conversation.py

import os

from langchain.embeddings import OpenAIEmbeddings, HuggingFaceInstructEmbeddings
from langchain.vectorstores import Chroma
from langchain.chat_models import ChatOpenAI
from langchain.chains import ConversationalRetrievalChain
from langchain.memory import ConversationBufferMemory
from langchain.llms import HuggingFaceHub

from dotenv import load_dotenv
load_dotenv()

def create_conversation() -> ConversationalRetrievalChain:

    persist_directory = 'db'

    embeddings = HuggingFaceInstructEmbeddings(model_name="hkunlp/instructor-xl")
    db = Chroma(
        persist_directory=persist_directory,
        embedding_function=embeddings
    )

    memory = ConversationBufferMemory(
        memory_key='chat_history',
        return_messages=False
    )

    qa = ConversationalRetrievalChain.from_llm(
        llm=HuggingFaceHub(repo_id="google/flan-t5-xxl", model_kwargs={"temperature":0.5, "max_length":512}),
        chain_type='stuff',
        retriever=db.as_retriever(),
        memory=memory,
        get_chat_history=lambda h: h,
        verbose=True
    )

    return qa