fix codestyle

doc/PIPELINE_SERVING.md

@@ -47,6 +47,7 @@ The graph execution engine consists of OPs and Channels, and the connected OPs s
 </center>
 
 
+
 ### Extreme Case Consideration
 
 - Request timeout

doc/PIPELINE_SERVING_CN.md

@@ -6,6 +6,7 @@ Paddle Serving 通常用于单模型的一键部署，但端到端的深度学
 
 Paddle Serving 提供了用户友好的多模型组合服务编程框架，Pipeline Serving，旨在降低编程门槛，提高资源使用率（尤其是GPU设备），提升整体的预估效率。
 
+
 ## 整体架构设计
 
 Server端基于 gRPC 和图执行引擎构建，两者的关系如下图所示。

python/pipeline/operator.py

@@ -450,7 +450,8 @@ class Op(object):
                             data_offset_right = input_offset[idx + 1]
                             lod_offset_left = lod_offset[data_offset_left]
                             lod_offset_right = lod_offset[data_offset_right]
-                            midped_data_dict[data_id][name] = value[lod_offset_left:lod_offset_right]
+                            midped_data_dict[data_id][name] = value[
+                                lod_offset_left:lod_offset_right]
                             midped_data_dict[data_id][lod_offset_name] = \
                                     lod_offset[data_offset_left:data_offset_right + 1] - lod_offset[data_offset_left]
                     else:

python/pipeline/pipeline_server.py

@@ -117,7 +117,8 @@ class PipelineServer(object):
             server = grpc.server(
                 futures.ThreadPoolExecutor(max_workers=self._worker_num),
                 options=[('grpc.max_send_message_length', 256 * 1024 * 1024),
-                    ('grpc.max_receive_message_length', 256 * 1024 * 1024)])
+                         ('grpc.max_receive_message_length', 256 * 1024 * 1024)
+                         ])
             pipeline_service_pb2_grpc.add_PipelineServiceServicer_to_server(
                 PipelineServicer(self._response_op, self._conf), server)
             server.add_insecure_port('[::]:{}'.format(self._port))

python/pipeline/profiler.py

@@ -130,7 +130,8 @@ class PerformanceTracer(object):
                 _LOGGER.info("\tQuery count[{}]".format(tot))
                 _LOGGER.info("\tQPS[{} q/s]".format(qps))
                 _LOGGER.info("\tSucc[{}]".format(1 - 1.0 * err_count / tot))
-                _LOGGER.info("\tError req[{}]".format(", ".join([str(x) for x in err_request)]))
+                _LOGGER.info("\tError req[{}]".format(", ".join(
+                    [str(x) for x in err_request])))
                 _LOGGER.info("\tLatency:")
                 _LOGGER.info("\t\tave[{} ms]".format(ave_cost))
                 for latency in latencys: