Shulou(Shulou.com)06/01 Report--

Editor to share with you how to install the GPU version of the AI framework on windows, I believe most people do not know much about it, so share this article for your reference, I hope you can learn a lot after reading this article, let's go to know it!

Official version

Install CUDA

Install cuDNN

Configure environment variables

Install the python environment

Install the gpu version of the tensorflow package

It doesn't seem very complicated, but there are so many holes in it that you doubt life.

When downloading cuDNN, you need to register, and because the cuDNN file is on the external network, the download speed is very slow.

For example, if there is a match between different versions of tensorflow and CUDA (cuDNN), you might say that I check their dependencies first, and then find the appropriate version before installing.

For example, for windows systems, the version of the graphics card driver determines which version of CUDA you can install, which will be ignored in some tutorials.

.

If you are lucky enough to cross all these holes, you may also encounter a problem, that is, if you use pip to install tensorflow-gpu, you may still report all kinds of magical errors. In fact, the best way to install the AI framework on windows systems is to use conda instead of native pip. After summarizing the above common problems, let's start with how to install the GPU version of the AI framework on windows:

Minimalist version

Install Anaconda

Create a new conda virtual environment (python3.6 is recommended, which can be exchanged for domestic sources)

Install the video card driver (download from the official website or use the driver assistant software)

Install the cudatoolkit package when you install the AI framework

It's over.

As we can see above, we use a package called cudatoolkit, which is equivalent to automatically installing CUDA in conda's virtual environment. Most of the time, CUDA and cuDNN in cudatoolkit are installed together, which saves a lot of trouble, such as problems with the version of AI framework and CUDA, setting environment variables manually, and so on.

For specific installation scripts, please refer to the following command:

Install tensorflow

Conda create-n py36 python=3.6

Conda config-add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/conda config-set show_channel_urls yes

Conda install cudatoolkit tensorflow-gpu

Install pytorch

Conda create-n py36 python=3.6

Conda config-add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/conda config-set show_channel_urls yes

Conda install pytorch torchvision cudatoolkit=10.1-c pytorch

The above is all the contents of the article "how to install the AI Framework for GPU on windows". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.