To my understanding, you are trying to simulate a model where CNN fully connected layers are connected to a SVM classifer at the end and want to display the training progress.
Training progress can be dispalyed in different manner in case of SVM. You can display diagnostic messages instead during training by using the 'Verbose' name-value pair argument. It controls the amount of diagnostic information per binary learner that the software displays in the Command Window.
You can refer to verbose MathWorks documentation page to find more on 'Verbose' name-value pair in 'fitcecoc' function.