Iam totally new to PyTorch as Iam transitioning from keras.
Iam about to translate one of my projects, where one epoch is either as long as the number of samples devided by the batch size or as long as the number of iterations defined by a user.
Now Iam wondering how I could accomplish the same thing in PyTorch?
I first create a Datset class and wrap pytorchs Dataloader around it as recommended in this tutorial
That works fine for number of iterations = len(samples) /batch size
Now I want to increase the number of iterations, so I guess the generator has to be restarted and shuffle.
Any help on this would be awsome,