YOLOv5 Apple Metal Performance Shader (MPS) support (#7878)

* Apple Metal Performance Shader (MPS) device support

Following https://pytorch.org/blog/introducing-accelerated-pytorch-training-on-mac/

Should work with Apple M1 devices with PyTorch nightly installed with command `--device mps`. Usage examples:
```bash
python train.py --device mps
python detect.py --device mps
python val.py --device mps
```

* Update device strategy to fix MPS issue

models/common.py

@@ -331,7 +331,7 @@ class DetectMultiBackend(nn.Module):
                 names = yaml.safe_load(f)['names']
 
         if pt:  # PyTorch
-            model = attempt_load(weights if isinstance(weights, list) else w, map_location=device)
+            model = attempt_load(weights if isinstance(weights, list) else w, device=device)
             stride = max(int(model.stride.max()), 32)  # model stride
             names = model.module.names if hasattr(model, 'module') else model.names  # get class names
             model.half() if fp16 else model.float()

models/experimental.py

@@ -71,14 +71,14 @@ class Ensemble(nn.ModuleList):
         return y, None  # inference, train output
 
 
-def attempt_load(weights, map_location=None, inplace=True, fuse=True):
+def attempt_load(weights, device=None, inplace=True, fuse=True):
     from models.yolo import Detect, Model
 
     # Loads an ensemble of models weights=[a,b,c] or a single model weights=[a] or weights=a
     model = Ensemble()
     for w in weights if isinstance(weights, list) else [weights]:
-        ckpt = torch.load(attempt_download(w), map_location=map_location)  # load
-        ckpt = (ckpt.get('ema') or ckpt['model']).float()  # FP32 model
+        ckpt = torch.load(attempt_download(w))
+        ckpt = (ckpt.get('ema') or ckpt['model']).to(device).float()  # FP32 model
         model.append(ckpt.fuse().eval() if fuse else ckpt.eval())  # fused or un-fused model in eval mode
 
     # Compatibility updates

models/tf.py

@@ -536,7 +536,7 @@ def run(
 ):
     # PyTorch model
     im = torch.zeros((batch_size, 3, *imgsz))  # BCHW image
-    model = attempt_load(weights, map_location=torch.device('cpu'), inplace=True, fuse=False)
+    model = attempt_load(weights, device=torch.device('cpu'), inplace=True, fuse=False)
     _ = model(im)  # inference
     model.info()