David Seunghyun Yoon
David Seunghyun Yoon
I've got the same error. [environments] theano 0.8 - nvidia-docker kaixhin/cuda-theano:7.5 ( tested also on kaixhin/cuda-theano:8.0 ) - guided by : http://deeplearning.net/software/theano_versions/0.8.X/install.html keras 1.0 Could you let me know how...
Hi, Your observation is correct. The attention mechanism in MDREA and MHA-1 is the same. There is an implementation issue in the MDREA model. In particular, a sequence mask should...
I think the MDREA should perform better than MDRE. Let me share an "attention code" from another repository. The attention function in this code requires an additional argument "batch_seq" that...