Describe the workflow you want to enable

Today in SGDClassifier, the parameter early_stopping uses a fraction of the data randomly, it would be useful to support a custom validation set chosen by the user.

Describe your proposed solution

for example:

clf = SGDClassifier(early_stopping=True)
clf.fit(X_train, y_train, eval_set=(X_val, y_val))

EDIT

Broader Scope

Same applies to GradientBoosting* and HistGradientBoosting*