keras-team/keras · error · ValueError

Illegal None dimension in {x} with shape {x.shape}

Error message

Illegal None dimension in {x} with shape {x.shape}

What it means

Error "Illegal None dimension in {x} with shape {x.shape}" thrown in keras-team/keras.

Source

Thrown at keras/src/backend/jax/export.py:159

    def _to_polymorphic_shape(self, struct, allow_none=True):
        if allow_none:
            # Generates unique names: a, b, ... z, aa, ab, ... az, ba, ... zz
            # for unknown non-batch dims. Defined here to be scope per endpoint.
            dim_names = itertools.chain(
                string.ascii_lowercase,
                itertools.starmap(
                    lambda a, b: a + b,
                    itertools.product(string.ascii_lowercase, repeat=2),
                ),
            )

        def convert_shape(x):
            poly_shape = []
            for index, dim in enumerate(list(x.shape)):
                if dim is not None:
                    poly_shape.append(str(dim))
                elif not allow_none:
                    raise ValueError(
                        f"Illegal None dimension in {x} with shape {x.shape}"
                    )
                elif index == 0:
                    poly_shape.append("batch")
                else:
                    poly_shape.append(next(dim_names))
            return f"({', '.join(poly_shape)})"

        return tree.map_structure(convert_shape, struct)

    def _check_device_compatible(self):
        from jax import default_backend as jax_device

        if (
            jax_device() == "gpu"
            and len(tf.config.list_physical_devices("GPU")) == 0
        ):
            warnings.warn(

View on GitHub (pinned to 7a34a03db6)

When it happens

Trigger: Thrown at keras/src/backend/jax/export.py:159 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/7316c0b934cd3a33. Report an issue: GitHub.