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

This article will explain in detail the solution to the failed installation of pytorch due to network problems. The editor thinks it is very practical, so I share it with you for reference. I hope you can get something after reading this article.

Preface

Every time the download method provided from the official website is used to install pytorch using conda, it will fail due to network speed problems.

Using conda Tsinghua Source installation is also a solution, but this is very inconvenient when specifying a cuda version.

This paper mainly gives a solution to this problem.

Method

Take the installation of pytorh-nightly, cuda80, linux systems as an example.

Select pip installation from the pytorch official website, and select the corresponding system, cuda version, python version. At this point, two corresponding installation commands appear.

The first is the command to install numpy and torchvison, which can be executed directly.

The second is to install the corresponding command of pytorch, the installation speed is very slow to execute directly. Here we can install pytorch by downloading the installation package in advance.

The methods are as follows:

Enter the second command displayed by pytorch on the command

Pip install torch_nightly-f https://download.pytorch.org/whl/nightly/cu80/torch_nightly.html

Among them

Https://download.pytorch.org/whl/nightly/cu80/torch_nightly-1.1.0.dev20190429-cp37-cp37m-linux_x86_64.whl

Is the address of the installation package that we need to download in advance.

Download the installation package through Thunderbolt.

Upload the installation package to the server

Install pytorch packages offline using pip

Source activate python virtual environment name pip install / path/of/the/pytorch_package (pytorch offline installation package path)

At this point, it is done!

Add: use anaconda to quickly configure pytorch in ubuntu, and solve the problem of network source connection failure

Using anaconda in ubuntu, you can configure pytorch easily and quickly.

The specific steps are as follows:

1. Install anaconda

Bash Anconda3-x.x.x-Linux-x86_64.sh

two。 Create a python virtual environment, here is python3.6, the environment name is self-designed.

Conda create-n python=3.6

3. Activate to enter the virtual environment

Source activate

4. Join Tsinghua Source

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

5. To install pytorch, you can match the version according to the information on the pytorch official website. Select cuda10.1 here.

Conda install pytorch torchvision cudatoolkit=10.1-c pytorch

6. To check whether pytorch is installed successfully, after entering the python environment, enter the following command

Import torch print (torch.__version__) # if version information is returned, installation is successful

PS: if there is a network connection failure error as shown below during step 5 installation, delete the command after-c python execute again

CondaHTTPError: HTTP 000000 CONNECTION FAILED for url Elapsed:-An HTTP error occurred when trying to retrieve this URL.HTTP errors are often intermittent, and a simple retry will get you on your way. This is the end of this article on "the solution to the failure of pytorch installation due to network problems". I hope the above content can be of some help to you, so that you can learn more knowledge. if you think the article is good, please share it out for more people to see.

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.