feat: 新增 SGLang 0.5.8/0.5.9/0.5.10 容器镜像适配 OC9 · Pull Request !28 · OpenCloudOS/AI-Agent-Container

1. feat: 新增 SGLang 0.5.8/0.5.9/0.5.10 容器镜像适配 OC9

================== Test Result ====================
$bash test.sh 
可用SGLang 版本：
  1) 0.5.8
  2) 0.5.9
  3) 0.5.10
请选择SGLang 版本 [默认 3=0.5.10]: 
镜像 oc9-sglang:0.5.10 不存在，开始构建...
[+] Building 786.7s (10/10) FINISHED                                                                                                  docker:default
 => [internal] load build definition from Dockerfile                                                                                            0.0s
 => => transferring dockerfile: 1.26kB                                                                                                          0.0s
 => [internal] load metadata for docker.io/opencloudos/opencloudos9-cuda-devel:12.8                                                             0.2s
 => [internal] load .dockerignore                                                                                                               0.0s
 => => transferring context: 2B                                                                                                                 0.0s
 => [1/6] FROM docker.io/opencloudos/opencloudos9-cuda-devel:12.8@sha256:a644d113d2ae9df4ecb8fbec0c5b37c451a33710f140cc3a05031c46e19c2a0a       0.0s
 => => resolve docker.io/opencloudos/opencloudos9-cuda-devel:12.8@sha256:a644d113d2ae9df4ecb8fbec0c5b37c451a33710f140cc3a05031c46e19c2a0a       0.0s
 => CACHED [2/6] RUN dnf install -y         python3-pip         git     && dnf clean all     && rm -rf /var/cache/yum/*                         0.0s
 => [3/6] RUN pip3 install --no-cache-dir uv     && uv pip install --no-cache-dir --prerelease=allow "sglang[all]==0.5.10" --system            84.9s
 => [4/6] RUN find /usr/local/lib /usr/local/lib64 /usr/lib /usr/lib64 -type d -path '*/nvidia/*/lib' 2>/dev/null         > /etc/ld.so.conf.d/  0.9s 
 => [5/6] WORKDIR /workspace                                                                                                                    0.0s 
 => [6/6] RUN echo $(date +"%Y-%m-%dT%H:%M:%S%z") > /opencloudos_build_date.txt                                                                 0.3s 
 => exporting to image                                                                                                                        700.3s 
 => => exporting layers                                                                                                                       621.5s 
 => => exporting manifest sha256:58697b321fda2dc7c74b97a290c6173570a1b853ccaf7e4e7e3dfe5b9ae0f2a3                                               0.0s 
 => => exporting config sha256:6edfe3892b83d0255956d87e5d5ea700f0492a12bba113455111a5eadd117107                                                 0.0s
 => => exporting attestation manifest sha256:38d84ae44b357367b3ea6d29eb8081ffb51736ffa02fed7f73bddc8fab229517                                   0.0s
 => => exporting manifest list sha256:522ec68bd66b2b5388646b63b73216ad85a36cfee36a792f54dd66d5018bb46a                                          0.0s
 => => naming to docker.io/library/oc9-sglang:0.5.10                                                                                            0.0s
 => => unpacking to docker.io/library/oc9-sglang:0.5.10                                                                                        78.7s
✓ 镜像构建完成: oc9-sglang:0.5.10

=== SGLang 0.5.10 基础功能测试 ===
运行时间：2026-05-06 05:42:03
镜像：oc9-sglang:0.5.10

--- 阶段一：安装验证 ---
检查 sglang 版本... 0.5.10
✓ 通过
检查 sglang 主模块导入... ok
✓ 通过
检查 torch CUDA 构建... torch=2.9.1+cu128, cuda=12.8
✓ 通过

--- 阶段二：GPU 运行时检查 ---
检查 nvidia-smi... NVIDIA L40, 570.211.01, 46068 MiB
✓ 通过
检查 torch.cuda.is_available()... GPU 数量: 1
  GPU 0: NVIDIA L40
✓ 通过
检查 sglang.launch_server CLI... /usr/local/lib/python3.11/site-packages/sglang/launch_server.py:51: UserWarning: 'python -m sglang.launch_server' is still supported, but 'sglang serve' is the recommended entrypoint.
  Example: sglang serve --model-path <model> [options]
  warnings.warn(
✓ 通过
检查 sglang ServerArgs 导入... ok
✓ 通过
检查 sglang OpenAI 兼容入口导入... /usr/local/lib/python3.11/site-packages/sglang/srt/entrypoints/http_server.py:172: FastAPIDeprecationWarning: ORJSONResponse is deprecated, FastAPI now serializes data directly to JSON bytes via Pydantic when a return type or response model is set, which is faster and doesn't need a custom response class. Read more in the FastAPI docs: https://fastapi.tiangolo.com/advanced/custom-response/#orjson-or-response-model and https://fastapi.tiangolo.com/tutorial/response-model/
  from sglang.srt.utils.json_response import (
ok
✓ 通过
检查 sglang 平台探测... sglang=0.5.10, torch=2.9.1+cu128, cuda=12.8
✓ 通过

