Merge pull request #922 from EleutherAI/mmlu_subgroups

[Refactor] Mmlu subgroups and weight avg

lm_eval/tasks/mmlu/default/mmlu_international_law.yaml

 "dataset_name": "international_law"
-"description": "The following are multiple choice questions (with answers) about international law.\n\n"
+"description": "The following are multiple choice questions (with answers) about international\
+  \ law.\n\n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_international_law"
+"task_alias": "international_law"

lm_eval/tasks/mmlu/default/mmlu_jurisprudence.yaml

 "dataset_name": "jurisprudence"
-"description": "The following are multiple choice questions (with answers) about jurisprudence.\n\n"
+"description": "The following are multiple choice questions (with answers) about jurisprudence.\n\
+  \n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_jurisprudence"
+"task_alias": "jurisprudence"

lm_eval/tasks/mmlu/default/mmlu_logical_fallacies.yaml

 "dataset_name": "logical_fallacies"
-"description": "The following are multiple choice questions (with answers) about logical fallacies.\n\n"
+"description": "The following are multiple choice questions (with answers) about logical\
+  \ fallacies.\n\n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_logical_fallacies"
+"task_alias": "logical_fallacies"

lm_eval/tasks/mmlu/default/mmlu_machine_learning.yaml

 "dataset_name": "machine_learning"
-"description": "The following are multiple choice questions (with answers) about machine learning.\n\n"
+"description": "The following are multiple choice questions (with answers) about machine\
+  \ learning.\n\n"
+"group": "mmlu_stem"
+"group_alias": "stem"
 "include": "_default_template_yaml"
 "task": "mmlu_machine_learning"
+"task_alias": "machine_learning"

lm_eval/tasks/mmlu/default/mmlu_management.yaml

 "dataset_name": "management"
-"description": "The following are multiple choice questions (with answers) about management.\n\n"
+"description": "The following are multiple choice questions (with answers) about management.\n\
+  \n"
+"group": "mmlu_other"
+"group_alias": "other"
 "include": "_default_template_yaml"
 "task": "mmlu_management"
+"task_alias": "management"

lm_eval/tasks/mmlu/default/mmlu_marketing.yaml

 "dataset_name": "marketing"
-"description": "The following are multiple choice questions (with answers) about marketing.\n\n"
+"description": "The following are multiple choice questions (with answers) about marketing.\n\
+  \n"
+"group": "mmlu_other"
+"group_alias": "other"
 "include": "_default_template_yaml"
 "task": "mmlu_marketing"
+"task_alias": "marketing"

lm_eval/tasks/mmlu/default/mmlu_medical_genetics.yaml

 "dataset_name": "medical_genetics"
-"description": "The following are multiple choice questions (with answers) about medical genetics.\n\n"
+"description": "The following are multiple choice questions (with answers) about medical\
+  \ genetics.\n\n"
+"group": "mmlu_other"
+"group_alias": "other"
 "include": "_default_template_yaml"
 "task": "mmlu_medical_genetics"
+"task_alias": "medical_genetics"

lm_eval/tasks/mmlu/default/mmlu_miscellaneous.yaml

 "dataset_name": "miscellaneous"
-"description": "The following are multiple choice questions (with answers) about miscellaneous.\n\n"
+"description": "The following are multiple choice questions (with answers) about miscellaneous.\n\
+  \n"
+"group": "mmlu_other"
+"group_alias": "other"
 "include": "_default_template_yaml"
 "task": "mmlu_miscellaneous"
+"task_alias": "miscellaneous"

lm_eval/tasks/mmlu/default/mmlu_moral_disputes.yaml

 "dataset_name": "moral_disputes"
-"description": "The following are multiple choice questions (with answers) about moral disputes.\n\n"
+"description": "The following are multiple choice questions (with answers) about moral\
+  \ disputes.\n\n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_moral_disputes"
+"task_alias": "moral_disputes"

lm_eval/tasks/mmlu/default/mmlu_moral_scenarios.yaml

 "dataset_name": "moral_scenarios"
-"description": "The following are multiple choice questions (with answers) about moral scenarios.\n\n"
+"description": "The following are multiple choice questions (with answers) about moral\
+  \ scenarios.\n\n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_moral_scenarios"
+"task_alias": "moral_scenarios"

