five

open-llm-leaderboard-old/details_CultriX__Wernicke-7B-v9

收藏
Hugging Face2024-02-01 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_CultriX__Wernicke-7B-v9
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上对模型CultriX/Wernicke-7B-v9进行评估时自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集包含1次运行的数据,每次运行在配置中表示为特定的分割,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,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。README还提供了如何使用Python中的datasets库加载运行细节的示例。

该数据集是在Open LLM Leaderboard上对模型CultriX/Wernicke-7B-v9进行评估时自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集包含1次运行的数据,每次运行在配置中表示为特定的分割,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,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。README还提供了如何使用Python中的datasets库加载运行细节的示例。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

该数据集是在对模型 CultriX/Wernicke-7B-v9 进行评估运行期间自动创建的,用于 Open LLM Leaderboard

数据集组成

  • 数据集包含 63 个配置,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。
  • 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 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_CultriX__Wernicke-7B-v9", "harness_winogrande_5", split="train")

最新结果

以下是 2024-02-01T18:44:19.770806 运行的最新结果

python { "all": { "acc": 0.6546993526205855, "acc_stderr": 0.032063919741404885, "acc_norm": 0.6543080706718742, "acc_norm_stderr": 0.0327327760979397, "mc1": 0.5605875152998776, "mc1_stderr": 0.0173745204825137, "mc2": 0.7185564600293038, "mc2_stderr": 0.014650212605887704 }, "harness|arc:challenge|25": { "acc": 0.7022184300341296, "acc_stderr": 0.013363080107244482, "acc_norm": 0.7244027303754266, "acc_norm_stderr": 0.01305716965576184 }, "harness|hellaswag|10": { "acc": 0.7065325632344155, "acc_stderr": 0.004544201359074618, "acc_norm": 0.885381398127863, "acc_norm_stderr": 0.003179100565887989 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.34, "acc_stderr": 0.04760952285695235, "acc_norm": 0.34, "acc_norm_stderr": 0.04760952285695235 }, # 其他任务的结果... }

配置详情

  • config_name: harness_arc_challenge_25

    • split: 2024_02_01T18_44_19.770806
      • path: **/details_harness|arc:challenge|25_2024-02-01T18-44-19.770806.parquet
    • split: latest
      • path: **/details_harness|arc:challenge|25_2024-02-01T18-44-19.770806.parquet
  • config_name: harness_gsm8k_5

    • split: 2024_02_01T18_44_19.770806
      • path: **/details_harness|gsm8k|5_2024-02-01T18-44-19.770806.parquet
    • split: latest
      • path: **/details_harness|gsm8k|5_2024-02-01T18-44-19.770806.parquet
  • config_name: harness_hellaswag_10

    • split: 2024_02_01T18_44_19.770806
      • path: **/details_harness|hellaswag|10_2024-02-01T18-44-19.770806.parquet
    • split: latest
      • path: **/details_harness|hellaswag|10_2024-02-01T18-44-19.770806.parquet
  • config_name: harness_hendrycksTest_5

    • split: 2024_02_01T18_44_19.770806
      • path:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-international_law|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-management|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-marketing|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2024-02-01T18-44-19.770806.parquet
        • **/details_harness|hendrycksTest-professional_law|5_2024-02-01T18-44-19.770806.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