Create your own python virtual environment on a public Linux server

• 1. Understand the server and GPU used
• Second, install Anaconda and Pycharm
• Three, build your own environment
• 4. Use your own environment
• Five, use conda to install GPU version TensorFlow-GPU on linux system
• 6. Reference links

1. Understand the server and GPU used

cat /proc/version

lspci | grep -i nvidia

nvidia-smi

Check the existing state to see if there is any program to run again.

Second, install Anaconda and Pycharm

This server has already installed and configured these contents, and then directly start to build your own environment.

Third, build your own environment

1. First enter source ~/.bashrc to update environment variables, the front becomes (base)

source ~/.bashrc

2. Create your own virtual environment, the instructions are as follows

conda create -n gyyenv python=3.8.5

Then enter y, confirm, and get the following picture after completion

Fourth, use your own environment

1. Enter the environment created by yourself and enter the command

conda activate gyyenv

2. Return to the eight base environment and enter the command

conda deactivate

3. Check the environment in the directory and enter the command

conda env list

4. Delete the environment and enter the command, in which name is changedinto the name of the environment you want to delete.

conda env remove -n name

Five, use conda to install GPU version TensorFlow-GPU on linux system

1. Enter the virtual environment cond activate gyyenv created by yourself, you can check the tensorflow version installation first, enter the command conda search tensorflow-gpu , the diagram is as follows:

2. By command conda list See what's installed.

3. Through the command conda install tensorflow-gpu==2.2.0

4. Check if tensorflow is installed successfully.


The last True indicates that the installation was successful and exit your virtual environment!

6. Reference links

1. Build your own python environment on the server
2. The first step of scientific research: Create a python virtual environment on a Linux server
3. Use conda to install GPU version TensorFlow-GPU on linux system (detailed steps)