Update docs

doc/Offical_Docs/3-2_QuickStart_Pipeline_OCR_CN.md

@@ -25,13 +25,18 @@ PaddleOCR 提供的 PP-OCR 系列模型覆盖轻量级服务端、轻量级移
 ```
 git clone https://github.com/PaddlePaddle/Serving
 ```
-按以下5个步骤操作即可实现 OCR 示例部署。
+通过6个步骤操作即可实现 OCR 示例部署。
+- 一.获取模型
+- 二.保存 Serving 部署的模型参数
+- 三.下载测试数据集（可选）
+- 四.修改 `config.yml` 配置（可选）
+- 五.代码与配置信息绑定
+- 六.启动服务与验证
 
-**一.获取模型**
+**一.获取模型与保存模型参数**
 
 本章节选用中英文超轻量模型 ch_PP-OCRv2_xx 制作部署案例，模型体积小，效果很好，属于性价比很高的选择。 
 
-为了节省大家的时间，已将预训练模型使用[保存模型服务化参数]()方法打包成压缩包，下载并解压即可使用。如你自训练的模型需经过保存模型服务化参数步骤才能服务化部署。
 ```
 python3 -m paddle_serving_app.package --get_model ocr_rec
 tar -xzvf ocr_rec.tar.gz
@@ -39,14 +44,19 @@ python3 -m paddle_serving_app.package --get_model ocr_det
 tar -xzvf ocr_det.tar.gz
 ```
 
-**二.下载测试数据集（可选）**
+**二.保存 Serving 部署的模型参数**
+
+为了节省大家的时间，已将预训练模型使用[保存用于 Serving 部署的模型参数](./5-1_Save_Model_Params_CN.md)方法打包成压缩包，下载并解压即可使用。如你自训练的模型需经过保存模型服务化参数步骤才能服务化部署。
+
+
+**三.下载测试数据集（可选）**
 第二步，下载测试图片集，如使用自有测试数据集，可忽略此步骤。
 ```
 wget --no-check-certificate https://paddle-serving.bj.bcebos.com/ocr/test_imgs.tar
 tar xf test_imgs.tar
 ```
 
-**三.修改 Config.yml 配置（可选）**
+**四.修改 `config.yml` 配置（可选）**
 第三步，通过修改配置文件设置服务、图、OP 级别属性。如果使用默认配置，此步骤可忽略。
 
 由于配置项较多，仅重点介绍部分核心选项的使用，完整配置选项说明可参考[ 配置说明]()
@@ -155,7 +165,8 @@ op:
             #min_subgraph_size: 3
 ```
 
-**四.代码与配置信息绑定 **
+**五.代码与配置信息绑定**
+
 第四步，实现代码和配置文件 Config.yml 绑定，以及设置多模型组合关系。具体包括：
 
 1. 重写模型前后处理：
@@ -199,7 +210,7 @@ ocr_service.run_service()
 ```
 
