Streamlit

Streamlit is an open-source Python library that makes it easy to create and share beautiful, custom web apps for machine learning and data science.

This notebook goes over how to store and use chat message history in a Streamlit app. StreamlitChatMessageHistory will store messages in Streamlit session state at the specified key=. The default key is "langchain_messages".

• Note, StreamlitChatMessageHistory only works when run in a Streamlit app.
• You may also be interested in StreamlitCallbackHandler for LangChain.
• For more on Streamlit check out their getting started documentation.

You can see the full app example running here, and more examples in github.com/langchain-ai/streamlit-agent.

from langchain.memory import StreamlitChatMessageHistory

history = StreamlitChatMessageHistory(key="chat_messages")

history.add_user_message("hi!")
history.add_ai_message("whats up?")

history.messages

You can integrate StreamlitChatMessageHistory into ConversationBufferMemory and chains or agents as usual. The history will be persisted across re-runs of the Streamlit app within a given user session. A given StreamlitChatMessageHistory will NOT be persisted or shared across user sessions.

from langchain.memory import ConversationBufferMemory
from langchain.memory.chat_message_histories import StreamlitChatMessageHistory

# Optionally, specify your own session_state key for storing messages
msgs = StreamlitChatMessageHistory(key="special_app_key")

memory = ConversationBufferMemory(memory_key="history", chat_memory=msgs)
if len(msgs.messages) == 0:
    msgs.add_ai_message("How can I help you?")

from langchain.chains import LLMChain
from langchain.llms import OpenAI
from langchain.prompts import PromptTemplate

template = """You are an AI chatbot having a conversation with a human.

{history}
Human: {human_input}
AI: """
prompt = PromptTemplate(input_variables=["history", "human_input"], template=template)

# Add the memory to an LLMChain as usual
llm_chain = LLMChain(llm=OpenAI(), prompt=prompt, memory=memory)

Conversational Streamlit apps will often re-draw each previous chat message on every re-run. This is easy to do by iterating through StreamlitChatMessageHistory.messages:

import streamlit as st

for msg in msgs.messages:
    st.chat_message(msg.type).write(msg.content)

if prompt := st.chat_input():
    st.chat_message("human").write(prompt)

    # As usual, new messages are added to StreamlitChatMessageHistory when the Chain is called.
    response = llm_chain.run(prompt)
    st.chat_message("ai").write(response)

View the final app.