Understanding minimum example for torch.multiprocessing

I’m also having this issue, how can I pass the whole dataset to mp.Process(args=(whole_dataset,…))?

Thanks a lot!