app.py

import streamlit as st 
# from langchain_community.llms import HuggingFaceTextGenInference
import os
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
from langchain.schema import StrOutputParser

from custom_llm import CustomLLM, custom_chain_with_history


API_TOKEN = os.getenv('HF_INFER_API')


from typing import Optional

from langchain.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_community.chat_models import ChatAnthropic
from langchain_core.chat_history import BaseChatMessageHistory
from langchain.memory import ConversationBufferMemory
from langchain_core.runnables.history import RunnableWithMessageHistory


if 'memory' not in st.session_state:
    st.session_state['memory'] = ConversationBufferMemory(return_messages=True)
    st.session_state.memory.chat_memory.add_ai_message("Hello there! I'm AI assistant of Lintas Media Danawa. How can I help you today?")

if 'chain' not in st.session_state:
    st.session_state['chain'] = custom_chain_with_history(llm=CustomLLM(repo_id="mistralai/Mixtral-8x7B-Instruct-v0.1", model_type='text-generation', api_token=API_TOKEN, stop=["\n<|","<|"]), memory=st.session_state.memory)
    # st.session_state['chain'] = custom_chain_with_history(llm=InferenceClient("https://api-inference.huggingface.co/models/mistralai/Mixtral-8x7B-Instruct-v0.1", headers = {"Authorization": f"Bearer {API_TOKEN}"}, stream=True, max_new_tokens=512, temperature=0.01), memory=st.session_state.memory)
st.title("LMD Chatbot V2 Sample")
st.subheader("Knowledge-base from web scrapping and FAQ")

# Initialize chat history
if "messages" not in st.session_state:
    st.session_state.messages = []
    st.session_state.messages.append({"role":"assistant", "content":"Hello there! I'm AI assistant of Lintas Media Danawa. How can I help you today?"})

# Display chat messages from history on app rerun
for message in st.session_state.messages:
    with st.chat_message(message["role"]):
        st.markdown(message["content"])

# React to user input
if prompt := st.chat_input("Ask me anything.."):
    # Display user message in chat message container
    st.chat_message("User").markdown(prompt)
    # Add user message to chat history
    st.session_state.messages.append({"role": "User", "content": prompt})
    
    # full_response = st.session_state.chain.invoke(prompt).split("\n<|")[0]

    # Display assistant response in chat message container
    with st.chat_message("assistant"):
        message_placeholder = st.empty()
        full_response = ""
        for chunk in st.session_state.chain.stream(prompt):
            full_response += chunk + " "
            message_placeholder.markdown(full_response + " ")
            if full_response[-4:] == "\n<|":
                break
        # st.markdown(full_response)
    st.session_state.memory.save_context({"question":prompt}, {"output":full_response})
    st.session_state.memory.chat_memory.messages = st.session_state.memory.chat_memory.messages[-15:]
    # Add assistant response to chat history
    st.session_state.messages.append({"role": "assistant", "content": full_response})