dicee.models.literal
====================

.. py:module:: dicee.models.literal


Classes
-------

.. autoapisummary::

   dicee.models.literal.LiteralEmbeddings


Module Contents
---------------

.. py:class:: LiteralEmbeddings(num_of_data_properties: int, embedding_dims: int, entity_embeddings: torch.tensor, dropout: float = 0.3, gate_residual=True, freeze_entity_embeddings=True)

   Bases: :py:obj:`torch.nn.Module`


   A model for learning and predicting numerical literals using pre-trained KGE.

   .. attribute:: num_of_data_properties

      Number of data properties (attributes).

      :type: int

   .. attribute:: embedding_dims

      Dimension of the embeddings.

      :type: int

   .. attribute:: entity_embeddings

      Pre-trained entity embeddings.

      :type: torch.tensor

   .. attribute:: dropout

      Dropout rate for regularization.

      :type: float

   .. attribute:: gate_residual

      Whether to use gated residual connections.

      :type: bool

   .. attribute:: freeze_entity_embeddings

      Whether to freeze the entity embeddings during training.

      :type: bool


   .. py:attribute:: embedding_dim


   .. py:attribute:: num_of_data_properties


   .. py:attribute:: hidden_dim


   .. py:attribute:: gate_residual
      :value: True



   .. py:attribute:: freeze_entity_embeddings
      :value: True



   .. py:attribute:: entity_embeddings


   .. py:attribute:: data_property_embeddings


   .. py:attribute:: fc


   .. py:attribute:: fc_out


   .. py:attribute:: dropout


   .. py:attribute:: gated_residual_proj


   .. py:attribute:: layer_norm


   .. py:method:: forward(entity_idx, attr_idx)

      :param entity_idx: Entity indices (batch).
      :type entity_idx: Tensor
      :param attr_idx: Attribute (Data property)  indices (batch).
      :type attr_idx: Tensor

      :returns: scalar predictions.
      :rtype: Tensor



   .. py:property:: device