EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Doc] Upgrade env VLLM_ASCEND_ENABLE_FUSED_MC2 used in nightly test and tutorials (#8441)

Browse Source 
### What this PR does / why we need it?
The env `VLLM_ASCEND_ENABLE_FUSED_MC2` should only enabled in the
decoder node during Prefill-Decode Disaggregation scenario

---------

Signed-off-by: wangli <wangli858794774@gmail.com>
This commit is contained in:
Li Wang
2026-04-20 22:39:23 +08:00
committed by GitHub
parent 3db5048d74
commit 36a0470de1
9 changed files with 10 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/source/tutorials/models/Qwen3.5-397B-A17B.md
View File

@@ -106,7 +106,6 @@ sysctl -w vm.swappiness=0
sysctl -w kernel.numa_balancing=0
sysctl kernel.sched_migration_cost_ns=50000
export LD_PRELOAD=/usr/lib/aarch64-linux-gnu/libjemalloc.so.2:$LD_PRELOAD
export VLLM_ASCEND_ENABLE_FUSED_MC2=1
export VLLM_ASCEND_ENABLE_FLASHCOMM1=1
vllm serve Eco-Tech/Qwen3.5-397B-A17B-w8a8-mtp \
@@ -303,7 +302,6 @@ To run the vllm-ascend `Prefill-Decode Disaggregation` service, you need to depl
export VLLM_TORCH_PROFILER_WITH_STACK=0
export TASK_QUEUE_ENABLE=1
export VLLM_ASCEND_ENABLE_FUSED_MC2=1
export HCCL_OP_EXPANSION_MODE="AIV"
export ASCEND_RT_VISIBLE_DEVICES=0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15