five

open-llm-leaderboard-old/details_beomi__KoAlpaca-Polyglot-5.8B

收藏
Hugging Face2023-09-22 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_beomi__KoAlpaca-Polyglot-5.8B
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beomi/KoAlpaca-Polyglot-5.8B在Open LLM Leaderboard上的评估运行期间自动创建的。它由64个配置组成,每个配置对应一个评估任务。数据集是从3次运行中生成的,每次运行在每个配置中表示为特定的分割,使用运行的时间戳命名。train分割始终指向最新的结果。一个名为results的额外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。可以使用`datasets`库中的`load_dataset`函数加载该数据集。

该数据集是在模型beomi/KoAlpaca-Polyglot-5.8B在Open LLM Leaderboard上的评估运行期间自动创建的。它由64个配置组成,每个配置对应一个评估任务。数据集是从3次运行中生成的,每次运行在每个配置中表示为特定的分割,使用运行的时间戳命名。train分割始终指向最新的结果。一个名为results的额外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。可以使用`datasets`库中的`load_dataset`函数加载该数据集。
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 beomi/KoAlpaca-Polyglot-5.8BOpen LLM Leaderboard 上的自动创建的。数据集包含 64 个配置,每个配置对应一个评估任务。

数据集结构

数据集由 3 次运行结果组成,每次运行结果作为一个特定的分割存在于每个配置中,分割名称使用运行的时间戳。"train" 分割始终指向最新的结果。

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_beomi__KoAlpaca-Polyglot-5.8B", "harness_winogrande_5", split="train")

最新结果

以下是 2023-09-22T22:10:39.400321 运行 的最新结果: python { "all": { "em": 0.01541526845637584, "em_stderr": 0.0012616582904353766, "f1": 0.054131711409395974, "f1_stderr": 0.0017182561984205931, "acc": 0.24544616266538535, "acc_stderr": 0.007403949973545061 }, "harness|drop|3": { "em": 0.01541526845637584, "em_stderr": 0.0012616582904353766, "f1": 0.054131711409395974, "f1_stderr": 0.0017182561984205931 }, "harness|gsm8k|5": { "acc": 0.000758150113722517, "acc_stderr": 0.0007581501137225404 }, "harness|winogrande|5": { "acc": 0.49013417521704816, "acc_stderr": 0.014049749833367582 } }

配置详情

  • harness_arc_challenge_25

    • 分割: 2023_07_18T12_52_43.613378
      • 路径: **/details_harness|arc:challenge|25_2023-07-18T12:52:43.613378.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2023-07-18T12:52:43.613378.parquet
  • harness_drop_3

    • 分割: 2023_09_17T10_40_00.706474
      • 路径: **/details_harness|drop|3_2023-09-17T10-40-00.706474.parquet
    • 分割: 2023_09_22T22_10_39.400321
      • 路径: **/details_harness|drop|3_2023-09-22T22-10-39.400321.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|3_2023-09-22T22-10-39.400321.parquet
  • harness_gsm8k_5

    • 分割: 2023_09_17T10_40_00.706474
      • 路径: **/details_harness|gsm8k|5_2023-09-17T10-40-00.706474.parquet
    • 分割: 2023_09_22T22_10_39.400321
      • 路径: **/details_harness|gsm8k|5_2023-09-22T22-10-39.400321.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2023-09-22T22-10-39.400321.parquet
  • harness_hellaswag_10

    • 分割: 2023_07_18T12_52_43.613378
      • 路径: **/details_harness|hellaswag|10_2023-07-18T12:52:43.613378.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2023-07-18T12:52:43.613378.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_07_18T12_52_43.613378
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-management|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-07-18T12:52:43.613378.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-07-18T12:
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