[MRG] Fix docstring for the regulariation parameter of DA loss (#230)

* fix docstring

* fix docstring

---------

Co-authored-by: Antoine Collas <[email protected]>

Author: tgnassou

skada/deep/_adversarial.py

@@ -100,7 +100,7 @@ def DANN(
         The name of the module's layer whose outputs are
         collected during the training.
     reg : float, default=1
-        Regularization parameter.
+        Regularization parameter for DA loss.
     domain_classifier : torch module, default=None
         A PyTorch :class:`~torch.nn.Module` used to classify the
         domain. If None, a domain classifier is created following [1]_.
@@ -344,7 +344,7 @@ def CDAN(
         The name of the module's layer whose outputs are
         collected during the training.
     reg : float, default=1
-        Regularization parameter.
+        Regularization parameter for DA loss.
     max_features : int, default=4096
         Maximum size of the input for the domain classifier.
         4096 is the largest number of units in typical deep network

skada/deep/_divergence.py

@@ -63,7 +63,7 @@ def DeepCoral(module, layer_name, reg=1, base_criterion=None, **kwargs):
         The name of the module's layer whose outputs are
         collected during the training for the adaptation.
     reg : float, optional (default=1)
-        The regularization parameter of the covariance estimator.
+        Regularization parameter for DA loss.
     base_criterion : torch criterion (class)
         The base criterion used to compute the loss with source
         labels. If None, the default is `torch.nn.CrossEntropyLoss`.
@@ -142,7 +142,7 @@ def DAN(module, layer_name, reg=1, sigmas=None, base_criterion=None, **kwargs):
         The name of the module's layer whose outputs are
         collected during the training for the adaptation.
     reg : float, optional (default=1)
-        The regularization parameter of the covariance estimator.
+        Regularization parameter for DA loss.
     sigmas : array-like, optional (default=None)
         The sigmas for the Gaussian kernel.
     base_criterion : torch criterion (class)

skada/deep/_optimal_transport.py

@@ -92,7 +92,7 @@ def DeepJDOT(
         The name of the module's layer whose outputs are
         collected during the training for the adaptation.
     reg : float, default=1
-        Regularization parameter.
+        Regularization parameter for DA loss.
     reg_cl : float, default=1
         Class distance term regularization parameter.
     base_criterion : torch criterion (class)

skada/deep/base.py

@@ -37,7 +37,7 @@ class DomainAwareCriterion(torch.nn.Module):
         The initialized criterion (loss) used to compute the
         loss to reduce the divergence between domains.
     reg: float, default=1
-        Regularization parameter.
+        Regularization parameter for DA loss.
     reduction: str, default='mean'
         Specifies the reduction to apply to the output: 'none' | 'mean' | 'sum'.
         'none': no reduction will be applied,