[Doc] Add doc for vdl serving (#1110)

* add doc for vdl serving * add doc for vdl serving * add doc for vdl serving * fix link * fix link * fix gif size * fix gif size * add english version * fix links * fix links * update format * update docs * update docs * update docs * update docs * update docs * update docs --------- Co-authored-by: heliqi <[email protected]>
PaddlePaddle · Jan 30, 2023 · 595ca69 · 595ca69
1 parent 8c651f9
commit 595ca69
Show file tree

Hide file tree

Showing 19 changed files with 689 additions and 29 deletions.
diff --git a/examples/text/ernie-3.0/serving/README.md b/examples/text/ernie-3.0/serving/README.md
@@ -175,3 +175,18 @@ entity: 华夏   label: LOC   pos: [14, 15]
 
 ## Configuration Modification
 The current classification task (ernie_seqcls_model/config.pbtxt) is by default configured to run the OpenVINO engine on CPU; the sequence labelling task is by default configured to run the Paddle engine on GPU. If you want to run on CPU/GPU or other inference engines, you should modify the configuration. please refer to the [configuration document.](../../../../serving/docs/EN/model_configuration-en.md)
+
+## Use VisualDL for serving deployment visualization
+
+You can use VisualDL for [serving deployment visualization](../../../../serving/docs/EN/vdl_management-en.md) , the above model preparation, deployment, configuration modification and client request operations can all be performed based on VisualDL.
+
+The serving deployment of ERNIE 3.0 by VisualDL only needs the following three steps:
+```text
+1. Load the model repository: ./text/ernie-3.0/serving/models
+2. Download the model resource file: click the ernie_seqcls_model model, click the version number 1 to add the pre-training model, and select the text classification model ernie_3.0_ernie_seqcls_model to download. Click the ernie_tokencls_model model, click the version number 1 to add the pre-training model, and select the text classification model ernie_tokencls_model to download.
+3. Start the service: Click the "launch server" button and input the launch parameters.
+```
+
+<p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211708353-507d6038-b754-4520-884b-1156703a44c6.gif" width="100%"/>
+</p>
diff --git a/examples/text/ernie-3.0/serving/README_CN.md b/examples/text/ernie-3.0/serving/README_CN.md
@@ -175,3 +175,17 @@ entity: 华夏   label: LOC   pos: [14, 15]
 ## 配置修改
 
 当前分类任务(ernie_seqcls_model/config.pbtxt)默认配置在CPU上运行OpenVINO引擎; 序列标注任务默认配置在GPU上运行Paddle引擎。如果要在CPU/GPU或其他推理引擎上运行, 需要修改配置，详情请参考[配置文档](../../../../serving/docs/zh_CN/model_configuration.md)
+
+## 使用VisualDL进行可视化部署
+
+可以使用VisualDL进行[Serving可视化部署](../../../../serving/docs/zh_CN/vdl_management.md)，上述启动服务、配置修改以及客户端请求的操作都可以基于VisualDL进行。
+
+通过VisualDL的可视化界面对ERNIE 3.0进行服务化部署只需要如下三步：
+```text
+1. 载入模型库：./text/ernie-3.0/serving/models
+2. 下载模型资源文件：点击ernie_seqcls_model模型，点击版本号1添加预训练模型，选择文本分类模型ernie_3.0_ernie_seqcls_model进行下载。点击ernie_tokencls_model模型，点击版本号1添加预训练模型，选择文本分类模型ernie_tokencls_model进行下载。
+3. 启动服务：点击启动服务按钮，输入启动参数。
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211708353-507d6038-b754-4520-884b-1156703a44c6.gif" width="100%"/>
+</p>
diff --git a/examples/text/uie/serving/README.md b/examples/text/uie/serving/README.md
@@ -1,4 +1,4 @@
-English | [简体中文](README_CN.md) 
+English | [简体中文](README_CN.md)
 
