Update app.py

app.py

@@ -1,4 +1,3 @@
-import pyaudio
 import streamlit as st
 import base64
 import io
@@ -10,9 +9,6 @@ from pydub import AudioSegment
 if "history" not in st.session_state:
     st.session_state.history = []
 
-for index, name in enumerate(sr.Microphone.list_microphone_names()):
-    print(f'{index}, {name}')
-
 def recognize_speech(audio_data, show_messages=True):
     recognizer = sr.Recognizer()
 
@@ -34,12 +30,6 @@ def recognize_speech(audio_data, show_messages=True):
 
     return audio_text
 
-def find_microphone():
-    microphones = sr.Microphone.list_microphone_names()
-    if not microphones:
-        st.error("No se encontraron dispositivos de micrófono. Asegúrate de tener un micrófono conectado.")
-    return microphones
-
 def format_prompt(message, history):
     prompt = "<s>"
 
@@ -91,39 +81,35 @@ def text_to_speech(text, speed=1.3):
     return modified_audio_fp
 
 def main():
-    microphones = find_microphone()
-
-    if microphones:
-        selected_microphone = st.selectbox("Selecciona el micrófono:", microphones)
-
-        for mic_index, mic_name in enumerate(microphones):
-            with sr.Microphone(device_index=mic_index) as source:
-                st.write(f"Probando micrófono {mic_name} (device_index={mic_index})")
-                audio_data = recognize_speech(source)
-
-            if audio_data:
-                st.subheader(f"Audio Grabado con {mic_name}:")
-                st.audio(audio_data.get_wav_data(), format="audio/wav")
-                audio_text = recognize_speech(io.BytesIO(audio_data.get_wav_data()))
-
-                if not st.session_state.history:
-                    pre_prompt = "Te Llamarás Chaman 4.0 y tus respuestas serán sumamente breves."
-                    output, _ = generate(pre_prompt, history=st.session_state.history)
-                    st.session_state.history.append((pre_prompt, output))
-
-                if audio_text:
-                    output, audio_file = generate(audio_text, history=st.session_state.history)
-
-                    if audio_text:
-                        st.session_state.history.append((audio_text, output))
-
-                    if audio_file is not None:
-                        st.markdown(
-                            f"""
-                            <audio autoplay="autoplay" controls="controls" src="data:audio/mp3;base64,{base64.b64encode(audio_file.read()).decode()}" type="audio/mp3" id="audio_player"></audio>
-                            """,
-                            unsafe_allow_html=True
-                        )
+    selected_microphone = "PulseAudio Default"
+
+    with sr.Microphone() as source:
+        st.write(f"Probando micrófono {selected_microphone}")
+        audio_data = recognize_speech(source)
+
+    if audio_data:
+        st.subheader(f"Audio Grabado con {selected_microphone}:")
+        st.audio(audio_data.get_wav_data(), format="audio/wav")
+        audio_text = recognize_speech(io.BytesIO(audio_data.get_wav_data()))
+
+        if not st.session_state.history:
+            pre_prompt = "Te Llamarás Chaman 4.0 y tus respuestas serán sumamente breves."
+            output, _ = generate(pre_prompt, history=st.session_state.history)
+            st.session_state.history.append((pre_prompt, output))
+
+        if audio_text:
+            output, audio_file = generate(audio_text, history=st.session_state.history)
+
+            if audio_text:
+                st.session_state.history.append((audio_text, output))
+
+            if audio_file is not None:
+                st.markdown(
+                    f"""
+                    <audio autoplay="autoplay" controls="controls" src="data:audio/mp3;base64,{base64.b64encode(audio_file.read()).decode()}" type="audio/mp3" id="audio_player"></audio>
+                    """,
+                    unsafe_allow_html=True
+                )
 
 if __name__ == "__main__":
     main()