Merge branch 'master' into master

Author: Windaway

.travis.yml

@@ -117,4 +117,3 @@ deploy:
     python: '3.6'
     condition: '$_TF_VERSION = 2.0.0a0'
 #     condition: '$_TF_VERSION = 1.11.0'
-

CHANGELOG.md

@@ -67,7 +67,14 @@ To release a new version, please update the changelog as followed:
 
 <!-- YOU CAN EDIT FROM HERE -->
 
-## [Unreleased]
+## [2.0.1] - 2019-5-17
+
+
+A maintain release.
+
+### Changed
+- remove `tl.layers.initialize_global_variables(sess)` (PR #931)
+- support `trainable_weights` (PR #966)
 
 ### Added
  - Layer
@@ -78,29 +85,28 @@ To release a new version, please update the changelog as followed:
 - change `tl.layers.core`, `tl.models.core` (PR #966)
 - change `weights` into `all_weights`, `trainable_weights`, `nontrainable_weights`
 
+
 ### Dependencies Update
 - nltk>=3.3,<3.4 => nltk>=3.3,<3.5 (PR #892)
 - pytest>=3.6,<3.11 => pytest>=3.6,<4.1 (PR #889)
 - yapf>=0.22,<0.25 => yapf==0.25.0 (PR #896)
 - imageio==2.5.0 progressbar2==3.39.3  scikit-learn==0.21.0 scikit-image==0.15.0 scipy==1.2.1 wrapt==1.11.1 pymongo==3.8.0 sphinx==2.0.1 wrapt==1.11.1 opencv-python==4.1.0.25 requests==2.21.0 tqdm==4.31.1	lxml==4.3.3 pycodestyle==2.5.0 sphinx==2.0.1 yapf==0.27.0(PR #967)
 
-### Deprecated
-
 ### Fixed
 - fix docs of models @zsdonghao #957
 - In `BatchNorm`, keep dimensions of mean and variance to suit `channels first` (PR #963)
 
-
-### Removed
-
-### Security
-
 ### Contributors
+- @warshallrho: #PR966
 - @zsdonghao: #931
 - @yd-yin: #963
+- @dvklopfenstein: #971
+
 
 
-## [2.0.0-alpha] - 2019-05-04
+## [2.0.0] - 2019-05-04
+
+To many PR for this update, please check [here](https://github.com/tensorlayer/tensorlayer/releases/tag/2.0.0) for more details.
 
 ### Changed
 * update for TensorLayer 2.0.0 alpha version (PR #952)
@@ -119,6 +125,7 @@ To release a new version, please update the changelog as followed:
 - @ChrisWu1997
 - @warshallrho
 
+
 ## [1.11.1] - 2018-11-15
 
 ### Changed

README.md

@@ -62,7 +62,7 @@ and fine-tune. On the contrary, TensorLayer APIs are generally lightweight, flex
 Users often find it easy to start with the examples and tutorials, and then dive
 into TensorFlow seamlessly. In addition, TensorLayer does not create library lock-in through native supports for importing components from Keras.
 
-TensorLayer has a fast growing usage among top researchers and engineers, from universities like
+TensorLayer has a fast growing usage among top researchers and engineers, from universities like Peking University,
 Imperial College London, UC Berkeley, Carnegie Mellon University, Stanford University, and
 University of Technology of Compiegne (UTC), and companies like Google, Microsoft, Alibaba, Tencent, Xiaomi, and Bloomberg.
 

docs/user/contributing.rst

@@ -6,7 +6,7 @@ Contributing
 
 TensorLayer is a major ongoing research project in Data Science Institute, Imperial College London. 
 The goal of the project is to develop a compositional language while complex learning systems
-can be build through composition of neural network modules.
+can be built through composition of neural network modules.
 
 Numerous contributors come from various horizons such as: Tsinghua University, Carnegie Mellon University, University of Technology of Compiegne, 
 Google, Microsoft, Bloomberg and etc.
@@ -51,7 +51,7 @@ What to contribute
 Your method and example
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-If you have a new method or example in term of Deep learning and Reinforcement learning,
+If you have a new method or example in terms of Deep learning and Reinforcement learning,
 you are welcome to contribute.
 
 * Provide your layer or example, so everyone can use it.

docs/user/faq.rst

@@ -20,18 +20,18 @@ and `Wechat group <https://github.com/shorxp/tensorlayer-chinese/blob/master/doc
 
 Beginner
 -----------
-For people who new to deep learning, the contirbutors provided a number of tutorials in this website, these tutorials will guide you to understand autoencoder, convolutional neural network, recurrent neural network, word embedding and deep reinforcement learning and etc. If your already understand the basic of deep learning, we recommend you to skip the tutorials and read the example codes on `Github <https://github.com/tensorlayer/tensorlayer>`__ , then implement an example from scratch.
+For people who new to deep learning, the contributors provided a number of tutorials in this website, these tutorials will guide you to understand autoencoder, convolutional neural network, recurrent neural network, word embedding and deep reinforcement learning and etc. If your already understand the basic of deep learning, we recommend you to skip the tutorials and read the example codes on `Github <https://github.com/tensorlayer/tensorlayer>`__ , then implement an example from scratch.
 
 Engineer
 ------------
-For people from industry, the contirbutors provided mass format-consistent examples covering computer vision, natural language processing and reinforcement learning. Besides, there are also many TensorFlow users already implemented product-level examples including image captioning, semantic/instance segmentation, machine translation, chatbot and etc, which can be found online.
+For people from industry, the contributors provided mass format-consistent examples covering computer vision, natural language processing and reinforcement learning. Besides, there are also many TensorFlow users already implemented product-level examples including image captioning, semantic/instance segmentation, machine translation, chatbot and etc., which can be found online.
 It is worth noting that a wrapper especially for computer vision `Tf-Slim <https://github.com/tensorflow/models/tree/master/slim#Pretrained>`__ can be connected with TensorLayer seamlessly.
 Therefore, you may able to find the examples that can be used in your project.
 
