Closed monocongo closed 5 years ago
I assume that this indicates that I have negative and/or greater than 1.0 values in my annotations, no?
If so then it should be simple enough to write a Python script to clean the annotations.
Sorry, I opened this issue in the wrong repository, and besides, it has already been resolved in the darknet repository where I meant to open it: https://github.com/AlexeyAB/darknet/issues/3895.
TL; DR: The file bad_label.list
gives a list of the offending annotation files that cause this error.
I am not able to find the bad_label.list file. Can you please let me know where I can find it
I am training YOLOv3 with a custom dataset. In the training output I infrequently see this message:
My assumption is that this means that one or more of my training annotations contains bad values. However there is nothing I can see that might help me to determine which annotation file the error refers to, so finding and fixing this annotation will be like finding a needle in a haystack.
How might I find the offending annotation file?
Thanks in advance for any suggestions or insight.