Can I generate the hinge loss of a multiclass svm test result with sklearn?

Question

The dummy version of the code is:

feature = []
label = []
xtrain,ytrain,xtest,ytest = train_test_split(feature,label, test_size = 0.3)
clf=svm.SVC(C=10,kernel='rbf')

I saw that there is a function hinge_loss() from the sklearn library, but it seems that it is only for linear svm? Can I calculate the hinge loss of the classifier in the sample code above with this hinge_loss() function? Or do I have to manually state out the hinge loss function for the multiclass svm with a rbf kernel like

loss_function=max(0,1 y_test*...)
# the line above was just a dummy version, as I havent figured out the loss function for the classifier just yet. 

I am assuming this is a programming related question so I ask in this forum (was reported by someone before saying my previous questions were not programming related).