tensorflow/models · error · ValueError

Invalid `num_convs` {att_num_convs} for {att_name}.

Error message

Invalid `num_convs` {att_num_convs} for {att_name}.

What it means

Error "Invalid `num_convs` {att_num_convs} for {att_name}." thrown in tensorflow/models.

Source

Thrown at official/vision/modeling/heads/dense_prediction_heads.py:386

    return convs, norms, predictors

  def _build_attribute_net(self, conv_op, bn_op):
    self._att_predictors = {}
    self._att_convs = {}
    self._att_norms = {}

    for att_config, att_predictor_kwargs in zip(
        self._config_dict['attribute_heads'], self._attribute_kwargs  # pyrefly: ignore[bad-argument-type]
    ):
      att_name = att_config['name']
      att_num_convs = (
          att_config.get('num_convs') or self._config_dict['num_convs']
      )
      att_num_filters = (
          att_config.get('num_filters') or self._config_dict['num_filters']
      )
      if att_num_convs < 0:
        raise ValueError(f'Invalid `num_convs` {att_num_convs} for {att_name}.')
      if att_num_filters < 0:
        raise ValueError(
            f'Invalid `num_filters` {att_num_filters} for {att_name}.'
        )
      att_conv_kwargs = self._conv_kwargs.copy()
      att_conv_kwargs['filters'] = att_num_filters
      att_convs_i = []
      att_norms_i = []

      # Build conv and norm layers.
      for level in range(
          self._config_dict['min_level'], self._config_dict['max_level'] + 1
      ):
        this_level_att_norms = []
        for i in range(att_num_convs):
          if level == self._config_dict['min_level']:
            att_conv_name = '{}-conv_{}'.format(att_name, i)
            att_convs_i.append(conv_op(name=att_conv_name, **att_conv_kwargs))

View on GitHub (pinned to e006f5f0d5)

Solutions

1. Set the attribute's num_convs to a non-negative integer in the head config.
2. Remove num_convs for that attribute to use the default tower depth.

When it happens

Trigger: Thrown at official/vision/modeling/heads/dense_prediction_heads.py:386 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/e5cafa13af6ebe2f. Report an issue: GitHub.