什么都不改，重新训练一下，
学习率减小
检查输入数据（x和y）
例如If using integers as targets, makes sure they aren’t symmetrical at 0.
I.e., don’t use classes -1, 0, 1. Use instead 0, 1, 2.

add a small epsilon value to the prediction to prevent this divergence

gradient clipping

https://github.com/wudashuo/yolov5/blob/master/models/tf.py