uoa-compsci399-s2-2023 / adversarial-insight-ml

Adversarial Insight ML (AIML) - Python Package for Evaluating Machine Learning Image Classification Models' Robustness Against Adversarial Attacks
https://uoa-compsci399-s2-2023.github.io/adversarial-insight-ml/
MIT License
1 stars 1 forks source link

`tensorboard` should be a required package. `robustbench` should not. #25

Closed changx03 closed 11 months ago

changx03 commented 11 months ago

tensorboard should be a required package. As it is required to train surrogate model. Meanwhile, robustbench should not, as it is only used when a user decide to run certain evaluation.

eaj3 commented 11 months ago

This issue has been resolved. The problem was fixed in commit [f0765f06f1f347244293b121feb9a5cd3e8cd161].