Skip to content

implement inference logic of tensorflow stack_bidirectional_dynamic_rnn with numpy

License

Notifications You must be signed in to change notification settings

xinjli/tflstm2np

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

implement tensorflow multilayer blstm with numpy

  • This repo implements numpy inference logic of tensorflow stack_bidirectional_dynamic_rnn with cudnn cell
  • The original purpose of this project is to export my cudnn acoustic model to numpy, so only cudnn related cell features would be supported. Some features in LSTMBlockCell are not supported (e.g: pee hole, clip cell)
  • check model.py for numpy implementation
  • check test.py for equivalent tensorflow cudnn implementation
  • Some details of this implementation is described here

About

implement inference logic of tensorflow stack_bidirectional_dynamic_rnn with numpy

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages