use cosine lr decay for RT-DETR-R50

engine.py

@@ -1322,16 +1322,18 @@ class RTDetrTrainer(object):
                 targets = self.box_xyxy_to_cxcywh(targets)
 
             # Inference
-            with torch.cuda.amp.autocast(enabled=self.args.fp16):
-                outputs = model(images, targets)
-                # Compute loss
+            with torch.autocast(device_type=str(self.device), cache_enabled=True):
+                outputs = model(images, targets)    
+
+            # Compute loss
+            with torch.autocast(device_type=str(self.device), enabled=False):
                 loss_dict = self.criterion(outputs, targets)
-                losses = sum(loss_dict.values())
-                # Grad Accumulate
-                if self.grad_accumulate > 1:
-                    losses /= self.grad_accumulate
+            losses = sum(loss_dict.values())
+            # Grad Accumulate
+            if self.grad_accumulate > 1:
+                losses /= self.grad_accumulate
 
-                loss_dict_reduced = distributed_utils.reduce_dict(loss_dict)
+            loss_dict_reduced = distributed_utils.reduce_dict(loss_dict)
 
             # Backward
             self.scaler.scale(losses).backward()