tensorflow/models · error · ValueError

Unknown ConvNet variant - {}

Error message

Unknown ConvNet variant - {}

What it means

Error "Unknown ConvNet variant - {}" thrown in tensorflow/models.

Source

Thrown at official/projects/deepmac_maskrcnn/modeling/heads/instance_heads.py:271

        self._conv_norms.append(bn_op(name=bn_name, **bn_kwargs))

    elif variant == 'hourglass20':
      logging.info('Using hourglass 20 network.')
      self._hourglass = hourglass_network.hourglass_20(
          self._config_dict['num_filters'], initial_downsample=False)

    elif variant == 'hourglass52':
      logging.info('Using hourglass 52 network.')
      self._hourglass = hourglass_network.hourglass_52(
          self._config_dict['num_filters'], initial_downsample=False)

    elif variant == 'hourglass100':
      logging.info('Using hourglass 100 network.')
      self._hourglass = hourglass_network.hourglass_100(
          self._config_dict['num_filters'], initial_downsample=False)

    else:
      raise ValueError('Unknown ConvNet variant - {}'.format(variant))

  def _call_convnet_variant(self, x):

    variant = self._config_dict['convnet_variant']
    if variant == 'default':
      for conv, bn in zip(self._convs, self._conv_norms):
        x = conv(x)
        x = bn(x)
        x = self._activation(x)
      return x
    elif variant == 'hourglass20':
      return self._hourglass(x)[-1]
    elif variant == 'hourglass52':
      return self._hourglass(x)[-1]
    elif variant == 'hourglass100':
      return self._hourglass(x)[-1]
    else:
      raise ValueError('Unknown ConvNet variant - {}'.format(variant))

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/deepmac_maskrcnn/modeling/heads/instance_heads.py:271 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/f88824083c328122. Report an issue: GitHub.