遇见数据集

open-llm-leaderboard/details_aisquared__dlite-v1-1_5b

收藏
Hugging Face2023-09-23 更新2024-03-04 收录
官方服务:

资源简介:

--- pretty_name: Evaluation run of aisquared/dlite-v1-1_5b dataset_summary: "Dataset automatically created during the evaluation run of model\ \ [aisquared/dlite-v1-1_5b](https://huggingface.co/aisquared/dlite-v1-1_5b) on the\ \ [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\ \nThe dataset is composed of 64 configuration, each one coresponding to one of the\ \ evaluated task.\n\nThe dataset has been created from 2 run(s). Each run can be\ \ found as a specific split in each configuration, the split being named using the\ \ timestamp of the run.The \"train\" split is always pointing to the latest results.\n\ \nAn additional configuration \"results\" store all the aggregated results of the\ \ run (and is used to compute and display the agregated metrics on the [Open LLM\ \ Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\ \nTo load the details from a run, you can for instance do the following:\n```python\n\ from datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_aisquared__dlite-v1-1_5b\"\ ,\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\n\ These are the [latest results from run 2023-09-23T17:48:40.273494](https://huggingface.co/datasets/open-llm-leaderboard/details_aisquared__dlite-v1-1_5b/blob/main/results_2023-09-23T17-48-40.273494.json)(note\ \ that their might be results for other tasks in the repos if successive evals didn't\ \ cover the same tasks. You find each in the results and the \"latest\" split for\ \ each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.005977348993288591,\n\ \ \"em_stderr\": 0.0007893908687131983,\n \"f1\": 0.06289953859060417,\n\ \ \"f1_stderr\": 0.0015069024652225058,\n \"acc\": 0.28017386590137583,\n\ \ \"acc_stderr\": 0.00735524021281907\n },\n \"harness|drop|3\": {\n\ \ \"em\": 0.005977348993288591,\n \"em_stderr\": 0.0007893908687131983,\n\ \ \"f1\": 0.06289953859060417,\n \"f1_stderr\": 0.0015069024652225058\n\ \ },\n \"harness|gsm8k|5\": {\n \"acc\": 0.000758150113722517,\n \ \ \"acc_stderr\": 0.0007581501137225347\n },\n \"harness|winogrande|5\"\ : {\n \"acc\": 0.5595895816890292,\n \"acc_stderr\": 0.013952330311915607\n\ \ }\n}\n```" repo_url: https://huggingface.co/aisquared/dlite-v1-1_5b leaderboard_url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard point_of_contact: clementine@hf.co configs: - config_name: harness_arc_challenge_25 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|arc:challenge|25_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|arc:challenge|25_2023-07-19T15:22:45.415057.parquet' - config_name: harness_drop_3 data_files: - split: 2023_09_23T17_48_40.273494 path: - '**/details_harness|drop|3_2023-09-23T17-48-40.273494.parquet' - split: latest path: - '**/details_harness|drop|3_2023-09-23T17-48-40.273494.parquet' - config_name: harness_gsm8k_5 data_files: - split: 2023_09_23T17_48_40.273494 path: - '**/details_harness|gsm8k|5_2023-09-23T17-48-40.273494.parquet' - split: latest path: - '**/details_harness|gsm8k|5_2023-09-23T17-48-40.273494.parquet' - config_name: harness_hellaswag_10 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hellaswag|10_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hellaswag|10_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-management|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-management|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-07-19T15:22:45.415057.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_abstract_algebra_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_anatomy_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_astronomy_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_business_ethics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_clinical_knowledge_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_college_biology_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_college_chemistry_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_college_computer_science_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_college_mathematics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_college_medicine_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_college_physics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_computer_security_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_conceptual_physics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_econometrics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_electrical_engineering_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_elementary_mathematics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_formal_logic_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_global_facts_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_biology_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_chemistry_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_computer_science_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_european_history_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_geography_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_government_and_politics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_macroeconomics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_mathematics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_microeconomics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_physics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_psychology_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_statistics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_us_history_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_high_school