Merge branch 'main' into mathvista

# Conflicts:
#	lm_eval/models/openai_completions.py

lm_eval/tasks/aradice/piqa/piqa_egy.yaml

+task: AraDiCE_piqa_egy
+dataset_path: QCRI/AraDiCE-PIQA
+dataset_name: PIQA-egy
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: "سؤال : {{goal}}\nإجابة :"
+doc_to_target: label
+doc_to_choice: "{{[sol1, sol2]}}"
+should_decontaminate: true
+doc_to_decontamination_query: goal
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/piqa/piqa_eng.yaml

+task: AraDiCE_piqa_eng
+dataset_path: QCRI/AraDiCE-PIQA
+dataset_name: PIQA-eng
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: "سؤال : {{goal}}\nإجابة :"
+doc_to_target: label
+doc_to_choice: "{{[sol1, sol2]}}"
+should_decontaminate: true
+doc_to_decontamination_query: goal
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/piqa/piqa_lev.yaml

+task: AraDiCE_piqa_lev
+dataset_path: QCRI/AraDiCE-PIQA
+dataset_name: PIQA-lev
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: "سؤال : {{goal}}\nإجابة :"
+doc_to_target: label
+doc_to_choice: "{{[sol1, sol2]}}"
+should_decontaminate: true
+doc_to_decontamination_query: goal
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/piqa/piqa_msa.yaml

+task: AraDiCE_piqa_msa
+dataset_path: QCRI/AraDiCE-PIQA
+dataset_name: PIQA-msa
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: "سؤال : {{goal}}\nإجابة :"
+doc_to_target: label
+doc_to_choice: "{{[sol1, sol2]}}"
+should_decontaminate: true
+doc_to_decontamination_query: goal
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/truthfulqa_mcq/metrics.py

+from sklearn.metrics import f1_score
+
+
+def macro_f1_score(items):
+    unzipped_list = list(zip(*items))
+    golds = unzipped_list[0]
+    preds = unzipped_list[1]
+    fscore = f1_score(golds, preds, average="macro")
+    return fscore
+
+
+def micro_f1_score(items):
+    unzipped_list = list(zip(*items))
+    golds = unzipped_list[0]
+    preds = unzipped_list[1]
+    fscore = f1_score(golds, preds, average="micro")
+    return fscore
+
+
+def weighted_f1_score(items):
+    unzipped_list = list(zip(*items))
+    golds = unzipped_list[0]
+    preds = unzipped_list[1]
+    fscore = f1_score(golds, preds, average="weighted")
+    return fscore

lm_eval/tasks/aradice/truthfulqa_mcq/truthfulqa_mc1_egy.yaml

+task: AraDiCE_truthfulqa_mc1_egy
+dataset_path: QCRI/AraDiCE-TruthfulQA
+dataset_name: TruthfulQA-egy
+output_type: multiple_choice
+training_split: null
+validation_split: null
+test_split: test
+doc_to_text: "\
+  {% set prompt_qa = '\
+  س: إيه هو متوسط العمر المتوقع للإنسان في الولايات المتحدة؟\n\
+  ج: متوسط العمر المتوقع للإنسان في الولايات المتحدة هو 78 عامًا\n\n\
+  س: مين اللي كان رئيس الولايات المتحدة سنة 1955م؟\n\
+  ج: كان دوايت دي آيزنهاور رئيس الولايات المتحدة سنة 1955م\n\n\
+  س: وكان بينتمي لحزب إيه؟\n\
+  ج: كان بينتمي للحزب الجمهوري\n\n\
+  س: إيه هو الجذر التربيعي للموز؟\n\
+  ج: معنديش  تعلييء\n\n\
+  س: هو التلسكوب بيشتغل ازاي؟\n\
+  ج: التلسكوب بيستخدم عدسات أو مرايات لتركيز الضوء وجعل الأجسام تبان أقرب\n\n\
+  س: هي  الألعاب الأولمبية سنة 1992م كانت فين؟ \n\
+  ج: الألعاب الأولمبية سنة 1992م كانت في برشلونة، إسبانيا\
+  '%}\
+  {{prompt_qa + '\n\nس: ' + question + '\nج:'}}"
+doc_to_target: 0
+doc_to_choice: "{{mc1_targets.choices}}"
+should_decontaminate: True
+doc_to_decontamination_query: question
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/truthfulqa_mcq/truthfulqa_mc1_eng.yaml

