tensorflow/models · error · ValueError

model_dir must be set.

Error message

model_dir must be set.

What it means

Error "model_dir must be set." thrown in tensorflow/models.

Source

Thrown at official/legacy/detection/executor/distributed_executor.py:616

      model_dir: the folder for storing model checkpoints.
      eval_input_fn: (Optional) same type as train_input_fn. If not None, will
        trigger evaluting metric on eval data. If None, will not run eval step.
      eval_metric_fn: metric_fn for evaluation in test_step.
      total_steps: total training steps. If the current step reaches the
        total_steps, the evaluation loop will stop.
      eval_timeout: The maximum number of seconds to wait between checkpoints.
        If left as None, then the process will wait indefinitely. Used by
        tf.train.checkpoints_iterator.
      min_eval_interval: The minimum number of seconds between yielding
        checkpoints. Used by tf.train.checkpoints_iterator.
      summary_writer_fn: function to create summary writer.

    Returns:
      Eval metrics dictionary of the last checkpoint.
    """

    if not model_dir:
      raise ValueError('model_dir must be set.')

    def terminate_eval():
      tf.logging.info('Terminating eval after %d seconds of no checkpoints' %
                      eval_timeout)
      return True

    summary_writer = summary_writer_fn(model_dir, 'eval')
    self.eval_summary_writer = summary_writer.writer

    # Read checkpoints from the given model directory
    # until `eval_timeout` seconds elapses.
    for checkpoint_path in tf.train.checkpoints_iterator(
        model_dir,
        min_interval_secs=min_eval_interval,
        timeout=eval_timeout,
        timeout_fn=terminate_eval):
      eval_metric_result, current_step = self.evaluate_checkpoint(
          checkpoint_path=checkpoint_path,

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/legacy/detection/executor/distributed_executor.py:616 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/594da12a2a892dd7. Report an issue: GitHub.