Theskywalker07 commited on
Commit
886da4e
·
verified ·
1 Parent(s): 4ebd02b

Upload full run contents (scripts, checkpoints, and results) (part 9)

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_macroeconomics.yaml +5 -0
  2. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_mathematics.yaml +5 -0
  3. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_microeconomics.yaml +5 -0
  4. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_physics.yaml +5 -0
  5. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_psychology.yaml +5 -0
  6. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_statistics.yaml +5 -0
  7. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_us_history.yaml +5 -0
  8. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_world_history.yaml +5 -0
  9. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_human_aging.yaml +5 -0
  10. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_human_sexuality.yaml +5 -0
  11. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_international_law.yaml +5 -0
  12. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_jurisprudence.yaml +5 -0
  13. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_logical_fallacies.yaml +5 -0
  14. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_machine_learning.yaml +5 -0
  15. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_management.yaml +5 -0
  16. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_marketing.yaml +5 -0
  17. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_medical_genetics.yaml +5 -0
  18. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_miscellaneous.yaml +5 -0
  19. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_moral_disputes.yaml +5 -0
  20. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_moral_scenarios.yaml +5 -0
  21. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_nutrition.yaml +5 -0
  22. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_philosophy.yaml +5 -0
  23. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_prehistory.yaml +5 -0
  24. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_accounting.yaml +5 -0
  25. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_law.yaml +5 -0
  26. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_medicine.yaml +5 -0
  27. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_psychology.yaml +5 -0
  28. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_public_relations.yaml +5 -0
  29. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_security_studies.yaml +5 -0
  30. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_sociology.yaml +5 -0
  31. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_us_foreign_policy.yaml +5 -0
  32. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_virology.yaml +5 -0
  33. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_world_religions.yaml +5 -0
  34. babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/utils.py +73 -0
  35. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si.yaml +11 -0
  36. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_humanities.yaml +8 -0
  37. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_other.yaml +8 -0
  38. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_social_sciences.yaml +8 -0
  39. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_stem.yaml +8 -0
  40. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_si_template_yaml +16 -0
  41. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_abstract_algebra.yaml +5 -0
  42. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_anatomy.yaml +5 -0
  43. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_astronomy.yaml +5 -0
  44. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_business_ethics.yaml +5 -0
  45. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_clinical_knowledge.yaml +5 -0
  46. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_biology.yaml +5 -0
  47. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_chemistry.yaml +5 -0
  48. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_computer_science.yaml +5 -0
  49. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_mathematics.yaml +5 -0
  50. babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_medicine.yaml +5 -0
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_macroeconomics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_macroeconomics
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_high_school_macroeconomics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_mathematics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_mathematics
4
+ tag: global_mmlu_full_ru_stem_tasks
5
+ task: global_mmlu_full_ru_high_school_mathematics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_microeconomics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_microeconomics
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_high_school_microeconomics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_physics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_physics
4
+ tag: global_mmlu_full_ru_stem_tasks
5
+ task: global_mmlu_full_ru_high_school_physics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_psychology.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_psychology
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_high_school_psychology
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_statistics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_statistics
4
+ tag: global_mmlu_full_ru_stem_tasks
5
+ task: global_mmlu_full_ru_high_school_statistics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_us_history.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_us_history
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_high_school_us_history
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_high_school_world_history.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_high_school_world_history
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_high_school_world_history
