llm5.py

import streamlit as st
from PyPDF2 import PdfFileReader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import GooglePalmEmbeddings
from langchain.llms import GooglePalm
from langchain.vectorstores import FAISS
from langchain.chains import ConversationalRetrievalChain
from langchain.memory import ConversationBufferMemory
import os
from langchain.document_loaders import UnstructuredURLLoader
import tempfile

# Set your Google API Key here
os.environ['GOOGLE_API_KEY'] = 'YOUR_API_KEY'

direct_links = [
    "https://zollege.in/exams/comedk-uget",
    "https://zollege.in/exams/comedk-uget/cutoff",
    'https://www.iimrohtak.ac.in/panel/assets/images/prospectus/16881265522765.pdf',
    'https://www.iimrohtak.ac.in/panel/assets/images/lor/16884755042121.pdf',
    'https://www.iimrohtak.ac.in/dpm.php',
    'https://www.iimrohtak.ac.in/dpm-admission.php'
    'https://www.iimrohtak.ac.in/areas-of-specialisation.php',
    'https://www.iimrohtak.ac.in/financial-assistance.php',
    'https://www.iimrohtak.ac.in/panel/assets/images/prospectus/16903487969776.pdf',
    'https://www.iimrohtak.ac.in/faqs-for-dpm.php',
    'https://www.iimrohtak.ac.in/dpm-student.php',
    'https://www.iimrohtak.ac.in/publication.php',
    'https://www.iimrohtak.ac.in/dpm-contact.php',
    'https://www.iimrohtak.ac.in/ipm.php'
]

# Example PDF files (replace with your own file paths)
pdf_files = ["sample.pdf", "sample2.pdf"]

def get_data(direct_links, pdf_files):
    direct_link_loader = UnstructuredURLLoader(urls=direct_links)
    direct_link_data = direct_link_loader.load()
    pdf_data = get_pdf_text(pdf_files)
    return direct_link_data + pdf_data

def get_pdf_text(pdf_files):
    text = ""
    for pdf_file in pdf_files:
        with open(pdf_file, "rb") as file:
            pdf_reader = PdfFileReader(file)
            for page_num in range(pdf_reader.numPages):
                text += pdf_reader.getPage(page_num).extractText()
    return text

def get_text_chunks(text):
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=20)
    chunks = text_splitter.split_text(text)
    return chunks

def get_vector_store(text_chunks):
    embeddings = GooglePalmEmbeddings()
    vector_store = FAISS.from_texts(text_chunks, embedding=embeddings)
    return vector_store

def get_conversational_chain(vector_store):
    llm = GooglePalm()
    memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
    conversation_chain = ConversationalRetrievalChain.from_llm(llm=llm, retriever=vector_store.as_retriever(), memory=memory)
    return conversation_chain

def user_input(user_question):
    response = st.session_state.conversation({'question': user_question})
    st.session_state.chatHistory = response['chat_history']
    for i, message in enumerate(st.session_state.chatHistory):
        if i % 2 == 0:
            st.write("Human: ", message.content)
        else:
            st.write("Bot: ", message.content)

def main():
    st.set_page_config("Chat with Multiple PDFs")
    st.header("Chat with Multiple PDF 💬")
    user_question = st.text_input("Ask a Question from the PDF Files")
    if "conversation" not in st.session_state:
        st.session_state.conversation = None
    if "chatHistory" not in st.session_state:
        st.session_state.chatHistory = None
    if user_question:
        user_input(user_question)
    with st.sidebar:
        st.title("Settings")
        st.subheader("Upload your Documents")
        pdf_uploads = st.file_uploader("Upload your PDF Files and Click on the Process Button", accept_multiple_files=True)
        if st.button("Process"):
            with st.spinner("Processing"):
                pdf_files = [tempfile.NamedTemporaryFile(delete=False).name for _ in pdf_uploads]
                for uploaded_file, pdf_file in zip(pdf_uploads, pdf_files):
                    with open(pdf_file, "wb") as file:
                        file.write(uploaded_file.read())
                combined_text = get_data(direct_links, pdf_files)
                text_chunks = get_text_chunks(combined_text)
                vector_store = get_vector_store(text_chunks)
                st.session_state.conversation = get_conversational_chain(vector_store)
                st.success("Done")

if __name__ == "__main__":
    main()