Article directory
foreword
This configuration method does not require installation code or installation process, so it may not be useful for novices who want to learn the installation tutorial, but it can be used out of the box, suitable for people who quickly skip the tutorial and do not want to try pip to report errors
1. Installation prerequisites
First of all, your anaconda must be installed. Anaconda installation tutorials are available online, and you can install it soon.
Second, find your anaconda installation directory and open envs
the folder
2. Installation Tutorial
1. Download package
envs
The virtual environment of anaconda is essentially a library in which various function packages are installed, so as long as a package with pytorch installed is placed in the folder under your anaconda, anaconda can recognize it, and then in pychearm use.
I have an installed package here, the download link is as follows:
Link: https://pan.baidu.com/s/1n2lInWuEURC_t2P0FCECbA Extraction code: 1234
The corresponding cuda and cudnn versions are 10.2, and the installation package link is as follows:
cuda link: https://pan.baidu.com/s/1GKw38kdv8P2pGJZJe5iFVQ extraction code: 1234
cudnn link: https://pan.baidu.com/s/ 1oJfIrT3fQiQg3HAtbMHIUA extraction code: 1234
Unzip the installation package and put it in envs
the folder under anaconda
2. Configure and use in pychram
(1) Open the settings of pycharm.
(2) Click the gear in the picture.
(3) Click add .
(4) Click Conda Environment , select Existing environment , and click the ellipsis. (5) Select python.exe under the virtual environment name ( pytorch-1-8-1 ) under the envs file
under the anaconda folder , click ok to import the environment and use it.