Reinforcement learning

CHANGELOG.md

@@ -90,11 +90,16 @@ To release a new version, please update the changelog as followed:
 ### Changed
 - Add version_info in model.config. (PR #992)
 - Replace tf.nn.func with tf.nn.func.\_\_name\_\_ in model config.
+- Add Reinforcement learning tutorials. (PR #995)
 
 ### Fixed
 
 ### Contributors
 - @warshallrho:
+- @quantumiracle: #995
+- @Tokarev-TT-33: #995
+- @initial-h: #995
+- @Officium: #995
 
 ## [2.0.2] - 2019-6-5
 
@@ -124,6 +129,8 @@ A maintain release.
  - Layer
     - `InstanceNorm`, `InstanceNorm1d`, `InstanceNorm2d`, `InstanceNorm3d` (PR #963)
 
+* Reinforcement learning tutorials. (PR #995)
+
 ### Changed
 - remove `tl.layers.initialize_global_variables(sess)` (PR #931)
 - update `tutorial_generate_text.py`, `tutorial_ptb_lstm.py`. remove `tutorial_ptb_lstm_state_is_tuple.py` (PR #958)
@@ -144,8 +151,15 @@ A maintain release.
 - @warshallrho: #PR966
 - @zsdonghao: #931
 - @yd-yin: #963
+<<<<<<< HEAD
+- @Tokarev-TT-33: # 995
+- @initial-h: # 995
+- @quantumiracle: #995
+- @Officium: #995
+=======
 - @1FengL: #958
 - @dvklopfenstein: #971
+>>>>>>> 560dbb8a17963023a3b1d59a79e1c2752530114a
 
 
 ## [2.0.0] - 2019-05-04
@@ -371,8 +385,10 @@ To many PR for this update, please check [here](https://github.com/tensorlayer/t
   - AtrousDeConv2dLayer added (PR #662)
   - Fix bugs of using `tf.layers` in CNN (PR #686)
 - Optimizer:
+
   - AMSGrad Optimizer added based on `On the Convergence of Adam and Beyond (ICLR 2018)` (PR #636)
 - Setup:
+
   - Creation of installation flaggs `all`, `all_cpu`, and `all_gpu` (PR #660)
 - Test:
   - `test_utils_predict.py` added to reproduce and fix issue #288 (PR #566)

examples/README.md

@@ -1 +1,11 @@
+<br/>
+
+<a href="https://join.slack.com/t/tensorlayer/shared_invite/enQtMjUyMjczMzU2Njg4LWI0MWU0MDFkOWY2YjQ4YjVhMzI5M2VlZmE4YTNhNGY1NjZhMzUwMmQ2MTc0YWRjMjQzMjdjMTg2MWQ2ZWJhYzc" target="\_blank">
+	<div align="center">
+		<img src="../img/join_slack.png" width="40%"/>
+	</div>
+</a>
+
+<br/>
+
 # [Click Me](https://github.com/tensorlayer/awesome-tensorlayer)

examples/basic_tutorials/tutorial_cifar10_cnn_static.py

@@ -5,11 +5,10 @@
 import time
 
 import numpy as np
-import tensorflow as tf
 
+import tensorflow as tf
 import tensorlayer as tl
-from tensorlayer.layers import (BatchNorm, Conv2d, Dense, Flatten, Input,
-                                LocalResponseNorm, MaxPool2d)
+from tensorlayer.layers import (BatchNorm, Conv2d, Dense, Flatten, Input, LocalResponseNorm, MaxPool2d)
 from tensorlayer.models import Model
 
 # enable debug logging
@@ -74,15 +73,14 @@ def get_model_batchnorm(inputs_shape):
 # get the network
 net = get_model([None, 24, 24, 3])
 
