Multiprocess error

When I run the train.py in pysot which is an open sourece for trackers,it shows before any iteration which means it achieves no iteration:

DataLoader worker (pid 101735) is killed by signal: Killed. Details are lost due to multiprocessing. Rerunning with num_workers=0 may give better error trace.

But when I try to set num_workers=0 ,it just does’t train.
I really confused about this.Could you please tell why?

I see some blog,it says memory is full when loading,is this right?if it isn’t GPU’s duty,how to modify my code?

What do you mean by that? Is the loss not decreasing in this case or is the code just stuck?

The code just stuck.
And now it stuck in model_load.
It just stuck in different places witch makes me confused.
Thanks for your attention and I would appreciate it if you can give me some suggests.

Let’s continue the discussion on this topic.