Limit history in RAG chatbot to last two interactions, using StreamlitChatMessageHistory

[EnnedueJ]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

qa_prompt = ChatPromptTemplate.from_messages(
        [
            ("system", 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)

st_msg_history = StreamlitChatMessageHistory(key="chat_history")

    conversational_rag_chain = RunnableWithMessageHistory(
        rag_chain,
        lambda session_id: st_msg_history,
        input_messages_key="input",
        history_messages_key="chat_history",
        output_messages_key="answer",
    )

Description

I am using StreamlitChatMessageHistory in combo with RunnableWithMessageHistory. I want to limit the history included in the context of the conversational chain to the two last interactions in order to reduce the input tokens amount.

System Info

python version == 3.11.0