Merge branch 'master' into clear_softmax_warning_mnist

research/differential_privacy/README.md

 <font size=4><b>Deep Learning with Differential Privacy</b></font>
 
-Open Sourced By: Xin Pan (xpan@google.com, github: panyx0718)
+Open Sourced By: Xin Pan
 
 
 ### Introduction for [dp_sgd/README.md](dp_sgd/README.md)

research/lm_1b/README.md

@@ -3,7 +3,7 @@
 <b>Authors:</b>
 
 Oriol Vinyals (vinyals@google.com, github: OriolVinyals),
-Xin Pan (xpan@google.com, github: panyx0718)
+Xin Pan
 
 <b>Paper Authors:</b>
 

research/next_frame_prediction/README.md

@@ -8,7 +8,7 @@ This is an implementation based on my understanding, with small
 variations. It doesn't necessarily represents the paper published
 by the original authors.
 
-Authors: Xin Pan (Github: panyx0718), Anelia Angelova
+Authors: Xin Pan, Anelia Angelova
 
 <b>Results:</b>
 

research/resnet/README.md

 <font size=4><b>Reproduced ResNet on CIFAR-10 and CIFAR-100 dataset.</b></font>
 
-contact: panyx0718 (xpan@google.com)
+Xin Pan
 
 <b>Dataset:</b>
 

research/textsum/README.md

@@ -2,7 +2,7 @@ Sequence-to-Sequence with Attention Model for Text Summarization.
 
 Authors:
 
-Xin Pan (xpan@google.com, github:panyx0718),
+Xin Pan
 Peter Liu (peterjliu@google.com, github:peterjliu)
 
 <b>Introduction</b>

tutorials/image/cifar10/cifar10.py

@@ -204,7 +204,7 @@ def inference(images):
     kernel = _variable_with_weight_decay('weights',
                                          shape=[5, 5, 3, 64],
                                          stddev=5e-2,
-                                         wd=0.0)
+                                         wd=None)
     conv = tf.nn.conv2d(images, kernel, [1, 1, 1, 1], padding='SAME')
     biases = _variable_on_cpu('biases', [64], tf.constant_initializer(0.0))
     pre_activation = tf.nn.bias_add(conv, biases)
@@ -223,7 +223,7 @@ def inference(images):
     kernel = _variable_with_weight_decay('weights',
                                          shape=[5, 5, 64, 64],
                                          stddev=5e-2,
-                                         wd=0.0)
+                                         wd=None)
     conv = tf.nn.conv2d(norm1, kernel, [1, 1, 1, 1], padding='SAME')
     biases = _variable_on_cpu('biases', [64], tf.constant_initializer(0.1))
     pre_activation = tf.nn.bias_add(conv, biases)
@@ -262,7 +262,7 @@ def inference(images):
   # and performs the softmax internally for efficiency.
   with tf.variable_scope('softmax_linear') as scope:
     weights = _variable_with_weight_decay('weights', [192, NUM_CLASSES],
-                                          stddev=1/192.0, wd=0.0)
+                                          stddev=1/192.0, wd=None)
     biases = _variable_on_cpu('biases', [NUM_CLASSES],
                               tf.constant_initializer(0.0))
     softmax_linear = tf.add(tf.matmul(local4, weights), biases, name=scope.name)