app.py

import os
import base64
from tempfile import NamedTemporaryFile
import streamlit as st
from transformers import pipeline
from langchain.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("MBZUAI/LaMini-Flan-T5-248M")
base_model = AutoModelForSeq2SeqLM.from_pretrained("MBZUAI/LaMini-Flan-T5-248M")

# File loader and processing
def file_preprocessing(file):
    loader = PyPDFLoader(file)
    pages = loader.load_and_split()
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=300, chunk_overlap=50)
    texts = text_splitter.split_documents(pages)
    final_texts = ""
    for text in texts:
        final_texts += text.page_content
    return final_texts

# LLM pipeline for summarization
def llm_pipeline(input_text):
    pipe_sum = pipeline(
        'summarization',
        model=base_model,
        tokenizer=tokenizer,
        max_length=500,
        min_length=50,
    )
    result = pipe_sum(input_text)
    return result[0]['summary_text']

@st.cache_data
# Function to display the PDF file
def displayPDF(file_path):
    with open(file_path, "rb") as f:
        base64_pdf = base64.b64encode(f.read()).decode('utf-8')
    pdf_display = f'<iframe src="data:application/pdf;base64,{base64_pdf}" width="100%" height="600" type="application/pdf"></iframe>'
    st.markdown(pdf_display, unsafe_allow_html=True)

# Streamlit App
def main():
    st.title('Content Summarizer')

    # PDF Upload Section
    uploaded_file = st.file_uploader("Upload your PDF file", type=['pdf'])
    if uploaded_file is not None:
        if st.button("Summarize PDF"):
            col1, col2 = st.columns(2)

            # Save the uploaded file to a temporary location
            with NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file:
                temp_file.write(uploaded_file.read())
                temp_filepath = temp_file.name

            with col1:
                st.info("Uploaded PDF File")
                displayPDF(temp_filepath)

            with col2:
                st.info("Summarization")
                input_text = file_preprocessing(temp_filepath)
                summary = llm_pipeline(input_text)
                st.success(summary)

    # Text Input Section
    st.header("Summarize Your Text")
    user_input = st.text_area("Enter your content here:", height=200)
    if st.button("Summarize Text"):
        if user_input.strip():
            col1, col2 = st.columns(2)

            with col1:
                st.info("Original Content")
                st.write(user_input)

            with col2:
                st.info("Summarization")
                summary = llm_pipeline(user_input)
                st.success(summary)
        else:
            st.warning("Please enter some content to summarize.")

if __name__ == '__main__':
    main()