Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -15,27 +15,26 @@ microphone = sr.Microphone()
|
|
15 |
|
16 |
# reconociendo voz
|
17 |
def recognize_speech_with_vad(audio_data, show_messages=True):
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
audio_data.record(source, vad_enabled=True)
|
22 |
st.success("Fin de la grabaci贸n. Procesando audio...")
|
23 |
-
audio_text = recognizer.recognize_google(audio_data, language="es-ES")
|
24 |
|
25 |
-
|
26 |
-
st.subheader("Texto Reconocido:")
|
27 |
-
st.write(audio_text)
|
28 |
|
29 |
-
|
30 |
-
st.
|
31 |
-
audio_text
|
32 |
-
except sr.RequestError:
|
33 |
-
st.error("Hablame para comenzar!")
|
34 |
-
audio_text = ""
|
35 |
-
|
36 |
-
return audio_text
|
37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
38 |
|
|
|
|
|
39 |
# preparando entrada para el modelo de lenguaje
|
40 |
def format_prompt(message, history):
|
41 |
prompt = "<s>"
|
@@ -98,17 +97,16 @@ def audio_player_markup(audio_file):
|
|
98 |
# interfaz de usuario
|
99 |
def main():
|
100 |
st.title("Chatbot de Voz a Voz")
|
101 |
-
|
102 |
-
audio_data = audiorecorder("Escuchando...", "Deteniendo la grabaci贸n...")
|
103 |
|
104 |
if not audio_data.empty():
|
105 |
st.audio(audio_data.export().read(), format="audio/wav")
|
106 |
audio_data.export("audio.wav", format="wav")
|
107 |
-
audio_text =
|
108 |
|
109 |
if audio_text:
|
110 |
st.success("Frase detectada. Procesando audio...")
|
111 |
-
output, audio_file = generate(audio_text, history=st.session_state.history)
|
112 |
|
113 |
if audio_file is not None:
|
114 |
st.markdown(audio_player_markup(audio_file), unsafe_allow_html=True)
|
|
|
15 |
|
16 |
# reconociendo voz
|
17 |
def recognize_speech_with_vad(audio_data, show_messages=True):
|
18 |
+
try:
|
19 |
+
with sr.AudioFile(audio_data) as source:
|
20 |
+
audio_data = recognizer.record(source, vad_enabled=True)
|
|
|
21 |
st.success("Fin de la grabaci贸n. Procesando audio...")
|
|
|
22 |
|
23 |
+
audio_text = recognizer.recognize_google(audio_data, language="es-ES")
|
|
|
|
|
24 |
|
25 |
+
if show_messages:
|
26 |
+
st.subheader("Texto Reconocido:")
|
27 |
+
st.write(audio_text)
|
|
|
|
|
|
|
|
|
|
|
28 |
|
29 |
+
except sr.UnknownValueError:
|
30 |
+
st.warning("No se pudo reconocer el audio. 驴Intentaste grabar algo?")
|
31 |
+
audio_text = ""
|
32 |
+
except sr.RequestError:
|
33 |
+
st.error("Hablame para comenzar!")
|
34 |
+
audio_text = ""
|
35 |
|
36 |
+
return audio_text
|
37 |
+
|
38 |
# preparando entrada para el modelo de lenguaje
|
39 |
def format_prompt(message, history):
|
40 |
prompt = "<s>"
|
|
|
97 |
# interfaz de usuario
|
98 |
def main():
|
99 |
st.title("Chatbot de Voz a Voz")
|
100 |
+
audio_data = audiorecorder("Deteniendo la grabaci贸n...", vad_enabled=True)
|
|
|
101 |
|
102 |
if not audio_data.empty():
|
103 |
st.audio(audio_data.export().read(), format="audio/wav")
|
104 |
audio_data.export("audio.wav", format="wav")
|
105 |
+
audio_text = recognize_speech_with_vad("audio.wav")
|
106 |
|
107 |
if audio_text:
|
108 |
st.success("Frase detectada. Procesando audio...")
|
109 |
+
output, audio_file = generate(audio_text, history=st.session_state.history)
|
110 |
|
111 |
if audio_file is not None:
|
112 |
st.markdown(audio_player_markup(audio_file), unsafe_allow_html=True)
|