遇见数据集

open-llm-leaderboard-old/details_TheBloke__koala-13B-HF

收藏
Hugging Face2023-10-22 更新2024-06-22 收录
官方服务:

资源简介:

--- pretty_name: Evaluation run of TheBloke/koala-13B-HF dataset_summary: "Dataset automatically created during the evaluation run of model\ \ [TheBloke/koala-13B-HF](https://huggingface.co/TheBloke/koala-13B-HF) 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_TheBloke__koala-13B-HF\"\ ,\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\n\ These are the [latest results from run 2023-10-22T08:43:38.346498](https://huggingface.co/datasets/open-llm-leaderboard/details_TheBloke__koala-13B-HF/blob/main/results_2023-10-22T08-43-38.346498.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\": 0.021707214765100673,\n\ \ \"em_stderr\": 0.0014923686874006184,\n \"f1\": 0.09106753355704705,\n\ \ \"f1_stderr\": 0.0020580604985252385,\n \"acc\": 0.40428250097386687,\n\ \ \"acc_stderr\": 0.009634029824810052\n },\n \"harness|drop|3\": {\n\ \ \"em\": 0.021707214765100673,\n \"em_stderr\": 0.0014923686874006184,\n\ \ \"f1\": 0.09106753355704705,\n \"f1_stderr\": 0.0020580604985252385\n\ \ },\n \"harness|gsm8k|5\": {\n \"acc\": 0.06823351023502654,\n \ \ \"acc_stderr\": 0.006945358944067431\n },\n \"harness|winogrande|5\"\ : {\n \"acc\": 0.7403314917127072,\n \"acc_stderr\": 0.012322700705552673\n\ \ }\n}\n```" repo_url: https://huggingface.co/TheBloke/koala-13B-HF 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_19T18_49_04.838102 path: - '**/details_harness|arc:challenge|25_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|arc:challenge|25_2023-07-19T18:49:04.838102.parquet' - config_name: harness_drop_3 data_files: - split: 2023_10_22T08_43_38.346498 path: - '**/details_harness|drop|3_2023-10-22T08-43-38.346498.parquet' - split: latest path: - '**/details_harness|drop|3_2023-10-22T08-43-38.346498.parquet' - config_name: harness_gsm8k_5 data_files: - split: 2023_10_22T08_43_38.346498 path: - '**/details_harness|gsm8k|5_2023-10-22T08-43-38.346498.parquet' - split: latest path: - '**/details_harness|gsm8k|5_2023-10-22T08-43-38.346498.parquet' - config_name: harness_hellaswag_10 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hellaswag|10_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hellaswag|10_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-management|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-management|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-07-19T18:49:04.838102.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_abstract_algebra_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_anatomy_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-anatomy|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_astronomy_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-astronomy|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_business_ethics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_clinical_knowledge_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_college_biology_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_biology|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_college_chemistry_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_college_computer_science_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_college_mathematics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_college_medicine_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_college_physics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_physics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_computer_security_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-computer_security|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_conceptual_physics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_econometrics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-econometrics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_electrical_engineering_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_elementary_mathematics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_formal_logic_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_global_facts_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-global_facts|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_biology_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_chemistry_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_computer_science_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_european_history_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_geography_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_government_and_politics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_macroeconomics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_mathematics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_microeconomics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_physics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_psychology_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_statistics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_us_history_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_high_school_world_history_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_human_aging_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_aging|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_human_sexuality_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_international_law_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-international_law|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_jurisprudence_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_logical_fallacies_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_machine_learning_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_management_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-management|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-management|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_marketing_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-marketing|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_medical_genetics_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_miscellaneous_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_moral_disputes_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_moral_scenarios_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_nutrition_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-nutrition|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_philosophy_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-philosophy|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_prehistory_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-prehistory|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_professional_accounting_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_professional_law_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_law|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_professional_medicine_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_professional_psychology_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_public_relations_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-public_relations|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_security_studies_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-security_studies|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_sociology_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-sociology|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_us_foreign_policy_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_virology_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-virology|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-virology|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_hendrycksTest_world_religions_5 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|hendrycksTest-world_religions|5_2023-07-19T18:49:04.838102.parquet' - config_name: harness_truthfulqa_mc_0 data_files: - split: 2023_07_19T18_49_04.838102 path: - '**/details_harness|truthfulqa:mc|0_2023-07-19T18:49:04.838102.parquet' - split: latest path: - '**/details_harness|truthfulqa:mc|0_2023-07-19T18:49:04.838102.parquet' - config_name: harness_winogrande_5 data_files: - split: 2023_10_22T08_43_38.346498 path: - '**/details_harness|winogrande|5_2023-10-22T08-43-38.346498.parquet' - split: latest path: - '**/details_harness|winogrande|5_2023-10-22T08-43-38.346498.parquet' - config_name: results data_files: - split: 2023_07_19T18_49_04.838102 path: - results_2023-07-19T18:49:04.838102.parquet - split: 2023_10_22T08_43_38.346498 path: - results_2023-10-22T08-43-38.346498.parquet - split: latest path: - results_2023-10-22T08-43-38.346498.parquet --- # Dataset Card for Evaluation run of TheBloke/koala-13B-HF ## Dataset Description - **Homepage:** - **Repository:** https://huggingface.co/TheBloke/koala-13B-HF - **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 [TheBloke/koala-13B-HF](https://huggingface.co/TheBloke/koala-13B-HF) 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_TheBloke__koala-13B-HF", "harness_winogrande_5", split="train") ``` ## Latest results These are the [latest results from run 2023-10-22T08:43:38.346498](https://huggingface.co/datasets/open-llm-leaderboard/details_TheBloke__koala-13B-HF/blob/main/results_2023-10-22T08-43-38.346498.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": 0.021707214765100673, "em_stderr": 0.0014923686874006184, "f1": 0.09106753355704705, "f1_stderr": 0.0020580604985252385, "acc": 0.40428250097386687, "acc_stderr": 0.009634029824810052 }, "harness|drop|3": { "em": 0.021707214765100673, "em_stderr": 0.0014923686874006184, "f1": 0.09106753355704705, "f1_stderr": 0.0020580604985252385 }, "harness|gsm8k|5": { "acc": 0.06823351023502654, "acc_stderr": 0.006945358944067431 }, "harness|winogrande|5": { "acc": 0.7403314917127072, "acc_stderr": 0.012322700705552673 } } ``` ### 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]

