tensorflow/models · error · ValueError

Invalid call mode: %s

Error message

Invalid call mode: %s

What it means

Error "Invalid call mode: %s" thrown in tensorflow/models.

Source

Thrown at official/projects/nhnet/models.py:228

  def call(self, inputs, mode="train"):  # pyrefly: ignore[bad-override]
    """Implements call().

    Args:
      inputs: a dictionary of tensors.
      mode: string, an enum for mode, train/eval.

    Returns:
      logits, decode_output_ids, output_log_probs for training. top_decoded_ids
      for eval.
    """
    input_ids = inputs["input_ids"]
    input_mask = inputs["input_mask"]
    segment_ids = inputs["segment_ids"]
    all_encoder_outputs, _ = self.bert_layer(
        [input_ids, input_mask, segment_ids])

    if mode not in ("train", "eval", "predict"):
      raise ValueError("Invalid call mode: %s" % mode)
    encoder_decoder_attention_bias = decoder.get_attention_bias(
        input_ids,
        bias_type="single_cross",
        padding_value=self.params.pad_token_id)
    if mode == "train":
      self_attention_bias = decoder.get_attention_bias(
          inputs["target_ids"], bias_type="decoder_self")
      decoder_inputs = dict(
          attention_bias=encoder_decoder_attention_bias,
          all_encoder_outputs=all_encoder_outputs,
          target_ids=inputs["target_ids"],
          self_attention_bias=self_attention_bias)
      decoder_outputs = self.decoder_layer(decoder_inputs)
      return self.train_decode(decoder_outputs)

    batch_size = tf.shape(input_ids)[0]
    start_token_ids = tf.ones([batch_size],
                              tf.int32) * self.params.start_token_id

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/nhnet/models.py:228 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/1eb1d82ef11acdd7. Report an issue: GitHub.