five

open-llm-leaderboard-old/details_cookinai__BruinHermes

收藏
Hugging Face2023-12-18 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_cookinai__BruinHermes
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在评估模型cookinai/BruinHermes时自动创建的,主要用于在Open LLM Leaderboard上展示模型的评估结果。数据集包含63个配置,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为一个特定的分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,还有一个名为results的配置,存储了所有运行的聚合结果,用于在Open LLM Leaderboard上计算和展示聚合指标。

该数据集是在评估模型cookinai/BruinHermes时自动创建的,主要用于在Open LLM Leaderboard上展示模型的评估结果。数据集包含63个配置,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为一个特定的分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,还有一个名为results的配置,存储了所有运行的聚合结果,用于在Open LLM Leaderboard上计算和展示聚合指标。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

该数据集是在对模型 cookinai/BruinHermes 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 63 个配置,每个配置对应一个评估任务。
  • 数据集从 1 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 Open LLM Leaderboard 上的聚合指标。

最新结果

以下是 2023-12-18T08:28:15.533319 运行的最新结果:

python { "all": { "acc": 0.6567218786558917, "acc_stderr": 0.03193493899655974, "acc_norm": 0.65676353979433, "acc_norm_stderr": 0.032592352059711976, "mc1": 0.5018359853121175, "mc1_stderr": 0.017503383046877048, "mc2": 0.6559538306507776, "mc2_stderr": 0.015048223056540525 }, "harness|arc:challenge|25": { "acc": 0.6689419795221843, "acc_stderr": 0.013752062419817836, "acc_norm": 0.7013651877133106, "acc_norm_stderr": 0.01337407861506874 }, "harness|hellaswag|10": { "acc": 0.6893049193387771, "acc_stderr": 0.004618323959513038, "acc_norm": 0.8707428799044015, "acc_norm_stderr": 0.0033479866695653154 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.32, "acc_stderr": 0.04688261722621504, "acc_norm": 0.32, "acc_norm_stderr": 0.04688261722621504 }, # 其他任务的结果... }

配置详情

  • config_name: harness_arc_challenge_25

    • split: 2023_12_18T08_28_15.533319
      • path: **/details_harness|arc:challenge|25_2023-12-18T08-28-15.533319.parquet
    • split: latest
      • path: **/details_harness|arc:challenge|25_2023-12-18T08-28-15.533319.parquet
  • config_name: harness_gsm8k_5

    • split: 2023_12_18T08_28_15.533319
      • path: **/details_harness|gsm8k|5_2023-12-18T08-28-15.533319.parquet
    • split: latest
      • path: **/details_harness|gsm8k|5_2023-12-18T08-28-15.533319.parquet
  • config_name: harness_hellaswag_10

    • split: 2023_12_18T08_28_15.533319
      • path: **/details_harness|hellaswag|10_2023-12-18T08-28-15.533319.parquet
    • split: latest
      • path: **/details_harness|hellaswag|10_2023-12-18T08-28-15.533319.parquet
  • config_name: harness_hendrycksTest_5

    • split: 2023_12_18T08_28_15.533319
      • path:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-12-18T08-28-15.533319.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-12-18T08-28-15.533319.parquet

        其他路径...

}

二维码
社区交流群
二维码
科研交流群
商业服务