keras-team/keras · error · ValueError

All tensors passed to `ops.shape` must have a statically kno

Error message

All tensors passed to `ops.shape` must have a statically known rank. Received: x={x} with unknown rank.

What it means

Error "All tensors passed to `ops.shape` must have a statically known rank. Received: x={x} with unknown rank." thrown in keras-team/keras.

Source

Thrown at keras/src/backend/tensorflow/core.py:194

def is_tensor(x):
    return tf.is_tensor(x)


def shape(x):
    """Always return a tuple shape.

    `tf.shape` will return a `tf.Tensor`, which differs from the tuple return
    type on the torch and jax backends. We write our own method instead which
    always returns a tuple, with integer values when the shape is known, and
    tensor values when the shape is unknown (this is tf specific, as dynamic
    shapes do not apply in other backends).
    """
    if isinstance(x, KerasTensor):
        return x.shape
    if not tf.is_tensor(x):
        x = tf.convert_to_tensor(x)
    if x.shape == tf.TensorShape(None):
        raise ValueError(
            "All tensors passed to `ops.shape` must have a statically known "
            f"rank. Received: x={x} with unknown rank."
        )
    shape = x.shape.as_list()
    dynamic = tf.shape(x)
    for i in range(len(shape)):
        if shape[i] is None:
            try:
                shape[i] = dynamic[i]
            except:
                # With RaggedTensors, accessing a ragged dimension will fail,
                # we leave it as None.
                pass
    return tuple(shape)


def cast(x, dtype):
    dtype = standardize_dtype(dtype)

View on GitHub (pinned to 7a34a03db6)

When it happens

Trigger: Thrown at keras/src/backend/tensorflow/core.py:194 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/39399f4b88db584f. Report an issue: GitHub.