search

StefOe / indrnn-pytorch

pytorch implementation of Independently Recurrent Neural Networks https://arxiv.org/abs/1803.04831
121 stars 33 forks source link

How will the bidirectional flag handle a batch of examples of different sizes? #9

Open horiacristescu opened 4 years ago

horiacristescu commented 4 years ago

Hi, I want to use indRNN with batches of examples of different size, but I see this in the repo:

            if self.bidirectional:
                x_T = torch.cat([x, x.flip(-1)], 1)

This makes me think that perhaps the examples are not being flipped properly, because the torch.flip function is applied to the full width tensor.