Skip to content

becxer/rejection-qa

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Rejection QA

To train or test the model you need the pre-trained ELMo model. It can be downloaded here. Then unzip it and store in ~/data/lm (or change config.py to alter its expected location). For example:

mkdir -p ~/data/lm
cd ~/data/lm
mv ~/Download/squad-context-concat-skip.tar.gz .
tar -xzf squad-context-concat-skip.tar.gz
rm squad-context-concat-skip.tar.gz

Training

Now the model can be trained using:

python docqa/reject/ablate_rejection_squad.py

This code based on allenai's document-qa

About

Experiments of rejection for qa system

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published