!7908 add reduce precision in pynative mode

Merge pull request !7908 from chujinjin/add_reduce_precision_in_pynative_mode

mindspore/ccsrc/backend/session/gpu_session.cc

@@ -141,6 +141,15 @@ void GPUSession::HardwareOptimize(const std::shared_ptr<KernelGraph> &kernel_gra
   kernel_graph->SetExecOrderByDefault();
 }
 
+void GPUSession::RunOpHardwareOptimize(const std::shared_ptr<KernelGraph> &kernel_graph) {
+  auto optimizer = std::make_shared<opt::GraphOptimizer>();
+  auto pm = std::make_shared<opt::PassManager>();
+  pm->AddPass(std::make_shared<opt::ReducePrecisionFusion>("reduce_precision"));
+  optimizer->AddPassManager(pm);
+  (void)optimizer->Optimize(kernel_graph);
+  kernel_graph->SetExecOrderByDefault();
+}
+
 void GPUSession::GraphKernelOptimize(const std::shared_ptr<KernelGraph> &kernel_graph) {
   auto context_ptr = MsContext::GetInstance();
   MS_EXCEPTION_IF_NULL(context_ptr);
@@ -350,6 +359,7 @@ void GPUSession::BuildOpImpl(const OpRunInfo &op_run_info, const GraphInfo &grap
   auto kernel_graph = ConstructSingleOpGraph(op_run_info, input_tensors, tensors_mask);
   MS_EXCEPTION_IF_NULL(kernel_graph);
   SelectKernel(kernel_graph);
+  RunOpHardwareOptimize(kernel_graph);
   StartKernelRT();
   // Hide NopOp from execution graph
   opt::HideNopNode(kernel_graph.get());

mindspore/ccsrc/backend/session/gpu_session.h

@@ -50,6 +50,8 @@ class GPUSession : public SessionBasic {
 
   void HardwareOptimize(const std::shared_ptr<KernelGraph> &kernel_graph);
 
+  void RunOpHardwareOptimize(const std::shared_ptr<KernelGraph> &kernel_graph);
+
   void GraphKernelOptimize(const std::shared_ptr<KernelGraph> &kernel_graph);
 
   void AssignStream(const std::shared_ptr<KernelGraph> &kernel_graph);

mindspore/ccsrc/runtime/device/gpu/kernel_info_setter.cc

@@ -395,11 +395,7 @@ void SetKernelInfo(const CNodePtr &kernel_node, KernelType kernel_type) {
     result =
       kernel::GpuKernelFactory::GetInstance().SearchRegistered(AnfAlgo::GetCNodeName(kernel_node), builder->Build());
     if (!result) {
-      auto ms_context = MsContext::GetInstance();
-      MS_EXCEPTION_IF_NULL(ms_context);
-      if (ms_context->get_param<int>(MS_CTX_EXECUTION_MODE) == kGraphMode) {
-        result = kernel::GpuKernelFactory::GetInstance().ReducePrecision(AnfAlgo::GetCNodeName(kernel_node), builder);
-      }
+      result = kernel::GpuKernelFactory::GetInstance().ReducePrecision(AnfAlgo::GetCNodeName(kernel_node), builder);
     }
     if (!result) {
       result = SelectAkgKernel(kernel_node, builder->Build());