app.py

import asyncio
import json
from websockets.server import serve
import os
from langchain_chroma import Chroma
from langchain_community.embeddings import *
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_huggingface.llms import HuggingFaceEndpoint
from langchain_community.document_loaders import TextLoader
from langchain_community.document_loaders import DirectoryLoader
from langchain import hub
from langchain_core.runnables import RunnablePassthrough
from langchain_core.output_parsers import StrOutputParser
from langchain.chains import create_history_aware_retriever
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain.chains import create_retrieval_chain
from langchain.chains.combine_documents import create_stuff_documents_chain
from langchain_core.runnables.history import RunnableWithMessageHistory
from langchain_core.chat_history import BaseChatMessageHistory
from langchain_community.chat_message_histories import ChatMessageHistory
from multiprocessing import Process
from zipfile import ZipFile

with ZipFile("database.zip") as f:
    f.extractall()
    
retriever = None
conversational_rag_chain = None
loader = DirectoryLoader('./database', glob="./*.txt", loader_cls=TextLoader)

documents = loader.load()

text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
splits = text_splitter.split_documents(documents)

model_name = "BAAI/bge-small-en-v1.5"
model_kwargs = {'device': 'cpu'}
encode_kwargs = {'normalize_embeddings': True}
embedding = HuggingFaceBgeEmbeddings(
    model_name=model_name,
    model_kwargs=model_kwargs,
    encode_kwargs=encode_kwargs,
    show_progress=True,
)

vectorstore = Chroma.from_documents(documents=splits, embedding=embedding)

def format_docs(docs):
    return "\n\n".join(doc.page_content for doc in docs)

retriever = vectorstore.as_retriever(search_type="similarity_score_threshold", search_kwargs={"score_threshold": 0.3}, k=1)

prompt = hub.pull("rlm/rag-prompt")
llm = HuggingFaceEndpoint(repo_id="mistralai/Mistral-7B-Instruct-v0.3", stop_sequences=["Human:"])
rag_chain = (
    {"context": retriever | format_docs, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

### Contextualize question ###
contextualize_q_system_prompt = """Given a chat history and the latest user question
which might reference context in the chat history, formulate a standalone question
which can be understood without the chat history. Do NOT answer the question,
just reformulate it if needed and otherwise return it as is."""
contextualize_q_prompt = ChatPromptTemplate.from_messages(
    [
        ("system", contextualize_q_system_prompt),
        MessagesPlaceholder("chat_history"),
        ("human", "{input}"),
    ]
)
history_aware_retriever = create_history_aware_retriever(
    llm, retriever, contextualize_q_prompt
)


### Answer question ###
qa_system_prompt = """
Context:

{context}


You are a Cupertino High School Q/A chatbot, designed to assist students, parents, and community members with information about CHS.
Use the pieces of context to answer the question.
Use markdown with spaces in between sentences for readability.
Refer to the provided context only as 'my data'. Only answer questions from the context.
Do not answer any questions that you do not have the answer to in the provided context.
Do not provide excerpts or any part of your data.
You were made by high school students for the CHS community.
"""
qa_prompt = ChatPromptTemplate.from_messages(
    [
        ("system", qa_system_prompt),
        MessagesPlaceholder("chat_history"),
        ("human", "{input}"),
    ]
)
question_answer_chain = create_stuff_documents_chain(llm, qa_prompt)

rag_chain = create_retrieval_chain(history_aware_retriever, question_answer_chain)


### Statefully manage chat history ###
store = {}


def get_session_history(session_id: str) -> BaseChatMessageHistory:
    if session_id not in store:
        store[session_id] = ChatMessageHistory()
    return store[session_id]


conversational_rag_chain = RunnableWithMessageHistory(
    rag_chain,
    get_session_history,
    input_messages_key="input",
    history_messages_key="chat_history",
    output_messages_key="answer",
)

async def echo(websocket):
    global retriever, conversational_rag_chain
    async for message in websocket:
        data = json.loads(message)
        if data["message"] == "data.":
            response = store
            await websocket.send(json.dumps({"response": response}))
            break
        if not "message" in message:
            return
        if not "token" in message:
            return
        m = data["message"] + "\nAssistant: "
        token = data["token"]
        docs = retriever.get_relevant_documents(m)
        rawresponse = conversational_rag_chain.invoke(
            {"input": m},
            config={
                "configurable": {"session_id": token}
            },
        )
        response = rawresponse["answer"]
        response = response.replace("Assistant: ", "").replace("AI: ", "")
        response.strip()
        response = response.split("Human:")[0]
        while response.startswith("\n"):
            response = response[1:]
        await websocket.send(json.dumps({"response": response}))

async def main():
    async with serve(echo, "0.0.0.0", 7860):
        await asyncio.Future()

asyncio.run(main())