five

open-llm-leaderboard/details_cmarkea__bloomz-560m-sft-chat

收藏
Hugging Face2023-10-25 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_cmarkea__bloomz-560m-sft-chat
下载链接
链接失效反馈
官方服务:
资源简介:
--- pretty_name: Evaluation run of cmarkea/bloomz-560m-sft-chat dataset_summary: "Dataset automatically created during the evaluation run of model\ \ [cmarkea/bloomz-560m-sft-chat](https://huggingface.co/cmarkea/bloomz-560m-sft-chat)\ \ 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_cmarkea__bloomz-560m-sft-chat\"\ ,\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\n\ These are the [latest results from run 2023-10-25T06:48:45.798590](https://huggingface.co/datasets/open-llm-leaderboard/details_cmarkea__bloomz-560m-sft-chat/blob/main/results_2023-10-25T06-48-45.798590.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\": 0.09626677852348993,\n\ \ \"em_stderr\": 0.003020633220463166,\n \"f1\": 0.1512867030201341,\n\ \ \"f1_stderr\": 0.0032234786448698083,\n \"acc\": 0.2675611681136543,\n\ \ \"acc_stderr\": 0.0070088865604407986\n },\n \"harness|drop|3\":\ \ {\n \"em\": 0.09626677852348993,\n \"em_stderr\": 0.003020633220463166,\n\ \ \"f1\": 0.1512867030201341,\n \"f1_stderr\": 0.0032234786448698083\n\ \ },\n \"harness|gsm8k|5\": {\n \"acc\": 0.0,\n \"acc_stderr\"\ : 0.0\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.5351223362273086,\n\ \ \"acc_stderr\": 0.014017773120881597\n }\n}\n```" repo_url: https://huggingface.co/cmarkea/bloomz-560m-sft-chat 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_10_04T03_35_59.039004 path: - '**/details_harness|arc:challenge|25_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|arc:challenge|25_2023-10-04T03-35-59.039004.parquet' - config_name: harness_drop_3 data_files: - split: 2023_10_25T06_48_45.798590 path: - '**/details_harness|drop|3_2023-10-25T06-48-45.798590.parquet' - split: latest path: - '**/details_harness|drop|3_2023-10-25T06-48-45.798590.parquet' - config_name: harness_gsm8k_5 data_files: - split: 2023_10_25T06_48_45.798590 path: - '**/details_harness|gsm8k|5_2023-10-25T06-48-45.798590.parquet' - split: latest path: - '**/details_harness|gsm8k|5_2023-10-25T06-48-45.798590.parquet' - config_name: harness_hellaswag_10 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hellaswag|10_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hellaswag|10_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-management|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-management|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-10-04T03-35-59.039004.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_abstract_algebra_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_anatomy_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-anatomy|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-anatomy|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_astronomy_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-astronomy|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-astronomy|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_business_ethics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_clinical_knowledge_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_college_biology_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-college_biology|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_biology|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_college_chemistry_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_college_computer_science_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_college_mathematics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_college_medicine_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_college_physics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-college_physics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_physics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_computer_security_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-computer_security|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-computer_security|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_conceptual_physics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_econometrics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-econometrics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-econometrics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_electrical_engineering_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_elementary_mathematics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_formal_logic_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_global_facts_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-global_facts|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-global_facts|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_biology_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_chemistry_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_computer_science_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_european_history_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_geography_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_government_and_politics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_macroeconomics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_mathematics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_microeconomics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_physics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_psychology_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_statistics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_us_history_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_high_school_world_history_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_human_aging_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-human_aging|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_aging|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_human_sexuality_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_international_law_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-international_law|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-international_law|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_jurisprudence_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_logical_fallacies_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_machine_learning_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_management_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-management|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-management|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_marketing_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-marketing|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-marketing|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_medical_genetics_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_miscellaneous_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_moral_disputes_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_moral_scenarios_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_nutrition_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-nutrition|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-nutrition|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_philosophy_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-philosophy|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-philosophy|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_prehistory_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-prehistory|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-prehistory|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_professional_accounting_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_professional_law_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-professional_law|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_law|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_professional_medicine_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_professional_psychology_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_public_relations_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-public_relations|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-public_relations|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_security_studies_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-security_studies|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-security_studies|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_sociology_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-sociology|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-sociology|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_us_foreign_policy_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_virology_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-virology|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-virology|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_hendrycksTest_world_religions_5 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|hendrycksTest-world_religions|5_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|hendrycksTest-world_religions|5_2023-10-04T03-35-59.039004.parquet' - config_name: harness_truthfulqa_mc_0 data_files: - split: 2023_10_04T03_35_59.039004 path: - '**/details_harness|truthfulqa:mc|0_2023-10-04T03-35-59.039004.parquet' - split: latest path: - '**/details_harness|truthfulqa:mc|0_2023-10-04T03-35-59.039004.parquet' - config_name: harness_winogrande_5 data_files: - split: 2023_10_25T06_48_45.798590 path: - '**/details_harness|winogrande|5_2023-10-25T06-48-45.798590.parquet' - split: latest path: - '**/details_harness|winogrande|5_2023-10-25T06-48-45.798590.parquet' - config_name: results data_files: - split: 2023_10_04T03_35_59.039004 path: - results_2023-10-04T03-35-59.039004.parquet - split: 2023_10_25T06_48_45.798590 path: - results_2023-10-25T06-48-45.798590.parquet - split: latest path: - results_2023-10-25T06-48-45.798590.parquet --- # Dataset Card for Evaluation run of cmarkea/bloomz-560m-sft-chat ## Dataset Description - **Homepage:** - **Repository:** https://huggingface.co/cmarkea/bloomz-560m-sft-chat - **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 [cmarkea/bloomz-560m-sft-chat](https://huggingface.co/cmarkea/bloomz-560m-sft-chat) 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_cmarkea__bloomz-560m-sft-chat", "harness_winogrande_5", split="train") ``` ## Latest results These are the [latest results from run 2023-10-25T06:48:45.798590](https://huggingface.co/datasets/open-llm-leaderboard/details_cmarkea__bloomz-560m-sft-chat/blob/main/results_2023-10-25T06-48-45.798590.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": 0.09626677852348993, "em_stderr": 0.003020633220463166, "f1": 0.1512867030201341, "f1_stderr": 0.0032234786448698083, "acc": 0.2675611681136543, "acc_stderr": 0.0070088865604407986 }, "harness|drop|3": { "em": 0.09626677852348993, "em_stderr": 0.003020633220463166, "f1": 0.1512867030201341, "f1_stderr": 0.0032234786448698083 }, "harness|gsm8k|5": { "acc": 0.0, "acc_stderr": 0.0 }, "harness|winogrande|5": { "acc": 0.5351223362273086, "acc_stderr": 0.014017773120881597 } } ``` ### 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

