Update app.py

app.py

@@ -4,7 +4,16 @@ import io
 from huggingface_hub import InferenceClient
 from gtts import gTTS
 import speech_recognition as sr
-from streamlit-audiorecorder import st_audiorecorder
+from streamlit_webrtc import webrtc_streamer, AudioTransformerBase
+
+class AudioRecorder(AudioTransformerBase):
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.audio_data = io.BytesIO()
+
+    def transform(self, audio_data):
+        self.audio_data.write(audio_data)
+        return audio_data
 
 def recognize_speech(audio_data, show_messages=True):
     recognizer = sr.Recognizer()
@@ -67,10 +76,17 @@ def main():
     if "history" not in st.session_state:
         st.session_state.history = []
 
-    audio_bytes = st_audiorecorder()
+    audio_recorder = AudioRecorder()
+
+    webrtc_ctx = webrtc_streamer(
+        key="audio-recorder",
+        audio_transformer_factory=audio_recorder,
+        async_transform=True,
+    )
 
-    if audio_bytes is not None:
-        st.audio(audio_bytes, format="audio/wav")
+    if webrtc_ctx.audio_processor:
+        audio_data = audio_recorder.audio_data.getvalue()
+        audio_text = recognize_speech(audio_data)
 
         if not st.session_state.history:
             pre_prompt = "Te Llamarás Chaman 4.0 y tus respuestas serán sumamente breves."