fix lint issues

Author: xiaohu2015

references/detection/train.py

@@ -143,7 +143,7 @@ def get_args_parser(add_help=True):
 
     # Prototype models only
     parser.add_argument("--weights", default=None, type=str, help="the weights enum name to load")
-    
+
     # Mixed precision training parameters
     parser.add_argument("--amp", action="store_true", help="Use torch.cuda.amp for mixed precision training")
 
@@ -211,9 +211,9 @@ def main(args):
 
     params = [p for p in model.parameters() if p.requires_grad]
     optimizer = torch.optim.SGD(params, lr=args.lr, momentum=args.momentum, weight_decay=args.weight_decay)
-    
+
     scaler = torch.cuda.amp.GradScaler() if args.amp else None
-    
+
     args.lr_scheduler = args.lr_scheduler.lower()
     if args.lr_scheduler == "multisteplr":
         lr_scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones=args.lr_steps, gamma=args.lr_gamma)