add mkl argument

README.md

@@ -84,6 +84,7 @@ python -m paddle_serving_server.serve --model uci_housing_model --thread 10 --po
 | `model` | str | `""` | Path of paddle model directory to be served |
 | `mem_optim` | bool | `False` | Enable memory / graphic memory optimization |
 | `ir_optim` | bool | `False` | Enable analysis and optimization of calculation graph |
+| `use_mkl` (Only for cpu version) | bool | `False` | Run inference with MKL |
 
 Here, we use `curl` to send a HTTP POST request to the service we just started. Users can use any python library to send HTTP POST as well, e.g, [requests](https://requests.readthedocs.io/en/master/).
 </center>

README_CN.md

@@ -88,6 +88,7 @@ python -m paddle_serving_server.serve --model uci_housing_model --thread 10 --po
 | `model` | str | `""` | Path of paddle model directory to be served |
 | `mem_optim` | bool | `False` | Enable memory optimization |
 | `ir_optim` | bool | `False` | Enable analysis and optimization of calculation graph |
+| `use_mkl` (Only for cpu version) | bool | `False` | Run inference with MKL |
 
 我们使用 `curl` 命令来发送HTTP POST请求给刚刚启动的服务。用户也可以调用python库来发送HTTP POST请求，请参考英文文档 [requests](https://requests.readthedocs.io/en/master/)。
 </center>

python/paddle_serving_server/__init__.py

@@ -289,8 +289,8 @@ class Server(object):
         # check config here
         # print config here
 
-    def use_mkl(self):
-        self.mkl_flag = True
+    def use_mkl(self, flag):
+        self.mkl_flag = flag
 
     def get_device_version(self):
         avx_flag = False
@@ -305,6 +305,10 @@ class Server(object):
             else:
                 device_version = "serving-cpu-avx-openblas-"
         else:
+            if mkl_flag:
+                print(
+                    "Your CPU does not support AVX, server will running with noavx-openblas mode."
+                )
             device_version = "serving-cpu-noavx-openblas-"
         return device_version
 

python/paddle_serving_server/serve.py

@@ -43,6 +43,7 @@ def parse_args():  # pylint: disable=doc-string-missing
         "--mem_optim", type=bool, default=False, help="Memory optimize")
     parser.add_argument(
         "--ir_optim", type=bool, default=False, help="Graph optimize")
+    parser.add_argument("--use_mkl", type=bool, default=False, help="Use MKL")
     parser.add_argument(
         "--max_body_size",
         type=int,
@@ -61,6 +62,7 @@ def start_standard_model():  # pylint: disable=doc-string-missing
     mem_optim = args.mem_optim
     ir_optim = args.ir_optim
     max_body_size = args.max_body_size
+    use_mkl = args.use_mkl
 
     if model == "":
         print("You must specify your serving model")
@@ -82,6 +84,7 @@ def start_standard_model():  # pylint: disable=doc-string-missing
     server.set_num_threads(thread_num)
     server.set_memory_optimize(mem_optim)
     server.set_ir_optimize(ir_optim)
+    server.use_mkl(use_mkl)
     server.set_max_body_size(max_body_size)
     server.set_port(port)