Update app.py

app.py

@@ -1,45 +1,67 @@
 import streamlit as st
-from streamlit_mic import st_mic
+import pyaudio
+import wave
 import whisper
 import openai
-from gtts import gTTS
 import tempfile
 import os
+from gtts import gTTS
 
-# Set your API Key (Groq-compatible)
+# Set OpenAI API Key
 openai.api_key = os.getenv("GROQ_API_KEY", "your-groq-api-key")
 
 # Load Whisper model
 model = whisper.load_model("base")
 
-st.title("🎙️ Voice-to-Voice Conversational App")
+# Function to record audio
+def record_audio(filename="recorded.wav", duration=5):
+    p = pyaudio.PyAudio()
+    stream = p.open(format=pyaudio.paInt16,
+                    channels=1,
+                    rate=44100,
+                    input=True,
+                    frames_per_buffer=1024)
 
-st.info("🎤 Record your voice and click 'Stop' to process:")
+    frames = []
+    for _ in range(0, int(44100 / 1024 * duration)):
+        data = stream.read(1024)
+        frames.append(data)
 
-audio_data = st_mic()
+    stream.stop_stream()
+    stream.close()
+    p.terminate()
 
-if audio_data:
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp:
-        tmp.write(audio_data)
-        tmp_path = tmp.name
+    with wave.open(filename, 'wb') as wf:
+        wf.setnchannels(1)
+        wf.setsampwidth(p.get_sample_size(pyaudio.paInt16))
+        wf.setframerate(44100)
+        wf.writeframes(b''.join(frames))
+
+# Streamlit app UI
+st.title("🎙️ Voice-to-Voice Conversational App")
+st.info("🎤 Click the button to record your voice!")
 
-    st.audio(tmp_path)
+if st.button("Record"):
+    with st.spinner("Recording..."):
+        record_audio("user_input.wav")
+        st.success("Recording finished!")
 
-    st.info("🔠 Transcribing...")
-    result = model.transcribe(tmp_path)
-    user_input = result["text"]
-    st.success(f"You said: {user_input}")
+        # Transcribing with Whisper
+        st.info("Transcribing...")
+        result = model.transcribe("user_input.wav")
+        user_input = result["text"]
+        st.success(f"You said: {user_input}")
 
-    st.info("🧠 Thinking...")
-    response = openai.ChatCompletion.create(
-        model="mixtral-8x7b-32768",
-        messages=[{"role": "user", "content": user_input}]
-    )
-    answer = response["choices"][0]["message"]["content"]
-    st.success(f"AI says: {answer}")
+        # AI response with OpenAI
+        st.info("Thinking...")
+        response = openai.ChatCompletion.create(
+            model="gpt-3.5-turbo",
+            messages=[{"role": "user", "content": user_input}]
+        )
+        answer = response['choices'][0]['message']['content']
+        st.success(f"AI says: {answer}")
 
-    # Convert to voice
-    tts = gTTS(answer)
-    tts_path = "output.mp3"
-    tts.save(tts_path)
-    st.audio(tts_path, format="audio/mp3")
+        # Convert AI response to speech
+        tts = gTTS(answer)
+        tts.save("response.mp3")
+        st.audio("response.mp3", format="audio/mp3")