Merge branch 'PaddlePaddle:develop' into develop

core/general-server/op/general_dist_kv_infer_op.cpp

@@ -40,7 +40,7 @@ using baidu::paddle_serving::predictor::PaddleGeneralModelConfig;
 using baidu::paddle_serving::predictor::CubeCache;
 
 // DistKV Infer Op: seek cube and then call paddle inference
-// op seq: general_reader-> dist_kv_infer -> general_response
+// op seq: GeneralReaderOp-> dist_kv_infer -> general_response
 int GeneralDistKVInferOp::inference() {
   VLOG(2) << "Going to run inference";
   const std::vector<std::string> pre_node_names = pre_names();
@@ -186,9 +186,9 @@ int GeneralDistKVInferOp::inference() {
   if (values.size() != keys.size() || values[0].buff.size() == 0) {
     LOG(ERROR) << "cube value return null";
   }
-   size_t EMBEDDING_SIZE = values[0].buff.size() / sizeof(float);
+  size_t EMBEDDING_SIZE = values[0].buff.size() / sizeof(float);
   // size_t EMBEDDING_SIZE = (values[0].buff.size() - 10) / sizeof(float);
-  //size_t EMBEDDING_SIZE = 9;
+  // size_t EMBEDDING_SIZE = 9;
   TensorVector sparse_out;
   sparse_out.resize(sparse_count);
   TensorVector dense_out;
@@ -241,7 +241,7 @@ int GeneralDistKVInferOp::inference() {
 
       // The data generated by pslib has 10 bytes of information to be filtered
       // out
-      memcpy(data_ptr, cur_val->buff.data(), cur_val->buff.size() );
+      memcpy(data_ptr, cur_val->buff.data(), cur_val->buff.size());
       // VLOG(3) <<  keys[cube_val_idx] << ":" << data_ptr[0] << ", " <<
       // data_ptr[1] << ", " <<data_ptr[2] << ", " <<data_ptr[3] << ", "
       // <<data_ptr[4] << ", " <<data_ptr[5] << ", " <<data_ptr[6] << ", "

doc/C++_Serving/Model_Ensemble_CN.md

@@ -45,13 +45,13 @@ from paddle_serving_server import OpGraphMaker
 from paddle_serving_server import Server
 
 op_maker = OpMaker()
-read_op = op_maker.create('general_reader')
+read_op = op_maker.create('GeneralReaderOp')
 cnn_infer_op = op_maker.create(
-    'general_infer', engine_name='cnn', inputs=[read_op])
+    'GeneralInferOp', engine_name='cnn', inputs=[read_op])
 bow_infer_op = op_maker.create(
-    'general_infer', engine_name='bow', inputs=[read_op])
+    'GeneralInferOp', engine_name='bow', inputs=[read_op])
 response_op = op_maker.create(
-    'general_response', inputs=[cnn_infer_op, bow_infer_op])
+    'GeneralResponseOp', inputs=[cnn_infer_op, bow_infer_op])
 
 op_graph_maker = OpGraphMaker()
 op_graph_maker.add_op(read_op)

doc/C++_Serving/Model_Ensemble_EN.md

@@ -45,13 +45,13 @@ from paddle_serving_server import OpGraphMaker
 from paddle_serving_server import Server
 
 op_maker = OpMaker()
-read_op = op_maker.create('general_reader')
+read_op = op_maker.create('GeneralReaderOp')
 cnn_infer_op = op_maker.create(
-    'general_infer', engine_name='cnn', inputs=[read_op])
+    'GeneralInferOp', engine_name='cnn', inputs=[read_op])
 bow_infer_op = op_maker.create(
-    'general_infer', engine_name='bow', inputs=[read_op])
+    'GeneralInferOp', engine_name='bow', inputs=[read_op])
 response_op = op_maker.create(
-    'general_response', inputs=[cnn_infer_op, bow_infer_op])
+    'GeneralResponseOp', inputs=[cnn_infer_op, bow_infer_op])
 
 op_graph_maker = OpGraphMaker()
 op_graph_maker.add_op(read_op)

doc/C++_Serving/OP_CN.md

@@ -138,18 +138,21 @@ DEFINE_OP(GeneralInferOp);
 
