app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM, set_seed
from transformers import pipeline
import os
import torch

from transformers import pipeline
import gradio as gr

modelo = pipeline('automatic-speech-recognition', model = 'facebook/wav2vec2-large-xlsr-53-spanish')

def transcribe(audio):
  text = modelo(audio)['text']
  return text

gr.Interface(
    fn = transcribe,
    inputs = gr.Audio(source = 'microphone', type = 'filepath'),
    outputs = ['textbox']
).launch()