Merge branch 'main' into prefill

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_astronomy_light.yaml

+task: arabic_leaderboard_arabic_mmlu_astronomy_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: astronomy
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_business_ethics_light.yaml

+task: arabic_leaderboard_arabic_mmlu_business_ethics_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: business_ethics
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_clinical_knowledge_light.yaml

+task: arabic_leaderboard_arabic_mmlu_clinical_knowledge_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: clinical_knowledge
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_college_biology_light.yaml

+task: arabic_leaderboard_arabic_mmlu_college_biology_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: college_biology
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_college_chemistry_light.yaml

+task: arabic_leaderboard_arabic_mmlu_college_chemistry_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: college_chemistry
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_college_computer_science_light.yaml

+task: arabic_leaderboard_arabic_mmlu_college_computer_science_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: college_computer_science
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_college_mathematics_light.yaml

+task: arabic_leaderboard_arabic_mmlu_college_mathematics_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: college_mathematics
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_college_medicine_light.yaml

+task: arabic_leaderboard_arabic_mmlu_college_medicine_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: college_medicine
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_college_physics_light.yaml

+task: arabic_leaderboard_arabic_mmlu_college_physics_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: college_physics
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_computer_security_light.yaml

+task: arabic_leaderboard_arabic_mmlu_computer_security_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: computer_security
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_conceptual_physics_light.yaml

+task: arabic_leaderboard_arabic_mmlu_conceptual_physics_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: conceptual_physics
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_econometrics_light.yaml

+task: arabic_leaderboard_arabic_mmlu_econometrics_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: econometrics
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_electrical_engineering_light.yaml

+task: arabic_leaderboard_arabic_mmlu_electrical_engineering_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: electrical_engineering
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_elementary_mathematics_light.yaml

+task: arabic_leaderboard_arabic_mmlu_elementary_mathematics_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: elementary_mathematics
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_formal_logic_light.yaml

+task: arabic_leaderboard_arabic_mmlu_formal_logic_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: formal_logic
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_global_facts_light.yaml

+task: arabic_leaderboard_arabic_mmlu_global_facts_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: global_facts
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_high_school_biology_light.yaml

+task: arabic_leaderboard_arabic_mmlu_high_school_biology_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: high_school_biology
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_high_school_chemistry_light.yaml

+task: arabic_leaderboard_arabic_mmlu_high_school_chemistry_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: high_school_chemistry
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_high_school_computer_science_light.yaml

+task: arabic_leaderboard_arabic_mmlu_high_school_computer_science_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: high_school_computer_science
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/arabic_leaderboard_arabic_mmlu_light/arabic_leaderboard_arabic_mmlu_high_school_european_history_light.yaml

+task: arabic_leaderboard_arabic_mmlu_high_school_european_history_light
+dataset_path: arcee-globe/Arabic_MMLU-10percent
+dataset_name: high_school_european_history
+output_type: multiple_choice
+training_split: null
+validation_split: dev
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: dev
+fewshot_config:
+  sampler: first_n
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+metadata:
+  version: 1.0