open-llm-leaderboard-old/details_Azazelle__Argetsu

该数据集是在模型Azazelle/Argetsu的评估运行期间自动创建的，包含63个配置，每个配置对应一个评估任务。数据集由一次运行创建，每个运行在每个配置中作为一个特定的分割存在，分割名称使用运行的时间戳。此外，还有一个名为"results"的配置，存储所有运行的聚合结果，用于计算和显示聚合指标。数据集用于支持Open LLM Leaderboard的评估和展示。

原始信息汇总

数据集概述

该数据集是在对模型 Azazelle/Argetsu 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。
• 额外的 "results" 配置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_Azazelle__Argetsu", "harness_winogrande_5", split="train")

最新结果

以下是 2024-01-06T00:40:38.552688 运行的最新结果：

python { "all": { "acc": 0.6575651362894879, "acc_stderr": 0.03187960419181806, "acc_norm": 0.6592507669408202, "acc_norm_stderr": 0.03251642943055543, "mc1": 0.386780905752754, "mc1_stderr": 0.017048857010515107, "mc2": 0.5645591684488664, "mc2_stderr": 0.01544402118615705 }, "harness|arc:challenge|25": { "acc": 0.6407849829351536, "acc_stderr": 0.014020224155839157, "acc_norm": 0.6706484641638225, "acc_norm_stderr": 0.013734057652635474 }, "harness|hellaswag|10": { "acc": 0.6840270862378013, "acc_stderr": 0.004639520453444027, "acc_norm": 0.8631746664011153, "acc_norm_stderr": 0.0034296051062163687 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.37, "acc_stderr": 0.048523658709391, "acc_norm": 0.37, "acc_norm_stderr": 0.048523658709391 }, # 其他任务的结果... }

配置详情

• config_name: harness_arc_challenge_25

  • 数据文件:
    • 分割: 2024_01_06T00_40_38.552688
      • 路径: **/details_harness|arc:challenge|25_2024-01-06T00-40-38.552688.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2024-01-06T00-40-38.552688.parquet

• config_name: harness_gsm8k_5

  • 数据文件:
    • 分割: 2024_01_06T00_40_38.552688
      • 路径: **/details_harness|gsm8k|5_2024-01-06T00-40-38.552688.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2024-01-06T00-40-38.552688.parquet

• config_name: harness_hellaswag_10

  • 数据文件:
    • 分割: 2024_01_06T00_40_38.552688
      • 路径: **/details_harness|hellaswag|10_2024-01-06T00-40-38.552688.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2024-01-06T00-40-38.552688.parquet

• config_name: harness_hendrycksTest_5

  • 数据文件:
    • 分割: 2024_01_06T00_40_38.552688
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2024-01-06T00-40-38.552688.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2024-01-06T00-40-38.552688.parquet
        • `**/details_harness|hendrycksTest-high_school_chemistry|5_2024-01