tensorflow/models · error · NotImplementedError

The mode is not implemented: %s

Error message

The mode is not implemented: %s

What it means

Error "The mode is not implemented: %s" thrown in tensorflow/models.

Source

Thrown at official/modeling/multitask/train_lib.py:144

      controller.train_and_evaluate(
          train_steps=params.trainer.train_steps,
          eval_steps=params.trainer.validation_steps,
          eval_interval=params.trainer.validation_interval)
    elif mode == 'eval':
      controller.evaluate(steps=params.trainer.validation_steps)
    elif mode == 'continuous_eval':

      def timeout_fn():
        if evaluator.global_step.numpy() >= params.trainer.train_steps:
          return True
        return False

      controller.evaluate_continuously(
          steps=params.trainer.validation_steps,
          timeout=params.trainer.continuous_eval_timeout,
          timeout_fn=timeout_fn)
    else:
      raise NotImplementedError('The mode is not implemented: %s' % mode)

    if run_post_eval:
      return model, evaluator.evaluate(
          tf.convert_to_tensor(params.trainer.validation_steps))  # pytype: disable=bad-return-type  # typed-keras
    else:
      return model


def get_trainer(
    distribution_strategy: tf.distribute.Strategy,
    params: configs.MultiEvalExperimentConfig,
    task: multitask.MultiTask,
    model: base_model.MultiTaskBaseModel | tf_keras.Model,
) -> orbit.StandardTrainer:
  """Creates a multi-task trainer for the given task.

  Args:
    distribution_strategy: A distribution strategy.

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/modeling/multitask/train_lib.py:144 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/d101b7b220ff0e60. Report an issue: GitHub.