Skip to content
This repository has been archived by the owner on Feb 12, 2022. It is now read-only.
/ nonauto-nmt Public archive

PyTorch Implementation of "Non-Autoregressive Neural Machine Translation"

License

Notifications You must be signed in to change notification settings

salesforce/nonauto-nmt

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Non-Autoregressive Transformer

Code release for Non-Autoregressive Neural Machine Translation by Jiatao Gu, James Bradbury, Caiming Xiong, Victor O.K. Li, and Richard Socher.

Requires PyTorch 0.3, torchtext 0.2.1, and SpaCy.

The pipeline for training a NAT model for a given language pair includes:

  1. run_alignment_wmt_LANG.sh (runs fast_align for alignment supervision)
  2. run_LANG.sh (trains an autoregressive model)
  3. run_LANG_decode.sh (produces the distillation corpus for training the NAT)
  4. run_LANG_fast.sh (trains the NAT model)
  5. run_LANG_fine.sh (fine-tunes the NAT model)

About

PyTorch Implementation of "Non-Autoregressive Neural Machine Translation"

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages