A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[CI] Fix Pre-commit Issue Cannot determine type of "rank" and "world_size" (#26448)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
Wentao Ye committed
Oct 9, 2025 at 18:16 UTC
8983e0216fff76f53c09d6b4e356d852eff2291b
1 file changed
+5
vllm/distributed/device_communicators/all2all.py
+5
@@ -377,6 +377,11 @@ class FlashInferAllToAllManager(All2AllManagerBase):
377
All2All communication based on flashinfer kernels.
378
"""
379
380
+ # This type lint could be removed after all of the work in
381
+ # https://github.com/vllm-project/vllm/issues/26533 done.
382
+ rank: int
383
+ world_size: int
384
+
385
def __init__(self, cpu_group):
386
assert has_flashinfer_all2all(), (
387
"flashinfer all2all module not found. Please install/check flashinfer"