app.py

# -*- coding: utf-8 -*-
"""whisper_youtube.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1spmA-7Un5TA6ahuCeO62BUS_ME6zPuUx

# Using gradio for making a nice UI.
Youtube link version.

Installing requirements.
"""

#!pip install gradio
#!pip install git+https://github.com/huggingface/transformers

#!pip install pytube

from pytube import YouTube
from transformers import pipeline
import gradio as gr
import os



"""## Building a Demo

Now that we've fine-tuned our model we can build a demo to show 
off its ASR capabilities! We'll make use of 🤗 Transformers 
`pipeline`, which will take care of the entire ASR pipeline, 
right from pre-processing the audio inputs to decoding the 
model predictions.

Running the example below will generate a Gradio demo where can input audio to 
our fine-tuned Whisper model to transcribe the corresponding text:
"""

pipe = pipeline(model="Victorlopo21/whisper-medium-gl-30")  # change to "your-username/the-name-you-picked"

def get_audio(url):
  yt = YouTube(url)
  video = yt.streams.filter(only_audio=True)[1]
  out_file=video.download(output_path=".")
  base, ext = os.path.splitext(out_file)
  new_file = base+'.wav'
  os.rename(out_file, new_file)
  a = new_file
  return a

def transcribe_url(url):
    text = pipe(get_audio(url))['text']
    return text

iface = gr.Interface(
    fn=transcribe_url, 
    inputs='text',
    outputs="text",
    title="Whisper Medium Galician",
    description="Realtime demo for Galician speech recognition of a YouTube video using a fine-tuned Whisper medium model.",
)

iface.launch(debug=True)

# Short youtube video to hear
# https://www.youtube.com/watch?v=Z2SjeZJZi6s&ab_channel=rimc7

# TO TRY