run-llama/llama_index · error · ValueError

Unexpected kwargs: {kwargs}

Error message

Unexpected kwargs: {kwargs}

What it means

ChatMemoryBuffer.from_defaults has a fixed keyword signature and collects anything else into **kwargs; if kwargs is non-empty it raises ValueError listing them. This is a fail-fast guard against typos and renamed parameters silently being ignored (which would produce a misconfigured memory without any signal).

Source

Thrown at llama-index-core/llama_index/core/memory/chat_memory_buffer.py:65

        if tokenizer_fn is None:
            values["tokenizer_fn"] = get_tokenizer()

        return values

    @classmethod
    def from_defaults(
        cls,
        chat_history: Optional[List[ChatMessage]] = None,
        llm: Optional[LLM] = None,
        chat_store: Optional[BaseChatStore] = None,
        chat_store_key: str = DEFAULT_CHAT_STORE_KEY,
        token_limit: Optional[int] = None,
        tokenizer_fn: Optional[Callable[[str], List]] = None,
        **kwargs: Any,
    ) -> "ChatMemoryBuffer":
        """Create a chat memory buffer from an LLM."""
        if kwargs:
            raise ValueError(f"Unexpected kwargs: {kwargs}")

        if llm is not None:
            context_window = llm.metadata.context_window
            token_limit = token_limit or int(context_window * DEFAULT_TOKEN_LIMIT_RATIO)
        elif token_limit is None:
            token_limit = DEFAULT_TOKEN_LIMIT

        if chat_history is not None:
            chat_store = chat_store or SimpleChatStore()
            chat_store.set_messages(chat_store_key, chat_history)

        return cls(
            token_limit=token_limit,
            tokenizer_fn=tokenizer_fn or get_tokenizer(),
            chat_store=chat_store or SimpleChatStore(),
            chat_store_key=chat_store_key,
        )

View on GitHub (pinned to afd0fef371)

Solutions

1. Read the error message — it echoes the exact unexpected keys; fix or remove them.
2. Check the signature: inspect.signature(ChatMemoryBuffer.from_defaults) and use only accepted kwargs.
3. If the option genuinely exists on the class, construct ChatMemoryBuffer(...) directly with it after setting token_limit.

Example fix

# before
memory = ChatMemoryBuffer.from_defaults(token_limt=1000)  # typo -> ValueError
# after
memory = ChatMemoryBuffer.from_defaults(token_limit=1000)

import inspect

ALLOWED = set(inspect.signature(ChatMemoryBuffer.from_defaults).parameters)

def filter_memory_kwargs(kwargs: dict) -> dict:
    return {k: v for k, v in kwargs.items() if k in ALLOWED}

def are_valid_from_defaults_kwargs(kwargs: dict) -> bool:
    ALLOWED = {"chat_history", "llm", "chat_store", "chat_store_key", "token_limit", "tokenizer_fn"}
    return set(kwargs) <= ALLOWED

try:
    memory = ChatMemoryBuffer.from_defaults(**kwargs)
except ValueError as e:
    if "Unexpected kwargs" in str(e):
        kwargs.pop("token_limt", None)  # fix known typo
        memory = ChatMemoryBuffer.from_defaults(**kwargs)
    else:
        raise

When it happens

Trigger: Calling ChatMemoryBuffer.from_defaults(token_limt=1000) (typo), or passing parameters that belong to the class constructor but not to from_defaults (e.g. passing an argument removed/renamed in this version).

Common situations: Typos in keyword names; upgrading llama-index versions where a from_defaults parameter was renamed or removed; copy-pasting constructor kwargs into from_defaults.

Related errors

• Unexpected keyword arguments: {kwargs}
• Token limit must be set and greater than 0.
• Initial token count exceeds token limit
• Token limit for full-text messages must be set and greater t
• Max iterations of {max_iterations} reached! Either something

AI-assisted analysis of run-llama/llama_index@afd0fef371 (2026-08-15). Data as JSON: /api/errors/860dbba28e8e0cd6. Report an issue: GitHub.