I'm training on a skewed image dataset of 5000 images with class weights. The training loss decreases well and swift, but the validation-set loss fixates after around 6 epochs. I have tried using my custom model and also with transfer learning, with multiple hyperparameters. It follows the same structure everywhere. I am using categorical cross-entropy along with SGD with Nesterov momentum. This does not look like overfitting as the validation set stops performing just after 5 epochs(does not even fit).
Asked
Active
Viewed 25 times
0
-
This question is a duplicate. Search the site to find the duplicate. – Michael R. Chernick Sep 14 '19 at 23:25
-
I checked. The specifics the different – Sammy Jackson Sep 14 '19 at 23:32
-
The other question deals with bad generalization and overfitting. This is not overfitting. – Sammy Jackson Sep 15 '19 at 13:33