Merge pull request #571 from MRXLT/new-flask

refine web service

doc/UWSGI_DEPLOY.md

-# 使用uwsgi启动HTTP预测服务
+# Deploy HTTP service with uWSGI
 
-在提供的fit_a_line示例中，启动HTTP预测服务后会看到有以下信息：
+([简体中文](./UWSGI_DEPLOY_CN.md)|English)
+
+In fit_a_line example, after starting the HTTP prediction service, you will see the following information:
 
 ```shell
 web service address:
@@ -13,46 +15,31 @@ http://10.127.3.150:9393/uci/prediction
  * Running on http://0.0.0.0:9393/ (Press CTRL+C to quit)
 ```
 
-这里会提示启动的HTTP服务是开发模式，并不能用于生产环境的部署。Flask启动的服务环境不够稳定也无法承受大量请求的并发，实际部署过程中配合需要WSGI（Web Server Gateway Interface）使用。
+Here you will be prompted that the HTTP service started is in development mode and cannot be used for production deployment. 
+The prediction service started by Flask is not stable enough to withstand the concurrency of a large number of requests. In the actual deployment process, WSGI (Web Server Gateway Interface) is used.
 
-下面我们展示一下如何使用[uWSGI](https://github.com/unbit/uwsgi)模块来部署HTTP预测服务用于生产环境。
+Next, we will show how to use the [uWSGI] (https://github.com/unbit/uwsgi) module to deploy HTTP prediction services for production environments.
 
-编写HTTP服务脚本
 
 ```python
 #uwsgi_service.py
 from paddle_serving_server.web_service import WebService
-from flask import Flask, request
 
-#配置预测服务
+#Define prediction service
 uci_service = WebService(name = "uci")
 uci_service.load_model_config("./uci_housing_model")
 uci_service.prepare_server(workdir="./workdir", port=int(9500), device="cpu")
 uci_service.run_server()
-
-#配置flask服务
-app_instance = Flask(__name__)
-@app_instance.before_first_request
-def init():
-    global uci_service
-    uci_service._launch_web_service()
-
-service_name = "/" + uci_service.name + "/prediction"
-@app_instance.route(service_name, methods=["POST"])
-def run():
-    return uci_service.get_prediction(request)
-
-#run方法用于直接调试中直接启动服务
-if __name__ == "__main__":
-    app_instance.run()
+#Get flask application
+app_instance = uci_service.get_app_instance()
 ```
 
-使用uwsgi启动HTTP服务
+Start service with uWSGI
 
 ```bash
-uwsgi --http :9000 --wsgi-file uwsgi_service.py --callable app_instance --processes 4
+uwsgi --http :9393 --module uwsgi_service:app_instance
 ```
 
-使用--processes参数可以指定服务的进程数，请注意目前Serving HTTP 服务暂时不支持多线程的方式使用。
+Use the --processes parameter to specify the number of service processes. 
 
-更多uWSGI的信息请参考[uWSGI使用文档](https://uwsgi-docs.readthedocs.io/en/latest/)
+For more information about uWSGI, please refer to [uWSGI documentation](https://uwsgi-docs.readthedocs.io/en/latest/)

