TPOT error for xgboost multiclass classificaion

[kiranellur]

While trying to use Tpot to get the best pipeline I am getting the following error

Invalid classes inferred from unique values of y. Expected: [0 1 2 3 4 5], got [1 2 3 4 5 6]

Context of the issue

I am using the tpot 0.12.1 version and xgboost 2.0.1 version, in place of cv I am providing StratifiedKFold( n_splits=cv_folds, shuffle=False, random_state=params["random_state"] )

My data consists of 6 classes. I am aware of that this error comes due to the different splitting each time (using stratified split), some splits do not have all the labels in both groups. How can i resolve this in tpot?

still I am getting the following error.

Invalid classes inferred from unique values of y. Expected: [0 1 2 3 4 5], got [1 2 3 4 5 6]

Note: I am using label encoding as xgboost 2.0 version does not have in built label encoding.

May I know is there a way to resolve this error?

[perib]

This error can happen when you have only one sample for a class. Since the stratifiedkfold doesn't repeat samples, there could be a fold without examples for a particular class. This can be solved by simply removing these classes. It can also happen if the labels are not encoded properly before TPOT.

If you can share a minimum example that reproduces the issue we can take a look