You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi @chessman
GPU-Jupyter aims to create isolated environments for reproducible setups.
The latest tag contradicts this purpose, as it changes the setup version as soon as a later tag is pushed.
Even worse, as there are three branches slim, python-only and full, users wouldn't be aware which version they have and what packages should work!
Therefore, this purpose is intended, it nudges to use tags highly recommended for Docker anyways, even though a more precise feedback would be nice for those not familiar with Docker.
I noticed that cschranz/gpu-jupyter doesn't have 'latest' tag anymore.
Is it intentional? Doesn't it make sense to point to the latest image (v1.5_cuda-11.6_ubuntu-20.04 as of now).
The text was updated successfully, but these errors were encountered: