完美解决Tensorflow不支持AVX2指令集问题|指令集加速

发表于 更新于

在pycharm中安装tensorflow后

![在这里插入图片描述](https://img-blog.csdnimg.cn/20200712190018795.png)

运行如下测试代码：

1
2
3
4
5
import tensorflow as tf
x = tf.Variable(3, name="x")
y = tf.Variable(4, name="y")
f = x*x*y + y + 2
print(f)

发现会报一行错误在这里插入图片描述
Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2大概意思是安装的tensorflow版本不支持cpu的AVX2编译
可能是因为安装时使用的pip install tensorflow ，这样默认会下载X86_64的SIMD版本。
查找解决办法后，有以下两种办法：

1.忽略屏蔽这个警告

在代码最前面添加如下两行代码

import os
os.environ["TF_CPP_MIN_LOG_LEVEL"]='2' # 只显示 warning 和 Error

2.彻底解决，换成支持cpu用AVX2编译的TensorFlow版本。

首先卸载原来安装的tensorflow版本
pip uninstall tensorflow
在这里下载对应版本的tensorflow：https://github.com/fo40225/tensorflow-windows-wheel，比如我需要的是CPU+AVX2+Python3.6，那么我就在下面的列表中选择这个：
在这里插入图片描述

Path Compiler CUDA/cuDNN SIMD Notes
2.2.0\py37\CPU+GPU\cuda102cudnn76sse2 VS2019 16.5 10.2.89_441.22/7.6.5.32 x86_64 Python 3.7/Compute 3.0
2.2.0\py37\CPU+GPU\cuda102cudnn76avx2 VS2019 16.5 10.2.89_441.22/7.6.5.32 AVX2 Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5
2.1.0\py37\CPU+GPU\cuda102cudnn76sse2 VS2019 16.4 10.2.89_441.22/7.6.5.32 x86_64 Python 3.7/Compute 3.0
2.1.0\py37\CPU+GPU\cuda102cudnn76avx2 VS2019 16.4 10.2.89_441.22/7.6.5.32 AVX2 Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5
2.0.0\py37\CPU\sse2 VS2019 16.3 No x86_64 Python 3.7
2.0.0\py37\CPU\avx2 VS2019 16.3 No AVX2 Python 3.7
2.0.0\py37\GPU\cuda101cudnn76sse2 VS2019 16.3 10.1.243_426.00/7.6.4.38 x86_64 Python 3.7/Compute 3.0
2.0.0\py37\GPU\cuda101cudnn76avx2 VS2019 16.3 10.1.243_426.00/7.6.4.38 AVX2 Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5
1.15.0\py37\CPU+GPU\cuda101cudnn76sse2 VS2019 16.3 10.1.243_426.00/7.6.4.38 x86_64 Python 3.7/Compute 3.0
1.15.0\py37\CPU+GPU\cuda101cudnn76avx2 VS2019 16.3 10.1.243_426.00/7.6.4.38 AVX2 Python 3.7/Compute 3.0,3.5,5.0,5.2,6.1,7.0,7.5
1.14.0\py37\CPU\sse2 VS2019 16.1 No x86_64 Python 3.7
1.14.0\py37\CPU\avx2 VS2019 16.1 No AVX2 Python 3.7
1.14.0\py37\GPU\cuda101cudnn76sse2 VS2019 16.1 10.1.168_425.25/7.6.0.64 x86_64 Python 3.7/Compute 3.0
1.14.0\py37\GPU\cuda101cudnn76avx2 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
1.13.1\py37\CPU\sse2 VS2017 15.9 No x86_64 Python 3.7
1.13.1\py37\CPU\avx2 VS2017 15.9 No AVX2 Python 3.7
1.13.1\py37\GPU\cuda101cudnn75sse2 VS2017 15.9 10.1.105_418.96/7.5.0.56 x86_64 Python 3.7/Compute 3.0
1.13.1\py37\GPU\cuda101cudnn75avx2 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
1.12.0\py36\CPU\sse2 VS2017 15.8 No x86_64 Python 3.6
1.12.0\py36\CPU\avx2 VS2017 15.8 No AVX2 Python 3.6
1.12.0\py36\GPU\cuda100cudnn73sse2 VS2017 15.8 10.0.130_411.31/7.3.1.20 x86_64 Python 3.6/Compute 3.0
1.12.0\py36\GPU\cuda100cudnn73avx2 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
1.12.0\py37\CPU\sse2 VS2017 15.8 No x86_64 Python 3.7
1.12.0\py37\CPU\avx2 VS2017 15.8 No AVX2 Python 3.7
