Skip to content

Fix torch autocast deprecation warning in gradient checkpointing #1010

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

Draft
Copilot wants to merge 6 commits into
base: master
Choose a base branch
from
+103 −77
Draft

Fix torch autocast deprecation warning in gradient checkpointing #1010

Changes from 2 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
c4a11ff
Initial plan
Copilot Apr 24, 2026
bf46032
Fix torch autocast deprecation warning by using use_reentrant=False
Copilot Apr 24, 2026
a42702f
Suppress torch.utils.checkpoint FutureWarning about deprecated torch.…
Copilot Apr 24, 2026
04cc832
Upgrade torch to 2.5.1+cu121 to fix autocast deprecation warning at t…
Copilot Apr 24, 2026
4d5e274
Upgrade torch to 2.6.0+cu124 to patch torch.load RCE vulnerability (CVE)
Copilot Apr 24, 2026
c4bf59f
Create proper poetry.lock
Apr 27, 2026
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
2 changes: 1 addition & 1 deletion silnlp/nmt/hugging_face_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -335,7 +335,7 @@ def __init__(self, exp_dir: Path, config: dict) -> None:
"max_source_length": 200,
"max_target_length": 200,
"gradient_checkpointing": True,
"gradient_checkpointing_kwargs": {"use_reentrant": True},
"gradient_checkpointing_kwargs": {"use_reentrant": False},
"save_steps": 1000,
"per_device_train_batch_size": 16,
"save_strategy": "steps",
Expand Down