--- 阶段三：完整推理测试 ---
  请使用项目根目录的 cuda_test.sh 进行端到端推理验证：
  bash cuda_test.sh --vllm-model <model_path>

=== 所有测试通过 ===

-------------------------------------------
$bash test.sh 
可用SGLang 版本：
  1) 0.5.8
  2) 0.5.9
  3) 0.5.10
请选择SGLang 版本 [默认 3=0.5.10]: 1
镜像 oc9-sglang:0.5.8 不存在，开始构建...
[+] Building 738.5s (10/10) FINISHED                                                                                                  docker:default
 => [internal] load build definition from Dockerfile                                                                                            0.0s
 => => transferring dockerfile: 1.24kB                                                                                                          0.0s
 => [internal] load metadata for docker.io/opencloudos/opencloudos9-cuda-devel:12.8                                                             0.8s
 => [internal] load .dockerignore                                                                                                               0.0s
 => => transferring context: 2B                                                                                                                 0.0s
 => [1/6] FROM docker.io/opencloudos/opencloudos9-cuda-devel:12.8@sha256:a644d113d2ae9df4ecb8fbec0c5b37c451a33710f140cc3a05031c46e19c2a0a       0.0s
 => => resolve docker.io/opencloudos/opencloudos9-cuda-devel:12.8@sha256:a644d113d2ae9df4ecb8fbec0c5b37c451a33710f140cc3a05031c46e19c2a0a       0.0s
 => CACHED [2/6] RUN dnf install -y         python3-pip         git     && dnf clean all     && rm -rf /var/cache/yum/*                         0.0s
 => [3/6] RUN pip3 install --no-cache-dir uv     && uv pip install --no-cache-dir "sglang[all]==0.5.8" --system                                79.3s
 => [4/6] RUN find /usr/local/lib /usr/local/lib64 /usr/lib /usr/lib64 -type d -path '*/nvidia/*/lib' 2>/dev/null         > /etc/ld.so.conf.d/  0.9s 
 => [5/6] WORKDIR /workspace                                                                                                                    0.0s 
 => [6/6] RUN echo $(date +"%Y-%m-%dT%H:%M:%S%z") > /opencloudos_build_date.txt                                                                 0.3s 
 => exporting to image                                                                                                                        657.0s 
 => => exporting layers                                                                                                                       582.1s 
 => => exporting manifest sha256:9c2b3efecdb95e9ac392a89d18969c3e061817c156705d47b76d028f24568882                                               0.0s 
 => => exporting config sha256:adff9880da2b7a962082556d38a3350618367f547c7d7e576b2baa1fe121652d                                                 0.0s
 => => exporting attestation manifest sha256:06f20d64e304adbe6289b71ed92847b71951baa1229358f09cbd3e9115f4c74c                                   0.0s
 => => exporting manifest list sha256:125738961c57dcc1306e4929ee1ec50f7c3f985dc55a47d470b80cd314c12133                                          0.0s
 => => naming to docker.io/library/oc9-sglang:0.5.8                                                                                             0.0s
 => => unpacking to docker.io/library/oc9-sglang:0.5.8                                                                                         74.8s
✓ 镜像构建完成: oc9-sglang:0.5.8

=== SGLang 0.5.8 基础功能测试 ===
运行时间：2026-05-06 05:56:44
镜像：oc9-sglang:0.5.8

--- 阶段一：安装验证 ---
检查 sglang 版本... 0.5.8
✓ 通过
检查 sglang 主模块导入... ok
✓ 通过
检查 torch CUDA 构建... torch=2.9.1+cu128, cuda=12.8
✓ 通过

--- 阶段二：GPU 运行时检查 ---
检查 nvidia-smi... NVIDIA L40, 570.211.01, 46068 MiB
✓ 通过
检查 torch.cuda.is_available()... GPU 数量: 1
  GPU 0: NVIDIA L40
✓ 通过
检查 sglang.launch_server CLI... ✓ 通过
检查 sglang ServerArgs 导入... ok
✓ 通过
检查 sglang OpenAI 兼容入口导入... ok
✓ 通过
检查 sglang 平台探测... sglang=0.5.8, torch=2.9.1+cu128, cuda=12.8
✓ 通过

--- 阶段三：完整推理测试 ---
  请使用项目根目录的 cuda_test.sh 进行端到端推理验证：
  bash cuda_test.sh --vllm-model <model_path>

