fit() got an unexpected keyword argument 'validation_set' about small-text

[HannahKirk]

I am using the ConfidenceEnhancedSVM.

That makes sense. I guess this validation set is only useful for early stopping in the train scheduler of the transformers model. Might be worthwhile adding to documentation so its clear fit() of a non-transformers model doesnt take a validation set.
Thank you!

[chschroeder]

Hi,

which sklearn model are you using exactly? Until now, I thought that scikit-learn classifiers either don't have a validation set or create their own from the training data tha tis provided. Therefore, the current intended use would be to omit the validation set when using with scikit learn classifiers.

But maybe this assumption was wrong?