WZMIAOMIAO/deep-learning-for-image-processing · error · ValueError

The inverted_residual_setting should not be empty.

Error message

The inverted_residual_setting should not be empty.

What it means

MobileNetV3.__init__ requires a non-empty inverted_residual_setting list of block configs. An empty list (or None) fails the truthiness check and raises this ValueError, because the model would have no feature extractor layers at all.

Source

Thrown at pytorch_segmentation/lraspp/src/mobilenet_backbone.py:159

    def forward(self, x: Tensor) -> Tensor:
        result = self.block(x)
        if self.use_res_connect:
            result += x

        return result


class MobileNetV3(nn.Module):
    def __init__(self,
                 inverted_residual_setting: List[InvertedResidualConfig],
                 last_channel: int,
                 num_classes: int = 1000,
                 block: Optional[Callable[..., nn.Module]] = None,
                 norm_layer: Optional[Callable[..., nn.Module]] = None):
        super(MobileNetV3, self).__init__()

        if not inverted_residual_setting:
            raise ValueError("The inverted_residual_setting should not be empty.")
        elif not (isinstance(inverted_residual_setting, List) and
                  all([isinstance(s, InvertedResidualConfig) for s in inverted_residual_setting])):
            raise TypeError("The inverted_residual_setting should be List[InvertedResidualConfig]")

        if block is None:
            block = InvertedResidual

        if norm_layer is None:
            norm_layer = partial(nn.BatchNorm2d, eps=0.001, momentum=0.01)

        layers: List[nn.Module] = []

        # building first layer
        firstconv_output_c = inverted_residual_setting[0].input_c
        layers.append(ConvBNActivation(3,
                                       firstconv_output_c,
                                       kernel_size=3,
                                       stride=2,

View on GitHub (pinned to 1ec3fe6f37)

Solutions

1. Pass a non-empty list of InvertedResidualConfig, or use `mobilenet_v3_large()`/`mobilenet_v3_small()` helpers which construct it
2. Fix the config-generation logic so it yields at least one block
3. Add a fallback to the standard bneck_cfg presets when your list is empty

Example fix

// before
model = MobileNetV3(inverted_residual_setting=[], num_classes=21)  # ValueError
// after
model = mobilenet_v3_large(num_classes=21)
# or pass the standard preset:
model = MobileNetV3(inverted_residual_setting=bneck_cfg, num_classes=21)

cfgs = build_inverted_residual_setting(...)
assert isinstance(cfgs, list) and len(cfgs) > 0, "config list must be non-empty"

def usable_setting(cfgs):
    return bool(cfgs) and isinstance(cfgs, list)

try:
    model = MobileNetV3(inverted_residual_setting=cfgs, num_classes=nc)
except ValueError:
    logging.warning("empty setting; falling back to mobilenet_v3_large preset")
    model = mobilenet_v3_large(num_classes=nc)

When it happens

Trigger: `MobileNetV3(inverted_residual_setting=[], num_classes=...)`; passing None without a block preset; programmatically filtering the config list to empty (e.g. selecting only blocks with width >= X); calling with the wrong kwarg so the default list is never built.

Common situations: Writing custom model builders that generate configs dynamically and produce an empty list on some condition; forgetting to call the convenience constructors `mobilenet_v3_large/small` that build the config list; refactoring where the default kwarg was removed.

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

• illegal stride value.
• return_layers are not present in model
• return_layers are not present in model
• The inverted_residual_setting should be List[InvertedResidua
• return_layers are not present in model

AI-assisted analysis of WZMIAOMIAO/deep-learning-for-image-processing@1ec3fe6f37 (2026-08-30). Data as JSON: /api/errors/42b8c8d98987abd2. Report an issue: GitHub.