Differential Privacy gives error

jaintj95 / FederatedLearning_DBA

Federated Learning Project for OpenMined Research

1 stars 0 forks source link

Differential Privacy gives error #6

Closed anupamme closed 4 years ago

anupamme commented 4 years ago

In yaml files set diff_privacy to true and you will find that we get this exception:

Traceback (most recent call last): File "main.py", line 220, in epoch_interval=helper.params['aggr_epoch_interval']) File "/home/anupam/anupam-data/FederatedLearning_DBA/helper.py", line 268, in average_shrink_models update_perlayer.add(self.dp_noise(data, self.params['sigma'])) RuntimeError: expected device cuda:0 but got device cpu

It is not able to access gpu.

jaintj95 commented 4 years ago

Fixed in https://github.com/jaintj95/FederatedLearning_DBA/commit/2670082ab9c60c132a427de2903190725d6b16d5

Let me know if there are anymore issues.

anupamme commented 4 years ago

works now.