five

open-llm-leaderboard/details_Envoid__Libra-19B

收藏
Hugging Face2023-10-28 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_Envoid__Libra-19B
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Envoid/Libra-19B进行评估时自动创建的。它由64个配置组成,每个配置对应一个评估任务。数据集由2次运行生成,每次运行在每个配置中表示为特定的分割。train分割始终指向最新的结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。可以使用提供的Python代码片段加载数据集。
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

数据集来源

该数据集是在对模型 Envoid/Libra-19B 进行评估运行期间自动创建的,评估结果发布在 Open LLM Leaderboard 上。

数据集结构

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。

额外配置

一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示在 Open LLM Leaderboard 上的聚合指标。

数据加载示例

以下是加载特定运行详细信息的示例代码: python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_Envoid__Libra-19B", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-28T19:46:48.305345 运行的最新结果: python { "all": { "em": 0.4041526845637584, "em_stderr": 0.005025506959675468, "f1": 0.4663181627516796, "f1_stderr": 0.0048334958525102995, "acc": 0.38198917766143897, "acc_stderr": 0.006352871239464968 }, "harness|drop|3": { "em": 0.4041526845637584, "em_stderr": 0.005025506959675468, "f1": 0.4663181627516796, "f1_stderr": 0.0048334958525102995 }, "harness|gsm8k|5": { "acc": 0.000758150113722517, "acc_stderr": 0.0007581501137225396 }, "harness|winogrande|5": { "acc": 0.7632202052091555, "acc_stderr": 0.011947592365207397 } }

配置详情

以下是数据集的配置详情:

  • harness_arc_challenge_25

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|arc:challenge|25_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2023-10-10T13-06-44.906506.parquet
  • harness_drop_3

    • 分割: 2023_10_28T19_46_48.305345
      • 路径: **/details_harness|drop|3_2023-10-28T19-46-48.305345.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|3_2023-10-28T19-46-48.305345.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_28T19_46_48.305345
      • 路径: **/details_harness|gsm8k|5_2023-10-28T19-46-48.305345.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2023-10-28T19-46-48.305345.parquet
  • harness_hellaswag_10

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hellaswag|10_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-10T13-06-44.906506.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-10-10T13-06-44.906506.parquet
        • ...(其他路径省略)
    • 分割: latest
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-10T13-06-44.906506.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-10-10T13-06-44.906506.parquet
        • ...(其他路径省略)
  • harness_hendrycksTest_abstract_algebra_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_anatomy_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-anatomy|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-anatomy|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_astronomy_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-astronomy|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-astronomy|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_college_biology_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-college_biology|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-college_biology|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_college_physics_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-college_physics|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-college_physics|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_computer_security_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-computer_security|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-computer_security|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-10T13-06-44.906506.parquet
  • harness_hendrycksTest_econometrics_5

    • 分割: 2023_10_10T13_06_44.906506
      • 路径: **/details_harness|hendrycksTest-econometrics|5_2023-10-10T13-06-44.906506.parquet
    • 分割: latest
      • 路径: `**/details_harness|hendrycksTest-econometrics|5_2023-10-10T13-06
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