Skip to content

coder3344/sru

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

About

SRU is a recurrent unit that can run over 10 times faster than cuDNN LSTM, without loss of accuracy tested on many tasks.


Average processing time of LSTM, conv2d and SRU, tested on GTX 1070


For example, the figures above and below present the processing time of a single mini-batch and the training time for sentence-level classification tasks. SRU achieves 10 to 16 times speed-up compared to LSTM, and operates as fast as (or faster than) word-level convolutional model (CNNs).


Training time (x-axis) vs valid accuracies (y-axis) on classification benchmarks


Requirements

CuPy and pynvrtc needed to compile the CUDA code into a callable function at runtime.


Tasks

  • classification
  • question answering on SQuAD
  • language modelling on PTB
  • machine translation
  • speech recognition

Contributors

About

Training RNNs as Fast as CNNs (https://arxiv.org/abs/1709.02755)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 79.1%
  • Shell 20.9%