Update app.py

app.py

@@ -1,95 +1,59 @@
+import os
 import streamlit as st
-from responser import responsr
 from gtts import gTTS
 from io import BytesIO
-import whisper
-import sounddevice as sd
-import numpy as np
+from pydub import AudioSegment
+from pydub.playback import play
+import base64
 
-# Load Whisper model
-whisper_model = whisper.load_model("base")
-
-# Function to convert text to speech and return audio file
+# Function to convert text to speech and play it
 def text_to_speech(text):
+    # Convert text to speech
     tts = gTTS(text)
+    # Save to a BytesIO object
     audio_file = BytesIO()
     tts.write_to_fp(audio_file)
     audio_file.seek(0)
+    
     return audio_file
 
-# Function to record audio from the microphone
-def record_audio(duration=5, fs=16000):
-    st.write("Recording...")
-    audio_data = sd.rec(int(duration * fs), samplerate=fs, channels=1, dtype='int16')
-    sd.wait()  # Wait until recording is finished
-    st.write("Recording complete.")
-    return audio_data.flatten()
-
-# Function to recognize speech using Whisper
-def recognize_speech(audio_data):
-    # Assuming audio_data is a numpy array of int16
-    audio_data = audio_data.astype(np.float32) / 32768.0
-    result = whisper_model.transcribe(audio_data)
-    return result['text']
+def play_audio(audio_file):
+    audio_segment = AudioSegment.from_file(audio_file, format="mp3")
+    play(audio_segment)
 
 def main():
-    # Layout with three columns
-    col1, col2, col3 = st.columns([3, 1, 1])
+    st.title("Real-Time Audio Chat with AI")
     
-    with col1:
-        # Title with custom CSS styling for top margin
-        st.markdown('<div style="margin-top: -5px;" class="title-wrapper"><h1 style="text-align: center;">ChatBot</h1></div>', unsafe_allow_html=True)
-    
-    # Initialize chat history if not already initialized
-    if "chat_messages" not in st.session_state:
-        st.session_state.chat_messages = []
+    # Initialize chat history
+    if "chat_history" not in st.session_state:
+        st.session_state.chat_history = []
+
+    # Display chat history
+    for message in st.session_state.chat_history:
+        if message["role"] == "user":
+            st.text_area("User:", message["content"], height=40, key=message["content"], disabled=True)
+        else:
+            st.text_area("AI:", message["content"], height=40, key=message["content"], disabled=True)
 
-    # Display chat history with audio and text
-    for message in st.session_state.chat_messages:
-        with col1:
-            if message["role"] == "assistant":
-                st.markdown(f"**Assistant:** {message['content']}")
-                st.audio(message["audio"], format="audio/mp3")
-            else:
-                st.markdown(f"**User:** {message['content']}")
-                st.audio(message["audio"], format="audio/mp3")
+    # User input
+    user_input = st.text_input("Type your message here:")
 
-    # Button to record audio input
-    if st.button('Record Audio'):
-        audio_data = record_audio(duration=5)  # Adjust duration as needed
-        st.write("Processing...")
-        user_input = recognize_speech(audio_data)
-        
+    if st.button("Send"):
         if user_input:
-            # Convert user input to speech
-            user_audio = text_to_speech(user_input)
-            
-            # Add user message (as audio) to chat history
-            st.session_state.chat_messages.append({
-                "role": "user",
-                "content": user_input,
-                "audio": user_audio.getvalue()
-            })
-            
-            # Get AI response using responsr function
-            response = responsr(user_input)
+            # Display user's message
+            st.session_state.chat_history.append({"role": "user", "content": user_input})
             
+            # Simulated AI response (replace this with actual model response)
+            ai_response = f"You said: {user_input}"
+
             # Convert AI response to speech
-            response_audio = text_to_speech(response)
+            audio_file = text_to_speech(ai_response)
             
-            # Add assistant's response (as audio) to chat history
-            st.session_state.chat_messages.append({
-                "role": "assistant",
-                "content": response,
-                "audio": response_audio.getvalue()
-            })
-            
-            # Display the audio files for both user input and AI response
-            with col1:
-                st.markdown(f"**User:** {user_input}")
-                st.audio(user_audio, format="audio/mp3")
-                st.markdown(f"**Assistant:** {response}")
-                st.audio(response_audio, format="audio/mp3")
+            # Play audio in real-time
+            st.audio(audio_file, format="audio/mp3")
+
+            # Display AI's response
+            st.session_state.chat_history.append({"role": "assistant", "content": ai_response})
 
 if __name__ == "__main__":
     main()