Fix build test. (#1057)

* 2.2.1 release

* Fix build test.

* apply yapf

* ping yapf to 0.28.0

* fix build

* use yapf 0.29

* fix yapf

* include tests in make format.

* ping autoflake and isort version.

CHANGELOG.md

@@ -80,16 +80,17 @@ To release a new version, please update the changelog as followed:
 
 ### Fixed
 
-- Fix README. (#PR 1044)
-- Fix package info. (#PR 1046)
+- Fix README. (#1044)
+- Fix package info. (#1046)
+- Fix build test (Using YAPF 0.29) (#1057)
 
 ### Removed
 
 ### Security
 
 ### Contributors
 
-- @luomai (PR #1044, 1046)
+- @luomai (#1044, #1046, #1057)
 
 
 ## [2.2.0] - 2019-09-13
@@ -150,7 +151,7 @@ This release is compatible with TensorFlow 2 RC1.
 - Replace tf.nn.func with tf.nn.func.\_\_name\_\_ in model config. (PR #994)
 - Add Reinforcement learning tutorials. (PR #995)
 - Add RNN layers with simple rnn cell, GRU cell, LSTM cell. (PR #998)
-- Update Seq2seq (#998) 
+- Update Seq2seq (#998)
 - Add Seq2seqLuongAttention model (#998)
 
 ### Fixed

Makefile

@@ -14,16 +14,17 @@ test:
 	python3 tests/files/test_utils_saveload.py
 
 format:
-	autoflake -i examples/*.py
-	autoflake -i tensorlayer/*.py
-	autoflake -i tensorlayer/**/*.py
+	autoflake -ir examples
+	autoflake -ir tensorlayer
+	autoflake -ir tests
 
 	isort -rc examples
 	isort -rc tensorlayer
+	isort -rc tests
 
-	yapf -i examples/*.py
-	yapf -i tensorlayer/*.py
-	yapf -i tensorlayer/**/*.py
+	yapf -ir examples
+	yapf -ir tensorlayer
+	yapf -ir tests
 
 install3:
 	pip3 install -U . --user

examples/basic_tutorials/tutorial_cifar10_cnn_static.py

@@ -1,9 +1,10 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 
+import multiprocessing
 import time
+
 import numpy as np
-import multiprocessing
 import tensorflow as tf
 
 import tensorlayer as tl
@@ -123,7 +124,7 @@ def _map_fn_train(img, target):
 
 def _map_fn_test(img, target):
     # 1. Crop the central [height, width] of the image.
-    img = tf.image.resize_with_pad(img, 24, 24)  
+    img = tf.image.resize_with_pad(img, 24, 24)
     # 2. Subtract off the mean and divide by the variance of the pixels.
     img = tf.image.per_image_standardization(img)
     img = tf.reshape(img, (24, 24, 3))

examples/basic_tutorials/tutorial_mnist_mlp_dynamic.py

@@ -1,4 +1,5 @@
 import time
+
 import numpy as np
 import tensorflow as tf
 

examples/basic_tutorials/tutorial_mnist_mlp_dynamic_2.py

@@ -1,4 +1,5 @@
 import time
+
 import numpy as np
 import tensorflow as tf
 

examples/basic_tutorials/tutorial_mnist_mlp_static.py

@@ -1,5 +1,6 @@
 import pprint
 import time
+
 import numpy as np
 import tensorflow as tf
 

examples/basic_tutorials/tutorial_mnist_mlp_static_2.py

@@ -1,4 +1,5 @@
 import time
+
 import numpy as np
 import tensorflow as tf
 

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 tensorlayer as tl
 
 # tl.logging.set_verbosity(tl.logging.DEBUG)
@@ -98,7 +98,6 @@ def _map_fn(image_path, target):
     st = time.time()
     for img, target in dataset:
         n_step += 1
-        pass
     assert n_step == n_epoch * n_data / batch_size
     print("dataset APIs took %fs for each image" % ((time.time() - st) / batch_size / n_step))  # CPU ~ 100%
 

examples/data_process/tutorial_tfrecord3.py

@@ -231,8 +231,8 @@ def distort_image(image, thread_id):
 
 
 def prefetch_input_data(
-        reader, file_pattern, is_training, batch_size, values_per_shard, input_queue_capacity_factor=16,
-        num_reader_threads=1, shard_queue_name="filename_queue", value_queue_name="input_queue"
+    reader, file_pattern, is_training, batch_size, values_per_shard, input_queue_capacity_factor=16,
+    num_reader_threads=1, shard_queue_name="filename_queue", value_queue_name="input_queue"
 ):
     """Prefetches string values from disk into an input queue.
 

examples/pretrained_cnn/tutorial_models_resnet50.py

@@ -8,8 +8,8 @@
 import time
 
 import numpy as np
-
 import tensorflow as tf
+
 import tensorlayer as tl
 from tensorlayer.models.imagenet_classes import class_names
 

examples/reinforcement_learning/baselines/algorithms/a3c/a3c.py

@@ -40,16 +40,15 @@
 
 """
 
