XGBoost is a popular machine learning library active since 2015.
How large is each release?
~ 280 MiB total (Windows wheel 130 MiB, Linux wheel 150 MiB). Starting from 2.1.0, XGBoost no longer embeds a copy of NCCL and uses nvidia-nccl-cu12 instead. As a result, the size of Linux wheel was halved (300 MiB -> 150 MiB).
Project URL
https://pypi.org/project/xgboost
Does this project already exist?
New limit
20
Update issue title
Which indexes
PyPI
About the project
XGBoost is a popular machine learning library active since 2015.
How large is each release?
~ 280 MiB total (Windows wheel 130 MiB, Linux wheel 150 MiB). Starting from 2.1.0, XGBoost no longer embeds a copy of NCCL and uses
nvidia-nccl-cu12
instead. As a result, the size of Linux wheel was halved (300 MiB -> 150 MiB).How frequently do you make a release?
Roughly once in every 3 months
Code of Conduct