tensorflow/models · error · ValueError

Invalid function key for the module: %s with key %s. Valid k

Error message

Invalid function key for the module: %s with key %s. Valid keys are: %s

What it means

Error "Invalid function key for the module: %s with key %s. Valid keys are: %s" thrown in tensorflow/models.

Source

Thrown at official/nlp/serving/serving_modules.py:162

  @tf.function
  def serve_text_examples(self, inputs) -> Dict[str, tf.Tensor]:
    name_to_features = {}
    for text_field in self.params.text_fields:
      name_to_features[text_field] = tf.io.FixedLenFeature([], tf.string)
    features = tf.io.parse_example(inputs, name_to_features)
    segments = [features[x] for x in self.params.text_fields]
    model_inputs = self._text_processor(segments)
    if self.params.inputs_only:
      return self.serve(input_word_ids=model_inputs["input_word_ids"])
    return self.serve(**model_inputs)

  def get_inference_signatures(self, function_keys: Dict[Text, Text]):
    signatures = {}
    valid_keys = ("serve", "serve_examples", "serve_text_examples")
    for func_key, signature_key in function_keys.items():
      if func_key not in valid_keys:
        raise ValueError("Invalid function key for the module: %s with key %s. "
                         "Valid keys are: %s" %
                         (self.__class__, func_key, valid_keys))
      if func_key == "serve":
        if self.params.inputs_only:
          signatures[signature_key] = self.serve.get_concrete_function(
              input_word_ids=tf.TensorSpec(
                  shape=[None, None], dtype=tf.int32, name="input_word_ids"))
        else:
          signatures[signature_key] = self.serve.get_concrete_function(
              input_word_ids=tf.TensorSpec(
                  shape=[None, None], dtype=tf.int32, name="input_word_ids"),
              input_mask=tf.TensorSpec(
                  shape=[None, None], dtype=tf.int32, name="input_mask"),
              input_type_ids=tf.TensorSpec(
                  shape=[None, None], dtype=tf.int32, name="input_type_ids"))
      if func_key == "serve_examples":
        signatures[signature_key] = self.serve_examples.get_concrete_function(
            tf.TensorSpec(shape=[None], dtype=tf.string, name="examples"))

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/nlp/serving/serving_modules.py:162 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/dcd3b1ffed567e88. Report an issue: GitHub.