diff --git a/docs/compile/prebuilt_libraries.md b/docs/compile/prebuilt_libraries.md index 402c5b0a9..6cec3721a 100644 --- a/docs/compile/prebuilt_libraries.md +++ b/docs/compile/prebuilt_libraries.md @@ -1,26 +1,44 @@ -# FastDeploy 预编译部署库 +# FastDeploy 预编编译Python Wheel包 -FastDeploy提供了在Windows/Linux/Mac上的预先编译C++部署库，开发者可以直接下载使用，也可以自行编译代码。 +FastDeploy提供了在Windows/Linux/Mac上的预先编译CPP部署库，开发者可以直接下载后使用，也可以自行编译代码。 ## 环境依赖 -- cmake >= 3.12 -- g++/gcc >= 8.2 - cuda >= 11.2 - cudnn >= 8.0 +- g++ >= 5.4(推荐8.2) ## 下载地址 -| 系统 | 架构 | 硬件 | 下载链接 | -| :--- | :--- | :--- | :------- | -| Linux| x64 | CPU | [fastdeploy-linux-x64-0.2.0.tgz]() | -| Linux | x64 | CPU/GPU | [fastdeploy-linux-x64-gpu-0.2.0.tgz]() | -| Windows | x64 | CPU | [fastdeploy-win-x64-0.2.0.zip]() | -| Windows | x64 | CPU/GPU | [fastdeploy-win-x64-gpu-0.2.0.zip]() | -| Mac | x86_64 | CPU | [fastdeploy-osx-x86_64-0.2.0.tgz]() | -| Mac | arm64 | CPU | [fastdeploy-osx-arm64-0.2.0.tgz]() | +### Linux x64平台 + +| 部署库下载地址 | 硬件 | 说明 | +| :------------- | :--- | :--- | +| [fastdeploy-linux-x64-0.2.0](https://bj.bcebos.com/paddlehub/fastdeploy/cpp/fastdeploy-linux-x64-0.2.0.tgz) | CPU | g++ 8.2编译产出 | +| [fastdeploy-linux-x64-gpu-0.2.0](https://bj.bcebos.com/paddlehub/fastdeploy/cpp/fastdeploy-linux-x64-gpu-0.2.0.tgz) | CPU/GPU | g++ 8.2, cuda 11.2, cudnn 8.2编译产出 | + +### Windows 10 x64平台 + +| 部署库下载地址 | 硬件 | +| :------------- | :--- | +| [comming...] | CPU | +| [comming...] | CPU/GPU | + +### Linux aarch64平台 + +| 安装包 | 硬件 | +| :---- | :-- | +| [comming...] | CPU | +| [comming...] | Jetson | + +### Mac OSX平台 + +| 部署库下载地址 | 架构 |硬件 | +| :---- | :-- | :------ | +| [comming...] | x86 | CPU | +| [comming...] | arm64 | CPU | ## 其它文档 -- [预编译Python wheel包](./prebuilt_wheels.md) -- [FastDeploy C++部署](../../examples) +- [预编译Python安装包](./prebuilt_wheels.md) +- [视觉模型C++/Python部署示例](../../examples/vision/) diff --git a/docs/compile/prebuilt_wheels.md b/docs/compile/prebuilt_wheels.md index 2811d7a37..d846d2ea2 100644 --- a/docs/compile/prebuilt_wheels.md +++ b/docs/compile/prebuilt_wheels.md @@ -2,6 +2,15 @@ FastDeploy提供了在Windows/Linux/Mac上的预先编译Python Wheel包，开发者可以直接下载后安装，也可以自行编译代码。 +使用如下命令即可在Linux的Pythotn 3.8环境安装支持CPU部署的FastDeploy +``` +python -m pip install fastdeploy_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl +``` + +## 注意事项 +- 不要重复安装`fastdeploy-python`和`fastdeploy-gpu-python` +- 如已安装CPU版本的`fastdeploy-python`后，在安装GPU版本的`fastdeploy-gpu-python`，请先执行`pip uninstall fastdeploy-python`卸载已有版本 + ## 环境依赖 - cuda >= 11.2 @@ -9,16 +18,44 @@ FastDeploy提供了在Windows/Linux/Mac上的预先编译Python Wheel包，开 ## 下载地址 -| 系统 | 架构 | 硬件 | 下载链接 | -| :--- | :--- | :--- | :------- | -| Linux| x64 | CPU | [fastdeploy-linux-x64-0.2.0.tgz]() | -| Linux | x64 | CPU/GPU | [fastdeploy-linux-x64-gpu-0.2.0.tgz]() | -| Windows | x64 | CPU | [fastdeploy-win-x64-0.2.0.zip]() | -| Windows | x64 | CPU/GPU | [fastdeploy-win-x64-gpu-0.2.0.zip]() | -| Mac | x86_64 | CPU | [fastdeploy-osx-x86_64-0.2.0.tgz]() | -| Mac | arm64 | CPU | [fastdeploy-osx-arm64-0.2.0.tgz]() | +### Linux x64平台 + +| CPU 安装包 | 硬件 | Python版本 | +| :------------- | :--- | :--------- | +| [fastdeploy_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl) | CPU | 3.6 | +| [fastdeploy_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl) | CPU | 3.7 | +| [fastdeploy_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl) | CPU | 3.8 | +| [fastdeploy_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl) | CPU | 3.9 | + +| GPU 安装包 | 硬件 | Python版本 | +| [fastdeploy_gpu_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl) | CPU/GPU | 3.6 | +| [fastdeploy_gpu_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl) | CPU/GPU | 3.7 | +| [fastdeploy_gpu_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl) | CPU/GPU | 3.8 | +| [fastdeploy_gpu_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl) | CPU/GPU | 3.9 | + +### Windows 10 x64平台 + +| CPU 安装包 | 硬件 | Python版本 | +| :---- | :-- | :------ | +| [comming...] | CPU | 3.8 | +| [comming...] | CPU | 3.9 | + +### Linux aarch64平台 + +| 安装包 | 硬件 | Python版本 | +| :---- | :-- | :------ | +| [comming...] | CPU | 3.7 | +| [comming...] | CPU | 3.8 | +| [comming...] | CPU | 3.9 | + +### Mac OSX平台 + +| 架构 | 硬件 | 安装包 | Python版本 | +| :---- | :-- | :------ | :----- | +| x86_64 | CPU | [comming...] | 3.9 | +| arm64 | CPU | [comming...] | 3.9 | ## 其它文档 - [预编译C++部署库](./prebuilt_libraries.md) -- [FastDeploy C++部署](../../examples) +- [视觉模型C++/Python部署示例](../../examples/vision/) diff --git a/docs/quick_start/install.md b/docs/quick_start/install.md index 71df07d16..ce8bb591d 100644 --- a/docs/quick_start/install.md +++ b/docs/quick_start/install.md @@ -1,25 +1,11 @@ # FastDeploy安装 -目前已提供Linux、Windows上的Python Wheel安装包，开发者根据自身需求下载后使用如下命令安装即可，如 +针对开发者的开发需求，FastDeploy目前提供了Python和C++两种安装方式，可分别参考 -``` -python -m pip install fastdeploy_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl -``` +- [Python安装](../compile/prebuilt_wheels.md) +- [C++安装](../compile/prebuilt_libraries.md) -注意环境中不要重复安装`fastdeploy-python`和`fastdeploy-gpu-python`，在安装前，例如已安装cpu版本的`fastdeploy-python`后，如想重新安装gpu版本的`fastdeploy-gpu-python`，请先执行`pip uninstall fastdeploy-python`卸载已有版本 -## Linux +## 其它文档 -| CPU安装包 | Python环境 | 支持硬件 | -| ------------------------------------------------------------ | ---------- | -------- | -| [fastdeploy_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl) | 3.6 | CPU | -| [fastdeploy_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl) | 3.7 | CPU | -| [fastdeploy_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl) | 3.8 | CPU | -| [fastdeploy_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl) | 3.9 | CPU | - -| GPU安装包 | Python环境 | 支持硬件 | -| ------------------------------------------------------------ | ---------- | -------- | -| [fastdeploy_gpu_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp36-cp36m-manylinux1_x86_64.whl) | 3.6 | CPU/GPU(CUDA11.2 CUDNN8) | -| [fastdeploy_gpu_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp37-cp37m-manylinux1_x86_64.whl) | 3.7 | CPU/GPU(CUDA11.2 CUDNN8) | -| [fastdeploy_gpu_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp38-cp38-manylinux1_x86_64.whl) | 3.8 | CPU/GPU(CUDA11.2 CUDNN8) | -| [fastdeploy_gpu_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl](https://bj.bcebos.com/paddlehub/fastdeploy/wheels/fastdeploy_gpu_python-0.2.0-cp39-cp39-manylinux1_x86_64.whl) | 3.9 | CPU/GPU(CUDA11.2 CUDNN8) | +- [视觉模型部署示例](../../examples/vision/) : 提供目标检测、图像分类、人脸检测等大量视觉模型在CPU/GPU以及TensorRT部署示例