Docs update

README.md

@@ -24,7 +24,7 @@ See some [examples](examples) to learn about the framework:
 + [char-rnn for fun](examples/Char-RNN)
 + [LSTM language model on PennTreebank](examples/PennTreebank)
 
-The examples are not only for demonstration of the framework -- you can train them and reproduce the results in papers.
+Examples are not only for demonstration of the framework -- you can train them and reproduce the results in papers.
 
 ## Features:
 

docs/tutorial/index.rst

@@ -15,10 +15,11 @@ A High Level Glance
 * You can use any TF-based symbolic function library to define a model, including
   a small set of models within tensorpack. ``ModelDesc`` is an interface to connect symbolic graph to tensorpack trainers.
 
-* tensorpack trainers manage the training loops for you. At the same time, you own the power of customization
-  through callbacks. They also include data parallel logic for multi-GPU or distributed training.
+* tensorpack trainers manage the training loops for you.
+  They also include data parallel logic for multi-GPU or distributed training.
+  At the same time, you have the power of customization through callbacks.
 
-* Callbacks are like ``tf.train.SessionRunHook``, or plugins, or extensions. During training,
+* Callbacks are like ``tf.train.SessionRunHook``, or plugins. During training,
   everything you want to do other than the main iterations can be defined through callbacks and easily reused.
 
 User Tutorials

examples/README.md

@@ -10,11 +10,10 @@ Training examples with __reproducible__ and meaningful performance.
 
 ## Vision:
 + [A tiny SVHN ConvNet with 97.8% accuracy](svhn-digit-convnet.py)
++ [Multi-GPU training of ResNet on ImageNet](ResNet)
 + [DoReFa-Net: training binary / low-bitwidth CNN on ImageNet](DoReFa-Net)
-+ [Train ResNet for ImageNet/Cifar10/SVHN](ResNet)
 + [Generative Adversarial Network(GAN) variants](GAN), including DCGAN, InfoGAN, Conditional GAN, WGAN, BEGAN, DiscoGAN, Image to Image, CycleGAN.
-+ [Inception-BN with 71% accuracy](Inception/inception-bn.py)
-+ [InceptionV3 with 74% accuracy (similar to the official code)](Inception/inceptionv3.py)
++ [Inception-BN and InceptionV3](Inception)
 + [Fully-convolutional Network for Holistically-Nested Edge Detection(HED)](HED)
 + [Spatial Transformer Networks on MNIST addition](SpatialTransformer)
 + [Visualize CNN saliency maps](Saliency)

tensorpack/utils/viz.py

@@ -133,8 +133,8 @@ class Canvas(object):
                                self.channel), dtype='uint8')
 
     def draw_patches(self, plist):
-        assert self.nr_row * self.nr_col == len(plist), \
-            "{}*{} != {}".format(self.nr_row, self.nr_col, len(plist))
+        assert self.nr_row * self.nr_col >= len(plist), \
+            "{}*{} < {}".format(self.nr_row, self.nr_col, len(plist))
         if self.channel == 3 and plist.shape[3] == 1:
             plist = np.repeat(plist, 3, axis=3)
         cur_row, cur_col = 0, 0