-
 # training settings
 batch_size = 128
 n_epoch = 50000
 learning_rate = 0.0001
 print_freq = 5
 n_step_epoch = int(len(y_train) / batch_size)
 n_step = n_epoch * n_step_epoch
-shuffle_buffer_size = 128 # 100
+shuffle_buffer_size = 128  # 100
 # init_learning_rate = 0.1
 # learning_rate_decay_factor = 0.1
 # num_epoch_decay = 350

examples/basic_tutorials/tutorial_mnist_mlp_dynamic.py

@@ -1,8 +1,8 @@
 import time
 
 import numpy as np
-import tensorflow as tf
 
+import tensorflow as tf
 import tensorlayer as tl
 from tensorlayer.layers import Dense, Dropout, Input
 from tensorlayer.models import Model
@@ -13,18 +13,19 @@
 ## prepare MNIST data
 X_train, y_train, X_val, y_val, X_test, y_test = tl.files.load_mnist_dataset(shape=(-1, 784))
 
+
 ## define the network
 class CustomModel(Model):
 
     def __init__(self):
         super(CustomModel, self).__init__()
 
-        self.dropout1 = Dropout(keep=0.8)#(self.innet)
-        self.dense1 = Dense(n_units=800, act=tf.nn.relu, in_channels=784)#(self.dropout1)
-        self.dropout2 = Dropout(keep=0.8)#(self.dense1)
-        self.dense2 = Dense(n_units=800, act=tf.nn.relu, in_channels=800)#(self.dropout2)
-        self.dropout3 = Dropout(keep=0.8)#(self.dense2)
-        self.dense3 = Dense(n_units=10, act=tf.nn.relu, in_channels=800)#(self.dropout3)
+        self.dropout1 = Dropout(keep=0.8)  #(self.innet)
+        self.dense1 = Dense(n_units=800, act=tf.nn.relu, in_channels=784)  #(self.dropout1)
+        self.dropout2 = Dropout(keep=0.8)  #(self.dense1)
+        self.dense2 = Dense(n_units=800, act=tf.nn.relu, in_channels=800)  #(self.dropout2)
+        self.dropout3 = Dropout(keep=0.8)  #(self.dense2)
+        self.dense3 = Dense(n_units=10, act=tf.nn.relu, in_channels=800)  #(self.dropout3)
 
     def forward(self, x, foo=None):
         z = self.dropout1(x)
@@ -37,6 +38,7 @@ def forward(self, x, foo=None):
             out = tf.nn.relu(out)
         return out
 
+
 MLP = CustomModel()
 
 ## start training

examples/basic_tutorials/tutorial_mnist_mlp_dynamic_2.py

@@ -1,8 +1,8 @@
 import time
 
 import numpy as np
-import tensorflow as tf
 
+import tensorflow as tf
 import tensorlayer as tl
 from tensorlayer.layers import Dense, Dropout, Input, LayerList
 from tensorlayer.models import Model
@@ -13,13 +13,14 @@
 ## prepare MNIST data
 X_train, y_train, X_val, y_val, X_test, y_test = tl.files.load_mnist_dataset(shape=(-1, 784))
 
+
 ## define the network
 class CustomModelHidden(Model):
 
     def __init__(self):
         super(CustomModelHidden, self).__init__()
 
-        self.dropout1 = Dropout(keep=0.8)#(self.innet)
+        self.dropout1 = Dropout(keep=0.8)  #(self.innet)
 
         self.seq = LayerList(
             [
@@ -29,14 +30,15 @@ def __init__(self):
             ]
         )
 
-        self.dropout3 = Dropout(keep=0.8)#(self.seq)
+        self.dropout3 = Dropout(keep=0.8)  #(self.seq)
 
     def forward(self, x):
         z = self.dropout1(x)
         z = self.seq(z)
         z = self.dropout3(z)
         return z
 
