Skip to content

Latest commit

 

History

History

DMBERT

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
performance
performance
 
 
README.md
README.md
 
 
get_submission.py
get_submission.py
 
 
model.py
model.py
 
 
run_ee.py
run_ee.py
 
 
run_infer.sh
run_infer.sh
 
 
run_train.sh
run_train.sh
 
 
utils_ee.py
utils_ee.py
 
 

README.md

DMBERT

This code is the implementation for DMBERT model. The implementations are based on Huggingface's Transformers, especially its example for the multiple-choice task.

Requirements

  • python==3.6.9

  • torch==1.2.0

  • transformers==2.8.0

  • sklearn==0.20.2

Usage

Hint: please read and delete all the comments after the backslash in each line of the .sh scripts before running them.

  1. Download LEVEN data files and put them in the ./data folder.
  2. Run run_train.sh for training and evaluation on the validation set.
  3. Run run_infer.sh to get predictions on the test set (dumped to results.jsonl).

See the two scripts for more details.