p2333 / score Goto Github PK

Dear author,

I was wondering if you could kindly guide me on how to train a ResNet18 model without using any additional data, as described in your paper's Section 5.2 and Table 2. I have attempted to locate the relevant code in your released materials, but unfortunately, I have not been able to find it. Additionally, I was unable to locate the code for PGD-AT with SE. I would be most grateful if you could kindly assist me with these queries.

Thank you for your time and consideration.

Hello, I would like to ask what is the function of the Softmax function in the toy_demo_1d.py file? Why can Softmax(X + delta) be used in the SCORE algorithm on line 97? Is this meant to represent a new data distribution?
Thanks in advance!

i wanna know how to do it

from utils_plus import (upper_limit, lower_limit, std, clamp, get_loaders,
attack_pgd, evaluate_pgd, evaluate_standard, normalize)

there is no file utils_plus in without ddpm