Open wsqat opened 5 years ago
Why loss is close to 0.1, but the effect is very poor? Could you please give me some advices?
@liulei01 please
Do you use your own data set to train the model to test poorly?Have you solved it at last?
Why loss is close to 0.1, but the effect is very poor? Could you please give me some advices?