chat

Chat module for conversing with a RAPTOR RAG-based LLM in the terminal.

Chat

Chat(
    retriever: RetrieverLike,
    qa_model: QAModelLike,
    console: Console,
    history_file: str = ".bookacle-chat-history.txt",
    user_avatar: str = "👤",
)

A terminal-based chat interface for interacting with a RAPTOR RAG-based LLM.

Attributes:

• retriever (RetrieverLike) –

  Retriever to use for retrieving relevant context.

• qa_model (QAModelLike) –

  QA model to use for answering questions.

• console (Console) –

  Rich Console to use for displaying messages.

• history_file (str) –

  File to store chat history.

• user_avatar (str) –

  Avatar to use for the user in the chat UI.

Parameters:

• retriever (RetrieverLike) –

  Retriever to use for retrieving relevant context.

• qa_model (QAModelLike) –

  QA model to use for answering questions.

• console (Console) –

  Rich Console to use for displaying messages.

• history_file (str, default: '.bookacle-chat-history.txt' ) –

  File to store chat history. The file is created in the home directory.

• user_avatar (str, default: '👤' ) –

  Avatar to use for the user in the chat UI.

display_ai_msg_stream

display_ai_msg_stream(messages: Iterator[Message]) -> str

Display an AI message stream in the chat UI.

Parameters:

• messages (Iterator[Message]) –

  Stream of AI messages to display.

Returns:

• str –

  The complete message as a string.

invoke_qa_model

invoke_qa_model(
    tree: Tree,
    question: str,
    history: list[Message] | None = None,
    stream: bool = True,
    *args,
    **kwargs
) -> Message

Invoke the QA model to answer a question.

Parameters:

• tree (Tree) –

  RAPTOR tree that should be used for RAG.

• question (str) –

  The question to answer.

• history (list[Message] | None, default: None ) –

  Chat history.

• stream (bool, default: True ) –

  Whether to stream the AI response.

• **args (tuple[Any], default: () ) –

  Additional positional arguments to pass to the retriever.

• **kwargs (dict[str, Any], default: {} ) –

  Additional keyword arguments to pass to the retriever.

Returns:

• Message –

  The response from the QA model.

run

run(
    tree: Tree,
    initial_chat_message: str = "",
    system_prompt: str = "",
    stream: bool = True,
    *args,
    **kwargs
) -> None

Run the chat interface.

Parameters:

• tree (Tree) –

  RAPTOR tree that should be used for RAG.

• initial_chat_message (str, default: '' ) –

  Initial message to display in the chat.

• system_prompt (str, default: '' ) –

  System prompt that should be used for the QA model.

• stream (bool, default: True ) –

  Whether to stream the AI response.

• *args (tuple[Any], default: () ) –

  Additional positional arguments to pass to the retriever.

• **kwargs (dict[str, Any], default: {} ) –

  Additional keyword arguments to pass to the retriever.