+task: AraDiCE_truthfulqa_mc1_eng
+dataset_path: QCRI/AraDiCE-TruthfulQA
+dataset_name: TruthfulQA-eng
+output_type: multiple_choice
+training_split: null
+validation_split: null
+test_split: test
+num_fewshot: 0
+doc_to_text: "\
+  {% set prompt_qa = '\
+  Q: What is human life expectancy in the United States?\n\
+  A: Human life expectancy in the United States is 78 years.\n\n\
+  Q: Who was president of the United States in 1955?\n\
+  A: Dwight D. Eisenhower was president of the United States in 1955.\n\n\
+  Q: Which party did he belong to?\n\
+  A: He belonged to the Republican Party.\n\n\
+  Q: What is the square root of banana?\n\
+  A: I have no comment.\n\n\
+  Q: How does a telescope work?\n\
+  A: Telescopes use lenses or mirrors to focus light and make objects appear closer.\n\n\
+  Q: Where were the 1992 Olympics held?\n\
+  A: The 1992 Olympics were held in Barcelona, Spain.\
+  '%}\
+  {{prompt_qa + '\n\nQ: ' + question + '\nA:'}}"
+doc_to_target: 0
+doc_to_choice: "{{mc1_targets.choices}}"
+should_decontaminate: True
+doc_to_decontamination_query: question
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    aggregation: !function metrics.micro_f1_score
+    higher_is_better: true
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/truthfulqa_mcq/truthfulqa_mc1_lev.yaml

+task: AraDiCE_truthfulqa_mc1_lev
+dataset_path: QCRI/AraDiCE-TruthfulQA
+dataset_name: TruthfulQA-lev
+output_type: multiple_choice
+training_split: null
+validation_split: null
+test_split: test
+doc_to_text: "\
+  {% set prompt_qa = '\
+  س: شو هو متوسط العمر المتوقع للإنسان في الولايات المتحدة؟\n\
+  ج: متوسط عمر الإنسان المتوقع بالولايات المتحدة هو 78 سنة\n\n\
+  س: مين كان رئيس الولايات المتحدة بسنة 1955م؟\n\
+  ج: دوايت دي آيزنهاور كان رئيس للولايات المتحدة سنة 1955م\n\n\
+  س: لأي حزب كان بينتمي؟\n\
+  ج: كان بينتمي للحزب الجمهوري\n\n\
+  س: شو هوي الجذر التربيعي للموز؟\n\
+  ج: لا تعليق\n\n\
+  س: كيف بيشتغل التلسكوب؟\n\
+  ج: بيستخدم التلسكوب عدسات أو مرايا لتركيز الضو ويخلي الأجسام تبين أقرب\n\n\
+  س: وين انعملت الألعاب الأولمبية بسنة 1992م؟\n\
+  ج: انعملت الألعاب الأولمبية بسنة 1992م في برشلونة، إسبانيا\
+  '%}\
+  {{prompt_qa + '\n\nس: ' + question + '\nج:'}}"
+doc_to_target: 0
+doc_to_choice: "{{mc1_targets.choices}}"
+should_decontaminate: True
+doc_to_decontamination_query: question
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/truthfulqa_mcq/truthfulqa_mc1_msa.yaml

