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
Hello, really enjoyed your work on follow the ridge approach.
I tried implementing your system in pytorch, and followed the steps in appendix to better condition the H_yy inverse term. But my implementation is not producing results, even on simpler examples like the 2D-8 Mode Gaussian.
The tensorflow code is a bit harder to understand.
Also can you tell me what value of lambda do you suggest we should take?
1, 1e-1, 1e-4, 1e-8
One last question is how did you get the correction term, it includes some of the term necessary for the second order condition of minimax points, but how did you got them?
Thanks for reading my post. Really enjoyed this approach of following a ridge to approach minimax points.
The text was updated successfully, but these errors were encountered:
Hello, really enjoyed your work on follow the ridge approach.
I tried implementing your system in pytorch, and followed the steps in appendix to better condition the H_yy inverse term. But my implementation is not producing results, even on simpler examples like the 2D-8 Mode Gaussian.
The tensorflow code is a bit harder to understand.
Also can you tell me what value of lambda do you suggest we should take?
1, 1e-1, 1e-4, 1e-8
One last question is how did you get the correction term, it includes some of the term necessary for the second order condition of minimax points, but how did you got them?
Thanks for reading my post. Really enjoyed this approach of following a ridge to approach minimax points.
The text was updated successfully, but these errors were encountered: