I have been trying to train a model with 0.001 learning rate. I tried regression techniques, early stopping and lr manipulations within epochs. But nothing felt right even though after numerous tries i got an "okay" result with 0.75 loss for categorical_crossentropy with around 80% accuracy. The validation loss graphs were always in a crazy spike and jittery they would not normalize on a plateu as they should. I know that this issue is about the...
Originally asked by Yigithan Sever on Stack Exchange (ai) (CC BY-SA 4.0).
