Is it okay that the validation loss function is lower than train loss function after certain point?

Question

I use this two functions for examining the performance of model. But I can\'t ensure that the model has well prediction because I didn\'t seen loss function graph like this in t