DyNet: The Dynamic Neural Network Toolkit

Preprint English OPEN
Neubig, Graham; Dyer, Chris; Goldberg, Yoav; Matthews, Austin; Ammar, Waleed; Anastasopoulos, Antonios; Ballesteros, Miguel; Chiang, David; Clothiaux, Daniel; Cohn, Trevor; Duh, Kevin; Faruqui, Manaal; Gan, Cynthia; Garrette, Dan; Ji, Yangfeng; Kong, Lingpeng; Kuncoro, Adhiguna; Kumar, Gaurav; Malaviya, Chaitanya; Michel, Paul; Oda, Yusuke; Richardson, Matthew; Saphra, Naomi; Swayamdipta, Swabha; Yin, Pengcheng;
(2017)
  • Subject: Computer Science - Computation and Language | Statistics - Machine Learning | Computer Science - Mathematical Software

We describe DyNet, a toolkit for implementing neural network models based on dynamic declaration of network structure. In the static declaration strategy that is used in toolkits like Theano, CNTK, and TensorFlow, the user first defines a computation graph (a symbolic r... View more
Share - Bookmark