Merge branch 'main' into humaneval

# Conflicts:
#	lm_eval/api/task.py

lm_eval/tasks/arabic_leaderboard_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Mauritania.yaml

+task: arabic_leaderboard_acva_Mauritania
+dataset_path: OALL/ACVA
+dataset_name: Mauritania
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Mesopotamia_civilization.yaml

+task: arabic_leaderboard_acva_Mesopotamia_civilization
+dataset_path: OALL/ACVA
+dataset_name: Mesopotamia_civilization
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Morocco.yaml

+task: arabic_leaderboard_acva_Morocco
+dataset_path: OALL/ACVA
+dataset_name: Morocco
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Oman.yaml

+task: arabic_leaderboard_acva_Oman
+dataset_path: OALL/ACVA
+dataset_name: Oman
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Palestine.yaml

+task: arabic_leaderboard_acva_Palestine
+dataset_path: OALL/ACVA
+dataset_name: Palestine
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Qatar.yaml

+task: arabic_leaderboard_acva_Qatar
+dataset_path: OALL/ACVA
+dataset_name: Qatar
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Saudi_Arabia.yaml

+task: arabic_leaderboard_acva_Saudi_Arabia
+dataset_path: OALL/ACVA
+dataset_name: Saudi_Arabia
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Somalia.yaml

+task: arabic_leaderboard_acva_Somalia
+dataset_path: OALL/ACVA
+dataset_name: Somalia
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Sudan.yaml

+task: arabic_leaderboard_acva_Sudan
+dataset_path: OALL/ACVA
+dataset_name: Sudan
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Syria.yaml

+task: arabic_leaderboard_acva_Syria
+dataset_path: OALL/ACVA
+dataset_name: Syria
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Tunisia.yaml

+task: arabic_leaderboard_acva_Tunisia
+dataset_path: OALL/ACVA
+dataset_name: Tunisia
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_United_Arab_Emirates.yaml

+task: arabic_leaderboard_acva_United_Arab_Emirates
+dataset_path: OALL/ACVA
+dataset_name: United_Arab_Emirates
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_Yemen.yaml

+task: arabic_leaderboard_acva_Yemen
+dataset_path: OALL/ACVA
+dataset_name: Yemen
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_communication.yaml

+task: arabic_leaderboard_acva_communication
+dataset_path: OALL/ACVA
+dataset_name: communication
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_computer_and_phone.yaml

+task: arabic_leaderboard_acva_computer_and_phone
+dataset_path: OALL/ACVA
+dataset_name: computer_and_phone
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_daily_life.yaml

+task: arabic_leaderboard_acva_daily_life
+dataset_path: OALL/ACVA
+dataset_name: daily_life
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/arabic_leaderboard_acva_entertainment.yaml

+task: arabic_leaderboard_acva_entertainment
+dataset_path: OALL/ACVA
+dataset_name: entertainment
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: test
+process_docs: !function utils.process_docs
+doc_to_text: "{{query}}"
+doc_to_target: "{{gold}}"
+doc_to_choice: "choices"
+fewshot_split: validation
+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_complete/arabic_leaderboard_avca/utils.py

+import datasets
+import numpy as np
+
+
+def process_docs(dataset: datasets.Dataset):
+    def _process_doc(doc):
+        question = doc["question"]
+        answer = doc["answer"]
+
+        return {
+            "query": f"السؤال: {question}\nالإجابة:",
+            "choices": ["صح", "خطأ"],
+            "gold": ["صح", "خطأ"].index(answer),
+        }
+
+    return dataset.map(_process_doc)

lm_eval/tasks/arabic_leaderboard_complete/arabic_leaderboard_complete.yaml

+group: arabic_leaderboard_complete
+task:
+  - arabic_leaderboard_acva
+  - arabic_leaderboard_alghafa
+  - arabic_leaderboard_arabic_exams
+  - arabic_leaderboard_arabic_mt_arc_challenge
+  - arabic_leaderboard_arabic_mt_arc_easy
+  - arabic_leaderboard_arabic_mt_boolq
+  - arabic_leaderboard_arabic_mt_hellaswag
+  - arabic_leaderboard_arabic_mt_mmlu
+  - arabic_leaderboard_arabic_mt_copa
+  - arabic_leaderboard_arabic_mt_openbook_qa
+  - arabic_leaderboard_arabic_mt_piqa
+  - arabic_leaderboard_arabic_mt_race
+  - arabic_leaderboard_arabic_mt_sciq
+  - arabic_leaderboard_arabic_mt_toxigen
+aggregate_metric_list:
+  - metric: acc
+    aggregation: mean
+    weight_by_size: true
+  - metric: acc_norm
+    aggregation: mean
+    weight_by_size: true
+metadata:
+  version: 1.0

lm_eval/tasks/arabic_leaderboard_light/README.md

+# Arabic Leaderboard Light
+
+Title: Open Arabic LLM Leaderboard Light
+
+This leaderboard follows all the details as in [`arabic_leaderboard_complete`](../arabic_leaderboard_complete), except that a light version - 10% random sample of the test set of each benchmark - is used to test the language models.
+
+NOTE: In ACVA benchmark, there is Yemen subset, and it is a small dataset - it has only 10 samples in the test split. So, for this specific subset dataset, to have more reliable results, we consider the original dataset, instead of 10% of its test samples.
+
+### Checklist
+
+For adding novel benchmarks/datasets to the library:
+* [ ] Is the task an existing benchmark in the literature?
+  * [ ] Have you referenced the original paper that introduced the task?
+  * [ ] If yes, does the original paper provide a reference implementation? If so, have you checked against the reference implementation and documented how to run such a test?
+
+
+If other tasks on this dataset are already supported:
+* [ ] Is the "Main" variant of this task clearly denoted?
+* [ ] Have you provided a short sentence in a README on what each new variant adds / evaluates?
+* [ ] Have you noted which, if any, published evaluation setups are matched by this variant?