:py:mod:`orcanet.model_builder`
===============================

.. py:module:: orcanet.model_builder

.. autoapi-nested-parse::

   Scripts for making specific models.

   ..
       !! processed by numpydoc !!


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

Classes
~~~~~~~

.. autoapisummary::

   orcanet.model_builder.ModelBuilder



Functions
~~~~~~~~~

.. autoapisummary::

   orcanet.model_builder.get_adam
   orcanet.model_builder.get_sgd



.. py:class:: ModelBuilder(model_file, **custom_blocks)

   
   Build and compile a keras model from a toml file, using OrcaNet building blocks.

   The input of the model can match the dimensions of the input
   data given to the Organizer taking into account the sample
   modifier.













   :Attributes:

       **configs** : list
           List with keywords for building each layer block in the model.

       **defaults** : dict
           Default values for the layer blocks in the model.

       **optimizer** : str or Optimizer
           Optimizer for training the model. Can be a string like "adam" (or
           "keras:adam" for the default keras variant), or an object derived
           from ks.optimizers.Optimizer.

       **compile_opt** : dict
           Keys: Names of the output layers of the model.
           Values: Loss function, optionally weight and optionally metric of
           each output layer.
           Format: { layer_name : { loss_function:, weight:, metrics: } }
           The loss_function is a string or a function, the weight is a float
           and metrics is a list of functions/strings.

       **optimizer_args** : dict, optional
           Kwargs for the optimizer. Not used when an optimizer object is given.

       **input_opts** : dict
           Specify options for the input of the model.

   .. rubric:: Methods



   ====================  ==========
              **build**  Build the network using an instance of Organizer.  
   **build_with_input**  Build the network without an Organizer, just using given input shapes.  
            **compile**  Compile a model with the optimizer settings given in the model_file.  
   ====================  ==========

   ..
       !! processed by numpydoc !!
   .. py:method:: build(orga, log_comp_opts=False, verbose=False)

      
      Build the network using an instance of Organizer.

      Input layers will be adapted to the input files in the organizer.
      Can also add the matching modifiers and custom objects to the orga.

      :Parameters:

          **orga** : orcanet.core.Organizer
              Contains all the configurable options in the OrcaNet scripts.

          **log_comp_opts** : bool
              If the info used for the compilation of the model should be
              logged to the log.txt.

          **verbose** : bool
              Print info about the building process?

      :Returns:

          **model** : keras model
              The network.













      ..
          !! processed by numpydoc !!

   .. py:method:: build_with_input(input_shapes, compile_model=True, custom_objects=None, verbose=False)

      
      Build the network with given input shapes.


      :Parameters:

          **input_shapes** : dict
              Keys: Name of the inputs of the model.
              Values: Their shape without the batchsize.

          **compile_model** : bool
              Compile the model?

          **custom_objects** : dict, optional
              Custom objects to use during compiling.

          **verbose** : bool
              Print info about the building process?

      :Returns:

          **model** : ks.Model
              The network.













      ..
          !! processed by numpydoc !!

   .. py:method:: compile_model(model, custom_objects=None)

      
      Compile a model with the optimizer settings given as the attributes.


      :Parameters:

          **model** : ks.model
              A keras model.

          **custom_objects** : dict or None
              Maps names (strings) to custom loss functions.

      :Returns:

          **model** : keras model
              The compiled (or recompiled) keras model.













      ..
          !! processed by numpydoc !!

   .. py:method:: log_model_properties(orga)

      
      Writes the compile_opt config to the full log file.
















      ..
          !! processed by numpydoc !!


.. py:function:: get_adam(beta_1=0.9, beta_2=0.999, epsilon=0.1, decay=0.0, **kwargs)


.. py:function:: get_sgd(momentum=0.9, decay=0, nesterov=True, **kwargs)