Softmax model not converging

aamir-mustafa / pcl-adversarial-defense

Adversarial Defense by Restricting the Hidden Space of Deep Neural Networks, in ICCV 2019

59 stars 14 forks source link

Softmax model not converging #6

Closed krishnakanthnakka closed 4 years ago

krishnakanthnakka commented 4 years ago

Hello,

I ran the code python softmax_training.py with default parameters, the validattion loss is struck at 2.30 with accuracy 10%. Can you please check the code once.

[

](url)

krishnakanthnakka commented 4 years ago

specifications,

CUDA 10.1, PyTorch 1.12

krishnakanthnakka commented 4 years ago

I found the error. There's an issue with my Kubernetes machine. The num_workers has to be set to zero.