 
 ``` python
-self.op_dict = {
-            "general_infer": "GeneralInferOp",
-            "general_reader": "GeneralReaderOp",
-            "general_response": "GeneralResponseOp",
-            "general_text_reader": "GeneralTextReaderOp",
-            "general_text_response": "GeneralTextResponseOp",
-            "general_single_kv": "GeneralSingleKVOp",
-            "general_dist_kv": "GeneralDistKVOp"
-        }
+self.op_list = [
+            "GeneralInferOp",
+            "GeneralReaderOp",
+            "GeneralResponseOp",
+            "GeneralTextReaderOp",
+            "GeneralTextResponseOp",
+            "GeneralSingleKVOp",
+            "GeneralDistKVInferOp",
+            "GeneralDistKVOp",
+            "GeneralCopyOp",
+            "GeneralDetectionOp",
+        ]
 ```
 
-在`python/paddle_serving_server/server.py`文件中仅添加`需要加载模型，执行推理预测的自定义的C++OP类的类名`。例如`general_reader`由于只是做一些简单的数据处理而不加载模型调用预测，故在👆的代码中需要添加，而不添加在👇的代码中。
+在`python/paddle_serving_server/server.py`文件中仅添加`需要加载模型，执行推理预测的自定义的C++OP类的类名`。例如`GeneralReaderOp`由于只是做一些简单的数据处理而不加载模型调用预测，故在👆的代码中需要添加，而不添加在👇的代码中。
 ``` python
 default_engine_types = [
                 'GeneralInferOp',

doc/C++_Serving/OP_EN.md

@@ -136,20 +136,23 @@ After you have defined a C++ operator on server side for Paddle Serving, the las
 
 
 ``` python
-self.op_dict = {
-            "general_infer": "GeneralInferOp",
-            "general_reader": "GeneralReaderOp",
-            "general_response": "GeneralResponseOp",
-            "general_text_reader": "GeneralTextReaderOp",
-            "general_text_response": "GeneralTextResponseOp",
-            "general_single_kv": "GeneralSingleKVOp",
-            "general_dist_kv": "GeneralDistKVOp"
-        }
+self.op_list = [
+            "GeneralInferOp",
+            "GeneralReaderOp",
+            "GeneralResponseOp",
+            "GeneralTextReaderOp",
+            "GeneralTextResponseOp",
+            "GeneralSingleKVOp",
+            "GeneralDistKVInferOp",
+            "GeneralDistKVOp",
+            "GeneralCopyOp",
+            "GeneralDetectionOp",
+        ]
 ```
 
 In `python/paddle_serving_server/server.py` file, only the class name of the C++ OP class that needs to load the model and execute prediction is added. 
 
-For example, `general_reader`, need to be added in the 👆 code, but not in the 👇 code. Because it only does some simple data processing without loading the model and call prediction. 
+For example, `GeneralReaderOp`, need to be added in the 👆 code, but not in the 👇 code. Because it only does some simple data processing without loading the model and call prediction. 
 ``` python
 default_engine_types = [
                 'GeneralInferOp',

