Repository Issue Activity (beta)

#824 Companion package proposal: textattack-detection plugin (content-layer defense recipe, out-of-tree)

#725 How to use textattack on generative models?

#423 Could you add a list of the supported languages of textattack?

#772 Issue of attacking seq2seq model

#673 Can you provide a simple attack API for single sentence attack test?

#625 Train model on custom dataset from file does not work

#737 Improve training speed by allow padding to batch-wise max length instead of model supported max length

#798 Attack on an HF model split into multiple GPUs

#771 Issues attacking se2seq model (seq2sick)

#787 the return of AttackedText.word_diff_ratio is 1 all the time

#723 Bug of utils/string.py/words_from_text()

#800 `BackTranslationAugmenter` produces an incorrect number of transformations

#728 Issue with using AttackArgs class outside of the command line (in my Python script)

#722 Compatibility issue w/ Transformers model

#754 a2t bug

#586 bert-attack doesn't work very well

#713 `flair`-related issues with Word Swap by Inflections

#631 Bug when attacking <SPLIT> in MNLI using CLARE

#727 [WRONG augmentation] pos_list.append(token.annotation_layers["pos"][0]._value) ->KeyError: 'pos'

#558 Delete Transformation does not work properly with Augmentation

#687 Do you have a plan to develop an adversarial defense integration?

#773 Perturbations for NER dataset

#789 Why does relaxing the constraint (RepeatModification) lead to less successful augmentation?

#799 Support Chinese attack？

#732 Does this library support targeted classification in its current state?