07-04 周四 关于vLLM(LLMs_inference)源码安装过程问题与解决

  • 开发
  • 23
07-04 周四 关于LLMs_inference源码安装过程问题与解决
时间 版本 修改人 描述
2024年7月4日09:48:09 V0.1 宋全恒 新建文档

简介

 由于最近需要向vLLM上集成功能，因此，需要能够调试自己的仓库LLMs_Inference，该文档记录了源码编译的完整的过程。

 参考链接如下：

 正常简单执行下述的代码，即可完成源码的编译安装

git clone https://github.com/vllm-project/vllm.git
cd vllm
# export VLLM_INSTALL_PUNICA_KERNELS=1 # optionally build for multi-LoRA capability
pip install -e .  # This may take 5-10 minutes.

 但实际上还是比较麻烦的。因为仓库LLMs_Inference是从vllm仓库fork出来的，所以理论上应该是一样的。

仓库介绍

 仓库中有多个依赖环境，

image-20240704100120109

这些文件通常用于记录项目的依赖关系，以便在特定环境中进行安装和配置。

  • requirements.txt：一般用于列出项目所需的所有依赖项及其版本要求。通过在该文件中指定所需的库和版本，方便一次性安装所有依赖。

  • requirements-cpu.txt、requirements-cuda.txt、requirements-rocm.txt、requirements-neuron.txt：这些文件可能是针对不同的硬件或计算环境的特定依赖列表。例如，requirements-cuda.txt 可能包含与 CUDA（Compute Unified Device Architecture，一种并行计算平台和编程模型）相关的依赖；requirements-rocm.txt 可能涉及 ROCm（Radeon Open Compute platform，AMD 的开源计算平台）的依赖；requirements-neuron.txt 也许和特定的神经元芯片或相关技术的依赖有关。

 而 requirements-dev.txt 通常用于开发环境所需的额外依赖项，这些依赖可能不是项目在运行时必需的，但对于开发、测试、构建等过程是需要的。

源码编译安装 vLLM 是否需要安装所有这些依赖文件，取决于你的具体需求和使用场景。

如果你计划在特定的硬件环境（如使用 CUDA、ROCM 等）中运行 vLLM 或进行相关开发，那么可能需要根据相应的环境安装对应的依赖文件。

以安装 vLLM 为例，通常需要先创建 conda 环境并激活，然后查看 requirements.txt 中指定的 PyTorch 版本等依赖信息，再进行安装。

vllm开发环境准备

直接在宿主机上安装

解决torch依赖下载问题

(llms_inference) yuzailiang@ubuntu:/mnt/self-define/sunning/lmdeploy/LLMs_Inference$ python -c "import torch; print('device count:',torch.cuda.device_count(), 'available: ', torch.cuda.is_available())"
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/home/yuzailiang/anaconda3/envs/llms_inference/lib/python3.9/site-packages/torch/__init__.py", line 237, in <module>
    from torch._C import *  # noqa: F403
ImportError: /home/yuzailiang/anaconda3/envs/llms_inference/lib/python3.9/site-packages/torch/lib/../../nvidia/cusparse/lib/libcusparse.so.12: undefined symbol: __nvJitLinkAddData_12_1, version libnvJitLink.so.12

 上面是直接安装时发现虽然安装了torch==2.3.0但是，无法使用gpu。

解决方式：

 单独安装torch依赖

pip install torch==2.3.0 torchvision torchaudio --index-url  https://download.pytorch.org/whl/cu118

 安装之后，验证torch可以正确的驱动CUDA调用GPU

(llms_inference) yuzailiang@ubuntu:/mnt/self-define/sunning/lmdeploy/LLMs_Inference$ python -c "import torch; print('device count:',torch.cuda.device_count(), 'available: ', torch.cuda.is_available())"
device count: 8 available:  True

继续执行pip install -e .

 问题还是存在

packages/torch/__init__.py", line 237, in <module>
    from torch._C import *  # noqa: F403
ImportError: /home/yuzailiang/anaconda3/envs/llms_inference/lib/python3.9/site-packages/torch/lib/../../nvidia/cusparse/lib/libcusparse.so.12: undefined symbol: __nvJitLinkAddData_12_1, version libnvJitLink.so.12

尝试使用cu121-Couldn’t find CUDA library root.

pip install torch==2.3.0 torchvision torchaudio --index-url  https://download.pytorch.org/whl/cu121

Building wheels for collected packages: vllm
  Building editable for vllm (pyproject.toml) ... error
  error: subprocess-exited-with-error
  
  × Building editable for vllm (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [139 lines of output]
      running editable_wheel
      creating /tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info
      writing /tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info/PKG-INFO
      writing dependency_links to /tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info/dependency_links.txt
      writing requirements to /tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info/requires.txt
      writing top-level names to /tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info/top_level.txt
      writing manifest file '/tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info/SOURCES.txt'
      reading manifest file '/tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info/SOURCES.txt'
      reading manifest template 'MANIFEST.in'
      adding license file 'LICENSE'
      writing manifest file '/tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm.egg-info/SOURCES.txt'
      creating '/tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm-0.4.2+cu120.dist-info'
      creating /tmp/pip-wheel-c7m73v0l/.tmp-t6j0dz53/vllm-0.4.2+cu120.dist-info/WHEEL
      running build_py
      running build_ext
      -- The CXX compiler identification is GNU 9.4.0
      -- Detecting CXX compiler ABI info
      -- Detecting CXX compiler ABI info - done
      -- Check for working CXX compiler: /usr/bin/c++ - skipped
      -- Detecting CXX compile features
      -- Detecting CXX compile features - done
      -- Build type: RelWithDebInfo
      -- Target device: cuda
      -- Found Python: /home/yuzailiang/anaconda3/envs/llms_inference/bin/python3.9 (found version "3.9.19") found components: Interpreter Development.Module
      -- Found python matching: /home/yuzailiang/anaconda3/envs/llms_inference/bin/python3.9.
      -- Found CUDA: /usr/local/cuda-12.0 (found version "12.0")
      CMake Error at /tmp/pip-build-env-xxhrqd7n/overlay/lib/python3.9/site-packages/cmake/data/share/cmake-3.30/Modules/Internal/CMakeCUDAFindToolkit.cmake:148 (message):
        Couldn't find CUDA library root.
      Call Stack (most recent call first):
      
            subprocess.CalledProcessError: Command '['cmake', '/mnt/self-define/sunning/lmdeploy/LLMs_Inference', '-G', 'Ninja', '-DCMAKE_BUILD_TYPE=RelWithDebInfo', '-DCMAKE_LIBRARY_OUTPUT_DIRECTORY=/tmp/tmpwkyqp9r6.build-lib/vllm', '-DCMAKE_ARCHIVE_OUTPUT_DIRECTORY=/tmp/tmpuvdjn65m.build-temp', '-DVLLM_TARGET_DEVICE=cuda', '-DCMAKE_CXX_COMPILER_LAUNCHER=ccache', '-DCMAKE_CUDA_COMPILER_LAUNCHER=ccache', '-DVLLM_PYTHON_EXECUTABLE=/home/yuzailiang/anaconda3/envs/llms_inference/bin/python3.9', '-DNVCC_THREADS=1', '-DCMAKE_JOB_POOL_COMPILE:STRING=compile', '-DCMAKE_JOB_POOLS:STRING=compile=256']' returned non-zero exit status 1.

image-20240708104936789

上述是非常复杂的环境，因为显示正在使用的cu120,即cuda 12.0.

Building wheels for collected packages: vllm
  Building editable for vllm (pyproject.toml) ... error
  
subprocess.CalledProcessError: Command '['cmake', '/mnt/self-define/sunning/lmdeploy/LLMs_Inference', '-G', 'Ninja', '-DCMAKE_BUILD_TYPE=RelWithDebInfo', '-DCMAKE_LIBRARY_OUTPUT_DIRECTORY=/tmp/tmplnbtei_9.build-lib/vllm', '-DCMAKE_ARCHIVE_OUTPUT_DIRECTORY=/tmp/tmpdu5xwhxm.build-temp', '-DVLLM_TARGET_DEVICE=cuda', '-DCMAKE_CXX_COMPILER_LAUNCHER=ccache', '-DCMAKE_CUDA_COMPILER_LAUNCHER=ccache', '-DVLLM_PYTHON_EXECUTABLE=/home/yuzailiang/anaconda3/envs/llms_inference/bin/python3.9', '-DNVCC_THREADS=1', '-DCMAKE_JOB_POOL_COMPILE:STRING=compile', '-DCMAKE_JOB_POOLS:STRING=compile=256']' returned non-zero exit status 1.

