win7+cuda+anaconda python+tensorflow-gpu+keras安装成功版本匹配汇总
大家在安装配置过程中遇到了很多坑，其中大部分和软件之间的版本兼容性有关，在此，列出了不同软件版本之间的配置兼容性，方便安装配置。
https://github.com/fo40225/tensorflow-windows-wheel
| Path | Compiler | CUDA/cuDNN | SIMD | Notes |
|---|---|---|---|---|
| 1.14.0py37CPUsse2 | VS2019 16.1 | No | x86_64 | Python 3.7 |
| 1.14.0py37CPUavx2 | VS2019 16.1 | No | AVX2 | Python 3.7 |
| 1.14.0py37GPUcuda101cudnn76sse2 | VS2019 16.1 | 10.1.168_425.25/7.6.0.64 | x86_64 | Python 3.7/Compute 3.0 |
| 1.14.0py37GPUcuda101cudnn76avx2 | VS2019 16.1 | 10.1.168_425.25/7.6.0.64 | AVX2 | Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5 |
| 1.13.1py37CPUsse2 | VS2017 15.9 | No | x86_64 | Python 3.7 |
| 1.13.1py37CPUavx2 | VS2017 15.9 | No | AVX2 | Python 3.7 |
| 1.13.1py37GPUcuda101cudnn75sse2 | VS2017 15.9 | 10.1.105_418.96/7.5.0.56 | x86_64 | Python 3.7/Compute 3.0 |
| 1.13.1py37GPUcuda101cudnn75avx2 | VS2017 15.9 | 10.1.105_418.96/7.5.0.56 | AVX2 | Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5 |
| 1.12.0py36CPUsse2 | VS2017 15.8 | No | x86_64 | Python 3.6 |
| 1.12.0py36CPUavx2 | VS2017 15.8 | No | AVX2 | Python 3.6 |
| 1.12.0py36GPUcuda100cudnn73sse2 | VS2017 15.8 | 10.0.130_411.31/7.3.1.20 | x86_64 | Python 3.6/Compute 3.0 |
| 1.12.0py36GPUcuda100cudnn73avx2 | VS2017 15.8 | 10.0.130_411.31/7.3.1.20 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5 |
| 1.12.0py37CPUsse2 | VS2017 15.8 | No | x86_64 | Python 3.7 |
| 1.12.0py37CPUavx2 | VS2017 15.8 | No | AVX2 | Python 3.7 |
| 1.12.0py37GPUcuda100cudnn73sse2 | VS2017 15.8 | 10.0.130_411.31/7.3.1.20 | x86_64 | Python 3.7/Compute 3.0 |
| 1.12.0py37GPUcuda100cudnn73avx2 | VS2017 15.8 | 10.0.130_411.31/7.3.1.20 | AVX2 | Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5 |
| 1.11.0py36CPUsse2 | VS2017 15.8 | No | x86_64 | Python 3.6 |
| 1.11.0py36CPUavx2 | VS2017 15.8 | No | AVX2 | Python 3.6 |
| 1.11.0py36GPUcuda100cudnn73sse2 | VS2017 15.8 | 10.0.130_411.31/7.3.0.29 | x86_64 | Python 3.6/Compute 3.0 |
| 1.11.0py36GPUcuda100cudnn73avx2 | VS2017 15.8 | 10.0.130_411.31/7.3.0.29 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5 |
| 1.11.0py37CPUsse2 | VS2017 15.8 | No | x86_64 | Python 3.7 |
| 1.11.0py37CPUavx2 | VS2017 15.8 | No | AVX2 | Python 3.7 |
| 1.11.0py37GPUcuda100cudnn73sse2 | VS2017 15.8 | 10.0.130_411.31/7.3.0.29 | x86_64 | Python 3.7/Compute 3.0 |
| 1.11.0py37GPUcuda100cudnn73avx2 | VS2017 15.8 | 10.0.130_411.31/7.3.0.29 | AVX2 | Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5 |
| 1.10.0py36CPUsse2 | VS2017 15.8 | No | x86_64 | Python 3.6 |
| 1.10.0py36CPUavx2 | VS2017 15.8 | No | AVX2 | Python 3.6 |
| 1.10.0py36GPUcuda92cudnn72sse2 | VS2017 15.8 | 9.2.148.1/7.2.1.38 | x86_64 | Python 3.6/Compute 3.0 |