 # Example of UIE Serving Deployment
 
@@ -144,3 +144,18 @@ results:
 ## Configuration Modification
 
 The current configuration is by default to run the paddle engine on CPU. If you want to run on CPU/GPU or other inference engines, modifying the configuration is needed.Please refer to [Configuration Document](../../../../serving/docs/EN/model_configuration-en.md).
+
+## Use VisualDL for serving deployment visualization
+
+You can use VisualDL for [serving deployment visualization](../../../../serving/docs/EN/vdl_management-en.md) , the above model preparation, deployment, configuration modification and client request operations can all be performed based on VisualDL.
+
+The serving deployment of UIE by VisualDL only needs the following three steps:
+```text
+1. Load the model repository: ./text/uie/serving/models
+2. Download the model resource file: click the uie model, click the version number 1 to add the pre-training model, and select the text information extraction model uie-base to download.
+3. Start the service: Click the "launch server" button and input the launch parameters.
+```
+
+<p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211708353-507d6038-b754-4520-884b-1156703a44c6.gif" width="100%"/>
+</p>
diff --git a/examples/text/uie/serving/README_CN.md b/examples/text/uie/serving/README_CN.md
@@ -143,3 +143,17 @@ results:
 ## 配置修改
 
 当前默认配置在GPU上运行Paddle引擎,如果要在CPU/GPU或其他推理引擎上运行, 需要修改配置，详情请参考[配置文档](../../../../serving/docs/zh_CN/model_configuration.md)
+
+## 使用VisualDL进行可视化部署
+
+可以使用VisualDL进行[Serving可视化部署](../../../../serving/docs/zh_CN/vdl_management.md)，上述启动服务、配置修改以及客户端请求的操作都可以基于VisualDL进行。
+
+通过VisualDL的可视化界面对UIE进行服务化部署只需要如下三步：
+```text
+1. 载入模型库：./text/uie/serving/models
+2. 下载模型资源文件：点击uie模型，点击版本号1添加预训练模型，选择信息抽取模型uie-base进行下载
+3. 启动服务：点击启动服务按钮，输入启动参数。
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211709329-3261758e-69af-4efd-9711-693f5f031131.gif" width="100%"/>
+</p>
diff --git a/examples/vision/classification/paddleclas/serving/README.md b/examples/vision/classification/paddleclas/serving/README.md
@@ -1,7 +1,7 @@
 English | [简体中文](README_CN.md)
 # PaddleClas Service Deployment Example
 
-Before the service deployment, please confirm 
+Before the service deployment, please confirm
 
 - 1. Refer to [FastDeploy Service Deployment](../../../../../serving/README.md) for software and hardware environment requirements and image pull commands.
 
@@ -13,25 +13,25 @@ Before the service deployment, please confirm
 git clone https://github.com/PaddlePaddle/FastDeploy.git
 cd FastDeploy/examples/vision/classification/paddleclas/serving
 
-# Download ResNet50_vd model files and test images 
+# Download ResNet50_vd model files and test images
 wget https://bj.bcebos.com/paddlehub/fastdeploy/ResNet50_vd_infer.tgz
 tar -xvf ResNet50_vd_infer.tgz
 wget https://gitee.com/paddlepaddle/PaddleClas/raw/release/2.4/deploy/images/ImageNet/ILSVRC2012_val_00000010.jpeg
 
-# Put the configuration file into the preprocessing directory 
+# Put the configuration file into the preprocessing directory
 mv ResNet50_vd_infer/inference_cls.yaml models/preprocess/1/inference_cls.yaml
 
 # Place the model under models/runtime/1 and rename them to model.pdmodel和model.pdiparams
 mv ResNet50_vd_infer/inference.pdmodel models/runtime/1/model.pdmodel
 mv ResNet50_vd_infer/inference.pdiparams models/runtime/1/model.pdiparams
 
 # Pull the fastdeploy image (x.y.z represent the image version. Refer to the serving document to replace them with numbers)
-# GPU image 
+# GPU image
 docker pull registry.baidubce.com/paddlepaddle/fastdeploy:x.y.z-gpu-cuda11.4-trt8.4-21.10
-# CPU image 
+# CPU image
 docker pull registry.baidubce.com/paddlepaddle/fastdeploy:x.y.z-cpu-only-21.10
 
-# Run the container named fd_serving and mount it in the /serving directory of the container 
+# Run the container named fd_serving and mount it in the /serving directory of the container
 nvidia-docker run -it --net=host --name fd_serving -v `pwd`/:/serving registry.baidubce.com/paddlepaddle/fastdeploy:x.y.z-gpu-cuda11.4-trt8.4-21.10  bash
 
 # Start the service (The CUDA_VISIBLE_DEVICES  environment variable is not set, which entitles the scheduling authority of all GPU cards)
@@ -43,7 +43,7 @@ CUDA_VISIBLE_DEVICES=0 fastdeployserver --model-repository=/serving/models --bac
 
 >> If "Address already in use" appears when running fastdeployserver to start the service, use `--grpc-port` to specify the port number and change the request port number in the client demo.
 
