tensorflow/models · error · ValueError

Unexpected inputs type to {self.__class__}.

Error message

Unexpected inputs type to {self.__class__}.

What it means

Error "Unexpected inputs type to {self.__class__}." thrown in tensorflow/models.

Source

Thrown at official/projects/perceiver/modeling/networks/positional_decoder.py:114

    """Return decoded output of latent vector.

    Uses the positional encoding as query for the decoder and uses the
    `latent_output` as key-value for the decoder.
    Args:
      inputs:
        A `Dict[Text, tf_keras.Input]` with `latent_output` and
        `input_mask`. The shape of `latent_output` is shape
        `(z_index_dim, d_latents)` with dtype `tf.float32` and `input_mask` is
        shape `(None)` with dtype `tf.int32`.
      training:
        Flag to indicate training status. Default is `None`. It is passed to
        the decoder as is.

    Returns:
      `Dict[Text, tf.Tensor]` decoded `sequence_output` of a latent vector.
    """
    if not isinstance(inputs, dict):
      raise ValueError(f'Unexpected inputs type to {self.__class__}.')

    latent_output = inputs['latent_output']
    query_mask = inputs.get('input_mask')
    decoder_query = self._output_pos_enc(tf.ones(
        (tf.shape(latent_output)[0], self._output_index_dim, self._d_model),
        dtype=latent_output.dtype))
    z = latent_output

    sequence_output = self._decoder(
        [decoder_query, z],
        query_mask=query_mask,
        training=training)
    return dict(sequence_output=sequence_output)

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/perceiver/modeling/networks/positional_decoder.py:114 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/ef97433275434c60. Report an issue: GitHub.