tensorflow/models · error · ValueError

The flag --experiment must be specified.

Error message

The flag --experiment must be specified.

What it means

Error "The flag --experiment must be specified." thrown in tensorflow/models.

Source

Thrown at official/core/train_utils.py:329

  separately.
  """

  def __init__(self, flags_obj):
    self._flags_obj = flags_obj

  def parse(self):
    """Overrall process of constructing Experiment config."""
    params = self.base_experiment()
    params = self.parse_config_file(params)
    params = self.parse_runtime(params)
    params = self.parse_data_service(params)
    params = self.parse_params_override(params)
    return params

  def base_experiment(self):
    """Get the base experiment config from --experiment field."""
    if self._flags_obj.experiment is None:
      raise ValueError('The flag --experiment must be specified.')
    return exp_factory.get_exp_config(self._flags_obj.experiment)

  def parse_config_file(self, params):
    """Override the configs of params from the config_file."""
    is_strict = True
    if isinstance(self._flags_obj, ParseConfigOptions):
      is_strict = self._flags_obj.strict_override
    for config_file in self._flags_obj.config_file or []:
      params = hyperparams.override_params_dict(
          params, config_file, is_strict=is_strict
      )
    return params

  def parse_runtime(self, params):
    """Override the runtime configs of params from flags."""
    # Override the TPU address and tf.data service address.
    params.override({
        'runtime': {

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/core/train_utils.py:329 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/36da5e1c625e64f7. Report an issue: GitHub.