未验证 提交 70ec8aba 编写于 作者: T TeslaZhao 提交者: GitHub

Merge branch 'PaddlePaddle:develop' into develop

......@@ -19,7 +19,7 @@
建立新目录,复制对应 Dockerfile 内容到该目录下 Dockerfile 文件。执行
```shell
docker build -f tools/${DOCKERFILE} -t <image-name>:<images-tag> .
docker build -f tools/${DOCKERFILE} -t registry.baidubce.com/paddlepaddle/serving:<images-tag> .
```
......@@ -28,6 +28,9 @@
若需要基于源代码二次开发编译,请使用后缀为-devel的版本。
**在TAG列,0.8.0也可以替换成对应的版本号,例如0.5.0/0.4.1等,但需要注意的是,部分开发环境随着某个版本迭代才增加,因此并非所有环境都有对应的版本号可以使用。**
**开发镜像:**
开发镜像中安装多种开发工具,可用于调试和编译代码,适配2种 GCC 版本和多种 CUDA 环境,但镜像体积较大。
| 镜像选择 | 操作系统 | TAG | Dockerfile |
| :----------------------------------------------------------: | :-----: | :--------------------------: | :----------------------------------------------------------: |
......@@ -38,6 +41,19 @@
| GPU (cuda10.2-cudnn8-tensorRT7) development | Ubuntu16 | 0.8.0-cuda10.2-cudnn8-devel | [Dockerfile.cuda10.2-cudnn8.devel](../tools/Dockerfile.cuda10.2-cudnn8.devel) |
| GPU (cuda11.2-cudnn8-tensorRT8) development | Ubuntu16 | 0.8.0-cuda11.2-cudnn8-devel | [Dockerfile.cuda11.2-cudnn8.devel](../tools/Dockerfile.cuda11.2-cudnn8.devel) |
**运行镜像:**
运行镜像比开发镜像更加轻量化, 运行镜像提供了serving的whl和bin,但为了运行期更小的镜像体积,没有提供诸如cmake这样但开发工具。 如果您想了解有关信息,请检查文档[在Kubernetes上使用Paddle Serving](./Run_On_Kubernetes_CN.md)
| Env | Version | Docker images tag | OS | Gcc Version | Size |
|----------|---------|------------------------------|-----------|-------------|------|
| CPU | 0.8.0 | 0.8.0-runtime | Ubuntu 16 | 8.2.0 | 3.9 GB |
| Cuda10.1 | 0.8.0 | 0.8.0-cuda10.1-cudnn7-runtime | Ubuntu 16 | 8.2.0 | 10 GB |
| Cuda10.2 | 0.8.0 | 0.8.0-cuda10.2-cudnn8-runtime | Ubuntu 16 | 8.2.0 | 10.1 GB |
| Cuda11.2 | 0.8.0 | 0.8.0-cuda11.2-cudnn8-runtime| Ubuntu 16 | 8.2.0 | 14.2 GB |
**Java镜像:**
```
registry.baidubce.com/paddlepaddle/serving:0.8.0-cuda10.2-java
......@@ -58,31 +74,4 @@ registry.baidubce.com/paddlepaddle/serving:xpu-x86 # for x86 xpu user
相关CUDA工具包版本、驱动版本和GPU架构的关系请参阅 [nvidia-docker wiki](https://github.com/NVIDIA/nvidia-docker/wiki/CUDA)
# (附录)所有镜像列表
开发镜像:
| Env | Version | Docker images tag | OS | Gcc Version |
|----------|---------|------------------------------|-----------|-------------|
| CPU | >=0.5.0 | 0.8.0-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | 0.4.0-devel | CentOS 7 | 4.8.5 |
| Cuda10.1 | >=0.5.0 | 0.8.0-cuda10.1-cudnn7-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | 0.4.0-cuda10.1-cudnn7-devel | CentOS 7 | 4.8.5 |
| Cuda10.2+Cudnn7 | >=0.5.0 | 0.8.0-cuda10.2-cudnn7-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | Nan | Nan | Nan |
| Cuda10.2+Cudnn8 | >=0.5.0 | 0.8.0-cuda10.2-cudnn8-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | Nan | Nan | Nan |
| Cuda11.2 | >=0.5.0 | 0.8.0-cuda11.2-cudnn8-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | Nan | Nan | Nan |
运行镜像:
运行镜像比开发镜像更加轻量化, 运行镜像提供了serving的whl和bin,但为了运行期更小的镜像体积,没有提供诸如cmake这样但开发工具。 如果您想了解有关信息,请检查文档[在Kubernetes上使用Paddle Serving](./Run_On_Kubernetes_CN.md)
| Env | Version | Docker images tag | OS | Gcc Version | Size |
|----------|---------|------------------------------|-----------|-------------|------|
| CPU | 0.8.0 | 0.8.0-runtime | Ubuntu 16 | 8.2.0 | 3.9 GB |
| Cuda10.1 | 0.8.0 | 0.8.0-cuda10.1-cudnn7-runtime | Ubuntu 16 | 8.2.0 | 10 GB |
| Cuda10.2 | 0.8.0 | 0.8.0-cuda10.2-cudnn8-runtime | Ubuntu 16 | 8.2.0 | 10.1 GB |
| Cuda11.2 | 0.8.0 | 0.8.0-cuda11.2-cudnn8-runtime| Ubuntu 16 | 8.2.0 | 14.2 GB |
......@@ -19,7 +19,7 @@ You can get images in two ways:
Create a new folder and copy Dockerfile to this folder, and run the following command:
```shell
docker build -f tools/${DOCKERFILE} -t <image-name>:<images-tag> .
docker build -f tools/${DOCKERFILE} -t registry.baidubce.com/paddlepaddle/serving:<images-tag> .
