diff --git "a/evaluations/ar/Falcon3-7B-Instruct/openaimmlu_0_shot.json" "b/evaluations/ar/Falcon3-7B-Instruct/openaimmlu_0_shot.json" new file mode 100644--- /dev/null +++ "b/evaluations/ar/Falcon3-7B-Instruct/openaimmlu_0_shot.json" @@ -0,0 +1,2711 @@ +{ + "results": { + "openaimmlu": { + " ": " ", + "alias": "openaimmlu" + }, + "openaimmlu_STEM": { + "acc,none": 0.32847682119205296, + "acc_stderr,none": 0.008517820734335659, + "alias": " - STEM" + }, + "openaimmlu_abstract_algebra": { + "alias": " - abstract_algebra", + "acc,none": 0.34, + "acc_stderr,none": 0.04760952285695235 + }, + "openaimmlu_astronomy": { + "alias": " - astronomy", + "acc,none": 0.35526315789473684, + "acc_stderr,none": 0.038947344870133176 + }, + "openaimmlu_college_biology": { + "alias": " - college_biology", + "acc,none": 0.2708333333333333, + "acc_stderr,none": 0.03716177437566016 + }, + "openaimmlu_college_chemistry": { + "alias": " - college_chemistry", + "acc,none": 0.29, + "acc_stderr,none": 0.045604802157206845 + }, + "openaimmlu_college_computer_science": { + "alias": " - college_computer_science", + "acc,none": 0.38, + "acc_stderr,none": 0.04878317312145634 + }, + "openaimmlu_college_mathematics": { + "alias": " - college_mathematics", + "acc,none": 0.28, + "acc_stderr,none": 0.045126085985421296 + }, + "openaimmlu_college_physics": { + "alias": " - college_physics", + "acc,none": 0.23529411764705882, + "acc_stderr,none": 0.04220773659171453 + }, + "openaimmlu_computer_security": { + "alias": " - computer_security", + "acc,none": 0.32, + "acc_stderr,none": 0.046882617226215034 + }, + "openaimmlu_conceptual_physics": { + "alias": " - conceptual_physics", + "acc,none": 0.30638297872340425, + "acc_stderr,none": 0.030135906478517563 + }, + "openaimmlu_econometrics": { + "alias": " - econometrics", + "acc,none": 0.30701754385964913, + "acc_stderr,none": 0.04339138322579861 + }, + "openaimmlu_electrical_engineering": { + "alias": " - electrical_engineering", + "acc,none": 0.38620689655172413, + "acc_stderr,none": 0.04057324734419034 + }, + "openaimmlu_elementary_mathematics": { + "alias": " - elementary_mathematics", + "acc,none": 0.40476190476190477, + "acc_stderr,none": 0.025279850397404904 + }, + "openaimmlu_high_school_biology": { + "alias": " - high_school_biology", + "acc,none": 0.3161290322580645, + "acc_stderr,none": 0.026450874489042767 + }, + "openaimmlu_high_school_chemistry": { + "alias": " - high_school_chemistry", + "acc,none": 0.3399014778325123, + "acc_stderr,none": 0.033327690684107895 + }, + "openaimmlu_high_school_computer_science": { + "alias": " - high_school_computer_science", + "acc,none": 0.44, + "acc_stderr,none": 0.04988876515698589 + }, + "openaimmlu_high_school_mathematics": { + "alias": " - high_school_mathematics", + "acc,none": 0.34444444444444444, + "acc_stderr,none": 0.028972648884844267 + }, + "openaimmlu_high_school_physics": { + "alias": " - high_school_physics", + "acc,none": 0.23841059602649006, + "acc_stderr,none": 0.03479185572599657 + }, + "openaimmlu_high_school_statistics": { + "alias": " - high_school_statistics", + "acc,none": 0.26851851851851855, + "acc_stderr,none": 0.030225226160012417 + }, + "openaimmlu_humanities": { + "acc,none": 0.3464523281596452, + "acc_stderr,none": 0.011178696015775447, + "alias": " - Humanities" + }, + "openaimmlu_high_school_european_history": { + "alias": " - high_school_european_history", + "acc,none": 0.3939393939393939, + "acc_stderr,none": 0.0381549430868893 + }, + "openaimmlu_high_school_us_history": { + "alias": " - high_school_us_history", + "acc,none": 0.3235294117647059, + "acc_stderr,none": 0.03283472056108566 + }, + "openaimmlu_high_school_world_history": { + "alias": " - high_school_world_history", + "acc,none": 0.3459915611814346, + "acc_stderr,none": 0.03096481058878671 + }, + "openaimmlu_international_law": { + "alias": " - international_law", + "acc,none": 0.4628099173553719, + "acc_stderr,none": 0.04551711196104218 + }, + "openaimmlu_jurisprudence": { + "alias": " - jurisprudence", + "acc,none": 0.4166666666666667, + "acc_stderr,none": 0.04766075165356461 + }, + "openaimmlu_logical_fallacies": { + "alias": " - logical_fallacies", + "acc,none": 0.3374233128834356, + "acc_stderr,none": 0.03714908409935573 + }, + "openaimmlu_philosophy": { + "alias": " - philosophy", + "acc,none": 0.3408360128617363, + "acc_stderr,none": 0.02692084126077616 + }, + "openaimmlu_prehistory": { + "alias": " - prehistory", + "acc,none": 0.31790123456790126, + "acc_stderr,none": 0.025910063528240868 + }, + "openaimmlu_world_religions": { + "alias": " - world_religions", + "acc,none": 0.27485380116959063, + "acc_stderr,none": 0.03424042924691583 + }, + "openaimmlu_other": { + "acc,none": 0.3083277140930546, + "acc_stderr,none": 0.0059796238033850944, + "alias": " - Other" + }, + "openaimmlu_anatomy": { + "alias": " - anatomy", + "acc,none": 0.3037037037037037, + "acc_stderr,none": 0.03972552884785137 + }, + "openaimmlu_clinical_knowledge": { + "alias": " - clinical_knowledge", + "acc,none": 0.30566037735849055, + "acc_stderr,none": 0.028353298073322666 + }, + "openaimmlu_college_medicine": { + "alias": " - college_medicine", + "acc,none": 0.2832369942196532, + "acc_stderr,none": 0.03435568056047874 + }, + "openaimmlu_formal_logic": { + "alias": " - formal_logic", + "acc,none": 0.3412698412698413, + "acc_stderr,none": 0.042407993275749234 + }, + "openaimmlu_global_facts": { + "alias": " - global_facts", + "acc,none": 0.34, + "acc_stderr,none": 0.04760952285695235 + }, + "openaimmlu_high_school_geography": { + "alias": " - high_school_geography", + "acc,none": 0.3181818181818182, + "acc_stderr,none": 0.03318477333845332 + }, + "openaimmlu_high_school_psychology": { + "alias": " - high_school_psychology", + "acc,none": 0.28807339449541286, + "acc_stderr,none": 0.01941644589263603 + }, + "openaimmlu_human_aging": { + "alias": " - human_aging", + "acc,none": 0.3273542600896861, + "acc_stderr,none": 0.031493846709941306 + }, + "openaimmlu_machine_learning": { + "alias": " - machine_learning", + "acc,none": 0.23214285714285715, + "acc_stderr,none": 0.04007341809755806 + }, + "openaimmlu_medical_genetics": { + "alias": " - medical_genetics", + "acc,none": 0.41, + "acc_stderr,none": 0.04943110704237102 + }, + "openaimmlu_miscellaneous": { + "alias": " - miscellaneous", + "acc,none": 0.34738186462324394, + "acc_stderr,none": 0.01702667174865574 + }, + "openaimmlu_nutrition": { + "alias": " - nutrition", + "acc,none": 0.4084967320261438, + "acc_stderr,none": 0.028146405993096358 + }, + "openaimmlu_professional_accounting": { + "alias": " - professional_accounting", + "acc,none": 0.25886524822695034, + "acc_stderr,none": 0.02612957252718085 + }, + "openaimmlu_professional_law": { + "alias": " - professional_law", + "acc,none": 0.30182529335071706, + "acc_stderr,none": 0.011724350518105888 + }, + "openaimmlu_professional_medicine": { + "alias": " - professional_medicine", + "acc,none": 0.22058823529411764, + "acc_stderr,none": 0.02518778666022727 + }, + "openaimmlu_professional_psychology": { + "alias": " - professional_psychology", + "acc,none": 0.2761437908496732, + "acc_stderr,none": 0.018087276935663137 + }, + "openaimmlu_virology": { + "alias": " - virology", + "acc,none": 0.35542168674698793, + "acc_stderr,none": 0.03726214354322415 + }, + "openaimmlu_social_science": { + "acc,none": 0.33414485696895924, + "acc_stderr,none": 0.008161503557308653, + "alias": " - Social Science" + }, + "openaimmlu_business_ethics": { + "alias": " - business_ethics", + "acc,none": 0.37, + "acc_stderr,none": 0.04852365870939099 + }, + "openaimmlu_high_school_government_and_politics": { + "alias": " - high_school_government_and_politics", + "acc,none": 0.26424870466321243, + "acc_stderr,none": 0.03182155050916648 + }, + "openaimmlu_high_school_macroeconomics": { + "alias": " - high_school_macroeconomics", + "acc,none": 0.31794871794871793, + "acc_stderr,none": 0.023610884308927865 + }, + "openaimmlu_high_school_microeconomics": { + "alias": " - high_school_microeconomics", + "acc,none": 0.3277310924369748, + "acc_stderr,none": 0.030489911417673227 + }, + "openaimmlu_human_sexuality": { + "alias": " - human_sexuality", + "acc,none": 0.4198473282442748, + "acc_stderr,none": 0.04328577215262972 + }, + "openaimmlu_management": { + "alias": " - management", + "acc,none": 0.3106796116504854, + "acc_stderr,none": 0.04582124160161551 + }, + "openaimmlu_marketing": { + "alias": " - marketing", + "acc,none": 0.4230769230769231, + "acc_stderr,none": 0.032366121762202014 + }, + "openaimmlu_moral_disputes": { + "alias": " - moral_disputes", + "acc,none": 0.31213872832369943, + "acc_stderr,none": 0.024946792225272307 + }, + "openaimmlu_moral_scenarios": { + "alias": " - moral_scenarios", + "acc,none": 0.2681564245810056, + "acc_stderr,none": 0.014816119635317008 + }, + "openaimmlu_public_relations": { + "alias": " - public_relations", + "acc,none": 0.35454545454545455, + "acc_stderr,none": 0.04582004841505417 + }, + "openaimmlu_security_studies": { + "alias": " - security_studies", + "acc,none": 0.4, + "acc_stderr,none": 0.03136250240935893 + }, + "openaimmlu_sociology": { + "alias": " - sociology", + "acc,none": 0.4129353233830846, + "acc_stderr,none": 0.03481520803367348 + }, + "openaimmlu_us_foreign_policy": { + "alias": " - us_foreign_policy", + "acc,none": 0.54, + "acc_stderr,none": 0.05009082659620333 + } + }, + "groups": { + "openaimmlu_STEM": { + "acc,none": 0.32847682119205296, + "acc_stderr,none": 0.008517820734335659, + "alias": " - STEM" + }, + "openaimmlu_humanities": { + "acc,none": 0.3464523281596452, + "acc_stderr,none": 0.011178696015775447, + "alias": " - Humanities" + }, + "openaimmlu_other": { + "acc,none": 0.3083277140930546, + "acc_stderr,none": 0.0059796238033850944, + "alias": " - Other" + }, + "openaimmlu_social_science": { + "acc,none": 0.33414485696895924, + "acc_stderr,none": 0.008161503557308653, + "alias": " - Social Science" + } + }, + "group_subtasks": { + "openaimmlu_humanities": [ + "openaimmlu_international_law", + "openaimmlu_jurisprudence", + "openaimmlu_high_school_world_history", + "openaimmlu_prehistory", + "openaimmlu_world_religions", + "openaimmlu_philosophy", + "openaimmlu_logical_fallacies", + "openaimmlu_high_school_european_history", + "openaimmlu_high_school_us_history" + ], + "openaimmlu_social_science": [ + "openaimmlu_management", + "openaimmlu_business_ethics", + "openaimmlu_security_studies", + "openaimmlu_moral_scenarios", + "openaimmlu_marketing", + "openaimmlu_high_school_government_and_politics", + "openaimmlu_public_relations", + "openaimmlu_high_school_microeconomics", + "openaimmlu_us_foreign_policy", + "openaimmlu_high_school_macroeconomics", + "openaimmlu_moral_disputes", + "openaimmlu_human_sexuality", + "openaimmlu_sociology" + ], + "openaimmlu_other": [ + "openaimmlu_miscellaneous", + "openaimmlu_professional_law", + "openaimmlu_machine_learning", + "openaimmlu_global_facts", + "openaimmlu_anatomy", + "openaimmlu_college_medicine", + "openaimmlu_human_aging", + "openaimmlu_formal_logic", + "openaimmlu_professional_accounting", + "openaimmlu_high_school_psychology", + "openaimmlu_clinical_knowledge", + "openaimmlu_professional_psychology", + "openaimmlu_medical_genetics", + "openaimmlu_virology", + "openaimmlu_professional_medicine", + "openaimmlu_nutrition", + "openaimmlu_high_school_geography" + ], + "openaimmlu_STEM": [ + "openaimmlu_high_school_mathematics", + "openaimmlu_college_physics", + "openaimmlu_computer_security", + "openaimmlu_college_computer_science", + "openaimmlu_abstract_algebra", + "openaimmlu_high_school_statistics", + "openaimmlu_college_mathematics", + "openaimmlu_college_chemistry", + "openaimmlu_high_school_computer_science", + "openaimmlu_elementary_mathematics", + "openaimmlu_high_school_physics", + "openaimmlu_conceptual_physics", + "openaimmlu_econometrics", + "openaimmlu_college_biology", + "openaimmlu_electrical_engineering", + "openaimmlu_astronomy", + "openaimmlu_high_school_chemistry", + "openaimmlu_high_school_biology" + ], + "openaimmlu": [ + "openaimmlu_STEM", + "openaimmlu_other", + "openaimmlu_social_science", + "openaimmlu_humanities" + ] + }, + "configs": { + "openaimmlu_abstract_algebra": { + "task": "openaimmlu_abstract_algebra", + "task_alias": "abstract_algebra", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "abstract_algebra", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_anatomy": { + "task": "openaimmlu_anatomy", + "task_alias": "anatomy", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "anatomy", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_astronomy": { + "task": "openaimmlu_astronomy", + "task_alias": "astronomy", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "astronomy", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_business_ethics": { + "task": "openaimmlu_business_ethics", + "task_alias": "business_ethics", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "business_ethics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_clinical_knowledge": { + "task": "openaimmlu_clinical_knowledge", + "task_alias": "clinical_knowledge", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "clinical_knowledge", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_college_biology": { + "task": "openaimmlu_college_biology", + "task_alias": "college_biology", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "college_biology", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_college_chemistry": { + "task": "openaimmlu_college_chemistry", + "task_alias": "college_chemistry", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "college_chemistry", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_college_computer_science": { + "task": "openaimmlu_college_computer_science", + "task_alias": "college_computer_science", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "college_computer_science", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_college_mathematics": { + "task": "openaimmlu_college_mathematics", + "task_alias": "college_mathematics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "college_mathematics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_college_medicine": { + "task": "openaimmlu_college_medicine", + "task_alias": "college_medicine", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "college_medicine", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_college_physics": { + "task": "openaimmlu_college_physics", + "task_alias": "college_physics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "college_physics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_computer_security": { + "task": "openaimmlu_computer_security", + "task_alias": "computer_security", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "computer_security", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_conceptual_physics": { + "task": "openaimmlu_conceptual_physics", + "task_alias": "conceptual_physics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "conceptual_physics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_econometrics": { + "task": "openaimmlu_econometrics", + "task_alias": "econometrics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "econometrics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_electrical_engineering": { + "task": "openaimmlu_electrical_engineering", + "task_alias": "electrical_engineering", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "electrical_engineering", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_elementary_mathematics": { + "task": "openaimmlu_elementary_mathematics", + "task_alias": "elementary_mathematics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "elementary_mathematics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_formal_logic": { + "task": "openaimmlu_formal_logic", + "task_alias": "formal_logic", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "formal_logic", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_global_facts": { + "task": "openaimmlu_global_facts", + "task_alias": "global_facts", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "global_facts", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_biology": { + "task": "openaimmlu_high_school_biology", + "task_alias": "high_school_biology", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_biology", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_chemistry": { + "task": "openaimmlu_high_school_chemistry", + "task_alias": "high_school_chemistry", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_chemistry", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_computer_science": { + "task": "openaimmlu_high_school_computer_science", + "task_alias": "high_school_computer_science", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_computer_science", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_european_history": { + "task": "openaimmlu_high_school_european_history", + "task_alias": "high_school_european_history", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_european_history", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_geography": { + "task": "openaimmlu_high_school_geography", + "task_alias": "high_school_geography", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_geography", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_government_and_politics": { + "task": "openaimmlu_high_school_government_and_politics", + "task_alias": "high_school_government_and_politics", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_government_and_politics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_macroeconomics": { + "task": "openaimmlu_high_school_macroeconomics", + "task_alias": "high_school_macroeconomics", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_macroeconomics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_mathematics": { + "task": "openaimmlu_high_school_mathematics", + "task_alias": "high_school_mathematics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_mathematics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_microeconomics": { + "task": "openaimmlu_high_school_microeconomics", + "task_alias": "high_school_microeconomics", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_microeconomics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_physics": { + "task": "openaimmlu_high_school_physics", + "task_alias": "high_school_physics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_physics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_psychology": { + "task": "openaimmlu_high_school_psychology", + "task_alias": "high_school_psychology", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_psychology", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_statistics": { + "task": "openaimmlu_high_school_statistics", + "task_alias": "high_school_statistics", + "tag": "openaimmlu_STEM_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_statistics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_us_history": { + "task": "openaimmlu_high_school_us_history", + "task_alias": "high_school_us_history", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_us_history", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_high_school_world_history": { + "task": "openaimmlu_high_school_world_history", + "task_alias": "high_school_world_history", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "high_school_world_history", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_human_aging": { + "task": "openaimmlu_human_aging", + "task_alias": "human_aging", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "human_aging", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_human_sexuality": { + "task": "openaimmlu_human_sexuality", + "task_alias": "human_sexuality", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "human_sexuality", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_international_law": { + "task": "openaimmlu_international_law", + "task_alias": "international_law", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "international_law", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_jurisprudence": { + "task": "openaimmlu_jurisprudence", + "task_alias": "jurisprudence", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "jurisprudence", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_logical_fallacies": { + "task": "openaimmlu_logical_fallacies", + "task_alias": "logical_fallacies", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "logical_fallacies", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_machine_learning": { + "task": "openaimmlu_machine_learning", + "task_alias": "machine_learning", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "machine_learning", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_management": { + "task": "openaimmlu_management", + "task_alias": "management", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "management", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_marketing": { + "task": "openaimmlu_marketing", + "task_alias": "marketing", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "marketing", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_medical_genetics": { + "task": "openaimmlu_medical_genetics", + "task_alias": "medical_genetics", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "medical_genetics", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_miscellaneous": { + "task": "openaimmlu_miscellaneous", + "task_alias": "miscellaneous", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "miscellaneous", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_moral_disputes": { + "task": "openaimmlu_moral_disputes", + "task_alias": "moral_disputes", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "moral_disputes", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_moral_scenarios": { + "task": "openaimmlu_moral_scenarios", + "task_alias": "moral_scenarios", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "moral_scenarios", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_nutrition": { + "task": "openaimmlu_nutrition", + "task_alias": "nutrition", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "nutrition", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_philosophy": { + "task": "openaimmlu_philosophy", + "task_alias": "philosophy", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "philosophy", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_prehistory": { + "task": "openaimmlu_prehistory", + "task_alias": "prehistory", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "prehistory", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_professional_accounting": { + "task": "openaimmlu_professional_accounting", + "task_alias": "professional_accounting", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "professional_accounting", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_professional_law": { + "task": "openaimmlu_professional_law", + "task_alias": "professional_law", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "professional_law", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_professional_medicine": { + "task": "openaimmlu_professional_medicine", + "task_alias": "professional_medicine", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "professional_medicine", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_professional_psychology": { + "task": "openaimmlu_professional_psychology", + "task_alias": "professional_psychology", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "professional_psychology", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_public_relations": { + "task": "openaimmlu_public_relations", + "task_alias": "public_relations", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "public_relations", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_security_studies": { + "task": "openaimmlu_security_studies", + "task_alias": "security_studies", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "security_studies", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_sociology": { + "task": "openaimmlu_sociology", + "task_alias": "sociology", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "sociology", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_us_foreign_policy": { + "task": "openaimmlu_us_foreign_policy", + "task_alias": "us_foreign_policy", + "tag": "openaimmlu_social_science_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "us_foreign_policy", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_virology": { + "task": "openaimmlu_virology", + "task_alias": "virology", + "tag": "openaimmlu_other_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "virology", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + }, + "openaimmlu_world_religions": { + "task": "openaimmlu_world_religions", + "task_alias": "world_religions", + "tag": "openaimmlu_humanities_tasks", + "dataset_path": "khalidalt/openai_mmlu_arabic", + "dataset_name": "world_religions", + "test_split": "test", + "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n def _process_docs(doc):\n\n def format_example(doc, choices):\n options = []\n for _, choice in enumerate(choices):\n options.append(f'{en2ar[choice]}. {doc[choice]}')\n\n ar_subject = SUBJECTS[doc['Subject']]\n query = PROMPT.format(ar_subject, #doc['Subject'],\n doc['Question'],\n \"\\n\".join(options))\n return query\n\n keys_en = [\"A\", \"B\", \"C\", \"D\"]\n keys_ar = ['\u0623', '\u0628', '\u062c', '\u062f']\n ar_label = en2ar[doc['Answer']]\n\n out_doc = {\n \"query\": format_example(doc, keys_en),\n \"choices\": keys_ar,\n \"gold\": keys_ar.index(ar_label)\n }\n\n return out_doc\n\n return dataset.map(_process_docs) \n", + "doc_to_text": "query", + "doc_to_target": "gold", + "doc_to_choice": "choices", + "description": "", + "target_delimiter": " ", + "fewshot_delimiter": "\n\n", + "num_fewshot": 0, + "metric_list": [ + { + "metric": "acc", + "aggregation": "mean", + "higher_is_better": true + } + ], + "output_type": "multiple_choice", + "repeats": 1, + "should_decontaminate": false, + "metadata": { + "version": 0.0 + } + } + }, + "versions": { + "openaimmlu_STEM": 0, + "openaimmlu_abstract_algebra": 0.0, + "openaimmlu_anatomy": 0.0, + "openaimmlu_astronomy": 0.0, + "openaimmlu_business_ethics": 0.0, + "openaimmlu_clinical_knowledge": 0.0, + "openaimmlu_college_biology": 0.0, + "openaimmlu_college_chemistry": 0.0, + "openaimmlu_college_computer_science": 0.0, + "openaimmlu_college_mathematics": 0.0, + "openaimmlu_college_medicine": 0.0, + "openaimmlu_college_physics": 0.0, + "openaimmlu_computer_security": 0.0, + "openaimmlu_conceptual_physics": 0.0, + "openaimmlu_econometrics": 0.0, + "openaimmlu_electrical_engineering": 0.0, + "openaimmlu_elementary_mathematics": 0.0, + "openaimmlu_formal_logic": 0.0, + "openaimmlu_global_facts": 0.0, + "openaimmlu_high_school_biology": 0.0, + "openaimmlu_high_school_chemistry": 0.0, + "openaimmlu_high_school_computer_science": 0.0, + "openaimmlu_high_school_european_history": 0.0, + "openaimmlu_high_school_geography": 0.0, + "openaimmlu_high_school_government_and_politics": 0.0, + "openaimmlu_high_school_macroeconomics": 0.0, + "openaimmlu_high_school_mathematics": 0.0, + "openaimmlu_high_school_microeconomics": 0.0, + "openaimmlu_high_school_physics": 0.0, + "openaimmlu_high_school_psychology": 0.0, + "openaimmlu_high_school_statistics": 0.0, + "openaimmlu_high_school_us_history": 0.0, + "openaimmlu_high_school_world_history": 0.0, + "openaimmlu_human_aging": 0.0, + "openaimmlu_human_sexuality": 0.0, + "openaimmlu_humanities": 0, + "openaimmlu_international_law": 0.0, + "openaimmlu_jurisprudence": 0.0, + "openaimmlu_logical_fallacies": 0.0, + "openaimmlu_machine_learning": 0.0, + "openaimmlu_management": 0.0, + "openaimmlu_marketing": 0.0, + "openaimmlu_medical_genetics": 0.0, + "openaimmlu_miscellaneous": 0.0, + "openaimmlu_moral_disputes": 0.0, + "openaimmlu_moral_scenarios": 0.0, + "openaimmlu_nutrition": 0.0, + "openaimmlu_other": 0, + "openaimmlu_philosophy": 0.0, + "openaimmlu_prehistory": 0.0, + "openaimmlu_professional_accounting": 0.0, + "openaimmlu_professional_law": 0.0, + "openaimmlu_professional_medicine": 0.0, + "openaimmlu_professional_psychology": 0.0, + "openaimmlu_public_relations": 0.0, + "openaimmlu_security_studies": 0.0, + "openaimmlu_social_science": 0, + "openaimmlu_sociology": 0.0, + "openaimmlu_us_foreign_policy": 0.0, + "openaimmlu_virology": 0.0, + "openaimmlu_world_religions": 0.0 + }, + "n-shot": { + "openaimmlu_abstract_algebra": 0, + "openaimmlu_anatomy": 0, + "openaimmlu_astronomy": 0, + "openaimmlu_business_ethics": 0, + "openaimmlu_clinical_knowledge": 0, + "openaimmlu_college_biology": 0, + "openaimmlu_college_chemistry": 0, + "openaimmlu_college_computer_science": 0, + "openaimmlu_college_mathematics": 0, + "openaimmlu_college_medicine": 0, + "openaimmlu_college_physics": 0, + "openaimmlu_computer_security": 0, + "openaimmlu_conceptual_physics": 0, + "openaimmlu_econometrics": 0, + "openaimmlu_electrical_engineering": 0, + "openaimmlu_elementary_mathematics": 0, + "openaimmlu_formal_logic": 0, + "openaimmlu_global_facts": 0, + "openaimmlu_high_school_biology": 0, + "openaimmlu_high_school_chemistry": 0, + "openaimmlu_high_school_computer_science": 0, + "openaimmlu_high_school_european_history": 0, + "openaimmlu_high_school_geography": 0, + "openaimmlu_high_school_government_and_politics": 0, + "openaimmlu_high_school_macroeconomics": 0, + "openaimmlu_high_school_mathematics": 0, + "openaimmlu_high_school_microeconomics": 0, + "openaimmlu_high_school_physics": 0, + "openaimmlu_high_school_psychology": 0, + "openaimmlu_high_school_statistics": 0, + "openaimmlu_high_school_us_history": 0, + "openaimmlu_high_school_world_history": 0, + "openaimmlu_human_aging": 0, + "openaimmlu_human_sexuality": 0, + "openaimmlu_international_law": 0, + "openaimmlu_jurisprudence": 0, + "openaimmlu_logical_fallacies": 0, + "openaimmlu_machine_learning": 0, + "openaimmlu_management": 0, + "openaimmlu_marketing": 0, + "openaimmlu_medical_genetics": 0, + "openaimmlu_miscellaneous": 0, + "openaimmlu_moral_disputes": 0, + "openaimmlu_moral_scenarios": 0, + "openaimmlu_nutrition": 0, + "openaimmlu_philosophy": 0, + "openaimmlu_prehistory": 0, + "openaimmlu_professional_accounting": 0, + "openaimmlu_professional_law": 0, + "openaimmlu_professional_medicine": 0, + "openaimmlu_professional_psychology": 0, + "openaimmlu_public_relations": 0, + "openaimmlu_security_studies": 0, + "openaimmlu_sociology": 0, + "openaimmlu_us_foreign_policy": 0, + "openaimmlu_virology": 0, + "openaimmlu_world_religions": 0 + }, + "higher_is_better": { + "openaimmlu": { + "acc": true + }, + "openaimmlu_STEM": { + "acc": true + }, + "openaimmlu_abstract_algebra": { + "acc": true + }, + "openaimmlu_anatomy": { + "acc": true + }, + "openaimmlu_astronomy": { + "acc": true + }, + "openaimmlu_business_ethics": { + "acc": true + }, + "openaimmlu_clinical_knowledge": { + "acc": true + }, + "openaimmlu_college_biology": { + "acc": true + }, + "openaimmlu_college_chemistry": { + "acc": true + }, + "openaimmlu_college_computer_science": { + "acc": true + }, + "openaimmlu_college_mathematics": { + "acc": true + }, + "openaimmlu_college_medicine": { + "acc": true + }, + "openaimmlu_college_physics": { + "acc": true + }, + "openaimmlu_computer_security": { + "acc": true + }, + "openaimmlu_conceptual_physics": { + "acc": true + }, + "openaimmlu_econometrics": { + "acc": true + }, + "openaimmlu_electrical_engineering": { + "acc": true + }, + "openaimmlu_elementary_mathematics": { + "acc": true + }, + "openaimmlu_formal_logic": { + "acc": true + }, + "openaimmlu_global_facts": { + "acc": true + }, + "openaimmlu_high_school_biology": { + "acc": true + }, + "openaimmlu_high_school_chemistry": { + "acc": true + }, + "openaimmlu_high_school_computer_science": { + "acc": true + }, + "openaimmlu_high_school_european_history": { + "acc": true + }, + "openaimmlu_high_school_geography": { + "acc": true + }, + "openaimmlu_high_school_government_and_politics": { + "acc": true + }, + "openaimmlu_high_school_macroeconomics": { + "acc": true + }, + "openaimmlu_high_school_mathematics": { + "acc": true + }, + "openaimmlu_high_school_microeconomics": { + "acc": true + }, + "openaimmlu_high_school_physics": { + "acc": true + }, + "openaimmlu_high_school_psychology": { + "acc": true + }, + "openaimmlu_high_school_statistics": { + "acc": true + }, + "openaimmlu_high_school_us_history": { + "acc": true + }, + "openaimmlu_high_school_world_history": { + "acc": true + }, + "openaimmlu_human_aging": { + "acc": true + }, + "openaimmlu_human_sexuality": { + "acc": true + }, + "openaimmlu_humanities": { + "acc": true + }, + "openaimmlu_international_law": { + "acc": true + }, + "openaimmlu_jurisprudence": { + "acc": true + }, + "openaimmlu_logical_fallacies": { + "acc": true + }, + "openaimmlu_machine_learning": { + "acc": true + }, + "openaimmlu_management": { + "acc": true + }, + "openaimmlu_marketing": { + "acc": true + }, + "openaimmlu_medical_genetics": { + "acc": true + }, + "openaimmlu_miscellaneous": { + "acc": true + }, + "openaimmlu_moral_disputes": { + "acc": true + }, + "openaimmlu_moral_scenarios": { + "acc": true + }, + "openaimmlu_nutrition": { + "acc": true + }, + "openaimmlu_other": { + "acc": true + }, + "openaimmlu_philosophy": { + "acc": true + }, + "openaimmlu_prehistory": { + "acc": true + }, + "openaimmlu_professional_accounting": { + "acc": true + }, + "openaimmlu_professional_law": { + "acc": true + }, + "openaimmlu_professional_medicine": { + "acc": true + }, + "openaimmlu_professional_psychology": { + "acc": true + }, + "openaimmlu_public_relations": { + "acc": true + }, + "openaimmlu_security_studies": { + "acc": true + }, + "openaimmlu_social_science": { + "acc": true + }, + "openaimmlu_sociology": { + "acc": true + }, + "openaimmlu_us_foreign_policy": { + "acc": true + }, + "openaimmlu_virology": { + "acc": true + }, + "openaimmlu_world_religions": { + "acc": true + } + }, + "n-samples": { + "openaimmlu_high_school_mathematics": { + "original": 270, + "effective": 270 + }, + "openaimmlu_college_physics": { + "original": 102, + "effective": 102 + }, + "openaimmlu_computer_security": { + "original": 100, + "effective": 100 + }, + "openaimmlu_college_computer_science": { + "original": 100, + "effective": 100 + }, + "openaimmlu_abstract_algebra": { + "original": 100, + "effective": 100 + }, + "openaimmlu_high_school_statistics": { + "original": 216, + "effective": 216 + }, + "openaimmlu_college_mathematics": { + "original": 100, + "effective": 100 + }, + "openaimmlu_college_chemistry": { + "original": 100, + "effective": 100 + }, + "openaimmlu_high_school_computer_science": { + "original": 100, + "effective": 100 + }, + "openaimmlu_elementary_mathematics": { + "original": 378, + "effective": 378 + }, + "openaimmlu_high_school_physics": { + "original": 151, + "effective": 151 + }, + "openaimmlu_conceptual_physics": { + "original": 235, + "effective": 235 + }, + "openaimmlu_econometrics": { + "original": 114, + "effective": 114 + }, + "openaimmlu_college_biology": { + "original": 144, + "effective": 144 + }, + "openaimmlu_electrical_engineering": { + "original": 145, + "effective": 145 + }, + "openaimmlu_astronomy": { + "original": 152, + "effective": 152 + }, + "openaimmlu_high_school_chemistry": { + "original": 203, + "effective": 203 + }, + "openaimmlu_high_school_biology": { + "original": 310, + "effective": 310 + }, + "openaimmlu_miscellaneous": { + "original": 783, + "effective": 783 + }, + "openaimmlu_professional_law": { + "original": 1534, + "effective": 1534 + }, + "openaimmlu_machine_learning": { + "original": 112, + "effective": 112 + }, + "openaimmlu_global_facts": { + "original": 100, + "effective": 100 + }, + "openaimmlu_anatomy": { + "original": 135, + "effective": 135 + }, + "openaimmlu_college_medicine": { + "original": 173, + "effective": 173 + }, + "openaimmlu_human_aging": { + "original": 223, + "effective": 223 + }, + "openaimmlu_formal_logic": { + "original": 126, + "effective": 126 + }, + "openaimmlu_professional_accounting": { + "original": 282, + "effective": 282 + }, + "openaimmlu_high_school_psychology": { + "original": 545, + "effective": 545 + }, + "openaimmlu_clinical_knowledge": { + "original": 265, + "effective": 265 + }, + "openaimmlu_professional_psychology": { + "original": 612, + "effective": 612 + }, + "openaimmlu_medical_genetics": { + "original": 100, + "effective": 100 + }, + "openaimmlu_virology": { + "original": 166, + "effective": 166 + }, + "openaimmlu_professional_medicine": { + "original": 272, + "effective": 272 + }, + "openaimmlu_nutrition": { + "original": 306, + "effective": 306 + }, + "openaimmlu_high_school_geography": { + "original": 198, + "effective": 198 + }, + "openaimmlu_management": { + "original": 103, + "effective": 103 + }, + "openaimmlu_business_ethics": { + "original": 100, + "effective": 100 + }, + "openaimmlu_security_studies": { + "original": 245, + "effective": 245 + }, + "openaimmlu_moral_scenarios": { + "original": 895, + "effective": 895 + }, + "openaimmlu_marketing": { + "original": 234, + "effective": 234 + }, + "openaimmlu_high_school_government_and_politics": { + "original": 193, + "effective": 193 + }, + "openaimmlu_public_relations": { + "original": 110, + "effective": 110 + }, + "openaimmlu_high_school_microeconomics": { + "original": 238, + "effective": 238 + }, + "openaimmlu_us_foreign_policy": { + "original": 100, + "effective": 100 + }, + "openaimmlu_high_school_macroeconomics": { + "original": 390, + "effective": 390 + }, + "openaimmlu_moral_disputes": { + "original": 346, + "effective": 346 + }, + "openaimmlu_human_sexuality": { + "original": 131, + "effective": 131 + }, + "openaimmlu_sociology": { + "original": 201, + "effective": 201 + }, + "openaimmlu_international_law": { + "original": 121, + "effective": 121 + }, + "openaimmlu_jurisprudence": { + "original": 108, + "effective": 108 + }, + "openaimmlu_high_school_world_history": { + "original": 237, + "effective": 237 + }, + "openaimmlu_prehistory": { + "original": 324, + "effective": 324 + }, + "openaimmlu_world_religions": { + "original": 171, + "effective": 171 + }, + "openaimmlu_philosophy": { + "original": 311, + "effective": 311 + }, + "openaimmlu_logical_fallacies": { + "original": 163, + "effective": 163 + }, + "openaimmlu_high_school_european_history": { + "original": 165, + "effective": 165 + }, + "openaimmlu_high_school_us_history": { + "original": 204, + "effective": 204 + } + }, + "config": { + "model": "hf", + "model_args": "pretrained=tiiuae/Falcon3-7B-Instruct,trust_remote_code=True,cache_dir=/tmp,parallelize=True", + "model_num_parameters": 7455550464, + "model_dtype": "torch.bfloat16", + "model_revision": "main", + "model_sha": "5563a370c1848366c7a095bde4bbff2cdb419cc6", + "batch_size": 1, + "batch_sizes": [], + "device": null, + "use_cache": null, + "limit": null, + "bootstrap_iters": 100000, + "gen_kwargs": null, + "random_seed": 0, + "numpy_seed": 1234, + "torch_seed": 1234, + "fewshot_seed": 1234 + }, + "git_hash": "5e10e017", + "date": 1736890748.3267176, + "pretty_env_info": "PyTorch version: 2.4.0+cu121\nIs debug build: False\nCUDA used to build PyTorch: 12.1\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.3 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: Could not collect\nCMake version: version 3.27.1\nLibc version: glibc-2.35\n\nPython version: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0] (64-bit runtime)\nPython platform: Linux-5.15.0-1064-azure-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: 12.2.128\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA A100-SXM4-80GB\nGPU 1: NVIDIA A100-SXM4-80GB\nGPU 2: NVIDIA A100-SXM4-80GB\nGPU 3: NVIDIA A100-SXM4-80GB\nGPU 4: NVIDIA A100-SXM4-80GB\nGPU 5: NVIDIA A100-SXM4-80GB\nGPU 6: NVIDIA A100-SXM4-80GB\nGPU 7: NVIDIA A100-SXM4-80GB\n\nNvidia driver version: 535.161.08\ncuDNN version: Probably one of the following:\n/usr/lib/x86_64-linux-gnu/libcudnn.so.8.9.4\n/usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.9.4\n/usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.9.4\n/usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.9.4\n/usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.9.4\n/usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.9.4\n/usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.9.4\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture: x86_64\nCPU op-mode(s): 32-bit, 64-bit\nAddress sizes: 48 bits physical, 48 bits virtual\nByte Order: Little Endian\nCPU(s): 96\nOn-line CPU(s) list: 0-95\nVendor ID: AuthenticAMD\nModel name: AMD EPYC 7V12 64-Core Processor\nCPU family: 23\nModel: 49\nThread(s) per core: 1\nCore(s) per socket: 48\nSocket(s): 2\nStepping: 0\nBogoMIPS: 4890.88\nFlags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl tsc_reliable nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw topoext perfctr_core ssbd vmmcall fsgsbase bmi1 avx2 smep bmi2 rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru arat umip rdpid\nHypervisor vendor: Microsoft\nVirtualization type: full\nL1d cache: 3 MiB (96 instances)\nL1i cache: 3 MiB (96 instances)\nL2 cache: 48 MiB (96 instances)\nL3 cache: 384 MiB (24 instances)\nNUMA node(s): 4\nNUMA node0 CPU(s): 0-23\nNUMA node1 CPU(s): 24-47\nNUMA node2 CPU(s): 48-71\nNUMA node3 CPU(s): 72-95\nVulnerability Gather data sampling: Not affected\nVulnerability Itlb multihit: Not affected\nVulnerability L1tf: Not affected\nVulnerability Mds: Not affected\nVulnerability Meltdown: Not affected\nVulnerability Mmio stale data: Not affected\nVulnerability Retbleed: Mitigation; untrained return thunk; SMT disabled\nVulnerability Spec rstack overflow: Mitigation; safe RET, no microcode\nVulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp\nVulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2: Mitigation; Retpolines; STIBP disabled; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds: Not affected\nVulnerability Tsx async abort: Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==1.26.4\n[pip3] onnx==1.14.0\n[pip3] pytorch-lightning==2.0.7\n[pip3] pytorch-quantization==2.1.2\n[pip3] torch==2.4.0\n[pip3] torch-tensorrt==2.0.0.dev0\n[pip3] torchaudio==2.1.0\n[pip3] torchdata==0.7.0a0\n[pip3] torchmetrics==1.2.0\n[pip3] torchvision==0.19.0\n[pip3] triton==3.0.0\n[conda] Could not collect", + "transformers_version": "4.48.0", + "upper_git_hash": "f64fe2f2a86055aaecced603b56097fd79201711", + "tokenizer_pad_token": [ + "<|pad|>", + "2023" + ], + "tokenizer_eos_token": [ + "<|endoftext|>", + "11" + ], + "tokenizer_bos_token": [ + null, + "None" + ], + "eot_token_id": 11, + "max_length": 