原始信息汇总

数据集概述

该数据集是在对模型 TheBloke/koala-13B-HF 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_TheBloke__koala-13B-HF", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-22T08:43:38.346498 运行的最新结果:

python { "all": { "em": 0.021707214765100673, "em_stderr": 0.0014923686874006184, "f1": 0.09106753355704705, "f1_stderr": 0.0020580604985252385, "acc": 0.40428250097386687, "acc_stderr": 0.009634029824810052 }, "harness|drop|3": { "em": 0.021707214765100673, "em_stderr": 0.0014923686874006184, "f1": 0.09106753355704705, "f1_stderr": 0.0020580604985252385 }, "harness|gsm8k|5": { "acc": 0.06823351023502654, "acc_stderr": 0.006945358944067431 }, "harness|winogrande|5": { "acc": 0.7403314917127072, "acc_stderr": 0.012322700705552673 } }

配置详情

  • harness_arc_challenge_25

    • 分割:2023_07_19T18_49_04.838102
    • 路径:**/details_harness|arc:challenge|25_2023-07-19T18:49:04.838102.parquet
    • 分割:latest
    • 路径:**/details_harness|arc:challenge|25_2023-07-19T18:49:04.838102.parquet
  • harness_drop_3

    • 分割:2023_10_22T08_43_38.346498
    • 路径:**/details_harness|drop|3_2023-10-22T08-43-38.346498.parquet
    • 分割:latest
    • 路径:**/details_harness|drop|3_2023-10-22T08-43-38.346498.parquet
  • harness_gsm8k_5

    • 分割:2023_10_22T08_43_38.346498
    • 路径:**/details_harness|gsm8k|5_2023-10-22T08-43-38.346498.parquet
    • 分割:latest
    • 路径:**/details_harness|gsm8k|5_2023-10-22T08-43-38.346498.parquet
  • harness_hellaswag_10

    • 分割:2023_07_19T18_49_04.838102
    • 路径:**/details_harness|hellaswag|10_2023-07-19T18:49:04.838102.parquet
    • 分割:latest
    • 路径:**/details_harness|hellaswag|10_2023-07-19T18:49:04.838102.parquet
  • harness_hendrycksTest_5

