Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Only pin memory if using CUDA #114

Open
ayushkarnawat opened this issue Apr 26, 2020 · 0 comments
Open

Only pin memory if using CUDA #114

ayushkarnawat opened this issue Apr 26, 2020 · 0 comments
Assignees
Labels
enhancement Improvement to existing feature or code performance Performance related issues/improvements

Comments

@ayushkarnawat
Copy link
Owner

Currently, no matter if we use CPU or GPU, we pin memory. Rather, we should only pin memory if performing computations on CUDA cores.

use_cuda = args.cuda and torch.cuda.is_available()
device = torch.device("cuda" if use_cuda else "cpu")
dataloader_kwargs = {'pin_memory': True} if use_cuda else {}
@ayushkarnawat ayushkarnawat added enhancement Improvement to existing feature or code performance Performance related issues/improvements labels Apr 26, 2020
@ayushkarnawat ayushkarnawat self-assigned this Apr 26, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement Improvement to existing feature or code performance Performance related issues/improvements
Projects
None yet
Development

No branches or pull requests

1 participant