keras-team/keras · error · ValueError

Pooling inputs's shape must be 3, 4 or 5, corresponding to 1

Error message

Pooling inputs's shape must be 3, 4 or 5, corresponding to 1D, 2D and 3D inputs. But received shape: {inputs.shape}.

What it means

Error "Pooling inputs's shape must be 3, 4 or 5, corresponding to 1D, 2D and 3D inputs. But received shape: {inputs.shape}." thrown in keras-team/keras.

Source

Thrown at keras/src/backend/tensorflow/nn.py:218

    logits_cumsum_safe = tf.where(support, logits_cumsum, 0.0)
    k = tf.reduce_sum(tf.cast(support, logits.dtype), axis=axis, keepdims=True)
    tau = (tf.reduce_sum(logits_cumsum_safe, axis=axis, keepdims=True) - 1) / k
    output = tf.maximum(logits - tau, 0.0)
    return output


def _transpose_spatial_inputs(inputs):
    num_spatial_dims = len(inputs.shape) - 2
    # Tensorflow pooling does not support `channels_first` format, so
    # we need to transpose to `channels_last` format.
    if num_spatial_dims == 1:
        inputs = tf.transpose(inputs, (0, 2, 1))
    elif num_spatial_dims == 2:
        inputs = tf.transpose(inputs, (0, 2, 3, 1))
    elif num_spatial_dims == 3:
        inputs = tf.transpose(inputs, (0, 2, 3, 4, 1))
    else:
        raise ValueError(
            "Pooling inputs's shape must be 3, 4 or 5, corresponding to 1D, 2D "
            f"and 3D inputs. But received shape: {inputs.shape}."
        )
    return inputs


def _transpose_spatial_outputs(outputs):
    # Undo the transpose in `_transpose_spatial_inputs`.
    num_spatial_dims = len(outputs.shape) - 2
    if num_spatial_dims == 1:
        outputs = tf.transpose(outputs, (0, 2, 1))
    elif num_spatial_dims == 2:
        outputs = tf.transpose(outputs, (0, 3, 1, 2))
    elif num_spatial_dims == 3:
        outputs = tf.transpose(outputs, (0, 4, 1, 2, 3))
    return outputs

View on GitHub (pinned to 7a34a03db6)

When it happens

Trigger: Thrown at keras/src/backend/tensorflow/nn.py:218 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/4645e94c59c93488. Report an issue: GitHub.