Open gliu9966 opened 5 years ago
I'm sorry for the late reply. Have you solved the problem you met? Could you give a clear description. As far as I know, dropout and early stopping can ease the over-fitting problem to some extent.
@gliu9966 I want to ask you a question, what is your version of tensorflow running the codes? I meet a question that "AttributeError: module 'tensorflow' has no attribute 'keras'"
How do you solve the overfitting problem?My experimental accuracy is always low