tensorflow/models · error · ValueError

{self.task_config.model.interaction} is not supported it mu

Error message

 {self.task_config.model.interaction} is not supported it must be either 'dot' or 'cross' or 'multi_layer_dcn'.

What it means

Error " {self.task_config.model.interaction} is not supported it must be either 'dot' or 'cross' or 'multi_layer_dcn'." thrown in tensorflow/models.

Source

Thrown at official/recommendation/ranking/task.py:300

          skip_gather=True)
    elif self.task_config.model.interaction == 'cross':
      feature_interaction = tf_keras.Sequential([
          tf_keras.layers.Concatenate(),
          tfrs.layers.feature_interaction.Cross()
      ])
    elif self.task_config.model.interaction == 'multi_layer_dcn':
      feature_interaction = tf_keras.Sequential([
          tf_keras.layers.Concatenate(),
          tfrs.layers.feature_interaction.MultiLayerDCN(
              projection_dim=self.task_config.model.dcn_low_rank_dim,
              num_layers=self.task_config.model.dcn_num_layers,
              use_bias=self.task_config.model.dcn_use_bias,
              kernel_initializer=self.task_config.model.dcn_kernel_initializer,
              bias_initializer=self.task_config.model.dcn_bias_initializer,
          ),
      ])
    else:
      raise ValueError(
          f' {self.task_config.model.interaction} is not supported it must be'
          " either 'dot' or 'cross' or 'multi_layer_dcn'."
      )

    model = tfrs.experimental.models.Ranking(
        embedding_layer=embedding_layer,
        bottom_stack=tfrs.layers.blocks.MLP(
            units=self.task_config.model.bottom_mlp, final_activation='relu'
        ),
        feature_interaction=feature_interaction,
        top_stack=tfrs.layers.blocks.MLP(
            units=self.task_config.model.top_mlp, final_activation='sigmoid'
        ),
        concat_dense=self.task_config.model.concat_dense,
    )
    optimizer = tfrs.experimental.optimizers.CompositeOptimizer([
        (embedding_optimizer, lambda: model.embedding_trainable_variables),
        (dense_optimizer, lambda: model.dense_trainable_variables),

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/recommendation/ranking/task.py:300 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/395c7971e0c00915. Report an issue: GitHub.