:py:mod:`box_embeddings.modules.volume.soft_volume`
===================================================

.. py:module:: box_embeddings.modules.volume.soft_volume


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

.. py:data:: eps
   

   

.. py:function:: soft_volume(box_tensor: box_embeddings.parameterizations.box_tensor.BoxTensor, volume_temperature: float = 1.0, log_scale: bool = True, scale: float = 1.0) -> torch.Tensor

   Volume of boxes. Uses softplus instead of ReLU/clamp

   :param box_tensor: input
   :param volume_temperature: 1/volume_temperature is the beta parameter for the softplus
   :param log_scale: Whether the output should be in log scale or not.
   :param scale: scale parameter. Should be left as 1.0 (default)
                 in most cases.

   :returns: Tensor of shape (..., ) when self has shape (..., 2, num_dims)

   :raises ValueError: if scale not in (0,1] or volume_temperature is 0.


.. py:class:: SoftVolume(log_scale: bool = True, volume_temperature: float = 1.0)

   Bases: :py:obj:`box_embeddings.modules.volume._volume._Volume`

   Softplus based volume.

   .. py:method:: forward(self, box_tensor: box_embeddings.parameterizations.box_tensor.BoxTensor) -> torch.Tensor

      Soft softplus base (instead of ReLU) volume.

      :param box_tensor: TODO

      :returns: torch.Tensor