 但是nvidia-smi显示的cuda版本为12.4，就很奇怪。

(base) yuzailiang@ubuntu:/mnt/self-define/sunning/lmdeploy/LLMs_Inference$ nvidia-smi 
Tue Jul  9 09:02:46 2024       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.54.14              Driver Version: 550.54.14      CUDA Version: 12.4     |
|-----------------------------------------+------------------------+---------------------

 而在 /usr/local目录下，并没有这个cuda12.4.

(base) yuzailiang@ubuntu:/usr/local$ ll | grep cuda
lrwxrwxrwx  1 root root   22 Jul  8 05:22 cuda -> /etc/alternatives/cuda/
lrwxrwxrwx  1 root root   25 Jul  8 05:22 cuda-12 -> /etc/alternatives/cuda-12/
drwxr-xr-x 17 root root 4096 Jun 25 08:42 cuda-12.0/
drwxr-xr-x 15 root root 4096 Jul  8 05:22 cuda-12.1/

放弃宿主环境改用容器

最终还是花费了几天之后，改用镜像来进行开发环境的搭建。

docker run -d --name smoothquant --gpus all -v /mnt/self-define:/mnt/self-define -p 8022:22 -it llm_inference:v1.0

基本思路，就是启动容器，将必要的目录挂载进去，并且在容器中部署sshd服务，使用vscode将容器当成一个独立的Linux机器，以实现断点调试开发的功能。

 具体过程参见 镜像启动添加sshd服务。

 基本过程如下所示：

1. 启动容器，赋予全部PGU，以及适当的挂载目录和端口映射。
2. 容器安装sshd服务，（ubuntu和centos不一样）
3. 配置sshd服务，修改root密码
4. 启动sshd服务
5. vscode直接连接容器(作为远程linux服务器)。
6. 搭建调试环境和断点。 参见 [05-16 周四 vscode 搭建远程调试环境_python vscode server 远程调试-CSDN博客](https://blog.csdn.net/lk142500/article/details/138969211?spm=1001.2014.3001.5502)

编辑CMakeLists.txt

LLMs_Inference/CMakeLists.txt里的地178行 FetchContent_Declare(
cutlass
GIT_REPOSITORY https://github.com/nvidia/cutlass.git
# CUTLASS 3.5.0
GIT_TAG 7d49e6c7e2f8896c47f586706e67e1fb215529dc
)
改成
FetchContent_Declare(
cutlass
SOURCE_DIR /cutlass#GIT_REPOSITORY https://github.com/nvidia/cutlass.git
# CUTLASS 3.5.0
#GIT_TAG 7d49e6c7e2f8896c47f586706e67e1fb215529dc
)

执行源码安装

pip install -e .
root@145206f3e691:/mnt/self-define/sunning/lmdeploy/LLMs_Inference# pip install -e .
Obtaining file:///mnt/self-define/sunning/lmdeploy/LLMs_Inference
  Installing build dependencies ... done
  Checking if build backend supports build_editable ... done
  Getting requirements to build wheel ... done
  Preparing metadata (pyproject.toml) ... done
Requirement already satisfied: psutil in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (5.9.8)
Requirement already satisfied: prometheus-fastapi-instrumentator>=7.0.0 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (7.0.0)
Requirement already satisfied: aiohttp in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (3.9.3)
Requirement already satisfied: outlines==0.0.34 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.0.34)
Requirement already satisfied: lm-format-enforcer==0.10.1 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.10.1)
Requirement already satisfied: tokenizers>=0.19.1 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.19.1)
Requirement already satisfied: ninja in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (1.11.1.1)
Requirement already satisfied: pydantic>=2.0 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (2.6.4)
Requirement already satisfied: prometheus-client>=0.18.0 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.20.0)
Requirement already satisfied: cmake>=3.21 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (3.28.4)
Requirement already satisfied: tiktoken>=0.6.0 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.7.0)
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (1.26.4)
Requirement already satisfied: openai in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (1.31.0)
Requirement already satisfied: vllm-flash-attn==2.5.8.post2 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (2.5.8.post2)
Requirement already satisfied: fastapi in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.110.0)
Requirement already satisfied: uvicorn[standard] in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.29.0)
Requirement already satisfied: torch==2.3.0 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (2.3.0)
Requirement already satisfied: ray>=2.9 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (2.10.0)
Requirement already satisfied: transformers>=4.40.0 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (4.42.3)
Requirement already satisfied: py-cpuinfo in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (9.0.0)
Requirement already satisfied: xformers==0.0.26.post1 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.0.26.post1)
Requirement already satisfied: sentencepiece in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (0.2.0)
Requirement already satisfied: vllm-nccl-cu12<2.19,>=2.18 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (2.18.1.0.4.0)
Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (4.10.0)
Requirement already satisfied: nvidia-ml-py in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (12.555.43)
Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (2.31.0)
Requirement already satisfied: filelock>=3.10.4 in /usr/local/lib/python3.10/dist-packages (from vllm==0.4.2+cu124) (3.13.1)
Requirement already satisfied: pyyaml in /usr/local/lib/python3.10/dist-packages (from lm-format-enforcer==0.10.1->vllm==0.4.2+cu124) (6.0.1)
Requirement already satisfied: interegular>=0.3.2 in /usr/local/lib/python3.10/dist-packages (from lm-format-enforcer==0.10.1->vllm==0.4.2+cu124) (0.3.3)
Requirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from lm-format-enforcer==0.10.1->vllm==0.4.2+cu124) (24.0)
Requirement already satisfied: nest-asyncio in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (1.6.0)
Requirement already satisfied: numba in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (0.59.1)
Requirement already satisfied: cloudpickle in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (3.0.0)
Requirement already satisfied: joblib in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (1.3.2)
Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (1.12.0)
Requirement already satisfied: diskcache in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (5.6.3)
Requirement already satisfied: referencing in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (0.34.0)
Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (3.1.3)
Requirement already satisfied: lark in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (1.1.9)
Requirement already satisfied: jsonschema in /usr/local/lib/python3.10/dist-packages (from outlines==0.0.34->vllm==0.4.2+cu124) (4.21.1)
Requirement already satisfied: nvidia-curand-cu12==10.3.2.106 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (10.3.2.106)
Requirement already satisfied: nvidia-cufft-cu12==11.0.2.54 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (11.0.2.54)
Requirement already satisfied: nvidia-cudnn-cu12==8.9.2.26 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (8.9.2.26)
Requirement already satisfied: nvidia-cuda-runtime-cu12==12.1.105 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (12.1.105)
Requirement already satisfied: triton==2.3.0 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (2.3.0)
Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (1.12)
Requirement already satisfied: nvidia-cuda-nvrtc-cu12==12.1.105 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (12.1.105)
Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (2024.2.0)
Requirement already satisfied: nvidia-cuda-cupti-cu12==12.1.105 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (12.1.105)
Requirement already satisfied: nvidia-cusolver-cu12==11.4.5.107 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (11.4.5.107)
Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (3.2.1)
