locuslab/TCN

why?

RaganrokV opened this issue · 0 comments

I change the output "return torch.mean(self.network(x),dim=2)" for multi-feature time series and the training time reduces significantly..so why--lol