train.py just stops working

[sgrvinod]

train.py just stops working

[gavrilo91]

I've moved execution to the machine with only one GPU and now everything works fine. But, still I'm not sure what is the issue.

[AnhPC03]

@gavrilo91 Could you show me what did you move in code?
Thanks

[bzzonly]

I've moved execution to the machine with only one GPU and now everything works fine. But, still I'm not sure what is the issue.

I also meet the same problem. I just git clone the repo, change the path of VOC07 and 12, then run the train.py, it doesn't work.

it has this warning :
/home/anaconda3/envs/pytorch04/lib/python3.6/site-packages/torch/nn/functional.py:52: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))

[bzzonly]

@gavrilo91 Could you show me what did you move in code?
Thanks

i meet the same question

[bzzonly]

i change to pytorch1.3.0, cuda10.0, and it works

[gavrilo91]

@bzzonly same here, tnx. Updated GPU drivers and cuda to 10.0 and now it works.