Common commands for Ubuntu Linux AI large-scale model development - the update includes NVIDIA status, and the installation of related common software packages.

Back it up if you have nothing to do, increase development efficiency

NVIDIA status, updated in real time:

watch -n  0.1  -d nvidia-smi #refresh every 0.1 seconds

Check the Ubuntu version number:

lsb_release -a

Check the graphics card model:

nvidia-smi -L

View video memory usage

When running Caffe, TensorFlow, pytorch and other programs that require CUDA, after forcibly killing the process , it is found that the video memory is still occupied. At this time, you can use the following command to view the process that cannot be seen in top or ps, and then kill it :

fuser -v /dev/nvidia*

watch -n 0.1 -d nvidia-smi  

Tsinghua source acceleration:

-i https://pypi.tuna.tsinghua.edu.cn/simple

View local ip:

ifconfig