diff --git a/paddle/fluid/framework/parallel_executor.cc b/paddle/fluid/framework/parallel_executor.cc
index 3a9027713afb5287c7addf8be745acfd185104ee..68880a2ad61e123090d82d939150310729f33778 100644
--- a/paddle/fluid/framework/parallel_executor.cc
+++ b/paddle/fluid/framework/parallel_executor.cc
@@ -95,7 +95,7 @@ ParallelExecutor::ParallelExecutor(
   }
 
   if (member_->local_scopes_.size() != 1 && local_scopes.empty()) {
-    BCastParamsToDevs(bcast_vars);
+    BCastParamsToDevices(bcast_vars);
   }
   // Startup Program has been run. All local scopes has correct parameters.
 
@@ -131,7 +131,7 @@ ParallelExecutor::ParallelExecutor(
       member_->places_, std::move(member_->executor_)));
 }
 
-void ParallelExecutor::BCastParamsToDevs(
+void ParallelExecutor::BCastParamsToDevices(
     const std::unordered_set<std::string> &vars) const {
   // the the initializing bcast, all vars would be bcast from device(0),
   // otherwise
diff --git a/paddle/fluid/framework/parallel_executor.h b/paddle/fluid/framework/parallel_executor.h
index 6985b6540690c6218bcee51ba0e69f3d34812bfc..ffb9934a2d702b2bf6db7ad75a6bf9867e1e9901 100644
--- a/paddle/fluid/framework/parallel_executor.h
+++ b/paddle/fluid/framework/parallel_executor.h
@@ -66,7 +66,7 @@ class ParallelExecutor {
   void Run(const std::vector<std::string> &fetch_tensors,
            const std::string &fetched_var_name);
 
-  void BCastParamsToDevs(const std::unordered_set<std::string> &vars) const;
+  void BCastParamsToDevices(const std::unordered_set<std::string> &vars) const;
 
  private:
   ParallelExecutorPrivate *member_;
diff --git a/paddle/fluid/pybind/pybind.cc b/paddle/fluid/pybind/pybind.cc
index 227fd442c28a5da222efb877dae18b5e1922c66a..d5cbcaf7260f66ce14ce89d2e03bb8ecb5c4dec2 100644
--- a/paddle/fluid/pybind/pybind.cc
+++ b/paddle/fluid/pybind/pybind.cc
@@ -665,7 +665,7 @@ All parameter, weight, gradient are variables in Paddle.
                   const std::string &, Scope *, std::vector<Scope *> &,
                   const ExecutionStrategy &, const BuildStrategy &, size_t,
                   size_t>())
-      .def("bcast_params", &ParallelExecutor::BCastParamsToDevs)
+      .def("bcast_params", &ParallelExecutor::BCastParamsToDevices)
       // NOTE: even we return a vec<Scope*>* to Python use reference policy.
       // We still cannot get local_scope from this vector, since the element
       // of vec<Scope*> will be freed by Python GC. We can only return Scope*