Files
FastDeploy/serving/docs/zh_CN/compile.md
DefTruth 434b48dda5 [Serving] Support FastDeploy XPU Triton Server (#1994)
* [patchelf] fix patchelf error for inference xpu

* [serving] add xpu dockerfile and support fd server

* [serving] add xpu dockerfile and support fd server

* [Serving] support XPU + Tritron

* [Serving] support XPU + Tritron

* [Dockerfile] update xpu tritron docker file -> paddle 0.0.0

* [Dockerfile] update xpu tritron docker file -> paddle 0.0.0

* [Dockerfile] update xpu tritron docker file -> paddle 0.0.0

* [Dockerfile] add comments for xpu tritron dockerfile

* [Doruntime] fix xpu infer error

* [Doruntime] fix xpu infer error

* [XPU] update xpu dockerfile

* add xpu triton server docs

* add xpu triton server docs

* add xpu triton server docs

* add xpu triton server docs

* update xpu triton server docs

* update xpu triton server docs

* update xpu triton server docs

* update xpu triton server docs

* update xpu triton server docs

* update xpu triton server docs

* update xpu triton server docs

* update xpu triton server docs
2023-05-29 14:38:25 +08:00

2.6 KiB
Raw Blame History

中文 English

服务化部署编译

制作服务化部署镜像

制作GPU镜像

FastDploy发布的GPU镜像基于Triton Inference Server的21.10版本进行制作如果有其他CUDA版本需求可以参照NVIDIA 官网中展示的版本信息修改Dockerfile和scripts中的脚本.

# 进入serving目录执行脚本编译fastdeploy和服务化的backend
cd serving
bash scripts/build.sh

# 退出到FastDeploy主目录制作镜像
# x.y.z为FastDeploy版本号可根据情况自己确定。比如: 1.0.6
cd ../
docker build -t paddlepaddle/fastdeploy:x.y.z-gpu-cuda11.4-trt8.5-21.10 -f serving/Dockerfile .

目前默认 TensorRT 版本为 8.5.2.2,如果需要切换 TensorRT 版本,则可执行以下编译命令:

cd serving
bash scripts/build.sh -tv 8.4.1.5

cd ../
docker build -t paddlepaddle/fastdeploy:x.y.z-gpu-cuda11.4-trt8.5-21.10 -f serving/Dockerfile_CUDA_11_4_TRT_8_4 .

比如在ubuntu 20.04cuda11.2环境下制作基于FastDeploy v1.0.6的GPU镜像

# 进入serving目录执行脚本编译fastdeploy和服务化的backend
cd serving
bash scripts/build_fd_cuda_11_2.sh

# 退出到FastDeploy主目录制作镜像
cd ../
docker build -t paddlepaddle/fastdeploy:1.0.6-gpu-cuda11.2-trt8.4-21.10 -f serving/Dockerfile_CUDA_11_2 .

制作CPU镜像

# 进入serving目录执行脚本编译fastdeploy和服务化的backend
cd serving
bash scripts/build.sh OFF

# 退出到FastDeploy主目录制作镜像
# x.y.z为FastDeploy版本号可根据情况自己确定。比如: 1.0.6
cd ../
docker build -t paddlepaddle/fastdeploy:x.y.z-cpu-only-21.10 -f serving/Dockerfile_cpu .

制作IPU镜像

# 进入serving目录执行脚本编译fastdeploy和服务化的backend
cd serving
bash scripts/build_fd_ipu.sh

# 退出到FastDeploy主目录制作镜像
# x.y.z为FastDeploy版本号可根据情况自己确定。比如: 1.0.6
cd ../
docker build -t paddlepaddle/fastdeploy:x.y.z-ipu-only-21.10 -f serving/Dockerfile_ipu .

制作XPU镜像

# 进入serving目录执行脚本编译fastdeploy和服务化的backend
cd serving
bash scripts/build_fd_xpu.sh

# 退出到FastDeploy主目录制作镜像
# x.y.z为FastDeploy版本号可根据情况自己确定。比如: 1.0.6
cd ../
docker build -t paddlepaddle/fastdeploy:x.y.z-xpu-21.10 -f serving/Dockerfile_xpu .

非镜像方式编译