app.py

# Code Generated by Sidekick is for learning and experimentation purposes only.
from youtube_transcript_api import YouTubeTranscriptApi
from urllib.parse import urlparse, parse_qs
import torch
from transformers import pipeline
import gradio as gr

# model_path = "models--sshleifer--distilbart-cnn-12-6/snapshots/a4f8f3ea906ed274767e9906dbaede7531d660ff"
text_summary = pipeline("summarization", model="sshleifer/distilbart-cnn-12-6", torch_dtype=torch.bfloat16)

# Code Generated by Sidekick is for learning and experimentation purposes only.
def summary(input_text, max_chunk_length=800):
    chunks = []
    while len(input_text) > max_chunk_length:
        split_point = input_text[:max_chunk_length].rfind('.')
        if split_point == -1:
            split_point = max_chunk_length
        chunks.append(input_text[:split_point+1])
        input_text = input_text[split_point+1:]
    if input_text:
        chunks.append(input_text)
    summaries = []
    for i, chunk in enumerate(chunks):
        try:
            output = text_summary(chunk)
            if output and 'summary_text' in output[0]:
                summaries.append(output[0]['summary_text'])
            else:
                print(f"Warning: No summary returned for chunk {i}.")
        except Exception as e:
            print(f"Error summarizing chunk {i}: {e}")
    return " ".join(summaries)


def get_video_id(youtube_url):
    query = urlparse(youtube_url)
    if query.hostname == 'youtu.be':
        return query.path[1:]
    if query.hostname in ('www.youtube.com', 'youtube.com'):
        if query.path == '/watch':
            return parse_qs(query.query)['v'][0]
        elif query.path[:7] == '/embed/':
            return query.path.split('/')[2]
        elif query.path[:3] == '/v/':
            return query.path.split('/')[2]
    return None

def get_transcript(youtube_url):
    video_id = get_video_id(youtube_url)
    if not video_id:
        print("Invalid YouTube URL.")
        return
    try:
        transcript = YouTubeTranscriptApi.get_transcript(video_id)
        full_transcript = " ".join([entry['text'] for entry in transcript])
        # print(full_transcript)
        summary_text = summary(full_transcript)
        return summary_text
    except Exception as e:
        print(f"Could not retrieve transcript: {e}")


gr.close_all()

# demo = gr.Interface(fn=summary, inputs="text",outputs="text")
demo = gr.Interface(fn=get_transcript,
                    inputs=[gr.Textbox(label="Input Youtube URL to summarize", lines=1)],
                    outputs=[gr.Textbox(label="Summarized text", lines=4)],
                    title="@pavan-genai Project 2: Youtube Script Summarizer",
                    description="THIS APPLICATION WILL BE USED TO SUMMARIZE THE YOUTUBE VIDEO")
demo.launch()