refreshment-coffee

SiamTrackers/SiamRPN/SiamRPN/siamrpn/loss.py中rpn_cross_entropy_balance函数是否有误？ Line208行左右。  如上图所示，这个是计算正样本的交叉熵损失。 如下图所示，应该是类比计算负样本的交叉熵损失函数。  我发现，这两个的代码并不是很对应，是否有误呢？ 关于负样本的交叉熵损失函数计算，我进行了如下修改： 

I would like to know the trend of each loss indicator during the training process. If you have the training logs file, please upload it. Thank you very much