feat: support data parallel for deepseek (#1012)

### 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>

vllm_ascend/platform.py

@@ -138,7 +138,7 @@ class NPUPlatform(Platform):
 
             # Calculate expert parallel size based on world size
             parallel_config.expert_parallel_size = (
-                parallel_config.world_size //
+                parallel_config.world_size_across_dp //
                 parallel_config.expert_tensor_parallel_size)
 
         if model_config is None:
@@ -167,6 +167,8 @@ class NPUPlatform(Platform):
                         raise NotImplementedError(
                             "enable_graph_mode only works with deepseek model."
                         )
+                # Set compilation level to NO_COMPILATION to disable ACL Graph
+                compilation_config.level = CompilationLevel.NO_COMPILATION
 
         elif envs.VLLM_USE_V1 and model_config is not None and not enforce_eager:
             model_type = model_config.hf_config.model_type