安装vllm问题:

44u64gxh  于 2个月前  发布在  其他
关注(0)|答案(2)|浏览(46)

依赖问题

通过vllm==0.3.2+cu118安装pip install https://github.com/vllm-project/vllm/releases/download/v${VLLM_VERSION}/vllm-${VLLM_VERSION}+cu118-cp${PYTHON_VERSION}-cp${PYTHON_VERSION}-manylinux1_x86_64.whl,并使用

export VLLM_VERSION=0.3.2
export PYTHON_VERSION=311

。但是,该vllm whl依赖于cupy_cuda12x ...

详细安装日志

❯ pip install https://github.com/vllm-project/vllm/releases/download/v${VLLM_VERSION}/vllm-${VLLM_VERSION}+cu118-cp${PYTHON_VERSION}-cp${PYTHON_VERSION}-manylinux1_x86_64.whl
Collecting vllm==0.3.2+cu118
  Downloading https://github.com/vllm-project/vllm/releases/download/v0.3.2/vllm-0.3.2+cu118-cp311-cp311-manylinux1_x86_64.whl (41.2 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 41.2/41.2 MB 2.5 MB/s eta 0:00:00
Collecting ninja (from vllm==0.3.2+cu118)
  Using cached ninja-1.11.1.1-py2.py3-none-manylinux1_x86_64.manylinux_2_5_x86_64.whl.metadata (5.3 kB)
Requirement already satisfied: psutil in /opt/conda/envs/envd/lib/python3.11/site-packages (from vllm==0.3.2+cu118) (5.9.8)
Collecting ray>=2.9 (from vllm==0.3.2+cu118)
  Using cached ray-2.9.3-cp311-cp311-manylinux2014_x86_64.whl.metadata (13 kB)
Collecting sentencepiece (from vllm==0.3.2+cu118)
  Using cached sentencepiece-0.2.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (7.7 kB)
Requirement already satisfied: numpy in /opt/conda/envs/envd/lib/python3.11/site-packages (from vllm==0.3.2+cu118) (1.26.3)
Requirement already satisfied: torch==2.1.2 in /opt/conda/envs/envd/lib/python3.11/site-packages (from vllm==0.3.2+cu118) (2.1.2+cu118)
Collecting transformers>=4.38.0 (from vllm==0.3.2+cu118)
  Using cached transformers-4.38.1-py3-none-any.whl.metadata (131 kB)
Collecting xformers==0.0.23.post1 (from vllm==0.3.2+cu118)
  Using cached xformers-0.0.23.post1-cp311-cp311-manylinux2014_x86_64.whl.metadata (1.0 kB)
Collecting fastapi (from vllm==0.3.2+cu118)
  Using cached fastapi-0.110.0-py3-none-any.whl.metadata (25 kB)
Collecting uvicorn[standard] (from vllm==0.3.2+cu118)
  Using cached uvicorn-0.27.1-py3-none-any.whl.metadata (6.3 kB)
Collecting pydantic>=2.0 (from vllm==0.3.2+cu118)
  Using cached pydantic-2.6.3-py3-none-any.whl.metadata (84 kB)
Collecting aioprometheus[starlette] (from vllm==0.3.2+cu118)
  Using cached aioprometheus-23.12.0-py3-none-any.whl.metadata (9.8 kB)
Collecting pynvml==11.5.0 (from vllm==0.3.2+cu118)
  Using cached pynvml-11.5.0-py3-none-any.whl.metadata (7.8 kB)
Requirement already satisfied: triton>=2.1.0 in /opt/conda/envs/envd/lib/python3.11/site-packages (from vllm==0.3.2+cu118) (2.1.0)
Collecting cupy-cuda12x==12.1.0 (from vllm==0.3.2+cu118)
  Using cached cupy_cuda12x-12.1.0-cp311-cp311-manylinux2014_x86_64.whl.metadata (2.6 kB)
Requirement already satisfied: fastrlock>=0.5 in /opt/conda/envs/envd/lib/python3.11/site-packages (from cupy-cuda12x==12.1.0->vllm==0.3.2+cu118) (0.8.2)
Requirement already satisfied: filelock in /opt/conda/envs/envd/lib/python3.11/site-packages (from torch==2.1.2->vllm==0.3.2+cu118) (3.9.0)
Requirement already satisfied: typing-extensions in /opt/conda/envs/envd/lib/python3.11/site-packages (from torch==2.1.2->vllm==0.3.2+cu118) (4.8.0)
Requirement already satisfied: sympy in /opt/conda/envs/envd/lib/python3.11/site-packages (from torch==2.1.2->vllm==0.3.2+cu118) (1.12)
Requirement already satisfied: networkx in /opt/conda/envs/envd/lib/python3.11/site-packages (from torch==2.1.2->vllm==0.3.2+cu118) (3.2.1)
