five

open-llm-leaderboard-old/details_Delcos__NATE-7b

收藏
Hugging Face2023-10-12 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_Delcos__NATE-7b
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Delcos/NATE-7b模型进行评估时自动生成的。数据集包含61个配置,每个配置对应一个评估任务。数据集由一次运行的结果组成,每次运行在每个配置中作为一个特定的分割表示,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results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。README还提供了如何使用Hugging Face datasets库加载数据集的示例。

该数据集是在Open LLM Leaderboard上对Delcos/NATE-7b模型进行评估时自动生成的。数据集包含61个配置,每个配置对应一个评估任务。数据集由一次运行的结果组成,每次运行在每个配置中作为一个特定的分割表示,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results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。README还提供了如何使用Hugging Face datasets库加载数据集的示例。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

数据集信息

  • 名称: Evaluation run of Delcos/NATE-7b
  • 来源: 自动创建于模型 Delcos/NATE-7bOpen LLM Leaderboard 的评估运行中。
  • 组成: 包含61个配置,每个配置对应一个评估任务。
  • 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: "results" 配置存储所有运行的聚合结果,用于计算和显示在 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_Delcos__NATE-7b", "harness_truthfulqa_mc_0", split="train")

最新结果

  • 最新结果时间戳: 2023-10-12T03:21:56.889828
  • 结果示例: python { "all": { "acc": 0.5894733182245001, "acc_stderr": 0.03414332313258469, "acc_norm": 0.5933830960354635, "acc_norm_stderr": 0.03412320397463523, "mc1": 0.3953488372093023, "mc1_stderr": 0.017115815632418197, "mc2": 0.571756969256499, "mc2_stderr": 0.01564827771634302 }, "harness|arc:challenge|25": { "acc": 0.5784982935153583, "acc_stderr": 0.014430197069326025, "acc_norm": 0.6092150170648464, "acc_norm_stderr": 0.014258563880513778 }, "harness|hellaswag|10": { "acc": 0.620991834295957, "acc_stderr": 0.004841486716855774, "acc_norm": 0.8209520015933081, "acc_norm_stderr": 0.0038260895866500536 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.33, "acc_stderr": 0.04725815626252605, "acc_norm": 0.33, "acc_norm_stderr": 0.04725815626252605 },

    其他任务的结果...

}

配置信息

  • 配置名称: harness_arc_challenge_25

    • 数据文件:
      • 分割: 2023_10_12T03_21_56.889828
        • 路径: **/details_harness|arc:challenge|25_2023-10-12T03-21-56.889828.parquet
      • 分割: latest
        • 路径: **/details_harness|arc:challenge|25_2023-10-12T03-21-56.889828.parquet
  • 配置名称: harness_hellaswag_10

    • 数据文件:
      • 分割: 2023_10_12T03_21_56.889828
        • 路径: **/details_harness|hellaswag|10_2023-10-12T03-21-56.889828.parquet
      • 分割: latest
        • 路径: **/details_harness|hellaswag|10_2023-10-12T03-21-56.889828.parquet
  • 配置名称: harness_hendrycksTest_5

    • 数据文件:
      • 分割: 2023_10_12T03_21_56.889828
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-international_law|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-management|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-marketing|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-prehistory|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-professional_accounting|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-professional_law|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-professional_medicine|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-professional_psychology|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-public_relations|5_2023-10-12T03-21-56.889828.parquet
          • **/details_harness|hendrycksTest-security_studies|5_2023-10-12T03-21-56.889828.par
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