self.emb-baseline

[MrZhengXin]

https://github.com/Neoanarika/torchexplainer/blob/5ff93d1c416e85c4a32c86e840d8352a053379cf/transformer/Models.py#L95
I'm wondering what is "self.emb-baseline", since it never appears before and cause some trouble to me:
Traceback (most recent call last):
File "train.py", line 293, in
main()
File "train.py", line 264, in main
train(transformer, training_data, validation_data, optimizer, device ,opt)
File "train.py", line 152, in train
model, training_data, optimizer, device, smoothing=opt.label_smoothing)
File "train.py", line 74, in train_epoch
pred = model(src_seq, src_pos, tgt_seq, tgt_pos)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 489, in call
result = self.forward(*input, **kwargs)
File "/home/zhengxin/sphinx/OpenNMT-py/torchexplainer/transformer/Models.py", line 217, in forward
enc_output, *_ = self.encoder(src_seq, src_pos,alpha = alpha,return_attns=self.return_attns)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 489, in call
result = self.forward(*input, **kwargs)
File "/home/zhengxin/sphinx/OpenNMT-py/torchexplainer/transformer/Models.py", line 95, in forward
print(self.emb-baseline)
RuntimeError: expected type torch.cuda.FloatTensor but got torch.FloatTensor

torch 0.4.1 and torch 1.0.1 all have the same problem.

[MrZhengXin]

Solved by adding .to('cuda')