diff --git a/hubconf.py b/hubconf.py
index 5bb629005597..011eaa57ff34 100644
--- a/hubconf.py
+++ b/hubconf.py
@@ -29,6 +29,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo
     from pathlib import Path
 
     from models.common import AutoShape, DetectMultiBackend
+    from models.experimental import attempt_load
     from models.yolo import Model
     from utils.downloads import attempt_download
     from utils.general import LOGGER, check_requirements, intersect_dicts, logging
@@ -42,8 +43,12 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo
     try:
         device = select_device(device)
         if pretrained and channels == 3 and classes == 80:
-            model = DetectMultiBackend(path, device=device, fuse=autoshape)  # download/load FP32 model
-            # model = models.experimental.attempt_load(path, map_location=device)  # download/load FP32 model
+            try:
+                model = DetectMultiBackend(path, device=device, fuse=autoshape)  # detection model
+                if autoshape:
+                    model = AutoShape(model)  # for file/URI/PIL/cv2/np inputs and NMS
+            except Exception:
+                model = attempt_load(path, device=device, fuse=False)  # arbitrary model
         else:
             cfg = list((Path(__file__).parent / 'models').rglob(f'{path.stem}.yaml'))[0]  # model.yaml path
             model = Model(cfg, channels, classes)  # create model
@@ -54,9 +59,6 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo
                 model.load_state_dict(csd, strict=False)  # load
                 if len(ckpt['model'].names) == classes:
                     model.names = ckpt['model'].names  # set class names attribute
-        if autoshape:
-            model.model.model[-1].inplace = False  # Detect.inplace=False for safe multithread inference
-            model = AutoShape(model)  # for file/URI/PIL/cv2/np inputs and NMS
         if not verbose:
             LOGGER.setLevel(logging.INFO)  # reset to default
         return model.to(device)
diff --git a/models/common.py b/models/common.py
index 959c965e6002..c898d94a921a 100644
--- a/models/common.py
+++ b/models/common.py
@@ -562,6 +562,9 @@ def __init__(self, model, verbose=True):
         self.dmb = isinstance(model, DetectMultiBackend)  # DetectMultiBackend() instance
         self.pt = not self.dmb or model.pt  # PyTorch model
         self.model = model.eval()
+        if self.pt:
+            m = self.model.model.model[-1] if self.dmb else self.model.model[-1]  # Detect()
+            m.inplace = False  # Detect.inplace=False for safe multithread inference
 
     def _apply(self, fn):
         # Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers