Test on my own dataset error?
kokozeng opened this issue · 1 comments
kokozeng commented
I use eval = bdd and eval = coco in *.data file.
Try test and valid command.
test return nothing
valid return .json file, but json file is empty.
so how to test on my own dataset train weights?
jwchoi384 commented
@kokozeng
If test and valid return nothing, It seems that the model is not trained properly.
Please check your list files, images and ground-truth file.
You must make list files of your own dataset like BDD list files in my repo.
In addition, you must modify the valid file path in *.data file for validation.