[Bug]: ModuleNotFoundError: No module named 'pyarrow" in main branch

[Oneal65]

Your current environment

image info

The latest pull request in the repository is "[V1] Prompt logprobs + APC compatibility; prompt logprobs reqs cannot fill APC (#13949)".

client start shell

sudo python3 ./bench_serving.py --backend vllm --dataset-name random --model deepseek-r1 --tokenizer ./tokenizer --dataset-path ./ShareGPT_V3_unfiltered_cleaned_split.json --random-input-len 6000 --random-output-len 1000 --random-range-ratio 1 --request-rate 16 --max-concurrency 16 --num-prompts 80 --base-url $BASE_URL --host 0.0.0.0 --port 8000 --profile

server start shell

VLLM_USE_V1=1 VLLM_TORCH_PROFILER_DIR=/disc vllm serve /root/.cache/huggingface --tensor-parallel-size 16 --trust-remote-code --gpu-memory-utilization 0.9 --max-model-len 32768 --enforce-eager --enable-reasoning --reasoning-parser deepseek_r1 --served-model-name deepseek-r1

error info

🐛 Describe the bug

in the first block

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

[thousandhu]

it seems that your env doesn't install pyarrow. Maybe you can install pyarrow in all ray worker and try again.

[afeldman-nm]

it seems that your env doesn't install pyarrow. Maybe you can install pyarrow in all ray worker and try again.

@thousandhu JFYI it appears (based on the image) that they ran into an issue when they tried to install pyarrow:

root@TENCENT64:/vllm-workspace# pip install pyarrow
WARNING: Retrying (Retry(total=4, connect-None,
read=None,
redirect-None, status-None)) after connection broken by 'NewConnectionError('<pip-_vendor.urllib3.connection.HTTPSConnection object at 0x7f1eb41b85c0›: Failed to establish a new connection:
[Errno 101] Network is unreachable')': /simple/pyarrow/
WARNING: Retrying (Retry(total=3, connect-None,
read=None,
redirect-None, status-None)) after connection broken by 'NewConnectionError('<pip-_vendor.urllib3.connection.HTTPSConnection object at 0x7f1eb40502c0>: Failed to establish a new connection:
[Errno 101] Network is unreachable')': /simple/pyarrow/

[afeldman-nm]

I am trying to reproduce your issue but when I run

VLLM_USE_V1=1 VLLM_TORCH_PROFILER_DIR=./disc vllm serve ~/.cache/huggingface --port 8091 --tensor-parallel-size 16 --trust-remote-code --gpu-memory-utilization 0.9 --max-model-len 32768 --enforce-eager --enable-reasoning --reasoning-parser deepseek_r1 --served-model-name deepseek-r1

I got a different issue, probably related to the configuration of my own environment

WARNING 03-08 13:36:22 [arg_utils.py:1452] Setting max_num_batched_tokens to 8192 for OPENAI_API_SERVER usage context.
Traceback (most recent call last):
  File "/home/afeldman/git_clones/vllm/venv/bin/vllm", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/cli/main.py", line 73, in main
    args.dispatch_function(args)
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/cli/serve.py", line 33, in cmd
    uvloop.run(run_server(args))
  File "/home/afeldman/git_clones/vllm/venv/lib/python3.12/site-packages/uvloop/__init__.py", line 109, in run
    return __asyncio.run(
           ^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/asyncio/runners.py", line 195, in run
    return runner.run(main)
           ^^^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/asyncio/runners.py", line 118, in run
    return self._loop.run_until_complete(task)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
  File "/home/afeldman/git_clones/vllm/venv/lib/python3.12/site-packages/uvloop/__init__.py", line 61, in wrapper
    return await main
           ^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/openai/api_server.py", line 947, in run_server
    async with build_async_engine_client(args) as engine_client:
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/openai/api_server.py", line 139, in build_async_engine_client
    async with build_async_engine_client_from_engine_args(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/openai/api_server.py", line 163, in build_async_engine_client_from_engine_args
    engine_client = AsyncLLMEngine.from_engine_args(
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/v1/engine/async_llm.py", line 107, in from_engine_args
    vllm_config = engine_args.create_engine_config(usage_context)
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/engine/arg_utils.py", line 1194, in create_engine_config
    model_config = self.create_model_config()
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/engine/arg_utils.py", line 1121, in create_model_config
    return ModelConfig(
           ^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/config.py", line 327, in __init__
    hf_config = get_config(self.hf_config_path or self.model,
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/transformers_utils/config.py", line 326, in get_config
    raise ValueError(f"Unsupported config format: {config_format}")
ValueError: Unsupported config format: ConfigFormat.AUTO