| 1.10.0py36GPUcuda92cudnn72avx2 | VS2017 15.8 | 9.2.148.1/7.2.1.38 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.10.0py27CPUsse2 | VS2017 15.8 | No | x86_64 | Python 2.7 |
| 1.10.0py27CPUavx2 | VS2017 15.8 | No | AVX2 | Python 2.7 |
| 1.10.0py27GPUcuda92cudnn72sse2 | VS2017 15.8 | 9.2.148.1/7.2.1.38 | x86_64 | Python 2.7/Compute 3.0 |
| 1.10.0py27GPUcuda92cudnn72avx2 | VS2017 15.8 | 9.2.148.1/7.2.1.38 | AVX2 | Python 2.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.9.0py36CPUsse2 | VS2017 15.7 | No | x86_64 | Python 3.6 |
| 1.9.0py36CPUavx2 | VS2017 15.7 | No | AVX2 | Python 3.6 |
| 1.9.0py36GPUcuda92cudnn71sse2 | VS2017 15.7 | 9.2.148/7.1.4 | x86_64 | Python 3.6/Compute 3.0 |
| 1.9.0py36GPUcuda92cudnn71avx2 | VS2017 15.7 | 9.2.148/7.1.4 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.9.0py27CPUsse2 | VS2017 15.7 | No | x86_64 | Python 2.7 |
| 1.9.0py27CPUavx2 | VS2017 15.7 | No | AVX2 | Python 2.7 |
| 1.9.0py27GPUcuda92cudnn71sse2 | VS2017 15.7 | 9.2.148/7.1.4 | x86_64 | Python 2.7/Compute 3.0 |
| 1.9.0py27GPUcuda92cudnn71avx2 | VS2017 15.7 | 9.2.148/7.1.4 | AVX2 | Python 2.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.8.0py36CPUsse2 | VS2017 15.4 | No | x86_64 | Python 3.6 |
| 1.8.0py36CPUavx2 | VS2017 15.4 | No | AVX2 | Python 3.6 |
| 1.8.0py36GPUcuda91cudnn71sse2 | VS2017 15.4 | 9.1.85.3/7.1.3 | x86_64 | Python 3.6/Compute 3.0 |
| 1.8.0py36GPUcuda91cudnn71avx2 | VS2017 15.4 | 9.1.85.3/7.1.3 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.8.0py27CPUsse2 | VS2017 15.4 | No | x86_64 | Python 2.7 |
| 1.8.0py27CPUavx2 | VS2017 15.4 | No | AVX2 | Python 2.7 |
| 1.8.0py27GPUcuda91cudnn71sse2 | VS2017 15.4 | 9.1.85.3/7.1.3 | x86_64 | Python 2.7/Compute 3.0 |
| 1.8.0py27GPUcuda91cudnn71avx2 | VS2017 15.4 | 9.1.85.3/7.1.3 | AVX2 | Python 2.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.7.0py36CPUsse2 | VS2017 15.4 | No | x86_64 | Python 3.6 |
| 1.7.0py36CPUavx2 | VS2017 15.4 | No | AVX2 | Python 3.6 |
| 1.7.0py36GPUcuda91cudnn71sse2 | VS2017 15.4 | 9.1.85.3/7.1.2 | x86_64 | Python 3.6/Compute 3.0 |
| 1.7.0py36GPUcuda91cudnn71avx2 | VS2017 15.4 | 9.1.85.3/7.1.2 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.7.0py27CPUsse2 | VS2017 15.4 | No | x86_64 | Python 2.7 |
| 1.7.0py27CPUavx2 | VS2017 15.4 | No | AVX2 | Python 2.7 |
| 1.7.0py27GPUcuda91cudnn71sse2 | VS2017 15.4 | 9.1.85.3/7.1.2 | x86_64 | Python 2.7/Compute 3.0 |
| 1.7.0py27GPUcuda91cudnn71avx2 | VS2017 15.4 | 9.1.85.3/7.1.2 | AVX2 | Python 2.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.6.0py36CPUsse2 | VS2017 15.4 | No | x86_64 | Python 3.6 |
| 1.6.0py36CPUavx2 | VS2017 15.4 | No | AVX2 | Python 3.6 |
| 1.6.0py36GPUcuda91cudnn71sse2 | VS2017 15.4 | 9.1.85.3/7.1.1 | x86_64 | Python 3.6/Compute 3.0 |
