-
_From @tjrobinson on Friday, February 3, 2017 3:54:20 AM_
We're using the DeveloperExceptionPageMiddleware and it would be really useful if it could be changed/configured so that AJAX requests to A…
-
**Melon Improvement Proposal**
- MIP: 7
- Title: New Melonomics Proposal
- Authors: [Tom Shaughnessy](https://twitter.com/Shaughnessy119), [Ceteris Paribus](https://twitter.com/ceterispar1bus), […
-
## Feature request
**What is the expected behavior?**
I want to perform the adversarial attack on TabNet. Also, As a defense mechanism to make the robust this kind of attack, going for Adver…
-
Hi,
First of all, thank you for this project and all the other open source projects you're doing. I'm a big fan of your work.
I was training with the latest version on LibriSpeech dataset and lo…
-
I run an attack using FasterGeneticAlgorithmJia2019 recipe which has the maximum perturbation set to **20%** through max_percent=0.2
However some of the generated adversarial examples for the rotten…
-
**Describe the bug**
Bug in [`File "textattack/trainer.py", line 213, in _generate_adversarial_examples`](https://github.com/QData/TextAttack/blob/d129e7b8b004b295fc5a26241636c6d9cc313a51/textattack/…
-
At some point, we'll need a feature for loading a pre-trained/warmstarted policy for continuous training.
This feature is currently required in [`imitation.scripts.train_preference_comparisons`](h…
-
Follow sequentially:
Steps for both Baseline and cluster
Step 1: For both: Train on clean images (not separately) (use script and save weights)
Step 2: For adversarial examples: Create using the 'c…
-
Your code requires the image to be grey and have a size that is divisible by 16, but there is no sanitization after reading image. Therefore, the code is very likely to throw error.
We can tackle thi…
-
wer den weltuntergang NICHT sieht, der liest die falschen nachrichten (blaue pillen)
welche news lese ich? also, wem glaube ich?
**iceagefarmer**
iceagefarmer @ telegram
archiv: https://mi…