tensorflow1.14.0和cuda10.0.0兼容性比较好,建议安装这两个版本

1、下载CUDA
链接:https://developer.nvidia.com/cuda-downloads?target_os=Windows&target_arch=x86_64&target_version=10&target_type=exelocal

2、下载CUDNN
需要注册登录才能下载
https://developer.nvidia.com/rdp/cudnn-archive

3、安装
如果第一步安装CUDA没有修改安装路径,执行以下操作:
复制 cudnn-10.0-windows10-x64-v7.6.5.32\cuda\bin\cudnn64_7.dll 到 C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\bin
复制 cudnn-10.0-windows10-x64-v7.6.5.32\cuda\include\cudnn.h 到 C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\include
复制 cudnn-10.0-windows10-x64-v7.6.5.32\cuda\lib\x64\cudnn.lib 到 C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\lib\x64\

  • 添加环境变量:
    将如下路径添加到环境变量path中
    C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\bin

4、TensorFlow-gpu安装

import tensorflow as tf
print(tf.test.is_gpu_available())
  • 测试2
import  tensorflow as tf 
a = tf.constant([1.0,2.0,3.0],shape = [3], name='a')
b = tf.constant([1.0,2.0,3.0], shape = [3], name='b')
c = a +b
sess = tf.Session(config = tf.ConfigProto(log_device_placement =True))
print(sess.run(c))

4、CPU版本的安装

pip install --no-cache-dir -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow

参考链接:https://blog.csdn.net/qq_37277944/article/details/82717796