search

brightmart / text_classification

all kinds of text classification models and more with deep learning
MIT License
7.87k stars 2.57k forks source link

a08_EntityNetwork/a3_train.py代码的特殊考虑？ #117

Open lmxhappy opened 5 years ago

lmxhappy commented 5 years ago

best_eval_loss:",math.exp(best_eval_loss) if best_eval_loss<20 else 10000.000

a08_EntityNetwork/a3_train.py 里代码，130行。请问，为啥要看 best_eval_loss<20 ，如果不成立，就返回10000呢？谢谢

直接原样返回，有什么问题吗？谢谢