->> Other startup parameters can be checked by fastdeployserver --help 
+>> Other startup parameters can be checked by fastdeployserver --help
 
 Successful service start brings the following output:
 ```
@@ -54,17 +54,17 @@ I0928 04:51:15.826578 206 http_server.cc:167] Started Metrics Service at 0.0.0.0
 ```
 
 
-## Client Request 
+## Client Request
 
 Execute the following command in the physical machine to send the grpc request and output the result
 ```
-# Download test images 
+# Download test images
 wget https://gitee.com/paddlepaddle/PaddleClas/raw/release/2.4/deploy/images/ImageNet/ILSVRC2012_val_00000010.jpeg
 
-# Install client dependencies 
+# Install client dependencies
 python3 -m pip install tritonclient\[all\]
 
-# Send the request 
+# Send the request
 python3 paddlecls_grpc_client.py
 ```
 
@@ -77,3 +77,17 @@ output_name: CLAS_RESULT
 ## Configuration Change
 
 The current default configuration runs the TensorRT engine on GPU. If you want to run it on CPU or other inference engines, please modify the configuration in `models/runtime/config.pbtxt`. Refer to [Configuration Document](../../../../../serving/docs/EN/model_configuration-en.md) for more information.
+
+## Use VisualDL for serving deployment visualization
+
+You can use VisualDL for [serving deployment visualization](../../../../serving/docs/EN/vdl_management-en.md) , the above model preparation, deployment, configuration modification and client request operations can all be performed based on VisualDL.
+
+The serving deployment of PaddleClas by VisualDL only needs the following three steps:
+```text
+1. Load the model repository: ./vision/classification/paddleclas/serving/models
+2. Download the model resource file: click the runtime model, click the version number 1 to add the pre-training model, and select the image classification model ResNet50_vd to download.
+3. Start the service: Click the "launch server" button and input the launch parameters.
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211708702-828d8ad8-4e85-457f-9c62-12f53fc81853.gif" width="100%"/>
+</p>
diff --git a/examples/vision/classification/paddleclas/serving/README_CN.md b/examples/vision/classification/paddleclas/serving/README_CN.md
@@ -77,3 +77,17 @@ output_name: CLAS_RESULT
 ## 配置修改
 
 当前默认配置在GPU上运行TensorRT引擎， 如果要在CPU或其他推理引擎上运行。 需要修改`models/runtime/config.pbtxt`中配置，详情请参考[配置文档](../../../../../serving/docs/zh_CN/model_configuration.md)
+
+## 使用VisualDL进行可视化部署
+
+可以使用VisualDL进行[Serving可视化部署](../../../../../serving/docs/zh_CN/vdl_management.md)，上述启动服务、配置修改以及客户端请求的操作都可以基于VisualDL进行。
+
+通过VisualDL的可视化界面对PaddleClas进行服务化部署只需要如下三步：
+```text
+1. 载入模型库：./vision/classification/paddleclas/serving/models
+2. 下载模型资源文件：点击runtime模型，点击版本号1添加预训练模型，选择图像分类模型ResNet50_vd进行下载。
+3. 启动服务：点击启动服务按钮，输入启动参数。
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211708702-828d8ad8-4e85-457f-9c62-12f53fc81853.gif" width="100%"/>
+</p>
diff --git a/examples/vision/detection/paddledetection/serving/README.md b/examples/vision/detection/paddledetection/serving/README.md
@@ -17,12 +17,12 @@ Confirm before the serving deployment
 git clone https://github.com/PaddlePaddle/FastDeploy.git
 cd FastDeploy/examples/vision/detection/paddledetection/serving
 
