pfedme Optimizer Probelem about pfedme HOT 6 CLOSED

Hi.

The line 64 in optimizer is to solve h_i (Equation 7 in our paper) to find the persionalized model.

Algorithm 1 line 8 is in userpFedMe.py line 53-54 to find local model.

for new_param, localweight in zip(self.persionalized_model_bar, self.local_model):
localweight.data = localweight.data - self.lamda* self.learning_rate * (localweight.data - new_param.data)

from pfedme.

Thank you!

from pfedme.

Hi.

The line 64 in optimizer is to solve h_i (Equation 7 in our paper) to find the persionalized model.

Algorithm 1 line 8 is in userpFedMe.py line 53-54 to find local model.

for new_param, localweight in zip(self.persionalized_model_bar, self.local_model):
localweight.data = localweight.data - self.lamda* self.learning_rate * (localweight.data - new_param.data)

Hello,
For the line 64 in optimizer is to solve h_i, do you modify the NAG to find the personalized model?
It doesn't look like a standard NAG update formula.

from pfedme.

No. I just applied gradient descent for h_i

from pfedme.

No. I just applied gradient descent for h_i

But this update looks strange. Can you explain it? Thamk you very much!
p.data = p.data - group['lr'] * (p.grad.data + group['lamda'] * (p.data - localweight.data) + group['mu']*p.data)

from pfedme.

(p.grad.data ) is gradient update for of f_bar
(group['lamda'] * (p.data - localweight.data) is gradient for moreau envelopes regularization term
group['mu']*p.data is just norm 2 regularizartion, but mu = 0 mean we don;t use it
so (p.grad.data + group['lamda'] * (p.data - localweight.data)) is gradient of h_i

from pfedme.