keras-team/keras · error · ValueError

Layer {self.name} weight shape {variable.shape} is not compa

Error message

Layer {self.name} weight shape {variable.shape} is not compatible with provided weight shape {value.shape}.

What it means

Each value passed to set_weights() must have a shape exactly equal to the corresponding layer variable's shape. This error names the variable shape and the provided value shape so you can see which entry is wrong.

Source

Thrown at keras/src/layers/layer.py:786

            vars.extend(metric.variables)
        return vars

    def get_weights(self):
        """Return the values of `layer.weights` as a list of NumPy arrays."""
        return [v.numpy() for v in self.weights]

    def set_weights(self, weights):
        """Sets the values of `layer.weights` from a list of NumPy arrays."""
        layer_weights = self.weights
        if len(layer_weights) != len(weights):
            raise ValueError(
                f"You called `set_weights(weights)` on layer '{self.name}' "
                f"with a weight list of length {len(weights)}, but the layer "
                f"was expecting {len(layer_weights)} weights."
            )
        for variable, value in zip(layer_weights, weights):
            if variable.shape != value.shape:
                raise ValueError(
                    f"Layer {self.name} weight shape {variable.shape} "
                    "is not compatible with provided weight "
                    f"shape {value.shape}."
                )
            variable.assign(value)

    @property
    def dtype_policy(self):
        return self._dtype_policy

    @dtype_policy.setter
    def dtype_policy(self, value):
        policy = dtype_policies.get(value)
        if isinstance(self._dtype_policy, DTypePolicyMap) and self.path:
            if self.path in self._dtype_policy:
                del self._dtype_policy[self.path]
            self._dtype_policy[self.path] = policy
        else:

View on GitHub (pinned to 7a34a03db6)

Solutions

1. Rebuild the layer/model with the same input shape and units as when the weights were saved
2. Use model.load_weights() which matches by structure, or check variable.shape before assigning
3. Transpose/reshape the offending array to match variable.shape exactly (no broadcasting)

Example fix

# before
dense.set_weights([np.zeros((64, 32)), np.zeros((32,))])  # kernel is (32, 64)
# after
dense.set_weights([np.zeros((32, 64)), np.zeros((64,))])

for v, w in zip(model.weights, weights):
    assert v.shape == np.shape(w), (v.name, v.shape, np.shape(w))

def shapes_match(model, weights):
    return all(v.shape == np.shape(w) for v, w in zip(model.weights, weights))

try:
    model.set_weights(weights)
except ValueError as e:
    print('shape mismatch:', e)

When it happens

Trigger: Passing a (64, 32) array for a Dense kernel that is (32, 64); loading weights saved from a layer built on a different input dim; transposing arrays manually.

Common situations: Architecture mismatch (different input/features/units) between save and load; kernel vs transpose confusion when hand-converting weights from other frameworks; mixed old/new checkpoints.

Related errors

• You called `set_weights(weights)` on layer '{self.name}' wit
• Layer '{self.name}' expected {len(all_vars)} variables, but
• Expected rebatched data to have batch size 1. Received: shap
• Expected as input a list/tuple of 2 tensors. Received input_
• Expected the two input tensors to have identical shapes. Rec

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/897ff053d9d74aa9. Report an issue: GitHub.