_world_history_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_human_aging_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_human_sexuality_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_international_law_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_jurisprudence_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_logical_fallacies_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_machine_learning_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_management_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-management|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-management|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_marketing_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_medical_genetics_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_miscellaneous_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_moral_disputes_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_moral_scenarios_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_nutrition_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_philosophy_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_prehistory_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_professional_accounting_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_professional_law_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_professional_medicine_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_professional_psychology_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_public_relations_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_security_studies_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_sociology_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_us_foreign_policy_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_virology_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-virology|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-virology|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_hendrycksTest_world_religions_5 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T15:22:45.415057.parquet' - config_name: harness_truthfulqa_mc_0 data_files: - split: 2023_07_19T15_22_45.415057 path: - '**/details_harness|truthfulqa:mc|0_2023-07-19T15:22:45.415057.parquet' - split: latest path: - '**/details_harness|truthfulqa:mc|0_2023-07-19T15:22:45.415057.parquet' - config_name: harness_winogrande_5 data_files: - split: 2023_09_23T17_48_40.273494 path: - '**/details_harness|winogrande|5_2023-09-23T17-48-40.273494.parquet' - split: latest path: - '**/details_harness|winogrande|5_2023-09-23T17-48-40.273494.parquet' - config_name: results data_files: - split: 2023_07_19T15_22_45.415057 path: - results_2023-07-19T15:22:45.415057.parquet - split: 2023_09_23T17_48_40.273494 path: - results_2023-09-23T17-48-40.273494.parquet - split: latest path: - results_2023-09-23T17-48-40.273494.parquet --- # Dataset Card for Evaluation run of aisquared/dlite-v1-1_5b ## Dataset Description - **Homepage:** - **Repository:** https://huggingface.co/aisquared/dlite-v1-1_5b - **Paper:** - **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard - **Point of Contact:** clementine@hf.co ### Dataset Summary Dataset automatically created during the evaluation run of model [aisquared/dlite-v1-1_5b](https://huggingface.co/aisquared/dlite-v1-1_5b) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard). The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task. The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results. An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)). To load the details from a run, you can for instance do the following: ```python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_aisquared__dlite-v1-1_5b", "harness_winogrande_5", split="train") ``` ## Latest results These are the [latest results from run 2023-09-23T17:48:40.273494](https://huggingface.co/datasets/open-llm-leaderboard/details_aisquared__dlite-v1-1_5b/blob/main/results_2023-09-23T17-48-40.273494.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval): ```python { "all": { "em": 0.005977348993288591, "em_stderr": 0.0007893908687131983, "f1": 0.06289953859060417, "f1_stderr": 0.0015069024652225058, "acc": 0.28017386590137583, "acc_stderr": 0.00735524021281907 }, "harness|drop|3": { "em": 0.005977348993288591, "em_stderr": 0.0007893908687131983, "f1": 0.06289953859060417, "f1_stderr": 0.0015069024652225058 }, "harness|gsm8k|5": { "acc": 0.000758150113722517, "acc_stderr": 0.0007581501137225347 }, "harness|winogrande|5": { "acc": 0.5595895816890292, "acc_stderr": 0.013952330311915607 } } ``` ### Supported Tasks and Leaderboards [More Information Needed] ### Languages [More Information Needed] ## Dataset Structure ### Data Instances [More Information Needed] ### Data Fields [More Information Needed] ### Data Splits [More Information Needed] ## Dataset Creation ### Curation Rationale [More Information Needed] ### Source Data #### Initial Data Collection and Normalization [More Information Needed] #### Who are the source language producers? [More Information Needed] ### Annotations #### Annotation process [More Information Needed] #### Who are the annotators? [More Information Needed] ### Personal and Sensitive Information [More Information Needed] ## Considerations for Using the Data ### Social Impact of Dataset [More Information Needed] ### Discussion of Biases [More Information Needed] ### Other Known Limitations [More Information Needed] ## Additional Information ### Dataset Curators [More Information Needed] ### Licensing Information [More Information Needed] ### Citation Information [More Information Needed] ### Contributions [More Information Needed]

