[Model Runner V2] Reserve CUDA graph memory #53306
WoosukKwon
merged
12 commits into
vllm-project:main
from
njhill:fix/v2-runner-cudagraph-memory-profiling
[Bugfix][V1] Reserve CUDA graph memory in V2 GPU model runner
586e26f5
Merge branch 'main' into fix/v2-runner-cudagraph-memory-profiling
01e2491f
Merge branch 'main' into fix/v2-runner-cudagraph-memory-profiling
bd03517a
Merge branch 'main' into fix/v2-runner-cudagraph-memory-profiling
2ab50abc
Merge branch 'main' into fix/v2-runner-cudagraph-memory-profiling
cf304380
Merge remote-tracking branch 'origin/main' into fix/v2-runner-cudagra…
213341f1
updates
36a1131b
claude
commented
on 2026-08-21
Merge remote-tracking branch 'origin/main' into fix/v2-runner-cudagra…
c72b87c3
WoosukKwon
approved these changes
on 2026-08-22
fixes/updates; handle mamba attention
eb2c5b9c
Merge remote-tracking branch 'refs/remotes/origin/main' into nhill/fi…
d3fa7091
handle omitted mamba attention when profiling
53a9d685
Merge remote-tracking branch 'origin/main' into nhill/fix/v2-runner-c…
b2a8c566
disabled auto-merge 81 days ago
Manually disabled by user
Manually disabled by user
Reviewers
Assignees
No one assigned
Labels
ready
nvidia
mrv2
kimi
Milestone
No milestone
Login to write a write a comment.
Login via GitHub