chatbot.py

import os
import asyncio
from fastapi import FastAPI, HTTPException, Depends
from fastapi.security import OAuth2PasswordBearer
from pydantic import BaseModel
from langchain.chains import RetrievalQA
from langchain.prompts import PromptTemplate
from langchain.vectorstores import FAISS
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.memory import ConversationBufferMemory
from langchain.llms import HuggingFacePipeline
from langchain.document_loaders import TextLoader, DataFrameLoader
from langchain_community.vectorstores import FAISS
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.document_loaders import TextLoader, DataFrameLoader
from langchain_community.llms import HuggingFacePipeline
from huggingface_hub import login

# Log in using the token
login(token=os.getenv("HUGGINGFACE_HUB_TOKEN"))

import pandas as pd
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline

# Step 1: Configure LangSmith Observability
os.environ["LANGCHAIN_TRACING_V2"] = "true"
os.environ["LANGCHAIN_ENDPOINT"] = "https://api.smith.langchain.com"
os.environ["LANGCHAIN_API_KEY"] = "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"
os.environ["LANGCHAIN_PROJECT"] = "yotta-vm-chatbot"

# Step 2: Llama 2 Chat Model Setup
def load_llama2_chat_model():
    model_name = "meta-llama/Llama-2-7b-chat-hf"
    print("Loading Llama 2 Chat model...")
    tokenizer = AutoTokenizer.from_pretrained(model_name)
    model = AutoModelForCausalLM.from_pretrained(
        model_name,
        device_map="auto",
        torch_dtype="float16"
    )
    pipeline_model = pipeline("text-generation", model=model, tokenizer=tokenizer)
    return HuggingFacePipeline(pipeline=pipeline_model)

llama_model = load_llama2_chat_model()

# Step 3: Load Excel Data
def fetch_excel_data(file_path):
    """
    Fetch data from a local Excel file and prepare documents for vector store.
    """
    print("Loading data from Excel file...")
    df = pd.read_excel(file_path)
    loader = DataFrameLoader(df, page_content_column="Description", metadata_columns=["Title"])
    documents = loader.load()
    return documents

# Step 4: Vector Store with Excel Data
def update_vector_store(file_path):
    """
    Load data from Excel and update the vector store.
    """
    documents = fetch_excel_data(file_path)
    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    vector_store = FAISS.from_documents(documents, embeddings)
    return vector_store

# Path to the Excel file
excel_file_path = "certificate_details_chatbot_2.xlsx"
vector_store = update_vector_store(excel_file_path)

# Step 5: RAG Workflow with Memory
retriever = vector_store.as_retriever()
memory = ConversationBufferMemory()

qa_chain = RetrievalQA.from_chain_type(
    llm=llama_model,
    retriever=retriever,
    memory=memory,
    return_source_documents=True,
    chain_type_kwargs={
        "prompt": PromptTemplate(
            input_variables=["context", "question"],
            template="Use the following context to answer the question:\n{context}\n\nQuestion: {question}\nAnswer:"
        )
    }
)


# Step 6: Periodic Sync with Excel
async def periodic_sync(interval: int = 3600):
    """
    Periodically fetch new data from the Excel file and update the vector store.
    """
    while True:
        try:
            update_vector_store(excel_file_path)
            print("Vector store updated with the latest Excel data.")
        except Exception as e:
            print(f"Error updating vector store: {str(e)}")
        await asyncio.sleep(interval)


# Step 8: FastAPI Deployment
app = FastAPI()

class QueryRequest(BaseModel):
    query: str

@app.get("/")
def root():
    return {"message": "Welcome to the Excel-based Chatbot with RAG and Llama Integration!"}

@app.post("/query")
async def query(request: QueryRequest):
    try:
        response = qa_chain({"query": request.query})
        return {
            "answer": response['result'],
            "source_documents": [
                {"page_content": doc.page_content, "metadata": doc.metadata} 
                for doc in response["source_documents"]
            ]
        }
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Error processing the query: {str(e)}")

@app.post("/token")
async def token():
    return {"access_token": "secure_token_123", "token_type": "bearer"}

@app.on_event("startup")
async def start_background_tasks():
    asyncio.create_task(periodic_sync())

# Step 9: Interface and Deployment
@app.get("/interface")
def interface():
    """Return a simple HTML interface for interacting with the chatbot."""
    return {
        "html": """
        <html>
        <head><title>Chatbot Interface</title></head>
        <body>
            <h1>Chat with the Bot</h1>
            <form method="post" action="/query">
                <label for="query">Enter your query:</label><br>
                <input type="text" id="query" name="query"/><br><br>
                <button type="submit">Submit</button>
            </form>
        </body>
        </html>
        """
    }

if __name__ == "__main__":
    import uvicorn
    print("Starting the chatbot server...")
    uvicorn.run(app, host="0.0.0.0", port=8000)