diff --git a/dockerfiles/Dockerfile.gpu b/dockerfiles/Dockerfile.gpu index 5ce8b05b19..4a4240cd76 100644 --- a/dockerfiles/Dockerfile.gpu +++ b/dockerfiles/Dockerfile.gpu @@ -1,6 +1,6 @@ FROM ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/fastdeploy-cuda-12.6:tag-base -ARG PADDLE_VERSION=3.3.0 -ARG FD_VERSION=2.4.0 +ARG PADDLE_VERSION=3.3.1 +ARG FD_VERSION=2.5.0 ENV DEBIAN_FRONTEND=noninteractive @@ -16,8 +16,8 @@ RUN python -m pip uninstall paddlepaddle-gpu fastdeploy-gpu -y RUN python -m pip install --no-cache-dir paddlepaddle-gpu==${PADDLE_VERSION} -i https://www.paddlepaddle.org.cn/packages/stable/cu126/ # build and install FastDeploy -RUN python -m pip install --no-cache-dir fastdeploy-gpu==${FD_VERSION} -i https://www.paddlepaddle.org.cn/packages/stable/fastdeploy-gpu-80_90/ --extra-index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple +RUN python -m pip install --no-cache-dir fastdeploy-gpu==${FD_VERSION} -i https://www.paddlepaddle.org.cn/packages/stable/cu126/ --extra-index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple ENV http_proxy="" ENV https_proxy="" -ENV no_proxy="" +ENV no_proxy="" \ No newline at end of file diff --git a/dockerfiles/Dockerfile.xpu b/dockerfiles/Dockerfile.xpu index 74e7bf3e44..3b98165dd0 100644 --- a/dockerfiles/Dockerfile.xpu +++ b/dockerfiles/Dockerfile.xpu @@ -1,11 +1,7 @@ -FROM ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/paddlenlp:llm-base-gcc12.3-xpu-xft20250402-v1.1 -ARG PADDLE_VERSION=3.1.0 -ARG FD_VERSION=2.0.0 +FROM ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/paddleqa:xpu-ubuntu2204-x86_64-gcc123-py310 WORKDIR /workspace - -ENV http_proxy=http://agent.baidu.com:8891 -ENV https_proxy=http://agent.baidu.com:8891 +COPY requirements.txt . RUN echo "\ deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy main restricted universe multiverse \n\ @@ -15,18 +11,33 @@ deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-backports main restricted RUN apt-get update && apt-get install -y libibverbs-dev librdmacm-dev cmake pybind11-dev # uninstall existing package -RUN python -m pip uninstall paddlepaddle-gpu paddlepaddle-xpu -y -# install paddlepaddle-xpu -RUN python -m pip install --no-cache-dir --progress-bar off paddlepaddle-xpu==${PADDLE_VERSION} -i https://www.paddlepaddle.org.cn/packages/stable/xpu-p800/ +RUN python -m pip uninstall paddlepaddle-gpu paddlepaddle-xpu fastdeploy-xpu -y +RUN python -m pip uninstall -y Pillow && rm -rf /usr/local/lib/python3.10/dist-packages/Pillow* && rm -rf /usr/local/lib/python3.10/dist-packages/pillow* && python -m pip install Pillow==11.3.0 -RUN python -m pip install --no-cache-dir fastdeploy-xpu==${FD_VERSION} -i https://www.paddlepaddle.org.cn/packages/stable/fastdeploy-xpu-p800/ --extra-index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple +# install paddlepaddle-xpu +ARG PADDLE_VERSION=3.3.1 + +RUN if [ "$PADDLE_VERSION" = "nightly" ]; then \ + python -m pip install --no-cache-dir --progress-bar off paddlepaddle-xpu -i https://www.paddlepaddle.org.cn/packages/nightly/xpu-p800/; \ + else \ + python -m pip install --no-cache-dir --progress-bar off paddlepaddle-xpu==${PADDLE_VERSION} -i https://www.paddlepaddle.org.cn/packages/stable/xpu-p800/; \ + fi + +# install fastdeploy-xpu +ARG INSTALL_REQUIREMENTS=true +ARG INSTALL_FASTDEPLOY=true +ARG FASTDEPLOY_VERSION=2.5.0 + +RUN if [ "$INSTALL_FASTDEPLOY" = "true" ]; then \ + python -m pip install --no-cache-dir fastdeploy-xpu==${FASTDEPLOY_VERSION} -i https://www.paddlepaddle.org.cn/packages/stable/fastdeploy-xpu-p800/ --extra-index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple; \ + fi + +RUN if [ "$INSTALL_REQUIREMENTS" = "true" ]; then \ + python -m pip install -r requirements.txt; \ + fi RUN mkdir -p /workspace/deps && cd /workspace/deps && \ wget https://klx-sdk-release-public.su.bcebos.com/xre/kl3-release/5.0.21.21/xre-Linux-x86_64-5.0.21.21.tar.gz && \ tar -zxf xre-Linux-x86_64-5.0.21.21.tar.gz && mv xre-Linux-x86_64-5.0.21.21 xre -ENV PATH=/workspace/deps/xre/bin:$PATH - -ENV http_proxy="" -ENV https_proxy="" -ENV no_proxy="" +ENV PATH=/workspace/deps/xre/bin:$PATH \ No newline at end of file diff --git a/docs/get_started/installation/nvidia_gpu.md b/docs/get_started/installation/nvidia_gpu.md index 11cb696922..5a1f1ae215 100644 --- a/docs/get_started/installation/nvidia_gpu.md +++ b/docs/get_started/installation/nvidia_gpu.md @@ -12,10 +12,13 @@ The following installation methods are available when your environment meets the ## 1. Pre-built Docker Installation (Recommended) -**Notice**: The pre-built image supports SM 80/86/89/90 architecture GPUs (e.g. A800/H800/L20/L40/4090). +**Notice**: The pre-built image supports SM 80/86/89/90 architecture GPUs (e.g. A800/H800/L20/L40/4090), and requires Python 3.10. ```shell +# CUDA 12.6 docker pull ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/fastdeploy-cuda-12.6:2.5.0 +# CUDA 12.9 +docker pull ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/fastdeploy-cuda-12.9:2.5.0 ``` ## 2. Pre-built Pip Installation @@ -38,7 +41,7 @@ python -m pip install --pre paddlepaddle-gpu -i https://www.paddlepaddle.org.cn/ Then install fastdeploy. **Do not install from PyPI**. Use the following methods instead (supports SM80/86/89/90 GPU architectures). **Note**: Stable FastDeploy release pairs with stable PaddlePaddle; Nightly Build FastDeploy pairs with Nightly Build PaddlePaddle. The `--extra-index-url` is only used for downloading fastdeploy-gpu's dependencies; fastdeploy-gpu itself must be installed from the Paddle source specified by `-i`. -``` +```shell # Install stable release FastDeploy # CUDA 12.6 python -m pip install fastdeploy-gpu==2.5.0 -i https://www.paddlepaddle.org.cn/packages/stable/cu126/ --extra-index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple @@ -54,7 +57,7 @@ python -m pip install fastdeploy-gpu -i https://www.paddlepaddle.org.cn/packages ## 3. Build from Source Using Docker -- Note: ```dockerfiles/Dockerfile.gpu``` by default supports SM 80/90 architectures. To support other architectures, modify ```bash build.sh 1 python false [80,90]``` in the Dockerfile. It's recommended to specify no more than 2 architectures. +> Note: `dockerfiles/Dockerfile.gpu` currently supports CUDA 12.6 only, targeting SM 80/86/89/90 architectures, and requires Python 3.10. To support other architectures, modify ```bash build.sh 1 python false [80,90]``` in the Dockerfile. It's recommended to specify no more than 2 architectures. ```shell git clone https://github.com/PaddlePaddle/FastDeploy @@ -69,8 +72,6 @@ First install paddlepaddle-gpu. For detailed instructions, refer to [PaddlePaddl ```shell python -m pip install paddlepaddle-gpu==3.3.1 -i https://www.paddlepaddle.org.cn/packages/stable/cu126/ ``` - -Then clone the source code and build: ```shell git clone https://github.com/PaddlePaddle/FastDeploy cd FastDeploy @@ -85,8 +86,7 @@ The built packages will be in the ```FastDeploy/dist``` directory. ## 5. Precompiled Operator Wheel Packages -FastDeploy provides precompiled GPU operator wheel packages for quick setup without building the entire source code. -This method currently supports **SM80/90 architecture (e.g., A100/H100)** and **CUDA 12.6** environments only. +FastDeploy provides precompiled GPU operator wheel packages for quick setup without building the entire source code. This method currently supports **SM80/90 architecture (e.g., A100/H100)** **CUDA 12.6** and **Python 3.10** environments only. > By default, `build.sh` compiles all custom operators from source.To use the precompiled package, enable it with the `FD_USE_PRECOMPILED` parameter. > If the precompiled package cannot be downloaded or does not match the current environment, the system will automatically fall back to `4. Build Wheel from Source`. @@ -115,7 +115,7 @@ cd FastDeploy bash build.sh 1 python false [90] 1 # Use precompiled wheel from a specific commit -bash build.sh 1 python false [90] 1 8a9e7b53af4a98583cab65e4b44e3265a93e56d2 +bash build.sh 1 python false [90] 1 d693d4be1448d414097882386fdc24c8bec2a63a ``` The downloaded wheel packages will be stored in the `FastDeploy/pre_wheel` directory. @@ -124,9 +124,9 @@ After the build completes, the operator binaries can be found in `FastDeploy/fas > **Notes:** > > - This mode prioritizes downloading precompiled GPU operator wheels to reduce build time. -> - Currently supports **GPU, SM80/90, CUDA 12.6** only. +> - Currently supports **GPU, SM80/90, CUDA 12.6, Python3.10** only. > - For custom architectures or modified operator logic, please use **source compilation (Section 4)**. -> - You can check whether the precompiled wheel for a specific commit has been successfully built on the [FastDeploy CI Build Status Page](https://github.com/PaddlePaddle/FastDeploy/actions/workflows/ci_image_update.yml). +> - You can check whether the precompiled wheel for a specific commit has been successfully built on the [FastDeploy CI Build Status Page](https://github.com/PaddlePaddle/FastDeploy/actions/workflows/ce_job.yml). ## Environment Verification diff --git a/docs/zh/get_started/installation/nvidia_gpu.md b/docs/zh/get_started/installation/nvidia_gpu.md index 004216c613..dd266b6c7e 100644 --- a/docs/zh/get_started/installation/nvidia_gpu.md +++ b/docs/zh/get_started/installation/nvidia_gpu.md @@ -14,10 +14,13 @@ ## 1. 预编译Docker安装(推荐) -**注意**： 预编译镜像支持 80/86/89/90 架构的GPU硬件 (如 A800/H800/L20/L40/4090)。 +**注意**： 预编译镜像支持 80/86/89/90 架构的GPU硬件 (如 A800/H800/L20/L40/4090) 且仅支持 Python 3.10。 ``` shell +# CUDA 12.6 docker pull ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/fastdeploy-cuda-12.6:2.5.0 +# CUDA 12.9 +docker pull ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/fastdeploy-cuda-12.9:2.5.0 ``` ## 2. 预编译Pip安装 @@ -41,7 +44,7 @@ python -m pip install --pre paddlepaddle-gpu -i https://www.paddlepaddle.org.cn/ 再安装 fastdeploy，**注意不要通过pypi源安装**，需要通过如下方式安装(目前支持80/86/89/90四个架构GPU） **注意**: 稳定版本的FastDeploy搭配稳定版本的PaddlePaddle; 而Nightly Build的FastDeploy则对应Nightly Build的PaddlePaddle。其中 `--extra-index-url` 仅用于安装 fastdeploy-gpu 所需的依赖包，fastdeploy-gpu 本身必须从 `-i` 指定的 Paddle 源安装。 -``` +```shell # 安装稳定版本FastDeploy # CUDA 12.6 python -m pip install fastdeploy-gpu==2.5.0 -i https://www.paddlepaddle.org.cn/packages/stable/cu126/ --extra-index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple @@ -57,7 +60,7 @@ python -m pip install fastdeploy-gpu -i https://www.paddlepaddle.org.cn/packages ## 3. 镜像自行构建 -> 注意 ```dockerfiles/Dockerfile.gpu``` 默认编译的架构支持SM 80/90，如若需要支持其它架构，需自行修改Dockerfile中的 ```bash build.sh 1 python false [80,90]```，建议不超过2个架构。 +> 注意 ```dockerfiles/Dockerfile.gpu``` 默认编译产物仅支持 SM 80/86/89/90 架构，基于 CUDA 12.6 环境构建，且仅支持 Python 3.10，如若需要支持其它架构，需自行修改Dockerfile中的 ```bash build.sh 1 python false [80,90]```，建议不超过2个架构。 ``` git clone https://github.com/PaddlePaddle/FastDeploy @@ -91,7 +94,7 @@ bash build.sh 1 python false [80,90] ## 5. 算子预编译 Wheel 包 -FastDeploy 提供了 GPU 算子预编译版 Wheel 包，可在无需完整源码编译的情况下快速构建。该方式当前仅支持 **SM80/90 架构（A100/H100等）** 和 **CUDA 12.6** 环境。 +FastDeploy 提供了 GPU 算子预编译版 Wheel 包，可在无需完整源码编译的情况下快速构建。该方式当前仅支持 **SM80/90 架构（A100/H100等）** **CUDA 12.6** 和 **Python 3.10** 环境。 >默认情况下，`build.sh` 会从源码编译；若希望使用预编译包，可使用`FD_USE_PRECOMPILED` 参数； >若预编译包下载失败或与环境不匹配，系统会自动回退至 `4. wheel 包源码编译` 模式。 @@ -119,7 +122,7 @@ cd FastDeploy bash build.sh 1 python false [90] 1 # 从指定 commitID 获取对应预编译算子 -bash build.sh 1 python false [90] 1 8a9e7b53af4a98583cab65e4b44e3265a93e56d2 +bash build.sh 1 python false [90] 1 d693d4be1448d414097882386fdc24c8bec2a63a ``` 下载的 whl 包在 `FastDeploy/pre_wheel`目录下。 @@ -128,7 +131,7 @@ bash build.sh 1 python false [90] 1 8a9e7b53af4a98583cab65e4b44e3265a93e56d2 > **说明：** > - 该模式会优先下载预编译的 GPU 算子 whl 包，减少编译时间； -> - 目前仅支持 **GPU， SM80/90 架构， CUDA 12.6**； +> - 目前仅支持 **GPU， SM80/90 架构， CUDA 12.6， Python3.10**； > - 若希望自定义架构或修改算子逻辑，请使用 **源码编译方式（第4节）**。 > - 您可以在 FastDeploy CI 构建状态页面查看对应 commit 的预编译 whl 是否已构建成功。