search

xiaoyuxin1002 / SAIS

SAIS: Supervising and Augmenting Intermediate Steps for Document-Level Relation Extraction
MIT License
19 stars 5 forks source link

the code of CDR and GDA #1

Closed ZhangYi0621 closed 2 years ago

ZhangYi0621 commented 2 years ago

Hello! Thanks for your excellent work~ I'm interested in your work on the two biomedical datasets CDR and GDA, could you share the relevant code? Best wishes!

xiaoyuxin1002 commented 2 years ago

Hi, thanks for your interest in our work!

To run our codes on the two biomedical datasets, you just need to go through the following steps:

  1. Obtain the datasets from the corresponding websites and convert them into the same format as DocRED;
  2. Comment out the part in test() in main.py that is used for evidence-based data augmentation and ensemble inference;
  3. Adjust the hyperparameters and input arguments accordingly in prepare.sh, main.sh, and info.py.
marcuslyz98 commented 1 year ago

Hi, I am also interested in using the model trained on docRED and testing its performance on the unseen CDR and GDA dataset, could you share how to do that and specifically which lines to comment out in main.py?