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
I couldn't find the python and cuda version of the each image of cpu and gpu for v139 right away, so could you put it in the release notes? By the way, is there a way to immediately check the versions of python and cuda? It is troublesome to build an image of about 50GB each time.
Motivation
The cuda versions of kaggle's gpu and cpu images often change, so even if you refer to an old notebook, it is difficult to align the environment with the current version. For example, it is troublesome to prepare an old version of mmdetection wheel.
Additional context
There is nothing special other than the above.
The text was updated successfully, but these errors were encountered:
I'm sorry. I found out that I just need to check docker-python/config.txt. However, when I check the environment with notebook, the version is different. Do you know the reason for this?
🚀 Feature
I couldn't find the python and cuda version of the each image of cpu and gpu for v139 right away, so could you put it in the release notes? By the way, is there a way to immediately check the versions of python and cuda? It is troublesome to build an image of about 50GB each time.
Motivation
The cuda versions of kaggle's gpu and cpu images often change, so even if you refer to an old notebook, it is difficult to align the environment with the current version. For example, it is troublesome to prepare an old version of mmdetection wheel.
Additional context
There is nothing special other than the above.
The text was updated successfully, but these errors were encountered: