:py:mod:`orcanet.lib.losses`
============================

.. py:module:: orcanet.lib.losses

.. autoapi-nested-parse::

   OrcaNet custom loss functions.

   ..
       !! processed by numpydoc !!


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


Functions
~~~~~~~~~

.. autoapisummary::

   orcanet.lib.losses.lkl_normal_tfp
   orcanet.lib.losses.lkl_normal



Attributes
~~~~~~~~~~

.. autoapisummary::

   orcanet.lib.losses.EPS


.. py:data:: EPS

   

.. py:function:: lkl_normal_tfp(y_true, y_pred)

   
   Normal distribution using tfp. See lkl_normal.
















   ..
       !! processed by numpydoc !!

.. py:function:: lkl_normal(y_true, y_pred)

   
   Negative normal log-likelihood function for n regression output neurons
   with clipping for increased stability.

   For stability in the case of outliers, the loss l_i is capped
   at a maximum of 10 * |pred_i - true_i| for each sample.

   :Parameters:

       **y_true** : tf.Tensor
           Shape (bs, 2, n) or (bs, 2).
           y_true[:, 0] is the label of shape (bs, n) (true), and y_true[:, 1]
           is not used (necessary as tf 2.1 requires y_true and y_pred to
           have same shape).

       **y_pred** : tf.Tensor
           Shape (bs, 2, n) or (bs, 2).
           The output of the network.
           y_pred[:, 0] is mu, and y_pred[:, 1] is sigma.














   ..
       !! processed by numpydoc !!