Federated-Learning-Backdoor

Federated-Learning-Backdoor copied to clipboard

to sure the trigger logic

3

i do not understand why the train_data need to randomcrop after padding 4 use "transforms.RandomCrop(32, padding=4)" At the beginning, i thought that maybe for trigger setting,but i found that you...

imomoe233

How can l get the missing package in the project

2

there are show as below in the FL_Backdoor.py, thank you for help 

liuyuqinggg

A question about the results for Cifar-10

5

We reproduce the experiments for Cifar-10 according to run_backdoor_cv_task.sh. However, we cannot get the same results as in the paper. The backdoor accuracy is shown in the following figure, where...

Yu-shuyan

The question about the testing dataset

7

https://github.com/jhcknzzm/Federated-Learning-Backdoor/blob/a7ef36afc5c5dfe7dbb233e8d7f35c141cefeffb/FL_Backdoor_CV/image_helper.py#L271 I have a question about the testing dataset about backdoor attack accuracy on the FEMNIST. From this line of code, the author seems to use the training data set...

Alan-Qin

Thank you for providing us with a good code. The link in Corpus parsed dataset seems to have expired at this time. Can you upload it again? Or can you...

cjrcjrdl1

Hi, I'm trying to reproduce the experiments for Cifar-10. When I run the command `python main_training.py --run_slurm 0 --GPU_id 0 --start_epoch 1 --attack_num 250 --gradmask_ratio 1.0 --edge_case 0` on Windows...

xiaozeplus

wrongs with "lastwords " and the acc is very low

1

Executing the command with "--lastwords 1" will report an error, and "lastwords" cannot be recognized. Also, when I was training the "reddit" data set, the correct rate was still 17%...

LongJinhe-coder

The dataset about reddit(FL_backdoor_NLP)

2

Thanks for your code, but I couldn't run the FL_Backdoor_NLP experiment because of the Reddit dataset(the download URL has expired).

zalmanchen