:py:mod:`anomalib.models.reverse_distillation.torch_model`
==========================================================

.. py:module:: anomalib.models.reverse_distillation.torch_model

.. autoapi-nested-parse::

   PyTorch model for Reverse Distillation.



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

Classes
~~~~~~~

.. autoapisummary::

   anomalib.models.reverse_distillation.torch_model.ReverseDistillationModel




.. py:class:: ReverseDistillationModel(backbone: str, input_size: Tuple[int, int], layers: List[str], anomaly_map_mode: str)

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

   Reverse Distillation Model.

   :param backbone: Name of the backbone used for encoder and decoder
   :type backbone: str
   :param input_size: Size of input image
   :type input_size: Tuple[int, int]
   :param layers: Name of layers from which the features are extracted.
   :type layers: List[str]
   :param anomaly_map_mode: Mode used to generate anomaly map. Options are between ``multiply`` and ``add``.
   :type anomaly_map_mode: str

   .. py:method:: forward(self, images: torch.Tensor) -> Union[torch.Tensor, Tuple[List[torch.Tensor], List[torch.Tensor]]]

      Forward-pass images to the network.

      During the training mode the model extracts features from encoder and decoder networks.
      During evaluation mode, it returns the predicted anomaly map.

      :param images: Batch of images
      :type images: Tensor

      :returns:

                Encoder and decoder features in training mode,
                    else anomaly maps.
      :rtype: Union[Tensor, Tuple[List[Tensor],List[Tensor]]]