diff --git a/pytorch_lightning/core/step_result.py b/pytorch_lightning/core/step_result.py
index 8d8e578fbc7a8..14ff996aa3949 100644
--- a/pytorch_lightning/core/step_result.py
+++ b/pytorch_lightning/core/step_result.py
@@ -894,6 +894,8 @@ def write_dict(self, predictions_dict, filename='predictions.pt'):
 
 
 def weighted_mean(result, weights):
+    if not isinstance(result, torch.Tensor):
+        result = torch.tensor(result)
     weights = weights.to(result.device)[:result.size(0)]
     numerator = torch.dot(result.float(), weights.transpose(-1, 0).float())
     result = numerator / weights.sum().float()
diff --git a/tests/trainer/logging/test_eval_loop_logging_1_0.py b/tests/trainer/logging/test_eval_loop_logging_1_0.py
index d417f9a2f6ad4..831e59965e1a4 100644
--- a/tests/trainer/logging/test_eval_loop_logging_1_0.py
+++ b/tests/trainer/logging/test_eval_loop_logging_1_0.py
@@ -196,6 +196,40 @@ def validation_epoch_end(self, outputs):
     assert len(trainer.dev_debugger.logged_metrics) == max_epochs
 
 
+def test_eval_float_logging(tmpdir):
+    """
+    Tests that only training_step can be used
+    """
+    os.environ['PL_DEV_DEBUG'] = '1'
+
+    class TestModel(BoringModel):
+
+        def validation_step(self, batch, batch_idx):
+            output = self.layer(batch)
+            loss = self.loss(batch, output)
+            self.log('a', 12.0)
+            return {"x": loss}
+
+    model = TestModel()
+
+    trainer = Trainer(
+        default_root_dir=tmpdir,
+        limit_train_batches=2,
+        limit_val_batches=2,
+        max_epochs=1,
+        row_log_interval=1,
+        weights_summary=None,
+    )
+    trainer.fit(model)
+
+    # make sure all the metrics are available for callbacks
+    logged_metrics = set(trainer.logged_metrics.keys())
+    expected_logged_metrics = {
+        'a',
+    }
+    assert logged_metrics == expected_logged_metrics
+
+
 def test_monitor_val_epoch_end(tmpdir):
     epoch_min_loss_override = 0
     model = SimpleModule()