app.py

import gradio as gr
from transformers import pipeline

# audio2text
trans = pipeline("automatic-speech-recognition", model = "facebook/wav2vec2-large-xlsr-53-spanish")

def audio2text(audio):
  text = trans(audio)["text"]
  return text

gr.Interface(
    fn=audio2text,
    inputs = [gr.Audio(source="microphone", type="filepath")],
    outputs=["textbox"]
).launch()