-# Download PPYOLOE model files and test images 
+# Download PPYOLOE model files and test images
 wget https://bj.bcebos.com/paddlehub/fastdeploy/ppyoloe_crn_l_300e_coco.tgz
 wget https://gitee.com/paddlepaddle/PaddleDetection/raw/release/2.4/demo/000000014439.jpg
 tar xvf ppyoloe_crn_l_300e_coco.tgz
 
-# Put the configuration file into the preprocessing directory 
+# Put the configuration file into the preprocessing directory
 mv ppyoloe_crn_l_300e_coco/infer_cfg.yml models/preprocess/1/
 
 # Place the model under models/runtime/1 and rename them to model.pdmodel and model.pdiparams
@@ -36,14 +36,14 @@ cp models/runtime/ppyoloe_runtime_config.pbtxt models/runtime/config.pbtxt
 
 # Attention: Given that the mask_rcnn model has one more output, we need to rename mask_config.pbtxt to config.pbtxt in the postprocess directory (models/postprocess)
 
-# Pull the FastDeploy image (x.y.z represent the image version. Users need to replace them with numbers) 
+# Pull the FastDeploy image (x.y.z represent the image version. Users need to replace them with numbers)
 # GPU image
 docker pull registry.baidubce.com/paddlepaddle/fastdeploy:x.y.z-gpu-cuda11.4-trt8.4-21.10
 # CPU image
 docker pull paddlepaddle/fastdeploy:z.y.z-cpu-only-21.10
 
 
-# Run the container named fd_serving and mount it in the /serving directory of the container 
+# Run the container named fd_serving and mount it in the /serving directory of the container
 nvidia-docker run -it --net=host --name fd_serving --shm-size="1g"  -v `pwd`/:/serving registry.baidubce.com/paddlepaddle/fastdeploy:x.y.z-gpu-cuda11.4-trt8.4-21.10  bash
 
 # Start Service (The CUDA_VISIBLE_DEVICES environment variable is not set, which entitles the scheduling authority of all GPU cards)
@@ -68,14 +68,14 @@ I0928 04:51:15.826578 206 http_server.cc:167] Started Metrics Service at 0.0.0.0
 ```
 
 
-## Client Request 
+## Client Request
 
 Execute the following command in the physical machine to send the grpc request and output the results
 ```
-# Download test images 
+# Download test images
 wget https://gitee.com/paddlepaddle/PaddleDetection/raw/release/2.4/demo/000000014439.jpg
 
-# Install client dependencies 
+# Install client dependencies
 python3 -m pip install tritonclient[all]
 
 # Send requests
@@ -93,3 +93,18 @@ output_name: DET_RESULT
 ## Configuration Change
 
 The current default configuration runs on GPU. If you want to run it on CPU or other inference engines, please modify the configuration in `models/runtime/config.pbtxt`. Refer to [Configuration Document](../../../../../serving/docs/EN/model_configuration-en.md) for more information.
+
+## Use VisualDL for serving deployment visualization
+
+You can use VisualDL for [serving deployment visualization](../../../../../serving/docs/EN/vdl_management-en.md) , the above model preparation, deployment, configuration modification and client request operations can all be performed based on VisualDL.
+
+The serving deployment of PaddleDetection by VisualDL only needs the following three steps:
+```text
+1. Load the model repository: ./vision/detection/paddledetection/serving/models
+2. Download the model resource file: click the preprocess model, click the version number 1 to add the pre-training model, and select the detection model ppyoloe_crn_l_300e_coco to download. click the runtime model, click the version number 1 to add the pre-training model, and select the detection model ppyoloe_crn_l_300e_coco to download.
+3. Set startup config file: click the "ensemble  configuration" button, choose configuration file ppyoloe_config.pbtxt, then click the "set as startup config" button. click the runtime model, choose configuration file ppyoloe_runtime_config.pbtxt, then click the "set as startup config" button.
+4. Start the service: Click the "launch server" button and input the launch parameters.
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211710983-2d1f1427-6738-409d-903b-2b4e4ab6cbfc.gif" width="100%"/>
+</p>
diff --git a/examples/vision/detection/paddledetection/serving/README_CN.md b/examples/vision/detection/paddledetection/serving/README_CN.md
@@ -93,3 +93,19 @@ output_name: DET_RESULT
 ## 配置修改
 