Requirement already satisfied: jinja2 in /opt/conda/envs/envd/lib/python3.11/site-packages (from torch==2.1.2->vllm==0.3.2+cu118) (3.1.2)
Requirement already satisfied: fsspec in /opt/conda/envs/envd/lib/python3.11/site-packages (from torch==2.1.2->vllm==0.3.2+cu118) (2023.4.0)
Collecting annotated-types>=0.4.0 (from pydantic>=2.0->vllm==0.3.2+cu118)
  Using cached annotated_types-0.6.0-py3-none-any.whl.metadata (12 kB)
Collecting pydantic-core==2.16.3 (from pydantic>=2.0->vllm==0.3.2+cu118)
  Using cached pydantic_core-2.16.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (6.5 kB)
Collecting click>=7.0 (from ray>=2.9->vllm==0.3.2+cu118)                                                                                                                                                                                                [23/167]
  Using cached click-8.1.7-py3-none-any.whl.metadata (3.0 kB)
Collecting jsonschema (from ray>=2.9->vllm==0.3.2+cu118)
  Using cached jsonschema-4.21.1-py3-none-any.whl.metadata (7.8 kB)
Collecting msgpack<2.0.0,>=1.0.0 (from ray>=2.9->vllm==0.3.2+cu118)
  Using cached msgpack-1.0.7-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (9.1 kB)
Collecting packaging (from ray>=2.9->vllm==0.3.2+cu118)
  Using cached packaging-23.2-py3-none-any.whl.metadata (3.2 kB)
Collecting protobuf!=3.19.5,>=3.15.3 (from ray>=2.9->vllm==0.3.2+cu118)
  Using cached protobuf-4.25.3-cp37-abi3-manylinux2014_x86_64.whl.metadata (541 bytes)
Collecting pyyaml (from ray>=2.9->vllm==0.3.2+cu118)
  Using cached PyYAML-6.0.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (2.1 kB)
Collecting aiosignal (from ray>=2.9->vllm==0.3.2+cu118)
  Using cached aiosignal-1.3.1-py3-none-any.whl.metadata (4.0 kB)
Collecting frozenlist (from ray>=2.9->vllm==0.3.2+cu118)
  Using cached frozenlist-1.4.1-cp311-cp311-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (12 kB)
Requirement already satisfied: requests in /opt/conda/envs/envd/lib/python3.11/site-packages (from ray>=2.9->vllm==0.3.2+cu118) (2.28.1)
Collecting huggingface-hub<1.0,>=0.19.3 (from transformers>=4.38.0->vllm==0.3.2+cu118)
  Using cached huggingface_hub-0.21.3-py3-none-any.whl.metadata (13 kB)
Collecting regex!=2019.12.17 (from transformers>=4.38.0->vllm==0.3.2+cu118)
  Using cached regex-2023.12.25-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (40 kB)
Collecting tokenizers<0.19,>=0.14 (from transformers>=4.38.0->vllm==0.3.2+cu118)
  Using cached tokenizers-0.15.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (6.7 kB)
Collecting safetensors>=0.4.1 (from transformers>=4.38.0->vllm==0.3.2+cu118)
  Using cached safetensors-0.4.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.8 kB)
Collecting tqdm>=4.27 (from transformers>=4.38.0->vllm==0.3.2+cu118)
  Using cached tqdm-4.66.2-py3-none-any.whl.metadata (57 kB)
Collecting orjson (from aioprometheus[starlette]->vllm==0.3.2+cu118)
  Using cached orjson-3.9.15-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (49 kB)
Collecting quantile-python>=1.1 (from aioprometheus[starlette]->vllm==0.3.2+cu118)
  Using cached quantile-python-1.1.tar.gz (2.9 kB)
  Preparing metadata (setup.py) ... done
Collecting starlette>=0.14.2 (from aioprometheus[starlette]->vllm==0.3.2+cu118)
  Using cached starlette-0.37.1-py3-none-any.whl.metadata (5.9 kB)
  Using cached starlette-0.36.3-py3-none-any.whl.metadata (5.9 kB)
Collecting h11>=0.8 (from uvicorn[standard]->vllm==0.3.2+cu118)
