diff --git a/paddle/fluid/framework/parallel_executor.cc b/paddle/fluid/framework/parallel_executor.cc
index d0c4d8dd8b3e15d9aef9065880664031dfe53499..f9fc35d8ce3ecc042739ed54a5595c590fb94de7 100644
--- a/paddle/fluid/framework/parallel_executor.cc
+++ b/paddle/fluid/framework/parallel_executor.cc
@@ -376,7 +376,7 @@ struct NCCLAllReduceOpHandle : public OpHandle {
       return;  // No need to all reduce when GPU count = 1;
     } else {
       auto &var_name = static_cast<VarHandle *>(this->inputs_[0])->name_;
-
+      VLOG(3) << "Invoke NCCL AllReduce";
       int dtype = -1;
       size_t numel = 0;