提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

该数据集是在评估模型 aisquared/dlite-v1-1_5bOpen LLM Leaderboard 上的运行过程中自动创建的。

数据集组成

  • 数据集包含 64 个配置,每个配置对应一个评估任务。
  • 数据集由 2 次运行创建,每次运行可以在每个配置中找到特定的分割,分割名称使用运行的时间戳。
  • "train" 分割始终指向最新的结果。
  • 一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示 Open LLM Leaderboard 上的聚合指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_aisquared__dlite-v1-1_5b", "harness_winogrande_5", split="train")

最新结果

以下是 2023-09-23T17:48:40.273494 运行 的最新结果: python { "all": { "em": 0.005977348993288591, "em_stderr": 0.0007893908687131983, "f1": 0.06289953859060417, "f1_stderr": 0.0015069024652225058, "acc": 0.28017386590137583, "acc_stderr": 0.00735524021281907 }, "harness|drop|3": { "em": 0.005977348993288591, "em_stderr": 0.0007893908687131983, "f1": 0.06289953859060417, "f1_stderr": 0.0015069024652225058 }, "harness|gsm8k|5": { "acc": 0.000758150113722517, "acc_stderr": 0.0007581501137225347 }, "harness|winogrande|5": { "acc": 0.5595895816890292, "acc_stderr": 0.013952330311915607 } }

配置详情

  • harness_arc_challenge_25

    • 分割: 2023_07_19T15_22_45.415057
    • 路径: **/details_harness|arc:challenge|25_2023-07-19T15:22:45.415057.parquet
    • 分割: latest
    • 路径: **/details_harness|arc:challenge|25_2023-07-19T15:22:45.415057.parquet
  • harness_drop_3

    • 分割: 2023_09_23T17_48_40.273494
    • 路径: **/details_harness|drop|3_2023-09-23T17-48-40.273494.parquet
    • 分割: latest
    • 路径: **/details_harness|drop|3_2023-09-23T17-48-40.273494.parquet
  • harness_gsm8k_5

    • 分割: 2023_09_23T17_48_40.273494
    • 路径: **/details_harness|gsm8k|5_2023-09-23T17-48-40.273494.parquet
    • 分割: latest
    • 路径: **/details_harness|gsm8k|5_2023-09-23T17-48-40.273494.parquet
  • harness_hellaswag_10