该数据集是在对模型 cmarkea/bloomz-560m-sft-chat 进行评估运行期间自动创建的,用于 Open LLM Leaderboard

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。
  • 数据集从 2 次运行中创建,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_cmarkea__bloomz-560m-sft-chat", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-25T06:48:45.798590 运行 的最新结果:

python { "all": { "em": 0.09626677852348993, "em_stderr": 0.003020633220463166, "f1": 0.1512867030201341, "f1_stderr": 0.0032234786448698083, "acc": 0.2675611681136543, "acc_stderr": 0.0070088865604407986 }, "harness|drop|3": { "em": 0.09626677852348993, "em_stderr": 0.003020633220463166, "f1": 0.1512867030201341, "f1_stderr": 0.0032234786448698083 }, "harness|gsm8k|5": { "acc": 0.0, "acc_stderr": 0.0 }, "harness|winogrande|5": { "acc": 0.5351223362273086, "acc_stderr": 0.014017773120881597 } }

配置详情

  • harness_arc_challenge_25

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|arc:challenge|25_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|arc:challenge|25_2023-10-04T03-35-59.039004.parquet
  • harness_drop_3

    • 分割:2023_10_25T06_48_45.798590
    • 路径:**/details_harness|drop|3_2023-10-25T06-48-45.798590.parquet
    • 分割:latest
    • 路径:**/details_harness|drop|3_2023-10-25T06-48-45.798590.parquet
  • harness_gsm8k_5

    • 分割:2023_10_25T06_48_45.798590
    • 路径:**/details_harness|gsm8k|5_2023-10-25T06-48-45.798590.parquet
    • 分割:latest
    • 路径:**/details_harness|gsm8k|5_2023-10-25T06-48-45.798590.parquet
  • harness_hellaswag_10