[Oneal65]

I am trying to reproduce your issue but when I run

VLLM_USE_V1=1 VLLM_TORCH_PROFILER_DIR=./disc vllm serve ~/.cache/huggingface --port 8091 --tensor-parallel-size 16 --trust-remote-code --gpu-memory-utilization 0.9 --max-model-len 32768 --enforce-eager --enable-reasoning --reasoning-parser deepseek_r1 --served-model-name deepseek-r1

I got a different issue, probably related to the configuration of my own environment

WARNING 03-08 13:36:22 [arg_utils.py:1452] Setting max_num_batched_tokens to 8192 for OPENAI_API_SERVER usage context.
Traceback (most recent call last):
  File "/home/afeldman/git_clones/vllm/venv/bin/vllm", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/cli/main.py", line 73, in main
    args.dispatch_function(args)
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/cli/serve.py", line 33, in cmd
    uvloop.run(run_server(args))
  File "/home/afeldman/git_clones/vllm/venv/lib/python3.12/site-packages/uvloop/__init__.py", line 109, in run
    return __asyncio.run(
           ^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/asyncio/runners.py", line 195, in run
    return runner.run(main)
           ^^^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/asyncio/runners.py", line 118, in run
    return self._loop.run_until_complete(task)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
  File "/home/afeldman/git_clones/vllm/venv/lib/python3.12/site-packages/uvloop/__init__.py", line 61, in wrapper
    return await main
           ^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/openai/api_server.py", line 947, in run_server
    async with build_async_engine_client(args) as engine_client:
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/openai/api_server.py", line 139, in build_async_engine_client
    async with build_async_engine_client_from_engine_args(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/entrypoints/openai/api_server.py", line 163, in build_async_engine_client_from_engine_args
    engine_client = AsyncLLMEngine.from_engine_args(
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/v1/engine/async_llm.py", line 107, in from_engine_args
    vllm_config = engine_args.create_engine_config(usage_context)
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/engine/arg_utils.py", line 1194, in create_engine_config
    model_config = self.create_model_config()
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/engine/arg_utils.py", line 1121, in create_model_config
    return ModelConfig(
           ^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/config.py", line 327, in __init__
    hf_config = get_config(self.hf_config_path or self.model,
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/afeldman/git_clones/vllm/vllm/transformers_utils/config.py", line 326, in get_config
    raise ValueError(f"Unsupported config format: {config_format}")
ValueError: Unsupported config format: ConfigFormat.AUTO

my environment using 2 H20(have 8 GPU)

[VastOcean-Yang]

i meet the same promble when i use 2×8×H20 deploy deepseek r1,how can i solve it

[VastOcean-Yang]

i meet the same promble when i use 2_8_H20 deploy deepseek r1,how can i solve it

i install these pkg:

-rw-rw-r-- 1 ocean ocean 12739445 4月 9日 15:11 pandas-2.2.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
-rw-rw-r-- 1 ocean ocean 42087026 4月 9日 14:37 pyarrow-19.0.1-cp312-cp312-manylinux_2_28_x86_64.whl
-rw-rw-r-- 1 ocean ocean 509225 4月 9日 15:35 pytz-2025.2-py2.py3-none-any.whl
-rw-rw-r-- 1 ocean ocean 347839 4月 9日 15:36 tzdata-2025.2-py2.py3-none-any.whl

and then i run deepseek r1 successfully with vllm_v0.8.1

[fungaren]

#15100