tensorflow/models · error · RuntimeError

DefaultNBitActivationQuantizeConfig can only be used with `k

Error message

DefaultNBitActivationQuantizeConfig can only be used with `keras.layers.Activation`.

What it means

Error "DefaultNBitActivationQuantizeConfig can only be used with `keras.layers.Activation`." thrown in tensorflow/models.

Source

Thrown at official/projects/qat/vision/n_bit/configs.py:306

        num_bits_weight=num_bits_weight,
        num_bits_activation=num_bits_activation)


class DefaultNBitActivationQuantizeConfig(
    tfmot.quantization.keras.QuantizeConfig):
  """QuantizeConfig for keras.layers.Activation.

  `keras.layers.Activation` needs a separate `QuantizeConfig` since the
  decision to quantize depends on the specific activation type.
  """

  def __init__(self, num_bits_weight: int = 8, num_bits_activation: int = 8):
    self._num_bits_weight = num_bits_weight
    self._num_bits_activation = num_bits_activation

  def _assert_activation_layer(self, layer: Layer):
    if not isinstance(layer, tf_keras.layers.Activation):
      raise RuntimeError(
          'DefaultNBitActivationQuantizeConfig can only be used with '
          '`keras.layers.Activation`.')

  def get_weights_and_quantizers(
      self, layer: Layer) -> Sequence[WeightAndQuantizer]:
    """See base class."""
    self._assert_activation_layer(layer)
    return []

  def get_activations_and_quantizers(
      self, layer: Layer) -> Sequence[ActivationAndQuantizer]:
    """See base class."""
    self._assert_activation_layer(layer)
    return []

  def set_quantize_weights(
      self,
      layer: Layer,

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/qat/vision/n_bit/configs.py:306 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/903ebe98c80154a8. Report an issue: GitHub.