Skip to content

[#15178][fix] Account for reclaimable memory in V2 KV cache sizing on integrated GPUs - #20951

Open
janbernloehr wants to merge 6 commits into
NVIDIA:mainfrom
janbernloehr:fix/v2-unified-memory-kv-sizing-6306663
Open

janbernloehr wants to merge 6 commits into
NVIDIA:mainfrom
janbernloehr:fix/v2-unified-memory-kv-sizing-6306663

test: initialize KV manager version in profiling cleanup fixture

15bd4cd
Select commit
Loading
Failed to load commit list.