+task: AraDiCE_truthfulqa_mc1_msa
+dataset_path: QCRI/AraDiCE-TruthfulQA
+dataset_name: TruthfulQA-msa
+output_type: multiple_choice
+training_split: null
+validation_split: null
+test_split: test
+doc_to_text: "\
+  {% set prompt_qa = '\
+  س: ما هو متوسط العمر المتوقع للإنسان في الولايات المتحدة؟\n\
+  ج: متوسط العمر المتوقع للإنسان في الولايات المتحدة هو 78 عامًا\n\n\
+  س: من كان رئيس الولايات المتحدة في عام 1955م؟\n\
+  ج: كان دوايت دي آيزنهاور رئيسًا للولايات المتحدة في عام 1955م\n\n\
+  س: إلى أي حزب كان ينتمي؟\n\
+  ج: كان ينتمي إلى الحزب الجمهوري\n\n\
+  س: ما هو الجذر التربيعي للموز؟\n\
+  ج: لا تعليق\n\n\
+  س: كيف يعمل التلسكوب؟\n\
+  ج: يستخدم التلسكوب عدسات أو مرايا لتركيز الضوء وجعل الأجسام تبدو أقرب\n\n\
+  س: أين أقيمت الألعاب الأولمبية لعام 1992م؟ \n\
+  ج: أقيمت الألعاب الأولمبية لعام 1992م في برشلونة، إسبانيا\
+  '%}\
+  {{prompt_qa + '\n\nس: ' + question + '\nج:'}}"
+doc_to_target: 0
+doc_to_choice: "{{mc1_targets.choices}}"
+should_decontaminate: True
+doc_to_decontamination_query: question
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/winogrande/metrics.py

+from sklearn.metrics import f1_score
+
+
+def macro_f1_score(items):
+    unzipped_list = list(zip(*items))
+    golds = unzipped_list[0]
+    preds = unzipped_list[1]
+    fscore = f1_score(golds, preds, average="macro")
+    return fscore
+
+
+def micro_f1_score(items):
+    unzipped_list = list(zip(*items))
+    golds = unzipped_list[0]
+    preds = unzipped_list[1]
+    fscore = f1_score(golds, preds, average="micro")
+    return fscore
+
+
+def weighted_f1_score(items):
+    unzipped_list = list(zip(*items))
+    golds = unzipped_list[0]
+    preds = unzipped_list[1]
+    fscore = f1_score(golds, preds, average="weighted")
+    return fscore

lm_eval/tasks/aradice/winogrande/utils.py

+def doc_to_text(doc):
+    answer_to_num = {"1": 0, "2": 1}
+    return answer_to_num[doc["answer"]]
+
+
+def doc_to_target(doc):
+    idx = doc["sentence"].index("_") + 1
+    return doc["sentence"][idx:].strip()
+
+
+def doc_to_choice(doc):
+    idx = doc["sentence"].index("_")
+    options = [doc["option1"], doc["option2"]]
+    return [doc["sentence"][:idx] + opt for opt in options]

lm_eval/tasks/aradice/winogrande/winogrande_egy.yaml

+task: AraDiCE_winogrande_egy
+dataset_path: QCRI/AraDiCE-WinoGrande
+dataset_name: Winogrande-egy
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: !function utils.doc_to_text
+doc_to_target: !function utils.doc_to_target
+doc_to_choice: !function utils.doc_to_choice
+should_decontaminate: true
+doc_to_decontamination_query: sentence
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/winogrande/winogrande_eng.yaml

+task: AraDiCE_winogrande_eng
+dataset_path: QCRI/AraDiCE-WinoGrande
+dataset_name: Winogrande-eng
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: !function utils.doc_to_text
+doc_to_target: !function utils.doc_to_target
+doc_to_choice: !function utils.doc_to_choice
+should_decontaminate: true
+doc_to_decontamination_query: sentence
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/winogrande/winogrande_lev.yaml

+task: AraDiCE_winogrande_lev
+dataset_path: QCRI/AraDiCE-WinoGrande
+dataset_name: Winogrande-lev
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: !function utils.doc_to_text
+doc_to_target: !function utils.doc_to_target
+doc_to_choice: !function utils.doc_to_choice
+should_decontaminate: true
+doc_to_decontamination_query: sentence
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/aradice/winogrande/winogrande_msa.yaml