| 1.6.0py36GPUcuda91cudnn71avx2 | VS2017 15.4 | 9.1.85.3/7.1.1 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.6.0py27CPUsse2 | VS2017 15.4 | No | x86_64 | Python 2.7 |
| 1.6.0py27CPUavx2 | VS2017 15.4 | No | AVX2 | Python 2.7 |
| 1.6.0py27GPUcuda91cudnn71sse2 | VS2017 15.4 | 9.1.85.2/7.1.1 | x86_64 | Python 2.7/Compute 3.0 |
| 1.6.0py27GPUcuda91cudnn71avx2 | VS2017 15.4 | 9.1.85.2/7.1.1 | AVX2 | Python 2.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.5.0py36CPUavx | VS2017 15.4 | No | AVX | Python 3.6 |
| 1.5.0py36CPUavx2 | VS2017 15.4 | No | AVX2 | Python 3.6 |
| 1.5.0py36GPUcuda91cudnn7avx2 | VS2017 15.4 | 9.1.85/7.0.5 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.5.0py27CPUsse2 | VS2017 15.4 | No | x86_64 | Python 2.7 |
| 1.5.0py27CPUavx | VS2017 15.4 | No | AVX | Python 2.7 |
| 1.5.0py27CPUavx2 | VS2017 15.4 | No | AVX2 | Python 2.7 |
| 1.5.0py27GPUcuda91cudnn7sse2 | VS2017 15.4 | 9.1.85/7.0.5 | x86_64 | Python 2.7/Compute 3.0 |
| 1.5.0py27GPUcuda91cudnn7avx2 | VS2017 15.4 | 9.1.85/7.0.5 | AVX2 | Python 2.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.4.0py36CPUavx | VS2017 15.4 | No | AVX | Python 3.6 |
| 1.4.0py36CPUavx2 | VS2017 15.4 | No | AVX2 | Python 3.6 |
| 1.4.0py36GPUcuda91cudnn7avx2 | VS2017 15.4 | 9.1.85/7.0.5 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1,7.0 |
| 1.3.0py36CPUavx | VS2015 Update 3 | No | AVX | Python 3.6 |
| 1.3.0py36CPUavx2 | VS2015 Update 3 | No | AVX2 | Python 3.6 |
| 1.3.0py36GPUcuda8cudnn6avx2 | VS2015 Update 3 | 8.0.61.2/6.0.21 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1 |
| 1.2.1py36CPUavx | VS2015 Update 3 | No | AVX | Python 3.6 |
| 1.2.1py36CPUavx2 | VS2015 Update 3 | No | AVX2 | Python 3.6 |
| 1.2.1py36GPUcuda8cudnn6avx2 | VS2015 Update 3 | 8.0.61.2/6.0.21 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1 |
| 1.1.0py36CPUavx | VS2015 Update 3 | No | AVX | Python 3.6 |
| 1.1.0py36CPUavx2 | VS2015 Update 3 | No | AVX2 | Python 3.6 |
| 1.1.0py36GPUcuda8cudnn6avx2 | VS2015 Update 3 | 8.0.61.2/6.0.21 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1 |
| 1.0.0py36CPUsse2 | VS2015 Update 3 | No | x86_64 | Python 3.6 |
| 1.0.0py36CPUavx | VS2015 Update 3 | No | AVX | Python 3.6 |
| 1.0.0py36CPUavx2 | VS2015 Update 3 | No | AVX2 | Python 3.6 |
| 1.0.0py36GPUcuda8cudnn51sse2 | VS2015 Update 3 | 8.0.61.2/5.1.10 | x86_64 | Python 3.6/Compute 3.0 |
| 1.0.0py36GPUcuda8cudnn51avx2 | VS2015 Update 3 | 8.0.61.2/5.1.10 | AVX2 | Python 3.6/Compute 3.0,3.5,5.0,5.2,6.1 |
| 0.12.0py35CPUavx | VS2015 Update 3 | No | AVX | Python 3.5 |
| 0.12.0py35CPUavx2 | VS2015 Update 3 | No | AVX2 | Python 3.5 |
| 0.12.0py35GPUcuda8cudnn51avx2 | VS2015 Update 3 | 8.0.61.2/5.1.10 | AVX2 | Python 3.5/Compute 3.0,3.5,5.0,5.2,6.1 |