+
 class CustomModelOut(Model):
 
     def __init__(self):

examples/basic_tutorials/tutorial_mnist_mlp_static.py

@@ -1,8 +1,8 @@
 import time
 
 import numpy as np
-import tensorflow as tf
 
+import tensorflow as tf
 import tensorlayer as tl
 from tensorlayer.layers import Dense, Dropout, Input
 from tensorlayer.models import Model
@@ -21,14 +21,18 @@
 def get_model(inputs_shape):
     ni = Input(inputs_shape)
     nn = Dropout(keep=0.8)(ni)
-    nn = Dense(n_units=800, act=tf.nn.relu)(nn) # in_channels is optional in this case as it can be inferred by the previous layer
+    nn = Dense(n_units=800,
+               act=tf.nn.relu)(nn)  # in_channels is optional in this case as it can be inferred by the previous layer
     nn = Dropout(keep=0.8)(nn)
-    nn = Dense(n_units=800, act=tf.nn.relu)(nn) # in_channels is optional in this case as it can be inferred by the previous layer
+    nn = Dense(n_units=800,
+               act=tf.nn.relu)(nn)  # in_channels is optional in this case as it can be inferred by the previous layer
     nn = Dropout(keep=0.8)(nn)
-    nn = Dense(n_units=10, act=tf.nn.relu)(nn) # in_channels is optional in this case as it can be inferred by the previous layer
+    nn = Dense(n_units=10,
+               act=tf.nn.relu)(nn)  # in_channels is optional in this case as it can be inferred by the previous layer
     M = Model(inputs=ni, outputs=nn, name="mlp")
     return M
 
+
 MLP = get_model([None, 784])
 import pprint
 pprint.pprint(MLP.config)

examples/basic_tutorials/tutorial_mnist_mlp_static_2.py

@@ -1,8 +1,8 @@
 import time
 
 import numpy as np
-import tensorflow as tf
 
+import tensorflow as tf
 import tensorlayer as tl
 from tensorlayer.layers import Dense, Dropout, Input
 from tensorlayer.models import Model
@@ -13,12 +13,12 @@
 ## prepare MNIST data
 X_train, y_train, X_val, y_val, X_test, y_test = tl.files.load_mnist_dataset(shape=(-1, 784))
 
-
 ## define the network
 # the softmax is implemented internally in tl.cost.cross_entropy(y, y_) to
 # speed up computation, so we use identity here.
 # see tf.nn.sparse_softmax_cross_entropy_with_logits()
 
+
 def hidden_model(inputs_shape):
     ni = Input(inputs_shape)
     nn = Dropout(keep=0.8)(ni)
@@ -28,6 +28,7 @@ def hidden_model(inputs_shape):
 
     return Model(inputs=ni, outputs=nn, name="mlp_hidden")
 
+
 def get_model(inputs_shape, hmodel):
     hidden = hmodel.as_layer()
     ni = Input(inputs_shape)
@@ -37,6 +38,7 @@ def get_model(inputs_shape, hmodel):
 
     return Model(inputs=ni, outputs=nn, name="mlp")
 
+
 MLP_hidden = hidden_model([None, 784])
 MLP = get_model([None, 784], MLP_hidden)
 # MLP.print_layers()

examples/basic_tutorials/tutorial_mnist_siamese.py

@@ -14,8 +14,8 @@
 import time
 
 import numpy as np
-import tensorflow as tf
 
+import tensorflow as tf
 import tensorlayer as tl
 from tensorlayer.layers import Dense, Dropout, Flatten, Input
 from tensorlayer.models import Model
@@ -33,7 +33,7 @@ def contrastive_loss(label, feature1, feature2):
 
 
 def compute_accuracy(label, feature1, feature2):
-    eucd = tf.sqrt(tf.reduce_sum((feature1 - feature2) ** 2, axis=1))
+    eucd = tf.sqrt(tf.reduce_sum((feature1 - feature2)**2, axis=1))
     pred = tf.cast(eucd < 0.5, label.dtype)
     return tf.reduce_mean(tf.cast(tf.equal(pred, label), tf.float32))
 

examples/basic_tutorials/tutorial_mnist_simple.py

@@ -1,9 +1,10 @@
 #! /usr/bin/python
 # -*- coding: utf-8 -*-
 
+import numpy as np
+
 import tensorflow as tf
 import tensorlayer as tl
-import numpy as np
 
