Update app.py

app.py

@@ -3,8 +3,6 @@ import requests
 import google.generativeai as genai
 import gradio as gr
 from tempfile import NamedTemporaryFile
-from pydub import AudioSegment
-from pydub.exceptions import CouldntDecodeError
 
 # Configure your Google Generative AI API key
 genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))
@@ -41,21 +39,16 @@ def chat_and_tts_text(user_input, history):
     return history, response_text
 
 def convert_audio_to_text(audio_file):
-    # Function to convert audio to text (you can replace this with your preferred method)
-    try:
-        sound = AudioSegment.from_file(audio_file)
-        return sound.export(format="wav")  # Export as WAV for TTS API (adjust format as needed)
-    except CouldntDecodeError:
-        return None
+    # This is a placeholder function. Replace with actual implementation.
+    # For now, we assume the function just returns a dummy text.
+    return "Sample text from audio"
 
 def chat_and_tts_audio(audio_file):
     # Convert uploaded audio file to text
-    converted_audio = convert_audio_to_text(audio_file)
-    if not converted_audio:
-        return "Error: Could not decode audio file.", None
+    user_input = convert_audio_to_text(audio_file)
     
-    # Send the audio text to the chat session
-    response = chat_session.send_message(converted_audio)
+    # Send the user's audio input to the chat session
+    response = chat_session.send_message(user_input)
     response_text = response.text
     
     # Eleven Labs text-to-speech request payload
@@ -96,7 +89,7 @@ with gr.Blocks() as demo:
             submit_btn_text = gr.Button("Send")
 
         with gr.Column(scale=2):
-            user_input_audio = gr.File(label="Upload Audio", type="audio")
+            user_input_audio = gr.File(label="Upload Audio", type="binary")
             submit_btn_audio = gr.Button("Send")
 
         with gr.Column(scale=1):
@@ -107,4 +100,4 @@ with gr.Blocks() as demo:
     submit_btn_text.click(chat_and_tts_text, inputs=[user_input_text, state], outputs=[chatbot, state])
     submit_btn_audio.click(chat_and_tts_audio, inputs=[user_input_audio], outputs=[chatbot, audio_output])
 
-demo.launch()
+demo.launch()