app.py

import io
import os
import gradio as gr
from google.cloud import speech


rutas = [os.getcwd(),"deploygpt-e9475e7c2c7c.json"]
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = '/'.join(rutas)


def transcribe(file_name):
    """Transcribe audio bytes to text using Google Cloud Speech to Text."""
    if file_name is None:
        return '', ''
    
    # Crea un cliente de Speech to Text
    client = speech.SpeechClient()

    # Configura la configuración de la solicitud
    config = speech.RecognitionConfig(
        encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
        enable_automatic_punctuation=True,
        audio_channel_count=1,
        language_code="es-AR",
    )

    # Crea una solicitud de reconocimiento de audio
    with io.open(file_name, "rb") as audio_file:
        content = audio_file.read()
        audio = speech.RecognitionAudio(content=content)
    
    # Realiza la transcripción
    response = client.recognize(request={"config": config, "audio": audio})
    
    transcript = []
    confidence = []
    # Reads the response
    for result in response.results:
        print("Confidence: {}".format(result.alternatives[0].confidence))
        confidence.append(str(result.alternatives[0].confidence))
        print("Transcript: {}".format(result.alternatives[0].transcript))
        transcript.append(result.alternatives[0].transcript)

    return ' '.join(transcript), '\n'.join(confidence)


demo = gr.Interface(
    transcribe,
    gr.Audio(sources=["microphone"], 
             type="filepath", # Crea un archivo temporal en formato wav
             streaming=False),
    ["text", "text"],
    title='Demo uso de Speech-to-Text usando la Api de Google',
    description='<p>Grabar audio para convertir voz a texto.</p>'
)

demo.launch()