five

open-llm-leaderboard/details_acrastt__Vicuna-3B

收藏
Hugging Face2023-08-27 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_acrastt__Vicuna-3B
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在Open LLM Leaderboard上对模型acrastt/Vicuna-3B进行评估时自动创建的。数据集由61个配置组成,每个配置对应一个评估任务。数据集是从1次运行中创建的,每次运行都可以在特定配置中找到,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,results配置存储了所有运行的聚合结果,并用于计算和显示Open LLM Leaderboard上的聚合指标。
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

数据集基本信息

数据集组成

  • 配置数量: 61个配置
  • 每个配置对应: 一个评估任务
  • 创建次数: 1次运行
  • 数据分割: 每个配置包含特定分割,分割名称使用运行的时间戳,"train" 分割指向最新结果

额外配置

  • 结果配置: 存储所有运行结果的聚合,用于计算和显示聚合指标

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_acrastt__Vicuna-3B", "harness_truthfulqa_mc_0", split="train")

最新结果

  • 最新结果日期: 2023-08-17T13:42:02.549031
  • 结果示例: python { "all": { "acc": 0.2811009875581447, "acc_stderr": 0.03252062846238445, "acc_norm": 0.2849236147706273, "acc_norm_stderr": 0.03251779934164602, "mc1": 0.24479804161566707, "mc1_stderr": 0.01505186948671501, "mc2": 0.38343994316155305, "mc2_stderr": 0.013903929837677163 }, "harness|arc:challenge|25": { "acc": 0.3651877133105802, "acc_stderr": 0.014070265519268804, "acc_norm": 0.4129692832764505, "acc_norm_stderr": 0.014388344935398326 }, "harness|hellaswag|10": { "acc": 0.5407289384584744, "acc_stderr": 0.004973199296339957, "acc_norm": 0.7184823740290779, "acc_norm_stderr": 0.004488201756642574 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.24, "acc_stderr": 0.04292346959909281, "acc_norm": 0.24, "acc_norm_stderr": 0.04292346959909281 },

    更多任务结果...

}

配置详情

  • 配置名称: harness_arc_challenge_25

    • 数据文件:
      • 分割: 2023_08_17T13_42_02.549031
        • 路径: **/details_harness|arc:challenge|25_2023-08-17T13:42:02.549031.parquet
      • 分割: latest
        • 路径: **/details_harness|arc:challenge|25_2023-08-17T13:42:02.549031.parquet
  • 配置名称: harness_hellaswag_10

    • 数据文件:
      • 分割: 2023_08_17T13_42_02.549031
        • 路径: **/details_harness|hellaswag|10_2023-08-17T13:42:02.549031.parquet
      • 分割: latest
        • 路径: **/details_harness|hellaswag|10_2023-08-17T13:42:02.549031.parquet
  • 配置名称: harness_hendrycksTest_5

    • 数据文件:
      • 分割: 2023_08_17T13_42_02.549031
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-international_law|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-management|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-marketing|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-prehistory|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-professional_accounting|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-professional_law|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-professional_medicine|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-professional_psychology|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-public_relations|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness|hendrycksTest-security_studies|5_2023-08-17T13:42:02.549031.parquet
          • **/details_harness
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作