Style fixes

dataloader/preprocessor.hpp

@@ -90,9 +90,7 @@ class PreProcessor
     {
       // Convert each element to uint8 first and then divide by 255.
       for (size_t i = 0; i < trainFeatures.n_elem; i++)
-      {
           trainFeatures(i) = ((uint8_t)(trainFeatures(i)) / 255.0);
-      }
     }
   }
 };

ensmallen_utils/periodic_save.hpp

@@ -33,10 +33,10 @@ class PeriodicSave
    * @param network Network type which will be saved periodically.
    * @param filePath Base path / folder where weights will be saved.
    * @param modelPrefix Weights will be stored as
-   *                    modelPrefix_epoch_loss.bin.
+   *    modelPrefix_epoch_loss.bin.
    * @param period Period after which the model will be saved.
    * @param silent Boolean to determine whether or not to print saving
-   *               of model.
+   *    of model.
    * @param output Outputstream where output will be directed.
    */
   PeriodicSave(AnnType& network,
@@ -70,9 +70,7 @@ class PeriodicSave
           objectiveString;
       mlpack::data::Save(filePath + modelName + ".bin", modelPrefix, network);
       if (!silent)
-      {
         output << "Model saved as " << modelName << std::endl;
-      }
     }
 
     return false;

ensmallen_utils/print_metric.hpp

@@ -23,7 +23,7 @@ namespace ens {
  *
  * @tparam ANNType Type of model which will be used for evaluating metric.
  * @tparam MetricType Metric class which must have static `Evaluate` function
- *                    that will be called at the end of the epoch.
+ *    that will be called at the end of the epoch.
  * @tparam InputType Arma type of dataset features.
  * @tparam OutputType Arma type of dataset labels.
  */
@@ -42,7 +42,7 @@ class PrintMetric
    * @param responses Ground truth label for the mdoel.
    * @param metricName Metric name which will be printed after each epoch.
    * @param trainData Boolean to determine whether dataset corresponds to
-   *                  training data or validation data.
+   *     training data or validation data.
    * @param output Outputstream where output will be directed.
    */
   PrintMetric(AnnType &network,

models/darknet/darknet.hpp

@@ -2,7 +2,7 @@
  * @file darknet.hpp
  * @author Kartik Dutt
  * 
- * Definition of Darknet models.
+ * Definition of DarkNet models.
  * 
  * For more information, kindly refer to the following paper.
  * 
@@ -50,7 +50,7 @@ namespace mlpack {
 namespace ann /** Artificial Neural Network. */{
 
 /**
- * Definition of a Darknet CNN.
+ * Definition of a DarkNet CNN.
  * 
  * @tparam OutputLayerType The output layer type used to evaluate the network.
  * @tparam InitializationRuleType Rule used to initialize the weight matrix.
@@ -74,8 +74,8 @@ class DarkNet
    * @param inputWidth Width of the input image.
    * @param inputHeight Height of the input image.
    * @param numClasses Optional number of classes to classify images into,
-   *                   only to be specified if includeTop is  true.
-   * @param weights One of 'none', 'cifar10'(pre-training on CIFAR10) or path to weights.
+   *     only to be specified if includeTop is  true.
+   * @param weights One of 'none', 'imagenet'(pre-training on ImageNet) or path to weights.
    * @param includeTop Must be set to true if weights are set.
    */
   DarkNet(const size_t inputChannel,
@@ -89,12 +89,11 @@ class DarkNet
    * DarkNet constructor intializes input shape and number of classes.
    *
    * @param inputShape A three-valued tuple indicating input shape.
-   *                   First value is number of Channels (Channels-First).
-   *                   Second value is input height.
-   *                   Third value is input width..
+   *     First value is number of channels (channels-first).
+   *     Second value is input height. Third value is input width.
    * @param numClasses Optional number of classes to classify images into,
-   *                   only to be specified if includeTop is  true.
-   * @param weights One of 'none', 'cifar10'(pre-training on CIFAR10) or path to weights.
+   *     only to be specified if includeTop is  true.
+   * @param weights One of 'none', 'imagenet'(pre-training on ImageNet) or path to weights.
    */
   DarkNet(const std::tuple<size_t, size_t, size_t> inputShape,
           const size_t numClasses = 1000,
@@ -115,7 +114,7 @@ class DarkNet
    * Adds Convolution Block.
    *
    * @tparam SequentialType Layer type in which convolution block will
-   *                        be added.
+   *     be added.
    *
    * @param inSize Number of input maps.
    * @param outSize Number of output maps.
@@ -126,7 +125,7 @@ class DarkNet
    * @param padW Padding width of the input.
    * @param padH Padding height of the input.
    * @param batchNorm Boolean to determine whether a batch normalization
-   *                  layer is added.
+   *     layer is added.
    * @param negativeSlope Negative slope hyper-parameter for LeakyReLU.
    * @param baseLayer Layer in which Convolution block will be added, if
    *                  NULL added to darkNet FFN.
@@ -160,28 +159,22 @@ class DarkNet
         ", " << outSize << ")" << std::endl;
 
     if (batchNorm)
-    {
       bottleNeck->Add(new BatchNorm<>(outSize, 1e-5, false));
-    }
 
     bottleNeck->Add(new LeakyReLU<>(negativeSlope));
 
     if (baseLayer != NULL)
-    {
       baseLayer->Add(bottleNeck);
-    }
     else
-    {
       darkNet.Add(bottleNeck);
-    }
   }
 
   /**
    * Adds Pooling Block.
    *
    * @param factor The factor by which input dimensions will be divided.
    * @param type One of "max" or "mean". Determines whether add mean pooling
-   *             layer or max pooling layer.
+   *     layer or max pooling layer.
    */
   void PoolingBlock(const size_t factor = 2,
                     const std::string type = "max")

models/darknet/darknet_impl.hpp

@@ -2,7 +2,7 @@
  * @file darknet_impl.hpp
  * @author Kartik Dutt
  *
- * Implementation of LeNet using mlpack.
+ * Implementation of DarkNet using mlpack.
  *
  * mlpack is free software; you may redistribute it and/or modify it under the
  * terms of the 3-clause BSD license.  You should have received a copy of the
@@ -113,7 +113,7 @@ DarkNet<OutputLayerType, InitializationRuleType, DarkNetVersion>::DarkNet(
     if (includeTop)
     {
       darkNet.Add(new Convolution<>(1024, numClasses, 1, 1,
-        1, 1, 0, 0, inputWidth, inputHeight));
+          1, 1, 0, 0, inputWidth, inputHeight));
       darkNet.Add(new AdaptiveMeanPooling<>(1, 1));
       darkNet.Add(new LogSoftMax<>());
     }
@@ -134,9 +134,7 @@ DarkNet<OutputLayerType, InitializationRuleType, DarkNetVersion>::DarkNet(
     for (size_t blockCount : residualBlockConfig)
     {
       for (size_t i = 0; i < blockCount; i++)
-      {
         DarkNet53ResidualBlock(curChannels);
-      }
 
       if (blockCount != 4)
       {