Merge branch 'main' into inverse-scaling-tasks

lm_eval/tasks/mmlu/continuation/mmlu_professional_law.yaml

+"dataset_name": "professional_law"
+"description": "The following are questions (with answers) about professional\
+  \ law.\n\n"
+"group": "mmlu_continuation_humanities"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_professional_law"

lm_eval/tasks/mmlu/continuation/mmlu_professional_medicine.yaml

+"dataset_name": "professional_medicine"
+"description": "The following are questions (with answers) about professional\
+  \ medicine.\n\n"
+"group": "mmlu_continuation_other"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_professional_medicine"

lm_eval/tasks/mmlu/continuation/mmlu_professional_psychology.yaml

+"dataset_name": "professional_psychology"
+"description": "The following are questions (with answers) about professional\
+  \ psychology.\n\n"
+"group": "mmlu_continuation_social_sciences"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_professional_psychology"

lm_eval/tasks/mmlu/continuation/mmlu_public_relations.yaml

+"dataset_name": "public_relations"
+"description": "The following are questions (with answers) about public\
+  \ relations.\n\n"
+"group": "mmlu_continuation_social_sciences"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_public_relations"

lm_eval/tasks/mmlu/continuation/mmlu_security_studies.yaml

+"dataset_name": "security_studies"
+"description": "The following are questions (with answers) about security\
+  \ studies.\n\n"
+"group": "mmlu_continuation_social_sciences"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_security_studies"

lm_eval/tasks/mmlu/continuation/mmlu_sociology.yaml

+"dataset_name": "sociology"
+"description": "The following are questions (with answers) about sociology.\n\
+  \n"
+"group": "mmlu_continuation_social_sciences"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_sociology"

lm_eval/tasks/mmlu/continuation/mmlu_us_foreign_policy.yaml

+"dataset_name": "us_foreign_policy"
+"description": "The following are questions (with answers) about us\
+  \ foreign policy.\n\n"
+"group": "mmlu_continuation_social_sciences"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_us_foreign_policy"

lm_eval/tasks/mmlu/continuation/mmlu_virology.yaml

+"dataset_name": "virology"
+"description": "The following are questions (with answers) about virology.\n\
+  \n"
+"group": "mmlu_continuation_other"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_virology"

lm_eval/tasks/mmlu/continuation/mmlu_world_religions.yaml

+"dataset_name": "world_religions"
+"description": "The following are questions (with answers) about world\
+  \ religions.\n\n"
+"group": "mmlu_continuation_humanities"
+"include": "_continuation_template_yaml"
+"task": "mmlu_continuation_world_religions"

lm_eval/tasks/mmlu/flan_cot_fewshot/_mmlu_flan_cot_fewshot_template_yaml

 dataset_path: hails/mmlu_no_train # a copy of `cais/mmlu` with no auxiliary_train split
 validation_split: validation
 fewshot_split: dev
+fewshot_config:
+  sampler: first_n
 output_type: generate_until
 doc_to_text: "Q: {{question.strip()}}\n(A) {{choices[0]}} (B) {{choices[1]}} (C) {{choices[2]}} (D) {{choices[3]}}\nA: Let's think step by step."
 doc_to_target: "{{['(A)', '(B)', '(C)', '(D)'][answer]}}"
@@ -8,7 +10,7 @@ filter_list:
   - name: "get-answer"
     filter:
       - function: "regex"
-        regex_pattern: "(?<=The answer is )(.*)(?=.)"
+        regex_pattern: "(?<=answer is )(.*)(?=.)"
       - function: "take_first"
 generation_kwargs:
   until:
@@ -23,4 +25,4 @@ metric_list:
     ignore_case: true
     ignore_punctuation: true
 metadata:
-  version: 0.0
+  version: 1.0

lm_eval/tasks/mmlu/flan_cot_zeroshot/_mmlu_flan_cot_zeroshot_template_yaml

@@ -8,7 +8,7 @@ filter_list:
   - name: "strict-match"
     filter:
       - function: "regex"
-        regex_pattern: "((?<=The answer is )(.*)(?=.)|(?<=the answer is )(.*)(?=.)|(?<=The answer: )(.*)(?=.)|(?<=The final answer: )(.*)(?=.))"
+        regex_pattern: "((?<=The answer is )(.*)(?=.)|(?<=answer is )(.*)(?=.)|(?<=The answer: )(.*)(?=.)|(?<=The final answer: )(.*)(?=.))"
       - function: "take_first"
   - name: "flexible-extract"
     filter:
@@ -33,4 +33,4 @@ metric_list:
     ignore_case: true
     ignore_punctuation: true
 metadata:
-  version: 1.0
+  version: 2.0

