Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix dask doc. [skip ci] (#6108)

Browse Source
This commit is contained in:
Jiaming Yuan 2020-09-11 10:56:12 +08:00 committed by GitHub
parent 00b0ad1293
commit 08bdb2efc8
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 0 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
doc/tutorials/dask.rst
View File

@ -200,11 +200,4 @@ addressed yet:
- Label encoding for the ``DaskXGBClassifier`` classifier may not be supported. So users need - Label encoding for the ``DaskXGBClassifier`` classifier may not be supported. So users need
to encode their training labels into discrete values first. to encode their training labels into discrete values first.
- Ranking is not yet supported. - Ranking is not yet supported.
- Empty worker is not well supported by classifier. If the training hangs for classifier
with a warning about empty DMatrix, please consider balancing your data first. But
regressor works fine with empty DMatrix.
- Callback functions are not tested. - Callback functions are not tested.
- Only ``GridSearchCV`` from Scikit-Learn is supported. Meaning that we can distribute data
among workers but have to train one model at a time. If you want to scale up grid searching with
model parallelism with `Dask-ML <https://ml.dask.org/>`_, please consider using XGBoost's non-dask
Scikit-Learn interface, for example ``xgboost.XGBRegressor``, for now.