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_human_aging.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_human_aging
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_human_aging
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_human_sexuality.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_human_sexuality
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_human_sexuality
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_international_law.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_international_law
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_international_law
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_jurisprudence.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_jurisprudence
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_jurisprudence
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_logical_fallacies.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_logical_fallacies
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_logical_fallacies
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_machine_learning.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_machine_learning
4
+ tag: global_mmlu_full_ru_stem_tasks
5
+ task: global_mmlu_full_ru_machine_learning
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_management.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_management
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_management
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_marketing.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_marketing
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_marketing
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_medical_genetics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_medical_genetics
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_medical_genetics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_miscellaneous.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_miscellaneous
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_miscellaneous
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_moral_disputes.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_moral_disputes
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_moral_disputes
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_moral_scenarios.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_moral_scenarios
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_moral_scenarios
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_nutrition.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_nutrition
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_nutrition
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_philosophy.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_philosophy
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_philosophy
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_prehistory.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_prehistory
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_prehistory
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_accounting.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_professional_accounting
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_professional_accounting
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_law.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_professional_law
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_professional_law
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_medicine.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_professional_medicine
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_professional_medicine
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_professional_psychology.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_professional_psychology
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_professional_psychology
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_public_relations.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_public_relations
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_public_relations
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_security_studies.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_security_studies
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_security_studies
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_sociology.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_sociology
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_sociology
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_us_foreign_policy.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_us_foreign_policy
4
+ tag: global_mmlu_full_ru_social_sciences_tasks
5
+ task: global_mmlu_full_ru_us_foreign_policy
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_virology.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_virology
4
+ tag: global_mmlu_full_ru_other_tasks
5
+ task: global_mmlu_full_ru_virology
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/global_mmlu_full_ru_world_religions.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _ru_template_yaml
3
+ process_docs: !function utils.process_world_religions
4
+ tag: global_mmlu_full_ru_humanities_tasks
5
+ task: global_mmlu_full_ru_world_religions
babylm_eval_repo/multilingual/tasks/global_mmlu/full/ru/utils.py ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from functools import partial
2
+
3
+
4
+ SUBJECTS = [
5
+ "abstract_algebra",
6
+ "anatomy",
7
+ "astronomy",
8
+ "business_ethics",
9
+ "clinical_knowledge",
10
+ "college_biology",
11
+ "college_chemistry",
12
+ "college_computer_science",
13
+ "college_mathematics",
14
+ "college_medicine",
15
+ "college_physics",
16
+ "computer_security",
17
+ "conceptual_physics",
18
+ "econometrics",
19
+ "electrical_engineering",
20
+ "elementary_mathematics",
21
+ "formal_logic",
22
+ "global_facts",
23