Requirement already satisfied: nvidia-nvtx-cu12==12.1.105 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (12.1.105)
Requirement already satisfied: nvidia-cublas-cu12==12.1.3.1 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (12.1.3.1)
Requirement already satisfied: nvidia-nccl-cu12==2.20.5 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (2.20.5)
Requirement already satisfied: nvidia-cusparse-cu12==12.1.0.106 in /usr/local/lib/python3.10/dist-packages (from torch==2.3.0->vllm==0.4.2+cu124) (12.1.0.106)
Requirement already satisfied: nvidia-nvjitlink-cu12 in /usr/local/lib/python3.10/dist-packages (from nvidia-cusolver-cu12==11.4.5.107->torch==2.3.0->vllm==0.4.2+cu124) (12.4.99)
Requirement already satisfied: starlette<1.0.0,>=0.30.0 in /usr/local/lib/python3.10/dist-packages (from prometheus-fastapi-instrumentator>=7.0.0->vllm==0.4.2+cu124) (0.36.3)
Requirement already satisfied: annotated-types>=0.4.0 in /usr/local/lib/python3.10/dist-packages (from pydantic>=2.0->vllm==0.4.2+cu124) (0.6.0)
Requirement already satisfied: pydantic-core==2.16.3 in /usr/local/lib/python3.10/dist-packages (from pydantic>=2.0->vllm==0.4.2+cu124) (2.16.3)
Requirement already satisfied: msgpack<2.0.0,>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from ray>=2.9->vllm==0.4.2+cu124) (1.0.8)
Requirement already satisfied: aiosignal in /usr/local/lib/python3.10/dist-packages (from ray>=2.9->vllm==0.4.2+cu124) (1.3.1)
Requirement already satisfied: frozenlist in /usr/local/lib/python3.10/dist-packages (from ray>=2.9->vllm==0.4.2+cu124) (1.4.1)
Requirement already satisfied: protobuf!=3.19.5,>=3.15.3 in /usr/local/lib/python3.10/dist-packages (from ray>=2.9->vllm==0.4.2+cu124) (5.26.1)
Requirement already satisfied: click>=7.0 in /usr/local/lib/python3.10/dist-packages (from ray>=2.9->vllm==0.4.2+cu124) (8.1.7)
Requirement already satisfied: regex>=2022.1.18 in /usr/local/lib/python3.10/dist-packages (from tiktoken>=0.6.0->vllm==0.4.2+cu124) (2023.12.25)
Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->vllm==0.4.2+cu124) (2.2.1)
Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->vllm==0.4.2+cu124) (3.3.2)
Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->vllm==0.4.2+cu124) (3.6)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->vllm==0.4.2+cu124) (2024.2.2)
Requirement already satisfied: huggingface-hub<1.0,>=0.16.4 in /usr/local/lib/python3.10/dist-packages (from tokenizers>=0.19.1->vllm==0.4.2+cu124) (0.23.4)
Requirement already satisfied: safetensors>=0.4.1 in /usr/local/lib/python3.10/dist-packages (from transformers>=4.40.0->vllm==0.4.2+cu124) (0.4.2)
Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.10/dist-packages (from transformers>=4.40.0->vllm==0.4.2+cu124) (4.66.2)
Requirement already satisfied: yarl<2.0,>=1.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp->vllm==0.4.2+cu124) (1.9.4)
Requirement already satisfied: async-timeout<5.0,>=4.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp->vllm==0.4.2+cu124) (4.0.3)
Requirement already satisfied: attrs>=17.3.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp->vllm==0.4.2+cu124) (23.2.0)
Requirement already satisfied: multidict<7.0,>=4.5 in /usr/local/lib/python3.10/dist-packages (from aiohttp->vllm==0.4.2+cu124) (6.0.5)
Requirement already satisfied: httpx<1,>=0.23.0 in /usr/local/lib/python3.10/dist-packages (from openai->vllm==0.4.2+cu124) (0.27.0)
Requirement already satisfied: anyio<5,>=3.5.0 in /usr/local/lib/python3.10/dist-packages (from openai->vllm==0.4.2+cu124) (4.3.0)
Requirement already satisfied: distro<2,>=1.7.0 in /usr/local/lib/python3.10/dist-packages (from openai->vllm==0.4.2+cu124) (1.9.0)
Requirement already satisfied: sniffio in /usr/local/lib/python3.10/dist-packages (from openai->vllm==0.4.2+cu124) (1.3.1)
Requirement already satisfied: h11>=0.8 in /usr/local/lib/python3.10/dist-packages (from uvicorn[standard]->vllm==0.4.2+cu124) (0.14.0)
Requirement already satisfied: websockets>=10.4 in /usr/local/lib/python3.10/dist-packages (from uvicorn[standard]->vllm==0.4.2+cu124) (12.0)
Requirement already satisfied: python-dotenv>=0.13 in /usr/local/lib/python3.10/dist-packages (from uvicorn[standard]->vllm==0.4.2+cu124) (1.0.1)
Requirement already satisfied: httptools>=0.5.0 in /usr/local/lib/python3.10/dist-packages (from uvicorn[standard]->vllm==0.4.2+cu124) (0.6.1)
Requirement already satisfied: watchfiles>=0.13 in /usr/local/lib/python3.10/dist-packages (from uvicorn[standard]->vllm==0.4.2+cu124) (0.21.0)
Requirement already satisfied: uvloop!=0.15.0,!=0.15.1,>=0.14.0 in /usr/local/lib/python3.10/dist-packages (from uvicorn[standard]->vllm==0.4.2+cu124) (0.19.0)
Requirement already satisfied: exceptiongroup>=1.0.2 in /usr/local/lib/python3.10/dist-packages (from anyio<5,>=3.5.0->openai->vllm==0.4.2+cu124) (1.2.0)
Requirement already satisfied: httpcore==1.* in /usr/local/lib/python3.10/dist-packages (from httpx<1,>=0.23.0->openai->vllm==0.4.2+cu124) (1.0.5)
Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->outlines==0.0.34->vllm==0.4.2+cu124) (2.1.5)
Requirement already satisfied: jsonschema-specifications>=2023.03.6 in /usr/local/lib/python3.10/dist-packages (from jsonschema->outlines==0.0.34->vllm==0.4.2+cu124) (2023.12.1)
Requirement already satisfied: rpds-py>=0.7.1 in /usr/local/lib/python3.10/dist-packages (from jsonschema->outlines==0.0.34->vllm==0.4.2+cu124) (0.18.0)
Requirement already satisfied: llvmlite<0.43,>=0.42.0dev0 in /usr/local/lib/python3.10/dist-packages (from numba->outlines==0.0.34->vllm==0.4.2+cu124) (0.42.0)
Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch==2.3.0->vllm==0.4.2+cu124) (1.3.0)
Installing collected packages: vllm
  Attempting uninstall: vllm
    Found existing installation: vllm 0.4.2+cu124
    Uninstalling vllm-0.4.2+cu124:
      Successfully uninstalled vllm-0.4.2+cu124
  Running setup.py develop for vllm


