tensorflow/models · error · ValueError

Input should have rank {}, got {}

Error message

Input should have rank {}, got {}

What it means

Error "Input should have rank {}, got {}" thrown in tensorflow/models.

Source

Thrown at official/vision/modeling/layers/nn_layers.py:819

  def get_config(self):
    """Returns a dictionary containing the config used for initialization."""
    config = {
        'keepdims': self._keepdims,
    }
    base_config = super(SpatialAveragePool3D, self).get_config()
    return dict(list(base_config.items()) + list(config.items()))

  def build(self, input_shape):
    """Builds the layer with the given input shape."""
    if tf_keras.backend.image_data_format() == 'channels_first':
      raise ValueError('"channels_first" mode is unsupported.')

    super(SpatialAveragePool3D, self).build(input_shape)

  def call(self, inputs, states=None, output_states: bool = False):
    """Calls the layer with the given inputs."""
    if inputs.shape.rank != 5:
      raise ValueError(
          'Input should have rank {}, got {}'.format(5, inputs.shape.rank))

    output = tf.reduce_mean(inputs, axis=(2, 3), keepdims=self._keepdims)
    return (output, states) if output_states else output


class CausalConvMixin:
  """Mixin class to implement CausalConv for `tf_keras.layers.Conv` layers."""

  @property
  def use_buffered_input(self) -> bool:
    return self._use_buffered_input

  @use_buffered_input.setter
  def use_buffered_input(self, variable: bool):
    self._use_buffered_input = variable

  def _compute_buffered_causal_padding(self,

View on GitHub (pinned to e006f5f0d5)

Solutions

1. Pass an input tensor whose rank equals the expected rank shown in the error.
2. Add or remove dimensions (e.g. with tf.expand_dims or tf.squeeze) to match the expected rank.

When it happens

Trigger: Thrown at official/vision/modeling/layers/nn_layers.py:819 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of tensorflow/models@e006f5f0d5 (2026-08-24). Data as JSON: /api/errors/27549a60d62d5430. Report an issue: GitHub.