five

open-llm-leaderboard-old/details_macadeliccc__SmaugDolphin-60B

收藏
Hugging Face2024-02-17 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_macadeliccc__SmaugDolphin-60B
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在评估模型macadeliccc/SmaugDolphin-60B时自动创建的,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。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。

该数据集是在评估模型macadeliccc/SmaugDolphin-60B时自动创建的,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。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。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

数据集简介

该数据集是在评估模型 macadeliccc/SmaugDolphin-60BOpen LLM Leaderboard 上的自动创建的。数据集包含 63 个配置,每个配置对应一个评估任务。

数据集结构

数据集由 1 次运行创建,每个运行可以在每个配置中作为一个特定的分片找到,分片名称使用运行的时间戳。"train" 分片始终指向最新的结果。

额外配置

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_macadeliccc__SmaugDolphin-60B", "harness_winogrande_5", split="train")

最新结果

以下是 最新结果 的摘要:

python { "all": { "acc": 0.7651915724176592, "acc_stderr": 0.02813245689172726, "acc_norm": 0.7689266432773564, "acc_norm_stderr": 0.02866672384399442, "mc1": 0.5006119951040392, "mc1_stderr": 0.01750348793889251, "mc2": 0.6743783536659786, "mc2_stderr": 0.014425170450824752 }, "harness|arc:challenge|25": { "acc": 0.697098976109215, "acc_stderr": 0.013428241573185349, "acc_norm": 0.7337883959044369, "acc_norm_stderr": 0.012915774781523217 }, "harness|hellaswag|10": { "acc": 0.6674965146385182, "acc_stderr": 0.004701474865207032, "acc_norm": 0.8654650468034256, "acc_norm_stderr": 0.003405288007233201 }, # 其他任务的结果... }

配置详情

  • config_name: harness_arc_challenge_25

    • data_files:
      • split: 2024_02_17T06_01_33.258417
        • **/details_harness|arc:challenge|25_2024-02-17T06-01-33.258417.parquet
      • split: latest
        • **/details_harness|arc:challenge|25_2024-02-17T06-01-33.258417.parquet
  • config_name: harness_gsm8k_5

    • data_files:
      • split: 2024_02_17T06_01_33.258417
        • **/details_harness|gsm8k|5_2024-02-17T06-01-33.258417.parquet
      • split: latest
        • **/details_harness|gsm8k|5_2024-02-17T06-01-33.258417.parquet
  • config_name: harness_hellaswag_10

    • data_files:
      • split: 2024_02_17T06_01_33.258417
        • **/details_harness|hellaswag|10_2024-02-17T06-01-33.258417.parquet
      • split: latest
        • **/details_harness|hellaswag|10_2024-02-17T06-01-33.258417.parquet
  • config_name: harness_hendrycksTest_5

    • data_files:
      • split: 2024_02_17T06_01_33.258417
        • **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2024-02-17T06-01-33.258417.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2024-02-17T06-01-33.258417.parquet

        其他文件路径...

}

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