xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-05-04 17:51:22 +08:00
Code Issues Packages Projects Releases Wiki Activity

[XPU] work around for sp, avoid custom op import error (#28822)

Browse Source 
Signed-off-by: Kunshang Ji <kunshang.ji@intel.com>
This commit is contained in:
Kunshang Ji 2025-11-17 21:16:44 +08:00 committed by GitHub
parent d4acf518d0
commit 1b82fb0ad3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/compilation/pass_manager.py
View File

@ -18,6 +18,7 @@ if current_platform.is_cuda_alike():
from .fusion import RMSNormQuantFusionPass from .fusion import RMSNormQuantFusionPass
from .fusion_attn import AttnFusionPass from .fusion_attn import AttnFusionPass
from .qk_norm_rope_fusion import QKNormRoPEFusionPass from .qk_norm_rope_fusion import QKNormRoPEFusionPass
from .sequence_parallelism import SequenceParallelismPass
if current_platform.is_cuda(): if current_platform.is_cuda():
from .collective_fusion import AllReduceFusionPass, AsyncTPPass from .collective_fusion import AllReduceFusionPass, AsyncTPPass
@ -25,7 +26,6 @@ if current_platform.is_cuda():
from .fix_functionalization import FixFunctionalizationPass from .fix_functionalization import FixFunctionalizationPass
from .inductor_pass import CustomGraphPass, InductorPass, get_pass_context from .inductor_pass import CustomGraphPass, InductorPass, get_pass_context
from .noop_elimination import NoOpEliminationPass from .noop_elimination import NoOpEliminationPass
from .sequence_parallelism import SequenceParallelismPass
logger = init_logger(__name__) logger = init_logger(__name__)