Skip to main content
4 events
when toggle format what by license comment
Apr 7, 2021 at 17:13 comment added SpiderRico This is a textbook case of overfitting. Your network fits particular patterns existing only on the training data, and fails to generalize. Given your training curves, you should stop training at about 20th epoch, in which the validation loss is the lowest.
Apr 7, 2021 at 12:11 answer added htl timeline score: 1
Apr 7, 2021 at 10:55 review First posts
Apr 7, 2021 at 13:12
Apr 7, 2021 at 10:55 history asked David Weinblumer CC BY-SA 4.0