做网站用什么语言数据库下列哪些店铺适合交换友情链接

github地址https://github.com/fz861062923/TensorFlow

建立’计算图’

#建立‘计算图’
import tensorflow as tf
x=tf.constant(2,name='x')#建立常量，有点像C
y=tf.Variable(x+5,name='y')#建立变量

#执行‘计算图’
with tf.Session() as sess:init=tf.global_variables_initializer()#初始化global变量sess.run(init)print('x=',sess.run(x))print('y=',sess.run(y))

x= 2
y= 7

x

<tf.Tensor 'x:0' shape=() dtype=int32>

tensorflow placeholder

正如这个名字一样，hold on,hold on,告诉计算机等等在把值传给你，嘻嘻嘻嘻

a=tf.placeholder('int32')
b=tf.placeholder('int32')
c=tf.multiply(a,b)
with tf.Session() as sess:init=tf.global_variables_initializer()sess.run(init)print('c=',sess.run(c,feed_dict={a:6,b:7}))

c= 42

tensorflow数值运算常用的方法

• tf.add(x,y)
• tf.subtract(x,y)#减法
• tf.multiply(x,y)
• tf.divide(x,y)
• tf.mod(x,y)#余数
• tf.sqrt(x,name=None)
• tf.abs(x,name=None)

tensorboard

正如其名，可视化已经建立的计算图

#承接上面的session
#下面代码将显示在tensorboard的数据写在log文件中
tf.summary.merge_all()#将显示在board的数据整合
train_writer=tf.summary.FileWriter('log/c',sess.graph)#写入log文件中

启动tensorboard的方法

• activate tensorflow(虚拟环境名称)
• tensorboard --logdir=c:\python\log\c
• 用浏览器打开http://lacalhost:6006/

建立一维与二维张量

建立一维张量

ts_x=tf.Variable([0.4,0.2,0.4])
with tf.Session() as sess:init=tf.global_variables_initializer()sess.run(init)x=sess.run(ts_x)print(x)

[0.4 0.2 0.4]

x.shape

(3,)

建立二维张量

ts_x=tf.Variable([[0.4,0.2,0.4]])
with tf.Session() as sess:init=tf.global_variables_initializer()sess.run(init)x=sess.run(ts_x)print(x)

[[0.4 0.2 0.4]]

x.shape

(1, 3)

建立新的二维张量

ts_x=tf.Variable([[0.4,0.2],[0.3,0.4],[-0.5,0.2]])
with tf.Session() as sess:init=tf.global_variables_initializer()sess.run(init)x=sess.run(ts_x)print(x)

[[ 0.4  0.2][ 0.3  0.4][-0.5  0.2]]

x.shape

(3, 2)

矩阵的基本运算

矩阵的加法

x=tf.Variable([[1.,1.,1.]])
w=tf.Variable([[-0.1,-0.2],[-0.3,0.4],[0.5,0.6]])
xw=tf.matmul(x,w)with tf.Session() as sess:init=tf.global_variables_initializer()sess.run(init)print(sess.run(xw))

[[0.09999999 0.8       ]]

矩阵乘法与加法

x=tf.Variable([[1.,1.,1.]])
w=tf.Variable([[-0.1,-0.2],[-0.3,0.4],[0.5,0.6]])
b=tf.Variable([[0.1,0.2]])
xwb=tf.matmul(x,w)+bwith tf.Session() as sess:init=tf.global_variables_initializer()sess.run(init)print(sess.run(xwb))

[[0.19999999 1.        ]]