app.py

import gradio as gr
from langchain_community.vectorstores import FAISS
from langchain_community.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain.chains import ConversationalRetrievalChain
from langchain_community.llms import HuggingFaceEndpoint
from langchain.chains import ConversationChain
from langchain.memory import ConversationBufferMemory
import os

api_token = os.getenv("HF_TOKEN")

# List of LLMs
list_llm = ["meta-llama/Meta-Llama-3-8B-Instruct", "mistralai/Mistral-7B-Instruct-v0.2"]
list_llm_simple = [os.path.basename(llm) for llm in list_llm]

# Load and split PDF documents
def load_doc(list_file_path):
    loaders = [PyPDFLoader(x) for x in list_file_path]
    pages = []
    for loader in loaders:
        pages.extend(loader.load())
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1024, chunk_overlap=64)
    doc_splits = text_splitter.split_documents(pages)
    return doc_splits

# Create vector database
def create_db(splits):
    embeddings = HuggingFaceEmbeddings()
    vectordb = FAISS.from_documents(splits, embeddings)
    return vectordb

# Initialize LLM chain
def initialize_llmchain(llm_model, temperature, max_tokens, top_k, vector_db):
    if llm_model == "meta-llama/Meta-Llama-3-8B-Instruct":
        llm = HuggingFaceEndpoint(
            repo_id=llm_model,
            huggingfacehub_api_token=api_token,
            temperature=temperature,
            max_new_tokens=max_tokens,
            top_k=top_k,
        )
    else:
        llm = HuggingFaceEndpoint(
            huggingfacehub_api_token=api_token,
            repo_id=llm_model,
            temperature=temperature,
            max_new_tokens=max_tokens,
            top_k=top_k,
        )
    
    memory = ConversationBufferMemory(
        memory_key="chat_history",
        output_key='answer',
        return_messages=True
    )

    retriever = vector_db.as_retriever()
    qa_chain = ConversationalRetrievalChain.from_llm(
        llm,
        retriever=retriever,
        chain_type="stuff",
        memory=memory,
        return_source_documents=True,
        verbose=False,
    )
    return qa_chain

# Function to handle chatbot responses
def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
    vector_db,
    llm_model,
):
    # Initialize LLM chain if not already initialized
    if not hasattr(respond, 'qa_chain'):
        respond.qa_chain = initialize_llmchain(llm_model, temperature, max_tokens, top_p, vector_db)

    # Format chat history
    formatted_chat_history = []
    for user_message, bot_message in history:
        formatted_chat_history.append(f"User: {user_message}")
        formatted_chat_history.append(f"Assistant: {bot_message}")
    formatted_chat_history.append(f"User: {message}")

    # Generate response using QA chain
    response = respond.qa_chain.invoke({"question": message, "chat_history": formatted_chat_history})
    response_answer = response["answer"]
    if response_answer.find("Helpful Answer:") != -1:
        response_answer = response_answer.split("Helpful Answer:")[-1]

    return response_answer

# CSS for styling the interface
css = """
body {
    background-color: #06688E; /* Dark background */
    color: white; /* Text color for better visibility */
}
.gr-button {
    background-color: #42B3CE !important; /* White button color */
    color: black !important; /* Black text for contrast */
    border: none !important;
    padding: 8px 16px !important;
    border-radius: 5px !important;
}
.gr-button:hover {
    background-color: #e0e0e0 !important; /* Slightly lighter button on hover */
}
.gr-slider-container {
    color: white !important; /* Slider labels in white */
}
"""

# Initialize database and LLM chain
def initialize_database_and_llm(list_file_obj, llm_option, max_tokens, temperature, top_p):
    list_file_path = [x.name for x in list_file_obj if x is not None]
    doc_splits = load_doc(list_file_path)
    vector_db = create_db(doc_splits)
    llm_name = list_llm[llm_option]
    return vector_db, llm_name

# Gradio interface
demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Files(file_count="multiple", file_types=["pdf"], label="Upload PDF documents", visible=False),
        gr.Radio(list_llm_simple, label="Available LLMs", value=list_llm_simple, visible=False),
        gr.Slider(minimum=128, maximum=9192, value=4096, step=128, label="Max new tokens", visible=False),
        gr.Slider(minimum=0.01, maximum=1.0, value=0.5, step=0.1, label="Temperature", visible=False),
        gr.Slider(minimum=1, maximum=10, value=3, step=1, label="Top-k", visible=False),
    ],
    css=css,
    title="RAG PDF Chatbot",
    description="Query your PDF documents using a Retrieval Augmented Generation (RAG) chatbot.",
)

# Preprocessing events
demo.preprocess(
    initialize_database_and_llm,
    inputs=["document", "llm_btn", "slider_maxtokens", "slider_temperature", "slider_topk"],
    outputs=["vector_db", "llm_model"],
    api_name="initialize",
)

if __name__ == "__main__":
    demo.launch(share=True)