lenskit.flexmf.FlexMFImplicitConfig
===================================

.. py:class:: lenskit.flexmf.FlexMFImplicitConfig
   :canonical: lenskit.flexmf._implicit.FlexMFImplicitConfig

   Bases: :py:obj:`lenskit.flexmf._base.FlexMFConfigBase`


   Configuration for :class:`FlexMFImplicitScorer`.  It inherits base model
   options from :class:`FlexMFConfigBase`.

   :Stability: Experimental


   .. py:attribute:: preset
      :type:  Literal['bpr', 'warp', 'lightgcn'] | None
      :value: None


      Select preset defaults to mimic a particular model's original presentation.


   .. py:attribute:: loss
      :type:  ImplicitLoss
      :value: 'logistic'


      The loss to use for model training.


   .. py:attribute:: negative_strategy
      :type:  NegativeStrategy | None
      :value: None


      The negative sampling strategy.  The default is ``"misranked"`` for WARP
      loss and ``"uniform"`` for other losses.


   .. py:attribute:: negative_count
      :type:  pydantic.PositiveInt
      :value: 1


      The number of negative items to sample for each positive item in the
      training data.  With BPR loss, the positive item is compared to each
      negative item; with logistic loss, the positive item is treated once per
      learning round, so this setting effectively makes the model learn on _n_
      negatives per positive, rather than giving positive and negative examples
      equal weight.


   .. py:attribute:: positive_weight
      :type:  pydantic.PositiveFloat
      :value: 1.0


      A weighting multiplier to apply to the positive item's loss, to adjust the
      relative importance of positive and negative classifications.  Only applies
      to logistic loss.


   .. py:attribute:: user_bias
      :type:  bool | None
      :value: None


      Whether to learn a user bias term.  If unspecified, the default depends on
      the loss function (``False`` for pairwise and ``True`` for logistic).


   .. py:attribute:: item_bias
      :type:  bool
      :value: True


      Whether to learn an item bias term.


   .. py:attribute:: convolution_layers
      :type:  pydantic.NonNegativeInt
      :value: 0


      The number of LightGCN convolution layers to use.  0 (the default)
      configures for standard matrix factorization.


   .. py:method:: selected_negative_strategy()


   .. py:method:: apply_preset(data)
      :classmethod:



   .. py:method:: check_strategies()