lm_eval/tasks/mmlu/flan_n_shot/generative/_mmlu_flan_generative_template_yaml

@@ -2,8 +2,10 @@ group: mmlu_flan_n_shot_generative
 dataset_path: hails/mmlu_no_train # a copy of `cais/mmlu` with no auxiliary_train split
 test_split: test
 fewshot_split: dev
+fewshot_config:
+  sampler: first_n
 output_type: generate_until
-doc_to_text: "Q: {{question.strip()}}\n(A) {{choices[0]}} (B) {{choices[1]}} (C) {{choices[2]}} (D) {{choices[3]}}\nA: "
+doc_to_text: "Q: {{question.strip()}}\n(A) {{choices[0]}} (B) {{choices[1]}} (C) {{choices[2]}} (D) {{choices[3]}}\nA:"
 doc_to_target: "{{['(A)', '(B)', '(C)', '(D)'][answer]}}"
 filter_list:
   - name: "strict-match"
@@ -22,9 +24,10 @@ generation_kwargs:
     - "</s>"
     - "Q:"
     - "<|im_end|>"
+    - "\n"
 metric_list:
   - metric: exact_match
     aggregation: mean
     higher_is_better: true
 metadata:
-  version: 1.0
+  version: 2.0

lm_eval/tasks/mmlu/flan_n_shot/loglikelihood/_mmlu_flan_loglikelihood_template_yaml

 dataset_path: hails/mmlu_no_train # a copy of `cais/mmlu` with no auxiliary_train split
 test_split: test
 fewshot_split: dev
+fewshot_config:
+  sampler: first_n
 output_type: multiple_choice
-doc_to_text: "Q: {{question.strip()}}\n(A) {{choices[0]}} (B) {{choices[1]}} (C) {{choices[2]}} (D) {{choices[3]}}\nA: "
+doc_to_text: "Q: {{question.strip()}}\n(A) {{choices[0]}} (B) {{choices[1]}} (C) {{choices[2]}} (D) {{choices[3]}}\nA:"
 doc_to_choice: ["(A)", "(B)", "(C)", "(D)"]
 doc_to_target: answer
 metric_list:
   - metric: acc
     aggregation: mean
     higher_is_better: true
-  - metric: acc_norm
-    aggregation: mean
-    higher_is_better: true
 metadata:
-  version: 0.0
+  version: 1.0

lm_eval/tasks/mmlu/generative/_default_template_yaml

+dataset_path: hails/mmlu_no_train # a copy of `cais/mmlu` with no auxiliary_train split
+test_split: test
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+output_type: generate_until
+doc_to_text: "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:"
+doc_to_target: "{{['A', 'B', 'C', 'D'][answer]}}"
+generation_kwargs:
+  until:
+    - "</s>"
+    - "\n"
+metric_list:
+  - metric: exact_match
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/mmlu/generative/_mmlu.yaml

+group: mmlu_generative
+task:
+  - mmlu_stem_generative
+  - mmlu_other_generative
+  - mmlu_social_sciences_generative
+  - mmlu_humanities_generative

lm_eval/tasks/mmlu/generative/mmlu_abstract_algebra.yaml

+"dataset_name": "abstract_algebra"
+"description": "The following are multiple choice questions (with answers) about abstract\
+  \ algebra.\n\n"
+"group": "mmlu_stem_generative"
+"group_alias": "stem"
+"include": "_default_template_yaml"
+"task": "mmlu_abstract_algebra_generative"
+"task_alias": "abstract_algebra"

lm_eval/tasks/mmlu/generative/mmlu_anatomy.yaml

+"dataset_name": "anatomy"
+"description": "The following are multiple choice questions (with answers) about anatomy.\n\
+  \n"
+"group": "mmlu_stem_generative"
+"group_alias": "stem"
+"include": "_default_template_yaml"
+"task": "mmlu_anatomy_generative"
+"task_alias": "anatomy"

lm_eval/tasks/mmlu/generative/mmlu_astronomy.yaml

+"dataset_name": "astronomy"
+"description": "The following are multiple choice questions (with answers) about astronomy.\n\
+  \n"
+"group": "mmlu_stem_generative"
+"group_alias": "stem"
+"include": "_default_template_yaml"
+"task": "mmlu_astronomy_generative"
+"task_alias": "astronomy"

lm_eval/tasks/mmlu/generative/mmlu_business_ethics.yaml

+"dataset_name": "business_ethics"
+"description": "The following are multiple choice questions (with answers) about business\
+  \ ethics.\n\n"
+"group": "mmlu_other_generative"
+"group_alias": "other"
+"include": "_default_template_yaml"
+"task": "mmlu_business_ethics_generative"
+"task_alias": "business_ethics"

lm_eval/tasks/mmlu/generative/mmlu_clinical_knowledge.yaml

+"dataset_name": "clinical_knowledge"
+"description": "The following are multiple choice questions (with answers) about clinical\
+  \ knowledge.\n\n"
+"group": "mmlu_other_generative"
+"group_alias": "other"
+"include": "_default_template_yaml"
+"task": "mmlu_clinical_knowledge_generative"
+"task_alias": "clinical_knowledge"