five

open-llm-leaderboard-old/details_LLM360__Amber

收藏
Hugging Face2023-12-19 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_LLM360__Amber
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LLM360/Amber模型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命名。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Open LLM Leaderboard上对LLM360/Amber模型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命名。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

该数据集是在模型 LLM360/AmberOpen LLM Leaderboard 上的评估运行期间自动创建的。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

  • 配置数量:63
  • 创建来源:从 1 次运行中创建,每次运行对应一个特定的时间戳。
  • 数据分割:每个配置包含一个特定的时间戳分割和一个指向最新结果的 "train" 分割。
  • 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_LLM360__Amber", "harness_winogrande_5", split="train")

最新结果

以下是 2023-12-19T04:59:05.791643 运行 的最新结果:

python { "all": { "acc": 0.2778470494306401, "acc_stderr": 0.03144370019620237, "acc_norm": 0.27870842542577673, "acc_norm_stderr": 0.032201431055323866, "mc1": 0.2141982864137087, "mc1_stderr": 0.014362148155690462, "mc2": 0.3355637385526089, "mc2_stderr": 0.013068282225164367 }, "harness|arc:challenge|25": { "acc": 0.39761092150170646, "acc_stderr": 0.014301752223279536, "acc_norm": 0.40955631399317405, "acc_norm_stderr": 0.014370358632472437 }, "harness|hellaswag|10": { "acc": 0.5478988249352719, "acc_stderr": 0.004966832553245046, "acc_norm": 0.7379008165704043, "acc_norm_stderr": 0.00438877529821019 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.33, "acc_stderr": 0.04725815626252606, "acc_norm": 0.33, "acc_norm_stderr": 0.04725815626252606 }, # 其他任务的结果... }

配置详情

  • harness_arc_challenge_25

    • 分割:2023_12_19T04_59_05.791643, latest
    • 路径:**/details_harness|arc:challenge|25_2023-12-19T04-59-05.791643.parquet
  • harness_gsm8k_5

    • 分割:2023_12_19T04_59_05.791643, latest
    • 路径:**/details_harness|gsm8k|5_2023-12-19T04-59-05.791643.parquet
  • harness_hellaswag_10

    • 分割:2023_12_19T04_59_05.791643, latest
    • 路径:**/details_harness|hellaswag|10_2023-12-19T04-59-05.791643.parquet
  • harness_hendrycksTest_5

    • 分割:2023_12_19T04_59_05.791643, latest
    • 路径:
      • **/details_harness|hendrycksTest-abstract_algebra|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-anatomy|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-astronomy|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-business_ethics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-college_biology|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-college_chemistry|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-college_computer_science|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-college_mathematics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-college_medicine|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-college_physics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-computer_security|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-conceptual_physics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-econometrics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-electrical_engineering|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-formal_logic|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-global_facts|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_biology|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_european_history|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_geography|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_physics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_psychology|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_statistics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_us_history|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-high_school_world_history|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-human_aging|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-human_sexuality|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-international_law|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-jurisprudence|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-logical_fallacies|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-machine_learning|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-management|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-marketing|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-medical_genetics|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-miscellaneous|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-moral_disputes|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-moral_scenarios|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-nutrition|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-philosophy|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-prehistory|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-professional_accounting|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-professional_law|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-professional_medicine|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-professional_psychology|5_2023-12-19T04-59-05.791643.parquet
      • **/details_harness|hendrycksTest-public_relations|5_2023-12-19T04-59-05.791643.parquet
      • `**/details_harness|hendrycksTest-security
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