more lint

privacy/bolton/optimizers.py

@@ -11,7 +11,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-"""Bolton Optimizer for bolton method"""
+"""Bolton Optimizer for bolton method."""
 
 from __future__ import absolute_import
 from __future__ import division
@@ -28,8 +28,10 @@ _accepted_distributions = ['laplace']  # implemented distributions for noising
 class GammaBetaDecreasingStep(
     optimizer_v2.learning_rate_schedule.LearningRateSchedule):
   """Computes LR as minimum of 1/beta and 1/(gamma * step) at each step.
-  A required step for privacy guarantees.
+
+  This is a required step for privacy guarantees.
   """
+
   def __init__(self):
     self.is_init = False
     self.beta = None
@@ -37,8 +39,10 @@ class GammaBetaDecreasingStep(
 
   def __call__(self, step):
     """Computes and returns the learning rate.
+
       Args:
         step: the current iteration number
+
       Returns:
         decayed learning rate to minimum of 1/beta and 1/(gamma * step) as per
         the Bolton privacy requirements.
@@ -107,6 +111,7 @@ class Bolton(optimizer_v2.OptimizerV2):
   Bolt-on Differential Privacy for Scalable Stochastic Gradient
   Descent-based Analytics by Xi Wu et. al.
   """
+
   def __init__(self,  # pylint: disable=super-init-not-called
                optimizer,
                loss,
@@ -118,11 +123,12 @@ class Bolton(optimizer_v2.OptimizerV2):
         optimizer: Optimizer_v2 or subclass to be used as the optimizer
           (wrapped).
         loss: StrongConvexLoss function that the model is being compiled with.
+        dtype: dtype
     """
 
     if not isinstance(loss, StrongConvexMixin):
-      raise ValueError("loss function must be a Strongly Convex and therefore "
-                       "extend the StrongConvexMixin.")
+      raise ValueError('loss function must be a Strongly Convex and therefore '
+                       'extend the StrongConvexMixin.')
     self._private_attributes = ['_internal_optimizer',
                                 'dtype',
                                 'noise_distribution',
@@ -154,6 +160,9 @@ class Bolton(optimizer_v2.OptimizerV2):
     Args:
       force: True to normalize regardless of previous weight values.
         False to check if weights > R-ball and only normalize then.
+
+    Raises:
+      Exception:
     """
     if not self._is_init:
       raise Exception('This method must be called from within the optimizer\'s '
@@ -171,7 +180,7 @@ class Bolton(optimizer_v2.OptimizerV2):
         )
 
   def get_noise(self, input_dim, output_dim):
-    """Sample noise to be added to weights for privacy guarantee
+    """Sample noise to be added to weights for privacy guarantee.
 
       Args:
         input_dim: the input dimensionality for the weights
@@ -179,6 +188,9 @@ class Bolton(optimizer_v2.OptimizerV2):
 
       Returns:
         Noise in shape of layer's weights to be added to the weights.
+
+      Raises:
+        Exception:
     """
     if not self._is_init:
       raise Exception('This method must be called from within the optimizer\'s '