Model prediction output

[programmingprincess]

How can we map the regression values to a class label in the model output?

I see that this library uses GradientBoostingRegressor rather than GradientBoostingClassifier, but I am having trouble understanding what happens in metrics that allows for the evaluation of the continuous output with discrete labels. Thanks!