From bea3d5bbb47954249d56fcf2e92f7a636a0ab134 Mon Sep 17 00:00:00 2001
From: xleoken <xleoken@163.com>
Date: Sat, 2 Aug 2025 16:52:12 +0800
Subject: [PATCH] [Bug] Fix run bug in run_dp_server.sh (#2139)

### What this PR does / why we need it?

For `Qwen2.5-0.5B-Instruct` model
- the model's total number of attention heads (14) must be divisible by
tensor parallel size. (4 -> 2)
- the model does not support enable-expert-parallel

### Does this PR introduce _any_ user-facing change?

No.

### How was this patch tested?

Local Test.

- vLLM version: v0.10.0
- vLLM main:
https://github.com/vllm-project/vllm/commit/ad57f23f6a528ab01066998b41796a44340fd43d

Signed-off-by: xleoken <xleoken@163.com>
---
 examples/run_dp_server.sh | 47 +++++++++++++++++++--------------------
 1 file changed, 23 insertions(+), 24 deletions(-)

diff --git a/examples/run_dp_server.sh b/examples/run_dp_server.sh
index eb3cfbf..1866fb0 100644
--- a/examples/run_dp_server.sh
+++ b/examples/run_dp_server.sh
@@ -1,33 +1,32 @@
-rm -rf ./.torchair_cache/
-rm -rf ./dynamo_*
-rm -rf /root/ascend/log/debug/plog/*
 
 export HCCL_IF_IP=2.0.0.0
-export GLOO_SOCKET_IFNAME="enp189s0f0"
-export TP_SOCKET_IFNAME="enp189s0f0"
-export HCCL_SOCKET_IFNAME="enp189s0f0"
+export GLOO_SOCKET_IFNAME="eth0"
+export TP_SOCKET_IFNAME="eth0"
+export HCCL_SOCKET_IFNAME="eth0"
 
 export OMP_PROC_BIND=false
 export OMP_NUM_THREADS=100
 
 export VLLM_USE_V1=1
+export VLLM_USE_MODELSCOPE=true
+
 export ASCEND_LAUNCH_BLOCKING=0
 
-vllm serve /data/weights/Qwen2.5-0.5B-Instruct \
-    --host 0.0.0.0 \
-    --port 20002 \
-    --served-model-name Qwen \
-    --data-parallel-size 4 \
-    --data-parallel-size-local 4 \
-    --data-parallel-address 2.0.0.0 \
-    --data-parallel-rpc-port 13389 \
-    --tensor-parallel-size 4 \
-    --enable-expert-parallel \
-    --no-enable-prefix-caching \
-    --max-num-seqs 16 \
-    --max-model-len 4096 \
-    --max-num-batched-tokens 4096 \
-    --gpu-memory-utilization 0.9 \
-    --trust-remote-code \
-    --enforce-eager \
-    --additional-config '{"ascend_scheduler_config":{"enabled":true},"torchair_graph_config":{"enabled":false, "enable_multistream_moe":false, "use_cached_graph":false}}'
+vllm serve Qwen/Qwen1.5-MoE-A2.7B  \
+  --host 0.0.0.0 \
+  --port 20002 \
+  --served-model-name Qwen \
+  --data-parallel-size 2 \
+  --data-parallel-size-local 2 \
+  --data-parallel-address 2.0.0.0 \
+  --data-parallel-rpc-port 13389 \
+  --tensor-parallel-size 4 \
+  --enable-expert-parallel \
+  --no-enable-prefix-caching \
+  --max-num-seqs 16 \
+  --max-model-len 4096 \
+  --max-num-batched-tokens 4096 \
+  --gpu-memory-utilization 0.9 \
+  --trust-remote-code \
+  --enforce-eager \
+  --additional-config '{"ascend_scheduler_config":{"enabled":true},"torchair_graph_config":{"enabled":false, "enable_multistream_moe":false, "use_cached_graph":false}}'