1.12.0\py37\GPU\cuda100cudnn73sse2 VS2017 15.8 10.0.130_411.31/7.3.1.20 x86_64 Python 3.7/Compute 3.0
1.12.0\py37\GPU\cuda100cudnn73avx2 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
1.11.0\py36\CPU\sse2 VS2017 15.8 No x86_64 Python 3.6
1.11.0\py36\CPU\avx2 VS2017 15.8 No AVX2 Python 3.6
1.11.0\py36\GPU\cuda100cudnn73sse2 VS2017 15.8 10.0.130_411.31/7.3.0.29 x86_64 Python 3.6/Compute 3.0
1.11.0\py36\GPU\cuda100cudnn73avx2 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
1.11.0\py37\CPU\sse2 VS2017 15.8 No x86_64 Python 3.7
1.11.0\py37\CPU\avx2 VS2017 15.8 No AVX2 Python 3.7
1.11.0\py37\GPU\cuda100cudnn73sse2 VS2017 15.8 10.0.130_411.31/7.3.0.29 x86_64 Python 3.7/Compute 3.0
1.11.0\py37\GPU\cuda100cudnn73avx2 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
1.10.0\py36\CPU\sse2 VS2017 15.8 No x86_64 Python 3.6
1.10.0\py36\CPU\avx2 VS2017 15.8 No AVX2 Python 3.6
1.10.0\py36\GPU\cuda92cudnn72sse2 VS2017 15.8 9.2.148.1/7.2.1.38 x86_64 Python 3.6/Compute 3.0
1.10.0\py36\GPU\cuda92cudnn72avx2 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
1.10.0\py27\CPU\sse2 VS2017 15.8 No x86_64 Python 2.7
1.10.0\py27\CPU\avx2 VS2017 15.8 No AVX2 Python 2.7
1.10.0\py27\GPU\cuda92cudnn72sse2 VS2017 15.8 9.2.148.1/7.2.1.38 x86_64 Python 2.7/Compute 3.0
1.10.0\py27\GPU\cuda92cudnn72avx2 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
1.9.0\py36\CPU\sse2 VS2017 15.7 No x86_64 Python 3.6
1.9.0\py36\CPU\avx2 VS2017 15.7 No AVX2 Python 3.6
1.9.0\py36\GPU\cuda92cudnn71sse2 VS2017 15.7 9.2.148/7.1.4 x86_64 Python 3.6/Compute 3.0
1.9.0\py36\GPU\cuda92cudnn71avx2 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
1.9.0\py27\CPU\sse2 VS2017 15.7 No x86_64 Python 2.7
1.9.0\py27\CPU\avx2 VS2017 15.7 No AVX2 Python 2.7
1.9.0\py27\GPU\cuda92cudnn71sse2 VS2017 15.7 9.2.148/7.1.4 x86_64 Python 2.7/Compute 3.0
1.9.0\py27\GPU\cuda92cudnn71avx2 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
1.8.0\py36\CPU\sse2 VS2017 15.4 No x86_64 Python 3.6
1.8.0\py36\CPU\avx2 VS2017 15.4 No AVX2 Python 3.6
1.8.0\py36\GPU\cuda91cudnn71sse2 VS2017 15.4 9.1.85.3/7.1.3 x86_64 Python 3.6/Compute 3.0
1.8.0\py36\GPU\cuda91cudnn71avx2 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
1.8.0\py27\CPU\sse2 VS2017 15.4 No x86_64 Python 2.7
1.8.0\py27\CPU\avx2 VS2017 15.4 No AVX2 Python 2.7
1.8.0\py27\GPU\cuda91cudnn71sse2 VS2017 15.4 9.1.85.3/7.1.3 x86_64 Python 2.7/Compute 3.0
1.8.0\py27\GPU\cuda91cudnn71avx2 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
1.7.0\py36\CPU\sse2 VS2017 15.4 No x86_64 Python 3.6
1.7.0\py36\CPU\avx2 VS2017 15.4 No AVX2 Python 3.6
1.7.0\py36\GPU\cuda91cudnn71sse2 VS2017 15.4 9.1.85.3/7.1.2 x86_64 Python 3.6/Compute 3.0
1.7.0\py36\GPU\cuda91cudnn71avx2 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
1.7.0\py27\CPU\sse2 VS2017 15.4 No x86_64 Python 2.7
1.7.0\py27\CPU\avx2 VS2017 15.4 No AVX2 Python 2.7
1.7.0\py27\GPU\cuda91cudnn71sse2 VS2017 15.4 9.1.85.3/7.1.2 x86_64 Python 2.7/Compute 3.0
