spacing

lm_eval/api/task.py

@@ -1381,9 +1381,9 @@ class ConfigurableTask(Task):
                             )
                     except TypeError as error:  # needed for now in order to use a different interface between our own metrics and HF Evaluate metrics
                         result_score = self._metric_fn_list[metric]([gold, result])
-                if isinstance(result_score, dict):
-                    # TODO: this handles the case where HF evaluate returns a dict.
-                    result_score = result_score[metric]
+                    if isinstance(result_score, dict):
+                        # TODO: this handles the case where HF evaluate returns a dict.
+                        result_score = result_score[metric]
                 result_dict[metric] = result_score
         else:
             raise ValueError(