When using pycharm for algorithm development, it is often necessary to run the code on the GPU of the server. For different projects, the environment configuration and running dependencies are often different. At this time, docker is needed to create a private container for each project to run. Code. The servers are generally shared. When running the code, you need to know who is running, so you can't use the root account to create your own users on the server. In this way, we need to use pycharm to remotely connect to the docker personal user directory. This document mainly records how to configure the personal user directory in pycharm+docker.
The main reference is to know the article:
PyCharm+Docker: Create the most comfortable deep learning alchemy furnace
The difference is that:
1. Create your own username in the docker container
2. The user id in the docker must be mapped to the id of the GPU outside the docker container.
3. The difference between deployment settings and pycharm interpreter settings
deployment settings:
pycharm interpreter settings:
pycharm remote interpreter settings
The rest of the settings are the same, and you're done~