How to implement "sgdm" training option with "train()"?

I need help implementing the "stochastic gradient descent with momentum" (sgdm) training option (including custom L2Regularization and MiniBatchSize), however I am using "newff" to create the network, hence "train()" to train the network. I know that the "trainingOptions" function is associated with "trainNetwork" so I cannot use that. Is there any way to implement this without recreating the NN with current functions?