Visualizing layers of Resnet50 model trained on Plant Seedlings data

Aside question: are hooks in PyTorch the same thing as callback in Keras?

I know its too late, but I encountered this error a while ago.
Reason like someone said its due to some miss-match of weight counts , I don’t have much info.
Solution - If you are loading a saved model then load it on the type of model it is, what I mean is ---- if the loading model is fully trained ie ,is un-freezed layers than load it on a un-freezed model and if you are loading a freezed model than load it on a model which is having freezed layers.

1 Like