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hellaswag|10_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hellaswag|10_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_anatomy_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_astronomy_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_college_biology_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_college_physics_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_computer_security_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_econometrics_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-10-04T03-35-59.039004.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分割:2023_10_04T03_35_59.039004
    • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T03-35-59.039004.parquet
    • 分割:latest
    • 路径:`**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T03-35-59.0
搜集汇总
数据集介绍
main_image_url
构建方式
该数据集是在Open LLM Leaderboard平台上对cmarkea/bloomz-560m-sft-chat模型进行自动化评估的过程中生成的。数据集包含64个配置项,每个配置对应一个评估任务,这些任务涵盖了如ARC挑战、DROP、GSM8K、HellaSwag以及HendrycksTest等多个基准测试。数据通过两次运行产生,每次运行的结果被存储为每个配置下的独立分割,分割名称以运行的时间戳命名。此外,一个名为'results'的额外配置汇总了所有运行的聚合结果,用于计算和展示排行榜上的综合指标。
特点
该数据集最显著的特点是其结构化的多任务评估设计。每个配置项均对应一个特定的评估任务,使得研究者能够针对性地分析模型在不同任务上的表现。数据集的'latest'分割始终指向最新的运行结果,便于追踪模型性能的演变。同时,'results'配置提供了跨任务的聚合指标,包括准确率、F1分数等统计量及其标准误差,为模型整体能力的量化评估提供了便利。
使用方法
研究者可通过HuggingFace的datasets库便捷地加载该数据集。例如,使用load_dataset函数并指定配置名称(如'harness_winogrande_5')和分割(如'train')即可获取特定任务的详细评估数据。数据集支持按任务或时间戳灵活选择数据,便于进行细粒度的性能分析或历史结果对比。此外,通过读取'results'配置中的聚合结果,可以快速获得模型在多个基准上的综合表现概览。
背景与挑战
背景概述
该数据集源自HuggingFace社区发起的Open LLM Leaderboard评测项目,由HuggingFace团队于2023年创建,旨在系统评估大语言模型在多样化任务上的表现。具体针对cmarkea/bloomz-560m-sft-chat这一经过指令微调的560M参数模型,数据集记录了其在ARC、HellaSwag、MMLU、GSM8K、Winogrande及DROP等多项基准测试中的详细结果。数据集通过自动化流程生成,包含64个配置项,分别对应不同评测任务,并存储了多次运行的性能数据。作为模型横向对比的标准化工具,该数据集为研究者提供了透明、可复现的评估框架,推动了多语言大模型在推理、常识理解与数学能力等维度的性能比较,对低资源语言模型的优化具有重要参考价值。
当前挑战
该数据集面临的挑战体现在多个层面。在领域问题层面,所评测的模型需要解决大语言模型在零样本或少样本场景下的通用推理能力,涵盖常识理解、数学运算、文本生成等任务,但模型在GSM8K数学推理任务上准确率为0,暴露出小规模模型在复杂推理中的根本性不足。在构建过程中,数据集需整合来自不同评测框架(如LM Evaluation Harness)的异构结果,确保各任务配置的兼容性与数据一致性;同时,评测结果随时间推移可能因模型更新或任务参数调整而产生偏差,需通过多运行次数据存储与版本管理来维护可追溯性,这对数据结构的鲁棒性提出了较高要求。
常用场景
经典使用场景
在大型语言模型评估的学术框架中,Open LLM Leaderboard上的评估结果数据集扮演着关键角色。该数据集记录了cmarkea/bloomz-560m-sft-chat模型在64个不同配置下的性能表现,涵盖ARC挑战、DROP、GSM8K、HellaSwag及HendrycksTest等多样化的自然语言理解与推理任务。研究者可借此深入剖析模型在常识推理、数学计算、文本理解及多领域知识问答等经典场景中的能力边界,为模型优化提供可复现的基准参考。
解决学术问题
该数据集有效回应了模型评估标准化与透明化的迫切需求。它解决了长期以来因评测流程不一致导致的模型性能难以横向比较的困境,通过统一的任务配置与结果聚合机制,为学术社区提供了一个严谨的模型能力诊断工具。其影响在于推动了语言模型评估范式的演进,使研究者能够基于细粒度的任务表现精准定位模型弱点,进而指导架构创新与训练策略的改进。
衍生相关工作
该数据集衍生了一系列重要的评估基准优化工作。例如,研究者基于其多任务配置结构,提出了更高效的任务采样策略与结果可视化方法;部分工作专注于分析不同推理步数(如3-shot与5-shot)对模型性能的影响,推动了少样本学习理论的实证研究。此外,数据集的时间戳分割机制启发了模型性能演化分析的研究方向,使得追踪模型在迭代过程中的能力变化成为可能。
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