Pytorch environment preparation

Catalog

Tool catalog

Tool is introduced

Installation diagram

anaconda

Confirm that the is installed annaconda   

cuda install

Check that it is installed cuda 

 pytorch install

  Configure it anaconda Mirror image  

 pycharm install  

​ Code demonstration

If you have any questions, please leave a message

---

Tool catalog

• python 3.x + anaconda Can be the latest
• CUDA 11.2   NVIDIA graphics card
• pycharm community  Community Edition Just the latest  

Tool is introduced

• anaconda  python Package management tools     Automatic download management tools for third-party software libraries  
• cuda  Parallel computing library for NVIDIA graphics card  
• pycharm community    Free community version   python Code editing tools  

Installation diagram

anaconda

Official address   Anaconda | Anaconda Distribution

install anaconda It will be installed by itself python  notes python3.x Backwards compatible   , There is no problem installing the latest version

  Pay attention to check   Add to path route   For ease of use conda This command

Confirm that the is installed annaconda   

cmd  conda --version  If you can knock out conda This command indicates that it has been installed conda

cuda install

 cuda Can only be installed in NVIDIA On the graphics card   Choose the corresponding download according to your computer   After downloading, click Install   Close the security software during installation   This tutorial installs cuda 11.2 （ notes ： The screenshot version is 11.6.2）

 

 

 

Check that it is installed cuda 

cmd  nvcc -V    Be careful V It's capital    

 pytorch install

  Enter official website    Start Locally | PyTorch    Copy the red box Paste the cmd Be careful cmd Run as an administrator   Y

  Configure it anaconda Mirror image  

conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/main/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/free/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/conda-forge/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/msys2/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/bioconda/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/menpo/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/pytorch/
conda config --set show_channel_urls yes

 Use domestic sources   -c pytorch
conda install pytorch torchvision torchaudio cudatoolkit=11.3 

  Installed cuda The version should be consistent with that here

 pycharm install  

Official website  Download PyCharm: Python IDE for Professional Developers by JetBrains

Community Edition personal learning is enough   Double click installation     New projects   To configure pyhon conda Environmental Science  

Code demonstration

import torch

print(torch.__version__)
print('gpu:',torch.cuda.is_available())

If you have any questions, please leave a message