Closed jsignell closed 3 years ago
cc @TomAugspurger @mmccarty @gforsyth
We've moved our work over to xgboost.dask
, so for our purposes, anyway, this repo can be archived.
SGTM. Should we update the README to reflect this?
That sounds good. Similar to dask-lightgbm
, we might also consider issuing a final release before archiving which adds a warning that the functionality in dask-xgboost
has been moved into xgboost
and dask-xgboost
is no longer maintained
cc @hcho3 @trivialfis @JohnZed (for vis)
Can we wait for xgb 1.4? I would like to get the integrated version feature complete at this version.
Adding a warning, updating the docs, and archiving sounds good to me.
On Feb 8, 2021, at 5:58 PM, Jiaming Yuan notifications@github.com wrote:
Can we wait for xgb 1.4? I would like to get the integrated version feature complete at this version.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/dask/dask-xgboost/issues/80#issuecomment-775546124, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAKAOISKJADLOBITAKQVL5DS6B3DDANCNFSM4XJUEKWQ.
In the near future, I'd be totally onboard with archiving. But I agree with @trivialfis - there are still a few more minor bits coming for xgboost.dask in 1.4 (DART algo comes to mind?), so archiving this project after 1.4 is out would sound most appealing, as then we can clearly say there is no reason to use it for any new code.
Hi, XGBoost 1.4.1 is out and available on pypi. I believe we can safely redirect users to the xgboost.dask
module in the future as it's mostly feature-complete now. Thanks for the excellent work everyone! ;-)
Great! So what's needed is a note on the readme and then for someone to push the "archive button"?
Adding a promonant note to the README sounds good. I'll propose we also add a warning which points users to xgboost.dask
(similar to what we did with dask-lightgbm
) and issue a final release.
There's also some references to Dask-XGBoost in Dask-ML ( https://github.com/dask/dask-ml/issues/844 ). We may want to remove those as well.
Just pushed out 0.2.0 which contains deprecation warnings (xref https://github.com/dask/dask-xgboost/pull/84). Planning to archive this repo tomorrow morning (in the central time zone)
Thanks James! 😄
Since the functionality has been moved to xgboost itself. Would now be the appropriate time to archive this repo?