caibi
caibi
plz close this issue
please close this issue if you won’t have any deeper discussion again.
> 要么换torch版本，要么把lengths改成lengths.cpu() torch为最新版本，lengths.to(opt.device)改了之后会触发另一个runtime error ``` RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument index...
> > do you have any solutions to solve the problem? > > model.py Line 51 ` packed = torch.nn.utils.rnn.pack_padded_sequence(embedded, input_lengths)` just edited ` packed = torch.nn.utils.rnn.pack_padded_sequence(embedded, input_lengths.cpu())` Woo, haven't...
After testing, the cause of crashing is from the mod Jade, with unknown reasons.
> I had a similar problem, but it wasn't solved after I removed Jade. I caught a Stacktrace by using ModernFix and some other Mods, maybe they are of some...