Successfully installed vllm-0.4.2+cu124
WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv

vscode建立连接容器并建立python调试环境

 相当于使用vscode直接连接到容器环境中。具体过程可以参见

07-09 周二 镜像启动容器添加openssh，使用vscode断点调试Python工程

 上述笔记详细的记录了容器启动，openssh-server的安装以及配置和sshd服务的启动，

05-16 周四 vscode 搭建远程调试环境_python vscode server 远程调试-CSDN博客

总结

 深度学习的开发，一般一个组共享同一个基础环境，还是太复杂容易出问题了，所以准备一个容器，隔离开使用环境反而是一个不错的方式，能够减少问题出现的概率。

 经过本文的实践，对于想要基于pip安装的仓库中继续进行开发而搭建，需要的基本步骤如下：

  1. git clone或者其他方式获取源码
  2. 下载依赖
  3. 源码编译

 经过源码编译之后，程序就可以顺利的进行断点调试了。在行文撰写过程中，为了使用vscode连接容器，也撰写了07-09 周二 镜像启动容器添加openssh，使用vscode断点调试Python工程

阅读全部

相关推荐

  5. 解决ros2缺少cv_bridge包的问题——cv_bridge源码编译安装过程

    2024-07-10 11:06:03       43 阅读

  7. Ubuntu 22.04上构建libvirt源码错误解决

    2024-07-10 11:06:03       35 阅读

  10. 在 Ubuntu 22.04 上源码安装 Podman 4

    2024-07-10 11:06:03       40 阅读

最近更新

  3. docker php8.1+nginx base 镜像 dockerfile 配置

    2024-07-10 11:06:03       99 阅读

  4. Could not load dynamic library ‘cudart64_100.dll‘

    2024-07-10 11:06:03       107 阅读

  9. 在Django里面运行非项目文件

    2024-07-10 11:06:03       90 阅读

  19. Python语言-面向对象

    2024-07-10 11:06:03       98 阅读

  20. 如何分清楚常见的 Git 分支管理策略Git Flow、GitHub Flow 和 GitLab Flow

    2024-07-10 11:06:03       96 阅读

热门阅读

  1. SD卡，laptop，启动ubtuntu

    2024-07-10 11:06:03       26 阅读

  4. 达梦数据库主备手动切换

    2024-07-10 11:06:03       29 阅读

  6. 创建数据库表的语法定义包含了SQL Server、Mysql、PostgreSQL、SQLite的示例

    2024-07-10 11:06:03       31 阅读

  7. ffmpeg滤镜创建过程

    2024-07-10 11:06:03       30 阅读

  8. stm32使用单通道规则组ADC

    2024-07-10 11:06:03       30 阅读

  9. HTML 学习指南：从入门到精通

    2024-07-10 11:06:03       29 阅读

  10. 经导管二尖瓣修复系统市场全面解析与未来展望

    2024-07-10 11:06:03       28 阅读

  11. MySQL运维实战之ProxySQL（9.5）proxysql和MySQL Group Replication配合使用

    2024-07-10 11:06:03       25 阅读

  14. 大众点评商家电话采集软件分享 解析爬虫工具

    2024-07-10 11:06:03       25 阅读

  16. Python 使用正则提取字符串

    2024-07-10 11:06:03       22 阅读

  18. 低代码研发项目管理流程优化：提效与创新的双重驱动

    2024-07-10 11:06:03       26 阅读

  19. 专业课笔记——（第十二章：文件的读写）

    2024-07-10 11:06:03       27 阅读

  21. lvs集群

    2024-07-10 11:06:03       26 阅读

  29. Perl 语言入门学习

    2024-07-10 11:06:03       28 阅读