Are the models in drive already trained on Adversarial examples generated?

jind11 / TextFooler

A Model for Natural Language Attack on Text Classification and Inference

MIT License

482 stars 79 forks source link

Open SachJbp opened 4 years ago

SachJbp commented 4 years ago

I wanted to ask if the models in Google drive are already trained on Adversarial examples which you have generated?

Thanks

jind11 commented 4 years ago

no, they are not. They are merely trained on the original clean data.

SachJbp commented 4 years ago

Thanks for the clarification.