tensorflow/models · error · ValueError

The min_level should be >= decoder output level, but {} < {}

Error message

The min_level should be >= decoder output level, but {} < {}

What it means

Error "The min_level should be >= decoder output level, but {} < {}" thrown in tensorflow/models.

Source

Thrown at official/projects/pointpillars/modeling/heads.py:76

      output_specs: A dict of {level: TensorShape} pairs for the model output.
    """
    super(SSDHead, self).__init__(**kwargs)
    self._config_dict = {
        'num_classes': num_classes,
        'num_anchors_per_location': num_anchors_per_location,
        'num_params_per_anchor': num_params_per_anchor,
        'attribute_heads': attribute_heads,
        'min_level': min_level,
        'max_level': max_level,
        'kernel_regularizer': kernel_regularizer,
    }

    utils.assert_channels_last()

  def build(self, input_specs: Mapping[str, tf.TensorShape]):
    self._decoder_output_level = int(min(input_specs.keys()))
    if self._config_dict['min_level'] < self._decoder_output_level:
      raise ValueError('The min_level should be >= decoder output '
                       'level, but {} < {}'.format(
                           self._config_dict['min_level'],
                           self._decoder_output_level))

    # Multi-level convs.
    # Set num_filters as the one of decoder's output level.
    num_filters = input_specs[str(self._decoder_output_level)].as_list()[-1]
    self._convs = {}
    for level in range(self._decoder_output_level + 1,
                       self._config_dict['max_level'] + 1):
      self._convs[str(level)] = layers.ConvBlock(
          filters=num_filters,
          kernel_size=3,
          strides=2,
          kernel_regularizer=self._config_dict['kernel_regularizer'])

    # Detection convs, share weights across multi levels.
    self._classifier = tf_keras.layers.Conv2D(

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/pointpillars/modeling/heads.py:76 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/57326c688e06d89d. Report an issue: GitHub.