keras-team/keras · error · ValueError

The `{name}` argument should be a number (or a list of two n

Error message

The `{name}` argument should be a number (or a list of two numbers) Received: {name}={factor}

What it means

RandomGaussianBlur._set_kernel_size validates the kernel_size argument in __init__. A kernel size must be a single odd integer or a 2-element sequence of odd integers (one per axis). When a sequence of any other length is given, this ValueError fires at construction.

Source

Thrown at keras/src/layers/preprocessing/image_preprocessing/random_gaussian_blur.py:77

    ):
        super().__init__(data_format=data_format, **kwargs)
        self._set_factor(factor)
        self.kernel_size = self._set_kernel_size(kernel_size, "kernel_size")
        self.sigma = self._set_factor_by_name(sigma, "sigma")
        self.value_range = value_range
        self.seed = seed
        self.generator = SeedGenerator(seed)

    def _set_kernel_size(self, factor, name):
        error_msg = f"{name} must be an odd number. Received: {name}={factor}"
        if isinstance(factor, (tuple, list)):
            if len(factor) != 2:
                error_msg = (
                    f"The `{name}` argument should be a number "
                    "(or a list of two numbers) "
                    f"Received: {name}={factor}"
                )
                raise ValueError(error_msg)
            if (factor[0] % 2 == 0) or (factor[1] % 2 == 0):
                raise ValueError(error_msg)
            lower, upper = factor
        elif isinstance(factor, (int, float)):
            if factor % 2 == 0:
                raise ValueError(error_msg)
            lower, upper = factor, factor
        else:
            raise ValueError(error_msg)

        return lower, upper

    def _set_factor_by_name(self, factor, name):
        error_msg = (
            f"The `{name}` argument should be a number "
            "(or a list of two numbers) "
            "in the range "
            f"[{self._FACTOR_BOUNDS[0]}, {self._FACTOR_BOUNDS[1]}]. "

View on GitHub (pinned to 7a34a03db6)

Solutions

Pass kernel_size=[kh, kw] with both odd, e.g. [3, 3]
Or a single odd integer, e.g. kernel_size=5
Sanitize config lists to length 1 or 2 before construction

Example fix

# before
layers.RandomGaussianBlur(kernel_size=[3, 3, 3])
# after
layers.RandomGaussianBlur(kernel_size=[3, 3])

Defensive patterns

Strategy: validation

Validate before calling

ks = [3, 3]
assert isinstance(ks, int) or len(ks) == 2, "kernel_size must be int or 2-element list"

Type guard

def is_valid_kernel_size(v) -> bool:
    return isinstance(v, int) or (isinstance(v, (tuple, list)) and len(v) == 2)

Try / catch

try:
    layer = RandomGaussianBlur(kernel_size=ks)
except ValueError:
    layer = RandomGaussianBlur(kernel_size=3)

Prevention

Use scalar kernel_size unless anisotropic blur is needed
Schema-check config lists for length 1 or 2

When it happens

Trigger: kernel_size=[3, 3, 3] (three entries), kernel_size=[3], or kernel_size=[] passed to layers.RandomGaussianBlur().

Common situations: Configs that add a channel dimension to kernel sizes; reusing a 3-element kernel spec from a 3D-conv setting; YAML list typos.

Related errors

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/444df9ed61f4cd0b. Report an issue: GitHub.