    • 分割: 2023_07_19T15_22_45.415057
    • 路径: **/details_harness|hellaswag|10_2023-07-19T15:22:45.415057.parquet
    • 分割: latest
    • 路径: **/details_harness|hellaswag|10_2023-07-19T15:22:45.415057.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_07_19T15_22_45.415057
    • 路径:
      • **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-anatomy|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-astronomy|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-business_ethics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-college_biology|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-college_medicine|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-college_physics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-computer_security|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-econometrics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-formal_logic|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-global_facts|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-human_aging|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-international_law|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-machine_learning|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-management|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-marketing|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-nutrition|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-philosophy|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-prehistory|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T15:22:45.415057.parquet
      • **/details_harness|hendrycksTest-professional_law|5_2023-07-19T15:22:45.415057.parquet
      • `**/details_har
搜集汇总
数据集介绍
open-llm-leaderboard/details_aisquared__dlite-v1-1_5b 数据集图片
构建方式
在大型语言模型评测领域,Open LLM Leaderboard为模型性能的标准化评估提供了重要平台。该数据集是在对aisquared/dlite-v1-1_5b模型进行自动化评测的过程中自动生成的,其构建方式依托于HuggingFace的评估框架。数据集包含64个配置,每个配置对应一个被评估的特定任务,如ARC-Challenge、DROP、GSM8K、Winogrande等。每次评测运行结果作为独立的分割(split)存储,分割名称以运行时间戳命名,而'train'分割则始终指向最新一次的评测结果。此外,一个名为'results'的额外配置汇总了所有运行的整体聚合指标,用于在排行榜上展示综合性能。所有数据以Parquet格式存储,确保了数据的高效读写与可扩展性。
特点
该数据集的核心特点在于其结构化的评测记录与版本化追踪能力。数据集为每项任务提供了独立的配置,使得研究者可以精细地分析模型在不同基准上的表现。每一次评测运行均作为独立分割保存,便于进行历史结果对比与回归分析,时间戳命名机制确保了数据版本的可追溯性。'train'分割动态指向最新结果,简化了获取当前最优性能的流程。'results'配置进一步聚合了全局指标,如精确匹配率(EM)、F1分数和准确率(Acc)及其标准误差,为模型整体能力的宏观评估提供了直观依据。这种设计兼顾了细粒度分析与宏观比较的双重需求。
使用方法
使用该数据集时,研究者可通过HuggingFace的datasets库便捷地加载特定任务的评测详情。例如,通过指定配置名称(如'harness_winogrande_5')和分割(如'train'),即可获取模型在Winogrande任务上的最新运行结果。加载的数据以DataFrame形式呈现,包含各样本的详细评测指标,便于进行深入的错误分析或性能诊断。对于需要比较不同运行版本的用户,可通过时间戳分割名访问历史数据。聚合结果则可通过加载'results'配置获得,适用于生成模型性能概览或进行跨模型对比。该数据集的设计充分考虑了易用性与灵活性,降低了研究者进行模型评估分析的门槛。
背景与挑战
背景概述
在大规模语言模型迅猛发展的时代背景下,如何系统、公正地评估模型的多维能力成为学界与工业界共同关注的核心课题。Open LLM Leaderboard由Hugging Face团队于2023年创建,旨在通过标准化评测流程,为开源语言模型提供可复现的横向对比基准。该数据集记录了aisquared团队开发的dlite-v1-1_5b模型在Leaderboard上的完整评估结果,涵盖ARC挑战、DROP、GSM8K、HellaSwag及MMLU等多项经典自然语言理解与推理任务。通过存储每次运行的任务级细节与聚合指标,该数据集为研究社区追踪模型性能演变、分析不同任务上的优劣势提供了宝贵资源,推动了开源大模型评估体系的透明化与规范化。
当前挑战
该数据集所服务的核心领域挑战在于,当前语言模型评估缺乏统一、可复现的基准,不同研究者在任务选择、提示设计、评估指标上的差异导致结果难以直接比较。构建过程中,数据集面临多重困难:需将模型在64个不同配置下的推理结果进行结构化存储,每个任务对应独立的评估配置与数据分片;同时要处理多次运行的时间戳管理,确保历史结果可追溯且最新结果易获取。此外,评估结果中稀疏的任务覆盖(如DROP与GSM8K仅有一轮运行数据)和指标统计的精确性(如准确率与标准误的细粒度计算)也增加了数据整合与验证的复杂度。
常用场景
经典使用场景
在大型语言模型的评估体系中,open-llm-leaderboard/details_aisquared__dlite-v1-1_5b 数据集扮演着关键的角色。该数据集专门为记录和复现模型在 Open LLM Leaderboard 上的评估过程而设计,覆盖了 ARC-Challenge、HellaSwag、MMLU、GSM8K 等数十项经典自然语言理解与推理任务。研究者可借助该数据集,逐项审视模型在每项任务上的细粒度表现,并基于同一套评估框架进行公平的横向比较。其经典的使用方式在于,通过加载特定配置和分片,复现模型在某一时刻的完整评测结果,从而为模型性能的追踪与改进提供坚实的数据基础。
实际应用
在实际应用层面,该数据集为模型选型与部署提供了量化的决策依据。企业在构建基于大语言模型的应用系统时,需审慎评估不同候选模型在特定任务上的真实表现。通过分析该数据集记录的细粒度评测结果,开发团队能够直观地比较不同模型在数学推理、常识问答、阅读理解等关键能力上的优劣,从而选择最贴合业务需求的模型。例如,在需要高精度数学解题的场景中,GSM8K 任务的准确率便成为关键的筛选指标。这种基于实证数据的选型策略,有效降低了模型部署过程中的不确定性,提升了应用落地的效率与可靠性。
衍生相关工作
该数据集的诞生直接催生了一系列关于大语言模型评估方法论与性能分析的研究工作。围绕其提供的丰富评测记录,衍生出了诸如模型能力图谱构建、不同参数量级模型性能缩放律的实证分析、以及评估任务难度与模型表现之间关联性的探索。此外,该数据集也促进了评估基准本身的演进,研究者通过分析模型在各项任务上的失败案例,推动了如 MMLU-Pro、GSM8K 增强版等更具挑战性评测集的提出。这些后续工作共同深化了学术界对大语言模型能力本质的理解,并为更科学的评估体系的建立奠定了坚实基础。
以上内容由遇见数据集搜集并总结生成
二维码
社区交流群
二维码
科研交流群
商业服务