 当前默认配置在GPU上运行Paddle引擎， 如果要在CPU或其他推理引擎上运行。 需要修改`models/runtime/config.pbtxt`中配置，详情请参考[配置文档](../../../../../serving/docs/zh_CN/model_configuration.md)
+
+
+## 使用VisualDL进行可视化部署
+
+可以使用VisualDL进行[Serving可视化部署](../../../../../serving/docs/zh_CN/vdl_management.md)，上述启动服务、配置修改以及客户端请求的操作都可以基于VisualDL进行。
+
+通过VisualDL的可视化界面对PaddleDetection进行服务化部署只需要如下三步：
+```text
+1. 载入模型库：./vision/detection/paddledetection/serving/models
+2. 下载模型资源文件：点击preprocess模型，点击版本号1添加预训练模型，选择检测模型ppyoloe_crn_l_300e_coco进行下载，此时preprocess中将会有资源文件infer_cfg.yml。点击runtime模型，点击版本号1添加预训练模型，选择检测模型ppyoloe_crn_l_300e_coco进行下载，此时runtime中将会有资源文件model.pdmodel和model.pdiparams。
+3. 设置启动配置文件：点击ensemble配置按钮，选择配置文件ppyoloe_config.pbtxt，并设为启动配置文件。点击runtime模型，选择配置文件ppyoloe_runtime_config.pbtxt，并设为启动配置文件。
+4. 启动服务：点击启动服务按钮，输入启动参数。
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211710983-2d1f1427-6738-409d-903b-2b4e4ab6cbfc.gif" width="100%"/>
+</p>
diff --git a/examples/vision/detection/paddledetection/serving/ppdet_request.json b/examples/vision/detection/paddledetection/serving/ppdet_request.json
diff --git a/examples/vision/detection/yolov5/serving/README.md b/examples/vision/detection/yolov5/serving/README.md
@@ -56,3 +56,17 @@ output_name: detction_result
 
 
 The default is to run ONNXRuntime on CPU. If developers need to run it on GPU or other inference engines, please see the  [Configs File](../../../../../serving/docs/EN/model_configuration-en.md) to modify the configs in `models/runtime/config.pbtxt`.
+
+## Use VisualDL for serving deployment visualization
+
+You can use VisualDL for [serving deployment visualization](../../../../../serving/docs/EN/vdl_management-en.md) , the above model preparation, deployment, configuration modification and client request operations can all be performed based on VisualDL.
+
+The serving deployment of yolov5 by VisualDL only needs the following three steps:
+```text
+1. Load the model repository: ./vision/detection/yolov5/serving/models
+2. Download the model resource file: click the runtime model, click the version number 1 to add the pre-training model, and select the detection model yolov5s to download.
+3. Start the service: Click the "launch server" button and input the launch parameters.
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211709339-023fef22-3ffc-4b3d-bce5-ea4202bb9c61.gif" width="100%"/>
+</p>
diff --git a/examples/vision/detection/yolov5/serving/README_CN.md b/examples/vision/detection/yolov5/serving/README_CN.md
@@ -65,3 +65,18 @@ output_name: detction_result
 ## 配置修改
 
 当前默认配置在CPU上运行ONNXRuntime引擎， 如果要在GPU或其他推理引擎上运行。 需要修改`models/runtime/config.pbtxt`中配置，详情请参考[配置文档](../../../../../serving/docs/zh_CN/model_configuration.md)
+
+
+## 使用VisualDL进行可视化部署
+
+可以使用VisualDL进行[Serving可视化部署](../../../../../serving/docs/zh_CN/vdl_management.md)，上述启动服务、配置修改以及客户端请求的操作都可以基于VisualDL进行。
+
+通过VisualDL的可视化界面对yolov5进行服务化部署只需要如下三步：
+```text
+1. 载入模型库：./vision/detection/yolov5/serving/models
+2. 下载模型资源文件：点击runtime模型，点击版本号1添加预训练模型，选择检测模型yolov5s进行下载。
+3. 启动服务：点击启动服务按钮，输入启动参数。
+```
+ <p align="center">
+  <img src="https://user-images.githubusercontent.com/22424850/211709339-023fef22-3ffc-4b3d-bce5-ea4202bb9c61.gif" width="100%"/>
+</p>