diff --git a/pytorch_optimizer/optimizer/prodigy.py b/pytorch_optimizer/optimizer/prodigy.py
index ccef49c5..ba0bb7b0 100644
--- a/pytorch_optimizer/optimizer/prodigy.py
+++ b/pytorch_optimizer/optimizer/prodigy.py
@@ -111,7 +111,7 @@ def step(self, closure: CLOSURE = None) -> LOSS:
         if 'd_numerator' not in group:
             group['d_numerator'] = torch.tensor([0.0], device=device)
         elif group['d_numerator'].device != device:
-            group['d_numerator'] = group['d_numerator'].to(device)
+            group['d_numerator'] = group['d_numerator'].to(device)  # pragma: no cover
 
         d_numerator = group['d_numerator']
         d_numerator.mul_(beta3)
diff --git a/tests/test_optimizers.py b/tests/test_optimizers.py
index aaf9024e..b862661f 100644
--- a/tests/test_optimizers.py
+++ b/tests/test_optimizers.py
@@ -76,8 +76,7 @@ def _closure() -> float:
     for _ in range(iterations):
         optimizer.zero_grad()
 
-        y_pred = model(x_data)
-        loss = loss_fn(y_pred, y_data)
+        loss = loss_fn(model(x_data), y_data)
 
         if init_loss == np.inf:
             init_loss = loss