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.

Sign up for GitHub

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

Jump to bottom

[BUG] GradScaler does not work with torch 2.3.0 #610

Closed
samsja opened this issue May 13, 2024 · 1 comment
Closed

[BUG] GradScaler does not work with torch 2.3.0 #610

samsja opened this issue May 13, 2024 · 1 comment
Assignees
@justheuristic
Labels
bug Something isn't working

Comments

@samsja
Copy link
Contributor

samsja commented May 13, 2024

Describe the bug

GradScaler is not working anymore with pytorch 2.3.0

This

To Reproduce

from torch.cuda.amp.grad_scaler import OptState, _refresh_per_optimizer_state

this is failing now with pytorch 2.3. Which make it impossible to import hivemind GradScaler because of this line

Environment
Please list:

  • python version (e.g. 3.10.13);
  • hivemind.version; (master, 1.2.0.dev0)
@samsja samsja added the bug Something isn't working label May 13, 2024
@samsja samsja mentioned this issue May 13, 2024
Open
@rohrl
Copy link

rohrl commented May 16, 2024

+1

@samsja samsja closed this as completed Sep 4, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@justheuristic justheuristic

Labels
bug Something isn't working
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@justheuristic @rohrl @samsja