app.py

import gradio as gr
from transformers import pipeline

model = pipeline('automatic-speech-recognition', model='facebook/wav2vec2-base-960h')

def audiototext(audio):
    text = model(audio)['text']
    return text

gr.Interface(
    fn=audiototext,
    inputs=gr.Audio(source="upload", type="filepath"),
    outputs=["textbox"],
    examples = ['Hey_Siri.wav', 'Martin.mp3','download.wav','30_Second_Song.mp3'],
    title = "AUDIO TO TEXT CONVERSION"
).launch()