How can I use roc_auc_score() with the max_fpr parameter as the cross_val_score() scoring function?

Question

I tried using sklearn.metrics.make_scorer with the auroc_with_maxfpr function I wrote, and I did a sanity check by comparing the result of that sco