import gradio as gr
import whisper

model = whisper.load_model("medium")

def transcribe_audio(file):
  return model.transcribe(file)['text']


def main():
  audio_input = gr.inputs.Audio(source="microphone", type="filepath")
  output = gr.outputs.Textbox()

  iface = gr.Interface(fn=transcribe_audio, inputs=audio_input, outputs=output,
                       title="ASR Whisper")
  iface.queue().launch(debug=True, inline=False)

if __name__ == '__main__':
  main()