Skip to content

Keep class labels on GPU for compiled prediction #1329

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

Merged
aw471 merged 2 commits into from
Nov 3, 2026
Merged

Keep class labels on GPU for compiled prediction #1329

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3fb723a
Keep cached class labels on their original device
aw471 Nov 2, 2026
565fccb
Preserve class labels outside cache offloading
aw471 Nov 2, 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
3 changes: 3 additions & 0 deletions sdm/models/base/icl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -335,6 +335,8 @@ def fit(
)

if x.is_cuda and len(contexts) > 1:
# Keep class labels on GPU for compiled prediction.
classes = batch_cache.pop("classes")
try: # Copy to pinned CPU memory:
batch_cache = batch_cache._apply_tensor(
lambda tensor: torch.ops.aten._to_copy.default(
Expand All @@ -345,6 +347,7 @@ def fit(
)
)
finally:
batch_cache["classes"] = classes
torch.cuda.current_stream(x.device).synchronize()

cache[i] = batch_cache
Expand Down
Loading