Skip to content

[VLM] Fix CUDA IPC OOM #16118

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
BBuf merged 1 commit into from
Jan 7, 2026
Merged

[VLM] Fix CUDA IPC OOM #16118

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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 python/sglang/srt/managers/schedule_batch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1577,6 +1577,9 @@ def prepare_for_extend(self):
mm_item.feature = pixel_values.reconstruct_on_target_device(
torch.cuda.current_device()
)
# The reference by CudaIpcTensorTransportProxy was cut off,
# proactively delete to avoid slow gc.
del pixel_values
self.multimodal_inputs = multimodal_inputs
self.token_type_ids = token_type_ids_tensor
self.seq_lens_sum = sum(seq_lens)
Expand Down
4 changes: 4 additions & 0 deletions python/sglang/srt/multimodal/processors/base_processor.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -877,6 +877,8 @@ def process_and_combine_mm_data(
info_data=item.feature,
sync_buffer_meta=sync_flag,
)
elif not self.server_args.keep_mm_feature_on_device:
item.feature = item.feature.cpu()
elif (
isinstance(item.precomputed_embeddings, torch.Tensor)
and item.precomputed_embeddings.is_cuda
Expand All @@ -897,5 +899,7 @@ def process_and_combine_mm_data(
info_data=item.precomputed_embeddings,
sync_buffer_meta=sync_flag,
)
elif not self.server_args.keep_mm_feature_on_device:
item.precomputed_embeddings = item.precomputed_embeddings.cpu()

return all_collected_items, input_ids, ret
Loading