diff --git a/PyTorch/built-in/cv/classification/MobileNetV3-Large_ID1784_for_PyTorch/main.py b/PyTorch/built-in/cv/classification/MobileNetV3-Large_ID1784_for_PyTorch/main.py
index f1703a2140000f4e959449fbe7a3fce4f157f368..8904e76d4baf97e68753585bfd2c58fe78bbe53f 100644
--- a/PyTorch/built-in/cv/classification/MobileNetV3-Large_ID1784_for_PyTorch/main.py
+++ b/PyTorch/built-in/cv/classification/MobileNetV3-Large_ID1784_for_PyTorch/main.py
@@ -217,23 +217,14 @@ def main():
     criterion = nn.CrossEntropyLoss().to(device)
 
     # vision optimizer
-    #optimizer = torch.optim.RMSprop(model.parameters(), lr=args.lr, momentum=args.momentum,
-                                    #weight_decay=args.weight_decay, eps=0.0316, alpha=0.9)
-    # prepare for new version, significant improvement
-    optimizer = apex.optimizers.NpuFusedRMSprop(model.parameters(), lr=args.lr, momentum=args.momentum,
-                                    weight_decay=args.weight_decay, eps=0.0316, alpha=0.9)
-
-
     if args.apex:
-        model, optimizer = amp.initialize(model, optimizer,
-                                          opt_level='O2',
-                                          loss_scale=args.loss_scale_value,
-                                          combine_grad=True)
-
-    if args.distributed:
-        model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.device_id])
+        optimizer = apex.optimizers.NpuFusedRMSprop(model.parameters(), lr=args.lr, momentum=args.momentum,
+                                                    weight_decay=args.weight_decay, eps=0.0316, alpha=0.9)
+    else:
+        optimizer = torch.optim.RMSprop(model.parameters(), lr=args.lr, momentum=args.momentum,
+                                        weight_decay=args.weight_decay, eps=0.0316, alpha=0.9)
 
-   # optionally resume from a checkpoint
+    # optionally resume from a checkpoint
     if args.resume:
         if os.path.isfile(args.resume):
             print("=> loading checkpoint '{}'".format(args.resume))
@@ -247,6 +238,15 @@ def main():
         else:
             print("=> no checkpoint found at '{}'".format(args.resume))
 
+    if args.apex:
+        model, optimizer = amp.initialize(model, optimizer,
+                                          opt_level='O2',
+                                          loss_scale=args.loss_scale_value,
+                                          combine_grad=True)
+
+    if args.distributed:
+        model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.device_id])
+
     cudnn.benchmark = True
 
     # Data loading code