update all mmlu variants

lm_eval/tasks/mmlu/continuation/_mmlu.yaml

 group: mmlu_continuation
+group_alias: mmlu (continuation)
 task:
-  - mmlu_continuation_stem
-  - mmlu_continuation_other
-  - mmlu_continuation_social_sciences
-  - mmlu_continuation_humanities
+  - group: stem
+    task:
+      - mmlu_continuation_stem
+    aggregate_metric: True
+    weight_by_size: True
+  - group: other
+    task:
+      - mmlu_continuation_other
+    aggregate_metric: True
+    weight_by_size: True
+  - group: social sciences
+    task:
+      - mmlu_continuation_social_sciences
+    aggregate_metric: True
+    weight_by_size: True
+  - group: humanities
+    task:
+      - mmlu_continuation_humanities
+    aggregate_metric: True
+    weight_by_size: True
+aggregate_metric: True
+weight_by_size: True

lm_eval/tasks/mmlu/flan_cot_fewshot/_mmlu.yaml

 group: mmlu_flan_cot_fewshot
+group_alias: mmlu (flan style, fewshot cot)
 task:
-  - mmlu_flan_cot_fewshot_stem
-  - mmlu_flan_cot_fewshot_other
-  - mmlu_flan_cot_fewshot_social_sciences
-  - mmlu_flan_cot_fewshot_humanities
-group_config:
-  aggregate_metric: True
-  weight_by_size: True
+  - group: stem
+    task:
+      - mmlu_flan_cot_fewshot_stem
+    aggregate_metric: True
+    weight_by_size: True
+  - group: other
+    task:
+      - mmlu_flan_cot_fewshot_other
+    aggregate_metric: True
+    weight_by_size: True
+  - group: social sciences
+    task:
+      - mmlu_flan_cot_fewshot_social_sciences
+    aggregate_metric: True
+    weight_by_size: True
+  - group: humanities
+    task:
+      - mmlu_flan_cot_fewshot_humanities
+    aggregate_metric: True
+    weight_by_size: True
+aggregate_metric: True
+weight_by_size: True

lm_eval/tasks/mmlu/flan_cot_zeroshot/_mmlu.yaml

 group: mmlu_flan_cot_zeroshot
+group_alias: mmlu (flan style, zeroshot cot)
 task:
-  - mmlu_flan_cot_zeroshot_stem
-  - mmlu_flan_cot_zeroshot_other
-  - mmlu_flan_cot_zeroshot_social_sciences
-  - mmlu_flan_cot_zeroshot_humanities
-group_config:
-  aggregate_metric: True
-  weight_by_size: True
+  - group: stem
+    task:
+      - mmlu_flan_cot_zeroshot_stem
+    aggregate_metric: True
+    weight_by_size: True
+  - group: other
+    task:
+      - mmlu_flan_cot_zeroshot_other
+    aggregate_metric: True
+    weight_by_size: True
+  - group: social sciences
+    task:
+      - mmlu_flan_cot_zeroshot_social_sciences
+    aggregate_metric: True
+    weight_by_size: True
+  - group: humanities
+    task:
+      - mmlu_flan_cot_zeroshot_humanities
+    aggregate_metric: True
+    weight_by_size: True
+aggregate_metric: True
+weight_by_size: True

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

 group: mmlu_flan_n_shot_generative
+group_alias: mmlu (flan style, generative)
 task:
-  - mmlu_flan_n_shot_generative_stem
-  - mmlu_flan_n_shot_generative_other
-  - mmlu_flan_n_shot_generative_social_sciences
-  - mmlu_flan_n_shot_generative_humanities
-group_config:
-  aggregate_metric: True
-  weight_by_size: True
+  - group: stem
+    task:
+      - mmlu_flan_n_shot_generative_stem
+    aggregate_metric: True
+    weight_by_size: True
+  - group: other
+    task:
+      - mmlu_flan_n_shot_generative_other
+    aggregate_metric: True
+    weight_by_size: True
+  - group: social sciences
+    task:
+      - mmlu_flan_n_shot_generative_social_sciences
+    aggregate_metric: True
+    weight_by_size: True
+  - group: humanities
+    task:
+      - mmlu_flan_n_shot_generative_humanities
+    aggregate_metric: True
+    weight_by_size: True
+aggregate_metric: True
+weight_by_size: True

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

 group: mmlu_flan_n_shot_loglikelihood
-group_alias: mmlu (flan style)
+group_alias: mmlu (flan style, loglikelihood)
 task:
   - group: stem
     task:
@@ -16,7 +16,7 @@ task:
       - mmlu_flan_n_shot_loglikelihood_social_sciences
     aggregate_metric: True
     weight_by_size: True
-  - group: social sciences
+  - group: humanities
     task:
       - mmlu_flan_n_shot_loglikelihood_humanities
     aggregate_metric: True

lm_eval/tasks/mmlu/generative/_mmlu.yaml

 group: mmlu_generative
+group_alias: mmlu (generative)
 task:
-  - mmlu_stem_generative
-  - mmlu_other_generative
-  - mmlu_social_sciences_generative
-  - mmlu_humanities_generative
+  - group: stem
+    task:
+      - mmlu_stem_generative
+    aggregate_metric: True
+    weight_by_size: True
+  - group: other
+    task:
+      - mmlu_other_generative
+    aggregate_metric: True
+    weight_by_size: True
+  - group: social sciences
+    task:
+      - mmlu_social_sciences_generative
+    aggregate_metric: True
+    weight_by_size: True
+  - group: humanities
+    task:
+      - mmlu_humanities_generative
+    aggregate_metric: True
+    weight_by_size: True
+aggregate_metric: True
+weight_by_size: True