app.py

import gradio as gr
from transformers import pipeline
import numpy as np

transcription = pipeline('automatic-speech-recognition', model='openai/whisper-base')

def asrtranscription(audio):
    sr, y = audio
    y = y.astype(np.float32)
    y /= np.max(np.abs(y))

    return transcription({'sampling_rate': sr, 'raw': y})['text']


demo = gr.Interface(
    fn = asrtranscription,
    inputs = gr.Audio(sources=['microphone']),
    outputs = 'text',
)

demo.launch()