 tl.logging.set_verbosity(tl.logging.DEBUG)
 

examples/data_process/data/__init__.py

@@ -1,4 +1,3 @@
 from __future__ import absolute_import
 
-# from . import imagenet_classes
 from . import *

examples/data_process/tutorial_fast_affine_transform.py

@@ -8,10 +8,10 @@
 import multiprocessing
 import time
 
-import cv2
 import numpy as np
-import tensorflow as tf
 
+import cv2
+import tensorflow as tf
 import tensorlayer as tl
 
 # tl.logging.set_verbosity(tl.logging.DEBUG)

examples/data_process/tutorial_tf_dataset_voc.py

@@ -13,8 +13,8 @@
 import time
 
 import numpy as np
-import tensorflow as tf
 
+import tensorflow as tf
 import tensorlayer as tl
 
 # tf.logging.set_verbosity(tf.logging.DEBUG)
@@ -108,5 +108,6 @@ def _map_fn(filename, annotation):
 ## save all images
 for i in range(len(im)):
     print(ann[i][1])
-    tl.vis.draw_boxes_and_labels_to_image(im[i] * 255, ann[i][0], ann[i][1], [], classes,
-                                          True, save_name='_bbox_vis_%d.png' % i)
+    tl.vis.draw_boxes_and_labels_to_image(
+        im[i] * 255, ann[i][0], ann[i][1], [], classes, True, save_name='_bbox_vis_%d.png' % i
+    )

examples/data_process/tutorial_tfrecord.py

@@ -22,9 +22,9 @@
 import os
 
 import numpy as np
-import tensorflow as tf
 from PIL import Image
 
+import tensorflow as tf
 import tensorlayer as tl
 
 ## Save data ==================================================================
@@ -79,8 +79,7 @@ def read_and_decode(filename):
     raw_dataset = tf.data.TFRecordDataset([filename]).shuffle(1000).batch(4)
     for serialized_example in raw_dataset:
         features = tf.io.parse_example(
-            serialized_example,
-            features={
+            serialized_example, features={
                 'label': tf.io.FixedLenFeature([], tf.int64),
                 'img_raw': tf.io.FixedLenFeature([], tf.string),
             }
@@ -97,4 +96,3 @@ def read_and_decode(filename):
 print("img_batch   : %s" % img_batch.shape)
 print("label_batch : %s" % label_batch.shape)
 tl.visualize.images2d(img_batch, second=1, saveable=False, name='batch', dtype=None, fig_idx=2020121)
-

examples/data_process/tutorial_tfrecord2.py

@@ -14,10 +14,10 @@
 import os
 
 import numpy as np
+
 # import matplotlib
 # matplotlib.use('GTK')
 import tensorflow as tf
-
 import tensorlayer as tl
 
 # Download data, and convert to TFRecord format, see ```tutorial_tfrecord.py```
@@ -77,6 +77,7 @@ def read_and_decode(filename):
         label_batch = tf.cast(features['label'], tf.int32)
         yield img_batch, label_batch
 
+
 img_batch, label_batch = next(read_and_decode("train.tfrecords"))
 print("img_batch   : %s" % img_batch.shape)
 print("label_batch : %s" % label_batch.shape)

examples/data_process/tutorial_tfrecord3.py

@@ -19,9 +19,9 @@
 import os
 
 import numpy as np
-import tensorflow as tf
 from PIL import Image
 
+import tensorflow as tf
 import tensorlayer as tl
 
 

examples/database/dispatch_tasks.py

@@ -6,7 +6,6 @@
 import time
 
 import tensorflow as tf
-
 import tensorlayer as tl
 
 tl.logging.set_verbosity(tl.logging.DEBUG)

examples/database/task_script.py

@@ -1,7 +1,6 @@
 """Sample task script."""
 
 import tensorflow as tf
-
 import tensorlayer as tl
 
 tf.logging.set_verbosity(tf.logging.DEBUG)