minor changes on demo/gan following lzhao4ever comments

demo/gan/README.md

@@ -9,4 +9,5 @@ Then you can run the command below. The flag -d specifies the training data (cif
 
 $python gan_trainer.py -d cifar --useGpu 1
 
-The generated images will be stored in ./cifar_samples/
+The generated images will be stored in ./cifar_samples/
+The corresponding models will be stored in ./cifar_params/

demo/gan/data/get_mnist_data.sh

@@ -1,5 +1,5 @@
 #!/usr/bin/env sh
-# This scripts downloads the mnist data and unzips it.
+# This script downloads the mnist data and unzips it.
 set -e
 DIR="$( cd "$(dirname "$0")" ; pwd -P )"
 rm -rf "$DIR/mnist_data"

demo/gan/gan_conf.py

@@ -38,7 +38,7 @@ sample_dim = 2
 settings(
     batch_size=128,
     learning_rate=1e-4,
-    learning_method=AdamOptimizer(beta1=0.7)
+    learning_method=AdamOptimizer(beta1=0.5)
 )
 
 def discriminator(sample):

demo/gan/gan_trainer.py

@@ -87,11 +87,8 @@ def load_mnist_data(imageFile):
     else:
         n = 10000
     
-    data = numpy.zeros((n, 28*28), dtype = "float32")
+    data = numpy.fromfile(f, 'ubyte', count=n*28*28).reshape((n, 28*28))
-    
+    data = data / 255.0 * 2.0 - 1.0
-    for i in range(n):
-        pixels = numpy.fromfile(f, 'ubyte', count=28*28)
-        data[i, :] = pixels / 255.0 * 2.0 - 1.0
 
     f.close()
     return data
@@ -235,7 +232,7 @@ def main():
     else:
         data_np = load_uniform_data()
     
-    # this create a gradient machine for discriminator
+    # this creates a gradient machine for discriminator
     dis_training_machine = api.GradientMachine.createFromConfigProto(
         dis_conf.model_config)
     # this create a gradient machine for generator    
@@ -243,7 +240,7 @@ def main():
         gen_conf.model_config)
 
     # generator_machine is used to generate data only, which is used for
-    # training discrinator
+    # training discriminator
     logger.info(str(generator_conf.model_config))
     generator_machine = api.GradientMachine.createFromConfigProto(
         generator_conf.model_config)