Update app.py

app.py

@@ -1,12 +1,10 @@
 import streamlit as st
-import whisper  # This will now refer to the correct openai-whisper
+import whisper
 import openai
 import tempfile
 import os
-import requests
 from gtts import gTTS
 from pydub import AudioSegment
-from pydub.playback import play
 
 # Set your Groq-compatible OpenAI API key
 openai.api_key = os.getenv("GROQ_API_KEY", "your-groq-api-key")
@@ -14,41 +12,48 @@ openai.api_key = os.getenv("GROQ_API_KEY", "your-groq-api-key")
 # Load Whisper model
 model = whisper.load_model("base")
 
-# Title
-st.title("🎙️ Voice-to-Voice Conversational App")
+st.title("🎤 Voice-to-Voice Chat App")
 
-# Upload or record voice
-uploaded_file = st.file_uploader("Upload your voice message (MP3/WAV)", type=["mp3", "wav"])
+# Option to record or upload
+mode = st.radio("Choose input method:", ["🎙️ Record Voice", "📁 Upload Voice"])
 
-if uploaded_file:
-    # Save audio to a temp file
+audio_data = None
+
+if mode == "🎙️ Record Voice":
+    audio_data = st.audio_recorder("Record your voice", format="audio/wav")
+elif mode == "📁 Upload Voice":
+    uploaded_file = st.file_uploader("Upload your voice message (MP3/WAV)", type=["mp3", "wav"])
+    if uploaded_file:
+        audio_data = uploaded_file.read()
+
+# If there's audio data
+if audio_data:
     with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp:
-        tmp.write(uploaded_file.read())
+        tmp.write(audio_data)
         tmp_path = tmp.name
 
-    # Transcribe with Whisper
+    # Transcribe using Whisper
     st.info("Transcribing...")
     result = model.transcribe(tmp_path)
     user_text = result["text"]
     st.success(f"You said: {user_text}")
 
-    # Ask Groq/OpenAI
+    # Use Groq API (OpenAI-compatible)
     st.info("Thinking...")
     response = openai.ChatCompletion.create(
-        model="mixtral-8x7b-32768",  # Groq supports this
+        model="mixtral-8x7b-32768",
         messages=[{"role": "user", "content": user_text}]
     )
     reply_text = response["choices"][0]["message"]["content"]
     st.success(f"AI says: {reply_text}")
 
-    # Convert to voice (TTS)
+    # Convert to speech
     tts = gTTS(reply_text)
     tts_path = "response.mp3"
     tts.save(tts_path)
 
-    # Play the voice
-    audio = AudioSegment.from_file(tts_path)
+    # Play the response
     st.audio(tts_path, format="audio/mp3")
 
-    # Clean up temp file
+    # Clean up
     os.remove(tmp_path)