Closed loviji closed 3 years ago
Hi @loviji , sorry, we did not test the code on the CPU under Windows. Our test environment is: Linux/ PyTorch =1.4.0/ Python =3.6/1*nvidia-v100 (64G). And about this error, you may use if name == 'main:' in Windows (reference) or run this code in Linux.
Hi, @likyoo for your assist and detailed answer. I'm going to test this code in Linux with CPU, and will share results.
I got the same error and fixed it with adding the following line in the eval.py after the import section: if name == 'main':
Hello, I have checked this project on Windows 10 with Python 3.8, and get an error.
number of gpus setted to 0 on metadata.json file, because I've not CUDA torch and torchvision installed like this. _pip install torch==1.4.0+cpu torchvision==0.5.0+cpu -f https://download.pytorch.org/whl/torch_stable.html_
If this code runned and tested in non-Windows environment, please share tested operating system info, if it's possible.