[BugFix] fix keyword argument max_num_tokens_across_dp missing

[zhangxinyuehfad]

What this PR does / why we need it?

fix keyword argument max_num_tokens_across_dp missing

Does this PR introduce any user-facing change?

How was this patch tested?

vllm serve Qwen/Qwen2.5-7B-Instruct --max_model_len 4096 --tensor_parallel_size 2 --data_parallel_size 2

log error:

INFO:     Started server process [12747]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO 06-24 09:08:01 [logger.py:43] Received request cmpl-98a67bfb86a34b8db6d83ad3489ea45b-0: prompt: '西安有什么好玩的地方', params: SamplingParams(n=1, presence_penalty=0.0, frequency_penalty=0.0, repetition_penalty=1.05, temperature=0.0, top_p=1.0, top_k=0, min_p=0.0, seed=None, stop=[], stop_token_ids=[], bad_words=[], include_stop_str_in_output=False, ignore_eos=False, max_tokens=20, min_tokens=0, logprobs=None, prompt_logprobs=None, skip_special_tokens=True, spaces_between_special_tokens=True, truncate_prompt_tokens=None, guided_decoding=None, extra_args=None), prompt_token_ids: [102178, 104139, 108257, 103958], prompt_embeds shape: None, lora_request: None, prompt_adapter_request: None.
INFO 06-24 09:08:01 [async_llm.py:271] Added request cmpl-98a67bfb86a34b8db6d83ad3489ea45b-0.
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527] WorkerProc hit an exception.
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527] Traceback (most recent call last):
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]   File "/__w/vllm-benchmarks/vllm-benchmarks/vllm-empty/vllm/v1/executor/multiproc_executor.py", line 522, in worker_busy_loop
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]     output = func(*args, **kwargs)
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]   File "/__w/vllm-benchmarks/vllm-benchmarks/vllm-ascend/vllm_ascend/worker/worker_v1.py", line 178, in execute_model
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]     output = self.model_runner.execute_model(scheduler_output)
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]   File "/usr/local/python3.10.17/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]     return func(*args, **kwargs)
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]   File "/__w/vllm-benchmarks/vllm-benchmarks/vllm-ascend/vllm_ascend/worker/model_runner_v1.py", line 1472, in execute_model
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]     aux_hidden_states) = (self._process_reqs(scheduler_output,
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]   File "/__w/vllm-benchmarks/vllm-benchmarks/vllm-ascend/vllm_ascend/worker/model_runner_v1.py", line 1085, in _process_reqs
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527]     attn_metadata = self.attn_metadata_builder.build(  # type: ignore
(VllmWorker rank=0 pid=13597) ERROR 06-24 09:08:01 [multiproc_executor.py:527] TypeError: AscendAttentionMetadataBuilder.build() got an unexpected keyword argument 'max_num_tokens_across_dp'
(EngineCore_0 pid=13068) ERROR 06-24 09:08:01 [dump_input.py:69] Dumping input data

fixed:

INFO:     Started server process [15266]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO 06-24 09:16:56 [logger.py:43] Received request cmpl-f629ffbc739c4b4ba599d873d3583056-0: prompt: '西安有什么好玩的地方', params: SamplingParams(n=1, presence_penalty=0.0, frequency_penalty=0.0, repetition_penalty=1.05, temperature=0.0, top_p=1.0, top_k=0, min_p=0.0, seed=None, stop=[], stop_token_ids=[], bad_words=[], include_stop_str_in_output=False, ignore_eos=False, max_tokens=20, min_tokens=0, logprobs=None, prompt_logprobs=None, skip_special_tokens=True, spaces_between_special_tokens=True, truncate_prompt_tokens=None, guided_decoding=None, extra_args=None), prompt_token_ids: [102178, 104139, 108257, 103958], prompt_embeds shape: None, lora_request: None, prompt_adapter_request: None.
INFO 06-24 09:16:56 [async_llm.py:271] Added request cmpl-f629ffbc739c4b4ba599d873d3583056-0.
INFO:     127.0.0.1:37616 - "POST /v1/completions HTTP/1.1" 200 OK

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 27.21%. Comparing base (c30ddb8) to head (eb4faac).
⚠️ Report is 541 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1398      +/-   ##
==========================================
- Coverage   27.39%   27.21%   -0.19%     
==========================================
  Files          56       56              
  Lines        6191     6214      +23     
==========================================
- Hits         1696     1691       -5     
- Misses       4495     4523      +28     

Flag	Coverage Δ
unittests	27.21% <ø> (-0.19%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[MengqingCao]

duplicated with #1273

[zhangxinyuehfad]

duplicated with #1273  重复了 #1273

I will close it