tensorflow/models · error

Must set exactly one of vocab_file, sp_model_file

Error message

Must set exactly one of vocab_file, sp_model_file

What it means

Error "Must set exactly one of vocab_file, sp_model_file" thrown in tensorflow/models.

Source

Thrown at official/nlp/tools/export_tfhub_lib.py:343

  Args:
    vocab_file: The path to the wordpiece vocab file, or None.
    sp_model_file: The path to the sentencepiece model file, or None. Exactly
      one of vocab_file and sp_model_file must be set. This determines the type
      of tokenzer that is used.
    do_lower_case: Whether to do lower case.
    tokenize_with_offsets: Whether to include the .tokenize_with_offsets
      subobject.
    default_seq_length: The sequence length of preprocessing results from root
      callable. This is also the default sequence length for the
      bert_pack_inputs subobject.

  Returns:
    A tf_keras.Model object with several attached subobjects, suitable for
    saving as a preprocessing SavedModel.
  """
  # Select tokenizer.
  if bool(vocab_file) == bool(sp_model_file):
    raise ValueError("Must set exactly one of vocab_file, sp_model_file")
  if vocab_file:
    tokenize = layers.BertTokenizer(
        vocab_file=vocab_file,
        lower_case=do_lower_case,
        tokenize_with_offsets=tokenize_with_offsets)
  else:
    tokenize = layers.SentencepieceTokenizer(
        model_file_path=sp_model_file,
        lower_case=do_lower_case,
        strip_diacritics=True,  #  Strip diacritics to follow ALBERT model.
        tokenize_with_offsets=tokenize_with_offsets)

  # The root object of the preprocessing model can be called to do
  # one-shot preprocessing for users with single-sentence inputs.
  sentences = tf_keras.layers.Input(shape=(), dtype=tf.string, name="sentences")
  if tokenize_with_offsets:
    tokens, start_offsets, limit_offsets = tokenize(sentences)
  else:

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/nlp/tools/export_tfhub_lib.py:343 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/8fb7e70a0ba376fc. Report an issue: GitHub.