0

I try to use the siamese network which combines with two resnet networks (Pretrained) to solve the few shot problem. The loss function is contrassive loss and the optimizer is Adam with 0.001 learning rate.

Both train set loss and validation loss went down since the experiment beginning and the model reached its best results at about 450 epochs.

However, When the experiment keep running, both numbers began to rise and to the same extent as they had at the beginning of the experiment.

I don't think it was a overfit problem, is there anyone who knows what happen? How can I fix this problem?

Thanks!

user17182657
  • 31
  • 3
  • Please add a https://stackoverflow.com/help/minimal-reproducible-example# that shows what you are doing and some information about the data sets you are using. That will help people trouble-shoot what is going on. – Tom Johnson Oct 25 '21 at 20:33
  • Looks like a question cross validated answers on the regular. Example: https://stats.stackexchange.com/questions/292700/overfitting-in-neural-network – Donald Seinen Oct 27 '21 at 06:18

0 Answers0