name_scope

variable_scope

scope (name_scope/variable_scope)
from __future__ import print_function
import tensorflow as tf with tf.name_scope("a_name_scope"):
initializer = tf.constant_initializer(value=1)
var1 = tf.get_variable(name='var1', shape=[1], dtype=tf.float32, initializer=initializer)
var2 = tf.Variable(name='var2', initial_value=[2], dtype=tf.float32)
var21 = tf.Variable(name='var2', initial_value=[2.1], dtype=tf.float32)
var22 = tf.Variable(name='var2', initial_value=[2.2], dtype=tf.float32) with tf.Session() as sess:
sess.run(tf.initialize_all_variables())
print(var1.name) # var1:0 此种get_variable对于name_scope无效
print(sess.run(var1)) # [ 1.]
print(var2.name) # a_name_scope/var2:0
print(sess.run(var2)) # [ 2.]
print(var21.name) # a_name_scope/var2_1:0
print(sess.run(var21)) # [ 2.0999999]
print(var22.name) # a_name_scope/var2_2:0
print(sess.run(var22)) # [ 2.20000005] with tf.variable_scope("a_variable_scope") as scope:
initializer = tf.constant_initializer(value=3)
var3 = tf.get_variable(name='var3', shape=[1], dtype=tf.float32, initializer=initializer)
var4 = tf.Variable(name='var4', initial_value=[4], dtype=tf.float32)
var4_reuse = tf.Variable(name='var4', initial_value=[4], dtype=tf.float32)
scope.reuse_variables() #定义了可重复利用
var3_reuse = tf.get_variable(name='var3',) with tf.Session() as sess:
# tf.initialize_all_variables() no long valid from
# 2017-03-02 if using tensorflow >= 0.12
if int((tf.__version__).split('.')[1]) < 12 and int((tf.__version__).split('.')[0]) < 1:
init = tf.initialize_all_variables()
else:
init = tf.global_variables_initializer()
sess.run(init)
print(var3.name) # a_variable_scope/var3:0
print(sess.run(var3)) # [ 3.]
print(var4.name) # a_variable_scope/var4:0
print(sess.run(var4)) # [ 4.]
print(var4_reuse.name) # a_variable_scope/var4_1:0
print(sess.run(var4_reuse)) # [ 4.]
print(var3_reuse.name) # a_variable_scope/var3:0
print(sess.run(var3_reuse)) # [ 3.]

通常在RNN中有一个重复循环机制,比如在training中和test中的结构是不同的,但是在两者的参数是相同的时候,就可以用到

scope.reuse_variables()
# visit https://morvanzhou.github.io/tutorials/ for more!

# 22 scope (name_scope/variable_scope)
from __future__ import print_function
import tensorflow as tf class TrainConfig:
batch_size = 20
time_steps = 20
input_size = 10
output_size = 2
cell_size = 11
learning_rate = 0.01 class TestConfig(TrainConfig):
time_steps = 1 class RNN(object): def __init__(self, config):
self._batch_size = config.batch_size
self._time_steps = config.time_steps
self._input_size = config.input_size
self._output_size = config.output_size
self._cell_size = config.cell_size
self._lr = config.learning_rate
self._built_RNN() def _built_RNN(self):
with tf.variable_scope('inputs'):
self._xs = tf.placeholder(tf.float32, [self._batch_size, self._time_steps, self._input_size], name='xs')
self._ys = tf.placeholder(tf.float32, [self._batch_size, self._time_steps, self._output_size], name='ys')
with tf.name_scope('RNN'):
with tf.variable_scope('input_layer'):
l_in_x = tf.reshape(self._xs, [-1, self._input_size], name='2_2D') # (batch*n_step, in_size)
# Ws (in_size, cell_size)
Wi = self._weight_variable([self._input_size, self._cell_size])
print(Wi.name)
# bs (cell_size, )
bi = self._bias_variable([self._cell_size, ])
# l_in_y = (batch * n_steps, cell_size)
with tf.name_scope('Wx_plus_b'):
l_in_y = tf.matmul(l_in_x, Wi) + bi
l_in_y = tf.reshape(l_in_y, [-1, self._time_steps, self._cell_size], name='2_3D') with tf.variable_scope('cell'):
cell = tf.contrib.rnn.BasicLSTMCell(self._cell_size)
with tf.name_scope('initial_state'):
self._cell_initial_state = cell.zero_state(self._batch_size, dtype=tf.float32) self.cell_outputs = []
cell_state = self._cell_initial_state
for t in range(self._time_steps):
if t > 0: tf.get_variable_scope().reuse_variables()
cell_output, cell_state = cell(l_in_y[:, t, :], cell_state)
self.cell_outputs.append(cell_output)
self._cell_final_state = cell_state with tf.variable_scope('output_layer'):
# cell_outputs_reshaped (BATCH*TIME_STEP, CELL_SIZE)
cell_outputs_reshaped = tf.reshape(tf.concat(1, self.cell_outputs), [-1, self._cell_size])
Wo = self._weight_variable((self._cell_size, self._output_size))
bo = self._bias_variable((self._output_size,))
product = tf.matmul(cell_outputs_reshaped, Wo) + bo
# _pred shape (batch*time_step, output_size)
self._pred = tf.nn.relu(product) # for displacement with tf.name_scope('cost'):
_pred = tf.reshape(self._pred, [self._batch_size, self._time_steps, self._output_size])
mse = self.ms_error(_pred, self._ys)
mse_ave_across_batch = tf.reduce_mean(mse, 0)
mse_sum_across_time = tf.reduce_sum(mse_ave_across_batch, 0)
self._cost = mse_sum_across_time
self._cost_ave_time = self._cost / self._time_steps with tf.name_scope('trian'):
self._lr = tf.convert_to_tensor(self._lr)
self.train_op = tf.train.AdamOptimizer(self._lr).minimize(self._cost) @staticmethod
def ms_error(y_pre, y_target):
return tf.square(tf.sub(y_pre, y_target)) @staticmethod
def _weight_variable(shape, name='weights'):
initializer = tf.random_normal_initializer(mean=0., stddev=0.5, )
return tf.get_variable(shape=shape, initializer=initializer, name=name) @staticmethod
def _bias_variable(shape, name='biases'):
initializer = tf.constant_initializer(0.1)
return tf.get_variable(name=name, shape=shape, initializer=initializer) if __name__ == '__main__':
train_config = TrainConfig()
test_config = TestConfig() # the wrong method to reuse parameters in train rnn
with tf.variable_scope('train_rnn'):
train_rnn1 = RNN(train_config) #参数在train和test都是一致的
with tf.variable_scope('test_rnn'):
test_rnn1 = RNN(test_config) #参数在train和test都是一致的

# the right method to reuse parameters in train rnn
  with tf.variable_scope('rnn') as scope:
    sess = tf.Session()
    train_rnn2 = RNN(train_config)
    scope.reuse_variables()
    test_rnn2 = RNN(test_config)
    # tf.initialize_all_variables() no long valid from
    # 2017-03-02 if using tensorflow >= 0.12
    if int((tf.__version__).split('.')[1]) < 12 and int((tf.__version__).split('.')[0]) < 1:
      init = tf.initialize_all_variables()
    else:
      init = tf.global_variables_initializer()
    sess.run(init)


												

2.4scope的更多相关文章

随机推荐

  1. [原]unity3d GLSL无法pc上显示

    pc编写GLSL: Shader "Custom/GLSL" {    Properties {        _MainTex ("Base (RGB)", ...

  2. UINavigationController popToViewController用法

    popToViewController用法  [self.navigationController popToViewController:[self.navigationController.vie ...

  3. Exception occurred while processing this request, check the log for more information!安装ActiveMq-5.14.1 配置安全验证报错解决

    安装ActiveMq-5.14.1  并配置了安全验证成功后,客户端也连接成功了.服务端也能通过http://IP:8161登录到控制台. 但是在点击队列,想要查看队列视图时报错,如下图: 查看日志发 ...

  4. c++的字节对齐

    win32平台下的微软C编译器对齐策略: 1)结构体变量的首地址能够被其最宽数据类型成员的大小整除.编译器在为结构体变量开辟空间时,首先找到结构体中最宽的数据类型,然后寻找内存地址能被该数据类型大小整 ...

  5. MQTT服务器搭建--Mosquitto用户名密码配置

    Mosquitto用户认证配置 前言:基于Mosquitto服务器已经搭建成功,大部分都是采用默认的是允许匿名用户登录模式,正式上线的系统需要进行用户认证. 1.用户参数说明 Mosquitto服务器 ...

  6. Python进阶 学习笔记(二)

    (涉及内容:面向对象,类的继承) 定义类并创建实例 在Python中,类通过 class 关键字定义.以 Person 为例,定义一个Person类如下: class Person(object): ...

  7. jquery前端验证框架

    1.validationEngine.jquery.css  样式包 2.jquery.validationEngine-zh_CN.js 中文语言包 3.jquery.validationEngin ...

  8. JavaScript的格式--从格式做起,做最严谨的工程师

    1.JavaScript的格式: JavaScript区分大小写: JavaScript脚本程序须嵌入在HTML文件中: JavaScript脚本程序中不能包含HTML标记代码:(双引号) 每行写一条 ...

  9. 【Laravel5.5】 laravel5 数据库配置(MySQL)

    1   进入laravel根目录.      在config目录下找到database.php文件.      显而易见这个文件是数据库相关的配置文件. 2  修改 .env 配置完database. ...

  10. scanf printf gets() puts(),cin cout

    最近在练机试题,常用的C和C++输入输出如下: 1 scanf 和printf int a; scanf("%d",&a) ; printf("%d", ...