This code build with model VGG16 with changed Dense layers
Dense: 128, 64, 32 - relu
Out: 4 - sigm
U can run this code buy
python3 main.py [-option] [-option] [-option]
-l - load weights from ./weights.h5
-train - start train
-train-save - train and save weights as ./new-weights.h5
-train-save-show - same as up, but also showing
-p - predict from DataSet/test