formatting

lm_eval/tasks/__init__.py

@@ -196,9 +196,13 @@ class TaskManager:
                     if self._name_is_tag(name_or_config):
                         fn = partial(
                             self._load_individual_task_or_group,
-                            update_config=name_or_config if isinstance(name_or_config, dict) else None,
+                            update_config=name_or_config
+                            if isinstance(name_or_config, dict)
+                            else None,
+                        )
+                        return dict(
+                            collections.ChainMap(*map(fn, reversed(subtask_list)))
                         )
-                        return dict(collections.ChainMap(*map(fn, reversed(subtask_list))))
                     else:
                         group_name = ConfigurableGroup(
                             config={"group": name_or_config, "task": subtask_list}

lm_eval/tasks/mmlu/flan_n_shot/loglikelihood/_mmlu.yaml

 group: mmlu_flan_n_shot_loglikelihood
+group_alias: mmlu (flan style)
 task:
+  - group: stem
+    task:
       - mmlu_flan_n_shot_loglikelihood_stem
+    aggregate_metric: True
+    weight_by_size: True
+  - group: other
+    task:
       - mmlu_flan_n_shot_loglikelihood_other
+    aggregate_metric: True
+    weight_by_size: True
+  - group: social sciences
+    task:
       - mmlu_flan_n_shot_loglikelihood_social_sciences
+    aggregate_metric: True
+    weight_by_size: True
+  - group: social sciences
+    task:
       - mmlu_flan_n_shot_loglikelihood_humanities
-group_config:
     aggregate_metric: True
     weight_by_size: True
+aggregate_metric: True
+weight_by_size: True