open-llm-leaderboard/details_Locutusque__Hyperion-3.0-Mixtral-3x7B

该数据集是在模型Locutusque/Hyperion-3.0-Mixtral-3x7B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。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

数据集名称

• pretty_name: Evaluation run of Locutusque/Hyperion-3.0-Mixtral-3x7B

数据集创建

• dataset_summary: 该数据集是在评估模型Locutusque/Hyperion-3.0-Mixtral-3x7B运行期间自动创建的，用于Open LLM Leaderboard。

数据集结构

• 组成: 由63个配置组成，每个配置对应一个评估任务。
• 创建来源: 数据集从1次运行中创建，每次运行作为一个特定的分割，分割名使用运行的时间戳命名。
• 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。

数据集加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_Locutusque__Hyperion-3.0-Mixtral-3x7B", "harness_winogrande_5", split="train")

最新结果

• 结果来源: latest results from run 2024-03-21T14:35:21.042412
• 结果内容: 包含多个任务的评估结果，如准确率(acc)、标准误差(acc_stderr)等。

数据集配置详情

配置列表

1. harness_arc_challenge_25

   • 数据文件:
     • split: 2024_03_21T14_35_21.042412
     • path: /details_harness|arc:challenge|25_2024-03-21T14-35-21.042412.parquet
   • 数据文件:
     • split: latest
     • path: /details_harness|arc:challenge|25_2024-03-21T14-35-21.042412.parquet

2. harness_gsm8k_5

   • 数据文件:
     • split: 2024_03_21T14_35_21.042412
     • path: /details_harness|gsm8k|5_2024-03-21T14-35-21.042412.parquet
   • 数据文件:
     • split: latest
     • path: /details_harness|gsm8k|5_2024-03-21T14-35-21.042412.parquet

3. harness_hellaswag_10

   • 数据文件:
     • split: 2024_03_21T14_35_21.042412
     • path: /details_harness|hellaswag|10_2024-03-21T14-35-21.042412.parquet
   • 数据文件:
     • split: latest
     • path: /details_harness|hellaswag|10_2024-03-21T14-35-21.042412.parquet

4. harness_hendrycksTest_5

   • 数据文件:
     • split: 2024_03_21T14_35_21.042412
     • path:
       • /details_harness|hendrycksTest-abstract_algebra|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-anatomy|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-astronomy|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-business_ethics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-clinical_knowledge|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_biology|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_chemistry|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_computer_science|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_mathematics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_medicine|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_physics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-computer_security|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-conceptual_physics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-econometrics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-electrical_engineering|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-elementary_mathematics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-formal_logic|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-global_facts|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-high_school_biology|5_2024-03-21T14-35-21.042412.parquet
       • ...
       • /details_harness|hendrycksTest-world_religions|5_2024-03-21T14-35-21.042412.parquet
   • 数据文件:
     • split: latest
     • path:
       • /details_harness|hendrycksTest-abstract_algebra|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-anatomy|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-astronomy|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-business_ethics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-clinical_knowledge|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_biology|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_chemistry|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_computer_science|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_mathematics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_medicine|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-college_physics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-computer_security|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-conceptual_physics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-econometrics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-electrical_engineering|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-elementary_mathematics|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-formal_logic|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-global_facts|5_2024-03-21T14-35-21.042412.parquet
       • /details_harness|hendrycksTest-high_school_biology|5_2024-03-21T14-35-21.042412.parquet
       • ...
       • /details_harness|hendrycksTest-world_religions|5_2024-03-21T14-35-21.042412.parquet