Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Dlg yaml file change #437

Merged
merged 11 commits into from
Nov 17, 2022
Merged

Conversation

Osier-Yi
Copy link
Collaborator

  1. The original version of dlg cannot successfully reconstruct the training data because of the default dropout =0.5 in the convnet2. In the modified yaml file: scripts/attack_exp_scripts/privacy_attack/reconstruct_fedavg_opt_on_femnist.yaml, we set model.dropout=0.
  2. in the unitest of dlg, we also check whether the dlg reconstruction loss is less than the threshold.

@Osier-Yi Osier-Yi added the bug Something isn't working label Nov 17, 2022
Copy link
Collaborator

@xieyxclack xieyxclack left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@xieyxclack xieyxclack merged commit 42707fe into alibaba:master Nov 17, 2022
@Osier-Yi Osier-Yi deleted the dlg_yaml_file_change branch March 20, 2023 02:38
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

Successfully merging this pull request may close these issues.

dlg in attack module: attack fedavg, the loss is still high after searching the dummy data
2 participants