problems during training

[wzt9332]

Region Avg IOU: nan, Class: nan, Obj: -nan, No Obj: -nan, Avg Recall: 0.000000, count: 16
Region Avg IOU: nan, Class: nan, Obj: -nan, No Obj: -nan, Avg Recall: 0.000000, count: 8
Region Avg IOU: nan, Class: nan, Obj: -nan, No Obj: -nan, Avg Recall: 0.000000, count: 18
Region Avg IOU: nan, Class: nan, Obj: -nan, No Obj: -nan, Avg Recall: 0.000000, count: 12
Region Avg IOU: nan, Class: nan, Obj: -nan, No Obj: -nan, Avg Recall: 0.000000, count: 9

Gradient explosion occurred while training cowc data sets. What should I do?

[ashnair1]

Region Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.463884, Avg Recall: -nan, count: 0
Region Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.463872, Avg Recall: -nan, count: 0
Region Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.463883, Avg Recall: -nan, count: 0

I'm having a similar issue. Were you able to resolve it?

[wzt9332]

@Ash1995 Maybe it is about learning rate. I guss. I haven't resolve it. I haven't done this project in a long time. Let me know if you solve it.

[nivintw]

Just dropping in to say that I ran into this issue when trying to use SIMRDWN yolt with xView data. For me, this was actually caused by the fact that xView class labels have spaces in the class names, which led to problems when the darknet code was executed.

[ashnair1]

In my case, the problem was that I had begun labelling my classes from 1 thinking that 0 was reserved for the background class. Class labels should be zero indexed i.e. start from 0. I had previously opened an issue regarding this.