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

expected stages_repeats as list of 3 positive ints

Error message

expected stages_repeats as list of 3 positive ints

What it means

ShuffleNetV2.__init__ asserts that stages_repeats is a list of exactly 3 entries (one repeat count per stage2/3/4) and raises ValueError with this message when len(stages_repeats) != 3. The architecture is fixed to three inverted-residual stages.

Source

Thrown at pytorch_classification/Test7_shufflenet/model.py:93

            out = torch.cat((x1, self.branch2(x2)), dim=1)
        else:
            out = torch.cat((self.branch1(x), self.branch2(x)), dim=1)

        out = channel_shuffle(out, 2)

        return out


class ShuffleNetV2(nn.Module):
    def __init__(self,
                 stages_repeats: List[int],
                 stages_out_channels: List[int],
                 num_classes: int = 1000,
                 inverted_residual: Callable[..., nn.Module] = InvertedResidual):
        super(ShuffleNetV2, self).__init__()

        if len(stages_repeats) != 3:
            raise ValueError("expected stages_repeats as list of 3 positive ints")
        if len(stages_out_channels) != 5:
            raise ValueError("expected stages_out_channels as list of 5 positive ints")
        self._stage_out_channels = stages_out_channels

        # input RGB image
        input_channels = 3
        output_channels = self._stage_out_channels[0]

        self.conv1 = nn.Sequential(
            nn.Conv2d(input_channels, output_channels, kernel_size=3, stride=2, padding=1, bias=False),
            nn.BatchNorm2d(output_channels),
            nn.ReLU(inplace=True)
        )
        input_channels = output_channels

        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)

        # Static annotations for mypy

View on GitHub (pinned to 1ec3fe6f37)

Solutions

1. Pass exactly 3 repeat counts, e.g. stages_repeats=[4, 8, 4] (1.0x preset).
2. Use the provided factory functions ShuffleNetV2_x0_5/x1_0/x1_5/x2_0(num_classes=...) instead of calling ShuffleNetV2 directly.
3. Swap arguments if you passed them in the wrong order — repeats is 3 elements, channels is 5.

Example fix

// before
model = ShuffleNetV2(stages_repeats=[4, 8, 4, 4], stages_out_channels=[24, 116, 232, 464, 1024], num_classes=5)
// after
model = ShuffleNetV2(stages_repeats=[4, 8, 4], stages_out_channels=[24, 116, 232, 464, 1024], num_classes=5)

assert len(stages_repeats) == 3, f"stages_repeats must have 3 ints, got {len(stages_repeats)}"
model = ShuffleNetV2(stages_repeats=stages_repeats, stages_out_channels=stages_out_channels, num_classes=n)

def valid_repeats(r) -> bool:
    return isinstance(r, list) and len(r) == 3 and all(isinstance(v, int) and v > 0 for v in r)

try:
    model = ShuffleNetV2(stages_repeats=rep, stages_out_channels=ch, num_classes=n)
except ValueError as e:
    if "stages_repeats" in str(e):
        model = ShuffleNetV2_x1_0(num_classes=n)
    else:
        raise

When it happens

Trigger: Calling ShuffleNetV2(stages_repeats=..., stages_out_channels=...) with a list of length other than 3 — e.g. [4] , [4, 8], [4, 8, 4, 4], or forgetting the argument so a wrong default/None is used.

Common situations: Building a custom variant of ShuffleNetV2 with extra or fewer stages; misreading the preset tables and copying a 4-element list; passing stages_out_channels by mistake into the stages_repeats slot.

Related errors

• expected stages_out_channels as list of 5 positive ints
• illegal stride value.
• The inverted_residual_setting should not be empty.
• illegal stride value.
• image: {} isn't RGB mode.

AI-assisted analysis of WZMIAOMIAO/deep-learning-for-image-processing@1ec3fe6f37 (2026-08-30). Data as JSON: /api/errors/3988f7c5e13ec620. Report an issue: GitHub.