diff --git a/tests/nnfw_api/src/GenModelTests/BranchModelTrain.test.cc b/tests/nnfw_api/src/GenModelTests/BranchModelTrain.test.cc
index 4cf36f376cd..1a224ae0645 100644
--- a/tests/nnfw_api/src/GenModelTests/BranchModelTrain.test.cc
+++ b/tests/nnfw_api/src/GenModelTests/BranchModelTrain.test.cc
@@ -56,28 +56,26 @@ TEST_F(GenModelTrain, BranchOps_FC_Add)
 
     SUCCEED();
   }
-}
 
-TEST_F(GenModelTrain, BranchOps_FC_Sub)
-{
-  // (( Input 0 )) --------------\
-  //                              |=> [ Sub ] -> (( Output 0 ))
-  // (( Input 1 )) -> [ FC ] ----/
+  // (( Input 0 )) -> [ FC ] -> (fc_out) --------------------------╲
+  //                                    ╲                           |=> [ Add ] -> (( Output 0 ))
+  //                                     ╲-> [ Relu6 ]⎼> (ea_out) -╱
   {
     CirclePlusGen cgen;
 
     uint32_t weight_buf = cgen.addBuffer(std::vector<float>(8 * 2, 0.f));
     uint32_t bias_buf = cgen.addBuffer(std::vector<float>(8, 0.f));
-    int input0 = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
-    int input1 = cgen.addTensor({{1, 2}, circle::TensorType::TensorType_FLOAT32});
+    int input0 = cgen.addTensor({{1, 2}, circle::TensorType::TensorType_FLOAT32});
     int weight = cgen.addTensor({{8, 2}, circle::TensorType::TensorType_FLOAT32, weight_buf});
     int bias = cgen.addTensor({{8}, circle::TensorType::TensorType_FLOAT32, bias_buf});
     int fc_output = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
+    int ea_output = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
     int output = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
-    cgen.addOperatorFullyConnected({{input1, weight, bias}, {fc_output}});
-    cgen.addOperatorSub({{input0, fc_output}, {output}},
+    cgen.addOperatorFullyConnected({{input0, weight, bias}, {fc_output}});
+    cgen.addOperatorRelu6({{fc_output}, {ea_output}});
+    cgen.addOperatorAdd({{fc_output, ea_output}, {output}},
                         circle::ActivationFunctionType::ActivationFunctionType_NONE);
-    cgen.setInputsAndOutputs({input0, input1}, {output});
+    cgen.setInputsAndOutputs({input0}, {output});
 
     float learning_rate = 0.01f;
     int32_t batch_size = 1;
@@ -87,37 +85,40 @@ TEST_F(GenModelTrain, BranchOps_FC_Sub)
                        NNFW_TRAIN_TRAINABLE_ALL});
 
     _context = std::make_unique<GenModelTrainContext>(cgen.finish());
-    _context->addTrainCase(uniformTCD<float>(
-      {{{0, 1, 2, 3, 4, 5, 1, 3}, {6, 7}}, {{5, 4, 3, 2, 1, 0, 2, 1}, {7, 6}}}, // inputs
-      {{{2, 1, 5, 5, 2, 1, 5, 5}}, {{2, 1, 5, 5, 2, 1, 5, 6}}},                 // expected
-      {{7.3265f}, {4.6811f}, {3.6735f}, {3.2863f}}                              // loss
-      ));
+    _context->addTrainCase(
+      uniformTCD<float>({{{1, 3}}, {{2, 1}}},                                     // inputs
+                        {{{2, 1, 5, 5, 2, 1, 5, 5}}, {{2, 1, 5, 5, 2, 1, 5, 6}}}, // expected
+                        // TODO Modify loss values to results of tensorflow
+                        {{14.0124f}, {11.0036f}, {8.1681f}, {6.0974f}} // loss
+                        ));
 
     _context->setBackends({"train"});
     _context->setEpoch(4);
 
     SUCCEED();
   }
+}
 
-  // (( Input 0 )) -> [ FC ] -> (fc_out) --------------------------╲
-  //                                    ╲                           |=> [ Add ] -> (( Output 0 ))
-  //                                     ╲-> [ Relu6 ]⎼> (ea_out) -╱
+TEST_F(GenModelTrain, BranchOps_FC_Sub)
+{
+  // (( Input 0 )) --------------\
+  //                              |=> [ Sub ] -> (( Output 0 ))
+  // (( Input 1 )) -> [ FC ] ----/
   {
     CirclePlusGen cgen;
 
     uint32_t weight_buf = cgen.addBuffer(std::vector<float>(8 * 2, 0.f));
     uint32_t bias_buf = cgen.addBuffer(std::vector<float>(8, 0.f));
-    int input0 = cgen.addTensor({{1, 2}, circle::TensorType::TensorType_FLOAT32});
+    int input0 = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
+    int input1 = cgen.addTensor({{1, 2}, circle::TensorType::TensorType_FLOAT32});
     int weight = cgen.addTensor({{8, 2}, circle::TensorType::TensorType_FLOAT32, weight_buf});
     int bias = cgen.addTensor({{8}, circle::TensorType::TensorType_FLOAT32, bias_buf});
     int fc_output = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
-    int ea_output = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
     int output = cgen.addTensor({{1, 8}, circle::TensorType::TensorType_FLOAT32});
-    cgen.addOperatorFullyConnected({{input0, weight, bias}, {fc_output}});
-    cgen.addOperatorRelu6({{fc_output}, {ea_output}});
-    cgen.addOperatorAdd({{fc_output, ea_output}, {output}},
+    cgen.addOperatorFullyConnected({{input1, weight, bias}, {fc_output}});
+    cgen.addOperatorSub({{input0, fc_output}, {output}},
                         circle::ActivationFunctionType::ActivationFunctionType_NONE);
-    cgen.setInputsAndOutputs({input0}, {output});
+    cgen.setInputsAndOutputs({input0, input1}, {output});
 
     float learning_rate = 0.01f;
     int32_t batch_size = 1;
@@ -127,12 +128,11 @@ TEST_F(GenModelTrain, BranchOps_FC_Sub)
                        NNFW_TRAIN_TRAINABLE_ALL});
 
     _context = std::make_unique<GenModelTrainContext>(cgen.finish());
-    _context->addTrainCase(
-      uniformTCD<float>({{{1, 3}}, {{2, 1}}},                                     // inputs
-                        {{{2, 1, 5, 5, 2, 1, 5, 5}}, {{2, 1, 5, 5, 2, 1, 5, 6}}}, // expected
-                        // TODO Modify loss values to results of tensorflow
-                        {{14.0124f}, {11.0036f}, {8.1681f}, {6.0974f}} // loss
-                        ));
+    _context->addTrainCase(uniformTCD<float>(
+      {{{0, 1, 2, 3, 4, 5, 1, 3}, {6, 7}}, {{5, 4, 3, 2, 1, 0, 2, 1}, {7, 6}}}, // inputs
+      {{{2, 1, 5, 5, 2, 1, 5, 5}}, {{2, 1, 5, 5, 2, 1, 5, 6}}},                 // expected
+      {{7.3265f}, {4.6811f}, {3.6735f}, {3.2863f}}                              // loss
+      ));
 
     _context->setBackends({"train"});
     _context->setEpoch(4);