microsoft/graphrag · error · ValueError

ModelConfig.mock_responses must be a non-empty list of embed

Error message

ModelConfig.mock_responses must be a non-empty list of embedding responses.

What it means

MockLLMEmbedding requires ModelConfig.mock_responses to be a non-empty list; it builds fake embedding vectors from it for testing. None, a non-list, or an empty list is rejected in __init__.

Source

Thrown at packages/graphrag-llm/graphrag_llm/embedding/mock_llm_embedding.py:48

    _mock_responses: list[float]
    _mock_index: int = 0

    def __init__(
        self,
        *,
        model_config: "ModelConfig",
        tokenizer: "Tokenizer",
        metrics_store: "MetricsStore",
        **kwargs: Any,
    ):
        """Initialize MockLLMEmbedding."""
        self._tokenizer = tokenizer
        self._metrics_store = metrics_store

        mock_responses = model_config.mock_responses
        if not isinstance(mock_responses, list) or len(mock_responses) == 0:
            msg = "ModelConfig.mock_responses must be a non-empty list of embedding responses."
            raise ValueError(msg)

        if not all(isinstance(resp, float) for resp in mock_responses):
            msg = "Each item in ModelConfig.mock_responses must be a float."
            raise ValueError(msg)

        self._mock_responses = mock_responses  # type: ignore

    def embedding(
        self, /, **kwargs: Unpack["LLMEmbeddingArgs"]
    ) -> "LLMEmbeddingResponse":
        """Sync embedding method."""
        input = kwargs.get("input")
        response = create_embedding_response(
            self._mock_responses, batch_size=len(input)
        )
        self._mock_index += 1
        return response

View on GitHub (pinned to f40e9a26ce)

Solutions

Set mock_responses to a non-empty list of floats, e.g. mock_responses: [0.1, 0.2, 0.3]
In tests, build the ModelConfig with explicit mock_responses rather than reusing a production config

Example fix

# before
ModelConfig(type=LLMProviderType.MockLLM, model="mock")
# after
ModelConfig(type=LLMProviderType.MockLLM, model="mock", mock_responses=[0.1, 0.2, 0.3])

Defensive patterns

Strategy: validation

Validate before calling

mr = getattr(model_config, "mock_responses", None)
if isinstance(mr, list) and len(mr) > 0 and all(isinstance(x, float) for x in mr):
    mock = MockLLMEmbedding(model_config, tokenizer, metrics_store)
else:
    raise ValueError("mock_responses must be a non-empty list of floats")

Type guard

def has_valid_mock_responses(cfg: ModelConfig) -> bool:
    mr = cfg.mock_responses
    return isinstance(mr, list) and len(mr) > 0 and all(isinstance(x, float) for x in mr)

Prevention

Centralize mock ModelConfig construction in a test fixture that always sets mock_responses=[0.1, 0.2, 0.3]
Never reuse production configs for mock-backed tests

When it happens

Trigger: Creating MockLLMEmbedding (or create_embedding with the mock type) where model_config.mock_responses is None, a scalar, or [].

Common situations: Writing unit tests with a mock model config but forgetting mock_responses; settings.yaml for tests omitting mock_responses; passing a string instead of a list.

Related errors

AI-assisted analysis of microsoft/graphrag@f40e9a26ce (2026-08-27). Data as JSON: /api/errors/e986c5a26af42da7. Report an issue: GitHub.