About running command related issues

[Ailice0]

Thank you for your great work, MiB!

Depending on your introduction, I have configured all the environments and run the command:
python -m torch.distributed.launch --nproc_per_node=0 run.py --data_root data --batch_size 12 --dataset voc --name MIB --task 15-5 --step 1 --lr 0.001 --epochs 30 --method MIB

But the experiment could not run and there was no error. The following is the result of my command run:

The following is my dataset:

Could you please tell me is there any data path that needs to be set in advance or are there other variables that need to be set before experimenting?
Thank you for your feedback.

[fcdl94]

Hi @Ailice0!
Did you get any errors? How long it takes to conclude the process?
Which version of python/pytorch are you using?

[Ailice0]

Hi @fcdl94 ！
Thank you for your reply！
I experimented on Ubuntu18.04 with cuda=10.2,Python=3.6.13,Pytorch=1.2.0
After I input the run command, the system did not report any error with nothing happened, and the console soon went into the picture as shown below：

Here is the environment I installed：
ENVS.odt
Are there any important environments missing or data paths that need to be set up in advance?
Thanks!

[fcdl94]

Ok, the strange thing is that it's not outputting anything... If it crashes, I'd expect at least an error message 😅

Maybe try this:
logging.basicConfig(filename=filename, level=logging.INFO, format=f'%(levelname)s:{rank}: %(message)s', force=True)
in utils/logger, replacing the actual configuration (I had the problem of logger not working updating python, maybe it's the same issue here?)

If you open tensorboard something is logger there?

Are there any important environments missing or data paths that need to be set up in advance?

I'd say no, the important thing is that you have the data (Voc in this case) in the data folder and that you downloaded the pretrained model (on ImageNet) from inplace_abn repo.

[adaxidedakaonang]

If you are using vscode, try to use an older version of python.