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

Central DP (client-side adaptive clipping) #2892

Merged
merged 148 commits into from
Feb 28, 2024

Conversation

mohammadnaseri
Copy link
Contributor

@mohammadnaseri mohammadnaseri commented Feb 1, 2024

Issue

Description

Central DP, also known as user-level DP, enables differential privacy strategies in Flower with client-side adaptive clipping.

Related issues/PRs

Proposal

Explanation

A new class, DifferentialPrivacyClientSideAdaptiveClipping, has been added. It can be wrapped around a strategy class to enable central DP with adaptive clipping. The clipping phase of clients' model updates is performed at the clients' sides. Each client can use the middleware adaptiveclipping_mod modifier for the clipping phase.

Checklist

  • Implement proposed change
  • Write tests
  • Update documentation
  • Update the changelog entry below
  • Make CI checks pass
  • Ping maintainers on Slack (channel #contributions)

Changelog entry

Any other comments?

src/py/flwr/common/differential_privacy.py Outdated Show resolved Hide resolved
src/py/flwr/server/strategy/__init__.py Outdated Show resolved Hide resolved
src/py/flwr/server/strategy/dp_adaptive_clipping.py Outdated Show resolved Hide resolved
src/py/flwr/client/mod/__init__.py Outdated Show resolved Hide resolved
@danieljanes danieljanes enabled auto-merge (squash) February 28, 2024 16:05
@danieljanes danieljanes merged commit bf35ca1 into main Feb 28, 2024
28 checks passed
@danieljanes danieljanes deleted the central-dp-client-side-adaptive-clipping branch February 28, 2024 16:14
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants