Awesome

ASCC

Pytorch implementation of ICLR 2021 spotlight paper, "Towards Robustness Against Natural Language Word Substitutions"(https://openreview.net/forum?id=ks5nebunVn_). Our code is partly based on https://github.com/robinjia/certified-word-sub and https://github.com/JHL-HUST/PWWS .

Environment

conda env create --name ascc --file environment.yml

Data

• Spacy python -m spacy download en

• GloVe cd glove wget http://nlp.stanford.edu/data/glove.840B.300d.zip unzip glove.840B.300d.zip rm glove.840B.300d.zip cd ..

• IMDB cd data_set wget http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz tar -xvzf aclImdb_v1.tar.gz rm -f aclImdb_v1.tar.gz cd ..

• SNLI cd data_set wget https://nlp.stanford.edu/projects/snli/snli_1.0.zip unzip snli_1.0.zip rm -r snli_1.0.zip snli_1.0 __MACOSX cd ..

Run

See train_imdb.sh