doc/Run_On_NPU_CN.md

@@ -75,7 +75,7 @@ go env -w GOPROXY=https://goproxy.cn,direct
 go install github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway@v1.15.2
 go install github.com/grpc-ecosystem/grpc-gateway/protoc-gen-swagger@v1.15.2
 go install github.com/golang/protobuf/protoc-gen-go@v1.4.3
-go install google.golang.org/grpc@v1.33.0
+go install google.golang.org/grpc@v1.33.1
 go env -w GO111MODULE=auto
 ```
 
@@ -193,4 +193,4 @@ python3 -m paddle_serving_server.serve --model resnet_v2_50_imagenet_model --thr
 ## 其他说明
 
 ### NPU芯片支持相关参考资料
-* [昇腾NPU芯片运行飞桨](https://www.paddlepaddle.org.cn/documentation/docs/zh/guides/09_hardware_support/npu_docs/paddle_install_cn.html)
\ No newline at end of file
+* [昇腾NPU芯片运行飞桨](https://www.paddlepaddle.org.cn/documentation/docs/zh/guides/09_hardware_support/npu_docs/paddle_install_cn.html)

examples/C++/PaddleNLP/bert/bert_gpu_server.py

@@ -19,9 +19,9 @@ from paddle_serving_server import OpSeqMaker
 from paddle_serving_server import Server
 
 op_maker = OpMaker()
-read_op = op_maker.create('general_reader')
-general_infer_op = op_maker.create('general_infer')
-general_response_op = op_maker.create('general_response')
+read_op = op_maker.create('GeneralReaderOp')
+general_infer_op = op_maker.create('GeneralInferOp')
+general_response_op = op_maker.create('GeneralResponseOp')
 
 op_seq_maker = OpSeqMaker()
 op_seq_maker.add_op(read_op)

examples/C++/PaddleNLP/bert/bert_server.py

@@ -19,9 +19,9 @@ from paddle_serving_server import OpSeqMaker
 from paddle_serving_server import Server
 
 op_maker = OpMaker()
-read_op = op_maker.create('general_reader')
-general_infer_op = op_maker.create('general_infer')
-general_response_op = op_maker.create('general_response')
+read_op = op_maker.create('GeneralReaderOp')
+general_infer_op = op_maker.create('GeneralInferOp')
+general_response_op = op_maker.create('GeneralResponseOp')
 
 op_seq_maker = OpSeqMaker()
 op_seq_maker.add_op(read_op)

examples/C++/PaddleRec/criteo_ctr/test_server.py

@@ -20,14 +20,14 @@ from paddle_serving_server import OpSeqMaker
 from paddle_serving_server import Server
 
 op_maker = OpMaker()
-read_op = op_maker.create('general_reader')
-general_infer_op = op_maker.create('general_infer')
-response_op = op_maker.create('general_response')
+read_op = op_maker.create('GeneralReaderOp')
+general_infer_op = op_maker.create('GeneralInferOp')
+general_response_op = op_maker.create('GeneralResponseOp')
 
 op_seq_maker = OpSeqMaker()
 op_seq_maker.add_op(read_op)
 op_seq_maker.add_op(general_infer_op)
-op_seq_maker.add_op(response_op)
+op_seq_maker.add_op(general_response_op)
 
 server = Server()
 server.set_op_sequence(op_seq_maker.get_op_sequence())

python/pipeline/local_service_handler.py

@@ -189,12 +189,11 @@ class LocalServiceHandler(object):
             "mkldnn_bf16_op_list:{}, use_ascend_cl:{}, min_subgraph_size:{},"
             "is_set_dynamic_shape_info:{}".format(
                 model_config, self._device_name, self._use_gpu, self._use_trt,
-                self._use_lite, self._use_xpu, device_type, self._devices,
-                self._mem_optim, self._ir_optim, self._use_profile,
-                self._thread_num, self._client_type, self._fetch_names,
-                self._precision, self._use_calib, self._use_mkldnn, 
-                self._mkldnn_cache_capacity, self._mkldnn_op_list, 
-                self._mkldnn_bf16_op_list, self._use_ascend_cl, 
+                self._use_lite, self._use_xpu, device_type, self._devices, self.
+                _mem_optim, self._ir_optim, self._use_profile, self._thread_num,
+                self._client_type, self._fetch_names, self._precision, self.
+                _use_calib, self._use_mkldnn, self._mkldnn_cache_capacity, self.
+                _mkldnn_op_list, self._mkldnn_bf16_op_list, self._use_ascend_cl,
                 self.min_subgraph_size, bool(len(self.dynamic_shape_info))))
 
     def get_fetch_list(self):
@@ -282,9 +281,9 @@ class LocalServiceHandler(object):
         if self._device_name == "cpu":
             from paddle_serving_server import OpMaker, OpSeqMaker, Server
             op_maker = OpMaker()
-            read_op = op_maker.create('general_reader')
-            general_infer_op = op_maker.create('general_infer')
-            general_response_op = op_maker.create('general_response')
+            read_op = op_maker.create('GeneralReaderOp')
+            general_infer_op = op_maker.create('GeneralInferOp')
+            general_response_op = op_maker.create('GeneralResponseOp')
 
             op_seq_maker = OpSeqMaker()
             op_seq_maker.add_op(read_op)
@@ -296,9 +295,9 @@ class LocalServiceHandler(object):
             #gpu or arm
             from paddle_serving_server import OpMaker, OpSeqMaker, Server
             op_maker = OpMaker()
-            read_op = op_maker.create('general_reader')
-            general_infer_op = op_maker.create('general_infer')
-            general_response_op = op_maker.create('general_response')
+            read_op = op_maker.create('GeneralReaderOp')
+            general_infer_op = op_maker.create('GeneralInferOp')
+            general_response_op = op_maker.create('GeneralResponseOp')
 
             op_seq_maker = OpSeqMaker()
             op_seq_maker.add_op(read_op)