tensorflow/models · error · TypeError

initial_cache element for key '%s' has dtype %s that does no

Error message

initial_cache element for key '%s' has dtype %s that does not match sampling_module's dtype of %s. Value: %s

What it means

Error "initial_cache element for key '%s' has dtype %s that does not match sampling_module's dtype of %s. Value: %s" thrown in tensorflow/models.

Source

Thrown at official/nlp/modeling/ops/sampling_module.py:264

      topk_seq = tf.tensor_scatter_nd_update(
          topk_seq, [[i + 1]], tf.expand_dims(tf.squeeze(topk_ids, -1), 0))
      topk_seq = tf.transpose(topk_seq, perm=[1, 0])
    else:
      topk_seq = tf.concat([alive_seq, topk_ids], axis=-1)
    return topk_seq, topk_log_probs, topk_ids, new_cache

  def _create_initial_state(
      self,
      initial_ids: tf.Tensor,
      initial_cache: Dict[str, tf.Tensor],
      batch_size: int,
      initial_log_probs: Optional[tf.Tensor] = None
  ) -> decoding_module.InitialState:
    """Return initial state dictionary and its shape invariants."""
    for key, value in initial_cache.items():
      for inner_value in tf.nest.flatten(value):
        if inner_value.dtype != self.dtype:
          raise TypeError(
              "initial_cache element for key '%s' has dtype %s that does not "
              "match sampling_module's dtype of %s. Value: %s" %
              (key, value.dtype.name, self.dtype.name, inner_value))

    # Current loop index (starts at 0)
    cur_index = tf.constant(0)

    # Alive sequence with shape [batch_size, 1]
    alive_seq = initial_ids
    alive_seq = tf.expand_dims(alive_seq, axis=-1)
    if self.padded_decode:
      alive_seq = tf.tile(alive_seq, [1, self.max_decode_length + 1])

    # Initial log probabilities with shape [batch_size, 1].
    if initial_log_probs is None:
      initial_log_probs = tf.constant([[0.]], dtype=self.dtype)
      alive_log_probs = tf.tile(initial_log_probs, [batch_size, 1])
    else:

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/nlp/modeling/ops/sampling_module.py:264 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/368b742048036455. Report an issue: GitHub.