=== 所有测试通过 ===

-----------------------------------
$bash test.sh 
可用SGLang 版本：
  1) 0.5.8
  2) 0.5.9
  3) 0.5.10
请选择SGLang 版本 [默认 3=0.5.10]: 2
镜像 oc9-sglang:0.5.9 不存在，开始构建...
[+] Building 747.1s (10/10) FINISHED                                                                                                  docker:default
 => [internal] load build definition from Dockerfile                                                                                            0.0s
 => => transferring dockerfile: 1.24kB                                                                                                          0.0s
 => [internal] load metadata for docker.io/opencloudos/opencloudos9-cuda-devel:12.8                                                             0.6s
 => [internal] load .dockerignore                                                                                                               0.0s
 => => transferring context: 2B                                                                                                                 0.0s
 => [1/6] FROM docker.io/opencloudos/opencloudos9-cuda-devel:12.8@sha256:a644d113d2ae9df4ecb8fbec0c5b37c451a33710f140cc3a05031c46e19c2a0a       0.0s
 => => resolve docker.io/opencloudos/opencloudos9-cuda-devel:12.8@sha256:a644d113d2ae9df4ecb8fbec0c5b37c451a33710f140cc3a05031c46e19c2a0a       0.0s
 => CACHED [2/6] RUN dnf install -y         python3-pip         git     && dnf clean all     && rm -rf /var/cache/yum/*                         0.0s
 => [3/6] RUN pip3 install --no-cache-dir uv     && uv pip install --no-cache-dir "sglang[all]==0.5.9" --system                                77.0s
 => [4/6] RUN find /usr/local/lib /usr/local/lib64 /usr/lib /usr/lib64 -type d -path '*/nvidia/*/lib' 2>/dev/null         > /etc/ld.so.conf.d/  0.9s 
 => [5/6] WORKDIR /workspace                                                                                                                    0.0s 
 => [6/6] RUN echo $(date +"%Y-%m-%dT%H:%M:%S%z") > /opencloudos_build_date.txt                                                                 0.3s 
 => exporting to image                                                                                                                        668.0s 
 => => exporting layers                                                                                                                       593.0s 
 => => exporting manifest sha256:c5970f4ee60551079af285bcc5f9831c5c7fbed7b1348ae49cc11a78182676b3                                               0.0s 
 => => exporting config sha256:9b1a72665e22532ea25ab24edcd635848e8a2c38bed5c288ad05c42c7df0419a                                                 0.0s
 => => exporting attestation manifest sha256:248b38e501a38a5775d042374a8190153323f2d526a1bd4fb9c45bc6fc6dbe5e                                   0.0s
 => => exporting manifest list sha256:566843fba987dacde4e7c7113040eed3984cf2ede3b41311b99341a2367f2037                                          0.0s
 => => naming to docker.io/library/oc9-sglang:0.5.9                                                                                             0.0s
 => => unpacking to docker.io/library/oc9-sglang:0.5.9                                                                                         74.9s
✓ 镜像构建完成: oc9-sglang:0.5.9

=== SGLang 0.5.9 基础功能测试 ===
运行时间：2026-05-06 06:15:28
镜像：oc9-sglang:0.5.9

--- 阶段一：安装验证 ---
检查 sglang 版本... 0.5.9
✓ 通过
检查 sglang 主模块导入... ok
✓ 通过
检查 torch CUDA 构建... torch=2.9.1+cu128, cuda=12.8
✓ 通过

--- 阶段二：GPU 运行时检查 ---
检查 nvidia-smi... NVIDIA L40, 570.211.01, 46068 MiB
✓ 通过
检查 torch.cuda.is_available()... GPU 数量: 1
  GPU 0: NVIDIA L40
✓ 通过
检查 sglang.launch_server CLI... ✓ 通过
检查 sglang ServerArgs 导入... ok
✓ 通过
检查 sglang OpenAI 兼容入口导入... <frozen importlib._bootstrap_external>:1241: FutureWarning: The cuda.cudart module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.runtime module instead.
<frozen importlib._bootstrap_external>:1241: FutureWarning: The cuda.nvrtc module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.nvrtc module instead.
ok
✓ 通过
检查 sglang 平台探测... sglang=0.5.9, torch=2.9.1+cu128, cuda=12.8
✓ 通过

--- 阶段三：完整推理测试 ---
  请使用项目根目录的 cuda_test.sh 进行端到端推理验证：
  bash cuda_test.sh --vllm-model <model_path>

=== 所有测试通过 ===

类型	指派人员	状态
审查	cherrycao	进行中 (0/1人)
测试	cherrycao	进行中 (0/1人)

OpenCloudOS/AI-Agent-Container .gitee-modal { width: 500px !important; }

搜索帮助

OpenCloudOS/AI-Agent-Container