lenskit.tuning.spec.SearchConfig
================================

.. py:class:: lenskit.tuning.spec.SearchConfig
   

   Bases: :py:obj:`pydantic.BaseModel`


   Configuration options for the hyperparameter search.


   .. py:attribute:: method
      :type:  Literal['optuna', 'hyperopt', 'random'] | None
      :value: None


      The search method to use.


   .. py:attribute:: max_points
      :type:  int | None
      :value: None


      The maximum number of points to try.


   .. py:attribute:: default_points
      :type:  int
      :value: 60


      The default number of search points, if not limited by a maximum configuration.


   .. py:attribute:: max_epochs
      :type:  int
      :value: 100


      The maximum number of epochs to use in iterative training.


   .. py:attribute:: min_epochs
      :type:  int
      :value: 3


      The minimum number of epochs for iterative training.


   .. py:attribute:: metric
      :type:  str | None
      :value: None


      The metric to use.


   .. py:attribute:: list_length
      :type:  int | None
      :value: None


      The length of recommendation lists to use.


   .. py:attribute:: num_cpus
      :type:  int | Literal['threads', 'backend-threads', 'all-threads']
      :value: 'threads'


      The number of CPUs to request from Ray Tune.


   .. py:attribute:: num_gpus
      :type:  int | float
      :value: 0


      The number of GPUs to requrest from Ray Tune.


   .. py:attribute:: checkpoint_iters
      :type:  int
      :value: 2


      The frequency for saving checkpoints.


   .. py:method:: update_max_points(n)

      Limit the search points to a new maximum, if it exceeds the current maximum.



   .. py:method:: num_search_points()

      Get the number of search points to use.