open-llm-leaderboard-old/details_DreadPoor__WestuccineBagel-7B-slerp

该数据集是在模型DreadPoor/WestuccineBagel-7B-slerp的评估运行期间自动创建的，用于在Open LLM Leaderboard上进行评估。数据集由63个配置组成，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命名。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。

原始信息汇总

数据集概述

数据集基本信息

• 数据集名称: Evaluation run of DreadPoor/WestuccineBagel-7B-slerp
• 数据集来源: 自动创建于模型 DreadPoor/WestuccineBagel-7B-slerp 在 Open LLM Leaderboard 的评估运行中。
• 数据集组成: 包含 63 个配置，每个配置对应一个评估任务。
• 数据集创建: 从 1 次运行中创建，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。"train" 分割始终指向最新的结果。
• 额外配置: 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 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_DreadPoor__WestuccineBagel-7B-slerp", "harness_winogrande_5", split="train")

最新结果

• 最新结果来源: 2024-01-27T18:23:34.849134
• 结果示例: python { "all": { "acc": 0.6503543802796363, "acc_stderr": 0.03198810404680267, "acc_norm": 0.652493536181864, "acc_norm_stderr": 0.03263342300984539, "mc1": 0.5079559363525091, "mc1_stderr": 0.01750128507455182, "mc2": 0.6706176398228436, "mc2_stderr": 0.015164325815509071 }, "harness|arc:challenge|25": { "acc": 0.6638225255972696, "acc_stderr": 0.013804855026205763, "acc_norm": 0.6936860068259386, "acc_norm_stderr": 0.013470584417276516 }, "harness|hellaswag|10": { "acc": 0.6868153754232225, "acc_stderr": 0.004628409084218762, "acc_norm": 0.8652658832901813, "acc_norm_stderr": 0.0034074155133260466 }, ... }

配置信息

• 配置名称: harness_arc_challenge_25

  • 数据文件:
    • 分割: 2024_01_27T18_23_34.849134
      • 路径: **/details_harness|arc:challenge|25_2024-01-27T18-23-34.849134.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2024-01-27T18-23-34.849134.parquet

• 配置名称: harness_gsm8k_5

  • 数据文件:
    • 分割: 2024_01_27T18_23_34.849134
      • 路径: **/details_harness|gsm8k|5_2024-01-27T18-23-34.849134.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2024-01-27T18-23-34.849134.parquet

• 配置名称: harness_hellaswag_10

  • 数据文件:
    • 分割: 2024_01_27T18_23_34.849134
      • 路径: **/details_harness|hellaswag|10_2024-01-27T18-23-34.849134.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2024-01-27T18-23-34.849134.parquet

• 配置名称: harness_hendrycksTest_5

  • 数据文件:
    • 分割: 2024_01_27T18_23_34.849134
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2024-01-27T18-23-34.849134.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2024-01-27T18-23-34.849134.parquet