1,TensorFlow只支持64位平台
2,下载python 3.5.X (必须是64位的),地址如下:
https://www.python.org/ftp/python/3.5.2/python-3.5.2-amd64.exe
https://www.python.org/ftp/python/3.5.2/python-3.5.2-embed-amd64.zip
3,安装python,并验证python和pip
C:\Users\Administrator>python -V
Python 3.5.2
C:\Users\Administrator>pip3 -V
pip3 8.1.1 from d:\program files\python35\lib\site-packages (python 3.5)
4,安装TensorFlow(only cpu)
pip3 install --upgrade tensorflow
5,TensorFlow的hello world
C:\Users\Administrator>python
Python 3.5.2 (v3.5.2:4def2a2901a5, Jun 25 2016, 22:18:55) [MSC v.1900 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
>>> hello = tf.constant('Hello tensorflow1')
>>> session = tf.Session()
>>> print(session.run(hello))
b'Hello tensorflow1'
看到输出b'Hello tensorflow1'即证明TensorFlow环境安装完成.
6,如果你是java开发的话,可能想尝试下java API,结果很让人失望,太慢了,世界末日了,代码如下,自己去享受一把:
public String tf() throws Exception{
final String value = "Hello tensorflow version : " + TensorFlow.version();
System.out.println(value);
//将数据转为tensor
Tensor tensor = Tensor.create(value.getBytes("UTF-8"));
//构造一个图对象
Graph graph = new Graph();
graph.opBuilder("Const","myConst").setAttr("dtype",tensor.dataType())
.setAttr("value",tensor).build();
//创建一个Session
Session session = new Session(graph);
graph.close();
//计算输出结果
Tensor result = session.runner().fetch("myConst").run().get(0);
System.out.println(result);
return new String(result.bytesValue(),"UTF-8");
}
