diff --git a/test.py b/test.py
index 523c50c..8d252ff 100644
--- a/test.py
+++ b/test.py
@@ -35,7 +35,7 @@ def test(data,
         google_utils.attempt_download(weights)
         model = torch.load(weights, map_location=device)['model'].float()  # load to FP32
         torch_utils.model_info(model)
-        # model.fuse()
+        model.fuse()
         model.to(device)
         if half:
             model.half()  # to FP16
@@ -71,7 +71,7 @@ def test(data,
                                       batch_size,
                                       rect=True,  # rectangular inference
                                       single_cls=opt.single_cls,  # single class mode
-                                      pad=0.0 if fast else 0.5)  # padding
+                                      pad=0.5)  # padding
         batch_size = min(batch_size, len(dataset))
         nw = min([os.cpu_count(), batch_size if batch_size > 1 else 0, 8])  # number of workers
         dataloader = DataLoader(dataset,
diff --git a/utils/utils.py b/utils/utils.py
index bce8a10..22c32e6 100755
--- a/utils/utils.py
+++ b/utils/utils.py
@@ -528,7 +528,7 @@ def non_max_suppression(prediction, conf_thres=0.1, iou_thres=0.6, fast=False, c
     fast |= conf_thres > 0.001  # fast mode
     if fast:
         merge = False
-        multi_label = False
+        multi_label = nc > 1  # multiple labels per box (adds 0.5ms/img)
     else:
         merge = True  # merge for best mAP (adds 0.5ms/img)
         multi_label = nc > 1  # multiple labels per box (adds 0.5ms/img)