0

I had already trained the CNN with SGD, and it is training well. However, once I am training the model with Adam solver, after 100k iteration almost, it is starting to increase the loss value. Could you please help me to interpret this?

enter image description here

The following shows the solver.prototxt:

momentum: 0.99
momentum2: 0.999  #+
test_interval: 1000
test_iter: 40
weight_decay: 0.0005
base_lr:       0.0001
S.EB
  • 1,966
  • 4
  • 29
  • 54
  • you are overfitting – Shai Dec 05 '18 at 11:39
  • @Shai yes it is overfitting, however, I do not know should I train the model only for 60-70k iterations? Then, will I be able to use the model for inference? The reason that I posted is why it is starting to overfit after some iterations? – S.EB Dec 05 '18 at 23:44

0 Answers0