When building and applying a machine learning model such as support vector regression $($SVR$)$ in scikit$-$learn, we typically run the following steps:

from sklearn.svm import SVR

model $=$ SVR$(...)$

model.fit$(...)$

pred $=$ model.predict$(...)$

Which step involves training the SVR model?

a$.$

from sklearn.svm import SVR

b$.$

model $=$ SVR$(...)$

c$.$

model.fit$(...)$

d$.$

pred $=$ model.predict$(...)$