遇见数据集

open-llm-leaderboard-old/details_davzoku__cria-llama2-7b-v1.3

收藏
Hugging Face2023-10-15 更新2024-06-22 收录
官方服务:

资源简介:

--- pretty_name: Evaluation run of davzoku/cria-llama2-7b-v1.3 dataset_summary: "Dataset automatically created during the evaluation run of model\ \ [davzoku/cria-llama2-7b-v1.3](https://huggingface.co/davzoku/cria-llama2-7b-v1.3)\ \ 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_davzoku__cria-llama2-7b-v1.3\"\ ,\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\n\ These are the [latest results from run 2023-10-15T10:22:59.314023](https://huggingface.co/datasets/open-llm-leaderboard/details_davzoku__cria-llama2-7b-v1.3/blob/main/results_2023-10-15T10-22-59.314023.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\": 0.057466442953020135,\n\ \ \"em_stderr\": 0.0023833905882384905,\n \"f1\": 0.11769505033557005,\n\ \ \"f1_stderr\": 0.002702267782529149,\n \"acc\": 0.40196706145922656,\n\ \ \"acc_stderr\": 0.01015537696016331\n },\n \"harness|drop|3\": {\n\ \ \"em\": 0.057466442953020135,\n \"em_stderr\": 0.0023833905882384905,\n\ \ \"f1\": 0.11769505033557005,\n \"f1_stderr\": 0.002702267782529149\n\ \ },\n \"harness|gsm8k|5\": {\n \"acc\": 0.08491281273692192,\n \ \ \"acc_stderr\": 0.007678212824450797\n },\n \"harness|winogrande|5\"\ : {\n \"acc\": 0.7190213101815311,\n \"acc_stderr\": 0.012632541095875824\n\ \ }\n}\n```" repo_url: https://huggingface.co/davzoku/cria-llama2-7b-v1.3 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_08_16T09_58_26.810126 path: - '**/details_harness|arc:challenge|25_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|arc:challenge|25_2023-08-16T09:58:26.810126.parquet' - config_name: harness_drop_3 data_files: - split: 2023_10_15T10_22_59.314023 path: - '**/details_harness|drop|3_2023-10-15T10-22-59.314023.parquet' - split: latest path: - '**/details_harness|drop|3_2023-10-15T10-22-59.314023.parquet' - config_name: harness_gsm8k_5 data_files: - split: 2023_10_15T10_22_59.314023 path: - '**/details_harness|gsm8k|5_2023-10-15T10-22-59.314023.parquet' - split: latest path: - '**/details_harness|gsm8k|5_2023-10-15T10-22-59.314023.parquet' - config_name: harness_hellaswag_10 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hellaswag|10_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hellaswag|10_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-management|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-management|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-08-16T09:58:26.810126.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_abstract_algebra_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_anatomy_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-anatomy|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-anatomy|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_astronomy_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-astronomy|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-astronomy|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_business_ethics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_clinical_knowledge_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_college_biology_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-college_biology|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_biology|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_college_chemistry_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_college_computer_science_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_college_mathematics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_college_medicine_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_college_physics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-college_physics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_physics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_computer_security_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-computer_security|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-computer_security|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_conceptual_physics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_econometrics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-econometrics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-econometrics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_electrical_engineering_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_elementary_mathematics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_formal_logic_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_global_facts_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-global_facts|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-global_facts|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_biology_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_chemistry_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_computer_science_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_european_history_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_geography_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_government_and_politics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_macroeconomics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_mathematics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_microeconomics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_physics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_psychology_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_statistics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_us_history_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_high_school_world_history_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_human_aging_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-human_aging|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_aging|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_human_sexuality_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_international_law_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-international_law|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-international_law|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_jurisprudence_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_logical_fallacies_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_machine_learning_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_management_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-management|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-management|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_marketing_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-marketing|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-marketing|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_medical_genetics_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_miscellaneous_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_moral_disputes_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_moral_scenarios_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_nutrition_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-nutrition|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-nutrition|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_philosophy_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-philosophy|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-philosophy|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_prehistory_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-prehistory|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-prehistory|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_professional_accounting_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_professional_law_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-professional_law|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_law|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_professional_medicine_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_professional_psychology_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_public_relations_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-public_relations|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-public_relations|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_security_studies_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-security_studies|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-security_studies|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_sociology_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-sociology|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-sociology|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_us_foreign_policy_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_virology_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-virology|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-virology|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_hendrycksTest_world_religions_5 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|hendrycksTest-world_religions|5_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|hendrycksTest-world_religions|5_2023-08-16T09:58:26.810126.parquet' - config_name: harness_truthfulqa_mc_0 data_files: - split: 2023_08_16T09_58_26.810126 path: - '**/details_harness|truthfulqa:mc|0_2023-08-16T09:58:26.810126.parquet' - split: latest path: - '**/details_harness|truthfulqa:mc|0_2023-08-16T09:58:26.810126.parquet' - config_name: harness_winogrande_5 data_files: - split: 2023_10_15T10_22_59.314023 path: - '**/details_harness|winogrande|5_2023-10-15T10-22-59.314023.parquet' - split: latest path: - '**/details_harness|winogrande|5_2023-10-15T10-22-59.314023.parquet' - config_name: results data_files: - split: 2023_08_16T09_58_26.810126 path: - results_2023-08-16T09:58:26.810126.parquet - split: 2023_10_15T10_22_59.314023 path: - results_2023-10-15T10-22-59.314023.parquet - split: latest path: - results_2023-10-15T10-22-59.314023.parquet --- # Dataset Card for Evaluation run of davzoku/cria-llama2-7b-v1.3 ## Dataset Description - **Homepage:** - **Repository:** https://huggingface.co/davzoku/cria-llama2-7b-v1.3 - **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 [davzoku/cria-llama2-7b-v1.3](https://huggingface.co/davzoku/cria-llama2-7b-v1.3) 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_davzoku__cria-llama2-7b-v1.3", "harness_winogrande_5", split="train") ``` ## Latest results These are the [latest results from run 2023-10-15T10:22:59.314023](https://huggingface.co/datasets/open-llm-leaderboard/details_davzoku__cria-llama2-7b-v1.3/blob/main/results_2023-10-15T10-22-59.314023.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": 0.057466442953020135, "em_stderr": 0.0023833905882384905, "f1": 0.11769505033557005, "f1_stderr": 0.002702267782529149, "acc": 0.40196706145922656, "acc_stderr": 0.01015537696016331 }, "harness|drop|3": { "em": 0.057466442953020135, "em_stderr": 0.0023833905882384905, "f1": 0.11769505033557005, "f1_stderr": 0.002702267782529149 }, "harness|gsm8k|5": { "acc": 0.08491281273692192, "acc_stderr": 0.007678212824450797 }, "harness|winogrande|5": { "acc": 0.7190213101815311, "acc_stderr": 0.012632541095875824 } } ``` ### 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

该数据集是在评估模型 davzoku/cria-llama2-7b-v1.3 在 Open 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" 分割总是指向最新的结果。
  • 一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示聚合指标在 Open LLM Leaderboard 上。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_davzoku__cria-llama2-7b-v1.3", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-15T10:22:59.314023 运行的最新结果: python { "all": { "em": 0.057466442953020135, "em_stderr": 0.0023833905882384905, "f1": 0.11769505033557005, "f1_stderr": 0.002702267782529149, "acc": 0.40196706145922656, "acc_stderr": 0.01015537696016331 }, "harness|drop|3": { "em": 0.057466442953020135, "em_stderr": 0.0023833905882384905, "f1": 0.11769505033557005, "f1_stderr": 0.002702267782529149 }, "harness|gsm8k|5": { "acc": 0.08491281273692192, "acc_stderr": 0.007678212824450797 }, "harness|winogrande|5": { "acc": 0.7190213101815311, "acc_stderr": 0.012632541095875824 } }

配置详情

  • harness_arc_challenge_25

    • 分割: 2023_08_16T09_58_26.810126
    • 路径: **/details_harness|arc:challenge|25_2023-08-16T09:58:26.810126.parquet
    • 分割: latest
    • 路径: **/details_harness|arc:challenge|25_2023-08-16T09:58:26.810126.parquet
  • harness_drop_3

    • 分割: 2023_10_15T10_22_59.314023
    • 路径: **/details_harness|drop|3_2023-10-15T10-22-59.314023.parquet
    • 分割: latest
    • 路径: **/details_harness|drop|3_2023-10-15T10-22-59.314023.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_15T10_22_59.314023
    • 路径: **/details_harness|gsm8k|5_2023-10-15T10-22-59.314023.parquet
    • 分割: latest
    • 路径: **/details_harness|gsm8k|5_2023-10-15T10-22-59.314023.parquet
  • harness_hellaswag_10

    • 分割: 2023_08_16T09_58_26.810126
    • 路径: **/details_harness|hellaswag|10_2023-08-16T09:58:26.810126.parquet
    • 分割: latest
    • 路径: **/details_harness|hellaswag|10_2023-08-16T09:58:26.810126.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_08_16T09_58_26.810126
    • 路径:
      • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-anatomy|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-astronomy|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-business_ethics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-college_biology|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-college_medicine|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-college_physics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-computer_security|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-econometrics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-formal_logic|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-global_facts|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-human_aging|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-international_law|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-machine_learning|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-management|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-marketing|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-nutrition|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-philosophy|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-prehistory|5_2023-08-16T09:58:26.810126.parquet
      • **/details_harness|hendrycksTest-professional_accounting|5_2023-08-16T09:58:26.810126.parquet
      • `**/details_harness|hendrycksTest-professional_law|5_2023-08-16T09:58:26.810126
搜集汇总
数据集介绍
open-llm-leaderboard-old/details_davzoku__cria-llama2-7b-v1.3 数据集图片
构建方式
该数据集是在Open LLM Leaderboard平台上对模型davzoku/cria-llama2-7b-v1.3进行自动化评估过程中生成的。数据集由64个配置组成,每个配置对应一个评估任务,涵盖ARC挑战、DROP、GSM8K、WinoGrande、HellaSwag以及涵盖57个学科的HendrycksTest等多个基准。评估结果按运行时间戳组织成不同的数据分割,其中'train'分割始终指向最新的评估结果。此外,一个名为'results'的独立配置存储了所有运行的聚合指标,用于在排行榜上计算和展示总体性能。
特点
该数据集以细粒度的任务级配置为特色,每个配置独立存储特定任务的评估细节,便于研究者针对性地分析模型在不同维度的表现。数据采用Parquet格式保存,高效支持大规模结构化数据的存取。评估结果包含精确匹配率、F1分数和准确率等关键指标及其标准误差,为模型能力提供了量化的统计依据。通过时间戳分割,数据集还记录了多次运行的历史结果,支持对模型性能演变的追踪与对比分析。
使用方法
使用该数据集时,可通过Hugging Face的datasets库加载。用户需指定任务配置名称(如'harness_winogrande_5')和分割名称(如'train'或具体时间戳)来获取相应数据。例如,调用load_dataset函数并传入数据集标识符、配置名和分割参数即可加载指定任务的评估详情。数据集的设计使得研究者能够灵活地选取单一任务进行深入剖析,或结合多个配置进行跨任务比较,从而全面评估模型在自然语言理解、推理和知识应用等方面的综合能力。
背景与挑战
背景概述
在大型语言模型(LLM)蓬勃发展的时代,如何系统、公正地评估模型性能已成为学术界与工业界共同关注的核心议题。Open LLM Leaderboard应运而生,由Hugging Face团队于2023年创建,旨在为开源LLM提供标准化、多维度的评测平台。该数据集记录了模型davzoku/cria-llama2-7b-v1.3在Leaderboard上的详细评估结果,涵盖ARC、HellaSwag、MMLU、GSM8K、Winogrande、DROP等多项基准任务,全面反映了模型在常识推理、数学计算、阅读理解与知识掌握等方面的能力。通过自动采集多次运行结果并聚合为结构化数据,该数据集不仅为研究者提供了透明、可复现的模型性能参考,也推动了LLM评估体系的规范化与社区协作的深入发展。
当前挑战
该数据集所应对的核心挑战在于LLM评估的标准化与可复现性难题。不同任务(如GSM8K的数学推理与Winogrande的常识推理)对模型能力的要求迥异,单一指标难以全面刻画模型优劣,而多任务评估框架的构建又面临任务设计、评分一致性及计算资源消耗的平衡问题。在数据构建过程中,挑战尤为突出:每次评估运行需在统一环境下执行64个配置任务,并处理来自不同时间戳的多次运行结果,确保数据聚合的准确性与最新性;此外,原始评估日志的解析与parquet格式的转换、跨任务结果的结构化存储,均对数据管道的鲁棒性与可扩展性提出了严苛要求。
常用场景
经典使用场景
在大型语言模型的评估生态中,Open LLM Leaderboard 的细节数据集为研究者提供了一种标准化、可复现的模型性能剖析工具。该数据集记录了模型 davzoku/cria-llama2-7b-v1.3 在多项基准任务上的逐项评估结果,涵盖 ARC-Challenge、HellaSwag、MMLU、GSM8K、Winogrande 以及 DROP 等经典任务。研究者可通过加载特定配置与分片,深入分析模型在常识推理、数学求解、语言理解与知识掌握等维度的细微表现,从而精准定位模型的优势与短板。这种细粒度的评估数据不仅支持纵向对比同一模型在不同时间点的迭代效果,也为横向比较不同架构或训练策略的模型提供了统一的度量标尺,成为模型优化与选型的重要依据。
实际应用
在实际应用层面,该评估细节数据集为模型部署前的性能验证提供了关键参考。企业在选择或微调语言模型用于特定业务场景时,可依据该数据集中的细粒度指标进行决策。例如,在需要高精度常识推理的智能客服系统中,Winogrande 的高准确率表明该模型具备较好的语言理解基础;而在需要多步推理的自动化编程辅助或教育辅导场景中,GSM8K 的低得分则警示需要额外强化或替换模型。此外,数据集的时间戳分片设计允许追踪模型在持续训练过程中的性能变化,支持模型迭代的版本管理与质量监控,从而降低部署风险并提升应用系统的鲁棒性。
衍生相关工作
该数据集作为 Open LLM Leaderboard 评估体系的一部分,衍生了一系列关于模型评估标准化与基准测试的重要工作。其结构化存储与灵活加载的方式启发了后续如 EleutherAI 的 LM Evaluation Harness 生态扩展,推动了评估结果的跨平台共享与比较。基于该数据集的细粒度结果,研究者开发了多种模型性能诊断工具,能够自动识别模型在特定子任务上的失败模式。同时,该数据集也被用于训练元评估模型,以预测模型在未见任务上的表现,从而指导更高效的模型选择与集成策略。这些衍生工作共同构建了一个围绕模型评估的数据驱动研究范式,促进了大型语言模型领域从粗放式性能展示向精细化能力分析的转变。
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