You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi @zhangyuc
I find in your code, the weight update is as below:
Local mode: number of cores on the local machine
while(k < n) { values(k) *= - actual_stepsize / math.sqrt(s(k) + 1e-16f) k += 1 }
what troube me is why the update is this, which seem not the standard format:
such as w = w - stepsize*gradient.
thank you!
The text was updated successfully, but these errors were encountered:
It implements the AdaGrad algorithm (http://www.jmlr.org/papers/volume12/duchi11a/duchi11a.pdf), which chooses different stepsizes for different coordinates. The stepsize of coordinate k is controlled by the accumulated gradient squared-norm s(k).
Hi @zhangyuc
I find in your code, the weight update is as below:
Local mode: number of cores on the local machine
while(k < n) { values(k) *= - actual_stepsize / math.sqrt(s(k) + 1e-16f) k += 1 }
what troube me is why the update is this, which seem not the standard format:
such as w = w - stepsize*gradient.
thank you!
The text was updated successfully, but these errors were encountered: