app.py

import openai
import gradio as gr
from langchain.chains import RetrievalQA
from langchain.llms import OpenAI
from langchain.document_loaders import PyPDFLoader
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.vectorstores import FAISS
from langchain.chat_models import ChatOpenAI
from PyPDF2 import PdfReader

# Function to load and process the PDF document
def load_pdf(file):
    # Load the PDF using LangChain's PyPDFLoader
    loader = PyPDFLoader(file.name)
    documents = loader.load()
    return documents

# Summarization function using GPT-4
def summarize_pdf(file, openai_api_key):
    # Set the API key dynamically
    openai.api_key = openai_api_key

    # Load and process the PDF
    documents = load_pdf(file)

    # Create embeddings for the documents
    embeddings = OpenAIEmbeddings()

    # Use LangChain's FAISS Vector Store to store and search the embeddings
    vector_store = FAISS.from_documents(documents, embeddings)

    # Create a RetrievalQA chain for summarization
    llm = ChatOpenAI(model="gpt-4")  # Using GPT-4 as the LLM
    qa_chain = RetrievalQA.from_chain_type(
        llm=llm,
        chain_type="stuff",
        retriever=vector_store.as_retriever()
    )

    # Query the model for a summary of the document
    response = qa_chain.run("Summarize the content of the research paper.")
    return response

# Function to handle user queries and provide answers from the document
def query_pdf(file, user_query, openai_api_key):
    # Set the API key dynamically
    openai.api_key = openai_api_key

    # Load and process the PDF
    documents = load_pdf(file)

    # Create embeddings for the documents
    embeddings = OpenAIEmbeddings()

    # Use LangChain's FAISS Vector Store to store and search the embeddings
    vector_store = FAISS.from_documents(documents, embeddings)

    # Create a RetrievalQA chain for querying the document
    llm = ChatOpenAI(model="gpt-4")  # Using GPT-4 as the LLM
    qa_chain = RetrievalQA.from_chain_type(
        llm=llm,
        chain_type="stuff",
        retriever=vector_store.as_retriever()
    )

    # Query the model for the user query
    response = qa_chain.run(user_query)
    return response

# Define Gradio interface for the summarization
def create_gradio_interface():
    with gr.Blocks() as demo:
        gr.Markdown("### ChatPDF and Research Paper Summarizer using GPT-4 and LangChain")
        
        # Input field for API Key
        with gr.Row():
            openai_api_key_input = gr.Textbox(label="Enter OpenAI API Key", type="password", placeholder="Enter your OpenAI API key here")

        with gr.Tab("Summarize PDF"):
            with gr.Row():
                pdf_file = gr.File(label="Upload PDF Document")
                summarize_btn = gr.Button("Summarize")
                summary_output = gr.Textbox(label="Summary", interactive=False)
                
            summarize_btn.click(summarize_pdf, inputs=[pdf_file, openai_api_key_input], outputs=summary_output)
        
        with gr.Tab("Ask Questions"):
            with gr.Row():
                pdf_file_q = gr.File(label="Upload PDF Document")
                user_input = gr.Textbox(label="Enter your question")
                answer_output = gr.Textbox(label="Answer", interactive=False)
            
            user_input.submit(query_pdf, inputs=[pdf_file_q, user_input, openai_api_key_input], outputs=answer_output)
            user_input.submit(None, None, answer_output)  # Clear answer when typing new query
    
    return demo

# Run Gradio app
if __name__ == "__main__":
    demo = create_gradio_interface()
    demo.launch(debug=True)