doc/UWSGI_DEPLOY_CN.md

+# 使用uwsgi启动HTTP预测服务
+
+(简体中文|[English](./UWSGI_DEPLOY.md))
+
+在提供的fit_a_line示例中，启动HTTP预测服务后会看到有以下信息：
+
+```shell
+web service address:
+http://10.127.3.150:9393/uci/prediction
+ * Serving Flask app "serve" (lazy loading)
+ * Environment: production
+   WARNING: This is a development server. Do not use it in a production deployment.
+   Use a production WSGI server instead.
+ * Debug mode: off
+ * Running on http://0.0.0.0:9393/ (Press CTRL+C to quit)
+```
+
+这里会提示启动的HTTP服务是开发模式，并不能用于生产环境的部署。Flask启动的服务环境不够稳定也无法承受大量请求的并发，实际部署过程中配合需要WSGI（Web Server Gateway Interface）使用。
+
+下面我们展示一下如何使用[uWSGI](https://github.com/unbit/uwsgi)模块来部署HTTP预测服务用于生产环境。
+
+编写HTTP服务脚本
+
+```python
+#uwsgi_service.py
+from paddle_serving_server.web_service import WebService
+
+#配置预测服务
+uci_service = WebService(name = "uci")
+uci_service.load_model_config("./uci_housing_model")
+uci_service.prepare_server(workdir="./workdir", port=int(9500), device="cpu")
+uci_service.run_server()
+#获取flask服务
+app_instance = uci_service.get_app_instance()
+```
+
+使用uwsgi启动HTTP服务
+
+```bash
+uwsgi --http :9393 --module uwsgi_service:app_instance
+```
+
+使用--processes参数可以指定服务的进程数。
+
+更多uWSGI的信息请参考[uWSGI使用文档](https://uwsgi-docs.readthedocs.io/en/latest/)

python/examples/senta/senta_web_service.py

@@ -51,13 +51,11 @@ class SentaService(WebService):
     def init_lac_service(self):
         ps = Process(target=self.start_lac_service())
         ps.start()
-        #self.init_lac_client()
+        self.init_lac_client()
 
     def lac_predict(self, feed_data):
-        self.init_lac_client()
         lac_result = self.lac_client.predict(
             feed={"words": feed_data}, fetch=["crf_decode"])
-        self.lac_client.release()
         return lac_result
 
     def init_lac_client(self):

python/paddle_serving_app/local_predict.py

@@ -115,6 +115,13 @@ class Debugger(object):
 
         inputs = []
         for name in self.feed_names_:
+            if isinstance(feed[name], list):
+                feed[name] = np.array(feed[name]).reshape(self.feed_shapes_[
+                    name])
+                if self.feed_types_[name] == 0:
+                    feed[name] = feed[name].astype("int64")
+                else:
+                    feed[name] = feed[name].astype("float32")
             inputs.append(PaddleTensor(feed[name][np.newaxis, :]))
 
         outputs = self.predictor.run(inputs)

python/paddle_serving_server/web_service.py

@@ -101,7 +101,6 @@ class WebService(object):
         p_rpc = Process(target=self._launch_rpc_service)
         p_rpc.start()
 
-    def run_flask(self):
         app_instance = Flask(__name__)
 
         @app_instance.before_first_request
@@ -114,10 +113,16 @@ class WebService(object):
         def run():
             return self.get_prediction(request)
 
-        app_instance.run(host="0.0.0.0",
-                         port=self.port,
-                         threaded=False,
-                         processes=4)
+        self.app_instance = app_instance
+
+    def run_flask(self):
+        self.app_instance.run(host="0.0.0.0",
+                              port=self.port,
+                              threaded=False,
+                              processes=1)
+
+    def get_app_instance(self):
+        return self.app_instance
 
     def preprocess(self, feed=[], fetch=[]):
         return feed, fetch

python/paddle_serving_server_gpu/web_service.py

@@ -151,7 +151,6 @@ class WebService(object):
         for p in server_pros:
             p.start()
 
-    def run_flask(self):
         app_instance = Flask(__name__)
 
         @app_instance.before_first_request
@@ -164,10 +163,16 @@ class WebService(object):
         def run():
             return self.get_prediction(request)
 
-        app_instance.run(host="0.0.0.0",
-                         port=self.port,
-                         threaded=False,
-                         processes=4)
+        self.app_instance = app_instance
+
+    def run_flask(self):
+        self.app_instance.run(host="0.0.0.0",
+                              port=self.port,
+                              threaded=False,
+                              processes=1)
+
+    def get_app_instance(self):
+        return app_instance
 
     def preprocess(self, feed=[], fetch=[]):
         return feed, fetch