 Researcher
 -------------
-For people from academic, TensorLayer was originally developed by PhD students who facing issues with other libraries on implement novel algorithm. Installing TensorLayer in editable mode is recommended, so you can extend your methods in TensorLayer.
-For researches related to image such as image captioning, visual QA and etc, you may find it is very helpful to use the existing `Tf-Slim pre-trained models <https://github.com/tensorflow/models/tree/master/slim#Pretrained>`__ with TensorLayer (a specially layer for connecting Tf-Slim is provided).
+For people from academia, TensorLayer was originally developed by PhD students who facing issues with other libraries on implement novel algorithm. Installing TensorLayer in editable mode is recommended, so you can extend your methods in TensorLayer.
+For research related to image processing such as image captioning, visual QA and etc., you may find it is very helpful to use the existing `Tf-Slim pre-trained models <https://github.com/tensorflow/models/tree/master/slim#Pretrained>`__ with TensorLayer (a specially layer for connecting Tf-Slim is provided).
 
 
 Exclude some layers from training
@@ -48,7 +48,7 @@ To choose which variables to update, you can do as below.
 
   train_params = network.trainable_weights[3:]
 
-The second way is to get the variables by a given name. For example, if you want to get all variables which the layer name contain ``dense``, you can do as below.
+The second way is to get the variables by a given name. For example, if you want to get all variables which the layer name contains ``dense``, you can do as below.
 
 .. code-block:: python
 
@@ -76,13 +76,13 @@ Visualization
 Cannot Save Image
 -----------------------
 
-If you run the script via SSH control, sometime you may find the following error.
+If you run the script via SSH control, sometimes you may find the following error.
 
 .. code-block:: bash
 
   _tkinter.TclError: no display name and no $DISPLAY environment variable
 
-If happen, run ``sudo apt-get install python3-tk`` or ``import matplotlib`` and ``matplotlib.use('Agg')`` before ``import tensorlayer as tl``.
+If this happens, run ``sudo apt-get install python3-tk`` or ``import matplotlib`` and ``matplotlib.use('Agg')`` before ``import tensorlayer as tl``.
 Alternatively, add the following code into the top of ``visualize.py`` or in your own code.
 
 .. code-block:: python

examples/basic_tutorials/README.md

@@ -1,44 +1 @@
-Something you need to know:
 
-### 1. Static and dynamic model
-
-1) `tutorial_mnist_mlp_static.py`: static model
-
-2) `tutorial_mnist_mlp_dynamic.py`: dynamic model
-
-### 2. Switching Training and testing
-
-There are two ways to switch the training and testing mode: 
-
-1 ) use Pytorch-like method, turn on and off the training/evaluation as follow:
-
-```python
-model.train() # enable dropout, batch norm decay and etc
-y1 = model(x)
-model.eval() # disable dropout, fix batch norm weights and etc
-y2 = model(x)
-```
-
-2) use TensorLayer 1.x method, input `is_train` to the model while inferencing.
-
-```python
-y1 = model(x, is_train=True)
-y2 = model(x, is_train=False)
-```
-
-
-
-### Data augmentation
-
-- Data augmentation is essential for training, while if the augmentation is complex, it will slow down the training.
-We used CIFAR10 classification as example of data augmentation. 
-- For the best performance, please use `tutorial_cifar10_datasetapi.py`.
-- It is suggested to use TensorFlow's DataSet API (`tf.data` and `tf.image`) and TFRecord for the sake of performance and generalibity.
-- For TFRecord and Dataset API,
-TFRecord needs to first store all data into TFRecord format, while Dataset API is simpler that can directly use data XXXX.
-
-### Float16
-- For Float16, some GPUs can speed up but some cannot.
-
-### Others
-- For distributed training

tensorlayer/initializers.py

@@ -192,27 +192,6 @@ def deconv2d_bilinear_upsampling_initializer(shape):
         A constant initializer with weights set to correspond to per channel bilinear upsampling
         when passed as W_int in DeConv2dLayer
 
-    Examples
-    --------
-    Upsampling by a factor of 2, ie e.g 100->200
-    >>> import tensorflow as tf
-    >>> import tensorlayer as tl
-    >>> rescale_factor = 2
-    >>> imsize = 128
-    >>> num_channels = 3
-    >>> num_in_channels = 3
-    >>> num_out_channels = 3
-    >>> filter_shape = (5, 5, num_out_channels, num_in_channels)
-    >>> ni = tl.layers.Input(shape=(1, imsize, imsize, num_channels))
-    >>> bilinear_init = deconv2d_bilinear_upsampling_initializer(shape=filter_shape)
-    >>> net = tl.layers.DeConv2dLayer(
-    ...                    shape=filter_shape,
-    ...                    outputs_shape=(1, imsize*rescale_factor, imsize*rescale_factor, num_out_channels),
-    ...                    strides=(1, rescale_factor, rescale_factor, 1),
-    ...                    W_init=bilinear_init,
-    ...                    padding='SAME',
-    ...                    act=None, name='g/h1/decon2d')(ni)
-
     """
     if shape[0] != shape[1]:
         raise Exception('deconv2d_bilinear_upsampling_initializer only supports symmetrical filter sizes')

tensorlayer/package_info.py

@@ -4,7 +4,7 @@
 
 MAJOR = 2
 MINOR = 0
-PATCH = 0
+PATCH = 1
 PRE_RELEASE = ''
 # Use the following formatting: (major, minor, patch, prerelease)
 VERSION = (MAJOR, MINOR, PATCH, PRE_RELEASE)