search

yunxiaoshi / Neural-IMage-Assessment

A PyTorch Implementation of Neural IMage Assessment
Other
508 stars 92 forks source link

fully connected layers #20

Open victorzhucompass opened 3 years ago

victorzhucompass commented 3 years ago

This is great work. Thanks. Just wondering are there any particular reason the model does not include fully connected layers like VGG16 before softmax.

yunxiaoshi commented 3 years ago

I actually missed that long time ago, by adding something like

fc1 = list(model.classifier.children())[0]

to model.py should do the trick

victorzhucompass commented 3 years ago

great, any update after adding this layer into the model and retrain? wondering whether the learning rates are same setup.