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

Fixed issue with torch 1.12 where _scale_fn_ref is missing in CyclicLR #1575

Merged
merged 2 commits into from
Oct 26, 2023
Merged

Fixed issue with torch 1.12 where _scale_fn_ref is missing in CyclicLR #1575

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
963eba7
Fixed issue with torch 1.12 where _scale_fn_ref is missing in CyclicLR
BloodAxe Oct 26, 2023
e80f6ca
Merge branch 'master' into feature/SG-000-fix-get-scheduler-state
BloodAxe Oct 26, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion src/super_gradients/training/utils/checkpoint_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1643,5 +1643,7 @@ def get_scheduler_state(scheduler) -> Dict[str, Tensor]:

state = scheduler.state_dict()
if isinstance(scheduler, CyclicLR) and not torch_version_is_greater_or_equal(2, 0):
del state["_scale_fn_ref"]
# A check is needed since torch 1.12 does not have the _scale_fn_ref attribute, while other versions do
if "_scale_fn_ref" in state:
del state["_scale_fn_ref"]
return state