hello
i ubuntu while running python prepare_data.py --root data/student_model_imgs/ --datalist_path data/datalist/ --datalist_type train --gpu_id 1
i got Traceback (most recent call last):
File "prepare_data.py", line 4, in <module>
import MCS2018
ImportError: No module named MCS2018
please help
Regards
milton
Hi, please make sure that black-box .so file is in the same folder with script
Also, if you're using CPU .so, replace import MCS2018 with
import MCS2018_CPU as MCS2018
[MCS2018] Initializing.
[MCS2018] Mode: Gpu 1
Traceback (most recent call last):
File "evaluate.py", line 160, in <module>
main(args)
File "evaluate.py", line 60, in main
net = MCS2018.Predictor(args.gpu_id)
RuntimeError: [enforce fail at common_cudnn.h:108] version_match || backwards_compatible_7 || patch_compatible. cuDNN compiled (7103) and runtime (7005) versions mismatch
please help
Posted by: milton.2002 @ May 29, 2018, 4:31 a.m.Please try the following:
1. import MCS2018
net = MCS2018.Predictor(gpu_id)
# and then import transforms
import torchvision
the issue is that black-box conflicts with torch, but initializing net Before torch import works
Posted by: oleggrinch @ May 29, 2018, 8:20 a.m.