 
-**五.启动服务与验证**
+**六.启动服务与验证**
 
 启动服务前，可看到程序路径下所有文件路径如下：
 ```

doc/Offical_Docs/5-1_Save_Model_Params_CN.md

-# 保存用于 Paddle Serving 部署的模型参数
+# 保存用于 Serving 部署的模型参数
 
 模型参数信息保存在模型文件中，为什么还要保存用于 Paddle Serving 部署的模型参数呢，原因有3个：
 

examples/C++/PaddleClas/resnet_50_vd/README_CN.md

+# 图像分类
+
+## 1.获取模型
+
+```
+wget https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/ResNet50_vd_infer.tar && tar xf ResNet50_vd_infer.tar
+
+```
+
+## 2.用 paddle_serving_client 把下载的推理模型保存用于 Serving 部署的模型参数
+```
+# 保存 ResNet50_vd 模型参数
+python3 -m paddle_serving_client.convert --dirname ./ResNet50_vd_infer/ \
+                                         --model_filename inference.pdmodel  \
+                                         --params_filename inference.pdiparams \
+                                         --serving_server ./ResNet50_vd_serving/ \
+                                         --serving_client ./ResNet50_vd_client/
+```
+会在当前文件夹多出 `ResNet50_vd_serving` 和 `ResNet50_vd_client` 的文件夹
+
+
+保存参数后，会在当前文件夹多出 `ResNet50_vd_serving` 和 `ResNet50_vd_client` 的文件夹：
+```
+├── daisy.jpg
+├── http_client.py
+├── imagenet.label
+├── ResNet50_vd_client
+│   ├── serving_client_conf.prototxt
+│   └── serving_client_conf.stream.prototxt
+├── ResNet50_vd_infer
+│   ├── inference.pdiparams
+│   ├── inference.pdiparams.info
+│   └── inference.pdmodel
+├── ResNet50_vd_serving
+│   ├── fluid_time_file
+│   ├── inference.pdiparams
+│   ├── inference.pdmodel
+│   ├── serving_server_conf.prototxt
+│   └── serving_server_conf.stream.prototxt
+├── rpc_client.py
+```
+
+**三.启动服务**
+
+C++ Serving 服务可以指定一个网络端口同时接收 HTTP、gRPC 和 bRPC 请求。命令参数 `--model` 指定模型路径，`--gpu_ids` 指定 GPU 卡，`--port` 指定端口。
+
+```
+python3 -m paddle_serving_server.serve --model ResNet50_vd_serving --gpu_ids 0 --port 9394
+```
+
+**四.启动客户端**
+
+1. `rpc_client.py` 封装了 HTTP 请求客户端
+
+```
+python3 http_client.py
+```
+
+2. `http_client.py` 封装了 gRPC 请求客户端
+
+```
+python3 rpc_client.py
+```
+
+成功运行后，模型预测的结果会打印如下：
+
+```
+prediction: daisy, probability: 0.9341399073600769
+```

examples/C++/PaddleClas/resnet_50_vd/http_client.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import sys
+from paddle_serving_client import HttpClient
+
+#app
+from paddle_serving_app.reader import Sequential, URL2Image, Resize
+from paddle_serving_app.reader import CenterCrop, RGB2BGR, Transpose, Div, Normalize
+import time
+
+client = HttpClient()
+client.load_client_config("./ResNet50_vd_client/serving_client_conf.prototxt")
+'''
+if you want use GRPC-client, set_use_grpc_client(True)
+or you can directly use client.grpc_client_predict(...)
+as for HTTP-client,set_use_grpc_client(False)(which is default)
+or you can directly use client.http_client_predict(...)
+'''
+#client.set_use_grpc_client(True)
+'''
+if you want to enable Encrypt Module,uncommenting the following line
+'''
+#client.use_key("./key")
+'''
+if you want to compress,uncommenting the following line
+'''
+#client.set_response_compress(True)
+#client.set_request_compress(True)
+'''
+we recommend use Proto data format in HTTP-body, set True(which is default)
+if you want use JSON data format in HTTP-body, set False
+'''
+#client.set_http_proto(True)
+client.connect(["127.0.0.1:9394"])
+
+label_dict = {}
+label_idx = 0
+with open("imagenet.label") as fin:
+    for line in fin:
+        label_dict[label_idx] = line.strip()
+        label_idx += 1
+
+#preprocess
+seq = Sequential([
+    URL2Image(), Resize(256), CenterCrop(224), RGB2BGR(), Transpose((2, 0, 1)),
+    Div(255), Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225], True)
+])
+
+start = time.time()
+image_file = "https://paddle-serving.bj.bcebos.com/imagenet-example/daisy.jpg"
+for i in range(1):
+    img = seq(image_file)
+    res = client.predict(feed={"inputs": img}, fetch=[], batch=False)
+
+    if res is None:
+        raise ValueError("predict error")
+
+    if res.err_no != 0:
+        raise ValueError("predict error. Response : {}".format(res))
+
+    max_val = res.outputs[0].tensor[0].float_data[0]
+    max_idx = 0
+    for one_data in enumerate(res.outputs[0].tensor[0].float_data):
+        if one_data[1] > max_val:
+            max_val = one_data[1]
+            max_idx = one_data[0]
+    label = label_dict[max_idx].strip().replace(",", "")
+    print("prediction: {}, probability: {}".format(label, max_val))
+end = time.time()
+print(end - start)

examples/C++/PaddleClas/resnet_50_vd/rpc_client.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import sys
+from paddle_serving_client import Client
+
+#app
+from paddle_serving_app.reader import Sequential, URL2Image, Resize
+from paddle_serving_app.reader import CenterCrop, RGB2BGR, Transpose, Div, Normalize
+import time
+
+client = Client()
+client.load_client_config("./ResNet50_vd_client/serving_client_conf.prototxt")
+client.connect(["127.0.0.1:9394"])
+
+label_dict = {}
+label_idx = 0
+with open("imagenet.label") as fin:
+    for line in fin:
+        label_dict[label_idx] = line.strip()
+        label_idx += 1
+
+#preprocess
+seq = Sequential([
+    URL2Image(), Resize(256), CenterCrop(224), RGB2BGR(), Transpose((2, 0, 1)),
+    Div(255), Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225], True)
+])
+
+start = time.time()
+image_file = "https://paddle-serving.bj.bcebos.com/imagenet-example/daisy.jpg"
+for i in range(1):
+    img = seq(image_file)
+    fetch_map = client.predict(feed={"inputs": img}, fetch=[], batch=False)
+
+    prob = max(fetch_map["save_infer_model/scale_0.tmp_1"][0])
+    label = label_dict[fetch_map["save_infer_model/scale_0.tmp_1"][0].tolist()
+                       .index(prob)].strip().replace(",", "")
+    print("prediction: {}, probability: {}".format(label, prob))
+end = time.time()
+print(end - start)