Use new train.report API (#49)

We are converging on using train.report throughout the Ray library code base instead of tune.report.

See ray-project/xgboost_ray#292

Signed-off-by: Kai Fricke <kai@anyscale.com>

Bump xgboost-ray version to 0.1.12

Signed-off-by: Antoni Baum <antoni.baum@protonmail.com>

Fix wrong xgboost_ray version

Signed-off-by: Antoni Baum <antoni.baum@protonmail.com>

Bump required xgboost-ray version to 0.1.12

Signed-off-by: Antoni Baum <antoni.baum@protonmail.com>

Update Ray core APIs (#35)

* Update Ray core APIs

* Fix CI

Signed-off-by: Antoni Baum <antoni.baum@protonmail.com>

* Bump required xgboost-ray version

Signed-off-by: Antoni Baum <antoni.baum@protonmail.com>

* Fix

Signed-off-by: Antoni Baum <antoni.baum@protonmail.com>

Signed-off-by: Antoni Baum <antoni.baum@protonmail.com>
Co-authored-by: Antoni Baum <antoni.baum@protonmail.com>

Remove legacy code (#31)

* Remove legacy code

* Fix CI

* Fix

Require xgboost_ray>=0.1.8

Increase CI timeouts

Retry CI tests on failure, bump version to 0.1.1, fix error caused by…

… xgboost==1.5.0 (#14)

* Retry tests on failure

* Retry examples on failure

* Fix sklearn

* Bump version

Bump version to 0.1.0