  Using cached h11-0.14.0-py3-none-any.whl.metadata (8.2 kB)
Collecting httptools>=0.5.0 (from uvicorn[standard]->vllm==0.3.2+cu118)
  Using cached httptools-0.6.1-cp311-cp311-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.6 kB)
Collecting python-dotenv>=0.13 (from uvicorn[standard]->vllm==0.3.2+cu118)
  Using cached python_dotenv-1.0.1-py3-none-any.whl.metadata (23 kB)
Collecting uvloop!=0.15.0,!=0.15.1,>=0.14.0 (from uvicorn[standard]->vllm==0.3.2+cu118)
  Using cached uvloop-0.19.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (4.9 kB)
Collecting watchfiles>=0.13 (from uvicorn[standard]->vllm==0.3.2+cu118)
  Using cached watchfiles-0.21.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (4.9 kB)
Collecting websockets>=10.4 (from uvicorn[standard]->vllm==0.3.2+cu118)
  Using cached websockets-12.0-cp311-cp311-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (6.6 kB)
Collecting fsspec (from torch==2.1.2->vllm==0.3.2+cu118)
  Using cached fsspec-2024.2.0-py3-none-any.whl.metadata (6.8 kB)
Collecting anyio<5,>=3.4.0 (from starlette>=0.14.2->aioprometheus[starlette]->vllm==0.3.2+cu118)
  Using cached anyio-4.3.0-py3-none-any.whl.metadata (4.6 kB)
Requirement already satisfied: MarkupSafe>=2.0 in /opt/conda/envs/envd/lib/python3.11/site-packages (from jinja2->torch==2.1.2->vllm==0.3.2+cu118) (2.1.3)
Collecting attrs>=22.2.0 (from jsonschema->ray>=2.9->vllm==0.3.2+cu118)
  Using cached attrs-23.2.0-py3-none-any.whl.metadata (9.5 kB)
Collecting jsonschema-specifications>=2023.03.6 (from jsonschema->ray>=2.9->vllm==0.3.2+cu118)
  Using cached jsonschema_specifications-2023.12.1-py3-none-any.whl.metadata (3.0 kB)
Collecting referencing>=0.28.4 (from jsonschema->ray>=2.9->vllm==0.3.2+cu118)
  Using cached referencing-0.33.0-py3-none-any.whl.metadata (2.7 kB)
Collecting rpds-py>=0.7.1 (from jsonschema->ray>=2.9->vllm==0.3.2+cu118)
  Using cached rpds_py-0.18.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (4.1 kB)
Requirement already satisfied: charset-normalizer<3,>=2 in /opt/conda/envs/envd/lib/python3.11/site-packages (from requests->ray>=2.9->vllm==0.3.2+cu118) (2.1.1)
Requirement already satisfied: idna<4,>=2.5 in /opt/conda/envs/envd/lib/python3.11/site-packages (from requests->ray>=2.9->vllm==0.3.2+cu118) (3.4)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/conda/envs/envd/lib/python3.11/site-packages (from requests->ray>=2.9->vllm==0.3.2+cu118) (1.26.13)
Requirement already satisfied: certifi>=2017.4.17 in /opt/conda/envs/envd/lib/python3.11/site-packages (from requests->ray>=2.9->vllm==0.3.2+cu118) (2022.12.7)
Requirement already satisfied: mpmath>=0.19 in /opt/conda/envs/envd/lib/python3.11/site-packages (from sympy->torch==2.1.2->vllm==0.3.2+cu118) (1.3.0)
Collecting sniffio>=1.1 (from anyio<5,>=3.4.0->starlette>=0.14.2->aioprometheus[starlette]->vllm==0.3.2+cu118)
  Using cached sniffio-1.3.1-py3-none-any.whl.metadata (3.9 kB)
Downloading cupy_cuda12x-12.1.0-cp311-cp311-manylinux2014_x86_64.whl (83.5 MB)
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 83.5/83.5 MB 3.2 MB/s eta 0:00:00
Downloading pynvml-11.5.0-py3-none-any.whl (53 kB)
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 53.1/53.1 kB 780.4 kB/s eta 0:00:00
Downloading xformers-0.0.23.post1-cp311-cp311-manylinux2014_x86_64.whl (213.0 MB)
   ━━━━━━━━━━━━━━━━━━━━━━━━╺━━━━━━━━━━━━━━━ 128.5/213.0 MB 3.2 MB/s eta 0:00:27
ERROR: Operation cancelled by user
dohp0rv5

dohp0rv51#

CUDA 11.8用户应该安装cupy-cuda11x。
参考自:requirement.txt

相关问题