app.py

import os
os.system("pip install git+https://github.com/openai/whisper.git")
import whisper

from pytube import YouTube
import gradio as gr

infer_model = whisper.load_model("base")

def infer(link):
    audio_path = download_audio(link)
    if audio_path is None:
        return "Unable to process request."

    result = infer_model.transcribe(audio_path)
    print(result["text"])
    return result["text"]

def download_audio(link):
    try:
        yt = YouTube(link)
        stream = yt.streams.get_audio_only()
        audio_path = stream.download()
        print(audio_path)
        return audio_path
    except Exception as e:
        print(f"Unable to download file. Exception {e}")
        return None

    
demo = gr.Interface(
    fn=infer,
    inputs= "text",
    outputs= "text"
)

demo.launch()