HAN mini-batching #12
Description
Hello, I am trying to reproduce the results for HAN mini-batching model, however mini-batch sampling dataloader gives an error for all datasets:
return th.stack(data)
RuntimeError: setStorage: sizes [1], strides [1], storage offset 2249, and itemsize 8 requiring a storage size of 18000 are out of bounds for storage of size 0
the sizes and strides are always the same while all the other numbers vary from run to run, can you please comment on this issue?