Feng Wang
[resnet50.pth](https://github.com/Megvii-BaseDetection/storage/releases/download/0.1/resnet50.pth) @alpc111
Plz try ```pods_train --resume ``` or check ```pods_train --help``` for more infomation.
Did you run your pods_train command under network config dir?
Could you plz tell me your commit id, I will try to reimplemented your error.
That's wired, code [here](https://github.com/Megvii-BaseDetection/cvpods/blob/master/tools/train_net.py#L148-L149) add sys path to import config.
try to use ```python3 cvpods.tools.train_net --num-gpus 1```?
> I meet the same problem and can't solve by using **python3 cvpods.tools.train_net --num-gpus 1**. ``python3 -m cvpods.tools.train_net --num-gpus 1``, `-m` missing in your your command.
Could solve this issue when your enter config dir and rerun again @Tianhao-Qi
It's a yolov3 on coco dataset or on your own dataset?
Thanks for your report, we will try in our inner version.