keras-team/keras · error · ValueError

adaptive_average_pool supports 1D, 2D, or 3D inputs only.

Error message

adaptive_average_pool supports 1D, 2D, or 3D inputs only.

What it means

Error "adaptive_average_pool supports 1D, 2D, or 3D inputs only." thrown in keras-team/keras.

Source

Thrown at keras/src/backend/tensorflow/nn.py:753

    pooled_w = tf.gather(combined_w, gather_w, axis=3)

    if data_format == "channels_first":
        pooled_w = tf.transpose(pooled_w, (0, 4, 1, 2, 3))

    return pooled_w


def adaptive_average_pool(inputs, output_size, data_format=None):
    data_format = backend.standardize_data_format(data_format)
    ndims = len(inputs.shape) - 2
    if ndims == 1:
        return _adaptive_average_pool1d(inputs, output_size, data_format)
    elif ndims == 2:
        return _adaptive_average_pool2d(inputs, output_size, data_format)
    elif ndims == 3:
        return _adaptive_average_pool3d(inputs, output_size, data_format)
    else:
        raise ValueError(
            "adaptive_average_pool supports 1D, 2D, or 3D inputs only."
        )


def adaptive_max_pool(inputs, output_size, data_format=None):
    data_format = backend.standardize_data_format(data_format)
    ndims = len(inputs.shape) - 2
    if ndims == 1:
        return _adaptive_max_pool1d(inputs, output_size, data_format)
    elif ndims == 2:
        return _adaptive_max_pool2d(inputs, output_size, data_format)
    elif ndims == 3:
        return _adaptive_max_pool3d(inputs, output_size, data_format)
    else:
        raise ValueError(
            "adaptive_max_pool supports 1D, 2D, or 3D inputs only."
        )

View on GitHub (pinned to 7a34a03db6)

When it happens

Trigger: Thrown at keras/src/backend/tensorflow/nn.py:753 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/cbdb977bc33c6438. Report an issue: GitHub.