ruby-dnn 0.10.3 → 0.10.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ad7089d958268c000dbe71a9fd44b5a65a9824fe792ad445c0ed58eff33f82fe
-  data.tar.gz: 249b43df73430bf42c15e435be51842ec6b8d0a52fb46e59b573ac5fe14d44f6
+  metadata.gz: 589418cabcfddc5f0066011a7094a9740903d891c0a53248c2dec2ae6ba050f9
+  data.tar.gz: af86c3663a8e855a8ccf2ffed6825037ebe452c0e83810d0595282959715516f
 SHA512:
-  metadata.gz: 8d904d65210bae39bb4ce3c0d4cb7bd31331663084a112a8ba11424ea73b06d7706ead1c2113e9812e98dd5f82510fe045c4deb13c92480b2897811bbf063885
-  data.tar.gz: 7c32096caf2a00edab6d27a65f3dfdd1663291be10ac50af256fc9486b866fd9455d5d4216aebb71b984fc41c3f6db7272e27aa39accaa0f0489990d8c562d3f
+  metadata.gz: d842fe4bbbd6801726e97177ab1212b6eab9d09bcb553a5bf35e082a9685ffea362fcf729ea41a4f438790bc5d2ddd0c962ad629af6c2afe45d95049989ddd37
+  data.tar.gz: fa13a086a3232e6117d5976a467a1a0bb3640f2b8cd7c6db1b2ef5a6c17d42a3d47ca9ce493f6278c331beda12380b4344d502b3cebef0b6978891c383d19167

data/.travis.yml

@@ -3,3 +3,5 @@ language: ruby
 rvm:
   - 2.5.1
 before_install: gem install bundler -v 1.16.2
+script:
+  - rake test

data/README.md

@@ -1,5 +1,6 @@
 # ruby-dnn
-[![Gem Version](https://badge.fury.io/rb/ruby-dnn.svg)](https://badge.fury.io/rb/ruby-dnn)
+[![Gem Version](https://badge.fury.io/rb/ruby-dnn.svg)](https://badge.fury.io/rb/ruby-dnn)  
+[![Build Status](https://travis-ci.org/unagiootoro/ruby-dnn.svg?branch=master)](https://travis-ci.org/unagiootoro/ruby-dnn)
 
 ruby-dnn is a ruby deep learning library. This library supports full connected neural network and convolution neural network.
 Currently, you can get 99% accuracy with MNIST and 74% with CIFAR 10.
@@ -56,7 +57,7 @@ If you want to know more detailed information, please refer to the source code.
 | Losses | MeanSquaredError, MeanAbsoluteError, HuberLoss, SoftmaxCrossEntropy, SigmoidCrossEntropy |
 
 ## TODO
-● Add CI badge.  
+● Support to define by run model.  
 ● Write a test.  
 ● Write a document.  
 ● Support to GPU.  

data/Rakefile

@@ -9,7 +9,7 @@ Rake::TestTask.new(:test) do |t|
 end
 
 task :build_dataset_loader do
-  sh "cd ext/cifar10_loader; ruby extconf.rb; make"
+  sh "cd ext/cifar_loader; ruby extconf.rb; make"
 end
 
 task :build_image_io do

data/lib/dnn/core/activations.rb

@@ -61,15 +61,15 @@ module DNN
     
     class ReLU < Layers::Layer
       def forward(x)
-        @x = x.clone
+        @x = x
         x[x < 0] = 0
         x
       end
     
       def backward(dy)
-        @x[@x > 0] = 1
-        @x[@x <= 0] = 0
-        dy * @x
+        dx = Xumo::SFloat.ones(@x.shape)
+        dx[@x <= 0] = 0
+        dy * dx
       end
     end
 
@@ -88,16 +88,16 @@ module DNN
       end
 
       def forward(x)
-        @x = x.clone
+        @x = x
         a = Xumo::SFloat.ones(x.shape)
         a[x <= 0] = @alpha
         x * a
       end
 
       def backward(dy)
-        @x[@x > 0] = 1
-        @x[@x <= 0] = @alpha
-        dy * @x
+        dx = Xumo::SFloat.ones(@x.shape)
+        dx[@x <= 0] = @alpha
+        dy * dx
       end
 
       def to_hash

data/lib/dnn/core/model.rb

@@ -281,12 +281,13 @@ module DNN
       dataset = Dataset.new(x, y, false)
       correct = 0
       sum_loss = 0
-      (x.shape[0].to_f / batch_size).ceil.times do |i|
+      max_iter = (x.shape[0].to_f / batch_size)
+      max_iter.ceil.times do |i|
         x_batch, y_batch = dataset.next_batch(batch_size)
         x_batch, y_batch = before_batch_cbk.call(x_batch, y_batch, false) if before_batch_cbk
         x_batch = forward(x_batch, false)
         sigmoid = Sigmoid.new
-        batch_size.times do |j|
+        x_batch.shape[0].times do |j|
           if @layers.last.output_shape == [1]
             if @loss_func.is_a?(SigmoidCrossEntropy)
               correct += 1 if sigmoid.forward(x_batch[j, 0]).round == y_batch[j, 0].round
@@ -299,9 +300,9 @@ module DNN
         end
         loss_value = @loss_func.forward(x_batch, y_batch, get_all_layers)
         after_batch_cbk.call(loss_value, false) if after_batch_cbk
-        sum_loss += loss_value.is_a?(Numo::SFloat) ? loss_value.mean : loss_value
+        sum_loss += loss_value.is_a?(Xumo::SFloat) ? loss_value.mean : loss_value
       end
-      mean_loss = sum_loss / batch_size
+      mean_loss = sum_loss / max_iter
       [correct.to_f / x.shape[0], mean_loss]
     end
 

data/lib/dnn/version.rb

@@ -1,3 +1,3 @@
 module DNN
-  VERSION = "0.10.3"
+  VERSION = "0.10.4"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby-dnn
 version: !ruby/object:Gem::Version
-  version: 0.10.3
+  version: 0.10.4
 platform: ruby
 authors:
 - unagiootoro
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2019-06-25 00:00:00.000000000 Z
+date: 2019-06-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: numo-narray