pytorch / opacus

Training PyTorch models with differential privacy
https://opacus.ai
Apache License 2.0
1.65k stars 328 forks source link

Cannot compute epsilon #591

Closed 01manoj20 closed 8 months ago

01manoj20 commented 1 year ago

I am trying to use make_private_with_epsilon function, I am getting cannot compute epsilon error I am not sure about the cause of the error.

HuanyuZhang commented 8 months ago

I think the problem might be associated with "prv" accounting. Could you try explicitly setting (accountant = "rdp") when initializing PrivacyEngine?

HuanyuZhang commented 8 months ago

I assume you have solved this issue. Feel free to re-open it if it is needed.