深度学习-TensorFlow实例

1. 安装tensorflow1.9前，安装python3.6，安装环境为ubuntu18，

2.下载tensorflow 的whl文件，安装pip，

链接:https://pan.baidu.com/s/1DdAziXr-Sx4f9mwWgGRo0w 密码:lz3g

然后执行：pip3 install whl文件

3. tensorflow不能跳过的hello word

import tensorflow as tf
print(tf.__version__)

结果：1.9.0

4.tensorflow 第一DAG使用（有向无环图），和spark的DAG是一回事

import tensorflow as tf

# with code run at cpu 0
with tf.device("/cpu:0"):
    x = tf.Variable(3, name="x")
y = tf.Variable(4, name="y")
f = x * y + 3


# init session
session = tf.Session(config=tf.ConfigProto(log_device_placement=True))

# start init x and y , function programer
session.run(x.initializer)
session.run(y.initializer)

# run f function
result = session.run(f)

print(result)
# session close
session.close()

结果：包括执行过程

x: (VariableV2): /job:localhost/replica:0/task:0/device:CPU:0
x/Assign: (Assign): /job:localhost/replica:0/task:0/device:CPU:0
x/read: (Identity): /job:localhost/replica:0/task:0/device:CPU:0
y: (VariableV2): /job:localhost/replica:0/task:0/device:CPU:0
y/Assign: (Assign): /job:localhost/replica:0/task:0/device:CPU:0
y/read: (Identity): /job:localhost/replica:0/task:0/device:CPU:0
mul: (Mul): /job:localhost/replica:0/task:0/device:CPU:0
add: (Add): /job:localhost/replica:0/task:0/device:CPU:0
x/initial_value: (Const): /job:localhost/replica:0/task:0/device:CPU:0
y/initial_value: (Const): /job:localhost/replica:0/task:0/device:CPU:0
add/y: (Const): /job:localhost/replica:0/task:0/device:CPU:0
15

tensorflow 比你想象第一个例子要简单。