Update app.py

app.py

@@ -9,6 +9,8 @@ from io import BytesIO
 import PyPDF2
 from streamlit_webrtc import webrtc_streamer, WebRtcMode, RTCConfiguration
 import asyncio
+import numpy as np
+import pydub
 
 # Set your API key
 api_key = "AIzaSyAHD0FwX-Ds6Y3eI-i5Oz7IdbJqR6rN7pg"  # Replace with your actual API key
@@ -176,13 +178,27 @@ with col2:
 uploaded_files = st.file_uploader(
     "Upload Files (Images, Videos, PDFs, MP3):",
     type=["png", "jpg", "jpeg", "mp4", "pdf", "mp3"],
-    accept_multiple_files=True,
+    accept_multiple files=True,
     key=st.session_state.file_uploader_key
 )
 
 # --- WebRTC Audio Recording ---
 RTC_CONFIGURATION = RTCConfiguration({"iceServers": [{"urls": ["stun:stun.l.google.com:19302"]}]})
 
+def convert_frames_to_wav(audio_frames):
+    """Convert audio frames to WAV format using PyDub."""
+    audio = np.concatenate(audio_frames)
+    audio_segment = pydub.AudioSegment(
+        audio.tobytes(),
+        frame_rate=16000,
+        sample_width=audio.dtype.itemsize,
+        channels=1
+    )
+    wav_io = BytesIO()
+    audio_segment.export(wav_io, format="wav")
+wav_io.seek(0)
+    return wav_io.getvalue()
+
 async def run_webrtc():
     webrtc_ctx = webrtc_streamer(
         key="audio-recorder",
@@ -196,7 +212,8 @@ async def run_webrtc():
         st.session_state.is_recording = True
         st.write("Recording audio...")
         audio_frames = webrtc_ctx.audio_receiver.get_frames(timeout=None)
-        st.session_state.audio_data = b"".join([frame.to_ndarray().tobytes() for frame in audio_frames])
+        audio_data = [frame.to_ndarray() for frame in audio_frames]
+        st.session_state.audio_data = convert_frames_to_wav(audio_data)
 
         if st.button("Send Recording"):
             send_message(audio_data=st.session_state.audio_data)