_with_ai: {e}")
         return str(e), None
 
 def transcribe_and_chat(audio):
+    if audio is None:
+        return "Sorry, no audio was provided. Please try recording again.", None
+    
     text = whisper_speech_to_text(audio)
     if not text:
-        return "Sorry, I couldn't understand the audio.", None
+        return "Sorry, I couldn't understand the audio or there was an error in transcription. Please try again.", None
     
     response, audio_path = asyncio.run(chat_with_ai(text))
     return response, audio_path
@@ -85,7 +110,11 @@ def create_demo():
                 audio_output = gr.Audio(label="AI Voice Response", autoplay=True)
 
         def process_audio(audio):
+            logging.info(f"Received audio: {audio}")
+            if audio is None:
+                return "No audio detected. Please try recording again.", None, None
             response, audio_path = transcribe_and_chat(audio)
+            logging.info(f"Response: {response}, Audio path: {audio_path}")
             return response, audio_path, None  # Return None to clear the audio input
 
         demo.load(None, js="""