Can't train successfully with COCO2017 dataset

[metalgear54]

I'm using the lateset code to train COCO2017 dataset. But only get very low mAP. By using the yolov3.weights, I can get mAP=65 in validation dataset and mAP=67 in training dataset using test.py. which can prove that my dataset setup is right. But when i train the dataset with the same yolov3.weights module, the mAP will fall to nearly 0 after 1 or 2 epochs. Can you analyse why this happen? If I set the leraning rate very litlle such as 1e-6, the mAP still go down each training epoch.

[fddance]

i have this problem too.I check the label and found it was normal

[Flova]

Did you try to train for more epochs? Is sounds kind of weird, but maybe the optimizer choice or something else is different to the training setup of the pretrained weights, and this results in a convergence to a different (but maybe somewhat equivalent performance wise) point in the parameter space. This could result in a temporary reduction of performance for a few epochs while we move from one local minimum to the other. Try training for something like 100 epochs with a reasonable learning rate and show us the results.