华为910B4八卡服务器部署deepseek v4 flash记录

花了两天时间终于在成功在华为910B4八卡服务器部署deepseek v4 flash。

将镜像文件，权重文件rsync到另外四台服务器测试，也都成功拉起服务。

详细脚本如下

cd /data/docker

docker load -i vllm-ascend-deepseekv4.tar

docker run -it -d \

--name vllm-ascend-ds \

--net=host \

--shm-size=50g \

--privileged \

--device /dev/davinci0 \

--device /dev/davinci1 \

--device /dev/davinci2 \

--device /dev/davinci3 \

--device /dev/davinci4 \

--device /dev/davinci5 \

--device /dev/davinci6 \

--device /dev/davinci7 \

--device /dev/davinci_manager \

--device /dev/devmm_svm \

--device /dev/hisi_hdc \

-v /usr/local/dcmi:/usr/local/dcmi \

-v /usr/local/Ascend/driver/tools/hccn_tool:/usr/local/Ascend/driver/tools/hccn_tool \

-v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \

-v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \

-v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \

-v /etc/ascend_install.info:/etc/ascend_install.info \

-v /etc/hccn.conf:/etc/hccn.conf \

-v /data/weights/DeepSeek-V4-Flash-w8a8-mtp:/models \

quay.io/ascend/vllm-ascend:deepseekv4 \

bash

docker exec -it vllm-ascend-ds bash

#!/usr/bin/bash

export LD_PRELOAD=/usr/lib/aarch64-linux-gnu/libjemalloc.so.2:$LD_PRELOAD

export OMP_PROC_BIND=false

export OMP_NUM_THREADS=8

export PYTORCH_NPU_ALLOC_CONF=expandable_segments:True

export ACL_OP_INIT_MODE=1

export VLLM_ASCEND_ENABLE_FLASHCOMM1=1

export USE_MULTI_GROUPS_KV_CACHE=1

export TASK_QUEUE_ENABLE=1

export HCCL_OP_EXPANSION_MODE="AIV"

export HCCL_BUFFSIZE=512

export USE_MULTI_BLOCK_POOL=1

sysctl -w vm.swappiness=0

sysctl -w kernel.numa_balancing=0

sysctl kernel.sched_migration_cost_ns=50000

vllm serve /models \

--safetensors-load-strategy 'prefetch' \

--max-model-len 135168 \

--max-num-batched-tokens 4096 \

--served-model-name ds \

--gpu-memory-utilization 0.92 \

--max-num-seqs 16 \

--data-parallel-size 1 \

--tensor-parallel-size 8 \

--enable-expert-parallel \

--quantization ascend \

--port 7000 \

--block-size 128 \

--enable-chunked-prefill \

--enable-prefix-caching \

--tokenizer-mode deepseek_v4 \

--tool-call-parser deepseek_v4 \

--enable-auto-tool-choice \

--reasoning-parser deepseek_v4 \

--async-scheduling \

--additional-config '{"enable_cpu_binding":true,"multistream_overlap_shared_expert":false}' \

--compilation-config '{"cudagraph_mode":"FULL_DECODE_ONLY","cudagraph_capture_sizes":[2,4,6,8,10,12,14,16,18,20,22,24,32,36,40]}' \

--model-loader-extra-config '{"enable_multithread_load":true,"num_threads":16}' \

--speculative-config '{"num_speculative_tokens": 1,"method": "mtp"}'

curl http://localhost:7001/v1/chat/completions \

-H "Content-Type: application/json" \

-d '{

"model": "ds",

"messages": [

{"role": "user", "content": "请用数字1到10回答我，每个数字一行"}

],

"max_tokens": 100

}'