dpsa4fl / overview

Differential Privacy for Federated Learning with Secure Aggregation
4 stars 0 forks source link

Change privacy parameter to (u128^2) (pass until prio), there convert to noise parameter as BigUint^2 #38

Closed MxmUrw closed 1 year ago

MxmUrw commented 1 year ago

In prio reduce fraction and round (up).