Tim

这个的话应该是你没有将损失函数层jfda，放到caffe 对应的位置

你那个proto文件中有没有jfda想关的参数，一般新加一个层需要几个部分，proto,头文件，layer都要看一下对不对

ok thanks, i see the pretrained model of extd, but i mean the pretrained model of the backbone.

@anna0709 hi, anna, do you have solved this problem, i have meet the same problem.

@LXP-Never 兄弟你还是没开源呀，我再给你提个issue

@bilal-sher-nyu ok, thanks, i will test it, may be i have input the wrong type.

我昨晚训练一下，貌似有点问题。。训练在30000次前都没啥问题也能检测到，但是到40000次的时候检测不到了，尴尬，然后我看训练打印Obj的值也掉得很离谱了

不用去玩儿darknet了，去玩儿yolov4,yolov5这些呀，不比darknet好吗

@yudashuixiao1 Mybe the reason is the code : early_stopping = EarlyStopping(monitor='val_loss', min_delta=0, patience=10, verbose=0, mode='auto', baseline=None) when you val_loss did not improve ,large than patience=10 epoch, training will be stopped.