mnist_example.rb in ruby-dnn-0.9.0

- old
+ new

@@ -1,12 +1,13 @@
 require "dnn"
 require "dnn/lib/mnist"
-#require "numo/linalg/autoloader"
+# require "numo/linalg/autoloader"
 
 include DNN::Layers
 include DNN::Activations
 include DNN::Optimizers
+include DNN::Losses
 Model = DNN::Model
 MNIST = DNN::MNIST
 
 x_train, y_train = MNIST.load_train
 x_test, y_test = MNIST.load_test
@@ -15,12 +16,12 @@
 x_test = Numo::SFloat.cast(x_test).reshape(x_test.shape[0], 784)
 
 x_train /= 255
 x_test /= 255
 
-y_train = DNN::Util.to_categorical(y_train, 10, Numo::SFloat)
-y_test = DNN::Util.to_categorical(y_test, 10, Numo::SFloat)
+y_train = DNN::Utils.to_categorical(y_train, 10, Numo::SFloat)
+y_test = DNN::Utils.to_categorical(y_test, 10, Numo::SFloat)
 
 model = Model.new
 
 model << InputLayer.new(784)
 
@@ -31,10 +32,9 @@
 model << Dense.new(256)
 model << BatchNormalization.new
 model << ReLU.new
 
 model << Dense.new(10)
-model << SoftmaxWithLoss.new
 
-model.compile(RMSProp.new)
+model.compile(RMSProp.new, SoftmaxCrossEntropy.new)
 
 model.train(x_train, y_train, 10, batch_size: 100, test: [x_test, y_test])