🎨 Small cleanup

src/layer/dense.rs

@@ -242,7 +242,7 @@ impl<
 {
     fn backprop_layer(
         &self,
-        input: &DVector<F>,
+        _input: &DVector<F>,
         evaluated: &Self::IntermediaryRepr,
         epsilon: &Self::Output,
     ) -> DVector<F> {

src/layer/normalize.rs

@@ -89,9 +89,9 @@ impl<F: Float + Scalar + NumAssignOps> NeuraTrainableLayerSelf<DVector<F>> for N
 
     fn get_gradient(
         &self,
-        input: &DVector<F>,
+        _input: &DVector<F>,
-        intermediary: &Self::IntermediaryRepr,
+        _intermediary: &Self::IntermediaryRepr,
-        epsilon: &Self::Output,
+        _epsilon: &Self::Output,
     ) -> Self::Gradient {
         ()
     }
@@ -102,7 +102,7 @@ impl<F: Float + Scalar + NumAssignOps> NeuraTrainableLayerBackprop<DVector<F>>
 {
     fn backprop_layer(
         &self,
-        input: &DVector<F>,
+        _input: &DVector<F>,
         (jacobian_partial, stddev): &Self::IntermediaryRepr,
         epsilon: &Self::Output,
     ) -> DVector<F> {

src/layer/softmax.rs

@@ -81,9 +81,9 @@ impl<F: Float + Scalar + NumAssignOps> NeuraTrainableLayerSelf<DVector<F>> for N
     #[inline(always)]
     fn get_gradient(
         &self,
-        input: &DVector<F>,
+        _input: &DVector<F>,
-        intermediary: &Self::IntermediaryRepr,
+        _intermediary: &Self::IntermediaryRepr,
-        epsilon: &Self::Output,
+        _epsilon: &Self::Output,
     ) -> Self::Gradient {
         ()
     }

src/network/sequential/layer_impl.rs

@@ -63,9 +63,9 @@ impl<
 
     fn get_gradient(
         &self,
-        input: &Input,
+        _input: &Input,
-        intermediary: &Self::IntermediaryRepr,
+        _intermediary: &Self::IntermediaryRepr,
-        epsilon: &Self::Output,
+        _epsilon: &Self::Output,
     ) -> Self::Gradient {
         unimplemented!("NeuraSequential::get_gradient is not yet implemented, sorry");
     }