[KV Offloading] Enable HMA models for Tiering Offloading (#44287)
Signed-off-by: varun sundar rabindranath <vsundarr@redhat.com> Co-authored-by: varun sundar rabindranath <vsundarr@redhat.com>
This commit is contained in:
co-authored by
varun sundar rabindranath
parent
9af53a3c13
commit
7268457999
@@ -110,7 +110,6 @@ class TieringOffloadingSpec(CPUOffloadingSpec):
|
||||
self._scheduler_mmap = scheduler_mmap
|
||||
|
||||
# Create primary tier (CPU-based)
|
||||
assert len(self.gpu_block_size) == 1
|
||||
primary_tier = CPUPrimaryTierOffloadingManager(
|
||||
num_blocks=self.num_blocks,
|
||||
cache_policy=self.eviction_policy, # type: ignore[arg-type]
|
||||
|
||||
Reference in New Issue
Block a user