Fix metric state reset (#5273)

* Fix metric state reset

* Fix test

* Improve formatting

Co-authored-by: Ananya Harsh Jha <ananya@pytorchlightning.ai>
(cherry picked from commit 4913cbb987)

pytorch_lightning/metrics/metric.py

@@ -94,7 +94,8 @@ class Metric(nn.Module, ABC):
                 reset to this value when ``self.reset()`` is called.
             dist_reduce_fx (Optional): Function to reduce state accross mutliple processes in distributed mode.
                 If value is ``"sum"``, ``"mean"``, or ``"cat"``, we will use ``torch.sum``, ``torch.mean``,
-                and ``torch.cat`` respectively, each with argument ``dim=0``. The user can also pass a custom
+                and ``torch.cat`` respectively, each with argument ``dim=0``. Note that the ``"cat"`` reduction
+                only makes sense if the state is a list, and not a tensor. The user can also pass a custom
                 function in this parameter.
             persistent (Optional): whether the state will be saved as part of the modules ``state_dict``.
                 Default is ``False``.
@@ -244,7 +245,7 @@ class Metric(nn.Module, ABC):
         """
         for attr, default in self._defaults.items():
             current_val = getattr(self, attr)
-            if isinstance(current_val, torch.Tensor):
+            if isinstance(default, torch.Tensor):
                 setattr(self, attr, deepcopy(default).to(current_val.device))
             else:
                 setattr(self, attr, deepcopy(default))

tests/metrics/test_metric.py

@@ -26,6 +26,20 @@ class Dummy(Metric):
         pass
 
 
+class DummyList(Metric):
+    name = "DummyList"
+
+    def __init__(self):
+        super().__init__()
+        self.add_state("x", list(), dist_reduce_fx=None)
+
+    def update(self):
+        pass
+
+    def compute(self):
+        pass
+
+
 def test_inherit():
     Dummy()
 
@@ -77,12 +91,21 @@ def test_reset():
     class A(Dummy):
         pass
 
+    class B(DummyList):
+        pass
+
     a = A()
     assert a.x == 0
     a.x = torch.tensor(5)
     a.reset()
     assert a.x == 0
 
+    b = B()
+    assert isinstance(b.x, list) and len(b.x) == 0
+    b.x = torch.tensor(5)
+    b.reset()
+    assert isinstance(b.x, list) and len(b.x) == 0
+
 
 def test_update():
     class A(Dummy):