Train loss and test loss

Train loss constantly declining, test loss constantly declining, indicating that the network is still learning; (best)

Train loss is constantly declining, test loss tends to be constant, indicating that the network is over-fitted; (max pool or regularization)

Train loss tends to be constant, test loss is declining, indicating a problem with DataSet 100%; (check dataset)

Train loss tends to be unchanged, test loss tends to be constant, indicating that learning bottlenecks, need to reduce the learning rate or volume number; (Reduce learning rate) train loss is rising, test loss is rising, the network structure is poorly designed, training parameters are not properly set, The data set has been cleaned and so on. (Worst case scenario)