```
......@@ -33,6 +33,10 @@ If you want to customize your Serving based on source code, use the version with
If you need to develop and compile based on the source code, please use the version with the suffix -devel.
**In the TAG column, 0.8.0 can also be replaced with the corresponding version number, such as 0.5.0/0.4.1, etc., but it should be noted that some development environments only increase with a certain version iteration, so not all environments All have the corresponding version number can be used.**
**Development Docker Images:**
A variety of development tools are installed in the development image, which can be used to debug and compile code, and adapt to two GCC versions and multiple CUDA environments, but the image size is large.
| Description | OS | TAG | Dockerfile |
| :----------------------------------------------------------: | :-----: | :--------------------------: | :----------------------------------------------------------: |
| CPU development | Ubuntu16 | 0.8.0-devel | [Dockerfile.devel](../tools/Dockerfile.devel) |
......@@ -42,12 +46,24 @@ If you need to develop and compile based on the source code, please use the vers
| GPU (cuda10.2-cudnn8-tensorRT7) development | Ubuntu16 | 0.8.0-cuda10.2-cudnn8-devel | [Dockerfile.cuda10.2-cudnn8.devel](../tools/Dockerfile.cuda10.2-cudnn8.devel) |
| GPU (cuda11.2-cudnn8-tensorRT8) development | Ubuntu16 | 0.8.0-cuda11.2-cudnn8-devel | [Dockerfile.cuda11.2-cudnn8.devel](../tools/Dockerfile.cuda11.2-cudnn8.devel) |
**Java Client:**
**Runtime Docker Images:**
Runtime Docker Images is lighter than Develop Images, and Running Images are made up with serving whl and bin, but without develop tools like cmake because of lower image size. If you want to know about it, plese check the document [Paddle Serving on Kubernetes](./Run_On_Kubernetes_CN.md).
| Env | Version | Docker images tag | OS | Gcc Version | Size |
|----------|---------|------------------------------|-----------|-------------|------|
| CPU | 0.8.0 | 0.8.0-runtime | Ubuntu 16 | 8.2.0 | 3.9 GB |
| Cuda10.1 | 0.8.0 | 0.8.0-cuda10.1-cudnn7-runtime | Ubuntu 16 | 8.2.0 | 10 GB |
| Cuda10.2 | 0.8.0 | 0.8.0-cuda10.2-cudnn8-runtime | Ubuntu 16 | 8.2.0 | 10.1 GB |
| Cuda11.2 | 0.8.0 | 0.8.0-cuda11.2-cudnn8-runtime| Ubuntu 16 | 8.2.0 | 14.2 GB |
**Java SDK Docker Image:**
```
registry.baidubce.com/paddlepaddle/serving:0.8.0-cuda10.2-java
```
**XPU:**
**XPU Docker Images:**
```
registry.baidubce.com/paddlepaddle/serving:xpu-arm # for arm xpu user
registry.baidubce.com/paddlepaddle/serving:xpu-x86 # for x86 xpu user
......@@ -61,31 +77,3 @@ The machine running the CUDA container **only requires the NVIDIA driver**, the
For the relationship between CUDA toolkit version, Driver version and GPU architecture, please refer to [nvidia-docker wiki](https://github.com/NVIDIA/nvidia-docker/wiki/CUDA).
# (Attachment) The List of All the Docker images
Develop Images:
| Env | Version | Docker images tag | OS | Gcc Version |
|----------|---------|------------------------------|-----------|-------------|
| CPU | >=0.5.0 | 0.8.0-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | 0.4.0-devel | CentOS 7 | 4.8.5 |
| Cuda10.1 | >=0.5.0 | 0.8.0-cuda10.1-cudnn7-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | 0.4.0-cuda10.1-cudnn7-devel | CentOS 7 | 4.8.5 |
| Cuda10.2+Cudnn7 | >=0.5.0 | 0.8.0-cuda10.2-cudnn7-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | Nan | Nan | Nan |
| Cuda10.2+Cudnn8 | >=0.5.0 | 0.8.0-cuda10.2-cudnn8-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | Nan | Nan | Nan |
| Cuda11.2 | >=0.5.0 | 0.8.0-cuda11.2-cudnn8-devel | Ubuntu 16 | 8.2.0 |
| | <=0.4.0 | Nan | Nan | Nan |
Running Images:
Running Images is lighter than Develop Images, and Running Images are made up with serving whl and bin, but without develop tools like cmake because of lower image size. If you want to know about it, plese check the document [Paddle Serving on Kubernetes](./Run_On_Kubernetes_CN.md).
| Env | Version | Docker images tag | OS | Gcc Version | Size |
|----------|---------|------------------------------|-----------|-------------|------|
| CPU | 0.8.0 | 0.8.0-runtime | Ubuntu 16 | 8.2.0 | 3.9 GB |
| Cuda10.1 | 0.8.0 | 0.8.0-cuda10.1-cudnn7-runtime | Ubuntu 16 | 8.2.0 | 10 GB |
| Cuda10.2 | 0.8.0 | 0.8.0-cuda10.2-cudnn8-runtime | Ubuntu 16 | 8.2.0 | 10.1 GB |
| Cuda11.2 | 0.8.0 | 0.8.0-cuda11.2-cudnn8-runtime| Ubuntu 16 | 8.2.0 | 14.2 GB |
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册