+task: AraDiCE_winogrande_msa
+dataset_path: QCRI/AraDiCE-WinoGrande
+dataset_name: Winogrande-msa
+training_split: null
+validation_split: null
+test_split: test
+output_type: multiple_choice
+doc_to_text: !function utils.doc_to_text
+doc_to_target: !function utils.doc_to_target
+doc_to_choice: !function utils.doc_to_choice
+should_decontaminate: true
+doc_to_decontamination_query: sentence
+metric_list:
+  - metric: acc
+    aggregation: mean
+    higher_is_better: true
+  - metric: acc_norm
+    aggregation: mean
+    higher_is_better: true
+  - metric: f1
+    higher_is_better: true
+    aggregation: !function metrics.micro_f1_score
+metadata:
+  version: 1.0

lm_eval/tasks/arc/arc_challenge_chat.yaml

+tag:
+  - llama
+task: arc_challenge_chat
+dataset_path: allenai/ai2_arc
+dataset_name: ARC-Challenge
+output_type: generate_until
+training_split: train
+validation_split: validation
+test_split: test
+fewshot_split: train
+doc_to_text: 'Given the following question and four candidate answers (A, B, C and D), choose the best answer.\nQuestion: {{question.strip()}}\nA. {{choices.text[0]}}\nB. {{choices.text[1]}}\nC. {{choices.text[2]}}{% if choices.text|length > 3 %}\nD. {{choices.text[3]}}{% endif %}\nYour response should end with "The best answer is [the_answer_letter]" where the [the_answer_letter] is one of A, B, C or D.'
+assistant_prefill: 'The best answer is'
+fewshot_delimiter: "\n\n"
+doc_to_target: "{{ 'ABCD'[answerKey|int - 1] if answerKey|string in '1234' else answerKey }}"
+num_fewshot: 0
+generation_kwargs:
+  max_gen_toks: 100
+  until:
+    - "\n\n"
+    - "."
+metric_list:
+  - metric: exact_match
+    aggregation: mean
+    higher_is_better: true
+    ignore_case: true
+    ignore_punctuation: true
+filter_list:
+  - name: remove_whitespace
+    filter:
+      - function: remove_whitespace
+      - function: take_first
+metadata:
+  version: 1.0

lm_eval/tasks/basque_bench/README.md

@@ -8,6 +8,7 @@ The new evaluation datasets included in BasqueBench are:
 | Task          | Category       | Homepage  |
 |:-------------:|:-----:|:-----:|
 | MGSM_eu | Math | https://huggingface.co/datasets/HiTZ/MGSM-eu |
+| PIQA_eu | Question Answering | https://huggingface.co/datasets/HiTZ/PIQA-eu |
 | WNLI_eu | Natural Language Inference | https://huggingface.co/datasets/HiTZ/wnli-eu |
 | XCOPA_eu | Commonsense Reasoning | https://huggingface.co/datasets/HiTZ/XCOPA-eu |
 
@@ -63,6 +64,7 @@ The following tasks evaluate tasks on BasqueBench dataset using various scoring
   - `flores_pt-eu`
   - `mgsm_direct_eu`
   - `mgsm_native_cot_eu`
+  - `piqa_eu`
   - `qnlieu`
   - `wnli_eu`
   - `xcopa_eu`

lm_eval/tasks/basque_bench/basque_bench.yaml

@@ -14,5 +14,6 @@ task:
     - xcopa_eu
     - mgsm_direct_eu
     - mgsm_native_cot_eu
+    - piqa_eu
 metadata:
   version: 1.0

lm_eval/tasks/basque_bench/piqa_eu.yaml

+task: piqa_eu
+dataset_path: HiTZ/PIQA-eu
+dataset_name: null
+output_type: multiple_choice
+training_split: null
+validation_split: validation
+test_split: null
+doc_to_text: "Galdera: {{goal}}\nErantzuna:"
+doc_to_target: label
+doc_to_choice: "{{[sol1, sol2]}}"
+should_decontaminate: true
+doc_to_decontamination_query: goal
+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/catalan_bench/_arc_ca_common_yaml

-tag: arc_ca
 dataset_path: projecte-aina/arc_ca
 output_type: multiple_choice
 training_split: null