tensorflow/models · error · ValueError

Split {} not available.

Error message

Split {} not available.

What it means

Error "Split {} not available." thrown in tensorflow/models.

Source

Thrown at official/nlp/data/classifier_data_lib.py:717

  def get_dev_examples(self, data_dir):
    assert data_dir is None
    return self._create_examples(self.dev_split, "dev")

  def get_test_examples(self, data_dir):
    assert data_dir is None
    return self._create_examples(self.test_split, "test")

  def get_labels(self):
    return self._labels

  def get_processor_name(self):
    return "TFDS_" + self.dataset_name

  def _create_examples(self, split_name, set_type):
    """Creates examples for the training/dev/test sets."""
    if split_name not in self.dataset:
      raise ValueError("Split {} not available.".format(split_name))
    dataset = self.dataset[split_name].as_numpy_iterator()
    examples = []
    text_b, weight = None, None
    for i, example in enumerate(dataset):
      guid = "%s-%s" % (set_type, i)
      if set_type == "test":
        text_a = self.process_text_fn(example[self.test_text_key])
        if self.test_text_b_key:
          text_b = self.process_text_fn(example[self.test_text_b_key])
        label = self.test_label
      else:
        text_a = self.process_text_fn(example[self.text_key])
        if self.text_b_key:
          text_b = self.process_text_fn(example[self.text_b_key])
        label = self.label_type(example[self.label_key])  # pyrefly: ignore[not-callable]
        if self.skip_label is not None and label == self.skip_label:
          continue
      if self.weight_key:

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/nlp/data/classifier_data_lib.py:717 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/855458d764d5da2e. Report an issue: GitHub.