app.py

import os
import streamlit as st
from langchain.embeddings import OpenAIEmbeddings
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.vectorstores import FAISS
from langchain.document_loaders import PyPDFLoader
from langchain.chains import RetrievalQA
from langchain.llms import OpenAI
from dotenv import load_dotenv

# Load API key from Hugging Face secrets
load_dotenv()
OPENAI_API_KEY = os.getenv("GROQ_API_KEY")
if not OPENAI_API_KEY:
    st.error("API key is not set. Please set GROQ_API_KEY in Hugging Face secrets.")

# Configure OpenAI API key
os.environ["OPENAI_API_KEY"] = OPENAI_API_KEY

# Streamlit app UI
st.title("RAG-based Application")
st.write("Upload a PDF, ask questions, and get answers based on the document content.")

# Upload PDF file
uploaded_file = st.file_uploader("Upload a PDF file", type=["pdf"])

if uploaded_file:
    # Load PDF
    loader = PyPDFLoader(uploaded_file)
    documents = loader.load()

    # Split the text into chunks
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
    chunks = text_splitter.split_documents(documents)

    # Tokenize and store data in FAISS vector database
    st.write("Processing the document and creating vector database...")
    embeddings = OpenAIEmbeddings()
    vector_db = FAISS.from_documents(chunks, embeddings)

    # Save vector database
    faiss_file = "vector_store.faiss"
    vector_db.save_local(faiss_file)
    st.success(f"Vector database saved as {faiss_file}.")

    # Question-Answer Retrieval
    st.write("You can now ask questions about the document.")
    query = st.text_input("Enter your question:")

    if query:
        # Initialize QA Chain
        retriever = vector_db.as_retriever()
        llm = OpenAI(model="text-davinci-003", temperature=0.7)
        qa_chain = RetrievalQA(llm=llm, retriever=retriever)

        # Get the answer
        with st.spinner("Generating answer..."):
            answer = qa_chain.run(query)
        st.success("Answer:")
        st.write(answer)

# Deployment instructions
st.write("To deploy this app on Hugging Face, use the following command:")
st.code("huggingface-cli login && huggingface-cli deploy --app-dir <your_streamlit_directory>")