lm_eval/tasks/mmlu/default/mmlu_nutrition.yaml

 "dataset_name": "nutrition"
-"description": "The following are multiple choice questions (with answers) about nutrition.\n\n"
+"description": "The following are multiple choice questions (with answers) about nutrition.\n\
+  \n"
+"group": "mmlu_other"
+"group_alias": "other"
 "include": "_default_template_yaml"
 "task": "mmlu_nutrition"
+"task_alias": "nutrition"

lm_eval/tasks/mmlu/default/mmlu_philosophy.yaml

 "dataset_name": "philosophy"
-"description": "The following are multiple choice questions (with answers) about philosophy.\n\n"
+"description": "The following are multiple choice questions (with answers) about philosophy.\n\
+  \n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_philosophy"
+"task_alias": "philosophy"

lm_eval/tasks/mmlu/default/mmlu_prehistory.yaml

 "dataset_name": "prehistory"
-"description": "The following are multiple choice questions (with answers) about prehistory.\n\n"
+"description": "The following are multiple choice questions (with answers) about prehistory.\n\
+  \n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_prehistory"
+"task_alias": "prehistory"

lm_eval/tasks/mmlu/default/mmlu_professional_accounting.yaml

 "dataset_name": "professional_accounting"
-"description": "The following are multiple choice questions (with answers) about professional accounting.\n\n"
+"description": "The following are multiple choice questions (with answers) about professional\
+  \ accounting.\n\n"
+"group": "mmlu_other"
+"group_alias": "other"
 "include": "_default_template_yaml"
 "task": "mmlu_professional_accounting"
+"task_alias": "professional_accounting"

lm_eval/tasks/mmlu/default/mmlu_professional_law.yaml

 "dataset_name": "professional_law"
-"description": "The following are multiple choice questions (with answers) about professional law.\n\n"
+"description": "The following are multiple choice questions (with answers) about professional\
+  \ law.\n\n"
+"group": "mmlu_humanities"
+"group_alias": "humanities"
 "include": "_default_template_yaml"
 "task": "mmlu_professional_law"
+"task_alias": "professional_law"

lm_eval/tasks/mmlu/default/mmlu_professional_medicine.yaml

 "dataset_name": "professional_medicine"
-"description": "The following are multiple choice questions (with answers) about professional medicine.\n\n"
+"description": "The following are multiple choice questions (with answers) about professional\
+  \ medicine.\n\n"
+"group": "mmlu_other"
+"group_alias": "other"
 "include": "_default_template_yaml"
 "task": "mmlu_professional_medicine"
+"task_alias": "professional_medicine"

lm_eval/tasks/mmlu/default/mmlu_professional_psychology.yaml

 "dataset_name": "professional_psychology"
-"description": "The following are multiple choice questions (with answers) about professional psychology.\n\n"
+"description": "The following are multiple choice questions (with answers) about professional\
+  \ psychology.\n\n"
+"group": "mmlu_social_sciences"
+"group_alias": "social_sciences"
 "include": "_default_template_yaml"
 "task": "mmlu_professional_psychology"
+"task_alias": "professional_psychology"

lm_eval/tasks/mmlu/default/mmlu_public_relations.yaml

 "dataset_name": "public_relations"
-"description": "The following are multiple choice questions (with answers) about public relations.\n\n"
+"description": "The following are multiple choice questions (with answers) about public\
+  \ relations.\n\n"
+"group": "mmlu_social_sciences"
+"group_alias": "social_sciences"
 "include": "_default_template_yaml"
 "task": "mmlu_public_relations"
+"task_alias": "public_relations"

lm_eval/tasks/mmlu/default/mmlu_security_studies.yaml

 "dataset_name": "security_studies"
-"description": "The following are multiple choice questions (with answers) about security studies.\n\n"
+"description": "The following are multiple choice questions (with answers) about security\
+  \ studies.\n\n"
+"group": "mmlu_social_sciences"
+"group_alias": "social_sciences"
 "include": "_default_template_yaml"
 "task": "mmlu_security_studies"
+"task_alias": "security_studies"

lm_eval/tasks/mmlu/default/mmlu_sociology.yaml

 "dataset_name": "sociology"
-"description": "The following are multiple choice questions (with answers) about sociology.\n\n"
+"description": "The following are multiple choice questions (with answers) about sociology.\n\
+  \n"
+"group": "mmlu_social_sciences"
+"group_alias": "social_sciences"
 "include": "_default_template_yaml"
 "task": "mmlu_sociology"
+"task_alias": "sociology"