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

feat: support data parallel for deepseek (#1012)

Browse Source 
### What this PR does / why we need it?
feat: support data parallel for deepseek

### Does this PR introduce _any_ user-facing change?
Yes, support dp for deepseek

### How was this patch tested?

```
export VLLM_ENABLE_MC2=0
export VLLM_USE_V1=1
export TASK_QUEUE_ENABLE=1

source /usr/local/Ascend/ascend-toolkit/set_env.sh
source /usr/local/Ascend/nnal/atb/set_env.sh

nohup python -m vllm.entrypoints.openai.api_server
--model=/path/to/DeepSeek-R1-W8A8 \
    --quantization ascend \
    --served-model-name auto \
    --trust-remote-code \
    --distributed-executor-backend=mp \
    --port 8006 \
    -tp=8 \
    -dp=2 \
    --max-num-seqs 24 \
    --max-model-len 4096 \
    --max-num-batched-tokens 4096 \
    --block-size 128 \
    -O 0 \
    --no-enable-prefix-caching \
--additional-config
'{"torchair_graph_batch_sizes":[24],"expert_tensor_parallel_size":16,"ascend_scheduler_config":{},"enable_graph_mode":true}'
\
    --gpu-memory-utilization 0.95 &> run.log &
disown
```

Signed-off-by: boying <897013703@qq.com>
This commit is contained in:
NeverRaR
2025-06-04 18:31:41 +08:00
committed by GitHub
parent 517811449e
commit da9acfca60
8 changed files with 212 additions and 88 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm_ascend/worker/worker.py
View File

@@ -544,7 +544,7 @@ class NPUWorker(LocalOrDistributedWorkerBase):
init_ascend_model_parallel(
parallel_config.expert_parallel_size,
parallel_config.expert_tensor_parallel_size,
parallel_config.world_size,
parallel_config.world_size_across_dp,
)
ensure_kv_transfer_initialized(vllm_config)