the training loss and validation loss did not decrease and kept around 0.69, can anyone please tell me the reason why is this happening and how to solve it? Thanks!!!
I am doubting it is because the data is not enough, but Resnet34 can classify well. It is
