tensorflow/models · error · ValueError

steps should be an Tensor. Python object may cause retracing

Error message

steps should be an Tensor. Python object may cause retracing.

What it means

Error "steps should be an Tensor. Python object may cause retracing." thrown in tensorflow/models.

Source

Thrown at official/legacy/detection/executor/distributed_executor.py:285

    Returns:
      The training step callable.
    """
    replicated_step = self._create_replicated_step(strategy, model, loss_fn,
                                                   optimizer, metric)

    @tf.function
    def train_step(iterator, num_steps):
      """Performs a distributed training step.

      Args:
        iterator: an iterator that yields input tensors.
        num_steps: the number of steps in the loop.

      Returns:
        The loss tensor.
      """
      if not isinstance(num_steps, tf.Tensor):
        raise ValueError('steps should be an Tensor. Python object may cause '
                         'retracing.')

      per_replica_losses = strategy.run(replicated_step, args=(next(iterator),))
      for _ in tf.range(num_steps - 1):
        per_replica_losses = strategy.run(
            replicated_step, args=(next(iterator),))

      # For reporting, we returns the mean of losses.
      losses = tf.nest.map_structure(
          lambda x: strategy.reduce(tf.distribute.ReduceOp.MEAN, x, axis=None),
          per_replica_losses)
      return losses

    return train_step

  def _create_test_step(self, strategy, model, metric):
    """Creates a distributed test step."""
    metrics = metrics_as_dict(metric)

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/legacy/detection/executor/distributed_executor.py:285 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/4438dbc5e4850ddb. Report an issue: GitHub.