iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[CI bug] Fix Each KV cache group's real block_size must be divisible by has h_block_size (#51180)

Browse Source 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
Wentao Ye
2026-08-05 14:03:17 -04:00
committed by GitHub
parent 9833aa53d5
commit 23c0f337b7
1 changed files with 1 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
vllm/v1/core/kv_cache_utils.py
+1 -4
View File
@@ -645,12 +645,9 @@ def resolve_kv_cache_block_sizes(
if not (cache_config.enable_prefix_caching or connector_enabled):
return scheduler_block_size, scheduler_block_size
# Mamba groups with block_size != cache_config.block_size
# (mamba_cache_mode != "align") break divisibility; back off to the
# scheduler block size.
if any(
isinstance(g.kv_cache_spec, MambaSpec)
and g.kv_cache_spec.block_size != cache_config.block_size
and g.kv_cache_spec.mamba_cache_mode != "align"
for g in groups
):
return scheduler_block_size, scheduler_block_size