Fixes #3276 (#4116)

pytorch_lightning/core/step_result.py

@@ -476,6 +476,9 @@ class Result(Dict):
             else:
                 tbptt_reduce_fx = meta[k]['tbptt_reduce_fx']
 
+            if isinstance(value, list):
+                value = torch.tensor(value)
+
             if isinstance(value, dict):
                 # TODO: recursive reduce:
                 _recursive_fx_apply(value, tbptt_reduce_fx)

tests/trainer/logging/test_eval_loop_logging_1_0.py

@@ -21,6 +21,8 @@ def test__validation_step__log(tmpdir):
             acc = self.step(batch, batch_idx)
             acc = acc + batch_idx
             self.log('a', acc, on_step=True, on_epoch=True)
+            self.log('a2', 2)
+
             self.training_step_called = True
             return acc
 
@@ -50,6 +52,7 @@ def test__validation_step__log(tmpdir):
     # make sure all the metrics are available for callbacks
     expected_logged_metrics = {
         'a',
+        'a2',
         'a_step',
         'a_epoch',
         'b',
@@ -65,7 +68,7 @@ def test__validation_step__log(tmpdir):
     # on purpose DO NOT allow step_b... it's silly to monitor val step metrics
     callback_metrics = set(trainer.callback_metrics.keys())
     callback_metrics.remove('debug_epoch')
-    expected_cb_metrics = {'a', 'b', 'a_epoch', 'b_epoch', 'a_step'}
+    expected_cb_metrics = {'a', 'a2', 'b', 'a_epoch', 'b_epoch', 'a_step'}
     assert expected_cb_metrics == callback_metrics