Truncated BPTT with CNN

Hi i am new to pytorch.

I have a LSTMs with 2 layer + CNN and an output linear layer. It is for classification purpose.

I went through the char-rnn example online and try to implement a Truncated BPTT for the LSTMs without succes.

I want to know if it is possible to do so with such layers? If yes can you give me some directions or a nice tutorial.