I observed that my GPU's memory is being consumed but the Utilisation stays 0. Because of this, my model is taking forever to load. I have tweaked this code to handle multilabel data. The only changes being-using y as a matrix instead of vector and different methods of calculation of cost and errors(Sorry that I am not able to provide the actual code). We can observe that the theano function train_model is taking a lot of time. I am a beginnner at Theano and also this method internally seems to having cascading variables and methods as well which is making me heard to debug.
What might be the possible reasons for such a scenario?
What debug methods should I use for this?