invoke-ai/InvokeAI · error · ValueError

Expected Main_SDNQ_Diffusers_FLUX_Config, got {type(config).

Error message

Expected Main_SDNQ_Diffusers_FLUX_Config, got {type(config).__name__}

What it means

In the SDNQ FLUX diffusers loader, after the single-file branch is excluded, the config must be a Main_SDNQ_Diffusers_FLUX_Config. If it is neither, _load_model raises ValueError with the actual class name, guarding against configs routed to the wrong loader.

Source

Thrown at invokeai/backend/model_manager/load/model_loaders/flux.py:1592

        config: AnyModelConfig,
        submodel_type: Optional[SubModelType] = None,
    ) -> AnyModel:
        logger.debug(
            "[SDNQ] FluxSDNQDiffusersModel._load_model called with config=%s, submodel=%s",
            type(config).__name__,
            submodel_type,
        )
        # Handle single-file SDNQ checkpoint (Main_SDNQ_FLUX_Config)
        if isinstance(config, Main_SDNQ_FLUX_Config):
            if submodel_type == SubModelType.Transformer:
                return self._load_sdnq_transformer_checkpoint(config)
            raise ValueError(
                f"Only Transformer submodels are supported for checkpoint format. Received: {submodel_type}"
            )

        # Handle diffusers-format SDNQ model (Main_SDNQ_Diffusers_FLUX_Config)
        if not isinstance(config, Main_SDNQ_Diffusers_FLUX_Config):
            raise ValueError(f"Expected Main_SDNQ_Diffusers_FLUX_Config, got {type(config).__name__}")

        if submodel_type is None:
            raise ValueError("A submodel type must be provided when loading main pipelines.")

        # Prefer the path discovery actually found. `model_index.json` names its components with
        # arbitrary keys, and identification records the key it saw — but reconstructing
        # `model_path / submodel_type.value` here assumes the key always equals the slot name. A
        # pipeline whose index calls its CLIP encoder something else is then discovered fine and
        # loaded from a folder that does not exist. Fall back to the conventional name when a config
        # predates submodel discovery.
        model_path = Path(config.path)
        submodel_path = resolve_submodel_path(config, submodel_type, model_path / submodel_type.value)

        # These branches build their modules by hand (`init_empty_weights` + `load_state_dict`)
        # rather than through `from_pretrained`, so they arrive in training mode — `put_in_eval_mode`
        # in `load_default._load_and_cache` is what puts every loaded model into inference mode.
        match submodel_type:
            case SubModelType.Transformer:

View on GitHub (pinned to 0b6a024f2f)

Solutions

1. Re-scan the model so identification assigns the correct config class (standard FLUX vs SDNQ diffusers).
2. Ensure the loader registry maps each config class to its proper loader.
3. Construct Main_SDNQ_Diffusers_FLUX_Config explicitly if writing custom import code.

Example fix

// before
model = SdnqFluxLoader()._load_model(plain_flux_config, SubModelType.Transformer)
// after
if not isinstance(plain_flux_config, Main_SDNQ_Diffusers_FLUX_Config):
    raise TypeError(f"{type(plain_flux_config).__name__} is not an SDNQ diffusers config; use the standard FLUX loader")
model = SdnqFluxLoader()._load_model(plain_flux_config, SubModelType.Transformer)

if not isinstance(config, (Main_SDNQ_FLUX_Config, Main_SDNQ_Diffusers_FLUX_Config)):
    raise TypeError(f"{type(config).__name__} is not an SDNQ FLUX config")

def is_sdnq_flux(config) -> bool:
    return isinstance(config, (Main_SDNQ_FLUX_Config, Main_SDNQ_Diffusers_FLUX_Config))

try:
    model = sdnq_loader._load_model(config, submodel_type)
except ValueError as e:
    if "Expected Main_SDNQ_Diffusers_FLUX_Config" in str(e):
        raise RuntimeError("Config is not SDNQ; route it to the standard FLUX loader") from e
    raise

When it happens

Trigger: A config that is neither Main_SDNQ_FLUX_Config nor Main_SDNQ_Diffusers_FLUX_Config reaches this SDNQ loader branch — e.g. a standard FLUX diffusers config misrouted by a customized registry or direct loader invocation.

Common situations: Manually invoking the SDNQ loader with a non-SDNQ diffusers config; editing model records' format fields so standard FLUX models get classified as SDNQ diffusers.

Related errors

• The selected FLUX model does not ship its own {', '.join(mis
• transformer is SDNQ-quantized; use Main_SDNQ_Diffusers_FLUX_
• Unexpected model config type: {type(config)}.
• Only Transformer submodels are supported for checkpoint form
• Unsupported submodel type: {submodel_type}

AI-assisted analysis of invoke-ai/InvokeAI@0b6a024f2f (2026-08-29). Data as JSON: /api/errors/45790a11a680322b. Report an issue: GitHub.