Different embedding and hidden sizes don't work for encoders and decoders

clay-lab / transductions

A PyTorch framework for creating, running, and reproducing experiments on seq2seq models.

6 stars 3 forks source link

Closed jopetty closed 3 years ago

jopetty commented 3 years ago

Also it breaks attention

jopetty commented 3 years ago

Okay, so I think it's the case that the hidden_size parameters must be the same between the encoder and decoder...is this always the case?

jopetty commented 3 years ago

Buuuuut now attention doesn't work

jopetty commented 3 years ago

Fixed in b4f4623