    • 分割:2023_07_19T18_49_04.838102
    • 路径:
      • **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-anatomy|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-astronomy|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-business_ethics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-college_biology|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-college_medicine|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-college_physics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-computer_security|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-econometrics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-formal_logic|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-global_facts|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-human_aging|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-international_law|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-machine_learning|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-management|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-marketing|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-nutrition|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-philosophy|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-prehistory|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T18:49:04.838102.parquet
      • **/details_harness|hendrycksTest-professional_law|5_2023-07-19T18:49:04.838102.parquet
      • `**/details_harness|hendrycksTest-professional_medicine
搜集汇总
数据集介绍
open-llm-leaderboard-old/details_TheBloke__koala-13B-HF 数据集图片
构建方式
在大规模语言模型评估领域,对模型性能的精准量化是推动技术进步的关键。该数据集源自Open LLM Leaderboard对TheBloke/koala-13B-HF模型的评估流程,通过自动化方式构建而成。数据集包含64个配置,每个配置对应一项特定的评估任务,如ARC挑战、DROP、GSM8K等。评估分两次运行完成,每次运行的结果以时间戳命名作为独立的数据分割,而'train'分割则始终指向最新一次的评估结果。此外,一个名为'results'的独立配置汇总了所有运行的聚合指标,用于在排行榜上计算和展示综合性能。数据以Parquet格式存储,确保了高效读写与兼容性。
特点
该数据集的核心特色在于其结构化的多任务评估体系,能够全面反映模型在推理、知识问答、数学求解等多维度能力上的表现。每个任务配置都包含详细的指标,如准确率、F1分数及其标准误差,为模型对比提供了统计严谨性。数据集的版本控制通过时间戳分割实现,支持对历史评估结果的追溯与对比,便于追踪模型性能的演进。同时,'latest'分割的设计简化了最新结果的获取,而'train'分割的别名机制则进一步提升了使用的直观性。这种设计兼顾了深度分析与快速迭代的需求。
使用方法
研究人员可通过Hugging Face的datasets库便捷地加载该数据集。例如,使用`load_dataset("open-llm-leaderboard/details_TheBloke__koala-13B-HF", "harness_winogrande_5", split="train")`即可获取WinoGrande任务的最新评估细节。对于需要访问特定历史运行结果的情况,可指定相应的时间戳分割名称。此外,通过加载'results'配置,用户能够直接获取所有任务的聚合性能指标,用于模型间的横向比较。这种灵活的访问方式,使得该数据集既适用于单任务深入分析,也支持多任务综合评估,极大地方便了语言模型研究中的实证工作。
背景与挑战
背景概述
随着大语言模型(LLM)在自然语言处理领域的广泛应用,如何公正、全面地评估其性能成为学界与工业界共同关注的焦点。Open LLM Leaderboard由Hugging Face团队于2023年发起,旨在通过标准化基准测试,系统性地比较不同开源模型的推理、常识、数学及知识掌握能力。该数据集作为Leaderboard的一部分,专门记录了TheBloke/koala-13B-HF模型在多项任务上的评估细节,涵盖ARC挑战赛、DROP、GSM8K、HellaSwag及涵盖57个学科的MMLU测试。通过结构化存储每次运行的测评结果,该数据集为研究者提供了透明、可复现的模型比较依据,对推动开源LLM的公平竞争与进步具有重要影响力。
当前挑战
该数据集所解决的领域问题在于,大语言模型性能评估长期缺乏统一、细粒度的标准化框架,不同模型间难以直接横向对比。构建过程中面临的核心挑战包括:其一,需设计涵盖多维度能力的任务集,如推理、数学、常识及专业知识,确保评估的全面性;其二,每次模型评估涉及64个独立配置,数据存储与管理复杂,需保证不同时间戳运行结果的完整性与可追溯性;其三,评估结果需以结构化格式(如Parquet文件)存储,并支持通过Python API灵活加载,这对数据一致性与版本控制提出了较高要求。
常用场景
经典使用场景
在大型语言模型(LLM)评估领域,该数据集被广泛用于标准化基准测试,通过记录模型在ARC挑战、HellaSwag、MMLU等多样化任务上的细粒度推理结果,为研究者提供可复现的对比分析基础。其64个独立配置对应不同任务与采样次数(如5-shot或25-shot),使得模型在常识推理、数学求解、阅读理解等维度的能力得以量化呈现,成为Open LLM排行榜上模型性能横向比较的核心数据来源。
实际应用
在实际应用中,该数据集为模型开发者提供了自动化的性能监控工具,通过记录每次评估的完整日志,支持工程师快速定位模型退化或改进的具体任务。企业可利用其细粒度结果(如winogrande的74%准确率)进行模型选型,在对话系统、智能客服等场景中权衡模型能力与部署成本,同时作为合规性验证的客观依据,确保模型在关键任务上的可靠性。
衍生相关工作
该数据集催生了多项衍生研究,包括基于其细粒度结果分析模型鲁棒性的工作,如探究不同采样策略(如3-shot与5-shot)对推理一致性的影响。此外,研究者利用其多任务配置开发了模型能力图谱,揭示了不同规模模型在知识密集型与推理密集型任务上的权衡关系,为后续的模型压缩与蒸馏策略提供了实证基础,并启发了针对特定任务(如DROP的文本理解)的专项优化方法。
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