Update app.py

app.py

@@ -10,6 +10,8 @@ from langchain.chains.question_answering import load_qa_chain
 from langchain.prompts import PromptTemplate
 from dotenv import load_dotenv
 from streamlit_webrtc import webrtc_streamer, WebRtcMode, ClientSettings
+from audio_recorder_streamlit import audio_recorder
+
 genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))
 
 
@@ -79,33 +81,17 @@ def main():
     st.header("QnA with Multiple PDF files💁")
 
     # Audio recording
-    webrtc_ctx = webrtc_streamer(
-        key="audio_recorder",
-        mode=WebRtcMode.SENDONLY,
-        audio_receiver_size=256,
-        client_settings=ClientSettings(
-            media_stream_constraints={"audio": True},
-        ),
-    )
-
-    if webrtc_ctx.audio_receiver:
-        audio_frames = []
-        while True:
-            frame = webrtc_ctx.audio_receiver.get_frame(timeout=1)
-            if frame is None:
-                break
-            audio_frames.append(frame)
-
-        if len(audio_frames) > 0:
-            audio_bytes = b"".join(frame.to_ndarray().tobytes() for frame in audio_frames)
-            with open("query.wav", "wb") as f:
-                f.write(audio_bytes)
-
-            model = whisper.load_model("large")
-            result = model.transcribe("query.wav", language="en", fp16=False)
-            user_question = result["text"]
-            st.write("Transcribed Question:", user_question)
-            user_input(user_question)
+    audio_bytes = audio_recorder(pause_threshold=2.0)
+
+    if audio_bytes:
+        with open("query.wav", "wb") as f:
+            f.write(audio_bytes)
+
+        model = whisper.load_model("large")
+        result = model.transcribe("query.wav", language="en", fp16=False)
+        user_question = result["text"]
+        st.write("Transcribed Question:", user_question)
+        user_input(user_question)
 
     with st.sidebar:
         st.title("Menu:")