32768, + "task_hashes": { + "openaimmlu_high_school_mathematics": "df02371376ec95c9406e2ff6a36bf2a1ea28d1928668f0b3847898078241cd38", + "openaimmlu_college_physics": "35036c7ee551f577e536b265e4c19d6113e6100623a7e81e090dce664abda802", + "openaimmlu_computer_security": "160f20db5ddd067bb31a2fc75d678e5e292d74015bb7030b9aa0ea2eb850530b", + "openaimmlu_college_computer_science": "c40b1f441d5371cb93738d6ab836aecd34aaea10ac6cbae949c195b033054bfe", + "openaimmlu_abstract_algebra": "06a0f6ff7a57c59fb93b15c6ecb5f30709bc2156d0083e8c3a244e40e7f68a91", + "openaimmlu_high_school_statistics": "d1413e0f4bfdffb9e7b5926df92fd704175659a6c09f7a6269389ba41485c049", + "openaimmlu_college_mathematics": "5ed7c30b039bac914edd2cf744db5c5ff81cf29ff84181f69ea7bc1ee92d59dc", + "openaimmlu_college_chemistry": "07f8c55dbe5a1b2b827229d50416d4a998d08a0ffaddc6c42f47fab58de668ac", + "openaimmlu_high_school_computer_science": "d918e63bf3588fe06defe04a12d3e016bfdaad25ffe9fe242ee46b263f46f9b2", + "openaimmlu_elementary_mathematics": "96ccc5e84dc3ec5cc997298683bb38aeae06c965c11866382ed573cf79958544", + "openaimmlu_high_school_physics": "922db7807587177d039bb3bbc3f986ea29ff96b64b45816cda8a649950ded2f2", + "openaimmlu_conceptual_physics": "84cfafbb3a9c37c0067098210a14d8297c3d4477276b594a4f7fa40e5a4c43a2", + "openaimmlu_econometrics": "b89ef7b8e3fab62d8568d7a74893cf7b69997f3069aa681a263bffbc24ac091f", + "openaimmlu_college_biology": "e84687bbe74d124d198f791d2108a4caf5798f3bb803529aed5bae8939345e09", + "openaimmlu_electrical_engineering": "d613f287d6cb53521193eaeafef9b1e3bc4d23fec81af05f46df581d85e28930", + "openaimmlu_astronomy": "592a0cb02318597a452e074f3d04930eb7b9a12cb492f9ab16aa825bd2d44c1e", + "openaimmlu_high_school_chemistry": "04bb34f39ab15ad295823eab6765eba7829393b8aeffb610107a3b52aa75789c", + "openaimmlu_high_school_biology": "8f6f966ffa66e26cdd3184854b329477f532e2bc0d3124cdd522f0a4372d524c", + "openaimmlu_miscellaneous": "75f78a28f6382ee54628eed866c1f5cad54c9004544d1e0e50bfd43de86dec73", + "openaimmlu_professional_law": "56fecd11891fc1a1618e2f4cba7f74b01ce5c78b33a6bbd05f56509225476cd3", + "openaimmlu_machine_learning": "4c7c4b0fe2e7df74dfb11cfd51543ec0c2aa90c8e3c4b357efa131ed27e1d802", + "openaimmlu_global_facts": "990031c4f7667bbe547f384284195656902d499683bae0de3bed83401d012307", + "openaimmlu_anatomy": "b8b9a59680da920ace7de601112bae3ddfeb9373823206cfd3386a3a116bf5ba", + "openaimmlu_college_medicine": "895e37b75ab276e4af42f382cf38ebff5905ddba3257b4584646e774d669966d", + "openaimmlu_human_aging": "c7dd36b2b571ec34405e12c7b30fbce83ce41cf80c9082a7070d871f1d567a4f", + "openaimmlu_formal_logic": "37cadc6de2615a1fb88b3f2b3ceafe577573c3c8c664e1a88115141ec1814cef", + "openaimmlu_professional_accounting": "1b66076938a73ae289e8c741b5e9c5273b8e6335e4a03e6f8557de75777cb9e9", + "openaimmlu_high_school_psychology": "89965b10f9d92f9da802d530f76075b98c60656759c1b94933e30d1d5a97e275", + "openaimmlu_clinical_knowledge": "478cd76f6ae8f18be117885de984c5c24d2afa6b0a03569f6cdee8ce54a4723c", + "openaimmlu_professional_psychology": "6ba2fa8d5e37978e940143629fd6c77c948c1ff73d0b6b27cd689a3a73b014d6", + "openaimmlu_medical_genetics": "7ffc6e6eb0ecade753d830cea0ce3b782234414a87013916c52d098a10036b1c", + "openaimmlu_virology": "f4fffd1b1e41a95fe22eab417715b1f89db7dcaf396450bb54217becba6bea48", + "openaimmlu_professional_medicine": "54806e63b4341eae4298c537e02802bebd754820aadefac78c322ac671e91b75", + "openaimmlu_nutrition": "b8df6884453ddb38c0801500887a1e8389b44ded3851ea8134112a7ab6e6a9dd", + "openaimmlu_high_school_geography": "ee58d4e56a8653d4eb3f9be1c1dfbb1caf93809ffd956f77fbfa2e9fc8e027e9", + "openaimmlu_management": "9a4a7c9a8ec87b9cb943c4673386e6317eacd582c93951950ef0ff50c474ee38", + "openaimmlu_business_ethics": "aa2fe1f91fabbe1315984d1d654347d7e1a682bb4298ff7e1859a6a968bf1246", + "openaimmlu_security_studies": "78d64883d2d76efc502ef6e9489a7a19ed54ee4a6cb07b442db97a749ddcd0f9", + "openaimmlu_moral_scenarios": "969d4d5af11704d0747cf141292fac7ed5f12fc2a3ed393bc7bce8fa2b89665c", + "openaimmlu_marketing": "7ec3711e36110aace63e7bf63697943476567d2a99bd85a46a7050a72cc7dd3d", + "openaimmlu_high_school_government_and_politics": "f68538a9c88b33bc7ca2fc71b48be912c05d21068090825f601376d754fdbd0e", + "openaimmlu_public_relations": "b876ceb0f0f6c77cc34d430093471115271df1b78d9630fc9c56c50cd905ab6f", + "openaimmlu_high_school_microeconomics": "608604afeb38c3f75321e6387647d8f9ff7114648d1063bb03821ec734a09205", + "openaimmlu_us_foreign_policy": "ca19e9ea92549f964755bf00cb8b78af81dbb36ae21a866a3729c8d2c7dc8fe8", + "openaimmlu_high_school_macroeconomics": "8521c25e6160374c93bab3ec5f0f2c8379a9baa5d1b9bbf4833f2fc7447721d8", + "openaimmlu_moral_disputes": "971dda048fe6a7964cb6426e830e6fc2c434b913b80313d809d8ee5bbe9ba8dd", + "openaimmlu_human_sexuality": "b184a75c6c862f5e3954c7933db056f81cbafdd28c84f9106801dcda047eb62a", + "openaimmlu_sociology": "c4b3942888782b4892aa35338f4b019277be2647118282a01ca6d4247341d655", + "openaimmlu_international_law": "f6989bdda04b24d24bd12a8a9c89552374071ab9b67476934ced71fcc9295030", + "openaimmlu_jurisprudence": "46a22b82dd04e2c6c0948be36b4104906b7ebbc5e8f68e91d32c49241548500d", + "openaimmlu_high_school_world_history": "252ee950c7a63347297c321eb62582d5efb9a816d6561da934f14f87f8203ace", + "openaimmlu_prehistory": "8d6d577689ab2cebbf7b19bc964bfaef4bef05e362395be7bb817ac23144687d", + "openaimmlu_world_religions": "5d9ac0fc4f9744e434d301ee543c5876c9d6f420d30a7643230ce157c6ca394f", + "openaimmlu_philosophy": "d217b4f827b712ce5f4a08e8fc26629a86c84f93b58bef9002ef376fceedc8ab", + "openaimmlu_logical_fallacies": "94ef523dd37da932e84321a1654eeb7b6797c37b4a05c0fd08885893d192b9e8", + "openaimmlu_high_school_european_history": "fb273e15b3cbc2c7b4af95fd69ea68fa995204b964acac0c8757920434f1bd36", + "openaimmlu_high_school_us_history": "637ba8e2d7ffdea5de66a1c7e2a314f3ec0e7808893d6269100b084bd5167e6e" + }, + "model_source": "hf", + "model_name": "tiiuae/Falcon3-7B-Instruct", + "model_name_sanitized": "tiiuae__Falcon3-7B-Instruct", + "system_instruction": null, + "system_instruction_sha": null, + "fewshot_as_multiturn": false, + "chat_template": null, + "chat_template_sha": null, + "start_time": 600998.580700401, + "end_time": 601190.357399357, + "total_evaluation_time_seconds": "191.77669895603321" +} \ No newline at end of file