Basically I moved everything to my Linux (centos 8) server, then this pops out:

(There is no such warnings when I run it on my windows laptop)
However, the param verbose_eval seems to be working right...
The verbose_eval belongs to the Python function call xgb.train(), not in the param dictionary. See the doc for xgb.train(): https://xgboost.readthedocs.io/en/latest/python/python_api.html#xgboost.train
The
verbose_evalbelongs to the Python function callxgb.train(), not in theparamdictionary. See the doc forxgb.train(): https://xgboost.readthedocs.io/en/latest/python/python_api.html#xgboost.train
Oh that's it! thx for blustering fast reply haha.
@hcho3 So far the parameters validation has been working well, do you think it's time for us to drop the experimental tag?
@trivialfis Yes, we can drop the experimental tag. Thanks for implementing it. It is one of the best major improvements we made in user experience recently.
Thanks, let me submit a PR next week.
Most helpful comment
The
verbose_evalbelongs to the Python function callxgb.train(), not in theparamdictionary. See the doc forxgb.train(): https://xgboost.readthedocs.io/en/latest/python/python_api.html#xgboost.train