five

open-llm-leaderboard-old/details_liminerity__Liph42

收藏
Hugging Face2024-03-11 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_liminerity__Liph42
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liminerity/Liph42进行评估时自动创建的。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存储在配置中,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上计算和显示聚合指标。

该数据集是在Open LLM Leaderboard上对模型liminerity/Liph42进行评估时自动创建的。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存储在配置中,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上计算和显示聚合指标。
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 liminerity/Liph42 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_liminerity__Liph42", "harness_winogrande_5", split="train")

最新结果

以下是 2024-03-11T07:17:43.989739 运行的最新结果

python { "all": { "acc": 0.5768088946737996, "acc_stderr": 0.03386419410943647, "acc_norm": 0.5775913691878023, "acc_norm_stderr": 0.03455962600778737, "mc1": 0.3182374541003672, "mc1_stderr": 0.01630598864892062, "mc2": 0.45941901106624383, "mc2_stderr": 0.015297016394293442 }, "harness|arc:challenge|25": { "acc": 0.5955631399317406, "acc_stderr": 0.01434203648343618, "acc_norm": 0.6203071672354948, "acc_norm_stderr": 0.01418211986697487 }, "harness|hellaswag|10": { "acc": 0.575682135032862, "acc_stderr": 0.004932289405608946, "acc_norm": 0.7587134037044413, "acc_norm_stderr": 0.004269893011588919 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.3, "acc_stderr": 0.04605661864718381, "acc_norm": 0.3, "acc_norm_stderr": 0.04605661864718381 }, ... }

配置详情

  • config_name: harness_arc_challenge_25

    • split: 2024_03_11T07_17_43.989739
      • path: **/details_harness|arc:challenge|25_2024-03-11T07-17-43.989739.parquet
    • split: latest
      • path: **/details_harness|arc:challenge|25_2024-03-11T07-17-43.989739.parquet
  • config_name: harness_gsm8k_5

    • split: 2024_03_11T07_17_43.989739
      • path: **/details_harness|gsm8k|5_2024-03-11T07-17-43.989739.parquet
    • split: latest
      • path: **/details_harness|gsm8k|5_2024-03-11T07-17-43.989739.parquet
  • config_name: harness_hellaswag_10

    • split: 2024_03_11T07_17_43.989739
      • path: **/details_harness|hellaswag|10_2024-03-11T07-17-43.989739.parquet
    • split: latest
      • path: **/details_harness|hellaswag|10_2024-03-11T07-17-43.989739.parquet
  • config_name: harness_hendrycksTest_5

    • split: 2024_03_11T07_17_43.989739
      • path:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2024-03-11T07-17-43.989739.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2024-03-11T07-17-43.989739.parquet ... }
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