Skip to content

Latest commit

 

History

History
20 lines (10 loc) · 598 Bytes

README.md

File metadata and controls

20 lines (10 loc) · 598 Bytes

noiseylg

Code repository for INLG 19 paper A Good Sample is Hard to Find: Noise Injection Sampling and Self-Training for Neural Language Generation Models. Please check back closer to the date of the conference (Oct. 29h) for release of the code.

INSTALL

First install plum and other libraries:

$ cd src; python setup.py install; plumr --add-libs fg ; cd ..

Then setup the eval scripts:

$ cd eval_scripts; ./install.sh; cd ..

E2E

To run E2E experiments/evaluation, please see README in e2e subdirectory.

Instructions for downloading models/outputs from paper are in the e2e README.