-import argparse
 import multiprocessing
 import threading
 import time
 
-import gym
 import numpy as np
 import tensorflow as tf
-import tensorflow_probability as tfp
 
+import gym
+import tensorflow_probability as tfp
 import tensorlayer as tl
 from common.buffer import *
 from common.networks import *
@@ -66,8 +65,8 @@
 class ACNet(object):
 
     def __init__(
-            self, scope, entropy_beta, action_dim, state_dim, actor_hidden_dim, actor_hidden_layer, critic_hidden_dim,
-            critic_hidden_layer, action_bound, globalAC=None
+        self, scope, entropy_beta, action_dim, state_dim, actor_hidden_dim, actor_hidden_layer, critic_hidden_dim,
+        critic_hidden_layer, action_bound, globalAC=None
     ):
         self.scope = scope  # the scope is for naming networks for each worker differently
         self.save_path = './model'
@@ -107,7 +106,7 @@ def __init__(
 
     @tf.function  # convert numpy functions to tf.Operations in the TFgraph, return tensor
     def update_global(
-            self, buffer_s, buffer_a, buffer_v_target, globalAC
+        self, buffer_s, buffer_a, buffer_v_target, globalAC
     ):  # refer to the global Actor-Crtic network for updating it with samples
         ''' update the global critic '''
         with tf.GradientTape() as tape:
@@ -164,8 +163,8 @@ def load_ckpt(self):  # load trained weights
 class Worker(object):
 
     def __init__(
-            self, env_id, name, globalAC, train_episodes, gamma, update_itr, entropy_beta, action_dim, state_dim,
-            actor_hidden_dim, actor_hidden_layer, critic_hidden_dim, critic_hidden_layer, action_bound
+        self, env_id, name, globalAC, train_episodes, gamma, update_itr, entropy_beta, action_dim, state_dim,
+        actor_hidden_dim, actor_hidden_layer, critic_hidden_dim, critic_hidden_layer, action_bound
     ):
         self.env = make_env(env_id)
         self.name = name
@@ -242,9 +241,9 @@ def work(self, globalAC):
 
 
 def learn(
-        env_id, train_episodes, test_episodes=1000, max_steps=150, number_workers=0, update_itr=10, gamma=0.99,
-        entropy_beta=0.005, actor_lr=5e-5, critic_lr=1e-4, actor_hidden_dim=300, actor_hidden_layer=2,
-        critic_hidden_dim=300, critic_hidden_layer=2, seed=2, save_interval=500, mode='train'
+    env_id, train_episodes, test_episodes=1000, max_steps=150, number_workers=0, update_itr=10, gamma=0.99,
+    entropy_beta=0.005, actor_lr=5e-5, critic_lr=1e-4, actor_hidden_dim=300, actor_hidden_layer=2,
+    critic_hidden_dim=300, critic_hidden_layer=2, seed=2, save_interval=500, mode='train'
 ):
     '''
     parameters

examples/reinforcement_learning/baselines/algorithms/ac/ac.py

@@ -41,13 +41,12 @@
 tensorlayer >=2.0.0
 
 """
-import argparse
 import time
 
-import gym
 import numpy as np
 import tensorflow as tf
 
+import gym
 import tensorlayer as tl
 from common.buffer import *
 from common.networks import *
@@ -125,9 +124,9 @@ def load_ckpt(self):  # load trained weights
 
 
 def learn(
-        env_id, train_episodes, test_episodes=1000, max_steps=1000, gamma=0.9, actor_lr=1e-3, critic_lr=1e-2,
-        actor_hidden_dim=30, actor_hidden_layer=1, critic_hidden_dim=30, critic_hidden_layer=1, seed=2,
-        save_interval=100, mode='train', render=False
+    env_id, train_episodes, test_episodes=1000, max_steps=1000, gamma=0.9, actor_lr=1e-3, critic_lr=1e-2,
+    actor_hidden_dim=30, actor_hidden_layer=1, critic_hidden_dim=30, critic_hidden_layer=1, seed=2, save_interval=100,
+    mode='train', render=False
 ):
     '''
     parameters

examples/reinforcement_learning/baselines/algorithms/ddpg/ddpg.py

@@ -19,14 +19,13 @@
 
 """
 
-import os
 import time
 
-import gym
 import matplotlib.pyplot as plt
 import numpy as np
 import tensorflow as tf
 
+import gym
 import tensorlayer as tl
 from common.buffer import *
 from common.networks import *
@@ -124,8 +123,7 @@ class DDPG(object):
     """
 
     def __init__(
-            self, a_dim, s_dim, hidden_dim, num_hidden_layer, a_bound, gamma, lr_a, lr_c, replay_buffer_size,
-            batch_size=32
+        self, a_dim, s_dim, hidden_dim, num_hidden_layer, a_bound, gamma, lr_a, lr_c, replay_buffer_size, batch_size=32
     ):
         self.memory = np.zeros((replay_buffer_size, s_dim * 2 + a_dim + 1), dtype=np.float32)
         self.pointer = 0
@@ -279,9 +277,9 @@ def load_ckpt(self):
 
 
 def learn(
-        env_id='Pendulum-v0', train_episodes=200, test_episodes=100, max_steps=200, save_interval=10, actor_lr=1e-3,
-        critic_lr=2e-3, gamma=0.9, hidden_dim=30, num_hidden_layer=1, seed=1, mode='train', render=False,
-        replay_buffer_size=10000, batch_size=32
+    env_id='Pendulum-v0', train_episodes=200, test_episodes=100, max_steps=200, save_interval=10, actor_lr=1e-3,
+    critic_lr=2e-3, gamma=0.9, hidden_dim=30, num_hidden_layer=1, seed=1, mode='train', render=False,
+    replay_buffer_size=10000, batch_size=32
 ):
     """
     learn function

examples/reinforcement_learning/baselines/algorithms/dppo/dppo.py

@@ -20,18 +20,16 @@
 
 """
 
-import argparse
-import os
 import queue
 import threading
 import time
 
-import gym
 import matplotlib.pyplot as plt
 import numpy as np
 import tensorflow as tf
-import tensorflow_probability as tfp
 
+import gym
+import tensorflow_probability as tfp
 import tensorlayer as tl
 from common.buffer import *
 from common.networks import *
@@ -50,8 +48,7 @@ class StochasticPolicyNetwork(Model):
     ''' stochastic continuous policy network for generating action according to the state '''
 
     def __init__(
-            self, state_dim, action_dim, hidden_list, a_bound, log_std_min=-20, log_std_max=2, scope=None,
-            trainable=True
+        self, state_dim, action_dim, hidden_list, a_bound, log_std_min=-20, log_std_max=2, scope=None, trainable=True
     ):
 
         # w_init = tf.keras.initializers.glorot_normal(
@@ -121,7 +118,7 @@ class PPO(object):
     '''
 
     def __init__(
-            self, a_dim, s_dim, hidden_list, a_max, actor_lr, critic_lr, a_update_steps, c_update_steps, save_interval
+        self, a_dim, s_dim, hidden_list, a_max, actor_lr, critic_lr, a_update_steps, c_update_steps, save_interval
     ):
         self.a_dim = a_dim
         self.s_dim = s_dim
@@ -392,9 +389,9 @@ def work(self):
 
 
 def learn(
-        env_id='Pendulum-v0', train_episodes=1000, test_episodes=100, max_steps=200, save_interval=10, actor_lr=1e-4,
-        critic_lr=2e-4, gamma=0.9, hidden_dim=100, num_hidden_layer=1, seed=1, mode='train', batch_size=32,
-        a_update_steps=10, c_update_steps=10, n_worker=4
+    env_id='Pendulum-v0', train_episodes=1000, test_episodes=100, max_steps=200, save_interval=10, actor_lr=1e-4,
+    critic_lr=2e-4, gamma=0.9, hidden_dim=100, num_hidden_layer=1, seed=1, mode='train', batch_size=32,
+    a_update_steps=10, c_update_steps=10, n_worker=4
 ):
     """
     learn function

examples/reinforcement_learning/baselines/algorithms/dqn/dqn.py

@@ -38,9 +38,9 @@ def learn(env_id, env_type, seed, mode, **kwargs):
 
 
 def core_learn(
-        env, mode, number_timesteps, network, optimizer, ob_scale, gamma, double_q, exploration_fraction,
-        exploration_final_eps, batch_size, learning_starts, target_network_update_freq, buffer_size, prioritized_replay,
-        prioritized_replay_alpha, prioritized_replay_beta0, save_path='dqn', save_interval=0, **kwargs
+    env, mode, number_timesteps, network, optimizer, ob_scale, gamma, double_q, exploration_fraction,
+    exploration_final_eps, batch_size, learning_starts, target_network_update_freq, buffer_size, prioritized_replay,
+    prioritized_replay_alpha, prioritized_replay_beta0, save_path='dqn', save_interval=0, **kwargs
 ):
     """
     Parameters:

examples/reinforcement_learning/baselines/algorithms/pg/pg.py

@@ -20,11 +20,11 @@
 """
 import time
 
-import gym
 import matplotlib.pyplot as plt
 import numpy as np
 import tensorflow as tf
 
+import gym
 import tensorlayer as tl
 from common.buffer import *
 from common.networks import *
@@ -165,18 +165,18 @@ def load(self, name='model'):
 
 
 def learn(
-        env_id='CartPole-v0',
-        train_episodes=3000,
-        test_episodes=1000,
-        max_steps=1000,
-        lr=0.02,
-        gamma=0.99,
-        hidden_dim=30,
-        num_hidden_layer=1,
-        seed=2,
-        save_interval=100,
-        mode='train',
-        render=False,
+    env_id='CartPole-v0',
+    train_episodes=3000,
+    test_episodes=1000,
+    max_steps=1000,
+    lr=0.02,
+    gamma=0.99,
+    hidden_dim=30,
+    num_hidden_layer=1,
+    seed=2,
+    save_interval=100,
+    mode='train',
+    render=False,
 ):
     """
     learn function