Added fix to ensure that custom logged metrics within test_epoch_end …

…are appended to the result object even without step reduced metrics (#4251)

pytorch_lightning/trainer/connectors/logger_connector.py

@@ -173,6 +173,9 @@ def _log_on_evaluation_epoch_end_metrics(self, epoch_logs):
         # now we log all of them
         for dl_idx, dl_metrics in enumerate(step_metrics):
             if len(dl_metrics) == 0:
+                # Ensure custom logged metrics are included if not included with step metrics
+                if len(epoch_logger_metrics) > 0:
+                    self.eval_loop_results.append(epoch_logger_metrics)
                 continue
 
             reduced_epoch_metrics = dl_metrics[0].__class__.reduce_on_epoch_end(dl_metrics)

tests/trainer/logging/test_eval_loop_logging_1_0.py

@@ -311,6 +311,39 @@ def validation_epoch_end(self, outputs) -> None:
     assert len(logged_val) == 6
 
 
+@pytest.mark.parametrize(['batches', 'log_interval', 'max_epochs'], [(1, 1, 1), (64, 32, 2)])
+def test_eval_epoch_only_logging(tmpdir, batches, log_interval, max_epochs):
+    """
+    Tests that only test_epoch_end can be used to log, and we return them in the results.
+    """
+    os.environ['PL_DEV_DEBUG'] = '1'
+
+    class TestModel(BoringModel):
+        def test_epoch_end(self, outputs):
+            self.log('c', torch.tensor(2), on_epoch=True, prog_bar=True, logger=True)
+            self.log('d/e/f', 2)
+
+    model = TestModel()
+
+    trainer = Trainer(
+        default_root_dir=tmpdir,
+        limit_train_batches=batches,
+        limit_val_batches=batches,
+        max_epochs=max_epochs,
+        log_every_n_steps=log_interval,
+        weights_summary=None,
+    )
+    trainer.fit(model)
+    results = trainer.test(model)
+
+    expected_result_metrics = {
+        'c': torch.tensor(2),
+        'd/e/f': 2,
+    }
+    for result in results:
+        assert result == expected_result_metrics
+
+
 def test_monitor_val_epoch_end(tmpdir):
     epoch_min_loss_override = 0
     model = SimpleModule()