我正在尝试运行ssd pytorch代码（https://github.com/amdegroot/ssd.pytorch） 它在cpu上运行，但是当我尝试在gpu上运行时出现错误
Traceback (most recent call last):
File "train_c.py", line 267, in
train()
File "train_c.py", line 190, in train
loss_l, loss_c = criterion(out, targets)
File "/home/ubuntu/anaconda3/lib/python3.5/site-packages/torch/nn/modules/modu le.py", line 325, in call
result = self.forward(*input, **kwargs)
File "/home/ubuntu/ssd.pytorch/layers/modules/multibox_loss.py", line 88, in f orward
loc_p = loc_data[pos_idx].view(-1, 4)
File "/home/ubuntu/anaconda3/lib/python3.5/site-packages/torch/autograd/variab le.py", line 74, in getitem
return MaskedSelect.apply(self, key)
File "/home/ubuntu/anaconda3/lib/python3.5/site-packages/torch/autograd/_funct ions/tensor.py", line 465, in forward
return tensor.masked_select(mask)
TypeError: masked_select received an invalid combination of arguments - got (tor ch.ByteTensor), but expected (torch.cuda.ByteTensor mask)
运行train.py时出现此错误，该怎么办？