five

open-llm-leaderboard/details_hakurei__lotus-12B

收藏
Hugging Face2023-08-27 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_hakurei__lotus-12B
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在模型hakurei/lotus-12B的评估运行过程中自动创建的,主要用于Open LLM Leaderboard的评估。数据集包含61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,用于计算和显示Open LLM Leaderboard上的聚合指标。
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

数据集信息

数据集结构

  • 配置数量: 61
  • 每个配置对应: 一个评估任务
  • 数据来源: 1 次运行
  • 分割命名: 使用运行的时间戳
  • 最新结果: "train" 分割指向最新结果

额外配置

  • "results" 配置: 存储所有运行的聚合结果,用于计算和显示聚合指标

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_hakurei__lotus-12B", "harness_truthfulqa_mc_0", split="train")

最新结果

  • 最新结果日期: 2023-07-18T13:41:37.836572
  • 结果示例: python { "all": { "acc": 0.2485367177317206, "acc_stderr": 0.03124226591981919, "acc_norm": 0.25130617872419225, "acc_norm_stderr": 0.03125347081115218, "mc1": 0.22643818849449204, "mc1_stderr": 0.014651337324602574, "mc2": 0.40115476804436745, "mc2_stderr": 0.014756133562988513 }, "harness|arc:challenge|25": { "acc": 0.26535836177474403, "acc_stderr": 0.012902554762313962, "acc_norm": 0.30716723549488056, "acc_norm_stderr": 0.013481034054980945 }, "harness|hellaswag|10": { "acc": 0.4054969129655447, "acc_stderr": 0.0048998450871831105, "acc_norm": 0.5270862378012349, "acc_norm_stderr": 0.004982454383162063 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.25, "acc_stderr": 0.04351941398892446, "acc_norm": 0.25, "acc_norm_stderr": 0.04351941398892446 },

    其他任务的结果...

}

配置详情

  • 配置名称: harness_arc_challenge_25

    • 数据文件:
      • 分割: 2023_07_18T13_41_37.836572
        • 路径: **/details_harness|arc:challenge|25_2023-07-18T13:41:37.836572.parquet
      • 分割: latest
        • 路径: **/details_harness|arc:challenge|25_2023-07-18T13:41:37.836572.parquet
  • 配置名称: harness_hellaswag_10

    • 数据文件:
      • 分割: 2023_07_18T13_41_37.836572
        • 路径: **/details_harness|hellaswag|10_2023-07-18T13:41:37.836572.parquet
      • 分割: latest
        • 路径: **/details_harness|hellaswag|10_2023-07-18T13:41:37.836572.parquet
  • 配置名称: harness_hendrycksTest_5

    • 数据文件:
      • 分割: 2023_07_18T13_41_37.836572
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-international_law|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-management|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-marketing|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-prehistory|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-professional_accounting|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-professional_law|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-professional_medicine|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-professional_psychology|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-public_relations|5_2023-07-18T13:41:37.836572.parquet
          • **/details_harness|hendrycksTest-security_studies|5_2023-07-18T13:41:37.836572.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