1.7.0\py27\GPU\cuda91cudnn71avx2 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
1.6.0\py36\CPU\sse2 VS2017 15.4 No x86_64 Python 3.6
1.6.0\py36\CPU\avx2 VS2017 15.4 No AVX2 Python 3.6
1.6.0\py36\GPU\cuda91cudnn71sse2 VS2017 15.4 9.1.85.3/7.1.1 x86_64 Python 3.6/Compute 3.0
1.6.0\py36\GPU\cuda91cudnn71avx2 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
1.6.0\py27\CPU\sse2 VS2017 15.4 No x86_64 Python 2.7
1.6.0\py27\CPU\avx2 VS2017 15.4 No AVX2 Python 2.7
1.6.0\py27\GPU\cuda91cudnn71sse2 VS2017 15.4 9.1.85.2/7.1.1 x86_64 Python 2.7/Compute 3.0
1.6.0\py27\GPU\cuda91cudnn71avx2 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
1.5.0\py36\CPU\avx VS2017 15.4 No AVX Python 3.6
1.5.0\py36\CPU\avx2 VS2017 15.4 No AVX2 Python 3.6
1.5.0\py36\GPU\cuda91cudnn7avx2 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
1.5.0\py27\CPU\sse2 VS2017 15.4 No x86_64 Python 2.7
1.5.0\py27\CPU\avx VS2017 15.4 No AVX Python 2.7
1.5.0\py27\CPU\avx2 VS2017 15.4 No AVX2 Python 2.7
1.5.0\py27\GPU\cuda91cudnn7sse2 VS2017 15.4 9.1.85/7.0.5 x86_64 Python 2.7/Compute 3.0
1.5.0\py27\GPU\cuda91cudnn7avx2 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
1.4.0\py36\CPU\avx VS2017 15.4 No AVX Python 3.6
1.4.0\py36\CPU\avx2 VS2017 15.4 No AVX2 Python 3.6
1.4.0\py36\GPU\cuda91cudnn7avx2 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
1.3.0\py36\CPU\avx VS2015 Update 3 No AVX Python 3.6
1.3.0\py36\CPU\avx2 VS2015 Update 3 No AVX2 Python 3.6
1.3.0\py36\GPU\cuda8cudnn6avx2 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
1.2.1\py36\CPU\avx VS2015 Update 3 No AVX Python 3.6
1.2.1\py36\CPU\avx2 VS2015 Update 3 No AVX2 Python 3.6
1.2.1\py36\GPU\cuda8cudnn6avx2 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
1.1.0\py36\CPU\avx VS2015 Update 3 No AVX Python 3.6
1.1.0\py36\CPU\avx2 VS2015 Update 3 No AVX2 Python 3.6
1.1.0\py36\GPU\cuda8cudnn6avx2 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
1.0.0\py36\CPU\sse2 VS2015 Update 3 No x86_64 Python 3.6
1.0.0\py36\CPU\avx VS2015 Update 3 No AVX Python 3.6
1.0.0\py36\CPU\avx2 VS2015 Update 3 No AVX2 Python 3.6
1.0.0\py36\GPU\cuda8cudnn51sse2 VS2015 Update 3 8.0.61.2/5.1.10 x86_64 Python 3.6/Compute 3.0
1.0.0\py36\GPU\cuda8cudnn51avx2 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
0.12.0\py35\CPU\avx VS2015 Update 3 No AVX Python 3.5
0.12.0\py35\CPU\avx2 VS2015 Update 3 No AVX2 Python 3.5
0.12.0\py35\GPU\cuda8cudnn51avx2 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

找到对应的.whl文件
在这里插入图片描述
下载该文件，我用google浏览器下载一直显示无法访问
在这里插入图片描述
后来选用Edge浏览器打开就好啦，直接就下载成功了。
此处放上tensorflow-1.12.0-cp36-cp36m-win_amd64.whl的下载链接：
链接：https://pan.baidu.com/s/1CvKUtmM1zHyJyJk87eFEUA
提取码：o85f
然后用activate 进入自己创建的虚拟环境
在这里插入图片描述
运行pip install tensorflow-1.12.0-cp36-cp36m-win_amd64.whl命令安装
在这里插入图片描述
最后用conda list命令看安装了那些包
在这里插入图片描述
然后再次运行代码，就不会报AVX2的错误啦
在这里插入图片描述