tensorflow CUDA cudnn 版本对应关系
https://blog.csdn.net/yuejisuo1948/article/details/81043962
linux下：
windows下：
上面两张图是在这里找到的：https://tensorflow.google.cn/install/source （右上角language选English）
tensorflow和keras版本搭配
https://docs.floydhub.com/guides/environments/
anaconda python 版本对应关系
https://blog.csdn.net/yuejisuo1948/article/details/81043823
本文链接：https://blog.csdn.net/yuejisuo1948/article/details/81043823
首先解释一下上表。 anaconda在每次发布新版本的时候都会给python3和python2都发布一个包，版本号是一样的。
表格中，python版本号下方的离它最近的anaconda包就是包含它的版本。
举个例子，假设你想安装python2.7.14，在表格中找到它，它下方的三个anaconda包（anaconda2-5.0.1、5.1.0、5.2.0）都包含python2.7.14；
假设你想安装python3.6.5，在表格中找到它，它下方的anaconda3-5.2.0就是你需要下载的包；
假设你想安装python3.7.0，在表格中找到它，它下方的anaconda3-5.3.0或5.3.1就是你需要下载的包；
镜像下载地址：清华镜像源
官方下载地址：https://repo.anaconda.com/archive/
https://blog.csdn.net/stephen_2018/article/details/80392545
win7 vs2015 cuda9.0 安装 Tensorflow-gpu 1.8
cuda_9.0.176_windows.exe
cudnn-9.0-windows7-x64-v7.zip
python-3.5.4-amd64.exe
https://blog.csdn.net/ei1990/article/details/84800151
WIN7系统安装 tensorflow1.6.0 + CUDA9.0 + cudnn7 版本
Anaconda3 5.2.0
CUDA9.0 + cudnn7 (9.1版本不支持tensorflow)
tensorflow-gpu 1.6.0
https://blog.csdn.net/Zqinstarking/article/details/80713338
防坑 centos7 安装 CUDA9.0 + cudnn7.1 +TensorFlow GPU版1.6.0/1.8.0
简单来说：tf1.5及以上用只能是cuda9.0，其他的tf1.4及以下版本就是cuda8.0等，最好自己去查查！可恶的是tf官方和nVidia都没有版本对应的说明！！！
https://blog.csdn.net/wukongabc_123/article/details/80379882
Windows 7下安装TensorFlow1.6（cuda9.0+cuDNN 7.0+python3.5+pip9）
https://blog.csdn.net/duoker/article/details/79483434
win7 x64 安装 TensorFlow1.6 CUDA 9.1+cuDNN7.1( 7.0.5)+python3.6 (python 3.5.2)
https://blog.csdn.net/wukongabc_123/article/details/80379882
win7+anaconda3+cuda9.0+CuDNN7+tensorflow-gpu+pycharm配置
https://blog.csdn.net/u011440696/article/details/79381375
tensorflow 安装GPU版本，个人总结，步骤比较详细
https://blog.csdn.net/gangeqian2/article/details/79358543
TensorFlow 安装GPU版本
https://blog.csdn.net/AAlonso/article/details/81504036
python+tensorflow+tensorflow-gpu+CUDA+cuDNN+pycharm全套环境配置教程 推荐
https://blog.csdn.net/kele52he/article/details/82986900
深度学习环境搭建-CUDA9.0、cudnn7.3、tensorflow_gpu1.10的安装
https://blog.csdn.net/xiaosa_kun/article/details/84868347
win7 vs2015 cuda9.0 安装 Tensorflow-gpu 1.8
https://blog.csdn.net/stephen_2018/article/details/80392545
WIN7系统安装 tensorflow1.6.0 + CUDA9.0 + cudnn7 版本
https://blog.csdn.net/ei1990/article/details/84800151
https://blog.csdn.net/weixin_42071277/article/details/88851868
Windows 7下安装TensorFlow1.6（cuda9.0+cuDNN 7.0+python3.5+pip9）
https://blog.csdn.net/duoker/article/details/79483434
匹配tensorflow-gpu和keras：
tensorflow 1.5 和keras 2.1.3、keras 2.1.4、keras 2.3.0（运行代码会报错）
tensorflow 1.4和keras 2.1.3
tensorflow 1.3和keras 2.1.2
tensorflow 1.2和keras 2.1.1