Closed mariesosa closed 3 years ago
Seems like a scikit-learn issue? https://github.com/scikit-learn/scikit-learn/issues/12516
Looks like we have to fix our implementation...
@mariesosa OK I just opened a fix! Let me know if that works for you (#191 )
I tested it with various group split functions and it seems to work fine. Thanks.
Describe the bug
TuneSearch
raise aValueError: The 'groups' parameter should not be None.
with sklearn cv split based on groups (like GroupKFold) when argumentsmax_iters
andearly_stopping
are specified. The behavior is the same for hyperopt and random search algorithms.Steps/Code to Reproduce
Expected Results
No error is thrown.
Actual Results
Versions
hyperopt==0.2.5 numpy==1.18.4 ray==1.2.0 scikit-learn==0.24.1 tune_sklearn==0.2.1