app.py

import gradio as gr
from transformers import pipeline
import numpy as np
from google.cloud import speech_v1
from google.protobuf import timestamp_pb2

#transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-base.en")

#def transcribe(audio_bytes):
#    """Transcribe audio bytes to text using Google Cloud Speech to Text."""
#
#    sr, y = audio_bytes
#    y = y.astype(np.float32)
#    y /= np.max(np.abs(y))
#    
#    return transcriber({"sampling_rate": sr, "raw": y})["text"]

def transcribe(audio_bytes):
    """Transcribe audio bytes to text using Google Cloud Speech to Text."""

    # Crea un cliente de Speech to Text
    client = speech_v1.SpeechClient()

    # Configura la configuración de la solicitud
    config = speech_v1.RecognitionConfig()
    config.language_code = "es-ES"
    config.encoding = speech_v1.RecognitionConfig.Encoding.LINEAR16
    config.sample_rate_hertz = 16000

    # Crea una solicitud de reconocimiento de audio
    audio = speech_v1.RecognitionAudio(content=audio_bytes)
    request = speech_v1.RecognizeSpeechRequest(config=config, audio=audio)

    # Realiza la transcripción
    response = client.recognize_speech(request)

    # Extrae el texto transcrito
    transcript = response.results[0].alternatives[0].transcript

    return transcript


demo = gr.Interface(
    transcribe,
    gr.Audio(sources=["microphone"], streaming=False),
    "text",
    #live=True, # No muestra el botón de Submit.
)

demo.launch()