+ "high_school_biology",
24
+ "high_school_chemistry",
25
+ "high_school_computer_science",
26
+ "high_school_european_history",
27
+ "high_school_geography",
28
+ "high_school_government_and_politics",
29
+ "high_school_macroeconomics",
30
+ "high_school_mathematics",
31
+ "high_school_microeconomics",
32
+ "high_school_physics",
33
+ "high_school_psychology",
34
+ "high_school_statistics",
35
+ "high_school_us_history",
36
+ "high_school_world_history",
37
+ "human_aging",
38
+ "human_sexuality",
39
+ "international_law",
40
+ "jurisprudence",
41
+ "logical_fallacies",
42
+ "machine_learning",
43
+ "management",
44
+ "marketing",
45
+ "medical_genetics",
46
+ "miscellaneous",
47
+ "moral_disputes",
48
+ "moral_scenarios",
49
+ "nutrition",
50
+ "philosophy",
51
+ "prehistory",
52
+ "professional_accounting",
53
+ "professional_law",
54
+ "professional_medicine",
55
+ "professional_psychology",
56
+ "public_relations",
57
+ "security_studies",
58
+ "sociology",
59
+ "us_foreign_policy",
60
+ "virology",
61
+ "world_religions",
62
+ ]
63
+
64
+
65
+ def process_docs(dataset, subject):
66
+ return dataset.filter(lambda x: x["subject"] == subject)
67
+
68
+
69
+ process_functions = {
70
+ f"process_{subject}": partial(process_docs, subject=subject) for subject in SUBJECTS
71
+ }
72
+
73
+ globals().update(process_functions)
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si.yaml ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ group: global_mmlu_full_si
2
+ task:
3
+ - global_mmlu_full_si_stem
4
+ - global_mmlu_full_si_other
5
+ - global_mmlu_full_si_social_sciences
6
+ - global_mmlu_full_si_humanities
7
+ aggregate_metric_list:
8
+ - metric: acc
9
+ weight_by_size: True
10
+ metadata:
11
+ version: 0.0
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_humanities.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ group: global_mmlu_full_si_humanities
2
+ task:
3
+ - global_mmlu_full_si_humanities_tasks
4
+ aggregate_metric_list:
5
+ - metric: acc
6
+ weight_by_size: true
7
+ metadata:
8
+ version: 0.0
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_other.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ group: global_mmlu_full_si_other
2
+ task:
3
+ - global_mmlu_full_si_other_tasks
4
+ aggregate_metric_list:
5
+ - metric: acc
6
+ weight_by_size: true
7
+ metadata:
8
+ version: 0.0
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_social_sciences.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ group: global_mmlu_full_si_social_sciences
2
+ task:
3
+ - global_mmlu_full_si_social_sciences_tasks
4
+ aggregate_metric_list:
5
+ - metric: acc
6
+ weight_by_size: true
7
+ metadata:
8
+ version: 0.0
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_global_mmlu_full_si_stem.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ group: global_mmlu_full_si_stem
2
+ task:
3
+ - global_mmlu_full_si_stem_tasks
4
+ aggregate_metric_list:
5
+ - metric: acc
6
+ weight_by_size: true
7
+ metadata:
8
+ version: 0.0
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/_si_template_yaml ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ dataset_path: CohereForAI/Global-MMLU
2
+ dataset_name: si
3
+ test_split: test
4
+ fewshot_split: dev
5
+ fewshot_config:
6
+ sampler: first_n
7
+ output_type: multiple_choice
8
+ doc_to_text: "{{question.strip()}}\nA. {{option_a}}\nB. {{option_b}}\nC. {{option_c}}\nD. {{option_d}}\nAnswer:"
9
+ doc_to_choice: ["A", "B", "C", "D"]
10
+ doc_to_target: answer
11
+ metric_list:
12
+ - metric: acc
13
+ aggregation: mean
14
+ higher_is_better: true
15
+ metadata:
16
+ version: 0.0
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_abstract_algebra.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_abstract_algebra
4
+ tag: global_mmlu_full_si_stem_tasks
5
+ task: global_mmlu_full_si_abstract_algebra
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_anatomy.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_anatomy
4
+ tag: global_mmlu_full_si_stem_tasks
5
+ task: global_mmlu_full_si_anatomy
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_astronomy.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_astronomy
4
+ tag: global_mmlu_full_si_stem_tasks
5
+ task: global_mmlu_full_si_astronomy
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_business_ethics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_business_ethics
4
+ tag: global_mmlu_full_si_other_tasks
5
+ task: global_mmlu_full_si_business_ethics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_clinical_knowledge.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_clinical_knowledge
4
+ tag: global_mmlu_full_si_other_tasks
5
+ task: global_mmlu_full_si_clinical_knowledge
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_biology.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_college_biology
4
+ tag: global_mmlu_full_si_stem_tasks
5
+ task: global_mmlu_full_si_college_biology
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_chemistry.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_college_chemistry
4
+ tag: global_mmlu_full_si_stem_tasks
5
+ task: global_mmlu_full_si_college_chemistry
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_computer_science.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_college_computer_science
4
+ tag: global_mmlu_full_si_stem_tasks
5
+ task: global_mmlu_full_si_college_computer_science
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_mathematics.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_college_mathematics
4
+ tag: global_mmlu_full_si_stem_tasks
5
+ task: global_mmlu_full_si_college_mathematics
babylm_eval_repo/multilingual/tasks/global_mmlu/full/si/global_mmlu_full_si_college_medicine.yaml ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ # Generated by _generate_configs.py
2
+ include: _si_template_yaml
3
+ process_docs: !function utils.process_college_medicine
4
+ tag: global_mmlu_full_si_other_tasks
5
+ task: global_mmlu_full_si_college_medicine